diff --git "a/trainer_state.json" "b/trainer_state.json" new file mode 100644--- /dev/null +++ "b/trainer_state.json" @@ -0,0 +1,403662 @@ +{ + "best_metric": null, + "best_model_checkpoint": null, + "epoch": 4.0, + "eval_steps": 4000, + "global_step": 57644, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 6.939143709666228e-05, + "grad_norm": 21.405803680419922, + "learning_rate": 1.1560693641618498e-07, + "loss": 3.5331, + "step": 1 + }, + { + "epoch": 0.00013878287419332455, + "grad_norm": 21.701461791992188, + "learning_rate": 2.3121387283236997e-07, + "loss": 3.8302, + "step": 2 + }, + { + "epoch": 0.0002081743112899868, + "grad_norm": 24.85260772705078, + "learning_rate": 3.4682080924855494e-07, + "loss": 4.5432, + "step": 3 + }, + { + "epoch": 0.0002775657483866491, + "grad_norm": 31.991670608520508, + "learning_rate": 4.6242774566473993e-07, + "loss": 4.1037, + "step": 4 + }, + { + "epoch": 0.00034695718548331133, + "grad_norm": 23.006437301635742, + "learning_rate": 5.780346820809248e-07, + "loss": 4.0821, + "step": 5 + }, + { + "epoch": 0.0004163486225799736, + "grad_norm": 21.958019256591797, + "learning_rate": 6.936416184971099e-07, + "loss": 3.9642, + "step": 6 + }, + { + "epoch": 0.0004857400596766359, + "grad_norm": 30.58041763305664, + "learning_rate": 8.092485549132948e-07, + "loss": 3.9866, + "step": 7 + }, + { + "epoch": 0.0005551314967732982, + "grad_norm": 23.328380584716797, + "learning_rate": 9.248554913294799e-07, + "loss": 4.3913, + "step": 8 + }, + { + "epoch": 0.0006245229338699605, + "grad_norm": 39.51573181152344, + "learning_rate": 1.0404624277456647e-06, + "loss": 3.8524, + "step": 9 + }, + { + "epoch": 0.0006939143709666227, + "grad_norm": 22.557479858398438, + "learning_rate": 1.1560693641618497e-06, + "loss": 4.2079, + "step": 10 + }, + { + "epoch": 0.0007633058080632849, + "grad_norm": 21.46837615966797, + "learning_rate": 1.2716763005780348e-06, + "loss": 4.1383, + "step": 11 + }, + { + "epoch": 0.0008326972451599472, + "grad_norm": 31.970840454101562, + "learning_rate": 1.3872832369942197e-06, + "loss": 4.1979, + "step": 12 + }, + { + "epoch": 0.0009020886822566095, + "grad_norm": 18.563188552856445, + "learning_rate": 1.5028901734104047e-06, + "loss": 3.6442, + "step": 13 + }, + { + "epoch": 0.0009714801193532718, + "grad_norm": 23.51290512084961, + "learning_rate": 1.6184971098265896e-06, + "loss": 3.6751, + "step": 14 + }, + { + "epoch": 0.0010408715564499342, + "grad_norm": 36.687904357910156, + "learning_rate": 1.7341040462427746e-06, + "loss": 3.6484, + "step": 15 + }, + { + "epoch": 0.0011102629935465964, + "grad_norm": 21.93012237548828, + "learning_rate": 1.8497109826589597e-06, + "loss": 3.5314, + "step": 16 + }, + { + "epoch": 0.0011796544306432587, + "grad_norm": 22.330028533935547, + "learning_rate": 1.9653179190751445e-06, + "loss": 3.6127, + "step": 17 + }, + { + "epoch": 0.001249045867739921, + "grad_norm": 25.791728973388672, + "learning_rate": 2.0809248554913294e-06, + "loss": 3.9995, + "step": 18 + }, + { + "epoch": 0.0013184373048365833, + "grad_norm": 32.43659973144531, + "learning_rate": 2.1965317919075144e-06, + "loss": 4.1515, + "step": 19 + }, + { + "epoch": 0.0013878287419332453, + "grad_norm": 17.843576431274414, + "learning_rate": 2.3121387283236993e-06, + "loss": 3.9444, + "step": 20 + }, + { + "epoch": 0.0014572201790299076, + "grad_norm": 23.111494064331055, + "learning_rate": 2.4277456647398842e-06, + "loss": 4.2122, + "step": 21 + }, + { + "epoch": 0.0015266116161265699, + "grad_norm": 18.68913459777832, + "learning_rate": 2.5433526011560696e-06, + "loss": 3.6723, + "step": 22 + }, + { + "epoch": 0.0015960030532232322, + "grad_norm": 35.17980194091797, + "learning_rate": 2.6589595375722546e-06, + "loss": 3.5381, + "step": 23 + }, + { + "epoch": 0.0016653944903198944, + "grad_norm": 20.540578842163086, + "learning_rate": 2.7745664739884395e-06, + "loss": 2.7859, + "step": 24 + }, + { + "epoch": 0.0017347859274165567, + "grad_norm": 17.03854751586914, + "learning_rate": 2.8901734104046244e-06, + "loss": 3.964, + "step": 25 + }, + { + "epoch": 0.001804177364513219, + "grad_norm": 21.13581085205078, + "learning_rate": 3.0057803468208094e-06, + "loss": 2.81, + "step": 26 + }, + { + "epoch": 0.0018735688016098813, + "grad_norm": 20.59122657775879, + "learning_rate": 3.1213872832369943e-06, + "loss": 4.0552, + "step": 27 + }, + { + "epoch": 0.0019429602387065436, + "grad_norm": 20.962318420410156, + "learning_rate": 3.2369942196531793e-06, + "loss": 3.2017, + "step": 28 + }, + { + "epoch": 0.002012351675803206, + "grad_norm": 19.67106056213379, + "learning_rate": 3.3526011560693642e-06, + "loss": 3.3833, + "step": 29 + }, + { + "epoch": 0.0020817431128998683, + "grad_norm": 16.151103973388672, + "learning_rate": 3.468208092485549e-06, + "loss": 2.9145, + "step": 30 + }, + { + "epoch": 0.0021511345499965304, + "grad_norm": 16.021282196044922, + "learning_rate": 3.5838150289017345e-06, + "loss": 2.9823, + "step": 31 + }, + { + "epoch": 0.002220525987093193, + "grad_norm": 23.42172622680664, + "learning_rate": 3.6994219653179195e-06, + "loss": 3.1267, + "step": 32 + }, + { + "epoch": 0.002289917424189855, + "grad_norm": 15.452238082885742, + "learning_rate": 3.815028901734104e-06, + "loss": 2.9552, + "step": 33 + }, + { + "epoch": 0.0023593088612865174, + "grad_norm": 13.868646621704102, + "learning_rate": 3.930635838150289e-06, + "loss": 2.7846, + "step": 34 + }, + { + "epoch": 0.0024287002983831795, + "grad_norm": 12.976654052734375, + "learning_rate": 4.046242774566474e-06, + "loss": 2.7113, + "step": 35 + }, + { + "epoch": 0.002498091735479842, + "grad_norm": 18.781455993652344, + "learning_rate": 4.161849710982659e-06, + "loss": 3.5297, + "step": 36 + }, + { + "epoch": 0.002567483172576504, + "grad_norm": 24.201580047607422, + "learning_rate": 4.277456647398844e-06, + "loss": 3.4332, + "step": 37 + }, + { + "epoch": 0.0026368746096731665, + "grad_norm": 18.725160598754883, + "learning_rate": 4.393063583815029e-06, + "loss": 2.6188, + "step": 38 + }, + { + "epoch": 0.0027062660467698286, + "grad_norm": 21.01583480834961, + "learning_rate": 4.508670520231214e-06, + "loss": 2.6996, + "step": 39 + }, + { + "epoch": 0.0027756574838664907, + "grad_norm": 16.526838302612305, + "learning_rate": 4.624277456647399e-06, + "loss": 2.8379, + "step": 40 + }, + { + "epoch": 0.002845048920963153, + "grad_norm": 31.108890533447266, + "learning_rate": 4.739884393063584e-06, + "loss": 3.2132, + "step": 41 + }, + { + "epoch": 0.0029144403580598152, + "grad_norm": 28.432260513305664, + "learning_rate": 4.8554913294797685e-06, + "loss": 3.4209, + "step": 42 + }, + { + "epoch": 0.0029838317951564777, + "grad_norm": 21.320701599121094, + "learning_rate": 4.971098265895955e-06, + "loss": 3.0635, + "step": 43 + }, + { + "epoch": 0.0030532232322531398, + "grad_norm": 12.298962593078613, + "learning_rate": 5.086705202312139e-06, + "loss": 2.399, + "step": 44 + }, + { + "epoch": 0.0031226146693498023, + "grad_norm": 11.910674095153809, + "learning_rate": 5.202312138728324e-06, + "loss": 2.5043, + "step": 45 + }, + { + "epoch": 0.0031920061064464643, + "grad_norm": 11.899988174438477, + "learning_rate": 5.317919075144509e-06, + "loss": 2.2943, + "step": 46 + }, + { + "epoch": 0.003261397543543127, + "grad_norm": 16.49997329711914, + "learning_rate": 5.433526011560694e-06, + "loss": 2.8523, + "step": 47 + }, + { + "epoch": 0.003330788980639789, + "grad_norm": 13.096921920776367, + "learning_rate": 5.549132947976879e-06, + "loss": 2.3065, + "step": 48 + }, + { + "epoch": 0.0034001804177364514, + "grad_norm": 17.366668701171875, + "learning_rate": 5.6647398843930635e-06, + "loss": 3.2924, + "step": 49 + }, + { + "epoch": 0.0034695718548331134, + "grad_norm": 14.120378494262695, + "learning_rate": 5.780346820809249e-06, + "loss": 2.4673, + "step": 50 + }, + { + "epoch": 0.003538963291929776, + "grad_norm": 14.282581329345703, + "learning_rate": 5.895953757225433e-06, + "loss": 2.6974, + "step": 51 + }, + { + "epoch": 0.003608354729026438, + "grad_norm": 14.23493480682373, + "learning_rate": 6.011560693641619e-06, + "loss": 2.3254, + "step": 52 + }, + { + "epoch": 0.0036777461661231005, + "grad_norm": 13.290118217468262, + "learning_rate": 6.127167630057803e-06, + "loss": 2.3469, + "step": 53 + }, + { + "epoch": 0.0037471376032197625, + "grad_norm": 11.081518173217773, + "learning_rate": 6.242774566473989e-06, + "loss": 2.1172, + "step": 54 + }, + { + "epoch": 0.003816529040316425, + "grad_norm": 14.52530288696289, + "learning_rate": 6.358381502890173e-06, + "loss": 2.6392, + "step": 55 + }, + { + "epoch": 0.003885920477413087, + "grad_norm": 9.958911895751953, + "learning_rate": 6.4739884393063585e-06, + "loss": 2.1577, + "step": 56 + }, + { + "epoch": 0.00395531191450975, + "grad_norm": 10.681069374084473, + "learning_rate": 6.589595375722543e-06, + "loss": 2.3108, + "step": 57 + }, + { + "epoch": 0.004024703351606412, + "grad_norm": 9.534003257751465, + "learning_rate": 6.7052023121387284e-06, + "loss": 2.1154, + "step": 58 + }, + { + "epoch": 0.004094094788703074, + "grad_norm": 12.165912628173828, + "learning_rate": 6.820809248554913e-06, + "loss": 2.3595, + "step": 59 + }, + { + "epoch": 0.004163486225799737, + "grad_norm": 17.796062469482422, + "learning_rate": 6.936416184971098e-06, + "loss": 2.0305, + "step": 60 + }, + { + "epoch": 0.004232877662896399, + "grad_norm": 9.182536125183105, + "learning_rate": 7.0520231213872845e-06, + "loss": 2.0943, + "step": 61 + }, + { + "epoch": 0.004302269099993061, + "grad_norm": 9.690338134765625, + "learning_rate": 7.167630057803469e-06, + "loss": 2.1382, + "step": 62 + }, + { + "epoch": 0.004371660537089723, + "grad_norm": 21.167312622070312, + "learning_rate": 7.2832369942196544e-06, + "loss": 3.091, + "step": 63 + }, + { + "epoch": 0.004441051974186386, + "grad_norm": 11.394143104553223, + "learning_rate": 7.398843930635839e-06, + "loss": 2.3607, + "step": 64 + }, + { + "epoch": 0.004510443411283048, + "grad_norm": 10.324246406555176, + "learning_rate": 7.514450867052024e-06, + "loss": 2.3499, + "step": 65 + }, + { + "epoch": 0.00457983484837971, + "grad_norm": 12.272679328918457, + "learning_rate": 7.630057803468209e-06, + "loss": 1.9029, + "step": 66 + }, + { + "epoch": 0.004649226285476372, + "grad_norm": 12.47921371459961, + "learning_rate": 7.745664739884393e-06, + "loss": 2.1124, + "step": 67 + }, + { + "epoch": 0.004718617722573035, + "grad_norm": 9.999581336975098, + "learning_rate": 7.861271676300578e-06, + "loss": 1.8648, + "step": 68 + }, + { + "epoch": 0.004788009159669697, + "grad_norm": 9.418438911437988, + "learning_rate": 7.976878612716764e-06, + "loss": 2.2876, + "step": 69 + }, + { + "epoch": 0.004857400596766359, + "grad_norm": 8.409302711486816, + "learning_rate": 8.092485549132949e-06, + "loss": 1.9028, + "step": 70 + }, + { + "epoch": 0.004926792033863021, + "grad_norm": 11.724508285522461, + "learning_rate": 8.208092485549133e-06, + "loss": 2.3033, + "step": 71 + }, + { + "epoch": 0.004996183470959684, + "grad_norm": 8.99478530883789, + "learning_rate": 8.323699421965318e-06, + "loss": 1.8088, + "step": 72 + }, + { + "epoch": 0.005065574908056346, + "grad_norm": 11.932640075683594, + "learning_rate": 8.439306358381504e-06, + "loss": 2.1263, + "step": 73 + }, + { + "epoch": 0.005134966345153008, + "grad_norm": 8.19472885131836, + "learning_rate": 8.554913294797688e-06, + "loss": 1.7291, + "step": 74 + }, + { + "epoch": 0.00520435778224967, + "grad_norm": 8.931638717651367, + "learning_rate": 8.670520231213873e-06, + "loss": 2.1478, + "step": 75 + }, + { + "epoch": 0.005273749219346333, + "grad_norm": 9.888033866882324, + "learning_rate": 8.786127167630057e-06, + "loss": 2.1114, + "step": 76 + }, + { + "epoch": 0.005343140656442995, + "grad_norm": 13.911140441894531, + "learning_rate": 8.901734104046244e-06, + "loss": 1.9168, + "step": 77 + }, + { + "epoch": 0.005412532093539657, + "grad_norm": 10.847952842712402, + "learning_rate": 9.017341040462428e-06, + "loss": 1.6876, + "step": 78 + }, + { + "epoch": 0.005481923530636319, + "grad_norm": 9.171956062316895, + "learning_rate": 9.132947976878613e-06, + "loss": 1.7244, + "step": 79 + }, + { + "epoch": 0.005551314967732981, + "grad_norm": 10.875395774841309, + "learning_rate": 9.248554913294797e-06, + "loss": 1.965, + "step": 80 + }, + { + "epoch": 0.005620706404829644, + "grad_norm": 14.410478591918945, + "learning_rate": 9.364161849710983e-06, + "loss": 2.1919, + "step": 81 + }, + { + "epoch": 0.005690097841926306, + "grad_norm": 9.450847625732422, + "learning_rate": 9.479768786127168e-06, + "loss": 1.8006, + "step": 82 + }, + { + "epoch": 0.005759489279022968, + "grad_norm": 13.713593482971191, + "learning_rate": 9.595375722543352e-06, + "loss": 2.2685, + "step": 83 + }, + { + "epoch": 0.0058288807161196304, + "grad_norm": 22.478927612304688, + "learning_rate": 9.710982658959537e-06, + "loss": 1.8721, + "step": 84 + }, + { + "epoch": 0.005898272153216293, + "grad_norm": 11.288623809814453, + "learning_rate": 9.826589595375723e-06, + "loss": 1.6012, + "step": 85 + }, + { + "epoch": 0.005967663590312955, + "grad_norm": 9.906073570251465, + "learning_rate": 9.94219653179191e-06, + "loss": 2.0033, + "step": 86 + }, + { + "epoch": 0.0060370550274096175, + "grad_norm": 10.015321731567383, + "learning_rate": 1.0057803468208094e-05, + "loss": 1.8181, + "step": 87 + }, + { + "epoch": 0.0061064464645062795, + "grad_norm": 11.706305503845215, + "learning_rate": 1.0173410404624278e-05, + "loss": 1.9071, + "step": 88 + }, + { + "epoch": 0.0061758379016029425, + "grad_norm": 9.542607307434082, + "learning_rate": 1.0289017341040463e-05, + "loss": 2.0284, + "step": 89 + }, + { + "epoch": 0.0062452293386996045, + "grad_norm": 10.177284240722656, + "learning_rate": 1.0404624277456647e-05, + "loss": 2.0203, + "step": 90 + }, + { + "epoch": 0.006314620775796267, + "grad_norm": 9.66942310333252, + "learning_rate": 1.0520231213872834e-05, + "loss": 1.8465, + "step": 91 + }, + { + "epoch": 0.006384012212892929, + "grad_norm": 10.712051391601562, + "learning_rate": 1.0635838150289018e-05, + "loss": 1.9304, + "step": 92 + }, + { + "epoch": 0.006453403649989592, + "grad_norm": 8.443596839904785, + "learning_rate": 1.0751445086705203e-05, + "loss": 1.7216, + "step": 93 + }, + { + "epoch": 0.006522795087086254, + "grad_norm": 12.554067611694336, + "learning_rate": 1.0867052023121387e-05, + "loss": 1.5931, + "step": 94 + }, + { + "epoch": 0.006592186524182916, + "grad_norm": 9.786715507507324, + "learning_rate": 1.0982658959537573e-05, + "loss": 1.8932, + "step": 95 + }, + { + "epoch": 0.006661577961279578, + "grad_norm": 13.634934425354004, + "learning_rate": 1.1098265895953758e-05, + "loss": 1.8779, + "step": 96 + }, + { + "epoch": 0.006730969398376241, + "grad_norm": 11.504860877990723, + "learning_rate": 1.1213872832369943e-05, + "loss": 2.0311, + "step": 97 + }, + { + "epoch": 0.006800360835472903, + "grad_norm": 18.475435256958008, + "learning_rate": 1.1329479768786127e-05, + "loss": 1.9719, + "step": 98 + }, + { + "epoch": 0.006869752272569565, + "grad_norm": 8.988012313842773, + "learning_rate": 1.1445086705202313e-05, + "loss": 1.4971, + "step": 99 + }, + { + "epoch": 0.006939143709666227, + "grad_norm": 8.822566032409668, + "learning_rate": 1.1560693641618498e-05, + "loss": 1.7043, + "step": 100 + }, + { + "epoch": 0.00700853514676289, + "grad_norm": 8.75823974609375, + "learning_rate": 1.1676300578034682e-05, + "loss": 1.7598, + "step": 101 + }, + { + "epoch": 0.007077926583859552, + "grad_norm": 10.1179838180542, + "learning_rate": 1.1791907514450867e-05, + "loss": 1.5154, + "step": 102 + }, + { + "epoch": 0.007147318020956214, + "grad_norm": 9.336431503295898, + "learning_rate": 1.1907514450867053e-05, + "loss": 2.0539, + "step": 103 + }, + { + "epoch": 0.007216709458052876, + "grad_norm": 12.446418762207031, + "learning_rate": 1.2023121387283238e-05, + "loss": 1.5678, + "step": 104 + }, + { + "epoch": 0.007286100895149539, + "grad_norm": 26.802261352539062, + "learning_rate": 1.2138728323699422e-05, + "loss": 2.2281, + "step": 105 + }, + { + "epoch": 0.007355492332246201, + "grad_norm": 10.438783645629883, + "learning_rate": 1.2254335260115607e-05, + "loss": 1.7254, + "step": 106 + }, + { + "epoch": 0.007424883769342863, + "grad_norm": 8.524996757507324, + "learning_rate": 1.2369942196531791e-05, + "loss": 1.8557, + "step": 107 + }, + { + "epoch": 0.007494275206439525, + "grad_norm": 9.33948802947998, + "learning_rate": 1.2485549132947977e-05, + "loss": 2.0662, + "step": 108 + }, + { + "epoch": 0.007563666643536188, + "grad_norm": 10.075897216796875, + "learning_rate": 1.2601156069364162e-05, + "loss": 1.7686, + "step": 109 + }, + { + "epoch": 0.00763305808063285, + "grad_norm": 8.926899909973145, + "learning_rate": 1.2716763005780346e-05, + "loss": 1.8185, + "step": 110 + }, + { + "epoch": 0.007702449517729512, + "grad_norm": 10.227436065673828, + "learning_rate": 1.2832369942196531e-05, + "loss": 2.1119, + "step": 111 + }, + { + "epoch": 0.007771840954826174, + "grad_norm": 10.701436996459961, + "learning_rate": 1.2947976878612717e-05, + "loss": 1.6204, + "step": 112 + }, + { + "epoch": 0.007841232391922836, + "grad_norm": 10.688033103942871, + "learning_rate": 1.3063583815028902e-05, + "loss": 1.5687, + "step": 113 + }, + { + "epoch": 0.0079106238290195, + "grad_norm": 8.889094352722168, + "learning_rate": 1.3179190751445086e-05, + "loss": 1.5599, + "step": 114 + }, + { + "epoch": 0.007980015266116162, + "grad_norm": 8.637164115905762, + "learning_rate": 1.329479768786127e-05, + "loss": 1.5802, + "step": 115 + }, + { + "epoch": 0.008049406703212823, + "grad_norm": 11.159412384033203, + "learning_rate": 1.3410404624277457e-05, + "loss": 1.8794, + "step": 116 + }, + { + "epoch": 0.008118798140309486, + "grad_norm": 7.845677852630615, + "learning_rate": 1.3526011560693641e-05, + "loss": 1.7936, + "step": 117 + }, + { + "epoch": 0.008188189577406147, + "grad_norm": 8.232104301452637, + "learning_rate": 1.3641618497109826e-05, + "loss": 1.5647, + "step": 118 + }, + { + "epoch": 0.00825758101450281, + "grad_norm": 8.499692916870117, + "learning_rate": 1.375722543352601e-05, + "loss": 1.6153, + "step": 119 + }, + { + "epoch": 0.008326972451599473, + "grad_norm": 11.761832237243652, + "learning_rate": 1.3872832369942197e-05, + "loss": 1.8789, + "step": 120 + }, + { + "epoch": 0.008396363888696134, + "grad_norm": 14.316187858581543, + "learning_rate": 1.3988439306358383e-05, + "loss": 2.0409, + "step": 121 + }, + { + "epoch": 0.008465755325792797, + "grad_norm": 9.68807315826416, + "learning_rate": 1.4104046242774569e-05, + "loss": 1.8207, + "step": 122 + }, + { + "epoch": 0.008535146762889459, + "grad_norm": 10.616473197937012, + "learning_rate": 1.4219653179190754e-05, + "loss": 1.8743, + "step": 123 + }, + { + "epoch": 0.008604538199986122, + "grad_norm": 8.670979499816895, + "learning_rate": 1.4335260115606938e-05, + "loss": 1.4093, + "step": 124 + }, + { + "epoch": 0.008673929637082784, + "grad_norm": 9.06837272644043, + "learning_rate": 1.4450867052023123e-05, + "loss": 1.4811, + "step": 125 + }, + { + "epoch": 0.008743321074179446, + "grad_norm": 8.885125160217285, + "learning_rate": 1.4566473988439309e-05, + "loss": 1.7678, + "step": 126 + }, + { + "epoch": 0.008812712511276109, + "grad_norm": 9.30599594116211, + "learning_rate": 1.4682080924855493e-05, + "loss": 1.6383, + "step": 127 + }, + { + "epoch": 0.008882103948372772, + "grad_norm": 10.367547035217285, + "learning_rate": 1.4797687861271678e-05, + "loss": 1.9181, + "step": 128 + }, + { + "epoch": 0.008951495385469433, + "grad_norm": 9.598464012145996, + "learning_rate": 1.4913294797687862e-05, + "loss": 1.9022, + "step": 129 + }, + { + "epoch": 0.009020886822566096, + "grad_norm": 11.928574562072754, + "learning_rate": 1.5028901734104049e-05, + "loss": 1.7005, + "step": 130 + }, + { + "epoch": 0.009090278259662757, + "grad_norm": 12.424153327941895, + "learning_rate": 1.5144508670520233e-05, + "loss": 1.951, + "step": 131 + }, + { + "epoch": 0.00915966969675942, + "grad_norm": 9.01201057434082, + "learning_rate": 1.5260115606936418e-05, + "loss": 1.8258, + "step": 132 + }, + { + "epoch": 0.009229061133856083, + "grad_norm": 8.762784004211426, + "learning_rate": 1.5375722543352604e-05, + "loss": 1.3982, + "step": 133 + }, + { + "epoch": 0.009298452570952744, + "grad_norm": 10.207000732421875, + "learning_rate": 1.5491329479768787e-05, + "loss": 1.7673, + "step": 134 + }, + { + "epoch": 0.009367844008049407, + "grad_norm": 9.866923332214355, + "learning_rate": 1.5606936416184973e-05, + "loss": 1.7076, + "step": 135 + }, + { + "epoch": 0.00943723544514607, + "grad_norm": 9.177887916564941, + "learning_rate": 1.5722543352601156e-05, + "loss": 1.6273, + "step": 136 + }, + { + "epoch": 0.009506626882242731, + "grad_norm": 10.254706382751465, + "learning_rate": 1.5838150289017342e-05, + "loss": 1.5612, + "step": 137 + }, + { + "epoch": 0.009576018319339394, + "grad_norm": 8.982560157775879, + "learning_rate": 1.5953757225433528e-05, + "loss": 1.7797, + "step": 138 + }, + { + "epoch": 0.009645409756436055, + "grad_norm": 9.847992897033691, + "learning_rate": 1.606936416184971e-05, + "loss": 1.8197, + "step": 139 + }, + { + "epoch": 0.009714801193532718, + "grad_norm": 11.586702346801758, + "learning_rate": 1.6184971098265897e-05, + "loss": 1.7728, + "step": 140 + }, + { + "epoch": 0.009784192630629381, + "grad_norm": 11.041914939880371, + "learning_rate": 1.6300578034682083e-05, + "loss": 1.7861, + "step": 141 + }, + { + "epoch": 0.009853584067726042, + "grad_norm": 11.699503898620605, + "learning_rate": 1.6416184971098266e-05, + "loss": 1.8414, + "step": 142 + }, + { + "epoch": 0.009922975504822705, + "grad_norm": 10.147649765014648, + "learning_rate": 1.6531791907514452e-05, + "loss": 1.8663, + "step": 143 + }, + { + "epoch": 0.009992366941919368, + "grad_norm": 46.290191650390625, + "learning_rate": 1.6647398843930635e-05, + "loss": 1.4955, + "step": 144 + }, + { + "epoch": 0.01006175837901603, + "grad_norm": 29.462751388549805, + "learning_rate": 1.676300578034682e-05, + "loss": 1.8194, + "step": 145 + }, + { + "epoch": 0.010131149816112692, + "grad_norm": 10.138164520263672, + "learning_rate": 1.6878612716763008e-05, + "loss": 1.8053, + "step": 146 + }, + { + "epoch": 0.010200541253209353, + "grad_norm": 7.995537281036377, + "learning_rate": 1.699421965317919e-05, + "loss": 1.5569, + "step": 147 + }, + { + "epoch": 0.010269932690306016, + "grad_norm": 10.719072341918945, + "learning_rate": 1.7109826589595377e-05, + "loss": 1.8439, + "step": 148 + }, + { + "epoch": 0.010339324127402679, + "grad_norm": 11.06162166595459, + "learning_rate": 1.722543352601156e-05, + "loss": 1.4631, + "step": 149 + }, + { + "epoch": 0.01040871556449934, + "grad_norm": 8.472554206848145, + "learning_rate": 1.7341040462427746e-05, + "loss": 1.6392, + "step": 150 + }, + { + "epoch": 0.010478107001596003, + "grad_norm": 8.977538108825684, + "learning_rate": 1.7456647398843932e-05, + "loss": 1.3619, + "step": 151 + }, + { + "epoch": 0.010547498438692666, + "grad_norm": 8.8873291015625, + "learning_rate": 1.7572254335260115e-05, + "loss": 1.7563, + "step": 152 + }, + { + "epoch": 0.010616889875789327, + "grad_norm": 9.12559986114502, + "learning_rate": 1.76878612716763e-05, + "loss": 1.3549, + "step": 153 + }, + { + "epoch": 0.01068628131288599, + "grad_norm": 11.218527793884277, + "learning_rate": 1.7803468208092487e-05, + "loss": 1.7129, + "step": 154 + }, + { + "epoch": 0.010755672749982651, + "grad_norm": 8.389120101928711, + "learning_rate": 1.791907514450867e-05, + "loss": 1.3956, + "step": 155 + }, + { + "epoch": 0.010825064187079314, + "grad_norm": 10.026215553283691, + "learning_rate": 1.8034682080924856e-05, + "loss": 1.711, + "step": 156 + }, + { + "epoch": 0.010894455624175977, + "grad_norm": 8.314817428588867, + "learning_rate": 1.815028901734104e-05, + "loss": 1.5063, + "step": 157 + }, + { + "epoch": 0.010963847061272639, + "grad_norm": 9.53980827331543, + "learning_rate": 1.8265895953757225e-05, + "loss": 1.4263, + "step": 158 + }, + { + "epoch": 0.011033238498369301, + "grad_norm": 12.609450340270996, + "learning_rate": 1.838150289017341e-05, + "loss": 1.6982, + "step": 159 + }, + { + "epoch": 0.011102629935465963, + "grad_norm": 9.495781898498535, + "learning_rate": 1.8497109826589594e-05, + "loss": 1.3135, + "step": 160 + }, + { + "epoch": 0.011172021372562626, + "grad_norm": 9.880600929260254, + "learning_rate": 1.861271676300578e-05, + "loss": 1.5445, + "step": 161 + }, + { + "epoch": 0.011241412809659289, + "grad_norm": 9.236615180969238, + "learning_rate": 1.8728323699421967e-05, + "loss": 1.6645, + "step": 162 + }, + { + "epoch": 0.01131080424675595, + "grad_norm": 8.853983879089355, + "learning_rate": 1.884393063583815e-05, + "loss": 1.5906, + "step": 163 + }, + { + "epoch": 0.011380195683852613, + "grad_norm": 10.062235832214355, + "learning_rate": 1.8959537572254336e-05, + "loss": 1.6866, + "step": 164 + }, + { + "epoch": 0.011449587120949276, + "grad_norm": 8.774020195007324, + "learning_rate": 1.907514450867052e-05, + "loss": 1.2946, + "step": 165 + }, + { + "epoch": 0.011518978558045937, + "grad_norm": 8.261381149291992, + "learning_rate": 1.9190751445086705e-05, + "loss": 1.3808, + "step": 166 + }, + { + "epoch": 0.0115883699951426, + "grad_norm": 8.384431838989258, + "learning_rate": 1.930635838150289e-05, + "loss": 1.331, + "step": 167 + }, + { + "epoch": 0.011657761432239261, + "grad_norm": 8.580657005310059, + "learning_rate": 1.9421965317919074e-05, + "loss": 1.179, + "step": 168 + }, + { + "epoch": 0.011727152869335924, + "grad_norm": 8.538634300231934, + "learning_rate": 1.9537572254335264e-05, + "loss": 1.3276, + "step": 169 + }, + { + "epoch": 0.011796544306432587, + "grad_norm": 8.824459075927734, + "learning_rate": 1.9653179190751446e-05, + "loss": 1.4692, + "step": 170 + }, + { + "epoch": 0.011865935743529248, + "grad_norm": 9.195901870727539, + "learning_rate": 1.9768786127167633e-05, + "loss": 1.5967, + "step": 171 + }, + { + "epoch": 0.01193532718062591, + "grad_norm": 9.597426414489746, + "learning_rate": 1.988439306358382e-05, + "loss": 1.781, + "step": 172 + }, + { + "epoch": 0.012004718617722574, + "grad_norm": 10.142624855041504, + "learning_rate": 2e-05, + "loss": 1.596, + "step": 173 + }, + { + "epoch": 0.012074110054819235, + "grad_norm": 11.721856117248535, + "learning_rate": 2.0115606936416188e-05, + "loss": 1.4976, + "step": 174 + }, + { + "epoch": 0.012143501491915898, + "grad_norm": 10.75423812866211, + "learning_rate": 2.023121387283237e-05, + "loss": 1.7573, + "step": 175 + }, + { + "epoch": 0.012212892929012559, + "grad_norm": 9.122659683227539, + "learning_rate": 2.0346820809248557e-05, + "loss": 1.5061, + "step": 176 + }, + { + "epoch": 0.012282284366109222, + "grad_norm": 8.880846977233887, + "learning_rate": 2.0462427745664743e-05, + "loss": 1.4314, + "step": 177 + }, + { + "epoch": 0.012351675803205885, + "grad_norm": 9.597685813903809, + "learning_rate": 2.0578034682080926e-05, + "loss": 1.6675, + "step": 178 + }, + { + "epoch": 0.012421067240302546, + "grad_norm": 9.041475296020508, + "learning_rate": 2.0693641618497112e-05, + "loss": 1.4333, + "step": 179 + }, + { + "epoch": 0.012490458677399209, + "grad_norm": 10.062479972839355, + "learning_rate": 2.0809248554913295e-05, + "loss": 1.6159, + "step": 180 + }, + { + "epoch": 0.012559850114495872, + "grad_norm": 14.622681617736816, + "learning_rate": 2.092485549132948e-05, + "loss": 1.5262, + "step": 181 + }, + { + "epoch": 0.012629241551592533, + "grad_norm": 11.729297637939453, + "learning_rate": 2.1040462427745667e-05, + "loss": 2.0052, + "step": 182 + }, + { + "epoch": 0.012698632988689196, + "grad_norm": 27.700241088867188, + "learning_rate": 2.115606936416185e-05, + "loss": 1.5718, + "step": 183 + }, + { + "epoch": 0.012768024425785857, + "grad_norm": 7.543630123138428, + "learning_rate": 2.1271676300578036e-05, + "loss": 1.2299, + "step": 184 + }, + { + "epoch": 0.01283741586288252, + "grad_norm": 9.457754135131836, + "learning_rate": 2.1387283236994223e-05, + "loss": 1.4731, + "step": 185 + }, + { + "epoch": 0.012906807299979183, + "grad_norm": 11.178112030029297, + "learning_rate": 2.1502890173410405e-05, + "loss": 1.7208, + "step": 186 + }, + { + "epoch": 0.012976198737075844, + "grad_norm": 9.227314949035645, + "learning_rate": 2.1618497109826592e-05, + "loss": 1.8119, + "step": 187 + }, + { + "epoch": 0.013045590174172507, + "grad_norm": 8.095595359802246, + "learning_rate": 2.1734104046242775e-05, + "loss": 1.1759, + "step": 188 + }, + { + "epoch": 0.01311498161126917, + "grad_norm": 7.867910385131836, + "learning_rate": 2.184971098265896e-05, + "loss": 1.2199, + "step": 189 + }, + { + "epoch": 0.013184373048365831, + "grad_norm": 11.053202629089355, + "learning_rate": 2.1965317919075147e-05, + "loss": 1.3166, + "step": 190 + }, + { + "epoch": 0.013253764485462494, + "grad_norm": 7.716185569763184, + "learning_rate": 2.208092485549133e-05, + "loss": 1.4072, + "step": 191 + }, + { + "epoch": 0.013323155922559156, + "grad_norm": 6.833305835723877, + "learning_rate": 2.2196531791907516e-05, + "loss": 1.1163, + "step": 192 + }, + { + "epoch": 0.013392547359655818, + "grad_norm": 8.708356857299805, + "learning_rate": 2.23121387283237e-05, + "loss": 1.2528, + "step": 193 + }, + { + "epoch": 0.013461938796752481, + "grad_norm": 8.735877990722656, + "learning_rate": 2.2427745664739885e-05, + "loss": 1.5087, + "step": 194 + }, + { + "epoch": 0.013531330233849143, + "grad_norm": 7.234625339508057, + "learning_rate": 2.254335260115607e-05, + "loss": 1.4721, + "step": 195 + }, + { + "epoch": 0.013600721670945806, + "grad_norm": 9.805716514587402, + "learning_rate": 2.2658959537572254e-05, + "loss": 1.6266, + "step": 196 + }, + { + "epoch": 0.013670113108042467, + "grad_norm": 7.779221534729004, + "learning_rate": 2.277456647398844e-05, + "loss": 1.5678, + "step": 197 + }, + { + "epoch": 0.01373950454513913, + "grad_norm": 8.847970008850098, + "learning_rate": 2.2890173410404627e-05, + "loss": 1.6803, + "step": 198 + }, + { + "epoch": 0.013808895982235793, + "grad_norm": 9.314676284790039, + "learning_rate": 2.300578034682081e-05, + "loss": 1.6696, + "step": 199 + }, + { + "epoch": 0.013878287419332454, + "grad_norm": 10.508227348327637, + "learning_rate": 2.3121387283236996e-05, + "loss": 1.5508, + "step": 200 + }, + { + "epoch": 0.013947678856429117, + "grad_norm": 8.208969116210938, + "learning_rate": 2.323699421965318e-05, + "loss": 1.4328, + "step": 201 + }, + { + "epoch": 0.01401707029352578, + "grad_norm": 8.359672546386719, + "learning_rate": 2.3352601156069365e-05, + "loss": 1.67, + "step": 202 + }, + { + "epoch": 0.01408646173062244, + "grad_norm": 8.936582565307617, + "learning_rate": 2.346820809248555e-05, + "loss": 1.2994, + "step": 203 + }, + { + "epoch": 0.014155853167719104, + "grad_norm": 7.023438930511475, + "learning_rate": 2.3583815028901734e-05, + "loss": 1.3315, + "step": 204 + }, + { + "epoch": 0.014225244604815765, + "grad_norm": 9.36505126953125, + "learning_rate": 2.369942196531792e-05, + "loss": 1.6447, + "step": 205 + }, + { + "epoch": 0.014294636041912428, + "grad_norm": 9.281452178955078, + "learning_rate": 2.3815028901734106e-05, + "loss": 1.6708, + "step": 206 + }, + { + "epoch": 0.01436402747900909, + "grad_norm": 10.77733325958252, + "learning_rate": 2.393063583815029e-05, + "loss": 1.553, + "step": 207 + }, + { + "epoch": 0.014433418916105752, + "grad_norm": 8.460607528686523, + "learning_rate": 2.4046242774566475e-05, + "loss": 1.4569, + "step": 208 + }, + { + "epoch": 0.014502810353202415, + "grad_norm": 7.788821220397949, + "learning_rate": 2.4161849710982658e-05, + "loss": 1.2505, + "step": 209 + }, + { + "epoch": 0.014572201790299078, + "grad_norm": 10.125301361083984, + "learning_rate": 2.4277456647398844e-05, + "loss": 1.7113, + "step": 210 + }, + { + "epoch": 0.014641593227395739, + "grad_norm": 10.241157531738281, + "learning_rate": 2.439306358381503e-05, + "loss": 1.4214, + "step": 211 + }, + { + "epoch": 0.014710984664492402, + "grad_norm": 7.281369686126709, + "learning_rate": 2.4508670520231213e-05, + "loss": 1.5223, + "step": 212 + }, + { + "epoch": 0.014780376101589063, + "grad_norm": 8.853864669799805, + "learning_rate": 2.46242774566474e-05, + "loss": 1.5475, + "step": 213 + }, + { + "epoch": 0.014849767538685726, + "grad_norm": 6.984564304351807, + "learning_rate": 2.4739884393063582e-05, + "loss": 1.3217, + "step": 214 + }, + { + "epoch": 0.014919158975782389, + "grad_norm": 10.25341510772705, + "learning_rate": 2.485549132947977e-05, + "loss": 1.4607, + "step": 215 + }, + { + "epoch": 0.01498855041287905, + "grad_norm": 9.323490142822266, + "learning_rate": 2.4971098265895955e-05, + "loss": 1.4162, + "step": 216 + }, + { + "epoch": 0.015057941849975713, + "grad_norm": 9.71323013305664, + "learning_rate": 2.5086705202312137e-05, + "loss": 1.6696, + "step": 217 + }, + { + "epoch": 0.015127333287072376, + "grad_norm": 6.9760870933532715, + "learning_rate": 2.5202312138728324e-05, + "loss": 1.7619, + "step": 218 + }, + { + "epoch": 0.015196724724169037, + "grad_norm": 7.084021091461182, + "learning_rate": 2.531791907514451e-05, + "loss": 1.427, + "step": 219 + }, + { + "epoch": 0.0152661161612657, + "grad_norm": 10.844748497009277, + "learning_rate": 2.5433526011560693e-05, + "loss": 1.507, + "step": 220 + }, + { + "epoch": 0.015335507598362361, + "grad_norm": 7.95432186126709, + "learning_rate": 2.554913294797688e-05, + "loss": 1.2086, + "step": 221 + }, + { + "epoch": 0.015404899035459024, + "grad_norm": 8.049032211303711, + "learning_rate": 2.5664739884393062e-05, + "loss": 1.3043, + "step": 222 + }, + { + "epoch": 0.015474290472555687, + "grad_norm": 10.40876293182373, + "learning_rate": 2.5780346820809248e-05, + "loss": 1.8058, + "step": 223 + }, + { + "epoch": 0.015543681909652348, + "grad_norm": 7.883240222930908, + "learning_rate": 2.5895953757225434e-05, + "loss": 1.5035, + "step": 224 + }, + { + "epoch": 0.015613073346749011, + "grad_norm": 13.374274253845215, + "learning_rate": 2.6011560693641617e-05, + "loss": 1.6509, + "step": 225 + }, + { + "epoch": 0.015682464783845673, + "grad_norm": 9.203773498535156, + "learning_rate": 2.6127167630057803e-05, + "loss": 1.4267, + "step": 226 + }, + { + "epoch": 0.015751856220942335, + "grad_norm": 9.000197410583496, + "learning_rate": 2.6242774566473986e-05, + "loss": 1.5673, + "step": 227 + }, + { + "epoch": 0.015821247658039, + "grad_norm": 9.094684600830078, + "learning_rate": 2.6358381502890172e-05, + "loss": 1.478, + "step": 228 + }, + { + "epoch": 0.01589063909513566, + "grad_norm": 8.799365997314453, + "learning_rate": 2.647398843930636e-05, + "loss": 1.5376, + "step": 229 + }, + { + "epoch": 0.015960030532232324, + "grad_norm": 8.421977043151855, + "learning_rate": 2.658959537572254e-05, + "loss": 1.3581, + "step": 230 + }, + { + "epoch": 0.016029421969328984, + "grad_norm": 8.079449653625488, + "learning_rate": 2.6705202312138728e-05, + "loss": 1.387, + "step": 231 + }, + { + "epoch": 0.016098813406425647, + "grad_norm": 16.44562530517578, + "learning_rate": 2.6820809248554914e-05, + "loss": 1.5029, + "step": 232 + }, + { + "epoch": 0.01616820484352231, + "grad_norm": 9.084707260131836, + "learning_rate": 2.6936416184971097e-05, + "loss": 1.5706, + "step": 233 + }, + { + "epoch": 0.016237596280618972, + "grad_norm": 11.46006965637207, + "learning_rate": 2.7052023121387283e-05, + "loss": 1.6278, + "step": 234 + }, + { + "epoch": 0.016306987717715635, + "grad_norm": 7.8552117347717285, + "learning_rate": 2.7167630057803466e-05, + "loss": 1.59, + "step": 235 + }, + { + "epoch": 0.016376379154812295, + "grad_norm": 9.276106834411621, + "learning_rate": 2.7283236994219652e-05, + "loss": 1.7408, + "step": 236 + }, + { + "epoch": 0.016445770591908958, + "grad_norm": 7.083552360534668, + "learning_rate": 2.7398843930635838e-05, + "loss": 1.1816, + "step": 237 + }, + { + "epoch": 0.01651516202900562, + "grad_norm": 7.653381824493408, + "learning_rate": 2.751445086705202e-05, + "loss": 1.7395, + "step": 238 + }, + { + "epoch": 0.016584553466102284, + "grad_norm": 7.880470275878906, + "learning_rate": 2.7630057803468207e-05, + "loss": 1.2688, + "step": 239 + }, + { + "epoch": 0.016653944903198947, + "grad_norm": 9.05591106414795, + "learning_rate": 2.7745664739884393e-05, + "loss": 1.4227, + "step": 240 + }, + { + "epoch": 0.016723336340295606, + "grad_norm": 7.048729419708252, + "learning_rate": 2.7861271676300583e-05, + "loss": 1.2682, + "step": 241 + }, + { + "epoch": 0.01679272777739227, + "grad_norm": 10.156103134155273, + "learning_rate": 2.7976878612716766e-05, + "loss": 1.6087, + "step": 242 + }, + { + "epoch": 0.016862119214488932, + "grad_norm": 9.055037498474121, + "learning_rate": 2.8092485549132952e-05, + "loss": 1.5301, + "step": 243 + }, + { + "epoch": 0.016931510651585595, + "grad_norm": 9.359967231750488, + "learning_rate": 2.8208092485549138e-05, + "loss": 1.3034, + "step": 244 + }, + { + "epoch": 0.017000902088682258, + "grad_norm": 9.756706237792969, + "learning_rate": 2.832369942196532e-05, + "loss": 1.5412, + "step": 245 + }, + { + "epoch": 0.017070293525778917, + "grad_norm": 9.309571266174316, + "learning_rate": 2.8439306358381507e-05, + "loss": 1.5253, + "step": 246 + }, + { + "epoch": 0.01713968496287558, + "grad_norm": 7.381655216217041, + "learning_rate": 2.855491329479769e-05, + "loss": 1.5509, + "step": 247 + }, + { + "epoch": 0.017209076399972243, + "grad_norm": 11.220626831054688, + "learning_rate": 2.8670520231213876e-05, + "loss": 1.495, + "step": 248 + }, + { + "epoch": 0.017278467837068906, + "grad_norm": 9.800970077514648, + "learning_rate": 2.8786127167630062e-05, + "loss": 1.5405, + "step": 249 + }, + { + "epoch": 0.01734785927416557, + "grad_norm": 6.749784469604492, + "learning_rate": 2.8901734104046245e-05, + "loss": 1.385, + "step": 250 + }, + { + "epoch": 0.017417250711262232, + "grad_norm": 8.788182258605957, + "learning_rate": 2.901734104046243e-05, + "loss": 1.4415, + "step": 251 + }, + { + "epoch": 0.01748664214835889, + "grad_norm": 9.811159133911133, + "learning_rate": 2.9132947976878618e-05, + "loss": 1.724, + "step": 252 + }, + { + "epoch": 0.017556033585455554, + "grad_norm": 7.410183429718018, + "learning_rate": 2.92485549132948e-05, + "loss": 1.4805, + "step": 253 + }, + { + "epoch": 0.017625425022552217, + "grad_norm": 8.544808387756348, + "learning_rate": 2.9364161849710987e-05, + "loss": 1.9162, + "step": 254 + }, + { + "epoch": 0.01769481645964888, + "grad_norm": 7.9746575355529785, + "learning_rate": 2.947976878612717e-05, + "loss": 1.5677, + "step": 255 + }, + { + "epoch": 0.017764207896745543, + "grad_norm": 7.656627178192139, + "learning_rate": 2.9595375722543356e-05, + "loss": 1.2679, + "step": 256 + }, + { + "epoch": 0.017833599333842202, + "grad_norm": 6.255059719085693, + "learning_rate": 2.9710982658959542e-05, + "loss": 1.053, + "step": 257 + }, + { + "epoch": 0.017902990770938865, + "grad_norm": 13.07886028289795, + "learning_rate": 2.9826589595375725e-05, + "loss": 1.5853, + "step": 258 + }, + { + "epoch": 0.01797238220803553, + "grad_norm": 9.59166431427002, + "learning_rate": 2.994219653179191e-05, + "loss": 1.3743, + "step": 259 + }, + { + "epoch": 0.01804177364513219, + "grad_norm": 7.841953277587891, + "learning_rate": 3.0057803468208097e-05, + "loss": 1.2733, + "step": 260 + }, + { + "epoch": 0.018111165082228854, + "grad_norm": 10.283413887023926, + "learning_rate": 3.017341040462428e-05, + "loss": 1.5229, + "step": 261 + }, + { + "epoch": 0.018180556519325514, + "grad_norm": 7.448075771331787, + "learning_rate": 3.0289017341040466e-05, + "loss": 1.3038, + "step": 262 + }, + { + "epoch": 0.018249947956422177, + "grad_norm": 8.793776512145996, + "learning_rate": 3.040462427745665e-05, + "loss": 1.4821, + "step": 263 + }, + { + "epoch": 0.01831933939351884, + "grad_norm": 8.291214942932129, + "learning_rate": 3.0520231213872835e-05, + "loss": 1.3335, + "step": 264 + }, + { + "epoch": 0.018388730830615502, + "grad_norm": 6.914064884185791, + "learning_rate": 3.063583815028902e-05, + "loss": 1.0479, + "step": 265 + }, + { + "epoch": 0.018458122267712165, + "grad_norm": 8.128153800964355, + "learning_rate": 3.075144508670521e-05, + "loss": 1.348, + "step": 266 + }, + { + "epoch": 0.01852751370480883, + "grad_norm": 8.50592041015625, + "learning_rate": 3.086705202312139e-05, + "loss": 1.2762, + "step": 267 + }, + { + "epoch": 0.018596905141905488, + "grad_norm": 9.933420181274414, + "learning_rate": 3.0982658959537573e-05, + "loss": 1.2689, + "step": 268 + }, + { + "epoch": 0.01866629657900215, + "grad_norm": 8.8480863571167, + "learning_rate": 3.109826589595376e-05, + "loss": 1.2992, + "step": 269 + }, + { + "epoch": 0.018735688016098814, + "grad_norm": 9.920266151428223, + "learning_rate": 3.1213872832369946e-05, + "loss": 1.7103, + "step": 270 + }, + { + "epoch": 0.018805079453195477, + "grad_norm": 7.160682201385498, + "learning_rate": 3.132947976878613e-05, + "loss": 1.2611, + "step": 271 + }, + { + "epoch": 0.01887447089029214, + "grad_norm": 8.31418514251709, + "learning_rate": 3.144508670520231e-05, + "loss": 1.6002, + "step": 272 + }, + { + "epoch": 0.0189438623273888, + "grad_norm": 8.697206497192383, + "learning_rate": 3.15606936416185e-05, + "loss": 1.4058, + "step": 273 + }, + { + "epoch": 0.019013253764485462, + "grad_norm": 7.4854254722595215, + "learning_rate": 3.1676300578034684e-05, + "loss": 1.3983, + "step": 274 + }, + { + "epoch": 0.019082645201582125, + "grad_norm": 15.1348237991333, + "learning_rate": 3.179190751445087e-05, + "loss": 0.9598, + "step": 275 + }, + { + "epoch": 0.019152036638678788, + "grad_norm": 7.185125350952148, + "learning_rate": 3.1907514450867056e-05, + "loss": 1.5804, + "step": 276 + }, + { + "epoch": 0.01922142807577545, + "grad_norm": 7.001240253448486, + "learning_rate": 3.202312138728324e-05, + "loss": 1.1868, + "step": 277 + }, + { + "epoch": 0.01929081951287211, + "grad_norm": 7.819228649139404, + "learning_rate": 3.213872832369942e-05, + "loss": 1.434, + "step": 278 + }, + { + "epoch": 0.019360210949968773, + "grad_norm": 8.119030952453613, + "learning_rate": 3.225433526011561e-05, + "loss": 1.484, + "step": 279 + }, + { + "epoch": 0.019429602387065436, + "grad_norm": 8.481521606445312, + "learning_rate": 3.2369942196531794e-05, + "loss": 1.167, + "step": 280 + }, + { + "epoch": 0.0194989938241621, + "grad_norm": 10.350101470947266, + "learning_rate": 3.248554913294798e-05, + "loss": 1.7312, + "step": 281 + }, + { + "epoch": 0.019568385261258762, + "grad_norm": 7.093286037445068, + "learning_rate": 3.260115606936417e-05, + "loss": 1.2612, + "step": 282 + }, + { + "epoch": 0.01963777669835542, + "grad_norm": 8.93200969696045, + "learning_rate": 3.2716763005780346e-05, + "loss": 1.4002, + "step": 283 + }, + { + "epoch": 0.019707168135452084, + "grad_norm": 9.990497589111328, + "learning_rate": 3.283236994219653e-05, + "loss": 1.0727, + "step": 284 + }, + { + "epoch": 0.019776559572548747, + "grad_norm": 8.13919448852539, + "learning_rate": 3.294797687861272e-05, + "loss": 1.2284, + "step": 285 + }, + { + "epoch": 0.01984595100964541, + "grad_norm": 11.466733932495117, + "learning_rate": 3.3063583815028905e-05, + "loss": 1.5922, + "step": 286 + }, + { + "epoch": 0.019915342446742073, + "grad_norm": 8.529134750366211, + "learning_rate": 3.317919075144509e-05, + "loss": 1.2574, + "step": 287 + }, + { + "epoch": 0.019984733883838736, + "grad_norm": 11.819015502929688, + "learning_rate": 3.329479768786127e-05, + "loss": 1.7099, + "step": 288 + }, + { + "epoch": 0.020054125320935395, + "grad_norm": 7.0236124992370605, + "learning_rate": 3.341040462427746e-05, + "loss": 1.0931, + "step": 289 + }, + { + "epoch": 0.02012351675803206, + "grad_norm": 7.128871917724609, + "learning_rate": 3.352601156069364e-05, + "loss": 1.1847, + "step": 290 + }, + { + "epoch": 0.02019290819512872, + "grad_norm": 10.233290672302246, + "learning_rate": 3.364161849710983e-05, + "loss": 1.2739, + "step": 291 + }, + { + "epoch": 0.020262299632225384, + "grad_norm": 14.914508819580078, + "learning_rate": 3.3757225433526015e-05, + "loss": 1.6147, + "step": 292 + }, + { + "epoch": 0.020331691069322047, + "grad_norm": 8.006830215454102, + "learning_rate": 3.3872832369942195e-05, + "loss": 1.1682, + "step": 293 + }, + { + "epoch": 0.020401082506418707, + "grad_norm": 7.947633266448975, + "learning_rate": 3.398843930635838e-05, + "loss": 1.306, + "step": 294 + }, + { + "epoch": 0.02047047394351537, + "grad_norm": 8.381584167480469, + "learning_rate": 3.410404624277457e-05, + "loss": 1.567, + "step": 295 + }, + { + "epoch": 0.020539865380612032, + "grad_norm": 9.315205574035645, + "learning_rate": 3.4219653179190754e-05, + "loss": 1.5195, + "step": 296 + }, + { + "epoch": 0.020609256817708695, + "grad_norm": 7.573694229125977, + "learning_rate": 3.433526011560694e-05, + "loss": 1.4812, + "step": 297 + }, + { + "epoch": 0.020678648254805358, + "grad_norm": 7.974030494689941, + "learning_rate": 3.445086705202312e-05, + "loss": 1.0627, + "step": 298 + }, + { + "epoch": 0.020748039691902018, + "grad_norm": 9.96582317352295, + "learning_rate": 3.4566473988439305e-05, + "loss": 1.2275, + "step": 299 + }, + { + "epoch": 0.02081743112899868, + "grad_norm": 6.986239433288574, + "learning_rate": 3.468208092485549e-05, + "loss": 0.9275, + "step": 300 + }, + { + "epoch": 0.020886822566095344, + "grad_norm": 7.219080448150635, + "learning_rate": 3.479768786127168e-05, + "loss": 1.4205, + "step": 301 + }, + { + "epoch": 0.020956214003192006, + "grad_norm": 7.968305587768555, + "learning_rate": 3.4913294797687864e-05, + "loss": 1.3569, + "step": 302 + }, + { + "epoch": 0.02102560544028867, + "grad_norm": 9.210378646850586, + "learning_rate": 3.502890173410405e-05, + "loss": 1.2616, + "step": 303 + }, + { + "epoch": 0.021094996877385332, + "grad_norm": 8.32539176940918, + "learning_rate": 3.514450867052023e-05, + "loss": 1.4583, + "step": 304 + }, + { + "epoch": 0.021164388314481992, + "grad_norm": 9.323540687561035, + "learning_rate": 3.5260115606936416e-05, + "loss": 1.4487, + "step": 305 + }, + { + "epoch": 0.021233779751578655, + "grad_norm": 7.314230918884277, + "learning_rate": 3.53757225433526e-05, + "loss": 1.2243, + "step": 306 + }, + { + "epoch": 0.021303171188675318, + "grad_norm": 8.251792907714844, + "learning_rate": 3.549132947976879e-05, + "loss": 1.0206, + "step": 307 + }, + { + "epoch": 0.02137256262577198, + "grad_norm": 7.845508098602295, + "learning_rate": 3.5606936416184975e-05, + "loss": 1.3471, + "step": 308 + }, + { + "epoch": 0.021441954062868643, + "grad_norm": 7.738494396209717, + "learning_rate": 3.5722543352601154e-05, + "loss": 1.3694, + "step": 309 + }, + { + "epoch": 0.021511345499965303, + "grad_norm": 8.963193893432617, + "learning_rate": 3.583815028901734e-05, + "loss": 1.6028, + "step": 310 + }, + { + "epoch": 0.021580736937061966, + "grad_norm": 6.752964019775391, + "learning_rate": 3.5953757225433526e-05, + "loss": 1.1423, + "step": 311 + }, + { + "epoch": 0.02165012837415863, + "grad_norm": 6.809983730316162, + "learning_rate": 3.606936416184971e-05, + "loss": 1.2884, + "step": 312 + }, + { + "epoch": 0.021719519811255292, + "grad_norm": 9.428483963012695, + "learning_rate": 3.61849710982659e-05, + "loss": 1.4135, + "step": 313 + }, + { + "epoch": 0.021788911248351955, + "grad_norm": 12.471772193908691, + "learning_rate": 3.630057803468208e-05, + "loss": 1.3746, + "step": 314 + }, + { + "epoch": 0.021858302685448614, + "grad_norm": 6.644201755523682, + "learning_rate": 3.6416184971098265e-05, + "loss": 1.1402, + "step": 315 + }, + { + "epoch": 0.021927694122545277, + "grad_norm": 9.581330299377441, + "learning_rate": 3.653179190751445e-05, + "loss": 1.2496, + "step": 316 + }, + { + "epoch": 0.02199708555964194, + "grad_norm": 9.743836402893066, + "learning_rate": 3.664739884393064e-05, + "loss": 1.56, + "step": 317 + }, + { + "epoch": 0.022066476996738603, + "grad_norm": 8.122932434082031, + "learning_rate": 3.676300578034682e-05, + "loss": 1.3686, + "step": 318 + }, + { + "epoch": 0.022135868433835266, + "grad_norm": 8.98464584350586, + "learning_rate": 3.6878612716763e-05, + "loss": 0.9764, + "step": 319 + }, + { + "epoch": 0.022205259870931925, + "grad_norm": 7.877935409545898, + "learning_rate": 3.699421965317919e-05, + "loss": 1.4949, + "step": 320 + }, + { + "epoch": 0.022274651308028588, + "grad_norm": 12.786532402038574, + "learning_rate": 3.7109826589595375e-05, + "loss": 2.1724, + "step": 321 + }, + { + "epoch": 0.02234404274512525, + "grad_norm": 6.780661106109619, + "learning_rate": 3.722543352601156e-05, + "loss": 0.9474, + "step": 322 + }, + { + "epoch": 0.022413434182221914, + "grad_norm": 8.371500015258789, + "learning_rate": 3.734104046242775e-05, + "loss": 1.5355, + "step": 323 + }, + { + "epoch": 0.022482825619318577, + "grad_norm": 7.253002643585205, + "learning_rate": 3.7456647398843934e-05, + "loss": 1.1442, + "step": 324 + }, + { + "epoch": 0.02255221705641524, + "grad_norm": 8.8172607421875, + "learning_rate": 3.757225433526011e-05, + "loss": 1.3707, + "step": 325 + }, + { + "epoch": 0.0226216084935119, + "grad_norm": 8.36879825592041, + "learning_rate": 3.76878612716763e-05, + "loss": 1.4176, + "step": 326 + }, + { + "epoch": 0.022690999930608562, + "grad_norm": 8.265416145324707, + "learning_rate": 3.7803468208092486e-05, + "loss": 1.2948, + "step": 327 + }, + { + "epoch": 0.022760391367705225, + "grad_norm": 8.152652740478516, + "learning_rate": 3.791907514450867e-05, + "loss": 1.3295, + "step": 328 + }, + { + "epoch": 0.022829782804801888, + "grad_norm": 7.184534549713135, + "learning_rate": 3.803468208092486e-05, + "loss": 1.377, + "step": 329 + }, + { + "epoch": 0.02289917424189855, + "grad_norm": 6.826780796051025, + "learning_rate": 3.815028901734104e-05, + "loss": 1.0455, + "step": 330 + }, + { + "epoch": 0.02296856567899521, + "grad_norm": 7.111767292022705, + "learning_rate": 3.8265895953757224e-05, + "loss": 1.3094, + "step": 331 + }, + { + "epoch": 0.023037957116091873, + "grad_norm": 11.068435668945312, + "learning_rate": 3.838150289017341e-05, + "loss": 1.176, + "step": 332 + }, + { + "epoch": 0.023107348553188536, + "grad_norm": 9.089969635009766, + "learning_rate": 3.8497109826589596e-05, + "loss": 1.2722, + "step": 333 + }, + { + "epoch": 0.0231767399902852, + "grad_norm": 9.449152946472168, + "learning_rate": 3.861271676300578e-05, + "loss": 1.2284, + "step": 334 + }, + { + "epoch": 0.023246131427381862, + "grad_norm": 8.346184730529785, + "learning_rate": 3.872832369942196e-05, + "loss": 1.0908, + "step": 335 + }, + { + "epoch": 0.023315522864478522, + "grad_norm": 9.131814002990723, + "learning_rate": 3.884393063583815e-05, + "loss": 1.435, + "step": 336 + }, + { + "epoch": 0.023384914301575185, + "grad_norm": 10.11670970916748, + "learning_rate": 3.895953757225434e-05, + "loss": 1.5617, + "step": 337 + }, + { + "epoch": 0.023454305738671848, + "grad_norm": 9.10450553894043, + "learning_rate": 3.907514450867053e-05, + "loss": 1.4059, + "step": 338 + }, + { + "epoch": 0.02352369717576851, + "grad_norm": 8.070188522338867, + "learning_rate": 3.9190751445086707e-05, + "loss": 1.2795, + "step": 339 + }, + { + "epoch": 0.023593088612865173, + "grad_norm": 9.955215454101562, + "learning_rate": 3.930635838150289e-05, + "loss": 1.0977, + "step": 340 + }, + { + "epoch": 0.023662480049961836, + "grad_norm": 8.840959548950195, + "learning_rate": 3.942196531791908e-05, + "loss": 1.2033, + "step": 341 + }, + { + "epoch": 0.023731871487058496, + "grad_norm": 11.950058937072754, + "learning_rate": 3.9537572254335265e-05, + "loss": 1.3593, + "step": 342 + }, + { + "epoch": 0.02380126292415516, + "grad_norm": 7.211612701416016, + "learning_rate": 3.965317919075145e-05, + "loss": 1.2123, + "step": 343 + }, + { + "epoch": 0.02387065436125182, + "grad_norm": 8.410841941833496, + "learning_rate": 3.976878612716764e-05, + "loss": 1.4763, + "step": 344 + }, + { + "epoch": 0.023940045798348485, + "grad_norm": 8.64093017578125, + "learning_rate": 3.988439306358382e-05, + "loss": 1.2624, + "step": 345 + }, + { + "epoch": 0.024009437235445148, + "grad_norm": 8.846633911132812, + "learning_rate": 4e-05, + "loss": 1.0592, + "step": 346 + }, + { + "epoch": 0.024078828672541807, + "grad_norm": 8.834220886230469, + "learning_rate": 4.011560693641619e-05, + "loss": 1.8233, + "step": 347 + }, + { + "epoch": 0.02414822010963847, + "grad_norm": 12.323884010314941, + "learning_rate": 4.0231213872832376e-05, + "loss": 1.5971, + "step": 348 + }, + { + "epoch": 0.024217611546735133, + "grad_norm": 11.278389930725098, + "learning_rate": 4.034682080924856e-05, + "loss": 1.5341, + "step": 349 + }, + { + "epoch": 0.024287002983831796, + "grad_norm": 7.0283074378967285, + "learning_rate": 4.046242774566474e-05, + "loss": 1.1549, + "step": 350 + }, + { + "epoch": 0.02435639442092846, + "grad_norm": 8.903668403625488, + "learning_rate": 4.057803468208093e-05, + "loss": 1.5307, + "step": 351 + }, + { + "epoch": 0.024425785858025118, + "grad_norm": 7.922576904296875, + "learning_rate": 4.0693641618497114e-05, + "loss": 1.4153, + "step": 352 + }, + { + "epoch": 0.02449517729512178, + "grad_norm": 8.038032531738281, + "learning_rate": 4.08092485549133e-05, + "loss": 1.2572, + "step": 353 + }, + { + "epoch": 0.024564568732218444, + "grad_norm": 8.196563720703125, + "learning_rate": 4.0924855491329486e-05, + "loss": 1.2865, + "step": 354 + }, + { + "epoch": 0.024633960169315107, + "grad_norm": 10.058512687683105, + "learning_rate": 4.1040462427745666e-05, + "loss": 1.2123, + "step": 355 + }, + { + "epoch": 0.02470335160641177, + "grad_norm": 9.173090934753418, + "learning_rate": 4.115606936416185e-05, + "loss": 1.5001, + "step": 356 + }, + { + "epoch": 0.02477274304350843, + "grad_norm": 8.364827156066895, + "learning_rate": 4.127167630057804e-05, + "loss": 1.472, + "step": 357 + }, + { + "epoch": 0.024842134480605092, + "grad_norm": 9.770203590393066, + "learning_rate": 4.1387283236994224e-05, + "loss": 1.218, + "step": 358 + }, + { + "epoch": 0.024911525917701755, + "grad_norm": 7.633975505828857, + "learning_rate": 4.150289017341041e-05, + "loss": 1.4782, + "step": 359 + }, + { + "epoch": 0.024980917354798418, + "grad_norm": 8.202336311340332, + "learning_rate": 4.161849710982659e-05, + "loss": 1.5881, + "step": 360 + }, + { + "epoch": 0.02505030879189508, + "grad_norm": 7.964822292327881, + "learning_rate": 4.1734104046242776e-05, + "loss": 1.5151, + "step": 361 + }, + { + "epoch": 0.025119700228991744, + "grad_norm": 7.7173871994018555, + "learning_rate": 4.184971098265896e-05, + "loss": 1.4847, + "step": 362 + }, + { + "epoch": 0.025189091666088403, + "grad_norm": 7.018986701965332, + "learning_rate": 4.196531791907515e-05, + "loss": 1.3224, + "step": 363 + }, + { + "epoch": 0.025258483103185066, + "grad_norm": 6.36120080947876, + "learning_rate": 4.2080924855491335e-05, + "loss": 1.0451, + "step": 364 + }, + { + "epoch": 0.02532787454028173, + "grad_norm": 7.655276298522949, + "learning_rate": 4.2196531791907514e-05, + "loss": 1.1698, + "step": 365 + }, + { + "epoch": 0.025397265977378392, + "grad_norm": 7.203214645385742, + "learning_rate": 4.23121387283237e-05, + "loss": 1.4568, + "step": 366 + }, + { + "epoch": 0.025466657414475055, + "grad_norm": 5.308102607727051, + "learning_rate": 4.242774566473989e-05, + "loss": 1.0518, + "step": 367 + }, + { + "epoch": 0.025536048851571715, + "grad_norm": 7.520718097686768, + "learning_rate": 4.254335260115607e-05, + "loss": 1.4556, + "step": 368 + }, + { + "epoch": 0.025605440288668378, + "grad_norm": 8.533038139343262, + "learning_rate": 4.265895953757226e-05, + "loss": 1.5887, + "step": 369 + }, + { + "epoch": 0.02567483172576504, + "grad_norm": 7.773636817932129, + "learning_rate": 4.2774566473988445e-05, + "loss": 1.2177, + "step": 370 + }, + { + "epoch": 0.025744223162861703, + "grad_norm": 7.447896480560303, + "learning_rate": 4.2890173410404625e-05, + "loss": 1.2411, + "step": 371 + }, + { + "epoch": 0.025813614599958366, + "grad_norm": 7.492413520812988, + "learning_rate": 4.300578034682081e-05, + "loss": 1.2102, + "step": 372 + }, + { + "epoch": 0.025883006037055026, + "grad_norm": 8.21690559387207, + "learning_rate": 4.3121387283237e-05, + "loss": 1.0782, + "step": 373 + }, + { + "epoch": 0.02595239747415169, + "grad_norm": 6.735875129699707, + "learning_rate": 4.3236994219653183e-05, + "loss": 1.12, + "step": 374 + }, + { + "epoch": 0.02602178891124835, + "grad_norm": 7.129016399383545, + "learning_rate": 4.335260115606937e-05, + "loss": 0.9489, + "step": 375 + }, + { + "epoch": 0.026091180348345015, + "grad_norm": 7.7948150634765625, + "learning_rate": 4.346820809248555e-05, + "loss": 1.398, + "step": 376 + }, + { + "epoch": 0.026160571785441677, + "grad_norm": 9.328797340393066, + "learning_rate": 4.3583815028901735e-05, + "loss": 1.4012, + "step": 377 + }, + { + "epoch": 0.02622996322253834, + "grad_norm": 9.744799613952637, + "learning_rate": 4.369942196531792e-05, + "loss": 1.3258, + "step": 378 + }, + { + "epoch": 0.026299354659635, + "grad_norm": 6.667611598968506, + "learning_rate": 4.381502890173411e-05, + "loss": 1.3931, + "step": 379 + }, + { + "epoch": 0.026368746096731663, + "grad_norm": 10.830458641052246, + "learning_rate": 4.3930635838150294e-05, + "loss": 1.3151, + "step": 380 + }, + { + "epoch": 0.026438137533828326, + "grad_norm": 6.436756134033203, + "learning_rate": 4.404624277456647e-05, + "loss": 1.1524, + "step": 381 + }, + { + "epoch": 0.02650752897092499, + "grad_norm": 9.624730110168457, + "learning_rate": 4.416184971098266e-05, + "loss": 1.5549, + "step": 382 + }, + { + "epoch": 0.02657692040802165, + "grad_norm": 8.216463088989258, + "learning_rate": 4.4277456647398846e-05, + "loss": 1.3934, + "step": 383 + }, + { + "epoch": 0.02664631184511831, + "grad_norm": 8.346541404724121, + "learning_rate": 4.439306358381503e-05, + "loss": 1.3547, + "step": 384 + }, + { + "epoch": 0.026715703282214974, + "grad_norm": 6.576135158538818, + "learning_rate": 4.450867052023122e-05, + "loss": 1.3841, + "step": 385 + }, + { + "epoch": 0.026785094719311637, + "grad_norm": 7.489294528961182, + "learning_rate": 4.46242774566474e-05, + "loss": 1.4014, + "step": 386 + }, + { + "epoch": 0.0268544861564083, + "grad_norm": 8.063232421875, + "learning_rate": 4.4739884393063584e-05, + "loss": 1.4095, + "step": 387 + }, + { + "epoch": 0.026923877593504963, + "grad_norm": 7.395705699920654, + "learning_rate": 4.485549132947977e-05, + "loss": 1.1777, + "step": 388 + }, + { + "epoch": 0.026993269030601622, + "grad_norm": 7.2524189949035645, + "learning_rate": 4.4971098265895956e-05, + "loss": 1.4102, + "step": 389 + }, + { + "epoch": 0.027062660467698285, + "grad_norm": 6.5494842529296875, + "learning_rate": 4.508670520231214e-05, + "loss": 1.2629, + "step": 390 + }, + { + "epoch": 0.027132051904794948, + "grad_norm": 8.47513484954834, + "learning_rate": 4.520231213872833e-05, + "loss": 1.1084, + "step": 391 + }, + { + "epoch": 0.02720144334189161, + "grad_norm": 9.662997245788574, + "learning_rate": 4.531791907514451e-05, + "loss": 1.2753, + "step": 392 + }, + { + "epoch": 0.027270834778988274, + "grad_norm": 7.143062114715576, + "learning_rate": 4.5433526011560694e-05, + "loss": 1.3287, + "step": 393 + }, + { + "epoch": 0.027340226216084933, + "grad_norm": 6.802155494689941, + "learning_rate": 4.554913294797688e-05, + "loss": 1.1162, + "step": 394 + }, + { + "epoch": 0.027409617653181596, + "grad_norm": 9.34698486328125, + "learning_rate": 4.566473988439307e-05, + "loss": 1.3365, + "step": 395 + }, + { + "epoch": 0.02747900909027826, + "grad_norm": 9.462413787841797, + "learning_rate": 4.578034682080925e-05, + "loss": 1.3541, + "step": 396 + }, + { + "epoch": 0.027548400527374922, + "grad_norm": 6.404652118682861, + "learning_rate": 4.589595375722543e-05, + "loss": 1.2287, + "step": 397 + }, + { + "epoch": 0.027617791964471585, + "grad_norm": 6.4934258460998535, + "learning_rate": 4.601156069364162e-05, + "loss": 1.242, + "step": 398 + }, + { + "epoch": 0.027687183401568248, + "grad_norm": 7.677034378051758, + "learning_rate": 4.6127167630057805e-05, + "loss": 1.2833, + "step": 399 + }, + { + "epoch": 0.027756574838664907, + "grad_norm": 7.806981563568115, + "learning_rate": 4.624277456647399e-05, + "loss": 1.513, + "step": 400 + }, + { + "epoch": 0.02782596627576157, + "grad_norm": 7.7379889488220215, + "learning_rate": 4.635838150289018e-05, + "loss": 1.3042, + "step": 401 + }, + { + "epoch": 0.027895357712858233, + "grad_norm": 8.278910636901855, + "learning_rate": 4.647398843930636e-05, + "loss": 1.1377, + "step": 402 + }, + { + "epoch": 0.027964749149954896, + "grad_norm": 8.863214492797852, + "learning_rate": 4.658959537572254e-05, + "loss": 1.4773, + "step": 403 + }, + { + "epoch": 0.02803414058705156, + "grad_norm": 10.159278869628906, + "learning_rate": 4.670520231213873e-05, + "loss": 1.2265, + "step": 404 + }, + { + "epoch": 0.02810353202414822, + "grad_norm": 9.939448356628418, + "learning_rate": 4.6820809248554915e-05, + "loss": 1.0279, + "step": 405 + }, + { + "epoch": 0.02817292346124488, + "grad_norm": 8.642842292785645, + "learning_rate": 4.69364161849711e-05, + "loss": 1.187, + "step": 406 + }, + { + "epoch": 0.028242314898341545, + "grad_norm": 8.012290000915527, + "learning_rate": 4.705202312138728e-05, + "loss": 1.3038, + "step": 407 + }, + { + "epoch": 0.028311706335438207, + "grad_norm": 10.756101608276367, + "learning_rate": 4.716763005780347e-05, + "loss": 1.4832, + "step": 408 + }, + { + "epoch": 0.02838109777253487, + "grad_norm": 11.81616497039795, + "learning_rate": 4.7283236994219653e-05, + "loss": 1.1345, + "step": 409 + }, + { + "epoch": 0.02845048920963153, + "grad_norm": 7.05643367767334, + "learning_rate": 4.739884393063584e-05, + "loss": 1.1295, + "step": 410 + }, + { + "epoch": 0.028519880646728193, + "grad_norm": 7.196845531463623, + "learning_rate": 4.7514450867052026e-05, + "loss": 1.2044, + "step": 411 + }, + { + "epoch": 0.028589272083824856, + "grad_norm": 9.162343978881836, + "learning_rate": 4.763005780346821e-05, + "loss": 1.3586, + "step": 412 + }, + { + "epoch": 0.02865866352092152, + "grad_norm": 6.7770771980285645, + "learning_rate": 4.774566473988439e-05, + "loss": 1.3023, + "step": 413 + }, + { + "epoch": 0.02872805495801818, + "grad_norm": 8.87976360321045, + "learning_rate": 4.786127167630058e-05, + "loss": 1.0513, + "step": 414 + }, + { + "epoch": 0.028797446395114844, + "grad_norm": 5.673719882965088, + "learning_rate": 4.7976878612716764e-05, + "loss": 1.0946, + "step": 415 + }, + { + "epoch": 0.028866837832211504, + "grad_norm": 8.886439323425293, + "learning_rate": 4.809248554913295e-05, + "loss": 1.3069, + "step": 416 + }, + { + "epoch": 0.028936229269308167, + "grad_norm": 7.717283248901367, + "learning_rate": 4.8208092485549136e-05, + "loss": 1.2986, + "step": 417 + }, + { + "epoch": 0.02900562070640483, + "grad_norm": 9.30689525604248, + "learning_rate": 4.8323699421965316e-05, + "loss": 1.287, + "step": 418 + }, + { + "epoch": 0.029075012143501493, + "grad_norm": 7.44536828994751, + "learning_rate": 4.84393063583815e-05, + "loss": 1.1664, + "step": 419 + }, + { + "epoch": 0.029144403580598156, + "grad_norm": 6.739874362945557, + "learning_rate": 4.855491329479769e-05, + "loss": 1.3671, + "step": 420 + }, + { + "epoch": 0.029213795017694815, + "grad_norm": 8.688088417053223, + "learning_rate": 4.8670520231213874e-05, + "loss": 1.4489, + "step": 421 + }, + { + "epoch": 0.029283186454791478, + "grad_norm": 6.528660774230957, + "learning_rate": 4.878612716763006e-05, + "loss": 0.9946, + "step": 422 + }, + { + "epoch": 0.02935257789188814, + "grad_norm": 7.350855350494385, + "learning_rate": 4.890173410404624e-05, + "loss": 1.1971, + "step": 423 + }, + { + "epoch": 0.029421969328984804, + "grad_norm": 7.9066009521484375, + "learning_rate": 4.9017341040462426e-05, + "loss": 1.1394, + "step": 424 + }, + { + "epoch": 0.029491360766081467, + "grad_norm": 11.922992706298828, + "learning_rate": 4.913294797687861e-05, + "loss": 1.189, + "step": 425 + }, + { + "epoch": 0.029560752203178126, + "grad_norm": 11.748053550720215, + "learning_rate": 4.92485549132948e-05, + "loss": 1.5058, + "step": 426 + }, + { + "epoch": 0.02963014364027479, + "grad_norm": 9.480945587158203, + "learning_rate": 4.9364161849710985e-05, + "loss": 1.0642, + "step": 427 + }, + { + "epoch": 0.029699535077371452, + "grad_norm": 6.968897342681885, + "learning_rate": 4.9479768786127164e-05, + "loss": 1.1351, + "step": 428 + }, + { + "epoch": 0.029768926514468115, + "grad_norm": 8.315743446350098, + "learning_rate": 4.959537572254335e-05, + "loss": 1.201, + "step": 429 + }, + { + "epoch": 0.029838317951564778, + "grad_norm": 7.697081089019775, + "learning_rate": 4.971098265895954e-05, + "loss": 1.1971, + "step": 430 + }, + { + "epoch": 0.02990770938866144, + "grad_norm": 7.559666156768799, + "learning_rate": 4.982658959537572e-05, + "loss": 1.3203, + "step": 431 + }, + { + "epoch": 0.0299771008257581, + "grad_norm": 8.65117073059082, + "learning_rate": 4.994219653179191e-05, + "loss": 1.5184, + "step": 432 + }, + { + "epoch": 0.030046492262854763, + "grad_norm": 7.873926162719727, + "learning_rate": 5.0057803468208096e-05, + "loss": 1.6422, + "step": 433 + }, + { + "epoch": 0.030115883699951426, + "grad_norm": 5.550638675689697, + "learning_rate": 5.0173410404624275e-05, + "loss": 1.1381, + "step": 434 + }, + { + "epoch": 0.03018527513704809, + "grad_norm": 7.212305068969727, + "learning_rate": 5.028901734104047e-05, + "loss": 1.1956, + "step": 435 + }, + { + "epoch": 0.030254666574144752, + "grad_norm": 7.6314826011657715, + "learning_rate": 5.040462427745665e-05, + "loss": 1.382, + "step": 436 + }, + { + "epoch": 0.03032405801124141, + "grad_norm": 8.081480026245117, + "learning_rate": 5.052023121387284e-05, + "loss": 1.3422, + "step": 437 + }, + { + "epoch": 0.030393449448338074, + "grad_norm": 8.378664016723633, + "learning_rate": 5.063583815028902e-05, + "loss": 1.2388, + "step": 438 + }, + { + "epoch": 0.030462840885434737, + "grad_norm": 6.990529537200928, + "learning_rate": 5.0751445086705206e-05, + "loss": 1.1586, + "step": 439 + }, + { + "epoch": 0.0305322323225314, + "grad_norm": 8.561406135559082, + "learning_rate": 5.0867052023121385e-05, + "loss": 1.2937, + "step": 440 + }, + { + "epoch": 0.030601623759628063, + "grad_norm": 8.689889907836914, + "learning_rate": 5.098265895953758e-05, + "loss": 1.4866, + "step": 441 + }, + { + "epoch": 0.030671015196724723, + "grad_norm": 7.031777858734131, + "learning_rate": 5.109826589595376e-05, + "loss": 1.3245, + "step": 442 + }, + { + "epoch": 0.030740406633821386, + "grad_norm": 7.261953353881836, + "learning_rate": 5.1213872832369944e-05, + "loss": 1.3107, + "step": 443 + }, + { + "epoch": 0.03080979807091805, + "grad_norm": 7.250613212585449, + "learning_rate": 5.1329479768786124e-05, + "loss": 1.135, + "step": 444 + }, + { + "epoch": 0.03087918950801471, + "grad_norm": 6.978677272796631, + "learning_rate": 5.1445086705202317e-05, + "loss": 1.2775, + "step": 445 + }, + { + "epoch": 0.030948580945111374, + "grad_norm": 6.782436370849609, + "learning_rate": 5.1560693641618496e-05, + "loss": 1.0459, + "step": 446 + }, + { + "epoch": 0.031017972382208034, + "grad_norm": 6.918546199798584, + "learning_rate": 5.167630057803469e-05, + "loss": 0.973, + "step": 447 + }, + { + "epoch": 0.031087363819304697, + "grad_norm": 8.727575302124023, + "learning_rate": 5.179190751445087e-05, + "loss": 1.4979, + "step": 448 + }, + { + "epoch": 0.03115675525640136, + "grad_norm": 7.26622200012207, + "learning_rate": 5.1907514450867055e-05, + "loss": 1.3809, + "step": 449 + }, + { + "epoch": 0.031226146693498023, + "grad_norm": 6.667756080627441, + "learning_rate": 5.2023121387283234e-05, + "loss": 1.3766, + "step": 450 + }, + { + "epoch": 0.031295538130594686, + "grad_norm": 5.791069030761719, + "learning_rate": 5.213872832369943e-05, + "loss": 1.1495, + "step": 451 + }, + { + "epoch": 0.031364929567691345, + "grad_norm": 7.450632572174072, + "learning_rate": 5.2254335260115606e-05, + "loss": 1.5343, + "step": 452 + }, + { + "epoch": 0.03143432100478801, + "grad_norm": 6.587392807006836, + "learning_rate": 5.236994219653179e-05, + "loss": 1.2425, + "step": 453 + }, + { + "epoch": 0.03150371244188467, + "grad_norm": 8.119291305541992, + "learning_rate": 5.248554913294797e-05, + "loss": 1.5283, + "step": 454 + }, + { + "epoch": 0.03157310387898133, + "grad_norm": 6.48348331451416, + "learning_rate": 5.2601156069364165e-05, + "loss": 1.0778, + "step": 455 + }, + { + "epoch": 0.031642495316078, + "grad_norm": 7.047346591949463, + "learning_rate": 5.2716763005780345e-05, + "loss": 1.1142, + "step": 456 + }, + { + "epoch": 0.031711886753174656, + "grad_norm": 6.584650039672852, + "learning_rate": 5.283236994219654e-05, + "loss": 1.2548, + "step": 457 + }, + { + "epoch": 0.03178127819027132, + "grad_norm": 8.073479652404785, + "learning_rate": 5.294797687861272e-05, + "loss": 1.6026, + "step": 458 + }, + { + "epoch": 0.03185066962736798, + "grad_norm": 8.384836196899414, + "learning_rate": 5.30635838150289e-05, + "loss": 1.1022, + "step": 459 + }, + { + "epoch": 0.03192006106446465, + "grad_norm": 6.675930500030518, + "learning_rate": 5.317919075144508e-05, + "loss": 0.9579, + "step": 460 + }, + { + "epoch": 0.03198945250156131, + "grad_norm": 6.450620651245117, + "learning_rate": 5.3294797687861276e-05, + "loss": 1.0305, + "step": 461 + }, + { + "epoch": 0.03205884393865797, + "grad_norm": 7.937978267669678, + "learning_rate": 5.3410404624277455e-05, + "loss": 1.2351, + "step": 462 + }, + { + "epoch": 0.032128235375754634, + "grad_norm": 8.336834907531738, + "learning_rate": 5.352601156069365e-05, + "loss": 1.54, + "step": 463 + }, + { + "epoch": 0.03219762681285129, + "grad_norm": 5.875189304351807, + "learning_rate": 5.364161849710983e-05, + "loss": 1.1322, + "step": 464 + }, + { + "epoch": 0.03226701824994796, + "grad_norm": 7.092771053314209, + "learning_rate": 5.3757225433526014e-05, + "loss": 1.344, + "step": 465 + }, + { + "epoch": 0.03233640968704462, + "grad_norm": 7.129488468170166, + "learning_rate": 5.387283236994219e-05, + "loss": 1.6918, + "step": 466 + }, + { + "epoch": 0.03240580112414128, + "grad_norm": 5.659474849700928, + "learning_rate": 5.3988439306358386e-05, + "loss": 1.2367, + "step": 467 + }, + { + "epoch": 0.032475192561237945, + "grad_norm": 7.378965854644775, + "learning_rate": 5.4104046242774566e-05, + "loss": 1.2715, + "step": 468 + }, + { + "epoch": 0.032544583998334604, + "grad_norm": 6.137615203857422, + "learning_rate": 5.421965317919075e-05, + "loss": 1.1469, + "step": 469 + }, + { + "epoch": 0.03261397543543127, + "grad_norm": 7.655120849609375, + "learning_rate": 5.433526011560693e-05, + "loss": 1.1216, + "step": 470 + }, + { + "epoch": 0.03268336687252793, + "grad_norm": 7.611056804656982, + "learning_rate": 5.4450867052023124e-05, + "loss": 1.3248, + "step": 471 + }, + { + "epoch": 0.03275275830962459, + "grad_norm": 8.748945236206055, + "learning_rate": 5.4566473988439304e-05, + "loss": 1.4837, + "step": 472 + }, + { + "epoch": 0.032822149746721256, + "grad_norm": 6.202244758605957, + "learning_rate": 5.46820809248555e-05, + "loss": 1.0285, + "step": 473 + }, + { + "epoch": 0.032891541183817916, + "grad_norm": 7.103924751281738, + "learning_rate": 5.4797687861271676e-05, + "loss": 1.0544, + "step": 474 + }, + { + "epoch": 0.03296093262091458, + "grad_norm": 8.627496719360352, + "learning_rate": 5.491329479768786e-05, + "loss": 1.1143, + "step": 475 + }, + { + "epoch": 0.03303032405801124, + "grad_norm": 8.103438377380371, + "learning_rate": 5.502890173410404e-05, + "loss": 1.0424, + "step": 476 + }, + { + "epoch": 0.0330997154951079, + "grad_norm": 11.440688133239746, + "learning_rate": 5.5144508670520235e-05, + "loss": 0.9413, + "step": 477 + }, + { + "epoch": 0.03316910693220457, + "grad_norm": 8.538552284240723, + "learning_rate": 5.5260115606936414e-05, + "loss": 1.2453, + "step": 478 + }, + { + "epoch": 0.03323849836930123, + "grad_norm": 10.647181510925293, + "learning_rate": 5.537572254335261e-05, + "loss": 1.4858, + "step": 479 + }, + { + "epoch": 0.03330788980639789, + "grad_norm": 13.84502124786377, + "learning_rate": 5.5491329479768787e-05, + "loss": 1.4097, + "step": 480 + }, + { + "epoch": 0.03337728124349455, + "grad_norm": 9.436966896057129, + "learning_rate": 5.560693641618497e-05, + "loss": 1.1904, + "step": 481 + }, + { + "epoch": 0.03344667268059121, + "grad_norm": 6.597962856292725, + "learning_rate": 5.5722543352601166e-05, + "loss": 0.9751, + "step": 482 + }, + { + "epoch": 0.03351606411768788, + "grad_norm": 7.739300727844238, + "learning_rate": 5.5838150289017345e-05, + "loss": 1.3457, + "step": 483 + }, + { + "epoch": 0.03358545555478454, + "grad_norm": 8.624950408935547, + "learning_rate": 5.595375722543353e-05, + "loss": 1.1828, + "step": 484 + }, + { + "epoch": 0.033654846991881204, + "grad_norm": 7.7601728439331055, + "learning_rate": 5.606936416184971e-05, + "loss": 1.3072, + "step": 485 + }, + { + "epoch": 0.033724238428977864, + "grad_norm": 7.528622627258301, + "learning_rate": 5.6184971098265904e-05, + "loss": 1.2899, + "step": 486 + }, + { + "epoch": 0.03379362986607452, + "grad_norm": 6.689788818359375, + "learning_rate": 5.630057803468208e-05, + "loss": 1.1417, + "step": 487 + }, + { + "epoch": 0.03386302130317119, + "grad_norm": 5.563075542449951, + "learning_rate": 5.6416184971098276e-05, + "loss": 0.969, + "step": 488 + }, + { + "epoch": 0.03393241274026785, + "grad_norm": 12.955428123474121, + "learning_rate": 5.6531791907514456e-05, + "loss": 1.4175, + "step": 489 + }, + { + "epoch": 0.034001804177364515, + "grad_norm": 9.051423072814941, + "learning_rate": 5.664739884393064e-05, + "loss": 1.4247, + "step": 490 + }, + { + "epoch": 0.034071195614461175, + "grad_norm": 8.81102466583252, + "learning_rate": 5.676300578034682e-05, + "loss": 1.3312, + "step": 491 + }, + { + "epoch": 0.034140587051557834, + "grad_norm": 7.270226955413818, + "learning_rate": 5.6878612716763014e-05, + "loss": 1.1807, + "step": 492 + }, + { + "epoch": 0.0342099784886545, + "grad_norm": 9.581392288208008, + "learning_rate": 5.6994219653179194e-05, + "loss": 1.2338, + "step": 493 + }, + { + "epoch": 0.03427936992575116, + "grad_norm": 6.770020008087158, + "learning_rate": 5.710982658959538e-05, + "loss": 1.1437, + "step": 494 + }, + { + "epoch": 0.03434876136284783, + "grad_norm": 7.853590965270996, + "learning_rate": 5.722543352601156e-05, + "loss": 1.5171, + "step": 495 + }, + { + "epoch": 0.034418152799944486, + "grad_norm": 7.39157772064209, + "learning_rate": 5.734104046242775e-05, + "loss": 1.4823, + "step": 496 + }, + { + "epoch": 0.03448754423704115, + "grad_norm": 9.550525665283203, + "learning_rate": 5.745664739884393e-05, + "loss": 1.09, + "step": 497 + }, + { + "epoch": 0.03455693567413781, + "grad_norm": 6.795161247253418, + "learning_rate": 5.7572254335260125e-05, + "loss": 1.1201, + "step": 498 + }, + { + "epoch": 0.03462632711123447, + "grad_norm": 13.278371810913086, + "learning_rate": 5.7687861271676304e-05, + "loss": 1.1262, + "step": 499 + }, + { + "epoch": 0.03469571854833114, + "grad_norm": 6.571694850921631, + "learning_rate": 5.780346820809249e-05, + "loss": 1.2631, + "step": 500 + }, + { + "epoch": 0.0347651099854278, + "grad_norm": 9.4550142288208, + "learning_rate": 5.791907514450867e-05, + "loss": 1.285, + "step": 501 + }, + { + "epoch": 0.034834501422524464, + "grad_norm": 10.346406936645508, + "learning_rate": 5.803468208092486e-05, + "loss": 1.444, + "step": 502 + }, + { + "epoch": 0.03490389285962112, + "grad_norm": 12.434240341186523, + "learning_rate": 5.815028901734104e-05, + "loss": 1.4609, + "step": 503 + }, + { + "epoch": 0.03497328429671778, + "grad_norm": 6.481371879577637, + "learning_rate": 5.8265895953757235e-05, + "loss": 1.1135, + "step": 504 + }, + { + "epoch": 0.03504267573381445, + "grad_norm": 7.1974406242370605, + "learning_rate": 5.8381502890173415e-05, + "loss": 1.3156, + "step": 505 + }, + { + "epoch": 0.03511206717091111, + "grad_norm": 14.45538330078125, + "learning_rate": 5.84971098265896e-05, + "loss": 1.2123, + "step": 506 + }, + { + "epoch": 0.035181458608007775, + "grad_norm": 7.230411052703857, + "learning_rate": 5.861271676300578e-05, + "loss": 1.3912, + "step": 507 + }, + { + "epoch": 0.035250850045104434, + "grad_norm": 7.975244522094727, + "learning_rate": 5.8728323699421973e-05, + "loss": 1.3261, + "step": 508 + }, + { + "epoch": 0.035320241482201094, + "grad_norm": 6.677006721496582, + "learning_rate": 5.884393063583815e-05, + "loss": 1.1558, + "step": 509 + }, + { + "epoch": 0.03538963291929776, + "grad_norm": 6.691371917724609, + "learning_rate": 5.895953757225434e-05, + "loss": 1.4777, + "step": 510 + }, + { + "epoch": 0.03545902435639442, + "grad_norm": 5.63232946395874, + "learning_rate": 5.907514450867052e-05, + "loss": 0.9707, + "step": 511 + }, + { + "epoch": 0.035528415793491086, + "grad_norm": 7.7168097496032715, + "learning_rate": 5.919075144508671e-05, + "loss": 1.1531, + "step": 512 + }, + { + "epoch": 0.035597807230587745, + "grad_norm": 6.73607063293457, + "learning_rate": 5.930635838150289e-05, + "loss": 0.9996, + "step": 513 + }, + { + "epoch": 0.035667198667684405, + "grad_norm": 7.968959808349609, + "learning_rate": 5.9421965317919084e-05, + "loss": 1.4952, + "step": 514 + }, + { + "epoch": 0.03573659010478107, + "grad_norm": 7.523364067077637, + "learning_rate": 5.9537572254335263e-05, + "loss": 1.1289, + "step": 515 + }, + { + "epoch": 0.03580598154187773, + "grad_norm": 7.520701885223389, + "learning_rate": 5.965317919075145e-05, + "loss": 1.0813, + "step": 516 + }, + { + "epoch": 0.0358753729789744, + "grad_norm": 11.999600410461426, + "learning_rate": 5.976878612716763e-05, + "loss": 0.8826, + "step": 517 + }, + { + "epoch": 0.03594476441607106, + "grad_norm": 6.602665424346924, + "learning_rate": 5.988439306358382e-05, + "loss": 1.5214, + "step": 518 + }, + { + "epoch": 0.036014155853167716, + "grad_norm": 11.963011741638184, + "learning_rate": 6e-05, + "loss": 1.1375, + "step": 519 + }, + { + "epoch": 0.03608354729026438, + "grad_norm": 7.499159812927246, + "learning_rate": 6.0115606936416195e-05, + "loss": 1.1736, + "step": 520 + }, + { + "epoch": 0.03615293872736104, + "grad_norm": 7.962111949920654, + "learning_rate": 6.023121387283237e-05, + "loss": 1.101, + "step": 521 + }, + { + "epoch": 0.03622233016445771, + "grad_norm": 6.992148399353027, + "learning_rate": 6.034682080924856e-05, + "loss": 1.0267, + "step": 522 + }, + { + "epoch": 0.03629172160155437, + "grad_norm": 8.088050842285156, + "learning_rate": 6.046242774566474e-05, + "loss": 1.2949, + "step": 523 + }, + { + "epoch": 0.03636111303865103, + "grad_norm": 7.081723690032959, + "learning_rate": 6.057803468208093e-05, + "loss": 1.0988, + "step": 524 + }, + { + "epoch": 0.036430504475747694, + "grad_norm": 7.5683207511901855, + "learning_rate": 6.069364161849711e-05, + "loss": 1.0169, + "step": 525 + }, + { + "epoch": 0.03649989591284435, + "grad_norm": 9.369019508361816, + "learning_rate": 6.08092485549133e-05, + "loss": 1.2287, + "step": 526 + }, + { + "epoch": 0.03656928734994102, + "grad_norm": 8.55128288269043, + "learning_rate": 6.092485549132948e-05, + "loss": 1.2261, + "step": 527 + }, + { + "epoch": 0.03663867878703768, + "grad_norm": 7.649362564086914, + "learning_rate": 6.104046242774567e-05, + "loss": 1.2793, + "step": 528 + }, + { + "epoch": 0.03670807022413434, + "grad_norm": 9.0997896194458, + "learning_rate": 6.115606936416184e-05, + "loss": 1.2066, + "step": 529 + }, + { + "epoch": 0.036777461661231005, + "grad_norm": 8.913084983825684, + "learning_rate": 6.127167630057804e-05, + "loss": 1.2641, + "step": 530 + }, + { + "epoch": 0.036846853098327664, + "grad_norm": 7.123581409454346, + "learning_rate": 6.138728323699422e-05, + "loss": 0.9324, + "step": 531 + }, + { + "epoch": 0.03691624453542433, + "grad_norm": 6.753522872924805, + "learning_rate": 6.150289017341042e-05, + "loss": 1.2039, + "step": 532 + }, + { + "epoch": 0.03698563597252099, + "grad_norm": 9.21194839477539, + "learning_rate": 6.161849710982659e-05, + "loss": 0.9455, + "step": 533 + }, + { + "epoch": 0.03705502740961766, + "grad_norm": 5.679974555969238, + "learning_rate": 6.173410404624277e-05, + "loss": 1.0404, + "step": 534 + }, + { + "epoch": 0.037124418846714316, + "grad_norm": 7.350649356842041, + "learning_rate": 6.184971098265896e-05, + "loss": 1.1588, + "step": 535 + }, + { + "epoch": 0.037193810283810975, + "grad_norm": 7.175858974456787, + "learning_rate": 6.196531791907515e-05, + "loss": 1.1113, + "step": 536 + }, + { + "epoch": 0.03726320172090764, + "grad_norm": 9.372313499450684, + "learning_rate": 6.208092485549133e-05, + "loss": 1.4575, + "step": 537 + }, + { + "epoch": 0.0373325931580043, + "grad_norm": 7.593583106994629, + "learning_rate": 6.219653179190752e-05, + "loss": 1.2951, + "step": 538 + }, + { + "epoch": 0.03740198459510097, + "grad_norm": 6.6288886070251465, + "learning_rate": 6.23121387283237e-05, + "loss": 1.3643, + "step": 539 + }, + { + "epoch": 0.03747137603219763, + "grad_norm": 12.328807830810547, + "learning_rate": 6.242774566473989e-05, + "loss": 1.2979, + "step": 540 + }, + { + "epoch": 0.03754076746929429, + "grad_norm": 8.3024320602417, + "learning_rate": 6.254335260115606e-05, + "loss": 1.1729, + "step": 541 + }, + { + "epoch": 0.03761015890639095, + "grad_norm": 6.182681560516357, + "learning_rate": 6.265895953757226e-05, + "loss": 1.1388, + "step": 542 + }, + { + "epoch": 0.03767955034348761, + "grad_norm": 7.6043572425842285, + "learning_rate": 6.277456647398844e-05, + "loss": 1.1901, + "step": 543 + }, + { + "epoch": 0.03774894178058428, + "grad_norm": 10.835043907165527, + "learning_rate": 6.289017341040462e-05, + "loss": 1.6231, + "step": 544 + }, + { + "epoch": 0.03781833321768094, + "grad_norm": 6.410502910614014, + "learning_rate": 6.300578034682081e-05, + "loss": 1.0764, + "step": 545 + }, + { + "epoch": 0.0378877246547776, + "grad_norm": 7.3326497077941895, + "learning_rate": 6.3121387283237e-05, + "loss": 0.9653, + "step": 546 + }, + { + "epoch": 0.037957116091874264, + "grad_norm": 6.890195369720459, + "learning_rate": 6.323699421965318e-05, + "loss": 1.0489, + "step": 547 + }, + { + "epoch": 0.038026507528970924, + "grad_norm": 7.009410381317139, + "learning_rate": 6.335260115606937e-05, + "loss": 1.0647, + "step": 548 + }, + { + "epoch": 0.03809589896606759, + "grad_norm": 5.996044158935547, + "learning_rate": 6.346820809248555e-05, + "loss": 1.0976, + "step": 549 + }, + { + "epoch": 0.03816529040316425, + "grad_norm": 7.704802513122559, + "learning_rate": 6.358381502890174e-05, + "loss": 1.0053, + "step": 550 + }, + { + "epoch": 0.03823468184026091, + "grad_norm": 6.278261184692383, + "learning_rate": 6.369942196531791e-05, + "loss": 1.3456, + "step": 551 + }, + { + "epoch": 0.038304073277357575, + "grad_norm": 7.341904640197754, + "learning_rate": 6.381502890173411e-05, + "loss": 1.5124, + "step": 552 + }, + { + "epoch": 0.038373464714454235, + "grad_norm": 7.042769908905029, + "learning_rate": 6.393063583815029e-05, + "loss": 1.3938, + "step": 553 + }, + { + "epoch": 0.0384428561515509, + "grad_norm": 6.952353477478027, + "learning_rate": 6.404624277456649e-05, + "loss": 1.2296, + "step": 554 + }, + { + "epoch": 0.03851224758864756, + "grad_norm": 6.50119161605835, + "learning_rate": 6.416184971098266e-05, + "loss": 1.192, + "step": 555 + }, + { + "epoch": 0.03858163902574422, + "grad_norm": 8.832952499389648, + "learning_rate": 6.427745664739884e-05, + "loss": 1.1796, + "step": 556 + }, + { + "epoch": 0.03865103046284089, + "grad_norm": 9.319852828979492, + "learning_rate": 6.439306358381503e-05, + "loss": 1.2966, + "step": 557 + }, + { + "epoch": 0.038720421899937546, + "grad_norm": 7.486120700836182, + "learning_rate": 6.450867052023122e-05, + "loss": 1.0758, + "step": 558 + }, + { + "epoch": 0.03878981333703421, + "grad_norm": 8.900551795959473, + "learning_rate": 6.46242774566474e-05, + "loss": 1.7377, + "step": 559 + }, + { + "epoch": 0.03885920477413087, + "grad_norm": 6.740694522857666, + "learning_rate": 6.473988439306359e-05, + "loss": 1.2397, + "step": 560 + }, + { + "epoch": 0.03892859621122753, + "grad_norm": 9.365303993225098, + "learning_rate": 6.485549132947976e-05, + "loss": 1.4384, + "step": 561 + }, + { + "epoch": 0.0389979876483242, + "grad_norm": 7.64422607421875, + "learning_rate": 6.497109826589596e-05, + "loss": 1.0375, + "step": 562 + }, + { + "epoch": 0.03906737908542086, + "grad_norm": 7.384373664855957, + "learning_rate": 6.508670520231213e-05, + "loss": 1.2452, + "step": 563 + }, + { + "epoch": 0.039136770522517524, + "grad_norm": 7.0529465675354, + "learning_rate": 6.520231213872833e-05, + "loss": 1.0258, + "step": 564 + }, + { + "epoch": 0.03920616195961418, + "grad_norm": 6.8748064041137695, + "learning_rate": 6.53179190751445e-05, + "loss": 1.201, + "step": 565 + }, + { + "epoch": 0.03927555339671084, + "grad_norm": 7.9478840827941895, + "learning_rate": 6.543352601156069e-05, + "loss": 1.2403, + "step": 566 + }, + { + "epoch": 0.03934494483380751, + "grad_norm": 7.467204570770264, + "learning_rate": 6.554913294797688e-05, + "loss": 1.305, + "step": 567 + }, + { + "epoch": 0.03941433627090417, + "grad_norm": 7.371156692504883, + "learning_rate": 6.566473988439307e-05, + "loss": 1.1762, + "step": 568 + }, + { + "epoch": 0.039483727708000835, + "grad_norm": 8.263506889343262, + "learning_rate": 6.578034682080925e-05, + "loss": 1.3115, + "step": 569 + }, + { + "epoch": 0.039553119145097494, + "grad_norm": 5.943240165710449, + "learning_rate": 6.589595375722544e-05, + "loss": 1.0482, + "step": 570 + }, + { + "epoch": 0.03962251058219416, + "grad_norm": 7.655527591705322, + "learning_rate": 6.601156069364161e-05, + "loss": 1.3146, + "step": 571 + }, + { + "epoch": 0.03969190201929082, + "grad_norm": 6.683849334716797, + "learning_rate": 6.612716763005781e-05, + "loss": 1.168, + "step": 572 + }, + { + "epoch": 0.03976129345638748, + "grad_norm": 7.48051643371582, + "learning_rate": 6.624277456647398e-05, + "loss": 1.205, + "step": 573 + }, + { + "epoch": 0.039830684893484146, + "grad_norm": 8.229850769042969, + "learning_rate": 6.635838150289018e-05, + "loss": 1.322, + "step": 574 + }, + { + "epoch": 0.039900076330580805, + "grad_norm": 8.585726737976074, + "learning_rate": 6.647398843930635e-05, + "loss": 1.3498, + "step": 575 + }, + { + "epoch": 0.03996946776767747, + "grad_norm": 7.948550224304199, + "learning_rate": 6.658959537572254e-05, + "loss": 1.1093, + "step": 576 + }, + { + "epoch": 0.04003885920477413, + "grad_norm": 7.889479637145996, + "learning_rate": 6.670520231213874e-05, + "loss": 1.371, + "step": 577 + }, + { + "epoch": 0.04010825064187079, + "grad_norm": 6.470623970031738, + "learning_rate": 6.682080924855491e-05, + "loss": 1.318, + "step": 578 + }, + { + "epoch": 0.04017764207896746, + "grad_norm": 6.131725788116455, + "learning_rate": 6.693641618497111e-05, + "loss": 0.9352, + "step": 579 + }, + { + "epoch": 0.04024703351606412, + "grad_norm": 6.3577985763549805, + "learning_rate": 6.705202312138729e-05, + "loss": 1.1046, + "step": 580 + }, + { + "epoch": 0.04031642495316078, + "grad_norm": 5.772814750671387, + "learning_rate": 6.716763005780347e-05, + "loss": 1.0667, + "step": 581 + }, + { + "epoch": 0.04038581639025744, + "grad_norm": 5.596307754516602, + "learning_rate": 6.728323699421966e-05, + "loss": 0.9177, + "step": 582 + }, + { + "epoch": 0.0404552078273541, + "grad_norm": 9.894975662231445, + "learning_rate": 6.739884393063584e-05, + "loss": 1.2731, + "step": 583 + }, + { + "epoch": 0.04052459926445077, + "grad_norm": 8.596860885620117, + "learning_rate": 6.751445086705203e-05, + "loss": 1.5144, + "step": 584 + }, + { + "epoch": 0.04059399070154743, + "grad_norm": 6.214570999145508, + "learning_rate": 6.763005780346822e-05, + "loss": 1.1868, + "step": 585 + }, + { + "epoch": 0.040663382138644094, + "grad_norm": 6.6037516593933105, + "learning_rate": 6.774566473988439e-05, + "loss": 1.3651, + "step": 586 + }, + { + "epoch": 0.040732773575740754, + "grad_norm": 6.401592254638672, + "learning_rate": 6.786127167630059e-05, + "loss": 0.9604, + "step": 587 + }, + { + "epoch": 0.04080216501283741, + "grad_norm": 5.830338001251221, + "learning_rate": 6.797687861271676e-05, + "loss": 1.1424, + "step": 588 + }, + { + "epoch": 0.04087155644993408, + "grad_norm": 5.8388800621032715, + "learning_rate": 6.809248554913296e-05, + "loss": 0.9749, + "step": 589 + }, + { + "epoch": 0.04094094788703074, + "grad_norm": 9.343720436096191, + "learning_rate": 6.820809248554913e-05, + "loss": 1.6504, + "step": 590 + }, + { + "epoch": 0.041010339324127405, + "grad_norm": 9.195332527160645, + "learning_rate": 6.832369942196532e-05, + "loss": 0.946, + "step": 591 + }, + { + "epoch": 0.041079730761224065, + "grad_norm": 7.665297508239746, + "learning_rate": 6.843930635838151e-05, + "loss": 1.3583, + "step": 592 + }, + { + "epoch": 0.041149122198320724, + "grad_norm": 7.807668209075928, + "learning_rate": 6.85549132947977e-05, + "loss": 1.3627, + "step": 593 + }, + { + "epoch": 0.04121851363541739, + "grad_norm": 7.689777851104736, + "learning_rate": 6.867052023121388e-05, + "loss": 1.3568, + "step": 594 + }, + { + "epoch": 0.04128790507251405, + "grad_norm": 7.011950492858887, + "learning_rate": 6.878612716763007e-05, + "loss": 0.931, + "step": 595 + }, + { + "epoch": 0.041357296509610716, + "grad_norm": 11.051543235778809, + "learning_rate": 6.890173410404624e-05, + "loss": 1.1866, + "step": 596 + }, + { + "epoch": 0.041426687946707376, + "grad_norm": 7.111940383911133, + "learning_rate": 6.901734104046244e-05, + "loss": 1.0091, + "step": 597 + }, + { + "epoch": 0.041496079383804035, + "grad_norm": 6.54121208190918, + "learning_rate": 6.913294797687861e-05, + "loss": 1.2498, + "step": 598 + }, + { + "epoch": 0.0415654708209007, + "grad_norm": 5.517703056335449, + "learning_rate": 6.924855491329481e-05, + "loss": 1.0096, + "step": 599 + }, + { + "epoch": 0.04163486225799736, + "grad_norm": 5.913308143615723, + "learning_rate": 6.936416184971098e-05, + "loss": 1.0315, + "step": 600 + }, + { + "epoch": 0.04170425369509403, + "grad_norm": 10.083365440368652, + "learning_rate": 6.947976878612717e-05, + "loss": 1.4897, + "step": 601 + }, + { + "epoch": 0.04177364513219069, + "grad_norm": 7.0428032875061035, + "learning_rate": 6.959537572254336e-05, + "loss": 1.5086, + "step": 602 + }, + { + "epoch": 0.04184303656928735, + "grad_norm": 5.885694980621338, + "learning_rate": 6.971098265895954e-05, + "loss": 1.1298, + "step": 603 + }, + { + "epoch": 0.04191242800638401, + "grad_norm": 5.060995101928711, + "learning_rate": 6.982658959537573e-05, + "loss": 1.0707, + "step": 604 + }, + { + "epoch": 0.04198181944348067, + "grad_norm": 6.754761219024658, + "learning_rate": 6.994219653179191e-05, + "loss": 1.6139, + "step": 605 + }, + { + "epoch": 0.04205121088057734, + "grad_norm": 6.220276832580566, + "learning_rate": 7.00578034682081e-05, + "loss": 0.968, + "step": 606 + }, + { + "epoch": 0.042120602317674, + "grad_norm": 6.722576141357422, + "learning_rate": 7.017341040462429e-05, + "loss": 1.0882, + "step": 607 + }, + { + "epoch": 0.042189993754770665, + "grad_norm": 8.653365135192871, + "learning_rate": 7.028901734104046e-05, + "loss": 1.1089, + "step": 608 + }, + { + "epoch": 0.042259385191867324, + "grad_norm": 11.922430992126465, + "learning_rate": 7.040462427745666e-05, + "loss": 1.1799, + "step": 609 + }, + { + "epoch": 0.042328776628963984, + "grad_norm": 8.366812705993652, + "learning_rate": 7.052023121387283e-05, + "loss": 1.413, + "step": 610 + }, + { + "epoch": 0.04239816806606065, + "grad_norm": 6.361443996429443, + "learning_rate": 7.063583815028902e-05, + "loss": 1.2061, + "step": 611 + }, + { + "epoch": 0.04246755950315731, + "grad_norm": 6.83425760269165, + "learning_rate": 7.07514450867052e-05, + "loss": 1.0762, + "step": 612 + }, + { + "epoch": 0.042536950940253976, + "grad_norm": 7.7701334953308105, + "learning_rate": 7.086705202312139e-05, + "loss": 1.2414, + "step": 613 + }, + { + "epoch": 0.042606342377350635, + "grad_norm": 7.617910385131836, + "learning_rate": 7.098265895953758e-05, + "loss": 1.3569, + "step": 614 + }, + { + "epoch": 0.042675733814447295, + "grad_norm": 7.494248390197754, + "learning_rate": 7.109826589595376e-05, + "loss": 1.3533, + "step": 615 + }, + { + "epoch": 0.04274512525154396, + "grad_norm": 12.272274017333984, + "learning_rate": 7.121387283236995e-05, + "loss": 1.3282, + "step": 616 + }, + { + "epoch": 0.04281451668864062, + "grad_norm": 6.362443447113037, + "learning_rate": 7.132947976878614e-05, + "loss": 1.1332, + "step": 617 + }, + { + "epoch": 0.04288390812573729, + "grad_norm": 6.031251430511475, + "learning_rate": 7.144508670520231e-05, + "loss": 1.1831, + "step": 618 + }, + { + "epoch": 0.042953299562833946, + "grad_norm": 6.286535263061523, + "learning_rate": 7.156069364161851e-05, + "loss": 1.3812, + "step": 619 + }, + { + "epoch": 0.043022690999930606, + "grad_norm": 6.320926189422607, + "learning_rate": 7.167630057803468e-05, + "loss": 1.1595, + "step": 620 + }, + { + "epoch": 0.04309208243702727, + "grad_norm": 7.262394905090332, + "learning_rate": 7.179190751445088e-05, + "loss": 1.3419, + "step": 621 + }, + { + "epoch": 0.04316147387412393, + "grad_norm": 4.9367570877075195, + "learning_rate": 7.190751445086705e-05, + "loss": 0.8973, + "step": 622 + }, + { + "epoch": 0.0432308653112206, + "grad_norm": 7.778924942016602, + "learning_rate": 7.202312138728324e-05, + "loss": 1.3827, + "step": 623 + }, + { + "epoch": 0.04330025674831726, + "grad_norm": 6.198282718658447, + "learning_rate": 7.213872832369943e-05, + "loss": 1.0845, + "step": 624 + }, + { + "epoch": 0.04336964818541392, + "grad_norm": 6.227132797241211, + "learning_rate": 7.225433526011561e-05, + "loss": 1.0851, + "step": 625 + }, + { + "epoch": 0.043439039622510583, + "grad_norm": 6.430438041687012, + "learning_rate": 7.23699421965318e-05, + "loss": 1.2142, + "step": 626 + }, + { + "epoch": 0.04350843105960724, + "grad_norm": 6.026603698730469, + "learning_rate": 7.248554913294798e-05, + "loss": 1.1746, + "step": 627 + }, + { + "epoch": 0.04357782249670391, + "grad_norm": 9.12875747680664, + "learning_rate": 7.260115606936416e-05, + "loss": 1.3835, + "step": 628 + }, + { + "epoch": 0.04364721393380057, + "grad_norm": 6.904024600982666, + "learning_rate": 7.271676300578036e-05, + "loss": 1.2184, + "step": 629 + }, + { + "epoch": 0.04371660537089723, + "grad_norm": 6.912041664123535, + "learning_rate": 7.283236994219653e-05, + "loss": 1.0196, + "step": 630 + }, + { + "epoch": 0.043785996807993895, + "grad_norm": 7.439908504486084, + "learning_rate": 7.294797687861273e-05, + "loss": 1.0714, + "step": 631 + }, + { + "epoch": 0.043855388245090554, + "grad_norm": 7.971566200256348, + "learning_rate": 7.30635838150289e-05, + "loss": 1.348, + "step": 632 + }, + { + "epoch": 0.04392477968218722, + "grad_norm": 5.541843414306641, + "learning_rate": 7.317919075144509e-05, + "loss": 0.8752, + "step": 633 + }, + { + "epoch": 0.04399417111928388, + "grad_norm": 6.461187362670898, + "learning_rate": 7.329479768786127e-05, + "loss": 1.2082, + "step": 634 + }, + { + "epoch": 0.04406356255638054, + "grad_norm": 7.966387748718262, + "learning_rate": 7.341040462427746e-05, + "loss": 1.2481, + "step": 635 + }, + { + "epoch": 0.044132953993477206, + "grad_norm": 5.839384078979492, + "learning_rate": 7.352601156069365e-05, + "loss": 1.2438, + "step": 636 + }, + { + "epoch": 0.044202345430573865, + "grad_norm": 7.331144332885742, + "learning_rate": 7.364161849710983e-05, + "loss": 1.4232, + "step": 637 + }, + { + "epoch": 0.04427173686767053, + "grad_norm": 7.3672637939453125, + "learning_rate": 7.3757225433526e-05, + "loss": 1.2181, + "step": 638 + }, + { + "epoch": 0.04434112830476719, + "grad_norm": 6.3997721672058105, + "learning_rate": 7.38728323699422e-05, + "loss": 1.3001, + "step": 639 + }, + { + "epoch": 0.04441051974186385, + "grad_norm": 6.646839141845703, + "learning_rate": 7.398843930635838e-05, + "loss": 1.1821, + "step": 640 + }, + { + "epoch": 0.04447991117896052, + "grad_norm": 7.495395183563232, + "learning_rate": 7.410404624277458e-05, + "loss": 1.0278, + "step": 641 + }, + { + "epoch": 0.044549302616057176, + "grad_norm": 8.46640682220459, + "learning_rate": 7.421965317919075e-05, + "loss": 1.1377, + "step": 642 + }, + { + "epoch": 0.04461869405315384, + "grad_norm": 6.172717094421387, + "learning_rate": 7.433526011560694e-05, + "loss": 1.0292, + "step": 643 + }, + { + "epoch": 0.0446880854902505, + "grad_norm": 6.957774639129639, + "learning_rate": 7.445086705202312e-05, + "loss": 1.5544, + "step": 644 + }, + { + "epoch": 0.04475747692734717, + "grad_norm": 8.123444557189941, + "learning_rate": 7.456647398843931e-05, + "loss": 0.8583, + "step": 645 + }, + { + "epoch": 0.04482686836444383, + "grad_norm": 7.511685848236084, + "learning_rate": 7.46820809248555e-05, + "loss": 1.2373, + "step": 646 + }, + { + "epoch": 0.04489625980154049, + "grad_norm": 7.554117679595947, + "learning_rate": 7.479768786127168e-05, + "loss": 1.3675, + "step": 647 + }, + { + "epoch": 0.044965651238637154, + "grad_norm": 6.9201812744140625, + "learning_rate": 7.491329479768787e-05, + "loss": 1.2595, + "step": 648 + }, + { + "epoch": 0.045035042675733813, + "grad_norm": 16.531089782714844, + "learning_rate": 7.502890173410405e-05, + "loss": 1.2777, + "step": 649 + }, + { + "epoch": 0.04510443411283048, + "grad_norm": 8.26958179473877, + "learning_rate": 7.514450867052023e-05, + "loss": 1.3637, + "step": 650 + }, + { + "epoch": 0.04517382554992714, + "grad_norm": 5.79640531539917, + "learning_rate": 7.526011560693643e-05, + "loss": 1.0895, + "step": 651 + }, + { + "epoch": 0.0452432169870238, + "grad_norm": 6.312289237976074, + "learning_rate": 7.53757225433526e-05, + "loss": 1.1155, + "step": 652 + }, + { + "epoch": 0.045312608424120465, + "grad_norm": 15.082344055175781, + "learning_rate": 7.549132947976878e-05, + "loss": 1.3, + "step": 653 + }, + { + "epoch": 0.045381999861217125, + "grad_norm": 8.063645362854004, + "learning_rate": 7.560693641618497e-05, + "loss": 1.1264, + "step": 654 + }, + { + "epoch": 0.04545139129831379, + "grad_norm": 10.499456405639648, + "learning_rate": 7.572254335260116e-05, + "loss": 1.6645, + "step": 655 + }, + { + "epoch": 0.04552078273541045, + "grad_norm": 5.668798923492432, + "learning_rate": 7.583815028901734e-05, + "loss": 1.226, + "step": 656 + }, + { + "epoch": 0.04559017417250711, + "grad_norm": 6.77095890045166, + "learning_rate": 7.595375722543353e-05, + "loss": 1.3134, + "step": 657 + }, + { + "epoch": 0.045659565609603776, + "grad_norm": 6.754121780395508, + "learning_rate": 7.606936416184972e-05, + "loss": 1.0695, + "step": 658 + }, + { + "epoch": 0.045728957046700436, + "grad_norm": 6.236030101776123, + "learning_rate": 7.61849710982659e-05, + "loss": 0.9423, + "step": 659 + }, + { + "epoch": 0.0457983484837971, + "grad_norm": 7.2759833335876465, + "learning_rate": 7.630057803468207e-05, + "loss": 1.0095, + "step": 660 + }, + { + "epoch": 0.04586773992089376, + "grad_norm": 10.282370567321777, + "learning_rate": 7.641618497109827e-05, + "loss": 1.2352, + "step": 661 + }, + { + "epoch": 0.04593713135799042, + "grad_norm": 5.669065475463867, + "learning_rate": 7.653179190751445e-05, + "loss": 1.2222, + "step": 662 + }, + { + "epoch": 0.04600652279508709, + "grad_norm": 8.642940521240234, + "learning_rate": 7.664739884393063e-05, + "loss": 1.2417, + "step": 663 + }, + { + "epoch": 0.04607591423218375, + "grad_norm": 8.508056640625, + "learning_rate": 7.676300578034682e-05, + "loss": 1.1894, + "step": 664 + }, + { + "epoch": 0.04614530566928041, + "grad_norm": 7.039309024810791, + "learning_rate": 7.6878612716763e-05, + "loss": 1.0211, + "step": 665 + }, + { + "epoch": 0.04621469710637707, + "grad_norm": 6.098023414611816, + "learning_rate": 7.699421965317919e-05, + "loss": 1.2617, + "step": 666 + }, + { + "epoch": 0.04628408854347373, + "grad_norm": 6.591110706329346, + "learning_rate": 7.710982658959538e-05, + "loss": 1.2603, + "step": 667 + }, + { + "epoch": 0.0463534799805704, + "grad_norm": 6.48390007019043, + "learning_rate": 7.722543352601156e-05, + "loss": 1.1764, + "step": 668 + }, + { + "epoch": 0.04642287141766706, + "grad_norm": 5.684776306152344, + "learning_rate": 7.734104046242775e-05, + "loss": 1.1038, + "step": 669 + }, + { + "epoch": 0.046492262854763725, + "grad_norm": 6.152649879455566, + "learning_rate": 7.745664739884392e-05, + "loss": 1.0634, + "step": 670 + }, + { + "epoch": 0.046561654291860384, + "grad_norm": 7.924769878387451, + "learning_rate": 7.757225433526012e-05, + "loss": 1.5222, + "step": 671 + }, + { + "epoch": 0.046631045728957043, + "grad_norm": 8.07640552520752, + "learning_rate": 7.76878612716763e-05, + "loss": 1.4017, + "step": 672 + }, + { + "epoch": 0.04670043716605371, + "grad_norm": 6.311427116394043, + "learning_rate": 7.78034682080925e-05, + "loss": 1.1312, + "step": 673 + }, + { + "epoch": 0.04676982860315037, + "grad_norm": 6.716381072998047, + "learning_rate": 7.791907514450868e-05, + "loss": 1.161, + "step": 674 + }, + { + "epoch": 0.046839220040247036, + "grad_norm": 8.13567066192627, + "learning_rate": 7.803468208092485e-05, + "loss": 1.3678, + "step": 675 + }, + { + "epoch": 0.046908611477343695, + "grad_norm": 6.003815174102783, + "learning_rate": 7.815028901734105e-05, + "loss": 1.032, + "step": 676 + }, + { + "epoch": 0.046978002914440355, + "grad_norm": 8.093462944030762, + "learning_rate": 7.826589595375723e-05, + "loss": 1.1221, + "step": 677 + }, + { + "epoch": 0.04704739435153702, + "grad_norm": 6.115405082702637, + "learning_rate": 7.838150289017341e-05, + "loss": 1.3283, + "step": 678 + }, + { + "epoch": 0.04711678578863368, + "grad_norm": 5.690483570098877, + "learning_rate": 7.84971098265896e-05, + "loss": 1.0729, + "step": 679 + }, + { + "epoch": 0.04718617722573035, + "grad_norm": 8.314352989196777, + "learning_rate": 7.861271676300579e-05, + "loss": 1.1857, + "step": 680 + }, + { + "epoch": 0.047255568662827006, + "grad_norm": 6.085377216339111, + "learning_rate": 7.872832369942197e-05, + "loss": 1.2221, + "step": 681 + }, + { + "epoch": 0.04732496009992367, + "grad_norm": 6.710886001586914, + "learning_rate": 7.884393063583816e-05, + "loss": 1.1637, + "step": 682 + }, + { + "epoch": 0.04739435153702033, + "grad_norm": 6.668133735656738, + "learning_rate": 7.895953757225434e-05, + "loss": 1.1794, + "step": 683 + }, + { + "epoch": 0.04746374297411699, + "grad_norm": 6.651283264160156, + "learning_rate": 7.907514450867053e-05, + "loss": 1.3379, + "step": 684 + }, + { + "epoch": 0.04753313441121366, + "grad_norm": 6.023229598999023, + "learning_rate": 7.91907514450867e-05, + "loss": 1.2157, + "step": 685 + }, + { + "epoch": 0.04760252584831032, + "grad_norm": 6.664051532745361, + "learning_rate": 7.93063583815029e-05, + "loss": 1.1472, + "step": 686 + }, + { + "epoch": 0.047671917285406984, + "grad_norm": 6.820763111114502, + "learning_rate": 7.942196531791908e-05, + "loss": 1.174, + "step": 687 + }, + { + "epoch": 0.04774130872250364, + "grad_norm": 6.888080596923828, + "learning_rate": 7.953757225433528e-05, + "loss": 0.9835, + "step": 688 + }, + { + "epoch": 0.0478107001596003, + "grad_norm": 6.291824817657471, + "learning_rate": 7.965317919075145e-05, + "loss": 1.1095, + "step": 689 + }, + { + "epoch": 0.04788009159669697, + "grad_norm": 6.278719902038574, + "learning_rate": 7.976878612716763e-05, + "loss": 1.2197, + "step": 690 + }, + { + "epoch": 0.04794948303379363, + "grad_norm": 6.234030723571777, + "learning_rate": 7.988439306358382e-05, + "loss": 1.1799, + "step": 691 + }, + { + "epoch": 0.048018874470890295, + "grad_norm": 7.930813312530518, + "learning_rate": 8e-05, + "loss": 1.0486, + "step": 692 + }, + { + "epoch": 0.048088265907986955, + "grad_norm": 6.172576904296875, + "learning_rate": 8.011560693641619e-05, + "loss": 0.8811, + "step": 693 + }, + { + "epoch": 0.048157657345083614, + "grad_norm": 5.816160202026367, + "learning_rate": 8.023121387283238e-05, + "loss": 0.8725, + "step": 694 + }, + { + "epoch": 0.04822704878218028, + "grad_norm": 14.719770431518555, + "learning_rate": 8.034682080924855e-05, + "loss": 1.3043, + "step": 695 + }, + { + "epoch": 0.04829644021927694, + "grad_norm": 6.711781978607178, + "learning_rate": 8.046242774566475e-05, + "loss": 1.321, + "step": 696 + }, + { + "epoch": 0.048365831656373606, + "grad_norm": 6.062890529632568, + "learning_rate": 8.057803468208092e-05, + "loss": 0.9024, + "step": 697 + }, + { + "epoch": 0.048435223093470266, + "grad_norm": 7.339975833892822, + "learning_rate": 8.069364161849712e-05, + "loss": 1.2583, + "step": 698 + }, + { + "epoch": 0.048504614530566925, + "grad_norm": 7.769414901733398, + "learning_rate": 8.08092485549133e-05, + "loss": 1.2646, + "step": 699 + }, + { + "epoch": 0.04857400596766359, + "grad_norm": 6.069802284240723, + "learning_rate": 8.092485549132948e-05, + "loss": 1.0212, + "step": 700 + }, + { + "epoch": 0.04864339740476025, + "grad_norm": 5.208179950714111, + "learning_rate": 8.104046242774567e-05, + "loss": 1.2241, + "step": 701 + }, + { + "epoch": 0.04871278884185692, + "grad_norm": 8.646373748779297, + "learning_rate": 8.115606936416186e-05, + "loss": 1.1415, + "step": 702 + }, + { + "epoch": 0.04878218027895358, + "grad_norm": 6.652426719665527, + "learning_rate": 8.127167630057804e-05, + "loss": 1.2569, + "step": 703 + }, + { + "epoch": 0.048851571716050236, + "grad_norm": 5.848877429962158, + "learning_rate": 8.138728323699423e-05, + "loss": 1.0853, + "step": 704 + }, + { + "epoch": 0.0489209631531469, + "grad_norm": 6.01759147644043, + "learning_rate": 8.15028901734104e-05, + "loss": 1.2817, + "step": 705 + }, + { + "epoch": 0.04899035459024356, + "grad_norm": 5.217519283294678, + "learning_rate": 8.16184971098266e-05, + "loss": 0.8917, + "step": 706 + }, + { + "epoch": 0.04905974602734023, + "grad_norm": 5.567752361297607, + "learning_rate": 8.173410404624277e-05, + "loss": 1.2297, + "step": 707 + }, + { + "epoch": 0.04912913746443689, + "grad_norm": 8.241035461425781, + "learning_rate": 8.184971098265897e-05, + "loss": 0.8622, + "step": 708 + }, + { + "epoch": 0.04919852890153355, + "grad_norm": 6.88716459274292, + "learning_rate": 8.196531791907515e-05, + "loss": 1.1641, + "step": 709 + }, + { + "epoch": 0.049267920338630214, + "grad_norm": 7.534900665283203, + "learning_rate": 8.208092485549133e-05, + "loss": 1.2459, + "step": 710 + }, + { + "epoch": 0.04933731177572687, + "grad_norm": 6.924213886260986, + "learning_rate": 8.219653179190752e-05, + "loss": 1.09, + "step": 711 + }, + { + "epoch": 0.04940670321282354, + "grad_norm": 6.802692413330078, + "learning_rate": 8.23121387283237e-05, + "loss": 1.0141, + "step": 712 + }, + { + "epoch": 0.0494760946499202, + "grad_norm": 6.493343830108643, + "learning_rate": 8.242774566473989e-05, + "loss": 1.4442, + "step": 713 + }, + { + "epoch": 0.04954548608701686, + "grad_norm": 6.8420023918151855, + "learning_rate": 8.254335260115608e-05, + "loss": 1.2239, + "step": 714 + }, + { + "epoch": 0.049614877524113525, + "grad_norm": 7.298171520233154, + "learning_rate": 8.265895953757226e-05, + "loss": 1.2198, + "step": 715 + }, + { + "epoch": 0.049684268961210185, + "grad_norm": 6.626737117767334, + "learning_rate": 8.277456647398845e-05, + "loss": 1.3301, + "step": 716 + }, + { + "epoch": 0.04975366039830685, + "grad_norm": 5.873114585876465, + "learning_rate": 8.289017341040462e-05, + "loss": 1.1564, + "step": 717 + }, + { + "epoch": 0.04982305183540351, + "grad_norm": 5.953824996948242, + "learning_rate": 8.300578034682082e-05, + "loss": 1.006, + "step": 718 + }, + { + "epoch": 0.04989244327250018, + "grad_norm": 5.802441596984863, + "learning_rate": 8.3121387283237e-05, + "loss": 1.1104, + "step": 719 + }, + { + "epoch": 0.049961834709596836, + "grad_norm": 6.563821315765381, + "learning_rate": 8.323699421965318e-05, + "loss": 1.1342, + "step": 720 + }, + { + "epoch": 0.050031226146693496, + "grad_norm": 5.940591812133789, + "learning_rate": 8.335260115606937e-05, + "loss": 1.0024, + "step": 721 + }, + { + "epoch": 0.05010061758379016, + "grad_norm": 5.948383808135986, + "learning_rate": 8.346820809248555e-05, + "loss": 1.6797, + "step": 722 + }, + { + "epoch": 0.05017000902088682, + "grad_norm": 5.607241630554199, + "learning_rate": 8.358381502890174e-05, + "loss": 1.2321, + "step": 723 + }, + { + "epoch": 0.05023940045798349, + "grad_norm": 5.658578872680664, + "learning_rate": 8.369942196531792e-05, + "loss": 1.0878, + "step": 724 + }, + { + "epoch": 0.05030879189508015, + "grad_norm": 5.765727996826172, + "learning_rate": 8.381502890173411e-05, + "loss": 1.152, + "step": 725 + }, + { + "epoch": 0.05037818333217681, + "grad_norm": 5.7814788818359375, + "learning_rate": 8.39306358381503e-05, + "loss": 1.148, + "step": 726 + }, + { + "epoch": 0.05044757476927347, + "grad_norm": 6.830914497375488, + "learning_rate": 8.404624277456647e-05, + "loss": 1.2732, + "step": 727 + }, + { + "epoch": 0.05051696620637013, + "grad_norm": 5.615469455718994, + "learning_rate": 8.416184971098267e-05, + "loss": 1.1127, + "step": 728 + }, + { + "epoch": 0.0505863576434668, + "grad_norm": 6.720466613769531, + "learning_rate": 8.427745664739884e-05, + "loss": 1.331, + "step": 729 + }, + { + "epoch": 0.05065574908056346, + "grad_norm": 5.826695442199707, + "learning_rate": 8.439306358381503e-05, + "loss": 1.3326, + "step": 730 + }, + { + "epoch": 0.05072514051766012, + "grad_norm": 10.105927467346191, + "learning_rate": 8.450867052023121e-05, + "loss": 1.1079, + "step": 731 + }, + { + "epoch": 0.050794531954756784, + "grad_norm": 5.612820148468018, + "learning_rate": 8.46242774566474e-05, + "loss": 0.9664, + "step": 732 + }, + { + "epoch": 0.050863923391853444, + "grad_norm": 5.9514641761779785, + "learning_rate": 8.473988439306359e-05, + "loss": 1.2526, + "step": 733 + }, + { + "epoch": 0.05093331482895011, + "grad_norm": 6.590364933013916, + "learning_rate": 8.485549132947977e-05, + "loss": 1.2196, + "step": 734 + }, + { + "epoch": 0.05100270626604677, + "grad_norm": 5.83685302734375, + "learning_rate": 8.497109826589596e-05, + "loss": 0.8484, + "step": 735 + }, + { + "epoch": 0.05107209770314343, + "grad_norm": 6.378995418548584, + "learning_rate": 8.508670520231215e-05, + "loss": 1.1671, + "step": 736 + }, + { + "epoch": 0.051141489140240096, + "grad_norm": 5.945015907287598, + "learning_rate": 8.520231213872832e-05, + "loss": 0.9584, + "step": 737 + }, + { + "epoch": 0.051210880577336755, + "grad_norm": 6.598106384277344, + "learning_rate": 8.531791907514452e-05, + "loss": 1.1682, + "step": 738 + }, + { + "epoch": 0.05128027201443342, + "grad_norm": 5.591760158538818, + "learning_rate": 8.543352601156069e-05, + "loss": 1.165, + "step": 739 + }, + { + "epoch": 0.05134966345153008, + "grad_norm": 6.1352057456970215, + "learning_rate": 8.554913294797689e-05, + "loss": 1.0055, + "step": 740 + }, + { + "epoch": 0.05141905488862674, + "grad_norm": 5.597328186035156, + "learning_rate": 8.566473988439306e-05, + "loss": 1.2035, + "step": 741 + }, + { + "epoch": 0.05148844632572341, + "grad_norm": 9.253616333007812, + "learning_rate": 8.578034682080925e-05, + "loss": 1.4107, + "step": 742 + }, + { + "epoch": 0.051557837762820066, + "grad_norm": 7.3413405418396, + "learning_rate": 8.589595375722544e-05, + "loss": 1.1567, + "step": 743 + }, + { + "epoch": 0.05162722919991673, + "grad_norm": 7.370278835296631, + "learning_rate": 8.601156069364162e-05, + "loss": 1.1782, + "step": 744 + }, + { + "epoch": 0.05169662063701339, + "grad_norm": 6.934401512145996, + "learning_rate": 8.612716763005781e-05, + "loss": 1.0456, + "step": 745 + }, + { + "epoch": 0.05176601207411005, + "grad_norm": 6.292669773101807, + "learning_rate": 8.6242774566474e-05, + "loss": 1.1504, + "step": 746 + }, + { + "epoch": 0.05183540351120672, + "grad_norm": 7.378996849060059, + "learning_rate": 8.635838150289017e-05, + "loss": 0.9354, + "step": 747 + }, + { + "epoch": 0.05190479494830338, + "grad_norm": 5.659523010253906, + "learning_rate": 8.647398843930637e-05, + "loss": 1.2284, + "step": 748 + }, + { + "epoch": 0.051974186385400044, + "grad_norm": 6.131499767303467, + "learning_rate": 8.658959537572254e-05, + "loss": 1.1344, + "step": 749 + }, + { + "epoch": 0.0520435778224967, + "grad_norm": 7.258898735046387, + "learning_rate": 8.670520231213874e-05, + "loss": 1.298, + "step": 750 + }, + { + "epoch": 0.05211296925959336, + "grad_norm": 8.475727081298828, + "learning_rate": 8.682080924855491e-05, + "loss": 1.006, + "step": 751 + }, + { + "epoch": 0.05218236069669003, + "grad_norm": 7.1479811668396, + "learning_rate": 8.69364161849711e-05, + "loss": 0.9802, + "step": 752 + }, + { + "epoch": 0.05225175213378669, + "grad_norm": 7.222642421722412, + "learning_rate": 8.705202312138728e-05, + "loss": 1.0965, + "step": 753 + }, + { + "epoch": 0.052321143570883355, + "grad_norm": 4.579987525939941, + "learning_rate": 8.716763005780347e-05, + "loss": 0.7435, + "step": 754 + }, + { + "epoch": 0.052390535007980014, + "grad_norm": 9.159570693969727, + "learning_rate": 8.728323699421966e-05, + "loss": 1.1296, + "step": 755 + }, + { + "epoch": 0.05245992644507668, + "grad_norm": 5.19547176361084, + "learning_rate": 8.739884393063584e-05, + "loss": 1.1029, + "step": 756 + }, + { + "epoch": 0.05252931788217334, + "grad_norm": 6.152987957000732, + "learning_rate": 8.751445086705203e-05, + "loss": 1.0701, + "step": 757 + }, + { + "epoch": 0.05259870931927, + "grad_norm": 6.496377468109131, + "learning_rate": 8.763005780346822e-05, + "loss": 1.1251, + "step": 758 + }, + { + "epoch": 0.052668100756366666, + "grad_norm": 7.582714557647705, + "learning_rate": 8.774566473988439e-05, + "loss": 1.0478, + "step": 759 + }, + { + "epoch": 0.052737492193463326, + "grad_norm": 7.916019916534424, + "learning_rate": 8.786127167630059e-05, + "loss": 0.9976, + "step": 760 + }, + { + "epoch": 0.05280688363055999, + "grad_norm": 6.22142219543457, + "learning_rate": 8.797687861271676e-05, + "loss": 0.8756, + "step": 761 + }, + { + "epoch": 0.05287627506765665, + "grad_norm": 7.2916789054870605, + "learning_rate": 8.809248554913295e-05, + "loss": 1.234, + "step": 762 + }, + { + "epoch": 0.05294566650475331, + "grad_norm": 7.1681623458862305, + "learning_rate": 8.820809248554913e-05, + "loss": 1.154, + "step": 763 + }, + { + "epoch": 0.05301505794184998, + "grad_norm": 6.886358261108398, + "learning_rate": 8.832369942196532e-05, + "loss": 1.3712, + "step": 764 + }, + { + "epoch": 0.05308444937894664, + "grad_norm": 7.967360973358154, + "learning_rate": 8.84393063583815e-05, + "loss": 1.4382, + "step": 765 + }, + { + "epoch": 0.0531538408160433, + "grad_norm": 4.9677348136901855, + "learning_rate": 8.855491329479769e-05, + "loss": 1.0019, + "step": 766 + }, + { + "epoch": 0.05322323225313996, + "grad_norm": 5.50779914855957, + "learning_rate": 8.867052023121388e-05, + "loss": 1.0028, + "step": 767 + }, + { + "epoch": 0.05329262369023662, + "grad_norm": 6.199459552764893, + "learning_rate": 8.878612716763006e-05, + "loss": 1.1425, + "step": 768 + }, + { + "epoch": 0.05336201512733329, + "grad_norm": 6.5877604484558105, + "learning_rate": 8.890173410404625e-05, + "loss": 1.12, + "step": 769 + }, + { + "epoch": 0.05343140656442995, + "grad_norm": 7.61715841293335, + "learning_rate": 8.901734104046244e-05, + "loss": 1.1624, + "step": 770 + }, + { + "epoch": 0.053500798001526614, + "grad_norm": 7.188614368438721, + "learning_rate": 8.913294797687862e-05, + "loss": 1.4155, + "step": 771 + }, + { + "epoch": 0.053570189438623274, + "grad_norm": 5.234808444976807, + "learning_rate": 8.92485549132948e-05, + "loss": 1.3025, + "step": 772 + }, + { + "epoch": 0.05363958087571993, + "grad_norm": 6.628874778747559, + "learning_rate": 8.9364161849711e-05, + "loss": 1.2441, + "step": 773 + }, + { + "epoch": 0.0537089723128166, + "grad_norm": 8.16806697845459, + "learning_rate": 8.947976878612717e-05, + "loss": 1.5413, + "step": 774 + }, + { + "epoch": 0.05377836374991326, + "grad_norm": 5.475348949432373, + "learning_rate": 8.959537572254337e-05, + "loss": 0.9402, + "step": 775 + }, + { + "epoch": 0.053847755187009926, + "grad_norm": 6.238165378570557, + "learning_rate": 8.971098265895954e-05, + "loss": 0.923, + "step": 776 + }, + { + "epoch": 0.053917146624106585, + "grad_norm": 6.768294811248779, + "learning_rate": 8.982658959537573e-05, + "loss": 1.1367, + "step": 777 + }, + { + "epoch": 0.053986538061203244, + "grad_norm": 8.414069175720215, + "learning_rate": 8.994219653179191e-05, + "loss": 1.2354, + "step": 778 + }, + { + "epoch": 0.05405592949829991, + "grad_norm": 6.110108852386475, + "learning_rate": 9.00578034682081e-05, + "loss": 0.7972, + "step": 779 + }, + { + "epoch": 0.05412532093539657, + "grad_norm": 6.391795635223389, + "learning_rate": 9.017341040462428e-05, + "loss": 1.1191, + "step": 780 + }, + { + "epoch": 0.05419471237249324, + "grad_norm": 5.249054431915283, + "learning_rate": 9.028901734104047e-05, + "loss": 0.8917, + "step": 781 + }, + { + "epoch": 0.054264103809589896, + "grad_norm": 6.916954517364502, + "learning_rate": 9.040462427745666e-05, + "loss": 1.1407, + "step": 782 + }, + { + "epoch": 0.054333495246686556, + "grad_norm": 5.581148147583008, + "learning_rate": 9.052023121387284e-05, + "loss": 1.0354, + "step": 783 + }, + { + "epoch": 0.05440288668378322, + "grad_norm": 7.4984354972839355, + "learning_rate": 9.063583815028902e-05, + "loss": 1.3667, + "step": 784 + }, + { + "epoch": 0.05447227812087988, + "grad_norm": 5.28513240814209, + "learning_rate": 9.075144508670522e-05, + "loss": 1.033, + "step": 785 + }, + { + "epoch": 0.05454166955797655, + "grad_norm": 7.554996013641357, + "learning_rate": 9.086705202312139e-05, + "loss": 1.165, + "step": 786 + }, + { + "epoch": 0.05461106099507321, + "grad_norm": 8.1099214553833, + "learning_rate": 9.098265895953757e-05, + "loss": 1.67, + "step": 787 + }, + { + "epoch": 0.05468045243216987, + "grad_norm": 6.848842620849609, + "learning_rate": 9.109826589595376e-05, + "loss": 1.1722, + "step": 788 + }, + { + "epoch": 0.05474984386926653, + "grad_norm": 6.454953670501709, + "learning_rate": 9.121387283236995e-05, + "loss": 1.0817, + "step": 789 + }, + { + "epoch": 0.05481923530636319, + "grad_norm": 7.364956378936768, + "learning_rate": 9.132947976878613e-05, + "loss": 1.1743, + "step": 790 + }, + { + "epoch": 0.05488862674345986, + "grad_norm": 13.135334014892578, + "learning_rate": 9.144508670520232e-05, + "loss": 1.0832, + "step": 791 + }, + { + "epoch": 0.05495801818055652, + "grad_norm": 5.161467552185059, + "learning_rate": 9.15606936416185e-05, + "loss": 0.9656, + "step": 792 + }, + { + "epoch": 0.055027409617653185, + "grad_norm": 7.815332412719727, + "learning_rate": 9.167630057803469e-05, + "loss": 1.3301, + "step": 793 + }, + { + "epoch": 0.055096801054749844, + "grad_norm": 6.59698486328125, + "learning_rate": 9.179190751445086e-05, + "loss": 1.3082, + "step": 794 + }, + { + "epoch": 0.055166192491846504, + "grad_norm": 7.163899898529053, + "learning_rate": 9.190751445086706e-05, + "loss": 1.3717, + "step": 795 + }, + { + "epoch": 0.05523558392894317, + "grad_norm": 5.83208703994751, + "learning_rate": 9.202312138728324e-05, + "loss": 1.139, + "step": 796 + }, + { + "epoch": 0.05530497536603983, + "grad_norm": 6.256465911865234, + "learning_rate": 9.213872832369944e-05, + "loss": 1.4401, + "step": 797 + }, + { + "epoch": 0.055374366803136496, + "grad_norm": 8.379135131835938, + "learning_rate": 9.225433526011561e-05, + "loss": 1.2332, + "step": 798 + }, + { + "epoch": 0.055443758240233156, + "grad_norm": 5.324146747589111, + "learning_rate": 9.23699421965318e-05, + "loss": 1.1272, + "step": 799 + }, + { + "epoch": 0.055513149677329815, + "grad_norm": 5.1640496253967285, + "learning_rate": 9.248554913294798e-05, + "loss": 1.101, + "step": 800 + }, + { + "epoch": 0.05558254111442648, + "grad_norm": 7.482885837554932, + "learning_rate": 9.260115606936417e-05, + "loss": 1.2877, + "step": 801 + }, + { + "epoch": 0.05565193255152314, + "grad_norm": 5.627552509307861, + "learning_rate": 9.271676300578035e-05, + "loss": 1.0165, + "step": 802 + }, + { + "epoch": 0.05572132398861981, + "grad_norm": 5.219727039337158, + "learning_rate": 9.283236994219654e-05, + "loss": 0.985, + "step": 803 + }, + { + "epoch": 0.05579071542571647, + "grad_norm": 6.12885046005249, + "learning_rate": 9.294797687861271e-05, + "loss": 1.1393, + "step": 804 + }, + { + "epoch": 0.055860106862813126, + "grad_norm": 6.639613151550293, + "learning_rate": 9.306358381502891e-05, + "loss": 0.9747, + "step": 805 + }, + { + "epoch": 0.05592949829990979, + "grad_norm": 8.42887020111084, + "learning_rate": 9.317919075144509e-05, + "loss": 0.931, + "step": 806 + }, + { + "epoch": 0.05599888973700645, + "grad_norm": 5.657486915588379, + "learning_rate": 9.329479768786129e-05, + "loss": 1.2718, + "step": 807 + }, + { + "epoch": 0.05606828117410312, + "grad_norm": 11.213651657104492, + "learning_rate": 9.341040462427746e-05, + "loss": 1.4845, + "step": 808 + }, + { + "epoch": 0.05613767261119978, + "grad_norm": 7.240329265594482, + "learning_rate": 9.352601156069364e-05, + "loss": 1.1334, + "step": 809 + }, + { + "epoch": 0.05620706404829644, + "grad_norm": 7.276681900024414, + "learning_rate": 9.364161849710983e-05, + "loss": 1.3353, + "step": 810 + }, + { + "epoch": 0.056276455485393104, + "grad_norm": 6.131457805633545, + "learning_rate": 9.375722543352602e-05, + "loss": 1.0343, + "step": 811 + }, + { + "epoch": 0.05634584692248976, + "grad_norm": 8.27358341217041, + "learning_rate": 9.38728323699422e-05, + "loss": 1.1124, + "step": 812 + }, + { + "epoch": 0.05641523835958643, + "grad_norm": 7.653395175933838, + "learning_rate": 9.398843930635839e-05, + "loss": 1.2113, + "step": 813 + }, + { + "epoch": 0.05648462979668309, + "grad_norm": 5.845747470855713, + "learning_rate": 9.410404624277456e-05, + "loss": 1.088, + "step": 814 + }, + { + "epoch": 0.05655402123377975, + "grad_norm": 6.134220123291016, + "learning_rate": 9.421965317919076e-05, + "loss": 1.4531, + "step": 815 + }, + { + "epoch": 0.056623412670876415, + "grad_norm": 5.392806053161621, + "learning_rate": 9.433526011560693e-05, + "loss": 1.0023, + "step": 816 + }, + { + "epoch": 0.056692804107973074, + "grad_norm": 6.585524082183838, + "learning_rate": 9.445086705202313e-05, + "loss": 1.2098, + "step": 817 + }, + { + "epoch": 0.05676219554506974, + "grad_norm": 4.945289134979248, + "learning_rate": 9.456647398843931e-05, + "loss": 1.0043, + "step": 818 + }, + { + "epoch": 0.0568315869821664, + "grad_norm": 7.974738121032715, + "learning_rate": 9.468208092485549e-05, + "loss": 1.1296, + "step": 819 + }, + { + "epoch": 0.05690097841926306, + "grad_norm": 6.3792314529418945, + "learning_rate": 9.479768786127168e-05, + "loss": 1.216, + "step": 820 + }, + { + "epoch": 0.056970369856359726, + "grad_norm": 6.148492813110352, + "learning_rate": 9.491329479768787e-05, + "loss": 0.9839, + "step": 821 + }, + { + "epoch": 0.057039761293456386, + "grad_norm": 8.577494621276855, + "learning_rate": 9.502890173410405e-05, + "loss": 1.2931, + "step": 822 + }, + { + "epoch": 0.05710915273055305, + "grad_norm": 9.200427055358887, + "learning_rate": 9.514450867052024e-05, + "loss": 1.1031, + "step": 823 + }, + { + "epoch": 0.05717854416764971, + "grad_norm": 6.413390636444092, + "learning_rate": 9.526011560693642e-05, + "loss": 0.9716, + "step": 824 + }, + { + "epoch": 0.05724793560474637, + "grad_norm": 5.897536754608154, + "learning_rate": 9.537572254335261e-05, + "loss": 1.1257, + "step": 825 + }, + { + "epoch": 0.05731732704184304, + "grad_norm": 5.835260391235352, + "learning_rate": 9.549132947976878e-05, + "loss": 0.9781, + "step": 826 + }, + { + "epoch": 0.0573867184789397, + "grad_norm": 8.361189842224121, + "learning_rate": 9.560693641618498e-05, + "loss": 1.4804, + "step": 827 + }, + { + "epoch": 0.05745610991603636, + "grad_norm": 5.532903671264648, + "learning_rate": 9.572254335260116e-05, + "loss": 1.1425, + "step": 828 + }, + { + "epoch": 0.05752550135313302, + "grad_norm": 6.5775065422058105, + "learning_rate": 9.583815028901734e-05, + "loss": 1.073, + "step": 829 + }, + { + "epoch": 0.05759489279022969, + "grad_norm": 7.116223335266113, + "learning_rate": 9.595375722543353e-05, + "loss": 1.2731, + "step": 830 + }, + { + "epoch": 0.05766428422732635, + "grad_norm": 4.290632247924805, + "learning_rate": 9.606936416184971e-05, + "loss": 0.976, + "step": 831 + }, + { + "epoch": 0.05773367566442301, + "grad_norm": 6.379570484161377, + "learning_rate": 9.61849710982659e-05, + "loss": 1.5432, + "step": 832 + }, + { + "epoch": 0.057803067101519674, + "grad_norm": 6.793046474456787, + "learning_rate": 9.630057803468209e-05, + "loss": 1.3489, + "step": 833 + }, + { + "epoch": 0.057872458538616334, + "grad_norm": 17.812660217285156, + "learning_rate": 9.641618497109827e-05, + "loss": 1.3952, + "step": 834 + }, + { + "epoch": 0.057941849975713, + "grad_norm": 6.892359733581543, + "learning_rate": 9.653179190751446e-05, + "loss": 0.8042, + "step": 835 + }, + { + "epoch": 0.05801124141280966, + "grad_norm": 6.651463508605957, + "learning_rate": 9.664739884393063e-05, + "loss": 1.4757, + "step": 836 + }, + { + "epoch": 0.05808063284990632, + "grad_norm": 10.628656387329102, + "learning_rate": 9.676300578034683e-05, + "loss": 1.4394, + "step": 837 + }, + { + "epoch": 0.058150024287002985, + "grad_norm": 6.82949161529541, + "learning_rate": 9.6878612716763e-05, + "loss": 1.1796, + "step": 838 + }, + { + "epoch": 0.058219415724099645, + "grad_norm": 6.134788513183594, + "learning_rate": 9.699421965317919e-05, + "loss": 0.8566, + "step": 839 + }, + { + "epoch": 0.05828880716119631, + "grad_norm": 6.816523551940918, + "learning_rate": 9.710982658959538e-05, + "loss": 1.1251, + "step": 840 + }, + { + "epoch": 0.05835819859829297, + "grad_norm": 5.2859015464782715, + "learning_rate": 9.722543352601156e-05, + "loss": 1.2099, + "step": 841 + }, + { + "epoch": 0.05842759003538963, + "grad_norm": 6.437781810760498, + "learning_rate": 9.734104046242775e-05, + "loss": 1.3275, + "step": 842 + }, + { + "epoch": 0.0584969814724863, + "grad_norm": 6.587648391723633, + "learning_rate": 9.745664739884394e-05, + "loss": 1.2565, + "step": 843 + }, + { + "epoch": 0.058566372909582956, + "grad_norm": 5.468827724456787, + "learning_rate": 9.757225433526012e-05, + "loss": 1.0565, + "step": 844 + }, + { + "epoch": 0.05863576434667962, + "grad_norm": 5.360687732696533, + "learning_rate": 9.768786127167631e-05, + "loss": 1.0094, + "step": 845 + }, + { + "epoch": 0.05870515578377628, + "grad_norm": 6.480072975158691, + "learning_rate": 9.780346820809248e-05, + "loss": 1.2104, + "step": 846 + }, + { + "epoch": 0.05877454722087294, + "grad_norm": 7.292744159698486, + "learning_rate": 9.791907514450868e-05, + "loss": 1.3065, + "step": 847 + }, + { + "epoch": 0.05884393865796961, + "grad_norm": 6.169072151184082, + "learning_rate": 9.803468208092485e-05, + "loss": 1.2539, + "step": 848 + }, + { + "epoch": 0.05891333009506627, + "grad_norm": 10.052629470825195, + "learning_rate": 9.815028901734105e-05, + "loss": 1.1691, + "step": 849 + }, + { + "epoch": 0.058982721532162934, + "grad_norm": 6.97867488861084, + "learning_rate": 9.826589595375723e-05, + "loss": 1.1114, + "step": 850 + }, + { + "epoch": 0.05905211296925959, + "grad_norm": 7.43370246887207, + "learning_rate": 9.838150289017341e-05, + "loss": 1.1635, + "step": 851 + }, + { + "epoch": 0.05912150440635625, + "grad_norm": 18.632598876953125, + "learning_rate": 9.84971098265896e-05, + "loss": 1.3149, + "step": 852 + }, + { + "epoch": 0.05919089584345292, + "grad_norm": 9.65411376953125, + "learning_rate": 9.861271676300578e-05, + "loss": 1.0844, + "step": 853 + }, + { + "epoch": 0.05926028728054958, + "grad_norm": 7.443455219268799, + "learning_rate": 9.872832369942197e-05, + "loss": 1.4471, + "step": 854 + }, + { + "epoch": 0.059329678717646245, + "grad_norm": 8.323163032531738, + "learning_rate": 9.884393063583816e-05, + "loss": 1.4509, + "step": 855 + }, + { + "epoch": 0.059399070154742904, + "grad_norm": 7.2459821701049805, + "learning_rate": 9.895953757225433e-05, + "loss": 1.148, + "step": 856 + }, + { + "epoch": 0.059468461591839564, + "grad_norm": 7.907388210296631, + "learning_rate": 9.907514450867053e-05, + "loss": 1.3672, + "step": 857 + }, + { + "epoch": 0.05953785302893623, + "grad_norm": 8.557660102844238, + "learning_rate": 9.91907514450867e-05, + "loss": 1.3701, + "step": 858 + }, + { + "epoch": 0.05960724446603289, + "grad_norm": 11.708367347717285, + "learning_rate": 9.93063583815029e-05, + "loss": 1.0848, + "step": 859 + }, + { + "epoch": 0.059676635903129556, + "grad_norm": 6.669211387634277, + "learning_rate": 9.942196531791907e-05, + "loss": 1.276, + "step": 860 + }, + { + "epoch": 0.059746027340226215, + "grad_norm": 5.340997219085693, + "learning_rate": 9.953757225433526e-05, + "loss": 1.0473, + "step": 861 + }, + { + "epoch": 0.05981541877732288, + "grad_norm": 7.248746395111084, + "learning_rate": 9.965317919075145e-05, + "loss": 0.9901, + "step": 862 + }, + { + "epoch": 0.05988481021441954, + "grad_norm": 6.330059051513672, + "learning_rate": 9.976878612716763e-05, + "loss": 1.4335, + "step": 863 + }, + { + "epoch": 0.0599542016515162, + "grad_norm": 6.406825542449951, + "learning_rate": 9.988439306358382e-05, + "loss": 1.3255, + "step": 864 + }, + { + "epoch": 0.06002359308861287, + "grad_norm": 8.152541160583496, + "learning_rate": 0.0001, + "loss": 1.3326, + "step": 865 + }, + { + "epoch": 0.06009298452570953, + "grad_norm": 7.8393940925598145, + "learning_rate": 0.00010011560693641619, + "loss": 1.7044, + "step": 866 + }, + { + "epoch": 0.06016237596280619, + "grad_norm": 5.38602876663208, + "learning_rate": 0.00010023121387283239, + "loss": 1.115, + "step": 867 + }, + { + "epoch": 0.06023176739990285, + "grad_norm": 4.524908065795898, + "learning_rate": 0.00010034682080924855, + "loss": 0.9718, + "step": 868 + }, + { + "epoch": 0.06030115883699951, + "grad_norm": 6.429388046264648, + "learning_rate": 0.00010046242774566475, + "loss": 1.5296, + "step": 869 + }, + { + "epoch": 0.06037055027409618, + "grad_norm": 7.4378581047058105, + "learning_rate": 0.00010057803468208094, + "loss": 1.3455, + "step": 870 + }, + { + "epoch": 0.06043994171119284, + "grad_norm": 8.678258895874023, + "learning_rate": 0.00010069364161849712, + "loss": 1.2474, + "step": 871 + }, + { + "epoch": 0.060509333148289504, + "grad_norm": 5.248087406158447, + "learning_rate": 0.0001008092485549133, + "loss": 0.9911, + "step": 872 + }, + { + "epoch": 0.060578724585386164, + "grad_norm": 5.474868297576904, + "learning_rate": 0.00010092485549132948, + "loss": 1.1676, + "step": 873 + }, + { + "epoch": 0.06064811602248282, + "grad_norm": 6.164037227630615, + "learning_rate": 0.00010104046242774568, + "loss": 1.3606, + "step": 874 + }, + { + "epoch": 0.06071750745957949, + "grad_norm": 7.988519668579102, + "learning_rate": 0.00010115606936416187, + "loss": 1.1812, + "step": 875 + }, + { + "epoch": 0.06078689889667615, + "grad_norm": 6.476558208465576, + "learning_rate": 0.00010127167630057804, + "loss": 1.7219, + "step": 876 + }, + { + "epoch": 0.060856290333772815, + "grad_norm": 4.907296657562256, + "learning_rate": 0.00010138728323699423, + "loss": 0.9978, + "step": 877 + }, + { + "epoch": 0.060925681770869475, + "grad_norm": 5.121548175811768, + "learning_rate": 0.00010150289017341041, + "loss": 1.1589, + "step": 878 + }, + { + "epoch": 0.060995073207966134, + "grad_norm": 6.029032230377197, + "learning_rate": 0.0001016184971098266, + "loss": 1.1517, + "step": 879 + }, + { + "epoch": 0.0610644646450628, + "grad_norm": 4.927402496337891, + "learning_rate": 0.00010173410404624277, + "loss": 0.9722, + "step": 880 + }, + { + "epoch": 0.06113385608215946, + "grad_norm": 7.581357955932617, + "learning_rate": 0.00010184971098265896, + "loss": 0.9346, + "step": 881 + }, + { + "epoch": 0.061203247519256126, + "grad_norm": 8.653351783752441, + "learning_rate": 0.00010196531791907516, + "loss": 1.2881, + "step": 882 + }, + { + "epoch": 0.061272638956352786, + "grad_norm": 8.063896179199219, + "learning_rate": 0.00010208092485549134, + "loss": 1.182, + "step": 883 + }, + { + "epoch": 0.061342030393449445, + "grad_norm": 7.5197672843933105, + "learning_rate": 0.00010219653179190752, + "loss": 1.4159, + "step": 884 + }, + { + "epoch": 0.06141142183054611, + "grad_norm": 6.191446781158447, + "learning_rate": 0.0001023121387283237, + "loss": 1.2203, + "step": 885 + }, + { + "epoch": 0.06148081326764277, + "grad_norm": 5.887239456176758, + "learning_rate": 0.00010242774566473989, + "loss": 1.1302, + "step": 886 + }, + { + "epoch": 0.06155020470473944, + "grad_norm": 6.743189811706543, + "learning_rate": 0.00010254335260115609, + "loss": 1.3451, + "step": 887 + }, + { + "epoch": 0.0616195961418361, + "grad_norm": 6.834993362426758, + "learning_rate": 0.00010265895953757225, + "loss": 1.243, + "step": 888 + }, + { + "epoch": 0.06168898757893276, + "grad_norm": 5.50251579284668, + "learning_rate": 0.00010277456647398845, + "loss": 1.1551, + "step": 889 + }, + { + "epoch": 0.06175837901602942, + "grad_norm": 5.892317295074463, + "learning_rate": 0.00010289017341040463, + "loss": 1.0909, + "step": 890 + }, + { + "epoch": 0.06182777045312608, + "grad_norm": 5.736100196838379, + "learning_rate": 0.00010300578034682082, + "loss": 1.1636, + "step": 891 + }, + { + "epoch": 0.06189716189022275, + "grad_norm": 6.535713195800781, + "learning_rate": 0.00010312138728323699, + "loss": 1.4742, + "step": 892 + }, + { + "epoch": 0.06196655332731941, + "grad_norm": 6.307315349578857, + "learning_rate": 0.00010323699421965318, + "loss": 1.1478, + "step": 893 + }, + { + "epoch": 0.06203594476441607, + "grad_norm": 6.734628677368164, + "learning_rate": 0.00010335260115606938, + "loss": 1.3842, + "step": 894 + }, + { + "epoch": 0.062105336201512734, + "grad_norm": 5.578872203826904, + "learning_rate": 0.00010346820809248556, + "loss": 1.0868, + "step": 895 + }, + { + "epoch": 0.062174727638609394, + "grad_norm": 5.095592498779297, + "learning_rate": 0.00010358381502890174, + "loss": 1.0717, + "step": 896 + }, + { + "epoch": 0.06224411907570606, + "grad_norm": 5.256373405456543, + "learning_rate": 0.00010369942196531792, + "loss": 1.2357, + "step": 897 + }, + { + "epoch": 0.06231351051280272, + "grad_norm": 5.652446269989014, + "learning_rate": 0.00010381502890173411, + "loss": 1.3027, + "step": 898 + }, + { + "epoch": 0.062382901949899386, + "grad_norm": 6.656091213226318, + "learning_rate": 0.00010393063583815031, + "loss": 1.3556, + "step": 899 + }, + { + "epoch": 0.062452293386996045, + "grad_norm": 5.792941570281982, + "learning_rate": 0.00010404624277456647, + "loss": 1.064, + "step": 900 + }, + { + "epoch": 0.0625216848240927, + "grad_norm": 5.73951530456543, + "learning_rate": 0.00010416184971098267, + "loss": 1.1082, + "step": 901 + }, + { + "epoch": 0.06259107626118937, + "grad_norm": 5.599228382110596, + "learning_rate": 0.00010427745664739885, + "loss": 1.0079, + "step": 902 + }, + { + "epoch": 0.06266046769828604, + "grad_norm": 5.1977057456970215, + "learning_rate": 0.00010439306358381504, + "loss": 1.011, + "step": 903 + }, + { + "epoch": 0.06272985913538269, + "grad_norm": 5.309018135070801, + "learning_rate": 0.00010450867052023121, + "loss": 0.9624, + "step": 904 + }, + { + "epoch": 0.06279925057247936, + "grad_norm": 5.113820552825928, + "learning_rate": 0.0001046242774566474, + "loss": 0.8711, + "step": 905 + }, + { + "epoch": 0.06286864200957602, + "grad_norm": 7.552628993988037, + "learning_rate": 0.00010473988439306359, + "loss": 1.1991, + "step": 906 + }, + { + "epoch": 0.06293803344667268, + "grad_norm": 8.376805305480957, + "learning_rate": 0.00010485549132947979, + "loss": 1.6156, + "step": 907 + }, + { + "epoch": 0.06300742488376934, + "grad_norm": 6.625391960144043, + "learning_rate": 0.00010497109826589594, + "loss": 1.3359, + "step": 908 + }, + { + "epoch": 0.06307681632086601, + "grad_norm": 5.942075729370117, + "learning_rate": 0.00010508670520231214, + "loss": 1.2395, + "step": 909 + }, + { + "epoch": 0.06314620775796266, + "grad_norm": 6.282374858856201, + "learning_rate": 0.00010520231213872833, + "loss": 1.388, + "step": 910 + }, + { + "epoch": 0.06321559919505933, + "grad_norm": 7.248420238494873, + "learning_rate": 0.00010531791907514452, + "loss": 1.4, + "step": 911 + }, + { + "epoch": 0.063284990632156, + "grad_norm": 6.960402488708496, + "learning_rate": 0.00010543352601156069, + "loss": 1.2277, + "step": 912 + }, + { + "epoch": 0.06335438206925266, + "grad_norm": 7.479408264160156, + "learning_rate": 0.00010554913294797688, + "loss": 1.3187, + "step": 913 + }, + { + "epoch": 0.06342377350634931, + "grad_norm": 5.168111801147461, + "learning_rate": 0.00010566473988439308, + "loss": 1.1359, + "step": 914 + }, + { + "epoch": 0.06349316494344598, + "grad_norm": 5.400572299957275, + "learning_rate": 0.00010578034682080926, + "loss": 0.9537, + "step": 915 + }, + { + "epoch": 0.06356255638054265, + "grad_norm": 4.9346466064453125, + "learning_rate": 0.00010589595375722543, + "loss": 1.0444, + "step": 916 + }, + { + "epoch": 0.0636319478176393, + "grad_norm": 6.16260290145874, + "learning_rate": 0.00010601156069364162, + "loss": 1.0903, + "step": 917 + }, + { + "epoch": 0.06370133925473596, + "grad_norm": 6.241594314575195, + "learning_rate": 0.0001061271676300578, + "loss": 1.2787, + "step": 918 + }, + { + "epoch": 0.06377073069183263, + "grad_norm": 5.498849391937256, + "learning_rate": 0.000106242774566474, + "loss": 1.1237, + "step": 919 + }, + { + "epoch": 0.0638401221289293, + "grad_norm": 5.197683334350586, + "learning_rate": 0.00010635838150289017, + "loss": 0.9965, + "step": 920 + }, + { + "epoch": 0.06390951356602595, + "grad_norm": 6.743277072906494, + "learning_rate": 0.00010647398843930637, + "loss": 1.2598, + "step": 921 + }, + { + "epoch": 0.06397890500312262, + "grad_norm": 5.357138633728027, + "learning_rate": 0.00010658959537572255, + "loss": 1.0766, + "step": 922 + }, + { + "epoch": 0.06404829644021928, + "grad_norm": 8.184362411499023, + "learning_rate": 0.00010670520231213874, + "loss": 1.4859, + "step": 923 + }, + { + "epoch": 0.06411768787731593, + "grad_norm": 5.988158702850342, + "learning_rate": 0.00010682080924855491, + "loss": 1.0676, + "step": 924 + }, + { + "epoch": 0.0641870793144126, + "grad_norm": 6.053996562957764, + "learning_rate": 0.0001069364161849711, + "loss": 1.1132, + "step": 925 + }, + { + "epoch": 0.06425647075150927, + "grad_norm": 8.990402221679688, + "learning_rate": 0.0001070520231213873, + "loss": 1.1402, + "step": 926 + }, + { + "epoch": 0.06432586218860592, + "grad_norm": 6.056857585906982, + "learning_rate": 0.00010716763005780348, + "loss": 1.1471, + "step": 927 + }, + { + "epoch": 0.06439525362570259, + "grad_norm": 6.745326519012451, + "learning_rate": 0.00010728323699421965, + "loss": 1.0058, + "step": 928 + }, + { + "epoch": 0.06446464506279925, + "grad_norm": 4.0814127922058105, + "learning_rate": 0.00010739884393063584, + "loss": 0.9197, + "step": 929 + }, + { + "epoch": 0.06453403649989592, + "grad_norm": 4.70957612991333, + "learning_rate": 0.00010751445086705203, + "loss": 0.9212, + "step": 930 + }, + { + "epoch": 0.06460342793699257, + "grad_norm": 7.327798843383789, + "learning_rate": 0.00010763005780346823, + "loss": 1.2886, + "step": 931 + }, + { + "epoch": 0.06467281937408924, + "grad_norm": 6.017044544219971, + "learning_rate": 0.00010774566473988439, + "loss": 1.3174, + "step": 932 + }, + { + "epoch": 0.0647422108111859, + "grad_norm": 6.12582540512085, + "learning_rate": 0.00010786127167630057, + "loss": 1.3317, + "step": 933 + }, + { + "epoch": 0.06481160224828256, + "grad_norm": 6.394991397857666, + "learning_rate": 0.00010797687861271677, + "loss": 1.003, + "step": 934 + }, + { + "epoch": 0.06488099368537922, + "grad_norm": 8.742952346801758, + "learning_rate": 0.00010809248554913296, + "loss": 1.3759, + "step": 935 + }, + { + "epoch": 0.06495038512247589, + "grad_norm": 6.328347682952881, + "learning_rate": 0.00010820809248554913, + "loss": 1.1886, + "step": 936 + }, + { + "epoch": 0.06501977655957254, + "grad_norm": 5.139235496520996, + "learning_rate": 0.00010832369942196532, + "loss": 1.3099, + "step": 937 + }, + { + "epoch": 0.06508916799666921, + "grad_norm": 6.482737064361572, + "learning_rate": 0.0001084393063583815, + "loss": 1.3369, + "step": 938 + }, + { + "epoch": 0.06515855943376588, + "grad_norm": 3.7927403450012207, + "learning_rate": 0.0001085549132947977, + "loss": 0.7153, + "step": 939 + }, + { + "epoch": 0.06522795087086254, + "grad_norm": 7.439101219177246, + "learning_rate": 0.00010867052023121386, + "loss": 0.8634, + "step": 940 + }, + { + "epoch": 0.0652973423079592, + "grad_norm": 5.673219680786133, + "learning_rate": 0.00010878612716763006, + "loss": 1.2065, + "step": 941 + }, + { + "epoch": 0.06536673374505586, + "grad_norm": 5.8922834396362305, + "learning_rate": 0.00010890173410404625, + "loss": 1.1255, + "step": 942 + }, + { + "epoch": 0.06543612518215253, + "grad_norm": 5.8634419441223145, + "learning_rate": 0.00010901734104046243, + "loss": 1.1091, + "step": 943 + }, + { + "epoch": 0.06550551661924918, + "grad_norm": 6.611083030700684, + "learning_rate": 0.00010913294797687861, + "loss": 1.2807, + "step": 944 + }, + { + "epoch": 0.06557490805634585, + "grad_norm": 12.461170196533203, + "learning_rate": 0.0001092485549132948, + "loss": 1.5656, + "step": 945 + }, + { + "epoch": 0.06564429949344251, + "grad_norm": 6.513397693634033, + "learning_rate": 0.000109364161849711, + "loss": 1.2207, + "step": 946 + }, + { + "epoch": 0.06571369093053916, + "grad_norm": 5.571642875671387, + "learning_rate": 0.00010947976878612718, + "loss": 0.9579, + "step": 947 + }, + { + "epoch": 0.06578308236763583, + "grad_norm": 5.317357063293457, + "learning_rate": 0.00010959537572254335, + "loss": 0.9512, + "step": 948 + }, + { + "epoch": 0.0658524738047325, + "grad_norm": 7.664186000823975, + "learning_rate": 0.00010971098265895954, + "loss": 1.0936, + "step": 949 + }, + { + "epoch": 0.06592186524182916, + "grad_norm": 6.832206726074219, + "learning_rate": 0.00010982658959537572, + "loss": 1.5323, + "step": 950 + }, + { + "epoch": 0.06599125667892582, + "grad_norm": 5.505645751953125, + "learning_rate": 0.00010994219653179192, + "loss": 1.346, + "step": 951 + }, + { + "epoch": 0.06606064811602248, + "grad_norm": 5.296020984649658, + "learning_rate": 0.00011005780346820808, + "loss": 0.893, + "step": 952 + }, + { + "epoch": 0.06613003955311915, + "grad_norm": 5.063529014587402, + "learning_rate": 0.00011017341040462428, + "loss": 1.0867, + "step": 953 + }, + { + "epoch": 0.0661994309902158, + "grad_norm": 6.027496337890625, + "learning_rate": 0.00011028901734104047, + "loss": 1.1832, + "step": 954 + }, + { + "epoch": 0.06626882242731247, + "grad_norm": 6.625126361846924, + "learning_rate": 0.00011040462427745666, + "loss": 1.1719, + "step": 955 + }, + { + "epoch": 0.06633821386440913, + "grad_norm": 4.745388507843018, + "learning_rate": 0.00011052023121387283, + "loss": 1.1237, + "step": 956 + }, + { + "epoch": 0.0664076053015058, + "grad_norm": 6.2461748123168945, + "learning_rate": 0.00011063583815028901, + "loss": 0.9818, + "step": 957 + }, + { + "epoch": 0.06647699673860245, + "grad_norm": 9.020347595214844, + "learning_rate": 0.00011075144508670521, + "loss": 0.9836, + "step": 958 + }, + { + "epoch": 0.06654638817569912, + "grad_norm": 7.848173141479492, + "learning_rate": 0.0001108670520231214, + "loss": 1.1535, + "step": 959 + }, + { + "epoch": 0.06661577961279579, + "grad_norm": 6.484100341796875, + "learning_rate": 0.00011098265895953757, + "loss": 1.2506, + "step": 960 + }, + { + "epoch": 0.06668517104989244, + "grad_norm": 6.783143043518066, + "learning_rate": 0.00011109826589595376, + "loss": 1.116, + "step": 961 + }, + { + "epoch": 0.0667545624869891, + "grad_norm": 10.31851863861084, + "learning_rate": 0.00011121387283236995, + "loss": 1.6591, + "step": 962 + }, + { + "epoch": 0.06682395392408577, + "grad_norm": 5.5841193199157715, + "learning_rate": 0.00011132947976878613, + "loss": 1.2022, + "step": 963 + }, + { + "epoch": 0.06689334536118242, + "grad_norm": 6.311263561248779, + "learning_rate": 0.00011144508670520233, + "loss": 1.171, + "step": 964 + }, + { + "epoch": 0.06696273679827909, + "grad_norm": 6.867467880249023, + "learning_rate": 0.00011156069364161849, + "loss": 1.3425, + "step": 965 + }, + { + "epoch": 0.06703212823537576, + "grad_norm": 7.001997470855713, + "learning_rate": 0.00011167630057803469, + "loss": 1.1748, + "step": 966 + }, + { + "epoch": 0.06710151967247242, + "grad_norm": 6.335434913635254, + "learning_rate": 0.00011179190751445088, + "loss": 1.2366, + "step": 967 + }, + { + "epoch": 0.06717091110956908, + "grad_norm": 6.800333499908447, + "learning_rate": 0.00011190751445086706, + "loss": 1.3152, + "step": 968 + }, + { + "epoch": 0.06724030254666574, + "grad_norm": 5.708077430725098, + "learning_rate": 0.00011202312138728324, + "loss": 1.0958, + "step": 969 + }, + { + "epoch": 0.06730969398376241, + "grad_norm": 7.752738952636719, + "learning_rate": 0.00011213872832369942, + "loss": 1.1952, + "step": 970 + }, + { + "epoch": 0.06737908542085906, + "grad_norm": 7.714700698852539, + "learning_rate": 0.00011225433526011562, + "loss": 1.4082, + "step": 971 + }, + { + "epoch": 0.06744847685795573, + "grad_norm": 5.32179594039917, + "learning_rate": 0.00011236994219653181, + "loss": 1.0535, + "step": 972 + }, + { + "epoch": 0.0675178682950524, + "grad_norm": 6.186491966247559, + "learning_rate": 0.00011248554913294798, + "loss": 1.0231, + "step": 973 + }, + { + "epoch": 0.06758725973214905, + "grad_norm": 6.509041786193848, + "learning_rate": 0.00011260115606936417, + "loss": 1.0375, + "step": 974 + }, + { + "epoch": 0.06765665116924571, + "grad_norm": 6.459548473358154, + "learning_rate": 0.00011271676300578035, + "loss": 1.132, + "step": 975 + }, + { + "epoch": 0.06772604260634238, + "grad_norm": 5.226365566253662, + "learning_rate": 0.00011283236994219655, + "loss": 1.1069, + "step": 976 + }, + { + "epoch": 0.06779543404343905, + "grad_norm": 7.36810302734375, + "learning_rate": 0.00011294797687861271, + "loss": 1.4516, + "step": 977 + }, + { + "epoch": 0.0678648254805357, + "grad_norm": 5.848696231842041, + "learning_rate": 0.00011306358381502891, + "loss": 1.2222, + "step": 978 + }, + { + "epoch": 0.06793421691763236, + "grad_norm": 5.521559715270996, + "learning_rate": 0.0001131791907514451, + "loss": 1.1027, + "step": 979 + }, + { + "epoch": 0.06800360835472903, + "grad_norm": 7.584116458892822, + "learning_rate": 0.00011329479768786128, + "loss": 1.4505, + "step": 980 + }, + { + "epoch": 0.06807299979182568, + "grad_norm": 8.583520889282227, + "learning_rate": 0.00011341040462427746, + "loss": 1.1281, + "step": 981 + }, + { + "epoch": 0.06814239122892235, + "grad_norm": 7.128651142120361, + "learning_rate": 0.00011352601156069364, + "loss": 1.2604, + "step": 982 + }, + { + "epoch": 0.06821178266601902, + "grad_norm": 5.544323921203613, + "learning_rate": 0.00011364161849710984, + "loss": 1.2203, + "step": 983 + }, + { + "epoch": 0.06828117410311567, + "grad_norm": 7.091053485870361, + "learning_rate": 0.00011375722543352603, + "loss": 1.3996, + "step": 984 + }, + { + "epoch": 0.06835056554021234, + "grad_norm": 5.865147113800049, + "learning_rate": 0.0001138728323699422, + "loss": 1.3751, + "step": 985 + }, + { + "epoch": 0.068419956977309, + "grad_norm": 6.108989238739014, + "learning_rate": 0.00011398843930635839, + "loss": 0.9212, + "step": 986 + }, + { + "epoch": 0.06848934841440567, + "grad_norm": 6.331236362457275, + "learning_rate": 0.00011410404624277457, + "loss": 1.488, + "step": 987 + }, + { + "epoch": 0.06855873985150232, + "grad_norm": 5.796266078948975, + "learning_rate": 0.00011421965317919076, + "loss": 1.1085, + "step": 988 + }, + { + "epoch": 0.06862813128859899, + "grad_norm": 7.398229122161865, + "learning_rate": 0.00011433526011560693, + "loss": 1.5113, + "step": 989 + }, + { + "epoch": 0.06869752272569565, + "grad_norm": 6.6239914894104, + "learning_rate": 0.00011445086705202312, + "loss": 1.0638, + "step": 990 + }, + { + "epoch": 0.0687669141627923, + "grad_norm": 4.6780877113342285, + "learning_rate": 0.00011456647398843932, + "loss": 1.1259, + "step": 991 + }, + { + "epoch": 0.06883630559988897, + "grad_norm": 5.243964195251465, + "learning_rate": 0.0001146820809248555, + "loss": 1.1372, + "step": 992 + }, + { + "epoch": 0.06890569703698564, + "grad_norm": 6.090360641479492, + "learning_rate": 0.00011479768786127168, + "loss": 1.3078, + "step": 993 + }, + { + "epoch": 0.0689750884740823, + "grad_norm": 5.9732136726379395, + "learning_rate": 0.00011491329479768786, + "loss": 1.5242, + "step": 994 + }, + { + "epoch": 0.06904447991117896, + "grad_norm": 5.694429874420166, + "learning_rate": 0.00011502890173410405, + "loss": 1.2545, + "step": 995 + }, + { + "epoch": 0.06911387134827562, + "grad_norm": 5.901613235473633, + "learning_rate": 0.00011514450867052025, + "loss": 1.0189, + "step": 996 + }, + { + "epoch": 0.06918326278537229, + "grad_norm": 6.447208881378174, + "learning_rate": 0.00011526011560693641, + "loss": 1.4103, + "step": 997 + }, + { + "epoch": 0.06925265422246894, + "grad_norm": 4.838980197906494, + "learning_rate": 0.00011537572254335261, + "loss": 0.9363, + "step": 998 + }, + { + "epoch": 0.06932204565956561, + "grad_norm": 8.36479663848877, + "learning_rate": 0.0001154913294797688, + "loss": 1.2856, + "step": 999 + }, + { + "epoch": 0.06939143709666228, + "grad_norm": 5.8553972244262695, + "learning_rate": 0.00011560693641618498, + "loss": 1.2768, + "step": 1000 + }, + { + "epoch": 0.06946082853375893, + "grad_norm": 5.653506278991699, + "learning_rate": 0.00011572254335260115, + "loss": 1.056, + "step": 1001 + }, + { + "epoch": 0.0695302199708556, + "grad_norm": 6.043652057647705, + "learning_rate": 0.00011583815028901734, + "loss": 1.139, + "step": 1002 + }, + { + "epoch": 0.06959961140795226, + "grad_norm": 6.355461120605469, + "learning_rate": 0.00011595375722543354, + "loss": 1.1128, + "step": 1003 + }, + { + "epoch": 0.06966900284504893, + "grad_norm": 6.642773628234863, + "learning_rate": 0.00011606936416184973, + "loss": 0.9631, + "step": 1004 + }, + { + "epoch": 0.06973839428214558, + "grad_norm": 5.841923236846924, + "learning_rate": 0.0001161849710982659, + "loss": 0.9381, + "step": 1005 + }, + { + "epoch": 0.06980778571924225, + "grad_norm": 5.095273494720459, + "learning_rate": 0.00011630057803468208, + "loss": 1.1212, + "step": 1006 + }, + { + "epoch": 0.06987717715633891, + "grad_norm": 5.4190473556518555, + "learning_rate": 0.00011641618497109827, + "loss": 1.1545, + "step": 1007 + }, + { + "epoch": 0.06994656859343557, + "grad_norm": 6.1723952293396, + "learning_rate": 0.00011653179190751447, + "loss": 1.3056, + "step": 1008 + }, + { + "epoch": 0.07001596003053223, + "grad_norm": 5.920086860656738, + "learning_rate": 0.00011664739884393063, + "loss": 1.2229, + "step": 1009 + }, + { + "epoch": 0.0700853514676289, + "grad_norm": 4.131924152374268, + "learning_rate": 0.00011676300578034683, + "loss": 1.0821, + "step": 1010 + }, + { + "epoch": 0.07015474290472555, + "grad_norm": 4.7337846755981445, + "learning_rate": 0.00011687861271676302, + "loss": 1.1048, + "step": 1011 + }, + { + "epoch": 0.07022413434182222, + "grad_norm": 17.999500274658203, + "learning_rate": 0.0001169942196531792, + "loss": 1.1571, + "step": 1012 + }, + { + "epoch": 0.07029352577891888, + "grad_norm": 5.9278669357299805, + "learning_rate": 0.00011710982658959537, + "loss": 1.0701, + "step": 1013 + }, + { + "epoch": 0.07036291721601555, + "grad_norm": 4.832433223724365, + "learning_rate": 0.00011722543352601156, + "loss": 1.1064, + "step": 1014 + }, + { + "epoch": 0.0704323086531122, + "grad_norm": 5.754904747009277, + "learning_rate": 0.00011734104046242775, + "loss": 1.0055, + "step": 1015 + }, + { + "epoch": 0.07050170009020887, + "grad_norm": 7.070765018463135, + "learning_rate": 0.00011745664739884395, + "loss": 1.3401, + "step": 1016 + }, + { + "epoch": 0.07057109152730554, + "grad_norm": 5.286227226257324, + "learning_rate": 0.0001175722543352601, + "loss": 1.244, + "step": 1017 + }, + { + "epoch": 0.07064048296440219, + "grad_norm": 6.596480846405029, + "learning_rate": 0.0001176878612716763, + "loss": 1.2669, + "step": 1018 + }, + { + "epoch": 0.07070987440149885, + "grad_norm": 5.768551826477051, + "learning_rate": 0.00011780346820809249, + "loss": 1.244, + "step": 1019 + }, + { + "epoch": 0.07077926583859552, + "grad_norm": 5.8772664070129395, + "learning_rate": 0.00011791907514450868, + "loss": 1.3252, + "step": 1020 + }, + { + "epoch": 0.07084865727569217, + "grad_norm": 6.883822441101074, + "learning_rate": 0.00011803468208092485, + "loss": 1.2247, + "step": 1021 + }, + { + "epoch": 0.07091804871278884, + "grad_norm": 6.437276840209961, + "learning_rate": 0.00011815028901734104, + "loss": 1.3164, + "step": 1022 + }, + { + "epoch": 0.0709874401498855, + "grad_norm": 7.746216773986816, + "learning_rate": 0.00011826589595375724, + "loss": 1.3257, + "step": 1023 + }, + { + "epoch": 0.07105683158698217, + "grad_norm": 8.569561004638672, + "learning_rate": 0.00011838150289017342, + "loss": 1.2439, + "step": 1024 + }, + { + "epoch": 0.07112622302407882, + "grad_norm": 7.461762428283691, + "learning_rate": 0.0001184971098265896, + "loss": 1.1733, + "step": 1025 + }, + { + "epoch": 0.07119561446117549, + "grad_norm": 6.488585948944092, + "learning_rate": 0.00011861271676300578, + "loss": 1.3196, + "step": 1026 + }, + { + "epoch": 0.07126500589827216, + "grad_norm": 5.738556385040283, + "learning_rate": 0.00011872832369942197, + "loss": 1.0697, + "step": 1027 + }, + { + "epoch": 0.07133439733536881, + "grad_norm": 6.058094024658203, + "learning_rate": 0.00011884393063583817, + "loss": 1.2155, + "step": 1028 + }, + { + "epoch": 0.07140378877246548, + "grad_norm": 7.3612961769104, + "learning_rate": 0.00011895953757225433, + "loss": 1.1194, + "step": 1029 + }, + { + "epoch": 0.07147318020956214, + "grad_norm": 5.656105995178223, + "learning_rate": 0.00011907514450867053, + "loss": 1.0703, + "step": 1030 + }, + { + "epoch": 0.07154257164665881, + "grad_norm": 5.416797637939453, + "learning_rate": 0.00011919075144508671, + "loss": 1.2869, + "step": 1031 + }, + { + "epoch": 0.07161196308375546, + "grad_norm": 5.914897441864014, + "learning_rate": 0.0001193063583815029, + "loss": 0.9367, + "step": 1032 + }, + { + "epoch": 0.07168135452085213, + "grad_norm": 5.095796585083008, + "learning_rate": 0.00011942196531791907, + "loss": 1.188, + "step": 1033 + }, + { + "epoch": 0.0717507459579488, + "grad_norm": 4.308776378631592, + "learning_rate": 0.00011953757225433526, + "loss": 1.0896, + "step": 1034 + }, + { + "epoch": 0.07182013739504545, + "grad_norm": 6.08712911605835, + "learning_rate": 0.00011965317919075146, + "loss": 1.1725, + "step": 1035 + }, + { + "epoch": 0.07188952883214211, + "grad_norm": 4.715702056884766, + "learning_rate": 0.00011976878612716764, + "loss": 1.1634, + "step": 1036 + }, + { + "epoch": 0.07195892026923878, + "grad_norm": 6.68136739730835, + "learning_rate": 0.00011988439306358382, + "loss": 1.3291, + "step": 1037 + }, + { + "epoch": 0.07202831170633543, + "grad_norm": 5.3571248054504395, + "learning_rate": 0.00012, + "loss": 1.0976, + "step": 1038 + }, + { + "epoch": 0.0720977031434321, + "grad_norm": 4.40831995010376, + "learning_rate": 0.00012011560693641619, + "loss": 0.8578, + "step": 1039 + }, + { + "epoch": 0.07216709458052877, + "grad_norm": 4.56652307510376, + "learning_rate": 0.00012023121387283239, + "loss": 0.96, + "step": 1040 + }, + { + "epoch": 0.07223648601762543, + "grad_norm": 8.150485038757324, + "learning_rate": 0.00012034682080924855, + "loss": 1.3709, + "step": 1041 + }, + { + "epoch": 0.07230587745472208, + "grad_norm": 6.0152363777160645, + "learning_rate": 0.00012046242774566473, + "loss": 1.43, + "step": 1042 + }, + { + "epoch": 0.07237526889181875, + "grad_norm": 6.368424892425537, + "learning_rate": 0.00012057803468208093, + "loss": 1.1386, + "step": 1043 + }, + { + "epoch": 0.07244466032891542, + "grad_norm": 4.709599018096924, + "learning_rate": 0.00012069364161849712, + "loss": 1.0278, + "step": 1044 + }, + { + "epoch": 0.07251405176601207, + "grad_norm": 8.32509994506836, + "learning_rate": 0.00012080924855491329, + "loss": 1.1717, + "step": 1045 + }, + { + "epoch": 0.07258344320310874, + "grad_norm": 4.847778797149658, + "learning_rate": 0.00012092485549132948, + "loss": 1.0874, + "step": 1046 + }, + { + "epoch": 0.0726528346402054, + "grad_norm": 6.149078369140625, + "learning_rate": 0.00012104046242774567, + "loss": 1.2013, + "step": 1047 + }, + { + "epoch": 0.07272222607730205, + "grad_norm": 7.525489807128906, + "learning_rate": 0.00012115606936416187, + "loss": 1.4716, + "step": 1048 + }, + { + "epoch": 0.07279161751439872, + "grad_norm": 6.0584564208984375, + "learning_rate": 0.00012127167630057802, + "loss": 1.2827, + "step": 1049 + }, + { + "epoch": 0.07286100895149539, + "grad_norm": 4.799071788787842, + "learning_rate": 0.00012138728323699422, + "loss": 1.0814, + "step": 1050 + }, + { + "epoch": 0.07293040038859205, + "grad_norm": 5.704986572265625, + "learning_rate": 0.00012150289017341041, + "loss": 1.0569, + "step": 1051 + }, + { + "epoch": 0.0729997918256887, + "grad_norm": 7.508641242980957, + "learning_rate": 0.0001216184971098266, + "loss": 1.3846, + "step": 1052 + }, + { + "epoch": 0.07306918326278537, + "grad_norm": 6.663384437561035, + "learning_rate": 0.00012173410404624277, + "loss": 1.3279, + "step": 1053 + }, + { + "epoch": 0.07313857469988204, + "grad_norm": 6.319398880004883, + "learning_rate": 0.00012184971098265896, + "loss": 1.1048, + "step": 1054 + }, + { + "epoch": 0.07320796613697869, + "grad_norm": 4.571174144744873, + "learning_rate": 0.00012196531791907516, + "loss": 1.0009, + "step": 1055 + }, + { + "epoch": 0.07327735757407536, + "grad_norm": 5.259437084197998, + "learning_rate": 0.00012208092485549134, + "loss": 0.9286, + "step": 1056 + }, + { + "epoch": 0.07334674901117202, + "grad_norm": 6.291734218597412, + "learning_rate": 0.0001221965317919075, + "loss": 1.1865, + "step": 1057 + }, + { + "epoch": 0.07341614044826868, + "grad_norm": 6.44913911819458, + "learning_rate": 0.0001223121387283237, + "loss": 1.2403, + "step": 1058 + }, + { + "epoch": 0.07348553188536534, + "grad_norm": 8.820833206176758, + "learning_rate": 0.0001224277456647399, + "loss": 1.3194, + "step": 1059 + }, + { + "epoch": 0.07355492332246201, + "grad_norm": 7.50294828414917, + "learning_rate": 0.00012254335260115609, + "loss": 1.3658, + "step": 1060 + }, + { + "epoch": 0.07362431475955868, + "grad_norm": 6.8616042137146, + "learning_rate": 0.00012265895953757227, + "loss": 1.4236, + "step": 1061 + }, + { + "epoch": 0.07369370619665533, + "grad_norm": 9.249374389648438, + "learning_rate": 0.00012277456647398843, + "loss": 1.52, + "step": 1062 + }, + { + "epoch": 0.073763097633752, + "grad_norm": 9.66752815246582, + "learning_rate": 0.00012289017341040462, + "loss": 1.5902, + "step": 1063 + }, + { + "epoch": 0.07383248907084866, + "grad_norm": 5.629762649536133, + "learning_rate": 0.00012300578034682083, + "loss": 1.046, + "step": 1064 + }, + { + "epoch": 0.07390188050794531, + "grad_norm": 8.133699417114258, + "learning_rate": 0.00012312138728323702, + "loss": 0.8659, + "step": 1065 + }, + { + "epoch": 0.07397127194504198, + "grad_norm": 6.97988224029541, + "learning_rate": 0.00012323699421965318, + "loss": 1.1732, + "step": 1066 + }, + { + "epoch": 0.07404066338213865, + "grad_norm": 6.889465808868408, + "learning_rate": 0.00012335260115606936, + "loss": 1.2272, + "step": 1067 + }, + { + "epoch": 0.07411005481923531, + "grad_norm": 6.226183891296387, + "learning_rate": 0.00012346820809248555, + "loss": 1.2733, + "step": 1068 + }, + { + "epoch": 0.07417944625633197, + "grad_norm": 7.89928674697876, + "learning_rate": 0.00012358381502890176, + "loss": 1.2348, + "step": 1069 + }, + { + "epoch": 0.07424883769342863, + "grad_norm": 6.029826641082764, + "learning_rate": 0.00012369942196531792, + "loss": 1.2755, + "step": 1070 + }, + { + "epoch": 0.0743182291305253, + "grad_norm": 6.348960876464844, + "learning_rate": 0.0001238150289017341, + "loss": 1.1879, + "step": 1071 + }, + { + "epoch": 0.07438762056762195, + "grad_norm": 5.8169684410095215, + "learning_rate": 0.0001239306358381503, + "loss": 1.316, + "step": 1072 + }, + { + "epoch": 0.07445701200471862, + "grad_norm": 6.206691265106201, + "learning_rate": 0.00012404624277456648, + "loss": 1.3042, + "step": 1073 + }, + { + "epoch": 0.07452640344181528, + "grad_norm": 4.98428201675415, + "learning_rate": 0.00012416184971098267, + "loss": 0.9908, + "step": 1074 + }, + { + "epoch": 0.07459579487891194, + "grad_norm": 5.3473124504089355, + "learning_rate": 0.00012427745664739885, + "loss": 1.1588, + "step": 1075 + }, + { + "epoch": 0.0746651863160086, + "grad_norm": 5.5239386558532715, + "learning_rate": 0.00012439306358381504, + "loss": 1.1844, + "step": 1076 + }, + { + "epoch": 0.07473457775310527, + "grad_norm": 4.3472580909729, + "learning_rate": 0.00012450867052023122, + "loss": 0.9298, + "step": 1077 + }, + { + "epoch": 0.07480396919020194, + "grad_norm": 6.8451457023620605, + "learning_rate": 0.0001246242774566474, + "loss": 1.2584, + "step": 1078 + }, + { + "epoch": 0.07487336062729859, + "grad_norm": 4.792855739593506, + "learning_rate": 0.0001247398843930636, + "loss": 1.0547, + "step": 1079 + }, + { + "epoch": 0.07494275206439525, + "grad_norm": 5.904514789581299, + "learning_rate": 0.00012485549132947978, + "loss": 1.3926, + "step": 1080 + }, + { + "epoch": 0.07501214350149192, + "grad_norm": 4.372405529022217, + "learning_rate": 0.00012497109826589597, + "loss": 0.9641, + "step": 1081 + }, + { + "epoch": 0.07508153493858857, + "grad_norm": 6.949167728424072, + "learning_rate": 0.00012508670520231213, + "loss": 1.3846, + "step": 1082 + }, + { + "epoch": 0.07515092637568524, + "grad_norm": 5.649738311767578, + "learning_rate": 0.00012520231213872831, + "loss": 1.2091, + "step": 1083 + }, + { + "epoch": 0.0752203178127819, + "grad_norm": 5.193731307983398, + "learning_rate": 0.00012531791907514453, + "loss": 1.2221, + "step": 1084 + }, + { + "epoch": 0.07528970924987856, + "grad_norm": 5.981949806213379, + "learning_rate": 0.00012543352601156071, + "loss": 1.2548, + "step": 1085 + }, + { + "epoch": 0.07535910068697523, + "grad_norm": 5.1077752113342285, + "learning_rate": 0.00012554913294797687, + "loss": 1.0814, + "step": 1086 + }, + { + "epoch": 0.07542849212407189, + "grad_norm": 6.234776973724365, + "learning_rate": 0.00012566473988439306, + "loss": 1.0855, + "step": 1087 + }, + { + "epoch": 0.07549788356116856, + "grad_norm": 4.877274513244629, + "learning_rate": 0.00012578034682080925, + "loss": 1.1189, + "step": 1088 + }, + { + "epoch": 0.07556727499826521, + "grad_norm": 5.238943576812744, + "learning_rate": 0.00012589595375722546, + "loss": 1.0811, + "step": 1089 + }, + { + "epoch": 0.07563666643536188, + "grad_norm": 6.292196273803711, + "learning_rate": 0.00012601156069364162, + "loss": 1.3777, + "step": 1090 + }, + { + "epoch": 0.07570605787245854, + "grad_norm": 6.027429580688477, + "learning_rate": 0.0001261271676300578, + "loss": 0.9914, + "step": 1091 + }, + { + "epoch": 0.0757754493095552, + "grad_norm": 5.711855411529541, + "learning_rate": 0.000126242774566474, + "loss": 1.205, + "step": 1092 + }, + { + "epoch": 0.07584484074665186, + "grad_norm": 4.678946018218994, + "learning_rate": 0.00012635838150289018, + "loss": 1.0796, + "step": 1093 + }, + { + "epoch": 0.07591423218374853, + "grad_norm": 5.935581684112549, + "learning_rate": 0.00012647398843930636, + "loss": 1.3916, + "step": 1094 + }, + { + "epoch": 0.07598362362084518, + "grad_norm": 5.357654571533203, + "learning_rate": 0.00012658959537572255, + "loss": 0.9782, + "step": 1095 + }, + { + "epoch": 0.07605301505794185, + "grad_norm": 6.532570838928223, + "learning_rate": 0.00012670520231213874, + "loss": 1.0903, + "step": 1096 + }, + { + "epoch": 0.07612240649503851, + "grad_norm": 5.567784309387207, + "learning_rate": 0.00012682080924855492, + "loss": 1.2076, + "step": 1097 + }, + { + "epoch": 0.07619179793213518, + "grad_norm": 5.363899230957031, + "learning_rate": 0.0001269364161849711, + "loss": 1.293, + "step": 1098 + }, + { + "epoch": 0.07626118936923183, + "grad_norm": 5.091403961181641, + "learning_rate": 0.0001270520231213873, + "loss": 1.253, + "step": 1099 + }, + { + "epoch": 0.0763305808063285, + "grad_norm": 5.9075446128845215, + "learning_rate": 0.00012716763005780348, + "loss": 1.115, + "step": 1100 + }, + { + "epoch": 0.07639997224342517, + "grad_norm": 4.707104682922363, + "learning_rate": 0.00012728323699421967, + "loss": 1.2619, + "step": 1101 + }, + { + "epoch": 0.07646936368052182, + "grad_norm": 5.272624492645264, + "learning_rate": 0.00012739884393063583, + "loss": 1.0337, + "step": 1102 + }, + { + "epoch": 0.07653875511761848, + "grad_norm": 5.192031383514404, + "learning_rate": 0.00012751445086705204, + "loss": 1.1366, + "step": 1103 + }, + { + "epoch": 0.07660814655471515, + "grad_norm": 4.909294128417969, + "learning_rate": 0.00012763005780346823, + "loss": 0.9822, + "step": 1104 + }, + { + "epoch": 0.07667753799181182, + "grad_norm": 6.134764194488525, + "learning_rate": 0.0001277456647398844, + "loss": 1.2434, + "step": 1105 + }, + { + "epoch": 0.07674692942890847, + "grad_norm": 5.286313533782959, + "learning_rate": 0.00012786127167630057, + "loss": 1.4543, + "step": 1106 + }, + { + "epoch": 0.07681632086600514, + "grad_norm": 6.123839378356934, + "learning_rate": 0.00012797687861271676, + "loss": 1.2653, + "step": 1107 + }, + { + "epoch": 0.0768857123031018, + "grad_norm": 5.506866931915283, + "learning_rate": 0.00012809248554913297, + "loss": 1.3808, + "step": 1108 + }, + { + "epoch": 0.07695510374019846, + "grad_norm": 4.6750712394714355, + "learning_rate": 0.00012820809248554916, + "loss": 1.0003, + "step": 1109 + }, + { + "epoch": 0.07702449517729512, + "grad_norm": 5.431289196014404, + "learning_rate": 0.00012832369942196532, + "loss": 0.9083, + "step": 1110 + }, + { + "epoch": 0.07709388661439179, + "grad_norm": 6.252151012420654, + "learning_rate": 0.0001284393063583815, + "loss": 1.3507, + "step": 1111 + }, + { + "epoch": 0.07716327805148844, + "grad_norm": 4.965311050415039, + "learning_rate": 0.0001285549132947977, + "loss": 1.1908, + "step": 1112 + }, + { + "epoch": 0.0772326694885851, + "grad_norm": 7.224578380584717, + "learning_rate": 0.00012867052023121387, + "loss": 1.4403, + "step": 1113 + }, + { + "epoch": 0.07730206092568177, + "grad_norm": 5.822369575500488, + "learning_rate": 0.00012878612716763006, + "loss": 1.1393, + "step": 1114 + }, + { + "epoch": 0.07737145236277844, + "grad_norm": 6.875157833099365, + "learning_rate": 0.00012890173410404625, + "loss": 0.8504, + "step": 1115 + }, + { + "epoch": 0.07744084379987509, + "grad_norm": 4.633720397949219, + "learning_rate": 0.00012901734104046243, + "loss": 1.0245, + "step": 1116 + }, + { + "epoch": 0.07751023523697176, + "grad_norm": 5.585602283477783, + "learning_rate": 0.00012913294797687862, + "loss": 0.9513, + "step": 1117 + }, + { + "epoch": 0.07757962667406842, + "grad_norm": 5.0478997230529785, + "learning_rate": 0.0001292485549132948, + "loss": 1.0029, + "step": 1118 + }, + { + "epoch": 0.07764901811116508, + "grad_norm": 5.993742942810059, + "learning_rate": 0.000129364161849711, + "loss": 1.37, + "step": 1119 + }, + { + "epoch": 0.07771840954826174, + "grad_norm": 5.5457329750061035, + "learning_rate": 0.00012947976878612718, + "loss": 0.9084, + "step": 1120 + }, + { + "epoch": 0.07778780098535841, + "grad_norm": 4.932546138763428, + "learning_rate": 0.00012959537572254336, + "loss": 0.9448, + "step": 1121 + }, + { + "epoch": 0.07785719242245506, + "grad_norm": 13.737611770629883, + "learning_rate": 0.00012971098265895952, + "loss": 0.9673, + "step": 1122 + }, + { + "epoch": 0.07792658385955173, + "grad_norm": 6.109320640563965, + "learning_rate": 0.00012982658959537574, + "loss": 1.2915, + "step": 1123 + }, + { + "epoch": 0.0779959752966484, + "grad_norm": 5.631484508514404, + "learning_rate": 0.00012994219653179192, + "loss": 1.3823, + "step": 1124 + }, + { + "epoch": 0.07806536673374506, + "grad_norm": 5.402753829956055, + "learning_rate": 0.0001300578034682081, + "loss": 1.1908, + "step": 1125 + }, + { + "epoch": 0.07813475817084171, + "grad_norm": 6.7480316162109375, + "learning_rate": 0.00013017341040462427, + "loss": 1.0356, + "step": 1126 + }, + { + "epoch": 0.07820414960793838, + "grad_norm": 6.685706615447998, + "learning_rate": 0.00013028901734104045, + "loss": 1.3129, + "step": 1127 + }, + { + "epoch": 0.07827354104503505, + "grad_norm": 5.267488479614258, + "learning_rate": 0.00013040462427745667, + "loss": 0.9436, + "step": 1128 + }, + { + "epoch": 0.0783429324821317, + "grad_norm": 6.659542083740234, + "learning_rate": 0.00013052023121387285, + "loss": 1.1218, + "step": 1129 + }, + { + "epoch": 0.07841232391922837, + "grad_norm": 8.679286003112793, + "learning_rate": 0.000130635838150289, + "loss": 0.9406, + "step": 1130 + }, + { + "epoch": 0.07848171535632503, + "grad_norm": 6.3278656005859375, + "learning_rate": 0.0001307514450867052, + "loss": 1.4415, + "step": 1131 + }, + { + "epoch": 0.07855110679342169, + "grad_norm": 6.188020706176758, + "learning_rate": 0.00013086705202312139, + "loss": 1.3472, + "step": 1132 + }, + { + "epoch": 0.07862049823051835, + "grad_norm": 6.129924297332764, + "learning_rate": 0.0001309826589595376, + "loss": 1.2142, + "step": 1133 + }, + { + "epoch": 0.07868988966761502, + "grad_norm": 5.7158660888671875, + "learning_rate": 0.00013109826589595376, + "loss": 1.2194, + "step": 1134 + }, + { + "epoch": 0.07875928110471168, + "grad_norm": 7.711948394775391, + "learning_rate": 0.00013121387283236994, + "loss": 1.2537, + "step": 1135 + }, + { + "epoch": 0.07882867254180834, + "grad_norm": 4.738063812255859, + "learning_rate": 0.00013132947976878613, + "loss": 1.1639, + "step": 1136 + }, + { + "epoch": 0.078898063978905, + "grad_norm": 5.1506757736206055, + "learning_rate": 0.00013144508670520232, + "loss": 1.0862, + "step": 1137 + }, + { + "epoch": 0.07896745541600167, + "grad_norm": 5.251074314117432, + "learning_rate": 0.0001315606936416185, + "loss": 1.1378, + "step": 1138 + }, + { + "epoch": 0.07903684685309832, + "grad_norm": 5.092872619628906, + "learning_rate": 0.0001316763005780347, + "loss": 1.241, + "step": 1139 + }, + { + "epoch": 0.07910623829019499, + "grad_norm": 5.25932502746582, + "learning_rate": 0.00013179190751445087, + "loss": 1.3971, + "step": 1140 + }, + { + "epoch": 0.07917562972729165, + "grad_norm": 5.240997791290283, + "learning_rate": 0.00013190751445086706, + "loss": 1.2234, + "step": 1141 + }, + { + "epoch": 0.07924502116438832, + "grad_norm": 4.29115629196167, + "learning_rate": 0.00013202312138728322, + "loss": 1.0889, + "step": 1142 + }, + { + "epoch": 0.07931441260148497, + "grad_norm": 5.428481578826904, + "learning_rate": 0.00013213872832369943, + "loss": 1.0803, + "step": 1143 + }, + { + "epoch": 0.07938380403858164, + "grad_norm": 4.950887203216553, + "learning_rate": 0.00013225433526011562, + "loss": 1.3452, + "step": 1144 + }, + { + "epoch": 0.0794531954756783, + "grad_norm": 5.56185245513916, + "learning_rate": 0.0001323699421965318, + "loss": 1.3095, + "step": 1145 + }, + { + "epoch": 0.07952258691277496, + "grad_norm": 5.3682355880737305, + "learning_rate": 0.00013248554913294797, + "loss": 1.1845, + "step": 1146 + }, + { + "epoch": 0.07959197834987163, + "grad_norm": 5.524921417236328, + "learning_rate": 0.00013260115606936415, + "loss": 1.2841, + "step": 1147 + }, + { + "epoch": 0.07966136978696829, + "grad_norm": 5.585562229156494, + "learning_rate": 0.00013271676300578036, + "loss": 1.2403, + "step": 1148 + }, + { + "epoch": 0.07973076122406494, + "grad_norm": 5.1067795753479, + "learning_rate": 0.00013283236994219655, + "loss": 1.1607, + "step": 1149 + }, + { + "epoch": 0.07980015266116161, + "grad_norm": 4.398856163024902, + "learning_rate": 0.0001329479768786127, + "loss": 0.9719, + "step": 1150 + }, + { + "epoch": 0.07986954409825828, + "grad_norm": 3.9637410640716553, + "learning_rate": 0.0001330635838150289, + "loss": 1.0205, + "step": 1151 + }, + { + "epoch": 0.07993893553535494, + "grad_norm": 5.573578834533691, + "learning_rate": 0.00013317919075144508, + "loss": 1.1448, + "step": 1152 + }, + { + "epoch": 0.0800083269724516, + "grad_norm": 5.16171932220459, + "learning_rate": 0.0001332947976878613, + "loss": 1.2539, + "step": 1153 + }, + { + "epoch": 0.08007771840954826, + "grad_norm": 4.529061794281006, + "learning_rate": 0.00013341040462427748, + "loss": 1.0714, + "step": 1154 + }, + { + "epoch": 0.08014710984664493, + "grad_norm": 6.770246982574463, + "learning_rate": 0.00013352601156069364, + "loss": 1.1309, + "step": 1155 + }, + { + "epoch": 0.08021650128374158, + "grad_norm": 4.577332019805908, + "learning_rate": 0.00013364161849710983, + "loss": 1.212, + "step": 1156 + }, + { + "epoch": 0.08028589272083825, + "grad_norm": 4.659226417541504, + "learning_rate": 0.000133757225433526, + "loss": 1.0539, + "step": 1157 + }, + { + "epoch": 0.08035528415793491, + "grad_norm": 5.242716312408447, + "learning_rate": 0.00013387283236994223, + "loss": 1.0319, + "step": 1158 + }, + { + "epoch": 0.08042467559503157, + "grad_norm": 5.283103942871094, + "learning_rate": 0.00013398843930635839, + "loss": 1.0528, + "step": 1159 + }, + { + "epoch": 0.08049406703212823, + "grad_norm": 5.800968170166016, + "learning_rate": 0.00013410404624277457, + "loss": 1.095, + "step": 1160 + }, + { + "epoch": 0.0805634584692249, + "grad_norm": 6.503697872161865, + "learning_rate": 0.00013421965317919076, + "loss": 1.1197, + "step": 1161 + }, + { + "epoch": 0.08063284990632157, + "grad_norm": 4.625351905822754, + "learning_rate": 0.00013433526011560694, + "loss": 1.121, + "step": 1162 + }, + { + "epoch": 0.08070224134341822, + "grad_norm": 5.26969575881958, + "learning_rate": 0.00013445086705202313, + "loss": 1.1085, + "step": 1163 + }, + { + "epoch": 0.08077163278051488, + "grad_norm": 6.609395503997803, + "learning_rate": 0.00013456647398843932, + "loss": 1.4335, + "step": 1164 + }, + { + "epoch": 0.08084102421761155, + "grad_norm": 5.860116004943848, + "learning_rate": 0.0001346820809248555, + "loss": 1.2011, + "step": 1165 + }, + { + "epoch": 0.0809104156547082, + "grad_norm": 5.071766376495361, + "learning_rate": 0.0001347976878612717, + "loss": 1.1014, + "step": 1166 + }, + { + "epoch": 0.08097980709180487, + "grad_norm": 4.056862831115723, + "learning_rate": 0.00013491329479768785, + "loss": 0.9448, + "step": 1167 + }, + { + "epoch": 0.08104919852890154, + "grad_norm": 6.830517292022705, + "learning_rate": 0.00013502890173410406, + "loss": 1.1005, + "step": 1168 + }, + { + "epoch": 0.08111858996599819, + "grad_norm": 5.884715557098389, + "learning_rate": 0.00013514450867052025, + "loss": 1.1813, + "step": 1169 + }, + { + "epoch": 0.08118798140309486, + "grad_norm": 4.9995880126953125, + "learning_rate": 0.00013526011560693643, + "loss": 1.1326, + "step": 1170 + }, + { + "epoch": 0.08125737284019152, + "grad_norm": 4.691529273986816, + "learning_rate": 0.0001353757225433526, + "loss": 1.158, + "step": 1171 + }, + { + "epoch": 0.08132676427728819, + "grad_norm": 4.701956272125244, + "learning_rate": 0.00013549132947976878, + "loss": 1.3282, + "step": 1172 + }, + { + "epoch": 0.08139615571438484, + "grad_norm": 4.879031181335449, + "learning_rate": 0.000135606936416185, + "loss": 1.1492, + "step": 1173 + }, + { + "epoch": 0.08146554715148151, + "grad_norm": 6.221784591674805, + "learning_rate": 0.00013572254335260118, + "loss": 1.2134, + "step": 1174 + }, + { + "epoch": 0.08153493858857817, + "grad_norm": 6.411861896514893, + "learning_rate": 0.00013583815028901734, + "loss": 1.372, + "step": 1175 + }, + { + "epoch": 0.08160433002567483, + "grad_norm": 5.237519264221191, + "learning_rate": 0.00013595375722543352, + "loss": 1.0927, + "step": 1176 + }, + { + "epoch": 0.08167372146277149, + "grad_norm": 6.8781657218933105, + "learning_rate": 0.0001360693641618497, + "loss": 1.2598, + "step": 1177 + }, + { + "epoch": 0.08174311289986816, + "grad_norm": 4.836091041564941, + "learning_rate": 0.00013618497109826592, + "loss": 1.0636, + "step": 1178 + }, + { + "epoch": 0.08181250433696483, + "grad_norm": 5.9124979972839355, + "learning_rate": 0.00013630057803468208, + "loss": 1.1022, + "step": 1179 + }, + { + "epoch": 0.08188189577406148, + "grad_norm": 5.265321731567383, + "learning_rate": 0.00013641618497109827, + "loss": 1.0948, + "step": 1180 + }, + { + "epoch": 0.08195128721115814, + "grad_norm": 3.939805746078491, + "learning_rate": 0.00013653179190751446, + "loss": 1.1444, + "step": 1181 + }, + { + "epoch": 0.08202067864825481, + "grad_norm": 5.3233466148376465, + "learning_rate": 0.00013664739884393064, + "loss": 1.194, + "step": 1182 + }, + { + "epoch": 0.08209007008535146, + "grad_norm": 5.296302795410156, + "learning_rate": 0.00013676300578034683, + "loss": 1.3406, + "step": 1183 + }, + { + "epoch": 0.08215946152244813, + "grad_norm": 5.615617752075195, + "learning_rate": 0.00013687861271676301, + "loss": 1.1269, + "step": 1184 + }, + { + "epoch": 0.0822288529595448, + "grad_norm": 5.529454708099365, + "learning_rate": 0.0001369942196531792, + "loss": 1.046, + "step": 1185 + }, + { + "epoch": 0.08229824439664145, + "grad_norm": 5.6077799797058105, + "learning_rate": 0.0001371098265895954, + "loss": 1.0513, + "step": 1186 + }, + { + "epoch": 0.08236763583373811, + "grad_norm": 4.979215145111084, + "learning_rate": 0.00013722543352601157, + "loss": 1.1971, + "step": 1187 + }, + { + "epoch": 0.08243702727083478, + "grad_norm": 6.781510353088379, + "learning_rate": 0.00013734104046242776, + "loss": 1.4731, + "step": 1188 + }, + { + "epoch": 0.08250641870793145, + "grad_norm": 6.495598793029785, + "learning_rate": 0.00013745664739884395, + "loss": 1.1705, + "step": 1189 + }, + { + "epoch": 0.0825758101450281, + "grad_norm": 10.507830619812012, + "learning_rate": 0.00013757225433526013, + "loss": 1.2513, + "step": 1190 + }, + { + "epoch": 0.08264520158212477, + "grad_norm": 5.6850385665893555, + "learning_rate": 0.0001376878612716763, + "loss": 0.973, + "step": 1191 + }, + { + "epoch": 0.08271459301922143, + "grad_norm": 5.8427605628967285, + "learning_rate": 0.00013780346820809248, + "loss": 1.1617, + "step": 1192 + }, + { + "epoch": 0.08278398445631809, + "grad_norm": 4.8043413162231445, + "learning_rate": 0.0001379190751445087, + "loss": 1.1148, + "step": 1193 + }, + { + "epoch": 0.08285337589341475, + "grad_norm": 5.175867080688477, + "learning_rate": 0.00013803468208092488, + "loss": 1.1995, + "step": 1194 + }, + { + "epoch": 0.08292276733051142, + "grad_norm": 6.54948091506958, + "learning_rate": 0.00013815028901734104, + "loss": 1.3853, + "step": 1195 + }, + { + "epoch": 0.08299215876760807, + "grad_norm": 7.669018268585205, + "learning_rate": 0.00013826589595375722, + "loss": 1.142, + "step": 1196 + }, + { + "epoch": 0.08306155020470474, + "grad_norm": 6.167696475982666, + "learning_rate": 0.0001383815028901734, + "loss": 1.2212, + "step": 1197 + }, + { + "epoch": 0.0831309416418014, + "grad_norm": 4.956823825836182, + "learning_rate": 0.00013849710982658962, + "loss": 1.1854, + "step": 1198 + }, + { + "epoch": 0.08320033307889807, + "grad_norm": 4.690191745758057, + "learning_rate": 0.00013861271676300578, + "loss": 0.9667, + "step": 1199 + }, + { + "epoch": 0.08326972451599472, + "grad_norm": 4.800649166107178, + "learning_rate": 0.00013872832369942197, + "loss": 1.0079, + "step": 1200 + }, + { + "epoch": 0.08333911595309139, + "grad_norm": 4.192370891571045, + "learning_rate": 0.00013884393063583815, + "loss": 0.8117, + "step": 1201 + }, + { + "epoch": 0.08340850739018806, + "grad_norm": 4.6055216789245605, + "learning_rate": 0.00013895953757225434, + "loss": 1.1428, + "step": 1202 + }, + { + "epoch": 0.08347789882728471, + "grad_norm": 5.588587760925293, + "learning_rate": 0.00013907514450867053, + "loss": 1.3894, + "step": 1203 + }, + { + "epoch": 0.08354729026438137, + "grad_norm": 5.038575649261475, + "learning_rate": 0.0001391907514450867, + "loss": 1.4149, + "step": 1204 + }, + { + "epoch": 0.08361668170147804, + "grad_norm": 4.6054229736328125, + "learning_rate": 0.0001393063583815029, + "loss": 1.2591, + "step": 1205 + }, + { + "epoch": 0.0836860731385747, + "grad_norm": 6.086676597595215, + "learning_rate": 0.00013942196531791908, + "loss": 1.2544, + "step": 1206 + }, + { + "epoch": 0.08375546457567136, + "grad_norm": 5.383978366851807, + "learning_rate": 0.00013953757225433527, + "loss": 1.5034, + "step": 1207 + }, + { + "epoch": 0.08382485601276803, + "grad_norm": 4.268815517425537, + "learning_rate": 0.00013965317919075146, + "loss": 0.8815, + "step": 1208 + }, + { + "epoch": 0.08389424744986469, + "grad_norm": 4.658596992492676, + "learning_rate": 0.00013976878612716764, + "loss": 1.2, + "step": 1209 + }, + { + "epoch": 0.08396363888696134, + "grad_norm": 7.548584461212158, + "learning_rate": 0.00013988439306358383, + "loss": 1.6597, + "step": 1210 + }, + { + "epoch": 0.08403303032405801, + "grad_norm": 4.52902364730835, + "learning_rate": 0.00014, + "loss": 0.8499, + "step": 1211 + }, + { + "epoch": 0.08410242176115468, + "grad_norm": 4.539451599121094, + "learning_rate": 0.0001401156069364162, + "loss": 1.0762, + "step": 1212 + }, + { + "epoch": 0.08417181319825133, + "grad_norm": 4.498000144958496, + "learning_rate": 0.0001402312138728324, + "loss": 1.266, + "step": 1213 + }, + { + "epoch": 0.084241204635348, + "grad_norm": 3.7410802841186523, + "learning_rate": 0.00014034682080924857, + "loss": 0.9616, + "step": 1214 + }, + { + "epoch": 0.08431059607244466, + "grad_norm": 4.250897407531738, + "learning_rate": 0.00014046242774566473, + "loss": 1.1083, + "step": 1215 + }, + { + "epoch": 0.08437998750954133, + "grad_norm": 4.216724395751953, + "learning_rate": 0.00014057803468208092, + "loss": 1.1426, + "step": 1216 + }, + { + "epoch": 0.08444937894663798, + "grad_norm": 8.423447608947754, + "learning_rate": 0.0001406936416184971, + "loss": 1.1917, + "step": 1217 + }, + { + "epoch": 0.08451877038373465, + "grad_norm": 4.543856143951416, + "learning_rate": 0.00014080924855491332, + "loss": 1.142, + "step": 1218 + }, + { + "epoch": 0.08458816182083131, + "grad_norm": 4.766380786895752, + "learning_rate": 0.00014092485549132948, + "loss": 1.0828, + "step": 1219 + }, + { + "epoch": 0.08465755325792797, + "grad_norm": 5.2711944580078125, + "learning_rate": 0.00014104046242774566, + "loss": 1.1701, + "step": 1220 + }, + { + "epoch": 0.08472694469502463, + "grad_norm": 5.6436448097229, + "learning_rate": 0.00014115606936416185, + "loss": 1.3268, + "step": 1221 + }, + { + "epoch": 0.0847963361321213, + "grad_norm": 5.304614067077637, + "learning_rate": 0.00014127167630057804, + "loss": 1.1102, + "step": 1222 + }, + { + "epoch": 0.08486572756921795, + "grad_norm": 5.684450626373291, + "learning_rate": 0.00014138728323699422, + "loss": 1.3629, + "step": 1223 + }, + { + "epoch": 0.08493511900631462, + "grad_norm": 6.332671165466309, + "learning_rate": 0.0001415028901734104, + "loss": 1.251, + "step": 1224 + }, + { + "epoch": 0.08500451044341129, + "grad_norm": 4.472497463226318, + "learning_rate": 0.0001416184971098266, + "loss": 0.9316, + "step": 1225 + }, + { + "epoch": 0.08507390188050795, + "grad_norm": 16.165300369262695, + "learning_rate": 0.00014173410404624278, + "loss": 1.1147, + "step": 1226 + }, + { + "epoch": 0.0851432933176046, + "grad_norm": 4.571699142456055, + "learning_rate": 0.00014184971098265897, + "loss": 1.0428, + "step": 1227 + }, + { + "epoch": 0.08521268475470127, + "grad_norm": 5.272360801696777, + "learning_rate": 0.00014196531791907515, + "loss": 1.1623, + "step": 1228 + }, + { + "epoch": 0.08528207619179794, + "grad_norm": 5.535228252410889, + "learning_rate": 0.00014208092485549134, + "loss": 1.3242, + "step": 1229 + }, + { + "epoch": 0.08535146762889459, + "grad_norm": 5.13378381729126, + "learning_rate": 0.00014219653179190753, + "loss": 1.2546, + "step": 1230 + }, + { + "epoch": 0.08542085906599126, + "grad_norm": 4.343364238739014, + "learning_rate": 0.00014231213872832368, + "loss": 1.0075, + "step": 1231 + }, + { + "epoch": 0.08549025050308792, + "grad_norm": 5.719141960144043, + "learning_rate": 0.0001424277456647399, + "loss": 1.2824, + "step": 1232 + }, + { + "epoch": 0.08555964194018457, + "grad_norm": 7.381421089172363, + "learning_rate": 0.00014254335260115608, + "loss": 1.4378, + "step": 1233 + }, + { + "epoch": 0.08562903337728124, + "grad_norm": 4.885416507720947, + "learning_rate": 0.00014265895953757227, + "loss": 1.2418, + "step": 1234 + }, + { + "epoch": 0.08569842481437791, + "grad_norm": 4.9622907638549805, + "learning_rate": 0.00014277456647398843, + "loss": 1.0162, + "step": 1235 + }, + { + "epoch": 0.08576781625147457, + "grad_norm": 5.834571361541748, + "learning_rate": 0.00014289017341040462, + "loss": 1.0434, + "step": 1236 + }, + { + "epoch": 0.08583720768857123, + "grad_norm": 5.627358436584473, + "learning_rate": 0.00014300578034682083, + "loss": 1.0502, + "step": 1237 + }, + { + "epoch": 0.08590659912566789, + "grad_norm": 4.330741882324219, + "learning_rate": 0.00014312138728323702, + "loss": 1.0453, + "step": 1238 + }, + { + "epoch": 0.08597599056276456, + "grad_norm": 4.633363723754883, + "learning_rate": 0.00014323699421965317, + "loss": 0.9668, + "step": 1239 + }, + { + "epoch": 0.08604538199986121, + "grad_norm": 7.050557613372803, + "learning_rate": 0.00014335260115606936, + "loss": 1.405, + "step": 1240 + }, + { + "epoch": 0.08611477343695788, + "grad_norm": 5.76857328414917, + "learning_rate": 0.00014346820809248555, + "loss": 1.266, + "step": 1241 + }, + { + "epoch": 0.08618416487405454, + "grad_norm": 6.354617118835449, + "learning_rate": 0.00014358381502890176, + "loss": 1.3154, + "step": 1242 + }, + { + "epoch": 0.0862535563111512, + "grad_norm": 5.939563751220703, + "learning_rate": 0.00014369942196531792, + "loss": 1.2082, + "step": 1243 + }, + { + "epoch": 0.08632294774824786, + "grad_norm": 5.167966842651367, + "learning_rate": 0.0001438150289017341, + "loss": 1.1758, + "step": 1244 + }, + { + "epoch": 0.08639233918534453, + "grad_norm": 7.0755615234375, + "learning_rate": 0.0001439306358381503, + "loss": 1.2071, + "step": 1245 + }, + { + "epoch": 0.0864617306224412, + "grad_norm": 4.901465892791748, + "learning_rate": 0.00014404624277456648, + "loss": 1.1499, + "step": 1246 + }, + { + "epoch": 0.08653112205953785, + "grad_norm": 5.6904296875, + "learning_rate": 0.00014416184971098266, + "loss": 1.5357, + "step": 1247 + }, + { + "epoch": 0.08660051349663452, + "grad_norm": 5.030852317810059, + "learning_rate": 0.00014427745664739885, + "loss": 1.3139, + "step": 1248 + }, + { + "epoch": 0.08666990493373118, + "grad_norm": 7.034609794616699, + "learning_rate": 0.00014439306358381504, + "loss": 1.1036, + "step": 1249 + }, + { + "epoch": 0.08673929637082783, + "grad_norm": 4.375635147094727, + "learning_rate": 0.00014450867052023122, + "loss": 0.9336, + "step": 1250 + }, + { + "epoch": 0.0868086878079245, + "grad_norm": 5.944920539855957, + "learning_rate": 0.0001446242774566474, + "loss": 1.1269, + "step": 1251 + }, + { + "epoch": 0.08687807924502117, + "grad_norm": 5.073287487030029, + "learning_rate": 0.0001447398843930636, + "loss": 1.0187, + "step": 1252 + }, + { + "epoch": 0.08694747068211783, + "grad_norm": 5.985785961151123, + "learning_rate": 0.00014485549132947978, + "loss": 1.2732, + "step": 1253 + }, + { + "epoch": 0.08701686211921449, + "grad_norm": 5.133194446563721, + "learning_rate": 0.00014497109826589597, + "loss": 1.1462, + "step": 1254 + }, + { + "epoch": 0.08708625355631115, + "grad_norm": 5.297552108764648, + "learning_rate": 0.00014508670520231215, + "loss": 1.2078, + "step": 1255 + }, + { + "epoch": 0.08715564499340782, + "grad_norm": 4.632575511932373, + "learning_rate": 0.0001452023121387283, + "loss": 1.0067, + "step": 1256 + }, + { + "epoch": 0.08722503643050447, + "grad_norm": 5.4039082527160645, + "learning_rate": 0.00014531791907514453, + "loss": 1.5035, + "step": 1257 + }, + { + "epoch": 0.08729442786760114, + "grad_norm": 4.8483567237854, + "learning_rate": 0.0001454335260115607, + "loss": 1.0896, + "step": 1258 + }, + { + "epoch": 0.0873638193046978, + "grad_norm": 8.207198143005371, + "learning_rate": 0.0001455491329479769, + "loss": 1.3331, + "step": 1259 + }, + { + "epoch": 0.08743321074179446, + "grad_norm": 30.904008865356445, + "learning_rate": 0.00014566473988439306, + "loss": 1.2471, + "step": 1260 + }, + { + "epoch": 0.08750260217889112, + "grad_norm": 5.121526718139648, + "learning_rate": 0.00014578034682080924, + "loss": 1.1223, + "step": 1261 + }, + { + "epoch": 0.08757199361598779, + "grad_norm": 5.284627437591553, + "learning_rate": 0.00014589595375722546, + "loss": 1.0464, + "step": 1262 + }, + { + "epoch": 0.08764138505308446, + "grad_norm": 5.866982936859131, + "learning_rate": 0.00014601156069364164, + "loss": 1.2266, + "step": 1263 + }, + { + "epoch": 0.08771077649018111, + "grad_norm": 5.693636894226074, + "learning_rate": 0.0001461271676300578, + "loss": 1.0594, + "step": 1264 + }, + { + "epoch": 0.08778016792727777, + "grad_norm": 7.7707414627075195, + "learning_rate": 0.000146242774566474, + "loss": 1.1374, + "step": 1265 + }, + { + "epoch": 0.08784955936437444, + "grad_norm": 6.277050971984863, + "learning_rate": 0.00014635838150289018, + "loss": 1.3586, + "step": 1266 + }, + { + "epoch": 0.0879189508014711, + "grad_norm": 5.882453918457031, + "learning_rate": 0.0001464739884393064, + "loss": 0.9652, + "step": 1267 + }, + { + "epoch": 0.08798834223856776, + "grad_norm": 6.7996954917907715, + "learning_rate": 0.00014658959537572255, + "loss": 1.1688, + "step": 1268 + }, + { + "epoch": 0.08805773367566443, + "grad_norm": 7.965925693511963, + "learning_rate": 0.00014670520231213873, + "loss": 1.6037, + "step": 1269 + }, + { + "epoch": 0.08812712511276108, + "grad_norm": 6.003370761871338, + "learning_rate": 0.00014682080924855492, + "loss": 1.3764, + "step": 1270 + }, + { + "epoch": 0.08819651654985775, + "grad_norm": 6.7336859703063965, + "learning_rate": 0.0001469364161849711, + "loss": 1.4315, + "step": 1271 + }, + { + "epoch": 0.08826590798695441, + "grad_norm": 5.796169281005859, + "learning_rate": 0.0001470520231213873, + "loss": 1.139, + "step": 1272 + }, + { + "epoch": 0.08833529942405108, + "grad_norm": 8.186182975769043, + "learning_rate": 0.00014716763005780348, + "loss": 1.3808, + "step": 1273 + }, + { + "epoch": 0.08840469086114773, + "grad_norm": 6.831549167633057, + "learning_rate": 0.00014728323699421967, + "loss": 1.1156, + "step": 1274 + }, + { + "epoch": 0.0884740822982444, + "grad_norm": 17.650564193725586, + "learning_rate": 0.00014739884393063585, + "loss": 1.7393, + "step": 1275 + }, + { + "epoch": 0.08854347373534106, + "grad_norm": 8.898625373840332, + "learning_rate": 0.000147514450867052, + "loss": 1.1325, + "step": 1276 + }, + { + "epoch": 0.08861286517243772, + "grad_norm": 9.764200210571289, + "learning_rate": 0.00014763005780346822, + "loss": 1.4682, + "step": 1277 + }, + { + "epoch": 0.08868225660953438, + "grad_norm": 5.406944274902344, + "learning_rate": 0.0001477456647398844, + "loss": 1.3583, + "step": 1278 + }, + { + "epoch": 0.08875164804663105, + "grad_norm": 6.6537675857543945, + "learning_rate": 0.0001478612716763006, + "loss": 1.3873, + "step": 1279 + }, + { + "epoch": 0.0888210394837277, + "grad_norm": 4.9319844245910645, + "learning_rate": 0.00014797687861271676, + "loss": 1.0553, + "step": 1280 + }, + { + "epoch": 0.08889043092082437, + "grad_norm": 17.02796745300293, + "learning_rate": 0.00014809248554913294, + "loss": 1.1464, + "step": 1281 + }, + { + "epoch": 0.08895982235792103, + "grad_norm": 6.242152214050293, + "learning_rate": 0.00014820809248554915, + "loss": 1.2997, + "step": 1282 + }, + { + "epoch": 0.0890292137950177, + "grad_norm": 7.0243964195251465, + "learning_rate": 0.00014832369942196534, + "loss": 1.4634, + "step": 1283 + }, + { + "epoch": 0.08909860523211435, + "grad_norm": 4.602797508239746, + "learning_rate": 0.0001484393063583815, + "loss": 1.1432, + "step": 1284 + }, + { + "epoch": 0.08916799666921102, + "grad_norm": 10.998750686645508, + "learning_rate": 0.00014855491329479769, + "loss": 1.17, + "step": 1285 + }, + { + "epoch": 0.08923738810630769, + "grad_norm": 5.324678897857666, + "learning_rate": 0.00014867052023121387, + "loss": 1.2242, + "step": 1286 + }, + { + "epoch": 0.08930677954340434, + "grad_norm": 6.6154680252075195, + "learning_rate": 0.00014878612716763009, + "loss": 1.2495, + "step": 1287 + }, + { + "epoch": 0.089376170980501, + "grad_norm": 15.465399742126465, + "learning_rate": 0.00014890173410404624, + "loss": 1.2386, + "step": 1288 + }, + { + "epoch": 0.08944556241759767, + "grad_norm": 10.033003807067871, + "learning_rate": 0.00014901734104046243, + "loss": 1.8534, + "step": 1289 + }, + { + "epoch": 0.08951495385469434, + "grad_norm": 8.502917289733887, + "learning_rate": 0.00014913294797687862, + "loss": 1.1907, + "step": 1290 + }, + { + "epoch": 0.08958434529179099, + "grad_norm": 11.60770320892334, + "learning_rate": 0.0001492485549132948, + "loss": 1.6582, + "step": 1291 + }, + { + "epoch": 0.08965373672888766, + "grad_norm": 22.396060943603516, + "learning_rate": 0.000149364161849711, + "loss": 1.4706, + "step": 1292 + }, + { + "epoch": 0.08972312816598432, + "grad_norm": 15.168434143066406, + "learning_rate": 0.00014947976878612718, + "loss": 1.8131, + "step": 1293 + }, + { + "epoch": 0.08979251960308098, + "grad_norm": 38.22659683227539, + "learning_rate": 0.00014959537572254336, + "loss": 2.5848, + "step": 1294 + }, + { + "epoch": 0.08986191104017764, + "grad_norm": 40.17222213745117, + "learning_rate": 0.00014971098265895955, + "loss": 2.7063, + "step": 1295 + }, + { + "epoch": 0.08993130247727431, + "grad_norm": 19.239370346069336, + "learning_rate": 0.00014982658959537573, + "loss": 1.5834, + "step": 1296 + }, + { + "epoch": 0.09000069391437096, + "grad_norm": 12.907248497009277, + "learning_rate": 0.00014994219653179192, + "loss": 1.4779, + "step": 1297 + }, + { + "epoch": 0.09007008535146763, + "grad_norm": 8.66781234741211, + "learning_rate": 0.0001500578034682081, + "loss": 1.8574, + "step": 1298 + }, + { + "epoch": 0.0901394767885643, + "grad_norm": 26.666908264160156, + "learning_rate": 0.0001501734104046243, + "loss": 1.4267, + "step": 1299 + }, + { + "epoch": 0.09020886822566096, + "grad_norm": 8.825684547424316, + "learning_rate": 0.00015028901734104045, + "loss": 1.5221, + "step": 1300 + }, + { + "epoch": 0.09027825966275761, + "grad_norm": 6.084698677062988, + "learning_rate": 0.00015040462427745664, + "loss": 1.3491, + "step": 1301 + }, + { + "epoch": 0.09034765109985428, + "grad_norm": 7.395939826965332, + "learning_rate": 0.00015052023121387285, + "loss": 1.6298, + "step": 1302 + }, + { + "epoch": 0.09041704253695095, + "grad_norm": 6.626138210296631, + "learning_rate": 0.00015063583815028904, + "loss": 1.4573, + "step": 1303 + }, + { + "epoch": 0.0904864339740476, + "grad_norm": 8.051230430603027, + "learning_rate": 0.0001507514450867052, + "loss": 1.65, + "step": 1304 + }, + { + "epoch": 0.09055582541114426, + "grad_norm": 7.116861820220947, + "learning_rate": 0.00015086705202312138, + "loss": 1.4574, + "step": 1305 + }, + { + "epoch": 0.09062521684824093, + "grad_norm": 7.60538387298584, + "learning_rate": 0.00015098265895953757, + "loss": 1.8287, + "step": 1306 + }, + { + "epoch": 0.09069460828533758, + "grad_norm": 15.54906940460205, + "learning_rate": 0.00015109826589595378, + "loss": 1.6986, + "step": 1307 + }, + { + "epoch": 0.09076399972243425, + "grad_norm": 5.45102596282959, + "learning_rate": 0.00015121387283236994, + "loss": 1.205, + "step": 1308 + }, + { + "epoch": 0.09083339115953092, + "grad_norm": 8.123376846313477, + "learning_rate": 0.00015132947976878613, + "loss": 1.4611, + "step": 1309 + }, + { + "epoch": 0.09090278259662758, + "grad_norm": 8.201353073120117, + "learning_rate": 0.00015144508670520231, + "loss": 1.2586, + "step": 1310 + }, + { + "epoch": 0.09097217403372423, + "grad_norm": 11.440990447998047, + "learning_rate": 0.0001515606936416185, + "loss": 1.7106, + "step": 1311 + }, + { + "epoch": 0.0910415654708209, + "grad_norm": 8.670344352722168, + "learning_rate": 0.0001516763005780347, + "loss": 1.3978, + "step": 1312 + }, + { + "epoch": 0.09111095690791757, + "grad_norm": 6.780796051025391, + "learning_rate": 0.00015179190751445087, + "loss": 1.3435, + "step": 1313 + }, + { + "epoch": 0.09118034834501422, + "grad_norm": 11.943354606628418, + "learning_rate": 0.00015190751445086706, + "loss": 1.7682, + "step": 1314 + }, + { + "epoch": 0.09124973978211089, + "grad_norm": 9.439358711242676, + "learning_rate": 0.00015202312138728325, + "loss": 1.3216, + "step": 1315 + }, + { + "epoch": 0.09131913121920755, + "grad_norm": 6.019266605377197, + "learning_rate": 0.00015213872832369943, + "loss": 1.2737, + "step": 1316 + }, + { + "epoch": 0.0913885226563042, + "grad_norm": 7.059696197509766, + "learning_rate": 0.00015225433526011562, + "loss": 1.4069, + "step": 1317 + }, + { + "epoch": 0.09145791409340087, + "grad_norm": 6.514481544494629, + "learning_rate": 0.0001523699421965318, + "loss": 1.6957, + "step": 1318 + }, + { + "epoch": 0.09152730553049754, + "grad_norm": 6.223950386047363, + "learning_rate": 0.000152485549132948, + "loss": 1.746, + "step": 1319 + }, + { + "epoch": 0.0915966969675942, + "grad_norm": 7.124370098114014, + "learning_rate": 0.00015260115606936415, + "loss": 1.4623, + "step": 1320 + }, + { + "epoch": 0.09166608840469086, + "grad_norm": 8.08069133758545, + "learning_rate": 0.00015271676300578036, + "loss": 1.7085, + "step": 1321 + }, + { + "epoch": 0.09173547984178752, + "grad_norm": 6.038676738739014, + "learning_rate": 0.00015283236994219655, + "loss": 1.4877, + "step": 1322 + }, + { + "epoch": 0.09180487127888419, + "grad_norm": 6.170168876647949, + "learning_rate": 0.00015294797687861274, + "loss": 1.4634, + "step": 1323 + }, + { + "epoch": 0.09187426271598084, + "grad_norm": 7.082663536071777, + "learning_rate": 0.0001530635838150289, + "loss": 1.466, + "step": 1324 + }, + { + "epoch": 0.09194365415307751, + "grad_norm": 5.616428852081299, + "learning_rate": 0.00015317919075144508, + "loss": 1.3041, + "step": 1325 + }, + { + "epoch": 0.09201304559017418, + "grad_norm": 4.9802632331848145, + "learning_rate": 0.00015329479768786127, + "loss": 1.3086, + "step": 1326 + }, + { + "epoch": 0.09208243702727084, + "grad_norm": 5.107244491577148, + "learning_rate": 0.00015341040462427748, + "loss": 1.2763, + "step": 1327 + }, + { + "epoch": 0.0921518284643675, + "grad_norm": 7.242337226867676, + "learning_rate": 0.00015352601156069364, + "loss": 1.6315, + "step": 1328 + }, + { + "epoch": 0.09222121990146416, + "grad_norm": 7.366434574127197, + "learning_rate": 0.00015364161849710983, + "loss": 1.3379, + "step": 1329 + }, + { + "epoch": 0.09229061133856083, + "grad_norm": 6.978879928588867, + "learning_rate": 0.000153757225433526, + "loss": 1.3506, + "step": 1330 + }, + { + "epoch": 0.09236000277565748, + "grad_norm": 5.083397388458252, + "learning_rate": 0.0001538728323699422, + "loss": 1.1663, + "step": 1331 + }, + { + "epoch": 0.09242939421275415, + "grad_norm": 6.201581954956055, + "learning_rate": 0.00015398843930635838, + "loss": 1.4533, + "step": 1332 + }, + { + "epoch": 0.09249878564985081, + "grad_norm": 5.759652137756348, + "learning_rate": 0.00015410404624277457, + "loss": 1.1799, + "step": 1333 + }, + { + "epoch": 0.09256817708694746, + "grad_norm": 8.927604675292969, + "learning_rate": 0.00015421965317919076, + "loss": 1.5611, + "step": 1334 + }, + { + "epoch": 0.09263756852404413, + "grad_norm": 6.224050998687744, + "learning_rate": 0.00015433526011560694, + "loss": 1.1914, + "step": 1335 + }, + { + "epoch": 0.0927069599611408, + "grad_norm": 5.974073886871338, + "learning_rate": 0.00015445086705202313, + "loss": 1.4063, + "step": 1336 + }, + { + "epoch": 0.09277635139823746, + "grad_norm": 25.876239776611328, + "learning_rate": 0.00015456647398843932, + "loss": 1.3267, + "step": 1337 + }, + { + "epoch": 0.09284574283533412, + "grad_norm": 8.267107963562012, + "learning_rate": 0.0001546820809248555, + "loss": 1.6408, + "step": 1338 + }, + { + "epoch": 0.09291513427243078, + "grad_norm": 9.12010669708252, + "learning_rate": 0.0001547976878612717, + "loss": 1.7431, + "step": 1339 + }, + { + "epoch": 0.09298452570952745, + "grad_norm": 13.394031524658203, + "learning_rate": 0.00015491329479768785, + "loss": 1.8481, + "step": 1340 + }, + { + "epoch": 0.0930539171466241, + "grad_norm": 5.87708854675293, + "learning_rate": 0.00015502890173410406, + "loss": 1.3204, + "step": 1341 + }, + { + "epoch": 0.09312330858372077, + "grad_norm": 5.673741340637207, + "learning_rate": 0.00015514450867052025, + "loss": 1.4476, + "step": 1342 + }, + { + "epoch": 0.09319270002081743, + "grad_norm": 5.914341926574707, + "learning_rate": 0.00015526011560693643, + "loss": 1.5636, + "step": 1343 + }, + { + "epoch": 0.09326209145791409, + "grad_norm": 5.572535991668701, + "learning_rate": 0.0001553757225433526, + "loss": 1.6242, + "step": 1344 + }, + { + "epoch": 0.09333148289501075, + "grad_norm": 6.157727241516113, + "learning_rate": 0.00015549132947976878, + "loss": 1.5678, + "step": 1345 + }, + { + "epoch": 0.09340087433210742, + "grad_norm": 5.471911907196045, + "learning_rate": 0.000155606936416185, + "loss": 1.3591, + "step": 1346 + }, + { + "epoch": 0.09347026576920409, + "grad_norm": 5.704415321350098, + "learning_rate": 0.00015572254335260118, + "loss": 1.4541, + "step": 1347 + }, + { + "epoch": 0.09353965720630074, + "grad_norm": 4.950115203857422, + "learning_rate": 0.00015583815028901736, + "loss": 1.2094, + "step": 1348 + }, + { + "epoch": 0.0936090486433974, + "grad_norm": 4.635327339172363, + "learning_rate": 0.00015595375722543352, + "loss": 1.2385, + "step": 1349 + }, + { + "epoch": 0.09367844008049407, + "grad_norm": 5.394261837005615, + "learning_rate": 0.0001560693641618497, + "loss": 1.2312, + "step": 1350 + }, + { + "epoch": 0.09374783151759072, + "grad_norm": 5.814078330993652, + "learning_rate": 0.00015618497109826592, + "loss": 1.2376, + "step": 1351 + }, + { + "epoch": 0.09381722295468739, + "grad_norm": 6.0607476234436035, + "learning_rate": 0.0001563005780346821, + "loss": 1.3533, + "step": 1352 + }, + { + "epoch": 0.09388661439178406, + "grad_norm": 6.818049907684326, + "learning_rate": 0.00015641618497109827, + "loss": 1.3906, + "step": 1353 + }, + { + "epoch": 0.09395600582888071, + "grad_norm": 5.721499443054199, + "learning_rate": 0.00015653179190751445, + "loss": 1.2995, + "step": 1354 + }, + { + "epoch": 0.09402539726597738, + "grad_norm": 9.092040061950684, + "learning_rate": 0.00015664739884393064, + "loss": 1.4586, + "step": 1355 + }, + { + "epoch": 0.09409478870307404, + "grad_norm": 8.133210182189941, + "learning_rate": 0.00015676300578034683, + "loss": 1.5659, + "step": 1356 + }, + { + "epoch": 0.09416418014017071, + "grad_norm": 7.168988227844238, + "learning_rate": 0.000156878612716763, + "loss": 1.4217, + "step": 1357 + }, + { + "epoch": 0.09423357157726736, + "grad_norm": 7.198262691497803, + "learning_rate": 0.0001569942196531792, + "loss": 1.6593, + "step": 1358 + }, + { + "epoch": 0.09430296301436403, + "grad_norm": 7.284234523773193, + "learning_rate": 0.00015710982658959538, + "loss": 1.302, + "step": 1359 + }, + { + "epoch": 0.0943723544514607, + "grad_norm": 5.914739608764648, + "learning_rate": 0.00015722543352601157, + "loss": 1.348, + "step": 1360 + }, + { + "epoch": 0.09444174588855735, + "grad_norm": 5.390138149261475, + "learning_rate": 0.00015734104046242776, + "loss": 1.2839, + "step": 1361 + }, + { + "epoch": 0.09451113732565401, + "grad_norm": 4.927929878234863, + "learning_rate": 0.00015745664739884394, + "loss": 1.0676, + "step": 1362 + }, + { + "epoch": 0.09458052876275068, + "grad_norm": 6.4791998863220215, + "learning_rate": 0.00015757225433526013, + "loss": 1.5611, + "step": 1363 + }, + { + "epoch": 0.09464992019984735, + "grad_norm": 7.825900554656982, + "learning_rate": 0.00015768786127167632, + "loss": 1.5164, + "step": 1364 + }, + { + "epoch": 0.094719311636944, + "grad_norm": 5.854959964752197, + "learning_rate": 0.00015780346820809248, + "loss": 1.4054, + "step": 1365 + }, + { + "epoch": 0.09478870307404066, + "grad_norm": 4.845619201660156, + "learning_rate": 0.0001579190751445087, + "loss": 1.2006, + "step": 1366 + }, + { + "epoch": 0.09485809451113733, + "grad_norm": 4.4648823738098145, + "learning_rate": 0.00015803468208092487, + "loss": 1.325, + "step": 1367 + }, + { + "epoch": 0.09492748594823398, + "grad_norm": 5.087299346923828, + "learning_rate": 0.00015815028901734106, + "loss": 1.2893, + "step": 1368 + }, + { + "epoch": 0.09499687738533065, + "grad_norm": 4.464088439941406, + "learning_rate": 0.00015826589595375722, + "loss": 1.2811, + "step": 1369 + }, + { + "epoch": 0.09506626882242732, + "grad_norm": 4.502194404602051, + "learning_rate": 0.0001583815028901734, + "loss": 1.2615, + "step": 1370 + }, + { + "epoch": 0.09513566025952397, + "grad_norm": 5.9749016761779785, + "learning_rate": 0.00015849710982658962, + "loss": 1.4301, + "step": 1371 + }, + { + "epoch": 0.09520505169662064, + "grad_norm": 3.8806376457214355, + "learning_rate": 0.0001586127167630058, + "loss": 0.8754, + "step": 1372 + }, + { + "epoch": 0.0952744431337173, + "grad_norm": 7.55876350402832, + "learning_rate": 0.00015872832369942196, + "loss": 1.2575, + "step": 1373 + }, + { + "epoch": 0.09534383457081397, + "grad_norm": 6.3321614265441895, + "learning_rate": 0.00015884393063583815, + "loss": 1.3761, + "step": 1374 + }, + { + "epoch": 0.09541322600791062, + "grad_norm": 7.160039901733398, + "learning_rate": 0.00015895953757225434, + "loss": 1.1896, + "step": 1375 + }, + { + "epoch": 0.09548261744500729, + "grad_norm": 10.23210620880127, + "learning_rate": 0.00015907514450867055, + "loss": 1.192, + "step": 1376 + }, + { + "epoch": 0.09555200888210395, + "grad_norm": 5.887913703918457, + "learning_rate": 0.0001591907514450867, + "loss": 1.281, + "step": 1377 + }, + { + "epoch": 0.0956214003192006, + "grad_norm": 5.776708126068115, + "learning_rate": 0.0001593063583815029, + "loss": 1.4419, + "step": 1378 + }, + { + "epoch": 0.09569079175629727, + "grad_norm": 8.933030128479004, + "learning_rate": 0.00015942196531791908, + "loss": 1.4183, + "step": 1379 + }, + { + "epoch": 0.09576018319339394, + "grad_norm": 8.715253829956055, + "learning_rate": 0.00015953757225433527, + "loss": 1.4453, + "step": 1380 + }, + { + "epoch": 0.09582957463049059, + "grad_norm": 5.097191333770752, + "learning_rate": 0.00015965317919075145, + "loss": 1.1688, + "step": 1381 + }, + { + "epoch": 0.09589896606758726, + "grad_norm": 5.121007919311523, + "learning_rate": 0.00015976878612716764, + "loss": 1.1622, + "step": 1382 + }, + { + "epoch": 0.09596835750468392, + "grad_norm": 5.152527332305908, + "learning_rate": 0.00015988439306358383, + "loss": 1.3236, + "step": 1383 + }, + { + "epoch": 0.09603774894178059, + "grad_norm": 5.631650447845459, + "learning_rate": 0.00016, + "loss": 1.2459, + "step": 1384 + }, + { + "epoch": 0.09610714037887724, + "grad_norm": 5.486510276794434, + "learning_rate": 0.00016011560693641617, + "loss": 1.3355, + "step": 1385 + }, + { + "epoch": 0.09617653181597391, + "grad_norm": 5.495135307312012, + "learning_rate": 0.00016023121387283239, + "loss": 1.1327, + "step": 1386 + }, + { + "epoch": 0.09624592325307058, + "grad_norm": 5.38565731048584, + "learning_rate": 0.00016034682080924857, + "loss": 1.6248, + "step": 1387 + }, + { + "epoch": 0.09631531469016723, + "grad_norm": 4.923528671264648, + "learning_rate": 0.00016046242774566476, + "loss": 0.9039, + "step": 1388 + }, + { + "epoch": 0.0963847061272639, + "grad_norm": 5.319502353668213, + "learning_rate": 0.00016057803468208092, + "loss": 1.1741, + "step": 1389 + }, + { + "epoch": 0.09645409756436056, + "grad_norm": 7.401739597320557, + "learning_rate": 0.0001606936416184971, + "loss": 1.2401, + "step": 1390 + }, + { + "epoch": 0.09652348900145721, + "grad_norm": 6.671390533447266, + "learning_rate": 0.00016080924855491332, + "loss": 1.4089, + "step": 1391 + }, + { + "epoch": 0.09659288043855388, + "grad_norm": 6.527892589569092, + "learning_rate": 0.0001609248554913295, + "loss": 1.4994, + "step": 1392 + }, + { + "epoch": 0.09666227187565055, + "grad_norm": 5.698474884033203, + "learning_rate": 0.00016104046242774566, + "loss": 1.3075, + "step": 1393 + }, + { + "epoch": 0.09673166331274721, + "grad_norm": 4.532931804656982, + "learning_rate": 0.00016115606936416185, + "loss": 1.057, + "step": 1394 + }, + { + "epoch": 0.09680105474984387, + "grad_norm": 8.459033966064453, + "learning_rate": 0.00016127167630057803, + "loss": 1.3842, + "step": 1395 + }, + { + "epoch": 0.09687044618694053, + "grad_norm": 7.2036285400390625, + "learning_rate": 0.00016138728323699425, + "loss": 1.4946, + "step": 1396 + }, + { + "epoch": 0.0969398376240372, + "grad_norm": 5.990884780883789, + "learning_rate": 0.0001615028901734104, + "loss": 1.4231, + "step": 1397 + }, + { + "epoch": 0.09700922906113385, + "grad_norm": 6.035245418548584, + "learning_rate": 0.0001616184971098266, + "loss": 1.3119, + "step": 1398 + }, + { + "epoch": 0.09707862049823052, + "grad_norm": 7.5014495849609375, + "learning_rate": 0.00016173410404624278, + "loss": 1.285, + "step": 1399 + }, + { + "epoch": 0.09714801193532718, + "grad_norm": 8.235336303710938, + "learning_rate": 0.00016184971098265897, + "loss": 1.3686, + "step": 1400 + }, + { + "epoch": 0.09721740337242385, + "grad_norm": 6.047291278839111, + "learning_rate": 0.00016196531791907515, + "loss": 1.4024, + "step": 1401 + }, + { + "epoch": 0.0972867948095205, + "grad_norm": 4.864603519439697, + "learning_rate": 0.00016208092485549134, + "loss": 1.0969, + "step": 1402 + }, + { + "epoch": 0.09735618624661717, + "grad_norm": 5.849535942077637, + "learning_rate": 0.00016219653179190752, + "loss": 1.2079, + "step": 1403 + }, + { + "epoch": 0.09742557768371383, + "grad_norm": 4.917699337005615, + "learning_rate": 0.0001623121387283237, + "loss": 1.2198, + "step": 1404 + }, + { + "epoch": 0.09749496912081049, + "grad_norm": 6.565923690795898, + "learning_rate": 0.0001624277456647399, + "loss": 0.9606, + "step": 1405 + }, + { + "epoch": 0.09756436055790715, + "grad_norm": 3.783968687057495, + "learning_rate": 0.00016254335260115608, + "loss": 0.9533, + "step": 1406 + }, + { + "epoch": 0.09763375199500382, + "grad_norm": 6.075403690338135, + "learning_rate": 0.00016265895953757227, + "loss": 1.4217, + "step": 1407 + }, + { + "epoch": 0.09770314343210047, + "grad_norm": 4.0914740562438965, + "learning_rate": 0.00016277456647398846, + "loss": 1.1818, + "step": 1408 + }, + { + "epoch": 0.09777253486919714, + "grad_norm": 5.0565314292907715, + "learning_rate": 0.00016289017341040461, + "loss": 1.4055, + "step": 1409 + }, + { + "epoch": 0.0978419263062938, + "grad_norm": 4.598503112792969, + "learning_rate": 0.0001630057803468208, + "loss": 1.2772, + "step": 1410 + }, + { + "epoch": 0.09791131774339047, + "grad_norm": 4.622971534729004, + "learning_rate": 0.00016312138728323701, + "loss": 1.2879, + "step": 1411 + }, + { + "epoch": 0.09798070918048712, + "grad_norm": 4.192399978637695, + "learning_rate": 0.0001632369942196532, + "loss": 1.3096, + "step": 1412 + }, + { + "epoch": 0.09805010061758379, + "grad_norm": 4.833970546722412, + "learning_rate": 0.00016335260115606936, + "loss": 1.1431, + "step": 1413 + }, + { + "epoch": 0.09811949205468046, + "grad_norm": 4.36276912689209, + "learning_rate": 0.00016346820809248555, + "loss": 1.0916, + "step": 1414 + }, + { + "epoch": 0.09818888349177711, + "grad_norm": 4.6851348876953125, + "learning_rate": 0.00016358381502890173, + "loss": 0.8851, + "step": 1415 + }, + { + "epoch": 0.09825827492887378, + "grad_norm": 5.920718193054199, + "learning_rate": 0.00016369942196531794, + "loss": 1.3606, + "step": 1416 + }, + { + "epoch": 0.09832766636597044, + "grad_norm": 6.505650520324707, + "learning_rate": 0.0001638150289017341, + "loss": 1.3214, + "step": 1417 + }, + { + "epoch": 0.0983970578030671, + "grad_norm": 5.251327991485596, + "learning_rate": 0.0001639306358381503, + "loss": 1.4055, + "step": 1418 + }, + { + "epoch": 0.09846644924016376, + "grad_norm": 3.98667311668396, + "learning_rate": 0.00016404624277456648, + "loss": 1.0476, + "step": 1419 + }, + { + "epoch": 0.09853584067726043, + "grad_norm": 8.18868350982666, + "learning_rate": 0.00016416184971098266, + "loss": 1.4142, + "step": 1420 + }, + { + "epoch": 0.0986052321143571, + "grad_norm": 4.349497318267822, + "learning_rate": 0.00016427745664739885, + "loss": 1.0927, + "step": 1421 + }, + { + "epoch": 0.09867462355145375, + "grad_norm": 4.6050262451171875, + "learning_rate": 0.00016439306358381504, + "loss": 1.1347, + "step": 1422 + }, + { + "epoch": 0.09874401498855041, + "grad_norm": 5.281850337982178, + "learning_rate": 0.00016450867052023122, + "loss": 1.2856, + "step": 1423 + }, + { + "epoch": 0.09881340642564708, + "grad_norm": 4.947666168212891, + "learning_rate": 0.0001646242774566474, + "loss": 1.2998, + "step": 1424 + }, + { + "epoch": 0.09888279786274373, + "grad_norm": 5.408533573150635, + "learning_rate": 0.0001647398843930636, + "loss": 1.2198, + "step": 1425 + }, + { + "epoch": 0.0989521892998404, + "grad_norm": 5.291476249694824, + "learning_rate": 0.00016485549132947978, + "loss": 1.2756, + "step": 1426 + }, + { + "epoch": 0.09902158073693706, + "grad_norm": 5.492124080657959, + "learning_rate": 0.00016497109826589597, + "loss": 1.2235, + "step": 1427 + }, + { + "epoch": 0.09909097217403372, + "grad_norm": 5.098992347717285, + "learning_rate": 0.00016508670520231215, + "loss": 1.235, + "step": 1428 + }, + { + "epoch": 0.09916036361113038, + "grad_norm": 7.135852813720703, + "learning_rate": 0.0001652023121387283, + "loss": 1.4518, + "step": 1429 + }, + { + "epoch": 0.09922975504822705, + "grad_norm": 4.098690032958984, + "learning_rate": 0.00016531791907514452, + "loss": 1.072, + "step": 1430 + }, + { + "epoch": 0.09929914648532372, + "grad_norm": 5.74235200881958, + "learning_rate": 0.0001654335260115607, + "loss": 1.2555, + "step": 1431 + }, + { + "epoch": 0.09936853792242037, + "grad_norm": 5.183781147003174, + "learning_rate": 0.0001655491329479769, + "loss": 1.2263, + "step": 1432 + }, + { + "epoch": 0.09943792935951704, + "grad_norm": 7.083201885223389, + "learning_rate": 0.00016566473988439306, + "loss": 1.3646, + "step": 1433 + }, + { + "epoch": 0.0995073207966137, + "grad_norm": 4.7256035804748535, + "learning_rate": 0.00016578034682080924, + "loss": 1.1496, + "step": 1434 + }, + { + "epoch": 0.09957671223371035, + "grad_norm": 5.805156230926514, + "learning_rate": 0.00016589595375722543, + "loss": 0.9919, + "step": 1435 + }, + { + "epoch": 0.09964610367080702, + "grad_norm": 5.640675067901611, + "learning_rate": 0.00016601156069364164, + "loss": 1.365, + "step": 1436 + }, + { + "epoch": 0.09971549510790369, + "grad_norm": 4.65151834487915, + "learning_rate": 0.0001661271676300578, + "loss": 1.1084, + "step": 1437 + }, + { + "epoch": 0.09978488654500035, + "grad_norm": 4.759190082550049, + "learning_rate": 0.000166242774566474, + "loss": 1.1369, + "step": 1438 + }, + { + "epoch": 0.099854277982097, + "grad_norm": 5.140300273895264, + "learning_rate": 0.00016635838150289017, + "loss": 0.9804, + "step": 1439 + }, + { + "epoch": 0.09992366941919367, + "grad_norm": 4.659962177276611, + "learning_rate": 0.00016647398843930636, + "loss": 1.1596, + "step": 1440 + }, + { + "epoch": 0.09999306085629034, + "grad_norm": 5.394941329956055, + "learning_rate": 0.00016658959537572255, + "loss": 1.2453, + "step": 1441 + }, + { + "epoch": 0.10006245229338699, + "grad_norm": 4.959056854248047, + "learning_rate": 0.00016670520231213873, + "loss": 1.083, + "step": 1442 + }, + { + "epoch": 0.10013184373048366, + "grad_norm": 12.154365539550781, + "learning_rate": 0.00016682080924855492, + "loss": 1.3261, + "step": 1443 + }, + { + "epoch": 0.10020123516758032, + "grad_norm": 4.149370193481445, + "learning_rate": 0.0001669364161849711, + "loss": 1.1759, + "step": 1444 + }, + { + "epoch": 0.10027062660467698, + "grad_norm": 5.419303894042969, + "learning_rate": 0.0001670520231213873, + "loss": 1.009, + "step": 1445 + }, + { + "epoch": 0.10034001804177364, + "grad_norm": 5.864902496337891, + "learning_rate": 0.00016716763005780348, + "loss": 1.2971, + "step": 1446 + }, + { + "epoch": 0.10040940947887031, + "grad_norm": 12.582813262939453, + "learning_rate": 0.00016728323699421966, + "loss": 1.3779, + "step": 1447 + }, + { + "epoch": 0.10047880091596698, + "grad_norm": 5.4385223388671875, + "learning_rate": 0.00016739884393063585, + "loss": 1.1053, + "step": 1448 + }, + { + "epoch": 0.10054819235306363, + "grad_norm": 5.173016548156738, + "learning_rate": 0.00016751445086705204, + "loss": 1.1648, + "step": 1449 + }, + { + "epoch": 0.1006175837901603, + "grad_norm": 4.834924221038818, + "learning_rate": 0.00016763005780346822, + "loss": 1.082, + "step": 1450 + }, + { + "epoch": 0.10068697522725696, + "grad_norm": 5.702645301818848, + "learning_rate": 0.0001677456647398844, + "loss": 1.3462, + "step": 1451 + }, + { + "epoch": 0.10075636666435361, + "grad_norm": 6.031981945037842, + "learning_rate": 0.0001678612716763006, + "loss": 1.0261, + "step": 1452 + }, + { + "epoch": 0.10082575810145028, + "grad_norm": 4.742885589599609, + "learning_rate": 0.00016797687861271678, + "loss": 1.167, + "step": 1453 + }, + { + "epoch": 0.10089514953854695, + "grad_norm": 4.4435601234436035, + "learning_rate": 0.00016809248554913294, + "loss": 1.1127, + "step": 1454 + }, + { + "epoch": 0.1009645409756436, + "grad_norm": 4.986124038696289, + "learning_rate": 0.00016820809248554915, + "loss": 1.08, + "step": 1455 + }, + { + "epoch": 0.10103393241274027, + "grad_norm": 4.927673816680908, + "learning_rate": 0.00016832369942196534, + "loss": 1.1744, + "step": 1456 + }, + { + "epoch": 0.10110332384983693, + "grad_norm": 5.239799976348877, + "learning_rate": 0.00016843930635838153, + "loss": 1.2046, + "step": 1457 + }, + { + "epoch": 0.1011727152869336, + "grad_norm": 7.997239112854004, + "learning_rate": 0.00016855491329479768, + "loss": 1.3736, + "step": 1458 + }, + { + "epoch": 0.10124210672403025, + "grad_norm": 5.949313640594482, + "learning_rate": 0.00016867052023121387, + "loss": 1.0164, + "step": 1459 + }, + { + "epoch": 0.10131149816112692, + "grad_norm": 5.876352787017822, + "learning_rate": 0.00016878612716763006, + "loss": 1.2754, + "step": 1460 + }, + { + "epoch": 0.10138088959822358, + "grad_norm": 5.291408061981201, + "learning_rate": 0.00016890173410404627, + "loss": 1.1047, + "step": 1461 + }, + { + "epoch": 0.10145028103532024, + "grad_norm": 6.462136745452881, + "learning_rate": 0.00016901734104046243, + "loss": 1.4335, + "step": 1462 + }, + { + "epoch": 0.1015196724724169, + "grad_norm": 5.289731979370117, + "learning_rate": 0.00016913294797687862, + "loss": 1.4541, + "step": 1463 + }, + { + "epoch": 0.10158906390951357, + "grad_norm": 4.890068054199219, + "learning_rate": 0.0001692485549132948, + "loss": 1.2361, + "step": 1464 + }, + { + "epoch": 0.10165845534661022, + "grad_norm": 5.919886589050293, + "learning_rate": 0.000169364161849711, + "loss": 1.384, + "step": 1465 + }, + { + "epoch": 0.10172784678370689, + "grad_norm": 5.931599140167236, + "learning_rate": 0.00016947976878612717, + "loss": 1.4897, + "step": 1466 + }, + { + "epoch": 0.10179723822080355, + "grad_norm": 6.510693073272705, + "learning_rate": 0.00016959537572254336, + "loss": 1.4774, + "step": 1467 + }, + { + "epoch": 0.10186662965790022, + "grad_norm": 4.84014368057251, + "learning_rate": 0.00016971098265895955, + "loss": 1.1597, + "step": 1468 + }, + { + "epoch": 0.10193602109499687, + "grad_norm": 5.574368953704834, + "learning_rate": 0.00016982658959537573, + "loss": 1.4135, + "step": 1469 + }, + { + "epoch": 0.10200541253209354, + "grad_norm": 5.800242900848389, + "learning_rate": 0.00016994219653179192, + "loss": 1.6652, + "step": 1470 + }, + { + "epoch": 0.1020748039691902, + "grad_norm": 5.397841453552246, + "learning_rate": 0.0001700578034682081, + "loss": 1.2847, + "step": 1471 + }, + { + "epoch": 0.10214419540628686, + "grad_norm": 5.483976364135742, + "learning_rate": 0.0001701734104046243, + "loss": 1.1366, + "step": 1472 + }, + { + "epoch": 0.10221358684338352, + "grad_norm": 4.2113213539123535, + "learning_rate": 0.00017028901734104048, + "loss": 1.0925, + "step": 1473 + }, + { + "epoch": 0.10228297828048019, + "grad_norm": 5.813305854797363, + "learning_rate": 0.00017040462427745664, + "loss": 1.0303, + "step": 1474 + }, + { + "epoch": 0.10235236971757686, + "grad_norm": 4.404136657714844, + "learning_rate": 0.00017052023121387285, + "loss": 1.2793, + "step": 1475 + }, + { + "epoch": 0.10242176115467351, + "grad_norm": 4.830326080322266, + "learning_rate": 0.00017063583815028904, + "loss": 0.9738, + "step": 1476 + }, + { + "epoch": 0.10249115259177018, + "grad_norm": 7.108009338378906, + "learning_rate": 0.00017075144508670522, + "loss": 1.2535, + "step": 1477 + }, + { + "epoch": 0.10256054402886684, + "grad_norm": 4.972548007965088, + "learning_rate": 0.00017086705202312138, + "loss": 1.4455, + "step": 1478 + }, + { + "epoch": 0.1026299354659635, + "grad_norm": 5.6082072257995605, + "learning_rate": 0.00017098265895953757, + "loss": 1.2892, + "step": 1479 + }, + { + "epoch": 0.10269932690306016, + "grad_norm": 4.760672092437744, + "learning_rate": 0.00017109826589595378, + "loss": 1.341, + "step": 1480 + }, + { + "epoch": 0.10276871834015683, + "grad_norm": 4.96142053604126, + "learning_rate": 0.00017121387283236997, + "loss": 1.0659, + "step": 1481 + }, + { + "epoch": 0.10283810977725348, + "grad_norm": 6.2908854484558105, + "learning_rate": 0.00017132947976878613, + "loss": 1.2387, + "step": 1482 + }, + { + "epoch": 0.10290750121435015, + "grad_norm": 6.30145263671875, + "learning_rate": 0.0001714450867052023, + "loss": 0.9645, + "step": 1483 + }, + { + "epoch": 0.10297689265144681, + "grad_norm": 5.486726760864258, + "learning_rate": 0.0001715606936416185, + "loss": 1.3159, + "step": 1484 + }, + { + "epoch": 0.10304628408854348, + "grad_norm": 4.462891101837158, + "learning_rate": 0.0001716763005780347, + "loss": 1.4682, + "step": 1485 + }, + { + "epoch": 0.10311567552564013, + "grad_norm": 5.139286041259766, + "learning_rate": 0.00017179190751445087, + "loss": 0.9893, + "step": 1486 + }, + { + "epoch": 0.1031850669627368, + "grad_norm": 7.038825988769531, + "learning_rate": 0.00017190751445086706, + "loss": 1.3125, + "step": 1487 + }, + { + "epoch": 0.10325445839983347, + "grad_norm": 6.2188191413879395, + "learning_rate": 0.00017202312138728324, + "loss": 1.3055, + "step": 1488 + }, + { + "epoch": 0.10332384983693012, + "grad_norm": 4.372456073760986, + "learning_rate": 0.00017213872832369943, + "loss": 1.1024, + "step": 1489 + }, + { + "epoch": 0.10339324127402678, + "grad_norm": 4.381187438964844, + "learning_rate": 0.00017225433526011562, + "loss": 1.0906, + "step": 1490 + }, + { + "epoch": 0.10346263271112345, + "grad_norm": 3.7498536109924316, + "learning_rate": 0.0001723699421965318, + "loss": 0.9675, + "step": 1491 + }, + { + "epoch": 0.1035320241482201, + "grad_norm": 4.93365478515625, + "learning_rate": 0.000172485549132948, + "loss": 1.3096, + "step": 1492 + }, + { + "epoch": 0.10360141558531677, + "grad_norm": 5.01161527633667, + "learning_rate": 0.00017260115606936417, + "loss": 1.416, + "step": 1493 + }, + { + "epoch": 0.10367080702241344, + "grad_norm": 4.891099452972412, + "learning_rate": 0.00017271676300578033, + "loss": 1.1542, + "step": 1494 + }, + { + "epoch": 0.1037401984595101, + "grad_norm": 4.955638408660889, + "learning_rate": 0.00017283236994219655, + "loss": 1.1036, + "step": 1495 + }, + { + "epoch": 0.10380958989660675, + "grad_norm": 5.551807880401611, + "learning_rate": 0.00017294797687861273, + "loss": 1.3554, + "step": 1496 + }, + { + "epoch": 0.10387898133370342, + "grad_norm": 5.4071760177612305, + "learning_rate": 0.00017306358381502892, + "loss": 1.4719, + "step": 1497 + }, + { + "epoch": 0.10394837277080009, + "grad_norm": 4.5340986251831055, + "learning_rate": 0.00017317919075144508, + "loss": 1.1275, + "step": 1498 + }, + { + "epoch": 0.10401776420789674, + "grad_norm": 6.043724536895752, + "learning_rate": 0.00017329479768786127, + "loss": 1.3435, + "step": 1499 + }, + { + "epoch": 0.1040871556449934, + "grad_norm": 4.39643669128418, + "learning_rate": 0.00017341040462427748, + "loss": 1.2843, + "step": 1500 + }, + { + "epoch": 0.10415654708209007, + "grad_norm": 4.877024173736572, + "learning_rate": 0.00017352601156069366, + "loss": 0.9376, + "step": 1501 + }, + { + "epoch": 0.10422593851918673, + "grad_norm": 5.251949310302734, + "learning_rate": 0.00017364161849710982, + "loss": 1.1805, + "step": 1502 + }, + { + "epoch": 0.10429532995628339, + "grad_norm": 4.2061262130737305, + "learning_rate": 0.000173757225433526, + "loss": 0.9806, + "step": 1503 + }, + { + "epoch": 0.10436472139338006, + "grad_norm": 5.536079406738281, + "learning_rate": 0.0001738728323699422, + "loss": 1.2279, + "step": 1504 + }, + { + "epoch": 0.10443411283047672, + "grad_norm": 5.80854606628418, + "learning_rate": 0.0001739884393063584, + "loss": 1.3183, + "step": 1505 + }, + { + "epoch": 0.10450350426757338, + "grad_norm": 4.3310980796813965, + "learning_rate": 0.00017410404624277457, + "loss": 0.9947, + "step": 1506 + }, + { + "epoch": 0.10457289570467004, + "grad_norm": 5.460811614990234, + "learning_rate": 0.00017421965317919075, + "loss": 1.5019, + "step": 1507 + }, + { + "epoch": 0.10464228714176671, + "grad_norm": 5.083295822143555, + "learning_rate": 0.00017433526011560694, + "loss": 1.1226, + "step": 1508 + }, + { + "epoch": 0.10471167857886336, + "grad_norm": 4.947555065155029, + "learning_rate": 0.00017445086705202313, + "loss": 1.1586, + "step": 1509 + }, + { + "epoch": 0.10478107001596003, + "grad_norm": 4.69368839263916, + "learning_rate": 0.0001745664739884393, + "loss": 0.9631, + "step": 1510 + }, + { + "epoch": 0.1048504614530567, + "grad_norm": 4.638361930847168, + "learning_rate": 0.0001746820809248555, + "loss": 1.223, + "step": 1511 + }, + { + "epoch": 0.10491985289015336, + "grad_norm": 4.6971354484558105, + "learning_rate": 0.00017479768786127169, + "loss": 1.2015, + "step": 1512 + }, + { + "epoch": 0.10498924432725001, + "grad_norm": 5.238001823425293, + "learning_rate": 0.00017491329479768787, + "loss": 1.0765, + "step": 1513 + }, + { + "epoch": 0.10505863576434668, + "grad_norm": 4.888269901275635, + "learning_rate": 0.00017502890173410406, + "loss": 1.3478, + "step": 1514 + }, + { + "epoch": 0.10512802720144335, + "grad_norm": 5.293951034545898, + "learning_rate": 0.00017514450867052024, + "loss": 1.2335, + "step": 1515 + }, + { + "epoch": 0.10519741863854, + "grad_norm": 8.454535484313965, + "learning_rate": 0.00017526011560693643, + "loss": 1.1811, + "step": 1516 + }, + { + "epoch": 0.10526681007563667, + "grad_norm": 6.303637981414795, + "learning_rate": 0.00017537572254335262, + "loss": 1.1333, + "step": 1517 + }, + { + "epoch": 0.10533620151273333, + "grad_norm": 5.535228252410889, + "learning_rate": 0.00017549132947976878, + "loss": 0.8794, + "step": 1518 + }, + { + "epoch": 0.10540559294982998, + "grad_norm": 4.761526107788086, + "learning_rate": 0.00017560693641618496, + "loss": 1.3494, + "step": 1519 + }, + { + "epoch": 0.10547498438692665, + "grad_norm": 4.084989547729492, + "learning_rate": 0.00017572254335260118, + "loss": 1.1421, + "step": 1520 + }, + { + "epoch": 0.10554437582402332, + "grad_norm": 5.416632175445557, + "learning_rate": 0.00017583815028901736, + "loss": 1.6652, + "step": 1521 + }, + { + "epoch": 0.10561376726111998, + "grad_norm": 4.463350296020508, + "learning_rate": 0.00017595375722543352, + "loss": 1.3216, + "step": 1522 + }, + { + "epoch": 0.10568315869821664, + "grad_norm": 5.349869728088379, + "learning_rate": 0.0001760693641618497, + "loss": 1.3719, + "step": 1523 + }, + { + "epoch": 0.1057525501353133, + "grad_norm": 4.623307704925537, + "learning_rate": 0.0001761849710982659, + "loss": 0.9877, + "step": 1524 + }, + { + "epoch": 0.10582194157240997, + "grad_norm": 4.524658203125, + "learning_rate": 0.0001763005780346821, + "loss": 1.4404, + "step": 1525 + }, + { + "epoch": 0.10589133300950662, + "grad_norm": 3.9635372161865234, + "learning_rate": 0.00017641618497109827, + "loss": 0.9141, + "step": 1526 + }, + { + "epoch": 0.10596072444660329, + "grad_norm": 3.948498010635376, + "learning_rate": 0.00017653179190751445, + "loss": 1.0154, + "step": 1527 + }, + { + "epoch": 0.10603011588369995, + "grad_norm": 5.7547502517700195, + "learning_rate": 0.00017664739884393064, + "loss": 1.2023, + "step": 1528 + }, + { + "epoch": 0.10609950732079661, + "grad_norm": 5.365858554840088, + "learning_rate": 0.00017676300578034682, + "loss": 0.814, + "step": 1529 + }, + { + "epoch": 0.10616889875789327, + "grad_norm": 4.519698619842529, + "learning_rate": 0.000176878612716763, + "loss": 1.0126, + "step": 1530 + }, + { + "epoch": 0.10623829019498994, + "grad_norm": 4.855530261993408, + "learning_rate": 0.0001769942196531792, + "loss": 1.1122, + "step": 1531 + }, + { + "epoch": 0.1063076816320866, + "grad_norm": 4.648070335388184, + "learning_rate": 0.00017710982658959538, + "loss": 1.128, + "step": 1532 + }, + { + "epoch": 0.10637707306918326, + "grad_norm": 4.765169143676758, + "learning_rate": 0.00017722543352601157, + "loss": 1.0966, + "step": 1533 + }, + { + "epoch": 0.10644646450627993, + "grad_norm": 4.656566619873047, + "learning_rate": 0.00017734104046242776, + "loss": 1.2205, + "step": 1534 + }, + { + "epoch": 0.10651585594337659, + "grad_norm": 5.29012393951416, + "learning_rate": 0.00017745664739884394, + "loss": 0.9505, + "step": 1535 + }, + { + "epoch": 0.10658524738047324, + "grad_norm": 5.6948370933532715, + "learning_rate": 0.00017757225433526013, + "loss": 1.3473, + "step": 1536 + }, + { + "epoch": 0.10665463881756991, + "grad_norm": 5.6681694984436035, + "learning_rate": 0.00017768786127167631, + "loss": 1.1229, + "step": 1537 + }, + { + "epoch": 0.10672403025466658, + "grad_norm": 4.062856674194336, + "learning_rate": 0.0001778034682080925, + "loss": 1.0609, + "step": 1538 + }, + { + "epoch": 0.10679342169176323, + "grad_norm": 5.258630275726318, + "learning_rate": 0.0001779190751445087, + "loss": 1.0931, + "step": 1539 + }, + { + "epoch": 0.1068628131288599, + "grad_norm": 6.110175609588623, + "learning_rate": 0.00017803468208092487, + "loss": 1.4654, + "step": 1540 + }, + { + "epoch": 0.10693220456595656, + "grad_norm": 5.219849109649658, + "learning_rate": 0.00017815028901734106, + "loss": 1.1678, + "step": 1541 + }, + { + "epoch": 0.10700159600305323, + "grad_norm": 5.650516510009766, + "learning_rate": 0.00017826589595375725, + "loss": 1.302, + "step": 1542 + }, + { + "epoch": 0.10707098744014988, + "grad_norm": 5.267240047454834, + "learning_rate": 0.0001783815028901734, + "loss": 1.3404, + "step": 1543 + }, + { + "epoch": 0.10714037887724655, + "grad_norm": 5.606182098388672, + "learning_rate": 0.0001784971098265896, + "loss": 1.5077, + "step": 1544 + }, + { + "epoch": 0.10720977031434321, + "grad_norm": 4.650083541870117, + "learning_rate": 0.0001786127167630058, + "loss": 1.0531, + "step": 1545 + }, + { + "epoch": 0.10727916175143987, + "grad_norm": 4.066108703613281, + "learning_rate": 0.000178728323699422, + "loss": 1.2009, + "step": 1546 + }, + { + "epoch": 0.10734855318853653, + "grad_norm": 3.9384167194366455, + "learning_rate": 0.00017884393063583815, + "loss": 0.9849, + "step": 1547 + }, + { + "epoch": 0.1074179446256332, + "grad_norm": 4.660472869873047, + "learning_rate": 0.00017895953757225434, + "loss": 1.2307, + "step": 1548 + }, + { + "epoch": 0.10748733606272987, + "grad_norm": 4.141851902008057, + "learning_rate": 0.00017907514450867052, + "loss": 1.1037, + "step": 1549 + }, + { + "epoch": 0.10755672749982652, + "grad_norm": 4.904295921325684, + "learning_rate": 0.00017919075144508673, + "loss": 1.1753, + "step": 1550 + }, + { + "epoch": 0.10762611893692318, + "grad_norm": 4.125256538391113, + "learning_rate": 0.0001793063583815029, + "loss": 1.0949, + "step": 1551 + }, + { + "epoch": 0.10769551037401985, + "grad_norm": 4.3398213386535645, + "learning_rate": 0.00017942196531791908, + "loss": 1.3039, + "step": 1552 + }, + { + "epoch": 0.1077649018111165, + "grad_norm": 4.520273685455322, + "learning_rate": 0.00017953757225433527, + "loss": 1.0232, + "step": 1553 + }, + { + "epoch": 0.10783429324821317, + "grad_norm": 5.098697662353516, + "learning_rate": 0.00017965317919075145, + "loss": 1.2437, + "step": 1554 + }, + { + "epoch": 0.10790368468530984, + "grad_norm": 4.094663143157959, + "learning_rate": 0.00017976878612716764, + "loss": 0.973, + "step": 1555 + }, + { + "epoch": 0.10797307612240649, + "grad_norm": 4.970353126525879, + "learning_rate": 0.00017988439306358383, + "loss": 1.4037, + "step": 1556 + }, + { + "epoch": 0.10804246755950316, + "grad_norm": 4.609293460845947, + "learning_rate": 0.00018, + "loss": 1.1507, + "step": 1557 + }, + { + "epoch": 0.10811185899659982, + "grad_norm": 4.4987359046936035, + "learning_rate": 0.0001801156069364162, + "loss": 0.9245, + "step": 1558 + }, + { + "epoch": 0.10818125043369649, + "grad_norm": 5.944486141204834, + "learning_rate": 0.00018023121387283238, + "loss": 1.4086, + "step": 1559 + }, + { + "epoch": 0.10825064187079314, + "grad_norm": 4.422690391540527, + "learning_rate": 0.00018034682080924857, + "loss": 1.1549, + "step": 1560 + }, + { + "epoch": 0.10832003330788981, + "grad_norm": 3.734879493713379, + "learning_rate": 0.00018046242774566476, + "loss": 0.9998, + "step": 1561 + }, + { + "epoch": 0.10838942474498647, + "grad_norm": 7.14640474319458, + "learning_rate": 0.00018057803468208094, + "loss": 1.2369, + "step": 1562 + }, + { + "epoch": 0.10845881618208313, + "grad_norm": 5.6020917892456055, + "learning_rate": 0.0001806936416184971, + "loss": 1.2849, + "step": 1563 + }, + { + "epoch": 0.10852820761917979, + "grad_norm": 6.03865909576416, + "learning_rate": 0.00018080924855491331, + "loss": 0.901, + "step": 1564 + }, + { + "epoch": 0.10859759905627646, + "grad_norm": 4.770050525665283, + "learning_rate": 0.0001809248554913295, + "loss": 1.2409, + "step": 1565 + }, + { + "epoch": 0.10866699049337311, + "grad_norm": 5.508462905883789, + "learning_rate": 0.0001810404624277457, + "loss": 1.0815, + "step": 1566 + }, + { + "epoch": 0.10873638193046978, + "grad_norm": 4.943717002868652, + "learning_rate": 0.00018115606936416185, + "loss": 1.1083, + "step": 1567 + }, + { + "epoch": 0.10880577336756644, + "grad_norm": 5.439047336578369, + "learning_rate": 0.00018127167630057803, + "loss": 1.3307, + "step": 1568 + }, + { + "epoch": 0.10887516480466311, + "grad_norm": 4.737016201019287, + "learning_rate": 0.00018138728323699422, + "loss": 1.2376, + "step": 1569 + }, + { + "epoch": 0.10894455624175976, + "grad_norm": 5.679584503173828, + "learning_rate": 0.00018150289017341043, + "loss": 1.1054, + "step": 1570 + }, + { + "epoch": 0.10901394767885643, + "grad_norm": 4.683243274688721, + "learning_rate": 0.0001816184971098266, + "loss": 0.9711, + "step": 1571 + }, + { + "epoch": 0.1090833391159531, + "grad_norm": 4.853281497955322, + "learning_rate": 0.00018173410404624278, + "loss": 1.2582, + "step": 1572 + }, + { + "epoch": 0.10915273055304975, + "grad_norm": 5.002804279327393, + "learning_rate": 0.00018184971098265896, + "loss": 1.1891, + "step": 1573 + }, + { + "epoch": 0.10922212199014641, + "grad_norm": 6.552338123321533, + "learning_rate": 0.00018196531791907515, + "loss": 1.1967, + "step": 1574 + }, + { + "epoch": 0.10929151342724308, + "grad_norm": 4.752842426300049, + "learning_rate": 0.00018208092485549134, + "loss": 1.1749, + "step": 1575 + }, + { + "epoch": 0.10936090486433973, + "grad_norm": 6.406384468078613, + "learning_rate": 0.00018219653179190752, + "loss": 1.3077, + "step": 1576 + }, + { + "epoch": 0.1094302963014364, + "grad_norm": 4.226191997528076, + "learning_rate": 0.0001823121387283237, + "loss": 1.1079, + "step": 1577 + }, + { + "epoch": 0.10949968773853307, + "grad_norm": 8.268516540527344, + "learning_rate": 0.0001824277456647399, + "loss": 1.1977, + "step": 1578 + }, + { + "epoch": 0.10956907917562973, + "grad_norm": 5.6874213218688965, + "learning_rate": 0.00018254335260115608, + "loss": 1.3654, + "step": 1579 + }, + { + "epoch": 0.10963847061272639, + "grad_norm": 5.175532341003418, + "learning_rate": 0.00018265895953757227, + "loss": 0.994, + "step": 1580 + }, + { + "epoch": 0.10970786204982305, + "grad_norm": 4.592198848724365, + "learning_rate": 0.00018277456647398845, + "loss": 1.1129, + "step": 1581 + }, + { + "epoch": 0.10977725348691972, + "grad_norm": 4.619868755340576, + "learning_rate": 0.00018289017341040464, + "loss": 1.2024, + "step": 1582 + }, + { + "epoch": 0.10984664492401637, + "grad_norm": 4.327352523803711, + "learning_rate": 0.0001830057803468208, + "loss": 0.898, + "step": 1583 + }, + { + "epoch": 0.10991603636111304, + "grad_norm": 5.145925521850586, + "learning_rate": 0.000183121387283237, + "loss": 0.986, + "step": 1584 + }, + { + "epoch": 0.1099854277982097, + "grad_norm": 3.5017447471618652, + "learning_rate": 0.0001832369942196532, + "loss": 0.8381, + "step": 1585 + }, + { + "epoch": 0.11005481923530637, + "grad_norm": 12.782552719116211, + "learning_rate": 0.00018335260115606938, + "loss": 1.6214, + "step": 1586 + }, + { + "epoch": 0.11012421067240302, + "grad_norm": 4.6792216300964355, + "learning_rate": 0.00018346820809248554, + "loss": 1.2132, + "step": 1587 + }, + { + "epoch": 0.11019360210949969, + "grad_norm": 6.832906723022461, + "learning_rate": 0.00018358381502890173, + "loss": 1.4976, + "step": 1588 + }, + { + "epoch": 0.11026299354659636, + "grad_norm": 10.302786827087402, + "learning_rate": 0.00018369942196531794, + "loss": 1.314, + "step": 1589 + }, + { + "epoch": 0.11033238498369301, + "grad_norm": 5.829784870147705, + "learning_rate": 0.00018381502890173413, + "loss": 1.3222, + "step": 1590 + }, + { + "epoch": 0.11040177642078967, + "grad_norm": 4.677511692047119, + "learning_rate": 0.0001839306358381503, + "loss": 1.4305, + "step": 1591 + }, + { + "epoch": 0.11047116785788634, + "grad_norm": 12.242266654968262, + "learning_rate": 0.00018404624277456647, + "loss": 1.3487, + "step": 1592 + }, + { + "epoch": 0.11054055929498299, + "grad_norm": 4.12431526184082, + "learning_rate": 0.00018416184971098266, + "loss": 1.1368, + "step": 1593 + }, + { + "epoch": 0.11060995073207966, + "grad_norm": 8.229165077209473, + "learning_rate": 0.00018427745664739887, + "loss": 1.3531, + "step": 1594 + }, + { + "epoch": 0.11067934216917633, + "grad_norm": 5.755809783935547, + "learning_rate": 0.00018439306358381503, + "loss": 1.37, + "step": 1595 + }, + { + "epoch": 0.11074873360627299, + "grad_norm": 4.921316623687744, + "learning_rate": 0.00018450867052023122, + "loss": 1.1422, + "step": 1596 + }, + { + "epoch": 0.11081812504336964, + "grad_norm": 5.667410850524902, + "learning_rate": 0.0001846242774566474, + "loss": 1.0872, + "step": 1597 + }, + { + "epoch": 0.11088751648046631, + "grad_norm": 4.962009906768799, + "learning_rate": 0.0001847398843930636, + "loss": 1.2431, + "step": 1598 + }, + { + "epoch": 0.11095690791756298, + "grad_norm": 4.663984298706055, + "learning_rate": 0.00018485549132947978, + "loss": 1.3999, + "step": 1599 + }, + { + "epoch": 0.11102629935465963, + "grad_norm": 4.831454277038574, + "learning_rate": 0.00018497109826589596, + "loss": 1.0368, + "step": 1600 + }, + { + "epoch": 0.1110956907917563, + "grad_norm": 6.982776165008545, + "learning_rate": 0.00018508670520231215, + "loss": 1.1099, + "step": 1601 + }, + { + "epoch": 0.11116508222885296, + "grad_norm": 5.0587849617004395, + "learning_rate": 0.00018520231213872834, + "loss": 1.2952, + "step": 1602 + }, + { + "epoch": 0.11123447366594962, + "grad_norm": 4.273077964782715, + "learning_rate": 0.0001853179190751445, + "loss": 0.9836, + "step": 1603 + }, + { + "epoch": 0.11130386510304628, + "grad_norm": 6.5486016273498535, + "learning_rate": 0.0001854335260115607, + "loss": 1.4422, + "step": 1604 + }, + { + "epoch": 0.11137325654014295, + "grad_norm": 4.959231376647949, + "learning_rate": 0.0001855491329479769, + "loss": 1.4634, + "step": 1605 + }, + { + "epoch": 0.11144264797723961, + "grad_norm": 4.24554443359375, + "learning_rate": 0.00018566473988439308, + "loss": 1.4267, + "step": 1606 + }, + { + "epoch": 0.11151203941433627, + "grad_norm": 3.7120521068573, + "learning_rate": 0.00018578034682080924, + "loss": 1.2318, + "step": 1607 + }, + { + "epoch": 0.11158143085143293, + "grad_norm": 4.5050249099731445, + "learning_rate": 0.00018589595375722543, + "loss": 1.4026, + "step": 1608 + }, + { + "epoch": 0.1116508222885296, + "grad_norm": 5.62978458404541, + "learning_rate": 0.00018601156069364164, + "loss": 1.2606, + "step": 1609 + }, + { + "epoch": 0.11172021372562625, + "grad_norm": 4.770850658416748, + "learning_rate": 0.00018612716763005783, + "loss": 1.2836, + "step": 1610 + }, + { + "epoch": 0.11178960516272292, + "grad_norm": 4.694400787353516, + "learning_rate": 0.00018624277456647399, + "loss": 1.1707, + "step": 1611 + }, + { + "epoch": 0.11185899659981959, + "grad_norm": 4.735474109649658, + "learning_rate": 0.00018635838150289017, + "loss": 1.1138, + "step": 1612 + }, + { + "epoch": 0.11192838803691624, + "grad_norm": 4.376530647277832, + "learning_rate": 0.00018647398843930636, + "loss": 1.4713, + "step": 1613 + }, + { + "epoch": 0.1119977794740129, + "grad_norm": 4.793402671813965, + "learning_rate": 0.00018658959537572257, + "loss": 1.3229, + "step": 1614 + }, + { + "epoch": 0.11206717091110957, + "grad_norm": 4.4789652824401855, + "learning_rate": 0.00018670520231213873, + "loss": 1.2932, + "step": 1615 + }, + { + "epoch": 0.11213656234820624, + "grad_norm": 4.418912410736084, + "learning_rate": 0.00018682080924855492, + "loss": 1.1622, + "step": 1616 + }, + { + "epoch": 0.11220595378530289, + "grad_norm": 5.6709771156311035, + "learning_rate": 0.0001869364161849711, + "loss": 1.4381, + "step": 1617 + }, + { + "epoch": 0.11227534522239956, + "grad_norm": 4.2884039878845215, + "learning_rate": 0.0001870520231213873, + "loss": 1.0249, + "step": 1618 + }, + { + "epoch": 0.11234473665949622, + "grad_norm": 4.7682976722717285, + "learning_rate": 0.00018716763005780348, + "loss": 1.0948, + "step": 1619 + }, + { + "epoch": 0.11241412809659287, + "grad_norm": 5.118902206420898, + "learning_rate": 0.00018728323699421966, + "loss": 1.6224, + "step": 1620 + }, + { + "epoch": 0.11248351953368954, + "grad_norm": 7.612045764923096, + "learning_rate": 0.00018739884393063585, + "loss": 1.3327, + "step": 1621 + }, + { + "epoch": 0.11255291097078621, + "grad_norm": 5.160789966583252, + "learning_rate": 0.00018751445086705203, + "loss": 1.2384, + "step": 1622 + }, + { + "epoch": 0.11262230240788287, + "grad_norm": 4.652639865875244, + "learning_rate": 0.0001876300578034682, + "loss": 1.2654, + "step": 1623 + }, + { + "epoch": 0.11269169384497953, + "grad_norm": 5.502891540527344, + "learning_rate": 0.0001877456647398844, + "loss": 1.4459, + "step": 1624 + }, + { + "epoch": 0.11276108528207619, + "grad_norm": 5.714651584625244, + "learning_rate": 0.0001878612716763006, + "loss": 1.3706, + "step": 1625 + }, + { + "epoch": 0.11283047671917286, + "grad_norm": 4.559238433837891, + "learning_rate": 0.00018797687861271678, + "loss": 0.9931, + "step": 1626 + }, + { + "epoch": 0.11289986815626951, + "grad_norm": 6.018820285797119, + "learning_rate": 0.00018809248554913294, + "loss": 1.3887, + "step": 1627 + }, + { + "epoch": 0.11296925959336618, + "grad_norm": 4.331910133361816, + "learning_rate": 0.00018820809248554912, + "loss": 1.3409, + "step": 1628 + }, + { + "epoch": 0.11303865103046284, + "grad_norm": 5.233016014099121, + "learning_rate": 0.00018832369942196534, + "loss": 1.0751, + "step": 1629 + }, + { + "epoch": 0.1131080424675595, + "grad_norm": 4.454774856567383, + "learning_rate": 0.00018843930635838152, + "loss": 0.9595, + "step": 1630 + }, + { + "epoch": 0.11317743390465616, + "grad_norm": 4.624378204345703, + "learning_rate": 0.00018855491329479768, + "loss": 1.2631, + "step": 1631 + }, + { + "epoch": 0.11324682534175283, + "grad_norm": 4.560770511627197, + "learning_rate": 0.00018867052023121387, + "loss": 1.3022, + "step": 1632 + }, + { + "epoch": 0.1133162167788495, + "grad_norm": 4.10243558883667, + "learning_rate": 0.00018878612716763006, + "loss": 1.0816, + "step": 1633 + }, + { + "epoch": 0.11338560821594615, + "grad_norm": 4.916540622711182, + "learning_rate": 0.00018890173410404627, + "loss": 1.3255, + "step": 1634 + }, + { + "epoch": 0.11345499965304282, + "grad_norm": 4.575097560882568, + "learning_rate": 0.00018901734104046245, + "loss": 1.0048, + "step": 1635 + }, + { + "epoch": 0.11352439109013948, + "grad_norm": 7.047465801239014, + "learning_rate": 0.00018913294797687861, + "loss": 1.5505, + "step": 1636 + }, + { + "epoch": 0.11359378252723613, + "grad_norm": 4.393843650817871, + "learning_rate": 0.0001892485549132948, + "loss": 1.2132, + "step": 1637 + }, + { + "epoch": 0.1136631739643328, + "grad_norm": 5.415940284729004, + "learning_rate": 0.00018936416184971099, + "loss": 1.3439, + "step": 1638 + }, + { + "epoch": 0.11373256540142947, + "grad_norm": 4.4054179191589355, + "learning_rate": 0.0001894797687861272, + "loss": 1.2318, + "step": 1639 + }, + { + "epoch": 0.11380195683852612, + "grad_norm": 5.669269561767578, + "learning_rate": 0.00018959537572254336, + "loss": 1.1926, + "step": 1640 + }, + { + "epoch": 0.11387134827562279, + "grad_norm": 4.436100959777832, + "learning_rate": 0.00018971098265895954, + "loss": 1.1062, + "step": 1641 + }, + { + "epoch": 0.11394073971271945, + "grad_norm": 4.311570167541504, + "learning_rate": 0.00018982658959537573, + "loss": 1.0705, + "step": 1642 + }, + { + "epoch": 0.11401013114981612, + "grad_norm": 4.316382884979248, + "learning_rate": 0.00018994219653179192, + "loss": 1.0283, + "step": 1643 + }, + { + "epoch": 0.11407952258691277, + "grad_norm": 4.616250514984131, + "learning_rate": 0.0001900578034682081, + "loss": 1.1839, + "step": 1644 + }, + { + "epoch": 0.11414891402400944, + "grad_norm": 5.73546028137207, + "learning_rate": 0.0001901734104046243, + "loss": 0.975, + "step": 1645 + }, + { + "epoch": 0.1142183054611061, + "grad_norm": 6.285916805267334, + "learning_rate": 0.00019028901734104048, + "loss": 1.2804, + "step": 1646 + }, + { + "epoch": 0.11428769689820276, + "grad_norm": 4.6898956298828125, + "learning_rate": 0.00019040462427745666, + "loss": 1.0916, + "step": 1647 + }, + { + "epoch": 0.11435708833529942, + "grad_norm": 6.931802749633789, + "learning_rate": 0.00019052023121387285, + "loss": 0.9342, + "step": 1648 + }, + { + "epoch": 0.11442647977239609, + "grad_norm": 4.774214267730713, + "learning_rate": 0.00019063583815028903, + "loss": 1.1514, + "step": 1649 + }, + { + "epoch": 0.11449587120949274, + "grad_norm": 4.291598320007324, + "learning_rate": 0.00019075144508670522, + "loss": 1.0582, + "step": 1650 + }, + { + "epoch": 0.11456526264658941, + "grad_norm": 5.235130786895752, + "learning_rate": 0.0001908670520231214, + "loss": 1.146, + "step": 1651 + }, + { + "epoch": 0.11463465408368607, + "grad_norm": 4.094526290893555, + "learning_rate": 0.00019098265895953757, + "loss": 1.0217, + "step": 1652 + }, + { + "epoch": 0.11470404552078274, + "grad_norm": 4.735103607177734, + "learning_rate": 0.00019109826589595375, + "loss": 1.0607, + "step": 1653 + }, + { + "epoch": 0.1147734369578794, + "grad_norm": 3.8545892238616943, + "learning_rate": 0.00019121387283236997, + "loss": 0.9198, + "step": 1654 + }, + { + "epoch": 0.11484282839497606, + "grad_norm": 4.566482067108154, + "learning_rate": 0.00019132947976878615, + "loss": 1.2102, + "step": 1655 + }, + { + "epoch": 0.11491221983207273, + "grad_norm": 6.317206382751465, + "learning_rate": 0.0001914450867052023, + "loss": 1.1098, + "step": 1656 + }, + { + "epoch": 0.11498161126916938, + "grad_norm": 5.731967449188232, + "learning_rate": 0.0001915606936416185, + "loss": 0.9941, + "step": 1657 + }, + { + "epoch": 0.11505100270626605, + "grad_norm": 4.084891319274902, + "learning_rate": 0.00019167630057803468, + "loss": 1.1719, + "step": 1658 + }, + { + "epoch": 0.11512039414336271, + "grad_norm": 7.0105109214782715, + "learning_rate": 0.0001917919075144509, + "loss": 1.6145, + "step": 1659 + }, + { + "epoch": 0.11518978558045938, + "grad_norm": 5.558907508850098, + "learning_rate": 0.00019190751445086706, + "loss": 1.1661, + "step": 1660 + }, + { + "epoch": 0.11525917701755603, + "grad_norm": 5.555027008056641, + "learning_rate": 0.00019202312138728324, + "loss": 1.1873, + "step": 1661 + }, + { + "epoch": 0.1153285684546527, + "grad_norm": 6.358254432678223, + "learning_rate": 0.00019213872832369943, + "loss": 0.8901, + "step": 1662 + }, + { + "epoch": 0.11539795989174936, + "grad_norm": 4.983002185821533, + "learning_rate": 0.00019225433526011561, + "loss": 1.1872, + "step": 1663 + }, + { + "epoch": 0.11546735132884602, + "grad_norm": 6.2524638175964355, + "learning_rate": 0.0001923699421965318, + "loss": 1.2322, + "step": 1664 + }, + { + "epoch": 0.11553674276594268, + "grad_norm": 5.625535488128662, + "learning_rate": 0.000192485549132948, + "loss": 1.1956, + "step": 1665 + }, + { + "epoch": 0.11560613420303935, + "grad_norm": 4.0616350173950195, + "learning_rate": 0.00019260115606936417, + "loss": 0.9408, + "step": 1666 + }, + { + "epoch": 0.115675525640136, + "grad_norm": 4.270815849304199, + "learning_rate": 0.00019271676300578036, + "loss": 1.0944, + "step": 1667 + }, + { + "epoch": 0.11574491707723267, + "grad_norm": 6.048637866973877, + "learning_rate": 0.00019283236994219655, + "loss": 1.172, + "step": 1668 + }, + { + "epoch": 0.11581430851432933, + "grad_norm": 4.926621913909912, + "learning_rate": 0.00019294797687861273, + "loss": 1.117, + "step": 1669 + }, + { + "epoch": 0.115883699951426, + "grad_norm": 5.455409049987793, + "learning_rate": 0.00019306358381502892, + "loss": 1.4235, + "step": 1670 + }, + { + "epoch": 0.11595309138852265, + "grad_norm": 4.329158782958984, + "learning_rate": 0.0001931791907514451, + "loss": 1.2844, + "step": 1671 + }, + { + "epoch": 0.11602248282561932, + "grad_norm": 4.471502780914307, + "learning_rate": 0.00019329479768786126, + "loss": 1.0405, + "step": 1672 + }, + { + "epoch": 0.11609187426271599, + "grad_norm": 4.257415294647217, + "learning_rate": 0.00019341040462427748, + "loss": 1.2872, + "step": 1673 + }, + { + "epoch": 0.11616126569981264, + "grad_norm": 3.7341647148132324, + "learning_rate": 0.00019352601156069366, + "loss": 0.9729, + "step": 1674 + }, + { + "epoch": 0.1162306571369093, + "grad_norm": 4.023412227630615, + "learning_rate": 0.00019364161849710985, + "loss": 0.9531, + "step": 1675 + }, + { + "epoch": 0.11630004857400597, + "grad_norm": 5.580128192901611, + "learning_rate": 0.000193757225433526, + "loss": 1.5275, + "step": 1676 + }, + { + "epoch": 0.11636944001110262, + "grad_norm": 4.440971851348877, + "learning_rate": 0.0001938728323699422, + "loss": 1.2182, + "step": 1677 + }, + { + "epoch": 0.11643883144819929, + "grad_norm": 4.786009788513184, + "learning_rate": 0.00019398843930635838, + "loss": 1.423, + "step": 1678 + }, + { + "epoch": 0.11650822288529596, + "grad_norm": 4.675642967224121, + "learning_rate": 0.0001941040462427746, + "loss": 1.3747, + "step": 1679 + }, + { + "epoch": 0.11657761432239262, + "grad_norm": 5.025696754455566, + "learning_rate": 0.00019421965317919075, + "loss": 1.2304, + "step": 1680 + }, + { + "epoch": 0.11664700575948928, + "grad_norm": 5.906426906585693, + "learning_rate": 0.00019433526011560694, + "loss": 1.5213, + "step": 1681 + }, + { + "epoch": 0.11671639719658594, + "grad_norm": 6.198960781097412, + "learning_rate": 0.00019445086705202313, + "loss": 0.8825, + "step": 1682 + }, + { + "epoch": 0.11678578863368261, + "grad_norm": 4.66171407699585, + "learning_rate": 0.0001945664739884393, + "loss": 1.2983, + "step": 1683 + }, + { + "epoch": 0.11685518007077926, + "grad_norm": 4.505459308624268, + "learning_rate": 0.0001946820809248555, + "loss": 1.2738, + "step": 1684 + }, + { + "epoch": 0.11692457150787593, + "grad_norm": 5.741061210632324, + "learning_rate": 0.00019479768786127168, + "loss": 1.3236, + "step": 1685 + }, + { + "epoch": 0.1169939629449726, + "grad_norm": 4.126547813415527, + "learning_rate": 0.00019491329479768787, + "loss": 1.288, + "step": 1686 + }, + { + "epoch": 0.11706335438206926, + "grad_norm": 4.419602394104004, + "learning_rate": 0.00019502890173410406, + "loss": 1.0733, + "step": 1687 + }, + { + "epoch": 0.11713274581916591, + "grad_norm": 5.6675496101379395, + "learning_rate": 0.00019514450867052024, + "loss": 1.2725, + "step": 1688 + }, + { + "epoch": 0.11720213725626258, + "grad_norm": 5.722709655761719, + "learning_rate": 0.00019526011560693643, + "loss": 1.4484, + "step": 1689 + }, + { + "epoch": 0.11727152869335924, + "grad_norm": 4.8036041259765625, + "learning_rate": 0.00019537572254335262, + "loss": 1.3952, + "step": 1690 + }, + { + "epoch": 0.1173409201304559, + "grad_norm": 5.215405464172363, + "learning_rate": 0.0001954913294797688, + "loss": 1.2747, + "step": 1691 + }, + { + "epoch": 0.11741031156755256, + "grad_norm": 4.383759021759033, + "learning_rate": 0.00019560693641618496, + "loss": 1.2283, + "step": 1692 + }, + { + "epoch": 0.11747970300464923, + "grad_norm": 5.7893781661987305, + "learning_rate": 0.00019572254335260117, + "loss": 1.2551, + "step": 1693 + }, + { + "epoch": 0.11754909444174588, + "grad_norm": 9.186378479003906, + "learning_rate": 0.00019583815028901736, + "loss": 1.6512, + "step": 1694 + }, + { + "epoch": 0.11761848587884255, + "grad_norm": 4.997702598571777, + "learning_rate": 0.00019595375722543355, + "loss": 1.1698, + "step": 1695 + }, + { + "epoch": 0.11768787731593922, + "grad_norm": 3.4912078380584717, + "learning_rate": 0.0001960693641618497, + "loss": 1.0625, + "step": 1696 + }, + { + "epoch": 0.11775726875303588, + "grad_norm": 4.322317600250244, + "learning_rate": 0.0001961849710982659, + "loss": 1.0689, + "step": 1697 + }, + { + "epoch": 0.11782666019013253, + "grad_norm": 6.000354290008545, + "learning_rate": 0.0001963005780346821, + "loss": 1.0875, + "step": 1698 + }, + { + "epoch": 0.1178960516272292, + "grad_norm": 4.114992141723633, + "learning_rate": 0.0001964161849710983, + "loss": 1.1257, + "step": 1699 + }, + { + "epoch": 0.11796544306432587, + "grad_norm": 4.542341232299805, + "learning_rate": 0.00019653179190751445, + "loss": 1.2252, + "step": 1700 + }, + { + "epoch": 0.11803483450142252, + "grad_norm": 4.232224941253662, + "learning_rate": 0.00019664739884393064, + "loss": 1.0242, + "step": 1701 + }, + { + "epoch": 0.11810422593851919, + "grad_norm": 3.798354148864746, + "learning_rate": 0.00019676300578034682, + "loss": 1.0915, + "step": 1702 + }, + { + "epoch": 0.11817361737561585, + "grad_norm": 5.4250359535217285, + "learning_rate": 0.000196878612716763, + "loss": 1.2657, + "step": 1703 + }, + { + "epoch": 0.1182430088127125, + "grad_norm": 4.418467998504639, + "learning_rate": 0.0001969942196531792, + "loss": 1.3514, + "step": 1704 + }, + { + "epoch": 0.11831240024980917, + "grad_norm": 6.482585906982422, + "learning_rate": 0.00019710982658959538, + "loss": 1.5078, + "step": 1705 + }, + { + "epoch": 0.11838179168690584, + "grad_norm": 7.016510009765625, + "learning_rate": 0.00019722543352601157, + "loss": 1.378, + "step": 1706 + }, + { + "epoch": 0.1184511831240025, + "grad_norm": 5.77690315246582, + "learning_rate": 0.00019734104046242775, + "loss": 1.4068, + "step": 1707 + }, + { + "epoch": 0.11852057456109916, + "grad_norm": 4.287348747253418, + "learning_rate": 0.00019745664739884394, + "loss": 0.9507, + "step": 1708 + }, + { + "epoch": 0.11858996599819582, + "grad_norm": 4.255311489105225, + "learning_rate": 0.00019757225433526013, + "loss": 1.21, + "step": 1709 + }, + { + "epoch": 0.11865935743529249, + "grad_norm": 3.9349842071533203, + "learning_rate": 0.0001976878612716763, + "loss": 0.9742, + "step": 1710 + }, + { + "epoch": 0.11872874887238914, + "grad_norm": 4.598840236663818, + "learning_rate": 0.0001978034682080925, + "loss": 1.0981, + "step": 1711 + }, + { + "epoch": 0.11879814030948581, + "grad_norm": 6.435896396636963, + "learning_rate": 0.00019791907514450866, + "loss": 1.2213, + "step": 1712 + }, + { + "epoch": 0.11886753174658247, + "grad_norm": 3.88681697845459, + "learning_rate": 0.00019803468208092487, + "loss": 0.8552, + "step": 1713 + }, + { + "epoch": 0.11893692318367913, + "grad_norm": 4.375607490539551, + "learning_rate": 0.00019815028901734106, + "loss": 1.0917, + "step": 1714 + }, + { + "epoch": 0.1190063146207758, + "grad_norm": 6.752068519592285, + "learning_rate": 0.00019826589595375724, + "loss": 1.2589, + "step": 1715 + }, + { + "epoch": 0.11907570605787246, + "grad_norm": 4.952205181121826, + "learning_rate": 0.0001983815028901734, + "loss": 1.4352, + "step": 1716 + }, + { + "epoch": 0.11914509749496913, + "grad_norm": 5.684750080108643, + "learning_rate": 0.0001984971098265896, + "loss": 1.2564, + "step": 1717 + }, + { + "epoch": 0.11921448893206578, + "grad_norm": 4.4039387702941895, + "learning_rate": 0.0001986127167630058, + "loss": 1.2321, + "step": 1718 + }, + { + "epoch": 0.11928388036916245, + "grad_norm": 5.8380255699157715, + "learning_rate": 0.000198728323699422, + "loss": 1.4501, + "step": 1719 + }, + { + "epoch": 0.11935327180625911, + "grad_norm": 5.485328674316406, + "learning_rate": 0.00019884393063583815, + "loss": 1.1449, + "step": 1720 + }, + { + "epoch": 0.11942266324335576, + "grad_norm": 3.8743977546691895, + "learning_rate": 0.00019895953757225433, + "loss": 1.1265, + "step": 1721 + }, + { + "epoch": 0.11949205468045243, + "grad_norm": 4.739249229431152, + "learning_rate": 0.00019907514450867052, + "loss": 1.1834, + "step": 1722 + }, + { + "epoch": 0.1195614461175491, + "grad_norm": 5.05216646194458, + "learning_rate": 0.00019919075144508673, + "loss": 1.1324, + "step": 1723 + }, + { + "epoch": 0.11963083755464576, + "grad_norm": 4.8187174797058105, + "learning_rate": 0.0001993063583815029, + "loss": 1.2505, + "step": 1724 + }, + { + "epoch": 0.11970022899174242, + "grad_norm": 4.752727508544922, + "learning_rate": 0.00019942196531791908, + "loss": 1.0532, + "step": 1725 + }, + { + "epoch": 0.11976962042883908, + "grad_norm": 9.172891616821289, + "learning_rate": 0.00019953757225433526, + "loss": 1.4113, + "step": 1726 + }, + { + "epoch": 0.11983901186593575, + "grad_norm": 4.791268348693848, + "learning_rate": 0.00019965317919075145, + "loss": 1.1311, + "step": 1727 + }, + { + "epoch": 0.1199084033030324, + "grad_norm": 5.6515703201293945, + "learning_rate": 0.00019976878612716764, + "loss": 0.8822, + "step": 1728 + }, + { + "epoch": 0.11997779474012907, + "grad_norm": 4.655667304992676, + "learning_rate": 0.00019988439306358382, + "loss": 1.2328, + "step": 1729 + }, + { + "epoch": 0.12004718617722573, + "grad_norm": 7.154439449310303, + "learning_rate": 0.0002, + "loss": 1.1952, + "step": 1730 + }, + { + "epoch": 0.12011657761432239, + "grad_norm": 4.634467124938965, + "learning_rate": 0.0001999999998421558, + "loss": 1.3404, + "step": 1731 + }, + { + "epoch": 0.12018596905141905, + "grad_norm": 3.6990907192230225, + "learning_rate": 0.00019999999936862322, + "loss": 1.1331, + "step": 1732 + }, + { + "epoch": 0.12025536048851572, + "grad_norm": 4.708794593811035, + "learning_rate": 0.00019999999857940224, + "loss": 1.1546, + "step": 1733 + }, + { + "epoch": 0.12032475192561239, + "grad_norm": 5.192391872406006, + "learning_rate": 0.00019999999747449284, + "loss": 1.4318, + "step": 1734 + }, + { + "epoch": 0.12039414336270904, + "grad_norm": 4.027135848999023, + "learning_rate": 0.00019999999605389507, + "loss": 1.0918, + "step": 1735 + }, + { + "epoch": 0.1204635347998057, + "grad_norm": 4.134101867675781, + "learning_rate": 0.0001999999943176089, + "loss": 1.2115, + "step": 1736 + }, + { + "epoch": 0.12053292623690237, + "grad_norm": 5.601803302764893, + "learning_rate": 0.00019999999226563437, + "loss": 1.0119, + "step": 1737 + }, + { + "epoch": 0.12060231767399902, + "grad_norm": 4.343542575836182, + "learning_rate": 0.00019999998989797145, + "loss": 1.4219, + "step": 1738 + }, + { + "epoch": 0.12067170911109569, + "grad_norm": 5.5581817626953125, + "learning_rate": 0.00019999998721462016, + "loss": 1.2104, + "step": 1739 + }, + { + "epoch": 0.12074110054819236, + "grad_norm": 6.373262405395508, + "learning_rate": 0.00019999998421558053, + "loss": 1.3013, + "step": 1740 + }, + { + "epoch": 0.12081049198528901, + "grad_norm": 4.858711242675781, + "learning_rate": 0.00019999998090085253, + "loss": 1.3626, + "step": 1741 + }, + { + "epoch": 0.12087988342238568, + "grad_norm": 4.365847110748291, + "learning_rate": 0.00019999997727043623, + "loss": 1.4221, + "step": 1742 + }, + { + "epoch": 0.12094927485948234, + "grad_norm": 4.3113112449646, + "learning_rate": 0.00019999997332433158, + "loss": 1.1357, + "step": 1743 + }, + { + "epoch": 0.12101866629657901, + "grad_norm": 5.688600063323975, + "learning_rate": 0.00019999996906253862, + "loss": 1.3333, + "step": 1744 + }, + { + "epoch": 0.12108805773367566, + "grad_norm": 5.437167167663574, + "learning_rate": 0.00019999996448505732, + "loss": 1.5244, + "step": 1745 + }, + { + "epoch": 0.12115744917077233, + "grad_norm": 4.510493755340576, + "learning_rate": 0.0001999999595918878, + "loss": 1.4842, + "step": 1746 + }, + { + "epoch": 0.121226840607869, + "grad_norm": 4.976790904998779, + "learning_rate": 0.00019999995438302998, + "loss": 1.5562, + "step": 1747 + }, + { + "epoch": 0.12129623204496565, + "grad_norm": 3.8931503295898438, + "learning_rate": 0.0001999999488584839, + "loss": 1.2063, + "step": 1748 + }, + { + "epoch": 0.12136562348206231, + "grad_norm": 4.248028755187988, + "learning_rate": 0.00019999994301824958, + "loss": 1.0733, + "step": 1749 + }, + { + "epoch": 0.12143501491915898, + "grad_norm": 5.6794657707214355, + "learning_rate": 0.00019999993686232705, + "loss": 1.4553, + "step": 1750 + }, + { + "epoch": 0.12150440635625563, + "grad_norm": 3.816603899002075, + "learning_rate": 0.00019999993039071634, + "loss": 1.1278, + "step": 1751 + }, + { + "epoch": 0.1215737977933523, + "grad_norm": 4.838738918304443, + "learning_rate": 0.00019999992360341743, + "loss": 1.2358, + "step": 1752 + }, + { + "epoch": 0.12164318923044896, + "grad_norm": 5.040391445159912, + "learning_rate": 0.00019999991650043036, + "loss": 1.1776, + "step": 1753 + }, + { + "epoch": 0.12171258066754563, + "grad_norm": 6.3992109298706055, + "learning_rate": 0.00019999990908175515, + "loss": 1.1453, + "step": 1754 + }, + { + "epoch": 0.12178197210464228, + "grad_norm": 5.533609390258789, + "learning_rate": 0.00019999990134739187, + "loss": 1.0908, + "step": 1755 + }, + { + "epoch": 0.12185136354173895, + "grad_norm": 4.807620048522949, + "learning_rate": 0.00019999989329734047, + "loss": 1.09, + "step": 1756 + }, + { + "epoch": 0.12192075497883562, + "grad_norm": 6.759379863739014, + "learning_rate": 0.00019999988493160103, + "loss": 1.0544, + "step": 1757 + }, + { + "epoch": 0.12199014641593227, + "grad_norm": 4.997951507568359, + "learning_rate": 0.00019999987625017354, + "loss": 1.2597, + "step": 1758 + }, + { + "epoch": 0.12205953785302893, + "grad_norm": 5.111935615539551, + "learning_rate": 0.00019999986725305804, + "loss": 1.0413, + "step": 1759 + }, + { + "epoch": 0.1221289292901256, + "grad_norm": 5.466860771179199, + "learning_rate": 0.00019999985794025459, + "loss": 0.9709, + "step": 1760 + }, + { + "epoch": 0.12219832072722227, + "grad_norm": 5.01601505279541, + "learning_rate": 0.00019999984831176314, + "loss": 1.3932, + "step": 1761 + }, + { + "epoch": 0.12226771216431892, + "grad_norm": 4.798618793487549, + "learning_rate": 0.0001999998383675838, + "loss": 1.2012, + "step": 1762 + }, + { + "epoch": 0.12233710360141559, + "grad_norm": 5.827358245849609, + "learning_rate": 0.0001999998281077166, + "loss": 1.1808, + "step": 1763 + }, + { + "epoch": 0.12240649503851225, + "grad_norm": 4.377131462097168, + "learning_rate": 0.00019999981753216148, + "loss": 1.0072, + "step": 1764 + }, + { + "epoch": 0.1224758864756089, + "grad_norm": 5.4143171310424805, + "learning_rate": 0.0001999998066409186, + "loss": 1.4439, + "step": 1765 + }, + { + "epoch": 0.12254527791270557, + "grad_norm": 4.036984443664551, + "learning_rate": 0.00019999979543398788, + "loss": 0.9848, + "step": 1766 + }, + { + "epoch": 0.12261466934980224, + "grad_norm": 5.016209602355957, + "learning_rate": 0.00019999978391136943, + "loss": 1.2602, + "step": 1767 + }, + { + "epoch": 0.12268406078689889, + "grad_norm": 4.1098952293396, + "learning_rate": 0.00019999977207306327, + "loss": 1.0398, + "step": 1768 + }, + { + "epoch": 0.12275345222399556, + "grad_norm": 6.015614032745361, + "learning_rate": 0.00019999975991906945, + "loss": 1.0893, + "step": 1769 + }, + { + "epoch": 0.12282284366109222, + "grad_norm": 4.0042219161987305, + "learning_rate": 0.00019999974744938797, + "loss": 1.0487, + "step": 1770 + }, + { + "epoch": 0.12289223509818889, + "grad_norm": 4.739800930023193, + "learning_rate": 0.00019999973466401888, + "loss": 1.2532, + "step": 1771 + }, + { + "epoch": 0.12296162653528554, + "grad_norm": 5.005980968475342, + "learning_rate": 0.00019999972156296223, + "loss": 1.2426, + "step": 1772 + }, + { + "epoch": 0.12303101797238221, + "grad_norm": 4.746572971343994, + "learning_rate": 0.00019999970814621806, + "loss": 1.207, + "step": 1773 + }, + { + "epoch": 0.12310040940947888, + "grad_norm": 6.977224826812744, + "learning_rate": 0.00019999969441378642, + "loss": 1.6534, + "step": 1774 + }, + { + "epoch": 0.12316980084657553, + "grad_norm": 4.926757335662842, + "learning_rate": 0.00019999968036566736, + "loss": 1.3024, + "step": 1775 + }, + { + "epoch": 0.1232391922836722, + "grad_norm": 3.976806640625, + "learning_rate": 0.0001999996660018609, + "loss": 0.9004, + "step": 1776 + }, + { + "epoch": 0.12330858372076886, + "grad_norm": 4.404874324798584, + "learning_rate": 0.0001999996513223671, + "loss": 1.081, + "step": 1777 + }, + { + "epoch": 0.12337797515786551, + "grad_norm": 4.389124870300293, + "learning_rate": 0.000199999636327186, + "loss": 1.1337, + "step": 1778 + }, + { + "epoch": 0.12344736659496218, + "grad_norm": 5.315927028656006, + "learning_rate": 0.00019999962101631766, + "loss": 1.1819, + "step": 1779 + }, + { + "epoch": 0.12351675803205885, + "grad_norm": 7.160079002380371, + "learning_rate": 0.00019999960538976213, + "loss": 1.7469, + "step": 1780 + }, + { + "epoch": 0.12358614946915551, + "grad_norm": 4.22015905380249, + "learning_rate": 0.0001999995894475194, + "loss": 0.9496, + "step": 1781 + }, + { + "epoch": 0.12365554090625216, + "grad_norm": 4.3080668449401855, + "learning_rate": 0.00019999957318958962, + "loss": 1.0105, + "step": 1782 + }, + { + "epoch": 0.12372493234334883, + "grad_norm": 5.406104564666748, + "learning_rate": 0.00019999955661597276, + "loss": 1.5524, + "step": 1783 + }, + { + "epoch": 0.1237943237804455, + "grad_norm": 4.928981781005859, + "learning_rate": 0.00019999953972666892, + "loss": 1.2236, + "step": 1784 + }, + { + "epoch": 0.12386371521754215, + "grad_norm": 4.035689830780029, + "learning_rate": 0.0001999995225216781, + "loss": 1.2857, + "step": 1785 + }, + { + "epoch": 0.12393310665463882, + "grad_norm": 4.5347466468811035, + "learning_rate": 0.00019999950500100043, + "loss": 1.2703, + "step": 1786 + }, + { + "epoch": 0.12400249809173548, + "grad_norm": 4.754541397094727, + "learning_rate": 0.0001999994871646359, + "loss": 1.3532, + "step": 1787 + }, + { + "epoch": 0.12407188952883214, + "grad_norm": 5.977343559265137, + "learning_rate": 0.0001999994690125846, + "loss": 1.039, + "step": 1788 + }, + { + "epoch": 0.1241412809659288, + "grad_norm": 5.301023960113525, + "learning_rate": 0.00019999945054484657, + "loss": 1.2345, + "step": 1789 + }, + { + "epoch": 0.12421067240302547, + "grad_norm": 5.1552605628967285, + "learning_rate": 0.00019999943176142188, + "loss": 1.184, + "step": 1790 + }, + { + "epoch": 0.12428006384012213, + "grad_norm": 4.307857513427734, + "learning_rate": 0.0001999994126623106, + "loss": 1.1881, + "step": 1791 + }, + { + "epoch": 0.12434945527721879, + "grad_norm": 4.590317249298096, + "learning_rate": 0.00019999939324751275, + "loss": 1.2352, + "step": 1792 + }, + { + "epoch": 0.12441884671431545, + "grad_norm": 6.5115275382995605, + "learning_rate": 0.00019999937351702844, + "loss": 1.2186, + "step": 1793 + }, + { + "epoch": 0.12448823815141212, + "grad_norm": 11.047599792480469, + "learning_rate": 0.0001999993534708577, + "loss": 1.0435, + "step": 1794 + }, + { + "epoch": 0.12455762958850877, + "grad_norm": 7.909645080566406, + "learning_rate": 0.0001999993331090006, + "loss": 1.3687, + "step": 1795 + }, + { + "epoch": 0.12462702102560544, + "grad_norm": 7.4458327293396, + "learning_rate": 0.00019999931243145722, + "loss": 1.2761, + "step": 1796 + }, + { + "epoch": 0.1246964124627021, + "grad_norm": 4.4001288414001465, + "learning_rate": 0.0001999992914382276, + "loss": 1.0261, + "step": 1797 + }, + { + "epoch": 0.12476580389979877, + "grad_norm": 5.1328887939453125, + "learning_rate": 0.00019999927012931184, + "loss": 1.6494, + "step": 1798 + }, + { + "epoch": 0.12483519533689542, + "grad_norm": 3.882904529571533, + "learning_rate": 0.00019999924850470997, + "loss": 0.9388, + "step": 1799 + }, + { + "epoch": 0.12490458677399209, + "grad_norm": 3.6971843242645264, + "learning_rate": 0.00019999922656442208, + "loss": 1.0675, + "step": 1800 + }, + { + "epoch": 0.12497397821108876, + "grad_norm": 7.596257209777832, + "learning_rate": 0.00019999920430844822, + "loss": 0.7564, + "step": 1801 + }, + { + "epoch": 0.1250433696481854, + "grad_norm": 4.921849250793457, + "learning_rate": 0.00019999918173678848, + "loss": 1.3346, + "step": 1802 + }, + { + "epoch": 0.12511276108528208, + "grad_norm": 4.233412742614746, + "learning_rate": 0.00019999915884944292, + "loss": 1.1952, + "step": 1803 + }, + { + "epoch": 0.12518215252237874, + "grad_norm": 4.708458423614502, + "learning_rate": 0.00019999913564641163, + "loss": 1.3325, + "step": 1804 + }, + { + "epoch": 0.1252515439594754, + "grad_norm": 5.348845958709717, + "learning_rate": 0.00019999911212769468, + "loss": 1.4794, + "step": 1805 + }, + { + "epoch": 0.12532093539657208, + "grad_norm": 5.460000038146973, + "learning_rate": 0.00019999908829329213, + "loss": 1.0286, + "step": 1806 + }, + { + "epoch": 0.1253903268336687, + "grad_norm": 9.257468223571777, + "learning_rate": 0.00019999906414320403, + "loss": 1.2954, + "step": 1807 + }, + { + "epoch": 0.12545971827076538, + "grad_norm": 5.93720817565918, + "learning_rate": 0.00019999903967743052, + "loss": 1.2931, + "step": 1808 + }, + { + "epoch": 0.12552910970786205, + "grad_norm": 7.172335624694824, + "learning_rate": 0.00019999901489597163, + "loss": 1.4336, + "step": 1809 + }, + { + "epoch": 0.1255985011449587, + "grad_norm": 4.646325588226318, + "learning_rate": 0.00019999898979882749, + "loss": 1.4192, + "step": 1810 + }, + { + "epoch": 0.12566789258205538, + "grad_norm": 5.1333441734313965, + "learning_rate": 0.00019999896438599807, + "loss": 1.4944, + "step": 1811 + }, + { + "epoch": 0.12573728401915205, + "grad_norm": 4.619052410125732, + "learning_rate": 0.00019999893865748356, + "loss": 1.1759, + "step": 1812 + }, + { + "epoch": 0.1258066754562487, + "grad_norm": 4.269963264465332, + "learning_rate": 0.000199998912613284, + "loss": 1.2488, + "step": 1813 + }, + { + "epoch": 0.12587606689334535, + "grad_norm": 5.678380489349365, + "learning_rate": 0.00019999888625339948, + "loss": 1.3523, + "step": 1814 + }, + { + "epoch": 0.12594545833044202, + "grad_norm": 8.018916130065918, + "learning_rate": 0.00019999885957783006, + "loss": 0.9983, + "step": 1815 + }, + { + "epoch": 0.12601484976753868, + "grad_norm": 4.021700382232666, + "learning_rate": 0.00019999883258657586, + "loss": 1.0895, + "step": 1816 + }, + { + "epoch": 0.12608424120463535, + "grad_norm": 5.908872604370117, + "learning_rate": 0.00019999880527963698, + "loss": 1.2091, + "step": 1817 + }, + { + "epoch": 0.12615363264173202, + "grad_norm": 4.11758279800415, + "learning_rate": 0.00019999877765701342, + "loss": 0.8499, + "step": 1818 + }, + { + "epoch": 0.12622302407882868, + "grad_norm": 11.717463493347168, + "learning_rate": 0.00019999874971870534, + "loss": 1.388, + "step": 1819 + }, + { + "epoch": 0.12629241551592532, + "grad_norm": 5.868113994598389, + "learning_rate": 0.00019999872146471282, + "loss": 1.4117, + "step": 1820 + }, + { + "epoch": 0.126361806953022, + "grad_norm": 7.75643253326416, + "learning_rate": 0.00019999869289503592, + "loss": 1.1106, + "step": 1821 + }, + { + "epoch": 0.12643119839011865, + "grad_norm": 5.7160749435424805, + "learning_rate": 0.00019999866400967475, + "loss": 1.0166, + "step": 1822 + }, + { + "epoch": 0.12650058982721532, + "grad_norm": 7.559929370880127, + "learning_rate": 0.0001999986348086294, + "loss": 1.6058, + "step": 1823 + }, + { + "epoch": 0.126569981264312, + "grad_norm": 8.357643127441406, + "learning_rate": 0.00019999860529189994, + "loss": 1.2653, + "step": 1824 + }, + { + "epoch": 0.12663937270140865, + "grad_norm": 5.762580394744873, + "learning_rate": 0.00019999857545948652, + "loss": 1.3967, + "step": 1825 + }, + { + "epoch": 0.12670876413850532, + "grad_norm": 5.713204860687256, + "learning_rate": 0.0001999985453113892, + "loss": 1.2082, + "step": 1826 + }, + { + "epoch": 0.12677815557560196, + "grad_norm": 4.875223636627197, + "learning_rate": 0.00019999851484760805, + "loss": 1.1188, + "step": 1827 + }, + { + "epoch": 0.12684754701269862, + "grad_norm": 4.723285675048828, + "learning_rate": 0.0001999984840681432, + "loss": 1.3862, + "step": 1828 + }, + { + "epoch": 0.1269169384497953, + "grad_norm": 6.148767948150635, + "learning_rate": 0.00019999845297299476, + "loss": 1.461, + "step": 1829 + }, + { + "epoch": 0.12698632988689196, + "grad_norm": 5.5473198890686035, + "learning_rate": 0.0001999984215621628, + "loss": 1.2112, + "step": 1830 + }, + { + "epoch": 0.12705572132398862, + "grad_norm": 4.594205856323242, + "learning_rate": 0.0001999983898356474, + "loss": 1.065, + "step": 1831 + }, + { + "epoch": 0.1271251127610853, + "grad_norm": 6.269320011138916, + "learning_rate": 0.0001999983577934487, + "loss": 1.1991, + "step": 1832 + }, + { + "epoch": 0.12719450419818196, + "grad_norm": 4.188746452331543, + "learning_rate": 0.0001999983254355668, + "loss": 1.1792, + "step": 1833 + }, + { + "epoch": 0.1272638956352786, + "grad_norm": 6.314086437225342, + "learning_rate": 0.00019999829276200174, + "loss": 1.3125, + "step": 1834 + }, + { + "epoch": 0.12733328707237526, + "grad_norm": 5.618014812469482, + "learning_rate": 0.00019999825977275371, + "loss": 1.0303, + "step": 1835 + }, + { + "epoch": 0.12740267850947193, + "grad_norm": 5.195143699645996, + "learning_rate": 0.0001999982264678228, + "loss": 1.2896, + "step": 1836 + }, + { + "epoch": 0.1274720699465686, + "grad_norm": 4.088600158691406, + "learning_rate": 0.00019999819284720905, + "loss": 1.4769, + "step": 1837 + }, + { + "epoch": 0.12754146138366526, + "grad_norm": 6.2665486335754395, + "learning_rate": 0.00019999815891091262, + "loss": 1.2381, + "step": 1838 + }, + { + "epoch": 0.12761085282076193, + "grad_norm": 6.491621971130371, + "learning_rate": 0.00019999812465893363, + "loss": 1.5377, + "step": 1839 + }, + { + "epoch": 0.1276802442578586, + "grad_norm": 4.916938304901123, + "learning_rate": 0.00019999809009127212, + "loss": 1.2152, + "step": 1840 + }, + { + "epoch": 0.12774963569495523, + "grad_norm": 4.953880310058594, + "learning_rate": 0.00019999805520792824, + "loss": 1.2861, + "step": 1841 + }, + { + "epoch": 0.1278190271320519, + "grad_norm": 6.567453384399414, + "learning_rate": 0.00019999802000890212, + "loss": 1.0602, + "step": 1842 + }, + { + "epoch": 0.12788841856914857, + "grad_norm": 5.649509429931641, + "learning_rate": 0.00019999798449419387, + "loss": 1.2771, + "step": 1843 + }, + { + "epoch": 0.12795781000624523, + "grad_norm": 4.801772594451904, + "learning_rate": 0.00019999794866380354, + "loss": 1.2883, + "step": 1844 + }, + { + "epoch": 0.1280272014433419, + "grad_norm": 4.0963969230651855, + "learning_rate": 0.00019999791251773132, + "loss": 1.2966, + "step": 1845 + }, + { + "epoch": 0.12809659288043856, + "grad_norm": 3.8254971504211426, + "learning_rate": 0.00019999787605597732, + "loss": 1.0987, + "step": 1846 + }, + { + "epoch": 0.1281659843175352, + "grad_norm": 4.503141403198242, + "learning_rate": 0.0001999978392785416, + "loss": 1.1901, + "step": 1847 + }, + { + "epoch": 0.12823537575463187, + "grad_norm": 4.520689964294434, + "learning_rate": 0.00019999780218542428, + "loss": 1.2448, + "step": 1848 + }, + { + "epoch": 0.12830476719172854, + "grad_norm": 5.841653347015381, + "learning_rate": 0.00019999776477662554, + "loss": 1.3291, + "step": 1849 + }, + { + "epoch": 0.1283741586288252, + "grad_norm": 4.473609924316406, + "learning_rate": 0.00019999772705214542, + "loss": 1.3261, + "step": 1850 + }, + { + "epoch": 0.12844355006592187, + "grad_norm": 4.024745941162109, + "learning_rate": 0.00019999768901198407, + "loss": 1.2551, + "step": 1851 + }, + { + "epoch": 0.12851294150301854, + "grad_norm": 4.588242053985596, + "learning_rate": 0.00019999765065614167, + "loss": 0.995, + "step": 1852 + }, + { + "epoch": 0.1285823329401152, + "grad_norm": 4.298238754272461, + "learning_rate": 0.00019999761198461827, + "loss": 1.2404, + "step": 1853 + }, + { + "epoch": 0.12865172437721184, + "grad_norm": 4.487196922302246, + "learning_rate": 0.000199997572997414, + "loss": 1.0813, + "step": 1854 + }, + { + "epoch": 0.1287211158143085, + "grad_norm": 4.337038040161133, + "learning_rate": 0.00019999753369452896, + "loss": 0.9268, + "step": 1855 + }, + { + "epoch": 0.12879050725140517, + "grad_norm": 6.459969520568848, + "learning_rate": 0.00019999749407596335, + "loss": 1.2217, + "step": 1856 + }, + { + "epoch": 0.12885989868850184, + "grad_norm": 4.688127517700195, + "learning_rate": 0.00019999745414171725, + "loss": 1.1604, + "step": 1857 + }, + { + "epoch": 0.1289292901255985, + "grad_norm": 9.921586036682129, + "learning_rate": 0.00019999741389179078, + "loss": 1.0659, + "step": 1858 + }, + { + "epoch": 0.12899868156269517, + "grad_norm": 6.0233635902404785, + "learning_rate": 0.00019999737332618408, + "loss": 1.6549, + "step": 1859 + }, + { + "epoch": 0.12906807299979184, + "grad_norm": 3.9082212448120117, + "learning_rate": 0.00019999733244489728, + "loss": 1.0519, + "step": 1860 + }, + { + "epoch": 0.12913746443688848, + "grad_norm": 5.010197162628174, + "learning_rate": 0.00019999729124793052, + "loss": 1.5016, + "step": 1861 + }, + { + "epoch": 0.12920685587398514, + "grad_norm": 5.06138277053833, + "learning_rate": 0.00019999724973528387, + "loss": 1.2977, + "step": 1862 + }, + { + "epoch": 0.1292762473110818, + "grad_norm": 6.508810520172119, + "learning_rate": 0.00019999720790695756, + "loss": 0.8773, + "step": 1863 + }, + { + "epoch": 0.12934563874817848, + "grad_norm": 8.718884468078613, + "learning_rate": 0.0001999971657629516, + "loss": 1.2516, + "step": 1864 + }, + { + "epoch": 0.12941503018527514, + "grad_norm": 4.09283971786499, + "learning_rate": 0.00019999712330326624, + "loss": 1.1745, + "step": 1865 + }, + { + "epoch": 0.1294844216223718, + "grad_norm": 4.76751708984375, + "learning_rate": 0.00019999708052790153, + "loss": 0.9525, + "step": 1866 + }, + { + "epoch": 0.12955381305946845, + "grad_norm": 3.967249870300293, + "learning_rate": 0.00019999703743685764, + "loss": 1.2839, + "step": 1867 + }, + { + "epoch": 0.12962320449656511, + "grad_norm": 5.201671123504639, + "learning_rate": 0.00019999699403013474, + "loss": 1.2536, + "step": 1868 + }, + { + "epoch": 0.12969259593366178, + "grad_norm": 7.827946662902832, + "learning_rate": 0.00019999695030773292, + "loss": 1.2632, + "step": 1869 + }, + { + "epoch": 0.12976198737075845, + "grad_norm": 5.455713748931885, + "learning_rate": 0.00019999690626965234, + "loss": 1.3348, + "step": 1870 + }, + { + "epoch": 0.1298313788078551, + "grad_norm": 5.483146667480469, + "learning_rate": 0.0001999968619158931, + "loss": 1.4461, + "step": 1871 + }, + { + "epoch": 0.12990077024495178, + "grad_norm": 4.168876647949219, + "learning_rate": 0.00019999681724645538, + "loss": 1.1154, + "step": 1872 + }, + { + "epoch": 0.12997016168204845, + "grad_norm": 5.472257137298584, + "learning_rate": 0.0001999967722613393, + "loss": 1.2329, + "step": 1873 + }, + { + "epoch": 0.13003955311914508, + "grad_norm": 4.98310661315918, + "learning_rate": 0.000199996726960545, + "loss": 1.2138, + "step": 1874 + }, + { + "epoch": 0.13010894455624175, + "grad_norm": 4.77062463760376, + "learning_rate": 0.00019999668134407266, + "loss": 1.2253, + "step": 1875 + }, + { + "epoch": 0.13017833599333842, + "grad_norm": 4.301822662353516, + "learning_rate": 0.0001999966354119224, + "loss": 1.2705, + "step": 1876 + }, + { + "epoch": 0.13024772743043508, + "grad_norm": 5.227789402008057, + "learning_rate": 0.00019999658916409436, + "loss": 1.1772, + "step": 1877 + }, + { + "epoch": 0.13031711886753175, + "grad_norm": 4.395518779754639, + "learning_rate": 0.0001999965426005887, + "loss": 1.208, + "step": 1878 + }, + { + "epoch": 0.13038651030462842, + "grad_norm": 4.986953258514404, + "learning_rate": 0.00019999649572140552, + "loss": 1.2733, + "step": 1879 + }, + { + "epoch": 0.13045590174172508, + "grad_norm": 5.622124195098877, + "learning_rate": 0.00019999644852654503, + "loss": 1.5142, + "step": 1880 + }, + { + "epoch": 0.13052529317882172, + "grad_norm": 4.989194869995117, + "learning_rate": 0.00019999640101600738, + "loss": 1.3925, + "step": 1881 + }, + { + "epoch": 0.1305946846159184, + "grad_norm": 5.895884990692139, + "learning_rate": 0.00019999635318979265, + "loss": 1.41, + "step": 1882 + }, + { + "epoch": 0.13066407605301505, + "grad_norm": 4.63303279876709, + "learning_rate": 0.00019999630504790106, + "loss": 1.4428, + "step": 1883 + }, + { + "epoch": 0.13073346749011172, + "grad_norm": 4.362260341644287, + "learning_rate": 0.0001999962565903327, + "loss": 1.0993, + "step": 1884 + }, + { + "epoch": 0.1308028589272084, + "grad_norm": 4.457571506500244, + "learning_rate": 0.00019999620781708779, + "loss": 1.2907, + "step": 1885 + }, + { + "epoch": 0.13087225036430505, + "grad_norm": 5.502100467681885, + "learning_rate": 0.00019999615872816646, + "loss": 1.4988, + "step": 1886 + }, + { + "epoch": 0.13094164180140172, + "grad_norm": 3.8762197494506836, + "learning_rate": 0.00019999610932356883, + "loss": 1.056, + "step": 1887 + }, + { + "epoch": 0.13101103323849836, + "grad_norm": 4.4870686531066895, + "learning_rate": 0.00019999605960329512, + "loss": 1.0566, + "step": 1888 + }, + { + "epoch": 0.13108042467559503, + "grad_norm": 3.8319332599639893, + "learning_rate": 0.00019999600956734542, + "loss": 0.9488, + "step": 1889 + }, + { + "epoch": 0.1311498161126917, + "grad_norm": 4.602350234985352, + "learning_rate": 0.00019999595921571994, + "loss": 1.1169, + "step": 1890 + }, + { + "epoch": 0.13121920754978836, + "grad_norm": 5.479883193969727, + "learning_rate": 0.00019999590854841883, + "loss": 1.2287, + "step": 1891 + }, + { + "epoch": 0.13128859898688502, + "grad_norm": 4.453867435455322, + "learning_rate": 0.00019999585756544223, + "loss": 0.857, + "step": 1892 + }, + { + "epoch": 0.1313579904239817, + "grad_norm": 5.470523834228516, + "learning_rate": 0.0001999958062667903, + "loss": 1.3557, + "step": 1893 + }, + { + "epoch": 0.13142738186107833, + "grad_norm": 4.5039849281311035, + "learning_rate": 0.0001999957546524632, + "loss": 0.9239, + "step": 1894 + }, + { + "epoch": 0.131496773298175, + "grad_norm": 5.67169713973999, + "learning_rate": 0.00019999570272246113, + "loss": 1.2524, + "step": 1895 + }, + { + "epoch": 0.13156616473527166, + "grad_norm": 3.974195718765259, + "learning_rate": 0.00019999565047678418, + "loss": 1.326, + "step": 1896 + }, + { + "epoch": 0.13163555617236833, + "grad_norm": 3.7098028659820557, + "learning_rate": 0.00019999559791543262, + "loss": 1.2284, + "step": 1897 + }, + { + "epoch": 0.131704947609465, + "grad_norm": 5.281296253204346, + "learning_rate": 0.00019999554503840654, + "loss": 1.2924, + "step": 1898 + }, + { + "epoch": 0.13177433904656166, + "grad_norm": 4.742228031158447, + "learning_rate": 0.00019999549184570613, + "loss": 1.231, + "step": 1899 + }, + { + "epoch": 0.13184373048365833, + "grad_norm": 4.283082962036133, + "learning_rate": 0.00019999543833733153, + "loss": 1.1556, + "step": 1900 + }, + { + "epoch": 0.13191312192075497, + "grad_norm": 3.991654634475708, + "learning_rate": 0.00019999538451328296, + "loss": 1.0676, + "step": 1901 + }, + { + "epoch": 0.13198251335785163, + "grad_norm": 4.966555118560791, + "learning_rate": 0.00019999533037356055, + "loss": 1.4868, + "step": 1902 + }, + { + "epoch": 0.1320519047949483, + "grad_norm": 4.410780906677246, + "learning_rate": 0.00019999527591816447, + "loss": 1.098, + "step": 1903 + }, + { + "epoch": 0.13212129623204497, + "grad_norm": 4.355289936065674, + "learning_rate": 0.00019999522114709492, + "loss": 1.2038, + "step": 1904 + }, + { + "epoch": 0.13219068766914163, + "grad_norm": 3.639549970626831, + "learning_rate": 0.00019999516606035206, + "loss": 1.3195, + "step": 1905 + }, + { + "epoch": 0.1322600791062383, + "grad_norm": 4.416062831878662, + "learning_rate": 0.0001999951106579361, + "loss": 1.2846, + "step": 1906 + }, + { + "epoch": 0.13232947054333496, + "grad_norm": 4.9306721687316895, + "learning_rate": 0.00019999505493984712, + "loss": 1.1774, + "step": 1907 + }, + { + "epoch": 0.1323988619804316, + "grad_norm": 10.828607559204102, + "learning_rate": 0.00019999499890608537, + "loss": 1.3649, + "step": 1908 + }, + { + "epoch": 0.13246825341752827, + "grad_norm": 4.811316013336182, + "learning_rate": 0.000199994942556651, + "loss": 1.1741, + "step": 1909 + }, + { + "epoch": 0.13253764485462494, + "grad_norm": 5.03908634185791, + "learning_rate": 0.00019999488589154423, + "loss": 1.3128, + "step": 1910 + }, + { + "epoch": 0.1326070362917216, + "grad_norm": 9.11916732788086, + "learning_rate": 0.00019999482891076516, + "loss": 1.3419, + "step": 1911 + }, + { + "epoch": 0.13267642772881827, + "grad_norm": 5.292067050933838, + "learning_rate": 0.00019999477161431405, + "loss": 1.2909, + "step": 1912 + }, + { + "epoch": 0.13274581916591494, + "grad_norm": 4.493209362030029, + "learning_rate": 0.00019999471400219105, + "loss": 1.0153, + "step": 1913 + }, + { + "epoch": 0.1328152106030116, + "grad_norm": 5.8357415199279785, + "learning_rate": 0.00019999465607439633, + "loss": 1.4234, + "step": 1914 + }, + { + "epoch": 0.13288460204010824, + "grad_norm": 3.841648578643799, + "learning_rate": 0.00019999459783093007, + "loss": 1.0215, + "step": 1915 + }, + { + "epoch": 0.1329539934772049, + "grad_norm": 5.499293804168701, + "learning_rate": 0.00019999453927179247, + "loss": 1.3852, + "step": 1916 + }, + { + "epoch": 0.13302338491430157, + "grad_norm": 10.884364128112793, + "learning_rate": 0.00019999448039698375, + "loss": 1.2069, + "step": 1917 + }, + { + "epoch": 0.13309277635139824, + "grad_norm": 3.7074708938598633, + "learning_rate": 0.00019999442120650403, + "loss": 1.1165, + "step": 1918 + }, + { + "epoch": 0.1331621677884949, + "grad_norm": 4.100527763366699, + "learning_rate": 0.0001999943617003535, + "loss": 1.3588, + "step": 1919 + }, + { + "epoch": 0.13323155922559157, + "grad_norm": 5.117258071899414, + "learning_rate": 0.00019999430187853242, + "loss": 1.1865, + "step": 1920 + }, + { + "epoch": 0.1333009506626882, + "grad_norm": 4.04844856262207, + "learning_rate": 0.0001999942417410409, + "loss": 1.3806, + "step": 1921 + }, + { + "epoch": 0.13337034209978488, + "grad_norm": 4.7098870277404785, + "learning_rate": 0.00019999418128787915, + "loss": 1.0793, + "step": 1922 + }, + { + "epoch": 0.13343973353688154, + "grad_norm": 4.43313455581665, + "learning_rate": 0.00019999412051904738, + "loss": 1.2253, + "step": 1923 + }, + { + "epoch": 0.1335091249739782, + "grad_norm": 4.425708770751953, + "learning_rate": 0.0001999940594345458, + "loss": 1.1765, + "step": 1924 + }, + { + "epoch": 0.13357851641107488, + "grad_norm": 4.486696720123291, + "learning_rate": 0.00019999399803437453, + "loss": 1.0921, + "step": 1925 + }, + { + "epoch": 0.13364790784817154, + "grad_norm": 4.11106014251709, + "learning_rate": 0.00019999393631853384, + "loss": 1.0564, + "step": 1926 + }, + { + "epoch": 0.1337172992852682, + "grad_norm": 4.696727275848389, + "learning_rate": 0.0001999938742870239, + "loss": 1.0093, + "step": 1927 + }, + { + "epoch": 0.13378669072236485, + "grad_norm": 4.128929615020752, + "learning_rate": 0.00019999381193984487, + "loss": 0.9888, + "step": 1928 + }, + { + "epoch": 0.13385608215946151, + "grad_norm": 3.5292017459869385, + "learning_rate": 0.00019999374927699698, + "loss": 0.9329, + "step": 1929 + }, + { + "epoch": 0.13392547359655818, + "grad_norm": 4.07451057434082, + "learning_rate": 0.00019999368629848046, + "loss": 1.1751, + "step": 1930 + }, + { + "epoch": 0.13399486503365485, + "grad_norm": 4.426505088806152, + "learning_rate": 0.00019999362300429546, + "loss": 1.2561, + "step": 1931 + }, + { + "epoch": 0.1340642564707515, + "grad_norm": 4.370931625366211, + "learning_rate": 0.00019999355939444218, + "loss": 1.3071, + "step": 1932 + }, + { + "epoch": 0.13413364790784818, + "grad_norm": 5.209187030792236, + "learning_rate": 0.00019999349546892084, + "loss": 1.4225, + "step": 1933 + }, + { + "epoch": 0.13420303934494485, + "grad_norm": 4.360305309295654, + "learning_rate": 0.00019999343122773165, + "loss": 1.0951, + "step": 1934 + }, + { + "epoch": 0.13427243078204149, + "grad_norm": 4.096409797668457, + "learning_rate": 0.00019999336667087477, + "loss": 1.1086, + "step": 1935 + }, + { + "epoch": 0.13434182221913815, + "grad_norm": 4.195526123046875, + "learning_rate": 0.00019999330179835045, + "loss": 1.0812, + "step": 1936 + }, + { + "epoch": 0.13441121365623482, + "grad_norm": 3.696530818939209, + "learning_rate": 0.0001999932366101589, + "loss": 1.0093, + "step": 1937 + }, + { + "epoch": 0.13448060509333148, + "grad_norm": 4.710662841796875, + "learning_rate": 0.00019999317110630025, + "loss": 1.1631, + "step": 1938 + }, + { + "epoch": 0.13454999653042815, + "grad_norm": 5.062164306640625, + "learning_rate": 0.00019999310528677483, + "loss": 1.1079, + "step": 1939 + }, + { + "epoch": 0.13461938796752482, + "grad_norm": 3.3778815269470215, + "learning_rate": 0.0001999930391515827, + "loss": 0.7809, + "step": 1940 + }, + { + "epoch": 0.13468877940462146, + "grad_norm": 4.542576789855957, + "learning_rate": 0.00019999297270072419, + "loss": 1.0139, + "step": 1941 + }, + { + "epoch": 0.13475817084171812, + "grad_norm": 4.29608154296875, + "learning_rate": 0.00019999290593419948, + "loss": 1.3159, + "step": 1942 + }, + { + "epoch": 0.1348275622788148, + "grad_norm": 3.884351968765259, + "learning_rate": 0.00019999283885200874, + "loss": 1.1027, + "step": 1943 + }, + { + "epoch": 0.13489695371591146, + "grad_norm": 8.771533012390137, + "learning_rate": 0.0001999927714541522, + "loss": 1.0672, + "step": 1944 + }, + { + "epoch": 0.13496634515300812, + "grad_norm": 4.129704475402832, + "learning_rate": 0.00019999270374063013, + "loss": 1.3632, + "step": 1945 + }, + { + "epoch": 0.1350357365901048, + "grad_norm": 3.7568235397338867, + "learning_rate": 0.00019999263571144265, + "loss": 1.1667, + "step": 1946 + }, + { + "epoch": 0.13510512802720145, + "grad_norm": 9.030211448669434, + "learning_rate": 0.00019999256736659004, + "loss": 1.2754, + "step": 1947 + }, + { + "epoch": 0.1351745194642981, + "grad_norm": 4.192205429077148, + "learning_rate": 0.0001999924987060725, + "loss": 1.0899, + "step": 1948 + }, + { + "epoch": 0.13524391090139476, + "grad_norm": 4.960344314575195, + "learning_rate": 0.0001999924297298902, + "loss": 1.3003, + "step": 1949 + }, + { + "epoch": 0.13531330233849143, + "grad_norm": 18.08932876586914, + "learning_rate": 0.00019999236043804344, + "loss": 1.2482, + "step": 1950 + }, + { + "epoch": 0.1353826937755881, + "grad_norm": 5.243579387664795, + "learning_rate": 0.00019999229083053242, + "loss": 1.2312, + "step": 1951 + }, + { + "epoch": 0.13545208521268476, + "grad_norm": 5.075210094451904, + "learning_rate": 0.00019999222090735729, + "loss": 1.3067, + "step": 1952 + }, + { + "epoch": 0.13552147664978142, + "grad_norm": 5.2650299072265625, + "learning_rate": 0.00019999215066851837, + "loss": 1.2054, + "step": 1953 + }, + { + "epoch": 0.1355908680868781, + "grad_norm": 5.150532245635986, + "learning_rate": 0.0001999920801140158, + "loss": 1.2093, + "step": 1954 + }, + { + "epoch": 0.13566025952397473, + "grad_norm": 4.942793369293213, + "learning_rate": 0.00019999200924384982, + "loss": 1.0455, + "step": 1955 + }, + { + "epoch": 0.1357296509610714, + "grad_norm": 5.787795066833496, + "learning_rate": 0.0001999919380580207, + "loss": 1.1054, + "step": 1956 + }, + { + "epoch": 0.13579904239816806, + "grad_norm": 6.556314468383789, + "learning_rate": 0.0001999918665565286, + "loss": 1.4564, + "step": 1957 + }, + { + "epoch": 0.13586843383526473, + "grad_norm": 5.173015117645264, + "learning_rate": 0.00019999179473937381, + "loss": 1.014, + "step": 1958 + }, + { + "epoch": 0.1359378252723614, + "grad_norm": 4.2094621658325195, + "learning_rate": 0.00019999172260655654, + "loss": 1.2681, + "step": 1959 + }, + { + "epoch": 0.13600721670945806, + "grad_norm": 4.7948713302612305, + "learning_rate": 0.00019999165015807698, + "loss": 1.0925, + "step": 1960 + }, + { + "epoch": 0.13607660814655473, + "grad_norm": 6.236848831176758, + "learning_rate": 0.00019999157739393537, + "loss": 1.0834, + "step": 1961 + }, + { + "epoch": 0.13614599958365137, + "grad_norm": 5.214427471160889, + "learning_rate": 0.00019999150431413196, + "loss": 1.208, + "step": 1962 + }, + { + "epoch": 0.13621539102074803, + "grad_norm": 4.872783184051514, + "learning_rate": 0.00019999143091866696, + "loss": 1.4865, + "step": 1963 + }, + { + "epoch": 0.1362847824578447, + "grad_norm": 5.295464515686035, + "learning_rate": 0.00019999135720754062, + "loss": 1.0897, + "step": 1964 + }, + { + "epoch": 0.13635417389494137, + "grad_norm": 4.560580730438232, + "learning_rate": 0.0001999912831807532, + "loss": 1.1704, + "step": 1965 + }, + { + "epoch": 0.13642356533203803, + "grad_norm": 5.248902320861816, + "learning_rate": 0.00019999120883830486, + "loss": 1.24, + "step": 1966 + }, + { + "epoch": 0.1364929567691347, + "grad_norm": 4.322673797607422, + "learning_rate": 0.00019999113418019587, + "loss": 1.0987, + "step": 1967 + }, + { + "epoch": 0.13656234820623134, + "grad_norm": 5.0458197593688965, + "learning_rate": 0.00019999105920642652, + "loss": 1.443, + "step": 1968 + }, + { + "epoch": 0.136631739643328, + "grad_norm": 6.0327839851379395, + "learning_rate": 0.00019999098391699694, + "loss": 1.1824, + "step": 1969 + }, + { + "epoch": 0.13670113108042467, + "grad_norm": 6.179948806762695, + "learning_rate": 0.00019999090831190745, + "loss": 1.4952, + "step": 1970 + }, + { + "epoch": 0.13677052251752134, + "grad_norm": 5.085090637207031, + "learning_rate": 0.00019999083239115827, + "loss": 1.3364, + "step": 1971 + }, + { + "epoch": 0.136839913954618, + "grad_norm": 5.052151679992676, + "learning_rate": 0.00019999075615474962, + "loss": 1.0473, + "step": 1972 + }, + { + "epoch": 0.13690930539171467, + "grad_norm": 5.239447116851807, + "learning_rate": 0.00019999067960268176, + "loss": 1.3529, + "step": 1973 + }, + { + "epoch": 0.13697869682881134, + "grad_norm": 4.520820140838623, + "learning_rate": 0.00019999060273495492, + "loss": 1.0533, + "step": 1974 + }, + { + "epoch": 0.13704808826590797, + "grad_norm": 4.476208686828613, + "learning_rate": 0.00019999052555156937, + "loss": 1.1203, + "step": 1975 + }, + { + "epoch": 0.13711747970300464, + "grad_norm": 4.409936904907227, + "learning_rate": 0.00019999044805252533, + "loss": 1.32, + "step": 1976 + }, + { + "epoch": 0.1371868711401013, + "grad_norm": 6.3967976570129395, + "learning_rate": 0.000199990370237823, + "loss": 1.0108, + "step": 1977 + }, + { + "epoch": 0.13725626257719797, + "grad_norm": 4.008444309234619, + "learning_rate": 0.00019999029210746272, + "loss": 1.1808, + "step": 1978 + }, + { + "epoch": 0.13732565401429464, + "grad_norm": 5.183061122894287, + "learning_rate": 0.00019999021366144467, + "loss": 1.3023, + "step": 1979 + }, + { + "epoch": 0.1373950454513913, + "grad_norm": 7.1720380783081055, + "learning_rate": 0.00019999013489976913, + "loss": 1.0143, + "step": 1980 + }, + { + "epoch": 0.13746443688848797, + "grad_norm": 4.654735088348389, + "learning_rate": 0.00019999005582243631, + "loss": 1.3957, + "step": 1981 + }, + { + "epoch": 0.1375338283255846, + "grad_norm": 5.047652244567871, + "learning_rate": 0.00019998997642944652, + "loss": 1.1749, + "step": 1982 + }, + { + "epoch": 0.13760321976268128, + "grad_norm": 5.767385482788086, + "learning_rate": 0.00019998989672079996, + "loss": 1.4076, + "step": 1983 + }, + { + "epoch": 0.13767261119977794, + "grad_norm": 4.069961071014404, + "learning_rate": 0.0001999898166964969, + "loss": 1.0195, + "step": 1984 + }, + { + "epoch": 0.1377420026368746, + "grad_norm": 4.434694290161133, + "learning_rate": 0.00019998973635653754, + "loss": 1.1701, + "step": 1985 + }, + { + "epoch": 0.13781139407397128, + "grad_norm": 5.315025329589844, + "learning_rate": 0.00019998965570092224, + "loss": 1.2697, + "step": 1986 + }, + { + "epoch": 0.13788078551106794, + "grad_norm": 4.372592926025391, + "learning_rate": 0.00019998957472965118, + "loss": 0.9788, + "step": 1987 + }, + { + "epoch": 0.1379501769481646, + "grad_norm": 3.8697030544281006, + "learning_rate": 0.00019998949344272464, + "loss": 1.1954, + "step": 1988 + }, + { + "epoch": 0.13801956838526125, + "grad_norm": 4.7656450271606445, + "learning_rate": 0.00019998941184014285, + "loss": 1.3943, + "step": 1989 + }, + { + "epoch": 0.13808895982235792, + "grad_norm": 5.69412899017334, + "learning_rate": 0.00019998932992190615, + "loss": 1.1733, + "step": 1990 + }, + { + "epoch": 0.13815835125945458, + "grad_norm": 5.4713616371154785, + "learning_rate": 0.00019998924768801466, + "loss": 1.3908, + "step": 1991 + }, + { + "epoch": 0.13822774269655125, + "grad_norm": 4.946126937866211, + "learning_rate": 0.00019998916513846874, + "loss": 1.3848, + "step": 1992 + }, + { + "epoch": 0.13829713413364791, + "grad_norm": 4.813767910003662, + "learning_rate": 0.00019998908227326865, + "loss": 1.3714, + "step": 1993 + }, + { + "epoch": 0.13836652557074458, + "grad_norm": 4.054676532745361, + "learning_rate": 0.0001999889990924146, + "loss": 1.0821, + "step": 1994 + }, + { + "epoch": 0.13843591700784122, + "grad_norm": 4.376652717590332, + "learning_rate": 0.00019998891559590688, + "loss": 0.9636, + "step": 1995 + }, + { + "epoch": 0.13850530844493789, + "grad_norm": 5.305632591247559, + "learning_rate": 0.00019998883178374579, + "loss": 1.4341, + "step": 1996 + }, + { + "epoch": 0.13857469988203455, + "grad_norm": 5.503647804260254, + "learning_rate": 0.00019998874765593153, + "loss": 1.3687, + "step": 1997 + }, + { + "epoch": 0.13864409131913122, + "grad_norm": 3.9554293155670166, + "learning_rate": 0.0001999886632124644, + "loss": 1.1288, + "step": 1998 + }, + { + "epoch": 0.13871348275622788, + "grad_norm": 4.774226188659668, + "learning_rate": 0.00019998857845334467, + "loss": 1.2084, + "step": 1999 + }, + { + "epoch": 0.13878287419332455, + "grad_norm": 4.603846073150635, + "learning_rate": 0.0001999884933785726, + "loss": 1.1225, + "step": 2000 + }, + { + "epoch": 0.13885226563042122, + "grad_norm": 4.687699317932129, + "learning_rate": 0.00019998840798814843, + "loss": 1.102, + "step": 2001 + }, + { + "epoch": 0.13892165706751786, + "grad_norm": 3.7618069648742676, + "learning_rate": 0.00019998832228207246, + "loss": 1.3263, + "step": 2002 + }, + { + "epoch": 0.13899104850461452, + "grad_norm": 4.614494323730469, + "learning_rate": 0.00019998823626034498, + "loss": 1.1173, + "step": 2003 + }, + { + "epoch": 0.1390604399417112, + "grad_norm": 4.202536106109619, + "learning_rate": 0.00019998814992296625, + "loss": 1.059, + "step": 2004 + }, + { + "epoch": 0.13912983137880786, + "grad_norm": 4.410794258117676, + "learning_rate": 0.0001999880632699365, + "loss": 1.058, + "step": 2005 + }, + { + "epoch": 0.13919922281590452, + "grad_norm": 4.776449203491211, + "learning_rate": 0.00019998797630125603, + "loss": 1.208, + "step": 2006 + }, + { + "epoch": 0.1392686142530012, + "grad_norm": 3.9849185943603516, + "learning_rate": 0.00019998788901692515, + "loss": 1.0216, + "step": 2007 + }, + { + "epoch": 0.13933800569009785, + "grad_norm": 4.126650333404541, + "learning_rate": 0.0001999878014169441, + "loss": 1.0019, + "step": 2008 + }, + { + "epoch": 0.1394073971271945, + "grad_norm": 14.199785232543945, + "learning_rate": 0.00019998771350131315, + "loss": 1.2817, + "step": 2009 + }, + { + "epoch": 0.13947678856429116, + "grad_norm": 4.144847393035889, + "learning_rate": 0.00019998762527003258, + "loss": 1.2537, + "step": 2010 + }, + { + "epoch": 0.13954618000138783, + "grad_norm": 7.98456335067749, + "learning_rate": 0.0001999875367231027, + "loss": 0.9588, + "step": 2011 + }, + { + "epoch": 0.1396155714384845, + "grad_norm": 5.478836536407471, + "learning_rate": 0.00019998744786052375, + "loss": 1.3108, + "step": 2012 + }, + { + "epoch": 0.13968496287558116, + "grad_norm": 8.011372566223145, + "learning_rate": 0.00019998735868229604, + "loss": 1.2554, + "step": 2013 + }, + { + "epoch": 0.13975435431267783, + "grad_norm": 5.310598373413086, + "learning_rate": 0.00019998726918841984, + "loss": 1.1861, + "step": 2014 + }, + { + "epoch": 0.13982374574977446, + "grad_norm": 4.366217613220215, + "learning_rate": 0.0001999871793788954, + "loss": 1.243, + "step": 2015 + }, + { + "epoch": 0.13989313718687113, + "grad_norm": 4.401420593261719, + "learning_rate": 0.00019998708925372309, + "loss": 1.2206, + "step": 2016 + }, + { + "epoch": 0.1399625286239678, + "grad_norm": 5.623363494873047, + "learning_rate": 0.0001999869988129031, + "loss": 1.3028, + "step": 2017 + }, + { + "epoch": 0.14003192006106446, + "grad_norm": 4.767771244049072, + "learning_rate": 0.0001999869080564358, + "loss": 1.2452, + "step": 2018 + }, + { + "epoch": 0.14010131149816113, + "grad_norm": 6.032095909118652, + "learning_rate": 0.00019998681698432135, + "loss": 1.3216, + "step": 2019 + }, + { + "epoch": 0.1401707029352578, + "grad_norm": 4.281265735626221, + "learning_rate": 0.00019998672559656016, + "loss": 1.1041, + "step": 2020 + }, + { + "epoch": 0.14024009437235446, + "grad_norm": 4.236314296722412, + "learning_rate": 0.00019998663389315252, + "loss": 1.0153, + "step": 2021 + }, + { + "epoch": 0.1403094858094511, + "grad_norm": 5.267057418823242, + "learning_rate": 0.00019998654187409863, + "loss": 1.2702, + "step": 2022 + }, + { + "epoch": 0.14037887724654777, + "grad_norm": 4.377004146575928, + "learning_rate": 0.00019998644953939885, + "loss": 1.0177, + "step": 2023 + }, + { + "epoch": 0.14044826868364443, + "grad_norm": 4.448112964630127, + "learning_rate": 0.00019998635688905342, + "loss": 0.9372, + "step": 2024 + }, + { + "epoch": 0.1405176601207411, + "grad_norm": 4.735557556152344, + "learning_rate": 0.0001999862639230627, + "loss": 1.151, + "step": 2025 + }, + { + "epoch": 0.14058705155783777, + "grad_norm": 7.0353569984436035, + "learning_rate": 0.00019998617064142694, + "loss": 0.9981, + "step": 2026 + }, + { + "epoch": 0.14065644299493443, + "grad_norm": 4.023563385009766, + "learning_rate": 0.0001999860770441464, + "loss": 1.1022, + "step": 2027 + }, + { + "epoch": 0.1407258344320311, + "grad_norm": 4.724294185638428, + "learning_rate": 0.00019998598313122144, + "loss": 1.1719, + "step": 2028 + }, + { + "epoch": 0.14079522586912774, + "grad_norm": 8.864587783813477, + "learning_rate": 0.00019998588890265235, + "loss": 1.4579, + "step": 2029 + }, + { + "epoch": 0.1408646173062244, + "grad_norm": 7.824373245239258, + "learning_rate": 0.0001999857943584394, + "loss": 1.1587, + "step": 2030 + }, + { + "epoch": 0.14093400874332107, + "grad_norm": 7.095874309539795, + "learning_rate": 0.00019998569949858288, + "loss": 1.2455, + "step": 2031 + }, + { + "epoch": 0.14100340018041774, + "grad_norm": 4.882680892944336, + "learning_rate": 0.00019998560432308313, + "loss": 1.193, + "step": 2032 + }, + { + "epoch": 0.1410727916175144, + "grad_norm": 5.407138824462891, + "learning_rate": 0.0001999855088319404, + "loss": 1.1809, + "step": 2033 + }, + { + "epoch": 0.14114218305461107, + "grad_norm": 4.446072101593018, + "learning_rate": 0.00019998541302515506, + "loss": 1.1959, + "step": 2034 + }, + { + "epoch": 0.14121157449170774, + "grad_norm": 4.874587535858154, + "learning_rate": 0.00019998531690272736, + "loss": 1.7081, + "step": 2035 + }, + { + "epoch": 0.14128096592880438, + "grad_norm": 4.881026268005371, + "learning_rate": 0.0001999852204646576, + "loss": 1.4781, + "step": 2036 + }, + { + "epoch": 0.14135035736590104, + "grad_norm": 4.750429630279541, + "learning_rate": 0.00019998512371094607, + "loss": 1.3268, + "step": 2037 + }, + { + "epoch": 0.1414197488029977, + "grad_norm": 4.464096546173096, + "learning_rate": 0.00019998502664159316, + "loss": 1.3032, + "step": 2038 + }, + { + "epoch": 0.14148914024009437, + "grad_norm": 3.7499468326568604, + "learning_rate": 0.00019998492925659908, + "loss": 0.9169, + "step": 2039 + }, + { + "epoch": 0.14155853167719104, + "grad_norm": 6.248579502105713, + "learning_rate": 0.00019998483155596423, + "loss": 1.1691, + "step": 2040 + }, + { + "epoch": 0.1416279231142877, + "grad_norm": 3.852663516998291, + "learning_rate": 0.00019998473353968882, + "loss": 0.8711, + "step": 2041 + }, + { + "epoch": 0.14169731455138435, + "grad_norm": 5.160830497741699, + "learning_rate": 0.0001999846352077732, + "loss": 1.4839, + "step": 2042 + }, + { + "epoch": 0.141766705988481, + "grad_norm": 5.2086381912231445, + "learning_rate": 0.00019998453656021773, + "loss": 1.1052, + "step": 2043 + }, + { + "epoch": 0.14183609742557768, + "grad_norm": 5.797573089599609, + "learning_rate": 0.00019998443759702264, + "loss": 1.3476, + "step": 2044 + }, + { + "epoch": 0.14190548886267434, + "grad_norm": 4.784862041473389, + "learning_rate": 0.0001999843383181883, + "loss": 0.9774, + "step": 2045 + }, + { + "epoch": 0.141974880299771, + "grad_norm": 5.507254600524902, + "learning_rate": 0.00019998423872371498, + "loss": 1.4514, + "step": 2046 + }, + { + "epoch": 0.14204427173686768, + "grad_norm": 4.741978168487549, + "learning_rate": 0.00019998413881360306, + "loss": 1.1995, + "step": 2047 + }, + { + "epoch": 0.14211366317396434, + "grad_norm": 4.439135551452637, + "learning_rate": 0.00019998403858785278, + "loss": 0.8457, + "step": 2048 + }, + { + "epoch": 0.14218305461106098, + "grad_norm": 4.519104957580566, + "learning_rate": 0.00019998393804646452, + "loss": 1.3072, + "step": 2049 + }, + { + "epoch": 0.14225244604815765, + "grad_norm": 4.452001094818115, + "learning_rate": 0.00019998383718943853, + "loss": 1.3468, + "step": 2050 + }, + { + "epoch": 0.14232183748525432, + "grad_norm": 3.9186058044433594, + "learning_rate": 0.0001999837360167752, + "loss": 0.9535, + "step": 2051 + }, + { + "epoch": 0.14239122892235098, + "grad_norm": 4.174495697021484, + "learning_rate": 0.0001999836345284748, + "loss": 0.9979, + "step": 2052 + }, + { + "epoch": 0.14246062035944765, + "grad_norm": 4.5036163330078125, + "learning_rate": 0.00019998353272453767, + "loss": 1.1757, + "step": 2053 + }, + { + "epoch": 0.14253001179654431, + "grad_norm": 5.006237983703613, + "learning_rate": 0.0001999834306049641, + "loss": 1.1073, + "step": 2054 + }, + { + "epoch": 0.14259940323364098, + "grad_norm": 3.4418814182281494, + "learning_rate": 0.0001999833281697545, + "loss": 0.9523, + "step": 2055 + }, + { + "epoch": 0.14266879467073762, + "grad_norm": 4.287426471710205, + "learning_rate": 0.00019998322541890908, + "loss": 1.1347, + "step": 2056 + }, + { + "epoch": 0.1427381861078343, + "grad_norm": 6.070193290710449, + "learning_rate": 0.00019998312235242828, + "loss": 1.061, + "step": 2057 + }, + { + "epoch": 0.14280757754493095, + "grad_norm": 6.41237211227417, + "learning_rate": 0.00019998301897031229, + "loss": 1.2059, + "step": 2058 + }, + { + "epoch": 0.14287696898202762, + "grad_norm": 5.91712760925293, + "learning_rate": 0.00019998291527256154, + "loss": 1.4013, + "step": 2059 + }, + { + "epoch": 0.14294636041912429, + "grad_norm": 4.797826766967773, + "learning_rate": 0.00019998281125917632, + "loss": 1.4295, + "step": 2060 + }, + { + "epoch": 0.14301575185622095, + "grad_norm": 4.666703224182129, + "learning_rate": 0.00019998270693015696, + "loss": 1.1775, + "step": 2061 + }, + { + "epoch": 0.14308514329331762, + "grad_norm": 5.2972822189331055, + "learning_rate": 0.00019998260228550383, + "loss": 1.4137, + "step": 2062 + }, + { + "epoch": 0.14315453473041426, + "grad_norm": 4.495140075683594, + "learning_rate": 0.00019998249732521722, + "loss": 1.3293, + "step": 2063 + }, + { + "epoch": 0.14322392616751092, + "grad_norm": 4.680865287780762, + "learning_rate": 0.00019998239204929743, + "loss": 1.4473, + "step": 2064 + }, + { + "epoch": 0.1432933176046076, + "grad_norm": 3.387483835220337, + "learning_rate": 0.00019998228645774484, + "loss": 0.9313, + "step": 2065 + }, + { + "epoch": 0.14336270904170426, + "grad_norm": 7.470517635345459, + "learning_rate": 0.00019998218055055977, + "loss": 1.6598, + "step": 2066 + }, + { + "epoch": 0.14343210047880092, + "grad_norm": 4.224184513092041, + "learning_rate": 0.00019998207432774255, + "loss": 1.0154, + "step": 2067 + }, + { + "epoch": 0.1435014919158976, + "grad_norm": 4.439693927764893, + "learning_rate": 0.00019998196778929356, + "loss": 1.2148, + "step": 2068 + }, + { + "epoch": 0.14357088335299423, + "grad_norm": 5.3844146728515625, + "learning_rate": 0.00019998186093521308, + "loss": 1.2814, + "step": 2069 + }, + { + "epoch": 0.1436402747900909, + "grad_norm": 6.517559051513672, + "learning_rate": 0.00019998175376550147, + "loss": 1.202, + "step": 2070 + }, + { + "epoch": 0.14370966622718756, + "grad_norm": 4.098563194274902, + "learning_rate": 0.00019998164628015904, + "loss": 1.2303, + "step": 2071 + }, + { + "epoch": 0.14377905766428423, + "grad_norm": 5.174404144287109, + "learning_rate": 0.00019998153847918614, + "loss": 1.2165, + "step": 2072 + }, + { + "epoch": 0.1438484491013809, + "grad_norm": 3.9744606018066406, + "learning_rate": 0.00019998143036258314, + "loss": 1.0769, + "step": 2073 + }, + { + "epoch": 0.14391784053847756, + "grad_norm": 5.166172027587891, + "learning_rate": 0.0001999813219303504, + "loss": 1.4892, + "step": 2074 + }, + { + "epoch": 0.14398723197557423, + "grad_norm": 5.773413181304932, + "learning_rate": 0.00019998121318248816, + "loss": 1.3541, + "step": 2075 + }, + { + "epoch": 0.14405662341267086, + "grad_norm": 3.9939048290252686, + "learning_rate": 0.00019998110411899688, + "loss": 1.3307, + "step": 2076 + }, + { + "epoch": 0.14412601484976753, + "grad_norm": 4.780254364013672, + "learning_rate": 0.00019998099473987683, + "loss": 1.5693, + "step": 2077 + }, + { + "epoch": 0.1441954062868642, + "grad_norm": 6.922140121459961, + "learning_rate": 0.00019998088504512843, + "loss": 1.5304, + "step": 2078 + }, + { + "epoch": 0.14426479772396086, + "grad_norm": 5.252416610717773, + "learning_rate": 0.00019998077503475192, + "loss": 1.0963, + "step": 2079 + }, + { + "epoch": 0.14433418916105753, + "grad_norm": 5.068620681762695, + "learning_rate": 0.00019998066470874773, + "loss": 1.2048, + "step": 2080 + }, + { + "epoch": 0.1444035805981542, + "grad_norm": 5.34529972076416, + "learning_rate": 0.00019998055406711615, + "loss": 1.3089, + "step": 2081 + }, + { + "epoch": 0.14447297203525086, + "grad_norm": 4.641902923583984, + "learning_rate": 0.00019998044310985758, + "loss": 1.1532, + "step": 2082 + }, + { + "epoch": 0.1445423634723475, + "grad_norm": 5.060982704162598, + "learning_rate": 0.00019998033183697236, + "loss": 1.2315, + "step": 2083 + }, + { + "epoch": 0.14461175490944417, + "grad_norm": 5.224031925201416, + "learning_rate": 0.00019998022024846085, + "loss": 1.2833, + "step": 2084 + }, + { + "epoch": 0.14468114634654083, + "grad_norm": 5.54855489730835, + "learning_rate": 0.00019998010834432336, + "loss": 1.365, + "step": 2085 + }, + { + "epoch": 0.1447505377836375, + "grad_norm": 5.000036239624023, + "learning_rate": 0.00019997999612456026, + "loss": 1.0666, + "step": 2086 + }, + { + "epoch": 0.14481992922073417, + "grad_norm": 4.93722677230835, + "learning_rate": 0.0001999798835891719, + "loss": 1.3678, + "step": 2087 + }, + { + "epoch": 0.14488932065783083, + "grad_norm": 5.2589216232299805, + "learning_rate": 0.00019997977073815868, + "loss": 1.2132, + "step": 2088 + }, + { + "epoch": 0.14495871209492747, + "grad_norm": 4.081704139709473, + "learning_rate": 0.0001999796575715209, + "loss": 0.9131, + "step": 2089 + }, + { + "epoch": 0.14502810353202414, + "grad_norm": 4.456450462341309, + "learning_rate": 0.00019997954408925898, + "loss": 1.2307, + "step": 2090 + }, + { + "epoch": 0.1450974949691208, + "grad_norm": 4.714766025543213, + "learning_rate": 0.0001999794302913732, + "loss": 1.3671, + "step": 2091 + }, + { + "epoch": 0.14516688640621747, + "grad_norm": 4.41970157623291, + "learning_rate": 0.00019997931617786395, + "loss": 1.4229, + "step": 2092 + }, + { + "epoch": 0.14523627784331414, + "grad_norm": 3.988266944885254, + "learning_rate": 0.00019997920174873163, + "loss": 1.4623, + "step": 2093 + }, + { + "epoch": 0.1453056692804108, + "grad_norm": 4.691127777099609, + "learning_rate": 0.00019997908700397656, + "loss": 1.0341, + "step": 2094 + }, + { + "epoch": 0.14537506071750747, + "grad_norm": 4.468564033508301, + "learning_rate": 0.0001999789719435991, + "loss": 1.4893, + "step": 2095 + }, + { + "epoch": 0.1454444521546041, + "grad_norm": 6.692159175872803, + "learning_rate": 0.00019997885656759966, + "loss": 1.0883, + "step": 2096 + }, + { + "epoch": 0.14551384359170078, + "grad_norm": 5.415416240692139, + "learning_rate": 0.00019997874087597853, + "loss": 1.2111, + "step": 2097 + }, + { + "epoch": 0.14558323502879744, + "grad_norm": 6.090401649475098, + "learning_rate": 0.00019997862486873614, + "loss": 1.6776, + "step": 2098 + }, + { + "epoch": 0.1456526264658941, + "grad_norm": 6.2124409675598145, + "learning_rate": 0.0001999785085458728, + "loss": 1.4847, + "step": 2099 + }, + { + "epoch": 0.14572201790299077, + "grad_norm": 4.129336357116699, + "learning_rate": 0.00019997839190738894, + "loss": 1.2193, + "step": 2100 + }, + { + "epoch": 0.14579140934008744, + "grad_norm": 4.834132194519043, + "learning_rate": 0.00019997827495328487, + "loss": 1.1802, + "step": 2101 + }, + { + "epoch": 0.1458608007771841, + "grad_norm": 5.9290595054626465, + "learning_rate": 0.000199978157683561, + "loss": 1.4626, + "step": 2102 + }, + { + "epoch": 0.14593019221428075, + "grad_norm": 4.617031574249268, + "learning_rate": 0.00019997804009821768, + "loss": 1.2436, + "step": 2103 + }, + { + "epoch": 0.1459995836513774, + "grad_norm": 3.959986925125122, + "learning_rate": 0.0001999779221972553, + "loss": 1.1897, + "step": 2104 + }, + { + "epoch": 0.14606897508847408, + "grad_norm": 3.8701541423797607, + "learning_rate": 0.00019997780398067422, + "loss": 1.1706, + "step": 2105 + }, + { + "epoch": 0.14613836652557075, + "grad_norm": 10.017986297607422, + "learning_rate": 0.00019997768544847478, + "loss": 1.4005, + "step": 2106 + }, + { + "epoch": 0.1462077579626674, + "grad_norm": 8.590078353881836, + "learning_rate": 0.00019997756660065742, + "loss": 0.9943, + "step": 2107 + }, + { + "epoch": 0.14627714939976408, + "grad_norm": 11.519901275634766, + "learning_rate": 0.0001999774474372225, + "loss": 1.376, + "step": 2108 + }, + { + "epoch": 0.14634654083686074, + "grad_norm": 13.607233047485352, + "learning_rate": 0.00019997732795817036, + "loss": 1.9405, + "step": 2109 + }, + { + "epoch": 0.14641593227395738, + "grad_norm": 37.16494369506836, + "learning_rate": 0.00019997720816350136, + "loss": 1.7205, + "step": 2110 + }, + { + "epoch": 0.14648532371105405, + "grad_norm": 5.530763626098633, + "learning_rate": 0.00019997708805321592, + "loss": 1.1279, + "step": 2111 + }, + { + "epoch": 0.14655471514815072, + "grad_norm": 7.647805213928223, + "learning_rate": 0.00019997696762731444, + "loss": 1.5807, + "step": 2112 + }, + { + "epoch": 0.14662410658524738, + "grad_norm": 8.37769603729248, + "learning_rate": 0.0001999768468857973, + "loss": 1.2906, + "step": 2113 + }, + { + "epoch": 0.14669349802234405, + "grad_norm": 4.446035385131836, + "learning_rate": 0.00019997672582866482, + "loss": 1.1353, + "step": 2114 + }, + { + "epoch": 0.14676288945944072, + "grad_norm": 5.509897708892822, + "learning_rate": 0.0001999766044559174, + "loss": 1.3114, + "step": 2115 + }, + { + "epoch": 0.14683228089653735, + "grad_norm": 5.12455940246582, + "learning_rate": 0.00019997648276755544, + "loss": 1.2477, + "step": 2116 + }, + { + "epoch": 0.14690167233363402, + "grad_norm": 4.4513936042785645, + "learning_rate": 0.00019997636076357934, + "loss": 1.2955, + "step": 2117 + }, + { + "epoch": 0.1469710637707307, + "grad_norm": 8.372267723083496, + "learning_rate": 0.00019997623844398947, + "loss": 1.2718, + "step": 2118 + }, + { + "epoch": 0.14704045520782735, + "grad_norm": 4.310723781585693, + "learning_rate": 0.00019997611580878618, + "loss": 1.3318, + "step": 2119 + }, + { + "epoch": 0.14710984664492402, + "grad_norm": 5.6929168701171875, + "learning_rate": 0.00019997599285796996, + "loss": 1.4016, + "step": 2120 + }, + { + "epoch": 0.14717923808202069, + "grad_norm": 4.552768707275391, + "learning_rate": 0.00019997586959154108, + "loss": 1.1891, + "step": 2121 + }, + { + "epoch": 0.14724862951911735, + "grad_norm": 5.201392650604248, + "learning_rate": 0.00019997574600949997, + "loss": 1.4373, + "step": 2122 + }, + { + "epoch": 0.147318020956214, + "grad_norm": 5.42899751663208, + "learning_rate": 0.00019997562211184706, + "loss": 1.2892, + "step": 2123 + }, + { + "epoch": 0.14738741239331066, + "grad_norm": 5.177676200866699, + "learning_rate": 0.0001999754978985827, + "loss": 1.615, + "step": 2124 + }, + { + "epoch": 0.14745680383040732, + "grad_norm": 4.819845199584961, + "learning_rate": 0.0001999753733697073, + "loss": 1.0721, + "step": 2125 + }, + { + "epoch": 0.147526195267504, + "grad_norm": 4.570306777954102, + "learning_rate": 0.00019997524852522123, + "loss": 1.2147, + "step": 2126 + }, + { + "epoch": 0.14759558670460066, + "grad_norm": 8.639703750610352, + "learning_rate": 0.00019997512336512491, + "loss": 1.1921, + "step": 2127 + }, + { + "epoch": 0.14766497814169732, + "grad_norm": 6.451936721801758, + "learning_rate": 0.0001999749978894187, + "loss": 1.2423, + "step": 2128 + }, + { + "epoch": 0.147734369578794, + "grad_norm": 4.86361837387085, + "learning_rate": 0.00019997487209810308, + "loss": 1.2612, + "step": 2129 + }, + { + "epoch": 0.14780376101589063, + "grad_norm": 6.192412376403809, + "learning_rate": 0.00019997474599117837, + "loss": 1.0934, + "step": 2130 + }, + { + "epoch": 0.1478731524529873, + "grad_norm": 5.025325775146484, + "learning_rate": 0.00019997461956864493, + "loss": 1.131, + "step": 2131 + }, + { + "epoch": 0.14794254389008396, + "grad_norm": 5.357651710510254, + "learning_rate": 0.00019997449283050327, + "loss": 1.1634, + "step": 2132 + }, + { + "epoch": 0.14801193532718063, + "grad_norm": 4.562382221221924, + "learning_rate": 0.0001999743657767537, + "loss": 1.4139, + "step": 2133 + }, + { + "epoch": 0.1480813267642773, + "grad_norm": 5.786166191101074, + "learning_rate": 0.0001999742384073967, + "loss": 1.3508, + "step": 2134 + }, + { + "epoch": 0.14815071820137396, + "grad_norm": 6.1476240158081055, + "learning_rate": 0.00019997411072243259, + "loss": 1.0823, + "step": 2135 + }, + { + "epoch": 0.14822010963847063, + "grad_norm": 6.82174015045166, + "learning_rate": 0.0001999739827218618, + "loss": 1.1638, + "step": 2136 + }, + { + "epoch": 0.14828950107556726, + "grad_norm": 5.207771301269531, + "learning_rate": 0.00019997385440568478, + "loss": 1.0778, + "step": 2137 + }, + { + "epoch": 0.14835889251266393, + "grad_norm": 7.018359661102295, + "learning_rate": 0.00019997372577390193, + "loss": 1.2433, + "step": 2138 + }, + { + "epoch": 0.1484282839497606, + "grad_norm": 4.820852279663086, + "learning_rate": 0.00019997359682651355, + "loss": 1.3745, + "step": 2139 + }, + { + "epoch": 0.14849767538685726, + "grad_norm": 4.013251304626465, + "learning_rate": 0.00019997346756352017, + "loss": 1.0169, + "step": 2140 + }, + { + "epoch": 0.14856706682395393, + "grad_norm": 12.593973159790039, + "learning_rate": 0.00019997333798492216, + "loss": 1.3876, + "step": 2141 + }, + { + "epoch": 0.1486364582610506, + "grad_norm": 5.984450340270996, + "learning_rate": 0.00019997320809071989, + "loss": 1.1805, + "step": 2142 + }, + { + "epoch": 0.14870584969814724, + "grad_norm": 6.2396559715271, + "learning_rate": 0.0001999730778809138, + "loss": 1.1303, + "step": 2143 + }, + { + "epoch": 0.1487752411352439, + "grad_norm": 5.5874342918396, + "learning_rate": 0.00019997294735550435, + "loss": 1.3572, + "step": 2144 + }, + { + "epoch": 0.14884463257234057, + "grad_norm": 5.434574604034424, + "learning_rate": 0.00019997281651449185, + "loss": 1.2548, + "step": 2145 + }, + { + "epoch": 0.14891402400943723, + "grad_norm": 4.007925510406494, + "learning_rate": 0.00019997268535787682, + "loss": 1.2683, + "step": 2146 + }, + { + "epoch": 0.1489834154465339, + "grad_norm": 5.10413122177124, + "learning_rate": 0.00019997255388565956, + "loss": 1.3171, + "step": 2147 + }, + { + "epoch": 0.14905280688363057, + "grad_norm": 5.0256500244140625, + "learning_rate": 0.0001999724220978406, + "loss": 1.7846, + "step": 2148 + }, + { + "epoch": 0.14912219832072723, + "grad_norm": 4.7062883377075195, + "learning_rate": 0.00019997228999442027, + "loss": 1.1347, + "step": 2149 + }, + { + "epoch": 0.14919158975782387, + "grad_norm": 6.932957649230957, + "learning_rate": 0.00019997215757539902, + "loss": 1.4677, + "step": 2150 + }, + { + "epoch": 0.14926098119492054, + "grad_norm": 7.15814733505249, + "learning_rate": 0.00019997202484077728, + "loss": 1.504, + "step": 2151 + }, + { + "epoch": 0.1493303726320172, + "grad_norm": 4.8692450523376465, + "learning_rate": 0.00019997189179055545, + "loss": 1.1031, + "step": 2152 + }, + { + "epoch": 0.14939976406911387, + "grad_norm": 3.76432204246521, + "learning_rate": 0.00019997175842473397, + "loss": 1.3226, + "step": 2153 + }, + { + "epoch": 0.14946915550621054, + "grad_norm": 4.282259464263916, + "learning_rate": 0.00019997162474331325, + "loss": 1.2195, + "step": 2154 + }, + { + "epoch": 0.1495385469433072, + "grad_norm": 3.777074098587036, + "learning_rate": 0.00019997149074629369, + "loss": 1.2964, + "step": 2155 + }, + { + "epoch": 0.14960793838040387, + "grad_norm": 5.337639808654785, + "learning_rate": 0.00019997135643367572, + "loss": 1.0319, + "step": 2156 + }, + { + "epoch": 0.1496773298175005, + "grad_norm": 4.662404537200928, + "learning_rate": 0.0001999712218054598, + "loss": 1.0687, + "step": 2157 + }, + { + "epoch": 0.14974672125459718, + "grad_norm": 4.13316535949707, + "learning_rate": 0.0001999710868616463, + "loss": 1.0208, + "step": 2158 + }, + { + "epoch": 0.14981611269169384, + "grad_norm": 5.164332866668701, + "learning_rate": 0.00019997095160223573, + "loss": 1.3518, + "step": 2159 + }, + { + "epoch": 0.1498855041287905, + "grad_norm": 4.562839031219482, + "learning_rate": 0.00019997081602722845, + "loss": 1.1194, + "step": 2160 + }, + { + "epoch": 0.14995489556588718, + "grad_norm": 4.032911777496338, + "learning_rate": 0.0001999706801366249, + "loss": 1.1072, + "step": 2161 + }, + { + "epoch": 0.15002428700298384, + "grad_norm": 13.193571090698242, + "learning_rate": 0.0001999705439304255, + "loss": 1.1215, + "step": 2162 + }, + { + "epoch": 0.15009367844008048, + "grad_norm": 8.293042182922363, + "learning_rate": 0.0001999704074086307, + "loss": 1.2684, + "step": 2163 + }, + { + "epoch": 0.15016306987717715, + "grad_norm": 7.014968395233154, + "learning_rate": 0.00019997027057124092, + "loss": 1.0803, + "step": 2164 + }, + { + "epoch": 0.1502324613142738, + "grad_norm": 6.850681304931641, + "learning_rate": 0.00019997013341825656, + "loss": 1.5477, + "step": 2165 + }, + { + "epoch": 0.15030185275137048, + "grad_norm": 6.0852274894714355, + "learning_rate": 0.00019996999594967813, + "loss": 1.278, + "step": 2166 + }, + { + "epoch": 0.15037124418846715, + "grad_norm": 4.370278358459473, + "learning_rate": 0.000199969858165506, + "loss": 1.1836, + "step": 2167 + }, + { + "epoch": 0.1504406356255638, + "grad_norm": 5.674388885498047, + "learning_rate": 0.00019996972006574066, + "loss": 1.1631, + "step": 2168 + }, + { + "epoch": 0.15051002706266048, + "grad_norm": 8.119260787963867, + "learning_rate": 0.0001999695816503825, + "loss": 1.1579, + "step": 2169 + }, + { + "epoch": 0.15057941849975712, + "grad_norm": 5.199925899505615, + "learning_rate": 0.00019996944291943193, + "loss": 1.1631, + "step": 2170 + }, + { + "epoch": 0.15064880993685378, + "grad_norm": 4.369781970977783, + "learning_rate": 0.00019996930387288945, + "loss": 1.0246, + "step": 2171 + }, + { + "epoch": 0.15071820137395045, + "grad_norm": 4.461900234222412, + "learning_rate": 0.00019996916451075548, + "loss": 1.2785, + "step": 2172 + }, + { + "epoch": 0.15078759281104712, + "grad_norm": 5.5817551612854, + "learning_rate": 0.00019996902483303047, + "loss": 1.1162, + "step": 2173 + }, + { + "epoch": 0.15085698424814378, + "grad_norm": 8.634241104125977, + "learning_rate": 0.00019996888483971487, + "loss": 1.1691, + "step": 2174 + }, + { + "epoch": 0.15092637568524045, + "grad_norm": 3.9956836700439453, + "learning_rate": 0.00019996874453080901, + "loss": 1.2626, + "step": 2175 + }, + { + "epoch": 0.15099576712233712, + "grad_norm": 4.387819766998291, + "learning_rate": 0.00019996860390631351, + "loss": 1.4237, + "step": 2176 + }, + { + "epoch": 0.15106515855943375, + "grad_norm": 5.5990190505981445, + "learning_rate": 0.0001999684629662287, + "loss": 1.4476, + "step": 2177 + }, + { + "epoch": 0.15113454999653042, + "grad_norm": 5.941329002380371, + "learning_rate": 0.00019996832171055505, + "loss": 1.0848, + "step": 2178 + }, + { + "epoch": 0.1512039414336271, + "grad_norm": 4.076796054840088, + "learning_rate": 0.00019996818013929296, + "loss": 1.1978, + "step": 2179 + }, + { + "epoch": 0.15127333287072375, + "grad_norm": 5.016147136688232, + "learning_rate": 0.00019996803825244298, + "loss": 1.1782, + "step": 2180 + }, + { + "epoch": 0.15134272430782042, + "grad_norm": 5.641890525817871, + "learning_rate": 0.0001999678960500055, + "loss": 1.3386, + "step": 2181 + }, + { + "epoch": 0.1514121157449171, + "grad_norm": 5.09291410446167, + "learning_rate": 0.00019996775353198096, + "loss": 1.091, + "step": 2182 + }, + { + "epoch": 0.15148150718201375, + "grad_norm": 4.458150863647461, + "learning_rate": 0.0001999676106983698, + "loss": 1.3111, + "step": 2183 + }, + { + "epoch": 0.1515508986191104, + "grad_norm": 5.51511287689209, + "learning_rate": 0.0001999674675491725, + "loss": 1.0526, + "step": 2184 + }, + { + "epoch": 0.15162029005620706, + "grad_norm": 7.593070030212402, + "learning_rate": 0.0001999673240843895, + "loss": 1.5218, + "step": 2185 + }, + { + "epoch": 0.15168968149330372, + "grad_norm": 4.614365100860596, + "learning_rate": 0.00019996718030402128, + "loss": 1.3056, + "step": 2186 + }, + { + "epoch": 0.1517590729304004, + "grad_norm": 4.716847896575928, + "learning_rate": 0.0001999670362080683, + "loss": 1.2708, + "step": 2187 + }, + { + "epoch": 0.15182846436749706, + "grad_norm": 5.294297695159912, + "learning_rate": 0.00019996689179653093, + "loss": 1.195, + "step": 2188 + }, + { + "epoch": 0.15189785580459372, + "grad_norm": 3.5130512714385986, + "learning_rate": 0.0001999667470694097, + "loss": 1.0455, + "step": 2189 + }, + { + "epoch": 0.15196724724169036, + "grad_norm": 5.112082004547119, + "learning_rate": 0.00019996660202670503, + "loss": 1.1558, + "step": 2190 + }, + { + "epoch": 0.15203663867878703, + "grad_norm": 4.886985778808594, + "learning_rate": 0.0001999664566684174, + "loss": 1.321, + "step": 2191 + }, + { + "epoch": 0.1521060301158837, + "grad_norm": 3.9051976203918457, + "learning_rate": 0.00019996631099454727, + "loss": 1.0587, + "step": 2192 + }, + { + "epoch": 0.15217542155298036, + "grad_norm": 6.886911869049072, + "learning_rate": 0.0001999661650050951, + "loss": 0.8815, + "step": 2193 + }, + { + "epoch": 0.15224481299007703, + "grad_norm": 5.056756496429443, + "learning_rate": 0.00019996601870006135, + "loss": 1.6487, + "step": 2194 + }, + { + "epoch": 0.1523142044271737, + "grad_norm": 4.710788249969482, + "learning_rate": 0.00019996587207944647, + "loss": 1.2574, + "step": 2195 + }, + { + "epoch": 0.15238359586427036, + "grad_norm": 6.979928493499756, + "learning_rate": 0.00019996572514325092, + "loss": 1.0847, + "step": 2196 + }, + { + "epoch": 0.152452987301367, + "grad_norm": 3.942593574523926, + "learning_rate": 0.00019996557789147518, + "loss": 1.0729, + "step": 2197 + }, + { + "epoch": 0.15252237873846367, + "grad_norm": 4.878655433654785, + "learning_rate": 0.0001999654303241197, + "loss": 1.2716, + "step": 2198 + }, + { + "epoch": 0.15259177017556033, + "grad_norm": 5.240975856781006, + "learning_rate": 0.00019996528244118497, + "loss": 1.2182, + "step": 2199 + }, + { + "epoch": 0.152661161612657, + "grad_norm": 6.196033954620361, + "learning_rate": 0.00019996513424267148, + "loss": 1.1804, + "step": 2200 + }, + { + "epoch": 0.15273055304975366, + "grad_norm": 6.951417922973633, + "learning_rate": 0.0001999649857285796, + "loss": 1.1348, + "step": 2201 + }, + { + "epoch": 0.15279994448685033, + "grad_norm": 4.822819709777832, + "learning_rate": 0.00019996483689890986, + "loss": 1.2566, + "step": 2202 + }, + { + "epoch": 0.152869335923947, + "grad_norm": 4.276825904846191, + "learning_rate": 0.00019996468775366276, + "loss": 1.1695, + "step": 2203 + }, + { + "epoch": 0.15293872736104364, + "grad_norm": 5.168324947357178, + "learning_rate": 0.00019996453829283873, + "loss": 1.0999, + "step": 2204 + }, + { + "epoch": 0.1530081187981403, + "grad_norm": 4.009620189666748, + "learning_rate": 0.00019996438851643822, + "loss": 1.1337, + "step": 2205 + }, + { + "epoch": 0.15307751023523697, + "grad_norm": 7.201740741729736, + "learning_rate": 0.00019996423842446177, + "loss": 1.2243, + "step": 2206 + }, + { + "epoch": 0.15314690167233364, + "grad_norm": 4.57582950592041, + "learning_rate": 0.0001999640880169098, + "loss": 1.3095, + "step": 2207 + }, + { + "epoch": 0.1532162931094303, + "grad_norm": 6.26279354095459, + "learning_rate": 0.00019996393729378279, + "loss": 1.299, + "step": 2208 + }, + { + "epoch": 0.15328568454652697, + "grad_norm": 3.6303982734680176, + "learning_rate": 0.00019996378625508125, + "loss": 1.0071, + "step": 2209 + }, + { + "epoch": 0.15335507598362363, + "grad_norm": 4.817541122436523, + "learning_rate": 0.00019996363490080565, + "loss": 1.3753, + "step": 2210 + }, + { + "epoch": 0.15342446742072027, + "grad_norm": 3.9542384147644043, + "learning_rate": 0.0001999634832309564, + "loss": 1.2977, + "step": 2211 + }, + { + "epoch": 0.15349385885781694, + "grad_norm": 5.804667949676514, + "learning_rate": 0.0001999633312455341, + "loss": 1.6811, + "step": 2212 + }, + { + "epoch": 0.1535632502949136, + "grad_norm": 4.206712245941162, + "learning_rate": 0.00019996317894453911, + "loss": 1.0392, + "step": 2213 + }, + { + "epoch": 0.15363264173201027, + "grad_norm": 4.4636125564575195, + "learning_rate": 0.00019996302632797197, + "loss": 1.2847, + "step": 2214 + }, + { + "epoch": 0.15370203316910694, + "grad_norm": 4.827295303344727, + "learning_rate": 0.00019996287339583315, + "loss": 1.2967, + "step": 2215 + }, + { + "epoch": 0.1537714246062036, + "grad_norm": 5.285245895385742, + "learning_rate": 0.00019996272014812316, + "loss": 1.2564, + "step": 2216 + }, + { + "epoch": 0.15384081604330024, + "grad_norm": 4.894904613494873, + "learning_rate": 0.00019996256658484245, + "loss": 1.1028, + "step": 2217 + }, + { + "epoch": 0.1539102074803969, + "grad_norm": 14.77458381652832, + "learning_rate": 0.0001999624127059915, + "loss": 1.199, + "step": 2218 + }, + { + "epoch": 0.15397959891749358, + "grad_norm": 5.15801477432251, + "learning_rate": 0.00019996225851157083, + "loss": 0.9373, + "step": 2219 + }, + { + "epoch": 0.15404899035459024, + "grad_norm": 4.514685153961182, + "learning_rate": 0.00019996210400158087, + "loss": 1.1504, + "step": 2220 + }, + { + "epoch": 0.1541183817916869, + "grad_norm": 4.192120552062988, + "learning_rate": 0.0001999619491760222, + "loss": 1.1765, + "step": 2221 + }, + { + "epoch": 0.15418777322878358, + "grad_norm": 4.109709739685059, + "learning_rate": 0.00019996179403489523, + "loss": 1.0842, + "step": 2222 + }, + { + "epoch": 0.15425716466588024, + "grad_norm": 7.0612473487854, + "learning_rate": 0.00019996163857820048, + "loss": 1.1855, + "step": 2223 + }, + { + "epoch": 0.15432655610297688, + "grad_norm": 3.8424248695373535, + "learning_rate": 0.0001999614828059384, + "loss": 1.261, + "step": 2224 + }, + { + "epoch": 0.15439594754007355, + "grad_norm": 4.238419532775879, + "learning_rate": 0.00019996132671810956, + "loss": 0.9006, + "step": 2225 + }, + { + "epoch": 0.1544653389771702, + "grad_norm": 4.460824489593506, + "learning_rate": 0.0001999611703147144, + "loss": 1.3509, + "step": 2226 + }, + { + "epoch": 0.15453473041426688, + "grad_norm": 4.108325004577637, + "learning_rate": 0.00019996101359575342, + "loss": 1.0929, + "step": 2227 + }, + { + "epoch": 0.15460412185136355, + "grad_norm": 4.085665702819824, + "learning_rate": 0.00019996085656122712, + "loss": 1.233, + "step": 2228 + }, + { + "epoch": 0.1546735132884602, + "grad_norm": 4.87822961807251, + "learning_rate": 0.000199960699211136, + "loss": 1.1419, + "step": 2229 + }, + { + "epoch": 0.15474290472555688, + "grad_norm": 5.580414295196533, + "learning_rate": 0.00019996054154548053, + "loss": 1.1586, + "step": 2230 + }, + { + "epoch": 0.15481229616265352, + "grad_norm": 4.183828353881836, + "learning_rate": 0.00019996038356426122, + "loss": 1.259, + "step": 2231 + }, + { + "epoch": 0.15488168759975018, + "grad_norm": 4.093809127807617, + "learning_rate": 0.00019996022526747858, + "loss": 1.1084, + "step": 2232 + }, + { + "epoch": 0.15495107903684685, + "grad_norm": 5.348049640655518, + "learning_rate": 0.0001999600666551331, + "loss": 1.3253, + "step": 2233 + }, + { + "epoch": 0.15502047047394352, + "grad_norm": 4.621278285980225, + "learning_rate": 0.00019995990772722535, + "loss": 1.5839, + "step": 2234 + }, + { + "epoch": 0.15508986191104018, + "grad_norm": 5.198116779327393, + "learning_rate": 0.0001999597484837557, + "loss": 1.2365, + "step": 2235 + }, + { + "epoch": 0.15515925334813685, + "grad_norm": 4.76794958114624, + "learning_rate": 0.00019995958892472478, + "loss": 1.1834, + "step": 2236 + }, + { + "epoch": 0.15522864478523352, + "grad_norm": 4.9369072914123535, + "learning_rate": 0.00019995942905013297, + "loss": 1.0721, + "step": 2237 + }, + { + "epoch": 0.15529803622233015, + "grad_norm": 4.4533772468566895, + "learning_rate": 0.00019995926885998087, + "loss": 1.0959, + "step": 2238 + }, + { + "epoch": 0.15536742765942682, + "grad_norm": 5.01877498626709, + "learning_rate": 0.00019995910835426896, + "loss": 1.3491, + "step": 2239 + }, + { + "epoch": 0.1554368190965235, + "grad_norm": 4.563301086425781, + "learning_rate": 0.00019995894753299775, + "loss": 1.1294, + "step": 2240 + }, + { + "epoch": 0.15550621053362015, + "grad_norm": 4.4800615310668945, + "learning_rate": 0.00019995878639616775, + "loss": 1.1566, + "step": 2241 + }, + { + "epoch": 0.15557560197071682, + "grad_norm": 5.818432331085205, + "learning_rate": 0.00019995862494377942, + "loss": 1.2125, + "step": 2242 + }, + { + "epoch": 0.1556449934078135, + "grad_norm": 5.879472732543945, + "learning_rate": 0.00019995846317583333, + "loss": 1.0239, + "step": 2243 + }, + { + "epoch": 0.15571438484491013, + "grad_norm": 10.482563018798828, + "learning_rate": 0.00019995830109232997, + "loss": 1.2545, + "step": 2244 + }, + { + "epoch": 0.1557837762820068, + "grad_norm": 4.784552574157715, + "learning_rate": 0.00019995813869326984, + "loss": 1.1872, + "step": 2245 + }, + { + "epoch": 0.15585316771910346, + "grad_norm": 5.37955379486084, + "learning_rate": 0.00019995797597865345, + "loss": 1.2599, + "step": 2246 + }, + { + "epoch": 0.15592255915620012, + "grad_norm": 4.264970779418945, + "learning_rate": 0.00019995781294848138, + "loss": 1.1248, + "step": 2247 + }, + { + "epoch": 0.1559919505932968, + "grad_norm": 4.431340217590332, + "learning_rate": 0.00019995764960275403, + "loss": 1.4779, + "step": 2248 + }, + { + "epoch": 0.15606134203039346, + "grad_norm": 4.368905544281006, + "learning_rate": 0.00019995748594147202, + "loss": 1.0412, + "step": 2249 + }, + { + "epoch": 0.15613073346749012, + "grad_norm": 4.635721206665039, + "learning_rate": 0.0001999573219646358, + "loss": 1.2258, + "step": 2250 + }, + { + "epoch": 0.15620012490458676, + "grad_norm": 5.740652084350586, + "learning_rate": 0.00019995715767224593, + "loss": 1.0361, + "step": 2251 + }, + { + "epoch": 0.15626951634168343, + "grad_norm": 6.558762550354004, + "learning_rate": 0.0001999569930643029, + "loss": 1.2245, + "step": 2252 + }, + { + "epoch": 0.1563389077787801, + "grad_norm": 5.880007743835449, + "learning_rate": 0.00019995682814080725, + "loss": 1.2222, + "step": 2253 + }, + { + "epoch": 0.15640829921587676, + "grad_norm": 4.976191997528076, + "learning_rate": 0.00019995666290175947, + "loss": 1.0865, + "step": 2254 + }, + { + "epoch": 0.15647769065297343, + "grad_norm": 4.1556782722473145, + "learning_rate": 0.0001999564973471601, + "loss": 1.039, + "step": 2255 + }, + { + "epoch": 0.1565470820900701, + "grad_norm": 6.305001258850098, + "learning_rate": 0.0001999563314770097, + "loss": 1.5027, + "step": 2256 + }, + { + "epoch": 0.15661647352716676, + "grad_norm": 4.6795806884765625, + "learning_rate": 0.00019995616529130874, + "loss": 1.0415, + "step": 2257 + }, + { + "epoch": 0.1566858649642634, + "grad_norm": 4.185312271118164, + "learning_rate": 0.00019995599879005775, + "loss": 1.0365, + "step": 2258 + }, + { + "epoch": 0.15675525640136007, + "grad_norm": 6.4461989402771, + "learning_rate": 0.0001999558319732573, + "loss": 1.3902, + "step": 2259 + }, + { + "epoch": 0.15682464783845673, + "grad_norm": 3.8833541870117188, + "learning_rate": 0.00019995566484090786, + "loss": 1.1056, + "step": 2260 + }, + { + "epoch": 0.1568940392755534, + "grad_norm": 7.98937463760376, + "learning_rate": 0.00019995549739300998, + "loss": 1.0463, + "step": 2261 + }, + { + "epoch": 0.15696343071265006, + "grad_norm": 4.726850986480713, + "learning_rate": 0.0001999553296295642, + "loss": 1.4321, + "step": 2262 + }, + { + "epoch": 0.15703282214974673, + "grad_norm": 3.6985878944396973, + "learning_rate": 0.00019995516155057102, + "loss": 1.2256, + "step": 2263 + }, + { + "epoch": 0.15710221358684337, + "grad_norm": 4.413760185241699, + "learning_rate": 0.00019995499315603104, + "loss": 1.4534, + "step": 2264 + }, + { + "epoch": 0.15717160502394004, + "grad_norm": 5.496564865112305, + "learning_rate": 0.00019995482444594474, + "loss": 1.5935, + "step": 2265 + }, + { + "epoch": 0.1572409964610367, + "grad_norm": 4.8774847984313965, + "learning_rate": 0.00019995465542031263, + "loss": 1.2487, + "step": 2266 + }, + { + "epoch": 0.15731038789813337, + "grad_norm": 3.0246870517730713, + "learning_rate": 0.00019995448607913527, + "loss": 0.803, + "step": 2267 + }, + { + "epoch": 0.15737977933523004, + "grad_norm": 5.031855583190918, + "learning_rate": 0.0001999543164224132, + "loss": 1.1235, + "step": 2268 + }, + { + "epoch": 0.1574491707723267, + "grad_norm": 3.7840893268585205, + "learning_rate": 0.00019995414645014695, + "loss": 1.1917, + "step": 2269 + }, + { + "epoch": 0.15751856220942337, + "grad_norm": 4.504709243774414, + "learning_rate": 0.00019995397616233705, + "loss": 0.9582, + "step": 2270 + }, + { + "epoch": 0.15758795364652, + "grad_norm": 10.172966003417969, + "learning_rate": 0.00019995380555898405, + "loss": 1.2667, + "step": 2271 + }, + { + "epoch": 0.15765734508361667, + "grad_norm": 4.672337055206299, + "learning_rate": 0.00019995363464008848, + "loss": 1.4159, + "step": 2272 + }, + { + "epoch": 0.15772673652071334, + "grad_norm": 5.711227893829346, + "learning_rate": 0.00019995346340565088, + "loss": 1.0202, + "step": 2273 + }, + { + "epoch": 0.15779612795781, + "grad_norm": 5.706086158752441, + "learning_rate": 0.0001999532918556718, + "loss": 1.1448, + "step": 2274 + }, + { + "epoch": 0.15786551939490667, + "grad_norm": 4.6661906242370605, + "learning_rate": 0.00019995311999015177, + "loss": 1.2394, + "step": 2275 + }, + { + "epoch": 0.15793491083200334, + "grad_norm": 6.229702949523926, + "learning_rate": 0.00019995294780909132, + "loss": 1.4036, + "step": 2276 + }, + { + "epoch": 0.1580043022691, + "grad_norm": 5.110751152038574, + "learning_rate": 0.00019995277531249103, + "loss": 1.3637, + "step": 2277 + }, + { + "epoch": 0.15807369370619664, + "grad_norm": 4.380594730377197, + "learning_rate": 0.00019995260250035143, + "loss": 1.1186, + "step": 2278 + }, + { + "epoch": 0.1581430851432933, + "grad_norm": 4.845581531524658, + "learning_rate": 0.00019995242937267306, + "loss": 1.0556, + "step": 2279 + }, + { + "epoch": 0.15821247658038998, + "grad_norm": 4.920147895812988, + "learning_rate": 0.00019995225592945646, + "loss": 1.1275, + "step": 2280 + }, + { + "epoch": 0.15828186801748664, + "grad_norm": 4.579648494720459, + "learning_rate": 0.0001999520821707022, + "loss": 0.9759, + "step": 2281 + }, + { + "epoch": 0.1583512594545833, + "grad_norm": 4.615515232086182, + "learning_rate": 0.00019995190809641078, + "loss": 1.3425, + "step": 2282 + }, + { + "epoch": 0.15842065089167998, + "grad_norm": 6.148870944976807, + "learning_rate": 0.0001999517337065828, + "loss": 1.1846, + "step": 2283 + }, + { + "epoch": 0.15849004232877664, + "grad_norm": 5.313032627105713, + "learning_rate": 0.0001999515590012188, + "loss": 1.1689, + "step": 2284 + }, + { + "epoch": 0.15855943376587328, + "grad_norm": 4.086667060852051, + "learning_rate": 0.00019995138398031933, + "loss": 1.1307, + "step": 2285 + }, + { + "epoch": 0.15862882520296995, + "grad_norm": 5.032186508178711, + "learning_rate": 0.00019995120864388494, + "loss": 1.1912, + "step": 2286 + }, + { + "epoch": 0.15869821664006661, + "grad_norm": 3.6356756687164307, + "learning_rate": 0.00019995103299191614, + "loss": 0.9402, + "step": 2287 + }, + { + "epoch": 0.15876760807716328, + "grad_norm": 5.746941089630127, + "learning_rate": 0.00019995085702441358, + "loss": 1.3242, + "step": 2288 + }, + { + "epoch": 0.15883699951425995, + "grad_norm": 4.8551740646362305, + "learning_rate": 0.00019995068074137772, + "loss": 1.281, + "step": 2289 + }, + { + "epoch": 0.1589063909513566, + "grad_norm": 4.354562282562256, + "learning_rate": 0.00019995050414280916, + "loss": 1.1682, + "step": 2290 + }, + { + "epoch": 0.15897578238845325, + "grad_norm": 8.023040771484375, + "learning_rate": 0.00019995032722870847, + "loss": 1.54, + "step": 2291 + }, + { + "epoch": 0.15904517382554992, + "grad_norm": 4.705010414123535, + "learning_rate": 0.00019995014999907618, + "loss": 1.2392, + "step": 2292 + }, + { + "epoch": 0.15911456526264658, + "grad_norm": 3.973090410232544, + "learning_rate": 0.00019994997245391288, + "loss": 1.3641, + "step": 2293 + }, + { + "epoch": 0.15918395669974325, + "grad_norm": 3.951809883117676, + "learning_rate": 0.0001999497945932191, + "loss": 1.0801, + "step": 2294 + }, + { + "epoch": 0.15925334813683992, + "grad_norm": 4.202714920043945, + "learning_rate": 0.00019994961641699543, + "loss": 1.2942, + "step": 2295 + }, + { + "epoch": 0.15932273957393658, + "grad_norm": 5.050889015197754, + "learning_rate": 0.00019994943792524242, + "loss": 1.403, + "step": 2296 + }, + { + "epoch": 0.15939213101103325, + "grad_norm": 3.7386293411254883, + "learning_rate": 0.0001999492591179606, + "loss": 1.0477, + "step": 2297 + }, + { + "epoch": 0.1594615224481299, + "grad_norm": 3.607943534851074, + "learning_rate": 0.00019994907999515058, + "loss": 1.0957, + "step": 2298 + }, + { + "epoch": 0.15953091388522656, + "grad_norm": 3.832850933074951, + "learning_rate": 0.00019994890055681294, + "loss": 1.0106, + "step": 2299 + }, + { + "epoch": 0.15960030532232322, + "grad_norm": 7.0116682052612305, + "learning_rate": 0.00019994872080294815, + "loss": 1.12, + "step": 2300 + }, + { + "epoch": 0.1596696967594199, + "grad_norm": 5.980141639709473, + "learning_rate": 0.0001999485407335569, + "loss": 0.9727, + "step": 2301 + }, + { + "epoch": 0.15973908819651655, + "grad_norm": 7.114522457122803, + "learning_rate": 0.00019994836034863968, + "loss": 1.2052, + "step": 2302 + }, + { + "epoch": 0.15980847963361322, + "grad_norm": 6.303493976593018, + "learning_rate": 0.0001999481796481971, + "loss": 1.3213, + "step": 2303 + }, + { + "epoch": 0.1598778710707099, + "grad_norm": 4.003570079803467, + "learning_rate": 0.0001999479986322297, + "loss": 1.1833, + "step": 2304 + }, + { + "epoch": 0.15994726250780653, + "grad_norm": 4.687757968902588, + "learning_rate": 0.0001999478173007381, + "loss": 1.1974, + "step": 2305 + }, + { + "epoch": 0.1600166539449032, + "grad_norm": 4.378289222717285, + "learning_rate": 0.0001999476356537228, + "loss": 1.2572, + "step": 2306 + }, + { + "epoch": 0.16008604538199986, + "grad_norm": 5.007067680358887, + "learning_rate": 0.00019994745369118438, + "loss": 1.2076, + "step": 2307 + }, + { + "epoch": 0.16015543681909652, + "grad_norm": 4.141595363616943, + "learning_rate": 0.00019994727141312352, + "loss": 1.2382, + "step": 2308 + }, + { + "epoch": 0.1602248282561932, + "grad_norm": 3.1796696186065674, + "learning_rate": 0.00019994708881954066, + "loss": 1.0519, + "step": 2309 + }, + { + "epoch": 0.16029421969328986, + "grad_norm": 6.054597854614258, + "learning_rate": 0.00019994690591043646, + "loss": 1.5804, + "step": 2310 + }, + { + "epoch": 0.16036361113038652, + "grad_norm": 4.42647647857666, + "learning_rate": 0.00019994672268581148, + "loss": 1.6566, + "step": 2311 + }, + { + "epoch": 0.16043300256748316, + "grad_norm": 4.502488613128662, + "learning_rate": 0.00019994653914566626, + "loss": 1.5183, + "step": 2312 + }, + { + "epoch": 0.16050239400457983, + "grad_norm": 5.619850158691406, + "learning_rate": 0.00019994635529000144, + "loss": 1.2266, + "step": 2313 + }, + { + "epoch": 0.1605717854416765, + "grad_norm": 3.6724696159362793, + "learning_rate": 0.00019994617111881757, + "loss": 1.139, + "step": 2314 + }, + { + "epoch": 0.16064117687877316, + "grad_norm": 3.7904436588287354, + "learning_rate": 0.00019994598663211524, + "loss": 1.0449, + "step": 2315 + }, + { + "epoch": 0.16071056831586983, + "grad_norm": 4.276314735412598, + "learning_rate": 0.00019994580182989503, + "loss": 1.3375, + "step": 2316 + }, + { + "epoch": 0.1607799597529665, + "grad_norm": 7.324974536895752, + "learning_rate": 0.00019994561671215748, + "loss": 1.1614, + "step": 2317 + }, + { + "epoch": 0.16084935119006313, + "grad_norm": 4.581085205078125, + "learning_rate": 0.00019994543127890325, + "loss": 1.0649, + "step": 2318 + }, + { + "epoch": 0.1609187426271598, + "grad_norm": 4.062494277954102, + "learning_rate": 0.00019994524553013288, + "loss": 1.0212, + "step": 2319 + }, + { + "epoch": 0.16098813406425647, + "grad_norm": 6.220315456390381, + "learning_rate": 0.00019994505946584697, + "loss": 1.1625, + "step": 2320 + }, + { + "epoch": 0.16105752550135313, + "grad_norm": 5.419581413269043, + "learning_rate": 0.00019994487308604608, + "loss": 1.1164, + "step": 2321 + }, + { + "epoch": 0.1611269169384498, + "grad_norm": 7.105452537536621, + "learning_rate": 0.00019994468639073084, + "loss": 1.0975, + "step": 2322 + }, + { + "epoch": 0.16119630837554647, + "grad_norm": 5.648862361907959, + "learning_rate": 0.00019994449937990182, + "loss": 1.4999, + "step": 2323 + }, + { + "epoch": 0.16126569981264313, + "grad_norm": 4.023277759552002, + "learning_rate": 0.0001999443120535596, + "loss": 1.185, + "step": 2324 + }, + { + "epoch": 0.16133509124973977, + "grad_norm": 4.118855953216553, + "learning_rate": 0.00019994412441170484, + "loss": 1.4858, + "step": 2325 + }, + { + "epoch": 0.16140448268683644, + "grad_norm": 5.016485691070557, + "learning_rate": 0.00019994393645433802, + "loss": 1.4471, + "step": 2326 + }, + { + "epoch": 0.1614738741239331, + "grad_norm": 4.845316410064697, + "learning_rate": 0.0001999437481814598, + "loss": 1.0928, + "step": 2327 + }, + { + "epoch": 0.16154326556102977, + "grad_norm": 3.4662952423095703, + "learning_rate": 0.00019994355959307075, + "loss": 1.0987, + "step": 2328 + }, + { + "epoch": 0.16161265699812644, + "grad_norm": 4.859582901000977, + "learning_rate": 0.00019994337068917148, + "loss": 1.6416, + "step": 2329 + }, + { + "epoch": 0.1616820484352231, + "grad_norm": 6.301947593688965, + "learning_rate": 0.0001999431814697626, + "loss": 1.1411, + "step": 2330 + }, + { + "epoch": 0.16175143987231977, + "grad_norm": 4.361723899841309, + "learning_rate": 0.0001999429919348447, + "loss": 1.2622, + "step": 2331 + }, + { + "epoch": 0.1618208313094164, + "grad_norm": 3.3102803230285645, + "learning_rate": 0.00019994280208441832, + "loss": 1.0645, + "step": 2332 + }, + { + "epoch": 0.16189022274651307, + "grad_norm": 3.669574022293091, + "learning_rate": 0.0001999426119184842, + "loss": 1.2376, + "step": 2333 + }, + { + "epoch": 0.16195961418360974, + "grad_norm": 3.453591823577881, + "learning_rate": 0.00019994242143704275, + "loss": 1.2582, + "step": 2334 + }, + { + "epoch": 0.1620290056207064, + "grad_norm": 19.358667373657227, + "learning_rate": 0.00019994223064009471, + "loss": 1.0128, + "step": 2335 + }, + { + "epoch": 0.16209839705780307, + "grad_norm": 5.842452049255371, + "learning_rate": 0.00019994203952764065, + "loss": 0.995, + "step": 2336 + }, + { + "epoch": 0.16216778849489974, + "grad_norm": 25.8676700592041, + "learning_rate": 0.00019994184809968117, + "loss": 1.7153, + "step": 2337 + }, + { + "epoch": 0.16223717993199638, + "grad_norm": 5.743529796600342, + "learning_rate": 0.00019994165635621686, + "loss": 0.8773, + "step": 2338 + }, + { + "epoch": 0.16230657136909304, + "grad_norm": 6.062124729156494, + "learning_rate": 0.00019994146429724836, + "loss": 1.1952, + "step": 2339 + }, + { + "epoch": 0.1623759628061897, + "grad_norm": 5.318705081939697, + "learning_rate": 0.00019994127192277624, + "loss": 1.1029, + "step": 2340 + }, + { + "epoch": 0.16244535424328638, + "grad_norm": 5.776012897491455, + "learning_rate": 0.0001999410792328011, + "loss": 1.2034, + "step": 2341 + }, + { + "epoch": 0.16251474568038304, + "grad_norm": 4.883278846740723, + "learning_rate": 0.00019994088622732358, + "loss": 1.4288, + "step": 2342 + }, + { + "epoch": 0.1625841371174797, + "grad_norm": 17.35283851623535, + "learning_rate": 0.00019994069290634427, + "loss": 1.2445, + "step": 2343 + }, + { + "epoch": 0.16265352855457638, + "grad_norm": 6.9218363761901855, + "learning_rate": 0.00019994049926986383, + "loss": 1.1539, + "step": 2344 + }, + { + "epoch": 0.16272291999167302, + "grad_norm": 6.58875036239624, + "learning_rate": 0.00019994030531788278, + "loss": 1.3545, + "step": 2345 + }, + { + "epoch": 0.16279231142876968, + "grad_norm": 4.24623441696167, + "learning_rate": 0.00019994011105040177, + "loss": 1.1012, + "step": 2346 + }, + { + "epoch": 0.16286170286586635, + "grad_norm": 4.09136962890625, + "learning_rate": 0.00019993991646742148, + "loss": 0.9705, + "step": 2347 + }, + { + "epoch": 0.16293109430296301, + "grad_norm": 4.7717814445495605, + "learning_rate": 0.00019993972156894243, + "loss": 1.2088, + "step": 2348 + }, + { + "epoch": 0.16300048574005968, + "grad_norm": 3.966447591781616, + "learning_rate": 0.00019993952635496528, + "loss": 1.1811, + "step": 2349 + }, + { + "epoch": 0.16306987717715635, + "grad_norm": 6.842019081115723, + "learning_rate": 0.00019993933082549064, + "loss": 1.2805, + "step": 2350 + }, + { + "epoch": 0.163139268614253, + "grad_norm": 4.102146625518799, + "learning_rate": 0.00019993913498051914, + "loss": 1.169, + "step": 2351 + }, + { + "epoch": 0.16320866005134965, + "grad_norm": 4.9484357833862305, + "learning_rate": 0.00019993893882005139, + "loss": 1.4828, + "step": 2352 + }, + { + "epoch": 0.16327805148844632, + "grad_norm": 8.678337097167969, + "learning_rate": 0.000199938742344088, + "loss": 1.0399, + "step": 2353 + }, + { + "epoch": 0.16334744292554298, + "grad_norm": 5.271059989929199, + "learning_rate": 0.00019993854555262958, + "loss": 1.1389, + "step": 2354 + }, + { + "epoch": 0.16341683436263965, + "grad_norm": 5.289060115814209, + "learning_rate": 0.00019993834844567678, + "loss": 1.3029, + "step": 2355 + }, + { + "epoch": 0.16348622579973632, + "grad_norm": 8.141990661621094, + "learning_rate": 0.00019993815102323023, + "loss": 1.3372, + "step": 2356 + }, + { + "epoch": 0.16355561723683298, + "grad_norm": 11.125157356262207, + "learning_rate": 0.0001999379532852905, + "loss": 1.3502, + "step": 2357 + }, + { + "epoch": 0.16362500867392965, + "grad_norm": 5.329672336578369, + "learning_rate": 0.00019993775523185825, + "loss": 1.3452, + "step": 2358 + }, + { + "epoch": 0.1636944001110263, + "grad_norm": 4.8320207595825195, + "learning_rate": 0.00019993755686293412, + "loss": 1.1193, + "step": 2359 + }, + { + "epoch": 0.16376379154812296, + "grad_norm": 4.25202751159668, + "learning_rate": 0.00019993735817851868, + "loss": 1.1924, + "step": 2360 + }, + { + "epoch": 0.16383318298521962, + "grad_norm": 5.569819927215576, + "learning_rate": 0.0001999371591786126, + "loss": 1.2241, + "step": 2361 + }, + { + "epoch": 0.1639025744223163, + "grad_norm": 6.169849872589111, + "learning_rate": 0.00019993695986321656, + "loss": 1.2043, + "step": 2362 + }, + { + "epoch": 0.16397196585941295, + "grad_norm": 4.536948204040527, + "learning_rate": 0.00019993676023233109, + "loss": 1.1668, + "step": 2363 + }, + { + "epoch": 0.16404135729650962, + "grad_norm": 6.793606281280518, + "learning_rate": 0.00019993656028595688, + "loss": 1.7811, + "step": 2364 + }, + { + "epoch": 0.16411074873360626, + "grad_norm": 9.364388465881348, + "learning_rate": 0.00019993636002409453, + "loss": 1.0858, + "step": 2365 + }, + { + "epoch": 0.16418014017070293, + "grad_norm": 4.208866119384766, + "learning_rate": 0.00019993615944674468, + "loss": 1.3973, + "step": 2366 + }, + { + "epoch": 0.1642495316077996, + "grad_norm": 6.890744209289551, + "learning_rate": 0.00019993595855390798, + "loss": 1.5643, + "step": 2367 + }, + { + "epoch": 0.16431892304489626, + "grad_norm": 6.242718696594238, + "learning_rate": 0.00019993575734558502, + "loss": 1.0536, + "step": 2368 + }, + { + "epoch": 0.16438831448199293, + "grad_norm": 6.681387424468994, + "learning_rate": 0.00019993555582177653, + "loss": 1.24, + "step": 2369 + }, + { + "epoch": 0.1644577059190896, + "grad_norm": 5.375817775726318, + "learning_rate": 0.00019993535398248304, + "loss": 1.2955, + "step": 2370 + }, + { + "epoch": 0.16452709735618626, + "grad_norm": 5.004377365112305, + "learning_rate": 0.00019993515182770524, + "loss": 1.2795, + "step": 2371 + }, + { + "epoch": 0.1645964887932829, + "grad_norm": 6.226109504699707, + "learning_rate": 0.0001999349493574438, + "loss": 1.3659, + "step": 2372 + }, + { + "epoch": 0.16466588023037956, + "grad_norm": 4.3169684410095215, + "learning_rate": 0.00019993474657169926, + "loss": 1.0622, + "step": 2373 + }, + { + "epoch": 0.16473527166747623, + "grad_norm": 16.060321807861328, + "learning_rate": 0.00019993454347047235, + "loss": 1.1107, + "step": 2374 + }, + { + "epoch": 0.1648046631045729, + "grad_norm": 9.78807544708252, + "learning_rate": 0.00019993434005376369, + "loss": 1.4098, + "step": 2375 + }, + { + "epoch": 0.16487405454166956, + "grad_norm": 6.490010738372803, + "learning_rate": 0.00019993413632157388, + "loss": 1.4727, + "step": 2376 + }, + { + "epoch": 0.16494344597876623, + "grad_norm": 4.607236385345459, + "learning_rate": 0.00019993393227390362, + "loss": 1.1326, + "step": 2377 + }, + { + "epoch": 0.1650128374158629, + "grad_norm": 4.945221424102783, + "learning_rate": 0.00019993372791075352, + "loss": 1.2079, + "step": 2378 + }, + { + "epoch": 0.16508222885295953, + "grad_norm": 5.863796234130859, + "learning_rate": 0.00019993352323212425, + "loss": 1.5164, + "step": 2379 + }, + { + "epoch": 0.1651516202900562, + "grad_norm": 4.410670757293701, + "learning_rate": 0.00019993331823801644, + "loss": 1.2988, + "step": 2380 + }, + { + "epoch": 0.16522101172715287, + "grad_norm": 4.349693298339844, + "learning_rate": 0.00019993311292843072, + "loss": 1.4892, + "step": 2381 + }, + { + "epoch": 0.16529040316424953, + "grad_norm": 4.840583801269531, + "learning_rate": 0.00019993290730336777, + "loss": 1.3565, + "step": 2382 + }, + { + "epoch": 0.1653597946013462, + "grad_norm": 4.422557353973389, + "learning_rate": 0.00019993270136282822, + "loss": 1.2391, + "step": 2383 + }, + { + "epoch": 0.16542918603844287, + "grad_norm": 12.899394035339355, + "learning_rate": 0.00019993249510681275, + "loss": 1.3358, + "step": 2384 + }, + { + "epoch": 0.16549857747553953, + "grad_norm": 8.568646430969238, + "learning_rate": 0.00019993228853532196, + "loss": 1.2869, + "step": 2385 + }, + { + "epoch": 0.16556796891263617, + "grad_norm": 5.692032337188721, + "learning_rate": 0.00019993208164835653, + "loss": 1.1933, + "step": 2386 + }, + { + "epoch": 0.16563736034973284, + "grad_norm": 4.542912006378174, + "learning_rate": 0.00019993187444591713, + "loss": 1.4116, + "step": 2387 + }, + { + "epoch": 0.1657067517868295, + "grad_norm": 3.5921430587768555, + "learning_rate": 0.00019993166692800437, + "loss": 1.2889, + "step": 2388 + }, + { + "epoch": 0.16577614322392617, + "grad_norm": 4.721619606018066, + "learning_rate": 0.00019993145909461896, + "loss": 1.4757, + "step": 2389 + }, + { + "epoch": 0.16584553466102284, + "grad_norm": 4.406286716461182, + "learning_rate": 0.00019993125094576152, + "loss": 1.2608, + "step": 2390 + }, + { + "epoch": 0.1659149260981195, + "grad_norm": 4.177829265594482, + "learning_rate": 0.00019993104248143274, + "loss": 1.1051, + "step": 2391 + }, + { + "epoch": 0.16598431753521614, + "grad_norm": 5.039063930511475, + "learning_rate": 0.0001999308337016332, + "loss": 1.2403, + "step": 2392 + }, + { + "epoch": 0.1660537089723128, + "grad_norm": 4.916114330291748, + "learning_rate": 0.00019993062460636364, + "loss": 1.2925, + "step": 2393 + }, + { + "epoch": 0.16612310040940947, + "grad_norm": 3.881234645843506, + "learning_rate": 0.00019993041519562467, + "loss": 1.1901, + "step": 2394 + }, + { + "epoch": 0.16619249184650614, + "grad_norm": 3.944396734237671, + "learning_rate": 0.000199930205469417, + "loss": 1.0731, + "step": 2395 + }, + { + "epoch": 0.1662618832836028, + "grad_norm": 3.9239747524261475, + "learning_rate": 0.00019992999542774127, + "loss": 1.0142, + "step": 2396 + }, + { + "epoch": 0.16633127472069947, + "grad_norm": 4.432819366455078, + "learning_rate": 0.0001999297850705981, + "loss": 1.2565, + "step": 2397 + }, + { + "epoch": 0.16640066615779614, + "grad_norm": 4.428493499755859, + "learning_rate": 0.00019992957439798824, + "loss": 1.226, + "step": 2398 + }, + { + "epoch": 0.16647005759489278, + "grad_norm": 4.88680362701416, + "learning_rate": 0.00019992936340991227, + "loss": 1.2092, + "step": 2399 + }, + { + "epoch": 0.16653944903198944, + "grad_norm": 4.106476783752441, + "learning_rate": 0.00019992915210637093, + "loss": 1.2772, + "step": 2400 + }, + { + "epoch": 0.1666088404690861, + "grad_norm": 4.172744274139404, + "learning_rate": 0.0001999289404873648, + "loss": 1.3045, + "step": 2401 + }, + { + "epoch": 0.16667823190618278, + "grad_norm": 4.4927520751953125, + "learning_rate": 0.00019992872855289465, + "loss": 1.2733, + "step": 2402 + }, + { + "epoch": 0.16674762334327944, + "grad_norm": 4.573215484619141, + "learning_rate": 0.00019992851630296107, + "loss": 1.2716, + "step": 2403 + }, + { + "epoch": 0.1668170147803761, + "grad_norm": 4.326798439025879, + "learning_rate": 0.00019992830373756476, + "loss": 1.1616, + "step": 2404 + }, + { + "epoch": 0.16688640621747278, + "grad_norm": 4.172313213348389, + "learning_rate": 0.0001999280908567064, + "loss": 1.2132, + "step": 2405 + }, + { + "epoch": 0.16695579765456942, + "grad_norm": 3.125476598739624, + "learning_rate": 0.00019992787766038667, + "loss": 0.8609, + "step": 2406 + }, + { + "epoch": 0.16702518909166608, + "grad_norm": 4.98372745513916, + "learning_rate": 0.00019992766414860617, + "loss": 0.912, + "step": 2407 + }, + { + "epoch": 0.16709458052876275, + "grad_norm": 5.998163223266602, + "learning_rate": 0.00019992745032136567, + "loss": 1.3377, + "step": 2408 + }, + { + "epoch": 0.16716397196585941, + "grad_norm": 5.303796291351318, + "learning_rate": 0.0001999272361786658, + "loss": 1.2599, + "step": 2409 + }, + { + "epoch": 0.16723336340295608, + "grad_norm": 4.093478679656982, + "learning_rate": 0.00019992702172050723, + "loss": 1.0286, + "step": 2410 + }, + { + "epoch": 0.16730275484005275, + "grad_norm": 4.17225456237793, + "learning_rate": 0.00019992680694689067, + "loss": 1.2719, + "step": 2411 + }, + { + "epoch": 0.1673721462771494, + "grad_norm": 4.572384357452393, + "learning_rate": 0.00019992659185781673, + "loss": 1.3076, + "step": 2412 + }, + { + "epoch": 0.16744153771424605, + "grad_norm": 3.495453357696533, + "learning_rate": 0.00019992637645328617, + "loss": 1.0623, + "step": 2413 + }, + { + "epoch": 0.16751092915134272, + "grad_norm": 4.6958184242248535, + "learning_rate": 0.00019992616073329956, + "loss": 1.1808, + "step": 2414 + }, + { + "epoch": 0.16758032058843939, + "grad_norm": 3.908421039581299, + "learning_rate": 0.00019992594469785772, + "loss": 1.1216, + "step": 2415 + }, + { + "epoch": 0.16764971202553605, + "grad_norm": 4.025219917297363, + "learning_rate": 0.0001999257283469613, + "loss": 1.2696, + "step": 2416 + }, + { + "epoch": 0.16771910346263272, + "grad_norm": 5.026100158691406, + "learning_rate": 0.00019992551168061085, + "loss": 1.0, + "step": 2417 + }, + { + "epoch": 0.16778849489972938, + "grad_norm": 4.681856155395508, + "learning_rate": 0.00019992529469880724, + "loss": 1.2455, + "step": 2418 + }, + { + "epoch": 0.16785788633682602, + "grad_norm": 4.804892539978027, + "learning_rate": 0.00019992507740155104, + "loss": 1.2301, + "step": 2419 + }, + { + "epoch": 0.1679272777739227, + "grad_norm": 4.615351676940918, + "learning_rate": 0.00019992485978884296, + "loss": 1.3393, + "step": 2420 + }, + { + "epoch": 0.16799666921101936, + "grad_norm": 4.176395893096924, + "learning_rate": 0.00019992464186068367, + "loss": 1.2587, + "step": 2421 + }, + { + "epoch": 0.16806606064811602, + "grad_norm": 3.950364351272583, + "learning_rate": 0.0001999244236170739, + "loss": 1.0947, + "step": 2422 + }, + { + "epoch": 0.1681354520852127, + "grad_norm": 5.110362529754639, + "learning_rate": 0.00019992420505801436, + "loss": 1.363, + "step": 2423 + }, + { + "epoch": 0.16820484352230936, + "grad_norm": 4.143012523651123, + "learning_rate": 0.00019992398618350562, + "loss": 1.0953, + "step": 2424 + }, + { + "epoch": 0.16827423495940602, + "grad_norm": 4.190120220184326, + "learning_rate": 0.0001999237669935485, + "loss": 1.0141, + "step": 2425 + }, + { + "epoch": 0.16834362639650266, + "grad_norm": 6.288783550262451, + "learning_rate": 0.00019992354748814362, + "loss": 1.3136, + "step": 2426 + }, + { + "epoch": 0.16841301783359933, + "grad_norm": 3.7744343280792236, + "learning_rate": 0.0001999233276672917, + "loss": 1.0916, + "step": 2427 + }, + { + "epoch": 0.168482409270696, + "grad_norm": 4.1511454582214355, + "learning_rate": 0.00019992310753099343, + "loss": 1.2226, + "step": 2428 + }, + { + "epoch": 0.16855180070779266, + "grad_norm": 5.054905414581299, + "learning_rate": 0.00019992288707924952, + "loss": 1.3404, + "step": 2429 + }, + { + "epoch": 0.16862119214488933, + "grad_norm": 3.983553171157837, + "learning_rate": 0.00019992266631206062, + "loss": 1.2704, + "step": 2430 + }, + { + "epoch": 0.168690583581986, + "grad_norm": 4.96087646484375, + "learning_rate": 0.00019992244522942744, + "loss": 1.1856, + "step": 2431 + }, + { + "epoch": 0.16875997501908266, + "grad_norm": 3.775891065597534, + "learning_rate": 0.00019992222383135073, + "loss": 1.2416, + "step": 2432 + }, + { + "epoch": 0.1688293664561793, + "grad_norm": 4.278141021728516, + "learning_rate": 0.00019992200211783114, + "loss": 0.9748, + "step": 2433 + }, + { + "epoch": 0.16889875789327596, + "grad_norm": 3.4993770122528076, + "learning_rate": 0.00019992178008886938, + "loss": 1.24, + "step": 2434 + }, + { + "epoch": 0.16896814933037263, + "grad_norm": 3.122760534286499, + "learning_rate": 0.00019992155774446616, + "loss": 1.0786, + "step": 2435 + }, + { + "epoch": 0.1690375407674693, + "grad_norm": 4.924180507659912, + "learning_rate": 0.0001999213350846222, + "loss": 1.1227, + "step": 2436 + }, + { + "epoch": 0.16910693220456596, + "grad_norm": 3.6299850940704346, + "learning_rate": 0.00019992111210933813, + "loss": 1.1484, + "step": 2437 + }, + { + "epoch": 0.16917632364166263, + "grad_norm": 3.5744564533233643, + "learning_rate": 0.00019992088881861473, + "loss": 0.8554, + "step": 2438 + }, + { + "epoch": 0.16924571507875927, + "grad_norm": 5.15972375869751, + "learning_rate": 0.00019992066521245266, + "loss": 1.1515, + "step": 2439 + }, + { + "epoch": 0.16931510651585593, + "grad_norm": 4.538723945617676, + "learning_rate": 0.00019992044129085265, + "loss": 0.9113, + "step": 2440 + }, + { + "epoch": 0.1693844979529526, + "grad_norm": 4.2872161865234375, + "learning_rate": 0.0001999202170538154, + "loss": 1.2048, + "step": 2441 + }, + { + "epoch": 0.16945388939004927, + "grad_norm": 5.092288970947266, + "learning_rate": 0.00019991999250134162, + "loss": 1.2814, + "step": 2442 + }, + { + "epoch": 0.16952328082714593, + "grad_norm": 5.597603797912598, + "learning_rate": 0.00019991976763343203, + "loss": 1.1485, + "step": 2443 + }, + { + "epoch": 0.1695926722642426, + "grad_norm": 4.0022406578063965, + "learning_rate": 0.0001999195424500873, + "loss": 1.1624, + "step": 2444 + }, + { + "epoch": 0.16966206370133927, + "grad_norm": 4.801334857940674, + "learning_rate": 0.0001999193169513082, + "loss": 0.9226, + "step": 2445 + }, + { + "epoch": 0.1697314551384359, + "grad_norm": 7.226405620574951, + "learning_rate": 0.00019991909113709538, + "loss": 1.2251, + "step": 2446 + }, + { + "epoch": 0.16980084657553257, + "grad_norm": 4.389807224273682, + "learning_rate": 0.00019991886500744959, + "loss": 0.8952, + "step": 2447 + }, + { + "epoch": 0.16987023801262924, + "grad_norm": 5.165872097015381, + "learning_rate": 0.0001999186385623715, + "loss": 1.5518, + "step": 2448 + }, + { + "epoch": 0.1699396294497259, + "grad_norm": 4.6309099197387695, + "learning_rate": 0.00019991841180186192, + "loss": 1.4026, + "step": 2449 + }, + { + "epoch": 0.17000902088682257, + "grad_norm": 6.095749855041504, + "learning_rate": 0.0001999181847259215, + "loss": 1.2109, + "step": 2450 + }, + { + "epoch": 0.17007841232391924, + "grad_norm": 4.213238716125488, + "learning_rate": 0.00019991795733455094, + "loss": 1.1764, + "step": 2451 + }, + { + "epoch": 0.1701478037610159, + "grad_norm": 4.142665386199951, + "learning_rate": 0.000199917729627751, + "loss": 1.0369, + "step": 2452 + }, + { + "epoch": 0.17021719519811254, + "grad_norm": 4.267261028289795, + "learning_rate": 0.00019991750160552237, + "loss": 1.2589, + "step": 2453 + }, + { + "epoch": 0.1702865866352092, + "grad_norm": 4.75944709777832, + "learning_rate": 0.00019991727326786575, + "loss": 1.2042, + "step": 2454 + }, + { + "epoch": 0.17035597807230587, + "grad_norm": 5.238974571228027, + "learning_rate": 0.00019991704461478195, + "loss": 1.2451, + "step": 2455 + }, + { + "epoch": 0.17042536950940254, + "grad_norm": 5.176207542419434, + "learning_rate": 0.0001999168156462716, + "loss": 1.3176, + "step": 2456 + }, + { + "epoch": 0.1704947609464992, + "grad_norm": 4.88720178604126, + "learning_rate": 0.00019991658636233546, + "loss": 1.0883, + "step": 2457 + }, + { + "epoch": 0.17056415238359587, + "grad_norm": 4.831857204437256, + "learning_rate": 0.00019991635676297424, + "loss": 1.2025, + "step": 2458 + }, + { + "epoch": 0.17063354382069254, + "grad_norm": 4.462563514709473, + "learning_rate": 0.00019991612684818868, + "loss": 1.289, + "step": 2459 + }, + { + "epoch": 0.17070293525778918, + "grad_norm": 4.173858642578125, + "learning_rate": 0.0001999158966179795, + "loss": 1.1872, + "step": 2460 + }, + { + "epoch": 0.17077232669488585, + "grad_norm": 4.445329189300537, + "learning_rate": 0.00019991566607234745, + "loss": 1.2468, + "step": 2461 + }, + { + "epoch": 0.1708417181319825, + "grad_norm": 4.5589599609375, + "learning_rate": 0.00019991543521129322, + "loss": 1.4636, + "step": 2462 + }, + { + "epoch": 0.17091110956907918, + "grad_norm": 4.114573001861572, + "learning_rate": 0.00019991520403481752, + "loss": 1.0047, + "step": 2463 + }, + { + "epoch": 0.17098050100617584, + "grad_norm": 3.155254602432251, + "learning_rate": 0.00019991497254292115, + "loss": 0.9687, + "step": 2464 + }, + { + "epoch": 0.1710498924432725, + "grad_norm": 3.834451913833618, + "learning_rate": 0.00019991474073560478, + "loss": 1.0279, + "step": 2465 + }, + { + "epoch": 0.17111928388036915, + "grad_norm": 3.4978830814361572, + "learning_rate": 0.00019991450861286917, + "loss": 1.0701, + "step": 2466 + }, + { + "epoch": 0.17118867531746582, + "grad_norm": 3.2768144607543945, + "learning_rate": 0.00019991427617471508, + "loss": 1.3239, + "step": 2467 + }, + { + "epoch": 0.17125806675456248, + "grad_norm": 3.9301271438598633, + "learning_rate": 0.00019991404342114318, + "loss": 1.3501, + "step": 2468 + }, + { + "epoch": 0.17132745819165915, + "grad_norm": 3.902170181274414, + "learning_rate": 0.00019991381035215424, + "loss": 1.1514, + "step": 2469 + }, + { + "epoch": 0.17139684962875582, + "grad_norm": 4.42300271987915, + "learning_rate": 0.00019991357696774902, + "loss": 1.1471, + "step": 2470 + }, + { + "epoch": 0.17146624106585248, + "grad_norm": 3.621094226837158, + "learning_rate": 0.00019991334326792818, + "loss": 1.0602, + "step": 2471 + }, + { + "epoch": 0.17153563250294915, + "grad_norm": 3.81111216545105, + "learning_rate": 0.00019991310925269255, + "loss": 1.0457, + "step": 2472 + }, + { + "epoch": 0.1716050239400458, + "grad_norm": 4.727643966674805, + "learning_rate": 0.00019991287492204283, + "loss": 1.1887, + "step": 2473 + }, + { + "epoch": 0.17167441537714245, + "grad_norm": 4.14417028427124, + "learning_rate": 0.00019991264027597973, + "loss": 1.0662, + "step": 2474 + }, + { + "epoch": 0.17174380681423912, + "grad_norm": 3.2582952976226807, + "learning_rate": 0.00019991240531450403, + "loss": 1.0422, + "step": 2475 + }, + { + "epoch": 0.17181319825133579, + "grad_norm": 4.466821670532227, + "learning_rate": 0.00019991217003761645, + "loss": 1.0129, + "step": 2476 + }, + { + "epoch": 0.17188258968843245, + "grad_norm": 4.961310863494873, + "learning_rate": 0.00019991193444531775, + "loss": 1.3051, + "step": 2477 + }, + { + "epoch": 0.17195198112552912, + "grad_norm": 8.285223960876465, + "learning_rate": 0.00019991169853760866, + "loss": 1.6929, + "step": 2478 + }, + { + "epoch": 0.17202137256262579, + "grad_norm": 4.640775203704834, + "learning_rate": 0.00019991146231448993, + "loss": 1.2087, + "step": 2479 + }, + { + "epoch": 0.17209076399972242, + "grad_norm": 3.9456191062927246, + "learning_rate": 0.00019991122577596232, + "loss": 1.1892, + "step": 2480 + }, + { + "epoch": 0.1721601554368191, + "grad_norm": 4.844111442565918, + "learning_rate": 0.00019991098892202655, + "loss": 1.0587, + "step": 2481 + }, + { + "epoch": 0.17222954687391576, + "grad_norm": 3.414106607437134, + "learning_rate": 0.00019991075175268335, + "loss": 1.044, + "step": 2482 + }, + { + "epoch": 0.17229893831101242, + "grad_norm": 3.248030185699463, + "learning_rate": 0.00019991051426793352, + "loss": 1.1319, + "step": 2483 + }, + { + "epoch": 0.1723683297481091, + "grad_norm": 3.487084150314331, + "learning_rate": 0.00019991027646777782, + "loss": 0.9602, + "step": 2484 + }, + { + "epoch": 0.17243772118520576, + "grad_norm": 6.608536243438721, + "learning_rate": 0.00019991003835221694, + "loss": 1.2703, + "step": 2485 + }, + { + "epoch": 0.1725071126223024, + "grad_norm": 3.6766109466552734, + "learning_rate": 0.00019990979992125167, + "loss": 1.1401, + "step": 2486 + }, + { + "epoch": 0.17257650405939906, + "grad_norm": 3.8838131427764893, + "learning_rate": 0.00019990956117488273, + "loss": 1.424, + "step": 2487 + }, + { + "epoch": 0.17264589549649573, + "grad_norm": 4.91753625869751, + "learning_rate": 0.00019990932211311094, + "loss": 1.1382, + "step": 2488 + }, + { + "epoch": 0.1727152869335924, + "grad_norm": 4.426657676696777, + "learning_rate": 0.00019990908273593696, + "loss": 1.078, + "step": 2489 + }, + { + "epoch": 0.17278467837068906, + "grad_norm": 4.331727027893066, + "learning_rate": 0.00019990884304336163, + "loss": 1.167, + "step": 2490 + }, + { + "epoch": 0.17285406980778573, + "grad_norm": 4.340266227722168, + "learning_rate": 0.00019990860303538565, + "loss": 1.3201, + "step": 2491 + }, + { + "epoch": 0.1729234612448824, + "grad_norm": 4.225974082946777, + "learning_rate": 0.00019990836271200984, + "loss": 1.0897, + "step": 2492 + }, + { + "epoch": 0.17299285268197903, + "grad_norm": 4.118218421936035, + "learning_rate": 0.00019990812207323487, + "loss": 1.0521, + "step": 2493 + }, + { + "epoch": 0.1730622441190757, + "grad_norm": 3.4920856952667236, + "learning_rate": 0.0001999078811190616, + "loss": 1.246, + "step": 2494 + }, + { + "epoch": 0.17313163555617236, + "grad_norm": 4.24950647354126, + "learning_rate": 0.00019990763984949073, + "loss": 1.244, + "step": 2495 + }, + { + "epoch": 0.17320102699326903, + "grad_norm": 3.99815034866333, + "learning_rate": 0.00019990739826452304, + "loss": 1.1367, + "step": 2496 + }, + { + "epoch": 0.1732704184303657, + "grad_norm": 4.630021095275879, + "learning_rate": 0.00019990715636415923, + "loss": 1.0926, + "step": 2497 + }, + { + "epoch": 0.17333980986746236, + "grad_norm": 4.649616241455078, + "learning_rate": 0.00019990691414840017, + "loss": 1.1812, + "step": 2498 + }, + { + "epoch": 0.17340920130455903, + "grad_norm": 4.1069488525390625, + "learning_rate": 0.00019990667161724654, + "loss": 1.306, + "step": 2499 + }, + { + "epoch": 0.17347859274165567, + "grad_norm": 3.8867881298065186, + "learning_rate": 0.0001999064287706992, + "loss": 1.1126, + "step": 2500 + }, + { + "epoch": 0.17354798417875233, + "grad_norm": 5.697386264801025, + "learning_rate": 0.00019990618560875877, + "loss": 1.1679, + "step": 2501 + }, + { + "epoch": 0.173617375615849, + "grad_norm": 3.6615471839904785, + "learning_rate": 0.00019990594213142618, + "loss": 1.1013, + "step": 2502 + }, + { + "epoch": 0.17368676705294567, + "grad_norm": 3.923921823501587, + "learning_rate": 0.00019990569833870208, + "loss": 1.3808, + "step": 2503 + }, + { + "epoch": 0.17375615849004233, + "grad_norm": 4.148416519165039, + "learning_rate": 0.00019990545423058728, + "loss": 1.1848, + "step": 2504 + }, + { + "epoch": 0.173825549927139, + "grad_norm": 4.248214244842529, + "learning_rate": 0.00019990520980708257, + "loss": 1.2041, + "step": 2505 + }, + { + "epoch": 0.17389494136423567, + "grad_norm": 4.175629138946533, + "learning_rate": 0.0001999049650681887, + "loss": 1.3102, + "step": 2506 + }, + { + "epoch": 0.1739643328013323, + "grad_norm": 3.451975107192993, + "learning_rate": 0.00019990472001390645, + "loss": 0.9354, + "step": 2507 + }, + { + "epoch": 0.17403372423842897, + "grad_norm": 4.824532985687256, + "learning_rate": 0.00019990447464423658, + "loss": 0.8902, + "step": 2508 + }, + { + "epoch": 0.17410311567552564, + "grad_norm": 3.730186700820923, + "learning_rate": 0.0001999042289591799, + "loss": 1.1864, + "step": 2509 + }, + { + "epoch": 0.1741725071126223, + "grad_norm": 3.8411765098571777, + "learning_rate": 0.0001999039829587371, + "loss": 0.9183, + "step": 2510 + }, + { + "epoch": 0.17424189854971897, + "grad_norm": 4.542413234710693, + "learning_rate": 0.00019990373664290908, + "loss": 1.2879, + "step": 2511 + }, + { + "epoch": 0.17431128998681564, + "grad_norm": 4.380620002746582, + "learning_rate": 0.00019990349001169653, + "loss": 1.2094, + "step": 2512 + }, + { + "epoch": 0.17438068142391228, + "grad_norm": 4.503941059112549, + "learning_rate": 0.00019990324306510025, + "loss": 0.9709, + "step": 2513 + }, + { + "epoch": 0.17445007286100894, + "grad_norm": 3.576003313064575, + "learning_rate": 0.00019990299580312102, + "loss": 0.807, + "step": 2514 + }, + { + "epoch": 0.1745194642981056, + "grad_norm": 3.6145365238189697, + "learning_rate": 0.00019990274822575963, + "loss": 1.1459, + "step": 2515 + }, + { + "epoch": 0.17458885573520228, + "grad_norm": 3.5010945796966553, + "learning_rate": 0.00019990250033301686, + "loss": 1.4347, + "step": 2516 + }, + { + "epoch": 0.17465824717229894, + "grad_norm": 4.347161769866943, + "learning_rate": 0.00019990225212489348, + "loss": 1.042, + "step": 2517 + }, + { + "epoch": 0.1747276386093956, + "grad_norm": 3.13651967048645, + "learning_rate": 0.00019990200360139028, + "loss": 0.6486, + "step": 2518 + }, + { + "epoch": 0.17479703004649227, + "grad_norm": 3.8666088581085205, + "learning_rate": 0.00019990175476250802, + "loss": 1.2071, + "step": 2519 + }, + { + "epoch": 0.1748664214835889, + "grad_norm": 4.180232524871826, + "learning_rate": 0.00019990150560824757, + "loss": 1.088, + "step": 2520 + }, + { + "epoch": 0.17493581292068558, + "grad_norm": 3.4234619140625, + "learning_rate": 0.0001999012561386096, + "loss": 1.01, + "step": 2521 + }, + { + "epoch": 0.17500520435778225, + "grad_norm": 6.519861221313477, + "learning_rate": 0.000199901006353595, + "loss": 1.0619, + "step": 2522 + }, + { + "epoch": 0.1750745957948789, + "grad_norm": 5.7333807945251465, + "learning_rate": 0.0001999007562532045, + "loss": 1.0636, + "step": 2523 + }, + { + "epoch": 0.17514398723197558, + "grad_norm": 4.961284637451172, + "learning_rate": 0.0001999005058374389, + "loss": 0.9728, + "step": 2524 + }, + { + "epoch": 0.17521337866907225, + "grad_norm": 4.102304458618164, + "learning_rate": 0.00019990025510629897, + "loss": 1.0905, + "step": 2525 + }, + { + "epoch": 0.1752827701061689, + "grad_norm": 4.902465343475342, + "learning_rate": 0.00019990000405978556, + "loss": 1.1272, + "step": 2526 + }, + { + "epoch": 0.17535216154326555, + "grad_norm": 3.8400590419769287, + "learning_rate": 0.0001998997526978994, + "loss": 0.9343, + "step": 2527 + }, + { + "epoch": 0.17542155298036222, + "grad_norm": 5.844316482543945, + "learning_rate": 0.00019989950102064133, + "loss": 1.4193, + "step": 2528 + }, + { + "epoch": 0.17549094441745888, + "grad_norm": 4.69776725769043, + "learning_rate": 0.00019989924902801212, + "loss": 0.7714, + "step": 2529 + }, + { + "epoch": 0.17556033585455555, + "grad_norm": 4.185298919677734, + "learning_rate": 0.00019989899672001257, + "loss": 1.2332, + "step": 2530 + }, + { + "epoch": 0.17562972729165222, + "grad_norm": 4.7076945304870605, + "learning_rate": 0.00019989874409664346, + "loss": 1.0616, + "step": 2531 + }, + { + "epoch": 0.17569911872874888, + "grad_norm": 3.914311408996582, + "learning_rate": 0.00019989849115790564, + "loss": 1.1382, + "step": 2532 + }, + { + "epoch": 0.17576851016584555, + "grad_norm": 4.932521343231201, + "learning_rate": 0.00019989823790379988, + "loss": 1.4344, + "step": 2533 + }, + { + "epoch": 0.1758379016029422, + "grad_norm": 4.1848859786987305, + "learning_rate": 0.00019989798433432693, + "loss": 0.9905, + "step": 2534 + }, + { + "epoch": 0.17590729304003885, + "grad_norm": 4.534704685211182, + "learning_rate": 0.00019989773044948764, + "loss": 1.0227, + "step": 2535 + }, + { + "epoch": 0.17597668447713552, + "grad_norm": 3.898334264755249, + "learning_rate": 0.00019989747624928286, + "loss": 1.0061, + "step": 2536 + }, + { + "epoch": 0.1760460759142322, + "grad_norm": 4.5949859619140625, + "learning_rate": 0.00019989722173371327, + "loss": 0.959, + "step": 2537 + }, + { + "epoch": 0.17611546735132885, + "grad_norm": 4.583086013793945, + "learning_rate": 0.0001998969669027798, + "loss": 1.0208, + "step": 2538 + }, + { + "epoch": 0.17618485878842552, + "grad_norm": 4.0839948654174805, + "learning_rate": 0.00019989671175648315, + "loss": 1.3274, + "step": 2539 + }, + { + "epoch": 0.17625425022552216, + "grad_norm": 4.508045196533203, + "learning_rate": 0.0001998964562948242, + "loss": 1.2562, + "step": 2540 + }, + { + "epoch": 0.17632364166261882, + "grad_norm": 5.018561840057373, + "learning_rate": 0.0001998962005178037, + "loss": 0.9256, + "step": 2541 + }, + { + "epoch": 0.1763930330997155, + "grad_norm": 6.497440814971924, + "learning_rate": 0.0001998959444254225, + "loss": 1.3613, + "step": 2542 + }, + { + "epoch": 0.17646242453681216, + "grad_norm": 4.352246284484863, + "learning_rate": 0.00019989568801768138, + "loss": 1.3018, + "step": 2543 + }, + { + "epoch": 0.17653181597390882, + "grad_norm": 4.778977394104004, + "learning_rate": 0.0001998954312945812, + "loss": 1.3069, + "step": 2544 + }, + { + "epoch": 0.1766012074110055, + "grad_norm": 3.39452862739563, + "learning_rate": 0.0001998951742561227, + "loss": 0.9881, + "step": 2545 + }, + { + "epoch": 0.17667059884810216, + "grad_norm": 3.6873788833618164, + "learning_rate": 0.00019989491690230673, + "loss": 1.0884, + "step": 2546 + }, + { + "epoch": 0.1767399902851988, + "grad_norm": 3.8707737922668457, + "learning_rate": 0.0001998946592331341, + "loss": 1.2609, + "step": 2547 + }, + { + "epoch": 0.17680938172229546, + "grad_norm": 4.178804874420166, + "learning_rate": 0.00019989440124860562, + "loss": 1.2146, + "step": 2548 + }, + { + "epoch": 0.17687877315939213, + "grad_norm": 4.612890720367432, + "learning_rate": 0.0001998941429487221, + "loss": 1.5024, + "step": 2549 + }, + { + "epoch": 0.1769481645964888, + "grad_norm": 3.7037103176116943, + "learning_rate": 0.00019989388433348438, + "loss": 1.1259, + "step": 2550 + }, + { + "epoch": 0.17701755603358546, + "grad_norm": 3.0700182914733887, + "learning_rate": 0.00019989362540289324, + "loss": 0.9451, + "step": 2551 + }, + { + "epoch": 0.17708694747068213, + "grad_norm": 3.713658571243286, + "learning_rate": 0.00019989336615694955, + "loss": 1.192, + "step": 2552 + }, + { + "epoch": 0.1771563389077788, + "grad_norm": 4.530342102050781, + "learning_rate": 0.00019989310659565402, + "loss": 1.0492, + "step": 2553 + }, + { + "epoch": 0.17722573034487543, + "grad_norm": 3.3558080196380615, + "learning_rate": 0.0001998928467190076, + "loss": 1.1498, + "step": 2554 + }, + { + "epoch": 0.1772951217819721, + "grad_norm": 5.144802093505859, + "learning_rate": 0.00019989258652701103, + "loss": 1.4255, + "step": 2555 + }, + { + "epoch": 0.17736451321906876, + "grad_norm": 3.3052501678466797, + "learning_rate": 0.00019989232601966517, + "loss": 0.9513, + "step": 2556 + }, + { + "epoch": 0.17743390465616543, + "grad_norm": 3.8797965049743652, + "learning_rate": 0.0001998920651969708, + "loss": 1.341, + "step": 2557 + }, + { + "epoch": 0.1775032960932621, + "grad_norm": 4.548045635223389, + "learning_rate": 0.0001998918040589288, + "loss": 1.0944, + "step": 2558 + }, + { + "epoch": 0.17757268753035876, + "grad_norm": 4.832139492034912, + "learning_rate": 0.00019989154260553995, + "loss": 1.654, + "step": 2559 + }, + { + "epoch": 0.1776420789674554, + "grad_norm": 3.69661545753479, + "learning_rate": 0.00019989128083680512, + "loss": 1.0578, + "step": 2560 + }, + { + "epoch": 0.17771147040455207, + "grad_norm": 4.571906566619873, + "learning_rate": 0.00019989101875272507, + "loss": 1.1491, + "step": 2561 + }, + { + "epoch": 0.17778086184164874, + "grad_norm": 3.885798931121826, + "learning_rate": 0.00019989075635330067, + "loss": 1.2171, + "step": 2562 + }, + { + "epoch": 0.1778502532787454, + "grad_norm": 4.208108425140381, + "learning_rate": 0.0001998904936385327, + "loss": 1.237, + "step": 2563 + }, + { + "epoch": 0.17791964471584207, + "grad_norm": 4.617264270782471, + "learning_rate": 0.0001998902306084221, + "loss": 1.2504, + "step": 2564 + }, + { + "epoch": 0.17798903615293873, + "grad_norm": 2.920095205307007, + "learning_rate": 0.0001998899672629696, + "loss": 1.0728, + "step": 2565 + }, + { + "epoch": 0.1780584275900354, + "grad_norm": 6.8767876625061035, + "learning_rate": 0.00019988970360217606, + "loss": 1.3394, + "step": 2566 + }, + { + "epoch": 0.17812781902713204, + "grad_norm": 4.529168605804443, + "learning_rate": 0.00019988943962604233, + "loss": 1.2494, + "step": 2567 + }, + { + "epoch": 0.1781972104642287, + "grad_norm": 3.4642040729522705, + "learning_rate": 0.0001998891753345692, + "loss": 1.1358, + "step": 2568 + }, + { + "epoch": 0.17826660190132537, + "grad_norm": 4.354159832000732, + "learning_rate": 0.00019988891072775757, + "loss": 1.2361, + "step": 2569 + }, + { + "epoch": 0.17833599333842204, + "grad_norm": 2.9509143829345703, + "learning_rate": 0.00019988864580560822, + "loss": 0.8358, + "step": 2570 + }, + { + "epoch": 0.1784053847755187, + "grad_norm": 2.9681499004364014, + "learning_rate": 0.00019988838056812203, + "loss": 1.1217, + "step": 2571 + }, + { + "epoch": 0.17847477621261537, + "grad_norm": 3.975602149963379, + "learning_rate": 0.00019988811501529976, + "loss": 1.3193, + "step": 2572 + }, + { + "epoch": 0.17854416764971204, + "grad_norm": 4.555499076843262, + "learning_rate": 0.00019988784914714232, + "loss": 1.1583, + "step": 2573 + }, + { + "epoch": 0.17861355908680868, + "grad_norm": 3.8232481479644775, + "learning_rate": 0.00019988758296365054, + "loss": 1.0439, + "step": 2574 + }, + { + "epoch": 0.17868295052390534, + "grad_norm": 4.179656982421875, + "learning_rate": 0.00019988731646482526, + "loss": 1.3463, + "step": 2575 + }, + { + "epoch": 0.178752341961002, + "grad_norm": 5.600642204284668, + "learning_rate": 0.00019988704965066728, + "loss": 1.5362, + "step": 2576 + }, + { + "epoch": 0.17882173339809868, + "grad_norm": 3.950573682785034, + "learning_rate": 0.00019988678252117753, + "loss": 1.0267, + "step": 2577 + }, + { + "epoch": 0.17889112483519534, + "grad_norm": 3.781283378601074, + "learning_rate": 0.00019988651507635673, + "loss": 1.2736, + "step": 2578 + }, + { + "epoch": 0.178960516272292, + "grad_norm": 3.732323408126831, + "learning_rate": 0.00019988624731620585, + "loss": 1.0626, + "step": 2579 + }, + { + "epoch": 0.17902990770938867, + "grad_norm": 4.102365493774414, + "learning_rate": 0.00019988597924072563, + "loss": 1.2793, + "step": 2580 + }, + { + "epoch": 0.1790992991464853, + "grad_norm": 4.550371170043945, + "learning_rate": 0.00019988571084991698, + "loss": 1.4509, + "step": 2581 + }, + { + "epoch": 0.17916869058358198, + "grad_norm": 4.442364692687988, + "learning_rate": 0.00019988544214378072, + "loss": 0.9797, + "step": 2582 + }, + { + "epoch": 0.17923808202067865, + "grad_norm": 5.2534050941467285, + "learning_rate": 0.0001998851731223177, + "loss": 1.1772, + "step": 2583 + }, + { + "epoch": 0.1793074734577753, + "grad_norm": 4.321468830108643, + "learning_rate": 0.0001998849037855288, + "loss": 0.9888, + "step": 2584 + }, + { + "epoch": 0.17937686489487198, + "grad_norm": 4.4779839515686035, + "learning_rate": 0.00019988463413341486, + "loss": 1.0276, + "step": 2585 + }, + { + "epoch": 0.17944625633196865, + "grad_norm": 4.365760803222656, + "learning_rate": 0.00019988436416597667, + "loss": 1.2605, + "step": 2586 + }, + { + "epoch": 0.17951564776906528, + "grad_norm": 4.98548698425293, + "learning_rate": 0.00019988409388321517, + "loss": 1.0702, + "step": 2587 + }, + { + "epoch": 0.17958503920616195, + "grad_norm": 4.144817352294922, + "learning_rate": 0.0001998838232851312, + "loss": 1.2887, + "step": 2588 + }, + { + "epoch": 0.17965443064325862, + "grad_norm": 4.756958484649658, + "learning_rate": 0.00019988355237172552, + "loss": 0.9772, + "step": 2589 + }, + { + "epoch": 0.17972382208035528, + "grad_norm": 3.592231512069702, + "learning_rate": 0.00019988328114299908, + "loss": 1.1325, + "step": 2590 + }, + { + "epoch": 0.17979321351745195, + "grad_norm": 4.726509094238281, + "learning_rate": 0.00019988300959895272, + "loss": 1.2225, + "step": 2591 + }, + { + "epoch": 0.17986260495454862, + "grad_norm": 3.6058499813079834, + "learning_rate": 0.00019988273773958727, + "loss": 0.9929, + "step": 2592 + }, + { + "epoch": 0.17993199639164528, + "grad_norm": 3.37977933883667, + "learning_rate": 0.00019988246556490365, + "loss": 1.0088, + "step": 2593 + }, + { + "epoch": 0.18000138782874192, + "grad_norm": 3.5369668006896973, + "learning_rate": 0.00019988219307490263, + "loss": 0.89, + "step": 2594 + }, + { + "epoch": 0.1800707792658386, + "grad_norm": 4.6418914794921875, + "learning_rate": 0.00019988192026958512, + "loss": 1.371, + "step": 2595 + }, + { + "epoch": 0.18014017070293525, + "grad_norm": 4.046370983123779, + "learning_rate": 0.000199881647148952, + "loss": 0.9129, + "step": 2596 + }, + { + "epoch": 0.18020956214003192, + "grad_norm": 3.570603609085083, + "learning_rate": 0.00019988137371300407, + "loss": 1.0989, + "step": 2597 + }, + { + "epoch": 0.1802789535771286, + "grad_norm": 4.031813144683838, + "learning_rate": 0.00019988109996174226, + "loss": 0.9974, + "step": 2598 + }, + { + "epoch": 0.18034834501422525, + "grad_norm": 4.853709697723389, + "learning_rate": 0.0001998808258951674, + "loss": 1.3057, + "step": 2599 + }, + { + "epoch": 0.18041773645132192, + "grad_norm": 3.7120726108551025, + "learning_rate": 0.00019988055151328035, + "loss": 1.116, + "step": 2600 + }, + { + "epoch": 0.18048712788841856, + "grad_norm": 5.28965425491333, + "learning_rate": 0.000199880276816082, + "loss": 1.1811, + "step": 2601 + }, + { + "epoch": 0.18055651932551522, + "grad_norm": 6.33475399017334, + "learning_rate": 0.00019988000180357323, + "loss": 1.3513, + "step": 2602 + }, + { + "epoch": 0.1806259107626119, + "grad_norm": 4.521208763122559, + "learning_rate": 0.00019987972647575485, + "loss": 1.0077, + "step": 2603 + }, + { + "epoch": 0.18069530219970856, + "grad_norm": 3.87496018409729, + "learning_rate": 0.00019987945083262777, + "loss": 1.0641, + "step": 2604 + }, + { + "epoch": 0.18076469363680522, + "grad_norm": 4.420419692993164, + "learning_rate": 0.00019987917487419282, + "loss": 1.5769, + "step": 2605 + }, + { + "epoch": 0.1808340850739019, + "grad_norm": 4.139669895172119, + "learning_rate": 0.00019987889860045095, + "loss": 1.1375, + "step": 2606 + }, + { + "epoch": 0.18090347651099856, + "grad_norm": 3.8384487628936768, + "learning_rate": 0.00019987862201140298, + "loss": 1.0112, + "step": 2607 + }, + { + "epoch": 0.1809728679480952, + "grad_norm": 4.392655372619629, + "learning_rate": 0.00019987834510704978, + "loss": 1.6289, + "step": 2608 + }, + { + "epoch": 0.18104225938519186, + "grad_norm": 4.409202575683594, + "learning_rate": 0.0001998780678873922, + "loss": 1.1858, + "step": 2609 + }, + { + "epoch": 0.18111165082228853, + "grad_norm": 3.4376258850097656, + "learning_rate": 0.0001998777903524312, + "loss": 1.0121, + "step": 2610 + }, + { + "epoch": 0.1811810422593852, + "grad_norm": 4.105009078979492, + "learning_rate": 0.00019987751250216757, + "loss": 0.9975, + "step": 2611 + }, + { + "epoch": 0.18125043369648186, + "grad_norm": 4.050047874450684, + "learning_rate": 0.00019987723433660223, + "loss": 0.9827, + "step": 2612 + }, + { + "epoch": 0.18131982513357853, + "grad_norm": 3.766630172729492, + "learning_rate": 0.00019987695585573604, + "loss": 0.8546, + "step": 2613 + }, + { + "epoch": 0.18138921657067517, + "grad_norm": 3.825439929962158, + "learning_rate": 0.0001998766770595699, + "loss": 0.9044, + "step": 2614 + }, + { + "epoch": 0.18145860800777183, + "grad_norm": 3.799356698989868, + "learning_rate": 0.00019987639794810468, + "loss": 1.1991, + "step": 2615 + }, + { + "epoch": 0.1815279994448685, + "grad_norm": 4.173964023590088, + "learning_rate": 0.00019987611852134127, + "loss": 1.0733, + "step": 2616 + }, + { + "epoch": 0.18159739088196516, + "grad_norm": 3.4963951110839844, + "learning_rate": 0.0001998758387792805, + "loss": 0.9906, + "step": 2617 + }, + { + "epoch": 0.18166678231906183, + "grad_norm": 4.6169514656066895, + "learning_rate": 0.00019987555872192335, + "loss": 1.1769, + "step": 2618 + }, + { + "epoch": 0.1817361737561585, + "grad_norm": 3.983372926712036, + "learning_rate": 0.0001998752783492706, + "loss": 1.1318, + "step": 2619 + }, + { + "epoch": 0.18180556519325516, + "grad_norm": 4.736612796783447, + "learning_rate": 0.00019987499766132321, + "loss": 1.3488, + "step": 2620 + }, + { + "epoch": 0.1818749566303518, + "grad_norm": 3.681417942047119, + "learning_rate": 0.00019987471665808203, + "loss": 1.1912, + "step": 2621 + }, + { + "epoch": 0.18194434806744847, + "grad_norm": 3.595543146133423, + "learning_rate": 0.00019987443533954797, + "loss": 1.1684, + "step": 2622 + }, + { + "epoch": 0.18201373950454514, + "grad_norm": 3.593111515045166, + "learning_rate": 0.0001998741537057219, + "loss": 1.0216, + "step": 2623 + }, + { + "epoch": 0.1820831309416418, + "grad_norm": 4.132025718688965, + "learning_rate": 0.0001998738717566047, + "loss": 1.1308, + "step": 2624 + }, + { + "epoch": 0.18215252237873847, + "grad_norm": 4.228890895843506, + "learning_rate": 0.00019987358949219729, + "loss": 0.8472, + "step": 2625 + }, + { + "epoch": 0.18222191381583513, + "grad_norm": 3.9460840225219727, + "learning_rate": 0.00019987330691250056, + "loss": 1.2698, + "step": 2626 + }, + { + "epoch": 0.1822913052529318, + "grad_norm": 5.3094987869262695, + "learning_rate": 0.00019987302401751537, + "loss": 1.0149, + "step": 2627 + }, + { + "epoch": 0.18236069669002844, + "grad_norm": 4.20151424407959, + "learning_rate": 0.00019987274080724264, + "loss": 0.9482, + "step": 2628 + }, + { + "epoch": 0.1824300881271251, + "grad_norm": 4.241060733795166, + "learning_rate": 0.00019987245728168322, + "loss": 1.1528, + "step": 2629 + }, + { + "epoch": 0.18249947956422177, + "grad_norm": 3.8053855895996094, + "learning_rate": 0.00019987217344083805, + "loss": 1.1813, + "step": 2630 + }, + { + "epoch": 0.18256887100131844, + "grad_norm": 4.7284159660339355, + "learning_rate": 0.00019987188928470806, + "loss": 1.1962, + "step": 2631 + }, + { + "epoch": 0.1826382624384151, + "grad_norm": 3.9987618923187256, + "learning_rate": 0.00019987160481329408, + "loss": 1.1704, + "step": 2632 + }, + { + "epoch": 0.18270765387551177, + "grad_norm": 4.253393173217773, + "learning_rate": 0.000199871320026597, + "loss": 1.19, + "step": 2633 + }, + { + "epoch": 0.1827770453126084, + "grad_norm": 3.7769999504089355, + "learning_rate": 0.0001998710349246178, + "loss": 0.8749, + "step": 2634 + }, + { + "epoch": 0.18284643674970508, + "grad_norm": 4.204223155975342, + "learning_rate": 0.0001998707495073573, + "loss": 1.1297, + "step": 2635 + }, + { + "epoch": 0.18291582818680174, + "grad_norm": 4.415786266326904, + "learning_rate": 0.00019987046377481642, + "loss": 1.1723, + "step": 2636 + }, + { + "epoch": 0.1829852196238984, + "grad_norm": 3.2554023265838623, + "learning_rate": 0.0001998701777269961, + "loss": 0.8383, + "step": 2637 + }, + { + "epoch": 0.18305461106099508, + "grad_norm": 4.745659351348877, + "learning_rate": 0.00019986989136389717, + "loss": 1.1677, + "step": 2638 + }, + { + "epoch": 0.18312400249809174, + "grad_norm": 4.932706356048584, + "learning_rate": 0.00019986960468552063, + "loss": 1.3297, + "step": 2639 + }, + { + "epoch": 0.1831933939351884, + "grad_norm": 3.6644997596740723, + "learning_rate": 0.0001998693176918673, + "loss": 1.008, + "step": 2640 + }, + { + "epoch": 0.18326278537228505, + "grad_norm": 5.216314792633057, + "learning_rate": 0.00019986903038293817, + "loss": 1.3072, + "step": 2641 + }, + { + "epoch": 0.18333217680938171, + "grad_norm": 3.8633320331573486, + "learning_rate": 0.00019986874275873407, + "loss": 0.8044, + "step": 2642 + }, + { + "epoch": 0.18340156824647838, + "grad_norm": 3.4640650749206543, + "learning_rate": 0.00019986845481925592, + "loss": 0.8237, + "step": 2643 + }, + { + "epoch": 0.18347095968357505, + "grad_norm": 4.846642017364502, + "learning_rate": 0.00019986816656450467, + "loss": 1.1036, + "step": 2644 + }, + { + "epoch": 0.1835403511206717, + "grad_norm": 3.307433605194092, + "learning_rate": 0.00019986787799448118, + "loss": 0.9906, + "step": 2645 + }, + { + "epoch": 0.18360974255776838, + "grad_norm": 4.004006385803223, + "learning_rate": 0.0001998675891091864, + "loss": 1.1522, + "step": 2646 + }, + { + "epoch": 0.18367913399486505, + "grad_norm": 3.610759735107422, + "learning_rate": 0.00019986729990862124, + "loss": 1.1483, + "step": 2647 + }, + { + "epoch": 0.18374852543196168, + "grad_norm": 5.482921123504639, + "learning_rate": 0.00019986701039278657, + "loss": 1.4347, + "step": 2648 + }, + { + "epoch": 0.18381791686905835, + "grad_norm": 4.7890825271606445, + "learning_rate": 0.00019986672056168336, + "loss": 1.0813, + "step": 2649 + }, + { + "epoch": 0.18388730830615502, + "grad_norm": 4.590893268585205, + "learning_rate": 0.00019986643041531248, + "loss": 1.1668, + "step": 2650 + }, + { + "epoch": 0.18395669974325168, + "grad_norm": 3.4171030521392822, + "learning_rate": 0.00019986613995367488, + "loss": 1.032, + "step": 2651 + }, + { + "epoch": 0.18402609118034835, + "grad_norm": 4.675971031188965, + "learning_rate": 0.00019986584917677146, + "loss": 1.1788, + "step": 2652 + }, + { + "epoch": 0.18409548261744502, + "grad_norm": 3.42244553565979, + "learning_rate": 0.00019986555808460316, + "loss": 1.0145, + "step": 2653 + }, + { + "epoch": 0.18416487405454168, + "grad_norm": 4.285312175750732, + "learning_rate": 0.00019986526667717085, + "loss": 1.514, + "step": 2654 + }, + { + "epoch": 0.18423426549163832, + "grad_norm": 4.249584674835205, + "learning_rate": 0.00019986497495447548, + "loss": 0.9475, + "step": 2655 + }, + { + "epoch": 0.184303656928735, + "grad_norm": 4.8058648109436035, + "learning_rate": 0.000199864682916518, + "loss": 1.3033, + "step": 2656 + }, + { + "epoch": 0.18437304836583165, + "grad_norm": 3.708925485610962, + "learning_rate": 0.0001998643905632993, + "loss": 1.0012, + "step": 2657 + }, + { + "epoch": 0.18444243980292832, + "grad_norm": 5.571639060974121, + "learning_rate": 0.00019986409789482027, + "loss": 1.0228, + "step": 2658 + }, + { + "epoch": 0.184511831240025, + "grad_norm": 4.490061283111572, + "learning_rate": 0.0001998638049110819, + "loss": 0.994, + "step": 2659 + }, + { + "epoch": 0.18458122267712165, + "grad_norm": 4.758593559265137, + "learning_rate": 0.00019986351161208505, + "loss": 1.174, + "step": 2660 + }, + { + "epoch": 0.1846506141142183, + "grad_norm": 4.0169806480407715, + "learning_rate": 0.00019986321799783072, + "loss": 1.2345, + "step": 2661 + }, + { + "epoch": 0.18472000555131496, + "grad_norm": 4.0466814041137695, + "learning_rate": 0.00019986292406831978, + "loss": 1.3249, + "step": 2662 + }, + { + "epoch": 0.18478939698841162, + "grad_norm": 3.8644139766693115, + "learning_rate": 0.00019986262982355318, + "loss": 1.0299, + "step": 2663 + }, + { + "epoch": 0.1848587884255083, + "grad_norm": 3.620800495147705, + "learning_rate": 0.00019986233526353184, + "loss": 1.1901, + "step": 2664 + }, + { + "epoch": 0.18492817986260496, + "grad_norm": 4.2809977531433105, + "learning_rate": 0.00019986204038825668, + "loss": 1.0971, + "step": 2665 + }, + { + "epoch": 0.18499757129970162, + "grad_norm": 3.645590305328369, + "learning_rate": 0.00019986174519772865, + "loss": 1.1147, + "step": 2666 + }, + { + "epoch": 0.1850669627367983, + "grad_norm": 3.2058420181274414, + "learning_rate": 0.0001998614496919487, + "loss": 0.9908, + "step": 2667 + }, + { + "epoch": 0.18513635417389493, + "grad_norm": 3.3067617416381836, + "learning_rate": 0.00019986115387091773, + "loss": 1.1687, + "step": 2668 + }, + { + "epoch": 0.1852057456109916, + "grad_norm": 4.177734375, + "learning_rate": 0.00019986085773463668, + "loss": 1.3266, + "step": 2669 + }, + { + "epoch": 0.18527513704808826, + "grad_norm": 3.4879190921783447, + "learning_rate": 0.00019986056128310647, + "loss": 0.9086, + "step": 2670 + }, + { + "epoch": 0.18534452848518493, + "grad_norm": 4.041145324707031, + "learning_rate": 0.00019986026451632806, + "loss": 1.2636, + "step": 2671 + }, + { + "epoch": 0.1854139199222816, + "grad_norm": 4.022595405578613, + "learning_rate": 0.0001998599674343024, + "loss": 1.1033, + "step": 2672 + }, + { + "epoch": 0.18548331135937826, + "grad_norm": 3.8977346420288086, + "learning_rate": 0.0001998596700370304, + "loss": 1.0343, + "step": 2673 + }, + { + "epoch": 0.18555270279647493, + "grad_norm": 3.6976845264434814, + "learning_rate": 0.00019985937232451303, + "loss": 0.9715, + "step": 2674 + }, + { + "epoch": 0.18562209423357157, + "grad_norm": 3.621260166168213, + "learning_rate": 0.00019985907429675116, + "loss": 1.0285, + "step": 2675 + }, + { + "epoch": 0.18569148567066823, + "grad_norm": 3.2021827697753906, + "learning_rate": 0.0001998587759537458, + "loss": 0.8306, + "step": 2676 + }, + { + "epoch": 0.1857608771077649, + "grad_norm": 3.572904348373413, + "learning_rate": 0.00019985847729549788, + "loss": 1.0652, + "step": 2677 + }, + { + "epoch": 0.18583026854486157, + "grad_norm": 6.107828617095947, + "learning_rate": 0.00019985817832200833, + "loss": 1.3391, + "step": 2678 + }, + { + "epoch": 0.18589965998195823, + "grad_norm": 4.940911769866943, + "learning_rate": 0.0001998578790332781, + "loss": 1.2468, + "step": 2679 + }, + { + "epoch": 0.1859690514190549, + "grad_norm": 3.7159512042999268, + "learning_rate": 0.00019985757942930812, + "loss": 0.8461, + "step": 2680 + }, + { + "epoch": 0.18603844285615156, + "grad_norm": 3.712710380554199, + "learning_rate": 0.00019985727951009935, + "loss": 1.0291, + "step": 2681 + }, + { + "epoch": 0.1861078342932482, + "grad_norm": 3.934901714324951, + "learning_rate": 0.00019985697927565275, + "loss": 1.1544, + "step": 2682 + }, + { + "epoch": 0.18617722573034487, + "grad_norm": 8.209358215332031, + "learning_rate": 0.0001998566787259692, + "loss": 1.1834, + "step": 2683 + }, + { + "epoch": 0.18624661716744154, + "grad_norm": 2.994887590408325, + "learning_rate": 0.00019985637786104976, + "loss": 0.9276, + "step": 2684 + }, + { + "epoch": 0.1863160086045382, + "grad_norm": 3.5886576175689697, + "learning_rate": 0.0001998560766808953, + "loss": 1.316, + "step": 2685 + }, + { + "epoch": 0.18638540004163487, + "grad_norm": 4.693149089813232, + "learning_rate": 0.0001998557751855068, + "loss": 0.9818, + "step": 2686 + }, + { + "epoch": 0.18645479147873154, + "grad_norm": 3.6397695541381836, + "learning_rate": 0.00019985547337488517, + "loss": 1.0681, + "step": 2687 + }, + { + "epoch": 0.18652418291582817, + "grad_norm": 5.155881881713867, + "learning_rate": 0.00019985517124903142, + "loss": 1.1315, + "step": 2688 + }, + { + "epoch": 0.18659357435292484, + "grad_norm": 5.346250057220459, + "learning_rate": 0.00019985486880794647, + "loss": 0.9998, + "step": 2689 + }, + { + "epoch": 0.1866629657900215, + "grad_norm": 5.792126655578613, + "learning_rate": 0.0001998545660516313, + "loss": 1.1202, + "step": 2690 + }, + { + "epoch": 0.18673235722711817, + "grad_norm": 6.079196453094482, + "learning_rate": 0.00019985426298008686, + "loss": 1.2594, + "step": 2691 + }, + { + "epoch": 0.18680174866421484, + "grad_norm": 3.690753221511841, + "learning_rate": 0.00019985395959331407, + "loss": 0.9728, + "step": 2692 + }, + { + "epoch": 0.1868711401013115, + "grad_norm": 4.908682823181152, + "learning_rate": 0.00019985365589131392, + "loss": 1.2727, + "step": 2693 + }, + { + "epoch": 0.18694053153840817, + "grad_norm": 4.421628475189209, + "learning_rate": 0.00019985335187408735, + "loss": 1.1511, + "step": 2694 + }, + { + "epoch": 0.1870099229755048, + "grad_norm": 4.081517696380615, + "learning_rate": 0.00019985304754163537, + "loss": 1.0562, + "step": 2695 + }, + { + "epoch": 0.18707931441260148, + "grad_norm": 5.31699275970459, + "learning_rate": 0.0001998527428939589, + "loss": 1.0895, + "step": 2696 + }, + { + "epoch": 0.18714870584969814, + "grad_norm": 6.459442615509033, + "learning_rate": 0.00019985243793105889, + "loss": 1.1733, + "step": 2697 + }, + { + "epoch": 0.1872180972867948, + "grad_norm": 5.58380126953125, + "learning_rate": 0.0001998521326529363, + "loss": 1.1592, + "step": 2698 + }, + { + "epoch": 0.18728748872389148, + "grad_norm": 4.924901485443115, + "learning_rate": 0.00019985182705959217, + "loss": 1.1401, + "step": 2699 + }, + { + "epoch": 0.18735688016098814, + "grad_norm": 5.518888473510742, + "learning_rate": 0.00019985152115102736, + "loss": 0.9471, + "step": 2700 + }, + { + "epoch": 0.1874262715980848, + "grad_norm": 3.3189258575439453, + "learning_rate": 0.00019985121492724288, + "loss": 1.0011, + "step": 2701 + }, + { + "epoch": 0.18749566303518145, + "grad_norm": 3.4615321159362793, + "learning_rate": 0.0001998509083882397, + "loss": 1.0398, + "step": 2702 + }, + { + "epoch": 0.18756505447227811, + "grad_norm": 3.928008794784546, + "learning_rate": 0.0001998506015340188, + "loss": 0.9524, + "step": 2703 + }, + { + "epoch": 0.18763444590937478, + "grad_norm": 4.522210121154785, + "learning_rate": 0.00019985029436458114, + "loss": 1.3688, + "step": 2704 + }, + { + "epoch": 0.18770383734647145, + "grad_norm": 4.446191310882568, + "learning_rate": 0.00019984998687992768, + "loss": 1.171, + "step": 2705 + }, + { + "epoch": 0.1877732287835681, + "grad_norm": 8.151577949523926, + "learning_rate": 0.00019984967908005937, + "loss": 1.3341, + "step": 2706 + }, + { + "epoch": 0.18784262022066478, + "grad_norm": 4.424386024475098, + "learning_rate": 0.00019984937096497726, + "loss": 1.2389, + "step": 2707 + }, + { + "epoch": 0.18791201165776142, + "grad_norm": 3.6915323734283447, + "learning_rate": 0.00019984906253468223, + "loss": 1.112, + "step": 2708 + }, + { + "epoch": 0.18798140309485808, + "grad_norm": 4.812434673309326, + "learning_rate": 0.00019984875378917534, + "loss": 1.1687, + "step": 2709 + }, + { + "epoch": 0.18805079453195475, + "grad_norm": 4.28702449798584, + "learning_rate": 0.00019984844472845749, + "loss": 1.0864, + "step": 2710 + }, + { + "epoch": 0.18812018596905142, + "grad_norm": 3.7253689765930176, + "learning_rate": 0.00019984813535252967, + "loss": 1.2026, + "step": 2711 + }, + { + "epoch": 0.18818957740614808, + "grad_norm": 5.502496719360352, + "learning_rate": 0.0001998478256613929, + "loss": 1.3465, + "step": 2712 + }, + { + "epoch": 0.18825896884324475, + "grad_norm": 3.1544344425201416, + "learning_rate": 0.00019984751565504808, + "loss": 0.9836, + "step": 2713 + }, + { + "epoch": 0.18832836028034142, + "grad_norm": 4.6087870597839355, + "learning_rate": 0.00019984720533349629, + "loss": 1.0884, + "step": 2714 + }, + { + "epoch": 0.18839775171743806, + "grad_norm": 3.31532621383667, + "learning_rate": 0.0001998468946967384, + "loss": 0.9743, + "step": 2715 + }, + { + "epoch": 0.18846714315453472, + "grad_norm": 3.5845887660980225, + "learning_rate": 0.00019984658374477552, + "loss": 0.8104, + "step": 2716 + }, + { + "epoch": 0.1885365345916314, + "grad_norm": 5.215023517608643, + "learning_rate": 0.0001998462724776085, + "loss": 1.1691, + "step": 2717 + }, + { + "epoch": 0.18860592602872805, + "grad_norm": 6.435008525848389, + "learning_rate": 0.0001998459608952384, + "loss": 1.3889, + "step": 2718 + }, + { + "epoch": 0.18867531746582472, + "grad_norm": 4.278061389923096, + "learning_rate": 0.0001998456489976662, + "loss": 1.1123, + "step": 2719 + }, + { + "epoch": 0.1887447089029214, + "grad_norm": 6.334469318389893, + "learning_rate": 0.00019984533678489287, + "loss": 0.8988, + "step": 2720 + }, + { + "epoch": 0.18881410034001805, + "grad_norm": 5.429837703704834, + "learning_rate": 0.00019984502425691937, + "loss": 1.208, + "step": 2721 + }, + { + "epoch": 0.1888834917771147, + "grad_norm": 5.136967658996582, + "learning_rate": 0.00019984471141374675, + "loss": 1.0315, + "step": 2722 + }, + { + "epoch": 0.18895288321421136, + "grad_norm": 3.740553617477417, + "learning_rate": 0.00019984439825537594, + "loss": 0.9395, + "step": 2723 + }, + { + "epoch": 0.18902227465130803, + "grad_norm": 3.6439409255981445, + "learning_rate": 0.00019984408478180795, + "loss": 1.3374, + "step": 2724 + }, + { + "epoch": 0.1890916660884047, + "grad_norm": 3.5784354209899902, + "learning_rate": 0.00019984377099304377, + "loss": 0.9634, + "step": 2725 + }, + { + "epoch": 0.18916105752550136, + "grad_norm": 7.55778694152832, + "learning_rate": 0.00019984345688908438, + "loss": 1.1304, + "step": 2726 + }, + { + "epoch": 0.18923044896259802, + "grad_norm": 4.28811502456665, + "learning_rate": 0.0001998431424699308, + "loss": 1.1456, + "step": 2727 + }, + { + "epoch": 0.1892998403996947, + "grad_norm": 4.801279544830322, + "learning_rate": 0.00019984282773558397, + "loss": 1.2019, + "step": 2728 + }, + { + "epoch": 0.18936923183679133, + "grad_norm": 4.580173492431641, + "learning_rate": 0.00019984251268604495, + "loss": 1.0649, + "step": 2729 + }, + { + "epoch": 0.189438623273888, + "grad_norm": 4.023212432861328, + "learning_rate": 0.00019984219732131469, + "loss": 1.0272, + "step": 2730 + }, + { + "epoch": 0.18950801471098466, + "grad_norm": 4.664628982543945, + "learning_rate": 0.0001998418816413942, + "loss": 0.9587, + "step": 2731 + }, + { + "epoch": 0.18957740614808133, + "grad_norm": 4.258840560913086, + "learning_rate": 0.00019984156564628448, + "loss": 1.1316, + "step": 2732 + }, + { + "epoch": 0.189646797585178, + "grad_norm": 6.18887996673584, + "learning_rate": 0.00019984124933598648, + "loss": 1.1061, + "step": 2733 + }, + { + "epoch": 0.18971618902227466, + "grad_norm": 3.6990203857421875, + "learning_rate": 0.0001998409327105013, + "loss": 0.9536, + "step": 2734 + }, + { + "epoch": 0.1897855804593713, + "grad_norm": 6.186032295227051, + "learning_rate": 0.00019984061576982983, + "loss": 1.3438, + "step": 2735 + }, + { + "epoch": 0.18985497189646797, + "grad_norm": 4.27781343460083, + "learning_rate": 0.00019984029851397313, + "loss": 1.0556, + "step": 2736 + }, + { + "epoch": 0.18992436333356463, + "grad_norm": 4.652002334594727, + "learning_rate": 0.00019983998094293218, + "loss": 1.149, + "step": 2737 + }, + { + "epoch": 0.1899937547706613, + "grad_norm": 4.073798656463623, + "learning_rate": 0.00019983966305670806, + "loss": 1.0859, + "step": 2738 + }, + { + "epoch": 0.19006314620775797, + "grad_norm": 5.508663654327393, + "learning_rate": 0.00019983934485530162, + "loss": 1.2248, + "step": 2739 + }, + { + "epoch": 0.19013253764485463, + "grad_norm": 4.156840801239014, + "learning_rate": 0.000199839026338714, + "loss": 0.8214, + "step": 2740 + }, + { + "epoch": 0.1902019290819513, + "grad_norm": 3.208096742630005, + "learning_rate": 0.00019983870750694616, + "loss": 0.7078, + "step": 2741 + }, + { + "epoch": 0.19027132051904794, + "grad_norm": 3.8092284202575684, + "learning_rate": 0.00019983838835999911, + "loss": 1.0068, + "step": 2742 + }, + { + "epoch": 0.1903407119561446, + "grad_norm": 5.595735549926758, + "learning_rate": 0.00019983806889787382, + "loss": 1.225, + "step": 2743 + }, + { + "epoch": 0.19041010339324127, + "grad_norm": 3.9195563793182373, + "learning_rate": 0.00019983774912057137, + "loss": 1.2162, + "step": 2744 + }, + { + "epoch": 0.19047949483033794, + "grad_norm": 4.122901916503906, + "learning_rate": 0.0001998374290280927, + "loss": 1.0085, + "step": 2745 + }, + { + "epoch": 0.1905488862674346, + "grad_norm": 5.507256984710693, + "learning_rate": 0.00019983710862043883, + "loss": 1.3242, + "step": 2746 + }, + { + "epoch": 0.19061827770453127, + "grad_norm": 4.2197265625, + "learning_rate": 0.00019983678789761082, + "loss": 1.1628, + "step": 2747 + }, + { + "epoch": 0.19068766914162794, + "grad_norm": 5.0944085121154785, + "learning_rate": 0.00019983646685960965, + "loss": 1.0837, + "step": 2748 + }, + { + "epoch": 0.19075706057872457, + "grad_norm": 4.5272321701049805, + "learning_rate": 0.00019983614550643633, + "loss": 1.3375, + "step": 2749 + }, + { + "epoch": 0.19082645201582124, + "grad_norm": 5.273068904876709, + "learning_rate": 0.0001998358238380919, + "loss": 1.4196, + "step": 2750 + }, + { + "epoch": 0.1908958434529179, + "grad_norm": 5.5787506103515625, + "learning_rate": 0.00019983550185457736, + "loss": 1.106, + "step": 2751 + }, + { + "epoch": 0.19096523489001457, + "grad_norm": 4.120942115783691, + "learning_rate": 0.0001998351795558937, + "loss": 1.1981, + "step": 2752 + }, + { + "epoch": 0.19103462632711124, + "grad_norm": 4.269922256469727, + "learning_rate": 0.00019983485694204196, + "loss": 1.3168, + "step": 2753 + }, + { + "epoch": 0.1911040177642079, + "grad_norm": 5.444863796234131, + "learning_rate": 0.00019983453401302317, + "loss": 1.2519, + "step": 2754 + }, + { + "epoch": 0.19117340920130457, + "grad_norm": 4.378446578979492, + "learning_rate": 0.00019983421076883833, + "loss": 1.1781, + "step": 2755 + }, + { + "epoch": 0.1912428006384012, + "grad_norm": 5.565683841705322, + "learning_rate": 0.00019983388720948848, + "loss": 1.1453, + "step": 2756 + }, + { + "epoch": 0.19131219207549788, + "grad_norm": 3.3058416843414307, + "learning_rate": 0.00019983356333497465, + "loss": 0.9208, + "step": 2757 + }, + { + "epoch": 0.19138158351259454, + "grad_norm": 4.574859619140625, + "learning_rate": 0.0001998332391452978, + "loss": 1.33, + "step": 2758 + }, + { + "epoch": 0.1914509749496912, + "grad_norm": 9.008939743041992, + "learning_rate": 0.00019983291464045904, + "loss": 1.0656, + "step": 2759 + }, + { + "epoch": 0.19152036638678788, + "grad_norm": 3.527397871017456, + "learning_rate": 0.00019983258982045934, + "loss": 1.0318, + "step": 2760 + }, + { + "epoch": 0.19158975782388454, + "grad_norm": 4.723813056945801, + "learning_rate": 0.00019983226468529969, + "loss": 1.2295, + "step": 2761 + }, + { + "epoch": 0.19165914926098118, + "grad_norm": 4.02273416519165, + "learning_rate": 0.00019983193923498122, + "loss": 1.089, + "step": 2762 + }, + { + "epoch": 0.19172854069807785, + "grad_norm": 5.104701042175293, + "learning_rate": 0.0001998316134695049, + "loss": 1.4076, + "step": 2763 + }, + { + "epoch": 0.19179793213517451, + "grad_norm": 5.279738426208496, + "learning_rate": 0.00019983128738887174, + "loss": 1.0687, + "step": 2764 + }, + { + "epoch": 0.19186732357227118, + "grad_norm": 3.628584623336792, + "learning_rate": 0.00019983096099308276, + "loss": 1.2307, + "step": 2765 + }, + { + "epoch": 0.19193671500936785, + "grad_norm": 5.288729667663574, + "learning_rate": 0.00019983063428213903, + "loss": 1.3559, + "step": 2766 + }, + { + "epoch": 0.19200610644646451, + "grad_norm": 4.251145362854004, + "learning_rate": 0.00019983030725604162, + "loss": 1.2461, + "step": 2767 + }, + { + "epoch": 0.19207549788356118, + "grad_norm": 3.399825096130371, + "learning_rate": 0.00019982997991479148, + "loss": 0.9661, + "step": 2768 + }, + { + "epoch": 0.19214488932065782, + "grad_norm": 3.5539283752441406, + "learning_rate": 0.00019982965225838966, + "loss": 1.0899, + "step": 2769 + }, + { + "epoch": 0.19221428075775449, + "grad_norm": 4.131640911102295, + "learning_rate": 0.00019982932428683724, + "loss": 1.1614, + "step": 2770 + }, + { + "epoch": 0.19228367219485115, + "grad_norm": 3.2930986881256104, + "learning_rate": 0.00019982899600013522, + "loss": 1.1308, + "step": 2771 + }, + { + "epoch": 0.19235306363194782, + "grad_norm": 3.1224162578582764, + "learning_rate": 0.00019982866739828463, + "loss": 0.9829, + "step": 2772 + }, + { + "epoch": 0.19242245506904448, + "grad_norm": 3.6938374042510986, + "learning_rate": 0.0001998283384812865, + "loss": 1.252, + "step": 2773 + }, + { + "epoch": 0.19249184650614115, + "grad_norm": 3.446056365966797, + "learning_rate": 0.00019982800924914194, + "loss": 1.0922, + "step": 2774 + }, + { + "epoch": 0.19256123794323782, + "grad_norm": 3.7042908668518066, + "learning_rate": 0.0001998276797018519, + "loss": 1.0463, + "step": 2775 + }, + { + "epoch": 0.19263062938033446, + "grad_norm": 3.9580397605895996, + "learning_rate": 0.00019982734983941747, + "loss": 1.1075, + "step": 2776 + }, + { + "epoch": 0.19270002081743112, + "grad_norm": 4.750606536865234, + "learning_rate": 0.00019982701966183964, + "loss": 1.0905, + "step": 2777 + }, + { + "epoch": 0.1927694122545278, + "grad_norm": 4.088352203369141, + "learning_rate": 0.00019982668916911954, + "loss": 1.1071, + "step": 2778 + }, + { + "epoch": 0.19283880369162446, + "grad_norm": 7.897123336791992, + "learning_rate": 0.00019982635836125816, + "loss": 1.181, + "step": 2779 + }, + { + "epoch": 0.19290819512872112, + "grad_norm": 3.8607420921325684, + "learning_rate": 0.00019982602723825653, + "loss": 1.121, + "step": 2780 + }, + { + "epoch": 0.1929775865658178, + "grad_norm": 3.976132392883301, + "learning_rate": 0.00019982569580011573, + "loss": 1.3815, + "step": 2781 + }, + { + "epoch": 0.19304697800291443, + "grad_norm": 4.189416885375977, + "learning_rate": 0.00019982536404683678, + "loss": 1.3731, + "step": 2782 + }, + { + "epoch": 0.1931163694400111, + "grad_norm": 3.24933123588562, + "learning_rate": 0.00019982503197842072, + "loss": 1.22, + "step": 2783 + }, + { + "epoch": 0.19318576087710776, + "grad_norm": 3.8113930225372314, + "learning_rate": 0.00019982469959486864, + "loss": 1.0303, + "step": 2784 + }, + { + "epoch": 0.19325515231420443, + "grad_norm": 8.846112251281738, + "learning_rate": 0.00019982436689618157, + "loss": 1.1226, + "step": 2785 + }, + { + "epoch": 0.1933245437513011, + "grad_norm": 4.755606174468994, + "learning_rate": 0.00019982403388236052, + "loss": 1.3169, + "step": 2786 + }, + { + "epoch": 0.19339393518839776, + "grad_norm": 7.629700660705566, + "learning_rate": 0.00019982370055340661, + "loss": 1.08, + "step": 2787 + }, + { + "epoch": 0.19346332662549443, + "grad_norm": 3.916562795639038, + "learning_rate": 0.00019982336690932087, + "loss": 1.0044, + "step": 2788 + }, + { + "epoch": 0.19353271806259106, + "grad_norm": 5.04274845123291, + "learning_rate": 0.00019982303295010433, + "loss": 1.2023, + "step": 2789 + }, + { + "epoch": 0.19360210949968773, + "grad_norm": 10.242226600646973, + "learning_rate": 0.00019982269867575804, + "loss": 1.2206, + "step": 2790 + }, + { + "epoch": 0.1936715009367844, + "grad_norm": 10.681792259216309, + "learning_rate": 0.00019982236408628307, + "loss": 1.0368, + "step": 2791 + }, + { + "epoch": 0.19374089237388106, + "grad_norm": 4.844213008880615, + "learning_rate": 0.00019982202918168048, + "loss": 1.2246, + "step": 2792 + }, + { + "epoch": 0.19381028381097773, + "grad_norm": 4.380436897277832, + "learning_rate": 0.00019982169396195136, + "loss": 0.9388, + "step": 2793 + }, + { + "epoch": 0.1938796752480744, + "grad_norm": 3.3295493125915527, + "learning_rate": 0.00019982135842709672, + "loss": 1.0605, + "step": 2794 + }, + { + "epoch": 0.19394906668517106, + "grad_norm": 4.089224338531494, + "learning_rate": 0.0001998210225771176, + "loss": 1.0299, + "step": 2795 + }, + { + "epoch": 0.1940184581222677, + "grad_norm": 5.579251289367676, + "learning_rate": 0.0001998206864120151, + "loss": 1.3566, + "step": 2796 + }, + { + "epoch": 0.19408784955936437, + "grad_norm": 6.386904716491699, + "learning_rate": 0.0001998203499317903, + "loss": 1.3871, + "step": 2797 + }, + { + "epoch": 0.19415724099646103, + "grad_norm": 5.191259860992432, + "learning_rate": 0.00019982001313644424, + "loss": 1.3825, + "step": 2798 + }, + { + "epoch": 0.1942266324335577, + "grad_norm": 6.63644552230835, + "learning_rate": 0.000199819676025978, + "loss": 1.6882, + "step": 2799 + }, + { + "epoch": 0.19429602387065437, + "grad_norm": 4.354030609130859, + "learning_rate": 0.00019981933860039255, + "loss": 1.0834, + "step": 2800 + }, + { + "epoch": 0.19436541530775103, + "grad_norm": 4.39838171005249, + "learning_rate": 0.00019981900085968906, + "loss": 1.3034, + "step": 2801 + }, + { + "epoch": 0.1944348067448477, + "grad_norm": 4.352022171020508, + "learning_rate": 0.00019981866280386858, + "loss": 1.2715, + "step": 2802 + }, + { + "epoch": 0.19450419818194434, + "grad_norm": 3.7791783809661865, + "learning_rate": 0.00019981832443293217, + "loss": 1.2603, + "step": 2803 + }, + { + "epoch": 0.194573589619041, + "grad_norm": 3.4381253719329834, + "learning_rate": 0.00019981798574688085, + "loss": 0.8641, + "step": 2804 + }, + { + "epoch": 0.19464298105613767, + "grad_norm": 3.8844521045684814, + "learning_rate": 0.0001998176467457158, + "loss": 1.1683, + "step": 2805 + }, + { + "epoch": 0.19471237249323434, + "grad_norm": 4.488562107086182, + "learning_rate": 0.00019981730742943797, + "loss": 1.2923, + "step": 2806 + }, + { + "epoch": 0.194781763930331, + "grad_norm": 3.891145944595337, + "learning_rate": 0.00019981696779804845, + "loss": 1.2319, + "step": 2807 + }, + { + "epoch": 0.19485115536742767, + "grad_norm": 3.4516124725341797, + "learning_rate": 0.0001998166278515484, + "loss": 1.1697, + "step": 2808 + }, + { + "epoch": 0.1949205468045243, + "grad_norm": 3.049957036972046, + "learning_rate": 0.00019981628758993883, + "loss": 1.1343, + "step": 2809 + }, + { + "epoch": 0.19498993824162097, + "grad_norm": 5.135621547698975, + "learning_rate": 0.00019981594701322082, + "loss": 1.1126, + "step": 2810 + }, + { + "epoch": 0.19505932967871764, + "grad_norm": 3.624445676803589, + "learning_rate": 0.00019981560612139542, + "loss": 0.8767, + "step": 2811 + }, + { + "epoch": 0.1951287211158143, + "grad_norm": 4.464219093322754, + "learning_rate": 0.00019981526491446374, + "loss": 0.9138, + "step": 2812 + }, + { + "epoch": 0.19519811255291097, + "grad_norm": 3.7488441467285156, + "learning_rate": 0.00019981492339242687, + "loss": 1.3247, + "step": 2813 + }, + { + "epoch": 0.19526750399000764, + "grad_norm": 5.6115312576293945, + "learning_rate": 0.00019981458155528587, + "loss": 1.1407, + "step": 2814 + }, + { + "epoch": 0.1953368954271043, + "grad_norm": 3.5831458568573, + "learning_rate": 0.00019981423940304178, + "loss": 1.1434, + "step": 2815 + }, + { + "epoch": 0.19540628686420095, + "grad_norm": 2.840263843536377, + "learning_rate": 0.00019981389693569574, + "loss": 0.7553, + "step": 2816 + }, + { + "epoch": 0.1954756783012976, + "grad_norm": 5.688775062561035, + "learning_rate": 0.00019981355415324883, + "loss": 1.1729, + "step": 2817 + }, + { + "epoch": 0.19554506973839428, + "grad_norm": 4.751173973083496, + "learning_rate": 0.00019981321105570207, + "loss": 1.233, + "step": 2818 + }, + { + "epoch": 0.19561446117549094, + "grad_norm": 4.665483474731445, + "learning_rate": 0.00019981286764305663, + "loss": 1.0486, + "step": 2819 + }, + { + "epoch": 0.1956838526125876, + "grad_norm": 13.256823539733887, + "learning_rate": 0.0001998125239153135, + "loss": 1.4117, + "step": 2820 + }, + { + "epoch": 0.19575324404968428, + "grad_norm": 4.0172271728515625, + "learning_rate": 0.00019981217987247383, + "loss": 0.9796, + "step": 2821 + }, + { + "epoch": 0.19582263548678094, + "grad_norm": 4.306264400482178, + "learning_rate": 0.00019981183551453867, + "loss": 1.0978, + "step": 2822 + }, + { + "epoch": 0.19589202692387758, + "grad_norm": 5.196447849273682, + "learning_rate": 0.00019981149084150916, + "loss": 1.1593, + "step": 2823 + }, + { + "epoch": 0.19596141836097425, + "grad_norm": 4.376152992248535, + "learning_rate": 0.00019981114585338635, + "loss": 1.3084, + "step": 2824 + }, + { + "epoch": 0.19603080979807092, + "grad_norm": 4.133366584777832, + "learning_rate": 0.00019981080055017132, + "loss": 0.9477, + "step": 2825 + }, + { + "epoch": 0.19610020123516758, + "grad_norm": 4.159884929656982, + "learning_rate": 0.00019981045493186516, + "loss": 0.94, + "step": 2826 + }, + { + "epoch": 0.19616959267226425, + "grad_norm": 4.055069923400879, + "learning_rate": 0.000199810108998469, + "loss": 1.1314, + "step": 2827 + }, + { + "epoch": 0.19623898410936091, + "grad_norm": 3.934678077697754, + "learning_rate": 0.0001998097627499839, + "loss": 1.0532, + "step": 2828 + }, + { + "epoch": 0.19630837554645758, + "grad_norm": 4.051695346832275, + "learning_rate": 0.00019980941618641095, + "loss": 1.3386, + "step": 2829 + }, + { + "epoch": 0.19637776698355422, + "grad_norm": 4.617209434509277, + "learning_rate": 0.00019980906930775123, + "loss": 1.2329, + "step": 2830 + }, + { + "epoch": 0.19644715842065089, + "grad_norm": 4.7219743728637695, + "learning_rate": 0.0001998087221140059, + "loss": 1.2336, + "step": 2831 + }, + { + "epoch": 0.19651654985774755, + "grad_norm": 4.143470287322998, + "learning_rate": 0.00019980837460517598, + "loss": 1.3466, + "step": 2832 + }, + { + "epoch": 0.19658594129484422, + "grad_norm": 7.388866424560547, + "learning_rate": 0.00019980802678126261, + "loss": 1.0068, + "step": 2833 + }, + { + "epoch": 0.19665533273194089, + "grad_norm": 4.2708740234375, + "learning_rate": 0.0001998076786422669, + "loss": 1.5514, + "step": 2834 + }, + { + "epoch": 0.19672472416903755, + "grad_norm": 5.597128868103027, + "learning_rate": 0.00019980733018818995, + "loss": 1.3293, + "step": 2835 + }, + { + "epoch": 0.1967941156061342, + "grad_norm": 7.271239757537842, + "learning_rate": 0.00019980698141903278, + "loss": 1.2731, + "step": 2836 + }, + { + "epoch": 0.19686350704323086, + "grad_norm": 3.825519323348999, + "learning_rate": 0.0001998066323347966, + "loss": 0.9175, + "step": 2837 + }, + { + "epoch": 0.19693289848032752, + "grad_norm": 5.396665096282959, + "learning_rate": 0.0001998062829354824, + "loss": 1.1694, + "step": 2838 + }, + { + "epoch": 0.1970022899174242, + "grad_norm": 4.243141174316406, + "learning_rate": 0.0001998059332210914, + "loss": 0.9725, + "step": 2839 + }, + { + "epoch": 0.19707168135452086, + "grad_norm": 4.650017738342285, + "learning_rate": 0.0001998055831916246, + "loss": 1.1919, + "step": 2840 + }, + { + "epoch": 0.19714107279161752, + "grad_norm": 4.884612560272217, + "learning_rate": 0.00019980523284708322, + "loss": 1.1637, + "step": 2841 + }, + { + "epoch": 0.1972104642287142, + "grad_norm": 5.308966159820557, + "learning_rate": 0.00019980488218746824, + "loss": 1.1079, + "step": 2842 + }, + { + "epoch": 0.19727985566581083, + "grad_norm": 4.308828830718994, + "learning_rate": 0.00019980453121278086, + "loss": 1.2287, + "step": 2843 + }, + { + "epoch": 0.1973492471029075, + "grad_norm": 3.8677313327789307, + "learning_rate": 0.00019980417992302213, + "loss": 0.93, + "step": 2844 + }, + { + "epoch": 0.19741863854000416, + "grad_norm": 19.382593154907227, + "learning_rate": 0.0001998038283181932, + "loss": 1.0578, + "step": 2845 + }, + { + "epoch": 0.19748802997710083, + "grad_norm": 4.5245747566223145, + "learning_rate": 0.0001998034763982952, + "loss": 1.0871, + "step": 2846 + }, + { + "epoch": 0.1975574214141975, + "grad_norm": 3.267183542251587, + "learning_rate": 0.00019980312416332914, + "loss": 1.0799, + "step": 2847 + }, + { + "epoch": 0.19762681285129416, + "grad_norm": 3.9785315990448, + "learning_rate": 0.00019980277161329622, + "loss": 1.1464, + "step": 2848 + }, + { + "epoch": 0.19769620428839083, + "grad_norm": 3.5799052715301514, + "learning_rate": 0.00019980241874819753, + "loss": 0.9675, + "step": 2849 + }, + { + "epoch": 0.19776559572548746, + "grad_norm": 3.7747082710266113, + "learning_rate": 0.0001998020655680342, + "loss": 1.3337, + "step": 2850 + }, + { + "epoch": 0.19783498716258413, + "grad_norm": 5.32577657699585, + "learning_rate": 0.0001998017120728073, + "loss": 1.1536, + "step": 2851 + }, + { + "epoch": 0.1979043785996808, + "grad_norm": 4.769526481628418, + "learning_rate": 0.00019980135826251798, + "loss": 0.8297, + "step": 2852 + }, + { + "epoch": 0.19797377003677746, + "grad_norm": 3.6810147762298584, + "learning_rate": 0.00019980100413716735, + "loss": 0.9511, + "step": 2853 + }, + { + "epoch": 0.19804316147387413, + "grad_norm": 3.936094045639038, + "learning_rate": 0.00019980064969675655, + "loss": 1.3057, + "step": 2854 + }, + { + "epoch": 0.1981125529109708, + "grad_norm": 3.617790699005127, + "learning_rate": 0.00019980029494128668, + "loss": 0.9541, + "step": 2855 + }, + { + "epoch": 0.19818194434806743, + "grad_norm": 5.485921382904053, + "learning_rate": 0.0001997999398707588, + "loss": 1.4152, + "step": 2856 + }, + { + "epoch": 0.1982513357851641, + "grad_norm": 3.6947617530822754, + "learning_rate": 0.00019979958448517414, + "loss": 1.0875, + "step": 2857 + }, + { + "epoch": 0.19832072722226077, + "grad_norm": 4.22296142578125, + "learning_rate": 0.00019979922878453376, + "loss": 0.9264, + "step": 2858 + }, + { + "epoch": 0.19839011865935743, + "grad_norm": 4.859781742095947, + "learning_rate": 0.0001997988727688388, + "loss": 1.1725, + "step": 2859 + }, + { + "epoch": 0.1984595100964541, + "grad_norm": 4.572719573974609, + "learning_rate": 0.00019979851643809037, + "loss": 1.448, + "step": 2860 + }, + { + "epoch": 0.19852890153355077, + "grad_norm": 3.6442112922668457, + "learning_rate": 0.0001997981597922896, + "loss": 1.1365, + "step": 2861 + }, + { + "epoch": 0.19859829297064743, + "grad_norm": 6.426149368286133, + "learning_rate": 0.00019979780283143763, + "loss": 0.9794, + "step": 2862 + }, + { + "epoch": 0.19866768440774407, + "grad_norm": 3.860948324203491, + "learning_rate": 0.00019979744555553557, + "loss": 1.0529, + "step": 2863 + }, + { + "epoch": 0.19873707584484074, + "grad_norm": 6.112753868103027, + "learning_rate": 0.00019979708796458455, + "loss": 0.9347, + "step": 2864 + }, + { + "epoch": 0.1988064672819374, + "grad_norm": 4.283918857574463, + "learning_rate": 0.0001997967300585857, + "loss": 1.4487, + "step": 2865 + }, + { + "epoch": 0.19887585871903407, + "grad_norm": 5.501506328582764, + "learning_rate": 0.00019979637183754014, + "loss": 1.1637, + "step": 2866 + }, + { + "epoch": 0.19894525015613074, + "grad_norm": 3.5549893379211426, + "learning_rate": 0.00019979601330144905, + "loss": 0.9226, + "step": 2867 + }, + { + "epoch": 0.1990146415932274, + "grad_norm": 6.171341419219971, + "learning_rate": 0.0001997956544503135, + "loss": 1.0073, + "step": 2868 + }, + { + "epoch": 0.19908403303032407, + "grad_norm": 4.0019354820251465, + "learning_rate": 0.00019979529528413466, + "loss": 0.9408, + "step": 2869 + }, + { + "epoch": 0.1991534244674207, + "grad_norm": 3.9924049377441406, + "learning_rate": 0.0001997949358029136, + "loss": 1.2006, + "step": 2870 + }, + { + "epoch": 0.19922281590451738, + "grad_norm": 4.265446662902832, + "learning_rate": 0.00019979457600665157, + "loss": 1.3098, + "step": 2871 + }, + { + "epoch": 0.19929220734161404, + "grad_norm": 6.227621555328369, + "learning_rate": 0.0001997942158953496, + "loss": 1.0166, + "step": 2872 + }, + { + "epoch": 0.1993615987787107, + "grad_norm": 4.149374961853027, + "learning_rate": 0.0001997938554690089, + "loss": 0.9126, + "step": 2873 + }, + { + "epoch": 0.19943099021580737, + "grad_norm": 6.118232727050781, + "learning_rate": 0.00019979349472763058, + "loss": 1.3665, + "step": 2874 + }, + { + "epoch": 0.19950038165290404, + "grad_norm": 4.911560535430908, + "learning_rate": 0.00019979313367121575, + "loss": 1.0775, + "step": 2875 + }, + { + "epoch": 0.1995697730900007, + "grad_norm": 4.288971900939941, + "learning_rate": 0.00019979277229976558, + "loss": 1.0249, + "step": 2876 + }, + { + "epoch": 0.19963916452709735, + "grad_norm": 4.366656303405762, + "learning_rate": 0.0001997924106132812, + "loss": 0.9051, + "step": 2877 + }, + { + "epoch": 0.199708555964194, + "grad_norm": 5.383941650390625, + "learning_rate": 0.00019979204861176377, + "loss": 1.3006, + "step": 2878 + }, + { + "epoch": 0.19977794740129068, + "grad_norm": 3.5366156101226807, + "learning_rate": 0.0001997916862952144, + "loss": 1.0563, + "step": 2879 + }, + { + "epoch": 0.19984733883838735, + "grad_norm": 5.308741092681885, + "learning_rate": 0.00019979132366363428, + "loss": 1.1454, + "step": 2880 + }, + { + "epoch": 0.199916730275484, + "grad_norm": 3.3944578170776367, + "learning_rate": 0.00019979096071702453, + "loss": 0.8452, + "step": 2881 + }, + { + "epoch": 0.19998612171258068, + "grad_norm": 3.3004045486450195, + "learning_rate": 0.00019979059745538628, + "loss": 1.0694, + "step": 2882 + }, + { + "epoch": 0.20005551314967732, + "grad_norm": 3.711270570755005, + "learning_rate": 0.0001997902338787207, + "loss": 1.0763, + "step": 2883 + }, + { + "epoch": 0.20012490458677398, + "grad_norm": 7.178623199462891, + "learning_rate": 0.00019978986998702894, + "loss": 1.6496, + "step": 2884 + }, + { + "epoch": 0.20019429602387065, + "grad_norm": 4.166405200958252, + "learning_rate": 0.00019978950578031213, + "loss": 1.2444, + "step": 2885 + }, + { + "epoch": 0.20026368746096732, + "grad_norm": 4.254423141479492, + "learning_rate": 0.00019978914125857143, + "loss": 1.2662, + "step": 2886 + }, + { + "epoch": 0.20033307889806398, + "grad_norm": 4.55605936050415, + "learning_rate": 0.00019978877642180796, + "loss": 1.0104, + "step": 2887 + }, + { + "epoch": 0.20040247033516065, + "grad_norm": 4.4496541023254395, + "learning_rate": 0.00019978841127002293, + "loss": 1.4224, + "step": 2888 + }, + { + "epoch": 0.20047186177225731, + "grad_norm": 5.4234466552734375, + "learning_rate": 0.00019978804580321745, + "loss": 1.311, + "step": 2889 + }, + { + "epoch": 0.20054125320935395, + "grad_norm": 6.573587894439697, + "learning_rate": 0.0001997876800213927, + "loss": 1.1967, + "step": 2890 + }, + { + "epoch": 0.20061064464645062, + "grad_norm": 4.4038519859313965, + "learning_rate": 0.0001997873139245498, + "loss": 0.9847, + "step": 2891 + }, + { + "epoch": 0.2006800360835473, + "grad_norm": 7.437925815582275, + "learning_rate": 0.00019978694751268995, + "loss": 1.3397, + "step": 2892 + }, + { + "epoch": 0.20074942752064395, + "grad_norm": 5.122848033905029, + "learning_rate": 0.00019978658078581427, + "loss": 1.1387, + "step": 2893 + }, + { + "epoch": 0.20081881895774062, + "grad_norm": 4.721127033233643, + "learning_rate": 0.00019978621374392396, + "loss": 1.2949, + "step": 2894 + }, + { + "epoch": 0.20088821039483729, + "grad_norm": 8.487513542175293, + "learning_rate": 0.0001997858463870201, + "loss": 1.4699, + "step": 2895 + }, + { + "epoch": 0.20095760183193395, + "grad_norm": 2.9578394889831543, + "learning_rate": 0.00019978547871510395, + "loss": 1.0233, + "step": 2896 + }, + { + "epoch": 0.2010269932690306, + "grad_norm": 2.898258924484253, + "learning_rate": 0.0001997851107281766, + "loss": 0.9478, + "step": 2897 + }, + { + "epoch": 0.20109638470612726, + "grad_norm": 5.863919258117676, + "learning_rate": 0.00019978474242623924, + "loss": 1.4887, + "step": 2898 + }, + { + "epoch": 0.20116577614322392, + "grad_norm": 4.549709320068359, + "learning_rate": 0.000199784373809293, + "loss": 1.1065, + "step": 2899 + }, + { + "epoch": 0.2012351675803206, + "grad_norm": 3.5174601078033447, + "learning_rate": 0.0001997840048773391, + "loss": 1.0441, + "step": 2900 + }, + { + "epoch": 0.20130455901741726, + "grad_norm": 4.820686340332031, + "learning_rate": 0.00019978363563037866, + "loss": 1.4277, + "step": 2901 + }, + { + "epoch": 0.20137395045451392, + "grad_norm": 4.965310573577881, + "learning_rate": 0.00019978326606841286, + "loss": 1.3159, + "step": 2902 + }, + { + "epoch": 0.2014433418916106, + "grad_norm": 3.882844924926758, + "learning_rate": 0.00019978289619144285, + "loss": 1.1556, + "step": 2903 + }, + { + "epoch": 0.20151273332870723, + "grad_norm": 3.93060040473938, + "learning_rate": 0.00019978252599946986, + "loss": 1.0177, + "step": 2904 + }, + { + "epoch": 0.2015821247658039, + "grad_norm": 3.9781312942504883, + "learning_rate": 0.00019978215549249498, + "loss": 1.272, + "step": 2905 + }, + { + "epoch": 0.20165151620290056, + "grad_norm": 4.746148109436035, + "learning_rate": 0.0001997817846705194, + "loss": 1.3192, + "step": 2906 + }, + { + "epoch": 0.20172090763999723, + "grad_norm": 3.5049240589141846, + "learning_rate": 0.00019978141353354433, + "loss": 1.2943, + "step": 2907 + }, + { + "epoch": 0.2017902990770939, + "grad_norm": 4.282980442047119, + "learning_rate": 0.0001997810420815709, + "loss": 1.2649, + "step": 2908 + }, + { + "epoch": 0.20185969051419056, + "grad_norm": 5.83132791519165, + "learning_rate": 0.00019978067031460028, + "loss": 1.2903, + "step": 2909 + }, + { + "epoch": 0.2019290819512872, + "grad_norm": 4.42138147354126, + "learning_rate": 0.0001997802982326337, + "loss": 1.2966, + "step": 2910 + }, + { + "epoch": 0.20199847338838386, + "grad_norm": 3.7836406230926514, + "learning_rate": 0.00019977992583567223, + "loss": 1.0665, + "step": 2911 + }, + { + "epoch": 0.20206786482548053, + "grad_norm": 5.493934154510498, + "learning_rate": 0.00019977955312371716, + "loss": 1.3823, + "step": 2912 + }, + { + "epoch": 0.2021372562625772, + "grad_norm": 3.3592171669006348, + "learning_rate": 0.0001997791800967696, + "loss": 1.0964, + "step": 2913 + }, + { + "epoch": 0.20220664769967386, + "grad_norm": 5.51983118057251, + "learning_rate": 0.00019977880675483078, + "loss": 1.3425, + "step": 2914 + }, + { + "epoch": 0.20227603913677053, + "grad_norm": 4.549570560455322, + "learning_rate": 0.00019977843309790182, + "loss": 0.986, + "step": 2915 + }, + { + "epoch": 0.2023454305738672, + "grad_norm": 3.624535083770752, + "learning_rate": 0.00019977805912598393, + "loss": 1.141, + "step": 2916 + }, + { + "epoch": 0.20241482201096384, + "grad_norm": 3.9567527770996094, + "learning_rate": 0.00019977768483907826, + "loss": 1.2333, + "step": 2917 + }, + { + "epoch": 0.2024842134480605, + "grad_norm": 3.7012722492218018, + "learning_rate": 0.00019977731023718603, + "loss": 0.9881, + "step": 2918 + }, + { + "epoch": 0.20255360488515717, + "grad_norm": 4.387827396392822, + "learning_rate": 0.0001997769353203084, + "loss": 1.6147, + "step": 2919 + }, + { + "epoch": 0.20262299632225383, + "grad_norm": 3.6160600185394287, + "learning_rate": 0.0001997765600884466, + "loss": 1.0753, + "step": 2920 + }, + { + "epoch": 0.2026923877593505, + "grad_norm": 4.095770835876465, + "learning_rate": 0.00019977618454160173, + "loss": 1.1279, + "step": 2921 + }, + { + "epoch": 0.20276177919644717, + "grad_norm": 3.527186155319214, + "learning_rate": 0.00019977580867977504, + "loss": 0.9916, + "step": 2922 + }, + { + "epoch": 0.20283117063354383, + "grad_norm": 5.459439754486084, + "learning_rate": 0.0001997754325029677, + "loss": 1.3739, + "step": 2923 + }, + { + "epoch": 0.20290056207064047, + "grad_norm": 5.236429691314697, + "learning_rate": 0.0001997750560111809, + "loss": 1.0592, + "step": 2924 + }, + { + "epoch": 0.20296995350773714, + "grad_norm": 4.896883487701416, + "learning_rate": 0.0001997746792044158, + "loss": 1.1337, + "step": 2925 + }, + { + "epoch": 0.2030393449448338, + "grad_norm": 17.975616455078125, + "learning_rate": 0.00019977430208267368, + "loss": 1.1402, + "step": 2926 + }, + { + "epoch": 0.20310873638193047, + "grad_norm": 3.7449135780334473, + "learning_rate": 0.0001997739246459556, + "loss": 1.409, + "step": 2927 + }, + { + "epoch": 0.20317812781902714, + "grad_norm": 4.121929168701172, + "learning_rate": 0.00019977354689426283, + "loss": 1.2574, + "step": 2928 + }, + { + "epoch": 0.2032475192561238, + "grad_norm": 3.378237247467041, + "learning_rate": 0.00019977316882759653, + "loss": 0.9382, + "step": 2929 + }, + { + "epoch": 0.20331691069322044, + "grad_norm": 3.4862635135650635, + "learning_rate": 0.00019977279044595793, + "loss": 1.2673, + "step": 2930 + }, + { + "epoch": 0.2033863021303171, + "grad_norm": 3.2356960773468018, + "learning_rate": 0.0001997724117493482, + "loss": 1.1818, + "step": 2931 + }, + { + "epoch": 0.20345569356741378, + "grad_norm": 6.237930774688721, + "learning_rate": 0.00019977203273776858, + "loss": 1.5858, + "step": 2932 + }, + { + "epoch": 0.20352508500451044, + "grad_norm": 5.5749921798706055, + "learning_rate": 0.00019977165341122016, + "loss": 1.3983, + "step": 2933 + }, + { + "epoch": 0.2035944764416071, + "grad_norm": 3.907160520553589, + "learning_rate": 0.00019977127376970426, + "loss": 1.2469, + "step": 2934 + }, + { + "epoch": 0.20366386787870377, + "grad_norm": 4.982121467590332, + "learning_rate": 0.00019977089381322202, + "loss": 1.5293, + "step": 2935 + }, + { + "epoch": 0.20373325931580044, + "grad_norm": 3.8390581607818604, + "learning_rate": 0.00019977051354177464, + "loss": 1.1312, + "step": 2936 + }, + { + "epoch": 0.20380265075289708, + "grad_norm": 4.406023979187012, + "learning_rate": 0.00019977013295536332, + "loss": 1.1906, + "step": 2937 + }, + { + "epoch": 0.20387204218999375, + "grad_norm": 5.063147068023682, + "learning_rate": 0.00019976975205398924, + "loss": 1.5578, + "step": 2938 + }, + { + "epoch": 0.2039414336270904, + "grad_norm": 4.527091026306152, + "learning_rate": 0.00019976937083765367, + "loss": 1.2927, + "step": 2939 + }, + { + "epoch": 0.20401082506418708, + "grad_norm": 4.677942276000977, + "learning_rate": 0.00019976898930635772, + "loss": 1.3274, + "step": 2940 + }, + { + "epoch": 0.20408021650128375, + "grad_norm": 3.6374008655548096, + "learning_rate": 0.00019976860746010272, + "loss": 1.0835, + "step": 2941 + }, + { + "epoch": 0.2041496079383804, + "grad_norm": 4.718100070953369, + "learning_rate": 0.00019976822529888972, + "loss": 1.0635, + "step": 2942 + }, + { + "epoch": 0.20421899937547708, + "grad_norm": 4.701373100280762, + "learning_rate": 0.00019976784282272006, + "loss": 1.1901, + "step": 2943 + }, + { + "epoch": 0.20428839081257372, + "grad_norm": 4.320451736450195, + "learning_rate": 0.0001997674600315949, + "loss": 1.0867, + "step": 2944 + }, + { + "epoch": 0.20435778224967038, + "grad_norm": 4.1948394775390625, + "learning_rate": 0.0001997670769255154, + "loss": 1.1329, + "step": 2945 + }, + { + "epoch": 0.20442717368676705, + "grad_norm": 5.090679168701172, + "learning_rate": 0.00019976669350448286, + "loss": 1.155, + "step": 2946 + }, + { + "epoch": 0.20449656512386372, + "grad_norm": 5.817384719848633, + "learning_rate": 0.0001997663097684984, + "loss": 1.0645, + "step": 2947 + }, + { + "epoch": 0.20456595656096038, + "grad_norm": 5.919626235961914, + "learning_rate": 0.00019976592571756328, + "loss": 1.2046, + "step": 2948 + }, + { + "epoch": 0.20463534799805705, + "grad_norm": 5.274261951446533, + "learning_rate": 0.00019976554135167876, + "loss": 1.2352, + "step": 2949 + }, + { + "epoch": 0.20470473943515372, + "grad_norm": 4.714739799499512, + "learning_rate": 0.00019976515667084597, + "loss": 1.3381, + "step": 2950 + }, + { + "epoch": 0.20477413087225035, + "grad_norm": 4.052755832672119, + "learning_rate": 0.00019976477167506614, + "loss": 1.185, + "step": 2951 + }, + { + "epoch": 0.20484352230934702, + "grad_norm": 4.533092498779297, + "learning_rate": 0.0001997643863643405, + "loss": 1.3364, + "step": 2952 + }, + { + "epoch": 0.2049129137464437, + "grad_norm": 5.593474864959717, + "learning_rate": 0.0001997640007386703, + "loss": 1.61, + "step": 2953 + }, + { + "epoch": 0.20498230518354035, + "grad_norm": 5.13147497177124, + "learning_rate": 0.0001997636147980567, + "loss": 1.3455, + "step": 2954 + }, + { + "epoch": 0.20505169662063702, + "grad_norm": 4.448239326477051, + "learning_rate": 0.00019976322854250094, + "loss": 1.4401, + "step": 2955 + }, + { + "epoch": 0.20512108805773369, + "grad_norm": 7.991942882537842, + "learning_rate": 0.00019976284197200425, + "loss": 1.2253, + "step": 2956 + }, + { + "epoch": 0.20519047949483032, + "grad_norm": 4.220806121826172, + "learning_rate": 0.00019976245508656785, + "loss": 1.3949, + "step": 2957 + }, + { + "epoch": 0.205259870931927, + "grad_norm": 3.935274839401245, + "learning_rate": 0.00019976206788619293, + "loss": 1.4114, + "step": 2958 + }, + { + "epoch": 0.20532926236902366, + "grad_norm": 3.569674491882324, + "learning_rate": 0.00019976168037088077, + "loss": 1.17, + "step": 2959 + }, + { + "epoch": 0.20539865380612032, + "grad_norm": 3.7807071208953857, + "learning_rate": 0.00019976129254063255, + "loss": 1.1484, + "step": 2960 + }, + { + "epoch": 0.205468045243217, + "grad_norm": 3.7290680408477783, + "learning_rate": 0.0001997609043954495, + "loss": 1.0877, + "step": 2961 + }, + { + "epoch": 0.20553743668031366, + "grad_norm": 3.787341594696045, + "learning_rate": 0.00019976051593533284, + "loss": 1.0506, + "step": 2962 + }, + { + "epoch": 0.20560682811741032, + "grad_norm": 3.4180753231048584, + "learning_rate": 0.0001997601271602838, + "loss": 0.9859, + "step": 2963 + }, + { + "epoch": 0.20567621955450696, + "grad_norm": 4.67551851272583, + "learning_rate": 0.00019975973807030361, + "loss": 1.3793, + "step": 2964 + }, + { + "epoch": 0.20574561099160363, + "grad_norm": 3.6968164443969727, + "learning_rate": 0.00019975934866539352, + "loss": 1.0798, + "step": 2965 + }, + { + "epoch": 0.2058150024287003, + "grad_norm": 4.602205753326416, + "learning_rate": 0.00019975895894555474, + "loss": 1.1035, + "step": 2966 + }, + { + "epoch": 0.20588439386579696, + "grad_norm": 4.719020843505859, + "learning_rate": 0.0001997585689107885, + "loss": 1.285, + "step": 2967 + }, + { + "epoch": 0.20595378530289363, + "grad_norm": 5.109767436981201, + "learning_rate": 0.00019975817856109602, + "loss": 1.1486, + "step": 2968 + }, + { + "epoch": 0.2060231767399903, + "grad_norm": 3.559634208679199, + "learning_rate": 0.00019975778789647856, + "loss": 1.0774, + "step": 2969 + }, + { + "epoch": 0.20609256817708696, + "grad_norm": 3.5552377700805664, + "learning_rate": 0.00019975739691693734, + "loss": 1.1302, + "step": 2970 + }, + { + "epoch": 0.2061619596141836, + "grad_norm": 3.014047145843506, + "learning_rate": 0.00019975700562247357, + "loss": 1.0919, + "step": 2971 + }, + { + "epoch": 0.20623135105128027, + "grad_norm": 5.4810895919799805, + "learning_rate": 0.00019975661401308852, + "loss": 1.3667, + "step": 2972 + }, + { + "epoch": 0.20630074248837693, + "grad_norm": 4.253967761993408, + "learning_rate": 0.00019975622208878342, + "loss": 1.3126, + "step": 2973 + }, + { + "epoch": 0.2063701339254736, + "grad_norm": 4.966777324676514, + "learning_rate": 0.00019975582984955949, + "loss": 1.0569, + "step": 2974 + }, + { + "epoch": 0.20643952536257026, + "grad_norm": 5.79423713684082, + "learning_rate": 0.00019975543729541798, + "loss": 1.3674, + "step": 2975 + }, + { + "epoch": 0.20650891679966693, + "grad_norm": 4.157489776611328, + "learning_rate": 0.00019975504442636013, + "loss": 1.0256, + "step": 2976 + }, + { + "epoch": 0.2065783082367636, + "grad_norm": 3.2287566661834717, + "learning_rate": 0.00019975465124238718, + "loss": 1.1435, + "step": 2977 + }, + { + "epoch": 0.20664769967386024, + "grad_norm": 5.357308387756348, + "learning_rate": 0.00019975425774350036, + "loss": 1.3333, + "step": 2978 + }, + { + "epoch": 0.2067170911109569, + "grad_norm": 4.576251029968262, + "learning_rate": 0.00019975386392970091, + "loss": 1.2833, + "step": 2979 + }, + { + "epoch": 0.20678648254805357, + "grad_norm": 3.548330307006836, + "learning_rate": 0.00019975346980099015, + "loss": 1.1394, + "step": 2980 + }, + { + "epoch": 0.20685587398515023, + "grad_norm": 4.014204025268555, + "learning_rate": 0.00019975307535736919, + "loss": 1.0467, + "step": 2981 + }, + { + "epoch": 0.2069252654222469, + "grad_norm": 4.204156398773193, + "learning_rate": 0.00019975268059883937, + "loss": 1.1139, + "step": 2982 + }, + { + "epoch": 0.20699465685934357, + "grad_norm": 3.3912243843078613, + "learning_rate": 0.00019975228552540194, + "loss": 1.0931, + "step": 2983 + }, + { + "epoch": 0.2070640482964402, + "grad_norm": 3.807462692260742, + "learning_rate": 0.00019975189013705805, + "loss": 0.9509, + "step": 2984 + }, + { + "epoch": 0.20713343973353687, + "grad_norm": 3.8010852336883545, + "learning_rate": 0.00019975149443380907, + "loss": 0.9769, + "step": 2985 + }, + { + "epoch": 0.20720283117063354, + "grad_norm": 4.546963214874268, + "learning_rate": 0.00019975109841565618, + "loss": 1.1506, + "step": 2986 + }, + { + "epoch": 0.2072722226077302, + "grad_norm": 4.578310489654541, + "learning_rate": 0.00019975070208260064, + "loss": 1.3071, + "step": 2987 + }, + { + "epoch": 0.20734161404482687, + "grad_norm": 3.875134229660034, + "learning_rate": 0.00019975030543464373, + "loss": 1.3068, + "step": 2988 + }, + { + "epoch": 0.20741100548192354, + "grad_norm": 4.4185285568237305, + "learning_rate": 0.00019974990847178666, + "loss": 1.2715, + "step": 2989 + }, + { + "epoch": 0.2074803969190202, + "grad_norm": 4.311311721801758, + "learning_rate": 0.0001997495111940307, + "loss": 1.2982, + "step": 2990 + }, + { + "epoch": 0.20754978835611684, + "grad_norm": 3.4881842136383057, + "learning_rate": 0.00019974911360137713, + "loss": 1.1827, + "step": 2991 + }, + { + "epoch": 0.2076191797932135, + "grad_norm": 4.222468852996826, + "learning_rate": 0.00019974871569382715, + "loss": 1.6806, + "step": 2992 + }, + { + "epoch": 0.20768857123031018, + "grad_norm": 3.543072462081909, + "learning_rate": 0.00019974831747138203, + "loss": 0.9528, + "step": 2993 + }, + { + "epoch": 0.20775796266740684, + "grad_norm": 3.857567310333252, + "learning_rate": 0.00019974791893404307, + "loss": 1.1523, + "step": 2994 + }, + { + "epoch": 0.2078273541045035, + "grad_norm": 3.904134511947632, + "learning_rate": 0.00019974752008181153, + "loss": 1.348, + "step": 2995 + }, + { + "epoch": 0.20789674554160018, + "grad_norm": 3.8891429901123047, + "learning_rate": 0.00019974712091468862, + "loss": 1.2771, + "step": 2996 + }, + { + "epoch": 0.20796613697869684, + "grad_norm": 3.8353962898254395, + "learning_rate": 0.00019974672143267558, + "loss": 0.988, + "step": 2997 + }, + { + "epoch": 0.20803552841579348, + "grad_norm": 5.3516082763671875, + "learning_rate": 0.00019974632163577378, + "loss": 1.1472, + "step": 2998 + }, + { + "epoch": 0.20810491985289015, + "grad_norm": 4.194650173187256, + "learning_rate": 0.00019974592152398438, + "loss": 1.1482, + "step": 2999 + }, + { + "epoch": 0.2081743112899868, + "grad_norm": 4.406322956085205, + "learning_rate": 0.0001997455210973087, + "loss": 0.9168, + "step": 3000 + }, + { + "epoch": 0.20824370272708348, + "grad_norm": 8.059192657470703, + "learning_rate": 0.00019974512035574797, + "loss": 1.0856, + "step": 3001 + }, + { + "epoch": 0.20831309416418015, + "grad_norm": 3.698859214782715, + "learning_rate": 0.00019974471929930346, + "loss": 1.0622, + "step": 3002 + }, + { + "epoch": 0.2083824856012768, + "grad_norm": 7.973196983337402, + "learning_rate": 0.00019974431792797644, + "loss": 1.2689, + "step": 3003 + }, + { + "epoch": 0.20845187703837345, + "grad_norm": 4.443449020385742, + "learning_rate": 0.0001997439162417682, + "loss": 1.3073, + "step": 3004 + }, + { + "epoch": 0.20852126847547012, + "grad_norm": 4.003130912780762, + "learning_rate": 0.00019974351424068, + "loss": 1.4822, + "step": 3005 + }, + { + "epoch": 0.20859065991256678, + "grad_norm": 3.622300148010254, + "learning_rate": 0.00019974311192471307, + "loss": 1.0209, + "step": 3006 + }, + { + "epoch": 0.20866005134966345, + "grad_norm": 4.889700412750244, + "learning_rate": 0.00019974270929386872, + "loss": 1.354, + "step": 3007 + }, + { + "epoch": 0.20872944278676012, + "grad_norm": 3.911736011505127, + "learning_rate": 0.0001997423063481482, + "loss": 1.4763, + "step": 3008 + }, + { + "epoch": 0.20879883422385678, + "grad_norm": 3.6710002422332764, + "learning_rate": 0.0001997419030875528, + "loss": 0.7301, + "step": 3009 + }, + { + "epoch": 0.20886822566095345, + "grad_norm": 2.7689270973205566, + "learning_rate": 0.00019974149951208377, + "loss": 0.9784, + "step": 3010 + }, + { + "epoch": 0.2089376170980501, + "grad_norm": 2.8861751556396484, + "learning_rate": 0.0001997410956217424, + "loss": 0.753, + "step": 3011 + }, + { + "epoch": 0.20900700853514675, + "grad_norm": 4.08631706237793, + "learning_rate": 0.00019974069141652997, + "loss": 1.2242, + "step": 3012 + }, + { + "epoch": 0.20907639997224342, + "grad_norm": 3.530897617340088, + "learning_rate": 0.00019974028689644774, + "loss": 1.08, + "step": 3013 + }, + { + "epoch": 0.2091457914093401, + "grad_norm": 2.8176307678222656, + "learning_rate": 0.00019973988206149702, + "loss": 0.7552, + "step": 3014 + }, + { + "epoch": 0.20921518284643675, + "grad_norm": 3.2544541358947754, + "learning_rate": 0.00019973947691167906, + "loss": 1.0011, + "step": 3015 + }, + { + "epoch": 0.20928457428353342, + "grad_norm": 6.76078462600708, + "learning_rate": 0.00019973907144699512, + "loss": 1.0962, + "step": 3016 + }, + { + "epoch": 0.2093539657206301, + "grad_norm": 3.652488946914673, + "learning_rate": 0.00019973866566744652, + "loss": 1.1593, + "step": 3017 + }, + { + "epoch": 0.20942335715772672, + "grad_norm": 4.563908100128174, + "learning_rate": 0.00019973825957303454, + "loss": 1.2292, + "step": 3018 + }, + { + "epoch": 0.2094927485948234, + "grad_norm": 4.71771764755249, + "learning_rate": 0.0001997378531637604, + "loss": 1.34, + "step": 3019 + }, + { + "epoch": 0.20956214003192006, + "grad_norm": 3.750929117202759, + "learning_rate": 0.0001997374464396255, + "loss": 1.176, + "step": 3020 + }, + { + "epoch": 0.20963153146901672, + "grad_norm": 4.233251571655273, + "learning_rate": 0.000199737039400631, + "loss": 1.2949, + "step": 3021 + }, + { + "epoch": 0.2097009229061134, + "grad_norm": 3.654665946960449, + "learning_rate": 0.00019973663204677825, + "loss": 0.9466, + "step": 3022 + }, + { + "epoch": 0.20977031434321006, + "grad_norm": 3.0407960414886475, + "learning_rate": 0.0001997362243780685, + "loss": 0.9559, + "step": 3023 + }, + { + "epoch": 0.20983970578030672, + "grad_norm": 4.745643615722656, + "learning_rate": 0.00019973581639450312, + "loss": 1.3546, + "step": 3024 + }, + { + "epoch": 0.20990909721740336, + "grad_norm": 4.1548686027526855, + "learning_rate": 0.00019973540809608328, + "loss": 0.9962, + "step": 3025 + }, + { + "epoch": 0.20997848865450003, + "grad_norm": 3.6019062995910645, + "learning_rate": 0.00019973499948281037, + "loss": 1.0866, + "step": 3026 + }, + { + "epoch": 0.2100478800915967, + "grad_norm": 4.558901309967041, + "learning_rate": 0.0001997345905546856, + "loss": 1.0268, + "step": 3027 + }, + { + "epoch": 0.21011727152869336, + "grad_norm": 4.993839740753174, + "learning_rate": 0.00019973418131171032, + "loss": 1.1223, + "step": 3028 + }, + { + "epoch": 0.21018666296579003, + "grad_norm": 4.202278137207031, + "learning_rate": 0.00019973377175388582, + "loss": 1.1018, + "step": 3029 + }, + { + "epoch": 0.2102560544028867, + "grad_norm": 3.78668212890625, + "learning_rate": 0.00019973336188121334, + "loss": 1.058, + "step": 3030 + }, + { + "epoch": 0.21032544583998333, + "grad_norm": 5.826493263244629, + "learning_rate": 0.00019973295169369425, + "loss": 1.0948, + "step": 3031 + }, + { + "epoch": 0.21039483727708, + "grad_norm": 4.968171119689941, + "learning_rate": 0.00019973254119132972, + "loss": 1.0363, + "step": 3032 + }, + { + "epoch": 0.21046422871417667, + "grad_norm": 3.142033100128174, + "learning_rate": 0.0001997321303741212, + "loss": 0.889, + "step": 3033 + }, + { + "epoch": 0.21053362015127333, + "grad_norm": 3.8012309074401855, + "learning_rate": 0.00019973171924206993, + "loss": 1.1974, + "step": 3034 + }, + { + "epoch": 0.21060301158837, + "grad_norm": 3.6583549976348877, + "learning_rate": 0.00019973130779517715, + "loss": 0.7232, + "step": 3035 + }, + { + "epoch": 0.21067240302546666, + "grad_norm": 4.051336765289307, + "learning_rate": 0.0001997308960334442, + "loss": 1.1256, + "step": 3036 + }, + { + "epoch": 0.21074179446256333, + "grad_norm": 6.5153374671936035, + "learning_rate": 0.00019973048395687242, + "loss": 1.3731, + "step": 3037 + }, + { + "epoch": 0.21081118589965997, + "grad_norm": 3.637132167816162, + "learning_rate": 0.00019973007156546304, + "loss": 0.8698, + "step": 3038 + }, + { + "epoch": 0.21088057733675664, + "grad_norm": 4.759273052215576, + "learning_rate": 0.00019972965885921743, + "loss": 0.9946, + "step": 3039 + }, + { + "epoch": 0.2109499687738533, + "grad_norm": 3.876715660095215, + "learning_rate": 0.0001997292458381368, + "loss": 1.1266, + "step": 3040 + }, + { + "epoch": 0.21101936021094997, + "grad_norm": 5.390665531158447, + "learning_rate": 0.00019972883250222255, + "loss": 1.2123, + "step": 3041 + }, + { + "epoch": 0.21108875164804664, + "grad_norm": 3.7546565532684326, + "learning_rate": 0.00019972841885147595, + "loss": 1.0468, + "step": 3042 + }, + { + "epoch": 0.2111581430851433, + "grad_norm": 5.417148113250732, + "learning_rate": 0.00019972800488589825, + "loss": 1.1736, + "step": 3043 + }, + { + "epoch": 0.21122753452223997, + "grad_norm": 3.615832805633545, + "learning_rate": 0.00019972759060549086, + "loss": 1.0822, + "step": 3044 + }, + { + "epoch": 0.2112969259593366, + "grad_norm": 3.755284070968628, + "learning_rate": 0.00019972717601025506, + "loss": 0.9872, + "step": 3045 + }, + { + "epoch": 0.21136631739643327, + "grad_norm": 3.968933582305908, + "learning_rate": 0.0001997267611001921, + "loss": 1.0709, + "step": 3046 + }, + { + "epoch": 0.21143570883352994, + "grad_norm": 3.824019193649292, + "learning_rate": 0.0001997263458753033, + "loss": 1.1221, + "step": 3047 + }, + { + "epoch": 0.2115051002706266, + "grad_norm": 6.360177040100098, + "learning_rate": 0.00019972593033559005, + "loss": 1.0385, + "step": 3048 + }, + { + "epoch": 0.21157449170772327, + "grad_norm": 4.6055378913879395, + "learning_rate": 0.0001997255144810536, + "loss": 1.1781, + "step": 3049 + }, + { + "epoch": 0.21164388314481994, + "grad_norm": 4.44965124130249, + "learning_rate": 0.00019972509831169525, + "loss": 1.3657, + "step": 3050 + }, + { + "epoch": 0.2117132745819166, + "grad_norm": 4.176546573638916, + "learning_rate": 0.00019972468182751632, + "loss": 1.2039, + "step": 3051 + }, + { + "epoch": 0.21178266601901324, + "grad_norm": 4.829652786254883, + "learning_rate": 0.00019972426502851818, + "loss": 1.2228, + "step": 3052 + }, + { + "epoch": 0.2118520574561099, + "grad_norm": 5.424121379852295, + "learning_rate": 0.00019972384791470206, + "loss": 0.9206, + "step": 3053 + }, + { + "epoch": 0.21192144889320658, + "grad_norm": 21.494966506958008, + "learning_rate": 0.00019972343048606938, + "loss": 1.3292, + "step": 3054 + }, + { + "epoch": 0.21199084033030324, + "grad_norm": 4.2690348625183105, + "learning_rate": 0.00019972301274262137, + "loss": 1.1011, + "step": 3055 + }, + { + "epoch": 0.2120602317673999, + "grad_norm": 3.6559665203094482, + "learning_rate": 0.00019972259468435937, + "loss": 1.0236, + "step": 3056 + }, + { + "epoch": 0.21212962320449658, + "grad_norm": 4.068591117858887, + "learning_rate": 0.00019972217631128471, + "loss": 1.0541, + "step": 3057 + }, + { + "epoch": 0.21219901464159321, + "grad_norm": 5.556787014007568, + "learning_rate": 0.00019972175762339873, + "loss": 1.37, + "step": 3058 + }, + { + "epoch": 0.21226840607868988, + "grad_norm": 4.4398112297058105, + "learning_rate": 0.0001997213386207027, + "loss": 1.5848, + "step": 3059 + }, + { + "epoch": 0.21233779751578655, + "grad_norm": 4.322177410125732, + "learning_rate": 0.000199720919303198, + "loss": 1.1317, + "step": 3060 + }, + { + "epoch": 0.2124071889528832, + "grad_norm": 3.5861847400665283, + "learning_rate": 0.0001997204996708859, + "loss": 1.3008, + "step": 3061 + }, + { + "epoch": 0.21247658038997988, + "grad_norm": 4.300471782684326, + "learning_rate": 0.00019972007972376778, + "loss": 1.02, + "step": 3062 + }, + { + "epoch": 0.21254597182707655, + "grad_norm": 3.1777474880218506, + "learning_rate": 0.00019971965946184494, + "loss": 0.9895, + "step": 3063 + }, + { + "epoch": 0.2126153632641732, + "grad_norm": 6.267367839813232, + "learning_rate": 0.00019971923888511867, + "loss": 1.7554, + "step": 3064 + }, + { + "epoch": 0.21268475470126985, + "grad_norm": 4.416842937469482, + "learning_rate": 0.00019971881799359036, + "loss": 1.2812, + "step": 3065 + }, + { + "epoch": 0.21275414613836652, + "grad_norm": 3.725861072540283, + "learning_rate": 0.0001997183967872613, + "loss": 1.1481, + "step": 3066 + }, + { + "epoch": 0.21282353757546318, + "grad_norm": 7.084883213043213, + "learning_rate": 0.00019971797526613285, + "loss": 1.5425, + "step": 3067 + }, + { + "epoch": 0.21289292901255985, + "grad_norm": 3.7342708110809326, + "learning_rate": 0.0001997175534302063, + "loss": 1.2238, + "step": 3068 + }, + { + "epoch": 0.21296232044965652, + "grad_norm": 5.020965099334717, + "learning_rate": 0.00019971713127948302, + "loss": 1.1054, + "step": 3069 + }, + { + "epoch": 0.21303171188675318, + "grad_norm": 3.975226879119873, + "learning_rate": 0.00019971670881396432, + "loss": 1.0459, + "step": 3070 + }, + { + "epoch": 0.21310110332384985, + "grad_norm": 3.993682622909546, + "learning_rate": 0.00019971628603365153, + "loss": 1.1401, + "step": 3071 + }, + { + "epoch": 0.2131704947609465, + "grad_norm": 3.806103229522705, + "learning_rate": 0.000199715862938546, + "loss": 1.2971, + "step": 3072 + }, + { + "epoch": 0.21323988619804315, + "grad_norm": 6.105772018432617, + "learning_rate": 0.00019971543952864907, + "loss": 1.3646, + "step": 3073 + }, + { + "epoch": 0.21330927763513982, + "grad_norm": 3.565213918685913, + "learning_rate": 0.00019971501580396206, + "loss": 1.1254, + "step": 3074 + }, + { + "epoch": 0.2133786690722365, + "grad_norm": 4.600832939147949, + "learning_rate": 0.0001997145917644863, + "loss": 1.0968, + "step": 3075 + }, + { + "epoch": 0.21344806050933315, + "grad_norm": 3.6163392066955566, + "learning_rate": 0.00019971416741022314, + "loss": 1.0152, + "step": 3076 + }, + { + "epoch": 0.21351745194642982, + "grad_norm": 3.5673468112945557, + "learning_rate": 0.00019971374274117395, + "loss": 1.1597, + "step": 3077 + }, + { + "epoch": 0.21358684338352646, + "grad_norm": 4.041092872619629, + "learning_rate": 0.00019971331775734004, + "loss": 1.163, + "step": 3078 + }, + { + "epoch": 0.21365623482062313, + "grad_norm": 4.059386730194092, + "learning_rate": 0.00019971289245872273, + "loss": 1.1002, + "step": 3079 + }, + { + "epoch": 0.2137256262577198, + "grad_norm": 5.601782321929932, + "learning_rate": 0.00019971246684532341, + "loss": 1.5184, + "step": 3080 + }, + { + "epoch": 0.21379501769481646, + "grad_norm": 5.538463115692139, + "learning_rate": 0.0001997120409171434, + "loss": 1.1078, + "step": 3081 + }, + { + "epoch": 0.21386440913191312, + "grad_norm": 4.107760429382324, + "learning_rate": 0.00019971161467418401, + "loss": 1.2902, + "step": 3082 + }, + { + "epoch": 0.2139338005690098, + "grad_norm": 5.227799892425537, + "learning_rate": 0.00019971118811644667, + "loss": 1.237, + "step": 3083 + }, + { + "epoch": 0.21400319200610646, + "grad_norm": 3.762237071990967, + "learning_rate": 0.00019971076124393266, + "loss": 1.0509, + "step": 3084 + }, + { + "epoch": 0.2140725834432031, + "grad_norm": 4.3305816650390625, + "learning_rate": 0.00019971033405664335, + "loss": 1.0033, + "step": 3085 + }, + { + "epoch": 0.21414197488029976, + "grad_norm": 3.956852436065674, + "learning_rate": 0.00019970990655458008, + "loss": 1.4796, + "step": 3086 + }, + { + "epoch": 0.21421136631739643, + "grad_norm": 4.363866329193115, + "learning_rate": 0.0001997094787377442, + "loss": 1.1713, + "step": 3087 + }, + { + "epoch": 0.2142807577544931, + "grad_norm": 4.167169570922852, + "learning_rate": 0.00019970905060613707, + "loss": 1.1058, + "step": 3088 + }, + { + "epoch": 0.21435014919158976, + "grad_norm": 4.19512414932251, + "learning_rate": 0.00019970862215976004, + "loss": 1.2028, + "step": 3089 + }, + { + "epoch": 0.21441954062868643, + "grad_norm": 4.011014461517334, + "learning_rate": 0.00019970819339861442, + "loss": 1.0549, + "step": 3090 + }, + { + "epoch": 0.2144889320657831, + "grad_norm": 4.4555158615112305, + "learning_rate": 0.00019970776432270162, + "loss": 1.2174, + "step": 3091 + }, + { + "epoch": 0.21455832350287973, + "grad_norm": 4.037343502044678, + "learning_rate": 0.000199707334932023, + "loss": 1.1481, + "step": 3092 + }, + { + "epoch": 0.2146277149399764, + "grad_norm": 4.894351482391357, + "learning_rate": 0.00019970690522657985, + "loss": 1.0367, + "step": 3093 + }, + { + "epoch": 0.21469710637707307, + "grad_norm": 4.08265495300293, + "learning_rate": 0.00019970647520637362, + "loss": 1.0245, + "step": 3094 + }, + { + "epoch": 0.21476649781416973, + "grad_norm": 3.547255516052246, + "learning_rate": 0.0001997060448714056, + "loss": 0.9017, + "step": 3095 + }, + { + "epoch": 0.2148358892512664, + "grad_norm": 3.878047227859497, + "learning_rate": 0.00019970561422167715, + "loss": 1.2206, + "step": 3096 + }, + { + "epoch": 0.21490528068836307, + "grad_norm": 3.1555233001708984, + "learning_rate": 0.00019970518325718964, + "loss": 1.0309, + "step": 3097 + }, + { + "epoch": 0.21497467212545973, + "grad_norm": 3.9546492099761963, + "learning_rate": 0.00019970475197794447, + "loss": 1.2343, + "step": 3098 + }, + { + "epoch": 0.21504406356255637, + "grad_norm": 3.4174108505249023, + "learning_rate": 0.00019970432038394294, + "loss": 1.0099, + "step": 3099 + }, + { + "epoch": 0.21511345499965304, + "grad_norm": 3.645648717880249, + "learning_rate": 0.00019970388847518643, + "loss": 1.0339, + "step": 3100 + }, + { + "epoch": 0.2151828464367497, + "grad_norm": 3.4233975410461426, + "learning_rate": 0.00019970345625167633, + "loss": 1.1869, + "step": 3101 + }, + { + "epoch": 0.21525223787384637, + "grad_norm": 4.90287446975708, + "learning_rate": 0.00019970302371341396, + "loss": 1.0668, + "step": 3102 + }, + { + "epoch": 0.21532162931094304, + "grad_norm": 3.8524627685546875, + "learning_rate": 0.00019970259086040076, + "loss": 1.2459, + "step": 3103 + }, + { + "epoch": 0.2153910207480397, + "grad_norm": 3.823697328567505, + "learning_rate": 0.00019970215769263798, + "loss": 1.1579, + "step": 3104 + }, + { + "epoch": 0.21546041218513634, + "grad_norm": 4.7424726486206055, + "learning_rate": 0.0001997017242101271, + "loss": 1.1011, + "step": 3105 + }, + { + "epoch": 0.215529803622233, + "grad_norm": 3.8474504947662354, + "learning_rate": 0.00019970129041286944, + "loss": 1.2427, + "step": 3106 + }, + { + "epoch": 0.21559919505932967, + "grad_norm": 3.188915729522705, + "learning_rate": 0.0001997008563008664, + "loss": 1.0053, + "step": 3107 + }, + { + "epoch": 0.21566858649642634, + "grad_norm": 3.8552968502044678, + "learning_rate": 0.00019970042187411927, + "loss": 0.9758, + "step": 3108 + }, + { + "epoch": 0.215737977933523, + "grad_norm": 3.496976613998413, + "learning_rate": 0.0001996999871326295, + "loss": 1.3005, + "step": 3109 + }, + { + "epoch": 0.21580736937061967, + "grad_norm": 4.095437526702881, + "learning_rate": 0.00019969955207639845, + "loss": 1.2688, + "step": 3110 + }, + { + "epoch": 0.21587676080771634, + "grad_norm": 4.747699737548828, + "learning_rate": 0.00019969911670542745, + "loss": 1.0233, + "step": 3111 + }, + { + "epoch": 0.21594615224481298, + "grad_norm": 4.587327480316162, + "learning_rate": 0.00019969868101971794, + "loss": 0.9687, + "step": 3112 + }, + { + "epoch": 0.21601554368190964, + "grad_norm": 4.41202974319458, + "learning_rate": 0.00019969824501927125, + "loss": 1.0383, + "step": 3113 + }, + { + "epoch": 0.2160849351190063, + "grad_norm": 3.176185369491577, + "learning_rate": 0.00019969780870408876, + "loss": 0.8524, + "step": 3114 + }, + { + "epoch": 0.21615432655610298, + "grad_norm": 4.0029377937316895, + "learning_rate": 0.00019969737207417186, + "loss": 1.1575, + "step": 3115 + }, + { + "epoch": 0.21622371799319964, + "grad_norm": 3.491671562194824, + "learning_rate": 0.00019969693512952193, + "loss": 0.8858, + "step": 3116 + }, + { + "epoch": 0.2162931094302963, + "grad_norm": 5.11273193359375, + "learning_rate": 0.00019969649787014032, + "loss": 1.0185, + "step": 3117 + }, + { + "epoch": 0.21636250086739298, + "grad_norm": 3.912785768508911, + "learning_rate": 0.00019969606029602844, + "loss": 1.2197, + "step": 3118 + }, + { + "epoch": 0.21643189230448961, + "grad_norm": 4.161687850952148, + "learning_rate": 0.0001996956224071877, + "loss": 1.1207, + "step": 3119 + }, + { + "epoch": 0.21650128374158628, + "grad_norm": 3.3861472606658936, + "learning_rate": 0.0001996951842036194, + "loss": 1.0676, + "step": 3120 + }, + { + "epoch": 0.21657067517868295, + "grad_norm": 4.228462219238281, + "learning_rate": 0.000199694745685325, + "loss": 1.009, + "step": 3121 + }, + { + "epoch": 0.21664006661577961, + "grad_norm": 4.2208733558654785, + "learning_rate": 0.00019969430685230584, + "loss": 0.9891, + "step": 3122 + }, + { + "epoch": 0.21670945805287628, + "grad_norm": 4.09535026550293, + "learning_rate": 0.00019969386770456334, + "loss": 1.0282, + "step": 3123 + }, + { + "epoch": 0.21677884948997295, + "grad_norm": 2.9558775424957275, + "learning_rate": 0.00019969342824209885, + "loss": 0.9361, + "step": 3124 + }, + { + "epoch": 0.2168482409270696, + "grad_norm": 3.242819309234619, + "learning_rate": 0.00019969298846491377, + "loss": 1.3304, + "step": 3125 + }, + { + "epoch": 0.21691763236416625, + "grad_norm": 2.9274864196777344, + "learning_rate": 0.00019969254837300948, + "loss": 1.0149, + "step": 3126 + }, + { + "epoch": 0.21698702380126292, + "grad_norm": 3.997258424758911, + "learning_rate": 0.00019969210796638742, + "loss": 1.2876, + "step": 3127 + }, + { + "epoch": 0.21705641523835958, + "grad_norm": 5.546414375305176, + "learning_rate": 0.00019969166724504892, + "loss": 1.032, + "step": 3128 + }, + { + "epoch": 0.21712580667545625, + "grad_norm": 3.764484167098999, + "learning_rate": 0.00019969122620899538, + "loss": 1.0737, + "step": 3129 + }, + { + "epoch": 0.21719519811255292, + "grad_norm": 4.322377681732178, + "learning_rate": 0.00019969078485822822, + "loss": 1.0457, + "step": 3130 + }, + { + "epoch": 0.21726458954964958, + "grad_norm": 5.312538146972656, + "learning_rate": 0.00019969034319274883, + "loss": 1.0426, + "step": 3131 + }, + { + "epoch": 0.21733398098674622, + "grad_norm": 5.265117645263672, + "learning_rate": 0.00019968990121255856, + "loss": 1.417, + "step": 3132 + }, + { + "epoch": 0.2174033724238429, + "grad_norm": 3.31270170211792, + "learning_rate": 0.00019968945891765886, + "loss": 1.1427, + "step": 3133 + }, + { + "epoch": 0.21747276386093956, + "grad_norm": 4.52830696105957, + "learning_rate": 0.0001996890163080511, + "loss": 1.1944, + "step": 3134 + }, + { + "epoch": 0.21754215529803622, + "grad_norm": 3.506591796875, + "learning_rate": 0.0001996885733837367, + "loss": 0.8557, + "step": 3135 + }, + { + "epoch": 0.2176115467351329, + "grad_norm": 3.28827166557312, + "learning_rate": 0.000199688130144717, + "loss": 1.2502, + "step": 3136 + }, + { + "epoch": 0.21768093817222955, + "grad_norm": 3.5695624351501465, + "learning_rate": 0.0001996876865909935, + "loss": 1.0577, + "step": 3137 + }, + { + "epoch": 0.21775032960932622, + "grad_norm": 3.3362250328063965, + "learning_rate": 0.00019968724272256748, + "loss": 0.9478, + "step": 3138 + }, + { + "epoch": 0.21781972104642286, + "grad_norm": 4.378856182098389, + "learning_rate": 0.00019968679853944044, + "loss": 1.1155, + "step": 3139 + }, + { + "epoch": 0.21788911248351953, + "grad_norm": 3.461195230484009, + "learning_rate": 0.0001996863540416137, + "loss": 0.8543, + "step": 3140 + }, + { + "epoch": 0.2179585039206162, + "grad_norm": 3.475688934326172, + "learning_rate": 0.00019968590922908873, + "loss": 1.224, + "step": 3141 + }, + { + "epoch": 0.21802789535771286, + "grad_norm": 3.875675678253174, + "learning_rate": 0.00019968546410186692, + "loss": 1.337, + "step": 3142 + }, + { + "epoch": 0.21809728679480953, + "grad_norm": 4.668177127838135, + "learning_rate": 0.00019968501865994965, + "loss": 1.4186, + "step": 3143 + }, + { + "epoch": 0.2181666782319062, + "grad_norm": 3.9721405506134033, + "learning_rate": 0.00019968457290333834, + "loss": 1.3765, + "step": 3144 + }, + { + "epoch": 0.21823606966900286, + "grad_norm": 2.991224765777588, + "learning_rate": 0.00019968412683203443, + "loss": 0.9764, + "step": 3145 + }, + { + "epoch": 0.2183054611060995, + "grad_norm": 5.110465049743652, + "learning_rate": 0.00019968368044603925, + "loss": 1.2652, + "step": 3146 + }, + { + "epoch": 0.21837485254319616, + "grad_norm": 5.202994346618652, + "learning_rate": 0.0001996832337453543, + "loss": 1.3279, + "step": 3147 + }, + { + "epoch": 0.21844424398029283, + "grad_norm": 3.9133448600769043, + "learning_rate": 0.00019968278672998093, + "loss": 1.0101, + "step": 3148 + }, + { + "epoch": 0.2185136354173895, + "grad_norm": 3.0206916332244873, + "learning_rate": 0.00019968233939992057, + "loss": 0.9985, + "step": 3149 + }, + { + "epoch": 0.21858302685448616, + "grad_norm": 3.501222610473633, + "learning_rate": 0.0001996818917551746, + "loss": 1.0831, + "step": 3150 + }, + { + "epoch": 0.21865241829158283, + "grad_norm": 3.6121320724487305, + "learning_rate": 0.0001996814437957445, + "loss": 1.1218, + "step": 3151 + }, + { + "epoch": 0.21872180972867947, + "grad_norm": 4.222762107849121, + "learning_rate": 0.00019968099552163164, + "loss": 1.2075, + "step": 3152 + }, + { + "epoch": 0.21879120116577613, + "grad_norm": 4.25860595703125, + "learning_rate": 0.0001996805469328374, + "loss": 1.282, + "step": 3153 + }, + { + "epoch": 0.2188605926028728, + "grad_norm": 3.4862329959869385, + "learning_rate": 0.00019968009802936329, + "loss": 1.1053, + "step": 3154 + }, + { + "epoch": 0.21892998403996947, + "grad_norm": 3.6504476070404053, + "learning_rate": 0.00019967964881121067, + "loss": 1.0316, + "step": 3155 + }, + { + "epoch": 0.21899937547706613, + "grad_norm": 5.0253214836120605, + "learning_rate": 0.00019967919927838096, + "loss": 1.1958, + "step": 3156 + }, + { + "epoch": 0.2190687669141628, + "grad_norm": 3.3643994331359863, + "learning_rate": 0.00019967874943087557, + "loss": 1.0286, + "step": 3157 + }, + { + "epoch": 0.21913815835125947, + "grad_norm": 3.9395413398742676, + "learning_rate": 0.00019967829926869593, + "loss": 1.0615, + "step": 3158 + }, + { + "epoch": 0.2192075497883561, + "grad_norm": 3.234104871749878, + "learning_rate": 0.0001996778487918435, + "loss": 1.0634, + "step": 3159 + }, + { + "epoch": 0.21927694122545277, + "grad_norm": 3.117150068283081, + "learning_rate": 0.00019967739800031962, + "loss": 1.0553, + "step": 3160 + }, + { + "epoch": 0.21934633266254944, + "grad_norm": 3.0221364498138428, + "learning_rate": 0.0001996769468941258, + "loss": 1.0189, + "step": 3161 + }, + { + "epoch": 0.2194157240996461, + "grad_norm": 4.399488925933838, + "learning_rate": 0.00019967649547326341, + "loss": 1.2688, + "step": 3162 + }, + { + "epoch": 0.21948511553674277, + "grad_norm": 3.4013054370880127, + "learning_rate": 0.00019967604373773387, + "loss": 1.0077, + "step": 3163 + }, + { + "epoch": 0.21955450697383944, + "grad_norm": 3.5338356494903564, + "learning_rate": 0.00019967559168753863, + "loss": 0.8637, + "step": 3164 + }, + { + "epoch": 0.2196238984109361, + "grad_norm": 3.801093339920044, + "learning_rate": 0.00019967513932267914, + "loss": 1.1531, + "step": 3165 + }, + { + "epoch": 0.21969328984803274, + "grad_norm": 4.460262775421143, + "learning_rate": 0.00019967468664315675, + "loss": 1.4617, + "step": 3166 + }, + { + "epoch": 0.2197626812851294, + "grad_norm": 3.9225893020629883, + "learning_rate": 0.00019967423364897297, + "loss": 1.2288, + "step": 3167 + }, + { + "epoch": 0.21983207272222607, + "grad_norm": 4.812930583953857, + "learning_rate": 0.0001996737803401292, + "loss": 1.0049, + "step": 3168 + }, + { + "epoch": 0.21990146415932274, + "grad_norm": 5.009568214416504, + "learning_rate": 0.00019967332671662685, + "loss": 1.0674, + "step": 3169 + }, + { + "epoch": 0.2199708555964194, + "grad_norm": 3.5705783367156982, + "learning_rate": 0.0001996728727784674, + "loss": 1.0898, + "step": 3170 + }, + { + "epoch": 0.22004024703351607, + "grad_norm": 4.347940921783447, + "learning_rate": 0.00019967241852565225, + "loss": 1.2236, + "step": 3171 + }, + { + "epoch": 0.22010963847061274, + "grad_norm": 3.114293098449707, + "learning_rate": 0.0001996719639581828, + "loss": 1.1008, + "step": 3172 + }, + { + "epoch": 0.22017902990770938, + "grad_norm": 3.9255423545837402, + "learning_rate": 0.00019967150907606058, + "loss": 1.0772, + "step": 3173 + }, + { + "epoch": 0.22024842134480604, + "grad_norm": 3.6733553409576416, + "learning_rate": 0.00019967105387928692, + "loss": 0.8894, + "step": 3174 + }, + { + "epoch": 0.2203178127819027, + "grad_norm": 3.5624911785125732, + "learning_rate": 0.00019967059836786336, + "loss": 0.8287, + "step": 3175 + }, + { + "epoch": 0.22038720421899938, + "grad_norm": 4.063735485076904, + "learning_rate": 0.00019967014254179124, + "loss": 1.159, + "step": 3176 + }, + { + "epoch": 0.22045659565609604, + "grad_norm": 3.3346779346466064, + "learning_rate": 0.00019966968640107208, + "loss": 1.1856, + "step": 3177 + }, + { + "epoch": 0.2205259870931927, + "grad_norm": 3.2622509002685547, + "learning_rate": 0.00019966922994570725, + "loss": 1.028, + "step": 3178 + }, + { + "epoch": 0.22059537853028935, + "grad_norm": 3.632016658782959, + "learning_rate": 0.00019966877317569823, + "loss": 1.1392, + "step": 3179 + }, + { + "epoch": 0.22066476996738602, + "grad_norm": 3.7477331161499023, + "learning_rate": 0.00019966831609104647, + "loss": 1.0171, + "step": 3180 + }, + { + "epoch": 0.22073416140448268, + "grad_norm": 4.725825786590576, + "learning_rate": 0.0001996678586917534, + "loss": 1.1822, + "step": 3181 + }, + { + "epoch": 0.22080355284157935, + "grad_norm": 4.656135082244873, + "learning_rate": 0.00019966740097782047, + "loss": 1.0145, + "step": 3182 + }, + { + "epoch": 0.22087294427867601, + "grad_norm": 3.8370237350463867, + "learning_rate": 0.00019966694294924907, + "loss": 1.247, + "step": 3183 + }, + { + "epoch": 0.22094233571577268, + "grad_norm": 3.8279199600219727, + "learning_rate": 0.00019966648460604077, + "loss": 1.2594, + "step": 3184 + }, + { + "epoch": 0.22101172715286935, + "grad_norm": 3.2520570755004883, + "learning_rate": 0.0001996660259481969, + "loss": 0.8126, + "step": 3185 + }, + { + "epoch": 0.22108111858996599, + "grad_norm": 2.9955718517303467, + "learning_rate": 0.00019966556697571894, + "loss": 1.3323, + "step": 3186 + }, + { + "epoch": 0.22115051002706265, + "grad_norm": 3.4663872718811035, + "learning_rate": 0.00019966510768860838, + "loss": 1.0877, + "step": 3187 + }, + { + "epoch": 0.22121990146415932, + "grad_norm": 3.185986280441284, + "learning_rate": 0.0001996646480868666, + "loss": 0.9646, + "step": 3188 + }, + { + "epoch": 0.22128929290125599, + "grad_norm": 4.201290607452393, + "learning_rate": 0.00019966418817049511, + "loss": 1.1814, + "step": 3189 + }, + { + "epoch": 0.22135868433835265, + "grad_norm": 3.006370782852173, + "learning_rate": 0.00019966372793949532, + "loss": 1.2959, + "step": 3190 + }, + { + "epoch": 0.22142807577544932, + "grad_norm": 3.8924708366394043, + "learning_rate": 0.00019966326739386873, + "loss": 0.8416, + "step": 3191 + }, + { + "epoch": 0.22149746721254598, + "grad_norm": 2.986753463745117, + "learning_rate": 0.00019966280653361678, + "loss": 0.9077, + "step": 3192 + }, + { + "epoch": 0.22156685864964262, + "grad_norm": 4.041496276855469, + "learning_rate": 0.0001996623453587409, + "loss": 1.1016, + "step": 3193 + }, + { + "epoch": 0.2216362500867393, + "grad_norm": 3.4307875633239746, + "learning_rate": 0.00019966188386924254, + "loss": 0.7563, + "step": 3194 + }, + { + "epoch": 0.22170564152383596, + "grad_norm": 5.108754634857178, + "learning_rate": 0.0001996614220651232, + "loss": 1.3269, + "step": 3195 + }, + { + "epoch": 0.22177503296093262, + "grad_norm": 5.8633952140808105, + "learning_rate": 0.0001996609599463843, + "loss": 1.1918, + "step": 3196 + }, + { + "epoch": 0.2218444243980293, + "grad_norm": 4.362759113311768, + "learning_rate": 0.0001996604975130273, + "loss": 1.3979, + "step": 3197 + }, + { + "epoch": 0.22191381583512595, + "grad_norm": 3.662135362625122, + "learning_rate": 0.00019966003476505368, + "loss": 1.041, + "step": 3198 + }, + { + "epoch": 0.22198320727222262, + "grad_norm": 4.040679931640625, + "learning_rate": 0.00019965957170246492, + "loss": 1.3173, + "step": 3199 + }, + { + "epoch": 0.22205259870931926, + "grad_norm": 4.1544365882873535, + "learning_rate": 0.00019965910832526242, + "loss": 1.0604, + "step": 3200 + }, + { + "epoch": 0.22212199014641593, + "grad_norm": 4.1971025466918945, + "learning_rate": 0.0001996586446334477, + "loss": 1.1548, + "step": 3201 + }, + { + "epoch": 0.2221913815835126, + "grad_norm": 4.741885185241699, + "learning_rate": 0.00019965818062702217, + "loss": 0.7873, + "step": 3202 + }, + { + "epoch": 0.22226077302060926, + "grad_norm": 4.240386962890625, + "learning_rate": 0.00019965771630598735, + "loss": 1.0703, + "step": 3203 + }, + { + "epoch": 0.22233016445770593, + "grad_norm": 4.803123950958252, + "learning_rate": 0.0001996572516703447, + "loss": 1.2736, + "step": 3204 + }, + { + "epoch": 0.2223995558948026, + "grad_norm": 3.419869899749756, + "learning_rate": 0.00019965678672009566, + "loss": 1.0693, + "step": 3205 + }, + { + "epoch": 0.22246894733189923, + "grad_norm": 4.376048564910889, + "learning_rate": 0.0001996563214552417, + "loss": 1.1458, + "step": 3206 + }, + { + "epoch": 0.2225383387689959, + "grad_norm": 3.6780426502227783, + "learning_rate": 0.00019965585587578428, + "loss": 1.1936, + "step": 3207 + }, + { + "epoch": 0.22260773020609256, + "grad_norm": 3.5695643424987793, + "learning_rate": 0.0001996553899817249, + "loss": 0.9922, + "step": 3208 + }, + { + "epoch": 0.22267712164318923, + "grad_norm": 3.0388453006744385, + "learning_rate": 0.000199654923773065, + "loss": 0.9192, + "step": 3209 + }, + { + "epoch": 0.2227465130802859, + "grad_norm": 4.1581621170043945, + "learning_rate": 0.0001996544572498061, + "loss": 1.0555, + "step": 3210 + }, + { + "epoch": 0.22281590451738256, + "grad_norm": 3.6749379634857178, + "learning_rate": 0.0001996539904119496, + "loss": 1.2068, + "step": 3211 + }, + { + "epoch": 0.22288529595447923, + "grad_norm": 3.1380999088287354, + "learning_rate": 0.00019965352325949706, + "loss": 0.9913, + "step": 3212 + }, + { + "epoch": 0.22295468739157587, + "grad_norm": 3.296942710876465, + "learning_rate": 0.00019965305579244986, + "loss": 1.0399, + "step": 3213 + }, + { + "epoch": 0.22302407882867253, + "grad_norm": 4.543742656707764, + "learning_rate": 0.00019965258801080954, + "loss": 1.374, + "step": 3214 + }, + { + "epoch": 0.2230934702657692, + "grad_norm": 3.4196383953094482, + "learning_rate": 0.0001996521199145776, + "loss": 1.0506, + "step": 3215 + }, + { + "epoch": 0.22316286170286587, + "grad_norm": 4.902700901031494, + "learning_rate": 0.00019965165150375545, + "loss": 1.2472, + "step": 3216 + }, + { + "epoch": 0.22323225313996253, + "grad_norm": 3.800143241882324, + "learning_rate": 0.0001996511827783446, + "loss": 1.1191, + "step": 3217 + }, + { + "epoch": 0.2233016445770592, + "grad_norm": 3.946549892425537, + "learning_rate": 0.00019965071373834653, + "loss": 1.2186, + "step": 3218 + }, + { + "epoch": 0.22337103601415587, + "grad_norm": 3.3203988075256348, + "learning_rate": 0.0001996502443837627, + "loss": 1.0009, + "step": 3219 + }, + { + "epoch": 0.2234404274512525, + "grad_norm": 3.4002506732940674, + "learning_rate": 0.00019964977471459463, + "loss": 0.8834, + "step": 3220 + }, + { + "epoch": 0.22350981888834917, + "grad_norm": 4.06787633895874, + "learning_rate": 0.0001996493047308438, + "loss": 1.1177, + "step": 3221 + }, + { + "epoch": 0.22357921032544584, + "grad_norm": 3.9321062564849854, + "learning_rate": 0.00019964883443251164, + "loss": 1.3676, + "step": 3222 + }, + { + "epoch": 0.2236486017625425, + "grad_norm": 3.4594078063964844, + "learning_rate": 0.0001996483638195997, + "loss": 1.0952, + "step": 3223 + }, + { + "epoch": 0.22371799319963917, + "grad_norm": 3.916213274002075, + "learning_rate": 0.0001996478928921094, + "loss": 1.2393, + "step": 3224 + }, + { + "epoch": 0.22378738463673584, + "grad_norm": 3.943495035171509, + "learning_rate": 0.00019964742165004228, + "loss": 1.3149, + "step": 3225 + }, + { + "epoch": 0.22385677607383248, + "grad_norm": 3.400681257247925, + "learning_rate": 0.00019964695009339983, + "loss": 1.0292, + "step": 3226 + }, + { + "epoch": 0.22392616751092914, + "grad_norm": 2.8291468620300293, + "learning_rate": 0.0001996464782221835, + "loss": 1.0757, + "step": 3227 + }, + { + "epoch": 0.2239955589480258, + "grad_norm": 3.279057502746582, + "learning_rate": 0.00019964600603639482, + "loss": 0.8914, + "step": 3228 + }, + { + "epoch": 0.22406495038512247, + "grad_norm": 5.4898810386657715, + "learning_rate": 0.0001996455335360352, + "loss": 1.0737, + "step": 3229 + }, + { + "epoch": 0.22413434182221914, + "grad_norm": 3.8515994548797607, + "learning_rate": 0.00019964506072110626, + "loss": 1.0662, + "step": 3230 + }, + { + "epoch": 0.2242037332593158, + "grad_norm": 3.381058692932129, + "learning_rate": 0.0001996445875916094, + "loss": 1.2557, + "step": 3231 + }, + { + "epoch": 0.22427312469641247, + "grad_norm": 4.00201416015625, + "learning_rate": 0.00019964411414754617, + "loss": 1.2738, + "step": 3232 + }, + { + "epoch": 0.2243425161335091, + "grad_norm": 3.394080877304077, + "learning_rate": 0.00019964364038891797, + "loss": 0.8897, + "step": 3233 + }, + { + "epoch": 0.22441190757060578, + "grad_norm": 3.42189884185791, + "learning_rate": 0.00019964316631572638, + "loss": 1.0325, + "step": 3234 + }, + { + "epoch": 0.22448129900770245, + "grad_norm": 4.0525383949279785, + "learning_rate": 0.00019964269192797288, + "loss": 1.4225, + "step": 3235 + }, + { + "epoch": 0.2245506904447991, + "grad_norm": 3.7687509059906006, + "learning_rate": 0.000199642217225659, + "loss": 0.9121, + "step": 3236 + }, + { + "epoch": 0.22462008188189578, + "grad_norm": 7.598371505737305, + "learning_rate": 0.00019964174220878615, + "loss": 1.2522, + "step": 3237 + }, + { + "epoch": 0.22468947331899244, + "grad_norm": 3.0130982398986816, + "learning_rate": 0.0001996412668773559, + "loss": 0.9659, + "step": 3238 + }, + { + "epoch": 0.2247588647560891, + "grad_norm": 3.5470986366271973, + "learning_rate": 0.0001996407912313697, + "loss": 1.067, + "step": 3239 + }, + { + "epoch": 0.22482825619318575, + "grad_norm": 19.720211029052734, + "learning_rate": 0.00019964031527082914, + "loss": 1.1791, + "step": 3240 + }, + { + "epoch": 0.22489764763028242, + "grad_norm": 4.184660911560059, + "learning_rate": 0.00019963983899573564, + "loss": 1.1057, + "step": 3241 + }, + { + "epoch": 0.22496703906737908, + "grad_norm": 3.494917154312134, + "learning_rate": 0.00019963936240609072, + "loss": 1.0809, + "step": 3242 + }, + { + "epoch": 0.22503643050447575, + "grad_norm": 2.873105049133301, + "learning_rate": 0.00019963888550189588, + "loss": 0.9876, + "step": 3243 + }, + { + "epoch": 0.22510582194157241, + "grad_norm": 4.351190090179443, + "learning_rate": 0.00019963840828315268, + "loss": 1.0763, + "step": 3244 + }, + { + "epoch": 0.22517521337866908, + "grad_norm": 4.814836025238037, + "learning_rate": 0.00019963793074986256, + "loss": 0.929, + "step": 3245 + }, + { + "epoch": 0.22524460481576575, + "grad_norm": 3.2266433238983154, + "learning_rate": 0.00019963745290202707, + "loss": 0.9246, + "step": 3246 + }, + { + "epoch": 0.2253139962528624, + "grad_norm": 5.989284515380859, + "learning_rate": 0.0001996369747396477, + "loss": 1.0659, + "step": 3247 + }, + { + "epoch": 0.22538338768995905, + "grad_norm": 4.438343048095703, + "learning_rate": 0.00019963649626272592, + "loss": 1.1131, + "step": 3248 + }, + { + "epoch": 0.22545277912705572, + "grad_norm": 3.482283592224121, + "learning_rate": 0.0001996360174712633, + "loss": 0.9944, + "step": 3249 + }, + { + "epoch": 0.22552217056415239, + "grad_norm": 4.665992736816406, + "learning_rate": 0.00019963553836526137, + "loss": 1.0649, + "step": 3250 + }, + { + "epoch": 0.22559156200124905, + "grad_norm": 4.991840362548828, + "learning_rate": 0.00019963505894472156, + "loss": 1.0538, + "step": 3251 + }, + { + "epoch": 0.22566095343834572, + "grad_norm": 5.04885721206665, + "learning_rate": 0.00019963457920964546, + "loss": 1.0957, + "step": 3252 + }, + { + "epoch": 0.22573034487544236, + "grad_norm": 3.0655243396759033, + "learning_rate": 0.0001996340991600345, + "loss": 1.1352, + "step": 3253 + }, + { + "epoch": 0.22579973631253902, + "grad_norm": 4.624704360961914, + "learning_rate": 0.00019963361879589032, + "loss": 1.202, + "step": 3254 + }, + { + "epoch": 0.2258691277496357, + "grad_norm": 6.702390670776367, + "learning_rate": 0.00019963313811721434, + "loss": 1.0345, + "step": 3255 + }, + { + "epoch": 0.22593851918673236, + "grad_norm": 4.138983726501465, + "learning_rate": 0.00019963265712400808, + "loss": 1.0557, + "step": 3256 + }, + { + "epoch": 0.22600791062382902, + "grad_norm": 4.121763706207275, + "learning_rate": 0.0001996321758162731, + "loss": 1.2625, + "step": 3257 + }, + { + "epoch": 0.2260773020609257, + "grad_norm": 4.100123405456543, + "learning_rate": 0.0001996316941940109, + "loss": 0.914, + "step": 3258 + }, + { + "epoch": 0.22614669349802236, + "grad_norm": 5.4231791496276855, + "learning_rate": 0.000199631212257223, + "loss": 1.3084, + "step": 3259 + }, + { + "epoch": 0.226216084935119, + "grad_norm": 6.997437000274658, + "learning_rate": 0.0001996307300059109, + "loss": 0.9511, + "step": 3260 + }, + { + "epoch": 0.22628547637221566, + "grad_norm": 4.465384006500244, + "learning_rate": 0.00019963024744007617, + "loss": 1.4575, + "step": 3261 + }, + { + "epoch": 0.22635486780931233, + "grad_norm": 8.870593070983887, + "learning_rate": 0.0001996297645597203, + "loss": 1.1518, + "step": 3262 + }, + { + "epoch": 0.226424259246409, + "grad_norm": 7.600251197814941, + "learning_rate": 0.00019962928136484483, + "loss": 1.3123, + "step": 3263 + }, + { + "epoch": 0.22649365068350566, + "grad_norm": 9.534591674804688, + "learning_rate": 0.00019962879785545128, + "loss": 1.3229, + "step": 3264 + }, + { + "epoch": 0.22656304212060233, + "grad_norm": 3.9728777408599854, + "learning_rate": 0.00019962831403154115, + "loss": 0.8093, + "step": 3265 + }, + { + "epoch": 0.226632433557699, + "grad_norm": 4.744974613189697, + "learning_rate": 0.00019962782989311602, + "loss": 1.2461, + "step": 3266 + }, + { + "epoch": 0.22670182499479563, + "grad_norm": 9.700112342834473, + "learning_rate": 0.00019962734544017735, + "loss": 1.6894, + "step": 3267 + }, + { + "epoch": 0.2267712164318923, + "grad_norm": 5.934017658233643, + "learning_rate": 0.00019962686067272675, + "loss": 1.1963, + "step": 3268 + }, + { + "epoch": 0.22684060786898896, + "grad_norm": 9.753150939941406, + "learning_rate": 0.00019962637559076569, + "loss": 1.2122, + "step": 3269 + }, + { + "epoch": 0.22690999930608563, + "grad_norm": 6.970123767852783, + "learning_rate": 0.00019962589019429573, + "loss": 1.5826, + "step": 3270 + }, + { + "epoch": 0.2269793907431823, + "grad_norm": 5.063417434692383, + "learning_rate": 0.00019962540448331838, + "loss": 1.2464, + "step": 3271 + }, + { + "epoch": 0.22704878218027896, + "grad_norm": 7.641640663146973, + "learning_rate": 0.00019962491845783523, + "loss": 1.2293, + "step": 3272 + }, + { + "epoch": 0.22711817361737563, + "grad_norm": 5.357052803039551, + "learning_rate": 0.00019962443211784771, + "loss": 1.1744, + "step": 3273 + }, + { + "epoch": 0.22718756505447227, + "grad_norm": 4.156831741333008, + "learning_rate": 0.00019962394546335746, + "loss": 1.2686, + "step": 3274 + }, + { + "epoch": 0.22725695649156893, + "grad_norm": 4.080024242401123, + "learning_rate": 0.00019962345849436596, + "loss": 1.2043, + "step": 3275 + }, + { + "epoch": 0.2273263479286656, + "grad_norm": 4.088179111480713, + "learning_rate": 0.00019962297121087476, + "loss": 1.355, + "step": 3276 + }, + { + "epoch": 0.22739573936576227, + "grad_norm": 3.930896520614624, + "learning_rate": 0.0001996224836128854, + "loss": 1.2432, + "step": 3277 + }, + { + "epoch": 0.22746513080285893, + "grad_norm": 3.7678072452545166, + "learning_rate": 0.00019962199570039941, + "loss": 1.0915, + "step": 3278 + }, + { + "epoch": 0.2275345222399556, + "grad_norm": 4.806376934051514, + "learning_rate": 0.00019962150747341835, + "loss": 1.0781, + "step": 3279 + }, + { + "epoch": 0.22760391367705224, + "grad_norm": 4.497867107391357, + "learning_rate": 0.00019962101893194375, + "loss": 1.3559, + "step": 3280 + }, + { + "epoch": 0.2276733051141489, + "grad_norm": 4.81323766708374, + "learning_rate": 0.0001996205300759771, + "loss": 1.2688, + "step": 3281 + }, + { + "epoch": 0.22774269655124557, + "grad_norm": 3.6768157482147217, + "learning_rate": 0.00019962004090552006, + "loss": 1.1465, + "step": 3282 + }, + { + "epoch": 0.22781208798834224, + "grad_norm": 6.172396659851074, + "learning_rate": 0.00019961955142057408, + "loss": 1.3427, + "step": 3283 + }, + { + "epoch": 0.2278814794254389, + "grad_norm": 3.21452260017395, + "learning_rate": 0.00019961906162114078, + "loss": 1.048, + "step": 3284 + }, + { + "epoch": 0.22795087086253557, + "grad_norm": 5.753089904785156, + "learning_rate": 0.00019961857150722158, + "loss": 1.162, + "step": 3285 + }, + { + "epoch": 0.22802026229963224, + "grad_norm": 3.924358367919922, + "learning_rate": 0.00019961808107881818, + "loss": 1.1436, + "step": 3286 + }, + { + "epoch": 0.22808965373672888, + "grad_norm": 5.523597717285156, + "learning_rate": 0.00019961759033593202, + "loss": 1.4006, + "step": 3287 + }, + { + "epoch": 0.22815904517382554, + "grad_norm": 3.6589152812957764, + "learning_rate": 0.0001996170992785647, + "loss": 1.1928, + "step": 3288 + }, + { + "epoch": 0.2282284366109222, + "grad_norm": 5.430599212646484, + "learning_rate": 0.00019961660790671773, + "loss": 1.2294, + "step": 3289 + }, + { + "epoch": 0.22829782804801887, + "grad_norm": 4.693499565124512, + "learning_rate": 0.0001996161162203927, + "loss": 1.0475, + "step": 3290 + }, + { + "epoch": 0.22836721948511554, + "grad_norm": 4.1239824295043945, + "learning_rate": 0.00019961562421959116, + "loss": 1.17, + "step": 3291 + }, + { + "epoch": 0.2284366109222122, + "grad_norm": 3.4579732418060303, + "learning_rate": 0.0001996151319043146, + "loss": 1.2952, + "step": 3292 + }, + { + "epoch": 0.22850600235930887, + "grad_norm": 5.268435001373291, + "learning_rate": 0.00019961463927456468, + "loss": 1.1986, + "step": 3293 + }, + { + "epoch": 0.2285753937964055, + "grad_norm": 3.429006338119507, + "learning_rate": 0.0001996141463303429, + "loss": 0.9321, + "step": 3294 + }, + { + "epoch": 0.22864478523350218, + "grad_norm": 4.680913925170898, + "learning_rate": 0.0001996136530716508, + "loss": 1.3813, + "step": 3295 + }, + { + "epoch": 0.22871417667059885, + "grad_norm": 3.566216230392456, + "learning_rate": 0.00019961315949848998, + "loss": 0.8505, + "step": 3296 + }, + { + "epoch": 0.2287835681076955, + "grad_norm": 2.832744836807251, + "learning_rate": 0.00019961266561086195, + "loss": 1.0537, + "step": 3297 + }, + { + "epoch": 0.22885295954479218, + "grad_norm": 3.1489763259887695, + "learning_rate": 0.00019961217140876828, + "loss": 1.0921, + "step": 3298 + }, + { + "epoch": 0.22892235098188884, + "grad_norm": 4.968688488006592, + "learning_rate": 0.0001996116768922106, + "loss": 1.307, + "step": 3299 + }, + { + "epoch": 0.22899174241898548, + "grad_norm": 3.7392730712890625, + "learning_rate": 0.00019961118206119033, + "loss": 1.0298, + "step": 3300 + }, + { + "epoch": 0.22906113385608215, + "grad_norm": 4.700845718383789, + "learning_rate": 0.00019961068691570918, + "loss": 1.2179, + "step": 3301 + }, + { + "epoch": 0.22913052529317882, + "grad_norm": 3.9930269718170166, + "learning_rate": 0.0001996101914557686, + "loss": 1.4633, + "step": 3302 + }, + { + "epoch": 0.22919991673027548, + "grad_norm": 3.4001405239105225, + "learning_rate": 0.0001996096956813702, + "loss": 1.1445, + "step": 3303 + }, + { + "epoch": 0.22926930816737215, + "grad_norm": 3.205134391784668, + "learning_rate": 0.00019960919959251562, + "loss": 0.9754, + "step": 3304 + }, + { + "epoch": 0.22933869960446882, + "grad_norm": 3.4887702465057373, + "learning_rate": 0.0001996087031892063, + "loss": 1.1047, + "step": 3305 + }, + { + "epoch": 0.22940809104156548, + "grad_norm": 3.6983931064605713, + "learning_rate": 0.00019960820647144388, + "loss": 1.0594, + "step": 3306 + }, + { + "epoch": 0.22947748247866212, + "grad_norm": 4.160669326782227, + "learning_rate": 0.0001996077094392299, + "loss": 1.159, + "step": 3307 + }, + { + "epoch": 0.2295468739157588, + "grad_norm": 3.8099117279052734, + "learning_rate": 0.00019960721209256592, + "loss": 0.9161, + "step": 3308 + }, + { + "epoch": 0.22961626535285545, + "grad_norm": 3.6551921367645264, + "learning_rate": 0.00019960671443145354, + "loss": 1.1308, + "step": 3309 + }, + { + "epoch": 0.22968565678995212, + "grad_norm": 5.040383338928223, + "learning_rate": 0.00019960621645589432, + "loss": 1.3001, + "step": 3310 + }, + { + "epoch": 0.22975504822704879, + "grad_norm": 3.7836098670959473, + "learning_rate": 0.00019960571816588985, + "loss": 1.2538, + "step": 3311 + }, + { + "epoch": 0.22982443966414545, + "grad_norm": 2.995384693145752, + "learning_rate": 0.00019960521956144167, + "loss": 0.9248, + "step": 3312 + }, + { + "epoch": 0.22989383110124212, + "grad_norm": 4.838963508605957, + "learning_rate": 0.00019960472064255137, + "loss": 1.0992, + "step": 3313 + }, + { + "epoch": 0.22996322253833876, + "grad_norm": 3.714815139770508, + "learning_rate": 0.0001996042214092205, + "loss": 1.1862, + "step": 3314 + }, + { + "epoch": 0.23003261397543542, + "grad_norm": 4.879833221435547, + "learning_rate": 0.00019960372186145068, + "loss": 1.2756, + "step": 3315 + }, + { + "epoch": 0.2301020054125321, + "grad_norm": 4.333714485168457, + "learning_rate": 0.00019960322199924347, + "loss": 1.3318, + "step": 3316 + }, + { + "epoch": 0.23017139684962876, + "grad_norm": 5.378793716430664, + "learning_rate": 0.00019960272182260045, + "loss": 1.2107, + "step": 3317 + }, + { + "epoch": 0.23024078828672542, + "grad_norm": 3.96059250831604, + "learning_rate": 0.00019960222133152318, + "loss": 0.9591, + "step": 3318 + }, + { + "epoch": 0.2303101797238221, + "grad_norm": 3.4309723377227783, + "learning_rate": 0.00019960172052601327, + "loss": 1.2013, + "step": 3319 + }, + { + "epoch": 0.23037957116091876, + "grad_norm": 3.601592779159546, + "learning_rate": 0.00019960121940607226, + "loss": 0.976, + "step": 3320 + }, + { + "epoch": 0.2304489625980154, + "grad_norm": 4.598283767700195, + "learning_rate": 0.00019960071797170176, + "loss": 1.3215, + "step": 3321 + }, + { + "epoch": 0.23051835403511206, + "grad_norm": 5.07687520980835, + "learning_rate": 0.00019960021622290338, + "loss": 1.2086, + "step": 3322 + }, + { + "epoch": 0.23058774547220873, + "grad_norm": 3.2937543392181396, + "learning_rate": 0.00019959971415967864, + "loss": 1.1006, + "step": 3323 + }, + { + "epoch": 0.2306571369093054, + "grad_norm": 4.361332893371582, + "learning_rate": 0.00019959921178202915, + "loss": 1.7069, + "step": 3324 + }, + { + "epoch": 0.23072652834640206, + "grad_norm": 4.39301872253418, + "learning_rate": 0.00019959870908995653, + "loss": 1.0911, + "step": 3325 + }, + { + "epoch": 0.23079591978349873, + "grad_norm": 3.7755794525146484, + "learning_rate": 0.00019959820608346233, + "loss": 0.9148, + "step": 3326 + }, + { + "epoch": 0.23086531122059537, + "grad_norm": 2.6079862117767334, + "learning_rate": 0.00019959770276254814, + "loss": 0.8631, + "step": 3327 + }, + { + "epoch": 0.23093470265769203, + "grad_norm": 3.5683999061584473, + "learning_rate": 0.00019959719912721556, + "loss": 1.2828, + "step": 3328 + }, + { + "epoch": 0.2310040940947887, + "grad_norm": 4.74910831451416, + "learning_rate": 0.0001995966951774662, + "loss": 1.2715, + "step": 3329 + }, + { + "epoch": 0.23107348553188536, + "grad_norm": 4.067714214324951, + "learning_rate": 0.00019959619091330162, + "loss": 1.2039, + "step": 3330 + }, + { + "epoch": 0.23114287696898203, + "grad_norm": 3.5243093967437744, + "learning_rate": 0.00019959568633472343, + "loss": 1.0498, + "step": 3331 + }, + { + "epoch": 0.2312122684060787, + "grad_norm": 3.3855268955230713, + "learning_rate": 0.00019959518144173317, + "loss": 1.0473, + "step": 3332 + }, + { + "epoch": 0.23128165984317536, + "grad_norm": 3.8902363777160645, + "learning_rate": 0.00019959467623433251, + "loss": 1.1013, + "step": 3333 + }, + { + "epoch": 0.231351051280272, + "grad_norm": 3.294043779373169, + "learning_rate": 0.00019959417071252302, + "loss": 1.1269, + "step": 3334 + }, + { + "epoch": 0.23142044271736867, + "grad_norm": 4.002228260040283, + "learning_rate": 0.0001995936648763063, + "loss": 1.2968, + "step": 3335 + }, + { + "epoch": 0.23148983415446533, + "grad_norm": 3.668818235397339, + "learning_rate": 0.00019959315872568388, + "loss": 1.1001, + "step": 3336 + }, + { + "epoch": 0.231559225591562, + "grad_norm": 5.527250289916992, + "learning_rate": 0.00019959265226065744, + "loss": 1.0913, + "step": 3337 + }, + { + "epoch": 0.23162861702865867, + "grad_norm": 3.3967788219451904, + "learning_rate": 0.00019959214548122858, + "loss": 1.321, + "step": 3338 + }, + { + "epoch": 0.23169800846575533, + "grad_norm": 3.8015782833099365, + "learning_rate": 0.00019959163838739885, + "loss": 1.0401, + "step": 3339 + }, + { + "epoch": 0.231767399902852, + "grad_norm": 4.014288425445557, + "learning_rate": 0.00019959113097916986, + "loss": 1.2009, + "step": 3340 + }, + { + "epoch": 0.23183679133994864, + "grad_norm": 3.4025652408599854, + "learning_rate": 0.00019959062325654322, + "loss": 1.2399, + "step": 3341 + }, + { + "epoch": 0.2319061827770453, + "grad_norm": 3.4481050968170166, + "learning_rate": 0.0001995901152195206, + "loss": 1.1363, + "step": 3342 + }, + { + "epoch": 0.23197557421414197, + "grad_norm": 4.001935958862305, + "learning_rate": 0.00019958960686810348, + "loss": 0.8138, + "step": 3343 + }, + { + "epoch": 0.23204496565123864, + "grad_norm": 3.220085620880127, + "learning_rate": 0.00019958909820229354, + "loss": 1.0131, + "step": 3344 + }, + { + "epoch": 0.2321143570883353, + "grad_norm": 3.20125150680542, + "learning_rate": 0.00019958858922209234, + "loss": 0.9715, + "step": 3345 + }, + { + "epoch": 0.23218374852543197, + "grad_norm": 3.89939022064209, + "learning_rate": 0.00019958807992750158, + "loss": 1.0177, + "step": 3346 + }, + { + "epoch": 0.23225313996252864, + "grad_norm": 3.875528335571289, + "learning_rate": 0.00019958757031852273, + "loss": 0.9484, + "step": 3347 + }, + { + "epoch": 0.23232253139962528, + "grad_norm": 4.10870361328125, + "learning_rate": 0.00019958706039515753, + "loss": 1.1356, + "step": 3348 + }, + { + "epoch": 0.23239192283672194, + "grad_norm": 3.080613851547241, + "learning_rate": 0.00019958655015740753, + "loss": 1.0636, + "step": 3349 + }, + { + "epoch": 0.2324613142738186, + "grad_norm": 4.376034736633301, + "learning_rate": 0.00019958603960527432, + "loss": 0.9748, + "step": 3350 + }, + { + "epoch": 0.23253070571091528, + "grad_norm": 3.7981717586517334, + "learning_rate": 0.00019958552873875957, + "loss": 1.1312, + "step": 3351 + }, + { + "epoch": 0.23260009714801194, + "grad_norm": 3.2829980850219727, + "learning_rate": 0.00019958501755786484, + "loss": 0.9835, + "step": 3352 + }, + { + "epoch": 0.2326694885851086, + "grad_norm": 3.7914576530456543, + "learning_rate": 0.00019958450606259178, + "loss": 1.1584, + "step": 3353 + }, + { + "epoch": 0.23273888002220525, + "grad_norm": 3.3325507640838623, + "learning_rate": 0.00019958399425294197, + "loss": 1.0741, + "step": 3354 + }, + { + "epoch": 0.2328082714593019, + "grad_norm": 4.041749954223633, + "learning_rate": 0.00019958348212891703, + "loss": 0.8301, + "step": 3355 + }, + { + "epoch": 0.23287766289639858, + "grad_norm": 4.881182670593262, + "learning_rate": 0.00019958296969051865, + "loss": 1.0127, + "step": 3356 + }, + { + "epoch": 0.23294705433349525, + "grad_norm": 3.9431467056274414, + "learning_rate": 0.00019958245693774834, + "loss": 1.0322, + "step": 3357 + }, + { + "epoch": 0.2330164457705919, + "grad_norm": 3.545780897140503, + "learning_rate": 0.00019958194387060778, + "loss": 1.0519, + "step": 3358 + }, + { + "epoch": 0.23308583720768858, + "grad_norm": 3.0983262062072754, + "learning_rate": 0.00019958143048909858, + "loss": 1.0047, + "step": 3359 + }, + { + "epoch": 0.23315522864478525, + "grad_norm": 3.345020294189453, + "learning_rate": 0.00019958091679322235, + "loss": 0.9492, + "step": 3360 + }, + { + "epoch": 0.23322462008188188, + "grad_norm": 3.8966403007507324, + "learning_rate": 0.00019958040278298073, + "loss": 1.4204, + "step": 3361 + }, + { + "epoch": 0.23329401151897855, + "grad_norm": 4.73689079284668, + "learning_rate": 0.00019957988845837533, + "loss": 1.2589, + "step": 3362 + }, + { + "epoch": 0.23336340295607522, + "grad_norm": 6.009547710418701, + "learning_rate": 0.00019957937381940774, + "loss": 1.2953, + "step": 3363 + }, + { + "epoch": 0.23343279439317188, + "grad_norm": 3.44488263130188, + "learning_rate": 0.00019957885886607966, + "loss": 1.0725, + "step": 3364 + }, + { + "epoch": 0.23350218583026855, + "grad_norm": 3.548105001449585, + "learning_rate": 0.00019957834359839266, + "loss": 1.1151, + "step": 3365 + }, + { + "epoch": 0.23357157726736522, + "grad_norm": 3.359855890274048, + "learning_rate": 0.0001995778280163484, + "loss": 1.2154, + "step": 3366 + }, + { + "epoch": 0.23364096870446188, + "grad_norm": 2.8362011909484863, + "learning_rate": 0.00019957731211994845, + "loss": 0.8356, + "step": 3367 + }, + { + "epoch": 0.23371036014155852, + "grad_norm": 3.1658775806427, + "learning_rate": 0.00019957679590919454, + "loss": 0.9974, + "step": 3368 + }, + { + "epoch": 0.2337797515786552, + "grad_norm": 4.323858261108398, + "learning_rate": 0.00019957627938408819, + "loss": 1.0046, + "step": 3369 + }, + { + "epoch": 0.23384914301575185, + "grad_norm": 3.520967721939087, + "learning_rate": 0.0001995757625446311, + "loss": 0.982, + "step": 3370 + }, + { + "epoch": 0.23391853445284852, + "grad_norm": 4.4618706703186035, + "learning_rate": 0.00019957524539082486, + "loss": 1.0025, + "step": 3371 + }, + { + "epoch": 0.2339879258899452, + "grad_norm": 3.754770278930664, + "learning_rate": 0.00019957472792267114, + "loss": 1.1363, + "step": 3372 + }, + { + "epoch": 0.23405731732704185, + "grad_norm": 3.8541977405548096, + "learning_rate": 0.00019957421014017154, + "loss": 1.1387, + "step": 3373 + }, + { + "epoch": 0.23412670876413852, + "grad_norm": 3.607874870300293, + "learning_rate": 0.00019957369204332774, + "loss": 0.9011, + "step": 3374 + }, + { + "epoch": 0.23419610020123516, + "grad_norm": 3.880535125732422, + "learning_rate": 0.00019957317363214133, + "loss": 1.1172, + "step": 3375 + }, + { + "epoch": 0.23426549163833182, + "grad_norm": 3.568010091781616, + "learning_rate": 0.00019957265490661396, + "loss": 0.8804, + "step": 3376 + }, + { + "epoch": 0.2343348830754285, + "grad_norm": 4.106049537658691, + "learning_rate": 0.0001995721358667473, + "loss": 1.015, + "step": 3377 + }, + { + "epoch": 0.23440427451252516, + "grad_norm": 4.584962368011475, + "learning_rate": 0.00019957161651254293, + "loss": 1.276, + "step": 3378 + }, + { + "epoch": 0.23447366594962182, + "grad_norm": 3.517829656600952, + "learning_rate": 0.0001995710968440025, + "loss": 1.0537, + "step": 3379 + }, + { + "epoch": 0.2345430573867185, + "grad_norm": 3.3805134296417236, + "learning_rate": 0.0001995705768611277, + "loss": 0.9104, + "step": 3380 + }, + { + "epoch": 0.23461244882381513, + "grad_norm": 6.258752346038818, + "learning_rate": 0.00019957005656392015, + "loss": 1.2012, + "step": 3381 + }, + { + "epoch": 0.2346818402609118, + "grad_norm": 3.0841176509857178, + "learning_rate": 0.0001995695359523815, + "loss": 1.0723, + "step": 3382 + }, + { + "epoch": 0.23475123169800846, + "grad_norm": 4.519922733306885, + "learning_rate": 0.00019956901502651332, + "loss": 1.0851, + "step": 3383 + }, + { + "epoch": 0.23482062313510513, + "grad_norm": 3.2723135948181152, + "learning_rate": 0.00019956849378631732, + "loss": 1.0168, + "step": 3384 + }, + { + "epoch": 0.2348900145722018, + "grad_norm": 3.765976905822754, + "learning_rate": 0.00019956797223179517, + "loss": 1.04, + "step": 3385 + }, + { + "epoch": 0.23495940600929846, + "grad_norm": 3.5118038654327393, + "learning_rate": 0.00019956745036294845, + "loss": 1.0537, + "step": 3386 + }, + { + "epoch": 0.23502879744639513, + "grad_norm": 3.2976317405700684, + "learning_rate": 0.00019956692817977887, + "loss": 1.0985, + "step": 3387 + }, + { + "epoch": 0.23509818888349177, + "grad_norm": 4.160938262939453, + "learning_rate": 0.00019956640568228804, + "loss": 1.1666, + "step": 3388 + }, + { + "epoch": 0.23516758032058843, + "grad_norm": 3.204685688018799, + "learning_rate": 0.00019956588287047764, + "loss": 1.15, + "step": 3389 + }, + { + "epoch": 0.2352369717576851, + "grad_norm": 4.963780879974365, + "learning_rate": 0.00019956535974434926, + "loss": 1.1609, + "step": 3390 + }, + { + "epoch": 0.23530636319478176, + "grad_norm": 3.9945311546325684, + "learning_rate": 0.0001995648363039046, + "loss": 0.9173, + "step": 3391 + }, + { + "epoch": 0.23537575463187843, + "grad_norm": 4.194967269897461, + "learning_rate": 0.00019956431254914534, + "loss": 1.056, + "step": 3392 + }, + { + "epoch": 0.2354451460689751, + "grad_norm": 3.481981039047241, + "learning_rate": 0.00019956378848007308, + "loss": 1.0631, + "step": 3393 + }, + { + "epoch": 0.23551453750607176, + "grad_norm": 3.5536296367645264, + "learning_rate": 0.0001995632640966895, + "loss": 1.144, + "step": 3394 + }, + { + "epoch": 0.2355839289431684, + "grad_norm": 4.797996520996094, + "learning_rate": 0.00019956273939899618, + "loss": 1.2935, + "step": 3395 + }, + { + "epoch": 0.23565332038026507, + "grad_norm": 7.433028697967529, + "learning_rate": 0.0001995622143869949, + "loss": 1.0738, + "step": 3396 + }, + { + "epoch": 0.23572271181736174, + "grad_norm": 3.9147756099700928, + "learning_rate": 0.00019956168906068726, + "loss": 1.0659, + "step": 3397 + }, + { + "epoch": 0.2357921032544584, + "grad_norm": 4.457418918609619, + "learning_rate": 0.00019956116342007493, + "loss": 0.9986, + "step": 3398 + }, + { + "epoch": 0.23586149469155507, + "grad_norm": 3.665029525756836, + "learning_rate": 0.0001995606374651595, + "loss": 1.1662, + "step": 3399 + }, + { + "epoch": 0.23593088612865173, + "grad_norm": 8.223603248596191, + "learning_rate": 0.00019956011119594275, + "loss": 1.0031, + "step": 3400 + }, + { + "epoch": 0.23600027756574837, + "grad_norm": 3.9165749549865723, + "learning_rate": 0.00019955958461242626, + "loss": 1.116, + "step": 3401 + }, + { + "epoch": 0.23606966900284504, + "grad_norm": 3.514860153198242, + "learning_rate": 0.0001995590577146117, + "loss": 0.9725, + "step": 3402 + }, + { + "epoch": 0.2361390604399417, + "grad_norm": 4.183601379394531, + "learning_rate": 0.00019955853050250078, + "loss": 1.397, + "step": 3403 + }, + { + "epoch": 0.23620845187703837, + "grad_norm": 3.0487966537475586, + "learning_rate": 0.0001995580029760951, + "loss": 1.119, + "step": 3404 + }, + { + "epoch": 0.23627784331413504, + "grad_norm": 3.796976089477539, + "learning_rate": 0.00019955747513539636, + "loss": 1.2179, + "step": 3405 + }, + { + "epoch": 0.2363472347512317, + "grad_norm": 3.507960557937622, + "learning_rate": 0.0001995569469804062, + "loss": 1.0475, + "step": 3406 + }, + { + "epoch": 0.23641662618832837, + "grad_norm": 4.008047103881836, + "learning_rate": 0.00019955641851112637, + "loss": 1.0, + "step": 3407 + }, + { + "epoch": 0.236486017625425, + "grad_norm": 4.383129596710205, + "learning_rate": 0.00019955588972755842, + "loss": 1.1602, + "step": 3408 + }, + { + "epoch": 0.23655540906252168, + "grad_norm": 4.176469326019287, + "learning_rate": 0.0001995553606297041, + "loss": 1.1231, + "step": 3409 + }, + { + "epoch": 0.23662480049961834, + "grad_norm": 3.2748029232025146, + "learning_rate": 0.00019955483121756505, + "loss": 0.9196, + "step": 3410 + }, + { + "epoch": 0.236694191936715, + "grad_norm": 5.0485382080078125, + "learning_rate": 0.00019955430149114296, + "loss": 1.1744, + "step": 3411 + }, + { + "epoch": 0.23676358337381168, + "grad_norm": 2.852327823638916, + "learning_rate": 0.00019955377145043946, + "loss": 1.0541, + "step": 3412 + }, + { + "epoch": 0.23683297481090834, + "grad_norm": 3.0798397064208984, + "learning_rate": 0.0001995532410954563, + "loss": 1.0426, + "step": 3413 + }, + { + "epoch": 0.236902366248005, + "grad_norm": 3.0307109355926514, + "learning_rate": 0.0001995527104261951, + "loss": 0.9753, + "step": 3414 + }, + { + "epoch": 0.23697175768510165, + "grad_norm": 3.3083252906799316, + "learning_rate": 0.00019955217944265749, + "loss": 1.254, + "step": 3415 + }, + { + "epoch": 0.2370411491221983, + "grad_norm": 3.306077718734741, + "learning_rate": 0.00019955164814484526, + "loss": 1.1452, + "step": 3416 + }, + { + "epoch": 0.23711054055929498, + "grad_norm": 5.149942398071289, + "learning_rate": 0.00019955111653276002, + "loss": 1.2785, + "step": 3417 + }, + { + "epoch": 0.23717993199639165, + "grad_norm": 3.982743263244629, + "learning_rate": 0.0001995505846064034, + "loss": 1.1032, + "step": 3418 + }, + { + "epoch": 0.2372493234334883, + "grad_norm": 3.4840240478515625, + "learning_rate": 0.00019955005236577719, + "loss": 1.0348, + "step": 3419 + }, + { + "epoch": 0.23731871487058498, + "grad_norm": 3.680542230606079, + "learning_rate": 0.00019954951981088301, + "loss": 1.4795, + "step": 3420 + }, + { + "epoch": 0.23738810630768165, + "grad_norm": 5.225826740264893, + "learning_rate": 0.00019954898694172254, + "loss": 1.3218, + "step": 3421 + }, + { + "epoch": 0.23745749774477828, + "grad_norm": 3.3373825550079346, + "learning_rate": 0.00019954845375829745, + "loss": 0.9959, + "step": 3422 + }, + { + "epoch": 0.23752688918187495, + "grad_norm": 4.271003246307373, + "learning_rate": 0.00019954792026060948, + "loss": 1.1848, + "step": 3423 + }, + { + "epoch": 0.23759628061897162, + "grad_norm": 3.857607841491699, + "learning_rate": 0.00019954738644866024, + "loss": 0.9855, + "step": 3424 + }, + { + "epoch": 0.23766567205606828, + "grad_norm": 2.9715518951416016, + "learning_rate": 0.00019954685232245146, + "loss": 0.9103, + "step": 3425 + }, + { + "epoch": 0.23773506349316495, + "grad_norm": 3.321157217025757, + "learning_rate": 0.0001995463178819848, + "loss": 1.1992, + "step": 3426 + }, + { + "epoch": 0.23780445493026162, + "grad_norm": 3.0434064865112305, + "learning_rate": 0.000199545783127262, + "loss": 0.8378, + "step": 3427 + }, + { + "epoch": 0.23787384636735825, + "grad_norm": 3.1912899017333984, + "learning_rate": 0.00019954524805828466, + "loss": 0.8981, + "step": 3428 + }, + { + "epoch": 0.23794323780445492, + "grad_norm": 3.0712575912475586, + "learning_rate": 0.00019954471267505456, + "loss": 1.1671, + "step": 3429 + }, + { + "epoch": 0.2380126292415516, + "grad_norm": 4.9891438484191895, + "learning_rate": 0.00019954417697757337, + "loss": 1.011, + "step": 3430 + }, + { + "epoch": 0.23808202067864825, + "grad_norm": 3.3351967334747314, + "learning_rate": 0.00019954364096584274, + "loss": 1.0797, + "step": 3431 + }, + { + "epoch": 0.23815141211574492, + "grad_norm": 3.5693867206573486, + "learning_rate": 0.00019954310463986437, + "loss": 0.9988, + "step": 3432 + }, + { + "epoch": 0.2382208035528416, + "grad_norm": 4.008341312408447, + "learning_rate": 0.00019954256799964, + "loss": 1.008, + "step": 3433 + }, + { + "epoch": 0.23829019498993825, + "grad_norm": 3.945685625076294, + "learning_rate": 0.00019954203104517127, + "loss": 1.3727, + "step": 3434 + }, + { + "epoch": 0.2383595864270349, + "grad_norm": 4.777988910675049, + "learning_rate": 0.0001995414937764599, + "loss": 1.1184, + "step": 3435 + }, + { + "epoch": 0.23842897786413156, + "grad_norm": 3.3311684131622314, + "learning_rate": 0.0001995409561935076, + "loss": 0.9699, + "step": 3436 + }, + { + "epoch": 0.23849836930122822, + "grad_norm": 3.7560698986053467, + "learning_rate": 0.000199540418296316, + "loss": 1.1067, + "step": 3437 + }, + { + "epoch": 0.2385677607383249, + "grad_norm": 3.8560352325439453, + "learning_rate": 0.0001995398800848869, + "loss": 1.0251, + "step": 3438 + }, + { + "epoch": 0.23863715217542156, + "grad_norm": 3.612513303756714, + "learning_rate": 0.00019953934155922195, + "loss": 0.9331, + "step": 3439 + }, + { + "epoch": 0.23870654361251822, + "grad_norm": 5.658044338226318, + "learning_rate": 0.0001995388027193228, + "loss": 1.4469, + "step": 3440 + }, + { + "epoch": 0.2387759350496149, + "grad_norm": 3.4160916805267334, + "learning_rate": 0.00019953826356519124, + "loss": 0.895, + "step": 3441 + }, + { + "epoch": 0.23884532648671153, + "grad_norm": 3.1803040504455566, + "learning_rate": 0.0001995377240968289, + "loss": 0.8302, + "step": 3442 + }, + { + "epoch": 0.2389147179238082, + "grad_norm": 4.451276779174805, + "learning_rate": 0.00019953718431423752, + "loss": 1.0489, + "step": 3443 + }, + { + "epoch": 0.23898410936090486, + "grad_norm": 5.13118314743042, + "learning_rate": 0.00019953664421741883, + "loss": 1.1995, + "step": 3444 + }, + { + "epoch": 0.23905350079800153, + "grad_norm": 3.741546869277954, + "learning_rate": 0.00019953610380637446, + "loss": 0.9904, + "step": 3445 + }, + { + "epoch": 0.2391228922350982, + "grad_norm": 4.503830432891846, + "learning_rate": 0.00019953556308110616, + "loss": 0.8474, + "step": 3446 + }, + { + "epoch": 0.23919228367219486, + "grad_norm": 3.9796595573425293, + "learning_rate": 0.00019953502204161566, + "loss": 1.2002, + "step": 3447 + }, + { + "epoch": 0.23926167510929153, + "grad_norm": 3.0768542289733887, + "learning_rate": 0.00019953448068790465, + "loss": 0.8437, + "step": 3448 + }, + { + "epoch": 0.23933106654638817, + "grad_norm": 3.6979432106018066, + "learning_rate": 0.00019953393901997482, + "loss": 1.0697, + "step": 3449 + }, + { + "epoch": 0.23940045798348483, + "grad_norm": 4.295778274536133, + "learning_rate": 0.00019953339703782788, + "loss": 1.14, + "step": 3450 + }, + { + "epoch": 0.2394698494205815, + "grad_norm": 4.67337703704834, + "learning_rate": 0.00019953285474146558, + "loss": 0.7078, + "step": 3451 + }, + { + "epoch": 0.23953924085767817, + "grad_norm": 3.5722429752349854, + "learning_rate": 0.0001995323121308896, + "loss": 1.1131, + "step": 3452 + }, + { + "epoch": 0.23960863229477483, + "grad_norm": 4.250100135803223, + "learning_rate": 0.00019953176920610164, + "loss": 1.2007, + "step": 3453 + }, + { + "epoch": 0.2396780237318715, + "grad_norm": 3.3485054969787598, + "learning_rate": 0.00019953122596710342, + "loss": 0.9927, + "step": 3454 + }, + { + "epoch": 0.23974741516896814, + "grad_norm": 3.712129592895508, + "learning_rate": 0.00019953068241389672, + "loss": 0.8409, + "step": 3455 + }, + { + "epoch": 0.2398168066060648, + "grad_norm": 2.639275550842285, + "learning_rate": 0.00019953013854648316, + "loss": 0.8789, + "step": 3456 + }, + { + "epoch": 0.23988619804316147, + "grad_norm": 3.4648022651672363, + "learning_rate": 0.0001995295943648645, + "loss": 0.9505, + "step": 3457 + }, + { + "epoch": 0.23995558948025814, + "grad_norm": 3.750455617904663, + "learning_rate": 0.00019952904986904248, + "loss": 1.0842, + "step": 3458 + }, + { + "epoch": 0.2400249809173548, + "grad_norm": 3.8571174144744873, + "learning_rate": 0.0001995285050590188, + "loss": 1.1673, + "step": 3459 + }, + { + "epoch": 0.24009437235445147, + "grad_norm": 2.6242928504943848, + "learning_rate": 0.00019952795993479516, + "loss": 0.7203, + "step": 3460 + }, + { + "epoch": 0.24016376379154813, + "grad_norm": 3.451505184173584, + "learning_rate": 0.0001995274144963733, + "loss": 1.0169, + "step": 3461 + }, + { + "epoch": 0.24023315522864477, + "grad_norm": 3.386063575744629, + "learning_rate": 0.00019952686874375494, + "loss": 0.9956, + "step": 3462 + }, + { + "epoch": 0.24030254666574144, + "grad_norm": 4.01813268661499, + "learning_rate": 0.0001995263226769418, + "loss": 1.0136, + "step": 3463 + }, + { + "epoch": 0.2403719381028381, + "grad_norm": 3.396890640258789, + "learning_rate": 0.00019952577629593564, + "loss": 1.1131, + "step": 3464 + }, + { + "epoch": 0.24044132953993477, + "grad_norm": 3.400387763977051, + "learning_rate": 0.00019952522960073812, + "loss": 0.9756, + "step": 3465 + }, + { + "epoch": 0.24051072097703144, + "grad_norm": 4.559964656829834, + "learning_rate": 0.000199524682591351, + "loss": 1.277, + "step": 3466 + }, + { + "epoch": 0.2405801124141281, + "grad_norm": 3.8075430393218994, + "learning_rate": 0.000199524135267776, + "loss": 1.3274, + "step": 3467 + }, + { + "epoch": 0.24064950385122477, + "grad_norm": 3.4807088375091553, + "learning_rate": 0.00019952358763001486, + "loss": 1.6221, + "step": 3468 + }, + { + "epoch": 0.2407188952883214, + "grad_norm": 3.5957181453704834, + "learning_rate": 0.0001995230396780693, + "loss": 0.9571, + "step": 3469 + }, + { + "epoch": 0.24078828672541808, + "grad_norm": 4.82681131362915, + "learning_rate": 0.00019952249141194102, + "loss": 1.1657, + "step": 3470 + }, + { + "epoch": 0.24085767816251474, + "grad_norm": 3.000309705734253, + "learning_rate": 0.0001995219428316318, + "loss": 1.0678, + "step": 3471 + }, + { + "epoch": 0.2409270695996114, + "grad_norm": 3.6231141090393066, + "learning_rate": 0.00019952139393714338, + "loss": 1.2127, + "step": 3472 + }, + { + "epoch": 0.24099646103670808, + "grad_norm": 5.1464152336120605, + "learning_rate": 0.00019952084472847745, + "loss": 1.0193, + "step": 3473 + }, + { + "epoch": 0.24106585247380474, + "grad_norm": 3.5336530208587646, + "learning_rate": 0.00019952029520563578, + "loss": 0.7664, + "step": 3474 + }, + { + "epoch": 0.24113524391090138, + "grad_norm": 3.415295124053955, + "learning_rate": 0.00019951974536862004, + "loss": 1.0111, + "step": 3475 + }, + { + "epoch": 0.24120463534799805, + "grad_norm": 3.512385368347168, + "learning_rate": 0.00019951919521743206, + "loss": 1.1158, + "step": 3476 + }, + { + "epoch": 0.24127402678509471, + "grad_norm": 2.6419408321380615, + "learning_rate": 0.00019951864475207346, + "loss": 1.0458, + "step": 3477 + }, + { + "epoch": 0.24134341822219138, + "grad_norm": 3.7553870677948, + "learning_rate": 0.00019951809397254612, + "loss": 1.1251, + "step": 3478 + }, + { + "epoch": 0.24141280965928805, + "grad_norm": 3.847503185272217, + "learning_rate": 0.00019951754287885165, + "loss": 0.9176, + "step": 3479 + }, + { + "epoch": 0.2414822010963847, + "grad_norm": 3.6143124103546143, + "learning_rate": 0.00019951699147099187, + "loss": 0.8085, + "step": 3480 + }, + { + "epoch": 0.24155159253348138, + "grad_norm": 3.4794204235076904, + "learning_rate": 0.00019951643974896847, + "loss": 1.1501, + "step": 3481 + }, + { + "epoch": 0.24162098397057802, + "grad_norm": 3.81398344039917, + "learning_rate": 0.00019951588771278324, + "loss": 0.995, + "step": 3482 + }, + { + "epoch": 0.24169037540767468, + "grad_norm": 3.2744853496551514, + "learning_rate": 0.0001995153353624379, + "loss": 0.93, + "step": 3483 + }, + { + "epoch": 0.24175976684477135, + "grad_norm": 3.8639869689941406, + "learning_rate": 0.0001995147826979342, + "loss": 1.2556, + "step": 3484 + }, + { + "epoch": 0.24182915828186802, + "grad_norm": 2.6599156856536865, + "learning_rate": 0.00019951422971927385, + "loss": 0.6623, + "step": 3485 + }, + { + "epoch": 0.24189854971896468, + "grad_norm": 3.452091932296753, + "learning_rate": 0.0001995136764264586, + "loss": 0.9971, + "step": 3486 + }, + { + "epoch": 0.24196794115606135, + "grad_norm": 4.032805442810059, + "learning_rate": 0.0001995131228194903, + "loss": 1.0699, + "step": 3487 + }, + { + "epoch": 0.24203733259315802, + "grad_norm": 3.4706151485443115, + "learning_rate": 0.00019951256889837052, + "loss": 1.2474, + "step": 3488 + }, + { + "epoch": 0.24210672403025466, + "grad_norm": 3.470019578933716, + "learning_rate": 0.00019951201466310116, + "loss": 1.12, + "step": 3489 + }, + { + "epoch": 0.24217611546735132, + "grad_norm": 3.883162021636963, + "learning_rate": 0.00019951146011368392, + "loss": 1.1057, + "step": 3490 + }, + { + "epoch": 0.242245506904448, + "grad_norm": 3.862604856491089, + "learning_rate": 0.00019951090525012052, + "loss": 1.0192, + "step": 3491 + }, + { + "epoch": 0.24231489834154465, + "grad_norm": 3.8017494678497314, + "learning_rate": 0.00019951035007241275, + "loss": 0.8258, + "step": 3492 + }, + { + "epoch": 0.24238428977864132, + "grad_norm": 2.6865854263305664, + "learning_rate": 0.00019950979458056235, + "loss": 0.7157, + "step": 3493 + }, + { + "epoch": 0.242453681215738, + "grad_norm": 3.568044662475586, + "learning_rate": 0.00019950923877457108, + "loss": 1.0439, + "step": 3494 + }, + { + "epoch": 0.24252307265283465, + "grad_norm": 3.5899579524993896, + "learning_rate": 0.0001995086826544407, + "loss": 0.8921, + "step": 3495 + }, + { + "epoch": 0.2425924640899313, + "grad_norm": 3.4703478813171387, + "learning_rate": 0.00019950812622017292, + "loss": 1.2259, + "step": 3496 + }, + { + "epoch": 0.24266185552702796, + "grad_norm": 3.810257911682129, + "learning_rate": 0.00019950756947176957, + "loss": 0.9777, + "step": 3497 + }, + { + "epoch": 0.24273124696412463, + "grad_norm": 3.94386887550354, + "learning_rate": 0.00019950701240923233, + "loss": 1.329, + "step": 3498 + }, + { + "epoch": 0.2428006384012213, + "grad_norm": 4.191793918609619, + "learning_rate": 0.00019950645503256302, + "loss": 1.1549, + "step": 3499 + }, + { + "epoch": 0.24287002983831796, + "grad_norm": 2.54777455329895, + "learning_rate": 0.00019950589734176336, + "loss": 0.7082, + "step": 3500 + }, + { + "epoch": 0.24293942127541462, + "grad_norm": 3.1745963096618652, + "learning_rate": 0.00019950533933683516, + "loss": 1.1007, + "step": 3501 + }, + { + "epoch": 0.24300881271251126, + "grad_norm": 3.945650339126587, + "learning_rate": 0.0001995047810177801, + "loss": 1.2747, + "step": 3502 + }, + { + "epoch": 0.24307820414960793, + "grad_norm": 5.888225555419922, + "learning_rate": 0.00019950422238460003, + "loss": 0.9553, + "step": 3503 + }, + { + "epoch": 0.2431475955867046, + "grad_norm": 3.709864854812622, + "learning_rate": 0.00019950366343729665, + "loss": 1.0245, + "step": 3504 + }, + { + "epoch": 0.24321698702380126, + "grad_norm": 3.63999342918396, + "learning_rate": 0.00019950310417587178, + "loss": 1.0748, + "step": 3505 + }, + { + "epoch": 0.24328637846089793, + "grad_norm": 5.998307228088379, + "learning_rate": 0.00019950254460032713, + "loss": 1.1558, + "step": 3506 + }, + { + "epoch": 0.2433557698979946, + "grad_norm": 3.6744234561920166, + "learning_rate": 0.0001995019847106645, + "loss": 1.0642, + "step": 3507 + }, + { + "epoch": 0.24342516133509126, + "grad_norm": 3.2565412521362305, + "learning_rate": 0.00019950142450688563, + "loss": 1.1521, + "step": 3508 + }, + { + "epoch": 0.2434945527721879, + "grad_norm": 3.6455013751983643, + "learning_rate": 0.00019950086398899233, + "loss": 1.0107, + "step": 3509 + }, + { + "epoch": 0.24356394420928457, + "grad_norm": 3.703680992126465, + "learning_rate": 0.00019950030315698633, + "loss": 0.9721, + "step": 3510 + }, + { + "epoch": 0.24363333564638123, + "grad_norm": 3.2352347373962402, + "learning_rate": 0.00019949974201086942, + "loss": 0.7451, + "step": 3511 + }, + { + "epoch": 0.2437027270834779, + "grad_norm": 6.015429496765137, + "learning_rate": 0.00019949918055064337, + "loss": 1.0538, + "step": 3512 + }, + { + "epoch": 0.24377211852057457, + "grad_norm": 4.818708419799805, + "learning_rate": 0.00019949861877630997, + "loss": 0.9128, + "step": 3513 + }, + { + "epoch": 0.24384150995767123, + "grad_norm": 3.706911087036133, + "learning_rate": 0.00019949805668787097, + "loss": 1.0196, + "step": 3514 + }, + { + "epoch": 0.2439109013947679, + "grad_norm": 3.517590284347534, + "learning_rate": 0.00019949749428532814, + "loss": 0.8923, + "step": 3515 + }, + { + "epoch": 0.24398029283186454, + "grad_norm": 3.434833526611328, + "learning_rate": 0.00019949693156868327, + "loss": 0.9739, + "step": 3516 + }, + { + "epoch": 0.2440496842689612, + "grad_norm": 3.4330615997314453, + "learning_rate": 0.0001994963685379381, + "loss": 1.1457, + "step": 3517 + }, + { + "epoch": 0.24411907570605787, + "grad_norm": 4.569477558135986, + "learning_rate": 0.00019949580519309446, + "loss": 1.0918, + "step": 3518 + }, + { + "epoch": 0.24418846714315454, + "grad_norm": 4.890532493591309, + "learning_rate": 0.00019949524153415414, + "loss": 1.2192, + "step": 3519 + }, + { + "epoch": 0.2442578585802512, + "grad_norm": 3.7970874309539795, + "learning_rate": 0.00019949467756111885, + "loss": 0.9861, + "step": 3520 + }, + { + "epoch": 0.24432725001734787, + "grad_norm": 3.3703017234802246, + "learning_rate": 0.00019949411327399038, + "loss": 0.9971, + "step": 3521 + }, + { + "epoch": 0.24439664145444454, + "grad_norm": 4.994119167327881, + "learning_rate": 0.00019949354867277057, + "loss": 1.2901, + "step": 3522 + }, + { + "epoch": 0.24446603289154117, + "grad_norm": 3.334357738494873, + "learning_rate": 0.0001994929837574612, + "loss": 0.9654, + "step": 3523 + }, + { + "epoch": 0.24453542432863784, + "grad_norm": 3.50935435295105, + "learning_rate": 0.00019949241852806397, + "loss": 1.1399, + "step": 3524 + }, + { + "epoch": 0.2446048157657345, + "grad_norm": 4.856400489807129, + "learning_rate": 0.00019949185298458075, + "loss": 0.9418, + "step": 3525 + }, + { + "epoch": 0.24467420720283117, + "grad_norm": 3.9250810146331787, + "learning_rate": 0.0001994912871270133, + "loss": 1.228, + "step": 3526 + }, + { + "epoch": 0.24474359863992784, + "grad_norm": 3.226517915725708, + "learning_rate": 0.00019949072095536336, + "loss": 1.3737, + "step": 3527 + }, + { + "epoch": 0.2448129900770245, + "grad_norm": 3.7732372283935547, + "learning_rate": 0.00019949015446963278, + "loss": 1.1218, + "step": 3528 + }, + { + "epoch": 0.24488238151412114, + "grad_norm": 3.5914156436920166, + "learning_rate": 0.00019948958766982333, + "loss": 1.2877, + "step": 3529 + }, + { + "epoch": 0.2449517729512178, + "grad_norm": 3.829010248184204, + "learning_rate": 0.0001994890205559368, + "loss": 1.1925, + "step": 3530 + }, + { + "epoch": 0.24502116438831448, + "grad_norm": 4.345839023590088, + "learning_rate": 0.00019948845312797495, + "loss": 1.4093, + "step": 3531 + }, + { + "epoch": 0.24509055582541114, + "grad_norm": 4.015350818634033, + "learning_rate": 0.00019948788538593962, + "loss": 0.923, + "step": 3532 + }, + { + "epoch": 0.2451599472625078, + "grad_norm": 3.3021011352539062, + "learning_rate": 0.00019948731732983256, + "loss": 1.1385, + "step": 3533 + }, + { + "epoch": 0.24522933869960448, + "grad_norm": 3.809717893600464, + "learning_rate": 0.0001994867489596556, + "loss": 1.3799, + "step": 3534 + }, + { + "epoch": 0.24529873013670114, + "grad_norm": 3.311636209487915, + "learning_rate": 0.00019948618027541053, + "loss": 1.1038, + "step": 3535 + }, + { + "epoch": 0.24536812157379778, + "grad_norm": 4.067187786102295, + "learning_rate": 0.0001994856112770991, + "loss": 0.9777, + "step": 3536 + }, + { + "epoch": 0.24543751301089445, + "grad_norm": 3.5336015224456787, + "learning_rate": 0.00019948504196472312, + "loss": 0.9904, + "step": 3537 + }, + { + "epoch": 0.24550690444799111, + "grad_norm": 4.719510078430176, + "learning_rate": 0.00019948447233828444, + "loss": 1.0529, + "step": 3538 + }, + { + "epoch": 0.24557629588508778, + "grad_norm": 2.764763593673706, + "learning_rate": 0.00019948390239778483, + "loss": 0.7037, + "step": 3539 + }, + { + "epoch": 0.24564568732218445, + "grad_norm": 3.512723445892334, + "learning_rate": 0.00019948333214322606, + "loss": 1.1442, + "step": 3540 + }, + { + "epoch": 0.2457150787592811, + "grad_norm": 3.956130027770996, + "learning_rate": 0.00019948276157460997, + "loss": 1.2383, + "step": 3541 + }, + { + "epoch": 0.24578447019637778, + "grad_norm": 4.513791561126709, + "learning_rate": 0.00019948219069193833, + "loss": 1.0735, + "step": 3542 + }, + { + "epoch": 0.24585386163347442, + "grad_norm": 3.57963490486145, + "learning_rate": 0.000199481619495213, + "loss": 1.0002, + "step": 3543 + }, + { + "epoch": 0.24592325307057109, + "grad_norm": 3.6866114139556885, + "learning_rate": 0.00019948104798443565, + "loss": 0.9439, + "step": 3544 + }, + { + "epoch": 0.24599264450766775, + "grad_norm": 3.6268951892852783, + "learning_rate": 0.00019948047615960825, + "loss": 0.8223, + "step": 3545 + }, + { + "epoch": 0.24606203594476442, + "grad_norm": 2.9243428707122803, + "learning_rate": 0.00019947990402073248, + "loss": 1.0766, + "step": 3546 + }, + { + "epoch": 0.24613142738186108, + "grad_norm": 3.6785483360290527, + "learning_rate": 0.00019947933156781023, + "loss": 1.4303, + "step": 3547 + }, + { + "epoch": 0.24620081881895775, + "grad_norm": 3.2352490425109863, + "learning_rate": 0.00019947875880084326, + "loss": 1.0344, + "step": 3548 + }, + { + "epoch": 0.2462702102560544, + "grad_norm": 2.7503855228424072, + "learning_rate": 0.0001994781857198334, + "loss": 1.1384, + "step": 3549 + }, + { + "epoch": 0.24633960169315106, + "grad_norm": 3.0422439575195312, + "learning_rate": 0.00019947761232478245, + "loss": 0.8283, + "step": 3550 + }, + { + "epoch": 0.24640899313024772, + "grad_norm": 3.638732433319092, + "learning_rate": 0.00019947703861569217, + "loss": 1.4077, + "step": 3551 + }, + { + "epoch": 0.2464783845673444, + "grad_norm": 3.226574182510376, + "learning_rate": 0.00019947646459256445, + "loss": 1.0882, + "step": 3552 + }, + { + "epoch": 0.24654777600444105, + "grad_norm": 3.984525680541992, + "learning_rate": 0.0001994758902554011, + "loss": 1.0697, + "step": 3553 + }, + { + "epoch": 0.24661716744153772, + "grad_norm": 4.816558361053467, + "learning_rate": 0.00019947531560420386, + "loss": 1.2022, + "step": 3554 + }, + { + "epoch": 0.2466865588786344, + "grad_norm": 2.8199660778045654, + "learning_rate": 0.00019947474063897465, + "loss": 0.8969, + "step": 3555 + }, + { + "epoch": 0.24675595031573103, + "grad_norm": 4.0441389083862305, + "learning_rate": 0.0001994741653597152, + "loss": 0.9804, + "step": 3556 + }, + { + "epoch": 0.2468253417528277, + "grad_norm": 3.188368082046509, + "learning_rate": 0.00019947358976642732, + "loss": 0.7974, + "step": 3557 + }, + { + "epoch": 0.24689473318992436, + "grad_norm": 3.4619531631469727, + "learning_rate": 0.0001994730138591129, + "loss": 1.0143, + "step": 3558 + }, + { + "epoch": 0.24696412462702103, + "grad_norm": 3.3757753372192383, + "learning_rate": 0.00019947243763777367, + "loss": 1.0763, + "step": 3559 + }, + { + "epoch": 0.2470335160641177, + "grad_norm": 3.7431349754333496, + "learning_rate": 0.00019947186110241155, + "loss": 1.2008, + "step": 3560 + }, + { + "epoch": 0.24710290750121436, + "grad_norm": 5.090317726135254, + "learning_rate": 0.0001994712842530283, + "loss": 1.0369, + "step": 3561 + }, + { + "epoch": 0.24717229893831102, + "grad_norm": 4.010122299194336, + "learning_rate": 0.0001994707070896257, + "loss": 1.0132, + "step": 3562 + }, + { + "epoch": 0.24724169037540766, + "grad_norm": 4.527193546295166, + "learning_rate": 0.00019947012961220564, + "loss": 1.0901, + "step": 3563 + }, + { + "epoch": 0.24731108181250433, + "grad_norm": 2.600006341934204, + "learning_rate": 0.00019946955182076993, + "loss": 0.6504, + "step": 3564 + }, + { + "epoch": 0.247380473249601, + "grad_norm": 6.207520008087158, + "learning_rate": 0.00019946897371532035, + "loss": 1.1016, + "step": 3565 + }, + { + "epoch": 0.24744986468669766, + "grad_norm": 3.6929948329925537, + "learning_rate": 0.0001994683952958588, + "loss": 0.9245, + "step": 3566 + }, + { + "epoch": 0.24751925612379433, + "grad_norm": 3.0079808235168457, + "learning_rate": 0.00019946781656238706, + "loss": 1.0669, + "step": 3567 + }, + { + "epoch": 0.247588647560891, + "grad_norm": 3.113098382949829, + "learning_rate": 0.00019946723751490696, + "loss": 0.9858, + "step": 3568 + }, + { + "epoch": 0.24765803899798766, + "grad_norm": 4.247239112854004, + "learning_rate": 0.00019946665815342033, + "loss": 1.2246, + "step": 3569 + }, + { + "epoch": 0.2477274304350843, + "grad_norm": 2.6810829639434814, + "learning_rate": 0.00019946607847792902, + "loss": 0.772, + "step": 3570 + }, + { + "epoch": 0.24779682187218097, + "grad_norm": 3.554161787033081, + "learning_rate": 0.00019946549848843482, + "loss": 0.8868, + "step": 3571 + }, + { + "epoch": 0.24786621330927763, + "grad_norm": 3.4472908973693848, + "learning_rate": 0.00019946491818493957, + "loss": 1.212, + "step": 3572 + }, + { + "epoch": 0.2479356047463743, + "grad_norm": 4.096602916717529, + "learning_rate": 0.00019946433756744513, + "loss": 1.3393, + "step": 3573 + }, + { + "epoch": 0.24800499618347097, + "grad_norm": 3.8425960540771484, + "learning_rate": 0.0001994637566359533, + "loss": 1.0415, + "step": 3574 + }, + { + "epoch": 0.24807438762056763, + "grad_norm": 4.355343341827393, + "learning_rate": 0.00019946317539046597, + "loss": 1.0679, + "step": 3575 + }, + { + "epoch": 0.24814377905766427, + "grad_norm": 4.2551589012146, + "learning_rate": 0.00019946259383098492, + "loss": 1.0538, + "step": 3576 + }, + { + "epoch": 0.24821317049476094, + "grad_norm": 3.429365873336792, + "learning_rate": 0.00019946201195751196, + "loss": 1.0704, + "step": 3577 + }, + { + "epoch": 0.2482825619318576, + "grad_norm": 2.8166890144348145, + "learning_rate": 0.000199461429770049, + "loss": 0.7922, + "step": 3578 + }, + { + "epoch": 0.24835195336895427, + "grad_norm": 3.5844109058380127, + "learning_rate": 0.00019946084726859783, + "loss": 0.8594, + "step": 3579 + }, + { + "epoch": 0.24842134480605094, + "grad_norm": 3.3173110485076904, + "learning_rate": 0.00019946026445316032, + "loss": 1.0314, + "step": 3580 + }, + { + "epoch": 0.2484907362431476, + "grad_norm": 4.8751678466796875, + "learning_rate": 0.00019945968132373832, + "loss": 1.3499, + "step": 3581 + }, + { + "epoch": 0.24856012768024427, + "grad_norm": 3.7398922443389893, + "learning_rate": 0.00019945909788033361, + "loss": 0.9882, + "step": 3582 + }, + { + "epoch": 0.2486295191173409, + "grad_norm": 3.8505773544311523, + "learning_rate": 0.00019945851412294809, + "loss": 1.1727, + "step": 3583 + }, + { + "epoch": 0.24869891055443757, + "grad_norm": 3.0206947326660156, + "learning_rate": 0.00019945793005158354, + "loss": 0.9369, + "step": 3584 + }, + { + "epoch": 0.24876830199153424, + "grad_norm": 2.793557643890381, + "learning_rate": 0.0001994573456662419, + "loss": 0.9111, + "step": 3585 + }, + { + "epoch": 0.2488376934286309, + "grad_norm": 2.7457146644592285, + "learning_rate": 0.00019945676096692493, + "loss": 0.8413, + "step": 3586 + }, + { + "epoch": 0.24890708486572757, + "grad_norm": 3.285252809524536, + "learning_rate": 0.0001994561759536345, + "loss": 0.9399, + "step": 3587 + }, + { + "epoch": 0.24897647630282424, + "grad_norm": 4.049124240875244, + "learning_rate": 0.00019945559062637246, + "loss": 0.88, + "step": 3588 + }, + { + "epoch": 0.2490458677399209, + "grad_norm": 3.0636086463928223, + "learning_rate": 0.00019945500498514066, + "loss": 0.8165, + "step": 3589 + }, + { + "epoch": 0.24911525917701755, + "grad_norm": 3.813300609588623, + "learning_rate": 0.00019945441902994096, + "loss": 1.0787, + "step": 3590 + }, + { + "epoch": 0.2491846506141142, + "grad_norm": 4.2317352294921875, + "learning_rate": 0.00019945383276077518, + "loss": 1.0569, + "step": 3591 + }, + { + "epoch": 0.24925404205121088, + "grad_norm": 3.259233236312866, + "learning_rate": 0.0001994532461776452, + "loss": 1.0029, + "step": 3592 + }, + { + "epoch": 0.24932343348830754, + "grad_norm": 3.0229151248931885, + "learning_rate": 0.00019945265928055287, + "loss": 0.8291, + "step": 3593 + }, + { + "epoch": 0.2493928249254042, + "grad_norm": 3.2431435585021973, + "learning_rate": 0.00019945207206950003, + "loss": 1.0201, + "step": 3594 + }, + { + "epoch": 0.24946221636250088, + "grad_norm": 4.829829692840576, + "learning_rate": 0.00019945148454448853, + "loss": 1.2387, + "step": 3595 + }, + { + "epoch": 0.24953160779959754, + "grad_norm": 2.8389275074005127, + "learning_rate": 0.00019945089670552024, + "loss": 1.1058, + "step": 3596 + }, + { + "epoch": 0.24960099923669418, + "grad_norm": 3.222689151763916, + "learning_rate": 0.000199450308552597, + "loss": 0.92, + "step": 3597 + }, + { + "epoch": 0.24967039067379085, + "grad_norm": 3.108664035797119, + "learning_rate": 0.00019944972008572066, + "loss": 1.1062, + "step": 3598 + }, + { + "epoch": 0.24973978211088751, + "grad_norm": 2.8446121215820312, + "learning_rate": 0.0001994491313048931, + "loss": 0.9193, + "step": 3599 + }, + { + "epoch": 0.24980917354798418, + "grad_norm": 5.01584005355835, + "learning_rate": 0.00019944854221011618, + "loss": 1.3226, + "step": 3600 + }, + { + "epoch": 0.24987856498508085, + "grad_norm": 2.6901886463165283, + "learning_rate": 0.00019944795280139177, + "loss": 1.0414, + "step": 3601 + }, + { + "epoch": 0.24994795642217751, + "grad_norm": 3.4535434246063232, + "learning_rate": 0.00019944736307872166, + "loss": 0.994, + "step": 3602 + }, + { + "epoch": 0.25001734785927415, + "grad_norm": 3.6534833908081055, + "learning_rate": 0.00019944677304210783, + "loss": 1.4894, + "step": 3603 + }, + { + "epoch": 0.2500867392963708, + "grad_norm": 3.8221828937530518, + "learning_rate": 0.000199446182691552, + "loss": 0.8733, + "step": 3604 + }, + { + "epoch": 0.2501561307334675, + "grad_norm": 3.0542893409729004, + "learning_rate": 0.00019944559202705615, + "loss": 1.0653, + "step": 3605 + }, + { + "epoch": 0.25022552217056415, + "grad_norm": 4.613558292388916, + "learning_rate": 0.0001994450010486221, + "loss": 1.1947, + "step": 3606 + }, + { + "epoch": 0.2502949136076608, + "grad_norm": 2.9151933193206787, + "learning_rate": 0.00019944440975625173, + "loss": 0.9269, + "step": 3607 + }, + { + "epoch": 0.2503643050447575, + "grad_norm": 2.875950574874878, + "learning_rate": 0.00019944381814994687, + "loss": 1.0219, + "step": 3608 + }, + { + "epoch": 0.25043369648185415, + "grad_norm": 3.3176698684692383, + "learning_rate": 0.00019944322622970947, + "loss": 0.7635, + "step": 3609 + }, + { + "epoch": 0.2505030879189508, + "grad_norm": 3.4090912342071533, + "learning_rate": 0.00019944263399554127, + "loss": 0.9653, + "step": 3610 + }, + { + "epoch": 0.2505724793560475, + "grad_norm": 2.8524508476257324, + "learning_rate": 0.00019944204144744425, + "loss": 0.9687, + "step": 3611 + }, + { + "epoch": 0.25064187079314415, + "grad_norm": 3.634891986846924, + "learning_rate": 0.0001994414485854202, + "loss": 0.9562, + "step": 3612 + }, + { + "epoch": 0.25071126223024076, + "grad_norm": 3.456604480743408, + "learning_rate": 0.0001994408554094711, + "loss": 1.4817, + "step": 3613 + }, + { + "epoch": 0.2507806536673374, + "grad_norm": 2.99407958984375, + "learning_rate": 0.00019944026191959875, + "loss": 0.9537, + "step": 3614 + }, + { + "epoch": 0.2508500451044341, + "grad_norm": 3.1079037189483643, + "learning_rate": 0.000199439668115805, + "loss": 1.1153, + "step": 3615 + }, + { + "epoch": 0.25091943654153076, + "grad_norm": 3.5964202880859375, + "learning_rate": 0.00019943907399809177, + "loss": 1.091, + "step": 3616 + }, + { + "epoch": 0.2509888279786274, + "grad_norm": 3.9606246948242188, + "learning_rate": 0.0001994384795664609, + "loss": 1.295, + "step": 3617 + }, + { + "epoch": 0.2510582194157241, + "grad_norm": 3.0068676471710205, + "learning_rate": 0.00019943788482091427, + "loss": 0.9292, + "step": 3618 + }, + { + "epoch": 0.25112761085282076, + "grad_norm": 3.1439759731292725, + "learning_rate": 0.00019943728976145383, + "loss": 0.8744, + "step": 3619 + }, + { + "epoch": 0.2511970022899174, + "grad_norm": 3.4575626850128174, + "learning_rate": 0.0001994366943880814, + "loss": 0.9925, + "step": 3620 + }, + { + "epoch": 0.2512663937270141, + "grad_norm": 3.2513010501861572, + "learning_rate": 0.00019943609870079883, + "loss": 0.8502, + "step": 3621 + }, + { + "epoch": 0.25133578516411076, + "grad_norm": 3.150669813156128, + "learning_rate": 0.00019943550269960802, + "loss": 0.818, + "step": 3622 + }, + { + "epoch": 0.2514051766012074, + "grad_norm": 3.338125705718994, + "learning_rate": 0.0001994349063845109, + "loss": 0.7766, + "step": 3623 + }, + { + "epoch": 0.2514745680383041, + "grad_norm": 3.643873691558838, + "learning_rate": 0.00019943430975550932, + "loss": 0.9789, + "step": 3624 + }, + { + "epoch": 0.25154395947540076, + "grad_norm": 3.7374024391174316, + "learning_rate": 0.0001994337128126051, + "loss": 0.994, + "step": 3625 + }, + { + "epoch": 0.2516133509124974, + "grad_norm": 4.819711208343506, + "learning_rate": 0.00019943311555580025, + "loss": 1.1925, + "step": 3626 + }, + { + "epoch": 0.25168274234959404, + "grad_norm": 3.234110116958618, + "learning_rate": 0.00019943251798509656, + "loss": 0.955, + "step": 3627 + }, + { + "epoch": 0.2517521337866907, + "grad_norm": 3.5095784664154053, + "learning_rate": 0.00019943192010049598, + "loss": 1.2062, + "step": 3628 + }, + { + "epoch": 0.25182152522378737, + "grad_norm": 4.05068826675415, + "learning_rate": 0.00019943132190200036, + "loss": 1.4236, + "step": 3629 + }, + { + "epoch": 0.25189091666088403, + "grad_norm": 6.347320079803467, + "learning_rate": 0.00019943072338961154, + "loss": 0.8617, + "step": 3630 + }, + { + "epoch": 0.2519603080979807, + "grad_norm": 3.7170205116271973, + "learning_rate": 0.00019943012456333153, + "loss": 1.1791, + "step": 3631 + }, + { + "epoch": 0.25202969953507737, + "grad_norm": 3.994637966156006, + "learning_rate": 0.00019942952542316208, + "loss": 1.1253, + "step": 3632 + }, + { + "epoch": 0.25209909097217403, + "grad_norm": 3.273106336593628, + "learning_rate": 0.00019942892596910521, + "loss": 1.2845, + "step": 3633 + }, + { + "epoch": 0.2521684824092707, + "grad_norm": 4.767282009124756, + "learning_rate": 0.00019942832620116274, + "loss": 1.0492, + "step": 3634 + }, + { + "epoch": 0.25223787384636737, + "grad_norm": 3.4060111045837402, + "learning_rate": 0.0001994277261193366, + "loss": 0.9733, + "step": 3635 + }, + { + "epoch": 0.25230726528346403, + "grad_norm": 4.08785343170166, + "learning_rate": 0.00019942712572362865, + "loss": 1.0142, + "step": 3636 + }, + { + "epoch": 0.2523766567205607, + "grad_norm": 3.3911805152893066, + "learning_rate": 0.0001994265250140408, + "loss": 1.072, + "step": 3637 + }, + { + "epoch": 0.25244604815765737, + "grad_norm": 3.523947238922119, + "learning_rate": 0.00019942592399057493, + "loss": 1.1233, + "step": 3638 + }, + { + "epoch": 0.25251543959475403, + "grad_norm": 3.4180819988250732, + "learning_rate": 0.00019942532265323294, + "loss": 1.3188, + "step": 3639 + }, + { + "epoch": 0.25258483103185064, + "grad_norm": 3.854011058807373, + "learning_rate": 0.0001994247210020168, + "loss": 1.066, + "step": 3640 + }, + { + "epoch": 0.2526542224689473, + "grad_norm": 3.2203941345214844, + "learning_rate": 0.0001994241190369283, + "loss": 0.8993, + "step": 3641 + }, + { + "epoch": 0.252723613906044, + "grad_norm": 3.35392165184021, + "learning_rate": 0.00019942351675796943, + "loss": 1.2068, + "step": 3642 + }, + { + "epoch": 0.25279300534314064, + "grad_norm": 4.201396465301514, + "learning_rate": 0.00019942291416514204, + "loss": 1.1009, + "step": 3643 + }, + { + "epoch": 0.2528623967802373, + "grad_norm": 3.5138449668884277, + "learning_rate": 0.00019942231125844804, + "loss": 0.8533, + "step": 3644 + }, + { + "epoch": 0.252931788217334, + "grad_norm": 3.0460400581359863, + "learning_rate": 0.00019942170803788933, + "loss": 1.0449, + "step": 3645 + }, + { + "epoch": 0.25300117965443064, + "grad_norm": 3.619330883026123, + "learning_rate": 0.00019942110450346784, + "loss": 1.1647, + "step": 3646 + }, + { + "epoch": 0.2530705710915273, + "grad_norm": 3.525946855545044, + "learning_rate": 0.00019942050065518543, + "loss": 1.1575, + "step": 3647 + }, + { + "epoch": 0.253139962528624, + "grad_norm": 3.5808348655700684, + "learning_rate": 0.00019941989649304405, + "loss": 1.3963, + "step": 3648 + }, + { + "epoch": 0.25320935396572064, + "grad_norm": 4.353100299835205, + "learning_rate": 0.00019941929201704558, + "loss": 1.2431, + "step": 3649 + }, + { + "epoch": 0.2532787454028173, + "grad_norm": 4.002392292022705, + "learning_rate": 0.00019941868722719196, + "loss": 1.1266, + "step": 3650 + }, + { + "epoch": 0.253348136839914, + "grad_norm": 7.531451225280762, + "learning_rate": 0.00019941808212348508, + "loss": 0.975, + "step": 3651 + }, + { + "epoch": 0.25341752827701064, + "grad_norm": 3.1201424598693848, + "learning_rate": 0.00019941747670592684, + "loss": 1.0018, + "step": 3652 + }, + { + "epoch": 0.2534869197141073, + "grad_norm": 4.0764360427856445, + "learning_rate": 0.00019941687097451915, + "loss": 0.9971, + "step": 3653 + }, + { + "epoch": 0.2535563111512039, + "grad_norm": 3.229218006134033, + "learning_rate": 0.00019941626492926394, + "loss": 0.8289, + "step": 3654 + }, + { + "epoch": 0.2536257025883006, + "grad_norm": 4.033497333526611, + "learning_rate": 0.00019941565857016313, + "loss": 0.9671, + "step": 3655 + }, + { + "epoch": 0.25369509402539725, + "grad_norm": 4.03755521774292, + "learning_rate": 0.0001994150518972186, + "loss": 1.044, + "step": 3656 + }, + { + "epoch": 0.2537644854624939, + "grad_norm": 4.862986087799072, + "learning_rate": 0.0001994144449104323, + "loss": 1.2207, + "step": 3657 + }, + { + "epoch": 0.2538338768995906, + "grad_norm": 3.109783411026001, + "learning_rate": 0.00019941383760980615, + "loss": 0.7956, + "step": 3658 + }, + { + "epoch": 0.25390326833668725, + "grad_norm": 3.6517441272735596, + "learning_rate": 0.000199413229995342, + "loss": 1.0396, + "step": 3659 + }, + { + "epoch": 0.2539726597737839, + "grad_norm": 3.2982661724090576, + "learning_rate": 0.00019941262206704186, + "loss": 1.1137, + "step": 3660 + }, + { + "epoch": 0.2540420512108806, + "grad_norm": 3.386521339416504, + "learning_rate": 0.0001994120138249076, + "loss": 0.8661, + "step": 3661 + }, + { + "epoch": 0.25411144264797725, + "grad_norm": 4.36583948135376, + "learning_rate": 0.00019941140526894113, + "loss": 1.2872, + "step": 3662 + }, + { + "epoch": 0.2541808340850739, + "grad_norm": 3.549475908279419, + "learning_rate": 0.0001994107963991444, + "loss": 0.9418, + "step": 3663 + }, + { + "epoch": 0.2542502255221706, + "grad_norm": 4.959924697875977, + "learning_rate": 0.0001994101872155193, + "loss": 1.1699, + "step": 3664 + }, + { + "epoch": 0.25431961695926725, + "grad_norm": 3.2760119438171387, + "learning_rate": 0.00019940957771806778, + "loss": 1.1078, + "step": 3665 + }, + { + "epoch": 0.2543890083963639, + "grad_norm": 3.3682050704956055, + "learning_rate": 0.00019940896790679178, + "loss": 1.1386, + "step": 3666 + }, + { + "epoch": 0.2544583998334605, + "grad_norm": 3.098271131515503, + "learning_rate": 0.00019940835778169317, + "loss": 0.7792, + "step": 3667 + }, + { + "epoch": 0.2545277912705572, + "grad_norm": 3.8819446563720703, + "learning_rate": 0.00019940774734277393, + "loss": 0.9265, + "step": 3668 + }, + { + "epoch": 0.25459718270765386, + "grad_norm": 4.3517680168151855, + "learning_rate": 0.00019940713659003596, + "loss": 1.1417, + "step": 3669 + }, + { + "epoch": 0.2546665741447505, + "grad_norm": 3.2556416988372803, + "learning_rate": 0.0001994065255234812, + "loss": 0.7846, + "step": 3670 + }, + { + "epoch": 0.2547359655818472, + "grad_norm": 3.4393749237060547, + "learning_rate": 0.00019940591414311157, + "loss": 1.1528, + "step": 3671 + }, + { + "epoch": 0.25480535701894386, + "grad_norm": 5.265082836151123, + "learning_rate": 0.00019940530244892899, + "loss": 0.9914, + "step": 3672 + }, + { + "epoch": 0.2548747484560405, + "grad_norm": 2.5782058238983154, + "learning_rate": 0.0001994046904409354, + "loss": 0.6957, + "step": 3673 + }, + { + "epoch": 0.2549441398931372, + "grad_norm": 3.20662522315979, + "learning_rate": 0.00019940407811913272, + "loss": 1.0759, + "step": 3674 + }, + { + "epoch": 0.25501353133023386, + "grad_norm": 3.5159425735473633, + "learning_rate": 0.00019940346548352294, + "loss": 1.3723, + "step": 3675 + }, + { + "epoch": 0.2550829227673305, + "grad_norm": 3.1071715354919434, + "learning_rate": 0.00019940285253410792, + "loss": 1.1652, + "step": 3676 + }, + { + "epoch": 0.2551523142044272, + "grad_norm": 3.04958176612854, + "learning_rate": 0.00019940223927088967, + "loss": 1.0936, + "step": 3677 + }, + { + "epoch": 0.25522170564152385, + "grad_norm": 3.165506362915039, + "learning_rate": 0.00019940162569387004, + "loss": 1.048, + "step": 3678 + }, + { + "epoch": 0.2552910970786205, + "grad_norm": 3.1636412143707275, + "learning_rate": 0.00019940101180305103, + "loss": 1.0307, + "step": 3679 + }, + { + "epoch": 0.2553604885157172, + "grad_norm": 2.8847975730895996, + "learning_rate": 0.00019940039759843456, + "loss": 0.7607, + "step": 3680 + }, + { + "epoch": 0.2554298799528138, + "grad_norm": 3.214221239089966, + "learning_rate": 0.00019939978308002257, + "loss": 1.0338, + "step": 3681 + }, + { + "epoch": 0.25549927138991047, + "grad_norm": 3.7814040184020996, + "learning_rate": 0.000199399168247817, + "loss": 0.9965, + "step": 3682 + }, + { + "epoch": 0.25556866282700713, + "grad_norm": 3.529507637023926, + "learning_rate": 0.00019939855310181976, + "loss": 1.2651, + "step": 3683 + }, + { + "epoch": 0.2556380542641038, + "grad_norm": 2.8920068740844727, + "learning_rate": 0.00019939793764203284, + "loss": 0.8092, + "step": 3684 + }, + { + "epoch": 0.25570744570120046, + "grad_norm": 4.898443698883057, + "learning_rate": 0.00019939732186845816, + "loss": 1.4073, + "step": 3685 + }, + { + "epoch": 0.25577683713829713, + "grad_norm": 2.9257707595825195, + "learning_rate": 0.00019939670578109767, + "loss": 1.1228, + "step": 3686 + }, + { + "epoch": 0.2558462285753938, + "grad_norm": 3.5435285568237305, + "learning_rate": 0.00019939608937995332, + "loss": 0.843, + "step": 3687 + }, + { + "epoch": 0.25591562001249046, + "grad_norm": 3.383101224899292, + "learning_rate": 0.00019939547266502702, + "loss": 0.9675, + "step": 3688 + }, + { + "epoch": 0.25598501144958713, + "grad_norm": 7.608818054199219, + "learning_rate": 0.00019939485563632076, + "loss": 0.8279, + "step": 3689 + }, + { + "epoch": 0.2560544028866838, + "grad_norm": 3.8429439067840576, + "learning_rate": 0.00019939423829383647, + "loss": 0.987, + "step": 3690 + }, + { + "epoch": 0.25612379432378046, + "grad_norm": 5.013918399810791, + "learning_rate": 0.00019939362063757613, + "loss": 1.0899, + "step": 3691 + }, + { + "epoch": 0.25619318576087713, + "grad_norm": 3.473099708557129, + "learning_rate": 0.00019939300266754164, + "loss": 1.0121, + "step": 3692 + }, + { + "epoch": 0.2562625771979738, + "grad_norm": 2.708664894104004, + "learning_rate": 0.00019939238438373494, + "loss": 0.916, + "step": 3693 + }, + { + "epoch": 0.2563319686350704, + "grad_norm": 5.054717063903809, + "learning_rate": 0.00019939176578615805, + "loss": 1.1045, + "step": 3694 + }, + { + "epoch": 0.2564013600721671, + "grad_norm": 3.1168973445892334, + "learning_rate": 0.00019939114687481292, + "loss": 0.9626, + "step": 3695 + }, + { + "epoch": 0.25647075150926374, + "grad_norm": 3.3108043670654297, + "learning_rate": 0.0001993905276497014, + "loss": 1.1106, + "step": 3696 + }, + { + "epoch": 0.2565401429463604, + "grad_norm": 3.032230854034424, + "learning_rate": 0.00019938990811082556, + "loss": 1.2366, + "step": 3697 + }, + { + "epoch": 0.25660953438345707, + "grad_norm": 4.052276134490967, + "learning_rate": 0.0001993892882581873, + "loss": 1.3054, + "step": 3698 + }, + { + "epoch": 0.25667892582055374, + "grad_norm": 3.8417041301727295, + "learning_rate": 0.0001993886680917886, + "loss": 1.3154, + "step": 3699 + }, + { + "epoch": 0.2567483172576504, + "grad_norm": 3.5405099391937256, + "learning_rate": 0.0001993880476116314, + "loss": 1.0907, + "step": 3700 + }, + { + "epoch": 0.25681770869474707, + "grad_norm": 3.784806966781616, + "learning_rate": 0.00019938742681771765, + "loss": 1.2555, + "step": 3701 + }, + { + "epoch": 0.25688710013184374, + "grad_norm": 3.706550359725952, + "learning_rate": 0.00019938680571004932, + "loss": 1.1277, + "step": 3702 + }, + { + "epoch": 0.2569564915689404, + "grad_norm": 3.8075718879699707, + "learning_rate": 0.00019938618428862838, + "loss": 1.188, + "step": 3703 + }, + { + "epoch": 0.25702588300603707, + "grad_norm": 4.12220573425293, + "learning_rate": 0.0001993855625534568, + "loss": 1.2977, + "step": 3704 + }, + { + "epoch": 0.25709527444313374, + "grad_norm": 4.487102508544922, + "learning_rate": 0.00019938494050453654, + "loss": 1.1229, + "step": 3705 + }, + { + "epoch": 0.2571646658802304, + "grad_norm": 2.8539068698883057, + "learning_rate": 0.00019938431814186953, + "loss": 0.6809, + "step": 3706 + }, + { + "epoch": 0.25723405731732707, + "grad_norm": 3.5491583347320557, + "learning_rate": 0.00019938369546545776, + "loss": 1.0787, + "step": 3707 + }, + { + "epoch": 0.2573034487544237, + "grad_norm": 3.3368918895721436, + "learning_rate": 0.00019938307247530324, + "loss": 1.0554, + "step": 3708 + }, + { + "epoch": 0.25737284019152035, + "grad_norm": 3.9917469024658203, + "learning_rate": 0.00019938244917140784, + "loss": 0.8067, + "step": 3709 + }, + { + "epoch": 0.257442231628617, + "grad_norm": 3.9892611503601074, + "learning_rate": 0.00019938182555377358, + "loss": 1.2013, + "step": 3710 + }, + { + "epoch": 0.2575116230657137, + "grad_norm": 2.60569429397583, + "learning_rate": 0.00019938120162240247, + "loss": 1.0026, + "step": 3711 + }, + { + "epoch": 0.25758101450281035, + "grad_norm": 3.769470691680908, + "learning_rate": 0.00019938057737729637, + "loss": 0.992, + "step": 3712 + }, + { + "epoch": 0.257650405939907, + "grad_norm": 3.01462459564209, + "learning_rate": 0.00019937995281845737, + "loss": 1.1233, + "step": 3713 + }, + { + "epoch": 0.2577197973770037, + "grad_norm": 3.2664945125579834, + "learning_rate": 0.00019937932794588737, + "loss": 1.0393, + "step": 3714 + }, + { + "epoch": 0.25778918881410035, + "grad_norm": 3.535313606262207, + "learning_rate": 0.00019937870275958837, + "loss": 0.9263, + "step": 3715 + }, + { + "epoch": 0.257858580251197, + "grad_norm": 3.6863999366760254, + "learning_rate": 0.00019937807725956233, + "loss": 1.1528, + "step": 3716 + }, + { + "epoch": 0.2579279716882937, + "grad_norm": 5.838938236236572, + "learning_rate": 0.00019937745144581122, + "loss": 1.1731, + "step": 3717 + }, + { + "epoch": 0.25799736312539034, + "grad_norm": 3.021812677383423, + "learning_rate": 0.00019937682531833702, + "loss": 1.0377, + "step": 3718 + }, + { + "epoch": 0.258066754562487, + "grad_norm": 4.053020477294922, + "learning_rate": 0.00019937619887714175, + "loss": 1.0767, + "step": 3719 + }, + { + "epoch": 0.2581361459995837, + "grad_norm": 3.0412673950195312, + "learning_rate": 0.00019937557212222731, + "loss": 1.0652, + "step": 3720 + }, + { + "epoch": 0.2582055374366803, + "grad_norm": 3.6762871742248535, + "learning_rate": 0.00019937494505359573, + "loss": 1.0345, + "step": 3721 + }, + { + "epoch": 0.25827492887377695, + "grad_norm": 4.6063008308410645, + "learning_rate": 0.000199374317671249, + "loss": 1.1069, + "step": 3722 + }, + { + "epoch": 0.2583443203108736, + "grad_norm": 3.235860586166382, + "learning_rate": 0.00019937368997518905, + "loss": 1.2487, + "step": 3723 + }, + { + "epoch": 0.2584137117479703, + "grad_norm": 2.5784056186676025, + "learning_rate": 0.0001993730619654179, + "loss": 0.8937, + "step": 3724 + }, + { + "epoch": 0.25848310318506695, + "grad_norm": 3.2642624378204346, + "learning_rate": 0.00019937243364193753, + "loss": 0.8879, + "step": 3725 + }, + { + "epoch": 0.2585524946221636, + "grad_norm": 2.987683057785034, + "learning_rate": 0.00019937180500474992, + "loss": 0.9282, + "step": 3726 + }, + { + "epoch": 0.2586218860592603, + "grad_norm": 3.0965359210968018, + "learning_rate": 0.00019937117605385704, + "loss": 0.7949, + "step": 3727 + }, + { + "epoch": 0.25869127749635695, + "grad_norm": 2.210616111755371, + "learning_rate": 0.0001993705467892609, + "loss": 0.7503, + "step": 3728 + }, + { + "epoch": 0.2587606689334536, + "grad_norm": 4.527703762054443, + "learning_rate": 0.00019936991721096345, + "loss": 0.9506, + "step": 3729 + }, + { + "epoch": 0.2588300603705503, + "grad_norm": 4.00172758102417, + "learning_rate": 0.0001993692873189667, + "loss": 1.29, + "step": 3730 + }, + { + "epoch": 0.25889945180764695, + "grad_norm": 3.937716484069824, + "learning_rate": 0.00019936865711327268, + "loss": 0.9666, + "step": 3731 + }, + { + "epoch": 0.2589688432447436, + "grad_norm": 3.8468480110168457, + "learning_rate": 0.0001993680265938833, + "loss": 1.1312, + "step": 3732 + }, + { + "epoch": 0.2590382346818403, + "grad_norm": 5.012948513031006, + "learning_rate": 0.0001993673957608006, + "loss": 1.3148, + "step": 3733 + }, + { + "epoch": 0.2591076261189369, + "grad_norm": 3.2295804023742676, + "learning_rate": 0.00019936676461402658, + "loss": 1.1703, + "step": 3734 + }, + { + "epoch": 0.25917701755603356, + "grad_norm": 3.147955894470215, + "learning_rate": 0.00019936613315356318, + "loss": 1.0546, + "step": 3735 + }, + { + "epoch": 0.25924640899313023, + "grad_norm": 2.9836459159851074, + "learning_rate": 0.00019936550137941242, + "loss": 0.8678, + "step": 3736 + }, + { + "epoch": 0.2593158004302269, + "grad_norm": 4.147982120513916, + "learning_rate": 0.00019936486929157633, + "loss": 1.0395, + "step": 3737 + }, + { + "epoch": 0.25938519186732356, + "grad_norm": 2.663539171218872, + "learning_rate": 0.00019936423689005686, + "loss": 0.9373, + "step": 3738 + }, + { + "epoch": 0.2594545833044202, + "grad_norm": 3.21981143951416, + "learning_rate": 0.00019936360417485601, + "loss": 0.9009, + "step": 3739 + }, + { + "epoch": 0.2595239747415169, + "grad_norm": 3.829648733139038, + "learning_rate": 0.00019936297114597583, + "loss": 1.3323, + "step": 3740 + }, + { + "epoch": 0.25959336617861356, + "grad_norm": 3.069566249847412, + "learning_rate": 0.00019936233780341826, + "loss": 1.2279, + "step": 3741 + }, + { + "epoch": 0.2596627576157102, + "grad_norm": 4.6706342697143555, + "learning_rate": 0.0001993617041471853, + "loss": 1.1316, + "step": 3742 + }, + { + "epoch": 0.2597321490528069, + "grad_norm": 3.923504114151001, + "learning_rate": 0.00019936107017727898, + "loss": 1.0507, + "step": 3743 + }, + { + "epoch": 0.25980154048990356, + "grad_norm": 3.16489315032959, + "learning_rate": 0.0001993604358937013, + "loss": 1.062, + "step": 3744 + }, + { + "epoch": 0.2598709319270002, + "grad_norm": 3.8881754875183105, + "learning_rate": 0.00019935980129645424, + "loss": 1.0867, + "step": 3745 + }, + { + "epoch": 0.2599403233640969, + "grad_norm": 3.7964422702789307, + "learning_rate": 0.0001993591663855398, + "loss": 1.2061, + "step": 3746 + }, + { + "epoch": 0.26000971480119356, + "grad_norm": 5.0767645835876465, + "learning_rate": 0.00019935853116096003, + "loss": 1.3711, + "step": 3747 + }, + { + "epoch": 0.26007910623829017, + "grad_norm": 3.862492322921753, + "learning_rate": 0.00019935789562271688, + "loss": 0.863, + "step": 3748 + }, + { + "epoch": 0.26014849767538684, + "grad_norm": 3.2564034461975098, + "learning_rate": 0.00019935725977081241, + "loss": 0.8445, + "step": 3749 + }, + { + "epoch": 0.2602178891124835, + "grad_norm": 3.5380890369415283, + "learning_rate": 0.00019935662360524858, + "loss": 1.1838, + "step": 3750 + }, + { + "epoch": 0.26028728054958017, + "grad_norm": 4.70656681060791, + "learning_rate": 0.00019935598712602742, + "loss": 1.0424, + "step": 3751 + }, + { + "epoch": 0.26035667198667684, + "grad_norm": 4.114934921264648, + "learning_rate": 0.00019935535033315094, + "loss": 1.0971, + "step": 3752 + }, + { + "epoch": 0.2604260634237735, + "grad_norm": 4.882083892822266, + "learning_rate": 0.00019935471322662113, + "loss": 1.0222, + "step": 3753 + }, + { + "epoch": 0.26049545486087017, + "grad_norm": 4.932382583618164, + "learning_rate": 0.00019935407580644004, + "loss": 0.9994, + "step": 3754 + }, + { + "epoch": 0.26056484629796683, + "grad_norm": 3.2112648487091064, + "learning_rate": 0.00019935343807260964, + "loss": 0.8984, + "step": 3755 + }, + { + "epoch": 0.2606342377350635, + "grad_norm": 3.574626922607422, + "learning_rate": 0.00019935280002513197, + "loss": 0.7989, + "step": 3756 + }, + { + "epoch": 0.26070362917216017, + "grad_norm": 4.428194522857666, + "learning_rate": 0.00019935216166400905, + "loss": 1.3471, + "step": 3757 + }, + { + "epoch": 0.26077302060925683, + "grad_norm": 3.6684834957122803, + "learning_rate": 0.00019935152298924288, + "loss": 0.9745, + "step": 3758 + }, + { + "epoch": 0.2608424120463535, + "grad_norm": 4.311000347137451, + "learning_rate": 0.00019935088400083547, + "loss": 1.0294, + "step": 3759 + }, + { + "epoch": 0.26091180348345017, + "grad_norm": 4.509703159332275, + "learning_rate": 0.00019935024469878884, + "loss": 1.2382, + "step": 3760 + }, + { + "epoch": 0.2609811949205468, + "grad_norm": 4.851202964782715, + "learning_rate": 0.00019934960508310502, + "loss": 1.3743, + "step": 3761 + }, + { + "epoch": 0.26105058635764344, + "grad_norm": 3.8669769763946533, + "learning_rate": 0.000199348965153786, + "loss": 1.3036, + "step": 3762 + }, + { + "epoch": 0.2611199777947401, + "grad_norm": 4.36993408203125, + "learning_rate": 0.00019934832491083385, + "loss": 1.2204, + "step": 3763 + }, + { + "epoch": 0.2611893692318368, + "grad_norm": 4.8059258460998535, + "learning_rate": 0.00019934768435425057, + "loss": 1.1878, + "step": 3764 + }, + { + "epoch": 0.26125876066893344, + "grad_norm": 3.2609620094299316, + "learning_rate": 0.00019934704348403818, + "loss": 1.0532, + "step": 3765 + }, + { + "epoch": 0.2613281521060301, + "grad_norm": 3.686630964279175, + "learning_rate": 0.0001993464023001987, + "loss": 0.8591, + "step": 3766 + }, + { + "epoch": 0.2613975435431268, + "grad_norm": 3.624081611633301, + "learning_rate": 0.00019934576080273412, + "loss": 1.0169, + "step": 3767 + }, + { + "epoch": 0.26146693498022344, + "grad_norm": 3.7764229774475098, + "learning_rate": 0.00019934511899164653, + "loss": 1.2444, + "step": 3768 + }, + { + "epoch": 0.2615363264173201, + "grad_norm": 3.788804292678833, + "learning_rate": 0.0001993444768669379, + "loss": 1.0991, + "step": 3769 + }, + { + "epoch": 0.2616057178544168, + "grad_norm": 3.34633731842041, + "learning_rate": 0.00019934383442861032, + "loss": 1.1744, + "step": 3770 + }, + { + "epoch": 0.26167510929151344, + "grad_norm": 4.140806674957275, + "learning_rate": 0.00019934319167666574, + "loss": 1.2703, + "step": 3771 + }, + { + "epoch": 0.2617445007286101, + "grad_norm": 5.859837532043457, + "learning_rate": 0.00019934254861110628, + "loss": 1.2505, + "step": 3772 + }, + { + "epoch": 0.2618138921657068, + "grad_norm": 3.0529465675354004, + "learning_rate": 0.0001993419052319339, + "loss": 0.9155, + "step": 3773 + }, + { + "epoch": 0.26188328360280344, + "grad_norm": 3.1921164989471436, + "learning_rate": 0.00019934126153915065, + "loss": 0.8485, + "step": 3774 + }, + { + "epoch": 0.26195267503990005, + "grad_norm": 4.935092449188232, + "learning_rate": 0.00019934061753275854, + "loss": 1.2816, + "step": 3775 + }, + { + "epoch": 0.2620220664769967, + "grad_norm": 5.017227649688721, + "learning_rate": 0.00019933997321275966, + "loss": 0.936, + "step": 3776 + }, + { + "epoch": 0.2620914579140934, + "grad_norm": 3.5585248470306396, + "learning_rate": 0.00019933932857915597, + "loss": 1.0545, + "step": 3777 + }, + { + "epoch": 0.26216084935119005, + "grad_norm": 3.345987558364868, + "learning_rate": 0.00019933868363194957, + "loss": 0.9529, + "step": 3778 + }, + { + "epoch": 0.2622302407882867, + "grad_norm": 4.154635429382324, + "learning_rate": 0.00019933803837114247, + "loss": 1.2526, + "step": 3779 + }, + { + "epoch": 0.2622996322253834, + "grad_norm": 4.5874924659729, + "learning_rate": 0.0001993373927967367, + "loss": 1.2935, + "step": 3780 + }, + { + "epoch": 0.26236902366248005, + "grad_norm": 4.0265326499938965, + "learning_rate": 0.00019933674690873431, + "loss": 0.9218, + "step": 3781 + }, + { + "epoch": 0.2624384150995767, + "grad_norm": 3.646057367324829, + "learning_rate": 0.00019933610070713736, + "loss": 1.3218, + "step": 3782 + }, + { + "epoch": 0.2625078065366734, + "grad_norm": 3.795004367828369, + "learning_rate": 0.00019933545419194783, + "loss": 0.8075, + "step": 3783 + }, + { + "epoch": 0.26257719797377005, + "grad_norm": 3.5258076190948486, + "learning_rate": 0.00019933480736316784, + "loss": 0.8137, + "step": 3784 + }, + { + "epoch": 0.2626465894108667, + "grad_norm": 5.300457954406738, + "learning_rate": 0.00019933416022079934, + "loss": 1.2203, + "step": 3785 + }, + { + "epoch": 0.2627159808479634, + "grad_norm": 3.3851230144500732, + "learning_rate": 0.00019933351276484444, + "loss": 0.9417, + "step": 3786 + }, + { + "epoch": 0.26278537228506005, + "grad_norm": 5.005582332611084, + "learning_rate": 0.00019933286499530516, + "loss": 1.162, + "step": 3787 + }, + { + "epoch": 0.26285476372215666, + "grad_norm": 3.211094379425049, + "learning_rate": 0.00019933221691218354, + "loss": 0.8235, + "step": 3788 + }, + { + "epoch": 0.2629241551592533, + "grad_norm": 3.9116032123565674, + "learning_rate": 0.00019933156851548165, + "loss": 0.9083, + "step": 3789 + }, + { + "epoch": 0.26299354659635, + "grad_norm": 3.4046125411987305, + "learning_rate": 0.00019933091980520155, + "loss": 1.067, + "step": 3790 + }, + { + "epoch": 0.26306293803344666, + "grad_norm": 3.421079397201538, + "learning_rate": 0.00019933027078134524, + "loss": 0.8411, + "step": 3791 + }, + { + "epoch": 0.2631323294705433, + "grad_norm": 4.173212051391602, + "learning_rate": 0.00019932962144391476, + "loss": 1.0656, + "step": 3792 + }, + { + "epoch": 0.26320172090764, + "grad_norm": 3.1682796478271484, + "learning_rate": 0.0001993289717929122, + "loss": 0.7473, + "step": 3793 + }, + { + "epoch": 0.26327111234473666, + "grad_norm": 3.6109044551849365, + "learning_rate": 0.0001993283218283396, + "loss": 1.2788, + "step": 3794 + }, + { + "epoch": 0.2633405037818333, + "grad_norm": 3.1715357303619385, + "learning_rate": 0.00019932767155019906, + "loss": 0.8762, + "step": 3795 + }, + { + "epoch": 0.26340989521893, + "grad_norm": 2.998180627822876, + "learning_rate": 0.00019932702095849255, + "loss": 0.8623, + "step": 3796 + }, + { + "epoch": 0.26347928665602666, + "grad_norm": 3.2704529762268066, + "learning_rate": 0.00019932637005322214, + "loss": 1.138, + "step": 3797 + }, + { + "epoch": 0.2635486780931233, + "grad_norm": 4.5656232833862305, + "learning_rate": 0.00019932571883438994, + "loss": 1.1344, + "step": 3798 + }, + { + "epoch": 0.26361806953022, + "grad_norm": 4.254693031311035, + "learning_rate": 0.00019932506730199794, + "loss": 1.1692, + "step": 3799 + }, + { + "epoch": 0.26368746096731666, + "grad_norm": 4.214611530303955, + "learning_rate": 0.00019932441545604824, + "loss": 1.1833, + "step": 3800 + }, + { + "epoch": 0.2637568524044133, + "grad_norm": 6.610746383666992, + "learning_rate": 0.00019932376329654286, + "loss": 0.7392, + "step": 3801 + }, + { + "epoch": 0.26382624384150993, + "grad_norm": 3.5148279666900635, + "learning_rate": 0.0001993231108234839, + "loss": 0.9526, + "step": 3802 + }, + { + "epoch": 0.2638956352786066, + "grad_norm": 3.4493606090545654, + "learning_rate": 0.00019932245803687344, + "loss": 1.0302, + "step": 3803 + }, + { + "epoch": 0.26396502671570327, + "grad_norm": 2.971811056137085, + "learning_rate": 0.00019932180493671347, + "loss": 1.0342, + "step": 3804 + }, + { + "epoch": 0.26403441815279993, + "grad_norm": 3.2211525440216064, + "learning_rate": 0.0001993211515230061, + "loss": 0.96, + "step": 3805 + }, + { + "epoch": 0.2641038095898966, + "grad_norm": 3.185149669647217, + "learning_rate": 0.00019932049779575336, + "loss": 0.8493, + "step": 3806 + }, + { + "epoch": 0.26417320102699327, + "grad_norm": 3.247252941131592, + "learning_rate": 0.0001993198437549573, + "loss": 1.2001, + "step": 3807 + }, + { + "epoch": 0.26424259246408993, + "grad_norm": 2.9938275814056396, + "learning_rate": 0.00019931918940062007, + "loss": 0.9775, + "step": 3808 + }, + { + "epoch": 0.2643119839011866, + "grad_norm": 3.4722745418548584, + "learning_rate": 0.00019931853473274366, + "loss": 1.3601, + "step": 3809 + }, + { + "epoch": 0.26438137533828326, + "grad_norm": 5.660181999206543, + "learning_rate": 0.00019931787975133018, + "loss": 1.2244, + "step": 3810 + }, + { + "epoch": 0.26445076677537993, + "grad_norm": 2.2952966690063477, + "learning_rate": 0.00019931722445638167, + "loss": 0.8226, + "step": 3811 + }, + { + "epoch": 0.2645201582124766, + "grad_norm": 2.5052263736724854, + "learning_rate": 0.00019931656884790018, + "loss": 0.9619, + "step": 3812 + }, + { + "epoch": 0.26458954964957326, + "grad_norm": 3.2129034996032715, + "learning_rate": 0.00019931591292588784, + "loss": 1.1925, + "step": 3813 + }, + { + "epoch": 0.26465894108666993, + "grad_norm": 3.910841941833496, + "learning_rate": 0.00019931525669034664, + "loss": 0.881, + "step": 3814 + }, + { + "epoch": 0.26472833252376654, + "grad_norm": 8.537650108337402, + "learning_rate": 0.00019931460014127874, + "loss": 1.1203, + "step": 3815 + }, + { + "epoch": 0.2647977239608632, + "grad_norm": 3.129019260406494, + "learning_rate": 0.00019931394327868615, + "loss": 0.7694, + "step": 3816 + }, + { + "epoch": 0.2648671153979599, + "grad_norm": 4.111045837402344, + "learning_rate": 0.00019931328610257095, + "loss": 1.2361, + "step": 3817 + }, + { + "epoch": 0.26493650683505654, + "grad_norm": 5.612663745880127, + "learning_rate": 0.00019931262861293527, + "loss": 0.9665, + "step": 3818 + }, + { + "epoch": 0.2650058982721532, + "grad_norm": 4.5804643630981445, + "learning_rate": 0.00019931197080978112, + "loss": 1.1403, + "step": 3819 + }, + { + "epoch": 0.2650752897092499, + "grad_norm": 3.5047357082366943, + "learning_rate": 0.00019931131269311058, + "loss": 0.9826, + "step": 3820 + }, + { + "epoch": 0.26514468114634654, + "grad_norm": 4.506104469299316, + "learning_rate": 0.00019931065426292578, + "loss": 1.3636, + "step": 3821 + }, + { + "epoch": 0.2652140725834432, + "grad_norm": 4.5272111892700195, + "learning_rate": 0.00019930999551922876, + "loss": 1.1729, + "step": 3822 + }, + { + "epoch": 0.26528346402053987, + "grad_norm": 3.796682596206665, + "learning_rate": 0.0001993093364620216, + "loss": 0.9488, + "step": 3823 + }, + { + "epoch": 0.26535285545763654, + "grad_norm": 4.545696258544922, + "learning_rate": 0.0001993086770913064, + "loss": 1.1583, + "step": 3824 + }, + { + "epoch": 0.2654222468947332, + "grad_norm": 4.475977897644043, + "learning_rate": 0.0001993080174070852, + "loss": 1.1864, + "step": 3825 + }, + { + "epoch": 0.26549163833182987, + "grad_norm": 3.738201856613159, + "learning_rate": 0.00019930735740936016, + "loss": 1.1073, + "step": 3826 + }, + { + "epoch": 0.26556102976892654, + "grad_norm": 3.827052354812622, + "learning_rate": 0.00019930669709813327, + "loss": 1.2857, + "step": 3827 + }, + { + "epoch": 0.2656304212060232, + "grad_norm": 4.010704517364502, + "learning_rate": 0.00019930603647340666, + "loss": 1.0556, + "step": 3828 + }, + { + "epoch": 0.2656998126431198, + "grad_norm": 4.048513889312744, + "learning_rate": 0.00019930537553518242, + "loss": 1.057, + "step": 3829 + }, + { + "epoch": 0.2657692040802165, + "grad_norm": 3.379070997238159, + "learning_rate": 0.00019930471428346265, + "loss": 1.149, + "step": 3830 + }, + { + "epoch": 0.26583859551731315, + "grad_norm": 3.0031697750091553, + "learning_rate": 0.0001993040527182494, + "loss": 1.2069, + "step": 3831 + }, + { + "epoch": 0.2659079869544098, + "grad_norm": 2.750767469406128, + "learning_rate": 0.0001993033908395448, + "loss": 0.8003, + "step": 3832 + }, + { + "epoch": 0.2659773783915065, + "grad_norm": 2.9584648609161377, + "learning_rate": 0.00019930272864735087, + "loss": 1.0986, + "step": 3833 + }, + { + "epoch": 0.26604676982860315, + "grad_norm": 5.008612632751465, + "learning_rate": 0.0001993020661416698, + "loss": 1.2102, + "step": 3834 + }, + { + "epoch": 0.2661161612656998, + "grad_norm": 4.317725658416748, + "learning_rate": 0.00019930140332250358, + "loss": 1.3015, + "step": 3835 + }, + { + "epoch": 0.2661855527027965, + "grad_norm": 3.0627524852752686, + "learning_rate": 0.0001993007401898544, + "loss": 1.1852, + "step": 3836 + }, + { + "epoch": 0.26625494413989315, + "grad_norm": 2.8463149070739746, + "learning_rate": 0.00019930007674372427, + "loss": 0.7768, + "step": 3837 + }, + { + "epoch": 0.2663243355769898, + "grad_norm": 2.4597885608673096, + "learning_rate": 0.0001992994129841153, + "loss": 0.7994, + "step": 3838 + }, + { + "epoch": 0.2663937270140865, + "grad_norm": 2.91518235206604, + "learning_rate": 0.00019929874891102967, + "loss": 1.1419, + "step": 3839 + }, + { + "epoch": 0.26646311845118315, + "grad_norm": 4.908204078674316, + "learning_rate": 0.00019929808452446936, + "loss": 1.1538, + "step": 3840 + }, + { + "epoch": 0.2665325098882798, + "grad_norm": 3.948665142059326, + "learning_rate": 0.00019929741982443653, + "loss": 1.3323, + "step": 3841 + }, + { + "epoch": 0.2666019013253764, + "grad_norm": 3.76802134513855, + "learning_rate": 0.00019929675481093325, + "loss": 1.1575, + "step": 3842 + }, + { + "epoch": 0.2666712927624731, + "grad_norm": 4.066969394683838, + "learning_rate": 0.00019929608948396165, + "loss": 1.2045, + "step": 3843 + }, + { + "epoch": 0.26674068419956976, + "grad_norm": 3.2822718620300293, + "learning_rate": 0.00019929542384352383, + "loss": 0.9895, + "step": 3844 + }, + { + "epoch": 0.2668100756366664, + "grad_norm": 3.7442188262939453, + "learning_rate": 0.00019929475788962186, + "loss": 1.1399, + "step": 3845 + }, + { + "epoch": 0.2668794670737631, + "grad_norm": 3.592768430709839, + "learning_rate": 0.00019929409162225788, + "loss": 1.0791, + "step": 3846 + }, + { + "epoch": 0.26694885851085975, + "grad_norm": 5.6375956535339355, + "learning_rate": 0.00019929342504143395, + "loss": 1.0956, + "step": 3847 + }, + { + "epoch": 0.2670182499479564, + "grad_norm": 4.811635971069336, + "learning_rate": 0.0001992927581471522, + "loss": 1.2249, + "step": 3848 + }, + { + "epoch": 0.2670876413850531, + "grad_norm": 2.9531781673431396, + "learning_rate": 0.00019929209093941475, + "loss": 1.0374, + "step": 3849 + }, + { + "epoch": 0.26715703282214975, + "grad_norm": 3.804356098175049, + "learning_rate": 0.00019929142341822367, + "loss": 1.3054, + "step": 3850 + }, + { + "epoch": 0.2672264242592464, + "grad_norm": 3.783818244934082, + "learning_rate": 0.0001992907555835811, + "loss": 1.0103, + "step": 3851 + }, + { + "epoch": 0.2672958156963431, + "grad_norm": 3.3869338035583496, + "learning_rate": 0.0001992900874354891, + "loss": 0.9728, + "step": 3852 + }, + { + "epoch": 0.26736520713343975, + "grad_norm": 3.3404529094696045, + "learning_rate": 0.00019928941897394986, + "loss": 0.9548, + "step": 3853 + }, + { + "epoch": 0.2674345985705364, + "grad_norm": 4.42744255065918, + "learning_rate": 0.00019928875019896542, + "loss": 1.1118, + "step": 3854 + }, + { + "epoch": 0.2675039900076331, + "grad_norm": 3.740314245223999, + "learning_rate": 0.00019928808111053793, + "loss": 1.0413, + "step": 3855 + }, + { + "epoch": 0.2675733814447297, + "grad_norm": 4.4759039878845215, + "learning_rate": 0.00019928741170866947, + "loss": 1.1005, + "step": 3856 + }, + { + "epoch": 0.26764277288182636, + "grad_norm": 4.109231472015381, + "learning_rate": 0.00019928674199336217, + "loss": 1.1769, + "step": 3857 + }, + { + "epoch": 0.26771216431892303, + "grad_norm": 3.435765266418457, + "learning_rate": 0.00019928607196461817, + "loss": 1.0031, + "step": 3858 + }, + { + "epoch": 0.2677815557560197, + "grad_norm": 3.509382724761963, + "learning_rate": 0.00019928540162243954, + "loss": 0.8656, + "step": 3859 + }, + { + "epoch": 0.26785094719311636, + "grad_norm": 2.5504274368286133, + "learning_rate": 0.00019928473096682842, + "loss": 1.098, + "step": 3860 + }, + { + "epoch": 0.26792033863021303, + "grad_norm": 3.6957173347473145, + "learning_rate": 0.00019928405999778692, + "loss": 1.3032, + "step": 3861 + }, + { + "epoch": 0.2679897300673097, + "grad_norm": 3.338993549346924, + "learning_rate": 0.00019928338871531715, + "loss": 1.0744, + "step": 3862 + }, + { + "epoch": 0.26805912150440636, + "grad_norm": 3.078434705734253, + "learning_rate": 0.00019928271711942128, + "loss": 0.716, + "step": 3863 + }, + { + "epoch": 0.268128512941503, + "grad_norm": 4.546936511993408, + "learning_rate": 0.00019928204521010138, + "loss": 0.9448, + "step": 3864 + }, + { + "epoch": 0.2681979043785997, + "grad_norm": 3.477739095687866, + "learning_rate": 0.00019928137298735955, + "loss": 1.1247, + "step": 3865 + }, + { + "epoch": 0.26826729581569636, + "grad_norm": 5.617781162261963, + "learning_rate": 0.00019928070045119794, + "loss": 1.017, + "step": 3866 + }, + { + "epoch": 0.268336687252793, + "grad_norm": 3.166886568069458, + "learning_rate": 0.00019928002760161872, + "loss": 0.9536, + "step": 3867 + }, + { + "epoch": 0.2684060786898897, + "grad_norm": 3.627246141433716, + "learning_rate": 0.00019927935443862395, + "loss": 1.0946, + "step": 3868 + }, + { + "epoch": 0.2684754701269863, + "grad_norm": 2.8498246669769287, + "learning_rate": 0.00019927868096221579, + "loss": 0.8021, + "step": 3869 + }, + { + "epoch": 0.26854486156408297, + "grad_norm": 3.2481956481933594, + "learning_rate": 0.00019927800717239633, + "loss": 1.199, + "step": 3870 + }, + { + "epoch": 0.26861425300117964, + "grad_norm": 3.265172243118286, + "learning_rate": 0.00019927733306916774, + "loss": 1.1758, + "step": 3871 + }, + { + "epoch": 0.2686836444382763, + "grad_norm": 3.047825336456299, + "learning_rate": 0.0001992766586525321, + "loss": 0.7389, + "step": 3872 + }, + { + "epoch": 0.26875303587537297, + "grad_norm": 3.03497314453125, + "learning_rate": 0.00019927598392249158, + "loss": 0.8769, + "step": 3873 + }, + { + "epoch": 0.26882242731246964, + "grad_norm": 2.959272623062134, + "learning_rate": 0.0001992753088790483, + "loss": 1.1441, + "step": 3874 + }, + { + "epoch": 0.2688918187495663, + "grad_norm": 3.3713290691375732, + "learning_rate": 0.0001992746335222044, + "loss": 0.9418, + "step": 3875 + }, + { + "epoch": 0.26896121018666297, + "grad_norm": 3.775909662246704, + "learning_rate": 0.000199273957851962, + "loss": 1.1634, + "step": 3876 + }, + { + "epoch": 0.26903060162375964, + "grad_norm": 3.622039318084717, + "learning_rate": 0.00019927328186832323, + "loss": 1.1407, + "step": 3877 + }, + { + "epoch": 0.2690999930608563, + "grad_norm": 3.3340084552764893, + "learning_rate": 0.00019927260557129023, + "loss": 1.2421, + "step": 3878 + }, + { + "epoch": 0.26916938449795297, + "grad_norm": 4.258200645446777, + "learning_rate": 0.0001992719289608651, + "loss": 1.1489, + "step": 3879 + }, + { + "epoch": 0.26923877593504963, + "grad_norm": 3.4168925285339355, + "learning_rate": 0.00019927125203705003, + "loss": 1.1826, + "step": 3880 + }, + { + "epoch": 0.2693081673721463, + "grad_norm": 3.585108757019043, + "learning_rate": 0.00019927057479984713, + "loss": 0.9669, + "step": 3881 + }, + { + "epoch": 0.2693775588092429, + "grad_norm": 3.515650987625122, + "learning_rate": 0.00019926989724925855, + "loss": 0.9525, + "step": 3882 + }, + { + "epoch": 0.2694469502463396, + "grad_norm": 2.7745611667633057, + "learning_rate": 0.00019926921938528643, + "loss": 0.9628, + "step": 3883 + }, + { + "epoch": 0.26951634168343624, + "grad_norm": 3.4276137351989746, + "learning_rate": 0.00019926854120793288, + "loss": 1.0955, + "step": 3884 + }, + { + "epoch": 0.2695857331205329, + "grad_norm": 3.9433937072753906, + "learning_rate": 0.00019926786271720008, + "loss": 1.4076, + "step": 3885 + }, + { + "epoch": 0.2696551245576296, + "grad_norm": 2.975775957107544, + "learning_rate": 0.00019926718391309015, + "loss": 0.7227, + "step": 3886 + }, + { + "epoch": 0.26972451599472624, + "grad_norm": 3.5578956604003906, + "learning_rate": 0.00019926650479560524, + "loss": 1.1074, + "step": 3887 + }, + { + "epoch": 0.2697939074318229, + "grad_norm": 2.4507832527160645, + "learning_rate": 0.0001992658253647475, + "loss": 0.9981, + "step": 3888 + }, + { + "epoch": 0.2698632988689196, + "grad_norm": 3.344695806503296, + "learning_rate": 0.00019926514562051904, + "loss": 1.0913, + "step": 3889 + }, + { + "epoch": 0.26993269030601624, + "grad_norm": 3.314547538757324, + "learning_rate": 0.00019926446556292206, + "loss": 1.216, + "step": 3890 + }, + { + "epoch": 0.2700020817431129, + "grad_norm": 3.830068588256836, + "learning_rate": 0.00019926378519195866, + "loss": 0.9042, + "step": 3891 + }, + { + "epoch": 0.2700714731802096, + "grad_norm": 3.2043187618255615, + "learning_rate": 0.00019926310450763102, + "loss": 0.6947, + "step": 3892 + }, + { + "epoch": 0.27014086461730624, + "grad_norm": 3.0196218490600586, + "learning_rate": 0.00019926242350994127, + "loss": 1.1671, + "step": 3893 + }, + { + "epoch": 0.2702102560544029, + "grad_norm": 3.4325437545776367, + "learning_rate": 0.00019926174219889157, + "loss": 1.2502, + "step": 3894 + }, + { + "epoch": 0.2702796474914996, + "grad_norm": 2.9328949451446533, + "learning_rate": 0.00019926106057448404, + "loss": 1.1228, + "step": 3895 + }, + { + "epoch": 0.2703490389285962, + "grad_norm": 3.0284416675567627, + "learning_rate": 0.00019926037863672087, + "loss": 1.2287, + "step": 3896 + }, + { + "epoch": 0.27041843036569285, + "grad_norm": 4.318721771240234, + "learning_rate": 0.0001992596963856042, + "loss": 1.0919, + "step": 3897 + }, + { + "epoch": 0.2704878218027895, + "grad_norm": 3.4554781913757324, + "learning_rate": 0.0001992590138211362, + "loss": 0.9885, + "step": 3898 + }, + { + "epoch": 0.2705572132398862, + "grad_norm": 3.9125561714172363, + "learning_rate": 0.000199258330943319, + "loss": 1.0272, + "step": 3899 + }, + { + "epoch": 0.27062660467698285, + "grad_norm": 3.3845670223236084, + "learning_rate": 0.00019925764775215478, + "loss": 1.0094, + "step": 3900 + }, + { + "epoch": 0.2706959961140795, + "grad_norm": 3.672781229019165, + "learning_rate": 0.00019925696424764566, + "loss": 1.0852, + "step": 3901 + }, + { + "epoch": 0.2707653875511762, + "grad_norm": 3.316335916519165, + "learning_rate": 0.00019925628042979384, + "loss": 0.937, + "step": 3902 + }, + { + "epoch": 0.27083477898827285, + "grad_norm": 3.6900601387023926, + "learning_rate": 0.00019925559629860144, + "loss": 1.0565, + "step": 3903 + }, + { + "epoch": 0.2709041704253695, + "grad_norm": 3.7490620613098145, + "learning_rate": 0.00019925491185407066, + "loss": 0.9319, + "step": 3904 + }, + { + "epoch": 0.2709735618624662, + "grad_norm": 3.5349879264831543, + "learning_rate": 0.00019925422709620363, + "loss": 1.0399, + "step": 3905 + }, + { + "epoch": 0.27104295329956285, + "grad_norm": 2.7419795989990234, + "learning_rate": 0.00019925354202500253, + "loss": 0.8651, + "step": 3906 + }, + { + "epoch": 0.2711123447366595, + "grad_norm": 3.39733624458313, + "learning_rate": 0.0001992528566404695, + "loss": 0.9502, + "step": 3907 + }, + { + "epoch": 0.2711817361737562, + "grad_norm": 3.035050868988037, + "learning_rate": 0.0001992521709426067, + "loss": 0.7975, + "step": 3908 + }, + { + "epoch": 0.2712511276108528, + "grad_norm": 5.676112174987793, + "learning_rate": 0.00019925148493141635, + "loss": 1.0286, + "step": 3909 + }, + { + "epoch": 0.27132051904794946, + "grad_norm": 3.5286006927490234, + "learning_rate": 0.00019925079860690054, + "loss": 0.9773, + "step": 3910 + }, + { + "epoch": 0.2713899104850461, + "grad_norm": 3.2764265537261963, + "learning_rate": 0.00019925011196906148, + "loss": 0.9658, + "step": 3911 + }, + { + "epoch": 0.2714593019221428, + "grad_norm": 3.8106696605682373, + "learning_rate": 0.00019924942501790133, + "loss": 0.839, + "step": 3912 + }, + { + "epoch": 0.27152869335923946, + "grad_norm": 3.548997402191162, + "learning_rate": 0.00019924873775342228, + "loss": 1.2087, + "step": 3913 + }, + { + "epoch": 0.2715980847963361, + "grad_norm": 3.4929611682891846, + "learning_rate": 0.0001992480501756265, + "loss": 1.1186, + "step": 3914 + }, + { + "epoch": 0.2716674762334328, + "grad_norm": 2.827775716781616, + "learning_rate": 0.00019924736228451608, + "loss": 0.8749, + "step": 3915 + }, + { + "epoch": 0.27173686767052946, + "grad_norm": 3.437845468521118, + "learning_rate": 0.0001992466740800933, + "loss": 1.0701, + "step": 3916 + }, + { + "epoch": 0.2718062591076261, + "grad_norm": 2.7950382232666016, + "learning_rate": 0.00019924598556236026, + "loss": 0.7987, + "step": 3917 + }, + { + "epoch": 0.2718756505447228, + "grad_norm": 4.259130954742432, + "learning_rate": 0.00019924529673131917, + "loss": 1.0189, + "step": 3918 + }, + { + "epoch": 0.27194504198181946, + "grad_norm": 3.045344114303589, + "learning_rate": 0.0001992446075869722, + "loss": 1.1599, + "step": 3919 + }, + { + "epoch": 0.2720144334189161, + "grad_norm": 3.1624767780303955, + "learning_rate": 0.0001992439181293215, + "loss": 0.9046, + "step": 3920 + }, + { + "epoch": 0.2720838248560128, + "grad_norm": 3.366907835006714, + "learning_rate": 0.00019924322835836929, + "loss": 1.0929, + "step": 3921 + }, + { + "epoch": 0.27215321629310946, + "grad_norm": 2.9017651081085205, + "learning_rate": 0.0001992425382741177, + "loss": 0.8531, + "step": 3922 + }, + { + "epoch": 0.27222260773020607, + "grad_norm": 4.526200771331787, + "learning_rate": 0.00019924184787656893, + "loss": 0.8912, + "step": 3923 + }, + { + "epoch": 0.27229199916730273, + "grad_norm": 4.237517833709717, + "learning_rate": 0.0001992411571657252, + "loss": 1.3233, + "step": 3924 + }, + { + "epoch": 0.2723613906043994, + "grad_norm": 3.8616082668304443, + "learning_rate": 0.0001992404661415886, + "loss": 1.262, + "step": 3925 + }, + { + "epoch": 0.27243078204149607, + "grad_norm": 3.060859441757202, + "learning_rate": 0.00019923977480416135, + "loss": 0.8127, + "step": 3926 + }, + { + "epoch": 0.27250017347859273, + "grad_norm": 5.442609786987305, + "learning_rate": 0.0001992390831534457, + "loss": 1.2278, + "step": 3927 + }, + { + "epoch": 0.2725695649156894, + "grad_norm": 3.9115397930145264, + "learning_rate": 0.00019923839118944376, + "loss": 0.9695, + "step": 3928 + }, + { + "epoch": 0.27263895635278607, + "grad_norm": 3.519395112991333, + "learning_rate": 0.00019923769891215773, + "loss": 1.0593, + "step": 3929 + }, + { + "epoch": 0.27270834778988273, + "grad_norm": 3.9693872928619385, + "learning_rate": 0.0001992370063215898, + "loss": 1.001, + "step": 3930 + }, + { + "epoch": 0.2727777392269794, + "grad_norm": 3.9552927017211914, + "learning_rate": 0.00019923631341774213, + "loss": 1.0337, + "step": 3931 + }, + { + "epoch": 0.27284713066407607, + "grad_norm": 3.836491107940674, + "learning_rate": 0.00019923562020061696, + "loss": 1.1114, + "step": 3932 + }, + { + "epoch": 0.27291652210117273, + "grad_norm": 4.8101701736450195, + "learning_rate": 0.00019923492667021643, + "loss": 0.9656, + "step": 3933 + }, + { + "epoch": 0.2729859135382694, + "grad_norm": 6.717007160186768, + "learning_rate": 0.00019923423282654278, + "loss": 1.2998, + "step": 3934 + }, + { + "epoch": 0.27305530497536606, + "grad_norm": 3.1727657318115234, + "learning_rate": 0.00019923353866959815, + "loss": 0.9, + "step": 3935 + }, + { + "epoch": 0.2731246964124627, + "grad_norm": 3.020327091217041, + "learning_rate": 0.00019923284419938475, + "loss": 0.7548, + "step": 3936 + }, + { + "epoch": 0.27319408784955934, + "grad_norm": 3.819084405899048, + "learning_rate": 0.00019923214941590479, + "loss": 1.01, + "step": 3937 + }, + { + "epoch": 0.273263479286656, + "grad_norm": 3.3985729217529297, + "learning_rate": 0.0001992314543191604, + "loss": 1.2666, + "step": 3938 + }, + { + "epoch": 0.2733328707237527, + "grad_norm": 4.303840637207031, + "learning_rate": 0.0001992307589091539, + "loss": 0.7989, + "step": 3939 + }, + { + "epoch": 0.27340226216084934, + "grad_norm": 3.3514089584350586, + "learning_rate": 0.00019923006318588735, + "loss": 0.8964, + "step": 3940 + }, + { + "epoch": 0.273471653597946, + "grad_norm": 4.5678229331970215, + "learning_rate": 0.00019922936714936304, + "loss": 1.3391, + "step": 3941 + }, + { + "epoch": 0.2735410450350427, + "grad_norm": 3.853269100189209, + "learning_rate": 0.0001992286707995831, + "loss": 1.2527, + "step": 3942 + }, + { + "epoch": 0.27361043647213934, + "grad_norm": 3.601008892059326, + "learning_rate": 0.00019922797413654975, + "loss": 1.1825, + "step": 3943 + }, + { + "epoch": 0.273679827909236, + "grad_norm": 4.357974052429199, + "learning_rate": 0.00019922727716026523, + "loss": 1.0263, + "step": 3944 + }, + { + "epoch": 0.2737492193463327, + "grad_norm": 3.5541043281555176, + "learning_rate": 0.00019922657987073167, + "loss": 1.1597, + "step": 3945 + }, + { + "epoch": 0.27381861078342934, + "grad_norm": 3.1879703998565674, + "learning_rate": 0.00019922588226795135, + "loss": 0.7655, + "step": 3946 + }, + { + "epoch": 0.273888002220526, + "grad_norm": 3.4267513751983643, + "learning_rate": 0.00019922518435192642, + "loss": 0.9518, + "step": 3947 + }, + { + "epoch": 0.27395739365762267, + "grad_norm": 4.357939720153809, + "learning_rate": 0.0001992244861226591, + "loss": 1.0124, + "step": 3948 + }, + { + "epoch": 0.27402678509471934, + "grad_norm": 3.6629416942596436, + "learning_rate": 0.00019922378758015157, + "loss": 0.8928, + "step": 3949 + }, + { + "epoch": 0.27409617653181595, + "grad_norm": 3.865980863571167, + "learning_rate": 0.00019922308872440608, + "loss": 0.9662, + "step": 3950 + }, + { + "epoch": 0.2741655679689126, + "grad_norm": 4.027549743652344, + "learning_rate": 0.0001992223895554248, + "loss": 1.1062, + "step": 3951 + }, + { + "epoch": 0.2742349594060093, + "grad_norm": 5.170684337615967, + "learning_rate": 0.00019922169007320991, + "loss": 0.9833, + "step": 3952 + }, + { + "epoch": 0.27430435084310595, + "grad_norm": 2.7509734630584717, + "learning_rate": 0.00019922099027776371, + "loss": 0.8866, + "step": 3953 + }, + { + "epoch": 0.2743737422802026, + "grad_norm": 4.384906768798828, + "learning_rate": 0.0001992202901690883, + "loss": 1.1594, + "step": 3954 + }, + { + "epoch": 0.2744431337172993, + "grad_norm": 4.23369026184082, + "learning_rate": 0.00019921958974718603, + "loss": 1.0564, + "step": 3955 + }, + { + "epoch": 0.27451252515439595, + "grad_norm": 2.5692508220672607, + "learning_rate": 0.00019921888901205896, + "loss": 0.8782, + "step": 3956 + }, + { + "epoch": 0.2745819165914926, + "grad_norm": 2.7907283306121826, + "learning_rate": 0.00019921818796370938, + "loss": 0.8402, + "step": 3957 + }, + { + "epoch": 0.2746513080285893, + "grad_norm": 5.480342388153076, + "learning_rate": 0.0001992174866021395, + "loss": 1.1669, + "step": 3958 + }, + { + "epoch": 0.27472069946568595, + "grad_norm": 3.4799437522888184, + "learning_rate": 0.00019921678492735155, + "loss": 1.0027, + "step": 3959 + }, + { + "epoch": 0.2747900909027826, + "grad_norm": 2.574784278869629, + "learning_rate": 0.0001992160829393477, + "loss": 0.8696, + "step": 3960 + }, + { + "epoch": 0.2748594823398793, + "grad_norm": 4.407836437225342, + "learning_rate": 0.00019921538063813016, + "loss": 1.0088, + "step": 3961 + }, + { + "epoch": 0.27492887377697595, + "grad_norm": 3.2159271240234375, + "learning_rate": 0.0001992146780237012, + "loss": 0.8994, + "step": 3962 + }, + { + "epoch": 0.27499826521407256, + "grad_norm": 2.6458141803741455, + "learning_rate": 0.000199213975096063, + "loss": 0.8982, + "step": 3963 + }, + { + "epoch": 0.2750676566511692, + "grad_norm": 3.9412498474121094, + "learning_rate": 0.0001992132718552178, + "loss": 1.1266, + "step": 3964 + }, + { + "epoch": 0.2751370480882659, + "grad_norm": 3.5731077194213867, + "learning_rate": 0.0001992125683011678, + "loss": 1.1091, + "step": 3965 + }, + { + "epoch": 0.27520643952536256, + "grad_norm": 3.947864532470703, + "learning_rate": 0.0001992118644339153, + "loss": 1.3346, + "step": 3966 + }, + { + "epoch": 0.2752758309624592, + "grad_norm": 5.752638816833496, + "learning_rate": 0.00019921116025346238, + "loss": 0.9926, + "step": 3967 + }, + { + "epoch": 0.2753452223995559, + "grad_norm": 4.138018608093262, + "learning_rate": 0.00019921045575981133, + "loss": 1.0482, + "step": 3968 + }, + { + "epoch": 0.27541461383665256, + "grad_norm": 2.866456985473633, + "learning_rate": 0.00019920975095296444, + "loss": 1.01, + "step": 3969 + }, + { + "epoch": 0.2754840052737492, + "grad_norm": 3.864849090576172, + "learning_rate": 0.00019920904583292382, + "loss": 1.003, + "step": 3970 + }, + { + "epoch": 0.2755533967108459, + "grad_norm": 3.2338759899139404, + "learning_rate": 0.0001992083403996918, + "loss": 1.193, + "step": 3971 + }, + { + "epoch": 0.27562278814794255, + "grad_norm": 2.8740274906158447, + "learning_rate": 0.00019920763465327055, + "loss": 0.9611, + "step": 3972 + }, + { + "epoch": 0.2756921795850392, + "grad_norm": 3.6587331295013428, + "learning_rate": 0.00019920692859366228, + "loss": 1.2732, + "step": 3973 + }, + { + "epoch": 0.2757615710221359, + "grad_norm": 3.4282455444335938, + "learning_rate": 0.00019920622222086926, + "loss": 0.959, + "step": 3974 + }, + { + "epoch": 0.27583096245923255, + "grad_norm": 3.432769298553467, + "learning_rate": 0.00019920551553489373, + "loss": 1.3135, + "step": 3975 + }, + { + "epoch": 0.2759003538963292, + "grad_norm": 3.7817459106445312, + "learning_rate": 0.00019920480853573787, + "loss": 1.2757, + "step": 3976 + }, + { + "epoch": 0.27596974533342583, + "grad_norm": 4.015946388244629, + "learning_rate": 0.00019920410122340395, + "loss": 0.9221, + "step": 3977 + }, + { + "epoch": 0.2760391367705225, + "grad_norm": 3.309220790863037, + "learning_rate": 0.0001992033935978942, + "loss": 1.2025, + "step": 3978 + }, + { + "epoch": 0.27610852820761916, + "grad_norm": 5.056514263153076, + "learning_rate": 0.00019920268565921083, + "loss": 0.8628, + "step": 3979 + }, + { + "epoch": 0.27617791964471583, + "grad_norm": 3.099342107772827, + "learning_rate": 0.0001992019774073561, + "loss": 0.8878, + "step": 3980 + }, + { + "epoch": 0.2762473110818125, + "grad_norm": 4.400445461273193, + "learning_rate": 0.00019920126884233225, + "loss": 0.8621, + "step": 3981 + }, + { + "epoch": 0.27631670251890916, + "grad_norm": 4.193173885345459, + "learning_rate": 0.00019920055996414145, + "loss": 1.2547, + "step": 3982 + }, + { + "epoch": 0.27638609395600583, + "grad_norm": 3.01908540725708, + "learning_rate": 0.00019919985077278604, + "loss": 0.7838, + "step": 3983 + }, + { + "epoch": 0.2764554853931025, + "grad_norm": 4.675381183624268, + "learning_rate": 0.0001991991412682682, + "loss": 1.2639, + "step": 3984 + }, + { + "epoch": 0.27652487683019916, + "grad_norm": 4.101699352264404, + "learning_rate": 0.0001991984314505902, + "loss": 1.3128, + "step": 3985 + }, + { + "epoch": 0.27659426826729583, + "grad_norm": 3.835914373397827, + "learning_rate": 0.00019919772131975422, + "loss": 1.0891, + "step": 3986 + }, + { + "epoch": 0.2766636597043925, + "grad_norm": 4.064837455749512, + "learning_rate": 0.00019919701087576258, + "loss": 1.0285, + "step": 3987 + }, + { + "epoch": 0.27673305114148916, + "grad_norm": 4.346231937408447, + "learning_rate": 0.00019919630011861744, + "loss": 1.0786, + "step": 3988 + }, + { + "epoch": 0.27680244257858583, + "grad_norm": 3.8856494426727295, + "learning_rate": 0.00019919558904832115, + "loss": 0.9806, + "step": 3989 + }, + { + "epoch": 0.27687183401568244, + "grad_norm": 4.032411098480225, + "learning_rate": 0.00019919487766487587, + "loss": 1.0708, + "step": 3990 + }, + { + "epoch": 0.2769412254527791, + "grad_norm": 3.71283221244812, + "learning_rate": 0.00019919416596828385, + "loss": 1.121, + "step": 3991 + }, + { + "epoch": 0.27701061688987577, + "grad_norm": 3.3053746223449707, + "learning_rate": 0.0001991934539585474, + "loss": 1.0037, + "step": 3992 + }, + { + "epoch": 0.27708000832697244, + "grad_norm": 3.0870449542999268, + "learning_rate": 0.00019919274163566865, + "loss": 1.0977, + "step": 3993 + }, + { + "epoch": 0.2771493997640691, + "grad_norm": 3.207928419113159, + "learning_rate": 0.00019919202899964996, + "loss": 1.274, + "step": 3994 + }, + { + "epoch": 0.27721879120116577, + "grad_norm": 3.423774003982544, + "learning_rate": 0.00019919131605049357, + "loss": 1.1272, + "step": 3995 + }, + { + "epoch": 0.27728818263826244, + "grad_norm": 3.5388033390045166, + "learning_rate": 0.0001991906027882017, + "loss": 1.0019, + "step": 3996 + }, + { + "epoch": 0.2773575740753591, + "grad_norm": 3.3840150833129883, + "learning_rate": 0.00019918988921277658, + "loss": 1.1736, + "step": 3997 + }, + { + "epoch": 0.27742696551245577, + "grad_norm": 6.163782119750977, + "learning_rate": 0.0001991891753242205, + "loss": 1.2524, + "step": 3998 + }, + { + "epoch": 0.27749635694955244, + "grad_norm": 3.9491846561431885, + "learning_rate": 0.00019918846112253568, + "loss": 1.1668, + "step": 3999 + }, + { + "epoch": 0.2775657483866491, + "grad_norm": 3.8621089458465576, + "learning_rate": 0.0001991877466077244, + "loss": 1.0807, + "step": 4000 + }, + { + "epoch": 0.2775657483866491, + "eval_loss": 1.0289721488952637, + "eval_runtime": 3769.8802, + "eval_samples_per_second": 6.438, + "eval_steps_per_second": 0.805, + "step": 4000 + }, + { + "epoch": 0.27763513982374577, + "grad_norm": 3.8478000164031982, + "learning_rate": 0.00019918703177978895, + "loss": 0.9663, + "step": 4001 + }, + { + "epoch": 0.27770453126084244, + "grad_norm": 3.882307529449463, + "learning_rate": 0.00019918631663873152, + "loss": 1.1223, + "step": 4002 + }, + { + "epoch": 0.2777739226979391, + "grad_norm": 3.410323143005371, + "learning_rate": 0.0001991856011845544, + "loss": 1.1989, + "step": 4003 + }, + { + "epoch": 0.2778433141350357, + "grad_norm": 2.9006216526031494, + "learning_rate": 0.00019918488541725986, + "loss": 0.902, + "step": 4004 + }, + { + "epoch": 0.2779127055721324, + "grad_norm": 3.316512107849121, + "learning_rate": 0.00019918416933685013, + "loss": 1.0781, + "step": 4005 + }, + { + "epoch": 0.27798209700922905, + "grad_norm": 4.203514099121094, + "learning_rate": 0.00019918345294332748, + "loss": 1.0621, + "step": 4006 + }, + { + "epoch": 0.2780514884463257, + "grad_norm": 5.275970458984375, + "learning_rate": 0.00019918273623669417, + "loss": 1.1233, + "step": 4007 + }, + { + "epoch": 0.2781208798834224, + "grad_norm": 3.0595016479492188, + "learning_rate": 0.0001991820192169525, + "loss": 0.9084, + "step": 4008 + }, + { + "epoch": 0.27819027132051904, + "grad_norm": 5.051469326019287, + "learning_rate": 0.00019918130188410467, + "loss": 1.2949, + "step": 4009 + }, + { + "epoch": 0.2782596627576157, + "grad_norm": 3.697718620300293, + "learning_rate": 0.00019918058423815297, + "loss": 1.1151, + "step": 4010 + }, + { + "epoch": 0.2783290541947124, + "grad_norm": 2.9702086448669434, + "learning_rate": 0.0001991798662790997, + "loss": 0.9041, + "step": 4011 + }, + { + "epoch": 0.27839844563180904, + "grad_norm": 3.5228536128997803, + "learning_rate": 0.0001991791480069471, + "loss": 0.9604, + "step": 4012 + }, + { + "epoch": 0.2784678370689057, + "grad_norm": 3.79758620262146, + "learning_rate": 0.00019917842942169744, + "loss": 1.0308, + "step": 4013 + }, + { + "epoch": 0.2785372285060024, + "grad_norm": 2.705355405807495, + "learning_rate": 0.00019917771052335297, + "loss": 0.7142, + "step": 4014 + }, + { + "epoch": 0.27860661994309904, + "grad_norm": 3.9058139324188232, + "learning_rate": 0.00019917699131191596, + "loss": 1.2091, + "step": 4015 + }, + { + "epoch": 0.2786760113801957, + "grad_norm": 3.782369613647461, + "learning_rate": 0.00019917627178738873, + "loss": 0.9717, + "step": 4016 + }, + { + "epoch": 0.2787454028172923, + "grad_norm": 4.7146196365356445, + "learning_rate": 0.00019917555194977351, + "loss": 0.8949, + "step": 4017 + }, + { + "epoch": 0.278814794254389, + "grad_norm": 4.0465922355651855, + "learning_rate": 0.00019917483179907256, + "loss": 1.2196, + "step": 4018 + }, + { + "epoch": 0.27888418569148565, + "grad_norm": 4.656719207763672, + "learning_rate": 0.00019917411133528817, + "loss": 1.4614, + "step": 4019 + }, + { + "epoch": 0.2789535771285823, + "grad_norm": 3.8514223098754883, + "learning_rate": 0.00019917339055842262, + "loss": 1.4145, + "step": 4020 + }, + { + "epoch": 0.279022968565679, + "grad_norm": 3.7992782592773438, + "learning_rate": 0.00019917266946847822, + "loss": 1.0762, + "step": 4021 + }, + { + "epoch": 0.27909236000277565, + "grad_norm": 3.5928385257720947, + "learning_rate": 0.00019917194806545716, + "loss": 0.7794, + "step": 4022 + }, + { + "epoch": 0.2791617514398723, + "grad_norm": 2.3518943786621094, + "learning_rate": 0.00019917122634936177, + "loss": 0.7593, + "step": 4023 + }, + { + "epoch": 0.279231142876969, + "grad_norm": 3.6021971702575684, + "learning_rate": 0.00019917050432019436, + "loss": 0.8836, + "step": 4024 + }, + { + "epoch": 0.27930053431406565, + "grad_norm": 3.279513120651245, + "learning_rate": 0.00019916978197795718, + "loss": 0.873, + "step": 4025 + }, + { + "epoch": 0.2793699257511623, + "grad_norm": 3.624993085861206, + "learning_rate": 0.00019916905932265247, + "loss": 0.9456, + "step": 4026 + }, + { + "epoch": 0.279439317188259, + "grad_norm": 3.494004011154175, + "learning_rate": 0.00019916833635428254, + "loss": 1.1022, + "step": 4027 + }, + { + "epoch": 0.27950870862535565, + "grad_norm": 3.1957359313964844, + "learning_rate": 0.0001991676130728497, + "loss": 1.2777, + "step": 4028 + }, + { + "epoch": 0.2795781000624523, + "grad_norm": 2.9828097820281982, + "learning_rate": 0.00019916688947835617, + "loss": 0.8283, + "step": 4029 + }, + { + "epoch": 0.27964749149954893, + "grad_norm": 2.9199459552764893, + "learning_rate": 0.00019916616557080432, + "loss": 0.8144, + "step": 4030 + }, + { + "epoch": 0.2797168829366456, + "grad_norm": 6.273212432861328, + "learning_rate": 0.0001991654413501964, + "loss": 1.11, + "step": 4031 + }, + { + "epoch": 0.27978627437374226, + "grad_norm": 2.79099178314209, + "learning_rate": 0.00019916471681653467, + "loss": 0.866, + "step": 4032 + }, + { + "epoch": 0.2798556658108389, + "grad_norm": 3.2993004322052, + "learning_rate": 0.00019916399196982142, + "loss": 0.9829, + "step": 4033 + }, + { + "epoch": 0.2799250572479356, + "grad_norm": 6.081720352172852, + "learning_rate": 0.00019916326681005897, + "loss": 1.2598, + "step": 4034 + }, + { + "epoch": 0.27999444868503226, + "grad_norm": 3.593524217605591, + "learning_rate": 0.00019916254133724956, + "loss": 1.1231, + "step": 4035 + }, + { + "epoch": 0.2800638401221289, + "grad_norm": 3.3057329654693604, + "learning_rate": 0.00019916181555139554, + "loss": 1.0853, + "step": 4036 + }, + { + "epoch": 0.2801332315592256, + "grad_norm": 3.2812750339508057, + "learning_rate": 0.00019916108945249917, + "loss": 1.2046, + "step": 4037 + }, + { + "epoch": 0.28020262299632226, + "grad_norm": 3.0233166217803955, + "learning_rate": 0.00019916036304056274, + "loss": 0.9691, + "step": 4038 + }, + { + "epoch": 0.2802720144334189, + "grad_norm": 3.87990403175354, + "learning_rate": 0.00019915963631558857, + "loss": 0.8969, + "step": 4039 + }, + { + "epoch": 0.2803414058705156, + "grad_norm": 4.390827178955078, + "learning_rate": 0.0001991589092775789, + "loss": 1.1114, + "step": 4040 + }, + { + "epoch": 0.28041079730761226, + "grad_norm": 4.0645751953125, + "learning_rate": 0.00019915818192653607, + "loss": 1.2224, + "step": 4041 + }, + { + "epoch": 0.2804801887447089, + "grad_norm": 4.626298904418945, + "learning_rate": 0.00019915745426246238, + "loss": 1.0969, + "step": 4042 + }, + { + "epoch": 0.2805495801818056, + "grad_norm": 3.542783498764038, + "learning_rate": 0.00019915672628536012, + "loss": 1.1182, + "step": 4043 + }, + { + "epoch": 0.2806189716189022, + "grad_norm": 3.3070995807647705, + "learning_rate": 0.00019915599799523155, + "loss": 0.9101, + "step": 4044 + }, + { + "epoch": 0.28068836305599887, + "grad_norm": 2.7596895694732666, + "learning_rate": 0.00019915526939207897, + "loss": 0.8722, + "step": 4045 + }, + { + "epoch": 0.28075775449309553, + "grad_norm": 3.8372185230255127, + "learning_rate": 0.00019915454047590475, + "loss": 1.2094, + "step": 4046 + }, + { + "epoch": 0.2808271459301922, + "grad_norm": 3.8847148418426514, + "learning_rate": 0.00019915381124671115, + "loss": 1.3046, + "step": 4047 + }, + { + "epoch": 0.28089653736728887, + "grad_norm": 3.344541549682617, + "learning_rate": 0.00019915308170450046, + "loss": 1.0525, + "step": 4048 + }, + { + "epoch": 0.28096592880438553, + "grad_norm": 3.330594062805176, + "learning_rate": 0.000199152351849275, + "loss": 1.0076, + "step": 4049 + }, + { + "epoch": 0.2810353202414822, + "grad_norm": 8.933518409729004, + "learning_rate": 0.00019915162168103709, + "loss": 1.0589, + "step": 4050 + }, + { + "epoch": 0.28110471167857887, + "grad_norm": 3.0201833248138428, + "learning_rate": 0.00019915089119978898, + "loss": 1.0243, + "step": 4051 + }, + { + "epoch": 0.28117410311567553, + "grad_norm": 3.395981550216675, + "learning_rate": 0.00019915016040553303, + "loss": 1.04, + "step": 4052 + }, + { + "epoch": 0.2812434945527722, + "grad_norm": 4.291260242462158, + "learning_rate": 0.00019914942929827155, + "loss": 1.1388, + "step": 4053 + }, + { + "epoch": 0.28131288598986887, + "grad_norm": 2.9085628986358643, + "learning_rate": 0.00019914869787800678, + "loss": 1.1225, + "step": 4054 + }, + { + "epoch": 0.28138227742696553, + "grad_norm": 3.382075786590576, + "learning_rate": 0.00019914796614474108, + "loss": 0.9757, + "step": 4055 + }, + { + "epoch": 0.2814516688640622, + "grad_norm": 3.3573837280273438, + "learning_rate": 0.00019914723409847676, + "loss": 1.1923, + "step": 4056 + }, + { + "epoch": 0.2815210603011588, + "grad_norm": 4.498620510101318, + "learning_rate": 0.00019914650173921613, + "loss": 1.0866, + "step": 4057 + }, + { + "epoch": 0.2815904517382555, + "grad_norm": 3.3931000232696533, + "learning_rate": 0.00019914576906696153, + "loss": 0.9535, + "step": 4058 + }, + { + "epoch": 0.28165984317535214, + "grad_norm": 3.846383810043335, + "learning_rate": 0.0001991450360817152, + "loss": 0.9305, + "step": 4059 + }, + { + "epoch": 0.2817292346124488, + "grad_norm": 2.998507499694824, + "learning_rate": 0.0001991443027834795, + "loss": 1.3277, + "step": 4060 + }, + { + "epoch": 0.2817986260495455, + "grad_norm": 3.3990478515625, + "learning_rate": 0.00019914356917225673, + "loss": 1.4077, + "step": 4061 + }, + { + "epoch": 0.28186801748664214, + "grad_norm": 3.1377100944519043, + "learning_rate": 0.00019914283524804924, + "loss": 1.1302, + "step": 4062 + }, + { + "epoch": 0.2819374089237388, + "grad_norm": 3.302509069442749, + "learning_rate": 0.00019914210101085932, + "loss": 1.023, + "step": 4063 + }, + { + "epoch": 0.2820068003608355, + "grad_norm": 3.7804129123687744, + "learning_rate": 0.00019914136646068925, + "loss": 0.953, + "step": 4064 + }, + { + "epoch": 0.28207619179793214, + "grad_norm": 3.9867260456085205, + "learning_rate": 0.00019914063159754143, + "loss": 1.1509, + "step": 4065 + }, + { + "epoch": 0.2821455832350288, + "grad_norm": 3.6138112545013428, + "learning_rate": 0.00019913989642141814, + "loss": 1.015, + "step": 4066 + }, + { + "epoch": 0.2822149746721255, + "grad_norm": 4.876156806945801, + "learning_rate": 0.00019913916093232165, + "loss": 0.7769, + "step": 4067 + }, + { + "epoch": 0.28228436610922214, + "grad_norm": 4.706211090087891, + "learning_rate": 0.00019913842513025437, + "loss": 1.0663, + "step": 4068 + }, + { + "epoch": 0.2823537575463188, + "grad_norm": 2.747326612472534, + "learning_rate": 0.0001991376890152186, + "loss": 0.9433, + "step": 4069 + }, + { + "epoch": 0.2824231489834155, + "grad_norm": 3.5159451961517334, + "learning_rate": 0.0001991369525872166, + "loss": 1.1758, + "step": 4070 + }, + { + "epoch": 0.2824925404205121, + "grad_norm": 3.856950044631958, + "learning_rate": 0.0001991362158462508, + "loss": 0.9536, + "step": 4071 + }, + { + "epoch": 0.28256193185760875, + "grad_norm": 4.6725077629089355, + "learning_rate": 0.0001991354787923234, + "loss": 1.1931, + "step": 4072 + }, + { + "epoch": 0.2826313232947054, + "grad_norm": 3.435523509979248, + "learning_rate": 0.00019913474142543687, + "loss": 0.7991, + "step": 4073 + }, + { + "epoch": 0.2827007147318021, + "grad_norm": 3.041260242462158, + "learning_rate": 0.00019913400374559342, + "loss": 0.9891, + "step": 4074 + }, + { + "epoch": 0.28277010616889875, + "grad_norm": 3.796086072921753, + "learning_rate": 0.00019913326575279544, + "loss": 1.0264, + "step": 4075 + }, + { + "epoch": 0.2828394976059954, + "grad_norm": 3.220496654510498, + "learning_rate": 0.00019913252744704522, + "loss": 1.0238, + "step": 4076 + }, + { + "epoch": 0.2829088890430921, + "grad_norm": 3.3307793140411377, + "learning_rate": 0.00019913178882834513, + "loss": 1.1379, + "step": 4077 + }, + { + "epoch": 0.28297828048018875, + "grad_norm": 3.478087902069092, + "learning_rate": 0.0001991310498966975, + "loss": 0.8619, + "step": 4078 + }, + { + "epoch": 0.2830476719172854, + "grad_norm": 3.1375136375427246, + "learning_rate": 0.0001991303106521046, + "loss": 0.9937, + "step": 4079 + }, + { + "epoch": 0.2831170633543821, + "grad_norm": 3.62019681930542, + "learning_rate": 0.00019912957109456884, + "loss": 1.0186, + "step": 4080 + }, + { + "epoch": 0.28318645479147875, + "grad_norm": 3.2656612396240234, + "learning_rate": 0.0001991288312240925, + "loss": 1.0175, + "step": 4081 + }, + { + "epoch": 0.2832558462285754, + "grad_norm": 3.6780357360839844, + "learning_rate": 0.00019912809104067795, + "loss": 1.1234, + "step": 4082 + }, + { + "epoch": 0.2833252376656721, + "grad_norm": 3.277637004852295, + "learning_rate": 0.00019912735054432754, + "loss": 0.9845, + "step": 4083 + }, + { + "epoch": 0.2833946291027687, + "grad_norm": 3.8432178497314453, + "learning_rate": 0.0001991266097350436, + "loss": 0.7917, + "step": 4084 + }, + { + "epoch": 0.28346402053986536, + "grad_norm": 4.119819641113281, + "learning_rate": 0.0001991258686128284, + "loss": 1.0873, + "step": 4085 + }, + { + "epoch": 0.283533411976962, + "grad_norm": 3.2660892009735107, + "learning_rate": 0.00019912512717768437, + "loss": 1.1372, + "step": 4086 + }, + { + "epoch": 0.2836028034140587, + "grad_norm": 4.908209323883057, + "learning_rate": 0.0001991243854296138, + "loss": 1.2009, + "step": 4087 + }, + { + "epoch": 0.28367219485115536, + "grad_norm": 2.978726387023926, + "learning_rate": 0.00019912364336861904, + "loss": 0.8746, + "step": 4088 + }, + { + "epoch": 0.283741586288252, + "grad_norm": 4.567093849182129, + "learning_rate": 0.00019912290099470247, + "loss": 1.0983, + "step": 4089 + }, + { + "epoch": 0.2838109777253487, + "grad_norm": 2.6434879302978516, + "learning_rate": 0.00019912215830786636, + "loss": 0.9556, + "step": 4090 + }, + { + "epoch": 0.28388036916244536, + "grad_norm": 3.795057535171509, + "learning_rate": 0.00019912141530811312, + "loss": 1.0956, + "step": 4091 + }, + { + "epoch": 0.283949760599542, + "grad_norm": 3.458695888519287, + "learning_rate": 0.00019912067199544508, + "loss": 0.8271, + "step": 4092 + }, + { + "epoch": 0.2840191520366387, + "grad_norm": 3.7669360637664795, + "learning_rate": 0.00019911992836986456, + "loss": 0.9961, + "step": 4093 + }, + { + "epoch": 0.28408854347373536, + "grad_norm": 3.6432337760925293, + "learning_rate": 0.00019911918443137394, + "loss": 1.1172, + "step": 4094 + }, + { + "epoch": 0.284157934910832, + "grad_norm": 3.329111337661743, + "learning_rate": 0.00019911844017997553, + "loss": 1.2275, + "step": 4095 + }, + { + "epoch": 0.2842273263479287, + "grad_norm": 3.7267005443573, + "learning_rate": 0.00019911769561567174, + "loss": 0.9068, + "step": 4096 + }, + { + "epoch": 0.28429671778502535, + "grad_norm": 4.0666728019714355, + "learning_rate": 0.0001991169507384649, + "loss": 1.0447, + "step": 4097 + }, + { + "epoch": 0.28436610922212197, + "grad_norm": 3.624513864517212, + "learning_rate": 0.00019911620554835727, + "loss": 1.0874, + "step": 4098 + }, + { + "epoch": 0.28443550065921863, + "grad_norm": 3.5978307723999023, + "learning_rate": 0.00019911546004535132, + "loss": 0.9928, + "step": 4099 + }, + { + "epoch": 0.2845048920963153, + "grad_norm": 2.6412956714630127, + "learning_rate": 0.0001991147142294494, + "loss": 0.8799, + "step": 4100 + }, + { + "epoch": 0.28457428353341196, + "grad_norm": 2.9648821353912354, + "learning_rate": 0.00019911396810065376, + "loss": 1.1162, + "step": 4101 + }, + { + "epoch": 0.28464367497050863, + "grad_norm": 3.7003586292266846, + "learning_rate": 0.00019911322165896687, + "loss": 1.0963, + "step": 4102 + }, + { + "epoch": 0.2847130664076053, + "grad_norm": 3.6567392349243164, + "learning_rate": 0.00019911247490439103, + "loss": 0.9586, + "step": 4103 + }, + { + "epoch": 0.28478245784470196, + "grad_norm": 3.0432870388031006, + "learning_rate": 0.00019911172783692858, + "loss": 0.9134, + "step": 4104 + }, + { + "epoch": 0.28485184928179863, + "grad_norm": 4.39106559753418, + "learning_rate": 0.00019911098045658193, + "loss": 0.8374, + "step": 4105 + }, + { + "epoch": 0.2849212407188953, + "grad_norm": 3.280324697494507, + "learning_rate": 0.00019911023276335341, + "loss": 1.0124, + "step": 4106 + }, + { + "epoch": 0.28499063215599196, + "grad_norm": 3.201376438140869, + "learning_rate": 0.00019910948475724538, + "loss": 1.034, + "step": 4107 + }, + { + "epoch": 0.28506002359308863, + "grad_norm": 3.350165843963623, + "learning_rate": 0.0001991087364382602, + "loss": 0.9681, + "step": 4108 + }, + { + "epoch": 0.2851294150301853, + "grad_norm": 3.6660590171813965, + "learning_rate": 0.00019910798780640025, + "loss": 1.1559, + "step": 4109 + }, + { + "epoch": 0.28519880646728196, + "grad_norm": 3.926104784011841, + "learning_rate": 0.00019910723886166787, + "loss": 1.0946, + "step": 4110 + }, + { + "epoch": 0.2852681979043786, + "grad_norm": 3.017836093902588, + "learning_rate": 0.00019910648960406543, + "loss": 0.8461, + "step": 4111 + }, + { + "epoch": 0.28533758934147524, + "grad_norm": 3.1829373836517334, + "learning_rate": 0.00019910574003359532, + "loss": 0.9284, + "step": 4112 + }, + { + "epoch": 0.2854069807785719, + "grad_norm": 3.84302020072937, + "learning_rate": 0.00019910499015025988, + "loss": 0.5742, + "step": 4113 + }, + { + "epoch": 0.2854763722156686, + "grad_norm": 3.1714320182800293, + "learning_rate": 0.00019910423995406147, + "loss": 1.1315, + "step": 4114 + }, + { + "epoch": 0.28554576365276524, + "grad_norm": 3.273087501525879, + "learning_rate": 0.00019910348944500247, + "loss": 0.8425, + "step": 4115 + }, + { + "epoch": 0.2856151550898619, + "grad_norm": 3.9383602142333984, + "learning_rate": 0.0001991027386230853, + "loss": 1.0641, + "step": 4116 + }, + { + "epoch": 0.28568454652695857, + "grad_norm": 3.3911304473876953, + "learning_rate": 0.00019910198748831223, + "loss": 1.1328, + "step": 4117 + }, + { + "epoch": 0.28575393796405524, + "grad_norm": 3.226754903793335, + "learning_rate": 0.00019910123604068573, + "loss": 1.3154, + "step": 4118 + }, + { + "epoch": 0.2858233294011519, + "grad_norm": 3.081587553024292, + "learning_rate": 0.0001991004842802081, + "loss": 0.8935, + "step": 4119 + }, + { + "epoch": 0.28589272083824857, + "grad_norm": 2.991973400115967, + "learning_rate": 0.00019909973220688171, + "loss": 0.9973, + "step": 4120 + }, + { + "epoch": 0.28596211227534524, + "grad_norm": 3.875445604324341, + "learning_rate": 0.000199098979820709, + "loss": 1.0897, + "step": 4121 + }, + { + "epoch": 0.2860315037124419, + "grad_norm": 2.625638961791992, + "learning_rate": 0.0001990982271216923, + "loss": 0.9422, + "step": 4122 + }, + { + "epoch": 0.28610089514953857, + "grad_norm": 4.093222141265869, + "learning_rate": 0.000199097474109834, + "loss": 0.789, + "step": 4123 + }, + { + "epoch": 0.28617028658663524, + "grad_norm": 3.0569565296173096, + "learning_rate": 0.00019909672078513647, + "loss": 1.1538, + "step": 4124 + }, + { + "epoch": 0.28623967802373185, + "grad_norm": 3.808865547180176, + "learning_rate": 0.0001990959671476021, + "loss": 0.9216, + "step": 4125 + }, + { + "epoch": 0.2863090694608285, + "grad_norm": 4.146356582641602, + "learning_rate": 0.00019909521319723321, + "loss": 1.0924, + "step": 4126 + }, + { + "epoch": 0.2863784608979252, + "grad_norm": 3.7992234230041504, + "learning_rate": 0.00019909445893403227, + "loss": 0.7709, + "step": 4127 + }, + { + "epoch": 0.28644785233502185, + "grad_norm": 3.68729305267334, + "learning_rate": 0.0001990937043580016, + "loss": 1.1101, + "step": 4128 + }, + { + "epoch": 0.2865172437721185, + "grad_norm": 3.469336748123169, + "learning_rate": 0.00019909294946914363, + "loss": 1.3067, + "step": 4129 + }, + { + "epoch": 0.2865866352092152, + "grad_norm": 2.288642168045044, + "learning_rate": 0.0001990921942674607, + "loss": 0.6407, + "step": 4130 + }, + { + "epoch": 0.28665602664631185, + "grad_norm": 2.6448004245758057, + "learning_rate": 0.00019909143875295523, + "loss": 0.872, + "step": 4131 + }, + { + "epoch": 0.2867254180834085, + "grad_norm": 3.52129864692688, + "learning_rate": 0.00019909068292562953, + "loss": 0.9373, + "step": 4132 + }, + { + "epoch": 0.2867948095205052, + "grad_norm": 3.3134541511535645, + "learning_rate": 0.0001990899267854861, + "loss": 1.0267, + "step": 4133 + }, + { + "epoch": 0.28686420095760184, + "grad_norm": 3.130727529525757, + "learning_rate": 0.00019908917033252722, + "loss": 1.1278, + "step": 4134 + }, + { + "epoch": 0.2869335923946985, + "grad_norm": 3.3872437477111816, + "learning_rate": 0.00019908841356675535, + "loss": 1.1649, + "step": 4135 + }, + { + "epoch": 0.2870029838317952, + "grad_norm": 3.066406011581421, + "learning_rate": 0.00019908765648817288, + "loss": 0.8831, + "step": 4136 + }, + { + "epoch": 0.28707237526889184, + "grad_norm": 2.6713707447052, + "learning_rate": 0.00019908689909678214, + "loss": 0.9337, + "step": 4137 + }, + { + "epoch": 0.28714176670598845, + "grad_norm": 3.0955493450164795, + "learning_rate": 0.00019908614139258556, + "loss": 0.9016, + "step": 4138 + }, + { + "epoch": 0.2872111581430851, + "grad_norm": 5.673596382141113, + "learning_rate": 0.0001990853833755855, + "loss": 0.8395, + "step": 4139 + }, + { + "epoch": 0.2872805495801818, + "grad_norm": 5.124567031860352, + "learning_rate": 0.00019908462504578442, + "loss": 1.2154, + "step": 4140 + }, + { + "epoch": 0.28734994101727845, + "grad_norm": 3.9909958839416504, + "learning_rate": 0.00019908386640318467, + "loss": 1.3042, + "step": 4141 + }, + { + "epoch": 0.2874193324543751, + "grad_norm": 4.815283298492432, + "learning_rate": 0.00019908310744778867, + "loss": 1.2169, + "step": 4142 + }, + { + "epoch": 0.2874887238914718, + "grad_norm": 2.830714702606201, + "learning_rate": 0.00019908234817959876, + "loss": 0.8821, + "step": 4143 + }, + { + "epoch": 0.28755811532856845, + "grad_norm": 4.608895778656006, + "learning_rate": 0.0001990815885986174, + "loss": 1.0397, + "step": 4144 + }, + { + "epoch": 0.2876275067656651, + "grad_norm": 4.636708736419678, + "learning_rate": 0.00019908082870484696, + "loss": 0.7672, + "step": 4145 + }, + { + "epoch": 0.2876968982027618, + "grad_norm": 3.867544174194336, + "learning_rate": 0.0001990800684982898, + "loss": 0.7506, + "step": 4146 + }, + { + "epoch": 0.28776628963985845, + "grad_norm": 3.073498487472534, + "learning_rate": 0.00019907930797894837, + "loss": 1.153, + "step": 4147 + }, + { + "epoch": 0.2878356810769551, + "grad_norm": 7.982140064239502, + "learning_rate": 0.0001990785471468251, + "loss": 0.9332, + "step": 4148 + }, + { + "epoch": 0.2879050725140518, + "grad_norm": 5.507485389709473, + "learning_rate": 0.0001990777860019223, + "loss": 0.8703, + "step": 4149 + }, + { + "epoch": 0.28797446395114845, + "grad_norm": 3.414015769958496, + "learning_rate": 0.00019907702454424246, + "loss": 0.8093, + "step": 4150 + }, + { + "epoch": 0.2880438553882451, + "grad_norm": 4.750209808349609, + "learning_rate": 0.00019907626277378793, + "loss": 0.7875, + "step": 4151 + }, + { + "epoch": 0.28811324682534173, + "grad_norm": 4.597024917602539, + "learning_rate": 0.00019907550069056117, + "loss": 1.1359, + "step": 4152 + }, + { + "epoch": 0.2881826382624384, + "grad_norm": 3.281597375869751, + "learning_rate": 0.0001990747382945645, + "loss": 0.9561, + "step": 4153 + }, + { + "epoch": 0.28825202969953506, + "grad_norm": 8.650919914245605, + "learning_rate": 0.0001990739755858004, + "loss": 1.0287, + "step": 4154 + }, + { + "epoch": 0.28832142113663173, + "grad_norm": 4.051912307739258, + "learning_rate": 0.00019907321256427125, + "loss": 0.8888, + "step": 4155 + }, + { + "epoch": 0.2883908125737284, + "grad_norm": 7.010167598724365, + "learning_rate": 0.00019907244922997948, + "loss": 1.0044, + "step": 4156 + }, + { + "epoch": 0.28846020401082506, + "grad_norm": 4.239618301391602, + "learning_rate": 0.00019907168558292747, + "loss": 1.1145, + "step": 4157 + }, + { + "epoch": 0.2885295954479217, + "grad_norm": 4.896151065826416, + "learning_rate": 0.00019907092162311762, + "loss": 1.492, + "step": 4158 + }, + { + "epoch": 0.2885989868850184, + "grad_norm": 3.8956117630004883, + "learning_rate": 0.00019907015735055238, + "loss": 1.2114, + "step": 4159 + }, + { + "epoch": 0.28866837832211506, + "grad_norm": 3.5417187213897705, + "learning_rate": 0.00019906939276523414, + "loss": 0.9964, + "step": 4160 + }, + { + "epoch": 0.2887377697592117, + "grad_norm": 3.532172918319702, + "learning_rate": 0.00019906862786716535, + "loss": 1.2381, + "step": 4161 + }, + { + "epoch": 0.2888071611963084, + "grad_norm": 5.372241497039795, + "learning_rate": 0.00019906786265634838, + "loss": 1.2036, + "step": 4162 + }, + { + "epoch": 0.28887655263340506, + "grad_norm": 2.3102381229400635, + "learning_rate": 0.00019906709713278566, + "loss": 0.8549, + "step": 4163 + }, + { + "epoch": 0.2889459440705017, + "grad_norm": 4.164895057678223, + "learning_rate": 0.0001990663312964796, + "loss": 1.2145, + "step": 4164 + }, + { + "epoch": 0.28901533550759834, + "grad_norm": 5.244839191436768, + "learning_rate": 0.00019906556514743266, + "loss": 1.0217, + "step": 4165 + }, + { + "epoch": 0.289084726944695, + "grad_norm": 3.9675207138061523, + "learning_rate": 0.00019906479868564718, + "loss": 1.5508, + "step": 4166 + }, + { + "epoch": 0.28915411838179167, + "grad_norm": 3.6134145259857178, + "learning_rate": 0.00019906403191112567, + "loss": 0.9563, + "step": 4167 + }, + { + "epoch": 0.28922350981888834, + "grad_norm": 4.128198623657227, + "learning_rate": 0.00019906326482387046, + "loss": 1.4481, + "step": 4168 + }, + { + "epoch": 0.289292901255985, + "grad_norm": 3.7842912673950195, + "learning_rate": 0.00019906249742388402, + "loss": 1.0433, + "step": 4169 + }, + { + "epoch": 0.28936229269308167, + "grad_norm": 5.144478797912598, + "learning_rate": 0.00019906172971116882, + "loss": 1.2908, + "step": 4170 + }, + { + "epoch": 0.28943168413017833, + "grad_norm": 3.729017496109009, + "learning_rate": 0.00019906096168572715, + "loss": 1.0758, + "step": 4171 + }, + { + "epoch": 0.289501075567275, + "grad_norm": 3.587515354156494, + "learning_rate": 0.00019906019334756158, + "loss": 0.8905, + "step": 4172 + }, + { + "epoch": 0.28957046700437167, + "grad_norm": 3.593043327331543, + "learning_rate": 0.00019905942469667447, + "loss": 1.1897, + "step": 4173 + }, + { + "epoch": 0.28963985844146833, + "grad_norm": 3.646028995513916, + "learning_rate": 0.00019905865573306822, + "loss": 0.9897, + "step": 4174 + }, + { + "epoch": 0.289709249878565, + "grad_norm": 3.9568819999694824, + "learning_rate": 0.0001990578864567453, + "loss": 1.2166, + "step": 4175 + }, + { + "epoch": 0.28977864131566167, + "grad_norm": 3.5810468196868896, + "learning_rate": 0.00019905711686770813, + "loss": 0.8703, + "step": 4176 + }, + { + "epoch": 0.28984803275275833, + "grad_norm": 3.7745754718780518, + "learning_rate": 0.00019905634696595914, + "loss": 1.1033, + "step": 4177 + }, + { + "epoch": 0.28991742418985494, + "grad_norm": 4.198090553283691, + "learning_rate": 0.00019905557675150074, + "loss": 1.2358, + "step": 4178 + }, + { + "epoch": 0.2899868156269516, + "grad_norm": 3.059020757675171, + "learning_rate": 0.0001990548062243354, + "loss": 1.0231, + "step": 4179 + }, + { + "epoch": 0.2900562070640483, + "grad_norm": 5.2585577964782715, + "learning_rate": 0.0001990540353844655, + "loss": 1.1771, + "step": 4180 + }, + { + "epoch": 0.29012559850114494, + "grad_norm": 4.638047695159912, + "learning_rate": 0.00019905326423189353, + "loss": 1.0598, + "step": 4181 + }, + { + "epoch": 0.2901949899382416, + "grad_norm": 3.958700656890869, + "learning_rate": 0.00019905249276662188, + "loss": 0.9108, + "step": 4182 + }, + { + "epoch": 0.2902643813753383, + "grad_norm": 7.445817470550537, + "learning_rate": 0.000199051720988653, + "loss": 1.2533, + "step": 4183 + }, + { + "epoch": 0.29033377281243494, + "grad_norm": 7.483765602111816, + "learning_rate": 0.00019905094889798933, + "loss": 1.2134, + "step": 4184 + }, + { + "epoch": 0.2904031642495316, + "grad_norm": 4.222705841064453, + "learning_rate": 0.00019905017649463332, + "loss": 0.9884, + "step": 4185 + }, + { + "epoch": 0.2904725556866283, + "grad_norm": 3.6815526485443115, + "learning_rate": 0.00019904940377858736, + "loss": 1.318, + "step": 4186 + }, + { + "epoch": 0.29054194712372494, + "grad_norm": 6.816771507263184, + "learning_rate": 0.000199048630749854, + "loss": 1.3216, + "step": 4187 + }, + { + "epoch": 0.2906113385608216, + "grad_norm": 3.8440229892730713, + "learning_rate": 0.00019904785740843553, + "loss": 0.9139, + "step": 4188 + }, + { + "epoch": 0.2906807299979183, + "grad_norm": 3.7840917110443115, + "learning_rate": 0.00019904708375433446, + "loss": 0.9773, + "step": 4189 + }, + { + "epoch": 0.29075012143501494, + "grad_norm": 5.64506196975708, + "learning_rate": 0.00019904630978755328, + "loss": 1.1799, + "step": 4190 + }, + { + "epoch": 0.2908195128721116, + "grad_norm": 3.530088186264038, + "learning_rate": 0.0001990455355080944, + "loss": 1.1864, + "step": 4191 + }, + { + "epoch": 0.2908889043092082, + "grad_norm": 3.7868051528930664, + "learning_rate": 0.00019904476091596023, + "loss": 1.3111, + "step": 4192 + }, + { + "epoch": 0.2909582957463049, + "grad_norm": 4.5732245445251465, + "learning_rate": 0.00019904398601115324, + "loss": 1.2501, + "step": 4193 + }, + { + "epoch": 0.29102768718340155, + "grad_norm": 3.622586250305176, + "learning_rate": 0.0001990432107936759, + "loss": 1.1892, + "step": 4194 + }, + { + "epoch": 0.2910970786204982, + "grad_norm": 3.9973371028900146, + "learning_rate": 0.0001990424352635306, + "loss": 1.207, + "step": 4195 + }, + { + "epoch": 0.2911664700575949, + "grad_norm": 6.212312698364258, + "learning_rate": 0.00019904165942071983, + "loss": 1.4134, + "step": 4196 + }, + { + "epoch": 0.29123586149469155, + "grad_norm": 2.8159875869750977, + "learning_rate": 0.00019904088326524606, + "loss": 0.8395, + "step": 4197 + }, + { + "epoch": 0.2913052529317882, + "grad_norm": 4.8120551109313965, + "learning_rate": 0.00019904010679711173, + "loss": 0.9242, + "step": 4198 + }, + { + "epoch": 0.2913746443688849, + "grad_norm": 6.074638366699219, + "learning_rate": 0.00019903933001631923, + "loss": 1.0072, + "step": 4199 + }, + { + "epoch": 0.29144403580598155, + "grad_norm": 4.464717864990234, + "learning_rate": 0.00019903855292287105, + "loss": 1.2443, + "step": 4200 + }, + { + "epoch": 0.2915134272430782, + "grad_norm": 2.8271303176879883, + "learning_rate": 0.00019903777551676964, + "loss": 1.1548, + "step": 4201 + }, + { + "epoch": 0.2915828186801749, + "grad_norm": 2.759239435195923, + "learning_rate": 0.0001990369977980175, + "loss": 0.9991, + "step": 4202 + }, + { + "epoch": 0.29165221011727155, + "grad_norm": 7.696215629577637, + "learning_rate": 0.00019903621976661705, + "loss": 0.9554, + "step": 4203 + }, + { + "epoch": 0.2917216015543682, + "grad_norm": 4.792369365692139, + "learning_rate": 0.00019903544142257072, + "loss": 1.2611, + "step": 4204 + }, + { + "epoch": 0.2917909929914648, + "grad_norm": 5.548574924468994, + "learning_rate": 0.000199034662765881, + "loss": 1.3559, + "step": 4205 + }, + { + "epoch": 0.2918603844285615, + "grad_norm": 4.113785266876221, + "learning_rate": 0.00019903388379655034, + "loss": 1.0986, + "step": 4206 + }, + { + "epoch": 0.29192977586565816, + "grad_norm": 3.4641337394714355, + "learning_rate": 0.00019903310451458122, + "loss": 1.0504, + "step": 4207 + }, + { + "epoch": 0.2919991673027548, + "grad_norm": 3.3622214794158936, + "learning_rate": 0.00019903232491997604, + "loss": 1.0989, + "step": 4208 + }, + { + "epoch": 0.2920685587398515, + "grad_norm": 4.442663192749023, + "learning_rate": 0.00019903154501273732, + "loss": 1.3214, + "step": 4209 + }, + { + "epoch": 0.29213795017694816, + "grad_norm": 3.1673378944396973, + "learning_rate": 0.00019903076479286754, + "loss": 0.9808, + "step": 4210 + }, + { + "epoch": 0.2922073416140448, + "grad_norm": 5.4794769287109375, + "learning_rate": 0.00019902998426036907, + "loss": 1.082, + "step": 4211 + }, + { + "epoch": 0.2922767330511415, + "grad_norm": 3.6721320152282715, + "learning_rate": 0.00019902920341524443, + "loss": 1.0173, + "step": 4212 + }, + { + "epoch": 0.29234612448823816, + "grad_norm": 3.884187698364258, + "learning_rate": 0.0001990284222574961, + "loss": 0.8599, + "step": 4213 + }, + { + "epoch": 0.2924155159253348, + "grad_norm": 2.9846436977386475, + "learning_rate": 0.00019902764078712654, + "loss": 0.9692, + "step": 4214 + }, + { + "epoch": 0.2924849073624315, + "grad_norm": 6.093502998352051, + "learning_rate": 0.0001990268590041382, + "loss": 1.0199, + "step": 4215 + }, + { + "epoch": 0.29255429879952816, + "grad_norm": 3.748849868774414, + "learning_rate": 0.00019902607690853357, + "loss": 1.1946, + "step": 4216 + }, + { + "epoch": 0.2926236902366248, + "grad_norm": 4.766956806182861, + "learning_rate": 0.00019902529450031508, + "loss": 1.097, + "step": 4217 + }, + { + "epoch": 0.2926930816737215, + "grad_norm": 3.0226056575775146, + "learning_rate": 0.00019902451177948526, + "loss": 0.9622, + "step": 4218 + }, + { + "epoch": 0.2927624731108181, + "grad_norm": 4.212645053863525, + "learning_rate": 0.0001990237287460465, + "loss": 1.1318, + "step": 4219 + }, + { + "epoch": 0.29283186454791477, + "grad_norm": 4.675900459289551, + "learning_rate": 0.00019902294540000132, + "loss": 1.3674, + "step": 4220 + }, + { + "epoch": 0.29290125598501143, + "grad_norm": 4.926464080810547, + "learning_rate": 0.00019902216174135222, + "loss": 1.1574, + "step": 4221 + }, + { + "epoch": 0.2929706474221081, + "grad_norm": 3.445854902267456, + "learning_rate": 0.00019902137777010162, + "loss": 0.818, + "step": 4222 + }, + { + "epoch": 0.29304003885920477, + "grad_norm": 3.56571888923645, + "learning_rate": 0.00019902059348625203, + "loss": 0.847, + "step": 4223 + }, + { + "epoch": 0.29310943029630143, + "grad_norm": 4.246292591094971, + "learning_rate": 0.00019901980888980593, + "loss": 1.1403, + "step": 4224 + }, + { + "epoch": 0.2931788217333981, + "grad_norm": 5.053463459014893, + "learning_rate": 0.00019901902398076576, + "loss": 1.1915, + "step": 4225 + }, + { + "epoch": 0.29324821317049476, + "grad_norm": 3.434906482696533, + "learning_rate": 0.00019901823875913402, + "loss": 0.9127, + "step": 4226 + }, + { + "epoch": 0.29331760460759143, + "grad_norm": 5.149567127227783, + "learning_rate": 0.00019901745322491317, + "loss": 1.1723, + "step": 4227 + }, + { + "epoch": 0.2933869960446881, + "grad_norm": 5.732105731964111, + "learning_rate": 0.00019901666737810576, + "loss": 1.1869, + "step": 4228 + }, + { + "epoch": 0.29345638748178476, + "grad_norm": 4.774272918701172, + "learning_rate": 0.00019901588121871416, + "loss": 1.3066, + "step": 4229 + }, + { + "epoch": 0.29352577891888143, + "grad_norm": 3.7609100341796875, + "learning_rate": 0.00019901509474674095, + "loss": 1.1537, + "step": 4230 + }, + { + "epoch": 0.2935951703559781, + "grad_norm": 3.3102304935455322, + "learning_rate": 0.0001990143079621886, + "loss": 1.0656, + "step": 4231 + }, + { + "epoch": 0.2936645617930747, + "grad_norm": 4.074265003204346, + "learning_rate": 0.0001990135208650595, + "loss": 1.2883, + "step": 4232 + }, + { + "epoch": 0.2937339532301714, + "grad_norm": 3.6229708194732666, + "learning_rate": 0.00019901273345535625, + "loss": 1.0389, + "step": 4233 + }, + { + "epoch": 0.29380334466726804, + "grad_norm": 3.1996490955352783, + "learning_rate": 0.00019901194573308128, + "loss": 1.1023, + "step": 4234 + }, + { + "epoch": 0.2938727361043647, + "grad_norm": 3.316676378250122, + "learning_rate": 0.00019901115769823706, + "loss": 0.9311, + "step": 4235 + }, + { + "epoch": 0.2939421275414614, + "grad_norm": 3.4313666820526123, + "learning_rate": 0.00019901036935082614, + "loss": 0.9282, + "step": 4236 + }, + { + "epoch": 0.29401151897855804, + "grad_norm": 4.057358741760254, + "learning_rate": 0.00019900958069085097, + "loss": 0.9398, + "step": 4237 + }, + { + "epoch": 0.2940809104156547, + "grad_norm": 14.422501564025879, + "learning_rate": 0.000199008791718314, + "loss": 1.2411, + "step": 4238 + }, + { + "epoch": 0.2941503018527514, + "grad_norm": 2.7167210578918457, + "learning_rate": 0.0001990080024332178, + "loss": 1.0493, + "step": 4239 + }, + { + "epoch": 0.29421969328984804, + "grad_norm": 14.151202201843262, + "learning_rate": 0.0001990072128355648, + "loss": 1.5336, + "step": 4240 + }, + { + "epoch": 0.2942890847269447, + "grad_norm": 4.298392295837402, + "learning_rate": 0.00019900642292535752, + "loss": 1.0252, + "step": 4241 + }, + { + "epoch": 0.29435847616404137, + "grad_norm": 3.321985960006714, + "learning_rate": 0.00019900563270259847, + "loss": 1.218, + "step": 4242 + }, + { + "epoch": 0.29442786760113804, + "grad_norm": 2.9848268032073975, + "learning_rate": 0.00019900484216729008, + "loss": 0.709, + "step": 4243 + }, + { + "epoch": 0.2944972590382347, + "grad_norm": 3.567831516265869, + "learning_rate": 0.00019900405131943494, + "loss": 1.1645, + "step": 4244 + }, + { + "epoch": 0.29456665047533137, + "grad_norm": 3.464597463607788, + "learning_rate": 0.0001990032601590355, + "loss": 1.0314, + "step": 4245 + }, + { + "epoch": 0.294636041912428, + "grad_norm": 3.8516342639923096, + "learning_rate": 0.0001990024686860942, + "loss": 1.1325, + "step": 4246 + }, + { + "epoch": 0.29470543334952465, + "grad_norm": 3.7099595069885254, + "learning_rate": 0.00019900167690061363, + "loss": 1.1087, + "step": 4247 + }, + { + "epoch": 0.2947748247866213, + "grad_norm": 4.405357360839844, + "learning_rate": 0.00019900088480259625, + "loss": 1.3626, + "step": 4248 + }, + { + "epoch": 0.294844216223718, + "grad_norm": 3.2206313610076904, + "learning_rate": 0.00019900009239204455, + "loss": 0.98, + "step": 4249 + }, + { + "epoch": 0.29491360766081465, + "grad_norm": 3.9787566661834717, + "learning_rate": 0.00019899929966896106, + "loss": 1.05, + "step": 4250 + }, + { + "epoch": 0.2949829990979113, + "grad_norm": 3.160195827484131, + "learning_rate": 0.00019899850663334824, + "loss": 0.9744, + "step": 4251 + }, + { + "epoch": 0.295052390535008, + "grad_norm": 4.2062087059021, + "learning_rate": 0.00019899771328520864, + "loss": 1.1889, + "step": 4252 + }, + { + "epoch": 0.29512178197210465, + "grad_norm": 3.9422407150268555, + "learning_rate": 0.00019899691962454473, + "loss": 1.3692, + "step": 4253 + }, + { + "epoch": 0.2951911734092013, + "grad_norm": 3.765930652618408, + "learning_rate": 0.00019899612565135906, + "loss": 1.1196, + "step": 4254 + }, + { + "epoch": 0.295260564846298, + "grad_norm": 3.5609097480773926, + "learning_rate": 0.0001989953313656541, + "loss": 1.0575, + "step": 4255 + }, + { + "epoch": 0.29532995628339465, + "grad_norm": 5.401237487792969, + "learning_rate": 0.00019899453676743237, + "loss": 1.3537, + "step": 4256 + }, + { + "epoch": 0.2953993477204913, + "grad_norm": 3.6982057094573975, + "learning_rate": 0.00019899374185669634, + "loss": 1.1194, + "step": 4257 + }, + { + "epoch": 0.295468739157588, + "grad_norm": 3.524474620819092, + "learning_rate": 0.0001989929466334486, + "loss": 1.1499, + "step": 4258 + }, + { + "epoch": 0.2955381305946846, + "grad_norm": 3.580286741256714, + "learning_rate": 0.00019899215109769156, + "loss": 1.026, + "step": 4259 + }, + { + "epoch": 0.29560752203178126, + "grad_norm": 2.7313859462738037, + "learning_rate": 0.0001989913552494278, + "loss": 0.8349, + "step": 4260 + }, + { + "epoch": 0.2956769134688779, + "grad_norm": 6.128452301025391, + "learning_rate": 0.00019899055908865983, + "loss": 1.0753, + "step": 4261 + }, + { + "epoch": 0.2957463049059746, + "grad_norm": 4.53357458114624, + "learning_rate": 0.00019898976261539014, + "loss": 1.2473, + "step": 4262 + }, + { + "epoch": 0.29581569634307125, + "grad_norm": 3.613823413848877, + "learning_rate": 0.00019898896582962126, + "loss": 0.929, + "step": 4263 + }, + { + "epoch": 0.2958850877801679, + "grad_norm": 3.378849744796753, + "learning_rate": 0.0001989881687313557, + "loss": 1.1322, + "step": 4264 + }, + { + "epoch": 0.2959544792172646, + "grad_norm": 3.22357439994812, + "learning_rate": 0.00019898737132059599, + "loss": 0.8971, + "step": 4265 + }, + { + "epoch": 0.29602387065436125, + "grad_norm": 2.780160903930664, + "learning_rate": 0.0001989865735973446, + "loss": 0.9351, + "step": 4266 + }, + { + "epoch": 0.2960932620914579, + "grad_norm": 3.5161941051483154, + "learning_rate": 0.0001989857755616041, + "loss": 0.9993, + "step": 4267 + }, + { + "epoch": 0.2961626535285546, + "grad_norm": 3.6340036392211914, + "learning_rate": 0.00019898497721337698, + "loss": 1.1161, + "step": 4268 + }, + { + "epoch": 0.29623204496565125, + "grad_norm": 3.611860513687134, + "learning_rate": 0.0001989841785526658, + "loss": 1.0671, + "step": 4269 + }, + { + "epoch": 0.2963014364027479, + "grad_norm": 3.373969793319702, + "learning_rate": 0.00019898337957947304, + "loss": 1.0369, + "step": 4270 + }, + { + "epoch": 0.2963708278398446, + "grad_norm": 3.4801788330078125, + "learning_rate": 0.00019898258029380124, + "loss": 0.8478, + "step": 4271 + }, + { + "epoch": 0.29644021927694125, + "grad_norm": 3.3907012939453125, + "learning_rate": 0.00019898178069565288, + "loss": 1.2762, + "step": 4272 + }, + { + "epoch": 0.29650961071403786, + "grad_norm": 4.176407814025879, + "learning_rate": 0.00019898098078503055, + "loss": 1.2535, + "step": 4273 + }, + { + "epoch": 0.29657900215113453, + "grad_norm": 3.2877180576324463, + "learning_rate": 0.00019898018056193676, + "loss": 1.0295, + "step": 4274 + }, + { + "epoch": 0.2966483935882312, + "grad_norm": 3.674363374710083, + "learning_rate": 0.000198979380026374, + "loss": 0.9687, + "step": 4275 + }, + { + "epoch": 0.29671778502532786, + "grad_norm": 3.3921070098876953, + "learning_rate": 0.00019897857917834487, + "loss": 1.0265, + "step": 4276 + }, + { + "epoch": 0.29678717646242453, + "grad_norm": 3.5535969734191895, + "learning_rate": 0.0001989777780178518, + "loss": 1.1986, + "step": 4277 + }, + { + "epoch": 0.2968565678995212, + "grad_norm": 3.148392915725708, + "learning_rate": 0.00019897697654489737, + "loss": 0.818, + "step": 4278 + }, + { + "epoch": 0.29692595933661786, + "grad_norm": 3.9076998233795166, + "learning_rate": 0.00019897617475948413, + "loss": 1.1235, + "step": 4279 + }, + { + "epoch": 0.29699535077371453, + "grad_norm": 3.0635223388671875, + "learning_rate": 0.0001989753726616146, + "loss": 1.0697, + "step": 4280 + }, + { + "epoch": 0.2970647422108112, + "grad_norm": 3.3057475090026855, + "learning_rate": 0.00019897457025129124, + "loss": 0.8416, + "step": 4281 + }, + { + "epoch": 0.29713413364790786, + "grad_norm": 4.315209865570068, + "learning_rate": 0.00019897376752851668, + "loss": 1.1793, + "step": 4282 + }, + { + "epoch": 0.2972035250850045, + "grad_norm": 3.2386398315429688, + "learning_rate": 0.00019897296449329342, + "loss": 1.1633, + "step": 4283 + }, + { + "epoch": 0.2972729165221012, + "grad_norm": 3.9466164112091064, + "learning_rate": 0.00019897216114562403, + "loss": 1.0976, + "step": 4284 + }, + { + "epoch": 0.29734230795919786, + "grad_norm": 3.306220054626465, + "learning_rate": 0.00019897135748551096, + "loss": 0.8114, + "step": 4285 + }, + { + "epoch": 0.29741169939629447, + "grad_norm": 3.0170340538024902, + "learning_rate": 0.00019897055351295682, + "loss": 1.2356, + "step": 4286 + }, + { + "epoch": 0.29748109083339114, + "grad_norm": 3.807319164276123, + "learning_rate": 0.00019896974922796413, + "loss": 1.0556, + "step": 4287 + }, + { + "epoch": 0.2975504822704878, + "grad_norm": 5.633152961730957, + "learning_rate": 0.0001989689446305354, + "loss": 1.1577, + "step": 4288 + }, + { + "epoch": 0.29761987370758447, + "grad_norm": 3.510206937789917, + "learning_rate": 0.00019896813972067323, + "loss": 1.0602, + "step": 4289 + }, + { + "epoch": 0.29768926514468114, + "grad_norm": 3.8492507934570312, + "learning_rate": 0.0001989673344983801, + "loss": 1.2771, + "step": 4290 + }, + { + "epoch": 0.2977586565817778, + "grad_norm": 2.846080780029297, + "learning_rate": 0.0001989665289636586, + "loss": 0.9146, + "step": 4291 + }, + { + "epoch": 0.29782804801887447, + "grad_norm": 3.1180615425109863, + "learning_rate": 0.00019896572311651123, + "loss": 0.9331, + "step": 4292 + }, + { + "epoch": 0.29789743945597114, + "grad_norm": 3.577604055404663, + "learning_rate": 0.00019896491695694054, + "loss": 0.9442, + "step": 4293 + }, + { + "epoch": 0.2979668308930678, + "grad_norm": 3.284097671508789, + "learning_rate": 0.0001989641104849491, + "loss": 0.9906, + "step": 4294 + }, + { + "epoch": 0.29803622233016447, + "grad_norm": 3.2460851669311523, + "learning_rate": 0.00019896330370053946, + "loss": 0.935, + "step": 4295 + }, + { + "epoch": 0.29810561376726114, + "grad_norm": 2.83512282371521, + "learning_rate": 0.00019896249660371416, + "loss": 0.883, + "step": 4296 + }, + { + "epoch": 0.2981750052043578, + "grad_norm": 3.3490889072418213, + "learning_rate": 0.0001989616891944757, + "loss": 0.9934, + "step": 4297 + }, + { + "epoch": 0.29824439664145447, + "grad_norm": 3.811851978302002, + "learning_rate": 0.0001989608814728267, + "loss": 1.2458, + "step": 4298 + }, + { + "epoch": 0.29831378807855113, + "grad_norm": 3.5197253227233887, + "learning_rate": 0.00019896007343876967, + "loss": 0.9094, + "step": 4299 + }, + { + "epoch": 0.29838317951564775, + "grad_norm": 3.737135171890259, + "learning_rate": 0.00019895926509230717, + "loss": 1.0334, + "step": 4300 + }, + { + "epoch": 0.2984525709527444, + "grad_norm": 3.8079185485839844, + "learning_rate": 0.00019895845643344174, + "loss": 1.0672, + "step": 4301 + }, + { + "epoch": 0.2985219623898411, + "grad_norm": 2.9282548427581787, + "learning_rate": 0.00019895764746217598, + "loss": 0.9277, + "step": 4302 + }, + { + "epoch": 0.29859135382693774, + "grad_norm": 2.477689266204834, + "learning_rate": 0.00019895683817851238, + "loss": 0.8792, + "step": 4303 + }, + { + "epoch": 0.2986607452640344, + "grad_norm": 2.708270311355591, + "learning_rate": 0.0001989560285824535, + "loss": 1.0671, + "step": 4304 + }, + { + "epoch": 0.2987301367011311, + "grad_norm": 2.7696940898895264, + "learning_rate": 0.00019895521867400198, + "loss": 1.0575, + "step": 4305 + }, + { + "epoch": 0.29879952813822774, + "grad_norm": 3.593106508255005, + "learning_rate": 0.00019895440845316026, + "loss": 1.1717, + "step": 4306 + }, + { + "epoch": 0.2988689195753244, + "grad_norm": 4.009599685668945, + "learning_rate": 0.000198953597919931, + "loss": 0.8953, + "step": 4307 + }, + { + "epoch": 0.2989383110124211, + "grad_norm": 3.157066583633423, + "learning_rate": 0.00019895278707431668, + "loss": 0.8663, + "step": 4308 + }, + { + "epoch": 0.29900770244951774, + "grad_norm": 2.9138004779815674, + "learning_rate": 0.0001989519759163199, + "loss": 1.0895, + "step": 4309 + }, + { + "epoch": 0.2990770938866144, + "grad_norm": 3.9623911380767822, + "learning_rate": 0.0001989511644459432, + "loss": 1.1034, + "step": 4310 + }, + { + "epoch": 0.2991464853237111, + "grad_norm": 3.163574457168579, + "learning_rate": 0.0001989503526631892, + "loss": 1.2182, + "step": 4311 + }, + { + "epoch": 0.29921587676080774, + "grad_norm": 3.3317954540252686, + "learning_rate": 0.00019894954056806036, + "loss": 0.9597, + "step": 4312 + }, + { + "epoch": 0.29928526819790435, + "grad_norm": 3.1026129722595215, + "learning_rate": 0.0001989487281605593, + "loss": 1.0702, + "step": 4313 + }, + { + "epoch": 0.299354659635001, + "grad_norm": 2.925673246383667, + "learning_rate": 0.00019894791544068863, + "loss": 0.802, + "step": 4314 + }, + { + "epoch": 0.2994240510720977, + "grad_norm": 3.4216651916503906, + "learning_rate": 0.00019894710240845088, + "loss": 1.2484, + "step": 4315 + }, + { + "epoch": 0.29949344250919435, + "grad_norm": 3.4916157722473145, + "learning_rate": 0.00019894628906384854, + "loss": 1.0931, + "step": 4316 + }, + { + "epoch": 0.299562833946291, + "grad_norm": 3.180067300796509, + "learning_rate": 0.00019894547540688428, + "loss": 1.0465, + "step": 4317 + }, + { + "epoch": 0.2996322253833877, + "grad_norm": 3.936983823776245, + "learning_rate": 0.00019894466143756065, + "loss": 1.0383, + "step": 4318 + }, + { + "epoch": 0.29970161682048435, + "grad_norm": 2.8539981842041016, + "learning_rate": 0.0001989438471558802, + "loss": 0.8462, + "step": 4319 + }, + { + "epoch": 0.299771008257581, + "grad_norm": 3.195103406906128, + "learning_rate": 0.0001989430325618455, + "loss": 1.0626, + "step": 4320 + }, + { + "epoch": 0.2998403996946777, + "grad_norm": 3.3901782035827637, + "learning_rate": 0.00019894221765545913, + "loss": 1.1677, + "step": 4321 + }, + { + "epoch": 0.29990979113177435, + "grad_norm": 4.293059349060059, + "learning_rate": 0.00019894140243672363, + "loss": 1.1076, + "step": 4322 + }, + { + "epoch": 0.299979182568871, + "grad_norm": 2.8418431282043457, + "learning_rate": 0.00019894058690564164, + "loss": 0.8736, + "step": 4323 + }, + { + "epoch": 0.3000485740059677, + "grad_norm": 3.9265549182891846, + "learning_rate": 0.00019893977106221565, + "loss": 1.0328, + "step": 4324 + }, + { + "epoch": 0.30011796544306435, + "grad_norm": 2.865351915359497, + "learning_rate": 0.0001989389549064483, + "loss": 0.7826, + "step": 4325 + }, + { + "epoch": 0.30018735688016096, + "grad_norm": 3.0942533016204834, + "learning_rate": 0.00019893813843834218, + "loss": 0.9244, + "step": 4326 + }, + { + "epoch": 0.3002567483172576, + "grad_norm": 3.8951079845428467, + "learning_rate": 0.0001989373216578998, + "loss": 1.0159, + "step": 4327 + }, + { + "epoch": 0.3003261397543543, + "grad_norm": 3.3720691204071045, + "learning_rate": 0.0001989365045651238, + "loss": 1.1592, + "step": 4328 + }, + { + "epoch": 0.30039553119145096, + "grad_norm": 3.057452917098999, + "learning_rate": 0.0001989356871600167, + "loss": 1.0039, + "step": 4329 + }, + { + "epoch": 0.3004649226285476, + "grad_norm": 3.8318045139312744, + "learning_rate": 0.00019893486944258117, + "loss": 1.0363, + "step": 4330 + }, + { + "epoch": 0.3005343140656443, + "grad_norm": 3.281409502029419, + "learning_rate": 0.0001989340514128197, + "loss": 0.911, + "step": 4331 + }, + { + "epoch": 0.30060370550274096, + "grad_norm": 5.250543594360352, + "learning_rate": 0.0001989332330707349, + "loss": 1.0516, + "step": 4332 + }, + { + "epoch": 0.3006730969398376, + "grad_norm": 3.113058567047119, + "learning_rate": 0.00019893241441632938, + "loss": 1.0257, + "step": 4333 + }, + { + "epoch": 0.3007424883769343, + "grad_norm": 2.9528634548187256, + "learning_rate": 0.0001989315954496057, + "loss": 0.9497, + "step": 4334 + }, + { + "epoch": 0.30081187981403096, + "grad_norm": 4.493294715881348, + "learning_rate": 0.00019893077617056643, + "loss": 0.8132, + "step": 4335 + }, + { + "epoch": 0.3008812712511276, + "grad_norm": 4.47160005569458, + "learning_rate": 0.0001989299565792142, + "loss": 0.9315, + "step": 4336 + }, + { + "epoch": 0.3009506626882243, + "grad_norm": 3.612006187438965, + "learning_rate": 0.00019892913667555158, + "loss": 0.9109, + "step": 4337 + }, + { + "epoch": 0.30102005412532096, + "grad_norm": 4.2310051918029785, + "learning_rate": 0.00019892831645958113, + "loss": 0.902, + "step": 4338 + }, + { + "epoch": 0.3010894455624176, + "grad_norm": 3.632319211959839, + "learning_rate": 0.0001989274959313055, + "loss": 0.9554, + "step": 4339 + }, + { + "epoch": 0.30115883699951423, + "grad_norm": 4.588902473449707, + "learning_rate": 0.00019892667509072723, + "loss": 1.2188, + "step": 4340 + }, + { + "epoch": 0.3012282284366109, + "grad_norm": 3.2486536502838135, + "learning_rate": 0.0001989258539378489, + "loss": 1.3121, + "step": 4341 + }, + { + "epoch": 0.30129761987370757, + "grad_norm": 2.880113363265991, + "learning_rate": 0.00019892503247267312, + "loss": 1.0852, + "step": 4342 + }, + { + "epoch": 0.30136701131080423, + "grad_norm": 3.2294392585754395, + "learning_rate": 0.00019892421069520253, + "loss": 0.9842, + "step": 4343 + }, + { + "epoch": 0.3014364027479009, + "grad_norm": 3.8159244060516357, + "learning_rate": 0.00019892338860543968, + "loss": 0.9889, + "step": 4344 + }, + { + "epoch": 0.30150579418499757, + "grad_norm": 4.153054237365723, + "learning_rate": 0.00019892256620338713, + "loss": 1.0796, + "step": 4345 + }, + { + "epoch": 0.30157518562209423, + "grad_norm": 3.4236929416656494, + "learning_rate": 0.00019892174348904755, + "loss": 0.9869, + "step": 4346 + }, + { + "epoch": 0.3016445770591909, + "grad_norm": 4.564944744110107, + "learning_rate": 0.0001989209204624235, + "loss": 1.2586, + "step": 4347 + }, + { + "epoch": 0.30171396849628757, + "grad_norm": 3.1387062072753906, + "learning_rate": 0.0001989200971235176, + "loss": 1.0044, + "step": 4348 + }, + { + "epoch": 0.30178335993338423, + "grad_norm": 2.9920928478240967, + "learning_rate": 0.00019891927347233236, + "loss": 0.8825, + "step": 4349 + }, + { + "epoch": 0.3018527513704809, + "grad_norm": 4.350956916809082, + "learning_rate": 0.0001989184495088705, + "loss": 1.1246, + "step": 4350 + }, + { + "epoch": 0.30192214280757756, + "grad_norm": 2.7855234146118164, + "learning_rate": 0.0001989176252331346, + "loss": 0.9879, + "step": 4351 + }, + { + "epoch": 0.30199153424467423, + "grad_norm": 2.792567491531372, + "learning_rate": 0.0001989168006451272, + "loss": 1.0579, + "step": 4352 + }, + { + "epoch": 0.30206092568177084, + "grad_norm": 2.9476206302642822, + "learning_rate": 0.0001989159757448509, + "loss": 0.9123, + "step": 4353 + }, + { + "epoch": 0.3021303171188675, + "grad_norm": 3.56624698638916, + "learning_rate": 0.0001989151505323084, + "loss": 1.096, + "step": 4354 + }, + { + "epoch": 0.3021997085559642, + "grad_norm": 2.966238498687744, + "learning_rate": 0.00019891432500750223, + "loss": 0.9163, + "step": 4355 + }, + { + "epoch": 0.30226909999306084, + "grad_norm": 3.980931282043457, + "learning_rate": 0.000198913499170435, + "loss": 1.1468, + "step": 4356 + }, + { + "epoch": 0.3023384914301575, + "grad_norm": 3.436593532562256, + "learning_rate": 0.0001989126730211093, + "loss": 1.0041, + "step": 4357 + }, + { + "epoch": 0.3024078828672542, + "grad_norm": 3.2371532917022705, + "learning_rate": 0.00019891184655952785, + "loss": 0.839, + "step": 4358 + }, + { + "epoch": 0.30247727430435084, + "grad_norm": 3.5475025177001953, + "learning_rate": 0.0001989110197856931, + "loss": 0.9774, + "step": 4359 + }, + { + "epoch": 0.3025466657414475, + "grad_norm": 4.01140022277832, + "learning_rate": 0.0001989101926996078, + "loss": 1.0074, + "step": 4360 + }, + { + "epoch": 0.3026160571785442, + "grad_norm": 4.152772426605225, + "learning_rate": 0.00019890936530127443, + "loss": 1.1707, + "step": 4361 + }, + { + "epoch": 0.30268544861564084, + "grad_norm": 2.9889137744903564, + "learning_rate": 0.00019890853759069573, + "loss": 0.9116, + "step": 4362 + }, + { + "epoch": 0.3027548400527375, + "grad_norm": 3.2169153690338135, + "learning_rate": 0.00019890770956787422, + "loss": 1.1201, + "step": 4363 + }, + { + "epoch": 0.3028242314898342, + "grad_norm": 4.733648777008057, + "learning_rate": 0.00019890688123281255, + "loss": 1.1459, + "step": 4364 + }, + { + "epoch": 0.30289362292693084, + "grad_norm": 2.943448066711426, + "learning_rate": 0.0001989060525855133, + "loss": 0.9974, + "step": 4365 + }, + { + "epoch": 0.3029630143640275, + "grad_norm": 3.5838088989257812, + "learning_rate": 0.00019890522362597918, + "loss": 1.1251, + "step": 4366 + }, + { + "epoch": 0.3030324058011241, + "grad_norm": 3.9644601345062256, + "learning_rate": 0.00019890439435421274, + "loss": 1.1155, + "step": 4367 + }, + { + "epoch": 0.3031017972382208, + "grad_norm": 2.8174424171447754, + "learning_rate": 0.00019890356477021657, + "loss": 0.9558, + "step": 4368 + }, + { + "epoch": 0.30317118867531745, + "grad_norm": 3.3826470375061035, + "learning_rate": 0.00019890273487399332, + "loss": 1.032, + "step": 4369 + }, + { + "epoch": 0.3032405801124141, + "grad_norm": 3.4515342712402344, + "learning_rate": 0.0001989019046655456, + "loss": 0.9593, + "step": 4370 + }, + { + "epoch": 0.3033099715495108, + "grad_norm": 2.887951374053955, + "learning_rate": 0.00019890107414487606, + "loss": 0.8055, + "step": 4371 + }, + { + "epoch": 0.30337936298660745, + "grad_norm": 4.115792751312256, + "learning_rate": 0.00019890024331198732, + "loss": 1.1207, + "step": 4372 + }, + { + "epoch": 0.3034487544237041, + "grad_norm": 3.8460826873779297, + "learning_rate": 0.00019889941216688194, + "loss": 0.9385, + "step": 4373 + }, + { + "epoch": 0.3035181458608008, + "grad_norm": 2.692486047744751, + "learning_rate": 0.00019889858070956264, + "loss": 0.8371, + "step": 4374 + }, + { + "epoch": 0.30358753729789745, + "grad_norm": 3.8513739109039307, + "learning_rate": 0.00019889774894003197, + "loss": 1.0549, + "step": 4375 + }, + { + "epoch": 0.3036569287349941, + "grad_norm": 2.5390918254852295, + "learning_rate": 0.0001988969168582926, + "loss": 0.8821, + "step": 4376 + }, + { + "epoch": 0.3037263201720908, + "grad_norm": 4.039330005645752, + "learning_rate": 0.0001988960844643471, + "loss": 1.1435, + "step": 4377 + }, + { + "epoch": 0.30379571160918745, + "grad_norm": 3.0979974269866943, + "learning_rate": 0.00019889525175819818, + "loss": 0.9653, + "step": 4378 + }, + { + "epoch": 0.3038651030462841, + "grad_norm": 4.378108978271484, + "learning_rate": 0.0001988944187398484, + "loss": 1.2083, + "step": 4379 + }, + { + "epoch": 0.3039344944833807, + "grad_norm": 2.224774122238159, + "learning_rate": 0.00019889358540930042, + "loss": 0.7332, + "step": 4380 + }, + { + "epoch": 0.3040038859204774, + "grad_norm": 4.157049179077148, + "learning_rate": 0.00019889275176655685, + "loss": 1.3968, + "step": 4381 + }, + { + "epoch": 0.30407327735757406, + "grad_norm": 3.0482044219970703, + "learning_rate": 0.00019889191781162036, + "loss": 0.9902, + "step": 4382 + }, + { + "epoch": 0.3041426687946707, + "grad_norm": 3.0972402095794678, + "learning_rate": 0.00019889108354449354, + "loss": 1.0969, + "step": 4383 + }, + { + "epoch": 0.3042120602317674, + "grad_norm": 2.9879837036132812, + "learning_rate": 0.00019889024896517905, + "loss": 0.7437, + "step": 4384 + }, + { + "epoch": 0.30428145166886406, + "grad_norm": 2.9560329914093018, + "learning_rate": 0.0001988894140736795, + "loss": 0.9141, + "step": 4385 + }, + { + "epoch": 0.3043508431059607, + "grad_norm": 3.7551329135894775, + "learning_rate": 0.0001988885788699976, + "loss": 1.1759, + "step": 4386 + }, + { + "epoch": 0.3044202345430574, + "grad_norm": 2.9069840908050537, + "learning_rate": 0.00019888774335413586, + "loss": 0.9589, + "step": 4387 + }, + { + "epoch": 0.30448962598015406, + "grad_norm": 3.22892689704895, + "learning_rate": 0.00019888690752609703, + "loss": 1.0539, + "step": 4388 + }, + { + "epoch": 0.3045590174172507, + "grad_norm": 3.062286853790283, + "learning_rate": 0.00019888607138588367, + "loss": 1.1064, + "step": 4389 + }, + { + "epoch": 0.3046284088543474, + "grad_norm": 4.100315093994141, + "learning_rate": 0.0001988852349334985, + "loss": 0.9545, + "step": 4390 + }, + { + "epoch": 0.30469780029144405, + "grad_norm": 4.004345893859863, + "learning_rate": 0.00019888439816894408, + "loss": 1.0392, + "step": 4391 + }, + { + "epoch": 0.3047671917285407, + "grad_norm": 7.18631649017334, + "learning_rate": 0.0001988835610922231, + "loss": 0.9864, + "step": 4392 + }, + { + "epoch": 0.3048365831656374, + "grad_norm": 4.278872013092041, + "learning_rate": 0.0001988827237033382, + "loss": 0.9759, + "step": 4393 + }, + { + "epoch": 0.304905974602734, + "grad_norm": 3.475792646408081, + "learning_rate": 0.000198881886002292, + "loss": 1.0224, + "step": 4394 + }, + { + "epoch": 0.30497536603983066, + "grad_norm": 8.788585662841797, + "learning_rate": 0.0001988810479890872, + "loss": 1.1001, + "step": 4395 + }, + { + "epoch": 0.30504475747692733, + "grad_norm": 3.9043455123901367, + "learning_rate": 0.00019888020966372634, + "loss": 1.1061, + "step": 4396 + }, + { + "epoch": 0.305114148914024, + "grad_norm": 7.7624592781066895, + "learning_rate": 0.00019887937102621217, + "loss": 1.0594, + "step": 4397 + }, + { + "epoch": 0.30518354035112066, + "grad_norm": 3.911916971206665, + "learning_rate": 0.0001988785320765473, + "loss": 0.9774, + "step": 4398 + }, + { + "epoch": 0.30525293178821733, + "grad_norm": 3.8841001987457275, + "learning_rate": 0.00019887769281473435, + "loss": 0.9365, + "step": 4399 + }, + { + "epoch": 0.305322323225314, + "grad_norm": 5.121184349060059, + "learning_rate": 0.000198876853240776, + "loss": 1.0197, + "step": 4400 + }, + { + "epoch": 0.30539171466241066, + "grad_norm": 4.1274895668029785, + "learning_rate": 0.0001988760133546749, + "loss": 1.0088, + "step": 4401 + }, + { + "epoch": 0.30546110609950733, + "grad_norm": 3.7116012573242188, + "learning_rate": 0.0001988751731564337, + "loss": 0.9283, + "step": 4402 + }, + { + "epoch": 0.305530497536604, + "grad_norm": 2.9586079120635986, + "learning_rate": 0.00019887433264605505, + "loss": 0.9192, + "step": 4403 + }, + { + "epoch": 0.30559988897370066, + "grad_norm": 6.841257572174072, + "learning_rate": 0.00019887349182354163, + "loss": 1.1184, + "step": 4404 + }, + { + "epoch": 0.30566928041079733, + "grad_norm": 3.0464773178100586, + "learning_rate": 0.00019887265068889605, + "loss": 1.1216, + "step": 4405 + }, + { + "epoch": 0.305738671847894, + "grad_norm": 3.6856613159179688, + "learning_rate": 0.00019887180924212095, + "loss": 0.9262, + "step": 4406 + }, + { + "epoch": 0.3058080632849906, + "grad_norm": 3.4290521144866943, + "learning_rate": 0.00019887096748321905, + "loss": 0.9044, + "step": 4407 + }, + { + "epoch": 0.30587745472208727, + "grad_norm": 3.6316821575164795, + "learning_rate": 0.00019887012541219297, + "loss": 0.9072, + "step": 4408 + }, + { + "epoch": 0.30594684615918394, + "grad_norm": 3.498518705368042, + "learning_rate": 0.00019886928302904536, + "loss": 0.9252, + "step": 4409 + }, + { + "epoch": 0.3060162375962806, + "grad_norm": 3.279813289642334, + "learning_rate": 0.00019886844033377893, + "loss": 1.1891, + "step": 4410 + }, + { + "epoch": 0.30608562903337727, + "grad_norm": 4.892227649688721, + "learning_rate": 0.0001988675973263963, + "loss": 1.2191, + "step": 4411 + }, + { + "epoch": 0.30615502047047394, + "grad_norm": 3.8988826274871826, + "learning_rate": 0.00019886675400690012, + "loss": 1.0187, + "step": 4412 + }, + { + "epoch": 0.3062244119075706, + "grad_norm": 3.5978403091430664, + "learning_rate": 0.00019886591037529308, + "loss": 0.9634, + "step": 4413 + }, + { + "epoch": 0.30629380334466727, + "grad_norm": 3.018052339553833, + "learning_rate": 0.00019886506643157782, + "loss": 1.2411, + "step": 4414 + }, + { + "epoch": 0.30636319478176394, + "grad_norm": 3.412436008453369, + "learning_rate": 0.00019886422217575703, + "loss": 1.0809, + "step": 4415 + }, + { + "epoch": 0.3064325862188606, + "grad_norm": 3.972097396850586, + "learning_rate": 0.00019886337760783334, + "loss": 1.1749, + "step": 4416 + }, + { + "epoch": 0.30650197765595727, + "grad_norm": 3.736173629760742, + "learning_rate": 0.00019886253272780944, + "loss": 1.3852, + "step": 4417 + }, + { + "epoch": 0.30657136909305394, + "grad_norm": 3.092869997024536, + "learning_rate": 0.000198861687535688, + "loss": 0.8463, + "step": 4418 + }, + { + "epoch": 0.3066407605301506, + "grad_norm": 4.248638153076172, + "learning_rate": 0.00019886084203147168, + "loss": 1.1746, + "step": 4419 + }, + { + "epoch": 0.30671015196724727, + "grad_norm": 2.643415689468384, + "learning_rate": 0.00019885999621516317, + "loss": 0.8968, + "step": 4420 + }, + { + "epoch": 0.3067795434043439, + "grad_norm": 3.0877528190612793, + "learning_rate": 0.0001988591500867651, + "loss": 1.0006, + "step": 4421 + }, + { + "epoch": 0.30684893484144055, + "grad_norm": 3.2489166259765625, + "learning_rate": 0.00019885830364628018, + "loss": 0.9244, + "step": 4422 + }, + { + "epoch": 0.3069183262785372, + "grad_norm": 3.52880859375, + "learning_rate": 0.00019885745689371105, + "loss": 1.0691, + "step": 4423 + }, + { + "epoch": 0.3069877177156339, + "grad_norm": 3.7250380516052246, + "learning_rate": 0.0001988566098290604, + "loss": 1.1509, + "step": 4424 + }, + { + "epoch": 0.30705710915273055, + "grad_norm": 2.9586329460144043, + "learning_rate": 0.0001988557624523309, + "loss": 1.0764, + "step": 4425 + }, + { + "epoch": 0.3071265005898272, + "grad_norm": 4.2440361976623535, + "learning_rate": 0.00019885491476352525, + "loss": 0.9712, + "step": 4426 + }, + { + "epoch": 0.3071958920269239, + "grad_norm": 4.119422435760498, + "learning_rate": 0.00019885406676264607, + "loss": 1.3242, + "step": 4427 + }, + { + "epoch": 0.30726528346402054, + "grad_norm": 2.964075803756714, + "learning_rate": 0.00019885321844969608, + "loss": 1.1061, + "step": 4428 + }, + { + "epoch": 0.3073346749011172, + "grad_norm": 4.063827991485596, + "learning_rate": 0.00019885236982467797, + "loss": 1.0344, + "step": 4429 + }, + { + "epoch": 0.3074040663382139, + "grad_norm": 3.296473264694214, + "learning_rate": 0.00019885152088759437, + "loss": 0.9938, + "step": 4430 + }, + { + "epoch": 0.30747345777531054, + "grad_norm": 2.79954195022583, + "learning_rate": 0.00019885067163844802, + "loss": 0.8566, + "step": 4431 + }, + { + "epoch": 0.3075428492124072, + "grad_norm": 3.4955062866210938, + "learning_rate": 0.00019884982207724153, + "loss": 1.2416, + "step": 4432 + }, + { + "epoch": 0.3076122406495039, + "grad_norm": 3.113213062286377, + "learning_rate": 0.00019884897220397765, + "loss": 1.0263, + "step": 4433 + }, + { + "epoch": 0.3076816320866005, + "grad_norm": 4.8048481941223145, + "learning_rate": 0.00019884812201865902, + "loss": 0.9712, + "step": 4434 + }, + { + "epoch": 0.30775102352369715, + "grad_norm": 2.932234287261963, + "learning_rate": 0.00019884727152128835, + "loss": 1.2293, + "step": 4435 + }, + { + "epoch": 0.3078204149607938, + "grad_norm": 6.3093695640563965, + "learning_rate": 0.0001988464207118683, + "loss": 1.2266, + "step": 4436 + }, + { + "epoch": 0.3078898063978905, + "grad_norm": 3.3219826221466064, + "learning_rate": 0.00019884556959040158, + "loss": 1.1418, + "step": 4437 + }, + { + "epoch": 0.30795919783498715, + "grad_norm": 3.601198673248291, + "learning_rate": 0.00019884471815689083, + "loss": 0.8442, + "step": 4438 + }, + { + "epoch": 0.3080285892720838, + "grad_norm": 3.6626482009887695, + "learning_rate": 0.0001988438664113388, + "loss": 1.046, + "step": 4439 + }, + { + "epoch": 0.3080979807091805, + "grad_norm": 3.3909757137298584, + "learning_rate": 0.00019884301435374818, + "loss": 1.1488, + "step": 4440 + }, + { + "epoch": 0.30816737214627715, + "grad_norm": 3.824542999267578, + "learning_rate": 0.00019884216198412158, + "loss": 0.8799, + "step": 4441 + }, + { + "epoch": 0.3082367635833738, + "grad_norm": 3.4560487270355225, + "learning_rate": 0.00019884130930246176, + "loss": 1.1127, + "step": 4442 + }, + { + "epoch": 0.3083061550204705, + "grad_norm": 3.351323366165161, + "learning_rate": 0.00019884045630877138, + "loss": 1.0306, + "step": 4443 + }, + { + "epoch": 0.30837554645756715, + "grad_norm": 3.5576670169830322, + "learning_rate": 0.00019883960300305317, + "loss": 1.4133, + "step": 4444 + }, + { + "epoch": 0.3084449378946638, + "grad_norm": 3.7649309635162354, + "learning_rate": 0.00019883874938530978, + "loss": 1.1411, + "step": 4445 + }, + { + "epoch": 0.3085143293317605, + "grad_norm": 3.3388190269470215, + "learning_rate": 0.00019883789545554396, + "loss": 1.0483, + "step": 4446 + }, + { + "epoch": 0.30858372076885715, + "grad_norm": 3.1031556129455566, + "learning_rate": 0.00019883704121375837, + "loss": 1.1953, + "step": 4447 + }, + { + "epoch": 0.30865311220595376, + "grad_norm": 2.6032934188842773, + "learning_rate": 0.0001988361866599557, + "loss": 1.0251, + "step": 4448 + }, + { + "epoch": 0.3087225036430504, + "grad_norm": 4.066292762756348, + "learning_rate": 0.00019883533179413862, + "loss": 1.3846, + "step": 4449 + }, + { + "epoch": 0.3087918950801471, + "grad_norm": 3.315275192260742, + "learning_rate": 0.0001988344766163099, + "loss": 1.0281, + "step": 4450 + }, + { + "epoch": 0.30886128651724376, + "grad_norm": 3.8551084995269775, + "learning_rate": 0.0001988336211264722, + "loss": 0.9763, + "step": 4451 + }, + { + "epoch": 0.3089306779543404, + "grad_norm": 2.9695091247558594, + "learning_rate": 0.0001988327653246282, + "loss": 0.9968, + "step": 4452 + }, + { + "epoch": 0.3090000693914371, + "grad_norm": 3.00398588180542, + "learning_rate": 0.00019883190921078067, + "loss": 1.1556, + "step": 4453 + }, + { + "epoch": 0.30906946082853376, + "grad_norm": 3.602830171585083, + "learning_rate": 0.00019883105278493225, + "loss": 0.9723, + "step": 4454 + }, + { + "epoch": 0.3091388522656304, + "grad_norm": 3.425234079360962, + "learning_rate": 0.00019883019604708565, + "loss": 1.2175, + "step": 4455 + }, + { + "epoch": 0.3092082437027271, + "grad_norm": 4.673014163970947, + "learning_rate": 0.0001988293389972436, + "loss": 1.2047, + "step": 4456 + }, + { + "epoch": 0.30927763513982376, + "grad_norm": 3.7197670936584473, + "learning_rate": 0.0001988284816354088, + "loss": 0.9136, + "step": 4457 + }, + { + "epoch": 0.3093470265769204, + "grad_norm": 4.05142068862915, + "learning_rate": 0.00019882762396158392, + "loss": 1.2311, + "step": 4458 + }, + { + "epoch": 0.3094164180140171, + "grad_norm": 2.859424114227295, + "learning_rate": 0.00019882676597577175, + "loss": 1.1109, + "step": 4459 + }, + { + "epoch": 0.30948580945111376, + "grad_norm": 3.5547070503234863, + "learning_rate": 0.0001988259076779749, + "loss": 1.179, + "step": 4460 + }, + { + "epoch": 0.30955520088821037, + "grad_norm": 3.9022738933563232, + "learning_rate": 0.00019882504906819612, + "loss": 0.956, + "step": 4461 + }, + { + "epoch": 0.30962459232530704, + "grad_norm": 3.26906418800354, + "learning_rate": 0.00019882419014643815, + "loss": 0.8462, + "step": 4462 + }, + { + "epoch": 0.3096939837624037, + "grad_norm": 3.111574172973633, + "learning_rate": 0.00019882333091270367, + "loss": 1.0781, + "step": 4463 + }, + { + "epoch": 0.30976337519950037, + "grad_norm": 3.6414523124694824, + "learning_rate": 0.0001988224713669954, + "loss": 1.0481, + "step": 4464 + }, + { + "epoch": 0.30983276663659703, + "grad_norm": 5.487741470336914, + "learning_rate": 0.00019882161150931607, + "loss": 1.1024, + "step": 4465 + }, + { + "epoch": 0.3099021580736937, + "grad_norm": 2.693664073944092, + "learning_rate": 0.00019882075133966835, + "loss": 0.8864, + "step": 4466 + }, + { + "epoch": 0.30997154951079037, + "grad_norm": 3.4253275394439697, + "learning_rate": 0.000198819890858055, + "loss": 1.1243, + "step": 4467 + }, + { + "epoch": 0.31004094094788703, + "grad_norm": 2.6221671104431152, + "learning_rate": 0.00019881903006447872, + "loss": 0.8374, + "step": 4468 + }, + { + "epoch": 0.3101103323849837, + "grad_norm": 3.3941664695739746, + "learning_rate": 0.0001988181689589422, + "loss": 1.0785, + "step": 4469 + }, + { + "epoch": 0.31017972382208037, + "grad_norm": 3.33652925491333, + "learning_rate": 0.00019881730754144823, + "loss": 1.0756, + "step": 4470 + }, + { + "epoch": 0.31024911525917703, + "grad_norm": 3.0924763679504395, + "learning_rate": 0.00019881644581199947, + "loss": 1.034, + "step": 4471 + }, + { + "epoch": 0.3103185066962737, + "grad_norm": 3.396047353744507, + "learning_rate": 0.00019881558377059865, + "loss": 1.1698, + "step": 4472 + }, + { + "epoch": 0.31038789813337037, + "grad_norm": 3.3566739559173584, + "learning_rate": 0.0001988147214172485, + "loss": 0.9968, + "step": 4473 + }, + { + "epoch": 0.31045728957046703, + "grad_norm": 3.471940279006958, + "learning_rate": 0.00019881385875195173, + "loss": 1.1632, + "step": 4474 + }, + { + "epoch": 0.31052668100756364, + "grad_norm": 3.3438236713409424, + "learning_rate": 0.0001988129957747111, + "loss": 1.2451, + "step": 4475 + }, + { + "epoch": 0.3105960724446603, + "grad_norm": 2.7223236560821533, + "learning_rate": 0.00019881213248552925, + "loss": 0.9095, + "step": 4476 + }, + { + "epoch": 0.310665463881757, + "grad_norm": 3.103987693786621, + "learning_rate": 0.000198811268884409, + "loss": 1.0286, + "step": 4477 + }, + { + "epoch": 0.31073485531885364, + "grad_norm": 3.4690120220184326, + "learning_rate": 0.00019881040497135303, + "loss": 0.9604, + "step": 4478 + }, + { + "epoch": 0.3108042467559503, + "grad_norm": 2.8113701343536377, + "learning_rate": 0.0001988095407463641, + "loss": 1.1025, + "step": 4479 + }, + { + "epoch": 0.310873638193047, + "grad_norm": 3.9171926975250244, + "learning_rate": 0.0001988086762094449, + "loss": 1.012, + "step": 4480 + }, + { + "epoch": 0.31094302963014364, + "grad_norm": 3.849820375442505, + "learning_rate": 0.00019880781136059813, + "loss": 1.2945, + "step": 4481 + }, + { + "epoch": 0.3110124210672403, + "grad_norm": 2.8950390815734863, + "learning_rate": 0.00019880694619982663, + "loss": 0.9196, + "step": 4482 + }, + { + "epoch": 0.311081812504337, + "grad_norm": 2.988781690597534, + "learning_rate": 0.00019880608072713303, + "loss": 1.0501, + "step": 4483 + }, + { + "epoch": 0.31115120394143364, + "grad_norm": 3.5369954109191895, + "learning_rate": 0.00019880521494252009, + "loss": 1.1956, + "step": 4484 + }, + { + "epoch": 0.3112205953785303, + "grad_norm": 3.210972309112549, + "learning_rate": 0.00019880434884599058, + "loss": 1.2862, + "step": 4485 + }, + { + "epoch": 0.311289986815627, + "grad_norm": 3.845564842224121, + "learning_rate": 0.0001988034824375472, + "loss": 0.8999, + "step": 4486 + }, + { + "epoch": 0.31135937825272364, + "grad_norm": 3.2635293006896973, + "learning_rate": 0.00019880261571719266, + "loss": 0.7672, + "step": 4487 + }, + { + "epoch": 0.31142876968982025, + "grad_norm": 6.029399394989014, + "learning_rate": 0.00019880174868492972, + "loss": 0.9902, + "step": 4488 + }, + { + "epoch": 0.3114981611269169, + "grad_norm": 3.839120388031006, + "learning_rate": 0.00019880088134076115, + "loss": 0.8496, + "step": 4489 + }, + { + "epoch": 0.3115675525640136, + "grad_norm": 2.91326904296875, + "learning_rate": 0.00019880001368468965, + "loss": 1.1793, + "step": 4490 + }, + { + "epoch": 0.31163694400111025, + "grad_norm": 2.540118932723999, + "learning_rate": 0.00019879914571671798, + "loss": 0.707, + "step": 4491 + }, + { + "epoch": 0.3117063354382069, + "grad_norm": 2.9754700660705566, + "learning_rate": 0.00019879827743684886, + "loss": 1.1549, + "step": 4492 + }, + { + "epoch": 0.3117757268753036, + "grad_norm": 2.3787622451782227, + "learning_rate": 0.00019879740884508505, + "loss": 0.7483, + "step": 4493 + }, + { + "epoch": 0.31184511831240025, + "grad_norm": 4.744495868682861, + "learning_rate": 0.0001987965399414293, + "loss": 1.0106, + "step": 4494 + }, + { + "epoch": 0.3119145097494969, + "grad_norm": 3.976792097091675, + "learning_rate": 0.0001987956707258843, + "loss": 1.191, + "step": 4495 + }, + { + "epoch": 0.3119839011865936, + "grad_norm": 2.3766391277313232, + "learning_rate": 0.00019879480119845285, + "loss": 0.8073, + "step": 4496 + }, + { + "epoch": 0.31205329262369025, + "grad_norm": 3.2649190425872803, + "learning_rate": 0.00019879393135913763, + "loss": 1.07, + "step": 4497 + }, + { + "epoch": 0.3121226840607869, + "grad_norm": 3.015321731567383, + "learning_rate": 0.0001987930612079415, + "loss": 0.7506, + "step": 4498 + }, + { + "epoch": 0.3121920754978836, + "grad_norm": 2.900686264038086, + "learning_rate": 0.0001987921907448671, + "loss": 0.963, + "step": 4499 + }, + { + "epoch": 0.31226146693498025, + "grad_norm": 3.155632734298706, + "learning_rate": 0.00019879131996991724, + "loss": 1.1563, + "step": 4500 + }, + { + "epoch": 0.31233085837207686, + "grad_norm": 4.489522457122803, + "learning_rate": 0.0001987904488830946, + "loss": 1.2597, + "step": 4501 + }, + { + "epoch": 0.3124002498091735, + "grad_norm": 3.8137564659118652, + "learning_rate": 0.000198789577484402, + "loss": 1.0058, + "step": 4502 + }, + { + "epoch": 0.3124696412462702, + "grad_norm": 3.640815019607544, + "learning_rate": 0.00019878870577384218, + "loss": 1.0476, + "step": 4503 + }, + { + "epoch": 0.31253903268336686, + "grad_norm": 2.7856264114379883, + "learning_rate": 0.00019878783375141786, + "loss": 1.0487, + "step": 4504 + }, + { + "epoch": 0.3126084241204635, + "grad_norm": 3.5064661502838135, + "learning_rate": 0.0001987869614171318, + "loss": 1.0933, + "step": 4505 + }, + { + "epoch": 0.3126778155575602, + "grad_norm": 3.1858201026916504, + "learning_rate": 0.0001987860887709868, + "loss": 0.9387, + "step": 4506 + }, + { + "epoch": 0.31274720699465686, + "grad_norm": 4.271132469177246, + "learning_rate": 0.00019878521581298558, + "loss": 1.0484, + "step": 4507 + }, + { + "epoch": 0.3128165984317535, + "grad_norm": 4.13946533203125, + "learning_rate": 0.00019878434254313085, + "loss": 0.9991, + "step": 4508 + }, + { + "epoch": 0.3128859898688502, + "grad_norm": 3.9074878692626953, + "learning_rate": 0.00019878346896142545, + "loss": 1.3112, + "step": 4509 + }, + { + "epoch": 0.31295538130594686, + "grad_norm": 2.304997444152832, + "learning_rate": 0.00019878259506787206, + "loss": 0.9011, + "step": 4510 + }, + { + "epoch": 0.3130247727430435, + "grad_norm": 3.854234218597412, + "learning_rate": 0.0001987817208624735, + "loss": 0.8369, + "step": 4511 + }, + { + "epoch": 0.3130941641801402, + "grad_norm": 3.353780746459961, + "learning_rate": 0.00019878084634523252, + "loss": 1.0894, + "step": 4512 + }, + { + "epoch": 0.31316355561723686, + "grad_norm": 3.667729377746582, + "learning_rate": 0.0001987799715161519, + "loss": 1.1711, + "step": 4513 + }, + { + "epoch": 0.3132329470543335, + "grad_norm": 3.367884635925293, + "learning_rate": 0.00019877909637523432, + "loss": 1.0167, + "step": 4514 + }, + { + "epoch": 0.31330233849143013, + "grad_norm": 3.531792163848877, + "learning_rate": 0.00019877822092248258, + "loss": 1.0188, + "step": 4515 + }, + { + "epoch": 0.3133717299285268, + "grad_norm": 3.4773998260498047, + "learning_rate": 0.0001987773451578995, + "loss": 0.8694, + "step": 4516 + }, + { + "epoch": 0.31344112136562347, + "grad_norm": 3.6949615478515625, + "learning_rate": 0.00019877646908148777, + "loss": 1.0889, + "step": 4517 + }, + { + "epoch": 0.31351051280272013, + "grad_norm": 4.451809883117676, + "learning_rate": 0.00019877559269325018, + "loss": 0.7865, + "step": 4518 + }, + { + "epoch": 0.3135799042398168, + "grad_norm": 4.853586673736572, + "learning_rate": 0.00019877471599318955, + "loss": 0.8028, + "step": 4519 + }, + { + "epoch": 0.31364929567691346, + "grad_norm": 3.503004312515259, + "learning_rate": 0.00019877383898130857, + "loss": 1.0107, + "step": 4520 + }, + { + "epoch": 0.31371868711401013, + "grad_norm": 4.058359146118164, + "learning_rate": 0.00019877296165761004, + "loss": 1.1149, + "step": 4521 + }, + { + "epoch": 0.3137880785511068, + "grad_norm": 4.613432884216309, + "learning_rate": 0.00019877208402209676, + "loss": 1.1804, + "step": 4522 + }, + { + "epoch": 0.31385746998820346, + "grad_norm": 4.251614570617676, + "learning_rate": 0.0001987712060747714, + "loss": 1.1485, + "step": 4523 + }, + { + "epoch": 0.31392686142530013, + "grad_norm": 4.562783718109131, + "learning_rate": 0.00019877032781563688, + "loss": 0.9414, + "step": 4524 + }, + { + "epoch": 0.3139962528623968, + "grad_norm": 3.608121156692505, + "learning_rate": 0.00019876944924469587, + "loss": 1.0718, + "step": 4525 + }, + { + "epoch": 0.31406564429949346, + "grad_norm": 4.5058369636535645, + "learning_rate": 0.00019876857036195116, + "loss": 1.172, + "step": 4526 + }, + { + "epoch": 0.31413503573659013, + "grad_norm": 3.565092086791992, + "learning_rate": 0.0001987676911674055, + "loss": 1.2505, + "step": 4527 + }, + { + "epoch": 0.31420442717368674, + "grad_norm": 3.143759250640869, + "learning_rate": 0.00019876681166106175, + "loss": 1.0238, + "step": 4528 + }, + { + "epoch": 0.3142738186107834, + "grad_norm": 3.7865970134735107, + "learning_rate": 0.0001987659318429226, + "loss": 0.9821, + "step": 4529 + }, + { + "epoch": 0.3143432100478801, + "grad_norm": 2.6564908027648926, + "learning_rate": 0.0001987650517129909, + "loss": 0.7827, + "step": 4530 + }, + { + "epoch": 0.31441260148497674, + "grad_norm": 2.7128889560699463, + "learning_rate": 0.00019876417127126937, + "loss": 0.9555, + "step": 4531 + }, + { + "epoch": 0.3144819929220734, + "grad_norm": 4.325357913970947, + "learning_rate": 0.0001987632905177608, + "loss": 1.2559, + "step": 4532 + }, + { + "epoch": 0.31455138435917007, + "grad_norm": 3.082151412963867, + "learning_rate": 0.000198762409452468, + "loss": 1.0152, + "step": 4533 + }, + { + "epoch": 0.31462077579626674, + "grad_norm": 3.219622850418091, + "learning_rate": 0.0001987615280753937, + "loss": 0.8595, + "step": 4534 + }, + { + "epoch": 0.3146901672333634, + "grad_norm": 2.9628968238830566, + "learning_rate": 0.00019876064638654076, + "loss": 1.0816, + "step": 4535 + }, + { + "epoch": 0.31475955867046007, + "grad_norm": 2.880220651626587, + "learning_rate": 0.0001987597643859119, + "loss": 1.0529, + "step": 4536 + }, + { + "epoch": 0.31482895010755674, + "grad_norm": 3.690671443939209, + "learning_rate": 0.0001987588820735099, + "loss": 1.0261, + "step": 4537 + }, + { + "epoch": 0.3148983415446534, + "grad_norm": 5.158776760101318, + "learning_rate": 0.0001987579994493376, + "loss": 1.1953, + "step": 4538 + }, + { + "epoch": 0.31496773298175007, + "grad_norm": 3.112961769104004, + "learning_rate": 0.00019875711651339773, + "loss": 1.0342, + "step": 4539 + }, + { + "epoch": 0.31503712441884674, + "grad_norm": 3.5792441368103027, + "learning_rate": 0.0001987562332656931, + "loss": 0.9935, + "step": 4540 + }, + { + "epoch": 0.3151065158559434, + "grad_norm": 3.4635448455810547, + "learning_rate": 0.0001987553497062265, + "loss": 1.0542, + "step": 4541 + }, + { + "epoch": 0.31517590729304, + "grad_norm": 4.891063690185547, + "learning_rate": 0.00019875446583500072, + "loss": 1.1804, + "step": 4542 + }, + { + "epoch": 0.3152452987301367, + "grad_norm": 4.630155563354492, + "learning_rate": 0.00019875358165201857, + "loss": 0.912, + "step": 4543 + }, + { + "epoch": 0.31531469016723335, + "grad_norm": 2.9719409942626953, + "learning_rate": 0.0001987526971572828, + "loss": 1.0633, + "step": 4544 + }, + { + "epoch": 0.31538408160433, + "grad_norm": 3.1701886653900146, + "learning_rate": 0.00019875181235079622, + "loss": 0.9933, + "step": 4545 + }, + { + "epoch": 0.3154534730414267, + "grad_norm": 4.728122711181641, + "learning_rate": 0.00019875092723256163, + "loss": 1.014, + "step": 4546 + }, + { + "epoch": 0.31552286447852335, + "grad_norm": 3.1162800788879395, + "learning_rate": 0.00019875004180258183, + "loss": 1.0261, + "step": 4547 + }, + { + "epoch": 0.31559225591562, + "grad_norm": 3.0477404594421387, + "learning_rate": 0.0001987491560608596, + "loss": 0.8575, + "step": 4548 + }, + { + "epoch": 0.3156616473527167, + "grad_norm": 2.5927231311798096, + "learning_rate": 0.00019874827000739776, + "loss": 0.9869, + "step": 4549 + }, + { + "epoch": 0.31573103878981335, + "grad_norm": 3.1427314281463623, + "learning_rate": 0.00019874738364219905, + "loss": 0.9337, + "step": 4550 + }, + { + "epoch": 0.31580043022691, + "grad_norm": 2.5341780185699463, + "learning_rate": 0.00019874649696526633, + "loss": 0.9164, + "step": 4551 + }, + { + "epoch": 0.3158698216640067, + "grad_norm": 3.3299145698547363, + "learning_rate": 0.00019874560997660236, + "loss": 0.8227, + "step": 4552 + }, + { + "epoch": 0.31593921310110334, + "grad_norm": 3.9313511848449707, + "learning_rate": 0.00019874472267621, + "loss": 1.4161, + "step": 4553 + }, + { + "epoch": 0.3160086045382, + "grad_norm": 3.8262691497802734, + "learning_rate": 0.00019874383506409196, + "loss": 1.0654, + "step": 4554 + }, + { + "epoch": 0.3160779959752966, + "grad_norm": 3.2259507179260254, + "learning_rate": 0.0001987429471402511, + "loss": 0.9942, + "step": 4555 + }, + { + "epoch": 0.3161473874123933, + "grad_norm": 3.5048084259033203, + "learning_rate": 0.0001987420589046902, + "loss": 1.0974, + "step": 4556 + }, + { + "epoch": 0.31621677884948995, + "grad_norm": 3.128458261489868, + "learning_rate": 0.0001987411703574121, + "loss": 1.0054, + "step": 4557 + }, + { + "epoch": 0.3162861702865866, + "grad_norm": 3.316422462463379, + "learning_rate": 0.00019874028149841956, + "loss": 0.9931, + "step": 4558 + }, + { + "epoch": 0.3163555617236833, + "grad_norm": 4.188426494598389, + "learning_rate": 0.00019873939232771542, + "loss": 1.2038, + "step": 4559 + }, + { + "epoch": 0.31642495316077995, + "grad_norm": 4.123785972595215, + "learning_rate": 0.00019873850284530246, + "loss": 0.9649, + "step": 4560 + }, + { + "epoch": 0.3164943445978766, + "grad_norm": 3.14243221282959, + "learning_rate": 0.00019873761305118354, + "loss": 0.9413, + "step": 4561 + }, + { + "epoch": 0.3165637360349733, + "grad_norm": 3.971142530441284, + "learning_rate": 0.00019873672294536137, + "loss": 0.8798, + "step": 4562 + }, + { + "epoch": 0.31663312747206995, + "grad_norm": 3.0276122093200684, + "learning_rate": 0.0001987358325278389, + "loss": 0.9411, + "step": 4563 + }, + { + "epoch": 0.3167025189091666, + "grad_norm": 3.1345691680908203, + "learning_rate": 0.0001987349417986188, + "loss": 1.0439, + "step": 4564 + }, + { + "epoch": 0.3167719103462633, + "grad_norm": 2.690321922302246, + "learning_rate": 0.00019873405075770397, + "loss": 0.8795, + "step": 4565 + }, + { + "epoch": 0.31684130178335995, + "grad_norm": 3.263361930847168, + "learning_rate": 0.0001987331594050972, + "loss": 1.0317, + "step": 4566 + }, + { + "epoch": 0.3169106932204566, + "grad_norm": 3.3324429988861084, + "learning_rate": 0.00019873226774080128, + "loss": 0.9896, + "step": 4567 + }, + { + "epoch": 0.3169800846575533, + "grad_norm": 2.7956273555755615, + "learning_rate": 0.0001987313757648191, + "loss": 0.8889, + "step": 4568 + }, + { + "epoch": 0.3170494760946499, + "grad_norm": 3.502159595489502, + "learning_rate": 0.00019873048347715335, + "loss": 1.1939, + "step": 4569 + }, + { + "epoch": 0.31711886753174656, + "grad_norm": 3.027878522872925, + "learning_rate": 0.00019872959087780694, + "loss": 1.0622, + "step": 4570 + }, + { + "epoch": 0.31718825896884323, + "grad_norm": 3.007949113845825, + "learning_rate": 0.0001987286979667827, + "loss": 1.1785, + "step": 4571 + }, + { + "epoch": 0.3172576504059399, + "grad_norm": 3.373533010482788, + "learning_rate": 0.00019872780474408338, + "loss": 1.0514, + "step": 4572 + }, + { + "epoch": 0.31732704184303656, + "grad_norm": 3.613291025161743, + "learning_rate": 0.00019872691120971183, + "loss": 0.9966, + "step": 4573 + }, + { + "epoch": 0.31739643328013323, + "grad_norm": 4.3197503089904785, + "learning_rate": 0.00019872601736367092, + "loss": 1.1287, + "step": 4574 + }, + { + "epoch": 0.3174658247172299, + "grad_norm": 3.1224844455718994, + "learning_rate": 0.00019872512320596338, + "loss": 1.0865, + "step": 4575 + }, + { + "epoch": 0.31753521615432656, + "grad_norm": 3.645570755004883, + "learning_rate": 0.00019872422873659212, + "loss": 1.1325, + "step": 4576 + }, + { + "epoch": 0.3176046075914232, + "grad_norm": 3.2805469036102295, + "learning_rate": 0.0001987233339555599, + "loss": 1.0425, + "step": 4577 + }, + { + "epoch": 0.3176739990285199, + "grad_norm": 2.9458231925964355, + "learning_rate": 0.00019872243886286955, + "loss": 0.8796, + "step": 4578 + }, + { + "epoch": 0.31774339046561656, + "grad_norm": 3.2348592281341553, + "learning_rate": 0.00019872154345852393, + "loss": 1.0606, + "step": 4579 + }, + { + "epoch": 0.3178127819027132, + "grad_norm": 2.835059404373169, + "learning_rate": 0.00019872064774252584, + "loss": 0.9985, + "step": 4580 + }, + { + "epoch": 0.3178821733398099, + "grad_norm": 2.949143409729004, + "learning_rate": 0.00019871975171487818, + "loss": 1.1522, + "step": 4581 + }, + { + "epoch": 0.3179515647769065, + "grad_norm": 3.7744877338409424, + "learning_rate": 0.00019871885537558367, + "loss": 0.9692, + "step": 4582 + }, + { + "epoch": 0.31802095621400317, + "grad_norm": 4.270084381103516, + "learning_rate": 0.0001987179587246452, + "loss": 1.0462, + "step": 4583 + }, + { + "epoch": 0.31809034765109984, + "grad_norm": 3.3940677642822266, + "learning_rate": 0.00019871706176206554, + "loss": 0.8328, + "step": 4584 + }, + { + "epoch": 0.3181597390881965, + "grad_norm": 2.952585220336914, + "learning_rate": 0.00019871616448784762, + "loss": 1.0605, + "step": 4585 + }, + { + "epoch": 0.31822913052529317, + "grad_norm": 3.1077044010162354, + "learning_rate": 0.00019871526690199424, + "loss": 1.0614, + "step": 4586 + }, + { + "epoch": 0.31829852196238984, + "grad_norm": 2.783047676086426, + "learning_rate": 0.00019871436900450816, + "loss": 0.9545, + "step": 4587 + }, + { + "epoch": 0.3183679133994865, + "grad_norm": 2.954874038696289, + "learning_rate": 0.00019871347079539227, + "loss": 0.6475, + "step": 4588 + }, + { + "epoch": 0.31843730483658317, + "grad_norm": 3.5663390159606934, + "learning_rate": 0.00019871257227464947, + "loss": 1.0407, + "step": 4589 + }, + { + "epoch": 0.31850669627367983, + "grad_norm": 3.5557024478912354, + "learning_rate": 0.00019871167344228248, + "loss": 1.2255, + "step": 4590 + }, + { + "epoch": 0.3185760877107765, + "grad_norm": 2.967137336730957, + "learning_rate": 0.00019871077429829418, + "loss": 0.932, + "step": 4591 + }, + { + "epoch": 0.31864547914787317, + "grad_norm": 3.3435821533203125, + "learning_rate": 0.00019870987484268745, + "loss": 1.119, + "step": 4592 + }, + { + "epoch": 0.31871487058496983, + "grad_norm": 2.609048366546631, + "learning_rate": 0.0001987089750754651, + "loss": 0.8206, + "step": 4593 + }, + { + "epoch": 0.3187842620220665, + "grad_norm": 3.2709670066833496, + "learning_rate": 0.00019870807499662997, + "loss": 0.7984, + "step": 4594 + }, + { + "epoch": 0.31885365345916317, + "grad_norm": 3.324880599975586, + "learning_rate": 0.00019870717460618488, + "loss": 1.0506, + "step": 4595 + }, + { + "epoch": 0.3189230448962598, + "grad_norm": 2.80951189994812, + "learning_rate": 0.00019870627390413268, + "loss": 0.8948, + "step": 4596 + }, + { + "epoch": 0.31899243633335644, + "grad_norm": 3.5313162803649902, + "learning_rate": 0.00019870537289047626, + "loss": 1.3012, + "step": 4597 + }, + { + "epoch": 0.3190618277704531, + "grad_norm": 3.336869239807129, + "learning_rate": 0.00019870447156521842, + "loss": 1.2602, + "step": 4598 + }, + { + "epoch": 0.3191312192075498, + "grad_norm": 2.5982189178466797, + "learning_rate": 0.00019870356992836202, + "loss": 0.7942, + "step": 4599 + }, + { + "epoch": 0.31920061064464644, + "grad_norm": 2.799548864364624, + "learning_rate": 0.00019870266797990987, + "loss": 0.9099, + "step": 4600 + }, + { + "epoch": 0.3192700020817431, + "grad_norm": 2.1470019817352295, + "learning_rate": 0.00019870176571986487, + "loss": 0.7664, + "step": 4601 + }, + { + "epoch": 0.3193393935188398, + "grad_norm": 3.063249111175537, + "learning_rate": 0.00019870086314822984, + "loss": 1.037, + "step": 4602 + }, + { + "epoch": 0.31940878495593644, + "grad_norm": 2.9386818408966064, + "learning_rate": 0.00019869996026500766, + "loss": 0.8268, + "step": 4603 + }, + { + "epoch": 0.3194781763930331, + "grad_norm": 3.001249074935913, + "learning_rate": 0.00019869905707020112, + "loss": 1.0569, + "step": 4604 + }, + { + "epoch": 0.3195475678301298, + "grad_norm": 3.0286755561828613, + "learning_rate": 0.00019869815356381314, + "loss": 0.9998, + "step": 4605 + }, + { + "epoch": 0.31961695926722644, + "grad_norm": 3.412543773651123, + "learning_rate": 0.00019869724974584652, + "loss": 0.9244, + "step": 4606 + }, + { + "epoch": 0.3196863507043231, + "grad_norm": 6.86044454574585, + "learning_rate": 0.00019869634561630415, + "loss": 0.8356, + "step": 4607 + }, + { + "epoch": 0.3197557421414198, + "grad_norm": 3.4858293533325195, + "learning_rate": 0.00019869544117518886, + "loss": 1.298, + "step": 4608 + }, + { + "epoch": 0.3198251335785164, + "grad_norm": 3.0177886486053467, + "learning_rate": 0.0001986945364225035, + "loss": 1.0485, + "step": 4609 + }, + { + "epoch": 0.31989452501561305, + "grad_norm": 3.0423007011413574, + "learning_rate": 0.00019869363135825095, + "loss": 1.0681, + "step": 4610 + }, + { + "epoch": 0.3199639164527097, + "grad_norm": 2.1849710941314697, + "learning_rate": 0.00019869272598243404, + "loss": 0.7432, + "step": 4611 + }, + { + "epoch": 0.3200333078898064, + "grad_norm": 2.925135850906372, + "learning_rate": 0.00019869182029505568, + "loss": 0.8758, + "step": 4612 + }, + { + "epoch": 0.32010269932690305, + "grad_norm": 2.4741625785827637, + "learning_rate": 0.00019869091429611867, + "loss": 0.9614, + "step": 4613 + }, + { + "epoch": 0.3201720907639997, + "grad_norm": 3.1915369033813477, + "learning_rate": 0.0001986900079856259, + "loss": 1.1524, + "step": 4614 + }, + { + "epoch": 0.3202414822010964, + "grad_norm": 3.5106704235076904, + "learning_rate": 0.00019868910136358023, + "loss": 1.0639, + "step": 4615 + }, + { + "epoch": 0.32031087363819305, + "grad_norm": 4.84419059753418, + "learning_rate": 0.0001986881944299845, + "loss": 1.1835, + "step": 4616 + }, + { + "epoch": 0.3203802650752897, + "grad_norm": 3.5641860961914062, + "learning_rate": 0.0001986872871848416, + "loss": 1.304, + "step": 4617 + }, + { + "epoch": 0.3204496565123864, + "grad_norm": 3.152381658554077, + "learning_rate": 0.0001986863796281544, + "loss": 1.1016, + "step": 4618 + }, + { + "epoch": 0.32051904794948305, + "grad_norm": 4.880514621734619, + "learning_rate": 0.0001986854717599257, + "loss": 1.1002, + "step": 4619 + }, + { + "epoch": 0.3205884393865797, + "grad_norm": 2.7034242153167725, + "learning_rate": 0.00019868456358015845, + "loss": 0.9335, + "step": 4620 + }, + { + "epoch": 0.3206578308236764, + "grad_norm": 3.150911569595337, + "learning_rate": 0.0001986836550888555, + "loss": 1.1353, + "step": 4621 + }, + { + "epoch": 0.32072722226077305, + "grad_norm": 2.4431068897247314, + "learning_rate": 0.00019868274628601968, + "loss": 0.8499, + "step": 4622 + }, + { + "epoch": 0.32079661369786966, + "grad_norm": 3.8629062175750732, + "learning_rate": 0.0001986818371716539, + "loss": 1.229, + "step": 4623 + }, + { + "epoch": 0.3208660051349663, + "grad_norm": 3.3199055194854736, + "learning_rate": 0.00019868092774576098, + "loss": 1.191, + "step": 4624 + }, + { + "epoch": 0.320935396572063, + "grad_norm": 2.700859546661377, + "learning_rate": 0.00019868001800834383, + "loss": 0.9475, + "step": 4625 + }, + { + "epoch": 0.32100478800915966, + "grad_norm": 2.3900089263916016, + "learning_rate": 0.0001986791079594053, + "loss": 0.8892, + "step": 4626 + }, + { + "epoch": 0.3210741794462563, + "grad_norm": 3.715404987335205, + "learning_rate": 0.0001986781975989483, + "loss": 0.9839, + "step": 4627 + }, + { + "epoch": 0.321143570883353, + "grad_norm": 3.8728082180023193, + "learning_rate": 0.00019867728692697566, + "loss": 1.1797, + "step": 4628 + }, + { + "epoch": 0.32121296232044966, + "grad_norm": 3.4523253440856934, + "learning_rate": 0.0001986763759434903, + "loss": 1.2251, + "step": 4629 + }, + { + "epoch": 0.3212823537575463, + "grad_norm": 2.8339693546295166, + "learning_rate": 0.00019867546464849505, + "loss": 0.9137, + "step": 4630 + }, + { + "epoch": 0.321351745194643, + "grad_norm": 3.879979372024536, + "learning_rate": 0.00019867455304199282, + "loss": 1.234, + "step": 4631 + }, + { + "epoch": 0.32142113663173966, + "grad_norm": 3.055455446243286, + "learning_rate": 0.00019867364112398646, + "loss": 1.0535, + "step": 4632 + }, + { + "epoch": 0.3214905280688363, + "grad_norm": 3.32818341255188, + "learning_rate": 0.00019867272889447884, + "loss": 1.0761, + "step": 4633 + }, + { + "epoch": 0.321559919505933, + "grad_norm": 2.8055245876312256, + "learning_rate": 0.0001986718163534729, + "loss": 1.0264, + "step": 4634 + }, + { + "epoch": 0.32162931094302966, + "grad_norm": 3.3412036895751953, + "learning_rate": 0.00019867090350097146, + "loss": 0.8627, + "step": 4635 + }, + { + "epoch": 0.32169870238012627, + "grad_norm": 3.2234158515930176, + "learning_rate": 0.00019866999033697744, + "loss": 0.7426, + "step": 4636 + }, + { + "epoch": 0.32176809381722293, + "grad_norm": 3.4756100177764893, + "learning_rate": 0.0001986690768614937, + "loss": 1.1833, + "step": 4637 + }, + { + "epoch": 0.3218374852543196, + "grad_norm": 2.5059378147125244, + "learning_rate": 0.00019866816307452314, + "loss": 0.8552, + "step": 4638 + }, + { + "epoch": 0.32190687669141627, + "grad_norm": 3.66745924949646, + "learning_rate": 0.00019866724897606863, + "loss": 0.8192, + "step": 4639 + }, + { + "epoch": 0.32197626812851293, + "grad_norm": 3.24941086769104, + "learning_rate": 0.00019866633456613308, + "loss": 1.0106, + "step": 4640 + }, + { + "epoch": 0.3220456595656096, + "grad_norm": 3.578880548477173, + "learning_rate": 0.00019866541984471937, + "loss": 1.0966, + "step": 4641 + }, + { + "epoch": 0.32211505100270627, + "grad_norm": 2.810762643814087, + "learning_rate": 0.0001986645048118303, + "loss": 1.1072, + "step": 4642 + }, + { + "epoch": 0.32218444243980293, + "grad_norm": 3.890200138092041, + "learning_rate": 0.00019866358946746891, + "loss": 0.9544, + "step": 4643 + }, + { + "epoch": 0.3222538338768996, + "grad_norm": 3.826913356781006, + "learning_rate": 0.000198662673811638, + "loss": 0.8775, + "step": 4644 + }, + { + "epoch": 0.32232322531399626, + "grad_norm": 6.296238422393799, + "learning_rate": 0.00019866175784434046, + "loss": 1.6264, + "step": 4645 + }, + { + "epoch": 0.32239261675109293, + "grad_norm": 7.142650127410889, + "learning_rate": 0.0001986608415655792, + "loss": 1.0554, + "step": 4646 + }, + { + "epoch": 0.3224620081881896, + "grad_norm": 3.5356600284576416, + "learning_rate": 0.00019865992497535711, + "loss": 1.0171, + "step": 4647 + }, + { + "epoch": 0.32253139962528626, + "grad_norm": 4.1360182762146, + "learning_rate": 0.00019865900807367713, + "loss": 1.0406, + "step": 4648 + }, + { + "epoch": 0.3226007910623829, + "grad_norm": 2.8266677856445312, + "learning_rate": 0.00019865809086054207, + "loss": 0.7049, + "step": 4649 + }, + { + "epoch": 0.32267018249947954, + "grad_norm": 3.099592447280884, + "learning_rate": 0.00019865717333595483, + "loss": 0.8987, + "step": 4650 + }, + { + "epoch": 0.3227395739365762, + "grad_norm": 3.669618606567383, + "learning_rate": 0.00019865625549991839, + "loss": 1.1386, + "step": 4651 + }, + { + "epoch": 0.3228089653736729, + "grad_norm": 3.1456894874572754, + "learning_rate": 0.00019865533735243557, + "loss": 1.0075, + "step": 4652 + }, + { + "epoch": 0.32287835681076954, + "grad_norm": 3.9617395401000977, + "learning_rate": 0.0001986544188935093, + "loss": 1.1065, + "step": 4653 + }, + { + "epoch": 0.3229477482478662, + "grad_norm": 3.9068360328674316, + "learning_rate": 0.00019865350012314247, + "loss": 1.2076, + "step": 4654 + }, + { + "epoch": 0.3230171396849629, + "grad_norm": 3.1899938583374023, + "learning_rate": 0.00019865258104133804, + "loss": 1.0842, + "step": 4655 + }, + { + "epoch": 0.32308653112205954, + "grad_norm": 3.9064180850982666, + "learning_rate": 0.0001986516616480988, + "loss": 1.1698, + "step": 4656 + }, + { + "epoch": 0.3231559225591562, + "grad_norm": 4.281027793884277, + "learning_rate": 0.00019865074194342773, + "loss": 1.2685, + "step": 4657 + }, + { + "epoch": 0.32322531399625287, + "grad_norm": 3.2564401626586914, + "learning_rate": 0.0001986498219273277, + "loss": 0.985, + "step": 4658 + }, + { + "epoch": 0.32329470543334954, + "grad_norm": 4.618688583374023, + "learning_rate": 0.00019864890159980165, + "loss": 0.9755, + "step": 4659 + }, + { + "epoch": 0.3233640968704462, + "grad_norm": 2.197619676589966, + "learning_rate": 0.00019864798096085244, + "loss": 0.925, + "step": 4660 + }, + { + "epoch": 0.32343348830754287, + "grad_norm": 2.592559814453125, + "learning_rate": 0.000198647060010483, + "loss": 1.0284, + "step": 4661 + }, + { + "epoch": 0.32350287974463954, + "grad_norm": 2.859186887741089, + "learning_rate": 0.00019864613874869623, + "loss": 0.9012, + "step": 4662 + }, + { + "epoch": 0.32357227118173615, + "grad_norm": 3.155900716781616, + "learning_rate": 0.00019864521717549507, + "loss": 0.9549, + "step": 4663 + }, + { + "epoch": 0.3236416626188328, + "grad_norm": 4.598819732666016, + "learning_rate": 0.00019864429529088242, + "loss": 1.0744, + "step": 4664 + }, + { + "epoch": 0.3237110540559295, + "grad_norm": 3.1988139152526855, + "learning_rate": 0.00019864337309486113, + "loss": 1.019, + "step": 4665 + }, + { + "epoch": 0.32378044549302615, + "grad_norm": 4.4413652420043945, + "learning_rate": 0.00019864245058743419, + "loss": 1.2002, + "step": 4666 + }, + { + "epoch": 0.3238498369301228, + "grad_norm": 2.751081943511963, + "learning_rate": 0.00019864152776860445, + "loss": 1.0902, + "step": 4667 + }, + { + "epoch": 0.3239192283672195, + "grad_norm": 3.873476982116699, + "learning_rate": 0.00019864060463837486, + "loss": 1.2417, + "step": 4668 + }, + { + "epoch": 0.32398861980431615, + "grad_norm": 3.6469645500183105, + "learning_rate": 0.00019863968119674833, + "loss": 1.1948, + "step": 4669 + }, + { + "epoch": 0.3240580112414128, + "grad_norm": 3.8258512020111084, + "learning_rate": 0.00019863875744372778, + "loss": 0.943, + "step": 4670 + }, + { + "epoch": 0.3241274026785095, + "grad_norm": 3.789234161376953, + "learning_rate": 0.0001986378333793161, + "loss": 1.2272, + "step": 4671 + }, + { + "epoch": 0.32419679411560615, + "grad_norm": 3.380955457687378, + "learning_rate": 0.00019863690900351623, + "loss": 1.1196, + "step": 4672 + }, + { + "epoch": 0.3242661855527028, + "grad_norm": 3.7191965579986572, + "learning_rate": 0.0001986359843163311, + "loss": 1.0682, + "step": 4673 + }, + { + "epoch": 0.3243355769897995, + "grad_norm": 2.461712121963501, + "learning_rate": 0.00019863505931776358, + "loss": 0.7859, + "step": 4674 + }, + { + "epoch": 0.32440496842689615, + "grad_norm": 3.6865291595458984, + "learning_rate": 0.00019863413400781664, + "loss": 1.0928, + "step": 4675 + }, + { + "epoch": 0.32447435986399276, + "grad_norm": 3.4784603118896484, + "learning_rate": 0.0001986332083864932, + "loss": 1.1911, + "step": 4676 + }, + { + "epoch": 0.3245437513010894, + "grad_norm": 3.9510021209716797, + "learning_rate": 0.00019863228245379613, + "loss": 1.1243, + "step": 4677 + }, + { + "epoch": 0.3246131427381861, + "grad_norm": 3.8686304092407227, + "learning_rate": 0.00019863135620972838, + "loss": 1.0054, + "step": 4678 + }, + { + "epoch": 0.32468253417528276, + "grad_norm": 4.272150993347168, + "learning_rate": 0.00019863042965429293, + "loss": 1.1803, + "step": 4679 + }, + { + "epoch": 0.3247519256123794, + "grad_norm": 3.467108726501465, + "learning_rate": 0.00019862950278749263, + "loss": 1.2755, + "step": 4680 + }, + { + "epoch": 0.3248213170494761, + "grad_norm": 3.157477378845215, + "learning_rate": 0.00019862857560933045, + "loss": 1.1699, + "step": 4681 + }, + { + "epoch": 0.32489070848657275, + "grad_norm": 4.2573981285095215, + "learning_rate": 0.00019862764811980928, + "loss": 1.0774, + "step": 4682 + }, + { + "epoch": 0.3249600999236694, + "grad_norm": 2.9474565982818604, + "learning_rate": 0.00019862672031893206, + "loss": 0.944, + "step": 4683 + }, + { + "epoch": 0.3250294913607661, + "grad_norm": 3.82789945602417, + "learning_rate": 0.00019862579220670173, + "loss": 1.2551, + "step": 4684 + }, + { + "epoch": 0.32509888279786275, + "grad_norm": 2.4826834201812744, + "learning_rate": 0.00019862486378312123, + "loss": 0.8308, + "step": 4685 + }, + { + "epoch": 0.3251682742349594, + "grad_norm": 4.457587718963623, + "learning_rate": 0.00019862393504819348, + "loss": 1.1614, + "step": 4686 + }, + { + "epoch": 0.3252376656720561, + "grad_norm": 3.8631644248962402, + "learning_rate": 0.0001986230060019214, + "loss": 1.0057, + "step": 4687 + }, + { + "epoch": 0.32530705710915275, + "grad_norm": 3.697157621383667, + "learning_rate": 0.00019862207664430792, + "loss": 1.2163, + "step": 4688 + }, + { + "epoch": 0.3253764485462494, + "grad_norm": 3.7761597633361816, + "learning_rate": 0.00019862114697535602, + "loss": 1.1465, + "step": 4689 + }, + { + "epoch": 0.32544583998334603, + "grad_norm": 2.488405227661133, + "learning_rate": 0.00019862021699506855, + "loss": 0.9876, + "step": 4690 + }, + { + "epoch": 0.3255152314204427, + "grad_norm": 3.0250039100646973, + "learning_rate": 0.00019861928670344856, + "loss": 0.8847, + "step": 4691 + }, + { + "epoch": 0.32558462285753936, + "grad_norm": 3.5899572372436523, + "learning_rate": 0.00019861835610049888, + "loss": 1.1654, + "step": 4692 + }, + { + "epoch": 0.32565401429463603, + "grad_norm": 3.144646167755127, + "learning_rate": 0.0001986174251862225, + "loss": 1.0532, + "step": 4693 + }, + { + "epoch": 0.3257234057317327, + "grad_norm": 3.2092583179473877, + "learning_rate": 0.00019861649396062236, + "loss": 0.9566, + "step": 4694 + }, + { + "epoch": 0.32579279716882936, + "grad_norm": 3.4772465229034424, + "learning_rate": 0.00019861556242370138, + "loss": 1.0794, + "step": 4695 + }, + { + "epoch": 0.32586218860592603, + "grad_norm": 3.4811606407165527, + "learning_rate": 0.0001986146305754625, + "loss": 0.9702, + "step": 4696 + }, + { + "epoch": 0.3259315800430227, + "grad_norm": 3.447772741317749, + "learning_rate": 0.0001986136984159087, + "loss": 0.9633, + "step": 4697 + }, + { + "epoch": 0.32600097148011936, + "grad_norm": 2.644841432571411, + "learning_rate": 0.0001986127659450429, + "loss": 0.8879, + "step": 4698 + }, + { + "epoch": 0.32607036291721603, + "grad_norm": 3.4800946712493896, + "learning_rate": 0.00019861183316286803, + "loss": 1.2443, + "step": 4699 + }, + { + "epoch": 0.3261397543543127, + "grad_norm": 3.223004102706909, + "learning_rate": 0.00019861090006938702, + "loss": 0.8434, + "step": 4700 + }, + { + "epoch": 0.32620914579140936, + "grad_norm": 3.286149024963379, + "learning_rate": 0.00019860996666460286, + "loss": 0.9165, + "step": 4701 + }, + { + "epoch": 0.326278537228506, + "grad_norm": 3.8360002040863037, + "learning_rate": 0.0001986090329485185, + "loss": 1.1029, + "step": 4702 + }, + { + "epoch": 0.32634792866560264, + "grad_norm": 3.9420924186706543, + "learning_rate": 0.00019860809892113682, + "loss": 1.3288, + "step": 4703 + }, + { + "epoch": 0.3264173201026993, + "grad_norm": 3.1279616355895996, + "learning_rate": 0.00019860716458246084, + "loss": 1.0748, + "step": 4704 + }, + { + "epoch": 0.32648671153979597, + "grad_norm": 2.9542016983032227, + "learning_rate": 0.00019860622993249344, + "loss": 0.8347, + "step": 4705 + }, + { + "epoch": 0.32655610297689264, + "grad_norm": 2.818114757537842, + "learning_rate": 0.00019860529497123765, + "loss": 0.9465, + "step": 4706 + }, + { + "epoch": 0.3266254944139893, + "grad_norm": 4.910421848297119, + "learning_rate": 0.0001986043596986964, + "loss": 1.4442, + "step": 4707 + }, + { + "epoch": 0.32669488585108597, + "grad_norm": 3.865408420562744, + "learning_rate": 0.00019860342411487258, + "loss": 1.1515, + "step": 4708 + }, + { + "epoch": 0.32676427728818264, + "grad_norm": 4.085899829864502, + "learning_rate": 0.00019860248821976923, + "loss": 1.1293, + "step": 4709 + }, + { + "epoch": 0.3268336687252793, + "grad_norm": 3.2859861850738525, + "learning_rate": 0.00019860155201338923, + "loss": 1.1121, + "step": 4710 + }, + { + "epoch": 0.32690306016237597, + "grad_norm": 3.8859870433807373, + "learning_rate": 0.0001986006154957356, + "loss": 1.0227, + "step": 4711 + }, + { + "epoch": 0.32697245159947264, + "grad_norm": 3.474365234375, + "learning_rate": 0.00019859967866681127, + "loss": 1.3381, + "step": 4712 + }, + { + "epoch": 0.3270418430365693, + "grad_norm": 3.2408225536346436, + "learning_rate": 0.00019859874152661916, + "loss": 1.0079, + "step": 4713 + }, + { + "epoch": 0.32711123447366597, + "grad_norm": 3.545104503631592, + "learning_rate": 0.00019859780407516228, + "loss": 1.1872, + "step": 4714 + }, + { + "epoch": 0.32718062591076263, + "grad_norm": 3.9090921878814697, + "learning_rate": 0.0001985968663124436, + "loss": 1.1898, + "step": 4715 + }, + { + "epoch": 0.3272500173478593, + "grad_norm": 3.4879202842712402, + "learning_rate": 0.000198595928238466, + "loss": 1.1639, + "step": 4716 + }, + { + "epoch": 0.3273194087849559, + "grad_norm": 3.1406357288360596, + "learning_rate": 0.00019859498985323252, + "loss": 1.0215, + "step": 4717 + }, + { + "epoch": 0.3273888002220526, + "grad_norm": 3.302733898162842, + "learning_rate": 0.00019859405115674612, + "loss": 0.9894, + "step": 4718 + }, + { + "epoch": 0.32745819165914924, + "grad_norm": 4.088558197021484, + "learning_rate": 0.0001985931121490097, + "loss": 1.1937, + "step": 4719 + }, + { + "epoch": 0.3275275830962459, + "grad_norm": 2.781628370285034, + "learning_rate": 0.00019859217283002626, + "loss": 1.0682, + "step": 4720 + }, + { + "epoch": 0.3275969745333426, + "grad_norm": 2.9078385829925537, + "learning_rate": 0.00019859123319979882, + "loss": 0.8817, + "step": 4721 + }, + { + "epoch": 0.32766636597043924, + "grad_norm": 2.9740700721740723, + "learning_rate": 0.00019859029325833025, + "loss": 1.0654, + "step": 4722 + }, + { + "epoch": 0.3277357574075359, + "grad_norm": 3.421959400177002, + "learning_rate": 0.00019858935300562359, + "loss": 0.8954, + "step": 4723 + }, + { + "epoch": 0.3278051488446326, + "grad_norm": 5.876124382019043, + "learning_rate": 0.00019858841244168175, + "loss": 1.1005, + "step": 4724 + }, + { + "epoch": 0.32787454028172924, + "grad_norm": 2.4204766750335693, + "learning_rate": 0.00019858747156650775, + "loss": 0.9236, + "step": 4725 + }, + { + "epoch": 0.3279439317188259, + "grad_norm": 3.0132858753204346, + "learning_rate": 0.00019858653038010455, + "loss": 1.2588, + "step": 4726 + }, + { + "epoch": 0.3280133231559226, + "grad_norm": 4.058139324188232, + "learning_rate": 0.00019858558888247513, + "loss": 0.9526, + "step": 4727 + }, + { + "epoch": 0.32808271459301924, + "grad_norm": 3.8118953704833984, + "learning_rate": 0.00019858464707362238, + "loss": 0.9711, + "step": 4728 + }, + { + "epoch": 0.3281521060301159, + "grad_norm": 3.010439872741699, + "learning_rate": 0.0001985837049535494, + "loss": 1.2128, + "step": 4729 + }, + { + "epoch": 0.3282214974672125, + "grad_norm": 3.644043445587158, + "learning_rate": 0.00019858276252225906, + "loss": 1.1577, + "step": 4730 + }, + { + "epoch": 0.3282908889043092, + "grad_norm": 2.601229190826416, + "learning_rate": 0.0001985818197797544, + "loss": 1.0267, + "step": 4731 + }, + { + "epoch": 0.32836028034140585, + "grad_norm": 3.2625961303710938, + "learning_rate": 0.00019858087672603833, + "loss": 1.1128, + "step": 4732 + }, + { + "epoch": 0.3284296717785025, + "grad_norm": 3.433253765106201, + "learning_rate": 0.00019857993336111393, + "loss": 1.0976, + "step": 4733 + }, + { + "epoch": 0.3284990632155992, + "grad_norm": 3.2048897743225098, + "learning_rate": 0.00019857898968498407, + "loss": 1.0996, + "step": 4734 + }, + { + "epoch": 0.32856845465269585, + "grad_norm": 2.6431102752685547, + "learning_rate": 0.0001985780456976518, + "loss": 0.8616, + "step": 4735 + }, + { + "epoch": 0.3286378460897925, + "grad_norm": 4.684629440307617, + "learning_rate": 0.0001985771013991201, + "loss": 1.1187, + "step": 4736 + }, + { + "epoch": 0.3287072375268892, + "grad_norm": 2.784736156463623, + "learning_rate": 0.00019857615678939189, + "loss": 0.9439, + "step": 4737 + }, + { + "epoch": 0.32877662896398585, + "grad_norm": 2.9092211723327637, + "learning_rate": 0.00019857521186847018, + "loss": 1.0033, + "step": 4738 + }, + { + "epoch": 0.3288460204010825, + "grad_norm": 4.728208065032959, + "learning_rate": 0.000198574266636358, + "loss": 1.0664, + "step": 4739 + }, + { + "epoch": 0.3289154118381792, + "grad_norm": 3.9074954986572266, + "learning_rate": 0.0001985733210930583, + "loss": 1.2115, + "step": 4740 + }, + { + "epoch": 0.32898480327527585, + "grad_norm": 3.406099796295166, + "learning_rate": 0.00019857237523857401, + "loss": 0.8625, + "step": 4741 + }, + { + "epoch": 0.3290541947123725, + "grad_norm": 3.396761417388916, + "learning_rate": 0.00019857142907290823, + "loss": 1.0422, + "step": 4742 + }, + { + "epoch": 0.3291235861494692, + "grad_norm": 2.8470301628112793, + "learning_rate": 0.00019857048259606382, + "loss": 0.9807, + "step": 4743 + }, + { + "epoch": 0.3291929775865658, + "grad_norm": 2.706334114074707, + "learning_rate": 0.0001985695358080439, + "loss": 0.9589, + "step": 4744 + }, + { + "epoch": 0.32926236902366246, + "grad_norm": 3.2794079780578613, + "learning_rate": 0.00019856858870885135, + "loss": 0.9578, + "step": 4745 + }, + { + "epoch": 0.3293317604607591, + "grad_norm": 2.8613054752349854, + "learning_rate": 0.00019856764129848922, + "loss": 1.0586, + "step": 4746 + }, + { + "epoch": 0.3294011518978558, + "grad_norm": 3.5982773303985596, + "learning_rate": 0.00019856669357696047, + "loss": 1.1762, + "step": 4747 + }, + { + "epoch": 0.32947054333495246, + "grad_norm": 3.610546350479126, + "learning_rate": 0.00019856574554426813, + "loss": 1.1846, + "step": 4748 + }, + { + "epoch": 0.3295399347720491, + "grad_norm": 3.4703283309936523, + "learning_rate": 0.00019856479720041512, + "loss": 1.1815, + "step": 4749 + }, + { + "epoch": 0.3296093262091458, + "grad_norm": 3.0467278957366943, + "learning_rate": 0.0001985638485454045, + "loss": 1.0329, + "step": 4750 + }, + { + "epoch": 0.32967871764624246, + "grad_norm": 2.910266637802124, + "learning_rate": 0.00019856289957923925, + "loss": 0.9881, + "step": 4751 + }, + { + "epoch": 0.3297481090833391, + "grad_norm": 3.0107967853546143, + "learning_rate": 0.0001985619503019224, + "loss": 0.9098, + "step": 4752 + }, + { + "epoch": 0.3298175005204358, + "grad_norm": 3.4561891555786133, + "learning_rate": 0.00019856100071345684, + "loss": 1.11, + "step": 4753 + }, + { + "epoch": 0.32988689195753246, + "grad_norm": 3.5796329975128174, + "learning_rate": 0.00019856005081384567, + "loss": 1.0125, + "step": 4754 + }, + { + "epoch": 0.3299562833946291, + "grad_norm": 2.796990156173706, + "learning_rate": 0.00019855910060309186, + "loss": 0.8081, + "step": 4755 + }, + { + "epoch": 0.3300256748317258, + "grad_norm": 2.9760866165161133, + "learning_rate": 0.0001985581500811984, + "loss": 0.9226, + "step": 4756 + }, + { + "epoch": 0.3300950662688224, + "grad_norm": 3.2398810386657715, + "learning_rate": 0.0001985571992481683, + "loss": 0.9595, + "step": 4757 + }, + { + "epoch": 0.33016445770591907, + "grad_norm": 3.4443624019622803, + "learning_rate": 0.00019855624810400456, + "loss": 1.0669, + "step": 4758 + }, + { + "epoch": 0.33023384914301573, + "grad_norm": 3.5521512031555176, + "learning_rate": 0.00019855529664871013, + "loss": 0.9771, + "step": 4759 + }, + { + "epoch": 0.3303032405801124, + "grad_norm": 2.929441452026367, + "learning_rate": 0.00019855434488228813, + "loss": 0.9314, + "step": 4760 + }, + { + "epoch": 0.33037263201720907, + "grad_norm": 2.819746971130371, + "learning_rate": 0.00019855339280474145, + "loss": 0.967, + "step": 4761 + }, + { + "epoch": 0.33044202345430573, + "grad_norm": 3.0972704887390137, + "learning_rate": 0.00019855244041607316, + "loss": 0.963, + "step": 4762 + }, + { + "epoch": 0.3305114148914024, + "grad_norm": 4.2714762687683105, + "learning_rate": 0.00019855148771628627, + "loss": 1.1301, + "step": 4763 + }, + { + "epoch": 0.33058080632849907, + "grad_norm": 2.596773147583008, + "learning_rate": 0.00019855053470538373, + "loss": 1.0799, + "step": 4764 + }, + { + "epoch": 0.33065019776559573, + "grad_norm": 3.562284469604492, + "learning_rate": 0.0001985495813833686, + "loss": 1.2519, + "step": 4765 + }, + { + "epoch": 0.3307195892026924, + "grad_norm": 3.605544328689575, + "learning_rate": 0.0001985486277502439, + "loss": 0.9876, + "step": 4766 + }, + { + "epoch": 0.33078898063978907, + "grad_norm": 3.3978707790374756, + "learning_rate": 0.00019854767380601257, + "loss": 0.9379, + "step": 4767 + }, + { + "epoch": 0.33085837207688573, + "grad_norm": 3.900113821029663, + "learning_rate": 0.00019854671955067771, + "loss": 1.0838, + "step": 4768 + }, + { + "epoch": 0.3309277635139824, + "grad_norm": 3.538590669631958, + "learning_rate": 0.0001985457649842423, + "loss": 1.0612, + "step": 4769 + }, + { + "epoch": 0.33099715495107906, + "grad_norm": 2.9986612796783447, + "learning_rate": 0.0001985448101067093, + "loss": 1.1243, + "step": 4770 + }, + { + "epoch": 0.3310665463881757, + "grad_norm": 4.044004440307617, + "learning_rate": 0.00019854385491808178, + "loss": 0.8017, + "step": 4771 + }, + { + "epoch": 0.33113593782527234, + "grad_norm": 2.846632719039917, + "learning_rate": 0.00019854289941836274, + "loss": 0.8263, + "step": 4772 + }, + { + "epoch": 0.331205329262369, + "grad_norm": 3.055474281311035, + "learning_rate": 0.0001985419436075552, + "loss": 1.0657, + "step": 4773 + }, + { + "epoch": 0.3312747206994657, + "grad_norm": 2.5601320266723633, + "learning_rate": 0.0001985409874856622, + "loss": 0.8057, + "step": 4774 + }, + { + "epoch": 0.33134411213656234, + "grad_norm": 2.3618404865264893, + "learning_rate": 0.00019854003105268671, + "loss": 0.9098, + "step": 4775 + }, + { + "epoch": 0.331413503573659, + "grad_norm": 3.3041770458221436, + "learning_rate": 0.0001985390743086318, + "loss": 1.0103, + "step": 4776 + }, + { + "epoch": 0.3314828950107557, + "grad_norm": 3.7057595252990723, + "learning_rate": 0.00019853811725350045, + "loss": 1.0226, + "step": 4777 + }, + { + "epoch": 0.33155228644785234, + "grad_norm": 2.9132401943206787, + "learning_rate": 0.0001985371598872957, + "loss": 0.8994, + "step": 4778 + }, + { + "epoch": 0.331621677884949, + "grad_norm": 3.36588191986084, + "learning_rate": 0.00019853620221002056, + "loss": 1.1313, + "step": 4779 + }, + { + "epoch": 0.3316910693220457, + "grad_norm": 3.233579397201538, + "learning_rate": 0.00019853524422167807, + "loss": 1.1352, + "step": 4780 + }, + { + "epoch": 0.33176046075914234, + "grad_norm": 3.982266664505005, + "learning_rate": 0.00019853428592227123, + "loss": 0.9997, + "step": 4781 + }, + { + "epoch": 0.331829852196239, + "grad_norm": 2.8598501682281494, + "learning_rate": 0.00019853332731180313, + "loss": 0.8515, + "step": 4782 + }, + { + "epoch": 0.33189924363333567, + "grad_norm": 3.8859786987304688, + "learning_rate": 0.0001985323683902767, + "loss": 1.0689, + "step": 4783 + }, + { + "epoch": 0.3319686350704323, + "grad_norm": 3.1344313621520996, + "learning_rate": 0.000198531409157695, + "loss": 0.9851, + "step": 4784 + }, + { + "epoch": 0.33203802650752895, + "grad_norm": 2.6864707469940186, + "learning_rate": 0.0001985304496140611, + "loss": 0.9893, + "step": 4785 + }, + { + "epoch": 0.3321074179446256, + "grad_norm": 3.3577303886413574, + "learning_rate": 0.00019852948975937801, + "loss": 1.2338, + "step": 4786 + }, + { + "epoch": 0.3321768093817223, + "grad_norm": 2.6908600330352783, + "learning_rate": 0.00019852852959364875, + "loss": 0.9289, + "step": 4787 + }, + { + "epoch": 0.33224620081881895, + "grad_norm": 2.871896743774414, + "learning_rate": 0.00019852756911687637, + "loss": 1.1051, + "step": 4788 + }, + { + "epoch": 0.3323155922559156, + "grad_norm": 3.394713878631592, + "learning_rate": 0.00019852660832906386, + "loss": 1.3289, + "step": 4789 + }, + { + "epoch": 0.3323849836930123, + "grad_norm": 3.293565511703491, + "learning_rate": 0.00019852564723021426, + "loss": 0.965, + "step": 4790 + }, + { + "epoch": 0.33245437513010895, + "grad_norm": 3.218824625015259, + "learning_rate": 0.00019852468582033065, + "loss": 0.9576, + "step": 4791 + }, + { + "epoch": 0.3325237665672056, + "grad_norm": 3.0724663734436035, + "learning_rate": 0.00019852372409941604, + "loss": 0.758, + "step": 4792 + }, + { + "epoch": 0.3325931580043023, + "grad_norm": 3.4234304428100586, + "learning_rate": 0.00019852276206747344, + "loss": 1.2214, + "step": 4793 + }, + { + "epoch": 0.33266254944139895, + "grad_norm": 2.9241745471954346, + "learning_rate": 0.00019852179972450594, + "loss": 1.0169, + "step": 4794 + }, + { + "epoch": 0.3327319408784956, + "grad_norm": 5.598458290100098, + "learning_rate": 0.00019852083707051654, + "loss": 0.9034, + "step": 4795 + }, + { + "epoch": 0.3328013323155923, + "grad_norm": 5.664482116699219, + "learning_rate": 0.00019851987410550827, + "loss": 1.0156, + "step": 4796 + }, + { + "epoch": 0.3328707237526889, + "grad_norm": 3.338611364364624, + "learning_rate": 0.00019851891082948422, + "loss": 0.9695, + "step": 4797 + }, + { + "epoch": 0.33294011518978556, + "grad_norm": 4.286836624145508, + "learning_rate": 0.0001985179472424474, + "loss": 1.0179, + "step": 4798 + }, + { + "epoch": 0.3330095066268822, + "grad_norm": 3.652686357498169, + "learning_rate": 0.00019851698334440082, + "loss": 1.1519, + "step": 4799 + }, + { + "epoch": 0.3330788980639789, + "grad_norm": 2.8832082748413086, + "learning_rate": 0.00019851601913534753, + "loss": 1.0885, + "step": 4800 + }, + { + "epoch": 0.33314828950107556, + "grad_norm": 3.66641902923584, + "learning_rate": 0.00019851505461529066, + "loss": 1.1222, + "step": 4801 + }, + { + "epoch": 0.3332176809381722, + "grad_norm": 4.8794989585876465, + "learning_rate": 0.00019851408978423317, + "loss": 1.0786, + "step": 4802 + }, + { + "epoch": 0.3332870723752689, + "grad_norm": 3.5061278343200684, + "learning_rate": 0.00019851312464217813, + "loss": 1.0255, + "step": 4803 + }, + { + "epoch": 0.33335646381236556, + "grad_norm": 2.4682137966156006, + "learning_rate": 0.00019851215918912857, + "loss": 1.0357, + "step": 4804 + }, + { + "epoch": 0.3334258552494622, + "grad_norm": 2.685072422027588, + "learning_rate": 0.00019851119342508756, + "loss": 0.7471, + "step": 4805 + }, + { + "epoch": 0.3334952466865589, + "grad_norm": 3.998924493789673, + "learning_rate": 0.00019851022735005813, + "loss": 1.2256, + "step": 4806 + }, + { + "epoch": 0.33356463812365555, + "grad_norm": 2.8016867637634277, + "learning_rate": 0.00019850926096404338, + "loss": 1.1316, + "step": 4807 + }, + { + "epoch": 0.3336340295607522, + "grad_norm": 3.7786810398101807, + "learning_rate": 0.0001985082942670463, + "loss": 0.9956, + "step": 4808 + }, + { + "epoch": 0.3337034209978489, + "grad_norm": 3.488844394683838, + "learning_rate": 0.00019850732725906994, + "loss": 0.8597, + "step": 4809 + }, + { + "epoch": 0.33377281243494555, + "grad_norm": 3.8123083114624023, + "learning_rate": 0.0001985063599401174, + "loss": 1.1341, + "step": 4810 + }, + { + "epoch": 0.33384220387204216, + "grad_norm": 2.8914413452148438, + "learning_rate": 0.0001985053923101917, + "loss": 0.9081, + "step": 4811 + }, + { + "epoch": 0.33391159530913883, + "grad_norm": 3.6090452671051025, + "learning_rate": 0.00019850442436929595, + "loss": 0.9897, + "step": 4812 + }, + { + "epoch": 0.3339809867462355, + "grad_norm": 3.5724220275878906, + "learning_rate": 0.00019850345611743312, + "loss": 1.1523, + "step": 4813 + }, + { + "epoch": 0.33405037818333216, + "grad_norm": 3.359771251678467, + "learning_rate": 0.00019850248755460628, + "loss": 1.1549, + "step": 4814 + }, + { + "epoch": 0.33411976962042883, + "grad_norm": 3.0480287075042725, + "learning_rate": 0.00019850151868081855, + "loss": 0.9159, + "step": 4815 + }, + { + "epoch": 0.3341891610575255, + "grad_norm": 2.941770076751709, + "learning_rate": 0.00019850054949607297, + "loss": 0.8492, + "step": 4816 + }, + { + "epoch": 0.33425855249462216, + "grad_norm": 2.6036570072174072, + "learning_rate": 0.00019849958000037256, + "loss": 0.8972, + "step": 4817 + }, + { + "epoch": 0.33432794393171883, + "grad_norm": 3.120499610900879, + "learning_rate": 0.00019849861019372042, + "loss": 0.9853, + "step": 4818 + }, + { + "epoch": 0.3343973353688155, + "grad_norm": 2.810248374938965, + "learning_rate": 0.00019849764007611958, + "loss": 0.8758, + "step": 4819 + }, + { + "epoch": 0.33446672680591216, + "grad_norm": 2.9185469150543213, + "learning_rate": 0.00019849666964757314, + "loss": 0.9465, + "step": 4820 + }, + { + "epoch": 0.33453611824300883, + "grad_norm": 3.313058853149414, + "learning_rate": 0.00019849569890808413, + "loss": 1.0439, + "step": 4821 + }, + { + "epoch": 0.3346055096801055, + "grad_norm": 2.361970901489258, + "learning_rate": 0.00019849472785765562, + "loss": 0.8303, + "step": 4822 + }, + { + "epoch": 0.33467490111720216, + "grad_norm": 2.6233818531036377, + "learning_rate": 0.00019849375649629068, + "loss": 0.9392, + "step": 4823 + }, + { + "epoch": 0.3347442925542988, + "grad_norm": 2.4938700199127197, + "learning_rate": 0.00019849278482399237, + "loss": 0.7693, + "step": 4824 + }, + { + "epoch": 0.33481368399139544, + "grad_norm": 2.801367998123169, + "learning_rate": 0.0001984918128407638, + "loss": 0.8888, + "step": 4825 + }, + { + "epoch": 0.3348830754284921, + "grad_norm": 4.05902624130249, + "learning_rate": 0.00019849084054660796, + "loss": 1.032, + "step": 4826 + }, + { + "epoch": 0.33495246686558877, + "grad_norm": 3.3209211826324463, + "learning_rate": 0.000198489867941528, + "loss": 0.9745, + "step": 4827 + }, + { + "epoch": 0.33502185830268544, + "grad_norm": 2.848332405090332, + "learning_rate": 0.00019848889502552692, + "loss": 0.9536, + "step": 4828 + }, + { + "epoch": 0.3350912497397821, + "grad_norm": 4.643240451812744, + "learning_rate": 0.00019848792179860783, + "loss": 1.2147, + "step": 4829 + }, + { + "epoch": 0.33516064117687877, + "grad_norm": 3.0195209980010986, + "learning_rate": 0.0001984869482607738, + "loss": 1.0535, + "step": 4830 + }, + { + "epoch": 0.33523003261397544, + "grad_norm": 3.112046003341675, + "learning_rate": 0.00019848597441202793, + "loss": 1.0884, + "step": 4831 + }, + { + "epoch": 0.3352994240510721, + "grad_norm": 3.302532196044922, + "learning_rate": 0.00019848500025237327, + "loss": 1.2399, + "step": 4832 + }, + { + "epoch": 0.33536881548816877, + "grad_norm": 2.7703850269317627, + "learning_rate": 0.00019848402578181285, + "loss": 0.8904, + "step": 4833 + }, + { + "epoch": 0.33543820692526544, + "grad_norm": 3.002723217010498, + "learning_rate": 0.0001984830510003498, + "loss": 0.7807, + "step": 4834 + }, + { + "epoch": 0.3355075983623621, + "grad_norm": 3.1276402473449707, + "learning_rate": 0.00019848207590798719, + "loss": 1.0842, + "step": 4835 + }, + { + "epoch": 0.33557698979945877, + "grad_norm": 3.2090795040130615, + "learning_rate": 0.0001984811005047281, + "loss": 1.0557, + "step": 4836 + }, + { + "epoch": 0.33564638123655544, + "grad_norm": 2.551647663116455, + "learning_rate": 0.0001984801247905756, + "loss": 0.9321, + "step": 4837 + }, + { + "epoch": 0.33571577267365205, + "grad_norm": 2.339859962463379, + "learning_rate": 0.00019847914876553276, + "loss": 0.7556, + "step": 4838 + }, + { + "epoch": 0.3357851641107487, + "grad_norm": 2.8377668857574463, + "learning_rate": 0.00019847817242960267, + "loss": 0.9674, + "step": 4839 + }, + { + "epoch": 0.3358545555478454, + "grad_norm": 4.277557373046875, + "learning_rate": 0.0001984771957827884, + "loss": 0.9757, + "step": 4840 + }, + { + "epoch": 0.33592394698494205, + "grad_norm": 2.570902109146118, + "learning_rate": 0.00019847621882509308, + "loss": 0.7381, + "step": 4841 + }, + { + "epoch": 0.3359933384220387, + "grad_norm": 3.2462778091430664, + "learning_rate": 0.00019847524155651978, + "loss": 1.2035, + "step": 4842 + }, + { + "epoch": 0.3360627298591354, + "grad_norm": 3.2714672088623047, + "learning_rate": 0.00019847426397707154, + "loss": 0.9252, + "step": 4843 + }, + { + "epoch": 0.33613212129623204, + "grad_norm": 2.8301429748535156, + "learning_rate": 0.00019847328608675147, + "loss": 0.929, + "step": 4844 + }, + { + "epoch": 0.3362015127333287, + "grad_norm": 4.0111284255981445, + "learning_rate": 0.00019847230788556266, + "loss": 0.9418, + "step": 4845 + }, + { + "epoch": 0.3362709041704254, + "grad_norm": 4.647679328918457, + "learning_rate": 0.00019847132937350818, + "loss": 1.1108, + "step": 4846 + }, + { + "epoch": 0.33634029560752204, + "grad_norm": 3.2207255363464355, + "learning_rate": 0.00019847035055059117, + "loss": 0.8091, + "step": 4847 + }, + { + "epoch": 0.3364096870446187, + "grad_norm": 3.6302478313446045, + "learning_rate": 0.0001984693714168147, + "loss": 0.9883, + "step": 4848 + }, + { + "epoch": 0.3364790784817154, + "grad_norm": 3.373408079147339, + "learning_rate": 0.00019846839197218183, + "loss": 1.0825, + "step": 4849 + }, + { + "epoch": 0.33654846991881204, + "grad_norm": 3.1939432621002197, + "learning_rate": 0.00019846741221669567, + "loss": 0.9378, + "step": 4850 + }, + { + "epoch": 0.33661786135590865, + "grad_norm": 3.8010952472686768, + "learning_rate": 0.00019846643215035931, + "loss": 1.1299, + "step": 4851 + }, + { + "epoch": 0.3366872527930053, + "grad_norm": 2.9832489490509033, + "learning_rate": 0.00019846545177317583, + "loss": 0.798, + "step": 4852 + }, + { + "epoch": 0.336756644230102, + "grad_norm": 2.847482204437256, + "learning_rate": 0.00019846447108514838, + "loss": 0.8957, + "step": 4853 + }, + { + "epoch": 0.33682603566719865, + "grad_norm": 5.462429046630859, + "learning_rate": 0.00019846349008628003, + "loss": 0.92, + "step": 4854 + }, + { + "epoch": 0.3368954271042953, + "grad_norm": 2.89941668510437, + "learning_rate": 0.00019846250877657382, + "loss": 1.0407, + "step": 4855 + }, + { + "epoch": 0.336964818541392, + "grad_norm": 3.5554051399230957, + "learning_rate": 0.00019846152715603292, + "loss": 0.7789, + "step": 4856 + }, + { + "epoch": 0.33703420997848865, + "grad_norm": 3.3701131343841553, + "learning_rate": 0.0001984605452246604, + "loss": 0.9946, + "step": 4857 + }, + { + "epoch": 0.3371036014155853, + "grad_norm": 3.552485227584839, + "learning_rate": 0.00019845956298245935, + "loss": 0.9469, + "step": 4858 + }, + { + "epoch": 0.337172992852682, + "grad_norm": 4.076583385467529, + "learning_rate": 0.0001984585804294329, + "loss": 1.0587, + "step": 4859 + }, + { + "epoch": 0.33724238428977865, + "grad_norm": 3.406876564025879, + "learning_rate": 0.00019845759756558412, + "loss": 1.0473, + "step": 4860 + }, + { + "epoch": 0.3373117757268753, + "grad_norm": 3.2475950717926025, + "learning_rate": 0.0001984566143909161, + "loss": 0.9567, + "step": 4861 + }, + { + "epoch": 0.337381167163972, + "grad_norm": 3.1912989616394043, + "learning_rate": 0.00019845563090543203, + "loss": 1.2632, + "step": 4862 + }, + { + "epoch": 0.33745055860106865, + "grad_norm": 3.6816494464874268, + "learning_rate": 0.00019845464710913492, + "loss": 0.8346, + "step": 4863 + }, + { + "epoch": 0.3375199500381653, + "grad_norm": 2.664206027984619, + "learning_rate": 0.00019845366300202795, + "loss": 0.9213, + "step": 4864 + }, + { + "epoch": 0.33758934147526193, + "grad_norm": 2.546757698059082, + "learning_rate": 0.00019845267858411414, + "loss": 0.8201, + "step": 4865 + }, + { + "epoch": 0.3376587329123586, + "grad_norm": 5.463809490203857, + "learning_rate": 0.00019845169385539668, + "loss": 1.4636, + "step": 4866 + }, + { + "epoch": 0.33772812434945526, + "grad_norm": 3.0140140056610107, + "learning_rate": 0.00019845070881587863, + "loss": 0.8932, + "step": 4867 + }, + { + "epoch": 0.3377975157865519, + "grad_norm": 3.210500717163086, + "learning_rate": 0.00019844972346556313, + "loss": 1.1245, + "step": 4868 + }, + { + "epoch": 0.3378669072236486, + "grad_norm": 3.0404069423675537, + "learning_rate": 0.00019844873780445324, + "loss": 0.8305, + "step": 4869 + }, + { + "epoch": 0.33793629866074526, + "grad_norm": 2.356090545654297, + "learning_rate": 0.00019844775183255214, + "loss": 0.8052, + "step": 4870 + }, + { + "epoch": 0.3380056900978419, + "grad_norm": 3.574016571044922, + "learning_rate": 0.0001984467655498629, + "loss": 1.0307, + "step": 4871 + }, + { + "epoch": 0.3380750815349386, + "grad_norm": 2.8512723445892334, + "learning_rate": 0.00019844577895638862, + "loss": 0.688, + "step": 4872 + }, + { + "epoch": 0.33814447297203526, + "grad_norm": 5.098503589630127, + "learning_rate": 0.0001984447920521325, + "loss": 1.55, + "step": 4873 + }, + { + "epoch": 0.3382138644091319, + "grad_norm": 6.197853088378906, + "learning_rate": 0.00019844380483709752, + "loss": 1.4214, + "step": 4874 + }, + { + "epoch": 0.3382832558462286, + "grad_norm": 4.066959857940674, + "learning_rate": 0.00019844281731128694, + "loss": 1.1409, + "step": 4875 + }, + { + "epoch": 0.33835264728332526, + "grad_norm": 3.138442277908325, + "learning_rate": 0.00019844182947470376, + "loss": 1.1374, + "step": 4876 + }, + { + "epoch": 0.3384220387204219, + "grad_norm": 3.698665142059326, + "learning_rate": 0.00019844084132735113, + "loss": 1.0138, + "step": 4877 + }, + { + "epoch": 0.33849143015751854, + "grad_norm": 2.330488920211792, + "learning_rate": 0.0001984398528692322, + "loss": 0.6726, + "step": 4878 + }, + { + "epoch": 0.3385608215946152, + "grad_norm": 4.9226155281066895, + "learning_rate": 0.0001984388641003501, + "loss": 1.055, + "step": 4879 + }, + { + "epoch": 0.33863021303171187, + "grad_norm": 2.9155073165893555, + "learning_rate": 0.00019843787502070792, + "loss": 1.2196, + "step": 4880 + }, + { + "epoch": 0.33869960446880853, + "grad_norm": 4.332104682922363, + "learning_rate": 0.0001984368856303088, + "loss": 1.1549, + "step": 4881 + }, + { + "epoch": 0.3387689959059052, + "grad_norm": 2.7951602935791016, + "learning_rate": 0.00019843589592915582, + "loss": 1.0286, + "step": 4882 + }, + { + "epoch": 0.33883838734300187, + "grad_norm": 3.17106032371521, + "learning_rate": 0.00019843490591725215, + "loss": 1.303, + "step": 4883 + }, + { + "epoch": 0.33890777878009853, + "grad_norm": 2.448443651199341, + "learning_rate": 0.00019843391559460093, + "loss": 0.9932, + "step": 4884 + }, + { + "epoch": 0.3389771702171952, + "grad_norm": 3.169863224029541, + "learning_rate": 0.00019843292496120522, + "loss": 0.9069, + "step": 4885 + }, + { + "epoch": 0.33904656165429187, + "grad_norm": 3.0078344345092773, + "learning_rate": 0.00019843193401706822, + "loss": 0.9314, + "step": 4886 + }, + { + "epoch": 0.33911595309138853, + "grad_norm": 3.3574435710906982, + "learning_rate": 0.000198430942762193, + "loss": 1.0671, + "step": 4887 + }, + { + "epoch": 0.3391853445284852, + "grad_norm": 3.002405881881714, + "learning_rate": 0.00019842995119658273, + "loss": 0.8752, + "step": 4888 + }, + { + "epoch": 0.33925473596558187, + "grad_norm": 3.034560203552246, + "learning_rate": 0.0001984289593202405, + "loss": 1.1918, + "step": 4889 + }, + { + "epoch": 0.33932412740267853, + "grad_norm": 2.6012051105499268, + "learning_rate": 0.00019842796713316946, + "loss": 0.9931, + "step": 4890 + }, + { + "epoch": 0.3393935188397752, + "grad_norm": 3.0142276287078857, + "learning_rate": 0.0001984269746353728, + "loss": 0.9077, + "step": 4891 + }, + { + "epoch": 0.3394629102768718, + "grad_norm": 3.643373966217041, + "learning_rate": 0.00019842598182685355, + "loss": 1.2735, + "step": 4892 + }, + { + "epoch": 0.3395323017139685, + "grad_norm": 3.7215981483459473, + "learning_rate": 0.00019842498870761493, + "loss": 1.246, + "step": 4893 + }, + { + "epoch": 0.33960169315106514, + "grad_norm": 3.5096333026885986, + "learning_rate": 0.00019842399527766, + "loss": 1.3581, + "step": 4894 + }, + { + "epoch": 0.3396710845881618, + "grad_norm": 2.617743492126465, + "learning_rate": 0.00019842300153699195, + "loss": 0.9787, + "step": 4895 + }, + { + "epoch": 0.3397404760252585, + "grad_norm": 3.339421033859253, + "learning_rate": 0.00019842200748561393, + "loss": 1.0821, + "step": 4896 + }, + { + "epoch": 0.33980986746235514, + "grad_norm": 3.0516114234924316, + "learning_rate": 0.00019842101312352904, + "loss": 1.0581, + "step": 4897 + }, + { + "epoch": 0.3398792588994518, + "grad_norm": 3.925884485244751, + "learning_rate": 0.0001984200184507404, + "loss": 0.9727, + "step": 4898 + }, + { + "epoch": 0.3399486503365485, + "grad_norm": 3.1953494548797607, + "learning_rate": 0.0001984190234672512, + "loss": 1.0747, + "step": 4899 + }, + { + "epoch": 0.34001804177364514, + "grad_norm": 3.302252769470215, + "learning_rate": 0.00019841802817306459, + "loss": 1.1286, + "step": 4900 + }, + { + "epoch": 0.3400874332107418, + "grad_norm": 3.152082681655884, + "learning_rate": 0.00019841703256818366, + "loss": 0.9903, + "step": 4901 + }, + { + "epoch": 0.3401568246478385, + "grad_norm": 4.428280353546143, + "learning_rate": 0.00019841603665261156, + "loss": 0.9709, + "step": 4902 + }, + { + "epoch": 0.34022621608493514, + "grad_norm": 3.7591845989227295, + "learning_rate": 0.0001984150404263515, + "loss": 0.9075, + "step": 4903 + }, + { + "epoch": 0.3402956075220318, + "grad_norm": 2.9109389781951904, + "learning_rate": 0.00019841404388940651, + "loss": 0.9512, + "step": 4904 + }, + { + "epoch": 0.3403649989591284, + "grad_norm": 2.689126968383789, + "learning_rate": 0.00019841304704177985, + "loss": 1.0548, + "step": 4905 + }, + { + "epoch": 0.3404343903962251, + "grad_norm": 2.777379035949707, + "learning_rate": 0.0001984120498834746, + "loss": 0.6841, + "step": 4906 + }, + { + "epoch": 0.34050378183332175, + "grad_norm": 3.070087432861328, + "learning_rate": 0.00019841105241449392, + "loss": 1.0506, + "step": 4907 + }, + { + "epoch": 0.3405731732704184, + "grad_norm": 3.713163375854492, + "learning_rate": 0.000198410054634841, + "loss": 1.0979, + "step": 4908 + }, + { + "epoch": 0.3406425647075151, + "grad_norm": 3.791754961013794, + "learning_rate": 0.0001984090565445189, + "loss": 0.7712, + "step": 4909 + }, + { + "epoch": 0.34071195614461175, + "grad_norm": 3.002696990966797, + "learning_rate": 0.00019840805814353086, + "loss": 0.8388, + "step": 4910 + }, + { + "epoch": 0.3407813475817084, + "grad_norm": 2.6177330017089844, + "learning_rate": 0.00019840705943188002, + "loss": 0.9649, + "step": 4911 + }, + { + "epoch": 0.3408507390188051, + "grad_norm": 2.7587225437164307, + "learning_rate": 0.00019840606040956945, + "loss": 1.1091, + "step": 4912 + }, + { + "epoch": 0.34092013045590175, + "grad_norm": 2.4809834957122803, + "learning_rate": 0.0001984050610766024, + "loss": 0.9961, + "step": 4913 + }, + { + "epoch": 0.3409895218929984, + "grad_norm": 2.9721930027008057, + "learning_rate": 0.000198404061432982, + "loss": 0.9472, + "step": 4914 + }, + { + "epoch": 0.3410589133300951, + "grad_norm": 2.8384759426116943, + "learning_rate": 0.00019840306147871138, + "loss": 0.9661, + "step": 4915 + }, + { + "epoch": 0.34112830476719175, + "grad_norm": 3.5225112438201904, + "learning_rate": 0.0001984020612137937, + "loss": 1.0849, + "step": 4916 + }, + { + "epoch": 0.3411976962042884, + "grad_norm": 3.9205806255340576, + "learning_rate": 0.00019840106063823216, + "loss": 1.2274, + "step": 4917 + }, + { + "epoch": 0.3412670876413851, + "grad_norm": 2.9845194816589355, + "learning_rate": 0.00019840005975202987, + "loss": 1.1879, + "step": 4918 + }, + { + "epoch": 0.3413364790784817, + "grad_norm": 4.325295448303223, + "learning_rate": 0.00019839905855519, + "loss": 1.127, + "step": 4919 + }, + { + "epoch": 0.34140587051557836, + "grad_norm": 2.5139575004577637, + "learning_rate": 0.00019839805704771575, + "loss": 1.1176, + "step": 4920 + }, + { + "epoch": 0.341475261952675, + "grad_norm": 3.876614570617676, + "learning_rate": 0.0001983970552296102, + "loss": 1.4287, + "step": 4921 + }, + { + "epoch": 0.3415446533897717, + "grad_norm": 2.7868459224700928, + "learning_rate": 0.0001983960531008766, + "loss": 1.0242, + "step": 4922 + }, + { + "epoch": 0.34161404482686836, + "grad_norm": 2.779245376586914, + "learning_rate": 0.00019839505066151806, + "loss": 1.0616, + "step": 4923 + }, + { + "epoch": 0.341683436263965, + "grad_norm": 5.570849895477295, + "learning_rate": 0.00019839404791153779, + "loss": 1.1111, + "step": 4924 + }, + { + "epoch": 0.3417528277010617, + "grad_norm": 4.6085076332092285, + "learning_rate": 0.0001983930448509389, + "loss": 1.2175, + "step": 4925 + }, + { + "epoch": 0.34182221913815836, + "grad_norm": 5.614877223968506, + "learning_rate": 0.00019839204147972458, + "loss": 1.0669, + "step": 4926 + }, + { + "epoch": 0.341891610575255, + "grad_norm": 2.746032476425171, + "learning_rate": 0.00019839103779789802, + "loss": 0.9399, + "step": 4927 + }, + { + "epoch": 0.3419610020123517, + "grad_norm": 4.140196800231934, + "learning_rate": 0.00019839003380546237, + "loss": 0.9185, + "step": 4928 + }, + { + "epoch": 0.34203039344944836, + "grad_norm": 3.067411184310913, + "learning_rate": 0.00019838902950242078, + "loss": 1.1414, + "step": 4929 + }, + { + "epoch": 0.342099784886545, + "grad_norm": 3.503218173980713, + "learning_rate": 0.00019838802488877642, + "loss": 0.9146, + "step": 4930 + }, + { + "epoch": 0.3421691763236417, + "grad_norm": 3.3210463523864746, + "learning_rate": 0.00019838701996453253, + "loss": 1.0088, + "step": 4931 + }, + { + "epoch": 0.3422385677607383, + "grad_norm": 4.363494873046875, + "learning_rate": 0.00019838601472969223, + "loss": 0.9274, + "step": 4932 + }, + { + "epoch": 0.34230795919783497, + "grad_norm": 2.4851717948913574, + "learning_rate": 0.00019838500918425867, + "loss": 0.9449, + "step": 4933 + }, + { + "epoch": 0.34237735063493163, + "grad_norm": 3.2444753646850586, + "learning_rate": 0.00019838400332823506, + "loss": 1.1259, + "step": 4934 + }, + { + "epoch": 0.3424467420720283, + "grad_norm": 5.0784382820129395, + "learning_rate": 0.00019838299716162455, + "loss": 0.7689, + "step": 4935 + }, + { + "epoch": 0.34251613350912496, + "grad_norm": 2.436328172683716, + "learning_rate": 0.00019838199068443035, + "loss": 0.5367, + "step": 4936 + }, + { + "epoch": 0.34258552494622163, + "grad_norm": 3.0083632469177246, + "learning_rate": 0.00019838098389665562, + "loss": 0.8637, + "step": 4937 + }, + { + "epoch": 0.3426549163833183, + "grad_norm": 3.2278733253479004, + "learning_rate": 0.00019837997679830353, + "loss": 1.0491, + "step": 4938 + }, + { + "epoch": 0.34272430782041496, + "grad_norm": 3.3602399826049805, + "learning_rate": 0.00019837896938937726, + "loss": 0.8512, + "step": 4939 + }, + { + "epoch": 0.34279369925751163, + "grad_norm": 3.664677143096924, + "learning_rate": 0.00019837796166988003, + "loss": 0.864, + "step": 4940 + }, + { + "epoch": 0.3428630906946083, + "grad_norm": 4.067848205566406, + "learning_rate": 0.00019837695363981496, + "loss": 0.6688, + "step": 4941 + }, + { + "epoch": 0.34293248213170496, + "grad_norm": 3.428861618041992, + "learning_rate": 0.00019837594529918527, + "loss": 1.1152, + "step": 4942 + }, + { + "epoch": 0.34300187356880163, + "grad_norm": 3.3559157848358154, + "learning_rate": 0.00019837493664799413, + "loss": 1.0485, + "step": 4943 + }, + { + "epoch": 0.3430712650058983, + "grad_norm": 3.3797988891601562, + "learning_rate": 0.00019837392768624472, + "loss": 1.2025, + "step": 4944 + }, + { + "epoch": 0.3431406564429949, + "grad_norm": 3.6614584922790527, + "learning_rate": 0.00019837291841394027, + "loss": 1.3162, + "step": 4945 + }, + { + "epoch": 0.3432100478800916, + "grad_norm": 3.7896459102630615, + "learning_rate": 0.0001983719088310839, + "loss": 0.9862, + "step": 4946 + }, + { + "epoch": 0.34327943931718824, + "grad_norm": 2.828364133834839, + "learning_rate": 0.0001983708989376788, + "loss": 1.0202, + "step": 4947 + }, + { + "epoch": 0.3433488307542849, + "grad_norm": 2.93410062789917, + "learning_rate": 0.00019836988873372824, + "loss": 1.2132, + "step": 4948 + }, + { + "epoch": 0.3434182221913816, + "grad_norm": 3.3041574954986572, + "learning_rate": 0.00019836887821923535, + "loss": 0.7319, + "step": 4949 + }, + { + "epoch": 0.34348761362847824, + "grad_norm": 2.9977524280548096, + "learning_rate": 0.0001983678673942033, + "loss": 1.0738, + "step": 4950 + }, + { + "epoch": 0.3435570050655749, + "grad_norm": 3.3891212940216064, + "learning_rate": 0.00019836685625863532, + "loss": 1.0688, + "step": 4951 + }, + { + "epoch": 0.34362639650267157, + "grad_norm": 3.2332913875579834, + "learning_rate": 0.00019836584481253456, + "loss": 1.1617, + "step": 4952 + }, + { + "epoch": 0.34369578793976824, + "grad_norm": 3.6917078495025635, + "learning_rate": 0.00019836483305590426, + "loss": 1.0637, + "step": 4953 + }, + { + "epoch": 0.3437651793768649, + "grad_norm": 4.184532165527344, + "learning_rate": 0.0001983638209887476, + "loss": 1.3882, + "step": 4954 + }, + { + "epoch": 0.34383457081396157, + "grad_norm": 3.2067525386810303, + "learning_rate": 0.00019836280861106778, + "loss": 0.9328, + "step": 4955 + }, + { + "epoch": 0.34390396225105824, + "grad_norm": 3.4144392013549805, + "learning_rate": 0.00019836179592286793, + "loss": 1.1711, + "step": 4956 + }, + { + "epoch": 0.3439733536881549, + "grad_norm": 2.545738458633423, + "learning_rate": 0.00019836078292415136, + "loss": 0.8282, + "step": 4957 + }, + { + "epoch": 0.34404274512525157, + "grad_norm": 2.774261474609375, + "learning_rate": 0.0001983597696149212, + "loss": 0.9411, + "step": 4958 + }, + { + "epoch": 0.3441121365623482, + "grad_norm": 2.719292402267456, + "learning_rate": 0.00019835875599518067, + "loss": 0.9631, + "step": 4959 + }, + { + "epoch": 0.34418152799944485, + "grad_norm": 2.852414608001709, + "learning_rate": 0.0001983577420649329, + "loss": 0.8738, + "step": 4960 + }, + { + "epoch": 0.3442509194365415, + "grad_norm": 2.714102029800415, + "learning_rate": 0.00019835672782418122, + "loss": 0.7496, + "step": 4961 + }, + { + "epoch": 0.3443203108736382, + "grad_norm": 2.8831307888031006, + "learning_rate": 0.00019835571327292872, + "loss": 0.9194, + "step": 4962 + }, + { + "epoch": 0.34438970231073485, + "grad_norm": 2.8378243446350098, + "learning_rate": 0.00019835469841117867, + "loss": 0.8222, + "step": 4963 + }, + { + "epoch": 0.3444590937478315, + "grad_norm": 3.5769131183624268, + "learning_rate": 0.00019835368323893425, + "loss": 0.8364, + "step": 4964 + }, + { + "epoch": 0.3445284851849282, + "grad_norm": 2.945960760116577, + "learning_rate": 0.00019835266775619862, + "loss": 0.7021, + "step": 4965 + }, + { + "epoch": 0.34459787662202485, + "grad_norm": 3.368274450302124, + "learning_rate": 0.00019835165196297507, + "loss": 0.8967, + "step": 4966 + }, + { + "epoch": 0.3446672680591215, + "grad_norm": 4.220859050750732, + "learning_rate": 0.00019835063585926674, + "loss": 0.9594, + "step": 4967 + }, + { + "epoch": 0.3447366594962182, + "grad_norm": 6.84867000579834, + "learning_rate": 0.0001983496194450769, + "loss": 0.8695, + "step": 4968 + }, + { + "epoch": 0.34480605093331484, + "grad_norm": 3.140885829925537, + "learning_rate": 0.00019834860272040868, + "loss": 0.9238, + "step": 4969 + }, + { + "epoch": 0.3448754423704115, + "grad_norm": 3.518634080886841, + "learning_rate": 0.00019834758568526535, + "loss": 0.9073, + "step": 4970 + }, + { + "epoch": 0.3449448338075082, + "grad_norm": 3.0576422214508057, + "learning_rate": 0.0001983465683396501, + "loss": 0.9152, + "step": 4971 + }, + { + "epoch": 0.3450142252446048, + "grad_norm": 2.9498982429504395, + "learning_rate": 0.00019834555068356616, + "loss": 0.982, + "step": 4972 + }, + { + "epoch": 0.34508361668170145, + "grad_norm": 3.722086191177368, + "learning_rate": 0.00019834453271701672, + "loss": 1.2981, + "step": 4973 + }, + { + "epoch": 0.3451530081187981, + "grad_norm": 4.809999465942383, + "learning_rate": 0.000198343514440005, + "loss": 1.3853, + "step": 4974 + }, + { + "epoch": 0.3452223995558948, + "grad_norm": 3.3933870792388916, + "learning_rate": 0.00019834249585253417, + "loss": 1.1563, + "step": 4975 + }, + { + "epoch": 0.34529179099299145, + "grad_norm": 3.6100285053253174, + "learning_rate": 0.00019834147695460755, + "loss": 1.0757, + "step": 4976 + }, + { + "epoch": 0.3453611824300881, + "grad_norm": 2.8579413890838623, + "learning_rate": 0.00019834045774622827, + "loss": 0.9556, + "step": 4977 + }, + { + "epoch": 0.3454305738671848, + "grad_norm": 4.1665191650390625, + "learning_rate": 0.00019833943822739959, + "loss": 1.2955, + "step": 4978 + }, + { + "epoch": 0.34549996530428145, + "grad_norm": 4.457631587982178, + "learning_rate": 0.00019833841839812468, + "loss": 1.1241, + "step": 4979 + }, + { + "epoch": 0.3455693567413781, + "grad_norm": 2.9648118019104004, + "learning_rate": 0.00019833739825840684, + "loss": 1.0324, + "step": 4980 + }, + { + "epoch": 0.3456387481784748, + "grad_norm": 2.79692006111145, + "learning_rate": 0.0001983363778082492, + "loss": 0.9002, + "step": 4981 + }, + { + "epoch": 0.34570813961557145, + "grad_norm": 3.179544687271118, + "learning_rate": 0.00019833535704765506, + "loss": 0.96, + "step": 4982 + }, + { + "epoch": 0.3457775310526681, + "grad_norm": 3.0599591732025146, + "learning_rate": 0.00019833433597662758, + "loss": 0.9812, + "step": 4983 + }, + { + "epoch": 0.3458469224897648, + "grad_norm": 2.971472978591919, + "learning_rate": 0.00019833331459517, + "loss": 0.9666, + "step": 4984 + }, + { + "epoch": 0.34591631392686145, + "grad_norm": 2.955357074737549, + "learning_rate": 0.0001983322929032856, + "loss": 0.9893, + "step": 4985 + }, + { + "epoch": 0.34598570536395806, + "grad_norm": 3.053558111190796, + "learning_rate": 0.0001983312709009775, + "loss": 1.1022, + "step": 4986 + }, + { + "epoch": 0.34605509680105473, + "grad_norm": 3.424154758453369, + "learning_rate": 0.00019833024858824906, + "loss": 0.9824, + "step": 4987 + }, + { + "epoch": 0.3461244882381514, + "grad_norm": 2.7270894050598145, + "learning_rate": 0.00019832922596510336, + "loss": 0.9168, + "step": 4988 + }, + { + "epoch": 0.34619387967524806, + "grad_norm": 3.3342976570129395, + "learning_rate": 0.00019832820303154378, + "loss": 1.1303, + "step": 4989 + }, + { + "epoch": 0.34626327111234473, + "grad_norm": 3.3830454349517822, + "learning_rate": 0.00019832717978757343, + "loss": 1.2584, + "step": 4990 + }, + { + "epoch": 0.3463326625494414, + "grad_norm": 8.64858627319336, + "learning_rate": 0.00019832615623319554, + "loss": 1.1589, + "step": 4991 + }, + { + "epoch": 0.34640205398653806, + "grad_norm": 3.955144166946411, + "learning_rate": 0.00019832513236841344, + "loss": 1.0994, + "step": 4992 + }, + { + "epoch": 0.3464714454236347, + "grad_norm": 3.281075954437256, + "learning_rate": 0.00019832410819323027, + "loss": 1.1765, + "step": 4993 + }, + { + "epoch": 0.3465408368607314, + "grad_norm": 2.4039223194122314, + "learning_rate": 0.00019832308370764932, + "loss": 0.693, + "step": 4994 + }, + { + "epoch": 0.34661022829782806, + "grad_norm": 5.1927056312561035, + "learning_rate": 0.00019832205891167381, + "loss": 0.8887, + "step": 4995 + }, + { + "epoch": 0.3466796197349247, + "grad_norm": 3.407104969024658, + "learning_rate": 0.00019832103380530694, + "loss": 0.7734, + "step": 4996 + }, + { + "epoch": 0.3467490111720214, + "grad_norm": 2.618608236312866, + "learning_rate": 0.000198320008388552, + "loss": 0.8236, + "step": 4997 + }, + { + "epoch": 0.34681840260911806, + "grad_norm": 3.8752858638763428, + "learning_rate": 0.0001983189826614122, + "loss": 1.1204, + "step": 4998 + }, + { + "epoch": 0.34688779404621467, + "grad_norm": 3.6090855598449707, + "learning_rate": 0.00019831795662389077, + "loss": 1.095, + "step": 4999 + }, + { + "epoch": 0.34695718548331134, + "grad_norm": 2.738651990890503, + "learning_rate": 0.00019831693027599095, + "loss": 1.0012, + "step": 5000 + }, + { + "epoch": 0.347026576920408, + "grad_norm": 3.6304142475128174, + "learning_rate": 0.00019831590361771599, + "loss": 0.8465, + "step": 5001 + }, + { + "epoch": 0.34709596835750467, + "grad_norm": 3.346151351928711, + "learning_rate": 0.00019831487664906915, + "loss": 0.9967, + "step": 5002 + }, + { + "epoch": 0.34716535979460134, + "grad_norm": 4.174840450286865, + "learning_rate": 0.00019831384937005363, + "loss": 1.1112, + "step": 5003 + }, + { + "epoch": 0.347234751231698, + "grad_norm": 4.154472351074219, + "learning_rate": 0.00019831282178067267, + "loss": 0.8109, + "step": 5004 + }, + { + "epoch": 0.34730414266879467, + "grad_norm": 4.236972808837891, + "learning_rate": 0.0001983117938809296, + "loss": 1.1086, + "step": 5005 + }, + { + "epoch": 0.34737353410589134, + "grad_norm": 3.144732713699341, + "learning_rate": 0.00019831076567082757, + "loss": 0.6578, + "step": 5006 + }, + { + "epoch": 0.347442925542988, + "grad_norm": 3.1070942878723145, + "learning_rate": 0.00019830973715036985, + "loss": 1.0501, + "step": 5007 + }, + { + "epoch": 0.34751231698008467, + "grad_norm": 3.96056866645813, + "learning_rate": 0.0001983087083195597, + "loss": 1.1412, + "step": 5008 + }, + { + "epoch": 0.34758170841718133, + "grad_norm": 2.900111675262451, + "learning_rate": 0.00019830767917840037, + "loss": 1.0551, + "step": 5009 + }, + { + "epoch": 0.347651099854278, + "grad_norm": 3.4722537994384766, + "learning_rate": 0.0001983066497268951, + "loss": 1.0157, + "step": 5010 + }, + { + "epoch": 0.34772049129137467, + "grad_norm": 2.485614061355591, + "learning_rate": 0.00019830561996504712, + "loss": 0.8586, + "step": 5011 + }, + { + "epoch": 0.34778988272847133, + "grad_norm": 3.3100879192352295, + "learning_rate": 0.00019830458989285977, + "loss": 0.9887, + "step": 5012 + }, + { + "epoch": 0.34785927416556794, + "grad_norm": 3.5141592025756836, + "learning_rate": 0.00019830355951033616, + "loss": 1.3245, + "step": 5013 + }, + { + "epoch": 0.3479286656026646, + "grad_norm": 3.0950064659118652, + "learning_rate": 0.00019830252881747966, + "loss": 1.0047, + "step": 5014 + }, + { + "epoch": 0.3479980570397613, + "grad_norm": 3.292567491531372, + "learning_rate": 0.00019830149781429346, + "loss": 0.9567, + "step": 5015 + }, + { + "epoch": 0.34806744847685794, + "grad_norm": 2.7604148387908936, + "learning_rate": 0.00019830046650078087, + "loss": 0.8673, + "step": 5016 + }, + { + "epoch": 0.3481368399139546, + "grad_norm": 3.6698172092437744, + "learning_rate": 0.0001982994348769451, + "loss": 1.0186, + "step": 5017 + }, + { + "epoch": 0.3482062313510513, + "grad_norm": 3.3541476726531982, + "learning_rate": 0.0001982984029427894, + "loss": 0.9276, + "step": 5018 + }, + { + "epoch": 0.34827562278814794, + "grad_norm": 3.4604225158691406, + "learning_rate": 0.00019829737069831704, + "loss": 1.1786, + "step": 5019 + }, + { + "epoch": 0.3483450142252446, + "grad_norm": 3.6683554649353027, + "learning_rate": 0.0001982963381435313, + "loss": 0.9994, + "step": 5020 + }, + { + "epoch": 0.3484144056623413, + "grad_norm": 2.9268250465393066, + "learning_rate": 0.00019829530527843542, + "loss": 1.0772, + "step": 5021 + }, + { + "epoch": 0.34848379709943794, + "grad_norm": 3.29203200340271, + "learning_rate": 0.00019829427210303266, + "loss": 0.9722, + "step": 5022 + }, + { + "epoch": 0.3485531885365346, + "grad_norm": 2.828754186630249, + "learning_rate": 0.0001982932386173263, + "loss": 0.8505, + "step": 5023 + }, + { + "epoch": 0.3486225799736313, + "grad_norm": 3.8326478004455566, + "learning_rate": 0.0001982922048213196, + "loss": 1.176, + "step": 5024 + }, + { + "epoch": 0.34869197141072794, + "grad_norm": 2.8388686180114746, + "learning_rate": 0.00019829117071501582, + "loss": 0.9628, + "step": 5025 + }, + { + "epoch": 0.34876136284782455, + "grad_norm": 3.0235414505004883, + "learning_rate": 0.0001982901362984182, + "loss": 0.9781, + "step": 5026 + }, + { + "epoch": 0.3488307542849212, + "grad_norm": 3.4757609367370605, + "learning_rate": 0.00019828910157153, + "loss": 0.8277, + "step": 5027 + }, + { + "epoch": 0.3489001457220179, + "grad_norm": 3.0966429710388184, + "learning_rate": 0.00019828806653435455, + "loss": 1.0846, + "step": 5028 + }, + { + "epoch": 0.34896953715911455, + "grad_norm": 5.0121684074401855, + "learning_rate": 0.00019828703118689503, + "loss": 0.9142, + "step": 5029 + }, + { + "epoch": 0.3490389285962112, + "grad_norm": 3.184746503829956, + "learning_rate": 0.0001982859955291548, + "loss": 1.0725, + "step": 5030 + }, + { + "epoch": 0.3491083200333079, + "grad_norm": 4.720214366912842, + "learning_rate": 0.00019828495956113708, + "loss": 1.0224, + "step": 5031 + }, + { + "epoch": 0.34917771147040455, + "grad_norm": 5.012705326080322, + "learning_rate": 0.00019828392328284514, + "loss": 1.0539, + "step": 5032 + }, + { + "epoch": 0.3492471029075012, + "grad_norm": 3.540424346923828, + "learning_rate": 0.00019828288669428224, + "loss": 1.065, + "step": 5033 + }, + { + "epoch": 0.3493164943445979, + "grad_norm": 3.381458282470703, + "learning_rate": 0.00019828184979545167, + "loss": 1.1232, + "step": 5034 + }, + { + "epoch": 0.34938588578169455, + "grad_norm": 3.114824056625366, + "learning_rate": 0.00019828081258635677, + "loss": 1.1343, + "step": 5035 + }, + { + "epoch": 0.3494552772187912, + "grad_norm": 3.2584383487701416, + "learning_rate": 0.0001982797750670007, + "loss": 1.3165, + "step": 5036 + }, + { + "epoch": 0.3495246686558879, + "grad_norm": 3.848726511001587, + "learning_rate": 0.00019827873723738675, + "loss": 1.1731, + "step": 5037 + }, + { + "epoch": 0.34959406009298455, + "grad_norm": 3.2959787845611572, + "learning_rate": 0.00019827769909751825, + "loss": 1.0906, + "step": 5038 + }, + { + "epoch": 0.3496634515300812, + "grad_norm": 3.0700063705444336, + "learning_rate": 0.00019827666064739846, + "loss": 0.9527, + "step": 5039 + }, + { + "epoch": 0.3497328429671778, + "grad_norm": 2.788647413253784, + "learning_rate": 0.00019827562188703065, + "loss": 1.1407, + "step": 5040 + }, + { + "epoch": 0.3498022344042745, + "grad_norm": 2.742973804473877, + "learning_rate": 0.00019827458281641814, + "loss": 0.9276, + "step": 5041 + }, + { + "epoch": 0.34987162584137116, + "grad_norm": 3.678187131881714, + "learning_rate": 0.00019827354343556413, + "loss": 1.3204, + "step": 5042 + }, + { + "epoch": 0.3499410172784678, + "grad_norm": 3.3478879928588867, + "learning_rate": 0.00019827250374447196, + "loss": 1.1391, + "step": 5043 + }, + { + "epoch": 0.3500104087155645, + "grad_norm": 2.823158025741577, + "learning_rate": 0.0001982714637431449, + "loss": 0.7497, + "step": 5044 + }, + { + "epoch": 0.35007980015266116, + "grad_norm": 3.8140313625335693, + "learning_rate": 0.00019827042343158625, + "loss": 1.0962, + "step": 5045 + }, + { + "epoch": 0.3501491915897578, + "grad_norm": 3.074880599975586, + "learning_rate": 0.00019826938280979924, + "loss": 1.1083, + "step": 5046 + }, + { + "epoch": 0.3502185830268545, + "grad_norm": 2.512014389038086, + "learning_rate": 0.00019826834187778718, + "loss": 0.811, + "step": 5047 + }, + { + "epoch": 0.35028797446395116, + "grad_norm": 3.186880111694336, + "learning_rate": 0.00019826730063555342, + "loss": 0.9433, + "step": 5048 + }, + { + "epoch": 0.3503573659010478, + "grad_norm": 2.9605767726898193, + "learning_rate": 0.0001982662590831011, + "loss": 0.9863, + "step": 5049 + }, + { + "epoch": 0.3504267573381445, + "grad_norm": 4.014979839324951, + "learning_rate": 0.00019826521722043368, + "loss": 1.2246, + "step": 5050 + }, + { + "epoch": 0.35049614877524116, + "grad_norm": 3.949603319168091, + "learning_rate": 0.00019826417504755436, + "loss": 1.1597, + "step": 5051 + }, + { + "epoch": 0.3505655402123378, + "grad_norm": 2.9745466709136963, + "learning_rate": 0.00019826313256446644, + "loss": 1.0126, + "step": 5052 + }, + { + "epoch": 0.35063493164943443, + "grad_norm": 2.9565176963806152, + "learning_rate": 0.0001982620897711732, + "loss": 0.9656, + "step": 5053 + }, + { + "epoch": 0.3507043230865311, + "grad_norm": 3.4015119075775146, + "learning_rate": 0.00019826104666767797, + "loss": 1.1149, + "step": 5054 + }, + { + "epoch": 0.35077371452362777, + "grad_norm": 3.5055205821990967, + "learning_rate": 0.00019826000325398399, + "loss": 0.9776, + "step": 5055 + }, + { + "epoch": 0.35084310596072443, + "grad_norm": 2.615628242492676, + "learning_rate": 0.00019825895953009458, + "loss": 0.8104, + "step": 5056 + }, + { + "epoch": 0.3509124973978211, + "grad_norm": 3.3660013675689697, + "learning_rate": 0.00019825791549601306, + "loss": 0.6768, + "step": 5057 + }, + { + "epoch": 0.35098188883491777, + "grad_norm": 2.9183685779571533, + "learning_rate": 0.00019825687115174266, + "loss": 0.8966, + "step": 5058 + }, + { + "epoch": 0.35105128027201443, + "grad_norm": 3.1460108757019043, + "learning_rate": 0.00019825582649728673, + "loss": 1.1961, + "step": 5059 + }, + { + "epoch": 0.3511206717091111, + "grad_norm": 3.54384446144104, + "learning_rate": 0.00019825478153264858, + "loss": 1.0433, + "step": 5060 + }, + { + "epoch": 0.35119006314620776, + "grad_norm": 2.9546165466308594, + "learning_rate": 0.00019825373625783146, + "loss": 0.9004, + "step": 5061 + }, + { + "epoch": 0.35125945458330443, + "grad_norm": 3.4314591884613037, + "learning_rate": 0.0001982526906728387, + "loss": 0.9236, + "step": 5062 + }, + { + "epoch": 0.3513288460204011, + "grad_norm": 5.598265171051025, + "learning_rate": 0.00019825164477767356, + "loss": 1.3238, + "step": 5063 + }, + { + "epoch": 0.35139823745749776, + "grad_norm": 2.639198064804077, + "learning_rate": 0.00019825059857233942, + "loss": 1.1654, + "step": 5064 + }, + { + "epoch": 0.35146762889459443, + "grad_norm": 3.2568001747131348, + "learning_rate": 0.00019824955205683954, + "loss": 1.0531, + "step": 5065 + }, + { + "epoch": 0.3515370203316911, + "grad_norm": 3.305663824081421, + "learning_rate": 0.00019824850523117725, + "loss": 1.0411, + "step": 5066 + }, + { + "epoch": 0.3516064117687877, + "grad_norm": 3.1001462936401367, + "learning_rate": 0.00019824745809535576, + "loss": 1.1545, + "step": 5067 + }, + { + "epoch": 0.3516758032058844, + "grad_norm": 2.3804283142089844, + "learning_rate": 0.00019824641064937847, + "loss": 0.7622, + "step": 5068 + }, + { + "epoch": 0.35174519464298104, + "grad_norm": 4.378282070159912, + "learning_rate": 0.00019824536289324868, + "loss": 1.2267, + "step": 5069 + }, + { + "epoch": 0.3518145860800777, + "grad_norm": 2.548062801361084, + "learning_rate": 0.00019824431482696965, + "loss": 0.9463, + "step": 5070 + }, + { + "epoch": 0.3518839775171744, + "grad_norm": 2.980347156524658, + "learning_rate": 0.00019824326645054475, + "loss": 0.7162, + "step": 5071 + }, + { + "epoch": 0.35195336895427104, + "grad_norm": 4.1256327629089355, + "learning_rate": 0.00019824221776397724, + "loss": 0.9889, + "step": 5072 + }, + { + "epoch": 0.3520227603913677, + "grad_norm": 4.021561622619629, + "learning_rate": 0.00019824116876727046, + "loss": 1.0848, + "step": 5073 + }, + { + "epoch": 0.3520921518284644, + "grad_norm": 3.1945297718048096, + "learning_rate": 0.00019824011946042766, + "loss": 0.8754, + "step": 5074 + }, + { + "epoch": 0.35216154326556104, + "grad_norm": 3.5630788803100586, + "learning_rate": 0.00019823906984345224, + "loss": 1.0171, + "step": 5075 + }, + { + "epoch": 0.3522309347026577, + "grad_norm": 3.035701274871826, + "learning_rate": 0.00019823801991634748, + "loss": 0.8572, + "step": 5076 + }, + { + "epoch": 0.35230032613975437, + "grad_norm": 2.5165250301361084, + "learning_rate": 0.0001982369696791167, + "loss": 0.6833, + "step": 5077 + }, + { + "epoch": 0.35236971757685104, + "grad_norm": 3.6064558029174805, + "learning_rate": 0.00019823591913176316, + "loss": 1.224, + "step": 5078 + }, + { + "epoch": 0.3524391090139477, + "grad_norm": 2.7724549770355225, + "learning_rate": 0.00019823486827429023, + "loss": 0.8518, + "step": 5079 + }, + { + "epoch": 0.3525085004510443, + "grad_norm": 4.264590263366699, + "learning_rate": 0.00019823381710670123, + "loss": 1.1263, + "step": 5080 + }, + { + "epoch": 0.352577891888141, + "grad_norm": 3.374124050140381, + "learning_rate": 0.00019823276562899947, + "loss": 0.773, + "step": 5081 + }, + { + "epoch": 0.35264728332523765, + "grad_norm": 2.8813982009887695, + "learning_rate": 0.00019823171384118825, + "loss": 1.0823, + "step": 5082 + }, + { + "epoch": 0.3527166747623343, + "grad_norm": 2.9827959537506104, + "learning_rate": 0.00019823066174327094, + "loss": 1.0693, + "step": 5083 + }, + { + "epoch": 0.352786066199431, + "grad_norm": 3.5238091945648193, + "learning_rate": 0.0001982296093352508, + "loss": 1.2592, + "step": 5084 + }, + { + "epoch": 0.35285545763652765, + "grad_norm": 2.593592882156372, + "learning_rate": 0.00019822855661713118, + "loss": 0.9029, + "step": 5085 + }, + { + "epoch": 0.3529248490736243, + "grad_norm": 3.9815165996551514, + "learning_rate": 0.00019822750358891543, + "loss": 1.0739, + "step": 5086 + }, + { + "epoch": 0.352994240510721, + "grad_norm": 3.405900239944458, + "learning_rate": 0.0001982264502506068, + "loss": 0.8674, + "step": 5087 + }, + { + "epoch": 0.35306363194781765, + "grad_norm": 4.189937591552734, + "learning_rate": 0.0001982253966022087, + "loss": 0.9301, + "step": 5088 + }, + { + "epoch": 0.3531330233849143, + "grad_norm": 2.661187171936035, + "learning_rate": 0.00019822434264372442, + "loss": 1.0334, + "step": 5089 + }, + { + "epoch": 0.353202414822011, + "grad_norm": 3.914494276046753, + "learning_rate": 0.00019822328837515728, + "loss": 1.1947, + "step": 5090 + }, + { + "epoch": 0.35327180625910765, + "grad_norm": 4.084868907928467, + "learning_rate": 0.0001982222337965106, + "loss": 0.9822, + "step": 5091 + }, + { + "epoch": 0.3533411976962043, + "grad_norm": 2.4992597103118896, + "learning_rate": 0.00019822117890778773, + "loss": 0.8711, + "step": 5092 + }, + { + "epoch": 0.3534105891333009, + "grad_norm": 2.7795352935791016, + "learning_rate": 0.000198220123708992, + "loss": 0.9155, + "step": 5093 + }, + { + "epoch": 0.3534799805703976, + "grad_norm": 3.5129313468933105, + "learning_rate": 0.00019821906820012675, + "loss": 1.0989, + "step": 5094 + }, + { + "epoch": 0.35354937200749426, + "grad_norm": 2.7865231037139893, + "learning_rate": 0.00019821801238119528, + "loss": 1.2242, + "step": 5095 + }, + { + "epoch": 0.3536187634445909, + "grad_norm": 3.6892778873443604, + "learning_rate": 0.00019821695625220093, + "loss": 1.0941, + "step": 5096 + }, + { + "epoch": 0.3536881548816876, + "grad_norm": 4.790486812591553, + "learning_rate": 0.00019821589981314705, + "loss": 1.1333, + "step": 5097 + }, + { + "epoch": 0.35375754631878426, + "grad_norm": 2.612006425857544, + "learning_rate": 0.00019821484306403698, + "loss": 0.7248, + "step": 5098 + }, + { + "epoch": 0.3538269377558809, + "grad_norm": 2.2830564975738525, + "learning_rate": 0.00019821378600487407, + "loss": 0.7825, + "step": 5099 + }, + { + "epoch": 0.3538963291929776, + "grad_norm": 2.9557347297668457, + "learning_rate": 0.0001982127286356616, + "loss": 0.7297, + "step": 5100 + }, + { + "epoch": 0.35396572063007425, + "grad_norm": 3.252589464187622, + "learning_rate": 0.00019821167095640294, + "loss": 0.9689, + "step": 5101 + }, + { + "epoch": 0.3540351120671709, + "grad_norm": 3.7473340034484863, + "learning_rate": 0.0001982106129671014, + "loss": 0.8975, + "step": 5102 + }, + { + "epoch": 0.3541045035042676, + "grad_norm": 2.397737503051758, + "learning_rate": 0.0001982095546677604, + "loss": 0.9191, + "step": 5103 + }, + { + "epoch": 0.35417389494136425, + "grad_norm": 2.89941143989563, + "learning_rate": 0.0001982084960583832, + "loss": 1.0779, + "step": 5104 + }, + { + "epoch": 0.3542432863784609, + "grad_norm": 2.6953861713409424, + "learning_rate": 0.0001982074371389732, + "loss": 1.1298, + "step": 5105 + }, + { + "epoch": 0.3543126778155576, + "grad_norm": 2.598609209060669, + "learning_rate": 0.0001982063779095337, + "loss": 0.9386, + "step": 5106 + }, + { + "epoch": 0.3543820692526542, + "grad_norm": 3.402564525604248, + "learning_rate": 0.00019820531837006806, + "loss": 1.3368, + "step": 5107 + }, + { + "epoch": 0.35445146068975086, + "grad_norm": 4.471917152404785, + "learning_rate": 0.00019820425852057962, + "loss": 1.2129, + "step": 5108 + }, + { + "epoch": 0.35452085212684753, + "grad_norm": 2.5412776470184326, + "learning_rate": 0.00019820319836107175, + "loss": 1.2203, + "step": 5109 + }, + { + "epoch": 0.3545902435639442, + "grad_norm": 3.7882375717163086, + "learning_rate": 0.00019820213789154773, + "loss": 0.9898, + "step": 5110 + }, + { + "epoch": 0.35465963500104086, + "grad_norm": 3.345891237258911, + "learning_rate": 0.00019820107711201099, + "loss": 1.0199, + "step": 5111 + }, + { + "epoch": 0.35472902643813753, + "grad_norm": 2.5265636444091797, + "learning_rate": 0.00019820001602246484, + "loss": 0.7579, + "step": 5112 + }, + { + "epoch": 0.3547984178752342, + "grad_norm": 3.0576205253601074, + "learning_rate": 0.00019819895462291264, + "loss": 1.0851, + "step": 5113 + }, + { + "epoch": 0.35486780931233086, + "grad_norm": 2.8403279781341553, + "learning_rate": 0.0001981978929133577, + "loss": 1.037, + "step": 5114 + }, + { + "epoch": 0.35493720074942753, + "grad_norm": 3.3561694622039795, + "learning_rate": 0.00019819683089380345, + "loss": 1.0459, + "step": 5115 + }, + { + "epoch": 0.3550065921865242, + "grad_norm": 3.521894693374634, + "learning_rate": 0.00019819576856425317, + "loss": 1.0509, + "step": 5116 + }, + { + "epoch": 0.35507598362362086, + "grad_norm": 3.17002010345459, + "learning_rate": 0.00019819470592471024, + "loss": 1.2517, + "step": 5117 + }, + { + "epoch": 0.3551453750607175, + "grad_norm": 3.019152879714966, + "learning_rate": 0.00019819364297517803, + "loss": 1.0763, + "step": 5118 + }, + { + "epoch": 0.3552147664978142, + "grad_norm": 4.10479736328125, + "learning_rate": 0.00019819257971565986, + "loss": 0.8377, + "step": 5119 + }, + { + "epoch": 0.3552841579349108, + "grad_norm": 3.379082441329956, + "learning_rate": 0.00019819151614615914, + "loss": 0.9688, + "step": 5120 + }, + { + "epoch": 0.35535354937200747, + "grad_norm": 3.426676034927368, + "learning_rate": 0.00019819045226667918, + "loss": 0.7818, + "step": 5121 + }, + { + "epoch": 0.35542294080910414, + "grad_norm": 3.7706458568573, + "learning_rate": 0.00019818938807722333, + "loss": 0.8781, + "step": 5122 + }, + { + "epoch": 0.3554923322462008, + "grad_norm": 4.515138626098633, + "learning_rate": 0.00019818832357779503, + "loss": 0.8878, + "step": 5123 + }, + { + "epoch": 0.35556172368329747, + "grad_norm": 3.230687379837036, + "learning_rate": 0.00019818725876839756, + "loss": 1.1205, + "step": 5124 + }, + { + "epoch": 0.35563111512039414, + "grad_norm": 3.186868190765381, + "learning_rate": 0.00019818619364903427, + "loss": 0.9748, + "step": 5125 + }, + { + "epoch": 0.3557005065574908, + "grad_norm": 3.5940845012664795, + "learning_rate": 0.00019818512821970861, + "loss": 1.0163, + "step": 5126 + }, + { + "epoch": 0.35576989799458747, + "grad_norm": 3.450605630874634, + "learning_rate": 0.00019818406248042384, + "loss": 0.8043, + "step": 5127 + }, + { + "epoch": 0.35583928943168414, + "grad_norm": 3.947892665863037, + "learning_rate": 0.00019818299643118342, + "loss": 0.987, + "step": 5128 + }, + { + "epoch": 0.3559086808687808, + "grad_norm": 3.817152261734009, + "learning_rate": 0.00019818193007199066, + "loss": 1.0619, + "step": 5129 + }, + { + "epoch": 0.35597807230587747, + "grad_norm": 4.640973091125488, + "learning_rate": 0.00019818086340284895, + "loss": 1.2654, + "step": 5130 + }, + { + "epoch": 0.35604746374297414, + "grad_norm": 3.1206891536712646, + "learning_rate": 0.0001981797964237616, + "loss": 1.0463, + "step": 5131 + }, + { + "epoch": 0.3561168551800708, + "grad_norm": 3.3433475494384766, + "learning_rate": 0.00019817872913473206, + "loss": 0.9428, + "step": 5132 + }, + { + "epoch": 0.35618624661716747, + "grad_norm": 2.72137188911438, + "learning_rate": 0.00019817766153576365, + "loss": 0.8789, + "step": 5133 + }, + { + "epoch": 0.3562556380542641, + "grad_norm": 3.4491076469421387, + "learning_rate": 0.00019817659362685977, + "loss": 1.2898, + "step": 5134 + }, + { + "epoch": 0.35632502949136075, + "grad_norm": 3.2096824645996094, + "learning_rate": 0.00019817552540802376, + "loss": 0.9323, + "step": 5135 + }, + { + "epoch": 0.3563944209284574, + "grad_norm": 2.5592923164367676, + "learning_rate": 0.000198174456879259, + "loss": 0.8449, + "step": 5136 + }, + { + "epoch": 0.3564638123655541, + "grad_norm": 3.5488779544830322, + "learning_rate": 0.0001981733880405689, + "loss": 0.8955, + "step": 5137 + }, + { + "epoch": 0.35653320380265074, + "grad_norm": 3.495450496673584, + "learning_rate": 0.00019817231889195677, + "loss": 0.9923, + "step": 5138 + }, + { + "epoch": 0.3566025952397474, + "grad_norm": 3.278414011001587, + "learning_rate": 0.00019817124943342603, + "loss": 1.0399, + "step": 5139 + }, + { + "epoch": 0.3566719866768441, + "grad_norm": 3.7646894454956055, + "learning_rate": 0.00019817017966498006, + "loss": 0.9507, + "step": 5140 + }, + { + "epoch": 0.35674137811394074, + "grad_norm": 3.2095630168914795, + "learning_rate": 0.0001981691095866222, + "loss": 0.9654, + "step": 5141 + }, + { + "epoch": 0.3568107695510374, + "grad_norm": 3.2263879776000977, + "learning_rate": 0.00019816803919835586, + "loss": 1.1876, + "step": 5142 + }, + { + "epoch": 0.3568801609881341, + "grad_norm": 3.241736888885498, + "learning_rate": 0.0001981669685001844, + "loss": 1.1568, + "step": 5143 + }, + { + "epoch": 0.35694955242523074, + "grad_norm": 3.349126100540161, + "learning_rate": 0.00019816589749211124, + "loss": 0.917, + "step": 5144 + }, + { + "epoch": 0.3570189438623274, + "grad_norm": 3.1510794162750244, + "learning_rate": 0.00019816482617413973, + "loss": 0.9865, + "step": 5145 + }, + { + "epoch": 0.3570883352994241, + "grad_norm": 3.668872594833374, + "learning_rate": 0.0001981637545462732, + "loss": 1.1375, + "step": 5146 + }, + { + "epoch": 0.3571577267365207, + "grad_norm": 2.778660297393799, + "learning_rate": 0.00019816268260851514, + "loss": 0.8657, + "step": 5147 + }, + { + "epoch": 0.35722711817361735, + "grad_norm": 3.927435874938965, + "learning_rate": 0.00019816161036086886, + "loss": 1.0817, + "step": 5148 + }, + { + "epoch": 0.357296509610714, + "grad_norm": 3.362290143966675, + "learning_rate": 0.00019816053780333777, + "loss": 1.0266, + "step": 5149 + }, + { + "epoch": 0.3573659010478107, + "grad_norm": 2.795541524887085, + "learning_rate": 0.00019815946493592526, + "loss": 0.9438, + "step": 5150 + }, + { + "epoch": 0.35743529248490735, + "grad_norm": 3.3627777099609375, + "learning_rate": 0.00019815839175863468, + "loss": 1.014, + "step": 5151 + }, + { + "epoch": 0.357504683922004, + "grad_norm": 3.5107877254486084, + "learning_rate": 0.00019815731827146945, + "loss": 1.1162, + "step": 5152 + }, + { + "epoch": 0.3575740753591007, + "grad_norm": 2.650210380554199, + "learning_rate": 0.00019815624447443297, + "loss": 0.947, + "step": 5153 + }, + { + "epoch": 0.35764346679619735, + "grad_norm": 3.817509651184082, + "learning_rate": 0.0001981551703675286, + "loss": 1.1529, + "step": 5154 + }, + { + "epoch": 0.357712858233294, + "grad_norm": 3.4876463413238525, + "learning_rate": 0.0001981540959507598, + "loss": 1.1727, + "step": 5155 + }, + { + "epoch": 0.3577822496703907, + "grad_norm": 3.5907278060913086, + "learning_rate": 0.00019815302122412983, + "loss": 1.3444, + "step": 5156 + }, + { + "epoch": 0.35785164110748735, + "grad_norm": 3.6132242679595947, + "learning_rate": 0.00019815194618764221, + "loss": 1.2248, + "step": 5157 + }, + { + "epoch": 0.357921032544584, + "grad_norm": 6.0165696144104, + "learning_rate": 0.00019815087084130027, + "loss": 1.0595, + "step": 5158 + }, + { + "epoch": 0.3579904239816807, + "grad_norm": 3.984933376312256, + "learning_rate": 0.0001981497951851074, + "loss": 1.0981, + "step": 5159 + }, + { + "epoch": 0.35805981541877735, + "grad_norm": 2.360292673110962, + "learning_rate": 0.00019814871921906705, + "loss": 0.8712, + "step": 5160 + }, + { + "epoch": 0.35812920685587396, + "grad_norm": 2.4562885761260986, + "learning_rate": 0.00019814764294318255, + "loss": 0.9202, + "step": 5161 + }, + { + "epoch": 0.3581985982929706, + "grad_norm": 2.420870780944824, + "learning_rate": 0.0001981465663574573, + "loss": 0.9489, + "step": 5162 + }, + { + "epoch": 0.3582679897300673, + "grad_norm": 2.760925531387329, + "learning_rate": 0.00019814548946189478, + "loss": 0.949, + "step": 5163 + }, + { + "epoch": 0.35833738116716396, + "grad_norm": 3.627135753631592, + "learning_rate": 0.00019814441225649834, + "loss": 1.1738, + "step": 5164 + }, + { + "epoch": 0.3584067726042606, + "grad_norm": 2.354177236557007, + "learning_rate": 0.00019814333474127137, + "loss": 0.7263, + "step": 5165 + }, + { + "epoch": 0.3584761640413573, + "grad_norm": 2.905393123626709, + "learning_rate": 0.00019814225691621725, + "loss": 0.9728, + "step": 5166 + }, + { + "epoch": 0.35854555547845396, + "grad_norm": 4.185401439666748, + "learning_rate": 0.0001981411787813394, + "loss": 1.4367, + "step": 5167 + }, + { + "epoch": 0.3586149469155506, + "grad_norm": 3.1030731201171875, + "learning_rate": 0.0001981401003366413, + "loss": 1.057, + "step": 5168 + }, + { + "epoch": 0.3586843383526473, + "grad_norm": 4.037734031677246, + "learning_rate": 0.00019813902158212626, + "loss": 0.8573, + "step": 5169 + }, + { + "epoch": 0.35875372978974396, + "grad_norm": 3.077927589416504, + "learning_rate": 0.0001981379425177977, + "loss": 1.0702, + "step": 5170 + }, + { + "epoch": 0.3588231212268406, + "grad_norm": 3.2431604862213135, + "learning_rate": 0.00019813686314365905, + "loss": 0.9453, + "step": 5171 + }, + { + "epoch": 0.3588925126639373, + "grad_norm": 3.2978131771087646, + "learning_rate": 0.00019813578345971372, + "loss": 0.9563, + "step": 5172 + }, + { + "epoch": 0.35896190410103396, + "grad_norm": 3.090585470199585, + "learning_rate": 0.00019813470346596507, + "loss": 0.8538, + "step": 5173 + }, + { + "epoch": 0.35903129553813057, + "grad_norm": 2.4293606281280518, + "learning_rate": 0.0001981336231624166, + "loss": 0.9293, + "step": 5174 + }, + { + "epoch": 0.35910068697522723, + "grad_norm": 4.063819885253906, + "learning_rate": 0.00019813254254907162, + "loss": 1.1435, + "step": 5175 + }, + { + "epoch": 0.3591700784123239, + "grad_norm": 3.5285727977752686, + "learning_rate": 0.00019813146162593362, + "loss": 0.9152, + "step": 5176 + }, + { + "epoch": 0.35923946984942057, + "grad_norm": 2.9652507305145264, + "learning_rate": 0.00019813038039300597, + "loss": 0.9862, + "step": 5177 + }, + { + "epoch": 0.35930886128651723, + "grad_norm": 3.0868887901306152, + "learning_rate": 0.0001981292988502921, + "loss": 1.0732, + "step": 5178 + }, + { + "epoch": 0.3593782527236139, + "grad_norm": 2.824221611022949, + "learning_rate": 0.00019812821699779544, + "loss": 1.1141, + "step": 5179 + }, + { + "epoch": 0.35944764416071057, + "grad_norm": 3.9236457347869873, + "learning_rate": 0.00019812713483551933, + "loss": 1.3173, + "step": 5180 + }, + { + "epoch": 0.35951703559780723, + "grad_norm": 4.492076396942139, + "learning_rate": 0.0001981260523634673, + "loss": 1.1144, + "step": 5181 + }, + { + "epoch": 0.3595864270349039, + "grad_norm": 2.683262586593628, + "learning_rate": 0.00019812496958164264, + "loss": 1.0459, + "step": 5182 + }, + { + "epoch": 0.35965581847200057, + "grad_norm": 3.1765716075897217, + "learning_rate": 0.0001981238864900489, + "loss": 1.1508, + "step": 5183 + }, + { + "epoch": 0.35972520990909723, + "grad_norm": 3.401026725769043, + "learning_rate": 0.0001981228030886894, + "loss": 1.0145, + "step": 5184 + }, + { + "epoch": 0.3597946013461939, + "grad_norm": 2.9044852256774902, + "learning_rate": 0.0001981217193775676, + "loss": 1.0617, + "step": 5185 + }, + { + "epoch": 0.35986399278329056, + "grad_norm": 2.6422252655029297, + "learning_rate": 0.0001981206353566869, + "loss": 0.9566, + "step": 5186 + }, + { + "epoch": 0.35993338422038723, + "grad_norm": 2.4827427864074707, + "learning_rate": 0.0001981195510260508, + "loss": 0.868, + "step": 5187 + }, + { + "epoch": 0.36000277565748384, + "grad_norm": 2.729417562484741, + "learning_rate": 0.00019811846638566257, + "loss": 0.9386, + "step": 5188 + }, + { + "epoch": 0.3600721670945805, + "grad_norm": 2.7519941329956055, + "learning_rate": 0.00019811738143552578, + "loss": 0.7735, + "step": 5189 + }, + { + "epoch": 0.3601415585316772, + "grad_norm": 2.8929717540740967, + "learning_rate": 0.0001981162961756438, + "loss": 0.8498, + "step": 5190 + }, + { + "epoch": 0.36021094996877384, + "grad_norm": 2.502877712249756, + "learning_rate": 0.00019811521060602005, + "loss": 0.8398, + "step": 5191 + }, + { + "epoch": 0.3602803414058705, + "grad_norm": 4.2168755531311035, + "learning_rate": 0.00019811412472665795, + "loss": 1.1965, + "step": 5192 + }, + { + "epoch": 0.3603497328429672, + "grad_norm": 5.280728816986084, + "learning_rate": 0.00019811303853756095, + "loss": 0.9694, + "step": 5193 + }, + { + "epoch": 0.36041912428006384, + "grad_norm": 2.6983489990234375, + "learning_rate": 0.0001981119520387325, + "loss": 0.7478, + "step": 5194 + }, + { + "epoch": 0.3604885157171605, + "grad_norm": 3.7588186264038086, + "learning_rate": 0.00019811086523017596, + "loss": 1.1251, + "step": 5195 + }, + { + "epoch": 0.3605579071542572, + "grad_norm": 2.3817386627197266, + "learning_rate": 0.00019810977811189485, + "loss": 0.881, + "step": 5196 + }, + { + "epoch": 0.36062729859135384, + "grad_norm": 3.177898406982422, + "learning_rate": 0.0001981086906838925, + "loss": 1.0647, + "step": 5197 + }, + { + "epoch": 0.3606966900284505, + "grad_norm": 2.6710431575775146, + "learning_rate": 0.00019810760294617242, + "loss": 0.9043, + "step": 5198 + }, + { + "epoch": 0.3607660814655472, + "grad_norm": 2.6246111392974854, + "learning_rate": 0.000198106514898738, + "loss": 1.0794, + "step": 5199 + }, + { + "epoch": 0.36083547290264384, + "grad_norm": 3.7942142486572266, + "learning_rate": 0.00019810542654159272, + "loss": 1.1384, + "step": 5200 + }, + { + "epoch": 0.36090486433974045, + "grad_norm": 4.349841117858887, + "learning_rate": 0.00019810433787473999, + "loss": 1.3739, + "step": 5201 + }, + { + "epoch": 0.3609742557768371, + "grad_norm": 3.541738510131836, + "learning_rate": 0.00019810324889818324, + "loss": 1.2584, + "step": 5202 + }, + { + "epoch": 0.3610436472139338, + "grad_norm": 3.6684656143188477, + "learning_rate": 0.0001981021596119259, + "loss": 1.1225, + "step": 5203 + }, + { + "epoch": 0.36111303865103045, + "grad_norm": 3.8348402976989746, + "learning_rate": 0.00019810107001597143, + "loss": 0.9889, + "step": 5204 + }, + { + "epoch": 0.3611824300881271, + "grad_norm": 2.947784662246704, + "learning_rate": 0.00019809998011032328, + "loss": 1.123, + "step": 5205 + }, + { + "epoch": 0.3612518215252238, + "grad_norm": 3.967036485671997, + "learning_rate": 0.0001980988898949849, + "loss": 0.9181, + "step": 5206 + }, + { + "epoch": 0.36132121296232045, + "grad_norm": 4.717576026916504, + "learning_rate": 0.00019809779936995968, + "loss": 1.2725, + "step": 5207 + }, + { + "epoch": 0.3613906043994171, + "grad_norm": 2.542691230773926, + "learning_rate": 0.00019809670853525107, + "loss": 0.8187, + "step": 5208 + }, + { + "epoch": 0.3614599958365138, + "grad_norm": 2.6268911361694336, + "learning_rate": 0.00019809561739086252, + "loss": 0.9433, + "step": 5209 + }, + { + "epoch": 0.36152938727361045, + "grad_norm": 3.1640570163726807, + "learning_rate": 0.00019809452593679754, + "loss": 0.9989, + "step": 5210 + }, + { + "epoch": 0.3615987787107071, + "grad_norm": 4.659074306488037, + "learning_rate": 0.00019809343417305947, + "loss": 0.9358, + "step": 5211 + }, + { + "epoch": 0.3616681701478038, + "grad_norm": 3.238140821456909, + "learning_rate": 0.00019809234209965183, + "loss": 0.9089, + "step": 5212 + }, + { + "epoch": 0.36173756158490045, + "grad_norm": 3.53230881690979, + "learning_rate": 0.00019809124971657807, + "loss": 0.8711, + "step": 5213 + }, + { + "epoch": 0.3618069530219971, + "grad_norm": 3.1034152507781982, + "learning_rate": 0.0001980901570238416, + "loss": 0.8616, + "step": 5214 + }, + { + "epoch": 0.3618763444590937, + "grad_norm": 3.7034597396850586, + "learning_rate": 0.00019808906402144587, + "loss": 1.0179, + "step": 5215 + }, + { + "epoch": 0.3619457358961904, + "grad_norm": 2.3479299545288086, + "learning_rate": 0.00019808797070939435, + "loss": 0.874, + "step": 5216 + }, + { + "epoch": 0.36201512733328706, + "grad_norm": 3.1114003658294678, + "learning_rate": 0.00019808687708769052, + "loss": 0.9867, + "step": 5217 + }, + { + "epoch": 0.3620845187703837, + "grad_norm": 6.154256820678711, + "learning_rate": 0.00019808578315633777, + "loss": 1.2239, + "step": 5218 + }, + { + "epoch": 0.3621539102074804, + "grad_norm": 3.3799784183502197, + "learning_rate": 0.00019808468891533955, + "loss": 0.8826, + "step": 5219 + }, + { + "epoch": 0.36222330164457706, + "grad_norm": 3.106943368911743, + "learning_rate": 0.00019808359436469938, + "loss": 0.8834, + "step": 5220 + }, + { + "epoch": 0.3622926930816737, + "grad_norm": 2.5624115467071533, + "learning_rate": 0.00019808249950442071, + "loss": 0.8227, + "step": 5221 + }, + { + "epoch": 0.3623620845187704, + "grad_norm": 2.6495914459228516, + "learning_rate": 0.00019808140433450695, + "loss": 0.9316, + "step": 5222 + }, + { + "epoch": 0.36243147595586706, + "grad_norm": 3.653291702270508, + "learning_rate": 0.00019808030885496157, + "loss": 1.275, + "step": 5223 + }, + { + "epoch": 0.3625008673929637, + "grad_norm": 3.622541904449463, + "learning_rate": 0.00019807921306578803, + "loss": 1.2723, + "step": 5224 + }, + { + "epoch": 0.3625702588300604, + "grad_norm": 3.2421305179595947, + "learning_rate": 0.0001980781169669898, + "loss": 1.0162, + "step": 5225 + }, + { + "epoch": 0.36263965026715705, + "grad_norm": 2.9944427013397217, + "learning_rate": 0.00019807702055857034, + "loss": 1.3253, + "step": 5226 + }, + { + "epoch": 0.3627090417042537, + "grad_norm": 5.5766801834106445, + "learning_rate": 0.00019807592384053308, + "loss": 1.0281, + "step": 5227 + }, + { + "epoch": 0.36277843314135033, + "grad_norm": 2.9487390518188477, + "learning_rate": 0.00019807482681288152, + "loss": 1.0818, + "step": 5228 + }, + { + "epoch": 0.362847824578447, + "grad_norm": 2.77988600730896, + "learning_rate": 0.0001980737294756191, + "loss": 0.9392, + "step": 5229 + }, + { + "epoch": 0.36291721601554366, + "grad_norm": 3.0732975006103516, + "learning_rate": 0.00019807263182874932, + "loss": 0.9715, + "step": 5230 + }, + { + "epoch": 0.36298660745264033, + "grad_norm": 3.447589874267578, + "learning_rate": 0.00019807153387227562, + "loss": 1.1207, + "step": 5231 + }, + { + "epoch": 0.363055998889737, + "grad_norm": 3.247098684310913, + "learning_rate": 0.00019807043560620145, + "loss": 1.0467, + "step": 5232 + }, + { + "epoch": 0.36312539032683366, + "grad_norm": 2.928682327270508, + "learning_rate": 0.0001980693370305303, + "loss": 1.3, + "step": 5233 + }, + { + "epoch": 0.36319478176393033, + "grad_norm": 3.5757994651794434, + "learning_rate": 0.00019806823814526563, + "loss": 0.9386, + "step": 5234 + }, + { + "epoch": 0.363264173201027, + "grad_norm": 2.7801902294158936, + "learning_rate": 0.00019806713895041092, + "loss": 0.7751, + "step": 5235 + }, + { + "epoch": 0.36333356463812366, + "grad_norm": 3.3931620121002197, + "learning_rate": 0.0001980660394459696, + "loss": 1.2902, + "step": 5236 + }, + { + "epoch": 0.36340295607522033, + "grad_norm": 3.200155019760132, + "learning_rate": 0.00019806493963194522, + "loss": 1.0678, + "step": 5237 + }, + { + "epoch": 0.363472347512317, + "grad_norm": 2.691201686859131, + "learning_rate": 0.00019806383950834116, + "loss": 0.933, + "step": 5238 + }, + { + "epoch": 0.36354173894941366, + "grad_norm": 3.0968501567840576, + "learning_rate": 0.00019806273907516095, + "loss": 0.9609, + "step": 5239 + }, + { + "epoch": 0.36361113038651033, + "grad_norm": 2.692688226699829, + "learning_rate": 0.00019806163833240806, + "loss": 1.0997, + "step": 5240 + }, + { + "epoch": 0.36368052182360694, + "grad_norm": 3.738682985305786, + "learning_rate": 0.00019806053728008594, + "loss": 1.0198, + "step": 5241 + }, + { + "epoch": 0.3637499132607036, + "grad_norm": 2.93464732170105, + "learning_rate": 0.00019805943591819807, + "loss": 1.1657, + "step": 5242 + }, + { + "epoch": 0.36381930469780027, + "grad_norm": 3.094968795776367, + "learning_rate": 0.00019805833424674796, + "loss": 1.0434, + "step": 5243 + }, + { + "epoch": 0.36388869613489694, + "grad_norm": 2.8575286865234375, + "learning_rate": 0.00019805723226573908, + "loss": 1.0523, + "step": 5244 + }, + { + "epoch": 0.3639580875719936, + "grad_norm": 3.2728445529937744, + "learning_rate": 0.00019805612997517486, + "loss": 1.0501, + "step": 5245 + }, + { + "epoch": 0.36402747900909027, + "grad_norm": 3.4142746925354004, + "learning_rate": 0.0001980550273750588, + "loss": 0.9295, + "step": 5246 + }, + { + "epoch": 0.36409687044618694, + "grad_norm": 4.081113338470459, + "learning_rate": 0.00019805392446539446, + "loss": 1.2237, + "step": 5247 + }, + { + "epoch": 0.3641662618832836, + "grad_norm": 2.7933237552642822, + "learning_rate": 0.0001980528212461852, + "loss": 1.0342, + "step": 5248 + }, + { + "epoch": 0.36423565332038027, + "grad_norm": 3.1661770343780518, + "learning_rate": 0.0001980517177174346, + "loss": 1.1069, + "step": 5249 + }, + { + "epoch": 0.36430504475747694, + "grad_norm": 2.6825478076934814, + "learning_rate": 0.00019805061387914607, + "loss": 0.7815, + "step": 5250 + }, + { + "epoch": 0.3643744361945736, + "grad_norm": 2.8856937885284424, + "learning_rate": 0.00019804950973132315, + "loss": 0.724, + "step": 5251 + }, + { + "epoch": 0.36444382763167027, + "grad_norm": 2.5917601585388184, + "learning_rate": 0.0001980484052739693, + "loss": 0.8304, + "step": 5252 + }, + { + "epoch": 0.36451321906876694, + "grad_norm": 3.0287559032440186, + "learning_rate": 0.000198047300507088, + "loss": 1.0658, + "step": 5253 + }, + { + "epoch": 0.3645826105058636, + "grad_norm": 3.0613722801208496, + "learning_rate": 0.00019804619543068274, + "loss": 1.0139, + "step": 5254 + }, + { + "epoch": 0.3646520019429602, + "grad_norm": 3.2186193466186523, + "learning_rate": 0.00019804509004475702, + "loss": 1.029, + "step": 5255 + }, + { + "epoch": 0.3647213933800569, + "grad_norm": 3.183694362640381, + "learning_rate": 0.00019804398434931434, + "loss": 1.0537, + "step": 5256 + }, + { + "epoch": 0.36479078481715355, + "grad_norm": 3.041588306427002, + "learning_rate": 0.00019804287834435818, + "loss": 0.8316, + "step": 5257 + }, + { + "epoch": 0.3648601762542502, + "grad_norm": 2.7833940982818604, + "learning_rate": 0.00019804177202989204, + "loss": 0.9037, + "step": 5258 + }, + { + "epoch": 0.3649295676913469, + "grad_norm": 2.7530643939971924, + "learning_rate": 0.00019804066540591938, + "loss": 0.8968, + "step": 5259 + }, + { + "epoch": 0.36499895912844355, + "grad_norm": 3.3418803215026855, + "learning_rate": 0.0001980395584724437, + "loss": 1.0469, + "step": 5260 + }, + { + "epoch": 0.3650683505655402, + "grad_norm": 4.012876033782959, + "learning_rate": 0.00019803845122946855, + "loss": 1.187, + "step": 5261 + }, + { + "epoch": 0.3651377420026369, + "grad_norm": 2.710322141647339, + "learning_rate": 0.00019803734367699734, + "loss": 1.1266, + "step": 5262 + }, + { + "epoch": 0.36520713343973354, + "grad_norm": 3.0095670223236084, + "learning_rate": 0.0001980362358150336, + "loss": 1.0339, + "step": 5263 + }, + { + "epoch": 0.3652765248768302, + "grad_norm": 3.169318675994873, + "learning_rate": 0.00019803512764358087, + "loss": 1.1315, + "step": 5264 + }, + { + "epoch": 0.3653459163139269, + "grad_norm": 2.850954294204712, + "learning_rate": 0.00019803401916264265, + "loss": 0.9126, + "step": 5265 + }, + { + "epoch": 0.36541530775102354, + "grad_norm": 3.6521801948547363, + "learning_rate": 0.00019803291037222235, + "loss": 1.3184, + "step": 5266 + }, + { + "epoch": 0.3654846991881202, + "grad_norm": 2.8757410049438477, + "learning_rate": 0.00019803180127232355, + "loss": 0.94, + "step": 5267 + }, + { + "epoch": 0.3655540906252168, + "grad_norm": 3.0372838973999023, + "learning_rate": 0.00019803069186294972, + "loss": 0.9059, + "step": 5268 + }, + { + "epoch": 0.3656234820623135, + "grad_norm": 3.21712064743042, + "learning_rate": 0.00019802958214410436, + "loss": 1.1923, + "step": 5269 + }, + { + "epoch": 0.36569287349941015, + "grad_norm": 4.284080982208252, + "learning_rate": 0.000198028472115791, + "loss": 1.009, + "step": 5270 + }, + { + "epoch": 0.3657622649365068, + "grad_norm": 3.1026527881622314, + "learning_rate": 0.0001980273617780131, + "loss": 1.0869, + "step": 5271 + }, + { + "epoch": 0.3658316563736035, + "grad_norm": 2.822648048400879, + "learning_rate": 0.00019802625113077423, + "loss": 0.7074, + "step": 5272 + }, + { + "epoch": 0.36590104781070015, + "grad_norm": 3.3526089191436768, + "learning_rate": 0.00019802514017407786, + "loss": 1.2884, + "step": 5273 + }, + { + "epoch": 0.3659704392477968, + "grad_norm": 3.9582908153533936, + "learning_rate": 0.00019802402890792747, + "loss": 1.0359, + "step": 5274 + }, + { + "epoch": 0.3660398306848935, + "grad_norm": 3.172348737716675, + "learning_rate": 0.0001980229173323266, + "loss": 1.1071, + "step": 5275 + }, + { + "epoch": 0.36610922212199015, + "grad_norm": 2.8776590824127197, + "learning_rate": 0.00019802180544727875, + "loss": 0.8756, + "step": 5276 + }, + { + "epoch": 0.3661786135590868, + "grad_norm": 3.601469039916992, + "learning_rate": 0.00019802069325278743, + "loss": 1.0411, + "step": 5277 + }, + { + "epoch": 0.3662480049961835, + "grad_norm": 3.2592251300811768, + "learning_rate": 0.00019801958074885617, + "loss": 0.9653, + "step": 5278 + }, + { + "epoch": 0.36631739643328015, + "grad_norm": 3.037276268005371, + "learning_rate": 0.00019801846793548843, + "loss": 0.8204, + "step": 5279 + }, + { + "epoch": 0.3663867878703768, + "grad_norm": 2.9725420475006104, + "learning_rate": 0.0001980173548126878, + "loss": 1.0905, + "step": 5280 + }, + { + "epoch": 0.3664561793074735, + "grad_norm": 4.0096940994262695, + "learning_rate": 0.00019801624138045775, + "loss": 0.882, + "step": 5281 + }, + { + "epoch": 0.3665255707445701, + "grad_norm": 3.702918291091919, + "learning_rate": 0.00019801512763880177, + "loss": 1.1454, + "step": 5282 + }, + { + "epoch": 0.36659496218166676, + "grad_norm": 4.727770805358887, + "learning_rate": 0.0001980140135877234, + "loss": 1.0666, + "step": 5283 + }, + { + "epoch": 0.36666435361876343, + "grad_norm": 2.6227521896362305, + "learning_rate": 0.0001980128992272262, + "loss": 0.9066, + "step": 5284 + }, + { + "epoch": 0.3667337450558601, + "grad_norm": 3.2816433906555176, + "learning_rate": 0.00019801178455731362, + "loss": 1.1484, + "step": 5285 + }, + { + "epoch": 0.36680313649295676, + "grad_norm": 2.7028000354766846, + "learning_rate": 0.00019801066957798923, + "loss": 0.9529, + "step": 5286 + }, + { + "epoch": 0.3668725279300534, + "grad_norm": 2.245023727416992, + "learning_rate": 0.00019800955428925652, + "loss": 0.5639, + "step": 5287 + }, + { + "epoch": 0.3669419193671501, + "grad_norm": 3.765406847000122, + "learning_rate": 0.000198008438691119, + "loss": 1.1027, + "step": 5288 + }, + { + "epoch": 0.36701131080424676, + "grad_norm": 2.86430025100708, + "learning_rate": 0.00019800732278358024, + "loss": 0.848, + "step": 5289 + }, + { + "epoch": 0.3670807022413434, + "grad_norm": 3.290618896484375, + "learning_rate": 0.0001980062065666437, + "loss": 1.2878, + "step": 5290 + }, + { + "epoch": 0.3671500936784401, + "grad_norm": 2.6763083934783936, + "learning_rate": 0.00019800509004031295, + "loss": 1.1149, + "step": 5291 + }, + { + "epoch": 0.36721948511553676, + "grad_norm": 3.046201229095459, + "learning_rate": 0.0001980039732045915, + "loss": 1.0135, + "step": 5292 + }, + { + "epoch": 0.3672888765526334, + "grad_norm": 2.3809499740600586, + "learning_rate": 0.00019800285605948287, + "loss": 0.7176, + "step": 5293 + }, + { + "epoch": 0.3673582679897301, + "grad_norm": 4.063191890716553, + "learning_rate": 0.0001980017386049906, + "loss": 1.047, + "step": 5294 + }, + { + "epoch": 0.3674276594268267, + "grad_norm": 2.9031901359558105, + "learning_rate": 0.00019800062084111822, + "loss": 0.8184, + "step": 5295 + }, + { + "epoch": 0.36749705086392337, + "grad_norm": 2.625164747238159, + "learning_rate": 0.00019799950276786925, + "loss": 0.939, + "step": 5296 + }, + { + "epoch": 0.36756644230102004, + "grad_norm": 2.833791732788086, + "learning_rate": 0.00019799838438524722, + "loss": 0.8754, + "step": 5297 + }, + { + "epoch": 0.3676358337381167, + "grad_norm": 3.1884615421295166, + "learning_rate": 0.00019799726569325564, + "loss": 0.936, + "step": 5298 + }, + { + "epoch": 0.36770522517521337, + "grad_norm": 2.5478715896606445, + "learning_rate": 0.00019799614669189808, + "loss": 0.8595, + "step": 5299 + }, + { + "epoch": 0.36777461661231003, + "grad_norm": 2.905287981033325, + "learning_rate": 0.00019799502738117804, + "loss": 0.7593, + "step": 5300 + }, + { + "epoch": 0.3678440080494067, + "grad_norm": 4.04644250869751, + "learning_rate": 0.0001979939077610991, + "loss": 1.1685, + "step": 5301 + }, + { + "epoch": 0.36791339948650337, + "grad_norm": 4.122577667236328, + "learning_rate": 0.00019799278783166476, + "loss": 1.0506, + "step": 5302 + }, + { + "epoch": 0.36798279092360003, + "grad_norm": 4.146118640899658, + "learning_rate": 0.00019799166759287853, + "loss": 1.3044, + "step": 5303 + }, + { + "epoch": 0.3680521823606967, + "grad_norm": 2.7414181232452393, + "learning_rate": 0.00019799054704474397, + "loss": 0.8219, + "step": 5304 + }, + { + "epoch": 0.36812157379779337, + "grad_norm": 3.8057808876037598, + "learning_rate": 0.00019798942618726466, + "loss": 1.1896, + "step": 5305 + }, + { + "epoch": 0.36819096523489003, + "grad_norm": 4.00393533706665, + "learning_rate": 0.00019798830502044409, + "loss": 1.1724, + "step": 5306 + }, + { + "epoch": 0.3682603566719867, + "grad_norm": 3.3438217639923096, + "learning_rate": 0.00019798718354428578, + "loss": 1.1048, + "step": 5307 + }, + { + "epoch": 0.36832974810908337, + "grad_norm": 3.5301613807678223, + "learning_rate": 0.00019798606175879332, + "loss": 0.9013, + "step": 5308 + }, + { + "epoch": 0.36839913954618, + "grad_norm": 3.487060070037842, + "learning_rate": 0.00019798493966397023, + "loss": 1.0727, + "step": 5309 + }, + { + "epoch": 0.36846853098327664, + "grad_norm": 3.311431646347046, + "learning_rate": 0.00019798381725982006, + "loss": 1.1796, + "step": 5310 + }, + { + "epoch": 0.3685379224203733, + "grad_norm": 2.9129233360290527, + "learning_rate": 0.00019798269454634633, + "loss": 1.0039, + "step": 5311 + }, + { + "epoch": 0.36860731385747, + "grad_norm": 3.975916862487793, + "learning_rate": 0.00019798157152355262, + "loss": 1.0435, + "step": 5312 + }, + { + "epoch": 0.36867670529456664, + "grad_norm": 3.780806303024292, + "learning_rate": 0.00019798044819144243, + "loss": 1.0507, + "step": 5313 + }, + { + "epoch": 0.3687460967316633, + "grad_norm": 3.2003657817840576, + "learning_rate": 0.00019797932455001936, + "loss": 0.9914, + "step": 5314 + }, + { + "epoch": 0.36881548816876, + "grad_norm": 3.299513101577759, + "learning_rate": 0.0001979782005992869, + "loss": 1.0787, + "step": 5315 + }, + { + "epoch": 0.36888487960585664, + "grad_norm": 2.848858594894409, + "learning_rate": 0.0001979770763392487, + "loss": 0.9641, + "step": 5316 + }, + { + "epoch": 0.3689542710429533, + "grad_norm": 2.722353219985962, + "learning_rate": 0.00019797595176990814, + "loss": 1.0155, + "step": 5317 + }, + { + "epoch": 0.36902366248005, + "grad_norm": 2.3596458435058594, + "learning_rate": 0.00019797482689126893, + "loss": 1.0202, + "step": 5318 + }, + { + "epoch": 0.36909305391714664, + "grad_norm": 2.5443081855773926, + "learning_rate": 0.00019797370170333457, + "loss": 0.7529, + "step": 5319 + }, + { + "epoch": 0.3691624453542433, + "grad_norm": 3.184566020965576, + "learning_rate": 0.00019797257620610855, + "loss": 0.868, + "step": 5320 + }, + { + "epoch": 0.36923183679134, + "grad_norm": 2.6375350952148438, + "learning_rate": 0.0001979714503995945, + "loss": 0.8746, + "step": 5321 + }, + { + "epoch": 0.3693012282284366, + "grad_norm": 2.731215000152588, + "learning_rate": 0.00019797032428379595, + "loss": 0.9299, + "step": 5322 + }, + { + "epoch": 0.36937061966553325, + "grad_norm": 3.261464834213257, + "learning_rate": 0.00019796919785871646, + "loss": 0.9909, + "step": 5323 + }, + { + "epoch": 0.3694400111026299, + "grad_norm": 3.0017271041870117, + "learning_rate": 0.00019796807112435955, + "loss": 0.9621, + "step": 5324 + }, + { + "epoch": 0.3695094025397266, + "grad_norm": 2.9646928310394287, + "learning_rate": 0.00019796694408072884, + "loss": 1.0758, + "step": 5325 + }, + { + "epoch": 0.36957879397682325, + "grad_norm": 2.4905364513397217, + "learning_rate": 0.00019796581672782782, + "loss": 0.8989, + "step": 5326 + }, + { + "epoch": 0.3696481854139199, + "grad_norm": 2.87701153755188, + "learning_rate": 0.00019796468906566013, + "loss": 0.9563, + "step": 5327 + }, + { + "epoch": 0.3697175768510166, + "grad_norm": 2.763648509979248, + "learning_rate": 0.00019796356109422924, + "loss": 0.9448, + "step": 5328 + }, + { + "epoch": 0.36978696828811325, + "grad_norm": 2.925708293914795, + "learning_rate": 0.00019796243281353877, + "loss": 1.0135, + "step": 5329 + }, + { + "epoch": 0.3698563597252099, + "grad_norm": 2.5681638717651367, + "learning_rate": 0.00019796130422359225, + "loss": 0.8834, + "step": 5330 + }, + { + "epoch": 0.3699257511623066, + "grad_norm": 3.8203089237213135, + "learning_rate": 0.00019796017532439328, + "loss": 1.0472, + "step": 5331 + }, + { + "epoch": 0.36999514259940325, + "grad_norm": 3.0069949626922607, + "learning_rate": 0.0001979590461159454, + "loss": 1.0469, + "step": 5332 + }, + { + "epoch": 0.3700645340364999, + "grad_norm": 3.1043853759765625, + "learning_rate": 0.00019795791659825215, + "loss": 1.0006, + "step": 5333 + }, + { + "epoch": 0.3701339254735966, + "grad_norm": 2.6929070949554443, + "learning_rate": 0.00019795678677131715, + "loss": 0.8009, + "step": 5334 + }, + { + "epoch": 0.37020331691069325, + "grad_norm": 2.7055318355560303, + "learning_rate": 0.00019795565663514392, + "loss": 0.9092, + "step": 5335 + }, + { + "epoch": 0.37027270834778986, + "grad_norm": 2.715202808380127, + "learning_rate": 0.00019795452618973607, + "loss": 0.9297, + "step": 5336 + }, + { + "epoch": 0.3703420997848865, + "grad_norm": 3.069448947906494, + "learning_rate": 0.0001979533954350971, + "loss": 1.0146, + "step": 5337 + }, + { + "epoch": 0.3704114912219832, + "grad_norm": 2.809715747833252, + "learning_rate": 0.00019795226437123068, + "loss": 0.901, + "step": 5338 + }, + { + "epoch": 0.37048088265907986, + "grad_norm": 3.364353656768799, + "learning_rate": 0.0001979511329981403, + "loss": 0.9396, + "step": 5339 + }, + { + "epoch": 0.3705502740961765, + "grad_norm": 2.8223838806152344, + "learning_rate": 0.00019795000131582957, + "loss": 1.0707, + "step": 5340 + }, + { + "epoch": 0.3706196655332732, + "grad_norm": 2.408378839492798, + "learning_rate": 0.00019794886932430203, + "loss": 0.6916, + "step": 5341 + }, + { + "epoch": 0.37068905697036986, + "grad_norm": 3.1693215370178223, + "learning_rate": 0.00019794773702356132, + "loss": 0.9373, + "step": 5342 + }, + { + "epoch": 0.3707584484074665, + "grad_norm": 2.4012162685394287, + "learning_rate": 0.0001979466044136109, + "loss": 1.1415, + "step": 5343 + }, + { + "epoch": 0.3708278398445632, + "grad_norm": 3.380441904067993, + "learning_rate": 0.00019794547149445447, + "loss": 0.984, + "step": 5344 + }, + { + "epoch": 0.37089723128165986, + "grad_norm": 2.8601603507995605, + "learning_rate": 0.00019794433826609553, + "loss": 0.8433, + "step": 5345 + }, + { + "epoch": 0.3709666227187565, + "grad_norm": 2.582111358642578, + "learning_rate": 0.00019794320472853766, + "loss": 1.106, + "step": 5346 + }, + { + "epoch": 0.3710360141558532, + "grad_norm": 2.689687490463257, + "learning_rate": 0.00019794207088178446, + "loss": 0.9673, + "step": 5347 + }, + { + "epoch": 0.37110540559294986, + "grad_norm": 3.335460662841797, + "learning_rate": 0.00019794093672583952, + "loss": 1.0363, + "step": 5348 + }, + { + "epoch": 0.37117479703004647, + "grad_norm": 3.053560972213745, + "learning_rate": 0.0001979398022607064, + "loss": 1.1713, + "step": 5349 + }, + { + "epoch": 0.37124418846714313, + "grad_norm": 2.718437433242798, + "learning_rate": 0.00019793866748638869, + "loss": 0.956, + "step": 5350 + }, + { + "epoch": 0.3713135799042398, + "grad_norm": 4.703678607940674, + "learning_rate": 0.00019793753240289, + "loss": 0.9069, + "step": 5351 + }, + { + "epoch": 0.37138297134133647, + "grad_norm": 2.9795076847076416, + "learning_rate": 0.00019793639701021386, + "loss": 0.9805, + "step": 5352 + }, + { + "epoch": 0.37145236277843313, + "grad_norm": 3.599067449569702, + "learning_rate": 0.00019793526130836386, + "loss": 1.0132, + "step": 5353 + }, + { + "epoch": 0.3715217542155298, + "grad_norm": 3.484410285949707, + "learning_rate": 0.0001979341252973436, + "loss": 1.2154, + "step": 5354 + }, + { + "epoch": 0.37159114565262646, + "grad_norm": 3.215038776397705, + "learning_rate": 0.0001979329889771567, + "loss": 0.8427, + "step": 5355 + }, + { + "epoch": 0.37166053708972313, + "grad_norm": 4.579373836517334, + "learning_rate": 0.0001979318523478067, + "loss": 1.4004, + "step": 5356 + }, + { + "epoch": 0.3717299285268198, + "grad_norm": 3.174192428588867, + "learning_rate": 0.0001979307154092972, + "loss": 1.0094, + "step": 5357 + }, + { + "epoch": 0.37179931996391646, + "grad_norm": 3.97552752494812, + "learning_rate": 0.0001979295781616318, + "loss": 0.9309, + "step": 5358 + }, + { + "epoch": 0.37186871140101313, + "grad_norm": 3.960853338241577, + "learning_rate": 0.00019792844060481407, + "loss": 1.1958, + "step": 5359 + }, + { + "epoch": 0.3719381028381098, + "grad_norm": 3.548008441925049, + "learning_rate": 0.00019792730273884764, + "loss": 1.2951, + "step": 5360 + }, + { + "epoch": 0.37200749427520646, + "grad_norm": 3.8560590744018555, + "learning_rate": 0.00019792616456373605, + "loss": 1.3521, + "step": 5361 + }, + { + "epoch": 0.37207688571230313, + "grad_norm": 2.503514289855957, + "learning_rate": 0.00019792502607948295, + "loss": 1.1567, + "step": 5362 + }, + { + "epoch": 0.37214627714939974, + "grad_norm": 2.8119332790374756, + "learning_rate": 0.00019792388728609187, + "loss": 0.8083, + "step": 5363 + }, + { + "epoch": 0.3722156685864964, + "grad_norm": 3.5165581703186035, + "learning_rate": 0.00019792274818356646, + "loss": 0.9547, + "step": 5364 + }, + { + "epoch": 0.3722850600235931, + "grad_norm": 3.4529876708984375, + "learning_rate": 0.0001979216087719103, + "loss": 1.1335, + "step": 5365 + }, + { + "epoch": 0.37235445146068974, + "grad_norm": 3.2937495708465576, + "learning_rate": 0.00019792046905112696, + "loss": 1.0559, + "step": 5366 + }, + { + "epoch": 0.3724238428977864, + "grad_norm": 4.14012336730957, + "learning_rate": 0.00019791932902122006, + "loss": 0.9323, + "step": 5367 + }, + { + "epoch": 0.37249323433488307, + "grad_norm": 2.5382871627807617, + "learning_rate": 0.00019791818868219324, + "loss": 0.7254, + "step": 5368 + }, + { + "epoch": 0.37256262577197974, + "grad_norm": 3.158247709274292, + "learning_rate": 0.00019791704803405003, + "loss": 0.7556, + "step": 5369 + }, + { + "epoch": 0.3726320172090764, + "grad_norm": 3.0984885692596436, + "learning_rate": 0.00019791590707679405, + "loss": 0.9501, + "step": 5370 + }, + { + "epoch": 0.37270140864617307, + "grad_norm": 2.885523796081543, + "learning_rate": 0.00019791476581042892, + "loss": 1.2959, + "step": 5371 + }, + { + "epoch": 0.37277080008326974, + "grad_norm": 3.0122311115264893, + "learning_rate": 0.00019791362423495824, + "loss": 1.1264, + "step": 5372 + }, + { + "epoch": 0.3728401915203664, + "grad_norm": 3.6260693073272705, + "learning_rate": 0.00019791248235038558, + "loss": 1.0322, + "step": 5373 + }, + { + "epoch": 0.37290958295746307, + "grad_norm": 2.9586832523345947, + "learning_rate": 0.00019791134015671461, + "loss": 1.0378, + "step": 5374 + }, + { + "epoch": 0.37297897439455974, + "grad_norm": 2.7915778160095215, + "learning_rate": 0.00019791019765394888, + "loss": 0.7952, + "step": 5375 + }, + { + "epoch": 0.37304836583165635, + "grad_norm": 2.889539957046509, + "learning_rate": 0.00019790905484209204, + "loss": 0.9979, + "step": 5376 + }, + { + "epoch": 0.373117757268753, + "grad_norm": 2.4220547676086426, + "learning_rate": 0.00019790791172114768, + "loss": 0.9906, + "step": 5377 + }, + { + "epoch": 0.3731871487058497, + "grad_norm": 3.0910682678222656, + "learning_rate": 0.00019790676829111936, + "loss": 0.9463, + "step": 5378 + }, + { + "epoch": 0.37325654014294635, + "grad_norm": 2.4603261947631836, + "learning_rate": 0.00019790562455201073, + "loss": 0.9135, + "step": 5379 + }, + { + "epoch": 0.373325931580043, + "grad_norm": 2.911553144454956, + "learning_rate": 0.00019790448050382544, + "loss": 1.3049, + "step": 5380 + }, + { + "epoch": 0.3733953230171397, + "grad_norm": 3.0511205196380615, + "learning_rate": 0.00019790333614656705, + "loss": 1.0756, + "step": 5381 + }, + { + "epoch": 0.37346471445423635, + "grad_norm": 2.1178476810455322, + "learning_rate": 0.00019790219148023915, + "loss": 0.7985, + "step": 5382 + }, + { + "epoch": 0.373534105891333, + "grad_norm": 3.8058857917785645, + "learning_rate": 0.00019790104650484547, + "loss": 1.036, + "step": 5383 + }, + { + "epoch": 0.3736034973284297, + "grad_norm": 4.00698184967041, + "learning_rate": 0.0001978999012203895, + "loss": 1.2606, + "step": 5384 + }, + { + "epoch": 0.37367288876552635, + "grad_norm": 3.1986382007598877, + "learning_rate": 0.00019789875562687488, + "loss": 1.0211, + "step": 5385 + }, + { + "epoch": 0.373742280202623, + "grad_norm": 3.6237971782684326, + "learning_rate": 0.00019789760972430527, + "loss": 0.9671, + "step": 5386 + }, + { + "epoch": 0.3738116716397197, + "grad_norm": 3.077089786529541, + "learning_rate": 0.00019789646351268427, + "loss": 1.0621, + "step": 5387 + }, + { + "epoch": 0.37388106307681634, + "grad_norm": 3.5000100135803223, + "learning_rate": 0.00019789531699201547, + "loss": 1.1151, + "step": 5388 + }, + { + "epoch": 0.37395045451391296, + "grad_norm": 2.8198070526123047, + "learning_rate": 0.00019789417016230254, + "loss": 1.1028, + "step": 5389 + }, + { + "epoch": 0.3740198459510096, + "grad_norm": 1.8359856605529785, + "learning_rate": 0.00019789302302354903, + "loss": 0.9192, + "step": 5390 + }, + { + "epoch": 0.3740892373881063, + "grad_norm": 3.1239030361175537, + "learning_rate": 0.00019789187557575866, + "loss": 1.1922, + "step": 5391 + }, + { + "epoch": 0.37415862882520295, + "grad_norm": 2.236635446548462, + "learning_rate": 0.00019789072781893494, + "loss": 1.0135, + "step": 5392 + }, + { + "epoch": 0.3742280202622996, + "grad_norm": 3.573028087615967, + "learning_rate": 0.0001978895797530816, + "loss": 1.2186, + "step": 5393 + }, + { + "epoch": 0.3742974116993963, + "grad_norm": 2.7732386589050293, + "learning_rate": 0.00019788843137820216, + "loss": 0.8894, + "step": 5394 + }, + { + "epoch": 0.37436680313649295, + "grad_norm": 2.5763766765594482, + "learning_rate": 0.00019788728269430034, + "loss": 1.0754, + "step": 5395 + }, + { + "epoch": 0.3744361945735896, + "grad_norm": 2.7697651386260986, + "learning_rate": 0.00019788613370137973, + "loss": 1.0022, + "step": 5396 + }, + { + "epoch": 0.3745055860106863, + "grad_norm": 2.844947576522827, + "learning_rate": 0.00019788498439944394, + "loss": 0.8415, + "step": 5397 + }, + { + "epoch": 0.37457497744778295, + "grad_norm": 3.008411407470703, + "learning_rate": 0.00019788383478849662, + "loss": 1.0989, + "step": 5398 + }, + { + "epoch": 0.3746443688848796, + "grad_norm": 2.589552879333496, + "learning_rate": 0.00019788268486854138, + "loss": 0.9192, + "step": 5399 + }, + { + "epoch": 0.3747137603219763, + "grad_norm": 2.858274221420288, + "learning_rate": 0.00019788153463958186, + "loss": 0.836, + "step": 5400 + }, + { + "epoch": 0.37478315175907295, + "grad_norm": 2.7403275966644287, + "learning_rate": 0.00019788038410162168, + "loss": 0.8336, + "step": 5401 + }, + { + "epoch": 0.3748525431961696, + "grad_norm": 3.32743239402771, + "learning_rate": 0.00019787923325466448, + "loss": 0.9372, + "step": 5402 + }, + { + "epoch": 0.37492193463326623, + "grad_norm": 3.194114923477173, + "learning_rate": 0.00019787808209871391, + "loss": 0.7989, + "step": 5403 + }, + { + "epoch": 0.3749913260703629, + "grad_norm": 2.869297742843628, + "learning_rate": 0.00019787693063377362, + "loss": 0.9527, + "step": 5404 + }, + { + "epoch": 0.37506071750745956, + "grad_norm": 3.7891976833343506, + "learning_rate": 0.0001978757788598472, + "loss": 1.0968, + "step": 5405 + }, + { + "epoch": 0.37513010894455623, + "grad_norm": 3.3396687507629395, + "learning_rate": 0.00019787462677693828, + "loss": 0.9564, + "step": 5406 + }, + { + "epoch": 0.3751995003816529, + "grad_norm": 3.0246410369873047, + "learning_rate": 0.00019787347438505052, + "loss": 0.5986, + "step": 5407 + }, + { + "epoch": 0.37526889181874956, + "grad_norm": 2.770062208175659, + "learning_rate": 0.00019787232168418757, + "loss": 1.1189, + "step": 5408 + }, + { + "epoch": 0.37533828325584623, + "grad_norm": 3.2813572883605957, + "learning_rate": 0.00019787116867435307, + "loss": 1.0769, + "step": 5409 + }, + { + "epoch": 0.3754076746929429, + "grad_norm": 3.2233028411865234, + "learning_rate": 0.0001978700153555506, + "loss": 0.9033, + "step": 5410 + }, + { + "epoch": 0.37547706613003956, + "grad_norm": 2.7766289710998535, + "learning_rate": 0.0001978688617277839, + "loss": 0.933, + "step": 5411 + }, + { + "epoch": 0.3755464575671362, + "grad_norm": 2.865216016769409, + "learning_rate": 0.0001978677077910565, + "loss": 1.0388, + "step": 5412 + }, + { + "epoch": 0.3756158490042329, + "grad_norm": 2.7475688457489014, + "learning_rate": 0.00019786655354537216, + "loss": 1.004, + "step": 5413 + }, + { + "epoch": 0.37568524044132956, + "grad_norm": 2.2087135314941406, + "learning_rate": 0.00019786539899073444, + "loss": 0.8975, + "step": 5414 + }, + { + "epoch": 0.3757546318784262, + "grad_norm": 2.8710246086120605, + "learning_rate": 0.00019786424412714698, + "loss": 0.88, + "step": 5415 + }, + { + "epoch": 0.37582402331552284, + "grad_norm": 2.7032904624938965, + "learning_rate": 0.00019786308895461353, + "loss": 0.7721, + "step": 5416 + }, + { + "epoch": 0.3758934147526195, + "grad_norm": 2.399064064025879, + "learning_rate": 0.0001978619334731376, + "loss": 0.7674, + "step": 5417 + }, + { + "epoch": 0.37596280618971617, + "grad_norm": 2.879467487335205, + "learning_rate": 0.0001978607776827229, + "loss": 0.9169, + "step": 5418 + }, + { + "epoch": 0.37603219762681284, + "grad_norm": 3.8489160537719727, + "learning_rate": 0.0001978596215833731, + "loss": 0.8517, + "step": 5419 + }, + { + "epoch": 0.3761015890639095, + "grad_norm": 3.2837483882904053, + "learning_rate": 0.00019785846517509183, + "loss": 1.0566, + "step": 5420 + }, + { + "epoch": 0.37617098050100617, + "grad_norm": 2.628955602645874, + "learning_rate": 0.00019785730845788276, + "loss": 0.8306, + "step": 5421 + }, + { + "epoch": 0.37624037193810284, + "grad_norm": 4.107940196990967, + "learning_rate": 0.0001978561514317495, + "loss": 1.008, + "step": 5422 + }, + { + "epoch": 0.3763097633751995, + "grad_norm": 2.476074695587158, + "learning_rate": 0.0001978549940966957, + "loss": 0.9115, + "step": 5423 + }, + { + "epoch": 0.37637915481229617, + "grad_norm": 2.4539527893066406, + "learning_rate": 0.00019785383645272507, + "loss": 1.0011, + "step": 5424 + }, + { + "epoch": 0.37644854624939283, + "grad_norm": 2.7257020473480225, + "learning_rate": 0.00019785267849984118, + "loss": 1.09, + "step": 5425 + }, + { + "epoch": 0.3765179376864895, + "grad_norm": 3.115788221359253, + "learning_rate": 0.00019785152023804782, + "loss": 0.9307, + "step": 5426 + }, + { + "epoch": 0.37658732912358617, + "grad_norm": 8.012001991271973, + "learning_rate": 0.0001978503616673485, + "loss": 0.9977, + "step": 5427 + }, + { + "epoch": 0.37665672056068283, + "grad_norm": 3.070941686630249, + "learning_rate": 0.00019784920278774696, + "loss": 0.9027, + "step": 5428 + }, + { + "epoch": 0.3767261119977795, + "grad_norm": 3.6905205249786377, + "learning_rate": 0.00019784804359924683, + "loss": 1.1732, + "step": 5429 + }, + { + "epoch": 0.3767955034348761, + "grad_norm": 3.1009278297424316, + "learning_rate": 0.0001978468841018518, + "loss": 1.1983, + "step": 5430 + }, + { + "epoch": 0.3768648948719728, + "grad_norm": 3.481027603149414, + "learning_rate": 0.00019784572429556546, + "loss": 1.0107, + "step": 5431 + }, + { + "epoch": 0.37693428630906944, + "grad_norm": 2.8256592750549316, + "learning_rate": 0.00019784456418039156, + "loss": 1.0064, + "step": 5432 + }, + { + "epoch": 0.3770036777461661, + "grad_norm": 2.7237253189086914, + "learning_rate": 0.00019784340375633374, + "loss": 0.7045, + "step": 5433 + }, + { + "epoch": 0.3770730691832628, + "grad_norm": 2.6562132835388184, + "learning_rate": 0.00019784224302339562, + "loss": 0.9332, + "step": 5434 + }, + { + "epoch": 0.37714246062035944, + "grad_norm": 3.335298538208008, + "learning_rate": 0.0001978410819815809, + "loss": 1.1181, + "step": 5435 + }, + { + "epoch": 0.3772118520574561, + "grad_norm": 2.753937005996704, + "learning_rate": 0.00019783992063089322, + "loss": 0.9022, + "step": 5436 + }, + { + "epoch": 0.3772812434945528, + "grad_norm": 3.3998799324035645, + "learning_rate": 0.00019783875897133628, + "loss": 1.1244, + "step": 5437 + }, + { + "epoch": 0.37735063493164944, + "grad_norm": 3.2307419776916504, + "learning_rate": 0.00019783759700291374, + "loss": 0.9496, + "step": 5438 + }, + { + "epoch": 0.3774200263687461, + "grad_norm": 4.521292209625244, + "learning_rate": 0.00019783643472562923, + "loss": 0.9491, + "step": 5439 + }, + { + "epoch": 0.3774894178058428, + "grad_norm": 3.476381540298462, + "learning_rate": 0.00019783527213948645, + "loss": 1.0151, + "step": 5440 + }, + { + "epoch": 0.37755880924293944, + "grad_norm": 3.2868237495422363, + "learning_rate": 0.00019783410924448908, + "loss": 1.0816, + "step": 5441 + }, + { + "epoch": 0.3776282006800361, + "grad_norm": 2.615797758102417, + "learning_rate": 0.00019783294604064076, + "loss": 0.9181, + "step": 5442 + }, + { + "epoch": 0.3776975921171327, + "grad_norm": 2.5930075645446777, + "learning_rate": 0.0001978317825279452, + "loss": 0.963, + "step": 5443 + }, + { + "epoch": 0.3777669835542294, + "grad_norm": 2.785062551498413, + "learning_rate": 0.00019783061870640604, + "loss": 0.7964, + "step": 5444 + }, + { + "epoch": 0.37783637499132605, + "grad_norm": 3.6524555683135986, + "learning_rate": 0.00019782945457602697, + "loss": 1.1362, + "step": 5445 + }, + { + "epoch": 0.3779057664284227, + "grad_norm": 3.220154047012329, + "learning_rate": 0.00019782829013681167, + "loss": 0.9348, + "step": 5446 + }, + { + "epoch": 0.3779751578655194, + "grad_norm": 3.037207841873169, + "learning_rate": 0.0001978271253887638, + "loss": 1.0194, + "step": 5447 + }, + { + "epoch": 0.37804454930261605, + "grad_norm": 3.488417863845825, + "learning_rate": 0.00019782596033188706, + "loss": 1.3102, + "step": 5448 + }, + { + "epoch": 0.3781139407397127, + "grad_norm": 2.950747489929199, + "learning_rate": 0.0001978247949661851, + "loss": 0.8576, + "step": 5449 + }, + { + "epoch": 0.3781833321768094, + "grad_norm": 3.224788188934326, + "learning_rate": 0.0001978236292916616, + "loss": 0.9661, + "step": 5450 + }, + { + "epoch": 0.37825272361390605, + "grad_norm": 3.8905582427978516, + "learning_rate": 0.0001978224633083203, + "loss": 1.0374, + "step": 5451 + }, + { + "epoch": 0.3783221150510027, + "grad_norm": 3.697401523590088, + "learning_rate": 0.00019782129701616477, + "loss": 1.1807, + "step": 5452 + }, + { + "epoch": 0.3783915064880994, + "grad_norm": 2.9593067169189453, + "learning_rate": 0.00019782013041519877, + "loss": 0.9806, + "step": 5453 + }, + { + "epoch": 0.37846089792519605, + "grad_norm": 3.877951145172119, + "learning_rate": 0.000197818963505426, + "loss": 0.8588, + "step": 5454 + }, + { + "epoch": 0.3785302893622927, + "grad_norm": 4.361673355102539, + "learning_rate": 0.00019781779628685008, + "loss": 1.3462, + "step": 5455 + }, + { + "epoch": 0.3785996807993894, + "grad_norm": 2.979909658432007, + "learning_rate": 0.00019781662875947476, + "loss": 1.0957, + "step": 5456 + }, + { + "epoch": 0.378669072236486, + "grad_norm": 2.916133165359497, + "learning_rate": 0.00019781546092330365, + "loss": 0.8694, + "step": 5457 + }, + { + "epoch": 0.37873846367358266, + "grad_norm": 2.877120018005371, + "learning_rate": 0.00019781429277834049, + "loss": 0.9232, + "step": 5458 + }, + { + "epoch": 0.3788078551106793, + "grad_norm": 3.151179552078247, + "learning_rate": 0.00019781312432458895, + "loss": 1.1106, + "step": 5459 + }, + { + "epoch": 0.378877246547776, + "grad_norm": 3.217268228530884, + "learning_rate": 0.00019781195556205276, + "loss": 0.7455, + "step": 5460 + }, + { + "epoch": 0.37894663798487266, + "grad_norm": 2.923732280731201, + "learning_rate": 0.00019781078649073555, + "loss": 0.7328, + "step": 5461 + }, + { + "epoch": 0.3790160294219693, + "grad_norm": 2.765397787094116, + "learning_rate": 0.00019780961711064103, + "loss": 1.1162, + "step": 5462 + }, + { + "epoch": 0.379085420859066, + "grad_norm": 3.518115758895874, + "learning_rate": 0.00019780844742177292, + "loss": 1.0385, + "step": 5463 + }, + { + "epoch": 0.37915481229616266, + "grad_norm": 2.9139938354492188, + "learning_rate": 0.0001978072774241349, + "loss": 0.9124, + "step": 5464 + }, + { + "epoch": 0.3792242037332593, + "grad_norm": 2.568963050842285, + "learning_rate": 0.00019780610711773062, + "loss": 0.7895, + "step": 5465 + }, + { + "epoch": 0.379293595170356, + "grad_norm": 3.422546148300171, + "learning_rate": 0.00019780493650256383, + "loss": 0.9184, + "step": 5466 + }, + { + "epoch": 0.37936298660745266, + "grad_norm": 3.2120113372802734, + "learning_rate": 0.00019780376557863817, + "loss": 1.1018, + "step": 5467 + }, + { + "epoch": 0.3794323780445493, + "grad_norm": 4.020591735839844, + "learning_rate": 0.0001978025943459574, + "loss": 1.0718, + "step": 5468 + }, + { + "epoch": 0.379501769481646, + "grad_norm": 2.8228979110717773, + "learning_rate": 0.00019780142280452517, + "loss": 0.9681, + "step": 5469 + }, + { + "epoch": 0.3795711609187426, + "grad_norm": 2.57670259475708, + "learning_rate": 0.0001978002509543452, + "loss": 0.7985, + "step": 5470 + }, + { + "epoch": 0.37964055235583927, + "grad_norm": 2.2659759521484375, + "learning_rate": 0.00019779907879542116, + "loss": 0.825, + "step": 5471 + }, + { + "epoch": 0.37970994379293593, + "grad_norm": 2.7515902519226074, + "learning_rate": 0.0001977979063277568, + "loss": 0.8586, + "step": 5472 + }, + { + "epoch": 0.3797793352300326, + "grad_norm": 3.478372812271118, + "learning_rate": 0.00019779673355135578, + "loss": 0.9648, + "step": 5473 + }, + { + "epoch": 0.37984872666712927, + "grad_norm": 2.286421060562134, + "learning_rate": 0.00019779556046622187, + "loss": 0.8888, + "step": 5474 + }, + { + "epoch": 0.37991811810422593, + "grad_norm": 3.3964548110961914, + "learning_rate": 0.0001977943870723587, + "loss": 0.9316, + "step": 5475 + }, + { + "epoch": 0.3799875095413226, + "grad_norm": 2.09287166595459, + "learning_rate": 0.00019779321336976997, + "loss": 0.7584, + "step": 5476 + }, + { + "epoch": 0.38005690097841927, + "grad_norm": 2.8696749210357666, + "learning_rate": 0.0001977920393584594, + "loss": 1.1531, + "step": 5477 + }, + { + "epoch": 0.38012629241551593, + "grad_norm": 3.097820997238159, + "learning_rate": 0.00019779086503843077, + "loss": 1.0819, + "step": 5478 + }, + { + "epoch": 0.3801956838526126, + "grad_norm": 4.565545082092285, + "learning_rate": 0.00019778969040968767, + "loss": 1.4199, + "step": 5479 + }, + { + "epoch": 0.38026507528970926, + "grad_norm": 3.184081792831421, + "learning_rate": 0.00019778851547223386, + "loss": 1.2218, + "step": 5480 + }, + { + "epoch": 0.38033446672680593, + "grad_norm": 2.7271671295166016, + "learning_rate": 0.00019778734022607309, + "loss": 0.8309, + "step": 5481 + }, + { + "epoch": 0.3804038581639026, + "grad_norm": 3.1248395442962646, + "learning_rate": 0.000197786164671209, + "loss": 1.0822, + "step": 5482 + }, + { + "epoch": 0.38047324960099926, + "grad_norm": 2.446047067642212, + "learning_rate": 0.00019778498880764538, + "loss": 0.7477, + "step": 5483 + }, + { + "epoch": 0.3805426410380959, + "grad_norm": 3.073068141937256, + "learning_rate": 0.00019778381263538584, + "loss": 0.9448, + "step": 5484 + }, + { + "epoch": 0.38061203247519254, + "grad_norm": 2.9032368659973145, + "learning_rate": 0.00019778263615443417, + "loss": 1.0433, + "step": 5485 + }, + { + "epoch": 0.3806814239122892, + "grad_norm": 2.355742931365967, + "learning_rate": 0.00019778145936479407, + "loss": 0.7584, + "step": 5486 + }, + { + "epoch": 0.3807508153493859, + "grad_norm": 2.55479097366333, + "learning_rate": 0.00019778028226646925, + "loss": 0.8287, + "step": 5487 + }, + { + "epoch": 0.38082020678648254, + "grad_norm": 2.908496379852295, + "learning_rate": 0.0001977791048594634, + "loss": 0.9324, + "step": 5488 + }, + { + "epoch": 0.3808895982235792, + "grad_norm": 3.5210509300231934, + "learning_rate": 0.0001977779271437803, + "loss": 0.9838, + "step": 5489 + }, + { + "epoch": 0.3809589896606759, + "grad_norm": 3.7827653884887695, + "learning_rate": 0.00019777674911942362, + "loss": 1.1326, + "step": 5490 + }, + { + "epoch": 0.38102838109777254, + "grad_norm": 3.843083620071411, + "learning_rate": 0.00019777557078639704, + "loss": 1.1431, + "step": 5491 + }, + { + "epoch": 0.3810977725348692, + "grad_norm": 2.7591469287872314, + "learning_rate": 0.00019777439214470436, + "loss": 0.8296, + "step": 5492 + }, + { + "epoch": 0.38116716397196587, + "grad_norm": 3.185314655303955, + "learning_rate": 0.00019777321319434926, + "loss": 0.8253, + "step": 5493 + }, + { + "epoch": 0.38123655540906254, + "grad_norm": 2.27020263671875, + "learning_rate": 0.0001977720339353355, + "loss": 0.7869, + "step": 5494 + }, + { + "epoch": 0.3813059468461592, + "grad_norm": 3.226301431655884, + "learning_rate": 0.00019777085436766673, + "loss": 0.9514, + "step": 5495 + }, + { + "epoch": 0.38137533828325587, + "grad_norm": 2.8729264736175537, + "learning_rate": 0.00019776967449134674, + "loss": 1.084, + "step": 5496 + }, + { + "epoch": 0.3814447297203525, + "grad_norm": 3.4336204528808594, + "learning_rate": 0.00019776849430637922, + "loss": 1.0721, + "step": 5497 + }, + { + "epoch": 0.38151412115744915, + "grad_norm": 3.2202091217041016, + "learning_rate": 0.00019776731381276793, + "loss": 1.0667, + "step": 5498 + }, + { + "epoch": 0.3815835125945458, + "grad_norm": 2.835336685180664, + "learning_rate": 0.00019776613301051654, + "loss": 1.2133, + "step": 5499 + }, + { + "epoch": 0.3816529040316425, + "grad_norm": 2.7105441093444824, + "learning_rate": 0.00019776495189962882, + "loss": 0.9262, + "step": 5500 + }, + { + "epoch": 0.38172229546873915, + "grad_norm": 3.1525561809539795, + "learning_rate": 0.00019776377048010853, + "loss": 1.0884, + "step": 5501 + }, + { + "epoch": 0.3817916869058358, + "grad_norm": 3.125659227371216, + "learning_rate": 0.00019776258875195933, + "loss": 0.8585, + "step": 5502 + }, + { + "epoch": 0.3818610783429325, + "grad_norm": 2.37534499168396, + "learning_rate": 0.00019776140671518496, + "loss": 0.8768, + "step": 5503 + }, + { + "epoch": 0.38193046978002915, + "grad_norm": 3.171870231628418, + "learning_rate": 0.0001977602243697892, + "loss": 0.8227, + "step": 5504 + }, + { + "epoch": 0.3819998612171258, + "grad_norm": 2.6560449600219727, + "learning_rate": 0.00019775904171577573, + "loss": 1.0204, + "step": 5505 + }, + { + "epoch": 0.3820692526542225, + "grad_norm": 4.542102336883545, + "learning_rate": 0.00019775785875314833, + "loss": 1.2866, + "step": 5506 + }, + { + "epoch": 0.38213864409131915, + "grad_norm": 3.031064987182617, + "learning_rate": 0.0001977566754819107, + "loss": 1.0766, + "step": 5507 + }, + { + "epoch": 0.3822080355284158, + "grad_norm": 2.4290578365325928, + "learning_rate": 0.00019775549190206658, + "loss": 0.8996, + "step": 5508 + }, + { + "epoch": 0.3822774269655125, + "grad_norm": 2.902376651763916, + "learning_rate": 0.00019775430801361973, + "loss": 0.8277, + "step": 5509 + }, + { + "epoch": 0.38234681840260915, + "grad_norm": 2.6004533767700195, + "learning_rate": 0.00019775312381657386, + "loss": 0.7276, + "step": 5510 + }, + { + "epoch": 0.38241620983970576, + "grad_norm": 2.5407462120056152, + "learning_rate": 0.00019775193931093272, + "loss": 0.9068, + "step": 5511 + }, + { + "epoch": 0.3824856012768024, + "grad_norm": 2.956570863723755, + "learning_rate": 0.00019775075449670007, + "loss": 1.0658, + "step": 5512 + }, + { + "epoch": 0.3825549927138991, + "grad_norm": 2.5890419483184814, + "learning_rate": 0.0001977495693738796, + "loss": 1.0445, + "step": 5513 + }, + { + "epoch": 0.38262438415099576, + "grad_norm": 3.499316453933716, + "learning_rate": 0.0001977483839424751, + "loss": 1.2975, + "step": 5514 + }, + { + "epoch": 0.3826937755880924, + "grad_norm": 3.1917524337768555, + "learning_rate": 0.00019774719820249027, + "loss": 0.8582, + "step": 5515 + }, + { + "epoch": 0.3827631670251891, + "grad_norm": 3.707989454269409, + "learning_rate": 0.0001977460121539289, + "loss": 1.0305, + "step": 5516 + }, + { + "epoch": 0.38283255846228575, + "grad_norm": 2.6668195724487305, + "learning_rate": 0.00019774482579679468, + "loss": 1.0625, + "step": 5517 + }, + { + "epoch": 0.3829019498993824, + "grad_norm": 2.5903730392456055, + "learning_rate": 0.00019774363913109143, + "loss": 0.9948, + "step": 5518 + }, + { + "epoch": 0.3829713413364791, + "grad_norm": 2.7596919536590576, + "learning_rate": 0.00019774245215682283, + "loss": 1.019, + "step": 5519 + }, + { + "epoch": 0.38304073277357575, + "grad_norm": 2.6664679050445557, + "learning_rate": 0.00019774126487399264, + "loss": 0.8622, + "step": 5520 + }, + { + "epoch": 0.3831101242106724, + "grad_norm": 3.1023013591766357, + "learning_rate": 0.00019774007728260462, + "loss": 0.997, + "step": 5521 + }, + { + "epoch": 0.3831795156477691, + "grad_norm": 2.796476125717163, + "learning_rate": 0.00019773888938266252, + "loss": 0.8891, + "step": 5522 + }, + { + "epoch": 0.38324890708486575, + "grad_norm": 3.303010940551758, + "learning_rate": 0.00019773770117417007, + "loss": 0.7814, + "step": 5523 + }, + { + "epoch": 0.38331829852196236, + "grad_norm": 3.8279712200164795, + "learning_rate": 0.00019773651265713108, + "loss": 1.3096, + "step": 5524 + }, + { + "epoch": 0.38338768995905903, + "grad_norm": 3.0202906131744385, + "learning_rate": 0.0001977353238315492, + "loss": 1.0155, + "step": 5525 + }, + { + "epoch": 0.3834570813961557, + "grad_norm": 2.497798442840576, + "learning_rate": 0.0001977341346974283, + "loss": 0.6795, + "step": 5526 + }, + { + "epoch": 0.38352647283325236, + "grad_norm": 3.4261672496795654, + "learning_rate": 0.00019773294525477205, + "loss": 1.2417, + "step": 5527 + }, + { + "epoch": 0.38359586427034903, + "grad_norm": 2.3652782440185547, + "learning_rate": 0.00019773175550358423, + "loss": 0.8823, + "step": 5528 + }, + { + "epoch": 0.3836652557074457, + "grad_norm": 2.6451549530029297, + "learning_rate": 0.00019773056544386859, + "loss": 0.9032, + "step": 5529 + }, + { + "epoch": 0.38373464714454236, + "grad_norm": 2.809701442718506, + "learning_rate": 0.00019772937507562893, + "loss": 0.9378, + "step": 5530 + }, + { + "epoch": 0.38380403858163903, + "grad_norm": 2.8242921829223633, + "learning_rate": 0.00019772818439886894, + "loss": 0.9122, + "step": 5531 + }, + { + "epoch": 0.3838734300187357, + "grad_norm": 3.041466474533081, + "learning_rate": 0.0001977269934135924, + "loss": 0.7002, + "step": 5532 + }, + { + "epoch": 0.38394282145583236, + "grad_norm": 2.7642173767089844, + "learning_rate": 0.00019772580211980308, + "loss": 0.9585, + "step": 5533 + }, + { + "epoch": 0.38401221289292903, + "grad_norm": 3.0687787532806396, + "learning_rate": 0.00019772461051750475, + "loss": 1.0517, + "step": 5534 + }, + { + "epoch": 0.3840816043300257, + "grad_norm": 2.9163265228271484, + "learning_rate": 0.00019772341860670119, + "loss": 1.0071, + "step": 5535 + }, + { + "epoch": 0.38415099576712236, + "grad_norm": 3.0039010047912598, + "learning_rate": 0.0001977222263873961, + "loss": 0.865, + "step": 5536 + }, + { + "epoch": 0.38422038720421897, + "grad_norm": 3.1879615783691406, + "learning_rate": 0.0001977210338595933, + "loss": 0.8982, + "step": 5537 + }, + { + "epoch": 0.38428977864131564, + "grad_norm": 2.86039400100708, + "learning_rate": 0.0001977198410232965, + "loss": 0.7638, + "step": 5538 + }, + { + "epoch": 0.3843591700784123, + "grad_norm": 2.973573684692383, + "learning_rate": 0.00019771864787850954, + "loss": 0.8112, + "step": 5539 + }, + { + "epoch": 0.38442856151550897, + "grad_norm": 3.4897377490997314, + "learning_rate": 0.00019771745442523613, + "loss": 1.1281, + "step": 5540 + }, + { + "epoch": 0.38449795295260564, + "grad_norm": 3.3784162998199463, + "learning_rate": 0.00019771626066348005, + "loss": 0.8452, + "step": 5541 + }, + { + "epoch": 0.3845673443897023, + "grad_norm": 2.892375946044922, + "learning_rate": 0.0001977150665932451, + "loss": 1.0932, + "step": 5542 + }, + { + "epoch": 0.38463673582679897, + "grad_norm": 3.1019082069396973, + "learning_rate": 0.000197713872214535, + "loss": 0.8793, + "step": 5543 + }, + { + "epoch": 0.38470612726389564, + "grad_norm": 2.717496871948242, + "learning_rate": 0.00019771267752735354, + "loss": 1.0098, + "step": 5544 + }, + { + "epoch": 0.3847755187009923, + "grad_norm": 4.154601573944092, + "learning_rate": 0.00019771148253170447, + "loss": 0.8929, + "step": 5545 + }, + { + "epoch": 0.38484491013808897, + "grad_norm": 2.658743381500244, + "learning_rate": 0.00019771028722759165, + "loss": 0.8937, + "step": 5546 + }, + { + "epoch": 0.38491430157518564, + "grad_norm": 2.7102479934692383, + "learning_rate": 0.00019770909161501873, + "loss": 0.894, + "step": 5547 + }, + { + "epoch": 0.3849836930122823, + "grad_norm": 6.090872287750244, + "learning_rate": 0.00019770789569398955, + "loss": 1.1984, + "step": 5548 + }, + { + "epoch": 0.38505308444937897, + "grad_norm": 4.146792411804199, + "learning_rate": 0.0001977066994645079, + "loss": 0.8993, + "step": 5549 + }, + { + "epoch": 0.38512247588647563, + "grad_norm": 3.199549674987793, + "learning_rate": 0.00019770550292657756, + "loss": 1.012, + "step": 5550 + }, + { + "epoch": 0.38519186732357225, + "grad_norm": 3.003122329711914, + "learning_rate": 0.00019770430608020222, + "loss": 0.8889, + "step": 5551 + }, + { + "epoch": 0.3852612587606689, + "grad_norm": 2.772876024246216, + "learning_rate": 0.00019770310892538577, + "loss": 0.7209, + "step": 5552 + }, + { + "epoch": 0.3853306501977656, + "grad_norm": 4.404020309448242, + "learning_rate": 0.00019770191146213192, + "loss": 1.1747, + "step": 5553 + }, + { + "epoch": 0.38540004163486224, + "grad_norm": 2.9654366970062256, + "learning_rate": 0.00019770071369044448, + "loss": 0.9808, + "step": 5554 + }, + { + "epoch": 0.3854694330719589, + "grad_norm": 3.8640358448028564, + "learning_rate": 0.0001976995156103272, + "loss": 1.1832, + "step": 5555 + }, + { + "epoch": 0.3855388245090556, + "grad_norm": 2.96543550491333, + "learning_rate": 0.00019769831722178395, + "loss": 1.109, + "step": 5556 + }, + { + "epoch": 0.38560821594615224, + "grad_norm": 3.4822583198547363, + "learning_rate": 0.0001976971185248184, + "loss": 0.9833, + "step": 5557 + }, + { + "epoch": 0.3856776073832489, + "grad_norm": 3.0976202487945557, + "learning_rate": 0.00019769591951943437, + "loss": 0.959, + "step": 5558 + }, + { + "epoch": 0.3857469988203456, + "grad_norm": 4.47316837310791, + "learning_rate": 0.00019769472020563566, + "loss": 1.1362, + "step": 5559 + }, + { + "epoch": 0.38581639025744224, + "grad_norm": 3.603473424911499, + "learning_rate": 0.00019769352058342606, + "loss": 0.834, + "step": 5560 + }, + { + "epoch": 0.3858857816945389, + "grad_norm": 5.240043640136719, + "learning_rate": 0.00019769232065280936, + "loss": 0.9913, + "step": 5561 + }, + { + "epoch": 0.3859551731316356, + "grad_norm": 2.8285012245178223, + "learning_rate": 0.00019769112041378932, + "loss": 1.0302, + "step": 5562 + }, + { + "epoch": 0.38602456456873224, + "grad_norm": 2.7988967895507812, + "learning_rate": 0.00019768991986636979, + "loss": 1.0305, + "step": 5563 + }, + { + "epoch": 0.38609395600582885, + "grad_norm": 3.365145206451416, + "learning_rate": 0.00019768871901055447, + "loss": 1.0253, + "step": 5564 + }, + { + "epoch": 0.3861633474429255, + "grad_norm": 3.5552921295166016, + "learning_rate": 0.00019768751784634723, + "loss": 1.1178, + "step": 5565 + }, + { + "epoch": 0.3862327388800222, + "grad_norm": 3.3712642192840576, + "learning_rate": 0.0001976863163737518, + "loss": 1.1187, + "step": 5566 + }, + { + "epoch": 0.38630213031711885, + "grad_norm": 3.465653657913208, + "learning_rate": 0.000197685114592772, + "loss": 0.8742, + "step": 5567 + }, + { + "epoch": 0.3863715217542155, + "grad_norm": 2.970733404159546, + "learning_rate": 0.00019768391250341166, + "loss": 1.0904, + "step": 5568 + }, + { + "epoch": 0.3864409131913122, + "grad_norm": 3.5431995391845703, + "learning_rate": 0.00019768271010567453, + "loss": 0.8853, + "step": 5569 + }, + { + "epoch": 0.38651030462840885, + "grad_norm": 2.6014654636383057, + "learning_rate": 0.0001976815073995644, + "loss": 0.9123, + "step": 5570 + }, + { + "epoch": 0.3865796960655055, + "grad_norm": 2.8281853199005127, + "learning_rate": 0.0001976803043850851, + "loss": 1.0273, + "step": 5571 + }, + { + "epoch": 0.3866490875026022, + "grad_norm": 2.566046953201294, + "learning_rate": 0.00019767910106224042, + "loss": 0.9012, + "step": 5572 + }, + { + "epoch": 0.38671847893969885, + "grad_norm": 3.039090633392334, + "learning_rate": 0.00019767789743103413, + "loss": 0.9674, + "step": 5573 + }, + { + "epoch": 0.3867878703767955, + "grad_norm": 2.9424140453338623, + "learning_rate": 0.00019767669349147007, + "loss": 0.9393, + "step": 5574 + }, + { + "epoch": 0.3868572618138922, + "grad_norm": 3.1172516345977783, + "learning_rate": 0.00019767548924355202, + "loss": 0.8526, + "step": 5575 + }, + { + "epoch": 0.38692665325098885, + "grad_norm": 3.192213296890259, + "learning_rate": 0.00019767428468728376, + "loss": 1.1474, + "step": 5576 + }, + { + "epoch": 0.3869960446880855, + "grad_norm": 3.3208587169647217, + "learning_rate": 0.00019767307982266913, + "loss": 0.8612, + "step": 5577 + }, + { + "epoch": 0.3870654361251821, + "grad_norm": 3.0142412185668945, + "learning_rate": 0.00019767187464971194, + "loss": 1.1248, + "step": 5578 + }, + { + "epoch": 0.3871348275622788, + "grad_norm": 3.1061179637908936, + "learning_rate": 0.00019767066916841593, + "loss": 0.8862, + "step": 5579 + }, + { + "epoch": 0.38720421899937546, + "grad_norm": 3.213437557220459, + "learning_rate": 0.00019766946337878498, + "loss": 1.0605, + "step": 5580 + }, + { + "epoch": 0.3872736104364721, + "grad_norm": 3.8500964641571045, + "learning_rate": 0.00019766825728082289, + "loss": 0.8634, + "step": 5581 + }, + { + "epoch": 0.3873430018735688, + "grad_norm": 3.032412528991699, + "learning_rate": 0.00019766705087453338, + "loss": 0.9212, + "step": 5582 + }, + { + "epoch": 0.38741239331066546, + "grad_norm": 3.289214849472046, + "learning_rate": 0.00019766584415992035, + "loss": 0.9102, + "step": 5583 + }, + { + "epoch": 0.3874817847477621, + "grad_norm": 3.0750982761383057, + "learning_rate": 0.0001976646371369876, + "loss": 0.7959, + "step": 5584 + }, + { + "epoch": 0.3875511761848588, + "grad_norm": 3.667062759399414, + "learning_rate": 0.0001976634298057389, + "loss": 1.0822, + "step": 5585 + }, + { + "epoch": 0.38762056762195546, + "grad_norm": 3.6381430625915527, + "learning_rate": 0.0001976622221661781, + "loss": 0.9756, + "step": 5586 + }, + { + "epoch": 0.3876899590590521, + "grad_norm": 2.535292863845825, + "learning_rate": 0.00019766101421830897, + "loss": 0.8776, + "step": 5587 + }, + { + "epoch": 0.3877593504961488, + "grad_norm": 2.8382601737976074, + "learning_rate": 0.00019765980596213537, + "loss": 0.8119, + "step": 5588 + }, + { + "epoch": 0.38782874193324546, + "grad_norm": 3.8564465045928955, + "learning_rate": 0.0001976585973976611, + "loss": 1.0462, + "step": 5589 + }, + { + "epoch": 0.3878981333703421, + "grad_norm": 2.5931496620178223, + "learning_rate": 0.00019765738852488992, + "loss": 0.6927, + "step": 5590 + }, + { + "epoch": 0.38796752480743874, + "grad_norm": 3.180088996887207, + "learning_rate": 0.00019765617934382575, + "loss": 0.7714, + "step": 5591 + }, + { + "epoch": 0.3880369162445354, + "grad_norm": 3.543596029281616, + "learning_rate": 0.00019765496985447233, + "loss": 0.8118, + "step": 5592 + }, + { + "epoch": 0.38810630768163207, + "grad_norm": 2.892176389694214, + "learning_rate": 0.00019765376005683352, + "loss": 0.8386, + "step": 5593 + }, + { + "epoch": 0.38817569911872873, + "grad_norm": 4.587162494659424, + "learning_rate": 0.00019765254995091308, + "loss": 0.9762, + "step": 5594 + }, + { + "epoch": 0.3882450905558254, + "grad_norm": 2.974163293838501, + "learning_rate": 0.0001976513395367149, + "loss": 0.9704, + "step": 5595 + }, + { + "epoch": 0.38831448199292207, + "grad_norm": 2.874701499938965, + "learning_rate": 0.00019765012881424275, + "loss": 0.7475, + "step": 5596 + }, + { + "epoch": 0.38838387343001873, + "grad_norm": 4.0696702003479, + "learning_rate": 0.00019764891778350049, + "loss": 0.9278, + "step": 5597 + }, + { + "epoch": 0.3884532648671154, + "grad_norm": 2.4984076023101807, + "learning_rate": 0.00019764770644449192, + "loss": 0.9261, + "step": 5598 + }, + { + "epoch": 0.38852265630421207, + "grad_norm": 2.7259464263916016, + "learning_rate": 0.00019764649479722088, + "loss": 1.1984, + "step": 5599 + }, + { + "epoch": 0.38859204774130873, + "grad_norm": 2.9382002353668213, + "learning_rate": 0.00019764528284169116, + "loss": 0.8752, + "step": 5600 + }, + { + "epoch": 0.3886614391784054, + "grad_norm": 2.322333335876465, + "learning_rate": 0.00019764407057790666, + "loss": 0.8492, + "step": 5601 + }, + { + "epoch": 0.38873083061550207, + "grad_norm": 2.9568495750427246, + "learning_rate": 0.00019764285800587112, + "loss": 1.0541, + "step": 5602 + }, + { + "epoch": 0.38880022205259873, + "grad_norm": 2.87218976020813, + "learning_rate": 0.00019764164512558842, + "loss": 0.812, + "step": 5603 + }, + { + "epoch": 0.3888696134896954, + "grad_norm": 2.6040730476379395, + "learning_rate": 0.00019764043193706236, + "loss": 0.9219, + "step": 5604 + }, + { + "epoch": 0.388939004926792, + "grad_norm": 2.8543474674224854, + "learning_rate": 0.00019763921844029677, + "loss": 1.0131, + "step": 5605 + }, + { + "epoch": 0.3890083963638887, + "grad_norm": 2.0552797317504883, + "learning_rate": 0.00019763800463529555, + "loss": 0.7375, + "step": 5606 + }, + { + "epoch": 0.38907778780098534, + "grad_norm": 3.3464839458465576, + "learning_rate": 0.00019763679052206242, + "loss": 0.9303, + "step": 5607 + }, + { + "epoch": 0.389147179238082, + "grad_norm": 2.8932175636291504, + "learning_rate": 0.0001976355761006013, + "loss": 0.8007, + "step": 5608 + }, + { + "epoch": 0.3892165706751787, + "grad_norm": 2.4276375770568848, + "learning_rate": 0.000197634361370916, + "loss": 0.9001, + "step": 5609 + }, + { + "epoch": 0.38928596211227534, + "grad_norm": 3.077834129333496, + "learning_rate": 0.00019763314633301034, + "loss": 1.0255, + "step": 5610 + }, + { + "epoch": 0.389355353549372, + "grad_norm": 2.5148258209228516, + "learning_rate": 0.00019763193098688815, + "loss": 0.7788, + "step": 5611 + }, + { + "epoch": 0.3894247449864687, + "grad_norm": 2.7122955322265625, + "learning_rate": 0.0001976307153325533, + "loss": 0.8342, + "step": 5612 + }, + { + "epoch": 0.38949413642356534, + "grad_norm": 2.970986843109131, + "learning_rate": 0.0001976294993700096, + "loss": 1.0376, + "step": 5613 + }, + { + "epoch": 0.389563527860662, + "grad_norm": 3.2118282318115234, + "learning_rate": 0.0001976282830992609, + "loss": 0.975, + "step": 5614 + }, + { + "epoch": 0.3896329192977587, + "grad_norm": 3.356868267059326, + "learning_rate": 0.00019762706652031104, + "loss": 0.9357, + "step": 5615 + }, + { + "epoch": 0.38970231073485534, + "grad_norm": 3.366473913192749, + "learning_rate": 0.00019762584963316387, + "loss": 1.0073, + "step": 5616 + }, + { + "epoch": 0.389771702171952, + "grad_norm": 2.356729030609131, + "learning_rate": 0.0001976246324378232, + "loss": 0.8796, + "step": 5617 + }, + { + "epoch": 0.3898410936090486, + "grad_norm": 2.920898675918579, + "learning_rate": 0.00019762341493429289, + "loss": 0.9467, + "step": 5618 + }, + { + "epoch": 0.3899104850461453, + "grad_norm": 2.8586535453796387, + "learning_rate": 0.00019762219712257678, + "loss": 0.9134, + "step": 5619 + }, + { + "epoch": 0.38997987648324195, + "grad_norm": 3.4171481132507324, + "learning_rate": 0.00019762097900267875, + "loss": 0.9966, + "step": 5620 + }, + { + "epoch": 0.3900492679203386, + "grad_norm": 2.5736920833587646, + "learning_rate": 0.00019761976057460257, + "loss": 1.0572, + "step": 5621 + }, + { + "epoch": 0.3901186593574353, + "grad_norm": 2.729675769805908, + "learning_rate": 0.00019761854183835217, + "loss": 0.9056, + "step": 5622 + }, + { + "epoch": 0.39018805079453195, + "grad_norm": 3.69140887260437, + "learning_rate": 0.00019761732279393134, + "loss": 0.9061, + "step": 5623 + }, + { + "epoch": 0.3902574422316286, + "grad_norm": 3.5118377208709717, + "learning_rate": 0.00019761610344134396, + "loss": 1.0214, + "step": 5624 + }, + { + "epoch": 0.3903268336687253, + "grad_norm": 3.421123743057251, + "learning_rate": 0.00019761488378059388, + "loss": 1.0544, + "step": 5625 + }, + { + "epoch": 0.39039622510582195, + "grad_norm": 3.5356674194335938, + "learning_rate": 0.00019761366381168492, + "loss": 0.9799, + "step": 5626 + }, + { + "epoch": 0.3904656165429186, + "grad_norm": 2.238673686981201, + "learning_rate": 0.00019761244353462093, + "loss": 0.9674, + "step": 5627 + }, + { + "epoch": 0.3905350079800153, + "grad_norm": 2.744825839996338, + "learning_rate": 0.0001976112229494058, + "loss": 0.7607, + "step": 5628 + }, + { + "epoch": 0.39060439941711195, + "grad_norm": 3.488657236099243, + "learning_rate": 0.00019761000205604335, + "loss": 1.1155, + "step": 5629 + }, + { + "epoch": 0.3906737908542086, + "grad_norm": 2.824197769165039, + "learning_rate": 0.0001976087808545375, + "loss": 0.79, + "step": 5630 + }, + { + "epoch": 0.3907431822913053, + "grad_norm": 4.067795753479004, + "learning_rate": 0.00019760755934489203, + "loss": 1.1361, + "step": 5631 + }, + { + "epoch": 0.3908125737284019, + "grad_norm": 3.0717735290527344, + "learning_rate": 0.0001976063375271108, + "loss": 1.1184, + "step": 5632 + }, + { + "epoch": 0.39088196516549856, + "grad_norm": 2.4113574028015137, + "learning_rate": 0.00019760511540119766, + "loss": 0.8504, + "step": 5633 + }, + { + "epoch": 0.3909513566025952, + "grad_norm": 2.4392757415771484, + "learning_rate": 0.00019760389296715657, + "loss": 0.7184, + "step": 5634 + }, + { + "epoch": 0.3910207480396919, + "grad_norm": 2.9418323040008545, + "learning_rate": 0.00019760267022499124, + "loss": 1.1853, + "step": 5635 + }, + { + "epoch": 0.39109013947678856, + "grad_norm": 3.4587011337280273, + "learning_rate": 0.00019760144717470562, + "loss": 1.0082, + "step": 5636 + }, + { + "epoch": 0.3911595309138852, + "grad_norm": 3.0125811100006104, + "learning_rate": 0.00019760022381630358, + "loss": 1.0792, + "step": 5637 + }, + { + "epoch": 0.3912289223509819, + "grad_norm": 2.5687873363494873, + "learning_rate": 0.00019759900014978897, + "loss": 1.144, + "step": 5638 + }, + { + "epoch": 0.39129831378807856, + "grad_norm": 1.8794277906417847, + "learning_rate": 0.0001975977761751656, + "loss": 0.6877, + "step": 5639 + }, + { + "epoch": 0.3913677052251752, + "grad_norm": 2.8147315979003906, + "learning_rate": 0.0001975965518924374, + "loss": 1.1603, + "step": 5640 + }, + { + "epoch": 0.3914370966622719, + "grad_norm": 2.523056983947754, + "learning_rate": 0.0001975953273016082, + "loss": 0.9961, + "step": 5641 + }, + { + "epoch": 0.39150648809936855, + "grad_norm": 3.9184048175811768, + "learning_rate": 0.00019759410240268187, + "loss": 1.0554, + "step": 5642 + }, + { + "epoch": 0.3915758795364652, + "grad_norm": 3.7920727729797363, + "learning_rate": 0.0001975928771956623, + "loss": 1.266, + "step": 5643 + }, + { + "epoch": 0.3916452709735619, + "grad_norm": 2.423632860183716, + "learning_rate": 0.00019759165168055332, + "loss": 0.8131, + "step": 5644 + }, + { + "epoch": 0.3917146624106585, + "grad_norm": 2.1850407123565674, + "learning_rate": 0.00019759042585735883, + "loss": 0.8062, + "step": 5645 + }, + { + "epoch": 0.39178405384775516, + "grad_norm": 2.5818965435028076, + "learning_rate": 0.00019758919972608267, + "loss": 1.2283, + "step": 5646 + }, + { + "epoch": 0.39185344528485183, + "grad_norm": 3.708627939224243, + "learning_rate": 0.00019758797328672876, + "loss": 1.0749, + "step": 5647 + }, + { + "epoch": 0.3919228367219485, + "grad_norm": 3.279130697250366, + "learning_rate": 0.00019758674653930092, + "loss": 0.851, + "step": 5648 + }, + { + "epoch": 0.39199222815904516, + "grad_norm": 2.5842881202697754, + "learning_rate": 0.00019758551948380305, + "loss": 0.9883, + "step": 5649 + }, + { + "epoch": 0.39206161959614183, + "grad_norm": 2.6186015605926514, + "learning_rate": 0.00019758429212023902, + "loss": 0.9269, + "step": 5650 + }, + { + "epoch": 0.3921310110332385, + "grad_norm": 3.03865647315979, + "learning_rate": 0.0001975830644486127, + "loss": 1.0283, + "step": 5651 + }, + { + "epoch": 0.39220040247033516, + "grad_norm": 3.699934720993042, + "learning_rate": 0.00019758183646892793, + "loss": 1.0434, + "step": 5652 + }, + { + "epoch": 0.39226979390743183, + "grad_norm": 2.136809825897217, + "learning_rate": 0.00019758060818118867, + "loss": 1.0888, + "step": 5653 + }, + { + "epoch": 0.3923391853445285, + "grad_norm": 3.7006876468658447, + "learning_rate": 0.00019757937958539873, + "loss": 1.1323, + "step": 5654 + }, + { + "epoch": 0.39240857678162516, + "grad_norm": 3.001185417175293, + "learning_rate": 0.00019757815068156202, + "loss": 0.9795, + "step": 5655 + }, + { + "epoch": 0.39247796821872183, + "grad_norm": 3.1832756996154785, + "learning_rate": 0.0001975769214696824, + "loss": 1.1795, + "step": 5656 + }, + { + "epoch": 0.3925473596558185, + "grad_norm": 2.412870407104492, + "learning_rate": 0.00019757569194976378, + "loss": 0.7763, + "step": 5657 + }, + { + "epoch": 0.39261675109291516, + "grad_norm": 2.650702714920044, + "learning_rate": 0.00019757446212181, + "loss": 1.0447, + "step": 5658 + }, + { + "epoch": 0.3926861425300118, + "grad_norm": 2.9268758296966553, + "learning_rate": 0.000197573231985825, + "loss": 0.8668, + "step": 5659 + }, + { + "epoch": 0.39275553396710844, + "grad_norm": 3.0832409858703613, + "learning_rate": 0.0001975720015418126, + "loss": 0.9709, + "step": 5660 + }, + { + "epoch": 0.3928249254042051, + "grad_norm": 2.643542766571045, + "learning_rate": 0.0001975707707897767, + "loss": 0.9175, + "step": 5661 + }, + { + "epoch": 0.39289431684130177, + "grad_norm": 2.2194082736968994, + "learning_rate": 0.0001975695397297212, + "loss": 0.7992, + "step": 5662 + }, + { + "epoch": 0.39296370827839844, + "grad_norm": 3.1787519454956055, + "learning_rate": 0.00019756830836164998, + "loss": 0.9379, + "step": 5663 + }, + { + "epoch": 0.3930330997154951, + "grad_norm": 4.338531017303467, + "learning_rate": 0.00019756707668556694, + "loss": 1.0509, + "step": 5664 + }, + { + "epoch": 0.39310249115259177, + "grad_norm": 2.5159990787506104, + "learning_rate": 0.00019756584470147597, + "loss": 0.8572, + "step": 5665 + }, + { + "epoch": 0.39317188258968844, + "grad_norm": 2.8896830081939697, + "learning_rate": 0.00019756461240938095, + "loss": 1.0578, + "step": 5666 + }, + { + "epoch": 0.3932412740267851, + "grad_norm": 3.147869110107422, + "learning_rate": 0.00019756337980928574, + "loss": 1.0021, + "step": 5667 + }, + { + "epoch": 0.39331066546388177, + "grad_norm": 2.7111611366271973, + "learning_rate": 0.0001975621469011943, + "loss": 0.8675, + "step": 5668 + }, + { + "epoch": 0.39338005690097844, + "grad_norm": 2.673610210418701, + "learning_rate": 0.00019756091368511043, + "loss": 0.9609, + "step": 5669 + }, + { + "epoch": 0.3934494483380751, + "grad_norm": 2.943162441253662, + "learning_rate": 0.00019755968016103812, + "loss": 0.9309, + "step": 5670 + }, + { + "epoch": 0.39351883977517177, + "grad_norm": 3.266859292984009, + "learning_rate": 0.0001975584463289812, + "loss": 0.6808, + "step": 5671 + }, + { + "epoch": 0.3935882312122684, + "grad_norm": 2.8731393814086914, + "learning_rate": 0.0001975572121889436, + "loss": 0.8856, + "step": 5672 + }, + { + "epoch": 0.39365762264936505, + "grad_norm": 3.1482739448547363, + "learning_rate": 0.00019755597774092918, + "loss": 0.9018, + "step": 5673 + }, + { + "epoch": 0.3937270140864617, + "grad_norm": 3.61320161819458, + "learning_rate": 0.00019755474298494187, + "loss": 1.2687, + "step": 5674 + }, + { + "epoch": 0.3937964055235584, + "grad_norm": 3.180262804031372, + "learning_rate": 0.00019755350792098551, + "loss": 0.9527, + "step": 5675 + }, + { + "epoch": 0.39386579696065505, + "grad_norm": 2.6112992763519287, + "learning_rate": 0.0001975522725490641, + "loss": 0.9852, + "step": 5676 + }, + { + "epoch": 0.3939351883977517, + "grad_norm": 4.760632038116455, + "learning_rate": 0.00019755103686918145, + "loss": 1.2765, + "step": 5677 + }, + { + "epoch": 0.3940045798348484, + "grad_norm": 2.5280330181121826, + "learning_rate": 0.0001975498008813415, + "loss": 0.919, + "step": 5678 + }, + { + "epoch": 0.39407397127194504, + "grad_norm": 3.4482667446136475, + "learning_rate": 0.00019754856458554818, + "loss": 0.9778, + "step": 5679 + }, + { + "epoch": 0.3941433627090417, + "grad_norm": 3.4320085048675537, + "learning_rate": 0.00019754732798180533, + "loss": 1.1985, + "step": 5680 + }, + { + "epoch": 0.3942127541461384, + "grad_norm": 2.6889939308166504, + "learning_rate": 0.00019754609107011688, + "loss": 0.8662, + "step": 5681 + }, + { + "epoch": 0.39428214558323504, + "grad_norm": 2.648012399673462, + "learning_rate": 0.00019754485385048675, + "loss": 0.9978, + "step": 5682 + }, + { + "epoch": 0.3943515370203317, + "grad_norm": 2.914581775665283, + "learning_rate": 0.0001975436163229188, + "loss": 0.9773, + "step": 5683 + }, + { + "epoch": 0.3944209284574284, + "grad_norm": 3.2135977745056152, + "learning_rate": 0.00019754237848741704, + "loss": 0.8707, + "step": 5684 + }, + { + "epoch": 0.39449031989452504, + "grad_norm": 2.4959185123443604, + "learning_rate": 0.00019754114034398526, + "loss": 0.9236, + "step": 5685 + }, + { + "epoch": 0.39455971133162165, + "grad_norm": 2.489081382751465, + "learning_rate": 0.00019753990189262744, + "loss": 0.8571, + "step": 5686 + }, + { + "epoch": 0.3946291027687183, + "grad_norm": 2.7393152713775635, + "learning_rate": 0.00019753866313334745, + "loss": 0.9772, + "step": 5687 + }, + { + "epoch": 0.394698494205815, + "grad_norm": 2.6141889095306396, + "learning_rate": 0.0001975374240661492, + "loss": 0.9117, + "step": 5688 + }, + { + "epoch": 0.39476788564291165, + "grad_norm": 2.7292628288269043, + "learning_rate": 0.00019753618469103663, + "loss": 0.8812, + "step": 5689 + }, + { + "epoch": 0.3948372770800083, + "grad_norm": 2.7504823207855225, + "learning_rate": 0.00019753494500801365, + "loss": 0.9349, + "step": 5690 + }, + { + "epoch": 0.394906668517105, + "grad_norm": 4.118539810180664, + "learning_rate": 0.00019753370501708416, + "loss": 0.9061, + "step": 5691 + }, + { + "epoch": 0.39497605995420165, + "grad_norm": 2.9985759258270264, + "learning_rate": 0.0001975324647182521, + "loss": 0.9295, + "step": 5692 + }, + { + "epoch": 0.3950454513912983, + "grad_norm": 2.69195556640625, + "learning_rate": 0.00019753122411152134, + "loss": 0.8808, + "step": 5693 + }, + { + "epoch": 0.395114842828395, + "grad_norm": 3.5722761154174805, + "learning_rate": 0.00019752998319689583, + "loss": 0.969, + "step": 5694 + }, + { + "epoch": 0.39518423426549165, + "grad_norm": 3.638442039489746, + "learning_rate": 0.00019752874197437946, + "loss": 1.0788, + "step": 5695 + }, + { + "epoch": 0.3952536257025883, + "grad_norm": 3.849363088607788, + "learning_rate": 0.00019752750044397618, + "loss": 1.4496, + "step": 5696 + }, + { + "epoch": 0.395323017139685, + "grad_norm": 2.681490421295166, + "learning_rate": 0.00019752625860568993, + "loss": 0.8646, + "step": 5697 + }, + { + "epoch": 0.39539240857678165, + "grad_norm": 3.1055145263671875, + "learning_rate": 0.00019752501645952454, + "loss": 1.074, + "step": 5698 + }, + { + "epoch": 0.39546180001387826, + "grad_norm": 2.597804307937622, + "learning_rate": 0.00019752377400548404, + "loss": 0.9599, + "step": 5699 + }, + { + "epoch": 0.39553119145097493, + "grad_norm": 3.553868293762207, + "learning_rate": 0.00019752253124357227, + "loss": 1.1154, + "step": 5700 + }, + { + "epoch": 0.3956005828880716, + "grad_norm": 2.969764232635498, + "learning_rate": 0.00019752128817379319, + "loss": 1.1103, + "step": 5701 + }, + { + "epoch": 0.39566997432516826, + "grad_norm": 3.0204408168792725, + "learning_rate": 0.00019752004479615072, + "loss": 1.1095, + "step": 5702 + }, + { + "epoch": 0.3957393657622649, + "grad_norm": 2.8192570209503174, + "learning_rate": 0.00019751880111064878, + "loss": 0.9189, + "step": 5703 + }, + { + "epoch": 0.3958087571993616, + "grad_norm": 3.684663772583008, + "learning_rate": 0.00019751755711729126, + "loss": 0.8691, + "step": 5704 + }, + { + "epoch": 0.39587814863645826, + "grad_norm": 3.2348620891571045, + "learning_rate": 0.00019751631281608218, + "loss": 0.8934, + "step": 5705 + }, + { + "epoch": 0.3959475400735549, + "grad_norm": 2.9472508430480957, + "learning_rate": 0.00019751506820702542, + "loss": 1.0007, + "step": 5706 + }, + { + "epoch": 0.3960169315106516, + "grad_norm": 3.006168842315674, + "learning_rate": 0.0001975138232901249, + "loss": 0.9401, + "step": 5707 + }, + { + "epoch": 0.39608632294774826, + "grad_norm": 3.138082265853882, + "learning_rate": 0.0001975125780653845, + "loss": 0.8647, + "step": 5708 + }, + { + "epoch": 0.3961557143848449, + "grad_norm": 4.0512542724609375, + "learning_rate": 0.00019751133253280827, + "loss": 1.2124, + "step": 5709 + }, + { + "epoch": 0.3962251058219416, + "grad_norm": 3.704251527786255, + "learning_rate": 0.00019751008669240003, + "loss": 1.012, + "step": 5710 + }, + { + "epoch": 0.39629449725903826, + "grad_norm": 2.4809062480926514, + "learning_rate": 0.00019750884054416375, + "loss": 0.8456, + "step": 5711 + }, + { + "epoch": 0.39636388869613487, + "grad_norm": 5.89408016204834, + "learning_rate": 0.00019750759408810345, + "loss": 1.2375, + "step": 5712 + }, + { + "epoch": 0.39643328013323154, + "grad_norm": 3.0067009925842285, + "learning_rate": 0.00019750634732422292, + "loss": 0.9021, + "step": 5713 + }, + { + "epoch": 0.3965026715703282, + "grad_norm": 8.030284881591797, + "learning_rate": 0.0001975051002525262, + "loss": 0.952, + "step": 5714 + }, + { + "epoch": 0.39657206300742487, + "grad_norm": 3.0635108947753906, + "learning_rate": 0.00019750385287301716, + "loss": 0.9837, + "step": 5715 + }, + { + "epoch": 0.39664145444452154, + "grad_norm": 3.2435531616210938, + "learning_rate": 0.0001975026051856998, + "loss": 0.9936, + "step": 5716 + }, + { + "epoch": 0.3967108458816182, + "grad_norm": 3.0887451171875, + "learning_rate": 0.000197501357190578, + "loss": 1.0322, + "step": 5717 + }, + { + "epoch": 0.39678023731871487, + "grad_norm": 5.0632429122924805, + "learning_rate": 0.00019750010888765573, + "loss": 0.9814, + "step": 5718 + }, + { + "epoch": 0.39684962875581153, + "grad_norm": 3.28439998626709, + "learning_rate": 0.00019749886027693693, + "loss": 0.9937, + "step": 5719 + }, + { + "epoch": 0.3969190201929082, + "grad_norm": 4.4836320877075195, + "learning_rate": 0.00019749761135842556, + "loss": 1.0692, + "step": 5720 + }, + { + "epoch": 0.39698841163000487, + "grad_norm": 10.958377838134766, + "learning_rate": 0.00019749636213212552, + "loss": 0.9854, + "step": 5721 + }, + { + "epoch": 0.39705780306710153, + "grad_norm": 2.9834344387054443, + "learning_rate": 0.0001974951125980408, + "loss": 0.9077, + "step": 5722 + }, + { + "epoch": 0.3971271945041982, + "grad_norm": 3.1592979431152344, + "learning_rate": 0.00019749386275617531, + "loss": 1.1223, + "step": 5723 + }, + { + "epoch": 0.39719658594129487, + "grad_norm": 2.9335687160491943, + "learning_rate": 0.000197492612606533, + "loss": 1.0326, + "step": 5724 + }, + { + "epoch": 0.39726597737839153, + "grad_norm": 3.1935880184173584, + "learning_rate": 0.00019749136214911782, + "loss": 1.1935, + "step": 5725 + }, + { + "epoch": 0.39733536881548814, + "grad_norm": 3.396420955657959, + "learning_rate": 0.00019749011138393371, + "loss": 1.1146, + "step": 5726 + }, + { + "epoch": 0.3974047602525848, + "grad_norm": 3.1880240440368652, + "learning_rate": 0.00019748886031098465, + "loss": 1.149, + "step": 5727 + }, + { + "epoch": 0.3974741516896815, + "grad_norm": 2.547642469406128, + "learning_rate": 0.00019748760893027457, + "loss": 0.8218, + "step": 5728 + }, + { + "epoch": 0.39754354312677814, + "grad_norm": 3.5989842414855957, + "learning_rate": 0.00019748635724180745, + "loss": 1.26, + "step": 5729 + }, + { + "epoch": 0.3976129345638748, + "grad_norm": 2.5463225841522217, + "learning_rate": 0.00019748510524558716, + "loss": 0.9241, + "step": 5730 + }, + { + "epoch": 0.3976823260009715, + "grad_norm": 2.8165624141693115, + "learning_rate": 0.00019748385294161773, + "loss": 0.9458, + "step": 5731 + }, + { + "epoch": 0.39775171743806814, + "grad_norm": 2.0625267028808594, + "learning_rate": 0.00019748260032990308, + "loss": 0.8333, + "step": 5732 + }, + { + "epoch": 0.3978211088751648, + "grad_norm": 6.043848514556885, + "learning_rate": 0.00019748134741044715, + "loss": 0.8907, + "step": 5733 + }, + { + "epoch": 0.3978905003122615, + "grad_norm": 2.8508901596069336, + "learning_rate": 0.00019748009418325393, + "loss": 0.9198, + "step": 5734 + }, + { + "epoch": 0.39795989174935814, + "grad_norm": 3.055645704269409, + "learning_rate": 0.00019747884064832735, + "loss": 1.0307, + "step": 5735 + }, + { + "epoch": 0.3980292831864548, + "grad_norm": 2.7425553798675537, + "learning_rate": 0.00019747758680567143, + "loss": 1.0719, + "step": 5736 + }, + { + "epoch": 0.3980986746235515, + "grad_norm": 4.812011241912842, + "learning_rate": 0.00019747633265529, + "loss": 0.9723, + "step": 5737 + }, + { + "epoch": 0.39816806606064814, + "grad_norm": 3.2209386825561523, + "learning_rate": 0.00019747507819718716, + "loss": 0.951, + "step": 5738 + }, + { + "epoch": 0.39823745749774475, + "grad_norm": 2.762064218521118, + "learning_rate": 0.0001974738234313668, + "loss": 0.8503, + "step": 5739 + }, + { + "epoch": 0.3983068489348414, + "grad_norm": 3.3663744926452637, + "learning_rate": 0.00019747256835783288, + "loss": 1.094, + "step": 5740 + }, + { + "epoch": 0.3983762403719381, + "grad_norm": 3.0396485328674316, + "learning_rate": 0.00019747131297658937, + "loss": 0.8115, + "step": 5741 + }, + { + "epoch": 0.39844563180903475, + "grad_norm": 3.123159170150757, + "learning_rate": 0.00019747005728764023, + "loss": 1.1415, + "step": 5742 + }, + { + "epoch": 0.3985150232461314, + "grad_norm": 3.0079734325408936, + "learning_rate": 0.0001974688012909894, + "loss": 0.9109, + "step": 5743 + }, + { + "epoch": 0.3985844146832281, + "grad_norm": 2.6802303791046143, + "learning_rate": 0.0001974675449866409, + "loss": 0.8405, + "step": 5744 + }, + { + "epoch": 0.39865380612032475, + "grad_norm": 3.184605836868286, + "learning_rate": 0.0001974662883745987, + "loss": 1.0598, + "step": 5745 + }, + { + "epoch": 0.3987231975574214, + "grad_norm": 2.7647101879119873, + "learning_rate": 0.0001974650314548667, + "loss": 0.7488, + "step": 5746 + }, + { + "epoch": 0.3987925889945181, + "grad_norm": 3.627868413925171, + "learning_rate": 0.0001974637742274489, + "loss": 1.3473, + "step": 5747 + }, + { + "epoch": 0.39886198043161475, + "grad_norm": 3.613060235977173, + "learning_rate": 0.0001974625166923493, + "loss": 0.9298, + "step": 5748 + }, + { + "epoch": 0.3989313718687114, + "grad_norm": 3.2316925525665283, + "learning_rate": 0.0001974612588495718, + "loss": 1.0578, + "step": 5749 + }, + { + "epoch": 0.3990007633058081, + "grad_norm": 3.101585626602173, + "learning_rate": 0.00019746000069912045, + "loss": 1.136, + "step": 5750 + }, + { + "epoch": 0.39907015474290475, + "grad_norm": 2.9403018951416016, + "learning_rate": 0.00019745874224099915, + "loss": 0.809, + "step": 5751 + }, + { + "epoch": 0.3991395461800014, + "grad_norm": 2.6319291591644287, + "learning_rate": 0.00019745748347521196, + "loss": 0.8578, + "step": 5752 + }, + { + "epoch": 0.399208937617098, + "grad_norm": 7.664699077606201, + "learning_rate": 0.00019745622440176276, + "loss": 1.297, + "step": 5753 + }, + { + "epoch": 0.3992783290541947, + "grad_norm": 3.105788469314575, + "learning_rate": 0.0001974549650206556, + "loss": 0.7975, + "step": 5754 + }, + { + "epoch": 0.39934772049129136, + "grad_norm": 3.476907730102539, + "learning_rate": 0.00019745370533189437, + "loss": 1.0869, + "step": 5755 + }, + { + "epoch": 0.399417111928388, + "grad_norm": 2.9884371757507324, + "learning_rate": 0.00019745244533548315, + "loss": 0.9869, + "step": 5756 + }, + { + "epoch": 0.3994865033654847, + "grad_norm": 3.309138774871826, + "learning_rate": 0.00019745118503142585, + "loss": 1.0267, + "step": 5757 + }, + { + "epoch": 0.39955589480258136, + "grad_norm": 2.7501375675201416, + "learning_rate": 0.00019744992441972646, + "loss": 0.9343, + "step": 5758 + }, + { + "epoch": 0.399625286239678, + "grad_norm": 2.5814459323883057, + "learning_rate": 0.000197448663500389, + "loss": 0.754, + "step": 5759 + }, + { + "epoch": 0.3996946776767747, + "grad_norm": 4.663064002990723, + "learning_rate": 0.00019744740227341736, + "loss": 1.2682, + "step": 5760 + }, + { + "epoch": 0.39976406911387136, + "grad_norm": 3.635572671890259, + "learning_rate": 0.0001974461407388156, + "loss": 0.99, + "step": 5761 + }, + { + "epoch": 0.399833460550968, + "grad_norm": 2.9141414165496826, + "learning_rate": 0.00019744487889658769, + "loss": 1.0521, + "step": 5762 + }, + { + "epoch": 0.3999028519880647, + "grad_norm": 2.7889158725738525, + "learning_rate": 0.00019744361674673758, + "loss": 1.0198, + "step": 5763 + }, + { + "epoch": 0.39997224342516136, + "grad_norm": 4.086861610412598, + "learning_rate": 0.0001974423542892693, + "loss": 1.1202, + "step": 5764 + }, + { + "epoch": 0.400041634862258, + "grad_norm": 3.223818302154541, + "learning_rate": 0.0001974410915241868, + "loss": 0.9733, + "step": 5765 + }, + { + "epoch": 0.40011102629935463, + "grad_norm": 3.636881113052368, + "learning_rate": 0.00019743982845149408, + "loss": 0.8359, + "step": 5766 + }, + { + "epoch": 0.4001804177364513, + "grad_norm": 2.8215975761413574, + "learning_rate": 0.00019743856507119513, + "loss": 1.0127, + "step": 5767 + }, + { + "epoch": 0.40024980917354797, + "grad_norm": 3.1366312503814697, + "learning_rate": 0.00019743730138329393, + "loss": 0.8277, + "step": 5768 + }, + { + "epoch": 0.40031920061064463, + "grad_norm": 2.879976511001587, + "learning_rate": 0.0001974360373877945, + "loss": 0.8432, + "step": 5769 + }, + { + "epoch": 0.4003885920477413, + "grad_norm": 3.351802349090576, + "learning_rate": 0.00019743477308470077, + "loss": 0.9793, + "step": 5770 + }, + { + "epoch": 0.40045798348483796, + "grad_norm": 3.2512545585632324, + "learning_rate": 0.00019743350847401678, + "loss": 0.7659, + "step": 5771 + }, + { + "epoch": 0.40052737492193463, + "grad_norm": 3.3059685230255127, + "learning_rate": 0.0001974322435557465, + "loss": 0.9641, + "step": 5772 + }, + { + "epoch": 0.4005967663590313, + "grad_norm": 3.050832748413086, + "learning_rate": 0.00019743097832989394, + "loss": 0.7834, + "step": 5773 + }, + { + "epoch": 0.40066615779612796, + "grad_norm": 2.6338560581207275, + "learning_rate": 0.0001974297127964631, + "loss": 0.7438, + "step": 5774 + }, + { + "epoch": 0.40073554923322463, + "grad_norm": 2.5963258743286133, + "learning_rate": 0.00019742844695545792, + "loss": 0.8023, + "step": 5775 + }, + { + "epoch": 0.4008049406703213, + "grad_norm": 3.5070693492889404, + "learning_rate": 0.00019742718080688248, + "loss": 1.1793, + "step": 5776 + }, + { + "epoch": 0.40087433210741796, + "grad_norm": 3.3423595428466797, + "learning_rate": 0.0001974259143507407, + "loss": 1.0086, + "step": 5777 + }, + { + "epoch": 0.40094372354451463, + "grad_norm": 3.533975124359131, + "learning_rate": 0.00019742464758703662, + "loss": 1.2236, + "step": 5778 + }, + { + "epoch": 0.4010131149816113, + "grad_norm": 4.435940742492676, + "learning_rate": 0.00019742338051577425, + "loss": 1.2885, + "step": 5779 + }, + { + "epoch": 0.4010825064187079, + "grad_norm": 2.9653759002685547, + "learning_rate": 0.00019742211313695756, + "loss": 0.9418, + "step": 5780 + }, + { + "epoch": 0.4011518978558046, + "grad_norm": 2.742964506149292, + "learning_rate": 0.00019742084545059058, + "loss": 1.0332, + "step": 5781 + }, + { + "epoch": 0.40122128929290124, + "grad_norm": 4.1759443283081055, + "learning_rate": 0.00019741957745667726, + "loss": 0.9296, + "step": 5782 + }, + { + "epoch": 0.4012906807299979, + "grad_norm": 3.0524792671203613, + "learning_rate": 0.00019741830915522167, + "loss": 1.0137, + "step": 5783 + }, + { + "epoch": 0.4013600721670946, + "grad_norm": 3.487760305404663, + "learning_rate": 0.00019741704054622775, + "loss": 1.1404, + "step": 5784 + }, + { + "epoch": 0.40142946360419124, + "grad_norm": 3.518846035003662, + "learning_rate": 0.00019741577162969953, + "loss": 0.735, + "step": 5785 + }, + { + "epoch": 0.4014988550412879, + "grad_norm": 2.8977582454681396, + "learning_rate": 0.00019741450240564102, + "loss": 0.956, + "step": 5786 + }, + { + "epoch": 0.40156824647838457, + "grad_norm": 2.915364980697632, + "learning_rate": 0.0001974132328740563, + "loss": 0.9351, + "step": 5787 + }, + { + "epoch": 0.40163763791548124, + "grad_norm": 3.7651827335357666, + "learning_rate": 0.00019741196303494921, + "loss": 0.9031, + "step": 5788 + }, + { + "epoch": 0.4017070293525779, + "grad_norm": 3.2662136554718018, + "learning_rate": 0.00019741069288832391, + "loss": 1.0243, + "step": 5789 + }, + { + "epoch": 0.40177642078967457, + "grad_norm": 3.0558722019195557, + "learning_rate": 0.00019740942243418435, + "loss": 0.9922, + "step": 5790 + }, + { + "epoch": 0.40184581222677124, + "grad_norm": 3.027853488922119, + "learning_rate": 0.00019740815167253452, + "loss": 0.8502, + "step": 5791 + }, + { + "epoch": 0.4019152036638679, + "grad_norm": 2.3106513023376465, + "learning_rate": 0.00019740688060337848, + "loss": 0.9104, + "step": 5792 + }, + { + "epoch": 0.4019845951009645, + "grad_norm": 2.675140619277954, + "learning_rate": 0.0001974056092267202, + "loss": 0.891, + "step": 5793 + }, + { + "epoch": 0.4020539865380612, + "grad_norm": 3.685138463973999, + "learning_rate": 0.00019740433754256373, + "loss": 1.0015, + "step": 5794 + }, + { + "epoch": 0.40212337797515785, + "grad_norm": 2.7529244422912598, + "learning_rate": 0.00019740306555091305, + "loss": 0.841, + "step": 5795 + }, + { + "epoch": 0.4021927694122545, + "grad_norm": 2.9762139320373535, + "learning_rate": 0.00019740179325177223, + "loss": 1.2465, + "step": 5796 + }, + { + "epoch": 0.4022621608493512, + "grad_norm": 2.7832560539245605, + "learning_rate": 0.00019740052064514522, + "loss": 0.9034, + "step": 5797 + }, + { + "epoch": 0.40233155228644785, + "grad_norm": 2.841068744659424, + "learning_rate": 0.00019739924773103605, + "loss": 0.8594, + "step": 5798 + }, + { + "epoch": 0.4024009437235445, + "grad_norm": 2.6153125762939453, + "learning_rate": 0.00019739797450944878, + "loss": 0.9378, + "step": 5799 + }, + { + "epoch": 0.4024703351606412, + "grad_norm": 3.1716384887695312, + "learning_rate": 0.00019739670098038742, + "loss": 1.1107, + "step": 5800 + }, + { + "epoch": 0.40253972659773785, + "grad_norm": 2.8290810585021973, + "learning_rate": 0.00019739542714385596, + "loss": 0.9243, + "step": 5801 + }, + { + "epoch": 0.4026091180348345, + "grad_norm": 4.603907585144043, + "learning_rate": 0.00019739415299985843, + "loss": 1.2765, + "step": 5802 + }, + { + "epoch": 0.4026785094719312, + "grad_norm": 3.0468332767486572, + "learning_rate": 0.00019739287854839886, + "loss": 0.8479, + "step": 5803 + }, + { + "epoch": 0.40274790090902785, + "grad_norm": 3.7519264221191406, + "learning_rate": 0.00019739160378948127, + "loss": 1.1169, + "step": 5804 + }, + { + "epoch": 0.4028172923461245, + "grad_norm": 3.1395797729492188, + "learning_rate": 0.00019739032872310972, + "loss": 0.9174, + "step": 5805 + }, + { + "epoch": 0.4028866837832212, + "grad_norm": 2.961164951324463, + "learning_rate": 0.00019738905334928814, + "loss": 1.0699, + "step": 5806 + }, + { + "epoch": 0.4029560752203178, + "grad_norm": 4.5434441566467285, + "learning_rate": 0.00019738777766802068, + "loss": 1.0142, + "step": 5807 + }, + { + "epoch": 0.40302546665741446, + "grad_norm": 2.8230180740356445, + "learning_rate": 0.00019738650167931126, + "loss": 0.9417, + "step": 5808 + }, + { + "epoch": 0.4030948580945111, + "grad_norm": 4.849161624908447, + "learning_rate": 0.00019738522538316396, + "loss": 0.9776, + "step": 5809 + }, + { + "epoch": 0.4031642495316078, + "grad_norm": 3.622274398803711, + "learning_rate": 0.00019738394877958282, + "loss": 1.1149, + "step": 5810 + }, + { + "epoch": 0.40323364096870445, + "grad_norm": 4.057389736175537, + "learning_rate": 0.00019738267186857183, + "loss": 1.2938, + "step": 5811 + }, + { + "epoch": 0.4033030324058011, + "grad_norm": 4.897439479827881, + "learning_rate": 0.00019738139465013505, + "loss": 1.3694, + "step": 5812 + }, + { + "epoch": 0.4033724238428978, + "grad_norm": 2.933774471282959, + "learning_rate": 0.00019738011712427654, + "loss": 0.9833, + "step": 5813 + }, + { + "epoch": 0.40344181527999445, + "grad_norm": 4.467594146728516, + "learning_rate": 0.00019737883929100028, + "loss": 1.183, + "step": 5814 + }, + { + "epoch": 0.4035112067170911, + "grad_norm": 4.063668727874756, + "learning_rate": 0.0001973775611503103, + "loss": 1.1696, + "step": 5815 + }, + { + "epoch": 0.4035805981541878, + "grad_norm": 7.660046577453613, + "learning_rate": 0.00019737628270221066, + "loss": 1.6198, + "step": 5816 + }, + { + "epoch": 0.40364998959128445, + "grad_norm": 3.109680414199829, + "learning_rate": 0.00019737500394670542, + "loss": 0.8573, + "step": 5817 + }, + { + "epoch": 0.4037193810283811, + "grad_norm": 3.3685903549194336, + "learning_rate": 0.00019737372488379858, + "loss": 1.0089, + "step": 5818 + }, + { + "epoch": 0.4037887724654778, + "grad_norm": 3.520503282546997, + "learning_rate": 0.00019737244551349417, + "loss": 0.9548, + "step": 5819 + }, + { + "epoch": 0.4038581639025744, + "grad_norm": 2.5393455028533936, + "learning_rate": 0.00019737116583579624, + "loss": 0.7934, + "step": 5820 + }, + { + "epoch": 0.40392755533967106, + "grad_norm": 3.612405776977539, + "learning_rate": 0.00019736988585070887, + "loss": 1.112, + "step": 5821 + }, + { + "epoch": 0.40399694677676773, + "grad_norm": 3.5680413246154785, + "learning_rate": 0.00019736860555823605, + "loss": 1.3514, + "step": 5822 + }, + { + "epoch": 0.4040663382138644, + "grad_norm": 3.453338146209717, + "learning_rate": 0.00019736732495838184, + "loss": 1.2422, + "step": 5823 + }, + { + "epoch": 0.40413572965096106, + "grad_norm": 3.592501640319824, + "learning_rate": 0.00019736604405115027, + "loss": 0.9506, + "step": 5824 + }, + { + "epoch": 0.40420512108805773, + "grad_norm": 2.79956316947937, + "learning_rate": 0.00019736476283654543, + "loss": 0.9738, + "step": 5825 + }, + { + "epoch": 0.4042745125251544, + "grad_norm": 2.8238027095794678, + "learning_rate": 0.0001973634813145713, + "loss": 0.7238, + "step": 5826 + }, + { + "epoch": 0.40434390396225106, + "grad_norm": 2.8462493419647217, + "learning_rate": 0.00019736219948523196, + "loss": 0.6464, + "step": 5827 + }, + { + "epoch": 0.4044132953993477, + "grad_norm": 3.519826650619507, + "learning_rate": 0.00019736091734853147, + "loss": 0.9328, + "step": 5828 + }, + { + "epoch": 0.4044826868364444, + "grad_norm": 3.187737464904785, + "learning_rate": 0.00019735963490447384, + "loss": 1.1033, + "step": 5829 + }, + { + "epoch": 0.40455207827354106, + "grad_norm": 3.2633914947509766, + "learning_rate": 0.00019735835215306313, + "loss": 1.1096, + "step": 5830 + }, + { + "epoch": 0.4046214697106377, + "grad_norm": 4.615447044372559, + "learning_rate": 0.00019735706909430341, + "loss": 1.05, + "step": 5831 + }, + { + "epoch": 0.4046908611477344, + "grad_norm": 3.4925343990325928, + "learning_rate": 0.00019735578572819873, + "loss": 1.1751, + "step": 5832 + }, + { + "epoch": 0.40476025258483106, + "grad_norm": 3.0665178298950195, + "learning_rate": 0.0001973545020547531, + "loss": 1.0925, + "step": 5833 + }, + { + "epoch": 0.40482964402192767, + "grad_norm": 3.557162046432495, + "learning_rate": 0.00019735321807397066, + "loss": 1.1654, + "step": 5834 + }, + { + "epoch": 0.40489903545902434, + "grad_norm": 4.69382381439209, + "learning_rate": 0.00019735193378585538, + "loss": 1.0448, + "step": 5835 + }, + { + "epoch": 0.404968426896121, + "grad_norm": 3.251180648803711, + "learning_rate": 0.0001973506491904113, + "loss": 1.1663, + "step": 5836 + }, + { + "epoch": 0.40503781833321767, + "grad_norm": 3.1397929191589355, + "learning_rate": 0.00019734936428764254, + "loss": 1.0156, + "step": 5837 + }, + { + "epoch": 0.40510720977031434, + "grad_norm": 2.9500691890716553, + "learning_rate": 0.00019734807907755312, + "loss": 0.7982, + "step": 5838 + }, + { + "epoch": 0.405176601207411, + "grad_norm": 4.036008358001709, + "learning_rate": 0.00019734679356014714, + "loss": 0.8517, + "step": 5839 + }, + { + "epoch": 0.40524599264450767, + "grad_norm": 2.960991144180298, + "learning_rate": 0.00019734550773542863, + "loss": 1.1715, + "step": 5840 + }, + { + "epoch": 0.40531538408160434, + "grad_norm": 4.38243293762207, + "learning_rate": 0.0001973442216034016, + "loss": 0.9943, + "step": 5841 + }, + { + "epoch": 0.405384775518701, + "grad_norm": 2.8185250759124756, + "learning_rate": 0.0001973429351640702, + "loss": 0.9146, + "step": 5842 + }, + { + "epoch": 0.40545416695579767, + "grad_norm": 3.732673168182373, + "learning_rate": 0.00019734164841743844, + "loss": 1.0705, + "step": 5843 + }, + { + "epoch": 0.40552355839289433, + "grad_norm": 4.098684787750244, + "learning_rate": 0.00019734036136351038, + "loss": 1.1256, + "step": 5844 + }, + { + "epoch": 0.405592949829991, + "grad_norm": 3.075211524963379, + "learning_rate": 0.0001973390740022901, + "loss": 0.9781, + "step": 5845 + }, + { + "epoch": 0.40566234126708767, + "grad_norm": 2.9149703979492188, + "learning_rate": 0.00019733778633378163, + "loss": 1.2556, + "step": 5846 + }, + { + "epoch": 0.4057317327041843, + "grad_norm": 3.239595890045166, + "learning_rate": 0.0001973364983579891, + "loss": 1.184, + "step": 5847 + }, + { + "epoch": 0.40580112414128094, + "grad_norm": 4.139148712158203, + "learning_rate": 0.0001973352100749165, + "loss": 0.9836, + "step": 5848 + }, + { + "epoch": 0.4058705155783776, + "grad_norm": 2.4272658824920654, + "learning_rate": 0.00019733392148456796, + "loss": 0.8492, + "step": 5849 + }, + { + "epoch": 0.4059399070154743, + "grad_norm": 2.4292349815368652, + "learning_rate": 0.00019733263258694752, + "loss": 0.7403, + "step": 5850 + }, + { + "epoch": 0.40600929845257094, + "grad_norm": 2.7640907764434814, + "learning_rate": 0.00019733134338205925, + "loss": 0.898, + "step": 5851 + }, + { + "epoch": 0.4060786898896676, + "grad_norm": 3.0816123485565186, + "learning_rate": 0.00019733005386990722, + "loss": 0.8952, + "step": 5852 + }, + { + "epoch": 0.4061480813267643, + "grad_norm": 3.1215808391571045, + "learning_rate": 0.00019732876405049548, + "loss": 0.8978, + "step": 5853 + }, + { + "epoch": 0.40621747276386094, + "grad_norm": 4.299373149871826, + "learning_rate": 0.00019732747392382818, + "loss": 1.0424, + "step": 5854 + }, + { + "epoch": 0.4062868642009576, + "grad_norm": 2.800471544265747, + "learning_rate": 0.0001973261834899093, + "loss": 0.915, + "step": 5855 + }, + { + "epoch": 0.4063562556380543, + "grad_norm": 3.055271625518799, + "learning_rate": 0.00019732489274874296, + "loss": 0.9145, + "step": 5856 + }, + { + "epoch": 0.40642564707515094, + "grad_norm": 3.4077494144439697, + "learning_rate": 0.00019732360170033323, + "loss": 1.2161, + "step": 5857 + }, + { + "epoch": 0.4064950385122476, + "grad_norm": 2.8723390102386475, + "learning_rate": 0.00019732231034468416, + "loss": 0.9727, + "step": 5858 + }, + { + "epoch": 0.4065644299493443, + "grad_norm": 2.3242955207824707, + "learning_rate": 0.00019732101868179987, + "loss": 0.778, + "step": 5859 + }, + { + "epoch": 0.4066338213864409, + "grad_norm": 3.0442862510681152, + "learning_rate": 0.0001973197267116844, + "loss": 1.1585, + "step": 5860 + }, + { + "epoch": 0.40670321282353755, + "grad_norm": 2.311073064804077, + "learning_rate": 0.00019731843443434184, + "loss": 0.7679, + "step": 5861 + }, + { + "epoch": 0.4067726042606342, + "grad_norm": 2.7152326107025146, + "learning_rate": 0.0001973171418497763, + "loss": 0.8624, + "step": 5862 + }, + { + "epoch": 0.4068419956977309, + "grad_norm": 3.7142391204833984, + "learning_rate": 0.0001973158489579918, + "loss": 0.8935, + "step": 5863 + }, + { + "epoch": 0.40691138713482755, + "grad_norm": 3.3825199604034424, + "learning_rate": 0.00019731455575899246, + "loss": 0.899, + "step": 5864 + }, + { + "epoch": 0.4069807785719242, + "grad_norm": 3.6041576862335205, + "learning_rate": 0.00019731326225278238, + "loss": 0.9019, + "step": 5865 + }, + { + "epoch": 0.4070501700090209, + "grad_norm": 2.9431638717651367, + "learning_rate": 0.0001973119684393656, + "loss": 0.8475, + "step": 5866 + }, + { + "epoch": 0.40711956144611755, + "grad_norm": 3.1986732482910156, + "learning_rate": 0.00019731067431874623, + "loss": 0.7689, + "step": 5867 + }, + { + "epoch": 0.4071889528832142, + "grad_norm": 2.810706853866577, + "learning_rate": 0.00019730937989092837, + "loss": 0.9844, + "step": 5868 + }, + { + "epoch": 0.4072583443203109, + "grad_norm": 4.515185356140137, + "learning_rate": 0.00019730808515591608, + "loss": 0.9468, + "step": 5869 + }, + { + "epoch": 0.40732773575740755, + "grad_norm": 3.052474021911621, + "learning_rate": 0.00019730679011371348, + "loss": 1.0059, + "step": 5870 + }, + { + "epoch": 0.4073971271945042, + "grad_norm": 3.9073472023010254, + "learning_rate": 0.00019730549476432456, + "loss": 1.0907, + "step": 5871 + }, + { + "epoch": 0.4074665186316009, + "grad_norm": 2.871588706970215, + "learning_rate": 0.00019730419910775354, + "loss": 0.83, + "step": 5872 + }, + { + "epoch": 0.40753591006869755, + "grad_norm": 3.4685404300689697, + "learning_rate": 0.00019730290314400444, + "loss": 0.9762, + "step": 5873 + }, + { + "epoch": 0.40760530150579416, + "grad_norm": 4.293680667877197, + "learning_rate": 0.00019730160687308137, + "loss": 1.0062, + "step": 5874 + }, + { + "epoch": 0.4076746929428908, + "grad_norm": 4.269751071929932, + "learning_rate": 0.0001973003102949884, + "loss": 0.9637, + "step": 5875 + }, + { + "epoch": 0.4077440843799875, + "grad_norm": 2.875115394592285, + "learning_rate": 0.00019729901340972966, + "loss": 0.7933, + "step": 5876 + }, + { + "epoch": 0.40781347581708416, + "grad_norm": 3.188511610031128, + "learning_rate": 0.0001972977162173092, + "loss": 1.0057, + "step": 5877 + }, + { + "epoch": 0.4078828672541808, + "grad_norm": 3.893714666366577, + "learning_rate": 0.00019729641871773116, + "loss": 1.0249, + "step": 5878 + }, + { + "epoch": 0.4079522586912775, + "grad_norm": 4.307023048400879, + "learning_rate": 0.0001972951209109996, + "loss": 1.0183, + "step": 5879 + }, + { + "epoch": 0.40802165012837416, + "grad_norm": 2.7347958087921143, + "learning_rate": 0.00019729382279711866, + "loss": 1.0001, + "step": 5880 + }, + { + "epoch": 0.4080910415654708, + "grad_norm": 2.8490498065948486, + "learning_rate": 0.00019729252437609236, + "loss": 0.812, + "step": 5881 + }, + { + "epoch": 0.4081604330025675, + "grad_norm": 2.9419548511505127, + "learning_rate": 0.00019729122564792487, + "loss": 0.9626, + "step": 5882 + }, + { + "epoch": 0.40822982443966416, + "grad_norm": 3.7838387489318848, + "learning_rate": 0.00019728992661262028, + "loss": 1.1125, + "step": 5883 + }, + { + "epoch": 0.4082992158767608, + "grad_norm": 2.851799249649048, + "learning_rate": 0.00019728862727018268, + "loss": 1.0418, + "step": 5884 + }, + { + "epoch": 0.4083686073138575, + "grad_norm": 3.774580717086792, + "learning_rate": 0.00019728732762061617, + "loss": 1.323, + "step": 5885 + }, + { + "epoch": 0.40843799875095416, + "grad_norm": 3.945965528488159, + "learning_rate": 0.00019728602766392483, + "loss": 0.8712, + "step": 5886 + }, + { + "epoch": 0.40850739018805077, + "grad_norm": 2.2762176990509033, + "learning_rate": 0.00019728472740011284, + "loss": 0.7471, + "step": 5887 + }, + { + "epoch": 0.40857678162514743, + "grad_norm": 2.0736823081970215, + "learning_rate": 0.0001972834268291842, + "loss": 0.8706, + "step": 5888 + }, + { + "epoch": 0.4086461730622441, + "grad_norm": 4.8327765464782715, + "learning_rate": 0.00019728212595114308, + "loss": 1.3895, + "step": 5889 + }, + { + "epoch": 0.40871556449934077, + "grad_norm": 5.983523368835449, + "learning_rate": 0.0001972808247659936, + "loss": 0.95, + "step": 5890 + }, + { + "epoch": 0.40878495593643743, + "grad_norm": 3.135042905807495, + "learning_rate": 0.00019727952327373982, + "loss": 0.8879, + "step": 5891 + }, + { + "epoch": 0.4088543473735341, + "grad_norm": 2.704960584640503, + "learning_rate": 0.00019727822147438587, + "loss": 0.9044, + "step": 5892 + }, + { + "epoch": 0.40892373881063077, + "grad_norm": 2.780640125274658, + "learning_rate": 0.0001972769193679359, + "loss": 1.1035, + "step": 5893 + }, + { + "epoch": 0.40899313024772743, + "grad_norm": 2.7716281414031982, + "learning_rate": 0.00019727561695439394, + "loss": 1.2582, + "step": 5894 + }, + { + "epoch": 0.4090625216848241, + "grad_norm": 2.7025339603424072, + "learning_rate": 0.00019727431423376414, + "loss": 0.8584, + "step": 5895 + }, + { + "epoch": 0.40913191312192076, + "grad_norm": 2.8087949752807617, + "learning_rate": 0.0001972730112060506, + "loss": 1.0543, + "step": 5896 + }, + { + "epoch": 0.40920130455901743, + "grad_norm": 2.7316908836364746, + "learning_rate": 0.00019727170787125752, + "loss": 1.0698, + "step": 5897 + }, + { + "epoch": 0.4092706959961141, + "grad_norm": 2.4379189014434814, + "learning_rate": 0.0001972704042293889, + "loss": 0.8747, + "step": 5898 + }, + { + "epoch": 0.40934008743321076, + "grad_norm": 2.358320474624634, + "learning_rate": 0.0001972691002804489, + "loss": 1.0779, + "step": 5899 + }, + { + "epoch": 0.40940947887030743, + "grad_norm": 2.663510322570801, + "learning_rate": 0.00019726779602444164, + "loss": 0.9023, + "step": 5900 + }, + { + "epoch": 0.40947887030740404, + "grad_norm": 2.891054153442383, + "learning_rate": 0.00019726649146137125, + "loss": 0.9285, + "step": 5901 + }, + { + "epoch": 0.4095482617445007, + "grad_norm": 3.549907922744751, + "learning_rate": 0.00019726518659124177, + "loss": 1.2015, + "step": 5902 + }, + { + "epoch": 0.4096176531815974, + "grad_norm": 2.9225473403930664, + "learning_rate": 0.00019726388141405745, + "loss": 1.0354, + "step": 5903 + }, + { + "epoch": 0.40968704461869404, + "grad_norm": 3.469822406768799, + "learning_rate": 0.0001972625759298223, + "loss": 0.8855, + "step": 5904 + }, + { + "epoch": 0.4097564360557907, + "grad_norm": 2.229200839996338, + "learning_rate": 0.0001972612701385405, + "loss": 0.8763, + "step": 5905 + }, + { + "epoch": 0.4098258274928874, + "grad_norm": 3.5745389461517334, + "learning_rate": 0.00019725996404021615, + "loss": 1.2391, + "step": 5906 + }, + { + "epoch": 0.40989521892998404, + "grad_norm": 5.091414928436279, + "learning_rate": 0.00019725865763485338, + "loss": 1.1835, + "step": 5907 + }, + { + "epoch": 0.4099646103670807, + "grad_norm": 3.5358564853668213, + "learning_rate": 0.00019725735092245632, + "loss": 0.8854, + "step": 5908 + }, + { + "epoch": 0.4100340018041774, + "grad_norm": 2.526045083999634, + "learning_rate": 0.00019725604390302906, + "loss": 0.8354, + "step": 5909 + }, + { + "epoch": 0.41010339324127404, + "grad_norm": 3.2496254444122314, + "learning_rate": 0.00019725473657657578, + "loss": 0.9716, + "step": 5910 + }, + { + "epoch": 0.4101727846783707, + "grad_norm": 2.7083451747894287, + "learning_rate": 0.00019725342894310057, + "loss": 0.7488, + "step": 5911 + }, + { + "epoch": 0.41024217611546737, + "grad_norm": 3.68369460105896, + "learning_rate": 0.00019725212100260757, + "loss": 0.9511, + "step": 5912 + }, + { + "epoch": 0.41031156755256404, + "grad_norm": 3.0170154571533203, + "learning_rate": 0.0001972508127551009, + "loss": 1.0254, + "step": 5913 + }, + { + "epoch": 0.41038095898966065, + "grad_norm": 2.908891439437866, + "learning_rate": 0.00019724950420058472, + "loss": 0.7601, + "step": 5914 + }, + { + "epoch": 0.4104503504267573, + "grad_norm": 4.4615888595581055, + "learning_rate": 0.00019724819533906311, + "loss": 1.0704, + "step": 5915 + }, + { + "epoch": 0.410519741863854, + "grad_norm": 2.5469346046447754, + "learning_rate": 0.00019724688617054023, + "loss": 0.828, + "step": 5916 + }, + { + "epoch": 0.41058913330095065, + "grad_norm": 2.8123791217803955, + "learning_rate": 0.00019724557669502027, + "loss": 0.8852, + "step": 5917 + }, + { + "epoch": 0.4106585247380473, + "grad_norm": 3.083697557449341, + "learning_rate": 0.00019724426691250724, + "loss": 1.2832, + "step": 5918 + }, + { + "epoch": 0.410727916175144, + "grad_norm": 2.5928025245666504, + "learning_rate": 0.00019724295682300538, + "loss": 0.8082, + "step": 5919 + }, + { + "epoch": 0.41079730761224065, + "grad_norm": 2.4554452896118164, + "learning_rate": 0.00019724164642651878, + "loss": 0.999, + "step": 5920 + }, + { + "epoch": 0.4108666990493373, + "grad_norm": 2.822383165359497, + "learning_rate": 0.00019724033572305155, + "loss": 0.7151, + "step": 5921 + }, + { + "epoch": 0.410936090486434, + "grad_norm": 2.72849702835083, + "learning_rate": 0.0001972390247126079, + "loss": 1.1215, + "step": 5922 + }, + { + "epoch": 0.41100548192353065, + "grad_norm": 2.885891914367676, + "learning_rate": 0.00019723771339519191, + "loss": 0.9171, + "step": 5923 + }, + { + "epoch": 0.4110748733606273, + "grad_norm": 3.260469913482666, + "learning_rate": 0.00019723640177080775, + "loss": 1.0271, + "step": 5924 + }, + { + "epoch": 0.411144264797724, + "grad_norm": 3.131571054458618, + "learning_rate": 0.00019723508983945955, + "loss": 1.0456, + "step": 5925 + }, + { + "epoch": 0.41121365623482065, + "grad_norm": 3.956766366958618, + "learning_rate": 0.00019723377760115147, + "loss": 0.9234, + "step": 5926 + }, + { + "epoch": 0.4112830476719173, + "grad_norm": 2.6819279193878174, + "learning_rate": 0.0001972324650558876, + "loss": 0.9381, + "step": 5927 + }, + { + "epoch": 0.4113524391090139, + "grad_norm": 3.4285666942596436, + "learning_rate": 0.00019723115220367213, + "loss": 0.945, + "step": 5928 + }, + { + "epoch": 0.4114218305461106, + "grad_norm": 3.365902900695801, + "learning_rate": 0.00019722983904450924, + "loss": 0.9949, + "step": 5929 + }, + { + "epoch": 0.41149122198320726, + "grad_norm": 2.803882598876953, + "learning_rate": 0.00019722852557840296, + "loss": 0.9773, + "step": 5930 + }, + { + "epoch": 0.4115606134203039, + "grad_norm": 3.4531912803649902, + "learning_rate": 0.00019722721180535753, + "loss": 1.164, + "step": 5931 + }, + { + "epoch": 0.4116300048574006, + "grad_norm": 2.8911211490631104, + "learning_rate": 0.0001972258977253771, + "loss": 1.0044, + "step": 5932 + }, + { + "epoch": 0.41169939629449726, + "grad_norm": 4.186860084533691, + "learning_rate": 0.00019722458333846577, + "loss": 1.1392, + "step": 5933 + }, + { + "epoch": 0.4117687877315939, + "grad_norm": 3.3418502807617188, + "learning_rate": 0.0001972232686446277, + "loss": 0.9606, + "step": 5934 + }, + { + "epoch": 0.4118381791686906, + "grad_norm": 3.7831919193267822, + "learning_rate": 0.00019722195364386705, + "loss": 1.1898, + "step": 5935 + }, + { + "epoch": 0.41190757060578725, + "grad_norm": 3.294335126876831, + "learning_rate": 0.000197220638336188, + "loss": 1.0918, + "step": 5936 + }, + { + "epoch": 0.4119769620428839, + "grad_norm": 3.32698655128479, + "learning_rate": 0.00019721932272159464, + "loss": 0.781, + "step": 5937 + }, + { + "epoch": 0.4120463534799806, + "grad_norm": 2.915720224380493, + "learning_rate": 0.0001972180068000912, + "loss": 1.2071, + "step": 5938 + }, + { + "epoch": 0.41211574491707725, + "grad_norm": 3.3471169471740723, + "learning_rate": 0.00019721669057168174, + "loss": 0.8745, + "step": 5939 + }, + { + "epoch": 0.4121851363541739, + "grad_norm": 4.734062671661377, + "learning_rate": 0.00019721537403637052, + "loss": 1.2355, + "step": 5940 + }, + { + "epoch": 0.41225452779127053, + "grad_norm": 3.0540082454681396, + "learning_rate": 0.0001972140571941616, + "loss": 0.9238, + "step": 5941 + }, + { + "epoch": 0.4123239192283672, + "grad_norm": 3.3220953941345215, + "learning_rate": 0.00019721274004505922, + "loss": 1.2026, + "step": 5942 + }, + { + "epoch": 0.41239331066546386, + "grad_norm": 2.6267971992492676, + "learning_rate": 0.00019721142258906745, + "loss": 0.7226, + "step": 5943 + }, + { + "epoch": 0.41246270210256053, + "grad_norm": 2.70650577545166, + "learning_rate": 0.00019721010482619052, + "loss": 0.9418, + "step": 5944 + }, + { + "epoch": 0.4125320935396572, + "grad_norm": 3.233412504196167, + "learning_rate": 0.00019720878675643259, + "loss": 0.862, + "step": 5945 + }, + { + "epoch": 0.41260148497675386, + "grad_norm": 3.3420166969299316, + "learning_rate": 0.00019720746837979778, + "loss": 1.0043, + "step": 5946 + }, + { + "epoch": 0.41267087641385053, + "grad_norm": 3.11397123336792, + "learning_rate": 0.00019720614969629026, + "loss": 1.1715, + "step": 5947 + }, + { + "epoch": 0.4127402678509472, + "grad_norm": 2.393156051635742, + "learning_rate": 0.00019720483070591422, + "loss": 0.9557, + "step": 5948 + }, + { + "epoch": 0.41280965928804386, + "grad_norm": 2.6082210540771484, + "learning_rate": 0.00019720351140867378, + "loss": 1.0452, + "step": 5949 + }, + { + "epoch": 0.41287905072514053, + "grad_norm": 3.184013605117798, + "learning_rate": 0.00019720219180457317, + "loss": 0.9749, + "step": 5950 + }, + { + "epoch": 0.4129484421622372, + "grad_norm": 2.646801471710205, + "learning_rate": 0.00019720087189361652, + "loss": 1.0347, + "step": 5951 + }, + { + "epoch": 0.41301783359933386, + "grad_norm": 2.6103055477142334, + "learning_rate": 0.00019719955167580796, + "loss": 1.0375, + "step": 5952 + }, + { + "epoch": 0.4130872250364305, + "grad_norm": 2.601834297180176, + "learning_rate": 0.00019719823115115173, + "loss": 1.044, + "step": 5953 + }, + { + "epoch": 0.4131566164735272, + "grad_norm": 2.9637858867645264, + "learning_rate": 0.0001971969103196519, + "loss": 0.9766, + "step": 5954 + }, + { + "epoch": 0.4132260079106238, + "grad_norm": 3.5635128021240234, + "learning_rate": 0.00019719558918131276, + "loss": 1.212, + "step": 5955 + }, + { + "epoch": 0.41329539934772047, + "grad_norm": 3.2499094009399414, + "learning_rate": 0.00019719426773613842, + "loss": 1.1396, + "step": 5956 + }, + { + "epoch": 0.41336479078481714, + "grad_norm": 3.1000988483428955, + "learning_rate": 0.00019719294598413302, + "loss": 1.1615, + "step": 5957 + }, + { + "epoch": 0.4134341822219138, + "grad_norm": 2.9278597831726074, + "learning_rate": 0.0001971916239253008, + "loss": 1.0101, + "step": 5958 + }, + { + "epoch": 0.41350357365901047, + "grad_norm": 3.2234365940093994, + "learning_rate": 0.00019719030155964585, + "loss": 1.1024, + "step": 5959 + }, + { + "epoch": 0.41357296509610714, + "grad_norm": 3.9558868408203125, + "learning_rate": 0.00019718897888717246, + "loss": 1.1294, + "step": 5960 + }, + { + "epoch": 0.4136423565332038, + "grad_norm": 2.7941911220550537, + "learning_rate": 0.00019718765590788468, + "loss": 0.936, + "step": 5961 + }, + { + "epoch": 0.41371174797030047, + "grad_norm": 2.958540201187134, + "learning_rate": 0.0001971863326217868, + "loss": 0.7896, + "step": 5962 + }, + { + "epoch": 0.41378113940739714, + "grad_norm": 2.3402984142303467, + "learning_rate": 0.0001971850090288829, + "loss": 0.8709, + "step": 5963 + }, + { + "epoch": 0.4138505308444938, + "grad_norm": 2.3606958389282227, + "learning_rate": 0.00019718368512917719, + "loss": 0.8924, + "step": 5964 + }, + { + "epoch": 0.41391992228159047, + "grad_norm": 2.0958738327026367, + "learning_rate": 0.0001971823609226739, + "loss": 0.7468, + "step": 5965 + }, + { + "epoch": 0.41398931371868714, + "grad_norm": 2.8261139392852783, + "learning_rate": 0.00019718103640937716, + "loss": 0.9026, + "step": 5966 + }, + { + "epoch": 0.4140587051557838, + "grad_norm": 3.4753189086914062, + "learning_rate": 0.00019717971158929117, + "loss": 0.9807, + "step": 5967 + }, + { + "epoch": 0.4141280965928804, + "grad_norm": 3.686185836791992, + "learning_rate": 0.00019717838646242012, + "loss": 1.1165, + "step": 5968 + }, + { + "epoch": 0.4141974880299771, + "grad_norm": 1.9174034595489502, + "learning_rate": 0.00019717706102876815, + "loss": 0.5983, + "step": 5969 + }, + { + "epoch": 0.41426687946707375, + "grad_norm": 3.513408899307251, + "learning_rate": 0.00019717573528833947, + "loss": 1.0804, + "step": 5970 + }, + { + "epoch": 0.4143362709041704, + "grad_norm": 2.2243125438690186, + "learning_rate": 0.00019717440924113828, + "loss": 0.9025, + "step": 5971 + }, + { + "epoch": 0.4144056623412671, + "grad_norm": 3.2180275917053223, + "learning_rate": 0.00019717308288716876, + "loss": 1.0473, + "step": 5972 + }, + { + "epoch": 0.41447505377836374, + "grad_norm": 4.112575531005859, + "learning_rate": 0.00019717175622643507, + "loss": 1.102, + "step": 5973 + }, + { + "epoch": 0.4145444452154604, + "grad_norm": 2.989763021469116, + "learning_rate": 0.00019717042925894145, + "loss": 1.0354, + "step": 5974 + }, + { + "epoch": 0.4146138366525571, + "grad_norm": 3.3067924976348877, + "learning_rate": 0.00019716910198469206, + "loss": 0.7702, + "step": 5975 + }, + { + "epoch": 0.41468322808965374, + "grad_norm": 3.806156873703003, + "learning_rate": 0.00019716777440369104, + "loss": 1.1533, + "step": 5976 + }, + { + "epoch": 0.4147526195267504, + "grad_norm": 2.457228660583496, + "learning_rate": 0.00019716644651594266, + "loss": 1.0338, + "step": 5977 + }, + { + "epoch": 0.4148220109638471, + "grad_norm": 2.609423875808716, + "learning_rate": 0.0001971651183214511, + "loss": 0.8831, + "step": 5978 + }, + { + "epoch": 0.41489140240094374, + "grad_norm": 3.092533826828003, + "learning_rate": 0.0001971637898202205, + "loss": 0.8218, + "step": 5979 + }, + { + "epoch": 0.4149607938380404, + "grad_norm": 2.460942029953003, + "learning_rate": 0.0001971624610122551, + "loss": 0.8557, + "step": 5980 + }, + { + "epoch": 0.4150301852751371, + "grad_norm": 2.7951393127441406, + "learning_rate": 0.00019716113189755912, + "loss": 0.8228, + "step": 5981 + }, + { + "epoch": 0.4150995767122337, + "grad_norm": 3.9287919998168945, + "learning_rate": 0.00019715980247613668, + "loss": 0.9993, + "step": 5982 + }, + { + "epoch": 0.41516896814933035, + "grad_norm": 3.565566301345825, + "learning_rate": 0.00019715847274799205, + "loss": 0.825, + "step": 5983 + }, + { + "epoch": 0.415238359586427, + "grad_norm": 3.4341189861297607, + "learning_rate": 0.00019715714271312934, + "loss": 0.9007, + "step": 5984 + }, + { + "epoch": 0.4153077510235237, + "grad_norm": 2.7878835201263428, + "learning_rate": 0.00019715581237155286, + "loss": 0.8609, + "step": 5985 + }, + { + "epoch": 0.41537714246062035, + "grad_norm": 2.6132736206054688, + "learning_rate": 0.00019715448172326672, + "loss": 0.9898, + "step": 5986 + }, + { + "epoch": 0.415446533897717, + "grad_norm": 3.2829179763793945, + "learning_rate": 0.00019715315076827513, + "loss": 0.8042, + "step": 5987 + }, + { + "epoch": 0.4155159253348137, + "grad_norm": 2.1417016983032227, + "learning_rate": 0.00019715181950658237, + "loss": 0.6179, + "step": 5988 + }, + { + "epoch": 0.41558531677191035, + "grad_norm": 3.2819643020629883, + "learning_rate": 0.00019715048793819256, + "loss": 0.886, + "step": 5989 + }, + { + "epoch": 0.415654708209007, + "grad_norm": 2.5632858276367188, + "learning_rate": 0.00019714915606310992, + "loss": 1.1112, + "step": 5990 + }, + { + "epoch": 0.4157240996461037, + "grad_norm": 3.6542673110961914, + "learning_rate": 0.00019714782388133867, + "loss": 0.8707, + "step": 5991 + }, + { + "epoch": 0.41579349108320035, + "grad_norm": 3.1525070667266846, + "learning_rate": 0.00019714649139288302, + "loss": 0.6839, + "step": 5992 + }, + { + "epoch": 0.415862882520297, + "grad_norm": 2.9439704418182373, + "learning_rate": 0.00019714515859774716, + "loss": 0.9524, + "step": 5993 + }, + { + "epoch": 0.4159322739573937, + "grad_norm": 3.074467182159424, + "learning_rate": 0.00019714382549593533, + "loss": 1.0257, + "step": 5994 + }, + { + "epoch": 0.4160016653944903, + "grad_norm": 3.482659101486206, + "learning_rate": 0.0001971424920874517, + "loss": 0.8475, + "step": 5995 + }, + { + "epoch": 0.41607105683158696, + "grad_norm": 2.830239772796631, + "learning_rate": 0.00019714115837230048, + "loss": 1.1534, + "step": 5996 + }, + { + "epoch": 0.4161404482686836, + "grad_norm": 2.6631550788879395, + "learning_rate": 0.00019713982435048591, + "loss": 0.9269, + "step": 5997 + }, + { + "epoch": 0.4162098397057803, + "grad_norm": 2.822930335998535, + "learning_rate": 0.00019713849002201217, + "loss": 0.925, + "step": 5998 + }, + { + "epoch": 0.41627923114287696, + "grad_norm": 4.3145012855529785, + "learning_rate": 0.0001971371553868835, + "loss": 1.1517, + "step": 5999 + }, + { + "epoch": 0.4163486225799736, + "grad_norm": 3.232234239578247, + "learning_rate": 0.00019713582044510408, + "loss": 0.8485, + "step": 6000 + }, + { + "epoch": 0.4164180140170703, + "grad_norm": 3.2413132190704346, + "learning_rate": 0.00019713448519667818, + "loss": 1.04, + "step": 6001 + }, + { + "epoch": 0.41648740545416696, + "grad_norm": 4.478813171386719, + "learning_rate": 0.00019713314964160995, + "loss": 1.4481, + "step": 6002 + }, + { + "epoch": 0.4165567968912636, + "grad_norm": 2.708874225616455, + "learning_rate": 0.00019713181377990365, + "loss": 0.8446, + "step": 6003 + }, + { + "epoch": 0.4166261883283603, + "grad_norm": 3.754491090774536, + "learning_rate": 0.00019713047761156348, + "loss": 1.237, + "step": 6004 + }, + { + "epoch": 0.41669557976545696, + "grad_norm": 3.7211577892303467, + "learning_rate": 0.00019712914113659364, + "loss": 1.1654, + "step": 6005 + }, + { + "epoch": 0.4167649712025536, + "grad_norm": 3.284529685974121, + "learning_rate": 0.0001971278043549984, + "loss": 1.1987, + "step": 6006 + }, + { + "epoch": 0.4168343626396503, + "grad_norm": 6.567990779876709, + "learning_rate": 0.00019712646726678193, + "loss": 1.0835, + "step": 6007 + }, + { + "epoch": 0.4169037540767469, + "grad_norm": 7.807351589202881, + "learning_rate": 0.0001971251298719485, + "loss": 1.2443, + "step": 6008 + }, + { + "epoch": 0.41697314551384357, + "grad_norm": 3.0601179599761963, + "learning_rate": 0.00019712379217050227, + "loss": 1.1633, + "step": 6009 + }, + { + "epoch": 0.41704253695094023, + "grad_norm": 2.506232738494873, + "learning_rate": 0.00019712245416244752, + "loss": 0.743, + "step": 6010 + }, + { + "epoch": 0.4171119283880369, + "grad_norm": 3.6179049015045166, + "learning_rate": 0.00019712111584778845, + "loss": 0.8769, + "step": 6011 + }, + { + "epoch": 0.41718131982513357, + "grad_norm": 3.5567593574523926, + "learning_rate": 0.00019711977722652928, + "loss": 0.9879, + "step": 6012 + }, + { + "epoch": 0.41725071126223023, + "grad_norm": 3.541358709335327, + "learning_rate": 0.00019711843829867424, + "loss": 0.7909, + "step": 6013 + }, + { + "epoch": 0.4173201026993269, + "grad_norm": 4.089737415313721, + "learning_rate": 0.00019711709906422754, + "loss": 1.2519, + "step": 6014 + }, + { + "epoch": 0.41738949413642357, + "grad_norm": 3.9706344604492188, + "learning_rate": 0.00019711575952319346, + "loss": 1.0158, + "step": 6015 + }, + { + "epoch": 0.41745888557352023, + "grad_norm": 3.5314395427703857, + "learning_rate": 0.0001971144196755762, + "loss": 1.2459, + "step": 6016 + }, + { + "epoch": 0.4175282770106169, + "grad_norm": 4.6913299560546875, + "learning_rate": 0.00019711307952137994, + "loss": 0.9253, + "step": 6017 + }, + { + "epoch": 0.41759766844771357, + "grad_norm": 3.2553787231445312, + "learning_rate": 0.00019711173906060897, + "loss": 0.9472, + "step": 6018 + }, + { + "epoch": 0.41766705988481023, + "grad_norm": 3.342379570007324, + "learning_rate": 0.00019711039829326752, + "loss": 1.0826, + "step": 6019 + }, + { + "epoch": 0.4177364513219069, + "grad_norm": 4.058983325958252, + "learning_rate": 0.0001971090572193598, + "loss": 0.8447, + "step": 6020 + }, + { + "epoch": 0.41780584275900357, + "grad_norm": 3.2887256145477295, + "learning_rate": 0.00019710771583889006, + "loss": 1.0869, + "step": 6021 + }, + { + "epoch": 0.4178752341961002, + "grad_norm": 3.601410388946533, + "learning_rate": 0.00019710637415186252, + "loss": 0.6927, + "step": 6022 + }, + { + "epoch": 0.41794462563319684, + "grad_norm": 3.550737142562866, + "learning_rate": 0.00019710503215828143, + "loss": 0.8262, + "step": 6023 + }, + { + "epoch": 0.4180140170702935, + "grad_norm": 3.3008053302764893, + "learning_rate": 0.00019710368985815103, + "loss": 1.0896, + "step": 6024 + }, + { + "epoch": 0.4180834085073902, + "grad_norm": 2.7198853492736816, + "learning_rate": 0.0001971023472514755, + "loss": 1.011, + "step": 6025 + }, + { + "epoch": 0.41815279994448684, + "grad_norm": 3.5310590267181396, + "learning_rate": 0.0001971010043382592, + "loss": 1.142, + "step": 6026 + }, + { + "epoch": 0.4182221913815835, + "grad_norm": 3.4228267669677734, + "learning_rate": 0.00019709966111850624, + "loss": 0.8962, + "step": 6027 + }, + { + "epoch": 0.4182915828186802, + "grad_norm": 4.557342529296875, + "learning_rate": 0.00019709831759222093, + "loss": 1.2163, + "step": 6028 + }, + { + "epoch": 0.41836097425577684, + "grad_norm": 3.3362011909484863, + "learning_rate": 0.0001970969737594075, + "loss": 0.7324, + "step": 6029 + }, + { + "epoch": 0.4184303656928735, + "grad_norm": 3.4070098400115967, + "learning_rate": 0.0001970956296200702, + "loss": 0.9913, + "step": 6030 + }, + { + "epoch": 0.4184997571299702, + "grad_norm": 2.59580659866333, + "learning_rate": 0.00019709428517421324, + "loss": 1.0712, + "step": 6031 + }, + { + "epoch": 0.41856914856706684, + "grad_norm": 3.052476406097412, + "learning_rate": 0.0001970929404218409, + "loss": 0.9848, + "step": 6032 + }, + { + "epoch": 0.4186385400041635, + "grad_norm": 3.135343074798584, + "learning_rate": 0.0001970915953629574, + "loss": 1.0269, + "step": 6033 + }, + { + "epoch": 0.4187079314412602, + "grad_norm": 3.1391713619232178, + "learning_rate": 0.000197090249997567, + "loss": 1.201, + "step": 6034 + }, + { + "epoch": 0.4187773228783568, + "grad_norm": 3.376195192337036, + "learning_rate": 0.00019708890432567396, + "loss": 1.2452, + "step": 6035 + }, + { + "epoch": 0.41884671431545345, + "grad_norm": 3.356720447540283, + "learning_rate": 0.00019708755834728253, + "loss": 1.1307, + "step": 6036 + }, + { + "epoch": 0.4189161057525501, + "grad_norm": 3.3014848232269287, + "learning_rate": 0.0001970862120623969, + "loss": 1.0088, + "step": 6037 + }, + { + "epoch": 0.4189854971896468, + "grad_norm": 2.6666321754455566, + "learning_rate": 0.0001970848654710214, + "loss": 1.0497, + "step": 6038 + }, + { + "epoch": 0.41905488862674345, + "grad_norm": 2.5621285438537598, + "learning_rate": 0.0001970835185731602, + "loss": 0.8921, + "step": 6039 + }, + { + "epoch": 0.4191242800638401, + "grad_norm": 3.1833627223968506, + "learning_rate": 0.0001970821713688176, + "loss": 1.0609, + "step": 6040 + }, + { + "epoch": 0.4191936715009368, + "grad_norm": 2.7341651916503906, + "learning_rate": 0.00019708082385799789, + "loss": 1.0966, + "step": 6041 + }, + { + "epoch": 0.41926306293803345, + "grad_norm": 2.7857563495635986, + "learning_rate": 0.00019707947604070527, + "loss": 0.8672, + "step": 6042 + }, + { + "epoch": 0.4193324543751301, + "grad_norm": 2.829226493835449, + "learning_rate": 0.000197078127916944, + "loss": 1.1176, + "step": 6043 + }, + { + "epoch": 0.4194018458122268, + "grad_norm": 2.981151819229126, + "learning_rate": 0.00019707677948671833, + "loss": 0.9857, + "step": 6044 + }, + { + "epoch": 0.41947123724932345, + "grad_norm": 2.187150001525879, + "learning_rate": 0.00019707543075003257, + "loss": 0.6607, + "step": 6045 + }, + { + "epoch": 0.4195406286864201, + "grad_norm": 3.7336645126342773, + "learning_rate": 0.0001970740817068909, + "loss": 0.8377, + "step": 6046 + }, + { + "epoch": 0.4196100201235168, + "grad_norm": 2.7303316593170166, + "learning_rate": 0.00019707273235729762, + "loss": 1.0475, + "step": 6047 + }, + { + "epoch": 0.41967941156061345, + "grad_norm": 3.327167510986328, + "learning_rate": 0.000197071382701257, + "loss": 0.9735, + "step": 6048 + }, + { + "epoch": 0.41974880299771006, + "grad_norm": 3.5620627403259277, + "learning_rate": 0.00019707003273877326, + "loss": 0.9885, + "step": 6049 + }, + { + "epoch": 0.4198181944348067, + "grad_norm": 2.9839298725128174, + "learning_rate": 0.00019706868246985074, + "loss": 1.055, + "step": 6050 + }, + { + "epoch": 0.4198875858719034, + "grad_norm": 2.916848659515381, + "learning_rate": 0.0001970673318944936, + "loss": 0.9578, + "step": 6051 + }, + { + "epoch": 0.41995697730900006, + "grad_norm": 3.8106648921966553, + "learning_rate": 0.0001970659810127062, + "loss": 0.9742, + "step": 6052 + }, + { + "epoch": 0.4200263687460967, + "grad_norm": 3.6596601009368896, + "learning_rate": 0.0001970646298244927, + "loss": 0.951, + "step": 6053 + }, + { + "epoch": 0.4200957601831934, + "grad_norm": 3.1371519565582275, + "learning_rate": 0.00019706327832985746, + "loss": 0.924, + "step": 6054 + }, + { + "epoch": 0.42016515162029006, + "grad_norm": 2.562088966369629, + "learning_rate": 0.00019706192652880472, + "loss": 0.7384, + "step": 6055 + }, + { + "epoch": 0.4202345430573867, + "grad_norm": 2.979098320007324, + "learning_rate": 0.00019706057442133874, + "loss": 1.0502, + "step": 6056 + }, + { + "epoch": 0.4203039344944834, + "grad_norm": 3.435375690460205, + "learning_rate": 0.0001970592220074638, + "loss": 1.0169, + "step": 6057 + }, + { + "epoch": 0.42037332593158006, + "grad_norm": 3.2301437854766846, + "learning_rate": 0.0001970578692871841, + "loss": 1.0626, + "step": 6058 + }, + { + "epoch": 0.4204427173686767, + "grad_norm": 2.7236292362213135, + "learning_rate": 0.00019705651626050402, + "loss": 0.8516, + "step": 6059 + }, + { + "epoch": 0.4205121088057734, + "grad_norm": 2.3164596557617188, + "learning_rate": 0.00019705516292742775, + "loss": 0.8927, + "step": 6060 + }, + { + "epoch": 0.42058150024287005, + "grad_norm": 3.2599313259124756, + "learning_rate": 0.00019705380928795963, + "loss": 0.8649, + "step": 6061 + }, + { + "epoch": 0.42065089167996667, + "grad_norm": 2.7592742443084717, + "learning_rate": 0.00019705245534210386, + "loss": 0.9036, + "step": 6062 + }, + { + "epoch": 0.42072028311706333, + "grad_norm": 3.450486660003662, + "learning_rate": 0.00019705110108986476, + "loss": 0.7662, + "step": 6063 + }, + { + "epoch": 0.42078967455416, + "grad_norm": 2.5810861587524414, + "learning_rate": 0.0001970497465312466, + "loss": 0.8777, + "step": 6064 + }, + { + "epoch": 0.42085906599125666, + "grad_norm": 2.923102617263794, + "learning_rate": 0.00019704839166625364, + "loss": 1.117, + "step": 6065 + }, + { + "epoch": 0.42092845742835333, + "grad_norm": 3.9374277591705322, + "learning_rate": 0.00019704703649489017, + "loss": 0.9241, + "step": 6066 + }, + { + "epoch": 0.42099784886545, + "grad_norm": 5.8078694343566895, + "learning_rate": 0.00019704568101716046, + "loss": 0.9519, + "step": 6067 + }, + { + "epoch": 0.42106724030254666, + "grad_norm": 3.2277121543884277, + "learning_rate": 0.00019704432523306878, + "loss": 1.1873, + "step": 6068 + }, + { + "epoch": 0.42113663173964333, + "grad_norm": 2.6571013927459717, + "learning_rate": 0.00019704296914261943, + "loss": 1.0476, + "step": 6069 + }, + { + "epoch": 0.42120602317674, + "grad_norm": 3.363589286804199, + "learning_rate": 0.00019704161274581672, + "loss": 0.8729, + "step": 6070 + }, + { + "epoch": 0.42127541461383666, + "grad_norm": 3.1271708011627197, + "learning_rate": 0.00019704025604266488, + "loss": 1.1065, + "step": 6071 + }, + { + "epoch": 0.42134480605093333, + "grad_norm": 2.060199022293091, + "learning_rate": 0.0001970388990331682, + "loss": 0.9183, + "step": 6072 + }, + { + "epoch": 0.42141419748803, + "grad_norm": 3.2676315307617188, + "learning_rate": 0.00019703754171733096, + "loss": 1.0797, + "step": 6073 + }, + { + "epoch": 0.42148358892512666, + "grad_norm": 3.008570909500122, + "learning_rate": 0.00019703618409515748, + "loss": 0.9854, + "step": 6074 + }, + { + "epoch": 0.42155298036222333, + "grad_norm": 3.1024861335754395, + "learning_rate": 0.00019703482616665206, + "loss": 0.9621, + "step": 6075 + }, + { + "epoch": 0.42162237179931994, + "grad_norm": 3.211085319519043, + "learning_rate": 0.0001970334679318189, + "loss": 1.2004, + "step": 6076 + }, + { + "epoch": 0.4216917632364166, + "grad_norm": 2.633146286010742, + "learning_rate": 0.00019703210939066236, + "loss": 0.9532, + "step": 6077 + }, + { + "epoch": 0.42176115467351327, + "grad_norm": 2.9336936473846436, + "learning_rate": 0.0001970307505431867, + "loss": 0.9449, + "step": 6078 + }, + { + "epoch": 0.42183054611060994, + "grad_norm": 3.3949973583221436, + "learning_rate": 0.00019702939138939622, + "loss": 1.0376, + "step": 6079 + }, + { + "epoch": 0.4218999375477066, + "grad_norm": 3.151284694671631, + "learning_rate": 0.00019702803192929522, + "loss": 1.0006, + "step": 6080 + }, + { + "epoch": 0.42196932898480327, + "grad_norm": 2.9561469554901123, + "learning_rate": 0.00019702667216288798, + "loss": 0.8976, + "step": 6081 + }, + { + "epoch": 0.42203872042189994, + "grad_norm": 3.0273513793945312, + "learning_rate": 0.0001970253120901788, + "loss": 1.098, + "step": 6082 + }, + { + "epoch": 0.4221081118589966, + "grad_norm": 4.919175624847412, + "learning_rate": 0.00019702395171117192, + "loss": 1.1586, + "step": 6083 + }, + { + "epoch": 0.42217750329609327, + "grad_norm": 2.9556777477264404, + "learning_rate": 0.0001970225910258717, + "loss": 0.9717, + "step": 6084 + }, + { + "epoch": 0.42224689473318994, + "grad_norm": 3.1692750453948975, + "learning_rate": 0.00019702123003428244, + "loss": 1.0906, + "step": 6085 + }, + { + "epoch": 0.4223162861702866, + "grad_norm": 2.3081462383270264, + "learning_rate": 0.0001970198687364084, + "loss": 0.7803, + "step": 6086 + }, + { + "epoch": 0.42238567760738327, + "grad_norm": 2.5061001777648926, + "learning_rate": 0.0001970185071322539, + "loss": 0.8482, + "step": 6087 + }, + { + "epoch": 0.42245506904447994, + "grad_norm": 3.133549928665161, + "learning_rate": 0.00019701714522182324, + "loss": 1.1594, + "step": 6088 + }, + { + "epoch": 0.42252446048157655, + "grad_norm": 2.519164562225342, + "learning_rate": 0.00019701578300512069, + "loss": 0.8618, + "step": 6089 + }, + { + "epoch": 0.4225938519186732, + "grad_norm": 3.0924923419952393, + "learning_rate": 0.00019701442048215056, + "loss": 0.9858, + "step": 6090 + }, + { + "epoch": 0.4226632433557699, + "grad_norm": 3.3178646564483643, + "learning_rate": 0.00019701305765291717, + "loss": 0.9163, + "step": 6091 + }, + { + "epoch": 0.42273263479286655, + "grad_norm": 3.3581528663635254, + "learning_rate": 0.00019701169451742483, + "loss": 0.9403, + "step": 6092 + }, + { + "epoch": 0.4228020262299632, + "grad_norm": 2.4227523803710938, + "learning_rate": 0.00019701033107567782, + "loss": 0.7155, + "step": 6093 + }, + { + "epoch": 0.4228714176670599, + "grad_norm": 2.821296215057373, + "learning_rate": 0.00019700896732768042, + "loss": 0.9056, + "step": 6094 + }, + { + "epoch": 0.42294080910415655, + "grad_norm": 3.4891934394836426, + "learning_rate": 0.000197007603273437, + "loss": 1.0091, + "step": 6095 + }, + { + "epoch": 0.4230102005412532, + "grad_norm": 3.329277992248535, + "learning_rate": 0.0001970062389129518, + "loss": 1.0384, + "step": 6096 + }, + { + "epoch": 0.4230795919783499, + "grad_norm": 2.775672197341919, + "learning_rate": 0.00019700487424622917, + "loss": 0.8385, + "step": 6097 + }, + { + "epoch": 0.42314898341544654, + "grad_norm": 2.688293695449829, + "learning_rate": 0.0001970035092732734, + "loss": 0.7482, + "step": 6098 + }, + { + "epoch": 0.4232183748525432, + "grad_norm": 3.555255651473999, + "learning_rate": 0.00019700214399408884, + "loss": 0.8842, + "step": 6099 + }, + { + "epoch": 0.4232877662896399, + "grad_norm": 3.3801984786987305, + "learning_rate": 0.00019700077840867974, + "loss": 1.1558, + "step": 6100 + }, + { + "epoch": 0.42335715772673654, + "grad_norm": 3.5175986289978027, + "learning_rate": 0.00019699941251705044, + "loss": 1.0766, + "step": 6101 + }, + { + "epoch": 0.4234265491638332, + "grad_norm": 3.049835443496704, + "learning_rate": 0.00019699804631920524, + "loss": 1.0524, + "step": 6102 + }, + { + "epoch": 0.4234959406009298, + "grad_norm": 3.883544921875, + "learning_rate": 0.0001969966798151485, + "loss": 0.986, + "step": 6103 + }, + { + "epoch": 0.4235653320380265, + "grad_norm": 1.9785995483398438, + "learning_rate": 0.00019699531300488446, + "loss": 0.7305, + "step": 6104 + }, + { + "epoch": 0.42363472347512315, + "grad_norm": 2.6502559185028076, + "learning_rate": 0.00019699394588841747, + "loss": 0.8803, + "step": 6105 + }, + { + "epoch": 0.4237041149122198, + "grad_norm": 2.8886024951934814, + "learning_rate": 0.00019699257846575186, + "loss": 0.9608, + "step": 6106 + }, + { + "epoch": 0.4237735063493165, + "grad_norm": 2.8755619525909424, + "learning_rate": 0.0001969912107368919, + "loss": 0.8038, + "step": 6107 + }, + { + "epoch": 0.42384289778641315, + "grad_norm": 2.3251304626464844, + "learning_rate": 0.00019698984270184196, + "loss": 0.9608, + "step": 6108 + }, + { + "epoch": 0.4239122892235098, + "grad_norm": 3.780238389968872, + "learning_rate": 0.00019698847436060637, + "loss": 1.0403, + "step": 6109 + }, + { + "epoch": 0.4239816806606065, + "grad_norm": 2.744504451751709, + "learning_rate": 0.0001969871057131894, + "loss": 1.0669, + "step": 6110 + }, + { + "epoch": 0.42405107209770315, + "grad_norm": 3.461327075958252, + "learning_rate": 0.00019698573675959536, + "loss": 1.0606, + "step": 6111 + }, + { + "epoch": 0.4241204635347998, + "grad_norm": 3.0870795249938965, + "learning_rate": 0.00019698436749982862, + "loss": 0.9587, + "step": 6112 + }, + { + "epoch": 0.4241898549718965, + "grad_norm": 2.188856601715088, + "learning_rate": 0.0001969829979338935, + "loss": 0.7911, + "step": 6113 + }, + { + "epoch": 0.42425924640899315, + "grad_norm": 3.575716733932495, + "learning_rate": 0.0001969816280617943, + "loss": 1.4179, + "step": 6114 + }, + { + "epoch": 0.4243286378460898, + "grad_norm": 2.733428716659546, + "learning_rate": 0.00019698025788353536, + "loss": 0.8393, + "step": 6115 + }, + { + "epoch": 0.42439802928318643, + "grad_norm": 2.954387664794922, + "learning_rate": 0.00019697888739912098, + "loss": 1.0754, + "step": 6116 + }, + { + "epoch": 0.4244674207202831, + "grad_norm": 6.366195201873779, + "learning_rate": 0.0001969775166085555, + "loss": 1.2566, + "step": 6117 + }, + { + "epoch": 0.42453681215737976, + "grad_norm": 2.624629497528076, + "learning_rate": 0.0001969761455118433, + "loss": 1.0929, + "step": 6118 + }, + { + "epoch": 0.42460620359447643, + "grad_norm": 3.298185110092163, + "learning_rate": 0.0001969747741089886, + "loss": 0.9512, + "step": 6119 + }, + { + "epoch": 0.4246755950315731, + "grad_norm": 3.1147215366363525, + "learning_rate": 0.0001969734023999958, + "loss": 0.9361, + "step": 6120 + }, + { + "epoch": 0.42474498646866976, + "grad_norm": 2.8864824771881104, + "learning_rate": 0.00019697203038486927, + "loss": 1.0334, + "step": 6121 + }, + { + "epoch": 0.4248143779057664, + "grad_norm": 3.037407159805298, + "learning_rate": 0.00019697065806361324, + "loss": 0.9309, + "step": 6122 + }, + { + "epoch": 0.4248837693428631, + "grad_norm": 5.296887397766113, + "learning_rate": 0.0001969692854362321, + "loss": 0.8085, + "step": 6123 + }, + { + "epoch": 0.42495316077995976, + "grad_norm": 2.415527820587158, + "learning_rate": 0.00019696791250273016, + "loss": 0.8766, + "step": 6124 + }, + { + "epoch": 0.4250225522170564, + "grad_norm": 3.4570517539978027, + "learning_rate": 0.0001969665392631118, + "loss": 0.9543, + "step": 6125 + }, + { + "epoch": 0.4250919436541531, + "grad_norm": 2.594125747680664, + "learning_rate": 0.00019696516571738134, + "loss": 0.848, + "step": 6126 + }, + { + "epoch": 0.42516133509124976, + "grad_norm": 3.6909632682800293, + "learning_rate": 0.00019696379186554309, + "loss": 1.2966, + "step": 6127 + }, + { + "epoch": 0.4252307265283464, + "grad_norm": 2.17818546295166, + "learning_rate": 0.00019696241770760138, + "loss": 0.7672, + "step": 6128 + }, + { + "epoch": 0.4253001179654431, + "grad_norm": 3.179279088973999, + "learning_rate": 0.00019696104324356056, + "loss": 1.2329, + "step": 6129 + }, + { + "epoch": 0.4253695094025397, + "grad_norm": 2.824528217315674, + "learning_rate": 0.000196959668473425, + "loss": 0.9054, + "step": 6130 + }, + { + "epoch": 0.42543890083963637, + "grad_norm": 2.9904630184173584, + "learning_rate": 0.000196958293397199, + "loss": 0.9576, + "step": 6131 + }, + { + "epoch": 0.42550829227673304, + "grad_norm": 2.603685140609741, + "learning_rate": 0.00019695691801488693, + "loss": 0.8374, + "step": 6132 + }, + { + "epoch": 0.4255776837138297, + "grad_norm": 3.042240858078003, + "learning_rate": 0.00019695554232649314, + "loss": 0.9134, + "step": 6133 + }, + { + "epoch": 0.42564707515092637, + "grad_norm": 3.6199162006378174, + "learning_rate": 0.00019695416633202194, + "loss": 0.9459, + "step": 6134 + }, + { + "epoch": 0.42571646658802303, + "grad_norm": 4.12022066116333, + "learning_rate": 0.00019695279003147765, + "loss": 1.0176, + "step": 6135 + }, + { + "epoch": 0.4257858580251197, + "grad_norm": 3.7532105445861816, + "learning_rate": 0.00019695141342486466, + "loss": 0.7934, + "step": 6136 + }, + { + "epoch": 0.42585524946221637, + "grad_norm": 2.6831159591674805, + "learning_rate": 0.00019695003651218733, + "loss": 0.7028, + "step": 6137 + }, + { + "epoch": 0.42592464089931303, + "grad_norm": 3.2706665992736816, + "learning_rate": 0.00019694865929344998, + "loss": 0.8832, + "step": 6138 + }, + { + "epoch": 0.4259940323364097, + "grad_norm": 4.032263278961182, + "learning_rate": 0.00019694728176865693, + "loss": 0.7413, + "step": 6139 + }, + { + "epoch": 0.42606342377350637, + "grad_norm": 2.793184995651245, + "learning_rate": 0.0001969459039378126, + "loss": 0.8779, + "step": 6140 + }, + { + "epoch": 0.42613281521060303, + "grad_norm": 2.8454158306121826, + "learning_rate": 0.00019694452580092125, + "loss": 0.8623, + "step": 6141 + }, + { + "epoch": 0.4262022066476997, + "grad_norm": 2.8549537658691406, + "learning_rate": 0.00019694314735798735, + "loss": 0.8816, + "step": 6142 + }, + { + "epoch": 0.4262715980847963, + "grad_norm": 2.6669070720672607, + "learning_rate": 0.00019694176860901515, + "loss": 0.7734, + "step": 6143 + }, + { + "epoch": 0.426340989521893, + "grad_norm": 3.87622332572937, + "learning_rate": 0.000196940389554009, + "loss": 0.9737, + "step": 6144 + }, + { + "epoch": 0.42641038095898964, + "grad_norm": 3.1834609508514404, + "learning_rate": 0.00019693901019297332, + "loss": 0.8979, + "step": 6145 + }, + { + "epoch": 0.4264797723960863, + "grad_norm": 2.70670485496521, + "learning_rate": 0.00019693763052591244, + "loss": 0.844, + "step": 6146 + }, + { + "epoch": 0.426549163833183, + "grad_norm": 2.710456609725952, + "learning_rate": 0.00019693625055283067, + "loss": 1.0559, + "step": 6147 + }, + { + "epoch": 0.42661855527027964, + "grad_norm": 3.7604806423187256, + "learning_rate": 0.00019693487027373243, + "loss": 1.0436, + "step": 6148 + }, + { + "epoch": 0.4266879467073763, + "grad_norm": 3.497281074523926, + "learning_rate": 0.00019693348968862203, + "loss": 0.9519, + "step": 6149 + }, + { + "epoch": 0.426757338144473, + "grad_norm": 2.710728883743286, + "learning_rate": 0.0001969321087975039, + "loss": 0.9315, + "step": 6150 + }, + { + "epoch": 0.42682672958156964, + "grad_norm": 2.2658376693725586, + "learning_rate": 0.00019693072760038232, + "loss": 0.7247, + "step": 6151 + }, + { + "epoch": 0.4268961210186663, + "grad_norm": 3.0925965309143066, + "learning_rate": 0.00019692934609726165, + "loss": 1.0343, + "step": 6152 + }, + { + "epoch": 0.426965512455763, + "grad_norm": 2.655409574508667, + "learning_rate": 0.0001969279642881463, + "loss": 0.9284, + "step": 6153 + }, + { + "epoch": 0.42703490389285964, + "grad_norm": 2.7057406902313232, + "learning_rate": 0.00019692658217304064, + "loss": 1.0065, + "step": 6154 + }, + { + "epoch": 0.4271042953299563, + "grad_norm": 2.191617727279663, + "learning_rate": 0.00019692519975194895, + "loss": 1.03, + "step": 6155 + }, + { + "epoch": 0.4271736867670529, + "grad_norm": 3.2223927974700928, + "learning_rate": 0.00019692381702487573, + "loss": 0.8976, + "step": 6156 + }, + { + "epoch": 0.4272430782041496, + "grad_norm": 1.9601950645446777, + "learning_rate": 0.0001969224339918252, + "loss": 0.811, + "step": 6157 + }, + { + "epoch": 0.42731246964124625, + "grad_norm": 2.5894858837127686, + "learning_rate": 0.00019692105065280183, + "loss": 0.8495, + "step": 6158 + }, + { + "epoch": 0.4273818610783429, + "grad_norm": 2.8311140537261963, + "learning_rate": 0.0001969196670078099, + "loss": 1.098, + "step": 6159 + }, + { + "epoch": 0.4274512525154396, + "grad_norm": 2.59063982963562, + "learning_rate": 0.00019691828305685384, + "loss": 0.9187, + "step": 6160 + }, + { + "epoch": 0.42752064395253625, + "grad_norm": 3.1889946460723877, + "learning_rate": 0.00019691689879993804, + "loss": 0.9687, + "step": 6161 + }, + { + "epoch": 0.4275900353896329, + "grad_norm": 2.810166358947754, + "learning_rate": 0.0001969155142370668, + "loss": 0.9065, + "step": 6162 + }, + { + "epoch": 0.4276594268267296, + "grad_norm": 3.3184518814086914, + "learning_rate": 0.00019691412936824454, + "loss": 0.941, + "step": 6163 + }, + { + "epoch": 0.42772881826382625, + "grad_norm": 3.692425489425659, + "learning_rate": 0.00019691274419347563, + "loss": 0.9889, + "step": 6164 + }, + { + "epoch": 0.4277982097009229, + "grad_norm": 3.3161230087280273, + "learning_rate": 0.0001969113587127644, + "loss": 0.888, + "step": 6165 + }, + { + "epoch": 0.4278676011380196, + "grad_norm": 2.995816469192505, + "learning_rate": 0.0001969099729261153, + "loss": 0.9087, + "step": 6166 + }, + { + "epoch": 0.42793699257511625, + "grad_norm": 3.4038164615631104, + "learning_rate": 0.00019690858683353262, + "loss": 1.1419, + "step": 6167 + }, + { + "epoch": 0.4280063840122129, + "grad_norm": 3.0430145263671875, + "learning_rate": 0.00019690720043502076, + "loss": 0.9404, + "step": 6168 + }, + { + "epoch": 0.4280757754493096, + "grad_norm": 5.5619120597839355, + "learning_rate": 0.00019690581373058415, + "loss": 0.9188, + "step": 6169 + }, + { + "epoch": 0.4281451668864062, + "grad_norm": 2.603193998336792, + "learning_rate": 0.0001969044267202271, + "loss": 1.0848, + "step": 6170 + }, + { + "epoch": 0.42821455832350286, + "grad_norm": 2.6568007469177246, + "learning_rate": 0.00019690303940395404, + "loss": 0.8252, + "step": 6171 + }, + { + "epoch": 0.4282839497605995, + "grad_norm": 2.369166374206543, + "learning_rate": 0.00019690165178176932, + "loss": 0.8504, + "step": 6172 + }, + { + "epoch": 0.4283533411976962, + "grad_norm": 2.6914849281311035, + "learning_rate": 0.00019690026385367734, + "loss": 1.0498, + "step": 6173 + }, + { + "epoch": 0.42842273263479286, + "grad_norm": 2.6496529579162598, + "learning_rate": 0.00019689887561968246, + "loss": 0.8677, + "step": 6174 + }, + { + "epoch": 0.4284921240718895, + "grad_norm": 3.160055160522461, + "learning_rate": 0.00019689748707978907, + "loss": 0.8417, + "step": 6175 + }, + { + "epoch": 0.4285615155089862, + "grad_norm": 2.5076792240142822, + "learning_rate": 0.00019689609823400153, + "loss": 0.9427, + "step": 6176 + }, + { + "epoch": 0.42863090694608286, + "grad_norm": 3.1147372722625732, + "learning_rate": 0.0001968947090823243, + "loss": 1.0707, + "step": 6177 + }, + { + "epoch": 0.4287002983831795, + "grad_norm": 2.7763278484344482, + "learning_rate": 0.00019689331962476168, + "loss": 0.8691, + "step": 6178 + }, + { + "epoch": 0.4287696898202762, + "grad_norm": 3.181269884109497, + "learning_rate": 0.0001968919298613181, + "loss": 1.0538, + "step": 6179 + }, + { + "epoch": 0.42883908125737286, + "grad_norm": 2.778170585632324, + "learning_rate": 0.00019689053979199796, + "loss": 0.8328, + "step": 6180 + }, + { + "epoch": 0.4289084726944695, + "grad_norm": 4.8411688804626465, + "learning_rate": 0.00019688914941680562, + "loss": 0.9822, + "step": 6181 + }, + { + "epoch": 0.4289778641315662, + "grad_norm": 3.0742502212524414, + "learning_rate": 0.00019688775873574543, + "loss": 1.0853, + "step": 6182 + }, + { + "epoch": 0.4290472555686628, + "grad_norm": 2.5294342041015625, + "learning_rate": 0.00019688636774882188, + "loss": 0.7353, + "step": 6183 + }, + { + "epoch": 0.42911664700575947, + "grad_norm": 2.676541566848755, + "learning_rate": 0.0001968849764560393, + "loss": 0.8548, + "step": 6184 + }, + { + "epoch": 0.42918603844285613, + "grad_norm": 3.867705821990967, + "learning_rate": 0.00019688358485740208, + "loss": 1.1407, + "step": 6185 + }, + { + "epoch": 0.4292554298799528, + "grad_norm": 3.5285825729370117, + "learning_rate": 0.0001968821929529146, + "loss": 0.9109, + "step": 6186 + }, + { + "epoch": 0.42932482131704947, + "grad_norm": 3.742248773574829, + "learning_rate": 0.00019688080074258132, + "loss": 1.4518, + "step": 6187 + }, + { + "epoch": 0.42939421275414613, + "grad_norm": 3.2983341217041016, + "learning_rate": 0.00019687940822640657, + "loss": 0.8519, + "step": 6188 + }, + { + "epoch": 0.4294636041912428, + "grad_norm": 3.2639338970184326, + "learning_rate": 0.00019687801540439477, + "loss": 0.8513, + "step": 6189 + }, + { + "epoch": 0.42953299562833946, + "grad_norm": 2.7205092906951904, + "learning_rate": 0.0001968766222765503, + "loss": 0.8736, + "step": 6190 + }, + { + "epoch": 0.42960238706543613, + "grad_norm": 2.979531764984131, + "learning_rate": 0.0001968752288428776, + "loss": 0.7277, + "step": 6191 + }, + { + "epoch": 0.4296717785025328, + "grad_norm": 3.3908298015594482, + "learning_rate": 0.00019687383510338106, + "loss": 1.0103, + "step": 6192 + }, + { + "epoch": 0.42974116993962946, + "grad_norm": 2.514662742614746, + "learning_rate": 0.00019687244105806502, + "loss": 0.8404, + "step": 6193 + }, + { + "epoch": 0.42981056137672613, + "grad_norm": 3.068877935409546, + "learning_rate": 0.0001968710467069339, + "loss": 0.8828, + "step": 6194 + }, + { + "epoch": 0.4298799528138228, + "grad_norm": 2.9117910861968994, + "learning_rate": 0.0001968696520499922, + "loss": 1.2022, + "step": 6195 + }, + { + "epoch": 0.42994934425091946, + "grad_norm": 2.6746666431427, + "learning_rate": 0.0001968682570872442, + "loss": 0.9862, + "step": 6196 + }, + { + "epoch": 0.4300187356880161, + "grad_norm": 3.473644733428955, + "learning_rate": 0.00019686686181869437, + "loss": 0.8834, + "step": 6197 + }, + { + "epoch": 0.43008812712511274, + "grad_norm": 3.356349229812622, + "learning_rate": 0.00019686546624434706, + "loss": 0.9284, + "step": 6198 + }, + { + "epoch": 0.4301575185622094, + "grad_norm": 3.050150156021118, + "learning_rate": 0.00019686407036420675, + "loss": 1.0812, + "step": 6199 + }, + { + "epoch": 0.4302269099993061, + "grad_norm": 2.402195930480957, + "learning_rate": 0.0001968626741782778, + "loss": 0.8027, + "step": 6200 + }, + { + "epoch": 0.43029630143640274, + "grad_norm": 2.514846086502075, + "learning_rate": 0.0001968612776865646, + "loss": 0.9303, + "step": 6201 + }, + { + "epoch": 0.4303656928734994, + "grad_norm": 2.476498603820801, + "learning_rate": 0.00019685988088907164, + "loss": 0.9016, + "step": 6202 + }, + { + "epoch": 0.43043508431059607, + "grad_norm": 3.2896203994750977, + "learning_rate": 0.00019685848378580323, + "loss": 0.787, + "step": 6203 + }, + { + "epoch": 0.43050447574769274, + "grad_norm": 3.5042755603790283, + "learning_rate": 0.00019685708637676383, + "loss": 0.9592, + "step": 6204 + }, + { + "epoch": 0.4305738671847894, + "grad_norm": 3.5030646324157715, + "learning_rate": 0.00019685568866195785, + "loss": 1.2949, + "step": 6205 + }, + { + "epoch": 0.43064325862188607, + "grad_norm": 3.770313262939453, + "learning_rate": 0.0001968542906413897, + "loss": 1.0069, + "step": 6206 + }, + { + "epoch": 0.43071265005898274, + "grad_norm": 2.4113969802856445, + "learning_rate": 0.0001968528923150638, + "loss": 0.8355, + "step": 6207 + }, + { + "epoch": 0.4307820414960794, + "grad_norm": 3.3012776374816895, + "learning_rate": 0.00019685149368298452, + "loss": 1.0954, + "step": 6208 + }, + { + "epoch": 0.43085143293317607, + "grad_norm": 3.1991751194000244, + "learning_rate": 0.00019685009474515637, + "loss": 0.8739, + "step": 6209 + }, + { + "epoch": 0.4309208243702727, + "grad_norm": 2.7161340713500977, + "learning_rate": 0.00019684869550158366, + "loss": 0.9826, + "step": 6210 + }, + { + "epoch": 0.43099021580736935, + "grad_norm": 2.804100275039673, + "learning_rate": 0.00019684729595227088, + "loss": 0.9877, + "step": 6211 + }, + { + "epoch": 0.431059607244466, + "grad_norm": 2.611879348754883, + "learning_rate": 0.0001968458960972224, + "loss": 0.9819, + "step": 6212 + }, + { + "epoch": 0.4311289986815627, + "grad_norm": 3.5294246673583984, + "learning_rate": 0.00019684449593644268, + "loss": 1.2203, + "step": 6213 + }, + { + "epoch": 0.43119839011865935, + "grad_norm": 3.127340316772461, + "learning_rate": 0.00019684309546993611, + "loss": 0.9517, + "step": 6214 + }, + { + "epoch": 0.431267781555756, + "grad_norm": 2.0893378257751465, + "learning_rate": 0.00019684169469770714, + "loss": 0.9167, + "step": 6215 + }, + { + "epoch": 0.4313371729928527, + "grad_norm": 2.8395402431488037, + "learning_rate": 0.00019684029361976014, + "loss": 0.8979, + "step": 6216 + }, + { + "epoch": 0.43140656442994935, + "grad_norm": 3.3750569820404053, + "learning_rate": 0.0001968388922360996, + "loss": 1.0831, + "step": 6217 + }, + { + "epoch": 0.431475955867046, + "grad_norm": 2.972177505493164, + "learning_rate": 0.00019683749054672992, + "loss": 0.8262, + "step": 6218 + }, + { + "epoch": 0.4315453473041427, + "grad_norm": 4.334634780883789, + "learning_rate": 0.0001968360885516555, + "loss": 1.0999, + "step": 6219 + }, + { + "epoch": 0.43161473874123935, + "grad_norm": 3.3005683422088623, + "learning_rate": 0.0001968346862508808, + "loss": 1.1322, + "step": 6220 + }, + { + "epoch": 0.431684130178336, + "grad_norm": 3.260749101638794, + "learning_rate": 0.00019683328364441022, + "loss": 1.1743, + "step": 6221 + }, + { + "epoch": 0.4317535216154327, + "grad_norm": 2.7741050720214844, + "learning_rate": 0.00019683188073224819, + "loss": 1.0329, + "step": 6222 + }, + { + "epoch": 0.43182291305252934, + "grad_norm": 3.6526825428009033, + "learning_rate": 0.00019683047751439917, + "loss": 1.0373, + "step": 6223 + }, + { + "epoch": 0.43189230448962596, + "grad_norm": 2.6831085681915283, + "learning_rate": 0.00019682907399086753, + "loss": 0.6561, + "step": 6224 + }, + { + "epoch": 0.4319616959267226, + "grad_norm": 2.7123911380767822, + "learning_rate": 0.00019682767016165778, + "loss": 0.7965, + "step": 6225 + }, + { + "epoch": 0.4320310873638193, + "grad_norm": 3.2210633754730225, + "learning_rate": 0.00019682626602677427, + "loss": 1.1235, + "step": 6226 + }, + { + "epoch": 0.43210047880091595, + "grad_norm": 3.7424941062927246, + "learning_rate": 0.00019682486158622151, + "loss": 1.1684, + "step": 6227 + }, + { + "epoch": 0.4321698702380126, + "grad_norm": 3.5696322917938232, + "learning_rate": 0.00019682345684000384, + "loss": 0.9479, + "step": 6228 + }, + { + "epoch": 0.4322392616751093, + "grad_norm": 3.3676226139068604, + "learning_rate": 0.0001968220517881258, + "loss": 1.0018, + "step": 6229 + }, + { + "epoch": 0.43230865311220595, + "grad_norm": 3.6725919246673584, + "learning_rate": 0.00019682064643059178, + "loss": 0.9046, + "step": 6230 + }, + { + "epoch": 0.4323780445493026, + "grad_norm": 3.0949466228485107, + "learning_rate": 0.0001968192407674062, + "loss": 0.7584, + "step": 6231 + }, + { + "epoch": 0.4324474359863993, + "grad_norm": 3.489509105682373, + "learning_rate": 0.00019681783479857349, + "loss": 1.1458, + "step": 6232 + }, + { + "epoch": 0.43251682742349595, + "grad_norm": 3.243861675262451, + "learning_rate": 0.00019681642852409812, + "loss": 0.864, + "step": 6233 + }, + { + "epoch": 0.4325862188605926, + "grad_norm": 3.1255884170532227, + "learning_rate": 0.0001968150219439845, + "loss": 0.7566, + "step": 6234 + }, + { + "epoch": 0.4326556102976893, + "grad_norm": 3.840794801712036, + "learning_rate": 0.00019681361505823714, + "loss": 1.0327, + "step": 6235 + }, + { + "epoch": 0.43272500173478595, + "grad_norm": 3.6920745372772217, + "learning_rate": 0.00019681220786686037, + "loss": 1.3526, + "step": 6236 + }, + { + "epoch": 0.43279439317188256, + "grad_norm": 2.9851479530334473, + "learning_rate": 0.00019681080036985872, + "loss": 0.9091, + "step": 6237 + }, + { + "epoch": 0.43286378460897923, + "grad_norm": 2.9348292350769043, + "learning_rate": 0.00019680939256723663, + "loss": 1.0965, + "step": 6238 + }, + { + "epoch": 0.4329331760460759, + "grad_norm": 3.4595558643341064, + "learning_rate": 0.0001968079844589985, + "loss": 1.1168, + "step": 6239 + }, + { + "epoch": 0.43300256748317256, + "grad_norm": 3.657865047454834, + "learning_rate": 0.00019680657604514878, + "loss": 1.0836, + "step": 6240 + }, + { + "epoch": 0.43307195892026923, + "grad_norm": 3.3007116317749023, + "learning_rate": 0.00019680516732569197, + "loss": 1.014, + "step": 6241 + }, + { + "epoch": 0.4331413503573659, + "grad_norm": 3.224368095397949, + "learning_rate": 0.00019680375830063242, + "loss": 0.8682, + "step": 6242 + }, + { + "epoch": 0.43321074179446256, + "grad_norm": 3.0105934143066406, + "learning_rate": 0.00019680234896997468, + "loss": 0.9336, + "step": 6243 + }, + { + "epoch": 0.43328013323155923, + "grad_norm": 4.186560153961182, + "learning_rate": 0.00019680093933372312, + "loss": 1.1354, + "step": 6244 + }, + { + "epoch": 0.4333495246686559, + "grad_norm": 2.7659342288970947, + "learning_rate": 0.00019679952939188227, + "loss": 0.8567, + "step": 6245 + }, + { + "epoch": 0.43341891610575256, + "grad_norm": 3.7039527893066406, + "learning_rate": 0.00019679811914445652, + "loss": 1.1009, + "step": 6246 + }, + { + "epoch": 0.4334883075428492, + "grad_norm": 2.307990312576294, + "learning_rate": 0.00019679670859145033, + "loss": 0.7654, + "step": 6247 + }, + { + "epoch": 0.4335576989799459, + "grad_norm": 3.489137887954712, + "learning_rate": 0.00019679529773286816, + "loss": 0.9772, + "step": 6248 + }, + { + "epoch": 0.43362709041704256, + "grad_norm": 2.3904478549957275, + "learning_rate": 0.00019679388656871448, + "loss": 0.7818, + "step": 6249 + }, + { + "epoch": 0.4336964818541392, + "grad_norm": 2.52748966217041, + "learning_rate": 0.0001967924750989937, + "loss": 0.8282, + "step": 6250 + }, + { + "epoch": 0.43376587329123584, + "grad_norm": 2.3743839263916016, + "learning_rate": 0.00019679106332371032, + "loss": 0.7697, + "step": 6251 + }, + { + "epoch": 0.4338352647283325, + "grad_norm": 2.995438575744629, + "learning_rate": 0.0001967896512428688, + "loss": 0.8223, + "step": 6252 + }, + { + "epoch": 0.43390465616542917, + "grad_norm": 3.1091666221618652, + "learning_rate": 0.00019678823885647357, + "loss": 0.7414, + "step": 6253 + }, + { + "epoch": 0.43397404760252584, + "grad_norm": 3.2957632541656494, + "learning_rate": 0.0001967868261645291, + "loss": 0.8549, + "step": 6254 + }, + { + "epoch": 0.4340434390396225, + "grad_norm": 3.1136906147003174, + "learning_rate": 0.00019678541316703985, + "loss": 0.8328, + "step": 6255 + }, + { + "epoch": 0.43411283047671917, + "grad_norm": 2.900843858718872, + "learning_rate": 0.0001967839998640103, + "loss": 0.6816, + "step": 6256 + }, + { + "epoch": 0.43418222191381584, + "grad_norm": 2.883868455886841, + "learning_rate": 0.00019678258625544483, + "loss": 0.8127, + "step": 6257 + }, + { + "epoch": 0.4342516133509125, + "grad_norm": 2.5046117305755615, + "learning_rate": 0.00019678117234134802, + "loss": 0.7446, + "step": 6258 + }, + { + "epoch": 0.43432100478800917, + "grad_norm": 3.507103204727173, + "learning_rate": 0.00019677975812172426, + "loss": 1.0498, + "step": 6259 + }, + { + "epoch": 0.43439039622510583, + "grad_norm": 4.359395503997803, + "learning_rate": 0.00019677834359657802, + "loss": 0.8676, + "step": 6260 + }, + { + "epoch": 0.4344597876622025, + "grad_norm": 3.481874942779541, + "learning_rate": 0.0001967769287659138, + "loss": 0.8839, + "step": 6261 + }, + { + "epoch": 0.43452917909929917, + "grad_norm": 3.1315412521362305, + "learning_rate": 0.00019677551362973605, + "loss": 0.8347, + "step": 6262 + }, + { + "epoch": 0.43459857053639583, + "grad_norm": 3.0357506275177, + "learning_rate": 0.00019677409818804923, + "loss": 1.1528, + "step": 6263 + }, + { + "epoch": 0.43466796197349244, + "grad_norm": 2.7366795539855957, + "learning_rate": 0.0001967726824408578, + "loss": 0.991, + "step": 6264 + }, + { + "epoch": 0.4347373534105891, + "grad_norm": 3.3508670330047607, + "learning_rate": 0.0001967712663881662, + "loss": 1.0879, + "step": 6265 + }, + { + "epoch": 0.4348067448476858, + "grad_norm": 2.307746171951294, + "learning_rate": 0.00019676985002997898, + "loss": 0.7988, + "step": 6266 + }, + { + "epoch": 0.43487613628478244, + "grad_norm": 2.2748992443084717, + "learning_rate": 0.00019676843336630058, + "loss": 1.0197, + "step": 6267 + }, + { + "epoch": 0.4349455277218791, + "grad_norm": 2.7089836597442627, + "learning_rate": 0.00019676701639713545, + "loss": 0.8798, + "step": 6268 + }, + { + "epoch": 0.4350149191589758, + "grad_norm": 2.889962673187256, + "learning_rate": 0.00019676559912248806, + "loss": 0.9274, + "step": 6269 + }, + { + "epoch": 0.43508431059607244, + "grad_norm": 3.6539440155029297, + "learning_rate": 0.00019676418154236292, + "loss": 1.2519, + "step": 6270 + }, + { + "epoch": 0.4351537020331691, + "grad_norm": 2.4002132415771484, + "learning_rate": 0.00019676276365676446, + "loss": 1.1169, + "step": 6271 + }, + { + "epoch": 0.4352230934702658, + "grad_norm": 2.5547375679016113, + "learning_rate": 0.00019676134546569722, + "loss": 1.1451, + "step": 6272 + }, + { + "epoch": 0.43529248490736244, + "grad_norm": 2.7645812034606934, + "learning_rate": 0.00019675992696916564, + "loss": 1.085, + "step": 6273 + }, + { + "epoch": 0.4353618763444591, + "grad_norm": 2.7148706912994385, + "learning_rate": 0.00019675850816717417, + "loss": 1.0498, + "step": 6274 + }, + { + "epoch": 0.4354312677815558, + "grad_norm": 2.748059034347534, + "learning_rate": 0.00019675708905972733, + "loss": 1.0712, + "step": 6275 + }, + { + "epoch": 0.43550065921865244, + "grad_norm": 2.9057881832122803, + "learning_rate": 0.00019675566964682956, + "loss": 0.7178, + "step": 6276 + }, + { + "epoch": 0.4355700506557491, + "grad_norm": 3.7199387550354004, + "learning_rate": 0.0001967542499284854, + "loss": 0.7904, + "step": 6277 + }, + { + "epoch": 0.4356394420928457, + "grad_norm": 3.5022823810577393, + "learning_rate": 0.00019675282990469929, + "loss": 1.0685, + "step": 6278 + }, + { + "epoch": 0.4357088335299424, + "grad_norm": 3.698040008544922, + "learning_rate": 0.00019675140957547572, + "loss": 1.2745, + "step": 6279 + }, + { + "epoch": 0.43577822496703905, + "grad_norm": 3.5051119327545166, + "learning_rate": 0.00019674998894081918, + "loss": 1.1155, + "step": 6280 + }, + { + "epoch": 0.4358476164041357, + "grad_norm": 2.7389156818389893, + "learning_rate": 0.0001967485680007341, + "loss": 0.7228, + "step": 6281 + }, + { + "epoch": 0.4359170078412324, + "grad_norm": 3.2882115840911865, + "learning_rate": 0.0001967471467552251, + "loss": 0.9817, + "step": 6282 + }, + { + "epoch": 0.43598639927832905, + "grad_norm": 3.324869394302368, + "learning_rate": 0.00019674572520429653, + "loss": 0.9959, + "step": 6283 + }, + { + "epoch": 0.4360557907154257, + "grad_norm": 4.232753276824951, + "learning_rate": 0.00019674430334795293, + "loss": 0.8584, + "step": 6284 + }, + { + "epoch": 0.4361251821525224, + "grad_norm": 3.0656208992004395, + "learning_rate": 0.0001967428811861988, + "loss": 1.0462, + "step": 6285 + }, + { + "epoch": 0.43619457358961905, + "grad_norm": 2.6676347255706787, + "learning_rate": 0.00019674145871903864, + "loss": 0.7004, + "step": 6286 + }, + { + "epoch": 0.4362639650267157, + "grad_norm": 2.4072957038879395, + "learning_rate": 0.00019674003594647688, + "loss": 0.6634, + "step": 6287 + }, + { + "epoch": 0.4363333564638124, + "grad_norm": 2.587867021560669, + "learning_rate": 0.00019673861286851807, + "loss": 0.7641, + "step": 6288 + }, + { + "epoch": 0.43640274790090905, + "grad_norm": 2.544104814529419, + "learning_rate": 0.00019673718948516665, + "loss": 1.0034, + "step": 6289 + }, + { + "epoch": 0.4364721393380057, + "grad_norm": 2.898813486099243, + "learning_rate": 0.0001967357657964272, + "loss": 1.0797, + "step": 6290 + }, + { + "epoch": 0.4365415307751023, + "grad_norm": 2.700761079788208, + "learning_rate": 0.00019673434180230416, + "loss": 0.9846, + "step": 6291 + }, + { + "epoch": 0.436610922212199, + "grad_norm": 3.679743528366089, + "learning_rate": 0.00019673291750280198, + "loss": 0.9045, + "step": 6292 + }, + { + "epoch": 0.43668031364929566, + "grad_norm": 2.4663197994232178, + "learning_rate": 0.00019673149289792523, + "loss": 0.8397, + "step": 6293 + }, + { + "epoch": 0.4367497050863923, + "grad_norm": 2.715674638748169, + "learning_rate": 0.00019673006798767836, + "loss": 0.8613, + "step": 6294 + }, + { + "epoch": 0.436819096523489, + "grad_norm": 2.7579708099365234, + "learning_rate": 0.0001967286427720659, + "loss": 0.9308, + "step": 6295 + }, + { + "epoch": 0.43688848796058566, + "grad_norm": 2.265378713607788, + "learning_rate": 0.0001967272172510924, + "loss": 1.0128, + "step": 6296 + }, + { + "epoch": 0.4369578793976823, + "grad_norm": 2.8933658599853516, + "learning_rate": 0.0001967257914247622, + "loss": 1.1314, + "step": 6297 + }, + { + "epoch": 0.437027270834779, + "grad_norm": 2.656048059463501, + "learning_rate": 0.00019672436529307995, + "loss": 0.6588, + "step": 6298 + }, + { + "epoch": 0.43709666227187566, + "grad_norm": 2.252941846847534, + "learning_rate": 0.0001967229388560501, + "loss": 0.8715, + "step": 6299 + }, + { + "epoch": 0.4371660537089723, + "grad_norm": 3.1200602054595947, + "learning_rate": 0.00019672151211367714, + "loss": 0.8675, + "step": 6300 + }, + { + "epoch": 0.437235445146069, + "grad_norm": 2.9074604511260986, + "learning_rate": 0.00019672008506596558, + "loss": 0.8343, + "step": 6301 + }, + { + "epoch": 0.43730483658316566, + "grad_norm": 2.357402801513672, + "learning_rate": 0.00019671865771291993, + "loss": 0.8683, + "step": 6302 + }, + { + "epoch": 0.4373742280202623, + "grad_norm": 3.2006173133850098, + "learning_rate": 0.0001967172300545447, + "loss": 1.3848, + "step": 6303 + }, + { + "epoch": 0.43744361945735893, + "grad_norm": 2.926750659942627, + "learning_rate": 0.00019671580209084443, + "loss": 1.0093, + "step": 6304 + }, + { + "epoch": 0.4375130108944556, + "grad_norm": 3.634268283843994, + "learning_rate": 0.00019671437382182355, + "loss": 1.0329, + "step": 6305 + }, + { + "epoch": 0.43758240233155227, + "grad_norm": 2.8912060260772705, + "learning_rate": 0.00019671294524748666, + "loss": 0.9519, + "step": 6306 + }, + { + "epoch": 0.43765179376864893, + "grad_norm": 2.812741994857788, + "learning_rate": 0.00019671151636783818, + "loss": 0.7515, + "step": 6307 + }, + { + "epoch": 0.4377211852057456, + "grad_norm": 3.033843755722046, + "learning_rate": 0.0001967100871828827, + "loss": 1.0306, + "step": 6308 + }, + { + "epoch": 0.43779057664284227, + "grad_norm": 2.8868870735168457, + "learning_rate": 0.00019670865769262465, + "loss": 1.0395, + "step": 6309 + }, + { + "epoch": 0.43785996807993893, + "grad_norm": 3.328714609146118, + "learning_rate": 0.0001967072278970686, + "loss": 0.9855, + "step": 6310 + }, + { + "epoch": 0.4379293595170356, + "grad_norm": 2.8547637462615967, + "learning_rate": 0.0001967057977962191, + "loss": 1.028, + "step": 6311 + }, + { + "epoch": 0.43799875095413227, + "grad_norm": 3.4350709915161133, + "learning_rate": 0.0001967043673900806, + "loss": 0.9222, + "step": 6312 + }, + { + "epoch": 0.43806814239122893, + "grad_norm": 4.770970344543457, + "learning_rate": 0.0001967029366786576, + "loss": 0.9082, + "step": 6313 + }, + { + "epoch": 0.4381375338283256, + "grad_norm": 3.2764546871185303, + "learning_rate": 0.00019670150566195464, + "loss": 1.0937, + "step": 6314 + }, + { + "epoch": 0.43820692526542226, + "grad_norm": 3.216949224472046, + "learning_rate": 0.00019670007433997627, + "loss": 1.0151, + "step": 6315 + }, + { + "epoch": 0.43827631670251893, + "grad_norm": 2.271873950958252, + "learning_rate": 0.000196698642712727, + "loss": 0.6785, + "step": 6316 + }, + { + "epoch": 0.4383457081396156, + "grad_norm": 2.8924148082733154, + "learning_rate": 0.00019669721078021134, + "loss": 1.074, + "step": 6317 + }, + { + "epoch": 0.4384150995767122, + "grad_norm": 2.951045513153076, + "learning_rate": 0.00019669577854243374, + "loss": 1.2956, + "step": 6318 + }, + { + "epoch": 0.4384844910138089, + "grad_norm": 2.826594829559326, + "learning_rate": 0.00019669434599939883, + "loss": 0.8971, + "step": 6319 + }, + { + "epoch": 0.43855388245090554, + "grad_norm": 2.679032564163208, + "learning_rate": 0.0001966929131511111, + "loss": 0.7788, + "step": 6320 + }, + { + "epoch": 0.4386232738880022, + "grad_norm": 2.8369860649108887, + "learning_rate": 0.00019669147999757506, + "loss": 1.0807, + "step": 6321 + }, + { + "epoch": 0.4386926653250989, + "grad_norm": 3.2412331104278564, + "learning_rate": 0.0001966900465387952, + "loss": 1.1592, + "step": 6322 + }, + { + "epoch": 0.43876205676219554, + "grad_norm": 3.004262685775757, + "learning_rate": 0.0001966886127747761, + "loss": 0.8276, + "step": 6323 + }, + { + "epoch": 0.4388314481992922, + "grad_norm": 3.7899723052978516, + "learning_rate": 0.0001966871787055223, + "loss": 1.2079, + "step": 6324 + }, + { + "epoch": 0.4389008396363889, + "grad_norm": 2.8296992778778076, + "learning_rate": 0.00019668574433103826, + "loss": 0.8611, + "step": 6325 + }, + { + "epoch": 0.43897023107348554, + "grad_norm": 2.8311493396759033, + "learning_rate": 0.00019668430965132853, + "loss": 0.7597, + "step": 6326 + }, + { + "epoch": 0.4390396225105822, + "grad_norm": 2.8525497913360596, + "learning_rate": 0.00019668287466639769, + "loss": 0.874, + "step": 6327 + }, + { + "epoch": 0.4391090139476789, + "grad_norm": 2.8311564922332764, + "learning_rate": 0.0001966814393762502, + "loss": 0.7843, + "step": 6328 + }, + { + "epoch": 0.43917840538477554, + "grad_norm": 2.824716806411743, + "learning_rate": 0.00019668000378089063, + "loss": 1.0053, + "step": 6329 + }, + { + "epoch": 0.4392477968218722, + "grad_norm": 2.3043603897094727, + "learning_rate": 0.00019667856788032354, + "loss": 0.8275, + "step": 6330 + }, + { + "epoch": 0.4393171882589688, + "grad_norm": 4.228503227233887, + "learning_rate": 0.0001966771316745534, + "loss": 0.8101, + "step": 6331 + }, + { + "epoch": 0.4393865796960655, + "grad_norm": 3.44799542427063, + "learning_rate": 0.00019667569516358476, + "loss": 1.0975, + "step": 6332 + }, + { + "epoch": 0.43945597113316215, + "grad_norm": 2.6516497135162354, + "learning_rate": 0.00019667425834742218, + "loss": 0.9053, + "step": 6333 + }, + { + "epoch": 0.4395253625702588, + "grad_norm": 2.912773847579956, + "learning_rate": 0.00019667282122607017, + "loss": 1.0634, + "step": 6334 + }, + { + "epoch": 0.4395947540073555, + "grad_norm": 3.576354503631592, + "learning_rate": 0.00019667138379953332, + "loss": 1.0878, + "step": 6335 + }, + { + "epoch": 0.43966414544445215, + "grad_norm": 3.1363401412963867, + "learning_rate": 0.00019666994606781608, + "loss": 1.0275, + "step": 6336 + }, + { + "epoch": 0.4397335368815488, + "grad_norm": 4.2639946937561035, + "learning_rate": 0.00019666850803092304, + "loss": 0.9102, + "step": 6337 + }, + { + "epoch": 0.4398029283186455, + "grad_norm": 3.5386099815368652, + "learning_rate": 0.00019666706968885875, + "loss": 0.9088, + "step": 6338 + }, + { + "epoch": 0.43987231975574215, + "grad_norm": 3.172551393508911, + "learning_rate": 0.00019666563104162775, + "loss": 1.1988, + "step": 6339 + }, + { + "epoch": 0.4399417111928388, + "grad_norm": 3.2186949253082275, + "learning_rate": 0.00019666419208923456, + "loss": 0.9602, + "step": 6340 + }, + { + "epoch": 0.4400111026299355, + "grad_norm": 2.822301149368286, + "learning_rate": 0.00019666275283168372, + "loss": 1.0048, + "step": 6341 + }, + { + "epoch": 0.44008049406703215, + "grad_norm": 3.3266310691833496, + "learning_rate": 0.0001966613132689798, + "loss": 1.2794, + "step": 6342 + }, + { + "epoch": 0.4401498855041288, + "grad_norm": 2.895460605621338, + "learning_rate": 0.00019665987340112732, + "loss": 0.932, + "step": 6343 + }, + { + "epoch": 0.4402192769412255, + "grad_norm": 3.117680311203003, + "learning_rate": 0.00019665843322813083, + "loss": 0.9865, + "step": 6344 + }, + { + "epoch": 0.4402886683783221, + "grad_norm": 4.280239582061768, + "learning_rate": 0.00019665699274999486, + "loss": 1.1547, + "step": 6345 + }, + { + "epoch": 0.44035805981541876, + "grad_norm": 3.5718939304351807, + "learning_rate": 0.00019665555196672402, + "loss": 0.7461, + "step": 6346 + }, + { + "epoch": 0.4404274512525154, + "grad_norm": 2.978001117706299, + "learning_rate": 0.0001966541108783228, + "loss": 0.9313, + "step": 6347 + }, + { + "epoch": 0.4404968426896121, + "grad_norm": 3.32479190826416, + "learning_rate": 0.00019665266948479574, + "loss": 0.8714, + "step": 6348 + }, + { + "epoch": 0.44056623412670876, + "grad_norm": 2.859910488128662, + "learning_rate": 0.00019665122778614747, + "loss": 0.9502, + "step": 6349 + }, + { + "epoch": 0.4406356255638054, + "grad_norm": 3.8431482315063477, + "learning_rate": 0.00019664978578238243, + "loss": 0.7442, + "step": 6350 + }, + { + "epoch": 0.4407050170009021, + "grad_norm": 2.7129595279693604, + "learning_rate": 0.00019664834347350526, + "loss": 0.7357, + "step": 6351 + }, + { + "epoch": 0.44077440843799875, + "grad_norm": 3.559237480163574, + "learning_rate": 0.0001966469008595205, + "loss": 1.0818, + "step": 6352 + }, + { + "epoch": 0.4408437998750954, + "grad_norm": 3.500828266143799, + "learning_rate": 0.00019664545794043266, + "loss": 0.8981, + "step": 6353 + }, + { + "epoch": 0.4409131913121921, + "grad_norm": 3.2654836177825928, + "learning_rate": 0.0001966440147162463, + "loss": 1.0056, + "step": 6354 + }, + { + "epoch": 0.44098258274928875, + "grad_norm": 5.927623271942139, + "learning_rate": 0.00019664257118696604, + "loss": 1.2295, + "step": 6355 + }, + { + "epoch": 0.4410519741863854, + "grad_norm": 3.4610390663146973, + "learning_rate": 0.0001966411273525964, + "loss": 1.0274, + "step": 6356 + }, + { + "epoch": 0.4411213656234821, + "grad_norm": 3.5232608318328857, + "learning_rate": 0.0001966396832131419, + "loss": 1.0656, + "step": 6357 + }, + { + "epoch": 0.4411907570605787, + "grad_norm": 3.470651626586914, + "learning_rate": 0.00019663823876860716, + "loss": 0.9888, + "step": 6358 + }, + { + "epoch": 0.44126014849767536, + "grad_norm": 3.2403388023376465, + "learning_rate": 0.00019663679401899672, + "loss": 1.0104, + "step": 6359 + }, + { + "epoch": 0.44132953993477203, + "grad_norm": 2.5166494846343994, + "learning_rate": 0.0001966353489643151, + "loss": 0.7597, + "step": 6360 + }, + { + "epoch": 0.4413989313718687, + "grad_norm": 4.400936603546143, + "learning_rate": 0.00019663390360456692, + "loss": 1.2792, + "step": 6361 + }, + { + "epoch": 0.44146832280896536, + "grad_norm": 3.0235841274261475, + "learning_rate": 0.0001966324579397567, + "loss": 0.9673, + "step": 6362 + }, + { + "epoch": 0.44153771424606203, + "grad_norm": 3.1483089923858643, + "learning_rate": 0.000196631011969889, + "loss": 0.9526, + "step": 6363 + }, + { + "epoch": 0.4416071056831587, + "grad_norm": 3.542062282562256, + "learning_rate": 0.00019662956569496847, + "loss": 0.9969, + "step": 6364 + }, + { + "epoch": 0.44167649712025536, + "grad_norm": 3.1813652515411377, + "learning_rate": 0.00019662811911499956, + "loss": 1.1179, + "step": 6365 + }, + { + "epoch": 0.44174588855735203, + "grad_norm": 3.147944688796997, + "learning_rate": 0.0001966266722299869, + "loss": 1.0915, + "step": 6366 + }, + { + "epoch": 0.4418152799944487, + "grad_norm": 3.9200141429901123, + "learning_rate": 0.00019662522503993506, + "loss": 0.9797, + "step": 6367 + }, + { + "epoch": 0.44188467143154536, + "grad_norm": 3.1039044857025146, + "learning_rate": 0.0001966237775448486, + "loss": 1.0215, + "step": 6368 + }, + { + "epoch": 0.44195406286864203, + "grad_norm": 3.107693672180176, + "learning_rate": 0.00019662232974473206, + "loss": 1.0644, + "step": 6369 + }, + { + "epoch": 0.4420234543057387, + "grad_norm": 4.045411109924316, + "learning_rate": 0.00019662088163959008, + "loss": 1.1475, + "step": 6370 + }, + { + "epoch": 0.44209284574283536, + "grad_norm": 3.1856777667999268, + "learning_rate": 0.00019661943322942714, + "loss": 0.9011, + "step": 6371 + }, + { + "epoch": 0.44216223717993197, + "grad_norm": 3.405167579650879, + "learning_rate": 0.00019661798451424785, + "loss": 1.0586, + "step": 6372 + }, + { + "epoch": 0.44223162861702864, + "grad_norm": 2.586017608642578, + "learning_rate": 0.00019661653549405683, + "loss": 0.9159, + "step": 6373 + }, + { + "epoch": 0.4423010200541253, + "grad_norm": 3.9791994094848633, + "learning_rate": 0.00019661508616885862, + "loss": 0.8984, + "step": 6374 + }, + { + "epoch": 0.44237041149122197, + "grad_norm": 3.1435775756835938, + "learning_rate": 0.00019661363653865777, + "loss": 0.8728, + "step": 6375 + }, + { + "epoch": 0.44243980292831864, + "grad_norm": 3.2995972633361816, + "learning_rate": 0.00019661218660345888, + "loss": 0.8673, + "step": 6376 + }, + { + "epoch": 0.4425091943654153, + "grad_norm": 2.9704537391662598, + "learning_rate": 0.00019661073636326654, + "loss": 0.8735, + "step": 6377 + }, + { + "epoch": 0.44257858580251197, + "grad_norm": 2.576747179031372, + "learning_rate": 0.00019660928581808528, + "loss": 0.7822, + "step": 6378 + }, + { + "epoch": 0.44264797723960864, + "grad_norm": 4.3720011711120605, + "learning_rate": 0.00019660783496791977, + "loss": 0.9739, + "step": 6379 + }, + { + "epoch": 0.4427173686767053, + "grad_norm": 3.3074893951416016, + "learning_rate": 0.00019660638381277447, + "loss": 0.9581, + "step": 6380 + }, + { + "epoch": 0.44278676011380197, + "grad_norm": 3.3383734226226807, + "learning_rate": 0.00019660493235265408, + "loss": 0.8404, + "step": 6381 + }, + { + "epoch": 0.44285615155089864, + "grad_norm": 3.1090176105499268, + "learning_rate": 0.0001966034805875631, + "loss": 0.9142, + "step": 6382 + }, + { + "epoch": 0.4429255429879953, + "grad_norm": 3.624824047088623, + "learning_rate": 0.00019660202851750613, + "loss": 1.0207, + "step": 6383 + }, + { + "epoch": 0.44299493442509197, + "grad_norm": 3.0270960330963135, + "learning_rate": 0.00019660057614248778, + "loss": 0.9422, + "step": 6384 + }, + { + "epoch": 0.4430643258621886, + "grad_norm": 3.199841022491455, + "learning_rate": 0.00019659912346251264, + "loss": 1.293, + "step": 6385 + }, + { + "epoch": 0.44313371729928525, + "grad_norm": 3.3811917304992676, + "learning_rate": 0.00019659767047758525, + "loss": 0.8462, + "step": 6386 + }, + { + "epoch": 0.4432031087363819, + "grad_norm": 2.4019055366516113, + "learning_rate": 0.00019659621718771023, + "loss": 0.8889, + "step": 6387 + }, + { + "epoch": 0.4432725001734786, + "grad_norm": 4.279316425323486, + "learning_rate": 0.00019659476359289215, + "loss": 1.0116, + "step": 6388 + }, + { + "epoch": 0.44334189161057524, + "grad_norm": 3.3755812644958496, + "learning_rate": 0.00019659330969313562, + "loss": 0.9429, + "step": 6389 + }, + { + "epoch": 0.4434112830476719, + "grad_norm": 2.878573417663574, + "learning_rate": 0.0001965918554884452, + "loss": 1.0233, + "step": 6390 + }, + { + "epoch": 0.4434806744847686, + "grad_norm": 2.2736310958862305, + "learning_rate": 0.0001965904009788255, + "loss": 0.8619, + "step": 6391 + }, + { + "epoch": 0.44355006592186524, + "grad_norm": 2.531388521194458, + "learning_rate": 0.00019658894616428113, + "loss": 0.8042, + "step": 6392 + }, + { + "epoch": 0.4436194573589619, + "grad_norm": 3.2537591457366943, + "learning_rate": 0.00019658749104481665, + "loss": 0.7307, + "step": 6393 + }, + { + "epoch": 0.4436888487960586, + "grad_norm": 5.3523149490356445, + "learning_rate": 0.00019658603562043667, + "loss": 1.1405, + "step": 6394 + }, + { + "epoch": 0.44375824023315524, + "grad_norm": 3.4659759998321533, + "learning_rate": 0.0001965845798911458, + "loss": 0.9099, + "step": 6395 + }, + { + "epoch": 0.4438276316702519, + "grad_norm": 3.0697245597839355, + "learning_rate": 0.00019658312385694858, + "loss": 0.9009, + "step": 6396 + }, + { + "epoch": 0.4438970231073486, + "grad_norm": 2.7874755859375, + "learning_rate": 0.00019658166751784965, + "loss": 0.8466, + "step": 6397 + }, + { + "epoch": 0.44396641454444524, + "grad_norm": 2.37502121925354, + "learning_rate": 0.0001965802108738536, + "loss": 0.7465, + "step": 6398 + }, + { + "epoch": 0.44403580598154185, + "grad_norm": 3.9257988929748535, + "learning_rate": 0.00019657875392496506, + "loss": 0.979, + "step": 6399 + }, + { + "epoch": 0.4441051974186385, + "grad_norm": 3.1245853900909424, + "learning_rate": 0.00019657729667118856, + "loss": 0.8858, + "step": 6400 + }, + { + "epoch": 0.4441745888557352, + "grad_norm": 3.8410348892211914, + "learning_rate": 0.00019657583911252878, + "loss": 0.9401, + "step": 6401 + }, + { + "epoch": 0.44424398029283185, + "grad_norm": 2.7345528602600098, + "learning_rate": 0.00019657438124899026, + "loss": 0.8513, + "step": 6402 + }, + { + "epoch": 0.4443133717299285, + "grad_norm": 2.8082566261291504, + "learning_rate": 0.0001965729230805776, + "loss": 0.6066, + "step": 6403 + }, + { + "epoch": 0.4443827631670252, + "grad_norm": 2.862445116043091, + "learning_rate": 0.00019657146460729547, + "loss": 0.9163, + "step": 6404 + }, + { + "epoch": 0.44445215460412185, + "grad_norm": 3.479689359664917, + "learning_rate": 0.0001965700058291484, + "loss": 1.024, + "step": 6405 + }, + { + "epoch": 0.4445215460412185, + "grad_norm": 2.874321937561035, + "learning_rate": 0.00019656854674614106, + "loss": 0.9438, + "step": 6406 + }, + { + "epoch": 0.4445909374783152, + "grad_norm": 2.452221155166626, + "learning_rate": 0.00019656708735827798, + "loss": 0.9004, + "step": 6407 + }, + { + "epoch": 0.44466032891541185, + "grad_norm": 3.063018560409546, + "learning_rate": 0.00019656562766556384, + "loss": 0.7717, + "step": 6408 + }, + { + "epoch": 0.4447297203525085, + "grad_norm": 4.5110249519348145, + "learning_rate": 0.00019656416766800322, + "loss": 1.0939, + "step": 6409 + }, + { + "epoch": 0.4447991117896052, + "grad_norm": 2.7713074684143066, + "learning_rate": 0.0001965627073656007, + "loss": 0.8499, + "step": 6410 + }, + { + "epoch": 0.44486850322670185, + "grad_norm": 2.8839309215545654, + "learning_rate": 0.00019656124675836094, + "loss": 1.1899, + "step": 6411 + }, + { + "epoch": 0.44493789466379846, + "grad_norm": 3.330744981765747, + "learning_rate": 0.0001965597858462885, + "loss": 0.9359, + "step": 6412 + }, + { + "epoch": 0.4450072861008951, + "grad_norm": 2.966071844100952, + "learning_rate": 0.00019655832462938804, + "loss": 1.1342, + "step": 6413 + }, + { + "epoch": 0.4450766775379918, + "grad_norm": 3.239037036895752, + "learning_rate": 0.00019655686310766414, + "loss": 0.9586, + "step": 6414 + }, + { + "epoch": 0.44514606897508846, + "grad_norm": 3.2455499172210693, + "learning_rate": 0.00019655540128112145, + "loss": 1.0594, + "step": 6415 + }, + { + "epoch": 0.4452154604121851, + "grad_norm": 3.4658236503601074, + "learning_rate": 0.00019655393914976452, + "loss": 1.0748, + "step": 6416 + }, + { + "epoch": 0.4452848518492818, + "grad_norm": 3.0122478008270264, + "learning_rate": 0.00019655247671359805, + "loss": 0.9279, + "step": 6417 + }, + { + "epoch": 0.44535424328637846, + "grad_norm": 2.84993839263916, + "learning_rate": 0.00019655101397262657, + "loss": 1.1343, + "step": 6418 + }, + { + "epoch": 0.4454236347234751, + "grad_norm": 2.867680549621582, + "learning_rate": 0.00019654955092685477, + "loss": 1.0133, + "step": 6419 + }, + { + "epoch": 0.4454930261605718, + "grad_norm": 3.163024663925171, + "learning_rate": 0.00019654808757628723, + "loss": 0.8783, + "step": 6420 + }, + { + "epoch": 0.44556241759766846, + "grad_norm": 3.547650098800659, + "learning_rate": 0.0001965466239209286, + "loss": 1.2215, + "step": 6421 + }, + { + "epoch": 0.4456318090347651, + "grad_norm": 2.581817150115967, + "learning_rate": 0.00019654515996078345, + "loss": 0.6968, + "step": 6422 + }, + { + "epoch": 0.4457012004718618, + "grad_norm": 2.552859306335449, + "learning_rate": 0.00019654369569585645, + "loss": 0.8412, + "step": 6423 + }, + { + "epoch": 0.44577059190895846, + "grad_norm": 3.1641528606414795, + "learning_rate": 0.00019654223112615217, + "loss": 1.2326, + "step": 6424 + }, + { + "epoch": 0.4458399833460551, + "grad_norm": 3.9216794967651367, + "learning_rate": 0.0001965407662516753, + "loss": 1.0176, + "step": 6425 + }, + { + "epoch": 0.44590937478315174, + "grad_norm": 3.0747594833374023, + "learning_rate": 0.00019653930107243044, + "loss": 0.6734, + "step": 6426 + }, + { + "epoch": 0.4459787662202484, + "grad_norm": 3.2272634506225586, + "learning_rate": 0.0001965378355884222, + "loss": 1.1281, + "step": 6427 + }, + { + "epoch": 0.44604815765734507, + "grad_norm": 2.414320945739746, + "learning_rate": 0.00019653636979965523, + "loss": 0.7759, + "step": 6428 + }, + { + "epoch": 0.44611754909444173, + "grad_norm": 2.861849069595337, + "learning_rate": 0.0001965349037061341, + "loss": 1.1952, + "step": 6429 + }, + { + "epoch": 0.4461869405315384, + "grad_norm": 3.3168222904205322, + "learning_rate": 0.0001965334373078635, + "loss": 1.1067, + "step": 6430 + }, + { + "epoch": 0.44625633196863507, + "grad_norm": 4.744684219360352, + "learning_rate": 0.00019653197060484802, + "loss": 1.1365, + "step": 6431 + }, + { + "epoch": 0.44632572340573173, + "grad_norm": 3.597968101501465, + "learning_rate": 0.00019653050359709232, + "loss": 1.0716, + "step": 6432 + }, + { + "epoch": 0.4463951148428284, + "grad_norm": 3.4074463844299316, + "learning_rate": 0.00019652903628460103, + "loss": 0.9749, + "step": 6433 + }, + { + "epoch": 0.44646450627992507, + "grad_norm": 7.574217796325684, + "learning_rate": 0.00019652756866737875, + "loss": 0.8519, + "step": 6434 + }, + { + "epoch": 0.44653389771702173, + "grad_norm": 3.429797410964966, + "learning_rate": 0.00019652610074543015, + "loss": 1.1404, + "step": 6435 + }, + { + "epoch": 0.4466032891541184, + "grad_norm": 2.7513983249664307, + "learning_rate": 0.00019652463251875987, + "loss": 0.7952, + "step": 6436 + }, + { + "epoch": 0.44667268059121507, + "grad_norm": 3.4165244102478027, + "learning_rate": 0.0001965231639873725, + "loss": 0.9706, + "step": 6437 + }, + { + "epoch": 0.44674207202831173, + "grad_norm": 3.544552803039551, + "learning_rate": 0.0001965216951512727, + "loss": 0.9527, + "step": 6438 + }, + { + "epoch": 0.44681146346540834, + "grad_norm": 2.615483522415161, + "learning_rate": 0.0001965202260104651, + "loss": 0.9778, + "step": 6439 + }, + { + "epoch": 0.446880854902505, + "grad_norm": 2.8845362663269043, + "learning_rate": 0.00019651875656495437, + "loss": 1.0169, + "step": 6440 + }, + { + "epoch": 0.4469502463396017, + "grad_norm": 2.6480488777160645, + "learning_rate": 0.0001965172868147451, + "loss": 0.977, + "step": 6441 + }, + { + "epoch": 0.44701963777669834, + "grad_norm": 2.803410291671753, + "learning_rate": 0.00019651581675984197, + "loss": 1.0748, + "step": 6442 + }, + { + "epoch": 0.447089029213795, + "grad_norm": 2.582928419113159, + "learning_rate": 0.00019651434640024959, + "loss": 0.7597, + "step": 6443 + }, + { + "epoch": 0.4471584206508917, + "grad_norm": 3.145156145095825, + "learning_rate": 0.00019651287573597262, + "loss": 0.9558, + "step": 6444 + }, + { + "epoch": 0.44722781208798834, + "grad_norm": 3.3816232681274414, + "learning_rate": 0.0001965114047670157, + "loss": 1.0955, + "step": 6445 + }, + { + "epoch": 0.447297203525085, + "grad_norm": 3.356610059738159, + "learning_rate": 0.0001965099334933835, + "loss": 0.8315, + "step": 6446 + }, + { + "epoch": 0.4473665949621817, + "grad_norm": 3.2029025554656982, + "learning_rate": 0.0001965084619150806, + "loss": 0.7833, + "step": 6447 + }, + { + "epoch": 0.44743598639927834, + "grad_norm": 3.6411054134368896, + "learning_rate": 0.0001965069900321117, + "loss": 0.6153, + "step": 6448 + }, + { + "epoch": 0.447505377836375, + "grad_norm": 3.270850896835327, + "learning_rate": 0.0001965055178444814, + "loss": 0.9597, + "step": 6449 + }, + { + "epoch": 0.4475747692734717, + "grad_norm": 2.931417226791382, + "learning_rate": 0.00019650404535219443, + "loss": 1.3306, + "step": 6450 + }, + { + "epoch": 0.44764416071056834, + "grad_norm": 2.381889581680298, + "learning_rate": 0.00019650257255525536, + "loss": 0.84, + "step": 6451 + }, + { + "epoch": 0.44771355214766495, + "grad_norm": 4.749880790710449, + "learning_rate": 0.00019650109945366885, + "loss": 0.8218, + "step": 6452 + }, + { + "epoch": 0.4477829435847616, + "grad_norm": 4.090230464935303, + "learning_rate": 0.00019649962604743957, + "loss": 1.1279, + "step": 6453 + }, + { + "epoch": 0.4478523350218583, + "grad_norm": 2.812803268432617, + "learning_rate": 0.00019649815233657218, + "loss": 1.0164, + "step": 6454 + }, + { + "epoch": 0.44792172645895495, + "grad_norm": 3.0954363346099854, + "learning_rate": 0.0001964966783210713, + "loss": 0.9397, + "step": 6455 + }, + { + "epoch": 0.4479911178960516, + "grad_norm": 2.567108154296875, + "learning_rate": 0.0001964952040009416, + "loss": 0.9182, + "step": 6456 + }, + { + "epoch": 0.4480605093331483, + "grad_norm": 3.021280527114868, + "learning_rate": 0.0001964937293761878, + "loss": 1.2038, + "step": 6457 + }, + { + "epoch": 0.44812990077024495, + "grad_norm": 3.0196094512939453, + "learning_rate": 0.0001964922544468144, + "loss": 1.0707, + "step": 6458 + }, + { + "epoch": 0.4481992922073416, + "grad_norm": 2.595238208770752, + "learning_rate": 0.0001964907792128262, + "loss": 0.787, + "step": 6459 + }, + { + "epoch": 0.4482686836444383, + "grad_norm": 4.182779312133789, + "learning_rate": 0.0001964893036742278, + "loss": 1.1221, + "step": 6460 + }, + { + "epoch": 0.44833807508153495, + "grad_norm": 2.7546513080596924, + "learning_rate": 0.00019648782783102386, + "loss": 1.2138, + "step": 6461 + }, + { + "epoch": 0.4484074665186316, + "grad_norm": 2.255934238433838, + "learning_rate": 0.00019648635168321903, + "loss": 0.8789, + "step": 6462 + }, + { + "epoch": 0.4484768579557283, + "grad_norm": 3.7897400856018066, + "learning_rate": 0.000196484875230818, + "loss": 1.1823, + "step": 6463 + }, + { + "epoch": 0.44854624939282495, + "grad_norm": 2.9375851154327393, + "learning_rate": 0.00019648339847382542, + "loss": 0.8606, + "step": 6464 + }, + { + "epoch": 0.4486156408299216, + "grad_norm": 2.523829460144043, + "learning_rate": 0.00019648192141224593, + "loss": 0.9606, + "step": 6465 + }, + { + "epoch": 0.4486850322670182, + "grad_norm": 3.1992390155792236, + "learning_rate": 0.0001964804440460842, + "loss": 1.3076, + "step": 6466 + }, + { + "epoch": 0.4487544237041149, + "grad_norm": 3.2373507022857666, + "learning_rate": 0.0001964789663753449, + "loss": 1.2132, + "step": 6467 + }, + { + "epoch": 0.44882381514121156, + "grad_norm": 4.00462007522583, + "learning_rate": 0.0001964774884000327, + "loss": 0.9157, + "step": 6468 + }, + { + "epoch": 0.4488932065783082, + "grad_norm": 4.122476577758789, + "learning_rate": 0.00019647601012015228, + "loss": 1.2496, + "step": 6469 + }, + { + "epoch": 0.4489625980154049, + "grad_norm": 2.8863518238067627, + "learning_rate": 0.0001964745315357083, + "loss": 1.1016, + "step": 6470 + }, + { + "epoch": 0.44903198945250156, + "grad_norm": 2.3744215965270996, + "learning_rate": 0.0001964730526467054, + "loss": 1.0672, + "step": 6471 + }, + { + "epoch": 0.4491013808895982, + "grad_norm": 2.398233413696289, + "learning_rate": 0.00019647157345314823, + "loss": 0.8592, + "step": 6472 + }, + { + "epoch": 0.4491707723266949, + "grad_norm": 2.979234457015991, + "learning_rate": 0.00019647009395504155, + "loss": 1.1138, + "step": 6473 + }, + { + "epoch": 0.44924016376379156, + "grad_norm": 3.4491775035858154, + "learning_rate": 0.00019646861415238996, + "loss": 1.3878, + "step": 6474 + }, + { + "epoch": 0.4493095552008882, + "grad_norm": 2.3502402305603027, + "learning_rate": 0.00019646713404519815, + "loss": 0.8048, + "step": 6475 + }, + { + "epoch": 0.4493789466379849, + "grad_norm": 2.637204170227051, + "learning_rate": 0.00019646565363347075, + "loss": 0.8046, + "step": 6476 + }, + { + "epoch": 0.44944833807508155, + "grad_norm": 2.8014729022979736, + "learning_rate": 0.00019646417291721254, + "loss": 1.1295, + "step": 6477 + }, + { + "epoch": 0.4495177295121782, + "grad_norm": 2.6648385524749756, + "learning_rate": 0.00019646269189642807, + "loss": 1.0083, + "step": 6478 + }, + { + "epoch": 0.44958712094927483, + "grad_norm": 3.0523979663848877, + "learning_rate": 0.0001964612105711221, + "loss": 1.0119, + "step": 6479 + }, + { + "epoch": 0.4496565123863715, + "grad_norm": 2.701873779296875, + "learning_rate": 0.00019645972894129926, + "loss": 1.1405, + "step": 6480 + }, + { + "epoch": 0.44972590382346816, + "grad_norm": 2.9991538524627686, + "learning_rate": 0.0001964582470069643, + "loss": 0.9136, + "step": 6481 + }, + { + "epoch": 0.44979529526056483, + "grad_norm": 3.3130831718444824, + "learning_rate": 0.0001964567647681218, + "loss": 1.2593, + "step": 6482 + }, + { + "epoch": 0.4498646866976615, + "grad_norm": 2.7465922832489014, + "learning_rate": 0.00019645528222477648, + "loss": 1.2041, + "step": 6483 + }, + { + "epoch": 0.44993407813475816, + "grad_norm": 4.405238628387451, + "learning_rate": 0.00019645379937693305, + "loss": 1.2908, + "step": 6484 + }, + { + "epoch": 0.45000346957185483, + "grad_norm": 2.2848687171936035, + "learning_rate": 0.00019645231622459613, + "loss": 0.7558, + "step": 6485 + }, + { + "epoch": 0.4500728610089515, + "grad_norm": 2.299013614654541, + "learning_rate": 0.00019645083276777046, + "loss": 0.8947, + "step": 6486 + }, + { + "epoch": 0.45014225244604816, + "grad_norm": 3.399193525314331, + "learning_rate": 0.00019644934900646074, + "loss": 0.9989, + "step": 6487 + }, + { + "epoch": 0.45021164388314483, + "grad_norm": 2.4706764221191406, + "learning_rate": 0.00019644786494067156, + "loss": 0.8551, + "step": 6488 + }, + { + "epoch": 0.4502810353202415, + "grad_norm": 3.557370185852051, + "learning_rate": 0.00019644638057040768, + "loss": 1.2474, + "step": 6489 + }, + { + "epoch": 0.45035042675733816, + "grad_norm": 3.0286519527435303, + "learning_rate": 0.00019644489589567377, + "loss": 0.8128, + "step": 6490 + }, + { + "epoch": 0.45041981819443483, + "grad_norm": 2.8921942710876465, + "learning_rate": 0.00019644341091647452, + "loss": 0.9031, + "step": 6491 + }, + { + "epoch": 0.4504892096315315, + "grad_norm": 2.1854846477508545, + "learning_rate": 0.0001964419256328146, + "loss": 0.8323, + "step": 6492 + }, + { + "epoch": 0.4505586010686281, + "grad_norm": 2.563086748123169, + "learning_rate": 0.00019644044004469871, + "loss": 0.7295, + "step": 6493 + }, + { + "epoch": 0.4506279925057248, + "grad_norm": 3.2563984394073486, + "learning_rate": 0.00019643895415213152, + "loss": 0.9593, + "step": 6494 + }, + { + "epoch": 0.45069738394282144, + "grad_norm": 3.9080650806427, + "learning_rate": 0.00019643746795511776, + "loss": 0.724, + "step": 6495 + }, + { + "epoch": 0.4507667753799181, + "grad_norm": 3.0193705558776855, + "learning_rate": 0.00019643598145366211, + "loss": 0.9392, + "step": 6496 + }, + { + "epoch": 0.45083616681701477, + "grad_norm": 3.15173077583313, + "learning_rate": 0.00019643449464776925, + "loss": 0.799, + "step": 6497 + }, + { + "epoch": 0.45090555825411144, + "grad_norm": 3.450984239578247, + "learning_rate": 0.00019643300753744388, + "loss": 0.8772, + "step": 6498 + }, + { + "epoch": 0.4509749496912081, + "grad_norm": 2.4530909061431885, + "learning_rate": 0.0001964315201226907, + "loss": 1.0776, + "step": 6499 + }, + { + "epoch": 0.45104434112830477, + "grad_norm": 3.519817590713501, + "learning_rate": 0.00019643003240351438, + "loss": 0.9804, + "step": 6500 + }, + { + "epoch": 0.45111373256540144, + "grad_norm": 2.527507781982422, + "learning_rate": 0.00019642854437991965, + "loss": 0.9436, + "step": 6501 + }, + { + "epoch": 0.4511831240024981, + "grad_norm": 2.172227382659912, + "learning_rate": 0.0001964270560519112, + "loss": 0.8646, + "step": 6502 + }, + { + "epoch": 0.45125251543959477, + "grad_norm": 2.7451932430267334, + "learning_rate": 0.0001964255674194937, + "loss": 0.8165, + "step": 6503 + }, + { + "epoch": 0.45132190687669144, + "grad_norm": 2.675924777984619, + "learning_rate": 0.00019642407848267191, + "loss": 0.8063, + "step": 6504 + }, + { + "epoch": 0.4513912983137881, + "grad_norm": 3.5641517639160156, + "learning_rate": 0.00019642258924145044, + "loss": 1.2155, + "step": 6505 + }, + { + "epoch": 0.4514606897508847, + "grad_norm": 2.4928743839263916, + "learning_rate": 0.00019642109969583406, + "loss": 0.8949, + "step": 6506 + }, + { + "epoch": 0.4515300811879814, + "grad_norm": 4.174919128417969, + "learning_rate": 0.0001964196098458275, + "loss": 1.116, + "step": 6507 + }, + { + "epoch": 0.45159947262507805, + "grad_norm": 4.028798580169678, + "learning_rate": 0.00019641811969143535, + "loss": 1.1119, + "step": 6508 + }, + { + "epoch": 0.4516688640621747, + "grad_norm": 2.553462028503418, + "learning_rate": 0.0001964166292326624, + "loss": 1.0251, + "step": 6509 + }, + { + "epoch": 0.4517382554992714, + "grad_norm": 2.4914662837982178, + "learning_rate": 0.00019641513846951336, + "loss": 0.9271, + "step": 6510 + }, + { + "epoch": 0.45180764693636805, + "grad_norm": 2.4600887298583984, + "learning_rate": 0.0001964136474019929, + "loss": 0.8657, + "step": 6511 + }, + { + "epoch": 0.4518770383734647, + "grad_norm": 2.2105906009674072, + "learning_rate": 0.00019641215603010578, + "loss": 0.7462, + "step": 6512 + }, + { + "epoch": 0.4519464298105614, + "grad_norm": 3.0849483013153076, + "learning_rate": 0.00019641066435385662, + "loss": 0.9644, + "step": 6513 + }, + { + "epoch": 0.45201582124765805, + "grad_norm": 2.715071439743042, + "learning_rate": 0.00019640917237325018, + "loss": 0.9438, + "step": 6514 + }, + { + "epoch": 0.4520852126847547, + "grad_norm": 3.0020225048065186, + "learning_rate": 0.0001964076800882912, + "loss": 0.867, + "step": 6515 + }, + { + "epoch": 0.4521546041218514, + "grad_norm": 2.9757659435272217, + "learning_rate": 0.00019640618749898435, + "loss": 0.9943, + "step": 6516 + }, + { + "epoch": 0.45222399555894804, + "grad_norm": 3.600292921066284, + "learning_rate": 0.00019640469460533434, + "loss": 1.1913, + "step": 6517 + }, + { + "epoch": 0.4522933869960447, + "grad_norm": 3.4403371810913086, + "learning_rate": 0.0001964032014073459, + "loss": 1.0428, + "step": 6518 + }, + { + "epoch": 0.4523627784331414, + "grad_norm": 3.5980803966522217, + "learning_rate": 0.00019640170790502374, + "loss": 1.2129, + "step": 6519 + }, + { + "epoch": 0.452432169870238, + "grad_norm": 2.7566354274749756, + "learning_rate": 0.00019640021409837255, + "loss": 0.9162, + "step": 6520 + }, + { + "epoch": 0.45250156130733465, + "grad_norm": 2.642080783843994, + "learning_rate": 0.00019639871998739707, + "loss": 0.7616, + "step": 6521 + }, + { + "epoch": 0.4525709527444313, + "grad_norm": 2.417609214782715, + "learning_rate": 0.00019639722557210205, + "loss": 0.8695, + "step": 6522 + }, + { + "epoch": 0.452640344181528, + "grad_norm": 3.684617280960083, + "learning_rate": 0.00019639573085249214, + "loss": 1.1616, + "step": 6523 + }, + { + "epoch": 0.45270973561862465, + "grad_norm": 2.7583115100860596, + "learning_rate": 0.0001963942358285721, + "loss": 0.8398, + "step": 6524 + }, + { + "epoch": 0.4527791270557213, + "grad_norm": 1.9199895858764648, + "learning_rate": 0.00019639274050034666, + "loss": 0.8057, + "step": 6525 + }, + { + "epoch": 0.452848518492818, + "grad_norm": 3.4536662101745605, + "learning_rate": 0.0001963912448678205, + "loss": 1.1754, + "step": 6526 + }, + { + "epoch": 0.45291790992991465, + "grad_norm": 2.8627240657806396, + "learning_rate": 0.0001963897489309984, + "loss": 1.0275, + "step": 6527 + }, + { + "epoch": 0.4529873013670113, + "grad_norm": 2.6801695823669434, + "learning_rate": 0.000196388252689885, + "loss": 0.9332, + "step": 6528 + }, + { + "epoch": 0.453056692804108, + "grad_norm": 4.077589511871338, + "learning_rate": 0.0001963867561444851, + "loss": 1.3593, + "step": 6529 + }, + { + "epoch": 0.45312608424120465, + "grad_norm": 3.3329293727874756, + "learning_rate": 0.00019638525929480336, + "loss": 1.1131, + "step": 6530 + }, + { + "epoch": 0.4531954756783013, + "grad_norm": 2.501195192337036, + "learning_rate": 0.00019638376214084458, + "loss": 0.6323, + "step": 6531 + }, + { + "epoch": 0.453264867115398, + "grad_norm": 3.2498133182525635, + "learning_rate": 0.00019638226468261344, + "loss": 0.7958, + "step": 6532 + }, + { + "epoch": 0.4533342585524946, + "grad_norm": 2.4448866844177246, + "learning_rate": 0.00019638076692011464, + "loss": 0.9322, + "step": 6533 + }, + { + "epoch": 0.45340364998959126, + "grad_norm": 3.326361894607544, + "learning_rate": 0.00019637926885335296, + "loss": 0.7851, + "step": 6534 + }, + { + "epoch": 0.45347304142668793, + "grad_norm": 2.9900588989257812, + "learning_rate": 0.00019637777048233312, + "loss": 0.9703, + "step": 6535 + }, + { + "epoch": 0.4535424328637846, + "grad_norm": 2.5383310317993164, + "learning_rate": 0.00019637627180705983, + "loss": 1.0573, + "step": 6536 + }, + { + "epoch": 0.45361182430088126, + "grad_norm": 2.8224008083343506, + "learning_rate": 0.00019637477282753783, + "loss": 0.8629, + "step": 6537 + }, + { + "epoch": 0.4536812157379779, + "grad_norm": 2.573606491088867, + "learning_rate": 0.00019637327354377187, + "loss": 0.9739, + "step": 6538 + }, + { + "epoch": 0.4537506071750746, + "grad_norm": 2.7503440380096436, + "learning_rate": 0.00019637177395576662, + "loss": 0.8163, + "step": 6539 + }, + { + "epoch": 0.45381999861217126, + "grad_norm": 2.2338509559631348, + "learning_rate": 0.00019637027406352694, + "loss": 0.5653, + "step": 6540 + }, + { + "epoch": 0.4538893900492679, + "grad_norm": 2.6534318923950195, + "learning_rate": 0.00019636877386705742, + "loss": 1.1574, + "step": 6541 + }, + { + "epoch": 0.4539587814863646, + "grad_norm": 2.4119300842285156, + "learning_rate": 0.00019636727336636286, + "loss": 0.9549, + "step": 6542 + }, + { + "epoch": 0.45402817292346126, + "grad_norm": 2.799865484237671, + "learning_rate": 0.00019636577256144802, + "loss": 0.9084, + "step": 6543 + }, + { + "epoch": 0.4540975643605579, + "grad_norm": 2.597205400466919, + "learning_rate": 0.00019636427145231763, + "loss": 1.0312, + "step": 6544 + }, + { + "epoch": 0.4541669557976546, + "grad_norm": 2.6026065349578857, + "learning_rate": 0.0001963627700389764, + "loss": 0.8375, + "step": 6545 + }, + { + "epoch": 0.45423634723475126, + "grad_norm": 2.2953004837036133, + "learning_rate": 0.00019636126832142908, + "loss": 0.8485, + "step": 6546 + }, + { + "epoch": 0.45430573867184787, + "grad_norm": 2.6915297508239746, + "learning_rate": 0.00019635976629968043, + "loss": 0.6766, + "step": 6547 + }, + { + "epoch": 0.45437513010894454, + "grad_norm": 2.68912410736084, + "learning_rate": 0.00019635826397373517, + "loss": 0.9785, + "step": 6548 + }, + { + "epoch": 0.4544445215460412, + "grad_norm": 4.386141777038574, + "learning_rate": 0.00019635676134359807, + "loss": 0.8106, + "step": 6549 + }, + { + "epoch": 0.45451391298313787, + "grad_norm": 2.5295886993408203, + "learning_rate": 0.00019635525840927382, + "loss": 0.9799, + "step": 6550 + }, + { + "epoch": 0.45458330442023454, + "grad_norm": 3.725421190261841, + "learning_rate": 0.00019635375517076722, + "loss": 0.7395, + "step": 6551 + }, + { + "epoch": 0.4546526958573312, + "grad_norm": 3.137399911880493, + "learning_rate": 0.000196352251628083, + "loss": 1.0926, + "step": 6552 + }, + { + "epoch": 0.45472208729442787, + "grad_norm": 3.319084644317627, + "learning_rate": 0.00019635074778122586, + "loss": 1.1503, + "step": 6553 + }, + { + "epoch": 0.45479147873152453, + "grad_norm": 3.0481979846954346, + "learning_rate": 0.0001963492436302006, + "loss": 1.4489, + "step": 6554 + }, + { + "epoch": 0.4548608701686212, + "grad_norm": 3.9061150550842285, + "learning_rate": 0.000196347739175012, + "loss": 0.9458, + "step": 6555 + }, + { + "epoch": 0.45493026160571787, + "grad_norm": 2.835780382156372, + "learning_rate": 0.00019634623441566474, + "loss": 1.1469, + "step": 6556 + }, + { + "epoch": 0.45499965304281453, + "grad_norm": 2.653028726577759, + "learning_rate": 0.0001963447293521636, + "loss": 1.2029, + "step": 6557 + }, + { + "epoch": 0.4550690444799112, + "grad_norm": 3.2448155879974365, + "learning_rate": 0.00019634322398451334, + "loss": 1.206, + "step": 6558 + }, + { + "epoch": 0.45513843591700787, + "grad_norm": 3.0522329807281494, + "learning_rate": 0.00019634171831271866, + "loss": 0.8806, + "step": 6559 + }, + { + "epoch": 0.4552078273541045, + "grad_norm": 2.239858865737915, + "learning_rate": 0.0001963402123367844, + "loss": 1.0291, + "step": 6560 + }, + { + "epoch": 0.45527721879120114, + "grad_norm": 2.6042978763580322, + "learning_rate": 0.00019633870605671524, + "loss": 1.0584, + "step": 6561 + }, + { + "epoch": 0.4553466102282978, + "grad_norm": 2.6792614459991455, + "learning_rate": 0.00019633719947251594, + "loss": 0.8268, + "step": 6562 + }, + { + "epoch": 0.4554160016653945, + "grad_norm": 8.099387168884277, + "learning_rate": 0.00019633569258419131, + "loss": 1.0376, + "step": 6563 + }, + { + "epoch": 0.45548539310249114, + "grad_norm": 3.7025580406188965, + "learning_rate": 0.00019633418539174606, + "loss": 0.929, + "step": 6564 + }, + { + "epoch": 0.4555547845395878, + "grad_norm": 2.5913538932800293, + "learning_rate": 0.00019633267789518496, + "loss": 1.0096, + "step": 6565 + }, + { + "epoch": 0.4556241759766845, + "grad_norm": 1.945050597190857, + "learning_rate": 0.0001963311700945128, + "loss": 0.8092, + "step": 6566 + }, + { + "epoch": 0.45569356741378114, + "grad_norm": 2.638704299926758, + "learning_rate": 0.0001963296619897343, + "loss": 0.9891, + "step": 6567 + }, + { + "epoch": 0.4557629588508778, + "grad_norm": 3.3777568340301514, + "learning_rate": 0.00019632815358085423, + "loss": 1.1308, + "step": 6568 + }, + { + "epoch": 0.4558323502879745, + "grad_norm": 3.105618715286255, + "learning_rate": 0.00019632664486787736, + "loss": 1.0466, + "step": 6569 + }, + { + "epoch": 0.45590174172507114, + "grad_norm": 2.6205217838287354, + "learning_rate": 0.00019632513585080843, + "loss": 0.9909, + "step": 6570 + }, + { + "epoch": 0.4559711331621678, + "grad_norm": 2.9743752479553223, + "learning_rate": 0.00019632362652965223, + "loss": 0.9678, + "step": 6571 + }, + { + "epoch": 0.4560405245992645, + "grad_norm": 3.795966625213623, + "learning_rate": 0.00019632211690441353, + "loss": 1.1729, + "step": 6572 + }, + { + "epoch": 0.45610991603636114, + "grad_norm": 2.271069288253784, + "learning_rate": 0.0001963206069750971, + "loss": 0.991, + "step": 6573 + }, + { + "epoch": 0.45617930747345775, + "grad_norm": 3.1730692386627197, + "learning_rate": 0.00019631909674170764, + "loss": 0.7902, + "step": 6574 + }, + { + "epoch": 0.4562486989105544, + "grad_norm": 2.7799384593963623, + "learning_rate": 0.00019631758620425, + "loss": 0.976, + "step": 6575 + }, + { + "epoch": 0.4563180903476511, + "grad_norm": 2.479686737060547, + "learning_rate": 0.0001963160753627289, + "loss": 0.7759, + "step": 6576 + }, + { + "epoch": 0.45638748178474775, + "grad_norm": 2.786097288131714, + "learning_rate": 0.00019631456421714913, + "loss": 1.0387, + "step": 6577 + }, + { + "epoch": 0.4564568732218444, + "grad_norm": 2.5039970874786377, + "learning_rate": 0.00019631305276751547, + "loss": 0.8146, + "step": 6578 + }, + { + "epoch": 0.4565262646589411, + "grad_norm": 3.680265188217163, + "learning_rate": 0.00019631154101383268, + "loss": 1.0597, + "step": 6579 + }, + { + "epoch": 0.45659565609603775, + "grad_norm": 4.064315319061279, + "learning_rate": 0.0001963100289561055, + "loss": 1.0371, + "step": 6580 + }, + { + "epoch": 0.4566650475331344, + "grad_norm": 2.637289524078369, + "learning_rate": 0.00019630851659433872, + "loss": 0.9549, + "step": 6581 + }, + { + "epoch": 0.4567344389702311, + "grad_norm": 2.3779914379119873, + "learning_rate": 0.00019630700392853716, + "loss": 1.1047, + "step": 6582 + }, + { + "epoch": 0.45680383040732775, + "grad_norm": 3.4930880069732666, + "learning_rate": 0.00019630549095870557, + "loss": 0.9749, + "step": 6583 + }, + { + "epoch": 0.4568732218444244, + "grad_norm": 3.094738006591797, + "learning_rate": 0.00019630397768484868, + "loss": 1.0631, + "step": 6584 + }, + { + "epoch": 0.4569426132815211, + "grad_norm": 2.4199771881103516, + "learning_rate": 0.00019630246410697134, + "loss": 0.7852, + "step": 6585 + }, + { + "epoch": 0.45701200471861775, + "grad_norm": 2.606186628341675, + "learning_rate": 0.00019630095022507828, + "loss": 0.8884, + "step": 6586 + }, + { + "epoch": 0.45708139615571436, + "grad_norm": 4.021385669708252, + "learning_rate": 0.0001962994360391743, + "loss": 1.2574, + "step": 6587 + }, + { + "epoch": 0.457150787592811, + "grad_norm": 2.8892288208007812, + "learning_rate": 0.00019629792154926415, + "loss": 0.9347, + "step": 6588 + }, + { + "epoch": 0.4572201790299077, + "grad_norm": 2.7521042823791504, + "learning_rate": 0.00019629640675535265, + "loss": 1.1483, + "step": 6589 + }, + { + "epoch": 0.45728957046700436, + "grad_norm": 3.6856656074523926, + "learning_rate": 0.00019629489165744456, + "loss": 0.9824, + "step": 6590 + }, + { + "epoch": 0.457358961904101, + "grad_norm": 2.721217393875122, + "learning_rate": 0.0001962933762555447, + "loss": 0.7884, + "step": 6591 + }, + { + "epoch": 0.4574283533411977, + "grad_norm": 3.0882599353790283, + "learning_rate": 0.0001962918605496578, + "loss": 1.1231, + "step": 6592 + }, + { + "epoch": 0.45749774477829436, + "grad_norm": 3.1039135456085205, + "learning_rate": 0.00019629034453978866, + "loss": 1.3107, + "step": 6593 + }, + { + "epoch": 0.457567136215391, + "grad_norm": 2.6388673782348633, + "learning_rate": 0.0001962888282259421, + "loss": 0.7556, + "step": 6594 + }, + { + "epoch": 0.4576365276524877, + "grad_norm": 3.050318956375122, + "learning_rate": 0.00019628731160812286, + "loss": 0.997, + "step": 6595 + }, + { + "epoch": 0.45770591908958436, + "grad_norm": 2.966560125350952, + "learning_rate": 0.00019628579468633576, + "loss": 0.9576, + "step": 6596 + }, + { + "epoch": 0.457775310526681, + "grad_norm": 2.6109273433685303, + "learning_rate": 0.0001962842774605856, + "loss": 0.7471, + "step": 6597 + }, + { + "epoch": 0.4578447019637777, + "grad_norm": 2.7637782096862793, + "learning_rate": 0.00019628275993087708, + "loss": 0.8546, + "step": 6598 + }, + { + "epoch": 0.45791409340087436, + "grad_norm": 3.697324514389038, + "learning_rate": 0.0001962812420972151, + "loss": 0.951, + "step": 6599 + }, + { + "epoch": 0.45798348483797097, + "grad_norm": 3.2985424995422363, + "learning_rate": 0.00019627972395960442, + "loss": 0.9047, + "step": 6600 + }, + { + "epoch": 0.45805287627506763, + "grad_norm": 3.2183938026428223, + "learning_rate": 0.0001962782055180498, + "loss": 1.0415, + "step": 6601 + }, + { + "epoch": 0.4581222677121643, + "grad_norm": 3.280635118484497, + "learning_rate": 0.00019627668677255606, + "loss": 0.8917, + "step": 6602 + }, + { + "epoch": 0.45819165914926097, + "grad_norm": 2.927556276321411, + "learning_rate": 0.000196275167723128, + "loss": 0.9768, + "step": 6603 + }, + { + "epoch": 0.45826105058635763, + "grad_norm": 2.6582868099212646, + "learning_rate": 0.00019627364836977037, + "loss": 0.8293, + "step": 6604 + }, + { + "epoch": 0.4583304420234543, + "grad_norm": 2.4253923892974854, + "learning_rate": 0.00019627212871248806, + "loss": 0.9908, + "step": 6605 + }, + { + "epoch": 0.45839983346055097, + "grad_norm": 3.656200647354126, + "learning_rate": 0.00019627060875128578, + "loss": 0.9386, + "step": 6606 + }, + { + "epoch": 0.45846922489764763, + "grad_norm": 2.986926317214966, + "learning_rate": 0.00019626908848616834, + "loss": 1.0395, + "step": 6607 + }, + { + "epoch": 0.4585386163347443, + "grad_norm": 2.7035598754882812, + "learning_rate": 0.00019626756791714058, + "loss": 0.9733, + "step": 6608 + }, + { + "epoch": 0.45860800777184096, + "grad_norm": 2.785731792449951, + "learning_rate": 0.00019626604704420727, + "loss": 0.9048, + "step": 6609 + }, + { + "epoch": 0.45867739920893763, + "grad_norm": 2.759763240814209, + "learning_rate": 0.0001962645258673732, + "loss": 0.9811, + "step": 6610 + }, + { + "epoch": 0.4587467906460343, + "grad_norm": 2.4331393241882324, + "learning_rate": 0.00019626300438664322, + "loss": 1.062, + "step": 6611 + }, + { + "epoch": 0.45881618208313096, + "grad_norm": 2.7711713314056396, + "learning_rate": 0.0001962614826020221, + "loss": 0.9492, + "step": 6612 + }, + { + "epoch": 0.45888557352022763, + "grad_norm": 2.630572557449341, + "learning_rate": 0.00019625996051351465, + "loss": 0.7902, + "step": 6613 + }, + { + "epoch": 0.45895496495732424, + "grad_norm": 2.558973789215088, + "learning_rate": 0.00019625843812112566, + "loss": 0.885, + "step": 6614 + }, + { + "epoch": 0.4590243563944209, + "grad_norm": 3.515507936477661, + "learning_rate": 0.00019625691542485995, + "loss": 1.0485, + "step": 6615 + }, + { + "epoch": 0.4590937478315176, + "grad_norm": 2.5461630821228027, + "learning_rate": 0.00019625539242472233, + "loss": 1.0546, + "step": 6616 + }, + { + "epoch": 0.45916313926861424, + "grad_norm": 2.453585147857666, + "learning_rate": 0.00019625386912071758, + "loss": 0.805, + "step": 6617 + }, + { + "epoch": 0.4592325307057109, + "grad_norm": 3.0459413528442383, + "learning_rate": 0.00019625234551285057, + "loss": 0.7285, + "step": 6618 + }, + { + "epoch": 0.4593019221428076, + "grad_norm": 3.5103957653045654, + "learning_rate": 0.00019625082160112605, + "loss": 0.8229, + "step": 6619 + }, + { + "epoch": 0.45937131357990424, + "grad_norm": 2.299517869949341, + "learning_rate": 0.0001962492973855489, + "loss": 1.0219, + "step": 6620 + }, + { + "epoch": 0.4594407050170009, + "grad_norm": 3.333285331726074, + "learning_rate": 0.00019624777286612383, + "loss": 1.0799, + "step": 6621 + }, + { + "epoch": 0.45951009645409757, + "grad_norm": 2.4549760818481445, + "learning_rate": 0.00019624624804285572, + "loss": 0.9945, + "step": 6622 + }, + { + "epoch": 0.45957948789119424, + "grad_norm": 3.1488893032073975, + "learning_rate": 0.00019624472291574936, + "loss": 0.8656, + "step": 6623 + }, + { + "epoch": 0.4596488793282909, + "grad_norm": 2.4250354766845703, + "learning_rate": 0.00019624319748480962, + "loss": 0.8055, + "step": 6624 + }, + { + "epoch": 0.45971827076538757, + "grad_norm": 2.7750556468963623, + "learning_rate": 0.00019624167175004122, + "loss": 0.9681, + "step": 6625 + }, + { + "epoch": 0.45978766220248424, + "grad_norm": 3.281108856201172, + "learning_rate": 0.00019624014571144907, + "loss": 1.037, + "step": 6626 + }, + { + "epoch": 0.45985705363958085, + "grad_norm": 2.5303750038146973, + "learning_rate": 0.00019623861936903794, + "loss": 1.0401, + "step": 6627 + }, + { + "epoch": 0.4599264450766775, + "grad_norm": 3.018707513809204, + "learning_rate": 0.00019623709272281262, + "loss": 0.8965, + "step": 6628 + }, + { + "epoch": 0.4599958365137742, + "grad_norm": 3.373516082763672, + "learning_rate": 0.000196235565772778, + "loss": 1.4092, + "step": 6629 + }, + { + "epoch": 0.46006522795087085, + "grad_norm": 2.5964789390563965, + "learning_rate": 0.00019623403851893884, + "loss": 0.6055, + "step": 6630 + }, + { + "epoch": 0.4601346193879675, + "grad_norm": 3.6151628494262695, + "learning_rate": 0.0001962325109613, + "loss": 0.9699, + "step": 6631 + }, + { + "epoch": 0.4602040108250642, + "grad_norm": 2.6051485538482666, + "learning_rate": 0.00019623098309986628, + "loss": 1.1402, + "step": 6632 + }, + { + "epoch": 0.46027340226216085, + "grad_norm": 2.698078155517578, + "learning_rate": 0.00019622945493464252, + "loss": 0.9026, + "step": 6633 + }, + { + "epoch": 0.4603427936992575, + "grad_norm": 2.4782211780548096, + "learning_rate": 0.00019622792646563353, + "loss": 1.0155, + "step": 6634 + }, + { + "epoch": 0.4604121851363542, + "grad_norm": 2.3655221462249756, + "learning_rate": 0.00019622639769284415, + "loss": 0.7335, + "step": 6635 + }, + { + "epoch": 0.46048157657345085, + "grad_norm": 2.289057493209839, + "learning_rate": 0.0001962248686162792, + "loss": 0.694, + "step": 6636 + }, + { + "epoch": 0.4605509680105475, + "grad_norm": 3.258976459503174, + "learning_rate": 0.00019622333923594346, + "loss": 1.0073, + "step": 6637 + }, + { + "epoch": 0.4606203594476442, + "grad_norm": 2.805342674255371, + "learning_rate": 0.00019622180955184185, + "loss": 0.8012, + "step": 6638 + }, + { + "epoch": 0.46068975088474085, + "grad_norm": 3.261047601699829, + "learning_rate": 0.00019622027956397916, + "loss": 1.0974, + "step": 6639 + }, + { + "epoch": 0.4607591423218375, + "grad_norm": 2.5228681564331055, + "learning_rate": 0.00019621874927236018, + "loss": 1.025, + "step": 6640 + }, + { + "epoch": 0.4608285337589341, + "grad_norm": 3.140185594558716, + "learning_rate": 0.00019621721867698976, + "loss": 1.1463, + "step": 6641 + }, + { + "epoch": 0.4608979251960308, + "grad_norm": 3.4498584270477295, + "learning_rate": 0.00019621568777787276, + "loss": 1.0445, + "step": 6642 + }, + { + "epoch": 0.46096731663312746, + "grad_norm": 3.2819125652313232, + "learning_rate": 0.000196214156575014, + "loss": 1.2922, + "step": 6643 + }, + { + "epoch": 0.4610367080702241, + "grad_norm": 3.4082069396972656, + "learning_rate": 0.0001962126250684183, + "loss": 1.112, + "step": 6644 + }, + { + "epoch": 0.4611060995073208, + "grad_norm": 2.8586325645446777, + "learning_rate": 0.00019621109325809052, + "loss": 1.0636, + "step": 6645 + }, + { + "epoch": 0.46117549094441745, + "grad_norm": 2.4012696743011475, + "learning_rate": 0.00019620956114403549, + "loss": 0.7417, + "step": 6646 + }, + { + "epoch": 0.4612448823815141, + "grad_norm": 2.5255775451660156, + "learning_rate": 0.00019620802872625804, + "loss": 1.1016, + "step": 6647 + }, + { + "epoch": 0.4613142738186108, + "grad_norm": 3.9479520320892334, + "learning_rate": 0.00019620649600476296, + "loss": 1.1208, + "step": 6648 + }, + { + "epoch": 0.46138366525570745, + "grad_norm": 3.1973037719726562, + "learning_rate": 0.00019620496297955518, + "loss": 1.1311, + "step": 6649 + }, + { + "epoch": 0.4614530566928041, + "grad_norm": 2.341710090637207, + "learning_rate": 0.00019620342965063947, + "loss": 0.9409, + "step": 6650 + }, + { + "epoch": 0.4615224481299008, + "grad_norm": 3.390594959259033, + "learning_rate": 0.0001962018960180207, + "loss": 0.8891, + "step": 6651 + }, + { + "epoch": 0.46159183956699745, + "grad_norm": 2.9156711101531982, + "learning_rate": 0.0001962003620817037, + "loss": 1.0014, + "step": 6652 + }, + { + "epoch": 0.4616612310040941, + "grad_norm": 9.515340805053711, + "learning_rate": 0.00019619882784169336, + "loss": 0.8008, + "step": 6653 + }, + { + "epoch": 0.46173062244119073, + "grad_norm": 2.227454900741577, + "learning_rate": 0.00019619729329799443, + "loss": 0.8059, + "step": 6654 + }, + { + "epoch": 0.4618000138782874, + "grad_norm": 1.9091235399246216, + "learning_rate": 0.00019619575845061183, + "loss": 0.8632, + "step": 6655 + }, + { + "epoch": 0.46186940531538406, + "grad_norm": 3.3647563457489014, + "learning_rate": 0.00019619422329955036, + "loss": 1.2875, + "step": 6656 + }, + { + "epoch": 0.46193879675248073, + "grad_norm": 3.099292516708374, + "learning_rate": 0.00019619268784481491, + "loss": 1.0419, + "step": 6657 + }, + { + "epoch": 0.4620081881895774, + "grad_norm": 2.6796016693115234, + "learning_rate": 0.00019619115208641027, + "loss": 0.8645, + "step": 6658 + }, + { + "epoch": 0.46207757962667406, + "grad_norm": 2.7035067081451416, + "learning_rate": 0.00019618961602434136, + "loss": 1.0067, + "step": 6659 + }, + { + "epoch": 0.46214697106377073, + "grad_norm": 2.7538259029388428, + "learning_rate": 0.00019618807965861298, + "loss": 1.0418, + "step": 6660 + }, + { + "epoch": 0.4622163625008674, + "grad_norm": 3.910640239715576, + "learning_rate": 0.00019618654298923002, + "loss": 1.1568, + "step": 6661 + }, + { + "epoch": 0.46228575393796406, + "grad_norm": 2.4722094535827637, + "learning_rate": 0.00019618500601619726, + "loss": 1.1242, + "step": 6662 + }, + { + "epoch": 0.4623551453750607, + "grad_norm": 2.6328330039978027, + "learning_rate": 0.0001961834687395196, + "loss": 0.7598, + "step": 6663 + }, + { + "epoch": 0.4624245368121574, + "grad_norm": 3.551734685897827, + "learning_rate": 0.0001961819311592019, + "loss": 1.1886, + "step": 6664 + }, + { + "epoch": 0.46249392824925406, + "grad_norm": 2.9417295455932617, + "learning_rate": 0.00019618039327524902, + "loss": 0.8828, + "step": 6665 + }, + { + "epoch": 0.4625633196863507, + "grad_norm": 4.953401565551758, + "learning_rate": 0.0001961788550876658, + "loss": 1.0664, + "step": 6666 + }, + { + "epoch": 0.4626327111234474, + "grad_norm": 3.191923141479492, + "learning_rate": 0.00019617731659645707, + "loss": 1.2056, + "step": 6667 + }, + { + "epoch": 0.462702102560544, + "grad_norm": 2.7779829502105713, + "learning_rate": 0.00019617577780162772, + "loss": 1.1477, + "step": 6668 + }, + { + "epoch": 0.46277149399764067, + "grad_norm": 4.696334362030029, + "learning_rate": 0.0001961742387031826, + "loss": 0.7175, + "step": 6669 + }, + { + "epoch": 0.46284088543473734, + "grad_norm": 2.568023204803467, + "learning_rate": 0.00019617269930112655, + "loss": 0.9065, + "step": 6670 + }, + { + "epoch": 0.462910276871834, + "grad_norm": 2.7312188148498535, + "learning_rate": 0.00019617115959546445, + "loss": 0.8652, + "step": 6671 + }, + { + "epoch": 0.46297966830893067, + "grad_norm": 2.7479755878448486, + "learning_rate": 0.00019616961958620116, + "loss": 0.9162, + "step": 6672 + }, + { + "epoch": 0.46304905974602734, + "grad_norm": 3.0208706855773926, + "learning_rate": 0.00019616807927334156, + "loss": 0.8877, + "step": 6673 + }, + { + "epoch": 0.463118451183124, + "grad_norm": 3.223196029663086, + "learning_rate": 0.00019616653865689045, + "loss": 1.0969, + "step": 6674 + }, + { + "epoch": 0.46318784262022067, + "grad_norm": 2.618009090423584, + "learning_rate": 0.00019616499773685277, + "loss": 0.894, + "step": 6675 + }, + { + "epoch": 0.46325723405731734, + "grad_norm": 2.6077663898468018, + "learning_rate": 0.00019616345651323337, + "loss": 0.865, + "step": 6676 + }, + { + "epoch": 0.463326625494414, + "grad_norm": 2.782386541366577, + "learning_rate": 0.00019616191498603703, + "loss": 0.9232, + "step": 6677 + }, + { + "epoch": 0.46339601693151067, + "grad_norm": 2.814368486404419, + "learning_rate": 0.00019616037315526872, + "loss": 0.943, + "step": 6678 + }, + { + "epoch": 0.46346540836860733, + "grad_norm": 2.7093143463134766, + "learning_rate": 0.00019615883102093326, + "loss": 0.9518, + "step": 6679 + }, + { + "epoch": 0.463534799805704, + "grad_norm": 3.10850191116333, + "learning_rate": 0.00019615728858303553, + "loss": 1.0017, + "step": 6680 + }, + { + "epoch": 0.4636041912428006, + "grad_norm": 3.188304901123047, + "learning_rate": 0.0001961557458415804, + "loss": 0.8542, + "step": 6681 + }, + { + "epoch": 0.4636735826798973, + "grad_norm": 2.948753833770752, + "learning_rate": 0.00019615420279657273, + "loss": 0.9581, + "step": 6682 + }, + { + "epoch": 0.46374297411699394, + "grad_norm": 4.142132759094238, + "learning_rate": 0.0001961526594480174, + "loss": 0.7346, + "step": 6683 + }, + { + "epoch": 0.4638123655540906, + "grad_norm": 2.976065158843994, + "learning_rate": 0.00019615111579591928, + "loss": 1.0516, + "step": 6684 + }, + { + "epoch": 0.4638817569911873, + "grad_norm": 4.82448148727417, + "learning_rate": 0.00019614957184028325, + "loss": 0.9904, + "step": 6685 + }, + { + "epoch": 0.46395114842828394, + "grad_norm": 2.7478785514831543, + "learning_rate": 0.00019614802758111417, + "loss": 0.9433, + "step": 6686 + }, + { + "epoch": 0.4640205398653806, + "grad_norm": 3.281489372253418, + "learning_rate": 0.00019614648301841695, + "loss": 0.9179, + "step": 6687 + }, + { + "epoch": 0.4640899313024773, + "grad_norm": 3.0961666107177734, + "learning_rate": 0.0001961449381521964, + "loss": 0.9232, + "step": 6688 + }, + { + "epoch": 0.46415932273957394, + "grad_norm": 2.6337218284606934, + "learning_rate": 0.00019614339298245744, + "loss": 0.9994, + "step": 6689 + }, + { + "epoch": 0.4642287141766706, + "grad_norm": 2.9134955406188965, + "learning_rate": 0.00019614184750920495, + "loss": 1.0457, + "step": 6690 + }, + { + "epoch": 0.4642981056137673, + "grad_norm": 2.605541229248047, + "learning_rate": 0.00019614030173244377, + "loss": 0.8039, + "step": 6691 + }, + { + "epoch": 0.46436749705086394, + "grad_norm": 3.20090913772583, + "learning_rate": 0.00019613875565217885, + "loss": 1.2001, + "step": 6692 + }, + { + "epoch": 0.4644368884879606, + "grad_norm": 2.095729112625122, + "learning_rate": 0.00019613720926841502, + "loss": 0.8673, + "step": 6693 + }, + { + "epoch": 0.4645062799250573, + "grad_norm": 2.699666976928711, + "learning_rate": 0.00019613566258115718, + "loss": 1.1282, + "step": 6694 + }, + { + "epoch": 0.4645756713621539, + "grad_norm": 3.154738426208496, + "learning_rate": 0.00019613411559041023, + "loss": 1.0691, + "step": 6695 + }, + { + "epoch": 0.46464506279925055, + "grad_norm": 2.428719997406006, + "learning_rate": 0.000196132568296179, + "loss": 0.9036, + "step": 6696 + }, + { + "epoch": 0.4647144542363472, + "grad_norm": 3.567899227142334, + "learning_rate": 0.00019613102069846843, + "loss": 0.9981, + "step": 6697 + }, + { + "epoch": 0.4647838456734439, + "grad_norm": 3.27594256401062, + "learning_rate": 0.00019612947279728336, + "loss": 0.8777, + "step": 6698 + }, + { + "epoch": 0.46485323711054055, + "grad_norm": 2.549464702606201, + "learning_rate": 0.00019612792459262868, + "loss": 0.9663, + "step": 6699 + }, + { + "epoch": 0.4649226285476372, + "grad_norm": 3.6119046211242676, + "learning_rate": 0.00019612637608450932, + "loss": 1.0798, + "step": 6700 + }, + { + "epoch": 0.4649920199847339, + "grad_norm": 2.911907434463501, + "learning_rate": 0.00019612482727293014, + "loss": 1.2404, + "step": 6701 + }, + { + "epoch": 0.46506141142183055, + "grad_norm": 3.9273123741149902, + "learning_rate": 0.00019612327815789603, + "loss": 0.8961, + "step": 6702 + }, + { + "epoch": 0.4651308028589272, + "grad_norm": 3.265676975250244, + "learning_rate": 0.0001961217287394119, + "loss": 0.9954, + "step": 6703 + }, + { + "epoch": 0.4652001942960239, + "grad_norm": 2.338270664215088, + "learning_rate": 0.0001961201790174826, + "loss": 0.9595, + "step": 6704 + }, + { + "epoch": 0.46526958573312055, + "grad_norm": 2.5982232093811035, + "learning_rate": 0.00019611862899211307, + "loss": 0.7655, + "step": 6705 + }, + { + "epoch": 0.4653389771702172, + "grad_norm": 2.536384105682373, + "learning_rate": 0.00019611707866330815, + "loss": 0.8904, + "step": 6706 + }, + { + "epoch": 0.4654083686073139, + "grad_norm": 2.628779172897339, + "learning_rate": 0.0001961155280310728, + "loss": 0.8437, + "step": 6707 + }, + { + "epoch": 0.4654777600444105, + "grad_norm": 2.9244351387023926, + "learning_rate": 0.00019611397709541186, + "loss": 0.8839, + "step": 6708 + }, + { + "epoch": 0.46554715148150716, + "grad_norm": 3.698870897293091, + "learning_rate": 0.00019611242585633023, + "loss": 1.0005, + "step": 6709 + }, + { + "epoch": 0.4656165429186038, + "grad_norm": 2.5637104511260986, + "learning_rate": 0.00019611087431383286, + "loss": 0.8603, + "step": 6710 + }, + { + "epoch": 0.4656859343557005, + "grad_norm": 3.266883373260498, + "learning_rate": 0.00019610932246792457, + "loss": 0.9819, + "step": 6711 + }, + { + "epoch": 0.46575532579279716, + "grad_norm": 2.5515012741088867, + "learning_rate": 0.00019610777031861033, + "loss": 0.8488, + "step": 6712 + }, + { + "epoch": 0.4658247172298938, + "grad_norm": 2.571565628051758, + "learning_rate": 0.000196106217865895, + "loss": 0.9107, + "step": 6713 + }, + { + "epoch": 0.4658941086669905, + "grad_norm": 2.4480085372924805, + "learning_rate": 0.0001961046651097835, + "loss": 0.8275, + "step": 6714 + }, + { + "epoch": 0.46596350010408716, + "grad_norm": 2.344186544418335, + "learning_rate": 0.00019610311205028072, + "loss": 0.8262, + "step": 6715 + }, + { + "epoch": 0.4660328915411838, + "grad_norm": 2.936904191970825, + "learning_rate": 0.00019610155868739153, + "loss": 0.897, + "step": 6716 + }, + { + "epoch": 0.4661022829782805, + "grad_norm": 4.369073390960693, + "learning_rate": 0.00019610000502112088, + "loss": 1.3062, + "step": 6717 + }, + { + "epoch": 0.46617167441537716, + "grad_norm": 2.7046902179718018, + "learning_rate": 0.0001960984510514737, + "loss": 1.0028, + "step": 6718 + }, + { + "epoch": 0.4662410658524738, + "grad_norm": 3.256906270980835, + "learning_rate": 0.00019609689677845484, + "loss": 1.1048, + "step": 6719 + }, + { + "epoch": 0.4663104572895705, + "grad_norm": 2.1145870685577393, + "learning_rate": 0.00019609534220206922, + "loss": 0.853, + "step": 6720 + }, + { + "epoch": 0.46637984872666716, + "grad_norm": 2.2145164012908936, + "learning_rate": 0.00019609378732232178, + "loss": 0.7448, + "step": 6721 + }, + { + "epoch": 0.46644924016376377, + "grad_norm": 4.641626358032227, + "learning_rate": 0.00019609223213921736, + "loss": 1.0249, + "step": 6722 + }, + { + "epoch": 0.46651863160086043, + "grad_norm": 3.2394912242889404, + "learning_rate": 0.00019609067665276095, + "loss": 1.2233, + "step": 6723 + }, + { + "epoch": 0.4665880230379571, + "grad_norm": 3.7474563121795654, + "learning_rate": 0.00019608912086295738, + "loss": 0.7776, + "step": 6724 + }, + { + "epoch": 0.46665741447505377, + "grad_norm": 3.5882408618927, + "learning_rate": 0.00019608756476981165, + "loss": 1.4501, + "step": 6725 + }, + { + "epoch": 0.46672680591215043, + "grad_norm": 2.5483944416046143, + "learning_rate": 0.0001960860083733286, + "loss": 0.9245, + "step": 6726 + }, + { + "epoch": 0.4667961973492471, + "grad_norm": 3.2691524028778076, + "learning_rate": 0.0001960844516735132, + "loss": 1.0051, + "step": 6727 + }, + { + "epoch": 0.46686558878634377, + "grad_norm": 3.256169319152832, + "learning_rate": 0.0001960828946703703, + "loss": 1.1529, + "step": 6728 + }, + { + "epoch": 0.46693498022344043, + "grad_norm": 2.1997671127319336, + "learning_rate": 0.00019608133736390487, + "loss": 0.9575, + "step": 6729 + }, + { + "epoch": 0.4670043716605371, + "grad_norm": 2.731943130493164, + "learning_rate": 0.00019607977975412178, + "loss": 0.9633, + "step": 6730 + }, + { + "epoch": 0.46707376309763377, + "grad_norm": 5.463911056518555, + "learning_rate": 0.00019607822184102597, + "loss": 1.1047, + "step": 6731 + }, + { + "epoch": 0.46714315453473043, + "grad_norm": 2.581538677215576, + "learning_rate": 0.00019607666362462237, + "loss": 0.8952, + "step": 6732 + }, + { + "epoch": 0.4672125459718271, + "grad_norm": 3.6144332885742188, + "learning_rate": 0.00019607510510491594, + "loss": 1.0735, + "step": 6733 + }, + { + "epoch": 0.46728193740892376, + "grad_norm": 2.3858652114868164, + "learning_rate": 0.00019607354628191148, + "loss": 0.879, + "step": 6734 + }, + { + "epoch": 0.4673513288460204, + "grad_norm": 2.4727060794830322, + "learning_rate": 0.00019607198715561403, + "loss": 0.7995, + "step": 6735 + }, + { + "epoch": 0.46742072028311704, + "grad_norm": 2.368250846862793, + "learning_rate": 0.00019607042772602844, + "loss": 0.9047, + "step": 6736 + }, + { + "epoch": 0.4674901117202137, + "grad_norm": 2.4737792015075684, + "learning_rate": 0.00019606886799315964, + "loss": 0.8092, + "step": 6737 + }, + { + "epoch": 0.4675595031573104, + "grad_norm": 2.1931419372558594, + "learning_rate": 0.0001960673079570126, + "loss": 0.7206, + "step": 6738 + }, + { + "epoch": 0.46762889459440704, + "grad_norm": 2.4844813346862793, + "learning_rate": 0.0001960657476175922, + "loss": 0.8212, + "step": 6739 + }, + { + "epoch": 0.4676982860315037, + "grad_norm": 4.568305492401123, + "learning_rate": 0.00019606418697490339, + "loss": 1.2218, + "step": 6740 + }, + { + "epoch": 0.4677676774686004, + "grad_norm": 3.137082099914551, + "learning_rate": 0.00019606262602895106, + "loss": 1.1694, + "step": 6741 + }, + { + "epoch": 0.46783706890569704, + "grad_norm": 2.525424003601074, + "learning_rate": 0.0001960610647797402, + "loss": 0.8644, + "step": 6742 + }, + { + "epoch": 0.4679064603427937, + "grad_norm": 3.096817970275879, + "learning_rate": 0.0001960595032272757, + "loss": 1.1108, + "step": 6743 + }, + { + "epoch": 0.4679758517798904, + "grad_norm": 3.564061164855957, + "learning_rate": 0.00019605794137156246, + "loss": 1.2101, + "step": 6744 + }, + { + "epoch": 0.46804524321698704, + "grad_norm": 4.362951278686523, + "learning_rate": 0.00019605637921260547, + "loss": 0.7503, + "step": 6745 + }, + { + "epoch": 0.4681146346540837, + "grad_norm": 2.775665044784546, + "learning_rate": 0.00019605481675040963, + "loss": 0.9102, + "step": 6746 + }, + { + "epoch": 0.46818402609118037, + "grad_norm": 3.4755818843841553, + "learning_rate": 0.0001960532539849799, + "loss": 0.9525, + "step": 6747 + }, + { + "epoch": 0.46825341752827704, + "grad_norm": 3.816856861114502, + "learning_rate": 0.00019605169091632116, + "loss": 1.1265, + "step": 6748 + }, + { + "epoch": 0.46832280896537365, + "grad_norm": 3.8756613731384277, + "learning_rate": 0.00019605012754443836, + "loss": 0.9613, + "step": 6749 + }, + { + "epoch": 0.4683922004024703, + "grad_norm": 2.8270297050476074, + "learning_rate": 0.0001960485638693365, + "loss": 0.9809, + "step": 6750 + }, + { + "epoch": 0.468461591839567, + "grad_norm": 2.930982828140259, + "learning_rate": 0.00019604699989102044, + "loss": 0.9678, + "step": 6751 + }, + { + "epoch": 0.46853098327666365, + "grad_norm": 2.74638032913208, + "learning_rate": 0.00019604543560949514, + "loss": 1.1309, + "step": 6752 + }, + { + "epoch": 0.4686003747137603, + "grad_norm": 2.7267673015594482, + "learning_rate": 0.00019604387102476557, + "loss": 0.9741, + "step": 6753 + }, + { + "epoch": 0.468669766150857, + "grad_norm": 2.741405487060547, + "learning_rate": 0.00019604230613683662, + "loss": 0.8517, + "step": 6754 + }, + { + "epoch": 0.46873915758795365, + "grad_norm": 3.6332058906555176, + "learning_rate": 0.00019604074094571329, + "loss": 1.0314, + "step": 6755 + }, + { + "epoch": 0.4688085490250503, + "grad_norm": 3.526028871536255, + "learning_rate": 0.00019603917545140044, + "loss": 1.0471, + "step": 6756 + }, + { + "epoch": 0.468877940462147, + "grad_norm": 3.5102052688598633, + "learning_rate": 0.00019603760965390305, + "loss": 1.0187, + "step": 6757 + }, + { + "epoch": 0.46894733189924365, + "grad_norm": 2.823519229888916, + "learning_rate": 0.0001960360435532261, + "loss": 0.7428, + "step": 6758 + }, + { + "epoch": 0.4690167233363403, + "grad_norm": 3.8121120929718018, + "learning_rate": 0.00019603447714937447, + "loss": 0.733, + "step": 6759 + }, + { + "epoch": 0.469086114773437, + "grad_norm": 3.31160569190979, + "learning_rate": 0.00019603291044235314, + "loss": 0.9035, + "step": 6760 + }, + { + "epoch": 0.46915550621053365, + "grad_norm": 2.874586820602417, + "learning_rate": 0.00019603134343216708, + "loss": 0.9391, + "step": 6761 + }, + { + "epoch": 0.46922489764763026, + "grad_norm": 3.6496942043304443, + "learning_rate": 0.00019602977611882122, + "loss": 0.8173, + "step": 6762 + }, + { + "epoch": 0.4692942890847269, + "grad_norm": 3.6639106273651123, + "learning_rate": 0.00019602820850232047, + "loss": 1.0888, + "step": 6763 + }, + { + "epoch": 0.4693636805218236, + "grad_norm": 2.8593251705169678, + "learning_rate": 0.00019602664058266983, + "loss": 0.9119, + "step": 6764 + }, + { + "epoch": 0.46943307195892026, + "grad_norm": 3.7211384773254395, + "learning_rate": 0.0001960250723598742, + "loss": 0.8888, + "step": 6765 + }, + { + "epoch": 0.4695024633960169, + "grad_norm": 2.9851059913635254, + "learning_rate": 0.00019602350383393858, + "loss": 0.9, + "step": 6766 + }, + { + "epoch": 0.4695718548331136, + "grad_norm": 3.1369853019714355, + "learning_rate": 0.0001960219350048679, + "loss": 1.1439, + "step": 6767 + }, + { + "epoch": 0.46964124627021026, + "grad_norm": 4.245553970336914, + "learning_rate": 0.00019602036587266708, + "loss": 1.4693, + "step": 6768 + }, + { + "epoch": 0.4697106377073069, + "grad_norm": 3.246954917907715, + "learning_rate": 0.00019601879643734113, + "loss": 0.9234, + "step": 6769 + }, + { + "epoch": 0.4697800291444036, + "grad_norm": 3.1907105445861816, + "learning_rate": 0.00019601722669889497, + "loss": 0.7897, + "step": 6770 + }, + { + "epoch": 0.46984942058150025, + "grad_norm": 3.2793707847595215, + "learning_rate": 0.00019601565665733358, + "loss": 0.9453, + "step": 6771 + }, + { + "epoch": 0.4699188120185969, + "grad_norm": 3.1562955379486084, + "learning_rate": 0.0001960140863126619, + "loss": 0.9832, + "step": 6772 + }, + { + "epoch": 0.4699882034556936, + "grad_norm": 2.7589287757873535, + "learning_rate": 0.0001960125156648849, + "loss": 0.9555, + "step": 6773 + }, + { + "epoch": 0.47005759489279025, + "grad_norm": 2.4921584129333496, + "learning_rate": 0.0001960109447140075, + "loss": 0.9753, + "step": 6774 + }, + { + "epoch": 0.47012698632988686, + "grad_norm": 2.81583571434021, + "learning_rate": 0.0001960093734600347, + "loss": 0.6774, + "step": 6775 + }, + { + "epoch": 0.47019637776698353, + "grad_norm": 2.542879343032837, + "learning_rate": 0.0001960078019029714, + "loss": 1.0509, + "step": 6776 + }, + { + "epoch": 0.4702657692040802, + "grad_norm": 4.090960502624512, + "learning_rate": 0.0001960062300428227, + "loss": 0.9718, + "step": 6777 + }, + { + "epoch": 0.47033516064117686, + "grad_norm": 2.876368761062622, + "learning_rate": 0.00019600465787959342, + "loss": 0.9434, + "step": 6778 + }, + { + "epoch": 0.47040455207827353, + "grad_norm": 4.947535514831543, + "learning_rate": 0.00019600308541328854, + "loss": 1.0204, + "step": 6779 + }, + { + "epoch": 0.4704739435153702, + "grad_norm": 3.518946409225464, + "learning_rate": 0.00019600151264391311, + "loss": 0.97, + "step": 6780 + }, + { + "epoch": 0.47054333495246686, + "grad_norm": 3.649099111557007, + "learning_rate": 0.00019599993957147204, + "loss": 1.0802, + "step": 6781 + }, + { + "epoch": 0.47061272638956353, + "grad_norm": 3.0779898166656494, + "learning_rate": 0.00019599836619597027, + "loss": 1.1151, + "step": 6782 + }, + { + "epoch": 0.4706821178266602, + "grad_norm": 3.766951322555542, + "learning_rate": 0.00019599679251741282, + "loss": 1.0612, + "step": 6783 + }, + { + "epoch": 0.47075150926375686, + "grad_norm": 3.198359489440918, + "learning_rate": 0.00019599521853580463, + "loss": 0.8362, + "step": 6784 + }, + { + "epoch": 0.47082090070085353, + "grad_norm": 3.532526731491089, + "learning_rate": 0.00019599364425115066, + "loss": 1.045, + "step": 6785 + }, + { + "epoch": 0.4708902921379502, + "grad_norm": 3.5132253170013428, + "learning_rate": 0.0001959920696634559, + "loss": 1.0997, + "step": 6786 + }, + { + "epoch": 0.47095968357504686, + "grad_norm": 4.731918811798096, + "learning_rate": 0.0001959904947727253, + "loss": 0.9014, + "step": 6787 + }, + { + "epoch": 0.4710290750121435, + "grad_norm": 2.4380764961242676, + "learning_rate": 0.0001959889195789639, + "loss": 0.865, + "step": 6788 + }, + { + "epoch": 0.47109846644924014, + "grad_norm": 2.667926549911499, + "learning_rate": 0.00019598734408217655, + "loss": 0.8861, + "step": 6789 + }, + { + "epoch": 0.4711678578863368, + "grad_norm": 3.3288936614990234, + "learning_rate": 0.0001959857682823683, + "loss": 0.9157, + "step": 6790 + }, + { + "epoch": 0.47123724932343347, + "grad_norm": 2.597430467605591, + "learning_rate": 0.00019598419217954416, + "loss": 0.8243, + "step": 6791 + }, + { + "epoch": 0.47130664076053014, + "grad_norm": 2.8403093814849854, + "learning_rate": 0.00019598261577370902, + "loss": 1.0296, + "step": 6792 + }, + { + "epoch": 0.4713760321976268, + "grad_norm": 3.020427703857422, + "learning_rate": 0.00019598103906486792, + "loss": 1.2804, + "step": 6793 + }, + { + "epoch": 0.47144542363472347, + "grad_norm": 2.642486333847046, + "learning_rate": 0.00019597946205302582, + "loss": 0.8989, + "step": 6794 + }, + { + "epoch": 0.47151481507182014, + "grad_norm": 3.560279607772827, + "learning_rate": 0.00019597788473818767, + "loss": 0.9045, + "step": 6795 + }, + { + "epoch": 0.4715842065089168, + "grad_norm": 3.144524335861206, + "learning_rate": 0.0001959763071203585, + "loss": 0.9808, + "step": 6796 + }, + { + "epoch": 0.47165359794601347, + "grad_norm": 3.6675310134887695, + "learning_rate": 0.00019597472919954324, + "loss": 1.1219, + "step": 6797 + }, + { + "epoch": 0.47172298938311014, + "grad_norm": 2.6013340950012207, + "learning_rate": 0.0001959731509757469, + "loss": 0.7861, + "step": 6798 + }, + { + "epoch": 0.4717923808202068, + "grad_norm": 2.974424362182617, + "learning_rate": 0.00019597157244897448, + "loss": 0.8693, + "step": 6799 + }, + { + "epoch": 0.47186177225730347, + "grad_norm": 2.5373525619506836, + "learning_rate": 0.00019596999361923091, + "loss": 0.9778, + "step": 6800 + }, + { + "epoch": 0.47193116369440014, + "grad_norm": 2.53629469871521, + "learning_rate": 0.0001959684144865212, + "loss": 0.782, + "step": 6801 + }, + { + "epoch": 0.47200055513149675, + "grad_norm": 3.1166093349456787, + "learning_rate": 0.00019596683505085038, + "loss": 0.9438, + "step": 6802 + }, + { + "epoch": 0.4720699465685934, + "grad_norm": 3.0511553287506104, + "learning_rate": 0.0001959652553122234, + "loss": 1.2113, + "step": 6803 + }, + { + "epoch": 0.4721393380056901, + "grad_norm": 3.359754800796509, + "learning_rate": 0.0001959636752706452, + "loss": 0.9761, + "step": 6804 + }, + { + "epoch": 0.47220872944278675, + "grad_norm": 2.3080732822418213, + "learning_rate": 0.0001959620949261208, + "loss": 1.0464, + "step": 6805 + }, + { + "epoch": 0.4722781208798834, + "grad_norm": 3.399764060974121, + "learning_rate": 0.00019596051427865524, + "loss": 1.1889, + "step": 6806 + }, + { + "epoch": 0.4723475123169801, + "grad_norm": 3.1902730464935303, + "learning_rate": 0.00019595893332825348, + "loss": 0.8634, + "step": 6807 + }, + { + "epoch": 0.47241690375407674, + "grad_norm": 4.251091480255127, + "learning_rate": 0.00019595735207492045, + "loss": 0.8353, + "step": 6808 + }, + { + "epoch": 0.4724862951911734, + "grad_norm": 2.919276237487793, + "learning_rate": 0.00019595577051866126, + "loss": 0.8095, + "step": 6809 + }, + { + "epoch": 0.4725556866282701, + "grad_norm": 2.8930420875549316, + "learning_rate": 0.00019595418865948075, + "loss": 0.9508, + "step": 6810 + }, + { + "epoch": 0.47262507806536674, + "grad_norm": 2.782968044281006, + "learning_rate": 0.00019595260649738408, + "loss": 0.853, + "step": 6811 + }, + { + "epoch": 0.4726944695024634, + "grad_norm": 3.458974838256836, + "learning_rate": 0.00019595102403237614, + "loss": 0.6831, + "step": 6812 + }, + { + "epoch": 0.4727638609395601, + "grad_norm": 2.068138837814331, + "learning_rate": 0.00019594944126446193, + "loss": 0.6125, + "step": 6813 + }, + { + "epoch": 0.47283325237665674, + "grad_norm": 2.395139217376709, + "learning_rate": 0.00019594785819364646, + "loss": 0.8617, + "step": 6814 + }, + { + "epoch": 0.4729026438137534, + "grad_norm": 2.1026365756988525, + "learning_rate": 0.00019594627481993475, + "loss": 0.7854, + "step": 6815 + }, + { + "epoch": 0.47297203525085, + "grad_norm": 2.954573392868042, + "learning_rate": 0.00019594469114333178, + "loss": 1.0001, + "step": 6816 + }, + { + "epoch": 0.4730414266879467, + "grad_norm": 2.8936352729797363, + "learning_rate": 0.00019594310716384254, + "loss": 1.0872, + "step": 6817 + }, + { + "epoch": 0.47311081812504335, + "grad_norm": 2.2436342239379883, + "learning_rate": 0.00019594152288147207, + "loss": 0.8985, + "step": 6818 + }, + { + "epoch": 0.47318020956214, + "grad_norm": 3.8966572284698486, + "learning_rate": 0.00019593993829622534, + "loss": 1.4134, + "step": 6819 + }, + { + "epoch": 0.4732496009992367, + "grad_norm": 2.07395601272583, + "learning_rate": 0.00019593835340810732, + "loss": 0.6397, + "step": 6820 + }, + { + "epoch": 0.47331899243633335, + "grad_norm": 5.461615085601807, + "learning_rate": 0.0001959367682171231, + "loss": 1.0433, + "step": 6821 + }, + { + "epoch": 0.47338838387343, + "grad_norm": 2.8035855293273926, + "learning_rate": 0.0001959351827232776, + "loss": 0.9551, + "step": 6822 + }, + { + "epoch": 0.4734577753105267, + "grad_norm": 2.966890811920166, + "learning_rate": 0.00019593359692657587, + "loss": 1.0764, + "step": 6823 + }, + { + "epoch": 0.47352716674762335, + "grad_norm": 2.7833900451660156, + "learning_rate": 0.00019593201082702288, + "loss": 0.9742, + "step": 6824 + }, + { + "epoch": 0.47359655818472, + "grad_norm": 2.425645112991333, + "learning_rate": 0.0001959304244246237, + "loss": 0.6282, + "step": 6825 + }, + { + "epoch": 0.4736659496218167, + "grad_norm": 2.50527024269104, + "learning_rate": 0.00019592883771938325, + "loss": 0.9098, + "step": 6826 + }, + { + "epoch": 0.47373534105891335, + "grad_norm": 4.9715986251831055, + "learning_rate": 0.00019592725071130665, + "loss": 1.0891, + "step": 6827 + }, + { + "epoch": 0.47380473249601, + "grad_norm": 3.5401248931884766, + "learning_rate": 0.00019592566340039883, + "loss": 1.3047, + "step": 6828 + }, + { + "epoch": 0.47387412393310663, + "grad_norm": 3.6545252799987793, + "learning_rate": 0.0001959240757866648, + "loss": 0.9561, + "step": 6829 + }, + { + "epoch": 0.4739435153702033, + "grad_norm": 2.2309975624084473, + "learning_rate": 0.0001959224878701096, + "loss": 0.7643, + "step": 6830 + }, + { + "epoch": 0.47401290680729996, + "grad_norm": 2.9385039806365967, + "learning_rate": 0.00019592089965073822, + "loss": 0.8866, + "step": 6831 + }, + { + "epoch": 0.4740822982443966, + "grad_norm": 3.21191668510437, + "learning_rate": 0.00019591931112855572, + "loss": 0.8981, + "step": 6832 + }, + { + "epoch": 0.4741516896814933, + "grad_norm": 3.0583348274230957, + "learning_rate": 0.0001959177223035671, + "loss": 0.8052, + "step": 6833 + }, + { + "epoch": 0.47422108111858996, + "grad_norm": 2.8860976696014404, + "learning_rate": 0.00019591613317577732, + "loss": 0.9812, + "step": 6834 + }, + { + "epoch": 0.4742904725556866, + "grad_norm": 2.4937546253204346, + "learning_rate": 0.00019591454374519142, + "loss": 0.8708, + "step": 6835 + }, + { + "epoch": 0.4743598639927833, + "grad_norm": 3.8062081336975098, + "learning_rate": 0.0001959129540118145, + "loss": 1.1248, + "step": 6836 + }, + { + "epoch": 0.47442925542987996, + "grad_norm": 1.9161521196365356, + "learning_rate": 0.00019591136397565143, + "loss": 0.7049, + "step": 6837 + }, + { + "epoch": 0.4744986468669766, + "grad_norm": 3.394442558288574, + "learning_rate": 0.00019590977363670735, + "loss": 0.8959, + "step": 6838 + }, + { + "epoch": 0.4745680383040733, + "grad_norm": 3.575659990310669, + "learning_rate": 0.00019590818299498725, + "loss": 0.8726, + "step": 6839 + }, + { + "epoch": 0.47463742974116996, + "grad_norm": 3.908886194229126, + "learning_rate": 0.00019590659205049614, + "loss": 1.1577, + "step": 6840 + }, + { + "epoch": 0.4747068211782666, + "grad_norm": 3.6863176822662354, + "learning_rate": 0.00019590500080323906, + "loss": 1.0664, + "step": 6841 + }, + { + "epoch": 0.4747762126153633, + "grad_norm": 2.2731075286865234, + "learning_rate": 0.000195903409253221, + "loss": 0.8454, + "step": 6842 + }, + { + "epoch": 0.4748456040524599, + "grad_norm": 2.809846878051758, + "learning_rate": 0.000195901817400447, + "loss": 0.9244, + "step": 6843 + }, + { + "epoch": 0.47491499548955657, + "grad_norm": 2.8021371364593506, + "learning_rate": 0.0001959002252449221, + "loss": 0.9166, + "step": 6844 + }, + { + "epoch": 0.47498438692665323, + "grad_norm": 2.9091529846191406, + "learning_rate": 0.0001958986327866513, + "loss": 0.9859, + "step": 6845 + }, + { + "epoch": 0.4750537783637499, + "grad_norm": 3.1366591453552246, + "learning_rate": 0.00019589704002563967, + "loss": 1.1331, + "step": 6846 + }, + { + "epoch": 0.47512316980084657, + "grad_norm": 2.629429578781128, + "learning_rate": 0.0001958954469618922, + "loss": 0.8699, + "step": 6847 + }, + { + "epoch": 0.47519256123794323, + "grad_norm": 2.4022562503814697, + "learning_rate": 0.00019589385359541393, + "loss": 1.0382, + "step": 6848 + }, + { + "epoch": 0.4752619526750399, + "grad_norm": 3.5177860260009766, + "learning_rate": 0.00019589225992620987, + "loss": 0.8968, + "step": 6849 + }, + { + "epoch": 0.47533134411213657, + "grad_norm": 5.222490310668945, + "learning_rate": 0.0001958906659542851, + "loss": 1.0987, + "step": 6850 + }, + { + "epoch": 0.47540073554923323, + "grad_norm": 3.5586729049682617, + "learning_rate": 0.00019588907167964462, + "loss": 0.8483, + "step": 6851 + }, + { + "epoch": 0.4754701269863299, + "grad_norm": 2.151268720626831, + "learning_rate": 0.00019588747710229344, + "loss": 0.8019, + "step": 6852 + }, + { + "epoch": 0.47553951842342657, + "grad_norm": 2.2474372386932373, + "learning_rate": 0.00019588588222223664, + "loss": 0.8157, + "step": 6853 + }, + { + "epoch": 0.47560890986052323, + "grad_norm": 2.527343988418579, + "learning_rate": 0.00019588428703947922, + "loss": 0.8078, + "step": 6854 + }, + { + "epoch": 0.4756783012976199, + "grad_norm": 4.418898582458496, + "learning_rate": 0.00019588269155402626, + "loss": 1.0232, + "step": 6855 + }, + { + "epoch": 0.4757476927347165, + "grad_norm": 2.7665038108825684, + "learning_rate": 0.0001958810957658827, + "loss": 0.9918, + "step": 6856 + }, + { + "epoch": 0.4758170841718132, + "grad_norm": 3.046778917312622, + "learning_rate": 0.00019587949967505373, + "loss": 1.1663, + "step": 6857 + }, + { + "epoch": 0.47588647560890984, + "grad_norm": 2.871488332748413, + "learning_rate": 0.00019587790328154427, + "loss": 1.1712, + "step": 6858 + }, + { + "epoch": 0.4759558670460065, + "grad_norm": 4.282267093658447, + "learning_rate": 0.0001958763065853594, + "loss": 1.0039, + "step": 6859 + }, + { + "epoch": 0.4760252584831032, + "grad_norm": 2.651569128036499, + "learning_rate": 0.0001958747095865041, + "loss": 0.8078, + "step": 6860 + }, + { + "epoch": 0.47609464992019984, + "grad_norm": 3.2454512119293213, + "learning_rate": 0.00019587311228498353, + "loss": 1.2729, + "step": 6861 + }, + { + "epoch": 0.4761640413572965, + "grad_norm": 2.876049280166626, + "learning_rate": 0.00019587151468080264, + "loss": 1.0363, + "step": 6862 + }, + { + "epoch": 0.4762334327943932, + "grad_norm": 3.6295571327209473, + "learning_rate": 0.00019586991677396652, + "loss": 1.2667, + "step": 6863 + }, + { + "epoch": 0.47630282423148984, + "grad_norm": 2.3878116607666016, + "learning_rate": 0.0001958683185644802, + "loss": 0.9563, + "step": 6864 + }, + { + "epoch": 0.4763722156685865, + "grad_norm": 2.890874147415161, + "learning_rate": 0.0001958667200523487, + "loss": 0.9535, + "step": 6865 + }, + { + "epoch": 0.4764416071056832, + "grad_norm": 3.035268783569336, + "learning_rate": 0.0001958651212375771, + "loss": 0.8114, + "step": 6866 + }, + { + "epoch": 0.47651099854277984, + "grad_norm": 2.8366076946258545, + "learning_rate": 0.0001958635221201704, + "loss": 0.9568, + "step": 6867 + }, + { + "epoch": 0.4765803899798765, + "grad_norm": 3.1948249340057373, + "learning_rate": 0.0001958619227001337, + "loss": 0.991, + "step": 6868 + }, + { + "epoch": 0.4766497814169732, + "grad_norm": 2.867439031600952, + "learning_rate": 0.00019586032297747206, + "loss": 1.0172, + "step": 6869 + }, + { + "epoch": 0.4767191728540698, + "grad_norm": 2.6689956188201904, + "learning_rate": 0.00019585872295219047, + "loss": 0.7559, + "step": 6870 + }, + { + "epoch": 0.47678856429116645, + "grad_norm": 2.652292013168335, + "learning_rate": 0.00019585712262429404, + "loss": 0.88, + "step": 6871 + }, + { + "epoch": 0.4768579557282631, + "grad_norm": 3.86450457572937, + "learning_rate": 0.00019585552199378778, + "loss": 0.9968, + "step": 6872 + }, + { + "epoch": 0.4769273471653598, + "grad_norm": 4.386094093322754, + "learning_rate": 0.00019585392106067674, + "loss": 0.8754, + "step": 6873 + }, + { + "epoch": 0.47699673860245645, + "grad_norm": 2.7098562717437744, + "learning_rate": 0.00019585231982496604, + "loss": 0.9292, + "step": 6874 + }, + { + "epoch": 0.4770661300395531, + "grad_norm": 3.280984878540039, + "learning_rate": 0.00019585071828666064, + "loss": 1.0091, + "step": 6875 + }, + { + "epoch": 0.4771355214766498, + "grad_norm": 2.4392335414886475, + "learning_rate": 0.00019584911644576566, + "loss": 0.8835, + "step": 6876 + }, + { + "epoch": 0.47720491291374645, + "grad_norm": 4.137631893157959, + "learning_rate": 0.00019584751430228617, + "loss": 1.2182, + "step": 6877 + }, + { + "epoch": 0.4772743043508431, + "grad_norm": 2.6979315280914307, + "learning_rate": 0.00019584591185622716, + "loss": 1.0308, + "step": 6878 + }, + { + "epoch": 0.4773436957879398, + "grad_norm": 3.5864555835723877, + "learning_rate": 0.00019584430910759372, + "loss": 1.4726, + "step": 6879 + }, + { + "epoch": 0.47741308722503645, + "grad_norm": 2.66805362701416, + "learning_rate": 0.00019584270605639095, + "loss": 0.8433, + "step": 6880 + }, + { + "epoch": 0.4774824786621331, + "grad_norm": 4.261061668395996, + "learning_rate": 0.00019584110270262387, + "loss": 1.259, + "step": 6881 + }, + { + "epoch": 0.4775518700992298, + "grad_norm": 2.7684824466705322, + "learning_rate": 0.00019583949904629754, + "loss": 0.9391, + "step": 6882 + }, + { + "epoch": 0.4776212615363264, + "grad_norm": 9.341896057128906, + "learning_rate": 0.000195837895087417, + "loss": 0.746, + "step": 6883 + }, + { + "epoch": 0.47769065297342306, + "grad_norm": 2.6623361110687256, + "learning_rate": 0.00019583629082598738, + "loss": 0.8518, + "step": 6884 + }, + { + "epoch": 0.4777600444105197, + "grad_norm": 2.7651913166046143, + "learning_rate": 0.00019583468626201372, + "loss": 1.0354, + "step": 6885 + }, + { + "epoch": 0.4778294358476164, + "grad_norm": 2.8731799125671387, + "learning_rate": 0.00019583308139550104, + "loss": 1.0498, + "step": 6886 + }, + { + "epoch": 0.47789882728471306, + "grad_norm": 3.9662535190582275, + "learning_rate": 0.00019583147622645445, + "loss": 1.001, + "step": 6887 + }, + { + "epoch": 0.4779682187218097, + "grad_norm": 2.666656255722046, + "learning_rate": 0.00019582987075487905, + "loss": 0.9971, + "step": 6888 + }, + { + "epoch": 0.4780376101589064, + "grad_norm": 2.6531310081481934, + "learning_rate": 0.00019582826498077983, + "loss": 0.9424, + "step": 6889 + }, + { + "epoch": 0.47810700159600306, + "grad_norm": 2.8899664878845215, + "learning_rate": 0.00019582665890416188, + "loss": 0.8961, + "step": 6890 + }, + { + "epoch": 0.4781763930330997, + "grad_norm": 3.8073904514312744, + "learning_rate": 0.0001958250525250303, + "loss": 1.2324, + "step": 6891 + }, + { + "epoch": 0.4782457844701964, + "grad_norm": 3.2994587421417236, + "learning_rate": 0.00019582344584339017, + "loss": 0.9872, + "step": 6892 + }, + { + "epoch": 0.47831517590729306, + "grad_norm": 2.2613914012908936, + "learning_rate": 0.0001958218388592465, + "loss": 0.8737, + "step": 6893 + }, + { + "epoch": 0.4783845673443897, + "grad_norm": 3.0255792140960693, + "learning_rate": 0.00019582023157260444, + "loss": 1.0732, + "step": 6894 + }, + { + "epoch": 0.4784539587814864, + "grad_norm": 2.6986076831817627, + "learning_rate": 0.000195818623983469, + "loss": 1.1092, + "step": 6895 + }, + { + "epoch": 0.47852335021858305, + "grad_norm": 4.138306617736816, + "learning_rate": 0.0001958170160918453, + "loss": 1.0136, + "step": 6896 + }, + { + "epoch": 0.47859274165567967, + "grad_norm": 2.5392110347747803, + "learning_rate": 0.00019581540789773838, + "loss": 0.924, + "step": 6897 + }, + { + "epoch": 0.47866213309277633, + "grad_norm": 3.5115842819213867, + "learning_rate": 0.00019581379940115335, + "loss": 0.8653, + "step": 6898 + }, + { + "epoch": 0.478731524529873, + "grad_norm": 2.8713247776031494, + "learning_rate": 0.00019581219060209524, + "loss": 1.1125, + "step": 6899 + }, + { + "epoch": 0.47880091596696966, + "grad_norm": 2.8933634757995605, + "learning_rate": 0.0001958105815005692, + "loss": 0.9576, + "step": 6900 + }, + { + "epoch": 0.47887030740406633, + "grad_norm": 2.857875347137451, + "learning_rate": 0.00019580897209658024, + "loss": 0.8387, + "step": 6901 + }, + { + "epoch": 0.478939698841163, + "grad_norm": 3.102726459503174, + "learning_rate": 0.0001958073623901335, + "loss": 0.9679, + "step": 6902 + }, + { + "epoch": 0.47900909027825966, + "grad_norm": 3.4459354877471924, + "learning_rate": 0.000195805752381234, + "loss": 0.9309, + "step": 6903 + }, + { + "epoch": 0.47907848171535633, + "grad_norm": 2.3827617168426514, + "learning_rate": 0.00019580414206988689, + "loss": 0.7149, + "step": 6904 + }, + { + "epoch": 0.479147873152453, + "grad_norm": 2.7665798664093018, + "learning_rate": 0.00019580253145609717, + "loss": 0.8882, + "step": 6905 + }, + { + "epoch": 0.47921726458954966, + "grad_norm": 3.095118284225464, + "learning_rate": 0.00019580092053987001, + "loss": 1.0749, + "step": 6906 + }, + { + "epoch": 0.47928665602664633, + "grad_norm": 3.0560994148254395, + "learning_rate": 0.00019579930932121046, + "loss": 1.1168, + "step": 6907 + }, + { + "epoch": 0.479356047463743, + "grad_norm": 3.068967580795288, + "learning_rate": 0.00019579769780012361, + "loss": 1.0007, + "step": 6908 + }, + { + "epoch": 0.47942543890083966, + "grad_norm": 3.0448195934295654, + "learning_rate": 0.0001957960859766145, + "loss": 1.0096, + "step": 6909 + }, + { + "epoch": 0.4794948303379363, + "grad_norm": 2.636733293533325, + "learning_rate": 0.00019579447385068832, + "loss": 0.7973, + "step": 6910 + }, + { + "epoch": 0.47956422177503294, + "grad_norm": 3.6532793045043945, + "learning_rate": 0.00019579286142235006, + "loss": 0.9872, + "step": 6911 + }, + { + "epoch": 0.4796336132121296, + "grad_norm": 2.6867494583129883, + "learning_rate": 0.0001957912486916049, + "loss": 0.8866, + "step": 6912 + }, + { + "epoch": 0.4797030046492263, + "grad_norm": 2.430440902709961, + "learning_rate": 0.00019578963565845781, + "loss": 1.0065, + "step": 6913 + }, + { + "epoch": 0.47977239608632294, + "grad_norm": 2.4450273513793945, + "learning_rate": 0.000195788022322914, + "loss": 0.9248, + "step": 6914 + }, + { + "epoch": 0.4798417875234196, + "grad_norm": 3.1345055103302, + "learning_rate": 0.0001957864086849785, + "loss": 0.9755, + "step": 6915 + }, + { + "epoch": 0.47991117896051627, + "grad_norm": 2.766427755355835, + "learning_rate": 0.0001957847947446564, + "loss": 1.0378, + "step": 6916 + }, + { + "epoch": 0.47998057039761294, + "grad_norm": 2.7838101387023926, + "learning_rate": 0.00019578318050195287, + "loss": 0.888, + "step": 6917 + }, + { + "epoch": 0.4800499618347096, + "grad_norm": 3.2156503200531006, + "learning_rate": 0.0001957815659568729, + "loss": 1.1197, + "step": 6918 + }, + { + "epoch": 0.48011935327180627, + "grad_norm": 2.2609548568725586, + "learning_rate": 0.0001957799511094217, + "loss": 0.6669, + "step": 6919 + }, + { + "epoch": 0.48018874470890294, + "grad_norm": 3.5755796432495117, + "learning_rate": 0.00019577833595960423, + "loss": 1.091, + "step": 6920 + }, + { + "epoch": 0.4802581361459996, + "grad_norm": 2.8378853797912598, + "learning_rate": 0.00019577672050742574, + "loss": 1.0903, + "step": 6921 + }, + { + "epoch": 0.48032752758309627, + "grad_norm": 2.6742119789123535, + "learning_rate": 0.0001957751047528912, + "loss": 0.8918, + "step": 6922 + }, + { + "epoch": 0.4803969190201929, + "grad_norm": 2.408421516418457, + "learning_rate": 0.00019577348869600578, + "loss": 0.8742, + "step": 6923 + }, + { + "epoch": 0.48046631045728955, + "grad_norm": 2.7750725746154785, + "learning_rate": 0.00019577187233677459, + "loss": 1.1876, + "step": 6924 + }, + { + "epoch": 0.4805357018943862, + "grad_norm": 2.874396800994873, + "learning_rate": 0.00019577025567520272, + "loss": 0.9992, + "step": 6925 + }, + { + "epoch": 0.4806050933314829, + "grad_norm": 2.6308462619781494, + "learning_rate": 0.00019576863871129522, + "loss": 0.8314, + "step": 6926 + }, + { + "epoch": 0.48067448476857955, + "grad_norm": 2.463266134262085, + "learning_rate": 0.00019576702144505725, + "loss": 0.9674, + "step": 6927 + }, + { + "epoch": 0.4807438762056762, + "grad_norm": 2.544311046600342, + "learning_rate": 0.00019576540387649391, + "loss": 0.9492, + "step": 6928 + }, + { + "epoch": 0.4808132676427729, + "grad_norm": 2.656942129135132, + "learning_rate": 0.00019576378600561032, + "loss": 0.7375, + "step": 6929 + }, + { + "epoch": 0.48088265907986955, + "grad_norm": 2.651176691055298, + "learning_rate": 0.00019576216783241156, + "loss": 0.9759, + "step": 6930 + }, + { + "epoch": 0.4809520505169662, + "grad_norm": 3.3534128665924072, + "learning_rate": 0.00019576054935690272, + "loss": 1.0331, + "step": 6931 + }, + { + "epoch": 0.4810214419540629, + "grad_norm": 3.7904868125915527, + "learning_rate": 0.00019575893057908897, + "loss": 1.0624, + "step": 6932 + }, + { + "epoch": 0.48109083339115954, + "grad_norm": 2.3221375942230225, + "learning_rate": 0.00019575731149897537, + "loss": 0.8509, + "step": 6933 + }, + { + "epoch": 0.4811602248282562, + "grad_norm": 2.5354301929473877, + "learning_rate": 0.00019575569211656703, + "loss": 0.9022, + "step": 6934 + }, + { + "epoch": 0.4812296162653529, + "grad_norm": 3.135688543319702, + "learning_rate": 0.00019575407243186912, + "loss": 0.8863, + "step": 6935 + }, + { + "epoch": 0.48129900770244954, + "grad_norm": 2.5845627784729004, + "learning_rate": 0.0001957524524448867, + "loss": 0.9689, + "step": 6936 + }, + { + "epoch": 0.48136839913954615, + "grad_norm": 3.1358354091644287, + "learning_rate": 0.00019575083215562486, + "loss": 1.017, + "step": 6937 + }, + { + "epoch": 0.4814377905766428, + "grad_norm": 2.9306936264038086, + "learning_rate": 0.00019574921156408878, + "loss": 0.8984, + "step": 6938 + }, + { + "epoch": 0.4815071820137395, + "grad_norm": 2.6368439197540283, + "learning_rate": 0.00019574759067028357, + "loss": 0.9125, + "step": 6939 + }, + { + "epoch": 0.48157657345083615, + "grad_norm": 3.1517446041107178, + "learning_rate": 0.0001957459694742143, + "loss": 1.1366, + "step": 6940 + }, + { + "epoch": 0.4816459648879328, + "grad_norm": 3.0569639205932617, + "learning_rate": 0.00019574434797588612, + "loss": 1.0766, + "step": 6941 + }, + { + "epoch": 0.4817153563250295, + "grad_norm": 2.731978178024292, + "learning_rate": 0.0001957427261753041, + "loss": 0.9298, + "step": 6942 + }, + { + "epoch": 0.48178474776212615, + "grad_norm": 2.29412579536438, + "learning_rate": 0.00019574110407247347, + "loss": 1.0646, + "step": 6943 + }, + { + "epoch": 0.4818541391992228, + "grad_norm": 2.0916106700897217, + "learning_rate": 0.00019573948166739925, + "loss": 0.8205, + "step": 6944 + }, + { + "epoch": 0.4819235306363195, + "grad_norm": 2.7667131423950195, + "learning_rate": 0.0001957378589600866, + "loss": 1.0457, + "step": 6945 + }, + { + "epoch": 0.48199292207341615, + "grad_norm": 3.099285125732422, + "learning_rate": 0.00019573623595054064, + "loss": 1.0702, + "step": 6946 + }, + { + "epoch": 0.4820623135105128, + "grad_norm": 2.413900375366211, + "learning_rate": 0.00019573461263876647, + "loss": 0.9458, + "step": 6947 + }, + { + "epoch": 0.4821317049476095, + "grad_norm": 2.761505365371704, + "learning_rate": 0.00019573298902476926, + "loss": 0.8055, + "step": 6948 + }, + { + "epoch": 0.48220109638470615, + "grad_norm": 2.6803181171417236, + "learning_rate": 0.0001957313651085541, + "loss": 1.0435, + "step": 6949 + }, + { + "epoch": 0.48227048782180276, + "grad_norm": 3.9982006549835205, + "learning_rate": 0.00019572974089012611, + "loss": 1.0452, + "step": 6950 + }, + { + "epoch": 0.48233987925889943, + "grad_norm": 3.7215240001678467, + "learning_rate": 0.00019572811636949045, + "loss": 1.0567, + "step": 6951 + }, + { + "epoch": 0.4824092706959961, + "grad_norm": 2.8993947505950928, + "learning_rate": 0.00019572649154665225, + "loss": 1.1016, + "step": 6952 + }, + { + "epoch": 0.48247866213309276, + "grad_norm": 2.43839955329895, + "learning_rate": 0.0001957248664216166, + "loss": 0.8096, + "step": 6953 + }, + { + "epoch": 0.48254805357018943, + "grad_norm": 2.4755468368530273, + "learning_rate": 0.00019572324099438867, + "loss": 0.7321, + "step": 6954 + }, + { + "epoch": 0.4826174450072861, + "grad_norm": 4.458401203155518, + "learning_rate": 0.00019572161526497356, + "loss": 1.0293, + "step": 6955 + }, + { + "epoch": 0.48268683644438276, + "grad_norm": 1.9334803819656372, + "learning_rate": 0.00019571998923337643, + "loss": 0.6282, + "step": 6956 + }, + { + "epoch": 0.4827562278814794, + "grad_norm": 2.7271430492401123, + "learning_rate": 0.0001957183628996024, + "loss": 1.0229, + "step": 6957 + }, + { + "epoch": 0.4828256193185761, + "grad_norm": 3.283820390701294, + "learning_rate": 0.0001957167362636566, + "loss": 1.0206, + "step": 6958 + }, + { + "epoch": 0.48289501075567276, + "grad_norm": 2.512852907180786, + "learning_rate": 0.00019571510932554412, + "loss": 1.0316, + "step": 6959 + }, + { + "epoch": 0.4829644021927694, + "grad_norm": 3.754014730453491, + "learning_rate": 0.0001957134820852702, + "loss": 0.945, + "step": 6960 + }, + { + "epoch": 0.4830337936298661, + "grad_norm": 2.880720615386963, + "learning_rate": 0.00019571185454283994, + "loss": 1.248, + "step": 6961 + }, + { + "epoch": 0.48310318506696276, + "grad_norm": 3.4223601818084717, + "learning_rate": 0.0001957102266982584, + "loss": 0.8602, + "step": 6962 + }, + { + "epoch": 0.4831725765040594, + "grad_norm": 4.2105913162231445, + "learning_rate": 0.00019570859855153083, + "loss": 0.9665, + "step": 6963 + }, + { + "epoch": 0.48324196794115604, + "grad_norm": 3.656214952468872, + "learning_rate": 0.0001957069701026623, + "loss": 1.0643, + "step": 6964 + }, + { + "epoch": 0.4833113593782527, + "grad_norm": 1.8988789319992065, + "learning_rate": 0.00019570534135165794, + "loss": 0.689, + "step": 6965 + }, + { + "epoch": 0.48338075081534937, + "grad_norm": 2.5828702449798584, + "learning_rate": 0.00019570371229852296, + "loss": 0.8954, + "step": 6966 + }, + { + "epoch": 0.48345014225244604, + "grad_norm": 2.610020160675049, + "learning_rate": 0.00019570208294326244, + "loss": 0.7723, + "step": 6967 + }, + { + "epoch": 0.4835195336895427, + "grad_norm": 2.8270657062530518, + "learning_rate": 0.00019570045328588158, + "loss": 1.0013, + "step": 6968 + }, + { + "epoch": 0.48358892512663937, + "grad_norm": 3.717313766479492, + "learning_rate": 0.00019569882332638546, + "loss": 1.1102, + "step": 6969 + }, + { + "epoch": 0.48365831656373603, + "grad_norm": 3.1270945072174072, + "learning_rate": 0.00019569719306477927, + "loss": 0.8088, + "step": 6970 + }, + { + "epoch": 0.4837277080008327, + "grad_norm": 2.910475015640259, + "learning_rate": 0.00019569556250106812, + "loss": 0.9405, + "step": 6971 + }, + { + "epoch": 0.48379709943792937, + "grad_norm": 2.8363780975341797, + "learning_rate": 0.00019569393163525723, + "loss": 0.7974, + "step": 6972 + }, + { + "epoch": 0.48386649087502603, + "grad_norm": 2.582235813140869, + "learning_rate": 0.00019569230046735168, + "loss": 0.8359, + "step": 6973 + }, + { + "epoch": 0.4839358823121227, + "grad_norm": 3.1686606407165527, + "learning_rate": 0.00019569066899735664, + "loss": 1.1033, + "step": 6974 + }, + { + "epoch": 0.48400527374921937, + "grad_norm": 2.842482328414917, + "learning_rate": 0.00019568903722527725, + "loss": 0.9379, + "step": 6975 + }, + { + "epoch": 0.48407466518631603, + "grad_norm": 3.651130437850952, + "learning_rate": 0.00019568740515111868, + "loss": 1.1013, + "step": 6976 + }, + { + "epoch": 0.48414405662341264, + "grad_norm": 3.0751891136169434, + "learning_rate": 0.00019568577277488607, + "loss": 0.9494, + "step": 6977 + }, + { + "epoch": 0.4842134480605093, + "grad_norm": 2.584721565246582, + "learning_rate": 0.00019568414009658457, + "loss": 0.7931, + "step": 6978 + }, + { + "epoch": 0.484282839497606, + "grad_norm": 2.759387731552124, + "learning_rate": 0.00019568250711621935, + "loss": 0.9454, + "step": 6979 + }, + { + "epoch": 0.48435223093470264, + "grad_norm": 2.755974531173706, + "learning_rate": 0.00019568087383379558, + "loss": 1.1007, + "step": 6980 + }, + { + "epoch": 0.4844216223717993, + "grad_norm": 3.0538713932037354, + "learning_rate": 0.00019567924024931833, + "loss": 0.7961, + "step": 6981 + }, + { + "epoch": 0.484491013808896, + "grad_norm": 2.851701021194458, + "learning_rate": 0.00019567760636279288, + "loss": 1.0518, + "step": 6982 + }, + { + "epoch": 0.48456040524599264, + "grad_norm": 3.26621150970459, + "learning_rate": 0.00019567597217422431, + "loss": 0.851, + "step": 6983 + }, + { + "epoch": 0.4846297966830893, + "grad_norm": 2.8870584964752197, + "learning_rate": 0.0001956743376836178, + "loss": 1.0541, + "step": 6984 + }, + { + "epoch": 0.484699188120186, + "grad_norm": 2.6600918769836426, + "learning_rate": 0.00019567270289097847, + "loss": 0.7643, + "step": 6985 + }, + { + "epoch": 0.48476857955728264, + "grad_norm": 5.372522830963135, + "learning_rate": 0.00019567106779631158, + "loss": 1.1053, + "step": 6986 + }, + { + "epoch": 0.4848379709943793, + "grad_norm": 2.57777738571167, + "learning_rate": 0.00019566943239962217, + "loss": 0.8454, + "step": 6987 + }, + { + "epoch": 0.484907362431476, + "grad_norm": 2.8925838470458984, + "learning_rate": 0.00019566779670091548, + "loss": 0.9095, + "step": 6988 + }, + { + "epoch": 0.48497675386857264, + "grad_norm": 2.659862995147705, + "learning_rate": 0.00019566616070019666, + "loss": 0.7611, + "step": 6989 + }, + { + "epoch": 0.4850461453056693, + "grad_norm": 2.393817663192749, + "learning_rate": 0.00019566452439747088, + "loss": 0.9024, + "step": 6990 + }, + { + "epoch": 0.4851155367427659, + "grad_norm": 2.456648826599121, + "learning_rate": 0.00019566288779274328, + "loss": 0.8475, + "step": 6991 + }, + { + "epoch": 0.4851849281798626, + "grad_norm": 3.3947505950927734, + "learning_rate": 0.00019566125088601904, + "loss": 1.0745, + "step": 6992 + }, + { + "epoch": 0.48525431961695925, + "grad_norm": 2.4232935905456543, + "learning_rate": 0.00019565961367730336, + "loss": 0.9328, + "step": 6993 + }, + { + "epoch": 0.4853237110540559, + "grad_norm": 1.8145133256912231, + "learning_rate": 0.00019565797616660134, + "loss": 0.7073, + "step": 6994 + }, + { + "epoch": 0.4853931024911526, + "grad_norm": 2.8131966590881348, + "learning_rate": 0.0001956563383539182, + "loss": 0.9859, + "step": 6995 + }, + { + "epoch": 0.48546249392824925, + "grad_norm": 2.825935125350952, + "learning_rate": 0.00019565470023925912, + "loss": 0.8915, + "step": 6996 + }, + { + "epoch": 0.4855318853653459, + "grad_norm": 3.291032552719116, + "learning_rate": 0.00019565306182262922, + "loss": 0.8185, + "step": 6997 + }, + { + "epoch": 0.4856012768024426, + "grad_norm": 3.812203884124756, + "learning_rate": 0.00019565142310403372, + "loss": 1.2465, + "step": 6998 + }, + { + "epoch": 0.48567066823953925, + "grad_norm": 2.5472493171691895, + "learning_rate": 0.00019564978408347775, + "loss": 1.063, + "step": 6999 + }, + { + "epoch": 0.4857400596766359, + "grad_norm": 3.953779458999634, + "learning_rate": 0.00019564814476096654, + "loss": 1.0189, + "step": 7000 + }, + { + "epoch": 0.4858094511137326, + "grad_norm": 2.65140700340271, + "learning_rate": 0.00019564650513650522, + "loss": 0.7319, + "step": 7001 + }, + { + "epoch": 0.48587884255082925, + "grad_norm": 2.863016366958618, + "learning_rate": 0.00019564486521009896, + "loss": 0.8813, + "step": 7002 + }, + { + "epoch": 0.4859482339879259, + "grad_norm": 3.4313876628875732, + "learning_rate": 0.000195643224981753, + "loss": 0.889, + "step": 7003 + }, + { + "epoch": 0.4860176254250225, + "grad_norm": 3.10935115814209, + "learning_rate": 0.00019564158445147244, + "loss": 1.0073, + "step": 7004 + }, + { + "epoch": 0.4860870168621192, + "grad_norm": 1.9408479928970337, + "learning_rate": 0.0001956399436192625, + "loss": 0.7115, + "step": 7005 + }, + { + "epoch": 0.48615640829921586, + "grad_norm": 3.3983848094940186, + "learning_rate": 0.00019563830248512838, + "loss": 0.7181, + "step": 7006 + }, + { + "epoch": 0.4862257997363125, + "grad_norm": 2.9781548976898193, + "learning_rate": 0.00019563666104907524, + "loss": 0.8083, + "step": 7007 + }, + { + "epoch": 0.4862951911734092, + "grad_norm": 2.6195197105407715, + "learning_rate": 0.0001956350193111082, + "loss": 0.9529, + "step": 7008 + }, + { + "epoch": 0.48636458261050586, + "grad_norm": 3.321884870529175, + "learning_rate": 0.00019563337727123255, + "loss": 1.0944, + "step": 7009 + }, + { + "epoch": 0.4864339740476025, + "grad_norm": 2.8949689865112305, + "learning_rate": 0.0001956317349294534, + "loss": 0.8032, + "step": 7010 + }, + { + "epoch": 0.4865033654846992, + "grad_norm": 2.5912740230560303, + "learning_rate": 0.00019563009228577597, + "loss": 0.724, + "step": 7011 + }, + { + "epoch": 0.48657275692179586, + "grad_norm": 2.7235631942749023, + "learning_rate": 0.00019562844934020544, + "loss": 1.0081, + "step": 7012 + }, + { + "epoch": 0.4866421483588925, + "grad_norm": 3.107785940170288, + "learning_rate": 0.00019562680609274698, + "loss": 1.1678, + "step": 7013 + }, + { + "epoch": 0.4867115397959892, + "grad_norm": 2.8598861694335938, + "learning_rate": 0.0001956251625434058, + "loss": 1.0619, + "step": 7014 + }, + { + "epoch": 0.48678093123308586, + "grad_norm": 3.4843904972076416, + "learning_rate": 0.00019562351869218704, + "loss": 1.0022, + "step": 7015 + }, + { + "epoch": 0.4868503226701825, + "grad_norm": 2.4303267002105713, + "learning_rate": 0.00019562187453909594, + "loss": 0.7109, + "step": 7016 + }, + { + "epoch": 0.4869197141072792, + "grad_norm": 2.588097095489502, + "learning_rate": 0.0001956202300841377, + "loss": 1.0454, + "step": 7017 + }, + { + "epoch": 0.4869891055443758, + "grad_norm": 4.078109264373779, + "learning_rate": 0.00019561858532731748, + "loss": 0.9268, + "step": 7018 + }, + { + "epoch": 0.48705849698147247, + "grad_norm": 3.2783849239349365, + "learning_rate": 0.00019561694026864047, + "loss": 0.99, + "step": 7019 + }, + { + "epoch": 0.48712788841856913, + "grad_norm": 3.253783702850342, + "learning_rate": 0.00019561529490811188, + "loss": 0.9677, + "step": 7020 + }, + { + "epoch": 0.4871972798556658, + "grad_norm": 2.5838654041290283, + "learning_rate": 0.0001956136492457369, + "loss": 0.9757, + "step": 7021 + }, + { + "epoch": 0.48726667129276247, + "grad_norm": 3.156885862350464, + "learning_rate": 0.00019561200328152068, + "loss": 0.9865, + "step": 7022 + }, + { + "epoch": 0.48733606272985913, + "grad_norm": 2.1995482444763184, + "learning_rate": 0.00019561035701546852, + "loss": 0.7216, + "step": 7023 + }, + { + "epoch": 0.4874054541669558, + "grad_norm": 3.5948548316955566, + "learning_rate": 0.00019560871044758553, + "loss": 0.9985, + "step": 7024 + }, + { + "epoch": 0.48747484560405246, + "grad_norm": 2.987093210220337, + "learning_rate": 0.00019560706357787694, + "loss": 0.8327, + "step": 7025 + }, + { + "epoch": 0.48754423704114913, + "grad_norm": 3.6155290603637695, + "learning_rate": 0.00019560541640634794, + "loss": 1.1847, + "step": 7026 + }, + { + "epoch": 0.4876136284782458, + "grad_norm": 2.8021669387817383, + "learning_rate": 0.00019560376893300373, + "loss": 1.0477, + "step": 7027 + }, + { + "epoch": 0.48768301991534246, + "grad_norm": 4.657587051391602, + "learning_rate": 0.0001956021211578495, + "loss": 0.9864, + "step": 7028 + }, + { + "epoch": 0.48775241135243913, + "grad_norm": 2.3538269996643066, + "learning_rate": 0.0001956004730808905, + "loss": 0.8427, + "step": 7029 + }, + { + "epoch": 0.4878218027895358, + "grad_norm": 2.5114336013793945, + "learning_rate": 0.00019559882470213187, + "loss": 0.7694, + "step": 7030 + }, + { + "epoch": 0.4878911942266324, + "grad_norm": 3.518824338912964, + "learning_rate": 0.00019559717602157884, + "loss": 0.9262, + "step": 7031 + }, + { + "epoch": 0.4879605856637291, + "grad_norm": 2.314181327819824, + "learning_rate": 0.00019559552703923665, + "loss": 0.8903, + "step": 7032 + }, + { + "epoch": 0.48802997710082574, + "grad_norm": 3.4500977993011475, + "learning_rate": 0.00019559387775511044, + "loss": 1.156, + "step": 7033 + }, + { + "epoch": 0.4880993685379224, + "grad_norm": 2.550422191619873, + "learning_rate": 0.00019559222816920545, + "loss": 0.8298, + "step": 7034 + }, + { + "epoch": 0.4881687599750191, + "grad_norm": 2.8388748168945312, + "learning_rate": 0.00019559057828152693, + "loss": 0.6499, + "step": 7035 + }, + { + "epoch": 0.48823815141211574, + "grad_norm": 3.140693187713623, + "learning_rate": 0.00019558892809207998, + "loss": 1.314, + "step": 7036 + }, + { + "epoch": 0.4883075428492124, + "grad_norm": 6.144458770751953, + "learning_rate": 0.0001955872776008699, + "loss": 0.9702, + "step": 7037 + }, + { + "epoch": 0.4883769342863091, + "grad_norm": 3.2371015548706055, + "learning_rate": 0.00019558562680790188, + "loss": 1.0423, + "step": 7038 + }, + { + "epoch": 0.48844632572340574, + "grad_norm": 3.791917324066162, + "learning_rate": 0.0001955839757131811, + "loss": 0.9513, + "step": 7039 + }, + { + "epoch": 0.4885157171605024, + "grad_norm": 3.4536473751068115, + "learning_rate": 0.00019558232431671283, + "loss": 0.9434, + "step": 7040 + }, + { + "epoch": 0.48858510859759907, + "grad_norm": 2.7301573753356934, + "learning_rate": 0.00019558067261850229, + "loss": 1.0697, + "step": 7041 + }, + { + "epoch": 0.48865450003469574, + "grad_norm": 2.612724781036377, + "learning_rate": 0.00019557902061855459, + "loss": 0.8913, + "step": 7042 + }, + { + "epoch": 0.4887238914717924, + "grad_norm": 2.3271164894104004, + "learning_rate": 0.00019557736831687502, + "loss": 0.9899, + "step": 7043 + }, + { + "epoch": 0.48879328290888907, + "grad_norm": 3.1105854511260986, + "learning_rate": 0.00019557571571346882, + "loss": 0.891, + "step": 7044 + }, + { + "epoch": 0.4888626743459857, + "grad_norm": 2.9613990783691406, + "learning_rate": 0.00019557406280834114, + "loss": 1.2222, + "step": 7045 + }, + { + "epoch": 0.48893206578308235, + "grad_norm": 2.3106374740600586, + "learning_rate": 0.00019557240960149726, + "loss": 0.9265, + "step": 7046 + }, + { + "epoch": 0.489001457220179, + "grad_norm": 2.2418172359466553, + "learning_rate": 0.00019557075609294235, + "loss": 0.9126, + "step": 7047 + }, + { + "epoch": 0.4890708486572757, + "grad_norm": 3.137389898300171, + "learning_rate": 0.00019556910228268166, + "loss": 0.8688, + "step": 7048 + }, + { + "epoch": 0.48914024009437235, + "grad_norm": 3.9641518592834473, + "learning_rate": 0.0001955674481707204, + "loss": 1.1862, + "step": 7049 + }, + { + "epoch": 0.489209631531469, + "grad_norm": 2.8981895446777344, + "learning_rate": 0.0001955657937570638, + "loss": 0.852, + "step": 7050 + }, + { + "epoch": 0.4892790229685657, + "grad_norm": 2.9559431076049805, + "learning_rate": 0.00019556413904171706, + "loss": 1.0295, + "step": 7051 + }, + { + "epoch": 0.48934841440566235, + "grad_norm": 2.4396252632141113, + "learning_rate": 0.0001955624840246854, + "loss": 0.7309, + "step": 7052 + }, + { + "epoch": 0.489417805842759, + "grad_norm": 2.3827764987945557, + "learning_rate": 0.00019556082870597413, + "loss": 0.7921, + "step": 7053 + }, + { + "epoch": 0.4894871972798557, + "grad_norm": 3.754849672317505, + "learning_rate": 0.00019555917308558839, + "loss": 0.9522, + "step": 7054 + }, + { + "epoch": 0.48955658871695235, + "grad_norm": 2.7991292476654053, + "learning_rate": 0.00019555751716353342, + "loss": 0.9027, + "step": 7055 + }, + { + "epoch": 0.489625980154049, + "grad_norm": 2.301046133041382, + "learning_rate": 0.00019555586093981443, + "loss": 0.8198, + "step": 7056 + }, + { + "epoch": 0.4896953715911457, + "grad_norm": 3.7860469818115234, + "learning_rate": 0.0001955542044144367, + "loss": 1.3832, + "step": 7057 + }, + { + "epoch": 0.4897647630282423, + "grad_norm": 3.192967653274536, + "learning_rate": 0.00019555254758740542, + "loss": 0.924, + "step": 7058 + }, + { + "epoch": 0.48983415446533896, + "grad_norm": 2.391248941421509, + "learning_rate": 0.00019555089045872586, + "loss": 0.8222, + "step": 7059 + }, + { + "epoch": 0.4899035459024356, + "grad_norm": 2.4505820274353027, + "learning_rate": 0.00019554923302840318, + "loss": 0.7488, + "step": 7060 + }, + { + "epoch": 0.4899729373395323, + "grad_norm": 2.8351998329162598, + "learning_rate": 0.0001955475752964427, + "loss": 1.1149, + "step": 7061 + }, + { + "epoch": 0.49004232877662895, + "grad_norm": 3.5538313388824463, + "learning_rate": 0.0001955459172628496, + "loss": 1.1516, + "step": 7062 + }, + { + "epoch": 0.4901117202137256, + "grad_norm": 3.293196201324463, + "learning_rate": 0.00019554425892762912, + "loss": 0.9796, + "step": 7063 + }, + { + "epoch": 0.4901811116508223, + "grad_norm": 2.2991294860839844, + "learning_rate": 0.00019554260029078647, + "loss": 0.7583, + "step": 7064 + }, + { + "epoch": 0.49025050308791895, + "grad_norm": 3.730822801589966, + "learning_rate": 0.00019554094135232694, + "loss": 1.0564, + "step": 7065 + }, + { + "epoch": 0.4903198945250156, + "grad_norm": 5.394999027252197, + "learning_rate": 0.00019553928211225576, + "loss": 1.1788, + "step": 7066 + }, + { + "epoch": 0.4903892859621123, + "grad_norm": 2.3195853233337402, + "learning_rate": 0.00019553762257057815, + "loss": 0.7543, + "step": 7067 + }, + { + "epoch": 0.49045867739920895, + "grad_norm": 4.032540798187256, + "learning_rate": 0.00019553596272729932, + "loss": 1.2595, + "step": 7068 + }, + { + "epoch": 0.4905280688363056, + "grad_norm": 2.7108819484710693, + "learning_rate": 0.00019553430258242456, + "loss": 0.8497, + "step": 7069 + }, + { + "epoch": 0.4905974602734023, + "grad_norm": 2.453444004058838, + "learning_rate": 0.00019553264213595907, + "loss": 0.8651, + "step": 7070 + }, + { + "epoch": 0.4906668517104989, + "grad_norm": 2.5534231662750244, + "learning_rate": 0.00019553098138790816, + "loss": 0.8936, + "step": 7071 + }, + { + "epoch": 0.49073624314759556, + "grad_norm": 2.927046537399292, + "learning_rate": 0.000195529320338277, + "loss": 1.0851, + "step": 7072 + }, + { + "epoch": 0.49080563458469223, + "grad_norm": 2.793799877166748, + "learning_rate": 0.00019552765898707083, + "loss": 0.8991, + "step": 7073 + }, + { + "epoch": 0.4908750260217889, + "grad_norm": 2.6373131275177, + "learning_rate": 0.00019552599733429498, + "loss": 0.9346, + "step": 7074 + }, + { + "epoch": 0.49094441745888556, + "grad_norm": 3.1869406700134277, + "learning_rate": 0.00019552433537995458, + "loss": 1.1034, + "step": 7075 + }, + { + "epoch": 0.49101380889598223, + "grad_norm": 3.2193634510040283, + "learning_rate": 0.00019552267312405498, + "loss": 1.0441, + "step": 7076 + }, + { + "epoch": 0.4910832003330789, + "grad_norm": 3.2911174297332764, + "learning_rate": 0.00019552101056660135, + "loss": 1.1409, + "step": 7077 + }, + { + "epoch": 0.49115259177017556, + "grad_norm": 2.9853310585021973, + "learning_rate": 0.000195519347707599, + "loss": 0.752, + "step": 7078 + }, + { + "epoch": 0.49122198320727223, + "grad_norm": 2.5745043754577637, + "learning_rate": 0.00019551768454705314, + "loss": 0.7858, + "step": 7079 + }, + { + "epoch": 0.4912913746443689, + "grad_norm": 4.484233856201172, + "learning_rate": 0.00019551602108496905, + "loss": 1.4302, + "step": 7080 + }, + { + "epoch": 0.49136076608146556, + "grad_norm": 2.5477588176727295, + "learning_rate": 0.00019551435732135194, + "loss": 0.798, + "step": 7081 + }, + { + "epoch": 0.4914301575185622, + "grad_norm": 3.2483510971069336, + "learning_rate": 0.00019551269325620706, + "loss": 0.8456, + "step": 7082 + }, + { + "epoch": 0.4914995489556589, + "grad_norm": 3.3203516006469727, + "learning_rate": 0.00019551102888953973, + "loss": 0.963, + "step": 7083 + }, + { + "epoch": 0.49156894039275556, + "grad_norm": 6.166804313659668, + "learning_rate": 0.00019550936422135518, + "loss": 1.1086, + "step": 7084 + }, + { + "epoch": 0.49163833182985217, + "grad_norm": 2.730004072189331, + "learning_rate": 0.00019550769925165862, + "loss": 0.9612, + "step": 7085 + }, + { + "epoch": 0.49170772326694884, + "grad_norm": 3.68326473236084, + "learning_rate": 0.00019550603398045533, + "loss": 1.0639, + "step": 7086 + }, + { + "epoch": 0.4917771147040455, + "grad_norm": 4.263821125030518, + "learning_rate": 0.00019550436840775056, + "loss": 1.1335, + "step": 7087 + }, + { + "epoch": 0.49184650614114217, + "grad_norm": 4.411851406097412, + "learning_rate": 0.00019550270253354963, + "loss": 1.1823, + "step": 7088 + }, + { + "epoch": 0.49191589757823884, + "grad_norm": 3.8117568492889404, + "learning_rate": 0.0001955010363578577, + "loss": 0.8962, + "step": 7089 + }, + { + "epoch": 0.4919852890153355, + "grad_norm": 2.8599658012390137, + "learning_rate": 0.0001954993698806801, + "loss": 0.7367, + "step": 7090 + }, + { + "epoch": 0.49205468045243217, + "grad_norm": 5.3042731285095215, + "learning_rate": 0.00019549770310202207, + "loss": 1.2579, + "step": 7091 + }, + { + "epoch": 0.49212407188952884, + "grad_norm": 6.988046169281006, + "learning_rate": 0.00019549603602188887, + "loss": 1.3257, + "step": 7092 + }, + { + "epoch": 0.4921934633266255, + "grad_norm": 2.976957082748413, + "learning_rate": 0.00019549436864028575, + "loss": 0.9883, + "step": 7093 + }, + { + "epoch": 0.49226285476372217, + "grad_norm": 2.6126692295074463, + "learning_rate": 0.000195492700957218, + "loss": 0.8188, + "step": 7094 + }, + { + "epoch": 0.49233224620081883, + "grad_norm": 2.6299667358398438, + "learning_rate": 0.00019549103297269085, + "loss": 0.9679, + "step": 7095 + }, + { + "epoch": 0.4924016376379155, + "grad_norm": 3.5613977909088135, + "learning_rate": 0.00019548936468670962, + "loss": 1.2528, + "step": 7096 + }, + { + "epoch": 0.49247102907501217, + "grad_norm": 3.3054230213165283, + "learning_rate": 0.00019548769609927953, + "loss": 0.9087, + "step": 7097 + }, + { + "epoch": 0.4925404205121088, + "grad_norm": 2.5325281620025635, + "learning_rate": 0.00019548602721040586, + "loss": 0.8393, + "step": 7098 + }, + { + "epoch": 0.49260981194920545, + "grad_norm": 4.3148512840271, + "learning_rate": 0.00019548435802009388, + "loss": 0.95, + "step": 7099 + }, + { + "epoch": 0.4926792033863021, + "grad_norm": 3.941030263900757, + "learning_rate": 0.00019548268852834886, + "loss": 1.339, + "step": 7100 + }, + { + "epoch": 0.4927485948233988, + "grad_norm": 2.953124761581421, + "learning_rate": 0.0001954810187351761, + "loss": 0.981, + "step": 7101 + }, + { + "epoch": 0.49281798626049544, + "grad_norm": 2.4371681213378906, + "learning_rate": 0.00019547934864058082, + "loss": 1.275, + "step": 7102 + }, + { + "epoch": 0.4928873776975921, + "grad_norm": 3.147174596786499, + "learning_rate": 0.00019547767824456828, + "loss": 1.3273, + "step": 7103 + }, + { + "epoch": 0.4929567691346888, + "grad_norm": 4.038508415222168, + "learning_rate": 0.00019547600754714383, + "loss": 0.7654, + "step": 7104 + }, + { + "epoch": 0.49302616057178544, + "grad_norm": 2.6389071941375732, + "learning_rate": 0.0001954743365483127, + "loss": 1.1502, + "step": 7105 + }, + { + "epoch": 0.4930955520088821, + "grad_norm": 2.354085922241211, + "learning_rate": 0.00019547266524808015, + "loss": 0.7158, + "step": 7106 + }, + { + "epoch": 0.4931649434459788, + "grad_norm": 2.647859811782837, + "learning_rate": 0.00019547099364645147, + "loss": 0.8354, + "step": 7107 + }, + { + "epoch": 0.49323433488307544, + "grad_norm": 2.2304933071136475, + "learning_rate": 0.00019546932174343195, + "loss": 0.9271, + "step": 7108 + }, + { + "epoch": 0.4933037263201721, + "grad_norm": 2.388402223587036, + "learning_rate": 0.00019546764953902684, + "loss": 0.8114, + "step": 7109 + }, + { + "epoch": 0.4933731177572688, + "grad_norm": 2.387859582901001, + "learning_rate": 0.00019546597703324147, + "loss": 0.9511, + "step": 7110 + }, + { + "epoch": 0.49344250919436544, + "grad_norm": 3.356590747833252, + "learning_rate": 0.00019546430422608105, + "loss": 0.9684, + "step": 7111 + }, + { + "epoch": 0.49351190063146205, + "grad_norm": 2.439990282058716, + "learning_rate": 0.0001954626311175509, + "loss": 0.6635, + "step": 7112 + }, + { + "epoch": 0.4935812920685587, + "grad_norm": 3.31601619720459, + "learning_rate": 0.00019546095770765635, + "loss": 1.2395, + "step": 7113 + }, + { + "epoch": 0.4936506835056554, + "grad_norm": 3.2881245613098145, + "learning_rate": 0.00019545928399640258, + "loss": 0.6437, + "step": 7114 + }, + { + "epoch": 0.49372007494275205, + "grad_norm": 2.888715982437134, + "learning_rate": 0.00019545760998379496, + "loss": 1.1236, + "step": 7115 + }, + { + "epoch": 0.4937894663798487, + "grad_norm": 3.177738904953003, + "learning_rate": 0.0001954559356698387, + "loss": 1.035, + "step": 7116 + }, + { + "epoch": 0.4938588578169454, + "grad_norm": 2.662057876586914, + "learning_rate": 0.00019545426105453917, + "loss": 1.001, + "step": 7117 + }, + { + "epoch": 0.49392824925404205, + "grad_norm": 3.073631525039673, + "learning_rate": 0.0001954525861379016, + "loss": 1.0497, + "step": 7118 + }, + { + "epoch": 0.4939976406911387, + "grad_norm": 3.4891157150268555, + "learning_rate": 0.00019545091091993126, + "loss": 1.0093, + "step": 7119 + }, + { + "epoch": 0.4940670321282354, + "grad_norm": 3.122434616088867, + "learning_rate": 0.00019544923540063352, + "loss": 0.8605, + "step": 7120 + }, + { + "epoch": 0.49413642356533205, + "grad_norm": 4.829577445983887, + "learning_rate": 0.0001954475595800136, + "loss": 1.0921, + "step": 7121 + }, + { + "epoch": 0.4942058150024287, + "grad_norm": 2.4150009155273438, + "learning_rate": 0.00019544588345807678, + "loss": 0.8078, + "step": 7122 + }, + { + "epoch": 0.4942752064395254, + "grad_norm": 2.631464958190918, + "learning_rate": 0.0001954442070348284, + "loss": 1.144, + "step": 7123 + }, + { + "epoch": 0.49434459787662205, + "grad_norm": 3.857008695602417, + "learning_rate": 0.00019544253031027374, + "loss": 0.952, + "step": 7124 + }, + { + "epoch": 0.49441398931371866, + "grad_norm": 3.263554811477661, + "learning_rate": 0.00019544085328441807, + "loss": 1.1103, + "step": 7125 + }, + { + "epoch": 0.4944833807508153, + "grad_norm": 2.981323003768921, + "learning_rate": 0.00019543917595726668, + "loss": 0.841, + "step": 7126 + }, + { + "epoch": 0.494552772187912, + "grad_norm": 2.87442684173584, + "learning_rate": 0.00019543749832882492, + "loss": 0.794, + "step": 7127 + }, + { + "epoch": 0.49462216362500866, + "grad_norm": 3.585728406906128, + "learning_rate": 0.00019543582039909804, + "loss": 1.1833, + "step": 7128 + }, + { + "epoch": 0.4946915550621053, + "grad_norm": 2.6637091636657715, + "learning_rate": 0.00019543414216809135, + "loss": 0.9088, + "step": 7129 + }, + { + "epoch": 0.494760946499202, + "grad_norm": 3.0303173065185547, + "learning_rate": 0.00019543246363581013, + "loss": 1.0363, + "step": 7130 + }, + { + "epoch": 0.49483033793629866, + "grad_norm": 3.5217950344085693, + "learning_rate": 0.0001954307848022597, + "loss": 0.9831, + "step": 7131 + }, + { + "epoch": 0.4948997293733953, + "grad_norm": 3.0159780979156494, + "learning_rate": 0.00019542910566744536, + "loss": 1.067, + "step": 7132 + }, + { + "epoch": 0.494969120810492, + "grad_norm": 4.126817226409912, + "learning_rate": 0.00019542742623137243, + "loss": 1.0477, + "step": 7133 + }, + { + "epoch": 0.49503851224758866, + "grad_norm": 3.157489776611328, + "learning_rate": 0.00019542574649404615, + "loss": 1.0804, + "step": 7134 + }, + { + "epoch": 0.4951079036846853, + "grad_norm": 2.935662031173706, + "learning_rate": 0.00019542406645547188, + "loss": 1.0146, + "step": 7135 + }, + { + "epoch": 0.495177295121782, + "grad_norm": 3.281268358230591, + "learning_rate": 0.0001954223861156549, + "loss": 1.1534, + "step": 7136 + }, + { + "epoch": 0.49524668655887866, + "grad_norm": 2.302347421646118, + "learning_rate": 0.0001954207054746005, + "loss": 0.7572, + "step": 7137 + }, + { + "epoch": 0.4953160779959753, + "grad_norm": 5.2552666664123535, + "learning_rate": 0.000195419024532314, + "loss": 0.7941, + "step": 7138 + }, + { + "epoch": 0.49538546943307193, + "grad_norm": 3.921196699142456, + "learning_rate": 0.00019541734328880076, + "loss": 0.8698, + "step": 7139 + }, + { + "epoch": 0.4954548608701686, + "grad_norm": 2.6738054752349854, + "learning_rate": 0.000195415661744066, + "loss": 0.8027, + "step": 7140 + }, + { + "epoch": 0.49552425230726527, + "grad_norm": 3.0345659255981445, + "learning_rate": 0.00019541397989811506, + "loss": 0.8044, + "step": 7141 + }, + { + "epoch": 0.49559364374436193, + "grad_norm": 3.8823447227478027, + "learning_rate": 0.0001954122977509533, + "loss": 0.8008, + "step": 7142 + }, + { + "epoch": 0.4956630351814586, + "grad_norm": 2.891331672668457, + "learning_rate": 0.00019541061530258593, + "loss": 0.9405, + "step": 7143 + }, + { + "epoch": 0.49573242661855527, + "grad_norm": 2.614657163619995, + "learning_rate": 0.00019540893255301836, + "loss": 0.9338, + "step": 7144 + }, + { + "epoch": 0.49580181805565193, + "grad_norm": 2.900189161300659, + "learning_rate": 0.0001954072495022558, + "loss": 0.8788, + "step": 7145 + }, + { + "epoch": 0.4958712094927486, + "grad_norm": 2.2729711532592773, + "learning_rate": 0.00019540556615030367, + "loss": 0.9681, + "step": 7146 + }, + { + "epoch": 0.49594060092984527, + "grad_norm": 3.025552272796631, + "learning_rate": 0.00019540388249716725, + "loss": 0.9528, + "step": 7147 + }, + { + "epoch": 0.49600999236694193, + "grad_norm": 3.1977245807647705, + "learning_rate": 0.00019540219854285182, + "loss": 0.9899, + "step": 7148 + }, + { + "epoch": 0.4960793838040386, + "grad_norm": 2.714857339859009, + "learning_rate": 0.00019540051428736273, + "loss": 0.9489, + "step": 7149 + }, + { + "epoch": 0.49614877524113526, + "grad_norm": 3.453991413116455, + "learning_rate": 0.00019539882973070526, + "loss": 0.9851, + "step": 7150 + }, + { + "epoch": 0.49621816667823193, + "grad_norm": 2.7595055103302, + "learning_rate": 0.00019539714487288475, + "loss": 1.222, + "step": 7151 + }, + { + "epoch": 0.49628755811532854, + "grad_norm": 3.2063357830047607, + "learning_rate": 0.00019539545971390657, + "loss": 0.7629, + "step": 7152 + }, + { + "epoch": 0.4963569495524252, + "grad_norm": 2.483307361602783, + "learning_rate": 0.00019539377425377595, + "loss": 0.9738, + "step": 7153 + }, + { + "epoch": 0.4964263409895219, + "grad_norm": 2.7116005420684814, + "learning_rate": 0.00019539208849249826, + "loss": 1.2277, + "step": 7154 + }, + { + "epoch": 0.49649573242661854, + "grad_norm": 6.391916751861572, + "learning_rate": 0.00019539040243007883, + "loss": 1.2755, + "step": 7155 + }, + { + "epoch": 0.4965651238637152, + "grad_norm": 2.9140684604644775, + "learning_rate": 0.00019538871606652292, + "loss": 1.0151, + "step": 7156 + }, + { + "epoch": 0.4966345153008119, + "grad_norm": 3.5272164344787598, + "learning_rate": 0.00019538702940183596, + "loss": 1.1158, + "step": 7157 + }, + { + "epoch": 0.49670390673790854, + "grad_norm": 2.7512669563293457, + "learning_rate": 0.0001953853424360232, + "loss": 1.0303, + "step": 7158 + }, + { + "epoch": 0.4967732981750052, + "grad_norm": 3.47739839553833, + "learning_rate": 0.00019538365516908997, + "loss": 0.995, + "step": 7159 + }, + { + "epoch": 0.4968426896121019, + "grad_norm": 3.0044872760772705, + "learning_rate": 0.00019538196760104159, + "loss": 0.9059, + "step": 7160 + }, + { + "epoch": 0.49691208104919854, + "grad_norm": 3.1664059162139893, + "learning_rate": 0.00019538027973188343, + "loss": 0.7391, + "step": 7161 + }, + { + "epoch": 0.4969814724862952, + "grad_norm": 2.033332347869873, + "learning_rate": 0.00019537859156162077, + "loss": 0.7579, + "step": 7162 + }, + { + "epoch": 0.4970508639233919, + "grad_norm": 2.886054039001465, + "learning_rate": 0.00019537690309025897, + "loss": 1.1612, + "step": 7163 + }, + { + "epoch": 0.49712025536048854, + "grad_norm": 3.0279829502105713, + "learning_rate": 0.00019537521431780338, + "loss": 1.1421, + "step": 7164 + }, + { + "epoch": 0.4971896467975852, + "grad_norm": 2.8869099617004395, + "learning_rate": 0.00019537352524425926, + "loss": 1.0305, + "step": 7165 + }, + { + "epoch": 0.4972590382346818, + "grad_norm": 2.8566806316375732, + "learning_rate": 0.00019537183586963204, + "loss": 1.1228, + "step": 7166 + }, + { + "epoch": 0.4973284296717785, + "grad_norm": 3.684512138366699, + "learning_rate": 0.00019537014619392697, + "loss": 1.1312, + "step": 7167 + }, + { + "epoch": 0.49739782110887515, + "grad_norm": 3.3884458541870117, + "learning_rate": 0.0001953684562171494, + "loss": 1.0121, + "step": 7168 + }, + { + "epoch": 0.4974672125459718, + "grad_norm": 2.706016778945923, + "learning_rate": 0.0001953667659393047, + "loss": 1.0696, + "step": 7169 + }, + { + "epoch": 0.4975366039830685, + "grad_norm": 2.5728471279144287, + "learning_rate": 0.00019536507536039818, + "loss": 1.0395, + "step": 7170 + }, + { + "epoch": 0.49760599542016515, + "grad_norm": 3.848662853240967, + "learning_rate": 0.0001953633844804352, + "loss": 1.0279, + "step": 7171 + }, + { + "epoch": 0.4976753868572618, + "grad_norm": 2.84244441986084, + "learning_rate": 0.00019536169329942104, + "loss": 0.9354, + "step": 7172 + }, + { + "epoch": 0.4977447782943585, + "grad_norm": 2.732555627822876, + "learning_rate": 0.0001953600018173611, + "loss": 1.105, + "step": 7173 + }, + { + "epoch": 0.49781416973145515, + "grad_norm": 2.934938907623291, + "learning_rate": 0.00019535831003426068, + "loss": 0.8142, + "step": 7174 + }, + { + "epoch": 0.4978835611685518, + "grad_norm": 2.27536678314209, + "learning_rate": 0.00019535661795012514, + "loss": 0.9814, + "step": 7175 + }, + { + "epoch": 0.4979529526056485, + "grad_norm": 2.6613779067993164, + "learning_rate": 0.0001953549255649598, + "loss": 0.794, + "step": 7176 + }, + { + "epoch": 0.49802234404274515, + "grad_norm": 3.0690524578094482, + "learning_rate": 0.00019535323287877009, + "loss": 1.0456, + "step": 7177 + }, + { + "epoch": 0.4980917354798418, + "grad_norm": 3.5392019748687744, + "learning_rate": 0.00019535153989156125, + "loss": 0.8903, + "step": 7178 + }, + { + "epoch": 0.4981611269169384, + "grad_norm": 2.4456827640533447, + "learning_rate": 0.0001953498466033387, + "loss": 0.911, + "step": 7179 + }, + { + "epoch": 0.4982305183540351, + "grad_norm": 3.492659568786621, + "learning_rate": 0.0001953481530141077, + "loss": 0.8316, + "step": 7180 + }, + { + "epoch": 0.49829990979113176, + "grad_norm": 3.342062473297119, + "learning_rate": 0.00019534645912387363, + "loss": 1.0214, + "step": 7181 + }, + { + "epoch": 0.4983693012282284, + "grad_norm": 3.0170156955718994, + "learning_rate": 0.00019534476493264188, + "loss": 1.154, + "step": 7182 + }, + { + "epoch": 0.4984386926653251, + "grad_norm": 2.6846072673797607, + "learning_rate": 0.00019534307044041777, + "loss": 0.6315, + "step": 7183 + }, + { + "epoch": 0.49850808410242176, + "grad_norm": 4.144607067108154, + "learning_rate": 0.00019534137564720662, + "loss": 1.2205, + "step": 7184 + }, + { + "epoch": 0.4985774755395184, + "grad_norm": 2.806594133377075, + "learning_rate": 0.00019533968055301384, + "loss": 0.8082, + "step": 7185 + }, + { + "epoch": 0.4986468669766151, + "grad_norm": 2.14652419090271, + "learning_rate": 0.00019533798515784475, + "loss": 0.8631, + "step": 7186 + }, + { + "epoch": 0.49871625841371175, + "grad_norm": 2.799020290374756, + "learning_rate": 0.00019533628946170472, + "loss": 0.6315, + "step": 7187 + }, + { + "epoch": 0.4987856498508084, + "grad_norm": 4.151699542999268, + "learning_rate": 0.00019533459346459905, + "loss": 1.0379, + "step": 7188 + }, + { + "epoch": 0.4988550412879051, + "grad_norm": 3.0427544116973877, + "learning_rate": 0.00019533289716653314, + "loss": 0.9615, + "step": 7189 + }, + { + "epoch": 0.49892443272500175, + "grad_norm": 2.780872106552124, + "learning_rate": 0.0001953312005675123, + "loss": 1.0355, + "step": 7190 + }, + { + "epoch": 0.4989938241620984, + "grad_norm": 4.737658977508545, + "learning_rate": 0.00019532950366754197, + "loss": 1.2183, + "step": 7191 + }, + { + "epoch": 0.4990632155991951, + "grad_norm": 1.9345015287399292, + "learning_rate": 0.00019532780646662741, + "loss": 0.7792, + "step": 7192 + }, + { + "epoch": 0.4991326070362917, + "grad_norm": 3.1484854221343994, + "learning_rate": 0.00019532610896477409, + "loss": 0.9416, + "step": 7193 + }, + { + "epoch": 0.49920199847338836, + "grad_norm": 3.2233924865722656, + "learning_rate": 0.00019532441116198727, + "loss": 0.8392, + "step": 7194 + }, + { + "epoch": 0.49927138991048503, + "grad_norm": 2.802306652069092, + "learning_rate": 0.00019532271305827234, + "loss": 1.1374, + "step": 7195 + }, + { + "epoch": 0.4993407813475817, + "grad_norm": 3.205103635787964, + "learning_rate": 0.00019532101465363465, + "loss": 0.9536, + "step": 7196 + }, + { + "epoch": 0.49941017278467836, + "grad_norm": 3.0149178504943848, + "learning_rate": 0.0001953193159480796, + "loss": 1.1227, + "step": 7197 + }, + { + "epoch": 0.49947956422177503, + "grad_norm": 3.4067182540893555, + "learning_rate": 0.00019531761694161252, + "loss": 0.9857, + "step": 7198 + }, + { + "epoch": 0.4995489556588717, + "grad_norm": 2.6592154502868652, + "learning_rate": 0.00019531591763423877, + "loss": 0.9694, + "step": 7199 + }, + { + "epoch": 0.49961834709596836, + "grad_norm": 2.788121223449707, + "learning_rate": 0.00019531421802596377, + "loss": 1.0095, + "step": 7200 + }, + { + "epoch": 0.49968773853306503, + "grad_norm": 3.516937017440796, + "learning_rate": 0.0001953125181167928, + "loss": 1.0779, + "step": 7201 + }, + { + "epoch": 0.4997571299701617, + "grad_norm": 3.902592658996582, + "learning_rate": 0.0001953108179067313, + "loss": 1.0994, + "step": 7202 + }, + { + "epoch": 0.49982652140725836, + "grad_norm": 2.9556267261505127, + "learning_rate": 0.00019530911739578456, + "loss": 0.8519, + "step": 7203 + }, + { + "epoch": 0.49989591284435503, + "grad_norm": 2.6407151222229004, + "learning_rate": 0.00019530741658395804, + "loss": 1.0602, + "step": 7204 + }, + { + "epoch": 0.4999653042814517, + "grad_norm": 3.141526937484741, + "learning_rate": 0.00019530571547125703, + "loss": 0.9842, + "step": 7205 + }, + { + "epoch": 0.5000346957185483, + "grad_norm": 3.4398272037506104, + "learning_rate": 0.00019530401405768694, + "loss": 1.1453, + "step": 7206 + }, + { + "epoch": 0.500104087155645, + "grad_norm": 2.5798256397247314, + "learning_rate": 0.00019530231234325316, + "loss": 0.8136, + "step": 7207 + }, + { + "epoch": 0.5001734785927416, + "grad_norm": 2.940272808074951, + "learning_rate": 0.000195300610327961, + "loss": 1.0177, + "step": 7208 + }, + { + "epoch": 0.5002428700298384, + "grad_norm": 4.333984851837158, + "learning_rate": 0.00019529890801181588, + "loss": 1.1513, + "step": 7209 + }, + { + "epoch": 0.500312261466935, + "grad_norm": 3.470541477203369, + "learning_rate": 0.0001952972053948232, + "loss": 1.1257, + "step": 7210 + }, + { + "epoch": 0.5003816529040317, + "grad_norm": 2.7198572158813477, + "learning_rate": 0.0001952955024769883, + "loss": 0.6737, + "step": 7211 + }, + { + "epoch": 0.5004510443411283, + "grad_norm": 2.3151063919067383, + "learning_rate": 0.0001952937992583165, + "loss": 0.7689, + "step": 7212 + }, + { + "epoch": 0.5005204357782249, + "grad_norm": 3.223471164703369, + "learning_rate": 0.00019529209573881327, + "loss": 1.0196, + "step": 7213 + }, + { + "epoch": 0.5005898272153216, + "grad_norm": 2.606016159057617, + "learning_rate": 0.00019529039191848395, + "loss": 0.8591, + "step": 7214 + }, + { + "epoch": 0.5006592186524182, + "grad_norm": 2.873701333999634, + "learning_rate": 0.0001952886877973339, + "loss": 0.9751, + "step": 7215 + }, + { + "epoch": 0.500728610089515, + "grad_norm": 2.7618260383605957, + "learning_rate": 0.00019528698337536852, + "loss": 0.9245, + "step": 7216 + }, + { + "epoch": 0.5007980015266116, + "grad_norm": 3.1391797065734863, + "learning_rate": 0.0001952852786525932, + "loss": 1.2238, + "step": 7217 + }, + { + "epoch": 0.5008673929637083, + "grad_norm": 2.3734664916992188, + "learning_rate": 0.0001952835736290133, + "loss": 0.9021, + "step": 7218 + }, + { + "epoch": 0.5009367844008049, + "grad_norm": 2.7128918170928955, + "learning_rate": 0.00019528186830463425, + "loss": 1.2601, + "step": 7219 + }, + { + "epoch": 0.5010061758379016, + "grad_norm": 4.399260520935059, + "learning_rate": 0.00019528016267946133, + "loss": 1.075, + "step": 7220 + }, + { + "epoch": 0.5010755672749982, + "grad_norm": 3.4689815044403076, + "learning_rate": 0.00019527845675350002, + "loss": 1.2965, + "step": 7221 + }, + { + "epoch": 0.501144958712095, + "grad_norm": 3.984999656677246, + "learning_rate": 0.00019527675052675569, + "loss": 0.9823, + "step": 7222 + }, + { + "epoch": 0.5012143501491916, + "grad_norm": 3.693148612976074, + "learning_rate": 0.00019527504399923374, + "loss": 0.9989, + "step": 7223 + }, + { + "epoch": 0.5012837415862883, + "grad_norm": 3.0976736545562744, + "learning_rate": 0.00019527333717093948, + "loss": 0.8699, + "step": 7224 + }, + { + "epoch": 0.5013531330233849, + "grad_norm": 2.548882484436035, + "learning_rate": 0.00019527163004187835, + "loss": 0.8558, + "step": 7225 + }, + { + "epoch": 0.5014225244604815, + "grad_norm": 3.4797096252441406, + "learning_rate": 0.00019526992261205577, + "loss": 0.995, + "step": 7226 + }, + { + "epoch": 0.5014919158975782, + "grad_norm": 4.229465484619141, + "learning_rate": 0.00019526821488147707, + "loss": 0.9215, + "step": 7227 + }, + { + "epoch": 0.5015613073346749, + "grad_norm": 2.988804578781128, + "learning_rate": 0.00019526650685014766, + "loss": 0.9983, + "step": 7228 + }, + { + "epoch": 0.5016306987717716, + "grad_norm": 2.1900882720947266, + "learning_rate": 0.000195264798518073, + "loss": 0.7324, + "step": 7229 + }, + { + "epoch": 0.5017000902088682, + "grad_norm": 3.7597079277038574, + "learning_rate": 0.00019526308988525835, + "loss": 0.9085, + "step": 7230 + }, + { + "epoch": 0.5017694816459649, + "grad_norm": 2.692713975906372, + "learning_rate": 0.00019526138095170924, + "loss": 0.7381, + "step": 7231 + }, + { + "epoch": 0.5018388730830615, + "grad_norm": 3.3050661087036133, + "learning_rate": 0.00019525967171743095, + "loss": 0.8429, + "step": 7232 + }, + { + "epoch": 0.5019082645201582, + "grad_norm": 2.334510087966919, + "learning_rate": 0.00019525796218242895, + "loss": 0.8414, + "step": 7233 + }, + { + "epoch": 0.5019776559572549, + "grad_norm": 3.536895990371704, + "learning_rate": 0.00019525625234670862, + "loss": 0.9116, + "step": 7234 + }, + { + "epoch": 0.5020470473943516, + "grad_norm": 2.550712823867798, + "learning_rate": 0.00019525454221027536, + "loss": 0.8172, + "step": 7235 + }, + { + "epoch": 0.5021164388314482, + "grad_norm": 3.24277925491333, + "learning_rate": 0.00019525283177313453, + "loss": 0.9802, + "step": 7236 + }, + { + "epoch": 0.5021858302685449, + "grad_norm": 3.39526629447937, + "learning_rate": 0.0001952511210352916, + "loss": 1.1926, + "step": 7237 + }, + { + "epoch": 0.5022552217056415, + "grad_norm": 2.3326525688171387, + "learning_rate": 0.00019524940999675192, + "loss": 0.8928, + "step": 7238 + }, + { + "epoch": 0.5023246131427382, + "grad_norm": 2.5449302196502686, + "learning_rate": 0.0001952476986575209, + "loss": 0.839, + "step": 7239 + }, + { + "epoch": 0.5023940045798349, + "grad_norm": 2.7025582790374756, + "learning_rate": 0.00019524598701760394, + "loss": 0.9464, + "step": 7240 + }, + { + "epoch": 0.5024633960169315, + "grad_norm": 3.3298745155334473, + "learning_rate": 0.00019524427507700646, + "loss": 1.1044, + "step": 7241 + }, + { + "epoch": 0.5025327874540282, + "grad_norm": 3.165311813354492, + "learning_rate": 0.00019524256283573384, + "loss": 1.0433, + "step": 7242 + }, + { + "epoch": 0.5026021788911248, + "grad_norm": 4.139050483703613, + "learning_rate": 0.00019524085029379152, + "loss": 0.9288, + "step": 7243 + }, + { + "epoch": 0.5026715703282215, + "grad_norm": 3.439938545227051, + "learning_rate": 0.00019523913745118487, + "loss": 1.1827, + "step": 7244 + }, + { + "epoch": 0.5027409617653181, + "grad_norm": 2.9788992404937744, + "learning_rate": 0.00019523742430791933, + "loss": 0.9368, + "step": 7245 + }, + { + "epoch": 0.5028103532024149, + "grad_norm": 3.239457607269287, + "learning_rate": 0.00019523571086400024, + "loss": 1.117, + "step": 7246 + }, + { + "epoch": 0.5028797446395115, + "grad_norm": 2.9475743770599365, + "learning_rate": 0.00019523399711943312, + "loss": 0.9702, + "step": 7247 + }, + { + "epoch": 0.5029491360766082, + "grad_norm": 3.128176212310791, + "learning_rate": 0.00019523228307422329, + "loss": 1.1163, + "step": 7248 + }, + { + "epoch": 0.5030185275137048, + "grad_norm": 2.1770288944244385, + "learning_rate": 0.00019523056872837623, + "loss": 0.8427, + "step": 7249 + }, + { + "epoch": 0.5030879189508015, + "grad_norm": 2.6079564094543457, + "learning_rate": 0.00019522885408189726, + "loss": 1.2038, + "step": 7250 + }, + { + "epoch": 0.5031573103878981, + "grad_norm": 3.1364502906799316, + "learning_rate": 0.00019522713913479187, + "loss": 1.1617, + "step": 7251 + }, + { + "epoch": 0.5032267018249948, + "grad_norm": 2.5995874404907227, + "learning_rate": 0.00019522542388706544, + "loss": 0.8883, + "step": 7252 + }, + { + "epoch": 0.5032960932620915, + "grad_norm": 2.484318256378174, + "learning_rate": 0.00019522370833872344, + "loss": 0.8475, + "step": 7253 + }, + { + "epoch": 0.5033654846991881, + "grad_norm": 3.179717779159546, + "learning_rate": 0.00019522199248977117, + "loss": 1.072, + "step": 7254 + }, + { + "epoch": 0.5034348761362848, + "grad_norm": 3.4817047119140625, + "learning_rate": 0.00019522027634021418, + "loss": 1.0756, + "step": 7255 + }, + { + "epoch": 0.5035042675733814, + "grad_norm": 2.5531957149505615, + "learning_rate": 0.00019521855989005778, + "loss": 0.8577, + "step": 7256 + }, + { + "epoch": 0.5035736590104781, + "grad_norm": 2.4936306476593018, + "learning_rate": 0.00019521684313930746, + "loss": 0.9932, + "step": 7257 + }, + { + "epoch": 0.5036430504475747, + "grad_norm": 3.1782174110412598, + "learning_rate": 0.00019521512608796862, + "loss": 0.9124, + "step": 7258 + }, + { + "epoch": 0.5037124418846715, + "grad_norm": 4.572049140930176, + "learning_rate": 0.00019521340873604665, + "loss": 0.8974, + "step": 7259 + }, + { + "epoch": 0.5037818333217681, + "grad_norm": 2.651663303375244, + "learning_rate": 0.000195211691083547, + "loss": 0.7239, + "step": 7260 + }, + { + "epoch": 0.5038512247588648, + "grad_norm": 2.35186767578125, + "learning_rate": 0.00019520997313047508, + "loss": 0.8336, + "step": 7261 + }, + { + "epoch": 0.5039206161959614, + "grad_norm": 2.4260568618774414, + "learning_rate": 0.00019520825487683634, + "loss": 0.815, + "step": 7262 + }, + { + "epoch": 0.5039900076330581, + "grad_norm": 2.912114143371582, + "learning_rate": 0.00019520653632263616, + "loss": 0.7901, + "step": 7263 + }, + { + "epoch": 0.5040593990701547, + "grad_norm": 3.3876185417175293, + "learning_rate": 0.00019520481746788005, + "loss": 0.9298, + "step": 7264 + }, + { + "epoch": 0.5041287905072515, + "grad_norm": 2.4346866607666016, + "learning_rate": 0.0001952030983125733, + "loss": 0.9974, + "step": 7265 + }, + { + "epoch": 0.5041981819443481, + "grad_norm": 2.6955859661102295, + "learning_rate": 0.00019520137885672148, + "loss": 0.9386, + "step": 7266 + }, + { + "epoch": 0.5042675733814447, + "grad_norm": 3.779125213623047, + "learning_rate": 0.0001951996591003299, + "loss": 0.8829, + "step": 7267 + }, + { + "epoch": 0.5043369648185414, + "grad_norm": 3.378845453262329, + "learning_rate": 0.00019519793904340408, + "loss": 1.2025, + "step": 7268 + }, + { + "epoch": 0.504406356255638, + "grad_norm": 3.0199382305145264, + "learning_rate": 0.0001951962186859494, + "loss": 0.9463, + "step": 7269 + }, + { + "epoch": 0.5044757476927347, + "grad_norm": 2.4673147201538086, + "learning_rate": 0.0001951944980279713, + "loss": 0.8695, + "step": 7270 + }, + { + "epoch": 0.5045451391298313, + "grad_norm": 3.014299154281616, + "learning_rate": 0.0001951927770694752, + "loss": 0.9804, + "step": 7271 + }, + { + "epoch": 0.5046145305669281, + "grad_norm": 3.610011339187622, + "learning_rate": 0.00019519105581046657, + "loss": 0.9865, + "step": 7272 + }, + { + "epoch": 0.5046839220040247, + "grad_norm": 2.3572921752929688, + "learning_rate": 0.0001951893342509508, + "loss": 0.9182, + "step": 7273 + }, + { + "epoch": 0.5047533134411214, + "grad_norm": 2.9406261444091797, + "learning_rate": 0.00019518761239093336, + "loss": 1.0506, + "step": 7274 + }, + { + "epoch": 0.504822704878218, + "grad_norm": 2.5333826541900635, + "learning_rate": 0.0001951858902304197, + "loss": 0.8321, + "step": 7275 + }, + { + "epoch": 0.5048920963153147, + "grad_norm": 4.078226089477539, + "learning_rate": 0.00019518416776941516, + "loss": 0.8634, + "step": 7276 + }, + { + "epoch": 0.5049614877524113, + "grad_norm": 4.30327033996582, + "learning_rate": 0.00019518244500792527, + "loss": 1.3764, + "step": 7277 + }, + { + "epoch": 0.5050308791895081, + "grad_norm": 3.191317081451416, + "learning_rate": 0.00019518072194595548, + "loss": 0.8068, + "step": 7278 + }, + { + "epoch": 0.5051002706266047, + "grad_norm": 2.4951510429382324, + "learning_rate": 0.00019517899858351116, + "loss": 0.8703, + "step": 7279 + }, + { + "epoch": 0.5051696620637013, + "grad_norm": 3.8182783126831055, + "learning_rate": 0.0001951772749205978, + "loss": 1.2353, + "step": 7280 + }, + { + "epoch": 0.505239053500798, + "grad_norm": 2.976454734802246, + "learning_rate": 0.00019517555095722083, + "loss": 0.9105, + "step": 7281 + }, + { + "epoch": 0.5053084449378946, + "grad_norm": 2.2952911853790283, + "learning_rate": 0.00019517382669338566, + "loss": 0.8665, + "step": 7282 + }, + { + "epoch": 0.5053778363749913, + "grad_norm": 2.111943006515503, + "learning_rate": 0.0001951721021290978, + "loss": 0.8518, + "step": 7283 + }, + { + "epoch": 0.505447227812088, + "grad_norm": 2.7823023796081543, + "learning_rate": 0.00019517037726436262, + "loss": 0.9725, + "step": 7284 + }, + { + "epoch": 0.5055166192491847, + "grad_norm": 2.7507729530334473, + "learning_rate": 0.00019516865209918564, + "loss": 0.9389, + "step": 7285 + }, + { + "epoch": 0.5055860106862813, + "grad_norm": 2.6934545040130615, + "learning_rate": 0.00019516692663357224, + "loss": 0.9826, + "step": 7286 + }, + { + "epoch": 0.505655402123378, + "grad_norm": 3.2534422874450684, + "learning_rate": 0.0001951652008675279, + "loss": 1.083, + "step": 7287 + }, + { + "epoch": 0.5057247935604746, + "grad_norm": 2.710570812225342, + "learning_rate": 0.00019516347480105805, + "loss": 0.9362, + "step": 7288 + }, + { + "epoch": 0.5057941849975713, + "grad_norm": 2.719766855239868, + "learning_rate": 0.00019516174843416817, + "loss": 1.0203, + "step": 7289 + }, + { + "epoch": 0.505863576434668, + "grad_norm": 2.6078131198883057, + "learning_rate": 0.00019516002176686367, + "loss": 0.8076, + "step": 7290 + }, + { + "epoch": 0.5059329678717647, + "grad_norm": 2.9204485416412354, + "learning_rate": 0.00019515829479915003, + "loss": 0.7042, + "step": 7291 + }, + { + "epoch": 0.5060023593088613, + "grad_norm": 1.7849944829940796, + "learning_rate": 0.0001951565675310327, + "loss": 0.5867, + "step": 7292 + }, + { + "epoch": 0.5060717507459579, + "grad_norm": 3.983990430831909, + "learning_rate": 0.0001951548399625171, + "loss": 1.0599, + "step": 7293 + }, + { + "epoch": 0.5061411421830546, + "grad_norm": 2.705590009689331, + "learning_rate": 0.00019515311209360875, + "loss": 0.9266, + "step": 7294 + }, + { + "epoch": 0.5062105336201512, + "grad_norm": 2.6408896446228027, + "learning_rate": 0.00019515138392431304, + "loss": 1.1926, + "step": 7295 + }, + { + "epoch": 0.506279925057248, + "grad_norm": 4.073147773742676, + "learning_rate": 0.00019514965545463548, + "loss": 1.0409, + "step": 7296 + }, + { + "epoch": 0.5063493164943446, + "grad_norm": 2.5233957767486572, + "learning_rate": 0.00019514792668458148, + "loss": 0.7205, + "step": 7297 + }, + { + "epoch": 0.5064187079314413, + "grad_norm": 2.8409180641174316, + "learning_rate": 0.0001951461976141565, + "loss": 0.925, + "step": 7298 + }, + { + "epoch": 0.5064880993685379, + "grad_norm": 2.7059667110443115, + "learning_rate": 0.00019514446824336602, + "loss": 1.0148, + "step": 7299 + }, + { + "epoch": 0.5065574908056346, + "grad_norm": 2.982218027114868, + "learning_rate": 0.0001951427385722155, + "loss": 1.1505, + "step": 7300 + }, + { + "epoch": 0.5066268822427312, + "grad_norm": 3.0056326389312744, + "learning_rate": 0.0001951410086007104, + "loss": 0.7897, + "step": 7301 + }, + { + "epoch": 0.506696273679828, + "grad_norm": 3.2773540019989014, + "learning_rate": 0.00019513927832885615, + "loss": 1.1566, + "step": 7302 + }, + { + "epoch": 0.5067656651169246, + "grad_norm": 2.6681835651397705, + "learning_rate": 0.00019513754775665824, + "loss": 0.8935, + "step": 7303 + }, + { + "epoch": 0.5068350565540213, + "grad_norm": 3.5644240379333496, + "learning_rate": 0.00019513581688412215, + "loss": 1.1484, + "step": 7304 + }, + { + "epoch": 0.5069044479911179, + "grad_norm": 2.971548557281494, + "learning_rate": 0.0001951340857112533, + "loss": 0.7734, + "step": 7305 + }, + { + "epoch": 0.5069738394282146, + "grad_norm": 2.3699913024902344, + "learning_rate": 0.0001951323542380572, + "loss": 0.7848, + "step": 7306 + }, + { + "epoch": 0.5070432308653112, + "grad_norm": 2.4273030757904053, + "learning_rate": 0.00019513062246453931, + "loss": 0.9085, + "step": 7307 + }, + { + "epoch": 0.5071126223024078, + "grad_norm": 2.682098627090454, + "learning_rate": 0.00019512889039070505, + "loss": 1.0841, + "step": 7308 + }, + { + "epoch": 0.5071820137395046, + "grad_norm": 2.8797245025634766, + "learning_rate": 0.00019512715801655995, + "loss": 0.7892, + "step": 7309 + }, + { + "epoch": 0.5072514051766012, + "grad_norm": 2.451931953430176, + "learning_rate": 0.0001951254253421094, + "loss": 0.9108, + "step": 7310 + }, + { + "epoch": 0.5073207966136979, + "grad_norm": 2.0775046348571777, + "learning_rate": 0.00019512369236735896, + "loss": 0.6073, + "step": 7311 + }, + { + "epoch": 0.5073901880507945, + "grad_norm": 3.44708251953125, + "learning_rate": 0.00019512195909231406, + "loss": 1.0085, + "step": 7312 + }, + { + "epoch": 0.5074595794878912, + "grad_norm": 2.6451282501220703, + "learning_rate": 0.00019512022551698016, + "loss": 0.9159, + "step": 7313 + }, + { + "epoch": 0.5075289709249878, + "grad_norm": 2.7421398162841797, + "learning_rate": 0.00019511849164136272, + "loss": 0.7661, + "step": 7314 + }, + { + "epoch": 0.5075983623620846, + "grad_norm": 2.4666333198547363, + "learning_rate": 0.00019511675746546728, + "loss": 0.9898, + "step": 7315 + }, + { + "epoch": 0.5076677537991812, + "grad_norm": 2.8798744678497314, + "learning_rate": 0.00019511502298929926, + "loss": 0.9501, + "step": 7316 + }, + { + "epoch": 0.5077371452362779, + "grad_norm": 3.799736738204956, + "learning_rate": 0.00019511328821286412, + "loss": 0.9114, + "step": 7317 + }, + { + "epoch": 0.5078065366733745, + "grad_norm": 2.933877944946289, + "learning_rate": 0.0001951115531361674, + "loss": 0.7778, + "step": 7318 + }, + { + "epoch": 0.5078759281104712, + "grad_norm": 2.9906604290008545, + "learning_rate": 0.00019510981775921452, + "loss": 1.2481, + "step": 7319 + }, + { + "epoch": 0.5079453195475678, + "grad_norm": 2.3977465629577637, + "learning_rate": 0.000195108082082011, + "loss": 0.9169, + "step": 7320 + }, + { + "epoch": 0.5080147109846644, + "grad_norm": 2.897941827774048, + "learning_rate": 0.00019510634610456226, + "loss": 1.1029, + "step": 7321 + }, + { + "epoch": 0.5080841024217612, + "grad_norm": 2.181605100631714, + "learning_rate": 0.00019510460982687387, + "loss": 0.7881, + "step": 7322 + }, + { + "epoch": 0.5081534938588578, + "grad_norm": 2.9162256717681885, + "learning_rate": 0.00019510287324895123, + "loss": 0.9997, + "step": 7323 + }, + { + "epoch": 0.5082228852959545, + "grad_norm": 2.22004771232605, + "learning_rate": 0.00019510113637079987, + "loss": 0.8099, + "step": 7324 + }, + { + "epoch": 0.5082922767330511, + "grad_norm": 2.2480504512786865, + "learning_rate": 0.00019509939919242525, + "loss": 0.7088, + "step": 7325 + }, + { + "epoch": 0.5083616681701478, + "grad_norm": 2.9214696884155273, + "learning_rate": 0.00019509766171383282, + "loss": 0.8474, + "step": 7326 + }, + { + "epoch": 0.5084310596072444, + "grad_norm": 3.2466330528259277, + "learning_rate": 0.00019509592393502815, + "loss": 0.9467, + "step": 7327 + }, + { + "epoch": 0.5085004510443412, + "grad_norm": 2.715546131134033, + "learning_rate": 0.00019509418585601668, + "loss": 0.9616, + "step": 7328 + }, + { + "epoch": 0.5085698424814378, + "grad_norm": 2.396972179412842, + "learning_rate": 0.00019509244747680388, + "loss": 0.8798, + "step": 7329 + }, + { + "epoch": 0.5086392339185345, + "grad_norm": 2.41096568107605, + "learning_rate": 0.0001950907087973953, + "loss": 0.98, + "step": 7330 + }, + { + "epoch": 0.5087086253556311, + "grad_norm": 3.1078720092773438, + "learning_rate": 0.00019508896981779633, + "loss": 0.9856, + "step": 7331 + }, + { + "epoch": 0.5087780167927278, + "grad_norm": 2.7289505004882812, + "learning_rate": 0.00019508723053801255, + "loss": 0.8949, + "step": 7332 + }, + { + "epoch": 0.5088474082298244, + "grad_norm": 2.895784616470337, + "learning_rate": 0.0001950854909580494, + "loss": 1.0611, + "step": 7333 + }, + { + "epoch": 0.508916799666921, + "grad_norm": 3.485131025314331, + "learning_rate": 0.0001950837510779124, + "loss": 0.7715, + "step": 7334 + }, + { + "epoch": 0.5089861911040178, + "grad_norm": 2.9993209838867188, + "learning_rate": 0.000195082010897607, + "loss": 0.859, + "step": 7335 + }, + { + "epoch": 0.5090555825411144, + "grad_norm": 3.113882064819336, + "learning_rate": 0.00019508027041713875, + "loss": 1.1848, + "step": 7336 + }, + { + "epoch": 0.5091249739782111, + "grad_norm": 2.687861204147339, + "learning_rate": 0.0001950785296365131, + "loss": 1.1184, + "step": 7337 + }, + { + "epoch": 0.5091943654153077, + "grad_norm": 4.609041690826416, + "learning_rate": 0.0001950767885557356, + "loss": 0.7825, + "step": 7338 + }, + { + "epoch": 0.5092637568524044, + "grad_norm": 2.675279140472412, + "learning_rate": 0.00019507504717481168, + "loss": 0.9331, + "step": 7339 + }, + { + "epoch": 0.509333148289501, + "grad_norm": 3.082014560699463, + "learning_rate": 0.00019507330549374684, + "loss": 0.8509, + "step": 7340 + }, + { + "epoch": 0.5094025397265978, + "grad_norm": 2.78337025642395, + "learning_rate": 0.00019507156351254666, + "loss": 0.7166, + "step": 7341 + }, + { + "epoch": 0.5094719311636944, + "grad_norm": 3.6459438800811768, + "learning_rate": 0.00019506982123121655, + "loss": 1.2754, + "step": 7342 + }, + { + "epoch": 0.5095413226007911, + "grad_norm": 3.047672986984253, + "learning_rate": 0.00019506807864976206, + "loss": 0.7777, + "step": 7343 + }, + { + "epoch": 0.5096107140378877, + "grad_norm": 3.1285698413848877, + "learning_rate": 0.00019506633576818865, + "loss": 0.8499, + "step": 7344 + }, + { + "epoch": 0.5096801054749844, + "grad_norm": 2.2557168006896973, + "learning_rate": 0.00019506459258650187, + "loss": 0.9371, + "step": 7345 + }, + { + "epoch": 0.509749496912081, + "grad_norm": 2.8235888481140137, + "learning_rate": 0.0001950628491047072, + "loss": 0.9058, + "step": 7346 + }, + { + "epoch": 0.5098188883491777, + "grad_norm": 3.3244705200195312, + "learning_rate": 0.00019506110532281013, + "loss": 0.8935, + "step": 7347 + }, + { + "epoch": 0.5098882797862744, + "grad_norm": 2.1876325607299805, + "learning_rate": 0.0001950593612408162, + "loss": 0.8436, + "step": 7348 + }, + { + "epoch": 0.509957671223371, + "grad_norm": 2.714290142059326, + "learning_rate": 0.0001950576168587309, + "loss": 0.8576, + "step": 7349 + }, + { + "epoch": 0.5100270626604677, + "grad_norm": 2.7042043209075928, + "learning_rate": 0.00019505587217655973, + "loss": 0.6866, + "step": 7350 + }, + { + "epoch": 0.5100964540975643, + "grad_norm": 3.6873862743377686, + "learning_rate": 0.00019505412719430818, + "loss": 0.9062, + "step": 7351 + }, + { + "epoch": 0.510165845534661, + "grad_norm": 2.6487152576446533, + "learning_rate": 0.00019505238191198178, + "loss": 0.9131, + "step": 7352 + }, + { + "epoch": 0.5102352369717577, + "grad_norm": 3.045623779296875, + "learning_rate": 0.00019505063632958605, + "loss": 1.1672, + "step": 7353 + }, + { + "epoch": 0.5103046284088544, + "grad_norm": 2.811152458190918, + "learning_rate": 0.00019504889044712648, + "loss": 1.1152, + "step": 7354 + }, + { + "epoch": 0.510374019845951, + "grad_norm": 2.4497570991516113, + "learning_rate": 0.0001950471442646086, + "loss": 0.6518, + "step": 7355 + }, + { + "epoch": 0.5104434112830477, + "grad_norm": 2.8019065856933594, + "learning_rate": 0.00019504539778203789, + "loss": 1.1527, + "step": 7356 + }, + { + "epoch": 0.5105128027201443, + "grad_norm": 2.3406941890716553, + "learning_rate": 0.00019504365099941992, + "loss": 0.8453, + "step": 7357 + }, + { + "epoch": 0.510582194157241, + "grad_norm": 2.2893385887145996, + "learning_rate": 0.00019504190391676017, + "loss": 0.7113, + "step": 7358 + }, + { + "epoch": 0.5106515855943377, + "grad_norm": 3.4587714672088623, + "learning_rate": 0.00019504015653406414, + "loss": 0.8055, + "step": 7359 + }, + { + "epoch": 0.5107209770314344, + "grad_norm": 3.8686506748199463, + "learning_rate": 0.00019503840885133736, + "loss": 1.2975, + "step": 7360 + }, + { + "epoch": 0.510790368468531, + "grad_norm": 3.66035795211792, + "learning_rate": 0.00019503666086858535, + "loss": 0.8086, + "step": 7361 + }, + { + "epoch": 0.5108597599056276, + "grad_norm": 3.1798222064971924, + "learning_rate": 0.00019503491258581366, + "loss": 1.034, + "step": 7362 + }, + { + "epoch": 0.5109291513427243, + "grad_norm": 3.933762550354004, + "learning_rate": 0.00019503316400302774, + "loss": 1.1046, + "step": 7363 + }, + { + "epoch": 0.5109985427798209, + "grad_norm": 2.357680559158325, + "learning_rate": 0.00019503141512023316, + "loss": 0.7688, + "step": 7364 + }, + { + "epoch": 0.5110679342169177, + "grad_norm": 2.32159161567688, + "learning_rate": 0.00019502966593743542, + "loss": 0.746, + "step": 7365 + }, + { + "epoch": 0.5111373256540143, + "grad_norm": 2.355621576309204, + "learning_rate": 0.00019502791645464006, + "loss": 0.8643, + "step": 7366 + }, + { + "epoch": 0.511206717091111, + "grad_norm": 3.141777276992798, + "learning_rate": 0.0001950261666718526, + "loss": 1.0221, + "step": 7367 + }, + { + "epoch": 0.5112761085282076, + "grad_norm": 2.517091751098633, + "learning_rate": 0.00019502441658907852, + "loss": 0.9759, + "step": 7368 + }, + { + "epoch": 0.5113454999653043, + "grad_norm": 2.799668073654175, + "learning_rate": 0.00019502266620632342, + "loss": 0.7186, + "step": 7369 + }, + { + "epoch": 0.5114148914024009, + "grad_norm": 2.523007869720459, + "learning_rate": 0.0001950209155235928, + "loss": 0.9581, + "step": 7370 + }, + { + "epoch": 0.5114842828394977, + "grad_norm": 4.618261814117432, + "learning_rate": 0.00019501916454089214, + "loss": 1.3341, + "step": 7371 + }, + { + "epoch": 0.5115536742765943, + "grad_norm": 2.327066421508789, + "learning_rate": 0.00019501741325822698, + "loss": 0.7926, + "step": 7372 + }, + { + "epoch": 0.511623065713691, + "grad_norm": 1.8688853979110718, + "learning_rate": 0.00019501566167560292, + "loss": 0.7291, + "step": 7373 + }, + { + "epoch": 0.5116924571507876, + "grad_norm": 2.8886311054229736, + "learning_rate": 0.0001950139097930254, + "loss": 0.9643, + "step": 7374 + }, + { + "epoch": 0.5117618485878842, + "grad_norm": 2.6177308559417725, + "learning_rate": 0.00019501215761050003, + "loss": 0.9045, + "step": 7375 + }, + { + "epoch": 0.5118312400249809, + "grad_norm": 2.473508596420288, + "learning_rate": 0.00019501040512803227, + "loss": 0.8721, + "step": 7376 + }, + { + "epoch": 0.5119006314620775, + "grad_norm": 2.004551410675049, + "learning_rate": 0.0001950086523456277, + "loss": 0.8508, + "step": 7377 + }, + { + "epoch": 0.5119700228991743, + "grad_norm": 3.1369504928588867, + "learning_rate": 0.00019500689926329185, + "loss": 1.2136, + "step": 7378 + }, + { + "epoch": 0.5120394143362709, + "grad_norm": 2.840636968612671, + "learning_rate": 0.0001950051458810302, + "loss": 1.0478, + "step": 7379 + }, + { + "epoch": 0.5121088057733676, + "grad_norm": 2.1411516666412354, + "learning_rate": 0.00019500339219884837, + "loss": 0.8606, + "step": 7380 + }, + { + "epoch": 0.5121781972104642, + "grad_norm": 2.260155200958252, + "learning_rate": 0.00019500163821675184, + "loss": 0.9149, + "step": 7381 + }, + { + "epoch": 0.5122475886475609, + "grad_norm": 2.229948043823242, + "learning_rate": 0.00019499988393474618, + "loss": 0.7397, + "step": 7382 + }, + { + "epoch": 0.5123169800846575, + "grad_norm": 2.7698259353637695, + "learning_rate": 0.00019499812935283686, + "loss": 0.676, + "step": 7383 + }, + { + "epoch": 0.5123863715217543, + "grad_norm": 2.4457848072052, + "learning_rate": 0.0001949963744710295, + "loss": 0.9526, + "step": 7384 + }, + { + "epoch": 0.5124557629588509, + "grad_norm": 2.741461753845215, + "learning_rate": 0.00019499461928932957, + "loss": 1.1915, + "step": 7385 + }, + { + "epoch": 0.5125251543959476, + "grad_norm": 3.2626070976257324, + "learning_rate": 0.0001949928638077427, + "loss": 0.843, + "step": 7386 + }, + { + "epoch": 0.5125945458330442, + "grad_norm": 2.3955817222595215, + "learning_rate": 0.00019499110802627439, + "loss": 1.0319, + "step": 7387 + }, + { + "epoch": 0.5126639372701408, + "grad_norm": 3.253077983856201, + "learning_rate": 0.00019498935194493014, + "loss": 1.1343, + "step": 7388 + }, + { + "epoch": 0.5127333287072375, + "grad_norm": 2.5933985710144043, + "learning_rate": 0.0001949875955637155, + "loss": 0.9123, + "step": 7389 + }, + { + "epoch": 0.5128027201443341, + "grad_norm": 3.029813051223755, + "learning_rate": 0.00019498583888263608, + "loss": 1.1395, + "step": 7390 + }, + { + "epoch": 0.5128721115814309, + "grad_norm": 2.230039358139038, + "learning_rate": 0.00019498408190169738, + "loss": 0.9173, + "step": 7391 + }, + { + "epoch": 0.5129415030185275, + "grad_norm": 2.7164220809936523, + "learning_rate": 0.00019498232462090496, + "loss": 0.7309, + "step": 7392 + }, + { + "epoch": 0.5130108944556242, + "grad_norm": 1.7026668787002563, + "learning_rate": 0.00019498056704026433, + "loss": 0.5515, + "step": 7393 + }, + { + "epoch": 0.5130802858927208, + "grad_norm": 2.3741936683654785, + "learning_rate": 0.00019497880915978112, + "loss": 0.9014, + "step": 7394 + }, + { + "epoch": 0.5131496773298175, + "grad_norm": 3.6184144020080566, + "learning_rate": 0.0001949770509794608, + "loss": 1.1109, + "step": 7395 + }, + { + "epoch": 0.5132190687669141, + "grad_norm": 2.5921339988708496, + "learning_rate": 0.00019497529249930894, + "loss": 0.9091, + "step": 7396 + }, + { + "epoch": 0.5132884602040109, + "grad_norm": 3.43719220161438, + "learning_rate": 0.0001949735337193311, + "loss": 0.8559, + "step": 7397 + }, + { + "epoch": 0.5133578516411075, + "grad_norm": 3.013227939605713, + "learning_rate": 0.00019497177463953286, + "loss": 1.1261, + "step": 7398 + }, + { + "epoch": 0.5134272430782042, + "grad_norm": 2.4923129081726074, + "learning_rate": 0.00019497001525991972, + "loss": 0.7749, + "step": 7399 + }, + { + "epoch": 0.5134966345153008, + "grad_norm": 3.151650905609131, + "learning_rate": 0.00019496825558049728, + "loss": 0.7622, + "step": 7400 + }, + { + "epoch": 0.5135660259523974, + "grad_norm": 2.9586715698242188, + "learning_rate": 0.00019496649560127107, + "loss": 1.0431, + "step": 7401 + }, + { + "epoch": 0.5136354173894941, + "grad_norm": 2.799140691757202, + "learning_rate": 0.00019496473532224664, + "loss": 0.9683, + "step": 7402 + }, + { + "epoch": 0.5137048088265908, + "grad_norm": 2.471592664718628, + "learning_rate": 0.00019496297474342957, + "loss": 0.9122, + "step": 7403 + }, + { + "epoch": 0.5137742002636875, + "grad_norm": 3.958928108215332, + "learning_rate": 0.0001949612138648254, + "loss": 0.7212, + "step": 7404 + }, + { + "epoch": 0.5138435917007841, + "grad_norm": 2.8226685523986816, + "learning_rate": 0.00019495945268643969, + "loss": 1.1866, + "step": 7405 + }, + { + "epoch": 0.5139129831378808, + "grad_norm": 2.6153974533081055, + "learning_rate": 0.00019495769120827803, + "loss": 0.8502, + "step": 7406 + }, + { + "epoch": 0.5139823745749774, + "grad_norm": 2.151237964630127, + "learning_rate": 0.00019495592943034594, + "loss": 0.7496, + "step": 7407 + }, + { + "epoch": 0.5140517660120741, + "grad_norm": 2.9160637855529785, + "learning_rate": 0.00019495416735264897, + "loss": 1.2192, + "step": 7408 + }, + { + "epoch": 0.5141211574491708, + "grad_norm": 2.817514657974243, + "learning_rate": 0.00019495240497519276, + "loss": 0.6926, + "step": 7409 + }, + { + "epoch": 0.5141905488862675, + "grad_norm": 3.537520170211792, + "learning_rate": 0.0001949506422979828, + "loss": 0.8186, + "step": 7410 + }, + { + "epoch": 0.5142599403233641, + "grad_norm": 2.8784780502319336, + "learning_rate": 0.00019494887932102468, + "loss": 0.9568, + "step": 7411 + }, + { + "epoch": 0.5143293317604608, + "grad_norm": 3.079198122024536, + "learning_rate": 0.00019494711604432397, + "loss": 1.2095, + "step": 7412 + }, + { + "epoch": 0.5143987231975574, + "grad_norm": 3.604482412338257, + "learning_rate": 0.00019494535246788623, + "loss": 1.0499, + "step": 7413 + }, + { + "epoch": 0.5144681146346541, + "grad_norm": 2.203575611114502, + "learning_rate": 0.00019494358859171703, + "loss": 0.8927, + "step": 7414 + }, + { + "epoch": 0.5145375060717507, + "grad_norm": 2.821154832839966, + "learning_rate": 0.00019494182441582192, + "loss": 1.1905, + "step": 7415 + }, + { + "epoch": 0.5146068975088474, + "grad_norm": 2.300457239151001, + "learning_rate": 0.00019494005994020647, + "loss": 0.767, + "step": 7416 + }, + { + "epoch": 0.5146762889459441, + "grad_norm": 2.8387157917022705, + "learning_rate": 0.0001949382951648763, + "loss": 0.8885, + "step": 7417 + }, + { + "epoch": 0.5147456803830407, + "grad_norm": 2.794265031814575, + "learning_rate": 0.00019493653008983695, + "loss": 0.8901, + "step": 7418 + }, + { + "epoch": 0.5148150718201374, + "grad_norm": 1.8431639671325684, + "learning_rate": 0.00019493476471509397, + "loss": 0.6915, + "step": 7419 + }, + { + "epoch": 0.514884463257234, + "grad_norm": 2.268425941467285, + "learning_rate": 0.00019493299904065298, + "loss": 0.951, + "step": 7420 + }, + { + "epoch": 0.5149538546943307, + "grad_norm": 2.2741973400115967, + "learning_rate": 0.00019493123306651947, + "loss": 0.9117, + "step": 7421 + }, + { + "epoch": 0.5150232461314274, + "grad_norm": 2.440904140472412, + "learning_rate": 0.00019492946679269913, + "loss": 0.8863, + "step": 7422 + }, + { + "epoch": 0.5150926375685241, + "grad_norm": 3.170168399810791, + "learning_rate": 0.00019492770021919747, + "loss": 0.9714, + "step": 7423 + }, + { + "epoch": 0.5151620290056207, + "grad_norm": 2.8100101947784424, + "learning_rate": 0.00019492593334602005, + "loss": 0.8785, + "step": 7424 + }, + { + "epoch": 0.5152314204427174, + "grad_norm": 2.2501354217529297, + "learning_rate": 0.0001949241661731725, + "loss": 0.716, + "step": 7425 + }, + { + "epoch": 0.515300811879814, + "grad_norm": 2.683236837387085, + "learning_rate": 0.0001949223987006603, + "loss": 0.8419, + "step": 7426 + }, + { + "epoch": 0.5153702033169107, + "grad_norm": 2.912916421890259, + "learning_rate": 0.00019492063092848918, + "loss": 0.9083, + "step": 7427 + }, + { + "epoch": 0.5154395947540074, + "grad_norm": 3.51403546333313, + "learning_rate": 0.0001949188628566646, + "loss": 0.8171, + "step": 7428 + }, + { + "epoch": 0.515508986191104, + "grad_norm": 2.368734359741211, + "learning_rate": 0.0001949170944851922, + "loss": 0.7733, + "step": 7429 + }, + { + "epoch": 0.5155783776282007, + "grad_norm": 3.4703426361083984, + "learning_rate": 0.00019491532581407755, + "loss": 0.8871, + "step": 7430 + }, + { + "epoch": 0.5156477690652973, + "grad_norm": 3.799018621444702, + "learning_rate": 0.0001949135568433262, + "loss": 1.1894, + "step": 7431 + }, + { + "epoch": 0.515717160502394, + "grad_norm": 2.206907272338867, + "learning_rate": 0.00019491178757294377, + "loss": 0.6818, + "step": 7432 + }, + { + "epoch": 0.5157865519394906, + "grad_norm": 2.7396106719970703, + "learning_rate": 0.00019491001800293586, + "loss": 0.86, + "step": 7433 + }, + { + "epoch": 0.5158559433765874, + "grad_norm": 3.27091121673584, + "learning_rate": 0.00019490824813330805, + "loss": 1.0722, + "step": 7434 + }, + { + "epoch": 0.515925334813684, + "grad_norm": 2.6980690956115723, + "learning_rate": 0.00019490647796406588, + "loss": 0.8438, + "step": 7435 + }, + { + "epoch": 0.5159947262507807, + "grad_norm": 2.9922752380371094, + "learning_rate": 0.00019490470749521498, + "loss": 0.9572, + "step": 7436 + }, + { + "epoch": 0.5160641176878773, + "grad_norm": 2.653737783432007, + "learning_rate": 0.0001949029367267609, + "loss": 0.8448, + "step": 7437 + }, + { + "epoch": 0.516133509124974, + "grad_norm": 3.1304166316986084, + "learning_rate": 0.00019490116565870928, + "loss": 0.8825, + "step": 7438 + }, + { + "epoch": 0.5162029005620706, + "grad_norm": 3.085573196411133, + "learning_rate": 0.0001948993942910657, + "loss": 0.8312, + "step": 7439 + }, + { + "epoch": 0.5162722919991674, + "grad_norm": 4.933455467224121, + "learning_rate": 0.00019489762262383576, + "loss": 0.8765, + "step": 7440 + }, + { + "epoch": 0.516341683436264, + "grad_norm": 3.0150675773620605, + "learning_rate": 0.00019489585065702501, + "loss": 0.8968, + "step": 7441 + }, + { + "epoch": 0.5164110748733606, + "grad_norm": 4.920393466949463, + "learning_rate": 0.00019489407839063907, + "loss": 1.1734, + "step": 7442 + }, + { + "epoch": 0.5164804663104573, + "grad_norm": 2.6717495918273926, + "learning_rate": 0.00019489230582468353, + "loss": 0.9364, + "step": 7443 + }, + { + "epoch": 0.5165498577475539, + "grad_norm": 2.7825143337249756, + "learning_rate": 0.00019489053295916402, + "loss": 0.7209, + "step": 7444 + }, + { + "epoch": 0.5166192491846506, + "grad_norm": 4.218972682952881, + "learning_rate": 0.00019488875979408607, + "loss": 0.9747, + "step": 7445 + }, + { + "epoch": 0.5166886406217472, + "grad_norm": 5.4580206871032715, + "learning_rate": 0.00019488698632945536, + "loss": 0.6914, + "step": 7446 + }, + { + "epoch": 0.516758032058844, + "grad_norm": 3.3886635303497314, + "learning_rate": 0.0001948852125652774, + "loss": 0.9442, + "step": 7447 + }, + { + "epoch": 0.5168274234959406, + "grad_norm": 3.1535041332244873, + "learning_rate": 0.00019488343850155784, + "loss": 0.8761, + "step": 7448 + }, + { + "epoch": 0.5168968149330373, + "grad_norm": 2.473823308944702, + "learning_rate": 0.00019488166413830226, + "loss": 0.8811, + "step": 7449 + }, + { + "epoch": 0.5169662063701339, + "grad_norm": 2.1751222610473633, + "learning_rate": 0.0001948798894755163, + "loss": 0.9454, + "step": 7450 + }, + { + "epoch": 0.5170355978072306, + "grad_norm": 2.981003761291504, + "learning_rate": 0.00019487811451320553, + "loss": 0.8328, + "step": 7451 + }, + { + "epoch": 0.5171049892443272, + "grad_norm": 2.2803714275360107, + "learning_rate": 0.00019487633925137555, + "loss": 0.8407, + "step": 7452 + }, + { + "epoch": 0.517174380681424, + "grad_norm": 3.255462646484375, + "learning_rate": 0.00019487456369003198, + "loss": 1.1097, + "step": 7453 + }, + { + "epoch": 0.5172437721185206, + "grad_norm": 2.653779983520508, + "learning_rate": 0.00019487278782918046, + "loss": 0.9018, + "step": 7454 + }, + { + "epoch": 0.5173131635556172, + "grad_norm": 2.1298959255218506, + "learning_rate": 0.00019487101166882652, + "loss": 0.752, + "step": 7455 + }, + { + "epoch": 0.5173825549927139, + "grad_norm": 2.6920337677001953, + "learning_rate": 0.00019486923520897577, + "loss": 1.2085, + "step": 7456 + }, + { + "epoch": 0.5174519464298105, + "grad_norm": 2.702946186065674, + "learning_rate": 0.0001948674584496339, + "loss": 0.9414, + "step": 7457 + }, + { + "epoch": 0.5175213378669072, + "grad_norm": 3.1621949672698975, + "learning_rate": 0.00019486568139080648, + "loss": 1.0779, + "step": 7458 + }, + { + "epoch": 0.5175907293040038, + "grad_norm": 2.6848294734954834, + "learning_rate": 0.00019486390403249907, + "loss": 0.9413, + "step": 7459 + }, + { + "epoch": 0.5176601207411006, + "grad_norm": 2.6908071041107178, + "learning_rate": 0.00019486212637471732, + "loss": 0.8802, + "step": 7460 + }, + { + "epoch": 0.5177295121781972, + "grad_norm": 2.4951155185699463, + "learning_rate": 0.00019486034841746688, + "loss": 0.8435, + "step": 7461 + }, + { + "epoch": 0.5177989036152939, + "grad_norm": 2.886057138442993, + "learning_rate": 0.0001948585701607533, + "loss": 0.825, + "step": 7462 + }, + { + "epoch": 0.5178682950523905, + "grad_norm": 2.649959087371826, + "learning_rate": 0.00019485679160458224, + "loss": 0.8448, + "step": 7463 + }, + { + "epoch": 0.5179376864894872, + "grad_norm": 2.4312257766723633, + "learning_rate": 0.00019485501274895927, + "loss": 1.0325, + "step": 7464 + }, + { + "epoch": 0.5180070779265838, + "grad_norm": 2.663250207901001, + "learning_rate": 0.00019485323359389004, + "loss": 0.8283, + "step": 7465 + }, + { + "epoch": 0.5180764693636806, + "grad_norm": 2.9637575149536133, + "learning_rate": 0.00019485145413938016, + "loss": 0.7852, + "step": 7466 + }, + { + "epoch": 0.5181458608007772, + "grad_norm": 3.440882682800293, + "learning_rate": 0.00019484967438543525, + "loss": 1.0903, + "step": 7467 + }, + { + "epoch": 0.5182152522378738, + "grad_norm": 2.9913065433502197, + "learning_rate": 0.0001948478943320609, + "loss": 1.414, + "step": 7468 + }, + { + "epoch": 0.5182846436749705, + "grad_norm": 2.548530340194702, + "learning_rate": 0.00019484611397926278, + "loss": 1.1012, + "step": 7469 + }, + { + "epoch": 0.5183540351120671, + "grad_norm": 2.7733147144317627, + "learning_rate": 0.00019484433332704645, + "loss": 0.9531, + "step": 7470 + }, + { + "epoch": 0.5184234265491638, + "grad_norm": 3.0840938091278076, + "learning_rate": 0.0001948425523754176, + "loss": 1.0221, + "step": 7471 + }, + { + "epoch": 0.5184928179862605, + "grad_norm": 2.2508933544158936, + "learning_rate": 0.00019484077112438177, + "loss": 0.6574, + "step": 7472 + }, + { + "epoch": 0.5185622094233572, + "grad_norm": 2.310713529586792, + "learning_rate": 0.00019483898957394467, + "loss": 0.9348, + "step": 7473 + }, + { + "epoch": 0.5186316008604538, + "grad_norm": 2.5577659606933594, + "learning_rate": 0.00019483720772411187, + "loss": 0.8457, + "step": 7474 + }, + { + "epoch": 0.5187009922975505, + "grad_norm": 2.659904956817627, + "learning_rate": 0.000194835425574889, + "loss": 0.7846, + "step": 7475 + }, + { + "epoch": 0.5187703837346471, + "grad_norm": 3.493995189666748, + "learning_rate": 0.00019483364312628172, + "loss": 1.033, + "step": 7476 + }, + { + "epoch": 0.5188397751717438, + "grad_norm": 2.4693779945373535, + "learning_rate": 0.00019483186037829562, + "loss": 1.0512, + "step": 7477 + }, + { + "epoch": 0.5189091666088405, + "grad_norm": 4.437367916107178, + "learning_rate": 0.00019483007733093634, + "loss": 1.0436, + "step": 7478 + }, + { + "epoch": 0.5189785580459372, + "grad_norm": 2.6038568019866943, + "learning_rate": 0.00019482829398420948, + "loss": 1.0289, + "step": 7479 + }, + { + "epoch": 0.5190479494830338, + "grad_norm": 3.264569044113159, + "learning_rate": 0.00019482651033812074, + "loss": 1.2674, + "step": 7480 + }, + { + "epoch": 0.5191173409201305, + "grad_norm": 2.589052200317383, + "learning_rate": 0.00019482472639267568, + "loss": 0.6198, + "step": 7481 + }, + { + "epoch": 0.5191867323572271, + "grad_norm": 2.2817280292510986, + "learning_rate": 0.00019482294214787996, + "loss": 0.75, + "step": 7482 + }, + { + "epoch": 0.5192561237943237, + "grad_norm": 2.857245683670044, + "learning_rate": 0.00019482115760373922, + "loss": 0.9566, + "step": 7483 + }, + { + "epoch": 0.5193255152314205, + "grad_norm": 3.1104390621185303, + "learning_rate": 0.0001948193727602591, + "loss": 0.9587, + "step": 7484 + }, + { + "epoch": 0.5193949066685171, + "grad_norm": 2.4012978076934814, + "learning_rate": 0.0001948175876174452, + "loss": 0.8409, + "step": 7485 + }, + { + "epoch": 0.5194642981056138, + "grad_norm": 2.4754695892333984, + "learning_rate": 0.00019481580217530316, + "loss": 0.8875, + "step": 7486 + }, + { + "epoch": 0.5195336895427104, + "grad_norm": 2.838895320892334, + "learning_rate": 0.0001948140164338387, + "loss": 1.149, + "step": 7487 + }, + { + "epoch": 0.5196030809798071, + "grad_norm": 2.1556825637817383, + "learning_rate": 0.00019481223039305732, + "loss": 0.8035, + "step": 7488 + }, + { + "epoch": 0.5196724724169037, + "grad_norm": 2.438154697418213, + "learning_rate": 0.00019481044405296477, + "loss": 0.7935, + "step": 7489 + }, + { + "epoch": 0.5197418638540005, + "grad_norm": 2.6795732975006104, + "learning_rate": 0.00019480865741356665, + "loss": 1.0411, + "step": 7490 + }, + { + "epoch": 0.5198112552910971, + "grad_norm": 3.0593678951263428, + "learning_rate": 0.0001948068704748686, + "loss": 0.7916, + "step": 7491 + }, + { + "epoch": 0.5198806467281938, + "grad_norm": 2.5985817909240723, + "learning_rate": 0.00019480508323687623, + "loss": 0.9903, + "step": 7492 + }, + { + "epoch": 0.5199500381652904, + "grad_norm": 2.977983236312866, + "learning_rate": 0.00019480329569959525, + "loss": 1.2017, + "step": 7493 + }, + { + "epoch": 0.5200194296023871, + "grad_norm": 2.2362444400787354, + "learning_rate": 0.0001948015078630312, + "loss": 0.7939, + "step": 7494 + }, + { + "epoch": 0.5200888210394837, + "grad_norm": 2.1765246391296387, + "learning_rate": 0.00019479971972718985, + "loss": 0.8061, + "step": 7495 + }, + { + "epoch": 0.5201582124765803, + "grad_norm": 2.8703725337982178, + "learning_rate": 0.00019479793129207677, + "loss": 0.8926, + "step": 7496 + }, + { + "epoch": 0.5202276039136771, + "grad_norm": 2.4719204902648926, + "learning_rate": 0.0001947961425576976, + "loss": 0.868, + "step": 7497 + }, + { + "epoch": 0.5202969953507737, + "grad_norm": 2.9245476722717285, + "learning_rate": 0.00019479435352405804, + "loss": 0.8333, + "step": 7498 + }, + { + "epoch": 0.5203663867878704, + "grad_norm": 2.3414649963378906, + "learning_rate": 0.00019479256419116368, + "loss": 0.7968, + "step": 7499 + }, + { + "epoch": 0.520435778224967, + "grad_norm": 2.718313217163086, + "learning_rate": 0.0001947907745590202, + "loss": 1.0479, + "step": 7500 + }, + { + "epoch": 0.5205051696620637, + "grad_norm": 2.4637670516967773, + "learning_rate": 0.00019478898462763322, + "loss": 0.8461, + "step": 7501 + }, + { + "epoch": 0.5205745610991603, + "grad_norm": 2.218482255935669, + "learning_rate": 0.00019478719439700845, + "loss": 0.7547, + "step": 7502 + }, + { + "epoch": 0.5206439525362571, + "grad_norm": 3.849006175994873, + "learning_rate": 0.0001947854038671515, + "loss": 0.9317, + "step": 7503 + }, + { + "epoch": 0.5207133439733537, + "grad_norm": 2.3921022415161133, + "learning_rate": 0.00019478361303806798, + "loss": 0.9001, + "step": 7504 + }, + { + "epoch": 0.5207827354104504, + "grad_norm": 3.0241992473602295, + "learning_rate": 0.00019478182190976363, + "loss": 1.0036, + "step": 7505 + }, + { + "epoch": 0.520852126847547, + "grad_norm": 2.720964193344116, + "learning_rate": 0.00019478003048224406, + "loss": 0.6689, + "step": 7506 + }, + { + "epoch": 0.5209215182846437, + "grad_norm": 2.7842957973480225, + "learning_rate": 0.00019477823875551495, + "loss": 1.0207, + "step": 7507 + }, + { + "epoch": 0.5209909097217403, + "grad_norm": 3.4293746948242188, + "learning_rate": 0.0001947764467295819, + "loss": 1.0139, + "step": 7508 + }, + { + "epoch": 0.521060301158837, + "grad_norm": 2.420539140701294, + "learning_rate": 0.00019477465440445065, + "loss": 0.8261, + "step": 7509 + }, + { + "epoch": 0.5211296925959337, + "grad_norm": 3.3902063369750977, + "learning_rate": 0.00019477286178012676, + "loss": 0.9675, + "step": 7510 + }, + { + "epoch": 0.5211990840330303, + "grad_norm": 2.2896485328674316, + "learning_rate": 0.00019477106885661596, + "loss": 0.7634, + "step": 7511 + }, + { + "epoch": 0.521268475470127, + "grad_norm": 2.2871246337890625, + "learning_rate": 0.00019476927563392388, + "loss": 0.8458, + "step": 7512 + }, + { + "epoch": 0.5213378669072236, + "grad_norm": 3.469531774520874, + "learning_rate": 0.0001947674821120562, + "loss": 1.2182, + "step": 7513 + }, + { + "epoch": 0.5214072583443203, + "grad_norm": 2.2504115104675293, + "learning_rate": 0.00019476568829101862, + "loss": 0.8694, + "step": 7514 + }, + { + "epoch": 0.521476649781417, + "grad_norm": 3.156233549118042, + "learning_rate": 0.00019476389417081671, + "loss": 1.0026, + "step": 7515 + }, + { + "epoch": 0.5215460412185137, + "grad_norm": 3.4027843475341797, + "learning_rate": 0.0001947620997514562, + "loss": 1.1473, + "step": 7516 + }, + { + "epoch": 0.5216154326556103, + "grad_norm": 3.8543803691864014, + "learning_rate": 0.00019476030503294268, + "loss": 0.9464, + "step": 7517 + }, + { + "epoch": 0.521684824092707, + "grad_norm": 3.0747368335723877, + "learning_rate": 0.00019475851001528194, + "loss": 1.225, + "step": 7518 + }, + { + "epoch": 0.5217542155298036, + "grad_norm": 2.8299899101257324, + "learning_rate": 0.00019475671469847955, + "loss": 0.8888, + "step": 7519 + }, + { + "epoch": 0.5218236069669003, + "grad_norm": 3.8632278442382812, + "learning_rate": 0.0001947549190825412, + "loss": 1.0515, + "step": 7520 + }, + { + "epoch": 0.5218929984039969, + "grad_norm": 4.776130199432373, + "learning_rate": 0.0001947531231674726, + "loss": 1.308, + "step": 7521 + }, + { + "epoch": 0.5219623898410936, + "grad_norm": 7.759490966796875, + "learning_rate": 0.00019475132695327934, + "loss": 0.9183, + "step": 7522 + }, + { + "epoch": 0.5220317812781903, + "grad_norm": 2.763760805130005, + "learning_rate": 0.00019474953043996715, + "loss": 0.8471, + "step": 7523 + }, + { + "epoch": 0.5221011727152869, + "grad_norm": 3.8576831817626953, + "learning_rate": 0.0001947477336275417, + "loss": 0.8656, + "step": 7524 + }, + { + "epoch": 0.5221705641523836, + "grad_norm": 3.648998975753784, + "learning_rate": 0.00019474593651600862, + "loss": 0.975, + "step": 7525 + }, + { + "epoch": 0.5222399555894802, + "grad_norm": 4.480388164520264, + "learning_rate": 0.0001947441391053736, + "loss": 0.8156, + "step": 7526 + }, + { + "epoch": 0.5223093470265769, + "grad_norm": 3.4214370250701904, + "learning_rate": 0.0001947423413956424, + "loss": 0.804, + "step": 7527 + }, + { + "epoch": 0.5223787384636736, + "grad_norm": 4.011039733886719, + "learning_rate": 0.00019474054338682053, + "loss": 1.2974, + "step": 7528 + }, + { + "epoch": 0.5224481299007703, + "grad_norm": 3.39560866355896, + "learning_rate": 0.00019473874507891381, + "loss": 1.251, + "step": 7529 + }, + { + "epoch": 0.5225175213378669, + "grad_norm": 3.2113795280456543, + "learning_rate": 0.00019473694647192784, + "loss": 0.9084, + "step": 7530 + }, + { + "epoch": 0.5225869127749636, + "grad_norm": 3.1788721084594727, + "learning_rate": 0.00019473514756586835, + "loss": 0.94, + "step": 7531 + }, + { + "epoch": 0.5226563042120602, + "grad_norm": 16.366235733032227, + "learning_rate": 0.00019473334836074096, + "loss": 1.0718, + "step": 7532 + }, + { + "epoch": 0.5227256956491569, + "grad_norm": 3.4296927452087402, + "learning_rate": 0.0001947315488565514, + "loss": 1.2201, + "step": 7533 + }, + { + "epoch": 0.5227950870862536, + "grad_norm": 3.418564558029175, + "learning_rate": 0.00019472974905330532, + "loss": 1.1265, + "step": 7534 + }, + { + "epoch": 0.5228644785233503, + "grad_norm": 3.0347893238067627, + "learning_rate": 0.0001947279489510084, + "loss": 1.0087, + "step": 7535 + }, + { + "epoch": 0.5229338699604469, + "grad_norm": 3.7281432151794434, + "learning_rate": 0.00019472614854966635, + "loss": 1.125, + "step": 7536 + }, + { + "epoch": 0.5230032613975435, + "grad_norm": 2.7510628700256348, + "learning_rate": 0.00019472434784928487, + "loss": 0.8491, + "step": 7537 + }, + { + "epoch": 0.5230726528346402, + "grad_norm": 2.9350168704986572, + "learning_rate": 0.00019472254684986957, + "loss": 0.839, + "step": 7538 + }, + { + "epoch": 0.5231420442717368, + "grad_norm": 2.8238158226013184, + "learning_rate": 0.0001947207455514262, + "loss": 1.0787, + "step": 7539 + }, + { + "epoch": 0.5232114357088335, + "grad_norm": 2.3920233249664307, + "learning_rate": 0.0001947189439539604, + "loss": 1.0527, + "step": 7540 + }, + { + "epoch": 0.5232808271459302, + "grad_norm": 2.7589728832244873, + "learning_rate": 0.0001947171420574779, + "loss": 1.0904, + "step": 7541 + }, + { + "epoch": 0.5233502185830269, + "grad_norm": 2.7405765056610107, + "learning_rate": 0.0001947153398619844, + "loss": 0.8717, + "step": 7542 + }, + { + "epoch": 0.5234196100201235, + "grad_norm": 3.1810340881347656, + "learning_rate": 0.00019471353736748556, + "loss": 1.2777, + "step": 7543 + }, + { + "epoch": 0.5234890014572202, + "grad_norm": 3.089698314666748, + "learning_rate": 0.000194711734573987, + "loss": 0.9207, + "step": 7544 + }, + { + "epoch": 0.5235583928943168, + "grad_norm": 2.748230218887329, + "learning_rate": 0.00019470993148149455, + "loss": 1.0096, + "step": 7545 + }, + { + "epoch": 0.5236277843314135, + "grad_norm": 2.433124303817749, + "learning_rate": 0.0001947081280900138, + "loss": 0.7695, + "step": 7546 + }, + { + "epoch": 0.5236971757685102, + "grad_norm": 2.2147767543792725, + "learning_rate": 0.00019470632439955052, + "loss": 1.0139, + "step": 7547 + }, + { + "epoch": 0.5237665672056069, + "grad_norm": 3.451056718826294, + "learning_rate": 0.00019470452041011035, + "loss": 1.1832, + "step": 7548 + }, + { + "epoch": 0.5238359586427035, + "grad_norm": 3.524082899093628, + "learning_rate": 0.00019470271612169893, + "loss": 0.9528, + "step": 7549 + }, + { + "epoch": 0.5239053500798001, + "grad_norm": 2.636284351348877, + "learning_rate": 0.0001947009115343221, + "loss": 0.8143, + "step": 7550 + }, + { + "epoch": 0.5239747415168968, + "grad_norm": 4.142624855041504, + "learning_rate": 0.00019469910664798545, + "loss": 0.8614, + "step": 7551 + }, + { + "epoch": 0.5240441329539934, + "grad_norm": 3.1154472827911377, + "learning_rate": 0.0001946973014626947, + "loss": 1.0734, + "step": 7552 + }, + { + "epoch": 0.5241135243910902, + "grad_norm": 2.382262945175171, + "learning_rate": 0.0001946954959784556, + "loss": 0.7268, + "step": 7553 + }, + { + "epoch": 0.5241829158281868, + "grad_norm": 3.1739370822906494, + "learning_rate": 0.00019469369019527375, + "loss": 1.1556, + "step": 7554 + }, + { + "epoch": 0.5242523072652835, + "grad_norm": 2.346303701400757, + "learning_rate": 0.00019469188411315496, + "loss": 0.9405, + "step": 7555 + }, + { + "epoch": 0.5243216987023801, + "grad_norm": 2.8822531700134277, + "learning_rate": 0.00019469007773210485, + "loss": 0.8103, + "step": 7556 + }, + { + "epoch": 0.5243910901394768, + "grad_norm": 2.2481987476348877, + "learning_rate": 0.00019468827105212914, + "loss": 0.777, + "step": 7557 + }, + { + "epoch": 0.5244604815765734, + "grad_norm": 2.649613380432129, + "learning_rate": 0.00019468646407323355, + "loss": 1.2258, + "step": 7558 + }, + { + "epoch": 0.5245298730136702, + "grad_norm": 2.3812196254730225, + "learning_rate": 0.00019468465679542377, + "loss": 0.8994, + "step": 7559 + }, + { + "epoch": 0.5245992644507668, + "grad_norm": 2.5715014934539795, + "learning_rate": 0.00019468284921870553, + "loss": 0.8666, + "step": 7560 + }, + { + "epoch": 0.5246686558878635, + "grad_norm": 2.561227321624756, + "learning_rate": 0.00019468104134308452, + "loss": 0.9316, + "step": 7561 + }, + { + "epoch": 0.5247380473249601, + "grad_norm": 3.1741325855255127, + "learning_rate": 0.00019467923316856644, + "loss": 1.1185, + "step": 7562 + }, + { + "epoch": 0.5248074387620567, + "grad_norm": 3.5633907318115234, + "learning_rate": 0.000194677424695157, + "loss": 0.9618, + "step": 7563 + }, + { + "epoch": 0.5248768301991534, + "grad_norm": 3.76625919342041, + "learning_rate": 0.00019467561592286195, + "loss": 1.6836, + "step": 7564 + }, + { + "epoch": 0.52494622163625, + "grad_norm": 2.7893426418304443, + "learning_rate": 0.00019467380685168691, + "loss": 0.9575, + "step": 7565 + }, + { + "epoch": 0.5250156130733468, + "grad_norm": 2.5621767044067383, + "learning_rate": 0.0001946719974816377, + "loss": 0.9503, + "step": 7566 + }, + { + "epoch": 0.5250850045104434, + "grad_norm": 3.636038303375244, + "learning_rate": 0.00019467018781271994, + "loss": 1.031, + "step": 7567 + }, + { + "epoch": 0.5251543959475401, + "grad_norm": 3.015775442123413, + "learning_rate": 0.0001946683778449394, + "loss": 0.9256, + "step": 7568 + }, + { + "epoch": 0.5252237873846367, + "grad_norm": 2.4297125339508057, + "learning_rate": 0.00019466656757830176, + "loss": 0.8668, + "step": 7569 + }, + { + "epoch": 0.5252931788217334, + "grad_norm": 4.151389122009277, + "learning_rate": 0.00019466475701281275, + "loss": 0.9954, + "step": 7570 + }, + { + "epoch": 0.52536257025883, + "grad_norm": 2.6639959812164307, + "learning_rate": 0.00019466294614847807, + "loss": 0.8328, + "step": 7571 + }, + { + "epoch": 0.5254319616959268, + "grad_norm": 2.264252185821533, + "learning_rate": 0.0001946611349853035, + "loss": 0.8305, + "step": 7572 + }, + { + "epoch": 0.5255013531330234, + "grad_norm": 2.5847606658935547, + "learning_rate": 0.0001946593235232947, + "loss": 0.7234, + "step": 7573 + }, + { + "epoch": 0.5255707445701201, + "grad_norm": 2.5867559909820557, + "learning_rate": 0.00019465751176245738, + "loss": 1.1121, + "step": 7574 + }, + { + "epoch": 0.5256401360072167, + "grad_norm": 3.05863356590271, + "learning_rate": 0.0001946556997027973, + "loss": 1.1131, + "step": 7575 + }, + { + "epoch": 0.5257095274443133, + "grad_norm": 2.5400900840759277, + "learning_rate": 0.00019465388734432013, + "loss": 0.7166, + "step": 7576 + }, + { + "epoch": 0.52577891888141, + "grad_norm": 3.177849292755127, + "learning_rate": 0.00019465207468703162, + "loss": 1.1289, + "step": 7577 + }, + { + "epoch": 0.5258483103185067, + "grad_norm": 3.4508843421936035, + "learning_rate": 0.0001946502617309375, + "loss": 1.1057, + "step": 7578 + }, + { + "epoch": 0.5259177017556034, + "grad_norm": 2.9122304916381836, + "learning_rate": 0.00019464844847604348, + "loss": 0.9304, + "step": 7579 + }, + { + "epoch": 0.5259870931927, + "grad_norm": 2.599172353744507, + "learning_rate": 0.00019464663492235528, + "loss": 0.826, + "step": 7580 + }, + { + "epoch": 0.5260564846297967, + "grad_norm": 2.557584285736084, + "learning_rate": 0.00019464482106987867, + "loss": 0.9727, + "step": 7581 + }, + { + "epoch": 0.5261258760668933, + "grad_norm": 2.7314555644989014, + "learning_rate": 0.0001946430069186193, + "loss": 0.869, + "step": 7582 + }, + { + "epoch": 0.52619526750399, + "grad_norm": 3.4919567108154297, + "learning_rate": 0.00019464119246858296, + "loss": 1.0373, + "step": 7583 + }, + { + "epoch": 0.5262646589410866, + "grad_norm": 2.903911590576172, + "learning_rate": 0.00019463937771977536, + "loss": 0.9391, + "step": 7584 + }, + { + "epoch": 0.5263340503781834, + "grad_norm": 3.8847498893737793, + "learning_rate": 0.0001946375626722022, + "loss": 1.2028, + "step": 7585 + }, + { + "epoch": 0.52640344181528, + "grad_norm": 2.061387777328491, + "learning_rate": 0.00019463574732586926, + "loss": 0.9962, + "step": 7586 + }, + { + "epoch": 0.5264728332523767, + "grad_norm": 3.1331801414489746, + "learning_rate": 0.00019463393168078227, + "loss": 0.886, + "step": 7587 + }, + { + "epoch": 0.5265422246894733, + "grad_norm": 2.293212890625, + "learning_rate": 0.00019463211573694688, + "loss": 0.9386, + "step": 7588 + }, + { + "epoch": 0.5266116161265699, + "grad_norm": 2.0436556339263916, + "learning_rate": 0.00019463029949436892, + "loss": 0.7115, + "step": 7589 + }, + { + "epoch": 0.5266810075636666, + "grad_norm": 2.0928421020507812, + "learning_rate": 0.00019462848295305406, + "loss": 0.9966, + "step": 7590 + }, + { + "epoch": 0.5267503990007633, + "grad_norm": 2.2158312797546387, + "learning_rate": 0.00019462666611300806, + "loss": 0.904, + "step": 7591 + }, + { + "epoch": 0.52681979043786, + "grad_norm": 1.8523586988449097, + "learning_rate": 0.00019462484897423668, + "loss": 0.9858, + "step": 7592 + }, + { + "epoch": 0.5268891818749566, + "grad_norm": 1.877873182296753, + "learning_rate": 0.00019462303153674562, + "loss": 0.7671, + "step": 7593 + }, + { + "epoch": 0.5269585733120533, + "grad_norm": 2.9518842697143555, + "learning_rate": 0.0001946212138005406, + "loss": 0.8174, + "step": 7594 + }, + { + "epoch": 0.5270279647491499, + "grad_norm": 2.6153416633605957, + "learning_rate": 0.0001946193957656274, + "loss": 0.8177, + "step": 7595 + }, + { + "epoch": 0.5270973561862466, + "grad_norm": 2.4683518409729004, + "learning_rate": 0.00019461757743201178, + "loss": 1.1437, + "step": 7596 + }, + { + "epoch": 0.5271667476233433, + "grad_norm": 2.341864585876465, + "learning_rate": 0.0001946157587996994, + "loss": 0.9269, + "step": 7597 + }, + { + "epoch": 0.52723613906044, + "grad_norm": 1.9682321548461914, + "learning_rate": 0.00019461393986869606, + "loss": 0.8126, + "step": 7598 + }, + { + "epoch": 0.5273055304975366, + "grad_norm": 2.2823143005371094, + "learning_rate": 0.00019461212063900748, + "loss": 0.7272, + "step": 7599 + }, + { + "epoch": 0.5273749219346333, + "grad_norm": 2.194783926010132, + "learning_rate": 0.00019461030111063944, + "loss": 0.7753, + "step": 7600 + }, + { + "epoch": 0.5274443133717299, + "grad_norm": 3.3434770107269287, + "learning_rate": 0.00019460848128359764, + "loss": 0.9788, + "step": 7601 + }, + { + "epoch": 0.5275137048088266, + "grad_norm": 1.910061001777649, + "learning_rate": 0.00019460666115788785, + "loss": 0.8185, + "step": 7602 + }, + { + "epoch": 0.5275830962459233, + "grad_norm": 4.026634216308594, + "learning_rate": 0.0001946048407335158, + "loss": 0.9567, + "step": 7603 + }, + { + "epoch": 0.5276524876830199, + "grad_norm": 3.0786759853363037, + "learning_rate": 0.00019460302001048722, + "loss": 1.2703, + "step": 7604 + }, + { + "epoch": 0.5277218791201166, + "grad_norm": 3.0741043090820312, + "learning_rate": 0.0001946011989888079, + "loss": 0.8111, + "step": 7605 + }, + { + "epoch": 0.5277912705572132, + "grad_norm": 2.924715757369995, + "learning_rate": 0.0001945993776684836, + "loss": 0.887, + "step": 7606 + }, + { + "epoch": 0.5278606619943099, + "grad_norm": 2.452650308609009, + "learning_rate": 0.00019459755604952, + "loss": 0.9365, + "step": 7607 + }, + { + "epoch": 0.5279300534314065, + "grad_norm": 3.106668472290039, + "learning_rate": 0.0001945957341319229, + "loss": 1.0423, + "step": 7608 + }, + { + "epoch": 0.5279994448685033, + "grad_norm": 3.6496407985687256, + "learning_rate": 0.00019459391191569808, + "loss": 0.9863, + "step": 7609 + }, + { + "epoch": 0.5280688363055999, + "grad_norm": 2.965459108352661, + "learning_rate": 0.00019459208940085122, + "loss": 0.9298, + "step": 7610 + }, + { + "epoch": 0.5281382277426966, + "grad_norm": 2.5598714351654053, + "learning_rate": 0.00019459026658738812, + "loss": 0.7778, + "step": 7611 + }, + { + "epoch": 0.5282076191797932, + "grad_norm": 1.66963529586792, + "learning_rate": 0.0001945884434753145, + "loss": 0.7618, + "step": 7612 + }, + { + "epoch": 0.5282770106168899, + "grad_norm": 2.363067388534546, + "learning_rate": 0.00019458662006463615, + "loss": 1.0141, + "step": 7613 + }, + { + "epoch": 0.5283464020539865, + "grad_norm": 3.0720746517181396, + "learning_rate": 0.00019458479635535883, + "loss": 0.8212, + "step": 7614 + }, + { + "epoch": 0.5284157934910833, + "grad_norm": 2.9391772747039795, + "learning_rate": 0.00019458297234748825, + "loss": 1.0898, + "step": 7615 + }, + { + "epoch": 0.5284851849281799, + "grad_norm": 2.122279405593872, + "learning_rate": 0.0001945811480410302, + "loss": 0.7137, + "step": 7616 + }, + { + "epoch": 0.5285545763652765, + "grad_norm": 2.8581771850585938, + "learning_rate": 0.0001945793234359905, + "loss": 0.9871, + "step": 7617 + }, + { + "epoch": 0.5286239678023732, + "grad_norm": 2.4901504516601562, + "learning_rate": 0.0001945774985323748, + "loss": 1.0021, + "step": 7618 + }, + { + "epoch": 0.5286933592394698, + "grad_norm": 2.413785457611084, + "learning_rate": 0.00019457567333018894, + "loss": 0.9389, + "step": 7619 + }, + { + "epoch": 0.5287627506765665, + "grad_norm": 2.492356777191162, + "learning_rate": 0.00019457384782943862, + "loss": 0.7804, + "step": 7620 + }, + { + "epoch": 0.5288321421136631, + "grad_norm": 2.1143617630004883, + "learning_rate": 0.00019457202203012965, + "loss": 0.7675, + "step": 7621 + }, + { + "epoch": 0.5289015335507599, + "grad_norm": 2.2566301822662354, + "learning_rate": 0.00019457019593226775, + "loss": 0.9507, + "step": 7622 + }, + { + "epoch": 0.5289709249878565, + "grad_norm": 2.3311820030212402, + "learning_rate": 0.00019456836953585875, + "loss": 0.8806, + "step": 7623 + }, + { + "epoch": 0.5290403164249532, + "grad_norm": 3.3463313579559326, + "learning_rate": 0.0001945665428409084, + "loss": 0.9614, + "step": 7624 + }, + { + "epoch": 0.5291097078620498, + "grad_norm": 2.4324915409088135, + "learning_rate": 0.00019456471584742237, + "loss": 0.834, + "step": 7625 + }, + { + "epoch": 0.5291790992991465, + "grad_norm": 2.9101901054382324, + "learning_rate": 0.00019456288855540657, + "loss": 0.9703, + "step": 7626 + }, + { + "epoch": 0.5292484907362431, + "grad_norm": 2.6863508224487305, + "learning_rate": 0.00019456106096486667, + "loss": 1.242, + "step": 7627 + }, + { + "epoch": 0.5293178821733399, + "grad_norm": 2.4295151233673096, + "learning_rate": 0.00019455923307580848, + "loss": 0.8064, + "step": 7628 + }, + { + "epoch": 0.5293872736104365, + "grad_norm": 2.329798460006714, + "learning_rate": 0.00019455740488823775, + "loss": 1.0817, + "step": 7629 + }, + { + "epoch": 0.5294566650475331, + "grad_norm": 2.655022144317627, + "learning_rate": 0.00019455557640216028, + "loss": 0.9267, + "step": 7630 + }, + { + "epoch": 0.5295260564846298, + "grad_norm": 2.6881158351898193, + "learning_rate": 0.00019455374761758183, + "loss": 0.9866, + "step": 7631 + }, + { + "epoch": 0.5295954479217264, + "grad_norm": 2.3802335262298584, + "learning_rate": 0.00019455191853450817, + "loss": 0.7515, + "step": 7632 + }, + { + "epoch": 0.5296648393588231, + "grad_norm": 2.972104787826538, + "learning_rate": 0.00019455008915294505, + "loss": 0.9159, + "step": 7633 + }, + { + "epoch": 0.5297342307959197, + "grad_norm": 2.7534031867980957, + "learning_rate": 0.00019454825947289828, + "loss": 0.922, + "step": 7634 + }, + { + "epoch": 0.5298036222330165, + "grad_norm": 2.3375813961029053, + "learning_rate": 0.00019454642949437364, + "loss": 0.9784, + "step": 7635 + }, + { + "epoch": 0.5298730136701131, + "grad_norm": 2.6215310096740723, + "learning_rate": 0.0001945445992173769, + "loss": 0.972, + "step": 7636 + }, + { + "epoch": 0.5299424051072098, + "grad_norm": 2.4067187309265137, + "learning_rate": 0.00019454276864191378, + "loss": 0.9865, + "step": 7637 + }, + { + "epoch": 0.5300117965443064, + "grad_norm": 3.0515518188476562, + "learning_rate": 0.00019454093776799015, + "loss": 1.1908, + "step": 7638 + }, + { + "epoch": 0.5300811879814031, + "grad_norm": 3.0602970123291016, + "learning_rate": 0.00019453910659561173, + "loss": 0.8882, + "step": 7639 + }, + { + "epoch": 0.5301505794184997, + "grad_norm": 2.1330161094665527, + "learning_rate": 0.00019453727512478433, + "loss": 0.8372, + "step": 7640 + }, + { + "epoch": 0.5302199708555965, + "grad_norm": 2.257208824157715, + "learning_rate": 0.0001945354433555137, + "loss": 0.7512, + "step": 7641 + }, + { + "epoch": 0.5302893622926931, + "grad_norm": 2.999885320663452, + "learning_rate": 0.00019453361128780565, + "loss": 0.8664, + "step": 7642 + }, + { + "epoch": 0.5303587537297897, + "grad_norm": 2.470052719116211, + "learning_rate": 0.00019453177892166598, + "loss": 1.0501, + "step": 7643 + }, + { + "epoch": 0.5304281451668864, + "grad_norm": 2.444990634918213, + "learning_rate": 0.00019452994625710043, + "loss": 0.7743, + "step": 7644 + }, + { + "epoch": 0.530497536603983, + "grad_norm": 2.0908665657043457, + "learning_rate": 0.00019452811329411483, + "loss": 0.8562, + "step": 7645 + }, + { + "epoch": 0.5305669280410797, + "grad_norm": 3.278409242630005, + "learning_rate": 0.0001945262800327149, + "loss": 0.9444, + "step": 7646 + }, + { + "epoch": 0.5306363194781764, + "grad_norm": 2.3237087726593018, + "learning_rate": 0.00019452444647290653, + "loss": 0.8034, + "step": 7647 + }, + { + "epoch": 0.5307057109152731, + "grad_norm": 2.6049766540527344, + "learning_rate": 0.00019452261261469541, + "loss": 0.8794, + "step": 7648 + }, + { + "epoch": 0.5307751023523697, + "grad_norm": 2.670614004135132, + "learning_rate": 0.00019452077845808735, + "loss": 0.9151, + "step": 7649 + }, + { + "epoch": 0.5308444937894664, + "grad_norm": 2.9469950199127197, + "learning_rate": 0.0001945189440030882, + "loss": 1.1085, + "step": 7650 + }, + { + "epoch": 0.530913885226563, + "grad_norm": 2.361370801925659, + "learning_rate": 0.0001945171092497037, + "loss": 0.7746, + "step": 7651 + }, + { + "epoch": 0.5309832766636597, + "grad_norm": 3.4757893085479736, + "learning_rate": 0.00019451527419793964, + "loss": 1.0591, + "step": 7652 + }, + { + "epoch": 0.5310526681007564, + "grad_norm": 2.4777960777282715, + "learning_rate": 0.00019451343884780185, + "loss": 0.9072, + "step": 7653 + }, + { + "epoch": 0.5311220595378531, + "grad_norm": 2.376023530960083, + "learning_rate": 0.0001945116031992961, + "loss": 0.7452, + "step": 7654 + }, + { + "epoch": 0.5311914509749497, + "grad_norm": 2.790550947189331, + "learning_rate": 0.00019450976725242814, + "loss": 0.9564, + "step": 7655 + }, + { + "epoch": 0.5312608424120464, + "grad_norm": 2.247969627380371, + "learning_rate": 0.00019450793100720383, + "loss": 0.7819, + "step": 7656 + }, + { + "epoch": 0.531330233849143, + "grad_norm": 2.2937653064727783, + "learning_rate": 0.00019450609446362895, + "loss": 0.9002, + "step": 7657 + }, + { + "epoch": 0.5313996252862396, + "grad_norm": 2.0651538372039795, + "learning_rate": 0.00019450425762170933, + "loss": 0.6544, + "step": 7658 + }, + { + "epoch": 0.5314690167233364, + "grad_norm": 2.021460771560669, + "learning_rate": 0.0001945024204814507, + "loss": 0.728, + "step": 7659 + }, + { + "epoch": 0.531538408160433, + "grad_norm": 3.8441238403320312, + "learning_rate": 0.00019450058304285888, + "loss": 1.1542, + "step": 7660 + }, + { + "epoch": 0.5316077995975297, + "grad_norm": 3.397700071334839, + "learning_rate": 0.0001944987453059397, + "loss": 1.2255, + "step": 7661 + }, + { + "epoch": 0.5316771910346263, + "grad_norm": 2.376295328140259, + "learning_rate": 0.00019449690727069895, + "loss": 0.8716, + "step": 7662 + }, + { + "epoch": 0.531746582471723, + "grad_norm": 3.0310006141662598, + "learning_rate": 0.0001944950689371424, + "loss": 0.9117, + "step": 7663 + }, + { + "epoch": 0.5318159739088196, + "grad_norm": 2.869784116744995, + "learning_rate": 0.00019449323030527592, + "loss": 0.729, + "step": 7664 + }, + { + "epoch": 0.5318853653459163, + "grad_norm": 3.769583225250244, + "learning_rate": 0.00019449139137510524, + "loss": 1.0162, + "step": 7665 + }, + { + "epoch": 0.531954756783013, + "grad_norm": 3.533191680908203, + "learning_rate": 0.0001944895521466362, + "loss": 0.6512, + "step": 7666 + }, + { + "epoch": 0.5320241482201097, + "grad_norm": 2.283001184463501, + "learning_rate": 0.00019448771261987465, + "loss": 0.9634, + "step": 7667 + }, + { + "epoch": 0.5320935396572063, + "grad_norm": 3.17779803276062, + "learning_rate": 0.0001944858727948263, + "loss": 1.1038, + "step": 7668 + }, + { + "epoch": 0.532162931094303, + "grad_norm": 2.361614942550659, + "learning_rate": 0.00019448403267149705, + "loss": 0.8115, + "step": 7669 + }, + { + "epoch": 0.5322323225313996, + "grad_norm": 2.7022178173065186, + "learning_rate": 0.00019448219224989265, + "loss": 0.938, + "step": 7670 + }, + { + "epoch": 0.5323017139684962, + "grad_norm": 2.5864737033843994, + "learning_rate": 0.00019448035153001896, + "loss": 0.711, + "step": 7671 + }, + { + "epoch": 0.532371105405593, + "grad_norm": 3.813140392303467, + "learning_rate": 0.00019447851051188173, + "loss": 0.9183, + "step": 7672 + }, + { + "epoch": 0.5324404968426896, + "grad_norm": 3.012779712677002, + "learning_rate": 0.00019447666919548682, + "loss": 1.0876, + "step": 7673 + }, + { + "epoch": 0.5325098882797863, + "grad_norm": 3.2422401905059814, + "learning_rate": 0.00019447482758084005, + "loss": 0.741, + "step": 7674 + }, + { + "epoch": 0.5325792797168829, + "grad_norm": 3.07279109954834, + "learning_rate": 0.00019447298566794717, + "loss": 0.8023, + "step": 7675 + }, + { + "epoch": 0.5326486711539796, + "grad_norm": 2.7146506309509277, + "learning_rate": 0.00019447114345681404, + "loss": 0.9809, + "step": 7676 + }, + { + "epoch": 0.5327180625910762, + "grad_norm": 2.0960071086883545, + "learning_rate": 0.0001944693009474465, + "loss": 0.8573, + "step": 7677 + }, + { + "epoch": 0.532787454028173, + "grad_norm": 3.1646535396575928, + "learning_rate": 0.00019446745813985028, + "loss": 1.056, + "step": 7678 + }, + { + "epoch": 0.5328568454652696, + "grad_norm": 3.04921293258667, + "learning_rate": 0.0001944656150340313, + "loss": 0.9862, + "step": 7679 + }, + { + "epoch": 0.5329262369023663, + "grad_norm": 2.263594150543213, + "learning_rate": 0.00019446377162999533, + "loss": 0.8893, + "step": 7680 + }, + { + "epoch": 0.5329956283394629, + "grad_norm": 3.348217010498047, + "learning_rate": 0.0001944619279277482, + "loss": 1.0173, + "step": 7681 + }, + { + "epoch": 0.5330650197765596, + "grad_norm": 2.705122947692871, + "learning_rate": 0.0001944600839272957, + "loss": 0.8972, + "step": 7682 + }, + { + "epoch": 0.5331344112136562, + "grad_norm": 2.1310369968414307, + "learning_rate": 0.00019445823962864367, + "loss": 0.7611, + "step": 7683 + }, + { + "epoch": 0.5332038026507528, + "grad_norm": 3.0449156761169434, + "learning_rate": 0.00019445639503179796, + "loss": 1.19, + "step": 7684 + }, + { + "epoch": 0.5332731940878496, + "grad_norm": 2.4576282501220703, + "learning_rate": 0.00019445455013676437, + "loss": 0.832, + "step": 7685 + }, + { + "epoch": 0.5333425855249462, + "grad_norm": 3.8738107681274414, + "learning_rate": 0.00019445270494354873, + "loss": 1.1034, + "step": 7686 + }, + { + "epoch": 0.5334119769620429, + "grad_norm": 2.810328245162964, + "learning_rate": 0.00019445085945215683, + "loss": 0.8964, + "step": 7687 + }, + { + "epoch": 0.5334813683991395, + "grad_norm": 3.0258162021636963, + "learning_rate": 0.00019444901366259454, + "loss": 0.9516, + "step": 7688 + }, + { + "epoch": 0.5335507598362362, + "grad_norm": 2.413644313812256, + "learning_rate": 0.00019444716757486768, + "loss": 0.7778, + "step": 7689 + }, + { + "epoch": 0.5336201512733328, + "grad_norm": 2.1139397621154785, + "learning_rate": 0.00019444532118898205, + "loss": 0.6514, + "step": 7690 + }, + { + "epoch": 0.5336895427104296, + "grad_norm": 2.905470132827759, + "learning_rate": 0.0001944434745049435, + "loss": 0.9488, + "step": 7691 + }, + { + "epoch": 0.5337589341475262, + "grad_norm": 3.1516170501708984, + "learning_rate": 0.00019444162752275788, + "loss": 1.256, + "step": 7692 + }, + { + "epoch": 0.5338283255846229, + "grad_norm": 2.918811798095703, + "learning_rate": 0.00019443978024243097, + "loss": 1.1555, + "step": 7693 + }, + { + "epoch": 0.5338977170217195, + "grad_norm": 2.907925605773926, + "learning_rate": 0.00019443793266396865, + "loss": 0.8126, + "step": 7694 + }, + { + "epoch": 0.5339671084588162, + "grad_norm": 2.4061009883880615, + "learning_rate": 0.00019443608478737672, + "loss": 0.75, + "step": 7695 + }, + { + "epoch": 0.5340364998959128, + "grad_norm": 1.7133365869522095, + "learning_rate": 0.00019443423661266108, + "loss": 0.6793, + "step": 7696 + }, + { + "epoch": 0.5341058913330095, + "grad_norm": 2.5282340049743652, + "learning_rate": 0.00019443238813982746, + "loss": 0.7958, + "step": 7697 + }, + { + "epoch": 0.5341752827701062, + "grad_norm": 2.7589340209960938, + "learning_rate": 0.00019443053936888177, + "loss": 0.8848, + "step": 7698 + }, + { + "epoch": 0.5342446742072028, + "grad_norm": 2.7769033908843994, + "learning_rate": 0.0001944286902998298, + "loss": 0.8884, + "step": 7699 + }, + { + "epoch": 0.5343140656442995, + "grad_norm": 3.188004732131958, + "learning_rate": 0.0001944268409326774, + "loss": 1.2692, + "step": 7700 + }, + { + "epoch": 0.5343834570813961, + "grad_norm": 2.876664400100708, + "learning_rate": 0.00019442499126743045, + "loss": 0.8526, + "step": 7701 + }, + { + "epoch": 0.5344528485184928, + "grad_norm": 3.9547953605651855, + "learning_rate": 0.00019442314130409475, + "loss": 0.7681, + "step": 7702 + }, + { + "epoch": 0.5345222399555895, + "grad_norm": 3.2827231884002686, + "learning_rate": 0.00019442129104267614, + "loss": 0.9656, + "step": 7703 + }, + { + "epoch": 0.5345916313926862, + "grad_norm": 3.3565337657928467, + "learning_rate": 0.00019441944048318049, + "loss": 0.7546, + "step": 7704 + }, + { + "epoch": 0.5346610228297828, + "grad_norm": 3.382117986679077, + "learning_rate": 0.0001944175896256136, + "loss": 0.8133, + "step": 7705 + }, + { + "epoch": 0.5347304142668795, + "grad_norm": 2.72353196144104, + "learning_rate": 0.00019441573846998135, + "loss": 1.1732, + "step": 7706 + }, + { + "epoch": 0.5347998057039761, + "grad_norm": 3.5603630542755127, + "learning_rate": 0.00019441388701628954, + "loss": 1.0048, + "step": 7707 + }, + { + "epoch": 0.5348691971410728, + "grad_norm": 2.6676156520843506, + "learning_rate": 0.0001944120352645441, + "loss": 0.8434, + "step": 7708 + }, + { + "epoch": 0.5349385885781694, + "grad_norm": 4.287142753601074, + "learning_rate": 0.00019441018321475072, + "loss": 0.8265, + "step": 7709 + }, + { + "epoch": 0.5350079800152662, + "grad_norm": 2.989370107650757, + "learning_rate": 0.00019440833086691544, + "loss": 0.9387, + "step": 7710 + }, + { + "epoch": 0.5350773714523628, + "grad_norm": 6.085770606994629, + "learning_rate": 0.00019440647822104398, + "loss": 0.9195, + "step": 7711 + }, + { + "epoch": 0.5351467628894594, + "grad_norm": 2.870055675506592, + "learning_rate": 0.0001944046252771422, + "loss": 1.1451, + "step": 7712 + }, + { + "epoch": 0.5352161543265561, + "grad_norm": 2.8327224254608154, + "learning_rate": 0.000194402772035216, + "loss": 1.0075, + "step": 7713 + }, + { + "epoch": 0.5352855457636527, + "grad_norm": 3.7606379985809326, + "learning_rate": 0.00019440091849527117, + "loss": 1.398, + "step": 7714 + }, + { + "epoch": 0.5353549372007494, + "grad_norm": 2.397773504257202, + "learning_rate": 0.00019439906465731362, + "loss": 0.7237, + "step": 7715 + }, + { + "epoch": 0.5354243286378461, + "grad_norm": 3.4332170486450195, + "learning_rate": 0.00019439721052134918, + "loss": 0.9997, + "step": 7716 + }, + { + "epoch": 0.5354937200749428, + "grad_norm": 2.3667168617248535, + "learning_rate": 0.00019439535608738368, + "loss": 1.0325, + "step": 7717 + }, + { + "epoch": 0.5355631115120394, + "grad_norm": 2.8789374828338623, + "learning_rate": 0.000194393501355423, + "loss": 0.889, + "step": 7718 + }, + { + "epoch": 0.5356325029491361, + "grad_norm": 2.39511775970459, + "learning_rate": 0.000194391646325473, + "loss": 0.6977, + "step": 7719 + }, + { + "epoch": 0.5357018943862327, + "grad_norm": 3.2426021099090576, + "learning_rate": 0.00019438979099753952, + "loss": 0.915, + "step": 7720 + }, + { + "epoch": 0.5357712858233294, + "grad_norm": 3.279604911804199, + "learning_rate": 0.0001943879353716284, + "loss": 0.9462, + "step": 7721 + }, + { + "epoch": 0.5358406772604261, + "grad_norm": 2.4969167709350586, + "learning_rate": 0.00019438607944774553, + "loss": 0.8404, + "step": 7722 + }, + { + "epoch": 0.5359100686975228, + "grad_norm": 2.9066929817199707, + "learning_rate": 0.00019438422322589678, + "loss": 0.7781, + "step": 7723 + }, + { + "epoch": 0.5359794601346194, + "grad_norm": 2.5534496307373047, + "learning_rate": 0.00019438236670608796, + "loss": 0.6533, + "step": 7724 + }, + { + "epoch": 0.536048851571716, + "grad_norm": 3.543433904647827, + "learning_rate": 0.00019438050988832494, + "loss": 0.8952, + "step": 7725 + }, + { + "epoch": 0.5361182430088127, + "grad_norm": 3.29292893409729, + "learning_rate": 0.00019437865277261365, + "loss": 1.1763, + "step": 7726 + }, + { + "epoch": 0.5361876344459093, + "grad_norm": 2.191600799560547, + "learning_rate": 0.00019437679535895988, + "loss": 0.8075, + "step": 7727 + }, + { + "epoch": 0.536257025883006, + "grad_norm": 2.4629952907562256, + "learning_rate": 0.0001943749376473695, + "loss": 0.7622, + "step": 7728 + }, + { + "epoch": 0.5363264173201027, + "grad_norm": 2.680006980895996, + "learning_rate": 0.00019437307963784842, + "loss": 0.8093, + "step": 7729 + }, + { + "epoch": 0.5363958087571994, + "grad_norm": 2.7114672660827637, + "learning_rate": 0.00019437122133040245, + "loss": 0.8871, + "step": 7730 + }, + { + "epoch": 0.536465200194296, + "grad_norm": 4.120847225189209, + "learning_rate": 0.00019436936272503751, + "loss": 0.9623, + "step": 7731 + }, + { + "epoch": 0.5365345916313927, + "grad_norm": 3.336047887802124, + "learning_rate": 0.00019436750382175944, + "loss": 0.8537, + "step": 7732 + }, + { + "epoch": 0.5366039830684893, + "grad_norm": 3.089350700378418, + "learning_rate": 0.0001943656446205741, + "loss": 0.9759, + "step": 7733 + }, + { + "epoch": 0.536673374505586, + "grad_norm": 2.440704822540283, + "learning_rate": 0.00019436378512148735, + "loss": 0.8683, + "step": 7734 + }, + { + "epoch": 0.5367427659426827, + "grad_norm": 2.7879931926727295, + "learning_rate": 0.0001943619253245051, + "loss": 0.9108, + "step": 7735 + }, + { + "epoch": 0.5368121573797794, + "grad_norm": 2.1675496101379395, + "learning_rate": 0.00019436006522963322, + "loss": 0.9435, + "step": 7736 + }, + { + "epoch": 0.536881548816876, + "grad_norm": 2.747358798980713, + "learning_rate": 0.00019435820483687752, + "loss": 0.9636, + "step": 7737 + }, + { + "epoch": 0.5369509402539726, + "grad_norm": 2.908931255340576, + "learning_rate": 0.00019435634414624394, + "loss": 0.9711, + "step": 7738 + }, + { + "epoch": 0.5370203316910693, + "grad_norm": 2.51997971534729, + "learning_rate": 0.0001943544831577383, + "loss": 0.9744, + "step": 7739 + }, + { + "epoch": 0.5370897231281659, + "grad_norm": 4.0819220542907715, + "learning_rate": 0.00019435262187136653, + "loss": 1.1063, + "step": 7740 + }, + { + "epoch": 0.5371591145652627, + "grad_norm": 3.967824935913086, + "learning_rate": 0.00019435076028713447, + "loss": 1.038, + "step": 7741 + }, + { + "epoch": 0.5372285060023593, + "grad_norm": 3.452218770980835, + "learning_rate": 0.00019434889840504803, + "loss": 1.1424, + "step": 7742 + }, + { + "epoch": 0.537297897439456, + "grad_norm": 2.145859718322754, + "learning_rate": 0.00019434703622511304, + "loss": 0.9353, + "step": 7743 + }, + { + "epoch": 0.5373672888765526, + "grad_norm": 3.6096274852752686, + "learning_rate": 0.0001943451737473354, + "loss": 0.9259, + "step": 7744 + }, + { + "epoch": 0.5374366803136493, + "grad_norm": 2.5784857273101807, + "learning_rate": 0.000194343310971721, + "loss": 0.7771, + "step": 7745 + }, + { + "epoch": 0.5375060717507459, + "grad_norm": 2.927888870239258, + "learning_rate": 0.00019434144789827573, + "loss": 1.2456, + "step": 7746 + }, + { + "epoch": 0.5375754631878427, + "grad_norm": 3.130831003189087, + "learning_rate": 0.00019433958452700544, + "loss": 1.0954, + "step": 7747 + }, + { + "epoch": 0.5376448546249393, + "grad_norm": 2.3742833137512207, + "learning_rate": 0.00019433772085791602, + "loss": 0.954, + "step": 7748 + }, + { + "epoch": 0.537714246062036, + "grad_norm": 3.1227312088012695, + "learning_rate": 0.00019433585689101337, + "loss": 1.0734, + "step": 7749 + }, + { + "epoch": 0.5377836374991326, + "grad_norm": 2.6472623348236084, + "learning_rate": 0.00019433399262630335, + "loss": 0.7795, + "step": 7750 + }, + { + "epoch": 0.5378530289362292, + "grad_norm": 2.745694875717163, + "learning_rate": 0.00019433212806379188, + "loss": 1.0672, + "step": 7751 + }, + { + "epoch": 0.5379224203733259, + "grad_norm": 3.458644151687622, + "learning_rate": 0.00019433026320348482, + "loss": 0.9647, + "step": 7752 + }, + { + "epoch": 0.5379918118104225, + "grad_norm": 2.5226314067840576, + "learning_rate": 0.00019432839804538806, + "loss": 0.8629, + "step": 7753 + }, + { + "epoch": 0.5380612032475193, + "grad_norm": 3.781881809234619, + "learning_rate": 0.00019432653258950747, + "loss": 1.0201, + "step": 7754 + }, + { + "epoch": 0.5381305946846159, + "grad_norm": 3.9420957565307617, + "learning_rate": 0.00019432466683584897, + "loss": 0.8869, + "step": 7755 + }, + { + "epoch": 0.5381999861217126, + "grad_norm": 2.686047077178955, + "learning_rate": 0.00019432280078441849, + "loss": 1.0456, + "step": 7756 + }, + { + "epoch": 0.5382693775588092, + "grad_norm": 4.106922626495361, + "learning_rate": 0.00019432093443522182, + "loss": 0.9544, + "step": 7757 + }, + { + "epoch": 0.5383387689959059, + "grad_norm": 3.1368770599365234, + "learning_rate": 0.00019431906778826492, + "loss": 1.0155, + "step": 7758 + }, + { + "epoch": 0.5384081604330025, + "grad_norm": 2.430065870285034, + "learning_rate": 0.00019431720084355365, + "loss": 0.9009, + "step": 7759 + }, + { + "epoch": 0.5384775518700993, + "grad_norm": 4.229537487030029, + "learning_rate": 0.00019431533360109393, + "loss": 1.2089, + "step": 7760 + }, + { + "epoch": 0.5385469433071959, + "grad_norm": 2.6668999195098877, + "learning_rate": 0.00019431346606089167, + "loss": 0.9584, + "step": 7761 + }, + { + "epoch": 0.5386163347442926, + "grad_norm": 2.771172285079956, + "learning_rate": 0.00019431159822295272, + "loss": 0.8581, + "step": 7762 + }, + { + "epoch": 0.5386857261813892, + "grad_norm": 3.326586961746216, + "learning_rate": 0.00019430973008728296, + "loss": 0.972, + "step": 7763 + }, + { + "epoch": 0.5387551176184858, + "grad_norm": 2.670565128326416, + "learning_rate": 0.00019430786165388838, + "loss": 0.7479, + "step": 7764 + }, + { + "epoch": 0.5388245090555825, + "grad_norm": 3.141228199005127, + "learning_rate": 0.0001943059929227748, + "loss": 0.9549, + "step": 7765 + }, + { + "epoch": 0.5388939004926792, + "grad_norm": 2.5879781246185303, + "learning_rate": 0.00019430412389394815, + "loss": 0.752, + "step": 7766 + }, + { + "epoch": 0.5389632919297759, + "grad_norm": 2.687286853790283, + "learning_rate": 0.0001943022545674143, + "loss": 0.7423, + "step": 7767 + }, + { + "epoch": 0.5390326833668725, + "grad_norm": 3.131593704223633, + "learning_rate": 0.00019430038494317923, + "loss": 0.9661, + "step": 7768 + }, + { + "epoch": 0.5391020748039692, + "grad_norm": 2.5852601528167725, + "learning_rate": 0.0001942985150212487, + "loss": 0.9103, + "step": 7769 + }, + { + "epoch": 0.5391714662410658, + "grad_norm": 3.1291000843048096, + "learning_rate": 0.00019429664480162876, + "loss": 1.4282, + "step": 7770 + }, + { + "epoch": 0.5392408576781625, + "grad_norm": 3.2797398567199707, + "learning_rate": 0.00019429477428432523, + "loss": 0.9464, + "step": 7771 + }, + { + "epoch": 0.5393102491152592, + "grad_norm": 2.634031295776367, + "learning_rate": 0.00019429290346934406, + "loss": 0.8112, + "step": 7772 + }, + { + "epoch": 0.5393796405523559, + "grad_norm": 3.908417224884033, + "learning_rate": 0.0001942910323566911, + "loss": 1.0091, + "step": 7773 + }, + { + "epoch": 0.5394490319894525, + "grad_norm": 2.4465370178222656, + "learning_rate": 0.0001942891609463723, + "loss": 0.9013, + "step": 7774 + }, + { + "epoch": 0.5395184234265492, + "grad_norm": 3.047271966934204, + "learning_rate": 0.00019428728923839356, + "loss": 1.3151, + "step": 7775 + }, + { + "epoch": 0.5395878148636458, + "grad_norm": 2.2495415210723877, + "learning_rate": 0.00019428541723276078, + "loss": 0.6795, + "step": 7776 + }, + { + "epoch": 0.5396572063007425, + "grad_norm": 1.987898349761963, + "learning_rate": 0.00019428354492947994, + "loss": 0.7023, + "step": 7777 + }, + { + "epoch": 0.5397265977378392, + "grad_norm": 2.603764057159424, + "learning_rate": 0.00019428167232855682, + "loss": 0.656, + "step": 7778 + }, + { + "epoch": 0.5397959891749358, + "grad_norm": 3.875516176223755, + "learning_rate": 0.0001942797994299974, + "loss": 1.2489, + "step": 7779 + }, + { + "epoch": 0.5398653806120325, + "grad_norm": 2.778473138809204, + "learning_rate": 0.0001942779262338076, + "loss": 0.7864, + "step": 7780 + }, + { + "epoch": 0.5399347720491291, + "grad_norm": 2.6339149475097656, + "learning_rate": 0.00019427605273999334, + "loss": 0.9707, + "step": 7781 + }, + { + "epoch": 0.5400041634862258, + "grad_norm": 2.176767349243164, + "learning_rate": 0.00019427417894856054, + "loss": 0.8556, + "step": 7782 + }, + { + "epoch": 0.5400735549233224, + "grad_norm": 2.395383834838867, + "learning_rate": 0.00019427230485951507, + "loss": 0.767, + "step": 7783 + }, + { + "epoch": 0.5401429463604192, + "grad_norm": 1.988965392112732, + "learning_rate": 0.00019427043047286286, + "loss": 0.6308, + "step": 7784 + }, + { + "epoch": 0.5402123377975158, + "grad_norm": 2.1796529293060303, + "learning_rate": 0.00019426855578860984, + "loss": 0.9897, + "step": 7785 + }, + { + "epoch": 0.5402817292346125, + "grad_norm": 2.5869576930999756, + "learning_rate": 0.00019426668080676194, + "loss": 0.6138, + "step": 7786 + }, + { + "epoch": 0.5403511206717091, + "grad_norm": 2.8388516902923584, + "learning_rate": 0.00019426480552732505, + "loss": 0.9795, + "step": 7787 + }, + { + "epoch": 0.5404205121088058, + "grad_norm": 2.9657790660858154, + "learning_rate": 0.00019426292995030513, + "loss": 0.9444, + "step": 7788 + }, + { + "epoch": 0.5404899035459024, + "grad_norm": 3.0173704624176025, + "learning_rate": 0.00019426105407570804, + "loss": 1.1236, + "step": 7789 + }, + { + "epoch": 0.5405592949829992, + "grad_norm": 2.8750689029693604, + "learning_rate": 0.00019425917790353978, + "loss": 1.0309, + "step": 7790 + }, + { + "epoch": 0.5406286864200958, + "grad_norm": 2.693866491317749, + "learning_rate": 0.0001942573014338062, + "loss": 1.1304, + "step": 7791 + }, + { + "epoch": 0.5406980778571924, + "grad_norm": 2.41001033782959, + "learning_rate": 0.00019425542466651327, + "loss": 0.8406, + "step": 7792 + }, + { + "epoch": 0.5407674692942891, + "grad_norm": 2.5309627056121826, + "learning_rate": 0.00019425354760166692, + "loss": 0.8655, + "step": 7793 + }, + { + "epoch": 0.5408368607313857, + "grad_norm": 2.3589675426483154, + "learning_rate": 0.000194251670239273, + "loss": 1.0441, + "step": 7794 + }, + { + "epoch": 0.5409062521684824, + "grad_norm": 2.721423864364624, + "learning_rate": 0.00019424979257933752, + "loss": 1.0634, + "step": 7795 + }, + { + "epoch": 0.540975643605579, + "grad_norm": 2.4606497287750244, + "learning_rate": 0.0001942479146218664, + "loss": 0.9182, + "step": 7796 + }, + { + "epoch": 0.5410450350426758, + "grad_norm": 2.273066759109497, + "learning_rate": 0.00019424603636686554, + "loss": 1.082, + "step": 7797 + }, + { + "epoch": 0.5411144264797724, + "grad_norm": 1.9442390203475952, + "learning_rate": 0.00019424415781434088, + "loss": 0.7261, + "step": 7798 + }, + { + "epoch": 0.5411838179168691, + "grad_norm": 1.9194227457046509, + "learning_rate": 0.00019424227896429832, + "loss": 0.7624, + "step": 7799 + }, + { + "epoch": 0.5412532093539657, + "grad_norm": 2.510565757751465, + "learning_rate": 0.00019424039981674388, + "loss": 1.0982, + "step": 7800 + }, + { + "epoch": 0.5413226007910624, + "grad_norm": 2.073647975921631, + "learning_rate": 0.00019423852037168336, + "loss": 0.816, + "step": 7801 + }, + { + "epoch": 0.541391992228159, + "grad_norm": 2.3484559059143066, + "learning_rate": 0.0001942366406291228, + "loss": 0.6999, + "step": 7802 + }, + { + "epoch": 0.5414613836652558, + "grad_norm": 2.3726119995117188, + "learning_rate": 0.0001942347605890681, + "loss": 1.0349, + "step": 7803 + }, + { + "epoch": 0.5415307751023524, + "grad_norm": 2.953920364379883, + "learning_rate": 0.0001942328802515252, + "loss": 0.9368, + "step": 7804 + }, + { + "epoch": 0.541600166539449, + "grad_norm": 2.6581671237945557, + "learning_rate": 0.0001942309996165, + "loss": 0.8429, + "step": 7805 + }, + { + "epoch": 0.5416695579765457, + "grad_norm": 3.1442205905914307, + "learning_rate": 0.0001942291186839985, + "loss": 0.9608, + "step": 7806 + }, + { + "epoch": 0.5417389494136423, + "grad_norm": 2.33958101272583, + "learning_rate": 0.00019422723745402657, + "loss": 0.872, + "step": 7807 + }, + { + "epoch": 0.541808340850739, + "grad_norm": 2.5156781673431396, + "learning_rate": 0.00019422535592659022, + "loss": 0.8881, + "step": 7808 + }, + { + "epoch": 0.5418777322878356, + "grad_norm": 2.6505181789398193, + "learning_rate": 0.00019422347410169533, + "loss": 0.9861, + "step": 7809 + }, + { + "epoch": 0.5419471237249324, + "grad_norm": 2.0571634769439697, + "learning_rate": 0.00019422159197934788, + "loss": 0.7775, + "step": 7810 + }, + { + "epoch": 0.542016515162029, + "grad_norm": 2.7676327228546143, + "learning_rate": 0.00019421970955955378, + "loss": 1.1194, + "step": 7811 + }, + { + "epoch": 0.5420859065991257, + "grad_norm": 2.925877332687378, + "learning_rate": 0.000194217826842319, + "loss": 0.819, + "step": 7812 + }, + { + "epoch": 0.5421552980362223, + "grad_norm": 2.1250057220458984, + "learning_rate": 0.00019421594382764944, + "loss": 0.6534, + "step": 7813 + }, + { + "epoch": 0.542224689473319, + "grad_norm": 4.593454360961914, + "learning_rate": 0.00019421406051555113, + "loss": 1.1459, + "step": 7814 + }, + { + "epoch": 0.5422940809104156, + "grad_norm": 2.2989978790283203, + "learning_rate": 0.00019421217690602992, + "loss": 0.8161, + "step": 7815 + }, + { + "epoch": 0.5423634723475124, + "grad_norm": 3.291973114013672, + "learning_rate": 0.00019421029299909184, + "loss": 0.8391, + "step": 7816 + }, + { + "epoch": 0.542432863784609, + "grad_norm": 2.1516072750091553, + "learning_rate": 0.00019420840879474276, + "loss": 0.8495, + "step": 7817 + }, + { + "epoch": 0.5425022552217056, + "grad_norm": 2.5226187705993652, + "learning_rate": 0.00019420652429298868, + "loss": 0.903, + "step": 7818 + }, + { + "epoch": 0.5425716466588023, + "grad_norm": 3.304164171218872, + "learning_rate": 0.0001942046394938355, + "loss": 0.9388, + "step": 7819 + }, + { + "epoch": 0.5426410380958989, + "grad_norm": 2.6653974056243896, + "learning_rate": 0.00019420275439728925, + "loss": 1.203, + "step": 7820 + }, + { + "epoch": 0.5427104295329956, + "grad_norm": 4.335971355438232, + "learning_rate": 0.00019420086900335582, + "loss": 1.0505, + "step": 7821 + }, + { + "epoch": 0.5427798209700923, + "grad_norm": 2.328871250152588, + "learning_rate": 0.00019419898331204117, + "loss": 0.7956, + "step": 7822 + }, + { + "epoch": 0.542849212407189, + "grad_norm": 11.247387886047363, + "learning_rate": 0.00019419709732335124, + "loss": 0.9881, + "step": 7823 + }, + { + "epoch": 0.5429186038442856, + "grad_norm": 2.6078240871429443, + "learning_rate": 0.00019419521103729203, + "loss": 0.855, + "step": 7824 + }, + { + "epoch": 0.5429879952813823, + "grad_norm": 2.5710628032684326, + "learning_rate": 0.00019419332445386945, + "loss": 0.9001, + "step": 7825 + }, + { + "epoch": 0.5430573867184789, + "grad_norm": 7.087853908538818, + "learning_rate": 0.0001941914375730895, + "loss": 0.869, + "step": 7826 + }, + { + "epoch": 0.5431267781555756, + "grad_norm": 2.4258391857147217, + "learning_rate": 0.0001941895503949581, + "loss": 0.8491, + "step": 7827 + }, + { + "epoch": 0.5431961695926723, + "grad_norm": 3.5351312160491943, + "learning_rate": 0.0001941876629194812, + "loss": 0.9963, + "step": 7828 + }, + { + "epoch": 0.543265561029769, + "grad_norm": 2.9824984073638916, + "learning_rate": 0.00019418577514666478, + "loss": 0.6947, + "step": 7829 + }, + { + "epoch": 0.5433349524668656, + "grad_norm": 2.78938627243042, + "learning_rate": 0.0001941838870765148, + "loss": 0.9531, + "step": 7830 + }, + { + "epoch": 0.5434043439039623, + "grad_norm": 3.4643287658691406, + "learning_rate": 0.0001941819987090372, + "loss": 0.7572, + "step": 7831 + }, + { + "epoch": 0.5434737353410589, + "grad_norm": 2.379114866256714, + "learning_rate": 0.00019418011004423798, + "loss": 0.9649, + "step": 7832 + }, + { + "epoch": 0.5435431267781555, + "grad_norm": 4.368305683135986, + "learning_rate": 0.00019417822108212308, + "loss": 0.9704, + "step": 7833 + }, + { + "epoch": 0.5436125182152522, + "grad_norm": 2.745948076248169, + "learning_rate": 0.00019417633182269845, + "loss": 0.9234, + "step": 7834 + }, + { + "epoch": 0.5436819096523489, + "grad_norm": 3.0490429401397705, + "learning_rate": 0.00019417444226597005, + "loss": 0.9188, + "step": 7835 + }, + { + "epoch": 0.5437513010894456, + "grad_norm": 2.990180730819702, + "learning_rate": 0.0001941725524119439, + "loss": 1.0003, + "step": 7836 + }, + { + "epoch": 0.5438206925265422, + "grad_norm": 2.8272175788879395, + "learning_rate": 0.00019417066226062592, + "loss": 0.9495, + "step": 7837 + }, + { + "epoch": 0.5438900839636389, + "grad_norm": 2.5709452629089355, + "learning_rate": 0.00019416877181202208, + "loss": 0.7603, + "step": 7838 + }, + { + "epoch": 0.5439594754007355, + "grad_norm": 2.4120564460754395, + "learning_rate": 0.00019416688106613835, + "loss": 0.9693, + "step": 7839 + }, + { + "epoch": 0.5440288668378322, + "grad_norm": 2.2156929969787598, + "learning_rate": 0.0001941649900229807, + "loss": 0.9193, + "step": 7840 + }, + { + "epoch": 0.5440982582749289, + "grad_norm": 2.4349217414855957, + "learning_rate": 0.00019416309868255512, + "loss": 0.9147, + "step": 7841 + }, + { + "epoch": 0.5441676497120256, + "grad_norm": 3.1563072204589844, + "learning_rate": 0.00019416120704486754, + "loss": 0.8109, + "step": 7842 + }, + { + "epoch": 0.5442370411491222, + "grad_norm": 2.3939924240112305, + "learning_rate": 0.00019415931510992395, + "loss": 1.0694, + "step": 7843 + }, + { + "epoch": 0.5443064325862189, + "grad_norm": 2.3467063903808594, + "learning_rate": 0.00019415742287773034, + "loss": 0.7453, + "step": 7844 + }, + { + "epoch": 0.5443758240233155, + "grad_norm": 2.695704460144043, + "learning_rate": 0.00019415553034829265, + "loss": 0.8907, + "step": 7845 + }, + { + "epoch": 0.5444452154604121, + "grad_norm": 3.7276480197906494, + "learning_rate": 0.0001941536375216169, + "loss": 1.0431, + "step": 7846 + }, + { + "epoch": 0.5445146068975089, + "grad_norm": 2.47238826751709, + "learning_rate": 0.00019415174439770905, + "loss": 0.8867, + "step": 7847 + }, + { + "epoch": 0.5445839983346055, + "grad_norm": 3.1139323711395264, + "learning_rate": 0.00019414985097657503, + "loss": 1.3032, + "step": 7848 + }, + { + "epoch": 0.5446533897717022, + "grad_norm": 2.5691585540771484, + "learning_rate": 0.0001941479572582209, + "loss": 1.2387, + "step": 7849 + }, + { + "epoch": 0.5447227812087988, + "grad_norm": 2.5476813316345215, + "learning_rate": 0.00019414606324265254, + "loss": 0.8607, + "step": 7850 + }, + { + "epoch": 0.5447921726458955, + "grad_norm": 2.6082918643951416, + "learning_rate": 0.00019414416892987602, + "loss": 0.8804, + "step": 7851 + }, + { + "epoch": 0.5448615640829921, + "grad_norm": 2.764204740524292, + "learning_rate": 0.00019414227431989724, + "loss": 0.8542, + "step": 7852 + }, + { + "epoch": 0.5449309555200889, + "grad_norm": 3.5453908443450928, + "learning_rate": 0.00019414037941272225, + "loss": 1.0021, + "step": 7853 + }, + { + "epoch": 0.5450003469571855, + "grad_norm": 2.567636251449585, + "learning_rate": 0.00019413848420835703, + "loss": 0.8752, + "step": 7854 + }, + { + "epoch": 0.5450697383942822, + "grad_norm": 2.628082275390625, + "learning_rate": 0.00019413658870680748, + "loss": 0.7332, + "step": 7855 + }, + { + "epoch": 0.5451391298313788, + "grad_norm": 2.7761359214782715, + "learning_rate": 0.0001941346929080797, + "loss": 0.8722, + "step": 7856 + }, + { + "epoch": 0.5452085212684755, + "grad_norm": 2.710452079772949, + "learning_rate": 0.0001941327968121796, + "loss": 0.7245, + "step": 7857 + }, + { + "epoch": 0.5452779127055721, + "grad_norm": 2.413156509399414, + "learning_rate": 0.00019413090041911314, + "loss": 0.7378, + "step": 7858 + }, + { + "epoch": 0.5453473041426687, + "grad_norm": 2.708282947540283, + "learning_rate": 0.0001941290037288864, + "loss": 0.9024, + "step": 7859 + }, + { + "epoch": 0.5454166955797655, + "grad_norm": 3.2103006839752197, + "learning_rate": 0.0001941271067415053, + "loss": 1.1699, + "step": 7860 + }, + { + "epoch": 0.5454860870168621, + "grad_norm": 3.291412353515625, + "learning_rate": 0.00019412520945697584, + "loss": 0.886, + "step": 7861 + }, + { + "epoch": 0.5455554784539588, + "grad_norm": 3.3849287033081055, + "learning_rate": 0.000194123311875304, + "loss": 1.0982, + "step": 7862 + }, + { + "epoch": 0.5456248698910554, + "grad_norm": 2.7452306747436523, + "learning_rate": 0.00019412141399649578, + "loss": 0.759, + "step": 7863 + }, + { + "epoch": 0.5456942613281521, + "grad_norm": 2.774991512298584, + "learning_rate": 0.00019411951582055722, + "loss": 1.0312, + "step": 7864 + }, + { + "epoch": 0.5457636527652487, + "grad_norm": 3.416201591491699, + "learning_rate": 0.00019411761734749425, + "loss": 0.8618, + "step": 7865 + }, + { + "epoch": 0.5458330442023455, + "grad_norm": 2.8297882080078125, + "learning_rate": 0.00019411571857731288, + "loss": 0.6736, + "step": 7866 + }, + { + "epoch": 0.5459024356394421, + "grad_norm": 3.2652037143707275, + "learning_rate": 0.00019411381951001908, + "loss": 1.0016, + "step": 7867 + }, + { + "epoch": 0.5459718270765388, + "grad_norm": 3.6157166957855225, + "learning_rate": 0.00019411192014561892, + "loss": 1.1937, + "step": 7868 + }, + { + "epoch": 0.5460412185136354, + "grad_norm": 1.5300071239471436, + "learning_rate": 0.0001941100204841183, + "loss": 0.6104, + "step": 7869 + }, + { + "epoch": 0.5461106099507321, + "grad_norm": 2.9062280654907227, + "learning_rate": 0.00019410812052552328, + "loss": 0.905, + "step": 7870 + }, + { + "epoch": 0.5461800013878287, + "grad_norm": 2.8389337062835693, + "learning_rate": 0.00019410622026983987, + "loss": 0.9237, + "step": 7871 + }, + { + "epoch": 0.5462493928249254, + "grad_norm": 2.933769941329956, + "learning_rate": 0.00019410431971707398, + "loss": 0.6376, + "step": 7872 + }, + { + "epoch": 0.5463187842620221, + "grad_norm": 3.626769781112671, + "learning_rate": 0.00019410241886723176, + "loss": 1.0193, + "step": 7873 + }, + { + "epoch": 0.5463881756991187, + "grad_norm": 4.0702667236328125, + "learning_rate": 0.00019410051772031906, + "loss": 1.0666, + "step": 7874 + }, + { + "epoch": 0.5464575671362154, + "grad_norm": 3.2688522338867188, + "learning_rate": 0.00019409861627634194, + "loss": 1.175, + "step": 7875 + }, + { + "epoch": 0.546526958573312, + "grad_norm": 2.6724274158477783, + "learning_rate": 0.0001940967145353064, + "loss": 0.9269, + "step": 7876 + }, + { + "epoch": 0.5465963500104087, + "grad_norm": 2.4895896911621094, + "learning_rate": 0.00019409481249721847, + "loss": 0.7641, + "step": 7877 + }, + { + "epoch": 0.5466657414475053, + "grad_norm": 2.966721773147583, + "learning_rate": 0.00019409291016208414, + "loss": 0.8657, + "step": 7878 + }, + { + "epoch": 0.5467351328846021, + "grad_norm": 3.241046667098999, + "learning_rate": 0.00019409100752990942, + "loss": 0.8074, + "step": 7879 + }, + { + "epoch": 0.5468045243216987, + "grad_norm": 3.2341489791870117, + "learning_rate": 0.0001940891046007003, + "loss": 1.0217, + "step": 7880 + }, + { + "epoch": 0.5468739157587954, + "grad_norm": 2.6096713542938232, + "learning_rate": 0.00019408720137446275, + "loss": 1.0778, + "step": 7881 + }, + { + "epoch": 0.546943307195892, + "grad_norm": 3.120948076248169, + "learning_rate": 0.00019408529785120283, + "loss": 1.2588, + "step": 7882 + }, + { + "epoch": 0.5470126986329887, + "grad_norm": 2.5976033210754395, + "learning_rate": 0.0001940833940309266, + "loss": 0.9858, + "step": 7883 + }, + { + "epoch": 0.5470820900700853, + "grad_norm": 2.7166333198547363, + "learning_rate": 0.00019408148991363996, + "loss": 0.8622, + "step": 7884 + }, + { + "epoch": 0.5471514815071821, + "grad_norm": 3.145212173461914, + "learning_rate": 0.000194079585499349, + "loss": 0.9151, + "step": 7885 + }, + { + "epoch": 0.5472208729442787, + "grad_norm": 2.8002867698669434, + "learning_rate": 0.00019407768078805968, + "loss": 0.875, + "step": 7886 + }, + { + "epoch": 0.5472902643813753, + "grad_norm": 3.4321279525756836, + "learning_rate": 0.00019407577577977805, + "loss": 0.898, + "step": 7887 + }, + { + "epoch": 0.547359655818472, + "grad_norm": 3.161888837814331, + "learning_rate": 0.0001940738704745101, + "loss": 0.7769, + "step": 7888 + }, + { + "epoch": 0.5474290472555686, + "grad_norm": 3.156404733657837, + "learning_rate": 0.00019407196487226185, + "loss": 0.9955, + "step": 7889 + }, + { + "epoch": 0.5474984386926653, + "grad_norm": 2.4729833602905273, + "learning_rate": 0.00019407005897303936, + "loss": 0.9379, + "step": 7890 + }, + { + "epoch": 0.547567830129762, + "grad_norm": 2.345716714859009, + "learning_rate": 0.00019406815277684856, + "loss": 0.6748, + "step": 7891 + }, + { + "epoch": 0.5476372215668587, + "grad_norm": 2.474806785583496, + "learning_rate": 0.00019406624628369554, + "loss": 0.9831, + "step": 7892 + }, + { + "epoch": 0.5477066130039553, + "grad_norm": 3.247943162918091, + "learning_rate": 0.0001940643394935863, + "loss": 0.9789, + "step": 7893 + }, + { + "epoch": 0.547776004441052, + "grad_norm": 4.2522501945495605, + "learning_rate": 0.00019406243240652684, + "loss": 0.8356, + "step": 7894 + }, + { + "epoch": 0.5478453958781486, + "grad_norm": 3.1048882007598877, + "learning_rate": 0.0001940605250225232, + "loss": 1.1252, + "step": 7895 + }, + { + "epoch": 0.5479147873152453, + "grad_norm": 2.4015181064605713, + "learning_rate": 0.00019405861734158138, + "loss": 0.8633, + "step": 7896 + }, + { + "epoch": 0.547984178752342, + "grad_norm": 2.4546120166778564, + "learning_rate": 0.00019405670936370744, + "loss": 0.8835, + "step": 7897 + }, + { + "epoch": 0.5480535701894387, + "grad_norm": 2.16243314743042, + "learning_rate": 0.00019405480108890738, + "loss": 0.8211, + "step": 7898 + }, + { + "epoch": 0.5481229616265353, + "grad_norm": 2.9013302326202393, + "learning_rate": 0.00019405289251718722, + "loss": 0.8487, + "step": 7899 + }, + { + "epoch": 0.5481923530636319, + "grad_norm": 3.8599839210510254, + "learning_rate": 0.000194050983648553, + "loss": 1.1663, + "step": 7900 + }, + { + "epoch": 0.5482617445007286, + "grad_norm": 2.660583257675171, + "learning_rate": 0.0001940490744830107, + "loss": 1.0693, + "step": 7901 + }, + { + "epoch": 0.5483311359378252, + "grad_norm": 3.524284601211548, + "learning_rate": 0.00019404716502056642, + "loss": 0.7429, + "step": 7902 + }, + { + "epoch": 0.548400527374922, + "grad_norm": 3.2867400646209717, + "learning_rate": 0.00019404525526122614, + "loss": 0.94, + "step": 7903 + }, + { + "epoch": 0.5484699188120186, + "grad_norm": 2.879462718963623, + "learning_rate": 0.0001940433452049959, + "loss": 1.1124, + "step": 7904 + }, + { + "epoch": 0.5485393102491153, + "grad_norm": 2.6793019771575928, + "learning_rate": 0.00019404143485188174, + "loss": 0.8714, + "step": 7905 + }, + { + "epoch": 0.5486087016862119, + "grad_norm": 2.6991899013519287, + "learning_rate": 0.00019403952420188968, + "loss": 0.88, + "step": 7906 + }, + { + "epoch": 0.5486780931233086, + "grad_norm": 2.2214651107788086, + "learning_rate": 0.0001940376132550257, + "loss": 1.049, + "step": 7907 + }, + { + "epoch": 0.5487474845604052, + "grad_norm": 3.8108208179473877, + "learning_rate": 0.00019403570201129595, + "loss": 1.0586, + "step": 7908 + }, + { + "epoch": 0.548816875997502, + "grad_norm": 3.690464735031128, + "learning_rate": 0.0001940337904707064, + "loss": 1.1453, + "step": 7909 + }, + { + "epoch": 0.5488862674345986, + "grad_norm": 2.6858725547790527, + "learning_rate": 0.00019403187863326307, + "loss": 0.9072, + "step": 7910 + }, + { + "epoch": 0.5489556588716953, + "grad_norm": 3.016329765319824, + "learning_rate": 0.000194029966498972, + "loss": 1.0115, + "step": 7911 + }, + { + "epoch": 0.5490250503087919, + "grad_norm": 2.308082342147827, + "learning_rate": 0.00019402805406783922, + "loss": 0.703, + "step": 7912 + }, + { + "epoch": 0.5490944417458885, + "grad_norm": 3.6502749919891357, + "learning_rate": 0.00019402614133987077, + "loss": 0.892, + "step": 7913 + }, + { + "epoch": 0.5491638331829852, + "grad_norm": 2.83475399017334, + "learning_rate": 0.00019402422831507274, + "loss": 0.8216, + "step": 7914 + }, + { + "epoch": 0.5492332246200818, + "grad_norm": 3.769195318222046, + "learning_rate": 0.00019402231499345113, + "loss": 1.1425, + "step": 7915 + }, + { + "epoch": 0.5493026160571786, + "grad_norm": 2.727424383163452, + "learning_rate": 0.000194020401375012, + "loss": 0.8984, + "step": 7916 + }, + { + "epoch": 0.5493720074942752, + "grad_norm": 2.3201279640197754, + "learning_rate": 0.00019401848745976132, + "loss": 0.8171, + "step": 7917 + }, + { + "epoch": 0.5494413989313719, + "grad_norm": 2.6873390674591064, + "learning_rate": 0.0001940165732477052, + "loss": 0.746, + "step": 7918 + }, + { + "epoch": 0.5495107903684685, + "grad_norm": 2.3139572143554688, + "learning_rate": 0.00019401465873884969, + "loss": 0.7706, + "step": 7919 + }, + { + "epoch": 0.5495801818055652, + "grad_norm": 2.919308662414551, + "learning_rate": 0.00019401274393320077, + "loss": 0.7515, + "step": 7920 + }, + { + "epoch": 0.5496495732426618, + "grad_norm": 3.1952035427093506, + "learning_rate": 0.00019401082883076457, + "loss": 0.9648, + "step": 7921 + }, + { + "epoch": 0.5497189646797586, + "grad_norm": 2.3167057037353516, + "learning_rate": 0.00019400891343154702, + "loss": 0.94, + "step": 7922 + }, + { + "epoch": 0.5497883561168552, + "grad_norm": 2.4170515537261963, + "learning_rate": 0.00019400699773555432, + "loss": 1.0446, + "step": 7923 + }, + { + "epoch": 0.5498577475539519, + "grad_norm": 3.258338689804077, + "learning_rate": 0.00019400508174279238, + "loss": 1.0961, + "step": 7924 + }, + { + "epoch": 0.5499271389910485, + "grad_norm": 3.67775559425354, + "learning_rate": 0.00019400316545326734, + "loss": 0.9983, + "step": 7925 + }, + { + "epoch": 0.5499965304281451, + "grad_norm": 2.462223529815674, + "learning_rate": 0.0001940012488669852, + "loss": 0.7071, + "step": 7926 + }, + { + "epoch": 0.5500659218652418, + "grad_norm": 2.5255377292633057, + "learning_rate": 0.000193999331983952, + "loss": 0.8636, + "step": 7927 + }, + { + "epoch": 0.5501353133023384, + "grad_norm": 3.0596251487731934, + "learning_rate": 0.00019399741480417384, + "loss": 1.217, + "step": 7928 + }, + { + "epoch": 0.5502047047394352, + "grad_norm": 2.5770723819732666, + "learning_rate": 0.00019399549732765676, + "loss": 0.8801, + "step": 7929 + }, + { + "epoch": 0.5502740961765318, + "grad_norm": 3.0162997245788574, + "learning_rate": 0.00019399357955440677, + "loss": 1.0027, + "step": 7930 + }, + { + "epoch": 0.5503434876136285, + "grad_norm": 2.1894359588623047, + "learning_rate": 0.00019399166148442997, + "loss": 0.957, + "step": 7931 + }, + { + "epoch": 0.5504128790507251, + "grad_norm": 2.796586036682129, + "learning_rate": 0.0001939897431177324, + "loss": 1.0675, + "step": 7932 + }, + { + "epoch": 0.5504822704878218, + "grad_norm": 2.1428403854370117, + "learning_rate": 0.00019398782445432012, + "loss": 0.6379, + "step": 7933 + }, + { + "epoch": 0.5505516619249184, + "grad_norm": 2.7116105556488037, + "learning_rate": 0.00019398590549419916, + "loss": 0.8352, + "step": 7934 + }, + { + "epoch": 0.5506210533620152, + "grad_norm": 2.839585542678833, + "learning_rate": 0.00019398398623737562, + "loss": 0.766, + "step": 7935 + }, + { + "epoch": 0.5506904447991118, + "grad_norm": 3.652141571044922, + "learning_rate": 0.00019398206668385553, + "loss": 0.8793, + "step": 7936 + }, + { + "epoch": 0.5507598362362085, + "grad_norm": 1.9110232591629028, + "learning_rate": 0.00019398014683364497, + "loss": 0.7687, + "step": 7937 + }, + { + "epoch": 0.5508292276733051, + "grad_norm": 3.40004301071167, + "learning_rate": 0.00019397822668675, + "loss": 1.0777, + "step": 7938 + }, + { + "epoch": 0.5508986191104017, + "grad_norm": 2.9858202934265137, + "learning_rate": 0.00019397630624317662, + "loss": 0.7756, + "step": 7939 + }, + { + "epoch": 0.5509680105474984, + "grad_norm": 3.2941410541534424, + "learning_rate": 0.00019397438550293098, + "loss": 0.7279, + "step": 7940 + }, + { + "epoch": 0.551037401984595, + "grad_norm": 2.7215688228607178, + "learning_rate": 0.0001939724644660191, + "loss": 1.0662, + "step": 7941 + }, + { + "epoch": 0.5511067934216918, + "grad_norm": 4.165773391723633, + "learning_rate": 0.0001939705431324471, + "loss": 0.8412, + "step": 7942 + }, + { + "epoch": 0.5511761848587884, + "grad_norm": 3.03184175491333, + "learning_rate": 0.00019396862150222095, + "loss": 0.6935, + "step": 7943 + }, + { + "epoch": 0.5512455762958851, + "grad_norm": 2.4040424823760986, + "learning_rate": 0.00019396669957534677, + "loss": 0.7938, + "step": 7944 + }, + { + "epoch": 0.5513149677329817, + "grad_norm": 3.7894604206085205, + "learning_rate": 0.00019396477735183064, + "loss": 0.9888, + "step": 7945 + }, + { + "epoch": 0.5513843591700784, + "grad_norm": 3.3787131309509277, + "learning_rate": 0.00019396285483167858, + "loss": 0.9537, + "step": 7946 + }, + { + "epoch": 0.551453750607175, + "grad_norm": 2.4954683780670166, + "learning_rate": 0.0001939609320148967, + "loss": 0.9905, + "step": 7947 + }, + { + "epoch": 0.5515231420442718, + "grad_norm": 3.285595417022705, + "learning_rate": 0.00019395900890149107, + "loss": 0.9494, + "step": 7948 + }, + { + "epoch": 0.5515925334813684, + "grad_norm": 3.4998745918273926, + "learning_rate": 0.00019395708549146775, + "loss": 0.923, + "step": 7949 + }, + { + "epoch": 0.5516619249184651, + "grad_norm": 2.8485820293426514, + "learning_rate": 0.00019395516178483279, + "loss": 0.6465, + "step": 7950 + }, + { + "epoch": 0.5517313163555617, + "grad_norm": 3.259192705154419, + "learning_rate": 0.00019395323778159229, + "loss": 0.8134, + "step": 7951 + }, + { + "epoch": 0.5518007077926584, + "grad_norm": 2.922046422958374, + "learning_rate": 0.00019395131348175232, + "loss": 1.0754, + "step": 7952 + }, + { + "epoch": 0.551870099229755, + "grad_norm": 3.604283094406128, + "learning_rate": 0.000193949388885319, + "loss": 1.0098, + "step": 7953 + }, + { + "epoch": 0.5519394906668517, + "grad_norm": 3.7990853786468506, + "learning_rate": 0.0001939474639922983, + "loss": 0.9216, + "step": 7954 + }, + { + "epoch": 0.5520088821039484, + "grad_norm": 2.4528942108154297, + "learning_rate": 0.00019394553880269636, + "loss": 0.8407, + "step": 7955 + }, + { + "epoch": 0.552078273541045, + "grad_norm": 2.6893255710601807, + "learning_rate": 0.00019394361331651929, + "loss": 1.0108, + "step": 7956 + }, + { + "epoch": 0.5521476649781417, + "grad_norm": 4.033621311187744, + "learning_rate": 0.00019394168753377312, + "loss": 0.7995, + "step": 7957 + }, + { + "epoch": 0.5522170564152383, + "grad_norm": 2.500983238220215, + "learning_rate": 0.00019393976145446392, + "loss": 0.9109, + "step": 7958 + }, + { + "epoch": 0.552286447852335, + "grad_norm": 2.413640022277832, + "learning_rate": 0.0001939378350785978, + "loss": 0.9349, + "step": 7959 + }, + { + "epoch": 0.5523558392894317, + "grad_norm": 2.768824577331543, + "learning_rate": 0.00019393590840618088, + "loss": 1.0398, + "step": 7960 + }, + { + "epoch": 0.5524252307265284, + "grad_norm": 2.71174955368042, + "learning_rate": 0.00019393398143721915, + "loss": 0.7727, + "step": 7961 + }, + { + "epoch": 0.552494622163625, + "grad_norm": 2.2454631328582764, + "learning_rate": 0.00019393205417171875, + "loss": 0.8393, + "step": 7962 + }, + { + "epoch": 0.5525640136007217, + "grad_norm": 2.6001055240631104, + "learning_rate": 0.00019393012660968574, + "loss": 1.0344, + "step": 7963 + }, + { + "epoch": 0.5526334050378183, + "grad_norm": 3.8246052265167236, + "learning_rate": 0.00019392819875112624, + "loss": 0.9717, + "step": 7964 + }, + { + "epoch": 0.552702796474915, + "grad_norm": 3.3294126987457275, + "learning_rate": 0.0001939262705960463, + "loss": 1.2462, + "step": 7965 + }, + { + "epoch": 0.5527721879120117, + "grad_norm": 2.272508144378662, + "learning_rate": 0.000193924342144452, + "loss": 0.8592, + "step": 7966 + }, + { + "epoch": 0.5528415793491083, + "grad_norm": 2.2898149490356445, + "learning_rate": 0.00019392241339634948, + "loss": 0.9699, + "step": 7967 + }, + { + "epoch": 0.552910970786205, + "grad_norm": 2.7361931800842285, + "learning_rate": 0.00019392048435174481, + "loss": 0.877, + "step": 7968 + }, + { + "epoch": 0.5529803622233016, + "grad_norm": 2.687008857727051, + "learning_rate": 0.00019391855501064403, + "loss": 1.0451, + "step": 7969 + }, + { + "epoch": 0.5530497536603983, + "grad_norm": 2.881882429122925, + "learning_rate": 0.0001939166253730533, + "loss": 0.8634, + "step": 7970 + }, + { + "epoch": 0.5531191450974949, + "grad_norm": 2.181668758392334, + "learning_rate": 0.00019391469543897868, + "loss": 0.7011, + "step": 7971 + }, + { + "epoch": 0.5531885365345917, + "grad_norm": 3.4460248947143555, + "learning_rate": 0.00019391276520842621, + "loss": 0.7908, + "step": 7972 + }, + { + "epoch": 0.5532579279716883, + "grad_norm": 3.6272406578063965, + "learning_rate": 0.00019391083468140208, + "loss": 1.0739, + "step": 7973 + }, + { + "epoch": 0.553327319408785, + "grad_norm": 2.0759520530700684, + "learning_rate": 0.00019390890385791236, + "loss": 0.6484, + "step": 7974 + }, + { + "epoch": 0.5533967108458816, + "grad_norm": 2.7267472743988037, + "learning_rate": 0.00019390697273796308, + "loss": 1.0027, + "step": 7975 + }, + { + "epoch": 0.5534661022829783, + "grad_norm": 2.3552238941192627, + "learning_rate": 0.00019390504132156043, + "loss": 0.9138, + "step": 7976 + }, + { + "epoch": 0.5535354937200749, + "grad_norm": 3.1371428966522217, + "learning_rate": 0.00019390310960871042, + "loss": 1.012, + "step": 7977 + }, + { + "epoch": 0.5536048851571717, + "grad_norm": 2.5967063903808594, + "learning_rate": 0.0001939011775994192, + "loss": 0.8677, + "step": 7978 + }, + { + "epoch": 0.5536742765942683, + "grad_norm": 2.873370885848999, + "learning_rate": 0.00019389924529369286, + "loss": 0.9817, + "step": 7979 + }, + { + "epoch": 0.5537436680313649, + "grad_norm": 2.910989284515381, + "learning_rate": 0.00019389731269153752, + "loss": 0.9771, + "step": 7980 + }, + { + "epoch": 0.5538130594684616, + "grad_norm": 2.70967435836792, + "learning_rate": 0.0001938953797929592, + "loss": 0.9668, + "step": 7981 + }, + { + "epoch": 0.5538824509055582, + "grad_norm": 3.5387120246887207, + "learning_rate": 0.0001938934465979641, + "loss": 0.8257, + "step": 7982 + }, + { + "epoch": 0.5539518423426549, + "grad_norm": 4.069309711456299, + "learning_rate": 0.00019389151310655826, + "loss": 1.2711, + "step": 7983 + }, + { + "epoch": 0.5540212337797515, + "grad_norm": 3.2496557235717773, + "learning_rate": 0.00019388957931874785, + "loss": 1.1455, + "step": 7984 + }, + { + "epoch": 0.5540906252168483, + "grad_norm": 3.1475772857666016, + "learning_rate": 0.00019388764523453888, + "loss": 1.1591, + "step": 7985 + }, + { + "epoch": 0.5541600166539449, + "grad_norm": 3.2158212661743164, + "learning_rate": 0.00019388571085393754, + "loss": 0.9354, + "step": 7986 + }, + { + "epoch": 0.5542294080910416, + "grad_norm": 2.447359323501587, + "learning_rate": 0.00019388377617694992, + "loss": 0.7381, + "step": 7987 + }, + { + "epoch": 0.5542987995281382, + "grad_norm": 2.656437397003174, + "learning_rate": 0.00019388184120358205, + "loss": 0.714, + "step": 7988 + }, + { + "epoch": 0.5543681909652349, + "grad_norm": 3.227736473083496, + "learning_rate": 0.00019387990593384014, + "loss": 1.0147, + "step": 7989 + }, + { + "epoch": 0.5544375824023315, + "grad_norm": 4.2276225090026855, + "learning_rate": 0.00019387797036773025, + "loss": 1.2212, + "step": 7990 + }, + { + "epoch": 0.5545069738394283, + "grad_norm": 3.5099220275878906, + "learning_rate": 0.0001938760345052585, + "loss": 0.9762, + "step": 7991 + }, + { + "epoch": 0.5545763652765249, + "grad_norm": 2.4550113677978516, + "learning_rate": 0.000193874098346431, + "loss": 0.875, + "step": 7992 + }, + { + "epoch": 0.5546457567136215, + "grad_norm": 2.75378680229187, + "learning_rate": 0.0001938721618912539, + "loss": 1.0607, + "step": 7993 + }, + { + "epoch": 0.5547151481507182, + "grad_norm": 2.559828758239746, + "learning_rate": 0.00019387022513973323, + "loss": 0.7924, + "step": 7994 + }, + { + "epoch": 0.5547845395878148, + "grad_norm": 2.569800615310669, + "learning_rate": 0.00019386828809187515, + "loss": 1.0019, + "step": 7995 + }, + { + "epoch": 0.5548539310249115, + "grad_norm": 2.9947710037231445, + "learning_rate": 0.0001938663507476858, + "loss": 0.8019, + "step": 7996 + }, + { + "epoch": 0.5549233224620082, + "grad_norm": 3.0326292514801025, + "learning_rate": 0.00019386441310717127, + "loss": 0.9891, + "step": 7997 + }, + { + "epoch": 0.5549927138991049, + "grad_norm": 2.4928970336914062, + "learning_rate": 0.00019386247517033767, + "loss": 0.8237, + "step": 7998 + }, + { + "epoch": 0.5550621053362015, + "grad_norm": 2.4099225997924805, + "learning_rate": 0.00019386053693719112, + "loss": 0.9324, + "step": 7999 + }, + { + "epoch": 0.5551314967732982, + "grad_norm": 2.274334192276001, + "learning_rate": 0.00019385859840773774, + "loss": 0.7767, + "step": 8000 + }, + { + "epoch": 0.5551314967732982, + "eval_loss": 0.9277942180633545, + "eval_runtime": 3781.6806, + "eval_samples_per_second": 6.418, + "eval_steps_per_second": 0.802, + "step": 8000 + }, + { + "epoch": 0.5552008882103948, + "grad_norm": 2.394580125808716, + "learning_rate": 0.00019385665958198366, + "loss": 0.7873, + "step": 8001 + }, + { + "epoch": 0.5552702796474915, + "grad_norm": 2.7529664039611816, + "learning_rate": 0.00019385472045993501, + "loss": 0.9472, + "step": 8002 + }, + { + "epoch": 0.5553396710845881, + "grad_norm": 2.6358561515808105, + "learning_rate": 0.00019385278104159788, + "loss": 0.8349, + "step": 8003 + }, + { + "epoch": 0.5554090625216849, + "grad_norm": 2.8277437686920166, + "learning_rate": 0.0001938508413269784, + "loss": 0.8397, + "step": 8004 + }, + { + "epoch": 0.5554784539587815, + "grad_norm": 2.3979175090789795, + "learning_rate": 0.00019384890131608274, + "loss": 0.8883, + "step": 8005 + }, + { + "epoch": 0.5555478453958782, + "grad_norm": 3.124020576477051, + "learning_rate": 0.00019384696100891696, + "loss": 1.0238, + "step": 8006 + }, + { + "epoch": 0.5556172368329748, + "grad_norm": 2.6803908348083496, + "learning_rate": 0.00019384502040548722, + "loss": 0.9103, + "step": 8007 + }, + { + "epoch": 0.5556866282700714, + "grad_norm": 2.356659173965454, + "learning_rate": 0.00019384307950579964, + "loss": 0.904, + "step": 8008 + }, + { + "epoch": 0.5557560197071681, + "grad_norm": 2.4668290615081787, + "learning_rate": 0.00019384113830986033, + "loss": 0.7948, + "step": 8009 + }, + { + "epoch": 0.5558254111442648, + "grad_norm": 2.5657570362091064, + "learning_rate": 0.00019383919681767545, + "loss": 0.8936, + "step": 8010 + }, + { + "epoch": 0.5558948025813615, + "grad_norm": 3.3164470195770264, + "learning_rate": 0.0001938372550292511, + "loss": 0.9743, + "step": 8011 + }, + { + "epoch": 0.5559641940184581, + "grad_norm": 2.9404993057250977, + "learning_rate": 0.00019383531294459342, + "loss": 0.8761, + "step": 8012 + }, + { + "epoch": 0.5560335854555548, + "grad_norm": 2.4552414417266846, + "learning_rate": 0.00019383337056370857, + "loss": 0.7835, + "step": 8013 + }, + { + "epoch": 0.5561029768926514, + "grad_norm": 2.322554588317871, + "learning_rate": 0.00019383142788660264, + "loss": 0.8232, + "step": 8014 + }, + { + "epoch": 0.5561723683297481, + "grad_norm": 3.806366443634033, + "learning_rate": 0.00019382948491328177, + "loss": 0.8795, + "step": 8015 + }, + { + "epoch": 0.5562417597668448, + "grad_norm": 2.330097198486328, + "learning_rate": 0.0001938275416437521, + "loss": 0.91, + "step": 8016 + }, + { + "epoch": 0.5563111512039415, + "grad_norm": 2.4348948001861572, + "learning_rate": 0.0001938255980780198, + "loss": 1.1504, + "step": 8017 + }, + { + "epoch": 0.5563805426410381, + "grad_norm": 2.628551721572876, + "learning_rate": 0.00019382365421609095, + "loss": 1.0683, + "step": 8018 + }, + { + "epoch": 0.5564499340781348, + "grad_norm": 3.00150990486145, + "learning_rate": 0.0001938217100579717, + "loss": 0.9439, + "step": 8019 + }, + { + "epoch": 0.5565193255152314, + "grad_norm": 2.4807047843933105, + "learning_rate": 0.0001938197656036682, + "loss": 0.9085, + "step": 8020 + }, + { + "epoch": 0.556588716952328, + "grad_norm": 2.2051773071289062, + "learning_rate": 0.00019381782085318657, + "loss": 0.7655, + "step": 8021 + }, + { + "epoch": 0.5566581083894248, + "grad_norm": 3.6897637844085693, + "learning_rate": 0.00019381587580653302, + "loss": 1.1651, + "step": 8022 + }, + { + "epoch": 0.5567274998265214, + "grad_norm": 2.506798505783081, + "learning_rate": 0.0001938139304637136, + "loss": 0.7941, + "step": 8023 + }, + { + "epoch": 0.5567968912636181, + "grad_norm": 2.526745319366455, + "learning_rate": 0.0001938119848247345, + "loss": 0.9838, + "step": 8024 + }, + { + "epoch": 0.5568662827007147, + "grad_norm": 2.512265920639038, + "learning_rate": 0.00019381003888960181, + "loss": 0.8729, + "step": 8025 + }, + { + "epoch": 0.5569356741378114, + "grad_norm": 2.6299643516540527, + "learning_rate": 0.00019380809265832173, + "loss": 0.9312, + "step": 8026 + }, + { + "epoch": 0.557005065574908, + "grad_norm": 2.493222951889038, + "learning_rate": 0.0001938061461309004, + "loss": 0.8997, + "step": 8027 + }, + { + "epoch": 0.5570744570120048, + "grad_norm": 2.677225112915039, + "learning_rate": 0.00019380419930734396, + "loss": 0.8426, + "step": 8028 + }, + { + "epoch": 0.5571438484491014, + "grad_norm": 3.83404278755188, + "learning_rate": 0.00019380225218765853, + "loss": 1.0868, + "step": 8029 + }, + { + "epoch": 0.5572132398861981, + "grad_norm": 3.6845192909240723, + "learning_rate": 0.00019380030477185025, + "loss": 1.0872, + "step": 8030 + }, + { + "epoch": 0.5572826313232947, + "grad_norm": 2.410163640975952, + "learning_rate": 0.00019379835705992532, + "loss": 0.9436, + "step": 8031 + }, + { + "epoch": 0.5573520227603914, + "grad_norm": 2.565487861633301, + "learning_rate": 0.00019379640905188984, + "loss": 0.9178, + "step": 8032 + }, + { + "epoch": 0.557421414197488, + "grad_norm": 2.7604990005493164, + "learning_rate": 0.00019379446074775, + "loss": 0.6909, + "step": 8033 + }, + { + "epoch": 0.5574908056345846, + "grad_norm": 3.2503020763397217, + "learning_rate": 0.00019379251214751193, + "loss": 0.8005, + "step": 8034 + }, + { + "epoch": 0.5575601970716814, + "grad_norm": 3.3046200275421143, + "learning_rate": 0.00019379056325118175, + "loss": 1.2247, + "step": 8035 + }, + { + "epoch": 0.557629588508778, + "grad_norm": 3.065530776977539, + "learning_rate": 0.00019378861405876566, + "loss": 0.8167, + "step": 8036 + }, + { + "epoch": 0.5576989799458747, + "grad_norm": 2.388145685195923, + "learning_rate": 0.00019378666457026982, + "loss": 0.9512, + "step": 8037 + }, + { + "epoch": 0.5577683713829713, + "grad_norm": 2.4296417236328125, + "learning_rate": 0.00019378471478570031, + "loss": 1.0409, + "step": 8038 + }, + { + "epoch": 0.557837762820068, + "grad_norm": 2.761734962463379, + "learning_rate": 0.00019378276470506341, + "loss": 1.0039, + "step": 8039 + }, + { + "epoch": 0.5579071542571646, + "grad_norm": 2.199970245361328, + "learning_rate": 0.00019378081432836513, + "loss": 0.7575, + "step": 8040 + }, + { + "epoch": 0.5579765456942614, + "grad_norm": 3.748638868331909, + "learning_rate": 0.00019377886365561176, + "loss": 0.9597, + "step": 8041 + }, + { + "epoch": 0.558045937131358, + "grad_norm": 2.562927007675171, + "learning_rate": 0.00019377691268680934, + "loss": 0.7942, + "step": 8042 + }, + { + "epoch": 0.5581153285684547, + "grad_norm": 2.6208314895629883, + "learning_rate": 0.0001937749614219641, + "loss": 0.7788, + "step": 8043 + }, + { + "epoch": 0.5581847200055513, + "grad_norm": 3.6419191360473633, + "learning_rate": 0.00019377300986108225, + "loss": 1.0223, + "step": 8044 + }, + { + "epoch": 0.558254111442648, + "grad_norm": 2.137507677078247, + "learning_rate": 0.00019377105800416983, + "loss": 0.756, + "step": 8045 + }, + { + "epoch": 0.5583235028797446, + "grad_norm": 2.5584726333618164, + "learning_rate": 0.00019376910585123307, + "loss": 0.8191, + "step": 8046 + }, + { + "epoch": 0.5583928943168412, + "grad_norm": 2.572058916091919, + "learning_rate": 0.00019376715340227812, + "loss": 0.8374, + "step": 8047 + }, + { + "epoch": 0.558462285753938, + "grad_norm": 1.9331883192062378, + "learning_rate": 0.00019376520065731115, + "loss": 0.8554, + "step": 8048 + }, + { + "epoch": 0.5585316771910346, + "grad_norm": 2.7145068645477295, + "learning_rate": 0.0001937632476163383, + "loss": 0.8788, + "step": 8049 + }, + { + "epoch": 0.5586010686281313, + "grad_norm": 3.0196776390075684, + "learning_rate": 0.00019376129427936577, + "loss": 0.9285, + "step": 8050 + }, + { + "epoch": 0.5586704600652279, + "grad_norm": 2.4989609718322754, + "learning_rate": 0.00019375934064639973, + "loss": 0.9425, + "step": 8051 + }, + { + "epoch": 0.5587398515023246, + "grad_norm": 2.4625954627990723, + "learning_rate": 0.0001937573867174463, + "loss": 0.8367, + "step": 8052 + }, + { + "epoch": 0.5588092429394212, + "grad_norm": 2.6068155765533447, + "learning_rate": 0.0001937554324925117, + "loss": 0.7721, + "step": 8053 + }, + { + "epoch": 0.558878634376518, + "grad_norm": 4.261009216308594, + "learning_rate": 0.00019375347797160206, + "loss": 1.017, + "step": 8054 + }, + { + "epoch": 0.5589480258136146, + "grad_norm": 2.6565346717834473, + "learning_rate": 0.0001937515231547236, + "loss": 0.973, + "step": 8055 + }, + { + "epoch": 0.5590174172507113, + "grad_norm": 2.7081217765808105, + "learning_rate": 0.00019374956804188242, + "loss": 0.9163, + "step": 8056 + }, + { + "epoch": 0.5590868086878079, + "grad_norm": 2.3332529067993164, + "learning_rate": 0.00019374761263308475, + "loss": 0.6626, + "step": 8057 + }, + { + "epoch": 0.5591562001249046, + "grad_norm": 3.1579749584198, + "learning_rate": 0.00019374565692833676, + "loss": 1.129, + "step": 8058 + }, + { + "epoch": 0.5592255915620012, + "grad_norm": 2.7539026737213135, + "learning_rate": 0.00019374370092764458, + "loss": 0.8517, + "step": 8059 + }, + { + "epoch": 0.5592949829990979, + "grad_norm": 2.41792368888855, + "learning_rate": 0.0001937417446310144, + "loss": 0.9161, + "step": 8060 + }, + { + "epoch": 0.5593643744361946, + "grad_norm": 3.8383257389068604, + "learning_rate": 0.00019373978803845243, + "loss": 0.9924, + "step": 8061 + }, + { + "epoch": 0.5594337658732912, + "grad_norm": 2.671685218811035, + "learning_rate": 0.00019373783114996483, + "loss": 0.9731, + "step": 8062 + }, + { + "epoch": 0.5595031573103879, + "grad_norm": 2.771263837814331, + "learning_rate": 0.00019373587396555778, + "loss": 0.9617, + "step": 8063 + }, + { + "epoch": 0.5595725487474845, + "grad_norm": 2.964975118637085, + "learning_rate": 0.00019373391648523742, + "loss": 0.9939, + "step": 8064 + }, + { + "epoch": 0.5596419401845812, + "grad_norm": 2.195880889892578, + "learning_rate": 0.00019373195870900999, + "loss": 0.7582, + "step": 8065 + }, + { + "epoch": 0.5597113316216779, + "grad_norm": 2.7346551418304443, + "learning_rate": 0.00019373000063688164, + "loss": 0.954, + "step": 8066 + }, + { + "epoch": 0.5597807230587746, + "grad_norm": 3.182511329650879, + "learning_rate": 0.00019372804226885855, + "loss": 1.1834, + "step": 8067 + }, + { + "epoch": 0.5598501144958712, + "grad_norm": 2.7326717376708984, + "learning_rate": 0.00019372608360494688, + "loss": 0.7391, + "step": 8068 + }, + { + "epoch": 0.5599195059329679, + "grad_norm": 3.644040584564209, + "learning_rate": 0.00019372412464515283, + "loss": 1.0271, + "step": 8069 + }, + { + "epoch": 0.5599888973700645, + "grad_norm": 2.9138853549957275, + "learning_rate": 0.00019372216538948264, + "loss": 0.9132, + "step": 8070 + }, + { + "epoch": 0.5600582888071612, + "grad_norm": 1.9580130577087402, + "learning_rate": 0.0001937202058379424, + "loss": 0.9552, + "step": 8071 + }, + { + "epoch": 0.5601276802442579, + "grad_norm": 2.8117575645446777, + "learning_rate": 0.0001937182459905384, + "loss": 1.2333, + "step": 8072 + }, + { + "epoch": 0.5601970716813546, + "grad_norm": 2.9322268962860107, + "learning_rate": 0.00019371628584727676, + "loss": 0.9641, + "step": 8073 + }, + { + "epoch": 0.5602664631184512, + "grad_norm": 1.8567845821380615, + "learning_rate": 0.00019371432540816365, + "loss": 0.8052, + "step": 8074 + }, + { + "epoch": 0.5603358545555478, + "grad_norm": 3.06998348236084, + "learning_rate": 0.00019371236467320528, + "loss": 1.1546, + "step": 8075 + }, + { + "epoch": 0.5604052459926445, + "grad_norm": 3.138953447341919, + "learning_rate": 0.00019371040364240787, + "loss": 1.002, + "step": 8076 + }, + { + "epoch": 0.5604746374297411, + "grad_norm": 2.034409761428833, + "learning_rate": 0.00019370844231577757, + "loss": 0.9194, + "step": 8077 + }, + { + "epoch": 0.5605440288668379, + "grad_norm": 2.4948277473449707, + "learning_rate": 0.0001937064806933206, + "loss": 0.9456, + "step": 8078 + }, + { + "epoch": 0.5606134203039345, + "grad_norm": 4.464545726776123, + "learning_rate": 0.00019370451877504316, + "loss": 1.1476, + "step": 8079 + }, + { + "epoch": 0.5606828117410312, + "grad_norm": 2.6885199546813965, + "learning_rate": 0.00019370255656095143, + "loss": 0.9807, + "step": 8080 + }, + { + "epoch": 0.5607522031781278, + "grad_norm": 2.9827568531036377, + "learning_rate": 0.00019370059405105156, + "loss": 0.96, + "step": 8081 + }, + { + "epoch": 0.5608215946152245, + "grad_norm": 3.007678508758545, + "learning_rate": 0.0001936986312453498, + "loss": 0.9645, + "step": 8082 + }, + { + "epoch": 0.5608909860523211, + "grad_norm": 2.3021748065948486, + "learning_rate": 0.00019369666814385235, + "loss": 0.8488, + "step": 8083 + }, + { + "epoch": 0.5609603774894178, + "grad_norm": 3.506593942642212, + "learning_rate": 0.00019369470474656536, + "loss": 0.9796, + "step": 8084 + }, + { + "epoch": 0.5610297689265145, + "grad_norm": 2.5903680324554443, + "learning_rate": 0.00019369274105349507, + "loss": 1.0872, + "step": 8085 + }, + { + "epoch": 0.5610991603636112, + "grad_norm": 2.6436069011688232, + "learning_rate": 0.0001936907770646477, + "loss": 0.8777, + "step": 8086 + }, + { + "epoch": 0.5611685518007078, + "grad_norm": 2.863538980484009, + "learning_rate": 0.00019368881278002936, + "loss": 1.0099, + "step": 8087 + }, + { + "epoch": 0.5612379432378044, + "grad_norm": 2.250659704208374, + "learning_rate": 0.00019368684819964639, + "loss": 0.7937, + "step": 8088 + }, + { + "epoch": 0.5613073346749011, + "grad_norm": 2.3103325366973877, + "learning_rate": 0.00019368488332350482, + "loss": 1.1187, + "step": 8089 + }, + { + "epoch": 0.5613767261119977, + "grad_norm": 2.7776100635528564, + "learning_rate": 0.00019368291815161098, + "loss": 0.8435, + "step": 8090 + }, + { + "epoch": 0.5614461175490945, + "grad_norm": 2.240226984024048, + "learning_rate": 0.00019368095268397105, + "loss": 0.9549, + "step": 8091 + }, + { + "epoch": 0.5615155089861911, + "grad_norm": 2.3339672088623047, + "learning_rate": 0.0001936789869205912, + "loss": 0.9382, + "step": 8092 + }, + { + "epoch": 0.5615849004232878, + "grad_norm": 3.0781352519989014, + "learning_rate": 0.00019367702086147768, + "loss": 0.9096, + "step": 8093 + }, + { + "epoch": 0.5616542918603844, + "grad_norm": 3.202357769012451, + "learning_rate": 0.00019367505450663665, + "loss": 1.0958, + "step": 8094 + }, + { + "epoch": 0.5617236832974811, + "grad_norm": 2.951258420944214, + "learning_rate": 0.00019367308785607436, + "loss": 0.6884, + "step": 8095 + }, + { + "epoch": 0.5617930747345777, + "grad_norm": 2.2781004905700684, + "learning_rate": 0.00019367112090979698, + "loss": 0.7954, + "step": 8096 + }, + { + "epoch": 0.5618624661716745, + "grad_norm": 2.793929100036621, + "learning_rate": 0.00019366915366781075, + "loss": 0.9835, + "step": 8097 + }, + { + "epoch": 0.5619318576087711, + "grad_norm": 2.365536689758301, + "learning_rate": 0.00019366718613012188, + "loss": 0.7598, + "step": 8098 + }, + { + "epoch": 0.5620012490458678, + "grad_norm": 2.0690507888793945, + "learning_rate": 0.00019366521829673655, + "loss": 0.7512, + "step": 8099 + }, + { + "epoch": 0.5620706404829644, + "grad_norm": 2.897068500518799, + "learning_rate": 0.000193663250167661, + "loss": 0.8406, + "step": 8100 + }, + { + "epoch": 0.562140031920061, + "grad_norm": 1.8130314350128174, + "learning_rate": 0.00019366128174290144, + "loss": 1.0095, + "step": 8101 + }, + { + "epoch": 0.5622094233571577, + "grad_norm": 2.7864127159118652, + "learning_rate": 0.00019365931302246407, + "loss": 0.8312, + "step": 8102 + }, + { + "epoch": 0.5622788147942543, + "grad_norm": 2.726105213165283, + "learning_rate": 0.00019365734400635513, + "loss": 0.9432, + "step": 8103 + }, + { + "epoch": 0.5623482062313511, + "grad_norm": 2.608013153076172, + "learning_rate": 0.00019365537469458082, + "loss": 0.7695, + "step": 8104 + }, + { + "epoch": 0.5624175976684477, + "grad_norm": 2.4247515201568604, + "learning_rate": 0.00019365340508714735, + "loss": 0.921, + "step": 8105 + }, + { + "epoch": 0.5624869891055444, + "grad_norm": 2.49722957611084, + "learning_rate": 0.00019365143518406094, + "loss": 1.103, + "step": 8106 + }, + { + "epoch": 0.562556380542641, + "grad_norm": 2.958463430404663, + "learning_rate": 0.00019364946498532786, + "loss": 0.8244, + "step": 8107 + }, + { + "epoch": 0.5626257719797377, + "grad_norm": 2.5267176628112793, + "learning_rate": 0.00019364749449095423, + "loss": 0.8014, + "step": 8108 + }, + { + "epoch": 0.5626951634168343, + "grad_norm": 4.171938896179199, + "learning_rate": 0.00019364552370094632, + "loss": 0.7489, + "step": 8109 + }, + { + "epoch": 0.5627645548539311, + "grad_norm": 3.7872917652130127, + "learning_rate": 0.0001936435526153104, + "loss": 1.1348, + "step": 8110 + }, + { + "epoch": 0.5628339462910277, + "grad_norm": 3.173095226287842, + "learning_rate": 0.00019364158123405263, + "loss": 0.9957, + "step": 8111 + }, + { + "epoch": 0.5629033377281244, + "grad_norm": 2.9283342361450195, + "learning_rate": 0.00019363960955717926, + "loss": 1.0816, + "step": 8112 + }, + { + "epoch": 0.562972729165221, + "grad_norm": 3.2138381004333496, + "learning_rate": 0.0001936376375846965, + "loss": 0.8849, + "step": 8113 + }, + { + "epoch": 0.5630421206023176, + "grad_norm": 3.4591550827026367, + "learning_rate": 0.00019363566531661056, + "loss": 0.7832, + "step": 8114 + }, + { + "epoch": 0.5631115120394143, + "grad_norm": 6.716202735900879, + "learning_rate": 0.00019363369275292774, + "loss": 0.943, + "step": 8115 + }, + { + "epoch": 0.563180903476511, + "grad_norm": 5.25673246383667, + "learning_rate": 0.00019363171989365417, + "loss": 0.7417, + "step": 8116 + }, + { + "epoch": 0.5632502949136077, + "grad_norm": 2.800534725189209, + "learning_rate": 0.00019362974673879614, + "loss": 0.7512, + "step": 8117 + }, + { + "epoch": 0.5633196863507043, + "grad_norm": 2.631495952606201, + "learning_rate": 0.00019362777328835987, + "loss": 0.8332, + "step": 8118 + }, + { + "epoch": 0.563389077787801, + "grad_norm": 2.60174822807312, + "learning_rate": 0.00019362579954235157, + "loss": 0.9405, + "step": 8119 + }, + { + "epoch": 0.5634584692248976, + "grad_norm": 2.6448307037353516, + "learning_rate": 0.0001936238255007775, + "loss": 0.827, + "step": 8120 + }, + { + "epoch": 0.5635278606619943, + "grad_norm": 2.2070865631103516, + "learning_rate": 0.00019362185116364385, + "loss": 0.6679, + "step": 8121 + }, + { + "epoch": 0.563597252099091, + "grad_norm": 2.3356943130493164, + "learning_rate": 0.00019361987653095688, + "loss": 0.9312, + "step": 8122 + }, + { + "epoch": 0.5636666435361877, + "grad_norm": 3.8612873554229736, + "learning_rate": 0.00019361790160272283, + "loss": 0.9533, + "step": 8123 + }, + { + "epoch": 0.5637360349732843, + "grad_norm": 3.661313533782959, + "learning_rate": 0.00019361592637894793, + "loss": 1.0152, + "step": 8124 + }, + { + "epoch": 0.563805426410381, + "grad_norm": 2.5798208713531494, + "learning_rate": 0.0001936139508596384, + "loss": 0.9545, + "step": 8125 + }, + { + "epoch": 0.5638748178474776, + "grad_norm": 2.1936442852020264, + "learning_rate": 0.00019361197504480052, + "loss": 0.7765, + "step": 8126 + }, + { + "epoch": 0.5639442092845743, + "grad_norm": 2.4669320583343506, + "learning_rate": 0.00019360999893444048, + "loss": 0.867, + "step": 8127 + }, + { + "epoch": 0.564013600721671, + "grad_norm": 2.258206844329834, + "learning_rate": 0.00019360802252856452, + "loss": 0.9898, + "step": 8128 + }, + { + "epoch": 0.5640829921587676, + "grad_norm": 2.293498992919922, + "learning_rate": 0.0001936060458271789, + "loss": 0.6934, + "step": 8129 + }, + { + "epoch": 0.5641523835958643, + "grad_norm": 2.9374818801879883, + "learning_rate": 0.00019360406883028985, + "loss": 1.0239, + "step": 8130 + }, + { + "epoch": 0.5642217750329609, + "grad_norm": 2.750725269317627, + "learning_rate": 0.00019360209153790363, + "loss": 1.2925, + "step": 8131 + }, + { + "epoch": 0.5642911664700576, + "grad_norm": 2.6907989978790283, + "learning_rate": 0.00019360011395002643, + "loss": 0.8467, + "step": 8132 + }, + { + "epoch": 0.5643605579071542, + "grad_norm": 2.9166393280029297, + "learning_rate": 0.0001935981360666646, + "loss": 1.0055, + "step": 8133 + }, + { + "epoch": 0.564429949344251, + "grad_norm": 4.527013778686523, + "learning_rate": 0.00019359615788782427, + "loss": 1.1825, + "step": 8134 + }, + { + "epoch": 0.5644993407813476, + "grad_norm": 2.9226701259613037, + "learning_rate": 0.0001935941794135117, + "loss": 0.8979, + "step": 8135 + }, + { + "epoch": 0.5645687322184443, + "grad_norm": 3.3387961387634277, + "learning_rate": 0.0001935922006437332, + "loss": 0.8002, + "step": 8136 + }, + { + "epoch": 0.5646381236555409, + "grad_norm": 2.8665761947631836, + "learning_rate": 0.00019359022157849497, + "loss": 0.9138, + "step": 8137 + }, + { + "epoch": 0.5647075150926376, + "grad_norm": 3.507481098175049, + "learning_rate": 0.0001935882422178033, + "loss": 0.7366, + "step": 8138 + }, + { + "epoch": 0.5647769065297342, + "grad_norm": 2.7054500579833984, + "learning_rate": 0.00019358626256166437, + "loss": 0.8272, + "step": 8139 + }, + { + "epoch": 0.564846297966831, + "grad_norm": 3.1541225910186768, + "learning_rate": 0.00019358428261008449, + "loss": 0.8894, + "step": 8140 + }, + { + "epoch": 0.5649156894039276, + "grad_norm": 3.1628730297088623, + "learning_rate": 0.00019358230236306986, + "loss": 0.7788, + "step": 8141 + }, + { + "epoch": 0.5649850808410242, + "grad_norm": 2.4505624771118164, + "learning_rate": 0.0001935803218206268, + "loss": 0.8587, + "step": 8142 + }, + { + "epoch": 0.5650544722781209, + "grad_norm": 3.0035781860351562, + "learning_rate": 0.0001935783409827615, + "loss": 0.975, + "step": 8143 + }, + { + "epoch": 0.5651238637152175, + "grad_norm": 2.984875440597534, + "learning_rate": 0.0001935763598494802, + "loss": 0.767, + "step": 8144 + }, + { + "epoch": 0.5651932551523142, + "grad_norm": 1.9927504062652588, + "learning_rate": 0.00019357437842078924, + "loss": 0.6594, + "step": 8145 + }, + { + "epoch": 0.5652626465894108, + "grad_norm": 3.2723658084869385, + "learning_rate": 0.0001935723966966948, + "loss": 0.9151, + "step": 8146 + }, + { + "epoch": 0.5653320380265076, + "grad_norm": 2.468515634536743, + "learning_rate": 0.00019357041467720317, + "loss": 0.9208, + "step": 8147 + }, + { + "epoch": 0.5654014294636042, + "grad_norm": 3.496546506881714, + "learning_rate": 0.0001935684323623206, + "loss": 1.0222, + "step": 8148 + }, + { + "epoch": 0.5654708209007009, + "grad_norm": 3.1005146503448486, + "learning_rate": 0.0001935664497520533, + "loss": 0.8131, + "step": 8149 + }, + { + "epoch": 0.5655402123377975, + "grad_norm": 2.844923496246338, + "learning_rate": 0.0001935644668464076, + "loss": 0.8193, + "step": 8150 + }, + { + "epoch": 0.5656096037748942, + "grad_norm": 3.0385007858276367, + "learning_rate": 0.00019356248364538976, + "loss": 0.9756, + "step": 8151 + }, + { + "epoch": 0.5656789952119908, + "grad_norm": 3.327211618423462, + "learning_rate": 0.000193560500149006, + "loss": 1.0346, + "step": 8152 + }, + { + "epoch": 0.5657483866490876, + "grad_norm": 2.6028528213500977, + "learning_rate": 0.0001935585163572626, + "loss": 0.9, + "step": 8153 + }, + { + "epoch": 0.5658177780861842, + "grad_norm": 2.628812551498413, + "learning_rate": 0.00019355653227016582, + "loss": 0.7888, + "step": 8154 + }, + { + "epoch": 0.5658871695232808, + "grad_norm": 2.222228527069092, + "learning_rate": 0.00019355454788772192, + "loss": 0.7368, + "step": 8155 + }, + { + "epoch": 0.5659565609603775, + "grad_norm": 4.984938621520996, + "learning_rate": 0.00019355256320993715, + "loss": 0.874, + "step": 8156 + }, + { + "epoch": 0.5660259523974741, + "grad_norm": 2.6346218585968018, + "learning_rate": 0.0001935505782368178, + "loss": 0.7273, + "step": 8157 + }, + { + "epoch": 0.5660953438345708, + "grad_norm": 3.0524065494537354, + "learning_rate": 0.00019354859296837014, + "loss": 1.1015, + "step": 8158 + }, + { + "epoch": 0.5661647352716674, + "grad_norm": 2.639500141143799, + "learning_rate": 0.0001935466074046004, + "loss": 0.8981, + "step": 8159 + }, + { + "epoch": 0.5662341267087642, + "grad_norm": 3.6764917373657227, + "learning_rate": 0.00019354462154551493, + "loss": 1.1218, + "step": 8160 + }, + { + "epoch": 0.5663035181458608, + "grad_norm": 2.5742225646972656, + "learning_rate": 0.00019354263539111987, + "loss": 0.7282, + "step": 8161 + }, + { + "epoch": 0.5663729095829575, + "grad_norm": 3.0646908283233643, + "learning_rate": 0.00019354064894142162, + "loss": 1.0121, + "step": 8162 + }, + { + "epoch": 0.5664423010200541, + "grad_norm": 3.7643039226531982, + "learning_rate": 0.00019353866219642634, + "loss": 0.8584, + "step": 8163 + }, + { + "epoch": 0.5665116924571508, + "grad_norm": 2.47790265083313, + "learning_rate": 0.0001935366751561404, + "loss": 0.9877, + "step": 8164 + }, + { + "epoch": 0.5665810838942474, + "grad_norm": 2.4152791500091553, + "learning_rate": 0.00019353468782057004, + "loss": 1.0303, + "step": 8165 + }, + { + "epoch": 0.5666504753313442, + "grad_norm": 3.0142834186553955, + "learning_rate": 0.0001935327001897215, + "loss": 1.233, + "step": 8166 + }, + { + "epoch": 0.5667198667684408, + "grad_norm": 3.219521999359131, + "learning_rate": 0.00019353071226360107, + "loss": 1.0653, + "step": 8167 + }, + { + "epoch": 0.5667892582055374, + "grad_norm": 3.1098170280456543, + "learning_rate": 0.00019352872404221504, + "loss": 1.1555, + "step": 8168 + }, + { + "epoch": 0.5668586496426341, + "grad_norm": 3.2114384174346924, + "learning_rate": 0.00019352673552556967, + "loss": 0.9457, + "step": 8169 + }, + { + "epoch": 0.5669280410797307, + "grad_norm": 2.617830753326416, + "learning_rate": 0.00019352474671367124, + "loss": 0.8489, + "step": 8170 + }, + { + "epoch": 0.5669974325168274, + "grad_norm": 3.1760101318359375, + "learning_rate": 0.00019352275760652608, + "loss": 0.9365, + "step": 8171 + }, + { + "epoch": 0.567066823953924, + "grad_norm": 2.1850221157073975, + "learning_rate": 0.0001935207682041404, + "loss": 0.9094, + "step": 8172 + }, + { + "epoch": 0.5671362153910208, + "grad_norm": 2.879516124725342, + "learning_rate": 0.0001935187785065205, + "loss": 0.8956, + "step": 8173 + }, + { + "epoch": 0.5672056068281174, + "grad_norm": 3.8990302085876465, + "learning_rate": 0.00019351678851367265, + "loss": 0.7216, + "step": 8174 + }, + { + "epoch": 0.5672749982652141, + "grad_norm": 3.6159064769744873, + "learning_rate": 0.00019351479822560317, + "loss": 1.0038, + "step": 8175 + }, + { + "epoch": 0.5673443897023107, + "grad_norm": 2.9157567024230957, + "learning_rate": 0.00019351280764231833, + "loss": 0.8837, + "step": 8176 + }, + { + "epoch": 0.5674137811394074, + "grad_norm": 2.8344340324401855, + "learning_rate": 0.0001935108167638244, + "loss": 0.8595, + "step": 8177 + }, + { + "epoch": 0.567483172576504, + "grad_norm": 2.6379284858703613, + "learning_rate": 0.00019350882559012765, + "loss": 0.8581, + "step": 8178 + }, + { + "epoch": 0.5675525640136008, + "grad_norm": 2.8777685165405273, + "learning_rate": 0.0001935068341212344, + "loss": 0.7893, + "step": 8179 + }, + { + "epoch": 0.5676219554506974, + "grad_norm": 2.6769962310791016, + "learning_rate": 0.0001935048423571509, + "loss": 0.8783, + "step": 8180 + }, + { + "epoch": 0.5676913468877941, + "grad_norm": 2.8571531772613525, + "learning_rate": 0.0001935028502978835, + "loss": 0.8356, + "step": 8181 + }, + { + "epoch": 0.5677607383248907, + "grad_norm": 2.746001720428467, + "learning_rate": 0.00019350085794343845, + "loss": 0.8022, + "step": 8182 + }, + { + "epoch": 0.5678301297619873, + "grad_norm": 2.539762496948242, + "learning_rate": 0.000193498865293822, + "loss": 0.8897, + "step": 8183 + }, + { + "epoch": 0.567899521199084, + "grad_norm": 2.770881414413452, + "learning_rate": 0.00019349687234904053, + "loss": 0.9256, + "step": 8184 + }, + { + "epoch": 0.5679689126361807, + "grad_norm": 2.8013060092926025, + "learning_rate": 0.00019349487910910024, + "loss": 1.0584, + "step": 8185 + }, + { + "epoch": 0.5680383040732774, + "grad_norm": 2.472464084625244, + "learning_rate": 0.00019349288557400751, + "loss": 0.8837, + "step": 8186 + }, + { + "epoch": 0.568107695510374, + "grad_norm": 2.4555652141571045, + "learning_rate": 0.00019349089174376855, + "loss": 0.9087, + "step": 8187 + }, + { + "epoch": 0.5681770869474707, + "grad_norm": 3.143381357192993, + "learning_rate": 0.00019348889761838968, + "loss": 1.0314, + "step": 8188 + }, + { + "epoch": 0.5682464783845673, + "grad_norm": 3.5019524097442627, + "learning_rate": 0.00019348690319787726, + "loss": 0.8185, + "step": 8189 + }, + { + "epoch": 0.568315869821664, + "grad_norm": 2.643444538116455, + "learning_rate": 0.00019348490848223747, + "loss": 1.0895, + "step": 8190 + }, + { + "epoch": 0.5683852612587607, + "grad_norm": 3.0515522956848145, + "learning_rate": 0.0001934829134714767, + "loss": 0.8866, + "step": 8191 + }, + { + "epoch": 0.5684546526958574, + "grad_norm": 2.605943441390991, + "learning_rate": 0.00019348091816560125, + "loss": 0.8621, + "step": 8192 + }, + { + "epoch": 0.568524044132954, + "grad_norm": 2.8629791736602783, + "learning_rate": 0.00019347892256461736, + "loss": 0.9282, + "step": 8193 + }, + { + "epoch": 0.5685934355700507, + "grad_norm": 2.2003133296966553, + "learning_rate": 0.0001934769266685313, + "loss": 0.8084, + "step": 8194 + }, + { + "epoch": 0.5686628270071473, + "grad_norm": 2.555095911026001, + "learning_rate": 0.0001934749304773495, + "loss": 0.8259, + "step": 8195 + }, + { + "epoch": 0.5687322184442439, + "grad_norm": 2.387305498123169, + "learning_rate": 0.00019347293399107817, + "loss": 0.8337, + "step": 8196 + }, + { + "epoch": 0.5688016098813407, + "grad_norm": 3.054473400115967, + "learning_rate": 0.00019347093720972365, + "loss": 1.029, + "step": 8197 + }, + { + "epoch": 0.5688710013184373, + "grad_norm": 2.7035067081451416, + "learning_rate": 0.00019346894013329223, + "loss": 0.9884, + "step": 8198 + }, + { + "epoch": 0.568940392755534, + "grad_norm": 2.898646831512451, + "learning_rate": 0.00019346694276179015, + "loss": 1.1109, + "step": 8199 + }, + { + "epoch": 0.5690097841926306, + "grad_norm": 2.88297963142395, + "learning_rate": 0.00019346494509522383, + "loss": 1.1246, + "step": 8200 + }, + { + "epoch": 0.5690791756297273, + "grad_norm": 2.6456239223480225, + "learning_rate": 0.0001934629471335995, + "loss": 0.8064, + "step": 8201 + }, + { + "epoch": 0.5691485670668239, + "grad_norm": 2.462104320526123, + "learning_rate": 0.0001934609488769235, + "loss": 0.8739, + "step": 8202 + }, + { + "epoch": 0.5692179585039207, + "grad_norm": 2.7909164428710938, + "learning_rate": 0.00019345895032520213, + "loss": 0.9712, + "step": 8203 + }, + { + "epoch": 0.5692873499410173, + "grad_norm": 4.834717750549316, + "learning_rate": 0.0001934569514784417, + "loss": 1.0084, + "step": 8204 + }, + { + "epoch": 0.569356741378114, + "grad_norm": 3.8437280654907227, + "learning_rate": 0.0001934549523366485, + "loss": 1.0673, + "step": 8205 + }, + { + "epoch": 0.5694261328152106, + "grad_norm": 2.689606189727783, + "learning_rate": 0.00019345295289982888, + "loss": 0.7707, + "step": 8206 + }, + { + "epoch": 0.5694955242523073, + "grad_norm": 2.4533491134643555, + "learning_rate": 0.0001934509531679891, + "loss": 1.0062, + "step": 8207 + }, + { + "epoch": 0.5695649156894039, + "grad_norm": 2.3908939361572266, + "learning_rate": 0.00019344895314113553, + "loss": 0.909, + "step": 8208 + }, + { + "epoch": 0.5696343071265005, + "grad_norm": 3.093163013458252, + "learning_rate": 0.00019344695281927444, + "loss": 1.0352, + "step": 8209 + }, + { + "epoch": 0.5697036985635973, + "grad_norm": 3.3900253772735596, + "learning_rate": 0.00019344495220241216, + "loss": 1.2204, + "step": 8210 + }, + { + "epoch": 0.5697730900006939, + "grad_norm": 2.549466371536255, + "learning_rate": 0.00019344295129055503, + "loss": 1.0049, + "step": 8211 + }, + { + "epoch": 0.5698424814377906, + "grad_norm": 2.364375591278076, + "learning_rate": 0.00019344095008370932, + "loss": 1.0389, + "step": 8212 + }, + { + "epoch": 0.5699118728748872, + "grad_norm": 2.4206383228302, + "learning_rate": 0.0001934389485818814, + "loss": 0.9782, + "step": 8213 + }, + { + "epoch": 0.5699812643119839, + "grad_norm": 2.585893154144287, + "learning_rate": 0.00019343694678507754, + "loss": 0.853, + "step": 8214 + }, + { + "epoch": 0.5700506557490805, + "grad_norm": 1.9888672828674316, + "learning_rate": 0.0001934349446933041, + "loss": 0.7897, + "step": 8215 + }, + { + "epoch": 0.5701200471861773, + "grad_norm": 2.253683090209961, + "learning_rate": 0.00019343294230656736, + "loss": 0.9048, + "step": 8216 + }, + { + "epoch": 0.5701894386232739, + "grad_norm": 2.2724549770355225, + "learning_rate": 0.00019343093962487367, + "loss": 0.7099, + "step": 8217 + }, + { + "epoch": 0.5702588300603706, + "grad_norm": 2.6814091205596924, + "learning_rate": 0.00019342893664822932, + "loss": 0.9472, + "step": 8218 + }, + { + "epoch": 0.5703282214974672, + "grad_norm": 2.732969045639038, + "learning_rate": 0.0001934269333766407, + "loss": 0.9477, + "step": 8219 + }, + { + "epoch": 0.5703976129345639, + "grad_norm": 2.7604825496673584, + "learning_rate": 0.00019342492981011403, + "loss": 0.806, + "step": 8220 + }, + { + "epoch": 0.5704670043716605, + "grad_norm": 4.836461544036865, + "learning_rate": 0.00019342292594865575, + "loss": 1.2152, + "step": 8221 + }, + { + "epoch": 0.5705363958087571, + "grad_norm": 3.3504137992858887, + "learning_rate": 0.00019342092179227213, + "loss": 1.0402, + "step": 8222 + }, + { + "epoch": 0.5706057872458539, + "grad_norm": 2.7140908241271973, + "learning_rate": 0.00019341891734096946, + "loss": 0.9392, + "step": 8223 + }, + { + "epoch": 0.5706751786829505, + "grad_norm": 2.7551231384277344, + "learning_rate": 0.00019341691259475413, + "loss": 0.8891, + "step": 8224 + }, + { + "epoch": 0.5707445701200472, + "grad_norm": 2.71144700050354, + "learning_rate": 0.00019341490755363244, + "loss": 0.877, + "step": 8225 + }, + { + "epoch": 0.5708139615571438, + "grad_norm": 2.0610334873199463, + "learning_rate": 0.00019341290221761072, + "loss": 0.9844, + "step": 8226 + }, + { + "epoch": 0.5708833529942405, + "grad_norm": 3.007530927658081, + "learning_rate": 0.00019341089658669533, + "loss": 0.9519, + "step": 8227 + }, + { + "epoch": 0.5709527444313371, + "grad_norm": 2.947441577911377, + "learning_rate": 0.00019340889066089256, + "loss": 0.8568, + "step": 8228 + }, + { + "epoch": 0.5710221358684339, + "grad_norm": 3.2568609714508057, + "learning_rate": 0.00019340688444020879, + "loss": 0.7822, + "step": 8229 + }, + { + "epoch": 0.5710915273055305, + "grad_norm": 2.4525279998779297, + "learning_rate": 0.00019340487792465026, + "loss": 0.8035, + "step": 8230 + }, + { + "epoch": 0.5711609187426272, + "grad_norm": 3.030278444290161, + "learning_rate": 0.00019340287111422341, + "loss": 1.1552, + "step": 8231 + }, + { + "epoch": 0.5712303101797238, + "grad_norm": 2.6678357124328613, + "learning_rate": 0.0001934008640089345, + "loss": 0.9214, + "step": 8232 + }, + { + "epoch": 0.5712997016168205, + "grad_norm": 3.1764166355133057, + "learning_rate": 0.00019339885660878996, + "loss": 0.8516, + "step": 8233 + }, + { + "epoch": 0.5713690930539171, + "grad_norm": 2.170938014984131, + "learning_rate": 0.000193396848913796, + "loss": 0.8078, + "step": 8234 + }, + { + "epoch": 0.5714384844910138, + "grad_norm": 2.9851784706115723, + "learning_rate": 0.00019339484092395907, + "loss": 0.9284, + "step": 8235 + }, + { + "epoch": 0.5715078759281105, + "grad_norm": 3.02701735496521, + "learning_rate": 0.00019339283263928545, + "loss": 0.7446, + "step": 8236 + }, + { + "epoch": 0.5715772673652071, + "grad_norm": 2.8270156383514404, + "learning_rate": 0.00019339082405978148, + "loss": 0.7881, + "step": 8237 + }, + { + "epoch": 0.5716466588023038, + "grad_norm": 2.3729171752929688, + "learning_rate": 0.00019338881518545352, + "loss": 0.995, + "step": 8238 + }, + { + "epoch": 0.5717160502394004, + "grad_norm": 2.981358528137207, + "learning_rate": 0.00019338680601630793, + "loss": 0.992, + "step": 8239 + }, + { + "epoch": 0.5717854416764971, + "grad_norm": 3.8236029148101807, + "learning_rate": 0.00019338479655235097, + "loss": 1.024, + "step": 8240 + }, + { + "epoch": 0.5718548331135938, + "grad_norm": 2.1075022220611572, + "learning_rate": 0.0001933827867935891, + "loss": 0.8979, + "step": 8241 + }, + { + "epoch": 0.5719242245506905, + "grad_norm": 2.389854669570923, + "learning_rate": 0.0001933807767400286, + "loss": 0.5743, + "step": 8242 + }, + { + "epoch": 0.5719936159877871, + "grad_norm": 2.643488645553589, + "learning_rate": 0.00019337876639167578, + "loss": 0.9155, + "step": 8243 + }, + { + "epoch": 0.5720630074248838, + "grad_norm": 2.8758552074432373, + "learning_rate": 0.0001933767557485371, + "loss": 1.0053, + "step": 8244 + }, + { + "epoch": 0.5721323988619804, + "grad_norm": 2.7469851970672607, + "learning_rate": 0.00019337474481061876, + "loss": 0.8873, + "step": 8245 + }, + { + "epoch": 0.5722017902990771, + "grad_norm": 2.6719107627868652, + "learning_rate": 0.00019337273357792722, + "loss": 0.7799, + "step": 8246 + }, + { + "epoch": 0.5722711817361738, + "grad_norm": 2.704444169998169, + "learning_rate": 0.0001933707220504688, + "loss": 1.0922, + "step": 8247 + }, + { + "epoch": 0.5723405731732705, + "grad_norm": 2.992896795272827, + "learning_rate": 0.0001933687102282498, + "loss": 0.8883, + "step": 8248 + }, + { + "epoch": 0.5724099646103671, + "grad_norm": 2.7922401428222656, + "learning_rate": 0.00019336669811127668, + "loss": 0.9728, + "step": 8249 + }, + { + "epoch": 0.5724793560474637, + "grad_norm": 2.8461785316467285, + "learning_rate": 0.0001933646856995557, + "loss": 0.7227, + "step": 8250 + }, + { + "epoch": 0.5725487474845604, + "grad_norm": 2.387021541595459, + "learning_rate": 0.0001933626729930932, + "loss": 0.943, + "step": 8251 + }, + { + "epoch": 0.572618138921657, + "grad_norm": 3.102421283721924, + "learning_rate": 0.0001933606599918956, + "loss": 0.7125, + "step": 8252 + }, + { + "epoch": 0.5726875303587537, + "grad_norm": 2.8481225967407227, + "learning_rate": 0.00019335864669596926, + "loss": 0.9117, + "step": 8253 + }, + { + "epoch": 0.5727569217958504, + "grad_norm": 2.4291882514953613, + "learning_rate": 0.00019335663310532047, + "loss": 0.7079, + "step": 8254 + }, + { + "epoch": 0.5728263132329471, + "grad_norm": 2.6792030334472656, + "learning_rate": 0.00019335461921995563, + "loss": 0.8776, + "step": 8255 + }, + { + "epoch": 0.5728957046700437, + "grad_norm": 3.2592267990112305, + "learning_rate": 0.0001933526050398811, + "loss": 1.1484, + "step": 8256 + }, + { + "epoch": 0.5729650961071404, + "grad_norm": 2.616373300552368, + "learning_rate": 0.0001933505905651032, + "loss": 0.8176, + "step": 8257 + }, + { + "epoch": 0.573034487544237, + "grad_norm": 2.401266098022461, + "learning_rate": 0.00019334857579562833, + "loss": 0.7572, + "step": 8258 + }, + { + "epoch": 0.5731038789813337, + "grad_norm": 2.436253309249878, + "learning_rate": 0.00019334656073146284, + "loss": 0.9206, + "step": 8259 + }, + { + "epoch": 0.5731732704184304, + "grad_norm": 3.208585739135742, + "learning_rate": 0.0001933445453726131, + "loss": 0.9914, + "step": 8260 + }, + { + "epoch": 0.5732426618555271, + "grad_norm": 2.2324321269989014, + "learning_rate": 0.00019334252971908543, + "loss": 0.9099, + "step": 8261 + }, + { + "epoch": 0.5733120532926237, + "grad_norm": 2.563685178756714, + "learning_rate": 0.00019334051377088626, + "loss": 0.7649, + "step": 8262 + }, + { + "epoch": 0.5733814447297203, + "grad_norm": 3.9630379676818848, + "learning_rate": 0.00019333849752802186, + "loss": 1.1524, + "step": 8263 + }, + { + "epoch": 0.573450836166817, + "grad_norm": 2.895993709564209, + "learning_rate": 0.0001933364809904987, + "loss": 1.0148, + "step": 8264 + }, + { + "epoch": 0.5735202276039136, + "grad_norm": 2.7287025451660156, + "learning_rate": 0.0001933344641583231, + "loss": 0.9893, + "step": 8265 + }, + { + "epoch": 0.5735896190410104, + "grad_norm": 2.236076593399048, + "learning_rate": 0.0001933324470315014, + "loss": 0.9265, + "step": 8266 + }, + { + "epoch": 0.573659010478107, + "grad_norm": 2.503632068634033, + "learning_rate": 0.00019333042961004003, + "loss": 1.1046, + "step": 8267 + }, + { + "epoch": 0.5737284019152037, + "grad_norm": 3.404134750366211, + "learning_rate": 0.0001933284118939453, + "loss": 1.1209, + "step": 8268 + }, + { + "epoch": 0.5737977933523003, + "grad_norm": 2.9090957641601562, + "learning_rate": 0.00019332639388322358, + "loss": 1.0337, + "step": 8269 + }, + { + "epoch": 0.573867184789397, + "grad_norm": 2.398442029953003, + "learning_rate": 0.0001933243755778813, + "loss": 0.8555, + "step": 8270 + }, + { + "epoch": 0.5739365762264936, + "grad_norm": 2.7621114253997803, + "learning_rate": 0.0001933223569779248, + "loss": 0.9333, + "step": 8271 + }, + { + "epoch": 0.5740059676635904, + "grad_norm": 2.598327159881592, + "learning_rate": 0.00019332033808336038, + "loss": 0.848, + "step": 8272 + }, + { + "epoch": 0.574075359100687, + "grad_norm": 2.893481731414795, + "learning_rate": 0.00019331831889419455, + "loss": 1.1877, + "step": 8273 + }, + { + "epoch": 0.5741447505377837, + "grad_norm": 2.5933568477630615, + "learning_rate": 0.00019331629941043356, + "loss": 0.7626, + "step": 8274 + }, + { + "epoch": 0.5742141419748803, + "grad_norm": 2.703976631164551, + "learning_rate": 0.00019331427963208389, + "loss": 1.1258, + "step": 8275 + }, + { + "epoch": 0.5742835334119769, + "grad_norm": 2.783668279647827, + "learning_rate": 0.00019331225955915185, + "loss": 0.9701, + "step": 8276 + }, + { + "epoch": 0.5743529248490736, + "grad_norm": 2.4431421756744385, + "learning_rate": 0.0001933102391916438, + "loss": 0.8069, + "step": 8277 + }, + { + "epoch": 0.5744223162861702, + "grad_norm": 2.7866923809051514, + "learning_rate": 0.0001933082185295662, + "loss": 1.1158, + "step": 8278 + }, + { + "epoch": 0.574491707723267, + "grad_norm": 2.4446089267730713, + "learning_rate": 0.00019330619757292533, + "loss": 0.9564, + "step": 8279 + }, + { + "epoch": 0.5745610991603636, + "grad_norm": 2.2523250579833984, + "learning_rate": 0.00019330417632172764, + "loss": 0.8825, + "step": 8280 + }, + { + "epoch": 0.5746304905974603, + "grad_norm": 2.2961058616638184, + "learning_rate": 0.0001933021547759795, + "loss": 0.9522, + "step": 8281 + }, + { + "epoch": 0.5746998820345569, + "grad_norm": 2.6035075187683105, + "learning_rate": 0.0001933001329356873, + "loss": 0.8125, + "step": 8282 + }, + { + "epoch": 0.5747692734716536, + "grad_norm": 2.073591470718384, + "learning_rate": 0.00019329811080085735, + "loss": 0.6058, + "step": 8283 + }, + { + "epoch": 0.5748386649087502, + "grad_norm": 3.663728713989258, + "learning_rate": 0.00019329608837149613, + "loss": 0.866, + "step": 8284 + }, + { + "epoch": 0.574908056345847, + "grad_norm": 2.7161920070648193, + "learning_rate": 0.00019329406564760998, + "loss": 0.8267, + "step": 8285 + }, + { + "epoch": 0.5749774477829436, + "grad_norm": 2.421454906463623, + "learning_rate": 0.0001932920426292053, + "loss": 0.8689, + "step": 8286 + }, + { + "epoch": 0.5750468392200403, + "grad_norm": 2.873652696609497, + "learning_rate": 0.00019329001931628844, + "loss": 0.7295, + "step": 8287 + }, + { + "epoch": 0.5751162306571369, + "grad_norm": 3.529205799102783, + "learning_rate": 0.00019328799570886583, + "loss": 0.8246, + "step": 8288 + }, + { + "epoch": 0.5751856220942335, + "grad_norm": 2.8107056617736816, + "learning_rate": 0.00019328597180694383, + "loss": 0.909, + "step": 8289 + }, + { + "epoch": 0.5752550135313302, + "grad_norm": 2.875889539718628, + "learning_rate": 0.00019328394761052885, + "loss": 0.8879, + "step": 8290 + }, + { + "epoch": 0.5753244049684269, + "grad_norm": 3.3829123973846436, + "learning_rate": 0.00019328192311962724, + "loss": 1.281, + "step": 8291 + }, + { + "epoch": 0.5753937964055236, + "grad_norm": 3.8567771911621094, + "learning_rate": 0.00019327989833424546, + "loss": 0.9156, + "step": 8292 + }, + { + "epoch": 0.5754631878426202, + "grad_norm": 3.1083197593688965, + "learning_rate": 0.00019327787325438983, + "loss": 1.009, + "step": 8293 + }, + { + "epoch": 0.5755325792797169, + "grad_norm": 2.3691937923431396, + "learning_rate": 0.00019327584788006682, + "loss": 0.9008, + "step": 8294 + }, + { + "epoch": 0.5756019707168135, + "grad_norm": 3.5205702781677246, + "learning_rate": 0.00019327382221128275, + "loss": 1.2261, + "step": 8295 + }, + { + "epoch": 0.5756713621539102, + "grad_norm": 3.3965201377868652, + "learning_rate": 0.00019327179624804405, + "loss": 0.8832, + "step": 8296 + }, + { + "epoch": 0.5757407535910068, + "grad_norm": 2.659839153289795, + "learning_rate": 0.0001932697699903571, + "loss": 0.9755, + "step": 8297 + }, + { + "epoch": 0.5758101450281036, + "grad_norm": 3.218705177307129, + "learning_rate": 0.00019326774343822833, + "loss": 1.0936, + "step": 8298 + }, + { + "epoch": 0.5758795364652002, + "grad_norm": 2.843200922012329, + "learning_rate": 0.0001932657165916641, + "loss": 1.1234, + "step": 8299 + }, + { + "epoch": 0.5759489279022969, + "grad_norm": 2.370300769805908, + "learning_rate": 0.00019326368945067084, + "loss": 0.8896, + "step": 8300 + }, + { + "epoch": 0.5760183193393935, + "grad_norm": 3.084298849105835, + "learning_rate": 0.00019326166201525492, + "loss": 0.9752, + "step": 8301 + }, + { + "epoch": 0.5760877107764902, + "grad_norm": 1.968093752861023, + "learning_rate": 0.00019325963428542276, + "loss": 0.6916, + "step": 8302 + }, + { + "epoch": 0.5761571022135868, + "grad_norm": 2.587982416152954, + "learning_rate": 0.00019325760626118075, + "loss": 0.966, + "step": 8303 + }, + { + "epoch": 0.5762264936506835, + "grad_norm": 3.8081555366516113, + "learning_rate": 0.00019325557794253528, + "loss": 1.2224, + "step": 8304 + }, + { + "epoch": 0.5762958850877802, + "grad_norm": 2.7145180702209473, + "learning_rate": 0.00019325354932949278, + "loss": 0.8827, + "step": 8305 + }, + { + "epoch": 0.5763652765248768, + "grad_norm": 2.352846384048462, + "learning_rate": 0.00019325152042205966, + "loss": 0.7447, + "step": 8306 + }, + { + "epoch": 0.5764346679619735, + "grad_norm": 2.444950580596924, + "learning_rate": 0.0001932494912202423, + "loss": 0.841, + "step": 8307 + }, + { + "epoch": 0.5765040593990701, + "grad_norm": 2.458221435546875, + "learning_rate": 0.0001932474617240471, + "loss": 0.7192, + "step": 8308 + }, + { + "epoch": 0.5765734508361668, + "grad_norm": 2.875020980834961, + "learning_rate": 0.0001932454319334805, + "loss": 0.6719, + "step": 8309 + }, + { + "epoch": 0.5766428422732635, + "grad_norm": 3.7867136001586914, + "learning_rate": 0.0001932434018485489, + "loss": 0.9435, + "step": 8310 + }, + { + "epoch": 0.5767122337103602, + "grad_norm": 2.8721563816070557, + "learning_rate": 0.00019324137146925868, + "loss": 1.0331, + "step": 8311 + }, + { + "epoch": 0.5767816251474568, + "grad_norm": 2.3684420585632324, + "learning_rate": 0.00019323934079561625, + "loss": 0.8324, + "step": 8312 + }, + { + "epoch": 0.5768510165845535, + "grad_norm": 3.3490231037139893, + "learning_rate": 0.00019323730982762806, + "loss": 0.8453, + "step": 8313 + }, + { + "epoch": 0.5769204080216501, + "grad_norm": 4.924532413482666, + "learning_rate": 0.00019323527856530054, + "loss": 0.6933, + "step": 8314 + }, + { + "epoch": 0.5769897994587468, + "grad_norm": 2.520470142364502, + "learning_rate": 0.00019323324700864, + "loss": 0.7756, + "step": 8315 + }, + { + "epoch": 0.5770591908958435, + "grad_norm": 2.234856128692627, + "learning_rate": 0.00019323121515765293, + "loss": 0.7181, + "step": 8316 + }, + { + "epoch": 0.5771285823329401, + "grad_norm": 2.890957832336426, + "learning_rate": 0.00019322918301234576, + "loss": 0.9774, + "step": 8317 + }, + { + "epoch": 0.5771979737700368, + "grad_norm": 2.4609975814819336, + "learning_rate": 0.00019322715057272486, + "loss": 0.8848, + "step": 8318 + }, + { + "epoch": 0.5772673652071334, + "grad_norm": 2.3843600749969482, + "learning_rate": 0.00019322511783879666, + "loss": 0.8648, + "step": 8319 + }, + { + "epoch": 0.5773367566442301, + "grad_norm": 2.7062244415283203, + "learning_rate": 0.00019322308481056758, + "loss": 1.0886, + "step": 8320 + }, + { + "epoch": 0.5774061480813267, + "grad_norm": 2.2123353481292725, + "learning_rate": 0.00019322105148804404, + "loss": 0.9492, + "step": 8321 + }, + { + "epoch": 0.5774755395184235, + "grad_norm": 2.2141268253326416, + "learning_rate": 0.00019321901787123244, + "loss": 0.8228, + "step": 8322 + }, + { + "epoch": 0.5775449309555201, + "grad_norm": 3.1597747802734375, + "learning_rate": 0.00019321698396013924, + "loss": 0.9877, + "step": 8323 + }, + { + "epoch": 0.5776143223926168, + "grad_norm": 2.341153383255005, + "learning_rate": 0.00019321494975477083, + "loss": 1.0055, + "step": 8324 + }, + { + "epoch": 0.5776837138297134, + "grad_norm": 2.7653629779815674, + "learning_rate": 0.00019321291525513362, + "loss": 0.9041, + "step": 8325 + }, + { + "epoch": 0.5777531052668101, + "grad_norm": 2.8915886878967285, + "learning_rate": 0.0001932108804612341, + "loss": 0.8273, + "step": 8326 + }, + { + "epoch": 0.5778224967039067, + "grad_norm": 2.9290006160736084, + "learning_rate": 0.00019320884537307858, + "loss": 0.8622, + "step": 8327 + }, + { + "epoch": 0.5778918881410035, + "grad_norm": 2.7457897663116455, + "learning_rate": 0.00019320680999067357, + "loss": 0.813, + "step": 8328 + }, + { + "epoch": 0.5779612795781001, + "grad_norm": 2.3117547035217285, + "learning_rate": 0.00019320477431402548, + "loss": 0.7474, + "step": 8329 + }, + { + "epoch": 0.5780306710151967, + "grad_norm": 2.8612499237060547, + "learning_rate": 0.00019320273834314073, + "loss": 1.0095, + "step": 8330 + }, + { + "epoch": 0.5781000624522934, + "grad_norm": 2.4260637760162354, + "learning_rate": 0.00019320070207802577, + "loss": 0.6745, + "step": 8331 + }, + { + "epoch": 0.57816945388939, + "grad_norm": 3.6307363510131836, + "learning_rate": 0.00019319866551868698, + "loss": 0.9561, + "step": 8332 + }, + { + "epoch": 0.5782388453264867, + "grad_norm": 3.1280229091644287, + "learning_rate": 0.00019319662866513083, + "loss": 1.0832, + "step": 8333 + }, + { + "epoch": 0.5783082367635833, + "grad_norm": 4.382002830505371, + "learning_rate": 0.00019319459151736373, + "loss": 1.0485, + "step": 8334 + }, + { + "epoch": 0.5783776282006801, + "grad_norm": 2.880842924118042, + "learning_rate": 0.00019319255407539214, + "loss": 1.0817, + "step": 8335 + }, + { + "epoch": 0.5784470196377767, + "grad_norm": 2.1708714962005615, + "learning_rate": 0.00019319051633922242, + "loss": 0.8025, + "step": 8336 + }, + { + "epoch": 0.5785164110748734, + "grad_norm": 5.034912109375, + "learning_rate": 0.00019318847830886109, + "loss": 1.0829, + "step": 8337 + }, + { + "epoch": 0.57858580251197, + "grad_norm": 2.5666282176971436, + "learning_rate": 0.00019318643998431453, + "loss": 0.8929, + "step": 8338 + }, + { + "epoch": 0.5786551939490667, + "grad_norm": 2.9232661724090576, + "learning_rate": 0.0001931844013655892, + "loss": 1.151, + "step": 8339 + }, + { + "epoch": 0.5787245853861633, + "grad_norm": 2.4414196014404297, + "learning_rate": 0.00019318236245269153, + "loss": 0.916, + "step": 8340 + }, + { + "epoch": 0.5787939768232601, + "grad_norm": 3.3248963356018066, + "learning_rate": 0.00019318032324562795, + "loss": 1.0595, + "step": 8341 + }, + { + "epoch": 0.5788633682603567, + "grad_norm": 2.300658941268921, + "learning_rate": 0.0001931782837444049, + "loss": 0.7807, + "step": 8342 + }, + { + "epoch": 0.5789327596974533, + "grad_norm": 2.68717622756958, + "learning_rate": 0.0001931762439490288, + "loss": 0.8348, + "step": 8343 + }, + { + "epoch": 0.57900215113455, + "grad_norm": 2.5546188354492188, + "learning_rate": 0.00019317420385950613, + "loss": 1.1212, + "step": 8344 + }, + { + "epoch": 0.5790715425716466, + "grad_norm": 2.149773120880127, + "learning_rate": 0.0001931721634758433, + "loss": 0.6647, + "step": 8345 + }, + { + "epoch": 0.5791409340087433, + "grad_norm": 2.3354270458221436, + "learning_rate": 0.0001931701227980468, + "loss": 1.054, + "step": 8346 + }, + { + "epoch": 0.57921032544584, + "grad_norm": 3.030155897140503, + "learning_rate": 0.00019316808182612296, + "loss": 0.9529, + "step": 8347 + }, + { + "epoch": 0.5792797168829367, + "grad_norm": 2.6858744621276855, + "learning_rate": 0.0001931660405600783, + "loss": 1.2019, + "step": 8348 + }, + { + "epoch": 0.5793491083200333, + "grad_norm": 2.616173028945923, + "learning_rate": 0.0001931639989999193, + "loss": 1.1305, + "step": 8349 + }, + { + "epoch": 0.57941849975713, + "grad_norm": 2.697603225708008, + "learning_rate": 0.00019316195714565238, + "loss": 1.0685, + "step": 8350 + }, + { + "epoch": 0.5794878911942266, + "grad_norm": 3.684781551361084, + "learning_rate": 0.00019315991499728395, + "loss": 0.9683, + "step": 8351 + }, + { + "epoch": 0.5795572826313233, + "grad_norm": 2.4596123695373535, + "learning_rate": 0.00019315787255482045, + "loss": 0.7701, + "step": 8352 + }, + { + "epoch": 0.57962667406842, + "grad_norm": 1.8019459247589111, + "learning_rate": 0.00019315582981826838, + "loss": 0.7974, + "step": 8353 + }, + { + "epoch": 0.5796960655055167, + "grad_norm": 3.408327102661133, + "learning_rate": 0.00019315378678763417, + "loss": 1.0765, + "step": 8354 + }, + { + "epoch": 0.5797654569426133, + "grad_norm": 2.9254448413848877, + "learning_rate": 0.00019315174346292425, + "loss": 1.0901, + "step": 8355 + }, + { + "epoch": 0.5798348483797099, + "grad_norm": 1.8195663690567017, + "learning_rate": 0.00019314969984414506, + "loss": 0.6597, + "step": 8356 + }, + { + "epoch": 0.5799042398168066, + "grad_norm": 3.0349793434143066, + "learning_rate": 0.0001931476559313031, + "loss": 1.1255, + "step": 8357 + }, + { + "epoch": 0.5799736312539032, + "grad_norm": 2.044898271560669, + "learning_rate": 0.0001931456117244048, + "loss": 0.7256, + "step": 8358 + }, + { + "epoch": 0.5800430226909999, + "grad_norm": 2.4955334663391113, + "learning_rate": 0.00019314356722345663, + "loss": 0.694, + "step": 8359 + }, + { + "epoch": 0.5801124141280966, + "grad_norm": 2.434739351272583, + "learning_rate": 0.000193141522428465, + "loss": 1.0481, + "step": 8360 + }, + { + "epoch": 0.5801818055651933, + "grad_norm": 2.930882453918457, + "learning_rate": 0.0001931394773394364, + "loss": 0.9056, + "step": 8361 + }, + { + "epoch": 0.5802511970022899, + "grad_norm": 2.4209189414978027, + "learning_rate": 0.00019313743195637723, + "loss": 0.8861, + "step": 8362 + }, + { + "epoch": 0.5803205884393866, + "grad_norm": 2.1382405757904053, + "learning_rate": 0.00019313538627929405, + "loss": 0.6916, + "step": 8363 + }, + { + "epoch": 0.5803899798764832, + "grad_norm": 2.8242568969726562, + "learning_rate": 0.00019313334030819323, + "loss": 0.9803, + "step": 8364 + }, + { + "epoch": 0.5804593713135799, + "grad_norm": 2.257397174835205, + "learning_rate": 0.0001931312940430813, + "loss": 0.8399, + "step": 8365 + }, + { + "epoch": 0.5805287627506766, + "grad_norm": 2.391620397567749, + "learning_rate": 0.00019312924748396464, + "loss": 1.0066, + "step": 8366 + }, + { + "epoch": 0.5805981541877733, + "grad_norm": 2.391188621520996, + "learning_rate": 0.00019312720063084977, + "loss": 0.965, + "step": 8367 + }, + { + "epoch": 0.5806675456248699, + "grad_norm": 2.2820587158203125, + "learning_rate": 0.00019312515348374315, + "loss": 0.9782, + "step": 8368 + }, + { + "epoch": 0.5807369370619666, + "grad_norm": 2.2882258892059326, + "learning_rate": 0.00019312310604265118, + "loss": 0.8461, + "step": 8369 + }, + { + "epoch": 0.5808063284990632, + "grad_norm": 2.5109786987304688, + "learning_rate": 0.00019312105830758037, + "loss": 0.9208, + "step": 8370 + }, + { + "epoch": 0.5808757199361598, + "grad_norm": 2.5882487297058105, + "learning_rate": 0.00019311901027853722, + "loss": 0.9251, + "step": 8371 + }, + { + "epoch": 0.5809451113732566, + "grad_norm": 2.8084115982055664, + "learning_rate": 0.00019311696195552816, + "loss": 0.9496, + "step": 8372 + }, + { + "epoch": 0.5810145028103532, + "grad_norm": 2.781125783920288, + "learning_rate": 0.0001931149133385596, + "loss": 1.0717, + "step": 8373 + }, + { + "epoch": 0.5810838942474499, + "grad_norm": 2.08567214012146, + "learning_rate": 0.0001931128644276381, + "loss": 0.8065, + "step": 8374 + }, + { + "epoch": 0.5811532856845465, + "grad_norm": 3.4992573261260986, + "learning_rate": 0.0001931108152227701, + "loss": 1.0516, + "step": 8375 + }, + { + "epoch": 0.5812226771216432, + "grad_norm": 2.9169557094573975, + "learning_rate": 0.00019310876572396204, + "loss": 1.0426, + "step": 8376 + }, + { + "epoch": 0.5812920685587398, + "grad_norm": 2.6706314086914062, + "learning_rate": 0.00019310671593122044, + "loss": 1.0927, + "step": 8377 + }, + { + "epoch": 0.5813614599958365, + "grad_norm": 4.709832668304443, + "learning_rate": 0.0001931046658445517, + "loss": 1.1312, + "step": 8378 + }, + { + "epoch": 0.5814308514329332, + "grad_norm": 2.674438714981079, + "learning_rate": 0.00019310261546396236, + "loss": 1.0218, + "step": 8379 + }, + { + "epoch": 0.5815002428700299, + "grad_norm": 2.5376338958740234, + "learning_rate": 0.00019310056478945883, + "loss": 0.9229, + "step": 8380 + }, + { + "epoch": 0.5815696343071265, + "grad_norm": 2.772789239883423, + "learning_rate": 0.00019309851382104767, + "loss": 0.9012, + "step": 8381 + }, + { + "epoch": 0.5816390257442232, + "grad_norm": 2.4934020042419434, + "learning_rate": 0.00019309646255873526, + "loss": 0.9821, + "step": 8382 + }, + { + "epoch": 0.5817084171813198, + "grad_norm": 2.952693223953247, + "learning_rate": 0.00019309441100252812, + "loss": 1.0276, + "step": 8383 + }, + { + "epoch": 0.5817778086184164, + "grad_norm": 2.706639528274536, + "learning_rate": 0.00019309235915243274, + "loss": 0.8541, + "step": 8384 + }, + { + "epoch": 0.5818472000555132, + "grad_norm": 2.2507452964782715, + "learning_rate": 0.0001930903070084556, + "loss": 0.7641, + "step": 8385 + }, + { + "epoch": 0.5819165914926098, + "grad_norm": 2.0264174938201904, + "learning_rate": 0.00019308825457060313, + "loss": 0.8357, + "step": 8386 + }, + { + "epoch": 0.5819859829297065, + "grad_norm": 2.57651686668396, + "learning_rate": 0.00019308620183888187, + "loss": 0.8867, + "step": 8387 + }, + { + "epoch": 0.5820553743668031, + "grad_norm": 2.003552198410034, + "learning_rate": 0.00019308414881329827, + "loss": 0.6707, + "step": 8388 + }, + { + "epoch": 0.5821247658038998, + "grad_norm": 2.467785120010376, + "learning_rate": 0.00019308209549385877, + "loss": 0.8573, + "step": 8389 + }, + { + "epoch": 0.5821941572409964, + "grad_norm": 2.1648459434509277, + "learning_rate": 0.0001930800418805699, + "loss": 0.7136, + "step": 8390 + }, + { + "epoch": 0.5822635486780932, + "grad_norm": 2.49802827835083, + "learning_rate": 0.00019307798797343816, + "loss": 0.9223, + "step": 8391 + }, + { + "epoch": 0.5823329401151898, + "grad_norm": 2.5714638233184814, + "learning_rate": 0.00019307593377247004, + "loss": 0.8765, + "step": 8392 + }, + { + "epoch": 0.5824023315522865, + "grad_norm": 2.309406280517578, + "learning_rate": 0.00019307387927767195, + "loss": 0.993, + "step": 8393 + }, + { + "epoch": 0.5824717229893831, + "grad_norm": 2.9248600006103516, + "learning_rate": 0.00019307182448905045, + "loss": 0.8333, + "step": 8394 + }, + { + "epoch": 0.5825411144264798, + "grad_norm": 2.6307952404022217, + "learning_rate": 0.00019306976940661197, + "loss": 0.69, + "step": 8395 + }, + { + "epoch": 0.5826105058635764, + "grad_norm": 2.7034614086151123, + "learning_rate": 0.00019306771403036302, + "loss": 0.9051, + "step": 8396 + }, + { + "epoch": 0.582679897300673, + "grad_norm": 2.355919361114502, + "learning_rate": 0.00019306565836031014, + "loss": 0.7102, + "step": 8397 + }, + { + "epoch": 0.5827492887377698, + "grad_norm": 2.45509672164917, + "learning_rate": 0.00019306360239645975, + "loss": 0.705, + "step": 8398 + }, + { + "epoch": 0.5828186801748664, + "grad_norm": 3.2895309925079346, + "learning_rate": 0.00019306154613881837, + "loss": 0.8442, + "step": 8399 + }, + { + "epoch": 0.5828880716119631, + "grad_norm": 2.6835131645202637, + "learning_rate": 0.00019305948958739247, + "loss": 1.2527, + "step": 8400 + }, + { + "epoch": 0.5829574630490597, + "grad_norm": 2.4997758865356445, + "learning_rate": 0.00019305743274218858, + "loss": 1.0839, + "step": 8401 + }, + { + "epoch": 0.5830268544861564, + "grad_norm": 2.90895414352417, + "learning_rate": 0.00019305537560321316, + "loss": 0.8457, + "step": 8402 + }, + { + "epoch": 0.583096245923253, + "grad_norm": 2.761276960372925, + "learning_rate": 0.0001930533181704727, + "loss": 0.7249, + "step": 8403 + }, + { + "epoch": 0.5831656373603498, + "grad_norm": 3.0163755416870117, + "learning_rate": 0.00019305126044397375, + "loss": 0.9881, + "step": 8404 + }, + { + "epoch": 0.5832350287974464, + "grad_norm": 2.8862366676330566, + "learning_rate": 0.00019304920242372276, + "loss": 1.0367, + "step": 8405 + }, + { + "epoch": 0.5833044202345431, + "grad_norm": 3.066749095916748, + "learning_rate": 0.0001930471441097262, + "loss": 1.0718, + "step": 8406 + }, + { + "epoch": 0.5833738116716397, + "grad_norm": 2.5772953033447266, + "learning_rate": 0.0001930450855019906, + "loss": 0.9871, + "step": 8407 + }, + { + "epoch": 0.5834432031087364, + "grad_norm": 2.5319361686706543, + "learning_rate": 0.00019304302660052249, + "loss": 0.7523, + "step": 8408 + }, + { + "epoch": 0.583512594545833, + "grad_norm": 2.297734498977661, + "learning_rate": 0.00019304096740532833, + "loss": 1.0236, + "step": 8409 + }, + { + "epoch": 0.5835819859829297, + "grad_norm": 2.071794271469116, + "learning_rate": 0.00019303890791641463, + "loss": 0.8177, + "step": 8410 + }, + { + "epoch": 0.5836513774200264, + "grad_norm": 2.983093500137329, + "learning_rate": 0.00019303684813378792, + "loss": 0.8398, + "step": 8411 + }, + { + "epoch": 0.583720768857123, + "grad_norm": 2.2015161514282227, + "learning_rate": 0.0001930347880574546, + "loss": 0.8, + "step": 8412 + }, + { + "epoch": 0.5837901602942197, + "grad_norm": 2.9628570079803467, + "learning_rate": 0.0001930327276874213, + "loss": 0.8521, + "step": 8413 + }, + { + "epoch": 0.5838595517313163, + "grad_norm": 2.8136301040649414, + "learning_rate": 0.00019303066702369448, + "loss": 1.1283, + "step": 8414 + }, + { + "epoch": 0.583928943168413, + "grad_norm": 2.611520767211914, + "learning_rate": 0.0001930286060662806, + "loss": 1.0576, + "step": 8415 + }, + { + "epoch": 0.5839983346055097, + "grad_norm": 2.225564479827881, + "learning_rate": 0.00019302654481518624, + "loss": 0.7615, + "step": 8416 + }, + { + "epoch": 0.5840677260426064, + "grad_norm": 2.4304182529449463, + "learning_rate": 0.00019302448327041788, + "loss": 0.8184, + "step": 8417 + }, + { + "epoch": 0.584137117479703, + "grad_norm": 2.417982816696167, + "learning_rate": 0.000193022421431982, + "loss": 0.8999, + "step": 8418 + }, + { + "epoch": 0.5842065089167997, + "grad_norm": 2.6137514114379883, + "learning_rate": 0.0001930203592998851, + "loss": 0.8257, + "step": 8419 + }, + { + "epoch": 0.5842759003538963, + "grad_norm": 3.443821907043457, + "learning_rate": 0.00019301829687413377, + "loss": 0.8552, + "step": 8420 + }, + { + "epoch": 0.584345291790993, + "grad_norm": 2.6890363693237305, + "learning_rate": 0.00019301623415473443, + "loss": 1.0352, + "step": 8421 + }, + { + "epoch": 0.5844146832280896, + "grad_norm": 3.046147584915161, + "learning_rate": 0.00019301417114169365, + "loss": 0.8992, + "step": 8422 + }, + { + "epoch": 0.5844840746651864, + "grad_norm": 2.2076222896575928, + "learning_rate": 0.0001930121078350179, + "loss": 0.7844, + "step": 8423 + }, + { + "epoch": 0.584553466102283, + "grad_norm": 3.4916884899139404, + "learning_rate": 0.00019301004423471376, + "loss": 1.1864, + "step": 8424 + }, + { + "epoch": 0.5846228575393796, + "grad_norm": 3.490507125854492, + "learning_rate": 0.0001930079803407877, + "loss": 1.1941, + "step": 8425 + }, + { + "epoch": 0.5846922489764763, + "grad_norm": 2.2478737831115723, + "learning_rate": 0.00019300591615324618, + "loss": 0.8421, + "step": 8426 + }, + { + "epoch": 0.5847616404135729, + "grad_norm": 2.2921061515808105, + "learning_rate": 0.00019300385167209583, + "loss": 0.8238, + "step": 8427 + }, + { + "epoch": 0.5848310318506696, + "grad_norm": 2.005128860473633, + "learning_rate": 0.00019300178689734307, + "loss": 0.674, + "step": 8428 + }, + { + "epoch": 0.5849004232877663, + "grad_norm": 3.4698503017425537, + "learning_rate": 0.00019299972182899446, + "loss": 1.2127, + "step": 8429 + }, + { + "epoch": 0.584969814724863, + "grad_norm": 2.4753754138946533, + "learning_rate": 0.00019299765646705653, + "loss": 1.0968, + "step": 8430 + }, + { + "epoch": 0.5850392061619596, + "grad_norm": 2.750419855117798, + "learning_rate": 0.00019299559081153582, + "loss": 0.974, + "step": 8431 + }, + { + "epoch": 0.5851085975990563, + "grad_norm": 2.5785138607025146, + "learning_rate": 0.00019299352486243878, + "loss": 1.058, + "step": 8432 + }, + { + "epoch": 0.5851779890361529, + "grad_norm": 2.7292072772979736, + "learning_rate": 0.00019299145861977198, + "loss": 1.1058, + "step": 8433 + }, + { + "epoch": 0.5852473804732496, + "grad_norm": 2.9402835369110107, + "learning_rate": 0.00019298939208354195, + "loss": 1.0482, + "step": 8434 + }, + { + "epoch": 0.5853167719103463, + "grad_norm": 2.525700330734253, + "learning_rate": 0.00019298732525375518, + "loss": 0.9059, + "step": 8435 + }, + { + "epoch": 0.585386163347443, + "grad_norm": 3.397273540496826, + "learning_rate": 0.0001929852581304182, + "loss": 1.0769, + "step": 8436 + }, + { + "epoch": 0.5854555547845396, + "grad_norm": 2.4460644721984863, + "learning_rate": 0.00019298319071353758, + "loss": 0.8049, + "step": 8437 + }, + { + "epoch": 0.5855249462216362, + "grad_norm": 2.0959606170654297, + "learning_rate": 0.00019298112300311982, + "loss": 0.7787, + "step": 8438 + }, + { + "epoch": 0.5855943376587329, + "grad_norm": 2.607088565826416, + "learning_rate": 0.00019297905499917142, + "loss": 0.9546, + "step": 8439 + }, + { + "epoch": 0.5856637290958295, + "grad_norm": 2.841944932937622, + "learning_rate": 0.00019297698670169892, + "loss": 1.0764, + "step": 8440 + }, + { + "epoch": 0.5857331205329263, + "grad_norm": 2.345325231552124, + "learning_rate": 0.0001929749181107089, + "loss": 0.7255, + "step": 8441 + }, + { + "epoch": 0.5858025119700229, + "grad_norm": 2.06709623336792, + "learning_rate": 0.00019297284922620782, + "loss": 0.7401, + "step": 8442 + }, + { + "epoch": 0.5858719034071196, + "grad_norm": 1.9915173053741455, + "learning_rate": 0.00019297078004820224, + "loss": 0.7445, + "step": 8443 + }, + { + "epoch": 0.5859412948442162, + "grad_norm": 2.9179012775421143, + "learning_rate": 0.0001929687105766987, + "loss": 1.0101, + "step": 8444 + }, + { + "epoch": 0.5860106862813129, + "grad_norm": 2.5837724208831787, + "learning_rate": 0.00019296664081170373, + "loss": 0.7911, + "step": 8445 + }, + { + "epoch": 0.5860800777184095, + "grad_norm": 2.5387368202209473, + "learning_rate": 0.00019296457075322385, + "loss": 0.8085, + "step": 8446 + }, + { + "epoch": 0.5861494691555063, + "grad_norm": 2.944399356842041, + "learning_rate": 0.00019296250040126562, + "loss": 0.5959, + "step": 8447 + }, + { + "epoch": 0.5862188605926029, + "grad_norm": 3.1086175441741943, + "learning_rate": 0.00019296042975583557, + "loss": 1.1019, + "step": 8448 + }, + { + "epoch": 0.5862882520296996, + "grad_norm": 2.579967737197876, + "learning_rate": 0.0001929583588169402, + "loss": 0.8133, + "step": 8449 + }, + { + "epoch": 0.5863576434667962, + "grad_norm": 3.0208613872528076, + "learning_rate": 0.0001929562875845861, + "loss": 0.9207, + "step": 8450 + }, + { + "epoch": 0.5864270349038928, + "grad_norm": 2.9948792457580566, + "learning_rate": 0.0001929542160587798, + "loss": 1.2036, + "step": 8451 + }, + { + "epoch": 0.5864964263409895, + "grad_norm": 2.2985072135925293, + "learning_rate": 0.0001929521442395278, + "loss": 0.7829, + "step": 8452 + }, + { + "epoch": 0.5865658177780861, + "grad_norm": 2.5561792850494385, + "learning_rate": 0.00019295007212683665, + "loss": 1.0046, + "step": 8453 + }, + { + "epoch": 0.5866352092151829, + "grad_norm": 2.1802783012390137, + "learning_rate": 0.00019294799972071293, + "loss": 0.616, + "step": 8454 + }, + { + "epoch": 0.5867046006522795, + "grad_norm": 2.8151345252990723, + "learning_rate": 0.00019294592702116315, + "loss": 0.847, + "step": 8455 + }, + { + "epoch": 0.5867739920893762, + "grad_norm": 3.1053810119628906, + "learning_rate": 0.00019294385402819387, + "loss": 1.0196, + "step": 8456 + }, + { + "epoch": 0.5868433835264728, + "grad_norm": 2.8367533683776855, + "learning_rate": 0.0001929417807418116, + "loss": 0.8784, + "step": 8457 + }, + { + "epoch": 0.5869127749635695, + "grad_norm": 3.901236057281494, + "learning_rate": 0.00019293970716202297, + "loss": 0.8199, + "step": 8458 + }, + { + "epoch": 0.5869821664006661, + "grad_norm": 3.2112679481506348, + "learning_rate": 0.00019293763328883442, + "loss": 1.2013, + "step": 8459 + }, + { + "epoch": 0.5870515578377629, + "grad_norm": 2.728687047958374, + "learning_rate": 0.00019293555912225255, + "loss": 0.8619, + "step": 8460 + }, + { + "epoch": 0.5871209492748595, + "grad_norm": 3.2166974544525146, + "learning_rate": 0.0001929334846622839, + "loss": 0.8418, + "step": 8461 + }, + { + "epoch": 0.5871903407119562, + "grad_norm": 3.5076065063476562, + "learning_rate": 0.00019293140990893503, + "loss": 1.1076, + "step": 8462 + }, + { + "epoch": 0.5872597321490528, + "grad_norm": 3.80605411529541, + "learning_rate": 0.0001929293348622125, + "loss": 1.2622, + "step": 8463 + }, + { + "epoch": 0.5873291235861494, + "grad_norm": 2.599869966506958, + "learning_rate": 0.0001929272595221228, + "loss": 0.9251, + "step": 8464 + }, + { + "epoch": 0.5873985150232461, + "grad_norm": 4.407098293304443, + "learning_rate": 0.00019292518388867253, + "loss": 1.0012, + "step": 8465 + }, + { + "epoch": 0.5874679064603427, + "grad_norm": 3.4281492233276367, + "learning_rate": 0.00019292310796186824, + "loss": 0.9647, + "step": 8466 + }, + { + "epoch": 0.5875372978974395, + "grad_norm": 2.2732300758361816, + "learning_rate": 0.00019292103174171652, + "loss": 0.7907, + "step": 8467 + }, + { + "epoch": 0.5876066893345361, + "grad_norm": 2.6784183979034424, + "learning_rate": 0.00019291895522822384, + "loss": 1.0042, + "step": 8468 + }, + { + "epoch": 0.5876760807716328, + "grad_norm": 2.2573463916778564, + "learning_rate": 0.00019291687842139683, + "loss": 0.8395, + "step": 8469 + }, + { + "epoch": 0.5877454722087294, + "grad_norm": 3.6184158325195312, + "learning_rate": 0.00019291480132124198, + "loss": 0.9509, + "step": 8470 + }, + { + "epoch": 0.5878148636458261, + "grad_norm": 4.417115211486816, + "learning_rate": 0.0001929127239277659, + "loss": 1.0758, + "step": 8471 + }, + { + "epoch": 0.5878842550829227, + "grad_norm": 2.1360373497009277, + "learning_rate": 0.00019291064624097514, + "loss": 0.7353, + "step": 8472 + }, + { + "epoch": 0.5879536465200195, + "grad_norm": 3.7275922298431396, + "learning_rate": 0.00019290856826087625, + "loss": 1.0983, + "step": 8473 + }, + { + "epoch": 0.5880230379571161, + "grad_norm": 9.687392234802246, + "learning_rate": 0.0001929064899874758, + "loss": 0.935, + "step": 8474 + }, + { + "epoch": 0.5880924293942128, + "grad_norm": 4.306230068206787, + "learning_rate": 0.0001929044114207803, + "loss": 1.0313, + "step": 8475 + }, + { + "epoch": 0.5881618208313094, + "grad_norm": 3.434490442276001, + "learning_rate": 0.00019290233256079637, + "loss": 1.0712, + "step": 8476 + }, + { + "epoch": 0.5882312122684061, + "grad_norm": 2.879146099090576, + "learning_rate": 0.00019290025340753055, + "loss": 0.9569, + "step": 8477 + }, + { + "epoch": 0.5883006037055027, + "grad_norm": 3.16650652885437, + "learning_rate": 0.0001928981739609894, + "loss": 0.9218, + "step": 8478 + }, + { + "epoch": 0.5883699951425994, + "grad_norm": 3.8374667167663574, + "learning_rate": 0.00019289609422117951, + "loss": 0.8463, + "step": 8479 + }, + { + "epoch": 0.5884393865796961, + "grad_norm": 5.756945610046387, + "learning_rate": 0.00019289401418810744, + "loss": 1.1391, + "step": 8480 + }, + { + "epoch": 0.5885087780167927, + "grad_norm": 3.2231504917144775, + "learning_rate": 0.0001928919338617797, + "loss": 0.7753, + "step": 8481 + }, + { + "epoch": 0.5885781694538894, + "grad_norm": 3.50022292137146, + "learning_rate": 0.00019288985324220295, + "loss": 0.9629, + "step": 8482 + }, + { + "epoch": 0.588647560890986, + "grad_norm": 3.5229079723358154, + "learning_rate": 0.0001928877723293837, + "loss": 1.0837, + "step": 8483 + }, + { + "epoch": 0.5887169523280827, + "grad_norm": 4.654405117034912, + "learning_rate": 0.0001928856911233285, + "loss": 1.1513, + "step": 8484 + }, + { + "epoch": 0.5887863437651794, + "grad_norm": 2.7413687705993652, + "learning_rate": 0.00019288360962404398, + "loss": 0.9327, + "step": 8485 + }, + { + "epoch": 0.5888557352022761, + "grad_norm": 2.7196881771087646, + "learning_rate": 0.0001928815278315367, + "loss": 0.8017, + "step": 8486 + }, + { + "epoch": 0.5889251266393727, + "grad_norm": 3.4911718368530273, + "learning_rate": 0.0001928794457458132, + "loss": 1.0183, + "step": 8487 + }, + { + "epoch": 0.5889945180764694, + "grad_norm": 3.0946426391601562, + "learning_rate": 0.00019287736336688003, + "loss": 0.7096, + "step": 8488 + }, + { + "epoch": 0.589063909513566, + "grad_norm": 3.8004961013793945, + "learning_rate": 0.0001928752806947438, + "loss": 1.1319, + "step": 8489 + }, + { + "epoch": 0.5891333009506627, + "grad_norm": 2.4168813228607178, + "learning_rate": 0.0001928731977294111, + "loss": 0.5754, + "step": 8490 + }, + { + "epoch": 0.5892026923877594, + "grad_norm": 2.3140456676483154, + "learning_rate": 0.00019287111447088853, + "loss": 0.7661, + "step": 8491 + }, + { + "epoch": 0.589272083824856, + "grad_norm": 3.57504940032959, + "learning_rate": 0.00019286903091918258, + "loss": 1.1778, + "step": 8492 + }, + { + "epoch": 0.5893414752619527, + "grad_norm": 2.9068033695220947, + "learning_rate": 0.0001928669470742999, + "loss": 1.0704, + "step": 8493 + }, + { + "epoch": 0.5894108666990493, + "grad_norm": 2.812718629837036, + "learning_rate": 0.000192864862936247, + "loss": 0.8329, + "step": 8494 + }, + { + "epoch": 0.589480258136146, + "grad_norm": 2.9196383953094482, + "learning_rate": 0.00019286277850503054, + "loss": 0.7854, + "step": 8495 + }, + { + "epoch": 0.5895496495732426, + "grad_norm": 3.481131076812744, + "learning_rate": 0.00019286069378065705, + "loss": 1.064, + "step": 8496 + }, + { + "epoch": 0.5896190410103394, + "grad_norm": 2.9259872436523438, + "learning_rate": 0.00019285860876313313, + "loss": 1.0909, + "step": 8497 + }, + { + "epoch": 0.589688432447436, + "grad_norm": 2.94732928276062, + "learning_rate": 0.00019285652345246539, + "loss": 0.9447, + "step": 8498 + }, + { + "epoch": 0.5897578238845327, + "grad_norm": 2.365687608718872, + "learning_rate": 0.0001928544378486603, + "loss": 0.8467, + "step": 8499 + }, + { + "epoch": 0.5898272153216293, + "grad_norm": 2.3896467685699463, + "learning_rate": 0.0001928523519517246, + "loss": 1.1787, + "step": 8500 + }, + { + "epoch": 0.589896606758726, + "grad_norm": 2.6713597774505615, + "learning_rate": 0.00019285026576166474, + "loss": 0.7569, + "step": 8501 + }, + { + "epoch": 0.5899659981958226, + "grad_norm": 2.6218807697296143, + "learning_rate": 0.0001928481792784874, + "loss": 0.8823, + "step": 8502 + }, + { + "epoch": 0.5900353896329193, + "grad_norm": 2.3560664653778076, + "learning_rate": 0.00019284609250219911, + "loss": 0.938, + "step": 8503 + }, + { + "epoch": 0.590104781070016, + "grad_norm": 2.0977084636688232, + "learning_rate": 0.0001928440054328065, + "loss": 0.7935, + "step": 8504 + }, + { + "epoch": 0.5901741725071126, + "grad_norm": 1.8214640617370605, + "learning_rate": 0.0001928419180703161, + "loss": 0.7787, + "step": 8505 + }, + { + "epoch": 0.5902435639442093, + "grad_norm": 2.378657817840576, + "learning_rate": 0.00019283983041473458, + "loss": 1.1028, + "step": 8506 + }, + { + "epoch": 0.5903129553813059, + "grad_norm": 2.489722728729248, + "learning_rate": 0.00019283774246606848, + "loss": 0.8929, + "step": 8507 + }, + { + "epoch": 0.5903823468184026, + "grad_norm": 2.56048583984375, + "learning_rate": 0.00019283565422432438, + "loss": 0.9472, + "step": 8508 + }, + { + "epoch": 0.5904517382554992, + "grad_norm": 2.760193109512329, + "learning_rate": 0.00019283356568950888, + "loss": 0.8468, + "step": 8509 + }, + { + "epoch": 0.590521129692596, + "grad_norm": 2.4606094360351562, + "learning_rate": 0.00019283147686162858, + "loss": 0.8527, + "step": 8510 + }, + { + "epoch": 0.5905905211296926, + "grad_norm": 2.91841983795166, + "learning_rate": 0.0001928293877406901, + "loss": 0.7893, + "step": 8511 + }, + { + "epoch": 0.5906599125667893, + "grad_norm": 2.3690884113311768, + "learning_rate": 0.0001928272983267, + "loss": 0.937, + "step": 8512 + }, + { + "epoch": 0.5907293040038859, + "grad_norm": 2.4188947677612305, + "learning_rate": 0.0001928252086196649, + "loss": 0.7808, + "step": 8513 + }, + { + "epoch": 0.5907986954409826, + "grad_norm": 3.126373767852783, + "learning_rate": 0.0001928231186195914, + "loss": 0.8661, + "step": 8514 + }, + { + "epoch": 0.5908680868780792, + "grad_norm": 2.346328020095825, + "learning_rate": 0.00019282102832648604, + "loss": 0.7265, + "step": 8515 + }, + { + "epoch": 0.590937478315176, + "grad_norm": 2.774747371673584, + "learning_rate": 0.00019281893774035554, + "loss": 0.8496, + "step": 8516 + }, + { + "epoch": 0.5910068697522726, + "grad_norm": 2.4441514015197754, + "learning_rate": 0.00019281684686120633, + "loss": 0.8873, + "step": 8517 + }, + { + "epoch": 0.5910762611893692, + "grad_norm": 2.559532403945923, + "learning_rate": 0.00019281475568904517, + "loss": 0.9286, + "step": 8518 + }, + { + "epoch": 0.5911456526264659, + "grad_norm": 2.2462317943573, + "learning_rate": 0.00019281266422387856, + "loss": 0.7784, + "step": 8519 + }, + { + "epoch": 0.5912150440635625, + "grad_norm": 2.0331971645355225, + "learning_rate": 0.00019281057246571317, + "loss": 0.6973, + "step": 8520 + }, + { + "epoch": 0.5912844355006592, + "grad_norm": 3.0400846004486084, + "learning_rate": 0.00019280848041455554, + "loss": 0.895, + "step": 8521 + }, + { + "epoch": 0.5913538269377558, + "grad_norm": 2.3638410568237305, + "learning_rate": 0.00019280638807041232, + "loss": 0.9894, + "step": 8522 + }, + { + "epoch": 0.5914232183748526, + "grad_norm": 2.9177606105804443, + "learning_rate": 0.0001928042954332901, + "loss": 0.801, + "step": 8523 + }, + { + "epoch": 0.5914926098119492, + "grad_norm": 2.9773459434509277, + "learning_rate": 0.00019280220250319547, + "loss": 0.8207, + "step": 8524 + }, + { + "epoch": 0.5915620012490459, + "grad_norm": 2.3210232257843018, + "learning_rate": 0.00019280010928013507, + "loss": 0.7583, + "step": 8525 + }, + { + "epoch": 0.5916313926861425, + "grad_norm": 2.77803635597229, + "learning_rate": 0.00019279801576411546, + "loss": 0.8499, + "step": 8526 + }, + { + "epoch": 0.5917007841232392, + "grad_norm": 2.806184768676758, + "learning_rate": 0.0001927959219551433, + "loss": 0.7077, + "step": 8527 + }, + { + "epoch": 0.5917701755603358, + "grad_norm": 2.2276759147644043, + "learning_rate": 0.00019279382785322523, + "loss": 0.8261, + "step": 8528 + }, + { + "epoch": 0.5918395669974326, + "grad_norm": 2.72516131401062, + "learning_rate": 0.00019279173345836776, + "loss": 0.6546, + "step": 8529 + }, + { + "epoch": 0.5919089584345292, + "grad_norm": 2.596099853515625, + "learning_rate": 0.00019278963877057757, + "loss": 0.9956, + "step": 8530 + }, + { + "epoch": 0.5919783498716258, + "grad_norm": 3.2704896926879883, + "learning_rate": 0.00019278754378986124, + "loss": 0.9266, + "step": 8531 + }, + { + "epoch": 0.5920477413087225, + "grad_norm": 2.2735674381256104, + "learning_rate": 0.00019278544851622539, + "loss": 0.7411, + "step": 8532 + }, + { + "epoch": 0.5921171327458191, + "grad_norm": 2.802419900894165, + "learning_rate": 0.00019278335294967667, + "loss": 1.1361, + "step": 8533 + }, + { + "epoch": 0.5921865241829158, + "grad_norm": 3.074763059616089, + "learning_rate": 0.00019278125709022167, + "loss": 1.1727, + "step": 8534 + }, + { + "epoch": 0.5922559156200125, + "grad_norm": 2.8359827995300293, + "learning_rate": 0.000192779160937867, + "loss": 0.8657, + "step": 8535 + }, + { + "epoch": 0.5923253070571092, + "grad_norm": 2.4271979331970215, + "learning_rate": 0.00019277706449261927, + "loss": 0.8867, + "step": 8536 + }, + { + "epoch": 0.5923946984942058, + "grad_norm": 2.31729793548584, + "learning_rate": 0.00019277496775448513, + "loss": 0.7946, + "step": 8537 + }, + { + "epoch": 0.5924640899313025, + "grad_norm": 3.015618085861206, + "learning_rate": 0.00019277287072347118, + "loss": 0.8587, + "step": 8538 + }, + { + "epoch": 0.5925334813683991, + "grad_norm": 4.0741777420043945, + "learning_rate": 0.00019277077339958405, + "loss": 1.0986, + "step": 8539 + }, + { + "epoch": 0.5926028728054958, + "grad_norm": 2.6298232078552246, + "learning_rate": 0.00019276867578283034, + "loss": 0.8765, + "step": 8540 + }, + { + "epoch": 0.5926722642425925, + "grad_norm": 2.941629409790039, + "learning_rate": 0.00019276657787321665, + "loss": 0.8892, + "step": 8541 + }, + { + "epoch": 0.5927416556796892, + "grad_norm": 2.599637746810913, + "learning_rate": 0.00019276447967074968, + "loss": 0.9529, + "step": 8542 + }, + { + "epoch": 0.5928110471167858, + "grad_norm": 2.242844581604004, + "learning_rate": 0.000192762381175436, + "loss": 0.7468, + "step": 8543 + }, + { + "epoch": 0.5928804385538825, + "grad_norm": 3.68053936958313, + "learning_rate": 0.00019276028238728222, + "loss": 0.8659, + "step": 8544 + }, + { + "epoch": 0.5929498299909791, + "grad_norm": 3.115936040878296, + "learning_rate": 0.00019275818330629502, + "loss": 0.8565, + "step": 8545 + }, + { + "epoch": 0.5930192214280757, + "grad_norm": 1.9932069778442383, + "learning_rate": 0.00019275608393248098, + "loss": 0.7436, + "step": 8546 + }, + { + "epoch": 0.5930886128651724, + "grad_norm": 2.5649845600128174, + "learning_rate": 0.0001927539842658468, + "loss": 0.6548, + "step": 8547 + }, + { + "epoch": 0.5931580043022691, + "grad_norm": 2.745090961456299, + "learning_rate": 0.00019275188430639898, + "loss": 0.9753, + "step": 8548 + }, + { + "epoch": 0.5932273957393658, + "grad_norm": 2.301892042160034, + "learning_rate": 0.00019274978405414426, + "loss": 0.8117, + "step": 8549 + }, + { + "epoch": 0.5932967871764624, + "grad_norm": 2.890622615814209, + "learning_rate": 0.00019274768350908923, + "loss": 0.6529, + "step": 8550 + }, + { + "epoch": 0.5933661786135591, + "grad_norm": 2.424933910369873, + "learning_rate": 0.0001927455826712405, + "loss": 0.8255, + "step": 8551 + }, + { + "epoch": 0.5934355700506557, + "grad_norm": 2.830141067504883, + "learning_rate": 0.00019274348154060475, + "loss": 0.8277, + "step": 8552 + }, + { + "epoch": 0.5935049614877524, + "grad_norm": 2.404308557510376, + "learning_rate": 0.00019274138011718858, + "loss": 0.9021, + "step": 8553 + }, + { + "epoch": 0.5935743529248491, + "grad_norm": 3.097979784011841, + "learning_rate": 0.00019273927840099866, + "loss": 0.9505, + "step": 8554 + }, + { + "epoch": 0.5936437443619458, + "grad_norm": 2.3861615657806396, + "learning_rate": 0.00019273717639204155, + "loss": 0.6535, + "step": 8555 + }, + { + "epoch": 0.5937131357990424, + "grad_norm": 2.4518444538116455, + "learning_rate": 0.00019273507409032397, + "loss": 0.7939, + "step": 8556 + }, + { + "epoch": 0.5937825272361391, + "grad_norm": 2.5114798545837402, + "learning_rate": 0.0001927329714958525, + "loss": 0.9237, + "step": 8557 + }, + { + "epoch": 0.5938519186732357, + "grad_norm": 2.1452715396881104, + "learning_rate": 0.0001927308686086338, + "loss": 0.7917, + "step": 8558 + }, + { + "epoch": 0.5939213101103323, + "grad_norm": 2.299055337905884, + "learning_rate": 0.00019272876542867452, + "loss": 0.8225, + "step": 8559 + }, + { + "epoch": 0.5939907015474291, + "grad_norm": 3.3238353729248047, + "learning_rate": 0.0001927266619559813, + "loss": 1.1691, + "step": 8560 + }, + { + "epoch": 0.5940600929845257, + "grad_norm": 3.8132741451263428, + "learning_rate": 0.00019272455819056072, + "loss": 1.0383, + "step": 8561 + }, + { + "epoch": 0.5941294844216224, + "grad_norm": 3.395245313644409, + "learning_rate": 0.00019272245413241948, + "loss": 0.8179, + "step": 8562 + }, + { + "epoch": 0.594198875858719, + "grad_norm": 2.780064344406128, + "learning_rate": 0.00019272034978156424, + "loss": 0.7733, + "step": 8563 + }, + { + "epoch": 0.5942682672958157, + "grad_norm": 2.46128249168396, + "learning_rate": 0.0001927182451380016, + "loss": 0.7307, + "step": 8564 + }, + { + "epoch": 0.5943376587329123, + "grad_norm": 1.7650001049041748, + "learning_rate": 0.00019271614020173822, + "loss": 0.7615, + "step": 8565 + }, + { + "epoch": 0.594407050170009, + "grad_norm": 2.5683515071868896, + "learning_rate": 0.00019271403497278075, + "loss": 1.0578, + "step": 8566 + }, + { + "epoch": 0.5944764416071057, + "grad_norm": 2.837825298309326, + "learning_rate": 0.0001927119294511358, + "loss": 0.9772, + "step": 8567 + }, + { + "epoch": 0.5945458330442024, + "grad_norm": 2.2808151245117188, + "learning_rate": 0.00019270982363681001, + "loss": 0.8498, + "step": 8568 + }, + { + "epoch": 0.594615224481299, + "grad_norm": 2.9916446208953857, + "learning_rate": 0.00019270771752981014, + "loss": 0.9489, + "step": 8569 + }, + { + "epoch": 0.5946846159183957, + "grad_norm": 3.279041290283203, + "learning_rate": 0.00019270561113014274, + "loss": 1.0308, + "step": 8570 + }, + { + "epoch": 0.5947540073554923, + "grad_norm": 3.821135997772217, + "learning_rate": 0.00019270350443781448, + "loss": 0.945, + "step": 8571 + }, + { + "epoch": 0.5948233987925889, + "grad_norm": 3.8915035724639893, + "learning_rate": 0.000192701397452832, + "loss": 0.9591, + "step": 8572 + }, + { + "epoch": 0.5948927902296857, + "grad_norm": 2.6162233352661133, + "learning_rate": 0.00019269929017520196, + "loss": 0.8202, + "step": 8573 + }, + { + "epoch": 0.5949621816667823, + "grad_norm": 2.5115957260131836, + "learning_rate": 0.00019269718260493104, + "loss": 0.8485, + "step": 8574 + }, + { + "epoch": 0.595031573103879, + "grad_norm": 2.5380337238311768, + "learning_rate": 0.00019269507474202583, + "loss": 0.9696, + "step": 8575 + }, + { + "epoch": 0.5951009645409756, + "grad_norm": 3.040891170501709, + "learning_rate": 0.00019269296658649306, + "loss": 1.0741, + "step": 8576 + }, + { + "epoch": 0.5951703559780723, + "grad_norm": 2.7395594120025635, + "learning_rate": 0.00019269085813833936, + "loss": 0.8977, + "step": 8577 + }, + { + "epoch": 0.5952397474151689, + "grad_norm": 3.561901807785034, + "learning_rate": 0.00019268874939757136, + "loss": 0.942, + "step": 8578 + }, + { + "epoch": 0.5953091388522657, + "grad_norm": 1.9902565479278564, + "learning_rate": 0.00019268664036419572, + "loss": 0.8483, + "step": 8579 + }, + { + "epoch": 0.5953785302893623, + "grad_norm": 2.3323802947998047, + "learning_rate": 0.0001926845310382191, + "loss": 0.5912, + "step": 8580 + }, + { + "epoch": 0.595447921726459, + "grad_norm": 2.6591291427612305, + "learning_rate": 0.00019268242141964818, + "loss": 0.9443, + "step": 8581 + }, + { + "epoch": 0.5955173131635556, + "grad_norm": 2.3556480407714844, + "learning_rate": 0.00019268031150848964, + "loss": 0.7861, + "step": 8582 + }, + { + "epoch": 0.5955867046006523, + "grad_norm": 3.0372378826141357, + "learning_rate": 0.00019267820130475008, + "loss": 0.9194, + "step": 8583 + }, + { + "epoch": 0.5956560960377489, + "grad_norm": 2.6540260314941406, + "learning_rate": 0.00019267609080843618, + "loss": 0.902, + "step": 8584 + }, + { + "epoch": 0.5957254874748455, + "grad_norm": 2.5173261165618896, + "learning_rate": 0.00019267398001955463, + "loss": 0.9006, + "step": 8585 + }, + { + "epoch": 0.5957948789119423, + "grad_norm": 4.151260852813721, + "learning_rate": 0.00019267186893811208, + "loss": 1.0128, + "step": 8586 + }, + { + "epoch": 0.5958642703490389, + "grad_norm": 3.4122748374938965, + "learning_rate": 0.0001926697575641152, + "loss": 0.8406, + "step": 8587 + }, + { + "epoch": 0.5959336617861356, + "grad_norm": 3.0176196098327637, + "learning_rate": 0.0001926676458975706, + "loss": 0.8917, + "step": 8588 + }, + { + "epoch": 0.5960030532232322, + "grad_norm": 2.812944173812866, + "learning_rate": 0.00019266553393848505, + "loss": 1.246, + "step": 8589 + }, + { + "epoch": 0.5960724446603289, + "grad_norm": 2.5741240978240967, + "learning_rate": 0.00019266342168686512, + "loss": 1.0403, + "step": 8590 + }, + { + "epoch": 0.5961418360974255, + "grad_norm": 3.4694924354553223, + "learning_rate": 0.00019266130914271753, + "loss": 1.0942, + "step": 8591 + }, + { + "epoch": 0.5962112275345223, + "grad_norm": 3.4078989028930664, + "learning_rate": 0.00019265919630604897, + "loss": 1.0508, + "step": 8592 + }, + { + "epoch": 0.5962806189716189, + "grad_norm": 2.8989477157592773, + "learning_rate": 0.00019265708317686605, + "loss": 0.8906, + "step": 8593 + }, + { + "epoch": 0.5963500104087156, + "grad_norm": 2.2913074493408203, + "learning_rate": 0.00019265496975517547, + "loss": 0.8624, + "step": 8594 + }, + { + "epoch": 0.5964194018458122, + "grad_norm": 2.8205623626708984, + "learning_rate": 0.0001926528560409839, + "loss": 1.0125, + "step": 8595 + }, + { + "epoch": 0.5964887932829089, + "grad_norm": 2.816572666168213, + "learning_rate": 0.00019265074203429801, + "loss": 0.7629, + "step": 8596 + }, + { + "epoch": 0.5965581847200055, + "grad_norm": 3.1388211250305176, + "learning_rate": 0.00019264862773512448, + "loss": 0.9171, + "step": 8597 + }, + { + "epoch": 0.5966275761571023, + "grad_norm": 2.576235294342041, + "learning_rate": 0.00019264651314347, + "loss": 1.0671, + "step": 8598 + }, + { + "epoch": 0.5966969675941989, + "grad_norm": 2.56272554397583, + "learning_rate": 0.0001926443982593412, + "loss": 0.877, + "step": 8599 + }, + { + "epoch": 0.5967663590312955, + "grad_norm": 2.161564826965332, + "learning_rate": 0.0001926422830827448, + "loss": 0.7786, + "step": 8600 + }, + { + "epoch": 0.5968357504683922, + "grad_norm": 2.371515989303589, + "learning_rate": 0.00019264016761368744, + "loss": 0.8134, + "step": 8601 + }, + { + "epoch": 0.5969051419054888, + "grad_norm": 2.861745595932007, + "learning_rate": 0.00019263805185217584, + "loss": 0.8572, + "step": 8602 + }, + { + "epoch": 0.5969745333425855, + "grad_norm": 2.708872079849243, + "learning_rate": 0.00019263593579821664, + "loss": 0.8932, + "step": 8603 + }, + { + "epoch": 0.5970439247796822, + "grad_norm": 3.6494338512420654, + "learning_rate": 0.00019263381945181657, + "loss": 1.0689, + "step": 8604 + }, + { + "epoch": 0.5971133162167789, + "grad_norm": 2.654083728790283, + "learning_rate": 0.00019263170281298226, + "loss": 0.9147, + "step": 8605 + }, + { + "epoch": 0.5971827076538755, + "grad_norm": 2.397747278213501, + "learning_rate": 0.0001926295858817204, + "loss": 0.7397, + "step": 8606 + }, + { + "epoch": 0.5972520990909722, + "grad_norm": 3.248490333557129, + "learning_rate": 0.00019262746865803768, + "loss": 1.0306, + "step": 8607 + }, + { + "epoch": 0.5973214905280688, + "grad_norm": 3.084585666656494, + "learning_rate": 0.00019262535114194083, + "loss": 0.9592, + "step": 8608 + }, + { + "epoch": 0.5973908819651655, + "grad_norm": 1.9862781763076782, + "learning_rate": 0.00019262323333343645, + "loss": 0.8651, + "step": 8609 + }, + { + "epoch": 0.5974602734022622, + "grad_norm": 2.767059087753296, + "learning_rate": 0.00019262111523253128, + "loss": 1.0443, + "step": 8610 + }, + { + "epoch": 0.5975296648393589, + "grad_norm": 2.2440474033355713, + "learning_rate": 0.000192618996839232, + "loss": 0.8668, + "step": 8611 + }, + { + "epoch": 0.5975990562764555, + "grad_norm": 2.6431729793548584, + "learning_rate": 0.0001926168781535453, + "loss": 1.1543, + "step": 8612 + }, + { + "epoch": 0.5976684477135521, + "grad_norm": 3.056018352508545, + "learning_rate": 0.00019261475917547785, + "loss": 1.0368, + "step": 8613 + }, + { + "epoch": 0.5977378391506488, + "grad_norm": 2.6882619857788086, + "learning_rate": 0.00019261263990503634, + "loss": 1.0235, + "step": 8614 + }, + { + "epoch": 0.5978072305877454, + "grad_norm": 2.7637062072753906, + "learning_rate": 0.00019261052034222752, + "loss": 0.823, + "step": 8615 + }, + { + "epoch": 0.5978766220248422, + "grad_norm": 2.9252777099609375, + "learning_rate": 0.000192608400487058, + "loss": 0.9834, + "step": 8616 + }, + { + "epoch": 0.5979460134619388, + "grad_norm": 3.4067766666412354, + "learning_rate": 0.0001926062803395345, + "loss": 0.9044, + "step": 8617 + }, + { + "epoch": 0.5980154048990355, + "grad_norm": 2.887831926345825, + "learning_rate": 0.0001926041598996637, + "loss": 1.0386, + "step": 8618 + }, + { + "epoch": 0.5980847963361321, + "grad_norm": 2.5258655548095703, + "learning_rate": 0.00019260203916745235, + "loss": 1.027, + "step": 8619 + }, + { + "epoch": 0.5981541877732288, + "grad_norm": 2.708315134048462, + "learning_rate": 0.00019259991814290708, + "loss": 0.9944, + "step": 8620 + }, + { + "epoch": 0.5982235792103254, + "grad_norm": 2.6315228939056396, + "learning_rate": 0.00019259779682603462, + "loss": 1.0749, + "step": 8621 + }, + { + "epoch": 0.5982929706474222, + "grad_norm": 2.18939208984375, + "learning_rate": 0.00019259567521684166, + "loss": 0.9329, + "step": 8622 + }, + { + "epoch": 0.5983623620845188, + "grad_norm": 3.019688606262207, + "learning_rate": 0.0001925935533153349, + "loss": 0.8531, + "step": 8623 + }, + { + "epoch": 0.5984317535216155, + "grad_norm": 2.0321156978607178, + "learning_rate": 0.00019259143112152104, + "loss": 0.9613, + "step": 8624 + }, + { + "epoch": 0.5985011449587121, + "grad_norm": 2.3211443424224854, + "learning_rate": 0.00019258930863540674, + "loss": 0.9304, + "step": 8625 + }, + { + "epoch": 0.5985705363958087, + "grad_norm": 2.384030818939209, + "learning_rate": 0.00019258718585699874, + "loss": 0.9227, + "step": 8626 + }, + { + "epoch": 0.5986399278329054, + "grad_norm": 3.244215965270996, + "learning_rate": 0.00019258506278630377, + "loss": 1.1946, + "step": 8627 + }, + { + "epoch": 0.598709319270002, + "grad_norm": 3.278104782104492, + "learning_rate": 0.0001925829394233285, + "loss": 0.8358, + "step": 8628 + }, + { + "epoch": 0.5987787107070988, + "grad_norm": 2.3331401348114014, + "learning_rate": 0.0001925808157680796, + "loss": 1.0246, + "step": 8629 + }, + { + "epoch": 0.5988481021441954, + "grad_norm": 2.8400278091430664, + "learning_rate": 0.00019257869182056383, + "loss": 0.8293, + "step": 8630 + }, + { + "epoch": 0.5989174935812921, + "grad_norm": 3.3349361419677734, + "learning_rate": 0.00019257656758078786, + "loss": 0.7831, + "step": 8631 + }, + { + "epoch": 0.5989868850183887, + "grad_norm": 2.4614739418029785, + "learning_rate": 0.0001925744430487584, + "loss": 0.9197, + "step": 8632 + }, + { + "epoch": 0.5990562764554854, + "grad_norm": 7.028626918792725, + "learning_rate": 0.00019257231822448215, + "loss": 0.6846, + "step": 8633 + }, + { + "epoch": 0.599125667892582, + "grad_norm": 2.381258249282837, + "learning_rate": 0.00019257019310796588, + "loss": 0.8882, + "step": 8634 + }, + { + "epoch": 0.5991950593296788, + "grad_norm": 3.835378646850586, + "learning_rate": 0.00019256806769921622, + "loss": 1.1943, + "step": 8635 + }, + { + "epoch": 0.5992644507667754, + "grad_norm": 3.537931442260742, + "learning_rate": 0.00019256594199823992, + "loss": 1.0632, + "step": 8636 + }, + { + "epoch": 0.5993338422038721, + "grad_norm": 2.513598680496216, + "learning_rate": 0.00019256381600504368, + "loss": 0.9509, + "step": 8637 + }, + { + "epoch": 0.5994032336409687, + "grad_norm": 2.651071310043335, + "learning_rate": 0.0001925616897196342, + "loss": 0.9039, + "step": 8638 + }, + { + "epoch": 0.5994726250780653, + "grad_norm": 2.0649867057800293, + "learning_rate": 0.00019255956314201823, + "loss": 0.8671, + "step": 8639 + }, + { + "epoch": 0.599542016515162, + "grad_norm": 2.3095664978027344, + "learning_rate": 0.0001925574362722024, + "loss": 0.6094, + "step": 8640 + }, + { + "epoch": 0.5996114079522586, + "grad_norm": 2.8457515239715576, + "learning_rate": 0.00019255530911019357, + "loss": 0.9646, + "step": 8641 + }, + { + "epoch": 0.5996807993893554, + "grad_norm": 3.010164737701416, + "learning_rate": 0.0001925531816559983, + "loss": 1.0021, + "step": 8642 + }, + { + "epoch": 0.599750190826452, + "grad_norm": 2.803126811981201, + "learning_rate": 0.0001925510539096234, + "loss": 1.0047, + "step": 8643 + }, + { + "epoch": 0.5998195822635487, + "grad_norm": 2.404017925262451, + "learning_rate": 0.00019254892587107557, + "loss": 0.7318, + "step": 8644 + }, + { + "epoch": 0.5998889737006453, + "grad_norm": 2.739051103591919, + "learning_rate": 0.00019254679754036148, + "loss": 0.8552, + "step": 8645 + }, + { + "epoch": 0.599958365137742, + "grad_norm": 4.083184719085693, + "learning_rate": 0.00019254466891748792, + "loss": 1.081, + "step": 8646 + }, + { + "epoch": 0.6000277565748386, + "grad_norm": 2.93285870552063, + "learning_rate": 0.00019254254000246158, + "loss": 0.8476, + "step": 8647 + }, + { + "epoch": 0.6000971480119354, + "grad_norm": 2.4315710067749023, + "learning_rate": 0.00019254041079528914, + "loss": 1.0935, + "step": 8648 + }, + { + "epoch": 0.600166539449032, + "grad_norm": 2.4350650310516357, + "learning_rate": 0.00019253828129597742, + "loss": 0.8892, + "step": 8649 + }, + { + "epoch": 0.6002359308861287, + "grad_norm": 1.9727524518966675, + "learning_rate": 0.00019253615150453303, + "loss": 0.9792, + "step": 8650 + }, + { + "epoch": 0.6003053223232253, + "grad_norm": 3.228764295578003, + "learning_rate": 0.00019253402142096278, + "loss": 0.8557, + "step": 8651 + }, + { + "epoch": 0.6003747137603219, + "grad_norm": 2.8151862621307373, + "learning_rate": 0.00019253189104527334, + "loss": 0.9495, + "step": 8652 + }, + { + "epoch": 0.6004441051974186, + "grad_norm": 2.3406128883361816, + "learning_rate": 0.00019252976037747145, + "loss": 0.7764, + "step": 8653 + }, + { + "epoch": 0.6005134966345153, + "grad_norm": 2.1304125785827637, + "learning_rate": 0.0001925276294175639, + "loss": 0.7207, + "step": 8654 + }, + { + "epoch": 0.600582888071612, + "grad_norm": 2.46999192237854, + "learning_rate": 0.00019252549816555729, + "loss": 0.8729, + "step": 8655 + }, + { + "epoch": 0.6006522795087086, + "grad_norm": 3.89235520362854, + "learning_rate": 0.00019252336662145845, + "loss": 0.9505, + "step": 8656 + }, + { + "epoch": 0.6007216709458053, + "grad_norm": 2.81807541847229, + "learning_rate": 0.00019252123478527406, + "loss": 0.8891, + "step": 8657 + }, + { + "epoch": 0.6007910623829019, + "grad_norm": 2.4878077507019043, + "learning_rate": 0.00019251910265701084, + "loss": 0.8806, + "step": 8658 + }, + { + "epoch": 0.6008604538199986, + "grad_norm": 2.0625417232513428, + "learning_rate": 0.00019251697023667559, + "loss": 0.8856, + "step": 8659 + }, + { + "epoch": 0.6009298452570953, + "grad_norm": 2.353959083557129, + "learning_rate": 0.000192514837524275, + "loss": 0.9761, + "step": 8660 + }, + { + "epoch": 0.600999236694192, + "grad_norm": 2.0964415073394775, + "learning_rate": 0.0001925127045198158, + "loss": 0.8416, + "step": 8661 + }, + { + "epoch": 0.6010686281312886, + "grad_norm": 3.4563403129577637, + "learning_rate": 0.00019251057122330473, + "loss": 0.8474, + "step": 8662 + }, + { + "epoch": 0.6011380195683853, + "grad_norm": 2.1190922260284424, + "learning_rate": 0.0001925084376347485, + "loss": 0.7288, + "step": 8663 + }, + { + "epoch": 0.6012074110054819, + "grad_norm": 2.420196533203125, + "learning_rate": 0.00019250630375415386, + "loss": 0.9431, + "step": 8664 + }, + { + "epoch": 0.6012768024425786, + "grad_norm": 3.0374631881713867, + "learning_rate": 0.00019250416958152756, + "loss": 0.9306, + "step": 8665 + }, + { + "epoch": 0.6013461938796753, + "grad_norm": 2.7104804515838623, + "learning_rate": 0.00019250203511687633, + "loss": 0.7588, + "step": 8666 + }, + { + "epoch": 0.6014155853167719, + "grad_norm": 2.2305703163146973, + "learning_rate": 0.0001924999003602069, + "loss": 0.6811, + "step": 8667 + }, + { + "epoch": 0.6014849767538686, + "grad_norm": 3.4133412837982178, + "learning_rate": 0.00019249776531152604, + "loss": 0.9248, + "step": 8668 + }, + { + "epoch": 0.6015543681909652, + "grad_norm": 2.20440673828125, + "learning_rate": 0.00019249562997084043, + "loss": 0.7688, + "step": 8669 + }, + { + "epoch": 0.6016237596280619, + "grad_norm": 3.4306323528289795, + "learning_rate": 0.00019249349433815688, + "loss": 1.1614, + "step": 8670 + }, + { + "epoch": 0.6016931510651585, + "grad_norm": 2.516871929168701, + "learning_rate": 0.00019249135841348206, + "loss": 0.8403, + "step": 8671 + }, + { + "epoch": 0.6017625425022552, + "grad_norm": 2.83561372756958, + "learning_rate": 0.00019248922219682282, + "loss": 1.0007, + "step": 8672 + }, + { + "epoch": 0.6018319339393519, + "grad_norm": 2.4848880767822266, + "learning_rate": 0.0001924870856881858, + "loss": 0.9108, + "step": 8673 + }, + { + "epoch": 0.6019013253764486, + "grad_norm": 3.1105737686157227, + "learning_rate": 0.0001924849488875778, + "loss": 1.173, + "step": 8674 + }, + { + "epoch": 0.6019707168135452, + "grad_norm": 2.245549440383911, + "learning_rate": 0.0001924828117950055, + "loss": 0.8129, + "step": 8675 + }, + { + "epoch": 0.6020401082506419, + "grad_norm": 2.292330265045166, + "learning_rate": 0.0001924806744104757, + "loss": 0.7905, + "step": 8676 + }, + { + "epoch": 0.6021094996877385, + "grad_norm": 2.6397948265075684, + "learning_rate": 0.00019247853673399516, + "loss": 0.7791, + "step": 8677 + }, + { + "epoch": 0.6021788911248352, + "grad_norm": 2.426846981048584, + "learning_rate": 0.00019247639876557063, + "loss": 0.8342, + "step": 8678 + }, + { + "epoch": 0.6022482825619319, + "grad_norm": 2.424926519393921, + "learning_rate": 0.0001924742605052088, + "loss": 0.7188, + "step": 8679 + }, + { + "epoch": 0.6023176739990285, + "grad_norm": 3.1369948387145996, + "learning_rate": 0.00019247212195291647, + "loss": 0.9035, + "step": 8680 + }, + { + "epoch": 0.6023870654361252, + "grad_norm": 1.7543915510177612, + "learning_rate": 0.0001924699831087004, + "loss": 0.8754, + "step": 8681 + }, + { + "epoch": 0.6024564568732218, + "grad_norm": 3.668217897415161, + "learning_rate": 0.0001924678439725673, + "loss": 0.8799, + "step": 8682 + }, + { + "epoch": 0.6025258483103185, + "grad_norm": 2.9693593978881836, + "learning_rate": 0.00019246570454452394, + "loss": 0.8252, + "step": 8683 + }, + { + "epoch": 0.6025952397474151, + "grad_norm": 2.843371868133545, + "learning_rate": 0.00019246356482457711, + "loss": 1.116, + "step": 8684 + }, + { + "epoch": 0.6026646311845119, + "grad_norm": 2.410991907119751, + "learning_rate": 0.0001924614248127335, + "loss": 0.9889, + "step": 8685 + }, + { + "epoch": 0.6027340226216085, + "grad_norm": 2.2312533855438232, + "learning_rate": 0.0001924592845089999, + "loss": 1.1274, + "step": 8686 + }, + { + "epoch": 0.6028034140587052, + "grad_norm": 2.8626863956451416, + "learning_rate": 0.0001924571439133831, + "loss": 0.9331, + "step": 8687 + }, + { + "epoch": 0.6028728054958018, + "grad_norm": 2.4054489135742188, + "learning_rate": 0.0001924550030258898, + "loss": 0.7793, + "step": 8688 + }, + { + "epoch": 0.6029421969328985, + "grad_norm": 2.6123483180999756, + "learning_rate": 0.0001924528618465268, + "loss": 0.9918, + "step": 8689 + }, + { + "epoch": 0.6030115883699951, + "grad_norm": 2.7332730293273926, + "learning_rate": 0.0001924507203753008, + "loss": 1.0842, + "step": 8690 + }, + { + "epoch": 0.6030809798070919, + "grad_norm": 3.0480339527130127, + "learning_rate": 0.00019244857861221864, + "loss": 0.7742, + "step": 8691 + }, + { + "epoch": 0.6031503712441885, + "grad_norm": 2.51713228225708, + "learning_rate": 0.00019244643655728702, + "loss": 0.9229, + "step": 8692 + }, + { + "epoch": 0.6032197626812851, + "grad_norm": 2.53135347366333, + "learning_rate": 0.00019244429421051275, + "loss": 0.7652, + "step": 8693 + }, + { + "epoch": 0.6032891541183818, + "grad_norm": 2.9416866302490234, + "learning_rate": 0.00019244215157190254, + "loss": 0.9253, + "step": 8694 + }, + { + "epoch": 0.6033585455554784, + "grad_norm": 2.4590063095092773, + "learning_rate": 0.00019244000864146324, + "loss": 0.8814, + "step": 8695 + }, + { + "epoch": 0.6034279369925751, + "grad_norm": 2.7757561206817627, + "learning_rate": 0.0001924378654192015, + "loss": 0.881, + "step": 8696 + }, + { + "epoch": 0.6034973284296717, + "grad_norm": 3.016094446182251, + "learning_rate": 0.00019243572190512416, + "loss": 0.7525, + "step": 8697 + }, + { + "epoch": 0.6035667198667685, + "grad_norm": 4.216963768005371, + "learning_rate": 0.00019243357809923796, + "loss": 0.9431, + "step": 8698 + }, + { + "epoch": 0.6036361113038651, + "grad_norm": 2.741237163543701, + "learning_rate": 0.0001924314340015497, + "loss": 1.0379, + "step": 8699 + }, + { + "epoch": 0.6037055027409618, + "grad_norm": 2.837207555770874, + "learning_rate": 0.0001924292896120661, + "loss": 0.989, + "step": 8700 + }, + { + "epoch": 0.6037748941780584, + "grad_norm": 2.2561373710632324, + "learning_rate": 0.000192427144930794, + "loss": 0.7974, + "step": 8701 + }, + { + "epoch": 0.6038442856151551, + "grad_norm": 2.652878522872925, + "learning_rate": 0.00019242499995774008, + "loss": 0.8228, + "step": 8702 + }, + { + "epoch": 0.6039136770522517, + "grad_norm": 5.753514289855957, + "learning_rate": 0.00019242285469291118, + "loss": 0.9588, + "step": 8703 + }, + { + "epoch": 0.6039830684893485, + "grad_norm": 2.7503161430358887, + "learning_rate": 0.00019242070913631407, + "loss": 0.9363, + "step": 8704 + }, + { + "epoch": 0.6040524599264451, + "grad_norm": 2.0348241329193115, + "learning_rate": 0.00019241856328795548, + "loss": 0.7064, + "step": 8705 + }, + { + "epoch": 0.6041218513635417, + "grad_norm": 2.9515762329101562, + "learning_rate": 0.0001924164171478422, + "loss": 0.8218, + "step": 8706 + }, + { + "epoch": 0.6041912428006384, + "grad_norm": 3.179600477218628, + "learning_rate": 0.000192414270715981, + "loss": 0.7348, + "step": 8707 + }, + { + "epoch": 0.604260634237735, + "grad_norm": 2.7007336616516113, + "learning_rate": 0.00019241212399237872, + "loss": 0.8149, + "step": 8708 + }, + { + "epoch": 0.6043300256748317, + "grad_norm": 2.2725307941436768, + "learning_rate": 0.00019240997697704207, + "loss": 0.7195, + "step": 8709 + }, + { + "epoch": 0.6043994171119283, + "grad_norm": 2.3945460319519043, + "learning_rate": 0.00019240782966997784, + "loss": 0.8455, + "step": 8710 + }, + { + "epoch": 0.6044688085490251, + "grad_norm": 3.0040364265441895, + "learning_rate": 0.0001924056820711928, + "loss": 0.7994, + "step": 8711 + }, + { + "epoch": 0.6045381999861217, + "grad_norm": 3.1874985694885254, + "learning_rate": 0.00019240353418069375, + "loss": 1.361, + "step": 8712 + }, + { + "epoch": 0.6046075914232184, + "grad_norm": 2.826084613800049, + "learning_rate": 0.00019240138599848746, + "loss": 0.8828, + "step": 8713 + }, + { + "epoch": 0.604676982860315, + "grad_norm": 4.687851428985596, + "learning_rate": 0.0001923992375245807, + "loss": 1.0547, + "step": 8714 + }, + { + "epoch": 0.6047463742974117, + "grad_norm": 5.082951068878174, + "learning_rate": 0.0001923970887589803, + "loss": 0.9801, + "step": 8715 + }, + { + "epoch": 0.6048157657345083, + "grad_norm": 2.9874253273010254, + "learning_rate": 0.000192394939701693, + "loss": 0.8011, + "step": 8716 + }, + { + "epoch": 0.6048851571716051, + "grad_norm": 3.4663825035095215, + "learning_rate": 0.0001923927903527256, + "loss": 0.7953, + "step": 8717 + }, + { + "epoch": 0.6049545486087017, + "grad_norm": 2.161860704421997, + "learning_rate": 0.0001923906407120849, + "loss": 0.726, + "step": 8718 + }, + { + "epoch": 0.6050239400457984, + "grad_norm": 2.6717143058776855, + "learning_rate": 0.0001923884907797776, + "loss": 1.0, + "step": 8719 + }, + { + "epoch": 0.605093331482895, + "grad_norm": 2.2447643280029297, + "learning_rate": 0.00019238634055581059, + "loss": 0.8133, + "step": 8720 + }, + { + "epoch": 0.6051627229199916, + "grad_norm": 3.791893482208252, + "learning_rate": 0.00019238419004019065, + "loss": 1.2622, + "step": 8721 + }, + { + "epoch": 0.6052321143570883, + "grad_norm": 18.400876998901367, + "learning_rate": 0.00019238203923292447, + "loss": 1.0513, + "step": 8722 + }, + { + "epoch": 0.605301505794185, + "grad_norm": 2.814845561981201, + "learning_rate": 0.000192379888134019, + "loss": 0.7319, + "step": 8723 + }, + { + "epoch": 0.6053708972312817, + "grad_norm": 4.8144402503967285, + "learning_rate": 0.00019237773674348087, + "loss": 0.7914, + "step": 8724 + }, + { + "epoch": 0.6054402886683783, + "grad_norm": 3.5931873321533203, + "learning_rate": 0.00019237558506131697, + "loss": 0.9617, + "step": 8725 + }, + { + "epoch": 0.605509680105475, + "grad_norm": 3.1671299934387207, + "learning_rate": 0.00019237343308753407, + "loss": 0.7596, + "step": 8726 + }, + { + "epoch": 0.6055790715425716, + "grad_norm": 3.6311893463134766, + "learning_rate": 0.00019237128082213894, + "loss": 1.353, + "step": 8727 + }, + { + "epoch": 0.6056484629796683, + "grad_norm": 3.874181032180786, + "learning_rate": 0.00019236912826513843, + "loss": 0.8299, + "step": 8728 + }, + { + "epoch": 0.605717854416765, + "grad_norm": 3.2182822227478027, + "learning_rate": 0.00019236697541653928, + "loss": 0.9153, + "step": 8729 + }, + { + "epoch": 0.6057872458538617, + "grad_norm": 6.019847393035889, + "learning_rate": 0.0001923648222763483, + "loss": 1.1333, + "step": 8730 + }, + { + "epoch": 0.6058566372909583, + "grad_norm": 4.552515506744385, + "learning_rate": 0.00019236266884457227, + "loss": 1.2362, + "step": 8731 + }, + { + "epoch": 0.605926028728055, + "grad_norm": 7.222696304321289, + "learning_rate": 0.00019236051512121804, + "loss": 1.1962, + "step": 8732 + }, + { + "epoch": 0.6059954201651516, + "grad_norm": 2.486361503601074, + "learning_rate": 0.00019235836110629236, + "loss": 0.796, + "step": 8733 + }, + { + "epoch": 0.6060648116022482, + "grad_norm": 3.831401824951172, + "learning_rate": 0.00019235620679980208, + "loss": 0.9055, + "step": 8734 + }, + { + "epoch": 0.606134203039345, + "grad_norm": 3.8401026725769043, + "learning_rate": 0.00019235405220175396, + "loss": 1.0826, + "step": 8735 + }, + { + "epoch": 0.6062035944764416, + "grad_norm": 3.510389804840088, + "learning_rate": 0.00019235189731215478, + "loss": 0.9568, + "step": 8736 + }, + { + "epoch": 0.6062729859135383, + "grad_norm": 3.037325143814087, + "learning_rate": 0.00019234974213101143, + "loss": 1.3341, + "step": 8737 + }, + { + "epoch": 0.6063423773506349, + "grad_norm": 2.793166399002075, + "learning_rate": 0.00019234758665833063, + "loss": 0.9477, + "step": 8738 + }, + { + "epoch": 0.6064117687877316, + "grad_norm": 3.122684955596924, + "learning_rate": 0.00019234543089411923, + "loss": 1.0316, + "step": 8739 + }, + { + "epoch": 0.6064811602248282, + "grad_norm": 3.4056060314178467, + "learning_rate": 0.000192343274838384, + "loss": 1.1813, + "step": 8740 + }, + { + "epoch": 0.606550551661925, + "grad_norm": 2.9263906478881836, + "learning_rate": 0.00019234111849113176, + "loss": 1.0328, + "step": 8741 + }, + { + "epoch": 0.6066199430990216, + "grad_norm": 2.97538685798645, + "learning_rate": 0.0001923389618523693, + "loss": 1.1759, + "step": 8742 + }, + { + "epoch": 0.6066893345361183, + "grad_norm": 3.193545341491699, + "learning_rate": 0.00019233680492210348, + "loss": 0.6898, + "step": 8743 + }, + { + "epoch": 0.6067587259732149, + "grad_norm": 2.9301328659057617, + "learning_rate": 0.00019233464770034112, + "loss": 0.9294, + "step": 8744 + }, + { + "epoch": 0.6068281174103116, + "grad_norm": 2.9655306339263916, + "learning_rate": 0.00019233249018708892, + "loss": 0.9489, + "step": 8745 + }, + { + "epoch": 0.6068975088474082, + "grad_norm": 3.5435547828674316, + "learning_rate": 0.00019233033238235381, + "loss": 0.8925, + "step": 8746 + }, + { + "epoch": 0.6069669002845048, + "grad_norm": 3.235360622406006, + "learning_rate": 0.0001923281742861425, + "loss": 1.3241, + "step": 8747 + }, + { + "epoch": 0.6070362917216016, + "grad_norm": 2.7915291786193848, + "learning_rate": 0.00019232601589846192, + "loss": 0.7523, + "step": 8748 + }, + { + "epoch": 0.6071056831586982, + "grad_norm": 2.6140999794006348, + "learning_rate": 0.00019232385721931877, + "loss": 1.0887, + "step": 8749 + }, + { + "epoch": 0.6071750745957949, + "grad_norm": 3.429749011993408, + "learning_rate": 0.00019232169824871996, + "loss": 0.937, + "step": 8750 + }, + { + "epoch": 0.6072444660328915, + "grad_norm": 2.65861177444458, + "learning_rate": 0.00019231953898667223, + "loss": 0.9317, + "step": 8751 + }, + { + "epoch": 0.6073138574699882, + "grad_norm": 2.3300116062164307, + "learning_rate": 0.00019231737943318242, + "loss": 0.7085, + "step": 8752 + }, + { + "epoch": 0.6073832489070848, + "grad_norm": 5.092051029205322, + "learning_rate": 0.00019231521958825739, + "loss": 1.0206, + "step": 8753 + }, + { + "epoch": 0.6074526403441816, + "grad_norm": 3.2131237983703613, + "learning_rate": 0.00019231305945190388, + "loss": 1.0531, + "step": 8754 + }, + { + "epoch": 0.6075220317812782, + "grad_norm": 2.4793295860290527, + "learning_rate": 0.00019231089902412878, + "loss": 0.7376, + "step": 8755 + }, + { + "epoch": 0.6075914232183749, + "grad_norm": 2.4975929260253906, + "learning_rate": 0.00019230873830493886, + "loss": 0.7817, + "step": 8756 + }, + { + "epoch": 0.6076608146554715, + "grad_norm": 2.8568079471588135, + "learning_rate": 0.00019230657729434096, + "loss": 0.8868, + "step": 8757 + }, + { + "epoch": 0.6077302060925682, + "grad_norm": 3.5569703578948975, + "learning_rate": 0.0001923044159923419, + "loss": 1.0589, + "step": 8758 + }, + { + "epoch": 0.6077995975296648, + "grad_norm": 3.398672103881836, + "learning_rate": 0.00019230225439894857, + "loss": 1.0296, + "step": 8759 + }, + { + "epoch": 0.6078689889667614, + "grad_norm": 3.1835761070251465, + "learning_rate": 0.00019230009251416767, + "loss": 1.0626, + "step": 8760 + }, + { + "epoch": 0.6079383804038582, + "grad_norm": 3.0530545711517334, + "learning_rate": 0.0001922979303380061, + "loss": 0.9611, + "step": 8761 + }, + { + "epoch": 0.6080077718409548, + "grad_norm": 3.414353370666504, + "learning_rate": 0.0001922957678704707, + "loss": 0.8884, + "step": 8762 + }, + { + "epoch": 0.6080771632780515, + "grad_norm": 4.008413314819336, + "learning_rate": 0.00019229360511156823, + "loss": 1.1508, + "step": 8763 + }, + { + "epoch": 0.6081465547151481, + "grad_norm": 3.170171022415161, + "learning_rate": 0.00019229144206130556, + "loss": 1.326, + "step": 8764 + }, + { + "epoch": 0.6082159461522448, + "grad_norm": 2.325917959213257, + "learning_rate": 0.00019228927871968953, + "loss": 0.614, + "step": 8765 + }, + { + "epoch": 0.6082853375893414, + "grad_norm": 2.7330408096313477, + "learning_rate": 0.00019228711508672696, + "loss": 0.9949, + "step": 8766 + }, + { + "epoch": 0.6083547290264382, + "grad_norm": 3.4195897579193115, + "learning_rate": 0.00019228495116242468, + "loss": 0.9606, + "step": 8767 + }, + { + "epoch": 0.6084241204635348, + "grad_norm": 3.3307712078094482, + "learning_rate": 0.0001922827869467895, + "loss": 1.0975, + "step": 8768 + }, + { + "epoch": 0.6084935119006315, + "grad_norm": 3.059220314025879, + "learning_rate": 0.00019228062243982827, + "loss": 0.811, + "step": 8769 + }, + { + "epoch": 0.6085629033377281, + "grad_norm": 2.4586732387542725, + "learning_rate": 0.0001922784576415478, + "loss": 0.9251, + "step": 8770 + }, + { + "epoch": 0.6086322947748248, + "grad_norm": 3.745824098587036, + "learning_rate": 0.000192276292551955, + "loss": 0.9506, + "step": 8771 + }, + { + "epoch": 0.6087016862119214, + "grad_norm": 2.8352558612823486, + "learning_rate": 0.0001922741271710566, + "loss": 0.8303, + "step": 8772 + }, + { + "epoch": 0.6087710776490182, + "grad_norm": 3.867917537689209, + "learning_rate": 0.0001922719614988595, + "loss": 1.2025, + "step": 8773 + }, + { + "epoch": 0.6088404690861148, + "grad_norm": 2.562304973602295, + "learning_rate": 0.00019226979553537056, + "loss": 0.801, + "step": 8774 + }, + { + "epoch": 0.6089098605232114, + "grad_norm": 2.7517638206481934, + "learning_rate": 0.00019226762928059655, + "loss": 1.0461, + "step": 8775 + }, + { + "epoch": 0.6089792519603081, + "grad_norm": 2.4721508026123047, + "learning_rate": 0.00019226546273454434, + "loss": 0.6595, + "step": 8776 + }, + { + "epoch": 0.6090486433974047, + "grad_norm": 3.1475014686584473, + "learning_rate": 0.00019226329589722076, + "loss": 0.9, + "step": 8777 + }, + { + "epoch": 0.6091180348345014, + "grad_norm": 3.04937744140625, + "learning_rate": 0.00019226112876863266, + "loss": 0.8863, + "step": 8778 + }, + { + "epoch": 0.609187426271598, + "grad_norm": 2.6657803058624268, + "learning_rate": 0.0001922589613487869, + "loss": 0.9433, + "step": 8779 + }, + { + "epoch": 0.6092568177086948, + "grad_norm": 2.2000908851623535, + "learning_rate": 0.00019225679363769026, + "loss": 0.9191, + "step": 8780 + }, + { + "epoch": 0.6093262091457914, + "grad_norm": 2.550342321395874, + "learning_rate": 0.00019225462563534967, + "loss": 0.8469, + "step": 8781 + }, + { + "epoch": 0.6093956005828881, + "grad_norm": 16.863210678100586, + "learning_rate": 0.00019225245734177192, + "loss": 0.9109, + "step": 8782 + }, + { + "epoch": 0.6094649920199847, + "grad_norm": 3.180744171142578, + "learning_rate": 0.00019225028875696388, + "loss": 0.8813, + "step": 8783 + }, + { + "epoch": 0.6095343834570814, + "grad_norm": 6.304474353790283, + "learning_rate": 0.00019224811988093236, + "loss": 0.8137, + "step": 8784 + }, + { + "epoch": 0.609603774894178, + "grad_norm": 4.247776508331299, + "learning_rate": 0.00019224595071368422, + "loss": 1.1546, + "step": 8785 + }, + { + "epoch": 0.6096731663312748, + "grad_norm": 3.005488872528076, + "learning_rate": 0.00019224378125522632, + "loss": 0.9812, + "step": 8786 + }, + { + "epoch": 0.6097425577683714, + "grad_norm": 5.6712236404418945, + "learning_rate": 0.00019224161150556554, + "loss": 1.2317, + "step": 8787 + }, + { + "epoch": 0.609811949205468, + "grad_norm": 4.518888473510742, + "learning_rate": 0.00019223944146470868, + "loss": 1.3846, + "step": 8788 + }, + { + "epoch": 0.6098813406425647, + "grad_norm": 4.868253707885742, + "learning_rate": 0.0001922372711326626, + "loss": 0.8912, + "step": 8789 + }, + { + "epoch": 0.6099507320796613, + "grad_norm": 3.6278722286224365, + "learning_rate": 0.00019223510050943413, + "loss": 0.8183, + "step": 8790 + }, + { + "epoch": 0.610020123516758, + "grad_norm": 2.9051480293273926, + "learning_rate": 0.00019223292959503016, + "loss": 0.9599, + "step": 8791 + }, + { + "epoch": 0.6100895149538547, + "grad_norm": 2.9295451641082764, + "learning_rate": 0.00019223075838945754, + "loss": 0.655, + "step": 8792 + }, + { + "epoch": 0.6101589063909514, + "grad_norm": 2.5944035053253174, + "learning_rate": 0.00019222858689272313, + "loss": 0.7956, + "step": 8793 + }, + { + "epoch": 0.610228297828048, + "grad_norm": 3.5938374996185303, + "learning_rate": 0.00019222641510483374, + "loss": 1.1096, + "step": 8794 + }, + { + "epoch": 0.6102976892651447, + "grad_norm": 2.5901474952697754, + "learning_rate": 0.0001922242430257963, + "loss": 0.8255, + "step": 8795 + }, + { + "epoch": 0.6103670807022413, + "grad_norm": 2.417933702468872, + "learning_rate": 0.0001922220706556176, + "loss": 0.7896, + "step": 8796 + }, + { + "epoch": 0.610436472139338, + "grad_norm": 2.889192819595337, + "learning_rate": 0.00019221989799430452, + "loss": 1.1008, + "step": 8797 + }, + { + "epoch": 0.6105058635764347, + "grad_norm": 2.8108389377593994, + "learning_rate": 0.00019221772504186393, + "loss": 1.0813, + "step": 8798 + }, + { + "epoch": 0.6105752550135314, + "grad_norm": 2.463736057281494, + "learning_rate": 0.0001922155517983027, + "loss": 0.5937, + "step": 8799 + }, + { + "epoch": 0.610644646450628, + "grad_norm": 3.233696699142456, + "learning_rate": 0.0001922133782636276, + "loss": 0.8588, + "step": 8800 + }, + { + "epoch": 0.6107140378877246, + "grad_norm": 2.8643617630004883, + "learning_rate": 0.00019221120443784563, + "loss": 1.0442, + "step": 8801 + }, + { + "epoch": 0.6107834293248213, + "grad_norm": 5.060187339782715, + "learning_rate": 0.00019220903032096356, + "loss": 0.8964, + "step": 8802 + }, + { + "epoch": 0.6108528207619179, + "grad_norm": 2.6633691787719727, + "learning_rate": 0.0001922068559129883, + "loss": 0.9041, + "step": 8803 + }, + { + "epoch": 0.6109222121990147, + "grad_norm": 3.336963415145874, + "learning_rate": 0.0001922046812139267, + "loss": 0.9257, + "step": 8804 + }, + { + "epoch": 0.6109916036361113, + "grad_norm": 4.408467769622803, + "learning_rate": 0.0001922025062237856, + "loss": 1.2901, + "step": 8805 + }, + { + "epoch": 0.611060995073208, + "grad_norm": 3.0886669158935547, + "learning_rate": 0.00019220033094257186, + "loss": 0.9817, + "step": 8806 + }, + { + "epoch": 0.6111303865103046, + "grad_norm": 3.810621738433838, + "learning_rate": 0.00019219815537029243, + "loss": 0.9197, + "step": 8807 + }, + { + "epoch": 0.6111997779474013, + "grad_norm": 3.373727560043335, + "learning_rate": 0.00019219597950695407, + "loss": 1.1904, + "step": 8808 + }, + { + "epoch": 0.6112691693844979, + "grad_norm": 2.635469436645508, + "learning_rate": 0.0001921938033525637, + "loss": 0.8042, + "step": 8809 + }, + { + "epoch": 0.6113385608215947, + "grad_norm": 2.9894042015075684, + "learning_rate": 0.0001921916269071282, + "loss": 0.8342, + "step": 8810 + }, + { + "epoch": 0.6114079522586913, + "grad_norm": 2.9790220260620117, + "learning_rate": 0.0001921894501706545, + "loss": 1.0006, + "step": 8811 + }, + { + "epoch": 0.611477343695788, + "grad_norm": 4.290069103240967, + "learning_rate": 0.00019218727314314932, + "loss": 0.9599, + "step": 8812 + }, + { + "epoch": 0.6115467351328846, + "grad_norm": 2.5180091857910156, + "learning_rate": 0.00019218509582461963, + "loss": 0.9204, + "step": 8813 + }, + { + "epoch": 0.6116161265699812, + "grad_norm": 2.453892946243286, + "learning_rate": 0.00019218291821507228, + "loss": 0.7664, + "step": 8814 + }, + { + "epoch": 0.6116855180070779, + "grad_norm": 3.189378499984741, + "learning_rate": 0.00019218074031451417, + "loss": 1.2487, + "step": 8815 + }, + { + "epoch": 0.6117549094441745, + "grad_norm": 2.7704858779907227, + "learning_rate": 0.00019217856212295215, + "loss": 1.0026, + "step": 8816 + }, + { + "epoch": 0.6118243008812713, + "grad_norm": 2.51055908203125, + "learning_rate": 0.0001921763836403931, + "loss": 0.7905, + "step": 8817 + }, + { + "epoch": 0.6118936923183679, + "grad_norm": 2.6881587505340576, + "learning_rate": 0.00019217420486684393, + "loss": 0.8232, + "step": 8818 + }, + { + "epoch": 0.6119630837554646, + "grad_norm": 4.21259069442749, + "learning_rate": 0.00019217202580231146, + "loss": 1.038, + "step": 8819 + }, + { + "epoch": 0.6120324751925612, + "grad_norm": 2.6658833026885986, + "learning_rate": 0.0001921698464468026, + "loss": 0.899, + "step": 8820 + }, + { + "epoch": 0.6121018666296579, + "grad_norm": 3.0652599334716797, + "learning_rate": 0.00019216766680032424, + "loss": 1.1594, + "step": 8821 + }, + { + "epoch": 0.6121712580667545, + "grad_norm": 2.170106887817383, + "learning_rate": 0.00019216548686288323, + "loss": 0.875, + "step": 8822 + }, + { + "epoch": 0.6122406495038513, + "grad_norm": 2.80009388923645, + "learning_rate": 0.0001921633066344865, + "loss": 0.7817, + "step": 8823 + }, + { + "epoch": 0.6123100409409479, + "grad_norm": 2.562289237976074, + "learning_rate": 0.0001921611261151409, + "loss": 0.7972, + "step": 8824 + }, + { + "epoch": 0.6123794323780446, + "grad_norm": 2.5407917499542236, + "learning_rate": 0.0001921589453048533, + "loss": 0.8767, + "step": 8825 + }, + { + "epoch": 0.6124488238151412, + "grad_norm": 2.637592077255249, + "learning_rate": 0.0001921567642036306, + "loss": 0.8917, + "step": 8826 + }, + { + "epoch": 0.6125182152522378, + "grad_norm": 2.1897146701812744, + "learning_rate": 0.00019215458281147968, + "loss": 1.0221, + "step": 8827 + }, + { + "epoch": 0.6125876066893345, + "grad_norm": 2.475902795791626, + "learning_rate": 0.00019215240112840747, + "loss": 0.9986, + "step": 8828 + }, + { + "epoch": 0.6126569981264312, + "grad_norm": 4.959930896759033, + "learning_rate": 0.0001921502191544208, + "loss": 1.1456, + "step": 8829 + }, + { + "epoch": 0.6127263895635279, + "grad_norm": 2.611025333404541, + "learning_rate": 0.0001921480368895266, + "loss": 0.9123, + "step": 8830 + }, + { + "epoch": 0.6127957810006245, + "grad_norm": 2.580944299697876, + "learning_rate": 0.0001921458543337317, + "loss": 0.8424, + "step": 8831 + }, + { + "epoch": 0.6128651724377212, + "grad_norm": 3.3231706619262695, + "learning_rate": 0.00019214367148704303, + "loss": 1.0588, + "step": 8832 + }, + { + "epoch": 0.6129345638748178, + "grad_norm": 2.0945723056793213, + "learning_rate": 0.00019214148834946753, + "loss": 0.799, + "step": 8833 + }, + { + "epoch": 0.6130039553119145, + "grad_norm": 2.2959585189819336, + "learning_rate": 0.000192139304921012, + "loss": 1.0104, + "step": 8834 + }, + { + "epoch": 0.6130733467490111, + "grad_norm": 2.693765163421631, + "learning_rate": 0.00019213712120168337, + "loss": 1.0875, + "step": 8835 + }, + { + "epoch": 0.6131427381861079, + "grad_norm": 2.8890187740325928, + "learning_rate": 0.00019213493719148856, + "loss": 0.7624, + "step": 8836 + }, + { + "epoch": 0.6132121296232045, + "grad_norm": 2.4356493949890137, + "learning_rate": 0.00019213275289043445, + "loss": 0.8637, + "step": 8837 + }, + { + "epoch": 0.6132815210603012, + "grad_norm": 2.2767210006713867, + "learning_rate": 0.00019213056829852788, + "loss": 0.8593, + "step": 8838 + }, + { + "epoch": 0.6133509124973978, + "grad_norm": 2.7071642875671387, + "learning_rate": 0.00019212838341577584, + "loss": 0.9836, + "step": 8839 + }, + { + "epoch": 0.6134203039344945, + "grad_norm": 2.188400983810425, + "learning_rate": 0.00019212619824218518, + "loss": 0.8174, + "step": 8840 + }, + { + "epoch": 0.6134896953715911, + "grad_norm": 3.5151495933532715, + "learning_rate": 0.0001921240127777628, + "loss": 1.025, + "step": 8841 + }, + { + "epoch": 0.6135590868086878, + "grad_norm": 2.9069535732269287, + "learning_rate": 0.0001921218270225156, + "loss": 0.8188, + "step": 8842 + }, + { + "epoch": 0.6136284782457845, + "grad_norm": 3.0800318717956543, + "learning_rate": 0.00019211964097645047, + "loss": 0.7866, + "step": 8843 + }, + { + "epoch": 0.6136978696828811, + "grad_norm": 2.489750862121582, + "learning_rate": 0.0001921174546395743, + "loss": 0.9956, + "step": 8844 + }, + { + "epoch": 0.6137672611199778, + "grad_norm": 2.6515696048736572, + "learning_rate": 0.00019211526801189406, + "loss": 0.8788, + "step": 8845 + }, + { + "epoch": 0.6138366525570744, + "grad_norm": 2.5247082710266113, + "learning_rate": 0.00019211308109341656, + "loss": 0.893, + "step": 8846 + }, + { + "epoch": 0.6139060439941711, + "grad_norm": 1.9537502527236938, + "learning_rate": 0.0001921108938841488, + "loss": 0.5003, + "step": 8847 + }, + { + "epoch": 0.6139754354312678, + "grad_norm": 2.4168906211853027, + "learning_rate": 0.00019210870638409762, + "loss": 0.8801, + "step": 8848 + }, + { + "epoch": 0.6140448268683645, + "grad_norm": 3.91800594329834, + "learning_rate": 0.0001921065185932699, + "loss": 0.8125, + "step": 8849 + }, + { + "epoch": 0.6141142183054611, + "grad_norm": 2.7315399646759033, + "learning_rate": 0.00019210433051167263, + "loss": 0.9048, + "step": 8850 + }, + { + "epoch": 0.6141836097425578, + "grad_norm": 2.5858724117279053, + "learning_rate": 0.00019210214213931265, + "loss": 1.1072, + "step": 8851 + }, + { + "epoch": 0.6142530011796544, + "grad_norm": 2.9477009773254395, + "learning_rate": 0.00019209995347619691, + "loss": 1.1135, + "step": 8852 + }, + { + "epoch": 0.6143223926167511, + "grad_norm": 2.0429258346557617, + "learning_rate": 0.0001920977645223323, + "loss": 0.8615, + "step": 8853 + }, + { + "epoch": 0.6143917840538478, + "grad_norm": 4.193000316619873, + "learning_rate": 0.00019209557527772572, + "loss": 1.516, + "step": 8854 + }, + { + "epoch": 0.6144611754909444, + "grad_norm": 3.13566255569458, + "learning_rate": 0.0001920933857423841, + "loss": 0.9663, + "step": 8855 + }, + { + "epoch": 0.6145305669280411, + "grad_norm": 3.065969944000244, + "learning_rate": 0.00019209119591631434, + "loss": 0.8083, + "step": 8856 + }, + { + "epoch": 0.6145999583651377, + "grad_norm": 2.515949010848999, + "learning_rate": 0.00019208900579952337, + "loss": 0.907, + "step": 8857 + }, + { + "epoch": 0.6146693498022344, + "grad_norm": 2.7611610889434814, + "learning_rate": 0.00019208681539201812, + "loss": 0.8968, + "step": 8858 + }, + { + "epoch": 0.614738741239331, + "grad_norm": 2.9880924224853516, + "learning_rate": 0.00019208462469380542, + "loss": 1.1038, + "step": 8859 + }, + { + "epoch": 0.6148081326764278, + "grad_norm": 2.602465867996216, + "learning_rate": 0.00019208243370489228, + "loss": 0.9331, + "step": 8860 + }, + { + "epoch": 0.6148775241135244, + "grad_norm": 2.830137252807617, + "learning_rate": 0.00019208024242528557, + "loss": 1.0371, + "step": 8861 + }, + { + "epoch": 0.6149469155506211, + "grad_norm": 2.4329721927642822, + "learning_rate": 0.00019207805085499224, + "loss": 0.8863, + "step": 8862 + }, + { + "epoch": 0.6150163069877177, + "grad_norm": 2.889404296875, + "learning_rate": 0.00019207585899401916, + "loss": 0.9362, + "step": 8863 + }, + { + "epoch": 0.6150856984248144, + "grad_norm": 2.748750686645508, + "learning_rate": 0.00019207366684237327, + "loss": 0.9299, + "step": 8864 + }, + { + "epoch": 0.615155089861911, + "grad_norm": 2.7011518478393555, + "learning_rate": 0.00019207147440006154, + "loss": 1.0102, + "step": 8865 + }, + { + "epoch": 0.6152244812990078, + "grad_norm": 1.9393963813781738, + "learning_rate": 0.00019206928166709078, + "loss": 0.7833, + "step": 8866 + }, + { + "epoch": 0.6152938727361044, + "grad_norm": 2.7550570964813232, + "learning_rate": 0.00019206708864346803, + "loss": 1.0587, + "step": 8867 + }, + { + "epoch": 0.615363264173201, + "grad_norm": 2.165316581726074, + "learning_rate": 0.00019206489532920014, + "loss": 0.7538, + "step": 8868 + }, + { + "epoch": 0.6154326556102977, + "grad_norm": 4.479246139526367, + "learning_rate": 0.00019206270172429407, + "loss": 1.1339, + "step": 8869 + }, + { + "epoch": 0.6155020470473943, + "grad_norm": 6.8708906173706055, + "learning_rate": 0.00019206050782875672, + "loss": 1.2532, + "step": 8870 + }, + { + "epoch": 0.615571438484491, + "grad_norm": 2.8644232749938965, + "learning_rate": 0.00019205831364259505, + "loss": 0.9881, + "step": 8871 + }, + { + "epoch": 0.6156408299215876, + "grad_norm": 4.490549564361572, + "learning_rate": 0.00019205611916581593, + "loss": 1.2582, + "step": 8872 + }, + { + "epoch": 0.6157102213586844, + "grad_norm": 2.752859592437744, + "learning_rate": 0.00019205392439842638, + "loss": 0.7076, + "step": 8873 + }, + { + "epoch": 0.615779612795781, + "grad_norm": 2.9023990631103516, + "learning_rate": 0.00019205172934043323, + "loss": 0.8602, + "step": 8874 + }, + { + "epoch": 0.6158490042328777, + "grad_norm": 2.358184337615967, + "learning_rate": 0.00019204953399184344, + "loss": 0.7763, + "step": 8875 + }, + { + "epoch": 0.6159183956699743, + "grad_norm": 2.7128119468688965, + "learning_rate": 0.000192047338352664, + "loss": 0.996, + "step": 8876 + }, + { + "epoch": 0.615987787107071, + "grad_norm": 2.720366954803467, + "learning_rate": 0.00019204514242290171, + "loss": 0.6994, + "step": 8877 + }, + { + "epoch": 0.6160571785441676, + "grad_norm": 3.2431833744049072, + "learning_rate": 0.00019204294620256364, + "loss": 0.9146, + "step": 8878 + }, + { + "epoch": 0.6161265699812644, + "grad_norm": 4.572152137756348, + "learning_rate": 0.00019204074969165667, + "loss": 0.9941, + "step": 8879 + }, + { + "epoch": 0.616195961418361, + "grad_norm": 2.2905399799346924, + "learning_rate": 0.00019203855289018771, + "loss": 0.7967, + "step": 8880 + }, + { + "epoch": 0.6162653528554576, + "grad_norm": 2.792806386947632, + "learning_rate": 0.00019203635579816375, + "loss": 1.0187, + "step": 8881 + }, + { + "epoch": 0.6163347442925543, + "grad_norm": 3.7552847862243652, + "learning_rate": 0.00019203415841559168, + "loss": 1.0981, + "step": 8882 + }, + { + "epoch": 0.6164041357296509, + "grad_norm": 2.2374250888824463, + "learning_rate": 0.00019203196074247842, + "loss": 0.5938, + "step": 8883 + }, + { + "epoch": 0.6164735271667476, + "grad_norm": 2.490267038345337, + "learning_rate": 0.00019202976277883096, + "loss": 0.9488, + "step": 8884 + }, + { + "epoch": 0.6165429186038442, + "grad_norm": 3.140934467315674, + "learning_rate": 0.0001920275645246562, + "loss": 0.9637, + "step": 8885 + }, + { + "epoch": 0.616612310040941, + "grad_norm": 3.837002992630005, + "learning_rate": 0.00019202536597996112, + "loss": 0.9078, + "step": 8886 + }, + { + "epoch": 0.6166817014780376, + "grad_norm": 3.2107207775115967, + "learning_rate": 0.00019202316714475264, + "loss": 0.8251, + "step": 8887 + }, + { + "epoch": 0.6167510929151343, + "grad_norm": 3.2907307147979736, + "learning_rate": 0.00019202096801903767, + "loss": 0.9217, + "step": 8888 + }, + { + "epoch": 0.6168204843522309, + "grad_norm": 2.329284429550171, + "learning_rate": 0.0001920187686028232, + "loss": 0.7523, + "step": 8889 + }, + { + "epoch": 0.6168898757893276, + "grad_norm": 2.5059282779693604, + "learning_rate": 0.00019201656889611615, + "loss": 0.8052, + "step": 8890 + }, + { + "epoch": 0.6169592672264242, + "grad_norm": 2.750916004180908, + "learning_rate": 0.0001920143688989235, + "loss": 0.8997, + "step": 8891 + }, + { + "epoch": 0.617028658663521, + "grad_norm": 2.7647171020507812, + "learning_rate": 0.0001920121686112521, + "loss": 1.0539, + "step": 8892 + }, + { + "epoch": 0.6170980501006176, + "grad_norm": 3.138134002685547, + "learning_rate": 0.000192009968033109, + "loss": 0.9283, + "step": 8893 + }, + { + "epoch": 0.6171674415377143, + "grad_norm": 3.1609513759613037, + "learning_rate": 0.0001920077671645011, + "loss": 0.9824, + "step": 8894 + }, + { + "epoch": 0.6172368329748109, + "grad_norm": 2.7222156524658203, + "learning_rate": 0.00019200556600543537, + "loss": 0.844, + "step": 8895 + }, + { + "epoch": 0.6173062244119075, + "grad_norm": 2.3333945274353027, + "learning_rate": 0.0001920033645559187, + "loss": 0.7765, + "step": 8896 + }, + { + "epoch": 0.6173756158490042, + "grad_norm": 2.096911668777466, + "learning_rate": 0.0001920011628159581, + "loss": 0.7041, + "step": 8897 + }, + { + "epoch": 0.6174450072861009, + "grad_norm": 2.5650486946105957, + "learning_rate": 0.0001919989607855605, + "loss": 0.9876, + "step": 8898 + }, + { + "epoch": 0.6175143987231976, + "grad_norm": 2.336853265762329, + "learning_rate": 0.00019199675846473286, + "loss": 0.8665, + "step": 8899 + }, + { + "epoch": 0.6175837901602942, + "grad_norm": 2.5582454204559326, + "learning_rate": 0.00019199455585348213, + "loss": 1.0383, + "step": 8900 + }, + { + "epoch": 0.6176531815973909, + "grad_norm": 2.0262563228607178, + "learning_rate": 0.00019199235295181528, + "loss": 0.7276, + "step": 8901 + }, + { + "epoch": 0.6177225730344875, + "grad_norm": 2.157749652862549, + "learning_rate": 0.0001919901497597392, + "loss": 0.8795, + "step": 8902 + }, + { + "epoch": 0.6177919644715842, + "grad_norm": 2.261934995651245, + "learning_rate": 0.00019198794627726094, + "loss": 0.7853, + "step": 8903 + }, + { + "epoch": 0.6178613559086809, + "grad_norm": 2.4592323303222656, + "learning_rate": 0.0001919857425043874, + "loss": 0.788, + "step": 8904 + }, + { + "epoch": 0.6179307473457776, + "grad_norm": 2.742793083190918, + "learning_rate": 0.00019198353844112552, + "loss": 0.9749, + "step": 8905 + }, + { + "epoch": 0.6180001387828742, + "grad_norm": 2.4119274616241455, + "learning_rate": 0.00019198133408748226, + "loss": 0.8343, + "step": 8906 + }, + { + "epoch": 0.6180695302199709, + "grad_norm": 4.605307102203369, + "learning_rate": 0.00019197912944346463, + "loss": 0.9436, + "step": 8907 + }, + { + "epoch": 0.6181389216570675, + "grad_norm": 3.1773998737335205, + "learning_rate": 0.00019197692450907957, + "loss": 0.8147, + "step": 8908 + }, + { + "epoch": 0.6182083130941641, + "grad_norm": 3.4574217796325684, + "learning_rate": 0.00019197471928433403, + "loss": 0.8784, + "step": 8909 + }, + { + "epoch": 0.6182777045312609, + "grad_norm": 2.5128066539764404, + "learning_rate": 0.00019197251376923492, + "loss": 0.6877, + "step": 8910 + }, + { + "epoch": 0.6183470959683575, + "grad_norm": 2.418034076690674, + "learning_rate": 0.0001919703079637893, + "loss": 0.5599, + "step": 8911 + }, + { + "epoch": 0.6184164874054542, + "grad_norm": 2.416872501373291, + "learning_rate": 0.0001919681018680041, + "loss": 0.7253, + "step": 8912 + }, + { + "epoch": 0.6184858788425508, + "grad_norm": 2.5357978343963623, + "learning_rate": 0.00019196589548188623, + "loss": 0.8154, + "step": 8913 + }, + { + "epoch": 0.6185552702796475, + "grad_norm": 4.062382698059082, + "learning_rate": 0.00019196368880544274, + "loss": 0.8728, + "step": 8914 + }, + { + "epoch": 0.6186246617167441, + "grad_norm": 2.728466272354126, + "learning_rate": 0.00019196148183868051, + "loss": 0.9205, + "step": 8915 + }, + { + "epoch": 0.6186940531538409, + "grad_norm": 6.529786586761475, + "learning_rate": 0.00019195927458160662, + "loss": 0.9214, + "step": 8916 + }, + { + "epoch": 0.6187634445909375, + "grad_norm": 2.778557777404785, + "learning_rate": 0.00019195706703422792, + "loss": 0.7816, + "step": 8917 + }, + { + "epoch": 0.6188328360280342, + "grad_norm": 2.4806480407714844, + "learning_rate": 0.00019195485919655143, + "loss": 0.9183, + "step": 8918 + }, + { + "epoch": 0.6189022274651308, + "grad_norm": 1.8495181798934937, + "learning_rate": 0.00019195265106858414, + "loss": 0.7062, + "step": 8919 + }, + { + "epoch": 0.6189716189022275, + "grad_norm": 2.2023022174835205, + "learning_rate": 0.00019195044265033298, + "loss": 0.7951, + "step": 8920 + }, + { + "epoch": 0.6190410103393241, + "grad_norm": 2.7586607933044434, + "learning_rate": 0.00019194823394180492, + "loss": 1.0528, + "step": 8921 + }, + { + "epoch": 0.6191104017764207, + "grad_norm": 2.95408034324646, + "learning_rate": 0.000191946024943007, + "loss": 0.7905, + "step": 8922 + }, + { + "epoch": 0.6191797932135175, + "grad_norm": 2.585618257522583, + "learning_rate": 0.0001919438156539461, + "loss": 0.903, + "step": 8923 + }, + { + "epoch": 0.6192491846506141, + "grad_norm": 2.1044466495513916, + "learning_rate": 0.0001919416060746293, + "loss": 1.0251, + "step": 8924 + }, + { + "epoch": 0.6193185760877108, + "grad_norm": 2.7516655921936035, + "learning_rate": 0.00019193939620506346, + "loss": 0.9247, + "step": 8925 + }, + { + "epoch": 0.6193879675248074, + "grad_norm": 3.6426990032196045, + "learning_rate": 0.00019193718604525563, + "loss": 1.0416, + "step": 8926 + }, + { + "epoch": 0.6194573589619041, + "grad_norm": 2.367664098739624, + "learning_rate": 0.0001919349755952128, + "loss": 0.7883, + "step": 8927 + }, + { + "epoch": 0.6195267503990007, + "grad_norm": 2.4348416328430176, + "learning_rate": 0.00019193276485494188, + "loss": 0.7837, + "step": 8928 + }, + { + "epoch": 0.6195961418360975, + "grad_norm": 2.2608659267425537, + "learning_rate": 0.0001919305538244499, + "loss": 0.8111, + "step": 8929 + }, + { + "epoch": 0.6196655332731941, + "grad_norm": 2.5482935905456543, + "learning_rate": 0.00019192834250374383, + "loss": 0.7471, + "step": 8930 + }, + { + "epoch": 0.6197349247102908, + "grad_norm": 1.9541435241699219, + "learning_rate": 0.00019192613089283068, + "loss": 0.5825, + "step": 8931 + }, + { + "epoch": 0.6198043161473874, + "grad_norm": 2.601238489151001, + "learning_rate": 0.00019192391899171735, + "loss": 0.8799, + "step": 8932 + }, + { + "epoch": 0.6198737075844841, + "grad_norm": 2.4885358810424805, + "learning_rate": 0.00019192170680041088, + "loss": 1.0411, + "step": 8933 + }, + { + "epoch": 0.6199430990215807, + "grad_norm": 2.3251099586486816, + "learning_rate": 0.00019191949431891824, + "loss": 0.6965, + "step": 8934 + }, + { + "epoch": 0.6200124904586773, + "grad_norm": 2.4201982021331787, + "learning_rate": 0.00019191728154724645, + "loss": 0.8939, + "step": 8935 + }, + { + "epoch": 0.6200818818957741, + "grad_norm": 3.320150852203369, + "learning_rate": 0.00019191506848540246, + "loss": 0.818, + "step": 8936 + }, + { + "epoch": 0.6201512733328707, + "grad_norm": 2.6427948474884033, + "learning_rate": 0.00019191285513339324, + "loss": 0.7529, + "step": 8937 + }, + { + "epoch": 0.6202206647699674, + "grad_norm": 2.8252038955688477, + "learning_rate": 0.00019191064149122586, + "loss": 0.9051, + "step": 8938 + }, + { + "epoch": 0.620290056207064, + "grad_norm": 2.7465126514434814, + "learning_rate": 0.0001919084275589072, + "loss": 0.8677, + "step": 8939 + }, + { + "epoch": 0.6203594476441607, + "grad_norm": 2.8586392402648926, + "learning_rate": 0.00019190621333644428, + "loss": 0.7563, + "step": 8940 + }, + { + "epoch": 0.6204288390812573, + "grad_norm": 3.123000383377075, + "learning_rate": 0.00019190399882384416, + "loss": 0.8868, + "step": 8941 + }, + { + "epoch": 0.6204982305183541, + "grad_norm": 5.959898948669434, + "learning_rate": 0.00019190178402111373, + "loss": 0.9125, + "step": 8942 + }, + { + "epoch": 0.6205676219554507, + "grad_norm": 2.537717580795288, + "learning_rate": 0.00019189956892826007, + "loss": 0.7218, + "step": 8943 + }, + { + "epoch": 0.6206370133925474, + "grad_norm": 2.7013044357299805, + "learning_rate": 0.0001918973535452901, + "loss": 0.7754, + "step": 8944 + }, + { + "epoch": 0.620706404829644, + "grad_norm": 2.9470434188842773, + "learning_rate": 0.00019189513787221089, + "loss": 1.1182, + "step": 8945 + }, + { + "epoch": 0.6207757962667407, + "grad_norm": 2.592466115951538, + "learning_rate": 0.00019189292190902935, + "loss": 0.9464, + "step": 8946 + }, + { + "epoch": 0.6208451877038373, + "grad_norm": 2.5800321102142334, + "learning_rate": 0.00019189070565575253, + "loss": 0.8506, + "step": 8947 + }, + { + "epoch": 0.6209145791409341, + "grad_norm": 2.1694016456604004, + "learning_rate": 0.00019188848911238742, + "loss": 1.1169, + "step": 8948 + }, + { + "epoch": 0.6209839705780307, + "grad_norm": 2.1973612308502197, + "learning_rate": 0.00019188627227894098, + "loss": 0.9451, + "step": 8949 + }, + { + "epoch": 0.6210533620151273, + "grad_norm": 2.374368667602539, + "learning_rate": 0.00019188405515542026, + "loss": 0.9881, + "step": 8950 + }, + { + "epoch": 0.621122753452224, + "grad_norm": 2.3920187950134277, + "learning_rate": 0.00019188183774183223, + "loss": 0.9474, + "step": 8951 + }, + { + "epoch": 0.6211921448893206, + "grad_norm": 2.8504748344421387, + "learning_rate": 0.0001918796200381839, + "loss": 1.0641, + "step": 8952 + }, + { + "epoch": 0.6212615363264173, + "grad_norm": 2.567817211151123, + "learning_rate": 0.00019187740204448233, + "loss": 0.9012, + "step": 8953 + }, + { + "epoch": 0.621330927763514, + "grad_norm": 2.553905725479126, + "learning_rate": 0.0001918751837607344, + "loss": 0.8024, + "step": 8954 + }, + { + "epoch": 0.6214003192006107, + "grad_norm": 3.1706795692443848, + "learning_rate": 0.00019187296518694718, + "loss": 0.9123, + "step": 8955 + }, + { + "epoch": 0.6214697106377073, + "grad_norm": 2.191941261291504, + "learning_rate": 0.00019187074632312767, + "loss": 0.7804, + "step": 8956 + }, + { + "epoch": 0.621539102074804, + "grad_norm": 3.6207072734832764, + "learning_rate": 0.00019186852716928288, + "loss": 0.9268, + "step": 8957 + }, + { + "epoch": 0.6216084935119006, + "grad_norm": 2.3843932151794434, + "learning_rate": 0.0001918663077254198, + "loss": 0.9268, + "step": 8958 + }, + { + "epoch": 0.6216778849489973, + "grad_norm": 2.1530256271362305, + "learning_rate": 0.00019186408799154545, + "loss": 0.6352, + "step": 8959 + }, + { + "epoch": 0.621747276386094, + "grad_norm": 3.458899974822998, + "learning_rate": 0.00019186186796766683, + "loss": 0.8764, + "step": 8960 + }, + { + "epoch": 0.6218166678231907, + "grad_norm": 2.1836884021759033, + "learning_rate": 0.00019185964765379096, + "loss": 0.6987, + "step": 8961 + }, + { + "epoch": 0.6218860592602873, + "grad_norm": 2.0372092723846436, + "learning_rate": 0.00019185742704992483, + "loss": 0.7061, + "step": 8962 + }, + { + "epoch": 0.6219554506973839, + "grad_norm": 2.2313578128814697, + "learning_rate": 0.00019185520615607547, + "loss": 0.7438, + "step": 8963 + }, + { + "epoch": 0.6220248421344806, + "grad_norm": 2.6149282455444336, + "learning_rate": 0.00019185298497224987, + "loss": 0.8387, + "step": 8964 + }, + { + "epoch": 0.6220942335715772, + "grad_norm": 2.6155319213867188, + "learning_rate": 0.00019185076349845508, + "loss": 0.8046, + "step": 8965 + }, + { + "epoch": 0.622163625008674, + "grad_norm": 3.9393997192382812, + "learning_rate": 0.00019184854173469806, + "loss": 0.8054, + "step": 8966 + }, + { + "epoch": 0.6222330164457706, + "grad_norm": 3.2764854431152344, + "learning_rate": 0.00019184631968098583, + "loss": 1.0465, + "step": 8967 + }, + { + "epoch": 0.6223024078828673, + "grad_norm": 1.8734489679336548, + "learning_rate": 0.00019184409733732544, + "loss": 0.6867, + "step": 8968 + }, + { + "epoch": 0.6223717993199639, + "grad_norm": 3.3721468448638916, + "learning_rate": 0.00019184187470372392, + "loss": 0.9972, + "step": 8969 + }, + { + "epoch": 0.6224411907570606, + "grad_norm": 2.192429304122925, + "learning_rate": 0.00019183965178018824, + "loss": 0.8368, + "step": 8970 + }, + { + "epoch": 0.6225105821941572, + "grad_norm": 2.9099862575531006, + "learning_rate": 0.00019183742856672544, + "loss": 0.8868, + "step": 8971 + }, + { + "epoch": 0.622579973631254, + "grad_norm": 3.7605113983154297, + "learning_rate": 0.0001918352050633425, + "loss": 0.9152, + "step": 8972 + }, + { + "epoch": 0.6226493650683506, + "grad_norm": 1.9960936307907104, + "learning_rate": 0.00019183298127004652, + "loss": 0.6797, + "step": 8973 + }, + { + "epoch": 0.6227187565054473, + "grad_norm": 2.1189146041870117, + "learning_rate": 0.00019183075718684443, + "loss": 1.005, + "step": 8974 + }, + { + "epoch": 0.6227881479425439, + "grad_norm": 2.4498095512390137, + "learning_rate": 0.00019182853281374333, + "loss": 1.0378, + "step": 8975 + }, + { + "epoch": 0.6228575393796405, + "grad_norm": 3.1275789737701416, + "learning_rate": 0.00019182630815075017, + "loss": 0.8347, + "step": 8976 + }, + { + "epoch": 0.6229269308167372, + "grad_norm": 2.6515495777130127, + "learning_rate": 0.00019182408319787207, + "loss": 0.7264, + "step": 8977 + }, + { + "epoch": 0.6229963222538338, + "grad_norm": 2.63248348236084, + "learning_rate": 0.00019182185795511595, + "loss": 0.8321, + "step": 8978 + }, + { + "epoch": 0.6230657136909306, + "grad_norm": 2.1203408241271973, + "learning_rate": 0.00019181963242248886, + "loss": 0.8263, + "step": 8979 + }, + { + "epoch": 0.6231351051280272, + "grad_norm": 2.457195520401001, + "learning_rate": 0.00019181740659999786, + "loss": 0.8337, + "step": 8980 + }, + { + "epoch": 0.6232044965651239, + "grad_norm": 2.633174180984497, + "learning_rate": 0.00019181518048764996, + "loss": 0.9746, + "step": 8981 + }, + { + "epoch": 0.6232738880022205, + "grad_norm": 3.319136619567871, + "learning_rate": 0.00019181295408545222, + "loss": 0.8789, + "step": 8982 + }, + { + "epoch": 0.6233432794393172, + "grad_norm": 2.98030161857605, + "learning_rate": 0.00019181072739341158, + "loss": 0.971, + "step": 8983 + }, + { + "epoch": 0.6234126708764138, + "grad_norm": 2.1243293285369873, + "learning_rate": 0.00019180850041153517, + "loss": 0.8977, + "step": 8984 + }, + { + "epoch": 0.6234820623135106, + "grad_norm": 3.875333786010742, + "learning_rate": 0.00019180627313982994, + "loss": 0.695, + "step": 8985 + }, + { + "epoch": 0.6235514537506072, + "grad_norm": 2.2223663330078125, + "learning_rate": 0.000191804045578303, + "loss": 0.7009, + "step": 8986 + }, + { + "epoch": 0.6236208451877039, + "grad_norm": 3.304042100906372, + "learning_rate": 0.0001918018177269613, + "loss": 1.23, + "step": 8987 + }, + { + "epoch": 0.6236902366248005, + "grad_norm": 2.4755749702453613, + "learning_rate": 0.00019179958958581193, + "loss": 0.7915, + "step": 8988 + }, + { + "epoch": 0.6237596280618971, + "grad_norm": 2.2463135719299316, + "learning_rate": 0.0001917973611548619, + "loss": 0.7123, + "step": 8989 + }, + { + "epoch": 0.6238290194989938, + "grad_norm": 2.3963747024536133, + "learning_rate": 0.00019179513243411825, + "loss": 0.9592, + "step": 8990 + }, + { + "epoch": 0.6238984109360904, + "grad_norm": 2.5999298095703125, + "learning_rate": 0.00019179290342358804, + "loss": 1.0253, + "step": 8991 + }, + { + "epoch": 0.6239678023731872, + "grad_norm": 2.307976722717285, + "learning_rate": 0.00019179067412327828, + "loss": 0.992, + "step": 8992 + }, + { + "epoch": 0.6240371938102838, + "grad_norm": 2.5163214206695557, + "learning_rate": 0.00019178844453319598, + "loss": 0.8539, + "step": 8993 + }, + { + "epoch": 0.6241065852473805, + "grad_norm": 2.3415842056274414, + "learning_rate": 0.00019178621465334823, + "loss": 0.9249, + "step": 8994 + }, + { + "epoch": 0.6241759766844771, + "grad_norm": 2.3179266452789307, + "learning_rate": 0.00019178398448374204, + "loss": 0.7046, + "step": 8995 + }, + { + "epoch": 0.6242453681215738, + "grad_norm": 2.310607671737671, + "learning_rate": 0.0001917817540243845, + "loss": 1.0263, + "step": 8996 + }, + { + "epoch": 0.6243147595586704, + "grad_norm": 2.6595330238342285, + "learning_rate": 0.00019177952327528256, + "loss": 1.1906, + "step": 8997 + }, + { + "epoch": 0.6243841509957672, + "grad_norm": 2.6913013458251953, + "learning_rate": 0.00019177729223644331, + "loss": 0.9444, + "step": 8998 + }, + { + "epoch": 0.6244535424328638, + "grad_norm": 2.1216065883636475, + "learning_rate": 0.00019177506090787384, + "loss": 0.9399, + "step": 8999 + }, + { + "epoch": 0.6245229338699605, + "grad_norm": 2.7565886974334717, + "learning_rate": 0.00019177282928958112, + "loss": 0.8174, + "step": 9000 + }, + { + "epoch": 0.6245923253070571, + "grad_norm": 2.346510171890259, + "learning_rate": 0.00019177059738157223, + "loss": 0.8297, + "step": 9001 + }, + { + "epoch": 0.6246617167441537, + "grad_norm": 2.7501065731048584, + "learning_rate": 0.00019176836518385422, + "loss": 0.7082, + "step": 9002 + }, + { + "epoch": 0.6247311081812504, + "grad_norm": 2.433074474334717, + "learning_rate": 0.00019176613269643414, + "loss": 0.7513, + "step": 9003 + }, + { + "epoch": 0.624800499618347, + "grad_norm": 2.07104229927063, + "learning_rate": 0.000191763899919319, + "loss": 0.8473, + "step": 9004 + }, + { + "epoch": 0.6248698910554438, + "grad_norm": 2.716318130493164, + "learning_rate": 0.0001917616668525159, + "loss": 1.0623, + "step": 9005 + }, + { + "epoch": 0.6249392824925404, + "grad_norm": 2.9585201740264893, + "learning_rate": 0.00019175943349603185, + "loss": 0.9579, + "step": 9006 + }, + { + "epoch": 0.6250086739296371, + "grad_norm": 3.0690934658050537, + "learning_rate": 0.0001917571998498739, + "loss": 0.8133, + "step": 9007 + }, + { + "epoch": 0.6250780653667337, + "grad_norm": 2.7131409645080566, + "learning_rate": 0.00019175496591404914, + "loss": 0.906, + "step": 9008 + }, + { + "epoch": 0.6251474568038304, + "grad_norm": 3.054006576538086, + "learning_rate": 0.00019175273168856458, + "loss": 1.1333, + "step": 9009 + }, + { + "epoch": 0.625216848240927, + "grad_norm": 2.2175354957580566, + "learning_rate": 0.0001917504971734273, + "loss": 0.7143, + "step": 9010 + }, + { + "epoch": 0.6252862396780238, + "grad_norm": 2.631147861480713, + "learning_rate": 0.00019174826236864436, + "loss": 0.8466, + "step": 9011 + }, + { + "epoch": 0.6253556311151204, + "grad_norm": 2.9898436069488525, + "learning_rate": 0.00019174602727422277, + "loss": 0.9732, + "step": 9012 + }, + { + "epoch": 0.6254250225522171, + "grad_norm": 2.623594284057617, + "learning_rate": 0.00019174379189016962, + "loss": 0.6994, + "step": 9013 + }, + { + "epoch": 0.6254944139893137, + "grad_norm": 2.0854523181915283, + "learning_rate": 0.000191741556216492, + "loss": 0.5715, + "step": 9014 + }, + { + "epoch": 0.6255638054264104, + "grad_norm": 2.294477701187134, + "learning_rate": 0.0001917393202531969, + "loss": 0.831, + "step": 9015 + }, + { + "epoch": 0.625633196863507, + "grad_norm": 2.825263500213623, + "learning_rate": 0.0001917370840002914, + "loss": 1.1042, + "step": 9016 + }, + { + "epoch": 0.6257025883006037, + "grad_norm": 2.638780355453491, + "learning_rate": 0.0001917348474577826, + "loss": 1.0007, + "step": 9017 + }, + { + "epoch": 0.6257719797377004, + "grad_norm": 2.6235616207122803, + "learning_rate": 0.0001917326106256775, + "loss": 0.8035, + "step": 9018 + }, + { + "epoch": 0.625841371174797, + "grad_norm": 2.769407033920288, + "learning_rate": 0.00019173037350398323, + "loss": 0.9066, + "step": 9019 + }, + { + "epoch": 0.6259107626118937, + "grad_norm": 2.2403600215911865, + "learning_rate": 0.00019172813609270677, + "loss": 0.9858, + "step": 9020 + }, + { + "epoch": 0.6259801540489903, + "grad_norm": 3.4983479976654053, + "learning_rate": 0.00019172589839185525, + "loss": 1.0353, + "step": 9021 + }, + { + "epoch": 0.626049545486087, + "grad_norm": 2.755525827407837, + "learning_rate": 0.0001917236604014357, + "loss": 1.0231, + "step": 9022 + }, + { + "epoch": 0.6261189369231837, + "grad_norm": 2.441209316253662, + "learning_rate": 0.0001917214221214552, + "loss": 0.7922, + "step": 9023 + }, + { + "epoch": 0.6261883283602804, + "grad_norm": 3.158534049987793, + "learning_rate": 0.0001917191835519208, + "loss": 0.8429, + "step": 9024 + }, + { + "epoch": 0.626257719797377, + "grad_norm": 2.7499382495880127, + "learning_rate": 0.0001917169446928396, + "loss": 0.5842, + "step": 9025 + }, + { + "epoch": 0.6263271112344737, + "grad_norm": 1.9865984916687012, + "learning_rate": 0.00019171470554421867, + "loss": 0.9575, + "step": 9026 + }, + { + "epoch": 0.6263965026715703, + "grad_norm": 2.001722574234009, + "learning_rate": 0.000191712466106065, + "loss": 0.8015, + "step": 9027 + }, + { + "epoch": 0.626465894108667, + "grad_norm": 2.4160807132720947, + "learning_rate": 0.00019171022637838576, + "loss": 0.7996, + "step": 9028 + }, + { + "epoch": 0.6265352855457637, + "grad_norm": 3.0524744987487793, + "learning_rate": 0.00019170798636118796, + "loss": 0.9938, + "step": 9029 + }, + { + "epoch": 0.6266046769828603, + "grad_norm": 2.322335958480835, + "learning_rate": 0.00019170574605447866, + "loss": 0.9067, + "step": 9030 + }, + { + "epoch": 0.626674068419957, + "grad_norm": 2.609659194946289, + "learning_rate": 0.000191703505458265, + "loss": 0.8082, + "step": 9031 + }, + { + "epoch": 0.6267434598570536, + "grad_norm": 2.4012796878814697, + "learning_rate": 0.00019170126457255398, + "loss": 0.7914, + "step": 9032 + }, + { + "epoch": 0.6268128512941503, + "grad_norm": 2.5096781253814697, + "learning_rate": 0.00019169902339735275, + "loss": 0.7639, + "step": 9033 + }, + { + "epoch": 0.6268822427312469, + "grad_norm": 3.8645477294921875, + "learning_rate": 0.00019169678193266832, + "loss": 0.7927, + "step": 9034 + }, + { + "epoch": 0.6269516341683437, + "grad_norm": 2.4377129077911377, + "learning_rate": 0.00019169454017850778, + "loss": 1.0052, + "step": 9035 + }, + { + "epoch": 0.6270210256054403, + "grad_norm": 1.9396281242370605, + "learning_rate": 0.0001916922981348782, + "loss": 0.6448, + "step": 9036 + }, + { + "epoch": 0.627090417042537, + "grad_norm": 3.5931711196899414, + "learning_rate": 0.00019169005580178672, + "loss": 0.9222, + "step": 9037 + }, + { + "epoch": 0.6271598084796336, + "grad_norm": 2.7080163955688477, + "learning_rate": 0.00019168781317924034, + "loss": 0.9116, + "step": 9038 + }, + { + "epoch": 0.6272291999167303, + "grad_norm": 3.373530149459839, + "learning_rate": 0.00019168557026724616, + "loss": 1.0082, + "step": 9039 + }, + { + "epoch": 0.6272985913538269, + "grad_norm": 2.58632755279541, + "learning_rate": 0.00019168332706581128, + "loss": 0.6618, + "step": 9040 + }, + { + "epoch": 0.6273679827909237, + "grad_norm": 4.087575435638428, + "learning_rate": 0.00019168108357494278, + "loss": 1.0215, + "step": 9041 + }, + { + "epoch": 0.6274373742280203, + "grad_norm": 2.3299803733825684, + "learning_rate": 0.00019167883979464775, + "loss": 0.6856, + "step": 9042 + }, + { + "epoch": 0.6275067656651169, + "grad_norm": 2.560014009475708, + "learning_rate": 0.00019167659572493325, + "loss": 0.8507, + "step": 9043 + }, + { + "epoch": 0.6275761571022136, + "grad_norm": 2.955315113067627, + "learning_rate": 0.00019167435136580637, + "loss": 0.8046, + "step": 9044 + }, + { + "epoch": 0.6276455485393102, + "grad_norm": 2.6485886573791504, + "learning_rate": 0.0001916721067172742, + "loss": 0.9319, + "step": 9045 + }, + { + "epoch": 0.6277149399764069, + "grad_norm": 3.1251933574676514, + "learning_rate": 0.00019166986177934382, + "loss": 0.8447, + "step": 9046 + }, + { + "epoch": 0.6277843314135035, + "grad_norm": 2.5123813152313232, + "learning_rate": 0.00019166761655202232, + "loss": 0.7624, + "step": 9047 + }, + { + "epoch": 0.6278537228506003, + "grad_norm": 3.5131285190582275, + "learning_rate": 0.0001916653710353168, + "loss": 1.1467, + "step": 9048 + }, + { + "epoch": 0.6279231142876969, + "grad_norm": 2.742980718612671, + "learning_rate": 0.00019166312522923433, + "loss": 1.0706, + "step": 9049 + }, + { + "epoch": 0.6279925057247936, + "grad_norm": 2.843921661376953, + "learning_rate": 0.00019166087913378204, + "loss": 0.725, + "step": 9050 + }, + { + "epoch": 0.6280618971618902, + "grad_norm": 2.7948246002197266, + "learning_rate": 0.00019165863274896692, + "loss": 0.9323, + "step": 9051 + }, + { + "epoch": 0.6281312885989869, + "grad_norm": 2.564244508743286, + "learning_rate": 0.00019165638607479622, + "loss": 0.9042, + "step": 9052 + }, + { + "epoch": 0.6282006800360835, + "grad_norm": 3.1430861949920654, + "learning_rate": 0.0001916541391112769, + "loss": 1.1242, + "step": 9053 + }, + { + "epoch": 0.6282700714731803, + "grad_norm": 2.720280408859253, + "learning_rate": 0.00019165189185841607, + "loss": 0.9453, + "step": 9054 + }, + { + "epoch": 0.6283394629102769, + "grad_norm": 2.85105037689209, + "learning_rate": 0.00019164964431622088, + "loss": 0.8557, + "step": 9055 + }, + { + "epoch": 0.6284088543473735, + "grad_norm": 2.5462899208068848, + "learning_rate": 0.0001916473964846984, + "loss": 0.8464, + "step": 9056 + }, + { + "epoch": 0.6284782457844702, + "grad_norm": 3.059483289718628, + "learning_rate": 0.00019164514836385573, + "loss": 1.0464, + "step": 9057 + }, + { + "epoch": 0.6285476372215668, + "grad_norm": 3.9078714847564697, + "learning_rate": 0.00019164289995369992, + "loss": 0.8024, + "step": 9058 + }, + { + "epoch": 0.6286170286586635, + "grad_norm": 2.308924674987793, + "learning_rate": 0.00019164065125423815, + "loss": 0.687, + "step": 9059 + }, + { + "epoch": 0.6286864200957601, + "grad_norm": 3.460455894470215, + "learning_rate": 0.00019163840226547744, + "loss": 0.9965, + "step": 9060 + }, + { + "epoch": 0.6287558115328569, + "grad_norm": 3.736628293991089, + "learning_rate": 0.00019163615298742494, + "loss": 1.0116, + "step": 9061 + }, + { + "epoch": 0.6288252029699535, + "grad_norm": 3.2361562252044678, + "learning_rate": 0.00019163390342008773, + "loss": 1.0055, + "step": 9062 + }, + { + "epoch": 0.6288945944070502, + "grad_norm": 5.215470790863037, + "learning_rate": 0.00019163165356347292, + "loss": 0.9646, + "step": 9063 + }, + { + "epoch": 0.6289639858441468, + "grad_norm": 2.7977325916290283, + "learning_rate": 0.00019162940341758763, + "loss": 1.0339, + "step": 9064 + }, + { + "epoch": 0.6290333772812435, + "grad_norm": 1.9607957601547241, + "learning_rate": 0.00019162715298243893, + "loss": 0.8911, + "step": 9065 + }, + { + "epoch": 0.6291027687183401, + "grad_norm": 3.2270185947418213, + "learning_rate": 0.00019162490225803396, + "loss": 1.0535, + "step": 9066 + }, + { + "epoch": 0.6291721601554369, + "grad_norm": 4.877711296081543, + "learning_rate": 0.00019162265124437976, + "loss": 1.2449, + "step": 9067 + }, + { + "epoch": 0.6292415515925335, + "grad_norm": 2.7772977352142334, + "learning_rate": 0.00019162039994148353, + "loss": 0.8599, + "step": 9068 + }, + { + "epoch": 0.6293109430296302, + "grad_norm": 2.6805005073547363, + "learning_rate": 0.0001916181483493523, + "loss": 0.9342, + "step": 9069 + }, + { + "epoch": 0.6293803344667268, + "grad_norm": 2.7065060138702393, + "learning_rate": 0.0001916158964679932, + "loss": 0.9517, + "step": 9070 + }, + { + "epoch": 0.6294497259038234, + "grad_norm": 2.7105395793914795, + "learning_rate": 0.00019161364429741336, + "loss": 0.7074, + "step": 9071 + }, + { + "epoch": 0.6295191173409201, + "grad_norm": 2.8264830112457275, + "learning_rate": 0.00019161139183761987, + "loss": 0.8934, + "step": 9072 + }, + { + "epoch": 0.6295885087780168, + "grad_norm": 2.7054359912872314, + "learning_rate": 0.00019160913908861983, + "loss": 0.9187, + "step": 9073 + }, + { + "epoch": 0.6296579002151135, + "grad_norm": 2.531010389328003, + "learning_rate": 0.00019160688605042038, + "loss": 0.7317, + "step": 9074 + }, + { + "epoch": 0.6297272916522101, + "grad_norm": 2.3957302570343018, + "learning_rate": 0.00019160463272302862, + "loss": 0.9033, + "step": 9075 + }, + { + "epoch": 0.6297966830893068, + "grad_norm": 1.931024432182312, + "learning_rate": 0.00019160237910645166, + "loss": 0.6782, + "step": 9076 + }, + { + "epoch": 0.6298660745264034, + "grad_norm": 2.262002944946289, + "learning_rate": 0.00019160012520069662, + "loss": 0.841, + "step": 9077 + }, + { + "epoch": 0.6299354659635001, + "grad_norm": 2.7901906967163086, + "learning_rate": 0.0001915978710057706, + "loss": 1.0271, + "step": 9078 + }, + { + "epoch": 0.6300048574005968, + "grad_norm": 3.168980360031128, + "learning_rate": 0.00019159561652168074, + "loss": 1.1818, + "step": 9079 + }, + { + "epoch": 0.6300742488376935, + "grad_norm": 2.822967052459717, + "learning_rate": 0.00019159336174843415, + "loss": 0.7949, + "step": 9080 + }, + { + "epoch": 0.6301436402747901, + "grad_norm": 3.131833791732788, + "learning_rate": 0.00019159110668603795, + "loss": 0.8704, + "step": 9081 + }, + { + "epoch": 0.6302130317118868, + "grad_norm": 2.402817726135254, + "learning_rate": 0.00019158885133449923, + "loss": 0.9318, + "step": 9082 + }, + { + "epoch": 0.6302824231489834, + "grad_norm": 2.403627872467041, + "learning_rate": 0.00019158659569382515, + "loss": 1.2313, + "step": 9083 + }, + { + "epoch": 0.63035181458608, + "grad_norm": 2.7086427211761475, + "learning_rate": 0.0001915843397640228, + "loss": 1.238, + "step": 9084 + }, + { + "epoch": 0.6304212060231768, + "grad_norm": 2.740888833999634, + "learning_rate": 0.00019158208354509934, + "loss": 0.8958, + "step": 9085 + }, + { + "epoch": 0.6304905974602734, + "grad_norm": 2.8838841915130615, + "learning_rate": 0.00019157982703706184, + "loss": 0.8273, + "step": 9086 + }, + { + "epoch": 0.6305599888973701, + "grad_norm": 2.555420398712158, + "learning_rate": 0.00019157757023991745, + "loss": 0.6907, + "step": 9087 + }, + { + "epoch": 0.6306293803344667, + "grad_norm": 2.440086841583252, + "learning_rate": 0.00019157531315367333, + "loss": 0.7373, + "step": 9088 + }, + { + "epoch": 0.6306987717715634, + "grad_norm": 2.918382406234741, + "learning_rate": 0.00019157305577833654, + "loss": 0.9947, + "step": 9089 + }, + { + "epoch": 0.63076816320866, + "grad_norm": 2.2427985668182373, + "learning_rate": 0.00019157079811391424, + "loss": 0.9523, + "step": 9090 + }, + { + "epoch": 0.6308375546457567, + "grad_norm": 2.5283637046813965, + "learning_rate": 0.00019156854016041356, + "loss": 0.8247, + "step": 9091 + }, + { + "epoch": 0.6309069460828534, + "grad_norm": 2.2781291007995605, + "learning_rate": 0.0001915662819178416, + "loss": 0.7199, + "step": 9092 + }, + { + "epoch": 0.6309763375199501, + "grad_norm": 2.1749982833862305, + "learning_rate": 0.00019156402338620554, + "loss": 0.7094, + "step": 9093 + }, + { + "epoch": 0.6310457289570467, + "grad_norm": 2.7068912982940674, + "learning_rate": 0.0001915617645655125, + "loss": 0.7192, + "step": 9094 + }, + { + "epoch": 0.6311151203941434, + "grad_norm": 2.6067590713500977, + "learning_rate": 0.00019155950545576955, + "loss": 1.0186, + "step": 9095 + }, + { + "epoch": 0.63118451183124, + "grad_norm": 2.8715291023254395, + "learning_rate": 0.0001915572460569839, + "loss": 1.0789, + "step": 9096 + }, + { + "epoch": 0.6312539032683366, + "grad_norm": 1.9414000511169434, + "learning_rate": 0.00019155498636916262, + "loss": 0.5745, + "step": 9097 + }, + { + "epoch": 0.6313232947054334, + "grad_norm": 2.860163450241089, + "learning_rate": 0.0001915527263923129, + "loss": 0.9685, + "step": 9098 + }, + { + "epoch": 0.63139268614253, + "grad_norm": 2.65067982673645, + "learning_rate": 0.00019155046612644177, + "loss": 1.0772, + "step": 9099 + }, + { + "epoch": 0.6314620775796267, + "grad_norm": 2.1077754497528076, + "learning_rate": 0.00019154820557155654, + "loss": 0.7229, + "step": 9100 + }, + { + "epoch": 0.6315314690167233, + "grad_norm": 2.3959028720855713, + "learning_rate": 0.00019154594472766418, + "loss": 0.7917, + "step": 9101 + }, + { + "epoch": 0.63160086045382, + "grad_norm": 2.6432862281799316, + "learning_rate": 0.00019154368359477193, + "loss": 0.8413, + "step": 9102 + }, + { + "epoch": 0.6316702518909166, + "grad_norm": 3.504206895828247, + "learning_rate": 0.00019154142217288686, + "loss": 0.9761, + "step": 9103 + }, + { + "epoch": 0.6317396433280134, + "grad_norm": 2.3903677463531494, + "learning_rate": 0.00019153916046201616, + "loss": 0.7946, + "step": 9104 + }, + { + "epoch": 0.63180903476511, + "grad_norm": 2.4851760864257812, + "learning_rate": 0.00019153689846216696, + "loss": 0.7571, + "step": 9105 + }, + { + "epoch": 0.6318784262022067, + "grad_norm": 2.849536657333374, + "learning_rate": 0.00019153463617334638, + "loss": 1.0297, + "step": 9106 + }, + { + "epoch": 0.6319478176393033, + "grad_norm": 2.321171760559082, + "learning_rate": 0.00019153237359556156, + "loss": 0.7754, + "step": 9107 + }, + { + "epoch": 0.6320172090764, + "grad_norm": 2.683687448501587, + "learning_rate": 0.00019153011072881967, + "loss": 0.9547, + "step": 9108 + }, + { + "epoch": 0.6320866005134966, + "grad_norm": 2.2671496868133545, + "learning_rate": 0.00019152784757312785, + "loss": 0.7841, + "step": 9109 + }, + { + "epoch": 0.6321559919505932, + "grad_norm": 2.2345969676971436, + "learning_rate": 0.00019152558412849322, + "loss": 0.847, + "step": 9110 + }, + { + "epoch": 0.63222538338769, + "grad_norm": 2.3726117610931396, + "learning_rate": 0.00019152332039492294, + "loss": 0.8491, + "step": 9111 + }, + { + "epoch": 0.6322947748247866, + "grad_norm": 3.1087255477905273, + "learning_rate": 0.00019152105637242417, + "loss": 1.0285, + "step": 9112 + }, + { + "epoch": 0.6323641662618833, + "grad_norm": 2.45654559135437, + "learning_rate": 0.00019151879206100403, + "loss": 0.9792, + "step": 9113 + }, + { + "epoch": 0.6324335576989799, + "grad_norm": 2.309743642807007, + "learning_rate": 0.00019151652746066968, + "loss": 0.8121, + "step": 9114 + }, + { + "epoch": 0.6325029491360766, + "grad_norm": 2.313112735748291, + "learning_rate": 0.00019151426257142827, + "loss": 0.7779, + "step": 9115 + }, + { + "epoch": 0.6325723405731732, + "grad_norm": 3.736380100250244, + "learning_rate": 0.00019151199739328697, + "loss": 1.1827, + "step": 9116 + }, + { + "epoch": 0.63264173201027, + "grad_norm": 2.174448251724243, + "learning_rate": 0.00019150973192625287, + "loss": 0.6842, + "step": 9117 + }, + { + "epoch": 0.6327111234473666, + "grad_norm": 2.485269546508789, + "learning_rate": 0.0001915074661703332, + "loss": 0.8454, + "step": 9118 + }, + { + "epoch": 0.6327805148844633, + "grad_norm": 2.80537748336792, + "learning_rate": 0.00019150520012553508, + "loss": 1.1463, + "step": 9119 + }, + { + "epoch": 0.6328499063215599, + "grad_norm": 2.263347625732422, + "learning_rate": 0.00019150293379186563, + "loss": 0.8097, + "step": 9120 + }, + { + "epoch": 0.6329192977586566, + "grad_norm": 2.533679485321045, + "learning_rate": 0.00019150066716933206, + "loss": 0.9854, + "step": 9121 + }, + { + "epoch": 0.6329886891957532, + "grad_norm": 1.9834171533584595, + "learning_rate": 0.0001914984002579415, + "loss": 0.5972, + "step": 9122 + }, + { + "epoch": 0.6330580806328499, + "grad_norm": 2.783750534057617, + "learning_rate": 0.0001914961330577011, + "loss": 0.7189, + "step": 9123 + }, + { + "epoch": 0.6331274720699466, + "grad_norm": 3.339254856109619, + "learning_rate": 0.00019149386556861805, + "loss": 0.8964, + "step": 9124 + }, + { + "epoch": 0.6331968635070432, + "grad_norm": 2.3409359455108643, + "learning_rate": 0.00019149159779069944, + "loss": 1.07, + "step": 9125 + }, + { + "epoch": 0.6332662549441399, + "grad_norm": 2.3314080238342285, + "learning_rate": 0.00019148932972395247, + "loss": 1.027, + "step": 9126 + }, + { + "epoch": 0.6333356463812365, + "grad_norm": 2.518779993057251, + "learning_rate": 0.0001914870613683843, + "loss": 0.9839, + "step": 9127 + }, + { + "epoch": 0.6334050378183332, + "grad_norm": 2.126234292984009, + "learning_rate": 0.00019148479272400212, + "loss": 0.7863, + "step": 9128 + }, + { + "epoch": 0.6334744292554298, + "grad_norm": 1.9751349687576294, + "learning_rate": 0.00019148252379081306, + "loss": 0.7342, + "step": 9129 + }, + { + "epoch": 0.6335438206925266, + "grad_norm": 2.5994629859924316, + "learning_rate": 0.00019148025456882425, + "loss": 0.9739, + "step": 9130 + }, + { + "epoch": 0.6336132121296232, + "grad_norm": 2.6081361770629883, + "learning_rate": 0.00019147798505804293, + "loss": 0.9051, + "step": 9131 + }, + { + "epoch": 0.6336826035667199, + "grad_norm": 2.3517425060272217, + "learning_rate": 0.0001914757152584762, + "loss": 0.9041, + "step": 9132 + }, + { + "epoch": 0.6337519950038165, + "grad_norm": 2.6989307403564453, + "learning_rate": 0.00019147344517013126, + "loss": 1.1789, + "step": 9133 + }, + { + "epoch": 0.6338213864409132, + "grad_norm": 2.7233457565307617, + "learning_rate": 0.0001914711747930153, + "loss": 1.2679, + "step": 9134 + }, + { + "epoch": 0.6338907778780098, + "grad_norm": 2.418156862258911, + "learning_rate": 0.00019146890412713541, + "loss": 0.8214, + "step": 9135 + }, + { + "epoch": 0.6339601693151066, + "grad_norm": 2.312093496322632, + "learning_rate": 0.0001914666331724988, + "loss": 1.1339, + "step": 9136 + }, + { + "epoch": 0.6340295607522032, + "grad_norm": 1.7052397727966309, + "learning_rate": 0.00019146436192911265, + "loss": 0.819, + "step": 9137 + }, + { + "epoch": 0.6340989521892998, + "grad_norm": 1.9400089979171753, + "learning_rate": 0.00019146209039698413, + "loss": 0.7391, + "step": 9138 + }, + { + "epoch": 0.6341683436263965, + "grad_norm": 2.7957472801208496, + "learning_rate": 0.0001914598185761204, + "loss": 0.8906, + "step": 9139 + }, + { + "epoch": 0.6342377350634931, + "grad_norm": 2.6726768016815186, + "learning_rate": 0.00019145754646652865, + "loss": 0.8441, + "step": 9140 + }, + { + "epoch": 0.6343071265005898, + "grad_norm": 2.523322343826294, + "learning_rate": 0.00019145527406821603, + "loss": 0.5806, + "step": 9141 + }, + { + "epoch": 0.6343765179376865, + "grad_norm": 2.292444944381714, + "learning_rate": 0.00019145300138118966, + "loss": 0.9896, + "step": 9142 + }, + { + "epoch": 0.6344459093747832, + "grad_norm": 2.4725027084350586, + "learning_rate": 0.00019145072840545685, + "loss": 0.7475, + "step": 9143 + }, + { + "epoch": 0.6345153008118798, + "grad_norm": 1.9270713329315186, + "learning_rate": 0.00019144845514102468, + "loss": 0.8665, + "step": 9144 + }, + { + "epoch": 0.6345846922489765, + "grad_norm": 2.9293155670166016, + "learning_rate": 0.00019144618158790033, + "loss": 0.8822, + "step": 9145 + }, + { + "epoch": 0.6346540836860731, + "grad_norm": 2.74513840675354, + "learning_rate": 0.000191443907746091, + "loss": 0.7763, + "step": 9146 + }, + { + "epoch": 0.6347234751231698, + "grad_norm": 2.2603440284729004, + "learning_rate": 0.00019144163361560387, + "loss": 0.7762, + "step": 9147 + }, + { + "epoch": 0.6347928665602665, + "grad_norm": 2.423574209213257, + "learning_rate": 0.00019143935919644611, + "loss": 0.8783, + "step": 9148 + }, + { + "epoch": 0.6348622579973632, + "grad_norm": 2.2498693466186523, + "learning_rate": 0.00019143708448862492, + "loss": 0.8089, + "step": 9149 + }, + { + "epoch": 0.6349316494344598, + "grad_norm": 2.197631359100342, + "learning_rate": 0.00019143480949214742, + "loss": 0.9109, + "step": 9150 + }, + { + "epoch": 0.6350010408715564, + "grad_norm": 2.419316291809082, + "learning_rate": 0.00019143253420702088, + "loss": 0.9634, + "step": 9151 + }, + { + "epoch": 0.6350704323086531, + "grad_norm": 2.255864381790161, + "learning_rate": 0.0001914302586332524, + "loss": 0.8036, + "step": 9152 + }, + { + "epoch": 0.6351398237457497, + "grad_norm": 1.840802788734436, + "learning_rate": 0.00019142798277084925, + "loss": 0.6371, + "step": 9153 + }, + { + "epoch": 0.6352092151828465, + "grad_norm": 2.190035581588745, + "learning_rate": 0.00019142570661981854, + "loss": 0.8211, + "step": 9154 + }, + { + "epoch": 0.6352786066199431, + "grad_norm": 2.0281200408935547, + "learning_rate": 0.00019142343018016748, + "loss": 0.8099, + "step": 9155 + }, + { + "epoch": 0.6353479980570398, + "grad_norm": 2.4722506999969482, + "learning_rate": 0.00019142115345190323, + "loss": 0.8476, + "step": 9156 + }, + { + "epoch": 0.6354173894941364, + "grad_norm": 2.7523467540740967, + "learning_rate": 0.00019141887643503306, + "loss": 0.868, + "step": 9157 + }, + { + "epoch": 0.6354867809312331, + "grad_norm": 3.259320020675659, + "learning_rate": 0.00019141659912956405, + "loss": 1.2234, + "step": 9158 + }, + { + "epoch": 0.6355561723683297, + "grad_norm": 2.659684419631958, + "learning_rate": 0.00019141432153550348, + "loss": 0.9764, + "step": 9159 + }, + { + "epoch": 0.6356255638054265, + "grad_norm": 2.4128451347351074, + "learning_rate": 0.00019141204365285847, + "loss": 1.2731, + "step": 9160 + }, + { + "epoch": 0.6356949552425231, + "grad_norm": 2.8232908248901367, + "learning_rate": 0.00019140976548163627, + "loss": 0.9329, + "step": 9161 + }, + { + "epoch": 0.6357643466796198, + "grad_norm": 2.1511361598968506, + "learning_rate": 0.00019140748702184404, + "loss": 0.957, + "step": 9162 + }, + { + "epoch": 0.6358337381167164, + "grad_norm": 1.8160756826400757, + "learning_rate": 0.000191405208273489, + "loss": 0.852, + "step": 9163 + }, + { + "epoch": 0.635903129553813, + "grad_norm": 2.210921049118042, + "learning_rate": 0.00019140292923657828, + "loss": 1.0416, + "step": 9164 + }, + { + "epoch": 0.6359725209909097, + "grad_norm": 1.7553517818450928, + "learning_rate": 0.00019140064991111913, + "loss": 0.6975, + "step": 9165 + }, + { + "epoch": 0.6360419124280063, + "grad_norm": 2.322591781616211, + "learning_rate": 0.00019139837029711875, + "loss": 1.0603, + "step": 9166 + }, + { + "epoch": 0.6361113038651031, + "grad_norm": 2.4603264331817627, + "learning_rate": 0.0001913960903945843, + "loss": 0.8728, + "step": 9167 + }, + { + "epoch": 0.6361806953021997, + "grad_norm": 3.416226863861084, + "learning_rate": 0.000191393810203523, + "loss": 1.0824, + "step": 9168 + }, + { + "epoch": 0.6362500867392964, + "grad_norm": 2.352030038833618, + "learning_rate": 0.00019139152972394205, + "loss": 0.7742, + "step": 9169 + }, + { + "epoch": 0.636319478176393, + "grad_norm": 2.135793447494507, + "learning_rate": 0.00019138924895584867, + "loss": 0.6389, + "step": 9170 + }, + { + "epoch": 0.6363888696134897, + "grad_norm": 3.6283414363861084, + "learning_rate": 0.00019138696789925002, + "loss": 1.0597, + "step": 9171 + }, + { + "epoch": 0.6364582610505863, + "grad_norm": 2.2377560138702393, + "learning_rate": 0.0001913846865541533, + "loss": 0.7992, + "step": 9172 + }, + { + "epoch": 0.6365276524876831, + "grad_norm": 1.8198257684707642, + "learning_rate": 0.00019138240492056575, + "loss": 0.5608, + "step": 9173 + }, + { + "epoch": 0.6365970439247797, + "grad_norm": 3.20052433013916, + "learning_rate": 0.0001913801229984945, + "loss": 0.7974, + "step": 9174 + }, + { + "epoch": 0.6366664353618764, + "grad_norm": 1.8941186666488647, + "learning_rate": 0.00019137784078794681, + "loss": 0.6863, + "step": 9175 + }, + { + "epoch": 0.636735826798973, + "grad_norm": 4.2864251136779785, + "learning_rate": 0.00019137555828892993, + "loss": 0.9603, + "step": 9176 + }, + { + "epoch": 0.6368052182360696, + "grad_norm": 2.7503232955932617, + "learning_rate": 0.000191373275501451, + "loss": 0.9911, + "step": 9177 + }, + { + "epoch": 0.6368746096731663, + "grad_norm": 2.1813931465148926, + "learning_rate": 0.00019137099242551723, + "loss": 0.662, + "step": 9178 + }, + { + "epoch": 0.636944001110263, + "grad_norm": 2.522475004196167, + "learning_rate": 0.00019136870906113584, + "loss": 0.8886, + "step": 9179 + }, + { + "epoch": 0.6370133925473597, + "grad_norm": 2.2284107208251953, + "learning_rate": 0.00019136642540831401, + "loss": 0.7404, + "step": 9180 + }, + { + "epoch": 0.6370827839844563, + "grad_norm": 2.7479324340820312, + "learning_rate": 0.000191364141467059, + "loss": 0.9612, + "step": 9181 + }, + { + "epoch": 0.637152175421553, + "grad_norm": 2.485902786254883, + "learning_rate": 0.000191361857237378, + "loss": 0.8962, + "step": 9182 + }, + { + "epoch": 0.6372215668586496, + "grad_norm": 2.413825750350952, + "learning_rate": 0.0001913595727192782, + "loss": 0.8085, + "step": 9183 + }, + { + "epoch": 0.6372909582957463, + "grad_norm": 2.7646420001983643, + "learning_rate": 0.00019135728791276683, + "loss": 0.9251, + "step": 9184 + }, + { + "epoch": 0.637360349732843, + "grad_norm": 2.2034804821014404, + "learning_rate": 0.00019135500281785114, + "loss": 0.8687, + "step": 9185 + }, + { + "epoch": 0.6374297411699397, + "grad_norm": 2.5305516719818115, + "learning_rate": 0.00019135271743453826, + "loss": 0.8922, + "step": 9186 + }, + { + "epoch": 0.6374991326070363, + "grad_norm": 2.2064766883850098, + "learning_rate": 0.00019135043176283546, + "loss": 0.9388, + "step": 9187 + }, + { + "epoch": 0.637568524044133, + "grad_norm": 4.547635555267334, + "learning_rate": 0.00019134814580274996, + "loss": 0.9425, + "step": 9188 + }, + { + "epoch": 0.6376379154812296, + "grad_norm": 2.2728588581085205, + "learning_rate": 0.00019134585955428896, + "loss": 0.4871, + "step": 9189 + }, + { + "epoch": 0.6377073069183263, + "grad_norm": 2.6712300777435303, + "learning_rate": 0.00019134357301745967, + "loss": 0.9057, + "step": 9190 + }, + { + "epoch": 0.6377766983554229, + "grad_norm": 1.9331518411636353, + "learning_rate": 0.00019134128619226933, + "loss": 0.7543, + "step": 9191 + }, + { + "epoch": 0.6378460897925196, + "grad_norm": 3.184574842453003, + "learning_rate": 0.00019133899907872512, + "loss": 0.797, + "step": 9192 + }, + { + "epoch": 0.6379154812296163, + "grad_norm": 2.516566514968872, + "learning_rate": 0.0001913367116768343, + "loss": 0.7517, + "step": 9193 + }, + { + "epoch": 0.6379848726667129, + "grad_norm": 2.4422545433044434, + "learning_rate": 0.00019133442398660408, + "loss": 0.9496, + "step": 9194 + }, + { + "epoch": 0.6380542641038096, + "grad_norm": 2.314793825149536, + "learning_rate": 0.0001913321360080417, + "loss": 1.0763, + "step": 9195 + }, + { + "epoch": 0.6381236555409062, + "grad_norm": 2.805800437927246, + "learning_rate": 0.00019132984774115433, + "loss": 0.8745, + "step": 9196 + }, + { + "epoch": 0.6381930469780029, + "grad_norm": 2.7581777572631836, + "learning_rate": 0.00019132755918594926, + "loss": 1.0208, + "step": 9197 + }, + { + "epoch": 0.6382624384150996, + "grad_norm": 1.9896231889724731, + "learning_rate": 0.00019132527034243366, + "loss": 0.8024, + "step": 9198 + }, + { + "epoch": 0.6383318298521963, + "grad_norm": 4.544515609741211, + "learning_rate": 0.00019132298121061478, + "loss": 0.8628, + "step": 9199 + }, + { + "epoch": 0.6384012212892929, + "grad_norm": 3.1994800567626953, + "learning_rate": 0.00019132069179049986, + "loss": 0.7038, + "step": 9200 + }, + { + "epoch": 0.6384706127263896, + "grad_norm": 2.0083999633789062, + "learning_rate": 0.00019131840208209608, + "loss": 0.888, + "step": 9201 + }, + { + "epoch": 0.6385400041634862, + "grad_norm": 2.847498893737793, + "learning_rate": 0.00019131611208541075, + "loss": 0.9282, + "step": 9202 + }, + { + "epoch": 0.6386093956005829, + "grad_norm": 2.7606945037841797, + "learning_rate": 0.000191313821800451, + "loss": 0.9801, + "step": 9203 + }, + { + "epoch": 0.6386787870376796, + "grad_norm": 3.532650947570801, + "learning_rate": 0.00019131153122722413, + "loss": 1.2898, + "step": 9204 + }, + { + "epoch": 0.6387481784747762, + "grad_norm": 3.0237555503845215, + "learning_rate": 0.00019130924036573732, + "loss": 1.2175, + "step": 9205 + }, + { + "epoch": 0.6388175699118729, + "grad_norm": 2.959573268890381, + "learning_rate": 0.00019130694921599788, + "loss": 1.0704, + "step": 9206 + }, + { + "epoch": 0.6388869613489695, + "grad_norm": 2.5089235305786133, + "learning_rate": 0.00019130465777801296, + "loss": 0.8319, + "step": 9207 + }, + { + "epoch": 0.6389563527860662, + "grad_norm": 2.0548391342163086, + "learning_rate": 0.00019130236605178984, + "loss": 0.8055, + "step": 9208 + }, + { + "epoch": 0.6390257442231628, + "grad_norm": 4.495128154754639, + "learning_rate": 0.00019130007403733574, + "loss": 0.9342, + "step": 9209 + }, + { + "epoch": 0.6390951356602596, + "grad_norm": 2.3855388164520264, + "learning_rate": 0.0001912977817346579, + "loss": 0.7472, + "step": 9210 + }, + { + "epoch": 0.6391645270973562, + "grad_norm": 2.793071746826172, + "learning_rate": 0.00019129548914376357, + "loss": 0.7993, + "step": 9211 + }, + { + "epoch": 0.6392339185344529, + "grad_norm": 3.1189990043640137, + "learning_rate": 0.00019129319626465995, + "loss": 0.7237, + "step": 9212 + }, + { + "epoch": 0.6393033099715495, + "grad_norm": 2.7327260971069336, + "learning_rate": 0.00019129090309735432, + "loss": 0.8906, + "step": 9213 + }, + { + "epoch": 0.6393727014086462, + "grad_norm": 3.9466018676757812, + "learning_rate": 0.0001912886096418539, + "loss": 1.1615, + "step": 9214 + }, + { + "epoch": 0.6394420928457428, + "grad_norm": 2.1881721019744873, + "learning_rate": 0.00019128631589816593, + "loss": 0.824, + "step": 9215 + }, + { + "epoch": 0.6395114842828395, + "grad_norm": 3.283890962600708, + "learning_rate": 0.00019128402186629763, + "loss": 0.989, + "step": 9216 + }, + { + "epoch": 0.6395808757199362, + "grad_norm": 2.5915424823760986, + "learning_rate": 0.00019128172754625627, + "loss": 0.8755, + "step": 9217 + }, + { + "epoch": 0.6396502671570328, + "grad_norm": 2.074918031692505, + "learning_rate": 0.00019127943293804908, + "loss": 0.8075, + "step": 9218 + }, + { + "epoch": 0.6397196585941295, + "grad_norm": 2.395172119140625, + "learning_rate": 0.00019127713804168334, + "loss": 0.7248, + "step": 9219 + }, + { + "epoch": 0.6397890500312261, + "grad_norm": 3.2002997398376465, + "learning_rate": 0.00019127484285716626, + "loss": 0.9124, + "step": 9220 + }, + { + "epoch": 0.6398584414683228, + "grad_norm": 2.639404773712158, + "learning_rate": 0.00019127254738450508, + "loss": 1.2768, + "step": 9221 + }, + { + "epoch": 0.6399278329054194, + "grad_norm": 2.1749119758605957, + "learning_rate": 0.00019127025162370706, + "loss": 0.9685, + "step": 9222 + }, + { + "epoch": 0.6399972243425162, + "grad_norm": 2.362011671066284, + "learning_rate": 0.00019126795557477944, + "loss": 1.0392, + "step": 9223 + }, + { + "epoch": 0.6400666157796128, + "grad_norm": 1.9068639278411865, + "learning_rate": 0.0001912656592377295, + "loss": 0.7531, + "step": 9224 + }, + { + "epoch": 0.6401360072167095, + "grad_norm": 2.3344316482543945, + "learning_rate": 0.00019126336261256443, + "loss": 0.5933, + "step": 9225 + }, + { + "epoch": 0.6402053986538061, + "grad_norm": 1.8507776260375977, + "learning_rate": 0.00019126106569929152, + "loss": 0.8888, + "step": 9226 + }, + { + "epoch": 0.6402747900909028, + "grad_norm": 2.191457509994507, + "learning_rate": 0.00019125876849791803, + "loss": 0.6224, + "step": 9227 + }, + { + "epoch": 0.6403441815279994, + "grad_norm": 2.046685218811035, + "learning_rate": 0.00019125647100845118, + "loss": 0.7675, + "step": 9228 + }, + { + "epoch": 0.6404135729650962, + "grad_norm": 3.058178663253784, + "learning_rate": 0.00019125417323089824, + "loss": 1.3325, + "step": 9229 + }, + { + "epoch": 0.6404829644021928, + "grad_norm": 2.2763547897338867, + "learning_rate": 0.00019125187516526646, + "loss": 1.1186, + "step": 9230 + }, + { + "epoch": 0.6405523558392894, + "grad_norm": 3.603092670440674, + "learning_rate": 0.0001912495768115631, + "loss": 1.0085, + "step": 9231 + }, + { + "epoch": 0.6406217472763861, + "grad_norm": 3.108917713165283, + "learning_rate": 0.0001912472781697954, + "loss": 0.9089, + "step": 9232 + }, + { + "epoch": 0.6406911387134827, + "grad_norm": 2.4331841468811035, + "learning_rate": 0.00019124497923997063, + "loss": 0.9212, + "step": 9233 + }, + { + "epoch": 0.6407605301505794, + "grad_norm": 2.477053642272949, + "learning_rate": 0.0001912426800220961, + "loss": 0.99, + "step": 9234 + }, + { + "epoch": 0.640829921587676, + "grad_norm": 2.836259126663208, + "learning_rate": 0.00019124038051617897, + "loss": 1.0107, + "step": 9235 + }, + { + "epoch": 0.6408993130247728, + "grad_norm": 2.0718188285827637, + "learning_rate": 0.00019123808072222655, + "loss": 0.7506, + "step": 9236 + }, + { + "epoch": 0.6409687044618694, + "grad_norm": 2.157879590988159, + "learning_rate": 0.0001912357806402461, + "loss": 0.6663, + "step": 9237 + }, + { + "epoch": 0.6410380958989661, + "grad_norm": 1.9512070417404175, + "learning_rate": 0.00019123348027024487, + "loss": 0.7832, + "step": 9238 + }, + { + "epoch": 0.6411074873360627, + "grad_norm": 2.6016359329223633, + "learning_rate": 0.00019123117961223013, + "loss": 1.0968, + "step": 9239 + }, + { + "epoch": 0.6411768787731594, + "grad_norm": 1.745674967765808, + "learning_rate": 0.00019122887866620912, + "loss": 0.7422, + "step": 9240 + }, + { + "epoch": 0.641246270210256, + "grad_norm": 2.6553730964660645, + "learning_rate": 0.00019122657743218916, + "loss": 0.8034, + "step": 9241 + }, + { + "epoch": 0.6413156616473528, + "grad_norm": 2.514910936355591, + "learning_rate": 0.00019122427591017747, + "loss": 0.8734, + "step": 9242 + }, + { + "epoch": 0.6413850530844494, + "grad_norm": 2.093204975128174, + "learning_rate": 0.0001912219741001813, + "loss": 0.7981, + "step": 9243 + }, + { + "epoch": 0.6414544445215461, + "grad_norm": 3.0953497886657715, + "learning_rate": 0.000191219672002208, + "loss": 0.9866, + "step": 9244 + }, + { + "epoch": 0.6415238359586427, + "grad_norm": 2.525607109069824, + "learning_rate": 0.00019121736961626474, + "loss": 0.8712, + "step": 9245 + }, + { + "epoch": 0.6415932273957393, + "grad_norm": 2.8736753463745117, + "learning_rate": 0.00019121506694235883, + "loss": 1.0051, + "step": 9246 + }, + { + "epoch": 0.641662618832836, + "grad_norm": 2.622102737426758, + "learning_rate": 0.00019121276398049753, + "loss": 1.1236, + "step": 9247 + }, + { + "epoch": 0.6417320102699327, + "grad_norm": 2.0066018104553223, + "learning_rate": 0.00019121046073068812, + "loss": 0.7178, + "step": 9248 + }, + { + "epoch": 0.6418014017070294, + "grad_norm": 2.223649740219116, + "learning_rate": 0.00019120815719293784, + "loss": 0.7095, + "step": 9249 + }, + { + "epoch": 0.641870793144126, + "grad_norm": 3.288966417312622, + "learning_rate": 0.00019120585336725404, + "loss": 0.936, + "step": 9250 + }, + { + "epoch": 0.6419401845812227, + "grad_norm": 1.9400554895401, + "learning_rate": 0.00019120354925364393, + "loss": 0.9355, + "step": 9251 + }, + { + "epoch": 0.6420095760183193, + "grad_norm": 2.6360416412353516, + "learning_rate": 0.00019120124485211477, + "loss": 0.8234, + "step": 9252 + }, + { + "epoch": 0.642078967455416, + "grad_norm": 3.101052761077881, + "learning_rate": 0.00019119894016267388, + "loss": 1.0992, + "step": 9253 + }, + { + "epoch": 0.6421483588925126, + "grad_norm": 2.7924466133117676, + "learning_rate": 0.00019119663518532852, + "loss": 0.938, + "step": 9254 + }, + { + "epoch": 0.6422177503296094, + "grad_norm": 2.6265594959259033, + "learning_rate": 0.00019119432992008593, + "loss": 1.2551, + "step": 9255 + }, + { + "epoch": 0.642287141766706, + "grad_norm": 1.6304311752319336, + "learning_rate": 0.00019119202436695346, + "loss": 0.5954, + "step": 9256 + }, + { + "epoch": 0.6423565332038027, + "grad_norm": 2.3901801109313965, + "learning_rate": 0.00019118971852593835, + "loss": 0.7603, + "step": 9257 + }, + { + "epoch": 0.6424259246408993, + "grad_norm": 2.3567302227020264, + "learning_rate": 0.00019118741239704788, + "loss": 0.9961, + "step": 9258 + }, + { + "epoch": 0.6424953160779959, + "grad_norm": 2.309077739715576, + "learning_rate": 0.0001911851059802893, + "loss": 0.914, + "step": 9259 + }, + { + "epoch": 0.6425647075150926, + "grad_norm": 2.0896525382995605, + "learning_rate": 0.00019118279927566993, + "loss": 0.8374, + "step": 9260 + }, + { + "epoch": 0.6426340989521893, + "grad_norm": 3.035463571548462, + "learning_rate": 0.00019118049228319706, + "loss": 1.1381, + "step": 9261 + }, + { + "epoch": 0.642703490389286, + "grad_norm": 2.2652809619903564, + "learning_rate": 0.00019117818500287794, + "loss": 0.961, + "step": 9262 + }, + { + "epoch": 0.6427728818263826, + "grad_norm": 2.501157283782959, + "learning_rate": 0.00019117587743471988, + "loss": 0.9679, + "step": 9263 + }, + { + "epoch": 0.6428422732634793, + "grad_norm": 2.1785292625427246, + "learning_rate": 0.00019117356957873015, + "loss": 0.9069, + "step": 9264 + }, + { + "epoch": 0.6429116647005759, + "grad_norm": 1.7179737091064453, + "learning_rate": 0.00019117126143491602, + "loss": 0.7025, + "step": 9265 + }, + { + "epoch": 0.6429810561376726, + "grad_norm": 3.032750368118286, + "learning_rate": 0.00019116895300328482, + "loss": 1.1691, + "step": 9266 + }, + { + "epoch": 0.6430504475747693, + "grad_norm": 1.975191593170166, + "learning_rate": 0.0001911666442838438, + "loss": 0.6794, + "step": 9267 + }, + { + "epoch": 0.643119839011866, + "grad_norm": 2.776409864425659, + "learning_rate": 0.00019116433527660025, + "loss": 0.62, + "step": 9268 + }, + { + "epoch": 0.6431892304489626, + "grad_norm": 2.666630983352661, + "learning_rate": 0.0001911620259815615, + "loss": 1.0547, + "step": 9269 + }, + { + "epoch": 0.6432586218860593, + "grad_norm": 2.28013014793396, + "learning_rate": 0.00019115971639873479, + "loss": 0.7115, + "step": 9270 + }, + { + "epoch": 0.6433280133231559, + "grad_norm": 2.62007999420166, + "learning_rate": 0.00019115740652812744, + "loss": 0.8977, + "step": 9271 + }, + { + "epoch": 0.6433974047602525, + "grad_norm": 2.687130928039551, + "learning_rate": 0.00019115509636974675, + "loss": 0.7029, + "step": 9272 + }, + { + "epoch": 0.6434667961973493, + "grad_norm": 1.7904548645019531, + "learning_rate": 0.00019115278592359995, + "loss": 0.7266, + "step": 9273 + }, + { + "epoch": 0.6435361876344459, + "grad_norm": 2.58709716796875, + "learning_rate": 0.00019115047518969444, + "loss": 1.0063, + "step": 9274 + }, + { + "epoch": 0.6436055790715426, + "grad_norm": 2.6442410945892334, + "learning_rate": 0.0001911481641680374, + "loss": 0.9925, + "step": 9275 + }, + { + "epoch": 0.6436749705086392, + "grad_norm": 2.660731077194214, + "learning_rate": 0.0001911458528586362, + "loss": 1.0705, + "step": 9276 + }, + { + "epoch": 0.6437443619457359, + "grad_norm": 2.2859721183776855, + "learning_rate": 0.00019114354126149811, + "loss": 1.1056, + "step": 9277 + }, + { + "epoch": 0.6438137533828325, + "grad_norm": 2.1856038570404053, + "learning_rate": 0.00019114122937663046, + "loss": 0.7778, + "step": 9278 + }, + { + "epoch": 0.6438831448199293, + "grad_norm": 3.424314260482788, + "learning_rate": 0.00019113891720404049, + "loss": 0.9965, + "step": 9279 + }, + { + "epoch": 0.6439525362570259, + "grad_norm": 3.8459596633911133, + "learning_rate": 0.00019113660474373556, + "loss": 0.6975, + "step": 9280 + }, + { + "epoch": 0.6440219276941226, + "grad_norm": 2.3935904502868652, + "learning_rate": 0.00019113429199572296, + "loss": 0.7828, + "step": 9281 + }, + { + "epoch": 0.6440913191312192, + "grad_norm": 2.1058826446533203, + "learning_rate": 0.00019113197896000994, + "loss": 0.821, + "step": 9282 + }, + { + "epoch": 0.6441607105683159, + "grad_norm": 3.030477285385132, + "learning_rate": 0.0001911296656366038, + "loss": 1.042, + "step": 9283 + }, + { + "epoch": 0.6442301020054125, + "grad_norm": 3.1237661838531494, + "learning_rate": 0.00019112735202551196, + "loss": 0.7566, + "step": 9284 + }, + { + "epoch": 0.6442994934425091, + "grad_norm": 2.1391029357910156, + "learning_rate": 0.0001911250381267416, + "loss": 0.7069, + "step": 9285 + }, + { + "epoch": 0.6443688848796059, + "grad_norm": 2.5002877712249756, + "learning_rate": 0.00019112272394030007, + "loss": 1.0446, + "step": 9286 + }, + { + "epoch": 0.6444382763167025, + "grad_norm": 2.7602243423461914, + "learning_rate": 0.00019112040946619467, + "loss": 0.8011, + "step": 9287 + }, + { + "epoch": 0.6445076677537992, + "grad_norm": 3.7594189643859863, + "learning_rate": 0.00019111809470443271, + "loss": 1.1875, + "step": 9288 + }, + { + "epoch": 0.6445770591908958, + "grad_norm": 2.0258865356445312, + "learning_rate": 0.0001911157796550215, + "loss": 0.9983, + "step": 9289 + }, + { + "epoch": 0.6446464506279925, + "grad_norm": 1.7107446193695068, + "learning_rate": 0.00019111346431796835, + "loss": 0.6099, + "step": 9290 + }, + { + "epoch": 0.6447158420650891, + "grad_norm": 2.937790632247925, + "learning_rate": 0.00019111114869328058, + "loss": 0.8297, + "step": 9291 + }, + { + "epoch": 0.6447852335021859, + "grad_norm": 2.842172384262085, + "learning_rate": 0.00019110883278096544, + "loss": 0.7444, + "step": 9292 + }, + { + "epoch": 0.6448546249392825, + "grad_norm": 2.609729290008545, + "learning_rate": 0.0001911065165810303, + "loss": 0.7821, + "step": 9293 + }, + { + "epoch": 0.6449240163763792, + "grad_norm": 3.377532482147217, + "learning_rate": 0.0001911042000934825, + "loss": 0.9938, + "step": 9294 + }, + { + "epoch": 0.6449934078134758, + "grad_norm": 2.787601947784424, + "learning_rate": 0.00019110188331832926, + "loss": 0.975, + "step": 9295 + }, + { + "epoch": 0.6450627992505725, + "grad_norm": 3.29823637008667, + "learning_rate": 0.00019109956625557796, + "loss": 0.8093, + "step": 9296 + }, + { + "epoch": 0.6451321906876691, + "grad_norm": 2.5709550380706787, + "learning_rate": 0.0001910972489052359, + "loss": 0.9163, + "step": 9297 + }, + { + "epoch": 0.6452015821247657, + "grad_norm": 2.780564308166504, + "learning_rate": 0.0001910949312673104, + "loss": 1.1193, + "step": 9298 + }, + { + "epoch": 0.6452709735618625, + "grad_norm": 2.652294874191284, + "learning_rate": 0.00019109261334180876, + "loss": 0.7787, + "step": 9299 + }, + { + "epoch": 0.6453403649989591, + "grad_norm": 3.591848850250244, + "learning_rate": 0.00019109029512873833, + "loss": 1.1898, + "step": 9300 + }, + { + "epoch": 0.6454097564360558, + "grad_norm": 2.398667097091675, + "learning_rate": 0.00019108797662810637, + "loss": 0.9435, + "step": 9301 + }, + { + "epoch": 0.6454791478731524, + "grad_norm": 2.622014284133911, + "learning_rate": 0.00019108565783992028, + "loss": 1.0618, + "step": 9302 + }, + { + "epoch": 0.6455485393102491, + "grad_norm": 1.9249261617660522, + "learning_rate": 0.00019108333876418732, + "loss": 0.9374, + "step": 9303 + }, + { + "epoch": 0.6456179307473457, + "grad_norm": 3.1512818336486816, + "learning_rate": 0.0001910810194009148, + "loss": 1.0065, + "step": 9304 + }, + { + "epoch": 0.6456873221844425, + "grad_norm": 3.0980312824249268, + "learning_rate": 0.00019107869975011009, + "loss": 0.9982, + "step": 9305 + }, + { + "epoch": 0.6457567136215391, + "grad_norm": 2.408109426498413, + "learning_rate": 0.0001910763798117805, + "loss": 0.9849, + "step": 9306 + }, + { + "epoch": 0.6458261050586358, + "grad_norm": 2.651642084121704, + "learning_rate": 0.0001910740595859333, + "loss": 0.9745, + "step": 9307 + }, + { + "epoch": 0.6458954964957324, + "grad_norm": 2.1364638805389404, + "learning_rate": 0.00019107173907257594, + "loss": 0.8964, + "step": 9308 + }, + { + "epoch": 0.6459648879328291, + "grad_norm": 2.735915422439575, + "learning_rate": 0.0001910694182717156, + "loss": 0.846, + "step": 9309 + }, + { + "epoch": 0.6460342793699257, + "grad_norm": 2.610200881958008, + "learning_rate": 0.0001910670971833597, + "loss": 1.1921, + "step": 9310 + }, + { + "epoch": 0.6461036708070225, + "grad_norm": 2.8411128520965576, + "learning_rate": 0.00019106477580751555, + "loss": 1.1438, + "step": 9311 + }, + { + "epoch": 0.6461730622441191, + "grad_norm": 2.021474599838257, + "learning_rate": 0.00019106245414419044, + "loss": 0.8537, + "step": 9312 + }, + { + "epoch": 0.6462424536812157, + "grad_norm": 2.1151108741760254, + "learning_rate": 0.00019106013219339176, + "loss": 0.801, + "step": 9313 + }, + { + "epoch": 0.6463118451183124, + "grad_norm": 3.16910457611084, + "learning_rate": 0.0001910578099551268, + "loss": 0.8764, + "step": 9314 + }, + { + "epoch": 0.646381236555409, + "grad_norm": 3.157947540283203, + "learning_rate": 0.00019105548742940286, + "loss": 1.1485, + "step": 9315 + }, + { + "epoch": 0.6464506279925057, + "grad_norm": 2.9120521545410156, + "learning_rate": 0.00019105316461622734, + "loss": 1.0234, + "step": 9316 + }, + { + "epoch": 0.6465200194296024, + "grad_norm": 2.807812452316284, + "learning_rate": 0.00019105084151560756, + "loss": 1.0312, + "step": 9317 + }, + { + "epoch": 0.6465894108666991, + "grad_norm": 2.260559320449829, + "learning_rate": 0.0001910485181275508, + "loss": 0.729, + "step": 9318 + }, + { + "epoch": 0.6466588023037957, + "grad_norm": 2.670212507247925, + "learning_rate": 0.00019104619445206448, + "loss": 0.9917, + "step": 9319 + }, + { + "epoch": 0.6467281937408924, + "grad_norm": 2.6991817951202393, + "learning_rate": 0.0001910438704891558, + "loss": 0.9892, + "step": 9320 + }, + { + "epoch": 0.646797585177989, + "grad_norm": 2.2979705333709717, + "learning_rate": 0.00019104154623883227, + "loss": 0.6991, + "step": 9321 + }, + { + "epoch": 0.6468669766150857, + "grad_norm": 2.432030200958252, + "learning_rate": 0.0001910392217011011, + "loss": 0.7851, + "step": 9322 + }, + { + "epoch": 0.6469363680521824, + "grad_norm": 2.324154853820801, + "learning_rate": 0.0001910368968759697, + "loss": 1.0245, + "step": 9323 + }, + { + "epoch": 0.6470057594892791, + "grad_norm": 3.0859739780426025, + "learning_rate": 0.00019103457176344537, + "loss": 1.0014, + "step": 9324 + }, + { + "epoch": 0.6470751509263757, + "grad_norm": 3.0454673767089844, + "learning_rate": 0.00019103224636353544, + "loss": 0.9891, + "step": 9325 + }, + { + "epoch": 0.6471445423634723, + "grad_norm": 1.8514659404754639, + "learning_rate": 0.00019102992067624727, + "loss": 0.7391, + "step": 9326 + }, + { + "epoch": 0.647213933800569, + "grad_norm": 1.971319317817688, + "learning_rate": 0.0001910275947015882, + "loss": 0.7683, + "step": 9327 + }, + { + "epoch": 0.6472833252376656, + "grad_norm": 2.642620086669922, + "learning_rate": 0.0001910252684395656, + "loss": 1.137, + "step": 9328 + }, + { + "epoch": 0.6473527166747624, + "grad_norm": 2.4878995418548584, + "learning_rate": 0.00019102294189018678, + "loss": 0.6912, + "step": 9329 + }, + { + "epoch": 0.647422108111859, + "grad_norm": 1.995025873184204, + "learning_rate": 0.0001910206150534591, + "loss": 0.7658, + "step": 9330 + }, + { + "epoch": 0.6474914995489557, + "grad_norm": 2.805609703063965, + "learning_rate": 0.0001910182879293899, + "loss": 1.0417, + "step": 9331 + }, + { + "epoch": 0.6475608909860523, + "grad_norm": 2.9813196659088135, + "learning_rate": 0.0001910159605179865, + "loss": 0.9013, + "step": 9332 + }, + { + "epoch": 0.647630282423149, + "grad_norm": 2.86236572265625, + "learning_rate": 0.00019101363281925627, + "loss": 0.8408, + "step": 9333 + }, + { + "epoch": 0.6476996738602456, + "grad_norm": 2.329054832458496, + "learning_rate": 0.0001910113048332066, + "loss": 0.953, + "step": 9334 + }, + { + "epoch": 0.6477690652973424, + "grad_norm": 2.3969266414642334, + "learning_rate": 0.00019100897655984476, + "loss": 0.9136, + "step": 9335 + }, + { + "epoch": 0.647838456734439, + "grad_norm": 2.2183687686920166, + "learning_rate": 0.00019100664799917815, + "loss": 0.9592, + "step": 9336 + }, + { + "epoch": 0.6479078481715357, + "grad_norm": 1.977657437324524, + "learning_rate": 0.00019100431915121413, + "loss": 0.9481, + "step": 9337 + }, + { + "epoch": 0.6479772396086323, + "grad_norm": 2.6835598945617676, + "learning_rate": 0.00019100199001596, + "loss": 0.8218, + "step": 9338 + }, + { + "epoch": 0.6480466310457289, + "grad_norm": 1.7699354887008667, + "learning_rate": 0.0001909996605934232, + "loss": 0.6839, + "step": 9339 + }, + { + "epoch": 0.6481160224828256, + "grad_norm": 1.9457374811172485, + "learning_rate": 0.00019099733088361097, + "loss": 0.7043, + "step": 9340 + }, + { + "epoch": 0.6481854139199222, + "grad_norm": 2.0299651622772217, + "learning_rate": 0.00019099500088653074, + "loss": 0.7998, + "step": 9341 + }, + { + "epoch": 0.648254805357019, + "grad_norm": 2.222275495529175, + "learning_rate": 0.00019099267060218987, + "loss": 0.894, + "step": 9342 + }, + { + "epoch": 0.6483241967941156, + "grad_norm": 2.008970260620117, + "learning_rate": 0.00019099034003059568, + "loss": 0.9654, + "step": 9343 + }, + { + "epoch": 0.6483935882312123, + "grad_norm": 2.213747978210449, + "learning_rate": 0.00019098800917175557, + "loss": 0.8823, + "step": 9344 + }, + { + "epoch": 0.6484629796683089, + "grad_norm": 2.1916303634643555, + "learning_rate": 0.00019098567802567682, + "loss": 0.7246, + "step": 9345 + }, + { + "epoch": 0.6485323711054056, + "grad_norm": 2.509415864944458, + "learning_rate": 0.0001909833465923669, + "loss": 0.6854, + "step": 9346 + }, + { + "epoch": 0.6486017625425022, + "grad_norm": 1.8394144773483276, + "learning_rate": 0.00019098101487183307, + "loss": 0.6962, + "step": 9347 + }, + { + "epoch": 0.648671153979599, + "grad_norm": 2.801530599594116, + "learning_rate": 0.00019097868286408273, + "loss": 0.922, + "step": 9348 + }, + { + "epoch": 0.6487405454166956, + "grad_norm": 2.4538543224334717, + "learning_rate": 0.00019097635056912327, + "loss": 0.9804, + "step": 9349 + }, + { + "epoch": 0.6488099368537923, + "grad_norm": 2.8569235801696777, + "learning_rate": 0.000190974017986962, + "loss": 1.1359, + "step": 9350 + }, + { + "epoch": 0.6488793282908889, + "grad_norm": 2.763643503189087, + "learning_rate": 0.00019097168511760632, + "loss": 0.9548, + "step": 9351 + }, + { + "epoch": 0.6489487197279855, + "grad_norm": 2.2157273292541504, + "learning_rate": 0.00019096935196106357, + "loss": 0.6932, + "step": 9352 + }, + { + "epoch": 0.6490181111650822, + "grad_norm": 2.6398794651031494, + "learning_rate": 0.00019096701851734115, + "loss": 0.7933, + "step": 9353 + }, + { + "epoch": 0.6490875026021788, + "grad_norm": 2.484740972518921, + "learning_rate": 0.00019096468478644642, + "loss": 0.8128, + "step": 9354 + }, + { + "epoch": 0.6491568940392756, + "grad_norm": 3.664952039718628, + "learning_rate": 0.00019096235076838673, + "loss": 0.7767, + "step": 9355 + }, + { + "epoch": 0.6492262854763722, + "grad_norm": 2.953442335128784, + "learning_rate": 0.00019096001646316942, + "loss": 0.7748, + "step": 9356 + }, + { + "epoch": 0.6492956769134689, + "grad_norm": 3.60091233253479, + "learning_rate": 0.0001909576818708019, + "loss": 0.7892, + "step": 9357 + }, + { + "epoch": 0.6493650683505655, + "grad_norm": 2.955406904220581, + "learning_rate": 0.00019095534699129155, + "loss": 1.1644, + "step": 9358 + }, + { + "epoch": 0.6494344597876622, + "grad_norm": 2.177326202392578, + "learning_rate": 0.00019095301182464573, + "loss": 0.8098, + "step": 9359 + }, + { + "epoch": 0.6495038512247588, + "grad_norm": 1.9081826210021973, + "learning_rate": 0.0001909506763708718, + "loss": 1.0746, + "step": 9360 + }, + { + "epoch": 0.6495732426618556, + "grad_norm": 1.966964602470398, + "learning_rate": 0.00019094834062997713, + "loss": 0.7304, + "step": 9361 + }, + { + "epoch": 0.6496426340989522, + "grad_norm": 2.200354814529419, + "learning_rate": 0.00019094600460196909, + "loss": 0.7014, + "step": 9362 + }, + { + "epoch": 0.6497120255360489, + "grad_norm": 2.664357900619507, + "learning_rate": 0.00019094366828685506, + "loss": 0.9583, + "step": 9363 + }, + { + "epoch": 0.6497814169731455, + "grad_norm": 2.596498727798462, + "learning_rate": 0.00019094133168464245, + "loss": 0.6859, + "step": 9364 + }, + { + "epoch": 0.6498508084102422, + "grad_norm": 3.375818967819214, + "learning_rate": 0.0001909389947953386, + "loss": 0.8582, + "step": 9365 + }, + { + "epoch": 0.6499201998473388, + "grad_norm": 2.1771366596221924, + "learning_rate": 0.00019093665761895086, + "loss": 0.771, + "step": 9366 + }, + { + "epoch": 0.6499895912844355, + "grad_norm": 2.245802164077759, + "learning_rate": 0.0001909343201554867, + "loss": 0.7864, + "step": 9367 + }, + { + "epoch": 0.6500589827215322, + "grad_norm": 1.8183306455612183, + "learning_rate": 0.0001909319824049534, + "loss": 0.7184, + "step": 9368 + }, + { + "epoch": 0.6501283741586288, + "grad_norm": 2.913045883178711, + "learning_rate": 0.00019092964436735835, + "loss": 1.0667, + "step": 9369 + }, + { + "epoch": 0.6501977655957255, + "grad_norm": 2.5670838356018066, + "learning_rate": 0.00019092730604270902, + "loss": 1.0499, + "step": 9370 + }, + { + "epoch": 0.6502671570328221, + "grad_norm": 2.3500616550445557, + "learning_rate": 0.00019092496743101274, + "loss": 0.8958, + "step": 9371 + }, + { + "epoch": 0.6503365484699188, + "grad_norm": 2.41064715385437, + "learning_rate": 0.00019092262853227684, + "loss": 0.8208, + "step": 9372 + }, + { + "epoch": 0.6504059399070155, + "grad_norm": 2.4508159160614014, + "learning_rate": 0.0001909202893465088, + "loss": 0.8707, + "step": 9373 + }, + { + "epoch": 0.6504753313441122, + "grad_norm": 2.6485722064971924, + "learning_rate": 0.00019091794987371593, + "loss": 1.1091, + "step": 9374 + }, + { + "epoch": 0.6505447227812088, + "grad_norm": 2.010009765625, + "learning_rate": 0.00019091561011390563, + "loss": 0.6243, + "step": 9375 + }, + { + "epoch": 0.6506141142183055, + "grad_norm": 2.303386926651001, + "learning_rate": 0.0001909132700670853, + "loss": 0.9193, + "step": 9376 + }, + { + "epoch": 0.6506835056554021, + "grad_norm": 2.3863279819488525, + "learning_rate": 0.00019091092973326236, + "loss": 1.0428, + "step": 9377 + }, + { + "epoch": 0.6507528970924988, + "grad_norm": 2.384660243988037, + "learning_rate": 0.00019090858911244413, + "loss": 0.9146, + "step": 9378 + }, + { + "epoch": 0.6508222885295954, + "grad_norm": 2.4521749019622803, + "learning_rate": 0.00019090624820463804, + "loss": 0.9485, + "step": 9379 + }, + { + "epoch": 0.6508916799666921, + "grad_norm": 2.0419275760650635, + "learning_rate": 0.00019090390700985145, + "loss": 0.6443, + "step": 9380 + }, + { + "epoch": 0.6509610714037888, + "grad_norm": 2.381226062774658, + "learning_rate": 0.00019090156552809177, + "loss": 1.0915, + "step": 9381 + }, + { + "epoch": 0.6510304628408854, + "grad_norm": 2.9374873638153076, + "learning_rate": 0.00019089922375936644, + "loss": 0.8699, + "step": 9382 + }, + { + "epoch": 0.6510998542779821, + "grad_norm": 1.8320449590682983, + "learning_rate": 0.00019089688170368277, + "loss": 0.8097, + "step": 9383 + }, + { + "epoch": 0.6511692457150787, + "grad_norm": 2.510160207748413, + "learning_rate": 0.0001908945393610482, + "loss": 0.9392, + "step": 9384 + }, + { + "epoch": 0.6512386371521754, + "grad_norm": 1.9889675378799438, + "learning_rate": 0.00019089219673147013, + "loss": 0.6802, + "step": 9385 + }, + { + "epoch": 0.6513080285892721, + "grad_norm": 2.4766080379486084, + "learning_rate": 0.0001908898538149559, + "loss": 1.0982, + "step": 9386 + }, + { + "epoch": 0.6513774200263688, + "grad_norm": 2.511965751647949, + "learning_rate": 0.00019088751061151297, + "loss": 1.1455, + "step": 9387 + }, + { + "epoch": 0.6514468114634654, + "grad_norm": 1.9817183017730713, + "learning_rate": 0.00019088516712114872, + "loss": 0.6923, + "step": 9388 + }, + { + "epoch": 0.6515162029005621, + "grad_norm": 3.6322710514068604, + "learning_rate": 0.00019088282334387056, + "loss": 0.789, + "step": 9389 + }, + { + "epoch": 0.6515855943376587, + "grad_norm": 1.9416948556900024, + "learning_rate": 0.00019088047927968583, + "loss": 0.8306, + "step": 9390 + }, + { + "epoch": 0.6516549857747554, + "grad_norm": 2.6383492946624756, + "learning_rate": 0.000190878134928602, + "loss": 0.8236, + "step": 9391 + }, + { + "epoch": 0.6517243772118521, + "grad_norm": 2.0380725860595703, + "learning_rate": 0.00019087579029062644, + "loss": 0.8414, + "step": 9392 + }, + { + "epoch": 0.6517937686489487, + "grad_norm": 2.6294567584991455, + "learning_rate": 0.00019087344536576655, + "loss": 1.1584, + "step": 9393 + }, + { + "epoch": 0.6518631600860454, + "grad_norm": 2.338181257247925, + "learning_rate": 0.00019087110015402972, + "loss": 0.7391, + "step": 9394 + }, + { + "epoch": 0.651932551523142, + "grad_norm": 3.4072463512420654, + "learning_rate": 0.0001908687546554234, + "loss": 0.8396, + "step": 9395 + }, + { + "epoch": 0.6520019429602387, + "grad_norm": 1.9483951330184937, + "learning_rate": 0.00019086640886995494, + "loss": 0.7202, + "step": 9396 + }, + { + "epoch": 0.6520713343973353, + "grad_norm": 2.984365463256836, + "learning_rate": 0.0001908640627976318, + "loss": 0.8544, + "step": 9397 + }, + { + "epoch": 0.6521407258344321, + "grad_norm": 2.2329177856445312, + "learning_rate": 0.0001908617164384613, + "loss": 0.9027, + "step": 9398 + }, + { + "epoch": 0.6522101172715287, + "grad_norm": 2.4242615699768066, + "learning_rate": 0.00019085936979245096, + "loss": 1.1252, + "step": 9399 + }, + { + "epoch": 0.6522795087086254, + "grad_norm": 2.678583860397339, + "learning_rate": 0.0001908570228596081, + "loss": 0.9751, + "step": 9400 + }, + { + "epoch": 0.652348900145722, + "grad_norm": 2.2193291187286377, + "learning_rate": 0.00019085467563994017, + "loss": 0.6586, + "step": 9401 + }, + { + "epoch": 0.6524182915828187, + "grad_norm": 2.5800538063049316, + "learning_rate": 0.00019085232813345455, + "loss": 0.8674, + "step": 9402 + }, + { + "epoch": 0.6524876830199153, + "grad_norm": 2.5189290046691895, + "learning_rate": 0.0001908499803401587, + "loss": 0.864, + "step": 9403 + }, + { + "epoch": 0.652557074457012, + "grad_norm": 2.4906439781188965, + "learning_rate": 0.00019084763226006, + "loss": 1.1491, + "step": 9404 + }, + { + "epoch": 0.6526264658941087, + "grad_norm": 3.386023759841919, + "learning_rate": 0.00019084528389316585, + "loss": 0.8097, + "step": 9405 + }, + { + "epoch": 0.6526958573312053, + "grad_norm": 2.7408721446990967, + "learning_rate": 0.0001908429352394837, + "loss": 0.8699, + "step": 9406 + }, + { + "epoch": 0.652765248768302, + "grad_norm": 2.457174301147461, + "learning_rate": 0.0001908405862990209, + "loss": 0.852, + "step": 9407 + }, + { + "epoch": 0.6528346402053986, + "grad_norm": 1.7368212938308716, + "learning_rate": 0.00019083823707178495, + "loss": 0.6614, + "step": 9408 + }, + { + "epoch": 0.6529040316424953, + "grad_norm": 3.5847256183624268, + "learning_rate": 0.0001908358875577832, + "loss": 0.8743, + "step": 9409 + }, + { + "epoch": 0.6529734230795919, + "grad_norm": 2.2247018814086914, + "learning_rate": 0.0001908335377570231, + "loss": 0.6585, + "step": 9410 + }, + { + "epoch": 0.6530428145166887, + "grad_norm": 1.9684536457061768, + "learning_rate": 0.00019083118766951207, + "loss": 0.619, + "step": 9411 + }, + { + "epoch": 0.6531122059537853, + "grad_norm": 2.0893757343292236, + "learning_rate": 0.00019082883729525752, + "loss": 0.8856, + "step": 9412 + }, + { + "epoch": 0.653181597390882, + "grad_norm": 2.1204164028167725, + "learning_rate": 0.00019082648663426685, + "loss": 0.8357, + "step": 9413 + }, + { + "epoch": 0.6532509888279786, + "grad_norm": 2.2148337364196777, + "learning_rate": 0.0001908241356865475, + "loss": 0.8623, + "step": 9414 + }, + { + "epoch": 0.6533203802650753, + "grad_norm": 2.4018619060516357, + "learning_rate": 0.00019082178445210693, + "loss": 0.8122, + "step": 9415 + }, + { + "epoch": 0.6533897717021719, + "grad_norm": 2.7435402870178223, + "learning_rate": 0.00019081943293095248, + "loss": 0.864, + "step": 9416 + }, + { + "epoch": 0.6534591631392687, + "grad_norm": 2.5501248836517334, + "learning_rate": 0.00019081708112309164, + "loss": 1.1015, + "step": 9417 + }, + { + "epoch": 0.6535285545763653, + "grad_norm": 2.625990867614746, + "learning_rate": 0.0001908147290285318, + "loss": 1.1573, + "step": 9418 + }, + { + "epoch": 0.6535979460134619, + "grad_norm": 2.120476007461548, + "learning_rate": 0.0001908123766472804, + "loss": 0.587, + "step": 9419 + }, + { + "epoch": 0.6536673374505586, + "grad_norm": 2.7818520069122314, + "learning_rate": 0.00019081002397934487, + "loss": 0.8985, + "step": 9420 + }, + { + "epoch": 0.6537367288876552, + "grad_norm": 1.9466198682785034, + "learning_rate": 0.00019080767102473264, + "loss": 0.8218, + "step": 9421 + }, + { + "epoch": 0.6538061203247519, + "grad_norm": 3.015498638153076, + "learning_rate": 0.0001908053177834511, + "loss": 0.8091, + "step": 9422 + }, + { + "epoch": 0.6538755117618485, + "grad_norm": 3.412763833999634, + "learning_rate": 0.00019080296425550772, + "loss": 0.9408, + "step": 9423 + }, + { + "epoch": 0.6539449031989453, + "grad_norm": 2.4277701377868652, + "learning_rate": 0.0001908006104409099, + "loss": 0.8948, + "step": 9424 + }, + { + "epoch": 0.6540142946360419, + "grad_norm": 2.0623929500579834, + "learning_rate": 0.00019079825633966512, + "loss": 0.7469, + "step": 9425 + }, + { + "epoch": 0.6540836860731386, + "grad_norm": 1.8483128547668457, + "learning_rate": 0.00019079590195178074, + "loss": 0.7089, + "step": 9426 + }, + { + "epoch": 0.6541530775102352, + "grad_norm": 2.872204065322876, + "learning_rate": 0.00019079354727726425, + "loss": 0.746, + "step": 9427 + }, + { + "epoch": 0.6542224689473319, + "grad_norm": 2.438192129135132, + "learning_rate": 0.0001907911923161231, + "loss": 0.9442, + "step": 9428 + }, + { + "epoch": 0.6542918603844285, + "grad_norm": 2.190852642059326, + "learning_rate": 0.00019078883706836463, + "loss": 0.6996, + "step": 9429 + }, + { + "epoch": 0.6543612518215253, + "grad_norm": 2.2468507289886475, + "learning_rate": 0.00019078648153399638, + "loss": 0.7496, + "step": 9430 + }, + { + "epoch": 0.6544306432586219, + "grad_norm": 2.540320634841919, + "learning_rate": 0.00019078412571302573, + "loss": 0.947, + "step": 9431 + }, + { + "epoch": 0.6545000346957186, + "grad_norm": 2.8889224529266357, + "learning_rate": 0.00019078176960546012, + "loss": 1.0305, + "step": 9432 + }, + { + "epoch": 0.6545694261328152, + "grad_norm": 2.3298230171203613, + "learning_rate": 0.000190779413211307, + "loss": 0.9618, + "step": 9433 + }, + { + "epoch": 0.6546388175699118, + "grad_norm": 2.236396551132202, + "learning_rate": 0.0001907770565305738, + "loss": 0.9643, + "step": 9434 + }, + { + "epoch": 0.6547082090070085, + "grad_norm": 2.535813093185425, + "learning_rate": 0.000190774699563268, + "loss": 0.7963, + "step": 9435 + }, + { + "epoch": 0.6547776004441052, + "grad_norm": 3.410832166671753, + "learning_rate": 0.00019077234230939697, + "loss": 0.8403, + "step": 9436 + }, + { + "epoch": 0.6548469918812019, + "grad_norm": 2.3897852897644043, + "learning_rate": 0.00019076998476896817, + "loss": 0.9228, + "step": 9437 + }, + { + "epoch": 0.6549163833182985, + "grad_norm": 2.2086737155914307, + "learning_rate": 0.0001907676269419891, + "loss": 1.013, + "step": 9438 + }, + { + "epoch": 0.6549857747553952, + "grad_norm": 2.900676727294922, + "learning_rate": 0.00019076526882846713, + "loss": 1.1014, + "step": 9439 + }, + { + "epoch": 0.6550551661924918, + "grad_norm": 2.4956111907958984, + "learning_rate": 0.00019076291042840975, + "loss": 0.729, + "step": 9440 + }, + { + "epoch": 0.6551245576295885, + "grad_norm": 2.3147056102752686, + "learning_rate": 0.0001907605517418244, + "loss": 0.7639, + "step": 9441 + }, + { + "epoch": 0.6551939490666852, + "grad_norm": 2.384716033935547, + "learning_rate": 0.0001907581927687185, + "loss": 0.8787, + "step": 9442 + }, + { + "epoch": 0.6552633405037819, + "grad_norm": 2.0104987621307373, + "learning_rate": 0.00019075583350909955, + "loss": 1.0283, + "step": 9443 + }, + { + "epoch": 0.6553327319408785, + "grad_norm": 2.662475824356079, + "learning_rate": 0.00019075347396297493, + "loss": 1.1307, + "step": 9444 + }, + { + "epoch": 0.6554021233779752, + "grad_norm": 1.7809044122695923, + "learning_rate": 0.00019075111413035213, + "loss": 0.6852, + "step": 9445 + }, + { + "epoch": 0.6554715148150718, + "grad_norm": 2.339371919631958, + "learning_rate": 0.0001907487540112386, + "loss": 0.8233, + "step": 9446 + }, + { + "epoch": 0.6555409062521684, + "grad_norm": 2.1360316276550293, + "learning_rate": 0.00019074639360564178, + "loss": 0.7513, + "step": 9447 + }, + { + "epoch": 0.6556102976892652, + "grad_norm": 1.8391474485397339, + "learning_rate": 0.0001907440329135691, + "loss": 0.8358, + "step": 9448 + }, + { + "epoch": 0.6556796891263618, + "grad_norm": 4.753307342529297, + "learning_rate": 0.00019074167193502807, + "loss": 1.1191, + "step": 9449 + }, + { + "epoch": 0.6557490805634585, + "grad_norm": 2.3001577854156494, + "learning_rate": 0.00019073931067002607, + "loss": 0.6958, + "step": 9450 + }, + { + "epoch": 0.6558184720005551, + "grad_norm": 2.148670196533203, + "learning_rate": 0.00019073694911857065, + "loss": 0.8286, + "step": 9451 + }, + { + "epoch": 0.6558878634376518, + "grad_norm": 2.6264193058013916, + "learning_rate": 0.00019073458728066918, + "loss": 0.9184, + "step": 9452 + }, + { + "epoch": 0.6559572548747484, + "grad_norm": 2.3174095153808594, + "learning_rate": 0.00019073222515632914, + "loss": 0.8158, + "step": 9453 + }, + { + "epoch": 0.6560266463118452, + "grad_norm": 2.720118284225464, + "learning_rate": 0.000190729862745558, + "loss": 1.1771, + "step": 9454 + }, + { + "epoch": 0.6560960377489418, + "grad_norm": 2.816225051879883, + "learning_rate": 0.0001907275000483632, + "loss": 1.0606, + "step": 9455 + }, + { + "epoch": 0.6561654291860385, + "grad_norm": 2.5619499683380127, + "learning_rate": 0.00019072513706475222, + "loss": 0.6632, + "step": 9456 + }, + { + "epoch": 0.6562348206231351, + "grad_norm": 2.1378531455993652, + "learning_rate": 0.0001907227737947325, + "loss": 0.7681, + "step": 9457 + }, + { + "epoch": 0.6563042120602318, + "grad_norm": 2.1311895847320557, + "learning_rate": 0.00019072041023831152, + "loss": 0.7186, + "step": 9458 + }, + { + "epoch": 0.6563736034973284, + "grad_norm": 2.2615280151367188, + "learning_rate": 0.0001907180463954967, + "loss": 0.8509, + "step": 9459 + }, + { + "epoch": 0.656442994934425, + "grad_norm": 3.2346255779266357, + "learning_rate": 0.00019071568226629556, + "loss": 0.9787, + "step": 9460 + }, + { + "epoch": 0.6565123863715218, + "grad_norm": 2.6999950408935547, + "learning_rate": 0.00019071331785071554, + "loss": 0.8577, + "step": 9461 + }, + { + "epoch": 0.6565817778086184, + "grad_norm": 3.0546228885650635, + "learning_rate": 0.0001907109531487641, + "loss": 1.0765, + "step": 9462 + }, + { + "epoch": 0.6566511692457151, + "grad_norm": 2.3635413646698, + "learning_rate": 0.00019070858816044867, + "loss": 0.6932, + "step": 9463 + }, + { + "epoch": 0.6567205606828117, + "grad_norm": 2.0657081604003906, + "learning_rate": 0.00019070622288577677, + "loss": 0.8204, + "step": 9464 + }, + { + "epoch": 0.6567899521199084, + "grad_norm": 1.859310507774353, + "learning_rate": 0.00019070385732475589, + "loss": 0.6549, + "step": 9465 + }, + { + "epoch": 0.656859343557005, + "grad_norm": 1.9759560823440552, + "learning_rate": 0.0001907014914773934, + "loss": 0.8088, + "step": 9466 + }, + { + "epoch": 0.6569287349941018, + "grad_norm": 2.384885311126709, + "learning_rate": 0.00019069912534369686, + "loss": 0.8294, + "step": 9467 + }, + { + "epoch": 0.6569981264311984, + "grad_norm": 1.9955073595046997, + "learning_rate": 0.00019069675892367368, + "loss": 0.8651, + "step": 9468 + }, + { + "epoch": 0.6570675178682951, + "grad_norm": 3.882526159286499, + "learning_rate": 0.00019069439221733138, + "loss": 0.8903, + "step": 9469 + }, + { + "epoch": 0.6571369093053917, + "grad_norm": 2.433490753173828, + "learning_rate": 0.0001906920252246774, + "loss": 0.8568, + "step": 9470 + }, + { + "epoch": 0.6572063007424884, + "grad_norm": 2.212015390396118, + "learning_rate": 0.0001906896579457192, + "loss": 0.7805, + "step": 9471 + }, + { + "epoch": 0.657275692179585, + "grad_norm": 1.8215584754943848, + "learning_rate": 0.0001906872903804643, + "loss": 0.6662, + "step": 9472 + }, + { + "epoch": 0.6573450836166816, + "grad_norm": 2.1134610176086426, + "learning_rate": 0.0001906849225289201, + "loss": 0.8839, + "step": 9473 + }, + { + "epoch": 0.6574144750537784, + "grad_norm": 2.7124335765838623, + "learning_rate": 0.00019068255439109416, + "loss": 1.0128, + "step": 9474 + }, + { + "epoch": 0.657483866490875, + "grad_norm": 3.4437358379364014, + "learning_rate": 0.0001906801859669939, + "loss": 0.8209, + "step": 9475 + }, + { + "epoch": 0.6575532579279717, + "grad_norm": 2.630807876586914, + "learning_rate": 0.00019067781725662684, + "loss": 0.76, + "step": 9476 + }, + { + "epoch": 0.6576226493650683, + "grad_norm": 2.1341052055358887, + "learning_rate": 0.0001906754482600004, + "loss": 0.7721, + "step": 9477 + }, + { + "epoch": 0.657692040802165, + "grad_norm": 2.7487666606903076, + "learning_rate": 0.0001906730789771221, + "loss": 0.9936, + "step": 9478 + }, + { + "epoch": 0.6577614322392616, + "grad_norm": 2.230494499206543, + "learning_rate": 0.0001906707094079994, + "loss": 0.878, + "step": 9479 + }, + { + "epoch": 0.6578308236763584, + "grad_norm": 1.9608887434005737, + "learning_rate": 0.00019066833955263981, + "loss": 0.8114, + "step": 9480 + }, + { + "epoch": 0.657900215113455, + "grad_norm": 2.129179000854492, + "learning_rate": 0.0001906659694110508, + "loss": 0.7556, + "step": 9481 + }, + { + "epoch": 0.6579696065505517, + "grad_norm": 2.454969644546509, + "learning_rate": 0.0001906635989832398, + "loss": 0.8827, + "step": 9482 + }, + { + "epoch": 0.6580389979876483, + "grad_norm": 2.6127800941467285, + "learning_rate": 0.00019066122826921439, + "loss": 1.0177, + "step": 9483 + }, + { + "epoch": 0.658108389424745, + "grad_norm": 2.457404613494873, + "learning_rate": 0.00019065885726898197, + "loss": 0.8211, + "step": 9484 + }, + { + "epoch": 0.6581777808618416, + "grad_norm": 4.35467004776001, + "learning_rate": 0.00019065648598255006, + "loss": 0.8574, + "step": 9485 + }, + { + "epoch": 0.6582471722989384, + "grad_norm": 2.1218056678771973, + "learning_rate": 0.00019065411440992612, + "loss": 0.576, + "step": 9486 + }, + { + "epoch": 0.658316563736035, + "grad_norm": 3.1830289363861084, + "learning_rate": 0.0001906517425511177, + "loss": 0.8488, + "step": 9487 + }, + { + "epoch": 0.6583859551731316, + "grad_norm": 2.7939045429229736, + "learning_rate": 0.00019064937040613224, + "loss": 0.8842, + "step": 9488 + }, + { + "epoch": 0.6584553466102283, + "grad_norm": 3.7226054668426514, + "learning_rate": 0.0001906469979749772, + "loss": 1.0939, + "step": 9489 + }, + { + "epoch": 0.6585247380473249, + "grad_norm": 2.486093044281006, + "learning_rate": 0.00019064462525766014, + "loss": 0.8495, + "step": 9490 + }, + { + "epoch": 0.6585941294844216, + "grad_norm": 2.529953718185425, + "learning_rate": 0.0001906422522541885, + "loss": 0.5743, + "step": 9491 + }, + { + "epoch": 0.6586635209215183, + "grad_norm": 2.2134063243865967, + "learning_rate": 0.00019063987896456982, + "loss": 0.7628, + "step": 9492 + }, + { + "epoch": 0.658732912358615, + "grad_norm": 2.6042845249176025, + "learning_rate": 0.00019063750538881151, + "loss": 0.8355, + "step": 9493 + }, + { + "epoch": 0.6588023037957116, + "grad_norm": 2.4359374046325684, + "learning_rate": 0.00019063513152692112, + "loss": 0.8202, + "step": 9494 + }, + { + "epoch": 0.6588716952328083, + "grad_norm": 2.7207705974578857, + "learning_rate": 0.00019063275737890613, + "loss": 0.8632, + "step": 9495 + }, + { + "epoch": 0.6589410866699049, + "grad_norm": 2.548896312713623, + "learning_rate": 0.00019063038294477407, + "loss": 0.8079, + "step": 9496 + }, + { + "epoch": 0.6590104781070016, + "grad_norm": 2.0960888862609863, + "learning_rate": 0.00019062800822453236, + "loss": 0.8571, + "step": 9497 + }, + { + "epoch": 0.6590798695440983, + "grad_norm": 2.713881731033325, + "learning_rate": 0.00019062563321818862, + "loss": 0.9275, + "step": 9498 + }, + { + "epoch": 0.659149260981195, + "grad_norm": 2.484600067138672, + "learning_rate": 0.0001906232579257502, + "loss": 0.8784, + "step": 9499 + }, + { + "epoch": 0.6592186524182916, + "grad_norm": 3.6943838596343994, + "learning_rate": 0.00019062088234722468, + "loss": 0.8566, + "step": 9500 + }, + { + "epoch": 0.6592880438553882, + "grad_norm": 3.740257501602173, + "learning_rate": 0.00019061850648261958, + "loss": 1.023, + "step": 9501 + }, + { + "epoch": 0.6593574352924849, + "grad_norm": 2.3588733673095703, + "learning_rate": 0.00019061613033194235, + "loss": 0.8171, + "step": 9502 + }, + { + "epoch": 0.6594268267295815, + "grad_norm": 2.75038480758667, + "learning_rate": 0.0001906137538952005, + "loss": 0.9856, + "step": 9503 + }, + { + "epoch": 0.6594962181666782, + "grad_norm": 3.949540615081787, + "learning_rate": 0.00019061137717240153, + "loss": 1.1183, + "step": 9504 + }, + { + "epoch": 0.6595656096037749, + "grad_norm": 2.172715902328491, + "learning_rate": 0.000190609000163553, + "loss": 0.9316, + "step": 9505 + }, + { + "epoch": 0.6596350010408716, + "grad_norm": 3.2225165367126465, + "learning_rate": 0.00019060662286866237, + "loss": 0.8694, + "step": 9506 + }, + { + "epoch": 0.6597043924779682, + "grad_norm": 2.9207510948181152, + "learning_rate": 0.0001906042452877371, + "loss": 0.9384, + "step": 9507 + }, + { + "epoch": 0.6597737839150649, + "grad_norm": 2.5132222175598145, + "learning_rate": 0.00019060186742078474, + "loss": 0.794, + "step": 9508 + }, + { + "epoch": 0.6598431753521615, + "grad_norm": 2.052957534790039, + "learning_rate": 0.00019059948926781282, + "loss": 0.8206, + "step": 9509 + }, + { + "epoch": 0.6599125667892582, + "grad_norm": 2.1676647663116455, + "learning_rate": 0.00019059711082882884, + "loss": 0.6552, + "step": 9510 + }, + { + "epoch": 0.6599819582263549, + "grad_norm": 2.901174306869507, + "learning_rate": 0.00019059473210384024, + "loss": 0.9962, + "step": 9511 + }, + { + "epoch": 0.6600513496634516, + "grad_norm": 2.325002431869507, + "learning_rate": 0.00019059235309285463, + "loss": 0.9996, + "step": 9512 + }, + { + "epoch": 0.6601207411005482, + "grad_norm": 2.6307790279388428, + "learning_rate": 0.00019058997379587945, + "loss": 0.8458, + "step": 9513 + }, + { + "epoch": 0.6601901325376448, + "grad_norm": 2.771684169769287, + "learning_rate": 0.00019058759421292226, + "loss": 0.9152, + "step": 9514 + }, + { + "epoch": 0.6602595239747415, + "grad_norm": 2.759587049484253, + "learning_rate": 0.00019058521434399053, + "loss": 1.074, + "step": 9515 + }, + { + "epoch": 0.6603289154118381, + "grad_norm": 2.3124353885650635, + "learning_rate": 0.00019058283418909176, + "loss": 1.1975, + "step": 9516 + }, + { + "epoch": 0.6603983068489349, + "grad_norm": 2.5850977897644043, + "learning_rate": 0.00019058045374823352, + "loss": 0.8133, + "step": 9517 + }, + { + "epoch": 0.6604676982860315, + "grad_norm": 2.257513999938965, + "learning_rate": 0.00019057807302142328, + "loss": 1.0339, + "step": 9518 + }, + { + "epoch": 0.6605370897231282, + "grad_norm": 2.929926633834839, + "learning_rate": 0.00019057569200866859, + "loss": 1.153, + "step": 9519 + }, + { + "epoch": 0.6606064811602248, + "grad_norm": 2.4502766132354736, + "learning_rate": 0.00019057331070997696, + "loss": 0.6986, + "step": 9520 + }, + { + "epoch": 0.6606758725973215, + "grad_norm": 2.7817015647888184, + "learning_rate": 0.00019057092912535585, + "loss": 0.8986, + "step": 9521 + }, + { + "epoch": 0.6607452640344181, + "grad_norm": 3.294708013534546, + "learning_rate": 0.00019056854725481287, + "loss": 1.036, + "step": 9522 + }, + { + "epoch": 0.6608146554715149, + "grad_norm": 2.798823356628418, + "learning_rate": 0.00019056616509835548, + "loss": 1.0561, + "step": 9523 + }, + { + "epoch": 0.6608840469086115, + "grad_norm": 2.47725772857666, + "learning_rate": 0.00019056378265599122, + "loss": 1.105, + "step": 9524 + }, + { + "epoch": 0.6609534383457082, + "grad_norm": 4.267831325531006, + "learning_rate": 0.00019056139992772758, + "loss": 1.3051, + "step": 9525 + }, + { + "epoch": 0.6610228297828048, + "grad_norm": 2.789219379425049, + "learning_rate": 0.0001905590169135721, + "loss": 0.9518, + "step": 9526 + }, + { + "epoch": 0.6610922212199014, + "grad_norm": 3.192603826522827, + "learning_rate": 0.00019055663361353237, + "loss": 0.9134, + "step": 9527 + }, + { + "epoch": 0.6611616126569981, + "grad_norm": 3.402493476867676, + "learning_rate": 0.0001905542500276158, + "loss": 1.0572, + "step": 9528 + }, + { + "epoch": 0.6612310040940947, + "grad_norm": 2.81278657913208, + "learning_rate": 0.00019055186615583001, + "loss": 0.7688, + "step": 9529 + }, + { + "epoch": 0.6613003955311915, + "grad_norm": 2.7109785079956055, + "learning_rate": 0.00019054948199818245, + "loss": 0.7882, + "step": 9530 + }, + { + "epoch": 0.6613697869682881, + "grad_norm": 3.398810386657715, + "learning_rate": 0.0001905470975546807, + "loss": 0.9941, + "step": 9531 + }, + { + "epoch": 0.6614391784053848, + "grad_norm": 2.7076213359832764, + "learning_rate": 0.00019054471282533227, + "loss": 1.0902, + "step": 9532 + }, + { + "epoch": 0.6615085698424814, + "grad_norm": 2.4923057556152344, + "learning_rate": 0.00019054232781014468, + "loss": 0.8011, + "step": 9533 + }, + { + "epoch": 0.6615779612795781, + "grad_norm": 2.617279052734375, + "learning_rate": 0.00019053994250912545, + "loss": 0.8926, + "step": 9534 + }, + { + "epoch": 0.6616473527166747, + "grad_norm": 1.8390498161315918, + "learning_rate": 0.00019053755692228214, + "loss": 0.7728, + "step": 9535 + }, + { + "epoch": 0.6617167441537715, + "grad_norm": 2.334376335144043, + "learning_rate": 0.00019053517104962228, + "loss": 0.9494, + "step": 9536 + }, + { + "epoch": 0.6617861355908681, + "grad_norm": 2.3288915157318115, + "learning_rate": 0.00019053278489115338, + "loss": 1.0195, + "step": 9537 + }, + { + "epoch": 0.6618555270279648, + "grad_norm": 2.628883123397827, + "learning_rate": 0.00019053039844688297, + "loss": 0.8876, + "step": 9538 + }, + { + "epoch": 0.6619249184650614, + "grad_norm": 2.489572525024414, + "learning_rate": 0.00019052801171681863, + "loss": 0.9272, + "step": 9539 + }, + { + "epoch": 0.6619943099021581, + "grad_norm": 5.8741984367370605, + "learning_rate": 0.00019052562470096783, + "loss": 1.1921, + "step": 9540 + }, + { + "epoch": 0.6620637013392547, + "grad_norm": 2.063809633255005, + "learning_rate": 0.0001905232373993381, + "loss": 0.7525, + "step": 9541 + }, + { + "epoch": 0.6621330927763514, + "grad_norm": 2.6660475730895996, + "learning_rate": 0.0001905208498119371, + "loss": 1.0031, + "step": 9542 + }, + { + "epoch": 0.6622024842134481, + "grad_norm": 2.5306897163391113, + "learning_rate": 0.0001905184619387722, + "loss": 0.9081, + "step": 9543 + }, + { + "epoch": 0.6622718756505447, + "grad_norm": 2.5475594997406006, + "learning_rate": 0.00019051607377985105, + "loss": 0.9758, + "step": 9544 + }, + { + "epoch": 0.6623412670876414, + "grad_norm": 1.9828587770462036, + "learning_rate": 0.00019051368533518117, + "loss": 0.7348, + "step": 9545 + }, + { + "epoch": 0.662410658524738, + "grad_norm": 2.6248133182525635, + "learning_rate": 0.00019051129660477007, + "loss": 0.8635, + "step": 9546 + }, + { + "epoch": 0.6624800499618347, + "grad_norm": 2.569467067718506, + "learning_rate": 0.0001905089075886253, + "loss": 0.8221, + "step": 9547 + }, + { + "epoch": 0.6625494413989313, + "grad_norm": 3.762362480163574, + "learning_rate": 0.00019050651828675442, + "loss": 1.0732, + "step": 9548 + }, + { + "epoch": 0.6626188328360281, + "grad_norm": 2.3336355686187744, + "learning_rate": 0.000190504128699165, + "loss": 0.7177, + "step": 9549 + }, + { + "epoch": 0.6626882242731247, + "grad_norm": 2.511056661605835, + "learning_rate": 0.00019050173882586449, + "loss": 0.5997, + "step": 9550 + }, + { + "epoch": 0.6627576157102214, + "grad_norm": 2.8631367683410645, + "learning_rate": 0.0001904993486668605, + "loss": 0.7897, + "step": 9551 + }, + { + "epoch": 0.662827007147318, + "grad_norm": 3.1691548824310303, + "learning_rate": 0.0001904969582221606, + "loss": 0.9545, + "step": 9552 + }, + { + "epoch": 0.6628963985844147, + "grad_norm": 2.385420083999634, + "learning_rate": 0.00019049456749177228, + "loss": 0.8444, + "step": 9553 + }, + { + "epoch": 0.6629657900215113, + "grad_norm": 2.4347875118255615, + "learning_rate": 0.0001904921764757031, + "loss": 0.7297, + "step": 9554 + }, + { + "epoch": 0.663035181458608, + "grad_norm": 2.5816919803619385, + "learning_rate": 0.0001904897851739606, + "loss": 0.9674, + "step": 9555 + }, + { + "epoch": 0.6631045728957047, + "grad_norm": 3.3373072147369385, + "learning_rate": 0.00019048739358655238, + "loss": 1.2566, + "step": 9556 + }, + { + "epoch": 0.6631739643328013, + "grad_norm": 2.1509671211242676, + "learning_rate": 0.00019048500171348596, + "loss": 0.8655, + "step": 9557 + }, + { + "epoch": 0.663243355769898, + "grad_norm": 3.2381296157836914, + "learning_rate": 0.0001904826095547689, + "loss": 1.0177, + "step": 9558 + }, + { + "epoch": 0.6633127472069946, + "grad_norm": 3.012636661529541, + "learning_rate": 0.0001904802171104087, + "loss": 0.8043, + "step": 9559 + }, + { + "epoch": 0.6633821386440913, + "grad_norm": 2.3849809169769287, + "learning_rate": 0.00019047782438041296, + "loss": 0.9098, + "step": 9560 + }, + { + "epoch": 0.663451530081188, + "grad_norm": 2.472296953201294, + "learning_rate": 0.00019047543136478925, + "loss": 0.9376, + "step": 9561 + }, + { + "epoch": 0.6635209215182847, + "grad_norm": 2.6991584300994873, + "learning_rate": 0.00019047303806354508, + "loss": 1.0742, + "step": 9562 + }, + { + "epoch": 0.6635903129553813, + "grad_norm": 2.128965139389038, + "learning_rate": 0.00019047064447668804, + "loss": 0.8765, + "step": 9563 + }, + { + "epoch": 0.663659704392478, + "grad_norm": 2.6440296173095703, + "learning_rate": 0.00019046825060422566, + "loss": 0.7739, + "step": 9564 + }, + { + "epoch": 0.6637290958295746, + "grad_norm": 2.1633460521698, + "learning_rate": 0.00019046585644616552, + "loss": 0.5921, + "step": 9565 + }, + { + "epoch": 0.6637984872666713, + "grad_norm": 2.8784403800964355, + "learning_rate": 0.00019046346200251514, + "loss": 0.9107, + "step": 9566 + }, + { + "epoch": 0.663867878703768, + "grad_norm": 2.312561511993408, + "learning_rate": 0.00019046106727328214, + "loss": 0.9282, + "step": 9567 + }, + { + "epoch": 0.6639372701408646, + "grad_norm": 2.4648187160491943, + "learning_rate": 0.00019045867225847402, + "loss": 0.9651, + "step": 9568 + }, + { + "epoch": 0.6640066615779613, + "grad_norm": 2.172236919403076, + "learning_rate": 0.0001904562769580984, + "loss": 1.0496, + "step": 9569 + }, + { + "epoch": 0.6640760530150579, + "grad_norm": 3.3402013778686523, + "learning_rate": 0.00019045388137216277, + "loss": 0.9833, + "step": 9570 + }, + { + "epoch": 0.6641454444521546, + "grad_norm": 2.3630101680755615, + "learning_rate": 0.00019045148550067475, + "loss": 0.6332, + "step": 9571 + }, + { + "epoch": 0.6642148358892512, + "grad_norm": 2.2397096157073975, + "learning_rate": 0.0001904490893436419, + "loss": 0.8215, + "step": 9572 + }, + { + "epoch": 0.664284227326348, + "grad_norm": 3.993896484375, + "learning_rate": 0.00019044669290107176, + "loss": 0.8493, + "step": 9573 + }, + { + "epoch": 0.6643536187634446, + "grad_norm": 2.0946125984191895, + "learning_rate": 0.00019044429617297187, + "loss": 0.8728, + "step": 9574 + }, + { + "epoch": 0.6644230102005413, + "grad_norm": 2.351734161376953, + "learning_rate": 0.00019044189915934986, + "loss": 0.7638, + "step": 9575 + }, + { + "epoch": 0.6644924016376379, + "grad_norm": 1.900978446006775, + "learning_rate": 0.00019043950186021328, + "loss": 0.7684, + "step": 9576 + }, + { + "epoch": 0.6645617930747346, + "grad_norm": 1.703765869140625, + "learning_rate": 0.00019043710427556966, + "loss": 0.6713, + "step": 9577 + }, + { + "epoch": 0.6646311845118312, + "grad_norm": 2.0875754356384277, + "learning_rate": 0.0001904347064054266, + "loss": 0.7217, + "step": 9578 + }, + { + "epoch": 0.664700575948928, + "grad_norm": 1.7505464553833008, + "learning_rate": 0.00019043230824979167, + "loss": 0.8139, + "step": 9579 + }, + { + "epoch": 0.6647699673860246, + "grad_norm": 2.9734058380126953, + "learning_rate": 0.00019042990980867245, + "loss": 0.9868, + "step": 9580 + }, + { + "epoch": 0.6648393588231212, + "grad_norm": 2.166637420654297, + "learning_rate": 0.0001904275110820765, + "loss": 0.8228, + "step": 9581 + }, + { + "epoch": 0.6649087502602179, + "grad_norm": 3.420630693435669, + "learning_rate": 0.00019042511207001135, + "loss": 1.0063, + "step": 9582 + }, + { + "epoch": 0.6649781416973145, + "grad_norm": 2.350484609603882, + "learning_rate": 0.00019042271277248463, + "loss": 0.9199, + "step": 9583 + }, + { + "epoch": 0.6650475331344112, + "grad_norm": 2.9487996101379395, + "learning_rate": 0.00019042031318950388, + "loss": 1.0014, + "step": 9584 + }, + { + "epoch": 0.6651169245715078, + "grad_norm": 2.5318868160247803, + "learning_rate": 0.00019041791332107672, + "loss": 0.8201, + "step": 9585 + }, + { + "epoch": 0.6651863160086046, + "grad_norm": 2.81514573097229, + "learning_rate": 0.0001904155131672107, + "loss": 0.8679, + "step": 9586 + }, + { + "epoch": 0.6652557074457012, + "grad_norm": 2.3273072242736816, + "learning_rate": 0.00019041311272791338, + "loss": 1.0917, + "step": 9587 + }, + { + "epoch": 0.6653250988827979, + "grad_norm": 2.469269275665283, + "learning_rate": 0.00019041071200319237, + "loss": 0.8365, + "step": 9588 + }, + { + "epoch": 0.6653944903198945, + "grad_norm": 2.1399800777435303, + "learning_rate": 0.0001904083109930552, + "loss": 0.7528, + "step": 9589 + }, + { + "epoch": 0.6654638817569912, + "grad_norm": 2.6707491874694824, + "learning_rate": 0.0001904059096975095, + "loss": 0.9663, + "step": 9590 + }, + { + "epoch": 0.6655332731940878, + "grad_norm": 2.9664688110351562, + "learning_rate": 0.00019040350811656284, + "loss": 0.9091, + "step": 9591 + }, + { + "epoch": 0.6656026646311846, + "grad_norm": 2.5829803943634033, + "learning_rate": 0.0001904011062502228, + "loss": 0.8514, + "step": 9592 + }, + { + "epoch": 0.6656720560682812, + "grad_norm": 2.4795963764190674, + "learning_rate": 0.00019039870409849692, + "loss": 0.718, + "step": 9593 + }, + { + "epoch": 0.6657414475053778, + "grad_norm": 2.1370925903320312, + "learning_rate": 0.00019039630166139287, + "loss": 0.7178, + "step": 9594 + }, + { + "epoch": 0.6658108389424745, + "grad_norm": 2.2075517177581787, + "learning_rate": 0.00019039389893891814, + "loss": 0.7137, + "step": 9595 + }, + { + "epoch": 0.6658802303795711, + "grad_norm": 3.4710490703582764, + "learning_rate": 0.00019039149593108037, + "loss": 1.013, + "step": 9596 + }, + { + "epoch": 0.6659496218166678, + "grad_norm": 2.43026065826416, + "learning_rate": 0.00019038909263788714, + "loss": 0.8017, + "step": 9597 + }, + { + "epoch": 0.6660190132537644, + "grad_norm": 2.1188886165618896, + "learning_rate": 0.00019038668905934603, + "loss": 0.8676, + "step": 9598 + }, + { + "epoch": 0.6660884046908612, + "grad_norm": 2.6600182056427, + "learning_rate": 0.00019038428519546461, + "loss": 0.9231, + "step": 9599 + }, + { + "epoch": 0.6661577961279578, + "grad_norm": 2.3761918544769287, + "learning_rate": 0.00019038188104625052, + "loss": 0.7793, + "step": 9600 + }, + { + "epoch": 0.6662271875650545, + "grad_norm": 2.4768989086151123, + "learning_rate": 0.0001903794766117113, + "loss": 1.0023, + "step": 9601 + }, + { + "epoch": 0.6662965790021511, + "grad_norm": 3.145709991455078, + "learning_rate": 0.0001903770718918546, + "loss": 0.9206, + "step": 9602 + }, + { + "epoch": 0.6663659704392478, + "grad_norm": 2.735602378845215, + "learning_rate": 0.0001903746668866879, + "loss": 0.8635, + "step": 9603 + }, + { + "epoch": 0.6664353618763444, + "grad_norm": 2.489776134490967, + "learning_rate": 0.0001903722615962189, + "loss": 0.7755, + "step": 9604 + }, + { + "epoch": 0.6665047533134412, + "grad_norm": 2.356780529022217, + "learning_rate": 0.00019036985602045517, + "loss": 0.8737, + "step": 9605 + }, + { + "epoch": 0.6665741447505378, + "grad_norm": 2.8145318031311035, + "learning_rate": 0.00019036745015940426, + "loss": 0.9332, + "step": 9606 + }, + { + "epoch": 0.6666435361876345, + "grad_norm": 2.6540675163269043, + "learning_rate": 0.0001903650440130738, + "loss": 0.8688, + "step": 9607 + }, + { + "epoch": 0.6667129276247311, + "grad_norm": 2.8090715408325195, + "learning_rate": 0.00019036263758147136, + "loss": 0.8723, + "step": 9608 + }, + { + "epoch": 0.6667823190618277, + "grad_norm": 2.70324444770813, + "learning_rate": 0.0001903602308646046, + "loss": 0.9589, + "step": 9609 + }, + { + "epoch": 0.6668517104989244, + "grad_norm": 3.1051666736602783, + "learning_rate": 0.00019035782386248108, + "loss": 0.783, + "step": 9610 + }, + { + "epoch": 0.666921101936021, + "grad_norm": 2.856691598892212, + "learning_rate": 0.00019035541657510833, + "loss": 0.9332, + "step": 9611 + }, + { + "epoch": 0.6669904933731178, + "grad_norm": 2.7183260917663574, + "learning_rate": 0.00019035300900249407, + "loss": 0.802, + "step": 9612 + }, + { + "epoch": 0.6670598848102144, + "grad_norm": 2.520237684249878, + "learning_rate": 0.00019035060114464583, + "loss": 0.9039, + "step": 9613 + }, + { + "epoch": 0.6671292762473111, + "grad_norm": 2.2703537940979004, + "learning_rate": 0.00019034819300157123, + "loss": 0.618, + "step": 9614 + }, + { + "epoch": 0.6671986676844077, + "grad_norm": 2.0293524265289307, + "learning_rate": 0.00019034578457327784, + "loss": 0.7629, + "step": 9615 + }, + { + "epoch": 0.6672680591215044, + "grad_norm": 2.5831525325775146, + "learning_rate": 0.00019034337585977334, + "loss": 0.9234, + "step": 9616 + }, + { + "epoch": 0.667337450558601, + "grad_norm": 2.789212226867676, + "learning_rate": 0.00019034096686106522, + "loss": 0.9717, + "step": 9617 + }, + { + "epoch": 0.6674068419956978, + "grad_norm": 2.5697479248046875, + "learning_rate": 0.0001903385575771612, + "loss": 1.0332, + "step": 9618 + }, + { + "epoch": 0.6674762334327944, + "grad_norm": 2.3623106479644775, + "learning_rate": 0.00019033614800806882, + "loss": 0.7377, + "step": 9619 + }, + { + "epoch": 0.6675456248698911, + "grad_norm": 2.603193998336792, + "learning_rate": 0.0001903337381537957, + "loss": 0.6198, + "step": 9620 + }, + { + "epoch": 0.6676150163069877, + "grad_norm": 2.5833137035369873, + "learning_rate": 0.00019033132801434945, + "loss": 0.7649, + "step": 9621 + }, + { + "epoch": 0.6676844077440843, + "grad_norm": 3.1507225036621094, + "learning_rate": 0.00019032891758973767, + "loss": 1.2239, + "step": 9622 + }, + { + "epoch": 0.667753799181181, + "grad_norm": 2.6632273197174072, + "learning_rate": 0.00019032650687996798, + "loss": 0.7327, + "step": 9623 + }, + { + "epoch": 0.6678231906182777, + "grad_norm": 2.289780855178833, + "learning_rate": 0.000190324095885048, + "loss": 0.9531, + "step": 9624 + }, + { + "epoch": 0.6678925820553744, + "grad_norm": 2.4944870471954346, + "learning_rate": 0.0001903216846049853, + "loss": 0.8099, + "step": 9625 + }, + { + "epoch": 0.667961973492471, + "grad_norm": 2.426198720932007, + "learning_rate": 0.00019031927303978755, + "loss": 1.041, + "step": 9626 + }, + { + "epoch": 0.6680313649295677, + "grad_norm": 8.191691398620605, + "learning_rate": 0.00019031686118946232, + "loss": 0.9827, + "step": 9627 + }, + { + "epoch": 0.6681007563666643, + "grad_norm": 14.530017852783203, + "learning_rate": 0.00019031444905401722, + "loss": 1.0295, + "step": 9628 + }, + { + "epoch": 0.668170147803761, + "grad_norm": 2.8534908294677734, + "learning_rate": 0.00019031203663345993, + "loss": 0.8358, + "step": 9629 + }, + { + "epoch": 0.6682395392408577, + "grad_norm": 2.8250088691711426, + "learning_rate": 0.000190309623927798, + "loss": 0.8439, + "step": 9630 + }, + { + "epoch": 0.6683089306779544, + "grad_norm": 2.4649112224578857, + "learning_rate": 0.00019030721093703904, + "loss": 0.8227, + "step": 9631 + }, + { + "epoch": 0.668378322115051, + "grad_norm": 3.503098964691162, + "learning_rate": 0.00019030479766119073, + "loss": 1.1619, + "step": 9632 + }, + { + "epoch": 0.6684477135521477, + "grad_norm": 2.7655224800109863, + "learning_rate": 0.00019030238410026063, + "loss": 0.809, + "step": 9633 + }, + { + "epoch": 0.6685171049892443, + "grad_norm": 3.4923341274261475, + "learning_rate": 0.00019029997025425636, + "loss": 1.018, + "step": 9634 + }, + { + "epoch": 0.6685864964263409, + "grad_norm": 2.2312114238739014, + "learning_rate": 0.0001902975561231856, + "loss": 0.7622, + "step": 9635 + }, + { + "epoch": 0.6686558878634377, + "grad_norm": 3.7758498191833496, + "learning_rate": 0.0001902951417070559, + "loss": 1.0805, + "step": 9636 + }, + { + "epoch": 0.6687252793005343, + "grad_norm": 2.724245071411133, + "learning_rate": 0.00019029272700587496, + "loss": 1.0844, + "step": 9637 + }, + { + "epoch": 0.668794670737631, + "grad_norm": 2.8158316612243652, + "learning_rate": 0.00019029031201965032, + "loss": 1.1023, + "step": 9638 + }, + { + "epoch": 0.6688640621747276, + "grad_norm": 3.144920825958252, + "learning_rate": 0.00019028789674838966, + "loss": 1.0824, + "step": 9639 + }, + { + "epoch": 0.6689334536118243, + "grad_norm": 2.901526689529419, + "learning_rate": 0.00019028548119210054, + "loss": 1.1105, + "step": 9640 + }, + { + "epoch": 0.6690028450489209, + "grad_norm": 2.226346731185913, + "learning_rate": 0.0001902830653507907, + "loss": 0.9798, + "step": 9641 + }, + { + "epoch": 0.6690722364860177, + "grad_norm": 2.4613497257232666, + "learning_rate": 0.00019028064922446762, + "loss": 0.7814, + "step": 9642 + }, + { + "epoch": 0.6691416279231143, + "grad_norm": 2.979261875152588, + "learning_rate": 0.00019027823281313904, + "loss": 0.8591, + "step": 9643 + }, + { + "epoch": 0.669211019360211, + "grad_norm": 2.4113476276397705, + "learning_rate": 0.0001902758161168126, + "loss": 0.972, + "step": 9644 + }, + { + "epoch": 0.6692804107973076, + "grad_norm": 2.4276282787323, + "learning_rate": 0.0001902733991354958, + "loss": 0.9734, + "step": 9645 + }, + { + "epoch": 0.6693498022344043, + "grad_norm": 3.0657060146331787, + "learning_rate": 0.00019027098186919641, + "loss": 1.001, + "step": 9646 + }, + { + "epoch": 0.6694191936715009, + "grad_norm": 2.206092119216919, + "learning_rate": 0.00019026856431792198, + "loss": 0.9714, + "step": 9647 + }, + { + "epoch": 0.6694885851085975, + "grad_norm": 2.6833279132843018, + "learning_rate": 0.00019026614648168016, + "loss": 0.8115, + "step": 9648 + }, + { + "epoch": 0.6695579765456943, + "grad_norm": 1.982647180557251, + "learning_rate": 0.00019026372836047862, + "loss": 0.6618, + "step": 9649 + }, + { + "epoch": 0.6696273679827909, + "grad_norm": 2.911547899246216, + "learning_rate": 0.00019026130995432493, + "loss": 0.8718, + "step": 9650 + }, + { + "epoch": 0.6696967594198876, + "grad_norm": 2.0309829711914062, + "learning_rate": 0.00019025889126322678, + "loss": 0.813, + "step": 9651 + }, + { + "epoch": 0.6697661508569842, + "grad_norm": 2.341362237930298, + "learning_rate": 0.00019025647228719178, + "loss": 0.8138, + "step": 9652 + }, + { + "epoch": 0.6698355422940809, + "grad_norm": 2.980085611343384, + "learning_rate": 0.00019025405302622756, + "loss": 0.7379, + "step": 9653 + }, + { + "epoch": 0.6699049337311775, + "grad_norm": 2.5025012493133545, + "learning_rate": 0.00019025163348034176, + "loss": 0.8059, + "step": 9654 + }, + { + "epoch": 0.6699743251682743, + "grad_norm": 3.168292760848999, + "learning_rate": 0.00019024921364954204, + "loss": 1.2471, + "step": 9655 + }, + { + "epoch": 0.6700437166053709, + "grad_norm": 2.4340147972106934, + "learning_rate": 0.000190246793533836, + "loss": 0.7858, + "step": 9656 + }, + { + "epoch": 0.6701131080424676, + "grad_norm": 2.851841449737549, + "learning_rate": 0.00019024437313323135, + "loss": 1.1258, + "step": 9657 + }, + { + "epoch": 0.6701824994795642, + "grad_norm": 1.7244489192962646, + "learning_rate": 0.00019024195244773563, + "loss": 0.4819, + "step": 9658 + }, + { + "epoch": 0.6702518909166609, + "grad_norm": 2.7445547580718994, + "learning_rate": 0.00019023953147735657, + "loss": 0.8279, + "step": 9659 + }, + { + "epoch": 0.6703212823537575, + "grad_norm": 2.208195924758911, + "learning_rate": 0.00019023711022210177, + "loss": 0.7797, + "step": 9660 + }, + { + "epoch": 0.6703906737908543, + "grad_norm": 2.7395448684692383, + "learning_rate": 0.00019023468868197886, + "loss": 0.8395, + "step": 9661 + }, + { + "epoch": 0.6704600652279509, + "grad_norm": 2.033287763595581, + "learning_rate": 0.00019023226685699554, + "loss": 0.7426, + "step": 9662 + }, + { + "epoch": 0.6705294566650475, + "grad_norm": 2.173267126083374, + "learning_rate": 0.0001902298447471594, + "loss": 0.7612, + "step": 9663 + }, + { + "epoch": 0.6705988481021442, + "grad_norm": 2.6550652980804443, + "learning_rate": 0.0001902274223524781, + "loss": 0.8463, + "step": 9664 + }, + { + "epoch": 0.6706682395392408, + "grad_norm": 2.733083963394165, + "learning_rate": 0.00019022499967295933, + "loss": 0.8016, + "step": 9665 + }, + { + "epoch": 0.6707376309763375, + "grad_norm": 1.8864326477050781, + "learning_rate": 0.00019022257670861067, + "loss": 0.7282, + "step": 9666 + }, + { + "epoch": 0.6708070224134342, + "grad_norm": 2.736732006072998, + "learning_rate": 0.00019022015345943985, + "loss": 0.813, + "step": 9667 + }, + { + "epoch": 0.6708764138505309, + "grad_norm": 2.938350200653076, + "learning_rate": 0.00019021772992545443, + "loss": 0.8208, + "step": 9668 + }, + { + "epoch": 0.6709458052876275, + "grad_norm": 2.615056276321411, + "learning_rate": 0.0001902153061066621, + "loss": 0.9547, + "step": 9669 + }, + { + "epoch": 0.6710151967247242, + "grad_norm": 2.5482373237609863, + "learning_rate": 0.00019021288200307056, + "loss": 0.7685, + "step": 9670 + }, + { + "epoch": 0.6710845881618208, + "grad_norm": 2.6496686935424805, + "learning_rate": 0.0001902104576146874, + "loss": 0.7762, + "step": 9671 + }, + { + "epoch": 0.6711539795989175, + "grad_norm": 3.0173654556274414, + "learning_rate": 0.00019020803294152026, + "loss": 0.8834, + "step": 9672 + }, + { + "epoch": 0.6712233710360141, + "grad_norm": 2.4404537677764893, + "learning_rate": 0.00019020560798357685, + "loss": 0.7015, + "step": 9673 + }, + { + "epoch": 0.6712927624731109, + "grad_norm": 2.5143685340881348, + "learning_rate": 0.0001902031827408648, + "loss": 0.8491, + "step": 9674 + }, + { + "epoch": 0.6713621539102075, + "grad_norm": 3.423712968826294, + "learning_rate": 0.00019020075721339178, + "loss": 1.0308, + "step": 9675 + }, + { + "epoch": 0.6714315453473041, + "grad_norm": 2.207359790802002, + "learning_rate": 0.00019019833140116542, + "loss": 0.8407, + "step": 9676 + }, + { + "epoch": 0.6715009367844008, + "grad_norm": 3.9691779613494873, + "learning_rate": 0.0001901959053041934, + "loss": 0.7666, + "step": 9677 + }, + { + "epoch": 0.6715703282214974, + "grad_norm": 7.298901081085205, + "learning_rate": 0.00019019347892248336, + "loss": 0.7178, + "step": 9678 + }, + { + "epoch": 0.6716397196585941, + "grad_norm": 2.4044299125671387, + "learning_rate": 0.00019019105225604298, + "loss": 0.8597, + "step": 9679 + }, + { + "epoch": 0.6717091110956908, + "grad_norm": 2.557446002960205, + "learning_rate": 0.0001901886253048799, + "loss": 0.649, + "step": 9680 + }, + { + "epoch": 0.6717785025327875, + "grad_norm": 2.5591373443603516, + "learning_rate": 0.0001901861980690018, + "loss": 0.8776, + "step": 9681 + }, + { + "epoch": 0.6718478939698841, + "grad_norm": 2.395803689956665, + "learning_rate": 0.00019018377054841635, + "loss": 0.7376, + "step": 9682 + }, + { + "epoch": 0.6719172854069808, + "grad_norm": 1.8428099155426025, + "learning_rate": 0.0001901813427431312, + "loss": 0.7238, + "step": 9683 + }, + { + "epoch": 0.6719866768440774, + "grad_norm": 3.1623127460479736, + "learning_rate": 0.000190178914653154, + "loss": 1.1257, + "step": 9684 + }, + { + "epoch": 0.6720560682811741, + "grad_norm": 1.9431095123291016, + "learning_rate": 0.00019017648627849242, + "loss": 0.7551, + "step": 9685 + }, + { + "epoch": 0.6721254597182708, + "grad_norm": 2.554692506790161, + "learning_rate": 0.00019017405761915414, + "loss": 0.9561, + "step": 9686 + }, + { + "epoch": 0.6721948511553675, + "grad_norm": 4.385035991668701, + "learning_rate": 0.00019017162867514685, + "loss": 1.0249, + "step": 9687 + }, + { + "epoch": 0.6722642425924641, + "grad_norm": 2.831782102584839, + "learning_rate": 0.00019016919944647817, + "loss": 0.9635, + "step": 9688 + }, + { + "epoch": 0.6723336340295607, + "grad_norm": 2.0225512981414795, + "learning_rate": 0.00019016676993315577, + "loss": 0.886, + "step": 9689 + }, + { + "epoch": 0.6724030254666574, + "grad_norm": 2.389643669128418, + "learning_rate": 0.00019016434013518737, + "loss": 0.6841, + "step": 9690 + }, + { + "epoch": 0.672472416903754, + "grad_norm": 3.478163719177246, + "learning_rate": 0.00019016191005258063, + "loss": 0.9234, + "step": 9691 + }, + { + "epoch": 0.6725418083408508, + "grad_norm": 1.934086561203003, + "learning_rate": 0.00019015947968534314, + "loss": 0.7523, + "step": 9692 + }, + { + "epoch": 0.6726111997779474, + "grad_norm": 3.133470296859741, + "learning_rate": 0.00019015704903348267, + "loss": 0.9023, + "step": 9693 + }, + { + "epoch": 0.6726805912150441, + "grad_norm": 2.7918710708618164, + "learning_rate": 0.00019015461809700684, + "loss": 0.914, + "step": 9694 + }, + { + "epoch": 0.6727499826521407, + "grad_norm": 2.1152095794677734, + "learning_rate": 0.00019015218687592334, + "loss": 0.7155, + "step": 9695 + }, + { + "epoch": 0.6728193740892374, + "grad_norm": 2.9280216693878174, + "learning_rate": 0.00019014975537023988, + "loss": 1.2582, + "step": 9696 + }, + { + "epoch": 0.672888765526334, + "grad_norm": 2.589275360107422, + "learning_rate": 0.00019014732357996406, + "loss": 0.7947, + "step": 9697 + }, + { + "epoch": 0.6729581569634308, + "grad_norm": 2.392631769180298, + "learning_rate": 0.00019014489150510363, + "loss": 0.7763, + "step": 9698 + }, + { + "epoch": 0.6730275484005274, + "grad_norm": 3.0983688831329346, + "learning_rate": 0.0001901424591456662, + "loss": 0.7758, + "step": 9699 + }, + { + "epoch": 0.6730969398376241, + "grad_norm": 2.902287006378174, + "learning_rate": 0.0001901400265016595, + "loss": 1.1682, + "step": 9700 + }, + { + "epoch": 0.6731663312747207, + "grad_norm": 3.0090253353118896, + "learning_rate": 0.0001901375935730912, + "loss": 0.8558, + "step": 9701 + }, + { + "epoch": 0.6732357227118173, + "grad_norm": 2.270571231842041, + "learning_rate": 0.00019013516035996898, + "loss": 0.6779, + "step": 9702 + }, + { + "epoch": 0.673305114148914, + "grad_norm": 2.778698682785034, + "learning_rate": 0.0001901327268623005, + "loss": 0.9719, + "step": 9703 + }, + { + "epoch": 0.6733745055860106, + "grad_norm": 2.7000131607055664, + "learning_rate": 0.00019013029308009348, + "loss": 1.0808, + "step": 9704 + }, + { + "epoch": 0.6734438970231074, + "grad_norm": 3.273808002471924, + "learning_rate": 0.00019012785901335557, + "loss": 0.8528, + "step": 9705 + }, + { + "epoch": 0.673513288460204, + "grad_norm": 2.0010905265808105, + "learning_rate": 0.00019012542466209447, + "loss": 0.732, + "step": 9706 + }, + { + "epoch": 0.6735826798973007, + "grad_norm": 2.1486928462982178, + "learning_rate": 0.00019012299002631785, + "loss": 0.6595, + "step": 9707 + }, + { + "epoch": 0.6736520713343973, + "grad_norm": 2.176596164703369, + "learning_rate": 0.0001901205551060334, + "loss": 0.74, + "step": 9708 + }, + { + "epoch": 0.673721462771494, + "grad_norm": 2.494239091873169, + "learning_rate": 0.00019011811990124884, + "loss": 1.0126, + "step": 9709 + }, + { + "epoch": 0.6737908542085906, + "grad_norm": 3.012057304382324, + "learning_rate": 0.00019011568441197182, + "loss": 1.1648, + "step": 9710 + }, + { + "epoch": 0.6738602456456874, + "grad_norm": 2.5589258670806885, + "learning_rate": 0.00019011324863821004, + "loss": 1.1216, + "step": 9711 + }, + { + "epoch": 0.673929637082784, + "grad_norm": 2.3054535388946533, + "learning_rate": 0.0001901108125799712, + "loss": 0.7478, + "step": 9712 + }, + { + "epoch": 0.6739990285198807, + "grad_norm": 2.112039089202881, + "learning_rate": 0.00019010837623726295, + "loss": 0.8771, + "step": 9713 + }, + { + "epoch": 0.6740684199569773, + "grad_norm": 3.760329484939575, + "learning_rate": 0.00019010593961009303, + "loss": 1.246, + "step": 9714 + }, + { + "epoch": 0.6741378113940739, + "grad_norm": 2.125264883041382, + "learning_rate": 0.00019010350269846912, + "loss": 0.7898, + "step": 9715 + }, + { + "epoch": 0.6742072028311706, + "grad_norm": 2.4640514850616455, + "learning_rate": 0.0001901010655023989, + "loss": 0.7461, + "step": 9716 + }, + { + "epoch": 0.6742765942682672, + "grad_norm": 1.952412486076355, + "learning_rate": 0.00019009862802189006, + "loss": 0.6579, + "step": 9717 + }, + { + "epoch": 0.674345985705364, + "grad_norm": 1.856726050376892, + "learning_rate": 0.0001900961902569503, + "loss": 0.9024, + "step": 9718 + }, + { + "epoch": 0.6744153771424606, + "grad_norm": 2.508230209350586, + "learning_rate": 0.00019009375220758733, + "loss": 0.954, + "step": 9719 + }, + { + "epoch": 0.6744847685795573, + "grad_norm": 2.2156789302825928, + "learning_rate": 0.00019009131387380884, + "loss": 0.7621, + "step": 9720 + }, + { + "epoch": 0.6745541600166539, + "grad_norm": 2.6850316524505615, + "learning_rate": 0.00019008887525562252, + "loss": 0.672, + "step": 9721 + }, + { + "epoch": 0.6746235514537506, + "grad_norm": 1.958542823791504, + "learning_rate": 0.0001900864363530361, + "loss": 0.7545, + "step": 9722 + }, + { + "epoch": 0.6746929428908472, + "grad_norm": 3.112687587738037, + "learning_rate": 0.0001900839971660572, + "loss": 1.0531, + "step": 9723 + }, + { + "epoch": 0.674762334327944, + "grad_norm": 2.0408294200897217, + "learning_rate": 0.00019008155769469363, + "loss": 0.7305, + "step": 9724 + }, + { + "epoch": 0.6748317257650406, + "grad_norm": 2.5809106826782227, + "learning_rate": 0.000190079117938953, + "loss": 0.8288, + "step": 9725 + }, + { + "epoch": 0.6749011172021373, + "grad_norm": 2.7884433269500732, + "learning_rate": 0.00019007667789884306, + "loss": 0.7536, + "step": 9726 + }, + { + "epoch": 0.6749705086392339, + "grad_norm": 3.4694886207580566, + "learning_rate": 0.00019007423757437145, + "loss": 0.7805, + "step": 9727 + }, + { + "epoch": 0.6750399000763306, + "grad_norm": 2.418198347091675, + "learning_rate": 0.00019007179696554598, + "loss": 0.877, + "step": 9728 + }, + { + "epoch": 0.6751092915134272, + "grad_norm": 1.9595171213150024, + "learning_rate": 0.00019006935607237432, + "loss": 0.7935, + "step": 9729 + }, + { + "epoch": 0.6751786829505239, + "grad_norm": 2.7204806804656982, + "learning_rate": 0.00019006691489486407, + "loss": 0.9564, + "step": 9730 + }, + { + "epoch": 0.6752480743876206, + "grad_norm": 2.208289623260498, + "learning_rate": 0.00019006447343302308, + "loss": 0.8314, + "step": 9731 + }, + { + "epoch": 0.6753174658247172, + "grad_norm": 1.9053637981414795, + "learning_rate": 0.00019006203168685896, + "loss": 0.9062, + "step": 9732 + }, + { + "epoch": 0.6753868572618139, + "grad_norm": 2.6795783042907715, + "learning_rate": 0.0001900595896563795, + "loss": 1.031, + "step": 9733 + }, + { + "epoch": 0.6754562486989105, + "grad_norm": 6.029585838317871, + "learning_rate": 0.0001900571473415923, + "loss": 0.9291, + "step": 9734 + }, + { + "epoch": 0.6755256401360072, + "grad_norm": 2.288336753845215, + "learning_rate": 0.0001900547047425052, + "loss": 0.6766, + "step": 9735 + }, + { + "epoch": 0.6755950315731039, + "grad_norm": 3.6522278785705566, + "learning_rate": 0.00019005226185912582, + "loss": 0.7523, + "step": 9736 + }, + { + "epoch": 0.6756644230102006, + "grad_norm": 2.1491713523864746, + "learning_rate": 0.00019004981869146187, + "loss": 0.7223, + "step": 9737 + }, + { + "epoch": 0.6757338144472972, + "grad_norm": 2.5460903644561768, + "learning_rate": 0.00019004737523952114, + "loss": 0.7953, + "step": 9738 + }, + { + "epoch": 0.6758032058843939, + "grad_norm": 2.895578384399414, + "learning_rate": 0.00019004493150331125, + "loss": 1.0448, + "step": 9739 + }, + { + "epoch": 0.6758725973214905, + "grad_norm": 4.2252020835876465, + "learning_rate": 0.00019004248748284, + "loss": 0.9278, + "step": 9740 + }, + { + "epoch": 0.6759419887585872, + "grad_norm": 2.1432721614837646, + "learning_rate": 0.00019004004317811504, + "loss": 0.961, + "step": 9741 + }, + { + "epoch": 0.6760113801956839, + "grad_norm": 3.492260694503784, + "learning_rate": 0.0001900375985891441, + "loss": 0.7862, + "step": 9742 + }, + { + "epoch": 0.6760807716327805, + "grad_norm": 3.822903633117676, + "learning_rate": 0.0001900351537159349, + "loss": 1.1414, + "step": 9743 + }, + { + "epoch": 0.6761501630698772, + "grad_norm": 3.743776798248291, + "learning_rate": 0.0001900327085584952, + "loss": 0.94, + "step": 9744 + }, + { + "epoch": 0.6762195545069738, + "grad_norm": 2.4377384185791016, + "learning_rate": 0.00019003026311683268, + "loss": 0.8063, + "step": 9745 + }, + { + "epoch": 0.6762889459440705, + "grad_norm": 1.6628870964050293, + "learning_rate": 0.00019002781739095505, + "loss": 0.6191, + "step": 9746 + }, + { + "epoch": 0.6763583373811671, + "grad_norm": 2.856947183609009, + "learning_rate": 0.00019002537138087004, + "loss": 1.0407, + "step": 9747 + }, + { + "epoch": 0.6764277288182639, + "grad_norm": 3.5422496795654297, + "learning_rate": 0.00019002292508658538, + "loss": 1.1177, + "step": 9748 + }, + { + "epoch": 0.6764971202553605, + "grad_norm": 2.6126608848571777, + "learning_rate": 0.00019002047850810881, + "loss": 1.0179, + "step": 9749 + }, + { + "epoch": 0.6765665116924572, + "grad_norm": 3.0338315963745117, + "learning_rate": 0.00019001803164544803, + "loss": 1.1931, + "step": 9750 + }, + { + "epoch": 0.6766359031295538, + "grad_norm": 2.9232327938079834, + "learning_rate": 0.00019001558449861076, + "loss": 0.8287, + "step": 9751 + }, + { + "epoch": 0.6767052945666505, + "grad_norm": 2.162740468978882, + "learning_rate": 0.00019001313706760474, + "loss": 0.8574, + "step": 9752 + }, + { + "epoch": 0.6767746860037471, + "grad_norm": 2.527587652206421, + "learning_rate": 0.0001900106893524377, + "loss": 0.821, + "step": 9753 + }, + { + "epoch": 0.6768440774408438, + "grad_norm": 2.386244773864746, + "learning_rate": 0.00019000824135311732, + "loss": 0.6203, + "step": 9754 + }, + { + "epoch": 0.6769134688779405, + "grad_norm": 2.1415910720825195, + "learning_rate": 0.00019000579306965138, + "loss": 0.8066, + "step": 9755 + }, + { + "epoch": 0.6769828603150371, + "grad_norm": 2.4444563388824463, + "learning_rate": 0.0001900033445020476, + "loss": 0.9278, + "step": 9756 + }, + { + "epoch": 0.6770522517521338, + "grad_norm": 2.8159308433532715, + "learning_rate": 0.0001900008956503137, + "loss": 1.0254, + "step": 9757 + }, + { + "epoch": 0.6771216431892304, + "grad_norm": 2.3613455295562744, + "learning_rate": 0.0001899984465144574, + "loss": 1.0353, + "step": 9758 + }, + { + "epoch": 0.6771910346263271, + "grad_norm": 2.879774332046509, + "learning_rate": 0.00018999599709448647, + "loss": 0.972, + "step": 9759 + }, + { + "epoch": 0.6772604260634237, + "grad_norm": 2.549741744995117, + "learning_rate": 0.00018999354739040862, + "loss": 1.0759, + "step": 9760 + }, + { + "epoch": 0.6773298175005205, + "grad_norm": 7.003169536590576, + "learning_rate": 0.00018999109740223158, + "loss": 0.9362, + "step": 9761 + }, + { + "epoch": 0.6773992089376171, + "grad_norm": 2.7830817699432373, + "learning_rate": 0.00018998864712996307, + "loss": 1.301, + "step": 9762 + }, + { + "epoch": 0.6774686003747138, + "grad_norm": 2.2377402782440186, + "learning_rate": 0.00018998619657361086, + "loss": 0.7847, + "step": 9763 + }, + { + "epoch": 0.6775379918118104, + "grad_norm": 2.66288423538208, + "learning_rate": 0.00018998374573318266, + "loss": 0.8415, + "step": 9764 + }, + { + "epoch": 0.6776073832489071, + "grad_norm": 2.2557876110076904, + "learning_rate": 0.00018998129460868622, + "loss": 0.6894, + "step": 9765 + }, + { + "epoch": 0.6776767746860037, + "grad_norm": 2.8380279541015625, + "learning_rate": 0.00018997884320012926, + "loss": 1.0037, + "step": 9766 + }, + { + "epoch": 0.6777461661231005, + "grad_norm": 3.0101609230041504, + "learning_rate": 0.00018997639150751953, + "loss": 0.937, + "step": 9767 + }, + { + "epoch": 0.6778155575601971, + "grad_norm": 2.4613490104675293, + "learning_rate": 0.0001899739395308648, + "loss": 0.7084, + "step": 9768 + }, + { + "epoch": 0.6778849489972937, + "grad_norm": 2.5157926082611084, + "learning_rate": 0.00018997148727017274, + "loss": 0.6907, + "step": 9769 + }, + { + "epoch": 0.6779543404343904, + "grad_norm": 2.5452332496643066, + "learning_rate": 0.00018996903472545115, + "loss": 0.8181, + "step": 9770 + }, + { + "epoch": 0.678023731871487, + "grad_norm": 2.2476072311401367, + "learning_rate": 0.00018996658189670776, + "loss": 0.9422, + "step": 9771 + }, + { + "epoch": 0.6780931233085837, + "grad_norm": 2.6655960083007812, + "learning_rate": 0.00018996412878395031, + "loss": 0.9688, + "step": 9772 + }, + { + "epoch": 0.6781625147456803, + "grad_norm": 2.5166213512420654, + "learning_rate": 0.00018996167538718655, + "loss": 0.9641, + "step": 9773 + }, + { + "epoch": 0.6782319061827771, + "grad_norm": 2.6851556301116943, + "learning_rate": 0.0001899592217064242, + "loss": 0.7455, + "step": 9774 + }, + { + "epoch": 0.6783012976198737, + "grad_norm": 2.8452792167663574, + "learning_rate": 0.00018995676774167104, + "loss": 1.224, + "step": 9775 + }, + { + "epoch": 0.6783706890569704, + "grad_norm": 3.3034660816192627, + "learning_rate": 0.0001899543134929348, + "loss": 0.9907, + "step": 9776 + }, + { + "epoch": 0.678440080494067, + "grad_norm": 2.4834864139556885, + "learning_rate": 0.00018995185896022323, + "loss": 1.374, + "step": 9777 + }, + { + "epoch": 0.6785094719311637, + "grad_norm": 1.4293720722198486, + "learning_rate": 0.00018994940414354406, + "loss": 0.5245, + "step": 9778 + }, + { + "epoch": 0.6785788633682603, + "grad_norm": 2.6884806156158447, + "learning_rate": 0.00018994694904290505, + "loss": 0.8181, + "step": 9779 + }, + { + "epoch": 0.6786482548053571, + "grad_norm": 2.753617525100708, + "learning_rate": 0.00018994449365831398, + "loss": 1.0145, + "step": 9780 + }, + { + "epoch": 0.6787176462424537, + "grad_norm": 2.0204949378967285, + "learning_rate": 0.00018994203798977858, + "loss": 0.706, + "step": 9781 + }, + { + "epoch": 0.6787870376795504, + "grad_norm": 2.0456888675689697, + "learning_rate": 0.00018993958203730661, + "loss": 0.7045, + "step": 9782 + }, + { + "epoch": 0.678856429116647, + "grad_norm": 2.4788482189178467, + "learning_rate": 0.00018993712580090582, + "loss": 0.7083, + "step": 9783 + }, + { + "epoch": 0.6789258205537436, + "grad_norm": 3.2028093338012695, + "learning_rate": 0.00018993466928058394, + "loss": 0.9471, + "step": 9784 + }, + { + "epoch": 0.6789952119908403, + "grad_norm": 2.0541162490844727, + "learning_rate": 0.00018993221247634874, + "loss": 0.7301, + "step": 9785 + }, + { + "epoch": 0.679064603427937, + "grad_norm": 3.16056227684021, + "learning_rate": 0.00018992975538820796, + "loss": 1.3003, + "step": 9786 + }, + { + "epoch": 0.6791339948650337, + "grad_norm": 2.2998387813568115, + "learning_rate": 0.0001899272980161694, + "loss": 0.8053, + "step": 9787 + }, + { + "epoch": 0.6792033863021303, + "grad_norm": 2.538158893585205, + "learning_rate": 0.00018992484036024082, + "loss": 0.8941, + "step": 9788 + }, + { + "epoch": 0.679272777739227, + "grad_norm": 2.479717969894409, + "learning_rate": 0.00018992238242042993, + "loss": 0.7701, + "step": 9789 + }, + { + "epoch": 0.6793421691763236, + "grad_norm": 2.376783609390259, + "learning_rate": 0.00018991992419674448, + "loss": 0.7066, + "step": 9790 + }, + { + "epoch": 0.6794115606134203, + "grad_norm": 2.7268900871276855, + "learning_rate": 0.00018991746568919232, + "loss": 1.0801, + "step": 9791 + }, + { + "epoch": 0.679480952050517, + "grad_norm": 2.1106083393096924, + "learning_rate": 0.0001899150068977811, + "loss": 0.7618, + "step": 9792 + }, + { + "epoch": 0.6795503434876137, + "grad_norm": 2.1368632316589355, + "learning_rate": 0.00018991254782251867, + "loss": 0.7272, + "step": 9793 + }, + { + "epoch": 0.6796197349247103, + "grad_norm": 2.6783556938171387, + "learning_rate": 0.00018991008846341273, + "loss": 0.8131, + "step": 9794 + }, + { + "epoch": 0.679689126361807, + "grad_norm": 6.636091232299805, + "learning_rate": 0.00018990762882047107, + "loss": 1.2248, + "step": 9795 + }, + { + "epoch": 0.6797585177989036, + "grad_norm": 2.5395328998565674, + "learning_rate": 0.00018990516889370152, + "loss": 1.1841, + "step": 9796 + }, + { + "epoch": 0.6798279092360002, + "grad_norm": 3.616464614868164, + "learning_rate": 0.00018990270868311172, + "loss": 0.8786, + "step": 9797 + }, + { + "epoch": 0.679897300673097, + "grad_norm": 1.9883253574371338, + "learning_rate": 0.0001899002481887095, + "loss": 0.9249, + "step": 9798 + }, + { + "epoch": 0.6799666921101936, + "grad_norm": 1.8627535104751587, + "learning_rate": 0.00018989778741050265, + "loss": 0.7991, + "step": 9799 + }, + { + "epoch": 0.6800360835472903, + "grad_norm": 1.9591491222381592, + "learning_rate": 0.0001898953263484989, + "loss": 0.7247, + "step": 9800 + }, + { + "epoch": 0.6801054749843869, + "grad_norm": 2.216933012008667, + "learning_rate": 0.00018989286500270604, + "loss": 0.8979, + "step": 9801 + }, + { + "epoch": 0.6801748664214836, + "grad_norm": 2.474621295928955, + "learning_rate": 0.00018989040337313187, + "loss": 0.7133, + "step": 9802 + }, + { + "epoch": 0.6802442578585802, + "grad_norm": 2.278602361679077, + "learning_rate": 0.00018988794145978407, + "loss": 0.8393, + "step": 9803 + }, + { + "epoch": 0.680313649295677, + "grad_norm": 2.6376914978027344, + "learning_rate": 0.0001898854792626705, + "loss": 0.6738, + "step": 9804 + }, + { + "epoch": 0.6803830407327736, + "grad_norm": 3.177255153656006, + "learning_rate": 0.0001898830167817989, + "loss": 1.1114, + "step": 9805 + }, + { + "epoch": 0.6804524321698703, + "grad_norm": 2.561448335647583, + "learning_rate": 0.00018988055401717704, + "loss": 0.9224, + "step": 9806 + }, + { + "epoch": 0.6805218236069669, + "grad_norm": 2.8530032634735107, + "learning_rate": 0.0001898780909688127, + "loss": 0.9611, + "step": 9807 + }, + { + "epoch": 0.6805912150440636, + "grad_norm": 2.754936933517456, + "learning_rate": 0.00018987562763671365, + "loss": 0.9224, + "step": 9808 + }, + { + "epoch": 0.6806606064811602, + "grad_norm": 3.0570425987243652, + "learning_rate": 0.00018987316402088765, + "loss": 0.8262, + "step": 9809 + }, + { + "epoch": 0.6807299979182568, + "grad_norm": 2.431185483932495, + "learning_rate": 0.00018987070012134254, + "loss": 0.8792, + "step": 9810 + }, + { + "epoch": 0.6807993893553536, + "grad_norm": 2.141146421432495, + "learning_rate": 0.00018986823593808604, + "loss": 0.94, + "step": 9811 + }, + { + "epoch": 0.6808687807924502, + "grad_norm": 3.6607251167297363, + "learning_rate": 0.0001898657714711259, + "loss": 0.9946, + "step": 9812 + }, + { + "epoch": 0.6809381722295469, + "grad_norm": 2.166632890701294, + "learning_rate": 0.00018986330672047, + "loss": 0.9486, + "step": 9813 + }, + { + "epoch": 0.6810075636666435, + "grad_norm": 2.268657684326172, + "learning_rate": 0.00018986084168612607, + "loss": 0.8937, + "step": 9814 + }, + { + "epoch": 0.6810769551037402, + "grad_norm": 2.577988386154175, + "learning_rate": 0.00018985837636810188, + "loss": 0.9703, + "step": 9815 + }, + { + "epoch": 0.6811463465408368, + "grad_norm": 2.029878854751587, + "learning_rate": 0.00018985591076640522, + "loss": 0.7717, + "step": 9816 + }, + { + "epoch": 0.6812157379779336, + "grad_norm": 2.832129955291748, + "learning_rate": 0.00018985344488104387, + "loss": 0.8716, + "step": 9817 + }, + { + "epoch": 0.6812851294150302, + "grad_norm": 2.8444230556488037, + "learning_rate": 0.0001898509787120256, + "loss": 1.034, + "step": 9818 + }, + { + "epoch": 0.6813545208521269, + "grad_norm": 2.994675636291504, + "learning_rate": 0.00018984851225935825, + "loss": 0.9908, + "step": 9819 + }, + { + "epoch": 0.6814239122892235, + "grad_norm": 2.2965166568756104, + "learning_rate": 0.00018984604552304957, + "loss": 0.7691, + "step": 9820 + }, + { + "epoch": 0.6814933037263202, + "grad_norm": 2.8235228061676025, + "learning_rate": 0.00018984357850310732, + "loss": 0.9256, + "step": 9821 + }, + { + "epoch": 0.6815626951634168, + "grad_norm": 2.849841594696045, + "learning_rate": 0.00018984111119953932, + "loss": 1.1916, + "step": 9822 + }, + { + "epoch": 0.6816320866005134, + "grad_norm": 1.8250192403793335, + "learning_rate": 0.00018983864361235336, + "loss": 0.6017, + "step": 9823 + }, + { + "epoch": 0.6817014780376102, + "grad_norm": 2.3835391998291016, + "learning_rate": 0.00018983617574155724, + "loss": 0.9824, + "step": 9824 + }, + { + "epoch": 0.6817708694747068, + "grad_norm": 1.7383109331130981, + "learning_rate": 0.00018983370758715872, + "loss": 0.7895, + "step": 9825 + }, + { + "epoch": 0.6818402609118035, + "grad_norm": 2.1065053939819336, + "learning_rate": 0.00018983123914916562, + "loss": 0.7175, + "step": 9826 + }, + { + "epoch": 0.6819096523489001, + "grad_norm": 2.5734517574310303, + "learning_rate": 0.0001898287704275857, + "loss": 0.935, + "step": 9827 + }, + { + "epoch": 0.6819790437859968, + "grad_norm": 2.314059257507324, + "learning_rate": 0.00018982630142242682, + "loss": 0.9785, + "step": 9828 + }, + { + "epoch": 0.6820484352230934, + "grad_norm": 2.391294002532959, + "learning_rate": 0.00018982383213369668, + "loss": 0.8063, + "step": 9829 + }, + { + "epoch": 0.6821178266601902, + "grad_norm": 2.4316112995147705, + "learning_rate": 0.00018982136256140314, + "loss": 0.8182, + "step": 9830 + }, + { + "epoch": 0.6821872180972868, + "grad_norm": 2.305206537246704, + "learning_rate": 0.00018981889270555397, + "loss": 0.7168, + "step": 9831 + }, + { + "epoch": 0.6822566095343835, + "grad_norm": 3.040703773498535, + "learning_rate": 0.000189816422566157, + "loss": 0.9401, + "step": 9832 + }, + { + "epoch": 0.6823260009714801, + "grad_norm": 2.680757522583008, + "learning_rate": 0.00018981395214321997, + "loss": 0.9651, + "step": 9833 + }, + { + "epoch": 0.6823953924085768, + "grad_norm": 2.7613487243652344, + "learning_rate": 0.00018981148143675072, + "loss": 1.0073, + "step": 9834 + }, + { + "epoch": 0.6824647838456734, + "grad_norm": 1.668739914894104, + "learning_rate": 0.0001898090104467571, + "loss": 0.6454, + "step": 9835 + }, + { + "epoch": 0.6825341752827702, + "grad_norm": 2.6641037464141846, + "learning_rate": 0.0001898065391732468, + "loss": 0.9622, + "step": 9836 + }, + { + "epoch": 0.6826035667198668, + "grad_norm": 2.923081636428833, + "learning_rate": 0.00018980406761622768, + "loss": 1.0074, + "step": 9837 + }, + { + "epoch": 0.6826729581569634, + "grad_norm": 2.068121910095215, + "learning_rate": 0.00018980159577570754, + "loss": 0.8095, + "step": 9838 + }, + { + "epoch": 0.6827423495940601, + "grad_norm": 3.1775643825531006, + "learning_rate": 0.00018979912365169418, + "loss": 1.2637, + "step": 9839 + }, + { + "epoch": 0.6828117410311567, + "grad_norm": 3.3862807750701904, + "learning_rate": 0.00018979665124419542, + "loss": 0.7633, + "step": 9840 + }, + { + "epoch": 0.6828811324682534, + "grad_norm": 2.2197699546813965, + "learning_rate": 0.00018979417855321905, + "loss": 0.8751, + "step": 9841 + }, + { + "epoch": 0.68295052390535, + "grad_norm": 2.2694802284240723, + "learning_rate": 0.00018979170557877287, + "loss": 0.7385, + "step": 9842 + }, + { + "epoch": 0.6830199153424468, + "grad_norm": 2.1294968128204346, + "learning_rate": 0.00018978923232086466, + "loss": 0.7415, + "step": 9843 + }, + { + "epoch": 0.6830893067795434, + "grad_norm": 1.9987549781799316, + "learning_rate": 0.0001897867587795023, + "loss": 0.6392, + "step": 9844 + }, + { + "epoch": 0.6831586982166401, + "grad_norm": 2.149083375930786, + "learning_rate": 0.00018978428495469357, + "loss": 0.7698, + "step": 9845 + }, + { + "epoch": 0.6832280896537367, + "grad_norm": 2.234952211380005, + "learning_rate": 0.00018978181084644625, + "loss": 0.8517, + "step": 9846 + }, + { + "epoch": 0.6832974810908334, + "grad_norm": 3.160029172897339, + "learning_rate": 0.00018977933645476818, + "loss": 0.9495, + "step": 9847 + }, + { + "epoch": 0.68336687252793, + "grad_norm": 2.1451187133789062, + "learning_rate": 0.00018977686177966714, + "loss": 0.6319, + "step": 9848 + }, + { + "epoch": 0.6834362639650268, + "grad_norm": 2.363394260406494, + "learning_rate": 0.000189774386821151, + "loss": 0.7167, + "step": 9849 + }, + { + "epoch": 0.6835056554021234, + "grad_norm": 3.185861349105835, + "learning_rate": 0.00018977191157922748, + "loss": 1.0009, + "step": 9850 + }, + { + "epoch": 0.68357504683922, + "grad_norm": 2.5327394008636475, + "learning_rate": 0.0001897694360539045, + "loss": 0.864, + "step": 9851 + }, + { + "epoch": 0.6836444382763167, + "grad_norm": 2.3483176231384277, + "learning_rate": 0.0001897669602451898, + "loss": 0.8504, + "step": 9852 + }, + { + "epoch": 0.6837138297134133, + "grad_norm": 2.5587215423583984, + "learning_rate": 0.00018976448415309122, + "loss": 1.1384, + "step": 9853 + }, + { + "epoch": 0.68378322115051, + "grad_norm": 2.6565589904785156, + "learning_rate": 0.00018976200777761658, + "loss": 0.989, + "step": 9854 + }, + { + "epoch": 0.6838526125876067, + "grad_norm": 1.8599729537963867, + "learning_rate": 0.00018975953111877373, + "loss": 0.6125, + "step": 9855 + }, + { + "epoch": 0.6839220040247034, + "grad_norm": 2.331502914428711, + "learning_rate": 0.00018975705417657044, + "loss": 0.9998, + "step": 9856 + }, + { + "epoch": 0.6839913954618, + "grad_norm": 1.9574600458145142, + "learning_rate": 0.0001897545769510145, + "loss": 0.7165, + "step": 9857 + }, + { + "epoch": 0.6840607868988967, + "grad_norm": 2.915253162384033, + "learning_rate": 0.0001897520994421138, + "loss": 1.009, + "step": 9858 + }, + { + "epoch": 0.6841301783359933, + "grad_norm": 1.8260973691940308, + "learning_rate": 0.00018974962164987615, + "loss": 0.7869, + "step": 9859 + }, + { + "epoch": 0.68419956977309, + "grad_norm": 2.6859207153320312, + "learning_rate": 0.00018974714357430935, + "loss": 1.0712, + "step": 9860 + }, + { + "epoch": 0.6842689612101867, + "grad_norm": 2.513416290283203, + "learning_rate": 0.00018974466521542126, + "loss": 0.9591, + "step": 9861 + }, + { + "epoch": 0.6843383526472834, + "grad_norm": 2.6857073307037354, + "learning_rate": 0.00018974218657321962, + "loss": 1.0565, + "step": 9862 + }, + { + "epoch": 0.68440774408438, + "grad_norm": 2.4395527839660645, + "learning_rate": 0.0001897397076477123, + "loss": 0.8251, + "step": 9863 + }, + { + "epoch": 0.6844771355214766, + "grad_norm": 2.3358030319213867, + "learning_rate": 0.0001897372284389072, + "loss": 0.7, + "step": 9864 + }, + { + "epoch": 0.6845465269585733, + "grad_norm": 1.9657845497131348, + "learning_rate": 0.00018973474894681204, + "loss": 0.8331, + "step": 9865 + }, + { + "epoch": 0.6846159183956699, + "grad_norm": 2.20774245262146, + "learning_rate": 0.0001897322691714347, + "loss": 0.8976, + "step": 9866 + }, + { + "epoch": 0.6846853098327667, + "grad_norm": 2.498189926147461, + "learning_rate": 0.00018972978911278298, + "loss": 0.9388, + "step": 9867 + }, + { + "epoch": 0.6847547012698633, + "grad_norm": 2.2480876445770264, + "learning_rate": 0.00018972730877086475, + "loss": 0.7885, + "step": 9868 + }, + { + "epoch": 0.68482409270696, + "grad_norm": 3.196851968765259, + "learning_rate": 0.00018972482814568782, + "loss": 1.2544, + "step": 9869 + }, + { + "epoch": 0.6848934841440566, + "grad_norm": 3.0809853076934814, + "learning_rate": 0.00018972234723726, + "loss": 0.9832, + "step": 9870 + }, + { + "epoch": 0.6849628755811533, + "grad_norm": 2.3818466663360596, + "learning_rate": 0.00018971986604558914, + "loss": 0.956, + "step": 9871 + }, + { + "epoch": 0.6850322670182499, + "grad_norm": 3.020097255706787, + "learning_rate": 0.00018971738457068308, + "loss": 1.0238, + "step": 9872 + }, + { + "epoch": 0.6851016584553467, + "grad_norm": 2.795003652572632, + "learning_rate": 0.00018971490281254963, + "loss": 0.7243, + "step": 9873 + }, + { + "epoch": 0.6851710498924433, + "grad_norm": 2.372767448425293, + "learning_rate": 0.00018971242077119667, + "loss": 0.736, + "step": 9874 + }, + { + "epoch": 0.68524044132954, + "grad_norm": 1.9756301641464233, + "learning_rate": 0.00018970993844663196, + "loss": 0.8118, + "step": 9875 + }, + { + "epoch": 0.6853098327666366, + "grad_norm": 2.5821096897125244, + "learning_rate": 0.00018970745583886343, + "loss": 0.8647, + "step": 9876 + }, + { + "epoch": 0.6853792242037332, + "grad_norm": 2.930802583694458, + "learning_rate": 0.00018970497294789883, + "loss": 0.9538, + "step": 9877 + }, + { + "epoch": 0.6854486156408299, + "grad_norm": 2.108637571334839, + "learning_rate": 0.00018970248977374606, + "loss": 0.832, + "step": 9878 + }, + { + "epoch": 0.6855180070779265, + "grad_norm": 2.12275767326355, + "learning_rate": 0.00018970000631641294, + "loss": 0.6138, + "step": 9879 + }, + { + "epoch": 0.6855873985150233, + "grad_norm": 2.72135329246521, + "learning_rate": 0.0001896975225759073, + "loss": 1.0647, + "step": 9880 + }, + { + "epoch": 0.6856567899521199, + "grad_norm": 2.136345624923706, + "learning_rate": 0.000189695038552237, + "loss": 0.8269, + "step": 9881 + }, + { + "epoch": 0.6857261813892166, + "grad_norm": 2.5755014419555664, + "learning_rate": 0.00018969255424540986, + "loss": 0.6743, + "step": 9882 + }, + { + "epoch": 0.6857955728263132, + "grad_norm": 3.148834466934204, + "learning_rate": 0.00018969006965543375, + "loss": 0.8149, + "step": 9883 + }, + { + "epoch": 0.6858649642634099, + "grad_norm": 2.546908140182495, + "learning_rate": 0.00018968758478231647, + "loss": 0.9974, + "step": 9884 + }, + { + "epoch": 0.6859343557005065, + "grad_norm": 2.235473394393921, + "learning_rate": 0.0001896850996260659, + "loss": 0.7384, + "step": 9885 + }, + { + "epoch": 0.6860037471376033, + "grad_norm": 2.856708288192749, + "learning_rate": 0.00018968261418668986, + "loss": 0.9354, + "step": 9886 + }, + { + "epoch": 0.6860731385746999, + "grad_norm": 2.99552845954895, + "learning_rate": 0.00018968012846419624, + "loss": 1.0653, + "step": 9887 + }, + { + "epoch": 0.6861425300117966, + "grad_norm": 1.7315384149551392, + "learning_rate": 0.00018967764245859286, + "loss": 0.7353, + "step": 9888 + }, + { + "epoch": 0.6862119214488932, + "grad_norm": 2.3664612770080566, + "learning_rate": 0.00018967515616988756, + "loss": 0.9186, + "step": 9889 + }, + { + "epoch": 0.6862813128859898, + "grad_norm": 2.7746219635009766, + "learning_rate": 0.00018967266959808818, + "loss": 0.8198, + "step": 9890 + }, + { + "epoch": 0.6863507043230865, + "grad_norm": 2.4201276302337646, + "learning_rate": 0.0001896701827432026, + "loss": 0.6251, + "step": 9891 + }, + { + "epoch": 0.6864200957601831, + "grad_norm": 2.063469648361206, + "learning_rate": 0.00018966769560523864, + "loss": 0.8426, + "step": 9892 + }, + { + "epoch": 0.6864894871972799, + "grad_norm": 2.9951889514923096, + "learning_rate": 0.0001896652081842042, + "loss": 0.8261, + "step": 9893 + }, + { + "epoch": 0.6865588786343765, + "grad_norm": 2.018418550491333, + "learning_rate": 0.00018966272048010708, + "loss": 0.8824, + "step": 9894 + }, + { + "epoch": 0.6866282700714732, + "grad_norm": 2.189602851867676, + "learning_rate": 0.00018966023249295514, + "loss": 0.7777, + "step": 9895 + }, + { + "epoch": 0.6866976615085698, + "grad_norm": 2.718717575073242, + "learning_rate": 0.0001896577442227563, + "loss": 0.8437, + "step": 9896 + }, + { + "epoch": 0.6867670529456665, + "grad_norm": 1.76239812374115, + "learning_rate": 0.00018965525566951832, + "loss": 0.682, + "step": 9897 + }, + { + "epoch": 0.6868364443827631, + "grad_norm": 2.5185861587524414, + "learning_rate": 0.00018965276683324911, + "loss": 0.8877, + "step": 9898 + }, + { + "epoch": 0.6869058358198599, + "grad_norm": 1.8548197746276855, + "learning_rate": 0.00018965027771395652, + "loss": 0.7761, + "step": 9899 + }, + { + "epoch": 0.6869752272569565, + "grad_norm": 2.1698503494262695, + "learning_rate": 0.0001896477883116484, + "loss": 0.6314, + "step": 9900 + }, + { + "epoch": 0.6870446186940532, + "grad_norm": 1.8096610307693481, + "learning_rate": 0.0001896452986263326, + "loss": 0.5968, + "step": 9901 + }, + { + "epoch": 0.6871140101311498, + "grad_norm": 2.2606818675994873, + "learning_rate": 0.000189642808658017, + "loss": 0.7626, + "step": 9902 + }, + { + "epoch": 0.6871834015682465, + "grad_norm": 2.0752780437469482, + "learning_rate": 0.00018964031840670945, + "loss": 0.8974, + "step": 9903 + }, + { + "epoch": 0.6872527930053431, + "grad_norm": 2.9972317218780518, + "learning_rate": 0.00018963782787241785, + "loss": 0.9322, + "step": 9904 + }, + { + "epoch": 0.6873221844424398, + "grad_norm": 2.092752695083618, + "learning_rate": 0.00018963533705515, + "loss": 0.8753, + "step": 9905 + }, + { + "epoch": 0.6873915758795365, + "grad_norm": 2.8584067821502686, + "learning_rate": 0.0001896328459549138, + "loss": 1.0086, + "step": 9906 + }, + { + "epoch": 0.6874609673166331, + "grad_norm": 2.2432265281677246, + "learning_rate": 0.0001896303545717171, + "loss": 0.7702, + "step": 9907 + }, + { + "epoch": 0.6875303587537298, + "grad_norm": 1.7584933042526245, + "learning_rate": 0.00018962786290556777, + "loss": 0.6784, + "step": 9908 + }, + { + "epoch": 0.6875997501908264, + "grad_norm": 2.6063284873962402, + "learning_rate": 0.00018962537095647367, + "loss": 0.7355, + "step": 9909 + }, + { + "epoch": 0.6876691416279231, + "grad_norm": 2.283404588699341, + "learning_rate": 0.00018962287872444268, + "loss": 0.6446, + "step": 9910 + }, + { + "epoch": 0.6877385330650198, + "grad_norm": 2.331713914871216, + "learning_rate": 0.00018962038620948262, + "loss": 0.7339, + "step": 9911 + }, + { + "epoch": 0.6878079245021165, + "grad_norm": 2.0147039890289307, + "learning_rate": 0.00018961789341160145, + "loss": 0.7524, + "step": 9912 + }, + { + "epoch": 0.6878773159392131, + "grad_norm": 2.5623223781585693, + "learning_rate": 0.000189615400330807, + "loss": 1.0314, + "step": 9913 + }, + { + "epoch": 0.6879467073763098, + "grad_norm": 2.344327688217163, + "learning_rate": 0.0001896129069671071, + "loss": 0.8889, + "step": 9914 + }, + { + "epoch": 0.6880160988134064, + "grad_norm": 3.280574083328247, + "learning_rate": 0.00018961041332050962, + "loss": 0.8798, + "step": 9915 + }, + { + "epoch": 0.6880854902505031, + "grad_norm": 3.2295782566070557, + "learning_rate": 0.00018960791939102252, + "loss": 1.0215, + "step": 9916 + }, + { + "epoch": 0.6881548816875998, + "grad_norm": 2.9317269325256348, + "learning_rate": 0.00018960542517865357, + "loss": 1.0029, + "step": 9917 + }, + { + "epoch": 0.6882242731246964, + "grad_norm": 2.0388550758361816, + "learning_rate": 0.00018960293068341071, + "loss": 0.8071, + "step": 9918 + }, + { + "epoch": 0.6882936645617931, + "grad_norm": 1.9401218891143799, + "learning_rate": 0.0001896004359053018, + "loss": 0.9295, + "step": 9919 + }, + { + "epoch": 0.6883630559988897, + "grad_norm": 2.1811444759368896, + "learning_rate": 0.00018959794084433469, + "loss": 0.7752, + "step": 9920 + }, + { + "epoch": 0.6884324474359864, + "grad_norm": 2.4654173851013184, + "learning_rate": 0.00018959544550051728, + "loss": 0.8765, + "step": 9921 + }, + { + "epoch": 0.688501838873083, + "grad_norm": 2.507328748703003, + "learning_rate": 0.00018959294987385745, + "loss": 0.8886, + "step": 9922 + }, + { + "epoch": 0.6885712303101797, + "grad_norm": 1.6881047487258911, + "learning_rate": 0.00018959045396436307, + "loss": 0.7746, + "step": 9923 + }, + { + "epoch": 0.6886406217472764, + "grad_norm": 2.2610931396484375, + "learning_rate": 0.00018958795777204203, + "loss": 1.1527, + "step": 9924 + }, + { + "epoch": 0.6887100131843731, + "grad_norm": 2.622042417526245, + "learning_rate": 0.0001895854612969022, + "loss": 1.0001, + "step": 9925 + }, + { + "epoch": 0.6887794046214697, + "grad_norm": 2.2766876220703125, + "learning_rate": 0.00018958296453895144, + "loss": 0.829, + "step": 9926 + }, + { + "epoch": 0.6888487960585664, + "grad_norm": 2.7127413749694824, + "learning_rate": 0.00018958046749819766, + "loss": 1.0157, + "step": 9927 + }, + { + "epoch": 0.688918187495663, + "grad_norm": 2.5932536125183105, + "learning_rate": 0.00018957797017464878, + "loss": 0.9723, + "step": 9928 + }, + { + "epoch": 0.6889875789327597, + "grad_norm": 3.1806724071502686, + "learning_rate": 0.0001895754725683126, + "loss": 1.0983, + "step": 9929 + }, + { + "epoch": 0.6890569703698564, + "grad_norm": 1.958181619644165, + "learning_rate": 0.00018957297467919705, + "loss": 0.826, + "step": 9930 + }, + { + "epoch": 0.689126361806953, + "grad_norm": 2.878007650375366, + "learning_rate": 0.00018957047650731003, + "loss": 0.7805, + "step": 9931 + }, + { + "epoch": 0.6891957532440497, + "grad_norm": 4.4831342697143555, + "learning_rate": 0.00018956797805265937, + "loss": 0.914, + "step": 9932 + }, + { + "epoch": 0.6892651446811463, + "grad_norm": 1.7974950075149536, + "learning_rate": 0.00018956547931525303, + "loss": 0.7542, + "step": 9933 + }, + { + "epoch": 0.689334536118243, + "grad_norm": 2.22784686088562, + "learning_rate": 0.00018956298029509884, + "loss": 0.7609, + "step": 9934 + }, + { + "epoch": 0.6894039275553396, + "grad_norm": 2.3562865257263184, + "learning_rate": 0.00018956048099220475, + "loss": 1.102, + "step": 9935 + }, + { + "epoch": 0.6894733189924364, + "grad_norm": 2.5762956142425537, + "learning_rate": 0.00018955798140657856, + "loss": 1.0683, + "step": 9936 + }, + { + "epoch": 0.689542710429533, + "grad_norm": 2.417529582977295, + "learning_rate": 0.00018955548153822823, + "loss": 1.1799, + "step": 9937 + }, + { + "epoch": 0.6896121018666297, + "grad_norm": 2.2568581104278564, + "learning_rate": 0.00018955298138716167, + "loss": 0.8132, + "step": 9938 + }, + { + "epoch": 0.6896814933037263, + "grad_norm": 2.6073238849639893, + "learning_rate": 0.00018955048095338674, + "loss": 0.7454, + "step": 9939 + }, + { + "epoch": 0.689750884740823, + "grad_norm": 3.6135759353637695, + "learning_rate": 0.0001895479802369113, + "loss": 0.8716, + "step": 9940 + }, + { + "epoch": 0.6898202761779196, + "grad_norm": 1.6647067070007324, + "learning_rate": 0.00018954547923774327, + "loss": 0.7682, + "step": 9941 + }, + { + "epoch": 0.6898896676150164, + "grad_norm": 2.6076905727386475, + "learning_rate": 0.00018954297795589057, + "loss": 0.9278, + "step": 9942 + }, + { + "epoch": 0.689959059052113, + "grad_norm": 3.949589729309082, + "learning_rate": 0.00018954047639136104, + "loss": 1.1289, + "step": 9943 + }, + { + "epoch": 0.6900284504892096, + "grad_norm": 3.0680289268493652, + "learning_rate": 0.00018953797454416264, + "loss": 1.0598, + "step": 9944 + }, + { + "epoch": 0.6900978419263063, + "grad_norm": 2.3400979042053223, + "learning_rate": 0.00018953547241430327, + "loss": 1.0183, + "step": 9945 + }, + { + "epoch": 0.6901672333634029, + "grad_norm": 3.860549211502075, + "learning_rate": 0.00018953297000179076, + "loss": 1.0587, + "step": 9946 + }, + { + "epoch": 0.6902366248004996, + "grad_norm": 2.1520705223083496, + "learning_rate": 0.0001895304673066331, + "loss": 0.7185, + "step": 9947 + }, + { + "epoch": 0.6903060162375962, + "grad_norm": 2.148418426513672, + "learning_rate": 0.00018952796432883807, + "loss": 0.7796, + "step": 9948 + }, + { + "epoch": 0.690375407674693, + "grad_norm": 1.9361706972122192, + "learning_rate": 0.0001895254610684137, + "loss": 0.9362, + "step": 9949 + }, + { + "epoch": 0.6904447991117896, + "grad_norm": 1.9213858842849731, + "learning_rate": 0.00018952295752536785, + "loss": 0.7408, + "step": 9950 + }, + { + "epoch": 0.6905141905488863, + "grad_norm": 2.273242712020874, + "learning_rate": 0.00018952045369970838, + "loss": 0.8141, + "step": 9951 + }, + { + "epoch": 0.6905835819859829, + "grad_norm": 2.764063596725464, + "learning_rate": 0.0001895179495914432, + "loss": 0.8966, + "step": 9952 + }, + { + "epoch": 0.6906529734230796, + "grad_norm": 2.8024747371673584, + "learning_rate": 0.00018951544520058028, + "loss": 0.7043, + "step": 9953 + }, + { + "epoch": 0.6907223648601762, + "grad_norm": 2.698777437210083, + "learning_rate": 0.00018951294052712748, + "loss": 1.017, + "step": 9954 + }, + { + "epoch": 0.690791756297273, + "grad_norm": 2.4411604404449463, + "learning_rate": 0.0001895104355710927, + "loss": 0.7167, + "step": 9955 + }, + { + "epoch": 0.6908611477343696, + "grad_norm": 2.660750150680542, + "learning_rate": 0.00018950793033248388, + "loss": 0.7688, + "step": 9956 + }, + { + "epoch": 0.6909305391714663, + "grad_norm": 2.1806070804595947, + "learning_rate": 0.00018950542481130887, + "loss": 0.9289, + "step": 9957 + }, + { + "epoch": 0.6909999306085629, + "grad_norm": 2.238123655319214, + "learning_rate": 0.00018950291900757564, + "loss": 0.9695, + "step": 9958 + }, + { + "epoch": 0.6910693220456595, + "grad_norm": 2.0008628368377686, + "learning_rate": 0.00018950041292129207, + "loss": 0.5899, + "step": 9959 + }, + { + "epoch": 0.6911387134827562, + "grad_norm": 4.681221008300781, + "learning_rate": 0.0001894979065524661, + "loss": 1.0006, + "step": 9960 + }, + { + "epoch": 0.6912081049198529, + "grad_norm": 2.061012029647827, + "learning_rate": 0.0001894953999011056, + "loss": 0.926, + "step": 9961 + }, + { + "epoch": 0.6912774963569496, + "grad_norm": 2.2723164558410645, + "learning_rate": 0.00018949289296721851, + "loss": 0.9279, + "step": 9962 + }, + { + "epoch": 0.6913468877940462, + "grad_norm": 1.6533856391906738, + "learning_rate": 0.00018949038575081275, + "loss": 0.5954, + "step": 9963 + }, + { + "epoch": 0.6914162792311429, + "grad_norm": 2.5321905612945557, + "learning_rate": 0.00018948787825189624, + "loss": 1.1953, + "step": 9964 + }, + { + "epoch": 0.6914856706682395, + "grad_norm": 2.664482593536377, + "learning_rate": 0.00018948537047047687, + "loss": 0.9315, + "step": 9965 + }, + { + "epoch": 0.6915550621053362, + "grad_norm": 2.4476318359375, + "learning_rate": 0.00018948286240656254, + "loss": 0.7861, + "step": 9966 + }, + { + "epoch": 0.6916244535424328, + "grad_norm": 1.7792210578918457, + "learning_rate": 0.00018948035406016122, + "loss": 0.6271, + "step": 9967 + }, + { + "epoch": 0.6916938449795296, + "grad_norm": 2.2002193927764893, + "learning_rate": 0.0001894778454312808, + "loss": 0.6508, + "step": 9968 + }, + { + "epoch": 0.6917632364166262, + "grad_norm": 2.97243332862854, + "learning_rate": 0.0001894753365199292, + "loss": 1.1027, + "step": 9969 + }, + { + "epoch": 0.6918326278537229, + "grad_norm": 2.1884467601776123, + "learning_rate": 0.00018947282732611437, + "loss": 0.867, + "step": 9970 + }, + { + "epoch": 0.6919020192908195, + "grad_norm": 2.4512088298797607, + "learning_rate": 0.00018947031784984417, + "loss": 1.1412, + "step": 9971 + }, + { + "epoch": 0.6919714107279161, + "grad_norm": 2.638451099395752, + "learning_rate": 0.00018946780809112656, + "loss": 0.8355, + "step": 9972 + }, + { + "epoch": 0.6920408021650128, + "grad_norm": 2.4225986003875732, + "learning_rate": 0.0001894652980499695, + "loss": 0.8631, + "step": 9973 + }, + { + "epoch": 0.6921101936021095, + "grad_norm": 2.73146915435791, + "learning_rate": 0.0001894627877263808, + "loss": 1.1289, + "step": 9974 + }, + { + "epoch": 0.6921795850392062, + "grad_norm": 2.3728132247924805, + "learning_rate": 0.00018946027712036853, + "loss": 1.0182, + "step": 9975 + }, + { + "epoch": 0.6922489764763028, + "grad_norm": 2.1734843254089355, + "learning_rate": 0.00018945776623194053, + "loss": 0.8508, + "step": 9976 + }, + { + "epoch": 0.6923183679133995, + "grad_norm": 1.9184120893478394, + "learning_rate": 0.00018945525506110472, + "loss": 0.6051, + "step": 9977 + }, + { + "epoch": 0.6923877593504961, + "grad_norm": 1.9456300735473633, + "learning_rate": 0.00018945274360786908, + "loss": 0.7523, + "step": 9978 + }, + { + "epoch": 0.6924571507875928, + "grad_norm": 2.4800338745117188, + "learning_rate": 0.00018945023187224146, + "loss": 0.7308, + "step": 9979 + }, + { + "epoch": 0.6925265422246895, + "grad_norm": 2.360076665878296, + "learning_rate": 0.00018944771985422983, + "loss": 0.8347, + "step": 9980 + }, + { + "epoch": 0.6925959336617862, + "grad_norm": 2.154374599456787, + "learning_rate": 0.00018944520755384217, + "loss": 0.8912, + "step": 9981 + }, + { + "epoch": 0.6926653250988828, + "grad_norm": 2.7563557624816895, + "learning_rate": 0.00018944269497108636, + "loss": 0.8588, + "step": 9982 + }, + { + "epoch": 0.6927347165359795, + "grad_norm": 2.2645652294158936, + "learning_rate": 0.00018944018210597033, + "loss": 0.8204, + "step": 9983 + }, + { + "epoch": 0.6928041079730761, + "grad_norm": 4.843137741088867, + "learning_rate": 0.00018943766895850202, + "loss": 0.9139, + "step": 9984 + }, + { + "epoch": 0.6928734994101727, + "grad_norm": 2.8431899547576904, + "learning_rate": 0.00018943515552868938, + "loss": 0.8133, + "step": 9985 + }, + { + "epoch": 0.6929428908472695, + "grad_norm": 2.7886879444122314, + "learning_rate": 0.00018943264181654033, + "loss": 0.9783, + "step": 9986 + }, + { + "epoch": 0.6930122822843661, + "grad_norm": 2.810544013977051, + "learning_rate": 0.0001894301278220628, + "loss": 0.7916, + "step": 9987 + }, + { + "epoch": 0.6930816737214628, + "grad_norm": 1.7596077919006348, + "learning_rate": 0.0001894276135452647, + "loss": 0.7197, + "step": 9988 + }, + { + "epoch": 0.6931510651585594, + "grad_norm": 2.120802640914917, + "learning_rate": 0.00018942509898615403, + "loss": 0.6574, + "step": 9989 + }, + { + "epoch": 0.6932204565956561, + "grad_norm": 1.9676536321640015, + "learning_rate": 0.0001894225841447387, + "loss": 0.9351, + "step": 9990 + }, + { + "epoch": 0.6932898480327527, + "grad_norm": 2.0423972606658936, + "learning_rate": 0.00018942006902102665, + "loss": 0.6393, + "step": 9991 + }, + { + "epoch": 0.6933592394698495, + "grad_norm": 2.137193202972412, + "learning_rate": 0.00018941755361502583, + "loss": 0.7701, + "step": 9992 + }, + { + "epoch": 0.6934286309069461, + "grad_norm": 1.8892594575881958, + "learning_rate": 0.00018941503792674414, + "loss": 0.6447, + "step": 9993 + }, + { + "epoch": 0.6934980223440428, + "grad_norm": 3.2682204246520996, + "learning_rate": 0.00018941252195618953, + "loss": 1.0979, + "step": 9994 + }, + { + "epoch": 0.6935674137811394, + "grad_norm": 2.5762953758239746, + "learning_rate": 0.00018941000570337, + "loss": 0.7306, + "step": 9995 + }, + { + "epoch": 0.6936368052182361, + "grad_norm": 2.5403237342834473, + "learning_rate": 0.00018940748916829346, + "loss": 0.8297, + "step": 9996 + }, + { + "epoch": 0.6937061966553327, + "grad_norm": 2.1342272758483887, + "learning_rate": 0.00018940497235096781, + "loss": 0.5363, + "step": 9997 + }, + { + "epoch": 0.6937755880924293, + "grad_norm": 3.644958257675171, + "learning_rate": 0.00018940245525140106, + "loss": 0.6194, + "step": 9998 + }, + { + "epoch": 0.6938449795295261, + "grad_norm": 2.350559949874878, + "learning_rate": 0.00018939993786960113, + "loss": 0.9598, + "step": 9999 + }, + { + "epoch": 0.6939143709666227, + "grad_norm": 2.1697776317596436, + "learning_rate": 0.00018939742020557599, + "loss": 1.0878, + "step": 10000 + }, + { + "epoch": 0.6939837624037194, + "grad_norm": 2.268677234649658, + "learning_rate": 0.00018939490225933353, + "loss": 0.7533, + "step": 10001 + }, + { + "epoch": 0.694053153840816, + "grad_norm": 1.7714265584945679, + "learning_rate": 0.00018939238403088175, + "loss": 0.6116, + "step": 10002 + }, + { + "epoch": 0.6941225452779127, + "grad_norm": 2.4668753147125244, + "learning_rate": 0.00018938986552022858, + "loss": 0.9505, + "step": 10003 + }, + { + "epoch": 0.6941919367150093, + "grad_norm": 2.063065528869629, + "learning_rate": 0.00018938734672738198, + "loss": 0.5675, + "step": 10004 + }, + { + "epoch": 0.6942613281521061, + "grad_norm": 2.6240031719207764, + "learning_rate": 0.0001893848276523499, + "loss": 0.8062, + "step": 10005 + }, + { + "epoch": 0.6943307195892027, + "grad_norm": 2.0476396083831787, + "learning_rate": 0.0001893823082951403, + "loss": 1.0052, + "step": 10006 + }, + { + "epoch": 0.6944001110262994, + "grad_norm": 3.3669557571411133, + "learning_rate": 0.0001893797886557611, + "loss": 1.0082, + "step": 10007 + }, + { + "epoch": 0.694469502463396, + "grad_norm": 2.2865211963653564, + "learning_rate": 0.0001893772687342203, + "loss": 0.8603, + "step": 10008 + }, + { + "epoch": 0.6945388939004927, + "grad_norm": 2.1110966205596924, + "learning_rate": 0.0001893747485305258, + "loss": 0.8772, + "step": 10009 + }, + { + "epoch": 0.6946082853375893, + "grad_norm": 1.9713352918624878, + "learning_rate": 0.0001893722280446856, + "loss": 0.7769, + "step": 10010 + }, + { + "epoch": 0.6946776767746861, + "grad_norm": 2.51283860206604, + "learning_rate": 0.00018936970727670768, + "loss": 0.7355, + "step": 10011 + }, + { + "epoch": 0.6947470682117827, + "grad_norm": 2.570375680923462, + "learning_rate": 0.00018936718622659991, + "loss": 1.057, + "step": 10012 + }, + { + "epoch": 0.6948164596488793, + "grad_norm": 3.5933525562286377, + "learning_rate": 0.00018936466489437035, + "loss": 1.041, + "step": 10013 + }, + { + "epoch": 0.694885851085976, + "grad_norm": 2.252579689025879, + "learning_rate": 0.00018936214328002692, + "loss": 0.7784, + "step": 10014 + }, + { + "epoch": 0.6949552425230726, + "grad_norm": 2.69497013092041, + "learning_rate": 0.00018935962138357751, + "loss": 0.7829, + "step": 10015 + }, + { + "epoch": 0.6950246339601693, + "grad_norm": 2.188173770904541, + "learning_rate": 0.00018935709920503017, + "loss": 0.9736, + "step": 10016 + }, + { + "epoch": 0.695094025397266, + "grad_norm": 2.6129631996154785, + "learning_rate": 0.00018935457674439286, + "loss": 1.2528, + "step": 10017 + }, + { + "epoch": 0.6951634168343627, + "grad_norm": 2.0257275104522705, + "learning_rate": 0.0001893520540016735, + "loss": 0.9214, + "step": 10018 + }, + { + "epoch": 0.6952328082714593, + "grad_norm": 1.9090681076049805, + "learning_rate": 0.00018934953097688006, + "loss": 0.8886, + "step": 10019 + }, + { + "epoch": 0.695302199708556, + "grad_norm": 2.95182466506958, + "learning_rate": 0.00018934700767002052, + "loss": 1.0783, + "step": 10020 + }, + { + "epoch": 0.6953715911456526, + "grad_norm": 2.183349370956421, + "learning_rate": 0.00018934448408110285, + "loss": 0.7202, + "step": 10021 + }, + { + "epoch": 0.6954409825827493, + "grad_norm": 2.346961498260498, + "learning_rate": 0.000189341960210135, + "loss": 0.5988, + "step": 10022 + }, + { + "epoch": 0.695510374019846, + "grad_norm": 2.2833147048950195, + "learning_rate": 0.00018933943605712496, + "loss": 0.8732, + "step": 10023 + }, + { + "epoch": 0.6955797654569427, + "grad_norm": 2.49131441116333, + "learning_rate": 0.00018933691162208066, + "loss": 0.9582, + "step": 10024 + }, + { + "epoch": 0.6956491568940393, + "grad_norm": 2.2468719482421875, + "learning_rate": 0.00018933438690501016, + "loss": 0.765, + "step": 10025 + }, + { + "epoch": 0.6957185483311359, + "grad_norm": 2.322188138961792, + "learning_rate": 0.0001893318619059213, + "loss": 1.0067, + "step": 10026 + }, + { + "epoch": 0.6957879397682326, + "grad_norm": 2.090169668197632, + "learning_rate": 0.00018932933662482216, + "loss": 0.6361, + "step": 10027 + }, + { + "epoch": 0.6958573312053292, + "grad_norm": 1.6755527257919312, + "learning_rate": 0.00018932681106172063, + "loss": 0.5026, + "step": 10028 + }, + { + "epoch": 0.6959267226424259, + "grad_norm": 2.257439136505127, + "learning_rate": 0.00018932428521662472, + "loss": 0.8213, + "step": 10029 + }, + { + "epoch": 0.6959961140795226, + "grad_norm": 2.0536537170410156, + "learning_rate": 0.00018932175908954247, + "loss": 0.7139, + "step": 10030 + }, + { + "epoch": 0.6960655055166193, + "grad_norm": 2.582270860671997, + "learning_rate": 0.0001893192326804817, + "loss": 0.5679, + "step": 10031 + }, + { + "epoch": 0.6961348969537159, + "grad_norm": 2.4603543281555176, + "learning_rate": 0.00018931670598945055, + "loss": 0.706, + "step": 10032 + }, + { + "epoch": 0.6962042883908126, + "grad_norm": 2.845081329345703, + "learning_rate": 0.0001893141790164569, + "loss": 1.0858, + "step": 10033 + }, + { + "epoch": 0.6962736798279092, + "grad_norm": 2.2422666549682617, + "learning_rate": 0.0001893116517615087, + "loss": 0.6623, + "step": 10034 + }, + { + "epoch": 0.6963430712650059, + "grad_norm": 2.34972882270813, + "learning_rate": 0.00018930912422461402, + "loss": 0.756, + "step": 10035 + }, + { + "epoch": 0.6964124627021026, + "grad_norm": 2.0668628215789795, + "learning_rate": 0.0001893065964057808, + "loss": 0.9089, + "step": 10036 + }, + { + "epoch": 0.6964818541391993, + "grad_norm": 3.024843454360962, + "learning_rate": 0.00018930406830501703, + "loss": 1.0829, + "step": 10037 + }, + { + "epoch": 0.6965512455762959, + "grad_norm": 2.28787899017334, + "learning_rate": 0.00018930153992233065, + "loss": 0.75, + "step": 10038 + }, + { + "epoch": 0.6966206370133925, + "grad_norm": 1.9889389276504517, + "learning_rate": 0.00018929901125772967, + "loss": 1.0025, + "step": 10039 + }, + { + "epoch": 0.6966900284504892, + "grad_norm": 2.5303854942321777, + "learning_rate": 0.0001892964823112221, + "loss": 1.0779, + "step": 10040 + }, + { + "epoch": 0.6967594198875858, + "grad_norm": 1.936132550239563, + "learning_rate": 0.00018929395308281587, + "loss": 0.7791, + "step": 10041 + }, + { + "epoch": 0.6968288113246826, + "grad_norm": 2.5891685485839844, + "learning_rate": 0.00018929142357251898, + "loss": 0.7352, + "step": 10042 + }, + { + "epoch": 0.6968982027617792, + "grad_norm": 2.896768808364868, + "learning_rate": 0.00018928889378033945, + "loss": 1.1164, + "step": 10043 + }, + { + "epoch": 0.6969675941988759, + "grad_norm": 2.4172842502593994, + "learning_rate": 0.00018928636370628524, + "loss": 0.9096, + "step": 10044 + }, + { + "epoch": 0.6970369856359725, + "grad_norm": 2.086425542831421, + "learning_rate": 0.00018928383335036433, + "loss": 0.7011, + "step": 10045 + }, + { + "epoch": 0.6971063770730692, + "grad_norm": 1.8238061666488647, + "learning_rate": 0.0001892813027125847, + "loss": 0.6599, + "step": 10046 + }, + { + "epoch": 0.6971757685101658, + "grad_norm": 2.050679922103882, + "learning_rate": 0.0001892787717929544, + "loss": 0.6748, + "step": 10047 + }, + { + "epoch": 0.6972451599472625, + "grad_norm": 1.83005952835083, + "learning_rate": 0.00018927624059148133, + "loss": 0.8297, + "step": 10048 + }, + { + "epoch": 0.6973145513843592, + "grad_norm": 2.116492748260498, + "learning_rate": 0.00018927370910817355, + "loss": 0.6598, + "step": 10049 + }, + { + "epoch": 0.6973839428214559, + "grad_norm": 2.1814911365509033, + "learning_rate": 0.00018927117734303904, + "loss": 0.9495, + "step": 10050 + }, + { + "epoch": 0.6974533342585525, + "grad_norm": 1.949371337890625, + "learning_rate": 0.00018926864529608578, + "loss": 0.7731, + "step": 10051 + }, + { + "epoch": 0.6975227256956491, + "grad_norm": 2.1469788551330566, + "learning_rate": 0.0001892661129673217, + "loss": 0.732, + "step": 10052 + }, + { + "epoch": 0.6975921171327458, + "grad_norm": 2.086663007736206, + "learning_rate": 0.00018926358035675493, + "loss": 0.9251, + "step": 10053 + }, + { + "epoch": 0.6976615085698424, + "grad_norm": 2.621964693069458, + "learning_rate": 0.00018926104746439337, + "loss": 0.8669, + "step": 10054 + }, + { + "epoch": 0.6977309000069392, + "grad_norm": 1.5534377098083496, + "learning_rate": 0.00018925851429024503, + "loss": 0.5614, + "step": 10055 + }, + { + "epoch": 0.6978002914440358, + "grad_norm": 2.236862897872925, + "learning_rate": 0.00018925598083431794, + "loss": 0.7018, + "step": 10056 + }, + { + "epoch": 0.6978696828811325, + "grad_norm": 2.166663646697998, + "learning_rate": 0.00018925344709662005, + "loss": 0.9422, + "step": 10057 + }, + { + "epoch": 0.6979390743182291, + "grad_norm": 1.967741847038269, + "learning_rate": 0.00018925091307715942, + "loss": 0.808, + "step": 10058 + }, + { + "epoch": 0.6980084657553258, + "grad_norm": 2.599752902984619, + "learning_rate": 0.00018924837877594397, + "loss": 1.0004, + "step": 10059 + }, + { + "epoch": 0.6980778571924224, + "grad_norm": 1.8516663312911987, + "learning_rate": 0.00018924584419298174, + "loss": 1.2059, + "step": 10060 + }, + { + "epoch": 0.6981472486295192, + "grad_norm": 2.269763708114624, + "learning_rate": 0.00018924330932828075, + "loss": 0.8885, + "step": 10061 + }, + { + "epoch": 0.6982166400666158, + "grad_norm": 2.2848751544952393, + "learning_rate": 0.000189240774181849, + "loss": 0.8224, + "step": 10062 + }, + { + "epoch": 0.6982860315037125, + "grad_norm": 2.328162670135498, + "learning_rate": 0.0001892382387536945, + "loss": 0.8461, + "step": 10063 + }, + { + "epoch": 0.6983554229408091, + "grad_norm": 4.201348304748535, + "learning_rate": 0.00018923570304382517, + "loss": 1.0726, + "step": 10064 + }, + { + "epoch": 0.6984248143779057, + "grad_norm": 2.368911027908325, + "learning_rate": 0.00018923316705224912, + "loss": 1.2676, + "step": 10065 + }, + { + "epoch": 0.6984942058150024, + "grad_norm": 2.0045554637908936, + "learning_rate": 0.0001892306307789743, + "loss": 0.7964, + "step": 10066 + }, + { + "epoch": 0.698563597252099, + "grad_norm": 2.433490753173828, + "learning_rate": 0.00018922809422400872, + "loss": 0.9718, + "step": 10067 + }, + { + "epoch": 0.6986329886891958, + "grad_norm": 1.8548593521118164, + "learning_rate": 0.00018922555738736042, + "loss": 0.7643, + "step": 10068 + }, + { + "epoch": 0.6987023801262924, + "grad_norm": 1.8457276821136475, + "learning_rate": 0.0001892230202690374, + "loss": 0.6604, + "step": 10069 + }, + { + "epoch": 0.6987717715633891, + "grad_norm": 1.988672137260437, + "learning_rate": 0.00018922048286904762, + "loss": 0.8954, + "step": 10070 + }, + { + "epoch": 0.6988411630004857, + "grad_norm": 2.50154972076416, + "learning_rate": 0.00018921794518739917, + "loss": 1.1804, + "step": 10071 + }, + { + "epoch": 0.6989105544375824, + "grad_norm": 2.618150234222412, + "learning_rate": 0.00018921540722410002, + "loss": 0.9255, + "step": 10072 + }, + { + "epoch": 0.698979945874679, + "grad_norm": 2.4311211109161377, + "learning_rate": 0.00018921286897915813, + "loss": 0.9345, + "step": 10073 + }, + { + "epoch": 0.6990493373117758, + "grad_norm": 2.0033822059631348, + "learning_rate": 0.0001892103304525816, + "loss": 0.7629, + "step": 10074 + }, + { + "epoch": 0.6991187287488724, + "grad_norm": 2.3818912506103516, + "learning_rate": 0.00018920779164437838, + "loss": 0.8222, + "step": 10075 + }, + { + "epoch": 0.6991881201859691, + "grad_norm": 2.987346887588501, + "learning_rate": 0.00018920525255455655, + "loss": 1.0566, + "step": 10076 + }, + { + "epoch": 0.6992575116230657, + "grad_norm": 2.4216489791870117, + "learning_rate": 0.00018920271318312404, + "loss": 0.6897, + "step": 10077 + }, + { + "epoch": 0.6993269030601624, + "grad_norm": 2.1225738525390625, + "learning_rate": 0.00018920017353008897, + "loss": 0.8275, + "step": 10078 + }, + { + "epoch": 0.699396294497259, + "grad_norm": 2.4236197471618652, + "learning_rate": 0.00018919763359545925, + "loss": 0.8986, + "step": 10079 + }, + { + "epoch": 0.6994656859343557, + "grad_norm": 2.908444404602051, + "learning_rate": 0.00018919509337924297, + "loss": 0.8455, + "step": 10080 + }, + { + "epoch": 0.6995350773714524, + "grad_norm": 2.113524913787842, + "learning_rate": 0.00018919255288144812, + "loss": 0.8163, + "step": 10081 + }, + { + "epoch": 0.699604468808549, + "grad_norm": 2.19632887840271, + "learning_rate": 0.00018919001210208273, + "loss": 0.9297, + "step": 10082 + }, + { + "epoch": 0.6996738602456457, + "grad_norm": 2.4864373207092285, + "learning_rate": 0.00018918747104115484, + "loss": 0.8212, + "step": 10083 + }, + { + "epoch": 0.6997432516827423, + "grad_norm": 2.0169966220855713, + "learning_rate": 0.00018918492969867244, + "loss": 0.7038, + "step": 10084 + }, + { + "epoch": 0.699812643119839, + "grad_norm": 2.046450138092041, + "learning_rate": 0.00018918238807464358, + "loss": 0.8248, + "step": 10085 + }, + { + "epoch": 0.6998820345569357, + "grad_norm": 1.9870245456695557, + "learning_rate": 0.00018917984616907625, + "loss": 0.9106, + "step": 10086 + }, + { + "epoch": 0.6999514259940324, + "grad_norm": 2.2732760906219482, + "learning_rate": 0.0001891773039819785, + "loss": 1.0184, + "step": 10087 + }, + { + "epoch": 0.700020817431129, + "grad_norm": 2.437772035598755, + "learning_rate": 0.00018917476151335833, + "loss": 0.8623, + "step": 10088 + }, + { + "epoch": 0.7000902088682257, + "grad_norm": 2.3073129653930664, + "learning_rate": 0.0001891722187632238, + "loss": 0.8881, + "step": 10089 + }, + { + "epoch": 0.7001596003053223, + "grad_norm": 2.882862091064453, + "learning_rate": 0.00018916967573158293, + "loss": 1.0095, + "step": 10090 + }, + { + "epoch": 0.700228991742419, + "grad_norm": 1.8040862083435059, + "learning_rate": 0.0001891671324184437, + "loss": 0.8126, + "step": 10091 + }, + { + "epoch": 0.7002983831795156, + "grad_norm": 2.286616563796997, + "learning_rate": 0.00018916458882381422, + "loss": 0.9906, + "step": 10092 + }, + { + "epoch": 0.7003677746166123, + "grad_norm": 1.7964303493499756, + "learning_rate": 0.00018916204494770248, + "loss": 0.7495, + "step": 10093 + }, + { + "epoch": 0.700437166053709, + "grad_norm": 3.1358044147491455, + "learning_rate": 0.00018915950079011648, + "loss": 0.9598, + "step": 10094 + }, + { + "epoch": 0.7005065574908056, + "grad_norm": 9.2219820022583, + "learning_rate": 0.00018915695635106426, + "loss": 0.9234, + "step": 10095 + }, + { + "epoch": 0.7005759489279023, + "grad_norm": 2.3167264461517334, + "learning_rate": 0.00018915441163055391, + "loss": 0.7387, + "step": 10096 + }, + { + "epoch": 0.7006453403649989, + "grad_norm": 2.303075075149536, + "learning_rate": 0.00018915186662859346, + "loss": 0.8472, + "step": 10097 + }, + { + "epoch": 0.7007147318020956, + "grad_norm": 1.8199284076690674, + "learning_rate": 0.00018914932134519086, + "loss": 0.6475, + "step": 10098 + }, + { + "epoch": 0.7007841232391923, + "grad_norm": 2.6208746433258057, + "learning_rate": 0.0001891467757803542, + "loss": 1.0628, + "step": 10099 + }, + { + "epoch": 0.700853514676289, + "grad_norm": 2.372331380844116, + "learning_rate": 0.00018914422993409155, + "loss": 0.9221, + "step": 10100 + }, + { + "epoch": 0.7009229061133856, + "grad_norm": 2.655674934387207, + "learning_rate": 0.00018914168380641084, + "loss": 0.7904, + "step": 10101 + }, + { + "epoch": 0.7009922975504823, + "grad_norm": 2.6207451820373535, + "learning_rate": 0.00018913913739732023, + "loss": 0.855, + "step": 10102 + }, + { + "epoch": 0.7010616889875789, + "grad_norm": 2.311941146850586, + "learning_rate": 0.00018913659070682768, + "loss": 0.9621, + "step": 10103 + }, + { + "epoch": 0.7011310804246756, + "grad_norm": 2.0257315635681152, + "learning_rate": 0.00018913404373494127, + "loss": 0.6672, + "step": 10104 + }, + { + "epoch": 0.7012004718617723, + "grad_norm": 2.2502377033233643, + "learning_rate": 0.00018913149648166903, + "loss": 0.7191, + "step": 10105 + }, + { + "epoch": 0.7012698632988689, + "grad_norm": 2.119321823120117, + "learning_rate": 0.00018912894894701897, + "loss": 0.7882, + "step": 10106 + }, + { + "epoch": 0.7013392547359656, + "grad_norm": 2.695805311203003, + "learning_rate": 0.00018912640113099918, + "loss": 1.0182, + "step": 10107 + }, + { + "epoch": 0.7014086461730622, + "grad_norm": 1.9879544973373413, + "learning_rate": 0.00018912385303361767, + "loss": 0.7255, + "step": 10108 + }, + { + "epoch": 0.7014780376101589, + "grad_norm": 2.3613617420196533, + "learning_rate": 0.0001891213046548825, + "loss": 0.9257, + "step": 10109 + }, + { + "epoch": 0.7015474290472555, + "grad_norm": 2.2483341693878174, + "learning_rate": 0.00018911875599480173, + "loss": 0.8465, + "step": 10110 + }, + { + "epoch": 0.7016168204843523, + "grad_norm": 2.330672025680542, + "learning_rate": 0.0001891162070533834, + "loss": 0.7346, + "step": 10111 + }, + { + "epoch": 0.7016862119214489, + "grad_norm": 1.862907886505127, + "learning_rate": 0.00018911365783063548, + "loss": 0.6927, + "step": 10112 + }, + { + "epoch": 0.7017556033585456, + "grad_norm": 2.3077807426452637, + "learning_rate": 0.00018911110832656611, + "loss": 0.9967, + "step": 10113 + }, + { + "epoch": 0.7018249947956422, + "grad_norm": 2.7828755378723145, + "learning_rate": 0.00018910855854118334, + "loss": 1.1168, + "step": 10114 + }, + { + "epoch": 0.7018943862327389, + "grad_norm": 2.2550055980682373, + "learning_rate": 0.00018910600847449517, + "loss": 1.0225, + "step": 10115 + }, + { + "epoch": 0.7019637776698355, + "grad_norm": 2.1011462211608887, + "learning_rate": 0.00018910345812650964, + "loss": 0.9824, + "step": 10116 + }, + { + "epoch": 0.7020331691069323, + "grad_norm": 2.39278244972229, + "learning_rate": 0.00018910090749723487, + "loss": 0.7176, + "step": 10117 + }, + { + "epoch": 0.7021025605440289, + "grad_norm": 2.9230172634124756, + "learning_rate": 0.00018909835658667887, + "loss": 0.8988, + "step": 10118 + }, + { + "epoch": 0.7021719519811255, + "grad_norm": 2.263716697692871, + "learning_rate": 0.00018909580539484968, + "loss": 0.7919, + "step": 10119 + }, + { + "epoch": 0.7022413434182222, + "grad_norm": 1.6470719575881958, + "learning_rate": 0.00018909325392175535, + "loss": 0.6267, + "step": 10120 + }, + { + "epoch": 0.7023107348553188, + "grad_norm": 2.2300570011138916, + "learning_rate": 0.000189090702167404, + "loss": 0.7867, + "step": 10121 + }, + { + "epoch": 0.7023801262924155, + "grad_norm": 2.689854621887207, + "learning_rate": 0.00018908815013180363, + "loss": 0.7986, + "step": 10122 + }, + { + "epoch": 0.7024495177295121, + "grad_norm": 2.548421859741211, + "learning_rate": 0.0001890855978149623, + "loss": 0.6634, + "step": 10123 + }, + { + "epoch": 0.7025189091666089, + "grad_norm": 2.0867931842803955, + "learning_rate": 0.00018908304521688803, + "loss": 0.6995, + "step": 10124 + }, + { + "epoch": 0.7025883006037055, + "grad_norm": 2.819010019302368, + "learning_rate": 0.00018908049233758896, + "loss": 0.6696, + "step": 10125 + }, + { + "epoch": 0.7026576920408022, + "grad_norm": 2.5064244270324707, + "learning_rate": 0.00018907793917707314, + "loss": 0.8086, + "step": 10126 + }, + { + "epoch": 0.7027270834778988, + "grad_norm": 2.2560923099517822, + "learning_rate": 0.00018907538573534858, + "loss": 0.9327, + "step": 10127 + }, + { + "epoch": 0.7027964749149955, + "grad_norm": 2.6791460514068604, + "learning_rate": 0.00018907283201242335, + "loss": 1.1582, + "step": 10128 + }, + { + "epoch": 0.7028658663520921, + "grad_norm": 2.7447004318237305, + "learning_rate": 0.00018907027800830554, + "loss": 0.7347, + "step": 10129 + }, + { + "epoch": 0.7029352577891889, + "grad_norm": 2.3241870403289795, + "learning_rate": 0.00018906772372300317, + "loss": 0.7016, + "step": 10130 + }, + { + "epoch": 0.7030046492262855, + "grad_norm": 2.373220205307007, + "learning_rate": 0.00018906516915652435, + "loss": 0.8035, + "step": 10131 + }, + { + "epoch": 0.7030740406633822, + "grad_norm": 2.349360466003418, + "learning_rate": 0.0001890626143088771, + "loss": 0.7072, + "step": 10132 + }, + { + "epoch": 0.7031434321004788, + "grad_norm": 2.2886242866516113, + "learning_rate": 0.00018906005918006952, + "loss": 0.8757, + "step": 10133 + }, + { + "epoch": 0.7032128235375754, + "grad_norm": 1.8316482305526733, + "learning_rate": 0.00018905750377010968, + "loss": 0.6648, + "step": 10134 + }, + { + "epoch": 0.7032822149746721, + "grad_norm": 2.985057830810547, + "learning_rate": 0.00018905494807900563, + "loss": 0.787, + "step": 10135 + }, + { + "epoch": 0.7033516064117687, + "grad_norm": 2.5371949672698975, + "learning_rate": 0.00018905239210676545, + "loss": 0.9296, + "step": 10136 + }, + { + "epoch": 0.7034209978488655, + "grad_norm": 3.1826741695404053, + "learning_rate": 0.00018904983585339718, + "loss": 0.879, + "step": 10137 + }, + { + "epoch": 0.7034903892859621, + "grad_norm": 2.891836166381836, + "learning_rate": 0.0001890472793189089, + "loss": 0.8751, + "step": 10138 + }, + { + "epoch": 0.7035597807230588, + "grad_norm": 2.1386184692382812, + "learning_rate": 0.00018904472250330873, + "loss": 0.725, + "step": 10139 + }, + { + "epoch": 0.7036291721601554, + "grad_norm": 2.5615830421447754, + "learning_rate": 0.00018904216540660467, + "loss": 0.8979, + "step": 10140 + }, + { + "epoch": 0.7036985635972521, + "grad_norm": 5.76805305480957, + "learning_rate": 0.00018903960802880485, + "loss": 0.8865, + "step": 10141 + }, + { + "epoch": 0.7037679550343487, + "grad_norm": 1.767880916595459, + "learning_rate": 0.0001890370503699173, + "loss": 0.5915, + "step": 10142 + }, + { + "epoch": 0.7038373464714455, + "grad_norm": 2.174520969390869, + "learning_rate": 0.00018903449242995012, + "loss": 0.5843, + "step": 10143 + }, + { + "epoch": 0.7039067379085421, + "grad_norm": 2.245176315307617, + "learning_rate": 0.00018903193420891136, + "loss": 0.7693, + "step": 10144 + }, + { + "epoch": 0.7039761293456388, + "grad_norm": 2.506357431411743, + "learning_rate": 0.00018902937570680912, + "loss": 0.8377, + "step": 10145 + }, + { + "epoch": 0.7040455207827354, + "grad_norm": 2.337852716445923, + "learning_rate": 0.0001890268169236515, + "loss": 0.9136, + "step": 10146 + }, + { + "epoch": 0.704114912219832, + "grad_norm": 2.2075018882751465, + "learning_rate": 0.0001890242578594465, + "loss": 0.9648, + "step": 10147 + }, + { + "epoch": 0.7041843036569287, + "grad_norm": 3.3957395553588867, + "learning_rate": 0.00018902169851420228, + "loss": 0.7469, + "step": 10148 + }, + { + "epoch": 0.7042536950940254, + "grad_norm": 2.1637113094329834, + "learning_rate": 0.0001890191388879269, + "loss": 0.7764, + "step": 10149 + }, + { + "epoch": 0.7043230865311221, + "grad_norm": 2.231940984725952, + "learning_rate": 0.00018901657898062838, + "loss": 0.8664, + "step": 10150 + }, + { + "epoch": 0.7043924779682187, + "grad_norm": 1.763419508934021, + "learning_rate": 0.0001890140187923149, + "loss": 0.593, + "step": 10151 + }, + { + "epoch": 0.7044618694053154, + "grad_norm": 2.2891879081726074, + "learning_rate": 0.00018901145832299446, + "loss": 0.6756, + "step": 10152 + }, + { + "epoch": 0.704531260842412, + "grad_norm": 2.9185636043548584, + "learning_rate": 0.00018900889757267515, + "loss": 0.8728, + "step": 10153 + }, + { + "epoch": 0.7046006522795087, + "grad_norm": 1.8165366649627686, + "learning_rate": 0.00018900633654136511, + "loss": 0.6756, + "step": 10154 + }, + { + "epoch": 0.7046700437166054, + "grad_norm": 2.094989538192749, + "learning_rate": 0.0001890037752290724, + "loss": 0.7005, + "step": 10155 + }, + { + "epoch": 0.7047394351537021, + "grad_norm": 3.150071859359741, + "learning_rate": 0.00018900121363580507, + "loss": 0.9382, + "step": 10156 + }, + { + "epoch": 0.7048088265907987, + "grad_norm": 2.440859317779541, + "learning_rate": 0.00018899865176157126, + "loss": 0.6608, + "step": 10157 + }, + { + "epoch": 0.7048782180278954, + "grad_norm": 2.5496716499328613, + "learning_rate": 0.00018899608960637902, + "loss": 0.8606, + "step": 10158 + }, + { + "epoch": 0.704947609464992, + "grad_norm": 2.781778573989868, + "learning_rate": 0.00018899352717023643, + "loss": 1.0361, + "step": 10159 + }, + { + "epoch": 0.7050170009020886, + "grad_norm": 2.6056740283966064, + "learning_rate": 0.00018899096445315164, + "loss": 0.9102, + "step": 10160 + }, + { + "epoch": 0.7050863923391854, + "grad_norm": 2.3029825687408447, + "learning_rate": 0.0001889884014551327, + "loss": 0.8678, + "step": 10161 + }, + { + "epoch": 0.705155783776282, + "grad_norm": 2.2536416053771973, + "learning_rate": 0.00018898583817618766, + "loss": 0.814, + "step": 10162 + }, + { + "epoch": 0.7052251752133787, + "grad_norm": 3.200568199157715, + "learning_rate": 0.00018898327461632465, + "loss": 1.0896, + "step": 10163 + }, + { + "epoch": 0.7052945666504753, + "grad_norm": 2.660940170288086, + "learning_rate": 0.00018898071077555183, + "loss": 0.9304, + "step": 10164 + }, + { + "epoch": 0.705363958087572, + "grad_norm": 2.352841854095459, + "learning_rate": 0.00018897814665387719, + "loss": 0.7769, + "step": 10165 + }, + { + "epoch": 0.7054333495246686, + "grad_norm": 2.4909284114837646, + "learning_rate": 0.00018897558225130885, + "loss": 0.9612, + "step": 10166 + }, + { + "epoch": 0.7055027409617654, + "grad_norm": 1.9584496021270752, + "learning_rate": 0.00018897301756785493, + "loss": 0.9255, + "step": 10167 + }, + { + "epoch": 0.705572132398862, + "grad_norm": 1.9349738359451294, + "learning_rate": 0.00018897045260352354, + "loss": 0.8156, + "step": 10168 + }, + { + "epoch": 0.7056415238359587, + "grad_norm": 2.2489171028137207, + "learning_rate": 0.0001889678873583227, + "loss": 0.9562, + "step": 10169 + }, + { + "epoch": 0.7057109152730553, + "grad_norm": 3.0274219512939453, + "learning_rate": 0.00018896532183226063, + "loss": 0.7778, + "step": 10170 + }, + { + "epoch": 0.705780306710152, + "grad_norm": 2.914327621459961, + "learning_rate": 0.0001889627560253453, + "loss": 1.1788, + "step": 10171 + }, + { + "epoch": 0.7058496981472486, + "grad_norm": 2.1094377040863037, + "learning_rate": 0.0001889601899375849, + "loss": 0.7071, + "step": 10172 + }, + { + "epoch": 0.7059190895843452, + "grad_norm": 2.6285948753356934, + "learning_rate": 0.00018895762356898753, + "loss": 0.9789, + "step": 10173 + }, + { + "epoch": 0.705988481021442, + "grad_norm": 2.102358818054199, + "learning_rate": 0.00018895505691956122, + "loss": 0.8694, + "step": 10174 + }, + { + "epoch": 0.7060578724585386, + "grad_norm": 2.509817123413086, + "learning_rate": 0.0001889524899893141, + "loss": 0.8786, + "step": 10175 + }, + { + "epoch": 0.7061272638956353, + "grad_norm": 2.57157826423645, + "learning_rate": 0.00018894992277825433, + "loss": 0.992, + "step": 10176 + }, + { + "epoch": 0.7061966553327319, + "grad_norm": 2.4955689907073975, + "learning_rate": 0.00018894735528638998, + "loss": 0.7393, + "step": 10177 + }, + { + "epoch": 0.7062660467698286, + "grad_norm": 2.070044755935669, + "learning_rate": 0.00018894478751372914, + "loss": 0.9881, + "step": 10178 + }, + { + "epoch": 0.7063354382069252, + "grad_norm": 2.723144054412842, + "learning_rate": 0.0001889422194602799, + "loss": 0.9318, + "step": 10179 + }, + { + "epoch": 0.706404829644022, + "grad_norm": 2.8034446239471436, + "learning_rate": 0.00018893965112605043, + "loss": 0.6518, + "step": 10180 + }, + { + "epoch": 0.7064742210811186, + "grad_norm": 1.8596535921096802, + "learning_rate": 0.00018893708251104874, + "loss": 0.6541, + "step": 10181 + }, + { + "epoch": 0.7065436125182153, + "grad_norm": 2.517070770263672, + "learning_rate": 0.00018893451361528307, + "loss": 1.0213, + "step": 10182 + }, + { + "epoch": 0.7066130039553119, + "grad_norm": 2.5693585872650146, + "learning_rate": 0.00018893194443876142, + "loss": 1.0527, + "step": 10183 + }, + { + "epoch": 0.7066823953924086, + "grad_norm": 2.496985912322998, + "learning_rate": 0.00018892937498149194, + "loss": 0.7702, + "step": 10184 + }, + { + "epoch": 0.7067517868295052, + "grad_norm": 2.1392242908477783, + "learning_rate": 0.00018892680524348277, + "loss": 0.8061, + "step": 10185 + }, + { + "epoch": 0.7068211782666018, + "grad_norm": 2.4767587184906006, + "learning_rate": 0.00018892423522474196, + "loss": 0.942, + "step": 10186 + }, + { + "epoch": 0.7068905697036986, + "grad_norm": 2.714994430541992, + "learning_rate": 0.00018892166492527767, + "loss": 0.9604, + "step": 10187 + }, + { + "epoch": 0.7069599611407952, + "grad_norm": 2.766011953353882, + "learning_rate": 0.000188919094345098, + "loss": 0.781, + "step": 10188 + }, + { + "epoch": 0.7070293525778919, + "grad_norm": 2.2767999172210693, + "learning_rate": 0.00018891652348421105, + "loss": 0.7317, + "step": 10189 + }, + { + "epoch": 0.7070987440149885, + "grad_norm": 2.9345366954803467, + "learning_rate": 0.00018891395234262498, + "loss": 0.9164, + "step": 10190 + }, + { + "epoch": 0.7071681354520852, + "grad_norm": 2.7752974033355713, + "learning_rate": 0.00018891138092034786, + "loss": 0.9032, + "step": 10191 + }, + { + "epoch": 0.7072375268891818, + "grad_norm": 3.6461598873138428, + "learning_rate": 0.00018890880921738784, + "loss": 0.9216, + "step": 10192 + }, + { + "epoch": 0.7073069183262786, + "grad_norm": 2.2259650230407715, + "learning_rate": 0.000188906237233753, + "loss": 0.9107, + "step": 10193 + }, + { + "epoch": 0.7073763097633752, + "grad_norm": 1.966029405593872, + "learning_rate": 0.00018890366496945152, + "loss": 0.672, + "step": 10194 + }, + { + "epoch": 0.7074457012004719, + "grad_norm": 2.036689281463623, + "learning_rate": 0.00018890109242449145, + "loss": 0.8867, + "step": 10195 + }, + { + "epoch": 0.7075150926375685, + "grad_norm": 2.14591383934021, + "learning_rate": 0.00018889851959888096, + "loss": 0.8436, + "step": 10196 + }, + { + "epoch": 0.7075844840746652, + "grad_norm": 3.4763429164886475, + "learning_rate": 0.00018889594649262813, + "loss": 1.0333, + "step": 10197 + }, + { + "epoch": 0.7076538755117618, + "grad_norm": 2.1010656356811523, + "learning_rate": 0.0001888933731057412, + "loss": 0.9248, + "step": 10198 + }, + { + "epoch": 0.7077232669488586, + "grad_norm": 6.534814834594727, + "learning_rate": 0.00018889079943822812, + "loss": 0.9747, + "step": 10199 + }, + { + "epoch": 0.7077926583859552, + "grad_norm": 4.916985034942627, + "learning_rate": 0.0001888882254900971, + "loss": 0.7716, + "step": 10200 + }, + { + "epoch": 0.7078620498230518, + "grad_norm": 2.7625527381896973, + "learning_rate": 0.00018888565126135628, + "loss": 0.8126, + "step": 10201 + }, + { + "epoch": 0.7079314412601485, + "grad_norm": 3.4290058612823486, + "learning_rate": 0.00018888307675201378, + "loss": 0.9973, + "step": 10202 + }, + { + "epoch": 0.7080008326972451, + "grad_norm": 2.6235828399658203, + "learning_rate": 0.00018888050196207775, + "loss": 1.077, + "step": 10203 + }, + { + "epoch": 0.7080702241343418, + "grad_norm": 3.532877206802368, + "learning_rate": 0.00018887792689155624, + "loss": 0.8653, + "step": 10204 + }, + { + "epoch": 0.7081396155714385, + "grad_norm": 2.224785089492798, + "learning_rate": 0.00018887535154045744, + "loss": 0.9204, + "step": 10205 + }, + { + "epoch": 0.7082090070085352, + "grad_norm": 2.1573023796081543, + "learning_rate": 0.00018887277590878948, + "loss": 1.0235, + "step": 10206 + }, + { + "epoch": 0.7082783984456318, + "grad_norm": 3.5133910179138184, + "learning_rate": 0.00018887019999656044, + "loss": 0.8091, + "step": 10207 + }, + { + "epoch": 0.7083477898827285, + "grad_norm": 3.4118924140930176, + "learning_rate": 0.00018886762380377855, + "loss": 0.9527, + "step": 10208 + }, + { + "epoch": 0.7084171813198251, + "grad_norm": 2.8452670574188232, + "learning_rate": 0.00018886504733045182, + "loss": 1.0141, + "step": 10209 + }, + { + "epoch": 0.7084865727569218, + "grad_norm": 2.337085723876953, + "learning_rate": 0.00018886247057658847, + "loss": 0.8576, + "step": 10210 + }, + { + "epoch": 0.7085559641940185, + "grad_norm": 2.961966037750244, + "learning_rate": 0.00018885989354219663, + "loss": 1.0449, + "step": 10211 + }, + { + "epoch": 0.7086253556311152, + "grad_norm": 2.875492572784424, + "learning_rate": 0.0001888573162272844, + "loss": 0.8814, + "step": 10212 + }, + { + "epoch": 0.7086947470682118, + "grad_norm": 2.9656689167022705, + "learning_rate": 0.00018885473863185993, + "loss": 0.7948, + "step": 10213 + }, + { + "epoch": 0.7087641385053084, + "grad_norm": 2.820841073989868, + "learning_rate": 0.00018885216075593134, + "loss": 0.8704, + "step": 10214 + }, + { + "epoch": 0.7088335299424051, + "grad_norm": 2.560490369796753, + "learning_rate": 0.00018884958259950683, + "loss": 0.8434, + "step": 10215 + }, + { + "epoch": 0.7089029213795017, + "grad_norm": 4.0819902420043945, + "learning_rate": 0.00018884700416259447, + "loss": 0.7979, + "step": 10216 + }, + { + "epoch": 0.7089723128165984, + "grad_norm": 4.062870979309082, + "learning_rate": 0.00018884442544520242, + "loss": 0.712, + "step": 10217 + }, + { + "epoch": 0.7090417042536951, + "grad_norm": 2.28729248046875, + "learning_rate": 0.00018884184644733885, + "loss": 0.8651, + "step": 10218 + }, + { + "epoch": 0.7091110956907918, + "grad_norm": 2.5530378818511963, + "learning_rate": 0.00018883926716901185, + "loss": 0.942, + "step": 10219 + }, + { + "epoch": 0.7091804871278884, + "grad_norm": 3.1534664630889893, + "learning_rate": 0.0001888366876102296, + "loss": 0.8977, + "step": 10220 + }, + { + "epoch": 0.7092498785649851, + "grad_norm": 4.217967510223389, + "learning_rate": 0.00018883410777100027, + "loss": 1.1298, + "step": 10221 + }, + { + "epoch": 0.7093192700020817, + "grad_norm": 3.1148829460144043, + "learning_rate": 0.00018883152765133192, + "loss": 0.877, + "step": 10222 + }, + { + "epoch": 0.7093886614391784, + "grad_norm": 3.071357488632202, + "learning_rate": 0.00018882894725123276, + "loss": 1.2157, + "step": 10223 + }, + { + "epoch": 0.7094580528762751, + "grad_norm": 2.179337739944458, + "learning_rate": 0.00018882636657071094, + "loss": 0.7068, + "step": 10224 + }, + { + "epoch": 0.7095274443133718, + "grad_norm": 2.5971486568450928, + "learning_rate": 0.00018882378560977454, + "loss": 0.8764, + "step": 10225 + }, + { + "epoch": 0.7095968357504684, + "grad_norm": 2.4747397899627686, + "learning_rate": 0.00018882120436843178, + "loss": 0.8099, + "step": 10226 + }, + { + "epoch": 0.709666227187565, + "grad_norm": 3.3650803565979004, + "learning_rate": 0.0001888186228466908, + "loss": 0.9995, + "step": 10227 + }, + { + "epoch": 0.7097356186246617, + "grad_norm": 2.9997010231018066, + "learning_rate": 0.00018881604104455972, + "loss": 1.1939, + "step": 10228 + }, + { + "epoch": 0.7098050100617583, + "grad_norm": 3.219120740890503, + "learning_rate": 0.0001888134589620467, + "loss": 1.1525, + "step": 10229 + }, + { + "epoch": 0.7098744014988551, + "grad_norm": 2.4874331951141357, + "learning_rate": 0.0001888108765991599, + "loss": 0.7224, + "step": 10230 + }, + { + "epoch": 0.7099437929359517, + "grad_norm": 1.941628336906433, + "learning_rate": 0.00018880829395590744, + "loss": 0.6689, + "step": 10231 + }, + { + "epoch": 0.7100131843730484, + "grad_norm": 3.2791836261749268, + "learning_rate": 0.00018880571103229753, + "loss": 1.0568, + "step": 10232 + }, + { + "epoch": 0.710082575810145, + "grad_norm": 4.346436977386475, + "learning_rate": 0.0001888031278283383, + "loss": 0.9906, + "step": 10233 + }, + { + "epoch": 0.7101519672472417, + "grad_norm": 3.157073497772217, + "learning_rate": 0.0001888005443440379, + "loss": 0.9025, + "step": 10234 + }, + { + "epoch": 0.7102213586843383, + "grad_norm": 2.384796619415283, + "learning_rate": 0.00018879796057940445, + "loss": 0.7256, + "step": 10235 + }, + { + "epoch": 0.710290750121435, + "grad_norm": 2.6097705364227295, + "learning_rate": 0.00018879537653444616, + "loss": 1.0929, + "step": 10236 + }, + { + "epoch": 0.7103601415585317, + "grad_norm": 2.923435926437378, + "learning_rate": 0.00018879279220917117, + "loss": 1.0955, + "step": 10237 + }, + { + "epoch": 0.7104295329956284, + "grad_norm": 2.0359275341033936, + "learning_rate": 0.00018879020760358763, + "loss": 0.807, + "step": 10238 + }, + { + "epoch": 0.710498924432725, + "grad_norm": 2.6410365104675293, + "learning_rate": 0.0001887876227177037, + "loss": 1.0216, + "step": 10239 + }, + { + "epoch": 0.7105683158698216, + "grad_norm": 2.3025476932525635, + "learning_rate": 0.00018878503755152758, + "loss": 0.5735, + "step": 10240 + }, + { + "epoch": 0.7106377073069183, + "grad_norm": 3.116593360900879, + "learning_rate": 0.00018878245210506736, + "loss": 0.8418, + "step": 10241 + }, + { + "epoch": 0.7107070987440149, + "grad_norm": 2.27071475982666, + "learning_rate": 0.00018877986637833125, + "loss": 0.9043, + "step": 10242 + }, + { + "epoch": 0.7107764901811117, + "grad_norm": 2.6857287883758545, + "learning_rate": 0.0001887772803713274, + "loss": 1.0217, + "step": 10243 + }, + { + "epoch": 0.7108458816182083, + "grad_norm": 2.4346208572387695, + "learning_rate": 0.000188774694084064, + "loss": 0.8018, + "step": 10244 + }, + { + "epoch": 0.710915273055305, + "grad_norm": 1.8582597970962524, + "learning_rate": 0.00018877210751654916, + "loss": 0.7925, + "step": 10245 + }, + { + "epoch": 0.7109846644924016, + "grad_norm": 2.426751136779785, + "learning_rate": 0.00018876952066879108, + "loss": 0.8983, + "step": 10246 + }, + { + "epoch": 0.7110540559294983, + "grad_norm": 2.502028465270996, + "learning_rate": 0.00018876693354079796, + "loss": 0.9053, + "step": 10247 + }, + { + "epoch": 0.7111234473665949, + "grad_norm": 2.142763614654541, + "learning_rate": 0.00018876434613257788, + "loss": 0.8553, + "step": 10248 + }, + { + "epoch": 0.7111928388036917, + "grad_norm": 2.629693031311035, + "learning_rate": 0.00018876175844413908, + "loss": 0.7314, + "step": 10249 + }, + { + "epoch": 0.7112622302407883, + "grad_norm": 2.5363190174102783, + "learning_rate": 0.0001887591704754897, + "loss": 1.0167, + "step": 10250 + }, + { + "epoch": 0.711331621677885, + "grad_norm": 2.3232946395874023, + "learning_rate": 0.00018875658222663793, + "loss": 0.9281, + "step": 10251 + }, + { + "epoch": 0.7114010131149816, + "grad_norm": 2.210078239440918, + "learning_rate": 0.0001887539936975919, + "loss": 0.7826, + "step": 10252 + }, + { + "epoch": 0.7114704045520783, + "grad_norm": 2.0513362884521484, + "learning_rate": 0.00018875140488835984, + "loss": 0.7701, + "step": 10253 + }, + { + "epoch": 0.7115397959891749, + "grad_norm": 2.8517794609069824, + "learning_rate": 0.00018874881579894988, + "loss": 0.8384, + "step": 10254 + }, + { + "epoch": 0.7116091874262715, + "grad_norm": 2.409925699234009, + "learning_rate": 0.0001887462264293702, + "loss": 0.7289, + "step": 10255 + }, + { + "epoch": 0.7116785788633683, + "grad_norm": 1.9953595399856567, + "learning_rate": 0.000188743636779629, + "loss": 0.762, + "step": 10256 + }, + { + "epoch": 0.7117479703004649, + "grad_norm": 2.5445234775543213, + "learning_rate": 0.0001887410468497344, + "loss": 1.0875, + "step": 10257 + }, + { + "epoch": 0.7118173617375616, + "grad_norm": 2.8775858879089355, + "learning_rate": 0.00018873845663969465, + "loss": 0.8364, + "step": 10258 + }, + { + "epoch": 0.7118867531746582, + "grad_norm": 2.7306466102600098, + "learning_rate": 0.00018873586614951785, + "loss": 1.0282, + "step": 10259 + }, + { + "epoch": 0.7119561446117549, + "grad_norm": 2.723801851272583, + "learning_rate": 0.00018873327537921223, + "loss": 0.5877, + "step": 10260 + }, + { + "epoch": 0.7120255360488515, + "grad_norm": 2.3214292526245117, + "learning_rate": 0.00018873068432878598, + "loss": 0.8173, + "step": 10261 + }, + { + "epoch": 0.7120949274859483, + "grad_norm": 2.61039137840271, + "learning_rate": 0.00018872809299824722, + "loss": 0.8264, + "step": 10262 + }, + { + "epoch": 0.7121643189230449, + "grad_norm": 2.8913285732269287, + "learning_rate": 0.0001887255013876042, + "loss": 0.7871, + "step": 10263 + }, + { + "epoch": 0.7122337103601416, + "grad_norm": 2.719642162322998, + "learning_rate": 0.00018872290949686505, + "loss": 1.0712, + "step": 10264 + }, + { + "epoch": 0.7123031017972382, + "grad_norm": 2.7330851554870605, + "learning_rate": 0.00018872031732603795, + "loss": 0.8767, + "step": 10265 + }, + { + "epoch": 0.7123724932343349, + "grad_norm": 1.660404086112976, + "learning_rate": 0.0001887177248751311, + "loss": 0.6961, + "step": 10266 + }, + { + "epoch": 0.7124418846714315, + "grad_norm": 1.7069038152694702, + "learning_rate": 0.00018871513214415271, + "loss": 0.6004, + "step": 10267 + }, + { + "epoch": 0.7125112761085282, + "grad_norm": 2.798992395401001, + "learning_rate": 0.00018871253913311094, + "loss": 0.8628, + "step": 10268 + }, + { + "epoch": 0.7125806675456249, + "grad_norm": 2.1872379779815674, + "learning_rate": 0.000188709945842014, + "loss": 0.9422, + "step": 10269 + }, + { + "epoch": 0.7126500589827215, + "grad_norm": 3.2513670921325684, + "learning_rate": 0.00018870735227087, + "loss": 1.2331, + "step": 10270 + }, + { + "epoch": 0.7127194504198182, + "grad_norm": 3.4239134788513184, + "learning_rate": 0.00018870475841968717, + "loss": 1.0337, + "step": 10271 + }, + { + "epoch": 0.7127888418569148, + "grad_norm": 2.1112871170043945, + "learning_rate": 0.00018870216428847376, + "loss": 0.8306, + "step": 10272 + }, + { + "epoch": 0.7128582332940115, + "grad_norm": 1.7452455759048462, + "learning_rate": 0.0001886995698772379, + "loss": 0.7338, + "step": 10273 + }, + { + "epoch": 0.7129276247311082, + "grad_norm": 2.327021598815918, + "learning_rate": 0.00018869697518598776, + "loss": 0.8822, + "step": 10274 + }, + { + "epoch": 0.7129970161682049, + "grad_norm": 2.6589410305023193, + "learning_rate": 0.00018869438021473157, + "loss": 1.0801, + "step": 10275 + }, + { + "epoch": 0.7130664076053015, + "grad_norm": 2.5307796001434326, + "learning_rate": 0.00018869178496347755, + "loss": 0.8134, + "step": 10276 + }, + { + "epoch": 0.7131357990423982, + "grad_norm": 2.1594085693359375, + "learning_rate": 0.0001886891894322338, + "loss": 0.715, + "step": 10277 + }, + { + "epoch": 0.7132051904794948, + "grad_norm": 2.1930768489837646, + "learning_rate": 0.00018868659362100863, + "loss": 0.753, + "step": 10278 + }, + { + "epoch": 0.7132745819165915, + "grad_norm": 2.7522268295288086, + "learning_rate": 0.00018868399752981012, + "loss": 0.9601, + "step": 10279 + }, + { + "epoch": 0.7133439733536882, + "grad_norm": 2.543375253677368, + "learning_rate": 0.00018868140115864654, + "loss": 0.7483, + "step": 10280 + }, + { + "epoch": 0.7134133647907848, + "grad_norm": 2.1664037704467773, + "learning_rate": 0.00018867880450752607, + "loss": 0.7648, + "step": 10281 + }, + { + "epoch": 0.7134827562278815, + "grad_norm": 2.372450828552246, + "learning_rate": 0.0001886762075764569, + "loss": 0.8269, + "step": 10282 + }, + { + "epoch": 0.7135521476649781, + "grad_norm": 2.424628496170044, + "learning_rate": 0.00018867361036544722, + "loss": 0.9441, + "step": 10283 + }, + { + "epoch": 0.7136215391020748, + "grad_norm": 3.395388126373291, + "learning_rate": 0.00018867101287450524, + "loss": 1.0867, + "step": 10284 + }, + { + "epoch": 0.7136909305391714, + "grad_norm": 2.8174397945404053, + "learning_rate": 0.00018866841510363917, + "loss": 0.9018, + "step": 10285 + }, + { + "epoch": 0.7137603219762682, + "grad_norm": 3.144233226776123, + "learning_rate": 0.0001886658170528572, + "loss": 0.9506, + "step": 10286 + }, + { + "epoch": 0.7138297134133648, + "grad_norm": 2.8576717376708984, + "learning_rate": 0.0001886632187221675, + "loss": 0.991, + "step": 10287 + }, + { + "epoch": 0.7138991048504615, + "grad_norm": 2.935856342315674, + "learning_rate": 0.00018866062011157838, + "loss": 0.956, + "step": 10288 + }, + { + "epoch": 0.7139684962875581, + "grad_norm": 2.865290880203247, + "learning_rate": 0.0001886580212210979, + "loss": 0.8849, + "step": 10289 + }, + { + "epoch": 0.7140378877246548, + "grad_norm": 2.1945478916168213, + "learning_rate": 0.00018865542205073434, + "loss": 0.8129, + "step": 10290 + }, + { + "epoch": 0.7141072791617514, + "grad_norm": 3.3860743045806885, + "learning_rate": 0.00018865282260049595, + "loss": 0.7585, + "step": 10291 + }, + { + "epoch": 0.7141766705988482, + "grad_norm": 2.093874454498291, + "learning_rate": 0.00018865022287039085, + "loss": 0.8207, + "step": 10292 + }, + { + "epoch": 0.7142460620359448, + "grad_norm": 2.4538121223449707, + "learning_rate": 0.00018864762286042728, + "loss": 0.8371, + "step": 10293 + }, + { + "epoch": 0.7143154534730414, + "grad_norm": 2.1578681468963623, + "learning_rate": 0.00018864502257061346, + "loss": 0.9203, + "step": 10294 + }, + { + "epoch": 0.7143848449101381, + "grad_norm": 2.2255828380584717, + "learning_rate": 0.00018864242200095758, + "loss": 0.9448, + "step": 10295 + }, + { + "epoch": 0.7144542363472347, + "grad_norm": 2.3790998458862305, + "learning_rate": 0.00018863982115146786, + "loss": 0.9721, + "step": 10296 + }, + { + "epoch": 0.7145236277843314, + "grad_norm": 2.3184821605682373, + "learning_rate": 0.0001886372200221525, + "loss": 0.9199, + "step": 10297 + }, + { + "epoch": 0.714593019221428, + "grad_norm": 2.6311323642730713, + "learning_rate": 0.00018863461861301972, + "loss": 0.8673, + "step": 10298 + }, + { + "epoch": 0.7146624106585248, + "grad_norm": 3.4447803497314453, + "learning_rate": 0.00018863201692407775, + "loss": 0.9451, + "step": 10299 + }, + { + "epoch": 0.7147318020956214, + "grad_norm": 2.5737032890319824, + "learning_rate": 0.00018862941495533477, + "loss": 0.8405, + "step": 10300 + }, + { + "epoch": 0.7148011935327181, + "grad_norm": 1.9594606161117554, + "learning_rate": 0.00018862681270679903, + "loss": 0.773, + "step": 10301 + }, + { + "epoch": 0.7148705849698147, + "grad_norm": 2.529153347015381, + "learning_rate": 0.0001886242101784787, + "loss": 1.0262, + "step": 10302 + }, + { + "epoch": 0.7149399764069114, + "grad_norm": 2.2259809970855713, + "learning_rate": 0.00018862160737038202, + "loss": 0.6672, + "step": 10303 + }, + { + "epoch": 0.715009367844008, + "grad_norm": 3.060230016708374, + "learning_rate": 0.00018861900428251724, + "loss": 1.1828, + "step": 10304 + }, + { + "epoch": 0.7150787592811048, + "grad_norm": 2.4983019828796387, + "learning_rate": 0.00018861640091489252, + "loss": 0.9641, + "step": 10305 + }, + { + "epoch": 0.7151481507182014, + "grad_norm": 1.960315465927124, + "learning_rate": 0.0001886137972675161, + "loss": 0.7172, + "step": 10306 + }, + { + "epoch": 0.7152175421552981, + "grad_norm": 2.024799108505249, + "learning_rate": 0.0001886111933403962, + "loss": 0.7382, + "step": 10307 + }, + { + "epoch": 0.7152869335923947, + "grad_norm": 2.361042022705078, + "learning_rate": 0.00018860858913354107, + "loss": 1.0145, + "step": 10308 + }, + { + "epoch": 0.7153563250294913, + "grad_norm": 2.858431816101074, + "learning_rate": 0.00018860598464695889, + "loss": 1.0123, + "step": 10309 + }, + { + "epoch": 0.715425716466588, + "grad_norm": 2.6097230911254883, + "learning_rate": 0.0001886033798806579, + "loss": 0.9099, + "step": 10310 + }, + { + "epoch": 0.7154951079036846, + "grad_norm": 3.17368745803833, + "learning_rate": 0.00018860077483464632, + "loss": 0.9398, + "step": 10311 + }, + { + "epoch": 0.7155644993407814, + "grad_norm": 2.574333667755127, + "learning_rate": 0.00018859816950893236, + "loss": 0.9714, + "step": 10312 + }, + { + "epoch": 0.715633890777878, + "grad_norm": 2.208707809448242, + "learning_rate": 0.00018859556390352425, + "loss": 0.9635, + "step": 10313 + }, + { + "epoch": 0.7157032822149747, + "grad_norm": 2.379523277282715, + "learning_rate": 0.00018859295801843025, + "loss": 0.8332, + "step": 10314 + }, + { + "epoch": 0.7157726736520713, + "grad_norm": 2.939842939376831, + "learning_rate": 0.00018859035185365858, + "loss": 0.9055, + "step": 10315 + }, + { + "epoch": 0.715842065089168, + "grad_norm": 2.3691375255584717, + "learning_rate": 0.00018858774540921739, + "loss": 0.895, + "step": 10316 + }, + { + "epoch": 0.7159114565262646, + "grad_norm": 3.971944808959961, + "learning_rate": 0.000188585138685115, + "loss": 0.9116, + "step": 10317 + }, + { + "epoch": 0.7159808479633614, + "grad_norm": 2.414233922958374, + "learning_rate": 0.0001885825316813596, + "loss": 1.0126, + "step": 10318 + }, + { + "epoch": 0.716050239400458, + "grad_norm": 2.3990087509155273, + "learning_rate": 0.00018857992439795944, + "loss": 0.6189, + "step": 10319 + }, + { + "epoch": 0.7161196308375547, + "grad_norm": 2.096025228500366, + "learning_rate": 0.0001885773168349227, + "loss": 0.8998, + "step": 10320 + }, + { + "epoch": 0.7161890222746513, + "grad_norm": 2.746779203414917, + "learning_rate": 0.0001885747089922577, + "loss": 1.0988, + "step": 10321 + }, + { + "epoch": 0.7162584137117479, + "grad_norm": 2.786673069000244, + "learning_rate": 0.00018857210086997255, + "loss": 0.9205, + "step": 10322 + }, + { + "epoch": 0.7163278051488446, + "grad_norm": 2.758118152618408, + "learning_rate": 0.0001885694924680756, + "loss": 0.9313, + "step": 10323 + }, + { + "epoch": 0.7163971965859413, + "grad_norm": 2.1600019931793213, + "learning_rate": 0.00018856688378657505, + "loss": 0.8224, + "step": 10324 + }, + { + "epoch": 0.716466588023038, + "grad_norm": 2.1863667964935303, + "learning_rate": 0.00018856427482547908, + "loss": 0.9875, + "step": 10325 + }, + { + "epoch": 0.7165359794601346, + "grad_norm": 2.1096041202545166, + "learning_rate": 0.00018856166558479603, + "loss": 0.8062, + "step": 10326 + }, + { + "epoch": 0.7166053708972313, + "grad_norm": 2.392125368118286, + "learning_rate": 0.00018855905606453402, + "loss": 0.6596, + "step": 10327 + }, + { + "epoch": 0.7166747623343279, + "grad_norm": 2.108473539352417, + "learning_rate": 0.00018855644626470137, + "loss": 0.6869, + "step": 10328 + }, + { + "epoch": 0.7167441537714246, + "grad_norm": 3.2082645893096924, + "learning_rate": 0.0001885538361853063, + "loss": 1.1272, + "step": 10329 + }, + { + "epoch": 0.7168135452085213, + "grad_norm": 2.8682565689086914, + "learning_rate": 0.000188551225826357, + "loss": 0.9055, + "step": 10330 + }, + { + "epoch": 0.716882936645618, + "grad_norm": 2.796257495880127, + "learning_rate": 0.0001885486151878618, + "loss": 0.962, + "step": 10331 + }, + { + "epoch": 0.7169523280827146, + "grad_norm": 2.359484910964966, + "learning_rate": 0.0001885460042698289, + "loss": 0.7168, + "step": 10332 + }, + { + "epoch": 0.7170217195198113, + "grad_norm": 3.0420124530792236, + "learning_rate": 0.00018854339307226652, + "loss": 0.7921, + "step": 10333 + }, + { + "epoch": 0.7170911109569079, + "grad_norm": 2.2730674743652344, + "learning_rate": 0.00018854078159518292, + "loss": 0.8717, + "step": 10334 + }, + { + "epoch": 0.7171605023940045, + "grad_norm": 2.45089054107666, + "learning_rate": 0.0001885381698385863, + "loss": 0.5066, + "step": 10335 + }, + { + "epoch": 0.7172298938311013, + "grad_norm": 2.4787845611572266, + "learning_rate": 0.00018853555780248504, + "loss": 1.0267, + "step": 10336 + }, + { + "epoch": 0.7172992852681979, + "grad_norm": 2.7170300483703613, + "learning_rate": 0.00018853294548688726, + "loss": 0.7884, + "step": 10337 + }, + { + "epoch": 0.7173686767052946, + "grad_norm": 2.5046608448028564, + "learning_rate": 0.0001885303328918012, + "loss": 0.9062, + "step": 10338 + }, + { + "epoch": 0.7174380681423912, + "grad_norm": 1.9548014402389526, + "learning_rate": 0.0001885277200172352, + "loss": 0.7094, + "step": 10339 + }, + { + "epoch": 0.7175074595794879, + "grad_norm": 2.0009541511535645, + "learning_rate": 0.00018852510686319745, + "loss": 0.7309, + "step": 10340 + }, + { + "epoch": 0.7175768510165845, + "grad_norm": 2.2408294677734375, + "learning_rate": 0.0001885224934296962, + "loss": 0.9447, + "step": 10341 + }, + { + "epoch": 0.7176462424536812, + "grad_norm": 2.170668840408325, + "learning_rate": 0.0001885198797167397, + "loss": 0.7976, + "step": 10342 + }, + { + "epoch": 0.7177156338907779, + "grad_norm": 1.9438990354537964, + "learning_rate": 0.00018851726572433622, + "loss": 0.8034, + "step": 10343 + }, + { + "epoch": 0.7177850253278746, + "grad_norm": 2.581753969192505, + "learning_rate": 0.00018851465145249403, + "loss": 0.9183, + "step": 10344 + }, + { + "epoch": 0.7178544167649712, + "grad_norm": 5.93079948425293, + "learning_rate": 0.0001885120369012213, + "loss": 0.9953, + "step": 10345 + }, + { + "epoch": 0.7179238082020679, + "grad_norm": 2.6743545532226562, + "learning_rate": 0.00018850942207052638, + "loss": 1.0731, + "step": 10346 + }, + { + "epoch": 0.7179931996391645, + "grad_norm": 2.642860174179077, + "learning_rate": 0.00018850680696041747, + "loss": 0.9346, + "step": 10347 + }, + { + "epoch": 0.7180625910762611, + "grad_norm": 2.4015285968780518, + "learning_rate": 0.00018850419157090284, + "loss": 0.9591, + "step": 10348 + }, + { + "epoch": 0.7181319825133579, + "grad_norm": 3.2153940200805664, + "learning_rate": 0.00018850157590199074, + "loss": 0.887, + "step": 10349 + }, + { + "epoch": 0.7182013739504545, + "grad_norm": 2.897411823272705, + "learning_rate": 0.00018849895995368945, + "loss": 0.8953, + "step": 10350 + }, + { + "epoch": 0.7182707653875512, + "grad_norm": 2.154989719390869, + "learning_rate": 0.0001884963437260072, + "loss": 0.8847, + "step": 10351 + }, + { + "epoch": 0.7183401568246478, + "grad_norm": 2.404021978378296, + "learning_rate": 0.00018849372721895224, + "loss": 0.8568, + "step": 10352 + }, + { + "epoch": 0.7184095482617445, + "grad_norm": 2.3069591522216797, + "learning_rate": 0.00018849111043253288, + "loss": 0.9296, + "step": 10353 + }, + { + "epoch": 0.7184789396988411, + "grad_norm": 2.449571132659912, + "learning_rate": 0.00018848849336675734, + "loss": 0.7076, + "step": 10354 + }, + { + "epoch": 0.7185483311359379, + "grad_norm": 2.8628594875335693, + "learning_rate": 0.0001884858760216339, + "loss": 0.9463, + "step": 10355 + }, + { + "epoch": 0.7186177225730345, + "grad_norm": 3.3321709632873535, + "learning_rate": 0.0001884832583971708, + "loss": 0.8962, + "step": 10356 + }, + { + "epoch": 0.7186871140101312, + "grad_norm": 2.9123599529266357, + "learning_rate": 0.00018848064049337634, + "loss": 0.6562, + "step": 10357 + }, + { + "epoch": 0.7187565054472278, + "grad_norm": 2.6321494579315186, + "learning_rate": 0.00018847802231025876, + "loss": 1.1374, + "step": 10358 + }, + { + "epoch": 0.7188258968843245, + "grad_norm": 2.464588165283203, + "learning_rate": 0.0001884754038478263, + "loss": 0.8501, + "step": 10359 + }, + { + "epoch": 0.7188952883214211, + "grad_norm": 1.985077977180481, + "learning_rate": 0.0001884727851060873, + "loss": 0.6941, + "step": 10360 + }, + { + "epoch": 0.7189646797585177, + "grad_norm": 2.925393581390381, + "learning_rate": 0.00018847016608504995, + "loss": 0.9644, + "step": 10361 + }, + { + "epoch": 0.7190340711956145, + "grad_norm": 2.1911892890930176, + "learning_rate": 0.00018846754678472256, + "loss": 0.7683, + "step": 10362 + }, + { + "epoch": 0.7191034626327111, + "grad_norm": 2.4438493251800537, + "learning_rate": 0.00018846492720511336, + "loss": 0.8134, + "step": 10363 + }, + { + "epoch": 0.7191728540698078, + "grad_norm": 2.651085376739502, + "learning_rate": 0.00018846230734623069, + "loss": 0.8391, + "step": 10364 + }, + { + "epoch": 0.7192422455069044, + "grad_norm": 2.394049644470215, + "learning_rate": 0.0001884596872080828, + "loss": 0.8436, + "step": 10365 + }, + { + "epoch": 0.7193116369440011, + "grad_norm": 4.0341105461120605, + "learning_rate": 0.0001884570667906779, + "loss": 1.2023, + "step": 10366 + }, + { + "epoch": 0.7193810283810977, + "grad_norm": 3.1624207496643066, + "learning_rate": 0.0001884544460940243, + "loss": 1.0767, + "step": 10367 + }, + { + "epoch": 0.7194504198181945, + "grad_norm": 2.447112560272217, + "learning_rate": 0.00018845182511813028, + "loss": 0.9071, + "step": 10368 + }, + { + "epoch": 0.7195198112552911, + "grad_norm": 2.8368754386901855, + "learning_rate": 0.00018844920386300413, + "loss": 0.899, + "step": 10369 + }, + { + "epoch": 0.7195892026923878, + "grad_norm": 2.8693792819976807, + "learning_rate": 0.0001884465823286541, + "loss": 0.8268, + "step": 10370 + }, + { + "epoch": 0.7196585941294844, + "grad_norm": 3.6127264499664307, + "learning_rate": 0.00018844396051508844, + "loss": 1.0585, + "step": 10371 + }, + { + "epoch": 0.7197279855665811, + "grad_norm": 2.5204689502716064, + "learning_rate": 0.00018844133842231547, + "loss": 1.0264, + "step": 10372 + }, + { + "epoch": 0.7197973770036777, + "grad_norm": 2.381545066833496, + "learning_rate": 0.00018843871605034348, + "loss": 1.0525, + "step": 10373 + }, + { + "epoch": 0.7198667684407745, + "grad_norm": 1.9473812580108643, + "learning_rate": 0.00018843609339918072, + "loss": 0.6671, + "step": 10374 + }, + { + "epoch": 0.7199361598778711, + "grad_norm": 2.354987621307373, + "learning_rate": 0.00018843347046883544, + "loss": 0.7984, + "step": 10375 + }, + { + "epoch": 0.7200055513149677, + "grad_norm": 2.6425094604492188, + "learning_rate": 0.000188430847259316, + "loss": 1.3839, + "step": 10376 + }, + { + "epoch": 0.7200749427520644, + "grad_norm": 2.4089105129241943, + "learning_rate": 0.0001884282237706306, + "loss": 0.7736, + "step": 10377 + }, + { + "epoch": 0.720144334189161, + "grad_norm": 2.3396496772766113, + "learning_rate": 0.00018842560000278757, + "loss": 0.8433, + "step": 10378 + }, + { + "epoch": 0.7202137256262577, + "grad_norm": 2.514603614807129, + "learning_rate": 0.00018842297595579518, + "loss": 1.061, + "step": 10379 + }, + { + "epoch": 0.7202831170633544, + "grad_norm": 2.7064497470855713, + "learning_rate": 0.0001884203516296617, + "loss": 0.9282, + "step": 10380 + }, + { + "epoch": 0.7203525085004511, + "grad_norm": 2.5603246688842773, + "learning_rate": 0.00018841772702439546, + "loss": 0.8539, + "step": 10381 + }, + { + "epoch": 0.7204218999375477, + "grad_norm": 2.654176712036133, + "learning_rate": 0.0001884151021400047, + "loss": 1.0529, + "step": 10382 + }, + { + "epoch": 0.7204912913746444, + "grad_norm": 1.9748698472976685, + "learning_rate": 0.0001884124769764977, + "loss": 0.7525, + "step": 10383 + }, + { + "epoch": 0.720560682811741, + "grad_norm": 2.821901559829712, + "learning_rate": 0.0001884098515338828, + "loss": 0.758, + "step": 10384 + }, + { + "epoch": 0.7206300742488377, + "grad_norm": 2.222705125808716, + "learning_rate": 0.00018840722581216825, + "loss": 0.9034, + "step": 10385 + }, + { + "epoch": 0.7206994656859343, + "grad_norm": 2.776876449584961, + "learning_rate": 0.00018840459981136234, + "loss": 0.7931, + "step": 10386 + }, + { + "epoch": 0.7207688571230311, + "grad_norm": 2.6425397396087646, + "learning_rate": 0.00018840197353147336, + "loss": 0.8152, + "step": 10387 + }, + { + "epoch": 0.7208382485601277, + "grad_norm": 2.3288254737854004, + "learning_rate": 0.0001883993469725096, + "loss": 0.9947, + "step": 10388 + }, + { + "epoch": 0.7209076399972243, + "grad_norm": 2.5468664169311523, + "learning_rate": 0.00018839672013447938, + "loss": 1.0294, + "step": 10389 + }, + { + "epoch": 0.720977031434321, + "grad_norm": 2.2607226371765137, + "learning_rate": 0.00018839409301739095, + "loss": 0.797, + "step": 10390 + }, + { + "epoch": 0.7210464228714176, + "grad_norm": 2.2080535888671875, + "learning_rate": 0.00018839146562125265, + "loss": 0.7202, + "step": 10391 + }, + { + "epoch": 0.7211158143085143, + "grad_norm": 2.1282167434692383, + "learning_rate": 0.00018838883794607272, + "loss": 0.8383, + "step": 10392 + }, + { + "epoch": 0.721185205745611, + "grad_norm": 2.2036569118499756, + "learning_rate": 0.0001883862099918595, + "loss": 0.763, + "step": 10393 + }, + { + "epoch": 0.7212545971827077, + "grad_norm": 2.552056312561035, + "learning_rate": 0.00018838358175862126, + "loss": 1.0728, + "step": 10394 + }, + { + "epoch": 0.7213239886198043, + "grad_norm": 2.763857841491699, + "learning_rate": 0.0001883809532463663, + "loss": 1.126, + "step": 10395 + }, + { + "epoch": 0.721393380056901, + "grad_norm": 3.711425304412842, + "learning_rate": 0.00018837832445510295, + "loss": 1.0446, + "step": 10396 + }, + { + "epoch": 0.7214627714939976, + "grad_norm": 1.989491581916809, + "learning_rate": 0.00018837569538483946, + "loss": 0.8873, + "step": 10397 + }, + { + "epoch": 0.7215321629310943, + "grad_norm": 1.895813226699829, + "learning_rate": 0.00018837306603558415, + "loss": 0.7504, + "step": 10398 + }, + { + "epoch": 0.721601554368191, + "grad_norm": 2.0928430557250977, + "learning_rate": 0.00018837043640734533, + "loss": 0.7373, + "step": 10399 + }, + { + "epoch": 0.7216709458052877, + "grad_norm": 2.9919545650482178, + "learning_rate": 0.0001883678065001313, + "loss": 1.1868, + "step": 10400 + }, + { + "epoch": 0.7217403372423843, + "grad_norm": 3.0337460041046143, + "learning_rate": 0.00018836517631395035, + "loss": 1.0873, + "step": 10401 + }, + { + "epoch": 0.7218097286794809, + "grad_norm": 2.6179909706115723, + "learning_rate": 0.0001883625458488108, + "loss": 0.9522, + "step": 10402 + }, + { + "epoch": 0.7218791201165776, + "grad_norm": 2.107424259185791, + "learning_rate": 0.00018835991510472095, + "loss": 0.7826, + "step": 10403 + }, + { + "epoch": 0.7219485115536742, + "grad_norm": 3.73696231842041, + "learning_rate": 0.00018835728408168908, + "loss": 1.0807, + "step": 10404 + }, + { + "epoch": 0.722017902990771, + "grad_norm": 2.1193816661834717, + "learning_rate": 0.00018835465277972354, + "loss": 0.7813, + "step": 10405 + }, + { + "epoch": 0.7220872944278676, + "grad_norm": 2.083972454071045, + "learning_rate": 0.0001883520211988326, + "loss": 0.7921, + "step": 10406 + }, + { + "epoch": 0.7221566858649643, + "grad_norm": 2.8568100929260254, + "learning_rate": 0.0001883493893390246, + "loss": 0.968, + "step": 10407 + }, + { + "epoch": 0.7222260773020609, + "grad_norm": 1.979989767074585, + "learning_rate": 0.0001883467572003078, + "loss": 0.7716, + "step": 10408 + }, + { + "epoch": 0.7222954687391576, + "grad_norm": 2.4460573196411133, + "learning_rate": 0.0001883441247826905, + "loss": 0.9558, + "step": 10409 + }, + { + "epoch": 0.7223648601762542, + "grad_norm": 2.777461290359497, + "learning_rate": 0.00018834149208618112, + "loss": 0.8236, + "step": 10410 + }, + { + "epoch": 0.722434251613351, + "grad_norm": 2.0829946994781494, + "learning_rate": 0.00018833885911078788, + "loss": 1.0084, + "step": 10411 + }, + { + "epoch": 0.7225036430504476, + "grad_norm": 3.4812307357788086, + "learning_rate": 0.0001883362258565191, + "loss": 0.9574, + "step": 10412 + }, + { + "epoch": 0.7225730344875443, + "grad_norm": 1.923871397972107, + "learning_rate": 0.0001883335923233831, + "loss": 0.7969, + "step": 10413 + }, + { + "epoch": 0.7226424259246409, + "grad_norm": 2.897956371307373, + "learning_rate": 0.00018833095851138822, + "loss": 0.9473, + "step": 10414 + }, + { + "epoch": 0.7227118173617375, + "grad_norm": 2.2043304443359375, + "learning_rate": 0.00018832832442054273, + "loss": 0.8659, + "step": 10415 + }, + { + "epoch": 0.7227812087988342, + "grad_norm": 1.9579988718032837, + "learning_rate": 0.00018832569005085498, + "loss": 0.8983, + "step": 10416 + }, + { + "epoch": 0.7228506002359308, + "grad_norm": 2.2473678588867188, + "learning_rate": 0.00018832305540233328, + "loss": 0.8603, + "step": 10417 + }, + { + "epoch": 0.7229199916730276, + "grad_norm": 3.128770351409912, + "learning_rate": 0.0001883204204749859, + "loss": 0.8236, + "step": 10418 + }, + { + "epoch": 0.7229893831101242, + "grad_norm": 2.623871326446533, + "learning_rate": 0.00018831778526882123, + "loss": 0.7659, + "step": 10419 + }, + { + "epoch": 0.7230587745472209, + "grad_norm": 1.5508511066436768, + "learning_rate": 0.00018831514978384756, + "loss": 0.6804, + "step": 10420 + }, + { + "epoch": 0.7231281659843175, + "grad_norm": 3.0962138175964355, + "learning_rate": 0.00018831251402007322, + "loss": 0.8891, + "step": 10421 + }, + { + "epoch": 0.7231975574214142, + "grad_norm": 2.687670946121216, + "learning_rate": 0.0001883098779775065, + "loss": 0.8534, + "step": 10422 + }, + { + "epoch": 0.7232669488585108, + "grad_norm": 2.02974534034729, + "learning_rate": 0.00018830724165615578, + "loss": 0.7501, + "step": 10423 + }, + { + "epoch": 0.7233363402956076, + "grad_norm": 2.3606553077697754, + "learning_rate": 0.0001883046050560293, + "loss": 0.9322, + "step": 10424 + }, + { + "epoch": 0.7234057317327042, + "grad_norm": 2.152331590652466, + "learning_rate": 0.00018830196817713543, + "loss": 0.8588, + "step": 10425 + }, + { + "epoch": 0.7234751231698009, + "grad_norm": 2.3545382022857666, + "learning_rate": 0.0001882993310194825, + "loss": 0.7655, + "step": 10426 + }, + { + "epoch": 0.7235445146068975, + "grad_norm": 2.9358294010162354, + "learning_rate": 0.00018829669358307882, + "loss": 0.9687, + "step": 10427 + }, + { + "epoch": 0.7236139060439942, + "grad_norm": 3.38407826423645, + "learning_rate": 0.00018829405586793274, + "loss": 1.0165, + "step": 10428 + }, + { + "epoch": 0.7236832974810908, + "grad_norm": 2.1574604511260986, + "learning_rate": 0.00018829141787405254, + "loss": 0.9225, + "step": 10429 + }, + { + "epoch": 0.7237526889181874, + "grad_norm": 2.5088775157928467, + "learning_rate": 0.00018828877960144663, + "loss": 0.7347, + "step": 10430 + }, + { + "epoch": 0.7238220803552842, + "grad_norm": 2.0556447505950928, + "learning_rate": 0.00018828614105012323, + "loss": 0.8448, + "step": 10431 + }, + { + "epoch": 0.7238914717923808, + "grad_norm": 2.881258964538574, + "learning_rate": 0.00018828350222009074, + "loss": 1.0198, + "step": 10432 + }, + { + "epoch": 0.7239608632294775, + "grad_norm": 2.503751754760742, + "learning_rate": 0.0001882808631113575, + "loss": 0.9567, + "step": 10433 + }, + { + "epoch": 0.7240302546665741, + "grad_norm": 3.214446544647217, + "learning_rate": 0.0001882782237239318, + "loss": 1.0257, + "step": 10434 + }, + { + "epoch": 0.7240996461036708, + "grad_norm": 2.220362424850464, + "learning_rate": 0.000188275584057822, + "loss": 0.6105, + "step": 10435 + }, + { + "epoch": 0.7241690375407674, + "grad_norm": 1.9890518188476562, + "learning_rate": 0.0001882729441130364, + "loss": 0.6897, + "step": 10436 + }, + { + "epoch": 0.7242384289778642, + "grad_norm": 2.8019824028015137, + "learning_rate": 0.00018827030388958338, + "loss": 0.8974, + "step": 10437 + }, + { + "epoch": 0.7243078204149608, + "grad_norm": 2.0648627281188965, + "learning_rate": 0.00018826766338747125, + "loss": 0.7852, + "step": 10438 + }, + { + "epoch": 0.7243772118520575, + "grad_norm": 2.1870157718658447, + "learning_rate": 0.00018826502260670832, + "loss": 0.762, + "step": 10439 + }, + { + "epoch": 0.7244466032891541, + "grad_norm": 1.8166847229003906, + "learning_rate": 0.00018826238154730297, + "loss": 0.7462, + "step": 10440 + }, + { + "epoch": 0.7245159947262508, + "grad_norm": 1.7583272457122803, + "learning_rate": 0.00018825974020926354, + "loss": 0.6602, + "step": 10441 + }, + { + "epoch": 0.7245853861633474, + "grad_norm": 3.1642401218414307, + "learning_rate": 0.00018825709859259836, + "loss": 0.8557, + "step": 10442 + }, + { + "epoch": 0.7246547776004441, + "grad_norm": 2.4259157180786133, + "learning_rate": 0.0001882544566973157, + "loss": 1.0828, + "step": 10443 + }, + { + "epoch": 0.7247241690375408, + "grad_norm": 3.5714633464813232, + "learning_rate": 0.000188251814523424, + "loss": 1.3177, + "step": 10444 + }, + { + "epoch": 0.7247935604746374, + "grad_norm": 2.838019609451294, + "learning_rate": 0.00018824917207093157, + "loss": 1.0247, + "step": 10445 + }, + { + "epoch": 0.7248629519117341, + "grad_norm": 2.9066290855407715, + "learning_rate": 0.00018824652933984674, + "loss": 1.1179, + "step": 10446 + }, + { + "epoch": 0.7249323433488307, + "grad_norm": 2.5926826000213623, + "learning_rate": 0.00018824388633017784, + "loss": 0.9723, + "step": 10447 + }, + { + "epoch": 0.7250017347859274, + "grad_norm": 2.214829206466675, + "learning_rate": 0.00018824124304193322, + "loss": 0.7056, + "step": 10448 + }, + { + "epoch": 0.725071126223024, + "grad_norm": 2.012845516204834, + "learning_rate": 0.00018823859947512129, + "loss": 0.7481, + "step": 10449 + }, + { + "epoch": 0.7251405176601208, + "grad_norm": 2.5448994636535645, + "learning_rate": 0.00018823595562975028, + "loss": 0.8948, + "step": 10450 + }, + { + "epoch": 0.7252099090972174, + "grad_norm": 3.205112934112549, + "learning_rate": 0.00018823331150582862, + "loss": 0.8681, + "step": 10451 + }, + { + "epoch": 0.7252793005343141, + "grad_norm": 2.595618724822998, + "learning_rate": 0.0001882306671033646, + "loss": 0.8487, + "step": 10452 + }, + { + "epoch": 0.7253486919714107, + "grad_norm": 1.987733006477356, + "learning_rate": 0.00018822802242236664, + "loss": 0.7586, + "step": 10453 + }, + { + "epoch": 0.7254180834085074, + "grad_norm": 2.8014135360717773, + "learning_rate": 0.00018822537746284303, + "loss": 0.8276, + "step": 10454 + }, + { + "epoch": 0.725487474845604, + "grad_norm": 2.575265884399414, + "learning_rate": 0.00018822273222480215, + "loss": 0.8923, + "step": 10455 + }, + { + "epoch": 0.7255568662827007, + "grad_norm": 2.24676775932312, + "learning_rate": 0.00018822008670825232, + "loss": 1.0458, + "step": 10456 + }, + { + "epoch": 0.7256262577197974, + "grad_norm": 2.000361680984497, + "learning_rate": 0.00018821744091320194, + "loss": 0.8233, + "step": 10457 + }, + { + "epoch": 0.725695649156894, + "grad_norm": 2.4819648265838623, + "learning_rate": 0.0001882147948396593, + "loss": 0.8986, + "step": 10458 + }, + { + "epoch": 0.7257650405939907, + "grad_norm": 2.3208084106445312, + "learning_rate": 0.00018821214848763283, + "loss": 0.9043, + "step": 10459 + }, + { + "epoch": 0.7258344320310873, + "grad_norm": 2.108574390411377, + "learning_rate": 0.00018820950185713083, + "loss": 0.896, + "step": 10460 + }, + { + "epoch": 0.725903823468184, + "grad_norm": 2.3856961727142334, + "learning_rate": 0.00018820685494816164, + "loss": 0.8628, + "step": 10461 + }, + { + "epoch": 0.7259732149052807, + "grad_norm": 3.50302791595459, + "learning_rate": 0.00018820420776073367, + "loss": 0.7536, + "step": 10462 + }, + { + "epoch": 0.7260426063423774, + "grad_norm": 2.1602115631103516, + "learning_rate": 0.00018820156029485523, + "loss": 0.6599, + "step": 10463 + }, + { + "epoch": 0.726111997779474, + "grad_norm": 2.458719253540039, + "learning_rate": 0.0001881989125505347, + "loss": 0.8318, + "step": 10464 + }, + { + "epoch": 0.7261813892165707, + "grad_norm": 2.206193447113037, + "learning_rate": 0.00018819626452778044, + "loss": 0.7151, + "step": 10465 + }, + { + "epoch": 0.7262507806536673, + "grad_norm": 1.5202643871307373, + "learning_rate": 0.00018819361622660083, + "loss": 0.5094, + "step": 10466 + }, + { + "epoch": 0.726320172090764, + "grad_norm": 2.5012590885162354, + "learning_rate": 0.00018819096764700416, + "loss": 0.9814, + "step": 10467 + }, + { + "epoch": 0.7263895635278607, + "grad_norm": 3.0143747329711914, + "learning_rate": 0.00018818831878899888, + "loss": 0.9782, + "step": 10468 + }, + { + "epoch": 0.7264589549649573, + "grad_norm": 2.2294247150421143, + "learning_rate": 0.0001881856696525933, + "loss": 0.7942, + "step": 10469 + }, + { + "epoch": 0.726528346402054, + "grad_norm": 2.4116060733795166, + "learning_rate": 0.00018818302023779576, + "loss": 0.9453, + "step": 10470 + }, + { + "epoch": 0.7265977378391506, + "grad_norm": 1.7997045516967773, + "learning_rate": 0.00018818037054461472, + "loss": 0.7296, + "step": 10471 + }, + { + "epoch": 0.7266671292762473, + "grad_norm": 1.7684345245361328, + "learning_rate": 0.00018817772057305843, + "loss": 0.5148, + "step": 10472 + }, + { + "epoch": 0.7267365207133439, + "grad_norm": 3.0233092308044434, + "learning_rate": 0.00018817507032313532, + "loss": 0.8422, + "step": 10473 + }, + { + "epoch": 0.7268059121504407, + "grad_norm": 2.4320695400238037, + "learning_rate": 0.0001881724197948537, + "loss": 0.726, + "step": 10474 + }, + { + "epoch": 0.7268753035875373, + "grad_norm": 2.5110466480255127, + "learning_rate": 0.00018816976898822202, + "loss": 0.9353, + "step": 10475 + }, + { + "epoch": 0.726944695024634, + "grad_norm": 2.3539979457855225, + "learning_rate": 0.00018816711790324862, + "loss": 0.9091, + "step": 10476 + }, + { + "epoch": 0.7270140864617306, + "grad_norm": 2.122732162475586, + "learning_rate": 0.00018816446653994185, + "loss": 0.9173, + "step": 10477 + }, + { + "epoch": 0.7270834778988273, + "grad_norm": 2.259972333908081, + "learning_rate": 0.00018816181489831008, + "loss": 0.6357, + "step": 10478 + }, + { + "epoch": 0.7271528693359239, + "grad_norm": 2.072197914123535, + "learning_rate": 0.0001881591629783617, + "loss": 0.7867, + "step": 10479 + }, + { + "epoch": 0.7272222607730207, + "grad_norm": 2.6556828022003174, + "learning_rate": 0.00018815651078010507, + "loss": 0.76, + "step": 10480 + }, + { + "epoch": 0.7272916522101173, + "grad_norm": 2.078618288040161, + "learning_rate": 0.00018815385830354858, + "loss": 0.8521, + "step": 10481 + }, + { + "epoch": 0.7273610436472139, + "grad_norm": 2.7096729278564453, + "learning_rate": 0.00018815120554870055, + "loss": 1.3063, + "step": 10482 + }, + { + "epoch": 0.7274304350843106, + "grad_norm": 2.3539044857025146, + "learning_rate": 0.0001881485525155694, + "loss": 0.8099, + "step": 10483 + }, + { + "epoch": 0.7274998265214072, + "grad_norm": 2.1782166957855225, + "learning_rate": 0.0001881458992041635, + "loss": 0.7292, + "step": 10484 + }, + { + "epoch": 0.7275692179585039, + "grad_norm": 2.1373178958892822, + "learning_rate": 0.00018814324561449125, + "loss": 0.8588, + "step": 10485 + }, + { + "epoch": 0.7276386093956005, + "grad_norm": 1.9949970245361328, + "learning_rate": 0.00018814059174656096, + "loss": 0.817, + "step": 10486 + }, + { + "epoch": 0.7277080008326973, + "grad_norm": 2.180478096008301, + "learning_rate": 0.00018813793760038107, + "loss": 0.8556, + "step": 10487 + }, + { + "epoch": 0.7277773922697939, + "grad_norm": 1.937850832939148, + "learning_rate": 0.00018813528317595992, + "loss": 0.8498, + "step": 10488 + }, + { + "epoch": 0.7278467837068906, + "grad_norm": 1.9586410522460938, + "learning_rate": 0.00018813262847330593, + "loss": 0.7531, + "step": 10489 + }, + { + "epoch": 0.7279161751439872, + "grad_norm": 2.270960569381714, + "learning_rate": 0.00018812997349242744, + "loss": 0.7505, + "step": 10490 + }, + { + "epoch": 0.7279855665810839, + "grad_norm": 2.368387460708618, + "learning_rate": 0.00018812731823333287, + "loss": 0.8967, + "step": 10491 + }, + { + "epoch": 0.7280549580181805, + "grad_norm": 2.661961793899536, + "learning_rate": 0.00018812466269603056, + "loss": 0.5719, + "step": 10492 + }, + { + "epoch": 0.7281243494552773, + "grad_norm": 1.7893562316894531, + "learning_rate": 0.0001881220068805289, + "loss": 0.7047, + "step": 10493 + }, + { + "epoch": 0.7281937408923739, + "grad_norm": 2.620668888092041, + "learning_rate": 0.0001881193507868363, + "loss": 0.97, + "step": 10494 + }, + { + "epoch": 0.7282631323294706, + "grad_norm": 2.7704479694366455, + "learning_rate": 0.0001881166944149612, + "loss": 0.771, + "step": 10495 + }, + { + "epoch": 0.7283325237665672, + "grad_norm": 2.5175158977508545, + "learning_rate": 0.00018811403776491183, + "loss": 0.6848, + "step": 10496 + }, + { + "epoch": 0.7284019152036638, + "grad_norm": 1.8401955366134644, + "learning_rate": 0.00018811138083669668, + "loss": 0.6559, + "step": 10497 + }, + { + "epoch": 0.7284713066407605, + "grad_norm": 2.5936262607574463, + "learning_rate": 0.00018810872363032416, + "loss": 0.897, + "step": 10498 + }, + { + "epoch": 0.7285406980778572, + "grad_norm": 2.236295700073242, + "learning_rate": 0.0001881060661458026, + "loss": 0.7832, + "step": 10499 + }, + { + "epoch": 0.7286100895149539, + "grad_norm": 2.5752668380737305, + "learning_rate": 0.0001881034083831404, + "loss": 0.9421, + "step": 10500 + }, + { + "epoch": 0.7286794809520505, + "grad_norm": 2.036508083343506, + "learning_rate": 0.00018810075034234597, + "loss": 0.8406, + "step": 10501 + }, + { + "epoch": 0.7287488723891472, + "grad_norm": 2.8794052600860596, + "learning_rate": 0.0001880980920234277, + "loss": 0.8476, + "step": 10502 + }, + { + "epoch": 0.7288182638262438, + "grad_norm": 2.2339320182800293, + "learning_rate": 0.00018809543342639397, + "loss": 0.9808, + "step": 10503 + }, + { + "epoch": 0.7288876552633405, + "grad_norm": 2.808565855026245, + "learning_rate": 0.00018809277455125321, + "loss": 0.8591, + "step": 10504 + }, + { + "epoch": 0.7289570467004372, + "grad_norm": 2.5556371212005615, + "learning_rate": 0.00018809011539801373, + "loss": 1.0094, + "step": 10505 + }, + { + "epoch": 0.7290264381375339, + "grad_norm": 4.029879093170166, + "learning_rate": 0.000188087455966684, + "loss": 0.6901, + "step": 10506 + }, + { + "epoch": 0.7290958295746305, + "grad_norm": 2.255228281021118, + "learning_rate": 0.0001880847962572724, + "loss": 0.9643, + "step": 10507 + }, + { + "epoch": 0.7291652210117272, + "grad_norm": 2.431105136871338, + "learning_rate": 0.0001880821362697873, + "loss": 1.1674, + "step": 10508 + }, + { + "epoch": 0.7292346124488238, + "grad_norm": 2.533634901046753, + "learning_rate": 0.00018807947600423712, + "loss": 0.7808, + "step": 10509 + }, + { + "epoch": 0.7293040038859204, + "grad_norm": 2.7119460105895996, + "learning_rate": 0.00018807681546063028, + "loss": 1.2704, + "step": 10510 + }, + { + "epoch": 0.7293733953230171, + "grad_norm": 2.269742250442505, + "learning_rate": 0.00018807415463897512, + "loss": 0.9346, + "step": 10511 + }, + { + "epoch": 0.7294427867601138, + "grad_norm": 2.050942897796631, + "learning_rate": 0.00018807149353928008, + "loss": 0.8711, + "step": 10512 + }, + { + "epoch": 0.7295121781972105, + "grad_norm": 3.099778175354004, + "learning_rate": 0.00018806883216155354, + "loss": 0.8394, + "step": 10513 + }, + { + "epoch": 0.7295815696343071, + "grad_norm": 2.9162259101867676, + "learning_rate": 0.00018806617050580392, + "loss": 1.2499, + "step": 10514 + }, + { + "epoch": 0.7296509610714038, + "grad_norm": 2.412167549133301, + "learning_rate": 0.00018806350857203966, + "loss": 0.8393, + "step": 10515 + }, + { + "epoch": 0.7297203525085004, + "grad_norm": 4.569793224334717, + "learning_rate": 0.00018806084636026907, + "loss": 0.8565, + "step": 10516 + }, + { + "epoch": 0.7297897439455971, + "grad_norm": 2.4692280292510986, + "learning_rate": 0.00018805818387050062, + "loss": 0.7169, + "step": 10517 + }, + { + "epoch": 0.7298591353826938, + "grad_norm": 2.1107053756713867, + "learning_rate": 0.0001880555211027427, + "loss": 0.6952, + "step": 10518 + }, + { + "epoch": 0.7299285268197905, + "grad_norm": 2.430408239364624, + "learning_rate": 0.0001880528580570037, + "loss": 0.9148, + "step": 10519 + }, + { + "epoch": 0.7299979182568871, + "grad_norm": 2.523364543914795, + "learning_rate": 0.0001880501947332921, + "loss": 0.8511, + "step": 10520 + }, + { + "epoch": 0.7300673096939838, + "grad_norm": 2.42387318611145, + "learning_rate": 0.0001880475311316162, + "loss": 0.7813, + "step": 10521 + }, + { + "epoch": 0.7301367011310804, + "grad_norm": 1.9890086650848389, + "learning_rate": 0.00018804486725198448, + "loss": 0.7904, + "step": 10522 + }, + { + "epoch": 0.730206092568177, + "grad_norm": 2.8343613147735596, + "learning_rate": 0.0001880422030944053, + "loss": 0.9529, + "step": 10523 + }, + { + "epoch": 0.7302754840052738, + "grad_norm": 2.161895990371704, + "learning_rate": 0.00018803953865888711, + "loss": 0.8614, + "step": 10524 + }, + { + "epoch": 0.7303448754423704, + "grad_norm": 2.3762128353118896, + "learning_rate": 0.00018803687394543833, + "loss": 1.0836, + "step": 10525 + }, + { + "epoch": 0.7304142668794671, + "grad_norm": 2.158817768096924, + "learning_rate": 0.00018803420895406733, + "loss": 0.7817, + "step": 10526 + }, + { + "epoch": 0.7304836583165637, + "grad_norm": 2.8133292198181152, + "learning_rate": 0.00018803154368478257, + "loss": 1.03, + "step": 10527 + }, + { + "epoch": 0.7305530497536604, + "grad_norm": 2.0880539417266846, + "learning_rate": 0.00018802887813759245, + "loss": 0.9663, + "step": 10528 + }, + { + "epoch": 0.730622441190757, + "grad_norm": 1.9960137605667114, + "learning_rate": 0.00018802621231250535, + "loss": 0.7764, + "step": 10529 + }, + { + "epoch": 0.7306918326278538, + "grad_norm": 2.270308494567871, + "learning_rate": 0.0001880235462095297, + "loss": 1.0223, + "step": 10530 + }, + { + "epoch": 0.7307612240649504, + "grad_norm": 2.492954969406128, + "learning_rate": 0.00018802087982867395, + "loss": 0.7957, + "step": 10531 + }, + { + "epoch": 0.7308306155020471, + "grad_norm": 2.0080337524414062, + "learning_rate": 0.00018801821316994649, + "loss": 0.8335, + "step": 10532 + }, + { + "epoch": 0.7309000069391437, + "grad_norm": 1.9781324863433838, + "learning_rate": 0.00018801554623335576, + "loss": 0.7931, + "step": 10533 + }, + { + "epoch": 0.7309693983762404, + "grad_norm": 2.7336134910583496, + "learning_rate": 0.00018801287901891014, + "loss": 1.0661, + "step": 10534 + }, + { + "epoch": 0.731038789813337, + "grad_norm": 2.589512348175049, + "learning_rate": 0.00018801021152661806, + "loss": 1.1032, + "step": 10535 + }, + { + "epoch": 0.7311081812504336, + "grad_norm": 1.9678422212600708, + "learning_rate": 0.00018800754375648796, + "loss": 0.8354, + "step": 10536 + }, + { + "epoch": 0.7311775726875304, + "grad_norm": 2.357788562774658, + "learning_rate": 0.00018800487570852829, + "loss": 0.8211, + "step": 10537 + }, + { + "epoch": 0.731246964124627, + "grad_norm": 1.8733651638031006, + "learning_rate": 0.0001880022073827474, + "loss": 0.7885, + "step": 10538 + }, + { + "epoch": 0.7313163555617237, + "grad_norm": 2.404230833053589, + "learning_rate": 0.00018799953877915378, + "loss": 0.8231, + "step": 10539 + }, + { + "epoch": 0.7313857469988203, + "grad_norm": 3.3866100311279297, + "learning_rate": 0.0001879968698977558, + "loss": 0.9296, + "step": 10540 + }, + { + "epoch": 0.731455138435917, + "grad_norm": 2.120431423187256, + "learning_rate": 0.0001879942007385619, + "loss": 0.9266, + "step": 10541 + }, + { + "epoch": 0.7315245298730136, + "grad_norm": 3.2071194648742676, + "learning_rate": 0.00018799153130158054, + "loss": 1.0267, + "step": 10542 + }, + { + "epoch": 0.7315939213101104, + "grad_norm": 3.112313747406006, + "learning_rate": 0.00018798886158682013, + "loss": 1.0529, + "step": 10543 + }, + { + "epoch": 0.731663312747207, + "grad_norm": 2.5266480445861816, + "learning_rate": 0.00018798619159428908, + "loss": 0.8855, + "step": 10544 + }, + { + "epoch": 0.7317327041843037, + "grad_norm": 2.191730260848999, + "learning_rate": 0.00018798352132399585, + "loss": 0.8214, + "step": 10545 + }, + { + "epoch": 0.7318020956214003, + "grad_norm": 2.0903759002685547, + "learning_rate": 0.00018798085077594885, + "loss": 0.596, + "step": 10546 + }, + { + "epoch": 0.731871487058497, + "grad_norm": 2.7002737522125244, + "learning_rate": 0.0001879781799501565, + "loss": 1.0084, + "step": 10547 + }, + { + "epoch": 0.7319408784955936, + "grad_norm": 1.522042989730835, + "learning_rate": 0.00018797550884662724, + "loss": 0.5912, + "step": 10548 + }, + { + "epoch": 0.7320102699326904, + "grad_norm": 1.8952699899673462, + "learning_rate": 0.0001879728374653695, + "loss": 0.8265, + "step": 10549 + }, + { + "epoch": 0.732079661369787, + "grad_norm": 2.616522789001465, + "learning_rate": 0.00018797016580639173, + "loss": 0.937, + "step": 10550 + }, + { + "epoch": 0.7321490528068836, + "grad_norm": 2.4389634132385254, + "learning_rate": 0.00018796749386970233, + "loss": 1.0046, + "step": 10551 + }, + { + "epoch": 0.7322184442439803, + "grad_norm": 2.0557124614715576, + "learning_rate": 0.0001879648216553098, + "loss": 0.5647, + "step": 10552 + }, + { + "epoch": 0.7322878356810769, + "grad_norm": 3.3018689155578613, + "learning_rate": 0.0001879621491632225, + "loss": 0.9407, + "step": 10553 + }, + { + "epoch": 0.7323572271181736, + "grad_norm": 2.3445262908935547, + "learning_rate": 0.0001879594763934489, + "loss": 0.6386, + "step": 10554 + }, + { + "epoch": 0.7324266185552702, + "grad_norm": 2.1997852325439453, + "learning_rate": 0.00018795680334599747, + "loss": 0.7948, + "step": 10555 + }, + { + "epoch": 0.732496009992367, + "grad_norm": 2.20793080329895, + "learning_rate": 0.0001879541300208766, + "loss": 0.9257, + "step": 10556 + }, + { + "epoch": 0.7325654014294636, + "grad_norm": 1.826047658920288, + "learning_rate": 0.00018795145641809473, + "loss": 0.8243, + "step": 10557 + }, + { + "epoch": 0.7326347928665603, + "grad_norm": 1.8906362056732178, + "learning_rate": 0.00018794878253766033, + "loss": 0.6681, + "step": 10558 + }, + { + "epoch": 0.7327041843036569, + "grad_norm": 2.7945210933685303, + "learning_rate": 0.00018794610837958185, + "loss": 0.8851, + "step": 10559 + }, + { + "epoch": 0.7327735757407536, + "grad_norm": 1.8882594108581543, + "learning_rate": 0.00018794343394386767, + "loss": 0.7588, + "step": 10560 + }, + { + "epoch": 0.7328429671778502, + "grad_norm": 2.229332447052002, + "learning_rate": 0.0001879407592305263, + "loss": 1.1072, + "step": 10561 + }, + { + "epoch": 0.732912358614947, + "grad_norm": 5.58914041519165, + "learning_rate": 0.00018793808423956613, + "loss": 1.009, + "step": 10562 + }, + { + "epoch": 0.7329817500520436, + "grad_norm": 2.967334032058716, + "learning_rate": 0.00018793540897099565, + "loss": 0.748, + "step": 10563 + }, + { + "epoch": 0.7330511414891402, + "grad_norm": 3.7343645095825195, + "learning_rate": 0.0001879327334248233, + "loss": 1.0215, + "step": 10564 + }, + { + "epoch": 0.7331205329262369, + "grad_norm": 1.808488368988037, + "learning_rate": 0.0001879300576010575, + "loss": 0.6205, + "step": 10565 + }, + { + "epoch": 0.7331899243633335, + "grad_norm": 2.395580291748047, + "learning_rate": 0.00018792738149970667, + "loss": 1.0028, + "step": 10566 + }, + { + "epoch": 0.7332593158004302, + "grad_norm": 2.1614787578582764, + "learning_rate": 0.00018792470512077935, + "loss": 0.7692, + "step": 10567 + }, + { + "epoch": 0.7333287072375269, + "grad_norm": 2.97390079498291, + "learning_rate": 0.00018792202846428392, + "loss": 1.3212, + "step": 10568 + }, + { + "epoch": 0.7333980986746236, + "grad_norm": 2.909487724304199, + "learning_rate": 0.00018791935153022885, + "loss": 0.7464, + "step": 10569 + }, + { + "epoch": 0.7334674901117202, + "grad_norm": 2.2396650314331055, + "learning_rate": 0.00018791667431862258, + "loss": 0.8476, + "step": 10570 + }, + { + "epoch": 0.7335368815488169, + "grad_norm": 2.1419894695281982, + "learning_rate": 0.0001879139968294736, + "loss": 0.7483, + "step": 10571 + }, + { + "epoch": 0.7336062729859135, + "grad_norm": 3.301729679107666, + "learning_rate": 0.0001879113190627903, + "loss": 0.9958, + "step": 10572 + }, + { + "epoch": 0.7336756644230102, + "grad_norm": 2.380218744277954, + "learning_rate": 0.00018790864101858116, + "loss": 0.7528, + "step": 10573 + }, + { + "epoch": 0.7337450558601069, + "grad_norm": 2.161006212234497, + "learning_rate": 0.00018790596269685465, + "loss": 0.9337, + "step": 10574 + }, + { + "epoch": 0.7338144472972036, + "grad_norm": 2.059396505355835, + "learning_rate": 0.00018790328409761922, + "loss": 0.6609, + "step": 10575 + }, + { + "epoch": 0.7338838387343002, + "grad_norm": 2.4126923084259033, + "learning_rate": 0.00018790060522088332, + "loss": 1.0749, + "step": 10576 + }, + { + "epoch": 0.7339532301713968, + "grad_norm": 2.41717529296875, + "learning_rate": 0.0001878979260666554, + "loss": 0.9399, + "step": 10577 + }, + { + "epoch": 0.7340226216084935, + "grad_norm": 1.8839657306671143, + "learning_rate": 0.00018789524663494393, + "loss": 0.8997, + "step": 10578 + }, + { + "epoch": 0.7340920130455901, + "grad_norm": 2.875922441482544, + "learning_rate": 0.00018789256692575737, + "loss": 0.9124, + "step": 10579 + }, + { + "epoch": 0.7341614044826869, + "grad_norm": 2.6294243335723877, + "learning_rate": 0.00018788988693910417, + "loss": 0.889, + "step": 10580 + }, + { + "epoch": 0.7342307959197835, + "grad_norm": 2.726257562637329, + "learning_rate": 0.00018788720667499277, + "loss": 0.927, + "step": 10581 + }, + { + "epoch": 0.7343001873568802, + "grad_norm": 3.861823558807373, + "learning_rate": 0.0001878845261334317, + "loss": 1.0758, + "step": 10582 + }, + { + "epoch": 0.7343695787939768, + "grad_norm": 2.6882431507110596, + "learning_rate": 0.00018788184531442936, + "loss": 1.3154, + "step": 10583 + }, + { + "epoch": 0.7344389702310735, + "grad_norm": 3.3600778579711914, + "learning_rate": 0.00018787916421799422, + "loss": 0.8292, + "step": 10584 + }, + { + "epoch": 0.7345083616681701, + "grad_norm": 2.40492582321167, + "learning_rate": 0.00018787648284413475, + "loss": 0.6985, + "step": 10585 + }, + { + "epoch": 0.7345777531052669, + "grad_norm": 2.509521961212158, + "learning_rate": 0.00018787380119285945, + "loss": 0.7258, + "step": 10586 + }, + { + "epoch": 0.7346471445423635, + "grad_norm": 2.1933608055114746, + "learning_rate": 0.0001878711192641767, + "loss": 1.1863, + "step": 10587 + }, + { + "epoch": 0.7347165359794602, + "grad_norm": 2.216808319091797, + "learning_rate": 0.00018786843705809506, + "loss": 0.8647, + "step": 10588 + }, + { + "epoch": 0.7347859274165568, + "grad_norm": 2.2929847240448, + "learning_rate": 0.00018786575457462298, + "loss": 1.0076, + "step": 10589 + }, + { + "epoch": 0.7348553188536534, + "grad_norm": 2.5610859394073486, + "learning_rate": 0.00018786307181376888, + "loss": 0.6827, + "step": 10590 + }, + { + "epoch": 0.7349247102907501, + "grad_norm": 2.1352617740631104, + "learning_rate": 0.00018786038877554125, + "loss": 0.9493, + "step": 10591 + }, + { + "epoch": 0.7349941017278467, + "grad_norm": 2.148834466934204, + "learning_rate": 0.00018785770545994856, + "loss": 0.7931, + "step": 10592 + }, + { + "epoch": 0.7350634931649435, + "grad_norm": 2.368471384048462, + "learning_rate": 0.0001878550218669993, + "loss": 0.9267, + "step": 10593 + }, + { + "epoch": 0.7351328846020401, + "grad_norm": 2.245157241821289, + "learning_rate": 0.00018785233799670193, + "loss": 0.8583, + "step": 10594 + }, + { + "epoch": 0.7352022760391368, + "grad_norm": 3.1243045330047607, + "learning_rate": 0.00018784965384906492, + "loss": 0.892, + "step": 10595 + }, + { + "epoch": 0.7352716674762334, + "grad_norm": 1.6505768299102783, + "learning_rate": 0.00018784696942409675, + "loss": 0.6604, + "step": 10596 + }, + { + "epoch": 0.7353410589133301, + "grad_norm": 2.7021586894989014, + "learning_rate": 0.0001878442847218059, + "loss": 0.964, + "step": 10597 + }, + { + "epoch": 0.7354104503504267, + "grad_norm": 1.9421284198760986, + "learning_rate": 0.00018784159974220081, + "loss": 0.8789, + "step": 10598 + }, + { + "epoch": 0.7354798417875235, + "grad_norm": 1.9929847717285156, + "learning_rate": 0.00018783891448529, + "loss": 0.8424, + "step": 10599 + }, + { + "epoch": 0.7355492332246201, + "grad_norm": 1.841612696647644, + "learning_rate": 0.0001878362289510819, + "loss": 0.9249, + "step": 10600 + }, + { + "epoch": 0.7356186246617168, + "grad_norm": 2.370542287826538, + "learning_rate": 0.00018783354313958505, + "loss": 1.0951, + "step": 10601 + }, + { + "epoch": 0.7356880160988134, + "grad_norm": 2.1540944576263428, + "learning_rate": 0.00018783085705080787, + "loss": 0.6773, + "step": 10602 + }, + { + "epoch": 0.7357574075359101, + "grad_norm": 2.3967723846435547, + "learning_rate": 0.00018782817068475888, + "loss": 0.9265, + "step": 10603 + }, + { + "epoch": 0.7358267989730067, + "grad_norm": 2.441161632537842, + "learning_rate": 0.00018782548404144654, + "loss": 0.8106, + "step": 10604 + }, + { + "epoch": 0.7358961904101033, + "grad_norm": 2.8127689361572266, + "learning_rate": 0.00018782279712087935, + "loss": 0.7707, + "step": 10605 + }, + { + "epoch": 0.7359655818472001, + "grad_norm": 1.8676186800003052, + "learning_rate": 0.00018782010992306577, + "loss": 0.863, + "step": 10606 + }, + { + "epoch": 0.7360349732842967, + "grad_norm": 1.5913199186325073, + "learning_rate": 0.0001878174224480143, + "loss": 0.6434, + "step": 10607 + }, + { + "epoch": 0.7361043647213934, + "grad_norm": 2.2056658267974854, + "learning_rate": 0.00018781473469573338, + "loss": 0.7184, + "step": 10608 + }, + { + "epoch": 0.73617375615849, + "grad_norm": 1.802778959274292, + "learning_rate": 0.00018781204666623154, + "loss": 0.65, + "step": 10609 + }, + { + "epoch": 0.7362431475955867, + "grad_norm": 2.217766284942627, + "learning_rate": 0.0001878093583595173, + "loss": 0.7261, + "step": 10610 + }, + { + "epoch": 0.7363125390326833, + "grad_norm": 2.275109052658081, + "learning_rate": 0.00018780666977559907, + "loss": 0.9018, + "step": 10611 + }, + { + "epoch": 0.7363819304697801, + "grad_norm": 2.2889156341552734, + "learning_rate": 0.00018780398091448536, + "loss": 0.7642, + "step": 10612 + }, + { + "epoch": 0.7364513219068767, + "grad_norm": 1.9871973991394043, + "learning_rate": 0.0001878012917761847, + "loss": 0.9042, + "step": 10613 + }, + { + "epoch": 0.7365207133439734, + "grad_norm": 2.1747756004333496, + "learning_rate": 0.00018779860236070554, + "loss": 0.8648, + "step": 10614 + }, + { + "epoch": 0.73659010478107, + "grad_norm": 1.9182233810424805, + "learning_rate": 0.00018779591266805635, + "loss": 0.8023, + "step": 10615 + }, + { + "epoch": 0.7366594962181667, + "grad_norm": 2.6050117015838623, + "learning_rate": 0.00018779322269824568, + "loss": 0.9069, + "step": 10616 + }, + { + "epoch": 0.7367288876552633, + "grad_norm": 2.4397895336151123, + "learning_rate": 0.000187790532451282, + "loss": 1.0314, + "step": 10617 + }, + { + "epoch": 0.73679827909236, + "grad_norm": 2.0678176879882812, + "learning_rate": 0.00018778784192717378, + "loss": 0.705, + "step": 10618 + }, + { + "epoch": 0.7368676705294567, + "grad_norm": 2.241783380508423, + "learning_rate": 0.00018778515112592953, + "loss": 0.7814, + "step": 10619 + }, + { + "epoch": 0.7369370619665533, + "grad_norm": 1.918161153793335, + "learning_rate": 0.00018778246004755775, + "loss": 0.6227, + "step": 10620 + }, + { + "epoch": 0.73700645340365, + "grad_norm": 3.158442258834839, + "learning_rate": 0.0001877797686920669, + "loss": 0.804, + "step": 10621 + }, + { + "epoch": 0.7370758448407466, + "grad_norm": 2.080239772796631, + "learning_rate": 0.00018777707705946555, + "loss": 0.9195, + "step": 10622 + }, + { + "epoch": 0.7371452362778433, + "grad_norm": 1.9998306035995483, + "learning_rate": 0.0001877743851497621, + "loss": 0.8356, + "step": 10623 + }, + { + "epoch": 0.73721462771494, + "grad_norm": 2.269942283630371, + "learning_rate": 0.00018777169296296516, + "loss": 0.8566, + "step": 10624 + }, + { + "epoch": 0.7372840191520367, + "grad_norm": 1.9105033874511719, + "learning_rate": 0.00018776900049908313, + "loss": 0.7859, + "step": 10625 + }, + { + "epoch": 0.7373534105891333, + "grad_norm": 2.103870391845703, + "learning_rate": 0.00018776630775812456, + "loss": 0.7708, + "step": 10626 + }, + { + "epoch": 0.73742280202623, + "grad_norm": 2.3153069019317627, + "learning_rate": 0.0001877636147400979, + "loss": 0.9373, + "step": 10627 + }, + { + "epoch": 0.7374921934633266, + "grad_norm": 1.8517082929611206, + "learning_rate": 0.00018776092144501174, + "loss": 0.9558, + "step": 10628 + }, + { + "epoch": 0.7375615849004233, + "grad_norm": 2.4551355838775635, + "learning_rate": 0.00018775822787287451, + "loss": 0.9726, + "step": 10629 + }, + { + "epoch": 0.73763097633752, + "grad_norm": 2.6976892948150635, + "learning_rate": 0.00018775553402369474, + "loss": 0.8464, + "step": 10630 + }, + { + "epoch": 0.7377003677746166, + "grad_norm": 2.1748440265655518, + "learning_rate": 0.00018775283989748092, + "loss": 0.6332, + "step": 10631 + }, + { + "epoch": 0.7377697592117133, + "grad_norm": 2.612312078475952, + "learning_rate": 0.00018775014549424158, + "loss": 0.8036, + "step": 10632 + }, + { + "epoch": 0.7378391506488099, + "grad_norm": 2.3671412467956543, + "learning_rate": 0.00018774745081398523, + "loss": 0.8653, + "step": 10633 + }, + { + "epoch": 0.7379085420859066, + "grad_norm": 2.2003304958343506, + "learning_rate": 0.00018774475585672034, + "loss": 0.7119, + "step": 10634 + }, + { + "epoch": 0.7379779335230032, + "grad_norm": 2.1428725719451904, + "learning_rate": 0.00018774206062245545, + "loss": 0.7348, + "step": 10635 + }, + { + "epoch": 0.7380473249601, + "grad_norm": 2.328799247741699, + "learning_rate": 0.00018773936511119903, + "loss": 0.9333, + "step": 10636 + }, + { + "epoch": 0.7381167163971966, + "grad_norm": 1.9672648906707764, + "learning_rate": 0.0001877366693229596, + "loss": 0.8118, + "step": 10637 + }, + { + "epoch": 0.7381861078342933, + "grad_norm": 1.9671610593795776, + "learning_rate": 0.0001877339732577457, + "loss": 0.7207, + "step": 10638 + }, + { + "epoch": 0.7382554992713899, + "grad_norm": 1.7152864933013916, + "learning_rate": 0.00018773127691556587, + "loss": 0.6656, + "step": 10639 + }, + { + "epoch": 0.7383248907084866, + "grad_norm": 2.077894449234009, + "learning_rate": 0.00018772858029642852, + "loss": 0.6962, + "step": 10640 + }, + { + "epoch": 0.7383942821455832, + "grad_norm": 2.6635169982910156, + "learning_rate": 0.00018772588340034224, + "loss": 0.9853, + "step": 10641 + }, + { + "epoch": 0.73846367358268, + "grad_norm": 1.9995514154434204, + "learning_rate": 0.00018772318622731552, + "loss": 1.0652, + "step": 10642 + }, + { + "epoch": 0.7385330650197766, + "grad_norm": 2.334378719329834, + "learning_rate": 0.0001877204887773569, + "loss": 0.8362, + "step": 10643 + }, + { + "epoch": 0.7386024564568732, + "grad_norm": 1.708554744720459, + "learning_rate": 0.00018771779105047486, + "loss": 0.7441, + "step": 10644 + }, + { + "epoch": 0.7386718478939699, + "grad_norm": 2.080805778503418, + "learning_rate": 0.00018771509304667792, + "loss": 0.9568, + "step": 10645 + }, + { + "epoch": 0.7387412393310665, + "grad_norm": 3.2146506309509277, + "learning_rate": 0.0001877123947659746, + "loss": 1.1583, + "step": 10646 + }, + { + "epoch": 0.7388106307681632, + "grad_norm": 1.9094873666763306, + "learning_rate": 0.00018770969620837346, + "loss": 0.6903, + "step": 10647 + }, + { + "epoch": 0.7388800222052598, + "grad_norm": 2.4911725521087646, + "learning_rate": 0.00018770699737388297, + "loss": 0.7978, + "step": 10648 + }, + { + "epoch": 0.7389494136423566, + "grad_norm": 2.033928394317627, + "learning_rate": 0.00018770429826251168, + "loss": 0.6831, + "step": 10649 + }, + { + "epoch": 0.7390188050794532, + "grad_norm": 1.7261840105056763, + "learning_rate": 0.00018770159887426807, + "loss": 0.7304, + "step": 10650 + }, + { + "epoch": 0.7390881965165499, + "grad_norm": 2.0162086486816406, + "learning_rate": 0.0001876988992091607, + "loss": 0.7244, + "step": 10651 + }, + { + "epoch": 0.7391575879536465, + "grad_norm": 2.377361536026001, + "learning_rate": 0.0001876961992671981, + "loss": 0.838, + "step": 10652 + }, + { + "epoch": 0.7392269793907432, + "grad_norm": 1.7656898498535156, + "learning_rate": 0.00018769349904838872, + "loss": 0.6571, + "step": 10653 + }, + { + "epoch": 0.7392963708278398, + "grad_norm": 2.1589996814727783, + "learning_rate": 0.0001876907985527412, + "loss": 0.8714, + "step": 10654 + }, + { + "epoch": 0.7393657622649366, + "grad_norm": 2.276693105697632, + "learning_rate": 0.00018768809778026396, + "loss": 1.033, + "step": 10655 + }, + { + "epoch": 0.7394351537020332, + "grad_norm": 1.8004096746444702, + "learning_rate": 0.00018768539673096556, + "loss": 0.6695, + "step": 10656 + }, + { + "epoch": 0.7395045451391298, + "grad_norm": 1.9440149068832397, + "learning_rate": 0.00018768269540485457, + "loss": 0.8067, + "step": 10657 + }, + { + "epoch": 0.7395739365762265, + "grad_norm": 2.353698253631592, + "learning_rate": 0.00018767999380193948, + "loss": 0.9803, + "step": 10658 + }, + { + "epoch": 0.7396433280133231, + "grad_norm": 1.8429756164550781, + "learning_rate": 0.0001876772919222288, + "loss": 0.6932, + "step": 10659 + }, + { + "epoch": 0.7397127194504198, + "grad_norm": 2.55165958404541, + "learning_rate": 0.0001876745897657311, + "loss": 1.191, + "step": 10660 + }, + { + "epoch": 0.7397821108875164, + "grad_norm": 2.344127893447876, + "learning_rate": 0.0001876718873324549, + "loss": 0.8075, + "step": 10661 + }, + { + "epoch": 0.7398515023246132, + "grad_norm": 2.177003860473633, + "learning_rate": 0.0001876691846224087, + "loss": 0.96, + "step": 10662 + }, + { + "epoch": 0.7399208937617098, + "grad_norm": 2.799506425857544, + "learning_rate": 0.00018766648163560107, + "loss": 0.7397, + "step": 10663 + }, + { + "epoch": 0.7399902851988065, + "grad_norm": 2.912184476852417, + "learning_rate": 0.0001876637783720405, + "loss": 0.9136, + "step": 10664 + }, + { + "epoch": 0.7400596766359031, + "grad_norm": 1.897289514541626, + "learning_rate": 0.0001876610748317356, + "loss": 0.8693, + "step": 10665 + }, + { + "epoch": 0.7401290680729998, + "grad_norm": 2.489672899246216, + "learning_rate": 0.00018765837101469482, + "loss": 0.8583, + "step": 10666 + }, + { + "epoch": 0.7401984595100964, + "grad_norm": 2.201845169067383, + "learning_rate": 0.00018765566692092675, + "loss": 0.8639, + "step": 10667 + }, + { + "epoch": 0.7402678509471932, + "grad_norm": 2.026017904281616, + "learning_rate": 0.00018765296255043992, + "loss": 0.8998, + "step": 10668 + }, + { + "epoch": 0.7403372423842898, + "grad_norm": 2.248171091079712, + "learning_rate": 0.00018765025790324285, + "loss": 0.8916, + "step": 10669 + }, + { + "epoch": 0.7404066338213865, + "grad_norm": 2.7714903354644775, + "learning_rate": 0.00018764755297934408, + "loss": 0.8688, + "step": 10670 + }, + { + "epoch": 0.7404760252584831, + "grad_norm": 2.600010395050049, + "learning_rate": 0.00018764484777875214, + "loss": 0.9038, + "step": 10671 + }, + { + "epoch": 0.7405454166955797, + "grad_norm": 4.107349395751953, + "learning_rate": 0.00018764214230147558, + "loss": 0.9245, + "step": 10672 + }, + { + "epoch": 0.7406148081326764, + "grad_norm": 2.431777000427246, + "learning_rate": 0.00018763943654752298, + "loss": 0.7853, + "step": 10673 + }, + { + "epoch": 0.740684199569773, + "grad_norm": 1.8614115715026855, + "learning_rate": 0.00018763673051690286, + "loss": 0.7288, + "step": 10674 + }, + { + "epoch": 0.7407535910068698, + "grad_norm": 1.8599659204483032, + "learning_rate": 0.00018763402420962367, + "loss": 0.666, + "step": 10675 + }, + { + "epoch": 0.7408229824439664, + "grad_norm": 2.245086908340454, + "learning_rate": 0.00018763131762569412, + "loss": 0.8151, + "step": 10676 + }, + { + "epoch": 0.7408923738810631, + "grad_norm": 3.4302258491516113, + "learning_rate": 0.00018762861076512263, + "loss": 1.2653, + "step": 10677 + }, + { + "epoch": 0.7409617653181597, + "grad_norm": 2.3188960552215576, + "learning_rate": 0.00018762590362791777, + "loss": 0.7535, + "step": 10678 + }, + { + "epoch": 0.7410311567552564, + "grad_norm": 2.4880919456481934, + "learning_rate": 0.0001876231962140881, + "loss": 1.0719, + "step": 10679 + }, + { + "epoch": 0.741100548192353, + "grad_norm": 1.5859122276306152, + "learning_rate": 0.0001876204885236422, + "loss": 0.6893, + "step": 10680 + }, + { + "epoch": 0.7411699396294498, + "grad_norm": 3.2243130207061768, + "learning_rate": 0.00018761778055658855, + "loss": 0.8898, + "step": 10681 + }, + { + "epoch": 0.7412393310665464, + "grad_norm": 2.6213629245758057, + "learning_rate": 0.00018761507231293575, + "loss": 0.7104, + "step": 10682 + }, + { + "epoch": 0.7413087225036431, + "grad_norm": 2.7252955436706543, + "learning_rate": 0.00018761236379269232, + "loss": 1.199, + "step": 10683 + }, + { + "epoch": 0.7413781139407397, + "grad_norm": 2.478114128112793, + "learning_rate": 0.00018760965499586683, + "loss": 0.6681, + "step": 10684 + }, + { + "epoch": 0.7414475053778363, + "grad_norm": 2.3028931617736816, + "learning_rate": 0.00018760694592246782, + "loss": 0.9262, + "step": 10685 + }, + { + "epoch": 0.741516896814933, + "grad_norm": 2.23637056350708, + "learning_rate": 0.00018760423657250384, + "loss": 0.8084, + "step": 10686 + }, + { + "epoch": 0.7415862882520297, + "grad_norm": 2.3074092864990234, + "learning_rate": 0.00018760152694598345, + "loss": 0.9617, + "step": 10687 + }, + { + "epoch": 0.7416556796891264, + "grad_norm": 1.6826081275939941, + "learning_rate": 0.00018759881704291523, + "loss": 0.679, + "step": 10688 + }, + { + "epoch": 0.741725071126223, + "grad_norm": 2.4478001594543457, + "learning_rate": 0.0001875961068633077, + "loss": 0.9369, + "step": 10689 + }, + { + "epoch": 0.7417944625633197, + "grad_norm": 2.968252420425415, + "learning_rate": 0.0001875933964071694, + "loss": 0.9614, + "step": 10690 + }, + { + "epoch": 0.7418638540004163, + "grad_norm": 1.992161512374878, + "learning_rate": 0.0001875906856745089, + "loss": 0.8744, + "step": 10691 + }, + { + "epoch": 0.741933245437513, + "grad_norm": 2.660959243774414, + "learning_rate": 0.0001875879746653348, + "loss": 0.886, + "step": 10692 + }, + { + "epoch": 0.7420026368746097, + "grad_norm": 2.9052932262420654, + "learning_rate": 0.0001875852633796556, + "loss": 0.9319, + "step": 10693 + }, + { + "epoch": 0.7420720283117064, + "grad_norm": 2.2128543853759766, + "learning_rate": 0.0001875825518174799, + "loss": 0.8877, + "step": 10694 + }, + { + "epoch": 0.742141419748803, + "grad_norm": 2.187206983566284, + "learning_rate": 0.00018757983997881624, + "loss": 0.8535, + "step": 10695 + }, + { + "epoch": 0.7422108111858997, + "grad_norm": 2.9326415061950684, + "learning_rate": 0.0001875771278636732, + "loss": 0.7424, + "step": 10696 + }, + { + "epoch": 0.7422802026229963, + "grad_norm": 3.322206974029541, + "learning_rate": 0.00018757441547205935, + "loss": 1.0366, + "step": 10697 + }, + { + "epoch": 0.7423495940600929, + "grad_norm": 2.2656338214874268, + "learning_rate": 0.00018757170280398318, + "loss": 1.0726, + "step": 10698 + }, + { + "epoch": 0.7424189854971897, + "grad_norm": 2.6381161212921143, + "learning_rate": 0.0001875689898594533, + "loss": 0.7701, + "step": 10699 + }, + { + "epoch": 0.7424883769342863, + "grad_norm": 2.5195488929748535, + "learning_rate": 0.00018756627663847833, + "loss": 0.9488, + "step": 10700 + }, + { + "epoch": 0.742557768371383, + "grad_norm": 2.3791165351867676, + "learning_rate": 0.00018756356314106673, + "loss": 0.6815, + "step": 10701 + }, + { + "epoch": 0.7426271598084796, + "grad_norm": 2.2038333415985107, + "learning_rate": 0.00018756084936722719, + "loss": 1.1813, + "step": 10702 + }, + { + "epoch": 0.7426965512455763, + "grad_norm": 2.8293607234954834, + "learning_rate": 0.00018755813531696816, + "loss": 0.8556, + "step": 10703 + }, + { + "epoch": 0.7427659426826729, + "grad_norm": 5.109283447265625, + "learning_rate": 0.00018755542099029823, + "loss": 0.9431, + "step": 10704 + }, + { + "epoch": 0.7428353341197697, + "grad_norm": 2.5350332260131836, + "learning_rate": 0.00018755270638722604, + "loss": 0.9666, + "step": 10705 + }, + { + "epoch": 0.7429047255568663, + "grad_norm": 2.1461517810821533, + "learning_rate": 0.00018754999150776008, + "loss": 0.8923, + "step": 10706 + }, + { + "epoch": 0.742974116993963, + "grad_norm": 2.5999183654785156, + "learning_rate": 0.000187547276351909, + "loss": 0.8105, + "step": 10707 + }, + { + "epoch": 0.7430435084310596, + "grad_norm": 2.186631202697754, + "learning_rate": 0.0001875445609196813, + "loss": 1.0723, + "step": 10708 + }, + { + "epoch": 0.7431128998681563, + "grad_norm": 2.2076497077941895, + "learning_rate": 0.00018754184521108558, + "loss": 0.7523, + "step": 10709 + }, + { + "epoch": 0.7431822913052529, + "grad_norm": 2.060692071914673, + "learning_rate": 0.00018753912922613041, + "loss": 0.7511, + "step": 10710 + }, + { + "epoch": 0.7432516827423495, + "grad_norm": 2.690371513366699, + "learning_rate": 0.00018753641296482437, + "loss": 0.6701, + "step": 10711 + }, + { + "epoch": 0.7433210741794463, + "grad_norm": 2.256197690963745, + "learning_rate": 0.00018753369642717603, + "loss": 0.7599, + "step": 10712 + }, + { + "epoch": 0.7433904656165429, + "grad_norm": 2.010477304458618, + "learning_rate": 0.00018753097961319394, + "loss": 0.6368, + "step": 10713 + }, + { + "epoch": 0.7434598570536396, + "grad_norm": 2.9444363117218018, + "learning_rate": 0.00018752826252288674, + "loss": 0.9567, + "step": 10714 + }, + { + "epoch": 0.7435292484907362, + "grad_norm": 2.494098663330078, + "learning_rate": 0.00018752554515626295, + "loss": 0.8025, + "step": 10715 + }, + { + "epoch": 0.7435986399278329, + "grad_norm": 2.8418662548065186, + "learning_rate": 0.00018752282751333116, + "loss": 0.8627, + "step": 10716 + }, + { + "epoch": 0.7436680313649295, + "grad_norm": 2.554783821105957, + "learning_rate": 0.00018752010959409998, + "loss": 0.8835, + "step": 10717 + }, + { + "epoch": 0.7437374228020263, + "grad_norm": 2.530618667602539, + "learning_rate": 0.000187517391398578, + "loss": 1.0172, + "step": 10718 + }, + { + "epoch": 0.7438068142391229, + "grad_norm": 2.575355291366577, + "learning_rate": 0.0001875146729267737, + "loss": 0.6984, + "step": 10719 + }, + { + "epoch": 0.7438762056762196, + "grad_norm": 2.3052332401275635, + "learning_rate": 0.00018751195417869576, + "loss": 0.8579, + "step": 10720 + }, + { + "epoch": 0.7439455971133162, + "grad_norm": 2.4897871017456055, + "learning_rate": 0.00018750923515435273, + "loss": 0.8212, + "step": 10721 + }, + { + "epoch": 0.7440149885504129, + "grad_norm": 2.623108386993408, + "learning_rate": 0.0001875065158537532, + "loss": 1.0638, + "step": 10722 + }, + { + "epoch": 0.7440843799875095, + "grad_norm": 2.1473324298858643, + "learning_rate": 0.00018750379627690576, + "loss": 0.7702, + "step": 10723 + }, + { + "epoch": 0.7441537714246063, + "grad_norm": 2.4156229496002197, + "learning_rate": 0.00018750107642381898, + "loss": 0.8134, + "step": 10724 + }, + { + "epoch": 0.7442231628617029, + "grad_norm": 3.1371185779571533, + "learning_rate": 0.00018749835629450144, + "loss": 0.6604, + "step": 10725 + }, + { + "epoch": 0.7442925542987995, + "grad_norm": 1.6057863235473633, + "learning_rate": 0.00018749563588896177, + "loss": 0.6817, + "step": 10726 + }, + { + "epoch": 0.7443619457358962, + "grad_norm": 2.060213327407837, + "learning_rate": 0.0001874929152072085, + "loss": 0.8232, + "step": 10727 + }, + { + "epoch": 0.7444313371729928, + "grad_norm": 2.6276590824127197, + "learning_rate": 0.00018749019424925027, + "loss": 1.0724, + "step": 10728 + }, + { + "epoch": 0.7445007286100895, + "grad_norm": 2.39227032661438, + "learning_rate": 0.00018748747301509564, + "loss": 0.7116, + "step": 10729 + }, + { + "epoch": 0.7445701200471861, + "grad_norm": 3.7670998573303223, + "learning_rate": 0.00018748475150475323, + "loss": 0.8186, + "step": 10730 + }, + { + "epoch": 0.7446395114842829, + "grad_norm": 3.3573553562164307, + "learning_rate": 0.00018748202971823158, + "loss": 0.8541, + "step": 10731 + }, + { + "epoch": 0.7447089029213795, + "grad_norm": 3.0566110610961914, + "learning_rate": 0.0001874793076555393, + "loss": 0.9483, + "step": 10732 + }, + { + "epoch": 0.7447782943584762, + "grad_norm": 2.798586845397949, + "learning_rate": 0.00018747658531668505, + "loss": 0.6821, + "step": 10733 + }, + { + "epoch": 0.7448476857955728, + "grad_norm": 2.3382275104522705, + "learning_rate": 0.0001874738627016773, + "loss": 0.8111, + "step": 10734 + }, + { + "epoch": 0.7449170772326695, + "grad_norm": 2.337960720062256, + "learning_rate": 0.00018747113981052476, + "loss": 0.8045, + "step": 10735 + }, + { + "epoch": 0.7449864686697661, + "grad_norm": 2.9480464458465576, + "learning_rate": 0.00018746841664323599, + "loss": 1.0367, + "step": 10736 + }, + { + "epoch": 0.7450558601068629, + "grad_norm": 2.705533742904663, + "learning_rate": 0.00018746569319981954, + "loss": 0.5635, + "step": 10737 + }, + { + "epoch": 0.7451252515439595, + "grad_norm": 3.3554744720458984, + "learning_rate": 0.00018746296948028409, + "loss": 0.9861, + "step": 10738 + }, + { + "epoch": 0.7451946429810561, + "grad_norm": 1.813719630241394, + "learning_rate": 0.00018746024548463815, + "loss": 0.666, + "step": 10739 + }, + { + "epoch": 0.7452640344181528, + "grad_norm": 2.735391855239868, + "learning_rate": 0.00018745752121289034, + "loss": 0.751, + "step": 10740 + }, + { + "epoch": 0.7453334258552494, + "grad_norm": 3.0957818031311035, + "learning_rate": 0.00018745479666504932, + "loss": 0.8425, + "step": 10741 + }, + { + "epoch": 0.7454028172923461, + "grad_norm": 2.2820355892181396, + "learning_rate": 0.0001874520718411237, + "loss": 0.7579, + "step": 10742 + }, + { + "epoch": 0.7454722087294428, + "grad_norm": 3.009814739227295, + "learning_rate": 0.00018744934674112197, + "loss": 0.915, + "step": 10743 + }, + { + "epoch": 0.7455416001665395, + "grad_norm": 2.515671491622925, + "learning_rate": 0.0001874466213650528, + "loss": 0.9164, + "step": 10744 + }, + { + "epoch": 0.7456109916036361, + "grad_norm": 3.2680864334106445, + "learning_rate": 0.0001874438957129248, + "loss": 0.685, + "step": 10745 + }, + { + "epoch": 0.7456803830407328, + "grad_norm": 2.9215989112854004, + "learning_rate": 0.0001874411697847466, + "loss": 0.9385, + "step": 10746 + }, + { + "epoch": 0.7457497744778294, + "grad_norm": 2.7902638912200928, + "learning_rate": 0.00018743844358052672, + "loss": 0.9777, + "step": 10747 + }, + { + "epoch": 0.7458191659149261, + "grad_norm": 3.1507928371429443, + "learning_rate": 0.00018743571710027387, + "loss": 0.9447, + "step": 10748 + }, + { + "epoch": 0.7458885573520228, + "grad_norm": 3.0428547859191895, + "learning_rate": 0.00018743299034399655, + "loss": 1.0816, + "step": 10749 + }, + { + "epoch": 0.7459579487891195, + "grad_norm": 2.35988187789917, + "learning_rate": 0.00018743026331170345, + "loss": 0.7918, + "step": 10750 + }, + { + "epoch": 0.7460273402262161, + "grad_norm": 2.4419987201690674, + "learning_rate": 0.00018742753600340317, + "loss": 0.7644, + "step": 10751 + }, + { + "epoch": 0.7460967316633127, + "grad_norm": 2.1755621433258057, + "learning_rate": 0.00018742480841910428, + "loss": 0.9954, + "step": 10752 + }, + { + "epoch": 0.7461661231004094, + "grad_norm": 3.2723686695098877, + "learning_rate": 0.00018742208055881542, + "loss": 1.0627, + "step": 10753 + }, + { + "epoch": 0.746235514537506, + "grad_norm": 2.6220450401306152, + "learning_rate": 0.0001874193524225452, + "loss": 1.0808, + "step": 10754 + }, + { + "epoch": 0.7463049059746028, + "grad_norm": 2.463815450668335, + "learning_rate": 0.00018741662401030223, + "loss": 1.0802, + "step": 10755 + }, + { + "epoch": 0.7463742974116994, + "grad_norm": 1.7127935886383057, + "learning_rate": 0.0001874138953220951, + "loss": 0.5956, + "step": 10756 + }, + { + "epoch": 0.7464436888487961, + "grad_norm": 2.071021318435669, + "learning_rate": 0.0001874111663579325, + "loss": 0.7465, + "step": 10757 + }, + { + "epoch": 0.7465130802858927, + "grad_norm": 2.3413259983062744, + "learning_rate": 0.00018740843711782291, + "loss": 0.884, + "step": 10758 + }, + { + "epoch": 0.7465824717229894, + "grad_norm": 3.2934913635253906, + "learning_rate": 0.00018740570760177508, + "loss": 1.0216, + "step": 10759 + }, + { + "epoch": 0.746651863160086, + "grad_norm": 2.748748779296875, + "learning_rate": 0.00018740297780979755, + "loss": 0.9549, + "step": 10760 + }, + { + "epoch": 0.7467212545971827, + "grad_norm": 3.1743533611297607, + "learning_rate": 0.00018740024774189896, + "loss": 0.8548, + "step": 10761 + }, + { + "epoch": 0.7467906460342794, + "grad_norm": 2.354928493499756, + "learning_rate": 0.00018739751739808792, + "loss": 0.9521, + "step": 10762 + }, + { + "epoch": 0.7468600374713761, + "grad_norm": 2.183384418487549, + "learning_rate": 0.00018739478677837307, + "loss": 0.8511, + "step": 10763 + }, + { + "epoch": 0.7469294289084727, + "grad_norm": 2.133084297180176, + "learning_rate": 0.00018739205588276303, + "loss": 0.9131, + "step": 10764 + }, + { + "epoch": 0.7469988203455693, + "grad_norm": 2.922891139984131, + "learning_rate": 0.0001873893247112664, + "loss": 0.9473, + "step": 10765 + }, + { + "epoch": 0.747068211782666, + "grad_norm": 2.347961664199829, + "learning_rate": 0.00018738659326389181, + "loss": 0.7998, + "step": 10766 + }, + { + "epoch": 0.7471376032197626, + "grad_norm": 2.793379783630371, + "learning_rate": 0.00018738386154064788, + "loss": 0.8521, + "step": 10767 + }, + { + "epoch": 0.7472069946568594, + "grad_norm": 2.5880348682403564, + "learning_rate": 0.00018738112954154324, + "loss": 0.9234, + "step": 10768 + }, + { + "epoch": 0.747276386093956, + "grad_norm": 2.0487687587738037, + "learning_rate": 0.0001873783972665865, + "loss": 0.7329, + "step": 10769 + }, + { + "epoch": 0.7473457775310527, + "grad_norm": 2.731813669204712, + "learning_rate": 0.0001873756647157863, + "loss": 0.895, + "step": 10770 + }, + { + "epoch": 0.7474151689681493, + "grad_norm": 2.7033517360687256, + "learning_rate": 0.0001873729318891513, + "loss": 0.7636, + "step": 10771 + }, + { + "epoch": 0.747484560405246, + "grad_norm": 2.216996669769287, + "learning_rate": 0.00018737019878669003, + "loss": 0.8285, + "step": 10772 + }, + { + "epoch": 0.7475539518423426, + "grad_norm": 3.370598554611206, + "learning_rate": 0.00018736746540841124, + "loss": 1.0616, + "step": 10773 + }, + { + "epoch": 0.7476233432794394, + "grad_norm": 1.86750328540802, + "learning_rate": 0.00018736473175432345, + "loss": 0.6556, + "step": 10774 + }, + { + "epoch": 0.747692734716536, + "grad_norm": 2.744793653488159, + "learning_rate": 0.00018736199782443537, + "loss": 0.8521, + "step": 10775 + }, + { + "epoch": 0.7477621261536327, + "grad_norm": 2.0441343784332275, + "learning_rate": 0.0001873592636187556, + "loss": 0.7251, + "step": 10776 + }, + { + "epoch": 0.7478315175907293, + "grad_norm": 2.20754075050354, + "learning_rate": 0.00018735652913729272, + "loss": 0.7133, + "step": 10777 + }, + { + "epoch": 0.7479009090278259, + "grad_norm": 2.5778920650482178, + "learning_rate": 0.00018735379438005546, + "loss": 0.8325, + "step": 10778 + }, + { + "epoch": 0.7479703004649226, + "grad_norm": 3.7416131496429443, + "learning_rate": 0.00018735105934705238, + "loss": 0.8568, + "step": 10779 + }, + { + "epoch": 0.7480396919020192, + "grad_norm": 2.923321485519409, + "learning_rate": 0.00018734832403829218, + "loss": 0.79, + "step": 10780 + }, + { + "epoch": 0.748109083339116, + "grad_norm": 2.0668718814849854, + "learning_rate": 0.00018734558845378345, + "loss": 0.5667, + "step": 10781 + }, + { + "epoch": 0.7481784747762126, + "grad_norm": 1.6881080865859985, + "learning_rate": 0.0001873428525935348, + "loss": 0.663, + "step": 10782 + }, + { + "epoch": 0.7482478662133093, + "grad_norm": 2.112201690673828, + "learning_rate": 0.00018734011645755492, + "loss": 0.6434, + "step": 10783 + }, + { + "epoch": 0.7483172576504059, + "grad_norm": 2.6449525356292725, + "learning_rate": 0.00018733738004585239, + "loss": 0.7725, + "step": 10784 + }, + { + "epoch": 0.7483866490875026, + "grad_norm": 2.1992928981781006, + "learning_rate": 0.00018733464335843594, + "loss": 0.7318, + "step": 10785 + }, + { + "epoch": 0.7484560405245992, + "grad_norm": 2.1621923446655273, + "learning_rate": 0.0001873319063953141, + "loss": 0.7483, + "step": 10786 + }, + { + "epoch": 0.748525431961696, + "grad_norm": 2.051495313644409, + "learning_rate": 0.0001873291691564956, + "loss": 0.5796, + "step": 10787 + }, + { + "epoch": 0.7485948233987926, + "grad_norm": 2.546430826187134, + "learning_rate": 0.00018732643164198905, + "loss": 1.0084, + "step": 10788 + }, + { + "epoch": 0.7486642148358893, + "grad_norm": 2.5801751613616943, + "learning_rate": 0.00018732369385180306, + "loss": 0.7933, + "step": 10789 + }, + { + "epoch": 0.7487336062729859, + "grad_norm": 2.162576675415039, + "learning_rate": 0.0001873209557859463, + "loss": 0.8771, + "step": 10790 + }, + { + "epoch": 0.7488029977100826, + "grad_norm": 2.276015281677246, + "learning_rate": 0.00018731821744442744, + "loss": 0.8118, + "step": 10791 + }, + { + "epoch": 0.7488723891471792, + "grad_norm": 2.317826986312866, + "learning_rate": 0.0001873154788272551, + "loss": 0.8461, + "step": 10792 + }, + { + "epoch": 0.7489417805842759, + "grad_norm": 2.5883853435516357, + "learning_rate": 0.00018731273993443788, + "loss": 1.2068, + "step": 10793 + }, + { + "epoch": 0.7490111720213726, + "grad_norm": 1.8013477325439453, + "learning_rate": 0.00018731000076598452, + "loss": 0.7674, + "step": 10794 + }, + { + "epoch": 0.7490805634584692, + "grad_norm": 2.9138355255126953, + "learning_rate": 0.00018730726132190357, + "loss": 1.0465, + "step": 10795 + }, + { + "epoch": 0.7491499548955659, + "grad_norm": 2.072004556655884, + "learning_rate": 0.00018730452160220377, + "loss": 0.8076, + "step": 10796 + }, + { + "epoch": 0.7492193463326625, + "grad_norm": 2.3564069271087646, + "learning_rate": 0.0001873017816068937, + "loss": 0.7959, + "step": 10797 + }, + { + "epoch": 0.7492887377697592, + "grad_norm": 2.426696300506592, + "learning_rate": 0.00018729904133598204, + "loss": 0.7747, + "step": 10798 + }, + { + "epoch": 0.7493581292068558, + "grad_norm": 1.8915293216705322, + "learning_rate": 0.0001872963007894774, + "loss": 0.822, + "step": 10799 + }, + { + "epoch": 0.7494275206439526, + "grad_norm": 2.1565823554992676, + "learning_rate": 0.00018729355996738854, + "loss": 0.9219, + "step": 10800 + }, + { + "epoch": 0.7494969120810492, + "grad_norm": 2.295234203338623, + "learning_rate": 0.00018729081886972399, + "loss": 0.8589, + "step": 10801 + }, + { + "epoch": 0.7495663035181459, + "grad_norm": 2.343989849090576, + "learning_rate": 0.00018728807749649246, + "loss": 0.8583, + "step": 10802 + }, + { + "epoch": 0.7496356949552425, + "grad_norm": 2.2364909648895264, + "learning_rate": 0.0001872853358477026, + "loss": 0.8997, + "step": 10803 + }, + { + "epoch": 0.7497050863923392, + "grad_norm": 2.5681912899017334, + "learning_rate": 0.00018728259392336306, + "loss": 0.7575, + "step": 10804 + }, + { + "epoch": 0.7497744778294358, + "grad_norm": 2.128204345703125, + "learning_rate": 0.00018727985172348248, + "loss": 0.7807, + "step": 10805 + }, + { + "epoch": 0.7498438692665325, + "grad_norm": 2.576371669769287, + "learning_rate": 0.00018727710924806956, + "loss": 1.0133, + "step": 10806 + }, + { + "epoch": 0.7499132607036292, + "grad_norm": 3.0633656978607178, + "learning_rate": 0.00018727436649713293, + "loss": 0.8133, + "step": 10807 + }, + { + "epoch": 0.7499826521407258, + "grad_norm": 2.551030158996582, + "learning_rate": 0.00018727162347068123, + "loss": 0.79, + "step": 10808 + }, + { + "epoch": 0.7500520435778225, + "grad_norm": 2.417992353439331, + "learning_rate": 0.00018726888016872318, + "loss": 0.7524, + "step": 10809 + }, + { + "epoch": 0.7501214350149191, + "grad_norm": 2.7231335639953613, + "learning_rate": 0.00018726613659126736, + "loss": 0.8502, + "step": 10810 + }, + { + "epoch": 0.7501908264520158, + "grad_norm": 2.692666530609131, + "learning_rate": 0.0001872633927383225, + "loss": 0.8566, + "step": 10811 + }, + { + "epoch": 0.7502602178891125, + "grad_norm": 2.191901206970215, + "learning_rate": 0.0001872606486098972, + "loss": 0.8012, + "step": 10812 + }, + { + "epoch": 0.7503296093262092, + "grad_norm": 2.317131996154785, + "learning_rate": 0.00018725790420600017, + "loss": 0.8727, + "step": 10813 + }, + { + "epoch": 0.7503990007633058, + "grad_norm": 2.38215970993042, + "learning_rate": 0.00018725515952664008, + "loss": 0.8023, + "step": 10814 + }, + { + "epoch": 0.7504683922004025, + "grad_norm": 2.2568273544311523, + "learning_rate": 0.00018725241457182557, + "loss": 0.8079, + "step": 10815 + }, + { + "epoch": 0.7505377836374991, + "grad_norm": 2.5349957942962646, + "learning_rate": 0.00018724966934156528, + "loss": 0.8843, + "step": 10816 + }, + { + "epoch": 0.7506071750745958, + "grad_norm": 3.3193836212158203, + "learning_rate": 0.00018724692383586796, + "loss": 0.9566, + "step": 10817 + }, + { + "epoch": 0.7506765665116925, + "grad_norm": 2.1903560161590576, + "learning_rate": 0.00018724417805474216, + "loss": 0.8313, + "step": 10818 + }, + { + "epoch": 0.7507459579487891, + "grad_norm": 3.5669474601745605, + "learning_rate": 0.00018724143199819667, + "loss": 1.0077, + "step": 10819 + }, + { + "epoch": 0.7508153493858858, + "grad_norm": 2.5454800128936768, + "learning_rate": 0.00018723868566624007, + "loss": 0.9351, + "step": 10820 + }, + { + "epoch": 0.7508847408229824, + "grad_norm": 5.057884693145752, + "learning_rate": 0.00018723593905888106, + "loss": 1.0288, + "step": 10821 + }, + { + "epoch": 0.7509541322600791, + "grad_norm": 2.4502511024475098, + "learning_rate": 0.00018723319217612834, + "loss": 0.8996, + "step": 10822 + }, + { + "epoch": 0.7510235236971757, + "grad_norm": 2.216221332550049, + "learning_rate": 0.0001872304450179905, + "loss": 0.8228, + "step": 10823 + }, + { + "epoch": 0.7510929151342725, + "grad_norm": 2.1248230934143066, + "learning_rate": 0.0001872276975844763, + "loss": 0.8738, + "step": 10824 + }, + { + "epoch": 0.7511623065713691, + "grad_norm": 2.2603278160095215, + "learning_rate": 0.0001872249498755944, + "loss": 0.9918, + "step": 10825 + }, + { + "epoch": 0.7512316980084658, + "grad_norm": 2.051172971725464, + "learning_rate": 0.00018722220189135343, + "loss": 0.9639, + "step": 10826 + }, + { + "epoch": 0.7513010894455624, + "grad_norm": 3.440457582473755, + "learning_rate": 0.0001872194536317621, + "loss": 0.9679, + "step": 10827 + }, + { + "epoch": 0.7513704808826591, + "grad_norm": 1.7455254793167114, + "learning_rate": 0.00018721670509682905, + "loss": 0.8641, + "step": 10828 + }, + { + "epoch": 0.7514398723197557, + "grad_norm": 2.548616409301758, + "learning_rate": 0.000187213956286563, + "loss": 0.8163, + "step": 10829 + }, + { + "epoch": 0.7515092637568525, + "grad_norm": 1.975549578666687, + "learning_rate": 0.0001872112072009726, + "loss": 0.7904, + "step": 10830 + }, + { + "epoch": 0.7515786551939491, + "grad_norm": 3.3876686096191406, + "learning_rate": 0.00018720845784006652, + "loss": 0.8325, + "step": 10831 + }, + { + "epoch": 0.7516480466310457, + "grad_norm": 2.331289052963257, + "learning_rate": 0.0001872057082038535, + "loss": 0.7656, + "step": 10832 + }, + { + "epoch": 0.7517174380681424, + "grad_norm": 1.8163036108016968, + "learning_rate": 0.00018720295829234215, + "loss": 0.7754, + "step": 10833 + }, + { + "epoch": 0.751786829505239, + "grad_norm": 2.486130952835083, + "learning_rate": 0.00018720020810554118, + "loss": 0.8932, + "step": 10834 + }, + { + "epoch": 0.7518562209423357, + "grad_norm": 2.425184488296509, + "learning_rate": 0.0001871974576434593, + "loss": 0.7479, + "step": 10835 + }, + { + "epoch": 0.7519256123794323, + "grad_norm": 2.435168981552124, + "learning_rate": 0.0001871947069061051, + "loss": 1.0047, + "step": 10836 + }, + { + "epoch": 0.7519950038165291, + "grad_norm": 2.332958221435547, + "learning_rate": 0.00018719195589348737, + "loss": 0.6762, + "step": 10837 + }, + { + "epoch": 0.7520643952536257, + "grad_norm": 2.1122164726257324, + "learning_rate": 0.00018718920460561474, + "loss": 0.9045, + "step": 10838 + }, + { + "epoch": 0.7521337866907224, + "grad_norm": 2.1085920333862305, + "learning_rate": 0.0001871864530424959, + "loss": 0.853, + "step": 10839 + }, + { + "epoch": 0.752203178127819, + "grad_norm": 2.7311978340148926, + "learning_rate": 0.00018718370120413957, + "loss": 0.8118, + "step": 10840 + }, + { + "epoch": 0.7522725695649157, + "grad_norm": 3.227599859237671, + "learning_rate": 0.00018718094909055436, + "loss": 0.9807, + "step": 10841 + }, + { + "epoch": 0.7523419610020123, + "grad_norm": 2.0203511714935303, + "learning_rate": 0.00018717819670174905, + "loss": 0.9832, + "step": 10842 + }, + { + "epoch": 0.7524113524391091, + "grad_norm": 3.2019026279449463, + "learning_rate": 0.0001871754440377323, + "loss": 1.2454, + "step": 10843 + }, + { + "epoch": 0.7524807438762057, + "grad_norm": 2.470146894454956, + "learning_rate": 0.00018717269109851277, + "loss": 0.9051, + "step": 10844 + }, + { + "epoch": 0.7525501353133024, + "grad_norm": 2.2356138229370117, + "learning_rate": 0.00018716993788409916, + "loss": 0.8101, + "step": 10845 + }, + { + "epoch": 0.752619526750399, + "grad_norm": 2.2073211669921875, + "learning_rate": 0.00018716718439450015, + "loss": 0.8103, + "step": 10846 + }, + { + "epoch": 0.7526889181874956, + "grad_norm": 1.8920384645462036, + "learning_rate": 0.00018716443062972448, + "loss": 1.0494, + "step": 10847 + }, + { + "epoch": 0.7527583096245923, + "grad_norm": 2.6883702278137207, + "learning_rate": 0.00018716167658978084, + "loss": 0.8497, + "step": 10848 + }, + { + "epoch": 0.752827701061689, + "grad_norm": 2.2725753784179688, + "learning_rate": 0.00018715892227467787, + "loss": 0.8271, + "step": 10849 + }, + { + "epoch": 0.7528970924987857, + "grad_norm": 2.1626830101013184, + "learning_rate": 0.00018715616768442427, + "loss": 0.9531, + "step": 10850 + }, + { + "epoch": 0.7529664839358823, + "grad_norm": 2.543811082839966, + "learning_rate": 0.0001871534128190288, + "loss": 1.0524, + "step": 10851 + }, + { + "epoch": 0.753035875372979, + "grad_norm": 2.338050603866577, + "learning_rate": 0.00018715065767850012, + "loss": 0.7078, + "step": 10852 + }, + { + "epoch": 0.7531052668100756, + "grad_norm": 3.114834785461426, + "learning_rate": 0.00018714790226284688, + "loss": 0.832, + "step": 10853 + }, + { + "epoch": 0.7531746582471723, + "grad_norm": 2.6286427974700928, + "learning_rate": 0.00018714514657207788, + "loss": 0.7156, + "step": 10854 + }, + { + "epoch": 0.753244049684269, + "grad_norm": 3.0048537254333496, + "learning_rate": 0.0001871423906062017, + "loss": 1.1536, + "step": 10855 + }, + { + "epoch": 0.7533134411213657, + "grad_norm": 2.6601922512054443, + "learning_rate": 0.00018713963436522714, + "loss": 1.0265, + "step": 10856 + }, + { + "epoch": 0.7533828325584623, + "grad_norm": 2.3030593395233154, + "learning_rate": 0.0001871368778491629, + "loss": 0.6373, + "step": 10857 + }, + { + "epoch": 0.753452223995559, + "grad_norm": 1.80209481716156, + "learning_rate": 0.00018713412105801757, + "loss": 0.6018, + "step": 10858 + }, + { + "epoch": 0.7535216154326556, + "grad_norm": 2.023613691329956, + "learning_rate": 0.0001871313639918, + "loss": 0.7048, + "step": 10859 + }, + { + "epoch": 0.7535910068697522, + "grad_norm": 2.0602004528045654, + "learning_rate": 0.0001871286066505188, + "loss": 0.6929, + "step": 10860 + }, + { + "epoch": 0.7536603983068489, + "grad_norm": 2.021467924118042, + "learning_rate": 0.00018712584903418268, + "loss": 0.8482, + "step": 10861 + }, + { + "epoch": 0.7537297897439456, + "grad_norm": 3.0034689903259277, + "learning_rate": 0.00018712309114280034, + "loss": 0.7809, + "step": 10862 + }, + { + "epoch": 0.7537991811810423, + "grad_norm": 2.47271990776062, + "learning_rate": 0.00018712033297638056, + "loss": 0.8628, + "step": 10863 + }, + { + "epoch": 0.7538685726181389, + "grad_norm": 2.3422858715057373, + "learning_rate": 0.00018711757453493196, + "loss": 0.7838, + "step": 10864 + }, + { + "epoch": 0.7539379640552356, + "grad_norm": 3.0184566974639893, + "learning_rate": 0.00018711481581846332, + "loss": 0.753, + "step": 10865 + }, + { + "epoch": 0.7540073554923322, + "grad_norm": 2.9593756198883057, + "learning_rate": 0.0001871120568269833, + "loss": 1.0376, + "step": 10866 + }, + { + "epoch": 0.7540767469294289, + "grad_norm": 2.2654669284820557, + "learning_rate": 0.00018710929756050063, + "loss": 0.6858, + "step": 10867 + }, + { + "epoch": 0.7541461383665256, + "grad_norm": 1.9405945539474487, + "learning_rate": 0.000187106538019024, + "loss": 0.7348, + "step": 10868 + }, + { + "epoch": 0.7542155298036223, + "grad_norm": 2.302001953125, + "learning_rate": 0.00018710377820256217, + "loss": 0.8821, + "step": 10869 + }, + { + "epoch": 0.7542849212407189, + "grad_norm": 1.7896370887756348, + "learning_rate": 0.00018710101811112378, + "loss": 0.7276, + "step": 10870 + }, + { + "epoch": 0.7543543126778156, + "grad_norm": 2.901287794113159, + "learning_rate": 0.00018709825774471757, + "loss": 0.8108, + "step": 10871 + }, + { + "epoch": 0.7544237041149122, + "grad_norm": 2.446169137954712, + "learning_rate": 0.00018709549710335227, + "loss": 0.9414, + "step": 10872 + }, + { + "epoch": 0.7544930955520088, + "grad_norm": 2.0797829627990723, + "learning_rate": 0.00018709273618703663, + "loss": 0.7114, + "step": 10873 + }, + { + "epoch": 0.7545624869891056, + "grad_norm": 1.7620056867599487, + "learning_rate": 0.0001870899749957793, + "loss": 0.5787, + "step": 10874 + }, + { + "epoch": 0.7546318784262022, + "grad_norm": 2.651176691055298, + "learning_rate": 0.000187087213529589, + "loss": 0.9633, + "step": 10875 + }, + { + "epoch": 0.7547012698632989, + "grad_norm": 1.6950677633285522, + "learning_rate": 0.00018708445178847454, + "loss": 0.7156, + "step": 10876 + }, + { + "epoch": 0.7547706613003955, + "grad_norm": 2.2724180221557617, + "learning_rate": 0.0001870816897724445, + "loss": 0.8501, + "step": 10877 + }, + { + "epoch": 0.7548400527374922, + "grad_norm": 2.184890031814575, + "learning_rate": 0.0001870789274815077, + "loss": 0.8657, + "step": 10878 + }, + { + "epoch": 0.7549094441745888, + "grad_norm": 2.4344425201416016, + "learning_rate": 0.00018707616491567283, + "loss": 0.8409, + "step": 10879 + }, + { + "epoch": 0.7549788356116856, + "grad_norm": 2.069572687149048, + "learning_rate": 0.0001870734020749486, + "loss": 0.6352, + "step": 10880 + }, + { + "epoch": 0.7550482270487822, + "grad_norm": 2.8043618202209473, + "learning_rate": 0.00018707063895934372, + "loss": 0.9383, + "step": 10881 + }, + { + "epoch": 0.7551176184858789, + "grad_norm": 3.265994071960449, + "learning_rate": 0.000187067875568867, + "loss": 0.7876, + "step": 10882 + }, + { + "epoch": 0.7551870099229755, + "grad_norm": 2.326242685317993, + "learning_rate": 0.00018706511190352704, + "loss": 0.8357, + "step": 10883 + }, + { + "epoch": 0.7552564013600722, + "grad_norm": 1.8713353872299194, + "learning_rate": 0.00018706234796333265, + "loss": 0.6559, + "step": 10884 + }, + { + "epoch": 0.7553257927971688, + "grad_norm": 1.8725783824920654, + "learning_rate": 0.00018705958374829251, + "loss": 0.7444, + "step": 10885 + }, + { + "epoch": 0.7553951842342654, + "grad_norm": 2.2950730323791504, + "learning_rate": 0.00018705681925841537, + "loss": 0.8144, + "step": 10886 + }, + { + "epoch": 0.7554645756713622, + "grad_norm": 1.975334644317627, + "learning_rate": 0.00018705405449370997, + "loss": 0.9225, + "step": 10887 + }, + { + "epoch": 0.7555339671084588, + "grad_norm": 2.644031047821045, + "learning_rate": 0.00018705128945418502, + "loss": 1.0036, + "step": 10888 + }, + { + "epoch": 0.7556033585455555, + "grad_norm": 2.4820008277893066, + "learning_rate": 0.00018704852413984924, + "loss": 1.0124, + "step": 10889 + }, + { + "epoch": 0.7556727499826521, + "grad_norm": 1.9117834568023682, + "learning_rate": 0.00018704575855071136, + "loss": 0.8473, + "step": 10890 + }, + { + "epoch": 0.7557421414197488, + "grad_norm": 2.2589354515075684, + "learning_rate": 0.0001870429926867801, + "loss": 0.7507, + "step": 10891 + }, + { + "epoch": 0.7558115328568454, + "grad_norm": 2.712937593460083, + "learning_rate": 0.00018704022654806427, + "loss": 0.9836, + "step": 10892 + }, + { + "epoch": 0.7558809242939422, + "grad_norm": 2.695004463195801, + "learning_rate": 0.00018703746013457248, + "loss": 0.8486, + "step": 10893 + }, + { + "epoch": 0.7559503157310388, + "grad_norm": 2.6202991008758545, + "learning_rate": 0.00018703469344631357, + "loss": 1.0553, + "step": 10894 + }, + { + "epoch": 0.7560197071681355, + "grad_norm": 2.013854742050171, + "learning_rate": 0.00018703192648329622, + "loss": 0.5849, + "step": 10895 + }, + { + "epoch": 0.7560890986052321, + "grad_norm": 3.926093816757202, + "learning_rate": 0.00018702915924552919, + "loss": 1.1503, + "step": 10896 + }, + { + "epoch": 0.7561584900423288, + "grad_norm": 2.0471086502075195, + "learning_rate": 0.00018702639173302118, + "loss": 0.7617, + "step": 10897 + }, + { + "epoch": 0.7562278814794254, + "grad_norm": 2.4736363887786865, + "learning_rate": 0.00018702362394578095, + "loss": 1.057, + "step": 10898 + }, + { + "epoch": 0.7562972729165222, + "grad_norm": 2.143277406692505, + "learning_rate": 0.00018702085588381722, + "loss": 0.6938, + "step": 10899 + }, + { + "epoch": 0.7563666643536188, + "grad_norm": 2.7739241123199463, + "learning_rate": 0.00018701808754713876, + "loss": 0.8729, + "step": 10900 + }, + { + "epoch": 0.7564360557907154, + "grad_norm": 2.3036482334136963, + "learning_rate": 0.0001870153189357543, + "loss": 0.7764, + "step": 10901 + }, + { + "epoch": 0.7565054472278121, + "grad_norm": 1.9293437004089355, + "learning_rate": 0.00018701255004967258, + "loss": 0.8334, + "step": 10902 + }, + { + "epoch": 0.7565748386649087, + "grad_norm": 2.3716695308685303, + "learning_rate": 0.0001870097808889023, + "loss": 1.0583, + "step": 10903 + }, + { + "epoch": 0.7566442301020054, + "grad_norm": 2.0036094188690186, + "learning_rate": 0.00018700701145345228, + "loss": 0.5276, + "step": 10904 + }, + { + "epoch": 0.756713621539102, + "grad_norm": 3.2147610187530518, + "learning_rate": 0.0001870042417433312, + "loss": 0.8955, + "step": 10905 + }, + { + "epoch": 0.7567830129761988, + "grad_norm": 2.427762746810913, + "learning_rate": 0.00018700147175854782, + "loss": 0.8979, + "step": 10906 + }, + { + "epoch": 0.7568524044132954, + "grad_norm": 2.3255200386047363, + "learning_rate": 0.0001869987014991109, + "loss": 0.8687, + "step": 10907 + }, + { + "epoch": 0.7569217958503921, + "grad_norm": 2.192929267883301, + "learning_rate": 0.00018699593096502915, + "loss": 0.7745, + "step": 10908 + }, + { + "epoch": 0.7569911872874887, + "grad_norm": 2.4006502628326416, + "learning_rate": 0.00018699316015631135, + "loss": 0.8557, + "step": 10909 + }, + { + "epoch": 0.7570605787245854, + "grad_norm": 1.6015986204147339, + "learning_rate": 0.00018699038907296625, + "loss": 0.6511, + "step": 10910 + }, + { + "epoch": 0.757129970161682, + "grad_norm": 1.7753902673721313, + "learning_rate": 0.00018698761771500258, + "loss": 0.629, + "step": 10911 + }, + { + "epoch": 0.7571993615987788, + "grad_norm": 2.191725492477417, + "learning_rate": 0.00018698484608242906, + "loss": 0.9093, + "step": 10912 + }, + { + "epoch": 0.7572687530358754, + "grad_norm": 2.546114206314087, + "learning_rate": 0.00018698207417525454, + "loss": 0.9139, + "step": 10913 + }, + { + "epoch": 0.757338144472972, + "grad_norm": 2.230804204940796, + "learning_rate": 0.00018697930199348766, + "loss": 0.7892, + "step": 10914 + }, + { + "epoch": 0.7574075359100687, + "grad_norm": 1.9537382125854492, + "learning_rate": 0.0001869765295371372, + "loss": 0.7729, + "step": 10915 + }, + { + "epoch": 0.7574769273471653, + "grad_norm": 2.05888032913208, + "learning_rate": 0.00018697375680621196, + "loss": 0.7931, + "step": 10916 + }, + { + "epoch": 0.757546318784262, + "grad_norm": 2.7198116779327393, + "learning_rate": 0.00018697098380072064, + "loss": 1.1403, + "step": 10917 + }, + { + "epoch": 0.7576157102213587, + "grad_norm": 2.2928569316864014, + "learning_rate": 0.00018696821052067205, + "loss": 0.8079, + "step": 10918 + }, + { + "epoch": 0.7576851016584554, + "grad_norm": 2.825990915298462, + "learning_rate": 0.00018696543696607485, + "loss": 0.8141, + "step": 10919 + }, + { + "epoch": 0.757754493095552, + "grad_norm": 2.015855550765991, + "learning_rate": 0.00018696266313693788, + "loss": 0.6645, + "step": 10920 + }, + { + "epoch": 0.7578238845326487, + "grad_norm": 1.8910343647003174, + "learning_rate": 0.0001869598890332699, + "loss": 0.7918, + "step": 10921 + }, + { + "epoch": 0.7578932759697453, + "grad_norm": 2.204625129699707, + "learning_rate": 0.0001869571146550796, + "loss": 0.72, + "step": 10922 + }, + { + "epoch": 0.757962667406842, + "grad_norm": 2.3410348892211914, + "learning_rate": 0.0001869543400023758, + "loss": 0.8954, + "step": 10923 + }, + { + "epoch": 0.7580320588439386, + "grad_norm": 2.155090808868408, + "learning_rate": 0.00018695156507516725, + "loss": 0.8619, + "step": 10924 + }, + { + "epoch": 0.7581014502810354, + "grad_norm": 2.598933696746826, + "learning_rate": 0.00018694878987346268, + "loss": 0.8308, + "step": 10925 + }, + { + "epoch": 0.758170841718132, + "grad_norm": 2.1306233406066895, + "learning_rate": 0.00018694601439727087, + "loss": 0.8809, + "step": 10926 + }, + { + "epoch": 0.7582402331552286, + "grad_norm": 2.5102040767669678, + "learning_rate": 0.00018694323864660057, + "loss": 0.6927, + "step": 10927 + }, + { + "epoch": 0.7583096245923253, + "grad_norm": 2.028862476348877, + "learning_rate": 0.00018694046262146059, + "loss": 0.495, + "step": 10928 + }, + { + "epoch": 0.7583790160294219, + "grad_norm": 2.1777021884918213, + "learning_rate": 0.00018693768632185962, + "loss": 0.8429, + "step": 10929 + }, + { + "epoch": 0.7584484074665186, + "grad_norm": 2.3208484649658203, + "learning_rate": 0.00018693490974780645, + "loss": 0.8952, + "step": 10930 + }, + { + "epoch": 0.7585177989036153, + "grad_norm": 2.28924560546875, + "learning_rate": 0.0001869321328993099, + "loss": 0.8468, + "step": 10931 + }, + { + "epoch": 0.758587190340712, + "grad_norm": 2.2255733013153076, + "learning_rate": 0.00018692935577637867, + "loss": 0.6782, + "step": 10932 + }, + { + "epoch": 0.7586565817778086, + "grad_norm": 3.071363687515259, + "learning_rate": 0.00018692657837902153, + "loss": 0.8049, + "step": 10933 + }, + { + "epoch": 0.7587259732149053, + "grad_norm": 2.6334493160247803, + "learning_rate": 0.00018692380070724728, + "loss": 0.8953, + "step": 10934 + }, + { + "epoch": 0.7587953646520019, + "grad_norm": 2.820169448852539, + "learning_rate": 0.0001869210227610647, + "loss": 0.8362, + "step": 10935 + }, + { + "epoch": 0.7588647560890986, + "grad_norm": 2.3372514247894287, + "learning_rate": 0.0001869182445404825, + "loss": 0.822, + "step": 10936 + }, + { + "epoch": 0.7589341475261953, + "grad_norm": 3.7199997901916504, + "learning_rate": 0.0001869154660455095, + "loss": 0.8063, + "step": 10937 + }, + { + "epoch": 0.759003538963292, + "grad_norm": 3.162513017654419, + "learning_rate": 0.0001869126872761545, + "loss": 1.0034, + "step": 10938 + }, + { + "epoch": 0.7590729304003886, + "grad_norm": 1.6766585111618042, + "learning_rate": 0.00018690990823242617, + "loss": 0.6303, + "step": 10939 + }, + { + "epoch": 0.7591423218374852, + "grad_norm": 2.0570528507232666, + "learning_rate": 0.00018690712891433338, + "loss": 0.7259, + "step": 10940 + }, + { + "epoch": 0.7592117132745819, + "grad_norm": 2.84755539894104, + "learning_rate": 0.00018690434932188487, + "loss": 0.9319, + "step": 10941 + }, + { + "epoch": 0.7592811047116785, + "grad_norm": 1.8517611026763916, + "learning_rate": 0.00018690156945508937, + "loss": 0.6544, + "step": 10942 + }, + { + "epoch": 0.7593504961487753, + "grad_norm": 2.5396177768707275, + "learning_rate": 0.00018689878931395573, + "loss": 0.8992, + "step": 10943 + }, + { + "epoch": 0.7594198875858719, + "grad_norm": 2.169252872467041, + "learning_rate": 0.00018689600889849265, + "loss": 0.8932, + "step": 10944 + }, + { + "epoch": 0.7594892790229686, + "grad_norm": 1.7063051462173462, + "learning_rate": 0.000186893228208709, + "loss": 0.7607, + "step": 10945 + }, + { + "epoch": 0.7595586704600652, + "grad_norm": 2.354435682296753, + "learning_rate": 0.0001868904472446135, + "loss": 0.7242, + "step": 10946 + }, + { + "epoch": 0.7596280618971619, + "grad_norm": 2.1009857654571533, + "learning_rate": 0.00018688766600621494, + "loss": 0.7187, + "step": 10947 + }, + { + "epoch": 0.7596974533342585, + "grad_norm": 2.0375258922576904, + "learning_rate": 0.0001868848844935221, + "loss": 0.7015, + "step": 10948 + }, + { + "epoch": 0.7597668447713553, + "grad_norm": 2.240294933319092, + "learning_rate": 0.00018688210270654377, + "loss": 0.9849, + "step": 10949 + }, + { + "epoch": 0.7598362362084519, + "grad_norm": 3.1183133125305176, + "learning_rate": 0.0001868793206452887, + "loss": 0.9121, + "step": 10950 + }, + { + "epoch": 0.7599056276455486, + "grad_norm": 2.2805089950561523, + "learning_rate": 0.0001868765383097657, + "loss": 0.7297, + "step": 10951 + }, + { + "epoch": 0.7599750190826452, + "grad_norm": 2.05490779876709, + "learning_rate": 0.00018687375569998354, + "loss": 0.7417, + "step": 10952 + }, + { + "epoch": 0.7600444105197418, + "grad_norm": 2.4253556728363037, + "learning_rate": 0.00018687097281595104, + "loss": 0.8136, + "step": 10953 + }, + { + "epoch": 0.7601138019568385, + "grad_norm": 2.4150071144104004, + "learning_rate": 0.00018686818965767693, + "loss": 1.0698, + "step": 10954 + }, + { + "epoch": 0.7601831933939351, + "grad_norm": 1.9417067766189575, + "learning_rate": 0.00018686540622517005, + "loss": 0.9321, + "step": 10955 + }, + { + "epoch": 0.7602525848310319, + "grad_norm": 3.4313061237335205, + "learning_rate": 0.00018686262251843913, + "loss": 0.9354, + "step": 10956 + }, + { + "epoch": 0.7603219762681285, + "grad_norm": 2.211261034011841, + "learning_rate": 0.00018685983853749302, + "loss": 0.7809, + "step": 10957 + }, + { + "epoch": 0.7603913677052252, + "grad_norm": 2.1416003704071045, + "learning_rate": 0.00018685705428234047, + "loss": 0.803, + "step": 10958 + }, + { + "epoch": 0.7604607591423218, + "grad_norm": 2.8766028881073, + "learning_rate": 0.00018685426975299026, + "loss": 0.8187, + "step": 10959 + }, + { + "epoch": 0.7605301505794185, + "grad_norm": 2.261302947998047, + "learning_rate": 0.0001868514849494512, + "loss": 0.8321, + "step": 10960 + }, + { + "epoch": 0.7605995420165151, + "grad_norm": 2.361849784851074, + "learning_rate": 0.00018684869987173209, + "loss": 0.7564, + "step": 10961 + }, + { + "epoch": 0.7606689334536119, + "grad_norm": 2.4008922576904297, + "learning_rate": 0.0001868459145198417, + "loss": 1.0558, + "step": 10962 + }, + { + "epoch": 0.7607383248907085, + "grad_norm": 2.3172082901000977, + "learning_rate": 0.00018684312889378883, + "loss": 0.954, + "step": 10963 + }, + { + "epoch": 0.7608077163278052, + "grad_norm": 1.8848930597305298, + "learning_rate": 0.0001868403429935823, + "loss": 0.5899, + "step": 10964 + }, + { + "epoch": 0.7608771077649018, + "grad_norm": 2.616338014602661, + "learning_rate": 0.0001868375568192309, + "loss": 1.0388, + "step": 10965 + }, + { + "epoch": 0.7609464992019985, + "grad_norm": 2.2485671043395996, + "learning_rate": 0.00018683477037074333, + "loss": 0.8356, + "step": 10966 + }, + { + "epoch": 0.7610158906390951, + "grad_norm": 1.9967585802078247, + "learning_rate": 0.00018683198364812854, + "loss": 0.8799, + "step": 10967 + }, + { + "epoch": 0.7610852820761917, + "grad_norm": 2.5631446838378906, + "learning_rate": 0.0001868291966513952, + "loss": 0.9856, + "step": 10968 + }, + { + "epoch": 0.7611546735132885, + "grad_norm": 2.413618564605713, + "learning_rate": 0.0001868264093805522, + "loss": 0.6508, + "step": 10969 + }, + { + "epoch": 0.7612240649503851, + "grad_norm": 2.197986602783203, + "learning_rate": 0.00018682362183560826, + "loss": 0.7252, + "step": 10970 + }, + { + "epoch": 0.7612934563874818, + "grad_norm": 2.2507894039154053, + "learning_rate": 0.00018682083401657228, + "loss": 0.7821, + "step": 10971 + }, + { + "epoch": 0.7613628478245784, + "grad_norm": 2.925508975982666, + "learning_rate": 0.00018681804592345295, + "loss": 0.9515, + "step": 10972 + }, + { + "epoch": 0.7614322392616751, + "grad_norm": 2.6577208042144775, + "learning_rate": 0.00018681525755625912, + "loss": 1.2021, + "step": 10973 + }, + { + "epoch": 0.7615016306987717, + "grad_norm": 2.040403127670288, + "learning_rate": 0.00018681246891499962, + "loss": 0.7902, + "step": 10974 + }, + { + "epoch": 0.7615710221358685, + "grad_norm": 2.693328857421875, + "learning_rate": 0.0001868096799996832, + "loss": 0.8198, + "step": 10975 + }, + { + "epoch": 0.7616404135729651, + "grad_norm": 2.581591844558716, + "learning_rate": 0.0001868068908103187, + "loss": 0.9415, + "step": 10976 + }, + { + "epoch": 0.7617098050100618, + "grad_norm": 2.2986252307891846, + "learning_rate": 0.00018680410134691496, + "loss": 0.7138, + "step": 10977 + }, + { + "epoch": 0.7617791964471584, + "grad_norm": 1.9985243082046509, + "learning_rate": 0.0001868013116094807, + "loss": 0.624, + "step": 10978 + }, + { + "epoch": 0.7618485878842551, + "grad_norm": 2.496829032897949, + "learning_rate": 0.0001867985215980248, + "loss": 0.8489, + "step": 10979 + }, + { + "epoch": 0.7619179793213517, + "grad_norm": 3.84631085395813, + "learning_rate": 0.000186795731312556, + "loss": 0.9384, + "step": 10980 + }, + { + "epoch": 0.7619873707584484, + "grad_norm": 1.8580626249313354, + "learning_rate": 0.00018679294075308315, + "loss": 0.6845, + "step": 10981 + }, + { + "epoch": 0.7620567621955451, + "grad_norm": 2.3188865184783936, + "learning_rate": 0.00018679014991961507, + "loss": 0.8511, + "step": 10982 + }, + { + "epoch": 0.7621261536326417, + "grad_norm": 2.693817138671875, + "learning_rate": 0.00018678735881216057, + "loss": 0.9941, + "step": 10983 + }, + { + "epoch": 0.7621955450697384, + "grad_norm": 2.9526283740997314, + "learning_rate": 0.00018678456743072844, + "loss": 1.0874, + "step": 10984 + }, + { + "epoch": 0.762264936506835, + "grad_norm": 2.4038989543914795, + "learning_rate": 0.00018678177577532748, + "loss": 1.0538, + "step": 10985 + }, + { + "epoch": 0.7623343279439317, + "grad_norm": 1.9109563827514648, + "learning_rate": 0.00018677898384596655, + "loss": 0.747, + "step": 10986 + }, + { + "epoch": 0.7624037193810284, + "grad_norm": 1.8995649814605713, + "learning_rate": 0.0001867761916426544, + "loss": 0.5149, + "step": 10987 + }, + { + "epoch": 0.7624731108181251, + "grad_norm": 2.083730936050415, + "learning_rate": 0.00018677339916539992, + "loss": 0.8797, + "step": 10988 + }, + { + "epoch": 0.7625425022552217, + "grad_norm": 1.8011143207550049, + "learning_rate": 0.00018677060641421184, + "loss": 0.844, + "step": 10989 + }, + { + "epoch": 0.7626118936923184, + "grad_norm": 3.104691505432129, + "learning_rate": 0.00018676781338909909, + "loss": 0.9236, + "step": 10990 + }, + { + "epoch": 0.762681285129415, + "grad_norm": 3.3408961296081543, + "learning_rate": 0.00018676502009007035, + "loss": 1.1772, + "step": 10991 + }, + { + "epoch": 0.7627506765665117, + "grad_norm": 2.423534393310547, + "learning_rate": 0.00018676222651713454, + "loss": 0.8752, + "step": 10992 + }, + { + "epoch": 0.7628200680036084, + "grad_norm": 2.1250648498535156, + "learning_rate": 0.00018675943267030046, + "loss": 0.9332, + "step": 10993 + }, + { + "epoch": 0.762889459440705, + "grad_norm": 2.8994243144989014, + "learning_rate": 0.00018675663854957693, + "loss": 0.9012, + "step": 10994 + }, + { + "epoch": 0.7629588508778017, + "grad_norm": 2.845458507537842, + "learning_rate": 0.00018675384415497274, + "loss": 0.8295, + "step": 10995 + }, + { + "epoch": 0.7630282423148983, + "grad_norm": 2.224851608276367, + "learning_rate": 0.0001867510494864967, + "loss": 0.9213, + "step": 10996 + }, + { + "epoch": 0.763097633751995, + "grad_norm": 1.8721086978912354, + "learning_rate": 0.00018674825454415768, + "loss": 1.0097, + "step": 10997 + }, + { + "epoch": 0.7631670251890916, + "grad_norm": 2.137537717819214, + "learning_rate": 0.0001867454593279645, + "loss": 0.9735, + "step": 10998 + }, + { + "epoch": 0.7632364166261884, + "grad_norm": 1.7204477787017822, + "learning_rate": 0.00018674266383792592, + "loss": 0.6986, + "step": 10999 + }, + { + "epoch": 0.763305808063285, + "grad_norm": 1.9898900985717773, + "learning_rate": 0.00018673986807405085, + "loss": 0.8664, + "step": 11000 + }, + { + "epoch": 0.7633751995003817, + "grad_norm": 2.0892815589904785, + "learning_rate": 0.0001867370720363481, + "loss": 0.7436, + "step": 11001 + }, + { + "epoch": 0.7634445909374783, + "grad_norm": 2.0514166355133057, + "learning_rate": 0.00018673427572482644, + "loss": 0.7797, + "step": 11002 + }, + { + "epoch": 0.763513982374575, + "grad_norm": 3.3393006324768066, + "learning_rate": 0.00018673147913949477, + "loss": 0.9944, + "step": 11003 + }, + { + "epoch": 0.7635833738116716, + "grad_norm": 2.317737579345703, + "learning_rate": 0.00018672868228036187, + "loss": 0.9641, + "step": 11004 + }, + { + "epoch": 0.7636527652487684, + "grad_norm": 1.8807913064956665, + "learning_rate": 0.00018672588514743657, + "loss": 0.7189, + "step": 11005 + }, + { + "epoch": 0.763722156685865, + "grad_norm": 2.9810609817504883, + "learning_rate": 0.00018672308774072774, + "loss": 0.884, + "step": 11006 + }, + { + "epoch": 0.7637915481229616, + "grad_norm": 2.2158472537994385, + "learning_rate": 0.00018672029006024414, + "loss": 0.7307, + "step": 11007 + }, + { + "epoch": 0.7638609395600583, + "grad_norm": 2.5635788440704346, + "learning_rate": 0.0001867174921059947, + "loss": 0.6804, + "step": 11008 + }, + { + "epoch": 0.7639303309971549, + "grad_norm": 2.3452327251434326, + "learning_rate": 0.00018671469387798815, + "loss": 0.8611, + "step": 11009 + }, + { + "epoch": 0.7639997224342516, + "grad_norm": 2.2511537075042725, + "learning_rate": 0.0001867118953762334, + "loss": 0.6664, + "step": 11010 + }, + { + "epoch": 0.7640691138713482, + "grad_norm": 2.288475751876831, + "learning_rate": 0.00018670909660073925, + "loss": 0.7205, + "step": 11011 + }, + { + "epoch": 0.764138505308445, + "grad_norm": 1.698255181312561, + "learning_rate": 0.00018670629755151454, + "loss": 0.6986, + "step": 11012 + }, + { + "epoch": 0.7642078967455416, + "grad_norm": 3.236593246459961, + "learning_rate": 0.0001867034982285681, + "loss": 0.9126, + "step": 11013 + }, + { + "epoch": 0.7642772881826383, + "grad_norm": 2.6146090030670166, + "learning_rate": 0.00018670069863190876, + "loss": 1.1745, + "step": 11014 + }, + { + "epoch": 0.7643466796197349, + "grad_norm": 2.3769078254699707, + "learning_rate": 0.00018669789876154543, + "loss": 0.9075, + "step": 11015 + }, + { + "epoch": 0.7644160710568316, + "grad_norm": 2.379004955291748, + "learning_rate": 0.00018669509861748686, + "loss": 0.759, + "step": 11016 + }, + { + "epoch": 0.7644854624939282, + "grad_norm": 1.851352572441101, + "learning_rate": 0.0001866922981997419, + "loss": 0.7454, + "step": 11017 + }, + { + "epoch": 0.764554853931025, + "grad_norm": 3.2765984535217285, + "learning_rate": 0.00018668949750831945, + "loss": 0.8881, + "step": 11018 + }, + { + "epoch": 0.7646242453681216, + "grad_norm": 2.2735798358917236, + "learning_rate": 0.0001866866965432283, + "loss": 0.8569, + "step": 11019 + }, + { + "epoch": 0.7646936368052183, + "grad_norm": 1.7540024518966675, + "learning_rate": 0.0001866838953044773, + "loss": 0.6961, + "step": 11020 + }, + { + "epoch": 0.7647630282423149, + "grad_norm": 1.7266765832901, + "learning_rate": 0.00018668109379207533, + "loss": 0.6854, + "step": 11021 + }, + { + "epoch": 0.7648324196794115, + "grad_norm": 1.835447072982788, + "learning_rate": 0.00018667829200603116, + "loss": 0.5289, + "step": 11022 + }, + { + "epoch": 0.7649018111165082, + "grad_norm": 1.9721744060516357, + "learning_rate": 0.00018667548994635372, + "loss": 0.8023, + "step": 11023 + }, + { + "epoch": 0.7649712025536048, + "grad_norm": 2.5770788192749023, + "learning_rate": 0.0001866726876130518, + "loss": 0.7117, + "step": 11024 + }, + { + "epoch": 0.7650405939907016, + "grad_norm": 1.979734182357788, + "learning_rate": 0.00018666988500613422, + "loss": 0.7651, + "step": 11025 + }, + { + "epoch": 0.7651099854277982, + "grad_norm": 2.458188772201538, + "learning_rate": 0.0001866670821256099, + "loss": 0.921, + "step": 11026 + }, + { + "epoch": 0.7651793768648949, + "grad_norm": 1.8202414512634277, + "learning_rate": 0.0001866642789714877, + "loss": 0.6649, + "step": 11027 + }, + { + "epoch": 0.7652487683019915, + "grad_norm": 2.026901960372925, + "learning_rate": 0.0001866614755437764, + "loss": 0.8232, + "step": 11028 + }, + { + "epoch": 0.7653181597390882, + "grad_norm": 2.6360390186309814, + "learning_rate": 0.00018665867184248483, + "loss": 1.0027, + "step": 11029 + }, + { + "epoch": 0.7653875511761848, + "grad_norm": 2.3877484798431396, + "learning_rate": 0.00018665586786762193, + "loss": 0.814, + "step": 11030 + }, + { + "epoch": 0.7654569426132816, + "grad_norm": 2.0865375995635986, + "learning_rate": 0.00018665306361919648, + "loss": 0.6665, + "step": 11031 + }, + { + "epoch": 0.7655263340503782, + "grad_norm": 2.1879992485046387, + "learning_rate": 0.00018665025909721735, + "loss": 0.7958, + "step": 11032 + }, + { + "epoch": 0.7655957254874749, + "grad_norm": 2.3583791255950928, + "learning_rate": 0.00018664745430169345, + "loss": 0.8423, + "step": 11033 + }, + { + "epoch": 0.7656651169245715, + "grad_norm": 2.5675618648529053, + "learning_rate": 0.00018664464923263358, + "loss": 1.0018, + "step": 11034 + }, + { + "epoch": 0.7657345083616681, + "grad_norm": 2.6137986183166504, + "learning_rate": 0.00018664184389004663, + "loss": 0.877, + "step": 11035 + }, + { + "epoch": 0.7658038997987648, + "grad_norm": 2.0224673748016357, + "learning_rate": 0.0001866390382739414, + "loss": 0.6521, + "step": 11036 + }, + { + "epoch": 0.7658732912358615, + "grad_norm": 2.8428962230682373, + "learning_rate": 0.00018663623238432674, + "loss": 1.2297, + "step": 11037 + }, + { + "epoch": 0.7659426826729582, + "grad_norm": 2.930567502975464, + "learning_rate": 0.0001866334262212116, + "loss": 0.6869, + "step": 11038 + }, + { + "epoch": 0.7660120741100548, + "grad_norm": 2.5814239978790283, + "learning_rate": 0.0001866306197846048, + "loss": 0.7199, + "step": 11039 + }, + { + "epoch": 0.7660814655471515, + "grad_norm": 2.451979637145996, + "learning_rate": 0.0001866278130745151, + "loss": 0.7871, + "step": 11040 + }, + { + "epoch": 0.7661508569842481, + "grad_norm": 2.401322364807129, + "learning_rate": 0.00018662500609095152, + "loss": 0.8625, + "step": 11041 + }, + { + "epoch": 0.7662202484213448, + "grad_norm": 2.0015320777893066, + "learning_rate": 0.00018662219883392283, + "loss": 0.8156, + "step": 11042 + }, + { + "epoch": 0.7662896398584415, + "grad_norm": 2.4703614711761475, + "learning_rate": 0.0001866193913034379, + "loss": 0.9245, + "step": 11043 + }, + { + "epoch": 0.7663590312955382, + "grad_norm": 3.57639479637146, + "learning_rate": 0.00018661658349950562, + "loss": 0.7894, + "step": 11044 + }, + { + "epoch": 0.7664284227326348, + "grad_norm": 2.185988664627075, + "learning_rate": 0.00018661377542213484, + "loss": 0.9332, + "step": 11045 + }, + { + "epoch": 0.7664978141697315, + "grad_norm": 2.3018836975097656, + "learning_rate": 0.0001866109670713344, + "loss": 1.0324, + "step": 11046 + }, + { + "epoch": 0.7665672056068281, + "grad_norm": 2.5934810638427734, + "learning_rate": 0.00018660815844711315, + "loss": 0.826, + "step": 11047 + }, + { + "epoch": 0.7666365970439247, + "grad_norm": 1.686028003692627, + "learning_rate": 0.00018660534954948003, + "loss": 0.7923, + "step": 11048 + }, + { + "epoch": 0.7667059884810214, + "grad_norm": 1.9847311973571777, + "learning_rate": 0.00018660254037844388, + "loss": 0.7668, + "step": 11049 + }, + { + "epoch": 0.7667753799181181, + "grad_norm": 2.0425527095794678, + "learning_rate": 0.00018659973093401356, + "loss": 0.8654, + "step": 11050 + }, + { + "epoch": 0.7668447713552148, + "grad_norm": 2.2228214740753174, + "learning_rate": 0.0001865969212161979, + "loss": 0.7849, + "step": 11051 + }, + { + "epoch": 0.7669141627923114, + "grad_norm": 2.1466550827026367, + "learning_rate": 0.00018659411122500582, + "loss": 0.8972, + "step": 11052 + }, + { + "epoch": 0.7669835542294081, + "grad_norm": 3.0536794662475586, + "learning_rate": 0.00018659130096044618, + "loss": 0.9282, + "step": 11053 + }, + { + "epoch": 0.7670529456665047, + "grad_norm": 2.967320680618286, + "learning_rate": 0.00018658849042252787, + "loss": 0.771, + "step": 11054 + }, + { + "epoch": 0.7671223371036014, + "grad_norm": 1.54647696018219, + "learning_rate": 0.00018658567961125973, + "loss": 0.718, + "step": 11055 + }, + { + "epoch": 0.7671917285406981, + "grad_norm": 2.1235275268554688, + "learning_rate": 0.00018658286852665063, + "loss": 0.843, + "step": 11056 + }, + { + "epoch": 0.7672611199777948, + "grad_norm": 2.0470783710479736, + "learning_rate": 0.00018658005716870948, + "loss": 0.8981, + "step": 11057 + }, + { + "epoch": 0.7673305114148914, + "grad_norm": 2.6002445220947266, + "learning_rate": 0.00018657724553744514, + "loss": 0.7021, + "step": 11058 + }, + { + "epoch": 0.7673999028519881, + "grad_norm": 2.0573365688323975, + "learning_rate": 0.00018657443363286646, + "loss": 0.8258, + "step": 11059 + }, + { + "epoch": 0.7674692942890847, + "grad_norm": 2.1103737354278564, + "learning_rate": 0.00018657162145498235, + "loss": 0.7589, + "step": 11060 + }, + { + "epoch": 0.7675386857261813, + "grad_norm": 2.0063178539276123, + "learning_rate": 0.00018656880900380166, + "loss": 0.7194, + "step": 11061 + }, + { + "epoch": 0.7676080771632781, + "grad_norm": 2.4645252227783203, + "learning_rate": 0.0001865659962793333, + "loss": 1.0205, + "step": 11062 + }, + { + "epoch": 0.7676774686003747, + "grad_norm": 2.106276035308838, + "learning_rate": 0.00018656318328158614, + "loss": 0.7618, + "step": 11063 + }, + { + "epoch": 0.7677468600374714, + "grad_norm": 2.173022985458374, + "learning_rate": 0.00018656037001056906, + "loss": 0.6907, + "step": 11064 + }, + { + "epoch": 0.767816251474568, + "grad_norm": 3.1229135990142822, + "learning_rate": 0.0001865575564662909, + "loss": 1.2026, + "step": 11065 + }, + { + "epoch": 0.7678856429116647, + "grad_norm": 1.8289506435394287, + "learning_rate": 0.0001865547426487606, + "loss": 0.7499, + "step": 11066 + }, + { + "epoch": 0.7679550343487613, + "grad_norm": 2.1700127124786377, + "learning_rate": 0.000186551928557987, + "loss": 0.7517, + "step": 11067 + }, + { + "epoch": 0.7680244257858581, + "grad_norm": 2.8146533966064453, + "learning_rate": 0.00018654911419397903, + "loss": 1.0659, + "step": 11068 + }, + { + "epoch": 0.7680938172229547, + "grad_norm": 1.8788456916809082, + "learning_rate": 0.0001865462995567455, + "loss": 0.6415, + "step": 11069 + }, + { + "epoch": 0.7681632086600514, + "grad_norm": 1.9610744714736938, + "learning_rate": 0.0001865434846462954, + "loss": 0.6841, + "step": 11070 + }, + { + "epoch": 0.768232600097148, + "grad_norm": 2.437566041946411, + "learning_rate": 0.00018654066946263752, + "loss": 0.7104, + "step": 11071 + }, + { + "epoch": 0.7683019915342447, + "grad_norm": 2.1798532009124756, + "learning_rate": 0.0001865378540057808, + "loss": 0.6966, + "step": 11072 + }, + { + "epoch": 0.7683713829713413, + "grad_norm": 3.066359043121338, + "learning_rate": 0.00018653503827573413, + "loss": 0.7931, + "step": 11073 + }, + { + "epoch": 0.7684407744084379, + "grad_norm": 2.819352388381958, + "learning_rate": 0.00018653222227250639, + "loss": 0.7463, + "step": 11074 + }, + { + "epoch": 0.7685101658455347, + "grad_norm": 2.704796075820923, + "learning_rate": 0.00018652940599610645, + "loss": 0.7531, + "step": 11075 + }, + { + "epoch": 0.7685795572826313, + "grad_norm": 2.61283802986145, + "learning_rate": 0.0001865265894465432, + "loss": 0.9324, + "step": 11076 + }, + { + "epoch": 0.768648948719728, + "grad_norm": 2.5488979816436768, + "learning_rate": 0.00018652377262382556, + "loss": 1.0554, + "step": 11077 + }, + { + "epoch": 0.7687183401568246, + "grad_norm": 2.4567441940307617, + "learning_rate": 0.00018652095552796241, + "loss": 0.8533, + "step": 11078 + }, + { + "epoch": 0.7687877315939213, + "grad_norm": 1.759080171585083, + "learning_rate": 0.00018651813815896264, + "loss": 0.6058, + "step": 11079 + }, + { + "epoch": 0.7688571230310179, + "grad_norm": 2.6632747650146484, + "learning_rate": 0.00018651532051683516, + "loss": 0.9887, + "step": 11080 + }, + { + "epoch": 0.7689265144681147, + "grad_norm": 2.7135350704193115, + "learning_rate": 0.0001865125026015888, + "loss": 1.1738, + "step": 11081 + }, + { + "epoch": 0.7689959059052113, + "grad_norm": 1.840832233428955, + "learning_rate": 0.0001865096844132326, + "loss": 0.7924, + "step": 11082 + }, + { + "epoch": 0.769065297342308, + "grad_norm": 2.8262135982513428, + "learning_rate": 0.00018650686595177526, + "loss": 0.6657, + "step": 11083 + }, + { + "epoch": 0.7691346887794046, + "grad_norm": 2.2957561016082764, + "learning_rate": 0.00018650404721722583, + "loss": 0.7894, + "step": 11084 + }, + { + "epoch": 0.7692040802165013, + "grad_norm": 1.876253366470337, + "learning_rate": 0.00018650122820959317, + "loss": 0.7784, + "step": 11085 + }, + { + "epoch": 0.7692734716535979, + "grad_norm": 1.9037059545516968, + "learning_rate": 0.00018649840892888614, + "loss": 0.7298, + "step": 11086 + }, + { + "epoch": 0.7693428630906947, + "grad_norm": 2.116647243499756, + "learning_rate": 0.0001864955893751137, + "loss": 0.7071, + "step": 11087 + }, + { + "epoch": 0.7694122545277913, + "grad_norm": 4.808163642883301, + "learning_rate": 0.00018649276954828468, + "loss": 0.8434, + "step": 11088 + }, + { + "epoch": 0.7694816459648879, + "grad_norm": 2.391331195831299, + "learning_rate": 0.00018648994944840805, + "loss": 0.9027, + "step": 11089 + }, + { + "epoch": 0.7695510374019846, + "grad_norm": 1.9259886741638184, + "learning_rate": 0.0001864871290754927, + "loss": 0.826, + "step": 11090 + }, + { + "epoch": 0.7696204288390812, + "grad_norm": 2.1136109828948975, + "learning_rate": 0.00018648430842954747, + "loss": 0.632, + "step": 11091 + }, + { + "epoch": 0.7696898202761779, + "grad_norm": 2.285426139831543, + "learning_rate": 0.00018648148751058133, + "loss": 0.804, + "step": 11092 + }, + { + "epoch": 0.7697592117132745, + "grad_norm": 1.985087513923645, + "learning_rate": 0.0001864786663186032, + "loss": 0.7038, + "step": 11093 + }, + { + "epoch": 0.7698286031503713, + "grad_norm": 1.5711420774459839, + "learning_rate": 0.0001864758448536219, + "loss": 0.6515, + "step": 11094 + }, + { + "epoch": 0.7698979945874679, + "grad_norm": 1.6021393537521362, + "learning_rate": 0.00018647302311564643, + "loss": 0.707, + "step": 11095 + }, + { + "epoch": 0.7699673860245646, + "grad_norm": 2.1828629970550537, + "learning_rate": 0.00018647020110468566, + "loss": 1.1176, + "step": 11096 + }, + { + "epoch": 0.7700367774616612, + "grad_norm": 3.3030991554260254, + "learning_rate": 0.0001864673788207485, + "loss": 0.7438, + "step": 11097 + }, + { + "epoch": 0.7701061688987579, + "grad_norm": 2.332317352294922, + "learning_rate": 0.00018646455626384383, + "loss": 0.7422, + "step": 11098 + }, + { + "epoch": 0.7701755603358545, + "grad_norm": 2.252903938293457, + "learning_rate": 0.0001864617334339806, + "loss": 0.7188, + "step": 11099 + }, + { + "epoch": 0.7702449517729513, + "grad_norm": 1.9093632698059082, + "learning_rate": 0.0001864589103311677, + "loss": 0.7437, + "step": 11100 + }, + { + "epoch": 0.7703143432100479, + "grad_norm": 2.2578961849212646, + "learning_rate": 0.00018645608695541408, + "loss": 1.2044, + "step": 11101 + }, + { + "epoch": 0.7703837346471445, + "grad_norm": 2.3479459285736084, + "learning_rate": 0.00018645326330672858, + "loss": 0.951, + "step": 11102 + }, + { + "epoch": 0.7704531260842412, + "grad_norm": 2.1014654636383057, + "learning_rate": 0.00018645043938512023, + "loss": 0.7119, + "step": 11103 + }, + { + "epoch": 0.7705225175213378, + "grad_norm": 1.9919626712799072, + "learning_rate": 0.0001864476151905978, + "loss": 0.8117, + "step": 11104 + }, + { + "epoch": 0.7705919089584345, + "grad_norm": 2.2460556030273438, + "learning_rate": 0.00018644479072317028, + "loss": 0.742, + "step": 11105 + }, + { + "epoch": 0.7706613003955312, + "grad_norm": 2.213737726211548, + "learning_rate": 0.00018644196598284664, + "loss": 0.7134, + "step": 11106 + }, + { + "epoch": 0.7707306918326279, + "grad_norm": 2.5032799243927, + "learning_rate": 0.0001864391409696357, + "loss": 1.035, + "step": 11107 + }, + { + "epoch": 0.7708000832697245, + "grad_norm": 2.8718643188476562, + "learning_rate": 0.00018643631568354644, + "loss": 1.0436, + "step": 11108 + }, + { + "epoch": 0.7708694747068212, + "grad_norm": 2.014707565307617, + "learning_rate": 0.00018643349012458776, + "loss": 0.8601, + "step": 11109 + }, + { + "epoch": 0.7709388661439178, + "grad_norm": 2.433760404586792, + "learning_rate": 0.00018643066429276857, + "loss": 0.8755, + "step": 11110 + }, + { + "epoch": 0.7710082575810145, + "grad_norm": 2.3641769886016846, + "learning_rate": 0.0001864278381880978, + "loss": 0.789, + "step": 11111 + }, + { + "epoch": 0.7710776490181112, + "grad_norm": 2.757568120956421, + "learning_rate": 0.00018642501181058438, + "loss": 0.7851, + "step": 11112 + }, + { + "epoch": 0.7711470404552079, + "grad_norm": 2.3299942016601562, + "learning_rate": 0.0001864221851602372, + "loss": 0.758, + "step": 11113 + }, + { + "epoch": 0.7712164318923045, + "grad_norm": 1.563315510749817, + "learning_rate": 0.00018641935823706523, + "loss": 0.9326, + "step": 11114 + }, + { + "epoch": 0.7712858233294011, + "grad_norm": 1.842273235321045, + "learning_rate": 0.0001864165310410774, + "loss": 0.615, + "step": 11115 + }, + { + "epoch": 0.7713552147664978, + "grad_norm": 2.024441957473755, + "learning_rate": 0.00018641370357228257, + "loss": 0.5854, + "step": 11116 + }, + { + "epoch": 0.7714246062035944, + "grad_norm": 2.250988721847534, + "learning_rate": 0.00018641087583068974, + "loss": 0.7409, + "step": 11117 + }, + { + "epoch": 0.7714939976406912, + "grad_norm": 2.3834638595581055, + "learning_rate": 0.00018640804781630777, + "loss": 0.9049, + "step": 11118 + }, + { + "epoch": 0.7715633890777878, + "grad_norm": 1.9768372774124146, + "learning_rate": 0.0001864052195291456, + "loss": 0.5921, + "step": 11119 + }, + { + "epoch": 0.7716327805148845, + "grad_norm": 2.180182933807373, + "learning_rate": 0.00018640239096921223, + "loss": 0.7278, + "step": 11120 + }, + { + "epoch": 0.7717021719519811, + "grad_norm": 2.8038980960845947, + "learning_rate": 0.00018639956213651647, + "loss": 0.9341, + "step": 11121 + }, + { + "epoch": 0.7717715633890778, + "grad_norm": 3.3339412212371826, + "learning_rate": 0.00018639673303106737, + "loss": 0.8164, + "step": 11122 + }, + { + "epoch": 0.7718409548261744, + "grad_norm": 2.7077949047088623, + "learning_rate": 0.00018639390365287378, + "loss": 0.7776, + "step": 11123 + }, + { + "epoch": 0.7719103462632712, + "grad_norm": 2.1249988079071045, + "learning_rate": 0.00018639107400194465, + "loss": 0.8531, + "step": 11124 + }, + { + "epoch": 0.7719797377003678, + "grad_norm": 1.7261697053909302, + "learning_rate": 0.00018638824407828897, + "loss": 0.7028, + "step": 11125 + }, + { + "epoch": 0.7720491291374645, + "grad_norm": 1.9657574892044067, + "learning_rate": 0.0001863854138819156, + "loss": 0.7543, + "step": 11126 + }, + { + "epoch": 0.7721185205745611, + "grad_norm": 2.354642868041992, + "learning_rate": 0.00018638258341283347, + "loss": 0.7229, + "step": 11127 + }, + { + "epoch": 0.7721879120116577, + "grad_norm": 2.0144882202148438, + "learning_rate": 0.0001863797526710516, + "loss": 0.6762, + "step": 11128 + }, + { + "epoch": 0.7722573034487544, + "grad_norm": 1.970426082611084, + "learning_rate": 0.00018637692165657886, + "loss": 0.6239, + "step": 11129 + }, + { + "epoch": 0.772326694885851, + "grad_norm": 2.966323137283325, + "learning_rate": 0.00018637409036942416, + "loss": 0.8083, + "step": 11130 + }, + { + "epoch": 0.7723960863229478, + "grad_norm": 2.451101779937744, + "learning_rate": 0.0001863712588095965, + "loss": 1.0504, + "step": 11131 + }, + { + "epoch": 0.7724654777600444, + "grad_norm": 2.6291637420654297, + "learning_rate": 0.0001863684269771048, + "loss": 0.9735, + "step": 11132 + }, + { + "epoch": 0.7725348691971411, + "grad_norm": 1.7982473373413086, + "learning_rate": 0.000186365594871958, + "loss": 0.7942, + "step": 11133 + }, + { + "epoch": 0.7726042606342377, + "grad_norm": 2.1510674953460693, + "learning_rate": 0.00018636276249416506, + "loss": 0.7446, + "step": 11134 + }, + { + "epoch": 0.7726736520713344, + "grad_norm": 1.8347532749176025, + "learning_rate": 0.00018635992984373487, + "loss": 0.9009, + "step": 11135 + }, + { + "epoch": 0.772743043508431, + "grad_norm": 1.916070818901062, + "learning_rate": 0.0001863570969206764, + "loss": 0.648, + "step": 11136 + }, + { + "epoch": 0.7728124349455278, + "grad_norm": 2.237791061401367, + "learning_rate": 0.0001863542637249986, + "loss": 0.8784, + "step": 11137 + }, + { + "epoch": 0.7728818263826244, + "grad_norm": 1.9531989097595215, + "learning_rate": 0.00018635143025671043, + "loss": 0.676, + "step": 11138 + }, + { + "epoch": 0.7729512178197211, + "grad_norm": 3.1202170848846436, + "learning_rate": 0.0001863485965158208, + "loss": 0.6445, + "step": 11139 + }, + { + "epoch": 0.7730206092568177, + "grad_norm": 3.2911593914031982, + "learning_rate": 0.00018634576250233865, + "loss": 0.7474, + "step": 11140 + }, + { + "epoch": 0.7730900006939144, + "grad_norm": 2.753614664077759, + "learning_rate": 0.00018634292821627295, + "loss": 1.2239, + "step": 11141 + }, + { + "epoch": 0.773159392131011, + "grad_norm": 1.9639670848846436, + "learning_rate": 0.00018634009365763266, + "loss": 0.7385, + "step": 11142 + }, + { + "epoch": 0.7732287835681076, + "grad_norm": 2.1108901500701904, + "learning_rate": 0.0001863372588264267, + "loss": 0.6952, + "step": 11143 + }, + { + "epoch": 0.7732981750052044, + "grad_norm": 2.218656063079834, + "learning_rate": 0.00018633442372266402, + "loss": 0.7553, + "step": 11144 + }, + { + "epoch": 0.773367566442301, + "grad_norm": 2.3340003490448, + "learning_rate": 0.0001863315883463536, + "loss": 0.7731, + "step": 11145 + }, + { + "epoch": 0.7734369578793977, + "grad_norm": 2.2693393230438232, + "learning_rate": 0.00018632875269750438, + "loss": 0.9304, + "step": 11146 + }, + { + "epoch": 0.7735063493164943, + "grad_norm": 2.5735011100769043, + "learning_rate": 0.00018632591677612527, + "loss": 0.8692, + "step": 11147 + }, + { + "epoch": 0.773575740753591, + "grad_norm": 2.5609467029571533, + "learning_rate": 0.00018632308058222528, + "loss": 0.7993, + "step": 11148 + }, + { + "epoch": 0.7736451321906876, + "grad_norm": 2.0280537605285645, + "learning_rate": 0.00018632024411581336, + "loss": 1.0805, + "step": 11149 + }, + { + "epoch": 0.7737145236277844, + "grad_norm": 1.6879743337631226, + "learning_rate": 0.00018631740737689843, + "loss": 1.0011, + "step": 11150 + }, + { + "epoch": 0.773783915064881, + "grad_norm": 2.1543774604797363, + "learning_rate": 0.00018631457036548944, + "loss": 0.7813, + "step": 11151 + }, + { + "epoch": 0.7738533065019777, + "grad_norm": 2.1418097019195557, + "learning_rate": 0.00018631173308159539, + "loss": 0.7148, + "step": 11152 + }, + { + "epoch": 0.7739226979390743, + "grad_norm": 2.1519503593444824, + "learning_rate": 0.00018630889552522518, + "loss": 0.9174, + "step": 11153 + }, + { + "epoch": 0.773992089376171, + "grad_norm": 2.0570151805877686, + "learning_rate": 0.0001863060576963878, + "loss": 0.7517, + "step": 11154 + }, + { + "epoch": 0.7740614808132676, + "grad_norm": 2.5786969661712646, + "learning_rate": 0.0001863032195950922, + "loss": 0.9312, + "step": 11155 + }, + { + "epoch": 0.7741308722503643, + "grad_norm": 2.1436071395874023, + "learning_rate": 0.00018630038122134743, + "loss": 0.7591, + "step": 11156 + }, + { + "epoch": 0.774200263687461, + "grad_norm": 1.7098697423934937, + "learning_rate": 0.0001862975425751623, + "loss": 0.7373, + "step": 11157 + }, + { + "epoch": 0.7742696551245576, + "grad_norm": 2.175163507461548, + "learning_rate": 0.00018629470365654583, + "loss": 0.8217, + "step": 11158 + }, + { + "epoch": 0.7743390465616543, + "grad_norm": 2.7039618492126465, + "learning_rate": 0.00018629186446550704, + "loss": 0.7549, + "step": 11159 + }, + { + "epoch": 0.7744084379987509, + "grad_norm": 2.405268907546997, + "learning_rate": 0.00018628902500205483, + "loss": 0.7713, + "step": 11160 + }, + { + "epoch": 0.7744778294358476, + "grad_norm": 2.393782138824463, + "learning_rate": 0.00018628618526619816, + "loss": 0.7789, + "step": 11161 + }, + { + "epoch": 0.7745472208729443, + "grad_norm": 2.0043420791625977, + "learning_rate": 0.00018628334525794604, + "loss": 0.8864, + "step": 11162 + }, + { + "epoch": 0.774616612310041, + "grad_norm": 2.07328724861145, + "learning_rate": 0.0001862805049773074, + "loss": 0.9882, + "step": 11163 + }, + { + "epoch": 0.7746860037471376, + "grad_norm": 1.8909491300582886, + "learning_rate": 0.0001862776644242912, + "loss": 0.9506, + "step": 11164 + }, + { + "epoch": 0.7747553951842343, + "grad_norm": 2.4257144927978516, + "learning_rate": 0.00018627482359890647, + "loss": 0.9006, + "step": 11165 + }, + { + "epoch": 0.7748247866213309, + "grad_norm": 2.4735920429229736, + "learning_rate": 0.00018627198250116206, + "loss": 0.9341, + "step": 11166 + }, + { + "epoch": 0.7748941780584276, + "grad_norm": 1.8302912712097168, + "learning_rate": 0.0001862691411310671, + "loss": 0.8395, + "step": 11167 + }, + { + "epoch": 0.7749635694955243, + "grad_norm": 2.878370523452759, + "learning_rate": 0.00018626629948863038, + "loss": 0.9374, + "step": 11168 + }, + { + "epoch": 0.7750329609326209, + "grad_norm": 2.2005910873413086, + "learning_rate": 0.000186263457573861, + "loss": 0.7169, + "step": 11169 + }, + { + "epoch": 0.7751023523697176, + "grad_norm": 2.5328662395477295, + "learning_rate": 0.0001862606153867679, + "loss": 1.1263, + "step": 11170 + }, + { + "epoch": 0.7751717438068142, + "grad_norm": 2.345710039138794, + "learning_rate": 0.00018625777292736006, + "loss": 0.9299, + "step": 11171 + }, + { + "epoch": 0.7752411352439109, + "grad_norm": 2.6834299564361572, + "learning_rate": 0.0001862549301956464, + "loss": 0.8059, + "step": 11172 + }, + { + "epoch": 0.7753105266810075, + "grad_norm": 3.579852819442749, + "learning_rate": 0.00018625208719163597, + "loss": 1.0515, + "step": 11173 + }, + { + "epoch": 0.7753799181181042, + "grad_norm": 2.2414283752441406, + "learning_rate": 0.0001862492439153377, + "loss": 0.8451, + "step": 11174 + }, + { + "epoch": 0.7754493095552009, + "grad_norm": 1.8316351175308228, + "learning_rate": 0.00018624640036676054, + "loss": 0.7898, + "step": 11175 + }, + { + "epoch": 0.7755187009922976, + "grad_norm": 2.045569658279419, + "learning_rate": 0.00018624355654591352, + "loss": 0.959, + "step": 11176 + }, + { + "epoch": 0.7755880924293942, + "grad_norm": 2.2956531047821045, + "learning_rate": 0.0001862407124528056, + "loss": 0.7768, + "step": 11177 + }, + { + "epoch": 0.7756574838664909, + "grad_norm": 2.3503243923187256, + "learning_rate": 0.00018623786808744577, + "loss": 0.8044, + "step": 11178 + }, + { + "epoch": 0.7757268753035875, + "grad_norm": 2.2055470943450928, + "learning_rate": 0.000186235023449843, + "loss": 0.8785, + "step": 11179 + }, + { + "epoch": 0.7757962667406842, + "grad_norm": 2.430872917175293, + "learning_rate": 0.00018623217854000625, + "loss": 0.8729, + "step": 11180 + }, + { + "epoch": 0.7758656581777809, + "grad_norm": 2.277785062789917, + "learning_rate": 0.00018622933335794452, + "loss": 1.0074, + "step": 11181 + }, + { + "epoch": 0.7759350496148775, + "grad_norm": 1.6858835220336914, + "learning_rate": 0.00018622648790366677, + "loss": 0.5221, + "step": 11182 + }, + { + "epoch": 0.7760044410519742, + "grad_norm": 2.5640954971313477, + "learning_rate": 0.000186223642177182, + "loss": 0.9075, + "step": 11183 + }, + { + "epoch": 0.7760738324890708, + "grad_norm": 3.366684675216675, + "learning_rate": 0.00018622079617849923, + "loss": 0.7009, + "step": 11184 + }, + { + "epoch": 0.7761432239261675, + "grad_norm": 2.098151445388794, + "learning_rate": 0.0001862179499076274, + "loss": 0.8704, + "step": 11185 + }, + { + "epoch": 0.7762126153632641, + "grad_norm": 1.8191324472427368, + "learning_rate": 0.00018621510336457552, + "loss": 0.4777, + "step": 11186 + }, + { + "epoch": 0.7762820068003609, + "grad_norm": 1.863650918006897, + "learning_rate": 0.00018621225654935253, + "loss": 0.7927, + "step": 11187 + }, + { + "epoch": 0.7763513982374575, + "grad_norm": 2.100980281829834, + "learning_rate": 0.00018620940946196743, + "loss": 0.922, + "step": 11188 + }, + { + "epoch": 0.7764207896745542, + "grad_norm": 2.256972074508667, + "learning_rate": 0.00018620656210242929, + "loss": 0.6751, + "step": 11189 + }, + { + "epoch": 0.7764901811116508, + "grad_norm": 2.218820095062256, + "learning_rate": 0.00018620371447074696, + "loss": 0.7812, + "step": 11190 + }, + { + "epoch": 0.7765595725487475, + "grad_norm": 2.034008502960205, + "learning_rate": 0.00018620086656692957, + "loss": 0.8051, + "step": 11191 + }, + { + "epoch": 0.7766289639858441, + "grad_norm": 1.8174467086791992, + "learning_rate": 0.000186198018390986, + "loss": 0.6975, + "step": 11192 + }, + { + "epoch": 0.7766983554229409, + "grad_norm": 2.1344685554504395, + "learning_rate": 0.00018619516994292532, + "loss": 0.6832, + "step": 11193 + }, + { + "epoch": 0.7767677468600375, + "grad_norm": 1.8469805717468262, + "learning_rate": 0.0001861923212227565, + "loss": 0.758, + "step": 11194 + }, + { + "epoch": 0.7768371382971342, + "grad_norm": 2.2136285305023193, + "learning_rate": 0.00018618947223048848, + "loss": 0.8205, + "step": 11195 + }, + { + "epoch": 0.7769065297342308, + "grad_norm": 2.2811572551727295, + "learning_rate": 0.0001861866229661303, + "loss": 0.8504, + "step": 11196 + }, + { + "epoch": 0.7769759211713274, + "grad_norm": 2.487424612045288, + "learning_rate": 0.00018618377342969098, + "loss": 0.8511, + "step": 11197 + }, + { + "epoch": 0.7770453126084241, + "grad_norm": 1.9440537691116333, + "learning_rate": 0.00018618092362117947, + "loss": 0.664, + "step": 11198 + }, + { + "epoch": 0.7771147040455207, + "grad_norm": 2.0179734230041504, + "learning_rate": 0.0001861780735406048, + "loss": 0.6658, + "step": 11199 + }, + { + "epoch": 0.7771840954826175, + "grad_norm": 2.6002933979034424, + "learning_rate": 0.0001861752231879759, + "loss": 0.6998, + "step": 11200 + }, + { + "epoch": 0.7772534869197141, + "grad_norm": 1.7756505012512207, + "learning_rate": 0.00018617237256330187, + "loss": 0.7419, + "step": 11201 + }, + { + "epoch": 0.7773228783568108, + "grad_norm": 2.090768814086914, + "learning_rate": 0.00018616952166659165, + "loss": 0.8099, + "step": 11202 + }, + { + "epoch": 0.7773922697939074, + "grad_norm": 1.9168857336044312, + "learning_rate": 0.00018616667049785424, + "loss": 0.7143, + "step": 11203 + }, + { + "epoch": 0.7774616612310041, + "grad_norm": 2.5063772201538086, + "learning_rate": 0.00018616381905709863, + "loss": 0.8731, + "step": 11204 + }, + { + "epoch": 0.7775310526681007, + "grad_norm": 2.4576804637908936, + "learning_rate": 0.00018616096734433386, + "loss": 0.9326, + "step": 11205 + }, + { + "epoch": 0.7776004441051975, + "grad_norm": 1.93064284324646, + "learning_rate": 0.00018615811535956892, + "loss": 0.6695, + "step": 11206 + }, + { + "epoch": 0.7776698355422941, + "grad_norm": 2.8957912921905518, + "learning_rate": 0.0001861552631028128, + "loss": 0.6961, + "step": 11207 + }, + { + "epoch": 0.7777392269793908, + "grad_norm": 2.3645033836364746, + "learning_rate": 0.0001861524105740745, + "loss": 0.9055, + "step": 11208 + }, + { + "epoch": 0.7778086184164874, + "grad_norm": 1.9695687294006348, + "learning_rate": 0.00018614955777336304, + "loss": 0.9578, + "step": 11209 + }, + { + "epoch": 0.777878009853584, + "grad_norm": 1.755902647972107, + "learning_rate": 0.00018614670470068741, + "loss": 0.8553, + "step": 11210 + }, + { + "epoch": 0.7779474012906807, + "grad_norm": 1.874263882637024, + "learning_rate": 0.00018614385135605667, + "loss": 0.7624, + "step": 11211 + }, + { + "epoch": 0.7780167927277774, + "grad_norm": 1.9448610544204712, + "learning_rate": 0.00018614099773947974, + "loss": 0.8673, + "step": 11212 + }, + { + "epoch": 0.7780861841648741, + "grad_norm": 2.176398754119873, + "learning_rate": 0.0001861381438509657, + "loss": 0.7933, + "step": 11213 + }, + { + "epoch": 0.7781555756019707, + "grad_norm": 2.5308327674865723, + "learning_rate": 0.00018613528969052353, + "loss": 0.9152, + "step": 11214 + }, + { + "epoch": 0.7782249670390674, + "grad_norm": 2.5228846073150635, + "learning_rate": 0.00018613243525816223, + "loss": 1.0175, + "step": 11215 + }, + { + "epoch": 0.778294358476164, + "grad_norm": 2.188061237335205, + "learning_rate": 0.00018612958055389082, + "loss": 0.9242, + "step": 11216 + }, + { + "epoch": 0.7783637499132607, + "grad_norm": 1.7502572536468506, + "learning_rate": 0.00018612672557771832, + "loss": 0.6635, + "step": 11217 + }, + { + "epoch": 0.7784331413503573, + "grad_norm": 2.190568208694458, + "learning_rate": 0.00018612387032965378, + "loss": 0.7977, + "step": 11218 + }, + { + "epoch": 0.7785025327874541, + "grad_norm": 2.3711440563201904, + "learning_rate": 0.00018612101480970614, + "loss": 0.9423, + "step": 11219 + }, + { + "epoch": 0.7785719242245507, + "grad_norm": 2.720532178878784, + "learning_rate": 0.00018611815901788445, + "loss": 0.7464, + "step": 11220 + }, + { + "epoch": 0.7786413156616474, + "grad_norm": 2.117236852645874, + "learning_rate": 0.00018611530295419771, + "loss": 0.902, + "step": 11221 + }, + { + "epoch": 0.778710707098744, + "grad_norm": 3.0126163959503174, + "learning_rate": 0.00018611244661865498, + "loss": 0.8878, + "step": 11222 + }, + { + "epoch": 0.7787800985358406, + "grad_norm": 2.3167355060577393, + "learning_rate": 0.00018610959001126524, + "loss": 0.936, + "step": 11223 + }, + { + "epoch": 0.7788494899729373, + "grad_norm": 1.8287452459335327, + "learning_rate": 0.0001861067331320375, + "loss": 0.7341, + "step": 11224 + }, + { + "epoch": 0.778918881410034, + "grad_norm": 2.0508475303649902, + "learning_rate": 0.00018610387598098083, + "loss": 0.8596, + "step": 11225 + }, + { + "epoch": 0.7789882728471307, + "grad_norm": 2.6178503036499023, + "learning_rate": 0.00018610101855810415, + "loss": 0.7544, + "step": 11226 + }, + { + "epoch": 0.7790576642842273, + "grad_norm": 2.080565929412842, + "learning_rate": 0.0001860981608634166, + "loss": 0.9486, + "step": 11227 + }, + { + "epoch": 0.779127055721324, + "grad_norm": 2.104872703552246, + "learning_rate": 0.0001860953028969271, + "loss": 0.9366, + "step": 11228 + }, + { + "epoch": 0.7791964471584206, + "grad_norm": 2.5307605266571045, + "learning_rate": 0.00018609244465864476, + "loss": 0.9842, + "step": 11229 + }, + { + "epoch": 0.7792658385955173, + "grad_norm": 1.7894279956817627, + "learning_rate": 0.00018608958614857855, + "loss": 0.6745, + "step": 11230 + }, + { + "epoch": 0.779335230032614, + "grad_norm": 2.331890821456909, + "learning_rate": 0.0001860867273667375, + "loss": 0.686, + "step": 11231 + }, + { + "epoch": 0.7794046214697107, + "grad_norm": 2.4368643760681152, + "learning_rate": 0.00018608386831313065, + "loss": 0.5946, + "step": 11232 + }, + { + "epoch": 0.7794740129068073, + "grad_norm": 3.205843687057495, + "learning_rate": 0.00018608100898776696, + "loss": 0.7243, + "step": 11233 + }, + { + "epoch": 0.779543404343904, + "grad_norm": 2.7454946041107178, + "learning_rate": 0.00018607814939065558, + "loss": 0.8124, + "step": 11234 + }, + { + "epoch": 0.7796127957810006, + "grad_norm": 1.8433696031570435, + "learning_rate": 0.00018607528952180544, + "loss": 0.748, + "step": 11235 + }, + { + "epoch": 0.7796821872180972, + "grad_norm": 2.751735210418701, + "learning_rate": 0.0001860724293812256, + "loss": 0.6171, + "step": 11236 + }, + { + "epoch": 0.779751578655194, + "grad_norm": 2.247720718383789, + "learning_rate": 0.00018606956896892509, + "loss": 0.7725, + "step": 11237 + }, + { + "epoch": 0.7798209700922906, + "grad_norm": 2.111924648284912, + "learning_rate": 0.00018606670828491293, + "loss": 0.8567, + "step": 11238 + }, + { + "epoch": 0.7798903615293873, + "grad_norm": 2.2160162925720215, + "learning_rate": 0.00018606384732919816, + "loss": 0.7738, + "step": 11239 + }, + { + "epoch": 0.7799597529664839, + "grad_norm": 2.553760051727295, + "learning_rate": 0.00018606098610178981, + "loss": 0.9305, + "step": 11240 + }, + { + "epoch": 0.7800291444035806, + "grad_norm": 2.474292039871216, + "learning_rate": 0.00018605812460269693, + "loss": 0.8016, + "step": 11241 + }, + { + "epoch": 0.7800985358406772, + "grad_norm": 2.2186379432678223, + "learning_rate": 0.00018605526283192852, + "loss": 0.9222, + "step": 11242 + }, + { + "epoch": 0.780167927277774, + "grad_norm": 2.082017660140991, + "learning_rate": 0.0001860524007894936, + "loss": 0.7945, + "step": 11243 + }, + { + "epoch": 0.7802373187148706, + "grad_norm": 2.342637062072754, + "learning_rate": 0.0001860495384754013, + "loss": 0.9113, + "step": 11244 + }, + { + "epoch": 0.7803067101519673, + "grad_norm": 2.7382047176361084, + "learning_rate": 0.00018604667588966052, + "loss": 0.833, + "step": 11245 + }, + { + "epoch": 0.7803761015890639, + "grad_norm": 1.749027967453003, + "learning_rate": 0.0001860438130322804, + "loss": 0.7192, + "step": 11246 + }, + { + "epoch": 0.7804454930261606, + "grad_norm": 2.1068673133850098, + "learning_rate": 0.00018604094990326998, + "loss": 1.0114, + "step": 11247 + }, + { + "epoch": 0.7805148844632572, + "grad_norm": 2.351970672607422, + "learning_rate": 0.00018603808650263822, + "loss": 0.8936, + "step": 11248 + }, + { + "epoch": 0.7805842759003538, + "grad_norm": 1.9597054719924927, + "learning_rate": 0.0001860352228303942, + "loss": 0.778, + "step": 11249 + }, + { + "epoch": 0.7806536673374506, + "grad_norm": 2.900573492050171, + "learning_rate": 0.000186032358886547, + "loss": 0.8862, + "step": 11250 + }, + { + "epoch": 0.7807230587745472, + "grad_norm": 2.274672508239746, + "learning_rate": 0.0001860294946711056, + "loss": 0.5754, + "step": 11251 + }, + { + "epoch": 0.7807924502116439, + "grad_norm": 2.3598103523254395, + "learning_rate": 0.00018602663018407904, + "loss": 1.1265, + "step": 11252 + }, + { + "epoch": 0.7808618416487405, + "grad_norm": 2.3082165718078613, + "learning_rate": 0.00018602376542547645, + "loss": 0.8578, + "step": 11253 + }, + { + "epoch": 0.7809312330858372, + "grad_norm": 2.31160831451416, + "learning_rate": 0.00018602090039530675, + "loss": 0.6843, + "step": 11254 + }, + { + "epoch": 0.7810006245229338, + "grad_norm": 2.5687601566314697, + "learning_rate": 0.0001860180350935791, + "loss": 0.969, + "step": 11255 + }, + { + "epoch": 0.7810700159600306, + "grad_norm": 2.067593574523926, + "learning_rate": 0.00018601516952030246, + "loss": 0.7746, + "step": 11256 + }, + { + "epoch": 0.7811394073971272, + "grad_norm": 2.342003345489502, + "learning_rate": 0.00018601230367548593, + "loss": 0.7567, + "step": 11257 + }, + { + "epoch": 0.7812087988342239, + "grad_norm": 2.402573585510254, + "learning_rate": 0.00018600943755913853, + "loss": 1.1766, + "step": 11258 + }, + { + "epoch": 0.7812781902713205, + "grad_norm": 2.2460172176361084, + "learning_rate": 0.0001860065711712693, + "loss": 0.8869, + "step": 11259 + }, + { + "epoch": 0.7813475817084172, + "grad_norm": 1.7237025499343872, + "learning_rate": 0.00018600370451188732, + "loss": 0.6896, + "step": 11260 + }, + { + "epoch": 0.7814169731455138, + "grad_norm": 2.13326096534729, + "learning_rate": 0.00018600083758100162, + "loss": 0.8348, + "step": 11261 + }, + { + "epoch": 0.7814863645826106, + "grad_norm": 1.8023885488510132, + "learning_rate": 0.00018599797037862126, + "loss": 0.6645, + "step": 11262 + }, + { + "epoch": 0.7815557560197072, + "grad_norm": 1.7431381940841675, + "learning_rate": 0.00018599510290475527, + "loss": 0.8609, + "step": 11263 + }, + { + "epoch": 0.7816251474568038, + "grad_norm": 1.7116477489471436, + "learning_rate": 0.0001859922351594127, + "loss": 0.6236, + "step": 11264 + }, + { + "epoch": 0.7816945388939005, + "grad_norm": 2.20102596282959, + "learning_rate": 0.00018598936714260265, + "loss": 1.1482, + "step": 11265 + }, + { + "epoch": 0.7817639303309971, + "grad_norm": 2.6528162956237793, + "learning_rate": 0.00018598649885433413, + "loss": 0.9239, + "step": 11266 + }, + { + "epoch": 0.7818333217680938, + "grad_norm": 1.7237133979797363, + "learning_rate": 0.00018598363029461625, + "loss": 0.7701, + "step": 11267 + }, + { + "epoch": 0.7819027132051904, + "grad_norm": 2.255643129348755, + "learning_rate": 0.00018598076146345797, + "loss": 0.7711, + "step": 11268 + }, + { + "epoch": 0.7819721046422872, + "grad_norm": 2.1680779457092285, + "learning_rate": 0.00018597789236086842, + "loss": 0.7976, + "step": 11269 + }, + { + "epoch": 0.7820414960793838, + "grad_norm": 2.0813140869140625, + "learning_rate": 0.00018597502298685665, + "loss": 0.8618, + "step": 11270 + }, + { + "epoch": 0.7821108875164805, + "grad_norm": 2.2381865978240967, + "learning_rate": 0.00018597215334143172, + "loss": 0.6811, + "step": 11271 + }, + { + "epoch": 0.7821802789535771, + "grad_norm": 1.9602268934249878, + "learning_rate": 0.00018596928342460266, + "loss": 0.6421, + "step": 11272 + }, + { + "epoch": 0.7822496703906738, + "grad_norm": 1.729249119758606, + "learning_rate": 0.00018596641323637852, + "loss": 0.6586, + "step": 11273 + }, + { + "epoch": 0.7823190618277704, + "grad_norm": 2.386685848236084, + "learning_rate": 0.00018596354277676844, + "loss": 0.8075, + "step": 11274 + }, + { + "epoch": 0.7823884532648672, + "grad_norm": 1.7283170223236084, + "learning_rate": 0.00018596067204578137, + "loss": 0.7838, + "step": 11275 + }, + { + "epoch": 0.7824578447019638, + "grad_norm": 2.537081480026245, + "learning_rate": 0.0001859578010434265, + "loss": 0.7293, + "step": 11276 + }, + { + "epoch": 0.7825272361390604, + "grad_norm": 2.3540351390838623, + "learning_rate": 0.00018595492976971277, + "loss": 1.028, + "step": 11277 + }, + { + "epoch": 0.7825966275761571, + "grad_norm": 3.0542542934417725, + "learning_rate": 0.00018595205822464933, + "loss": 0.805, + "step": 11278 + }, + { + "epoch": 0.7826660190132537, + "grad_norm": 2.312023639678955, + "learning_rate": 0.00018594918640824518, + "loss": 0.6416, + "step": 11279 + }, + { + "epoch": 0.7827354104503504, + "grad_norm": 2.2073662281036377, + "learning_rate": 0.00018594631432050944, + "loss": 0.8929, + "step": 11280 + }, + { + "epoch": 0.782804801887447, + "grad_norm": 2.0780372619628906, + "learning_rate": 0.00018594344196145118, + "loss": 0.7308, + "step": 11281 + }, + { + "epoch": 0.7828741933245438, + "grad_norm": 2.555304527282715, + "learning_rate": 0.00018594056933107943, + "loss": 0.7727, + "step": 11282 + }, + { + "epoch": 0.7829435847616404, + "grad_norm": 2.1315371990203857, + "learning_rate": 0.00018593769642940327, + "loss": 0.871, + "step": 11283 + }, + { + "epoch": 0.7830129761987371, + "grad_norm": 2.1313681602478027, + "learning_rate": 0.00018593482325643176, + "loss": 0.8389, + "step": 11284 + }, + { + "epoch": 0.7830823676358337, + "grad_norm": 2.3943636417388916, + "learning_rate": 0.000185931949812174, + "loss": 0.8419, + "step": 11285 + }, + { + "epoch": 0.7831517590729304, + "grad_norm": 1.6552586555480957, + "learning_rate": 0.00018592907609663904, + "loss": 0.5121, + "step": 11286 + }, + { + "epoch": 0.783221150510027, + "grad_norm": 2.453580379486084, + "learning_rate": 0.00018592620210983594, + "loss": 0.7892, + "step": 11287 + }, + { + "epoch": 0.7832905419471238, + "grad_norm": 3.03131103515625, + "learning_rate": 0.0001859233278517738, + "loss": 0.9466, + "step": 11288 + }, + { + "epoch": 0.7833599333842204, + "grad_norm": 2.5187504291534424, + "learning_rate": 0.00018592045332246172, + "loss": 0.9107, + "step": 11289 + }, + { + "epoch": 0.783429324821317, + "grad_norm": 2.435236930847168, + "learning_rate": 0.00018591757852190867, + "loss": 0.836, + "step": 11290 + }, + { + "epoch": 0.7834987162584137, + "grad_norm": 2.521682024002075, + "learning_rate": 0.00018591470345012386, + "loss": 1.0437, + "step": 11291 + }, + { + "epoch": 0.7835681076955103, + "grad_norm": 3.057317018508911, + "learning_rate": 0.00018591182810711625, + "loss": 1.0599, + "step": 11292 + }, + { + "epoch": 0.783637499132607, + "grad_norm": 2.448071241378784, + "learning_rate": 0.000185908952492895, + "loss": 1.0505, + "step": 11293 + }, + { + "epoch": 0.7837068905697037, + "grad_norm": 1.979245901107788, + "learning_rate": 0.00018590607660746913, + "loss": 0.702, + "step": 11294 + }, + { + "epoch": 0.7837762820068004, + "grad_norm": 1.8296008110046387, + "learning_rate": 0.00018590320045084775, + "loss": 0.9161, + "step": 11295 + }, + { + "epoch": 0.783845673443897, + "grad_norm": 2.018608570098877, + "learning_rate": 0.00018590032402303992, + "loss": 0.6102, + "step": 11296 + }, + { + "epoch": 0.7839150648809937, + "grad_norm": 1.985021948814392, + "learning_rate": 0.00018589744732405475, + "loss": 0.9469, + "step": 11297 + }, + { + "epoch": 0.7839844563180903, + "grad_norm": 2.2998390197753906, + "learning_rate": 0.00018589457035390132, + "loss": 1.1664, + "step": 11298 + }, + { + "epoch": 0.784053847755187, + "grad_norm": 2.707120418548584, + "learning_rate": 0.00018589169311258869, + "loss": 0.9598, + "step": 11299 + }, + { + "epoch": 0.7841232391922837, + "grad_norm": 2.5528149604797363, + "learning_rate": 0.00018588881560012592, + "loss": 0.9362, + "step": 11300 + }, + { + "epoch": 0.7841926306293804, + "grad_norm": 1.8310480117797852, + "learning_rate": 0.00018588593781652214, + "loss": 0.6425, + "step": 11301 + }, + { + "epoch": 0.784262022066477, + "grad_norm": 1.7567251920700073, + "learning_rate": 0.00018588305976178644, + "loss": 0.8892, + "step": 11302 + }, + { + "epoch": 0.7843314135035736, + "grad_norm": 1.848697304725647, + "learning_rate": 0.00018588018143592783, + "loss": 0.8606, + "step": 11303 + }, + { + "epoch": 0.7844008049406703, + "grad_norm": 1.6610440015792847, + "learning_rate": 0.0001858773028389555, + "loss": 0.597, + "step": 11304 + }, + { + "epoch": 0.7844701963777669, + "grad_norm": 1.899948000907898, + "learning_rate": 0.0001858744239708785, + "loss": 0.7431, + "step": 11305 + }, + { + "epoch": 0.7845395878148637, + "grad_norm": 2.4425294399261475, + "learning_rate": 0.00018587154483170587, + "loss": 1.0214, + "step": 11306 + }, + { + "epoch": 0.7846089792519603, + "grad_norm": 1.9362157583236694, + "learning_rate": 0.00018586866542144673, + "loss": 0.7293, + "step": 11307 + }, + { + "epoch": 0.784678370689057, + "grad_norm": 2.322345018386841, + "learning_rate": 0.00018586578574011022, + "loss": 0.655, + "step": 11308 + }, + { + "epoch": 0.7847477621261536, + "grad_norm": 2.029982805252075, + "learning_rate": 0.00018586290578770536, + "loss": 0.5307, + "step": 11309 + }, + { + "epoch": 0.7848171535632503, + "grad_norm": 2.5180749893188477, + "learning_rate": 0.00018586002556424127, + "loss": 1.1047, + "step": 11310 + }, + { + "epoch": 0.7848865450003469, + "grad_norm": 2.4969279766082764, + "learning_rate": 0.00018585714506972703, + "loss": 0.8639, + "step": 11311 + }, + { + "epoch": 0.7849559364374437, + "grad_norm": 2.023409843444824, + "learning_rate": 0.00018585426430417176, + "loss": 0.6953, + "step": 11312 + }, + { + "epoch": 0.7850253278745403, + "grad_norm": 2.3899526596069336, + "learning_rate": 0.00018585138326758453, + "loss": 0.7117, + "step": 11313 + }, + { + "epoch": 0.785094719311637, + "grad_norm": 1.6788350343704224, + "learning_rate": 0.00018584850195997444, + "loss": 0.562, + "step": 11314 + }, + { + "epoch": 0.7851641107487336, + "grad_norm": 2.7442095279693604, + "learning_rate": 0.00018584562038135061, + "loss": 0.8766, + "step": 11315 + }, + { + "epoch": 0.7852335021858303, + "grad_norm": 2.2551839351654053, + "learning_rate": 0.0001858427385317221, + "loss": 0.6749, + "step": 11316 + }, + { + "epoch": 0.7853028936229269, + "grad_norm": 2.3003242015838623, + "learning_rate": 0.00018583985641109802, + "loss": 0.8028, + "step": 11317 + }, + { + "epoch": 0.7853722850600235, + "grad_norm": 2.396449327468872, + "learning_rate": 0.0001858369740194875, + "loss": 1.0488, + "step": 11318 + }, + { + "epoch": 0.7854416764971203, + "grad_norm": 2.8031816482543945, + "learning_rate": 0.00018583409135689957, + "loss": 0.8674, + "step": 11319 + }, + { + "epoch": 0.7855110679342169, + "grad_norm": 2.0533251762390137, + "learning_rate": 0.0001858312084233434, + "loss": 1.062, + "step": 11320 + }, + { + "epoch": 0.7855804593713136, + "grad_norm": 1.904259204864502, + "learning_rate": 0.00018582832521882807, + "loss": 0.6382, + "step": 11321 + }, + { + "epoch": 0.7856498508084102, + "grad_norm": 2.099128484725952, + "learning_rate": 0.00018582544174336265, + "loss": 0.9487, + "step": 11322 + }, + { + "epoch": 0.7857192422455069, + "grad_norm": 2.0826034545898438, + "learning_rate": 0.00018582255799695627, + "loss": 0.6754, + "step": 11323 + }, + { + "epoch": 0.7857886336826035, + "grad_norm": 3.247870922088623, + "learning_rate": 0.00018581967397961806, + "loss": 0.8099, + "step": 11324 + }, + { + "epoch": 0.7858580251197003, + "grad_norm": 1.958892822265625, + "learning_rate": 0.00018581678969135706, + "loss": 0.7377, + "step": 11325 + }, + { + "epoch": 0.7859274165567969, + "grad_norm": 2.378173828125, + "learning_rate": 0.00018581390513218245, + "loss": 0.7117, + "step": 11326 + }, + { + "epoch": 0.7859968079938936, + "grad_norm": 2.518986940383911, + "learning_rate": 0.0001858110203021033, + "loss": 0.9702, + "step": 11327 + }, + { + "epoch": 0.7860661994309902, + "grad_norm": 1.7241824865341187, + "learning_rate": 0.00018580813520112867, + "loss": 0.6177, + "step": 11328 + }, + { + "epoch": 0.7861355908680869, + "grad_norm": 2.0395610332489014, + "learning_rate": 0.00018580524982926772, + "loss": 0.9036, + "step": 11329 + }, + { + "epoch": 0.7862049823051835, + "grad_norm": 2.231588125228882, + "learning_rate": 0.0001858023641865296, + "loss": 0.697, + "step": 11330 + }, + { + "epoch": 0.7862743737422802, + "grad_norm": 2.3145296573638916, + "learning_rate": 0.00018579947827292333, + "loss": 0.8582, + "step": 11331 + }, + { + "epoch": 0.7863437651793769, + "grad_norm": 2.0555925369262695, + "learning_rate": 0.00018579659208845802, + "loss": 0.6449, + "step": 11332 + }, + { + "epoch": 0.7864131566164735, + "grad_norm": 1.7407068014144897, + "learning_rate": 0.00018579370563314291, + "loss": 0.6169, + "step": 11333 + }, + { + "epoch": 0.7864825480535702, + "grad_norm": 2.3991830348968506, + "learning_rate": 0.00018579081890698698, + "loss": 1.0438, + "step": 11334 + }, + { + "epoch": 0.7865519394906668, + "grad_norm": 2.5490517616271973, + "learning_rate": 0.00018578793190999942, + "loss": 0.957, + "step": 11335 + }, + { + "epoch": 0.7866213309277635, + "grad_norm": 1.914252519607544, + "learning_rate": 0.00018578504464218928, + "loss": 0.5721, + "step": 11336 + }, + { + "epoch": 0.7866907223648602, + "grad_norm": 2.8514912128448486, + "learning_rate": 0.0001857821571035657, + "loss": 1.1958, + "step": 11337 + }, + { + "epoch": 0.7867601138019569, + "grad_norm": 1.938369631767273, + "learning_rate": 0.00018577926929413784, + "loss": 0.7734, + "step": 11338 + }, + { + "epoch": 0.7868295052390535, + "grad_norm": 2.456954002380371, + "learning_rate": 0.00018577638121391473, + "loss": 1.0417, + "step": 11339 + }, + { + "epoch": 0.7868988966761502, + "grad_norm": 2.311859607696533, + "learning_rate": 0.00018577349286290557, + "loss": 0.7144, + "step": 11340 + }, + { + "epoch": 0.7869682881132468, + "grad_norm": 2.249498128890991, + "learning_rate": 0.00018577060424111944, + "loss": 0.8159, + "step": 11341 + }, + { + "epoch": 0.7870376795503435, + "grad_norm": 2.3643362522125244, + "learning_rate": 0.00018576771534856544, + "loss": 0.9667, + "step": 11342 + }, + { + "epoch": 0.7871070709874401, + "grad_norm": 1.9977116584777832, + "learning_rate": 0.00018576482618525274, + "loss": 0.7211, + "step": 11343 + }, + { + "epoch": 0.7871764624245368, + "grad_norm": 3.132139205932617, + "learning_rate": 0.00018576193675119045, + "loss": 0.7223, + "step": 11344 + }, + { + "epoch": 0.7872458538616335, + "grad_norm": 2.32918119430542, + "learning_rate": 0.00018575904704638764, + "loss": 0.704, + "step": 11345 + }, + { + "epoch": 0.7873152452987301, + "grad_norm": 1.9299753904342651, + "learning_rate": 0.0001857561570708535, + "loss": 0.9383, + "step": 11346 + }, + { + "epoch": 0.7873846367358268, + "grad_norm": 2.3877222537994385, + "learning_rate": 0.0001857532668245971, + "loss": 0.7214, + "step": 11347 + }, + { + "epoch": 0.7874540281729234, + "grad_norm": 1.5591245889663696, + "learning_rate": 0.0001857503763076276, + "loss": 0.8366, + "step": 11348 + }, + { + "epoch": 0.7875234196100201, + "grad_norm": 2.309842824935913, + "learning_rate": 0.0001857474855199541, + "loss": 1.1369, + "step": 11349 + }, + { + "epoch": 0.7875928110471168, + "grad_norm": 2.50959849357605, + "learning_rate": 0.00018574459446158575, + "loss": 0.8943, + "step": 11350 + }, + { + "epoch": 0.7876622024842135, + "grad_norm": 2.300187349319458, + "learning_rate": 0.00018574170313253162, + "loss": 0.9021, + "step": 11351 + }, + { + "epoch": 0.7877315939213101, + "grad_norm": 2.367753744125366, + "learning_rate": 0.00018573881153280094, + "loss": 0.747, + "step": 11352 + }, + { + "epoch": 0.7878009853584068, + "grad_norm": 2.379286050796509, + "learning_rate": 0.00018573591966240276, + "loss": 0.8486, + "step": 11353 + }, + { + "epoch": 0.7878703767955034, + "grad_norm": 2.255207061767578, + "learning_rate": 0.0001857330275213462, + "loss": 0.7661, + "step": 11354 + }, + { + "epoch": 0.7879397682326001, + "grad_norm": 2.425624132156372, + "learning_rate": 0.00018573013510964046, + "loss": 0.9116, + "step": 11355 + }, + { + "epoch": 0.7880091596696968, + "grad_norm": 1.616883397102356, + "learning_rate": 0.00018572724242729458, + "loss": 0.538, + "step": 11356 + }, + { + "epoch": 0.7880785511067934, + "grad_norm": 2.2262725830078125, + "learning_rate": 0.00018572434947431777, + "loss": 0.6665, + "step": 11357 + }, + { + "epoch": 0.7881479425438901, + "grad_norm": 2.709981918334961, + "learning_rate": 0.00018572145625071914, + "loss": 0.8105, + "step": 11358 + }, + { + "epoch": 0.7882173339809867, + "grad_norm": 2.170668363571167, + "learning_rate": 0.00018571856275650784, + "loss": 0.7393, + "step": 11359 + }, + { + "epoch": 0.7882867254180834, + "grad_norm": 2.4498181343078613, + "learning_rate": 0.00018571566899169294, + "loss": 0.8724, + "step": 11360 + }, + { + "epoch": 0.78835611685518, + "grad_norm": 3.276020050048828, + "learning_rate": 0.00018571277495628365, + "loss": 1.0472, + "step": 11361 + }, + { + "epoch": 0.7884255082922768, + "grad_norm": 2.6516427993774414, + "learning_rate": 0.00018570988065028904, + "loss": 1.0319, + "step": 11362 + }, + { + "epoch": 0.7884948997293734, + "grad_norm": 1.8182371854782104, + "learning_rate": 0.0001857069860737183, + "loss": 0.7306, + "step": 11363 + }, + { + "epoch": 0.7885642911664701, + "grad_norm": 1.9382374286651611, + "learning_rate": 0.00018570409122658057, + "loss": 0.668, + "step": 11364 + }, + { + "epoch": 0.7886336826035667, + "grad_norm": 2.1274359226226807, + "learning_rate": 0.00018570119610888494, + "loss": 0.8792, + "step": 11365 + }, + { + "epoch": 0.7887030740406634, + "grad_norm": 2.0713491439819336, + "learning_rate": 0.0001856983007206406, + "loss": 0.7081, + "step": 11366 + }, + { + "epoch": 0.78877246547776, + "grad_norm": 1.79682195186615, + "learning_rate": 0.00018569540506185663, + "loss": 0.7766, + "step": 11367 + }, + { + "epoch": 0.7888418569148568, + "grad_norm": 2.692244529724121, + "learning_rate": 0.00018569250913254225, + "loss": 0.8076, + "step": 11368 + }, + { + "epoch": 0.7889112483519534, + "grad_norm": 1.7170312404632568, + "learning_rate": 0.00018568961293270652, + "loss": 0.6765, + "step": 11369 + }, + { + "epoch": 0.7889806397890501, + "grad_norm": 1.8722507953643799, + "learning_rate": 0.00018568671646235866, + "loss": 0.7618, + "step": 11370 + }, + { + "epoch": 0.7890500312261467, + "grad_norm": 2.3760147094726562, + "learning_rate": 0.00018568381972150776, + "loss": 0.9488, + "step": 11371 + }, + { + "epoch": 0.7891194226632433, + "grad_norm": 1.8987441062927246, + "learning_rate": 0.00018568092271016297, + "loss": 0.698, + "step": 11372 + }, + { + "epoch": 0.78918881410034, + "grad_norm": 1.8686059713363647, + "learning_rate": 0.00018567802542833345, + "loss": 0.7419, + "step": 11373 + }, + { + "epoch": 0.7892582055374366, + "grad_norm": 2.7912850379943848, + "learning_rate": 0.0001856751278760284, + "loss": 0.786, + "step": 11374 + }, + { + "epoch": 0.7893275969745334, + "grad_norm": 2.2482001781463623, + "learning_rate": 0.00018567223005325683, + "loss": 0.8779, + "step": 11375 + }, + { + "epoch": 0.78939698841163, + "grad_norm": 1.7299647331237793, + "learning_rate": 0.00018566933196002803, + "loss": 0.7825, + "step": 11376 + }, + { + "epoch": 0.7894663798487267, + "grad_norm": 1.9939837455749512, + "learning_rate": 0.00018566643359635106, + "loss": 0.7072, + "step": 11377 + }, + { + "epoch": 0.7895357712858233, + "grad_norm": 2.622929096221924, + "learning_rate": 0.0001856635349622351, + "loss": 0.8427, + "step": 11378 + }, + { + "epoch": 0.78960516272292, + "grad_norm": 3.281904697418213, + "learning_rate": 0.0001856606360576893, + "loss": 0.7878, + "step": 11379 + }, + { + "epoch": 0.7896745541600166, + "grad_norm": 2.250494956970215, + "learning_rate": 0.0001856577368827228, + "loss": 0.866, + "step": 11380 + }, + { + "epoch": 0.7897439455971134, + "grad_norm": 2.036133289337158, + "learning_rate": 0.0001856548374373448, + "loss": 0.8228, + "step": 11381 + }, + { + "epoch": 0.78981333703421, + "grad_norm": 3.057269811630249, + "learning_rate": 0.0001856519377215644, + "loss": 0.939, + "step": 11382 + }, + { + "epoch": 0.7898827284713067, + "grad_norm": 2.3625683784484863, + "learning_rate": 0.00018564903773539076, + "loss": 0.8162, + "step": 11383 + }, + { + "epoch": 0.7899521199084033, + "grad_norm": 2.124150276184082, + "learning_rate": 0.00018564613747883306, + "loss": 0.9963, + "step": 11384 + }, + { + "epoch": 0.7900215113454999, + "grad_norm": 2.3711981773376465, + "learning_rate": 0.00018564323695190041, + "loss": 1.1409, + "step": 11385 + }, + { + "epoch": 0.7900909027825966, + "grad_norm": 2.2239620685577393, + "learning_rate": 0.000185640336154602, + "loss": 0.7747, + "step": 11386 + }, + { + "epoch": 0.7901602942196932, + "grad_norm": 2.5039222240448, + "learning_rate": 0.00018563743508694702, + "loss": 0.8409, + "step": 11387 + }, + { + "epoch": 0.79022968565679, + "grad_norm": 1.8808265924453735, + "learning_rate": 0.0001856345337489446, + "loss": 0.8798, + "step": 11388 + }, + { + "epoch": 0.7902990770938866, + "grad_norm": 2.3019790649414062, + "learning_rate": 0.00018563163214060384, + "loss": 1.1272, + "step": 11389 + }, + { + "epoch": 0.7903684685309833, + "grad_norm": 2.0423312187194824, + "learning_rate": 0.000185628730261934, + "loss": 0.8117, + "step": 11390 + }, + { + "epoch": 0.7904378599680799, + "grad_norm": 1.5646464824676514, + "learning_rate": 0.00018562582811294416, + "loss": 0.7422, + "step": 11391 + }, + { + "epoch": 0.7905072514051766, + "grad_norm": 2.5255587100982666, + "learning_rate": 0.00018562292569364352, + "loss": 0.8495, + "step": 11392 + }, + { + "epoch": 0.7905766428422732, + "grad_norm": 2.156726360321045, + "learning_rate": 0.00018562002300404125, + "loss": 0.9908, + "step": 11393 + }, + { + "epoch": 0.79064603427937, + "grad_norm": 1.773353934288025, + "learning_rate": 0.00018561712004414648, + "loss": 0.7479, + "step": 11394 + }, + { + "epoch": 0.7907154257164666, + "grad_norm": 1.8158715963363647, + "learning_rate": 0.00018561421681396843, + "loss": 0.6622, + "step": 11395 + }, + { + "epoch": 0.7907848171535633, + "grad_norm": 3.149911880493164, + "learning_rate": 0.0001856113133135162, + "loss": 1.1553, + "step": 11396 + }, + { + "epoch": 0.7908542085906599, + "grad_norm": 2.028012990951538, + "learning_rate": 0.000185608409542799, + "loss": 0.7384, + "step": 11397 + }, + { + "epoch": 0.7909236000277565, + "grad_norm": 1.8766295909881592, + "learning_rate": 0.00018560550550182598, + "loss": 0.7251, + "step": 11398 + }, + { + "epoch": 0.7909929914648532, + "grad_norm": 1.6240547895431519, + "learning_rate": 0.00018560260119060632, + "loss": 0.8648, + "step": 11399 + }, + { + "epoch": 0.7910623829019499, + "grad_norm": 2.225231409072876, + "learning_rate": 0.00018559969660914919, + "loss": 0.7641, + "step": 11400 + }, + { + "epoch": 0.7911317743390466, + "grad_norm": 2.180422782897949, + "learning_rate": 0.0001855967917574637, + "loss": 0.766, + "step": 11401 + }, + { + "epoch": 0.7912011657761432, + "grad_norm": 2.1908419132232666, + "learning_rate": 0.00018559388663555912, + "loss": 0.7579, + "step": 11402 + }, + { + "epoch": 0.7912705572132399, + "grad_norm": 2.2850656509399414, + "learning_rate": 0.00018559098124344456, + "loss": 0.8248, + "step": 11403 + }, + { + "epoch": 0.7913399486503365, + "grad_norm": 2.0625360012054443, + "learning_rate": 0.00018558807558112918, + "loss": 0.7099, + "step": 11404 + }, + { + "epoch": 0.7914093400874332, + "grad_norm": 2.1670374870300293, + "learning_rate": 0.0001855851696486222, + "loss": 0.6549, + "step": 11405 + }, + { + "epoch": 0.7914787315245299, + "grad_norm": 1.559556484222412, + "learning_rate": 0.00018558226344593277, + "loss": 0.6536, + "step": 11406 + }, + { + "epoch": 0.7915481229616266, + "grad_norm": 2.4417903423309326, + "learning_rate": 0.00018557935697307006, + "loss": 0.647, + "step": 11407 + }, + { + "epoch": 0.7916175143987232, + "grad_norm": 2.528402328491211, + "learning_rate": 0.00018557645023004323, + "loss": 0.7893, + "step": 11408 + }, + { + "epoch": 0.7916869058358199, + "grad_norm": 2.5603110790252686, + "learning_rate": 0.0001855735432168615, + "loss": 1.0204, + "step": 11409 + }, + { + "epoch": 0.7917562972729165, + "grad_norm": 1.827186107635498, + "learning_rate": 0.000185570635933534, + "loss": 0.8212, + "step": 11410 + }, + { + "epoch": 0.7918256887100131, + "grad_norm": 2.1592791080474854, + "learning_rate": 0.00018556772838006994, + "loss": 0.7777, + "step": 11411 + }, + { + "epoch": 0.7918950801471099, + "grad_norm": 2.091376543045044, + "learning_rate": 0.0001855648205564785, + "loss": 0.915, + "step": 11412 + }, + { + "epoch": 0.7919644715842065, + "grad_norm": 1.7915303707122803, + "learning_rate": 0.00018556191246276883, + "loss": 0.8158, + "step": 11413 + }, + { + "epoch": 0.7920338630213032, + "grad_norm": 2.0524911880493164, + "learning_rate": 0.00018555900409895015, + "loss": 0.7161, + "step": 11414 + }, + { + "epoch": 0.7921032544583998, + "grad_norm": 1.687852144241333, + "learning_rate": 0.00018555609546503157, + "loss": 0.6121, + "step": 11415 + }, + { + "epoch": 0.7921726458954965, + "grad_norm": 3.3344883918762207, + "learning_rate": 0.00018555318656102239, + "loss": 0.9671, + "step": 11416 + }, + { + "epoch": 0.7922420373325931, + "grad_norm": 2.3080174922943115, + "learning_rate": 0.00018555027738693169, + "loss": 0.5524, + "step": 11417 + }, + { + "epoch": 0.7923114287696899, + "grad_norm": 2.0621743202209473, + "learning_rate": 0.0001855473679427687, + "loss": 0.7896, + "step": 11418 + }, + { + "epoch": 0.7923808202067865, + "grad_norm": 1.8562674522399902, + "learning_rate": 0.00018554445822854262, + "loss": 0.9117, + "step": 11419 + }, + { + "epoch": 0.7924502116438832, + "grad_norm": 2.114469289779663, + "learning_rate": 0.00018554154824426255, + "loss": 0.7163, + "step": 11420 + }, + { + "epoch": 0.7925196030809798, + "grad_norm": 2.9679832458496094, + "learning_rate": 0.00018553863798993777, + "loss": 1.0825, + "step": 11421 + }, + { + "epoch": 0.7925889945180765, + "grad_norm": 2.135395050048828, + "learning_rate": 0.00018553572746557743, + "loss": 0.7586, + "step": 11422 + }, + { + "epoch": 0.7926583859551731, + "grad_norm": 3.112290382385254, + "learning_rate": 0.00018553281667119074, + "loss": 0.939, + "step": 11423 + }, + { + "epoch": 0.7927277773922697, + "grad_norm": 1.8206815719604492, + "learning_rate": 0.00018552990560678686, + "loss": 0.6922, + "step": 11424 + }, + { + "epoch": 0.7927971688293665, + "grad_norm": 2.1934406757354736, + "learning_rate": 0.00018552699427237498, + "loss": 0.8863, + "step": 11425 + }, + { + "epoch": 0.7928665602664631, + "grad_norm": 1.9225306510925293, + "learning_rate": 0.00018552408266796433, + "loss": 0.7929, + "step": 11426 + }, + { + "epoch": 0.7929359517035598, + "grad_norm": 2.0601353645324707, + "learning_rate": 0.00018552117079356404, + "loss": 0.7929, + "step": 11427 + }, + { + "epoch": 0.7930053431406564, + "grad_norm": 2.1445071697235107, + "learning_rate": 0.00018551825864918336, + "loss": 1.1419, + "step": 11428 + }, + { + "epoch": 0.7930747345777531, + "grad_norm": 1.8564472198486328, + "learning_rate": 0.00018551534623483144, + "loss": 0.9188, + "step": 11429 + }, + { + "epoch": 0.7931441260148497, + "grad_norm": 2.0317249298095703, + "learning_rate": 0.00018551243355051752, + "loss": 0.8377, + "step": 11430 + }, + { + "epoch": 0.7932135174519465, + "grad_norm": 1.5139169692993164, + "learning_rate": 0.00018550952059625075, + "loss": 0.6607, + "step": 11431 + }, + { + "epoch": 0.7932829088890431, + "grad_norm": 2.508833169937134, + "learning_rate": 0.00018550660737204035, + "loss": 0.8802, + "step": 11432 + }, + { + "epoch": 0.7933523003261398, + "grad_norm": 1.5978692770004272, + "learning_rate": 0.0001855036938778955, + "loss": 0.7022, + "step": 11433 + }, + { + "epoch": 0.7934216917632364, + "grad_norm": 2.4482126235961914, + "learning_rate": 0.0001855007801138254, + "loss": 0.908, + "step": 11434 + }, + { + "epoch": 0.7934910832003331, + "grad_norm": 2.0792288780212402, + "learning_rate": 0.0001854978660798393, + "loss": 0.5667, + "step": 11435 + }, + { + "epoch": 0.7935604746374297, + "grad_norm": 1.8469425439834595, + "learning_rate": 0.0001854949517759463, + "loss": 0.8695, + "step": 11436 + }, + { + "epoch": 0.7936298660745265, + "grad_norm": 1.8228517770767212, + "learning_rate": 0.0001854920372021557, + "loss": 0.7807, + "step": 11437 + }, + { + "epoch": 0.7936992575116231, + "grad_norm": 2.1913464069366455, + "learning_rate": 0.00018548912235847665, + "loss": 0.7314, + "step": 11438 + }, + { + "epoch": 0.7937686489487197, + "grad_norm": 2.3751723766326904, + "learning_rate": 0.00018548620724491834, + "loss": 0.7239, + "step": 11439 + }, + { + "epoch": 0.7938380403858164, + "grad_norm": 2.4007179737091064, + "learning_rate": 0.00018548329186149004, + "loss": 0.7762, + "step": 11440 + }, + { + "epoch": 0.793907431822913, + "grad_norm": 2.089280128479004, + "learning_rate": 0.00018548037620820084, + "loss": 1.0008, + "step": 11441 + }, + { + "epoch": 0.7939768232600097, + "grad_norm": 2.289360284805298, + "learning_rate": 0.00018547746028506007, + "loss": 0.7647, + "step": 11442 + }, + { + "epoch": 0.7940462146971063, + "grad_norm": 2.129201889038086, + "learning_rate": 0.00018547454409207682, + "loss": 0.844, + "step": 11443 + }, + { + "epoch": 0.7941156061342031, + "grad_norm": 1.5752296447753906, + "learning_rate": 0.00018547162762926037, + "loss": 0.6739, + "step": 11444 + }, + { + "epoch": 0.7941849975712997, + "grad_norm": 4.090980529785156, + "learning_rate": 0.0001854687108966199, + "loss": 1.2028, + "step": 11445 + }, + { + "epoch": 0.7942543890083964, + "grad_norm": 2.3237366676330566, + "learning_rate": 0.00018546579389416468, + "loss": 0.7452, + "step": 11446 + }, + { + "epoch": 0.794323780445493, + "grad_norm": 2.534569501876831, + "learning_rate": 0.0001854628766219038, + "loss": 0.8286, + "step": 11447 + }, + { + "epoch": 0.7943931718825897, + "grad_norm": 1.8277612924575806, + "learning_rate": 0.00018545995907984656, + "loss": 0.6957, + "step": 11448 + }, + { + "epoch": 0.7944625633196863, + "grad_norm": 2.492356061935425, + "learning_rate": 0.00018545704126800215, + "loss": 0.897, + "step": 11449 + }, + { + "epoch": 0.7945319547567831, + "grad_norm": 1.7464922666549683, + "learning_rate": 0.0001854541231863798, + "loss": 0.672, + "step": 11450 + }, + { + "epoch": 0.7946013461938797, + "grad_norm": 4.303979873657227, + "learning_rate": 0.00018545120483498865, + "loss": 0.7704, + "step": 11451 + }, + { + "epoch": 0.7946707376309763, + "grad_norm": 2.4222450256347656, + "learning_rate": 0.00018544828621383798, + "loss": 0.7908, + "step": 11452 + }, + { + "epoch": 0.794740129068073, + "grad_norm": 2.1188907623291016, + "learning_rate": 0.00018544536732293697, + "loss": 0.8026, + "step": 11453 + }, + { + "epoch": 0.7948095205051696, + "grad_norm": 1.9832812547683716, + "learning_rate": 0.00018544244816229486, + "loss": 0.7755, + "step": 11454 + }, + { + "epoch": 0.7948789119422663, + "grad_norm": 2.4280664920806885, + "learning_rate": 0.00018543952873192087, + "loss": 0.8297, + "step": 11455 + }, + { + "epoch": 0.794948303379363, + "grad_norm": 2.0026025772094727, + "learning_rate": 0.00018543660903182416, + "loss": 0.5842, + "step": 11456 + }, + { + "epoch": 0.7950176948164597, + "grad_norm": 2.547466278076172, + "learning_rate": 0.00018543368906201403, + "loss": 0.9752, + "step": 11457 + }, + { + "epoch": 0.7950870862535563, + "grad_norm": 2.2706305980682373, + "learning_rate": 0.00018543076882249963, + "loss": 0.8039, + "step": 11458 + }, + { + "epoch": 0.795156477690653, + "grad_norm": 2.467879056930542, + "learning_rate": 0.00018542784831329022, + "loss": 0.9084, + "step": 11459 + }, + { + "epoch": 0.7952258691277496, + "grad_norm": 1.9854580163955688, + "learning_rate": 0.00018542492753439503, + "loss": 0.6868, + "step": 11460 + }, + { + "epoch": 0.7952952605648463, + "grad_norm": 1.581386685371399, + "learning_rate": 0.0001854220064858232, + "loss": 0.711, + "step": 11461 + }, + { + "epoch": 0.795364652001943, + "grad_norm": 2.513878345489502, + "learning_rate": 0.000185419085167584, + "loss": 0.9289, + "step": 11462 + }, + { + "epoch": 0.7954340434390397, + "grad_norm": 2.602869987487793, + "learning_rate": 0.00018541616357968673, + "loss": 0.9793, + "step": 11463 + }, + { + "epoch": 0.7955034348761363, + "grad_norm": 2.368901252746582, + "learning_rate": 0.0001854132417221405, + "loss": 0.8517, + "step": 11464 + }, + { + "epoch": 0.7955728263132329, + "grad_norm": 2.641794443130493, + "learning_rate": 0.00018541031959495453, + "loss": 0.8673, + "step": 11465 + }, + { + "epoch": 0.7956422177503296, + "grad_norm": 2.1307082176208496, + "learning_rate": 0.00018540739719813816, + "loss": 0.7407, + "step": 11466 + }, + { + "epoch": 0.7957116091874262, + "grad_norm": 2.438920021057129, + "learning_rate": 0.00018540447453170052, + "loss": 0.6676, + "step": 11467 + }, + { + "epoch": 0.795781000624523, + "grad_norm": 2.5286622047424316, + "learning_rate": 0.00018540155159565088, + "loss": 0.8658, + "step": 11468 + }, + { + "epoch": 0.7958503920616196, + "grad_norm": 2.6637589931488037, + "learning_rate": 0.00018539862838999843, + "loss": 0.7694, + "step": 11469 + }, + { + "epoch": 0.7959197834987163, + "grad_norm": 3.0233452320098877, + "learning_rate": 0.00018539570491475242, + "loss": 1.1713, + "step": 11470 + }, + { + "epoch": 0.7959891749358129, + "grad_norm": 2.169957399368286, + "learning_rate": 0.00018539278116992209, + "loss": 0.9631, + "step": 11471 + }, + { + "epoch": 0.7960585663729096, + "grad_norm": 2.5949785709381104, + "learning_rate": 0.00018538985715551665, + "loss": 0.7283, + "step": 11472 + }, + { + "epoch": 0.7961279578100062, + "grad_norm": 1.788701057434082, + "learning_rate": 0.00018538693287154531, + "loss": 0.7661, + "step": 11473 + }, + { + "epoch": 0.796197349247103, + "grad_norm": 2.741783618927002, + "learning_rate": 0.00018538400831801736, + "loss": 0.8578, + "step": 11474 + }, + { + "epoch": 0.7962667406841996, + "grad_norm": 2.142299175262451, + "learning_rate": 0.000185381083494942, + "loss": 0.6112, + "step": 11475 + }, + { + "epoch": 0.7963361321212963, + "grad_norm": 2.351012706756592, + "learning_rate": 0.00018537815840232846, + "loss": 0.8266, + "step": 11476 + }, + { + "epoch": 0.7964055235583929, + "grad_norm": 2.331665515899658, + "learning_rate": 0.000185375233040186, + "loss": 0.7365, + "step": 11477 + }, + { + "epoch": 0.7964749149954895, + "grad_norm": 2.8063321113586426, + "learning_rate": 0.0001853723074085238, + "loss": 0.9452, + "step": 11478 + }, + { + "epoch": 0.7965443064325862, + "grad_norm": 2.353691577911377, + "learning_rate": 0.00018536938150735116, + "loss": 1.0391, + "step": 11479 + }, + { + "epoch": 0.7966136978696828, + "grad_norm": 2.5354201793670654, + "learning_rate": 0.00018536645533667726, + "loss": 1.0576, + "step": 11480 + }, + { + "epoch": 0.7966830893067796, + "grad_norm": 2.458247661590576, + "learning_rate": 0.00018536352889651138, + "loss": 0.687, + "step": 11481 + }, + { + "epoch": 0.7967524807438762, + "grad_norm": 1.8274693489074707, + "learning_rate": 0.00018536060218686274, + "loss": 0.7481, + "step": 11482 + }, + { + "epoch": 0.7968218721809729, + "grad_norm": 2.293365716934204, + "learning_rate": 0.0001853576752077406, + "loss": 0.867, + "step": 11483 + }, + { + "epoch": 0.7968912636180695, + "grad_norm": 2.335231304168701, + "learning_rate": 0.00018535474795915415, + "loss": 0.9458, + "step": 11484 + }, + { + "epoch": 0.7969606550551662, + "grad_norm": 2.780200719833374, + "learning_rate": 0.0001853518204411127, + "loss": 1.0141, + "step": 11485 + }, + { + "epoch": 0.7970300464922628, + "grad_norm": 2.3358612060546875, + "learning_rate": 0.00018534889265362544, + "loss": 0.8091, + "step": 11486 + }, + { + "epoch": 0.7970994379293596, + "grad_norm": 2.1918187141418457, + "learning_rate": 0.00018534596459670162, + "loss": 0.6175, + "step": 11487 + }, + { + "epoch": 0.7971688293664562, + "grad_norm": 2.3039276599884033, + "learning_rate": 0.0001853430362703505, + "loss": 0.8135, + "step": 11488 + }, + { + "epoch": 0.7972382208035529, + "grad_norm": 2.5207924842834473, + "learning_rate": 0.0001853401076745813, + "loss": 0.9929, + "step": 11489 + }, + { + "epoch": 0.7973076122406495, + "grad_norm": 2.6763975620269775, + "learning_rate": 0.00018533717880940328, + "loss": 1.1505, + "step": 11490 + }, + { + "epoch": 0.7973770036777462, + "grad_norm": 2.4160776138305664, + "learning_rate": 0.00018533424967482573, + "loss": 1.2025, + "step": 11491 + }, + { + "epoch": 0.7974463951148428, + "grad_norm": 1.9977936744689941, + "learning_rate": 0.00018533132027085783, + "loss": 0.6976, + "step": 11492 + }, + { + "epoch": 0.7975157865519394, + "grad_norm": 1.6186949014663696, + "learning_rate": 0.00018532839059750883, + "loss": 0.7428, + "step": 11493 + }, + { + "epoch": 0.7975851779890362, + "grad_norm": 2.2182090282440186, + "learning_rate": 0.00018532546065478802, + "loss": 0.8023, + "step": 11494 + }, + { + "epoch": 0.7976545694261328, + "grad_norm": 2.212310791015625, + "learning_rate": 0.00018532253044270465, + "loss": 0.6583, + "step": 11495 + }, + { + "epoch": 0.7977239608632295, + "grad_norm": 2.051690101623535, + "learning_rate": 0.00018531959996126793, + "loss": 0.9195, + "step": 11496 + }, + { + "epoch": 0.7977933523003261, + "grad_norm": 2.258617639541626, + "learning_rate": 0.00018531666921048711, + "loss": 0.9518, + "step": 11497 + }, + { + "epoch": 0.7978627437374228, + "grad_norm": 1.8967667818069458, + "learning_rate": 0.0001853137381903715, + "loss": 0.9073, + "step": 11498 + }, + { + "epoch": 0.7979321351745194, + "grad_norm": 2.2737042903900146, + "learning_rate": 0.0001853108069009303, + "loss": 0.7077, + "step": 11499 + }, + { + "epoch": 0.7980015266116162, + "grad_norm": 2.061155319213867, + "learning_rate": 0.00018530787534217278, + "loss": 0.7227, + "step": 11500 + }, + { + "epoch": 0.7980709180487128, + "grad_norm": 1.9119435548782349, + "learning_rate": 0.00018530494351410822, + "loss": 0.8281, + "step": 11501 + }, + { + "epoch": 0.7981403094858095, + "grad_norm": 2.7460477352142334, + "learning_rate": 0.0001853020114167458, + "loss": 1.0428, + "step": 11502 + }, + { + "epoch": 0.7982097009229061, + "grad_norm": 2.181162118911743, + "learning_rate": 0.00018529907905009488, + "loss": 0.7899, + "step": 11503 + }, + { + "epoch": 0.7982790923600028, + "grad_norm": 2.0217714309692383, + "learning_rate": 0.00018529614641416464, + "loss": 0.7996, + "step": 11504 + }, + { + "epoch": 0.7983484837970994, + "grad_norm": 2.118762493133545, + "learning_rate": 0.00018529321350896436, + "loss": 0.9909, + "step": 11505 + }, + { + "epoch": 0.798417875234196, + "grad_norm": 2.0734570026397705, + "learning_rate": 0.00018529028033450331, + "loss": 0.7331, + "step": 11506 + }, + { + "epoch": 0.7984872666712928, + "grad_norm": 2.977015733718872, + "learning_rate": 0.0001852873468907907, + "loss": 0.8473, + "step": 11507 + }, + { + "epoch": 0.7985566581083894, + "grad_norm": 2.1487174034118652, + "learning_rate": 0.00018528441317783588, + "loss": 0.8967, + "step": 11508 + }, + { + "epoch": 0.7986260495454861, + "grad_norm": 2.5860140323638916, + "learning_rate": 0.00018528147919564804, + "loss": 1.0077, + "step": 11509 + }, + { + "epoch": 0.7986954409825827, + "grad_norm": 2.759625196456909, + "learning_rate": 0.00018527854494423647, + "loss": 0.8349, + "step": 11510 + }, + { + "epoch": 0.7987648324196794, + "grad_norm": 2.1886587142944336, + "learning_rate": 0.00018527561042361047, + "loss": 0.8256, + "step": 11511 + }, + { + "epoch": 0.798834223856776, + "grad_norm": 1.7489299774169922, + "learning_rate": 0.0001852726756337792, + "loss": 0.6326, + "step": 11512 + }, + { + "epoch": 0.7989036152938728, + "grad_norm": 2.216632843017578, + "learning_rate": 0.00018526974057475203, + "loss": 0.7277, + "step": 11513 + }, + { + "epoch": 0.7989730067309694, + "grad_norm": 2.2220888137817383, + "learning_rate": 0.00018526680524653813, + "loss": 0.6263, + "step": 11514 + }, + { + "epoch": 0.7990423981680661, + "grad_norm": 1.6668791770935059, + "learning_rate": 0.00018526386964914685, + "loss": 0.6193, + "step": 11515 + }, + { + "epoch": 0.7991117896051627, + "grad_norm": 2.4270660877227783, + "learning_rate": 0.0001852609337825874, + "loss": 0.7987, + "step": 11516 + }, + { + "epoch": 0.7991811810422594, + "grad_norm": 2.2293713092803955, + "learning_rate": 0.00018525799764686912, + "loss": 0.8801, + "step": 11517 + }, + { + "epoch": 0.799250572479356, + "grad_norm": 2.7402098178863525, + "learning_rate": 0.00018525506124200122, + "loss": 0.9281, + "step": 11518 + }, + { + "epoch": 0.7993199639164527, + "grad_norm": 3.0406229496002197, + "learning_rate": 0.00018525212456799297, + "loss": 0.9582, + "step": 11519 + }, + { + "epoch": 0.7993893553535494, + "grad_norm": 2.4855101108551025, + "learning_rate": 0.00018524918762485366, + "loss": 0.8994, + "step": 11520 + }, + { + "epoch": 0.799458746790646, + "grad_norm": 2.1956138610839844, + "learning_rate": 0.00018524625041259257, + "loss": 0.7246, + "step": 11521 + }, + { + "epoch": 0.7995281382277427, + "grad_norm": 2.3488929271698, + "learning_rate": 0.00018524331293121892, + "loss": 0.6356, + "step": 11522 + }, + { + "epoch": 0.7995975296648393, + "grad_norm": 2.2003746032714844, + "learning_rate": 0.00018524037518074203, + "loss": 0.8355, + "step": 11523 + }, + { + "epoch": 0.799666921101936, + "grad_norm": 1.8052424192428589, + "learning_rate": 0.00018523743716117118, + "loss": 0.7405, + "step": 11524 + }, + { + "epoch": 0.7997363125390327, + "grad_norm": 1.713025450706482, + "learning_rate": 0.00018523449887251564, + "loss": 0.6816, + "step": 11525 + }, + { + "epoch": 0.7998057039761294, + "grad_norm": 1.9538700580596924, + "learning_rate": 0.00018523156031478467, + "loss": 0.7268, + "step": 11526 + }, + { + "epoch": 0.799875095413226, + "grad_norm": 2.3543219566345215, + "learning_rate": 0.00018522862148798752, + "loss": 0.695, + "step": 11527 + }, + { + "epoch": 0.7999444868503227, + "grad_norm": 3.285698652267456, + "learning_rate": 0.00018522568239213354, + "loss": 0.6822, + "step": 11528 + }, + { + "epoch": 0.8000138782874193, + "grad_norm": 2.7618181705474854, + "learning_rate": 0.00018522274302723197, + "loss": 0.9537, + "step": 11529 + }, + { + "epoch": 0.800083269724516, + "grad_norm": 2.3335511684417725, + "learning_rate": 0.00018521980339329205, + "loss": 0.8228, + "step": 11530 + }, + { + "epoch": 0.8001526611616127, + "grad_norm": 1.9792592525482178, + "learning_rate": 0.00018521686349032315, + "loss": 0.7276, + "step": 11531 + }, + { + "epoch": 0.8002220525987093, + "grad_norm": 2.8049933910369873, + "learning_rate": 0.00018521392331833446, + "loss": 0.9818, + "step": 11532 + }, + { + "epoch": 0.800291444035806, + "grad_norm": 2.360659122467041, + "learning_rate": 0.00018521098287733532, + "loss": 0.7319, + "step": 11533 + }, + { + "epoch": 0.8003608354729026, + "grad_norm": 2.123044967651367, + "learning_rate": 0.00018520804216733496, + "loss": 0.9566, + "step": 11534 + }, + { + "epoch": 0.8004302269099993, + "grad_norm": 2.102626085281372, + "learning_rate": 0.00018520510118834273, + "loss": 0.7648, + "step": 11535 + }, + { + "epoch": 0.8004996183470959, + "grad_norm": 2.1109225749969482, + "learning_rate": 0.0001852021599403679, + "loss": 0.8884, + "step": 11536 + }, + { + "epoch": 0.8005690097841927, + "grad_norm": 3.6617960929870605, + "learning_rate": 0.0001851992184234197, + "loss": 0.962, + "step": 11537 + }, + { + "epoch": 0.8006384012212893, + "grad_norm": 2.309803009033203, + "learning_rate": 0.00018519627663750744, + "loss": 0.8045, + "step": 11538 + }, + { + "epoch": 0.800707792658386, + "grad_norm": 1.9194709062576294, + "learning_rate": 0.00018519333458264046, + "loss": 0.7055, + "step": 11539 + }, + { + "epoch": 0.8007771840954826, + "grad_norm": 1.744250774383545, + "learning_rate": 0.00018519039225882797, + "loss": 0.7075, + "step": 11540 + }, + { + "epoch": 0.8008465755325793, + "grad_norm": 1.743483066558838, + "learning_rate": 0.00018518744966607932, + "loss": 0.6277, + "step": 11541 + }, + { + "epoch": 0.8009159669696759, + "grad_norm": 3.5148158073425293, + "learning_rate": 0.00018518450680440375, + "loss": 0.8012, + "step": 11542 + }, + { + "epoch": 0.8009853584067727, + "grad_norm": 2.797562837600708, + "learning_rate": 0.00018518156367381062, + "loss": 0.8937, + "step": 11543 + }, + { + "epoch": 0.8010547498438693, + "grad_norm": 2.168668270111084, + "learning_rate": 0.00018517862027430914, + "loss": 0.7477, + "step": 11544 + }, + { + "epoch": 0.8011241412809659, + "grad_norm": 2.6906545162200928, + "learning_rate": 0.00018517567660590863, + "loss": 0.9097, + "step": 11545 + }, + { + "epoch": 0.8011935327180626, + "grad_norm": 2.06866717338562, + "learning_rate": 0.00018517273266861842, + "loss": 0.5549, + "step": 11546 + }, + { + "epoch": 0.8012629241551592, + "grad_norm": 1.9510732889175415, + "learning_rate": 0.00018516978846244779, + "loss": 0.8416, + "step": 11547 + }, + { + "epoch": 0.8013323155922559, + "grad_norm": 2.1513724327087402, + "learning_rate": 0.00018516684398740599, + "loss": 0.9629, + "step": 11548 + }, + { + "epoch": 0.8014017070293525, + "grad_norm": 2.4771788120269775, + "learning_rate": 0.00018516389924350235, + "loss": 0.7004, + "step": 11549 + }, + { + "epoch": 0.8014710984664493, + "grad_norm": 2.3500006198883057, + "learning_rate": 0.00018516095423074615, + "loss": 0.8216, + "step": 11550 + }, + { + "epoch": 0.8015404899035459, + "grad_norm": 1.9776114225387573, + "learning_rate": 0.0001851580089491467, + "loss": 0.9067, + "step": 11551 + }, + { + "epoch": 0.8016098813406426, + "grad_norm": 2.294224262237549, + "learning_rate": 0.0001851550633987133, + "loss": 0.8565, + "step": 11552 + }, + { + "epoch": 0.8016792727777392, + "grad_norm": 2.2498888969421387, + "learning_rate": 0.00018515211757945526, + "loss": 0.7607, + "step": 11553 + }, + { + "epoch": 0.8017486642148359, + "grad_norm": 1.8356645107269287, + "learning_rate": 0.00018514917149138184, + "loss": 0.8358, + "step": 11554 + }, + { + "epoch": 0.8018180556519325, + "grad_norm": 2.1460964679718018, + "learning_rate": 0.00018514622513450237, + "loss": 0.7774, + "step": 11555 + }, + { + "epoch": 0.8018874470890293, + "grad_norm": 2.516386032104492, + "learning_rate": 0.00018514327850882616, + "loss": 0.8197, + "step": 11556 + }, + { + "epoch": 0.8019568385261259, + "grad_norm": 1.6558897495269775, + "learning_rate": 0.0001851403316143625, + "loss": 0.561, + "step": 11557 + }, + { + "epoch": 0.8020262299632226, + "grad_norm": 1.799789309501648, + "learning_rate": 0.00018513738445112066, + "loss": 0.6021, + "step": 11558 + }, + { + "epoch": 0.8020956214003192, + "grad_norm": 1.9575215578079224, + "learning_rate": 0.00018513443701911002, + "loss": 0.7228, + "step": 11559 + }, + { + "epoch": 0.8021650128374158, + "grad_norm": 2.526352882385254, + "learning_rate": 0.0001851314893183398, + "loss": 0.931, + "step": 11560 + }, + { + "epoch": 0.8022344042745125, + "grad_norm": 2.246206283569336, + "learning_rate": 0.00018512854134881938, + "loss": 0.8093, + "step": 11561 + }, + { + "epoch": 0.8023037957116091, + "grad_norm": 2.521148681640625, + "learning_rate": 0.00018512559311055802, + "loss": 0.613, + "step": 11562 + }, + { + "epoch": 0.8023731871487059, + "grad_norm": 3.412757635116577, + "learning_rate": 0.00018512264460356504, + "loss": 0.8332, + "step": 11563 + }, + { + "epoch": 0.8024425785858025, + "grad_norm": 2.623188018798828, + "learning_rate": 0.00018511969582784974, + "loss": 0.7989, + "step": 11564 + }, + { + "epoch": 0.8025119700228992, + "grad_norm": 1.7029153108596802, + "learning_rate": 0.00018511674678342143, + "loss": 0.5777, + "step": 11565 + }, + { + "epoch": 0.8025813614599958, + "grad_norm": 2.3917179107666016, + "learning_rate": 0.00018511379747028942, + "loss": 1.0066, + "step": 11566 + }, + { + "epoch": 0.8026507528970925, + "grad_norm": 2.2817461490631104, + "learning_rate": 0.00018511084788846305, + "loss": 0.8078, + "step": 11567 + }, + { + "epoch": 0.8027201443341891, + "grad_norm": 2.1634130477905273, + "learning_rate": 0.0001851078980379516, + "loss": 0.8341, + "step": 11568 + }, + { + "epoch": 0.8027895357712859, + "grad_norm": 2.2899887561798096, + "learning_rate": 0.00018510494791876438, + "loss": 0.7592, + "step": 11569 + }, + { + "epoch": 0.8028589272083825, + "grad_norm": 2.385890483856201, + "learning_rate": 0.0001851019975309107, + "loss": 0.8536, + "step": 11570 + }, + { + "epoch": 0.8029283186454792, + "grad_norm": 1.5994538068771362, + "learning_rate": 0.0001850990468743999, + "loss": 0.6343, + "step": 11571 + }, + { + "epoch": 0.8029977100825758, + "grad_norm": 1.950365662574768, + "learning_rate": 0.0001850960959492413, + "loss": 0.713, + "step": 11572 + }, + { + "epoch": 0.8030671015196724, + "grad_norm": 1.755776286125183, + "learning_rate": 0.0001850931447554442, + "loss": 0.7533, + "step": 11573 + }, + { + "epoch": 0.8031364929567691, + "grad_norm": 1.8257583379745483, + "learning_rate": 0.0001850901932930179, + "loss": 0.6025, + "step": 11574 + }, + { + "epoch": 0.8032058843938658, + "grad_norm": 1.828578233718872, + "learning_rate": 0.00018508724156197172, + "loss": 0.8909, + "step": 11575 + }, + { + "epoch": 0.8032752758309625, + "grad_norm": 2.0184061527252197, + "learning_rate": 0.00018508428956231504, + "loss": 0.7025, + "step": 11576 + }, + { + "epoch": 0.8033446672680591, + "grad_norm": 2.2100017070770264, + "learning_rate": 0.00018508133729405708, + "loss": 0.9027, + "step": 11577 + }, + { + "epoch": 0.8034140587051558, + "grad_norm": 2.163334369659424, + "learning_rate": 0.00018507838475720722, + "loss": 1.1518, + "step": 11578 + }, + { + "epoch": 0.8034834501422524, + "grad_norm": 2.3215928077697754, + "learning_rate": 0.00018507543195177476, + "loss": 0.6376, + "step": 11579 + }, + { + "epoch": 0.8035528415793491, + "grad_norm": 1.8904731273651123, + "learning_rate": 0.00018507247887776906, + "loss": 0.7968, + "step": 11580 + }, + { + "epoch": 0.8036222330164458, + "grad_norm": 1.5916770696640015, + "learning_rate": 0.0001850695255351994, + "loss": 0.5841, + "step": 11581 + }, + { + "epoch": 0.8036916244535425, + "grad_norm": 2.055088520050049, + "learning_rate": 0.00018506657192407514, + "loss": 0.7849, + "step": 11582 + }, + { + "epoch": 0.8037610158906391, + "grad_norm": 2.074284315109253, + "learning_rate": 0.00018506361804440554, + "loss": 0.6917, + "step": 11583 + }, + { + "epoch": 0.8038304073277358, + "grad_norm": 1.6713382005691528, + "learning_rate": 0.00018506066389619997, + "loss": 0.9044, + "step": 11584 + }, + { + "epoch": 0.8038997987648324, + "grad_norm": 1.9900784492492676, + "learning_rate": 0.0001850577094794678, + "loss": 0.7743, + "step": 11585 + }, + { + "epoch": 0.803969190201929, + "grad_norm": 2.930194616317749, + "learning_rate": 0.00018505475479421826, + "loss": 0.8537, + "step": 11586 + }, + { + "epoch": 0.8040385816390258, + "grad_norm": 2.3294336795806885, + "learning_rate": 0.00018505179984046075, + "loss": 0.709, + "step": 11587 + }, + { + "epoch": 0.8041079730761224, + "grad_norm": 3.006635904312134, + "learning_rate": 0.00018504884461820458, + "loss": 0.9842, + "step": 11588 + }, + { + "epoch": 0.8041773645132191, + "grad_norm": 1.8636314868927002, + "learning_rate": 0.00018504588912745905, + "loss": 0.5616, + "step": 11589 + }, + { + "epoch": 0.8042467559503157, + "grad_norm": 3.1281650066375732, + "learning_rate": 0.00018504293336823354, + "loss": 1.0375, + "step": 11590 + }, + { + "epoch": 0.8043161473874124, + "grad_norm": 2.088674306869507, + "learning_rate": 0.00018503997734053735, + "loss": 1.0219, + "step": 11591 + }, + { + "epoch": 0.804385538824509, + "grad_norm": 2.233553409576416, + "learning_rate": 0.00018503702104437985, + "loss": 0.9151, + "step": 11592 + }, + { + "epoch": 0.8044549302616057, + "grad_norm": 3.003264904022217, + "learning_rate": 0.0001850340644797703, + "loss": 0.8268, + "step": 11593 + }, + { + "epoch": 0.8045243216987024, + "grad_norm": 1.9617000818252563, + "learning_rate": 0.00018503110764671806, + "loss": 0.8501, + "step": 11594 + }, + { + "epoch": 0.8045937131357991, + "grad_norm": 3.114630699157715, + "learning_rate": 0.0001850281505452325, + "loss": 1.0213, + "step": 11595 + }, + { + "epoch": 0.8046631045728957, + "grad_norm": 2.045438766479492, + "learning_rate": 0.00018502519317532294, + "loss": 0.9011, + "step": 11596 + }, + { + "epoch": 0.8047324960099924, + "grad_norm": 2.933627128601074, + "learning_rate": 0.0001850222355369987, + "loss": 0.8064, + "step": 11597 + }, + { + "epoch": 0.804801887447089, + "grad_norm": 2.54046368598938, + "learning_rate": 0.00018501927763026914, + "loss": 1.0876, + "step": 11598 + }, + { + "epoch": 0.8048712788841856, + "grad_norm": 1.5096549987792969, + "learning_rate": 0.00018501631945514358, + "loss": 0.7285, + "step": 11599 + }, + { + "epoch": 0.8049406703212824, + "grad_norm": 1.8965222835540771, + "learning_rate": 0.00018501336101163135, + "loss": 0.8797, + "step": 11600 + }, + { + "epoch": 0.805010061758379, + "grad_norm": 2.5731499195098877, + "learning_rate": 0.00018501040229974182, + "loss": 0.9333, + "step": 11601 + }, + { + "epoch": 0.8050794531954757, + "grad_norm": 1.9473127126693726, + "learning_rate": 0.00018500744331948428, + "loss": 0.8796, + "step": 11602 + }, + { + "epoch": 0.8051488446325723, + "grad_norm": 2.556546449661255, + "learning_rate": 0.00018500448407086814, + "loss": 0.8879, + "step": 11603 + }, + { + "epoch": 0.805218236069669, + "grad_norm": 2.2133166790008545, + "learning_rate": 0.0001850015245539027, + "loss": 0.7689, + "step": 11604 + }, + { + "epoch": 0.8052876275067656, + "grad_norm": 1.5366100072860718, + "learning_rate": 0.0001849985647685973, + "loss": 0.6183, + "step": 11605 + }, + { + "epoch": 0.8053570189438624, + "grad_norm": 2.331383466720581, + "learning_rate": 0.0001849956047149613, + "loss": 0.8327, + "step": 11606 + }, + { + "epoch": 0.805426410380959, + "grad_norm": 1.7859052419662476, + "learning_rate": 0.00018499264439300402, + "loss": 0.7203, + "step": 11607 + }, + { + "epoch": 0.8054958018180557, + "grad_norm": 2.963505744934082, + "learning_rate": 0.00018498968380273485, + "loss": 0.8916, + "step": 11608 + }, + { + "epoch": 0.8055651932551523, + "grad_norm": 2.6092476844787598, + "learning_rate": 0.00018498672294416308, + "loss": 0.5824, + "step": 11609 + }, + { + "epoch": 0.805634584692249, + "grad_norm": 2.4253718852996826, + "learning_rate": 0.0001849837618172981, + "loss": 0.6772, + "step": 11610 + }, + { + "epoch": 0.8057039761293456, + "grad_norm": 1.8280202150344849, + "learning_rate": 0.00018498080042214925, + "loss": 0.6937, + "step": 11611 + }, + { + "epoch": 0.8057733675664424, + "grad_norm": 1.9700039625167847, + "learning_rate": 0.00018497783875872587, + "loss": 0.5911, + "step": 11612 + }, + { + "epoch": 0.805842759003539, + "grad_norm": 2.6740520000457764, + "learning_rate": 0.00018497487682703732, + "loss": 1.1128, + "step": 11613 + }, + { + "epoch": 0.8059121504406356, + "grad_norm": 2.5599992275238037, + "learning_rate": 0.00018497191462709292, + "loss": 0.7855, + "step": 11614 + }, + { + "epoch": 0.8059815418777323, + "grad_norm": 1.702648401260376, + "learning_rate": 0.00018496895215890205, + "loss": 0.5206, + "step": 11615 + }, + { + "epoch": 0.8060509333148289, + "grad_norm": 2.0368471145629883, + "learning_rate": 0.00018496598942247406, + "loss": 0.8149, + "step": 11616 + }, + { + "epoch": 0.8061203247519256, + "grad_norm": 2.5151894092559814, + "learning_rate": 0.00018496302641781833, + "loss": 0.8537, + "step": 11617 + }, + { + "epoch": 0.8061897161890222, + "grad_norm": 2.1469614505767822, + "learning_rate": 0.00018496006314494415, + "loss": 0.7189, + "step": 11618 + }, + { + "epoch": 0.806259107626119, + "grad_norm": 1.8603757619857788, + "learning_rate": 0.0001849570996038609, + "loss": 0.656, + "step": 11619 + }, + { + "epoch": 0.8063284990632156, + "grad_norm": 2.031195640563965, + "learning_rate": 0.00018495413579457795, + "loss": 0.7523, + "step": 11620 + }, + { + "epoch": 0.8063978905003123, + "grad_norm": 2.232036828994751, + "learning_rate": 0.00018495117171710469, + "loss": 0.6374, + "step": 11621 + }, + { + "epoch": 0.8064672819374089, + "grad_norm": 2.5724472999572754, + "learning_rate": 0.0001849482073714504, + "loss": 0.8506, + "step": 11622 + }, + { + "epoch": 0.8065366733745056, + "grad_norm": 1.992332935333252, + "learning_rate": 0.00018494524275762447, + "loss": 0.7037, + "step": 11623 + }, + { + "epoch": 0.8066060648116022, + "grad_norm": 1.877589225769043, + "learning_rate": 0.00018494227787563627, + "loss": 0.5916, + "step": 11624 + }, + { + "epoch": 0.806675456248699, + "grad_norm": 2.111966609954834, + "learning_rate": 0.00018493931272549517, + "loss": 0.5704, + "step": 11625 + }, + { + "epoch": 0.8067448476857956, + "grad_norm": 1.9754027128219604, + "learning_rate": 0.0001849363473072105, + "loss": 0.6528, + "step": 11626 + }, + { + "epoch": 0.8068142391228922, + "grad_norm": 2.1960959434509277, + "learning_rate": 0.0001849333816207916, + "loss": 0.6258, + "step": 11627 + }, + { + "epoch": 0.8068836305599889, + "grad_norm": 2.1373064517974854, + "learning_rate": 0.00018493041566624793, + "loss": 0.5682, + "step": 11628 + }, + { + "epoch": 0.8069530219970855, + "grad_norm": 2.405482769012451, + "learning_rate": 0.00018492744944358878, + "loss": 1.0985, + "step": 11629 + }, + { + "epoch": 0.8070224134341822, + "grad_norm": 2.3537240028381348, + "learning_rate": 0.0001849244829528235, + "loss": 0.6859, + "step": 11630 + }, + { + "epoch": 0.8070918048712789, + "grad_norm": 2.631857395172119, + "learning_rate": 0.0001849215161939615, + "loss": 0.9727, + "step": 11631 + }, + { + "epoch": 0.8071611963083756, + "grad_norm": 2.4269046783447266, + "learning_rate": 0.0001849185491670121, + "loss": 0.8881, + "step": 11632 + }, + { + "epoch": 0.8072305877454722, + "grad_norm": 3.354435682296753, + "learning_rate": 0.00018491558187198472, + "loss": 1.0955, + "step": 11633 + }, + { + "epoch": 0.8072999791825689, + "grad_norm": 2.978947639465332, + "learning_rate": 0.00018491261430888872, + "loss": 0.791, + "step": 11634 + }, + { + "epoch": 0.8073693706196655, + "grad_norm": 1.7471574544906616, + "learning_rate": 0.00018490964647773338, + "loss": 0.6222, + "step": 11635 + }, + { + "epoch": 0.8074387620567622, + "grad_norm": 2.0692641735076904, + "learning_rate": 0.00018490667837852816, + "loss": 0.8665, + "step": 11636 + }, + { + "epoch": 0.8075081534938588, + "grad_norm": 2.2778334617614746, + "learning_rate": 0.00018490371001128245, + "loss": 0.9564, + "step": 11637 + }, + { + "epoch": 0.8075775449309556, + "grad_norm": 2.4888243675231934, + "learning_rate": 0.00018490074137600556, + "loss": 0.9899, + "step": 11638 + }, + { + "epoch": 0.8076469363680522, + "grad_norm": 2.200134515762329, + "learning_rate": 0.00018489777247270687, + "loss": 0.8181, + "step": 11639 + }, + { + "epoch": 0.8077163278051488, + "grad_norm": 2.5136499404907227, + "learning_rate": 0.00018489480330139576, + "loss": 0.7195, + "step": 11640 + }, + { + "epoch": 0.8077857192422455, + "grad_norm": 2.459749221801758, + "learning_rate": 0.0001848918338620816, + "loss": 0.9135, + "step": 11641 + }, + { + "epoch": 0.8078551106793421, + "grad_norm": 2.1440396308898926, + "learning_rate": 0.0001848888641547738, + "loss": 0.5496, + "step": 11642 + }, + { + "epoch": 0.8079245021164388, + "grad_norm": 2.4268224239349365, + "learning_rate": 0.00018488589417948168, + "loss": 1.1091, + "step": 11643 + }, + { + "epoch": 0.8079938935535355, + "grad_norm": 2.674513339996338, + "learning_rate": 0.00018488292393621466, + "loss": 0.8179, + "step": 11644 + }, + { + "epoch": 0.8080632849906322, + "grad_norm": 2.05332088470459, + "learning_rate": 0.00018487995342498207, + "loss": 0.7301, + "step": 11645 + }, + { + "epoch": 0.8081326764277288, + "grad_norm": 2.4034461975097656, + "learning_rate": 0.00018487698264579332, + "loss": 0.6804, + "step": 11646 + }, + { + "epoch": 0.8082020678648255, + "grad_norm": 1.9913580417633057, + "learning_rate": 0.0001848740115986578, + "loss": 0.675, + "step": 11647 + }, + { + "epoch": 0.8082714593019221, + "grad_norm": 2.1528079509735107, + "learning_rate": 0.00018487104028358487, + "loss": 0.7847, + "step": 11648 + }, + { + "epoch": 0.8083408507390188, + "grad_norm": 1.8658199310302734, + "learning_rate": 0.00018486806870058388, + "loss": 0.9111, + "step": 11649 + }, + { + "epoch": 0.8084102421761155, + "grad_norm": 2.2475533485412598, + "learning_rate": 0.00018486509684966431, + "loss": 0.8302, + "step": 11650 + }, + { + "epoch": 0.8084796336132122, + "grad_norm": 1.6673749685287476, + "learning_rate": 0.0001848621247308354, + "loss": 0.6907, + "step": 11651 + }, + { + "epoch": 0.8085490250503088, + "grad_norm": 1.8712594509124756, + "learning_rate": 0.00018485915234410668, + "loss": 0.8109, + "step": 11652 + }, + { + "epoch": 0.8086184164874054, + "grad_norm": 2.2504701614379883, + "learning_rate": 0.00018485617968948742, + "loss": 0.7639, + "step": 11653 + }, + { + "epoch": 0.8086878079245021, + "grad_norm": 2.287419319152832, + "learning_rate": 0.00018485320676698704, + "loss": 0.7087, + "step": 11654 + }, + { + "epoch": 0.8087571993615987, + "grad_norm": 2.196472644805908, + "learning_rate": 0.00018485023357661494, + "loss": 0.7343, + "step": 11655 + }, + { + "epoch": 0.8088265907986955, + "grad_norm": 2.356778144836426, + "learning_rate": 0.0001848472601183805, + "loss": 0.7804, + "step": 11656 + }, + { + "epoch": 0.8088959822357921, + "grad_norm": 2.145259141921997, + "learning_rate": 0.0001848442863922931, + "loss": 0.8479, + "step": 11657 + }, + { + "epoch": 0.8089653736728888, + "grad_norm": 2.2148377895355225, + "learning_rate": 0.00018484131239836213, + "loss": 0.8189, + "step": 11658 + }, + { + "epoch": 0.8090347651099854, + "grad_norm": 1.9371248483657837, + "learning_rate": 0.00018483833813659698, + "loss": 0.7997, + "step": 11659 + }, + { + "epoch": 0.8091041565470821, + "grad_norm": 3.2033705711364746, + "learning_rate": 0.00018483536360700705, + "loss": 1.048, + "step": 11660 + }, + { + "epoch": 0.8091735479841787, + "grad_norm": 1.920444369316101, + "learning_rate": 0.00018483238880960174, + "loss": 0.7249, + "step": 11661 + }, + { + "epoch": 0.8092429394212755, + "grad_norm": 2.1862356662750244, + "learning_rate": 0.00018482941374439038, + "loss": 0.8261, + "step": 11662 + }, + { + "epoch": 0.8093123308583721, + "grad_norm": 2.083585262298584, + "learning_rate": 0.00018482643841138244, + "loss": 0.728, + "step": 11663 + }, + { + "epoch": 0.8093817222954688, + "grad_norm": 2.1204214096069336, + "learning_rate": 0.00018482346281058725, + "loss": 0.7917, + "step": 11664 + }, + { + "epoch": 0.8094511137325654, + "grad_norm": 1.7495027780532837, + "learning_rate": 0.00018482048694201423, + "loss": 0.5205, + "step": 11665 + }, + { + "epoch": 0.8095205051696621, + "grad_norm": 2.485957384109497, + "learning_rate": 0.00018481751080567276, + "loss": 1.098, + "step": 11666 + }, + { + "epoch": 0.8095898966067587, + "grad_norm": 2.3655552864074707, + "learning_rate": 0.00018481453440157225, + "loss": 0.9688, + "step": 11667 + }, + { + "epoch": 0.8096592880438553, + "grad_norm": 2.4427568912506104, + "learning_rate": 0.0001848115577297221, + "loss": 1.0256, + "step": 11668 + }, + { + "epoch": 0.8097286794809521, + "grad_norm": 2.11085844039917, + "learning_rate": 0.00018480858079013174, + "loss": 0.8879, + "step": 11669 + }, + { + "epoch": 0.8097980709180487, + "grad_norm": 1.4582329988479614, + "learning_rate": 0.0001848056035828105, + "loss": 0.6163, + "step": 11670 + }, + { + "epoch": 0.8098674623551454, + "grad_norm": 3.018171787261963, + "learning_rate": 0.00018480262610776781, + "loss": 0.9627, + "step": 11671 + }, + { + "epoch": 0.809936853792242, + "grad_norm": 2.7533228397369385, + "learning_rate": 0.00018479964836501308, + "loss": 0.8605, + "step": 11672 + }, + { + "epoch": 0.8100062452293387, + "grad_norm": 2.3073482513427734, + "learning_rate": 0.0001847966703545557, + "loss": 1.0853, + "step": 11673 + }, + { + "epoch": 0.8100756366664353, + "grad_norm": 1.570376992225647, + "learning_rate": 0.00018479369207640505, + "loss": 0.5944, + "step": 11674 + }, + { + "epoch": 0.8101450281035321, + "grad_norm": 1.6967048645019531, + "learning_rate": 0.00018479071353057056, + "loss": 0.6876, + "step": 11675 + }, + { + "epoch": 0.8102144195406287, + "grad_norm": 2.3480923175811768, + "learning_rate": 0.00018478773471706164, + "loss": 0.8457, + "step": 11676 + }, + { + "epoch": 0.8102838109777254, + "grad_norm": 2.3465352058410645, + "learning_rate": 0.00018478475563588764, + "loss": 0.7804, + "step": 11677 + }, + { + "epoch": 0.810353202414822, + "grad_norm": 2.268397569656372, + "learning_rate": 0.00018478177628705804, + "loss": 0.9206, + "step": 11678 + }, + { + "epoch": 0.8104225938519187, + "grad_norm": 2.2316267490386963, + "learning_rate": 0.0001847787966705822, + "loss": 0.7286, + "step": 11679 + }, + { + "epoch": 0.8104919852890153, + "grad_norm": 1.8227968215942383, + "learning_rate": 0.00018477581678646955, + "loss": 0.7609, + "step": 11680 + }, + { + "epoch": 0.810561376726112, + "grad_norm": 1.9083445072174072, + "learning_rate": 0.00018477283663472944, + "loss": 0.7822, + "step": 11681 + }, + { + "epoch": 0.8106307681632087, + "grad_norm": 2.6908183097839355, + "learning_rate": 0.00018476985621537133, + "loss": 1.0094, + "step": 11682 + }, + { + "epoch": 0.8107001596003053, + "grad_norm": 2.76271390914917, + "learning_rate": 0.00018476687552840466, + "loss": 0.9201, + "step": 11683 + }, + { + "epoch": 0.810769551037402, + "grad_norm": 2.259593963623047, + "learning_rate": 0.00018476389457383876, + "loss": 0.8116, + "step": 11684 + }, + { + "epoch": 0.8108389424744986, + "grad_norm": 2.264554023742676, + "learning_rate": 0.0001847609133516831, + "loss": 0.8162, + "step": 11685 + }, + { + "epoch": 0.8109083339115953, + "grad_norm": 2.147860050201416, + "learning_rate": 0.00018475793186194706, + "loss": 0.9818, + "step": 11686 + }, + { + "epoch": 0.810977725348692, + "grad_norm": 3.122234582901001, + "learning_rate": 0.00018475495010464006, + "loss": 0.9029, + "step": 11687 + }, + { + "epoch": 0.8110471167857887, + "grad_norm": 1.739938735961914, + "learning_rate": 0.00018475196807977152, + "loss": 0.7338, + "step": 11688 + }, + { + "epoch": 0.8111165082228853, + "grad_norm": 2.241379499435425, + "learning_rate": 0.00018474898578735085, + "loss": 0.996, + "step": 11689 + }, + { + "epoch": 0.811185899659982, + "grad_norm": 1.3546005487442017, + "learning_rate": 0.00018474600322738747, + "loss": 0.5497, + "step": 11690 + }, + { + "epoch": 0.8112552910970786, + "grad_norm": 1.7432504892349243, + "learning_rate": 0.00018474302039989078, + "loss": 0.8201, + "step": 11691 + }, + { + "epoch": 0.8113246825341753, + "grad_norm": 1.9106237888336182, + "learning_rate": 0.0001847400373048702, + "loss": 0.7515, + "step": 11692 + }, + { + "epoch": 0.811394073971272, + "grad_norm": 2.9640445709228516, + "learning_rate": 0.00018473705394233517, + "loss": 1.3965, + "step": 11693 + }, + { + "epoch": 0.8114634654083686, + "grad_norm": 1.9033355712890625, + "learning_rate": 0.00018473407031229509, + "loss": 0.8721, + "step": 11694 + }, + { + "epoch": 0.8115328568454653, + "grad_norm": 2.1633970737457275, + "learning_rate": 0.00018473108641475937, + "loss": 0.6543, + "step": 11695 + }, + { + "epoch": 0.8116022482825619, + "grad_norm": 2.516043186187744, + "learning_rate": 0.00018472810224973746, + "loss": 0.8321, + "step": 11696 + }, + { + "epoch": 0.8116716397196586, + "grad_norm": 1.7979124784469604, + "learning_rate": 0.00018472511781723877, + "loss": 0.607, + "step": 11697 + }, + { + "epoch": 0.8117410311567552, + "grad_norm": 2.905917167663574, + "learning_rate": 0.00018472213311727267, + "loss": 1.0276, + "step": 11698 + }, + { + "epoch": 0.8118104225938519, + "grad_norm": 2.0167040824890137, + "learning_rate": 0.00018471914814984866, + "loss": 0.6128, + "step": 11699 + }, + { + "epoch": 0.8118798140309486, + "grad_norm": 1.993158221244812, + "learning_rate": 0.0001847161629149761, + "loss": 0.7037, + "step": 11700 + }, + { + "epoch": 0.8119492054680453, + "grad_norm": 2.3460516929626465, + "learning_rate": 0.00018471317741266448, + "loss": 1.2118, + "step": 11701 + }, + { + "epoch": 0.8120185969051419, + "grad_norm": 2.7038559913635254, + "learning_rate": 0.00018471019164292317, + "loss": 1.0517, + "step": 11702 + }, + { + "epoch": 0.8120879883422386, + "grad_norm": 2.363283634185791, + "learning_rate": 0.00018470720560576157, + "loss": 0.8712, + "step": 11703 + }, + { + "epoch": 0.8121573797793352, + "grad_norm": 2.0618703365325928, + "learning_rate": 0.0001847042193011892, + "loss": 0.8973, + "step": 11704 + }, + { + "epoch": 0.8122267712164319, + "grad_norm": 1.8505491018295288, + "learning_rate": 0.00018470123272921544, + "loss": 0.7748, + "step": 11705 + }, + { + "epoch": 0.8122961626535286, + "grad_norm": 2.0936481952667236, + "learning_rate": 0.00018469824588984973, + "loss": 0.6759, + "step": 11706 + }, + { + "epoch": 0.8123655540906252, + "grad_norm": 2.1526012420654297, + "learning_rate": 0.00018469525878310147, + "loss": 0.7611, + "step": 11707 + }, + { + "epoch": 0.8124349455277219, + "grad_norm": 1.8450366258621216, + "learning_rate": 0.00018469227140898003, + "loss": 0.763, + "step": 11708 + }, + { + "epoch": 0.8125043369648185, + "grad_norm": 2.3567070960998535, + "learning_rate": 0.00018468928376749502, + "loss": 0.9286, + "step": 11709 + }, + { + "epoch": 0.8125737284019152, + "grad_norm": 2.2231791019439697, + "learning_rate": 0.0001846862958586557, + "loss": 0.7057, + "step": 11710 + }, + { + "epoch": 0.8126431198390118, + "grad_norm": 1.842575192451477, + "learning_rate": 0.00018468330768247162, + "loss": 0.759, + "step": 11711 + }, + { + "epoch": 0.8127125112761086, + "grad_norm": 1.5413891077041626, + "learning_rate": 0.00018468031923895211, + "loss": 0.4967, + "step": 11712 + }, + { + "epoch": 0.8127819027132052, + "grad_norm": 1.7089591026306152, + "learning_rate": 0.00018467733052810668, + "loss": 0.6254, + "step": 11713 + }, + { + "epoch": 0.8128512941503019, + "grad_norm": 1.6321150064468384, + "learning_rate": 0.00018467434154994475, + "loss": 0.6977, + "step": 11714 + }, + { + "epoch": 0.8129206855873985, + "grad_norm": 2.303391456604004, + "learning_rate": 0.00018467135230447576, + "loss": 0.6993, + "step": 11715 + }, + { + "epoch": 0.8129900770244952, + "grad_norm": 1.8595097064971924, + "learning_rate": 0.0001846683627917091, + "loss": 0.6191, + "step": 11716 + }, + { + "epoch": 0.8130594684615918, + "grad_norm": 2.434324264526367, + "learning_rate": 0.0001846653730116543, + "loss": 0.7933, + "step": 11717 + }, + { + "epoch": 0.8131288598986885, + "grad_norm": 2.324465036392212, + "learning_rate": 0.00018466238296432067, + "loss": 0.7217, + "step": 11718 + }, + { + "epoch": 0.8131982513357852, + "grad_norm": 2.348313808441162, + "learning_rate": 0.00018465939264971777, + "loss": 0.9496, + "step": 11719 + }, + { + "epoch": 0.8132676427728818, + "grad_norm": 2.526826858520508, + "learning_rate": 0.00018465640206785495, + "loss": 0.7996, + "step": 11720 + }, + { + "epoch": 0.8133370342099785, + "grad_norm": 2.1049346923828125, + "learning_rate": 0.0001846534112187417, + "loss": 0.6101, + "step": 11721 + }, + { + "epoch": 0.8134064256470751, + "grad_norm": 2.5691206455230713, + "learning_rate": 0.0001846504201023875, + "loss": 0.814, + "step": 11722 + }, + { + "epoch": 0.8134758170841718, + "grad_norm": 3.1533405780792236, + "learning_rate": 0.00018464742871880167, + "loss": 0.9678, + "step": 11723 + }, + { + "epoch": 0.8135452085212684, + "grad_norm": 2.1561386585235596, + "learning_rate": 0.00018464443706799378, + "loss": 0.6875, + "step": 11724 + }, + { + "epoch": 0.8136145999583652, + "grad_norm": 2.5957093238830566, + "learning_rate": 0.0001846414451499732, + "loss": 0.777, + "step": 11725 + }, + { + "epoch": 0.8136839913954618, + "grad_norm": 2.6630966663360596, + "learning_rate": 0.0001846384529647494, + "loss": 0.911, + "step": 11726 + }, + { + "epoch": 0.8137533828325585, + "grad_norm": 2.2867321968078613, + "learning_rate": 0.00018463546051233185, + "loss": 0.7309, + "step": 11727 + }, + { + "epoch": 0.8138227742696551, + "grad_norm": 2.3164165019989014, + "learning_rate": 0.00018463246779272993, + "loss": 0.9152, + "step": 11728 + }, + { + "epoch": 0.8138921657067518, + "grad_norm": 2.959470510482788, + "learning_rate": 0.00018462947480595316, + "loss": 0.8839, + "step": 11729 + }, + { + "epoch": 0.8139615571438484, + "grad_norm": 1.7773796319961548, + "learning_rate": 0.00018462648155201093, + "loss": 0.72, + "step": 11730 + }, + { + "epoch": 0.8140309485809452, + "grad_norm": 2.5473859310150146, + "learning_rate": 0.00018462348803091277, + "loss": 0.7668, + "step": 11731 + }, + { + "epoch": 0.8141003400180418, + "grad_norm": 2.012646198272705, + "learning_rate": 0.00018462049424266804, + "loss": 0.7849, + "step": 11732 + }, + { + "epoch": 0.8141697314551385, + "grad_norm": 1.8446028232574463, + "learning_rate": 0.00018461750018728621, + "loss": 0.7769, + "step": 11733 + }, + { + "epoch": 0.8142391228922351, + "grad_norm": 2.061318874359131, + "learning_rate": 0.00018461450586477676, + "loss": 0.9757, + "step": 11734 + }, + { + "epoch": 0.8143085143293317, + "grad_norm": 2.084820508956909, + "learning_rate": 0.00018461151127514914, + "loss": 0.8632, + "step": 11735 + }, + { + "epoch": 0.8143779057664284, + "grad_norm": 2.1884403228759766, + "learning_rate": 0.0001846085164184128, + "loss": 0.8373, + "step": 11736 + }, + { + "epoch": 0.814447297203525, + "grad_norm": 2.1513187885284424, + "learning_rate": 0.0001846055212945772, + "loss": 0.9887, + "step": 11737 + }, + { + "epoch": 0.8145166886406218, + "grad_norm": 2.5489680767059326, + "learning_rate": 0.00018460252590365175, + "loss": 0.8416, + "step": 11738 + }, + { + "epoch": 0.8145860800777184, + "grad_norm": 1.7500112056732178, + "learning_rate": 0.00018459953024564597, + "loss": 0.6198, + "step": 11739 + }, + { + "epoch": 0.8146554715148151, + "grad_norm": 2.7431509494781494, + "learning_rate": 0.0001845965343205693, + "loss": 0.798, + "step": 11740 + }, + { + "epoch": 0.8147248629519117, + "grad_norm": 1.997571587562561, + "learning_rate": 0.00018459353812843115, + "loss": 1.0289, + "step": 11741 + }, + { + "epoch": 0.8147942543890084, + "grad_norm": 1.8270641565322876, + "learning_rate": 0.000184590541669241, + "loss": 0.715, + "step": 11742 + }, + { + "epoch": 0.814863645826105, + "grad_norm": 2.086542844772339, + "learning_rate": 0.0001845875449430084, + "loss": 0.8292, + "step": 11743 + }, + { + "epoch": 0.8149330372632018, + "grad_norm": 2.1490187644958496, + "learning_rate": 0.00018458454794974266, + "loss": 0.7068, + "step": 11744 + }, + { + "epoch": 0.8150024287002984, + "grad_norm": 2.3011598587036133, + "learning_rate": 0.00018458155068945336, + "loss": 0.8995, + "step": 11745 + }, + { + "epoch": 0.8150718201373951, + "grad_norm": 2.319904327392578, + "learning_rate": 0.00018457855316214987, + "loss": 0.8527, + "step": 11746 + }, + { + "epoch": 0.8151412115744917, + "grad_norm": 1.7845362424850464, + "learning_rate": 0.00018457555536784175, + "loss": 0.7701, + "step": 11747 + }, + { + "epoch": 0.8152106030115883, + "grad_norm": 2.0113003253936768, + "learning_rate": 0.0001845725573065384, + "loss": 0.7761, + "step": 11748 + }, + { + "epoch": 0.815279994448685, + "grad_norm": 2.2025632858276367, + "learning_rate": 0.0001845695589782493, + "loss": 1.1817, + "step": 11749 + }, + { + "epoch": 0.8153493858857817, + "grad_norm": 2.0308914184570312, + "learning_rate": 0.0001845665603829839, + "loss": 0.7251, + "step": 11750 + }, + { + "epoch": 0.8154187773228784, + "grad_norm": 2.663541078567505, + "learning_rate": 0.0001845635615207517, + "loss": 0.8197, + "step": 11751 + }, + { + "epoch": 0.815488168759975, + "grad_norm": 1.920727014541626, + "learning_rate": 0.00018456056239156215, + "loss": 0.7196, + "step": 11752 + }, + { + "epoch": 0.8155575601970717, + "grad_norm": 2.894388198852539, + "learning_rate": 0.00018455756299542468, + "loss": 0.9083, + "step": 11753 + }, + { + "epoch": 0.8156269516341683, + "grad_norm": 2.3334577083587646, + "learning_rate": 0.00018455456333234884, + "loss": 0.8888, + "step": 11754 + }, + { + "epoch": 0.815696343071265, + "grad_norm": 1.9895906448364258, + "learning_rate": 0.000184551563402344, + "loss": 0.9055, + "step": 11755 + }, + { + "epoch": 0.8157657345083617, + "grad_norm": 2.563593864440918, + "learning_rate": 0.00018454856320541973, + "loss": 1.0182, + "step": 11756 + }, + { + "epoch": 0.8158351259454584, + "grad_norm": 1.9252504110336304, + "learning_rate": 0.00018454556274158544, + "loss": 0.5705, + "step": 11757 + }, + { + "epoch": 0.815904517382555, + "grad_norm": 2.998527765274048, + "learning_rate": 0.00018454256201085061, + "loss": 0.9902, + "step": 11758 + }, + { + "epoch": 0.8159739088196517, + "grad_norm": 1.9430252313613892, + "learning_rate": 0.00018453956101322473, + "loss": 0.6248, + "step": 11759 + }, + { + "epoch": 0.8160433002567483, + "grad_norm": 2.6627352237701416, + "learning_rate": 0.00018453655974871728, + "loss": 0.7827, + "step": 11760 + }, + { + "epoch": 0.8161126916938449, + "grad_norm": 3.4157121181488037, + "learning_rate": 0.00018453355821733771, + "loss": 0.7895, + "step": 11761 + }, + { + "epoch": 0.8161820831309416, + "grad_norm": 2.285862684249878, + "learning_rate": 0.00018453055641909547, + "loss": 0.7023, + "step": 11762 + }, + { + "epoch": 0.8162514745680383, + "grad_norm": 2.211914539337158, + "learning_rate": 0.0001845275543540001, + "loss": 0.8219, + "step": 11763 + }, + { + "epoch": 0.816320866005135, + "grad_norm": 2.1383063793182373, + "learning_rate": 0.00018452455202206105, + "loss": 0.8919, + "step": 11764 + }, + { + "epoch": 0.8163902574422316, + "grad_norm": 2.0924792289733887, + "learning_rate": 0.00018452154942328778, + "loss": 0.7801, + "step": 11765 + }, + { + "epoch": 0.8164596488793283, + "grad_norm": 2.2231099605560303, + "learning_rate": 0.0001845185465576898, + "loss": 0.7621, + "step": 11766 + }, + { + "epoch": 0.8165290403164249, + "grad_norm": 1.6586031913757324, + "learning_rate": 0.0001845155434252766, + "loss": 0.768, + "step": 11767 + }, + { + "epoch": 0.8165984317535216, + "grad_norm": 2.9511160850524902, + "learning_rate": 0.00018451254002605757, + "loss": 1.2324, + "step": 11768 + }, + { + "epoch": 0.8166678231906183, + "grad_norm": 2.1276798248291016, + "learning_rate": 0.00018450953636004228, + "loss": 0.8982, + "step": 11769 + }, + { + "epoch": 0.816737214627715, + "grad_norm": 2.643498659133911, + "learning_rate": 0.00018450653242724023, + "loss": 0.7899, + "step": 11770 + }, + { + "epoch": 0.8168066060648116, + "grad_norm": 1.8660115003585815, + "learning_rate": 0.0001845035282276608, + "loss": 0.8023, + "step": 11771 + }, + { + "epoch": 0.8168759975019083, + "grad_norm": 1.8777376413345337, + "learning_rate": 0.0001845005237613136, + "loss": 0.8403, + "step": 11772 + }, + { + "epoch": 0.8169453889390049, + "grad_norm": 2.202120304107666, + "learning_rate": 0.00018449751902820802, + "loss": 0.9174, + "step": 11773 + }, + { + "epoch": 0.8170147803761015, + "grad_norm": 1.9140007495880127, + "learning_rate": 0.00018449451402835357, + "loss": 0.5885, + "step": 11774 + }, + { + "epoch": 0.8170841718131983, + "grad_norm": 1.8519452810287476, + "learning_rate": 0.00018449150876175975, + "loss": 0.6848, + "step": 11775 + }, + { + "epoch": 0.8171535632502949, + "grad_norm": 2.2848026752471924, + "learning_rate": 0.00018448850322843604, + "loss": 0.6987, + "step": 11776 + }, + { + "epoch": 0.8172229546873916, + "grad_norm": 2.0533077716827393, + "learning_rate": 0.00018448549742839193, + "loss": 0.7911, + "step": 11777 + }, + { + "epoch": 0.8172923461244882, + "grad_norm": 3.0361883640289307, + "learning_rate": 0.0001844824913616369, + "loss": 0.9491, + "step": 11778 + }, + { + "epoch": 0.8173617375615849, + "grad_norm": 1.6788097620010376, + "learning_rate": 0.00018447948502818044, + "loss": 0.6085, + "step": 11779 + }, + { + "epoch": 0.8174311289986815, + "grad_norm": 2.892996072769165, + "learning_rate": 0.00018447647842803204, + "loss": 0.7951, + "step": 11780 + }, + { + "epoch": 0.8175005204357783, + "grad_norm": 2.516084671020508, + "learning_rate": 0.00018447347156120124, + "loss": 0.7025, + "step": 11781 + }, + { + "epoch": 0.8175699118728749, + "grad_norm": 2.2569687366485596, + "learning_rate": 0.00018447046442769745, + "loss": 0.6696, + "step": 11782 + }, + { + "epoch": 0.8176393033099716, + "grad_norm": 2.897984743118286, + "learning_rate": 0.00018446745702753022, + "loss": 0.9647, + "step": 11783 + }, + { + "epoch": 0.8177086947470682, + "grad_norm": 2.1096043586730957, + "learning_rate": 0.000184464449360709, + "loss": 0.8375, + "step": 11784 + }, + { + "epoch": 0.8177780861841649, + "grad_norm": 2.2897889614105225, + "learning_rate": 0.00018446144142724339, + "loss": 0.609, + "step": 11785 + }, + { + "epoch": 0.8178474776212615, + "grad_norm": 1.7285243272781372, + "learning_rate": 0.00018445843322714272, + "loss": 0.6938, + "step": 11786 + }, + { + "epoch": 0.8179168690583583, + "grad_norm": 2.29960560798645, + "learning_rate": 0.00018445542476041663, + "loss": 0.8653, + "step": 11787 + }, + { + "epoch": 0.8179862604954549, + "grad_norm": 2.5160417556762695, + "learning_rate": 0.0001844524160270745, + "loss": 1.1281, + "step": 11788 + }, + { + "epoch": 0.8180556519325515, + "grad_norm": 2.23310923576355, + "learning_rate": 0.00018444940702712597, + "loss": 0.8346, + "step": 11789 + }, + { + "epoch": 0.8181250433696482, + "grad_norm": 2.0122692584991455, + "learning_rate": 0.00018444639776058042, + "loss": 0.7348, + "step": 11790 + }, + { + "epoch": 0.8181944348067448, + "grad_norm": 1.6037218570709229, + "learning_rate": 0.0001844433882274474, + "loss": 0.6531, + "step": 11791 + }, + { + "epoch": 0.8182638262438415, + "grad_norm": 1.8264963626861572, + "learning_rate": 0.0001844403784277364, + "loss": 0.5859, + "step": 11792 + }, + { + "epoch": 0.8183332176809381, + "grad_norm": 1.8164538145065308, + "learning_rate": 0.00018443736836145692, + "loss": 0.9013, + "step": 11793 + }, + { + "epoch": 0.8184026091180349, + "grad_norm": 2.594989061355591, + "learning_rate": 0.00018443435802861847, + "loss": 0.8156, + "step": 11794 + }, + { + "epoch": 0.8184720005551315, + "grad_norm": 1.9208334684371948, + "learning_rate": 0.00018443134742923055, + "loss": 0.8604, + "step": 11795 + }, + { + "epoch": 0.8185413919922282, + "grad_norm": 1.8984755277633667, + "learning_rate": 0.00018442833656330266, + "loss": 0.7783, + "step": 11796 + }, + { + "epoch": 0.8186107834293248, + "grad_norm": 2.3545989990234375, + "learning_rate": 0.0001844253254308443, + "loss": 0.9487, + "step": 11797 + }, + { + "epoch": 0.8186801748664215, + "grad_norm": 2.542553663253784, + "learning_rate": 0.00018442231403186497, + "loss": 0.9537, + "step": 11798 + }, + { + "epoch": 0.8187495663035181, + "grad_norm": 2.216437339782715, + "learning_rate": 0.0001844193023663742, + "loss": 0.7661, + "step": 11799 + }, + { + "epoch": 0.8188189577406149, + "grad_norm": 1.8099184036254883, + "learning_rate": 0.00018441629043438148, + "loss": 0.8029, + "step": 11800 + }, + { + "epoch": 0.8188883491777115, + "grad_norm": 2.02712345123291, + "learning_rate": 0.00018441327823589637, + "loss": 0.9051, + "step": 11801 + }, + { + "epoch": 0.8189577406148081, + "grad_norm": 2.192504405975342, + "learning_rate": 0.00018441026577092827, + "loss": 0.9229, + "step": 11802 + }, + { + "epoch": 0.8190271320519048, + "grad_norm": 2.3983895778656006, + "learning_rate": 0.00018440725303948678, + "loss": 0.7891, + "step": 11803 + }, + { + "epoch": 0.8190965234890014, + "grad_norm": 2.6580309867858887, + "learning_rate": 0.00018440424004158138, + "loss": 1.0402, + "step": 11804 + }, + { + "epoch": 0.8191659149260981, + "grad_norm": 3.2309865951538086, + "learning_rate": 0.00018440122677722162, + "loss": 0.6494, + "step": 11805 + }, + { + "epoch": 0.8192353063631947, + "grad_norm": 2.0275144577026367, + "learning_rate": 0.00018439821324641695, + "loss": 0.8439, + "step": 11806 + }, + { + "epoch": 0.8193046978002915, + "grad_norm": 2.350790500640869, + "learning_rate": 0.0001843951994491769, + "loss": 0.856, + "step": 11807 + }, + { + "epoch": 0.8193740892373881, + "grad_norm": 1.741292119026184, + "learning_rate": 0.00018439218538551105, + "loss": 0.6493, + "step": 11808 + }, + { + "epoch": 0.8194434806744848, + "grad_norm": 2.1010801792144775, + "learning_rate": 0.0001843891710554288, + "loss": 0.8429, + "step": 11809 + }, + { + "epoch": 0.8195128721115814, + "grad_norm": 2.4723968505859375, + "learning_rate": 0.00018438615645893978, + "loss": 0.8971, + "step": 11810 + }, + { + "epoch": 0.8195822635486781, + "grad_norm": 1.8990670442581177, + "learning_rate": 0.00018438314159605344, + "loss": 0.6564, + "step": 11811 + }, + { + "epoch": 0.8196516549857747, + "grad_norm": 2.2355523109436035, + "learning_rate": 0.00018438012646677926, + "loss": 0.7543, + "step": 11812 + }, + { + "epoch": 0.8197210464228715, + "grad_norm": 1.9614818096160889, + "learning_rate": 0.00018437711107112688, + "loss": 0.9176, + "step": 11813 + }, + { + "epoch": 0.8197904378599681, + "grad_norm": 1.752223253250122, + "learning_rate": 0.00018437409540910573, + "loss": 0.6359, + "step": 11814 + }, + { + "epoch": 0.8198598292970647, + "grad_norm": 2.662476062774658, + "learning_rate": 0.00018437107948072532, + "loss": 1.023, + "step": 11815 + }, + { + "epoch": 0.8199292207341614, + "grad_norm": 2.331533193588257, + "learning_rate": 0.00018436806328599525, + "loss": 0.7119, + "step": 11816 + }, + { + "epoch": 0.819998612171258, + "grad_norm": 2.126375675201416, + "learning_rate": 0.00018436504682492496, + "loss": 0.7585, + "step": 11817 + }, + { + "epoch": 0.8200680036083547, + "grad_norm": 2.790458917617798, + "learning_rate": 0.000184362030097524, + "loss": 0.8189, + "step": 11818 + }, + { + "epoch": 0.8201373950454514, + "grad_norm": 2.226752281188965, + "learning_rate": 0.0001843590131038019, + "loss": 0.9786, + "step": 11819 + }, + { + "epoch": 0.8202067864825481, + "grad_norm": 1.886589765548706, + "learning_rate": 0.0001843559958437682, + "loss": 0.8473, + "step": 11820 + }, + { + "epoch": 0.8202761779196447, + "grad_norm": 2.276242256164551, + "learning_rate": 0.00018435297831743242, + "loss": 0.9879, + "step": 11821 + }, + { + "epoch": 0.8203455693567414, + "grad_norm": 2.620820999145508, + "learning_rate": 0.00018434996052480406, + "loss": 0.8494, + "step": 11822 + }, + { + "epoch": 0.820414960793838, + "grad_norm": 2.957136869430542, + "learning_rate": 0.00018434694246589265, + "loss": 0.8989, + "step": 11823 + }, + { + "epoch": 0.8204843522309347, + "grad_norm": 2.2363717555999756, + "learning_rate": 0.0001843439241407077, + "loss": 0.9567, + "step": 11824 + }, + { + "epoch": 0.8205537436680314, + "grad_norm": 2.267451524734497, + "learning_rate": 0.00018434090554925883, + "loss": 0.771, + "step": 11825 + }, + { + "epoch": 0.8206231351051281, + "grad_norm": 1.8978272676467896, + "learning_rate": 0.0001843378866915555, + "loss": 0.7303, + "step": 11826 + }, + { + "epoch": 0.8206925265422247, + "grad_norm": 2.103317975997925, + "learning_rate": 0.00018433486756760722, + "loss": 0.7798, + "step": 11827 + }, + { + "epoch": 0.8207619179793213, + "grad_norm": 2.8794848918914795, + "learning_rate": 0.00018433184817742354, + "loss": 0.8624, + "step": 11828 + }, + { + "epoch": 0.820831309416418, + "grad_norm": 2.081522226333618, + "learning_rate": 0.00018432882852101402, + "loss": 0.8296, + "step": 11829 + }, + { + "epoch": 0.8209007008535146, + "grad_norm": 2.9741461277008057, + "learning_rate": 0.00018432580859838814, + "loss": 0.8416, + "step": 11830 + }, + { + "epoch": 0.8209700922906114, + "grad_norm": 2.501002788543701, + "learning_rate": 0.00018432278840955552, + "loss": 0.8772, + "step": 11831 + }, + { + "epoch": 0.821039483727708, + "grad_norm": 2.435356855392456, + "learning_rate": 0.0001843197679545256, + "loss": 0.8097, + "step": 11832 + }, + { + "epoch": 0.8211088751648047, + "grad_norm": 1.9356077909469604, + "learning_rate": 0.00018431674723330796, + "loss": 0.816, + "step": 11833 + }, + { + "epoch": 0.8211782666019013, + "grad_norm": 2.0778186321258545, + "learning_rate": 0.00018431372624591217, + "loss": 0.6277, + "step": 11834 + }, + { + "epoch": 0.821247658038998, + "grad_norm": 2.0635533332824707, + "learning_rate": 0.00018431070499234768, + "loss": 0.7358, + "step": 11835 + }, + { + "epoch": 0.8213170494760946, + "grad_norm": 2.047649621963501, + "learning_rate": 0.0001843076834726241, + "loss": 0.5773, + "step": 11836 + }, + { + "epoch": 0.8213864409131914, + "grad_norm": 2.3304572105407715, + "learning_rate": 0.00018430466168675093, + "loss": 0.8054, + "step": 11837 + }, + { + "epoch": 0.821455832350288, + "grad_norm": 1.8956141471862793, + "learning_rate": 0.00018430163963473772, + "loss": 0.7469, + "step": 11838 + }, + { + "epoch": 0.8215252237873847, + "grad_norm": 2.3856585025787354, + "learning_rate": 0.00018429861731659406, + "loss": 0.8431, + "step": 11839 + }, + { + "epoch": 0.8215946152244813, + "grad_norm": 2.2522456645965576, + "learning_rate": 0.0001842955947323294, + "loss": 0.8411, + "step": 11840 + }, + { + "epoch": 0.8216640066615779, + "grad_norm": 1.8236018419265747, + "learning_rate": 0.00018429257188195335, + "loss": 0.8446, + "step": 11841 + }, + { + "epoch": 0.8217333980986746, + "grad_norm": 2.6670315265655518, + "learning_rate": 0.0001842895487654754, + "loss": 0.865, + "step": 11842 + }, + { + "epoch": 0.8218027895357712, + "grad_norm": 2.1346843242645264, + "learning_rate": 0.00018428652538290513, + "loss": 0.8407, + "step": 11843 + }, + { + "epoch": 0.821872180972868, + "grad_norm": 2.0610687732696533, + "learning_rate": 0.00018428350173425214, + "loss": 0.6693, + "step": 11844 + }, + { + "epoch": 0.8219415724099646, + "grad_norm": 2.210045099258423, + "learning_rate": 0.00018428047781952585, + "loss": 0.8952, + "step": 11845 + }, + { + "epoch": 0.8220109638470613, + "grad_norm": 2.188858985900879, + "learning_rate": 0.0001842774536387359, + "loss": 0.9156, + "step": 11846 + }, + { + "epoch": 0.8220803552841579, + "grad_norm": 2.2568304538726807, + "learning_rate": 0.00018427442919189178, + "loss": 1.0156, + "step": 11847 + }, + { + "epoch": 0.8221497467212546, + "grad_norm": 1.9606201648712158, + "learning_rate": 0.0001842714044790031, + "loss": 0.891, + "step": 11848 + }, + { + "epoch": 0.8222191381583512, + "grad_norm": 2.3546011447906494, + "learning_rate": 0.00018426837950007933, + "loss": 0.8596, + "step": 11849 + }, + { + "epoch": 0.822288529595448, + "grad_norm": 2.3341217041015625, + "learning_rate": 0.00018426535425513008, + "loss": 0.8101, + "step": 11850 + }, + { + "epoch": 0.8223579210325446, + "grad_norm": 2.017794370651245, + "learning_rate": 0.0001842623287441649, + "loss": 0.7324, + "step": 11851 + }, + { + "epoch": 0.8224273124696413, + "grad_norm": 2.253915309906006, + "learning_rate": 0.00018425930296719328, + "loss": 0.7997, + "step": 11852 + }, + { + "epoch": 0.8224967039067379, + "grad_norm": 2.1441807746887207, + "learning_rate": 0.00018425627692422486, + "loss": 0.6487, + "step": 11853 + }, + { + "epoch": 0.8225660953438346, + "grad_norm": 2.634632110595703, + "learning_rate": 0.00018425325061526915, + "loss": 0.9813, + "step": 11854 + }, + { + "epoch": 0.8226354867809312, + "grad_norm": 2.0401179790496826, + "learning_rate": 0.0001842502240403357, + "loss": 0.5972, + "step": 11855 + }, + { + "epoch": 0.8227048782180278, + "grad_norm": 2.2059249877929688, + "learning_rate": 0.00018424719719943404, + "loss": 0.8038, + "step": 11856 + }, + { + "epoch": 0.8227742696551246, + "grad_norm": 1.9837658405303955, + "learning_rate": 0.00018424417009257378, + "loss": 0.7456, + "step": 11857 + }, + { + "epoch": 0.8228436610922212, + "grad_norm": 2.0550363063812256, + "learning_rate": 0.00018424114271976442, + "loss": 0.6576, + "step": 11858 + }, + { + "epoch": 0.8229130525293179, + "grad_norm": 2.3823702335357666, + "learning_rate": 0.00018423811508101558, + "loss": 1.0311, + "step": 11859 + }, + { + "epoch": 0.8229824439664145, + "grad_norm": 1.9642168283462524, + "learning_rate": 0.00018423508717633677, + "loss": 0.7502, + "step": 11860 + }, + { + "epoch": 0.8230518354035112, + "grad_norm": 2.7718048095703125, + "learning_rate": 0.00018423205900573757, + "loss": 0.9556, + "step": 11861 + }, + { + "epoch": 0.8231212268406078, + "grad_norm": 3.597346305847168, + "learning_rate": 0.00018422903056922754, + "loss": 0.8908, + "step": 11862 + }, + { + "epoch": 0.8231906182777046, + "grad_norm": 2.7660250663757324, + "learning_rate": 0.0001842260018668162, + "loss": 1.0123, + "step": 11863 + }, + { + "epoch": 0.8232600097148012, + "grad_norm": 2.8259944915771484, + "learning_rate": 0.00018422297289851316, + "loss": 0.9981, + "step": 11864 + }, + { + "epoch": 0.8233294011518979, + "grad_norm": 2.607236385345459, + "learning_rate": 0.00018421994366432797, + "loss": 0.6758, + "step": 11865 + }, + { + "epoch": 0.8233987925889945, + "grad_norm": 1.7649955749511719, + "learning_rate": 0.00018421691416427017, + "loss": 0.7431, + "step": 11866 + }, + { + "epoch": 0.8234681840260912, + "grad_norm": 1.922518014907837, + "learning_rate": 0.00018421388439834936, + "loss": 0.8696, + "step": 11867 + }, + { + "epoch": 0.8235375754631878, + "grad_norm": 1.572533369064331, + "learning_rate": 0.00018421085436657513, + "loss": 0.7454, + "step": 11868 + }, + { + "epoch": 0.8236069669002845, + "grad_norm": 1.8332736492156982, + "learning_rate": 0.00018420782406895693, + "loss": 0.8177, + "step": 11869 + }, + { + "epoch": 0.8236763583373812, + "grad_norm": 2.4151663780212402, + "learning_rate": 0.00018420479350550445, + "loss": 0.7939, + "step": 11870 + }, + { + "epoch": 0.8237457497744778, + "grad_norm": 3.26690673828125, + "learning_rate": 0.00018420176267622718, + "loss": 0.6547, + "step": 11871 + }, + { + "epoch": 0.8238151412115745, + "grad_norm": 1.8341618776321411, + "learning_rate": 0.00018419873158113473, + "loss": 0.8872, + "step": 11872 + }, + { + "epoch": 0.8238845326486711, + "grad_norm": 2.546248435974121, + "learning_rate": 0.00018419570022023665, + "loss": 0.8846, + "step": 11873 + }, + { + "epoch": 0.8239539240857678, + "grad_norm": 1.4688966274261475, + "learning_rate": 0.0001841926685935425, + "loss": 0.7208, + "step": 11874 + }, + { + "epoch": 0.8240233155228645, + "grad_norm": 2.454183578491211, + "learning_rate": 0.00018418963670106185, + "loss": 0.9748, + "step": 11875 + }, + { + "epoch": 0.8240927069599612, + "grad_norm": 1.7219449281692505, + "learning_rate": 0.00018418660454280433, + "loss": 0.5664, + "step": 11876 + }, + { + "epoch": 0.8241620983970578, + "grad_norm": 2.346576690673828, + "learning_rate": 0.00018418357211877944, + "loss": 0.772, + "step": 11877 + }, + { + "epoch": 0.8242314898341545, + "grad_norm": 2.045138120651245, + "learning_rate": 0.00018418053942899677, + "loss": 0.7374, + "step": 11878 + }, + { + "epoch": 0.8243008812712511, + "grad_norm": 2.2940473556518555, + "learning_rate": 0.00018417750647346592, + "loss": 0.8135, + "step": 11879 + }, + { + "epoch": 0.8243702727083478, + "grad_norm": 1.767337679862976, + "learning_rate": 0.00018417447325219647, + "loss": 0.7292, + "step": 11880 + }, + { + "epoch": 0.8244396641454445, + "grad_norm": 1.7985420227050781, + "learning_rate": 0.00018417143976519795, + "loss": 0.5952, + "step": 11881 + }, + { + "epoch": 0.8245090555825411, + "grad_norm": 2.344364881515503, + "learning_rate": 0.00018416840601247995, + "loss": 0.8825, + "step": 11882 + }, + { + "epoch": 0.8245784470196378, + "grad_norm": 1.9162769317626953, + "learning_rate": 0.00018416537199405207, + "loss": 0.6192, + "step": 11883 + }, + { + "epoch": 0.8246478384567344, + "grad_norm": 2.125347852706909, + "learning_rate": 0.00018416233770992386, + "loss": 0.8885, + "step": 11884 + }, + { + "epoch": 0.8247172298938311, + "grad_norm": 1.6512982845306396, + "learning_rate": 0.00018415930316010494, + "loss": 0.5978, + "step": 11885 + }, + { + "epoch": 0.8247866213309277, + "grad_norm": 1.8485673666000366, + "learning_rate": 0.00018415626834460485, + "loss": 0.7138, + "step": 11886 + }, + { + "epoch": 0.8248560127680244, + "grad_norm": 2.328648328781128, + "learning_rate": 0.00018415323326343317, + "loss": 0.9747, + "step": 11887 + }, + { + "epoch": 0.8249254042051211, + "grad_norm": 2.067056179046631, + "learning_rate": 0.0001841501979165995, + "loss": 0.9976, + "step": 11888 + }, + { + "epoch": 0.8249947956422178, + "grad_norm": 2.308892250061035, + "learning_rate": 0.00018414716230411346, + "loss": 0.7672, + "step": 11889 + }, + { + "epoch": 0.8250641870793144, + "grad_norm": 2.1653318405151367, + "learning_rate": 0.00018414412642598458, + "loss": 0.8981, + "step": 11890 + }, + { + "epoch": 0.8251335785164111, + "grad_norm": 1.8089474439620972, + "learning_rate": 0.00018414109028222242, + "loss": 0.6424, + "step": 11891 + }, + { + "epoch": 0.8252029699535077, + "grad_norm": 1.7806373834609985, + "learning_rate": 0.0001841380538728366, + "loss": 0.5958, + "step": 11892 + }, + { + "epoch": 0.8252723613906044, + "grad_norm": 1.9038575887680054, + "learning_rate": 0.00018413501719783675, + "loss": 0.8058, + "step": 11893 + }, + { + "epoch": 0.8253417528277011, + "grad_norm": 1.939124584197998, + "learning_rate": 0.00018413198025723238, + "loss": 0.7198, + "step": 11894 + }, + { + "epoch": 0.8254111442647977, + "grad_norm": 1.434388518333435, + "learning_rate": 0.0001841289430510331, + "loss": 0.506, + "step": 11895 + }, + { + "epoch": 0.8254805357018944, + "grad_norm": 2.198162317276001, + "learning_rate": 0.00018412590557924854, + "loss": 0.6285, + "step": 11896 + }, + { + "epoch": 0.825549927138991, + "grad_norm": 1.9610363245010376, + "learning_rate": 0.00018412286784188825, + "loss": 0.8093, + "step": 11897 + }, + { + "epoch": 0.8256193185760877, + "grad_norm": 2.4469974040985107, + "learning_rate": 0.0001841198298389618, + "loss": 0.7471, + "step": 11898 + }, + { + "epoch": 0.8256887100131843, + "grad_norm": 1.9013906717300415, + "learning_rate": 0.0001841167915704788, + "loss": 0.8655, + "step": 11899 + }, + { + "epoch": 0.8257581014502811, + "grad_norm": 1.8977265357971191, + "learning_rate": 0.0001841137530364489, + "loss": 0.6377, + "step": 11900 + }, + { + "epoch": 0.8258274928873777, + "grad_norm": 2.2507457733154297, + "learning_rate": 0.0001841107142368816, + "loss": 0.7639, + "step": 11901 + }, + { + "epoch": 0.8258968843244744, + "grad_norm": 2.227088212966919, + "learning_rate": 0.00018410767517178657, + "loss": 0.808, + "step": 11902 + }, + { + "epoch": 0.825966275761571, + "grad_norm": 2.066431999206543, + "learning_rate": 0.00018410463584117332, + "loss": 0.8155, + "step": 11903 + }, + { + "epoch": 0.8260356671986677, + "grad_norm": 2.611868381500244, + "learning_rate": 0.00018410159624505153, + "loss": 0.7813, + "step": 11904 + }, + { + "epoch": 0.8261050586357643, + "grad_norm": 2.173863649368286, + "learning_rate": 0.00018409855638343073, + "loss": 0.7497, + "step": 11905 + }, + { + "epoch": 0.826174450072861, + "grad_norm": 2.0610949993133545, + "learning_rate": 0.00018409551625632058, + "loss": 0.8938, + "step": 11906 + }, + { + "epoch": 0.8262438415099577, + "grad_norm": 2.1017093658447266, + "learning_rate": 0.00018409247586373062, + "loss": 0.7105, + "step": 11907 + }, + { + "epoch": 0.8263132329470544, + "grad_norm": 2.3581528663635254, + "learning_rate": 0.00018408943520567046, + "loss": 1.2701, + "step": 11908 + }, + { + "epoch": 0.826382624384151, + "grad_norm": 2.3110501766204834, + "learning_rate": 0.0001840863942821497, + "loss": 0.7412, + "step": 11909 + }, + { + "epoch": 0.8264520158212476, + "grad_norm": 2.618100881576538, + "learning_rate": 0.000184083353093178, + "loss": 0.9573, + "step": 11910 + }, + { + "epoch": 0.8265214072583443, + "grad_norm": 2.3531246185302734, + "learning_rate": 0.00018408031163876486, + "loss": 0.9195, + "step": 11911 + }, + { + "epoch": 0.8265907986954409, + "grad_norm": 3.070868730545044, + "learning_rate": 0.00018407726991891994, + "loss": 0.8765, + "step": 11912 + }, + { + "epoch": 0.8266601901325377, + "grad_norm": 1.8532403707504272, + "learning_rate": 0.00018407422793365285, + "loss": 0.8126, + "step": 11913 + }, + { + "epoch": 0.8267295815696343, + "grad_norm": 2.926719903945923, + "learning_rate": 0.00018407118568297318, + "loss": 1.1219, + "step": 11914 + }, + { + "epoch": 0.826798973006731, + "grad_norm": 2.1822943687438965, + "learning_rate": 0.00018406814316689052, + "loss": 0.9658, + "step": 11915 + }, + { + "epoch": 0.8268683644438276, + "grad_norm": 2.292617082595825, + "learning_rate": 0.0001840651003854145, + "loss": 0.7951, + "step": 11916 + }, + { + "epoch": 0.8269377558809243, + "grad_norm": 1.821008563041687, + "learning_rate": 0.0001840620573385547, + "loss": 0.5968, + "step": 11917 + }, + { + "epoch": 0.8270071473180209, + "grad_norm": 2.920318603515625, + "learning_rate": 0.00018405901402632072, + "loss": 1.0903, + "step": 11918 + }, + { + "epoch": 0.8270765387551177, + "grad_norm": 2.3596174716949463, + "learning_rate": 0.0001840559704487222, + "loss": 0.8126, + "step": 11919 + }, + { + "epoch": 0.8271459301922143, + "grad_norm": 2.8074216842651367, + "learning_rate": 0.00018405292660576876, + "loss": 1.0222, + "step": 11920 + }, + { + "epoch": 0.827215321629311, + "grad_norm": 1.6789228916168213, + "learning_rate": 0.00018404988249746994, + "loss": 0.6169, + "step": 11921 + }, + { + "epoch": 0.8272847130664076, + "grad_norm": 2.4615979194641113, + "learning_rate": 0.0001840468381238354, + "loss": 0.8467, + "step": 11922 + }, + { + "epoch": 0.8273541045035042, + "grad_norm": 2.229355573654175, + "learning_rate": 0.00018404379348487478, + "loss": 0.7371, + "step": 11923 + }, + { + "epoch": 0.8274234959406009, + "grad_norm": 2.214876413345337, + "learning_rate": 0.00018404074858059764, + "loss": 0.9597, + "step": 11924 + }, + { + "epoch": 0.8274928873776976, + "grad_norm": 2.1249043941497803, + "learning_rate": 0.00018403770341101359, + "loss": 0.5722, + "step": 11925 + }, + { + "epoch": 0.8275622788147943, + "grad_norm": 2.9217493534088135, + "learning_rate": 0.00018403465797613223, + "loss": 1.0498, + "step": 11926 + }, + { + "epoch": 0.8276316702518909, + "grad_norm": 1.7160325050354004, + "learning_rate": 0.00018403161227596327, + "loss": 0.5527, + "step": 11927 + }, + { + "epoch": 0.8277010616889876, + "grad_norm": 3.0285933017730713, + "learning_rate": 0.0001840285663105162, + "loss": 0.9353, + "step": 11928 + }, + { + "epoch": 0.8277704531260842, + "grad_norm": 2.294276237487793, + "learning_rate": 0.00018402552007980077, + "loss": 0.6828, + "step": 11929 + }, + { + "epoch": 0.8278398445631809, + "grad_norm": 2.509561061859131, + "learning_rate": 0.00018402247358382647, + "loss": 0.9868, + "step": 11930 + }, + { + "epoch": 0.8279092360002775, + "grad_norm": 1.7339965105056763, + "learning_rate": 0.00018401942682260295, + "loss": 0.5898, + "step": 11931 + }, + { + "epoch": 0.8279786274373743, + "grad_norm": 1.6927709579467773, + "learning_rate": 0.00018401637979613989, + "loss": 0.6935, + "step": 11932 + }, + { + "epoch": 0.8280480188744709, + "grad_norm": 2.1790332794189453, + "learning_rate": 0.00018401333250444684, + "loss": 0.7915, + "step": 11933 + }, + { + "epoch": 0.8281174103115676, + "grad_norm": 1.4707276821136475, + "learning_rate": 0.00018401028494753347, + "loss": 0.5985, + "step": 11934 + }, + { + "epoch": 0.8281868017486642, + "grad_norm": 2.3109045028686523, + "learning_rate": 0.00018400723712540937, + "loss": 0.8183, + "step": 11935 + }, + { + "epoch": 0.8282561931857608, + "grad_norm": 2.1823177337646484, + "learning_rate": 0.00018400418903808416, + "loss": 0.661, + "step": 11936 + }, + { + "epoch": 0.8283255846228575, + "grad_norm": 2.13037109375, + "learning_rate": 0.0001840011406855675, + "loss": 0.7327, + "step": 11937 + }, + { + "epoch": 0.8283949760599542, + "grad_norm": 2.460960865020752, + "learning_rate": 0.00018399809206786895, + "loss": 0.8202, + "step": 11938 + }, + { + "epoch": 0.8284643674970509, + "grad_norm": 2.5421316623687744, + "learning_rate": 0.00018399504318499817, + "loss": 1.1536, + "step": 11939 + }, + { + "epoch": 0.8285337589341475, + "grad_norm": 1.6120463609695435, + "learning_rate": 0.0001839919940369648, + "loss": 0.6845, + "step": 11940 + }, + { + "epoch": 0.8286031503712442, + "grad_norm": 1.9546228647232056, + "learning_rate": 0.00018398894462377845, + "loss": 0.753, + "step": 11941 + }, + { + "epoch": 0.8286725418083408, + "grad_norm": 2.285324811935425, + "learning_rate": 0.00018398589494544873, + "loss": 0.7478, + "step": 11942 + }, + { + "epoch": 0.8287419332454375, + "grad_norm": 1.8739303350448608, + "learning_rate": 0.00018398284500198528, + "loss": 0.65, + "step": 11943 + }, + { + "epoch": 0.8288113246825342, + "grad_norm": 2.1498196125030518, + "learning_rate": 0.00018397979479339776, + "loss": 0.9867, + "step": 11944 + }, + { + "epoch": 0.8288807161196309, + "grad_norm": 2.268582820892334, + "learning_rate": 0.00018397674431969575, + "loss": 0.7458, + "step": 11945 + }, + { + "epoch": 0.8289501075567275, + "grad_norm": 2.1639509201049805, + "learning_rate": 0.0001839736935808889, + "loss": 0.8233, + "step": 11946 + }, + { + "epoch": 0.8290194989938242, + "grad_norm": 2.2904908657073975, + "learning_rate": 0.0001839706425769869, + "loss": 0.7623, + "step": 11947 + }, + { + "epoch": 0.8290888904309208, + "grad_norm": 2.286991834640503, + "learning_rate": 0.00018396759130799922, + "loss": 0.9107, + "step": 11948 + }, + { + "epoch": 0.8291582818680174, + "grad_norm": 2.2722933292388916, + "learning_rate": 0.00018396453977393567, + "loss": 0.864, + "step": 11949 + }, + { + "epoch": 0.8292276733051142, + "grad_norm": 1.8228697776794434, + "learning_rate": 0.00018396148797480578, + "loss": 0.9035, + "step": 11950 + }, + { + "epoch": 0.8292970647422108, + "grad_norm": 3.289431095123291, + "learning_rate": 0.00018395843591061922, + "loss": 0.8495, + "step": 11951 + }, + { + "epoch": 0.8293664561793075, + "grad_norm": 2.221341609954834, + "learning_rate": 0.00018395538358138565, + "loss": 0.9038, + "step": 11952 + }, + { + "epoch": 0.8294358476164041, + "grad_norm": 2.2443432807922363, + "learning_rate": 0.00018395233098711465, + "loss": 0.8088, + "step": 11953 + }, + { + "epoch": 0.8295052390535008, + "grad_norm": 2.1707119941711426, + "learning_rate": 0.00018394927812781587, + "loss": 0.6851, + "step": 11954 + }, + { + "epoch": 0.8295746304905974, + "grad_norm": 2.0443365573883057, + "learning_rate": 0.00018394622500349895, + "loss": 0.6602, + "step": 11955 + }, + { + "epoch": 0.8296440219276942, + "grad_norm": 1.9570914506912231, + "learning_rate": 0.00018394317161417355, + "loss": 0.7804, + "step": 11956 + }, + { + "epoch": 0.8297134133647908, + "grad_norm": 2.0474934577941895, + "learning_rate": 0.0001839401179598493, + "loss": 0.7126, + "step": 11957 + }, + { + "epoch": 0.8297828048018875, + "grad_norm": 1.8275322914123535, + "learning_rate": 0.00018393706404053585, + "loss": 0.8141, + "step": 11958 + }, + { + "epoch": 0.8298521962389841, + "grad_norm": 2.085231304168701, + "learning_rate": 0.00018393400985624282, + "loss": 0.875, + "step": 11959 + }, + { + "epoch": 0.8299215876760808, + "grad_norm": 1.6758452653884888, + "learning_rate": 0.00018393095540697986, + "loss": 0.6639, + "step": 11960 + }, + { + "epoch": 0.8299909791131774, + "grad_norm": 2.40492582321167, + "learning_rate": 0.00018392790069275661, + "loss": 0.7636, + "step": 11961 + }, + { + "epoch": 0.8300603705502742, + "grad_norm": 2.093979835510254, + "learning_rate": 0.0001839248457135827, + "loss": 0.8663, + "step": 11962 + }, + { + "epoch": 0.8301297619873708, + "grad_norm": 1.5687817335128784, + "learning_rate": 0.0001839217904694678, + "loss": 0.7171, + "step": 11963 + }, + { + "epoch": 0.8301991534244674, + "grad_norm": 1.7251999378204346, + "learning_rate": 0.00018391873496042155, + "loss": 0.6363, + "step": 11964 + }, + { + "epoch": 0.8302685448615641, + "grad_norm": 1.8309541940689087, + "learning_rate": 0.00018391567918645357, + "loss": 0.7005, + "step": 11965 + }, + { + "epoch": 0.8303379362986607, + "grad_norm": 1.9010584354400635, + "learning_rate": 0.00018391262314757355, + "loss": 0.7979, + "step": 11966 + }, + { + "epoch": 0.8304073277357574, + "grad_norm": 2.286688804626465, + "learning_rate": 0.00018390956684379112, + "loss": 0.8489, + "step": 11967 + }, + { + "epoch": 0.830476719172854, + "grad_norm": 2.1096675395965576, + "learning_rate": 0.0001839065102751159, + "loss": 0.7905, + "step": 11968 + }, + { + "epoch": 0.8305461106099508, + "grad_norm": 2.4150736331939697, + "learning_rate": 0.00018390345344155757, + "loss": 0.6903, + "step": 11969 + }, + { + "epoch": 0.8306155020470474, + "grad_norm": 3.741640329360962, + "learning_rate": 0.0001839003963431258, + "loss": 1.0589, + "step": 11970 + }, + { + "epoch": 0.8306848934841441, + "grad_norm": 2.1549265384674072, + "learning_rate": 0.00018389733897983016, + "loss": 0.8882, + "step": 11971 + }, + { + "epoch": 0.8307542849212407, + "grad_norm": 1.4902206659317017, + "learning_rate": 0.00018389428135168038, + "loss": 0.4271, + "step": 11972 + }, + { + "epoch": 0.8308236763583374, + "grad_norm": 2.520249366760254, + "learning_rate": 0.0001838912234586861, + "loss": 0.8908, + "step": 11973 + }, + { + "epoch": 0.830893067795434, + "grad_norm": 2.6484742164611816, + "learning_rate": 0.00018388816530085693, + "loss": 0.967, + "step": 11974 + }, + { + "epoch": 0.8309624592325308, + "grad_norm": 2.0081863403320312, + "learning_rate": 0.00018388510687820258, + "loss": 0.7957, + "step": 11975 + }, + { + "epoch": 0.8310318506696274, + "grad_norm": 2.4686050415039062, + "learning_rate": 0.00018388204819073267, + "loss": 0.8563, + "step": 11976 + }, + { + "epoch": 0.831101242106724, + "grad_norm": 1.9606682062149048, + "learning_rate": 0.0001838789892384569, + "loss": 0.7107, + "step": 11977 + }, + { + "epoch": 0.8311706335438207, + "grad_norm": 11.22233772277832, + "learning_rate": 0.00018387593002138484, + "loss": 0.9707, + "step": 11978 + }, + { + "epoch": 0.8312400249809173, + "grad_norm": 2.028839588165283, + "learning_rate": 0.00018387287053952623, + "loss": 0.8391, + "step": 11979 + }, + { + "epoch": 0.831309416418014, + "grad_norm": 2.7440614700317383, + "learning_rate": 0.00018386981079289067, + "loss": 0.9937, + "step": 11980 + }, + { + "epoch": 0.8313788078551106, + "grad_norm": 2.7872676849365234, + "learning_rate": 0.00018386675078148787, + "loss": 0.8507, + "step": 11981 + }, + { + "epoch": 0.8314481992922074, + "grad_norm": 2.3635945320129395, + "learning_rate": 0.0001838636905053275, + "loss": 0.7401, + "step": 11982 + }, + { + "epoch": 0.831517590729304, + "grad_norm": 1.8221300840377808, + "learning_rate": 0.00018386062996441914, + "loss": 0.7519, + "step": 11983 + }, + { + "epoch": 0.8315869821664007, + "grad_norm": 1.7584726810455322, + "learning_rate": 0.0001838575691587725, + "loss": 0.8787, + "step": 11984 + }, + { + "epoch": 0.8316563736034973, + "grad_norm": 2.1851470470428467, + "learning_rate": 0.0001838545080883973, + "loss": 0.8642, + "step": 11985 + }, + { + "epoch": 0.831725765040594, + "grad_norm": 2.6711947917938232, + "learning_rate": 0.0001838514467533031, + "loss": 0.8273, + "step": 11986 + }, + { + "epoch": 0.8317951564776906, + "grad_norm": 2.5974385738372803, + "learning_rate": 0.0001838483851534996, + "loss": 1.0143, + "step": 11987 + }, + { + "epoch": 0.8318645479147874, + "grad_norm": 1.938469409942627, + "learning_rate": 0.0001838453232889965, + "loss": 0.6777, + "step": 11988 + }, + { + "epoch": 0.831933939351884, + "grad_norm": 3.3916609287261963, + "learning_rate": 0.00018384226115980342, + "loss": 0.9737, + "step": 11989 + }, + { + "epoch": 0.8320033307889806, + "grad_norm": 2.7131221294403076, + "learning_rate": 0.00018383919876593008, + "loss": 0.8738, + "step": 11990 + }, + { + "epoch": 0.8320727222260773, + "grad_norm": 3.098191022872925, + "learning_rate": 0.0001838361361073861, + "loss": 0.9206, + "step": 11991 + }, + { + "epoch": 0.8321421136631739, + "grad_norm": 1.9999338388442993, + "learning_rate": 0.00018383307318418118, + "loss": 0.7629, + "step": 11992 + }, + { + "epoch": 0.8322115051002706, + "grad_norm": 2.3355259895324707, + "learning_rate": 0.00018383000999632496, + "loss": 0.7372, + "step": 11993 + }, + { + "epoch": 0.8322808965373673, + "grad_norm": 2.0583789348602295, + "learning_rate": 0.0001838269465438271, + "loss": 0.8524, + "step": 11994 + }, + { + "epoch": 0.832350287974464, + "grad_norm": 2.299427032470703, + "learning_rate": 0.00018382388282669733, + "loss": 0.845, + "step": 11995 + }, + { + "epoch": 0.8324196794115606, + "grad_norm": 2.508981943130493, + "learning_rate": 0.00018382081884494528, + "loss": 0.7543, + "step": 11996 + }, + { + "epoch": 0.8324890708486573, + "grad_norm": 2.294755458831787, + "learning_rate": 0.00018381775459858063, + "loss": 0.8837, + "step": 11997 + }, + { + "epoch": 0.8325584622857539, + "grad_norm": 1.92831289768219, + "learning_rate": 0.00018381469008761308, + "loss": 0.7074, + "step": 11998 + }, + { + "epoch": 0.8326278537228506, + "grad_norm": 2.0830905437469482, + "learning_rate": 0.00018381162531205223, + "loss": 0.7559, + "step": 11999 + }, + { + "epoch": 0.8326972451599473, + "grad_norm": 3.6701500415802, + "learning_rate": 0.00018380856027190782, + "loss": 0.8673, + "step": 12000 + }, + { + "epoch": 0.8326972451599473, + "eval_loss": 0.8368859887123108, + "eval_runtime": 3816.947, + "eval_samples_per_second": 6.359, + "eval_steps_per_second": 0.795, + "step": 12000 + }, + { + "epoch": 0.832766636597044, + "grad_norm": 2.588080883026123, + "learning_rate": 0.00018380549496718951, + "loss": 0.7547, + "step": 12001 + }, + { + "epoch": 0.8328360280341406, + "grad_norm": 1.8149828910827637, + "learning_rate": 0.000183802429397907, + "loss": 0.6474, + "step": 12002 + }, + { + "epoch": 0.8329054194712372, + "grad_norm": 2.044100284576416, + "learning_rate": 0.00018379936356406992, + "loss": 0.7666, + "step": 12003 + }, + { + "epoch": 0.8329748109083339, + "grad_norm": 2.465888500213623, + "learning_rate": 0.00018379629746568797, + "loss": 1.0472, + "step": 12004 + }, + { + "epoch": 0.8330442023454305, + "grad_norm": 1.9043681621551514, + "learning_rate": 0.0001837932311027708, + "loss": 0.911, + "step": 12005 + }, + { + "epoch": 0.8331135937825273, + "grad_norm": 2.151090383529663, + "learning_rate": 0.0001837901644753282, + "loss": 0.8566, + "step": 12006 + }, + { + "epoch": 0.8331829852196239, + "grad_norm": 3.486088991165161, + "learning_rate": 0.00018378709758336972, + "loss": 1.0526, + "step": 12007 + }, + { + "epoch": 0.8332523766567206, + "grad_norm": 2.929922580718994, + "learning_rate": 0.0001837840304269051, + "loss": 1.1264, + "step": 12008 + }, + { + "epoch": 0.8333217680938172, + "grad_norm": 2.0183281898498535, + "learning_rate": 0.000183780963005944, + "loss": 0.6919, + "step": 12009 + }, + { + "epoch": 0.8333911595309139, + "grad_norm": 2.124366283416748, + "learning_rate": 0.00018377789532049614, + "loss": 0.768, + "step": 12010 + }, + { + "epoch": 0.8334605509680105, + "grad_norm": 2.3934643268585205, + "learning_rate": 0.0001837748273705712, + "loss": 0.8453, + "step": 12011 + }, + { + "epoch": 0.8335299424051072, + "grad_norm": 2.8956096172332764, + "learning_rate": 0.0001837717591561788, + "loss": 0.8306, + "step": 12012 + }, + { + "epoch": 0.8335993338422039, + "grad_norm": 2.3378405570983887, + "learning_rate": 0.00018376869067732873, + "loss": 0.7183, + "step": 12013 + }, + { + "epoch": 0.8336687252793006, + "grad_norm": 2.559112071990967, + "learning_rate": 0.0001837656219340306, + "loss": 0.8037, + "step": 12014 + }, + { + "epoch": 0.8337381167163972, + "grad_norm": 1.9395805597305298, + "learning_rate": 0.00018376255292629413, + "loss": 0.7782, + "step": 12015 + }, + { + "epoch": 0.8338075081534938, + "grad_norm": 2.968855619430542, + "learning_rate": 0.00018375948365412898, + "loss": 0.7388, + "step": 12016 + }, + { + "epoch": 0.8338768995905905, + "grad_norm": 2.198265314102173, + "learning_rate": 0.00018375641411754487, + "loss": 0.9153, + "step": 12017 + }, + { + "epoch": 0.8339462910276871, + "grad_norm": 1.7106685638427734, + "learning_rate": 0.00018375334431655147, + "loss": 0.5398, + "step": 12018 + }, + { + "epoch": 0.8340156824647839, + "grad_norm": 3.1354568004608154, + "learning_rate": 0.0001837502742511585, + "loss": 0.7284, + "step": 12019 + }, + { + "epoch": 0.8340850739018805, + "grad_norm": 2.1537132263183594, + "learning_rate": 0.0001837472039213756, + "loss": 0.6818, + "step": 12020 + }, + { + "epoch": 0.8341544653389772, + "grad_norm": 2.262723922729492, + "learning_rate": 0.00018374413332721253, + "loss": 0.7571, + "step": 12021 + }, + { + "epoch": 0.8342238567760738, + "grad_norm": 2.2920899391174316, + "learning_rate": 0.00018374106246867893, + "loss": 0.9031, + "step": 12022 + }, + { + "epoch": 0.8342932482131705, + "grad_norm": 1.6290411949157715, + "learning_rate": 0.0001837379913457845, + "loss": 0.7032, + "step": 12023 + }, + { + "epoch": 0.8343626396502671, + "grad_norm": 2.1723897457122803, + "learning_rate": 0.00018373491995853897, + "loss": 0.8666, + "step": 12024 + }, + { + "epoch": 0.8344320310873639, + "grad_norm": 2.1123197078704834, + "learning_rate": 0.000183731848306952, + "loss": 0.7109, + "step": 12025 + }, + { + "epoch": 0.8345014225244605, + "grad_norm": 4.182211399078369, + "learning_rate": 0.00018372877639103328, + "loss": 0.8953, + "step": 12026 + }, + { + "epoch": 0.8345708139615572, + "grad_norm": 2.27431058883667, + "learning_rate": 0.00018372570421079255, + "loss": 0.9201, + "step": 12027 + }, + { + "epoch": 0.8346402053986538, + "grad_norm": 1.8972177505493164, + "learning_rate": 0.00018372263176623946, + "loss": 0.8509, + "step": 12028 + }, + { + "epoch": 0.8347095968357505, + "grad_norm": 2.0069549083709717, + "learning_rate": 0.00018371955905738377, + "loss": 1.0096, + "step": 12029 + }, + { + "epoch": 0.8347789882728471, + "grad_norm": 1.9802916049957275, + "learning_rate": 0.00018371648608423512, + "loss": 0.776, + "step": 12030 + }, + { + "epoch": 0.8348483797099437, + "grad_norm": 1.8942095041275024, + "learning_rate": 0.00018371341284680325, + "loss": 0.6241, + "step": 12031 + }, + { + "epoch": 0.8349177711470405, + "grad_norm": 2.6025378704071045, + "learning_rate": 0.00018371033934509785, + "loss": 0.8207, + "step": 12032 + }, + { + "epoch": 0.8349871625841371, + "grad_norm": 1.7946934700012207, + "learning_rate": 0.00018370726557912862, + "loss": 0.6793, + "step": 12033 + }, + { + "epoch": 0.8350565540212338, + "grad_norm": 2.7143657207489014, + "learning_rate": 0.00018370419154890527, + "loss": 1.0424, + "step": 12034 + }, + { + "epoch": 0.8351259454583304, + "grad_norm": 2.6451098918914795, + "learning_rate": 0.00018370111725443748, + "loss": 0.869, + "step": 12035 + }, + { + "epoch": 0.8351953368954271, + "grad_norm": 2.4846129417419434, + "learning_rate": 0.00018369804269573498, + "loss": 0.904, + "step": 12036 + }, + { + "epoch": 0.8352647283325237, + "grad_norm": 2.012509822845459, + "learning_rate": 0.0001836949678728075, + "loss": 0.8044, + "step": 12037 + }, + { + "epoch": 0.8353341197696205, + "grad_norm": 2.46166729927063, + "learning_rate": 0.00018369189278566467, + "loss": 1.0017, + "step": 12038 + }, + { + "epoch": 0.8354035112067171, + "grad_norm": 2.6276395320892334, + "learning_rate": 0.00018368881743431627, + "loss": 0.7764, + "step": 12039 + }, + { + "epoch": 0.8354729026438138, + "grad_norm": 2.6948750019073486, + "learning_rate": 0.000183685741818772, + "loss": 0.8289, + "step": 12040 + }, + { + "epoch": 0.8355422940809104, + "grad_norm": 2.357119560241699, + "learning_rate": 0.00018368266593904152, + "loss": 0.8063, + "step": 12041 + }, + { + "epoch": 0.8356116855180071, + "grad_norm": 2.643005847930908, + "learning_rate": 0.00018367958979513457, + "loss": 0.8585, + "step": 12042 + }, + { + "epoch": 0.8356810769551037, + "grad_norm": 2.312062978744507, + "learning_rate": 0.00018367651338706086, + "loss": 0.8879, + "step": 12043 + }, + { + "epoch": 0.8357504683922004, + "grad_norm": 1.7273709774017334, + "learning_rate": 0.00018367343671483013, + "loss": 0.6722, + "step": 12044 + }, + { + "epoch": 0.8358198598292971, + "grad_norm": 2.055684804916382, + "learning_rate": 0.00018367035977845207, + "loss": 0.7824, + "step": 12045 + }, + { + "epoch": 0.8358892512663937, + "grad_norm": 2.462843418121338, + "learning_rate": 0.00018366728257793638, + "loss": 0.8658, + "step": 12046 + }, + { + "epoch": 0.8359586427034904, + "grad_norm": 2.60577392578125, + "learning_rate": 0.00018366420511329277, + "loss": 1.0974, + "step": 12047 + }, + { + "epoch": 0.836028034140587, + "grad_norm": 2.1058478355407715, + "learning_rate": 0.000183661127384531, + "loss": 0.7675, + "step": 12048 + }, + { + "epoch": 0.8360974255776837, + "grad_norm": 2.3180763721466064, + "learning_rate": 0.00018365804939166075, + "loss": 0.7601, + "step": 12049 + }, + { + "epoch": 0.8361668170147804, + "grad_norm": 2.5631580352783203, + "learning_rate": 0.00018365497113469173, + "loss": 0.9181, + "step": 12050 + }, + { + "epoch": 0.8362362084518771, + "grad_norm": 1.9734452962875366, + "learning_rate": 0.0001836518926136337, + "loss": 0.7202, + "step": 12051 + }, + { + "epoch": 0.8363055998889737, + "grad_norm": 3.9552948474884033, + "learning_rate": 0.00018364881382849633, + "loss": 1.0176, + "step": 12052 + }, + { + "epoch": 0.8363749913260704, + "grad_norm": 1.902398943901062, + "learning_rate": 0.00018364573477928935, + "loss": 0.6302, + "step": 12053 + }, + { + "epoch": 0.836444382763167, + "grad_norm": 2.5254836082458496, + "learning_rate": 0.0001836426554660225, + "loss": 0.6933, + "step": 12054 + }, + { + "epoch": 0.8365137742002637, + "grad_norm": 2.4305126667022705, + "learning_rate": 0.00018363957588870552, + "loss": 0.6559, + "step": 12055 + }, + { + "epoch": 0.8365831656373603, + "grad_norm": 2.545517921447754, + "learning_rate": 0.00018363649604734804, + "loss": 0.6848, + "step": 12056 + }, + { + "epoch": 0.836652557074457, + "grad_norm": 2.3732802867889404, + "learning_rate": 0.0001836334159419599, + "loss": 0.6246, + "step": 12057 + }, + { + "epoch": 0.8367219485115537, + "grad_norm": 2.292848587036133, + "learning_rate": 0.00018363033557255076, + "loss": 0.9611, + "step": 12058 + }, + { + "epoch": 0.8367913399486503, + "grad_norm": 1.7686551809310913, + "learning_rate": 0.00018362725493913034, + "loss": 0.7921, + "step": 12059 + }, + { + "epoch": 0.836860731385747, + "grad_norm": 2.5339338779449463, + "learning_rate": 0.0001836241740417084, + "loss": 1.1034, + "step": 12060 + }, + { + "epoch": 0.8369301228228436, + "grad_norm": 2.308530569076538, + "learning_rate": 0.0001836210928802946, + "loss": 0.8518, + "step": 12061 + }, + { + "epoch": 0.8369995142599403, + "grad_norm": 1.7550300359725952, + "learning_rate": 0.00018361801145489875, + "loss": 0.6406, + "step": 12062 + }, + { + "epoch": 0.837068905697037, + "grad_norm": 1.9805530309677124, + "learning_rate": 0.00018361492976553054, + "loss": 0.5824, + "step": 12063 + }, + { + "epoch": 0.8371382971341337, + "grad_norm": 1.9734305143356323, + "learning_rate": 0.00018361184781219968, + "loss": 0.6927, + "step": 12064 + }, + { + "epoch": 0.8372076885712303, + "grad_norm": 1.9519555568695068, + "learning_rate": 0.00018360876559491592, + "loss": 0.729, + "step": 12065 + }, + { + "epoch": 0.837277080008327, + "grad_norm": 2.0749282836914062, + "learning_rate": 0.000183605683113689, + "loss": 0.6754, + "step": 12066 + }, + { + "epoch": 0.8373464714454236, + "grad_norm": 2.2776315212249756, + "learning_rate": 0.00018360260036852862, + "loss": 0.5534, + "step": 12067 + }, + { + "epoch": 0.8374158628825203, + "grad_norm": 2.23645281791687, + "learning_rate": 0.00018359951735944457, + "loss": 0.8124, + "step": 12068 + }, + { + "epoch": 0.837485254319617, + "grad_norm": 2.3081204891204834, + "learning_rate": 0.00018359643408644648, + "loss": 0.8798, + "step": 12069 + }, + { + "epoch": 0.8375546457567136, + "grad_norm": 2.3752546310424805, + "learning_rate": 0.0001835933505495442, + "loss": 0.8273, + "step": 12070 + }, + { + "epoch": 0.8376240371938103, + "grad_norm": 2.707143545150757, + "learning_rate": 0.00018359026674874738, + "loss": 0.7664, + "step": 12071 + }, + { + "epoch": 0.8376934286309069, + "grad_norm": 1.9358503818511963, + "learning_rate": 0.0001835871826840658, + "loss": 0.7168, + "step": 12072 + }, + { + "epoch": 0.8377628200680036, + "grad_norm": 3.0067622661590576, + "learning_rate": 0.00018358409835550919, + "loss": 0.756, + "step": 12073 + }, + { + "epoch": 0.8378322115051002, + "grad_norm": 2.381165027618408, + "learning_rate": 0.00018358101376308728, + "loss": 0.8899, + "step": 12074 + }, + { + "epoch": 0.837901602942197, + "grad_norm": 2.778578996658325, + "learning_rate": 0.0001835779289068098, + "loss": 0.7691, + "step": 12075 + }, + { + "epoch": 0.8379709943792936, + "grad_norm": 2.858004331588745, + "learning_rate": 0.00018357484378668648, + "loss": 0.731, + "step": 12076 + }, + { + "epoch": 0.8380403858163903, + "grad_norm": 1.7013412714004517, + "learning_rate": 0.0001835717584027271, + "loss": 0.5312, + "step": 12077 + }, + { + "epoch": 0.8381097772534869, + "grad_norm": 2.0981454849243164, + "learning_rate": 0.00018356867275494133, + "loss": 0.7432, + "step": 12078 + }, + { + "epoch": 0.8381791686905836, + "grad_norm": 2.109527111053467, + "learning_rate": 0.000183565586843339, + "loss": 0.7831, + "step": 12079 + }, + { + "epoch": 0.8382485601276802, + "grad_norm": 2.2050745487213135, + "learning_rate": 0.0001835625006679298, + "loss": 0.9059, + "step": 12080 + }, + { + "epoch": 0.838317951564777, + "grad_norm": 2.695918321609497, + "learning_rate": 0.00018355941422872347, + "loss": 0.9227, + "step": 12081 + }, + { + "epoch": 0.8383873430018736, + "grad_norm": 2.325814962387085, + "learning_rate": 0.00018355632752572977, + "loss": 0.8076, + "step": 12082 + }, + { + "epoch": 0.8384567344389703, + "grad_norm": 1.7331355810165405, + "learning_rate": 0.00018355324055895845, + "loss": 0.6598, + "step": 12083 + }, + { + "epoch": 0.8385261258760669, + "grad_norm": 2.2886693477630615, + "learning_rate": 0.00018355015332841923, + "loss": 0.7865, + "step": 12084 + }, + { + "epoch": 0.8385955173131635, + "grad_norm": 2.4096362590789795, + "learning_rate": 0.00018354706583412184, + "loss": 0.799, + "step": 12085 + }, + { + "epoch": 0.8386649087502602, + "grad_norm": 2.7466776371002197, + "learning_rate": 0.00018354397807607612, + "loss": 0.9673, + "step": 12086 + }, + { + "epoch": 0.8387343001873568, + "grad_norm": 2.0978002548217773, + "learning_rate": 0.00018354089005429168, + "loss": 0.775, + "step": 12087 + }, + { + "epoch": 0.8388036916244536, + "grad_norm": 3.613942861557007, + "learning_rate": 0.0001835378017687784, + "loss": 0.828, + "step": 12088 + }, + { + "epoch": 0.8388730830615502, + "grad_norm": 2.032541036605835, + "learning_rate": 0.00018353471321954596, + "loss": 0.8532, + "step": 12089 + }, + { + "epoch": 0.8389424744986469, + "grad_norm": 1.8617230653762817, + "learning_rate": 0.0001835316244066041, + "loss": 0.6913, + "step": 12090 + }, + { + "epoch": 0.8390118659357435, + "grad_norm": 2.3984596729278564, + "learning_rate": 0.00018352853532996265, + "loss": 0.8575, + "step": 12091 + }, + { + "epoch": 0.8390812573728402, + "grad_norm": 2.1855297088623047, + "learning_rate": 0.00018352544598963124, + "loss": 0.6978, + "step": 12092 + }, + { + "epoch": 0.8391506488099368, + "grad_norm": 1.9342107772827148, + "learning_rate": 0.00018352235638561973, + "loss": 0.7499, + "step": 12093 + }, + { + "epoch": 0.8392200402470336, + "grad_norm": 1.9766103029251099, + "learning_rate": 0.0001835192665179378, + "loss": 0.7348, + "step": 12094 + }, + { + "epoch": 0.8392894316841302, + "grad_norm": 2.299771308898926, + "learning_rate": 0.00018351617638659527, + "loss": 0.7894, + "step": 12095 + }, + { + "epoch": 0.8393588231212269, + "grad_norm": 2.7316133975982666, + "learning_rate": 0.0001835130859916018, + "loss": 0.8649, + "step": 12096 + }, + { + "epoch": 0.8394282145583235, + "grad_norm": 1.9297596216201782, + "learning_rate": 0.00018350999533296727, + "loss": 0.7668, + "step": 12097 + }, + { + "epoch": 0.8394976059954201, + "grad_norm": 4.034580707550049, + "learning_rate": 0.00018350690441070136, + "loss": 0.8613, + "step": 12098 + }, + { + "epoch": 0.8395669974325168, + "grad_norm": 3.0305933952331543, + "learning_rate": 0.00018350381322481383, + "loss": 0.799, + "step": 12099 + }, + { + "epoch": 0.8396363888696134, + "grad_norm": 2.2671968936920166, + "learning_rate": 0.00018350072177531444, + "loss": 0.7163, + "step": 12100 + }, + { + "epoch": 0.8397057803067102, + "grad_norm": 3.3035120964050293, + "learning_rate": 0.00018349763006221298, + "loss": 0.8121, + "step": 12101 + }, + { + "epoch": 0.8397751717438068, + "grad_norm": 2.171281576156616, + "learning_rate": 0.00018349453808551918, + "loss": 0.7882, + "step": 12102 + }, + { + "epoch": 0.8398445631809035, + "grad_norm": 1.7902076244354248, + "learning_rate": 0.00018349144584524281, + "loss": 0.8357, + "step": 12103 + }, + { + "epoch": 0.8399139546180001, + "grad_norm": 2.2530558109283447, + "learning_rate": 0.00018348835334139362, + "loss": 0.9201, + "step": 12104 + }, + { + "epoch": 0.8399833460550968, + "grad_norm": 2.148850679397583, + "learning_rate": 0.00018348526057398141, + "loss": 1.0078, + "step": 12105 + }, + { + "epoch": 0.8400527374921934, + "grad_norm": 2.467437744140625, + "learning_rate": 0.00018348216754301592, + "loss": 0.9978, + "step": 12106 + }, + { + "epoch": 0.8401221289292902, + "grad_norm": 2.512934684753418, + "learning_rate": 0.0001834790742485069, + "loss": 0.8742, + "step": 12107 + }, + { + "epoch": 0.8401915203663868, + "grad_norm": 2.583005905151367, + "learning_rate": 0.0001834759806904641, + "loss": 0.9938, + "step": 12108 + }, + { + "epoch": 0.8402609118034835, + "grad_norm": 2.7794411182403564, + "learning_rate": 0.00018347288686889737, + "loss": 1.0006, + "step": 12109 + }, + { + "epoch": 0.8403303032405801, + "grad_norm": 2.1557135581970215, + "learning_rate": 0.0001834697927838164, + "loss": 0.8047, + "step": 12110 + }, + { + "epoch": 0.8403996946776767, + "grad_norm": 2.0101664066314697, + "learning_rate": 0.00018346669843523095, + "loss": 0.8154, + "step": 12111 + }, + { + "epoch": 0.8404690861147734, + "grad_norm": 1.770199179649353, + "learning_rate": 0.00018346360382315085, + "loss": 0.6347, + "step": 12112 + }, + { + "epoch": 0.8405384775518701, + "grad_norm": 2.4753212928771973, + "learning_rate": 0.00018346050894758583, + "loss": 1.0514, + "step": 12113 + }, + { + "epoch": 0.8406078689889668, + "grad_norm": 2.2889115810394287, + "learning_rate": 0.00018345741380854563, + "loss": 0.9155, + "step": 12114 + }, + { + "epoch": 0.8406772604260634, + "grad_norm": 1.9390267133712769, + "learning_rate": 0.00018345431840604014, + "loss": 0.7069, + "step": 12115 + }, + { + "epoch": 0.8407466518631601, + "grad_norm": 2.388625383377075, + "learning_rate": 0.00018345122274007896, + "loss": 0.9645, + "step": 12116 + }, + { + "epoch": 0.8408160433002567, + "grad_norm": 3.788306713104248, + "learning_rate": 0.00018344812681067197, + "loss": 0.6968, + "step": 12117 + }, + { + "epoch": 0.8408854347373534, + "grad_norm": 2.1446492671966553, + "learning_rate": 0.00018344503061782898, + "loss": 0.756, + "step": 12118 + }, + { + "epoch": 0.84095482617445, + "grad_norm": 2.3757615089416504, + "learning_rate": 0.0001834419341615597, + "loss": 0.7021, + "step": 12119 + }, + { + "epoch": 0.8410242176115468, + "grad_norm": 2.565131902694702, + "learning_rate": 0.00018343883744187386, + "loss": 0.8682, + "step": 12120 + }, + { + "epoch": 0.8410936090486434, + "grad_norm": 2.8650593757629395, + "learning_rate": 0.00018343574045878133, + "loss": 1.0432, + "step": 12121 + }, + { + "epoch": 0.8411630004857401, + "grad_norm": 1.9841946363449097, + "learning_rate": 0.00018343264321229183, + "loss": 0.7547, + "step": 12122 + }, + { + "epoch": 0.8412323919228367, + "grad_norm": 2.5654494762420654, + "learning_rate": 0.00018342954570241515, + "loss": 0.9219, + "step": 12123 + }, + { + "epoch": 0.8413017833599333, + "grad_norm": 2.383615016937256, + "learning_rate": 0.00018342644792916112, + "loss": 0.8095, + "step": 12124 + }, + { + "epoch": 0.84137117479703, + "grad_norm": 3.0147318840026855, + "learning_rate": 0.00018342334989253942, + "loss": 0.9115, + "step": 12125 + }, + { + "epoch": 0.8414405662341267, + "grad_norm": 2.157372236251831, + "learning_rate": 0.0001834202515925599, + "loss": 0.9558, + "step": 12126 + }, + { + "epoch": 0.8415099576712234, + "grad_norm": 1.6612657308578491, + "learning_rate": 0.00018341715302923233, + "loss": 0.6824, + "step": 12127 + }, + { + "epoch": 0.84157934910832, + "grad_norm": 1.6719398498535156, + "learning_rate": 0.0001834140542025665, + "loss": 0.5525, + "step": 12128 + }, + { + "epoch": 0.8416487405454167, + "grad_norm": 1.9336059093475342, + "learning_rate": 0.00018341095511257216, + "loss": 0.7571, + "step": 12129 + }, + { + "epoch": 0.8417181319825133, + "grad_norm": 2.663383960723877, + "learning_rate": 0.00018340785575925912, + "loss": 0.9531, + "step": 12130 + }, + { + "epoch": 0.84178752341961, + "grad_norm": 2.382178783416748, + "learning_rate": 0.00018340475614263716, + "loss": 0.6891, + "step": 12131 + }, + { + "epoch": 0.8418569148567067, + "grad_norm": 1.814487338066101, + "learning_rate": 0.00018340165626271605, + "loss": 0.73, + "step": 12132 + }, + { + "epoch": 0.8419263062938034, + "grad_norm": 2.471992254257202, + "learning_rate": 0.00018339855611950557, + "loss": 1.0018, + "step": 12133 + }, + { + "epoch": 0.8419956977309, + "grad_norm": 2.829453945159912, + "learning_rate": 0.00018339545571301554, + "loss": 1.005, + "step": 12134 + }, + { + "epoch": 0.8420650891679967, + "grad_norm": 2.772723913192749, + "learning_rate": 0.00018339235504325575, + "loss": 0.9384, + "step": 12135 + }, + { + "epoch": 0.8421344806050933, + "grad_norm": 2.5924432277679443, + "learning_rate": 0.00018338925411023595, + "loss": 1.0576, + "step": 12136 + }, + { + "epoch": 0.8422038720421899, + "grad_norm": 3.1223037242889404, + "learning_rate": 0.00018338615291396598, + "loss": 1.131, + "step": 12137 + }, + { + "epoch": 0.8422732634792867, + "grad_norm": 1.9688564538955688, + "learning_rate": 0.00018338305145445557, + "loss": 0.8738, + "step": 12138 + }, + { + "epoch": 0.8423426549163833, + "grad_norm": 2.323965311050415, + "learning_rate": 0.00018337994973171453, + "loss": 0.7037, + "step": 12139 + }, + { + "epoch": 0.84241204635348, + "grad_norm": 2.6814610958099365, + "learning_rate": 0.0001833768477457527, + "loss": 0.8906, + "step": 12140 + }, + { + "epoch": 0.8424814377905766, + "grad_norm": 2.8556156158447266, + "learning_rate": 0.00018337374549657983, + "loss": 1.1557, + "step": 12141 + }, + { + "epoch": 0.8425508292276733, + "grad_norm": 1.9141879081726074, + "learning_rate": 0.00018337064298420567, + "loss": 0.6459, + "step": 12142 + }, + { + "epoch": 0.8426202206647699, + "grad_norm": 1.6709222793579102, + "learning_rate": 0.00018336754020864012, + "loss": 0.7021, + "step": 12143 + }, + { + "epoch": 0.8426896121018667, + "grad_norm": 2.7562754154205322, + "learning_rate": 0.00018336443716989288, + "loss": 0.9911, + "step": 12144 + }, + { + "epoch": 0.8427590035389633, + "grad_norm": 1.664876103401184, + "learning_rate": 0.00018336133386797382, + "loss": 0.6665, + "step": 12145 + }, + { + "epoch": 0.84282839497606, + "grad_norm": 4.425276756286621, + "learning_rate": 0.00018335823030289267, + "loss": 0.8493, + "step": 12146 + }, + { + "epoch": 0.8428977864131566, + "grad_norm": 2.148512125015259, + "learning_rate": 0.00018335512647465926, + "loss": 0.8755, + "step": 12147 + }, + { + "epoch": 0.8429671778502533, + "grad_norm": 2.3166611194610596, + "learning_rate": 0.00018335202238328342, + "loss": 0.9962, + "step": 12148 + }, + { + "epoch": 0.8430365692873499, + "grad_norm": 1.8787864446640015, + "learning_rate": 0.00018334891802877486, + "loss": 0.7541, + "step": 12149 + }, + { + "epoch": 0.8431059607244467, + "grad_norm": 2.232759952545166, + "learning_rate": 0.00018334581341114347, + "loss": 0.671, + "step": 12150 + }, + { + "epoch": 0.8431753521615433, + "grad_norm": 2.2579004764556885, + "learning_rate": 0.000183342708530399, + "loss": 0.7959, + "step": 12151 + }, + { + "epoch": 0.8432447435986399, + "grad_norm": 2.8961966037750244, + "learning_rate": 0.0001833396033865513, + "loss": 0.9121, + "step": 12152 + }, + { + "epoch": 0.8433141350357366, + "grad_norm": 2.2016217708587646, + "learning_rate": 0.0001833364979796101, + "loss": 0.8454, + "step": 12153 + }, + { + "epoch": 0.8433835264728332, + "grad_norm": 2.175996780395508, + "learning_rate": 0.00018333339230958528, + "loss": 1.053, + "step": 12154 + }, + { + "epoch": 0.8434529179099299, + "grad_norm": 2.1594297885894775, + "learning_rate": 0.00018333028637648662, + "loss": 0.9205, + "step": 12155 + }, + { + "epoch": 0.8435223093470265, + "grad_norm": 2.021911382675171, + "learning_rate": 0.0001833271801803239, + "loss": 0.76, + "step": 12156 + }, + { + "epoch": 0.8435917007841233, + "grad_norm": 1.991857886314392, + "learning_rate": 0.00018332407372110692, + "loss": 0.9126, + "step": 12157 + }, + { + "epoch": 0.8436610922212199, + "grad_norm": 2.1484129428863525, + "learning_rate": 0.0001833209669988455, + "loss": 0.8947, + "step": 12158 + }, + { + "epoch": 0.8437304836583166, + "grad_norm": 2.5515763759613037, + "learning_rate": 0.0001833178600135495, + "loss": 1.0412, + "step": 12159 + }, + { + "epoch": 0.8437998750954132, + "grad_norm": 2.282977819442749, + "learning_rate": 0.00018331475276522864, + "loss": 0.8077, + "step": 12160 + }, + { + "epoch": 0.8438692665325099, + "grad_norm": 2.8135838508605957, + "learning_rate": 0.0001833116452538928, + "loss": 1.0124, + "step": 12161 + }, + { + "epoch": 0.8439386579696065, + "grad_norm": 1.6264679431915283, + "learning_rate": 0.00018330853747955175, + "loss": 0.7364, + "step": 12162 + }, + { + "epoch": 0.8440080494067033, + "grad_norm": 2.0972256660461426, + "learning_rate": 0.0001833054294422153, + "loss": 0.7207, + "step": 12163 + }, + { + "epoch": 0.8440774408437999, + "grad_norm": 1.8458263874053955, + "learning_rate": 0.00018330232114189328, + "loss": 0.7164, + "step": 12164 + }, + { + "epoch": 0.8441468322808965, + "grad_norm": 1.8373440504074097, + "learning_rate": 0.0001832992125785955, + "loss": 0.6774, + "step": 12165 + }, + { + "epoch": 0.8442162237179932, + "grad_norm": 2.450432300567627, + "learning_rate": 0.0001832961037523318, + "loss": 0.9616, + "step": 12166 + }, + { + "epoch": 0.8442856151550898, + "grad_norm": 2.0277106761932373, + "learning_rate": 0.00018329299466311193, + "loss": 0.7941, + "step": 12167 + }, + { + "epoch": 0.8443550065921865, + "grad_norm": 1.7430561780929565, + "learning_rate": 0.00018328988531094573, + "loss": 0.6317, + "step": 12168 + }, + { + "epoch": 0.8444243980292832, + "grad_norm": 1.8810561895370483, + "learning_rate": 0.00018328677569584304, + "loss": 0.7118, + "step": 12169 + }, + { + "epoch": 0.8444937894663799, + "grad_norm": 2.320035934448242, + "learning_rate": 0.00018328366581781368, + "loss": 0.9365, + "step": 12170 + }, + { + "epoch": 0.8445631809034765, + "grad_norm": 1.858393669128418, + "learning_rate": 0.00018328055567686744, + "loss": 0.7134, + "step": 12171 + }, + { + "epoch": 0.8446325723405732, + "grad_norm": 1.7281652688980103, + "learning_rate": 0.00018327744527301413, + "loss": 0.5439, + "step": 12172 + }, + { + "epoch": 0.8447019637776698, + "grad_norm": 2.6744301319122314, + "learning_rate": 0.00018327433460626363, + "loss": 0.724, + "step": 12173 + }, + { + "epoch": 0.8447713552147665, + "grad_norm": 1.8287551403045654, + "learning_rate": 0.00018327122367662568, + "loss": 0.7354, + "step": 12174 + }, + { + "epoch": 0.8448407466518632, + "grad_norm": 1.9514144659042358, + "learning_rate": 0.0001832681124841101, + "loss": 0.8121, + "step": 12175 + }, + { + "epoch": 0.8449101380889599, + "grad_norm": 2.202455759048462, + "learning_rate": 0.0001832650010287268, + "loss": 0.848, + "step": 12176 + }, + { + "epoch": 0.8449795295260565, + "grad_norm": 2.1168363094329834, + "learning_rate": 0.00018326188931048557, + "loss": 0.8744, + "step": 12177 + }, + { + "epoch": 0.8450489209631531, + "grad_norm": 2.7278189659118652, + "learning_rate": 0.0001832587773293962, + "loss": 0.9164, + "step": 12178 + }, + { + "epoch": 0.8451183124002498, + "grad_norm": 2.35656476020813, + "learning_rate": 0.00018325566508546853, + "loss": 0.969, + "step": 12179 + }, + { + "epoch": 0.8451877038373464, + "grad_norm": 2.5012295246124268, + "learning_rate": 0.00018325255257871237, + "loss": 0.8434, + "step": 12180 + }, + { + "epoch": 0.8452570952744431, + "grad_norm": 1.8957493305206299, + "learning_rate": 0.00018324943980913756, + "loss": 0.7902, + "step": 12181 + }, + { + "epoch": 0.8453264867115398, + "grad_norm": 1.950962781906128, + "learning_rate": 0.00018324632677675396, + "loss": 0.813, + "step": 12182 + }, + { + "epoch": 0.8453958781486365, + "grad_norm": 3.0516583919525146, + "learning_rate": 0.00018324321348157132, + "loss": 0.9324, + "step": 12183 + }, + { + "epoch": 0.8454652695857331, + "grad_norm": 2.1450893878936768, + "learning_rate": 0.00018324009992359955, + "loss": 0.5936, + "step": 12184 + }, + { + "epoch": 0.8455346610228298, + "grad_norm": 2.0731163024902344, + "learning_rate": 0.00018323698610284842, + "loss": 0.9679, + "step": 12185 + }, + { + "epoch": 0.8456040524599264, + "grad_norm": 2.594067096710205, + "learning_rate": 0.00018323387201932777, + "loss": 1.0918, + "step": 12186 + }, + { + "epoch": 0.8456734438970231, + "grad_norm": 2.386683225631714, + "learning_rate": 0.00018323075767304748, + "loss": 0.785, + "step": 12187 + }, + { + "epoch": 0.8457428353341198, + "grad_norm": 2.052244186401367, + "learning_rate": 0.00018322764306401732, + "loss": 0.697, + "step": 12188 + }, + { + "epoch": 0.8458122267712165, + "grad_norm": 2.541877269744873, + "learning_rate": 0.00018322452819224716, + "loss": 0.9284, + "step": 12189 + }, + { + "epoch": 0.8458816182083131, + "grad_norm": 2.040947437286377, + "learning_rate": 0.0001832214130577468, + "loss": 0.7417, + "step": 12190 + }, + { + "epoch": 0.8459510096454097, + "grad_norm": 3.02998948097229, + "learning_rate": 0.00018321829766052613, + "loss": 0.792, + "step": 12191 + }, + { + "epoch": 0.8460204010825064, + "grad_norm": 2.301194667816162, + "learning_rate": 0.0001832151820005949, + "loss": 0.8237, + "step": 12192 + }, + { + "epoch": 0.846089792519603, + "grad_norm": 2.1114261150360107, + "learning_rate": 0.00018321206607796304, + "loss": 0.6015, + "step": 12193 + }, + { + "epoch": 0.8461591839566998, + "grad_norm": 1.7367643117904663, + "learning_rate": 0.0001832089498926403, + "loss": 0.4695, + "step": 12194 + }, + { + "epoch": 0.8462285753937964, + "grad_norm": 3.5907371044158936, + "learning_rate": 0.00018320583344463658, + "loss": 0.8342, + "step": 12195 + }, + { + "epoch": 0.8462979668308931, + "grad_norm": 2.542480707168579, + "learning_rate": 0.0001832027167339617, + "loss": 1.0521, + "step": 12196 + }, + { + "epoch": 0.8463673582679897, + "grad_norm": 2.4560208320617676, + "learning_rate": 0.00018319959976062551, + "loss": 0.6735, + "step": 12197 + }, + { + "epoch": 0.8464367497050864, + "grad_norm": 3.2346417903900146, + "learning_rate": 0.0001831964825246378, + "loss": 0.9104, + "step": 12198 + }, + { + "epoch": 0.846506141142183, + "grad_norm": 2.2164146900177, + "learning_rate": 0.00018319336502600848, + "loss": 0.9158, + "step": 12199 + }, + { + "epoch": 0.8465755325792798, + "grad_norm": 2.3539490699768066, + "learning_rate": 0.00018319024726474737, + "loss": 0.6914, + "step": 12200 + }, + { + "epoch": 0.8466449240163764, + "grad_norm": 3.1924173831939697, + "learning_rate": 0.0001831871292408643, + "loss": 1.0408, + "step": 12201 + }, + { + "epoch": 0.8467143154534731, + "grad_norm": 1.8376796245574951, + "learning_rate": 0.00018318401095436908, + "loss": 0.7323, + "step": 12202 + }, + { + "epoch": 0.8467837068905697, + "grad_norm": 2.2048587799072266, + "learning_rate": 0.00018318089240527158, + "loss": 0.9516, + "step": 12203 + }, + { + "epoch": 0.8468530983276664, + "grad_norm": 1.991502046585083, + "learning_rate": 0.0001831777735935817, + "loss": 0.7144, + "step": 12204 + }, + { + "epoch": 0.846922489764763, + "grad_norm": 1.962799072265625, + "learning_rate": 0.0001831746545193092, + "loss": 0.7957, + "step": 12205 + }, + { + "epoch": 0.8469918812018596, + "grad_norm": 2.1031670570373535, + "learning_rate": 0.00018317153518246398, + "loss": 0.8379, + "step": 12206 + }, + { + "epoch": 0.8470612726389564, + "grad_norm": 2.6443777084350586, + "learning_rate": 0.0001831684155830559, + "loss": 0.8382, + "step": 12207 + }, + { + "epoch": 0.847130664076053, + "grad_norm": 1.9782379865646362, + "learning_rate": 0.00018316529572109475, + "loss": 0.6154, + "step": 12208 + }, + { + "epoch": 0.8472000555131497, + "grad_norm": 1.7790759801864624, + "learning_rate": 0.0001831621755965904, + "loss": 0.876, + "step": 12209 + }, + { + "epoch": 0.8472694469502463, + "grad_norm": 2.5326108932495117, + "learning_rate": 0.00018315905520955274, + "loss": 1.1069, + "step": 12210 + }, + { + "epoch": 0.847338838387343, + "grad_norm": 1.9786672592163086, + "learning_rate": 0.00018315593455999157, + "loss": 0.6816, + "step": 12211 + }, + { + "epoch": 0.8474082298244396, + "grad_norm": 1.8967527151107788, + "learning_rate": 0.00018315281364791677, + "loss": 0.7776, + "step": 12212 + }, + { + "epoch": 0.8474776212615364, + "grad_norm": 2.647808074951172, + "learning_rate": 0.0001831496924733382, + "loss": 1.0205, + "step": 12213 + }, + { + "epoch": 0.847547012698633, + "grad_norm": 2.434333562850952, + "learning_rate": 0.00018314657103626566, + "loss": 0.93, + "step": 12214 + }, + { + "epoch": 0.8476164041357297, + "grad_norm": 2.022622585296631, + "learning_rate": 0.0001831434493367091, + "loss": 0.753, + "step": 12215 + }, + { + "epoch": 0.8476857955728263, + "grad_norm": 2.043174982070923, + "learning_rate": 0.00018314032737467827, + "loss": 0.9754, + "step": 12216 + }, + { + "epoch": 0.847755187009923, + "grad_norm": 2.3437182903289795, + "learning_rate": 0.0001831372051501831, + "loss": 0.7038, + "step": 12217 + }, + { + "epoch": 0.8478245784470196, + "grad_norm": 2.4676313400268555, + "learning_rate": 0.00018313408266323338, + "loss": 0.8977, + "step": 12218 + }, + { + "epoch": 0.8478939698841162, + "grad_norm": 2.3958778381347656, + "learning_rate": 0.00018313095991383903, + "loss": 0.8836, + "step": 12219 + }, + { + "epoch": 0.847963361321213, + "grad_norm": 2.16038179397583, + "learning_rate": 0.00018312783690200988, + "loss": 0.7228, + "step": 12220 + }, + { + "epoch": 0.8480327527583096, + "grad_norm": 1.9877111911773682, + "learning_rate": 0.0001831247136277558, + "loss": 0.7132, + "step": 12221 + }, + { + "epoch": 0.8481021441954063, + "grad_norm": 1.839174509048462, + "learning_rate": 0.00018312159009108663, + "loss": 0.7996, + "step": 12222 + }, + { + "epoch": 0.8481715356325029, + "grad_norm": 1.8251521587371826, + "learning_rate": 0.0001831184662920123, + "loss": 0.5613, + "step": 12223 + }, + { + "epoch": 0.8482409270695996, + "grad_norm": 2.039602756500244, + "learning_rate": 0.00018311534223054254, + "loss": 0.7297, + "step": 12224 + }, + { + "epoch": 0.8483103185066962, + "grad_norm": 1.7714065313339233, + "learning_rate": 0.0001831122179066873, + "loss": 0.8975, + "step": 12225 + }, + { + "epoch": 0.848379709943793, + "grad_norm": 1.8792237043380737, + "learning_rate": 0.00018310909332045644, + "loss": 0.5832, + "step": 12226 + }, + { + "epoch": 0.8484491013808896, + "grad_norm": 1.995628833770752, + "learning_rate": 0.00018310596847185987, + "loss": 0.9337, + "step": 12227 + }, + { + "epoch": 0.8485184928179863, + "grad_norm": 2.2283031940460205, + "learning_rate": 0.00018310284336090732, + "loss": 0.9362, + "step": 12228 + }, + { + "epoch": 0.8485878842550829, + "grad_norm": 1.7646077871322632, + "learning_rate": 0.00018309971798760878, + "loss": 0.6107, + "step": 12229 + }, + { + "epoch": 0.8486572756921796, + "grad_norm": 2.480487585067749, + "learning_rate": 0.00018309659235197408, + "loss": 0.8093, + "step": 12230 + }, + { + "epoch": 0.8487266671292762, + "grad_norm": 2.2165091037750244, + "learning_rate": 0.00018309346645401303, + "loss": 0.708, + "step": 12231 + }, + { + "epoch": 0.8487960585663729, + "grad_norm": 1.9619008302688599, + "learning_rate": 0.00018309034029373558, + "loss": 0.7348, + "step": 12232 + }, + { + "epoch": 0.8488654500034696, + "grad_norm": 2.7569143772125244, + "learning_rate": 0.00018308721387115157, + "loss": 0.8107, + "step": 12233 + }, + { + "epoch": 0.8489348414405662, + "grad_norm": 2.62870454788208, + "learning_rate": 0.00018308408718627087, + "loss": 0.9357, + "step": 12234 + }, + { + "epoch": 0.8490042328776629, + "grad_norm": 2.3450002670288086, + "learning_rate": 0.00018308096023910334, + "loss": 1.0423, + "step": 12235 + }, + { + "epoch": 0.8490736243147595, + "grad_norm": 1.7762213945388794, + "learning_rate": 0.00018307783302965884, + "loss": 0.6605, + "step": 12236 + }, + { + "epoch": 0.8491430157518562, + "grad_norm": 2.4119038581848145, + "learning_rate": 0.00018307470555794727, + "loss": 0.9711, + "step": 12237 + }, + { + "epoch": 0.8492124071889529, + "grad_norm": 2.515894889831543, + "learning_rate": 0.00018307157782397848, + "loss": 0.8711, + "step": 12238 + }, + { + "epoch": 0.8492817986260496, + "grad_norm": 2.0793046951293945, + "learning_rate": 0.00018306844982776238, + "loss": 0.8639, + "step": 12239 + }, + { + "epoch": 0.8493511900631462, + "grad_norm": 1.909494400024414, + "learning_rate": 0.00018306532156930883, + "loss": 0.6776, + "step": 12240 + }, + { + "epoch": 0.8494205815002429, + "grad_norm": 2.411257266998291, + "learning_rate": 0.00018306219304862768, + "loss": 0.8179, + "step": 12241 + }, + { + "epoch": 0.8494899729373395, + "grad_norm": 2.055288314819336, + "learning_rate": 0.00018305906426572882, + "loss": 0.917, + "step": 12242 + }, + { + "epoch": 0.8495593643744362, + "grad_norm": 2.5753190517425537, + "learning_rate": 0.0001830559352206221, + "loss": 1.2084, + "step": 12243 + }, + { + "epoch": 0.8496287558115329, + "grad_norm": 2.422119617462158, + "learning_rate": 0.0001830528059133175, + "loss": 0.7059, + "step": 12244 + }, + { + "epoch": 0.8496981472486295, + "grad_norm": 1.9711970090866089, + "learning_rate": 0.0001830496763438248, + "loss": 0.5517, + "step": 12245 + }, + { + "epoch": 0.8497675386857262, + "grad_norm": 2.239150285720825, + "learning_rate": 0.00018304654651215389, + "loss": 0.8138, + "step": 12246 + }, + { + "epoch": 0.8498369301228228, + "grad_norm": 2.3812787532806396, + "learning_rate": 0.00018304341641831465, + "loss": 0.866, + "step": 12247 + }, + { + "epoch": 0.8499063215599195, + "grad_norm": 2.5496442317962646, + "learning_rate": 0.00018304028606231705, + "loss": 0.9857, + "step": 12248 + }, + { + "epoch": 0.8499757129970161, + "grad_norm": 2.3432345390319824, + "learning_rate": 0.00018303715544417084, + "loss": 0.7146, + "step": 12249 + }, + { + "epoch": 0.8500451044341129, + "grad_norm": 2.0266993045806885, + "learning_rate": 0.000183034024563886, + "loss": 0.901, + "step": 12250 + }, + { + "epoch": 0.8501144958712095, + "grad_norm": 1.884713888168335, + "learning_rate": 0.00018303089342147234, + "loss": 0.7292, + "step": 12251 + }, + { + "epoch": 0.8501838873083062, + "grad_norm": 2.2685115337371826, + "learning_rate": 0.0001830277620169398, + "loss": 0.8077, + "step": 12252 + }, + { + "epoch": 0.8502532787454028, + "grad_norm": 3.462045431137085, + "learning_rate": 0.00018302463035029827, + "loss": 0.8834, + "step": 12253 + }, + { + "epoch": 0.8503226701824995, + "grad_norm": 1.8807885646820068, + "learning_rate": 0.0001830214984215576, + "loss": 0.6095, + "step": 12254 + }, + { + "epoch": 0.8503920616195961, + "grad_norm": 1.9251676797866821, + "learning_rate": 0.00018301836623072764, + "loss": 0.6403, + "step": 12255 + }, + { + "epoch": 0.8504614530566929, + "grad_norm": 2.765245199203491, + "learning_rate": 0.0001830152337778184, + "loss": 0.7589, + "step": 12256 + }, + { + "epoch": 0.8505308444937895, + "grad_norm": 2.179999589920044, + "learning_rate": 0.00018301210106283969, + "loss": 0.8723, + "step": 12257 + }, + { + "epoch": 0.8506002359308862, + "grad_norm": 2.641509532928467, + "learning_rate": 0.00018300896808580138, + "loss": 1.2479, + "step": 12258 + }, + { + "epoch": 0.8506696273679828, + "grad_norm": 2.617835521697998, + "learning_rate": 0.0001830058348467134, + "loss": 0.7996, + "step": 12259 + }, + { + "epoch": 0.8507390188050794, + "grad_norm": 3.3382463455200195, + "learning_rate": 0.00018300270134558565, + "loss": 0.9183, + "step": 12260 + }, + { + "epoch": 0.8508084102421761, + "grad_norm": 2.2759971618652344, + "learning_rate": 0.00018299956758242798, + "loss": 0.8816, + "step": 12261 + }, + { + "epoch": 0.8508778016792727, + "grad_norm": 2.013439178466797, + "learning_rate": 0.00018299643355725032, + "loss": 0.9235, + "step": 12262 + }, + { + "epoch": 0.8509471931163695, + "grad_norm": 1.9291085004806519, + "learning_rate": 0.00018299329927006254, + "loss": 0.8475, + "step": 12263 + }, + { + "epoch": 0.8510165845534661, + "grad_norm": 1.8503414392471313, + "learning_rate": 0.00018299016472087454, + "loss": 0.6798, + "step": 12264 + }, + { + "epoch": 0.8510859759905628, + "grad_norm": 1.7068527936935425, + "learning_rate": 0.0001829870299096962, + "loss": 0.7956, + "step": 12265 + }, + { + "epoch": 0.8511553674276594, + "grad_norm": 2.156170129776001, + "learning_rate": 0.00018298389483653745, + "loss": 0.9878, + "step": 12266 + }, + { + "epoch": 0.8512247588647561, + "grad_norm": 1.7333858013153076, + "learning_rate": 0.00018298075950140818, + "loss": 0.7812, + "step": 12267 + }, + { + "epoch": 0.8512941503018527, + "grad_norm": 1.9168561697006226, + "learning_rate": 0.00018297762390431828, + "loss": 0.7495, + "step": 12268 + }, + { + "epoch": 0.8513635417389495, + "grad_norm": 1.8841636180877686, + "learning_rate": 0.00018297448804527764, + "loss": 0.5489, + "step": 12269 + }, + { + "epoch": 0.8514329331760461, + "grad_norm": 2.4900736808776855, + "learning_rate": 0.00018297135192429616, + "loss": 0.8083, + "step": 12270 + }, + { + "epoch": 0.8515023246131428, + "grad_norm": 2.0968620777130127, + "learning_rate": 0.0001829682155413838, + "loss": 0.8205, + "step": 12271 + }, + { + "epoch": 0.8515717160502394, + "grad_norm": 1.9961893558502197, + "learning_rate": 0.00018296507889655034, + "loss": 0.8308, + "step": 12272 + }, + { + "epoch": 0.851641107487336, + "grad_norm": 1.7516653537750244, + "learning_rate": 0.00018296194198980578, + "loss": 0.7092, + "step": 12273 + }, + { + "epoch": 0.8517104989244327, + "grad_norm": 2.0980374813079834, + "learning_rate": 0.00018295880482115999, + "loss": 0.6592, + "step": 12274 + }, + { + "epoch": 0.8517798903615293, + "grad_norm": 2.2644052505493164, + "learning_rate": 0.0001829556673906229, + "loss": 0.7628, + "step": 12275 + }, + { + "epoch": 0.8518492817986261, + "grad_norm": 1.8320670127868652, + "learning_rate": 0.00018295252969820439, + "loss": 0.7453, + "step": 12276 + }, + { + "epoch": 0.8519186732357227, + "grad_norm": 2.245431423187256, + "learning_rate": 0.00018294939174391437, + "loss": 0.9095, + "step": 12277 + }, + { + "epoch": 0.8519880646728194, + "grad_norm": 2.4534761905670166, + "learning_rate": 0.00018294625352776271, + "loss": 0.928, + "step": 12278 + }, + { + "epoch": 0.852057456109916, + "grad_norm": 2.5276825428009033, + "learning_rate": 0.00018294311504975935, + "loss": 0.6627, + "step": 12279 + }, + { + "epoch": 0.8521268475470127, + "grad_norm": 2.1514763832092285, + "learning_rate": 0.00018293997630991425, + "loss": 0.6376, + "step": 12280 + }, + { + "epoch": 0.8521962389841093, + "grad_norm": 2.2114150524139404, + "learning_rate": 0.00018293683730823723, + "loss": 0.7509, + "step": 12281 + }, + { + "epoch": 0.8522656304212061, + "grad_norm": 1.8306039571762085, + "learning_rate": 0.00018293369804473823, + "loss": 0.6632, + "step": 12282 + }, + { + "epoch": 0.8523350218583027, + "grad_norm": 2.4134390354156494, + "learning_rate": 0.0001829305585194272, + "loss": 1.0669, + "step": 12283 + }, + { + "epoch": 0.8524044132953994, + "grad_norm": 2.0520427227020264, + "learning_rate": 0.000182927418732314, + "loss": 0.738, + "step": 12284 + }, + { + "epoch": 0.852473804732496, + "grad_norm": 1.9885728359222412, + "learning_rate": 0.00018292427868340854, + "loss": 0.6541, + "step": 12285 + }, + { + "epoch": 0.8525431961695926, + "grad_norm": 2.2778799533843994, + "learning_rate": 0.00018292113837272079, + "loss": 0.7479, + "step": 12286 + }, + { + "epoch": 0.8526125876066893, + "grad_norm": 2.7351982593536377, + "learning_rate": 0.0001829179978002606, + "loss": 0.7342, + "step": 12287 + }, + { + "epoch": 0.852681979043786, + "grad_norm": 2.1723263263702393, + "learning_rate": 0.00018291485696603793, + "loss": 1.0222, + "step": 12288 + }, + { + "epoch": 0.8527513704808827, + "grad_norm": 2.0349056720733643, + "learning_rate": 0.00018291171587006265, + "loss": 0.9225, + "step": 12289 + }, + { + "epoch": 0.8528207619179793, + "grad_norm": 2.0206236839294434, + "learning_rate": 0.00018290857451234475, + "loss": 0.8762, + "step": 12290 + }, + { + "epoch": 0.852890153355076, + "grad_norm": 2.027702808380127, + "learning_rate": 0.00018290543289289408, + "loss": 0.616, + "step": 12291 + }, + { + "epoch": 0.8529595447921726, + "grad_norm": 2.4243505001068115, + "learning_rate": 0.00018290229101172053, + "loss": 0.8712, + "step": 12292 + }, + { + "epoch": 0.8530289362292693, + "grad_norm": 2.3240814208984375, + "learning_rate": 0.00018289914886883413, + "loss": 0.8917, + "step": 12293 + }, + { + "epoch": 0.853098327666366, + "grad_norm": 1.9559271335601807, + "learning_rate": 0.00018289600646424469, + "loss": 0.902, + "step": 12294 + }, + { + "epoch": 0.8531677191034627, + "grad_norm": 2.232415199279785, + "learning_rate": 0.0001828928637979622, + "loss": 0.7929, + "step": 12295 + }, + { + "epoch": 0.8532371105405593, + "grad_norm": 2.5551459789276123, + "learning_rate": 0.00018288972086999655, + "loss": 0.7342, + "step": 12296 + }, + { + "epoch": 0.853306501977656, + "grad_norm": 2.174248218536377, + "learning_rate": 0.00018288657768035768, + "loss": 0.8695, + "step": 12297 + }, + { + "epoch": 0.8533758934147526, + "grad_norm": 2.2167985439300537, + "learning_rate": 0.00018288343422905547, + "loss": 0.8631, + "step": 12298 + }, + { + "epoch": 0.8534452848518492, + "grad_norm": 2.707629442214966, + "learning_rate": 0.00018288029051609993, + "loss": 0.8054, + "step": 12299 + }, + { + "epoch": 0.853514676288946, + "grad_norm": 2.1519064903259277, + "learning_rate": 0.00018287714654150088, + "loss": 0.8869, + "step": 12300 + }, + { + "epoch": 0.8535840677260426, + "grad_norm": 2.6276323795318604, + "learning_rate": 0.0001828740023052683, + "loss": 0.7245, + "step": 12301 + }, + { + "epoch": 0.8536534591631393, + "grad_norm": 2.7145042419433594, + "learning_rate": 0.00018287085780741214, + "loss": 0.9708, + "step": 12302 + }, + { + "epoch": 0.8537228506002359, + "grad_norm": 2.1752374172210693, + "learning_rate": 0.00018286771304794225, + "loss": 0.8105, + "step": 12303 + }, + { + "epoch": 0.8537922420373326, + "grad_norm": 2.679164171218872, + "learning_rate": 0.00018286456802686863, + "loss": 0.9462, + "step": 12304 + }, + { + "epoch": 0.8538616334744292, + "grad_norm": 1.829775094985962, + "learning_rate": 0.0001828614227442012, + "loss": 0.6206, + "step": 12305 + }, + { + "epoch": 0.853931024911526, + "grad_norm": 2.0835824012756348, + "learning_rate": 0.0001828582771999498, + "loss": 0.6472, + "step": 12306 + }, + { + "epoch": 0.8540004163486226, + "grad_norm": 1.8241890668869019, + "learning_rate": 0.0001828551313941245, + "loss": 0.7747, + "step": 12307 + }, + { + "epoch": 0.8540698077857193, + "grad_norm": 1.8798081874847412, + "learning_rate": 0.00018285198532673515, + "loss": 0.8426, + "step": 12308 + }, + { + "epoch": 0.8541391992228159, + "grad_norm": 2.876415252685547, + "learning_rate": 0.0001828488389977917, + "loss": 0.9901, + "step": 12309 + }, + { + "epoch": 0.8542085906599126, + "grad_norm": 2.216580629348755, + "learning_rate": 0.00018284569240730407, + "loss": 0.9323, + "step": 12310 + }, + { + "epoch": 0.8542779820970092, + "grad_norm": 1.9595345258712769, + "learning_rate": 0.0001828425455552822, + "loss": 0.8351, + "step": 12311 + }, + { + "epoch": 0.8543473735341058, + "grad_norm": 2.2179694175720215, + "learning_rate": 0.00018283939844173603, + "loss": 0.9886, + "step": 12312 + }, + { + "epoch": 0.8544167649712026, + "grad_norm": 2.3263254165649414, + "learning_rate": 0.0001828362510666755, + "loss": 1.0079, + "step": 12313 + }, + { + "epoch": 0.8544861564082992, + "grad_norm": 2.5436623096466064, + "learning_rate": 0.0001828331034301105, + "loss": 0.7369, + "step": 12314 + }, + { + "epoch": 0.8545555478453959, + "grad_norm": 2.0572354793548584, + "learning_rate": 0.00018282995553205105, + "loss": 1.0238, + "step": 12315 + }, + { + "epoch": 0.8546249392824925, + "grad_norm": 1.8487049341201782, + "learning_rate": 0.000182826807372507, + "loss": 0.656, + "step": 12316 + }, + { + "epoch": 0.8546943307195892, + "grad_norm": 2.1820929050445557, + "learning_rate": 0.00018282365895148835, + "loss": 0.7899, + "step": 12317 + }, + { + "epoch": 0.8547637221566858, + "grad_norm": 2.1196324825286865, + "learning_rate": 0.000182820510269005, + "loss": 0.7478, + "step": 12318 + }, + { + "epoch": 0.8548331135937826, + "grad_norm": 2.0465574264526367, + "learning_rate": 0.00018281736132506692, + "loss": 0.8646, + "step": 12319 + }, + { + "epoch": 0.8549025050308792, + "grad_norm": 2.886453866958618, + "learning_rate": 0.00018281421211968407, + "loss": 1.0259, + "step": 12320 + }, + { + "epoch": 0.8549718964679759, + "grad_norm": 2.132868528366089, + "learning_rate": 0.00018281106265286633, + "loss": 0.9993, + "step": 12321 + }, + { + "epoch": 0.8550412879050725, + "grad_norm": 2.3773884773254395, + "learning_rate": 0.00018280791292462367, + "loss": 0.8507, + "step": 12322 + }, + { + "epoch": 0.8551106793421692, + "grad_norm": 2.154045581817627, + "learning_rate": 0.00018280476293496602, + "loss": 0.5048, + "step": 12323 + }, + { + "epoch": 0.8551800707792658, + "grad_norm": 1.9593111276626587, + "learning_rate": 0.00018280161268390335, + "loss": 0.7643, + "step": 12324 + }, + { + "epoch": 0.8552494622163626, + "grad_norm": 2.259516716003418, + "learning_rate": 0.00018279846217144564, + "loss": 0.7304, + "step": 12325 + }, + { + "epoch": 0.8553188536534592, + "grad_norm": 2.4130799770355225, + "learning_rate": 0.00018279531139760272, + "loss": 0.7195, + "step": 12326 + }, + { + "epoch": 0.8553882450905558, + "grad_norm": 1.965649962425232, + "learning_rate": 0.00018279216036238465, + "loss": 0.8104, + "step": 12327 + }, + { + "epoch": 0.8554576365276525, + "grad_norm": 1.9664171934127808, + "learning_rate": 0.00018278900906580135, + "loss": 0.658, + "step": 12328 + }, + { + "epoch": 0.8555270279647491, + "grad_norm": 2.7413699626922607, + "learning_rate": 0.00018278585750786274, + "loss": 0.7561, + "step": 12329 + }, + { + "epoch": 0.8555964194018458, + "grad_norm": 2.247117757797241, + "learning_rate": 0.0001827827056885788, + "loss": 0.865, + "step": 12330 + }, + { + "epoch": 0.8556658108389424, + "grad_norm": 2.0809433460235596, + "learning_rate": 0.0001827795536079594, + "loss": 0.6712, + "step": 12331 + }, + { + "epoch": 0.8557352022760392, + "grad_norm": 2.566035270690918, + "learning_rate": 0.0001827764012660146, + "loss": 1.09, + "step": 12332 + }, + { + "epoch": 0.8558045937131358, + "grad_norm": 1.8363282680511475, + "learning_rate": 0.0001827732486627543, + "loss": 0.8635, + "step": 12333 + }, + { + "epoch": 0.8558739851502325, + "grad_norm": 1.7440946102142334, + "learning_rate": 0.00018277009579818846, + "loss": 0.8706, + "step": 12334 + }, + { + "epoch": 0.8559433765873291, + "grad_norm": 2.090017318725586, + "learning_rate": 0.00018276694267232703, + "loss": 0.7594, + "step": 12335 + }, + { + "epoch": 0.8560127680244258, + "grad_norm": 2.59968900680542, + "learning_rate": 0.00018276378928517995, + "loss": 0.8783, + "step": 12336 + }, + { + "epoch": 0.8560821594615224, + "grad_norm": 2.346020460128784, + "learning_rate": 0.0001827606356367572, + "loss": 0.9946, + "step": 12337 + }, + { + "epoch": 0.8561515508986192, + "grad_norm": 1.8599146604537964, + "learning_rate": 0.00018275748172706874, + "loss": 0.7401, + "step": 12338 + }, + { + "epoch": 0.8562209423357158, + "grad_norm": 2.0513815879821777, + "learning_rate": 0.00018275432755612448, + "loss": 0.8999, + "step": 12339 + }, + { + "epoch": 0.8562903337728124, + "grad_norm": 2.041127920150757, + "learning_rate": 0.00018275117312393444, + "loss": 0.6294, + "step": 12340 + }, + { + "epoch": 0.8563597252099091, + "grad_norm": 2.58339786529541, + "learning_rate": 0.00018274801843050852, + "loss": 0.8417, + "step": 12341 + }, + { + "epoch": 0.8564291166470057, + "grad_norm": 2.2318429946899414, + "learning_rate": 0.00018274486347585672, + "loss": 0.8077, + "step": 12342 + }, + { + "epoch": 0.8564985080841024, + "grad_norm": 1.9122873544692993, + "learning_rate": 0.000182741708259989, + "loss": 0.7767, + "step": 12343 + }, + { + "epoch": 0.856567899521199, + "grad_norm": 4.616090774536133, + "learning_rate": 0.0001827385527829153, + "loss": 0.7414, + "step": 12344 + }, + { + "epoch": 0.8566372909582958, + "grad_norm": 1.9682379961013794, + "learning_rate": 0.00018273539704464557, + "loss": 0.8379, + "step": 12345 + }, + { + "epoch": 0.8567066823953924, + "grad_norm": 1.9688186645507812, + "learning_rate": 0.0001827322410451898, + "loss": 0.7874, + "step": 12346 + }, + { + "epoch": 0.8567760738324891, + "grad_norm": 2.89575457572937, + "learning_rate": 0.00018272908478455797, + "loss": 0.6817, + "step": 12347 + }, + { + "epoch": 0.8568454652695857, + "grad_norm": 1.9028838872909546, + "learning_rate": 0.00018272592826276, + "loss": 0.8205, + "step": 12348 + }, + { + "epoch": 0.8569148567066824, + "grad_norm": 2.3242697715759277, + "learning_rate": 0.00018272277147980584, + "loss": 0.6826, + "step": 12349 + }, + { + "epoch": 0.856984248143779, + "grad_norm": 2.5542144775390625, + "learning_rate": 0.00018271961443570552, + "loss": 0.9002, + "step": 12350 + }, + { + "epoch": 0.8570536395808758, + "grad_norm": 2.629462957382202, + "learning_rate": 0.00018271645713046895, + "loss": 0.9031, + "step": 12351 + }, + { + "epoch": 0.8571230310179724, + "grad_norm": 2.0775344371795654, + "learning_rate": 0.00018271329956410614, + "loss": 0.6996, + "step": 12352 + }, + { + "epoch": 0.857192422455069, + "grad_norm": 2.441927433013916, + "learning_rate": 0.00018271014173662706, + "loss": 0.8414, + "step": 12353 + }, + { + "epoch": 0.8572618138921657, + "grad_norm": 1.8394662141799927, + "learning_rate": 0.00018270698364804164, + "loss": 0.877, + "step": 12354 + }, + { + "epoch": 0.8573312053292623, + "grad_norm": 2.0813181400299072, + "learning_rate": 0.00018270382529835986, + "loss": 0.678, + "step": 12355 + }, + { + "epoch": 0.857400596766359, + "grad_norm": 2.3403549194335938, + "learning_rate": 0.0001827006666875917, + "loss": 0.7614, + "step": 12356 + }, + { + "epoch": 0.8574699882034557, + "grad_norm": 2.3138341903686523, + "learning_rate": 0.00018269750781574714, + "loss": 0.6879, + "step": 12357 + }, + { + "epoch": 0.8575393796405524, + "grad_norm": 1.8498461246490479, + "learning_rate": 0.00018269434868283614, + "loss": 0.7772, + "step": 12358 + }, + { + "epoch": 0.857608771077649, + "grad_norm": 1.9224475622177124, + "learning_rate": 0.00018269118928886867, + "loss": 0.6951, + "step": 12359 + }, + { + "epoch": 0.8576781625147457, + "grad_norm": 1.7746596336364746, + "learning_rate": 0.00018268802963385472, + "loss": 0.6797, + "step": 12360 + }, + { + "epoch": 0.8577475539518423, + "grad_norm": 1.928810954093933, + "learning_rate": 0.00018268486971780425, + "loss": 0.6931, + "step": 12361 + }, + { + "epoch": 0.857816945388939, + "grad_norm": 2.410553455352783, + "learning_rate": 0.00018268170954072725, + "loss": 0.7758, + "step": 12362 + }, + { + "epoch": 0.8578863368260357, + "grad_norm": 2.3539321422576904, + "learning_rate": 0.00018267854910263367, + "loss": 0.9238, + "step": 12363 + }, + { + "epoch": 0.8579557282631324, + "grad_norm": 1.5459800958633423, + "learning_rate": 0.00018267538840353352, + "loss": 0.505, + "step": 12364 + }, + { + "epoch": 0.858025119700229, + "grad_norm": 2.2604660987854004, + "learning_rate": 0.00018267222744343674, + "loss": 0.7167, + "step": 12365 + }, + { + "epoch": 0.8580945111373256, + "grad_norm": 1.8674246072769165, + "learning_rate": 0.00018266906622235333, + "loss": 0.5866, + "step": 12366 + }, + { + "epoch": 0.8581639025744223, + "grad_norm": 2.194047451019287, + "learning_rate": 0.00018266590474029328, + "loss": 0.6288, + "step": 12367 + }, + { + "epoch": 0.8582332940115189, + "grad_norm": 1.8637058734893799, + "learning_rate": 0.00018266274299726654, + "loss": 0.7438, + "step": 12368 + }, + { + "epoch": 0.8583026854486157, + "grad_norm": 2.372157096862793, + "learning_rate": 0.00018265958099328313, + "loss": 1.0594, + "step": 12369 + }, + { + "epoch": 0.8583720768857123, + "grad_norm": 2.5850775241851807, + "learning_rate": 0.00018265641872835302, + "loss": 1.1743, + "step": 12370 + }, + { + "epoch": 0.858441468322809, + "grad_norm": 1.8199467658996582, + "learning_rate": 0.00018265325620248616, + "loss": 0.5949, + "step": 12371 + }, + { + "epoch": 0.8585108597599056, + "grad_norm": 2.3245601654052734, + "learning_rate": 0.0001826500934156926, + "loss": 0.6444, + "step": 12372 + }, + { + "epoch": 0.8585802511970023, + "grad_norm": 2.1476259231567383, + "learning_rate": 0.00018264693036798227, + "loss": 0.8243, + "step": 12373 + }, + { + "epoch": 0.8586496426340989, + "grad_norm": 2.3837695121765137, + "learning_rate": 0.00018264376705936516, + "loss": 0.821, + "step": 12374 + }, + { + "epoch": 0.8587190340711957, + "grad_norm": 2.4701948165893555, + "learning_rate": 0.00018264060348985127, + "loss": 0.9496, + "step": 12375 + }, + { + "epoch": 0.8587884255082923, + "grad_norm": 2.349313259124756, + "learning_rate": 0.00018263743965945056, + "loss": 0.7088, + "step": 12376 + }, + { + "epoch": 0.858857816945389, + "grad_norm": 2.232804298400879, + "learning_rate": 0.00018263427556817306, + "loss": 0.7236, + "step": 12377 + }, + { + "epoch": 0.8589272083824856, + "grad_norm": 2.6192760467529297, + "learning_rate": 0.00018263111121602873, + "loss": 1.0441, + "step": 12378 + }, + { + "epoch": 0.8589965998195823, + "grad_norm": 4.770200252532959, + "learning_rate": 0.00018262794660302758, + "loss": 0.8456, + "step": 12379 + }, + { + "epoch": 0.8590659912566789, + "grad_norm": 1.8790209293365479, + "learning_rate": 0.00018262478172917962, + "loss": 0.7159, + "step": 12380 + }, + { + "epoch": 0.8591353826937755, + "grad_norm": 3.029266595840454, + "learning_rate": 0.00018262161659449475, + "loss": 1.0166, + "step": 12381 + }, + { + "epoch": 0.8592047741308723, + "grad_norm": 2.6706180572509766, + "learning_rate": 0.00018261845119898308, + "loss": 0.9077, + "step": 12382 + }, + { + "epoch": 0.8592741655679689, + "grad_norm": 1.9060462713241577, + "learning_rate": 0.00018261528554265448, + "loss": 0.7626, + "step": 12383 + }, + { + "epoch": 0.8593435570050656, + "grad_norm": 2.018472671508789, + "learning_rate": 0.00018261211962551906, + "loss": 0.8278, + "step": 12384 + }, + { + "epoch": 0.8594129484421622, + "grad_norm": 1.9513072967529297, + "learning_rate": 0.00018260895344758674, + "loss": 0.8303, + "step": 12385 + }, + { + "epoch": 0.8594823398792589, + "grad_norm": 2.0382282733917236, + "learning_rate": 0.00018260578700886754, + "loss": 0.8156, + "step": 12386 + }, + { + "epoch": 0.8595517313163555, + "grad_norm": 2.3178958892822266, + "learning_rate": 0.00018260262030937147, + "loss": 0.9325, + "step": 12387 + }, + { + "epoch": 0.8596211227534523, + "grad_norm": 3.5132336616516113, + "learning_rate": 0.00018259945334910848, + "loss": 0.7827, + "step": 12388 + }, + { + "epoch": 0.8596905141905489, + "grad_norm": 2.480677843093872, + "learning_rate": 0.00018259628612808863, + "loss": 0.7157, + "step": 12389 + }, + { + "epoch": 0.8597599056276456, + "grad_norm": 2.1798272132873535, + "learning_rate": 0.0001825931186463219, + "loss": 0.9289, + "step": 12390 + }, + { + "epoch": 0.8598292970647422, + "grad_norm": 2.4952805042266846, + "learning_rate": 0.00018258995090381826, + "loss": 1.0797, + "step": 12391 + }, + { + "epoch": 0.8598986885018389, + "grad_norm": 1.770788311958313, + "learning_rate": 0.0001825867829005877, + "loss": 0.6491, + "step": 12392 + }, + { + "epoch": 0.8599680799389355, + "grad_norm": 1.9237587451934814, + "learning_rate": 0.0001825836146366403, + "loss": 0.827, + "step": 12393 + }, + { + "epoch": 0.8600374713760321, + "grad_norm": 2.1737899780273438, + "learning_rate": 0.00018258044611198595, + "loss": 0.64, + "step": 12394 + }, + { + "epoch": 0.8601068628131289, + "grad_norm": 2.8951258659362793, + "learning_rate": 0.00018257727732663473, + "loss": 1.1439, + "step": 12395 + }, + { + "epoch": 0.8601762542502255, + "grad_norm": 1.5904839038848877, + "learning_rate": 0.00018257410828059664, + "loss": 0.5885, + "step": 12396 + }, + { + "epoch": 0.8602456456873222, + "grad_norm": 2.6505699157714844, + "learning_rate": 0.00018257093897388168, + "loss": 0.828, + "step": 12397 + }, + { + "epoch": 0.8603150371244188, + "grad_norm": 2.282970428466797, + "learning_rate": 0.00018256776940649982, + "loss": 0.7925, + "step": 12398 + }, + { + "epoch": 0.8603844285615155, + "grad_norm": 1.997707724571228, + "learning_rate": 0.00018256459957846112, + "loss": 0.8103, + "step": 12399 + }, + { + "epoch": 0.8604538199986121, + "grad_norm": 2.0447804927825928, + "learning_rate": 0.00018256142948977556, + "loss": 0.8571, + "step": 12400 + }, + { + "epoch": 0.8605232114357089, + "grad_norm": 2.8830175399780273, + "learning_rate": 0.00018255825914045314, + "loss": 0.8136, + "step": 12401 + }, + { + "epoch": 0.8605926028728055, + "grad_norm": 1.825191617012024, + "learning_rate": 0.00018255508853050385, + "loss": 0.6329, + "step": 12402 + }, + { + "epoch": 0.8606619943099022, + "grad_norm": 2.0552937984466553, + "learning_rate": 0.00018255191765993774, + "loss": 1.036, + "step": 12403 + }, + { + "epoch": 0.8607313857469988, + "grad_norm": 1.9402774572372437, + "learning_rate": 0.0001825487465287648, + "loss": 0.5807, + "step": 12404 + }, + { + "epoch": 0.8608007771840955, + "grad_norm": 2.286113739013672, + "learning_rate": 0.00018254557513699507, + "loss": 0.7019, + "step": 12405 + }, + { + "epoch": 0.8608701686211921, + "grad_norm": 2.583393096923828, + "learning_rate": 0.0001825424034846385, + "loss": 1.0318, + "step": 12406 + }, + { + "epoch": 0.8609395600582888, + "grad_norm": 2.0089216232299805, + "learning_rate": 0.00018253923157170513, + "loss": 0.6585, + "step": 12407 + }, + { + "epoch": 0.8610089514953855, + "grad_norm": 2.391119956970215, + "learning_rate": 0.000182536059398205, + "loss": 0.7757, + "step": 12408 + }, + { + "epoch": 0.8610783429324821, + "grad_norm": 3.780834197998047, + "learning_rate": 0.00018253288696414813, + "loss": 0.8862, + "step": 12409 + }, + { + "epoch": 0.8611477343695788, + "grad_norm": 3.04140043258667, + "learning_rate": 0.00018252971426954448, + "loss": 0.9005, + "step": 12410 + }, + { + "epoch": 0.8612171258066754, + "grad_norm": 2.6076486110687256, + "learning_rate": 0.00018252654131440413, + "loss": 0.7544, + "step": 12411 + }, + { + "epoch": 0.8612865172437721, + "grad_norm": 2.36083984375, + "learning_rate": 0.00018252336809873704, + "loss": 0.883, + "step": 12412 + }, + { + "epoch": 0.8613559086808688, + "grad_norm": 1.977431297302246, + "learning_rate": 0.00018252019462255324, + "loss": 0.6284, + "step": 12413 + }, + { + "epoch": 0.8614253001179655, + "grad_norm": 2.181727409362793, + "learning_rate": 0.00018251702088586277, + "loss": 0.834, + "step": 12414 + }, + { + "epoch": 0.8614946915550621, + "grad_norm": 3.0725815296173096, + "learning_rate": 0.00018251384688867565, + "loss": 0.9857, + "step": 12415 + }, + { + "epoch": 0.8615640829921588, + "grad_norm": 2.3858768939971924, + "learning_rate": 0.0001825106726310019, + "loss": 0.7884, + "step": 12416 + }, + { + "epoch": 0.8616334744292554, + "grad_norm": 2.4648258686065674, + "learning_rate": 0.00018250749811285152, + "loss": 0.7898, + "step": 12417 + }, + { + "epoch": 0.8617028658663521, + "grad_norm": 2.1149141788482666, + "learning_rate": 0.00018250432333423452, + "loss": 0.7922, + "step": 12418 + }, + { + "epoch": 0.8617722573034488, + "grad_norm": 2.280163049697876, + "learning_rate": 0.00018250114829516096, + "loss": 0.7376, + "step": 12419 + }, + { + "epoch": 0.8618416487405454, + "grad_norm": 3.1312530040740967, + "learning_rate": 0.00018249797299564084, + "loss": 0.7168, + "step": 12420 + }, + { + "epoch": 0.8619110401776421, + "grad_norm": 2.5677480697631836, + "learning_rate": 0.0001824947974356842, + "loss": 0.9027, + "step": 12421 + }, + { + "epoch": 0.8619804316147387, + "grad_norm": 2.836932897567749, + "learning_rate": 0.00018249162161530105, + "loss": 0.7877, + "step": 12422 + }, + { + "epoch": 0.8620498230518354, + "grad_norm": 2.1796209812164307, + "learning_rate": 0.00018248844553450143, + "loss": 0.6779, + "step": 12423 + }, + { + "epoch": 0.862119214488932, + "grad_norm": 1.745506763458252, + "learning_rate": 0.00018248526919329534, + "loss": 0.7028, + "step": 12424 + }, + { + "epoch": 0.8621886059260288, + "grad_norm": 2.022742509841919, + "learning_rate": 0.00018248209259169284, + "loss": 0.8734, + "step": 12425 + }, + { + "epoch": 0.8622579973631254, + "grad_norm": 2.7767114639282227, + "learning_rate": 0.00018247891572970395, + "loss": 0.7723, + "step": 12426 + }, + { + "epoch": 0.8623273888002221, + "grad_norm": 1.8338243961334229, + "learning_rate": 0.00018247573860733865, + "loss": 0.7569, + "step": 12427 + }, + { + "epoch": 0.8623967802373187, + "grad_norm": 1.7041938304901123, + "learning_rate": 0.00018247256122460705, + "loss": 0.7987, + "step": 12428 + }, + { + "epoch": 0.8624661716744154, + "grad_norm": 2.285238265991211, + "learning_rate": 0.00018246938358151916, + "loss": 1.1001, + "step": 12429 + }, + { + "epoch": 0.862535563111512, + "grad_norm": 2.2340171337127686, + "learning_rate": 0.00018246620567808494, + "loss": 0.7746, + "step": 12430 + }, + { + "epoch": 0.8626049545486087, + "grad_norm": 2.8141872882843018, + "learning_rate": 0.00018246302751431453, + "loss": 0.9185, + "step": 12431 + }, + { + "epoch": 0.8626743459857054, + "grad_norm": 2.270369529724121, + "learning_rate": 0.00018245984909021786, + "loss": 0.8846, + "step": 12432 + }, + { + "epoch": 0.8627437374228021, + "grad_norm": 2.6493401527404785, + "learning_rate": 0.00018245667040580506, + "loss": 0.8754, + "step": 12433 + }, + { + "epoch": 0.8628131288598987, + "grad_norm": 2.2439112663269043, + "learning_rate": 0.00018245349146108608, + "loss": 1.023, + "step": 12434 + }, + { + "epoch": 0.8628825202969953, + "grad_norm": 1.4919184446334839, + "learning_rate": 0.000182450312256071, + "loss": 0.6902, + "step": 12435 + }, + { + "epoch": 0.862951911734092, + "grad_norm": 1.8266565799713135, + "learning_rate": 0.00018244713279076989, + "loss": 0.7547, + "step": 12436 + }, + { + "epoch": 0.8630213031711886, + "grad_norm": 2.225069522857666, + "learning_rate": 0.00018244395306519273, + "loss": 0.8657, + "step": 12437 + }, + { + "epoch": 0.8630906946082854, + "grad_norm": 1.9730101823806763, + "learning_rate": 0.00018244077307934955, + "loss": 0.8425, + "step": 12438 + }, + { + "epoch": 0.863160086045382, + "grad_norm": 1.8090755939483643, + "learning_rate": 0.00018243759283325043, + "loss": 0.7341, + "step": 12439 + }, + { + "epoch": 0.8632294774824787, + "grad_norm": 2.0701680183410645, + "learning_rate": 0.0001824344123269054, + "loss": 0.7417, + "step": 12440 + }, + { + "epoch": 0.8632988689195753, + "grad_norm": 1.5726792812347412, + "learning_rate": 0.00018243123156032448, + "loss": 0.5964, + "step": 12441 + }, + { + "epoch": 0.863368260356672, + "grad_norm": 1.768463134765625, + "learning_rate": 0.00018242805053351771, + "loss": 0.6876, + "step": 12442 + }, + { + "epoch": 0.8634376517937686, + "grad_norm": 2.5405783653259277, + "learning_rate": 0.0001824248692464952, + "loss": 1.1221, + "step": 12443 + }, + { + "epoch": 0.8635070432308654, + "grad_norm": 1.8877960443496704, + "learning_rate": 0.0001824216876992669, + "loss": 0.8036, + "step": 12444 + }, + { + "epoch": 0.863576434667962, + "grad_norm": 2.523918390274048, + "learning_rate": 0.00018241850589184293, + "loss": 0.9937, + "step": 12445 + }, + { + "epoch": 0.8636458261050587, + "grad_norm": 1.9105141162872314, + "learning_rate": 0.00018241532382423325, + "loss": 0.6799, + "step": 12446 + }, + { + "epoch": 0.8637152175421553, + "grad_norm": 2.0025951862335205, + "learning_rate": 0.00018241214149644798, + "loss": 0.655, + "step": 12447 + }, + { + "epoch": 0.8637846089792519, + "grad_norm": 2.7445483207702637, + "learning_rate": 0.00018240895890849715, + "loss": 0.73, + "step": 12448 + }, + { + "epoch": 0.8638540004163486, + "grad_norm": 2.2317891120910645, + "learning_rate": 0.00018240577606039077, + "loss": 0.6419, + "step": 12449 + }, + { + "epoch": 0.8639233918534452, + "grad_norm": 2.0144283771514893, + "learning_rate": 0.00018240259295213896, + "loss": 0.6551, + "step": 12450 + }, + { + "epoch": 0.863992783290542, + "grad_norm": 2.5677812099456787, + "learning_rate": 0.0001823994095837517, + "loss": 0.8051, + "step": 12451 + }, + { + "epoch": 0.8640621747276386, + "grad_norm": 2.5819458961486816, + "learning_rate": 0.00018239622595523906, + "loss": 0.9616, + "step": 12452 + }, + { + "epoch": 0.8641315661647353, + "grad_norm": 3.082693099975586, + "learning_rate": 0.00018239304206661109, + "loss": 1.1154, + "step": 12453 + }, + { + "epoch": 0.8642009576018319, + "grad_norm": 2.5215065479278564, + "learning_rate": 0.00018238985791787786, + "loss": 1.0073, + "step": 12454 + }, + { + "epoch": 0.8642703490389286, + "grad_norm": 2.7451114654541016, + "learning_rate": 0.0001823866735090494, + "loss": 0.7866, + "step": 12455 + }, + { + "epoch": 0.8643397404760252, + "grad_norm": 2.223222017288208, + "learning_rate": 0.00018238348884013576, + "loss": 0.6778, + "step": 12456 + }, + { + "epoch": 0.864409131913122, + "grad_norm": 2.014843702316284, + "learning_rate": 0.00018238030391114703, + "loss": 0.7892, + "step": 12457 + }, + { + "epoch": 0.8644785233502186, + "grad_norm": 1.6521233320236206, + "learning_rate": 0.0001823771187220932, + "loss": 0.7345, + "step": 12458 + }, + { + "epoch": 0.8645479147873153, + "grad_norm": 1.8740514516830444, + "learning_rate": 0.0001823739332729844, + "loss": 0.767, + "step": 12459 + }, + { + "epoch": 0.8646173062244119, + "grad_norm": 1.978336215019226, + "learning_rate": 0.00018237074756383064, + "loss": 0.8033, + "step": 12460 + }, + { + "epoch": 0.8646866976615085, + "grad_norm": 2.185563325881958, + "learning_rate": 0.00018236756159464198, + "loss": 1.1828, + "step": 12461 + }, + { + "epoch": 0.8647560890986052, + "grad_norm": 1.9275163412094116, + "learning_rate": 0.0001823643753654285, + "loss": 0.6494, + "step": 12462 + }, + { + "epoch": 0.8648254805357019, + "grad_norm": 1.6191508769989014, + "learning_rate": 0.00018236118887620024, + "loss": 0.6691, + "step": 12463 + }, + { + "epoch": 0.8648948719727986, + "grad_norm": 1.9223781824111938, + "learning_rate": 0.00018235800212696725, + "loss": 0.6091, + "step": 12464 + }, + { + "epoch": 0.8649642634098952, + "grad_norm": 1.8231695890426636, + "learning_rate": 0.00018235481511773958, + "loss": 0.9467, + "step": 12465 + }, + { + "epoch": 0.8650336548469919, + "grad_norm": 1.9254027605056763, + "learning_rate": 0.00018235162784852735, + "loss": 0.7738, + "step": 12466 + }, + { + "epoch": 0.8651030462840885, + "grad_norm": 2.125443696975708, + "learning_rate": 0.0001823484403193406, + "loss": 0.7132, + "step": 12467 + }, + { + "epoch": 0.8651724377211852, + "grad_norm": 2.2257378101348877, + "learning_rate": 0.00018234525253018932, + "loss": 0.8717, + "step": 12468 + }, + { + "epoch": 0.8652418291582818, + "grad_norm": 2.049816370010376, + "learning_rate": 0.0001823420644810837, + "loss": 0.884, + "step": 12469 + }, + { + "epoch": 0.8653112205953786, + "grad_norm": 1.583207607269287, + "learning_rate": 0.00018233887617203368, + "loss": 0.702, + "step": 12470 + }, + { + "epoch": 0.8653806120324752, + "grad_norm": 1.9976263046264648, + "learning_rate": 0.0001823356876030494, + "loss": 0.6424, + "step": 12471 + }, + { + "epoch": 0.8654500034695719, + "grad_norm": 1.6013530492782593, + "learning_rate": 0.00018233249877414091, + "loss": 0.4968, + "step": 12472 + }, + { + "epoch": 0.8655193949066685, + "grad_norm": 2.1518702507019043, + "learning_rate": 0.00018232930968531828, + "loss": 0.7064, + "step": 12473 + }, + { + "epoch": 0.8655887863437651, + "grad_norm": 1.8493248224258423, + "learning_rate": 0.00018232612033659156, + "loss": 0.851, + "step": 12474 + }, + { + "epoch": 0.8656581777808618, + "grad_norm": 2.190554618835449, + "learning_rate": 0.00018232293072797083, + "loss": 0.7374, + "step": 12475 + }, + { + "epoch": 0.8657275692179585, + "grad_norm": 1.8929197788238525, + "learning_rate": 0.00018231974085946618, + "loss": 0.5844, + "step": 12476 + }, + { + "epoch": 0.8657969606550552, + "grad_norm": 3.0221450328826904, + "learning_rate": 0.00018231655073108763, + "loss": 0.9481, + "step": 12477 + }, + { + "epoch": 0.8658663520921518, + "grad_norm": 1.6654034852981567, + "learning_rate": 0.0001823133603428453, + "loss": 0.8254, + "step": 12478 + }, + { + "epoch": 0.8659357435292485, + "grad_norm": 1.7779550552368164, + "learning_rate": 0.0001823101696947492, + "loss": 0.6437, + "step": 12479 + }, + { + "epoch": 0.8660051349663451, + "grad_norm": 1.9613122940063477, + "learning_rate": 0.0001823069787868095, + "loss": 0.5555, + "step": 12480 + }, + { + "epoch": 0.8660745264034418, + "grad_norm": 2.2650911808013916, + "learning_rate": 0.00018230378761903619, + "loss": 0.8309, + "step": 12481 + }, + { + "epoch": 0.8661439178405385, + "grad_norm": 2.2644433975219727, + "learning_rate": 0.00018230059619143938, + "loss": 0.5292, + "step": 12482 + }, + { + "epoch": 0.8662133092776352, + "grad_norm": 2.4836676120758057, + "learning_rate": 0.00018229740450402912, + "loss": 1.0603, + "step": 12483 + }, + { + "epoch": 0.8662827007147318, + "grad_norm": 1.8865313529968262, + "learning_rate": 0.00018229421255681552, + "loss": 0.9281, + "step": 12484 + }, + { + "epoch": 0.8663520921518285, + "grad_norm": 2.475339412689209, + "learning_rate": 0.00018229102034980862, + "loss": 1.1852, + "step": 12485 + }, + { + "epoch": 0.8664214835889251, + "grad_norm": 1.9543753862380981, + "learning_rate": 0.00018228782788301854, + "loss": 0.6699, + "step": 12486 + }, + { + "epoch": 0.8664908750260217, + "grad_norm": 2.609827995300293, + "learning_rate": 0.00018228463515645532, + "loss": 0.9423, + "step": 12487 + }, + { + "epoch": 0.8665602664631185, + "grad_norm": 2.225247621536255, + "learning_rate": 0.00018228144217012906, + "loss": 0.6361, + "step": 12488 + }, + { + "epoch": 0.8666296579002151, + "grad_norm": 1.71830415725708, + "learning_rate": 0.00018227824892404982, + "loss": 0.6298, + "step": 12489 + }, + { + "epoch": 0.8666990493373118, + "grad_norm": 3.350383758544922, + "learning_rate": 0.0001822750554182277, + "loss": 0.7064, + "step": 12490 + }, + { + "epoch": 0.8667684407744084, + "grad_norm": 1.9286282062530518, + "learning_rate": 0.0001822718616526728, + "loss": 0.7878, + "step": 12491 + }, + { + "epoch": 0.8668378322115051, + "grad_norm": 2.1912529468536377, + "learning_rate": 0.00018226866762739515, + "loss": 0.8188, + "step": 12492 + }, + { + "epoch": 0.8669072236486017, + "grad_norm": 2.182093620300293, + "learning_rate": 0.00018226547334240488, + "loss": 0.8991, + "step": 12493 + }, + { + "epoch": 0.8669766150856985, + "grad_norm": 1.950215458869934, + "learning_rate": 0.00018226227879771206, + "loss": 0.9964, + "step": 12494 + }, + { + "epoch": 0.8670460065227951, + "grad_norm": 1.7222949266433716, + "learning_rate": 0.00018225908399332674, + "loss": 0.6071, + "step": 12495 + }, + { + "epoch": 0.8671153979598918, + "grad_norm": 2.0106992721557617, + "learning_rate": 0.00018225588892925907, + "loss": 0.9902, + "step": 12496 + }, + { + "epoch": 0.8671847893969884, + "grad_norm": 1.8430488109588623, + "learning_rate": 0.00018225269360551906, + "loss": 0.6018, + "step": 12497 + }, + { + "epoch": 0.8672541808340851, + "grad_norm": 1.9739693403244019, + "learning_rate": 0.00018224949802211687, + "loss": 0.5935, + "step": 12498 + }, + { + "epoch": 0.8673235722711817, + "grad_norm": 2.41042160987854, + "learning_rate": 0.00018224630217906256, + "loss": 0.9615, + "step": 12499 + }, + { + "epoch": 0.8673929637082785, + "grad_norm": 2.0864365100860596, + "learning_rate": 0.0001822431060763662, + "loss": 0.7928, + "step": 12500 + }, + { + "epoch": 0.8674623551453751, + "grad_norm": 2.109287738800049, + "learning_rate": 0.00018223990971403788, + "loss": 0.9726, + "step": 12501 + }, + { + "epoch": 0.8675317465824717, + "grad_norm": 1.9120895862579346, + "learning_rate": 0.00018223671309208777, + "loss": 0.7046, + "step": 12502 + }, + { + "epoch": 0.8676011380195684, + "grad_norm": 2.592918634414673, + "learning_rate": 0.00018223351621052584, + "loss": 0.7486, + "step": 12503 + }, + { + "epoch": 0.867670529456665, + "grad_norm": 2.3989551067352295, + "learning_rate": 0.00018223031906936226, + "loss": 1.0276, + "step": 12504 + }, + { + "epoch": 0.8677399208937617, + "grad_norm": 3.3559982776641846, + "learning_rate": 0.0001822271216686071, + "loss": 1.0443, + "step": 12505 + }, + { + "epoch": 0.8678093123308583, + "grad_norm": 2.0016441345214844, + "learning_rate": 0.00018222392400827047, + "loss": 0.757, + "step": 12506 + }, + { + "epoch": 0.8678787037679551, + "grad_norm": 2.725231647491455, + "learning_rate": 0.00018222072608836242, + "loss": 0.7436, + "step": 12507 + }, + { + "epoch": 0.8679480952050517, + "grad_norm": 2.135066509246826, + "learning_rate": 0.0001822175279088931, + "loss": 0.9189, + "step": 12508 + }, + { + "epoch": 0.8680174866421484, + "grad_norm": 2.0641071796417236, + "learning_rate": 0.00018221432946987257, + "loss": 0.7079, + "step": 12509 + }, + { + "epoch": 0.868086878079245, + "grad_norm": 2.5072429180145264, + "learning_rate": 0.00018221113077131096, + "loss": 0.7892, + "step": 12510 + }, + { + "epoch": 0.8681562695163417, + "grad_norm": 1.9362396001815796, + "learning_rate": 0.00018220793181321834, + "loss": 0.7367, + "step": 12511 + }, + { + "epoch": 0.8682256609534383, + "grad_norm": 1.5841652154922485, + "learning_rate": 0.0001822047325956048, + "loss": 0.9021, + "step": 12512 + }, + { + "epoch": 0.8682950523905351, + "grad_norm": 2.2087366580963135, + "learning_rate": 0.00018220153311848049, + "loss": 0.925, + "step": 12513 + }, + { + "epoch": 0.8683644438276317, + "grad_norm": 2.0256528854370117, + "learning_rate": 0.00018219833338185547, + "loss": 0.9655, + "step": 12514 + }, + { + "epoch": 0.8684338352647283, + "grad_norm": 1.9305418729782104, + "learning_rate": 0.0001821951333857398, + "loss": 0.7256, + "step": 12515 + }, + { + "epoch": 0.868503226701825, + "grad_norm": 1.9103597402572632, + "learning_rate": 0.0001821919331301437, + "loss": 0.921, + "step": 12516 + }, + { + "epoch": 0.8685726181389216, + "grad_norm": 2.0407230854034424, + "learning_rate": 0.00018218873261507715, + "loss": 0.778, + "step": 12517 + }, + { + "epoch": 0.8686420095760183, + "grad_norm": 1.9622366428375244, + "learning_rate": 0.00018218553184055035, + "loss": 0.7762, + "step": 12518 + }, + { + "epoch": 0.868711401013115, + "grad_norm": 4.713228225708008, + "learning_rate": 0.00018218233080657332, + "loss": 0.9118, + "step": 12519 + }, + { + "epoch": 0.8687807924502117, + "grad_norm": 3.347399950027466, + "learning_rate": 0.00018217912951315623, + "loss": 0.8248, + "step": 12520 + }, + { + "epoch": 0.8688501838873083, + "grad_norm": 2.339782476425171, + "learning_rate": 0.00018217592796030915, + "loss": 0.7948, + "step": 12521 + }, + { + "epoch": 0.868919575324405, + "grad_norm": 2.7206175327301025, + "learning_rate": 0.00018217272614804223, + "loss": 1.0039, + "step": 12522 + }, + { + "epoch": 0.8689889667615016, + "grad_norm": 2.187262535095215, + "learning_rate": 0.0001821695240763655, + "loss": 0.7364, + "step": 12523 + }, + { + "epoch": 0.8690583581985983, + "grad_norm": 1.9202880859375, + "learning_rate": 0.00018216632174528915, + "loss": 0.9042, + "step": 12524 + }, + { + "epoch": 0.869127749635695, + "grad_norm": 2.194622039794922, + "learning_rate": 0.00018216311915482324, + "loss": 0.7978, + "step": 12525 + }, + { + "epoch": 0.8691971410727917, + "grad_norm": 2.776184558868408, + "learning_rate": 0.0001821599163049779, + "loss": 0.8311, + "step": 12526 + }, + { + "epoch": 0.8692665325098883, + "grad_norm": 2.1178598403930664, + "learning_rate": 0.00018215671319576324, + "loss": 1.0069, + "step": 12527 + }, + { + "epoch": 0.8693359239469849, + "grad_norm": 2.502403497695923, + "learning_rate": 0.00018215350982718937, + "loss": 0.8758, + "step": 12528 + }, + { + "epoch": 0.8694053153840816, + "grad_norm": 1.8099929094314575, + "learning_rate": 0.00018215030619926642, + "loss": 0.7003, + "step": 12529 + }, + { + "epoch": 0.8694747068211782, + "grad_norm": 1.8974779844284058, + "learning_rate": 0.00018214710231200445, + "loss": 0.6111, + "step": 12530 + }, + { + "epoch": 0.869544098258275, + "grad_norm": 2.5292856693267822, + "learning_rate": 0.00018214389816541363, + "loss": 1.0423, + "step": 12531 + }, + { + "epoch": 0.8696134896953716, + "grad_norm": 1.8030269145965576, + "learning_rate": 0.00018214069375950404, + "loss": 0.9248, + "step": 12532 + }, + { + "epoch": 0.8696828811324683, + "grad_norm": 1.740323543548584, + "learning_rate": 0.00018213748909428582, + "loss": 0.7193, + "step": 12533 + }, + { + "epoch": 0.8697522725695649, + "grad_norm": 1.7486988306045532, + "learning_rate": 0.00018213428416976907, + "loss": 0.9737, + "step": 12534 + }, + { + "epoch": 0.8698216640066616, + "grad_norm": 2.025628089904785, + "learning_rate": 0.0001821310789859639, + "loss": 0.8005, + "step": 12535 + }, + { + "epoch": 0.8698910554437582, + "grad_norm": 2.880861759185791, + "learning_rate": 0.00018212787354288045, + "loss": 0.9328, + "step": 12536 + }, + { + "epoch": 0.8699604468808549, + "grad_norm": 1.7778081893920898, + "learning_rate": 0.00018212466784052887, + "loss": 0.6038, + "step": 12537 + }, + { + "epoch": 0.8700298383179516, + "grad_norm": 7.533321380615234, + "learning_rate": 0.0001821214618789192, + "loss": 1.0276, + "step": 12538 + }, + { + "epoch": 0.8700992297550483, + "grad_norm": 1.7644622325897217, + "learning_rate": 0.0001821182556580616, + "loss": 0.5973, + "step": 12539 + }, + { + "epoch": 0.8701686211921449, + "grad_norm": 1.7313110828399658, + "learning_rate": 0.00018211504917796622, + "loss": 0.7193, + "step": 12540 + }, + { + "epoch": 0.8702380126292415, + "grad_norm": 2.4053773880004883, + "learning_rate": 0.00018211184243864317, + "loss": 0.8724, + "step": 12541 + }, + { + "epoch": 0.8703074040663382, + "grad_norm": 2.619140148162842, + "learning_rate": 0.00018210863544010253, + "loss": 0.8168, + "step": 12542 + }, + { + "epoch": 0.8703767955034348, + "grad_norm": 2.043833017349243, + "learning_rate": 0.0001821054281823545, + "loss": 0.9175, + "step": 12543 + }, + { + "epoch": 0.8704461869405316, + "grad_norm": 2.6511716842651367, + "learning_rate": 0.00018210222066540908, + "loss": 0.7907, + "step": 12544 + }, + { + "epoch": 0.8705155783776282, + "grad_norm": 2.51072359085083, + "learning_rate": 0.0001820990128892765, + "loss": 1.0007, + "step": 12545 + }, + { + "epoch": 0.8705849698147249, + "grad_norm": 2.365651845932007, + "learning_rate": 0.00018209580485396688, + "loss": 0.9428, + "step": 12546 + }, + { + "epoch": 0.8706543612518215, + "grad_norm": 2.100951671600342, + "learning_rate": 0.00018209259655949033, + "loss": 0.8924, + "step": 12547 + }, + { + "epoch": 0.8707237526889182, + "grad_norm": 1.8533655405044556, + "learning_rate": 0.000182089388005857, + "loss": 0.6847, + "step": 12548 + }, + { + "epoch": 0.8707931441260148, + "grad_norm": 1.8895347118377686, + "learning_rate": 0.00018208617919307694, + "loss": 0.6814, + "step": 12549 + }, + { + "epoch": 0.8708625355631116, + "grad_norm": 1.6292566061019897, + "learning_rate": 0.00018208297012116037, + "loss": 0.6533, + "step": 12550 + }, + { + "epoch": 0.8709319270002082, + "grad_norm": 2.0484938621520996, + "learning_rate": 0.00018207976079011738, + "loss": 1.0244, + "step": 12551 + }, + { + "epoch": 0.8710013184373049, + "grad_norm": 3.194448709487915, + "learning_rate": 0.00018207655119995814, + "loss": 0.8371, + "step": 12552 + }, + { + "epoch": 0.8710707098744015, + "grad_norm": 2.364110231399536, + "learning_rate": 0.00018207334135069268, + "loss": 0.8316, + "step": 12553 + }, + { + "epoch": 0.8711401013114982, + "grad_norm": 1.601057529449463, + "learning_rate": 0.00018207013124233126, + "loss": 0.5536, + "step": 12554 + }, + { + "epoch": 0.8712094927485948, + "grad_norm": 2.888101100921631, + "learning_rate": 0.00018206692087488393, + "loss": 0.8002, + "step": 12555 + }, + { + "epoch": 0.8712788841856914, + "grad_norm": 2.0574445724487305, + "learning_rate": 0.00018206371024836088, + "loss": 0.6548, + "step": 12556 + }, + { + "epoch": 0.8713482756227882, + "grad_norm": 2.0151379108428955, + "learning_rate": 0.0001820604993627722, + "loss": 0.8601, + "step": 12557 + }, + { + "epoch": 0.8714176670598848, + "grad_norm": 2.483844757080078, + "learning_rate": 0.00018205728821812803, + "loss": 0.5708, + "step": 12558 + }, + { + "epoch": 0.8714870584969815, + "grad_norm": 2.2562453746795654, + "learning_rate": 0.00018205407681443856, + "loss": 0.6091, + "step": 12559 + }, + { + "epoch": 0.8715564499340781, + "grad_norm": 2.24324631690979, + "learning_rate": 0.00018205086515171387, + "loss": 0.9782, + "step": 12560 + }, + { + "epoch": 0.8716258413711748, + "grad_norm": 3.4147510528564453, + "learning_rate": 0.0001820476532299641, + "loss": 0.9973, + "step": 12561 + }, + { + "epoch": 0.8716952328082714, + "grad_norm": 2.370276927947998, + "learning_rate": 0.00018204444104919943, + "loss": 0.775, + "step": 12562 + }, + { + "epoch": 0.8717646242453682, + "grad_norm": 2.135622024536133, + "learning_rate": 0.00018204122860942997, + "loss": 0.7117, + "step": 12563 + }, + { + "epoch": 0.8718340156824648, + "grad_norm": 1.409454345703125, + "learning_rate": 0.00018203801591066586, + "loss": 0.5822, + "step": 12564 + }, + { + "epoch": 0.8719034071195615, + "grad_norm": 2.9068143367767334, + "learning_rate": 0.00018203480295291726, + "loss": 0.8154, + "step": 12565 + }, + { + "epoch": 0.8719727985566581, + "grad_norm": 2.71016263961792, + "learning_rate": 0.00018203158973619435, + "loss": 0.8865, + "step": 12566 + }, + { + "epoch": 0.8720421899937548, + "grad_norm": 2.37736177444458, + "learning_rate": 0.00018202837626050714, + "loss": 0.9696, + "step": 12567 + }, + { + "epoch": 0.8721115814308514, + "grad_norm": 2.1562445163726807, + "learning_rate": 0.00018202516252586594, + "loss": 0.7472, + "step": 12568 + }, + { + "epoch": 0.872180972867948, + "grad_norm": 1.9587435722351074, + "learning_rate": 0.00018202194853228078, + "loss": 0.4494, + "step": 12569 + }, + { + "epoch": 0.8722503643050448, + "grad_norm": 2.3693156242370605, + "learning_rate": 0.00018201873427976185, + "loss": 0.7935, + "step": 12570 + }, + { + "epoch": 0.8723197557421414, + "grad_norm": 2.2702016830444336, + "learning_rate": 0.0001820155197683193, + "loss": 0.8693, + "step": 12571 + }, + { + "epoch": 0.8723891471792381, + "grad_norm": 3.0384938716888428, + "learning_rate": 0.00018201230499796328, + "loss": 0.8753, + "step": 12572 + }, + { + "epoch": 0.8724585386163347, + "grad_norm": 2.9293391704559326, + "learning_rate": 0.00018200908996870389, + "loss": 0.8643, + "step": 12573 + }, + { + "epoch": 0.8725279300534314, + "grad_norm": 2.935704231262207, + "learning_rate": 0.00018200587468055135, + "loss": 0.9763, + "step": 12574 + }, + { + "epoch": 0.872597321490528, + "grad_norm": 2.9378538131713867, + "learning_rate": 0.00018200265913351576, + "loss": 0.5809, + "step": 12575 + }, + { + "epoch": 0.8726667129276248, + "grad_norm": 2.233747720718384, + "learning_rate": 0.00018199944332760733, + "loss": 0.6657, + "step": 12576 + }, + { + "epoch": 0.8727361043647214, + "grad_norm": 2.216723918914795, + "learning_rate": 0.00018199622726283612, + "loss": 0.7664, + "step": 12577 + }, + { + "epoch": 0.8728054958018181, + "grad_norm": 1.6213821172714233, + "learning_rate": 0.00018199301093921235, + "loss": 0.767, + "step": 12578 + }, + { + "epoch": 0.8728748872389147, + "grad_norm": 2.2251670360565186, + "learning_rate": 0.00018198979435674617, + "loss": 0.787, + "step": 12579 + }, + { + "epoch": 0.8729442786760114, + "grad_norm": 2.3509883880615234, + "learning_rate": 0.0001819865775154477, + "loss": 0.9181, + "step": 12580 + }, + { + "epoch": 0.873013670113108, + "grad_norm": 1.9009382724761963, + "learning_rate": 0.00018198336041532715, + "loss": 0.7887, + "step": 12581 + }, + { + "epoch": 0.8730830615502047, + "grad_norm": 2.022020101547241, + "learning_rate": 0.0001819801430563946, + "loss": 0.8156, + "step": 12582 + }, + { + "epoch": 0.8731524529873014, + "grad_norm": 2.406494617462158, + "learning_rate": 0.00018197692543866028, + "loss": 1.0065, + "step": 12583 + }, + { + "epoch": 0.873221844424398, + "grad_norm": 1.7449620962142944, + "learning_rate": 0.0001819737075621343, + "loss": 0.7524, + "step": 12584 + }, + { + "epoch": 0.8732912358614947, + "grad_norm": 1.9590848684310913, + "learning_rate": 0.00018197048942682686, + "loss": 0.5547, + "step": 12585 + }, + { + "epoch": 0.8733606272985913, + "grad_norm": 1.8677829504013062, + "learning_rate": 0.00018196727103274806, + "loss": 0.8101, + "step": 12586 + }, + { + "epoch": 0.873430018735688, + "grad_norm": 2.045203447341919, + "learning_rate": 0.00018196405237990812, + "loss": 0.6354, + "step": 12587 + }, + { + "epoch": 0.8734994101727847, + "grad_norm": 2.189237356185913, + "learning_rate": 0.0001819608334683172, + "loss": 0.7808, + "step": 12588 + }, + { + "epoch": 0.8735688016098814, + "grad_norm": 2.059425115585327, + "learning_rate": 0.0001819576142979854, + "loss": 0.8966, + "step": 12589 + }, + { + "epoch": 0.873638193046978, + "grad_norm": 1.8788866996765137, + "learning_rate": 0.00018195439486892295, + "loss": 0.6522, + "step": 12590 + }, + { + "epoch": 0.8737075844840747, + "grad_norm": 2.2319345474243164, + "learning_rate": 0.00018195117518113993, + "loss": 0.8006, + "step": 12591 + }, + { + "epoch": 0.8737769759211713, + "grad_norm": 2.5609145164489746, + "learning_rate": 0.0001819479552346466, + "loss": 0.6361, + "step": 12592 + }, + { + "epoch": 0.873846367358268, + "grad_norm": 2.18280291557312, + "learning_rate": 0.0001819447350294531, + "loss": 0.7599, + "step": 12593 + }, + { + "epoch": 0.8739157587953646, + "grad_norm": 1.9896440505981445, + "learning_rate": 0.00018194151456556954, + "loss": 0.6411, + "step": 12594 + }, + { + "epoch": 0.8739851502324613, + "grad_norm": 2.1818931102752686, + "learning_rate": 0.00018193829384300614, + "loss": 0.8605, + "step": 12595 + }, + { + "epoch": 0.874054541669558, + "grad_norm": 1.9465837478637695, + "learning_rate": 0.00018193507286177308, + "loss": 0.6102, + "step": 12596 + }, + { + "epoch": 0.8741239331066546, + "grad_norm": 1.825915813446045, + "learning_rate": 0.00018193185162188047, + "loss": 0.8748, + "step": 12597 + }, + { + "epoch": 0.8741933245437513, + "grad_norm": 2.382108688354492, + "learning_rate": 0.0001819286301233385, + "loss": 1.0815, + "step": 12598 + }, + { + "epoch": 0.8742627159808479, + "grad_norm": 2.45334792137146, + "learning_rate": 0.00018192540836615737, + "loss": 1.0521, + "step": 12599 + }, + { + "epoch": 0.8743321074179446, + "grad_norm": 2.544523239135742, + "learning_rate": 0.0001819221863503472, + "loss": 0.8278, + "step": 12600 + }, + { + "epoch": 0.8744014988550413, + "grad_norm": 1.6479886770248413, + "learning_rate": 0.00018191896407591823, + "loss": 0.7579, + "step": 12601 + }, + { + "epoch": 0.874470890292138, + "grad_norm": 2.50156307220459, + "learning_rate": 0.0001819157415428806, + "loss": 0.8518, + "step": 12602 + }, + { + "epoch": 0.8745402817292346, + "grad_norm": 2.3580963611602783, + "learning_rate": 0.00018191251875124444, + "loss": 0.6441, + "step": 12603 + }, + { + "epoch": 0.8746096731663313, + "grad_norm": 3.21043062210083, + "learning_rate": 0.00018190929570102, + "loss": 0.882, + "step": 12604 + }, + { + "epoch": 0.8746790646034279, + "grad_norm": 2.210526466369629, + "learning_rate": 0.0001819060723922174, + "loss": 0.7642, + "step": 12605 + }, + { + "epoch": 0.8747484560405246, + "grad_norm": 1.9265553951263428, + "learning_rate": 0.00018190284882484682, + "loss": 0.8812, + "step": 12606 + }, + { + "epoch": 0.8748178474776213, + "grad_norm": 2.323246717453003, + "learning_rate": 0.0001818996249989185, + "loss": 0.8277, + "step": 12607 + }, + { + "epoch": 0.8748872389147179, + "grad_norm": 2.281513214111328, + "learning_rate": 0.00018189640091444248, + "loss": 0.7892, + "step": 12608 + }, + { + "epoch": 0.8749566303518146, + "grad_norm": 2.526566982269287, + "learning_rate": 0.00018189317657142908, + "loss": 0.74, + "step": 12609 + }, + { + "epoch": 0.8750260217889112, + "grad_norm": 2.44935941696167, + "learning_rate": 0.00018188995196988844, + "loss": 0.7061, + "step": 12610 + }, + { + "epoch": 0.8750954132260079, + "grad_norm": 1.9641973972320557, + "learning_rate": 0.00018188672710983068, + "loss": 0.8835, + "step": 12611 + }, + { + "epoch": 0.8751648046631045, + "grad_norm": 2.227647304534912, + "learning_rate": 0.00018188350199126605, + "loss": 0.8131, + "step": 12612 + }, + { + "epoch": 0.8752341961002013, + "grad_norm": 2.113889217376709, + "learning_rate": 0.00018188027661420467, + "loss": 0.8176, + "step": 12613 + }, + { + "epoch": 0.8753035875372979, + "grad_norm": 2.025012493133545, + "learning_rate": 0.00018187705097865678, + "loss": 0.9706, + "step": 12614 + }, + { + "epoch": 0.8753729789743946, + "grad_norm": 3.0652074813842773, + "learning_rate": 0.00018187382508463255, + "loss": 0.8979, + "step": 12615 + }, + { + "epoch": 0.8754423704114912, + "grad_norm": 2.2257072925567627, + "learning_rate": 0.00018187059893214213, + "loss": 0.9042, + "step": 12616 + }, + { + "epoch": 0.8755117618485879, + "grad_norm": 2.7683773040771484, + "learning_rate": 0.00018186737252119577, + "loss": 0.8943, + "step": 12617 + }, + { + "epoch": 0.8755811532856845, + "grad_norm": 2.775110960006714, + "learning_rate": 0.00018186414585180357, + "loss": 0.8425, + "step": 12618 + }, + { + "epoch": 0.8756505447227813, + "grad_norm": 2.4908504486083984, + "learning_rate": 0.00018186091892397575, + "loss": 0.8725, + "step": 12619 + }, + { + "epoch": 0.8757199361598779, + "grad_norm": 2.2451508045196533, + "learning_rate": 0.00018185769173772255, + "loss": 0.8553, + "step": 12620 + }, + { + "epoch": 0.8757893275969746, + "grad_norm": 1.8445395231246948, + "learning_rate": 0.0001818544642930541, + "loss": 0.8079, + "step": 12621 + }, + { + "epoch": 0.8758587190340712, + "grad_norm": 2.137305498123169, + "learning_rate": 0.0001818512365899806, + "loss": 0.7199, + "step": 12622 + }, + { + "epoch": 0.8759281104711678, + "grad_norm": 2.4649088382720947, + "learning_rate": 0.00018184800862851224, + "loss": 0.8584, + "step": 12623 + }, + { + "epoch": 0.8759975019082645, + "grad_norm": 2.5827910900115967, + "learning_rate": 0.00018184478040865923, + "loss": 0.8195, + "step": 12624 + }, + { + "epoch": 0.8760668933453611, + "grad_norm": 2.3099164962768555, + "learning_rate": 0.00018184155193043172, + "loss": 0.7025, + "step": 12625 + }, + { + "epoch": 0.8761362847824579, + "grad_norm": 1.8132643699645996, + "learning_rate": 0.00018183832319383994, + "loss": 0.7355, + "step": 12626 + }, + { + "epoch": 0.8762056762195545, + "grad_norm": 2.2269885540008545, + "learning_rate": 0.00018183509419889406, + "loss": 1.0597, + "step": 12627 + }, + { + "epoch": 0.8762750676566512, + "grad_norm": 2.6724913120269775, + "learning_rate": 0.0001818318649456043, + "loss": 0.8775, + "step": 12628 + }, + { + "epoch": 0.8763444590937478, + "grad_norm": 1.5041455030441284, + "learning_rate": 0.00018182863543398078, + "loss": 0.6943, + "step": 12629 + }, + { + "epoch": 0.8764138505308445, + "grad_norm": 1.7666789293289185, + "learning_rate": 0.00018182540566403382, + "loss": 0.6581, + "step": 12630 + }, + { + "epoch": 0.8764832419679411, + "grad_norm": 4.068995475769043, + "learning_rate": 0.0001818221756357735, + "loss": 0.7913, + "step": 12631 + }, + { + "epoch": 0.8765526334050379, + "grad_norm": 1.6426293849945068, + "learning_rate": 0.0001818189453492101, + "loss": 0.6109, + "step": 12632 + }, + { + "epoch": 0.8766220248421345, + "grad_norm": 1.9500813484191895, + "learning_rate": 0.00018181571480435378, + "loss": 0.6262, + "step": 12633 + }, + { + "epoch": 0.8766914162792312, + "grad_norm": 2.172515869140625, + "learning_rate": 0.0001818124840012147, + "loss": 0.9268, + "step": 12634 + }, + { + "epoch": 0.8767608077163278, + "grad_norm": 2.115096092224121, + "learning_rate": 0.00018180925293980314, + "loss": 0.7417, + "step": 12635 + }, + { + "epoch": 0.8768301991534244, + "grad_norm": 2.764369249343872, + "learning_rate": 0.00018180602162012926, + "loss": 0.7075, + "step": 12636 + }, + { + "epoch": 0.8768995905905211, + "grad_norm": 2.122393846511841, + "learning_rate": 0.00018180279004220324, + "loss": 0.8774, + "step": 12637 + }, + { + "epoch": 0.8769689820276177, + "grad_norm": 2.508014440536499, + "learning_rate": 0.0001817995582060353, + "loss": 0.8737, + "step": 12638 + }, + { + "epoch": 0.8770383734647145, + "grad_norm": 2.4336936473846436, + "learning_rate": 0.00018179632611163568, + "loss": 1.0235, + "step": 12639 + }, + { + "epoch": 0.8771077649018111, + "grad_norm": 2.0864412784576416, + "learning_rate": 0.00018179309375901452, + "loss": 0.8547, + "step": 12640 + }, + { + "epoch": 0.8771771563389078, + "grad_norm": 2.6634671688079834, + "learning_rate": 0.00018178986114818205, + "loss": 0.8767, + "step": 12641 + }, + { + "epoch": 0.8772465477760044, + "grad_norm": 1.97787344455719, + "learning_rate": 0.00018178662827914848, + "loss": 0.7804, + "step": 12642 + }, + { + "epoch": 0.8773159392131011, + "grad_norm": 1.688357949256897, + "learning_rate": 0.00018178339515192404, + "loss": 0.6553, + "step": 12643 + }, + { + "epoch": 0.8773853306501977, + "grad_norm": 2.09606671333313, + "learning_rate": 0.0001817801617665189, + "loss": 0.7912, + "step": 12644 + }, + { + "epoch": 0.8774547220872945, + "grad_norm": 2.8582165241241455, + "learning_rate": 0.00018177692812294325, + "loss": 0.8152, + "step": 12645 + }, + { + "epoch": 0.8775241135243911, + "grad_norm": 2.0125362873077393, + "learning_rate": 0.00018177369422120736, + "loss": 0.8412, + "step": 12646 + }, + { + "epoch": 0.8775935049614878, + "grad_norm": 1.9474414587020874, + "learning_rate": 0.00018177046006132139, + "loss": 0.7301, + "step": 12647 + }, + { + "epoch": 0.8776628963985844, + "grad_norm": 1.9521814584732056, + "learning_rate": 0.00018176722564329555, + "loss": 0.6873, + "step": 12648 + }, + { + "epoch": 0.877732287835681, + "grad_norm": 1.9364619255065918, + "learning_rate": 0.00018176399096714007, + "loss": 0.73, + "step": 12649 + }, + { + "epoch": 0.8778016792727777, + "grad_norm": 2.641662836074829, + "learning_rate": 0.00018176075603286519, + "loss": 0.9731, + "step": 12650 + }, + { + "epoch": 0.8778710707098744, + "grad_norm": 3.120943784713745, + "learning_rate": 0.00018175752084048104, + "loss": 0.9326, + "step": 12651 + }, + { + "epoch": 0.8779404621469711, + "grad_norm": 2.334988594055176, + "learning_rate": 0.00018175428538999794, + "loss": 0.929, + "step": 12652 + }, + { + "epoch": 0.8780098535840677, + "grad_norm": 2.6817235946655273, + "learning_rate": 0.000181751049681426, + "loss": 1.2227, + "step": 12653 + }, + { + "epoch": 0.8780792450211644, + "grad_norm": 2.7452940940856934, + "learning_rate": 0.0001817478137147755, + "loss": 0.722, + "step": 12654 + }, + { + "epoch": 0.878148636458261, + "grad_norm": 2.0903849601745605, + "learning_rate": 0.00018174457749005663, + "loss": 0.793, + "step": 12655 + }, + { + "epoch": 0.8782180278953577, + "grad_norm": 1.7760542631149292, + "learning_rate": 0.0001817413410072796, + "loss": 0.8001, + "step": 12656 + }, + { + "epoch": 0.8782874193324544, + "grad_norm": 2.444199323654175, + "learning_rate": 0.00018173810426645468, + "loss": 0.937, + "step": 12657 + }, + { + "epoch": 0.8783568107695511, + "grad_norm": 2.1783416271209717, + "learning_rate": 0.000181734867267592, + "loss": 0.8071, + "step": 12658 + }, + { + "epoch": 0.8784262022066477, + "grad_norm": 2.4425604343414307, + "learning_rate": 0.00018173163001070186, + "loss": 0.9291, + "step": 12659 + }, + { + "epoch": 0.8784955936437444, + "grad_norm": 2.0551180839538574, + "learning_rate": 0.00018172839249579444, + "loss": 0.8097, + "step": 12660 + }, + { + "epoch": 0.878564985080841, + "grad_norm": 2.849541664123535, + "learning_rate": 0.00018172515472287995, + "loss": 1.0038, + "step": 12661 + }, + { + "epoch": 0.8786343765179376, + "grad_norm": 1.92014479637146, + "learning_rate": 0.00018172191669196865, + "loss": 0.8298, + "step": 12662 + }, + { + "epoch": 0.8787037679550344, + "grad_norm": 2.7527687549591064, + "learning_rate": 0.00018171867840307074, + "loss": 0.9935, + "step": 12663 + }, + { + "epoch": 0.878773159392131, + "grad_norm": 2.786311388015747, + "learning_rate": 0.00018171543985619643, + "loss": 0.9091, + "step": 12664 + }, + { + "epoch": 0.8788425508292277, + "grad_norm": 3.204841136932373, + "learning_rate": 0.00018171220105135595, + "loss": 0.9168, + "step": 12665 + }, + { + "epoch": 0.8789119422663243, + "grad_norm": 2.4401590824127197, + "learning_rate": 0.00018170896198855955, + "loss": 0.785, + "step": 12666 + }, + { + "epoch": 0.878981333703421, + "grad_norm": 2.2833640575408936, + "learning_rate": 0.00018170572266781743, + "loss": 0.8178, + "step": 12667 + }, + { + "epoch": 0.8790507251405176, + "grad_norm": 2.2440619468688965, + "learning_rate": 0.00018170248308913985, + "loss": 0.7841, + "step": 12668 + }, + { + "epoch": 0.8791201165776144, + "grad_norm": 3.6910459995269775, + "learning_rate": 0.00018169924325253697, + "loss": 1.1544, + "step": 12669 + }, + { + "epoch": 0.879189508014711, + "grad_norm": 2.083364248275757, + "learning_rate": 0.00018169600315801906, + "loss": 0.7407, + "step": 12670 + }, + { + "epoch": 0.8792588994518077, + "grad_norm": 2.063063144683838, + "learning_rate": 0.00018169276280559637, + "loss": 0.9101, + "step": 12671 + }, + { + "epoch": 0.8793282908889043, + "grad_norm": 1.7768365144729614, + "learning_rate": 0.0001816895221952791, + "loss": 0.8072, + "step": 12672 + }, + { + "epoch": 0.879397682326001, + "grad_norm": 2.3506481647491455, + "learning_rate": 0.0001816862813270775, + "loss": 0.7554, + "step": 12673 + }, + { + "epoch": 0.8794670737630976, + "grad_norm": 2.3328981399536133, + "learning_rate": 0.00018168304020100174, + "loss": 0.9028, + "step": 12674 + }, + { + "epoch": 0.8795364652001944, + "grad_norm": 1.718381643295288, + "learning_rate": 0.00018167979881706213, + "loss": 0.8828, + "step": 12675 + }, + { + "epoch": 0.879605856637291, + "grad_norm": 2.348280668258667, + "learning_rate": 0.00018167655717526887, + "loss": 0.9409, + "step": 12676 + }, + { + "epoch": 0.8796752480743876, + "grad_norm": 2.395860195159912, + "learning_rate": 0.00018167331527563221, + "loss": 0.8446, + "step": 12677 + }, + { + "epoch": 0.8797446395114843, + "grad_norm": 2.5764975547790527, + "learning_rate": 0.00018167007311816234, + "loss": 0.7974, + "step": 12678 + }, + { + "epoch": 0.8798140309485809, + "grad_norm": 2.4855682849884033, + "learning_rate": 0.00018166683070286953, + "loss": 0.9343, + "step": 12679 + }, + { + "epoch": 0.8798834223856776, + "grad_norm": 2.703223466873169, + "learning_rate": 0.00018166358802976405, + "loss": 0.812, + "step": 12680 + }, + { + "epoch": 0.8799528138227742, + "grad_norm": 1.627586007118225, + "learning_rate": 0.00018166034509885606, + "loss": 0.5373, + "step": 12681 + }, + { + "epoch": 0.880022205259871, + "grad_norm": 1.6653881072998047, + "learning_rate": 0.00018165710191015582, + "loss": 0.571, + "step": 12682 + }, + { + "epoch": 0.8800915966969676, + "grad_norm": 3.339479684829712, + "learning_rate": 0.00018165385846367362, + "loss": 0.7763, + "step": 12683 + }, + { + "epoch": 0.8801609881340643, + "grad_norm": 2.1336474418640137, + "learning_rate": 0.00018165061475941964, + "loss": 0.7886, + "step": 12684 + }, + { + "epoch": 0.8802303795711609, + "grad_norm": 1.8559110164642334, + "learning_rate": 0.00018164737079740415, + "loss": 0.691, + "step": 12685 + }, + { + "epoch": 0.8802997710082576, + "grad_norm": 1.9505800008773804, + "learning_rate": 0.00018164412657763738, + "loss": 0.8269, + "step": 12686 + }, + { + "epoch": 0.8803691624453542, + "grad_norm": 2.2989017963409424, + "learning_rate": 0.0001816408821001296, + "loss": 0.8808, + "step": 12687 + }, + { + "epoch": 0.880438553882451, + "grad_norm": 2.235044002532959, + "learning_rate": 0.000181637637364891, + "loss": 0.8894, + "step": 12688 + }, + { + "epoch": 0.8805079453195476, + "grad_norm": 2.767123222351074, + "learning_rate": 0.00018163439237193184, + "loss": 1.084, + "step": 12689 + }, + { + "epoch": 0.8805773367566442, + "grad_norm": 2.7513086795806885, + "learning_rate": 0.00018163114712126238, + "loss": 0.9003, + "step": 12690 + }, + { + "epoch": 0.8806467281937409, + "grad_norm": 1.853002905845642, + "learning_rate": 0.00018162790161289287, + "loss": 0.8819, + "step": 12691 + }, + { + "epoch": 0.8807161196308375, + "grad_norm": 1.78013014793396, + "learning_rate": 0.00018162465584683357, + "loss": 0.7089, + "step": 12692 + }, + { + "epoch": 0.8807855110679342, + "grad_norm": 2.2973828315734863, + "learning_rate": 0.00018162140982309465, + "loss": 0.9067, + "step": 12693 + }, + { + "epoch": 0.8808549025050308, + "grad_norm": 2.7523751258850098, + "learning_rate": 0.00018161816354168647, + "loss": 1.0407, + "step": 12694 + }, + { + "epoch": 0.8809242939421276, + "grad_norm": 2.256002426147461, + "learning_rate": 0.00018161491700261916, + "loss": 0.8943, + "step": 12695 + }, + { + "epoch": 0.8809936853792242, + "grad_norm": 1.7723125219345093, + "learning_rate": 0.00018161167020590307, + "loss": 0.7486, + "step": 12696 + }, + { + "epoch": 0.8810630768163209, + "grad_norm": 2.7639946937561035, + "learning_rate": 0.00018160842315154837, + "loss": 0.9963, + "step": 12697 + }, + { + "epoch": 0.8811324682534175, + "grad_norm": 3.04904842376709, + "learning_rate": 0.00018160517583956537, + "loss": 1.1179, + "step": 12698 + }, + { + "epoch": 0.8812018596905142, + "grad_norm": 1.8627158403396606, + "learning_rate": 0.00018160192826996427, + "loss": 0.8173, + "step": 12699 + }, + { + "epoch": 0.8812712511276108, + "grad_norm": 1.901469111442566, + "learning_rate": 0.00018159868044275536, + "loss": 0.7915, + "step": 12700 + }, + { + "epoch": 0.8813406425647076, + "grad_norm": 1.9866470098495483, + "learning_rate": 0.00018159543235794892, + "loss": 0.724, + "step": 12701 + }, + { + "epoch": 0.8814100340018042, + "grad_norm": 2.0858051776885986, + "learning_rate": 0.00018159218401555513, + "loss": 0.7759, + "step": 12702 + }, + { + "epoch": 0.8814794254389008, + "grad_norm": 2.276259183883667, + "learning_rate": 0.0001815889354155843, + "loss": 0.8657, + "step": 12703 + }, + { + "epoch": 0.8815488168759975, + "grad_norm": 1.766861081123352, + "learning_rate": 0.00018158568655804666, + "loss": 0.6896, + "step": 12704 + }, + { + "epoch": 0.8816182083130941, + "grad_norm": 2.0480291843414307, + "learning_rate": 0.00018158243744295245, + "loss": 0.7554, + "step": 12705 + }, + { + "epoch": 0.8816875997501908, + "grad_norm": 2.87650465965271, + "learning_rate": 0.00018157918807031197, + "loss": 0.8676, + "step": 12706 + }, + { + "epoch": 0.8817569911872875, + "grad_norm": 1.529594898223877, + "learning_rate": 0.00018157593844013548, + "loss": 0.7586, + "step": 12707 + }, + { + "epoch": 0.8818263826243842, + "grad_norm": 2.3644540309906006, + "learning_rate": 0.00018157268855243319, + "loss": 0.7249, + "step": 12708 + }, + { + "epoch": 0.8818957740614808, + "grad_norm": 3.697230577468872, + "learning_rate": 0.0001815694384072154, + "loss": 1.1778, + "step": 12709 + }, + { + "epoch": 0.8819651654985775, + "grad_norm": 2.080447196960449, + "learning_rate": 0.00018156618800449235, + "loss": 0.9017, + "step": 12710 + }, + { + "epoch": 0.8820345569356741, + "grad_norm": 2.5837960243225098, + "learning_rate": 0.0001815629373442743, + "loss": 0.8846, + "step": 12711 + }, + { + "epoch": 0.8821039483727708, + "grad_norm": 2.1760401725769043, + "learning_rate": 0.00018155968642657154, + "loss": 0.853, + "step": 12712 + }, + { + "epoch": 0.8821733398098675, + "grad_norm": 2.092843532562256, + "learning_rate": 0.0001815564352513943, + "loss": 0.6543, + "step": 12713 + }, + { + "epoch": 0.8822427312469642, + "grad_norm": 2.6339120864868164, + "learning_rate": 0.00018155318381875285, + "loss": 0.9154, + "step": 12714 + }, + { + "epoch": 0.8823121226840608, + "grad_norm": 4.654819965362549, + "learning_rate": 0.00018154993212865747, + "loss": 0.8276, + "step": 12715 + }, + { + "epoch": 0.8823815141211574, + "grad_norm": 2.5282092094421387, + "learning_rate": 0.0001815466801811184, + "loss": 0.9135, + "step": 12716 + }, + { + "epoch": 0.8824509055582541, + "grad_norm": 2.153064489364624, + "learning_rate": 0.00018154342797614594, + "loss": 0.8429, + "step": 12717 + }, + { + "epoch": 0.8825202969953507, + "grad_norm": 1.9197183847427368, + "learning_rate": 0.00018154017551375034, + "loss": 0.6143, + "step": 12718 + }, + { + "epoch": 0.8825896884324475, + "grad_norm": 2.3487701416015625, + "learning_rate": 0.00018153692279394186, + "loss": 1.2107, + "step": 12719 + }, + { + "epoch": 0.8826590798695441, + "grad_norm": 2.1768805980682373, + "learning_rate": 0.00018153366981673078, + "loss": 0.6589, + "step": 12720 + }, + { + "epoch": 0.8827284713066408, + "grad_norm": 2.3267006874084473, + "learning_rate": 0.00018153041658212736, + "loss": 0.7861, + "step": 12721 + }, + { + "epoch": 0.8827978627437374, + "grad_norm": 2.550407648086548, + "learning_rate": 0.00018152716309014183, + "loss": 0.9188, + "step": 12722 + }, + { + "epoch": 0.8828672541808341, + "grad_norm": 1.9044564962387085, + "learning_rate": 0.00018152390934078456, + "loss": 0.7059, + "step": 12723 + }, + { + "epoch": 0.8829366456179307, + "grad_norm": 1.788604497909546, + "learning_rate": 0.00018152065533406578, + "loss": 0.7137, + "step": 12724 + }, + { + "epoch": 0.8830060370550274, + "grad_norm": 2.2097134590148926, + "learning_rate": 0.0001815174010699957, + "loss": 0.7058, + "step": 12725 + }, + { + "epoch": 0.8830754284921241, + "grad_norm": 3.8530619144439697, + "learning_rate": 0.00018151414654858468, + "loss": 0.7344, + "step": 12726 + }, + { + "epoch": 0.8831448199292208, + "grad_norm": 1.8512057065963745, + "learning_rate": 0.00018151089176984294, + "loss": 0.7951, + "step": 12727 + }, + { + "epoch": 0.8832142113663174, + "grad_norm": 2.2243902683258057, + "learning_rate": 0.00018150763673378077, + "loss": 0.9839, + "step": 12728 + }, + { + "epoch": 0.8832836028034141, + "grad_norm": 1.912865400314331, + "learning_rate": 0.00018150438144040845, + "loss": 0.707, + "step": 12729 + }, + { + "epoch": 0.8833529942405107, + "grad_norm": 2.3187170028686523, + "learning_rate": 0.00018150112588973625, + "loss": 0.7265, + "step": 12730 + }, + { + "epoch": 0.8834223856776073, + "grad_norm": 1.900504469871521, + "learning_rate": 0.00018149787008177448, + "loss": 0.724, + "step": 12731 + }, + { + "epoch": 0.8834917771147041, + "grad_norm": 2.1051104068756104, + "learning_rate": 0.00018149461401653337, + "loss": 1.0128, + "step": 12732 + }, + { + "epoch": 0.8835611685518007, + "grad_norm": 2.194812774658203, + "learning_rate": 0.00018149135769402318, + "loss": 0.7983, + "step": 12733 + }, + { + "epoch": 0.8836305599888974, + "grad_norm": 2.6717774868011475, + "learning_rate": 0.00018148810111425428, + "loss": 0.955, + "step": 12734 + }, + { + "epoch": 0.883699951425994, + "grad_norm": 1.9834431409835815, + "learning_rate": 0.0001814848442772369, + "loss": 0.8267, + "step": 12735 + }, + { + "epoch": 0.8837693428630907, + "grad_norm": 2.0241904258728027, + "learning_rate": 0.0001814815871829813, + "loss": 0.6679, + "step": 12736 + }, + { + "epoch": 0.8838387343001873, + "grad_norm": 2.201843500137329, + "learning_rate": 0.0001814783298314978, + "loss": 0.7679, + "step": 12737 + }, + { + "epoch": 0.8839081257372841, + "grad_norm": 2.371840238571167, + "learning_rate": 0.00018147507222279662, + "loss": 0.6371, + "step": 12738 + }, + { + "epoch": 0.8839775171743807, + "grad_norm": 2.3618526458740234, + "learning_rate": 0.00018147181435688813, + "loss": 0.7954, + "step": 12739 + }, + { + "epoch": 0.8840469086114774, + "grad_norm": 2.0682179927825928, + "learning_rate": 0.00018146855623378256, + "loss": 0.7941, + "step": 12740 + }, + { + "epoch": 0.884116300048574, + "grad_norm": 1.908178687095642, + "learning_rate": 0.00018146529785349023, + "loss": 0.7244, + "step": 12741 + }, + { + "epoch": 0.8841856914856707, + "grad_norm": 2.5795373916625977, + "learning_rate": 0.00018146203921602137, + "loss": 0.8163, + "step": 12742 + }, + { + "epoch": 0.8842550829227673, + "grad_norm": 2.0589683055877686, + "learning_rate": 0.00018145878032138635, + "loss": 0.873, + "step": 12743 + }, + { + "epoch": 0.8843244743598639, + "grad_norm": 2.346733570098877, + "learning_rate": 0.0001814555211695954, + "loss": 0.719, + "step": 12744 + }, + { + "epoch": 0.8843938657969607, + "grad_norm": 2.5241172313690186, + "learning_rate": 0.0001814522617606588, + "loss": 1.1086, + "step": 12745 + }, + { + "epoch": 0.8844632572340573, + "grad_norm": 2.0106699466705322, + "learning_rate": 0.00018144900209458685, + "loss": 0.8078, + "step": 12746 + }, + { + "epoch": 0.884532648671154, + "grad_norm": 2.299276828765869, + "learning_rate": 0.0001814457421713899, + "loss": 0.8968, + "step": 12747 + }, + { + "epoch": 0.8846020401082506, + "grad_norm": 2.3191163539886475, + "learning_rate": 0.00018144248199107813, + "loss": 0.7299, + "step": 12748 + }, + { + "epoch": 0.8846714315453473, + "grad_norm": 2.5473363399505615, + "learning_rate": 0.00018143922155366195, + "loss": 1.0684, + "step": 12749 + }, + { + "epoch": 0.8847408229824439, + "grad_norm": 2.2890098094940186, + "learning_rate": 0.00018143596085915155, + "loss": 1.0731, + "step": 12750 + }, + { + "epoch": 0.8848102144195407, + "grad_norm": 2.504568576812744, + "learning_rate": 0.0001814326999075573, + "loss": 0.9518, + "step": 12751 + }, + { + "epoch": 0.8848796058566373, + "grad_norm": 2.1518046855926514, + "learning_rate": 0.00018142943869888946, + "loss": 0.7597, + "step": 12752 + }, + { + "epoch": 0.884948997293734, + "grad_norm": 1.891343593597412, + "learning_rate": 0.00018142617723315832, + "loss": 0.6987, + "step": 12753 + }, + { + "epoch": 0.8850183887308306, + "grad_norm": 2.3266525268554688, + "learning_rate": 0.0001814229155103742, + "loss": 0.9105, + "step": 12754 + }, + { + "epoch": 0.8850877801679273, + "grad_norm": 2.4388904571533203, + "learning_rate": 0.00018141965353054735, + "loss": 0.8479, + "step": 12755 + }, + { + "epoch": 0.8851571716050239, + "grad_norm": 1.996400237083435, + "learning_rate": 0.00018141639129368812, + "loss": 0.8002, + "step": 12756 + }, + { + "epoch": 0.8852265630421206, + "grad_norm": 2.0722646713256836, + "learning_rate": 0.0001814131287998068, + "loss": 0.9266, + "step": 12757 + }, + { + "epoch": 0.8852959544792173, + "grad_norm": 1.8640806674957275, + "learning_rate": 0.00018140986604891368, + "loss": 0.7445, + "step": 12758 + }, + { + "epoch": 0.8853653459163139, + "grad_norm": 2.7816600799560547, + "learning_rate": 0.00018140660304101903, + "loss": 0.9829, + "step": 12759 + }, + { + "epoch": 0.8854347373534106, + "grad_norm": 2.372399091720581, + "learning_rate": 0.0001814033397761332, + "loss": 0.8688, + "step": 12760 + }, + { + "epoch": 0.8855041287905072, + "grad_norm": 2.359273672103882, + "learning_rate": 0.00018140007625426648, + "loss": 0.7649, + "step": 12761 + }, + { + "epoch": 0.8855735202276039, + "grad_norm": 2.0340869426727295, + "learning_rate": 0.00018139681247542915, + "loss": 0.6898, + "step": 12762 + }, + { + "epoch": 0.8856429116647005, + "grad_norm": 2.3387954235076904, + "learning_rate": 0.00018139354843963153, + "loss": 0.8383, + "step": 12763 + }, + { + "epoch": 0.8857123031017973, + "grad_norm": 2.1565511226654053, + "learning_rate": 0.00018139028414688392, + "loss": 0.6031, + "step": 12764 + }, + { + "epoch": 0.8857816945388939, + "grad_norm": 1.7521198987960815, + "learning_rate": 0.00018138701959719662, + "loss": 0.7161, + "step": 12765 + }, + { + "epoch": 0.8858510859759906, + "grad_norm": 2.5015199184417725, + "learning_rate": 0.00018138375479057994, + "loss": 1.064, + "step": 12766 + }, + { + "epoch": 0.8859204774130872, + "grad_norm": 2.3766844272613525, + "learning_rate": 0.00018138048972704418, + "loss": 0.9514, + "step": 12767 + }, + { + "epoch": 0.8859898688501839, + "grad_norm": 1.9670145511627197, + "learning_rate": 0.00018137722440659968, + "loss": 0.7537, + "step": 12768 + }, + { + "epoch": 0.8860592602872805, + "grad_norm": 2.7169933319091797, + "learning_rate": 0.00018137395882925671, + "loss": 1.0367, + "step": 12769 + }, + { + "epoch": 0.8861286517243772, + "grad_norm": 1.384742021560669, + "learning_rate": 0.0001813706929950256, + "loss": 0.5655, + "step": 12770 + }, + { + "epoch": 0.8861980431614739, + "grad_norm": 2.267434597015381, + "learning_rate": 0.00018136742690391666, + "loss": 0.7702, + "step": 12771 + }, + { + "epoch": 0.8862674345985705, + "grad_norm": 1.8982340097427368, + "learning_rate": 0.0001813641605559402, + "loss": 0.8654, + "step": 12772 + }, + { + "epoch": 0.8863368260356672, + "grad_norm": 2.350872039794922, + "learning_rate": 0.00018136089395110648, + "loss": 0.8764, + "step": 12773 + }, + { + "epoch": 0.8864062174727638, + "grad_norm": 2.5001068115234375, + "learning_rate": 0.0001813576270894259, + "loss": 0.6522, + "step": 12774 + }, + { + "epoch": 0.8864756089098605, + "grad_norm": 2.1417548656463623, + "learning_rate": 0.0001813543599709087, + "loss": 0.838, + "step": 12775 + }, + { + "epoch": 0.8865450003469572, + "grad_norm": 2.512220859527588, + "learning_rate": 0.00018135109259556525, + "loss": 0.6841, + "step": 12776 + }, + { + "epoch": 0.8866143917840539, + "grad_norm": 2.1055028438568115, + "learning_rate": 0.00018134782496340583, + "loss": 0.7632, + "step": 12777 + }, + { + "epoch": 0.8866837832211505, + "grad_norm": 2.2203729152679443, + "learning_rate": 0.00018134455707444078, + "loss": 0.7891, + "step": 12778 + }, + { + "epoch": 0.8867531746582472, + "grad_norm": 1.4922043085098267, + "learning_rate": 0.0001813412889286804, + "loss": 0.5775, + "step": 12779 + }, + { + "epoch": 0.8868225660953438, + "grad_norm": 2.0157485008239746, + "learning_rate": 0.00018133802052613498, + "loss": 0.6136, + "step": 12780 + }, + { + "epoch": 0.8868919575324405, + "grad_norm": 2.211700677871704, + "learning_rate": 0.00018133475186681488, + "loss": 0.8108, + "step": 12781 + }, + { + "epoch": 0.8869613489695372, + "grad_norm": 1.8410670757293701, + "learning_rate": 0.0001813314829507304, + "loss": 0.65, + "step": 12782 + }, + { + "epoch": 0.8870307404066338, + "grad_norm": 2.7415566444396973, + "learning_rate": 0.00018132821377789187, + "loss": 0.7683, + "step": 12783 + }, + { + "epoch": 0.8871001318437305, + "grad_norm": 2.746328353881836, + "learning_rate": 0.0001813249443483096, + "loss": 0.7293, + "step": 12784 + }, + { + "epoch": 0.8871695232808271, + "grad_norm": 2.098543882369995, + "learning_rate": 0.00018132167466199394, + "loss": 0.7019, + "step": 12785 + }, + { + "epoch": 0.8872389147179238, + "grad_norm": 2.6178877353668213, + "learning_rate": 0.00018131840471895515, + "loss": 0.8629, + "step": 12786 + }, + { + "epoch": 0.8873083061550204, + "grad_norm": 2.305246591567993, + "learning_rate": 0.00018131513451920365, + "loss": 0.8461, + "step": 12787 + }, + { + "epoch": 0.8873776975921172, + "grad_norm": 2.4183714389801025, + "learning_rate": 0.00018131186406274963, + "loss": 0.7087, + "step": 12788 + }, + { + "epoch": 0.8874470890292138, + "grad_norm": 2.716844081878662, + "learning_rate": 0.00018130859334960353, + "loss": 1.069, + "step": 12789 + }, + { + "epoch": 0.8875164804663105, + "grad_norm": 2.6423733234405518, + "learning_rate": 0.00018130532237977564, + "loss": 0.925, + "step": 12790 + }, + { + "epoch": 0.8875858719034071, + "grad_norm": 2.5836918354034424, + "learning_rate": 0.00018130205115327624, + "loss": 0.8178, + "step": 12791 + }, + { + "epoch": 0.8876552633405038, + "grad_norm": 1.7073014974594116, + "learning_rate": 0.00018129877967011573, + "loss": 0.6533, + "step": 12792 + }, + { + "epoch": 0.8877246547776004, + "grad_norm": 2.2810428142547607, + "learning_rate": 0.00018129550793030442, + "loss": 0.7849, + "step": 12793 + }, + { + "epoch": 0.8877940462146972, + "grad_norm": 2.471832275390625, + "learning_rate": 0.0001812922359338526, + "loss": 0.9102, + "step": 12794 + }, + { + "epoch": 0.8878634376517938, + "grad_norm": 2.241511106491089, + "learning_rate": 0.00018128896368077062, + "loss": 0.766, + "step": 12795 + }, + { + "epoch": 0.8879328290888905, + "grad_norm": 1.910827398300171, + "learning_rate": 0.00018128569117106883, + "loss": 0.7747, + "step": 12796 + }, + { + "epoch": 0.8880022205259871, + "grad_norm": 1.8801387548446655, + "learning_rate": 0.00018128241840475754, + "loss": 0.8099, + "step": 12797 + }, + { + "epoch": 0.8880716119630837, + "grad_norm": 3.595400810241699, + "learning_rate": 0.00018127914538184706, + "loss": 0.8498, + "step": 12798 + }, + { + "epoch": 0.8881410034001804, + "grad_norm": 2.54111647605896, + "learning_rate": 0.00018127587210234778, + "loss": 0.7246, + "step": 12799 + }, + { + "epoch": 0.888210394837277, + "grad_norm": 1.784690022468567, + "learning_rate": 0.00018127259856627, + "loss": 0.6386, + "step": 12800 + }, + { + "epoch": 0.8882797862743738, + "grad_norm": 2.1457912921905518, + "learning_rate": 0.00018126932477362403, + "loss": 0.8132, + "step": 12801 + }, + { + "epoch": 0.8883491777114704, + "grad_norm": 1.9685425758361816, + "learning_rate": 0.00018126605072442024, + "loss": 0.9894, + "step": 12802 + }, + { + "epoch": 0.8884185691485671, + "grad_norm": 2.0747885704040527, + "learning_rate": 0.00018126277641866895, + "loss": 0.7842, + "step": 12803 + }, + { + "epoch": 0.8884879605856637, + "grad_norm": 2.0537781715393066, + "learning_rate": 0.00018125950185638053, + "loss": 0.6501, + "step": 12804 + }, + { + "epoch": 0.8885573520227604, + "grad_norm": 3.5555148124694824, + "learning_rate": 0.0001812562270375653, + "loss": 0.9461, + "step": 12805 + }, + { + "epoch": 0.888626743459857, + "grad_norm": 2.1347615718841553, + "learning_rate": 0.00018125295196223356, + "loss": 0.8143, + "step": 12806 + }, + { + "epoch": 0.8886961348969538, + "grad_norm": 2.108506202697754, + "learning_rate": 0.00018124967663039567, + "loss": 0.8137, + "step": 12807 + }, + { + "epoch": 0.8887655263340504, + "grad_norm": 1.8547322750091553, + "learning_rate": 0.00018124640104206198, + "loss": 0.715, + "step": 12808 + }, + { + "epoch": 0.8888349177711471, + "grad_norm": 2.2560250759124756, + "learning_rate": 0.00018124312519724286, + "loss": 0.6256, + "step": 12809 + }, + { + "epoch": 0.8889043092082437, + "grad_norm": 2.549222230911255, + "learning_rate": 0.00018123984909594859, + "loss": 0.9122, + "step": 12810 + }, + { + "epoch": 0.8889737006453403, + "grad_norm": 2.111262083053589, + "learning_rate": 0.00018123657273818954, + "loss": 0.8781, + "step": 12811 + }, + { + "epoch": 0.889043092082437, + "grad_norm": 2.565086603164673, + "learning_rate": 0.00018123329612397608, + "loss": 1.0145, + "step": 12812 + }, + { + "epoch": 0.8891124835195336, + "grad_norm": 2.7357256412506104, + "learning_rate": 0.0001812300192533185, + "loss": 0.7371, + "step": 12813 + }, + { + "epoch": 0.8891818749566304, + "grad_norm": 1.6783932447433472, + "learning_rate": 0.0001812267421262272, + "loss": 0.7526, + "step": 12814 + }, + { + "epoch": 0.889251266393727, + "grad_norm": 1.9601917266845703, + "learning_rate": 0.0001812234647427125, + "loss": 0.842, + "step": 12815 + }, + { + "epoch": 0.8893206578308237, + "grad_norm": 2.2427210807800293, + "learning_rate": 0.0001812201871027847, + "loss": 1.1121, + "step": 12816 + }, + { + "epoch": 0.8893900492679203, + "grad_norm": 3.4099180698394775, + "learning_rate": 0.00018121690920645423, + "loss": 0.9218, + "step": 12817 + }, + { + "epoch": 0.889459440705017, + "grad_norm": 1.6670442819595337, + "learning_rate": 0.00018121363105373138, + "loss": 0.5846, + "step": 12818 + }, + { + "epoch": 0.8895288321421136, + "grad_norm": 2.9187729358673096, + "learning_rate": 0.00018121035264462653, + "loss": 1.01, + "step": 12819 + }, + { + "epoch": 0.8895982235792104, + "grad_norm": 2.9326095581054688, + "learning_rate": 0.00018120707397915005, + "loss": 1.1899, + "step": 12820 + }, + { + "epoch": 0.889667615016307, + "grad_norm": 2.876291275024414, + "learning_rate": 0.00018120379505731223, + "loss": 0.86, + "step": 12821 + }, + { + "epoch": 0.8897370064534037, + "grad_norm": 1.7909907102584839, + "learning_rate": 0.00018120051587912345, + "loss": 0.623, + "step": 12822 + }, + { + "epoch": 0.8898063978905003, + "grad_norm": 2.238081216812134, + "learning_rate": 0.00018119723644459407, + "loss": 0.7787, + "step": 12823 + }, + { + "epoch": 0.8898757893275969, + "grad_norm": 2.5279946327209473, + "learning_rate": 0.00018119395675373444, + "loss": 0.9146, + "step": 12824 + }, + { + "epoch": 0.8899451807646936, + "grad_norm": 1.5922373533248901, + "learning_rate": 0.0001811906768065549, + "loss": 0.6553, + "step": 12825 + }, + { + "epoch": 0.8900145722017903, + "grad_norm": 2.2920353412628174, + "learning_rate": 0.00018118739660306582, + "loss": 0.8774, + "step": 12826 + }, + { + "epoch": 0.890083963638887, + "grad_norm": 1.7966445684432983, + "learning_rate": 0.0001811841161432775, + "loss": 0.7018, + "step": 12827 + }, + { + "epoch": 0.8901533550759836, + "grad_norm": 1.8556383848190308, + "learning_rate": 0.00018118083542720042, + "loss": 0.9598, + "step": 12828 + }, + { + "epoch": 0.8902227465130803, + "grad_norm": 1.9231786727905273, + "learning_rate": 0.0001811775544548448, + "loss": 0.8196, + "step": 12829 + }, + { + "epoch": 0.8902921379501769, + "grad_norm": 2.1486988067626953, + "learning_rate": 0.00018117427322622112, + "loss": 0.6575, + "step": 12830 + }, + { + "epoch": 0.8903615293872736, + "grad_norm": 1.7413511276245117, + "learning_rate": 0.00018117099174133963, + "loss": 0.6567, + "step": 12831 + }, + { + "epoch": 0.8904309208243703, + "grad_norm": 1.6954907178878784, + "learning_rate": 0.00018116771000021076, + "loss": 0.8372, + "step": 12832 + }, + { + "epoch": 0.890500312261467, + "grad_norm": 1.7703460454940796, + "learning_rate": 0.00018116442800284482, + "loss": 0.7203, + "step": 12833 + }, + { + "epoch": 0.8905697036985636, + "grad_norm": 2.289088249206543, + "learning_rate": 0.00018116114574925224, + "loss": 0.941, + "step": 12834 + }, + { + "epoch": 0.8906390951356603, + "grad_norm": 2.6858279705047607, + "learning_rate": 0.0001811578632394433, + "loss": 0.7533, + "step": 12835 + }, + { + "epoch": 0.8907084865727569, + "grad_norm": 1.9288389682769775, + "learning_rate": 0.0001811545804734284, + "loss": 0.7802, + "step": 12836 + }, + { + "epoch": 0.8907778780098535, + "grad_norm": 1.840222954750061, + "learning_rate": 0.0001811512974512179, + "loss": 0.7469, + "step": 12837 + }, + { + "epoch": 0.8908472694469503, + "grad_norm": 2.861605644226074, + "learning_rate": 0.0001811480141728222, + "loss": 0.8347, + "step": 12838 + }, + { + "epoch": 0.8909166608840469, + "grad_norm": 1.9147839546203613, + "learning_rate": 0.00018114473063825164, + "loss": 0.7175, + "step": 12839 + }, + { + "epoch": 0.8909860523211436, + "grad_norm": 2.830561637878418, + "learning_rate": 0.00018114144684751654, + "loss": 0.8612, + "step": 12840 + }, + { + "epoch": 0.8910554437582402, + "grad_norm": 2.3447718620300293, + "learning_rate": 0.00018113816280062738, + "loss": 0.7228, + "step": 12841 + }, + { + "epoch": 0.8911248351953369, + "grad_norm": 2.2251765727996826, + "learning_rate": 0.00018113487849759436, + "loss": 0.6284, + "step": 12842 + }, + { + "epoch": 0.8911942266324335, + "grad_norm": 2.038465738296509, + "learning_rate": 0.000181131593938428, + "loss": 0.7142, + "step": 12843 + }, + { + "epoch": 0.8912636180695303, + "grad_norm": 2.144684314727783, + "learning_rate": 0.00018112830912313862, + "loss": 0.801, + "step": 12844 + }, + { + "epoch": 0.8913330095066269, + "grad_norm": 2.393383264541626, + "learning_rate": 0.0001811250240517365, + "loss": 0.9034, + "step": 12845 + }, + { + "epoch": 0.8914024009437236, + "grad_norm": 1.6922131776809692, + "learning_rate": 0.0001811217387242322, + "loss": 0.7545, + "step": 12846 + }, + { + "epoch": 0.8914717923808202, + "grad_norm": 2.315903902053833, + "learning_rate": 0.0001811184531406359, + "loss": 0.7189, + "step": 12847 + }, + { + "epoch": 0.8915411838179169, + "grad_norm": 1.9851760864257812, + "learning_rate": 0.0001811151673009581, + "loss": 0.7906, + "step": 12848 + }, + { + "epoch": 0.8916105752550135, + "grad_norm": 2.261802911758423, + "learning_rate": 0.0001811118812052091, + "loss": 0.5037, + "step": 12849 + }, + { + "epoch": 0.8916799666921102, + "grad_norm": 1.9901645183563232, + "learning_rate": 0.00018110859485339931, + "loss": 0.8476, + "step": 12850 + }, + { + "epoch": 0.8917493581292069, + "grad_norm": 1.8207658529281616, + "learning_rate": 0.00018110530824553913, + "loss": 0.757, + "step": 12851 + }, + { + "epoch": 0.8918187495663035, + "grad_norm": 2.0952718257904053, + "learning_rate": 0.00018110202138163888, + "loss": 0.7371, + "step": 12852 + }, + { + "epoch": 0.8918881410034002, + "grad_norm": 2.1952290534973145, + "learning_rate": 0.00018109873426170892, + "loss": 0.864, + "step": 12853 + }, + { + "epoch": 0.8919575324404968, + "grad_norm": 1.9287382364273071, + "learning_rate": 0.0001810954468857597, + "loss": 0.7977, + "step": 12854 + }, + { + "epoch": 0.8920269238775935, + "grad_norm": 2.2018206119537354, + "learning_rate": 0.00018109215925380157, + "loss": 0.8797, + "step": 12855 + }, + { + "epoch": 0.8920963153146901, + "grad_norm": 1.978287696838379, + "learning_rate": 0.0001810888713658449, + "loss": 0.9257, + "step": 12856 + }, + { + "epoch": 0.8921657067517869, + "grad_norm": 2.2991020679473877, + "learning_rate": 0.00018108558322190005, + "loss": 1.0878, + "step": 12857 + }, + { + "epoch": 0.8922350981888835, + "grad_norm": 1.8800886869430542, + "learning_rate": 0.00018108229482197746, + "loss": 0.8932, + "step": 12858 + }, + { + "epoch": 0.8923044896259802, + "grad_norm": 1.959303379058838, + "learning_rate": 0.00018107900616608742, + "loss": 0.9013, + "step": 12859 + }, + { + "epoch": 0.8923738810630768, + "grad_norm": 1.8436084985733032, + "learning_rate": 0.00018107571725424043, + "loss": 0.7882, + "step": 12860 + }, + { + "epoch": 0.8924432725001735, + "grad_norm": 1.9553494453430176, + "learning_rate": 0.00018107242808644672, + "loss": 0.8054, + "step": 12861 + }, + { + "epoch": 0.8925126639372701, + "grad_norm": 4.587706565856934, + "learning_rate": 0.00018106913866271684, + "loss": 0.8511, + "step": 12862 + }, + { + "epoch": 0.8925820553743669, + "grad_norm": 1.9730266332626343, + "learning_rate": 0.00018106584898306105, + "loss": 0.8391, + "step": 12863 + }, + { + "epoch": 0.8926514468114635, + "grad_norm": 2.0295798778533936, + "learning_rate": 0.0001810625590474898, + "loss": 0.8374, + "step": 12864 + }, + { + "epoch": 0.8927208382485601, + "grad_norm": 1.9120440483093262, + "learning_rate": 0.00018105926885601345, + "loss": 0.6084, + "step": 12865 + }, + { + "epoch": 0.8927902296856568, + "grad_norm": 1.6153621673583984, + "learning_rate": 0.0001810559784086424, + "loss": 0.6103, + "step": 12866 + }, + { + "epoch": 0.8928596211227534, + "grad_norm": 2.430004119873047, + "learning_rate": 0.00018105268770538704, + "loss": 0.97, + "step": 12867 + }, + { + "epoch": 0.8929290125598501, + "grad_norm": 2.1224112510681152, + "learning_rate": 0.00018104939674625772, + "loss": 0.9381, + "step": 12868 + }, + { + "epoch": 0.8929984039969467, + "grad_norm": 2.076392412185669, + "learning_rate": 0.00018104610553126485, + "loss": 0.8981, + "step": 12869 + }, + { + "epoch": 0.8930677954340435, + "grad_norm": 1.5720620155334473, + "learning_rate": 0.00018104281406041887, + "loss": 0.7519, + "step": 12870 + }, + { + "epoch": 0.8931371868711401, + "grad_norm": 1.879548192024231, + "learning_rate": 0.0001810395223337301, + "loss": 0.6726, + "step": 12871 + }, + { + "epoch": 0.8932065783082368, + "grad_norm": 2.1111860275268555, + "learning_rate": 0.00018103623035120898, + "loss": 0.9267, + "step": 12872 + }, + { + "epoch": 0.8932759697453334, + "grad_norm": 3.829789161682129, + "learning_rate": 0.00018103293811286587, + "loss": 0.7495, + "step": 12873 + }, + { + "epoch": 0.8933453611824301, + "grad_norm": 2.3082435131073, + "learning_rate": 0.00018102964561871117, + "loss": 0.8588, + "step": 12874 + }, + { + "epoch": 0.8934147526195267, + "grad_norm": 1.6899006366729736, + "learning_rate": 0.00018102635286875528, + "loss": 0.4804, + "step": 12875 + }, + { + "epoch": 0.8934841440566235, + "grad_norm": 2.1431591510772705, + "learning_rate": 0.0001810230598630086, + "loss": 0.7445, + "step": 12876 + }, + { + "epoch": 0.8935535354937201, + "grad_norm": 2.0894131660461426, + "learning_rate": 0.00018101976660148152, + "loss": 0.8273, + "step": 12877 + }, + { + "epoch": 0.8936229269308167, + "grad_norm": 1.949476957321167, + "learning_rate": 0.00018101647308418445, + "loss": 0.6087, + "step": 12878 + }, + { + "epoch": 0.8936923183679134, + "grad_norm": 1.8081822395324707, + "learning_rate": 0.00018101317931112776, + "loss": 0.7033, + "step": 12879 + }, + { + "epoch": 0.89376170980501, + "grad_norm": 2.3214986324310303, + "learning_rate": 0.00018100988528232186, + "loss": 1.0171, + "step": 12880 + }, + { + "epoch": 0.8938311012421067, + "grad_norm": 2.18573260307312, + "learning_rate": 0.00018100659099777716, + "loss": 0.4945, + "step": 12881 + }, + { + "epoch": 0.8939004926792034, + "grad_norm": 2.42136549949646, + "learning_rate": 0.00018100329645750405, + "loss": 0.8051, + "step": 12882 + }, + { + "epoch": 0.8939698841163001, + "grad_norm": 2.1512060165405273, + "learning_rate": 0.00018100000166151292, + "loss": 0.7271, + "step": 12883 + }, + { + "epoch": 0.8940392755533967, + "grad_norm": 1.8036996126174927, + "learning_rate": 0.0001809967066098142, + "loss": 0.6257, + "step": 12884 + }, + { + "epoch": 0.8941086669904934, + "grad_norm": 1.7724964618682861, + "learning_rate": 0.00018099341130241826, + "loss": 0.7847, + "step": 12885 + }, + { + "epoch": 0.89417805842759, + "grad_norm": 2.843845844268799, + "learning_rate": 0.00018099011573933553, + "loss": 0.6882, + "step": 12886 + }, + { + "epoch": 0.8942474498646867, + "grad_norm": 1.5212041139602661, + "learning_rate": 0.00018098681992057637, + "loss": 0.5562, + "step": 12887 + }, + { + "epoch": 0.8943168413017833, + "grad_norm": 3.641148328781128, + "learning_rate": 0.00018098352384615126, + "loss": 0.8092, + "step": 12888 + }, + { + "epoch": 0.8943862327388801, + "grad_norm": 1.681068778038025, + "learning_rate": 0.00018098022751607054, + "loss": 0.7953, + "step": 12889 + }, + { + "epoch": 0.8944556241759767, + "grad_norm": 2.4361798763275146, + "learning_rate": 0.00018097693093034462, + "loss": 0.9358, + "step": 12890 + }, + { + "epoch": 0.8945250156130733, + "grad_norm": 1.9882234334945679, + "learning_rate": 0.00018097363408898395, + "loss": 0.6087, + "step": 12891 + }, + { + "epoch": 0.89459440705017, + "grad_norm": 2.5685763359069824, + "learning_rate": 0.0001809703369919989, + "loss": 0.8906, + "step": 12892 + }, + { + "epoch": 0.8946637984872666, + "grad_norm": 2.16953706741333, + "learning_rate": 0.0001809670396393999, + "loss": 1.0318, + "step": 12893 + }, + { + "epoch": 0.8947331899243633, + "grad_norm": 2.425687551498413, + "learning_rate": 0.00018096374203119733, + "loss": 1.0621, + "step": 12894 + }, + { + "epoch": 0.89480258136146, + "grad_norm": 2.5394389629364014, + "learning_rate": 0.00018096044416740165, + "loss": 0.9813, + "step": 12895 + }, + { + "epoch": 0.8948719727985567, + "grad_norm": 2.6915054321289062, + "learning_rate": 0.00018095714604802324, + "loss": 0.8796, + "step": 12896 + }, + { + "epoch": 0.8949413642356533, + "grad_norm": 1.8656532764434814, + "learning_rate": 0.00018095384767307247, + "loss": 0.6655, + "step": 12897 + }, + { + "epoch": 0.89501075567275, + "grad_norm": 2.0643186569213867, + "learning_rate": 0.00018095054904255982, + "loss": 0.633, + "step": 12898 + }, + { + "epoch": 0.8950801471098466, + "grad_norm": 2.2868688106536865, + "learning_rate": 0.00018094725015649568, + "loss": 1.0495, + "step": 12899 + }, + { + "epoch": 0.8951495385469433, + "grad_norm": 2.066697835922241, + "learning_rate": 0.0001809439510148905, + "loss": 0.8892, + "step": 12900 + }, + { + "epoch": 0.89521892998404, + "grad_norm": 2.116199254989624, + "learning_rate": 0.00018094065161775459, + "loss": 0.666, + "step": 12901 + }, + { + "epoch": 0.8952883214211367, + "grad_norm": 3.067660331726074, + "learning_rate": 0.00018093735196509847, + "loss": 1.028, + "step": 12902 + }, + { + "epoch": 0.8953577128582333, + "grad_norm": 2.442241907119751, + "learning_rate": 0.00018093405205693252, + "loss": 0.7341, + "step": 12903 + }, + { + "epoch": 0.8954271042953299, + "grad_norm": 2.270305871963501, + "learning_rate": 0.00018093075189326715, + "loss": 0.5956, + "step": 12904 + }, + { + "epoch": 0.8954964957324266, + "grad_norm": 2.6773109436035156, + "learning_rate": 0.00018092745147411279, + "loss": 0.7291, + "step": 12905 + }, + { + "epoch": 0.8955658871695232, + "grad_norm": 2.108123540878296, + "learning_rate": 0.00018092415079947984, + "loss": 0.737, + "step": 12906 + }, + { + "epoch": 0.89563527860662, + "grad_norm": 2.0893218517303467, + "learning_rate": 0.00018092084986937875, + "loss": 0.9783, + "step": 12907 + }, + { + "epoch": 0.8957046700437166, + "grad_norm": 2.5726981163024902, + "learning_rate": 0.00018091754868381993, + "loss": 1.1805, + "step": 12908 + }, + { + "epoch": 0.8957740614808133, + "grad_norm": 2.4299585819244385, + "learning_rate": 0.0001809142472428138, + "loss": 0.7068, + "step": 12909 + }, + { + "epoch": 0.8958434529179099, + "grad_norm": 2.1807432174682617, + "learning_rate": 0.00018091094554637075, + "loss": 0.5371, + "step": 12910 + }, + { + "epoch": 0.8959128443550066, + "grad_norm": 2.1097733974456787, + "learning_rate": 0.00018090764359450127, + "loss": 0.847, + "step": 12911 + }, + { + "epoch": 0.8959822357921032, + "grad_norm": 2.277506113052368, + "learning_rate": 0.00018090434138721574, + "loss": 0.6919, + "step": 12912 + }, + { + "epoch": 0.8960516272292, + "grad_norm": 2.453155040740967, + "learning_rate": 0.00018090103892452457, + "loss": 0.9416, + "step": 12913 + }, + { + "epoch": 0.8961210186662966, + "grad_norm": 1.903777003288269, + "learning_rate": 0.00018089773620643822, + "loss": 0.7692, + "step": 12914 + }, + { + "epoch": 0.8961904101033933, + "grad_norm": 1.9855846166610718, + "learning_rate": 0.0001808944332329671, + "loss": 0.6202, + "step": 12915 + }, + { + "epoch": 0.8962598015404899, + "grad_norm": 2.1583423614501953, + "learning_rate": 0.00018089113000412164, + "loss": 0.7002, + "step": 12916 + }, + { + "epoch": 0.8963291929775866, + "grad_norm": 2.081967830657959, + "learning_rate": 0.00018088782651991226, + "loss": 0.788, + "step": 12917 + }, + { + "epoch": 0.8963985844146832, + "grad_norm": 2.4484667778015137, + "learning_rate": 0.00018088452278034942, + "loss": 0.8738, + "step": 12918 + }, + { + "epoch": 0.8964679758517798, + "grad_norm": 2.0511043071746826, + "learning_rate": 0.0001808812187854435, + "loss": 0.8242, + "step": 12919 + }, + { + "epoch": 0.8965373672888766, + "grad_norm": 1.807228684425354, + "learning_rate": 0.00018087791453520494, + "loss": 0.8146, + "step": 12920 + }, + { + "epoch": 0.8966067587259732, + "grad_norm": 2.192586660385132, + "learning_rate": 0.00018087461002964423, + "loss": 0.7608, + "step": 12921 + }, + { + "epoch": 0.8966761501630699, + "grad_norm": 2.4848268032073975, + "learning_rate": 0.00018087130526877175, + "loss": 0.9646, + "step": 12922 + }, + { + "epoch": 0.8967455416001665, + "grad_norm": 2.016866445541382, + "learning_rate": 0.00018086800025259793, + "loss": 0.8356, + "step": 12923 + }, + { + "epoch": 0.8968149330372632, + "grad_norm": 2.8467190265655518, + "learning_rate": 0.00018086469498113322, + "loss": 0.9858, + "step": 12924 + }, + { + "epoch": 0.8968843244743598, + "grad_norm": 2.6273393630981445, + "learning_rate": 0.00018086138945438805, + "loss": 0.973, + "step": 12925 + }, + { + "epoch": 0.8969537159114566, + "grad_norm": 2.700279712677002, + "learning_rate": 0.00018085808367237287, + "loss": 0.8824, + "step": 12926 + }, + { + "epoch": 0.8970231073485532, + "grad_norm": 1.91414213180542, + "learning_rate": 0.0001808547776350981, + "loss": 0.8836, + "step": 12927 + }, + { + "epoch": 0.8970924987856499, + "grad_norm": 2.2753827571868896, + "learning_rate": 0.00018085147134257415, + "loss": 0.7707, + "step": 12928 + }, + { + "epoch": 0.8971618902227465, + "grad_norm": 1.8939366340637207, + "learning_rate": 0.00018084816479481148, + "loss": 0.5093, + "step": 12929 + }, + { + "epoch": 0.8972312816598432, + "grad_norm": 2.10254168510437, + "learning_rate": 0.00018084485799182057, + "loss": 1.036, + "step": 12930 + }, + { + "epoch": 0.8973006730969398, + "grad_norm": 2.474191427230835, + "learning_rate": 0.0001808415509336118, + "loss": 0.9233, + "step": 12931 + }, + { + "epoch": 0.8973700645340364, + "grad_norm": 2.0425684452056885, + "learning_rate": 0.00018083824362019564, + "loss": 0.8454, + "step": 12932 + }, + { + "epoch": 0.8974394559711332, + "grad_norm": 2.322953939437866, + "learning_rate": 0.00018083493605158251, + "loss": 0.868, + "step": 12933 + }, + { + "epoch": 0.8975088474082298, + "grad_norm": 2.337179183959961, + "learning_rate": 0.00018083162822778287, + "loss": 0.9367, + "step": 12934 + }, + { + "epoch": 0.8975782388453265, + "grad_norm": 2.314840793609619, + "learning_rate": 0.0001808283201488072, + "loss": 0.863, + "step": 12935 + }, + { + "epoch": 0.8976476302824231, + "grad_norm": 3.6357674598693848, + "learning_rate": 0.00018082501181466585, + "loss": 0.7651, + "step": 12936 + }, + { + "epoch": 0.8977170217195198, + "grad_norm": 1.6848305463790894, + "learning_rate": 0.00018082170322536936, + "loss": 0.5738, + "step": 12937 + }, + { + "epoch": 0.8977864131566164, + "grad_norm": 1.99782395362854, + "learning_rate": 0.00018081839438092808, + "loss": 0.7617, + "step": 12938 + }, + { + "epoch": 0.8978558045937132, + "grad_norm": 2.2717204093933105, + "learning_rate": 0.00018081508528135253, + "loss": 0.8884, + "step": 12939 + }, + { + "epoch": 0.8979251960308098, + "grad_norm": 1.9044371843338013, + "learning_rate": 0.00018081177592665313, + "loss": 0.567, + "step": 12940 + }, + { + "epoch": 0.8979945874679065, + "grad_norm": 1.9298791885375977, + "learning_rate": 0.00018080846631684033, + "loss": 0.7696, + "step": 12941 + }, + { + "epoch": 0.8980639789050031, + "grad_norm": 3.1583166122436523, + "learning_rate": 0.0001808051564519246, + "loss": 0.7004, + "step": 12942 + }, + { + "epoch": 0.8981333703420998, + "grad_norm": 2.448831081390381, + "learning_rate": 0.00018080184633191632, + "loss": 0.8353, + "step": 12943 + }, + { + "epoch": 0.8982027617791964, + "grad_norm": 2.4959092140197754, + "learning_rate": 0.000180798535956826, + "loss": 0.7551, + "step": 12944 + }, + { + "epoch": 0.8982721532162931, + "grad_norm": 2.7405319213867188, + "learning_rate": 0.0001807952253266641, + "loss": 0.7841, + "step": 12945 + }, + { + "epoch": 0.8983415446533898, + "grad_norm": 2.1791813373565674, + "learning_rate": 0.00018079191444144102, + "loss": 0.8024, + "step": 12946 + }, + { + "epoch": 0.8984109360904864, + "grad_norm": 2.087698221206665, + "learning_rate": 0.00018078860330116725, + "loss": 0.8745, + "step": 12947 + }, + { + "epoch": 0.8984803275275831, + "grad_norm": 2.1059229373931885, + "learning_rate": 0.00018078529190585326, + "loss": 0.8829, + "step": 12948 + }, + { + "epoch": 0.8985497189646797, + "grad_norm": 2.361262559890747, + "learning_rate": 0.00018078198025550944, + "loss": 0.6622, + "step": 12949 + }, + { + "epoch": 0.8986191104017764, + "grad_norm": 1.9574618339538574, + "learning_rate": 0.0001807786683501463, + "loss": 0.6966, + "step": 12950 + }, + { + "epoch": 0.898688501838873, + "grad_norm": 2.845177412033081, + "learning_rate": 0.00018077535618977425, + "loss": 0.8035, + "step": 12951 + }, + { + "epoch": 0.8987578932759698, + "grad_norm": 3.155965805053711, + "learning_rate": 0.00018077204377440383, + "loss": 0.9002, + "step": 12952 + }, + { + "epoch": 0.8988272847130664, + "grad_norm": 3.3126189708709717, + "learning_rate": 0.00018076873110404537, + "loss": 1.0502, + "step": 12953 + }, + { + "epoch": 0.8988966761501631, + "grad_norm": 2.296645402908325, + "learning_rate": 0.00018076541817870943, + "loss": 0.952, + "step": 12954 + }, + { + "epoch": 0.8989660675872597, + "grad_norm": 2.4018373489379883, + "learning_rate": 0.00018076210499840643, + "loss": 0.7503, + "step": 12955 + }, + { + "epoch": 0.8990354590243564, + "grad_norm": 2.1398448944091797, + "learning_rate": 0.00018075879156314684, + "loss": 0.6156, + "step": 12956 + }, + { + "epoch": 0.899104850461453, + "grad_norm": 2.7964928150177, + "learning_rate": 0.00018075547787294111, + "loss": 1.0249, + "step": 12957 + }, + { + "epoch": 0.8991742418985497, + "grad_norm": 2.396422863006592, + "learning_rate": 0.00018075216392779973, + "loss": 0.9513, + "step": 12958 + }, + { + "epoch": 0.8992436333356464, + "grad_norm": 2.6117453575134277, + "learning_rate": 0.0001807488497277331, + "loss": 0.8069, + "step": 12959 + }, + { + "epoch": 0.899313024772743, + "grad_norm": 2.5907583236694336, + "learning_rate": 0.00018074553527275172, + "loss": 0.7643, + "step": 12960 + }, + { + "epoch": 0.8993824162098397, + "grad_norm": 2.267948865890503, + "learning_rate": 0.00018074222056286608, + "loss": 0.7377, + "step": 12961 + }, + { + "epoch": 0.8994518076469363, + "grad_norm": 2.118445873260498, + "learning_rate": 0.0001807389055980866, + "loss": 0.9705, + "step": 12962 + }, + { + "epoch": 0.899521199084033, + "grad_norm": 2.4042069911956787, + "learning_rate": 0.00018073559037842378, + "loss": 0.8473, + "step": 12963 + }, + { + "epoch": 0.8995905905211297, + "grad_norm": 1.888630747795105, + "learning_rate": 0.00018073227490388805, + "loss": 0.8017, + "step": 12964 + }, + { + "epoch": 0.8996599819582264, + "grad_norm": 1.4637120962142944, + "learning_rate": 0.0001807289591744899, + "loss": 0.5584, + "step": 12965 + }, + { + "epoch": 0.899729373395323, + "grad_norm": 1.679092526435852, + "learning_rate": 0.0001807256431902398, + "loss": 0.5551, + "step": 12966 + }, + { + "epoch": 0.8997987648324197, + "grad_norm": 1.669195294380188, + "learning_rate": 0.00018072232695114823, + "loss": 0.6799, + "step": 12967 + }, + { + "epoch": 0.8998681562695163, + "grad_norm": 2.5811212062835693, + "learning_rate": 0.00018071901045722556, + "loss": 0.9803, + "step": 12968 + }, + { + "epoch": 0.899937547706613, + "grad_norm": 1.7160634994506836, + "learning_rate": 0.00018071569370848242, + "loss": 0.611, + "step": 12969 + }, + { + "epoch": 0.9000069391437097, + "grad_norm": 1.9837768077850342, + "learning_rate": 0.00018071237670492916, + "loss": 1.1303, + "step": 12970 + }, + { + "epoch": 0.9000763305808064, + "grad_norm": 1.8084721565246582, + "learning_rate": 0.0001807090594465763, + "loss": 0.7198, + "step": 12971 + }, + { + "epoch": 0.900145722017903, + "grad_norm": 1.8938237428665161, + "learning_rate": 0.0001807057419334343, + "loss": 0.9541, + "step": 12972 + }, + { + "epoch": 0.9002151134549996, + "grad_norm": 1.8647215366363525, + "learning_rate": 0.00018070242416551367, + "loss": 0.789, + "step": 12973 + }, + { + "epoch": 0.9002845048920963, + "grad_norm": 2.38143253326416, + "learning_rate": 0.00018069910614282482, + "loss": 0.7567, + "step": 12974 + }, + { + "epoch": 0.9003538963291929, + "grad_norm": 2.4264066219329834, + "learning_rate": 0.00018069578786537825, + "loss": 0.6363, + "step": 12975 + }, + { + "epoch": 0.9004232877662897, + "grad_norm": 2.0768206119537354, + "learning_rate": 0.00018069246933318444, + "loss": 1.0603, + "step": 12976 + }, + { + "epoch": 0.9004926792033863, + "grad_norm": 2.31744647026062, + "learning_rate": 0.0001806891505462539, + "loss": 0.948, + "step": 12977 + }, + { + "epoch": 0.900562070640483, + "grad_norm": 2.1147491931915283, + "learning_rate": 0.00018068583150459704, + "loss": 0.9095, + "step": 12978 + }, + { + "epoch": 0.9006314620775796, + "grad_norm": 2.072242498397827, + "learning_rate": 0.0001806825122082244, + "loss": 0.6317, + "step": 12979 + }, + { + "epoch": 0.9007008535146763, + "grad_norm": 1.9563357830047607, + "learning_rate": 0.00018067919265714636, + "loss": 0.7735, + "step": 12980 + }, + { + "epoch": 0.9007702449517729, + "grad_norm": 2.0976128578186035, + "learning_rate": 0.00018067587285137355, + "loss": 0.7882, + "step": 12981 + }, + { + "epoch": 0.9008396363888697, + "grad_norm": 1.9659242630004883, + "learning_rate": 0.00018067255279091633, + "loss": 0.7127, + "step": 12982 + }, + { + "epoch": 0.9009090278259663, + "grad_norm": 2.4926745891571045, + "learning_rate": 0.00018066923247578523, + "loss": 0.921, + "step": 12983 + }, + { + "epoch": 0.900978419263063, + "grad_norm": 1.958718180656433, + "learning_rate": 0.0001806659119059907, + "loss": 0.9535, + "step": 12984 + }, + { + "epoch": 0.9010478107001596, + "grad_norm": 2.847778558731079, + "learning_rate": 0.00018066259108154327, + "loss": 0.8728, + "step": 12985 + }, + { + "epoch": 0.9011172021372562, + "grad_norm": 2.101840019226074, + "learning_rate": 0.0001806592700024534, + "loss": 0.767, + "step": 12986 + }, + { + "epoch": 0.9011865935743529, + "grad_norm": 3.0642924308776855, + "learning_rate": 0.00018065594866873157, + "loss": 0.839, + "step": 12987 + }, + { + "epoch": 0.9012559850114495, + "grad_norm": 2.210251569747925, + "learning_rate": 0.00018065262708038827, + "loss": 0.705, + "step": 12988 + }, + { + "epoch": 0.9013253764485463, + "grad_norm": 2.793435573577881, + "learning_rate": 0.00018064930523743397, + "loss": 0.7943, + "step": 12989 + }, + { + "epoch": 0.9013947678856429, + "grad_norm": 2.1377763748168945, + "learning_rate": 0.0001806459831398792, + "loss": 0.7544, + "step": 12990 + }, + { + "epoch": 0.9014641593227396, + "grad_norm": 2.172619104385376, + "learning_rate": 0.00018064266078773438, + "loss": 0.67, + "step": 12991 + }, + { + "epoch": 0.9015335507598362, + "grad_norm": 2.469024419784546, + "learning_rate": 0.00018063933818101004, + "loss": 0.6756, + "step": 12992 + }, + { + "epoch": 0.9016029421969329, + "grad_norm": 2.3353917598724365, + "learning_rate": 0.00018063601531971668, + "loss": 0.9896, + "step": 12993 + }, + { + "epoch": 0.9016723336340295, + "grad_norm": 2.360396385192871, + "learning_rate": 0.00018063269220386477, + "loss": 0.7188, + "step": 12994 + }, + { + "epoch": 0.9017417250711263, + "grad_norm": 2.5954360961914062, + "learning_rate": 0.00018062936883346482, + "loss": 0.8405, + "step": 12995 + }, + { + "epoch": 0.9018111165082229, + "grad_norm": 4.225437164306641, + "learning_rate": 0.00018062604520852727, + "loss": 0.8345, + "step": 12996 + }, + { + "epoch": 0.9018805079453196, + "grad_norm": 2.8841240406036377, + "learning_rate": 0.00018062272132906272, + "loss": 0.984, + "step": 12997 + }, + { + "epoch": 0.9019498993824162, + "grad_norm": 1.9024611711502075, + "learning_rate": 0.00018061939719508154, + "loss": 0.878, + "step": 12998 + }, + { + "epoch": 0.9020192908195128, + "grad_norm": 1.722341537475586, + "learning_rate": 0.00018061607280659428, + "loss": 0.7422, + "step": 12999 + }, + { + "epoch": 0.9020886822566095, + "grad_norm": 2.1937403678894043, + "learning_rate": 0.0001806127481636114, + "loss": 1.0625, + "step": 13000 + }, + { + "epoch": 0.9021580736937062, + "grad_norm": 2.627758741378784, + "learning_rate": 0.0001806094232661435, + "loss": 0.8563, + "step": 13001 + }, + { + "epoch": 0.9022274651308029, + "grad_norm": 2.0784082412719727, + "learning_rate": 0.00018060609811420094, + "loss": 0.9004, + "step": 13002 + }, + { + "epoch": 0.9022968565678995, + "grad_norm": 2.696906805038452, + "learning_rate": 0.00018060277270779432, + "loss": 0.9051, + "step": 13003 + }, + { + "epoch": 0.9023662480049962, + "grad_norm": 2.5479471683502197, + "learning_rate": 0.0001805994470469341, + "loss": 0.9694, + "step": 13004 + }, + { + "epoch": 0.9024356394420928, + "grad_norm": 2.293870449066162, + "learning_rate": 0.00018059612113163076, + "loss": 1.2186, + "step": 13005 + }, + { + "epoch": 0.9025050308791895, + "grad_norm": 2.3043758869171143, + "learning_rate": 0.00018059279496189482, + "loss": 0.823, + "step": 13006 + }, + { + "epoch": 0.9025744223162862, + "grad_norm": 3.8263933658599854, + "learning_rate": 0.00018058946853773678, + "loss": 0.8244, + "step": 13007 + }, + { + "epoch": 0.9026438137533829, + "grad_norm": 2.6979074478149414, + "learning_rate": 0.00018058614185916712, + "loss": 1.0801, + "step": 13008 + }, + { + "epoch": 0.9027132051904795, + "grad_norm": 1.7986470460891724, + "learning_rate": 0.00018058281492619636, + "loss": 0.7011, + "step": 13009 + }, + { + "epoch": 0.9027825966275762, + "grad_norm": 2.5502116680145264, + "learning_rate": 0.00018057948773883504, + "loss": 0.86, + "step": 13010 + }, + { + "epoch": 0.9028519880646728, + "grad_norm": 2.670269012451172, + "learning_rate": 0.00018057616029709358, + "loss": 0.7197, + "step": 13011 + }, + { + "epoch": 0.9029213795017694, + "grad_norm": 2.3561041355133057, + "learning_rate": 0.00018057283260098256, + "loss": 0.7209, + "step": 13012 + }, + { + "epoch": 0.9029907709388661, + "grad_norm": 2.266296148300171, + "learning_rate": 0.00018056950465051248, + "loss": 0.6728, + "step": 13013 + }, + { + "epoch": 0.9030601623759628, + "grad_norm": 2.195366859436035, + "learning_rate": 0.00018056617644569378, + "loss": 0.8042, + "step": 13014 + }, + { + "epoch": 0.9031295538130595, + "grad_norm": 3.1154496669769287, + "learning_rate": 0.00018056284798653702, + "loss": 0.9673, + "step": 13015 + }, + { + "epoch": 0.9031989452501561, + "grad_norm": 2.413381814956665, + "learning_rate": 0.0001805595192730527, + "loss": 0.832, + "step": 13016 + }, + { + "epoch": 0.9032683366872528, + "grad_norm": 2.864196538925171, + "learning_rate": 0.0001805561903052513, + "loss": 0.8924, + "step": 13017 + }, + { + "epoch": 0.9033377281243494, + "grad_norm": 1.8191300630569458, + "learning_rate": 0.00018055286108314338, + "loss": 0.7632, + "step": 13018 + }, + { + "epoch": 0.9034071195614461, + "grad_norm": 2.2797794342041016, + "learning_rate": 0.00018054953160673943, + "loss": 1.0171, + "step": 13019 + }, + { + "epoch": 0.9034765109985428, + "grad_norm": 2.205977201461792, + "learning_rate": 0.00018054620187604994, + "loss": 0.6175, + "step": 13020 + }, + { + "epoch": 0.9035459024356395, + "grad_norm": 2.6467819213867188, + "learning_rate": 0.0001805428718910854, + "loss": 0.7076, + "step": 13021 + }, + { + "epoch": 0.9036152938727361, + "grad_norm": 2.1249282360076904, + "learning_rate": 0.0001805395416518564, + "loss": 0.7061, + "step": 13022 + }, + { + "epoch": 0.9036846853098328, + "grad_norm": 2.370896339416504, + "learning_rate": 0.00018053621115837341, + "loss": 0.7308, + "step": 13023 + }, + { + "epoch": 0.9037540767469294, + "grad_norm": 2.3180649280548096, + "learning_rate": 0.00018053288041064696, + "loss": 0.8606, + "step": 13024 + }, + { + "epoch": 0.9038234681840261, + "grad_norm": 2.3575336933135986, + "learning_rate": 0.0001805295494086875, + "loss": 0.9667, + "step": 13025 + }, + { + "epoch": 0.9038928596211228, + "grad_norm": 1.4001243114471436, + "learning_rate": 0.00018052621815250563, + "loss": 0.5681, + "step": 13026 + }, + { + "epoch": 0.9039622510582194, + "grad_norm": 3.363269805908203, + "learning_rate": 0.00018052288664211185, + "loss": 0.8118, + "step": 13027 + }, + { + "epoch": 0.9040316424953161, + "grad_norm": 2.725820541381836, + "learning_rate": 0.00018051955487751663, + "loss": 0.8657, + "step": 13028 + }, + { + "epoch": 0.9041010339324127, + "grad_norm": 3.839994192123413, + "learning_rate": 0.00018051622285873052, + "loss": 0.815, + "step": 13029 + }, + { + "epoch": 0.9041704253695094, + "grad_norm": 2.3354909420013428, + "learning_rate": 0.00018051289058576404, + "loss": 0.8604, + "step": 13030 + }, + { + "epoch": 0.904239816806606, + "grad_norm": 1.9964929819107056, + "learning_rate": 0.0001805095580586277, + "loss": 0.7428, + "step": 13031 + }, + { + "epoch": 0.9043092082437028, + "grad_norm": 1.9119912385940552, + "learning_rate": 0.00018050622527733206, + "loss": 0.631, + "step": 13032 + }, + { + "epoch": 0.9043785996807994, + "grad_norm": 2.8870537281036377, + "learning_rate": 0.00018050289224188758, + "loss": 1.0752, + "step": 13033 + }, + { + "epoch": 0.9044479911178961, + "grad_norm": 1.8966783285140991, + "learning_rate": 0.0001804995589523048, + "loss": 0.7445, + "step": 13034 + }, + { + "epoch": 0.9045173825549927, + "grad_norm": 1.9779702425003052, + "learning_rate": 0.00018049622540859426, + "loss": 0.8457, + "step": 13035 + }, + { + "epoch": 0.9045867739920894, + "grad_norm": 2.1227996349334717, + "learning_rate": 0.00018049289161076648, + "loss": 0.9197, + "step": 13036 + }, + { + "epoch": 0.904656165429186, + "grad_norm": 2.297941207885742, + "learning_rate": 0.00018048955755883196, + "loss": 0.7696, + "step": 13037 + }, + { + "epoch": 0.9047255568662828, + "grad_norm": 1.8757622241973877, + "learning_rate": 0.00018048622325280128, + "loss": 0.7384, + "step": 13038 + }, + { + "epoch": 0.9047949483033794, + "grad_norm": 2.479050874710083, + "learning_rate": 0.0001804828886926849, + "loss": 0.8122, + "step": 13039 + }, + { + "epoch": 0.904864339740476, + "grad_norm": 1.7776881456375122, + "learning_rate": 0.0001804795538784934, + "loss": 0.7121, + "step": 13040 + }, + { + "epoch": 0.9049337311775727, + "grad_norm": 1.9120861291885376, + "learning_rate": 0.00018047621881023726, + "loss": 0.7597, + "step": 13041 + }, + { + "epoch": 0.9050031226146693, + "grad_norm": 2.3166115283966064, + "learning_rate": 0.00018047288348792707, + "loss": 0.7506, + "step": 13042 + }, + { + "epoch": 0.905072514051766, + "grad_norm": 2.1571128368377686, + "learning_rate": 0.0001804695479115733, + "loss": 0.8132, + "step": 13043 + }, + { + "epoch": 0.9051419054888626, + "grad_norm": 1.5190889835357666, + "learning_rate": 0.0001804662120811865, + "loss": 0.7059, + "step": 13044 + }, + { + "epoch": 0.9052112969259594, + "grad_norm": 2.8361752033233643, + "learning_rate": 0.0001804628759967772, + "loss": 0.7217, + "step": 13045 + }, + { + "epoch": 0.905280688363056, + "grad_norm": 1.9657509326934814, + "learning_rate": 0.00018045953965835598, + "loss": 0.6657, + "step": 13046 + }, + { + "epoch": 0.9053500798001527, + "grad_norm": 2.203376531600952, + "learning_rate": 0.00018045620306593327, + "loss": 1.1038, + "step": 13047 + }, + { + "epoch": 0.9054194712372493, + "grad_norm": 2.2513697147369385, + "learning_rate": 0.0001804528662195197, + "loss": 0.776, + "step": 13048 + }, + { + "epoch": 0.905488862674346, + "grad_norm": 2.3396658897399902, + "learning_rate": 0.00018044952911912577, + "loss": 0.9904, + "step": 13049 + }, + { + "epoch": 0.9055582541114426, + "grad_norm": 2.090146064758301, + "learning_rate": 0.000180446191764762, + "loss": 0.8256, + "step": 13050 + }, + { + "epoch": 0.9056276455485394, + "grad_norm": 2.397414207458496, + "learning_rate": 0.00018044285415643892, + "loss": 0.9426, + "step": 13051 + }, + { + "epoch": 0.905697036985636, + "grad_norm": 3.011188507080078, + "learning_rate": 0.0001804395162941671, + "loss": 0.6516, + "step": 13052 + }, + { + "epoch": 0.9057664284227326, + "grad_norm": 2.7432522773742676, + "learning_rate": 0.00018043617817795703, + "loss": 0.8424, + "step": 13053 + }, + { + "epoch": 0.9058358198598293, + "grad_norm": 1.8115845918655396, + "learning_rate": 0.0001804328398078193, + "loss": 0.7684, + "step": 13054 + }, + { + "epoch": 0.9059052112969259, + "grad_norm": 1.5222280025482178, + "learning_rate": 0.00018042950118376446, + "loss": 0.6639, + "step": 13055 + }, + { + "epoch": 0.9059746027340226, + "grad_norm": 2.2503020763397217, + "learning_rate": 0.00018042616230580296, + "loss": 0.7649, + "step": 13056 + }, + { + "epoch": 0.9060439941711192, + "grad_norm": 1.5356500148773193, + "learning_rate": 0.00018042282317394542, + "loss": 0.6893, + "step": 13057 + }, + { + "epoch": 0.906113385608216, + "grad_norm": 2.3855481147766113, + "learning_rate": 0.00018041948378820236, + "loss": 0.9957, + "step": 13058 + }, + { + "epoch": 0.9061827770453126, + "grad_norm": 2.375760078430176, + "learning_rate": 0.00018041614414858432, + "loss": 0.598, + "step": 13059 + }, + { + "epoch": 0.9062521684824093, + "grad_norm": 2.9350883960723877, + "learning_rate": 0.00018041280425510183, + "loss": 0.9019, + "step": 13060 + }, + { + "epoch": 0.9063215599195059, + "grad_norm": 2.209588050842285, + "learning_rate": 0.00018040946410776546, + "loss": 0.8761, + "step": 13061 + }, + { + "epoch": 0.9063909513566026, + "grad_norm": 2.6746292114257812, + "learning_rate": 0.00018040612370658575, + "loss": 1.1776, + "step": 13062 + }, + { + "epoch": 0.9064603427936992, + "grad_norm": 1.8424885272979736, + "learning_rate": 0.00018040278305157323, + "loss": 0.6969, + "step": 13063 + }, + { + "epoch": 0.906529734230796, + "grad_norm": 2.2370269298553467, + "learning_rate": 0.00018039944214273844, + "loss": 0.7162, + "step": 13064 + }, + { + "epoch": 0.9065991256678926, + "grad_norm": 3.586980104446411, + "learning_rate": 0.00018039610098009195, + "loss": 1.12, + "step": 13065 + }, + { + "epoch": 0.9066685171049892, + "grad_norm": 2.633715867996216, + "learning_rate": 0.00018039275956364427, + "loss": 0.9295, + "step": 13066 + }, + { + "epoch": 0.9067379085420859, + "grad_norm": 2.536939859390259, + "learning_rate": 0.00018038941789340603, + "loss": 0.7457, + "step": 13067 + }, + { + "epoch": 0.9068072999791825, + "grad_norm": 3.024235486984253, + "learning_rate": 0.00018038607596938767, + "loss": 0.9627, + "step": 13068 + }, + { + "epoch": 0.9068766914162792, + "grad_norm": 2.0088486671447754, + "learning_rate": 0.00018038273379159984, + "loss": 1.0489, + "step": 13069 + }, + { + "epoch": 0.9069460828533759, + "grad_norm": 2.2322890758514404, + "learning_rate": 0.00018037939136005302, + "loss": 0.7862, + "step": 13070 + }, + { + "epoch": 0.9070154742904726, + "grad_norm": 2.9425735473632812, + "learning_rate": 0.0001803760486747578, + "loss": 0.828, + "step": 13071 + }, + { + "epoch": 0.9070848657275692, + "grad_norm": 1.8998820781707764, + "learning_rate": 0.0001803727057357247, + "loss": 0.6666, + "step": 13072 + }, + { + "epoch": 0.9071542571646659, + "grad_norm": 2.579625129699707, + "learning_rate": 0.00018036936254296428, + "loss": 0.8982, + "step": 13073 + }, + { + "epoch": 0.9072236486017625, + "grad_norm": 1.6000560522079468, + "learning_rate": 0.00018036601909648714, + "loss": 0.698, + "step": 13074 + }, + { + "epoch": 0.9072930400388592, + "grad_norm": 1.8536514043807983, + "learning_rate": 0.00018036267539630382, + "loss": 0.6522, + "step": 13075 + }, + { + "epoch": 0.9073624314759559, + "grad_norm": 2.2354984283447266, + "learning_rate": 0.00018035933144242483, + "loss": 0.8056, + "step": 13076 + }, + { + "epoch": 0.9074318229130526, + "grad_norm": 1.8519072532653809, + "learning_rate": 0.00018035598723486075, + "loss": 0.7832, + "step": 13077 + }, + { + "epoch": 0.9075012143501492, + "grad_norm": 1.7890746593475342, + "learning_rate": 0.00018035264277362214, + "loss": 0.6575, + "step": 13078 + }, + { + "epoch": 0.9075706057872458, + "grad_norm": 2.376865863800049, + "learning_rate": 0.00018034929805871956, + "loss": 0.7742, + "step": 13079 + }, + { + "epoch": 0.9076399972243425, + "grad_norm": 1.3307796716690063, + "learning_rate": 0.0001803459530901636, + "loss": 0.6341, + "step": 13080 + }, + { + "epoch": 0.9077093886614391, + "grad_norm": 2.34480881690979, + "learning_rate": 0.00018034260786796473, + "loss": 0.8549, + "step": 13081 + }, + { + "epoch": 0.9077787800985359, + "grad_norm": 2.2307536602020264, + "learning_rate": 0.0001803392623921336, + "loss": 0.9453, + "step": 13082 + }, + { + "epoch": 0.9078481715356325, + "grad_norm": 2.605336904525757, + "learning_rate": 0.00018033591666268074, + "loss": 0.9357, + "step": 13083 + }, + { + "epoch": 0.9079175629727292, + "grad_norm": 2.6905131340026855, + "learning_rate": 0.00018033257067961672, + "loss": 1.1861, + "step": 13084 + }, + { + "epoch": 0.9079869544098258, + "grad_norm": 2.055359125137329, + "learning_rate": 0.00018032922444295209, + "loss": 0.8117, + "step": 13085 + }, + { + "epoch": 0.9080563458469225, + "grad_norm": 1.909535527229309, + "learning_rate": 0.00018032587795269742, + "loss": 0.6637, + "step": 13086 + }, + { + "epoch": 0.9081257372840191, + "grad_norm": 2.1887741088867188, + "learning_rate": 0.00018032253120886323, + "loss": 1.1147, + "step": 13087 + }, + { + "epoch": 0.9081951287211159, + "grad_norm": 2.0370259284973145, + "learning_rate": 0.00018031918421146018, + "loss": 0.9764, + "step": 13088 + }, + { + "epoch": 0.9082645201582125, + "grad_norm": 1.9553158283233643, + "learning_rate": 0.00018031583696049877, + "loss": 0.93, + "step": 13089 + }, + { + "epoch": 0.9083339115953092, + "grad_norm": 1.9477407932281494, + "learning_rate": 0.0001803124894559896, + "loss": 0.6314, + "step": 13090 + }, + { + "epoch": 0.9084033030324058, + "grad_norm": 3.567622184753418, + "learning_rate": 0.00018030914169794317, + "loss": 1.1435, + "step": 13091 + }, + { + "epoch": 0.9084726944695025, + "grad_norm": 2.0049631595611572, + "learning_rate": 0.00018030579368637011, + "loss": 0.905, + "step": 13092 + }, + { + "epoch": 0.9085420859065991, + "grad_norm": 2.4678077697753906, + "learning_rate": 0.00018030244542128102, + "loss": 1.1383, + "step": 13093 + }, + { + "epoch": 0.9086114773436957, + "grad_norm": 2.0580742359161377, + "learning_rate": 0.0001802990969026864, + "loss": 0.8533, + "step": 13094 + }, + { + "epoch": 0.9086808687807925, + "grad_norm": 1.799981951713562, + "learning_rate": 0.00018029574813059685, + "loss": 0.6938, + "step": 13095 + }, + { + "epoch": 0.9087502602178891, + "grad_norm": 2.084841728210449, + "learning_rate": 0.00018029239910502293, + "loss": 0.8229, + "step": 13096 + }, + { + "epoch": 0.9088196516549858, + "grad_norm": 2.1069629192352295, + "learning_rate": 0.00018028904982597522, + "loss": 0.8102, + "step": 13097 + }, + { + "epoch": 0.9088890430920824, + "grad_norm": 1.5426727533340454, + "learning_rate": 0.00018028570029346433, + "loss": 0.6113, + "step": 13098 + }, + { + "epoch": 0.9089584345291791, + "grad_norm": 3.146090269088745, + "learning_rate": 0.00018028235050750074, + "loss": 1.1066, + "step": 13099 + }, + { + "epoch": 0.9090278259662757, + "grad_norm": 1.8576818704605103, + "learning_rate": 0.0001802790004680951, + "loss": 0.5142, + "step": 13100 + }, + { + "epoch": 0.9090972174033725, + "grad_norm": 2.7540476322174072, + "learning_rate": 0.000180275650175258, + "loss": 0.7285, + "step": 13101 + }, + { + "epoch": 0.9091666088404691, + "grad_norm": 1.5152587890625, + "learning_rate": 0.000180272299629, + "loss": 0.545, + "step": 13102 + }, + { + "epoch": 0.9092360002775658, + "grad_norm": 2.621377944946289, + "learning_rate": 0.00018026894882933162, + "loss": 0.9381, + "step": 13103 + }, + { + "epoch": 0.9093053917146624, + "grad_norm": 2.3326175212860107, + "learning_rate": 0.0001802655977762635, + "loss": 0.8567, + "step": 13104 + }, + { + "epoch": 0.9093747831517591, + "grad_norm": 2.9836673736572266, + "learning_rate": 0.0001802622464698062, + "loss": 0.8693, + "step": 13105 + }, + { + "epoch": 0.9094441745888557, + "grad_norm": 2.0622341632843018, + "learning_rate": 0.0001802588949099703, + "loss": 1.0994, + "step": 13106 + }, + { + "epoch": 0.9095135660259523, + "grad_norm": 2.5283470153808594, + "learning_rate": 0.00018025554309676639, + "loss": 0.9102, + "step": 13107 + }, + { + "epoch": 0.9095829574630491, + "grad_norm": 1.9090030193328857, + "learning_rate": 0.000180252191030205, + "loss": 0.7365, + "step": 13108 + }, + { + "epoch": 0.9096523489001457, + "grad_norm": 2.172739028930664, + "learning_rate": 0.00018024883871029678, + "loss": 0.9353, + "step": 13109 + }, + { + "epoch": 0.9097217403372424, + "grad_norm": 1.747469186782837, + "learning_rate": 0.0001802454861370523, + "loss": 0.7711, + "step": 13110 + }, + { + "epoch": 0.909791131774339, + "grad_norm": 1.6668318510055542, + "learning_rate": 0.00018024213331048215, + "loss": 0.6132, + "step": 13111 + }, + { + "epoch": 0.9098605232114357, + "grad_norm": 1.6860179901123047, + "learning_rate": 0.00018023878023059684, + "loss": 0.822, + "step": 13112 + }, + { + "epoch": 0.9099299146485323, + "grad_norm": 2.191225528717041, + "learning_rate": 0.00018023542689740704, + "loss": 0.8497, + "step": 13113 + }, + { + "epoch": 0.9099993060856291, + "grad_norm": 1.8822189569473267, + "learning_rate": 0.00018023207331092328, + "loss": 0.8816, + "step": 13114 + }, + { + "epoch": 0.9100686975227257, + "grad_norm": 1.8992987871170044, + "learning_rate": 0.00018022871947115623, + "loss": 0.5992, + "step": 13115 + }, + { + "epoch": 0.9101380889598224, + "grad_norm": 1.7031867504119873, + "learning_rate": 0.00018022536537811636, + "loss": 0.6513, + "step": 13116 + }, + { + "epoch": 0.910207480396919, + "grad_norm": 2.2480392456054688, + "learning_rate": 0.00018022201103181436, + "loss": 0.961, + "step": 13117 + }, + { + "epoch": 0.9102768718340157, + "grad_norm": 1.8428226709365845, + "learning_rate": 0.00018021865643226074, + "loss": 0.6842, + "step": 13118 + }, + { + "epoch": 0.9103462632711123, + "grad_norm": 2.3666274547576904, + "learning_rate": 0.0001802153015794662, + "loss": 1.0229, + "step": 13119 + }, + { + "epoch": 0.910415654708209, + "grad_norm": 1.9212270975112915, + "learning_rate": 0.00018021194647344118, + "loss": 0.8181, + "step": 13120 + }, + { + "epoch": 0.9104850461453057, + "grad_norm": 1.9225687980651855, + "learning_rate": 0.0001802085911141964, + "loss": 0.8051, + "step": 13121 + }, + { + "epoch": 0.9105544375824023, + "grad_norm": 1.7452812194824219, + "learning_rate": 0.0001802052355017424, + "loss": 0.7432, + "step": 13122 + }, + { + "epoch": 0.910623829019499, + "grad_norm": 2.0275533199310303, + "learning_rate": 0.00018020187963608972, + "loss": 0.5784, + "step": 13123 + }, + { + "epoch": 0.9106932204565956, + "grad_norm": 1.966993808746338, + "learning_rate": 0.00018019852351724907, + "loss": 0.7404, + "step": 13124 + }, + { + "epoch": 0.9107626118936923, + "grad_norm": 1.8812710046768188, + "learning_rate": 0.00018019516714523097, + "loss": 0.7259, + "step": 13125 + }, + { + "epoch": 0.910832003330789, + "grad_norm": 1.9682117700576782, + "learning_rate": 0.00018019181052004603, + "loss": 0.7591, + "step": 13126 + }, + { + "epoch": 0.9109013947678857, + "grad_norm": 2.542614221572876, + "learning_rate": 0.00018018845364170484, + "loss": 0.8726, + "step": 13127 + }, + { + "epoch": 0.9109707862049823, + "grad_norm": 2.162719488143921, + "learning_rate": 0.00018018509651021805, + "loss": 0.8196, + "step": 13128 + }, + { + "epoch": 0.911040177642079, + "grad_norm": 3.121056079864502, + "learning_rate": 0.00018018173912559616, + "loss": 0.7083, + "step": 13129 + }, + { + "epoch": 0.9111095690791756, + "grad_norm": 2.219210386276245, + "learning_rate": 0.00018017838148784985, + "loss": 0.5695, + "step": 13130 + }, + { + "epoch": 0.9111789605162723, + "grad_norm": 2.084118604660034, + "learning_rate": 0.00018017502359698968, + "loss": 0.8218, + "step": 13131 + }, + { + "epoch": 0.911248351953369, + "grad_norm": 2.023538827896118, + "learning_rate": 0.00018017166545302625, + "loss": 0.9314, + "step": 13132 + }, + { + "epoch": 0.9113177433904656, + "grad_norm": 1.6946543455123901, + "learning_rate": 0.0001801683070559702, + "loss": 0.7311, + "step": 13133 + }, + { + "epoch": 0.9113871348275623, + "grad_norm": 2.0292489528656006, + "learning_rate": 0.0001801649484058321, + "loss": 0.6813, + "step": 13134 + }, + { + "epoch": 0.9114565262646589, + "grad_norm": 1.8016679286956787, + "learning_rate": 0.00018016158950262255, + "loss": 0.476, + "step": 13135 + }, + { + "epoch": 0.9115259177017556, + "grad_norm": 2.431946039199829, + "learning_rate": 0.0001801582303463522, + "loss": 0.8382, + "step": 13136 + }, + { + "epoch": 0.9115953091388522, + "grad_norm": 1.9869710206985474, + "learning_rate": 0.00018015487093703156, + "loss": 0.5728, + "step": 13137 + }, + { + "epoch": 0.911664700575949, + "grad_norm": 1.8926922082901, + "learning_rate": 0.0001801515112746713, + "loss": 0.947, + "step": 13138 + }, + { + "epoch": 0.9117340920130456, + "grad_norm": 2.194992780685425, + "learning_rate": 0.00018014815135928208, + "loss": 0.8861, + "step": 13139 + }, + { + "epoch": 0.9118034834501423, + "grad_norm": 1.9762758016586304, + "learning_rate": 0.00018014479119087437, + "loss": 0.7692, + "step": 13140 + }, + { + "epoch": 0.9118728748872389, + "grad_norm": 1.857753872871399, + "learning_rate": 0.00018014143076945888, + "loss": 0.8124, + "step": 13141 + }, + { + "epoch": 0.9119422663243356, + "grad_norm": 1.8918834924697876, + "learning_rate": 0.00018013807009504618, + "loss": 0.7976, + "step": 13142 + }, + { + "epoch": 0.9120116577614322, + "grad_norm": 3.0554730892181396, + "learning_rate": 0.00018013470916764695, + "loss": 0.6861, + "step": 13143 + }, + { + "epoch": 0.912081049198529, + "grad_norm": 2.2437844276428223, + "learning_rate": 0.00018013134798727166, + "loss": 0.8006, + "step": 13144 + }, + { + "epoch": 0.9121504406356256, + "grad_norm": 2.1629581451416016, + "learning_rate": 0.00018012798655393106, + "loss": 0.8189, + "step": 13145 + }, + { + "epoch": 0.9122198320727223, + "grad_norm": 1.5300955772399902, + "learning_rate": 0.00018012462486763567, + "loss": 0.6466, + "step": 13146 + }, + { + "epoch": 0.9122892235098189, + "grad_norm": 2.4115564823150635, + "learning_rate": 0.00018012126292839614, + "loss": 0.9799, + "step": 13147 + }, + { + "epoch": 0.9123586149469155, + "grad_norm": 1.454581379890442, + "learning_rate": 0.00018011790073622309, + "loss": 0.5798, + "step": 13148 + }, + { + "epoch": 0.9124280063840122, + "grad_norm": 1.7560038566589355, + "learning_rate": 0.00018011453829112713, + "loss": 0.5908, + "step": 13149 + }, + { + "epoch": 0.9124973978211088, + "grad_norm": 2.177260398864746, + "learning_rate": 0.00018011117559311885, + "loss": 0.8628, + "step": 13150 + }, + { + "epoch": 0.9125667892582056, + "grad_norm": 1.8289326429367065, + "learning_rate": 0.00018010781264220888, + "loss": 0.8002, + "step": 13151 + }, + { + "epoch": 0.9126361806953022, + "grad_norm": 1.6818681955337524, + "learning_rate": 0.00018010444943840782, + "loss": 0.5823, + "step": 13152 + }, + { + "epoch": 0.9127055721323989, + "grad_norm": 2.60282564163208, + "learning_rate": 0.00018010108598172635, + "loss": 0.8243, + "step": 13153 + }, + { + "epoch": 0.9127749635694955, + "grad_norm": 2.5522804260253906, + "learning_rate": 0.00018009772227217503, + "loss": 0.8911, + "step": 13154 + }, + { + "epoch": 0.9128443550065922, + "grad_norm": 2.23537540435791, + "learning_rate": 0.0001800943583097645, + "loss": 0.5557, + "step": 13155 + }, + { + "epoch": 0.9129137464436888, + "grad_norm": 1.6872869729995728, + "learning_rate": 0.00018009099409450535, + "loss": 0.7163, + "step": 13156 + }, + { + "epoch": 0.9129831378807856, + "grad_norm": 2.2459888458251953, + "learning_rate": 0.00018008762962640827, + "loss": 0.8135, + "step": 13157 + }, + { + "epoch": 0.9130525293178822, + "grad_norm": 1.945448637008667, + "learning_rate": 0.00018008426490548379, + "loss": 0.7674, + "step": 13158 + }, + { + "epoch": 0.9131219207549789, + "grad_norm": 2.2942607402801514, + "learning_rate": 0.0001800808999317426, + "loss": 0.7696, + "step": 13159 + }, + { + "epoch": 0.9131913121920755, + "grad_norm": 2.5684707164764404, + "learning_rate": 0.00018007753470519527, + "loss": 0.7334, + "step": 13160 + }, + { + "epoch": 0.9132607036291721, + "grad_norm": 1.659276008605957, + "learning_rate": 0.00018007416922585247, + "loss": 0.4136, + "step": 13161 + }, + { + "epoch": 0.9133300950662688, + "grad_norm": 2.4026901721954346, + "learning_rate": 0.00018007080349372484, + "loss": 0.9924, + "step": 13162 + }, + { + "epoch": 0.9133994865033654, + "grad_norm": 2.4201056957244873, + "learning_rate": 0.00018006743750882293, + "loss": 0.6311, + "step": 13163 + }, + { + "epoch": 0.9134688779404622, + "grad_norm": 1.7838116884231567, + "learning_rate": 0.00018006407127115742, + "loss": 0.7231, + "step": 13164 + }, + { + "epoch": 0.9135382693775588, + "grad_norm": 2.1878836154937744, + "learning_rate": 0.00018006070478073896, + "loss": 0.8495, + "step": 13165 + }, + { + "epoch": 0.9136076608146555, + "grad_norm": 2.5493834018707275, + "learning_rate": 0.00018005733803757811, + "loss": 0.7316, + "step": 13166 + }, + { + "epoch": 0.9136770522517521, + "grad_norm": 1.9983409643173218, + "learning_rate": 0.00018005397104168554, + "loss": 0.7441, + "step": 13167 + }, + { + "epoch": 0.9137464436888488, + "grad_norm": 2.4541571140289307, + "learning_rate": 0.00018005060379307187, + "loss": 0.7806, + "step": 13168 + }, + { + "epoch": 0.9138158351259454, + "grad_norm": 2.40722393989563, + "learning_rate": 0.00018004723629174774, + "loss": 0.7778, + "step": 13169 + }, + { + "epoch": 0.9138852265630422, + "grad_norm": 2.8053932189941406, + "learning_rate": 0.00018004386853772374, + "loss": 0.7945, + "step": 13170 + }, + { + "epoch": 0.9139546180001388, + "grad_norm": 2.5161540508270264, + "learning_rate": 0.0001800405005310106, + "loss": 0.9449, + "step": 13171 + }, + { + "epoch": 0.9140240094372355, + "grad_norm": 2.152803897857666, + "learning_rate": 0.0001800371322716188, + "loss": 1.0035, + "step": 13172 + }, + { + "epoch": 0.9140934008743321, + "grad_norm": 2.1264946460723877, + "learning_rate": 0.00018003376375955914, + "loss": 0.6209, + "step": 13173 + }, + { + "epoch": 0.9141627923114287, + "grad_norm": 2.192068576812744, + "learning_rate": 0.00018003039499484214, + "loss": 0.7439, + "step": 13174 + }, + { + "epoch": 0.9142321837485254, + "grad_norm": 1.5687803030014038, + "learning_rate": 0.00018002702597747844, + "loss": 0.5261, + "step": 13175 + }, + { + "epoch": 0.914301575185622, + "grad_norm": 2.5199615955352783, + "learning_rate": 0.00018002365670747873, + "loss": 0.9688, + "step": 13176 + }, + { + "epoch": 0.9143709666227188, + "grad_norm": 2.246533155441284, + "learning_rate": 0.00018002028718485362, + "loss": 0.8507, + "step": 13177 + }, + { + "epoch": 0.9144403580598154, + "grad_norm": 1.934830904006958, + "learning_rate": 0.00018001691740961376, + "loss": 0.9056, + "step": 13178 + }, + { + "epoch": 0.9145097494969121, + "grad_norm": 2.184995651245117, + "learning_rate": 0.00018001354738176977, + "loss": 0.8174, + "step": 13179 + }, + { + "epoch": 0.9145791409340087, + "grad_norm": 2.4362053871154785, + "learning_rate": 0.00018001017710133228, + "loss": 0.8823, + "step": 13180 + }, + { + "epoch": 0.9146485323711054, + "grad_norm": 2.915259599685669, + "learning_rate": 0.00018000680656831195, + "loss": 0.7786, + "step": 13181 + }, + { + "epoch": 0.914717923808202, + "grad_norm": 2.364426851272583, + "learning_rate": 0.00018000343578271938, + "loss": 0.8502, + "step": 13182 + }, + { + "epoch": 0.9147873152452988, + "grad_norm": 1.9458822011947632, + "learning_rate": 0.0001800000647445653, + "loss": 0.7937, + "step": 13183 + }, + { + "epoch": 0.9148567066823954, + "grad_norm": 1.6309775114059448, + "learning_rate": 0.0001799966934538603, + "loss": 0.6029, + "step": 13184 + }, + { + "epoch": 0.9149260981194921, + "grad_norm": 2.5172808170318604, + "learning_rate": 0.00017999332191061498, + "loss": 1.2457, + "step": 13185 + }, + { + "epoch": 0.9149954895565887, + "grad_norm": 1.8705801963806152, + "learning_rate": 0.00017998995011484005, + "loss": 0.6055, + "step": 13186 + }, + { + "epoch": 0.9150648809936853, + "grad_norm": 2.656656503677368, + "learning_rate": 0.0001799865780665461, + "loss": 0.7521, + "step": 13187 + }, + { + "epoch": 0.915134272430782, + "grad_norm": 2.2578651905059814, + "learning_rate": 0.00017998320576574384, + "loss": 0.7606, + "step": 13188 + }, + { + "epoch": 0.9152036638678787, + "grad_norm": 1.7932944297790527, + "learning_rate": 0.00017997983321244384, + "loss": 0.7479, + "step": 13189 + }, + { + "epoch": 0.9152730553049754, + "grad_norm": 2.235337257385254, + "learning_rate": 0.00017997646040665682, + "loss": 0.8821, + "step": 13190 + }, + { + "epoch": 0.915342446742072, + "grad_norm": 2.099327325820923, + "learning_rate": 0.0001799730873483934, + "loss": 0.6925, + "step": 13191 + }, + { + "epoch": 0.9154118381791687, + "grad_norm": 2.1613047122955322, + "learning_rate": 0.00017996971403766418, + "loss": 0.8576, + "step": 13192 + }, + { + "epoch": 0.9154812296162653, + "grad_norm": 2.108738660812378, + "learning_rate": 0.0001799663404744799, + "loss": 0.5895, + "step": 13193 + }, + { + "epoch": 0.915550621053362, + "grad_norm": 1.9654889106750488, + "learning_rate": 0.00017996296665885112, + "loss": 1.0804, + "step": 13194 + }, + { + "epoch": 0.9156200124904587, + "grad_norm": 2.192396879196167, + "learning_rate": 0.00017995959259078855, + "loss": 0.7368, + "step": 13195 + }, + { + "epoch": 0.9156894039275554, + "grad_norm": 2.134876251220703, + "learning_rate": 0.0001799562182703028, + "loss": 0.8281, + "step": 13196 + }, + { + "epoch": 0.915758795364652, + "grad_norm": 1.721197485923767, + "learning_rate": 0.00017995284369740457, + "loss": 0.6353, + "step": 13197 + }, + { + "epoch": 0.9158281868017487, + "grad_norm": 2.101104497909546, + "learning_rate": 0.0001799494688721045, + "loss": 0.6036, + "step": 13198 + }, + { + "epoch": 0.9158975782388453, + "grad_norm": 1.9941034317016602, + "learning_rate": 0.0001799460937944132, + "loss": 0.6296, + "step": 13199 + }, + { + "epoch": 0.9159669696759419, + "grad_norm": 2.529223918914795, + "learning_rate": 0.00017994271846434137, + "loss": 0.7831, + "step": 13200 + }, + { + "epoch": 0.9160363611130387, + "grad_norm": 2.849902629852295, + "learning_rate": 0.00017993934288189966, + "loss": 0.8783, + "step": 13201 + }, + { + "epoch": 0.9161057525501353, + "grad_norm": 1.5754480361938477, + "learning_rate": 0.00017993596704709872, + "loss": 0.7054, + "step": 13202 + }, + { + "epoch": 0.916175143987232, + "grad_norm": 1.7262307405471802, + "learning_rate": 0.0001799325909599492, + "loss": 0.6401, + "step": 13203 + }, + { + "epoch": 0.9162445354243286, + "grad_norm": 2.0782439708709717, + "learning_rate": 0.00017992921462046176, + "loss": 0.8617, + "step": 13204 + }, + { + "epoch": 0.9163139268614253, + "grad_norm": 1.7353607416152954, + "learning_rate": 0.0001799258380286471, + "loss": 0.6772, + "step": 13205 + }, + { + "epoch": 0.9163833182985219, + "grad_norm": 1.902708888053894, + "learning_rate": 0.0001799224611845158, + "loss": 0.8053, + "step": 13206 + }, + { + "epoch": 0.9164527097356187, + "grad_norm": 1.5904278755187988, + "learning_rate": 0.00017991908408807855, + "loss": 0.622, + "step": 13207 + }, + { + "epoch": 0.9165221011727153, + "grad_norm": 1.5544575452804565, + "learning_rate": 0.00017991570673934606, + "loss": 0.4505, + "step": 13208 + }, + { + "epoch": 0.916591492609812, + "grad_norm": 1.461842656135559, + "learning_rate": 0.00017991232913832893, + "loss": 0.5507, + "step": 13209 + }, + { + "epoch": 0.9166608840469086, + "grad_norm": 2.6514835357666016, + "learning_rate": 0.00017990895128503786, + "loss": 0.9339, + "step": 13210 + }, + { + "epoch": 0.9167302754840053, + "grad_norm": 2.4317116737365723, + "learning_rate": 0.0001799055731794835, + "loss": 0.8314, + "step": 13211 + }, + { + "epoch": 0.9167996669211019, + "grad_norm": 2.1527762413024902, + "learning_rate": 0.00017990219482167655, + "loss": 0.6401, + "step": 13212 + }, + { + "epoch": 0.9168690583581987, + "grad_norm": 2.1962223052978516, + "learning_rate": 0.00017989881621162761, + "loss": 0.7989, + "step": 13213 + }, + { + "epoch": 0.9169384497952953, + "grad_norm": 2.5393733978271484, + "learning_rate": 0.00017989543734934737, + "loss": 0.735, + "step": 13214 + }, + { + "epoch": 0.9170078412323919, + "grad_norm": 1.6442012786865234, + "learning_rate": 0.00017989205823484652, + "loss": 0.474, + "step": 13215 + }, + { + "epoch": 0.9170772326694886, + "grad_norm": 1.8029248714447021, + "learning_rate": 0.0001798886788681357, + "loss": 0.7613, + "step": 13216 + }, + { + "epoch": 0.9171466241065852, + "grad_norm": 1.7011662721633911, + "learning_rate": 0.0001798852992492256, + "loss": 0.5523, + "step": 13217 + }, + { + "epoch": 0.9172160155436819, + "grad_norm": 2.4089620113372803, + "learning_rate": 0.00017988191937812686, + "loss": 0.9715, + "step": 13218 + }, + { + "epoch": 0.9172854069807785, + "grad_norm": 1.4966317415237427, + "learning_rate": 0.00017987853925485018, + "loss": 0.7783, + "step": 13219 + }, + { + "epoch": 0.9173547984178753, + "grad_norm": 2.4444406032562256, + "learning_rate": 0.00017987515887940624, + "loss": 1.0205, + "step": 13220 + }, + { + "epoch": 0.9174241898549719, + "grad_norm": 2.7209935188293457, + "learning_rate": 0.00017987177825180565, + "loss": 0.9831, + "step": 13221 + }, + { + "epoch": 0.9174935812920686, + "grad_norm": 2.130282402038574, + "learning_rate": 0.00017986839737205913, + "loss": 0.6225, + "step": 13222 + }, + { + "epoch": 0.9175629727291652, + "grad_norm": 2.6631813049316406, + "learning_rate": 0.0001798650162401774, + "loss": 0.9865, + "step": 13223 + }, + { + "epoch": 0.9176323641662619, + "grad_norm": 1.9069193601608276, + "learning_rate": 0.00017986163485617102, + "loss": 0.6558, + "step": 13224 + }, + { + "epoch": 0.9177017556033585, + "grad_norm": 2.0017166137695312, + "learning_rate": 0.00017985825322005075, + "loss": 0.8194, + "step": 13225 + }, + { + "epoch": 0.9177711470404553, + "grad_norm": 3.270219087600708, + "learning_rate": 0.00017985487133182724, + "loss": 0.985, + "step": 13226 + }, + { + "epoch": 0.9178405384775519, + "grad_norm": 2.0363357067108154, + "learning_rate": 0.00017985148919151116, + "loss": 0.7485, + "step": 13227 + }, + { + "epoch": 0.9179099299146485, + "grad_norm": 1.5425403118133545, + "learning_rate": 0.00017984810679911318, + "loss": 0.6453, + "step": 13228 + }, + { + "epoch": 0.9179793213517452, + "grad_norm": 2.004871129989624, + "learning_rate": 0.000179844724154644, + "loss": 0.7587, + "step": 13229 + }, + { + "epoch": 0.9180487127888418, + "grad_norm": 3.3776438236236572, + "learning_rate": 0.0001798413412581143, + "loss": 0.8197, + "step": 13230 + }, + { + "epoch": 0.9181181042259385, + "grad_norm": 1.9565863609313965, + "learning_rate": 0.00017983795810953474, + "loss": 0.6588, + "step": 13231 + }, + { + "epoch": 0.9181874956630351, + "grad_norm": 2.1253573894500732, + "learning_rate": 0.000179834574708916, + "loss": 0.6266, + "step": 13232 + }, + { + "epoch": 0.9182568871001319, + "grad_norm": 2.443418264389038, + "learning_rate": 0.0001798311910562688, + "loss": 0.9876, + "step": 13233 + }, + { + "epoch": 0.9183262785372285, + "grad_norm": 2.858463764190674, + "learning_rate": 0.00017982780715160377, + "loss": 0.8428, + "step": 13234 + }, + { + "epoch": 0.9183956699743252, + "grad_norm": 2.0712907314300537, + "learning_rate": 0.0001798244229949316, + "loss": 0.7426, + "step": 13235 + }, + { + "epoch": 0.9184650614114218, + "grad_norm": 1.5649642944335938, + "learning_rate": 0.00017982103858626302, + "loss": 0.7113, + "step": 13236 + }, + { + "epoch": 0.9185344528485185, + "grad_norm": 2.5241427421569824, + "learning_rate": 0.00017981765392560868, + "loss": 0.8987, + "step": 13237 + }, + { + "epoch": 0.9186038442856151, + "grad_norm": 2.3424954414367676, + "learning_rate": 0.00017981426901297925, + "loss": 1.0511, + "step": 13238 + }, + { + "epoch": 0.9186732357227119, + "grad_norm": 2.1850554943084717, + "learning_rate": 0.00017981088384838546, + "loss": 0.7574, + "step": 13239 + }, + { + "epoch": 0.9187426271598085, + "grad_norm": 1.9213007688522339, + "learning_rate": 0.00017980749843183794, + "loss": 0.7178, + "step": 13240 + }, + { + "epoch": 0.9188120185969051, + "grad_norm": 2.550929069519043, + "learning_rate": 0.00017980411276334742, + "loss": 0.7594, + "step": 13241 + }, + { + "epoch": 0.9188814100340018, + "grad_norm": 2.20328950881958, + "learning_rate": 0.00017980072684292456, + "loss": 0.7663, + "step": 13242 + }, + { + "epoch": 0.9189508014710984, + "grad_norm": 1.6828935146331787, + "learning_rate": 0.00017979734067058008, + "loss": 0.5892, + "step": 13243 + }, + { + "epoch": 0.9190201929081951, + "grad_norm": 2.5638864040374756, + "learning_rate": 0.00017979395424632465, + "loss": 0.9143, + "step": 13244 + }, + { + "epoch": 0.9190895843452918, + "grad_norm": 2.099440813064575, + "learning_rate": 0.00017979056757016893, + "loss": 0.7936, + "step": 13245 + }, + { + "epoch": 0.9191589757823885, + "grad_norm": 2.30830717086792, + "learning_rate": 0.00017978718064212367, + "loss": 0.7193, + "step": 13246 + }, + { + "epoch": 0.9192283672194851, + "grad_norm": 1.6727819442749023, + "learning_rate": 0.00017978379346219957, + "loss": 0.6056, + "step": 13247 + }, + { + "epoch": 0.9192977586565818, + "grad_norm": 3.1343162059783936, + "learning_rate": 0.00017978040603040726, + "loss": 0.7876, + "step": 13248 + }, + { + "epoch": 0.9193671500936784, + "grad_norm": 1.941311001777649, + "learning_rate": 0.00017977701834675746, + "loss": 0.7587, + "step": 13249 + }, + { + "epoch": 0.9194365415307751, + "grad_norm": 2.2643625736236572, + "learning_rate": 0.0001797736304112609, + "loss": 0.8041, + "step": 13250 + }, + { + "epoch": 0.9195059329678718, + "grad_norm": 1.6352993249893188, + "learning_rate": 0.00017977024222392822, + "loss": 0.7477, + "step": 13251 + }, + { + "epoch": 0.9195753244049685, + "grad_norm": 2.4733245372772217, + "learning_rate": 0.00017976685378477014, + "loss": 0.9784, + "step": 13252 + }, + { + "epoch": 0.9196447158420651, + "grad_norm": 3.0105743408203125, + "learning_rate": 0.00017976346509379737, + "loss": 0.7711, + "step": 13253 + }, + { + "epoch": 0.9197141072791617, + "grad_norm": 2.1860079765319824, + "learning_rate": 0.00017976007615102058, + "loss": 0.9348, + "step": 13254 + }, + { + "epoch": 0.9197834987162584, + "grad_norm": 1.9227523803710938, + "learning_rate": 0.00017975668695645048, + "loss": 0.7612, + "step": 13255 + }, + { + "epoch": 0.919852890153355, + "grad_norm": 1.3775596618652344, + "learning_rate": 0.00017975329751009777, + "loss": 0.6083, + "step": 13256 + }, + { + "epoch": 0.9199222815904518, + "grad_norm": 1.73227858543396, + "learning_rate": 0.00017974990781197318, + "loss": 0.6451, + "step": 13257 + }, + { + "epoch": 0.9199916730275484, + "grad_norm": 1.6249535083770752, + "learning_rate": 0.00017974651786208736, + "loss": 0.5768, + "step": 13258 + }, + { + "epoch": 0.9200610644646451, + "grad_norm": 2.163478374481201, + "learning_rate": 0.00017974312766045106, + "loss": 0.6413, + "step": 13259 + }, + { + "epoch": 0.9201304559017417, + "grad_norm": 2.1990036964416504, + "learning_rate": 0.00017973973720707492, + "loss": 0.8485, + "step": 13260 + }, + { + "epoch": 0.9201998473388384, + "grad_norm": 2.9493329524993896, + "learning_rate": 0.00017973634650196972, + "loss": 1.171, + "step": 13261 + }, + { + "epoch": 0.920269238775935, + "grad_norm": 2.106632709503174, + "learning_rate": 0.00017973295554514612, + "loss": 0.9265, + "step": 13262 + }, + { + "epoch": 0.9203386302130317, + "grad_norm": 2.0424981117248535, + "learning_rate": 0.00017972956433661482, + "loss": 0.8393, + "step": 13263 + }, + { + "epoch": 0.9204080216501284, + "grad_norm": 1.891987919807434, + "learning_rate": 0.0001797261728763865, + "loss": 0.582, + "step": 13264 + }, + { + "epoch": 0.9204774130872251, + "grad_norm": 2.4134063720703125, + "learning_rate": 0.00017972278116447195, + "loss": 0.8098, + "step": 13265 + }, + { + "epoch": 0.9205468045243217, + "grad_norm": 2.5255632400512695, + "learning_rate": 0.00017971938920088181, + "loss": 0.6818, + "step": 13266 + }, + { + "epoch": 0.9206161959614184, + "grad_norm": 1.9028291702270508, + "learning_rate": 0.00017971599698562684, + "loss": 0.9255, + "step": 13267 + }, + { + "epoch": 0.920685587398515, + "grad_norm": 1.8086049556732178, + "learning_rate": 0.00017971260451871767, + "loss": 0.7314, + "step": 13268 + }, + { + "epoch": 0.9207549788356116, + "grad_norm": 1.368874430656433, + "learning_rate": 0.0001797092118001651, + "loss": 0.6627, + "step": 13269 + }, + { + "epoch": 0.9208243702727084, + "grad_norm": 1.996521234512329, + "learning_rate": 0.00017970581882997973, + "loss": 0.7941, + "step": 13270 + }, + { + "epoch": 0.920893761709805, + "grad_norm": 3.815898895263672, + "learning_rate": 0.0001797024256081724, + "loss": 0.8849, + "step": 13271 + }, + { + "epoch": 0.9209631531469017, + "grad_norm": 2.516110897064209, + "learning_rate": 0.00017969903213475375, + "loss": 1.3163, + "step": 13272 + }, + { + "epoch": 0.9210325445839983, + "grad_norm": 2.1547603607177734, + "learning_rate": 0.00017969563840973448, + "loss": 0.8794, + "step": 13273 + }, + { + "epoch": 0.921101936021095, + "grad_norm": 2.979666233062744, + "learning_rate": 0.00017969224443312537, + "loss": 0.8105, + "step": 13274 + }, + { + "epoch": 0.9211713274581916, + "grad_norm": 2.194289207458496, + "learning_rate": 0.00017968885020493703, + "loss": 0.6407, + "step": 13275 + }, + { + "epoch": 0.9212407188952884, + "grad_norm": 2.199871301651001, + "learning_rate": 0.00017968545572518026, + "loss": 0.8395, + "step": 13276 + }, + { + "epoch": 0.921310110332385, + "grad_norm": 1.967193841934204, + "learning_rate": 0.00017968206099386577, + "loss": 0.6969, + "step": 13277 + }, + { + "epoch": 0.9213795017694817, + "grad_norm": 2.2031848430633545, + "learning_rate": 0.00017967866601100422, + "loss": 0.9045, + "step": 13278 + }, + { + "epoch": 0.9214488932065783, + "grad_norm": 2.268120288848877, + "learning_rate": 0.00017967527077660642, + "loss": 0.8539, + "step": 13279 + }, + { + "epoch": 0.921518284643675, + "grad_norm": 1.8312656879425049, + "learning_rate": 0.00017967187529068296, + "loss": 0.7286, + "step": 13280 + }, + { + "epoch": 0.9215876760807716, + "grad_norm": 1.9212560653686523, + "learning_rate": 0.0001796684795532447, + "loss": 0.8282, + "step": 13281 + }, + { + "epoch": 0.9216570675178682, + "grad_norm": 2.918647527694702, + "learning_rate": 0.00017966508356430226, + "loss": 1.1467, + "step": 13282 + }, + { + "epoch": 0.921726458954965, + "grad_norm": 2.552122116088867, + "learning_rate": 0.0001796616873238664, + "loss": 0.6033, + "step": 13283 + }, + { + "epoch": 0.9217958503920616, + "grad_norm": 2.1556572914123535, + "learning_rate": 0.00017965829083194785, + "loss": 0.6079, + "step": 13284 + }, + { + "epoch": 0.9218652418291583, + "grad_norm": 2.232501983642578, + "learning_rate": 0.00017965489408855733, + "loss": 0.666, + "step": 13285 + }, + { + "epoch": 0.9219346332662549, + "grad_norm": 1.8399156332015991, + "learning_rate": 0.0001796514970937055, + "loss": 0.7068, + "step": 13286 + }, + { + "epoch": 0.9220040247033516, + "grad_norm": 2.4016940593719482, + "learning_rate": 0.00017964809984740318, + "loss": 0.9288, + "step": 13287 + }, + { + "epoch": 0.9220734161404482, + "grad_norm": 2.657430648803711, + "learning_rate": 0.00017964470234966103, + "loss": 0.7975, + "step": 13288 + }, + { + "epoch": 0.922142807577545, + "grad_norm": 2.1012167930603027, + "learning_rate": 0.0001796413046004898, + "loss": 0.7843, + "step": 13289 + }, + { + "epoch": 0.9222121990146416, + "grad_norm": 1.9309300184249878, + "learning_rate": 0.00017963790659990017, + "loss": 0.867, + "step": 13290 + }, + { + "epoch": 0.9222815904517383, + "grad_norm": 2.1610727310180664, + "learning_rate": 0.000179634508347903, + "loss": 0.8234, + "step": 13291 + }, + { + "epoch": 0.9223509818888349, + "grad_norm": 1.429079294204712, + "learning_rate": 0.00017963110984450884, + "loss": 0.5121, + "step": 13292 + }, + { + "epoch": 0.9224203733259316, + "grad_norm": 2.0023717880249023, + "learning_rate": 0.00017962771108972857, + "loss": 0.775, + "step": 13293 + }, + { + "epoch": 0.9224897647630282, + "grad_norm": 1.9192299842834473, + "learning_rate": 0.0001796243120835728, + "loss": 0.9191, + "step": 13294 + }, + { + "epoch": 0.9225591562001249, + "grad_norm": 1.8613280057907104, + "learning_rate": 0.00017962091282605234, + "loss": 0.7133, + "step": 13295 + }, + { + "epoch": 0.9226285476372216, + "grad_norm": 1.7983736991882324, + "learning_rate": 0.00017961751331717788, + "loss": 0.6339, + "step": 13296 + }, + { + "epoch": 0.9226979390743182, + "grad_norm": 2.6770763397216797, + "learning_rate": 0.0001796141135569602, + "loss": 1.1844, + "step": 13297 + }, + { + "epoch": 0.9227673305114149, + "grad_norm": 2.217651844024658, + "learning_rate": 0.00017961071354540998, + "loss": 0.9709, + "step": 13298 + }, + { + "epoch": 0.9228367219485115, + "grad_norm": 1.886523962020874, + "learning_rate": 0.000179607313282538, + "loss": 0.6957, + "step": 13299 + }, + { + "epoch": 0.9229061133856082, + "grad_norm": 1.8919929265975952, + "learning_rate": 0.00017960391276835493, + "loss": 0.6174, + "step": 13300 + }, + { + "epoch": 0.9229755048227049, + "grad_norm": 1.926537275314331, + "learning_rate": 0.00017960051200287155, + "loss": 0.9732, + "step": 13301 + }, + { + "epoch": 0.9230448962598016, + "grad_norm": 2.197279930114746, + "learning_rate": 0.0001795971109860986, + "loss": 0.8193, + "step": 13302 + }, + { + "epoch": 0.9231142876968982, + "grad_norm": 2.0248539447784424, + "learning_rate": 0.0001795937097180468, + "loss": 0.9161, + "step": 13303 + }, + { + "epoch": 0.9231836791339949, + "grad_norm": 1.8868408203125, + "learning_rate": 0.0001795903081987269, + "loss": 0.6482, + "step": 13304 + }, + { + "epoch": 0.9232530705710915, + "grad_norm": 2.285958766937256, + "learning_rate": 0.00017958690642814963, + "loss": 0.9051, + "step": 13305 + }, + { + "epoch": 0.9233224620081882, + "grad_norm": 2.0955798625946045, + "learning_rate": 0.00017958350440632574, + "loss": 0.7935, + "step": 13306 + }, + { + "epoch": 0.9233918534452848, + "grad_norm": 2.1473751068115234, + "learning_rate": 0.00017958010213326591, + "loss": 0.8081, + "step": 13307 + }, + { + "epoch": 0.9234612448823815, + "grad_norm": 2.5535073280334473, + "learning_rate": 0.000179576699608981, + "loss": 0.9611, + "step": 13308 + }, + { + "epoch": 0.9235306363194782, + "grad_norm": 1.515424370765686, + "learning_rate": 0.00017957329683348166, + "loss": 0.5862, + "step": 13309 + }, + { + "epoch": 0.9236000277565748, + "grad_norm": 2.0599112510681152, + "learning_rate": 0.00017956989380677865, + "loss": 0.7369, + "step": 13310 + }, + { + "epoch": 0.9236694191936715, + "grad_norm": 2.898500442504883, + "learning_rate": 0.0001795664905288827, + "loss": 0.9809, + "step": 13311 + }, + { + "epoch": 0.9237388106307681, + "grad_norm": 2.000084638595581, + "learning_rate": 0.00017956308699980457, + "loss": 0.6476, + "step": 13312 + }, + { + "epoch": 0.9238082020678648, + "grad_norm": 2.8620359897613525, + "learning_rate": 0.00017955968321955502, + "loss": 0.9944, + "step": 13313 + }, + { + "epoch": 0.9238775935049615, + "grad_norm": 2.6780471801757812, + "learning_rate": 0.00017955627918814478, + "loss": 0.8298, + "step": 13314 + }, + { + "epoch": 0.9239469849420582, + "grad_norm": 1.8240561485290527, + "learning_rate": 0.0001795528749055846, + "loss": 0.7437, + "step": 13315 + }, + { + "epoch": 0.9240163763791548, + "grad_norm": 2.109323740005493, + "learning_rate": 0.0001795494703718852, + "loss": 0.7798, + "step": 13316 + }, + { + "epoch": 0.9240857678162515, + "grad_norm": 1.9246106147766113, + "learning_rate": 0.00017954606558705737, + "loss": 0.7581, + "step": 13317 + }, + { + "epoch": 0.9241551592533481, + "grad_norm": 1.459477186203003, + "learning_rate": 0.00017954266055111187, + "loss": 0.7255, + "step": 13318 + }, + { + "epoch": 0.9242245506904448, + "grad_norm": 2.2386486530303955, + "learning_rate": 0.00017953925526405936, + "loss": 0.972, + "step": 13319 + }, + { + "epoch": 0.9242939421275415, + "grad_norm": 1.832558512687683, + "learning_rate": 0.00017953584972591068, + "loss": 0.8582, + "step": 13320 + }, + { + "epoch": 0.9243633335646382, + "grad_norm": 1.9371618032455444, + "learning_rate": 0.00017953244393667655, + "loss": 0.9077, + "step": 13321 + }, + { + "epoch": 0.9244327250017348, + "grad_norm": 2.7476553916931152, + "learning_rate": 0.00017952903789636773, + "loss": 0.8081, + "step": 13322 + }, + { + "epoch": 0.9245021164388314, + "grad_norm": 1.6629397869110107, + "learning_rate": 0.00017952563160499496, + "loss": 0.6465, + "step": 13323 + }, + { + "epoch": 0.9245715078759281, + "grad_norm": 2.1422719955444336, + "learning_rate": 0.000179522225062569, + "loss": 0.8698, + "step": 13324 + }, + { + "epoch": 0.9246408993130247, + "grad_norm": 2.545509099960327, + "learning_rate": 0.0001795188182691006, + "loss": 0.9153, + "step": 13325 + }, + { + "epoch": 0.9247102907501215, + "grad_norm": 1.8788959980010986, + "learning_rate": 0.0001795154112246005, + "loss": 0.7937, + "step": 13326 + }, + { + "epoch": 0.9247796821872181, + "grad_norm": 2.5145599842071533, + "learning_rate": 0.00017951200392907947, + "loss": 0.8541, + "step": 13327 + }, + { + "epoch": 0.9248490736243148, + "grad_norm": 2.099443197250366, + "learning_rate": 0.0001795085963825483, + "loss": 0.696, + "step": 13328 + }, + { + "epoch": 0.9249184650614114, + "grad_norm": 2.6715846061706543, + "learning_rate": 0.0001795051885850177, + "loss": 0.7664, + "step": 13329 + }, + { + "epoch": 0.9249878564985081, + "grad_norm": 1.689658522605896, + "learning_rate": 0.00017950178053649843, + "loss": 0.6673, + "step": 13330 + }, + { + "epoch": 0.9250572479356047, + "grad_norm": 2.158780336380005, + "learning_rate": 0.00017949837223700128, + "loss": 0.8137, + "step": 13331 + }, + { + "epoch": 0.9251266393727015, + "grad_norm": 2.8774445056915283, + "learning_rate": 0.00017949496368653697, + "loss": 0.8801, + "step": 13332 + }, + { + "epoch": 0.9251960308097981, + "grad_norm": 3.325136661529541, + "learning_rate": 0.00017949155488511634, + "loss": 1.3648, + "step": 13333 + }, + { + "epoch": 0.9252654222468948, + "grad_norm": 2.123995780944824, + "learning_rate": 0.00017948814583275, + "loss": 1.0482, + "step": 13334 + }, + { + "epoch": 0.9253348136839914, + "grad_norm": 1.6423420906066895, + "learning_rate": 0.00017948473652944889, + "loss": 0.6634, + "step": 13335 + }, + { + "epoch": 0.925404205121088, + "grad_norm": 2.762800931930542, + "learning_rate": 0.00017948132697522367, + "loss": 0.8519, + "step": 13336 + }, + { + "epoch": 0.9254735965581847, + "grad_norm": 1.552945613861084, + "learning_rate": 0.0001794779171700851, + "loss": 0.5566, + "step": 13337 + }, + { + "epoch": 0.9255429879952813, + "grad_norm": 2.887824535369873, + "learning_rate": 0.00017947450711404398, + "loss": 0.9081, + "step": 13338 + }, + { + "epoch": 0.9256123794323781, + "grad_norm": 1.6773791313171387, + "learning_rate": 0.00017947109680711108, + "loss": 0.5649, + "step": 13339 + }, + { + "epoch": 0.9256817708694747, + "grad_norm": 1.8965660333633423, + "learning_rate": 0.0001794676862492971, + "loss": 0.9452, + "step": 13340 + }, + { + "epoch": 0.9257511623065714, + "grad_norm": 1.69387948513031, + "learning_rate": 0.0001794642754406129, + "loss": 0.6216, + "step": 13341 + }, + { + "epoch": 0.925820553743668, + "grad_norm": 2.2447357177734375, + "learning_rate": 0.00017946086438106918, + "loss": 0.5882, + "step": 13342 + }, + { + "epoch": 0.9258899451807647, + "grad_norm": 1.779778242111206, + "learning_rate": 0.00017945745307067673, + "loss": 0.7794, + "step": 13343 + }, + { + "epoch": 0.9259593366178613, + "grad_norm": 2.5255284309387207, + "learning_rate": 0.0001794540415094463, + "loss": 0.9725, + "step": 13344 + }, + { + "epoch": 0.9260287280549581, + "grad_norm": 1.9882838726043701, + "learning_rate": 0.0001794506296973887, + "loss": 0.6636, + "step": 13345 + }, + { + "epoch": 0.9260981194920547, + "grad_norm": 1.9690324068069458, + "learning_rate": 0.0001794472176345147, + "loss": 0.7013, + "step": 13346 + }, + { + "epoch": 0.9261675109291514, + "grad_norm": 2.0775036811828613, + "learning_rate": 0.00017944380532083502, + "loss": 0.6339, + "step": 13347 + }, + { + "epoch": 0.926236902366248, + "grad_norm": 1.8955086469650269, + "learning_rate": 0.0001794403927563605, + "loss": 0.6285, + "step": 13348 + }, + { + "epoch": 0.9263062938033446, + "grad_norm": 2.789695978164673, + "learning_rate": 0.00017943697994110184, + "loss": 1.0716, + "step": 13349 + }, + { + "epoch": 0.9263756852404413, + "grad_norm": 2.162301778793335, + "learning_rate": 0.00017943356687506986, + "loss": 0.8651, + "step": 13350 + }, + { + "epoch": 0.926445076677538, + "grad_norm": 2.4663665294647217, + "learning_rate": 0.00017943015355827534, + "loss": 0.7932, + "step": 13351 + }, + { + "epoch": 0.9265144681146347, + "grad_norm": 1.9847561120986938, + "learning_rate": 0.00017942673999072907, + "loss": 0.8826, + "step": 13352 + }, + { + "epoch": 0.9265838595517313, + "grad_norm": 1.9432162046432495, + "learning_rate": 0.0001794233261724417, + "loss": 0.649, + "step": 13353 + }, + { + "epoch": 0.926653250988828, + "grad_norm": 2.5105159282684326, + "learning_rate": 0.0001794199121034242, + "loss": 0.743, + "step": 13354 + }, + { + "epoch": 0.9267226424259246, + "grad_norm": 1.7977858781814575, + "learning_rate": 0.0001794164977836872, + "loss": 0.6982, + "step": 13355 + }, + { + "epoch": 0.9267920338630213, + "grad_norm": 2.30539870262146, + "learning_rate": 0.00017941308321324154, + "loss": 0.7381, + "step": 13356 + }, + { + "epoch": 0.926861425300118, + "grad_norm": 2.3165524005889893, + "learning_rate": 0.00017940966839209802, + "loss": 0.8183, + "step": 13357 + }, + { + "epoch": 0.9269308167372147, + "grad_norm": 1.8475959300994873, + "learning_rate": 0.00017940625332026736, + "loss": 0.8441, + "step": 13358 + }, + { + "epoch": 0.9270002081743113, + "grad_norm": 2.184023141860962, + "learning_rate": 0.00017940283799776038, + "loss": 0.8051, + "step": 13359 + }, + { + "epoch": 0.927069599611408, + "grad_norm": 2.3878684043884277, + "learning_rate": 0.00017939942242458786, + "loss": 0.8255, + "step": 13360 + }, + { + "epoch": 0.9271389910485046, + "grad_norm": 2.4974849224090576, + "learning_rate": 0.00017939600660076053, + "loss": 0.8025, + "step": 13361 + }, + { + "epoch": 0.9272083824856012, + "grad_norm": 2.0892937183380127, + "learning_rate": 0.00017939259052628928, + "loss": 0.8656, + "step": 13362 + }, + { + "epoch": 0.927277773922698, + "grad_norm": 1.9418810606002808, + "learning_rate": 0.0001793891742011848, + "loss": 0.9365, + "step": 13363 + }, + { + "epoch": 0.9273471653597946, + "grad_norm": 1.6437456607818604, + "learning_rate": 0.0001793857576254579, + "loss": 0.7099, + "step": 13364 + }, + { + "epoch": 0.9274165567968913, + "grad_norm": 2.3709282875061035, + "learning_rate": 0.0001793823407991194, + "loss": 0.7423, + "step": 13365 + }, + { + "epoch": 0.9274859482339879, + "grad_norm": 1.920894980430603, + "learning_rate": 0.00017937892372218005, + "loss": 0.6033, + "step": 13366 + }, + { + "epoch": 0.9275553396710846, + "grad_norm": 2.7236292362213135, + "learning_rate": 0.00017937550639465064, + "loss": 1.0048, + "step": 13367 + }, + { + "epoch": 0.9276247311081812, + "grad_norm": 1.6625571250915527, + "learning_rate": 0.00017937208881654197, + "loss": 0.807, + "step": 13368 + }, + { + "epoch": 0.9276941225452779, + "grad_norm": 2.3286397457122803, + "learning_rate": 0.00017936867098786483, + "loss": 0.666, + "step": 13369 + }, + { + "epoch": 0.9277635139823746, + "grad_norm": 1.914905071258545, + "learning_rate": 0.00017936525290863, + "loss": 0.5066, + "step": 13370 + }, + { + "epoch": 0.9278329054194713, + "grad_norm": 2.289620876312256, + "learning_rate": 0.00017936183457884828, + "loss": 0.872, + "step": 13371 + }, + { + "epoch": 0.9279022968565679, + "grad_norm": 1.9453855752944946, + "learning_rate": 0.00017935841599853042, + "loss": 0.7788, + "step": 13372 + }, + { + "epoch": 0.9279716882936646, + "grad_norm": 2.180833101272583, + "learning_rate": 0.0001793549971676873, + "loss": 0.8345, + "step": 13373 + }, + { + "epoch": 0.9280410797307612, + "grad_norm": 3.1326510906219482, + "learning_rate": 0.00017935157808632966, + "loss": 0.8499, + "step": 13374 + }, + { + "epoch": 0.9281104711678578, + "grad_norm": 1.773873209953308, + "learning_rate": 0.00017934815875446828, + "loss": 0.7321, + "step": 13375 + }, + { + "epoch": 0.9281798626049546, + "grad_norm": 2.6755950450897217, + "learning_rate": 0.00017934473917211393, + "loss": 1.0961, + "step": 13376 + }, + { + "epoch": 0.9282492540420512, + "grad_norm": 2.107109546661377, + "learning_rate": 0.0001793413193392775, + "loss": 0.7891, + "step": 13377 + }, + { + "epoch": 0.9283186454791479, + "grad_norm": 2.4298713207244873, + "learning_rate": 0.0001793378992559697, + "loss": 0.8552, + "step": 13378 + }, + { + "epoch": 0.9283880369162445, + "grad_norm": 1.6807849407196045, + "learning_rate": 0.00017933447892220134, + "loss": 0.4177, + "step": 13379 + }, + { + "epoch": 0.9284574283533412, + "grad_norm": 2.817824363708496, + "learning_rate": 0.00017933105833798328, + "loss": 0.792, + "step": 13380 + }, + { + "epoch": 0.9285268197904378, + "grad_norm": 2.2507474422454834, + "learning_rate": 0.00017932763750332624, + "loss": 0.882, + "step": 13381 + }, + { + "epoch": 0.9285962112275346, + "grad_norm": 2.076345205307007, + "learning_rate": 0.00017932421641824107, + "loss": 0.6768, + "step": 13382 + }, + { + "epoch": 0.9286656026646312, + "grad_norm": 1.7135059833526611, + "learning_rate": 0.00017932079508273856, + "loss": 0.599, + "step": 13383 + }, + { + "epoch": 0.9287349941017279, + "grad_norm": 2.727609395980835, + "learning_rate": 0.00017931737349682946, + "loss": 0.8948, + "step": 13384 + }, + { + "epoch": 0.9288043855388245, + "grad_norm": 3.0209901332855225, + "learning_rate": 0.00017931395166052467, + "loss": 1.4102, + "step": 13385 + }, + { + "epoch": 0.9288737769759212, + "grad_norm": 1.58987295627594, + "learning_rate": 0.00017931052957383491, + "loss": 0.6372, + "step": 13386 + }, + { + "epoch": 0.9289431684130178, + "grad_norm": 2.123441457748413, + "learning_rate": 0.000179307107236771, + "loss": 0.7086, + "step": 13387 + }, + { + "epoch": 0.9290125598501145, + "grad_norm": 1.642380952835083, + "learning_rate": 0.00017930368464934374, + "loss": 0.7063, + "step": 13388 + }, + { + "epoch": 0.9290819512872112, + "grad_norm": 1.7907259464263916, + "learning_rate": 0.00017930026181156399, + "loss": 0.7674, + "step": 13389 + }, + { + "epoch": 0.9291513427243078, + "grad_norm": 2.683117389678955, + "learning_rate": 0.0001792968387234425, + "loss": 0.8029, + "step": 13390 + }, + { + "epoch": 0.9292207341614045, + "grad_norm": 1.9057590961456299, + "learning_rate": 0.00017929341538499007, + "loss": 0.8892, + "step": 13391 + }, + { + "epoch": 0.9292901255985011, + "grad_norm": 1.5641779899597168, + "learning_rate": 0.00017928999179621754, + "loss": 0.6789, + "step": 13392 + }, + { + "epoch": 0.9293595170355978, + "grad_norm": 1.750501036643982, + "learning_rate": 0.00017928656795713574, + "loss": 0.7822, + "step": 13393 + }, + { + "epoch": 0.9294289084726944, + "grad_norm": 1.8291727304458618, + "learning_rate": 0.0001792831438677554, + "loss": 0.8835, + "step": 13394 + }, + { + "epoch": 0.9294982999097912, + "grad_norm": 2.0405309200286865, + "learning_rate": 0.0001792797195280874, + "loss": 0.7571, + "step": 13395 + }, + { + "epoch": 0.9295676913468878, + "grad_norm": 1.8661781549453735, + "learning_rate": 0.00017927629493814248, + "loss": 0.7796, + "step": 13396 + }, + { + "epoch": 0.9296370827839845, + "grad_norm": 2.1375980377197266, + "learning_rate": 0.00017927287009793154, + "loss": 1.0118, + "step": 13397 + }, + { + "epoch": 0.9297064742210811, + "grad_norm": 2.0673890113830566, + "learning_rate": 0.00017926944500746533, + "loss": 0.8559, + "step": 13398 + }, + { + "epoch": 0.9297758656581778, + "grad_norm": 2.0447590351104736, + "learning_rate": 0.0001792660196667547, + "loss": 0.8229, + "step": 13399 + }, + { + "epoch": 0.9298452570952744, + "grad_norm": 2.0198116302490234, + "learning_rate": 0.0001792625940758104, + "loss": 0.7277, + "step": 13400 + }, + { + "epoch": 0.9299146485323712, + "grad_norm": 1.4385658502578735, + "learning_rate": 0.0001792591682346433, + "loss": 0.6966, + "step": 13401 + }, + { + "epoch": 0.9299840399694678, + "grad_norm": 1.6646220684051514, + "learning_rate": 0.0001792557421432642, + "loss": 0.7306, + "step": 13402 + }, + { + "epoch": 0.9300534314065644, + "grad_norm": 2.331139326095581, + "learning_rate": 0.00017925231580168393, + "loss": 0.8648, + "step": 13403 + }, + { + "epoch": 0.9301228228436611, + "grad_norm": 1.3233399391174316, + "learning_rate": 0.0001792488892099133, + "loss": 0.5375, + "step": 13404 + }, + { + "epoch": 0.9301922142807577, + "grad_norm": 2.009798049926758, + "learning_rate": 0.00017924546236796314, + "loss": 0.5602, + "step": 13405 + }, + { + "epoch": 0.9302616057178544, + "grad_norm": 2.343393325805664, + "learning_rate": 0.0001792420352758442, + "loss": 1.0052, + "step": 13406 + }, + { + "epoch": 0.930330997154951, + "grad_norm": 1.8048182725906372, + "learning_rate": 0.00017923860793356736, + "loss": 0.6752, + "step": 13407 + }, + { + "epoch": 0.9304003885920478, + "grad_norm": 2.137634515762329, + "learning_rate": 0.00017923518034114344, + "loss": 0.8962, + "step": 13408 + }, + { + "epoch": 0.9304697800291444, + "grad_norm": 1.8355201482772827, + "learning_rate": 0.00017923175249858327, + "loss": 0.7077, + "step": 13409 + }, + { + "epoch": 0.9305391714662411, + "grad_norm": 2.491957426071167, + "learning_rate": 0.0001792283244058976, + "loss": 0.7088, + "step": 13410 + }, + { + "epoch": 0.9306085629033377, + "grad_norm": 2.3180038928985596, + "learning_rate": 0.00017922489606309736, + "loss": 0.8762, + "step": 13411 + }, + { + "epoch": 0.9306779543404344, + "grad_norm": 2.2168242931365967, + "learning_rate": 0.00017922146747019328, + "loss": 0.9907, + "step": 13412 + }, + { + "epoch": 0.930747345777531, + "grad_norm": 1.8844600915908813, + "learning_rate": 0.00017921803862719623, + "loss": 0.5265, + "step": 13413 + }, + { + "epoch": 0.9308167372146278, + "grad_norm": 2.2842860221862793, + "learning_rate": 0.000179214609534117, + "loss": 0.6593, + "step": 13414 + }, + { + "epoch": 0.9308861286517244, + "grad_norm": 1.6074670553207397, + "learning_rate": 0.00017921118019096645, + "loss": 0.4732, + "step": 13415 + }, + { + "epoch": 0.930955520088821, + "grad_norm": 2.169955015182495, + "learning_rate": 0.0001792077505977554, + "loss": 0.8772, + "step": 13416 + }, + { + "epoch": 0.9310249115259177, + "grad_norm": 2.235084056854248, + "learning_rate": 0.00017920432075449466, + "loss": 0.9074, + "step": 13417 + }, + { + "epoch": 0.9310943029630143, + "grad_norm": 2.952493190765381, + "learning_rate": 0.0001792008906611951, + "loss": 0.7766, + "step": 13418 + }, + { + "epoch": 0.931163694400111, + "grad_norm": 1.816744327545166, + "learning_rate": 0.00017919746031786748, + "loss": 0.7215, + "step": 13419 + }, + { + "epoch": 0.9312330858372077, + "grad_norm": 1.9864698648452759, + "learning_rate": 0.0001791940297245227, + "loss": 0.7155, + "step": 13420 + }, + { + "epoch": 0.9313024772743044, + "grad_norm": 2.7543301582336426, + "learning_rate": 0.00017919059888117157, + "loss": 0.7387, + "step": 13421 + }, + { + "epoch": 0.931371868711401, + "grad_norm": 2.0457561016082764, + "learning_rate": 0.00017918716778782482, + "loss": 0.5507, + "step": 13422 + }, + { + "epoch": 0.9314412601484977, + "grad_norm": 2.3880538940429688, + "learning_rate": 0.00017918373644449344, + "loss": 0.8393, + "step": 13423 + }, + { + "epoch": 0.9315106515855943, + "grad_norm": 1.7640646696090698, + "learning_rate": 0.00017918030485118819, + "loss": 0.7474, + "step": 13424 + }, + { + "epoch": 0.931580043022691, + "grad_norm": 1.8099583387374878, + "learning_rate": 0.00017917687300791988, + "loss": 0.7248, + "step": 13425 + }, + { + "epoch": 0.9316494344597877, + "grad_norm": 2.412588119506836, + "learning_rate": 0.0001791734409146994, + "loss": 0.748, + "step": 13426 + }, + { + "epoch": 0.9317188258968844, + "grad_norm": 1.810347557067871, + "learning_rate": 0.00017917000857153753, + "loss": 0.7194, + "step": 13427 + }, + { + "epoch": 0.931788217333981, + "grad_norm": 1.8579761981964111, + "learning_rate": 0.0001791665759784451, + "loss": 0.7234, + "step": 13428 + }, + { + "epoch": 0.9318576087710776, + "grad_norm": 1.69245183467865, + "learning_rate": 0.00017916314313543304, + "loss": 0.5449, + "step": 13429 + }, + { + "epoch": 0.9319270002081743, + "grad_norm": 2.369891405105591, + "learning_rate": 0.00017915971004251206, + "loss": 0.8088, + "step": 13430 + }, + { + "epoch": 0.9319963916452709, + "grad_norm": 2.0696616172790527, + "learning_rate": 0.00017915627669969308, + "loss": 0.8206, + "step": 13431 + }, + { + "epoch": 0.9320657830823676, + "grad_norm": 1.838043212890625, + "learning_rate": 0.00017915284310698692, + "loss": 0.5878, + "step": 13432 + }, + { + "epoch": 0.9321351745194643, + "grad_norm": 1.7872740030288696, + "learning_rate": 0.00017914940926440443, + "loss": 0.6394, + "step": 13433 + }, + { + "epoch": 0.932204565956561, + "grad_norm": 2.215691089630127, + "learning_rate": 0.00017914597517195642, + "loss": 0.9813, + "step": 13434 + }, + { + "epoch": 0.9322739573936576, + "grad_norm": 2.2553398609161377, + "learning_rate": 0.0001791425408296538, + "loss": 0.9482, + "step": 13435 + }, + { + "epoch": 0.9323433488307543, + "grad_norm": 2.185983180999756, + "learning_rate": 0.0001791391062375073, + "loss": 0.7751, + "step": 13436 + }, + { + "epoch": 0.9324127402678509, + "grad_norm": 1.8964170217514038, + "learning_rate": 0.00017913567139552785, + "loss": 0.6244, + "step": 13437 + }, + { + "epoch": 0.9324821317049476, + "grad_norm": 2.4681713581085205, + "learning_rate": 0.00017913223630372627, + "loss": 0.7495, + "step": 13438 + }, + { + "epoch": 0.9325515231420443, + "grad_norm": 1.952232003211975, + "learning_rate": 0.00017912880096211336, + "loss": 0.7799, + "step": 13439 + }, + { + "epoch": 0.932620914579141, + "grad_norm": 2.241105079650879, + "learning_rate": 0.00017912536537070003, + "loss": 0.8525, + "step": 13440 + }, + { + "epoch": 0.9326903060162376, + "grad_norm": 3.1852779388427734, + "learning_rate": 0.00017912192952949712, + "loss": 0.8358, + "step": 13441 + }, + { + "epoch": 0.9327596974533343, + "grad_norm": 2.202592134475708, + "learning_rate": 0.00017911849343851545, + "loss": 0.7865, + "step": 13442 + }, + { + "epoch": 0.9328290888904309, + "grad_norm": 2.7632029056549072, + "learning_rate": 0.00017911505709776587, + "loss": 0.9963, + "step": 13443 + }, + { + "epoch": 0.9328984803275275, + "grad_norm": 1.4822428226470947, + "learning_rate": 0.00017911162050725924, + "loss": 0.6375, + "step": 13444 + }, + { + "epoch": 0.9329678717646243, + "grad_norm": 1.948420524597168, + "learning_rate": 0.0001791081836670064, + "loss": 0.8884, + "step": 13445 + }, + { + "epoch": 0.9330372632017209, + "grad_norm": 2.1185803413391113, + "learning_rate": 0.0001791047465770182, + "loss": 0.735, + "step": 13446 + }, + { + "epoch": 0.9331066546388176, + "grad_norm": 2.149372100830078, + "learning_rate": 0.0001791013092373055, + "loss": 0.8837, + "step": 13447 + }, + { + "epoch": 0.9331760460759142, + "grad_norm": 2.2351317405700684, + "learning_rate": 0.00017909787164787912, + "loss": 0.8657, + "step": 13448 + }, + { + "epoch": 0.9332454375130109, + "grad_norm": 1.562161922454834, + "learning_rate": 0.00017909443380874994, + "loss": 0.7184, + "step": 13449 + }, + { + "epoch": 0.9333148289501075, + "grad_norm": 1.7559868097305298, + "learning_rate": 0.00017909099571992884, + "loss": 0.5565, + "step": 13450 + }, + { + "epoch": 0.9333842203872043, + "grad_norm": 1.6782350540161133, + "learning_rate": 0.0001790875573814266, + "loss": 0.6248, + "step": 13451 + }, + { + "epoch": 0.9334536118243009, + "grad_norm": 2.7389025688171387, + "learning_rate": 0.00017908411879325419, + "loss": 0.7203, + "step": 13452 + }, + { + "epoch": 0.9335230032613976, + "grad_norm": 2.4289193153381348, + "learning_rate": 0.00017908067995542232, + "loss": 0.7897, + "step": 13453 + }, + { + "epoch": 0.9335923946984942, + "grad_norm": 1.668221116065979, + "learning_rate": 0.00017907724086794197, + "loss": 0.6869, + "step": 13454 + }, + { + "epoch": 0.9336617861355909, + "grad_norm": 2.564931869506836, + "learning_rate": 0.0001790738015308239, + "loss": 0.8022, + "step": 13455 + }, + { + "epoch": 0.9337311775726875, + "grad_norm": 2.110732078552246, + "learning_rate": 0.00017907036194407904, + "loss": 0.7945, + "step": 13456 + }, + { + "epoch": 0.9338005690097841, + "grad_norm": 1.7833223342895508, + "learning_rate": 0.00017906692210771822, + "loss": 0.6652, + "step": 13457 + }, + { + "epoch": 0.9338699604468809, + "grad_norm": 3.58188796043396, + "learning_rate": 0.00017906348202175227, + "loss": 0.9016, + "step": 13458 + }, + { + "epoch": 0.9339393518839775, + "grad_norm": 1.402617335319519, + "learning_rate": 0.00017906004168619213, + "loss": 0.6588, + "step": 13459 + }, + { + "epoch": 0.9340087433210742, + "grad_norm": 1.9655131101608276, + "learning_rate": 0.00017905660110104855, + "loss": 0.6665, + "step": 13460 + }, + { + "epoch": 0.9340781347581708, + "grad_norm": 2.0428500175476074, + "learning_rate": 0.00017905316026633252, + "loss": 0.7797, + "step": 13461 + }, + { + "epoch": 0.9341475261952675, + "grad_norm": 1.9474817514419556, + "learning_rate": 0.0001790497191820548, + "loss": 0.6872, + "step": 13462 + }, + { + "epoch": 0.9342169176323641, + "grad_norm": 2.1906728744506836, + "learning_rate": 0.00017904627784822629, + "loss": 0.9769, + "step": 13463 + }, + { + "epoch": 0.9342863090694609, + "grad_norm": 2.0547924041748047, + "learning_rate": 0.00017904283626485787, + "loss": 0.7072, + "step": 13464 + }, + { + "epoch": 0.9343557005065575, + "grad_norm": 2.439312219619751, + "learning_rate": 0.00017903939443196035, + "loss": 0.8333, + "step": 13465 + }, + { + "epoch": 0.9344250919436542, + "grad_norm": 2.6586248874664307, + "learning_rate": 0.00017903595234954468, + "loss": 0.6421, + "step": 13466 + }, + { + "epoch": 0.9344944833807508, + "grad_norm": 1.9256339073181152, + "learning_rate": 0.00017903251001762162, + "loss": 0.7064, + "step": 13467 + }, + { + "epoch": 0.9345638748178475, + "grad_norm": 2.6123011112213135, + "learning_rate": 0.0001790290674362021, + "loss": 0.7565, + "step": 13468 + }, + { + "epoch": 0.9346332662549441, + "grad_norm": 2.2312564849853516, + "learning_rate": 0.00017902562460529704, + "loss": 1.1026, + "step": 13469 + }, + { + "epoch": 0.9347026576920408, + "grad_norm": 1.7196071147918701, + "learning_rate": 0.0001790221815249172, + "loss": 0.7886, + "step": 13470 + }, + { + "epoch": 0.9347720491291375, + "grad_norm": 3.5664165019989014, + "learning_rate": 0.00017901873819507354, + "loss": 1.0069, + "step": 13471 + }, + { + "epoch": 0.9348414405662341, + "grad_norm": 2.606412172317505, + "learning_rate": 0.00017901529461577685, + "loss": 0.9178, + "step": 13472 + }, + { + "epoch": 0.9349108320033308, + "grad_norm": 1.615336537361145, + "learning_rate": 0.0001790118507870381, + "loss": 0.6347, + "step": 13473 + }, + { + "epoch": 0.9349802234404274, + "grad_norm": 2.0806610584259033, + "learning_rate": 0.00017900840670886806, + "loss": 0.9943, + "step": 13474 + }, + { + "epoch": 0.9350496148775241, + "grad_norm": 1.986789345741272, + "learning_rate": 0.00017900496238127764, + "loss": 0.7643, + "step": 13475 + }, + { + "epoch": 0.9351190063146207, + "grad_norm": 1.9117581844329834, + "learning_rate": 0.00017900151780427773, + "loss": 0.842, + "step": 13476 + }, + { + "epoch": 0.9351883977517175, + "grad_norm": 2.203423023223877, + "learning_rate": 0.0001789980729778792, + "loss": 0.6659, + "step": 13477 + }, + { + "epoch": 0.9352577891888141, + "grad_norm": 2.0656986236572266, + "learning_rate": 0.00017899462790209292, + "loss": 0.6895, + "step": 13478 + }, + { + "epoch": 0.9353271806259108, + "grad_norm": 2.185467004776001, + "learning_rate": 0.00017899118257692977, + "loss": 0.9785, + "step": 13479 + }, + { + "epoch": 0.9353965720630074, + "grad_norm": 2.112070083618164, + "learning_rate": 0.00017898773700240062, + "loss": 0.8599, + "step": 13480 + }, + { + "epoch": 0.9354659635001041, + "grad_norm": 2.7459001541137695, + "learning_rate": 0.00017898429117851636, + "loss": 0.8394, + "step": 13481 + }, + { + "epoch": 0.9355353549372007, + "grad_norm": 1.97752046585083, + "learning_rate": 0.00017898084510528783, + "loss": 0.4696, + "step": 13482 + }, + { + "epoch": 0.9356047463742974, + "grad_norm": 2.143988609313965, + "learning_rate": 0.00017897739878272595, + "loss": 0.9236, + "step": 13483 + }, + { + "epoch": 0.9356741378113941, + "grad_norm": 1.8449347019195557, + "learning_rate": 0.00017897395221084156, + "loss": 0.7426, + "step": 13484 + }, + { + "epoch": 0.9357435292484907, + "grad_norm": 4.0978193283081055, + "learning_rate": 0.00017897050538964558, + "loss": 0.7768, + "step": 13485 + }, + { + "epoch": 0.9358129206855874, + "grad_norm": 2.433582067489624, + "learning_rate": 0.00017896705831914892, + "loss": 0.8304, + "step": 13486 + }, + { + "epoch": 0.935882312122684, + "grad_norm": 2.218398332595825, + "learning_rate": 0.00017896361099936235, + "loss": 0.9227, + "step": 13487 + }, + { + "epoch": 0.9359517035597807, + "grad_norm": 1.9888696670532227, + "learning_rate": 0.00017896016343029686, + "loss": 0.8619, + "step": 13488 + }, + { + "epoch": 0.9360210949968774, + "grad_norm": 1.7866824865341187, + "learning_rate": 0.0001789567156119633, + "loss": 0.7615, + "step": 13489 + }, + { + "epoch": 0.9360904864339741, + "grad_norm": 1.8002372980117798, + "learning_rate": 0.00017895326754437254, + "loss": 0.9065, + "step": 13490 + }, + { + "epoch": 0.9361598778710707, + "grad_norm": 2.407210111618042, + "learning_rate": 0.0001789498192275355, + "loss": 1.0502, + "step": 13491 + }, + { + "epoch": 0.9362292693081674, + "grad_norm": 3.4766995906829834, + "learning_rate": 0.000178946370661463, + "loss": 0.8051, + "step": 13492 + }, + { + "epoch": 0.936298660745264, + "grad_norm": 2.081775665283203, + "learning_rate": 0.00017894292184616597, + "loss": 0.8154, + "step": 13493 + }, + { + "epoch": 0.9363680521823607, + "grad_norm": 2.383373737335205, + "learning_rate": 0.0001789394727816553, + "loss": 0.941, + "step": 13494 + }, + { + "epoch": 0.9364374436194574, + "grad_norm": 2.184380531311035, + "learning_rate": 0.0001789360234679419, + "loss": 0.7275, + "step": 13495 + }, + { + "epoch": 0.9365068350565541, + "grad_norm": 2.225212335586548, + "learning_rate": 0.00017893257390503659, + "loss": 0.7922, + "step": 13496 + }, + { + "epoch": 0.9365762264936507, + "grad_norm": 1.86916184425354, + "learning_rate": 0.00017892912409295033, + "loss": 0.6382, + "step": 13497 + }, + { + "epoch": 0.9366456179307473, + "grad_norm": 2.593590497970581, + "learning_rate": 0.00017892567403169398, + "loss": 0.868, + "step": 13498 + }, + { + "epoch": 0.936715009367844, + "grad_norm": 1.8669236898422241, + "learning_rate": 0.00017892222372127843, + "loss": 0.7629, + "step": 13499 + }, + { + "epoch": 0.9367844008049406, + "grad_norm": 1.657619595527649, + "learning_rate": 0.00017891877316171458, + "loss": 0.6263, + "step": 13500 + }, + { + "epoch": 0.9368537922420374, + "grad_norm": 2.1358230113983154, + "learning_rate": 0.0001789153223530133, + "loss": 0.8436, + "step": 13501 + }, + { + "epoch": 0.936923183679134, + "grad_norm": 2.724534273147583, + "learning_rate": 0.00017891187129518552, + "loss": 0.8401, + "step": 13502 + }, + { + "epoch": 0.9369925751162307, + "grad_norm": 2.6310300827026367, + "learning_rate": 0.00017890841998824213, + "loss": 1.0153, + "step": 13503 + }, + { + "epoch": 0.9370619665533273, + "grad_norm": 2.5411949157714844, + "learning_rate": 0.000178904968432194, + "loss": 0.7799, + "step": 13504 + }, + { + "epoch": 0.937131357990424, + "grad_norm": 2.108160972595215, + "learning_rate": 0.000178901516627052, + "loss": 0.8763, + "step": 13505 + }, + { + "epoch": 0.9372007494275206, + "grad_norm": 1.957393765449524, + "learning_rate": 0.00017889806457282713, + "loss": 0.832, + "step": 13506 + }, + { + "epoch": 0.9372701408646174, + "grad_norm": 2.8327739238739014, + "learning_rate": 0.0001788946122695302, + "loss": 0.9414, + "step": 13507 + }, + { + "epoch": 0.937339532301714, + "grad_norm": 2.616527795791626, + "learning_rate": 0.00017889115971717214, + "loss": 0.8922, + "step": 13508 + }, + { + "epoch": 0.9374089237388107, + "grad_norm": 2.077094793319702, + "learning_rate": 0.00017888770691576382, + "loss": 0.6473, + "step": 13509 + }, + { + "epoch": 0.9374783151759073, + "grad_norm": 1.8804553747177124, + "learning_rate": 0.0001788842538653162, + "loss": 0.6648, + "step": 13510 + }, + { + "epoch": 0.9375477066130039, + "grad_norm": 1.839637279510498, + "learning_rate": 0.0001788808005658401, + "loss": 0.6202, + "step": 13511 + }, + { + "epoch": 0.9376170980501006, + "grad_norm": 2.3641562461853027, + "learning_rate": 0.00017887734701734647, + "loss": 0.753, + "step": 13512 + }, + { + "epoch": 0.9376864894871972, + "grad_norm": 2.2011215686798096, + "learning_rate": 0.00017887389321984625, + "loss": 0.7488, + "step": 13513 + }, + { + "epoch": 0.937755880924294, + "grad_norm": 1.8570117950439453, + "learning_rate": 0.00017887043917335027, + "loss": 0.7852, + "step": 13514 + }, + { + "epoch": 0.9378252723613906, + "grad_norm": 2.250002384185791, + "learning_rate": 0.00017886698487786946, + "loss": 0.8421, + "step": 13515 + }, + { + "epoch": 0.9378946637984873, + "grad_norm": 1.7521789073944092, + "learning_rate": 0.00017886353033341474, + "loss": 0.6467, + "step": 13516 + }, + { + "epoch": 0.9379640552355839, + "grad_norm": 2.1202504634857178, + "learning_rate": 0.000178860075539997, + "loss": 0.6286, + "step": 13517 + }, + { + "epoch": 0.9380334466726806, + "grad_norm": 2.0463287830352783, + "learning_rate": 0.00017885662049762715, + "loss": 0.6181, + "step": 13518 + }, + { + "epoch": 0.9381028381097772, + "grad_norm": 2.233412981033325, + "learning_rate": 0.00017885316520631608, + "loss": 1.0352, + "step": 13519 + }, + { + "epoch": 0.938172229546874, + "grad_norm": 2.6303658485412598, + "learning_rate": 0.00017884970966607473, + "loss": 0.8839, + "step": 13520 + }, + { + "epoch": 0.9382416209839706, + "grad_norm": 2.152285575866699, + "learning_rate": 0.00017884625387691401, + "loss": 0.8398, + "step": 13521 + }, + { + "epoch": 0.9383110124210673, + "grad_norm": 2.4299991130828857, + "learning_rate": 0.0001788427978388448, + "loss": 1.1346, + "step": 13522 + }, + { + "epoch": 0.9383804038581639, + "grad_norm": 2.004187822341919, + "learning_rate": 0.000178839341551878, + "loss": 0.6784, + "step": 13523 + }, + { + "epoch": 0.9384497952952605, + "grad_norm": 2.0906684398651123, + "learning_rate": 0.0001788358850160246, + "loss": 0.7416, + "step": 13524 + }, + { + "epoch": 0.9385191867323572, + "grad_norm": 1.5445196628570557, + "learning_rate": 0.00017883242823129542, + "loss": 0.6538, + "step": 13525 + }, + { + "epoch": 0.9385885781694538, + "grad_norm": 2.0968286991119385, + "learning_rate": 0.0001788289711977014, + "loss": 0.8506, + "step": 13526 + }, + { + "epoch": 0.9386579696065506, + "grad_norm": 2.0313336849212646, + "learning_rate": 0.0001788255139152535, + "loss": 0.821, + "step": 13527 + }, + { + "epoch": 0.9387273610436472, + "grad_norm": 2.198111057281494, + "learning_rate": 0.00017882205638396258, + "loss": 0.8373, + "step": 13528 + }, + { + "epoch": 0.9387967524807439, + "grad_norm": 1.7798981666564941, + "learning_rate": 0.0001788185986038396, + "loss": 0.6307, + "step": 13529 + }, + { + "epoch": 0.9388661439178405, + "grad_norm": 2.0080947875976562, + "learning_rate": 0.00017881514057489542, + "loss": 0.7505, + "step": 13530 + }, + { + "epoch": 0.9389355353549372, + "grad_norm": 1.239169716835022, + "learning_rate": 0.00017881168229714098, + "loss": 0.4789, + "step": 13531 + }, + { + "epoch": 0.9390049267920338, + "grad_norm": 1.9357209205627441, + "learning_rate": 0.0001788082237705872, + "loss": 0.7671, + "step": 13532 + }, + { + "epoch": 0.9390743182291306, + "grad_norm": 2.1776037216186523, + "learning_rate": 0.00017880476499524504, + "loss": 0.8046, + "step": 13533 + }, + { + "epoch": 0.9391437096662272, + "grad_norm": 1.694519281387329, + "learning_rate": 0.00017880130597112533, + "loss": 0.796, + "step": 13534 + }, + { + "epoch": 0.9392131011033239, + "grad_norm": 2.3445606231689453, + "learning_rate": 0.0001787978466982391, + "loss": 0.7788, + "step": 13535 + }, + { + "epoch": 0.9392824925404205, + "grad_norm": 1.9846981763839722, + "learning_rate": 0.00017879438717659712, + "loss": 0.8867, + "step": 13536 + }, + { + "epoch": 0.9393518839775171, + "grad_norm": 1.951723575592041, + "learning_rate": 0.0001787909274062105, + "loss": 0.6771, + "step": 13537 + }, + { + "epoch": 0.9394212754146138, + "grad_norm": 1.8560012578964233, + "learning_rate": 0.00017878746738709, + "loss": 0.632, + "step": 13538 + }, + { + "epoch": 0.9394906668517105, + "grad_norm": 2.309992790222168, + "learning_rate": 0.00017878400711924663, + "loss": 0.8538, + "step": 13539 + }, + { + "epoch": 0.9395600582888072, + "grad_norm": 2.3391997814178467, + "learning_rate": 0.00017878054660269127, + "loss": 0.8596, + "step": 13540 + }, + { + "epoch": 0.9396294497259038, + "grad_norm": 1.7914133071899414, + "learning_rate": 0.0001787770858374349, + "loss": 0.7807, + "step": 13541 + }, + { + "epoch": 0.9396988411630005, + "grad_norm": 2.1038217544555664, + "learning_rate": 0.00017877362482348838, + "loss": 0.8918, + "step": 13542 + }, + { + "epoch": 0.9397682326000971, + "grad_norm": 1.7262667417526245, + "learning_rate": 0.0001787701635608627, + "loss": 0.6868, + "step": 13543 + }, + { + "epoch": 0.9398376240371938, + "grad_norm": 2.567073106765747, + "learning_rate": 0.00017876670204956872, + "loss": 0.9943, + "step": 13544 + }, + { + "epoch": 0.9399070154742905, + "grad_norm": 2.120389938354492, + "learning_rate": 0.00017876324028961738, + "loss": 0.5799, + "step": 13545 + }, + { + "epoch": 0.9399764069113872, + "grad_norm": 2.1160714626312256, + "learning_rate": 0.00017875977828101967, + "loss": 0.6904, + "step": 13546 + }, + { + "epoch": 0.9400457983484838, + "grad_norm": 3.907782793045044, + "learning_rate": 0.00017875631602378645, + "loss": 0.9442, + "step": 13547 + }, + { + "epoch": 0.9401151897855805, + "grad_norm": 1.6221675872802734, + "learning_rate": 0.0001787528535179287, + "loss": 0.5663, + "step": 13548 + }, + { + "epoch": 0.9401845812226771, + "grad_norm": 1.5328415632247925, + "learning_rate": 0.00017874939076345734, + "loss": 0.6512, + "step": 13549 + }, + { + "epoch": 0.9402539726597737, + "grad_norm": 1.8012021780014038, + "learning_rate": 0.00017874592776038326, + "loss": 0.8378, + "step": 13550 + }, + { + "epoch": 0.9403233640968705, + "grad_norm": 1.8016499280929565, + "learning_rate": 0.00017874246450871745, + "loss": 0.809, + "step": 13551 + }, + { + "epoch": 0.9403927555339671, + "grad_norm": 2.2189810276031494, + "learning_rate": 0.00017873900100847078, + "loss": 0.6766, + "step": 13552 + }, + { + "epoch": 0.9404621469710638, + "grad_norm": 3.1343228816986084, + "learning_rate": 0.00017873553725965423, + "loss": 1.1581, + "step": 13553 + }, + { + "epoch": 0.9405315384081604, + "grad_norm": 3.3625121116638184, + "learning_rate": 0.00017873207326227875, + "loss": 0.9899, + "step": 13554 + }, + { + "epoch": 0.9406009298452571, + "grad_norm": 1.8683348894119263, + "learning_rate": 0.0001787286090163552, + "loss": 0.9408, + "step": 13555 + }, + { + "epoch": 0.9406703212823537, + "grad_norm": 2.603774070739746, + "learning_rate": 0.0001787251445218946, + "loss": 0.8405, + "step": 13556 + }, + { + "epoch": 0.9407397127194504, + "grad_norm": 2.6062676906585693, + "learning_rate": 0.00017872167977890787, + "loss": 0.7128, + "step": 13557 + }, + { + "epoch": 0.9408091041565471, + "grad_norm": 1.568172812461853, + "learning_rate": 0.00017871821478740592, + "loss": 0.7902, + "step": 13558 + }, + { + "epoch": 0.9408784955936438, + "grad_norm": 2.1649057865142822, + "learning_rate": 0.0001787147495473997, + "loss": 0.7961, + "step": 13559 + }, + { + "epoch": 0.9409478870307404, + "grad_norm": 2.3374154567718506, + "learning_rate": 0.00017871128405890012, + "loss": 0.8402, + "step": 13560 + }, + { + "epoch": 0.9410172784678371, + "grad_norm": 2.4257500171661377, + "learning_rate": 0.00017870781832191816, + "loss": 0.6605, + "step": 13561 + }, + { + "epoch": 0.9410866699049337, + "grad_norm": 2.5200204849243164, + "learning_rate": 0.00017870435233646473, + "loss": 1.1533, + "step": 13562 + }, + { + "epoch": 0.9411560613420304, + "grad_norm": 6.750542163848877, + "learning_rate": 0.00017870088610255085, + "loss": 0.7838, + "step": 13563 + }, + { + "epoch": 0.9412254527791271, + "grad_norm": 2.678269863128662, + "learning_rate": 0.00017869741962018738, + "loss": 1.0743, + "step": 13564 + }, + { + "epoch": 0.9412948442162237, + "grad_norm": 2.043128252029419, + "learning_rate": 0.00017869395288938527, + "loss": 0.9328, + "step": 13565 + }, + { + "epoch": 0.9413642356533204, + "grad_norm": 1.559792399406433, + "learning_rate": 0.00017869048591015547, + "loss": 0.5912, + "step": 13566 + }, + { + "epoch": 0.941433627090417, + "grad_norm": 2.877643346786499, + "learning_rate": 0.00017868701868250896, + "loss": 0.9797, + "step": 13567 + }, + { + "epoch": 0.9415030185275137, + "grad_norm": 1.9309746026992798, + "learning_rate": 0.00017868355120645665, + "loss": 0.6405, + "step": 13568 + }, + { + "epoch": 0.9415724099646103, + "grad_norm": 1.2797110080718994, + "learning_rate": 0.00017868008348200952, + "loss": 0.4592, + "step": 13569 + }, + { + "epoch": 0.9416418014017071, + "grad_norm": 3.016507148742676, + "learning_rate": 0.00017867661550917846, + "loss": 1.212, + "step": 13570 + }, + { + "epoch": 0.9417111928388037, + "grad_norm": 1.760959267616272, + "learning_rate": 0.00017867314728797448, + "loss": 0.6207, + "step": 13571 + }, + { + "epoch": 0.9417805842759004, + "grad_norm": 2.9485692977905273, + "learning_rate": 0.00017866967881840847, + "loss": 0.6908, + "step": 13572 + }, + { + "epoch": 0.941849975712997, + "grad_norm": 1.7985233068466187, + "learning_rate": 0.00017866621010049146, + "loss": 0.7146, + "step": 13573 + }, + { + "epoch": 0.9419193671500937, + "grad_norm": 2.0519070625305176, + "learning_rate": 0.0001786627411342343, + "loss": 0.8169, + "step": 13574 + }, + { + "epoch": 0.9419887585871903, + "grad_norm": 1.8727580308914185, + "learning_rate": 0.00017865927191964805, + "loss": 0.7261, + "step": 13575 + }, + { + "epoch": 0.942058150024287, + "grad_norm": 1.9991283416748047, + "learning_rate": 0.00017865580245674355, + "loss": 0.9288, + "step": 13576 + }, + { + "epoch": 0.9421275414613837, + "grad_norm": 1.9442936182022095, + "learning_rate": 0.00017865233274553185, + "loss": 1.0331, + "step": 13577 + }, + { + "epoch": 0.9421969328984803, + "grad_norm": 1.8860772848129272, + "learning_rate": 0.00017864886278602383, + "loss": 0.5034, + "step": 13578 + }, + { + "epoch": 0.942266324335577, + "grad_norm": 2.8942549228668213, + "learning_rate": 0.00017864539257823048, + "loss": 1.1412, + "step": 13579 + }, + { + "epoch": 0.9423357157726736, + "grad_norm": 2.8638617992401123, + "learning_rate": 0.00017864192212216278, + "loss": 1.1126, + "step": 13580 + }, + { + "epoch": 0.9424051072097703, + "grad_norm": 2.693394899368286, + "learning_rate": 0.0001786384514178316, + "loss": 0.7787, + "step": 13581 + }, + { + "epoch": 0.9424744986468669, + "grad_norm": 1.809117317199707, + "learning_rate": 0.00017863498046524797, + "loss": 0.7943, + "step": 13582 + }, + { + "epoch": 0.9425438900839637, + "grad_norm": 2.907522678375244, + "learning_rate": 0.00017863150926442287, + "loss": 1.1826, + "step": 13583 + }, + { + "epoch": 0.9426132815210603, + "grad_norm": 3.955052137374878, + "learning_rate": 0.0001786280378153672, + "loss": 0.7404, + "step": 13584 + }, + { + "epoch": 0.942682672958157, + "grad_norm": 3.517308235168457, + "learning_rate": 0.00017862456611809191, + "loss": 0.8795, + "step": 13585 + }, + { + "epoch": 0.9427520643952536, + "grad_norm": 4.103773593902588, + "learning_rate": 0.00017862109417260804, + "loss": 0.8919, + "step": 13586 + }, + { + "epoch": 0.9428214558323503, + "grad_norm": 2.1202969551086426, + "learning_rate": 0.00017861762197892645, + "loss": 0.5898, + "step": 13587 + }, + { + "epoch": 0.9428908472694469, + "grad_norm": 2.0927398204803467, + "learning_rate": 0.00017861414953705817, + "loss": 0.5887, + "step": 13588 + }, + { + "epoch": 0.9429602387065437, + "grad_norm": 2.6220147609710693, + "learning_rate": 0.00017861067684701412, + "loss": 0.7422, + "step": 13589 + }, + { + "epoch": 0.9430296301436403, + "grad_norm": 2.299136161804199, + "learning_rate": 0.00017860720390880532, + "loss": 0.7747, + "step": 13590 + }, + { + "epoch": 0.9430990215807369, + "grad_norm": 1.9838114976882935, + "learning_rate": 0.00017860373072244265, + "loss": 0.8211, + "step": 13591 + }, + { + "epoch": 0.9431684130178336, + "grad_norm": 1.932910442352295, + "learning_rate": 0.00017860025728793713, + "loss": 0.6088, + "step": 13592 + }, + { + "epoch": 0.9432378044549302, + "grad_norm": 2.161977529525757, + "learning_rate": 0.00017859678360529976, + "loss": 0.8384, + "step": 13593 + }, + { + "epoch": 0.9433071958920269, + "grad_norm": 1.6392905712127686, + "learning_rate": 0.00017859330967454144, + "loss": 0.6902, + "step": 13594 + }, + { + "epoch": 0.9433765873291236, + "grad_norm": 2.0901894569396973, + "learning_rate": 0.00017858983549567316, + "loss": 0.7103, + "step": 13595 + }, + { + "epoch": 0.9434459787662203, + "grad_norm": 2.126931667327881, + "learning_rate": 0.0001785863610687059, + "loss": 1.0629, + "step": 13596 + }, + { + "epoch": 0.9435153702033169, + "grad_norm": 3.148068904876709, + "learning_rate": 0.00017858288639365058, + "loss": 0.8915, + "step": 13597 + }, + { + "epoch": 0.9435847616404136, + "grad_norm": 2.75091814994812, + "learning_rate": 0.00017857941147051822, + "loss": 0.7339, + "step": 13598 + }, + { + "epoch": 0.9436541530775102, + "grad_norm": 2.9195544719696045, + "learning_rate": 0.00017857593629931978, + "loss": 0.9358, + "step": 13599 + }, + { + "epoch": 0.9437235445146069, + "grad_norm": 2.8371243476867676, + "learning_rate": 0.00017857246088006623, + "loss": 1.0057, + "step": 13600 + }, + { + "epoch": 0.9437929359517035, + "grad_norm": 2.6561713218688965, + "learning_rate": 0.00017856898521276853, + "loss": 0.8466, + "step": 13601 + }, + { + "epoch": 0.9438623273888003, + "grad_norm": 2.439138174057007, + "learning_rate": 0.00017856550929743766, + "loss": 0.891, + "step": 13602 + }, + { + "epoch": 0.9439317188258969, + "grad_norm": 1.826606035232544, + "learning_rate": 0.00017856203313408462, + "loss": 0.8441, + "step": 13603 + }, + { + "epoch": 0.9440011102629935, + "grad_norm": 1.8166753053665161, + "learning_rate": 0.00017855855672272033, + "loss": 0.6179, + "step": 13604 + }, + { + "epoch": 0.9440705017000902, + "grad_norm": 2.524885654449463, + "learning_rate": 0.00017855508006335577, + "loss": 0.6362, + "step": 13605 + }, + { + "epoch": 0.9441398931371868, + "grad_norm": 1.9566923379898071, + "learning_rate": 0.00017855160315600194, + "loss": 0.7737, + "step": 13606 + }, + { + "epoch": 0.9442092845742835, + "grad_norm": 3.0151095390319824, + "learning_rate": 0.00017854812600066983, + "loss": 1.011, + "step": 13607 + }, + { + "epoch": 0.9442786760113802, + "grad_norm": 2.0807244777679443, + "learning_rate": 0.00017854464859737044, + "loss": 0.8789, + "step": 13608 + }, + { + "epoch": 0.9443480674484769, + "grad_norm": 2.6108639240264893, + "learning_rate": 0.00017854117094611466, + "loss": 0.9006, + "step": 13609 + }, + { + "epoch": 0.9444174588855735, + "grad_norm": 3.9584591388702393, + "learning_rate": 0.0001785376930469135, + "loss": 0.9298, + "step": 13610 + }, + { + "epoch": 0.9444868503226702, + "grad_norm": 2.3219921588897705, + "learning_rate": 0.000178534214899778, + "loss": 0.8118, + "step": 13611 + }, + { + "epoch": 0.9445562417597668, + "grad_norm": 2.3203227519989014, + "learning_rate": 0.00017853073650471907, + "loss": 0.7506, + "step": 13612 + }, + { + "epoch": 0.9446256331968635, + "grad_norm": 2.3282721042633057, + "learning_rate": 0.0001785272578617477, + "loss": 0.7243, + "step": 13613 + }, + { + "epoch": 0.9446950246339602, + "grad_norm": 2.5339407920837402, + "learning_rate": 0.0001785237789708749, + "loss": 0.7104, + "step": 13614 + }, + { + "epoch": 0.9447644160710569, + "grad_norm": 2.577979564666748, + "learning_rate": 0.00017852029983211166, + "loss": 0.726, + "step": 13615 + }, + { + "epoch": 0.9448338075081535, + "grad_norm": 2.5466866493225098, + "learning_rate": 0.00017851682044546892, + "loss": 1.0905, + "step": 13616 + }, + { + "epoch": 0.9449031989452502, + "grad_norm": 2.41178297996521, + "learning_rate": 0.0001785133408109577, + "loss": 0.85, + "step": 13617 + }, + { + "epoch": 0.9449725903823468, + "grad_norm": 2.129786491394043, + "learning_rate": 0.00017850986092858898, + "loss": 0.6913, + "step": 13618 + }, + { + "epoch": 0.9450419818194434, + "grad_norm": 1.9366174936294556, + "learning_rate": 0.00017850638079837375, + "loss": 0.9333, + "step": 13619 + }, + { + "epoch": 0.9451113732565402, + "grad_norm": 1.7534927129745483, + "learning_rate": 0.00017850290042032297, + "loss": 0.787, + "step": 13620 + }, + { + "epoch": 0.9451807646936368, + "grad_norm": 4.227016448974609, + "learning_rate": 0.00017849941979444762, + "loss": 1.0063, + "step": 13621 + }, + { + "epoch": 0.9452501561307335, + "grad_norm": 2.175102472305298, + "learning_rate": 0.00017849593892075872, + "loss": 0.7387, + "step": 13622 + }, + { + "epoch": 0.9453195475678301, + "grad_norm": 1.8988059759140015, + "learning_rate": 0.00017849245779926726, + "loss": 0.5661, + "step": 13623 + }, + { + "epoch": 0.9453889390049268, + "grad_norm": 2.1054067611694336, + "learning_rate": 0.00017848897642998418, + "loss": 0.9108, + "step": 13624 + }, + { + "epoch": 0.9454583304420234, + "grad_norm": 2.348560333251953, + "learning_rate": 0.00017848549481292057, + "loss": 0.6519, + "step": 13625 + }, + { + "epoch": 0.9455277218791202, + "grad_norm": 2.867863655090332, + "learning_rate": 0.0001784820129480873, + "loss": 0.9098, + "step": 13626 + }, + { + "epoch": 0.9455971133162168, + "grad_norm": 3.244174003601074, + "learning_rate": 0.00017847853083549547, + "loss": 0.6215, + "step": 13627 + }, + { + "epoch": 0.9456665047533135, + "grad_norm": 2.5305373668670654, + "learning_rate": 0.000178475048475156, + "loss": 0.8315, + "step": 13628 + }, + { + "epoch": 0.9457358961904101, + "grad_norm": 2.577134847640991, + "learning_rate": 0.00017847156586707992, + "loss": 0.8296, + "step": 13629 + }, + { + "epoch": 0.9458052876275068, + "grad_norm": 2.064314842224121, + "learning_rate": 0.00017846808301127818, + "loss": 0.629, + "step": 13630 + }, + { + "epoch": 0.9458746790646034, + "grad_norm": 2.3616137504577637, + "learning_rate": 0.00017846459990776184, + "loss": 0.8422, + "step": 13631 + }, + { + "epoch": 0.9459440705017, + "grad_norm": 1.9075013399124146, + "learning_rate": 0.00017846111655654183, + "loss": 0.8159, + "step": 13632 + }, + { + "epoch": 0.9460134619387968, + "grad_norm": 1.7797491550445557, + "learning_rate": 0.0001784576329576292, + "loss": 0.4814, + "step": 13633 + }, + { + "epoch": 0.9460828533758934, + "grad_norm": 1.9642655849456787, + "learning_rate": 0.00017845414911103492, + "loss": 0.8838, + "step": 13634 + }, + { + "epoch": 0.9461522448129901, + "grad_norm": 2.0894181728363037, + "learning_rate": 0.00017845066501676998, + "loss": 0.8984, + "step": 13635 + }, + { + "epoch": 0.9462216362500867, + "grad_norm": 1.5515813827514648, + "learning_rate": 0.0001784471806748454, + "loss": 0.68, + "step": 13636 + }, + { + "epoch": 0.9462910276871834, + "grad_norm": 2.3785111904144287, + "learning_rate": 0.00017844369608527219, + "loss": 0.6847, + "step": 13637 + }, + { + "epoch": 0.94636041912428, + "grad_norm": 2.3017940521240234, + "learning_rate": 0.0001784402112480613, + "loss": 0.9886, + "step": 13638 + }, + { + "epoch": 0.9464298105613768, + "grad_norm": 4.033344268798828, + "learning_rate": 0.0001784367261632238, + "loss": 0.8978, + "step": 13639 + }, + { + "epoch": 0.9464992019984734, + "grad_norm": 2.6299867630004883, + "learning_rate": 0.00017843324083077064, + "loss": 1.097, + "step": 13640 + }, + { + "epoch": 0.9465685934355701, + "grad_norm": 2.6554505825042725, + "learning_rate": 0.00017842975525071285, + "loss": 0.8733, + "step": 13641 + }, + { + "epoch": 0.9466379848726667, + "grad_norm": 2.4711954593658447, + "learning_rate": 0.0001784262694230614, + "loss": 0.7301, + "step": 13642 + }, + { + "epoch": 0.9467073763097634, + "grad_norm": 2.2402634620666504, + "learning_rate": 0.00017842278334782733, + "loss": 0.7053, + "step": 13643 + }, + { + "epoch": 0.94677676774686, + "grad_norm": 2.0900018215179443, + "learning_rate": 0.00017841929702502163, + "loss": 0.7294, + "step": 13644 + }, + { + "epoch": 0.9468461591839566, + "grad_norm": 2.5138704776763916, + "learning_rate": 0.0001784158104546553, + "loss": 0.8832, + "step": 13645 + }, + { + "epoch": 0.9469155506210534, + "grad_norm": 1.8381481170654297, + "learning_rate": 0.00017841232363673934, + "loss": 0.6342, + "step": 13646 + }, + { + "epoch": 0.94698494205815, + "grad_norm": 1.884104609489441, + "learning_rate": 0.0001784088365712848, + "loss": 0.7866, + "step": 13647 + }, + { + "epoch": 0.9470543334952467, + "grad_norm": 2.5958456993103027, + "learning_rate": 0.00017840534925830267, + "loss": 0.7034, + "step": 13648 + }, + { + "epoch": 0.9471237249323433, + "grad_norm": 2.3081891536712646, + "learning_rate": 0.00017840186169780392, + "loss": 0.9136, + "step": 13649 + }, + { + "epoch": 0.94719311636944, + "grad_norm": 2.934802293777466, + "learning_rate": 0.0001783983738897996, + "loss": 0.8557, + "step": 13650 + }, + { + "epoch": 0.9472625078065366, + "grad_norm": 3.1497738361358643, + "learning_rate": 0.0001783948858343007, + "loss": 0.8482, + "step": 13651 + }, + { + "epoch": 0.9473318992436334, + "grad_norm": 2.333076238632202, + "learning_rate": 0.0001783913975313183, + "loss": 0.9083, + "step": 13652 + }, + { + "epoch": 0.94740129068073, + "grad_norm": 2.0924086570739746, + "learning_rate": 0.00017838790898086327, + "loss": 0.7487, + "step": 13653 + }, + { + "epoch": 0.9474706821178267, + "grad_norm": 1.8855705261230469, + "learning_rate": 0.00017838442018294675, + "loss": 0.6583, + "step": 13654 + }, + { + "epoch": 0.9475400735549233, + "grad_norm": 2.8146560192108154, + "learning_rate": 0.0001783809311375797, + "loss": 0.7403, + "step": 13655 + }, + { + "epoch": 0.94760946499202, + "grad_norm": 1.6629209518432617, + "learning_rate": 0.00017837744184477313, + "loss": 0.5796, + "step": 13656 + }, + { + "epoch": 0.9476788564291166, + "grad_norm": 1.8695423603057861, + "learning_rate": 0.00017837395230453808, + "loss": 0.6569, + "step": 13657 + }, + { + "epoch": 0.9477482478662133, + "grad_norm": 2.659494638442993, + "learning_rate": 0.00017837046251688556, + "loss": 0.8211, + "step": 13658 + }, + { + "epoch": 0.94781763930331, + "grad_norm": 2.1543641090393066, + "learning_rate": 0.0001783669724818266, + "loss": 0.6546, + "step": 13659 + }, + { + "epoch": 0.9478870307404066, + "grad_norm": 2.819467067718506, + "learning_rate": 0.0001783634821993722, + "loss": 0.9915, + "step": 13660 + }, + { + "epoch": 0.9479564221775033, + "grad_norm": 1.566247582435608, + "learning_rate": 0.00017835999166953333, + "loss": 0.7107, + "step": 13661 + }, + { + "epoch": 0.9480258136145999, + "grad_norm": 2.022674560546875, + "learning_rate": 0.0001783565008923211, + "loss": 0.7585, + "step": 13662 + }, + { + "epoch": 0.9480952050516966, + "grad_norm": 2.1040446758270264, + "learning_rate": 0.00017835300986774648, + "loss": 0.5999, + "step": 13663 + }, + { + "epoch": 0.9481645964887933, + "grad_norm": 2.3478269577026367, + "learning_rate": 0.00017834951859582048, + "loss": 0.7353, + "step": 13664 + }, + { + "epoch": 0.94823398792589, + "grad_norm": 2.157134532928467, + "learning_rate": 0.00017834602707655415, + "loss": 0.8631, + "step": 13665 + }, + { + "epoch": 0.9483033793629866, + "grad_norm": 2.5320348739624023, + "learning_rate": 0.0001783425353099585, + "loss": 0.9888, + "step": 13666 + }, + { + "epoch": 0.9483727708000833, + "grad_norm": 2.6321866512298584, + "learning_rate": 0.00017833904329604455, + "loss": 0.6595, + "step": 13667 + }, + { + "epoch": 0.9484421622371799, + "grad_norm": 2.491560697555542, + "learning_rate": 0.00017833555103482338, + "loss": 0.9227, + "step": 13668 + }, + { + "epoch": 0.9485115536742766, + "grad_norm": 2.3192267417907715, + "learning_rate": 0.0001783320585263059, + "loss": 0.9806, + "step": 13669 + }, + { + "epoch": 0.9485809451113733, + "grad_norm": 2.1627230644226074, + "learning_rate": 0.00017832856577050323, + "loss": 0.8218, + "step": 13670 + }, + { + "epoch": 0.9486503365484699, + "grad_norm": 2.0743868350982666, + "learning_rate": 0.00017832507276742637, + "loss": 0.8477, + "step": 13671 + }, + { + "epoch": 0.9487197279855666, + "grad_norm": 2.413945436477661, + "learning_rate": 0.0001783215795170863, + "loss": 0.859, + "step": 13672 + }, + { + "epoch": 0.9487891194226632, + "grad_norm": 2.148033857345581, + "learning_rate": 0.00017831808601949416, + "loss": 1.1417, + "step": 13673 + }, + { + "epoch": 0.9488585108597599, + "grad_norm": 2.019972801208496, + "learning_rate": 0.00017831459227466088, + "loss": 0.9098, + "step": 13674 + }, + { + "epoch": 0.9489279022968565, + "grad_norm": 2.0945162773132324, + "learning_rate": 0.0001783110982825975, + "loss": 0.7341, + "step": 13675 + }, + { + "epoch": 0.9489972937339533, + "grad_norm": 2.908447265625, + "learning_rate": 0.00017830760404331507, + "loss": 1.0497, + "step": 13676 + }, + { + "epoch": 0.9490666851710499, + "grad_norm": 2.0045440196990967, + "learning_rate": 0.00017830410955682464, + "loss": 0.7193, + "step": 13677 + }, + { + "epoch": 0.9491360766081466, + "grad_norm": 1.8626521825790405, + "learning_rate": 0.0001783006148231372, + "loss": 0.7816, + "step": 13678 + }, + { + "epoch": 0.9492054680452432, + "grad_norm": 2.4986751079559326, + "learning_rate": 0.00017829711984226383, + "loss": 0.8502, + "step": 13679 + }, + { + "epoch": 0.9492748594823399, + "grad_norm": 1.7177249193191528, + "learning_rate": 0.00017829362461421552, + "loss": 0.7007, + "step": 13680 + }, + { + "epoch": 0.9493442509194365, + "grad_norm": 2.0872414112091064, + "learning_rate": 0.00017829012913900329, + "loss": 0.7237, + "step": 13681 + }, + { + "epoch": 0.9494136423565332, + "grad_norm": 2.071950912475586, + "learning_rate": 0.00017828663341663825, + "loss": 0.862, + "step": 13682 + }, + { + "epoch": 0.9494830337936299, + "grad_norm": 2.073241710662842, + "learning_rate": 0.0001782831374471314, + "loss": 0.8971, + "step": 13683 + }, + { + "epoch": 0.9495524252307266, + "grad_norm": 2.0609419345855713, + "learning_rate": 0.00017827964123049372, + "loss": 0.751, + "step": 13684 + }, + { + "epoch": 0.9496218166678232, + "grad_norm": 3.6249985694885254, + "learning_rate": 0.00017827614476673636, + "loss": 1.0141, + "step": 13685 + }, + { + "epoch": 0.9496912081049198, + "grad_norm": 3.1848361492156982, + "learning_rate": 0.00017827264805587024, + "loss": 0.7586, + "step": 13686 + }, + { + "epoch": 0.9497605995420165, + "grad_norm": 1.651969075202942, + "learning_rate": 0.00017826915109790645, + "loss": 0.5763, + "step": 13687 + }, + { + "epoch": 0.9498299909791131, + "grad_norm": 2.464061975479126, + "learning_rate": 0.00017826565389285603, + "loss": 0.8075, + "step": 13688 + }, + { + "epoch": 0.9498993824162099, + "grad_norm": 2.344593048095703, + "learning_rate": 0.00017826215644073005, + "loss": 0.7998, + "step": 13689 + }, + { + "epoch": 0.9499687738533065, + "grad_norm": 1.5881965160369873, + "learning_rate": 0.00017825865874153952, + "loss": 0.5976, + "step": 13690 + }, + { + "epoch": 0.9500381652904032, + "grad_norm": 2.7664408683776855, + "learning_rate": 0.00017825516079529547, + "loss": 0.8797, + "step": 13691 + }, + { + "epoch": 0.9501075567274998, + "grad_norm": 4.1415486335754395, + "learning_rate": 0.00017825166260200897, + "loss": 0.9545, + "step": 13692 + }, + { + "epoch": 0.9501769481645965, + "grad_norm": 2.450834035873413, + "learning_rate": 0.00017824816416169101, + "loss": 0.975, + "step": 13693 + }, + { + "epoch": 0.9502463396016931, + "grad_norm": 2.7588794231414795, + "learning_rate": 0.00017824466547435272, + "loss": 0.6688, + "step": 13694 + }, + { + "epoch": 0.9503157310387899, + "grad_norm": 2.7968525886535645, + "learning_rate": 0.00017824116654000508, + "loss": 1.019, + "step": 13695 + }, + { + "epoch": 0.9503851224758865, + "grad_norm": 2.2991812229156494, + "learning_rate": 0.00017823766735865917, + "loss": 1.0609, + "step": 13696 + }, + { + "epoch": 0.9504545139129832, + "grad_norm": 7.538624286651611, + "learning_rate": 0.00017823416793032602, + "loss": 0.882, + "step": 13697 + }, + { + "epoch": 0.9505239053500798, + "grad_norm": 2.5852465629577637, + "learning_rate": 0.00017823066825501665, + "loss": 0.7272, + "step": 13698 + }, + { + "epoch": 0.9505932967871764, + "grad_norm": 1.8913285732269287, + "learning_rate": 0.00017822716833274215, + "loss": 0.587, + "step": 13699 + }, + { + "epoch": 0.9506626882242731, + "grad_norm": 2.741767644882202, + "learning_rate": 0.00017822366816351356, + "loss": 0.6582, + "step": 13700 + }, + { + "epoch": 0.9507320796613697, + "grad_norm": 3.9050400257110596, + "learning_rate": 0.00017822016774734193, + "loss": 1.0906, + "step": 13701 + }, + { + "epoch": 0.9508014710984665, + "grad_norm": 2.4297983646392822, + "learning_rate": 0.0001782166670842383, + "loss": 0.7424, + "step": 13702 + }, + { + "epoch": 0.9508708625355631, + "grad_norm": 1.9458906650543213, + "learning_rate": 0.00017821316617421372, + "loss": 1.0055, + "step": 13703 + }, + { + "epoch": 0.9509402539726598, + "grad_norm": 2.182995080947876, + "learning_rate": 0.00017820966501727928, + "loss": 0.8484, + "step": 13704 + }, + { + "epoch": 0.9510096454097564, + "grad_norm": 2.4475414752960205, + "learning_rate": 0.00017820616361344596, + "loss": 1.085, + "step": 13705 + }, + { + "epoch": 0.9510790368468531, + "grad_norm": 2.117281198501587, + "learning_rate": 0.00017820266196272487, + "loss": 0.7815, + "step": 13706 + }, + { + "epoch": 0.9511484282839497, + "grad_norm": 2.031365156173706, + "learning_rate": 0.00017819916006512707, + "loss": 0.8469, + "step": 13707 + }, + { + "epoch": 0.9512178197210465, + "grad_norm": 2.206669330596924, + "learning_rate": 0.00017819565792066358, + "loss": 0.8588, + "step": 13708 + }, + { + "epoch": 0.9512872111581431, + "grad_norm": 3.0879008769989014, + "learning_rate": 0.00017819215552934546, + "loss": 0.7678, + "step": 13709 + }, + { + "epoch": 0.9513566025952398, + "grad_norm": 2.19771409034729, + "learning_rate": 0.0001781886528911838, + "loss": 0.688, + "step": 13710 + }, + { + "epoch": 0.9514259940323364, + "grad_norm": 2.081299304962158, + "learning_rate": 0.0001781851500061896, + "loss": 0.7703, + "step": 13711 + }, + { + "epoch": 0.951495385469433, + "grad_norm": 2.451704502105713, + "learning_rate": 0.00017818164687437397, + "loss": 0.9255, + "step": 13712 + }, + { + "epoch": 0.9515647769065297, + "grad_norm": 2.463959217071533, + "learning_rate": 0.00017817814349574795, + "loss": 0.8969, + "step": 13713 + }, + { + "epoch": 0.9516341683436264, + "grad_norm": 2.4552056789398193, + "learning_rate": 0.0001781746398703226, + "loss": 0.8816, + "step": 13714 + }, + { + "epoch": 0.9517035597807231, + "grad_norm": 1.8779200315475464, + "learning_rate": 0.00017817113599810899, + "loss": 0.6842, + "step": 13715 + }, + { + "epoch": 0.9517729512178197, + "grad_norm": 2.1162469387054443, + "learning_rate": 0.00017816763187911817, + "loss": 0.739, + "step": 13716 + }, + { + "epoch": 0.9518423426549164, + "grad_norm": 2.3200719356536865, + "learning_rate": 0.0001781641275133612, + "loss": 1.1049, + "step": 13717 + }, + { + "epoch": 0.951911734092013, + "grad_norm": 2.1139984130859375, + "learning_rate": 0.00017816062290084916, + "loss": 0.7703, + "step": 13718 + }, + { + "epoch": 0.9519811255291097, + "grad_norm": 1.733342170715332, + "learning_rate": 0.0001781571180415931, + "loss": 0.5292, + "step": 13719 + }, + { + "epoch": 0.9520505169662064, + "grad_norm": 1.8769457340240479, + "learning_rate": 0.00017815361293560406, + "loss": 0.5709, + "step": 13720 + }, + { + "epoch": 0.9521199084033031, + "grad_norm": 1.85630202293396, + "learning_rate": 0.00017815010758289316, + "loss": 0.7046, + "step": 13721 + }, + { + "epoch": 0.9521892998403997, + "grad_norm": 3.4020395278930664, + "learning_rate": 0.00017814660198347144, + "loss": 1.1013, + "step": 13722 + }, + { + "epoch": 0.9522586912774964, + "grad_norm": 1.9235799312591553, + "learning_rate": 0.00017814309613734994, + "loss": 0.6425, + "step": 13723 + }, + { + "epoch": 0.952328082714593, + "grad_norm": 2.580552339553833, + "learning_rate": 0.00017813959004453977, + "loss": 0.8925, + "step": 13724 + }, + { + "epoch": 0.9523974741516896, + "grad_norm": 1.662161946296692, + "learning_rate": 0.000178136083705052, + "loss": 0.681, + "step": 13725 + }, + { + "epoch": 0.9524668655887863, + "grad_norm": 2.0511884689331055, + "learning_rate": 0.00017813257711889763, + "loss": 0.9162, + "step": 13726 + }, + { + "epoch": 0.952536257025883, + "grad_norm": 2.170663833618164, + "learning_rate": 0.00017812907028608784, + "loss": 0.754, + "step": 13727 + }, + { + "epoch": 0.9526056484629797, + "grad_norm": 2.5953807830810547, + "learning_rate": 0.00017812556320663357, + "loss": 1.2569, + "step": 13728 + }, + { + "epoch": 0.9526750399000763, + "grad_norm": 2.205906629562378, + "learning_rate": 0.000178122055880546, + "loss": 0.8766, + "step": 13729 + }, + { + "epoch": 0.952744431337173, + "grad_norm": 1.8719228506088257, + "learning_rate": 0.00017811854830783615, + "loss": 0.7913, + "step": 13730 + }, + { + "epoch": 0.9528138227742696, + "grad_norm": 2.016129732131958, + "learning_rate": 0.00017811504048851514, + "loss": 0.7272, + "step": 13731 + }, + { + "epoch": 0.9528832142113663, + "grad_norm": 1.834529161453247, + "learning_rate": 0.00017811153242259397, + "loss": 0.7595, + "step": 13732 + }, + { + "epoch": 0.952952605648463, + "grad_norm": 2.6910853385925293, + "learning_rate": 0.00017810802411008379, + "loss": 0.833, + "step": 13733 + }, + { + "epoch": 0.9530219970855597, + "grad_norm": 2.4883508682250977, + "learning_rate": 0.00017810451555099559, + "loss": 0.8299, + "step": 13734 + }, + { + "epoch": 0.9530913885226563, + "grad_norm": 2.363250732421875, + "learning_rate": 0.00017810100674534054, + "loss": 0.9296, + "step": 13735 + }, + { + "epoch": 0.953160779959753, + "grad_norm": 2.8288869857788086, + "learning_rate": 0.00017809749769312965, + "loss": 0.8892, + "step": 13736 + }, + { + "epoch": 0.9532301713968496, + "grad_norm": 2.1282222270965576, + "learning_rate": 0.00017809398839437407, + "loss": 0.991, + "step": 13737 + }, + { + "epoch": 0.9532995628339463, + "grad_norm": 1.819104552268982, + "learning_rate": 0.00017809047884908473, + "loss": 0.7741, + "step": 13738 + }, + { + "epoch": 0.953368954271043, + "grad_norm": 2.61358642578125, + "learning_rate": 0.0001780869690572729, + "loss": 0.8454, + "step": 13739 + }, + { + "epoch": 0.9534383457081396, + "grad_norm": 2.2491860389709473, + "learning_rate": 0.00017808345901894953, + "loss": 0.6357, + "step": 13740 + }, + { + "epoch": 0.9535077371452363, + "grad_norm": 2.4307618141174316, + "learning_rate": 0.00017807994873412575, + "loss": 0.9213, + "step": 13741 + }, + { + "epoch": 0.9535771285823329, + "grad_norm": 1.9118553400039673, + "learning_rate": 0.0001780764382028126, + "loss": 0.6484, + "step": 13742 + }, + { + "epoch": 0.9536465200194296, + "grad_norm": 2.2488179206848145, + "learning_rate": 0.00017807292742502123, + "loss": 0.8082, + "step": 13743 + }, + { + "epoch": 0.9537159114565262, + "grad_norm": 1.9076985120773315, + "learning_rate": 0.00017806941640076269, + "loss": 0.6322, + "step": 13744 + }, + { + "epoch": 0.953785302893623, + "grad_norm": 2.665055751800537, + "learning_rate": 0.00017806590513004804, + "loss": 0.9435, + "step": 13745 + }, + { + "epoch": 0.9538546943307196, + "grad_norm": 2.4160666465759277, + "learning_rate": 0.00017806239361288842, + "loss": 0.7552, + "step": 13746 + }, + { + "epoch": 0.9539240857678163, + "grad_norm": 2.252377986907959, + "learning_rate": 0.00017805888184929485, + "loss": 0.6766, + "step": 13747 + }, + { + "epoch": 0.9539934772049129, + "grad_norm": 2.3890960216522217, + "learning_rate": 0.0001780553698392784, + "loss": 0.664, + "step": 13748 + }, + { + "epoch": 0.9540628686420096, + "grad_norm": 2.5446670055389404, + "learning_rate": 0.00017805185758285028, + "loss": 0.8791, + "step": 13749 + }, + { + "epoch": 0.9541322600791062, + "grad_norm": 2.4161744117736816, + "learning_rate": 0.00017804834508002146, + "loss": 0.6449, + "step": 13750 + }, + { + "epoch": 0.954201651516203, + "grad_norm": 2.1246371269226074, + "learning_rate": 0.00017804483233080308, + "loss": 0.7505, + "step": 13751 + }, + { + "epoch": 0.9542710429532996, + "grad_norm": 2.993741750717163, + "learning_rate": 0.0001780413193352062, + "loss": 0.8669, + "step": 13752 + }, + { + "epoch": 0.9543404343903962, + "grad_norm": 3.295539379119873, + "learning_rate": 0.000178037806093242, + "loss": 0.9062, + "step": 13753 + }, + { + "epoch": 0.9544098258274929, + "grad_norm": 1.81743586063385, + "learning_rate": 0.00017803429260492145, + "loss": 0.685, + "step": 13754 + }, + { + "epoch": 0.9544792172645895, + "grad_norm": 2.5182621479034424, + "learning_rate": 0.00017803077887025567, + "loss": 0.9237, + "step": 13755 + }, + { + "epoch": 0.9545486087016862, + "grad_norm": 2.5306050777435303, + "learning_rate": 0.0001780272648892558, + "loss": 0.9271, + "step": 13756 + }, + { + "epoch": 0.9546180001387828, + "grad_norm": 1.671043872833252, + "learning_rate": 0.00017802375066193292, + "loss": 0.6054, + "step": 13757 + }, + { + "epoch": 0.9546873915758796, + "grad_norm": 3.2359912395477295, + "learning_rate": 0.0001780202361882981, + "loss": 0.7022, + "step": 13758 + }, + { + "epoch": 0.9547567830129762, + "grad_norm": 2.3644838333129883, + "learning_rate": 0.00017801672146836243, + "loss": 0.7402, + "step": 13759 + }, + { + "epoch": 0.9548261744500729, + "grad_norm": 2.117917537689209, + "learning_rate": 0.000178013206502137, + "loss": 0.7887, + "step": 13760 + }, + { + "epoch": 0.9548955658871695, + "grad_norm": 3.1725921630859375, + "learning_rate": 0.000178009691289633, + "loss": 1.0945, + "step": 13761 + }, + { + "epoch": 0.9549649573242662, + "grad_norm": 1.9444241523742676, + "learning_rate": 0.0001780061758308614, + "loss": 0.6923, + "step": 13762 + }, + { + "epoch": 0.9550343487613628, + "grad_norm": 2.1395974159240723, + "learning_rate": 0.00017800266012583338, + "loss": 0.633, + "step": 13763 + }, + { + "epoch": 0.9551037401984596, + "grad_norm": 2.7218775749206543, + "learning_rate": 0.00017799914417456001, + "loss": 0.7171, + "step": 13764 + }, + { + "epoch": 0.9551731316355562, + "grad_norm": 1.978512167930603, + "learning_rate": 0.00017799562797705238, + "loss": 0.5925, + "step": 13765 + }, + { + "epoch": 0.9552425230726528, + "grad_norm": 1.8717355728149414, + "learning_rate": 0.0001779921115333216, + "loss": 0.5432, + "step": 13766 + }, + { + "epoch": 0.9553119145097495, + "grad_norm": 1.6197236776351929, + "learning_rate": 0.00017798859484337878, + "loss": 0.567, + "step": 13767 + }, + { + "epoch": 0.9553813059468461, + "grad_norm": 2.2068934440612793, + "learning_rate": 0.00017798507790723502, + "loss": 0.679, + "step": 13768 + }, + { + "epoch": 0.9554506973839428, + "grad_norm": 2.617687225341797, + "learning_rate": 0.0001779815607249014, + "loss": 0.902, + "step": 13769 + }, + { + "epoch": 0.9555200888210394, + "grad_norm": 2.3378257751464844, + "learning_rate": 0.00017797804329638907, + "loss": 0.8225, + "step": 13770 + }, + { + "epoch": 0.9555894802581362, + "grad_norm": 2.272686243057251, + "learning_rate": 0.0001779745256217091, + "loss": 0.8718, + "step": 13771 + }, + { + "epoch": 0.9556588716952328, + "grad_norm": 2.571803331375122, + "learning_rate": 0.00017797100770087257, + "loss": 0.857, + "step": 13772 + }, + { + "epoch": 0.9557282631323295, + "grad_norm": 2.757683277130127, + "learning_rate": 0.00017796748953389062, + "loss": 0.8621, + "step": 13773 + }, + { + "epoch": 0.9557976545694261, + "grad_norm": 2.234109401702881, + "learning_rate": 0.00017796397112077438, + "loss": 0.8851, + "step": 13774 + }, + { + "epoch": 0.9558670460065228, + "grad_norm": 2.0072133541107178, + "learning_rate": 0.0001779604524615349, + "loss": 0.7261, + "step": 13775 + }, + { + "epoch": 0.9559364374436194, + "grad_norm": 2.6444528102874756, + "learning_rate": 0.00017795693355618331, + "loss": 0.7589, + "step": 13776 + }, + { + "epoch": 0.9560058288807162, + "grad_norm": 1.9439724683761597, + "learning_rate": 0.00017795341440473075, + "loss": 0.5753, + "step": 13777 + }, + { + "epoch": 0.9560752203178128, + "grad_norm": 2.4301257133483887, + "learning_rate": 0.00017794989500718832, + "loss": 1.1319, + "step": 13778 + }, + { + "epoch": 0.9561446117549094, + "grad_norm": 1.7598968744277954, + "learning_rate": 0.00017794637536356706, + "loss": 0.6755, + "step": 13779 + }, + { + "epoch": 0.9562140031920061, + "grad_norm": 2.479187250137329, + "learning_rate": 0.00017794285547387818, + "loss": 0.9743, + "step": 13780 + }, + { + "epoch": 0.9562833946291027, + "grad_norm": 1.7143656015396118, + "learning_rate": 0.00017793933533813273, + "loss": 0.6749, + "step": 13781 + }, + { + "epoch": 0.9563527860661994, + "grad_norm": 2.1684975624084473, + "learning_rate": 0.00017793581495634185, + "loss": 0.7568, + "step": 13782 + }, + { + "epoch": 0.9564221775032961, + "grad_norm": 2.664957284927368, + "learning_rate": 0.00017793229432851667, + "loss": 0.8567, + "step": 13783 + }, + { + "epoch": 0.9564915689403928, + "grad_norm": 2.4396252632141113, + "learning_rate": 0.00017792877345466823, + "loss": 0.8453, + "step": 13784 + }, + { + "epoch": 0.9565609603774894, + "grad_norm": 2.3683927059173584, + "learning_rate": 0.0001779252523348077, + "loss": 0.9709, + "step": 13785 + }, + { + "epoch": 0.9566303518145861, + "grad_norm": 4.008657932281494, + "learning_rate": 0.00017792173096894623, + "loss": 0.7188, + "step": 13786 + }, + { + "epoch": 0.9566997432516827, + "grad_norm": 1.9150549173355103, + "learning_rate": 0.00017791820935709484, + "loss": 0.7793, + "step": 13787 + }, + { + "epoch": 0.9567691346887794, + "grad_norm": 1.880492925643921, + "learning_rate": 0.00017791468749926476, + "loss": 0.683, + "step": 13788 + }, + { + "epoch": 0.956838526125876, + "grad_norm": 3.801593780517578, + "learning_rate": 0.00017791116539546701, + "loss": 0.9072, + "step": 13789 + }, + { + "epoch": 0.9569079175629728, + "grad_norm": 2.1800689697265625, + "learning_rate": 0.00017790764304571278, + "loss": 0.9982, + "step": 13790 + }, + { + "epoch": 0.9569773090000694, + "grad_norm": 1.834199070930481, + "learning_rate": 0.00017790412045001313, + "loss": 0.6765, + "step": 13791 + }, + { + "epoch": 0.9570467004371661, + "grad_norm": 2.1641852855682373, + "learning_rate": 0.00017790059760837922, + "loss": 0.7558, + "step": 13792 + }, + { + "epoch": 0.9571160918742627, + "grad_norm": 1.948663353919983, + "learning_rate": 0.00017789707452082215, + "loss": 0.8549, + "step": 13793 + }, + { + "epoch": 0.9571854833113593, + "grad_norm": 2.251657485961914, + "learning_rate": 0.0001778935511873531, + "loss": 0.8517, + "step": 13794 + }, + { + "epoch": 0.957254874748456, + "grad_norm": 2.1904683113098145, + "learning_rate": 0.0001778900276079831, + "loss": 0.6864, + "step": 13795 + }, + { + "epoch": 0.9573242661855527, + "grad_norm": 2.073613166809082, + "learning_rate": 0.00017788650378272337, + "loss": 0.6273, + "step": 13796 + }, + { + "epoch": 0.9573936576226494, + "grad_norm": 2.1786322593688965, + "learning_rate": 0.00017788297971158495, + "loss": 0.9816, + "step": 13797 + }, + { + "epoch": 0.957463049059746, + "grad_norm": 2.1558430194854736, + "learning_rate": 0.00017787945539457896, + "loss": 0.9252, + "step": 13798 + }, + { + "epoch": 0.9575324404968427, + "grad_norm": 2.1475369930267334, + "learning_rate": 0.00017787593083171664, + "loss": 0.8721, + "step": 13799 + }, + { + "epoch": 0.9576018319339393, + "grad_norm": 2.1305880546569824, + "learning_rate": 0.00017787240602300898, + "loss": 0.7888, + "step": 13800 + }, + { + "epoch": 0.957671223371036, + "grad_norm": 2.958116292953491, + "learning_rate": 0.0001778688809684672, + "loss": 1.038, + "step": 13801 + }, + { + "epoch": 0.9577406148081327, + "grad_norm": 1.8711429834365845, + "learning_rate": 0.0001778653556681024, + "loss": 0.7987, + "step": 13802 + }, + { + "epoch": 0.9578100062452294, + "grad_norm": 1.854575276374817, + "learning_rate": 0.0001778618301219257, + "loss": 0.6192, + "step": 13803 + }, + { + "epoch": 0.957879397682326, + "grad_norm": 2.073975086212158, + "learning_rate": 0.00017785830432994824, + "loss": 0.9242, + "step": 13804 + }, + { + "epoch": 0.9579487891194227, + "grad_norm": 2.049899101257324, + "learning_rate": 0.00017785477829218113, + "loss": 0.9399, + "step": 13805 + }, + { + "epoch": 0.9580181805565193, + "grad_norm": 1.759474754333496, + "learning_rate": 0.00017785125200863556, + "loss": 0.5176, + "step": 13806 + }, + { + "epoch": 0.9580875719936159, + "grad_norm": 1.9230058193206787, + "learning_rate": 0.00017784772547932255, + "loss": 0.8335, + "step": 13807 + }, + { + "epoch": 0.9581569634307127, + "grad_norm": 1.6917037963867188, + "learning_rate": 0.00017784419870425336, + "loss": 0.6383, + "step": 13808 + }, + { + "epoch": 0.9582263548678093, + "grad_norm": 1.5570564270019531, + "learning_rate": 0.00017784067168343903, + "loss": 0.8299, + "step": 13809 + }, + { + "epoch": 0.958295746304906, + "grad_norm": 2.0093913078308105, + "learning_rate": 0.00017783714441689074, + "loss": 0.7156, + "step": 13810 + }, + { + "epoch": 0.9583651377420026, + "grad_norm": 2.5304603576660156, + "learning_rate": 0.00017783361690461964, + "loss": 0.8722, + "step": 13811 + }, + { + "epoch": 0.9584345291790993, + "grad_norm": 2.921295642852783, + "learning_rate": 0.00017783008914663682, + "loss": 0.7329, + "step": 13812 + }, + { + "epoch": 0.9585039206161959, + "grad_norm": 2.2936618328094482, + "learning_rate": 0.00017782656114295343, + "loss": 0.6025, + "step": 13813 + }, + { + "epoch": 0.9585733120532927, + "grad_norm": 2.8783984184265137, + "learning_rate": 0.00017782303289358065, + "loss": 0.9181, + "step": 13814 + }, + { + "epoch": 0.9586427034903893, + "grad_norm": 1.7809576988220215, + "learning_rate": 0.00017781950439852958, + "loss": 0.6934, + "step": 13815 + }, + { + "epoch": 0.958712094927486, + "grad_norm": 2.6175246238708496, + "learning_rate": 0.00017781597565781133, + "loss": 0.7678, + "step": 13816 + }, + { + "epoch": 0.9587814863645826, + "grad_norm": 1.9410399198532104, + "learning_rate": 0.00017781244667143707, + "loss": 0.6738, + "step": 13817 + }, + { + "epoch": 0.9588508778016793, + "grad_norm": 2.066006898880005, + "learning_rate": 0.00017780891743941798, + "loss": 0.8898, + "step": 13818 + }, + { + "epoch": 0.9589202692387759, + "grad_norm": 2.103996992111206, + "learning_rate": 0.00017780538796176515, + "loss": 0.6646, + "step": 13819 + }, + { + "epoch": 0.9589896606758725, + "grad_norm": 3.091498851776123, + "learning_rate": 0.00017780185823848974, + "loss": 1.1459, + "step": 13820 + }, + { + "epoch": 0.9590590521129693, + "grad_norm": 2.316906213760376, + "learning_rate": 0.0001777983282696029, + "loss": 0.8695, + "step": 13821 + }, + { + "epoch": 0.9591284435500659, + "grad_norm": 1.7597260475158691, + "learning_rate": 0.00017779479805511575, + "loss": 0.8587, + "step": 13822 + }, + { + "epoch": 0.9591978349871626, + "grad_norm": 2.6247215270996094, + "learning_rate": 0.00017779126759503945, + "loss": 0.7556, + "step": 13823 + }, + { + "epoch": 0.9592672264242592, + "grad_norm": 2.5523297786712646, + "learning_rate": 0.00017778773688938512, + "loss": 0.6134, + "step": 13824 + }, + { + "epoch": 0.9593366178613559, + "grad_norm": 1.9260505437850952, + "learning_rate": 0.000177784205938164, + "loss": 0.5808, + "step": 13825 + }, + { + "epoch": 0.9594060092984525, + "grad_norm": 2.0367789268493652, + "learning_rate": 0.00017778067474138708, + "loss": 0.5677, + "step": 13826 + }, + { + "epoch": 0.9594754007355493, + "grad_norm": 1.9256168603897095, + "learning_rate": 0.00017777714329906562, + "loss": 0.7665, + "step": 13827 + }, + { + "epoch": 0.9595447921726459, + "grad_norm": 2.063812017440796, + "learning_rate": 0.00017777361161121078, + "loss": 0.7732, + "step": 13828 + }, + { + "epoch": 0.9596141836097426, + "grad_norm": 2.22515869140625, + "learning_rate": 0.00017777007967783367, + "loss": 1.0519, + "step": 13829 + }, + { + "epoch": 0.9596835750468392, + "grad_norm": 2.734450578689575, + "learning_rate": 0.0001777665474989454, + "loss": 0.9269, + "step": 13830 + }, + { + "epoch": 0.9597529664839359, + "grad_norm": 2.394179582595825, + "learning_rate": 0.00017776301507455716, + "loss": 0.8644, + "step": 13831 + }, + { + "epoch": 0.9598223579210325, + "grad_norm": 3.183865547180176, + "learning_rate": 0.00017775948240468013, + "loss": 0.8745, + "step": 13832 + }, + { + "epoch": 0.9598917493581292, + "grad_norm": 2.029320001602173, + "learning_rate": 0.00017775594948932547, + "loss": 0.5086, + "step": 13833 + }, + { + "epoch": 0.9599611407952259, + "grad_norm": 2.0336053371429443, + "learning_rate": 0.00017775241632850424, + "loss": 0.6943, + "step": 13834 + }, + { + "epoch": 0.9600305322323225, + "grad_norm": 2.019728660583496, + "learning_rate": 0.00017774888292222766, + "loss": 0.8713, + "step": 13835 + }, + { + "epoch": 0.9600999236694192, + "grad_norm": 2.802137851715088, + "learning_rate": 0.00017774534927050687, + "loss": 1.1958, + "step": 13836 + }, + { + "epoch": 0.9601693151065158, + "grad_norm": 2.373891592025757, + "learning_rate": 0.00017774181537335307, + "loss": 0.7674, + "step": 13837 + }, + { + "epoch": 0.9602387065436125, + "grad_norm": 2.126490831375122, + "learning_rate": 0.00017773828123077737, + "loss": 0.7683, + "step": 13838 + }, + { + "epoch": 0.9603080979807092, + "grad_norm": 3.094320058822632, + "learning_rate": 0.00017773474684279093, + "loss": 0.9204, + "step": 13839 + }, + { + "epoch": 0.9603774894178059, + "grad_norm": 2.1420295238494873, + "learning_rate": 0.0001777312122094049, + "loss": 0.8295, + "step": 13840 + }, + { + "epoch": 0.9604468808549025, + "grad_norm": 2.3616580963134766, + "learning_rate": 0.00017772767733063044, + "loss": 0.9701, + "step": 13841 + }, + { + "epoch": 0.9605162722919992, + "grad_norm": 1.8300036191940308, + "learning_rate": 0.00017772414220647878, + "loss": 0.8342, + "step": 13842 + }, + { + "epoch": 0.9605856637290958, + "grad_norm": 1.6069072484970093, + "learning_rate": 0.00017772060683696097, + "loss": 0.6415, + "step": 13843 + }, + { + "epoch": 0.9606550551661925, + "grad_norm": 1.8260122537612915, + "learning_rate": 0.00017771707122208824, + "loss": 0.6358, + "step": 13844 + }, + { + "epoch": 0.9607244466032892, + "grad_norm": 2.355164051055908, + "learning_rate": 0.0001777135353618717, + "loss": 0.6449, + "step": 13845 + }, + { + "epoch": 0.9607938380403858, + "grad_norm": 1.9099260568618774, + "learning_rate": 0.00017770999925632256, + "loss": 0.6431, + "step": 13846 + }, + { + "epoch": 0.9608632294774825, + "grad_norm": 1.9034183025360107, + "learning_rate": 0.00017770646290545198, + "loss": 0.7164, + "step": 13847 + }, + { + "epoch": 0.9609326209145791, + "grad_norm": 2.054810047149658, + "learning_rate": 0.00017770292630927114, + "loss": 0.8694, + "step": 13848 + }, + { + "epoch": 0.9610020123516758, + "grad_norm": 1.7773102521896362, + "learning_rate": 0.00017769938946779115, + "loss": 0.7398, + "step": 13849 + }, + { + "epoch": 0.9610714037887724, + "grad_norm": 2.7090766429901123, + "learning_rate": 0.0001776958523810232, + "loss": 0.8794, + "step": 13850 + }, + { + "epoch": 0.9611407952258691, + "grad_norm": 1.9394197463989258, + "learning_rate": 0.0001776923150489785, + "loss": 0.6853, + "step": 13851 + }, + { + "epoch": 0.9612101866629658, + "grad_norm": 2.228759527206421, + "learning_rate": 0.0001776887774716681, + "loss": 0.6561, + "step": 13852 + }, + { + "epoch": 0.9612795781000625, + "grad_norm": 2.9540512561798096, + "learning_rate": 0.0001776852396491033, + "loss": 0.9523, + "step": 13853 + }, + { + "epoch": 0.9613489695371591, + "grad_norm": 2.181926965713501, + "learning_rate": 0.00017768170158129516, + "loss": 0.9174, + "step": 13854 + }, + { + "epoch": 0.9614183609742558, + "grad_norm": 2.072885751724243, + "learning_rate": 0.00017767816326825492, + "loss": 1.0531, + "step": 13855 + }, + { + "epoch": 0.9614877524113524, + "grad_norm": 2.139305591583252, + "learning_rate": 0.00017767462470999376, + "loss": 0.7947, + "step": 13856 + }, + { + "epoch": 0.9615571438484491, + "grad_norm": 2.577965259552002, + "learning_rate": 0.0001776710859065228, + "loss": 0.714, + "step": 13857 + }, + { + "epoch": 0.9616265352855458, + "grad_norm": 2.400286912918091, + "learning_rate": 0.00017766754685785322, + "loss": 0.9931, + "step": 13858 + }, + { + "epoch": 0.9616959267226425, + "grad_norm": 2.617741107940674, + "learning_rate": 0.00017766400756399624, + "loss": 0.7519, + "step": 13859 + }, + { + "epoch": 0.9617653181597391, + "grad_norm": 2.571291923522949, + "learning_rate": 0.000177660468024963, + "loss": 0.6544, + "step": 13860 + }, + { + "epoch": 0.9618347095968357, + "grad_norm": 1.9783697128295898, + "learning_rate": 0.00017765692824076463, + "loss": 0.6686, + "step": 13861 + }, + { + "epoch": 0.9619041010339324, + "grad_norm": 2.8714377880096436, + "learning_rate": 0.00017765338821141237, + "loss": 1.2144, + "step": 13862 + }, + { + "epoch": 0.961973492471029, + "grad_norm": 2.2982583045959473, + "learning_rate": 0.0001776498479369174, + "loss": 0.81, + "step": 13863 + }, + { + "epoch": 0.9620428839081258, + "grad_norm": 2.4108102321624756, + "learning_rate": 0.00017764630741729081, + "loss": 0.822, + "step": 13864 + }, + { + "epoch": 0.9621122753452224, + "grad_norm": 1.7832194566726685, + "learning_rate": 0.0001776427666525439, + "loss": 0.8218, + "step": 13865 + }, + { + "epoch": 0.9621816667823191, + "grad_norm": 2.4932732582092285, + "learning_rate": 0.00017763922564268775, + "loss": 0.6762, + "step": 13866 + }, + { + "epoch": 0.9622510582194157, + "grad_norm": 1.8309324979782104, + "learning_rate": 0.00017763568438773354, + "loss": 0.505, + "step": 13867 + }, + { + "epoch": 0.9623204496565124, + "grad_norm": 2.0470094680786133, + "learning_rate": 0.00017763214288769253, + "loss": 0.6726, + "step": 13868 + }, + { + "epoch": 0.962389841093609, + "grad_norm": 2.442870855331421, + "learning_rate": 0.00017762860114257584, + "loss": 0.8099, + "step": 13869 + }, + { + "epoch": 0.9624592325307058, + "grad_norm": 1.8932024240493774, + "learning_rate": 0.00017762505915239467, + "loss": 0.8382, + "step": 13870 + }, + { + "epoch": 0.9625286239678024, + "grad_norm": 1.9119024276733398, + "learning_rate": 0.00017762151691716018, + "loss": 0.6729, + "step": 13871 + }, + { + "epoch": 0.9625980154048991, + "grad_norm": 2.695648193359375, + "learning_rate": 0.00017761797443688356, + "loss": 0.6726, + "step": 13872 + }, + { + "epoch": 0.9626674068419957, + "grad_norm": 1.5965843200683594, + "learning_rate": 0.00017761443171157603, + "loss": 0.4929, + "step": 13873 + }, + { + "epoch": 0.9627367982790923, + "grad_norm": 2.2873473167419434, + "learning_rate": 0.00017761088874124873, + "loss": 0.8418, + "step": 13874 + }, + { + "epoch": 0.962806189716189, + "grad_norm": 1.967934012413025, + "learning_rate": 0.00017760734552591285, + "loss": 0.6601, + "step": 13875 + }, + { + "epoch": 0.9628755811532856, + "grad_norm": 2.589203357696533, + "learning_rate": 0.00017760380206557962, + "loss": 1.0957, + "step": 13876 + }, + { + "epoch": 0.9629449725903824, + "grad_norm": 2.873858690261841, + "learning_rate": 0.00017760025836026014, + "loss": 0.844, + "step": 13877 + }, + { + "epoch": 0.963014364027479, + "grad_norm": 2.515645980834961, + "learning_rate": 0.00017759671440996565, + "loss": 0.7977, + "step": 13878 + }, + { + "epoch": 0.9630837554645757, + "grad_norm": 1.999129295349121, + "learning_rate": 0.00017759317021470738, + "loss": 0.6584, + "step": 13879 + }, + { + "epoch": 0.9631531469016723, + "grad_norm": 1.5718034505844116, + "learning_rate": 0.00017758962577449645, + "loss": 0.7007, + "step": 13880 + }, + { + "epoch": 0.963222538338769, + "grad_norm": 2.394383430480957, + "learning_rate": 0.00017758608108934406, + "loss": 1.0208, + "step": 13881 + }, + { + "epoch": 0.9632919297758656, + "grad_norm": 1.8548002243041992, + "learning_rate": 0.00017758253615926143, + "loss": 0.8022, + "step": 13882 + }, + { + "epoch": 0.9633613212129624, + "grad_norm": 2.0566160678863525, + "learning_rate": 0.00017757899098425973, + "loss": 0.6838, + "step": 13883 + }, + { + "epoch": 0.963430712650059, + "grad_norm": 2.382632255554199, + "learning_rate": 0.00017757544556435014, + "loss": 0.8544, + "step": 13884 + }, + { + "epoch": 0.9635001040871557, + "grad_norm": 2.0298140048980713, + "learning_rate": 0.0001775718998995439, + "loss": 0.554, + "step": 13885 + }, + { + "epoch": 0.9635694955242523, + "grad_norm": 2.3416662216186523, + "learning_rate": 0.00017756835398985215, + "loss": 0.9622, + "step": 13886 + }, + { + "epoch": 0.9636388869613489, + "grad_norm": 2.4556379318237305, + "learning_rate": 0.00017756480783528612, + "loss": 0.7445, + "step": 13887 + }, + { + "epoch": 0.9637082783984456, + "grad_norm": 2.4543097019195557, + "learning_rate": 0.00017756126143585697, + "loss": 1.0263, + "step": 13888 + }, + { + "epoch": 0.9637776698355422, + "grad_norm": 4.422319412231445, + "learning_rate": 0.00017755771479157595, + "loss": 0.7473, + "step": 13889 + }, + { + "epoch": 0.963847061272639, + "grad_norm": 3.3480849266052246, + "learning_rate": 0.00017755416790245422, + "loss": 0.858, + "step": 13890 + }, + { + "epoch": 0.9639164527097356, + "grad_norm": 2.6788058280944824, + "learning_rate": 0.00017755062076850293, + "loss": 1.0712, + "step": 13891 + }, + { + "epoch": 0.9639858441468323, + "grad_norm": 2.053576946258545, + "learning_rate": 0.00017754707338973335, + "loss": 0.7716, + "step": 13892 + }, + { + "epoch": 0.9640552355839289, + "grad_norm": 2.046635866165161, + "learning_rate": 0.00017754352576615666, + "loss": 0.7587, + "step": 13893 + }, + { + "epoch": 0.9641246270210256, + "grad_norm": 1.5479693412780762, + "learning_rate": 0.0001775399778977841, + "loss": 0.598, + "step": 13894 + }, + { + "epoch": 0.9641940184581222, + "grad_norm": 2.0338330268859863, + "learning_rate": 0.00017753642978462678, + "loss": 0.5786, + "step": 13895 + }, + { + "epoch": 0.964263409895219, + "grad_norm": 2.2183215618133545, + "learning_rate": 0.00017753288142669597, + "loss": 1.0945, + "step": 13896 + }, + { + "epoch": 0.9643328013323156, + "grad_norm": 2.094043493270874, + "learning_rate": 0.00017752933282400281, + "loss": 0.76, + "step": 13897 + }, + { + "epoch": 0.9644021927694123, + "grad_norm": 2.278254747390747, + "learning_rate": 0.0001775257839765586, + "loss": 1.0086, + "step": 13898 + }, + { + "epoch": 0.9644715842065089, + "grad_norm": 2.9109115600585938, + "learning_rate": 0.00017752223488437443, + "loss": 1.0834, + "step": 13899 + }, + { + "epoch": 0.9645409756436055, + "grad_norm": 2.5911824703216553, + "learning_rate": 0.0001775186855474616, + "loss": 1.0969, + "step": 13900 + }, + { + "epoch": 0.9646103670807022, + "grad_norm": 1.7868767976760864, + "learning_rate": 0.00017751513596583123, + "loss": 0.6216, + "step": 13901 + }, + { + "epoch": 0.9646797585177989, + "grad_norm": 2.1927573680877686, + "learning_rate": 0.00017751158613949462, + "loss": 0.8248, + "step": 13902 + }, + { + "epoch": 0.9647491499548956, + "grad_norm": 1.602418065071106, + "learning_rate": 0.00017750803606846288, + "loss": 0.567, + "step": 13903 + }, + { + "epoch": 0.9648185413919922, + "grad_norm": 2.2403461933135986, + "learning_rate": 0.00017750448575274728, + "loss": 0.7583, + "step": 13904 + }, + { + "epoch": 0.9648879328290889, + "grad_norm": 1.915834665298462, + "learning_rate": 0.00017750093519235902, + "loss": 0.7008, + "step": 13905 + }, + { + "epoch": 0.9649573242661855, + "grad_norm": 2.4902584552764893, + "learning_rate": 0.0001774973843873093, + "loss": 0.8501, + "step": 13906 + }, + { + "epoch": 0.9650267157032822, + "grad_norm": 2.0290958881378174, + "learning_rate": 0.0001774938333376093, + "loss": 0.7419, + "step": 13907 + }, + { + "epoch": 0.9650961071403789, + "grad_norm": 1.5403679609298706, + "learning_rate": 0.00017749028204327029, + "loss": 0.6208, + "step": 13908 + }, + { + "epoch": 0.9651654985774756, + "grad_norm": 1.9175527095794678, + "learning_rate": 0.00017748673050430342, + "loss": 0.8064, + "step": 13909 + }, + { + "epoch": 0.9652348900145722, + "grad_norm": 2.234694480895996, + "learning_rate": 0.00017748317872071995, + "loss": 0.9876, + "step": 13910 + }, + { + "epoch": 0.9653042814516689, + "grad_norm": 2.4114813804626465, + "learning_rate": 0.0001774796266925311, + "loss": 0.7945, + "step": 13911 + }, + { + "epoch": 0.9653736728887655, + "grad_norm": 2.1055798530578613, + "learning_rate": 0.00017747607441974804, + "loss": 0.7338, + "step": 13912 + }, + { + "epoch": 0.9654430643258622, + "grad_norm": 1.356134295463562, + "learning_rate": 0.00017747252190238196, + "loss": 0.4943, + "step": 13913 + }, + { + "epoch": 0.9655124557629589, + "grad_norm": 1.9492707252502441, + "learning_rate": 0.00017746896914044418, + "loss": 0.654, + "step": 13914 + }, + { + "epoch": 0.9655818472000555, + "grad_norm": 2.5000381469726562, + "learning_rate": 0.0001774654161339458, + "loss": 0.8232, + "step": 13915 + }, + { + "epoch": 0.9656512386371522, + "grad_norm": 1.706600546836853, + "learning_rate": 0.00017746186288289812, + "loss": 0.6224, + "step": 13916 + }, + { + "epoch": 0.9657206300742488, + "grad_norm": 1.6168763637542725, + "learning_rate": 0.0001774583093873123, + "loss": 0.8793, + "step": 13917 + }, + { + "epoch": 0.9657900215113455, + "grad_norm": 1.585378646850586, + "learning_rate": 0.0001774547556471996, + "loss": 0.5976, + "step": 13918 + }, + { + "epoch": 0.9658594129484421, + "grad_norm": 1.9947680234909058, + "learning_rate": 0.00017745120166257126, + "loss": 0.6656, + "step": 13919 + }, + { + "epoch": 0.9659288043855389, + "grad_norm": 2.148991346359253, + "learning_rate": 0.00017744764743343841, + "loss": 0.827, + "step": 13920 + }, + { + "epoch": 0.9659981958226355, + "grad_norm": 1.9698543548583984, + "learning_rate": 0.00017744409295981234, + "loss": 0.621, + "step": 13921 + }, + { + "epoch": 0.9660675872597322, + "grad_norm": 1.982476830482483, + "learning_rate": 0.00017744053824170428, + "loss": 0.7363, + "step": 13922 + }, + { + "epoch": 0.9661369786968288, + "grad_norm": 2.5449585914611816, + "learning_rate": 0.00017743698327912539, + "loss": 0.7581, + "step": 13923 + }, + { + "epoch": 0.9662063701339255, + "grad_norm": 2.8451120853424072, + "learning_rate": 0.00017743342807208693, + "loss": 0.5325, + "step": 13924 + }, + { + "epoch": 0.9662757615710221, + "grad_norm": 2.0436525344848633, + "learning_rate": 0.00017742987262060013, + "loss": 0.5988, + "step": 13925 + }, + { + "epoch": 0.9663451530081189, + "grad_norm": 2.1545629501342773, + "learning_rate": 0.0001774263169246762, + "loss": 0.7813, + "step": 13926 + }, + { + "epoch": 0.9664145444452155, + "grad_norm": 2.3589468002319336, + "learning_rate": 0.00017742276098432639, + "loss": 0.8813, + "step": 13927 + }, + { + "epoch": 0.9664839358823121, + "grad_norm": 3.366295337677002, + "learning_rate": 0.0001774192047995619, + "loss": 0.6868, + "step": 13928 + }, + { + "epoch": 0.9665533273194088, + "grad_norm": 2.0563771724700928, + "learning_rate": 0.00017741564837039396, + "loss": 0.8438, + "step": 13929 + }, + { + "epoch": 0.9666227187565054, + "grad_norm": 1.7952114343643188, + "learning_rate": 0.00017741209169683378, + "loss": 0.6288, + "step": 13930 + }, + { + "epoch": 0.9666921101936021, + "grad_norm": 2.48108172416687, + "learning_rate": 0.00017740853477889263, + "loss": 0.9586, + "step": 13931 + }, + { + "epoch": 0.9667615016306987, + "grad_norm": 1.7291204929351807, + "learning_rate": 0.00017740497761658172, + "loss": 0.582, + "step": 13932 + }, + { + "epoch": 0.9668308930677955, + "grad_norm": 2.246364116668701, + "learning_rate": 0.00017740142020991227, + "loss": 0.7387, + "step": 13933 + }, + { + "epoch": 0.9669002845048921, + "grad_norm": 2.308725118637085, + "learning_rate": 0.0001773978625588955, + "loss": 0.7394, + "step": 13934 + }, + { + "epoch": 0.9669696759419888, + "grad_norm": 1.908685564994812, + "learning_rate": 0.00017739430466354267, + "loss": 0.6785, + "step": 13935 + }, + { + "epoch": 0.9670390673790854, + "grad_norm": 3.277946710586548, + "learning_rate": 0.000177390746523865, + "loss": 0.9563, + "step": 13936 + }, + { + "epoch": 0.9671084588161821, + "grad_norm": 2.0884790420532227, + "learning_rate": 0.00017738718813987373, + "loss": 0.7816, + "step": 13937 + }, + { + "epoch": 0.9671778502532787, + "grad_norm": 2.238426446914673, + "learning_rate": 0.00017738362951158008, + "loss": 0.7987, + "step": 13938 + }, + { + "epoch": 0.9672472416903755, + "grad_norm": 2.175889730453491, + "learning_rate": 0.0001773800706389953, + "loss": 0.7305, + "step": 13939 + }, + { + "epoch": 0.9673166331274721, + "grad_norm": 2.2371768951416016, + "learning_rate": 0.0001773765115221306, + "loss": 0.9274, + "step": 13940 + }, + { + "epoch": 0.9673860245645687, + "grad_norm": 2.1679625511169434, + "learning_rate": 0.0001773729521609972, + "loss": 0.5575, + "step": 13941 + }, + { + "epoch": 0.9674554160016654, + "grad_norm": 2.6731278896331787, + "learning_rate": 0.0001773693925556064, + "loss": 0.6949, + "step": 13942 + }, + { + "epoch": 0.967524807438762, + "grad_norm": 2.1253392696380615, + "learning_rate": 0.00017736583270596944, + "loss": 0.949, + "step": 13943 + }, + { + "epoch": 0.9675941988758587, + "grad_norm": 2.223609209060669, + "learning_rate": 0.00017736227261209745, + "loss": 0.7087, + "step": 13944 + }, + { + "epoch": 0.9676635903129553, + "grad_norm": 2.130715847015381, + "learning_rate": 0.0001773587122740018, + "loss": 0.6752, + "step": 13945 + }, + { + "epoch": 0.9677329817500521, + "grad_norm": 2.8631348609924316, + "learning_rate": 0.00017735515169169363, + "loss": 1.1057, + "step": 13946 + }, + { + "epoch": 0.9678023731871487, + "grad_norm": 2.4356000423431396, + "learning_rate": 0.00017735159086518424, + "loss": 0.908, + "step": 13947 + }, + { + "epoch": 0.9678717646242454, + "grad_norm": 1.8279770612716675, + "learning_rate": 0.00017734802979448486, + "loss": 0.6242, + "step": 13948 + }, + { + "epoch": 0.967941156061342, + "grad_norm": 3.2274560928344727, + "learning_rate": 0.0001773444684796067, + "loss": 0.8329, + "step": 13949 + }, + { + "epoch": 0.9680105474984387, + "grad_norm": 1.8452839851379395, + "learning_rate": 0.00017734090692056106, + "loss": 0.5748, + "step": 13950 + }, + { + "epoch": 0.9680799389355353, + "grad_norm": 1.5980178117752075, + "learning_rate": 0.00017733734511735913, + "loss": 0.5311, + "step": 13951 + }, + { + "epoch": 0.9681493303726321, + "grad_norm": 2.043856382369995, + "learning_rate": 0.00017733378307001219, + "loss": 0.6353, + "step": 13952 + }, + { + "epoch": 0.9682187218097287, + "grad_norm": 1.778864860534668, + "learning_rate": 0.00017733022077853145, + "loss": 0.6933, + "step": 13953 + }, + { + "epoch": 0.9682881132468253, + "grad_norm": 1.8592146635055542, + "learning_rate": 0.00017732665824292818, + "loss": 0.6573, + "step": 13954 + }, + { + "epoch": 0.968357504683922, + "grad_norm": 2.1872012615203857, + "learning_rate": 0.00017732309546321365, + "loss": 0.9396, + "step": 13955 + }, + { + "epoch": 0.9684268961210186, + "grad_norm": 3.0913383960723877, + "learning_rate": 0.00017731953243939902, + "loss": 0.8843, + "step": 13956 + }, + { + "epoch": 0.9684962875581153, + "grad_norm": 1.8098715543746948, + "learning_rate": 0.00017731596917149563, + "loss": 0.575, + "step": 13957 + }, + { + "epoch": 0.968565678995212, + "grad_norm": 2.3826441764831543, + "learning_rate": 0.0001773124056595147, + "loss": 0.6905, + "step": 13958 + }, + { + "epoch": 0.9686350704323087, + "grad_norm": 1.5272661447525024, + "learning_rate": 0.00017730884190346745, + "loss": 0.6312, + "step": 13959 + }, + { + "epoch": 0.9687044618694053, + "grad_norm": 2.165893316268921, + "learning_rate": 0.00017730527790336517, + "loss": 0.7963, + "step": 13960 + }, + { + "epoch": 0.968773853306502, + "grad_norm": 2.162362575531006, + "learning_rate": 0.00017730171365921914, + "loss": 0.8728, + "step": 13961 + }, + { + "epoch": 0.9688432447435986, + "grad_norm": 2.0777764320373535, + "learning_rate": 0.0001772981491710405, + "loss": 0.7073, + "step": 13962 + }, + { + "epoch": 0.9689126361806953, + "grad_norm": 1.9166622161865234, + "learning_rate": 0.0001772945844388406, + "loss": 0.633, + "step": 13963 + }, + { + "epoch": 0.968982027617792, + "grad_norm": 1.7792853116989136, + "learning_rate": 0.00017729101946263067, + "loss": 0.698, + "step": 13964 + }, + { + "epoch": 0.9690514190548887, + "grad_norm": 2.6910452842712402, + "learning_rate": 0.00017728745424242198, + "loss": 1.0027, + "step": 13965 + }, + { + "epoch": 0.9691208104919853, + "grad_norm": 2.8149123191833496, + "learning_rate": 0.00017728388877822572, + "loss": 0.7977, + "step": 13966 + }, + { + "epoch": 0.9691902019290819, + "grad_norm": 2.2423558235168457, + "learning_rate": 0.00017728032307005322, + "loss": 0.7175, + "step": 13967 + }, + { + "epoch": 0.9692595933661786, + "grad_norm": 2.589578151702881, + "learning_rate": 0.0001772767571179157, + "loss": 0.9761, + "step": 13968 + }, + { + "epoch": 0.9693289848032752, + "grad_norm": 1.8108009099960327, + "learning_rate": 0.0001772731909218244, + "loss": 0.5754, + "step": 13969 + }, + { + "epoch": 0.969398376240372, + "grad_norm": 2.193345308303833, + "learning_rate": 0.00017726962448179063, + "loss": 1.0416, + "step": 13970 + }, + { + "epoch": 0.9694677676774686, + "grad_norm": 4.705147743225098, + "learning_rate": 0.0001772660577978256, + "loss": 0.9254, + "step": 13971 + }, + { + "epoch": 0.9695371591145653, + "grad_norm": 1.569785475730896, + "learning_rate": 0.0001772624908699406, + "loss": 0.7168, + "step": 13972 + }, + { + "epoch": 0.9696065505516619, + "grad_norm": 1.8501231670379639, + "learning_rate": 0.00017725892369814688, + "loss": 0.942, + "step": 13973 + }, + { + "epoch": 0.9696759419887586, + "grad_norm": 2.052847385406494, + "learning_rate": 0.0001772553562824557, + "loss": 0.6274, + "step": 13974 + }, + { + "epoch": 0.9697453334258552, + "grad_norm": 2.271765947341919, + "learning_rate": 0.00017725178862287833, + "loss": 0.7435, + "step": 13975 + }, + { + "epoch": 0.969814724862952, + "grad_norm": 1.6775763034820557, + "learning_rate": 0.00017724822071942602, + "loss": 0.7171, + "step": 13976 + }, + { + "epoch": 0.9698841163000486, + "grad_norm": 1.7416914701461792, + "learning_rate": 0.00017724465257211005, + "loss": 0.7524, + "step": 13977 + }, + { + "epoch": 0.9699535077371453, + "grad_norm": 2.3754680156707764, + "learning_rate": 0.00017724108418094164, + "loss": 0.881, + "step": 13978 + }, + { + "epoch": 0.9700228991742419, + "grad_norm": 1.743143916130066, + "learning_rate": 0.00017723751554593213, + "loss": 0.5979, + "step": 13979 + }, + { + "epoch": 0.9700922906113386, + "grad_norm": 2.2105536460876465, + "learning_rate": 0.00017723394666709273, + "loss": 0.7635, + "step": 13980 + }, + { + "epoch": 0.9701616820484352, + "grad_norm": 2.7174510955810547, + "learning_rate": 0.00017723037754443472, + "loss": 0.9109, + "step": 13981 + }, + { + "epoch": 0.9702310734855318, + "grad_norm": 4.543217182159424, + "learning_rate": 0.00017722680817796937, + "loss": 0.7413, + "step": 13982 + }, + { + "epoch": 0.9703004649226286, + "grad_norm": 1.5470491647720337, + "learning_rate": 0.00017722323856770794, + "loss": 0.6042, + "step": 13983 + }, + { + "epoch": 0.9703698563597252, + "grad_norm": 2.610508680343628, + "learning_rate": 0.0001772196687136617, + "loss": 0.8652, + "step": 13984 + }, + { + "epoch": 0.9704392477968219, + "grad_norm": 2.1569442749023438, + "learning_rate": 0.00017721609861584195, + "loss": 0.699, + "step": 13985 + }, + { + "epoch": 0.9705086392339185, + "grad_norm": 2.3775529861450195, + "learning_rate": 0.00017721252827425992, + "loss": 0.9305, + "step": 13986 + }, + { + "epoch": 0.9705780306710152, + "grad_norm": 2.712164878845215, + "learning_rate": 0.0001772089576889269, + "loss": 0.5923, + "step": 13987 + }, + { + "epoch": 0.9706474221081118, + "grad_norm": 2.242672920227051, + "learning_rate": 0.00017720538685985415, + "loss": 0.8683, + "step": 13988 + }, + { + "epoch": 0.9707168135452086, + "grad_norm": 3.1560449600219727, + "learning_rate": 0.00017720181578705297, + "loss": 0.8251, + "step": 13989 + }, + { + "epoch": 0.9707862049823052, + "grad_norm": 2.1819300651550293, + "learning_rate": 0.0001771982444705346, + "loss": 0.7684, + "step": 13990 + }, + { + "epoch": 0.9708555964194019, + "grad_norm": 2.115562677383423, + "learning_rate": 0.00017719467291031033, + "loss": 0.967, + "step": 13991 + }, + { + "epoch": 0.9709249878564985, + "grad_norm": 2.090606689453125, + "learning_rate": 0.00017719110110639143, + "loss": 0.6636, + "step": 13992 + }, + { + "epoch": 0.9709943792935952, + "grad_norm": 2.228847026824951, + "learning_rate": 0.00017718752905878917, + "loss": 0.7313, + "step": 13993 + }, + { + "epoch": 0.9710637707306918, + "grad_norm": 2.897822856903076, + "learning_rate": 0.00017718395676751488, + "loss": 0.9477, + "step": 13994 + }, + { + "epoch": 0.9711331621677884, + "grad_norm": 2.452773332595825, + "learning_rate": 0.00017718038423257973, + "loss": 0.969, + "step": 13995 + }, + { + "epoch": 0.9712025536048852, + "grad_norm": 1.9642401933670044, + "learning_rate": 0.00017717681145399512, + "loss": 0.6863, + "step": 13996 + }, + { + "epoch": 0.9712719450419818, + "grad_norm": 2.093691825866699, + "learning_rate": 0.00017717323843177222, + "loss": 0.7813, + "step": 13997 + }, + { + "epoch": 0.9713413364790785, + "grad_norm": 1.9893475770950317, + "learning_rate": 0.0001771696651659224, + "loss": 0.6284, + "step": 13998 + }, + { + "epoch": 0.9714107279161751, + "grad_norm": 2.684260606765747, + "learning_rate": 0.0001771660916564569, + "loss": 0.8694, + "step": 13999 + }, + { + "epoch": 0.9714801193532718, + "grad_norm": 2.5939342975616455, + "learning_rate": 0.00017716251790338697, + "loss": 0.9262, + "step": 14000 + }, + { + "epoch": 0.9715495107903684, + "grad_norm": 2.1592979431152344, + "learning_rate": 0.00017715894390672396, + "loss": 0.7794, + "step": 14001 + }, + { + "epoch": 0.9716189022274652, + "grad_norm": 2.524338960647583, + "learning_rate": 0.0001771553696664791, + "loss": 0.7277, + "step": 14002 + }, + { + "epoch": 0.9716882936645618, + "grad_norm": 2.343414306640625, + "learning_rate": 0.00017715179518266368, + "loss": 0.9317, + "step": 14003 + }, + { + "epoch": 0.9717576851016585, + "grad_norm": 7.05615758895874, + "learning_rate": 0.000177148220455289, + "loss": 0.8595, + "step": 14004 + }, + { + "epoch": 0.9718270765387551, + "grad_norm": 2.806063175201416, + "learning_rate": 0.00017714464548436633, + "loss": 0.9288, + "step": 14005 + }, + { + "epoch": 0.9718964679758518, + "grad_norm": 2.0473334789276123, + "learning_rate": 0.00017714107026990696, + "loss": 0.8941, + "step": 14006 + }, + { + "epoch": 0.9719658594129484, + "grad_norm": 2.7045552730560303, + "learning_rate": 0.0001771374948119222, + "loss": 0.6492, + "step": 14007 + }, + { + "epoch": 0.972035250850045, + "grad_norm": 2.577394485473633, + "learning_rate": 0.00017713391911042332, + "loss": 0.7269, + "step": 14008 + }, + { + "epoch": 0.9721046422871418, + "grad_norm": 2.630772113800049, + "learning_rate": 0.0001771303431654216, + "loss": 0.7761, + "step": 14009 + }, + { + "epoch": 0.9721740337242384, + "grad_norm": 2.4723212718963623, + "learning_rate": 0.00017712676697692836, + "loss": 1.0406, + "step": 14010 + }, + { + "epoch": 0.9722434251613351, + "grad_norm": 2.89349365234375, + "learning_rate": 0.00017712319054495482, + "loss": 0.7668, + "step": 14011 + }, + { + "epoch": 0.9723128165984317, + "grad_norm": 1.8818211555480957, + "learning_rate": 0.00017711961386951235, + "loss": 0.7233, + "step": 14012 + }, + { + "epoch": 0.9723822080355284, + "grad_norm": 3.346625566482544, + "learning_rate": 0.0001771160369506122, + "loss": 1.0066, + "step": 14013 + }, + { + "epoch": 0.972451599472625, + "grad_norm": 1.8122349977493286, + "learning_rate": 0.00017711245978826564, + "loss": 0.7035, + "step": 14014 + }, + { + "epoch": 0.9725209909097218, + "grad_norm": 2.9785420894622803, + "learning_rate": 0.00017710888238248401, + "loss": 0.6894, + "step": 14015 + }, + { + "epoch": 0.9725903823468184, + "grad_norm": 2.1753666400909424, + "learning_rate": 0.0001771053047332786, + "loss": 0.7876, + "step": 14016 + }, + { + "epoch": 0.9726597737839151, + "grad_norm": 2.4034018516540527, + "learning_rate": 0.0001771017268406607, + "loss": 0.7772, + "step": 14017 + }, + { + "epoch": 0.9727291652210117, + "grad_norm": 2.5729403495788574, + "learning_rate": 0.00017709814870464155, + "loss": 1.0485, + "step": 14018 + }, + { + "epoch": 0.9727985566581084, + "grad_norm": 2.725109815597534, + "learning_rate": 0.00017709457032523252, + "loss": 0.8765, + "step": 14019 + }, + { + "epoch": 0.972867948095205, + "grad_norm": 1.9484561681747437, + "learning_rate": 0.00017709099170244485, + "loss": 0.6268, + "step": 14020 + }, + { + "epoch": 0.9729373395323017, + "grad_norm": 3.094363212585449, + "learning_rate": 0.00017708741283628987, + "loss": 1.0608, + "step": 14021 + }, + { + "epoch": 0.9730067309693984, + "grad_norm": 2.1353342533111572, + "learning_rate": 0.00017708383372677888, + "loss": 1.0996, + "step": 14022 + }, + { + "epoch": 0.973076122406495, + "grad_norm": 3.4351234436035156, + "learning_rate": 0.00017708025437392316, + "loss": 0.9094, + "step": 14023 + }, + { + "epoch": 0.9731455138435917, + "grad_norm": 2.3876724243164062, + "learning_rate": 0.00017707667477773407, + "loss": 1.0065, + "step": 14024 + }, + { + "epoch": 0.9732149052806883, + "grad_norm": 2.790640115737915, + "learning_rate": 0.0001770730949382228, + "loss": 0.7194, + "step": 14025 + }, + { + "epoch": 0.973284296717785, + "grad_norm": 3.308774948120117, + "learning_rate": 0.00017706951485540074, + "loss": 0.7413, + "step": 14026 + }, + { + "epoch": 0.9733536881548817, + "grad_norm": 2.051605224609375, + "learning_rate": 0.00017706593452927914, + "loss": 0.6971, + "step": 14027 + }, + { + "epoch": 0.9734230795919784, + "grad_norm": 2.929028034210205, + "learning_rate": 0.00017706235395986936, + "loss": 0.8483, + "step": 14028 + }, + { + "epoch": 0.973492471029075, + "grad_norm": 1.8491710424423218, + "learning_rate": 0.00017705877314718265, + "loss": 0.6228, + "step": 14029 + }, + { + "epoch": 0.9735618624661717, + "grad_norm": 3.02115797996521, + "learning_rate": 0.0001770551920912303, + "loss": 0.8968, + "step": 14030 + }, + { + "epoch": 0.9736312539032683, + "grad_norm": 2.098451614379883, + "learning_rate": 0.0001770516107920237, + "loss": 0.869, + "step": 14031 + }, + { + "epoch": 0.973700645340365, + "grad_norm": 2.510328531265259, + "learning_rate": 0.00017704802924957408, + "loss": 0.8271, + "step": 14032 + }, + { + "epoch": 0.9737700367774617, + "grad_norm": 2.122645139694214, + "learning_rate": 0.0001770444474638928, + "loss": 0.6939, + "step": 14033 + }, + { + "epoch": 0.9738394282145584, + "grad_norm": 2.5381243228912354, + "learning_rate": 0.0001770408654349911, + "loss": 1.0952, + "step": 14034 + }, + { + "epoch": 0.973908819651655, + "grad_norm": 2.573472023010254, + "learning_rate": 0.00017703728316288036, + "loss": 0.8945, + "step": 14035 + }, + { + "epoch": 0.9739782110887516, + "grad_norm": 2.4337379932403564, + "learning_rate": 0.00017703370064757183, + "loss": 0.86, + "step": 14036 + }, + { + "epoch": 0.9740476025258483, + "grad_norm": 2.900285243988037, + "learning_rate": 0.00017703011788907684, + "loss": 0.7959, + "step": 14037 + }, + { + "epoch": 0.9741169939629449, + "grad_norm": 2.3079833984375, + "learning_rate": 0.00017702653488740671, + "loss": 0.8159, + "step": 14038 + }, + { + "epoch": 0.9741863854000417, + "grad_norm": 2.360240936279297, + "learning_rate": 0.00017702295164257275, + "loss": 0.8506, + "step": 14039 + }, + { + "epoch": 0.9742557768371383, + "grad_norm": 2.493068218231201, + "learning_rate": 0.0001770193681545863, + "loss": 0.8182, + "step": 14040 + }, + { + "epoch": 0.974325168274235, + "grad_norm": 2.7688956260681152, + "learning_rate": 0.0001770157844234586, + "loss": 1.2159, + "step": 14041 + }, + { + "epoch": 0.9743945597113316, + "grad_norm": 2.575073480606079, + "learning_rate": 0.000177012200449201, + "loss": 0.9738, + "step": 14042 + }, + { + "epoch": 0.9744639511484283, + "grad_norm": 1.559130311012268, + "learning_rate": 0.00017700861623182488, + "loss": 0.5479, + "step": 14043 + }, + { + "epoch": 0.9745333425855249, + "grad_norm": 2.8142755031585693, + "learning_rate": 0.00017700503177134143, + "loss": 0.9406, + "step": 14044 + }, + { + "epoch": 0.9746027340226217, + "grad_norm": 2.053544044494629, + "learning_rate": 0.00017700144706776205, + "loss": 0.692, + "step": 14045 + }, + { + "epoch": 0.9746721254597183, + "grad_norm": 2.628931760787964, + "learning_rate": 0.00017699786212109803, + "loss": 0.9106, + "step": 14046 + }, + { + "epoch": 0.974741516896815, + "grad_norm": 2.8003783226013184, + "learning_rate": 0.00017699427693136072, + "loss": 0.8516, + "step": 14047 + }, + { + "epoch": 0.9748109083339116, + "grad_norm": 2.4147262573242188, + "learning_rate": 0.00017699069149856134, + "loss": 0.6357, + "step": 14048 + }, + { + "epoch": 0.9748802997710082, + "grad_norm": 2.3378660678863525, + "learning_rate": 0.00017698710582271133, + "loss": 0.6952, + "step": 14049 + }, + { + "epoch": 0.9749496912081049, + "grad_norm": 1.482175350189209, + "learning_rate": 0.00017698351990382198, + "loss": 0.7317, + "step": 14050 + }, + { + "epoch": 0.9750190826452015, + "grad_norm": 1.7033767700195312, + "learning_rate": 0.00017697993374190456, + "loss": 0.797, + "step": 14051 + }, + { + "epoch": 0.9750884740822983, + "grad_norm": 2.48146653175354, + "learning_rate": 0.0001769763473369704, + "loss": 0.9152, + "step": 14052 + }, + { + "epoch": 0.9751578655193949, + "grad_norm": 2.4599039554595947, + "learning_rate": 0.00017697276068903088, + "loss": 0.8536, + "step": 14053 + }, + { + "epoch": 0.9752272569564916, + "grad_norm": 2.1623375415802, + "learning_rate": 0.00017696917379809727, + "loss": 0.916, + "step": 14054 + }, + { + "epoch": 0.9752966483935882, + "grad_norm": 1.9896968603134155, + "learning_rate": 0.00017696558666418092, + "loss": 0.653, + "step": 14055 + }, + { + "epoch": 0.9753660398306849, + "grad_norm": 1.7951170206069946, + "learning_rate": 0.0001769619992872931, + "loss": 0.6451, + "step": 14056 + }, + { + "epoch": 0.9754354312677815, + "grad_norm": 2.118149518966675, + "learning_rate": 0.00017695841166744526, + "loss": 0.7124, + "step": 14057 + }, + { + "epoch": 0.9755048227048783, + "grad_norm": 2.016263484954834, + "learning_rate": 0.00017695482380464855, + "loss": 1.0863, + "step": 14058 + }, + { + "epoch": 0.9755742141419749, + "grad_norm": 1.8034732341766357, + "learning_rate": 0.00017695123569891444, + "loss": 0.6236, + "step": 14059 + }, + { + "epoch": 0.9756436055790716, + "grad_norm": 1.6742727756500244, + "learning_rate": 0.0001769476473502542, + "loss": 0.6956, + "step": 14060 + }, + { + "epoch": 0.9757129970161682, + "grad_norm": 2.264859914779663, + "learning_rate": 0.00017694405875867916, + "loss": 1.0456, + "step": 14061 + }, + { + "epoch": 0.9757823884532648, + "grad_norm": 2.3153679370880127, + "learning_rate": 0.00017694046992420068, + "loss": 1.1017, + "step": 14062 + }, + { + "epoch": 0.9758517798903615, + "grad_norm": 2.2487409114837646, + "learning_rate": 0.00017693688084683003, + "loss": 0.7969, + "step": 14063 + }, + { + "epoch": 0.9759211713274581, + "grad_norm": 2.1592087745666504, + "learning_rate": 0.0001769332915265786, + "loss": 0.9115, + "step": 14064 + }, + { + "epoch": 0.9759905627645549, + "grad_norm": 2.133898973464966, + "learning_rate": 0.00017692970196345767, + "loss": 0.8718, + "step": 14065 + }, + { + "epoch": 0.9760599542016515, + "grad_norm": 2.0826594829559326, + "learning_rate": 0.0001769261121574786, + "loss": 0.6626, + "step": 14066 + }, + { + "epoch": 0.9761293456387482, + "grad_norm": 1.7641717195510864, + "learning_rate": 0.00017692252210865274, + "loss": 0.751, + "step": 14067 + }, + { + "epoch": 0.9761987370758448, + "grad_norm": 2.4159955978393555, + "learning_rate": 0.00017691893181699138, + "loss": 0.9436, + "step": 14068 + }, + { + "epoch": 0.9762681285129415, + "grad_norm": 2.4509427547454834, + "learning_rate": 0.00017691534128250587, + "loss": 0.6367, + "step": 14069 + }, + { + "epoch": 0.9763375199500381, + "grad_norm": 2.1142704486846924, + "learning_rate": 0.0001769117505052076, + "loss": 0.8681, + "step": 14070 + }, + { + "epoch": 0.9764069113871349, + "grad_norm": 2.063448429107666, + "learning_rate": 0.00017690815948510783, + "loss": 0.68, + "step": 14071 + }, + { + "epoch": 0.9764763028242315, + "grad_norm": 2.7785797119140625, + "learning_rate": 0.0001769045682222179, + "loss": 0.9191, + "step": 14072 + }, + { + "epoch": 0.9765456942613282, + "grad_norm": 2.395930051803589, + "learning_rate": 0.00017690097671654918, + "loss": 1.1204, + "step": 14073 + }, + { + "epoch": 0.9766150856984248, + "grad_norm": 2.1180825233459473, + "learning_rate": 0.000176897384968113, + "loss": 0.7995, + "step": 14074 + }, + { + "epoch": 0.9766844771355214, + "grad_norm": 2.346683979034424, + "learning_rate": 0.00017689379297692071, + "loss": 0.889, + "step": 14075 + }, + { + "epoch": 0.9767538685726181, + "grad_norm": 2.091287136077881, + "learning_rate": 0.00017689020074298362, + "loss": 0.6308, + "step": 14076 + }, + { + "epoch": 0.9768232600097148, + "grad_norm": 1.7371331453323364, + "learning_rate": 0.00017688660826631312, + "loss": 0.7809, + "step": 14077 + }, + { + "epoch": 0.9768926514468115, + "grad_norm": 3.6259920597076416, + "learning_rate": 0.00017688301554692046, + "loss": 1.0289, + "step": 14078 + }, + { + "epoch": 0.9769620428839081, + "grad_norm": 1.8475192785263062, + "learning_rate": 0.00017687942258481709, + "loss": 0.7719, + "step": 14079 + }, + { + "epoch": 0.9770314343210048, + "grad_norm": 1.9133414030075073, + "learning_rate": 0.00017687582938001427, + "loss": 0.7702, + "step": 14080 + }, + { + "epoch": 0.9771008257581014, + "grad_norm": 2.2132463455200195, + "learning_rate": 0.0001768722359325234, + "loss": 0.9547, + "step": 14081 + }, + { + "epoch": 0.9771702171951981, + "grad_norm": 1.9722137451171875, + "learning_rate": 0.0001768686422423558, + "loss": 0.6889, + "step": 14082 + }, + { + "epoch": 0.9772396086322948, + "grad_norm": 2.294312000274658, + "learning_rate": 0.00017686504830952279, + "loss": 0.6782, + "step": 14083 + }, + { + "epoch": 0.9773090000693915, + "grad_norm": 1.876211166381836, + "learning_rate": 0.0001768614541340358, + "loss": 0.8577, + "step": 14084 + }, + { + "epoch": 0.9773783915064881, + "grad_norm": 2.993523597717285, + "learning_rate": 0.00017685785971590604, + "loss": 1.0866, + "step": 14085 + }, + { + "epoch": 0.9774477829435848, + "grad_norm": 2.05020809173584, + "learning_rate": 0.00017685426505514499, + "loss": 0.8177, + "step": 14086 + }, + { + "epoch": 0.9775171743806814, + "grad_norm": 1.5491511821746826, + "learning_rate": 0.0001768506701517639, + "loss": 0.7055, + "step": 14087 + }, + { + "epoch": 0.9775865658177781, + "grad_norm": 2.3899190425872803, + "learning_rate": 0.0001768470750057742, + "loss": 0.8508, + "step": 14088 + }, + { + "epoch": 0.9776559572548748, + "grad_norm": 2.3847978115081787, + "learning_rate": 0.00017684347961718718, + "loss": 0.9243, + "step": 14089 + }, + { + "epoch": 0.9777253486919714, + "grad_norm": 1.5690335035324097, + "learning_rate": 0.00017683988398601422, + "loss": 0.5813, + "step": 14090 + }, + { + "epoch": 0.9777947401290681, + "grad_norm": 1.509518027305603, + "learning_rate": 0.00017683628811226664, + "loss": 0.5509, + "step": 14091 + }, + { + "epoch": 0.9778641315661647, + "grad_norm": 1.6132177114486694, + "learning_rate": 0.00017683269199595585, + "loss": 0.8286, + "step": 14092 + }, + { + "epoch": 0.9779335230032614, + "grad_norm": 2.4391119480133057, + "learning_rate": 0.00017682909563709314, + "loss": 0.7978, + "step": 14093 + }, + { + "epoch": 0.978002914440358, + "grad_norm": 1.3696073293685913, + "learning_rate": 0.00017682549903568986, + "loss": 0.6464, + "step": 14094 + }, + { + "epoch": 0.9780723058774548, + "grad_norm": 2.196054220199585, + "learning_rate": 0.00017682190219175744, + "loss": 0.7321, + "step": 14095 + }, + { + "epoch": 0.9781416973145514, + "grad_norm": 2.633225917816162, + "learning_rate": 0.00017681830510530718, + "loss": 0.9165, + "step": 14096 + }, + { + "epoch": 0.9782110887516481, + "grad_norm": 1.9747543334960938, + "learning_rate": 0.00017681470777635043, + "loss": 0.6753, + "step": 14097 + }, + { + "epoch": 0.9782804801887447, + "grad_norm": 2.2677741050720215, + "learning_rate": 0.00017681111020489856, + "loss": 0.8656, + "step": 14098 + }, + { + "epoch": 0.9783498716258414, + "grad_norm": 1.8789076805114746, + "learning_rate": 0.00017680751239096293, + "loss": 0.7755, + "step": 14099 + }, + { + "epoch": 0.978419263062938, + "grad_norm": 1.861351490020752, + "learning_rate": 0.0001768039143345549, + "loss": 0.6866, + "step": 14100 + }, + { + "epoch": 0.9784886545000347, + "grad_norm": 1.978823184967041, + "learning_rate": 0.00017680031603568584, + "loss": 0.7111, + "step": 14101 + }, + { + "epoch": 0.9785580459371314, + "grad_norm": 2.792861223220825, + "learning_rate": 0.00017679671749436707, + "loss": 0.8905, + "step": 14102 + }, + { + "epoch": 0.978627437374228, + "grad_norm": 2.4549429416656494, + "learning_rate": 0.00017679311871060996, + "loss": 0.8144, + "step": 14103 + }, + { + "epoch": 0.9786968288113247, + "grad_norm": 1.8266048431396484, + "learning_rate": 0.0001767895196844259, + "loss": 0.6355, + "step": 14104 + }, + { + "epoch": 0.9787662202484213, + "grad_norm": 2.107539653778076, + "learning_rate": 0.00017678592041582623, + "loss": 0.7045, + "step": 14105 + }, + { + "epoch": 0.978835611685518, + "grad_norm": 1.865914225578308, + "learning_rate": 0.00017678232090482233, + "loss": 0.7448, + "step": 14106 + }, + { + "epoch": 0.9789050031226146, + "grad_norm": 1.9220128059387207, + "learning_rate": 0.00017677872115142554, + "loss": 0.7902, + "step": 14107 + }, + { + "epoch": 0.9789743945597114, + "grad_norm": 1.5964897871017456, + "learning_rate": 0.00017677512115564726, + "loss": 0.5775, + "step": 14108 + }, + { + "epoch": 0.979043785996808, + "grad_norm": 2.4990339279174805, + "learning_rate": 0.0001767715209174988, + "loss": 0.9183, + "step": 14109 + }, + { + "epoch": 0.9791131774339047, + "grad_norm": 1.694413661956787, + "learning_rate": 0.00017676792043699155, + "loss": 0.6788, + "step": 14110 + }, + { + "epoch": 0.9791825688710013, + "grad_norm": 2.113827705383301, + "learning_rate": 0.0001767643197141369, + "loss": 0.7106, + "step": 14111 + }, + { + "epoch": 0.979251960308098, + "grad_norm": 2.1247050762176514, + "learning_rate": 0.00017676071874894622, + "loss": 1.019, + "step": 14112 + }, + { + "epoch": 0.9793213517451946, + "grad_norm": 1.924554467201233, + "learning_rate": 0.00017675711754143084, + "loss": 0.7446, + "step": 14113 + }, + { + "epoch": 0.9793907431822914, + "grad_norm": 3.0028085708618164, + "learning_rate": 0.0001767535160916021, + "loss": 0.8318, + "step": 14114 + }, + { + "epoch": 0.979460134619388, + "grad_norm": 1.7831896543502808, + "learning_rate": 0.00017674991439947148, + "loss": 0.627, + "step": 14115 + }, + { + "epoch": 0.9795295260564846, + "grad_norm": 1.986406683921814, + "learning_rate": 0.00017674631246505026, + "loss": 0.7896, + "step": 14116 + }, + { + "epoch": 0.9795989174935813, + "grad_norm": 2.491201639175415, + "learning_rate": 0.00017674271028834983, + "loss": 0.8208, + "step": 14117 + }, + { + "epoch": 0.9796683089306779, + "grad_norm": 2.79571533203125, + "learning_rate": 0.00017673910786938156, + "loss": 1.1377, + "step": 14118 + }, + { + "epoch": 0.9797377003677746, + "grad_norm": 1.9932068586349487, + "learning_rate": 0.00017673550520815685, + "loss": 0.8046, + "step": 14119 + }, + { + "epoch": 0.9798070918048712, + "grad_norm": 2.1133768558502197, + "learning_rate": 0.00017673190230468703, + "loss": 0.7137, + "step": 14120 + }, + { + "epoch": 0.979876483241968, + "grad_norm": 2.4064888954162598, + "learning_rate": 0.00017672829915898352, + "loss": 1.0487, + "step": 14121 + }, + { + "epoch": 0.9799458746790646, + "grad_norm": 2.1040425300598145, + "learning_rate": 0.00017672469577105767, + "loss": 0.9156, + "step": 14122 + }, + { + "epoch": 0.9800152661161613, + "grad_norm": 2.411703109741211, + "learning_rate": 0.00017672109214092086, + "loss": 0.639, + "step": 14123 + }, + { + "epoch": 0.9800846575532579, + "grad_norm": 2.6633079051971436, + "learning_rate": 0.00017671748826858443, + "loss": 1.0166, + "step": 14124 + }, + { + "epoch": 0.9801540489903546, + "grad_norm": 1.527295708656311, + "learning_rate": 0.00017671388415405983, + "loss": 0.4493, + "step": 14125 + }, + { + "epoch": 0.9802234404274512, + "grad_norm": 2.3066771030426025, + "learning_rate": 0.00017671027979735836, + "loss": 0.8055, + "step": 14126 + }, + { + "epoch": 0.980292831864548, + "grad_norm": 2.10496187210083, + "learning_rate": 0.00017670667519849146, + "loss": 0.6501, + "step": 14127 + }, + { + "epoch": 0.9803622233016446, + "grad_norm": 1.9602410793304443, + "learning_rate": 0.00017670307035747048, + "loss": 0.8976, + "step": 14128 + }, + { + "epoch": 0.9804316147387412, + "grad_norm": 6.028655052185059, + "learning_rate": 0.00017669946527430682, + "loss": 0.7884, + "step": 14129 + }, + { + "epoch": 0.9805010061758379, + "grad_norm": 2.441540002822876, + "learning_rate": 0.00017669585994901184, + "loss": 0.9535, + "step": 14130 + }, + { + "epoch": 0.9805703976129345, + "grad_norm": 2.372995138168335, + "learning_rate": 0.00017669225438159694, + "loss": 0.7883, + "step": 14131 + }, + { + "epoch": 0.9806397890500312, + "grad_norm": 1.737420916557312, + "learning_rate": 0.00017668864857207347, + "loss": 0.6189, + "step": 14132 + }, + { + "epoch": 0.9807091804871279, + "grad_norm": 2.8044283390045166, + "learning_rate": 0.00017668504252045284, + "loss": 0.8942, + "step": 14133 + }, + { + "epoch": 0.9807785719242246, + "grad_norm": 3.3505630493164062, + "learning_rate": 0.00017668143622674644, + "loss": 0.9038, + "step": 14134 + }, + { + "epoch": 0.9808479633613212, + "grad_norm": 2.7238941192626953, + "learning_rate": 0.00017667782969096562, + "loss": 0.818, + "step": 14135 + }, + { + "epoch": 0.9809173547984179, + "grad_norm": 1.8254214525222778, + "learning_rate": 0.0001766742229131218, + "loss": 0.5756, + "step": 14136 + }, + { + "epoch": 0.9809867462355145, + "grad_norm": 2.0586864948272705, + "learning_rate": 0.00017667061589322637, + "loss": 0.7195, + "step": 14137 + }, + { + "epoch": 0.9810561376726112, + "grad_norm": 1.7388492822647095, + "learning_rate": 0.00017666700863129068, + "loss": 0.9518, + "step": 14138 + }, + { + "epoch": 0.9811255291097079, + "grad_norm": 2.773258924484253, + "learning_rate": 0.00017666340112732612, + "loss": 0.9282, + "step": 14139 + }, + { + "epoch": 0.9811949205468046, + "grad_norm": 1.9450587034225464, + "learning_rate": 0.00017665979338134412, + "loss": 0.5257, + "step": 14140 + }, + { + "epoch": 0.9812643119839012, + "grad_norm": 2.662809371948242, + "learning_rate": 0.00017665618539335607, + "loss": 0.8026, + "step": 14141 + }, + { + "epoch": 0.9813337034209978, + "grad_norm": 2.725539207458496, + "learning_rate": 0.00017665257716337332, + "loss": 0.6905, + "step": 14142 + }, + { + "epoch": 0.9814030948580945, + "grad_norm": 3.677323818206787, + "learning_rate": 0.0001766489686914073, + "loss": 1.1882, + "step": 14143 + }, + { + "epoch": 0.9814724862951911, + "grad_norm": 2.301201105117798, + "learning_rate": 0.0001766453599774693, + "loss": 0.6804, + "step": 14144 + }, + { + "epoch": 0.9815418777322878, + "grad_norm": 1.9162434339523315, + "learning_rate": 0.00017664175102157087, + "loss": 0.8294, + "step": 14145 + }, + { + "epoch": 0.9816112691693845, + "grad_norm": 2.582551956176758, + "learning_rate": 0.00017663814182372327, + "loss": 0.6745, + "step": 14146 + }, + { + "epoch": 0.9816806606064812, + "grad_norm": 5.547406196594238, + "learning_rate": 0.00017663453238393796, + "loss": 0.7564, + "step": 14147 + }, + { + "epoch": 0.9817500520435778, + "grad_norm": 2.030123472213745, + "learning_rate": 0.00017663092270222633, + "loss": 0.939, + "step": 14148 + }, + { + "epoch": 0.9818194434806745, + "grad_norm": 2.5948569774627686, + "learning_rate": 0.0001766273127785998, + "loss": 0.8423, + "step": 14149 + }, + { + "epoch": 0.9818888349177711, + "grad_norm": 2.1133925914764404, + "learning_rate": 0.00017662370261306971, + "loss": 0.9932, + "step": 14150 + }, + { + "epoch": 0.9819582263548678, + "grad_norm": 1.7889246940612793, + "learning_rate": 0.00017662009220564747, + "loss": 0.5923, + "step": 14151 + }, + { + "epoch": 0.9820276177919645, + "grad_norm": 2.630470037460327, + "learning_rate": 0.00017661648155634451, + "loss": 0.7756, + "step": 14152 + }, + { + "epoch": 0.9820970092290612, + "grad_norm": 2.7955901622772217, + "learning_rate": 0.00017661287066517218, + "loss": 1.062, + "step": 14153 + }, + { + "epoch": 0.9821664006661578, + "grad_norm": 2.065886974334717, + "learning_rate": 0.00017660925953214194, + "loss": 0.7194, + "step": 14154 + }, + { + "epoch": 0.9822357921032545, + "grad_norm": 2.269534111022949, + "learning_rate": 0.00017660564815726512, + "loss": 0.7819, + "step": 14155 + }, + { + "epoch": 0.9823051835403511, + "grad_norm": 2.137655019760132, + "learning_rate": 0.00017660203654055318, + "loss": 0.8383, + "step": 14156 + }, + { + "epoch": 0.9823745749774477, + "grad_norm": 1.971085548400879, + "learning_rate": 0.00017659842468201748, + "loss": 0.8025, + "step": 14157 + }, + { + "epoch": 0.9824439664145445, + "grad_norm": 2.164716958999634, + "learning_rate": 0.00017659481258166946, + "loss": 0.8032, + "step": 14158 + }, + { + "epoch": 0.9825133578516411, + "grad_norm": 2.1094629764556885, + "learning_rate": 0.0001765912002395205, + "loss": 0.6853, + "step": 14159 + }, + { + "epoch": 0.9825827492887378, + "grad_norm": 1.7542170286178589, + "learning_rate": 0.000176587587655582, + "loss": 0.4708, + "step": 14160 + }, + { + "epoch": 0.9826521407258344, + "grad_norm": 3.684636354446411, + "learning_rate": 0.00017658397482986536, + "loss": 0.9831, + "step": 14161 + }, + { + "epoch": 0.9827215321629311, + "grad_norm": 2.243762493133545, + "learning_rate": 0.000176580361762382, + "loss": 1.0672, + "step": 14162 + }, + { + "epoch": 0.9827909236000277, + "grad_norm": 2.70983624458313, + "learning_rate": 0.00017657674845314333, + "loss": 0.8241, + "step": 14163 + }, + { + "epoch": 0.9828603150371245, + "grad_norm": 2.728748321533203, + "learning_rate": 0.00017657313490216078, + "loss": 0.6985, + "step": 14164 + }, + { + "epoch": 0.9829297064742211, + "grad_norm": 1.8336530923843384, + "learning_rate": 0.0001765695211094457, + "loss": 0.633, + "step": 14165 + }, + { + "epoch": 0.9829990979113178, + "grad_norm": 2.3175792694091797, + "learning_rate": 0.0001765659070750095, + "loss": 0.8409, + "step": 14166 + }, + { + "epoch": 0.9830684893484144, + "grad_norm": 2.2604432106018066, + "learning_rate": 0.00017656229279886367, + "loss": 0.7703, + "step": 14167 + }, + { + "epoch": 0.9831378807855111, + "grad_norm": 1.8186638355255127, + "learning_rate": 0.00017655867828101954, + "loss": 0.88, + "step": 14168 + }, + { + "epoch": 0.9832072722226077, + "grad_norm": 2.1338770389556885, + "learning_rate": 0.0001765550635214885, + "loss": 0.7601, + "step": 14169 + }, + { + "epoch": 0.9832766636597043, + "grad_norm": 2.593625068664551, + "learning_rate": 0.00017655144852028208, + "loss": 0.9922, + "step": 14170 + }, + { + "epoch": 0.9833460550968011, + "grad_norm": 1.8118836879730225, + "learning_rate": 0.00017654783327741158, + "loss": 0.9029, + "step": 14171 + }, + { + "epoch": 0.9834154465338977, + "grad_norm": 2.340657949447632, + "learning_rate": 0.00017654421779288843, + "loss": 0.7056, + "step": 14172 + }, + { + "epoch": 0.9834848379709944, + "grad_norm": 2.2542366981506348, + "learning_rate": 0.0001765406020667241, + "loss": 0.7794, + "step": 14173 + }, + { + "epoch": 0.983554229408091, + "grad_norm": 2.5396783351898193, + "learning_rate": 0.00017653698609892997, + "loss": 0.5534, + "step": 14174 + }, + { + "epoch": 0.9836236208451877, + "grad_norm": 2.1622087955474854, + "learning_rate": 0.00017653336988951744, + "loss": 0.7643, + "step": 14175 + }, + { + "epoch": 0.9836930122822843, + "grad_norm": 2.0402870178222656, + "learning_rate": 0.00017652975343849793, + "loss": 0.8695, + "step": 14176 + }, + { + "epoch": 0.9837624037193811, + "grad_norm": 2.233208417892456, + "learning_rate": 0.0001765261367458829, + "loss": 0.7038, + "step": 14177 + }, + { + "epoch": 0.9838317951564777, + "grad_norm": 2.346721887588501, + "learning_rate": 0.00017652251981168372, + "loss": 1.204, + "step": 14178 + }, + { + "epoch": 0.9839011865935744, + "grad_norm": 2.510138511657715, + "learning_rate": 0.0001765189026359118, + "loss": 1.1326, + "step": 14179 + }, + { + "epoch": 0.983970578030671, + "grad_norm": 2.231333017349243, + "learning_rate": 0.0001765152852185786, + "loss": 0.7397, + "step": 14180 + }, + { + "epoch": 0.9840399694677677, + "grad_norm": 2.5906052589416504, + "learning_rate": 0.00017651166755969556, + "loss": 1.0138, + "step": 14181 + }, + { + "epoch": 0.9841093609048643, + "grad_norm": 2.058290719985962, + "learning_rate": 0.000176508049659274, + "loss": 0.7926, + "step": 14182 + }, + { + "epoch": 0.984178752341961, + "grad_norm": 2.025102376937866, + "learning_rate": 0.00017650443151732544, + "loss": 0.7138, + "step": 14183 + }, + { + "epoch": 0.9842481437790577, + "grad_norm": 1.5768976211547852, + "learning_rate": 0.00017650081313386128, + "loss": 0.8072, + "step": 14184 + }, + { + "epoch": 0.9843175352161543, + "grad_norm": 1.915937900543213, + "learning_rate": 0.0001764971945088929, + "loss": 0.8734, + "step": 14185 + }, + { + "epoch": 0.984386926653251, + "grad_norm": 2.5249133110046387, + "learning_rate": 0.00017649357564243173, + "loss": 0.836, + "step": 14186 + }, + { + "epoch": 0.9844563180903476, + "grad_norm": 3.386936664581299, + "learning_rate": 0.00017648995653448925, + "loss": 0.8769, + "step": 14187 + }, + { + "epoch": 0.9845257095274443, + "grad_norm": 2.063106060028076, + "learning_rate": 0.00017648633718507685, + "loss": 0.5562, + "step": 14188 + }, + { + "epoch": 0.984595100964541, + "grad_norm": 1.8823010921478271, + "learning_rate": 0.00017648271759420595, + "loss": 0.7811, + "step": 14189 + }, + { + "epoch": 0.9846644924016377, + "grad_norm": 2.353951930999756, + "learning_rate": 0.00017647909776188802, + "loss": 0.9764, + "step": 14190 + }, + { + "epoch": 0.9847338838387343, + "grad_norm": 2.3522112369537354, + "learning_rate": 0.0001764754776881344, + "loss": 0.8126, + "step": 14191 + }, + { + "epoch": 0.984803275275831, + "grad_norm": 2.2512168884277344, + "learning_rate": 0.0001764718573729566, + "loss": 0.8418, + "step": 14192 + }, + { + "epoch": 0.9848726667129276, + "grad_norm": 1.9124683141708374, + "learning_rate": 0.00017646823681636602, + "loss": 0.7928, + "step": 14193 + }, + { + "epoch": 0.9849420581500243, + "grad_norm": 2.1435837745666504, + "learning_rate": 0.00017646461601837407, + "loss": 0.6307, + "step": 14194 + }, + { + "epoch": 0.985011449587121, + "grad_norm": 2.538102865219116, + "learning_rate": 0.0001764609949789922, + "loss": 0.6549, + "step": 14195 + }, + { + "epoch": 0.9850808410242176, + "grad_norm": 2.9491031169891357, + "learning_rate": 0.00017645737369823185, + "loss": 0.8509, + "step": 14196 + }, + { + "epoch": 0.9851502324613143, + "grad_norm": 2.4564273357391357, + "learning_rate": 0.00017645375217610444, + "loss": 1.253, + "step": 14197 + }, + { + "epoch": 0.9852196238984109, + "grad_norm": 1.4904824495315552, + "learning_rate": 0.0001764501304126214, + "loss": 0.6288, + "step": 14198 + }, + { + "epoch": 0.9852890153355076, + "grad_norm": 1.7830848693847656, + "learning_rate": 0.0001764465084077942, + "loss": 0.7878, + "step": 14199 + }, + { + "epoch": 0.9853584067726042, + "grad_norm": 3.446833372116089, + "learning_rate": 0.00017644288616163421, + "loss": 1.0048, + "step": 14200 + }, + { + "epoch": 0.985427798209701, + "grad_norm": 2.491124153137207, + "learning_rate": 0.0001764392636741529, + "loss": 0.9169, + "step": 14201 + }, + { + "epoch": 0.9854971896467976, + "grad_norm": 1.8069078922271729, + "learning_rate": 0.00017643564094536166, + "loss": 0.7787, + "step": 14202 + }, + { + "epoch": 0.9855665810838943, + "grad_norm": 1.6192584037780762, + "learning_rate": 0.00017643201797527204, + "loss": 0.6354, + "step": 14203 + }, + { + "epoch": 0.9856359725209909, + "grad_norm": 1.7506159543991089, + "learning_rate": 0.00017642839476389539, + "loss": 0.5706, + "step": 14204 + }, + { + "epoch": 0.9857053639580876, + "grad_norm": 2.8941829204559326, + "learning_rate": 0.00017642477131124315, + "loss": 1.0891, + "step": 14205 + }, + { + "epoch": 0.9857747553951842, + "grad_norm": 2.6354317665100098, + "learning_rate": 0.00017642114761732677, + "loss": 0.7722, + "step": 14206 + }, + { + "epoch": 0.9858441468322809, + "grad_norm": 2.276623010635376, + "learning_rate": 0.0001764175236821577, + "loss": 0.8536, + "step": 14207 + }, + { + "epoch": 0.9859135382693776, + "grad_norm": 1.8107084035873413, + "learning_rate": 0.0001764138995057474, + "loss": 0.835, + "step": 14208 + }, + { + "epoch": 0.9859829297064743, + "grad_norm": 3.249323606491089, + "learning_rate": 0.00017641027508810724, + "loss": 0.9995, + "step": 14209 + }, + { + "epoch": 0.9860523211435709, + "grad_norm": 2.364408254623413, + "learning_rate": 0.00017640665042924873, + "loss": 0.9617, + "step": 14210 + }, + { + "epoch": 0.9861217125806675, + "grad_norm": 15.363763809204102, + "learning_rate": 0.0001764030255291833, + "loss": 1.0234, + "step": 14211 + }, + { + "epoch": 0.9861911040177642, + "grad_norm": 1.8469809293746948, + "learning_rate": 0.00017639940038792235, + "loss": 0.8507, + "step": 14212 + }, + { + "epoch": 0.9862604954548608, + "grad_norm": 2.1500539779663086, + "learning_rate": 0.0001763957750054774, + "loss": 0.6799, + "step": 14213 + }, + { + "epoch": 0.9863298868919576, + "grad_norm": 2.2266697883605957, + "learning_rate": 0.00017639214938185982, + "loss": 0.7647, + "step": 14214 + }, + { + "epoch": 0.9863992783290542, + "grad_norm": 2.8433666229248047, + "learning_rate": 0.00017638852351708112, + "loss": 0.7289, + "step": 14215 + }, + { + "epoch": 0.9864686697661509, + "grad_norm": 2.3965210914611816, + "learning_rate": 0.00017638489741115266, + "loss": 0.8868, + "step": 14216 + }, + { + "epoch": 0.9865380612032475, + "grad_norm": 3.4975388050079346, + "learning_rate": 0.00017638127106408598, + "loss": 0.872, + "step": 14217 + }, + { + "epoch": 0.9866074526403442, + "grad_norm": 2.48150897026062, + "learning_rate": 0.0001763776444758925, + "loss": 0.5408, + "step": 14218 + }, + { + "epoch": 0.9866768440774408, + "grad_norm": 2.2306809425354004, + "learning_rate": 0.0001763740176465836, + "loss": 0.8495, + "step": 14219 + }, + { + "epoch": 0.9867462355145376, + "grad_norm": 2.538872718811035, + "learning_rate": 0.00017637039057617085, + "loss": 0.6801, + "step": 14220 + }, + { + "epoch": 0.9868156269516342, + "grad_norm": 2.013678789138794, + "learning_rate": 0.00017636676326466562, + "loss": 0.8118, + "step": 14221 + }, + { + "epoch": 0.9868850183887309, + "grad_norm": 3.3310587406158447, + "learning_rate": 0.00017636313571207937, + "loss": 0.7082, + "step": 14222 + }, + { + "epoch": 0.9869544098258275, + "grad_norm": 2.0229692459106445, + "learning_rate": 0.00017635950791842358, + "loss": 0.7066, + "step": 14223 + }, + { + "epoch": 0.9870238012629241, + "grad_norm": 3.2833032608032227, + "learning_rate": 0.00017635587988370962, + "loss": 0.6043, + "step": 14224 + }, + { + "epoch": 0.9870931927000208, + "grad_norm": 2.8801331520080566, + "learning_rate": 0.00017635225160794908, + "loss": 0.919, + "step": 14225 + }, + { + "epoch": 0.9871625841371174, + "grad_norm": 2.271918535232544, + "learning_rate": 0.0001763486230911533, + "loss": 0.7177, + "step": 14226 + }, + { + "epoch": 0.9872319755742142, + "grad_norm": 1.7864910364151, + "learning_rate": 0.0001763449943333338, + "loss": 0.7247, + "step": 14227 + }, + { + "epoch": 0.9873013670113108, + "grad_norm": 2.105023145675659, + "learning_rate": 0.000176341365334502, + "loss": 0.6911, + "step": 14228 + }, + { + "epoch": 0.9873707584484075, + "grad_norm": 2.125835418701172, + "learning_rate": 0.00017633773609466936, + "loss": 0.8065, + "step": 14229 + }, + { + "epoch": 0.9874401498855041, + "grad_norm": 2.950895071029663, + "learning_rate": 0.00017633410661384736, + "loss": 0.7729, + "step": 14230 + }, + { + "epoch": 0.9875095413226008, + "grad_norm": 2.3688037395477295, + "learning_rate": 0.00017633047689204743, + "loss": 0.9377, + "step": 14231 + }, + { + "epoch": 0.9875789327596974, + "grad_norm": 3.781606435775757, + "learning_rate": 0.000176326846929281, + "loss": 0.8698, + "step": 14232 + }, + { + "epoch": 0.9876483241967942, + "grad_norm": 1.757925271987915, + "learning_rate": 0.00017632321672555966, + "loss": 0.7186, + "step": 14233 + }, + { + "epoch": 0.9877177156338908, + "grad_norm": 3.0055360794067383, + "learning_rate": 0.0001763195862808947, + "loss": 1.0322, + "step": 14234 + }, + { + "epoch": 0.9877871070709875, + "grad_norm": 2.632319211959839, + "learning_rate": 0.0001763159555952977, + "loss": 0.8729, + "step": 14235 + }, + { + "epoch": 0.9878564985080841, + "grad_norm": 2.5019567012786865, + "learning_rate": 0.00017631232466878006, + "loss": 0.8342, + "step": 14236 + }, + { + "epoch": 0.9879258899451807, + "grad_norm": 2.088721752166748, + "learning_rate": 0.0001763086935013533, + "loss": 0.7906, + "step": 14237 + }, + { + "epoch": 0.9879952813822774, + "grad_norm": 2.757521867752075, + "learning_rate": 0.00017630506209302883, + "loss": 0.7356, + "step": 14238 + }, + { + "epoch": 0.988064672819374, + "grad_norm": 2.441770076751709, + "learning_rate": 0.0001763014304438181, + "loss": 0.8045, + "step": 14239 + }, + { + "epoch": 0.9881340642564708, + "grad_norm": 2.056110382080078, + "learning_rate": 0.00017629779855373264, + "loss": 0.8135, + "step": 14240 + }, + { + "epoch": 0.9882034556935674, + "grad_norm": 2.9346230030059814, + "learning_rate": 0.00017629416642278388, + "loss": 1.0889, + "step": 14241 + }, + { + "epoch": 0.9882728471306641, + "grad_norm": 2.5966978073120117, + "learning_rate": 0.00017629053405098327, + "loss": 1.1487, + "step": 14242 + }, + { + "epoch": 0.9883422385677607, + "grad_norm": 2.682279348373413, + "learning_rate": 0.00017628690143834232, + "loss": 1.1244, + "step": 14243 + }, + { + "epoch": 0.9884116300048574, + "grad_norm": 2.416064977645874, + "learning_rate": 0.00017628326858487247, + "loss": 0.8837, + "step": 14244 + }, + { + "epoch": 0.988481021441954, + "grad_norm": 2.2190349102020264, + "learning_rate": 0.0001762796354905852, + "loss": 0.7633, + "step": 14245 + }, + { + "epoch": 0.9885504128790508, + "grad_norm": 2.4336488246917725, + "learning_rate": 0.00017627600215549194, + "loss": 1.0777, + "step": 14246 + }, + { + "epoch": 0.9886198043161474, + "grad_norm": 2.1776797771453857, + "learning_rate": 0.0001762723685796042, + "loss": 0.6388, + "step": 14247 + }, + { + "epoch": 0.9886891957532441, + "grad_norm": 2.0746660232543945, + "learning_rate": 0.00017626873476293348, + "loss": 0.9497, + "step": 14248 + }, + { + "epoch": 0.9887585871903407, + "grad_norm": 1.81680428981781, + "learning_rate": 0.0001762651007054912, + "loss": 0.5939, + "step": 14249 + }, + { + "epoch": 0.9888279786274373, + "grad_norm": 1.9281036853790283, + "learning_rate": 0.0001762614664072888, + "loss": 0.8823, + "step": 14250 + }, + { + "epoch": 0.988897370064534, + "grad_norm": 2.9897055625915527, + "learning_rate": 0.0001762578318683378, + "loss": 1.1805, + "step": 14251 + }, + { + "epoch": 0.9889667615016307, + "grad_norm": 2.3476767539978027, + "learning_rate": 0.00017625419708864972, + "loss": 0.8285, + "step": 14252 + }, + { + "epoch": 0.9890361529387274, + "grad_norm": 2.382305383682251, + "learning_rate": 0.00017625056206823598, + "loss": 0.8806, + "step": 14253 + }, + { + "epoch": 0.989105544375824, + "grad_norm": 2.1381478309631348, + "learning_rate": 0.00017624692680710805, + "loss": 0.672, + "step": 14254 + }, + { + "epoch": 0.9891749358129207, + "grad_norm": 1.9952752590179443, + "learning_rate": 0.00017624329130527743, + "loss": 0.7245, + "step": 14255 + }, + { + "epoch": 0.9892443272500173, + "grad_norm": 2.177523612976074, + "learning_rate": 0.00017623965556275555, + "loss": 0.5807, + "step": 14256 + }, + { + "epoch": 0.989313718687114, + "grad_norm": 2.2139508724212646, + "learning_rate": 0.00017623601957955396, + "loss": 0.8721, + "step": 14257 + }, + { + "epoch": 0.9893831101242107, + "grad_norm": 2.6727235317230225, + "learning_rate": 0.00017623238335568407, + "loss": 0.7523, + "step": 14258 + }, + { + "epoch": 0.9894525015613074, + "grad_norm": 2.810324192047119, + "learning_rate": 0.0001762287468911574, + "loss": 0.809, + "step": 14259 + }, + { + "epoch": 0.989521892998404, + "grad_norm": 2.4271202087402344, + "learning_rate": 0.00017622511018598543, + "loss": 1.0321, + "step": 14260 + }, + { + "epoch": 0.9895912844355007, + "grad_norm": 2.098801851272583, + "learning_rate": 0.00017622147324017963, + "loss": 0.8443, + "step": 14261 + }, + { + "epoch": 0.9896606758725973, + "grad_norm": 1.7781933546066284, + "learning_rate": 0.00017621783605375145, + "loss": 0.6906, + "step": 14262 + }, + { + "epoch": 0.9897300673096939, + "grad_norm": 2.0578250885009766, + "learning_rate": 0.00017621419862671247, + "loss": 0.7507, + "step": 14263 + }, + { + "epoch": 0.9897994587467907, + "grad_norm": 2.2694568634033203, + "learning_rate": 0.00017621056095907405, + "loss": 0.9547, + "step": 14264 + }, + { + "epoch": 0.9898688501838873, + "grad_norm": 1.9868844747543335, + "learning_rate": 0.00017620692305084777, + "loss": 0.8424, + "step": 14265 + }, + { + "epoch": 0.989938241620984, + "grad_norm": 2.0208380222320557, + "learning_rate": 0.00017620328490204506, + "loss": 0.7252, + "step": 14266 + }, + { + "epoch": 0.9900076330580806, + "grad_norm": 2.1260740756988525, + "learning_rate": 0.0001761996465126774, + "loss": 0.8652, + "step": 14267 + }, + { + "epoch": 0.9900770244951773, + "grad_norm": 1.8033326864242554, + "learning_rate": 0.00017619600788275632, + "loss": 0.7873, + "step": 14268 + }, + { + "epoch": 0.9901464159322739, + "grad_norm": 2.447662591934204, + "learning_rate": 0.00017619236901229328, + "loss": 0.9262, + "step": 14269 + }, + { + "epoch": 0.9902158073693706, + "grad_norm": 1.7803980112075806, + "learning_rate": 0.00017618872990129978, + "loss": 0.6637, + "step": 14270 + }, + { + "epoch": 0.9902851988064673, + "grad_norm": 2.0185468196868896, + "learning_rate": 0.00017618509054978727, + "loss": 0.5888, + "step": 14271 + }, + { + "epoch": 0.990354590243564, + "grad_norm": 1.941312313079834, + "learning_rate": 0.0001761814509577673, + "loss": 0.9266, + "step": 14272 + }, + { + "epoch": 0.9904239816806606, + "grad_norm": 2.019179582595825, + "learning_rate": 0.00017617781112525132, + "loss": 0.891, + "step": 14273 + }, + { + "epoch": 0.9904933731177573, + "grad_norm": 1.797606348991394, + "learning_rate": 0.0001761741710522508, + "loss": 0.4894, + "step": 14274 + }, + { + "epoch": 0.9905627645548539, + "grad_norm": 1.8609848022460938, + "learning_rate": 0.00017617053073877732, + "loss": 0.8145, + "step": 14275 + }, + { + "epoch": 0.9906321559919506, + "grad_norm": 2.072763204574585, + "learning_rate": 0.00017616689018484224, + "loss": 0.9231, + "step": 14276 + }, + { + "epoch": 0.9907015474290473, + "grad_norm": 2.2466232776641846, + "learning_rate": 0.00017616324939045717, + "loss": 1.1442, + "step": 14277 + }, + { + "epoch": 0.9907709388661439, + "grad_norm": 2.2916953563690186, + "learning_rate": 0.00017615960835563356, + "loss": 0.7014, + "step": 14278 + }, + { + "epoch": 0.9908403303032406, + "grad_norm": 3.8655402660369873, + "learning_rate": 0.0001761559670803829, + "loss": 1.204, + "step": 14279 + }, + { + "epoch": 0.9909097217403372, + "grad_norm": 1.6862226724624634, + "learning_rate": 0.00017615232556471665, + "loss": 0.5801, + "step": 14280 + }, + { + "epoch": 0.9909791131774339, + "grad_norm": 2.191518783569336, + "learning_rate": 0.0001761486838086464, + "loss": 0.788, + "step": 14281 + }, + { + "epoch": 0.9910485046145305, + "grad_norm": 2.8662800788879395, + "learning_rate": 0.00017614504181218355, + "loss": 0.9313, + "step": 14282 + }, + { + "epoch": 0.9911178960516273, + "grad_norm": 1.839758038520813, + "learning_rate": 0.00017614139957533963, + "loss": 0.747, + "step": 14283 + }, + { + "epoch": 0.9911872874887239, + "grad_norm": 1.6377640962600708, + "learning_rate": 0.0001761377570981262, + "loss": 0.6662, + "step": 14284 + }, + { + "epoch": 0.9912566789258206, + "grad_norm": 2.407879114151001, + "learning_rate": 0.00017613411438055466, + "loss": 0.8293, + "step": 14285 + }, + { + "epoch": 0.9913260703629172, + "grad_norm": 1.9640058279037476, + "learning_rate": 0.00017613047142263656, + "loss": 0.7043, + "step": 14286 + }, + { + "epoch": 0.9913954618000139, + "grad_norm": 2.4092652797698975, + "learning_rate": 0.0001761268282243834, + "loss": 0.8943, + "step": 14287 + }, + { + "epoch": 0.9914648532371105, + "grad_norm": 1.9998317956924438, + "learning_rate": 0.00017612318478580667, + "loss": 0.8356, + "step": 14288 + }, + { + "epoch": 0.9915342446742073, + "grad_norm": 3.8176910877227783, + "learning_rate": 0.0001761195411069179, + "loss": 0.9151, + "step": 14289 + }, + { + "epoch": 0.9916036361113039, + "grad_norm": 2.2995827198028564, + "learning_rate": 0.00017611589718772853, + "loss": 0.7897, + "step": 14290 + }, + { + "epoch": 0.9916730275484005, + "grad_norm": 1.6504501104354858, + "learning_rate": 0.00017611225302825015, + "loss": 0.6081, + "step": 14291 + }, + { + "epoch": 0.9917424189854972, + "grad_norm": 2.112067937850952, + "learning_rate": 0.00017610860862849416, + "loss": 0.9258, + "step": 14292 + }, + { + "epoch": 0.9918118104225938, + "grad_norm": 2.2991576194763184, + "learning_rate": 0.00017610496398847217, + "loss": 0.766, + "step": 14293 + }, + { + "epoch": 0.9918812018596905, + "grad_norm": 2.224916934967041, + "learning_rate": 0.00017610131910819564, + "loss": 0.8241, + "step": 14294 + }, + { + "epoch": 0.9919505932967871, + "grad_norm": 2.4908266067504883, + "learning_rate": 0.00017609767398767606, + "loss": 0.9007, + "step": 14295 + }, + { + "epoch": 0.9920199847338839, + "grad_norm": 2.8542826175689697, + "learning_rate": 0.00017609402862692495, + "loss": 0.7363, + "step": 14296 + }, + { + "epoch": 0.9920893761709805, + "grad_norm": 2.3064987659454346, + "learning_rate": 0.00017609038302595378, + "loss": 0.9409, + "step": 14297 + }, + { + "epoch": 0.9921587676080772, + "grad_norm": 2.841993808746338, + "learning_rate": 0.00017608673718477418, + "loss": 0.7847, + "step": 14298 + }, + { + "epoch": 0.9922281590451738, + "grad_norm": 2.513624429702759, + "learning_rate": 0.00017608309110339753, + "loss": 1.0138, + "step": 14299 + }, + { + "epoch": 0.9922975504822705, + "grad_norm": 1.8556822538375854, + "learning_rate": 0.0001760794447818354, + "loss": 0.7092, + "step": 14300 + }, + { + "epoch": 0.9923669419193671, + "grad_norm": 1.5918498039245605, + "learning_rate": 0.0001760757982200993, + "loss": 0.6347, + "step": 14301 + }, + { + "epoch": 0.9924363333564639, + "grad_norm": 2.269240617752075, + "learning_rate": 0.0001760721514182007, + "loss": 0.9399, + "step": 14302 + }, + { + "epoch": 0.9925057247935605, + "grad_norm": 1.837280511856079, + "learning_rate": 0.00017606850437615118, + "loss": 0.7017, + "step": 14303 + }, + { + "epoch": 0.9925751162306571, + "grad_norm": 2.870358943939209, + "learning_rate": 0.00017606485709396218, + "loss": 0.8953, + "step": 14304 + }, + { + "epoch": 0.9926445076677538, + "grad_norm": 2.6626791954040527, + "learning_rate": 0.00017606120957164528, + "loss": 0.86, + "step": 14305 + }, + { + "epoch": 0.9927138991048504, + "grad_norm": 1.607809066772461, + "learning_rate": 0.00017605756180921197, + "loss": 0.4528, + "step": 14306 + }, + { + "epoch": 0.9927832905419471, + "grad_norm": 1.7528055906295776, + "learning_rate": 0.00017605391380667374, + "loss": 0.6964, + "step": 14307 + }, + { + "epoch": 0.9928526819790437, + "grad_norm": 1.7796988487243652, + "learning_rate": 0.00017605026556404213, + "loss": 0.9003, + "step": 14308 + }, + { + "epoch": 0.9929220734161405, + "grad_norm": 2.6299517154693604, + "learning_rate": 0.00017604661708132868, + "loss": 0.834, + "step": 14309 + }, + { + "epoch": 0.9929914648532371, + "grad_norm": 2.7975077629089355, + "learning_rate": 0.00017604296835854486, + "loss": 0.7768, + "step": 14310 + }, + { + "epoch": 0.9930608562903338, + "grad_norm": 2.0046212673187256, + "learning_rate": 0.00017603931939570223, + "loss": 0.7469, + "step": 14311 + }, + { + "epoch": 0.9931302477274304, + "grad_norm": 1.9754809141159058, + "learning_rate": 0.00017603567019281228, + "loss": 0.6987, + "step": 14312 + }, + { + "epoch": 0.9931996391645271, + "grad_norm": 2.467329502105713, + "learning_rate": 0.0001760320207498865, + "loss": 0.8277, + "step": 14313 + }, + { + "epoch": 0.9932690306016237, + "grad_norm": 2.1333858966827393, + "learning_rate": 0.0001760283710669365, + "loss": 0.8505, + "step": 14314 + }, + { + "epoch": 0.9933384220387205, + "grad_norm": 2.078214645385742, + "learning_rate": 0.00017602472114397374, + "loss": 0.6497, + "step": 14315 + }, + { + "epoch": 0.9934078134758171, + "grad_norm": 2.3373982906341553, + "learning_rate": 0.00017602107098100978, + "loss": 0.7129, + "step": 14316 + }, + { + "epoch": 0.9934772049129137, + "grad_norm": 1.8013325929641724, + "learning_rate": 0.00017601742057805607, + "loss": 0.6727, + "step": 14317 + }, + { + "epoch": 0.9935465963500104, + "grad_norm": 1.7643879652023315, + "learning_rate": 0.0001760137699351242, + "loss": 0.6326, + "step": 14318 + }, + { + "epoch": 0.993615987787107, + "grad_norm": 2.545196533203125, + "learning_rate": 0.00017601011905222572, + "loss": 0.7243, + "step": 14319 + }, + { + "epoch": 0.9936853792242037, + "grad_norm": 2.3058431148529053, + "learning_rate": 0.00017600646792937207, + "loss": 0.7564, + "step": 14320 + }, + { + "epoch": 0.9937547706613004, + "grad_norm": 2.1843910217285156, + "learning_rate": 0.00017600281656657486, + "loss": 0.8638, + "step": 14321 + }, + { + "epoch": 0.9938241620983971, + "grad_norm": 2.1182799339294434, + "learning_rate": 0.00017599916496384551, + "loss": 0.6818, + "step": 14322 + }, + { + "epoch": 0.9938935535354937, + "grad_norm": 2.326289653778076, + "learning_rate": 0.00017599551312119568, + "loss": 0.7964, + "step": 14323 + }, + { + "epoch": 0.9939629449725904, + "grad_norm": 2.859971046447754, + "learning_rate": 0.0001759918610386368, + "loss": 0.7196, + "step": 14324 + }, + { + "epoch": 0.994032336409687, + "grad_norm": 2.451517343521118, + "learning_rate": 0.00017598820871618043, + "loss": 0.7674, + "step": 14325 + }, + { + "epoch": 0.9941017278467837, + "grad_norm": 2.007509469985962, + "learning_rate": 0.0001759845561538381, + "loss": 0.7796, + "step": 14326 + }, + { + "epoch": 0.9941711192838804, + "grad_norm": 2.3677871227264404, + "learning_rate": 0.00017598090335162135, + "loss": 0.8784, + "step": 14327 + }, + { + "epoch": 0.9942405107209771, + "grad_norm": 2.367112159729004, + "learning_rate": 0.00017597725030954172, + "loss": 0.9139, + "step": 14328 + }, + { + "epoch": 0.9943099021580737, + "grad_norm": 2.089413642883301, + "learning_rate": 0.0001759735970276107, + "loss": 0.7901, + "step": 14329 + }, + { + "epoch": 0.9943792935951704, + "grad_norm": 2.7475392818450928, + "learning_rate": 0.00017596994350583986, + "loss": 1.0505, + "step": 14330 + }, + { + "epoch": 0.994448685032267, + "grad_norm": 2.0245182514190674, + "learning_rate": 0.00017596628974424073, + "loss": 0.8467, + "step": 14331 + }, + { + "epoch": 0.9945180764693636, + "grad_norm": 2.2156286239624023, + "learning_rate": 0.00017596263574282478, + "loss": 0.9611, + "step": 14332 + }, + { + "epoch": 0.9945874679064604, + "grad_norm": 1.8769160509109497, + "learning_rate": 0.00017595898150160366, + "loss": 0.7394, + "step": 14333 + }, + { + "epoch": 0.994656859343557, + "grad_norm": 2.2860097885131836, + "learning_rate": 0.00017595532702058882, + "loss": 0.7339, + "step": 14334 + }, + { + "epoch": 0.9947262507806537, + "grad_norm": 2.1802713871002197, + "learning_rate": 0.00017595167229979186, + "loss": 0.906, + "step": 14335 + }, + { + "epoch": 0.9947956422177503, + "grad_norm": 1.9794056415557861, + "learning_rate": 0.00017594801733922428, + "loss": 0.7449, + "step": 14336 + }, + { + "epoch": 0.994865033654847, + "grad_norm": 1.9119246006011963, + "learning_rate": 0.0001759443621388976, + "loss": 0.9845, + "step": 14337 + }, + { + "epoch": 0.9949344250919436, + "grad_norm": 1.706827163696289, + "learning_rate": 0.00017594070669882337, + "loss": 0.7119, + "step": 14338 + }, + { + "epoch": 0.9950038165290404, + "grad_norm": 1.7186861038208008, + "learning_rate": 0.00017593705101901316, + "loss": 0.5854, + "step": 14339 + }, + { + "epoch": 0.995073207966137, + "grad_norm": 1.8634063005447388, + "learning_rate": 0.0001759333950994785, + "loss": 0.9039, + "step": 14340 + }, + { + "epoch": 0.9951425994032337, + "grad_norm": 1.663009762763977, + "learning_rate": 0.00017592973894023088, + "loss": 0.6522, + "step": 14341 + }, + { + "epoch": 0.9952119908403303, + "grad_norm": 2.089362144470215, + "learning_rate": 0.00017592608254128194, + "loss": 0.7838, + "step": 14342 + }, + { + "epoch": 0.995281382277427, + "grad_norm": 1.9246821403503418, + "learning_rate": 0.0001759224259026431, + "loss": 0.8297, + "step": 14343 + }, + { + "epoch": 0.9953507737145236, + "grad_norm": 2.2041547298431396, + "learning_rate": 0.00017591876902432602, + "loss": 0.7739, + "step": 14344 + }, + { + "epoch": 0.9954201651516202, + "grad_norm": 2.069272041320801, + "learning_rate": 0.0001759151119063422, + "loss": 0.7806, + "step": 14345 + }, + { + "epoch": 0.995489556588717, + "grad_norm": 2.077373743057251, + "learning_rate": 0.00017591145454870315, + "loss": 0.8443, + "step": 14346 + }, + { + "epoch": 0.9955589480258136, + "grad_norm": 1.9572913646697998, + "learning_rate": 0.00017590779695142046, + "loss": 0.882, + "step": 14347 + }, + { + "epoch": 0.9956283394629103, + "grad_norm": 1.6963427066802979, + "learning_rate": 0.00017590413911450564, + "loss": 0.6631, + "step": 14348 + }, + { + "epoch": 0.9956977309000069, + "grad_norm": 1.745176911354065, + "learning_rate": 0.0001759004810379703, + "loss": 0.6669, + "step": 14349 + }, + { + "epoch": 0.9957671223371036, + "grad_norm": 1.5159778594970703, + "learning_rate": 0.00017589682272182593, + "loss": 0.6966, + "step": 14350 + }, + { + "epoch": 0.9958365137742002, + "grad_norm": 2.1255035400390625, + "learning_rate": 0.0001758931641660841, + "loss": 0.8231, + "step": 14351 + }, + { + "epoch": 0.995905905211297, + "grad_norm": 2.5757343769073486, + "learning_rate": 0.00017588950537075632, + "loss": 0.8866, + "step": 14352 + }, + { + "epoch": 0.9959752966483936, + "grad_norm": 2.4479565620422363, + "learning_rate": 0.00017588584633585418, + "loss": 0.7456, + "step": 14353 + }, + { + "epoch": 0.9960446880854903, + "grad_norm": 1.525644063949585, + "learning_rate": 0.00017588218706138928, + "loss": 0.5999, + "step": 14354 + }, + { + "epoch": 0.9961140795225869, + "grad_norm": 1.8907804489135742, + "learning_rate": 0.00017587852754737308, + "loss": 0.7265, + "step": 14355 + }, + { + "epoch": 0.9961834709596836, + "grad_norm": 1.725273609161377, + "learning_rate": 0.00017587486779381718, + "loss": 0.7689, + "step": 14356 + }, + { + "epoch": 0.9962528623967802, + "grad_norm": 2.0328369140625, + "learning_rate": 0.00017587120780073313, + "loss": 0.898, + "step": 14357 + }, + { + "epoch": 0.9963222538338768, + "grad_norm": 2.2225661277770996, + "learning_rate": 0.00017586754756813248, + "loss": 0.8658, + "step": 14358 + }, + { + "epoch": 0.9963916452709736, + "grad_norm": 1.4900050163269043, + "learning_rate": 0.0001758638870960268, + "loss": 0.6517, + "step": 14359 + }, + { + "epoch": 0.9964610367080702, + "grad_norm": 2.683424711227417, + "learning_rate": 0.0001758602263844276, + "loss": 0.7159, + "step": 14360 + }, + { + "epoch": 0.9965304281451669, + "grad_norm": 2.7011919021606445, + "learning_rate": 0.00017585656543334647, + "loss": 0.8409, + "step": 14361 + }, + { + "epoch": 0.9965998195822635, + "grad_norm": 1.7737793922424316, + "learning_rate": 0.000175852904242795, + "loss": 0.8281, + "step": 14362 + }, + { + "epoch": 0.9966692110193602, + "grad_norm": 2.0366291999816895, + "learning_rate": 0.00017584924281278467, + "loss": 0.8422, + "step": 14363 + }, + { + "epoch": 0.9967386024564568, + "grad_norm": 1.5615999698638916, + "learning_rate": 0.00017584558114332709, + "loss": 0.6997, + "step": 14364 + }, + { + "epoch": 0.9968079938935536, + "grad_norm": 1.8888897895812988, + "learning_rate": 0.0001758419192344338, + "loss": 0.6386, + "step": 14365 + }, + { + "epoch": 0.9968773853306502, + "grad_norm": 2.2206642627716064, + "learning_rate": 0.00017583825708611639, + "loss": 0.7082, + "step": 14366 + }, + { + "epoch": 0.9969467767677469, + "grad_norm": 2.5884079933166504, + "learning_rate": 0.0001758345946983864, + "loss": 0.7746, + "step": 14367 + }, + { + "epoch": 0.9970161682048435, + "grad_norm": 1.994383454322815, + "learning_rate": 0.0001758309320712554, + "loss": 0.8119, + "step": 14368 + }, + { + "epoch": 0.9970855596419402, + "grad_norm": 2.1714110374450684, + "learning_rate": 0.00017582726920473493, + "loss": 0.6235, + "step": 14369 + }, + { + "epoch": 0.9971549510790368, + "grad_norm": 2.351522922515869, + "learning_rate": 0.00017582360609883657, + "loss": 0.8314, + "step": 14370 + }, + { + "epoch": 0.9972243425161335, + "grad_norm": 1.9435144662857056, + "learning_rate": 0.00017581994275357187, + "loss": 0.7227, + "step": 14371 + }, + { + "epoch": 0.9972937339532302, + "grad_norm": 2.3959805965423584, + "learning_rate": 0.0001758162791689524, + "loss": 0.8848, + "step": 14372 + }, + { + "epoch": 0.9973631253903268, + "grad_norm": 2.250570297241211, + "learning_rate": 0.0001758126153449898, + "loss": 0.9019, + "step": 14373 + }, + { + "epoch": 0.9974325168274235, + "grad_norm": 1.7596410512924194, + "learning_rate": 0.00017580895128169551, + "loss": 0.7972, + "step": 14374 + }, + { + "epoch": 0.9975019082645201, + "grad_norm": 3.026216983795166, + "learning_rate": 0.0001758052869790812, + "loss": 0.7749, + "step": 14375 + }, + { + "epoch": 0.9975712997016168, + "grad_norm": 2.241203546524048, + "learning_rate": 0.00017580162243715835, + "loss": 0.5981, + "step": 14376 + }, + { + "epoch": 0.9976406911387135, + "grad_norm": 2.0277209281921387, + "learning_rate": 0.0001757979576559386, + "loss": 0.8236, + "step": 14377 + }, + { + "epoch": 0.9977100825758102, + "grad_norm": 1.826964259147644, + "learning_rate": 0.00017579429263543347, + "loss": 0.681, + "step": 14378 + }, + { + "epoch": 0.9977794740129068, + "grad_norm": 1.5577222108840942, + "learning_rate": 0.00017579062737565457, + "loss": 0.461, + "step": 14379 + }, + { + "epoch": 0.9978488654500035, + "grad_norm": 2.28053879737854, + "learning_rate": 0.00017578696187661342, + "loss": 1.0613, + "step": 14380 + }, + { + "epoch": 0.9979182568871001, + "grad_norm": 1.7129862308502197, + "learning_rate": 0.00017578329613832168, + "loss": 0.7352, + "step": 14381 + }, + { + "epoch": 0.9979876483241968, + "grad_norm": 2.3278849124908447, + "learning_rate": 0.00017577963016079083, + "loss": 0.8024, + "step": 14382 + }, + { + "epoch": 0.9980570397612935, + "grad_norm": 2.457320213317871, + "learning_rate": 0.00017577596394403248, + "loss": 0.6493, + "step": 14383 + }, + { + "epoch": 0.9981264311983902, + "grad_norm": 4.449375629425049, + "learning_rate": 0.0001757722974880582, + "loss": 1.0533, + "step": 14384 + }, + { + "epoch": 0.9981958226354868, + "grad_norm": 2.124559164047241, + "learning_rate": 0.0001757686307928796, + "loss": 0.7521, + "step": 14385 + }, + { + "epoch": 0.9982652140725834, + "grad_norm": 1.9552198648452759, + "learning_rate": 0.0001757649638585082, + "loss": 0.7173, + "step": 14386 + }, + { + "epoch": 0.9983346055096801, + "grad_norm": 2.1898488998413086, + "learning_rate": 0.0001757612966849556, + "loss": 0.8313, + "step": 14387 + }, + { + "epoch": 0.9984039969467767, + "grad_norm": 2.0185534954071045, + "learning_rate": 0.00017575762927223338, + "loss": 0.7937, + "step": 14388 + }, + { + "epoch": 0.9984733883838735, + "grad_norm": 1.6462832689285278, + "learning_rate": 0.0001757539616203531, + "loss": 0.6861, + "step": 14389 + }, + { + "epoch": 0.9985427798209701, + "grad_norm": 3.3566033840179443, + "learning_rate": 0.00017575029372932636, + "loss": 0.8301, + "step": 14390 + }, + { + "epoch": 0.9986121712580668, + "grad_norm": 1.99271559715271, + "learning_rate": 0.00017574662559916474, + "loss": 0.6422, + "step": 14391 + }, + { + "epoch": 0.9986815626951634, + "grad_norm": 1.7099164724349976, + "learning_rate": 0.00017574295722987978, + "loss": 0.7862, + "step": 14392 + }, + { + "epoch": 0.9987509541322601, + "grad_norm": 1.8273873329162598, + "learning_rate": 0.0001757392886214831, + "loss": 0.7423, + "step": 14393 + }, + { + "epoch": 0.9988203455693567, + "grad_norm": 3.5074660778045654, + "learning_rate": 0.00017573561977398633, + "loss": 0.8148, + "step": 14394 + }, + { + "epoch": 0.9988897370064534, + "grad_norm": 1.7900255918502808, + "learning_rate": 0.00017573195068740097, + "loss": 0.5533, + "step": 14395 + }, + { + "epoch": 0.9989591284435501, + "grad_norm": 2.791461944580078, + "learning_rate": 0.0001757282813617386, + "loss": 0.8686, + "step": 14396 + }, + { + "epoch": 0.9990285198806468, + "grad_norm": 2.5270440578460693, + "learning_rate": 0.00017572461179701085, + "loss": 0.8361, + "step": 14397 + }, + { + "epoch": 0.9990979113177434, + "grad_norm": 1.8924082517623901, + "learning_rate": 0.00017572094199322927, + "loss": 0.8675, + "step": 14398 + }, + { + "epoch": 0.99916730275484, + "grad_norm": 2.3922476768493652, + "learning_rate": 0.00017571727195040548, + "loss": 0.9179, + "step": 14399 + }, + { + "epoch": 0.9992366941919367, + "grad_norm": 1.5868417024612427, + "learning_rate": 0.00017571360166855103, + "loss": 0.5677, + "step": 14400 + }, + { + "epoch": 0.9993060856290333, + "grad_norm": 2.214553117752075, + "learning_rate": 0.00017570993114767754, + "loss": 0.9353, + "step": 14401 + }, + { + "epoch": 0.9993754770661301, + "grad_norm": 2.8221731185913086, + "learning_rate": 0.00017570626038779659, + "loss": 0.8558, + "step": 14402 + }, + { + "epoch": 0.9994448685032267, + "grad_norm": 2.966601610183716, + "learning_rate": 0.00017570258938891974, + "loss": 0.873, + "step": 14403 + }, + { + "epoch": 0.9995142599403234, + "grad_norm": 1.7366551160812378, + "learning_rate": 0.0001756989181510586, + "loss": 0.6747, + "step": 14404 + }, + { + "epoch": 0.99958365137742, + "grad_norm": 2.2091126441955566, + "learning_rate": 0.00017569524667422477, + "loss": 0.7086, + "step": 14405 + }, + { + "epoch": 0.9996530428145167, + "grad_norm": 2.244621992111206, + "learning_rate": 0.0001756915749584298, + "loss": 0.7216, + "step": 14406 + }, + { + "epoch": 0.9997224342516133, + "grad_norm": 1.7442823648452759, + "learning_rate": 0.00017568790300368536, + "loss": 0.7259, + "step": 14407 + }, + { + "epoch": 0.9997918256887101, + "grad_norm": 2.748220920562744, + "learning_rate": 0.00017568423081000293, + "loss": 0.9034, + "step": 14408 + }, + { + "epoch": 0.9998612171258067, + "grad_norm": 1.9858314990997314, + "learning_rate": 0.0001756805583773942, + "loss": 0.863, + "step": 14409 + }, + { + "epoch": 0.9999306085629034, + "grad_norm": 1.675145149230957, + "learning_rate": 0.00017567688570587072, + "loss": 0.8152, + "step": 14410 + }, + { + "epoch": 1.0, + "grad_norm": 2.3182125091552734, + "learning_rate": 0.0001756732127954441, + "loss": 0.776, + "step": 14411 + }, + { + "epoch": 1.0000693914370966, + "grad_norm": 2.4057953357696533, + "learning_rate": 0.00017566953964612592, + "loss": 0.7485, + "step": 14412 + }, + { + "epoch": 1.0001387828741932, + "grad_norm": 2.0094664096832275, + "learning_rate": 0.0001756658662579278, + "loss": 0.7121, + "step": 14413 + }, + { + "epoch": 1.00020817431129, + "grad_norm": 2.040484666824341, + "learning_rate": 0.0001756621926308613, + "loss": 0.9157, + "step": 14414 + }, + { + "epoch": 1.0002775657483867, + "grad_norm": 1.9691367149353027, + "learning_rate": 0.00017565851876493805, + "loss": 0.7378, + "step": 14415 + }, + { + "epoch": 1.0003469571854833, + "grad_norm": 1.7330613136291504, + "learning_rate": 0.0001756548446601696, + "loss": 0.7384, + "step": 14416 + }, + { + "epoch": 1.0004163486225799, + "grad_norm": 1.6530100107192993, + "learning_rate": 0.00017565117031656763, + "loss": 0.6428, + "step": 14417 + }, + { + "epoch": 1.0004857400596767, + "grad_norm": 2.020174503326416, + "learning_rate": 0.00017564749573414367, + "loss": 0.8353, + "step": 14418 + }, + { + "epoch": 1.0005551314967733, + "grad_norm": 2.3327043056488037, + "learning_rate": 0.00017564382091290933, + "loss": 0.6591, + "step": 14419 + }, + { + "epoch": 1.00062452293387, + "grad_norm": 1.5081731081008911, + "learning_rate": 0.00017564014585287624, + "loss": 0.5475, + "step": 14420 + }, + { + "epoch": 1.0006939143709666, + "grad_norm": 2.1254067420959473, + "learning_rate": 0.000175636470554056, + "loss": 0.8729, + "step": 14421 + }, + { + "epoch": 1.0007633058080634, + "grad_norm": 1.808518648147583, + "learning_rate": 0.0001756327950164602, + "loss": 0.6151, + "step": 14422 + }, + { + "epoch": 1.00083269724516, + "grad_norm": 1.561829924583435, + "learning_rate": 0.00017562911924010041, + "loss": 0.6863, + "step": 14423 + }, + { + "epoch": 1.0009020886822566, + "grad_norm": 1.7916663885116577, + "learning_rate": 0.00017562544322498826, + "loss": 0.5081, + "step": 14424 + }, + { + "epoch": 1.0009714801193532, + "grad_norm": 3.2363460063934326, + "learning_rate": 0.00017562176697113538, + "loss": 0.7765, + "step": 14425 + }, + { + "epoch": 1.0010408715564498, + "grad_norm": 1.6263220310211182, + "learning_rate": 0.00017561809047855336, + "loss": 0.5766, + "step": 14426 + }, + { + "epoch": 1.0011102629935467, + "grad_norm": 3.5224077701568604, + "learning_rate": 0.0001756144137472538, + "loss": 0.6722, + "step": 14427 + }, + { + "epoch": 1.0011796544306433, + "grad_norm": 2.1318976879119873, + "learning_rate": 0.0001756107367772483, + "loss": 0.6258, + "step": 14428 + }, + { + "epoch": 1.0012490458677399, + "grad_norm": 1.7212997674942017, + "learning_rate": 0.00017560705956854849, + "loss": 0.5826, + "step": 14429 + }, + { + "epoch": 1.0013184373048365, + "grad_norm": 1.7573730945587158, + "learning_rate": 0.00017560338212116593, + "loss": 0.6816, + "step": 14430 + }, + { + "epoch": 1.0013878287419333, + "grad_norm": 2.6081206798553467, + "learning_rate": 0.0001755997044351123, + "loss": 0.7257, + "step": 14431 + }, + { + "epoch": 1.00145722017903, + "grad_norm": 3.0668933391571045, + "learning_rate": 0.00017559602651039916, + "loss": 0.8738, + "step": 14432 + }, + { + "epoch": 1.0015266116161265, + "grad_norm": 2.2274603843688965, + "learning_rate": 0.00017559234834703814, + "loss": 0.6848, + "step": 14433 + }, + { + "epoch": 1.0015960030532232, + "grad_norm": 1.8161355257034302, + "learning_rate": 0.00017558866994504084, + "loss": 0.6156, + "step": 14434 + }, + { + "epoch": 1.00166539449032, + "grad_norm": 1.6216360330581665, + "learning_rate": 0.00017558499130441887, + "loss": 0.724, + "step": 14435 + }, + { + "epoch": 1.0017347859274166, + "grad_norm": 2.620194673538208, + "learning_rate": 0.00017558131242518388, + "loss": 0.8692, + "step": 14436 + }, + { + "epoch": 1.0018041773645132, + "grad_norm": 1.9195328950881958, + "learning_rate": 0.00017557763330734743, + "loss": 0.7377, + "step": 14437 + }, + { + "epoch": 1.0018735688016098, + "grad_norm": 2.626918077468872, + "learning_rate": 0.00017557395395092118, + "loss": 0.8473, + "step": 14438 + }, + { + "epoch": 1.0019429602387064, + "grad_norm": 2.067689895629883, + "learning_rate": 0.0001755702743559167, + "loss": 0.6603, + "step": 14439 + }, + { + "epoch": 1.0020123516758033, + "grad_norm": 1.9964183568954468, + "learning_rate": 0.00017556659452234565, + "loss": 0.7174, + "step": 14440 + }, + { + "epoch": 1.0020817431128999, + "grad_norm": 6.635228157043457, + "learning_rate": 0.00017556291445021963, + "loss": 0.5788, + "step": 14441 + }, + { + "epoch": 1.0021511345499965, + "grad_norm": 1.7975804805755615, + "learning_rate": 0.00017555923413955027, + "loss": 0.595, + "step": 14442 + }, + { + "epoch": 1.002220525987093, + "grad_norm": 2.1327054500579834, + "learning_rate": 0.00017555555359034913, + "loss": 0.5637, + "step": 14443 + }, + { + "epoch": 1.00228991742419, + "grad_norm": 1.8520902395248413, + "learning_rate": 0.00017555187280262791, + "loss": 0.5503, + "step": 14444 + }, + { + "epoch": 1.0023593088612865, + "grad_norm": 2.5264623165130615, + "learning_rate": 0.00017554819177639818, + "loss": 0.6706, + "step": 14445 + }, + { + "epoch": 1.0024287002983832, + "grad_norm": 1.7810508012771606, + "learning_rate": 0.0001755445105116716, + "loss": 0.5757, + "step": 14446 + }, + { + "epoch": 1.0024980917354798, + "grad_norm": 1.660447359085083, + "learning_rate": 0.00017554082900845974, + "loss": 0.6264, + "step": 14447 + }, + { + "epoch": 1.0025674831725766, + "grad_norm": 2.15138840675354, + "learning_rate": 0.00017553714726677424, + "loss": 0.5442, + "step": 14448 + }, + { + "epoch": 1.0026368746096732, + "grad_norm": 2.025178909301758, + "learning_rate": 0.00017553346528662673, + "loss": 0.554, + "step": 14449 + }, + { + "epoch": 1.0027062660467698, + "grad_norm": 2.1251168251037598, + "learning_rate": 0.00017552978306802886, + "loss": 0.6517, + "step": 14450 + }, + { + "epoch": 1.0027756574838664, + "grad_norm": 1.968063473701477, + "learning_rate": 0.0001755261006109922, + "loss": 0.7104, + "step": 14451 + }, + { + "epoch": 1.002845048920963, + "grad_norm": 2.511136054992676, + "learning_rate": 0.00017552241791552841, + "loss": 0.678, + "step": 14452 + }, + { + "epoch": 1.0029144403580599, + "grad_norm": 1.8046202659606934, + "learning_rate": 0.00017551873498164913, + "loss": 0.5909, + "step": 14453 + }, + { + "epoch": 1.0029838317951565, + "grad_norm": 3.1712799072265625, + "learning_rate": 0.00017551505180936594, + "loss": 1.0016, + "step": 14454 + }, + { + "epoch": 1.003053223232253, + "grad_norm": 2.3413476943969727, + "learning_rate": 0.0001755113683986905, + "loss": 0.8312, + "step": 14455 + }, + { + "epoch": 1.0031226146693497, + "grad_norm": 1.8872169256210327, + "learning_rate": 0.00017550768474963445, + "loss": 0.7006, + "step": 14456 + }, + { + "epoch": 1.0031920061064465, + "grad_norm": 1.5644716024398804, + "learning_rate": 0.00017550400086220938, + "loss": 0.4898, + "step": 14457 + }, + { + "epoch": 1.0032613975435432, + "grad_norm": 1.998327612876892, + "learning_rate": 0.00017550031673642692, + "loss": 0.8174, + "step": 14458 + }, + { + "epoch": 1.0033307889806398, + "grad_norm": 1.882520079612732, + "learning_rate": 0.00017549663237229876, + "loss": 0.6442, + "step": 14459 + }, + { + "epoch": 1.0034001804177364, + "grad_norm": 1.8010002374649048, + "learning_rate": 0.00017549294776983648, + "loss": 0.5774, + "step": 14460 + }, + { + "epoch": 1.0034695718548332, + "grad_norm": 1.990260362625122, + "learning_rate": 0.0001754892629290517, + "loss": 0.8051, + "step": 14461 + }, + { + "epoch": 1.0035389632919298, + "grad_norm": 1.8769185543060303, + "learning_rate": 0.0001754855778499561, + "loss": 0.7741, + "step": 14462 + }, + { + "epoch": 1.0036083547290264, + "grad_norm": 2.355213165283203, + "learning_rate": 0.00017548189253256127, + "loss": 0.5461, + "step": 14463 + }, + { + "epoch": 1.003677746166123, + "grad_norm": 2.3793609142303467, + "learning_rate": 0.00017547820697687888, + "loss": 0.8172, + "step": 14464 + }, + { + "epoch": 1.0037471376032197, + "grad_norm": 2.1449313163757324, + "learning_rate": 0.0001754745211829205, + "loss": 0.7486, + "step": 14465 + }, + { + "epoch": 1.0038165290403165, + "grad_norm": 1.8585619926452637, + "learning_rate": 0.00017547083515069786, + "loss": 0.6518, + "step": 14466 + }, + { + "epoch": 1.003885920477413, + "grad_norm": 3.3291378021240234, + "learning_rate": 0.00017546714888022255, + "loss": 0.5549, + "step": 14467 + }, + { + "epoch": 1.0039553119145097, + "grad_norm": 2.7744290828704834, + "learning_rate": 0.0001754634623715062, + "loss": 0.5966, + "step": 14468 + }, + { + "epoch": 1.0040247033516063, + "grad_norm": 1.7588036060333252, + "learning_rate": 0.00017545977562456044, + "loss": 0.5421, + "step": 14469 + }, + { + "epoch": 1.0040940947887032, + "grad_norm": 1.9540430307388306, + "learning_rate": 0.00017545608863939695, + "loss": 0.7506, + "step": 14470 + }, + { + "epoch": 1.0041634862257998, + "grad_norm": 1.5891884565353394, + "learning_rate": 0.00017545240141602736, + "loss": 0.5414, + "step": 14471 + }, + { + "epoch": 1.0042328776628964, + "grad_norm": 1.6922391653060913, + "learning_rate": 0.00017544871395446324, + "loss": 0.559, + "step": 14472 + }, + { + "epoch": 1.004302269099993, + "grad_norm": 2.744600534439087, + "learning_rate": 0.0001754450262547163, + "loss": 0.7142, + "step": 14473 + }, + { + "epoch": 1.0043716605370898, + "grad_norm": 2.055880308151245, + "learning_rate": 0.0001754413383167982, + "loss": 0.7424, + "step": 14474 + }, + { + "epoch": 1.0044410519741864, + "grad_norm": 1.9535562992095947, + "learning_rate": 0.0001754376501407205, + "loss": 0.5089, + "step": 14475 + }, + { + "epoch": 1.004510443411283, + "grad_norm": 1.8268256187438965, + "learning_rate": 0.00017543396172649492, + "loss": 0.4519, + "step": 14476 + }, + { + "epoch": 1.0045798348483796, + "grad_norm": 2.80241322517395, + "learning_rate": 0.00017543027307413304, + "loss": 0.5813, + "step": 14477 + }, + { + "epoch": 1.0046492262854765, + "grad_norm": 2.039778470993042, + "learning_rate": 0.00017542658418364656, + "loss": 0.6429, + "step": 14478 + }, + { + "epoch": 1.004718617722573, + "grad_norm": 2.535327196121216, + "learning_rate": 0.00017542289505504712, + "loss": 0.8284, + "step": 14479 + }, + { + "epoch": 1.0047880091596697, + "grad_norm": 1.5476484298706055, + "learning_rate": 0.00017541920568834636, + "loss": 0.4909, + "step": 14480 + }, + { + "epoch": 1.0048574005967663, + "grad_norm": 2.0911478996276855, + "learning_rate": 0.00017541551608355585, + "loss": 0.5864, + "step": 14481 + }, + { + "epoch": 1.004926792033863, + "grad_norm": 1.6176605224609375, + "learning_rate": 0.00017541182624068738, + "loss": 0.5632, + "step": 14482 + }, + { + "epoch": 1.0049961834709598, + "grad_norm": 2.0455996990203857, + "learning_rate": 0.00017540813615975253, + "loss": 0.6351, + "step": 14483 + }, + { + "epoch": 1.0050655749080564, + "grad_norm": 2.0542304515838623, + "learning_rate": 0.00017540444584076287, + "loss": 0.7687, + "step": 14484 + }, + { + "epoch": 1.005134966345153, + "grad_norm": 2.3462464809417725, + "learning_rate": 0.00017540075528373018, + "loss": 0.7175, + "step": 14485 + }, + { + "epoch": 1.0052043577822496, + "grad_norm": 2.161273956298828, + "learning_rate": 0.00017539706448866605, + "loss": 0.4773, + "step": 14486 + }, + { + "epoch": 1.0052737492193464, + "grad_norm": 1.4305877685546875, + "learning_rate": 0.0001753933734555821, + "loss": 0.5407, + "step": 14487 + }, + { + "epoch": 1.005343140656443, + "grad_norm": 1.9188553094863892, + "learning_rate": 0.00017538968218449005, + "loss": 0.5181, + "step": 14488 + }, + { + "epoch": 1.0054125320935396, + "grad_norm": 2.970547676086426, + "learning_rate": 0.00017538599067540148, + "loss": 0.8144, + "step": 14489 + }, + { + "epoch": 1.0054819235306363, + "grad_norm": 1.8916864395141602, + "learning_rate": 0.00017538229892832813, + "loss": 0.7408, + "step": 14490 + }, + { + "epoch": 1.005551314967733, + "grad_norm": 2.3655078411102295, + "learning_rate": 0.0001753786069432816, + "loss": 0.8785, + "step": 14491 + }, + { + "epoch": 1.0056207064048297, + "grad_norm": 1.9897828102111816, + "learning_rate": 0.00017537491472027354, + "loss": 0.626, + "step": 14492 + }, + { + "epoch": 1.0056900978419263, + "grad_norm": 1.7034964561462402, + "learning_rate": 0.00017537122225931562, + "loss": 0.509, + "step": 14493 + }, + { + "epoch": 1.005759489279023, + "grad_norm": 1.8896147012710571, + "learning_rate": 0.0001753675295604195, + "loss": 0.7353, + "step": 14494 + }, + { + "epoch": 1.0058288807161195, + "grad_norm": 2.965178966522217, + "learning_rate": 0.00017536383662359684, + "loss": 0.8145, + "step": 14495 + }, + { + "epoch": 1.0058982721532164, + "grad_norm": 1.61528742313385, + "learning_rate": 0.0001753601434488593, + "loss": 0.6382, + "step": 14496 + }, + { + "epoch": 1.005967663590313, + "grad_norm": 1.9443068504333496, + "learning_rate": 0.0001753564500362185, + "loss": 0.5909, + "step": 14497 + }, + { + "epoch": 1.0060370550274096, + "grad_norm": 2.0037267208099365, + "learning_rate": 0.00017535275638568614, + "loss": 0.8108, + "step": 14498 + }, + { + "epoch": 1.0061064464645062, + "grad_norm": 2.8104910850524902, + "learning_rate": 0.00017534906249727387, + "loss": 0.5179, + "step": 14499 + }, + { + "epoch": 1.006175837901603, + "grad_norm": 1.993272304534912, + "learning_rate": 0.00017534536837099339, + "loss": 0.8033, + "step": 14500 + }, + { + "epoch": 1.0062452293386996, + "grad_norm": 1.5355175733566284, + "learning_rate": 0.0001753416740068563, + "loss": 0.5429, + "step": 14501 + }, + { + "epoch": 1.0063146207757963, + "grad_norm": 1.8251503705978394, + "learning_rate": 0.00017533797940487425, + "loss": 0.4868, + "step": 14502 + }, + { + "epoch": 1.0063840122128929, + "grad_norm": 2.337721586227417, + "learning_rate": 0.000175334284565059, + "loss": 0.6582, + "step": 14503 + }, + { + "epoch": 1.0064534036499897, + "grad_norm": 2.1986539363861084, + "learning_rate": 0.0001753305894874221, + "loss": 0.7624, + "step": 14504 + }, + { + "epoch": 1.0065227950870863, + "grad_norm": 2.611370325088501, + "learning_rate": 0.00017532689417197527, + "loss": 0.8223, + "step": 14505 + }, + { + "epoch": 1.006592186524183, + "grad_norm": 2.054211139678955, + "learning_rate": 0.0001753231986187302, + "loss": 0.5925, + "step": 14506 + }, + { + "epoch": 1.0066615779612795, + "grad_norm": 2.0219368934631348, + "learning_rate": 0.00017531950282769854, + "loss": 0.8299, + "step": 14507 + }, + { + "epoch": 1.0067309693983761, + "grad_norm": 2.1783878803253174, + "learning_rate": 0.00017531580679889193, + "loss": 0.5048, + "step": 14508 + }, + { + "epoch": 1.006800360835473, + "grad_norm": 11.53846263885498, + "learning_rate": 0.00017531211053232207, + "loss": 0.7734, + "step": 14509 + }, + { + "epoch": 1.0068697522725696, + "grad_norm": 1.4782929420471191, + "learning_rate": 0.00017530841402800063, + "loss": 0.6015, + "step": 14510 + }, + { + "epoch": 1.0069391437096662, + "grad_norm": 2.3487961292266846, + "learning_rate": 0.00017530471728593924, + "loss": 0.7347, + "step": 14511 + }, + { + "epoch": 1.0070085351467628, + "grad_norm": 3.1981818675994873, + "learning_rate": 0.00017530102030614958, + "loss": 0.8277, + "step": 14512 + }, + { + "epoch": 1.0070779265838596, + "grad_norm": 3.591066598892212, + "learning_rate": 0.00017529732308864336, + "loss": 0.8949, + "step": 14513 + }, + { + "epoch": 1.0071473180209563, + "grad_norm": 3.1007282733917236, + "learning_rate": 0.00017529362563343222, + "loss": 0.8954, + "step": 14514 + }, + { + "epoch": 1.0072167094580529, + "grad_norm": 2.75793719291687, + "learning_rate": 0.00017528992794052782, + "loss": 0.6168, + "step": 14515 + }, + { + "epoch": 1.0072861008951495, + "grad_norm": 2.747849702835083, + "learning_rate": 0.00017528623000994187, + "loss": 0.7733, + "step": 14516 + }, + { + "epoch": 1.0073554923322463, + "grad_norm": 4.097131252288818, + "learning_rate": 0.00017528253184168603, + "loss": 0.894, + "step": 14517 + }, + { + "epoch": 1.007424883769343, + "grad_norm": 2.317099094390869, + "learning_rate": 0.00017527883343577196, + "loss": 0.7458, + "step": 14518 + }, + { + "epoch": 1.0074942752064395, + "grad_norm": 2.221115827560425, + "learning_rate": 0.00017527513479221135, + "loss": 0.8718, + "step": 14519 + }, + { + "epoch": 1.0075636666435361, + "grad_norm": 2.2612948417663574, + "learning_rate": 0.0001752714359110159, + "loss": 0.8606, + "step": 14520 + }, + { + "epoch": 1.0076330580806327, + "grad_norm": 2.3813164234161377, + "learning_rate": 0.00017526773679219724, + "loss": 0.6481, + "step": 14521 + }, + { + "epoch": 1.0077024495177296, + "grad_norm": 2.3498239517211914, + "learning_rate": 0.00017526403743576706, + "loss": 0.7944, + "step": 14522 + }, + { + "epoch": 1.0077718409548262, + "grad_norm": 3.476205587387085, + "learning_rate": 0.00017526033784173706, + "loss": 0.8207, + "step": 14523 + }, + { + "epoch": 1.0078412323919228, + "grad_norm": 2.4979088306427, + "learning_rate": 0.00017525663801011887, + "loss": 0.6107, + "step": 14524 + }, + { + "epoch": 1.0079106238290194, + "grad_norm": 3.3748044967651367, + "learning_rate": 0.00017525293794092424, + "loss": 0.6738, + "step": 14525 + }, + { + "epoch": 1.0079800152661162, + "grad_norm": 2.499159812927246, + "learning_rate": 0.00017524923763416483, + "loss": 0.6707, + "step": 14526 + }, + { + "epoch": 1.0080494067032129, + "grad_norm": 3.582801103591919, + "learning_rate": 0.00017524553708985228, + "loss": 1.0432, + "step": 14527 + }, + { + "epoch": 1.0081187981403095, + "grad_norm": 2.2552285194396973, + "learning_rate": 0.0001752418363079983, + "loss": 0.6675, + "step": 14528 + }, + { + "epoch": 1.008188189577406, + "grad_norm": 3.5260610580444336, + "learning_rate": 0.00017523813528861457, + "loss": 0.9874, + "step": 14529 + }, + { + "epoch": 1.008257581014503, + "grad_norm": 2.826760768890381, + "learning_rate": 0.00017523443403171277, + "loss": 0.9636, + "step": 14530 + }, + { + "epoch": 1.0083269724515995, + "grad_norm": 2.1510870456695557, + "learning_rate": 0.00017523073253730462, + "loss": 0.6768, + "step": 14531 + }, + { + "epoch": 1.0083963638886961, + "grad_norm": 3.6110482215881348, + "learning_rate": 0.00017522703080540176, + "loss": 0.7695, + "step": 14532 + }, + { + "epoch": 1.0084657553257927, + "grad_norm": 2.242021322250366, + "learning_rate": 0.00017522332883601588, + "loss": 0.8721, + "step": 14533 + }, + { + "epoch": 1.0085351467628894, + "grad_norm": 2.3018863201141357, + "learning_rate": 0.0001752196266291587, + "loss": 0.7253, + "step": 14534 + }, + { + "epoch": 1.0086045381999862, + "grad_norm": 2.4440505504608154, + "learning_rate": 0.00017521592418484188, + "loss": 0.8822, + "step": 14535 + }, + { + "epoch": 1.0086739296370828, + "grad_norm": 2.281730890274048, + "learning_rate": 0.0001752122215030771, + "loss": 0.8482, + "step": 14536 + }, + { + "epoch": 1.0087433210741794, + "grad_norm": 3.2089126110076904, + "learning_rate": 0.00017520851858387608, + "loss": 0.8798, + "step": 14537 + }, + { + "epoch": 1.008812712511276, + "grad_norm": 1.6131272315979004, + "learning_rate": 0.0001752048154272505, + "loss": 0.5694, + "step": 14538 + }, + { + "epoch": 1.0088821039483729, + "grad_norm": 2.0687785148620605, + "learning_rate": 0.00017520111203321202, + "loss": 0.7581, + "step": 14539 + }, + { + "epoch": 1.0089514953854695, + "grad_norm": 2.332812547683716, + "learning_rate": 0.00017519740840177237, + "loss": 0.701, + "step": 14540 + }, + { + "epoch": 1.009020886822566, + "grad_norm": 2.069267988204956, + "learning_rate": 0.00017519370453294322, + "loss": 0.7168, + "step": 14541 + }, + { + "epoch": 1.0090902782596627, + "grad_norm": 2.237891674041748, + "learning_rate": 0.00017519000042673626, + "loss": 0.6437, + "step": 14542 + }, + { + "epoch": 1.0091596696967595, + "grad_norm": 1.985019326210022, + "learning_rate": 0.0001751862960831632, + "loss": 0.7587, + "step": 14543 + }, + { + "epoch": 1.0092290611338561, + "grad_norm": 2.375441312789917, + "learning_rate": 0.00017518259150223572, + "loss": 0.8629, + "step": 14544 + }, + { + "epoch": 1.0092984525709527, + "grad_norm": 2.0644917488098145, + "learning_rate": 0.00017517888668396552, + "loss": 0.7974, + "step": 14545 + }, + { + "epoch": 1.0093678440080494, + "grad_norm": 3.570814371109009, + "learning_rate": 0.0001751751816283643, + "loss": 0.7477, + "step": 14546 + }, + { + "epoch": 1.009437235445146, + "grad_norm": 1.9483847618103027, + "learning_rate": 0.00017517147633544378, + "loss": 0.7064, + "step": 14547 + }, + { + "epoch": 1.0095066268822428, + "grad_norm": 2.4785449504852295, + "learning_rate": 0.0001751677708052156, + "loss": 0.8731, + "step": 14548 + }, + { + "epoch": 1.0095760183193394, + "grad_norm": 2.2888996601104736, + "learning_rate": 0.0001751640650376915, + "loss": 0.8035, + "step": 14549 + }, + { + "epoch": 1.009645409756436, + "grad_norm": 2.2707693576812744, + "learning_rate": 0.00017516035903288315, + "loss": 0.8854, + "step": 14550 + }, + { + "epoch": 1.0097148011935326, + "grad_norm": 2.489698648452759, + "learning_rate": 0.00017515665279080228, + "loss": 0.9441, + "step": 14551 + }, + { + "epoch": 1.0097841926306295, + "grad_norm": 2.459747076034546, + "learning_rate": 0.0001751529463114606, + "loss": 1.0252, + "step": 14552 + }, + { + "epoch": 1.009853584067726, + "grad_norm": 2.9175448417663574, + "learning_rate": 0.00017514923959486978, + "loss": 0.9867, + "step": 14553 + }, + { + "epoch": 1.0099229755048227, + "grad_norm": 3.002680540084839, + "learning_rate": 0.0001751455326410415, + "loss": 0.6624, + "step": 14554 + }, + { + "epoch": 1.0099923669419193, + "grad_norm": 1.9657294750213623, + "learning_rate": 0.00017514182544998752, + "loss": 0.5947, + "step": 14555 + }, + { + "epoch": 1.0100617583790161, + "grad_norm": 2.326817035675049, + "learning_rate": 0.0001751381180217195, + "loss": 0.8802, + "step": 14556 + }, + { + "epoch": 1.0101311498161127, + "grad_norm": 1.854224681854248, + "learning_rate": 0.00017513441035624916, + "loss": 0.582, + "step": 14557 + }, + { + "epoch": 1.0102005412532093, + "grad_norm": 2.8740808963775635, + "learning_rate": 0.0001751307024535882, + "loss": 0.9035, + "step": 14558 + }, + { + "epoch": 1.010269932690306, + "grad_norm": 1.7840757369995117, + "learning_rate": 0.00017512699431374834, + "loss": 1.0357, + "step": 14559 + }, + { + "epoch": 1.0103393241274026, + "grad_norm": 2.5569205284118652, + "learning_rate": 0.00017512328593674125, + "loss": 0.899, + "step": 14560 + }, + { + "epoch": 1.0104087155644994, + "grad_norm": 1.5775904655456543, + "learning_rate": 0.00017511957732257872, + "loss": 0.6517, + "step": 14561 + }, + { + "epoch": 1.010478107001596, + "grad_norm": 2.0765066146850586, + "learning_rate": 0.00017511586847127232, + "loss": 0.8375, + "step": 14562 + }, + { + "epoch": 1.0105474984386926, + "grad_norm": 1.4495809078216553, + "learning_rate": 0.0001751121593828339, + "loss": 0.6079, + "step": 14563 + }, + { + "epoch": 1.0106168898757892, + "grad_norm": 1.981370210647583, + "learning_rate": 0.0001751084500572751, + "loss": 0.8133, + "step": 14564 + }, + { + "epoch": 1.010686281312886, + "grad_norm": 2.111423969268799, + "learning_rate": 0.00017510474049460762, + "loss": 0.8982, + "step": 14565 + }, + { + "epoch": 1.0107556727499827, + "grad_norm": 2.1372194290161133, + "learning_rate": 0.0001751010306948432, + "loss": 0.5777, + "step": 14566 + }, + { + "epoch": 1.0108250641870793, + "grad_norm": 2.4565186500549316, + "learning_rate": 0.00017509732065799352, + "loss": 0.7185, + "step": 14567 + }, + { + "epoch": 1.010894455624176, + "grad_norm": 2.7127485275268555, + "learning_rate": 0.00017509361038407033, + "loss": 0.7945, + "step": 14568 + }, + { + "epoch": 1.0109638470612727, + "grad_norm": 1.7881457805633545, + "learning_rate": 0.0001750898998730853, + "loss": 0.8632, + "step": 14569 + }, + { + "epoch": 1.0110332384983693, + "grad_norm": 2.1283886432647705, + "learning_rate": 0.0001750861891250502, + "loss": 0.8951, + "step": 14570 + }, + { + "epoch": 1.011102629935466, + "grad_norm": 2.043874740600586, + "learning_rate": 0.0001750824781399767, + "loss": 0.7367, + "step": 14571 + }, + { + "epoch": 1.0111720213725626, + "grad_norm": 2.120896816253662, + "learning_rate": 0.00017507876691787651, + "loss": 0.75, + "step": 14572 + }, + { + "epoch": 1.0112414128096592, + "grad_norm": 4.944669723510742, + "learning_rate": 0.0001750750554587614, + "loss": 0.6954, + "step": 14573 + }, + { + "epoch": 1.011310804246756, + "grad_norm": 2.7554328441619873, + "learning_rate": 0.00017507134376264306, + "loss": 0.809, + "step": 14574 + }, + { + "epoch": 1.0113801956838526, + "grad_norm": 1.733963966369629, + "learning_rate": 0.00017506763182953313, + "loss": 0.6048, + "step": 14575 + }, + { + "epoch": 1.0114495871209492, + "grad_norm": 3.1139769554138184, + "learning_rate": 0.00017506391965944345, + "loss": 0.7244, + "step": 14576 + }, + { + "epoch": 1.0115189785580458, + "grad_norm": 2.5879807472229004, + "learning_rate": 0.00017506020725238566, + "loss": 0.8523, + "step": 14577 + }, + { + "epoch": 1.0115883699951427, + "grad_norm": 2.880571126937866, + "learning_rate": 0.00017505649460837152, + "loss": 0.7595, + "step": 14578 + }, + { + "epoch": 1.0116577614322393, + "grad_norm": 2.284306764602661, + "learning_rate": 0.00017505278172741273, + "loss": 0.7048, + "step": 14579 + }, + { + "epoch": 1.011727152869336, + "grad_norm": 2.125645875930786, + "learning_rate": 0.00017504906860952102, + "loss": 0.849, + "step": 14580 + }, + { + "epoch": 1.0117965443064325, + "grad_norm": 1.684417963027954, + "learning_rate": 0.00017504535525470809, + "loss": 0.5691, + "step": 14581 + }, + { + "epoch": 1.0118659357435293, + "grad_norm": 2.1394667625427246, + "learning_rate": 0.00017504164166298568, + "loss": 0.6745, + "step": 14582 + }, + { + "epoch": 1.011935327180626, + "grad_norm": 2.286250352859497, + "learning_rate": 0.00017503792783436556, + "loss": 0.9116, + "step": 14583 + }, + { + "epoch": 1.0120047186177226, + "grad_norm": 2.4670932292938232, + "learning_rate": 0.00017503421376885935, + "loss": 0.6769, + "step": 14584 + }, + { + "epoch": 1.0120741100548192, + "grad_norm": 1.844481110572815, + "learning_rate": 0.00017503049946647887, + "loss": 0.5496, + "step": 14585 + }, + { + "epoch": 1.012143501491916, + "grad_norm": 2.9041643142700195, + "learning_rate": 0.00017502678492723579, + "loss": 0.7233, + "step": 14586 + }, + { + "epoch": 1.0122128929290126, + "grad_norm": 1.8994362354278564, + "learning_rate": 0.00017502307015114185, + "loss": 0.6628, + "step": 14587 + }, + { + "epoch": 1.0122822843661092, + "grad_norm": 2.4780166149139404, + "learning_rate": 0.0001750193551382088, + "loss": 0.7599, + "step": 14588 + }, + { + "epoch": 1.0123516758032058, + "grad_norm": 2.2433114051818848, + "learning_rate": 0.00017501563988844834, + "loss": 0.7069, + "step": 14589 + }, + { + "epoch": 1.0124210672403025, + "grad_norm": 2.5540904998779297, + "learning_rate": 0.00017501192440187222, + "loss": 0.8574, + "step": 14590 + }, + { + "epoch": 1.0124904586773993, + "grad_norm": 3.2908902168273926, + "learning_rate": 0.00017500820867849215, + "loss": 0.8455, + "step": 14591 + }, + { + "epoch": 1.012559850114496, + "grad_norm": 2.113924741744995, + "learning_rate": 0.00017500449271831987, + "loss": 0.4886, + "step": 14592 + }, + { + "epoch": 1.0126292415515925, + "grad_norm": 2.1385533809661865, + "learning_rate": 0.0001750007765213671, + "loss": 0.8129, + "step": 14593 + }, + { + "epoch": 1.0126986329886891, + "grad_norm": 2.0068585872650146, + "learning_rate": 0.00017499706008764556, + "loss": 0.6482, + "step": 14594 + }, + { + "epoch": 1.012768024425786, + "grad_norm": 2.789210796356201, + "learning_rate": 0.00017499334341716701, + "loss": 0.8308, + "step": 14595 + }, + { + "epoch": 1.0128374158628826, + "grad_norm": 1.8947087526321411, + "learning_rate": 0.00017498962650994322, + "loss": 0.6226, + "step": 14596 + }, + { + "epoch": 1.0129068072999792, + "grad_norm": 5.049558162689209, + "learning_rate": 0.00017498590936598584, + "loss": 0.6791, + "step": 14597 + }, + { + "epoch": 1.0129761987370758, + "grad_norm": 2.529524087905884, + "learning_rate": 0.00017498219198530665, + "loss": 0.6955, + "step": 14598 + }, + { + "epoch": 1.0130455901741726, + "grad_norm": 2.4033102989196777, + "learning_rate": 0.0001749784743679174, + "loss": 0.8772, + "step": 14599 + }, + { + "epoch": 1.0131149816112692, + "grad_norm": 2.5665342807769775, + "learning_rate": 0.00017497475651382977, + "loss": 0.755, + "step": 14600 + }, + { + "epoch": 1.0131843730483658, + "grad_norm": 1.7491538524627686, + "learning_rate": 0.00017497103842305557, + "loss": 0.5783, + "step": 14601 + }, + { + "epoch": 1.0132537644854624, + "grad_norm": 2.3898868560791016, + "learning_rate": 0.00017496732009560648, + "loss": 0.6876, + "step": 14602 + }, + { + "epoch": 1.013323155922559, + "grad_norm": 1.7179120779037476, + "learning_rate": 0.00017496360153149428, + "loss": 0.7985, + "step": 14603 + }, + { + "epoch": 1.013392547359656, + "grad_norm": 1.9916707277297974, + "learning_rate": 0.00017495988273073066, + "loss": 0.7578, + "step": 14604 + }, + { + "epoch": 1.0134619387967525, + "grad_norm": 2.1261563301086426, + "learning_rate": 0.0001749561636933274, + "loss": 0.6917, + "step": 14605 + }, + { + "epoch": 1.0135313302338491, + "grad_norm": 1.896748661994934, + "learning_rate": 0.0001749524444192962, + "loss": 0.8472, + "step": 14606 + }, + { + "epoch": 1.0136007216709457, + "grad_norm": 1.7308979034423828, + "learning_rate": 0.00017494872490864886, + "loss": 0.4696, + "step": 14607 + }, + { + "epoch": 1.0136701131080426, + "grad_norm": 3.03376841545105, + "learning_rate": 0.00017494500516139708, + "loss": 0.6667, + "step": 14608 + }, + { + "epoch": 1.0137395045451392, + "grad_norm": 2.4594273567199707, + "learning_rate": 0.00017494128517755265, + "loss": 0.9038, + "step": 14609 + }, + { + "epoch": 1.0138088959822358, + "grad_norm": 1.8150994777679443, + "learning_rate": 0.0001749375649571272, + "loss": 0.7361, + "step": 14610 + }, + { + "epoch": 1.0138782874193324, + "grad_norm": 3.9821057319641113, + "learning_rate": 0.00017493384450013264, + "loss": 0.9736, + "step": 14611 + }, + { + "epoch": 1.0139476788564292, + "grad_norm": 1.4482612609863281, + "learning_rate": 0.00017493012380658055, + "loss": 0.4306, + "step": 14612 + }, + { + "epoch": 1.0140170702935258, + "grad_norm": 1.638054370880127, + "learning_rate": 0.0001749264028764828, + "loss": 0.7088, + "step": 14613 + }, + { + "epoch": 1.0140864617306224, + "grad_norm": 2.994302988052368, + "learning_rate": 0.00017492268170985108, + "loss": 0.6817, + "step": 14614 + }, + { + "epoch": 1.014155853167719, + "grad_norm": 2.4012889862060547, + "learning_rate": 0.00017491896030669715, + "loss": 0.6072, + "step": 14615 + }, + { + "epoch": 1.0142252446048157, + "grad_norm": 1.5964293479919434, + "learning_rate": 0.00017491523866703274, + "loss": 0.6165, + "step": 14616 + }, + { + "epoch": 1.0142946360419125, + "grad_norm": 2.7539782524108887, + "learning_rate": 0.0001749115167908696, + "loss": 0.7428, + "step": 14617 + }, + { + "epoch": 1.0143640274790091, + "grad_norm": 2.3355202674865723, + "learning_rate": 0.00017490779467821952, + "loss": 0.7414, + "step": 14618 + }, + { + "epoch": 1.0144334189161057, + "grad_norm": 2.3450818061828613, + "learning_rate": 0.00017490407232909422, + "loss": 0.935, + "step": 14619 + }, + { + "epoch": 1.0145028103532023, + "grad_norm": 2.1604130268096924, + "learning_rate": 0.00017490034974350547, + "loss": 0.7351, + "step": 14620 + }, + { + "epoch": 1.0145722017902992, + "grad_norm": 1.85299551486969, + "learning_rate": 0.00017489662692146495, + "loss": 0.6142, + "step": 14621 + }, + { + "epoch": 1.0146415932273958, + "grad_norm": 2.2292933464050293, + "learning_rate": 0.00017489290386298452, + "loss": 0.9787, + "step": 14622 + }, + { + "epoch": 1.0147109846644924, + "grad_norm": 1.5604842901229858, + "learning_rate": 0.00017488918056807586, + "loss": 0.4735, + "step": 14623 + }, + { + "epoch": 1.014780376101589, + "grad_norm": 3.1215858459472656, + "learning_rate": 0.00017488545703675076, + "loss": 0.7098, + "step": 14624 + }, + { + "epoch": 1.0148497675386858, + "grad_norm": 1.9763463735580444, + "learning_rate": 0.00017488173326902096, + "loss": 0.819, + "step": 14625 + }, + { + "epoch": 1.0149191589757824, + "grad_norm": 2.594911575317383, + "learning_rate": 0.0001748780092648982, + "loss": 0.7718, + "step": 14626 + }, + { + "epoch": 1.014988550412879, + "grad_norm": 2.263197183609009, + "learning_rate": 0.00017487428502439427, + "loss": 0.7526, + "step": 14627 + }, + { + "epoch": 1.0150579418499757, + "grad_norm": 2.0170652866363525, + "learning_rate": 0.0001748705605475209, + "loss": 0.4846, + "step": 14628 + }, + { + "epoch": 1.0151273332870723, + "grad_norm": 1.9736435413360596, + "learning_rate": 0.00017486683583428983, + "loss": 0.8281, + "step": 14629 + }, + { + "epoch": 1.015196724724169, + "grad_norm": 2.960867166519165, + "learning_rate": 0.00017486311088471288, + "loss": 0.8286, + "step": 14630 + }, + { + "epoch": 1.0152661161612657, + "grad_norm": 1.90788996219635, + "learning_rate": 0.00017485938569880176, + "loss": 0.5728, + "step": 14631 + }, + { + "epoch": 1.0153355075983623, + "grad_norm": 1.3394097089767456, + "learning_rate": 0.00017485566027656828, + "loss": 0.5927, + "step": 14632 + }, + { + "epoch": 1.015404899035459, + "grad_norm": 2.850369691848755, + "learning_rate": 0.00017485193461802415, + "loss": 0.963, + "step": 14633 + }, + { + "epoch": 1.0154742904725558, + "grad_norm": 2.088383674621582, + "learning_rate": 0.00017484820872318112, + "loss": 0.6555, + "step": 14634 + }, + { + "epoch": 1.0155436819096524, + "grad_norm": 1.5778307914733887, + "learning_rate": 0.000174844482592051, + "loss": 0.5903, + "step": 14635 + }, + { + "epoch": 1.015613073346749, + "grad_norm": 1.9911391735076904, + "learning_rate": 0.0001748407562246455, + "loss": 0.4911, + "step": 14636 + }, + { + "epoch": 1.0156824647838456, + "grad_norm": 1.5737625360488892, + "learning_rate": 0.00017483702962097644, + "loss": 0.4328, + "step": 14637 + }, + { + "epoch": 1.0157518562209424, + "grad_norm": 1.8484182357788086, + "learning_rate": 0.00017483330278105557, + "loss": 0.6783, + "step": 14638 + }, + { + "epoch": 1.015821247658039, + "grad_norm": 2.5990307331085205, + "learning_rate": 0.00017482957570489463, + "loss": 0.9247, + "step": 14639 + }, + { + "epoch": 1.0158906390951357, + "grad_norm": 2.2593743801116943, + "learning_rate": 0.00017482584839250543, + "loss": 0.8409, + "step": 14640 + }, + { + "epoch": 1.0159600305322323, + "grad_norm": 2.189218282699585, + "learning_rate": 0.00017482212084389968, + "loss": 0.8029, + "step": 14641 + }, + { + "epoch": 1.0160294219693289, + "grad_norm": 2.1814815998077393, + "learning_rate": 0.00017481839305908916, + "loss": 0.9445, + "step": 14642 + }, + { + "epoch": 1.0160988134064257, + "grad_norm": 1.7760770320892334, + "learning_rate": 0.0001748146650380857, + "loss": 0.5446, + "step": 14643 + }, + { + "epoch": 1.0161682048435223, + "grad_norm": 2.4224424362182617, + "learning_rate": 0.000174810936780901, + "loss": 0.8639, + "step": 14644 + }, + { + "epoch": 1.016237596280619, + "grad_norm": 1.9527604579925537, + "learning_rate": 0.00017480720828754682, + "loss": 0.7116, + "step": 14645 + }, + { + "epoch": 1.0163069877177155, + "grad_norm": 2.4908294677734375, + "learning_rate": 0.000174803479558035, + "loss": 0.7257, + "step": 14646 + }, + { + "epoch": 1.0163763791548124, + "grad_norm": 2.735536575317383, + "learning_rate": 0.00017479975059237727, + "loss": 0.8025, + "step": 14647 + }, + { + "epoch": 1.016445770591909, + "grad_norm": 2.6414923667907715, + "learning_rate": 0.0001747960213905854, + "loss": 0.6738, + "step": 14648 + }, + { + "epoch": 1.0165151620290056, + "grad_norm": 1.6666851043701172, + "learning_rate": 0.00017479229195267116, + "loss": 0.4989, + "step": 14649 + }, + { + "epoch": 1.0165845534661022, + "grad_norm": 2.3092126846313477, + "learning_rate": 0.00017478856227864635, + "loss": 0.5876, + "step": 14650 + }, + { + "epoch": 1.016653944903199, + "grad_norm": 1.961692452430725, + "learning_rate": 0.00017478483236852273, + "loss": 0.581, + "step": 14651 + }, + { + "epoch": 1.0167233363402957, + "grad_norm": 1.723120927810669, + "learning_rate": 0.00017478110222231204, + "loss": 0.8603, + "step": 14652 + }, + { + "epoch": 1.0167927277773923, + "grad_norm": 2.323873519897461, + "learning_rate": 0.00017477737184002612, + "loss": 0.8395, + "step": 14653 + }, + { + "epoch": 1.0168621192144889, + "grad_norm": 2.48838472366333, + "learning_rate": 0.00017477364122167667, + "loss": 0.589, + "step": 14654 + }, + { + "epoch": 1.0169315106515855, + "grad_norm": 1.9505462646484375, + "learning_rate": 0.00017476991036727553, + "loss": 0.6936, + "step": 14655 + }, + { + "epoch": 1.0170009020886823, + "grad_norm": 1.9527870416641235, + "learning_rate": 0.00017476617927683448, + "loss": 0.9047, + "step": 14656 + }, + { + "epoch": 1.017070293525779, + "grad_norm": 2.4122653007507324, + "learning_rate": 0.00017476244795036527, + "loss": 0.7671, + "step": 14657 + }, + { + "epoch": 1.0171396849628755, + "grad_norm": 2.1141977310180664, + "learning_rate": 0.00017475871638787966, + "loss": 0.5941, + "step": 14658 + }, + { + "epoch": 1.0172090763999722, + "grad_norm": 1.9315314292907715, + "learning_rate": 0.00017475498458938948, + "loss": 0.5394, + "step": 14659 + }, + { + "epoch": 1.017278467837069, + "grad_norm": 1.6977448463439941, + "learning_rate": 0.0001747512525549065, + "loss": 0.6801, + "step": 14660 + }, + { + "epoch": 1.0173478592741656, + "grad_norm": 1.4786396026611328, + "learning_rate": 0.00017474752028444245, + "loss": 0.6647, + "step": 14661 + }, + { + "epoch": 1.0174172507112622, + "grad_norm": 2.816472053527832, + "learning_rate": 0.00017474378777800918, + "loss": 0.724, + "step": 14662 + }, + { + "epoch": 1.0174866421483588, + "grad_norm": 3.6966209411621094, + "learning_rate": 0.00017474005503561845, + "loss": 0.7313, + "step": 14663 + }, + { + "epoch": 1.0175560335854557, + "grad_norm": 2.2939398288726807, + "learning_rate": 0.000174736322057282, + "loss": 0.4719, + "step": 14664 + }, + { + "epoch": 1.0176254250225523, + "grad_norm": 1.909502625465393, + "learning_rate": 0.00017473258884301166, + "loss": 0.9101, + "step": 14665 + }, + { + "epoch": 1.0176948164596489, + "grad_norm": 2.6158876419067383, + "learning_rate": 0.00017472885539281924, + "loss": 0.8266, + "step": 14666 + }, + { + "epoch": 1.0177642078967455, + "grad_norm": 2.0289394855499268, + "learning_rate": 0.0001747251217067165, + "loss": 0.6309, + "step": 14667 + }, + { + "epoch": 1.017833599333842, + "grad_norm": 1.8826607465744019, + "learning_rate": 0.0001747213877847152, + "loss": 0.7151, + "step": 14668 + }, + { + "epoch": 1.017902990770939, + "grad_norm": 2.3728182315826416, + "learning_rate": 0.00017471765362682713, + "loss": 0.6283, + "step": 14669 + }, + { + "epoch": 1.0179723822080355, + "grad_norm": 2.0820789337158203, + "learning_rate": 0.0001747139192330641, + "loss": 0.6869, + "step": 14670 + }, + { + "epoch": 1.0180417736451322, + "grad_norm": 1.9557452201843262, + "learning_rate": 0.00017471018460343788, + "loss": 0.5727, + "step": 14671 + }, + { + "epoch": 1.0181111650822288, + "grad_norm": 2.5339136123657227, + "learning_rate": 0.00017470644973796034, + "loss": 0.6435, + "step": 14672 + }, + { + "epoch": 1.0181805565193256, + "grad_norm": 2.277203321456909, + "learning_rate": 0.00017470271463664315, + "loss": 0.7276, + "step": 14673 + }, + { + "epoch": 1.0182499479564222, + "grad_norm": 2.5910394191741943, + "learning_rate": 0.0001746989792994982, + "loss": 0.8474, + "step": 14674 + }, + { + "epoch": 1.0183193393935188, + "grad_norm": 1.8647269010543823, + "learning_rate": 0.00017469524372653718, + "loss": 0.6515, + "step": 14675 + }, + { + "epoch": 1.0183887308306154, + "grad_norm": 2.2207248210906982, + "learning_rate": 0.00017469150791777198, + "loss": 0.6275, + "step": 14676 + }, + { + "epoch": 1.0184581222677123, + "grad_norm": 1.7729628086090088, + "learning_rate": 0.00017468777187321435, + "loss": 0.6169, + "step": 14677 + }, + { + "epoch": 1.0185275137048089, + "grad_norm": 2.917475938796997, + "learning_rate": 0.0001746840355928761, + "loss": 0.8351, + "step": 14678 + }, + { + "epoch": 1.0185969051419055, + "grad_norm": 3.2100722789764404, + "learning_rate": 0.000174680299076769, + "loss": 0.7521, + "step": 14679 + }, + { + "epoch": 1.018666296579002, + "grad_norm": 2.3627283573150635, + "learning_rate": 0.00017467656232490487, + "loss": 0.6122, + "step": 14680 + }, + { + "epoch": 1.0187356880160987, + "grad_norm": 2.3511745929718018, + "learning_rate": 0.00017467282533729548, + "loss": 0.6221, + "step": 14681 + }, + { + "epoch": 1.0188050794531955, + "grad_norm": 1.848154067993164, + "learning_rate": 0.00017466908811395265, + "loss": 0.7559, + "step": 14682 + }, + { + "epoch": 1.0188744708902921, + "grad_norm": 1.4552327394485474, + "learning_rate": 0.0001746653506548882, + "loss": 0.4597, + "step": 14683 + }, + { + "epoch": 1.0189438623273888, + "grad_norm": 2.106571674346924, + "learning_rate": 0.00017466161296011387, + "loss": 0.8297, + "step": 14684 + }, + { + "epoch": 1.0190132537644854, + "grad_norm": 2.025736093521118, + "learning_rate": 0.00017465787502964148, + "loss": 0.8954, + "step": 14685 + }, + { + "epoch": 1.0190826452015822, + "grad_norm": 2.762185573577881, + "learning_rate": 0.00017465413686348287, + "loss": 0.6926, + "step": 14686 + }, + { + "epoch": 1.0191520366386788, + "grad_norm": 1.8446199893951416, + "learning_rate": 0.0001746503984616498, + "loss": 0.6345, + "step": 14687 + }, + { + "epoch": 1.0192214280757754, + "grad_norm": 2.669971466064453, + "learning_rate": 0.00017464665982415406, + "loss": 0.9492, + "step": 14688 + }, + { + "epoch": 1.019290819512872, + "grad_norm": 2.025662422180176, + "learning_rate": 0.0001746429209510075, + "loss": 0.647, + "step": 14689 + }, + { + "epoch": 1.0193602109499689, + "grad_norm": 2.0786662101745605, + "learning_rate": 0.0001746391818422219, + "loss": 0.7309, + "step": 14690 + }, + { + "epoch": 1.0194296023870655, + "grad_norm": 2.8817291259765625, + "learning_rate": 0.0001746354424978091, + "loss": 0.8396, + "step": 14691 + }, + { + "epoch": 1.019498993824162, + "grad_norm": 2.2740182876586914, + "learning_rate": 0.00017463170291778083, + "loss": 0.6286, + "step": 14692 + }, + { + "epoch": 1.0195683852612587, + "grad_norm": 2.3563177585601807, + "learning_rate": 0.0001746279631021489, + "loss": 0.7489, + "step": 14693 + }, + { + "epoch": 1.0196377766983553, + "grad_norm": 2.019317388534546, + "learning_rate": 0.00017462422305092518, + "loss": 0.5869, + "step": 14694 + }, + { + "epoch": 1.0197071681354521, + "grad_norm": 2.200855255126953, + "learning_rate": 0.00017462048276412144, + "loss": 0.6054, + "step": 14695 + }, + { + "epoch": 1.0197765595725488, + "grad_norm": 2.3284409046173096, + "learning_rate": 0.0001746167422417495, + "loss": 0.7583, + "step": 14696 + }, + { + "epoch": 1.0198459510096454, + "grad_norm": 2.2143280506134033, + "learning_rate": 0.00017461300148382115, + "loss": 0.6213, + "step": 14697 + }, + { + "epoch": 1.019915342446742, + "grad_norm": 1.9386166334152222, + "learning_rate": 0.00017460926049034825, + "loss": 0.4567, + "step": 14698 + }, + { + "epoch": 1.0199847338838388, + "grad_norm": 2.1917083263397217, + "learning_rate": 0.00017460551926134254, + "loss": 0.8411, + "step": 14699 + }, + { + "epoch": 1.0200541253209354, + "grad_norm": 2.8559887409210205, + "learning_rate": 0.00017460177779681586, + "loss": 0.8447, + "step": 14700 + }, + { + "epoch": 1.020123516758032, + "grad_norm": 3.0641744136810303, + "learning_rate": 0.00017459803609678, + "loss": 0.5884, + "step": 14701 + }, + { + "epoch": 1.0201929081951286, + "grad_norm": 2.5465519428253174, + "learning_rate": 0.00017459429416124685, + "loss": 0.9347, + "step": 14702 + }, + { + "epoch": 1.0202622996322255, + "grad_norm": 1.8018620014190674, + "learning_rate": 0.00017459055199022812, + "loss": 0.5656, + "step": 14703 + }, + { + "epoch": 1.020331691069322, + "grad_norm": 1.6360960006713867, + "learning_rate": 0.0001745868095837357, + "loss": 0.6867, + "step": 14704 + }, + { + "epoch": 1.0204010825064187, + "grad_norm": 1.5852153301239014, + "learning_rate": 0.00017458306694178137, + "loss": 0.5168, + "step": 14705 + }, + { + "epoch": 1.0204704739435153, + "grad_norm": 2.0293612480163574, + "learning_rate": 0.00017457932406437693, + "loss": 0.7246, + "step": 14706 + }, + { + "epoch": 1.020539865380612, + "grad_norm": 1.8071293830871582, + "learning_rate": 0.00017457558095153422, + "loss": 0.5393, + "step": 14707 + }, + { + "epoch": 1.0206092568177088, + "grad_norm": 1.6532139778137207, + "learning_rate": 0.0001745718376032651, + "loss": 0.4626, + "step": 14708 + }, + { + "epoch": 1.0206786482548054, + "grad_norm": 1.8705482482910156, + "learning_rate": 0.0001745680940195813, + "loss": 0.5691, + "step": 14709 + }, + { + "epoch": 1.020748039691902, + "grad_norm": 2.5510847568511963, + "learning_rate": 0.00017456435020049468, + "loss": 0.8768, + "step": 14710 + }, + { + "epoch": 1.0208174311289986, + "grad_norm": 1.7638500928878784, + "learning_rate": 0.00017456060614601707, + "loss": 0.6368, + "step": 14711 + }, + { + "epoch": 1.0208868225660954, + "grad_norm": 1.8606352806091309, + "learning_rate": 0.00017455686185616028, + "loss": 0.807, + "step": 14712 + }, + { + "epoch": 1.020956214003192, + "grad_norm": 2.4109599590301514, + "learning_rate": 0.00017455311733093613, + "loss": 0.6528, + "step": 14713 + }, + { + "epoch": 1.0210256054402886, + "grad_norm": 2.147597551345825, + "learning_rate": 0.00017454937257035643, + "loss": 0.877, + "step": 14714 + }, + { + "epoch": 1.0210949968773853, + "grad_norm": 2.3566882610321045, + "learning_rate": 0.000174545627574433, + "loss": 0.6637, + "step": 14715 + }, + { + "epoch": 1.021164388314482, + "grad_norm": 1.973287582397461, + "learning_rate": 0.00017454188234317767, + "loss": 0.5708, + "step": 14716 + }, + { + "epoch": 1.0212337797515787, + "grad_norm": 2.028958320617676, + "learning_rate": 0.00017453813687660227, + "loss": 0.7483, + "step": 14717 + }, + { + "epoch": 1.0213031711886753, + "grad_norm": 2.0818517208099365, + "learning_rate": 0.00017453439117471863, + "loss": 0.7122, + "step": 14718 + }, + { + "epoch": 1.021372562625772, + "grad_norm": 1.7959694862365723, + "learning_rate": 0.00017453064523753855, + "loss": 0.7951, + "step": 14719 + }, + { + "epoch": 1.0214419540628688, + "grad_norm": 1.6075198650360107, + "learning_rate": 0.00017452689906507392, + "loss": 0.6354, + "step": 14720 + }, + { + "epoch": 1.0215113454999654, + "grad_norm": 2.9687559604644775, + "learning_rate": 0.00017452315265733645, + "loss": 0.8615, + "step": 14721 + }, + { + "epoch": 1.021580736937062, + "grad_norm": 1.938751459121704, + "learning_rate": 0.00017451940601433808, + "loss": 0.8459, + "step": 14722 + }, + { + "epoch": 1.0216501283741586, + "grad_norm": 2.100248336791992, + "learning_rate": 0.00017451565913609055, + "loss": 0.828, + "step": 14723 + }, + { + "epoch": 1.0217195198112552, + "grad_norm": 2.032850503921509, + "learning_rate": 0.00017451191202260576, + "loss": 0.7904, + "step": 14724 + }, + { + "epoch": 1.021788911248352, + "grad_norm": 1.5795753002166748, + "learning_rate": 0.0001745081646738955, + "loss": 0.6513, + "step": 14725 + }, + { + "epoch": 1.0218583026854486, + "grad_norm": 1.9230042695999146, + "learning_rate": 0.0001745044170899716, + "loss": 0.7717, + "step": 14726 + }, + { + "epoch": 1.0219276941225452, + "grad_norm": 1.5335866212844849, + "learning_rate": 0.00017450066927084592, + "loss": 0.5675, + "step": 14727 + }, + { + "epoch": 1.0219970855596419, + "grad_norm": 2.1497533321380615, + "learning_rate": 0.00017449692121653025, + "loss": 0.7185, + "step": 14728 + }, + { + "epoch": 1.0220664769967387, + "grad_norm": 1.7688435316085815, + "learning_rate": 0.00017449317292703644, + "loss": 0.7425, + "step": 14729 + }, + { + "epoch": 1.0221358684338353, + "grad_norm": 1.9602869749069214, + "learning_rate": 0.00017448942440237634, + "loss": 0.8459, + "step": 14730 + }, + { + "epoch": 1.022205259870932, + "grad_norm": 1.6280115842819214, + "learning_rate": 0.00017448567564256178, + "loss": 0.6579, + "step": 14731 + }, + { + "epoch": 1.0222746513080285, + "grad_norm": 1.6422085762023926, + "learning_rate": 0.00017448192664760458, + "loss": 0.6118, + "step": 14732 + }, + { + "epoch": 1.0223440427451254, + "grad_norm": 2.1070003509521484, + "learning_rate": 0.00017447817741751654, + "loss": 0.8395, + "step": 14733 + }, + { + "epoch": 1.022413434182222, + "grad_norm": 2.132561683654785, + "learning_rate": 0.00017447442795230956, + "loss": 0.6379, + "step": 14734 + }, + { + "epoch": 1.0224828256193186, + "grad_norm": 1.5291759967803955, + "learning_rate": 0.00017447067825199548, + "loss": 0.572, + "step": 14735 + }, + { + "epoch": 1.0225522170564152, + "grad_norm": 1.3952341079711914, + "learning_rate": 0.00017446692831658608, + "loss": 0.5299, + "step": 14736 + }, + { + "epoch": 1.0226216084935118, + "grad_norm": 1.9247729778289795, + "learning_rate": 0.00017446317814609325, + "loss": 0.5764, + "step": 14737 + }, + { + "epoch": 1.0226909999306086, + "grad_norm": 1.9140969514846802, + "learning_rate": 0.00017445942774052878, + "loss": 0.8536, + "step": 14738 + }, + { + "epoch": 1.0227603913677052, + "grad_norm": 1.9603097438812256, + "learning_rate": 0.00017445567709990453, + "loss": 0.6229, + "step": 14739 + }, + { + "epoch": 1.0228297828048019, + "grad_norm": 1.7875310182571411, + "learning_rate": 0.00017445192622423238, + "loss": 0.6574, + "step": 14740 + }, + { + "epoch": 1.0228991742418985, + "grad_norm": 2.042656660079956, + "learning_rate": 0.0001744481751135241, + "loss": 0.6925, + "step": 14741 + }, + { + "epoch": 1.0229685656789953, + "grad_norm": 2.3930773735046387, + "learning_rate": 0.0001744444237677916, + "loss": 0.7469, + "step": 14742 + }, + { + "epoch": 1.023037957116092, + "grad_norm": 1.7928487062454224, + "learning_rate": 0.00017444067218704668, + "loss": 0.8132, + "step": 14743 + }, + { + "epoch": 1.0231073485531885, + "grad_norm": 2.952996015548706, + "learning_rate": 0.00017443692037130118, + "loss": 0.7374, + "step": 14744 + }, + { + "epoch": 1.0231767399902851, + "grad_norm": 1.7912476062774658, + "learning_rate": 0.00017443316832056697, + "loss": 0.7564, + "step": 14745 + }, + { + "epoch": 1.023246131427382, + "grad_norm": 1.413756251335144, + "learning_rate": 0.0001744294160348559, + "loss": 0.3912, + "step": 14746 + }, + { + "epoch": 1.0233155228644786, + "grad_norm": 2.016160249710083, + "learning_rate": 0.00017442566351417977, + "loss": 0.7721, + "step": 14747 + }, + { + "epoch": 1.0233849143015752, + "grad_norm": 1.7577029466629028, + "learning_rate": 0.0001744219107585505, + "loss": 0.4759, + "step": 14748 + }, + { + "epoch": 1.0234543057386718, + "grad_norm": 1.5833826065063477, + "learning_rate": 0.00017441815776797983, + "loss": 0.6446, + "step": 14749 + }, + { + "epoch": 1.0235236971757684, + "grad_norm": 1.9563745260238647, + "learning_rate": 0.0001744144045424797, + "loss": 0.7285, + "step": 14750 + }, + { + "epoch": 1.0235930886128652, + "grad_norm": 1.9198524951934814, + "learning_rate": 0.00017441065108206191, + "loss": 0.7861, + "step": 14751 + }, + { + "epoch": 1.0236624800499619, + "grad_norm": 1.9887683391571045, + "learning_rate": 0.00017440689738673837, + "loss": 0.7963, + "step": 14752 + }, + { + "epoch": 1.0237318714870585, + "grad_norm": 2.429048538208008, + "learning_rate": 0.00017440314345652084, + "loss": 0.7786, + "step": 14753 + }, + { + "epoch": 1.023801262924155, + "grad_norm": 1.8444182872772217, + "learning_rate": 0.00017439938929142125, + "loss": 0.5936, + "step": 14754 + }, + { + "epoch": 1.023870654361252, + "grad_norm": 2.830064058303833, + "learning_rate": 0.0001743956348914514, + "loss": 0.9434, + "step": 14755 + }, + { + "epoch": 1.0239400457983485, + "grad_norm": 2.4042489528656006, + "learning_rate": 0.0001743918802566232, + "loss": 0.5839, + "step": 14756 + }, + { + "epoch": 1.0240094372354451, + "grad_norm": 1.7862588167190552, + "learning_rate": 0.00017438812538694842, + "loss": 0.5616, + "step": 14757 + }, + { + "epoch": 1.0240788286725417, + "grad_norm": 1.6231931447982788, + "learning_rate": 0.000174384370282439, + "loss": 0.5447, + "step": 14758 + }, + { + "epoch": 1.0241482201096386, + "grad_norm": 1.3088070154190063, + "learning_rate": 0.00017438061494310674, + "loss": 0.4387, + "step": 14759 + }, + { + "epoch": 1.0242176115467352, + "grad_norm": 2.0388386249542236, + "learning_rate": 0.00017437685936896348, + "loss": 0.5964, + "step": 14760 + }, + { + "epoch": 1.0242870029838318, + "grad_norm": 1.9468538761138916, + "learning_rate": 0.00017437310356002112, + "loss": 0.6917, + "step": 14761 + }, + { + "epoch": 1.0243563944209284, + "grad_norm": 2.124619960784912, + "learning_rate": 0.0001743693475162915, + "loss": 0.8701, + "step": 14762 + }, + { + "epoch": 1.024425785858025, + "grad_norm": 1.552708387374878, + "learning_rate": 0.0001743655912377865, + "loss": 0.5971, + "step": 14763 + }, + { + "epoch": 1.0244951772951219, + "grad_norm": 2.0694997310638428, + "learning_rate": 0.00017436183472451796, + "loss": 0.6335, + "step": 14764 + }, + { + "epoch": 1.0245645687322185, + "grad_norm": 1.7032058238983154, + "learning_rate": 0.00017435807797649768, + "loss": 0.5999, + "step": 14765 + }, + { + "epoch": 1.024633960169315, + "grad_norm": 2.126248836517334, + "learning_rate": 0.00017435432099373763, + "loss": 0.57, + "step": 14766 + }, + { + "epoch": 1.0247033516064117, + "grad_norm": 1.669064998626709, + "learning_rate": 0.0001743505637762496, + "loss": 0.5626, + "step": 14767 + }, + { + "epoch": 1.0247727430435085, + "grad_norm": 1.9638997316360474, + "learning_rate": 0.00017434680632404546, + "loss": 0.8521, + "step": 14768 + }, + { + "epoch": 1.0248421344806051, + "grad_norm": 1.4548070430755615, + "learning_rate": 0.00017434304863713705, + "loss": 0.4127, + "step": 14769 + }, + { + "epoch": 1.0249115259177017, + "grad_norm": 1.8838657140731812, + "learning_rate": 0.0001743392907155363, + "loss": 0.6041, + "step": 14770 + }, + { + "epoch": 1.0249809173547983, + "grad_norm": 2.2346861362457275, + "learning_rate": 0.00017433553255925504, + "loss": 0.761, + "step": 14771 + }, + { + "epoch": 1.0250503087918952, + "grad_norm": 2.0861008167266846, + "learning_rate": 0.0001743317741683051, + "loss": 0.8279, + "step": 14772 + }, + { + "epoch": 1.0251197002289918, + "grad_norm": 1.8502007722854614, + "learning_rate": 0.0001743280155426984, + "loss": 0.647, + "step": 14773 + }, + { + "epoch": 1.0251890916660884, + "grad_norm": 1.9411252737045288, + "learning_rate": 0.00017432425668244676, + "loss": 0.7503, + "step": 14774 + }, + { + "epoch": 1.025258483103185, + "grad_norm": 2.2978296279907227, + "learning_rate": 0.00017432049758756208, + "loss": 0.6148, + "step": 14775 + }, + { + "epoch": 1.0253278745402816, + "grad_norm": 2.118553638458252, + "learning_rate": 0.00017431673825805619, + "loss": 0.6924, + "step": 14776 + }, + { + "epoch": 1.0253972659773785, + "grad_norm": 1.9490790367126465, + "learning_rate": 0.00017431297869394097, + "loss": 0.6002, + "step": 14777 + }, + { + "epoch": 1.025466657414475, + "grad_norm": 2.0421969890594482, + "learning_rate": 0.00017430921889522835, + "loss": 0.5628, + "step": 14778 + }, + { + "epoch": 1.0255360488515717, + "grad_norm": 1.9569326639175415, + "learning_rate": 0.00017430545886193012, + "loss": 0.8125, + "step": 14779 + }, + { + "epoch": 1.0256054402886683, + "grad_norm": 1.856981635093689, + "learning_rate": 0.00017430169859405817, + "loss": 0.7038, + "step": 14780 + }, + { + "epoch": 1.0256748317257651, + "grad_norm": 1.7686107158660889, + "learning_rate": 0.0001742979380916244, + "loss": 0.6455, + "step": 14781 + }, + { + "epoch": 1.0257442231628617, + "grad_norm": 1.9306458234786987, + "learning_rate": 0.00017429417735464063, + "loss": 0.7194, + "step": 14782 + }, + { + "epoch": 1.0258136145999583, + "grad_norm": 1.8850996494293213, + "learning_rate": 0.00017429041638311875, + "loss": 0.6459, + "step": 14783 + }, + { + "epoch": 1.025883006037055, + "grad_norm": 2.0640201568603516, + "learning_rate": 0.00017428665517707067, + "loss": 0.6468, + "step": 14784 + }, + { + "epoch": 1.0259523974741518, + "grad_norm": 1.7699320316314697, + "learning_rate": 0.00017428289373650824, + "loss": 0.428, + "step": 14785 + }, + { + "epoch": 1.0260217889112484, + "grad_norm": 2.0962374210357666, + "learning_rate": 0.00017427913206144333, + "loss": 0.9093, + "step": 14786 + }, + { + "epoch": 1.026091180348345, + "grad_norm": 2.456446886062622, + "learning_rate": 0.0001742753701518878, + "loss": 0.6219, + "step": 14787 + }, + { + "epoch": 1.0261605717854416, + "grad_norm": 1.5454398393630981, + "learning_rate": 0.0001742716080078536, + "loss": 0.638, + "step": 14788 + }, + { + "epoch": 1.0262299632225382, + "grad_norm": 2.1527063846588135, + "learning_rate": 0.0001742678456293525, + "loss": 0.645, + "step": 14789 + }, + { + "epoch": 1.026299354659635, + "grad_norm": 1.870632529258728, + "learning_rate": 0.00017426408301639645, + "loss": 0.5792, + "step": 14790 + }, + { + "epoch": 1.0263687460967317, + "grad_norm": 1.6573431491851807, + "learning_rate": 0.0001742603201689973, + "loss": 0.5031, + "step": 14791 + }, + { + "epoch": 1.0264381375338283, + "grad_norm": 1.5810520648956299, + "learning_rate": 0.00017425655708716692, + "loss": 0.645, + "step": 14792 + }, + { + "epoch": 1.026507528970925, + "grad_norm": 2.131204843521118, + "learning_rate": 0.0001742527937709172, + "loss": 0.7861, + "step": 14793 + }, + { + "epoch": 1.0265769204080217, + "grad_norm": 1.8464161157608032, + "learning_rate": 0.00017424903022026007, + "loss": 0.6343, + "step": 14794 + }, + { + "epoch": 1.0266463118451183, + "grad_norm": 1.932602047920227, + "learning_rate": 0.00017424526643520732, + "loss": 0.5053, + "step": 14795 + }, + { + "epoch": 1.026715703282215, + "grad_norm": 1.655415654182434, + "learning_rate": 0.0001742415024157709, + "loss": 0.6573, + "step": 14796 + }, + { + "epoch": 1.0267850947193116, + "grad_norm": 2.5206820964813232, + "learning_rate": 0.00017423773816196267, + "loss": 0.888, + "step": 14797 + }, + { + "epoch": 1.0268544861564084, + "grad_norm": 2.2961697578430176, + "learning_rate": 0.00017423397367379451, + "loss": 0.6678, + "step": 14798 + }, + { + "epoch": 1.026923877593505, + "grad_norm": 2.0547256469726562, + "learning_rate": 0.0001742302089512783, + "loss": 0.6542, + "step": 14799 + }, + { + "epoch": 1.0269932690306016, + "grad_norm": 1.6609832048416138, + "learning_rate": 0.00017422644399442595, + "loss": 0.6706, + "step": 14800 + }, + { + "epoch": 1.0270626604676982, + "grad_norm": 1.7505155801773071, + "learning_rate": 0.00017422267880324931, + "loss": 0.6797, + "step": 14801 + }, + { + "epoch": 1.0271320519047948, + "grad_norm": 1.759889006614685, + "learning_rate": 0.0001742189133777603, + "loss": 0.749, + "step": 14802 + }, + { + "epoch": 1.0272014433418917, + "grad_norm": 1.9244098663330078, + "learning_rate": 0.0001742151477179708, + "loss": 0.7024, + "step": 14803 + }, + { + "epoch": 1.0272708347789883, + "grad_norm": 2.5629024505615234, + "learning_rate": 0.00017421138182389267, + "loss": 0.8462, + "step": 14804 + }, + { + "epoch": 1.027340226216085, + "grad_norm": 1.8190311193466187, + "learning_rate": 0.0001742076156955378, + "loss": 0.5282, + "step": 14805 + }, + { + "epoch": 1.0274096176531815, + "grad_norm": 1.5405861139297485, + "learning_rate": 0.00017420384933291812, + "loss": 0.669, + "step": 14806 + }, + { + "epoch": 1.0274790090902783, + "grad_norm": 2.7225425243377686, + "learning_rate": 0.00017420008273604548, + "loss": 0.6073, + "step": 14807 + }, + { + "epoch": 1.027548400527375, + "grad_norm": 1.7506698369979858, + "learning_rate": 0.00017419631590493185, + "loss": 0.7509, + "step": 14808 + }, + { + "epoch": 1.0276177919644716, + "grad_norm": 1.7777056694030762, + "learning_rate": 0.000174192548839589, + "loss": 0.8202, + "step": 14809 + }, + { + "epoch": 1.0276871834015682, + "grad_norm": 1.962000846862793, + "learning_rate": 0.00017418878154002888, + "loss": 0.5048, + "step": 14810 + }, + { + "epoch": 1.027756574838665, + "grad_norm": 1.6130071878433228, + "learning_rate": 0.0001741850140062634, + "loss": 0.4934, + "step": 14811 + }, + { + "epoch": 1.0278259662757616, + "grad_norm": 1.999584674835205, + "learning_rate": 0.00017418124623830444, + "loss": 0.5727, + "step": 14812 + }, + { + "epoch": 1.0278953577128582, + "grad_norm": 1.9745781421661377, + "learning_rate": 0.00017417747823616388, + "loss": 0.7448, + "step": 14813 + }, + { + "epoch": 1.0279647491499548, + "grad_norm": 1.6336774826049805, + "learning_rate": 0.00017417370999985363, + "loss": 0.6599, + "step": 14814 + }, + { + "epoch": 1.0280341405870517, + "grad_norm": 1.61984384059906, + "learning_rate": 0.0001741699415293856, + "loss": 0.7548, + "step": 14815 + }, + { + "epoch": 1.0281035320241483, + "grad_norm": 1.8075987100601196, + "learning_rate": 0.00017416617282477165, + "loss": 0.8888, + "step": 14816 + }, + { + "epoch": 1.028172923461245, + "grad_norm": 2.1125710010528564, + "learning_rate": 0.00017416240388602372, + "loss": 0.7, + "step": 14817 + }, + { + "epoch": 1.0282423148983415, + "grad_norm": 1.5224852561950684, + "learning_rate": 0.00017415863471315367, + "loss": 0.4072, + "step": 14818 + }, + { + "epoch": 1.0283117063354381, + "grad_norm": 1.8089346885681152, + "learning_rate": 0.0001741548653061734, + "loss": 0.7533, + "step": 14819 + }, + { + "epoch": 1.028381097772535, + "grad_norm": 1.9782987833023071, + "learning_rate": 0.00017415109566509484, + "loss": 0.5925, + "step": 14820 + }, + { + "epoch": 1.0284504892096316, + "grad_norm": 1.5622800588607788, + "learning_rate": 0.00017414732578992988, + "loss": 0.6895, + "step": 14821 + }, + { + "epoch": 1.0285198806467282, + "grad_norm": 1.955262303352356, + "learning_rate": 0.0001741435556806904, + "loss": 0.625, + "step": 14822 + }, + { + "epoch": 1.0285892720838248, + "grad_norm": 2.3050339221954346, + "learning_rate": 0.00017413978533738832, + "loss": 0.6728, + "step": 14823 + }, + { + "epoch": 1.0286586635209216, + "grad_norm": 2.2853317260742188, + "learning_rate": 0.00017413601476003553, + "loss": 0.9806, + "step": 14824 + }, + { + "epoch": 1.0287280549580182, + "grad_norm": 2.097916603088379, + "learning_rate": 0.000174132243948644, + "loss": 0.7218, + "step": 14825 + }, + { + "epoch": 1.0287974463951148, + "grad_norm": 2.174755096435547, + "learning_rate": 0.00017412847290322551, + "loss": 0.6338, + "step": 14826 + }, + { + "epoch": 1.0288668378322114, + "grad_norm": 2.313276767730713, + "learning_rate": 0.00017412470162379205, + "loss": 0.6784, + "step": 14827 + }, + { + "epoch": 1.0289362292693083, + "grad_norm": 3.3061468601226807, + "learning_rate": 0.0001741209301103555, + "loss": 0.7603, + "step": 14828 + }, + { + "epoch": 1.0290056207064049, + "grad_norm": 2.239680767059326, + "learning_rate": 0.0001741171583629278, + "loss": 0.791, + "step": 14829 + }, + { + "epoch": 1.0290750121435015, + "grad_norm": 1.2539726495742798, + "learning_rate": 0.00017411338638152085, + "loss": 0.4981, + "step": 14830 + }, + { + "epoch": 1.029144403580598, + "grad_norm": 1.789037823677063, + "learning_rate": 0.00017410961416614647, + "loss": 0.5152, + "step": 14831 + }, + { + "epoch": 1.0292137950176947, + "grad_norm": 1.783174753189087, + "learning_rate": 0.00017410584171681666, + "loss": 0.7241, + "step": 14832 + }, + { + "epoch": 1.0292831864547916, + "grad_norm": 2.432354211807251, + "learning_rate": 0.00017410206903354335, + "loss": 0.8584, + "step": 14833 + }, + { + "epoch": 1.0293525778918882, + "grad_norm": 1.9176911115646362, + "learning_rate": 0.00017409829611633834, + "loss": 0.8135, + "step": 14834 + }, + { + "epoch": 1.0294219693289848, + "grad_norm": 1.8988823890686035, + "learning_rate": 0.00017409452296521366, + "loss": 0.676, + "step": 14835 + }, + { + "epoch": 1.0294913607660814, + "grad_norm": 1.75678551197052, + "learning_rate": 0.00017409074958018114, + "loss": 0.6795, + "step": 14836 + }, + { + "epoch": 1.0295607522031782, + "grad_norm": 1.85078763961792, + "learning_rate": 0.00017408697596125273, + "loss": 0.6687, + "step": 14837 + }, + { + "epoch": 1.0296301436402748, + "grad_norm": 2.5698304176330566, + "learning_rate": 0.00017408320210844034, + "loss": 0.9399, + "step": 14838 + }, + { + "epoch": 1.0296995350773714, + "grad_norm": 2.3792166709899902, + "learning_rate": 0.00017407942802175584, + "loss": 0.493, + "step": 14839 + }, + { + "epoch": 1.029768926514468, + "grad_norm": 1.9001544713974, + "learning_rate": 0.0001740756537012112, + "loss": 0.6258, + "step": 14840 + }, + { + "epoch": 1.0298383179515649, + "grad_norm": 1.8440130949020386, + "learning_rate": 0.00017407187914681834, + "loss": 0.6194, + "step": 14841 + }, + { + "epoch": 1.0299077093886615, + "grad_norm": 1.3357799053192139, + "learning_rate": 0.00017406810435858914, + "loss": 0.5693, + "step": 14842 + }, + { + "epoch": 1.029977100825758, + "grad_norm": 1.730202078819275, + "learning_rate": 0.0001740643293365355, + "loss": 0.9837, + "step": 14843 + }, + { + "epoch": 1.0300464922628547, + "grad_norm": 2.1377792358398438, + "learning_rate": 0.0001740605540806694, + "loss": 0.7319, + "step": 14844 + }, + { + "epoch": 1.0301158836999513, + "grad_norm": 2.63767147064209, + "learning_rate": 0.0001740567785910027, + "loss": 0.9076, + "step": 14845 + }, + { + "epoch": 1.0301852751370482, + "grad_norm": 3.090348482131958, + "learning_rate": 0.00017405300286754734, + "loss": 0.7937, + "step": 14846 + }, + { + "epoch": 1.0302546665741448, + "grad_norm": 1.9152648448944092, + "learning_rate": 0.00017404922691031523, + "loss": 0.7533, + "step": 14847 + }, + { + "epoch": 1.0303240580112414, + "grad_norm": 1.849722146987915, + "learning_rate": 0.00017404545071931835, + "loss": 0.6161, + "step": 14848 + }, + { + "epoch": 1.030393449448338, + "grad_norm": 1.8509639501571655, + "learning_rate": 0.00017404167429456853, + "loss": 0.7397, + "step": 14849 + }, + { + "epoch": 1.0304628408854348, + "grad_norm": 2.060241937637329, + "learning_rate": 0.00017403789763607776, + "loss": 0.7039, + "step": 14850 + }, + { + "epoch": 1.0305322323225314, + "grad_norm": 1.8968569040298462, + "learning_rate": 0.0001740341207438579, + "loss": 0.5597, + "step": 14851 + }, + { + "epoch": 1.030601623759628, + "grad_norm": 1.8909012079238892, + "learning_rate": 0.00017403034361792097, + "loss": 0.7705, + "step": 14852 + }, + { + "epoch": 1.0306710151967247, + "grad_norm": 1.5611653327941895, + "learning_rate": 0.00017402656625827882, + "loss": 0.6466, + "step": 14853 + }, + { + "epoch": 1.0307404066338215, + "grad_norm": 3.3707475662231445, + "learning_rate": 0.00017402278866494335, + "loss": 0.7215, + "step": 14854 + }, + { + "epoch": 1.030809798070918, + "grad_norm": 1.7231042385101318, + "learning_rate": 0.00017401901083792656, + "loss": 0.6374, + "step": 14855 + }, + { + "epoch": 1.0308791895080147, + "grad_norm": 2.633831262588501, + "learning_rate": 0.00017401523277724032, + "loss": 0.8641, + "step": 14856 + }, + { + "epoch": 1.0309485809451113, + "grad_norm": 1.8695570230484009, + "learning_rate": 0.0001740114544828966, + "loss": 0.6613, + "step": 14857 + }, + { + "epoch": 1.031017972382208, + "grad_norm": 2.2239410877227783, + "learning_rate": 0.00017400767595490728, + "loss": 0.8643, + "step": 14858 + }, + { + "epoch": 1.0310873638193048, + "grad_norm": 1.6160080432891846, + "learning_rate": 0.00017400389719328431, + "loss": 0.6882, + "step": 14859 + }, + { + "epoch": 1.0311567552564014, + "grad_norm": 1.844388484954834, + "learning_rate": 0.00017400011819803965, + "loss": 0.6832, + "step": 14860 + }, + { + "epoch": 1.031226146693498, + "grad_norm": 1.8370864391326904, + "learning_rate": 0.00017399633896918518, + "loss": 0.5407, + "step": 14861 + }, + { + "epoch": 1.0312955381305946, + "grad_norm": 3.999936819076538, + "learning_rate": 0.00017399255950673287, + "loss": 0.8057, + "step": 14862 + }, + { + "epoch": 1.0313649295676914, + "grad_norm": 1.9530699253082275, + "learning_rate": 0.00017398877981069464, + "loss": 0.7236, + "step": 14863 + }, + { + "epoch": 1.031434321004788, + "grad_norm": 2.0334532260894775, + "learning_rate": 0.00017398499988108238, + "loss": 0.7785, + "step": 14864 + }, + { + "epoch": 1.0315037124418847, + "grad_norm": 1.8767452239990234, + "learning_rate": 0.0001739812197179081, + "loss": 0.6335, + "step": 14865 + }, + { + "epoch": 1.0315731038789813, + "grad_norm": 1.5219939947128296, + "learning_rate": 0.00017397743932118367, + "loss": 0.487, + "step": 14866 + }, + { + "epoch": 1.031642495316078, + "grad_norm": 1.7916289567947388, + "learning_rate": 0.00017397365869092104, + "loss": 0.7469, + "step": 14867 + }, + { + "epoch": 1.0317118867531747, + "grad_norm": 1.7560803890228271, + "learning_rate": 0.00017396987782713214, + "loss": 0.7045, + "step": 14868 + }, + { + "epoch": 1.0317812781902713, + "grad_norm": 1.98279869556427, + "learning_rate": 0.000173966096729829, + "loss": 0.6713, + "step": 14869 + }, + { + "epoch": 1.031850669627368, + "grad_norm": 1.775228500366211, + "learning_rate": 0.00017396231539902339, + "loss": 0.7219, + "step": 14870 + }, + { + "epoch": 1.0319200610644645, + "grad_norm": 2.933563470840454, + "learning_rate": 0.00017395853383472734, + "loss": 0.7466, + "step": 14871 + }, + { + "epoch": 1.0319894525015614, + "grad_norm": 1.7671842575073242, + "learning_rate": 0.0001739547520369528, + "loss": 0.6277, + "step": 14872 + }, + { + "epoch": 1.032058843938658, + "grad_norm": 2.2418408393859863, + "learning_rate": 0.00017395097000571169, + "loss": 0.5441, + "step": 14873 + }, + { + "epoch": 1.0321282353757546, + "grad_norm": 1.9945639371871948, + "learning_rate": 0.00017394718774101594, + "loss": 0.5715, + "step": 14874 + }, + { + "epoch": 1.0321976268128512, + "grad_norm": 2.8578414916992188, + "learning_rate": 0.0001739434052428775, + "loss": 0.6753, + "step": 14875 + }, + { + "epoch": 1.032267018249948, + "grad_norm": 3.6587953567504883, + "learning_rate": 0.0001739396225113083, + "loss": 0.7264, + "step": 14876 + }, + { + "epoch": 1.0323364096870447, + "grad_norm": 1.4752817153930664, + "learning_rate": 0.0001739358395463203, + "loss": 0.66, + "step": 14877 + }, + { + "epoch": 1.0324058011241413, + "grad_norm": 2.0751047134399414, + "learning_rate": 0.00017393205634792542, + "loss": 0.704, + "step": 14878 + }, + { + "epoch": 1.0324751925612379, + "grad_norm": 1.9742473363876343, + "learning_rate": 0.00017392827291613566, + "loss": 0.55, + "step": 14879 + }, + { + "epoch": 1.0325445839983347, + "grad_norm": 2.2293317317962646, + "learning_rate": 0.00017392448925096287, + "loss": 0.8648, + "step": 14880 + }, + { + "epoch": 1.0326139754354313, + "grad_norm": 1.9785881042480469, + "learning_rate": 0.00017392070535241904, + "loss": 0.7867, + "step": 14881 + }, + { + "epoch": 1.032683366872528, + "grad_norm": 2.1771621704101562, + "learning_rate": 0.00017391692122051614, + "loss": 0.7208, + "step": 14882 + }, + { + "epoch": 1.0327527583096245, + "grad_norm": 2.230212450027466, + "learning_rate": 0.00017391313685526612, + "loss": 0.9292, + "step": 14883 + }, + { + "epoch": 1.0328221497467212, + "grad_norm": 2.1112332344055176, + "learning_rate": 0.00017390935225668086, + "loss": 0.5209, + "step": 14884 + }, + { + "epoch": 1.032891541183818, + "grad_norm": 2.143770217895508, + "learning_rate": 0.00017390556742477237, + "loss": 0.8149, + "step": 14885 + }, + { + "epoch": 1.0329609326209146, + "grad_norm": 2.199000835418701, + "learning_rate": 0.0001739017823595526, + "loss": 0.9548, + "step": 14886 + }, + { + "epoch": 1.0330303240580112, + "grad_norm": 3.2871484756469727, + "learning_rate": 0.00017389799706103346, + "loss": 0.7962, + "step": 14887 + }, + { + "epoch": 1.0330997154951078, + "grad_norm": 1.9590272903442383, + "learning_rate": 0.0001738942115292269, + "loss": 0.6348, + "step": 14888 + }, + { + "epoch": 1.0331691069322047, + "grad_norm": 1.5336366891860962, + "learning_rate": 0.0001738904257641449, + "loss": 0.6959, + "step": 14889 + }, + { + "epoch": 1.0332384983693013, + "grad_norm": 1.9289600849151611, + "learning_rate": 0.0001738866397657994, + "loss": 0.7804, + "step": 14890 + }, + { + "epoch": 1.0333078898063979, + "grad_norm": 1.8399443626403809, + "learning_rate": 0.00017388285353420234, + "loss": 0.7122, + "step": 14891 + }, + { + "epoch": 1.0333772812434945, + "grad_norm": 2.124507188796997, + "learning_rate": 0.0001738790670693657, + "loss": 0.5983, + "step": 14892 + }, + { + "epoch": 1.0334466726805913, + "grad_norm": 1.6866514682769775, + "learning_rate": 0.00017387528037130138, + "loss": 0.5368, + "step": 14893 + }, + { + "epoch": 1.033516064117688, + "grad_norm": 1.9481033086776733, + "learning_rate": 0.0001738714934400214, + "loss": 0.8487, + "step": 14894 + }, + { + "epoch": 1.0335854555547845, + "grad_norm": 2.2198941707611084, + "learning_rate": 0.0001738677062755377, + "loss": 0.77, + "step": 14895 + }, + { + "epoch": 1.0336548469918811, + "grad_norm": 1.6543611288070679, + "learning_rate": 0.0001738639188778622, + "loss": 0.4883, + "step": 14896 + }, + { + "epoch": 1.0337242384289778, + "grad_norm": 1.9206241369247437, + "learning_rate": 0.0001738601312470069, + "loss": 0.7053, + "step": 14897 + }, + { + "epoch": 1.0337936298660746, + "grad_norm": 1.7919121980667114, + "learning_rate": 0.0001738563433829837, + "loss": 0.8076, + "step": 14898 + }, + { + "epoch": 1.0338630213031712, + "grad_norm": 2.0642030239105225, + "learning_rate": 0.00017385255528580462, + "loss": 0.6981, + "step": 14899 + }, + { + "epoch": 1.0339324127402678, + "grad_norm": 2.0508482456207275, + "learning_rate": 0.00017384876695548157, + "loss": 0.6546, + "step": 14900 + }, + { + "epoch": 1.0340018041773644, + "grad_norm": 2.2057433128356934, + "learning_rate": 0.0001738449783920266, + "loss": 0.6015, + "step": 14901 + }, + { + "epoch": 1.0340711956144613, + "grad_norm": 2.426286458969116, + "learning_rate": 0.00017384118959545153, + "loss": 0.5968, + "step": 14902 + }, + { + "epoch": 1.0341405870515579, + "grad_norm": 2.220848321914673, + "learning_rate": 0.00017383740056576842, + "loss": 0.8894, + "step": 14903 + }, + { + "epoch": 1.0342099784886545, + "grad_norm": 2.6112539768218994, + "learning_rate": 0.00017383361130298916, + "loss": 0.6975, + "step": 14904 + }, + { + "epoch": 1.034279369925751, + "grad_norm": 1.8916634321212769, + "learning_rate": 0.0001738298218071258, + "loss": 0.6143, + "step": 14905 + }, + { + "epoch": 1.034348761362848, + "grad_norm": 1.7485806941986084, + "learning_rate": 0.00017382603207819023, + "loss": 0.547, + "step": 14906 + }, + { + "epoch": 1.0344181527999445, + "grad_norm": 1.6807979345321655, + "learning_rate": 0.00017382224211619443, + "loss": 0.558, + "step": 14907 + }, + { + "epoch": 1.0344875442370411, + "grad_norm": 2.4811718463897705, + "learning_rate": 0.00017381845192115044, + "loss": 0.5988, + "step": 14908 + }, + { + "epoch": 1.0345569356741378, + "grad_norm": 1.4490630626678467, + "learning_rate": 0.00017381466149307012, + "loss": 0.6596, + "step": 14909 + }, + { + "epoch": 1.0346263271112344, + "grad_norm": 1.6730434894561768, + "learning_rate": 0.00017381087083196548, + "loss": 0.7955, + "step": 14910 + }, + { + "epoch": 1.0346957185483312, + "grad_norm": 2.182373046875, + "learning_rate": 0.0001738070799378485, + "loss": 0.8585, + "step": 14911 + }, + { + "epoch": 1.0347651099854278, + "grad_norm": 1.7281932830810547, + "learning_rate": 0.0001738032888107311, + "loss": 0.6944, + "step": 14912 + }, + { + "epoch": 1.0348345014225244, + "grad_norm": 2.130253553390503, + "learning_rate": 0.00017379949745062532, + "loss": 0.8233, + "step": 14913 + }, + { + "epoch": 1.034903892859621, + "grad_norm": 1.7512234449386597, + "learning_rate": 0.00017379570585754307, + "loss": 0.6859, + "step": 14914 + }, + { + "epoch": 1.0349732842967179, + "grad_norm": 1.5025237798690796, + "learning_rate": 0.00017379191403149631, + "loss": 0.5757, + "step": 14915 + }, + { + "epoch": 1.0350426757338145, + "grad_norm": 1.744010329246521, + "learning_rate": 0.00017378812197249707, + "loss": 0.6729, + "step": 14916 + }, + { + "epoch": 1.035112067170911, + "grad_norm": 1.6752632856369019, + "learning_rate": 0.00017378432968055728, + "loss": 0.7775, + "step": 14917 + }, + { + "epoch": 1.0351814586080077, + "grad_norm": 2.1634366512298584, + "learning_rate": 0.0001737805371556889, + "loss": 0.7094, + "step": 14918 + }, + { + "epoch": 1.0352508500451045, + "grad_norm": 2.2529497146606445, + "learning_rate": 0.00017377674439790397, + "loss": 0.7607, + "step": 14919 + }, + { + "epoch": 1.0353202414822011, + "grad_norm": 4.243276119232178, + "learning_rate": 0.00017377295140721442, + "loss": 0.7796, + "step": 14920 + }, + { + "epoch": 1.0353896329192978, + "grad_norm": 1.9998888969421387, + "learning_rate": 0.00017376915818363218, + "loss": 0.5224, + "step": 14921 + }, + { + "epoch": 1.0354590243563944, + "grad_norm": 2.188851833343506, + "learning_rate": 0.0001737653647271693, + "loss": 0.7019, + "step": 14922 + }, + { + "epoch": 1.035528415793491, + "grad_norm": 1.821403980255127, + "learning_rate": 0.0001737615710378377, + "loss": 0.8772, + "step": 14923 + }, + { + "epoch": 1.0355978072305878, + "grad_norm": 1.7057164907455444, + "learning_rate": 0.0001737577771156494, + "loss": 0.6112, + "step": 14924 + }, + { + "epoch": 1.0356671986676844, + "grad_norm": 1.6446962356567383, + "learning_rate": 0.00017375398296061635, + "loss": 0.4357, + "step": 14925 + }, + { + "epoch": 1.035736590104781, + "grad_norm": 2.5557568073272705, + "learning_rate": 0.0001737501885727505, + "loss": 0.6833, + "step": 14926 + }, + { + "epoch": 1.0358059815418776, + "grad_norm": 1.508507251739502, + "learning_rate": 0.0001737463939520639, + "loss": 0.5245, + "step": 14927 + }, + { + "epoch": 1.0358753729789745, + "grad_norm": 2.6975557804107666, + "learning_rate": 0.00017374259909856847, + "loss": 1.0633, + "step": 14928 + }, + { + "epoch": 1.035944764416071, + "grad_norm": 1.6549534797668457, + "learning_rate": 0.0001737388040122762, + "loss": 0.5173, + "step": 14929 + }, + { + "epoch": 1.0360141558531677, + "grad_norm": 2.2624595165252686, + "learning_rate": 0.00017373500869319913, + "loss": 0.8378, + "step": 14930 + }, + { + "epoch": 1.0360835472902643, + "grad_norm": 1.7353394031524658, + "learning_rate": 0.00017373121314134916, + "loss": 0.7519, + "step": 14931 + }, + { + "epoch": 1.0361529387273611, + "grad_norm": 1.6629383563995361, + "learning_rate": 0.0001737274173567383, + "loss": 0.6482, + "step": 14932 + }, + { + "epoch": 1.0362223301644578, + "grad_norm": 1.7652981281280518, + "learning_rate": 0.00017372362133937855, + "loss": 0.7125, + "step": 14933 + }, + { + "epoch": 1.0362917216015544, + "grad_norm": 2.085512638092041, + "learning_rate": 0.00017371982508928186, + "loss": 0.6501, + "step": 14934 + }, + { + "epoch": 1.036361113038651, + "grad_norm": 2.1495933532714844, + "learning_rate": 0.00017371602860646025, + "loss": 0.6881, + "step": 14935 + }, + { + "epoch": 1.0364305044757476, + "grad_norm": 2.37380051612854, + "learning_rate": 0.0001737122318909257, + "loss": 0.7461, + "step": 14936 + }, + { + "epoch": 1.0364998959128444, + "grad_norm": 2.4352567195892334, + "learning_rate": 0.00017370843494269017, + "loss": 1.0669, + "step": 14937 + }, + { + "epoch": 1.036569287349941, + "grad_norm": 1.979030966758728, + "learning_rate": 0.00017370463776176566, + "loss": 0.7203, + "step": 14938 + }, + { + "epoch": 1.0366386787870376, + "grad_norm": 2.4698266983032227, + "learning_rate": 0.00017370084034816418, + "loss": 0.7316, + "step": 14939 + }, + { + "epoch": 1.0367080702241342, + "grad_norm": 2.1343507766723633, + "learning_rate": 0.00017369704270189766, + "loss": 0.8259, + "step": 14940 + }, + { + "epoch": 1.036777461661231, + "grad_norm": 1.8293006420135498, + "learning_rate": 0.00017369324482297818, + "loss": 0.8133, + "step": 14941 + }, + { + "epoch": 1.0368468530983277, + "grad_norm": 1.9253275394439697, + "learning_rate": 0.00017368944671141762, + "loss": 0.7524, + "step": 14942 + }, + { + "epoch": 1.0369162445354243, + "grad_norm": 6.12160587310791, + "learning_rate": 0.00017368564836722807, + "loss": 0.6294, + "step": 14943 + }, + { + "epoch": 1.036985635972521, + "grad_norm": 1.6832163333892822, + "learning_rate": 0.00017368184979042145, + "loss": 0.5808, + "step": 14944 + }, + { + "epoch": 1.0370550274096177, + "grad_norm": 2.589195966720581, + "learning_rate": 0.0001736780509810098, + "loss": 0.8546, + "step": 14945 + }, + { + "epoch": 1.0371244188467144, + "grad_norm": 1.5748225450515747, + "learning_rate": 0.00017367425193900508, + "loss": 0.6975, + "step": 14946 + }, + { + "epoch": 1.037193810283811, + "grad_norm": 2.4397130012512207, + "learning_rate": 0.00017367045266441928, + "loss": 0.6796, + "step": 14947 + }, + { + "epoch": 1.0372632017209076, + "grad_norm": 1.8748990297317505, + "learning_rate": 0.00017366665315726444, + "loss": 0.7981, + "step": 14948 + }, + { + "epoch": 1.0373325931580042, + "grad_norm": 1.7586920261383057, + "learning_rate": 0.00017366285341755252, + "loss": 0.4868, + "step": 14949 + }, + { + "epoch": 1.037401984595101, + "grad_norm": 2.1557395458221436, + "learning_rate": 0.0001736590534452955, + "loss": 0.8808, + "step": 14950 + }, + { + "epoch": 1.0374713760321976, + "grad_norm": 1.8568909168243408, + "learning_rate": 0.0001736552532405054, + "loss": 0.5524, + "step": 14951 + }, + { + "epoch": 1.0375407674692942, + "grad_norm": 1.8527525663375854, + "learning_rate": 0.0001736514528031942, + "loss": 0.8162, + "step": 14952 + }, + { + "epoch": 1.0376101589063909, + "grad_norm": 1.8960871696472168, + "learning_rate": 0.00017364765213337394, + "loss": 0.6314, + "step": 14953 + }, + { + "epoch": 1.0376795503434877, + "grad_norm": 2.19226336479187, + "learning_rate": 0.00017364385123105655, + "loss": 0.6635, + "step": 14954 + }, + { + "epoch": 1.0377489417805843, + "grad_norm": 2.37205171585083, + "learning_rate": 0.00017364005009625407, + "loss": 0.9284, + "step": 14955 + }, + { + "epoch": 1.037818333217681, + "grad_norm": 2.703146457672119, + "learning_rate": 0.0001736362487289785, + "loss": 0.9128, + "step": 14956 + }, + { + "epoch": 1.0378877246547775, + "grad_norm": 1.7295632362365723, + "learning_rate": 0.00017363244712924187, + "loss": 0.8112, + "step": 14957 + }, + { + "epoch": 1.0379571160918744, + "grad_norm": 1.9229707717895508, + "learning_rate": 0.00017362864529705613, + "loss": 0.6114, + "step": 14958 + }, + { + "epoch": 1.038026507528971, + "grad_norm": 1.8568885326385498, + "learning_rate": 0.00017362484323243328, + "loss": 0.7484, + "step": 14959 + }, + { + "epoch": 1.0380958989660676, + "grad_norm": 1.5755927562713623, + "learning_rate": 0.00017362104093538537, + "loss": 0.5361, + "step": 14960 + }, + { + "epoch": 1.0381652904031642, + "grad_norm": 2.1637156009674072, + "learning_rate": 0.00017361723840592433, + "loss": 0.6644, + "step": 14961 + }, + { + "epoch": 1.0382346818402608, + "grad_norm": 2.207895517349243, + "learning_rate": 0.00017361343564406225, + "loss": 0.7949, + "step": 14962 + }, + { + "epoch": 1.0383040732773576, + "grad_norm": 1.817828893661499, + "learning_rate": 0.00017360963264981106, + "loss": 0.571, + "step": 14963 + }, + { + "epoch": 1.0383734647144542, + "grad_norm": 1.9603747129440308, + "learning_rate": 0.00017360582942318285, + "loss": 0.7955, + "step": 14964 + }, + { + "epoch": 1.0384428561515509, + "grad_norm": 1.7662266492843628, + "learning_rate": 0.00017360202596418956, + "loss": 0.7894, + "step": 14965 + }, + { + "epoch": 1.0385122475886475, + "grad_norm": 2.7191689014434814, + "learning_rate": 0.00017359822227284318, + "loss": 0.8655, + "step": 14966 + }, + { + "epoch": 1.0385816390257443, + "grad_norm": 2.2125444412231445, + "learning_rate": 0.0001735944183491558, + "loss": 0.4375, + "step": 14967 + }, + { + "epoch": 1.038651030462841, + "grad_norm": 2.324094295501709, + "learning_rate": 0.00017359061419313932, + "loss": 0.7788, + "step": 14968 + }, + { + "epoch": 1.0387204218999375, + "grad_norm": 1.7654856443405151, + "learning_rate": 0.00017358680980480586, + "loss": 0.7779, + "step": 14969 + }, + { + "epoch": 1.0387898133370341, + "grad_norm": 1.780844807624817, + "learning_rate": 0.00017358300518416738, + "loss": 0.8265, + "step": 14970 + }, + { + "epoch": 1.038859204774131, + "grad_norm": 2.385221481323242, + "learning_rate": 0.00017357920033123585, + "loss": 0.5925, + "step": 14971 + }, + { + "epoch": 1.0389285962112276, + "grad_norm": 2.2273383140563965, + "learning_rate": 0.00017357539524602333, + "loss": 0.4483, + "step": 14972 + }, + { + "epoch": 1.0389979876483242, + "grad_norm": 1.7309855222702026, + "learning_rate": 0.00017357158992854186, + "loss": 0.6625, + "step": 14973 + }, + { + "epoch": 1.0390673790854208, + "grad_norm": 2.1117937564849854, + "learning_rate": 0.00017356778437880338, + "loss": 0.5663, + "step": 14974 + }, + { + "epoch": 1.0391367705225176, + "grad_norm": 1.7178741693496704, + "learning_rate": 0.00017356397859681993, + "loss": 0.6888, + "step": 14975 + }, + { + "epoch": 1.0392061619596142, + "grad_norm": 1.9403311014175415, + "learning_rate": 0.00017356017258260357, + "loss": 0.7433, + "step": 14976 + }, + { + "epoch": 1.0392755533967108, + "grad_norm": 4.7524566650390625, + "learning_rate": 0.0001735563663361663, + "loss": 0.6794, + "step": 14977 + }, + { + "epoch": 1.0393449448338075, + "grad_norm": 2.794292449951172, + "learning_rate": 0.00017355255985752007, + "loss": 0.6972, + "step": 14978 + }, + { + "epoch": 1.039414336270904, + "grad_norm": 1.6273120641708374, + "learning_rate": 0.00017354875314667693, + "loss": 0.8315, + "step": 14979 + }, + { + "epoch": 1.039483727708001, + "grad_norm": 1.434116244316101, + "learning_rate": 0.00017354494620364897, + "loss": 0.5541, + "step": 14980 + }, + { + "epoch": 1.0395531191450975, + "grad_norm": 1.9781168699264526, + "learning_rate": 0.00017354113902844808, + "loss": 0.6669, + "step": 14981 + }, + { + "epoch": 1.0396225105821941, + "grad_norm": 2.4588146209716797, + "learning_rate": 0.0001735373316210864, + "loss": 0.7034, + "step": 14982 + }, + { + "epoch": 1.0396919020192907, + "grad_norm": 2.1441142559051514, + "learning_rate": 0.00017353352398157584, + "loss": 0.9789, + "step": 14983 + }, + { + "epoch": 1.0397612934563876, + "grad_norm": 2.105149984359741, + "learning_rate": 0.00017352971610992853, + "loss": 0.9086, + "step": 14984 + }, + { + "epoch": 1.0398306848934842, + "grad_norm": 2.6643829345703125, + "learning_rate": 0.00017352590800615643, + "loss": 0.5851, + "step": 14985 + }, + { + "epoch": 1.0399000763305808, + "grad_norm": 1.908864140510559, + "learning_rate": 0.00017352209967027156, + "loss": 0.8019, + "step": 14986 + }, + { + "epoch": 1.0399694677676774, + "grad_norm": 5.3275146484375, + "learning_rate": 0.00017351829110228594, + "loss": 0.7641, + "step": 14987 + }, + { + "epoch": 1.0400388592047742, + "grad_norm": 1.6728523969650269, + "learning_rate": 0.00017351448230221162, + "loss": 0.4507, + "step": 14988 + }, + { + "epoch": 1.0401082506418708, + "grad_norm": 1.9148505926132202, + "learning_rate": 0.0001735106732700606, + "loss": 0.8125, + "step": 14989 + }, + { + "epoch": 1.0401776420789675, + "grad_norm": 2.0217230319976807, + "learning_rate": 0.0001735068640058449, + "loss": 0.6814, + "step": 14990 + }, + { + "epoch": 1.040247033516064, + "grad_norm": 4.166388511657715, + "learning_rate": 0.00017350305450957658, + "loss": 0.7188, + "step": 14991 + }, + { + "epoch": 1.0403164249531607, + "grad_norm": 3.154322385787964, + "learning_rate": 0.00017349924478126766, + "loss": 0.8766, + "step": 14992 + }, + { + "epoch": 1.0403858163902575, + "grad_norm": 1.9800071716308594, + "learning_rate": 0.00017349543482093012, + "loss": 0.8162, + "step": 14993 + }, + { + "epoch": 1.0404552078273541, + "grad_norm": 1.6040657758712769, + "learning_rate": 0.00017349162462857603, + "loss": 0.68, + "step": 14994 + }, + { + "epoch": 1.0405245992644507, + "grad_norm": 2.2574753761291504, + "learning_rate": 0.0001734878142042174, + "loss": 0.6144, + "step": 14995 + }, + { + "epoch": 1.0405939907015473, + "grad_norm": 1.8214701414108276, + "learning_rate": 0.0001734840035478663, + "loss": 0.5964, + "step": 14996 + }, + { + "epoch": 1.0406633821386442, + "grad_norm": 1.948969841003418, + "learning_rate": 0.00017348019265953467, + "loss": 0.6085, + "step": 14997 + }, + { + "epoch": 1.0407327735757408, + "grad_norm": 2.68415904045105, + "learning_rate": 0.00017347638153923463, + "loss": 1.0679, + "step": 14998 + }, + { + "epoch": 1.0408021650128374, + "grad_norm": 2.037242889404297, + "learning_rate": 0.00017347257018697818, + "loss": 0.989, + "step": 14999 + }, + { + "epoch": 1.040871556449934, + "grad_norm": 1.712904453277588, + "learning_rate": 0.00017346875860277733, + "loss": 0.7137, + "step": 15000 + }, + { + "epoch": 1.0409409478870308, + "grad_norm": 2.0542590618133545, + "learning_rate": 0.00017346494678664417, + "loss": 0.6823, + "step": 15001 + }, + { + "epoch": 1.0410103393241275, + "grad_norm": 2.3034090995788574, + "learning_rate": 0.00017346113473859067, + "loss": 0.8993, + "step": 15002 + }, + { + "epoch": 1.041079730761224, + "grad_norm": 2.0762927532196045, + "learning_rate": 0.00017345732245862888, + "loss": 0.6342, + "step": 15003 + }, + { + "epoch": 1.0411491221983207, + "grad_norm": 1.6412456035614014, + "learning_rate": 0.00017345350994677088, + "loss": 0.4948, + "step": 15004 + }, + { + "epoch": 1.0412185136354173, + "grad_norm": 1.8381867408752441, + "learning_rate": 0.00017344969720302865, + "loss": 0.9346, + "step": 15005 + }, + { + "epoch": 1.0412879050725141, + "grad_norm": 1.1863747835159302, + "learning_rate": 0.00017344588422741426, + "loss": 0.5185, + "step": 15006 + }, + { + "epoch": 1.0413572965096107, + "grad_norm": 1.5569363832473755, + "learning_rate": 0.0001734420710199397, + "loss": 0.6523, + "step": 15007 + }, + { + "epoch": 1.0414266879467073, + "grad_norm": 1.9280009269714355, + "learning_rate": 0.0001734382575806171, + "loss": 0.5575, + "step": 15008 + }, + { + "epoch": 1.041496079383804, + "grad_norm": 1.7160227298736572, + "learning_rate": 0.0001734344439094584, + "loss": 0.6617, + "step": 15009 + }, + { + "epoch": 1.0415654708209008, + "grad_norm": 1.5930397510528564, + "learning_rate": 0.00017343063000647567, + "loss": 0.504, + "step": 15010 + }, + { + "epoch": 1.0416348622579974, + "grad_norm": 2.3361005783081055, + "learning_rate": 0.00017342681587168102, + "loss": 0.8671, + "step": 15011 + }, + { + "epoch": 1.041704253695094, + "grad_norm": 2.772191047668457, + "learning_rate": 0.00017342300150508636, + "loss": 0.9337, + "step": 15012 + }, + { + "epoch": 1.0417736451321906, + "grad_norm": 1.9630697965621948, + "learning_rate": 0.00017341918690670383, + "loss": 0.7447, + "step": 15013 + }, + { + "epoch": 1.0418430365692875, + "grad_norm": 2.9110114574432373, + "learning_rate": 0.00017341537207654547, + "loss": 0.6275, + "step": 15014 + }, + { + "epoch": 1.041912428006384, + "grad_norm": 1.6494461297988892, + "learning_rate": 0.00017341155701462328, + "loss": 0.5023, + "step": 15015 + }, + { + "epoch": 1.0419818194434807, + "grad_norm": 1.7252062559127808, + "learning_rate": 0.00017340774172094932, + "loss": 0.8966, + "step": 15016 + }, + { + "epoch": 1.0420512108805773, + "grad_norm": 2.1802074909210205, + "learning_rate": 0.00017340392619553564, + "loss": 0.5239, + "step": 15017 + }, + { + "epoch": 1.042120602317674, + "grad_norm": 1.737647533416748, + "learning_rate": 0.00017340011043839426, + "loss": 0.8298, + "step": 15018 + }, + { + "epoch": 1.0421899937547707, + "grad_norm": 1.8213526010513306, + "learning_rate": 0.00017339629444953727, + "loss": 0.796, + "step": 15019 + }, + { + "epoch": 1.0422593851918673, + "grad_norm": 1.7382760047912598, + "learning_rate": 0.00017339247822897668, + "loss": 0.5676, + "step": 15020 + }, + { + "epoch": 1.042328776628964, + "grad_norm": 2.9297072887420654, + "learning_rate": 0.00017338866177672458, + "loss": 0.5751, + "step": 15021 + }, + { + "epoch": 1.0423981680660606, + "grad_norm": 1.8560121059417725, + "learning_rate": 0.00017338484509279298, + "loss": 0.5985, + "step": 15022 + }, + { + "epoch": 1.0424675595031574, + "grad_norm": 1.6707566976547241, + "learning_rate": 0.00017338102817719393, + "loss": 0.7064, + "step": 15023 + }, + { + "epoch": 1.042536950940254, + "grad_norm": 2.3328356742858887, + "learning_rate": 0.00017337721102993946, + "loss": 0.8332, + "step": 15024 + }, + { + "epoch": 1.0426063423773506, + "grad_norm": 2.2923734188079834, + "learning_rate": 0.0001733733936510417, + "loss": 0.7413, + "step": 15025 + }, + { + "epoch": 1.0426757338144472, + "grad_norm": 2.314953327178955, + "learning_rate": 0.0001733695760405126, + "loss": 0.6181, + "step": 15026 + }, + { + "epoch": 1.042745125251544, + "grad_norm": 1.8456696271896362, + "learning_rate": 0.00017336575819836429, + "loss": 0.7464, + "step": 15027 + }, + { + "epoch": 1.0428145166886407, + "grad_norm": 1.9476237297058105, + "learning_rate": 0.0001733619401246088, + "loss": 0.5413, + "step": 15028 + }, + { + "epoch": 1.0428839081257373, + "grad_norm": 1.9339525699615479, + "learning_rate": 0.00017335812181925813, + "loss": 0.4682, + "step": 15029 + }, + { + "epoch": 1.042953299562834, + "grad_norm": 1.6216825246810913, + "learning_rate": 0.00017335430328232443, + "loss": 0.6223, + "step": 15030 + }, + { + "epoch": 1.0430226909999305, + "grad_norm": 1.8709921836853027, + "learning_rate": 0.0001733504845138197, + "loss": 0.7285, + "step": 15031 + }, + { + "epoch": 1.0430920824370273, + "grad_norm": 1.6279246807098389, + "learning_rate": 0.00017334666551375598, + "loss": 0.6637, + "step": 15032 + }, + { + "epoch": 1.043161473874124, + "grad_norm": 2.385206937789917, + "learning_rate": 0.00017334284628214538, + "loss": 0.6413, + "step": 15033 + }, + { + "epoch": 1.0432308653112206, + "grad_norm": 1.891688346862793, + "learning_rate": 0.00017333902681899988, + "loss": 0.6994, + "step": 15034 + }, + { + "epoch": 1.0433002567483172, + "grad_norm": 1.9035717248916626, + "learning_rate": 0.0001733352071243316, + "loss": 0.4058, + "step": 15035 + }, + { + "epoch": 1.043369648185414, + "grad_norm": 2.0641121864318848, + "learning_rate": 0.0001733313871981526, + "loss": 0.6152, + "step": 15036 + }, + { + "epoch": 1.0434390396225106, + "grad_norm": 2.5364186763763428, + "learning_rate": 0.0001733275670404749, + "loss": 0.5966, + "step": 15037 + }, + { + "epoch": 1.0435084310596072, + "grad_norm": 1.5506616830825806, + "learning_rate": 0.00017332374665131057, + "loss": 0.4958, + "step": 15038 + }, + { + "epoch": 1.0435778224967038, + "grad_norm": 2.174886703491211, + "learning_rate": 0.0001733199260306717, + "loss": 0.7617, + "step": 15039 + }, + { + "epoch": 1.0436472139338007, + "grad_norm": 2.120309352874756, + "learning_rate": 0.0001733161051785703, + "loss": 0.8393, + "step": 15040 + }, + { + "epoch": 1.0437166053708973, + "grad_norm": 1.8411109447479248, + "learning_rate": 0.0001733122840950185, + "loss": 0.5694, + "step": 15041 + }, + { + "epoch": 1.0437859968079939, + "grad_norm": 2.15095853805542, + "learning_rate": 0.00017330846278002828, + "loss": 0.7199, + "step": 15042 + }, + { + "epoch": 1.0438553882450905, + "grad_norm": 3.0108792781829834, + "learning_rate": 0.00017330464123361177, + "loss": 0.6684, + "step": 15043 + }, + { + "epoch": 1.0439247796821873, + "grad_norm": 1.4418110847473145, + "learning_rate": 0.00017330081945578102, + "loss": 0.5669, + "step": 15044 + }, + { + "epoch": 1.043994171119284, + "grad_norm": 3.0087156295776367, + "learning_rate": 0.0001732969974465481, + "loss": 0.9847, + "step": 15045 + }, + { + "epoch": 1.0440635625563806, + "grad_norm": 1.9876439571380615, + "learning_rate": 0.00017329317520592503, + "loss": 0.5229, + "step": 15046 + }, + { + "epoch": 1.0441329539934772, + "grad_norm": 2.1716079711914062, + "learning_rate": 0.00017328935273392394, + "loss": 0.6712, + "step": 15047 + }, + { + "epoch": 1.0442023454305738, + "grad_norm": 2.0710527896881104, + "learning_rate": 0.00017328553003055684, + "loss": 0.6531, + "step": 15048 + }, + { + "epoch": 1.0442717368676706, + "grad_norm": 1.9004751443862915, + "learning_rate": 0.00017328170709583583, + "loss": 0.7583, + "step": 15049 + }, + { + "epoch": 1.0443411283047672, + "grad_norm": 2.5108489990234375, + "learning_rate": 0.000173277883929773, + "loss": 0.8699, + "step": 15050 + }, + { + "epoch": 1.0444105197418638, + "grad_norm": 2.0169246196746826, + "learning_rate": 0.00017327406053238039, + "loss": 0.5707, + "step": 15051 + }, + { + "epoch": 1.0444799111789604, + "grad_norm": 2.192966938018799, + "learning_rate": 0.00017327023690367005, + "loss": 0.6361, + "step": 15052 + }, + { + "epoch": 1.0445493026160573, + "grad_norm": 2.0758209228515625, + "learning_rate": 0.00017326641304365404, + "loss": 0.6175, + "step": 15053 + }, + { + "epoch": 1.0446186940531539, + "grad_norm": 2.0253872871398926, + "learning_rate": 0.00017326258895234453, + "loss": 0.6227, + "step": 15054 + }, + { + "epoch": 1.0446880854902505, + "grad_norm": 2.104114532470703, + "learning_rate": 0.0001732587646297535, + "loss": 0.5941, + "step": 15055 + }, + { + "epoch": 1.044757476927347, + "grad_norm": 1.9119601249694824, + "learning_rate": 0.00017325494007589302, + "loss": 0.7688, + "step": 15056 + }, + { + "epoch": 1.044826868364444, + "grad_norm": 2.107532262802124, + "learning_rate": 0.00017325111529077524, + "loss": 0.825, + "step": 15057 + }, + { + "epoch": 1.0448962598015406, + "grad_norm": 1.9951870441436768, + "learning_rate": 0.0001732472902744122, + "loss": 0.6942, + "step": 15058 + }, + { + "epoch": 1.0449656512386372, + "grad_norm": 1.6023632287979126, + "learning_rate": 0.0001732434650268159, + "loss": 0.6768, + "step": 15059 + }, + { + "epoch": 1.0450350426757338, + "grad_norm": 2.071993827819824, + "learning_rate": 0.00017323963954799852, + "loss": 0.5898, + "step": 15060 + }, + { + "epoch": 1.0451044341128304, + "grad_norm": 2.8265774250030518, + "learning_rate": 0.00017323581383797207, + "loss": 0.6576, + "step": 15061 + }, + { + "epoch": 1.0451738255499272, + "grad_norm": 1.502506136894226, + "learning_rate": 0.00017323198789674868, + "loss": 0.5225, + "step": 15062 + }, + { + "epoch": 1.0452432169870238, + "grad_norm": 2.0665230751037598, + "learning_rate": 0.0001732281617243404, + "loss": 0.6092, + "step": 15063 + }, + { + "epoch": 1.0453126084241204, + "grad_norm": 2.233027219772339, + "learning_rate": 0.0001732243353207593, + "loss": 0.7059, + "step": 15064 + }, + { + "epoch": 1.045381999861217, + "grad_norm": 2.261535167694092, + "learning_rate": 0.00017322050868601746, + "loss": 0.5619, + "step": 15065 + }, + { + "epoch": 1.0454513912983139, + "grad_norm": 1.9910287857055664, + "learning_rate": 0.00017321668182012699, + "loss": 0.8579, + "step": 15066 + }, + { + "epoch": 1.0455207827354105, + "grad_norm": 1.9650280475616455, + "learning_rate": 0.00017321285472309992, + "loss": 0.5819, + "step": 15067 + }, + { + "epoch": 1.045590174172507, + "grad_norm": 1.4099615812301636, + "learning_rate": 0.00017320902739494839, + "loss": 0.5483, + "step": 15068 + }, + { + "epoch": 1.0456595656096037, + "grad_norm": 1.5554461479187012, + "learning_rate": 0.0001732051998356844, + "loss": 0.7319, + "step": 15069 + }, + { + "epoch": 1.0457289570467005, + "grad_norm": 1.5745372772216797, + "learning_rate": 0.00017320137204532014, + "loss": 0.6101, + "step": 15070 + }, + { + "epoch": 1.0457983484837972, + "grad_norm": 1.5894882678985596, + "learning_rate": 0.00017319754402386766, + "loss": 0.481, + "step": 15071 + }, + { + "epoch": 1.0458677399208938, + "grad_norm": 1.9107335805892944, + "learning_rate": 0.00017319371577133902, + "loss": 0.6316, + "step": 15072 + }, + { + "epoch": 1.0459371313579904, + "grad_norm": 2.330280303955078, + "learning_rate": 0.00017318988728774626, + "loss": 0.6326, + "step": 15073 + }, + { + "epoch": 1.046006522795087, + "grad_norm": 1.5370373725891113, + "learning_rate": 0.00017318605857310156, + "loss": 0.6612, + "step": 15074 + }, + { + "epoch": 1.0460759142321838, + "grad_norm": 1.9974650144577026, + "learning_rate": 0.00017318222962741696, + "loss": 0.6605, + "step": 15075 + }, + { + "epoch": 1.0461453056692804, + "grad_norm": 1.8377538919448853, + "learning_rate": 0.0001731784004507045, + "loss": 0.5072, + "step": 15076 + }, + { + "epoch": 1.046214697106377, + "grad_norm": 2.1399996280670166, + "learning_rate": 0.0001731745710429764, + "loss": 0.7353, + "step": 15077 + }, + { + "epoch": 1.0462840885434737, + "grad_norm": 1.953762412071228, + "learning_rate": 0.00017317074140424464, + "loss": 0.8281, + "step": 15078 + }, + { + "epoch": 1.0463534799805705, + "grad_norm": 1.6647521257400513, + "learning_rate": 0.00017316691153452133, + "loss": 0.7686, + "step": 15079 + }, + { + "epoch": 1.046422871417667, + "grad_norm": 1.8305248022079468, + "learning_rate": 0.00017316308143381853, + "loss": 0.658, + "step": 15080 + }, + { + "epoch": 1.0464922628547637, + "grad_norm": 1.7796125411987305, + "learning_rate": 0.00017315925110214845, + "loss": 0.8327, + "step": 15081 + }, + { + "epoch": 1.0465616542918603, + "grad_norm": 1.7984539270401, + "learning_rate": 0.00017315542053952305, + "loss": 0.7026, + "step": 15082 + }, + { + "epoch": 1.0466310457289572, + "grad_norm": 2.0365984439849854, + "learning_rate": 0.0001731515897459545, + "loss": 0.8402, + "step": 15083 + }, + { + "epoch": 1.0467004371660538, + "grad_norm": 1.5802644491195679, + "learning_rate": 0.00017314775872145485, + "loss": 0.4904, + "step": 15084 + }, + { + "epoch": 1.0467698286031504, + "grad_norm": 2.228656053543091, + "learning_rate": 0.00017314392746603621, + "loss": 0.6276, + "step": 15085 + }, + { + "epoch": 1.046839220040247, + "grad_norm": 1.5834304094314575, + "learning_rate": 0.0001731400959797107, + "loss": 0.4639, + "step": 15086 + }, + { + "epoch": 1.0469086114773436, + "grad_norm": 2.519421100616455, + "learning_rate": 0.00017313626426249037, + "loss": 0.8466, + "step": 15087 + }, + { + "epoch": 1.0469780029144404, + "grad_norm": 2.016896963119507, + "learning_rate": 0.00017313243231438733, + "loss": 0.6854, + "step": 15088 + }, + { + "epoch": 1.047047394351537, + "grad_norm": 1.460484504699707, + "learning_rate": 0.0001731286001354137, + "loss": 0.5452, + "step": 15089 + }, + { + "epoch": 1.0471167857886337, + "grad_norm": 1.8409197330474854, + "learning_rate": 0.0001731247677255816, + "loss": 0.6461, + "step": 15090 + }, + { + "epoch": 1.0471861772257303, + "grad_norm": 2.007214307785034, + "learning_rate": 0.00017312093508490306, + "loss": 0.7622, + "step": 15091 + }, + { + "epoch": 1.047255568662827, + "grad_norm": 2.7457408905029297, + "learning_rate": 0.0001731171022133902, + "loss": 0.6099, + "step": 15092 + }, + { + "epoch": 1.0473249600999237, + "grad_norm": 1.99260675907135, + "learning_rate": 0.00017311326911105513, + "loss": 0.7827, + "step": 15093 + }, + { + "epoch": 1.0473943515370203, + "grad_norm": 2.3756332397460938, + "learning_rate": 0.00017310943577791, + "loss": 0.792, + "step": 15094 + }, + { + "epoch": 1.047463742974117, + "grad_norm": 2.3593528270721436, + "learning_rate": 0.00017310560221396682, + "loss": 0.6606, + "step": 15095 + }, + { + "epoch": 1.0475331344112138, + "grad_norm": 2.7073171138763428, + "learning_rate": 0.00017310176841923774, + "loss": 0.9125, + "step": 15096 + }, + { + "epoch": 1.0476025258483104, + "grad_norm": 1.8309950828552246, + "learning_rate": 0.00017309793439373484, + "loss": 0.6991, + "step": 15097 + }, + { + "epoch": 1.047671917285407, + "grad_norm": 1.8705487251281738, + "learning_rate": 0.00017309410013747028, + "loss": 0.6225, + "step": 15098 + }, + { + "epoch": 1.0477413087225036, + "grad_norm": 2.086573600769043, + "learning_rate": 0.00017309026565045613, + "loss": 0.6794, + "step": 15099 + }, + { + "epoch": 1.0478107001596002, + "grad_norm": 2.6501047611236572, + "learning_rate": 0.00017308643093270445, + "loss": 0.7913, + "step": 15100 + }, + { + "epoch": 1.047880091596697, + "grad_norm": 1.6804687976837158, + "learning_rate": 0.0001730825959842274, + "loss": 0.5468, + "step": 15101 + }, + { + "epoch": 1.0479494830337936, + "grad_norm": 1.7591264247894287, + "learning_rate": 0.0001730787608050371, + "loss": 0.6288, + "step": 15102 + }, + { + "epoch": 1.0480188744708903, + "grad_norm": 3.0975582599639893, + "learning_rate": 0.0001730749253951456, + "loss": 1.0511, + "step": 15103 + }, + { + "epoch": 1.0480882659079869, + "grad_norm": 1.698501706123352, + "learning_rate": 0.00017307108975456506, + "loss": 0.5428, + "step": 15104 + }, + { + "epoch": 1.0481576573450837, + "grad_norm": 1.9023672342300415, + "learning_rate": 0.00017306725388330758, + "loss": 0.6746, + "step": 15105 + }, + { + "epoch": 1.0482270487821803, + "grad_norm": 1.4383618831634521, + "learning_rate": 0.00017306341778138524, + "loss": 0.6064, + "step": 15106 + }, + { + "epoch": 1.048296440219277, + "grad_norm": 2.302633762359619, + "learning_rate": 0.00017305958144881017, + "loss": 0.8576, + "step": 15107 + }, + { + "epoch": 1.0483658316563735, + "grad_norm": 1.6323069334030151, + "learning_rate": 0.00017305574488559445, + "loss": 0.7775, + "step": 15108 + }, + { + "epoch": 1.0484352230934704, + "grad_norm": 1.3650455474853516, + "learning_rate": 0.00017305190809175025, + "loss": 0.504, + "step": 15109 + }, + { + "epoch": 1.048504614530567, + "grad_norm": 1.7324779033660889, + "learning_rate": 0.00017304807106728964, + "loss": 0.692, + "step": 15110 + }, + { + "epoch": 1.0485740059676636, + "grad_norm": 2.170964002609253, + "learning_rate": 0.00017304423381222475, + "loss": 0.5876, + "step": 15111 + }, + { + "epoch": 1.0486433974047602, + "grad_norm": 2.0848028659820557, + "learning_rate": 0.0001730403963265677, + "loss": 0.6442, + "step": 15112 + }, + { + "epoch": 1.0487127888418568, + "grad_norm": 2.1335554122924805, + "learning_rate": 0.00017303655861033055, + "loss": 0.8139, + "step": 15113 + }, + { + "epoch": 1.0487821802789536, + "grad_norm": 1.8495521545410156, + "learning_rate": 0.00017303272066352548, + "loss": 0.6541, + "step": 15114 + }, + { + "epoch": 1.0488515717160503, + "grad_norm": 2.01851487159729, + "learning_rate": 0.0001730288824861646, + "loss": 0.7463, + "step": 15115 + }, + { + "epoch": 1.0489209631531469, + "grad_norm": 1.947568655014038, + "learning_rate": 0.00017302504407826002, + "loss": 0.6576, + "step": 15116 + }, + { + "epoch": 1.0489903545902435, + "grad_norm": 1.613006830215454, + "learning_rate": 0.00017302120543982383, + "loss": 0.5427, + "step": 15117 + }, + { + "epoch": 1.0490597460273403, + "grad_norm": 2.0385048389434814, + "learning_rate": 0.00017301736657086817, + "loss": 0.5011, + "step": 15118 + }, + { + "epoch": 1.049129137464437, + "grad_norm": 1.930215835571289, + "learning_rate": 0.00017301352747140517, + "loss": 0.9859, + "step": 15119 + }, + { + "epoch": 1.0491985289015335, + "grad_norm": 2.418020725250244, + "learning_rate": 0.00017300968814144688, + "loss": 0.7767, + "step": 15120 + }, + { + "epoch": 1.0492679203386301, + "grad_norm": 1.9490866661071777, + "learning_rate": 0.00017300584858100554, + "loss": 0.5747, + "step": 15121 + }, + { + "epoch": 1.049337311775727, + "grad_norm": 2.3136415481567383, + "learning_rate": 0.00017300200879009317, + "loss": 0.8219, + "step": 15122 + }, + { + "epoch": 1.0494067032128236, + "grad_norm": 1.7194764614105225, + "learning_rate": 0.00017299816876872192, + "loss": 0.7419, + "step": 15123 + }, + { + "epoch": 1.0494760946499202, + "grad_norm": 2.3059914112091064, + "learning_rate": 0.00017299432851690396, + "loss": 0.867, + "step": 15124 + }, + { + "epoch": 1.0495454860870168, + "grad_norm": 2.38615345954895, + "learning_rate": 0.00017299048803465136, + "loss": 0.7307, + "step": 15125 + }, + { + "epoch": 1.0496148775241134, + "grad_norm": 2.113469362258911, + "learning_rate": 0.00017298664732197624, + "loss": 0.5187, + "step": 15126 + }, + { + "epoch": 1.0496842689612103, + "grad_norm": 2.0934486389160156, + "learning_rate": 0.00017298280637889076, + "loss": 0.6067, + "step": 15127 + }, + { + "epoch": 1.0497536603983069, + "grad_norm": 2.041581392288208, + "learning_rate": 0.000172978965205407, + "loss": 0.7328, + "step": 15128 + }, + { + "epoch": 1.0498230518354035, + "grad_norm": 1.9939699172973633, + "learning_rate": 0.00017297512380153715, + "loss": 0.5251, + "step": 15129 + }, + { + "epoch": 1.0498924432725, + "grad_norm": 1.4308850765228271, + "learning_rate": 0.00017297128216729328, + "loss": 0.7885, + "step": 15130 + }, + { + "epoch": 1.049961834709597, + "grad_norm": 1.69549560546875, + "learning_rate": 0.00017296744030268754, + "loss": 0.726, + "step": 15131 + }, + { + "epoch": 1.0500312261466935, + "grad_norm": 2.2528820037841797, + "learning_rate": 0.00017296359820773204, + "loss": 0.7512, + "step": 15132 + }, + { + "epoch": 1.0501006175837901, + "grad_norm": 2.6828958988189697, + "learning_rate": 0.00017295975588243898, + "loss": 0.66, + "step": 15133 + }, + { + "epoch": 1.0501700090208868, + "grad_norm": 1.7392115592956543, + "learning_rate": 0.0001729559133268204, + "loss": 0.6497, + "step": 15134 + }, + { + "epoch": 1.0502394004579836, + "grad_norm": 1.7561486959457397, + "learning_rate": 0.00017295207054088846, + "loss": 0.635, + "step": 15135 + }, + { + "epoch": 1.0503087918950802, + "grad_norm": 1.847993016242981, + "learning_rate": 0.0001729482275246553, + "loss": 0.6408, + "step": 15136 + }, + { + "epoch": 1.0503781833321768, + "grad_norm": 1.5539374351501465, + "learning_rate": 0.00017294438427813307, + "loss": 0.6486, + "step": 15137 + }, + { + "epoch": 1.0504475747692734, + "grad_norm": 1.7590408325195312, + "learning_rate": 0.00017294054080133387, + "loss": 0.6937, + "step": 15138 + }, + { + "epoch": 1.05051696620637, + "grad_norm": 1.532958745956421, + "learning_rate": 0.00017293669709426984, + "loss": 0.558, + "step": 15139 + }, + { + "epoch": 1.0505863576434669, + "grad_norm": 1.4859340190887451, + "learning_rate": 0.00017293285315695312, + "loss": 0.6308, + "step": 15140 + }, + { + "epoch": 1.0506557490805635, + "grad_norm": 1.8387469053268433, + "learning_rate": 0.00017292900898939584, + "loss": 0.789, + "step": 15141 + }, + { + "epoch": 1.05072514051766, + "grad_norm": 2.1987364292144775, + "learning_rate": 0.00017292516459161013, + "loss": 0.6361, + "step": 15142 + }, + { + "epoch": 1.0507945319547567, + "grad_norm": 1.7388945817947388, + "learning_rate": 0.00017292131996360816, + "loss": 0.7221, + "step": 15143 + }, + { + "epoch": 1.0508639233918535, + "grad_norm": 1.4839845895767212, + "learning_rate": 0.00017291747510540204, + "loss": 0.5864, + "step": 15144 + }, + { + "epoch": 1.0509333148289501, + "grad_norm": 1.9960508346557617, + "learning_rate": 0.00017291363001700391, + "loss": 0.5611, + "step": 15145 + }, + { + "epoch": 1.0510027062660467, + "grad_norm": 2.663416624069214, + "learning_rate": 0.0001729097846984259, + "loss": 0.6538, + "step": 15146 + }, + { + "epoch": 1.0510720977031434, + "grad_norm": 1.5693658590316772, + "learning_rate": 0.00017290593914968017, + "loss": 0.6284, + "step": 15147 + }, + { + "epoch": 1.0511414891402402, + "grad_norm": 1.8103692531585693, + "learning_rate": 0.00017290209337077883, + "loss": 0.4929, + "step": 15148 + }, + { + "epoch": 1.0512108805773368, + "grad_norm": 2.2105231285095215, + "learning_rate": 0.00017289824736173407, + "loss": 0.685, + "step": 15149 + }, + { + "epoch": 1.0512802720144334, + "grad_norm": 1.9985811710357666, + "learning_rate": 0.00017289440112255798, + "loss": 0.7364, + "step": 15150 + }, + { + "epoch": 1.05134966345153, + "grad_norm": 1.7947829961776733, + "learning_rate": 0.00017289055465326272, + "loss": 0.7321, + "step": 15151 + }, + { + "epoch": 1.0514190548886266, + "grad_norm": 2.294316530227661, + "learning_rate": 0.00017288670795386044, + "loss": 0.7516, + "step": 15152 + }, + { + "epoch": 1.0514884463257235, + "grad_norm": 1.7992926836013794, + "learning_rate": 0.0001728828610243633, + "loss": 0.7072, + "step": 15153 + }, + { + "epoch": 1.05155783776282, + "grad_norm": 2.7046921253204346, + "learning_rate": 0.00017287901386478337, + "loss": 0.7666, + "step": 15154 + }, + { + "epoch": 1.0516272291999167, + "grad_norm": 1.9679679870605469, + "learning_rate": 0.0001728751664751329, + "loss": 0.7896, + "step": 15155 + }, + { + "epoch": 1.0516966206370133, + "grad_norm": 2.126814365386963, + "learning_rate": 0.00017287131885542398, + "loss": 0.5607, + "step": 15156 + }, + { + "epoch": 1.0517660120741101, + "grad_norm": 2.447225570678711, + "learning_rate": 0.00017286747100566876, + "loss": 0.6178, + "step": 15157 + }, + { + "epoch": 1.0518354035112067, + "grad_norm": 1.7162731885910034, + "learning_rate": 0.00017286362292587938, + "loss": 0.6312, + "step": 15158 + }, + { + "epoch": 1.0519047949483034, + "grad_norm": 2.4422521591186523, + "learning_rate": 0.00017285977461606796, + "loss": 0.6995, + "step": 15159 + }, + { + "epoch": 1.0519741863854, + "grad_norm": 1.989519476890564, + "learning_rate": 0.00017285592607624675, + "loss": 0.6325, + "step": 15160 + }, + { + "epoch": 1.0520435778224968, + "grad_norm": 2.114668130874634, + "learning_rate": 0.00017285207730642778, + "loss": 0.9832, + "step": 15161 + }, + { + "epoch": 1.0521129692595934, + "grad_norm": 1.5220069885253906, + "learning_rate": 0.00017284822830662328, + "loss": 0.343, + "step": 15162 + }, + { + "epoch": 1.05218236069669, + "grad_norm": 1.6217625141143799, + "learning_rate": 0.0001728443790768454, + "loss": 0.6009, + "step": 15163 + }, + { + "epoch": 1.0522517521337866, + "grad_norm": 1.8083195686340332, + "learning_rate": 0.00017284052961710623, + "loss": 0.7559, + "step": 15164 + }, + { + "epoch": 1.0523211435708832, + "grad_norm": 1.618082046508789, + "learning_rate": 0.00017283667992741797, + "loss": 0.5303, + "step": 15165 + }, + { + "epoch": 1.05239053500798, + "grad_norm": 4.2593994140625, + "learning_rate": 0.00017283283000779274, + "loss": 0.5602, + "step": 15166 + }, + { + "epoch": 1.0524599264450767, + "grad_norm": 1.9369086027145386, + "learning_rate": 0.00017282897985824275, + "loss": 0.8086, + "step": 15167 + }, + { + "epoch": 1.0525293178821733, + "grad_norm": 1.5295110940933228, + "learning_rate": 0.00017282512947878011, + "loss": 0.5999, + "step": 15168 + }, + { + "epoch": 1.05259870931927, + "grad_norm": 2.0121328830718994, + "learning_rate": 0.00017282127886941696, + "loss": 0.4337, + "step": 15169 + }, + { + "epoch": 1.0526681007563667, + "grad_norm": 8.63448715209961, + "learning_rate": 0.0001728174280301655, + "loss": 1.0488, + "step": 15170 + }, + { + "epoch": 1.0527374921934634, + "grad_norm": 3.2038190364837646, + "learning_rate": 0.0001728135769610379, + "loss": 0.7876, + "step": 15171 + }, + { + "epoch": 1.05280688363056, + "grad_norm": 2.3854193687438965, + "learning_rate": 0.00017280972566204622, + "loss": 0.8179, + "step": 15172 + }, + { + "epoch": 1.0528762750676566, + "grad_norm": 2.2313475608825684, + "learning_rate": 0.00017280587413320275, + "loss": 0.7645, + "step": 15173 + }, + { + "epoch": 1.0529456665047534, + "grad_norm": 2.069169044494629, + "learning_rate": 0.00017280202237451956, + "loss": 0.7119, + "step": 15174 + }, + { + "epoch": 1.05301505794185, + "grad_norm": 2.461259365081787, + "learning_rate": 0.00017279817038600882, + "loss": 0.7139, + "step": 15175 + }, + { + "epoch": 1.0530844493789466, + "grad_norm": 2.543156385421753, + "learning_rate": 0.0001727943181676827, + "loss": 0.5739, + "step": 15176 + }, + { + "epoch": 1.0531538408160432, + "grad_norm": 2.422107696533203, + "learning_rate": 0.00017279046571955337, + "loss": 0.8065, + "step": 15177 + }, + { + "epoch": 1.0532232322531399, + "grad_norm": 2.870603084564209, + "learning_rate": 0.000172786613041633, + "loss": 0.7181, + "step": 15178 + }, + { + "epoch": 1.0532926236902367, + "grad_norm": 2.1873817443847656, + "learning_rate": 0.0001727827601339337, + "loss": 0.8098, + "step": 15179 + }, + { + "epoch": 1.0533620151273333, + "grad_norm": 2.8102638721466064, + "learning_rate": 0.0001727789069964677, + "loss": 0.8028, + "step": 15180 + }, + { + "epoch": 1.05343140656443, + "grad_norm": 2.3300681114196777, + "learning_rate": 0.00017277505362924713, + "loss": 0.5201, + "step": 15181 + }, + { + "epoch": 1.0535007980015265, + "grad_norm": 1.9926000833511353, + "learning_rate": 0.00017277120003228417, + "loss": 0.6262, + "step": 15182 + }, + { + "epoch": 1.0535701894386234, + "grad_norm": 1.5599440336227417, + "learning_rate": 0.00017276734620559094, + "loss": 0.6867, + "step": 15183 + }, + { + "epoch": 1.05363958087572, + "grad_norm": 1.925918698310852, + "learning_rate": 0.00017276349214917968, + "loss": 0.6586, + "step": 15184 + }, + { + "epoch": 1.0537089723128166, + "grad_norm": 2.1426632404327393, + "learning_rate": 0.00017275963786306248, + "loss": 0.7735, + "step": 15185 + }, + { + "epoch": 1.0537783637499132, + "grad_norm": 1.775154709815979, + "learning_rate": 0.0001727557833472516, + "loss": 0.8351, + "step": 15186 + }, + { + "epoch": 1.05384775518701, + "grad_norm": 1.8521448373794556, + "learning_rate": 0.0001727519286017591, + "loss": 0.6684, + "step": 15187 + }, + { + "epoch": 1.0539171466241066, + "grad_norm": 1.7801629304885864, + "learning_rate": 0.00017274807362659722, + "loss": 0.5863, + "step": 15188 + }, + { + "epoch": 1.0539865380612032, + "grad_norm": 2.2859246730804443, + "learning_rate": 0.0001727442184217781, + "loss": 0.6476, + "step": 15189 + }, + { + "epoch": 1.0540559294982998, + "grad_norm": 2.408579111099243, + "learning_rate": 0.00017274036298731396, + "loss": 0.8177, + "step": 15190 + }, + { + "epoch": 1.0541253209353965, + "grad_norm": 2.30851411819458, + "learning_rate": 0.00017273650732321688, + "loss": 0.8457, + "step": 15191 + }, + { + "epoch": 1.0541947123724933, + "grad_norm": 2.682633876800537, + "learning_rate": 0.00017273265142949911, + "loss": 0.5958, + "step": 15192 + }, + { + "epoch": 1.05426410380959, + "grad_norm": 1.7490246295928955, + "learning_rate": 0.00017272879530617284, + "loss": 0.7642, + "step": 15193 + }, + { + "epoch": 1.0543334952466865, + "grad_norm": 1.771393895149231, + "learning_rate": 0.00017272493895325016, + "loss": 0.6905, + "step": 15194 + }, + { + "epoch": 1.0544028866837831, + "grad_norm": 1.3891202211380005, + "learning_rate": 0.00017272108237074328, + "loss": 0.5478, + "step": 15195 + }, + { + "epoch": 1.05447227812088, + "grad_norm": 1.553564190864563, + "learning_rate": 0.00017271722555866438, + "loss": 0.4718, + "step": 15196 + }, + { + "epoch": 1.0545416695579766, + "grad_norm": 2.5465493202209473, + "learning_rate": 0.00017271336851702564, + "loss": 0.84, + "step": 15197 + }, + { + "epoch": 1.0546110609950732, + "grad_norm": 1.6285288333892822, + "learning_rate": 0.00017270951124583923, + "loss": 0.6244, + "step": 15198 + }, + { + "epoch": 1.0546804524321698, + "grad_norm": 2.1370904445648193, + "learning_rate": 0.00017270565374511736, + "loss": 0.8493, + "step": 15199 + }, + { + "epoch": 1.0547498438692666, + "grad_norm": 1.9856752157211304, + "learning_rate": 0.00017270179601487214, + "loss": 0.8767, + "step": 15200 + }, + { + "epoch": 1.0548192353063632, + "grad_norm": 1.9018442630767822, + "learning_rate": 0.0001726979380551158, + "loss": 0.6179, + "step": 15201 + }, + { + "epoch": 1.0548886267434598, + "grad_norm": 2.753821611404419, + "learning_rate": 0.0001726940798658605, + "loss": 0.9623, + "step": 15202 + }, + { + "epoch": 1.0549580181805565, + "grad_norm": 2.793649911880493, + "learning_rate": 0.00017269022144711843, + "loss": 0.6004, + "step": 15203 + }, + { + "epoch": 1.0550274096176533, + "grad_norm": 1.9984074831008911, + "learning_rate": 0.00017268636279890175, + "loss": 0.492, + "step": 15204 + }, + { + "epoch": 1.05509680105475, + "grad_norm": 1.6147414445877075, + "learning_rate": 0.00017268250392122266, + "loss": 0.8323, + "step": 15205 + }, + { + "epoch": 1.0551661924918465, + "grad_norm": 1.4245414733886719, + "learning_rate": 0.00017267864481409334, + "loss": 0.5048, + "step": 15206 + }, + { + "epoch": 1.0552355839289431, + "grad_norm": 1.7267274856567383, + "learning_rate": 0.00017267478547752598, + "loss": 0.6378, + "step": 15207 + }, + { + "epoch": 1.0553049753660397, + "grad_norm": 2.219419240951538, + "learning_rate": 0.00017267092591153276, + "loss": 0.7327, + "step": 15208 + }, + { + "epoch": 1.0553743668031366, + "grad_norm": 1.868290662765503, + "learning_rate": 0.00017266706611612583, + "loss": 0.6386, + "step": 15209 + }, + { + "epoch": 1.0554437582402332, + "grad_norm": 1.977640986442566, + "learning_rate": 0.00017266320609131742, + "loss": 0.6827, + "step": 15210 + }, + { + "epoch": 1.0555131496773298, + "grad_norm": 1.7001501321792603, + "learning_rate": 0.00017265934583711969, + "loss": 0.5859, + "step": 15211 + }, + { + "epoch": 1.0555825411144264, + "grad_norm": 1.7757863998413086, + "learning_rate": 0.00017265548535354484, + "loss": 0.4703, + "step": 15212 + }, + { + "epoch": 1.0556519325515232, + "grad_norm": 2.1408655643463135, + "learning_rate": 0.00017265162464060507, + "loss": 0.7487, + "step": 15213 + }, + { + "epoch": 1.0557213239886198, + "grad_norm": 2.0764458179473877, + "learning_rate": 0.00017264776369831255, + "loss": 0.7769, + "step": 15214 + }, + { + "epoch": 1.0557907154257165, + "grad_norm": 2.139843463897705, + "learning_rate": 0.00017264390252667944, + "loss": 0.7053, + "step": 15215 + }, + { + "epoch": 1.055860106862813, + "grad_norm": 2.2295658588409424, + "learning_rate": 0.00017264004112571796, + "loss": 0.7136, + "step": 15216 + }, + { + "epoch": 1.05592949829991, + "grad_norm": 1.677254319190979, + "learning_rate": 0.0001726361794954403, + "loss": 0.7022, + "step": 15217 + }, + { + "epoch": 1.0559988897370065, + "grad_norm": 1.9520779848098755, + "learning_rate": 0.00017263231763585868, + "loss": 0.5974, + "step": 15218 + }, + { + "epoch": 1.0560682811741031, + "grad_norm": 1.1775598526000977, + "learning_rate": 0.00017262845554698523, + "loss": 0.4508, + "step": 15219 + }, + { + "epoch": 1.0561376726111997, + "grad_norm": 2.0128092765808105, + "learning_rate": 0.00017262459322883218, + "loss": 0.5932, + "step": 15220 + }, + { + "epoch": 1.0562070640482963, + "grad_norm": 1.9436283111572266, + "learning_rate": 0.00017262073068141174, + "loss": 0.6858, + "step": 15221 + }, + { + "epoch": 1.0562764554853932, + "grad_norm": 2.1655874252319336, + "learning_rate": 0.00017261686790473604, + "loss": 0.8213, + "step": 15222 + }, + { + "epoch": 1.0563458469224898, + "grad_norm": 2.179021120071411, + "learning_rate": 0.0001726130048988173, + "loss": 0.716, + "step": 15223 + }, + { + "epoch": 1.0564152383595864, + "grad_norm": 1.9666277170181274, + "learning_rate": 0.0001726091416636678, + "loss": 0.5156, + "step": 15224 + }, + { + "epoch": 1.056484629796683, + "grad_norm": 1.9076693058013916, + "learning_rate": 0.00017260527819929958, + "loss": 0.7147, + "step": 15225 + }, + { + "epoch": 1.0565540212337798, + "grad_norm": 1.6442408561706543, + "learning_rate": 0.00017260141450572497, + "loss": 0.4943, + "step": 15226 + }, + { + "epoch": 1.0566234126708764, + "grad_norm": 2.2452383041381836, + "learning_rate": 0.0001725975505829561, + "loss": 0.7887, + "step": 15227 + }, + { + "epoch": 1.056692804107973, + "grad_norm": 1.5288490056991577, + "learning_rate": 0.0001725936864310052, + "loss": 0.4768, + "step": 15228 + }, + { + "epoch": 1.0567621955450697, + "grad_norm": 2.901423692703247, + "learning_rate": 0.00017258982204988445, + "loss": 0.9423, + "step": 15229 + }, + { + "epoch": 1.0568315869821665, + "grad_norm": 3.9335734844207764, + "learning_rate": 0.0001725859574396061, + "loss": 0.7496, + "step": 15230 + }, + { + "epoch": 1.0569009784192631, + "grad_norm": 2.000256061553955, + "learning_rate": 0.00017258209260018222, + "loss": 0.6426, + "step": 15231 + }, + { + "epoch": 1.0569703698563597, + "grad_norm": 1.7438910007476807, + "learning_rate": 0.00017257822753162513, + "loss": 0.6074, + "step": 15232 + }, + { + "epoch": 1.0570397612934563, + "grad_norm": 1.5512081384658813, + "learning_rate": 0.000172574362233947, + "loss": 0.602, + "step": 15233 + }, + { + "epoch": 1.057109152730553, + "grad_norm": 2.3159468173980713, + "learning_rate": 0.00017257049670716, + "loss": 0.8231, + "step": 15234 + }, + { + "epoch": 1.0571785441676498, + "grad_norm": 2.03434681892395, + "learning_rate": 0.00017256663095127642, + "loss": 0.7016, + "step": 15235 + }, + { + "epoch": 1.0572479356047464, + "grad_norm": 1.8264496326446533, + "learning_rate": 0.00017256276496630836, + "loss": 0.5221, + "step": 15236 + }, + { + "epoch": 1.057317327041843, + "grad_norm": 1.6023792028427124, + "learning_rate": 0.00017255889875226807, + "loss": 0.5946, + "step": 15237 + }, + { + "epoch": 1.0573867184789396, + "grad_norm": 2.1225781440734863, + "learning_rate": 0.00017255503230916778, + "loss": 0.7452, + "step": 15238 + }, + { + "epoch": 1.0574561099160364, + "grad_norm": 2.316128730773926, + "learning_rate": 0.00017255116563701965, + "loss": 0.8605, + "step": 15239 + }, + { + "epoch": 1.057525501353133, + "grad_norm": 1.894959807395935, + "learning_rate": 0.00017254729873583595, + "loss": 0.7109, + "step": 15240 + }, + { + "epoch": 1.0575948927902297, + "grad_norm": 1.5954331159591675, + "learning_rate": 0.00017254343160562878, + "loss": 0.472, + "step": 15241 + }, + { + "epoch": 1.0576642842273263, + "grad_norm": 2.56298565864563, + "learning_rate": 0.00017253956424641046, + "loss": 0.7687, + "step": 15242 + }, + { + "epoch": 1.0577336756644231, + "grad_norm": 1.5142951011657715, + "learning_rate": 0.00017253569665819313, + "loss": 0.6666, + "step": 15243 + }, + { + "epoch": 1.0578030671015197, + "grad_norm": 3.319575071334839, + "learning_rate": 0.00017253182884098905, + "loss": 0.9111, + "step": 15244 + }, + { + "epoch": 1.0578724585386163, + "grad_norm": 2.391308069229126, + "learning_rate": 0.00017252796079481034, + "loss": 0.7787, + "step": 15245 + }, + { + "epoch": 1.057941849975713, + "grad_norm": 2.491403341293335, + "learning_rate": 0.00017252409251966932, + "loss": 0.8076, + "step": 15246 + }, + { + "epoch": 1.0580112414128096, + "grad_norm": 1.6970993280410767, + "learning_rate": 0.00017252022401557815, + "loss": 0.6329, + "step": 15247 + }, + { + "epoch": 1.0580806328499064, + "grad_norm": 1.6620701551437378, + "learning_rate": 0.00017251635528254907, + "loss": 0.5399, + "step": 15248 + }, + { + "epoch": 1.058150024287003, + "grad_norm": 1.9085272550582886, + "learning_rate": 0.00017251248632059422, + "loss": 0.8581, + "step": 15249 + }, + { + "epoch": 1.0582194157240996, + "grad_norm": 1.8697278499603271, + "learning_rate": 0.00017250861712972588, + "loss": 0.821, + "step": 15250 + }, + { + "epoch": 1.0582888071611962, + "grad_norm": 1.8385539054870605, + "learning_rate": 0.00017250474770995626, + "loss": 0.4559, + "step": 15251 + }, + { + "epoch": 1.058358198598293, + "grad_norm": 3.335031509399414, + "learning_rate": 0.00017250087806129755, + "loss": 1.0677, + "step": 15252 + }, + { + "epoch": 1.0584275900353897, + "grad_norm": 1.6803470849990845, + "learning_rate": 0.000172497008183762, + "loss": 0.7286, + "step": 15253 + }, + { + "epoch": 1.0584969814724863, + "grad_norm": 1.6617094278335571, + "learning_rate": 0.0001724931380773618, + "loss": 0.5095, + "step": 15254 + }, + { + "epoch": 1.0585663729095829, + "grad_norm": 2.0160882472991943, + "learning_rate": 0.00017248926774210917, + "loss": 0.6455, + "step": 15255 + }, + { + "epoch": 1.0586357643466797, + "grad_norm": 1.9622687101364136, + "learning_rate": 0.00017248539717801633, + "loss": 0.7945, + "step": 15256 + }, + { + "epoch": 1.0587051557837763, + "grad_norm": 1.6058611869812012, + "learning_rate": 0.0001724815263850955, + "loss": 0.5857, + "step": 15257 + }, + { + "epoch": 1.058774547220873, + "grad_norm": 2.0428574085235596, + "learning_rate": 0.0001724776553633589, + "loss": 0.7679, + "step": 15258 + }, + { + "epoch": 1.0588439386579696, + "grad_norm": 2.3162665367126465, + "learning_rate": 0.00017247378411281873, + "loss": 0.7093, + "step": 15259 + }, + { + "epoch": 1.0589133300950662, + "grad_norm": 2.196674346923828, + "learning_rate": 0.00017246991263348727, + "loss": 0.6169, + "step": 15260 + }, + { + "epoch": 1.058982721532163, + "grad_norm": 2.905724048614502, + "learning_rate": 0.0001724660409253767, + "loss": 0.7879, + "step": 15261 + }, + { + "epoch": 1.0590521129692596, + "grad_norm": 1.8840175867080688, + "learning_rate": 0.00017246216898849924, + "loss": 0.6846, + "step": 15262 + }, + { + "epoch": 1.0591215044063562, + "grad_norm": 2.2903153896331787, + "learning_rate": 0.0001724582968228671, + "loss": 1.0278, + "step": 15263 + }, + { + "epoch": 1.0591908958434528, + "grad_norm": 1.9869866371154785, + "learning_rate": 0.00017245442442849254, + "loss": 0.6694, + "step": 15264 + }, + { + "epoch": 1.0592602872805497, + "grad_norm": 1.7803679704666138, + "learning_rate": 0.00017245055180538775, + "loss": 0.7784, + "step": 15265 + }, + { + "epoch": 1.0593296787176463, + "grad_norm": 4.120851516723633, + "learning_rate": 0.000172446678953565, + "loss": 0.6069, + "step": 15266 + }, + { + "epoch": 1.0593990701547429, + "grad_norm": 1.9805155992507935, + "learning_rate": 0.00017244280587303646, + "loss": 0.6229, + "step": 15267 + }, + { + "epoch": 1.0594684615918395, + "grad_norm": 1.9579578638076782, + "learning_rate": 0.00017243893256381444, + "loss": 0.6936, + "step": 15268 + }, + { + "epoch": 1.0595378530289363, + "grad_norm": 1.8074994087219238, + "learning_rate": 0.00017243505902591108, + "loss": 0.6836, + "step": 15269 + }, + { + "epoch": 1.059607244466033, + "grad_norm": 1.9546509981155396, + "learning_rate": 0.00017243118525933865, + "loss": 0.7498, + "step": 15270 + }, + { + "epoch": 1.0596766359031295, + "grad_norm": 1.942847728729248, + "learning_rate": 0.00017242731126410936, + "loss": 0.8176, + "step": 15271 + }, + { + "epoch": 1.0597460273402262, + "grad_norm": 2.8243424892425537, + "learning_rate": 0.00017242343704023543, + "loss": 0.7057, + "step": 15272 + }, + { + "epoch": 1.059815418777323, + "grad_norm": 1.829677700996399, + "learning_rate": 0.00017241956258772916, + "loss": 0.645, + "step": 15273 + }, + { + "epoch": 1.0598848102144196, + "grad_norm": 2.1946794986724854, + "learning_rate": 0.0001724156879066027, + "loss": 0.6093, + "step": 15274 + }, + { + "epoch": 1.0599542016515162, + "grad_norm": 2.3728983402252197, + "learning_rate": 0.00017241181299686834, + "loss": 0.9217, + "step": 15275 + }, + { + "epoch": 1.0600235930886128, + "grad_norm": 1.7521544694900513, + "learning_rate": 0.00017240793785853827, + "loss": 0.5056, + "step": 15276 + }, + { + "epoch": 1.0600929845257094, + "grad_norm": 2.138310432434082, + "learning_rate": 0.00017240406249162475, + "loss": 1.0078, + "step": 15277 + }, + { + "epoch": 1.0601623759628063, + "grad_norm": 1.8591970205307007, + "learning_rate": 0.00017240018689614002, + "loss": 0.4618, + "step": 15278 + }, + { + "epoch": 1.0602317673999029, + "grad_norm": 1.8064513206481934, + "learning_rate": 0.00017239631107209626, + "loss": 0.8266, + "step": 15279 + }, + { + "epoch": 1.0603011588369995, + "grad_norm": 2.0079784393310547, + "learning_rate": 0.00017239243501950577, + "loss": 0.8567, + "step": 15280 + }, + { + "epoch": 1.060370550274096, + "grad_norm": 2.4401838779449463, + "learning_rate": 0.00017238855873838075, + "loss": 0.6932, + "step": 15281 + }, + { + "epoch": 1.060439941711193, + "grad_norm": 2.9031107425689697, + "learning_rate": 0.00017238468222873348, + "loss": 0.698, + "step": 15282 + }, + { + "epoch": 1.0605093331482895, + "grad_norm": 1.4132176637649536, + "learning_rate": 0.00017238080549057615, + "loss": 0.6117, + "step": 15283 + }, + { + "epoch": 1.0605787245853862, + "grad_norm": 1.837893009185791, + "learning_rate": 0.000172376928523921, + "loss": 0.6186, + "step": 15284 + }, + { + "epoch": 1.0606481160224828, + "grad_norm": 1.5561696290969849, + "learning_rate": 0.00017237305132878028, + "loss": 0.4456, + "step": 15285 + }, + { + "epoch": 1.0607175074595796, + "grad_norm": 1.6436729431152344, + "learning_rate": 0.00017236917390516622, + "loss": 0.5772, + "step": 15286 + }, + { + "epoch": 1.0607868988966762, + "grad_norm": 2.346426010131836, + "learning_rate": 0.00017236529625309113, + "loss": 0.7833, + "step": 15287 + }, + { + "epoch": 1.0608562903337728, + "grad_norm": 1.889575481414795, + "learning_rate": 0.00017236141837256717, + "loss": 0.773, + "step": 15288 + }, + { + "epoch": 1.0609256817708694, + "grad_norm": 2.4918463230133057, + "learning_rate": 0.00017235754026360657, + "loss": 0.723, + "step": 15289 + }, + { + "epoch": 1.060995073207966, + "grad_norm": 1.8374722003936768, + "learning_rate": 0.00017235366192622166, + "loss": 0.6027, + "step": 15290 + }, + { + "epoch": 1.0610644646450629, + "grad_norm": 2.371169090270996, + "learning_rate": 0.00017234978336042462, + "loss": 0.8025, + "step": 15291 + }, + { + "epoch": 1.0611338560821595, + "grad_norm": 2.0314300060272217, + "learning_rate": 0.0001723459045662277, + "loss": 0.6032, + "step": 15292 + }, + { + "epoch": 1.061203247519256, + "grad_norm": 2.126585006713867, + "learning_rate": 0.00017234202554364315, + "loss": 0.7801, + "step": 15293 + }, + { + "epoch": 1.0612726389563527, + "grad_norm": 2.193202018737793, + "learning_rate": 0.00017233814629268322, + "loss": 0.7305, + "step": 15294 + }, + { + "epoch": 1.0613420303934495, + "grad_norm": 1.4631811380386353, + "learning_rate": 0.0001723342668133602, + "loss": 0.7559, + "step": 15295 + }, + { + "epoch": 1.0614114218305462, + "grad_norm": 1.9214292764663696, + "learning_rate": 0.00017233038710568624, + "loss": 0.6247, + "step": 15296 + }, + { + "epoch": 1.0614808132676428, + "grad_norm": 2.204806089401245, + "learning_rate": 0.00017232650716967368, + "loss": 0.6345, + "step": 15297 + }, + { + "epoch": 1.0615502047047394, + "grad_norm": 1.8088879585266113, + "learning_rate": 0.00017232262700533468, + "loss": 0.5561, + "step": 15298 + }, + { + "epoch": 1.0616195961418362, + "grad_norm": 2.688891887664795, + "learning_rate": 0.00017231874661268156, + "loss": 1.0196, + "step": 15299 + }, + { + "epoch": 1.0616889875789328, + "grad_norm": 1.9103028774261475, + "learning_rate": 0.00017231486599172656, + "loss": 0.6019, + "step": 15300 + }, + { + "epoch": 1.0617583790160294, + "grad_norm": 2.2272119522094727, + "learning_rate": 0.0001723109851424819, + "loss": 0.6, + "step": 15301 + }, + { + "epoch": 1.061827770453126, + "grad_norm": 2.7489261627197266, + "learning_rate": 0.00017230710406495984, + "loss": 0.8414, + "step": 15302 + }, + { + "epoch": 1.0618971618902227, + "grad_norm": 2.2885236740112305, + "learning_rate": 0.0001723032227591727, + "loss": 0.696, + "step": 15303 + }, + { + "epoch": 1.0619665533273195, + "grad_norm": 1.8364653587341309, + "learning_rate": 0.0001722993412251326, + "loss": 0.4355, + "step": 15304 + }, + { + "epoch": 1.062035944764416, + "grad_norm": 2.508859395980835, + "learning_rate": 0.0001722954594628519, + "loss": 0.5641, + "step": 15305 + }, + { + "epoch": 1.0621053362015127, + "grad_norm": 1.9273771047592163, + "learning_rate": 0.00017229157747234282, + "loss": 0.5248, + "step": 15306 + }, + { + "epoch": 1.0621747276386093, + "grad_norm": 2.213559150695801, + "learning_rate": 0.00017228769525361762, + "loss": 0.5563, + "step": 15307 + }, + { + "epoch": 1.0622441190757062, + "grad_norm": 2.3444619178771973, + "learning_rate": 0.00017228381280668854, + "loss": 0.648, + "step": 15308 + }, + { + "epoch": 1.0623135105128028, + "grad_norm": 1.83638334274292, + "learning_rate": 0.00017227993013156789, + "loss": 0.6204, + "step": 15309 + }, + { + "epoch": 1.0623829019498994, + "grad_norm": 2.6977601051330566, + "learning_rate": 0.00017227604722826783, + "loss": 0.8025, + "step": 15310 + }, + { + "epoch": 1.062452293386996, + "grad_norm": 2.1418468952178955, + "learning_rate": 0.00017227216409680072, + "loss": 0.5928, + "step": 15311 + }, + { + "epoch": 1.0625216848240928, + "grad_norm": 1.8256584405899048, + "learning_rate": 0.00017226828073717875, + "loss": 0.7154, + "step": 15312 + }, + { + "epoch": 1.0625910762611894, + "grad_norm": 1.8296260833740234, + "learning_rate": 0.0001722643971494142, + "loss": 0.7267, + "step": 15313 + }, + { + "epoch": 1.062660467698286, + "grad_norm": 2.342895269393921, + "learning_rate": 0.00017226051333351934, + "loss": 0.8119, + "step": 15314 + }, + { + "epoch": 1.0627298591353826, + "grad_norm": 2.3660366535186768, + "learning_rate": 0.00017225662928950643, + "loss": 0.7107, + "step": 15315 + }, + { + "epoch": 1.0627992505724793, + "grad_norm": 1.9270570278167725, + "learning_rate": 0.00017225274501738772, + "loss": 0.7599, + "step": 15316 + }, + { + "epoch": 1.062868642009576, + "grad_norm": 1.9577000141143799, + "learning_rate": 0.00017224886051717548, + "loss": 0.8662, + "step": 15317 + }, + { + "epoch": 1.0629380334466727, + "grad_norm": 1.8471158742904663, + "learning_rate": 0.00017224497578888193, + "loss": 0.6257, + "step": 15318 + }, + { + "epoch": 1.0630074248837693, + "grad_norm": 2.086599111557007, + "learning_rate": 0.00017224109083251944, + "loss": 0.6312, + "step": 15319 + }, + { + "epoch": 1.063076816320866, + "grad_norm": 2.2923192977905273, + "learning_rate": 0.00017223720564810016, + "loss": 0.4418, + "step": 15320 + }, + { + "epoch": 1.0631462077579628, + "grad_norm": 1.7949494123458862, + "learning_rate": 0.00017223332023563642, + "loss": 0.7195, + "step": 15321 + }, + { + "epoch": 1.0632155991950594, + "grad_norm": 1.6705814599990845, + "learning_rate": 0.00017222943459514048, + "loss": 0.5832, + "step": 15322 + }, + { + "epoch": 1.063284990632156, + "grad_norm": 3.036048412322998, + "learning_rate": 0.00017222554872662458, + "loss": 0.7904, + "step": 15323 + }, + { + "epoch": 1.0633543820692526, + "grad_norm": 6.338029861450195, + "learning_rate": 0.000172221662630101, + "loss": 0.9703, + "step": 15324 + }, + { + "epoch": 1.0634237735063494, + "grad_norm": 2.3834455013275146, + "learning_rate": 0.00017221777630558203, + "loss": 0.736, + "step": 15325 + }, + { + "epoch": 1.063493164943446, + "grad_norm": 2.0250511169433594, + "learning_rate": 0.0001722138897530799, + "loss": 0.8509, + "step": 15326 + }, + { + "epoch": 1.0635625563805426, + "grad_norm": 2.0933749675750732, + "learning_rate": 0.0001722100029726069, + "loss": 0.8462, + "step": 15327 + }, + { + "epoch": 1.0636319478176393, + "grad_norm": 2.5884816646575928, + "learning_rate": 0.0001722061159641753, + "loss": 0.7528, + "step": 15328 + }, + { + "epoch": 1.0637013392547359, + "grad_norm": 1.8603904247283936, + "learning_rate": 0.00017220222872779738, + "loss": 0.6123, + "step": 15329 + }, + { + "epoch": 1.0637707306918327, + "grad_norm": 1.9572646617889404, + "learning_rate": 0.0001721983412634854, + "loss": 0.6431, + "step": 15330 + }, + { + "epoch": 1.0638401221289293, + "grad_norm": 1.6238394975662231, + "learning_rate": 0.0001721944535712516, + "loss": 0.5276, + "step": 15331 + }, + { + "epoch": 1.063909513566026, + "grad_norm": 1.9641950130462646, + "learning_rate": 0.00017219056565110833, + "loss": 0.6162, + "step": 15332 + }, + { + "epoch": 1.0639789050031225, + "grad_norm": 2.555307149887085, + "learning_rate": 0.0001721866775030678, + "loss": 0.8199, + "step": 15333 + }, + { + "epoch": 1.0640482964402194, + "grad_norm": 5.778421401977539, + "learning_rate": 0.00017218278912714227, + "loss": 0.6862, + "step": 15334 + }, + { + "epoch": 1.064117687877316, + "grad_norm": 1.869466781616211, + "learning_rate": 0.00017217890052334412, + "loss": 0.452, + "step": 15335 + }, + { + "epoch": 1.0641870793144126, + "grad_norm": 2.2408852577209473, + "learning_rate": 0.0001721750116916855, + "loss": 0.8438, + "step": 15336 + }, + { + "epoch": 1.0642564707515092, + "grad_norm": 2.474635124206543, + "learning_rate": 0.00017217112263217876, + "loss": 0.8301, + "step": 15337 + }, + { + "epoch": 1.064325862188606, + "grad_norm": 3.626396417617798, + "learning_rate": 0.00017216723334483616, + "loss": 0.6866, + "step": 15338 + }, + { + "epoch": 1.0643952536257026, + "grad_norm": 2.1890809535980225, + "learning_rate": 0.00017216334382966997, + "loss": 0.9657, + "step": 15339 + }, + { + "epoch": 1.0644646450627993, + "grad_norm": 1.7388904094696045, + "learning_rate": 0.00017215945408669253, + "loss": 0.7738, + "step": 15340 + }, + { + "epoch": 1.0645340364998959, + "grad_norm": 3.062124490737915, + "learning_rate": 0.000172155564115916, + "loss": 0.7213, + "step": 15341 + }, + { + "epoch": 1.0646034279369925, + "grad_norm": 1.7146544456481934, + "learning_rate": 0.00017215167391735275, + "loss": 0.7511, + "step": 15342 + }, + { + "epoch": 1.0646728193740893, + "grad_norm": 2.094365119934082, + "learning_rate": 0.000172147783491015, + "loss": 0.6845, + "step": 15343 + }, + { + "epoch": 1.064742210811186, + "grad_norm": 2.4690215587615967, + "learning_rate": 0.0001721438928369151, + "loss": 0.9181, + "step": 15344 + }, + { + "epoch": 1.0648116022482825, + "grad_norm": 3.81119966506958, + "learning_rate": 0.00017214000195506533, + "loss": 0.92, + "step": 15345 + }, + { + "epoch": 1.0648809936853791, + "grad_norm": 1.4664783477783203, + "learning_rate": 0.0001721361108454779, + "loss": 0.4154, + "step": 15346 + }, + { + "epoch": 1.064950385122476, + "grad_norm": 1.692368745803833, + "learning_rate": 0.00017213221950816517, + "loss": 0.731, + "step": 15347 + }, + { + "epoch": 1.0650197765595726, + "grad_norm": 2.4893996715545654, + "learning_rate": 0.00017212832794313935, + "loss": 0.8599, + "step": 15348 + }, + { + "epoch": 1.0650891679966692, + "grad_norm": 1.4822076559066772, + "learning_rate": 0.0001721244361504128, + "loss": 0.435, + "step": 15349 + }, + { + "epoch": 1.0651585594337658, + "grad_norm": 2.2684714794158936, + "learning_rate": 0.00017212054412999776, + "loss": 0.8097, + "step": 15350 + }, + { + "epoch": 1.0652279508708626, + "grad_norm": 2.6849350929260254, + "learning_rate": 0.00017211665188190654, + "loss": 0.6237, + "step": 15351 + }, + { + "epoch": 1.0652973423079592, + "grad_norm": 2.572847366333008, + "learning_rate": 0.00017211275940615141, + "loss": 1.0871, + "step": 15352 + }, + { + "epoch": 1.0653667337450559, + "grad_norm": 2.309187650680542, + "learning_rate": 0.00017210886670274467, + "loss": 0.7721, + "step": 15353 + }, + { + "epoch": 1.0654361251821525, + "grad_norm": 2.8145992755889893, + "learning_rate": 0.0001721049737716986, + "loss": 0.8467, + "step": 15354 + }, + { + "epoch": 1.065505516619249, + "grad_norm": 1.5390245914459229, + "learning_rate": 0.00017210108061302548, + "loss": 0.5269, + "step": 15355 + }, + { + "epoch": 1.065574908056346, + "grad_norm": 2.9958267211914062, + "learning_rate": 0.00017209718722673766, + "loss": 0.5877, + "step": 15356 + }, + { + "epoch": 1.0656442994934425, + "grad_norm": 2.458144187927246, + "learning_rate": 0.00017209329361284734, + "loss": 0.6843, + "step": 15357 + }, + { + "epoch": 1.0657136909305391, + "grad_norm": 2.4597952365875244, + "learning_rate": 0.00017208939977136683, + "loss": 0.7985, + "step": 15358 + }, + { + "epoch": 1.0657830823676357, + "grad_norm": 2.5029795169830322, + "learning_rate": 0.00017208550570230852, + "loss": 0.6307, + "step": 15359 + }, + { + "epoch": 1.0658524738047326, + "grad_norm": 3.1310479640960693, + "learning_rate": 0.0001720816114056846, + "loss": 0.7773, + "step": 15360 + }, + { + "epoch": 1.0659218652418292, + "grad_norm": 4.027185916900635, + "learning_rate": 0.0001720777168815074, + "loss": 0.7204, + "step": 15361 + }, + { + "epoch": 1.0659912566789258, + "grad_norm": 2.18843412399292, + "learning_rate": 0.0001720738221297892, + "loss": 0.5962, + "step": 15362 + }, + { + "epoch": 1.0660606481160224, + "grad_norm": 1.7732865810394287, + "learning_rate": 0.0001720699271505423, + "loss": 0.5025, + "step": 15363 + }, + { + "epoch": 1.0661300395531192, + "grad_norm": 2.517584800720215, + "learning_rate": 0.00017206603194377902, + "loss": 0.6766, + "step": 15364 + }, + { + "epoch": 1.0661994309902159, + "grad_norm": 2.135664224624634, + "learning_rate": 0.00017206213650951163, + "loss": 0.6228, + "step": 15365 + }, + { + "epoch": 1.0662688224273125, + "grad_norm": 2.0088648796081543, + "learning_rate": 0.00017205824084775244, + "loss": 0.7555, + "step": 15366 + }, + { + "epoch": 1.066338213864409, + "grad_norm": 1.7633923292160034, + "learning_rate": 0.00017205434495851373, + "loss": 0.5996, + "step": 15367 + }, + { + "epoch": 1.0664076053015057, + "grad_norm": 2.706064462661743, + "learning_rate": 0.00017205044884180783, + "loss": 0.7979, + "step": 15368 + }, + { + "epoch": 1.0664769967386025, + "grad_norm": 2.5595884323120117, + "learning_rate": 0.00017204655249764702, + "loss": 0.7089, + "step": 15369 + }, + { + "epoch": 1.0665463881756991, + "grad_norm": 1.8407810926437378, + "learning_rate": 0.00017204265592604361, + "loss": 0.5834, + "step": 15370 + }, + { + "epoch": 1.0666157796127957, + "grad_norm": 2.959622859954834, + "learning_rate": 0.00017203875912700988, + "loss": 0.5489, + "step": 15371 + }, + { + "epoch": 1.0666851710498924, + "grad_norm": 2.897937774658203, + "learning_rate": 0.00017203486210055812, + "loss": 0.7156, + "step": 15372 + }, + { + "epoch": 1.0667545624869892, + "grad_norm": 2.167663812637329, + "learning_rate": 0.0001720309648467007, + "loss": 0.6234, + "step": 15373 + }, + { + "epoch": 1.0668239539240858, + "grad_norm": 3.7843234539031982, + "learning_rate": 0.00017202706736544986, + "loss": 0.6123, + "step": 15374 + }, + { + "epoch": 1.0668933453611824, + "grad_norm": 3.5504488945007324, + "learning_rate": 0.00017202316965681793, + "loss": 0.7059, + "step": 15375 + }, + { + "epoch": 1.066962736798279, + "grad_norm": 3.5232787132263184, + "learning_rate": 0.00017201927172081724, + "loss": 0.5261, + "step": 15376 + }, + { + "epoch": 1.0670321282353759, + "grad_norm": 3.5883960723876953, + "learning_rate": 0.00017201537355746003, + "loss": 0.8965, + "step": 15377 + }, + { + "epoch": 1.0671015196724725, + "grad_norm": 2.481423854827881, + "learning_rate": 0.00017201147516675863, + "loss": 0.6991, + "step": 15378 + }, + { + "epoch": 1.067170911109569, + "grad_norm": 2.2155206203460693, + "learning_rate": 0.00017200757654872538, + "loss": 0.6261, + "step": 15379 + }, + { + "epoch": 1.0672403025466657, + "grad_norm": 4.098437786102295, + "learning_rate": 0.0001720036777033726, + "loss": 0.7794, + "step": 15380 + }, + { + "epoch": 1.0673096939837623, + "grad_norm": 2.226419448852539, + "learning_rate": 0.0001719997786307125, + "loss": 0.6854, + "step": 15381 + }, + { + "epoch": 1.0673790854208591, + "grad_norm": 3.448636054992676, + "learning_rate": 0.00017199587933075747, + "loss": 0.6184, + "step": 15382 + }, + { + "epoch": 1.0674484768579557, + "grad_norm": 1.5428770780563354, + "learning_rate": 0.00017199197980351979, + "loss": 0.5367, + "step": 15383 + }, + { + "epoch": 1.0675178682950524, + "grad_norm": 2.18365740776062, + "learning_rate": 0.0001719880800490118, + "loss": 0.7734, + "step": 15384 + }, + { + "epoch": 1.067587259732149, + "grad_norm": 2.803614616394043, + "learning_rate": 0.0001719841800672458, + "loss": 0.7358, + "step": 15385 + }, + { + "epoch": 1.0676566511692458, + "grad_norm": 2.118232250213623, + "learning_rate": 0.00017198027985823407, + "loss": 0.6373, + "step": 15386 + }, + { + "epoch": 1.0677260426063424, + "grad_norm": 1.9389822483062744, + "learning_rate": 0.00017197637942198898, + "loss": 0.5706, + "step": 15387 + }, + { + "epoch": 1.067795434043439, + "grad_norm": 1.9558521509170532, + "learning_rate": 0.00017197247875852282, + "loss": 0.6004, + "step": 15388 + }, + { + "epoch": 1.0678648254805356, + "grad_norm": 3.655109167098999, + "learning_rate": 0.00017196857786784788, + "loss": 0.8091, + "step": 15389 + }, + { + "epoch": 1.0679342169176325, + "grad_norm": 2.2264320850372314, + "learning_rate": 0.00017196467674997646, + "loss": 0.7016, + "step": 15390 + }, + { + "epoch": 1.068003608354729, + "grad_norm": 1.893069863319397, + "learning_rate": 0.00017196077540492094, + "loss": 0.6552, + "step": 15391 + }, + { + "epoch": 1.0680729997918257, + "grad_norm": 2.233269453048706, + "learning_rate": 0.0001719568738326936, + "loss": 0.7637, + "step": 15392 + }, + { + "epoch": 1.0681423912289223, + "grad_norm": 1.8932979106903076, + "learning_rate": 0.00017195297203330675, + "loss": 0.5395, + "step": 15393 + }, + { + "epoch": 1.068211782666019, + "grad_norm": 2.1184780597686768, + "learning_rate": 0.0001719490700067727, + "loss": 0.5969, + "step": 15394 + }, + { + "epoch": 1.0682811741031157, + "grad_norm": 2.230691432952881, + "learning_rate": 0.0001719451677531038, + "loss": 0.8335, + "step": 15395 + }, + { + "epoch": 1.0683505655402123, + "grad_norm": 2.237265110015869, + "learning_rate": 0.00017194126527231234, + "loss": 0.609, + "step": 15396 + }, + { + "epoch": 1.068419956977309, + "grad_norm": 2.5124778747558594, + "learning_rate": 0.00017193736256441067, + "loss": 0.6082, + "step": 15397 + }, + { + "epoch": 1.0684893484144056, + "grad_norm": 2.201338768005371, + "learning_rate": 0.00017193345962941106, + "loss": 0.846, + "step": 15398 + }, + { + "epoch": 1.0685587398515024, + "grad_norm": 2.1007883548736572, + "learning_rate": 0.0001719295564673259, + "loss": 0.5287, + "step": 15399 + }, + { + "epoch": 1.068628131288599, + "grad_norm": 1.7958550453186035, + "learning_rate": 0.00017192565307816746, + "loss": 0.6565, + "step": 15400 + }, + { + "epoch": 1.0686975227256956, + "grad_norm": 2.0785372257232666, + "learning_rate": 0.0001719217494619481, + "loss": 0.6653, + "step": 15401 + }, + { + "epoch": 1.0687669141627922, + "grad_norm": 3.5626931190490723, + "learning_rate": 0.0001719178456186801, + "loss": 0.7213, + "step": 15402 + }, + { + "epoch": 1.068836305599889, + "grad_norm": 2.1749300956726074, + "learning_rate": 0.0001719139415483758, + "loss": 0.6965, + "step": 15403 + }, + { + "epoch": 1.0689056970369857, + "grad_norm": 2.3091177940368652, + "learning_rate": 0.00017191003725104754, + "loss": 0.7595, + "step": 15404 + }, + { + "epoch": 1.0689750884740823, + "grad_norm": 2.776409864425659, + "learning_rate": 0.00017190613272670763, + "loss": 0.6383, + "step": 15405 + }, + { + "epoch": 1.069044479911179, + "grad_norm": 2.551168441772461, + "learning_rate": 0.0001719022279753684, + "loss": 0.6637, + "step": 15406 + }, + { + "epoch": 1.0691138713482755, + "grad_norm": 2.287956714630127, + "learning_rate": 0.00017189832299704217, + "loss": 0.6566, + "step": 15407 + }, + { + "epoch": 1.0691832627853723, + "grad_norm": 2.2092056274414062, + "learning_rate": 0.0001718944177917413, + "loss": 0.7583, + "step": 15408 + }, + { + "epoch": 1.069252654222469, + "grad_norm": 2.292983293533325, + "learning_rate": 0.0001718905123594781, + "loss": 0.6494, + "step": 15409 + }, + { + "epoch": 1.0693220456595656, + "grad_norm": 2.089179039001465, + "learning_rate": 0.00017188660670026487, + "loss": 0.5466, + "step": 15410 + }, + { + "epoch": 1.0693914370966622, + "grad_norm": 1.6089880466461182, + "learning_rate": 0.00017188270081411395, + "loss": 0.5762, + "step": 15411 + }, + { + "epoch": 1.069460828533759, + "grad_norm": 2.9985005855560303, + "learning_rate": 0.0001718787947010377, + "loss": 0.7137, + "step": 15412 + }, + { + "epoch": 1.0695302199708556, + "grad_norm": 4.191632270812988, + "learning_rate": 0.00017187488836104846, + "loss": 0.7208, + "step": 15413 + }, + { + "epoch": 1.0695996114079522, + "grad_norm": 1.7030400037765503, + "learning_rate": 0.00017187098179415852, + "loss": 0.585, + "step": 15414 + }, + { + "epoch": 1.0696690028450488, + "grad_norm": 3.9251866340637207, + "learning_rate": 0.00017186707500038023, + "loss": 1.1211, + "step": 15415 + }, + { + "epoch": 1.0697383942821457, + "grad_norm": 2.8744051456451416, + "learning_rate": 0.0001718631679797259, + "loss": 0.7166, + "step": 15416 + }, + { + "epoch": 1.0698077857192423, + "grad_norm": 2.0858047008514404, + "learning_rate": 0.00017185926073220791, + "loss": 0.7577, + "step": 15417 + }, + { + "epoch": 1.069877177156339, + "grad_norm": 2.4400479793548584, + "learning_rate": 0.00017185535325783854, + "loss": 0.7183, + "step": 15418 + }, + { + "epoch": 1.0699465685934355, + "grad_norm": 2.140880584716797, + "learning_rate": 0.00017185144555663018, + "loss": 0.905, + "step": 15419 + }, + { + "epoch": 1.0700159600305321, + "grad_norm": 2.139840602874756, + "learning_rate": 0.00017184753762859516, + "loss": 0.6329, + "step": 15420 + }, + { + "epoch": 1.070085351467629, + "grad_norm": 1.7064706087112427, + "learning_rate": 0.00017184362947374578, + "loss": 0.7247, + "step": 15421 + }, + { + "epoch": 1.0701547429047256, + "grad_norm": 1.7119039297103882, + "learning_rate": 0.0001718397210920944, + "loss": 0.6268, + "step": 15422 + }, + { + "epoch": 1.0702241343418222, + "grad_norm": 1.9116215705871582, + "learning_rate": 0.00017183581248365334, + "loss": 0.7127, + "step": 15423 + }, + { + "epoch": 1.0702935257789188, + "grad_norm": 2.34748911857605, + "learning_rate": 0.000171831903648435, + "loss": 0.5229, + "step": 15424 + }, + { + "epoch": 1.0703629172160156, + "grad_norm": 6.055117130279541, + "learning_rate": 0.00017182799458645166, + "loss": 0.7301, + "step": 15425 + }, + { + "epoch": 1.0704323086531122, + "grad_norm": 1.87600839138031, + "learning_rate": 0.00017182408529771564, + "loss": 0.6814, + "step": 15426 + }, + { + "epoch": 1.0705017000902088, + "grad_norm": 2.30619740486145, + "learning_rate": 0.00017182017578223931, + "loss": 0.676, + "step": 15427 + }, + { + "epoch": 1.0705710915273055, + "grad_norm": 2.50864839553833, + "learning_rate": 0.00017181626604003504, + "loss": 0.5849, + "step": 15428 + }, + { + "epoch": 1.0706404829644023, + "grad_norm": 2.5810887813568115, + "learning_rate": 0.00017181235607111517, + "loss": 0.7971, + "step": 15429 + }, + { + "epoch": 1.070709874401499, + "grad_norm": 1.8374927043914795, + "learning_rate": 0.000171808445875492, + "loss": 0.5652, + "step": 15430 + }, + { + "epoch": 1.0707792658385955, + "grad_norm": 3.221302032470703, + "learning_rate": 0.00017180453545317793, + "loss": 0.6749, + "step": 15431 + }, + { + "epoch": 1.0708486572756921, + "grad_norm": 2.5718867778778076, + "learning_rate": 0.00017180062480418525, + "loss": 0.6535, + "step": 15432 + }, + { + "epoch": 1.0709180487127887, + "grad_norm": 2.622213125228882, + "learning_rate": 0.0001717967139285263, + "loss": 0.6946, + "step": 15433 + }, + { + "epoch": 1.0709874401498856, + "grad_norm": 2.6062822341918945, + "learning_rate": 0.00017179280282621345, + "loss": 0.7995, + "step": 15434 + }, + { + "epoch": 1.0710568315869822, + "grad_norm": 2.108398675918579, + "learning_rate": 0.0001717888914972591, + "loss": 0.7497, + "step": 15435 + }, + { + "epoch": 1.0711262230240788, + "grad_norm": 2.3240914344787598, + "learning_rate": 0.00017178497994167551, + "loss": 0.7655, + "step": 15436 + }, + { + "epoch": 1.0711956144611754, + "grad_norm": 2.8961713314056396, + "learning_rate": 0.0001717810681594751, + "loss": 0.7316, + "step": 15437 + }, + { + "epoch": 1.0712650058982722, + "grad_norm": 2.790588140487671, + "learning_rate": 0.00017177715615067016, + "loss": 1.0309, + "step": 15438 + }, + { + "epoch": 1.0713343973353688, + "grad_norm": 3.0160226821899414, + "learning_rate": 0.0001717732439152731, + "loss": 0.8654, + "step": 15439 + }, + { + "epoch": 1.0714037887724654, + "grad_norm": 2.4517412185668945, + "learning_rate": 0.0001717693314532962, + "loss": 0.7592, + "step": 15440 + }, + { + "epoch": 1.071473180209562, + "grad_norm": 3.025946617126465, + "learning_rate": 0.00017176541876475184, + "loss": 0.7347, + "step": 15441 + }, + { + "epoch": 1.071542571646659, + "grad_norm": 2.133424997329712, + "learning_rate": 0.00017176150584965243, + "loss": 0.6104, + "step": 15442 + }, + { + "epoch": 1.0716119630837555, + "grad_norm": 1.8508843183517456, + "learning_rate": 0.00017175759270801022, + "loss": 0.6355, + "step": 15443 + }, + { + "epoch": 1.0716813545208521, + "grad_norm": 1.6538761854171753, + "learning_rate": 0.00017175367933983763, + "loss": 0.6004, + "step": 15444 + }, + { + "epoch": 1.0717507459579487, + "grad_norm": 1.7720119953155518, + "learning_rate": 0.00017174976574514702, + "loss": 0.6325, + "step": 15445 + }, + { + "epoch": 1.0718201373950453, + "grad_norm": 2.105168104171753, + "learning_rate": 0.0001717458519239507, + "loss": 0.9727, + "step": 15446 + }, + { + "epoch": 1.0718895288321422, + "grad_norm": 2.4815688133239746, + "learning_rate": 0.00017174193787626108, + "loss": 0.6687, + "step": 15447 + }, + { + "epoch": 1.0719589202692388, + "grad_norm": 2.414095878601074, + "learning_rate": 0.00017173802360209047, + "loss": 0.5168, + "step": 15448 + }, + { + "epoch": 1.0720283117063354, + "grad_norm": 1.7375701665878296, + "learning_rate": 0.00017173410910145123, + "loss": 0.4958, + "step": 15449 + }, + { + "epoch": 1.072097703143432, + "grad_norm": 2.29394268989563, + "learning_rate": 0.00017173019437435573, + "loss": 0.6539, + "step": 15450 + }, + { + "epoch": 1.0721670945805288, + "grad_norm": 1.8529274463653564, + "learning_rate": 0.00017172627942081635, + "loss": 0.6626, + "step": 15451 + }, + { + "epoch": 1.0722364860176254, + "grad_norm": 1.5402520895004272, + "learning_rate": 0.00017172236424084543, + "loss": 0.6836, + "step": 15452 + }, + { + "epoch": 1.072305877454722, + "grad_norm": 2.5465614795684814, + "learning_rate": 0.0001717184488344553, + "loss": 0.8542, + "step": 15453 + }, + { + "epoch": 1.0723752688918187, + "grad_norm": 2.370962381362915, + "learning_rate": 0.00017171453320165837, + "loss": 0.9132, + "step": 15454 + }, + { + "epoch": 1.0724446603289155, + "grad_norm": 2.0890121459960938, + "learning_rate": 0.00017171061734246694, + "loss": 0.494, + "step": 15455 + }, + { + "epoch": 1.072514051766012, + "grad_norm": 1.8410708904266357, + "learning_rate": 0.00017170670125689344, + "loss": 0.7464, + "step": 15456 + }, + { + "epoch": 1.0725834432031087, + "grad_norm": 2.141810894012451, + "learning_rate": 0.0001717027849449502, + "loss": 0.6033, + "step": 15457 + }, + { + "epoch": 1.0726528346402053, + "grad_norm": 2.075737237930298, + "learning_rate": 0.00017169886840664962, + "loss": 0.7777, + "step": 15458 + }, + { + "epoch": 1.072722226077302, + "grad_norm": 1.9186230897903442, + "learning_rate": 0.00017169495164200398, + "loss": 0.8482, + "step": 15459 + }, + { + "epoch": 1.0727916175143988, + "grad_norm": 1.7737646102905273, + "learning_rate": 0.00017169103465102572, + "loss": 0.6545, + "step": 15460 + }, + { + "epoch": 1.0728610089514954, + "grad_norm": 3.544059991836548, + "learning_rate": 0.00017168711743372717, + "loss": 1.0811, + "step": 15461 + }, + { + "epoch": 1.072930400388592, + "grad_norm": 2.038686513900757, + "learning_rate": 0.0001716831999901207, + "loss": 0.64, + "step": 15462 + }, + { + "epoch": 1.0729997918256886, + "grad_norm": 2.003666639328003, + "learning_rate": 0.00017167928232021874, + "loss": 0.517, + "step": 15463 + }, + { + "epoch": 1.0730691832627854, + "grad_norm": 2.855329751968384, + "learning_rate": 0.00017167536442403352, + "loss": 0.6986, + "step": 15464 + }, + { + "epoch": 1.073138574699882, + "grad_norm": 1.9928741455078125, + "learning_rate": 0.00017167144630157757, + "loss": 0.6811, + "step": 15465 + }, + { + "epoch": 1.0732079661369787, + "grad_norm": 2.2194597721099854, + "learning_rate": 0.00017166752795286314, + "loss": 0.6841, + "step": 15466 + }, + { + "epoch": 1.0732773575740753, + "grad_norm": 2.0023348331451416, + "learning_rate": 0.00017166360937790262, + "loss": 0.489, + "step": 15467 + }, + { + "epoch": 1.073346749011172, + "grad_norm": 1.776113748550415, + "learning_rate": 0.00017165969057670842, + "loss": 0.6695, + "step": 15468 + }, + { + "epoch": 1.0734161404482687, + "grad_norm": 1.9752246141433716, + "learning_rate": 0.00017165577154929288, + "loss": 0.5811, + "step": 15469 + }, + { + "epoch": 1.0734855318853653, + "grad_norm": 2.437614679336548, + "learning_rate": 0.0001716518522956684, + "loss": 0.6335, + "step": 15470 + }, + { + "epoch": 1.073554923322462, + "grad_norm": 1.5590991973876953, + "learning_rate": 0.00017164793281584736, + "loss": 0.709, + "step": 15471 + }, + { + "epoch": 1.0736243147595588, + "grad_norm": 1.937529444694519, + "learning_rate": 0.00017164401310984205, + "loss": 0.6295, + "step": 15472 + }, + { + "epoch": 1.0736937061966554, + "grad_norm": 1.8881993293762207, + "learning_rate": 0.00017164009317766494, + "loss": 0.6675, + "step": 15473 + }, + { + "epoch": 1.073763097633752, + "grad_norm": 1.6120113134384155, + "learning_rate": 0.00017163617301932838, + "loss": 0.6051, + "step": 15474 + }, + { + "epoch": 1.0738324890708486, + "grad_norm": 2.1135215759277344, + "learning_rate": 0.0001716322526348447, + "loss": 0.6129, + "step": 15475 + }, + { + "epoch": 1.0739018805079452, + "grad_norm": 2.0742077827453613, + "learning_rate": 0.0001716283320242263, + "loss": 0.6425, + "step": 15476 + }, + { + "epoch": 1.073971271945042, + "grad_norm": 2.117708444595337, + "learning_rate": 0.0001716244111874856, + "loss": 0.8872, + "step": 15477 + }, + { + "epoch": 1.0740406633821387, + "grad_norm": 1.4239833354949951, + "learning_rate": 0.00017162049012463496, + "loss": 0.5206, + "step": 15478 + }, + { + "epoch": 1.0741100548192353, + "grad_norm": 2.0306830406188965, + "learning_rate": 0.00017161656883568668, + "loss": 0.6501, + "step": 15479 + }, + { + "epoch": 1.0741794462563319, + "grad_norm": 1.5911331176757812, + "learning_rate": 0.00017161264732065324, + "loss": 0.4478, + "step": 15480 + }, + { + "epoch": 1.0742488376934287, + "grad_norm": 1.6978230476379395, + "learning_rate": 0.00017160872557954698, + "loss": 0.5106, + "step": 15481 + }, + { + "epoch": 1.0743182291305253, + "grad_norm": 2.2219340801239014, + "learning_rate": 0.00017160480361238029, + "loss": 0.6503, + "step": 15482 + }, + { + "epoch": 1.074387620567622, + "grad_norm": 2.295482873916626, + "learning_rate": 0.00017160088141916552, + "loss": 0.9297, + "step": 15483 + }, + { + "epoch": 1.0744570120047185, + "grad_norm": 1.7167309522628784, + "learning_rate": 0.00017159695899991508, + "loss": 0.5256, + "step": 15484 + }, + { + "epoch": 1.0745264034418154, + "grad_norm": 1.8739027976989746, + "learning_rate": 0.00017159303635464138, + "loss": 0.6973, + "step": 15485 + }, + { + "epoch": 1.074595794878912, + "grad_norm": 1.7965812683105469, + "learning_rate": 0.0001715891134833567, + "loss": 0.6948, + "step": 15486 + }, + { + "epoch": 1.0746651863160086, + "grad_norm": 1.3466427326202393, + "learning_rate": 0.00017158519038607356, + "loss": 0.5231, + "step": 15487 + }, + { + "epoch": 1.0747345777531052, + "grad_norm": 2.1240272521972656, + "learning_rate": 0.00017158126706280427, + "loss": 1.0487, + "step": 15488 + }, + { + "epoch": 1.074803969190202, + "grad_norm": 2.3931922912597656, + "learning_rate": 0.0001715773435135612, + "loss": 0.7154, + "step": 15489 + }, + { + "epoch": 1.0748733606272987, + "grad_norm": 3.0512032508850098, + "learning_rate": 0.00017157341973835676, + "loss": 0.6328, + "step": 15490 + }, + { + "epoch": 1.0749427520643953, + "grad_norm": 1.7838577032089233, + "learning_rate": 0.00017156949573720335, + "loss": 0.6969, + "step": 15491 + }, + { + "epoch": 1.0750121435014919, + "grad_norm": 1.6463713645935059, + "learning_rate": 0.00017156557151011334, + "loss": 0.619, + "step": 15492 + }, + { + "epoch": 1.0750815349385885, + "grad_norm": 1.8780672550201416, + "learning_rate": 0.00017156164705709915, + "loss": 0.5715, + "step": 15493 + }, + { + "epoch": 1.0751509263756853, + "grad_norm": 2.193481206893921, + "learning_rate": 0.00017155772237817312, + "loss": 1.0474, + "step": 15494 + }, + { + "epoch": 1.075220317812782, + "grad_norm": 2.8878955841064453, + "learning_rate": 0.00017155379747334765, + "loss": 0.7636, + "step": 15495 + }, + { + "epoch": 1.0752897092498785, + "grad_norm": 1.6794347763061523, + "learning_rate": 0.00017154987234263517, + "loss": 0.6742, + "step": 15496 + }, + { + "epoch": 1.0753591006869752, + "grad_norm": 5.214985370635986, + "learning_rate": 0.00017154594698604803, + "loss": 0.7595, + "step": 15497 + }, + { + "epoch": 1.075428492124072, + "grad_norm": 3.374499797821045, + "learning_rate": 0.00017154202140359865, + "loss": 0.8933, + "step": 15498 + }, + { + "epoch": 1.0754978835611686, + "grad_norm": 2.0738494396209717, + "learning_rate": 0.0001715380955952994, + "loss": 0.7429, + "step": 15499 + }, + { + "epoch": 1.0755672749982652, + "grad_norm": 3.447948455810547, + "learning_rate": 0.00017153416956116265, + "loss": 0.9208, + "step": 15500 + }, + { + "epoch": 1.0756366664353618, + "grad_norm": 2.4296083450317383, + "learning_rate": 0.0001715302433012009, + "loss": 0.6996, + "step": 15501 + }, + { + "epoch": 1.0757060578724587, + "grad_norm": 2.4661405086517334, + "learning_rate": 0.0001715263168154264, + "loss": 0.7533, + "step": 15502 + }, + { + "epoch": 1.0757754493095553, + "grad_norm": 2.3949098587036133, + "learning_rate": 0.00017152239010385166, + "loss": 0.9455, + "step": 15503 + }, + { + "epoch": 1.0758448407466519, + "grad_norm": 2.28294038772583, + "learning_rate": 0.000171518463166489, + "loss": 0.7463, + "step": 15504 + }, + { + "epoch": 1.0759142321837485, + "grad_norm": 1.7363284826278687, + "learning_rate": 0.0001715145360033509, + "loss": 0.5689, + "step": 15505 + }, + { + "epoch": 1.075983623620845, + "grad_norm": 2.024242639541626, + "learning_rate": 0.00017151060861444965, + "loss": 0.918, + "step": 15506 + }, + { + "epoch": 1.076053015057942, + "grad_norm": 2.4570999145507812, + "learning_rate": 0.00017150668099979774, + "loss": 0.6477, + "step": 15507 + }, + { + "epoch": 1.0761224064950385, + "grad_norm": 2.322946071624756, + "learning_rate": 0.0001715027531594075, + "loss": 0.8064, + "step": 15508 + }, + { + "epoch": 1.0761917979321352, + "grad_norm": 2.3207812309265137, + "learning_rate": 0.0001714988250932914, + "loss": 0.8093, + "step": 15509 + }, + { + "epoch": 1.0762611893692318, + "grad_norm": 2.214566230773926, + "learning_rate": 0.0001714948968014618, + "loss": 0.7251, + "step": 15510 + }, + { + "epoch": 1.0763305808063286, + "grad_norm": 2.4216175079345703, + "learning_rate": 0.0001714909682839311, + "loss": 0.8407, + "step": 15511 + }, + { + "epoch": 1.0763999722434252, + "grad_norm": 1.9273067712783813, + "learning_rate": 0.0001714870395407117, + "loss": 0.8743, + "step": 15512 + }, + { + "epoch": 1.0764693636805218, + "grad_norm": 1.8970181941986084, + "learning_rate": 0.00017148311057181602, + "loss": 0.7019, + "step": 15513 + }, + { + "epoch": 1.0765387551176184, + "grad_norm": 2.0273876190185547, + "learning_rate": 0.00017147918137725644, + "loss": 0.6256, + "step": 15514 + }, + { + "epoch": 1.0766081465547153, + "grad_norm": 2.227282762527466, + "learning_rate": 0.0001714752519570454, + "loss": 0.6853, + "step": 15515 + }, + { + "epoch": 1.0766775379918119, + "grad_norm": 4.962366580963135, + "learning_rate": 0.00017147132231119528, + "loss": 0.7268, + "step": 15516 + }, + { + "epoch": 1.0767469294289085, + "grad_norm": 1.7870962619781494, + "learning_rate": 0.00017146739243971846, + "loss": 0.523, + "step": 15517 + }, + { + "epoch": 1.076816320866005, + "grad_norm": 1.8196862936019897, + "learning_rate": 0.00017146346234262738, + "loss": 0.7808, + "step": 15518 + }, + { + "epoch": 1.0768857123031017, + "grad_norm": 2.147221803665161, + "learning_rate": 0.00017145953201993445, + "loss": 0.6506, + "step": 15519 + }, + { + "epoch": 1.0769551037401985, + "grad_norm": 1.9532400369644165, + "learning_rate": 0.00017145560147165207, + "loss": 0.4925, + "step": 15520 + }, + { + "epoch": 1.0770244951772951, + "grad_norm": 1.5741552114486694, + "learning_rate": 0.00017145167069779264, + "loss": 0.5134, + "step": 15521 + }, + { + "epoch": 1.0770938866143918, + "grad_norm": 1.9936786890029907, + "learning_rate": 0.0001714477396983686, + "loss": 0.6172, + "step": 15522 + }, + { + "epoch": 1.0771632780514884, + "grad_norm": 2.28279185295105, + "learning_rate": 0.0001714438084733923, + "loss": 0.9043, + "step": 15523 + }, + { + "epoch": 1.0772326694885852, + "grad_norm": 2.434687376022339, + "learning_rate": 0.0001714398770228762, + "loss": 0.8753, + "step": 15524 + }, + { + "epoch": 1.0773020609256818, + "grad_norm": 1.3982023000717163, + "learning_rate": 0.0001714359453468327, + "loss": 0.4213, + "step": 15525 + }, + { + "epoch": 1.0773714523627784, + "grad_norm": 2.467435836791992, + "learning_rate": 0.00017143201344527423, + "loss": 0.9422, + "step": 15526 + }, + { + "epoch": 1.077440843799875, + "grad_norm": 1.5540215969085693, + "learning_rate": 0.00017142808131821315, + "loss": 0.5643, + "step": 15527 + }, + { + "epoch": 1.0775102352369719, + "grad_norm": 1.8969210386276245, + "learning_rate": 0.0001714241489656619, + "loss": 0.7268, + "step": 15528 + }, + { + "epoch": 1.0775796266740685, + "grad_norm": 3.4681003093719482, + "learning_rate": 0.00017142021638763292, + "loss": 0.9096, + "step": 15529 + }, + { + "epoch": 1.077649018111165, + "grad_norm": 1.5866767168045044, + "learning_rate": 0.0001714162835841386, + "loss": 0.687, + "step": 15530 + }, + { + "epoch": 1.0777184095482617, + "grad_norm": 1.7857465744018555, + "learning_rate": 0.00017141235055519137, + "loss": 0.7023, + "step": 15531 + }, + { + "epoch": 1.0777878009853583, + "grad_norm": 1.8410499095916748, + "learning_rate": 0.0001714084173008036, + "loss": 0.6991, + "step": 15532 + }, + { + "epoch": 1.0778571924224551, + "grad_norm": 1.826348066329956, + "learning_rate": 0.00017140448382098775, + "loss": 0.7406, + "step": 15533 + }, + { + "epoch": 1.0779265838595518, + "grad_norm": 2.566535472869873, + "learning_rate": 0.00017140055011575624, + "loss": 0.5992, + "step": 15534 + }, + { + "epoch": 1.0779959752966484, + "grad_norm": 6.104402542114258, + "learning_rate": 0.0001713966161851215, + "loss": 0.8266, + "step": 15535 + }, + { + "epoch": 1.078065366733745, + "grad_norm": 1.6618759632110596, + "learning_rate": 0.00017139268202909585, + "loss": 0.6568, + "step": 15536 + }, + { + "epoch": 1.0781347581708418, + "grad_norm": 2.8442816734313965, + "learning_rate": 0.00017138874764769185, + "loss": 0.7924, + "step": 15537 + }, + { + "epoch": 1.0782041496079384, + "grad_norm": 1.7120652198791504, + "learning_rate": 0.00017138481304092187, + "loss": 0.7975, + "step": 15538 + }, + { + "epoch": 1.078273541045035, + "grad_norm": 1.6874780654907227, + "learning_rate": 0.00017138087820879827, + "loss": 0.6432, + "step": 15539 + }, + { + "epoch": 1.0783429324821316, + "grad_norm": 2.2455506324768066, + "learning_rate": 0.00017137694315133356, + "loss": 0.6866, + "step": 15540 + }, + { + "epoch": 1.0784123239192285, + "grad_norm": 2.1276402473449707, + "learning_rate": 0.0001713730078685401, + "loss": 0.7997, + "step": 15541 + }, + { + "epoch": 1.078481715356325, + "grad_norm": 1.9238934516906738, + "learning_rate": 0.00017136907236043033, + "loss": 0.6664, + "step": 15542 + }, + { + "epoch": 1.0785511067934217, + "grad_norm": 1.7126206159591675, + "learning_rate": 0.00017136513662701668, + "loss": 0.5109, + "step": 15543 + }, + { + "epoch": 1.0786204982305183, + "grad_norm": 2.095046043395996, + "learning_rate": 0.00017136120066831159, + "loss": 0.9508, + "step": 15544 + }, + { + "epoch": 1.078689889667615, + "grad_norm": 2.6286799907684326, + "learning_rate": 0.00017135726448432744, + "loss": 0.7241, + "step": 15545 + }, + { + "epoch": 1.0787592811047118, + "grad_norm": 2.305185079574585, + "learning_rate": 0.00017135332807507672, + "loss": 0.7462, + "step": 15546 + }, + { + "epoch": 1.0788286725418084, + "grad_norm": 1.9387297630310059, + "learning_rate": 0.0001713493914405718, + "loss": 0.6597, + "step": 15547 + }, + { + "epoch": 1.078898063978905, + "grad_norm": 1.792968988418579, + "learning_rate": 0.00017134545458082515, + "loss": 0.6254, + "step": 15548 + }, + { + "epoch": 1.0789674554160016, + "grad_norm": 2.0794153213500977, + "learning_rate": 0.00017134151749584916, + "loss": 0.6618, + "step": 15549 + }, + { + "epoch": 1.0790368468530984, + "grad_norm": 1.9064977169036865, + "learning_rate": 0.00017133758018565628, + "loss": 0.6377, + "step": 15550 + }, + { + "epoch": 1.079106238290195, + "grad_norm": 1.6234204769134521, + "learning_rate": 0.00017133364265025893, + "loss": 0.7444, + "step": 15551 + }, + { + "epoch": 1.0791756297272916, + "grad_norm": 1.7677197456359863, + "learning_rate": 0.00017132970488966954, + "loss": 0.6817, + "step": 15552 + }, + { + "epoch": 1.0792450211643883, + "grad_norm": 2.3252310752868652, + "learning_rate": 0.00017132576690390058, + "loss": 0.6163, + "step": 15553 + }, + { + "epoch": 1.079314412601485, + "grad_norm": 2.0265607833862305, + "learning_rate": 0.0001713218286929644, + "loss": 0.5683, + "step": 15554 + }, + { + "epoch": 1.0793838040385817, + "grad_norm": 2.2772951126098633, + "learning_rate": 0.00017131789025687352, + "loss": 0.8288, + "step": 15555 + }, + { + "epoch": 1.0794531954756783, + "grad_norm": 1.7483166456222534, + "learning_rate": 0.00017131395159564034, + "loss": 0.4449, + "step": 15556 + }, + { + "epoch": 1.079522586912775, + "grad_norm": 1.9072946310043335, + "learning_rate": 0.00017131001270927727, + "loss": 0.6513, + "step": 15557 + }, + { + "epoch": 1.0795919783498715, + "grad_norm": 2.008000135421753, + "learning_rate": 0.00017130607359779675, + "loss": 0.6678, + "step": 15558 + }, + { + "epoch": 1.0796613697869684, + "grad_norm": 1.867972731590271, + "learning_rate": 0.00017130213426121125, + "loss": 0.6649, + "step": 15559 + }, + { + "epoch": 1.079730761224065, + "grad_norm": 2.732090711593628, + "learning_rate": 0.00017129819469953317, + "loss": 0.6349, + "step": 15560 + }, + { + "epoch": 1.0798001526611616, + "grad_norm": 1.7728745937347412, + "learning_rate": 0.000171294254912775, + "loss": 0.7615, + "step": 15561 + }, + { + "epoch": 1.0798695440982582, + "grad_norm": 1.6661601066589355, + "learning_rate": 0.0001712903149009491, + "loss": 0.5036, + "step": 15562 + }, + { + "epoch": 1.079938935535355, + "grad_norm": 1.9065005779266357, + "learning_rate": 0.00017128637466406796, + "loss": 0.6769, + "step": 15563 + }, + { + "epoch": 1.0800083269724516, + "grad_norm": 1.7672498226165771, + "learning_rate": 0.000171282434202144, + "loss": 0.6734, + "step": 15564 + }, + { + "epoch": 1.0800777184095482, + "grad_norm": 1.885562777519226, + "learning_rate": 0.00017127849351518965, + "loss": 0.6935, + "step": 15565 + }, + { + "epoch": 1.0801471098466449, + "grad_norm": 1.872289776802063, + "learning_rate": 0.00017127455260321738, + "loss": 0.6306, + "step": 15566 + }, + { + "epoch": 1.0802165012837417, + "grad_norm": 2.263514518737793, + "learning_rate": 0.00017127061146623963, + "loss": 0.7798, + "step": 15567 + }, + { + "epoch": 1.0802858927208383, + "grad_norm": 2.0249087810516357, + "learning_rate": 0.00017126667010426884, + "loss": 0.5258, + "step": 15568 + }, + { + "epoch": 1.080355284157935, + "grad_norm": 2.797232151031494, + "learning_rate": 0.0001712627285173174, + "loss": 0.7413, + "step": 15569 + }, + { + "epoch": 1.0804246755950315, + "grad_norm": 3.0583157539367676, + "learning_rate": 0.0001712587867053978, + "loss": 0.8796, + "step": 15570 + }, + { + "epoch": 1.0804940670321281, + "grad_norm": 3.1690189838409424, + "learning_rate": 0.00017125484466852251, + "loss": 0.8326, + "step": 15571 + }, + { + "epoch": 1.080563458469225, + "grad_norm": 2.9868571758270264, + "learning_rate": 0.00017125090240670393, + "loss": 0.6626, + "step": 15572 + }, + { + "epoch": 1.0806328499063216, + "grad_norm": 1.5139245986938477, + "learning_rate": 0.0001712469599199545, + "loss": 0.4316, + "step": 15573 + }, + { + "epoch": 1.0807022413434182, + "grad_norm": 1.1902843713760376, + "learning_rate": 0.0001712430172082867, + "loss": 0.4014, + "step": 15574 + }, + { + "epoch": 1.0807716327805148, + "grad_norm": 2.094233989715576, + "learning_rate": 0.00017123907427171297, + "loss": 0.6781, + "step": 15575 + }, + { + "epoch": 1.0808410242176116, + "grad_norm": 2.4990220069885254, + "learning_rate": 0.0001712351311102457, + "loss": 0.8213, + "step": 15576 + }, + { + "epoch": 1.0809104156547082, + "grad_norm": 2.4452991485595703, + "learning_rate": 0.00017123118772389745, + "loss": 0.8812, + "step": 15577 + }, + { + "epoch": 1.0809798070918049, + "grad_norm": 2.0908596515655518, + "learning_rate": 0.0001712272441126806, + "loss": 0.6203, + "step": 15578 + }, + { + "epoch": 1.0810491985289015, + "grad_norm": 2.0510027408599854, + "learning_rate": 0.00017122330027660756, + "loss": 0.9839, + "step": 15579 + }, + { + "epoch": 1.0811185899659983, + "grad_norm": 1.8724995851516724, + "learning_rate": 0.00017121935621569083, + "loss": 0.6132, + "step": 15580 + }, + { + "epoch": 1.081187981403095, + "grad_norm": 1.748106598854065, + "learning_rate": 0.00017121541192994288, + "loss": 0.5258, + "step": 15581 + }, + { + "epoch": 1.0812573728401915, + "grad_norm": 2.0406603813171387, + "learning_rate": 0.00017121146741937615, + "loss": 0.484, + "step": 15582 + }, + { + "epoch": 1.0813267642772881, + "grad_norm": 4.952658653259277, + "learning_rate": 0.00017120752268400303, + "loss": 0.8173, + "step": 15583 + }, + { + "epoch": 1.0813961557143847, + "grad_norm": 1.9038937091827393, + "learning_rate": 0.00017120357772383606, + "loss": 0.8644, + "step": 15584 + }, + { + "epoch": 1.0814655471514816, + "grad_norm": 2.803130626678467, + "learning_rate": 0.00017119963253888764, + "loss": 0.9523, + "step": 15585 + }, + { + "epoch": 1.0815349385885782, + "grad_norm": 2.390671491622925, + "learning_rate": 0.00017119568712917026, + "loss": 0.6859, + "step": 15586 + }, + { + "epoch": 1.0816043300256748, + "grad_norm": 2.307239055633545, + "learning_rate": 0.00017119174149469633, + "loss": 0.6912, + "step": 15587 + }, + { + "epoch": 1.0816737214627714, + "grad_norm": 2.4886341094970703, + "learning_rate": 0.00017118779563547834, + "loss": 0.9222, + "step": 15588 + }, + { + "epoch": 1.0817431128998682, + "grad_norm": 1.8325037956237793, + "learning_rate": 0.00017118384955152873, + "loss": 0.5576, + "step": 15589 + }, + { + "epoch": 1.0818125043369649, + "grad_norm": 1.6709718704223633, + "learning_rate": 0.00017117990324286, + "loss": 0.5742, + "step": 15590 + }, + { + "epoch": 1.0818818957740615, + "grad_norm": 2.373616933822632, + "learning_rate": 0.00017117595670948453, + "loss": 0.7001, + "step": 15591 + }, + { + "epoch": 1.081951287211158, + "grad_norm": 2.6651484966278076, + "learning_rate": 0.00017117200995141486, + "loss": 0.9267, + "step": 15592 + }, + { + "epoch": 1.082020678648255, + "grad_norm": 3.0013957023620605, + "learning_rate": 0.00017116806296866338, + "loss": 0.8387, + "step": 15593 + }, + { + "epoch": 1.0820900700853515, + "grad_norm": 2.5136239528656006, + "learning_rate": 0.0001711641157612426, + "loss": 0.6994, + "step": 15594 + }, + { + "epoch": 1.0821594615224481, + "grad_norm": 2.4068984985351562, + "learning_rate": 0.00017116016832916496, + "loss": 0.9289, + "step": 15595 + }, + { + "epoch": 1.0822288529595447, + "grad_norm": 2.32421612739563, + "learning_rate": 0.00017115622067244294, + "loss": 0.5918, + "step": 15596 + }, + { + "epoch": 1.0822982443966414, + "grad_norm": 1.8747786283493042, + "learning_rate": 0.00017115227279108897, + "loss": 0.5417, + "step": 15597 + }, + { + "epoch": 1.0823676358337382, + "grad_norm": 1.8056548833847046, + "learning_rate": 0.00017114832468511555, + "loss": 0.9202, + "step": 15598 + }, + { + "epoch": 1.0824370272708348, + "grad_norm": 2.0223684310913086, + "learning_rate": 0.0001711443763545351, + "loss": 0.6255, + "step": 15599 + }, + { + "epoch": 1.0825064187079314, + "grad_norm": 2.911771297454834, + "learning_rate": 0.0001711404277993601, + "loss": 0.5937, + "step": 15600 + }, + { + "epoch": 1.082575810145028, + "grad_norm": 2.5233068466186523, + "learning_rate": 0.00017113647901960305, + "loss": 0.8445, + "step": 15601 + }, + { + "epoch": 1.0826452015821248, + "grad_norm": 2.3461997509002686, + "learning_rate": 0.00017113253001527635, + "loss": 0.9036, + "step": 15602 + }, + { + "epoch": 1.0827145930192215, + "grad_norm": 1.8697407245635986, + "learning_rate": 0.00017112858078639255, + "loss": 0.762, + "step": 15603 + }, + { + "epoch": 1.082783984456318, + "grad_norm": 2.6299116611480713, + "learning_rate": 0.00017112463133296405, + "loss": 0.9068, + "step": 15604 + }, + { + "epoch": 1.0828533758934147, + "grad_norm": 2.3066840171813965, + "learning_rate": 0.00017112068165500336, + "loss": 0.7673, + "step": 15605 + }, + { + "epoch": 1.0829227673305115, + "grad_norm": 2.2998695373535156, + "learning_rate": 0.0001711167317525229, + "loss": 0.5732, + "step": 15606 + }, + { + "epoch": 1.0829921587676081, + "grad_norm": 2.1859569549560547, + "learning_rate": 0.00017111278162553522, + "loss": 0.7385, + "step": 15607 + }, + { + "epoch": 1.0830615502047047, + "grad_norm": 1.9602922201156616, + "learning_rate": 0.00017110883127405271, + "loss": 0.6406, + "step": 15608 + }, + { + "epoch": 1.0831309416418013, + "grad_norm": 1.9100520610809326, + "learning_rate": 0.00017110488069808786, + "loss": 0.7541, + "step": 15609 + }, + { + "epoch": 1.083200333078898, + "grad_norm": 2.32669734954834, + "learning_rate": 0.00017110092989765314, + "loss": 0.7975, + "step": 15610 + }, + { + "epoch": 1.0832697245159948, + "grad_norm": 2.2764832973480225, + "learning_rate": 0.00017109697887276104, + "loss": 0.7804, + "step": 15611 + }, + { + "epoch": 1.0833391159530914, + "grad_norm": 2.1712656021118164, + "learning_rate": 0.00017109302762342404, + "loss": 0.6629, + "step": 15612 + }, + { + "epoch": 1.083408507390188, + "grad_norm": 2.719956636428833, + "learning_rate": 0.0001710890761496546, + "loss": 0.5927, + "step": 15613 + }, + { + "epoch": 1.0834778988272846, + "grad_norm": 1.7202837467193604, + "learning_rate": 0.00017108512445146518, + "loss": 0.6448, + "step": 15614 + }, + { + "epoch": 1.0835472902643815, + "grad_norm": 1.484214425086975, + "learning_rate": 0.0001710811725288683, + "loss": 0.786, + "step": 15615 + }, + { + "epoch": 1.083616681701478, + "grad_norm": 2.5055484771728516, + "learning_rate": 0.00017107722038187638, + "loss": 0.8438, + "step": 15616 + }, + { + "epoch": 1.0836860731385747, + "grad_norm": 1.8965920209884644, + "learning_rate": 0.00017107326801050191, + "loss": 0.6312, + "step": 15617 + }, + { + "epoch": 1.0837554645756713, + "grad_norm": 1.6866072416305542, + "learning_rate": 0.00017106931541475742, + "loss": 0.6274, + "step": 15618 + }, + { + "epoch": 1.0838248560127681, + "grad_norm": 2.139683723449707, + "learning_rate": 0.0001710653625946553, + "loss": 0.8737, + "step": 15619 + }, + { + "epoch": 1.0838942474498647, + "grad_norm": 1.8861562013626099, + "learning_rate": 0.0001710614095502081, + "loss": 0.6068, + "step": 15620 + }, + { + "epoch": 1.0839636388869613, + "grad_norm": 1.870115876197815, + "learning_rate": 0.00017105745628142826, + "loss": 0.5883, + "step": 15621 + }, + { + "epoch": 1.084033030324058, + "grad_norm": 2.0601422786712646, + "learning_rate": 0.0001710535027883283, + "loss": 0.7638, + "step": 15622 + }, + { + "epoch": 1.0841024217611546, + "grad_norm": 2.40460467338562, + "learning_rate": 0.00017104954907092067, + "loss": 0.7161, + "step": 15623 + }, + { + "epoch": 1.0841718131982514, + "grad_norm": 1.7182037830352783, + "learning_rate": 0.00017104559512921784, + "loss": 0.5432, + "step": 15624 + }, + { + "epoch": 1.084241204635348, + "grad_norm": 2.5393710136413574, + "learning_rate": 0.0001710416409632323, + "loss": 0.8369, + "step": 15625 + }, + { + "epoch": 1.0843105960724446, + "grad_norm": 1.8011341094970703, + "learning_rate": 0.00017103768657297658, + "loss": 0.582, + "step": 15626 + }, + { + "epoch": 1.0843799875095412, + "grad_norm": 1.7243828773498535, + "learning_rate": 0.00017103373195846312, + "loss": 0.5826, + "step": 15627 + }, + { + "epoch": 1.084449378946638, + "grad_norm": 1.8964531421661377, + "learning_rate": 0.0001710297771197044, + "loss": 0.6377, + "step": 15628 + }, + { + "epoch": 1.0845187703837347, + "grad_norm": 1.9747883081436157, + "learning_rate": 0.0001710258220567129, + "loss": 0.6334, + "step": 15629 + }, + { + "epoch": 1.0845881618208313, + "grad_norm": 1.7387598752975464, + "learning_rate": 0.0001710218667695011, + "loss": 0.6673, + "step": 15630 + }, + { + "epoch": 1.084657553257928, + "grad_norm": 2.3233911991119385, + "learning_rate": 0.00017101791125808156, + "loss": 0.7748, + "step": 15631 + }, + { + "epoch": 1.0847269446950247, + "grad_norm": 1.9649875164031982, + "learning_rate": 0.0001710139555224667, + "loss": 0.8789, + "step": 15632 + }, + { + "epoch": 1.0847963361321213, + "grad_norm": 1.8972183465957642, + "learning_rate": 0.000171009999562669, + "loss": 0.6296, + "step": 15633 + }, + { + "epoch": 1.084865727569218, + "grad_norm": 2.0212974548339844, + "learning_rate": 0.000171006043378701, + "loss": 0.8361, + "step": 15634 + }, + { + "epoch": 1.0849351190063146, + "grad_norm": 1.7678873538970947, + "learning_rate": 0.00017100208697057512, + "loss": 0.7718, + "step": 15635 + }, + { + "epoch": 1.0850045104434112, + "grad_norm": 2.2285256385803223, + "learning_rate": 0.00017099813033830393, + "loss": 0.6552, + "step": 15636 + }, + { + "epoch": 1.085073901880508, + "grad_norm": 1.9047340154647827, + "learning_rate": 0.00017099417348189985, + "loss": 0.7842, + "step": 15637 + }, + { + "epoch": 1.0851432933176046, + "grad_norm": 1.759472370147705, + "learning_rate": 0.00017099021640137543, + "loss": 0.5954, + "step": 15638 + }, + { + "epoch": 1.0852126847547012, + "grad_norm": 1.9341398477554321, + "learning_rate": 0.0001709862590967431, + "loss": 0.7319, + "step": 15639 + }, + { + "epoch": 1.0852820761917978, + "grad_norm": 1.6666240692138672, + "learning_rate": 0.00017098230156801538, + "loss": 0.6523, + "step": 15640 + }, + { + "epoch": 1.0853514676288947, + "grad_norm": 1.5672019720077515, + "learning_rate": 0.0001709783438152048, + "loss": 0.5335, + "step": 15641 + }, + { + "epoch": 1.0854208590659913, + "grad_norm": 1.8110727071762085, + "learning_rate": 0.00017097438583832384, + "loss": 0.6967, + "step": 15642 + }, + { + "epoch": 1.085490250503088, + "grad_norm": 1.8169991970062256, + "learning_rate": 0.00017097042763738495, + "loss": 0.6934, + "step": 15643 + }, + { + "epoch": 1.0855596419401845, + "grad_norm": 1.5616029500961304, + "learning_rate": 0.00017096646921240064, + "loss": 0.5463, + "step": 15644 + }, + { + "epoch": 1.0856290333772813, + "grad_norm": 1.7780084609985352, + "learning_rate": 0.00017096251056338344, + "loss": 0.638, + "step": 15645 + }, + { + "epoch": 1.085698424814378, + "grad_norm": 1.7560110092163086, + "learning_rate": 0.00017095855169034583, + "loss": 0.7013, + "step": 15646 + }, + { + "epoch": 1.0857678162514746, + "grad_norm": 1.4216477870941162, + "learning_rate": 0.0001709545925933003, + "loss": 0.5776, + "step": 15647 + }, + { + "epoch": 1.0858372076885712, + "grad_norm": 2.968212842941284, + "learning_rate": 0.00017095063327225932, + "loss": 0.7528, + "step": 15648 + }, + { + "epoch": 1.0859065991256678, + "grad_norm": 2.972468614578247, + "learning_rate": 0.00017094667372723549, + "loss": 0.6025, + "step": 15649 + }, + { + "epoch": 1.0859759905627646, + "grad_norm": 1.9633978605270386, + "learning_rate": 0.0001709427139582412, + "loss": 0.707, + "step": 15650 + }, + { + "epoch": 1.0860453819998612, + "grad_norm": 1.8495638370513916, + "learning_rate": 0.000170938753965289, + "loss": 0.5393, + "step": 15651 + }, + { + "epoch": 1.0861147734369578, + "grad_norm": 2.1200578212738037, + "learning_rate": 0.0001709347937483914, + "loss": 0.8842, + "step": 15652 + }, + { + "epoch": 1.0861841648740544, + "grad_norm": 2.162754535675049, + "learning_rate": 0.00017093083330756086, + "loss": 0.7812, + "step": 15653 + }, + { + "epoch": 1.0862535563111513, + "grad_norm": 2.0598912239074707, + "learning_rate": 0.0001709268726428099, + "loss": 0.9744, + "step": 15654 + }, + { + "epoch": 1.086322947748248, + "grad_norm": 1.9245790243148804, + "learning_rate": 0.00017092291175415105, + "loss": 0.5936, + "step": 15655 + }, + { + "epoch": 1.0863923391853445, + "grad_norm": 1.6542319059371948, + "learning_rate": 0.00017091895064159682, + "loss": 0.4911, + "step": 15656 + }, + { + "epoch": 1.0864617306224411, + "grad_norm": 2.703829526901245, + "learning_rate": 0.00017091498930515967, + "loss": 0.9141, + "step": 15657 + }, + { + "epoch": 1.086531122059538, + "grad_norm": 1.9061646461486816, + "learning_rate": 0.00017091102774485211, + "loss": 0.668, + "step": 15658 + }, + { + "epoch": 1.0866005134966346, + "grad_norm": 2.6824281215667725, + "learning_rate": 0.00017090706596068666, + "loss": 0.5705, + "step": 15659 + }, + { + "epoch": 1.0866699049337312, + "grad_norm": 1.6602128744125366, + "learning_rate": 0.00017090310395267587, + "loss": 0.4895, + "step": 15660 + }, + { + "epoch": 1.0867392963708278, + "grad_norm": 2.222705364227295, + "learning_rate": 0.0001708991417208322, + "loss": 0.6579, + "step": 15661 + }, + { + "epoch": 1.0868086878079244, + "grad_norm": 1.9457433223724365, + "learning_rate": 0.00017089517926516815, + "loss": 0.7458, + "step": 15662 + }, + { + "epoch": 1.0868780792450212, + "grad_norm": 1.771209716796875, + "learning_rate": 0.0001708912165856962, + "loss": 0.7893, + "step": 15663 + }, + { + "epoch": 1.0869474706821178, + "grad_norm": 2.1755189895629883, + "learning_rate": 0.00017088725368242898, + "loss": 1.1688, + "step": 15664 + }, + { + "epoch": 1.0870168621192144, + "grad_norm": 2.8715007305145264, + "learning_rate": 0.0001708832905553789, + "loss": 0.6707, + "step": 15665 + }, + { + "epoch": 1.087086253556311, + "grad_norm": 1.857366681098938, + "learning_rate": 0.00017087932720455848, + "loss": 0.7715, + "step": 15666 + }, + { + "epoch": 1.0871556449934079, + "grad_norm": 1.8281394243240356, + "learning_rate": 0.00017087536362998023, + "loss": 0.6621, + "step": 15667 + }, + { + "epoch": 1.0872250364305045, + "grad_norm": 1.5371571779251099, + "learning_rate": 0.00017087139983165673, + "loss": 0.639, + "step": 15668 + }, + { + "epoch": 1.087294427867601, + "grad_norm": 2.3341450691223145, + "learning_rate": 0.00017086743580960043, + "loss": 0.6431, + "step": 15669 + }, + { + "epoch": 1.0873638193046977, + "grad_norm": 1.5144171714782715, + "learning_rate": 0.00017086347156382384, + "loss": 0.5626, + "step": 15670 + }, + { + "epoch": 1.0874332107417946, + "grad_norm": 2.218498706817627, + "learning_rate": 0.00017085950709433949, + "loss": 0.6151, + "step": 15671 + }, + { + "epoch": 1.0875026021788912, + "grad_norm": 1.6767724752426147, + "learning_rate": 0.0001708555424011599, + "loss": 0.9229, + "step": 15672 + }, + { + "epoch": 1.0875719936159878, + "grad_norm": 1.9643445014953613, + "learning_rate": 0.0001708515774842976, + "loss": 0.5524, + "step": 15673 + }, + { + "epoch": 1.0876413850530844, + "grad_norm": 1.9873698949813843, + "learning_rate": 0.00017084761234376506, + "loss": 0.7541, + "step": 15674 + }, + { + "epoch": 1.087710776490181, + "grad_norm": 2.324004650115967, + "learning_rate": 0.00017084364697957487, + "loss": 0.8365, + "step": 15675 + }, + { + "epoch": 1.0877801679272778, + "grad_norm": 2.0715606212615967, + "learning_rate": 0.00017083968139173947, + "loss": 0.8993, + "step": 15676 + }, + { + "epoch": 1.0878495593643744, + "grad_norm": 1.983272910118103, + "learning_rate": 0.00017083571558027143, + "loss": 0.7343, + "step": 15677 + }, + { + "epoch": 1.087918950801471, + "grad_norm": 1.7502323389053345, + "learning_rate": 0.00017083174954518325, + "loss": 0.6585, + "step": 15678 + }, + { + "epoch": 1.0879883422385677, + "grad_norm": 1.7591220140457153, + "learning_rate": 0.00017082778328648748, + "loss": 0.8115, + "step": 15679 + }, + { + "epoch": 1.0880577336756645, + "grad_norm": 2.3273563385009766, + "learning_rate": 0.0001708238168041966, + "loss": 0.6915, + "step": 15680 + }, + { + "epoch": 1.088127125112761, + "grad_norm": 1.7889372110366821, + "learning_rate": 0.00017081985009832312, + "loss": 0.629, + "step": 15681 + }, + { + "epoch": 1.0881965165498577, + "grad_norm": 1.8689208030700684, + "learning_rate": 0.00017081588316887963, + "loss": 0.5077, + "step": 15682 + }, + { + "epoch": 1.0882659079869543, + "grad_norm": 2.052339553833008, + "learning_rate": 0.0001708119160158786, + "loss": 0.5827, + "step": 15683 + }, + { + "epoch": 1.0883352994240512, + "grad_norm": 2.267693042755127, + "learning_rate": 0.00017080794863933254, + "loss": 0.8223, + "step": 15684 + }, + { + "epoch": 1.0884046908611478, + "grad_norm": 1.816564917564392, + "learning_rate": 0.000170803981039254, + "loss": 0.6703, + "step": 15685 + }, + { + "epoch": 1.0884740822982444, + "grad_norm": 1.9884568452835083, + "learning_rate": 0.00017080001321565552, + "loss": 0.4808, + "step": 15686 + }, + { + "epoch": 1.088543473735341, + "grad_norm": 1.8400940895080566, + "learning_rate": 0.00017079604516854967, + "loss": 0.7041, + "step": 15687 + }, + { + "epoch": 1.0886128651724376, + "grad_norm": 2.0209434032440186, + "learning_rate": 0.00017079207689794882, + "loss": 0.5651, + "step": 15688 + }, + { + "epoch": 1.0886822566095344, + "grad_norm": 2.5667531490325928, + "learning_rate": 0.00017078810840386567, + "loss": 0.8812, + "step": 15689 + }, + { + "epoch": 1.088751648046631, + "grad_norm": 2.191739320755005, + "learning_rate": 0.00017078413968631267, + "loss": 0.7544, + "step": 15690 + }, + { + "epoch": 1.0888210394837277, + "grad_norm": 1.7969692945480347, + "learning_rate": 0.00017078017074530232, + "loss": 0.56, + "step": 15691 + }, + { + "epoch": 1.0888904309208243, + "grad_norm": 1.6156402826309204, + "learning_rate": 0.00017077620158084718, + "loss": 0.6208, + "step": 15692 + }, + { + "epoch": 1.088959822357921, + "grad_norm": 2.1060538291931152, + "learning_rate": 0.00017077223219295982, + "loss": 0.7238, + "step": 15693 + }, + { + "epoch": 1.0890292137950177, + "grad_norm": 2.40287709236145, + "learning_rate": 0.0001707682625816527, + "loss": 0.794, + "step": 15694 + }, + { + "epoch": 1.0890986052321143, + "grad_norm": 2.083427906036377, + "learning_rate": 0.0001707642927469384, + "loss": 0.6426, + "step": 15695 + }, + { + "epoch": 1.089167996669211, + "grad_norm": 2.4732484817504883, + "learning_rate": 0.00017076032268882946, + "loss": 0.5591, + "step": 15696 + }, + { + "epoch": 1.0892373881063078, + "grad_norm": 2.1254515647888184, + "learning_rate": 0.00017075635240733833, + "loss": 0.6656, + "step": 15697 + }, + { + "epoch": 1.0893067795434044, + "grad_norm": 2.199615955352783, + "learning_rate": 0.00017075238190247767, + "loss": 0.7083, + "step": 15698 + }, + { + "epoch": 1.089376170980501, + "grad_norm": 1.4606620073318481, + "learning_rate": 0.00017074841117425992, + "loss": 0.6614, + "step": 15699 + }, + { + "epoch": 1.0894455624175976, + "grad_norm": 2.413506031036377, + "learning_rate": 0.00017074444022269764, + "loss": 0.8344, + "step": 15700 + }, + { + "epoch": 1.0895149538546944, + "grad_norm": 1.9539566040039062, + "learning_rate": 0.00017074046904780337, + "loss": 0.6934, + "step": 15701 + }, + { + "epoch": 1.089584345291791, + "grad_norm": 1.6841788291931152, + "learning_rate": 0.00017073649764958967, + "loss": 0.6665, + "step": 15702 + }, + { + "epoch": 1.0896537367288877, + "grad_norm": 1.5080832242965698, + "learning_rate": 0.00017073252602806905, + "loss": 0.7429, + "step": 15703 + }, + { + "epoch": 1.0897231281659843, + "grad_norm": 1.8298466205596924, + "learning_rate": 0.00017072855418325404, + "loss": 0.7243, + "step": 15704 + }, + { + "epoch": 1.0897925196030809, + "grad_norm": 1.5109467506408691, + "learning_rate": 0.00017072458211515718, + "loss": 0.5324, + "step": 15705 + }, + { + "epoch": 1.0898619110401777, + "grad_norm": 2.2384088039398193, + "learning_rate": 0.00017072060982379104, + "loss": 0.843, + "step": 15706 + }, + { + "epoch": 1.0899313024772743, + "grad_norm": 2.084916830062866, + "learning_rate": 0.00017071663730916812, + "loss": 0.5425, + "step": 15707 + }, + { + "epoch": 1.090000693914371, + "grad_norm": 1.9592995643615723, + "learning_rate": 0.00017071266457130099, + "loss": 0.664, + "step": 15708 + }, + { + "epoch": 1.0900700853514675, + "grad_norm": 2.330875873565674, + "learning_rate": 0.00017070869161020218, + "loss": 0.7349, + "step": 15709 + }, + { + "epoch": 1.0901394767885644, + "grad_norm": 2.0983240604400635, + "learning_rate": 0.00017070471842588422, + "loss": 0.5904, + "step": 15710 + }, + { + "epoch": 1.090208868225661, + "grad_norm": 2.381634473800659, + "learning_rate": 0.0001707007450183597, + "loss": 0.834, + "step": 15711 + }, + { + "epoch": 1.0902782596627576, + "grad_norm": 2.107682466506958, + "learning_rate": 0.00017069677138764112, + "loss": 0.5422, + "step": 15712 + }, + { + "epoch": 1.0903476510998542, + "grad_norm": 1.5965123176574707, + "learning_rate": 0.00017069279753374105, + "loss": 0.7257, + "step": 15713 + }, + { + "epoch": 1.090417042536951, + "grad_norm": 2.7858622074127197, + "learning_rate": 0.000170688823456672, + "loss": 0.8596, + "step": 15714 + }, + { + "epoch": 1.0904864339740477, + "grad_norm": 1.3824998140335083, + "learning_rate": 0.00017068484915644651, + "loss": 0.6762, + "step": 15715 + }, + { + "epoch": 1.0905558254111443, + "grad_norm": 2.020829200744629, + "learning_rate": 0.00017068087463307718, + "loss": 0.946, + "step": 15716 + }, + { + "epoch": 1.0906252168482409, + "grad_norm": 1.4463661909103394, + "learning_rate": 0.0001706768998865765, + "loss": 0.5193, + "step": 15717 + }, + { + "epoch": 1.0906946082853375, + "grad_norm": 1.4118455648422241, + "learning_rate": 0.0001706729249169571, + "loss": 0.7233, + "step": 15718 + }, + { + "epoch": 1.0907639997224343, + "grad_norm": 2.103783369064331, + "learning_rate": 0.00017066894972423145, + "loss": 0.6152, + "step": 15719 + }, + { + "epoch": 1.090833391159531, + "grad_norm": 1.9166340827941895, + "learning_rate": 0.00017066497430841214, + "loss": 0.7375, + "step": 15720 + }, + { + "epoch": 1.0909027825966275, + "grad_norm": 1.6765848398208618, + "learning_rate": 0.0001706609986695117, + "loss": 0.6239, + "step": 15721 + }, + { + "epoch": 1.0909721740337242, + "grad_norm": 2.062817335128784, + "learning_rate": 0.00017065702280754267, + "loss": 0.7001, + "step": 15722 + }, + { + "epoch": 1.091041565470821, + "grad_norm": 1.586186408996582, + "learning_rate": 0.0001706530467225176, + "loss": 0.6511, + "step": 15723 + }, + { + "epoch": 1.0911109569079176, + "grad_norm": 1.752914309501648, + "learning_rate": 0.00017064907041444908, + "loss": 0.7285, + "step": 15724 + }, + { + "epoch": 1.0911803483450142, + "grad_norm": 1.5150768756866455, + "learning_rate": 0.00017064509388334967, + "loss": 0.5837, + "step": 15725 + }, + { + "epoch": 1.0912497397821108, + "grad_norm": 1.4536694288253784, + "learning_rate": 0.00017064111712923185, + "loss": 0.5302, + "step": 15726 + }, + { + "epoch": 1.0913191312192076, + "grad_norm": 1.5270261764526367, + "learning_rate": 0.00017063714015210822, + "loss": 0.5877, + "step": 15727 + }, + { + "epoch": 1.0913885226563043, + "grad_norm": 1.996640682220459, + "learning_rate": 0.00017063316295199137, + "loss": 0.699, + "step": 15728 + }, + { + "epoch": 1.0914579140934009, + "grad_norm": 2.0566391944885254, + "learning_rate": 0.0001706291855288938, + "loss": 0.7906, + "step": 15729 + }, + { + "epoch": 1.0915273055304975, + "grad_norm": 1.7486438751220703, + "learning_rate": 0.0001706252078828281, + "loss": 0.7706, + "step": 15730 + }, + { + "epoch": 1.0915966969675943, + "grad_norm": 2.1264283657073975, + "learning_rate": 0.00017062123001380678, + "loss": 0.7226, + "step": 15731 + }, + { + "epoch": 1.091666088404691, + "grad_norm": 2.112701177597046, + "learning_rate": 0.00017061725192184246, + "loss": 0.7796, + "step": 15732 + }, + { + "epoch": 1.0917354798417875, + "grad_norm": 2.4209108352661133, + "learning_rate": 0.00017061327360694763, + "loss": 0.6023, + "step": 15733 + }, + { + "epoch": 1.0918048712788841, + "grad_norm": 2.0004332065582275, + "learning_rate": 0.00017060929506913493, + "loss": 0.523, + "step": 15734 + }, + { + "epoch": 1.0918742627159808, + "grad_norm": 1.697336196899414, + "learning_rate": 0.00017060531630841687, + "loss": 0.6658, + "step": 15735 + }, + { + "epoch": 1.0919436541530776, + "grad_norm": 1.6757813692092896, + "learning_rate": 0.00017060133732480602, + "loss": 0.9883, + "step": 15736 + }, + { + "epoch": 1.0920130455901742, + "grad_norm": 1.3936185836791992, + "learning_rate": 0.0001705973581183149, + "loss": 0.4889, + "step": 15737 + }, + { + "epoch": 1.0920824370272708, + "grad_norm": 1.8806426525115967, + "learning_rate": 0.00017059337868895614, + "loss": 0.6863, + "step": 15738 + }, + { + "epoch": 1.0921518284643674, + "grad_norm": 2.250824451446533, + "learning_rate": 0.0001705893990367423, + "loss": 0.5821, + "step": 15739 + }, + { + "epoch": 1.0922212199014643, + "grad_norm": 2.204575300216675, + "learning_rate": 0.0001705854191616859, + "loss": 0.6747, + "step": 15740 + }, + { + "epoch": 1.0922906113385609, + "grad_norm": 1.9165189266204834, + "learning_rate": 0.0001705814390637995, + "loss": 0.8748, + "step": 15741 + }, + { + "epoch": 1.0923600027756575, + "grad_norm": 1.4908409118652344, + "learning_rate": 0.00017057745874309572, + "loss": 0.6237, + "step": 15742 + }, + { + "epoch": 1.092429394212754, + "grad_norm": 1.7158515453338623, + "learning_rate": 0.00017057347819958703, + "loss": 0.5694, + "step": 15743 + }, + { + "epoch": 1.092498785649851, + "grad_norm": 2.2051379680633545, + "learning_rate": 0.00017056949743328613, + "loss": 0.6228, + "step": 15744 + }, + { + "epoch": 1.0925681770869475, + "grad_norm": 1.7346218824386597, + "learning_rate": 0.00017056551644420548, + "loss": 0.5989, + "step": 15745 + }, + { + "epoch": 1.0926375685240441, + "grad_norm": 1.824520230293274, + "learning_rate": 0.00017056153523235768, + "loss": 0.6711, + "step": 15746 + }, + { + "epoch": 1.0927069599611408, + "grad_norm": 1.9684126377105713, + "learning_rate": 0.0001705575537977553, + "loss": 0.7126, + "step": 15747 + }, + { + "epoch": 1.0927763513982374, + "grad_norm": 1.6599452495574951, + "learning_rate": 0.0001705535721404109, + "loss": 0.4627, + "step": 15748 + }, + { + "epoch": 1.0928457428353342, + "grad_norm": 1.6897673606872559, + "learning_rate": 0.0001705495902603371, + "loss": 0.5894, + "step": 15749 + }, + { + "epoch": 1.0929151342724308, + "grad_norm": 2.2901179790496826, + "learning_rate": 0.00017054560815754638, + "loss": 0.8529, + "step": 15750 + }, + { + "epoch": 1.0929845257095274, + "grad_norm": 2.011085271835327, + "learning_rate": 0.0001705416258320514, + "loss": 0.5353, + "step": 15751 + }, + { + "epoch": 1.093053917146624, + "grad_norm": 2.1828701496124268, + "learning_rate": 0.00017053764328386466, + "loss": 0.7688, + "step": 15752 + }, + { + "epoch": 1.0931233085837209, + "grad_norm": 1.8091472387313843, + "learning_rate": 0.0001705336605129988, + "loss": 0.5653, + "step": 15753 + }, + { + "epoch": 1.0931927000208175, + "grad_norm": 1.9714080095291138, + "learning_rate": 0.0001705296775194663, + "loss": 0.6995, + "step": 15754 + }, + { + "epoch": 1.093262091457914, + "grad_norm": 1.600699782371521, + "learning_rate": 0.00017052569430327984, + "loss": 0.5682, + "step": 15755 + }, + { + "epoch": 1.0933314828950107, + "grad_norm": 1.5801821947097778, + "learning_rate": 0.00017052171086445191, + "loss": 0.6517, + "step": 15756 + }, + { + "epoch": 1.0934008743321075, + "grad_norm": 1.585212230682373, + "learning_rate": 0.00017051772720299513, + "loss": 0.7417, + "step": 15757 + }, + { + "epoch": 1.0934702657692041, + "grad_norm": 2.195077657699585, + "learning_rate": 0.0001705137433189221, + "loss": 0.7514, + "step": 15758 + }, + { + "epoch": 1.0935396572063008, + "grad_norm": 1.813480257987976, + "learning_rate": 0.00017050975921224535, + "loss": 0.6662, + "step": 15759 + }, + { + "epoch": 1.0936090486433974, + "grad_norm": 2.403369665145874, + "learning_rate": 0.00017050577488297747, + "loss": 0.823, + "step": 15760 + }, + { + "epoch": 1.093678440080494, + "grad_norm": 1.567542314529419, + "learning_rate": 0.000170501790331131, + "loss": 0.6041, + "step": 15761 + }, + { + "epoch": 1.0937478315175908, + "grad_norm": 3.399977445602417, + "learning_rate": 0.00017049780555671858, + "loss": 0.6824, + "step": 15762 + }, + { + "epoch": 1.0938172229546874, + "grad_norm": 2.3761167526245117, + "learning_rate": 0.0001704938205597528, + "loss": 0.9087, + "step": 15763 + }, + { + "epoch": 1.093886614391784, + "grad_norm": 2.719980239868164, + "learning_rate": 0.00017048983534024615, + "loss": 0.6342, + "step": 15764 + }, + { + "epoch": 1.0939560058288806, + "grad_norm": 1.48020339012146, + "learning_rate": 0.00017048584989821133, + "loss": 0.6389, + "step": 15765 + }, + { + "epoch": 1.0940253972659775, + "grad_norm": 1.7484761476516724, + "learning_rate": 0.00017048186423366084, + "loss": 0.4633, + "step": 15766 + }, + { + "epoch": 1.094094788703074, + "grad_norm": 2.1688146591186523, + "learning_rate": 0.00017047787834660727, + "loss": 0.7165, + "step": 15767 + }, + { + "epoch": 1.0941641801401707, + "grad_norm": 1.5845657587051392, + "learning_rate": 0.0001704738922370632, + "loss": 0.5554, + "step": 15768 + }, + { + "epoch": 1.0942335715772673, + "grad_norm": 2.025601387023926, + "learning_rate": 0.00017046990590504127, + "loss": 0.7263, + "step": 15769 + }, + { + "epoch": 1.0943029630143641, + "grad_norm": 1.4420527219772339, + "learning_rate": 0.000170465919350554, + "loss": 0.6293, + "step": 15770 + }, + { + "epoch": 1.0943723544514607, + "grad_norm": 1.5837445259094238, + "learning_rate": 0.000170461932573614, + "loss": 0.6032, + "step": 15771 + }, + { + "epoch": 1.0944417458885574, + "grad_norm": 2.21014404296875, + "learning_rate": 0.00017045794557423382, + "loss": 0.7532, + "step": 15772 + }, + { + "epoch": 1.094511137325654, + "grad_norm": 2.5563876628875732, + "learning_rate": 0.00017045395835242613, + "loss": 0.6107, + "step": 15773 + }, + { + "epoch": 1.0945805287627506, + "grad_norm": 2.383181571960449, + "learning_rate": 0.00017044997090820342, + "loss": 0.7735, + "step": 15774 + }, + { + "epoch": 1.0946499201998474, + "grad_norm": 1.8861585855484009, + "learning_rate": 0.00017044598324157838, + "loss": 0.6555, + "step": 15775 + }, + { + "epoch": 1.094719311636944, + "grad_norm": 1.9699093103408813, + "learning_rate": 0.00017044199535256352, + "loss": 0.6777, + "step": 15776 + }, + { + "epoch": 1.0947887030740406, + "grad_norm": 1.871066689491272, + "learning_rate": 0.00017043800724117145, + "loss": 0.8541, + "step": 15777 + }, + { + "epoch": 1.0948580945111372, + "grad_norm": 1.648343801498413, + "learning_rate": 0.00017043401890741477, + "loss": 0.7757, + "step": 15778 + }, + { + "epoch": 1.094927485948234, + "grad_norm": 2.1257426738739014, + "learning_rate": 0.00017043003035130606, + "loss": 0.6059, + "step": 15779 + }, + { + "epoch": 1.0949968773853307, + "grad_norm": 2.801316022872925, + "learning_rate": 0.00017042604157285792, + "loss": 0.8054, + "step": 15780 + }, + { + "epoch": 1.0950662688224273, + "grad_norm": 1.682260513305664, + "learning_rate": 0.00017042205257208293, + "loss": 0.6297, + "step": 15781 + }, + { + "epoch": 1.095135660259524, + "grad_norm": 1.8848888874053955, + "learning_rate": 0.00017041806334899367, + "loss": 0.7061, + "step": 15782 + }, + { + "epoch": 1.0952050516966207, + "grad_norm": 1.9519226551055908, + "learning_rate": 0.0001704140739036028, + "loss": 0.7428, + "step": 15783 + }, + { + "epoch": 1.0952744431337174, + "grad_norm": 1.7216697931289673, + "learning_rate": 0.00017041008423592286, + "loss": 0.416, + "step": 15784 + }, + { + "epoch": 1.095343834570814, + "grad_norm": 1.8443443775177002, + "learning_rate": 0.00017040609434596643, + "loss": 0.6807, + "step": 15785 + }, + { + "epoch": 1.0954132260079106, + "grad_norm": 2.5626962184906006, + "learning_rate": 0.00017040210423374612, + "loss": 0.8564, + "step": 15786 + }, + { + "epoch": 1.0954826174450072, + "grad_norm": 2.830228805541992, + "learning_rate": 0.00017039811389927455, + "loss": 0.6614, + "step": 15787 + }, + { + "epoch": 1.095552008882104, + "grad_norm": 1.7060056924819946, + "learning_rate": 0.00017039412334256433, + "loss": 0.5093, + "step": 15788 + }, + { + "epoch": 1.0956214003192006, + "grad_norm": 1.7830358743667603, + "learning_rate": 0.000170390132563628, + "loss": 0.7498, + "step": 15789 + }, + { + "epoch": 1.0956907917562972, + "grad_norm": 1.7454545497894287, + "learning_rate": 0.0001703861415624782, + "loss": 0.5643, + "step": 15790 + }, + { + "epoch": 1.0957601831933939, + "grad_norm": 2.2465596199035645, + "learning_rate": 0.0001703821503391275, + "loss": 0.7465, + "step": 15791 + }, + { + "epoch": 1.0958295746304907, + "grad_norm": 1.6341227293014526, + "learning_rate": 0.00017037815889358853, + "loss": 0.5874, + "step": 15792 + }, + { + "epoch": 1.0958989660675873, + "grad_norm": 1.769797682762146, + "learning_rate": 0.00017037416722587387, + "loss": 0.6307, + "step": 15793 + }, + { + "epoch": 1.095968357504684, + "grad_norm": 2.444258689880371, + "learning_rate": 0.00017037017533599614, + "loss": 0.7789, + "step": 15794 + }, + { + "epoch": 1.0960377489417805, + "grad_norm": 1.681579351425171, + "learning_rate": 0.00017036618322396793, + "loss": 0.6361, + "step": 15795 + }, + { + "epoch": 1.0961071403788774, + "grad_norm": 1.5569021701812744, + "learning_rate": 0.00017036219088980184, + "loss": 0.513, + "step": 15796 + }, + { + "epoch": 1.096176531815974, + "grad_norm": 1.670816421508789, + "learning_rate": 0.0001703581983335105, + "loss": 0.7774, + "step": 15797 + }, + { + "epoch": 1.0962459232530706, + "grad_norm": 2.0518548488616943, + "learning_rate": 0.00017035420555510644, + "loss": 0.8142, + "step": 15798 + }, + { + "epoch": 1.0963153146901672, + "grad_norm": 2.4152562618255615, + "learning_rate": 0.00017035021255460233, + "loss": 0.9614, + "step": 15799 + }, + { + "epoch": 1.0963847061272638, + "grad_norm": 1.645691156387329, + "learning_rate": 0.0001703462193320108, + "loss": 0.7291, + "step": 15800 + }, + { + "epoch": 1.0964540975643606, + "grad_norm": 1.3658910989761353, + "learning_rate": 0.0001703422258873444, + "loss": 0.5913, + "step": 15801 + }, + { + "epoch": 1.0965234890014572, + "grad_norm": 1.9568337202072144, + "learning_rate": 0.00017033823222061572, + "loss": 0.7149, + "step": 15802 + }, + { + "epoch": 1.0965928804385539, + "grad_norm": 2.185598134994507, + "learning_rate": 0.0001703342383318374, + "loss": 0.7567, + "step": 15803 + }, + { + "epoch": 1.0966622718756505, + "grad_norm": 1.1896421909332275, + "learning_rate": 0.0001703302442210221, + "loss": 0.4809, + "step": 15804 + }, + { + "epoch": 1.0967316633127473, + "grad_norm": 2.0794622898101807, + "learning_rate": 0.00017032624988818235, + "loss": 0.6019, + "step": 15805 + }, + { + "epoch": 1.096801054749844, + "grad_norm": 2.377244472503662, + "learning_rate": 0.0001703222553333308, + "loss": 0.6164, + "step": 15806 + }, + { + "epoch": 1.0968704461869405, + "grad_norm": 2.516235589981079, + "learning_rate": 0.00017031826055648007, + "loss": 0.7472, + "step": 15807 + }, + { + "epoch": 1.0969398376240371, + "grad_norm": 1.692493200302124, + "learning_rate": 0.00017031426555764268, + "loss": 0.6642, + "step": 15808 + }, + { + "epoch": 1.097009229061134, + "grad_norm": 2.601475954055786, + "learning_rate": 0.00017031027033683139, + "loss": 0.84, + "step": 15809 + }, + { + "epoch": 1.0970786204982306, + "grad_norm": 1.762515664100647, + "learning_rate": 0.0001703062748940587, + "loss": 0.8027, + "step": 15810 + }, + { + "epoch": 1.0971480119353272, + "grad_norm": 2.838146924972534, + "learning_rate": 0.00017030227922933727, + "loss": 0.7009, + "step": 15811 + }, + { + "epoch": 1.0972174033724238, + "grad_norm": 2.2723171710968018, + "learning_rate": 0.00017029828334267966, + "loss": 0.7961, + "step": 15812 + }, + { + "epoch": 1.0972867948095204, + "grad_norm": 1.8363009691238403, + "learning_rate": 0.00017029428723409855, + "loss": 0.7154, + "step": 15813 + }, + { + "epoch": 1.0973561862466172, + "grad_norm": 1.5781543254852295, + "learning_rate": 0.00017029029090360654, + "loss": 0.6013, + "step": 15814 + }, + { + "epoch": 1.0974255776837138, + "grad_norm": 1.8894386291503906, + "learning_rate": 0.00017028629435121621, + "loss": 0.6189, + "step": 15815 + }, + { + "epoch": 1.0974949691208105, + "grad_norm": 1.662940502166748, + "learning_rate": 0.00017028229757694023, + "loss": 0.6674, + "step": 15816 + }, + { + "epoch": 1.097564360557907, + "grad_norm": 1.7183631658554077, + "learning_rate": 0.00017027830058079119, + "loss": 0.7392, + "step": 15817 + }, + { + "epoch": 1.097633751995004, + "grad_norm": 1.5572311878204346, + "learning_rate": 0.0001702743033627817, + "loss": 0.6228, + "step": 15818 + }, + { + "epoch": 1.0977031434321005, + "grad_norm": 1.8555800914764404, + "learning_rate": 0.0001702703059229244, + "loss": 0.7066, + "step": 15819 + }, + { + "epoch": 1.0977725348691971, + "grad_norm": 1.8384758234024048, + "learning_rate": 0.00017026630826123186, + "loss": 0.8146, + "step": 15820 + }, + { + "epoch": 1.0978419263062937, + "grad_norm": 2.378939151763916, + "learning_rate": 0.0001702623103777168, + "loss": 0.7189, + "step": 15821 + }, + { + "epoch": 1.0979113177433906, + "grad_norm": 1.7341140508651733, + "learning_rate": 0.00017025831227239174, + "loss": 0.4722, + "step": 15822 + }, + { + "epoch": 1.0979807091804872, + "grad_norm": 1.633195161819458, + "learning_rate": 0.0001702543139452693, + "loss": 0.5928, + "step": 15823 + }, + { + "epoch": 1.0980501006175838, + "grad_norm": 2.3691811561584473, + "learning_rate": 0.0001702503153963622, + "loss": 0.6045, + "step": 15824 + }, + { + "epoch": 1.0981194920546804, + "grad_norm": 1.3742916584014893, + "learning_rate": 0.000170246316625683, + "loss": 0.6605, + "step": 15825 + }, + { + "epoch": 1.098188883491777, + "grad_norm": 2.1739437580108643, + "learning_rate": 0.0001702423176332443, + "loss": 0.7961, + "step": 15826 + }, + { + "epoch": 1.0982582749288738, + "grad_norm": 2.0735738277435303, + "learning_rate": 0.0001702383184190588, + "loss": 0.8967, + "step": 15827 + }, + { + "epoch": 1.0983276663659705, + "grad_norm": 1.6432785987854004, + "learning_rate": 0.00017023431898313904, + "loss": 0.5989, + "step": 15828 + }, + { + "epoch": 1.098397057803067, + "grad_norm": 1.6342164278030396, + "learning_rate": 0.00017023031932549769, + "loss": 0.5429, + "step": 15829 + }, + { + "epoch": 1.0984664492401637, + "grad_norm": 2.169858694076538, + "learning_rate": 0.00017022631944614734, + "loss": 0.9334, + "step": 15830 + }, + { + "epoch": 1.0985358406772605, + "grad_norm": 1.630049467086792, + "learning_rate": 0.0001702223193451007, + "loss": 0.6157, + "step": 15831 + }, + { + "epoch": 1.0986052321143571, + "grad_norm": 2.2999658584594727, + "learning_rate": 0.00017021831902237035, + "loss": 0.8129, + "step": 15832 + }, + { + "epoch": 1.0986746235514537, + "grad_norm": 2.195190191268921, + "learning_rate": 0.00017021431847796887, + "loss": 0.6252, + "step": 15833 + }, + { + "epoch": 1.0987440149885503, + "grad_norm": 1.8200589418411255, + "learning_rate": 0.00017021031771190893, + "loss": 0.6807, + "step": 15834 + }, + { + "epoch": 1.0988134064256472, + "grad_norm": 2.195955991744995, + "learning_rate": 0.0001702063167242032, + "loss": 0.5763, + "step": 15835 + }, + { + "epoch": 1.0988827978627438, + "grad_norm": 2.5280096530914307, + "learning_rate": 0.00017020231551486426, + "loss": 0.7352, + "step": 15836 + }, + { + "epoch": 1.0989521892998404, + "grad_norm": 1.734026551246643, + "learning_rate": 0.00017019831408390473, + "loss": 0.6869, + "step": 15837 + }, + { + "epoch": 1.099021580736937, + "grad_norm": 1.8328452110290527, + "learning_rate": 0.0001701943124313373, + "loss": 0.6387, + "step": 15838 + }, + { + "epoch": 1.0990909721740336, + "grad_norm": 1.9798846244812012, + "learning_rate": 0.0001701903105571746, + "loss": 0.6777, + "step": 15839 + }, + { + "epoch": 1.0991603636111305, + "grad_norm": 2.0885345935821533, + "learning_rate": 0.00017018630846142916, + "loss": 0.6124, + "step": 15840 + }, + { + "epoch": 1.099229755048227, + "grad_norm": 2.043958902359009, + "learning_rate": 0.00017018230614411373, + "loss": 0.7231, + "step": 15841 + }, + { + "epoch": 1.0992991464853237, + "grad_norm": 1.6946392059326172, + "learning_rate": 0.0001701783036052409, + "loss": 0.5694, + "step": 15842 + }, + { + "epoch": 1.0993685379224203, + "grad_norm": 2.4660966396331787, + "learning_rate": 0.0001701743008448233, + "loss": 0.6539, + "step": 15843 + }, + { + "epoch": 1.0994379293595171, + "grad_norm": 1.8148071765899658, + "learning_rate": 0.00017017029786287356, + "loss": 0.5457, + "step": 15844 + }, + { + "epoch": 1.0995073207966137, + "grad_norm": 1.5301415920257568, + "learning_rate": 0.00017016629465940435, + "loss": 0.6593, + "step": 15845 + }, + { + "epoch": 1.0995767122337103, + "grad_norm": 1.5478140115737915, + "learning_rate": 0.00017016229123442827, + "loss": 0.5864, + "step": 15846 + }, + { + "epoch": 1.099646103670807, + "grad_norm": 2.4058547019958496, + "learning_rate": 0.000170158287587958, + "loss": 0.7801, + "step": 15847 + }, + { + "epoch": 1.0997154951079038, + "grad_norm": 1.6872966289520264, + "learning_rate": 0.00017015428372000612, + "loss": 0.5988, + "step": 15848 + }, + { + "epoch": 1.0997848865450004, + "grad_norm": 2.112675666809082, + "learning_rate": 0.00017015027963058536, + "loss": 0.7857, + "step": 15849 + }, + { + "epoch": 1.099854277982097, + "grad_norm": 1.8069874048233032, + "learning_rate": 0.00017014627531970828, + "loss": 1.0395, + "step": 15850 + }, + { + "epoch": 1.0999236694191936, + "grad_norm": 2.044630289077759, + "learning_rate": 0.00017014227078738755, + "loss": 0.7801, + "step": 15851 + }, + { + "epoch": 1.0999930608562902, + "grad_norm": 2.4743590354919434, + "learning_rate": 0.00017013826603363578, + "loss": 0.7392, + "step": 15852 + }, + { + "epoch": 1.100062452293387, + "grad_norm": 2.010525941848755, + "learning_rate": 0.00017013426105846565, + "loss": 0.8495, + "step": 15853 + }, + { + "epoch": 1.1001318437304837, + "grad_norm": 2.6365270614624023, + "learning_rate": 0.00017013025586188983, + "loss": 0.673, + "step": 15854 + }, + { + "epoch": 1.1002012351675803, + "grad_norm": 1.9679880142211914, + "learning_rate": 0.0001701262504439209, + "loss": 0.7597, + "step": 15855 + }, + { + "epoch": 1.100270626604677, + "grad_norm": 2.2278249263763428, + "learning_rate": 0.00017012224480457156, + "loss": 0.5755, + "step": 15856 + }, + { + "epoch": 1.1003400180417737, + "grad_norm": 2.033515691757202, + "learning_rate": 0.00017011823894385442, + "loss": 0.6299, + "step": 15857 + }, + { + "epoch": 1.1004094094788703, + "grad_norm": 2.1430089473724365, + "learning_rate": 0.00017011423286178212, + "loss": 0.7731, + "step": 15858 + }, + { + "epoch": 1.100478800915967, + "grad_norm": 2.132955551147461, + "learning_rate": 0.00017011022655836735, + "loss": 0.8122, + "step": 15859 + }, + { + "epoch": 1.1005481923530636, + "grad_norm": 2.4233832359313965, + "learning_rate": 0.0001701062200336227, + "loss": 0.5669, + "step": 15860 + }, + { + "epoch": 1.1006175837901604, + "grad_norm": 1.4781982898712158, + "learning_rate": 0.00017010221328756084, + "loss": 0.4418, + "step": 15861 + }, + { + "epoch": 1.100686975227257, + "grad_norm": 1.4763565063476562, + "learning_rate": 0.00017009820632019445, + "loss": 0.6504, + "step": 15862 + }, + { + "epoch": 1.1007563666643536, + "grad_norm": 1.7515270709991455, + "learning_rate": 0.00017009419913153613, + "loss": 0.8169, + "step": 15863 + }, + { + "epoch": 1.1008257581014502, + "grad_norm": 1.039827823638916, + "learning_rate": 0.00017009019172159858, + "loss": 0.3219, + "step": 15864 + }, + { + "epoch": 1.1008951495385468, + "grad_norm": 2.2540316581726074, + "learning_rate": 0.00017008618409039442, + "loss": 0.7167, + "step": 15865 + }, + { + "epoch": 1.1009645409756437, + "grad_norm": 2.235978364944458, + "learning_rate": 0.00017008217623793633, + "loss": 0.8222, + "step": 15866 + }, + { + "epoch": 1.1010339324127403, + "grad_norm": 1.7157214879989624, + "learning_rate": 0.0001700781681642369, + "loss": 0.5629, + "step": 15867 + }, + { + "epoch": 1.101103323849837, + "grad_norm": 1.70364248752594, + "learning_rate": 0.00017007415986930883, + "loss": 0.6189, + "step": 15868 + }, + { + "epoch": 1.1011727152869335, + "grad_norm": 1.8021799325942993, + "learning_rate": 0.00017007015135316479, + "loss": 0.4531, + "step": 15869 + }, + { + "epoch": 1.1012421067240303, + "grad_norm": 1.9260609149932861, + "learning_rate": 0.0001700661426158174, + "loss": 0.6896, + "step": 15870 + }, + { + "epoch": 1.101311498161127, + "grad_norm": 2.346593141555786, + "learning_rate": 0.00017006213365727932, + "loss": 0.8188, + "step": 15871 + }, + { + "epoch": 1.1013808895982236, + "grad_norm": 2.5093421936035156, + "learning_rate": 0.00017005812447756322, + "loss": 0.4485, + "step": 15872 + }, + { + "epoch": 1.1014502810353202, + "grad_norm": 1.7037452459335327, + "learning_rate": 0.00017005411507668177, + "loss": 0.5601, + "step": 15873 + }, + { + "epoch": 1.101519672472417, + "grad_norm": 1.7434650659561157, + "learning_rate": 0.00017005010545464758, + "loss": 0.6269, + "step": 15874 + }, + { + "epoch": 1.1015890639095136, + "grad_norm": 2.1178693771362305, + "learning_rate": 0.00017004609561147333, + "loss": 0.5669, + "step": 15875 + }, + { + "epoch": 1.1016584553466102, + "grad_norm": 1.6926733255386353, + "learning_rate": 0.0001700420855471717, + "loss": 0.7714, + "step": 15876 + }, + { + "epoch": 1.1017278467837068, + "grad_norm": 1.91959547996521, + "learning_rate": 0.00017003807526175534, + "loss": 0.7249, + "step": 15877 + }, + { + "epoch": 1.1017972382208034, + "grad_norm": 2.018733263015747, + "learning_rate": 0.00017003406475523685, + "loss": 0.6728, + "step": 15878 + }, + { + "epoch": 1.1018666296579003, + "grad_norm": 1.7752360105514526, + "learning_rate": 0.000170030054027629, + "loss": 0.7599, + "step": 15879 + }, + { + "epoch": 1.1019360210949969, + "grad_norm": 1.688805103302002, + "learning_rate": 0.00017002604307894437, + "loss": 1.0125, + "step": 15880 + }, + { + "epoch": 1.1020054125320935, + "grad_norm": 1.7358946800231934, + "learning_rate": 0.00017002203190919564, + "loss": 0.6252, + "step": 15881 + }, + { + "epoch": 1.10207480396919, + "grad_norm": 1.5168328285217285, + "learning_rate": 0.0001700180205183955, + "loss": 0.5286, + "step": 15882 + }, + { + "epoch": 1.102144195406287, + "grad_norm": 1.6437124013900757, + "learning_rate": 0.00017001400890655658, + "loss": 0.6735, + "step": 15883 + }, + { + "epoch": 1.1022135868433836, + "grad_norm": 2.9393880367279053, + "learning_rate": 0.00017000999707369156, + "loss": 0.7777, + "step": 15884 + }, + { + "epoch": 1.1022829782804802, + "grad_norm": 4.237025260925293, + "learning_rate": 0.00017000598501981312, + "loss": 0.786, + "step": 15885 + }, + { + "epoch": 1.1023523697175768, + "grad_norm": 1.8984462022781372, + "learning_rate": 0.0001700019727449339, + "loss": 0.7739, + "step": 15886 + }, + { + "epoch": 1.1024217611546736, + "grad_norm": 1.825365424156189, + "learning_rate": 0.00016999796024906653, + "loss": 0.6226, + "step": 15887 + }, + { + "epoch": 1.1024911525917702, + "grad_norm": 1.66318941116333, + "learning_rate": 0.00016999394753222378, + "loss": 0.6456, + "step": 15888 + }, + { + "epoch": 1.1025605440288668, + "grad_norm": 2.072840929031372, + "learning_rate": 0.00016998993459441822, + "loss": 0.7804, + "step": 15889 + }, + { + "epoch": 1.1026299354659634, + "grad_norm": 1.6173545122146606, + "learning_rate": 0.0001699859214356626, + "loss": 0.5586, + "step": 15890 + }, + { + "epoch": 1.10269932690306, + "grad_norm": 1.7405511140823364, + "learning_rate": 0.00016998190805596949, + "loss": 0.6568, + "step": 15891 + }, + { + "epoch": 1.1027687183401569, + "grad_norm": 1.4879306554794312, + "learning_rate": 0.00016997789445535164, + "loss": 0.8038, + "step": 15892 + }, + { + "epoch": 1.1028381097772535, + "grad_norm": 1.848060965538025, + "learning_rate": 0.0001699738806338217, + "loss": 0.5972, + "step": 15893 + }, + { + "epoch": 1.10290750121435, + "grad_norm": 1.6533244848251343, + "learning_rate": 0.00016996986659139233, + "loss": 0.6549, + "step": 15894 + }, + { + "epoch": 1.1029768926514467, + "grad_norm": 2.2549221515655518, + "learning_rate": 0.00016996585232807623, + "loss": 0.7177, + "step": 15895 + }, + { + "epoch": 1.1030462840885435, + "grad_norm": 1.875877857208252, + "learning_rate": 0.000169961837843886, + "loss": 0.6239, + "step": 15896 + }, + { + "epoch": 1.1031156755256402, + "grad_norm": 1.9167298078536987, + "learning_rate": 0.00016995782313883445, + "loss": 0.7311, + "step": 15897 + }, + { + "epoch": 1.1031850669627368, + "grad_norm": 2.2349612712860107, + "learning_rate": 0.0001699538082129341, + "loss": 0.4914, + "step": 15898 + }, + { + "epoch": 1.1032544583998334, + "grad_norm": 1.7511094808578491, + "learning_rate": 0.0001699497930661977, + "loss": 0.566, + "step": 15899 + }, + { + "epoch": 1.1033238498369302, + "grad_norm": 1.9149113893508911, + "learning_rate": 0.00016994577769863792, + "loss": 0.6889, + "step": 15900 + }, + { + "epoch": 1.1033932412740268, + "grad_norm": 1.9354674816131592, + "learning_rate": 0.00016994176211026748, + "loss": 0.8451, + "step": 15901 + }, + { + "epoch": 1.1034626327111234, + "grad_norm": 1.744589924812317, + "learning_rate": 0.00016993774630109899, + "loss": 0.5669, + "step": 15902 + }, + { + "epoch": 1.10353202414822, + "grad_norm": 1.9759063720703125, + "learning_rate": 0.0001699337302711451, + "loss": 0.5786, + "step": 15903 + }, + { + "epoch": 1.1036014155853167, + "grad_norm": 2.1536550521850586, + "learning_rate": 0.0001699297140204186, + "loss": 0.4861, + "step": 15904 + }, + { + "epoch": 1.1036708070224135, + "grad_norm": 1.8073179721832275, + "learning_rate": 0.00016992569754893205, + "loss": 0.6169, + "step": 15905 + }, + { + "epoch": 1.10374019845951, + "grad_norm": 2.0851664543151855, + "learning_rate": 0.00016992168085669824, + "loss": 0.7506, + "step": 15906 + }, + { + "epoch": 1.1038095898966067, + "grad_norm": 1.902851939201355, + "learning_rate": 0.00016991766394372978, + "loss": 0.7615, + "step": 15907 + }, + { + "epoch": 1.1038789813337033, + "grad_norm": 1.820243000984192, + "learning_rate": 0.00016991364681003936, + "loss": 0.6624, + "step": 15908 + }, + { + "epoch": 1.1039483727708002, + "grad_norm": 2.275770425796509, + "learning_rate": 0.00016990962945563967, + "loss": 0.9272, + "step": 15909 + }, + { + "epoch": 1.1040177642078968, + "grad_norm": 2.31280517578125, + "learning_rate": 0.00016990561188054336, + "loss": 0.7939, + "step": 15910 + }, + { + "epoch": 1.1040871556449934, + "grad_norm": 2.0885369777679443, + "learning_rate": 0.0001699015940847632, + "loss": 0.789, + "step": 15911 + }, + { + "epoch": 1.10415654708209, + "grad_norm": 3.731292486190796, + "learning_rate": 0.00016989757606831178, + "loss": 0.723, + "step": 15912 + }, + { + "epoch": 1.1042259385191868, + "grad_norm": 1.4437545537948608, + "learning_rate": 0.00016989355783120184, + "loss": 0.6197, + "step": 15913 + }, + { + "epoch": 1.1042953299562834, + "grad_norm": 1.9639955759048462, + "learning_rate": 0.00016988953937344605, + "loss": 0.901, + "step": 15914 + }, + { + "epoch": 1.10436472139338, + "grad_norm": 1.4658000469207764, + "learning_rate": 0.00016988552069505708, + "loss": 0.5353, + "step": 15915 + }, + { + "epoch": 1.1044341128304767, + "grad_norm": 1.8101857900619507, + "learning_rate": 0.00016988150179604765, + "loss": 0.608, + "step": 15916 + }, + { + "epoch": 1.1045035042675733, + "grad_norm": 1.824747920036316, + "learning_rate": 0.0001698774826764304, + "loss": 0.7566, + "step": 15917 + }, + { + "epoch": 1.10457289570467, + "grad_norm": 2.052685260772705, + "learning_rate": 0.00016987346333621804, + "loss": 0.4271, + "step": 15918 + }, + { + "epoch": 1.1046422871417667, + "grad_norm": 2.0813798904418945, + "learning_rate": 0.0001698694437754233, + "loss": 0.7076, + "step": 15919 + }, + { + "epoch": 1.1047116785788633, + "grad_norm": 1.7133967876434326, + "learning_rate": 0.00016986542399405878, + "loss": 0.7363, + "step": 15920 + }, + { + "epoch": 1.10478107001596, + "grad_norm": 1.9736535549163818, + "learning_rate": 0.00016986140399213726, + "loss": 0.7877, + "step": 15921 + }, + { + "epoch": 1.1048504614530568, + "grad_norm": 1.552392840385437, + "learning_rate": 0.0001698573837696714, + "loss": 0.6502, + "step": 15922 + }, + { + "epoch": 1.1049198528901534, + "grad_norm": 2.1177079677581787, + "learning_rate": 0.00016985336332667384, + "loss": 0.5683, + "step": 15923 + }, + { + "epoch": 1.10498924432725, + "grad_norm": 1.5903167724609375, + "learning_rate": 0.00016984934266315736, + "loss": 0.4378, + "step": 15924 + }, + { + "epoch": 1.1050586357643466, + "grad_norm": 2.1025590896606445, + "learning_rate": 0.00016984532177913463, + "loss": 0.8053, + "step": 15925 + }, + { + "epoch": 1.1051280272014434, + "grad_norm": 2.0279858112335205, + "learning_rate": 0.00016984130067461826, + "loss": 0.8246, + "step": 15926 + }, + { + "epoch": 1.10519741863854, + "grad_norm": 1.9882335662841797, + "learning_rate": 0.000169837279349621, + "loss": 0.7951, + "step": 15927 + }, + { + "epoch": 1.1052668100756367, + "grad_norm": 1.5744296312332153, + "learning_rate": 0.0001698332578041556, + "loss": 0.5836, + "step": 15928 + }, + { + "epoch": 1.1053362015127333, + "grad_norm": 1.7235649824142456, + "learning_rate": 0.0001698292360382347, + "loss": 0.4826, + "step": 15929 + }, + { + "epoch": 1.1054055929498299, + "grad_norm": 2.1488611698150635, + "learning_rate": 0.00016982521405187103, + "loss": 0.9568, + "step": 15930 + }, + { + "epoch": 1.1054749843869267, + "grad_norm": 1.9533308744430542, + "learning_rate": 0.00016982119184507722, + "loss": 0.743, + "step": 15931 + }, + { + "epoch": 1.1055443758240233, + "grad_norm": 1.963475227355957, + "learning_rate": 0.000169817169417866, + "loss": 0.7515, + "step": 15932 + }, + { + "epoch": 1.10561376726112, + "grad_norm": 2.100658416748047, + "learning_rate": 0.00016981314677025008, + "loss": 0.7503, + "step": 15933 + }, + { + "epoch": 1.1056831586982165, + "grad_norm": 3.1570658683776855, + "learning_rate": 0.00016980912390224219, + "loss": 0.8209, + "step": 15934 + }, + { + "epoch": 1.1057525501353134, + "grad_norm": 1.5851118564605713, + "learning_rate": 0.00016980510081385496, + "loss": 0.9131, + "step": 15935 + }, + { + "epoch": 1.10582194157241, + "grad_norm": 2.9598376750946045, + "learning_rate": 0.00016980107750510116, + "loss": 0.7636, + "step": 15936 + }, + { + "epoch": 1.1058913330095066, + "grad_norm": 2.496095657348633, + "learning_rate": 0.00016979705397599342, + "loss": 0.6934, + "step": 15937 + }, + { + "epoch": 1.1059607244466032, + "grad_norm": 1.8048501014709473, + "learning_rate": 0.00016979303022654447, + "loss": 0.6689, + "step": 15938 + }, + { + "epoch": 1.1060301158837, + "grad_norm": 1.8427064418792725, + "learning_rate": 0.00016978900625676707, + "loss": 0.5777, + "step": 15939 + }, + { + "epoch": 1.1060995073207966, + "grad_norm": 2.1486940383911133, + "learning_rate": 0.00016978498206667383, + "loss": 0.8974, + "step": 15940 + }, + { + "epoch": 1.1061688987578933, + "grad_norm": 2.061898708343506, + "learning_rate": 0.00016978095765627754, + "loss": 0.8778, + "step": 15941 + }, + { + "epoch": 1.1062382901949899, + "grad_norm": 2.2438900470733643, + "learning_rate": 0.00016977693302559082, + "loss": 0.9185, + "step": 15942 + }, + { + "epoch": 1.1063076816320867, + "grad_norm": 2.251554012298584, + "learning_rate": 0.00016977290817462646, + "loss": 0.678, + "step": 15943 + }, + { + "epoch": 1.1063770730691833, + "grad_norm": 1.9112064838409424, + "learning_rate": 0.0001697688831033971, + "loss": 0.6228, + "step": 15944 + }, + { + "epoch": 1.10644646450628, + "grad_norm": 1.4977391958236694, + "learning_rate": 0.00016976485781191545, + "loss": 0.5834, + "step": 15945 + }, + { + "epoch": 1.1065158559433765, + "grad_norm": 1.783108115196228, + "learning_rate": 0.0001697608323001943, + "loss": 0.5714, + "step": 15946 + }, + { + "epoch": 1.1065852473804731, + "grad_norm": 1.6663403511047363, + "learning_rate": 0.0001697568065682462, + "loss": 0.7468, + "step": 15947 + }, + { + "epoch": 1.10665463881757, + "grad_norm": 3.7411844730377197, + "learning_rate": 0.00016975278061608404, + "loss": 0.7362, + "step": 15948 + }, + { + "epoch": 1.1067240302546666, + "grad_norm": 2.202188730239868, + "learning_rate": 0.00016974875444372042, + "loss": 0.9587, + "step": 15949 + }, + { + "epoch": 1.1067934216917632, + "grad_norm": 1.5939475297927856, + "learning_rate": 0.00016974472805116807, + "loss": 0.6228, + "step": 15950 + }, + { + "epoch": 1.1068628131288598, + "grad_norm": 2.1407341957092285, + "learning_rate": 0.0001697407014384397, + "loss": 0.7189, + "step": 15951 + }, + { + "epoch": 1.1069322045659566, + "grad_norm": 2.209465503692627, + "learning_rate": 0.00016973667460554802, + "loss": 0.5642, + "step": 15952 + }, + { + "epoch": 1.1070015960030533, + "grad_norm": 2.4666032791137695, + "learning_rate": 0.0001697326475525058, + "loss": 1.1146, + "step": 15953 + }, + { + "epoch": 1.1070709874401499, + "grad_norm": 1.7283127307891846, + "learning_rate": 0.00016972862027932566, + "loss": 0.6115, + "step": 15954 + }, + { + "epoch": 1.1071403788772465, + "grad_norm": 2.3916189670562744, + "learning_rate": 0.00016972459278602036, + "loss": 0.7255, + "step": 15955 + }, + { + "epoch": 1.1072097703143433, + "grad_norm": 3.166316509246826, + "learning_rate": 0.00016972056507260258, + "loss": 0.7615, + "step": 15956 + }, + { + "epoch": 1.10727916175144, + "grad_norm": 1.8235381841659546, + "learning_rate": 0.0001697165371390851, + "loss": 0.6551, + "step": 15957 + }, + { + "epoch": 1.1073485531885365, + "grad_norm": 2.4881479740142822, + "learning_rate": 0.0001697125089854806, + "loss": 0.8271, + "step": 15958 + }, + { + "epoch": 1.1074179446256331, + "grad_norm": 1.4303559064865112, + "learning_rate": 0.0001697084806118018, + "loss": 0.5668, + "step": 15959 + }, + { + "epoch": 1.10748733606273, + "grad_norm": 1.9895936250686646, + "learning_rate": 0.00016970445201806142, + "loss": 0.6254, + "step": 15960 + }, + { + "epoch": 1.1075567274998266, + "grad_norm": 1.8617690801620483, + "learning_rate": 0.00016970042320427214, + "loss": 0.7878, + "step": 15961 + }, + { + "epoch": 1.1076261189369232, + "grad_norm": 2.5935943126678467, + "learning_rate": 0.00016969639417044674, + "loss": 1.0613, + "step": 15962 + }, + { + "epoch": 1.1076955103740198, + "grad_norm": 1.588821530342102, + "learning_rate": 0.0001696923649165979, + "loss": 0.4884, + "step": 15963 + }, + { + "epoch": 1.1077649018111164, + "grad_norm": 2.294009208679199, + "learning_rate": 0.00016968833544273837, + "loss": 0.917, + "step": 15964 + }, + { + "epoch": 1.1078342932482133, + "grad_norm": 2.33524489402771, + "learning_rate": 0.0001696843057488808, + "loss": 0.8622, + "step": 15965 + }, + { + "epoch": 1.1079036846853099, + "grad_norm": 1.8638019561767578, + "learning_rate": 0.000169680275835038, + "loss": 0.5993, + "step": 15966 + }, + { + "epoch": 1.1079730761224065, + "grad_norm": 2.084453582763672, + "learning_rate": 0.00016967624570122264, + "loss": 0.6456, + "step": 15967 + }, + { + "epoch": 1.108042467559503, + "grad_norm": 1.6214631795883179, + "learning_rate": 0.00016967221534744747, + "loss": 0.5437, + "step": 15968 + }, + { + "epoch": 1.1081118589966, + "grad_norm": 2.46456241607666, + "learning_rate": 0.0001696681847737252, + "loss": 0.8487, + "step": 15969 + }, + { + "epoch": 1.1081812504336965, + "grad_norm": 1.942217469215393, + "learning_rate": 0.00016966415398006856, + "loss": 0.6932, + "step": 15970 + }, + { + "epoch": 1.1082506418707931, + "grad_norm": 1.5888534784317017, + "learning_rate": 0.00016966012296649026, + "loss": 0.6766, + "step": 15971 + }, + { + "epoch": 1.1083200333078898, + "grad_norm": 1.99362051486969, + "learning_rate": 0.00016965609173300302, + "loss": 0.7434, + "step": 15972 + }, + { + "epoch": 1.1083894247449866, + "grad_norm": 2.3703501224517822, + "learning_rate": 0.00016965206027961957, + "loss": 0.8323, + "step": 15973 + }, + { + "epoch": 1.1084588161820832, + "grad_norm": 1.6673004627227783, + "learning_rate": 0.00016964802860635266, + "loss": 0.8838, + "step": 15974 + }, + { + "epoch": 1.1085282076191798, + "grad_norm": 1.872517704963684, + "learning_rate": 0.00016964399671321503, + "loss": 0.562, + "step": 15975 + }, + { + "epoch": 1.1085975990562764, + "grad_norm": 2.0517594814300537, + "learning_rate": 0.00016963996460021935, + "loss": 0.7465, + "step": 15976 + }, + { + "epoch": 1.108666990493373, + "grad_norm": 1.6076347827911377, + "learning_rate": 0.0001696359322673784, + "loss": 0.5463, + "step": 15977 + }, + { + "epoch": 1.1087363819304699, + "grad_norm": 2.3572003841400146, + "learning_rate": 0.00016963189971470487, + "loss": 0.5479, + "step": 15978 + }, + { + "epoch": 1.1088057733675665, + "grad_norm": 2.102653980255127, + "learning_rate": 0.00016962786694221154, + "loss": 0.7085, + "step": 15979 + }, + { + "epoch": 1.108875164804663, + "grad_norm": 1.622867465019226, + "learning_rate": 0.0001696238339499111, + "loss": 0.6474, + "step": 15980 + }, + { + "epoch": 1.1089445562417597, + "grad_norm": 1.9173518419265747, + "learning_rate": 0.00016961980073781628, + "loss": 0.693, + "step": 15981 + }, + { + "epoch": 1.1090139476788565, + "grad_norm": 2.011219024658203, + "learning_rate": 0.00016961576730593984, + "loss": 0.9238, + "step": 15982 + }, + { + "epoch": 1.1090833391159531, + "grad_norm": 1.5956292152404785, + "learning_rate": 0.00016961173365429447, + "loss": 0.6567, + "step": 15983 + }, + { + "epoch": 1.1091527305530497, + "grad_norm": 2.2775166034698486, + "learning_rate": 0.00016960769978289299, + "loss": 0.7288, + "step": 15984 + }, + { + "epoch": 1.1092221219901464, + "grad_norm": 1.9021238088607788, + "learning_rate": 0.00016960366569174803, + "loss": 0.6738, + "step": 15985 + }, + { + "epoch": 1.1092915134272432, + "grad_norm": 1.6500027179718018, + "learning_rate": 0.0001695996313808724, + "loss": 0.6443, + "step": 15986 + }, + { + "epoch": 1.1093609048643398, + "grad_norm": 2.134242057800293, + "learning_rate": 0.00016959559685027884, + "loss": 0.9448, + "step": 15987 + }, + { + "epoch": 1.1094302963014364, + "grad_norm": 1.8595608472824097, + "learning_rate": 0.00016959156209998, + "loss": 0.8767, + "step": 15988 + }, + { + "epoch": 1.109499687738533, + "grad_norm": 1.4718191623687744, + "learning_rate": 0.00016958752712998865, + "loss": 0.6629, + "step": 15989 + }, + { + "epoch": 1.1095690791756296, + "grad_norm": 1.99508535861969, + "learning_rate": 0.00016958349194031762, + "loss": 0.6131, + "step": 15990 + }, + { + "epoch": 1.1096384706127265, + "grad_norm": 1.8730181455612183, + "learning_rate": 0.00016957945653097955, + "loss": 0.7634, + "step": 15991 + }, + { + "epoch": 1.109707862049823, + "grad_norm": 1.7910176515579224, + "learning_rate": 0.00016957542090198722, + "loss": 0.5887, + "step": 15992 + }, + { + "epoch": 1.1097772534869197, + "grad_norm": 1.8447368144989014, + "learning_rate": 0.00016957138505335333, + "loss": 0.6673, + "step": 15993 + }, + { + "epoch": 1.1098466449240163, + "grad_norm": 1.547207236289978, + "learning_rate": 0.0001695673489850907, + "loss": 0.8034, + "step": 15994 + }, + { + "epoch": 1.1099160363611131, + "grad_norm": 1.7553789615631104, + "learning_rate": 0.00016956331269721194, + "loss": 0.6825, + "step": 15995 + }, + { + "epoch": 1.1099854277982097, + "grad_norm": 1.855778455734253, + "learning_rate": 0.00016955927618972992, + "loss": 0.6803, + "step": 15996 + }, + { + "epoch": 1.1100548192353064, + "grad_norm": 1.9365589618682861, + "learning_rate": 0.00016955523946265736, + "loss": 0.6959, + "step": 15997 + }, + { + "epoch": 1.110124210672403, + "grad_norm": 1.6437312364578247, + "learning_rate": 0.00016955120251600694, + "loss": 0.554, + "step": 15998 + }, + { + "epoch": 1.1101936021094998, + "grad_norm": 2.072890281677246, + "learning_rate": 0.00016954716534979146, + "loss": 0.8149, + "step": 15999 + }, + { + "epoch": 1.1102629935465964, + "grad_norm": 1.8010828495025635, + "learning_rate": 0.00016954312796402363, + "loss": 0.6704, + "step": 16000 + }, + { + "epoch": 1.1102629935465964, + "eval_loss": 0.7437890768051147, + "eval_runtime": 3813.1256, + "eval_samples_per_second": 6.365, + "eval_steps_per_second": 0.796, + "step": 16000 + }, + { + "epoch": 1.110332384983693, + "grad_norm": 6.772910118103027, + "learning_rate": 0.00016953909035871621, + "loss": 0.9618, + "step": 16001 + }, + { + "epoch": 1.1104017764207896, + "grad_norm": 1.576694369316101, + "learning_rate": 0.000169535052533882, + "loss": 0.458, + "step": 16002 + }, + { + "epoch": 1.1104711678578862, + "grad_norm": 2.237835645675659, + "learning_rate": 0.00016953101448953365, + "loss": 0.8202, + "step": 16003 + }, + { + "epoch": 1.110540559294983, + "grad_norm": 1.5627386569976807, + "learning_rate": 0.00016952697622568395, + "loss": 0.5582, + "step": 16004 + }, + { + "epoch": 1.1106099507320797, + "grad_norm": 2.1824657917022705, + "learning_rate": 0.00016952293774234568, + "loss": 0.6889, + "step": 16005 + }, + { + "epoch": 1.1106793421691763, + "grad_norm": 2.107924222946167, + "learning_rate": 0.00016951889903953155, + "loss": 0.9852, + "step": 16006 + }, + { + "epoch": 1.110748733606273, + "grad_norm": 1.6743063926696777, + "learning_rate": 0.00016951486011725433, + "loss": 0.6865, + "step": 16007 + }, + { + "epoch": 1.1108181250433697, + "grad_norm": 1.6725846529006958, + "learning_rate": 0.00016951082097552679, + "loss": 0.6747, + "step": 16008 + }, + { + "epoch": 1.1108875164804664, + "grad_norm": 1.772179365158081, + "learning_rate": 0.00016950678161436157, + "loss": 0.7646, + "step": 16009 + }, + { + "epoch": 1.110956907917563, + "grad_norm": 1.6257480382919312, + "learning_rate": 0.00016950274203377156, + "loss": 0.4858, + "step": 16010 + }, + { + "epoch": 1.1110262993546596, + "grad_norm": 2.0876824855804443, + "learning_rate": 0.00016949870223376947, + "loss": 0.7535, + "step": 16011 + }, + { + "epoch": 1.1110956907917564, + "grad_norm": 1.8584997653961182, + "learning_rate": 0.00016949466221436802, + "loss": 0.671, + "step": 16012 + }, + { + "epoch": 1.111165082228853, + "grad_norm": 2.6197774410247803, + "learning_rate": 0.00016949062197557999, + "loss": 0.7084, + "step": 16013 + }, + { + "epoch": 1.1112344736659496, + "grad_norm": 2.144470453262329, + "learning_rate": 0.00016948658151741812, + "loss": 0.6207, + "step": 16014 + }, + { + "epoch": 1.1113038651030462, + "grad_norm": 2.3410933017730713, + "learning_rate": 0.00016948254083989518, + "loss": 0.8855, + "step": 16015 + }, + { + "epoch": 1.1113732565401429, + "grad_norm": 1.955442190170288, + "learning_rate": 0.0001694784999430239, + "loss": 0.6633, + "step": 16016 + }, + { + "epoch": 1.1114426479772397, + "grad_norm": 2.845179319381714, + "learning_rate": 0.0001694744588268171, + "loss": 0.7328, + "step": 16017 + }, + { + "epoch": 1.1115120394143363, + "grad_norm": 2.4192354679107666, + "learning_rate": 0.00016947041749128746, + "loss": 0.6701, + "step": 16018 + }, + { + "epoch": 1.111581430851433, + "grad_norm": 2.369917154312134, + "learning_rate": 0.00016946637593644778, + "loss": 0.7159, + "step": 16019 + }, + { + "epoch": 1.1116508222885295, + "grad_norm": 1.549509048461914, + "learning_rate": 0.0001694623341623108, + "loss": 0.611, + "step": 16020 + }, + { + "epoch": 1.1117202137256263, + "grad_norm": 1.9119594097137451, + "learning_rate": 0.00016945829216888928, + "loss": 0.816, + "step": 16021 + }, + { + "epoch": 1.111789605162723, + "grad_norm": 2.2436680793762207, + "learning_rate": 0.00016945424995619605, + "loss": 0.7388, + "step": 16022 + }, + { + "epoch": 1.1118589965998196, + "grad_norm": 2.5530073642730713, + "learning_rate": 0.00016945020752424377, + "loss": 0.738, + "step": 16023 + }, + { + "epoch": 1.1119283880369162, + "grad_norm": 1.8134154081344604, + "learning_rate": 0.00016944616487304522, + "loss": 0.5971, + "step": 16024 + }, + { + "epoch": 1.111997779474013, + "grad_norm": 2.0300209522247314, + "learning_rate": 0.00016944212200261323, + "loss": 0.6281, + "step": 16025 + }, + { + "epoch": 1.1120671709111096, + "grad_norm": 2.084049701690674, + "learning_rate": 0.0001694380789129605, + "loss": 0.5173, + "step": 16026 + }, + { + "epoch": 1.1121365623482062, + "grad_norm": 1.6722739934921265, + "learning_rate": 0.00016943403560409982, + "loss": 0.5401, + "step": 16027 + }, + { + "epoch": 1.1122059537853028, + "grad_norm": 3.059826374053955, + "learning_rate": 0.00016942999207604392, + "loss": 0.6775, + "step": 16028 + }, + { + "epoch": 1.1122753452223995, + "grad_norm": 3.320307493209839, + "learning_rate": 0.0001694259483288056, + "loss": 0.8201, + "step": 16029 + }, + { + "epoch": 1.1123447366594963, + "grad_norm": 2.0408263206481934, + "learning_rate": 0.0001694219043623976, + "loss": 0.6945, + "step": 16030 + }, + { + "epoch": 1.112414128096593, + "grad_norm": 2.5876307487487793, + "learning_rate": 0.00016941786017683275, + "loss": 0.5575, + "step": 16031 + }, + { + "epoch": 1.1124835195336895, + "grad_norm": 1.8874914646148682, + "learning_rate": 0.00016941381577212375, + "loss": 0.7746, + "step": 16032 + }, + { + "epoch": 1.1125529109707861, + "grad_norm": 0.9817914962768555, + "learning_rate": 0.0001694097711482834, + "loss": 0.418, + "step": 16033 + }, + { + "epoch": 1.112622302407883, + "grad_norm": 2.182610034942627, + "learning_rate": 0.00016940572630532442, + "loss": 0.7485, + "step": 16034 + }, + { + "epoch": 1.1126916938449796, + "grad_norm": 1.726538896560669, + "learning_rate": 0.00016940168124325963, + "loss": 0.698, + "step": 16035 + }, + { + "epoch": 1.1127610852820762, + "grad_norm": 1.9388816356658936, + "learning_rate": 0.00016939763596210175, + "loss": 0.5104, + "step": 16036 + }, + { + "epoch": 1.1128304767191728, + "grad_norm": 1.9339183568954468, + "learning_rate": 0.00016939359046186362, + "loss": 0.6101, + "step": 16037 + }, + { + "epoch": 1.1128998681562696, + "grad_norm": 1.6843440532684326, + "learning_rate": 0.00016938954474255796, + "loss": 0.6622, + "step": 16038 + }, + { + "epoch": 1.1129692595933662, + "grad_norm": 1.7294219732284546, + "learning_rate": 0.00016938549880419758, + "loss": 0.6082, + "step": 16039 + }, + { + "epoch": 1.1130386510304628, + "grad_norm": 1.6491141319274902, + "learning_rate": 0.00016938145264679526, + "loss": 0.6435, + "step": 16040 + }, + { + "epoch": 1.1131080424675595, + "grad_norm": 2.735765218734741, + "learning_rate": 0.0001693774062703637, + "loss": 0.7506, + "step": 16041 + }, + { + "epoch": 1.113177433904656, + "grad_norm": 2.2303755283355713, + "learning_rate": 0.00016937335967491573, + "loss": 0.6603, + "step": 16042 + }, + { + "epoch": 1.113246825341753, + "grad_norm": 1.5089678764343262, + "learning_rate": 0.0001693693128604641, + "loss": 0.6769, + "step": 16043 + }, + { + "epoch": 1.1133162167788495, + "grad_norm": 1.7298552989959717, + "learning_rate": 0.0001693652658270216, + "loss": 0.6957, + "step": 16044 + }, + { + "epoch": 1.1133856082159461, + "grad_norm": 1.9311397075653076, + "learning_rate": 0.000169361218574601, + "loss": 0.6466, + "step": 16045 + }, + { + "epoch": 1.1134549996530427, + "grad_norm": 1.379914402961731, + "learning_rate": 0.0001693571711032151, + "loss": 0.5212, + "step": 16046 + }, + { + "epoch": 1.1135243910901396, + "grad_norm": 1.8950309753417969, + "learning_rate": 0.00016935312341287664, + "loss": 0.7429, + "step": 16047 + }, + { + "epoch": 1.1135937825272362, + "grad_norm": 1.650924801826477, + "learning_rate": 0.00016934907550359842, + "loss": 0.5249, + "step": 16048 + }, + { + "epoch": 1.1136631739643328, + "grad_norm": 1.6592020988464355, + "learning_rate": 0.00016934502737539324, + "loss": 0.6695, + "step": 16049 + }, + { + "epoch": 1.1137325654014294, + "grad_norm": 2.5167417526245117, + "learning_rate": 0.00016934097902827382, + "loss": 0.8964, + "step": 16050 + }, + { + "epoch": 1.1138019568385262, + "grad_norm": 2.169029951095581, + "learning_rate": 0.00016933693046225297, + "loss": 0.7116, + "step": 16051 + }, + { + "epoch": 1.1138713482756228, + "grad_norm": 2.3616485595703125, + "learning_rate": 0.0001693328816773435, + "loss": 0.828, + "step": 16052 + }, + { + "epoch": 1.1139407397127195, + "grad_norm": 1.6650935411453247, + "learning_rate": 0.00016932883267355817, + "loss": 0.7389, + "step": 16053 + }, + { + "epoch": 1.114010131149816, + "grad_norm": 1.8850594758987427, + "learning_rate": 0.00016932478345090975, + "loss": 0.8922, + "step": 16054 + }, + { + "epoch": 1.1140795225869127, + "grad_norm": 2.058067560195923, + "learning_rate": 0.000169320734009411, + "loss": 0.7546, + "step": 16055 + }, + { + "epoch": 1.1141489140240095, + "grad_norm": 1.6529425382614136, + "learning_rate": 0.00016931668434907478, + "loss": 0.5253, + "step": 16056 + }, + { + "epoch": 1.1142183054611061, + "grad_norm": 1.5830130577087402, + "learning_rate": 0.0001693126344699138, + "loss": 0.4979, + "step": 16057 + }, + { + "epoch": 1.1142876968982027, + "grad_norm": 1.315726399421692, + "learning_rate": 0.0001693085843719409, + "loss": 0.4728, + "step": 16058 + }, + { + "epoch": 1.1143570883352993, + "grad_norm": 2.2612204551696777, + "learning_rate": 0.00016930453405516883, + "loss": 0.9098, + "step": 16059 + }, + { + "epoch": 1.1144264797723962, + "grad_norm": 1.8909263610839844, + "learning_rate": 0.00016930048351961036, + "loss": 0.6834, + "step": 16060 + }, + { + "epoch": 1.1144958712094928, + "grad_norm": 2.353689193725586, + "learning_rate": 0.00016929643276527835, + "loss": 0.9102, + "step": 16061 + }, + { + "epoch": 1.1145652626465894, + "grad_norm": 1.7553150653839111, + "learning_rate": 0.00016929238179218552, + "loss": 0.6831, + "step": 16062 + }, + { + "epoch": 1.114634654083686, + "grad_norm": 1.9339234828948975, + "learning_rate": 0.00016928833060034468, + "loss": 0.81, + "step": 16063 + }, + { + "epoch": 1.1147040455207828, + "grad_norm": 2.2448174953460693, + "learning_rate": 0.00016928427918976862, + "loss": 0.6501, + "step": 16064 + }, + { + "epoch": 1.1147734369578794, + "grad_norm": 1.8900476694107056, + "learning_rate": 0.00016928022756047012, + "loss": 0.539, + "step": 16065 + }, + { + "epoch": 1.114842828394976, + "grad_norm": 1.7155362367630005, + "learning_rate": 0.000169276175712462, + "loss": 0.6607, + "step": 16066 + }, + { + "epoch": 1.1149122198320727, + "grad_norm": 2.49626088142395, + "learning_rate": 0.00016927212364575703, + "loss": 0.871, + "step": 16067 + }, + { + "epoch": 1.1149816112691693, + "grad_norm": 1.8687556982040405, + "learning_rate": 0.00016926807136036798, + "loss": 0.7806, + "step": 16068 + }, + { + "epoch": 1.1150510027062661, + "grad_norm": 3.40517258644104, + "learning_rate": 0.00016926401885630768, + "loss": 0.8634, + "step": 16069 + }, + { + "epoch": 1.1151203941433627, + "grad_norm": 1.413211703300476, + "learning_rate": 0.0001692599661335889, + "loss": 0.4987, + "step": 16070 + }, + { + "epoch": 1.1151897855804593, + "grad_norm": 1.413206934928894, + "learning_rate": 0.00016925591319222444, + "loss": 0.5839, + "step": 16071 + }, + { + "epoch": 1.115259177017556, + "grad_norm": 1.5375044345855713, + "learning_rate": 0.0001692518600322271, + "loss": 0.7025, + "step": 16072 + }, + { + "epoch": 1.1153285684546528, + "grad_norm": 2.0867671966552734, + "learning_rate": 0.0001692478066536097, + "loss": 1.0087, + "step": 16073 + }, + { + "epoch": 1.1153979598917494, + "grad_norm": 1.9927130937576294, + "learning_rate": 0.00016924375305638498, + "loss": 0.5244, + "step": 16074 + }, + { + "epoch": 1.115467351328846, + "grad_norm": 1.4714674949645996, + "learning_rate": 0.00016923969924056577, + "loss": 0.6272, + "step": 16075 + }, + { + "epoch": 1.1155367427659426, + "grad_norm": 1.7372171878814697, + "learning_rate": 0.00016923564520616486, + "loss": 0.7161, + "step": 16076 + }, + { + "epoch": 1.1156061342030394, + "grad_norm": 2.284635305404663, + "learning_rate": 0.00016923159095319506, + "loss": 0.6759, + "step": 16077 + }, + { + "epoch": 1.115675525640136, + "grad_norm": 2.4138238430023193, + "learning_rate": 0.00016922753648166916, + "loss": 1.1502, + "step": 16078 + }, + { + "epoch": 1.1157449170772327, + "grad_norm": 2.063154697418213, + "learning_rate": 0.00016922348179159994, + "loss": 0.9541, + "step": 16079 + }, + { + "epoch": 1.1158143085143293, + "grad_norm": 1.9672152996063232, + "learning_rate": 0.00016921942688300024, + "loss": 0.5873, + "step": 16080 + }, + { + "epoch": 1.115883699951426, + "grad_norm": 2.1919138431549072, + "learning_rate": 0.00016921537175588282, + "loss": 0.7666, + "step": 16081 + }, + { + "epoch": 1.1159530913885227, + "grad_norm": 2.5230157375335693, + "learning_rate": 0.00016921131641026054, + "loss": 0.753, + "step": 16082 + }, + { + "epoch": 1.1160224828256193, + "grad_norm": 2.4759013652801514, + "learning_rate": 0.00016920726084614614, + "loss": 0.7747, + "step": 16083 + }, + { + "epoch": 1.116091874262716, + "grad_norm": 1.5779424905776978, + "learning_rate": 0.00016920320506355244, + "loss": 0.5718, + "step": 16084 + }, + { + "epoch": 1.1161612656998126, + "grad_norm": 2.0419559478759766, + "learning_rate": 0.00016919914906249225, + "loss": 0.7124, + "step": 16085 + }, + { + "epoch": 1.1162306571369094, + "grad_norm": 1.7410826683044434, + "learning_rate": 0.0001691950928429784, + "loss": 0.8223, + "step": 16086 + }, + { + "epoch": 1.116300048574006, + "grad_norm": 2.1728570461273193, + "learning_rate": 0.00016919103640502365, + "loss": 0.7056, + "step": 16087 + }, + { + "epoch": 1.1163694400111026, + "grad_norm": 2.136584758758545, + "learning_rate": 0.00016918697974864082, + "loss": 0.8031, + "step": 16088 + }, + { + "epoch": 1.1164388314481992, + "grad_norm": 1.8552422523498535, + "learning_rate": 0.00016918292287384273, + "loss": 0.5686, + "step": 16089 + }, + { + "epoch": 1.116508222885296, + "grad_norm": 1.5707534551620483, + "learning_rate": 0.00016917886578064216, + "loss": 0.5727, + "step": 16090 + }, + { + "epoch": 1.1165776143223927, + "grad_norm": 2.5951039791107178, + "learning_rate": 0.00016917480846905198, + "loss": 0.7295, + "step": 16091 + }, + { + "epoch": 1.1166470057594893, + "grad_norm": 1.7029328346252441, + "learning_rate": 0.00016917075093908493, + "loss": 0.678, + "step": 16092 + }, + { + "epoch": 1.1167163971965859, + "grad_norm": 2.610166311264038, + "learning_rate": 0.00016916669319075385, + "loss": 0.6723, + "step": 16093 + }, + { + "epoch": 1.1167857886336825, + "grad_norm": 2.136469602584839, + "learning_rate": 0.00016916263522407158, + "loss": 0.6139, + "step": 16094 + }, + { + "epoch": 1.1168551800707793, + "grad_norm": 1.6474623680114746, + "learning_rate": 0.00016915857703905083, + "loss": 0.4725, + "step": 16095 + }, + { + "epoch": 1.116924571507876, + "grad_norm": 1.7831737995147705, + "learning_rate": 0.00016915451863570448, + "loss": 0.5677, + "step": 16096 + }, + { + "epoch": 1.1169939629449726, + "grad_norm": 1.7093844413757324, + "learning_rate": 0.00016915046001404535, + "loss": 0.644, + "step": 16097 + }, + { + "epoch": 1.1170633543820692, + "grad_norm": 2.1799142360687256, + "learning_rate": 0.00016914640117408628, + "loss": 0.6918, + "step": 16098 + }, + { + "epoch": 1.117132745819166, + "grad_norm": 1.5145037174224854, + "learning_rate": 0.00016914234211584003, + "loss": 0.4713, + "step": 16099 + }, + { + "epoch": 1.1172021372562626, + "grad_norm": 1.8055082559585571, + "learning_rate": 0.0001691382828393194, + "loss": 0.7538, + "step": 16100 + }, + { + "epoch": 1.1172715286933592, + "grad_norm": 2.012340545654297, + "learning_rate": 0.00016913422334453724, + "loss": 0.759, + "step": 16101 + }, + { + "epoch": 1.1173409201304558, + "grad_norm": 1.0870866775512695, + "learning_rate": 0.00016913016363150635, + "loss": 0.4111, + "step": 16102 + }, + { + "epoch": 1.1174103115675527, + "grad_norm": 1.847985863685608, + "learning_rate": 0.00016912610370023958, + "loss": 0.5637, + "step": 16103 + }, + { + "epoch": 1.1174797030046493, + "grad_norm": 1.8152908086776733, + "learning_rate": 0.0001691220435507497, + "loss": 0.7374, + "step": 16104 + }, + { + "epoch": 1.1175490944417459, + "grad_norm": 2.332209587097168, + "learning_rate": 0.00016911798318304956, + "loss": 0.6969, + "step": 16105 + }, + { + "epoch": 1.1176184858788425, + "grad_norm": 2.158123731613159, + "learning_rate": 0.00016911392259715195, + "loss": 0.7954, + "step": 16106 + }, + { + "epoch": 1.117687877315939, + "grad_norm": 1.6636991500854492, + "learning_rate": 0.0001691098617930697, + "loss": 0.4069, + "step": 16107 + }, + { + "epoch": 1.117757268753036, + "grad_norm": 1.983849287033081, + "learning_rate": 0.0001691058007708157, + "loss": 0.806, + "step": 16108 + }, + { + "epoch": 1.1178266601901325, + "grad_norm": 2.064303398132324, + "learning_rate": 0.00016910173953040266, + "loss": 0.8649, + "step": 16109 + }, + { + "epoch": 1.1178960516272292, + "grad_norm": 1.3751250505447388, + "learning_rate": 0.00016909767807184344, + "loss": 0.513, + "step": 16110 + }, + { + "epoch": 1.1179654430643258, + "grad_norm": 1.9388772249221802, + "learning_rate": 0.00016909361639515087, + "loss": 0.5866, + "step": 16111 + }, + { + "epoch": 1.1180348345014226, + "grad_norm": 1.4795979261398315, + "learning_rate": 0.00016908955450033774, + "loss": 0.6015, + "step": 16112 + }, + { + "epoch": 1.1181042259385192, + "grad_norm": 1.7439113855361938, + "learning_rate": 0.00016908549238741694, + "loss": 0.8415, + "step": 16113 + }, + { + "epoch": 1.1181736173756158, + "grad_norm": 1.6811668872833252, + "learning_rate": 0.00016908143005640126, + "loss": 0.7291, + "step": 16114 + }, + { + "epoch": 1.1182430088127124, + "grad_norm": 2.7864811420440674, + "learning_rate": 0.0001690773675073035, + "loss": 0.8354, + "step": 16115 + }, + { + "epoch": 1.1183124002498093, + "grad_norm": 1.9144574403762817, + "learning_rate": 0.00016907330474013653, + "loss": 0.6361, + "step": 16116 + }, + { + "epoch": 1.1183817916869059, + "grad_norm": 1.268575668334961, + "learning_rate": 0.00016906924175491312, + "loss": 0.4752, + "step": 16117 + }, + { + "epoch": 1.1184511831240025, + "grad_norm": 1.6264523267745972, + "learning_rate": 0.00016906517855164613, + "loss": 0.5664, + "step": 16118 + }, + { + "epoch": 1.118520574561099, + "grad_norm": 1.4816077947616577, + "learning_rate": 0.00016906111513034842, + "loss": 0.4236, + "step": 16119 + }, + { + "epoch": 1.1185899659981957, + "grad_norm": 1.820052981376648, + "learning_rate": 0.00016905705149103276, + "loss": 0.4445, + "step": 16120 + }, + { + "epoch": 1.1186593574352925, + "grad_norm": 2.090268611907959, + "learning_rate": 0.00016905298763371198, + "loss": 0.7205, + "step": 16121 + }, + { + "epoch": 1.1187287488723892, + "grad_norm": 1.7216517925262451, + "learning_rate": 0.00016904892355839896, + "loss": 0.5276, + "step": 16122 + }, + { + "epoch": 1.1187981403094858, + "grad_norm": 1.9816226959228516, + "learning_rate": 0.00016904485926510647, + "loss": 0.7494, + "step": 16123 + }, + { + "epoch": 1.1188675317465824, + "grad_norm": 1.641957402229309, + "learning_rate": 0.0001690407947538474, + "loss": 0.6835, + "step": 16124 + }, + { + "epoch": 1.1189369231836792, + "grad_norm": 1.8750205039978027, + "learning_rate": 0.00016903673002463453, + "loss": 0.6483, + "step": 16125 + }, + { + "epoch": 1.1190063146207758, + "grad_norm": 1.56289541721344, + "learning_rate": 0.00016903266507748072, + "loss": 0.5894, + "step": 16126 + }, + { + "epoch": 1.1190757060578724, + "grad_norm": 1.797067403793335, + "learning_rate": 0.00016902859991239878, + "loss": 0.5356, + "step": 16127 + }, + { + "epoch": 1.119145097494969, + "grad_norm": 1.783265233039856, + "learning_rate": 0.0001690245345294016, + "loss": 0.8328, + "step": 16128 + }, + { + "epoch": 1.1192144889320659, + "grad_norm": 2.2722339630126953, + "learning_rate": 0.00016902046892850194, + "loss": 0.8696, + "step": 16129 + }, + { + "epoch": 1.1192838803691625, + "grad_norm": 2.1106607913970947, + "learning_rate": 0.0001690164031097127, + "loss": 0.9058, + "step": 16130 + }, + { + "epoch": 1.119353271806259, + "grad_norm": 2.281728506088257, + "learning_rate": 0.00016901233707304663, + "loss": 0.6477, + "step": 16131 + }, + { + "epoch": 1.1194226632433557, + "grad_norm": 2.089737892150879, + "learning_rate": 0.00016900827081851668, + "loss": 0.6742, + "step": 16132 + }, + { + "epoch": 1.1194920546804523, + "grad_norm": 1.6335698366165161, + "learning_rate": 0.00016900420434613558, + "loss": 0.908, + "step": 16133 + }, + { + "epoch": 1.1195614461175492, + "grad_norm": 2.3742835521698, + "learning_rate": 0.00016900013765591622, + "loss": 0.9556, + "step": 16134 + }, + { + "epoch": 1.1196308375546458, + "grad_norm": 2.7166013717651367, + "learning_rate": 0.00016899607074787146, + "loss": 0.8921, + "step": 16135 + }, + { + "epoch": 1.1197002289917424, + "grad_norm": 1.5102653503417969, + "learning_rate": 0.00016899200362201406, + "loss": 0.6851, + "step": 16136 + }, + { + "epoch": 1.119769620428839, + "grad_norm": 1.6285291910171509, + "learning_rate": 0.00016898793627835693, + "loss": 0.6887, + "step": 16137 + }, + { + "epoch": 1.1198390118659358, + "grad_norm": 1.9521327018737793, + "learning_rate": 0.00016898386871691293, + "loss": 0.7897, + "step": 16138 + }, + { + "epoch": 1.1199084033030324, + "grad_norm": 3.5222346782684326, + "learning_rate": 0.0001689798009376948, + "loss": 0.6837, + "step": 16139 + }, + { + "epoch": 1.119977794740129, + "grad_norm": 2.1185970306396484, + "learning_rate": 0.00016897573294071547, + "loss": 0.7991, + "step": 16140 + }, + { + "epoch": 1.1200471861772257, + "grad_norm": 1.534494161605835, + "learning_rate": 0.00016897166472598776, + "loss": 0.4652, + "step": 16141 + }, + { + "epoch": 1.1201165776143225, + "grad_norm": 1.8728829622268677, + "learning_rate": 0.00016896759629352447, + "loss": 0.6093, + "step": 16142 + }, + { + "epoch": 1.120185969051419, + "grad_norm": 1.9800009727478027, + "learning_rate": 0.00016896352764333853, + "loss": 0.7952, + "step": 16143 + }, + { + "epoch": 1.1202553604885157, + "grad_norm": 1.8255138397216797, + "learning_rate": 0.0001689594587754427, + "loss": 0.5735, + "step": 16144 + }, + { + "epoch": 1.1203247519256123, + "grad_norm": 1.9119441509246826, + "learning_rate": 0.00016895538968984987, + "loss": 0.7776, + "step": 16145 + }, + { + "epoch": 1.120394143362709, + "grad_norm": 2.3757283687591553, + "learning_rate": 0.0001689513203865729, + "loss": 0.7033, + "step": 16146 + }, + { + "epoch": 1.1204635347998058, + "grad_norm": 1.6831951141357422, + "learning_rate": 0.00016894725086562457, + "loss": 0.5755, + "step": 16147 + }, + { + "epoch": 1.1205329262369024, + "grad_norm": 2.4048824310302734, + "learning_rate": 0.0001689431811270178, + "loss": 0.9413, + "step": 16148 + }, + { + "epoch": 1.120602317673999, + "grad_norm": 2.3318755626678467, + "learning_rate": 0.00016893911117076539, + "loss": 0.6889, + "step": 16149 + }, + { + "epoch": 1.1206717091110956, + "grad_norm": 1.6017591953277588, + "learning_rate": 0.0001689350409968802, + "loss": 0.5096, + "step": 16150 + }, + { + "epoch": 1.1207411005481924, + "grad_norm": 2.508453369140625, + "learning_rate": 0.0001689309706053751, + "loss": 0.7701, + "step": 16151 + }, + { + "epoch": 1.120810491985289, + "grad_norm": 1.9162746667861938, + "learning_rate": 0.0001689268999962629, + "loss": 0.6801, + "step": 16152 + }, + { + "epoch": 1.1208798834223856, + "grad_norm": 1.8442710638046265, + "learning_rate": 0.00016892282916955648, + "loss": 0.5892, + "step": 16153 + }, + { + "epoch": 1.1209492748594823, + "grad_norm": 2.3707163333892822, + "learning_rate": 0.0001689187581252687, + "loss": 0.8594, + "step": 16154 + }, + { + "epoch": 1.121018666296579, + "grad_norm": 2.3408913612365723, + "learning_rate": 0.00016891468686341238, + "loss": 0.631, + "step": 16155 + }, + { + "epoch": 1.1210880577336757, + "grad_norm": 2.0705666542053223, + "learning_rate": 0.0001689106153840004, + "loss": 0.7589, + "step": 16156 + }, + { + "epoch": 1.1211574491707723, + "grad_norm": 2.1243975162506104, + "learning_rate": 0.0001689065436870456, + "loss": 0.7058, + "step": 16157 + }, + { + "epoch": 1.121226840607869, + "grad_norm": 1.6025497913360596, + "learning_rate": 0.00016890247177256084, + "loss": 0.6925, + "step": 16158 + }, + { + "epoch": 1.1212962320449655, + "grad_norm": 2.6515026092529297, + "learning_rate": 0.00016889839964055896, + "loss": 0.6797, + "step": 16159 + }, + { + "epoch": 1.1213656234820624, + "grad_norm": 2.576063871383667, + "learning_rate": 0.00016889432729105285, + "loss": 0.8145, + "step": 16160 + }, + { + "epoch": 1.121435014919159, + "grad_norm": 1.9737426042556763, + "learning_rate": 0.00016889025472405532, + "loss": 0.9061, + "step": 16161 + }, + { + "epoch": 1.1215044063562556, + "grad_norm": 5.179057598114014, + "learning_rate": 0.00016888618193957927, + "loss": 1.0125, + "step": 16162 + }, + { + "epoch": 1.1215737977933522, + "grad_norm": 2.005974769592285, + "learning_rate": 0.00016888210893763753, + "loss": 0.7936, + "step": 16163 + }, + { + "epoch": 1.121643189230449, + "grad_norm": 1.5585235357284546, + "learning_rate": 0.00016887803571824294, + "loss": 0.568, + "step": 16164 + }, + { + "epoch": 1.1217125806675456, + "grad_norm": 2.1788723468780518, + "learning_rate": 0.00016887396228140839, + "loss": 0.7912, + "step": 16165 + }, + { + "epoch": 1.1217819721046423, + "grad_norm": 2.240471839904785, + "learning_rate": 0.00016886988862714673, + "loss": 0.7223, + "step": 16166 + }, + { + "epoch": 1.1218513635417389, + "grad_norm": 1.9179682731628418, + "learning_rate": 0.00016886581475547083, + "loss": 0.5385, + "step": 16167 + }, + { + "epoch": 1.1219207549788357, + "grad_norm": 2.0311598777770996, + "learning_rate": 0.00016886174066639353, + "loss": 0.727, + "step": 16168 + }, + { + "epoch": 1.1219901464159323, + "grad_norm": 2.266434669494629, + "learning_rate": 0.00016885766635992773, + "loss": 0.6597, + "step": 16169 + }, + { + "epoch": 1.122059537853029, + "grad_norm": 2.2365641593933105, + "learning_rate": 0.00016885359183608625, + "loss": 0.7541, + "step": 16170 + }, + { + "epoch": 1.1221289292901255, + "grad_norm": 2.3409578800201416, + "learning_rate": 0.00016884951709488196, + "loss": 0.7473, + "step": 16171 + }, + { + "epoch": 1.1221983207272224, + "grad_norm": 2.1599998474121094, + "learning_rate": 0.00016884544213632775, + "loss": 0.6018, + "step": 16172 + }, + { + "epoch": 1.122267712164319, + "grad_norm": 1.732582926750183, + "learning_rate": 0.00016884136696043646, + "loss": 0.7077, + "step": 16173 + }, + { + "epoch": 1.1223371036014156, + "grad_norm": 1.7054266929626465, + "learning_rate": 0.000168837291567221, + "loss": 0.7074, + "step": 16174 + }, + { + "epoch": 1.1224064950385122, + "grad_norm": 1.2182320356369019, + "learning_rate": 0.00016883321595669414, + "loss": 0.5281, + "step": 16175 + }, + { + "epoch": 1.1224758864756088, + "grad_norm": 1.7918213605880737, + "learning_rate": 0.0001688291401288688, + "loss": 0.6257, + "step": 16176 + }, + { + "epoch": 1.1225452779127056, + "grad_norm": 1.9284933805465698, + "learning_rate": 0.0001688250640837579, + "loss": 0.7069, + "step": 16177 + }, + { + "epoch": 1.1226146693498023, + "grad_norm": 5.119495868682861, + "learning_rate": 0.00016882098782137423, + "loss": 0.7954, + "step": 16178 + }, + { + "epoch": 1.1226840607868989, + "grad_norm": 2.783090591430664, + "learning_rate": 0.0001688169113417307, + "loss": 0.5327, + "step": 16179 + }, + { + "epoch": 1.1227534522239955, + "grad_norm": 2.052110195159912, + "learning_rate": 0.00016881283464484015, + "loss": 0.7485, + "step": 16180 + }, + { + "epoch": 1.1228228436610923, + "grad_norm": 2.3084068298339844, + "learning_rate": 0.00016880875773071544, + "loss": 0.6612, + "step": 16181 + }, + { + "epoch": 1.122892235098189, + "grad_norm": 2.4806430339813232, + "learning_rate": 0.00016880468059936952, + "loss": 0.944, + "step": 16182 + }, + { + "epoch": 1.1229616265352855, + "grad_norm": 2.1620984077453613, + "learning_rate": 0.00016880060325081514, + "loss": 0.6996, + "step": 16183 + }, + { + "epoch": 1.1230310179723821, + "grad_norm": 3.0407657623291016, + "learning_rate": 0.0001687965256850653, + "loss": 0.5099, + "step": 16184 + }, + { + "epoch": 1.123100409409479, + "grad_norm": 1.946088433265686, + "learning_rate": 0.00016879244790213277, + "loss": 0.6107, + "step": 16185 + }, + { + "epoch": 1.1231698008465756, + "grad_norm": 1.9710010290145874, + "learning_rate": 0.0001687883699020305, + "loss": 0.7605, + "step": 16186 + }, + { + "epoch": 1.1232391922836722, + "grad_norm": 2.653749465942383, + "learning_rate": 0.00016878429168477134, + "loss": 0.7646, + "step": 16187 + }, + { + "epoch": 1.1233085837207688, + "grad_norm": 2.68743634223938, + "learning_rate": 0.0001687802132503681, + "loss": 0.7488, + "step": 16188 + }, + { + "epoch": 1.1233779751578654, + "grad_norm": 2.291670560836792, + "learning_rate": 0.00016877613459883372, + "loss": 0.7153, + "step": 16189 + }, + { + "epoch": 1.1234473665949622, + "grad_norm": 2.097799301147461, + "learning_rate": 0.00016877205573018108, + "loss": 0.6404, + "step": 16190 + }, + { + "epoch": 1.1235167580320589, + "grad_norm": 1.9200928211212158, + "learning_rate": 0.000168767976644423, + "loss": 0.7324, + "step": 16191 + }, + { + "epoch": 1.1235861494691555, + "grad_norm": 1.8709948062896729, + "learning_rate": 0.00016876389734157244, + "loss": 0.4453, + "step": 16192 + }, + { + "epoch": 1.123655540906252, + "grad_norm": 1.956957221031189, + "learning_rate": 0.0001687598178216422, + "loss": 0.8096, + "step": 16193 + }, + { + "epoch": 1.123724932343349, + "grad_norm": 2.1863608360290527, + "learning_rate": 0.00016875573808464522, + "loss": 0.7907, + "step": 16194 + }, + { + "epoch": 1.1237943237804455, + "grad_norm": 2.1408491134643555, + "learning_rate": 0.00016875165813059434, + "loss": 0.783, + "step": 16195 + }, + { + "epoch": 1.1238637152175421, + "grad_norm": 1.986379861831665, + "learning_rate": 0.00016874757795950246, + "loss": 0.8507, + "step": 16196 + }, + { + "epoch": 1.1239331066546387, + "grad_norm": 1.6711286306381226, + "learning_rate": 0.00016874349757138246, + "loss": 0.547, + "step": 16197 + }, + { + "epoch": 1.1240024980917356, + "grad_norm": 2.934093475341797, + "learning_rate": 0.0001687394169662472, + "loss": 0.8897, + "step": 16198 + }, + { + "epoch": 1.1240718895288322, + "grad_norm": 2.051455020904541, + "learning_rate": 0.00016873533614410958, + "loss": 0.6852, + "step": 16199 + }, + { + "epoch": 1.1241412809659288, + "grad_norm": 2.4777965545654297, + "learning_rate": 0.00016873125510498248, + "loss": 0.8355, + "step": 16200 + }, + { + "epoch": 1.1242106724030254, + "grad_norm": 1.6723976135253906, + "learning_rate": 0.00016872717384887876, + "loss": 0.5833, + "step": 16201 + }, + { + "epoch": 1.1242800638401222, + "grad_norm": 2.066460609436035, + "learning_rate": 0.00016872309237581136, + "loss": 0.6598, + "step": 16202 + }, + { + "epoch": 1.1243494552772189, + "grad_norm": 1.716312050819397, + "learning_rate": 0.0001687190106857931, + "loss": 0.5428, + "step": 16203 + }, + { + "epoch": 1.1244188467143155, + "grad_norm": 1.7893725633621216, + "learning_rate": 0.0001687149287788369, + "loss": 0.7261, + "step": 16204 + }, + { + "epoch": 1.124488238151412, + "grad_norm": 2.081921339035034, + "learning_rate": 0.00016871084665495567, + "loss": 0.8424, + "step": 16205 + }, + { + "epoch": 1.1245576295885087, + "grad_norm": 1.62126886844635, + "learning_rate": 0.00016870676431416225, + "loss": 0.5902, + "step": 16206 + }, + { + "epoch": 1.1246270210256055, + "grad_norm": 2.086649179458618, + "learning_rate": 0.00016870268175646954, + "loss": 0.7617, + "step": 16207 + }, + { + "epoch": 1.1246964124627021, + "grad_norm": 1.9005639553070068, + "learning_rate": 0.00016869859898189046, + "loss": 0.5863, + "step": 16208 + }, + { + "epoch": 1.1247658038997987, + "grad_norm": 2.790116310119629, + "learning_rate": 0.00016869451599043784, + "loss": 0.972, + "step": 16209 + }, + { + "epoch": 1.1248351953368954, + "grad_norm": 1.8887685537338257, + "learning_rate": 0.00016869043278212463, + "loss": 0.7063, + "step": 16210 + }, + { + "epoch": 1.1249045867739922, + "grad_norm": 1.8802947998046875, + "learning_rate": 0.0001686863493569637, + "loss": 0.764, + "step": 16211 + }, + { + "epoch": 1.1249739782110888, + "grad_norm": 1.924747109413147, + "learning_rate": 0.0001686822657149679, + "loss": 0.5669, + "step": 16212 + }, + { + "epoch": 1.1250433696481854, + "grad_norm": 1.6645723581314087, + "learning_rate": 0.00016867818185615016, + "loss": 0.6794, + "step": 16213 + }, + { + "epoch": 1.125112761085282, + "grad_norm": 1.7719000577926636, + "learning_rate": 0.0001686740977805234, + "loss": 0.7964, + "step": 16214 + }, + { + "epoch": 1.1251821525223789, + "grad_norm": 2.0291314125061035, + "learning_rate": 0.00016867001348810047, + "loss": 0.7444, + "step": 16215 + }, + { + "epoch": 1.1252515439594755, + "grad_norm": 1.6142888069152832, + "learning_rate": 0.00016866592897889427, + "loss": 0.5101, + "step": 16216 + }, + { + "epoch": 1.125320935396572, + "grad_norm": 1.9867902994155884, + "learning_rate": 0.00016866184425291767, + "loss": 0.7679, + "step": 16217 + }, + { + "epoch": 1.1253903268336687, + "grad_norm": 1.685838222503662, + "learning_rate": 0.00016865775931018364, + "loss": 0.5254, + "step": 16218 + }, + { + "epoch": 1.1254597182707653, + "grad_norm": 1.7907253503799438, + "learning_rate": 0.00016865367415070499, + "loss": 0.594, + "step": 16219 + }, + { + "epoch": 1.1255291097078621, + "grad_norm": 2.4425806999206543, + "learning_rate": 0.00016864958877449467, + "loss": 0.6812, + "step": 16220 + }, + { + "epoch": 1.1255985011449587, + "grad_norm": 2.088996648788452, + "learning_rate": 0.00016864550318156557, + "loss": 0.785, + "step": 16221 + }, + { + "epoch": 1.1256678925820554, + "grad_norm": 2.6292359828948975, + "learning_rate": 0.0001686414173719306, + "loss": 1.0077, + "step": 16222 + }, + { + "epoch": 1.125737284019152, + "grad_norm": 2.251047134399414, + "learning_rate": 0.0001686373313456026, + "loss": 0.5528, + "step": 16223 + }, + { + "epoch": 1.1258066754562488, + "grad_norm": 1.9739896059036255, + "learning_rate": 0.00016863324510259453, + "loss": 0.654, + "step": 16224 + }, + { + "epoch": 1.1258760668933454, + "grad_norm": 1.6395975351333618, + "learning_rate": 0.00016862915864291925, + "loss": 0.3661, + "step": 16225 + }, + { + "epoch": 1.125945458330442, + "grad_norm": 1.90915846824646, + "learning_rate": 0.00016862507196658972, + "loss": 0.6195, + "step": 16226 + }, + { + "epoch": 1.1260148497675386, + "grad_norm": 2.228878974914551, + "learning_rate": 0.00016862098507361875, + "loss": 0.7102, + "step": 16227 + }, + { + "epoch": 1.1260842412046355, + "grad_norm": 1.7437183856964111, + "learning_rate": 0.00016861689796401934, + "loss": 0.5852, + "step": 16228 + }, + { + "epoch": 1.126153632641732, + "grad_norm": 1.8872191905975342, + "learning_rate": 0.0001686128106378043, + "loss": 0.6603, + "step": 16229 + }, + { + "epoch": 1.1262230240788287, + "grad_norm": 1.7616132497787476, + "learning_rate": 0.0001686087230949866, + "loss": 0.7268, + "step": 16230 + }, + { + "epoch": 1.1262924155159253, + "grad_norm": 2.5985658168792725, + "learning_rate": 0.00016860463533557914, + "loss": 0.5639, + "step": 16231 + }, + { + "epoch": 1.126361806953022, + "grad_norm": 2.017631769180298, + "learning_rate": 0.00016860054735959477, + "loss": 0.6504, + "step": 16232 + }, + { + "epoch": 1.1264311983901187, + "grad_norm": 1.7671517133712769, + "learning_rate": 0.00016859645916704645, + "loss": 0.6376, + "step": 16233 + }, + { + "epoch": 1.1265005898272153, + "grad_norm": 1.836946725845337, + "learning_rate": 0.00016859237075794704, + "loss": 0.695, + "step": 16234 + }, + { + "epoch": 1.126569981264312, + "grad_norm": 2.3630692958831787, + "learning_rate": 0.00016858828213230952, + "loss": 0.6372, + "step": 16235 + }, + { + "epoch": 1.1266393727014086, + "grad_norm": 1.537699580192566, + "learning_rate": 0.00016858419329014672, + "loss": 0.6978, + "step": 16236 + }, + { + "epoch": 1.1267087641385054, + "grad_norm": 2.163849353790283, + "learning_rate": 0.0001685801042314716, + "loss": 0.6459, + "step": 16237 + }, + { + "epoch": 1.126778155575602, + "grad_norm": 1.7890338897705078, + "learning_rate": 0.00016857601495629703, + "loss": 0.5519, + "step": 16238 + }, + { + "epoch": 1.1268475470126986, + "grad_norm": 1.6828489303588867, + "learning_rate": 0.00016857192546463592, + "loss": 0.5972, + "step": 16239 + }, + { + "epoch": 1.1269169384497952, + "grad_norm": 1.9873003959655762, + "learning_rate": 0.00016856783575650122, + "loss": 0.715, + "step": 16240 + }, + { + "epoch": 1.126986329886892, + "grad_norm": 2.822967290878296, + "learning_rate": 0.00016856374583190583, + "loss": 0.8264, + "step": 16241 + }, + { + "epoch": 1.1270557213239887, + "grad_norm": 1.674769401550293, + "learning_rate": 0.00016855965569086262, + "loss": 0.6213, + "step": 16242 + }, + { + "epoch": 1.1271251127610853, + "grad_norm": 1.8594141006469727, + "learning_rate": 0.00016855556533338457, + "loss": 0.571, + "step": 16243 + }, + { + "epoch": 1.127194504198182, + "grad_norm": 1.977557897567749, + "learning_rate": 0.0001685514747594845, + "loss": 0.5956, + "step": 16244 + }, + { + "epoch": 1.1272638956352785, + "grad_norm": 2.330817699432373, + "learning_rate": 0.0001685473839691754, + "loss": 0.8694, + "step": 16245 + }, + { + "epoch": 1.1273332870723753, + "grad_norm": 1.3966647386550903, + "learning_rate": 0.00016854329296247018, + "loss": 0.6454, + "step": 16246 + }, + { + "epoch": 1.127402678509472, + "grad_norm": 2.2814910411834717, + "learning_rate": 0.00016853920173938173, + "loss": 0.6792, + "step": 16247 + }, + { + "epoch": 1.1274720699465686, + "grad_norm": 2.3131418228149414, + "learning_rate": 0.00016853511029992296, + "loss": 0.5279, + "step": 16248 + }, + { + "epoch": 1.1275414613836652, + "grad_norm": 1.8379440307617188, + "learning_rate": 0.00016853101864410682, + "loss": 0.5584, + "step": 16249 + }, + { + "epoch": 1.127610852820762, + "grad_norm": 1.9822604656219482, + "learning_rate": 0.0001685269267719462, + "loss": 0.5384, + "step": 16250 + }, + { + "epoch": 1.1276802442578586, + "grad_norm": 1.9731638431549072, + "learning_rate": 0.00016852283468345402, + "loss": 0.7053, + "step": 16251 + }, + { + "epoch": 1.1277496356949552, + "grad_norm": 1.9703953266143799, + "learning_rate": 0.00016851874237864317, + "loss": 0.6735, + "step": 16252 + }, + { + "epoch": 1.1278190271320518, + "grad_norm": 2.119300603866577, + "learning_rate": 0.00016851464985752666, + "loss": 0.7234, + "step": 16253 + }, + { + "epoch": 1.1278884185691487, + "grad_norm": 2.4654088020324707, + "learning_rate": 0.0001685105571201173, + "loss": 0.9063, + "step": 16254 + }, + { + "epoch": 1.1279578100062453, + "grad_norm": 2.812053918838501, + "learning_rate": 0.00016850646416642805, + "loss": 0.7367, + "step": 16255 + }, + { + "epoch": 1.128027201443342, + "grad_norm": 2.2233126163482666, + "learning_rate": 0.00016850237099647189, + "loss": 0.7482, + "step": 16256 + }, + { + "epoch": 1.1280965928804385, + "grad_norm": 2.3275976181030273, + "learning_rate": 0.00016849827761026165, + "loss": 0.6169, + "step": 16257 + }, + { + "epoch": 1.1281659843175351, + "grad_norm": 1.8768056631088257, + "learning_rate": 0.00016849418400781033, + "loss": 0.738, + "step": 16258 + }, + { + "epoch": 1.128235375754632, + "grad_norm": 1.369666337966919, + "learning_rate": 0.00016849009018913078, + "loss": 0.4961, + "step": 16259 + }, + { + "epoch": 1.1283047671917286, + "grad_norm": 1.7496124505996704, + "learning_rate": 0.000168485996154236, + "loss": 0.572, + "step": 16260 + }, + { + "epoch": 1.1283741586288252, + "grad_norm": 2.065788745880127, + "learning_rate": 0.00016848190190313887, + "loss": 0.6511, + "step": 16261 + }, + { + "epoch": 1.1284435500659218, + "grad_norm": 2.245260000228882, + "learning_rate": 0.0001684778074358523, + "loss": 0.6878, + "step": 16262 + }, + { + "epoch": 1.1285129415030186, + "grad_norm": 1.8733865022659302, + "learning_rate": 0.00016847371275238924, + "loss": 0.5512, + "step": 16263 + }, + { + "epoch": 1.1285823329401152, + "grad_norm": 2.1567330360412598, + "learning_rate": 0.00016846961785276263, + "loss": 0.5813, + "step": 16264 + }, + { + "epoch": 1.1286517243772118, + "grad_norm": 1.9223029613494873, + "learning_rate": 0.00016846552273698534, + "loss": 0.7654, + "step": 16265 + }, + { + "epoch": 1.1287211158143085, + "grad_norm": 1.5758064985275269, + "learning_rate": 0.0001684614274050704, + "loss": 0.6083, + "step": 16266 + }, + { + "epoch": 1.1287905072514053, + "grad_norm": 1.8701826333999634, + "learning_rate": 0.00016845733185703066, + "loss": 0.6574, + "step": 16267 + }, + { + "epoch": 1.128859898688502, + "grad_norm": 2.6216416358947754, + "learning_rate": 0.00016845323609287905, + "loss": 0.9673, + "step": 16268 + }, + { + "epoch": 1.1289292901255985, + "grad_norm": 2.3783211708068848, + "learning_rate": 0.0001684491401126285, + "loss": 0.7176, + "step": 16269 + }, + { + "epoch": 1.1289986815626951, + "grad_norm": 1.9291731119155884, + "learning_rate": 0.000168445043916292, + "loss": 0.7281, + "step": 16270 + }, + { + "epoch": 1.1290680729997917, + "grad_norm": 1.8955830335617065, + "learning_rate": 0.0001684409475038824, + "loss": 0.5522, + "step": 16271 + }, + { + "epoch": 1.1291374644368886, + "grad_norm": 1.7363824844360352, + "learning_rate": 0.00016843685087541266, + "loss": 0.5734, + "step": 16272 + }, + { + "epoch": 1.1292068558739852, + "grad_norm": 2.5241267681121826, + "learning_rate": 0.00016843275403089574, + "loss": 0.7747, + "step": 16273 + }, + { + "epoch": 1.1292762473110818, + "grad_norm": 2.523451566696167, + "learning_rate": 0.00016842865697034456, + "loss": 0.8618, + "step": 16274 + }, + { + "epoch": 1.1293456387481784, + "grad_norm": 1.7017823457717896, + "learning_rate": 0.00016842455969377204, + "loss": 0.5516, + "step": 16275 + }, + { + "epoch": 1.1294150301852752, + "grad_norm": 1.3065086603164673, + "learning_rate": 0.00016842046220119113, + "loss": 0.4619, + "step": 16276 + }, + { + "epoch": 1.1294844216223718, + "grad_norm": 1.3718421459197998, + "learning_rate": 0.00016841636449261476, + "loss": 0.6145, + "step": 16277 + }, + { + "epoch": 1.1295538130594684, + "grad_norm": 1.9883484840393066, + "learning_rate": 0.00016841226656805588, + "loss": 0.5483, + "step": 16278 + }, + { + "epoch": 1.129623204496565, + "grad_norm": 1.7690441608428955, + "learning_rate": 0.00016840816842752737, + "loss": 0.5226, + "step": 16279 + }, + { + "epoch": 1.129692595933662, + "grad_norm": 2.450174331665039, + "learning_rate": 0.00016840407007104224, + "loss": 0.7702, + "step": 16280 + }, + { + "epoch": 1.1297619873707585, + "grad_norm": 2.12300968170166, + "learning_rate": 0.0001683999714986134, + "loss": 0.7183, + "step": 16281 + }, + { + "epoch": 1.1298313788078551, + "grad_norm": 2.570847988128662, + "learning_rate": 0.00016839587271025376, + "loss": 0.6999, + "step": 16282 + }, + { + "epoch": 1.1299007702449517, + "grad_norm": 2.269545078277588, + "learning_rate": 0.00016839177370597632, + "loss": 0.8742, + "step": 16283 + }, + { + "epoch": 1.1299701616820483, + "grad_norm": 1.5137521028518677, + "learning_rate": 0.00016838767448579394, + "loss": 0.8182, + "step": 16284 + }, + { + "epoch": 1.1300395531191452, + "grad_norm": 1.8577079772949219, + "learning_rate": 0.00016838357504971968, + "loss": 0.7084, + "step": 16285 + }, + { + "epoch": 1.1301089445562418, + "grad_norm": 2.3174166679382324, + "learning_rate": 0.00016837947539776634, + "loss": 0.6772, + "step": 16286 + }, + { + "epoch": 1.1301783359933384, + "grad_norm": 2.0139265060424805, + "learning_rate": 0.00016837537552994694, + "loss": 0.5584, + "step": 16287 + }, + { + "epoch": 1.130247727430435, + "grad_norm": 1.8869402408599854, + "learning_rate": 0.00016837127544627444, + "loss": 0.5595, + "step": 16288 + }, + { + "epoch": 1.1303171188675318, + "grad_norm": 1.9441745281219482, + "learning_rate": 0.00016836717514676175, + "loss": 0.7381, + "step": 16289 + }, + { + "epoch": 1.1303865103046284, + "grad_norm": 1.7683318853378296, + "learning_rate": 0.0001683630746314218, + "loss": 0.619, + "step": 16290 + }, + { + "epoch": 1.130455901741725, + "grad_norm": 2.2215025424957275, + "learning_rate": 0.00016835897390026755, + "loss": 0.6041, + "step": 16291 + }, + { + "epoch": 1.1305252931788217, + "grad_norm": 2.1484944820404053, + "learning_rate": 0.000168354872953312, + "loss": 0.9784, + "step": 16292 + }, + { + "epoch": 1.1305946846159185, + "grad_norm": 2.6089694499969482, + "learning_rate": 0.000168350771790568, + "loss": 0.8165, + "step": 16293 + }, + { + "epoch": 1.130664076053015, + "grad_norm": 3.0498483180999756, + "learning_rate": 0.0001683466704120485, + "loss": 0.9198, + "step": 16294 + }, + { + "epoch": 1.1307334674901117, + "grad_norm": 1.9712141752243042, + "learning_rate": 0.0001683425688177666, + "loss": 0.5932, + "step": 16295 + }, + { + "epoch": 1.1308028589272083, + "grad_norm": 1.9999655485153198, + "learning_rate": 0.00016833846700773508, + "loss": 0.7362, + "step": 16296 + }, + { + "epoch": 1.130872250364305, + "grad_norm": 1.8977538347244263, + "learning_rate": 0.00016833436498196695, + "loss": 0.5634, + "step": 16297 + }, + { + "epoch": 1.1309416418014018, + "grad_norm": 1.9902783632278442, + "learning_rate": 0.00016833026274047517, + "loss": 0.7708, + "step": 16298 + }, + { + "epoch": 1.1310110332384984, + "grad_norm": 2.4889841079711914, + "learning_rate": 0.00016832616028327266, + "loss": 0.8033, + "step": 16299 + }, + { + "epoch": 1.131080424675595, + "grad_norm": 1.4809482097625732, + "learning_rate": 0.00016832205761037238, + "loss": 0.4981, + "step": 16300 + }, + { + "epoch": 1.1311498161126916, + "grad_norm": 2.10235595703125, + "learning_rate": 0.00016831795472178733, + "loss": 0.8182, + "step": 16301 + }, + { + "epoch": 1.1312192075497884, + "grad_norm": 2.8298892974853516, + "learning_rate": 0.00016831385161753041, + "loss": 1.0679, + "step": 16302 + }, + { + "epoch": 1.131288598986885, + "grad_norm": 1.6857722997665405, + "learning_rate": 0.00016830974829761458, + "loss": 0.6253, + "step": 16303 + }, + { + "epoch": 1.1313579904239817, + "grad_norm": 1.7542273998260498, + "learning_rate": 0.0001683056447620528, + "loss": 0.7574, + "step": 16304 + }, + { + "epoch": 1.1314273818610783, + "grad_norm": 2.185145616531372, + "learning_rate": 0.00016830154101085803, + "loss": 0.769, + "step": 16305 + }, + { + "epoch": 1.131496773298175, + "grad_norm": 2.0227415561676025, + "learning_rate": 0.0001682974370440432, + "loss": 0.6755, + "step": 16306 + }, + { + "epoch": 1.1315661647352717, + "grad_norm": 1.429578185081482, + "learning_rate": 0.00016829333286162132, + "loss": 0.4537, + "step": 16307 + }, + { + "epoch": 1.1316355561723683, + "grad_norm": 1.4098145961761475, + "learning_rate": 0.00016828922846360528, + "loss": 0.4996, + "step": 16308 + }, + { + "epoch": 1.131704947609465, + "grad_norm": 1.6655774116516113, + "learning_rate": 0.0001682851238500081, + "loss": 0.6541, + "step": 16309 + }, + { + "epoch": 1.1317743390465616, + "grad_norm": 1.5248830318450928, + "learning_rate": 0.00016828101902084268, + "loss": 0.6349, + "step": 16310 + }, + { + "epoch": 1.1318437304836584, + "grad_norm": 1.3597911596298218, + "learning_rate": 0.000168276913976122, + "loss": 0.5831, + "step": 16311 + }, + { + "epoch": 1.131913121920755, + "grad_norm": 1.8821437358856201, + "learning_rate": 0.00016827280871585905, + "loss": 0.6987, + "step": 16312 + }, + { + "epoch": 1.1319825133578516, + "grad_norm": 1.6311599016189575, + "learning_rate": 0.00016826870324006675, + "loss": 0.6017, + "step": 16313 + }, + { + "epoch": 1.1320519047949482, + "grad_norm": 1.8758409023284912, + "learning_rate": 0.00016826459754875806, + "loss": 0.6567, + "step": 16314 + }, + { + "epoch": 1.132121296232045, + "grad_norm": 1.8635863065719604, + "learning_rate": 0.000168260491641946, + "loss": 0.5566, + "step": 16315 + }, + { + "epoch": 1.1321906876691417, + "grad_norm": 1.6913717985153198, + "learning_rate": 0.00016825638551964343, + "loss": 0.5772, + "step": 16316 + }, + { + "epoch": 1.1322600791062383, + "grad_norm": 1.581079363822937, + "learning_rate": 0.0001682522791818634, + "loss": 0.5415, + "step": 16317 + }, + { + "epoch": 1.1323294705433349, + "grad_norm": 1.776729941368103, + "learning_rate": 0.00016824817262861883, + "loss": 0.7017, + "step": 16318 + }, + { + "epoch": 1.1323988619804317, + "grad_norm": 1.9132788181304932, + "learning_rate": 0.00016824406585992275, + "loss": 0.52, + "step": 16319 + }, + { + "epoch": 1.1324682534175283, + "grad_norm": 1.6789048910140991, + "learning_rate": 0.000168239958875788, + "loss": 0.4626, + "step": 16320 + }, + { + "epoch": 1.132537644854625, + "grad_norm": 1.791770100593567, + "learning_rate": 0.00016823585167622765, + "loss": 0.7933, + "step": 16321 + }, + { + "epoch": 1.1326070362917215, + "grad_norm": 1.4943872690200806, + "learning_rate": 0.00016823174426125465, + "loss": 0.654, + "step": 16322 + }, + { + "epoch": 1.1326764277288182, + "grad_norm": 1.7471963167190552, + "learning_rate": 0.00016822763663088192, + "loss": 0.7017, + "step": 16323 + }, + { + "epoch": 1.132745819165915, + "grad_norm": 2.191129684448242, + "learning_rate": 0.00016822352878512247, + "loss": 0.6466, + "step": 16324 + }, + { + "epoch": 1.1328152106030116, + "grad_norm": 1.2819106578826904, + "learning_rate": 0.00016821942072398925, + "loss": 0.4942, + "step": 16325 + }, + { + "epoch": 1.1328846020401082, + "grad_norm": 1.74800705909729, + "learning_rate": 0.00016821531244749522, + "loss": 0.7607, + "step": 16326 + }, + { + "epoch": 1.1329539934772048, + "grad_norm": 1.763747215270996, + "learning_rate": 0.00016821120395565337, + "loss": 0.5815, + "step": 16327 + }, + { + "epoch": 1.1330233849143017, + "grad_norm": 1.4199581146240234, + "learning_rate": 0.0001682070952484767, + "loss": 0.5952, + "step": 16328 + }, + { + "epoch": 1.1330927763513983, + "grad_norm": 1.7137014865875244, + "learning_rate": 0.0001682029863259781, + "loss": 0.6783, + "step": 16329 + }, + { + "epoch": 1.1331621677884949, + "grad_norm": 1.9540414810180664, + "learning_rate": 0.00016819887718817062, + "loss": 0.5072, + "step": 16330 + }, + { + "epoch": 1.1332315592255915, + "grad_norm": 1.7422691583633423, + "learning_rate": 0.00016819476783506718, + "loss": 0.604, + "step": 16331 + }, + { + "epoch": 1.1333009506626883, + "grad_norm": 1.8358328342437744, + "learning_rate": 0.00016819065826668075, + "loss": 0.5423, + "step": 16332 + }, + { + "epoch": 1.133370342099785, + "grad_norm": 1.7938401699066162, + "learning_rate": 0.00016818654848302434, + "loss": 0.7364, + "step": 16333 + }, + { + "epoch": 1.1334397335368815, + "grad_norm": 1.3518421649932861, + "learning_rate": 0.0001681824384841109, + "loss": 0.5455, + "step": 16334 + }, + { + "epoch": 1.1335091249739782, + "grad_norm": 2.3008062839508057, + "learning_rate": 0.00016817832826995343, + "loss": 0.7837, + "step": 16335 + }, + { + "epoch": 1.1335785164110748, + "grad_norm": 2.2185347080230713, + "learning_rate": 0.0001681742178405649, + "loss": 0.5984, + "step": 16336 + }, + { + "epoch": 1.1336479078481716, + "grad_norm": 1.534512996673584, + "learning_rate": 0.00016817010719595823, + "loss": 0.6133, + "step": 16337 + }, + { + "epoch": 1.1337172992852682, + "grad_norm": 1.6577250957489014, + "learning_rate": 0.00016816599633614647, + "loss": 0.8216, + "step": 16338 + }, + { + "epoch": 1.1337866907223648, + "grad_norm": 2.1562297344207764, + "learning_rate": 0.00016816188526114255, + "loss": 0.8667, + "step": 16339 + }, + { + "epoch": 1.1338560821594614, + "grad_norm": 2.494042158126831, + "learning_rate": 0.00016815777397095948, + "loss": 0.9036, + "step": 16340 + }, + { + "epoch": 1.1339254735965583, + "grad_norm": 2.4097096920013428, + "learning_rate": 0.00016815366246561024, + "loss": 1.1298, + "step": 16341 + }, + { + "epoch": 1.1339948650336549, + "grad_norm": 1.834869146347046, + "learning_rate": 0.00016814955074510776, + "loss": 0.4758, + "step": 16342 + }, + { + "epoch": 1.1340642564707515, + "grad_norm": 1.5941787958145142, + "learning_rate": 0.0001681454388094651, + "loss": 0.6454, + "step": 16343 + }, + { + "epoch": 1.134133647907848, + "grad_norm": 1.7598243951797485, + "learning_rate": 0.00016814132665869517, + "loss": 0.7189, + "step": 16344 + }, + { + "epoch": 1.134203039344945, + "grad_norm": 1.9980984926223755, + "learning_rate": 0.00016813721429281096, + "loss": 0.4406, + "step": 16345 + }, + { + "epoch": 1.1342724307820415, + "grad_norm": 1.4983344078063965, + "learning_rate": 0.0001681331017118255, + "loss": 0.6581, + "step": 16346 + }, + { + "epoch": 1.1343418222191382, + "grad_norm": 1.7987719774246216, + "learning_rate": 0.00016812898891575175, + "loss": 0.6067, + "step": 16347 + }, + { + "epoch": 1.1344112136562348, + "grad_norm": 2.1991028785705566, + "learning_rate": 0.00016812487590460266, + "loss": 0.9903, + "step": 16348 + }, + { + "epoch": 1.1344806050933314, + "grad_norm": 2.0475902557373047, + "learning_rate": 0.00016812076267839124, + "loss": 0.7753, + "step": 16349 + }, + { + "epoch": 1.1345499965304282, + "grad_norm": 1.4345885515213013, + "learning_rate": 0.00016811664923713047, + "loss": 0.5321, + "step": 16350 + }, + { + "epoch": 1.1346193879675248, + "grad_norm": 2.59226393699646, + "learning_rate": 0.00016811253558083336, + "loss": 0.5109, + "step": 16351 + }, + { + "epoch": 1.1346887794046214, + "grad_norm": 1.7531583309173584, + "learning_rate": 0.0001681084217095129, + "loss": 0.6271, + "step": 16352 + }, + { + "epoch": 1.134758170841718, + "grad_norm": 1.722711443901062, + "learning_rate": 0.000168104307623182, + "loss": 0.7301, + "step": 16353 + }, + { + "epoch": 1.1348275622788149, + "grad_norm": 1.5200080871582031, + "learning_rate": 0.00016810019332185374, + "loss": 0.713, + "step": 16354 + }, + { + "epoch": 1.1348969537159115, + "grad_norm": 1.8085145950317383, + "learning_rate": 0.00016809607880554107, + "loss": 0.5475, + "step": 16355 + }, + { + "epoch": 1.134966345153008, + "grad_norm": 2.168055295944214, + "learning_rate": 0.00016809196407425696, + "loss": 0.7064, + "step": 16356 + }, + { + "epoch": 1.1350357365901047, + "grad_norm": 2.023250102996826, + "learning_rate": 0.00016808784912801443, + "loss": 0.7765, + "step": 16357 + }, + { + "epoch": 1.1351051280272015, + "grad_norm": 1.661971926689148, + "learning_rate": 0.00016808373396682647, + "loss": 0.5921, + "step": 16358 + }, + { + "epoch": 1.1351745194642981, + "grad_norm": 1.7539864778518677, + "learning_rate": 0.00016807961859070604, + "loss": 0.6992, + "step": 16359 + }, + { + "epoch": 1.1352439109013948, + "grad_norm": 2.4736404418945312, + "learning_rate": 0.00016807550299966618, + "loss": 0.7828, + "step": 16360 + }, + { + "epoch": 1.1353133023384914, + "grad_norm": 1.419204592704773, + "learning_rate": 0.00016807138719371982, + "loss": 0.526, + "step": 16361 + }, + { + "epoch": 1.135382693775588, + "grad_norm": 2.2971136569976807, + "learning_rate": 0.00016806727117287998, + "loss": 0.7727, + "step": 16362 + }, + { + "epoch": 1.1354520852126848, + "grad_norm": 1.9571195840835571, + "learning_rate": 0.0001680631549371597, + "loss": 0.7158, + "step": 16363 + }, + { + "epoch": 1.1355214766497814, + "grad_norm": 1.6646913290023804, + "learning_rate": 0.00016805903848657195, + "loss": 0.6554, + "step": 16364 + }, + { + "epoch": 1.135590868086878, + "grad_norm": 2.077401638031006, + "learning_rate": 0.00016805492182112968, + "loss": 0.7754, + "step": 16365 + }, + { + "epoch": 1.1356602595239746, + "grad_norm": 2.52003812789917, + "learning_rate": 0.00016805080494084593, + "loss": 0.6729, + "step": 16366 + }, + { + "epoch": 1.1357296509610715, + "grad_norm": 2.7381229400634766, + "learning_rate": 0.00016804668784573367, + "loss": 0.8686, + "step": 16367 + }, + { + "epoch": 1.135799042398168, + "grad_norm": 2.815504550933838, + "learning_rate": 0.0001680425705358059, + "loss": 1.097, + "step": 16368 + }, + { + "epoch": 1.1358684338352647, + "grad_norm": 3.8126542568206787, + "learning_rate": 0.0001680384530110757, + "loss": 0.8486, + "step": 16369 + }, + { + "epoch": 1.1359378252723613, + "grad_norm": 2.1816320419311523, + "learning_rate": 0.00016803433527155592, + "loss": 0.72, + "step": 16370 + }, + { + "epoch": 1.1360072167094581, + "grad_norm": 2.407033681869507, + "learning_rate": 0.00016803021731725968, + "loss": 0.5955, + "step": 16371 + }, + { + "epoch": 1.1360766081465548, + "grad_norm": 2.0348997116088867, + "learning_rate": 0.00016802609914819992, + "loss": 0.7362, + "step": 16372 + }, + { + "epoch": 1.1361459995836514, + "grad_norm": 2.0347557067871094, + "learning_rate": 0.00016802198076438965, + "loss": 0.6595, + "step": 16373 + }, + { + "epoch": 1.136215391020748, + "grad_norm": 1.741895079612732, + "learning_rate": 0.0001680178621658419, + "loss": 0.6556, + "step": 16374 + }, + { + "epoch": 1.1362847824578446, + "grad_norm": 2.058698892593384, + "learning_rate": 0.00016801374335256964, + "loss": 0.7388, + "step": 16375 + }, + { + "epoch": 1.1363541738949414, + "grad_norm": 2.143956422805786, + "learning_rate": 0.00016800962432458587, + "loss": 0.6576, + "step": 16376 + }, + { + "epoch": 1.136423565332038, + "grad_norm": 2.3339104652404785, + "learning_rate": 0.0001680055050819036, + "loss": 0.5922, + "step": 16377 + }, + { + "epoch": 1.1364929567691346, + "grad_norm": 1.9906885623931885, + "learning_rate": 0.0001680013856245359, + "loss": 0.7259, + "step": 16378 + }, + { + "epoch": 1.1365623482062313, + "grad_norm": 2.0701918601989746, + "learning_rate": 0.00016799726595249565, + "loss": 0.7729, + "step": 16379 + }, + { + "epoch": 1.136631739643328, + "grad_norm": 1.8341646194458008, + "learning_rate": 0.00016799314606579595, + "loss": 0.5147, + "step": 16380 + }, + { + "epoch": 1.1367011310804247, + "grad_norm": 2.0595760345458984, + "learning_rate": 0.00016798902596444978, + "loss": 0.7662, + "step": 16381 + }, + { + "epoch": 1.1367705225175213, + "grad_norm": 2.1461665630340576, + "learning_rate": 0.00016798490564847013, + "loss": 0.547, + "step": 16382 + }, + { + "epoch": 1.136839913954618, + "grad_norm": 2.5423390865325928, + "learning_rate": 0.00016798078511787002, + "loss": 0.7653, + "step": 16383 + }, + { + "epoch": 1.1369093053917148, + "grad_norm": 1.222622036933899, + "learning_rate": 0.00016797666437266248, + "loss": 0.4474, + "step": 16384 + }, + { + "epoch": 1.1369786968288114, + "grad_norm": 2.1692516803741455, + "learning_rate": 0.00016797254341286046, + "loss": 0.7615, + "step": 16385 + }, + { + "epoch": 1.137048088265908, + "grad_norm": 1.5583223104476929, + "learning_rate": 0.000167968422238477, + "loss": 0.5449, + "step": 16386 + }, + { + "epoch": 1.1371174797030046, + "grad_norm": 2.5249314308166504, + "learning_rate": 0.00016796430084952516, + "loss": 0.6277, + "step": 16387 + }, + { + "epoch": 1.1371868711401012, + "grad_norm": 2.4133682250976562, + "learning_rate": 0.00016796017924601792, + "loss": 0.9248, + "step": 16388 + }, + { + "epoch": 1.137256262577198, + "grad_norm": 2.4995598793029785, + "learning_rate": 0.00016795605742796825, + "loss": 0.6538, + "step": 16389 + }, + { + "epoch": 1.1373256540142946, + "grad_norm": 2.339700937271118, + "learning_rate": 0.0001679519353953892, + "loss": 0.7436, + "step": 16390 + }, + { + "epoch": 1.1373950454513913, + "grad_norm": 1.9084327220916748, + "learning_rate": 0.00016794781314829376, + "loss": 0.6348, + "step": 16391 + }, + { + "epoch": 1.137464436888488, + "grad_norm": 1.7385905981063843, + "learning_rate": 0.00016794369068669495, + "loss": 0.4959, + "step": 16392 + }, + { + "epoch": 1.1375338283255847, + "grad_norm": 6.689008712768555, + "learning_rate": 0.00016793956801060581, + "loss": 0.6356, + "step": 16393 + }, + { + "epoch": 1.1376032197626813, + "grad_norm": 1.802636742591858, + "learning_rate": 0.00016793544512003932, + "loss": 0.495, + "step": 16394 + }, + { + "epoch": 1.137672611199778, + "grad_norm": 2.2997891902923584, + "learning_rate": 0.00016793132201500853, + "loss": 0.811, + "step": 16395 + }, + { + "epoch": 1.1377420026368745, + "grad_norm": 1.6735559701919556, + "learning_rate": 0.00016792719869552643, + "loss": 0.5615, + "step": 16396 + }, + { + "epoch": 1.1378113940739714, + "grad_norm": 3.04801607131958, + "learning_rate": 0.00016792307516160605, + "loss": 0.8937, + "step": 16397 + }, + { + "epoch": 1.137880785511068, + "grad_norm": 1.659971833229065, + "learning_rate": 0.0001679189514132604, + "loss": 0.7144, + "step": 16398 + }, + { + "epoch": 1.1379501769481646, + "grad_norm": 2.7853996753692627, + "learning_rate": 0.0001679148274505025, + "loss": 0.78, + "step": 16399 + }, + { + "epoch": 1.1380195683852612, + "grad_norm": 2.2941367626190186, + "learning_rate": 0.00016791070327334536, + "loss": 0.7609, + "step": 16400 + }, + { + "epoch": 1.1380889598223578, + "grad_norm": 1.6557185649871826, + "learning_rate": 0.000167906578881802, + "loss": 0.4298, + "step": 16401 + }, + { + "epoch": 1.1381583512594546, + "grad_norm": 1.8315820693969727, + "learning_rate": 0.00016790245427588548, + "loss": 0.7489, + "step": 16402 + }, + { + "epoch": 1.1382277426965512, + "grad_norm": 2.6128716468811035, + "learning_rate": 0.00016789832945560877, + "loss": 0.678, + "step": 16403 + }, + { + "epoch": 1.1382971341336479, + "grad_norm": 1.51139235496521, + "learning_rate": 0.00016789420442098494, + "loss": 0.4919, + "step": 16404 + }, + { + "epoch": 1.1383665255707447, + "grad_norm": 2.512213945388794, + "learning_rate": 0.00016789007917202695, + "loss": 0.7656, + "step": 16405 + }, + { + "epoch": 1.1384359170078413, + "grad_norm": 2.0302793979644775, + "learning_rate": 0.00016788595370874788, + "loss": 0.6294, + "step": 16406 + }, + { + "epoch": 1.138505308444938, + "grad_norm": 1.8607561588287354, + "learning_rate": 0.00016788182803116071, + "loss": 0.5218, + "step": 16407 + }, + { + "epoch": 1.1385746998820345, + "grad_norm": 2.2659666538238525, + "learning_rate": 0.00016787770213927847, + "loss": 0.8094, + "step": 16408 + }, + { + "epoch": 1.1386440913191311, + "grad_norm": 2.4550068378448486, + "learning_rate": 0.00016787357603311424, + "loss": 0.4895, + "step": 16409 + }, + { + "epoch": 1.138713482756228, + "grad_norm": 2.4328346252441406, + "learning_rate": 0.00016786944971268098, + "loss": 0.7724, + "step": 16410 + }, + { + "epoch": 1.1387828741933246, + "grad_norm": 2.411447763442993, + "learning_rate": 0.00016786532317799175, + "loss": 0.7021, + "step": 16411 + }, + { + "epoch": 1.1388522656304212, + "grad_norm": 2.0920395851135254, + "learning_rate": 0.0001678611964290596, + "loss": 0.9176, + "step": 16412 + }, + { + "epoch": 1.1389216570675178, + "grad_norm": 2.197514533996582, + "learning_rate": 0.00016785706946589746, + "loss": 0.7472, + "step": 16413 + }, + { + "epoch": 1.1389910485046144, + "grad_norm": 2.0631678104400635, + "learning_rate": 0.00016785294228851847, + "loss": 1.0354, + "step": 16414 + }, + { + "epoch": 1.1390604399417112, + "grad_norm": 1.8229464292526245, + "learning_rate": 0.00016784881489693558, + "loss": 0.8127, + "step": 16415 + }, + { + "epoch": 1.1391298313788079, + "grad_norm": 2.327402114868164, + "learning_rate": 0.0001678446872911619, + "loss": 0.5708, + "step": 16416 + }, + { + "epoch": 1.1391992228159045, + "grad_norm": 2.308260440826416, + "learning_rate": 0.00016784055947121037, + "loss": 0.7334, + "step": 16417 + }, + { + "epoch": 1.1392686142530013, + "grad_norm": 2.3625288009643555, + "learning_rate": 0.00016783643143709412, + "loss": 0.7829, + "step": 16418 + }, + { + "epoch": 1.139338005690098, + "grad_norm": 2.572782039642334, + "learning_rate": 0.00016783230318882607, + "loss": 0.6737, + "step": 16419 + }, + { + "epoch": 1.1394073971271945, + "grad_norm": 2.9816675186157227, + "learning_rate": 0.00016782817472641933, + "loss": 0.9782, + "step": 16420 + }, + { + "epoch": 1.1394767885642911, + "grad_norm": 2.9748504161834717, + "learning_rate": 0.0001678240460498869, + "loss": 0.6914, + "step": 16421 + }, + { + "epoch": 1.1395461800013877, + "grad_norm": 3.2469494342803955, + "learning_rate": 0.00016781991715924182, + "loss": 0.5791, + "step": 16422 + }, + { + "epoch": 1.1396155714384846, + "grad_norm": 3.3200528621673584, + "learning_rate": 0.00016781578805449718, + "loss": 0.727, + "step": 16423 + }, + { + "epoch": 1.1396849628755812, + "grad_norm": 2.009842872619629, + "learning_rate": 0.00016781165873566592, + "loss": 0.6267, + "step": 16424 + }, + { + "epoch": 1.1397543543126778, + "grad_norm": 2.088273048400879, + "learning_rate": 0.00016780752920276114, + "loss": 0.6379, + "step": 16425 + }, + { + "epoch": 1.1398237457497744, + "grad_norm": 2.218196153640747, + "learning_rate": 0.00016780339945579584, + "loss": 0.5933, + "step": 16426 + }, + { + "epoch": 1.139893137186871, + "grad_norm": 2.1410975456237793, + "learning_rate": 0.0001677992694947831, + "loss": 0.7414, + "step": 16427 + }, + { + "epoch": 1.1399625286239679, + "grad_norm": 2.70835018157959, + "learning_rate": 0.00016779513931973588, + "loss": 0.7185, + "step": 16428 + }, + { + "epoch": 1.1400319200610645, + "grad_norm": 2.429763078689575, + "learning_rate": 0.00016779100893066736, + "loss": 0.7972, + "step": 16429 + }, + { + "epoch": 1.140101311498161, + "grad_norm": 2.455477237701416, + "learning_rate": 0.0001677868783275904, + "loss": 0.8632, + "step": 16430 + }, + { + "epoch": 1.140170702935258, + "grad_norm": 3.942742109298706, + "learning_rate": 0.00016778274751051817, + "loss": 0.8266, + "step": 16431 + }, + { + "epoch": 1.1402400943723545, + "grad_norm": 2.5811829566955566, + "learning_rate": 0.00016777861647946366, + "loss": 0.7365, + "step": 16432 + }, + { + "epoch": 1.1403094858094511, + "grad_norm": 2.405597686767578, + "learning_rate": 0.00016777448523443996, + "loss": 0.8139, + "step": 16433 + }, + { + "epoch": 1.1403788772465477, + "grad_norm": 1.9870529174804688, + "learning_rate": 0.00016777035377546005, + "loss": 0.6889, + "step": 16434 + }, + { + "epoch": 1.1404482686836444, + "grad_norm": 1.7941210269927979, + "learning_rate": 0.00016776622210253698, + "loss": 0.6838, + "step": 16435 + }, + { + "epoch": 1.1405176601207412, + "grad_norm": 1.8133580684661865, + "learning_rate": 0.00016776209021568384, + "loss": 0.6088, + "step": 16436 + }, + { + "epoch": 1.1405870515578378, + "grad_norm": 2.6543943881988525, + "learning_rate": 0.0001677579581149136, + "loss": 0.833, + "step": 16437 + }, + { + "epoch": 1.1406564429949344, + "grad_norm": 1.7923303842544556, + "learning_rate": 0.00016775382580023934, + "loss": 0.6643, + "step": 16438 + }, + { + "epoch": 1.140725834432031, + "grad_norm": 2.279534101486206, + "learning_rate": 0.00016774969327167417, + "loss": 0.7221, + "step": 16439 + }, + { + "epoch": 1.1407952258691276, + "grad_norm": 3.2281079292297363, + "learning_rate": 0.00016774556052923104, + "loss": 0.7568, + "step": 16440 + }, + { + "epoch": 1.1408646173062245, + "grad_norm": 2.588592052459717, + "learning_rate": 0.00016774142757292304, + "loss": 0.8444, + "step": 16441 + }, + { + "epoch": 1.140934008743321, + "grad_norm": 1.4433289766311646, + "learning_rate": 0.00016773729440276323, + "loss": 0.6401, + "step": 16442 + }, + { + "epoch": 1.1410034001804177, + "grad_norm": 1.8271474838256836, + "learning_rate": 0.00016773316101876463, + "loss": 0.7801, + "step": 16443 + }, + { + "epoch": 1.1410727916175145, + "grad_norm": 1.4343467950820923, + "learning_rate": 0.0001677290274209403, + "loss": 0.4635, + "step": 16444 + }, + { + "epoch": 1.1411421830546111, + "grad_norm": 2.9838907718658447, + "learning_rate": 0.0001677248936093033, + "loss": 0.5426, + "step": 16445 + }, + { + "epoch": 1.1412115744917077, + "grad_norm": 2.07942795753479, + "learning_rate": 0.00016772075958386662, + "loss": 0.7154, + "step": 16446 + }, + { + "epoch": 1.1412809659288043, + "grad_norm": 2.902805805206299, + "learning_rate": 0.00016771662534464344, + "loss": 0.9327, + "step": 16447 + }, + { + "epoch": 1.141350357365901, + "grad_norm": 1.6328554153442383, + "learning_rate": 0.00016771249089164668, + "loss": 0.5328, + "step": 16448 + }, + { + "epoch": 1.1414197488029978, + "grad_norm": 1.927883505821228, + "learning_rate": 0.00016770835622488943, + "loss": 0.6572, + "step": 16449 + }, + { + "epoch": 1.1414891402400944, + "grad_norm": 1.7890006303787231, + "learning_rate": 0.00016770422134438476, + "loss": 0.6736, + "step": 16450 + }, + { + "epoch": 1.141558531677191, + "grad_norm": 1.751218318939209, + "learning_rate": 0.00016770008625014576, + "loss": 0.692, + "step": 16451 + }, + { + "epoch": 1.1416279231142876, + "grad_norm": 2.8562333583831787, + "learning_rate": 0.0001676959509421854, + "loss": 0.9023, + "step": 16452 + }, + { + "epoch": 1.1416973145513842, + "grad_norm": 4.0157999992370605, + "learning_rate": 0.00016769181542051678, + "loss": 0.6403, + "step": 16453 + }, + { + "epoch": 1.141766705988481, + "grad_norm": 2.9055063724517822, + "learning_rate": 0.00016768767968515296, + "loss": 0.7438, + "step": 16454 + }, + { + "epoch": 1.1418360974255777, + "grad_norm": 1.5268933773040771, + "learning_rate": 0.000167683543736107, + "loss": 0.5659, + "step": 16455 + }, + { + "epoch": 1.1419054888626743, + "grad_norm": 1.84392511844635, + "learning_rate": 0.00016767940757339193, + "loss": 0.8728, + "step": 16456 + }, + { + "epoch": 1.1419748802997711, + "grad_norm": 2.357015609741211, + "learning_rate": 0.0001676752711970208, + "loss": 1.1026, + "step": 16457 + }, + { + "epoch": 1.1420442717368677, + "grad_norm": 2.170503854751587, + "learning_rate": 0.00016767113460700674, + "loss": 0.4355, + "step": 16458 + }, + { + "epoch": 1.1421136631739643, + "grad_norm": 1.5732756853103638, + "learning_rate": 0.00016766699780336274, + "loss": 0.5228, + "step": 16459 + }, + { + "epoch": 1.142183054611061, + "grad_norm": 1.7296419143676758, + "learning_rate": 0.00016766286078610188, + "loss": 0.7814, + "step": 16460 + }, + { + "epoch": 1.1422524460481576, + "grad_norm": 2.0505294799804688, + "learning_rate": 0.0001676587235552372, + "loss": 0.6808, + "step": 16461 + }, + { + "epoch": 1.1423218374852544, + "grad_norm": 1.7303022146224976, + "learning_rate": 0.0001676545861107818, + "loss": 0.7823, + "step": 16462 + }, + { + "epoch": 1.142391228922351, + "grad_norm": 2.9153382778167725, + "learning_rate": 0.00016765044845274871, + "loss": 0.7801, + "step": 16463 + }, + { + "epoch": 1.1424606203594476, + "grad_norm": 1.4372285604476929, + "learning_rate": 0.00016764631058115102, + "loss": 0.7584, + "step": 16464 + }, + { + "epoch": 1.1425300117965442, + "grad_norm": 1.5460319519042969, + "learning_rate": 0.0001676421724960018, + "loss": 0.6889, + "step": 16465 + }, + { + "epoch": 1.142599403233641, + "grad_norm": 1.872855544090271, + "learning_rate": 0.00016763803419731402, + "loss": 0.6783, + "step": 16466 + }, + { + "epoch": 1.1426687946707377, + "grad_norm": 1.7813382148742676, + "learning_rate": 0.00016763389568510086, + "loss": 0.5509, + "step": 16467 + }, + { + "epoch": 1.1427381861078343, + "grad_norm": 1.9883904457092285, + "learning_rate": 0.00016762975695937534, + "loss": 0.6636, + "step": 16468 + }, + { + "epoch": 1.142807577544931, + "grad_norm": 2.1651172637939453, + "learning_rate": 0.00016762561802015049, + "loss": 1.0227, + "step": 16469 + }, + { + "epoch": 1.1428769689820277, + "grad_norm": 1.8861671686172485, + "learning_rate": 0.00016762147886743942, + "loss": 0.7539, + "step": 16470 + }, + { + "epoch": 1.1429463604191243, + "grad_norm": 2.0454819202423096, + "learning_rate": 0.0001676173395012552, + "loss": 0.8406, + "step": 16471 + }, + { + "epoch": 1.143015751856221, + "grad_norm": 1.7301772832870483, + "learning_rate": 0.0001676131999216109, + "loss": 0.711, + "step": 16472 + }, + { + "epoch": 1.1430851432933176, + "grad_norm": 1.4250487089157104, + "learning_rate": 0.00016760906012851953, + "loss": 0.6352, + "step": 16473 + }, + { + "epoch": 1.1431545347304142, + "grad_norm": 1.6107869148254395, + "learning_rate": 0.00016760492012199424, + "loss": 0.7483, + "step": 16474 + }, + { + "epoch": 1.143223926167511, + "grad_norm": 1.7001317739486694, + "learning_rate": 0.00016760077990204806, + "loss": 0.6928, + "step": 16475 + }, + { + "epoch": 1.1432933176046076, + "grad_norm": 1.6081843376159668, + "learning_rate": 0.00016759663946869403, + "loss": 0.629, + "step": 16476 + }, + { + "epoch": 1.1433627090417042, + "grad_norm": 1.5997424125671387, + "learning_rate": 0.00016759249882194527, + "loss": 0.6126, + "step": 16477 + }, + { + "epoch": 1.1434321004788008, + "grad_norm": 1.8532994985580444, + "learning_rate": 0.00016758835796181485, + "loss": 0.626, + "step": 16478 + }, + { + "epoch": 1.1435014919158977, + "grad_norm": 2.4411141872406006, + "learning_rate": 0.0001675842168883158, + "loss": 0.5988, + "step": 16479 + }, + { + "epoch": 1.1435708833529943, + "grad_norm": 2.1676149368286133, + "learning_rate": 0.0001675800756014612, + "loss": 0.6785, + "step": 16480 + }, + { + "epoch": 1.143640274790091, + "grad_norm": 2.1623170375823975, + "learning_rate": 0.0001675759341012642, + "loss": 0.7787, + "step": 16481 + }, + { + "epoch": 1.1437096662271875, + "grad_norm": 1.5149985551834106, + "learning_rate": 0.0001675717923877378, + "loss": 0.6158, + "step": 16482 + }, + { + "epoch": 1.1437790576642843, + "grad_norm": 2.080416679382324, + "learning_rate": 0.00016756765046089503, + "loss": 0.9562, + "step": 16483 + }, + { + "epoch": 1.143848449101381, + "grad_norm": 2.404127359390259, + "learning_rate": 0.0001675635083207491, + "loss": 0.6727, + "step": 16484 + }, + { + "epoch": 1.1439178405384776, + "grad_norm": 2.03405499458313, + "learning_rate": 0.000167559365967313, + "loss": 0.8459, + "step": 16485 + }, + { + "epoch": 1.1439872319755742, + "grad_norm": 1.8771436214447021, + "learning_rate": 0.00016755522340059978, + "loss": 0.8177, + "step": 16486 + }, + { + "epoch": 1.1440566234126708, + "grad_norm": 2.750617027282715, + "learning_rate": 0.0001675510806206226, + "loss": 0.9402, + "step": 16487 + }, + { + "epoch": 1.1441260148497676, + "grad_norm": 2.263756275177002, + "learning_rate": 0.00016754693762739448, + "loss": 0.7336, + "step": 16488 + }, + { + "epoch": 1.1441954062868642, + "grad_norm": 2.681732416152954, + "learning_rate": 0.0001675427944209285, + "loss": 0.6739, + "step": 16489 + }, + { + "epoch": 1.1442647977239608, + "grad_norm": 3.104639768600464, + "learning_rate": 0.0001675386510012378, + "loss": 0.5849, + "step": 16490 + }, + { + "epoch": 1.1443341891610574, + "grad_norm": 1.9460482597351074, + "learning_rate": 0.00016753450736833533, + "loss": 0.6694, + "step": 16491 + }, + { + "epoch": 1.1444035805981543, + "grad_norm": 1.9693324565887451, + "learning_rate": 0.00016753036352223433, + "loss": 0.6214, + "step": 16492 + }, + { + "epoch": 1.144472972035251, + "grad_norm": 2.161254644393921, + "learning_rate": 0.0001675262194629478, + "loss": 0.7781, + "step": 16493 + }, + { + "epoch": 1.1445423634723475, + "grad_norm": 1.945881962776184, + "learning_rate": 0.00016752207519048882, + "loss": 0.779, + "step": 16494 + }, + { + "epoch": 1.1446117549094441, + "grad_norm": 1.7687182426452637, + "learning_rate": 0.00016751793070487046, + "loss": 0.5695, + "step": 16495 + }, + { + "epoch": 1.144681146346541, + "grad_norm": 3.4236340522766113, + "learning_rate": 0.00016751378600610584, + "loss": 0.7098, + "step": 16496 + }, + { + "epoch": 1.1447505377836376, + "grad_norm": 1.728837013244629, + "learning_rate": 0.00016750964109420803, + "loss": 0.5661, + "step": 16497 + }, + { + "epoch": 1.1448199292207342, + "grad_norm": 2.4254708290100098, + "learning_rate": 0.00016750549596919015, + "loss": 0.7467, + "step": 16498 + }, + { + "epoch": 1.1448893206578308, + "grad_norm": 2.069997787475586, + "learning_rate": 0.0001675013506310652, + "loss": 0.77, + "step": 16499 + }, + { + "epoch": 1.1449587120949274, + "grad_norm": 1.7714924812316895, + "learning_rate": 0.00016749720507984637, + "loss": 0.746, + "step": 16500 + }, + { + "epoch": 1.1450281035320242, + "grad_norm": 1.9440950155258179, + "learning_rate": 0.00016749305931554665, + "loss": 0.4674, + "step": 16501 + }, + { + "epoch": 1.1450974949691208, + "grad_norm": 1.784018874168396, + "learning_rate": 0.00016748891333817918, + "loss": 0.8539, + "step": 16502 + }, + { + "epoch": 1.1451668864062174, + "grad_norm": 1.9724491834640503, + "learning_rate": 0.00016748476714775705, + "loss": 0.5373, + "step": 16503 + }, + { + "epoch": 1.145236277843314, + "grad_norm": 1.689682960510254, + "learning_rate": 0.00016748062074429333, + "loss": 0.5442, + "step": 16504 + }, + { + "epoch": 1.1453056692804109, + "grad_norm": 1.6245592832565308, + "learning_rate": 0.00016747647412780115, + "loss": 0.5559, + "step": 16505 + }, + { + "epoch": 1.1453750607175075, + "grad_norm": 2.2649762630462646, + "learning_rate": 0.00016747232729829353, + "loss": 0.8075, + "step": 16506 + }, + { + "epoch": 1.145444452154604, + "grad_norm": 2.0181901454925537, + "learning_rate": 0.00016746818025578362, + "loss": 0.5942, + "step": 16507 + }, + { + "epoch": 1.1455138435917007, + "grad_norm": 1.5770014524459839, + "learning_rate": 0.00016746403300028448, + "loss": 0.5545, + "step": 16508 + }, + { + "epoch": 1.1455832350287976, + "grad_norm": 1.815055251121521, + "learning_rate": 0.00016745988553180924, + "loss": 0.5561, + "step": 16509 + }, + { + "epoch": 1.1456526264658942, + "grad_norm": 2.814697742462158, + "learning_rate": 0.00016745573785037095, + "loss": 0.664, + "step": 16510 + }, + { + "epoch": 1.1457220179029908, + "grad_norm": 1.8460280895233154, + "learning_rate": 0.00016745158995598271, + "loss": 0.565, + "step": 16511 + }, + { + "epoch": 1.1457914093400874, + "grad_norm": 1.7913285493850708, + "learning_rate": 0.00016744744184865766, + "loss": 0.5864, + "step": 16512 + }, + { + "epoch": 1.145860800777184, + "grad_norm": 1.9833948612213135, + "learning_rate": 0.00016744329352840883, + "loss": 0.6037, + "step": 16513 + }, + { + "epoch": 1.1459301922142808, + "grad_norm": 1.494183897972107, + "learning_rate": 0.00016743914499524936, + "loss": 0.452, + "step": 16514 + }, + { + "epoch": 1.1459995836513774, + "grad_norm": 1.985646367073059, + "learning_rate": 0.00016743499624919235, + "loss": 0.4955, + "step": 16515 + }, + { + "epoch": 1.146068975088474, + "grad_norm": 1.911024808883667, + "learning_rate": 0.00016743084729025086, + "loss": 0.5901, + "step": 16516 + }, + { + "epoch": 1.1461383665255707, + "grad_norm": 1.9133377075195312, + "learning_rate": 0.00016742669811843804, + "loss": 0.7299, + "step": 16517 + }, + { + "epoch": 1.1462077579626675, + "grad_norm": 2.6398661136627197, + "learning_rate": 0.0001674225487337669, + "loss": 0.619, + "step": 16518 + }, + { + "epoch": 1.146277149399764, + "grad_norm": 2.5425896644592285, + "learning_rate": 0.00016741839913625062, + "loss": 0.7717, + "step": 16519 + }, + { + "epoch": 1.1463465408368607, + "grad_norm": 1.6062045097351074, + "learning_rate": 0.00016741424932590227, + "loss": 0.6966, + "step": 16520 + }, + { + "epoch": 1.1464159322739573, + "grad_norm": 3.972586154937744, + "learning_rate": 0.00016741009930273498, + "loss": 0.7906, + "step": 16521 + }, + { + "epoch": 1.1464853237110542, + "grad_norm": 2.1439034938812256, + "learning_rate": 0.00016740594906676184, + "loss": 0.7033, + "step": 16522 + }, + { + "epoch": 1.1465547151481508, + "grad_norm": 1.6734905242919922, + "learning_rate": 0.00016740179861799592, + "loss": 0.6681, + "step": 16523 + }, + { + "epoch": 1.1466241065852474, + "grad_norm": 1.6237053871154785, + "learning_rate": 0.00016739764795645033, + "loss": 0.7686, + "step": 16524 + }, + { + "epoch": 1.146693498022344, + "grad_norm": 2.6806399822235107, + "learning_rate": 0.00016739349708213817, + "loss": 0.5895, + "step": 16525 + }, + { + "epoch": 1.1467628894594406, + "grad_norm": 1.7165483236312866, + "learning_rate": 0.00016738934599507262, + "loss": 0.4836, + "step": 16526 + }, + { + "epoch": 1.1468322808965374, + "grad_norm": 1.7957762479782104, + "learning_rate": 0.00016738519469526665, + "loss": 0.5791, + "step": 16527 + }, + { + "epoch": 1.146901672333634, + "grad_norm": 3.415487289428711, + "learning_rate": 0.0001673810431827335, + "loss": 0.7639, + "step": 16528 + }, + { + "epoch": 1.1469710637707307, + "grad_norm": 2.9839823246002197, + "learning_rate": 0.00016737689145748615, + "loss": 0.8926, + "step": 16529 + }, + { + "epoch": 1.1470404552078273, + "grad_norm": 2.3039917945861816, + "learning_rate": 0.00016737273951953783, + "loss": 0.9451, + "step": 16530 + }, + { + "epoch": 1.147109846644924, + "grad_norm": 1.7971540689468384, + "learning_rate": 0.00016736858736890158, + "loss": 0.6533, + "step": 16531 + }, + { + "epoch": 1.1471792380820207, + "grad_norm": 1.6983351707458496, + "learning_rate": 0.0001673644350055905, + "loss": 0.7126, + "step": 16532 + }, + { + "epoch": 1.1472486295191173, + "grad_norm": 2.7056756019592285, + "learning_rate": 0.00016736028242961772, + "loss": 0.7244, + "step": 16533 + }, + { + "epoch": 1.147318020956214, + "grad_norm": 1.9663559198379517, + "learning_rate": 0.00016735612964099633, + "loss": 0.7642, + "step": 16534 + }, + { + "epoch": 1.1473874123933108, + "grad_norm": 2.590728998184204, + "learning_rate": 0.00016735197663973948, + "loss": 0.6343, + "step": 16535 + }, + { + "epoch": 1.1474568038304074, + "grad_norm": 1.9211995601654053, + "learning_rate": 0.00016734782342586022, + "loss": 0.666, + "step": 16536 + }, + { + "epoch": 1.147526195267504, + "grad_norm": 2.0147817134857178, + "learning_rate": 0.0001673436699993717, + "loss": 0.7592, + "step": 16537 + }, + { + "epoch": 1.1475955867046006, + "grad_norm": 2.5275895595550537, + "learning_rate": 0.00016733951636028707, + "loss": 0.9394, + "step": 16538 + }, + { + "epoch": 1.1476649781416972, + "grad_norm": 2.3139195442199707, + "learning_rate": 0.00016733536250861935, + "loss": 0.7926, + "step": 16539 + }, + { + "epoch": 1.147734369578794, + "grad_norm": 5.466318130493164, + "learning_rate": 0.00016733120844438174, + "loss": 0.7022, + "step": 16540 + }, + { + "epoch": 1.1478037610158907, + "grad_norm": 2.3690292835235596, + "learning_rate": 0.0001673270541675873, + "loss": 0.8577, + "step": 16541 + }, + { + "epoch": 1.1478731524529873, + "grad_norm": 2.099402904510498, + "learning_rate": 0.00016732289967824916, + "loss": 0.7382, + "step": 16542 + }, + { + "epoch": 1.1479425438900839, + "grad_norm": 2.3775312900543213, + "learning_rate": 0.00016731874497638046, + "loss": 1.0641, + "step": 16543 + }, + { + "epoch": 1.1480119353271807, + "grad_norm": 2.0380401611328125, + "learning_rate": 0.00016731459006199425, + "loss": 0.653, + "step": 16544 + }, + { + "epoch": 1.1480813267642773, + "grad_norm": 2.056143045425415, + "learning_rate": 0.0001673104349351037, + "loss": 0.7031, + "step": 16545 + }, + { + "epoch": 1.148150718201374, + "grad_norm": 2.4305033683776855, + "learning_rate": 0.00016730627959572191, + "loss": 0.773, + "step": 16546 + }, + { + "epoch": 1.1482201096384705, + "grad_norm": 1.6519663333892822, + "learning_rate": 0.00016730212404386205, + "loss": 0.5618, + "step": 16547 + }, + { + "epoch": 1.1482895010755674, + "grad_norm": 1.8071153163909912, + "learning_rate": 0.00016729796827953715, + "loss": 0.5826, + "step": 16548 + }, + { + "epoch": 1.148358892512664, + "grad_norm": 1.6466407775878906, + "learning_rate": 0.0001672938123027604, + "loss": 0.4512, + "step": 16549 + }, + { + "epoch": 1.1484282839497606, + "grad_norm": 2.3013648986816406, + "learning_rate": 0.00016728965611354486, + "loss": 0.7807, + "step": 16550 + }, + { + "epoch": 1.1484976753868572, + "grad_norm": 2.135488271713257, + "learning_rate": 0.0001672854997119037, + "loss": 0.7588, + "step": 16551 + }, + { + "epoch": 1.1485670668239538, + "grad_norm": 2.901700496673584, + "learning_rate": 0.00016728134309785002, + "loss": 0.8473, + "step": 16552 + }, + { + "epoch": 1.1486364582610507, + "grad_norm": 2.446382761001587, + "learning_rate": 0.00016727718627139695, + "loss": 0.7555, + "step": 16553 + }, + { + "epoch": 1.1487058496981473, + "grad_norm": 2.1166112422943115, + "learning_rate": 0.00016727302923255763, + "loss": 0.719, + "step": 16554 + }, + { + "epoch": 1.1487752411352439, + "grad_norm": 2.061730146408081, + "learning_rate": 0.0001672688719813451, + "loss": 0.7008, + "step": 16555 + }, + { + "epoch": 1.1488446325723405, + "grad_norm": 2.541459321975708, + "learning_rate": 0.00016726471451777262, + "loss": 0.8305, + "step": 16556 + }, + { + "epoch": 1.1489140240094373, + "grad_norm": 2.601517677307129, + "learning_rate": 0.0001672605568418532, + "loss": 0.7874, + "step": 16557 + }, + { + "epoch": 1.148983415446534, + "grad_norm": 1.7483657598495483, + "learning_rate": 0.0001672563989536, + "loss": 0.6576, + "step": 16558 + }, + { + "epoch": 1.1490528068836305, + "grad_norm": 2.5851171016693115, + "learning_rate": 0.00016725224085302617, + "loss": 0.6145, + "step": 16559 + }, + { + "epoch": 1.1491221983207272, + "grad_norm": 2.0105724334716797, + "learning_rate": 0.00016724808254014483, + "loss": 0.6676, + "step": 16560 + }, + { + "epoch": 1.149191589757824, + "grad_norm": 2.636979103088379, + "learning_rate": 0.00016724392401496906, + "loss": 0.5399, + "step": 16561 + }, + { + "epoch": 1.1492609811949206, + "grad_norm": 2.558770179748535, + "learning_rate": 0.00016723976527751203, + "loss": 0.6464, + "step": 16562 + }, + { + "epoch": 1.1493303726320172, + "grad_norm": 1.8552006483078003, + "learning_rate": 0.0001672356063277869, + "loss": 0.4754, + "step": 16563 + }, + { + "epoch": 1.1493997640691138, + "grad_norm": 2.3790395259857178, + "learning_rate": 0.00016723144716580674, + "loss": 0.6609, + "step": 16564 + }, + { + "epoch": 1.1494691555062104, + "grad_norm": 1.9819673299789429, + "learning_rate": 0.0001672272877915847, + "loss": 0.5871, + "step": 16565 + }, + { + "epoch": 1.1495385469433073, + "grad_norm": 3.6492409706115723, + "learning_rate": 0.00016722312820513388, + "loss": 0.7686, + "step": 16566 + }, + { + "epoch": 1.1496079383804039, + "grad_norm": 1.6780575513839722, + "learning_rate": 0.00016721896840646747, + "loss": 0.5819, + "step": 16567 + }, + { + "epoch": 1.1496773298175005, + "grad_norm": 2.318394660949707, + "learning_rate": 0.0001672148083955986, + "loss": 0.504, + "step": 16568 + }, + { + "epoch": 1.149746721254597, + "grad_norm": 1.8481159210205078, + "learning_rate": 0.00016721064817254035, + "loss": 0.6234, + "step": 16569 + }, + { + "epoch": 1.149816112691694, + "grad_norm": 1.4893945455551147, + "learning_rate": 0.0001672064877373059, + "loss": 0.4948, + "step": 16570 + }, + { + "epoch": 1.1498855041287905, + "grad_norm": 1.8795732259750366, + "learning_rate": 0.00016720232708990836, + "loss": 0.7136, + "step": 16571 + }, + { + "epoch": 1.1499548955658871, + "grad_norm": 1.9353749752044678, + "learning_rate": 0.00016719816623036085, + "loss": 0.6813, + "step": 16572 + }, + { + "epoch": 1.1500242870029838, + "grad_norm": 1.8928471803665161, + "learning_rate": 0.00016719400515867653, + "loss": 0.7341, + "step": 16573 + }, + { + "epoch": 1.1500936784400806, + "grad_norm": 1.9889122247695923, + "learning_rate": 0.00016718984387486854, + "loss": 0.8546, + "step": 16574 + }, + { + "epoch": 1.1501630698771772, + "grad_norm": 1.788987159729004, + "learning_rate": 0.00016718568237895004, + "loss": 0.8465, + "step": 16575 + }, + { + "epoch": 1.1502324613142738, + "grad_norm": 2.772676944732666, + "learning_rate": 0.0001671815206709341, + "loss": 0.7176, + "step": 16576 + }, + { + "epoch": 1.1503018527513704, + "grad_norm": 1.8677220344543457, + "learning_rate": 0.00016717735875083388, + "loss": 0.8115, + "step": 16577 + }, + { + "epoch": 1.150371244188467, + "grad_norm": 2.132054090499878, + "learning_rate": 0.00016717319661866255, + "loss": 0.673, + "step": 16578 + }, + { + "epoch": 1.1504406356255639, + "grad_norm": 1.5037193298339844, + "learning_rate": 0.00016716903427443325, + "loss": 0.5155, + "step": 16579 + }, + { + "epoch": 1.1505100270626605, + "grad_norm": 2.3840153217315674, + "learning_rate": 0.00016716487171815906, + "loss": 0.6403, + "step": 16580 + }, + { + "epoch": 1.150579418499757, + "grad_norm": 2.1864852905273438, + "learning_rate": 0.00016716070894985322, + "loss": 0.547, + "step": 16581 + }, + { + "epoch": 1.1506488099368537, + "grad_norm": 1.506309151649475, + "learning_rate": 0.00016715654596952874, + "loss": 0.5807, + "step": 16582 + }, + { + "epoch": 1.1507182013739505, + "grad_norm": 2.4199934005737305, + "learning_rate": 0.00016715238277719888, + "loss": 0.5204, + "step": 16583 + }, + { + "epoch": 1.1507875928110471, + "grad_norm": 1.5987111330032349, + "learning_rate": 0.00016714821937287674, + "loss": 0.5363, + "step": 16584 + }, + { + "epoch": 1.1508569842481438, + "grad_norm": 2.1974294185638428, + "learning_rate": 0.00016714405575657547, + "loss": 0.6193, + "step": 16585 + }, + { + "epoch": 1.1509263756852404, + "grad_norm": 2.1875452995300293, + "learning_rate": 0.00016713989192830822, + "loss": 0.6958, + "step": 16586 + }, + { + "epoch": 1.1509957671223372, + "grad_norm": 2.1180520057678223, + "learning_rate": 0.00016713572788808808, + "loss": 0.7287, + "step": 16587 + }, + { + "epoch": 1.1510651585594338, + "grad_norm": 2.2095189094543457, + "learning_rate": 0.00016713156363592826, + "loss": 0.6066, + "step": 16588 + }, + { + "epoch": 1.1511345499965304, + "grad_norm": 1.9599113464355469, + "learning_rate": 0.00016712739917184186, + "loss": 0.9321, + "step": 16589 + }, + { + "epoch": 1.151203941433627, + "grad_norm": 1.8161259889602661, + "learning_rate": 0.00016712323449584205, + "loss": 0.6335, + "step": 16590 + }, + { + "epoch": 1.1512733328707236, + "grad_norm": 1.840329647064209, + "learning_rate": 0.00016711906960794202, + "loss": 0.6703, + "step": 16591 + }, + { + "epoch": 1.1513427243078205, + "grad_norm": 1.616027593612671, + "learning_rate": 0.00016711490450815482, + "loss": 0.4962, + "step": 16592 + }, + { + "epoch": 1.151412115744917, + "grad_norm": 1.9983789920806885, + "learning_rate": 0.00016711073919649367, + "loss": 0.7117, + "step": 16593 + }, + { + "epoch": 1.1514815071820137, + "grad_norm": 2.1224100589752197, + "learning_rate": 0.00016710657367297174, + "loss": 0.7172, + "step": 16594 + }, + { + "epoch": 1.1515508986191103, + "grad_norm": 2.408642530441284, + "learning_rate": 0.0001671024079376021, + "loss": 0.625, + "step": 16595 + }, + { + "epoch": 1.1516202900562071, + "grad_norm": 2.490865707397461, + "learning_rate": 0.00016709824199039795, + "loss": 0.7874, + "step": 16596 + }, + { + "epoch": 1.1516896814933038, + "grad_norm": 2.2934670448303223, + "learning_rate": 0.00016709407583137241, + "loss": 0.7047, + "step": 16597 + }, + { + "epoch": 1.1517590729304004, + "grad_norm": 1.5384832620620728, + "learning_rate": 0.0001670899094605387, + "loss": 0.5957, + "step": 16598 + }, + { + "epoch": 1.151828464367497, + "grad_norm": 2.0057735443115234, + "learning_rate": 0.00016708574287790993, + "loss": 0.6106, + "step": 16599 + }, + { + "epoch": 1.1518978558045938, + "grad_norm": 1.5065584182739258, + "learning_rate": 0.00016708157608349926, + "loss": 0.5989, + "step": 16600 + }, + { + "epoch": 1.1519672472416904, + "grad_norm": 1.899043321609497, + "learning_rate": 0.0001670774090773198, + "loss": 0.6152, + "step": 16601 + }, + { + "epoch": 1.152036638678787, + "grad_norm": 2.5280120372772217, + "learning_rate": 0.00016707324185938476, + "loss": 0.8515, + "step": 16602 + }, + { + "epoch": 1.1521060301158836, + "grad_norm": 1.7312982082366943, + "learning_rate": 0.0001670690744297073, + "loss": 0.5707, + "step": 16603 + }, + { + "epoch": 1.1521754215529803, + "grad_norm": 1.950312614440918, + "learning_rate": 0.0001670649067883005, + "loss": 0.8363, + "step": 16604 + }, + { + "epoch": 1.152244812990077, + "grad_norm": 1.590216040611267, + "learning_rate": 0.00016706073893517763, + "loss": 0.6473, + "step": 16605 + }, + { + "epoch": 1.1523142044271737, + "grad_norm": 1.8448575735092163, + "learning_rate": 0.00016705657087035176, + "loss": 0.6472, + "step": 16606 + }, + { + "epoch": 1.1523835958642703, + "grad_norm": 1.4728854894638062, + "learning_rate": 0.00016705240259383606, + "loss": 0.5546, + "step": 16607 + }, + { + "epoch": 1.152452987301367, + "grad_norm": 2.146958827972412, + "learning_rate": 0.00016704823410564372, + "loss": 0.7058, + "step": 16608 + }, + { + "epoch": 1.1525223787384637, + "grad_norm": 2.257746934890747, + "learning_rate": 0.00016704406540578786, + "loss": 0.7516, + "step": 16609 + }, + { + "epoch": 1.1525917701755604, + "grad_norm": 1.7060061693191528, + "learning_rate": 0.0001670398964942817, + "loss": 0.8105, + "step": 16610 + }, + { + "epoch": 1.152661161612657, + "grad_norm": 1.9203777313232422, + "learning_rate": 0.00016703572737113834, + "loss": 0.6157, + "step": 16611 + }, + { + "epoch": 1.1527305530497536, + "grad_norm": 1.975275993347168, + "learning_rate": 0.000167031558036371, + "loss": 0.6657, + "step": 16612 + }, + { + "epoch": 1.1527999444868504, + "grad_norm": 2.095388412475586, + "learning_rate": 0.0001670273884899928, + "loss": 0.7599, + "step": 16613 + }, + { + "epoch": 1.152869335923947, + "grad_norm": 2.087726354598999, + "learning_rate": 0.00016702321873201688, + "loss": 0.4577, + "step": 16614 + }, + { + "epoch": 1.1529387273610436, + "grad_norm": 2.8295416831970215, + "learning_rate": 0.00016701904876245644, + "loss": 0.6075, + "step": 16615 + }, + { + "epoch": 1.1530081187981402, + "grad_norm": 1.636720895767212, + "learning_rate": 0.0001670148785813247, + "loss": 0.6353, + "step": 16616 + }, + { + "epoch": 1.1530775102352369, + "grad_norm": 2.062283515930176, + "learning_rate": 0.00016701070818863468, + "loss": 0.64, + "step": 16617 + }, + { + "epoch": 1.1531469016723337, + "grad_norm": 1.5030670166015625, + "learning_rate": 0.00016700653758439966, + "loss": 0.5897, + "step": 16618 + }, + { + "epoch": 1.1532162931094303, + "grad_norm": 2.043726682662964, + "learning_rate": 0.00016700236676863277, + "loss": 0.7589, + "step": 16619 + }, + { + "epoch": 1.153285684546527, + "grad_norm": 1.5409612655639648, + "learning_rate": 0.00016699819574134722, + "loss": 0.5924, + "step": 16620 + }, + { + "epoch": 1.1533550759836237, + "grad_norm": 2.0421407222747803, + "learning_rate": 0.00016699402450255608, + "loss": 0.8295, + "step": 16621 + }, + { + "epoch": 1.1534244674207204, + "grad_norm": 2.585226535797119, + "learning_rate": 0.00016698985305227263, + "loss": 0.668, + "step": 16622 + }, + { + "epoch": 1.153493858857817, + "grad_norm": 1.6360771656036377, + "learning_rate": 0.00016698568139050996, + "loss": 0.7169, + "step": 16623 + }, + { + "epoch": 1.1535632502949136, + "grad_norm": 1.9017645120620728, + "learning_rate": 0.00016698150951728128, + "loss": 0.6644, + "step": 16624 + }, + { + "epoch": 1.1536326417320102, + "grad_norm": 2.3781120777130127, + "learning_rate": 0.00016697733743259972, + "loss": 0.8502, + "step": 16625 + }, + { + "epoch": 1.153702033169107, + "grad_norm": 1.7883484363555908, + "learning_rate": 0.0001669731651364785, + "loss": 0.4768, + "step": 16626 + }, + { + "epoch": 1.1537714246062036, + "grad_norm": 2.05204701423645, + "learning_rate": 0.00016696899262893074, + "loss": 0.7287, + "step": 16627 + }, + { + "epoch": 1.1538408160433002, + "grad_norm": 1.8181506395339966, + "learning_rate": 0.00016696481990996967, + "loss": 0.769, + "step": 16628 + }, + { + "epoch": 1.1539102074803969, + "grad_norm": 1.5816125869750977, + "learning_rate": 0.0001669606469796084, + "loss": 0.6402, + "step": 16629 + }, + { + "epoch": 1.1539795989174935, + "grad_norm": 2.1958200931549072, + "learning_rate": 0.00016695647383786018, + "loss": 0.8715, + "step": 16630 + }, + { + "epoch": 1.1540489903545903, + "grad_norm": 1.9343326091766357, + "learning_rate": 0.0001669523004847381, + "loss": 0.593, + "step": 16631 + }, + { + "epoch": 1.154118381791687, + "grad_norm": 2.5714612007141113, + "learning_rate": 0.00016694812692025538, + "loss": 0.5961, + "step": 16632 + }, + { + "epoch": 1.1541877732287835, + "grad_norm": 1.9567291736602783, + "learning_rate": 0.00016694395314442525, + "loss": 0.6702, + "step": 16633 + }, + { + "epoch": 1.1542571646658804, + "grad_norm": 2.606998920440674, + "learning_rate": 0.00016693977915726075, + "loss": 0.6083, + "step": 16634 + }, + { + "epoch": 1.154326556102977, + "grad_norm": 2.0076839923858643, + "learning_rate": 0.00016693560495877519, + "loss": 0.7989, + "step": 16635 + }, + { + "epoch": 1.1543959475400736, + "grad_norm": 1.7708261013031006, + "learning_rate": 0.00016693143054898166, + "loss": 0.514, + "step": 16636 + }, + { + "epoch": 1.1544653389771702, + "grad_norm": 2.2280309200286865, + "learning_rate": 0.0001669272559278934, + "loss": 0.7068, + "step": 16637 + }, + { + "epoch": 1.1545347304142668, + "grad_norm": 2.2472145557403564, + "learning_rate": 0.0001669230810955235, + "loss": 0.8421, + "step": 16638 + }, + { + "epoch": 1.1546041218513636, + "grad_norm": 2.39437198638916, + "learning_rate": 0.00016691890605188522, + "loss": 0.487, + "step": 16639 + }, + { + "epoch": 1.1546735132884602, + "grad_norm": 1.6383631229400635, + "learning_rate": 0.00016691473079699174, + "loss": 0.6137, + "step": 16640 + }, + { + "epoch": 1.1547429047255569, + "grad_norm": 4.130270481109619, + "learning_rate": 0.0001669105553308562, + "loss": 0.812, + "step": 16641 + }, + { + "epoch": 1.1548122961626535, + "grad_norm": 2.0382943153381348, + "learning_rate": 0.0001669063796534918, + "loss": 0.5502, + "step": 16642 + }, + { + "epoch": 1.15488168759975, + "grad_norm": 1.8050856590270996, + "learning_rate": 0.00016690220376491173, + "loss": 0.6632, + "step": 16643 + }, + { + "epoch": 1.154951079036847, + "grad_norm": 2.5741326808929443, + "learning_rate": 0.00016689802766512915, + "loss": 0.6156, + "step": 16644 + }, + { + "epoch": 1.1550204704739435, + "grad_norm": 2.655364513397217, + "learning_rate": 0.00016689385135415733, + "loss": 0.7738, + "step": 16645 + }, + { + "epoch": 1.1550898619110401, + "grad_norm": 1.9200215339660645, + "learning_rate": 0.0001668896748320093, + "loss": 0.6959, + "step": 16646 + }, + { + "epoch": 1.155159253348137, + "grad_norm": 2.1169159412384033, + "learning_rate": 0.00016688549809869835, + "loss": 0.6498, + "step": 16647 + }, + { + "epoch": 1.1552286447852336, + "grad_norm": 1.7698906660079956, + "learning_rate": 0.00016688132115423763, + "loss": 0.6155, + "step": 16648 + }, + { + "epoch": 1.1552980362223302, + "grad_norm": 1.517744779586792, + "learning_rate": 0.00016687714399864036, + "loss": 0.6187, + "step": 16649 + }, + { + "epoch": 1.1553674276594268, + "grad_norm": 1.7121479511260986, + "learning_rate": 0.0001668729666319197, + "loss": 0.5309, + "step": 16650 + }, + { + "epoch": 1.1554368190965234, + "grad_norm": 5.005683898925781, + "learning_rate": 0.00016686878905408883, + "loss": 0.8172, + "step": 16651 + }, + { + "epoch": 1.1555062105336202, + "grad_norm": 2.0204055309295654, + "learning_rate": 0.00016686461126516099, + "loss": 0.7677, + "step": 16652 + }, + { + "epoch": 1.1555756019707168, + "grad_norm": 2.2684056758880615, + "learning_rate": 0.00016686043326514927, + "loss": 0.8541, + "step": 16653 + }, + { + "epoch": 1.1556449934078135, + "grad_norm": 2.2872745990753174, + "learning_rate": 0.00016685625505406698, + "loss": 0.7075, + "step": 16654 + }, + { + "epoch": 1.15571438484491, + "grad_norm": 1.9803472757339478, + "learning_rate": 0.00016685207663192722, + "loss": 0.6515, + "step": 16655 + }, + { + "epoch": 1.1557837762820067, + "grad_norm": 1.6349104642868042, + "learning_rate": 0.0001668478979987432, + "loss": 0.6281, + "step": 16656 + }, + { + "epoch": 1.1558531677191035, + "grad_norm": 2.1478376388549805, + "learning_rate": 0.00016684371915452815, + "loss": 0.6495, + "step": 16657 + }, + { + "epoch": 1.1559225591562001, + "grad_norm": 2.7195513248443604, + "learning_rate": 0.0001668395400992952, + "loss": 0.7843, + "step": 16658 + }, + { + "epoch": 1.1559919505932967, + "grad_norm": 1.572629451751709, + "learning_rate": 0.0001668353608330576, + "loss": 0.7887, + "step": 16659 + }, + { + "epoch": 1.1560613420303936, + "grad_norm": 1.7909497022628784, + "learning_rate": 0.00016683118135582848, + "loss": 0.6856, + "step": 16660 + }, + { + "epoch": 1.1561307334674902, + "grad_norm": 2.0753579139709473, + "learning_rate": 0.00016682700166762115, + "loss": 0.725, + "step": 16661 + }, + { + "epoch": 1.1562001249045868, + "grad_norm": 1.975461721420288, + "learning_rate": 0.00016682282176844868, + "loss": 0.7269, + "step": 16662 + }, + { + "epoch": 1.1562695163416834, + "grad_norm": 1.7617976665496826, + "learning_rate": 0.00016681864165832432, + "loss": 0.6698, + "step": 16663 + }, + { + "epoch": 1.15633890777878, + "grad_norm": 1.8949822187423706, + "learning_rate": 0.00016681446133726127, + "loss": 0.7637, + "step": 16664 + }, + { + "epoch": 1.1564082992158768, + "grad_norm": 2.5368611812591553, + "learning_rate": 0.00016681028080527274, + "loss": 0.8118, + "step": 16665 + }, + { + "epoch": 1.1564776906529735, + "grad_norm": 2.1509451866149902, + "learning_rate": 0.00016680610006237186, + "loss": 0.7671, + "step": 16666 + }, + { + "epoch": 1.15654708209007, + "grad_norm": 2.0265095233917236, + "learning_rate": 0.00016680191910857186, + "loss": 0.5406, + "step": 16667 + }, + { + "epoch": 1.1566164735271667, + "grad_norm": 1.9708259105682373, + "learning_rate": 0.000166797737943886, + "loss": 0.7529, + "step": 16668 + }, + { + "epoch": 1.1566858649642633, + "grad_norm": 2.943699359893799, + "learning_rate": 0.0001667935565683274, + "loss": 0.6873, + "step": 16669 + }, + { + "epoch": 1.1567552564013601, + "grad_norm": 2.088719129562378, + "learning_rate": 0.0001667893749819093, + "loss": 0.7311, + "step": 16670 + }, + { + "epoch": 1.1568246478384567, + "grad_norm": 1.8849300146102905, + "learning_rate": 0.00016678519318464493, + "loss": 0.8134, + "step": 16671 + }, + { + "epoch": 1.1568940392755533, + "grad_norm": 1.8356425762176514, + "learning_rate": 0.00016678101117654741, + "loss": 0.6699, + "step": 16672 + }, + { + "epoch": 1.1569634307126502, + "grad_norm": 1.873952865600586, + "learning_rate": 0.00016677682895763, + "loss": 0.7487, + "step": 16673 + }, + { + "epoch": 1.1570328221497468, + "grad_norm": 1.9032701253890991, + "learning_rate": 0.00016677264652790585, + "loss": 0.7164, + "step": 16674 + }, + { + "epoch": 1.1571022135868434, + "grad_norm": 1.7388521432876587, + "learning_rate": 0.00016676846388738827, + "loss": 0.6809, + "step": 16675 + }, + { + "epoch": 1.15717160502394, + "grad_norm": 1.5932857990264893, + "learning_rate": 0.00016676428103609033, + "loss": 0.6612, + "step": 16676 + }, + { + "epoch": 1.1572409964610366, + "grad_norm": 1.3515539169311523, + "learning_rate": 0.00016676009797402533, + "loss": 0.5337, + "step": 16677 + }, + { + "epoch": 1.1573103878981335, + "grad_norm": 2.8723912239074707, + "learning_rate": 0.00016675591470120647, + "loss": 0.8979, + "step": 16678 + }, + { + "epoch": 1.15737977933523, + "grad_norm": 1.4060041904449463, + "learning_rate": 0.0001667517312176469, + "loss": 0.5749, + "step": 16679 + }, + { + "epoch": 1.1574491707723267, + "grad_norm": 2.0549092292785645, + "learning_rate": 0.00016674754752335987, + "loss": 0.6237, + "step": 16680 + }, + { + "epoch": 1.1575185622094233, + "grad_norm": 1.6358709335327148, + "learning_rate": 0.00016674336361835858, + "loss": 0.382, + "step": 16681 + }, + { + "epoch": 1.15758795364652, + "grad_norm": 1.8112223148345947, + "learning_rate": 0.00016673917950265623, + "loss": 0.7577, + "step": 16682 + }, + { + "epoch": 1.1576573450836167, + "grad_norm": 1.6612263917922974, + "learning_rate": 0.00016673499517626605, + "loss": 0.6408, + "step": 16683 + }, + { + "epoch": 1.1577267365207133, + "grad_norm": 3.0853822231292725, + "learning_rate": 0.0001667308106392012, + "loss": 0.8901, + "step": 16684 + }, + { + "epoch": 1.15779612795781, + "grad_norm": 1.7963690757751465, + "learning_rate": 0.00016672662589147495, + "loss": 0.8546, + "step": 16685 + }, + { + "epoch": 1.1578655193949068, + "grad_norm": 1.5779091119766235, + "learning_rate": 0.00016672244093310048, + "loss": 0.586, + "step": 16686 + }, + { + "epoch": 1.1579349108320034, + "grad_norm": 1.8268476724624634, + "learning_rate": 0.000166718255764091, + "loss": 0.7705, + "step": 16687 + }, + { + "epoch": 1.1580043022691, + "grad_norm": 1.6772897243499756, + "learning_rate": 0.00016671407038445972, + "loss": 0.7017, + "step": 16688 + }, + { + "epoch": 1.1580736937061966, + "grad_norm": 2.631648063659668, + "learning_rate": 0.00016670988479421987, + "loss": 0.6392, + "step": 16689 + }, + { + "epoch": 1.1581430851432932, + "grad_norm": 1.7452551126480103, + "learning_rate": 0.00016670569899338467, + "loss": 0.5186, + "step": 16690 + }, + { + "epoch": 1.15821247658039, + "grad_norm": 2.1139910221099854, + "learning_rate": 0.0001667015129819673, + "loss": 0.5997, + "step": 16691 + }, + { + "epoch": 1.1582818680174867, + "grad_norm": 2.005969285964966, + "learning_rate": 0.00016669732675998096, + "loss": 0.569, + "step": 16692 + }, + { + "epoch": 1.1583512594545833, + "grad_norm": 2.3704967498779297, + "learning_rate": 0.00016669314032743894, + "loss": 0.8798, + "step": 16693 + }, + { + "epoch": 1.15842065089168, + "grad_norm": 1.9685224294662476, + "learning_rate": 0.0001666889536843544, + "loss": 0.6381, + "step": 16694 + }, + { + "epoch": 1.1584900423287767, + "grad_norm": 1.8454656600952148, + "learning_rate": 0.00016668476683074057, + "loss": 0.8465, + "step": 16695 + }, + { + "epoch": 1.1585594337658733, + "grad_norm": 1.4234477281570435, + "learning_rate": 0.00016668057976661067, + "loss": 0.7817, + "step": 16696 + }, + { + "epoch": 1.15862882520297, + "grad_norm": 2.2427940368652344, + "learning_rate": 0.0001666763924919779, + "loss": 0.7448, + "step": 16697 + }, + { + "epoch": 1.1586982166400666, + "grad_norm": 1.8998507261276245, + "learning_rate": 0.00016667220500685552, + "loss": 0.6314, + "step": 16698 + }, + { + "epoch": 1.1587676080771634, + "grad_norm": 2.709493398666382, + "learning_rate": 0.00016666801731125673, + "loss": 0.9641, + "step": 16699 + }, + { + "epoch": 1.15883699951426, + "grad_norm": 1.9972546100616455, + "learning_rate": 0.00016666382940519473, + "loss": 0.8035, + "step": 16700 + }, + { + "epoch": 1.1589063909513566, + "grad_norm": 1.6454088687896729, + "learning_rate": 0.00016665964128868277, + "loss": 0.7263, + "step": 16701 + }, + { + "epoch": 1.1589757823884532, + "grad_norm": 1.5689785480499268, + "learning_rate": 0.00016665545296173405, + "loss": 0.4567, + "step": 16702 + }, + { + "epoch": 1.1590451738255498, + "grad_norm": 2.1321704387664795, + "learning_rate": 0.00016665126442436177, + "loss": 0.5574, + "step": 16703 + }, + { + "epoch": 1.1591145652626467, + "grad_norm": 1.5977838039398193, + "learning_rate": 0.0001666470756765792, + "loss": 0.615, + "step": 16704 + }, + { + "epoch": 1.1591839566997433, + "grad_norm": 2.093273401260376, + "learning_rate": 0.00016664288671839955, + "loss": 0.5571, + "step": 16705 + }, + { + "epoch": 1.15925334813684, + "grad_norm": 2.148336172103882, + "learning_rate": 0.00016663869754983602, + "loss": 0.7044, + "step": 16706 + }, + { + "epoch": 1.1593227395739365, + "grad_norm": 1.5738941431045532, + "learning_rate": 0.0001666345081709019, + "loss": 0.5111, + "step": 16707 + }, + { + "epoch": 1.1593921310110333, + "grad_norm": 1.8223029375076294, + "learning_rate": 0.0001666303185816103, + "loss": 0.5223, + "step": 16708 + }, + { + "epoch": 1.15946152244813, + "grad_norm": 1.8754032850265503, + "learning_rate": 0.0001666261287819746, + "loss": 0.6545, + "step": 16709 + }, + { + "epoch": 1.1595309138852266, + "grad_norm": 2.487880229949951, + "learning_rate": 0.00016662193877200787, + "loss": 0.6755, + "step": 16710 + }, + { + "epoch": 1.1596003053223232, + "grad_norm": 1.8094466924667358, + "learning_rate": 0.00016661774855172343, + "loss": 0.764, + "step": 16711 + }, + { + "epoch": 1.15966969675942, + "grad_norm": 1.4911085367202759, + "learning_rate": 0.00016661355812113446, + "loss": 0.4499, + "step": 16712 + }, + { + "epoch": 1.1597390881965166, + "grad_norm": 1.5879966020584106, + "learning_rate": 0.00016660936748025428, + "loss": 0.6755, + "step": 16713 + }, + { + "epoch": 1.1598084796336132, + "grad_norm": 2.3451802730560303, + "learning_rate": 0.00016660517662909601, + "loss": 0.7337, + "step": 16714 + }, + { + "epoch": 1.1598778710707098, + "grad_norm": 2.0756754875183105, + "learning_rate": 0.00016660098556767295, + "loss": 0.8682, + "step": 16715 + }, + { + "epoch": 1.1599472625078064, + "grad_norm": 2.804374933242798, + "learning_rate": 0.00016659679429599828, + "loss": 0.8971, + "step": 16716 + }, + { + "epoch": 1.1600166539449033, + "grad_norm": 1.7925043106079102, + "learning_rate": 0.00016659260281408526, + "loss": 0.5674, + "step": 16717 + }, + { + "epoch": 1.1600860453819999, + "grad_norm": 1.2170428037643433, + "learning_rate": 0.00016658841112194713, + "loss": 0.4628, + "step": 16718 + }, + { + "epoch": 1.1601554368190965, + "grad_norm": 2.1595616340637207, + "learning_rate": 0.0001665842192195971, + "loss": 0.6711, + "step": 16719 + }, + { + "epoch": 1.160224828256193, + "grad_norm": 2.7754478454589844, + "learning_rate": 0.00016658002710704843, + "loss": 0.7203, + "step": 16720 + }, + { + "epoch": 1.16029421969329, + "grad_norm": 1.277854323387146, + "learning_rate": 0.00016657583478431432, + "loss": 0.5788, + "step": 16721 + }, + { + "epoch": 1.1603636111303866, + "grad_norm": 2.1028480529785156, + "learning_rate": 0.00016657164225140805, + "loss": 0.7547, + "step": 16722 + }, + { + "epoch": 1.1604330025674832, + "grad_norm": 1.893349289894104, + "learning_rate": 0.00016656744950834283, + "loss": 0.7015, + "step": 16723 + }, + { + "epoch": 1.1605023940045798, + "grad_norm": 1.6636104583740234, + "learning_rate": 0.00016656325655513185, + "loss": 0.6017, + "step": 16724 + }, + { + "epoch": 1.1605717854416766, + "grad_norm": 1.9345401525497437, + "learning_rate": 0.00016655906339178844, + "loss": 0.8411, + "step": 16725 + }, + { + "epoch": 1.1606411768787732, + "grad_norm": 2.140774965286255, + "learning_rate": 0.00016655487001832577, + "loss": 0.6642, + "step": 16726 + }, + { + "epoch": 1.1607105683158698, + "grad_norm": 1.5546271800994873, + "learning_rate": 0.0001665506764347571, + "loss": 0.578, + "step": 16727 + }, + { + "epoch": 1.1607799597529664, + "grad_norm": 1.950646162033081, + "learning_rate": 0.00016654648264109566, + "loss": 0.516, + "step": 16728 + }, + { + "epoch": 1.160849351190063, + "grad_norm": 2.2552146911621094, + "learning_rate": 0.0001665422886373547, + "loss": 0.6904, + "step": 16729 + }, + { + "epoch": 1.1609187426271599, + "grad_norm": 1.6613322496414185, + "learning_rate": 0.00016653809442354747, + "loss": 0.6008, + "step": 16730 + }, + { + "epoch": 1.1609881340642565, + "grad_norm": 1.7481355667114258, + "learning_rate": 0.00016653389999968715, + "loss": 0.6838, + "step": 16731 + }, + { + "epoch": 1.161057525501353, + "grad_norm": 1.7055096626281738, + "learning_rate": 0.00016652970536578708, + "loss": 0.7337, + "step": 16732 + }, + { + "epoch": 1.1611269169384497, + "grad_norm": 1.9369415044784546, + "learning_rate": 0.00016652551052186043, + "loss": 0.6708, + "step": 16733 + }, + { + "epoch": 1.1611963083755465, + "grad_norm": 2.791747808456421, + "learning_rate": 0.00016652131546792044, + "loss": 0.804, + "step": 16734 + }, + { + "epoch": 1.1612656998126432, + "grad_norm": 1.975290060043335, + "learning_rate": 0.0001665171202039804, + "loss": 0.729, + "step": 16735 + }, + { + "epoch": 1.1613350912497398, + "grad_norm": 1.956058382987976, + "learning_rate": 0.0001665129247300535, + "loss": 0.6491, + "step": 16736 + }, + { + "epoch": 1.1614044826868364, + "grad_norm": 2.4367334842681885, + "learning_rate": 0.00016650872904615306, + "loss": 0.8175, + "step": 16737 + }, + { + "epoch": 1.1614738741239332, + "grad_norm": 2.20259428024292, + "learning_rate": 0.00016650453315229227, + "loss": 0.7912, + "step": 16738 + }, + { + "epoch": 1.1615432655610298, + "grad_norm": 2.3338570594787598, + "learning_rate": 0.00016650033704848435, + "loss": 0.8063, + "step": 16739 + }, + { + "epoch": 1.1616126569981264, + "grad_norm": 1.757139801979065, + "learning_rate": 0.0001664961407347426, + "loss": 0.5256, + "step": 16740 + }, + { + "epoch": 1.161682048435223, + "grad_norm": 2.3183646202087402, + "learning_rate": 0.00016649194421108026, + "loss": 0.6941, + "step": 16741 + }, + { + "epoch": 1.1617514398723197, + "grad_norm": 1.4957133531570435, + "learning_rate": 0.00016648774747751055, + "loss": 0.5485, + "step": 16742 + }, + { + "epoch": 1.1618208313094165, + "grad_norm": 2.605501413345337, + "learning_rate": 0.00016648355053404675, + "loss": 1.017, + "step": 16743 + }, + { + "epoch": 1.161890222746513, + "grad_norm": 1.8430556058883667, + "learning_rate": 0.0001664793533807021, + "loss": 0.5061, + "step": 16744 + }, + { + "epoch": 1.1619596141836097, + "grad_norm": 1.8971189260482788, + "learning_rate": 0.0001664751560174898, + "loss": 0.6654, + "step": 16745 + }, + { + "epoch": 1.1620290056207063, + "grad_norm": 2.0205657482147217, + "learning_rate": 0.00016647095844442316, + "loss": 0.5676, + "step": 16746 + }, + { + "epoch": 1.1620983970578032, + "grad_norm": 2.254365921020508, + "learning_rate": 0.00016646676066151545, + "loss": 0.545, + "step": 16747 + }, + { + "epoch": 1.1621677884948998, + "grad_norm": 1.482308268547058, + "learning_rate": 0.00016646256266877987, + "loss": 0.4404, + "step": 16748 + }, + { + "epoch": 1.1622371799319964, + "grad_norm": 1.744685411453247, + "learning_rate": 0.00016645836446622968, + "loss": 0.5089, + "step": 16749 + }, + { + "epoch": 1.162306571369093, + "grad_norm": 1.9035210609436035, + "learning_rate": 0.00016645416605387816, + "loss": 0.653, + "step": 16750 + }, + { + "epoch": 1.1623759628061898, + "grad_norm": 2.9643959999084473, + "learning_rate": 0.00016644996743173852, + "loss": 0.7179, + "step": 16751 + }, + { + "epoch": 1.1624453542432864, + "grad_norm": 2.642669677734375, + "learning_rate": 0.0001664457685998241, + "loss": 0.7849, + "step": 16752 + }, + { + "epoch": 1.162514745680383, + "grad_norm": 2.0676653385162354, + "learning_rate": 0.00016644156955814804, + "loss": 0.878, + "step": 16753 + }, + { + "epoch": 1.1625841371174797, + "grad_norm": 2.308591604232788, + "learning_rate": 0.00016643737030672367, + "loss": 0.5204, + "step": 16754 + }, + { + "epoch": 1.1626535285545763, + "grad_norm": 1.7123444080352783, + "learning_rate": 0.00016643317084556424, + "loss": 0.645, + "step": 16755 + }, + { + "epoch": 1.162722919991673, + "grad_norm": 3.2647578716278076, + "learning_rate": 0.000166428971174683, + "loss": 0.6318, + "step": 16756 + }, + { + "epoch": 1.1627923114287697, + "grad_norm": 1.891352653503418, + "learning_rate": 0.0001664247712940932, + "loss": 0.5643, + "step": 16757 + }, + { + "epoch": 1.1628617028658663, + "grad_norm": 2.24286150932312, + "learning_rate": 0.0001664205712038081, + "loss": 0.6958, + "step": 16758 + }, + { + "epoch": 1.162931094302963, + "grad_norm": 1.6359537839889526, + "learning_rate": 0.00016641637090384097, + "loss": 0.807, + "step": 16759 + }, + { + "epoch": 1.1630004857400598, + "grad_norm": 2.088318109512329, + "learning_rate": 0.00016641217039420507, + "loss": 0.707, + "step": 16760 + }, + { + "epoch": 1.1630698771771564, + "grad_norm": 1.472144603729248, + "learning_rate": 0.00016640796967491364, + "loss": 0.7127, + "step": 16761 + }, + { + "epoch": 1.163139268614253, + "grad_norm": 2.414076328277588, + "learning_rate": 0.00016640376874597993, + "loss": 0.6681, + "step": 16762 + }, + { + "epoch": 1.1632086600513496, + "grad_norm": 1.5285029411315918, + "learning_rate": 0.00016639956760741725, + "loss": 0.6653, + "step": 16763 + }, + { + "epoch": 1.1632780514884464, + "grad_norm": 1.7511074542999268, + "learning_rate": 0.00016639536625923885, + "loss": 0.6466, + "step": 16764 + }, + { + "epoch": 1.163347442925543, + "grad_norm": 2.682115316390991, + "learning_rate": 0.00016639116470145798, + "loss": 0.937, + "step": 16765 + }, + { + "epoch": 1.1634168343626397, + "grad_norm": 1.550594687461853, + "learning_rate": 0.0001663869629340879, + "loss": 0.7139, + "step": 16766 + }, + { + "epoch": 1.1634862257997363, + "grad_norm": 2.5023951530456543, + "learning_rate": 0.00016638276095714188, + "loss": 0.891, + "step": 16767 + }, + { + "epoch": 1.1635556172368329, + "grad_norm": 1.9058088064193726, + "learning_rate": 0.0001663785587706332, + "loss": 0.6579, + "step": 16768 + }, + { + "epoch": 1.1636250086739297, + "grad_norm": 2.0828464031219482, + "learning_rate": 0.0001663743563745751, + "loss": 0.826, + "step": 16769 + }, + { + "epoch": 1.1636944001110263, + "grad_norm": 1.926567554473877, + "learning_rate": 0.00016637015376898083, + "loss": 0.7201, + "step": 16770 + }, + { + "epoch": 1.163763791548123, + "grad_norm": 1.6664711236953735, + "learning_rate": 0.00016636595095386372, + "loss": 0.6996, + "step": 16771 + }, + { + "epoch": 1.1638331829852195, + "grad_norm": 1.9871257543563843, + "learning_rate": 0.00016636174792923698, + "loss": 0.6634, + "step": 16772 + }, + { + "epoch": 1.1639025744223164, + "grad_norm": 2.067962884902954, + "learning_rate": 0.00016635754469511392, + "loss": 0.7241, + "step": 16773 + }, + { + "epoch": 1.163971965859413, + "grad_norm": 1.4758563041687012, + "learning_rate": 0.00016635334125150776, + "loss": 0.5075, + "step": 16774 + }, + { + "epoch": 1.1640413572965096, + "grad_norm": 2.268691301345825, + "learning_rate": 0.00016634913759843185, + "loss": 0.7425, + "step": 16775 + }, + { + "epoch": 1.1641107487336062, + "grad_norm": 1.9065966606140137, + "learning_rate": 0.00016634493373589936, + "loss": 0.6955, + "step": 16776 + }, + { + "epoch": 1.164180140170703, + "grad_norm": 1.9379853010177612, + "learning_rate": 0.00016634072966392364, + "loss": 0.5872, + "step": 16777 + }, + { + "epoch": 1.1642495316077996, + "grad_norm": 2.5436861515045166, + "learning_rate": 0.00016633652538251793, + "loss": 0.7097, + "step": 16778 + }, + { + "epoch": 1.1643189230448963, + "grad_norm": 2.230084180831909, + "learning_rate": 0.0001663323208916955, + "loss": 0.6289, + "step": 16779 + }, + { + "epoch": 1.1643883144819929, + "grad_norm": 1.7372455596923828, + "learning_rate": 0.00016632811619146965, + "loss": 0.7839, + "step": 16780 + }, + { + "epoch": 1.1644577059190895, + "grad_norm": 1.8233026266098022, + "learning_rate": 0.0001663239112818536, + "loss": 0.7592, + "step": 16781 + }, + { + "epoch": 1.1645270973561863, + "grad_norm": 1.8931572437286377, + "learning_rate": 0.00016631970616286067, + "loss": 0.7953, + "step": 16782 + }, + { + "epoch": 1.164596488793283, + "grad_norm": 1.5660958290100098, + "learning_rate": 0.00016631550083450413, + "loss": 0.6773, + "step": 16783 + }, + { + "epoch": 1.1646658802303795, + "grad_norm": 3.414883613586426, + "learning_rate": 0.00016631129529679726, + "loss": 0.7565, + "step": 16784 + }, + { + "epoch": 1.1647352716674761, + "grad_norm": 2.417417287826538, + "learning_rate": 0.00016630708954975328, + "loss": 0.6085, + "step": 16785 + }, + { + "epoch": 1.164804663104573, + "grad_norm": 2.4557526111602783, + "learning_rate": 0.00016630288359338556, + "loss": 0.8576, + "step": 16786 + }, + { + "epoch": 1.1648740545416696, + "grad_norm": 2.179269313812256, + "learning_rate": 0.00016629867742770727, + "loss": 0.9285, + "step": 16787 + }, + { + "epoch": 1.1649434459787662, + "grad_norm": 1.6594791412353516, + "learning_rate": 0.0001662944710527318, + "loss": 0.6542, + "step": 16788 + }, + { + "epoch": 1.1650128374158628, + "grad_norm": 1.8580812215805054, + "learning_rate": 0.00016629026446847239, + "loss": 0.6284, + "step": 16789 + }, + { + "epoch": 1.1650822288529596, + "grad_norm": 2.0542702674865723, + "learning_rate": 0.00016628605767494227, + "loss": 0.8609, + "step": 16790 + }, + { + "epoch": 1.1651516202900563, + "grad_norm": 1.6551235914230347, + "learning_rate": 0.00016628185067215476, + "loss": 0.6738, + "step": 16791 + }, + { + "epoch": 1.1652210117271529, + "grad_norm": 1.6865653991699219, + "learning_rate": 0.00016627764346012315, + "loss": 0.7641, + "step": 16792 + }, + { + "epoch": 1.1652904031642495, + "grad_norm": 1.5223050117492676, + "learning_rate": 0.00016627343603886068, + "loss": 0.7498, + "step": 16793 + }, + { + "epoch": 1.165359794601346, + "grad_norm": 2.1889889240264893, + "learning_rate": 0.0001662692284083807, + "loss": 0.6479, + "step": 16794 + }, + { + "epoch": 1.165429186038443, + "grad_norm": 1.771039366722107, + "learning_rate": 0.00016626502056869643, + "loss": 0.6744, + "step": 16795 + }, + { + "epoch": 1.1654985774755395, + "grad_norm": 2.052973508834839, + "learning_rate": 0.0001662608125198212, + "loss": 0.6512, + "step": 16796 + }, + { + "epoch": 1.1655679689126361, + "grad_norm": 1.7640868425369263, + "learning_rate": 0.00016625660426176824, + "loss": 0.5875, + "step": 16797 + }, + { + "epoch": 1.1656373603497328, + "grad_norm": 2.1197822093963623, + "learning_rate": 0.00016625239579455087, + "loss": 0.6682, + "step": 16798 + }, + { + "epoch": 1.1657067517868296, + "grad_norm": 2.346846580505371, + "learning_rate": 0.0001662481871181824, + "loss": 0.6426, + "step": 16799 + }, + { + "epoch": 1.1657761432239262, + "grad_norm": 1.9101617336273193, + "learning_rate": 0.00016624397823267607, + "loss": 0.7014, + "step": 16800 + }, + { + "epoch": 1.1658455346610228, + "grad_norm": 1.919179081916809, + "learning_rate": 0.0001662397691380452, + "loss": 0.7176, + "step": 16801 + }, + { + "epoch": 1.1659149260981194, + "grad_norm": 1.7652933597564697, + "learning_rate": 0.00016623555983430306, + "loss": 0.6736, + "step": 16802 + }, + { + "epoch": 1.1659843175352163, + "grad_norm": 1.5511651039123535, + "learning_rate": 0.00016623135032146292, + "loss": 0.664, + "step": 16803 + }, + { + "epoch": 1.1660537089723129, + "grad_norm": 1.9765362739562988, + "learning_rate": 0.00016622714059953814, + "loss": 0.7294, + "step": 16804 + }, + { + "epoch": 1.1661231004094095, + "grad_norm": 1.6443010568618774, + "learning_rate": 0.00016622293066854193, + "loss": 0.5445, + "step": 16805 + }, + { + "epoch": 1.166192491846506, + "grad_norm": 1.5011847019195557, + "learning_rate": 0.0001662187205284876, + "loss": 0.4801, + "step": 16806 + }, + { + "epoch": 1.1662618832836027, + "grad_norm": 1.916153907775879, + "learning_rate": 0.00016621451017938846, + "loss": 0.6876, + "step": 16807 + }, + { + "epoch": 1.1663312747206995, + "grad_norm": 1.7322129011154175, + "learning_rate": 0.0001662102996212578, + "loss": 0.6086, + "step": 16808 + }, + { + "epoch": 1.1664006661577961, + "grad_norm": 1.2422592639923096, + "learning_rate": 0.00016620608885410893, + "loss": 0.467, + "step": 16809 + }, + { + "epoch": 1.1664700575948928, + "grad_norm": 2.4336559772491455, + "learning_rate": 0.00016620187787795508, + "loss": 1.0341, + "step": 16810 + }, + { + "epoch": 1.1665394490319894, + "grad_norm": 1.614659070968628, + "learning_rate": 0.0001661976666928096, + "loss": 0.5588, + "step": 16811 + }, + { + "epoch": 1.1666088404690862, + "grad_norm": 1.5002994537353516, + "learning_rate": 0.00016619345529868577, + "loss": 0.4923, + "step": 16812 + }, + { + "epoch": 1.1666782319061828, + "grad_norm": 2.0447402000427246, + "learning_rate": 0.00016618924369559687, + "loss": 0.8638, + "step": 16813 + }, + { + "epoch": 1.1667476233432794, + "grad_norm": 1.5422440767288208, + "learning_rate": 0.00016618503188355624, + "loss": 0.5468, + "step": 16814 + }, + { + "epoch": 1.166817014780376, + "grad_norm": 2.354736566543579, + "learning_rate": 0.00016618081986257706, + "loss": 0.7025, + "step": 16815 + }, + { + "epoch": 1.1668864062174729, + "grad_norm": 1.914109468460083, + "learning_rate": 0.00016617660763267283, + "loss": 0.587, + "step": 16816 + }, + { + "epoch": 1.1669557976545695, + "grad_norm": 1.6413980722427368, + "learning_rate": 0.00016617239519385664, + "loss": 0.6233, + "step": 16817 + }, + { + "epoch": 1.167025189091666, + "grad_norm": 1.7932238578796387, + "learning_rate": 0.00016616818254614187, + "loss": 0.6713, + "step": 16818 + }, + { + "epoch": 1.1670945805287627, + "grad_norm": 1.859514594078064, + "learning_rate": 0.00016616396968954187, + "loss": 0.6276, + "step": 16819 + }, + { + "epoch": 1.1671639719658593, + "grad_norm": 2.4040961265563965, + "learning_rate": 0.00016615975662406988, + "loss": 0.6477, + "step": 16820 + }, + { + "epoch": 1.1672333634029561, + "grad_norm": 1.5052279233932495, + "learning_rate": 0.00016615554334973922, + "loss": 0.502, + "step": 16821 + }, + { + "epoch": 1.1673027548400527, + "grad_norm": 2.115391969680786, + "learning_rate": 0.00016615132986656317, + "loss": 0.6582, + "step": 16822 + }, + { + "epoch": 1.1673721462771494, + "grad_norm": 1.9971187114715576, + "learning_rate": 0.00016614711617455506, + "loss": 0.7032, + "step": 16823 + }, + { + "epoch": 1.167441537714246, + "grad_norm": 1.1409831047058105, + "learning_rate": 0.00016614290227372815, + "loss": 0.421, + "step": 16824 + }, + { + "epoch": 1.1675109291513428, + "grad_norm": 1.6679613590240479, + "learning_rate": 0.00016613868816409578, + "loss": 0.6688, + "step": 16825 + }, + { + "epoch": 1.1675803205884394, + "grad_norm": 1.9236458539962769, + "learning_rate": 0.0001661344738456713, + "loss": 0.5494, + "step": 16826 + }, + { + "epoch": 1.167649712025536, + "grad_norm": 1.775381326675415, + "learning_rate": 0.00016613025931846791, + "loss": 0.6517, + "step": 16827 + }, + { + "epoch": 1.1677191034626326, + "grad_norm": 1.8750487565994263, + "learning_rate": 0.00016612604458249896, + "loss": 0.617, + "step": 16828 + }, + { + "epoch": 1.1677884948997295, + "grad_norm": 1.7474743127822876, + "learning_rate": 0.00016612182963777778, + "loss": 0.5959, + "step": 16829 + }, + { + "epoch": 1.167857886336826, + "grad_norm": 2.455422878265381, + "learning_rate": 0.00016611761448431765, + "loss": 0.8272, + "step": 16830 + }, + { + "epoch": 1.1679272777739227, + "grad_norm": 1.5101007223129272, + "learning_rate": 0.0001661133991221319, + "loss": 0.6422, + "step": 16831 + }, + { + "epoch": 1.1679966692110193, + "grad_norm": 1.6464214324951172, + "learning_rate": 0.00016610918355123376, + "loss": 0.5362, + "step": 16832 + }, + { + "epoch": 1.168066060648116, + "grad_norm": 2.215541362762451, + "learning_rate": 0.00016610496777163665, + "loss": 0.8409, + "step": 16833 + }, + { + "epoch": 1.1681354520852127, + "grad_norm": 2.7268550395965576, + "learning_rate": 0.0001661007517833538, + "loss": 0.5513, + "step": 16834 + }, + { + "epoch": 1.1682048435223094, + "grad_norm": 1.5874202251434326, + "learning_rate": 0.00016609653558639854, + "loss": 0.6344, + "step": 16835 + }, + { + "epoch": 1.168274234959406, + "grad_norm": 1.8527323007583618, + "learning_rate": 0.00016609231918078421, + "loss": 0.6286, + "step": 16836 + }, + { + "epoch": 1.1683436263965026, + "grad_norm": 2.572284698486328, + "learning_rate": 0.00016608810256652412, + "loss": 0.5137, + "step": 16837 + }, + { + "epoch": 1.1684130178335994, + "grad_norm": 2.085191249847412, + "learning_rate": 0.00016608388574363152, + "loss": 0.5377, + "step": 16838 + }, + { + "epoch": 1.168482409270696, + "grad_norm": 1.8914614915847778, + "learning_rate": 0.00016607966871211977, + "loss": 0.7837, + "step": 16839 + }, + { + "epoch": 1.1685518007077926, + "grad_norm": 2.073028564453125, + "learning_rate": 0.00016607545147200216, + "loss": 0.5939, + "step": 16840 + }, + { + "epoch": 1.1686211921448892, + "grad_norm": 2.078094005584717, + "learning_rate": 0.00016607123402329204, + "loss": 0.6274, + "step": 16841 + }, + { + "epoch": 1.168690583581986, + "grad_norm": 1.8840562105178833, + "learning_rate": 0.00016606701636600268, + "loss": 0.6696, + "step": 16842 + }, + { + "epoch": 1.1687599750190827, + "grad_norm": 1.773586392402649, + "learning_rate": 0.0001660627985001474, + "loss": 0.5732, + "step": 16843 + }, + { + "epoch": 1.1688293664561793, + "grad_norm": 2.615950345993042, + "learning_rate": 0.00016605858042573955, + "loss": 0.9664, + "step": 16844 + }, + { + "epoch": 1.168898757893276, + "grad_norm": 1.7718747854232788, + "learning_rate": 0.00016605436214279243, + "loss": 0.5554, + "step": 16845 + }, + { + "epoch": 1.1689681493303725, + "grad_norm": 1.6077519655227661, + "learning_rate": 0.00016605014365131938, + "loss": 0.5327, + "step": 16846 + }, + { + "epoch": 1.1690375407674694, + "grad_norm": 1.9678621292114258, + "learning_rate": 0.00016604592495133365, + "loss": 0.5936, + "step": 16847 + }, + { + "epoch": 1.169106932204566, + "grad_norm": 1.9907881021499634, + "learning_rate": 0.0001660417060428486, + "loss": 0.8649, + "step": 16848 + }, + { + "epoch": 1.1691763236416626, + "grad_norm": 1.5688974857330322, + "learning_rate": 0.00016603748692587756, + "loss": 0.5793, + "step": 16849 + }, + { + "epoch": 1.1692457150787592, + "grad_norm": 3.800863027572632, + "learning_rate": 0.0001660332676004338, + "loss": 0.7341, + "step": 16850 + }, + { + "epoch": 1.169315106515856, + "grad_norm": 2.3404221534729004, + "learning_rate": 0.00016602904806653073, + "loss": 0.5839, + "step": 16851 + }, + { + "epoch": 1.1693844979529526, + "grad_norm": 1.4476240873336792, + "learning_rate": 0.0001660248283241816, + "loss": 0.5825, + "step": 16852 + }, + { + "epoch": 1.1694538893900492, + "grad_norm": 1.554936408996582, + "learning_rate": 0.0001660206083733997, + "loss": 0.6608, + "step": 16853 + }, + { + "epoch": 1.1695232808271459, + "grad_norm": 1.5581960678100586, + "learning_rate": 0.00016601638821419848, + "loss": 0.5136, + "step": 16854 + }, + { + "epoch": 1.1695926722642427, + "grad_norm": 3.059176445007324, + "learning_rate": 0.00016601216784659111, + "loss": 0.9942, + "step": 16855 + }, + { + "epoch": 1.1696620637013393, + "grad_norm": 1.6262391805648804, + "learning_rate": 0.000166007947270591, + "loss": 0.5511, + "step": 16856 + }, + { + "epoch": 1.169731455138436, + "grad_norm": 1.9796274900436401, + "learning_rate": 0.00016600372648621148, + "loss": 0.9691, + "step": 16857 + }, + { + "epoch": 1.1698008465755325, + "grad_norm": 1.7627030611038208, + "learning_rate": 0.00016599950549346586, + "loss": 0.6181, + "step": 16858 + }, + { + "epoch": 1.1698702380126291, + "grad_norm": 1.931044101715088, + "learning_rate": 0.00016599528429236742, + "loss": 0.8299, + "step": 16859 + }, + { + "epoch": 1.169939629449726, + "grad_norm": 3.170991897583008, + "learning_rate": 0.00016599106288292954, + "loss": 0.9373, + "step": 16860 + }, + { + "epoch": 1.1700090208868226, + "grad_norm": 1.9004403352737427, + "learning_rate": 0.00016598684126516554, + "loss": 0.727, + "step": 16861 + }, + { + "epoch": 1.1700784123239192, + "grad_norm": 1.401147484779358, + "learning_rate": 0.00016598261943908874, + "loss": 0.6755, + "step": 16862 + }, + { + "epoch": 1.170147803761016, + "grad_norm": 2.342885971069336, + "learning_rate": 0.0001659783974047125, + "loss": 0.5952, + "step": 16863 + }, + { + "epoch": 1.1702171951981126, + "grad_norm": 1.8886017799377441, + "learning_rate": 0.00016597417516205006, + "loss": 0.7745, + "step": 16864 + }, + { + "epoch": 1.1702865866352092, + "grad_norm": 1.953593134880066, + "learning_rate": 0.0001659699527111148, + "loss": 0.7297, + "step": 16865 + }, + { + "epoch": 1.1703559780723058, + "grad_norm": 1.928519606590271, + "learning_rate": 0.00016596573005192008, + "loss": 0.8257, + "step": 16866 + }, + { + "epoch": 1.1704253695094025, + "grad_norm": 1.9227615594863892, + "learning_rate": 0.0001659615071844792, + "loss": 0.7145, + "step": 16867 + }, + { + "epoch": 1.1704947609464993, + "grad_norm": 1.7343757152557373, + "learning_rate": 0.00016595728410880552, + "loss": 0.7179, + "step": 16868 + }, + { + "epoch": 1.170564152383596, + "grad_norm": 2.8131632804870605, + "learning_rate": 0.0001659530608249123, + "loss": 0.7354, + "step": 16869 + }, + { + "epoch": 1.1706335438206925, + "grad_norm": 1.6794387102127075, + "learning_rate": 0.0001659488373328129, + "loss": 0.6636, + "step": 16870 + }, + { + "epoch": 1.1707029352577891, + "grad_norm": 1.8816587924957275, + "learning_rate": 0.00016594461363252073, + "loss": 0.7074, + "step": 16871 + }, + { + "epoch": 1.1707723266948857, + "grad_norm": 1.724562644958496, + "learning_rate": 0.00016594038972404907, + "loss": 0.6318, + "step": 16872 + }, + { + "epoch": 1.1708417181319826, + "grad_norm": 2.769195318222046, + "learning_rate": 0.0001659361656074112, + "loss": 0.8514, + "step": 16873 + }, + { + "epoch": 1.1709111095690792, + "grad_norm": 1.8280798196792603, + "learning_rate": 0.00016593194128262053, + "loss": 0.7025, + "step": 16874 + }, + { + "epoch": 1.1709805010061758, + "grad_norm": 1.4885221719741821, + "learning_rate": 0.0001659277167496904, + "loss": 0.5881, + "step": 16875 + }, + { + "epoch": 1.1710498924432726, + "grad_norm": 2.0169084072113037, + "learning_rate": 0.00016592349200863406, + "loss": 0.7125, + "step": 16876 + }, + { + "epoch": 1.1711192838803692, + "grad_norm": 2.6340513229370117, + "learning_rate": 0.00016591926705946492, + "loss": 1.0122, + "step": 16877 + }, + { + "epoch": 1.1711886753174658, + "grad_norm": 3.2232825756073, + "learning_rate": 0.00016591504190219633, + "loss": 0.466, + "step": 16878 + }, + { + "epoch": 1.1712580667545625, + "grad_norm": 3.5056653022766113, + "learning_rate": 0.00016591081653684158, + "loss": 0.8594, + "step": 16879 + }, + { + "epoch": 1.171327458191659, + "grad_norm": 2.2931578159332275, + "learning_rate": 0.00016590659096341404, + "loss": 0.7846, + "step": 16880 + }, + { + "epoch": 1.171396849628756, + "grad_norm": 2.180130958557129, + "learning_rate": 0.00016590236518192704, + "loss": 0.7992, + "step": 16881 + }, + { + "epoch": 1.1714662410658525, + "grad_norm": 2.0437448024749756, + "learning_rate": 0.0001658981391923939, + "loss": 0.7075, + "step": 16882 + }, + { + "epoch": 1.1715356325029491, + "grad_norm": 1.840673565864563, + "learning_rate": 0.000165893912994828, + "loss": 0.6264, + "step": 16883 + }, + { + "epoch": 1.1716050239400457, + "grad_norm": 1.7660353183746338, + "learning_rate": 0.00016588968658924266, + "loss": 0.6278, + "step": 16884 + }, + { + "epoch": 1.1716744153771423, + "grad_norm": 2.1151514053344727, + "learning_rate": 0.0001658854599756512, + "loss": 0.6911, + "step": 16885 + }, + { + "epoch": 1.1717438068142392, + "grad_norm": 2.2609121799468994, + "learning_rate": 0.00016588123315406704, + "loss": 0.7591, + "step": 16886 + }, + { + "epoch": 1.1718131982513358, + "grad_norm": 3.128918170928955, + "learning_rate": 0.00016587700612450342, + "loss": 0.5991, + "step": 16887 + }, + { + "epoch": 1.1718825896884324, + "grad_norm": 1.8308429718017578, + "learning_rate": 0.00016587277888697376, + "loss": 0.5244, + "step": 16888 + }, + { + "epoch": 1.1719519811255292, + "grad_norm": 2.030027151107788, + "learning_rate": 0.00016586855144149135, + "loss": 0.7426, + "step": 16889 + }, + { + "epoch": 1.1720213725626258, + "grad_norm": 1.9422703981399536, + "learning_rate": 0.00016586432378806957, + "loss": 1.0, + "step": 16890 + }, + { + "epoch": 1.1720907639997225, + "grad_norm": 2.3786046504974365, + "learning_rate": 0.0001658600959267218, + "loss": 0.5769, + "step": 16891 + }, + { + "epoch": 1.172160155436819, + "grad_norm": 1.647910475730896, + "learning_rate": 0.00016585586785746131, + "loss": 0.5172, + "step": 16892 + }, + { + "epoch": 1.1722295468739157, + "grad_norm": 2.320570945739746, + "learning_rate": 0.0001658516395803015, + "loss": 0.7254, + "step": 16893 + }, + { + "epoch": 1.1722989383110125, + "grad_norm": 1.8827457427978516, + "learning_rate": 0.0001658474110952557, + "loss": 0.7093, + "step": 16894 + }, + { + "epoch": 1.1723683297481091, + "grad_norm": 1.691706657409668, + "learning_rate": 0.00016584318240233728, + "loss": 0.7025, + "step": 16895 + }, + { + "epoch": 1.1724377211852057, + "grad_norm": 1.8924121856689453, + "learning_rate": 0.00016583895350155955, + "loss": 0.7621, + "step": 16896 + }, + { + "epoch": 1.1725071126223023, + "grad_norm": 1.963342547416687, + "learning_rate": 0.0001658347243929359, + "loss": 0.6056, + "step": 16897 + }, + { + "epoch": 1.172576504059399, + "grad_norm": 2.163609027862549, + "learning_rate": 0.00016583049507647965, + "loss": 0.5642, + "step": 16898 + }, + { + "epoch": 1.1726458954964958, + "grad_norm": 1.7439700365066528, + "learning_rate": 0.00016582626555220416, + "loss": 0.7133, + "step": 16899 + }, + { + "epoch": 1.1727152869335924, + "grad_norm": 1.3572659492492676, + "learning_rate": 0.00016582203582012284, + "loss": 0.5587, + "step": 16900 + }, + { + "epoch": 1.172784678370689, + "grad_norm": 1.795747995376587, + "learning_rate": 0.00016581780588024894, + "loss": 0.6355, + "step": 16901 + }, + { + "epoch": 1.1728540698077858, + "grad_norm": 1.8192429542541504, + "learning_rate": 0.00016581357573259585, + "loss": 0.6688, + "step": 16902 + }, + { + "epoch": 1.1729234612448824, + "grad_norm": 1.6811676025390625, + "learning_rate": 0.00016580934537717697, + "loss": 0.7351, + "step": 16903 + }, + { + "epoch": 1.172992852681979, + "grad_norm": 1.5303293466567993, + "learning_rate": 0.00016580511481400562, + "loss": 0.6629, + "step": 16904 + }, + { + "epoch": 1.1730622441190757, + "grad_norm": 2.157353401184082, + "learning_rate": 0.00016580088404309512, + "loss": 0.7038, + "step": 16905 + }, + { + "epoch": 1.1731316355561723, + "grad_norm": 2.0273380279541016, + "learning_rate": 0.0001657966530644589, + "loss": 0.7582, + "step": 16906 + }, + { + "epoch": 1.1732010269932691, + "grad_norm": 1.9592019319534302, + "learning_rate": 0.00016579242187811027, + "loss": 0.7946, + "step": 16907 + }, + { + "epoch": 1.1732704184303657, + "grad_norm": 1.8660976886749268, + "learning_rate": 0.00016578819048406262, + "loss": 0.6106, + "step": 16908 + }, + { + "epoch": 1.1733398098674623, + "grad_norm": 2.6377415657043457, + "learning_rate": 0.00016578395888232925, + "loss": 0.8918, + "step": 16909 + }, + { + "epoch": 1.173409201304559, + "grad_norm": 1.9610191583633423, + "learning_rate": 0.00016577972707292358, + "loss": 0.9151, + "step": 16910 + }, + { + "epoch": 1.1734785927416556, + "grad_norm": 2.6034936904907227, + "learning_rate": 0.00016577549505585891, + "loss": 0.7198, + "step": 16911 + }, + { + "epoch": 1.1735479841787524, + "grad_norm": 3.016261100769043, + "learning_rate": 0.00016577126283114866, + "loss": 0.9038, + "step": 16912 + }, + { + "epoch": 1.173617375615849, + "grad_norm": 2.2681939601898193, + "learning_rate": 0.00016576703039880615, + "loss": 0.7365, + "step": 16913 + }, + { + "epoch": 1.1736867670529456, + "grad_norm": 2.129260540008545, + "learning_rate": 0.00016576279775884479, + "loss": 0.6214, + "step": 16914 + }, + { + "epoch": 1.1737561584900424, + "grad_norm": 2.1490190029144287, + "learning_rate": 0.0001657585649112779, + "loss": 0.7724, + "step": 16915 + }, + { + "epoch": 1.173825549927139, + "grad_norm": 1.454481840133667, + "learning_rate": 0.0001657543318561188, + "loss": 0.4908, + "step": 16916 + }, + { + "epoch": 1.1738949413642357, + "grad_norm": 1.8010296821594238, + "learning_rate": 0.00016575009859338093, + "loss": 0.5935, + "step": 16917 + }, + { + "epoch": 1.1739643328013323, + "grad_norm": 1.5500911474227905, + "learning_rate": 0.00016574586512307763, + "loss": 0.5673, + "step": 16918 + }, + { + "epoch": 1.1740337242384289, + "grad_norm": 2.306171417236328, + "learning_rate": 0.00016574163144522227, + "loss": 0.8113, + "step": 16919 + }, + { + "epoch": 1.1741031156755257, + "grad_norm": 1.8223121166229248, + "learning_rate": 0.00016573739755982821, + "loss": 0.7505, + "step": 16920 + }, + { + "epoch": 1.1741725071126223, + "grad_norm": 1.8399120569229126, + "learning_rate": 0.0001657331634669088, + "loss": 0.6741, + "step": 16921 + }, + { + "epoch": 1.174241898549719, + "grad_norm": 2.2039294242858887, + "learning_rate": 0.00016572892916647743, + "loss": 0.7614, + "step": 16922 + }, + { + "epoch": 1.1743112899868156, + "grad_norm": 1.38357412815094, + "learning_rate": 0.00016572469465854744, + "loss": 0.5321, + "step": 16923 + }, + { + "epoch": 1.1743806814239122, + "grad_norm": 1.700024962425232, + "learning_rate": 0.00016572045994313222, + "loss": 0.508, + "step": 16924 + }, + { + "epoch": 1.174450072861009, + "grad_norm": 2.535069227218628, + "learning_rate": 0.00016571622502024514, + "loss": 0.8036, + "step": 16925 + }, + { + "epoch": 1.1745194642981056, + "grad_norm": 1.4807751178741455, + "learning_rate": 0.00016571198988989954, + "loss": 0.7101, + "step": 16926 + }, + { + "epoch": 1.1745888557352022, + "grad_norm": 1.5972951650619507, + "learning_rate": 0.00016570775455210886, + "loss": 0.6427, + "step": 16927 + }, + { + "epoch": 1.174658247172299, + "grad_norm": 2.141679286956787, + "learning_rate": 0.0001657035190068864, + "loss": 0.7203, + "step": 16928 + }, + { + "epoch": 1.1747276386093957, + "grad_norm": 2.6364669799804688, + "learning_rate": 0.00016569928325424553, + "loss": 0.7092, + "step": 16929 + }, + { + "epoch": 1.1747970300464923, + "grad_norm": 1.8332029581069946, + "learning_rate": 0.00016569504729419966, + "loss": 0.6396, + "step": 16930 + }, + { + "epoch": 1.1748664214835889, + "grad_norm": 1.4692108631134033, + "learning_rate": 0.00016569081112676216, + "loss": 0.5437, + "step": 16931 + }, + { + "epoch": 1.1749358129206855, + "grad_norm": 1.433674693107605, + "learning_rate": 0.00016568657475194635, + "loss": 0.5919, + "step": 16932 + }, + { + "epoch": 1.1750052043577823, + "grad_norm": 1.6167643070220947, + "learning_rate": 0.00016568233816976569, + "loss": 0.6127, + "step": 16933 + }, + { + "epoch": 1.175074595794879, + "grad_norm": 2.114476442337036, + "learning_rate": 0.0001656781013802335, + "loss": 0.6184, + "step": 16934 + }, + { + "epoch": 1.1751439872319756, + "grad_norm": 1.6892529726028442, + "learning_rate": 0.00016567386438336317, + "loss": 0.4755, + "step": 16935 + }, + { + "epoch": 1.1752133786690722, + "grad_norm": 1.3280375003814697, + "learning_rate": 0.00016566962717916805, + "loss": 0.504, + "step": 16936 + }, + { + "epoch": 1.175282770106169, + "grad_norm": 1.5133113861083984, + "learning_rate": 0.00016566538976766154, + "loss": 0.7367, + "step": 16937 + }, + { + "epoch": 1.1753521615432656, + "grad_norm": 1.2478721141815186, + "learning_rate": 0.00016566115214885702, + "loss": 0.566, + "step": 16938 + }, + { + "epoch": 1.1754215529803622, + "grad_norm": 2.7224960327148438, + "learning_rate": 0.00016565691432276786, + "loss": 0.736, + "step": 16939 + }, + { + "epoch": 1.1754909444174588, + "grad_norm": 1.4496079683303833, + "learning_rate": 0.00016565267628940742, + "loss": 0.6493, + "step": 16940 + }, + { + "epoch": 1.1755603358545557, + "grad_norm": 1.6394729614257812, + "learning_rate": 0.00016564843804878915, + "loss": 0.5234, + "step": 16941 + }, + { + "epoch": 1.1756297272916523, + "grad_norm": 2.8758111000061035, + "learning_rate": 0.0001656441996009263, + "loss": 0.7881, + "step": 16942 + }, + { + "epoch": 1.1756991187287489, + "grad_norm": 1.4791010618209839, + "learning_rate": 0.00016563996094583238, + "loss": 0.5546, + "step": 16943 + }, + { + "epoch": 1.1757685101658455, + "grad_norm": 2.079305410385132, + "learning_rate": 0.00016563572208352074, + "loss": 0.6743, + "step": 16944 + }, + { + "epoch": 1.175837901602942, + "grad_norm": 2.228652000427246, + "learning_rate": 0.0001656314830140047, + "loss": 0.7845, + "step": 16945 + }, + { + "epoch": 1.175907293040039, + "grad_norm": 1.575354814529419, + "learning_rate": 0.0001656272437372977, + "loss": 0.6138, + "step": 16946 + }, + { + "epoch": 1.1759766844771355, + "grad_norm": 1.7044031620025635, + "learning_rate": 0.00016562300425341307, + "loss": 0.6033, + "step": 16947 + }, + { + "epoch": 1.1760460759142322, + "grad_norm": 1.6581274271011353, + "learning_rate": 0.00016561876456236426, + "loss": 0.6414, + "step": 16948 + }, + { + "epoch": 1.1761154673513288, + "grad_norm": 1.8662424087524414, + "learning_rate": 0.00016561452466416462, + "loss": 0.5209, + "step": 16949 + }, + { + "epoch": 1.1761848587884256, + "grad_norm": 2.414489507675171, + "learning_rate": 0.00016561028455882757, + "loss": 1.0244, + "step": 16950 + }, + { + "epoch": 1.1762542502255222, + "grad_norm": 1.4089624881744385, + "learning_rate": 0.00016560604424636644, + "loss": 0.6239, + "step": 16951 + }, + { + "epoch": 1.1763236416626188, + "grad_norm": 2.433349132537842, + "learning_rate": 0.00016560180372679462, + "loss": 0.6917, + "step": 16952 + }, + { + "epoch": 1.1763930330997154, + "grad_norm": 2.158940553665161, + "learning_rate": 0.00016559756300012554, + "loss": 0.8754, + "step": 16953 + }, + { + "epoch": 1.1764624245368123, + "grad_norm": 2.1946160793304443, + "learning_rate": 0.00016559332206637255, + "loss": 0.8261, + "step": 16954 + }, + { + "epoch": 1.1765318159739089, + "grad_norm": 2.261220932006836, + "learning_rate": 0.00016558908092554906, + "loss": 0.8132, + "step": 16955 + }, + { + "epoch": 1.1766012074110055, + "grad_norm": 1.7014796733856201, + "learning_rate": 0.00016558483957766843, + "loss": 0.6133, + "step": 16956 + }, + { + "epoch": 1.176670598848102, + "grad_norm": 1.7572910785675049, + "learning_rate": 0.00016558059802274411, + "loss": 0.6895, + "step": 16957 + }, + { + "epoch": 1.1767399902851987, + "grad_norm": 1.8015973567962646, + "learning_rate": 0.0001655763562607894, + "loss": 0.7993, + "step": 16958 + }, + { + "epoch": 1.1768093817222955, + "grad_norm": 1.8140937089920044, + "learning_rate": 0.0001655721142918178, + "loss": 0.4227, + "step": 16959 + }, + { + "epoch": 1.1768787731593922, + "grad_norm": 1.8122532367706299, + "learning_rate": 0.0001655678721158426, + "loss": 0.721, + "step": 16960 + }, + { + "epoch": 1.1769481645964888, + "grad_norm": 1.7195563316345215, + "learning_rate": 0.00016556362973287728, + "loss": 0.4876, + "step": 16961 + }, + { + "epoch": 1.1770175560335854, + "grad_norm": 2.4262895584106445, + "learning_rate": 0.00016555938714293516, + "loss": 0.9032, + "step": 16962 + }, + { + "epoch": 1.1770869474706822, + "grad_norm": 2.9091222286224365, + "learning_rate": 0.00016555514434602963, + "loss": 0.7166, + "step": 16963 + }, + { + "epoch": 1.1771563389077788, + "grad_norm": 2.4210617542266846, + "learning_rate": 0.00016555090134217414, + "loss": 0.8101, + "step": 16964 + }, + { + "epoch": 1.1772257303448754, + "grad_norm": 1.5921260118484497, + "learning_rate": 0.00016554665813138208, + "loss": 0.4912, + "step": 16965 + }, + { + "epoch": 1.177295121781972, + "grad_norm": 1.5828547477722168, + "learning_rate": 0.00016554241471366678, + "loss": 0.5059, + "step": 16966 + }, + { + "epoch": 1.1773645132190689, + "grad_norm": 2.0941717624664307, + "learning_rate": 0.0001655381710890417, + "loss": 0.6557, + "step": 16967 + }, + { + "epoch": 1.1774339046561655, + "grad_norm": 1.7410389184951782, + "learning_rate": 0.00016553392725752027, + "loss": 0.6035, + "step": 16968 + }, + { + "epoch": 1.177503296093262, + "grad_norm": 1.4038398265838623, + "learning_rate": 0.00016552968321911578, + "loss": 0.56, + "step": 16969 + }, + { + "epoch": 1.1775726875303587, + "grad_norm": 1.3539535999298096, + "learning_rate": 0.0001655254389738417, + "loss": 0.5322, + "step": 16970 + }, + { + "epoch": 1.1776420789674553, + "grad_norm": 2.197617530822754, + "learning_rate": 0.0001655211945217114, + "loss": 0.7768, + "step": 16971 + }, + { + "epoch": 1.1777114704045522, + "grad_norm": 1.6791983842849731, + "learning_rate": 0.0001655169498627383, + "loss": 0.5323, + "step": 16972 + }, + { + "epoch": 1.1777808618416488, + "grad_norm": 1.9109140634536743, + "learning_rate": 0.00016551270499693579, + "loss": 0.7438, + "step": 16973 + }, + { + "epoch": 1.1778502532787454, + "grad_norm": 2.991865634918213, + "learning_rate": 0.00016550845992431725, + "loss": 0.9283, + "step": 16974 + }, + { + "epoch": 1.177919644715842, + "grad_norm": 1.9521605968475342, + "learning_rate": 0.00016550421464489614, + "loss": 0.6171, + "step": 16975 + }, + { + "epoch": 1.1779890361529388, + "grad_norm": 2.125603199005127, + "learning_rate": 0.0001654999691586858, + "loss": 0.7038, + "step": 16976 + }, + { + "epoch": 1.1780584275900354, + "grad_norm": 2.494637966156006, + "learning_rate": 0.00016549572346569966, + "loss": 0.7053, + "step": 16977 + }, + { + "epoch": 1.178127819027132, + "grad_norm": 1.8216601610183716, + "learning_rate": 0.00016549147756595112, + "loss": 0.4984, + "step": 16978 + }, + { + "epoch": 1.1781972104642287, + "grad_norm": 1.7726982831954956, + "learning_rate": 0.0001654872314594536, + "loss": 0.612, + "step": 16979 + }, + { + "epoch": 1.1782666019013255, + "grad_norm": 1.7547742128372192, + "learning_rate": 0.00016548298514622046, + "loss": 0.6097, + "step": 16980 + }, + { + "epoch": 1.178335993338422, + "grad_norm": 2.462155342102051, + "learning_rate": 0.00016547873862626515, + "loss": 0.6244, + "step": 16981 + }, + { + "epoch": 1.1784053847755187, + "grad_norm": 1.8991144895553589, + "learning_rate": 0.00016547449189960106, + "loss": 0.7277, + "step": 16982 + }, + { + "epoch": 1.1784747762126153, + "grad_norm": 1.8579775094985962, + "learning_rate": 0.0001654702449662416, + "loss": 0.7178, + "step": 16983 + }, + { + "epoch": 1.178544167649712, + "grad_norm": 1.7016481161117554, + "learning_rate": 0.00016546599782620017, + "loss": 0.7406, + "step": 16984 + }, + { + "epoch": 1.1786135590868088, + "grad_norm": 1.7687315940856934, + "learning_rate": 0.00016546175047949015, + "loss": 0.5319, + "step": 16985 + }, + { + "epoch": 1.1786829505239054, + "grad_norm": 2.22163724899292, + "learning_rate": 0.00016545750292612502, + "loss": 0.7298, + "step": 16986 + }, + { + "epoch": 1.178752341961002, + "grad_norm": 2.049527406692505, + "learning_rate": 0.00016545325516611814, + "loss": 0.7051, + "step": 16987 + }, + { + "epoch": 1.1788217333980986, + "grad_norm": 2.1041676998138428, + "learning_rate": 0.00016544900719948292, + "loss": 0.6065, + "step": 16988 + }, + { + "epoch": 1.1788911248351954, + "grad_norm": 1.7550820112228394, + "learning_rate": 0.00016544475902623279, + "loss": 0.8171, + "step": 16989 + }, + { + "epoch": 1.178960516272292, + "grad_norm": 1.896552562713623, + "learning_rate": 0.00016544051064638113, + "loss": 0.6234, + "step": 16990 + }, + { + "epoch": 1.1790299077093886, + "grad_norm": 1.7550387382507324, + "learning_rate": 0.00016543626205994135, + "loss": 0.6607, + "step": 16991 + }, + { + "epoch": 1.1790992991464853, + "grad_norm": 1.958144187927246, + "learning_rate": 0.00016543201326692692, + "loss": 0.6864, + "step": 16992 + }, + { + "epoch": 1.179168690583582, + "grad_norm": 2.096402883529663, + "learning_rate": 0.0001654277642673512, + "loss": 0.8011, + "step": 16993 + }, + { + "epoch": 1.1792380820206787, + "grad_norm": 1.4981889724731445, + "learning_rate": 0.00016542351506122763, + "loss": 0.5334, + "step": 16994 + }, + { + "epoch": 1.1793074734577753, + "grad_norm": 2.317199468612671, + "learning_rate": 0.0001654192656485696, + "loss": 0.5636, + "step": 16995 + }, + { + "epoch": 1.179376864894872, + "grad_norm": 1.9126702547073364, + "learning_rate": 0.00016541501602939054, + "loss": 0.687, + "step": 16996 + }, + { + "epoch": 1.1794462563319685, + "grad_norm": 1.784036636352539, + "learning_rate": 0.0001654107662037039, + "loss": 0.747, + "step": 16997 + }, + { + "epoch": 1.1795156477690654, + "grad_norm": 2.063896417617798, + "learning_rate": 0.000165406516171523, + "loss": 0.7209, + "step": 16998 + }, + { + "epoch": 1.179585039206162, + "grad_norm": 2.1127214431762695, + "learning_rate": 0.00016540226593286136, + "loss": 0.8022, + "step": 16999 + }, + { + "epoch": 1.1796544306432586, + "grad_norm": 2.1061604022979736, + "learning_rate": 0.00016539801548773236, + "loss": 0.8993, + "step": 17000 + }, + { + "epoch": 1.1797238220803552, + "grad_norm": 2.5988247394561768, + "learning_rate": 0.00016539376483614936, + "loss": 0.8911, + "step": 17001 + }, + { + "epoch": 1.179793213517452, + "grad_norm": 1.3518123626708984, + "learning_rate": 0.00016538951397812587, + "loss": 0.5705, + "step": 17002 + }, + { + "epoch": 1.1798626049545486, + "grad_norm": 2.1603598594665527, + "learning_rate": 0.00016538526291367528, + "loss": 0.4395, + "step": 17003 + }, + { + "epoch": 1.1799319963916453, + "grad_norm": 1.7804704904556274, + "learning_rate": 0.000165381011642811, + "loss": 0.6495, + "step": 17004 + }, + { + "epoch": 1.1800013878287419, + "grad_norm": 2.3771166801452637, + "learning_rate": 0.0001653767601655464, + "loss": 0.6104, + "step": 17005 + }, + { + "epoch": 1.1800707792658387, + "grad_norm": 2.0933711528778076, + "learning_rate": 0.00016537250848189502, + "loss": 0.7557, + "step": 17006 + }, + { + "epoch": 1.1801401707029353, + "grad_norm": 1.723840594291687, + "learning_rate": 0.0001653682565918702, + "loss": 0.5407, + "step": 17007 + }, + { + "epoch": 1.180209562140032, + "grad_norm": 1.5522844791412354, + "learning_rate": 0.00016536400449548535, + "loss": 0.4841, + "step": 17008 + }, + { + "epoch": 1.1802789535771285, + "grad_norm": 2.563270092010498, + "learning_rate": 0.00016535975219275395, + "loss": 0.7408, + "step": 17009 + }, + { + "epoch": 1.1803483450142251, + "grad_norm": 2.3058910369873047, + "learning_rate": 0.00016535549968368937, + "loss": 0.6599, + "step": 17010 + }, + { + "epoch": 1.180417736451322, + "grad_norm": 2.374847412109375, + "learning_rate": 0.0001653512469683051, + "loss": 0.5238, + "step": 17011 + }, + { + "epoch": 1.1804871278884186, + "grad_norm": 1.371233582496643, + "learning_rate": 0.00016534699404661447, + "loss": 0.5171, + "step": 17012 + }, + { + "epoch": 1.1805565193255152, + "grad_norm": 2.4467029571533203, + "learning_rate": 0.000165342740918631, + "loss": 0.6222, + "step": 17013 + }, + { + "epoch": 1.1806259107626118, + "grad_norm": 1.9283065795898438, + "learning_rate": 0.00016533848758436807, + "loss": 0.7633, + "step": 17014 + }, + { + "epoch": 1.1806953021997086, + "grad_norm": 1.889547348022461, + "learning_rate": 0.00016533423404383913, + "loss": 0.5175, + "step": 17015 + }, + { + "epoch": 1.1807646936368053, + "grad_norm": 2.270991325378418, + "learning_rate": 0.00016532998029705756, + "loss": 0.6893, + "step": 17016 + }, + { + "epoch": 1.1808340850739019, + "grad_norm": 1.6728092432022095, + "learning_rate": 0.00016532572634403683, + "loss": 0.5861, + "step": 17017 + }, + { + "epoch": 1.1809034765109985, + "grad_norm": 2.406146287918091, + "learning_rate": 0.00016532147218479038, + "loss": 0.72, + "step": 17018 + }, + { + "epoch": 1.1809728679480953, + "grad_norm": 2.024411678314209, + "learning_rate": 0.00016531721781933159, + "loss": 0.6896, + "step": 17019 + }, + { + "epoch": 1.181042259385192, + "grad_norm": 2.258192539215088, + "learning_rate": 0.00016531296324767396, + "loss": 0.7466, + "step": 17020 + }, + { + "epoch": 1.1811116508222885, + "grad_norm": 1.6895424127578735, + "learning_rate": 0.00016530870846983087, + "loss": 0.6758, + "step": 17021 + }, + { + "epoch": 1.1811810422593851, + "grad_norm": 2.4077365398406982, + "learning_rate": 0.00016530445348581573, + "loss": 0.7902, + "step": 17022 + }, + { + "epoch": 1.1812504336964817, + "grad_norm": 3.0705642700195312, + "learning_rate": 0.00016530019829564204, + "loss": 0.7856, + "step": 17023 + }, + { + "epoch": 1.1813198251335786, + "grad_norm": 2.2222745418548584, + "learning_rate": 0.00016529594289932322, + "loss": 0.6474, + "step": 17024 + }, + { + "epoch": 1.1813892165706752, + "grad_norm": 2.6329760551452637, + "learning_rate": 0.00016529168729687265, + "loss": 0.7365, + "step": 17025 + }, + { + "epoch": 1.1814586080077718, + "grad_norm": 2.3798160552978516, + "learning_rate": 0.00016528743148830382, + "loss": 0.687, + "step": 17026 + }, + { + "epoch": 1.1815279994448684, + "grad_norm": 2.0410728454589844, + "learning_rate": 0.00016528317547363016, + "loss": 0.6147, + "step": 17027 + }, + { + "epoch": 1.1815973908819652, + "grad_norm": 1.4465337991714478, + "learning_rate": 0.00016527891925286503, + "loss": 0.6447, + "step": 17028 + }, + { + "epoch": 1.1816667823190619, + "grad_norm": 2.011228561401367, + "learning_rate": 0.00016527466282602196, + "loss": 0.7179, + "step": 17029 + }, + { + "epoch": 1.1817361737561585, + "grad_norm": 1.639431357383728, + "learning_rate": 0.00016527040619311432, + "loss": 0.5948, + "step": 17030 + }, + { + "epoch": 1.181805565193255, + "grad_norm": 1.5450645685195923, + "learning_rate": 0.00016526614935415563, + "loss": 0.6595, + "step": 17031 + }, + { + "epoch": 1.181874956630352, + "grad_norm": 2.196078300476074, + "learning_rate": 0.00016526189230915928, + "loss": 0.8089, + "step": 17032 + }, + { + "epoch": 1.1819443480674485, + "grad_norm": 1.7246019840240479, + "learning_rate": 0.00016525763505813868, + "loss": 0.7884, + "step": 17033 + }, + { + "epoch": 1.1820137395045451, + "grad_norm": 1.775648832321167, + "learning_rate": 0.0001652533776011073, + "loss": 0.6378, + "step": 17034 + }, + { + "epoch": 1.1820831309416417, + "grad_norm": 2.2427992820739746, + "learning_rate": 0.00016524911993807856, + "loss": 0.7604, + "step": 17035 + }, + { + "epoch": 1.1821525223787384, + "grad_norm": 2.7174057960510254, + "learning_rate": 0.00016524486206906595, + "loss": 0.8285, + "step": 17036 + }, + { + "epoch": 1.1822219138158352, + "grad_norm": 1.541300892829895, + "learning_rate": 0.0001652406039940829, + "loss": 0.5314, + "step": 17037 + }, + { + "epoch": 1.1822913052529318, + "grad_norm": 1.824359655380249, + "learning_rate": 0.0001652363457131428, + "loss": 0.6517, + "step": 17038 + }, + { + "epoch": 1.1823606966900284, + "grad_norm": 1.7297015190124512, + "learning_rate": 0.0001652320872262591, + "loss": 0.574, + "step": 17039 + }, + { + "epoch": 1.182430088127125, + "grad_norm": 1.4895597696304321, + "learning_rate": 0.00016522782853344532, + "loss": 0.6632, + "step": 17040 + }, + { + "epoch": 1.1824994795642219, + "grad_norm": 1.5224230289459229, + "learning_rate": 0.00016522356963471482, + "loss": 0.5941, + "step": 17041 + }, + { + "epoch": 1.1825688710013185, + "grad_norm": 1.868804931640625, + "learning_rate": 0.0001652193105300811, + "loss": 0.7923, + "step": 17042 + }, + { + "epoch": 1.182638262438415, + "grad_norm": 1.9942647218704224, + "learning_rate": 0.00016521505121955756, + "loss": 0.6121, + "step": 17043 + }, + { + "epoch": 1.1827076538755117, + "grad_norm": 1.5357404947280884, + "learning_rate": 0.0001652107917031577, + "loss": 0.6957, + "step": 17044 + }, + { + "epoch": 1.1827770453126085, + "grad_norm": 2.115591526031494, + "learning_rate": 0.0001652065319808949, + "loss": 0.7281, + "step": 17045 + }, + { + "epoch": 1.1828464367497051, + "grad_norm": 2.1543619632720947, + "learning_rate": 0.00016520227205278268, + "loss": 0.6319, + "step": 17046 + }, + { + "epoch": 1.1829158281868017, + "grad_norm": 1.6003165245056152, + "learning_rate": 0.0001651980119188344, + "loss": 0.6235, + "step": 17047 + }, + { + "epoch": 1.1829852196238984, + "grad_norm": 2.072816848754883, + "learning_rate": 0.00016519375157906364, + "loss": 0.6764, + "step": 17048 + }, + { + "epoch": 1.183054611060995, + "grad_norm": 1.911847472190857, + "learning_rate": 0.00016518949103348372, + "loss": 0.8059, + "step": 17049 + }, + { + "epoch": 1.1831240024980918, + "grad_norm": 1.945177435874939, + "learning_rate": 0.00016518523028210815, + "loss": 0.7016, + "step": 17050 + }, + { + "epoch": 1.1831933939351884, + "grad_norm": 1.688759446144104, + "learning_rate": 0.00016518096932495037, + "loss": 0.6445, + "step": 17051 + }, + { + "epoch": 1.183262785372285, + "grad_norm": 1.8221137523651123, + "learning_rate": 0.00016517670816202382, + "loss": 0.6338, + "step": 17052 + }, + { + "epoch": 1.1833321768093816, + "grad_norm": 1.3892267942428589, + "learning_rate": 0.00016517244679334196, + "loss": 0.4753, + "step": 17053 + }, + { + "epoch": 1.1834015682464785, + "grad_norm": 1.8130261898040771, + "learning_rate": 0.00016516818521891826, + "loss": 1.0085, + "step": 17054 + }, + { + "epoch": 1.183470959683575, + "grad_norm": 2.0850117206573486, + "learning_rate": 0.00016516392343876617, + "loss": 0.7791, + "step": 17055 + }, + { + "epoch": 1.1835403511206717, + "grad_norm": 1.834972858428955, + "learning_rate": 0.0001651596614528991, + "loss": 0.7681, + "step": 17056 + }, + { + "epoch": 1.1836097425577683, + "grad_norm": 3.944828748703003, + "learning_rate": 0.0001651553992613306, + "loss": 0.5993, + "step": 17057 + }, + { + "epoch": 1.1836791339948651, + "grad_norm": 1.588479995727539, + "learning_rate": 0.00016515113686407402, + "loss": 0.7343, + "step": 17058 + }, + { + "epoch": 1.1837485254319617, + "grad_norm": 2.268629789352417, + "learning_rate": 0.00016514687426114289, + "loss": 0.775, + "step": 17059 + }, + { + "epoch": 1.1838179168690584, + "grad_norm": 2.187037706375122, + "learning_rate": 0.00016514261145255058, + "loss": 0.8512, + "step": 17060 + }, + { + "epoch": 1.183887308306155, + "grad_norm": 1.4557307958602905, + "learning_rate": 0.00016513834843831062, + "loss": 0.5685, + "step": 17061 + }, + { + "epoch": 1.1839566997432516, + "grad_norm": 1.3110063076019287, + "learning_rate": 0.0001651340852184365, + "loss": 0.5464, + "step": 17062 + }, + { + "epoch": 1.1840260911803484, + "grad_norm": 1.466802954673767, + "learning_rate": 0.00016512982179294158, + "loss": 0.5945, + "step": 17063 + }, + { + "epoch": 1.184095482617445, + "grad_norm": 2.181473731994629, + "learning_rate": 0.00016512555816183938, + "loss": 0.7139, + "step": 17064 + }, + { + "epoch": 1.1841648740545416, + "grad_norm": 1.7803221940994263, + "learning_rate": 0.00016512129432514337, + "loss": 0.5974, + "step": 17065 + }, + { + "epoch": 1.1842342654916382, + "grad_norm": 2.1544315814971924, + "learning_rate": 0.00016511703028286698, + "loss": 0.8334, + "step": 17066 + }, + { + "epoch": 1.184303656928735, + "grad_norm": 1.967494010925293, + "learning_rate": 0.00016511276603502366, + "loss": 0.8351, + "step": 17067 + }, + { + "epoch": 1.1843730483658317, + "grad_norm": 1.8986008167266846, + "learning_rate": 0.00016510850158162692, + "loss": 0.9513, + "step": 17068 + }, + { + "epoch": 1.1844424398029283, + "grad_norm": 1.8431819677352905, + "learning_rate": 0.0001651042369226902, + "loss": 0.6584, + "step": 17069 + }, + { + "epoch": 1.184511831240025, + "grad_norm": 1.8086875677108765, + "learning_rate": 0.00016509997205822692, + "loss": 0.6161, + "step": 17070 + }, + { + "epoch": 1.1845812226771217, + "grad_norm": 1.7842390537261963, + "learning_rate": 0.0001650957069882506, + "loss": 0.6973, + "step": 17071 + }, + { + "epoch": 1.1846506141142183, + "grad_norm": 1.882445216178894, + "learning_rate": 0.0001650914417127747, + "loss": 0.5069, + "step": 17072 + }, + { + "epoch": 1.184720005551315, + "grad_norm": 1.9760113954544067, + "learning_rate": 0.00016508717623181266, + "loss": 0.7297, + "step": 17073 + }, + { + "epoch": 1.1847893969884116, + "grad_norm": 1.7337150573730469, + "learning_rate": 0.00016508291054537794, + "loss": 0.6645, + "step": 17074 + }, + { + "epoch": 1.1848587884255082, + "grad_norm": 2.129453659057617, + "learning_rate": 0.00016507864465348403, + "loss": 0.6121, + "step": 17075 + }, + { + "epoch": 1.184928179862605, + "grad_norm": 1.4979026317596436, + "learning_rate": 0.0001650743785561444, + "loss": 0.5589, + "step": 17076 + }, + { + "epoch": 1.1849975712997016, + "grad_norm": 1.3597880601882935, + "learning_rate": 0.00016507011225337252, + "loss": 0.6455, + "step": 17077 + }, + { + "epoch": 1.1850669627367982, + "grad_norm": 2.0165390968322754, + "learning_rate": 0.00016506584574518183, + "loss": 0.9567, + "step": 17078 + }, + { + "epoch": 1.1851363541738948, + "grad_norm": 2.1157398223876953, + "learning_rate": 0.0001650615790315858, + "loss": 0.7296, + "step": 17079 + }, + { + "epoch": 1.1852057456109917, + "grad_norm": 2.058243751525879, + "learning_rate": 0.00016505731211259792, + "loss": 0.891, + "step": 17080 + }, + { + "epoch": 1.1852751370480883, + "grad_norm": 1.9114797115325928, + "learning_rate": 0.00016505304498823167, + "loss": 0.9289, + "step": 17081 + }, + { + "epoch": 1.185344528485185, + "grad_norm": 1.7707051038742065, + "learning_rate": 0.0001650487776585005, + "loss": 0.6634, + "step": 17082 + }, + { + "epoch": 1.1854139199222815, + "grad_norm": 1.7571300268173218, + "learning_rate": 0.00016504451012341786, + "loss": 0.5535, + "step": 17083 + }, + { + "epoch": 1.1854833113593783, + "grad_norm": 1.4696784019470215, + "learning_rate": 0.00016504024238299728, + "loss": 0.5274, + "step": 17084 + }, + { + "epoch": 1.185552702796475, + "grad_norm": 2.179138660430908, + "learning_rate": 0.0001650359744372522, + "loss": 0.6668, + "step": 17085 + }, + { + "epoch": 1.1856220942335716, + "grad_norm": 2.8987395763397217, + "learning_rate": 0.00016503170628619606, + "loss": 0.8631, + "step": 17086 + }, + { + "epoch": 1.1856914856706682, + "grad_norm": 1.8137532472610474, + "learning_rate": 0.0001650274379298424, + "loss": 0.602, + "step": 17087 + }, + { + "epoch": 1.1857608771077648, + "grad_norm": 1.5219638347625732, + "learning_rate": 0.00016502316936820464, + "loss": 0.6209, + "step": 17088 + }, + { + "epoch": 1.1858302685448616, + "grad_norm": 1.6030681133270264, + "learning_rate": 0.0001650189006012963, + "loss": 0.7973, + "step": 17089 + }, + { + "epoch": 1.1858996599819582, + "grad_norm": 1.676472783088684, + "learning_rate": 0.0001650146316291308, + "loss": 0.602, + "step": 17090 + }, + { + "epoch": 1.1859690514190548, + "grad_norm": 1.3131675720214844, + "learning_rate": 0.0001650103624517217, + "loss": 0.6041, + "step": 17091 + }, + { + "epoch": 1.1860384428561517, + "grad_norm": 1.8565343618392944, + "learning_rate": 0.0001650060930690824, + "loss": 0.6955, + "step": 17092 + }, + { + "epoch": 1.1861078342932483, + "grad_norm": 2.090202569961548, + "learning_rate": 0.00016500182348122642, + "loss": 1.0534, + "step": 17093 + }, + { + "epoch": 1.186177225730345, + "grad_norm": 1.702944040298462, + "learning_rate": 0.0001649975536881672, + "loss": 0.5987, + "step": 17094 + }, + { + "epoch": 1.1862466171674415, + "grad_norm": 1.2778103351593018, + "learning_rate": 0.00016499328368991826, + "loss": 0.4421, + "step": 17095 + }, + { + "epoch": 1.1863160086045381, + "grad_norm": 1.9421308040618896, + "learning_rate": 0.0001649890134864931, + "loss": 0.5658, + "step": 17096 + }, + { + "epoch": 1.186385400041635, + "grad_norm": 1.7687352895736694, + "learning_rate": 0.0001649847430779051, + "loss": 0.7179, + "step": 17097 + }, + { + "epoch": 1.1864547914787316, + "grad_norm": 2.0477240085601807, + "learning_rate": 0.00016498047246416784, + "loss": 0.8362, + "step": 17098 + }, + { + "epoch": 1.1865241829158282, + "grad_norm": 2.709754467010498, + "learning_rate": 0.00016497620164529475, + "loss": 0.6999, + "step": 17099 + }, + { + "epoch": 1.1865935743529248, + "grad_norm": 7.347981929779053, + "learning_rate": 0.00016497193062129937, + "loss": 0.6513, + "step": 17100 + }, + { + "epoch": 1.1866629657900214, + "grad_norm": 1.6882094144821167, + "learning_rate": 0.0001649676593921951, + "loss": 0.7116, + "step": 17101 + }, + { + "epoch": 1.1867323572271182, + "grad_norm": 1.8848652839660645, + "learning_rate": 0.0001649633879579955, + "loss": 0.7412, + "step": 17102 + }, + { + "epoch": 1.1868017486642148, + "grad_norm": 2.039736032485962, + "learning_rate": 0.000164959116318714, + "loss": 0.6561, + "step": 17103 + }, + { + "epoch": 1.1868711401013115, + "grad_norm": 1.5525394678115845, + "learning_rate": 0.00016495484447436411, + "loss": 0.7003, + "step": 17104 + }, + { + "epoch": 1.1869405315384083, + "grad_norm": 2.012735605239868, + "learning_rate": 0.0001649505724249593, + "loss": 0.6632, + "step": 17105 + }, + { + "epoch": 1.187009922975505, + "grad_norm": 2.9231479167938232, + "learning_rate": 0.00016494630017051305, + "loss": 1.0332, + "step": 17106 + }, + { + "epoch": 1.1870793144126015, + "grad_norm": 2.723878860473633, + "learning_rate": 0.00016494202771103896, + "loss": 0.7289, + "step": 17107 + }, + { + "epoch": 1.1871487058496981, + "grad_norm": 1.9249407052993774, + "learning_rate": 0.00016493775504655033, + "loss": 0.6218, + "step": 17108 + }, + { + "epoch": 1.1872180972867947, + "grad_norm": 1.918297529220581, + "learning_rate": 0.00016493348217706076, + "loss": 0.6111, + "step": 17109 + }, + { + "epoch": 1.1872874887238916, + "grad_norm": 1.902786374092102, + "learning_rate": 0.00016492920910258376, + "loss": 0.5212, + "step": 17110 + }, + { + "epoch": 1.1873568801609882, + "grad_norm": 1.3290176391601562, + "learning_rate": 0.00016492493582313277, + "loss": 0.5625, + "step": 17111 + }, + { + "epoch": 1.1874262715980848, + "grad_norm": 1.624556064605713, + "learning_rate": 0.00016492066233872126, + "loss": 0.7099, + "step": 17112 + }, + { + "epoch": 1.1874956630351814, + "grad_norm": 2.0685155391693115, + "learning_rate": 0.00016491638864936275, + "loss": 0.7786, + "step": 17113 + }, + { + "epoch": 1.187565054472278, + "grad_norm": 1.8687262535095215, + "learning_rate": 0.00016491211475507076, + "loss": 0.7992, + "step": 17114 + }, + { + "epoch": 1.1876344459093748, + "grad_norm": 2.1638970375061035, + "learning_rate": 0.00016490784065585873, + "loss": 0.644, + "step": 17115 + }, + { + "epoch": 1.1877038373464714, + "grad_norm": 1.7634241580963135, + "learning_rate": 0.00016490356635174018, + "loss": 0.4502, + "step": 17116 + }, + { + "epoch": 1.187773228783568, + "grad_norm": 1.3647663593292236, + "learning_rate": 0.00016489929184272862, + "loss": 0.6336, + "step": 17117 + }, + { + "epoch": 1.187842620220665, + "grad_norm": 1.6846286058425903, + "learning_rate": 0.00016489501712883752, + "loss": 0.7585, + "step": 17118 + }, + { + "epoch": 1.1879120116577615, + "grad_norm": 1.9964922666549683, + "learning_rate": 0.00016489074221008037, + "loss": 0.8693, + "step": 17119 + }, + { + "epoch": 1.1879814030948581, + "grad_norm": 2.350916862487793, + "learning_rate": 0.00016488646708647068, + "loss": 0.7184, + "step": 17120 + }, + { + "epoch": 1.1880507945319547, + "grad_norm": 1.933142066001892, + "learning_rate": 0.00016488219175802196, + "loss": 0.5662, + "step": 17121 + }, + { + "epoch": 1.1881201859690513, + "grad_norm": 2.6394412517547607, + "learning_rate": 0.00016487791622474767, + "loss": 0.6028, + "step": 17122 + }, + { + "epoch": 1.1881895774061482, + "grad_norm": 2.2439281940460205, + "learning_rate": 0.00016487364048666133, + "loss": 0.493, + "step": 17123 + }, + { + "epoch": 1.1882589688432448, + "grad_norm": 2.0267536640167236, + "learning_rate": 0.00016486936454377643, + "loss": 0.8023, + "step": 17124 + }, + { + "epoch": 1.1883283602803414, + "grad_norm": 2.2271599769592285, + "learning_rate": 0.00016486508839610645, + "loss": 0.7455, + "step": 17125 + }, + { + "epoch": 1.188397751717438, + "grad_norm": 1.892532229423523, + "learning_rate": 0.00016486081204366494, + "loss": 0.7244, + "step": 17126 + }, + { + "epoch": 1.1884671431545346, + "grad_norm": 1.8714879751205444, + "learning_rate": 0.00016485653548646537, + "loss": 0.7087, + "step": 17127 + }, + { + "epoch": 1.1885365345916314, + "grad_norm": 2.562861680984497, + "learning_rate": 0.00016485225872452124, + "loss": 0.7454, + "step": 17128 + }, + { + "epoch": 1.188605926028728, + "grad_norm": 2.0878312587738037, + "learning_rate": 0.00016484798175784606, + "loss": 0.6635, + "step": 17129 + }, + { + "epoch": 1.1886753174658247, + "grad_norm": 1.7336277961730957, + "learning_rate": 0.0001648437045864533, + "loss": 0.5474, + "step": 17130 + }, + { + "epoch": 1.1887447089029215, + "grad_norm": 2.0168702602386475, + "learning_rate": 0.0001648394272103565, + "loss": 0.7202, + "step": 17131 + }, + { + "epoch": 1.188814100340018, + "grad_norm": 2.0703837871551514, + "learning_rate": 0.00016483514962956915, + "loss": 0.6959, + "step": 17132 + }, + { + "epoch": 1.1888834917771147, + "grad_norm": 1.8159908056259155, + "learning_rate": 0.00016483087184410478, + "loss": 0.6751, + "step": 17133 + }, + { + "epoch": 1.1889528832142113, + "grad_norm": 2.349400043487549, + "learning_rate": 0.00016482659385397684, + "loss": 0.8483, + "step": 17134 + }, + { + "epoch": 1.189022274651308, + "grad_norm": 1.8818631172180176, + "learning_rate": 0.0001648223156591989, + "loss": 0.8531, + "step": 17135 + }, + { + "epoch": 1.1890916660884048, + "grad_norm": 1.5305954217910767, + "learning_rate": 0.00016481803725978438, + "loss": 0.672, + "step": 17136 + }, + { + "epoch": 1.1891610575255014, + "grad_norm": 2.192197561264038, + "learning_rate": 0.00016481375865574688, + "loss": 0.8879, + "step": 17137 + }, + { + "epoch": 1.189230448962598, + "grad_norm": 1.6108500957489014, + "learning_rate": 0.00016480947984709985, + "loss": 0.5613, + "step": 17138 + }, + { + "epoch": 1.1892998403996946, + "grad_norm": 2.005842685699463, + "learning_rate": 0.0001648052008338568, + "loss": 0.7656, + "step": 17139 + }, + { + "epoch": 1.1893692318367912, + "grad_norm": 1.7389086484909058, + "learning_rate": 0.00016480092161603123, + "loss": 0.5834, + "step": 17140 + }, + { + "epoch": 1.189438623273888, + "grad_norm": 2.221973419189453, + "learning_rate": 0.00016479664219363674, + "loss": 0.7605, + "step": 17141 + }, + { + "epoch": 1.1895080147109847, + "grad_norm": 1.2987768650054932, + "learning_rate": 0.00016479236256668672, + "loss": 0.4461, + "step": 17142 + }, + { + "epoch": 1.1895774061480813, + "grad_norm": 1.7198063135147095, + "learning_rate": 0.00016478808273519472, + "loss": 0.6768, + "step": 17143 + }, + { + "epoch": 1.189646797585178, + "grad_norm": 1.8989261388778687, + "learning_rate": 0.00016478380269917427, + "loss": 0.8214, + "step": 17144 + }, + { + "epoch": 1.1897161890222747, + "grad_norm": 1.937057375907898, + "learning_rate": 0.00016477952245863887, + "loss": 0.8049, + "step": 17145 + }, + { + "epoch": 1.1897855804593713, + "grad_norm": 2.041447162628174, + "learning_rate": 0.00016477524201360204, + "loss": 0.6474, + "step": 17146 + }, + { + "epoch": 1.189854971896468, + "grad_norm": 2.1343066692352295, + "learning_rate": 0.00016477096136407728, + "loss": 0.7452, + "step": 17147 + }, + { + "epoch": 1.1899243633335645, + "grad_norm": 3.137604236602783, + "learning_rate": 0.00016476668051007814, + "loss": 0.6821, + "step": 17148 + }, + { + "epoch": 1.1899937547706614, + "grad_norm": 1.3886758089065552, + "learning_rate": 0.00016476239945161807, + "loss": 0.5764, + "step": 17149 + }, + { + "epoch": 1.190063146207758, + "grad_norm": 2.2562806606292725, + "learning_rate": 0.00016475811818871062, + "loss": 0.9789, + "step": 17150 + }, + { + "epoch": 1.1901325376448546, + "grad_norm": 1.72641921043396, + "learning_rate": 0.00016475383672136931, + "loss": 0.5637, + "step": 17151 + }, + { + "epoch": 1.1902019290819512, + "grad_norm": 1.9162975549697876, + "learning_rate": 0.00016474955504960765, + "loss": 0.5765, + "step": 17152 + }, + { + "epoch": 1.1902713205190478, + "grad_norm": 1.9479858875274658, + "learning_rate": 0.00016474527317343916, + "loss": 0.6285, + "step": 17153 + }, + { + "epoch": 1.1903407119561447, + "grad_norm": 2.017329454421997, + "learning_rate": 0.00016474099109287736, + "loss": 0.6971, + "step": 17154 + }, + { + "epoch": 1.1904101033932413, + "grad_norm": 2.389150381088257, + "learning_rate": 0.00016473670880793573, + "loss": 0.7011, + "step": 17155 + }, + { + "epoch": 1.1904794948303379, + "grad_norm": 2.1943252086639404, + "learning_rate": 0.00016473242631862784, + "loss": 0.7917, + "step": 17156 + }, + { + "epoch": 1.1905488862674347, + "grad_norm": 2.0636532306671143, + "learning_rate": 0.0001647281436249672, + "loss": 0.6525, + "step": 17157 + }, + { + "epoch": 1.1906182777045313, + "grad_norm": 1.1635665893554688, + "learning_rate": 0.00016472386072696735, + "loss": 0.3726, + "step": 17158 + }, + { + "epoch": 1.190687669141628, + "grad_norm": 1.9889720678329468, + "learning_rate": 0.00016471957762464174, + "loss": 0.7784, + "step": 17159 + }, + { + "epoch": 1.1907570605787245, + "grad_norm": 2.0425734519958496, + "learning_rate": 0.00016471529431800393, + "loss": 0.7642, + "step": 17160 + }, + { + "epoch": 1.1908264520158212, + "grad_norm": 1.822585940361023, + "learning_rate": 0.00016471101080706743, + "loss": 0.5767, + "step": 17161 + }, + { + "epoch": 1.190895843452918, + "grad_norm": 2.041245222091675, + "learning_rate": 0.00016470672709184583, + "loss": 0.715, + "step": 17162 + }, + { + "epoch": 1.1909652348900146, + "grad_norm": 2.2808313369750977, + "learning_rate": 0.00016470244317235254, + "loss": 0.6845, + "step": 17163 + }, + { + "epoch": 1.1910346263271112, + "grad_norm": 3.247220277786255, + "learning_rate": 0.0001646981590486012, + "loss": 0.8608, + "step": 17164 + }, + { + "epoch": 1.1911040177642078, + "grad_norm": 1.6226557493209839, + "learning_rate": 0.00016469387472060524, + "loss": 0.8268, + "step": 17165 + }, + { + "epoch": 1.1911734092013047, + "grad_norm": 1.7362219095230103, + "learning_rate": 0.00016468959018837827, + "loss": 0.6288, + "step": 17166 + }, + { + "epoch": 1.1912428006384013, + "grad_norm": 1.5065956115722656, + "learning_rate": 0.0001646853054519337, + "loss": 0.5792, + "step": 17167 + }, + { + "epoch": 1.1913121920754979, + "grad_norm": 1.2824910879135132, + "learning_rate": 0.00016468102051128515, + "loss": 0.5174, + "step": 17168 + }, + { + "epoch": 1.1913815835125945, + "grad_norm": 1.8333319425582886, + "learning_rate": 0.00016467673536644614, + "loss": 0.6716, + "step": 17169 + }, + { + "epoch": 1.1914509749496913, + "grad_norm": 1.61879563331604, + "learning_rate": 0.00016467245001743018, + "loss": 0.514, + "step": 17170 + }, + { + "epoch": 1.191520366386788, + "grad_norm": 1.8277379274368286, + "learning_rate": 0.0001646681644642508, + "loss": 0.6011, + "step": 17171 + }, + { + "epoch": 1.1915897578238845, + "grad_norm": 1.345044732093811, + "learning_rate": 0.00016466387870692152, + "loss": 0.7168, + "step": 17172 + }, + { + "epoch": 1.1916591492609812, + "grad_norm": 1.5373353958129883, + "learning_rate": 0.00016465959274545588, + "loss": 0.6921, + "step": 17173 + }, + { + "epoch": 1.1917285406980778, + "grad_norm": 1.8153818845748901, + "learning_rate": 0.0001646553065798674, + "loss": 0.7232, + "step": 17174 + }, + { + "epoch": 1.1917979321351746, + "grad_norm": 2.2563202381134033, + "learning_rate": 0.00016465102021016962, + "loss": 0.6979, + "step": 17175 + }, + { + "epoch": 1.1918673235722712, + "grad_norm": 2.2699503898620605, + "learning_rate": 0.00016464673363637608, + "loss": 0.799, + "step": 17176 + }, + { + "epoch": 1.1919367150093678, + "grad_norm": 2.445115089416504, + "learning_rate": 0.0001646424468585003, + "loss": 0.8693, + "step": 17177 + }, + { + "epoch": 1.1920061064464644, + "grad_norm": 1.994376301765442, + "learning_rate": 0.0001646381598765558, + "loss": 0.9224, + "step": 17178 + }, + { + "epoch": 1.1920754978835613, + "grad_norm": 1.9470041990280151, + "learning_rate": 0.00016463387269055615, + "loss": 0.9571, + "step": 17179 + }, + { + "epoch": 1.1921448893206579, + "grad_norm": 2.154257297515869, + "learning_rate": 0.00016462958530051485, + "loss": 0.6591, + "step": 17180 + }, + { + "epoch": 1.1922142807577545, + "grad_norm": 2.196195125579834, + "learning_rate": 0.00016462529770644544, + "loss": 0.8693, + "step": 17181 + }, + { + "epoch": 1.192283672194851, + "grad_norm": 1.4414407014846802, + "learning_rate": 0.00016462100990836148, + "loss": 0.638, + "step": 17182 + }, + { + "epoch": 1.192353063631948, + "grad_norm": 1.6077214479446411, + "learning_rate": 0.0001646167219062765, + "loss": 0.7114, + "step": 17183 + }, + { + "epoch": 1.1924224550690445, + "grad_norm": 2.226435899734497, + "learning_rate": 0.000164612433700204, + "loss": 0.6937, + "step": 17184 + }, + { + "epoch": 1.1924918465061412, + "grad_norm": 1.771842360496521, + "learning_rate": 0.00016460814529015754, + "loss": 0.6263, + "step": 17185 + }, + { + "epoch": 1.1925612379432378, + "grad_norm": 1.7707277536392212, + "learning_rate": 0.00016460385667615068, + "loss": 0.9111, + "step": 17186 + }, + { + "epoch": 1.1926306293803344, + "grad_norm": 1.5868449211120605, + "learning_rate": 0.00016459956785819693, + "loss": 0.5518, + "step": 17187 + }, + { + "epoch": 1.1927000208174312, + "grad_norm": 2.0035741329193115, + "learning_rate": 0.00016459527883630983, + "loss": 0.9735, + "step": 17188 + }, + { + "epoch": 1.1927694122545278, + "grad_norm": 2.247502088546753, + "learning_rate": 0.00016459098961050294, + "loss": 0.7226, + "step": 17189 + }, + { + "epoch": 1.1928388036916244, + "grad_norm": 1.7234951257705688, + "learning_rate": 0.0001645867001807898, + "loss": 0.5764, + "step": 17190 + }, + { + "epoch": 1.192908195128721, + "grad_norm": 1.9504214525222778, + "learning_rate": 0.00016458241054718395, + "loss": 0.8567, + "step": 17191 + }, + { + "epoch": 1.1929775865658179, + "grad_norm": 2.27580189704895, + "learning_rate": 0.00016457812070969888, + "loss": 0.7798, + "step": 17192 + }, + { + "epoch": 1.1930469780029145, + "grad_norm": 1.9570763111114502, + "learning_rate": 0.0001645738306683482, + "loss": 0.5735, + "step": 17193 + }, + { + "epoch": 1.193116369440011, + "grad_norm": 2.8115952014923096, + "learning_rate": 0.00016456954042314546, + "loss": 0.6955, + "step": 17194 + }, + { + "epoch": 1.1931857608771077, + "grad_norm": 1.305841088294983, + "learning_rate": 0.00016456524997410411, + "loss": 0.5019, + "step": 17195 + }, + { + "epoch": 1.1932551523142045, + "grad_norm": 1.542826771736145, + "learning_rate": 0.0001645609593212378, + "loss": 0.4189, + "step": 17196 + }, + { + "epoch": 1.1933245437513011, + "grad_norm": 1.4679332971572876, + "learning_rate": 0.00016455666846456002, + "loss": 0.6021, + "step": 17197 + }, + { + "epoch": 1.1933939351883978, + "grad_norm": 1.6657742261886597, + "learning_rate": 0.00016455237740408432, + "loss": 0.4724, + "step": 17198 + }, + { + "epoch": 1.1934633266254944, + "grad_norm": 2.36387038230896, + "learning_rate": 0.00016454808613982426, + "loss": 0.6777, + "step": 17199 + }, + { + "epoch": 1.193532718062591, + "grad_norm": 1.6218026876449585, + "learning_rate": 0.0001645437946717934, + "loss": 0.4782, + "step": 17200 + }, + { + "epoch": 1.1936021094996878, + "grad_norm": 1.3585960865020752, + "learning_rate": 0.00016453950300000526, + "loss": 0.6786, + "step": 17201 + }, + { + "epoch": 1.1936715009367844, + "grad_norm": 1.9855269193649292, + "learning_rate": 0.00016453521112447337, + "loss": 0.7457, + "step": 17202 + }, + { + "epoch": 1.193740892373881, + "grad_norm": 1.7693777084350586, + "learning_rate": 0.00016453091904521135, + "loss": 0.6069, + "step": 17203 + }, + { + "epoch": 1.1938102838109776, + "grad_norm": 2.71919846534729, + "learning_rate": 0.00016452662676223265, + "loss": 0.6678, + "step": 17204 + }, + { + "epoch": 1.1938796752480745, + "grad_norm": 1.6046947240829468, + "learning_rate": 0.0001645223342755509, + "loss": 0.5558, + "step": 17205 + }, + { + "epoch": 1.193949066685171, + "grad_norm": 2.2378644943237305, + "learning_rate": 0.00016451804158517966, + "loss": 0.8504, + "step": 17206 + }, + { + "epoch": 1.1940184581222677, + "grad_norm": 1.931854486465454, + "learning_rate": 0.0001645137486911324, + "loss": 0.8216, + "step": 17207 + }, + { + "epoch": 1.1940878495593643, + "grad_norm": 1.7492940425872803, + "learning_rate": 0.00016450945559342274, + "loss": 0.5763, + "step": 17208 + }, + { + "epoch": 1.1941572409964611, + "grad_norm": 1.8702293634414673, + "learning_rate": 0.0001645051622920642, + "loss": 0.689, + "step": 17209 + }, + { + "epoch": 1.1942266324335578, + "grad_norm": 2.4945056438446045, + "learning_rate": 0.00016450086878707038, + "loss": 0.7607, + "step": 17210 + }, + { + "epoch": 1.1942960238706544, + "grad_norm": 1.904314637184143, + "learning_rate": 0.00016449657507845478, + "loss": 0.6087, + "step": 17211 + }, + { + "epoch": 1.194365415307751, + "grad_norm": 2.2013378143310547, + "learning_rate": 0.00016449228116623097, + "loss": 0.6449, + "step": 17212 + }, + { + "epoch": 1.1944348067448476, + "grad_norm": 2.2740771770477295, + "learning_rate": 0.0001644879870504125, + "loss": 0.7968, + "step": 17213 + }, + { + "epoch": 1.1945041981819444, + "grad_norm": 2.251197099685669, + "learning_rate": 0.00016448369273101298, + "loss": 0.7216, + "step": 17214 + }, + { + "epoch": 1.194573589619041, + "grad_norm": 2.920933485031128, + "learning_rate": 0.00016447939820804586, + "loss": 0.8563, + "step": 17215 + }, + { + "epoch": 1.1946429810561376, + "grad_norm": 1.767590880393982, + "learning_rate": 0.00016447510348152484, + "loss": 0.7048, + "step": 17216 + }, + { + "epoch": 1.1947123724932343, + "grad_norm": 1.7696653604507446, + "learning_rate": 0.00016447080855146336, + "loss": 0.8316, + "step": 17217 + }, + { + "epoch": 1.194781763930331, + "grad_norm": 1.6640738248825073, + "learning_rate": 0.000164466513417875, + "loss": 0.5175, + "step": 17218 + }, + { + "epoch": 1.1948511553674277, + "grad_norm": 1.521328330039978, + "learning_rate": 0.00016446221808077337, + "loss": 0.6677, + "step": 17219 + }, + { + "epoch": 1.1949205468045243, + "grad_norm": 1.726689100265503, + "learning_rate": 0.00016445792254017193, + "loss": 0.607, + "step": 17220 + }, + { + "epoch": 1.194989938241621, + "grad_norm": 1.5095950365066528, + "learning_rate": 0.0001644536267960844, + "loss": 0.6193, + "step": 17221 + }, + { + "epoch": 1.1950593296787178, + "grad_norm": 2.402555227279663, + "learning_rate": 0.0001644493308485242, + "loss": 0.6901, + "step": 17222 + }, + { + "epoch": 1.1951287211158144, + "grad_norm": 1.9388538599014282, + "learning_rate": 0.00016444503469750492, + "loss": 0.5558, + "step": 17223 + }, + { + "epoch": 1.195198112552911, + "grad_norm": 1.7619999647140503, + "learning_rate": 0.00016444073834304018, + "loss": 0.8126, + "step": 17224 + }, + { + "epoch": 1.1952675039900076, + "grad_norm": 2.134925603866577, + "learning_rate": 0.00016443644178514347, + "loss": 0.6118, + "step": 17225 + }, + { + "epoch": 1.1953368954271042, + "grad_norm": 1.2358909845352173, + "learning_rate": 0.00016443214502382842, + "loss": 0.5139, + "step": 17226 + }, + { + "epoch": 1.195406286864201, + "grad_norm": 2.616546869277954, + "learning_rate": 0.0001644278480591086, + "loss": 0.6258, + "step": 17227 + }, + { + "epoch": 1.1954756783012976, + "grad_norm": 1.8578087091445923, + "learning_rate": 0.00016442355089099745, + "loss": 0.6387, + "step": 17228 + }, + { + "epoch": 1.1955450697383943, + "grad_norm": 1.955970048904419, + "learning_rate": 0.00016441925351950868, + "loss": 0.8693, + "step": 17229 + }, + { + "epoch": 1.1956144611754909, + "grad_norm": 1.6428816318511963, + "learning_rate": 0.00016441495594465578, + "loss": 0.6813, + "step": 17230 + }, + { + "epoch": 1.1956838526125877, + "grad_norm": 2.1703836917877197, + "learning_rate": 0.00016441065816645236, + "loss": 0.4868, + "step": 17231 + }, + { + "epoch": 1.1957532440496843, + "grad_norm": 1.7242573499679565, + "learning_rate": 0.00016440636018491192, + "loss": 0.6449, + "step": 17232 + }, + { + "epoch": 1.195822635486781, + "grad_norm": 1.926131010055542, + "learning_rate": 0.0001644020620000481, + "loss": 0.547, + "step": 17233 + }, + { + "epoch": 1.1958920269238775, + "grad_norm": 2.602508068084717, + "learning_rate": 0.0001643977636118745, + "loss": 0.6805, + "step": 17234 + }, + { + "epoch": 1.1959614183609744, + "grad_norm": 1.8204294443130493, + "learning_rate": 0.00016439346502040457, + "loss": 0.6916, + "step": 17235 + }, + { + "epoch": 1.196030809798071, + "grad_norm": 1.8029365539550781, + "learning_rate": 0.00016438916622565192, + "loss": 0.8116, + "step": 17236 + }, + { + "epoch": 1.1961002012351676, + "grad_norm": 2.041433572769165, + "learning_rate": 0.00016438486722763018, + "loss": 0.6755, + "step": 17237 + }, + { + "epoch": 1.1961695926722642, + "grad_norm": 1.962280035018921, + "learning_rate": 0.0001643805680263529, + "loss": 0.7279, + "step": 17238 + }, + { + "epoch": 1.1962389841093608, + "grad_norm": 2.001725912094116, + "learning_rate": 0.00016437626862183358, + "loss": 0.7025, + "step": 17239 + }, + { + "epoch": 1.1963083755464576, + "grad_norm": 1.7633832693099976, + "learning_rate": 0.0001643719690140859, + "loss": 0.6134, + "step": 17240 + }, + { + "epoch": 1.1963777669835542, + "grad_norm": 1.9115923643112183, + "learning_rate": 0.00016436766920312337, + "loss": 0.802, + "step": 17241 + }, + { + "epoch": 1.1964471584206509, + "grad_norm": 1.6053836345672607, + "learning_rate": 0.00016436336918895957, + "loss": 0.7476, + "step": 17242 + }, + { + "epoch": 1.1965165498577475, + "grad_norm": 1.8657952547073364, + "learning_rate": 0.0001643590689716081, + "loss": 0.8792, + "step": 17243 + }, + { + "epoch": 1.1965859412948443, + "grad_norm": 1.8773562908172607, + "learning_rate": 0.00016435476855108244, + "loss": 0.6712, + "step": 17244 + }, + { + "epoch": 1.196655332731941, + "grad_norm": 2.2689201831817627, + "learning_rate": 0.00016435046792739636, + "loss": 0.9599, + "step": 17245 + }, + { + "epoch": 1.1967247241690375, + "grad_norm": 1.853905200958252, + "learning_rate": 0.00016434616710056323, + "loss": 0.8289, + "step": 17246 + }, + { + "epoch": 1.1967941156061341, + "grad_norm": 1.7744725942611694, + "learning_rate": 0.00016434186607059673, + "loss": 0.9103, + "step": 17247 + }, + { + "epoch": 1.196863507043231, + "grad_norm": 1.7657191753387451, + "learning_rate": 0.00016433756483751047, + "loss": 0.6112, + "step": 17248 + }, + { + "epoch": 1.1969328984803276, + "grad_norm": 1.5887987613677979, + "learning_rate": 0.00016433326340131794, + "loss": 0.5588, + "step": 17249 + }, + { + "epoch": 1.1970022899174242, + "grad_norm": 1.9499397277832031, + "learning_rate": 0.00016432896176203277, + "loss": 0.7903, + "step": 17250 + }, + { + "epoch": 1.1970716813545208, + "grad_norm": 1.6704117059707642, + "learning_rate": 0.00016432465991966853, + "loss": 0.7197, + "step": 17251 + }, + { + "epoch": 1.1971410727916174, + "grad_norm": 1.7098320722579956, + "learning_rate": 0.00016432035787423881, + "loss": 0.4831, + "step": 17252 + }, + { + "epoch": 1.1972104642287142, + "grad_norm": 1.7110626697540283, + "learning_rate": 0.0001643160556257572, + "loss": 0.4901, + "step": 17253 + }, + { + "epoch": 1.1972798556658109, + "grad_norm": 1.964889407157898, + "learning_rate": 0.00016431175317423722, + "loss": 0.8208, + "step": 17254 + }, + { + "epoch": 1.1973492471029075, + "grad_norm": 1.6439735889434814, + "learning_rate": 0.00016430745051969253, + "loss": 0.4905, + "step": 17255 + }, + { + "epoch": 1.197418638540004, + "grad_norm": 1.7466795444488525, + "learning_rate": 0.00016430314766213667, + "loss": 0.5164, + "step": 17256 + }, + { + "epoch": 1.197488029977101, + "grad_norm": 1.870172381401062, + "learning_rate": 0.00016429884460158323, + "loss": 0.4839, + "step": 17257 + }, + { + "epoch": 1.1975574214141975, + "grad_norm": 2.037562370300293, + "learning_rate": 0.00016429454133804578, + "loss": 0.5985, + "step": 17258 + }, + { + "epoch": 1.1976268128512941, + "grad_norm": 1.3546066284179688, + "learning_rate": 0.00016429023787153794, + "loss": 0.4926, + "step": 17259 + }, + { + "epoch": 1.1976962042883907, + "grad_norm": 1.9323385953903198, + "learning_rate": 0.0001642859342020733, + "loss": 0.6954, + "step": 17260 + }, + { + "epoch": 1.1977655957254876, + "grad_norm": 1.3639122247695923, + "learning_rate": 0.0001642816303296654, + "loss": 0.5509, + "step": 17261 + }, + { + "epoch": 1.1978349871625842, + "grad_norm": 1.8045785427093506, + "learning_rate": 0.0001642773262543279, + "loss": 0.5055, + "step": 17262 + }, + { + "epoch": 1.1979043785996808, + "grad_norm": 1.9698387384414673, + "learning_rate": 0.00016427302197607427, + "loss": 0.6809, + "step": 17263 + }, + { + "epoch": 1.1979737700367774, + "grad_norm": 1.8090908527374268, + "learning_rate": 0.00016426871749491818, + "loss": 0.5969, + "step": 17264 + }, + { + "epoch": 1.198043161473874, + "grad_norm": 2.262876033782959, + "learning_rate": 0.00016426441281087322, + "loss": 1.0509, + "step": 17265 + }, + { + "epoch": 1.1981125529109709, + "grad_norm": 2.1484711170196533, + "learning_rate": 0.00016426010792395295, + "loss": 0.6568, + "step": 17266 + }, + { + "epoch": 1.1981819443480675, + "grad_norm": 2.0462427139282227, + "learning_rate": 0.000164255802834171, + "loss": 0.5906, + "step": 17267 + }, + { + "epoch": 1.198251335785164, + "grad_norm": 2.57863712310791, + "learning_rate": 0.00016425149754154092, + "loss": 0.822, + "step": 17268 + }, + { + "epoch": 1.1983207272222607, + "grad_norm": 2.024739980697632, + "learning_rate": 0.00016424719204607632, + "loss": 0.6548, + "step": 17269 + }, + { + "epoch": 1.1983901186593575, + "grad_norm": 1.5862900018692017, + "learning_rate": 0.00016424288634779078, + "loss": 0.4088, + "step": 17270 + }, + { + "epoch": 1.1984595100964541, + "grad_norm": 1.763519287109375, + "learning_rate": 0.0001642385804466979, + "loss": 0.6863, + "step": 17271 + }, + { + "epoch": 1.1985289015335507, + "grad_norm": 1.9580456018447876, + "learning_rate": 0.0001642342743428113, + "loss": 0.781, + "step": 17272 + }, + { + "epoch": 1.1985982929706473, + "grad_norm": 1.9835482835769653, + "learning_rate": 0.00016422996803614455, + "loss": 0.7771, + "step": 17273 + }, + { + "epoch": 1.1986676844077442, + "grad_norm": 1.6372768878936768, + "learning_rate": 0.00016422566152671122, + "loss": 0.7029, + "step": 17274 + }, + { + "epoch": 1.1987370758448408, + "grad_norm": 2.132517099380493, + "learning_rate": 0.00016422135481452491, + "loss": 0.6534, + "step": 17275 + }, + { + "epoch": 1.1988064672819374, + "grad_norm": 1.565116047859192, + "learning_rate": 0.00016421704789959928, + "loss": 0.5883, + "step": 17276 + }, + { + "epoch": 1.198875858719034, + "grad_norm": 1.6718679666519165, + "learning_rate": 0.00016421274078194783, + "loss": 0.8387, + "step": 17277 + }, + { + "epoch": 1.1989452501561306, + "grad_norm": 1.349711537361145, + "learning_rate": 0.00016420843346158425, + "loss": 0.3304, + "step": 17278 + }, + { + "epoch": 1.1990146415932275, + "grad_norm": 1.7069060802459717, + "learning_rate": 0.00016420412593852208, + "loss": 0.5706, + "step": 17279 + }, + { + "epoch": 1.199084033030324, + "grad_norm": 3.0924618244171143, + "learning_rate": 0.00016419981821277495, + "loss": 0.7002, + "step": 17280 + }, + { + "epoch": 1.1991534244674207, + "grad_norm": 2.578427791595459, + "learning_rate": 0.00016419551028435642, + "loss": 0.7957, + "step": 17281 + }, + { + "epoch": 1.1992228159045173, + "grad_norm": 1.8557485342025757, + "learning_rate": 0.0001641912021532801, + "loss": 0.6025, + "step": 17282 + }, + { + "epoch": 1.1992922073416141, + "grad_norm": 2.3490047454833984, + "learning_rate": 0.00016418689381955966, + "loss": 0.8011, + "step": 17283 + }, + { + "epoch": 1.1993615987787107, + "grad_norm": 2.680424690246582, + "learning_rate": 0.00016418258528320858, + "loss": 0.8141, + "step": 17284 + }, + { + "epoch": 1.1994309902158073, + "grad_norm": 1.8903943300247192, + "learning_rate": 0.00016417827654424054, + "loss": 0.6987, + "step": 17285 + }, + { + "epoch": 1.199500381652904, + "grad_norm": 1.6141782999038696, + "learning_rate": 0.00016417396760266915, + "loss": 0.5332, + "step": 17286 + }, + { + "epoch": 1.1995697730900008, + "grad_norm": 1.5580607652664185, + "learning_rate": 0.00016416965845850797, + "loss": 0.5524, + "step": 17287 + }, + { + "epoch": 1.1996391645270974, + "grad_norm": 1.8774470090866089, + "learning_rate": 0.00016416534911177063, + "loss": 0.5968, + "step": 17288 + }, + { + "epoch": 1.199708555964194, + "grad_norm": 1.6975643634796143, + "learning_rate": 0.00016416103956247072, + "loss": 0.5555, + "step": 17289 + }, + { + "epoch": 1.1997779474012906, + "grad_norm": 1.8790522813796997, + "learning_rate": 0.00016415672981062187, + "loss": 0.5813, + "step": 17290 + }, + { + "epoch": 1.1998473388383872, + "grad_norm": 2.713496685028076, + "learning_rate": 0.00016415241985623765, + "loss": 0.9347, + "step": 17291 + }, + { + "epoch": 1.199916730275484, + "grad_norm": 1.9256356954574585, + "learning_rate": 0.00016414810969933168, + "loss": 0.5511, + "step": 17292 + }, + { + "epoch": 1.1999861217125807, + "grad_norm": 1.5688117742538452, + "learning_rate": 0.00016414379933991758, + "loss": 0.8069, + "step": 17293 + }, + { + "epoch": 1.2000555131496773, + "grad_norm": 1.8731229305267334, + "learning_rate": 0.00016413948877800895, + "loss": 0.7815, + "step": 17294 + }, + { + "epoch": 1.200124904586774, + "grad_norm": 1.3980711698532104, + "learning_rate": 0.0001641351780136194, + "loss": 0.5095, + "step": 17295 + }, + { + "epoch": 1.2001942960238707, + "grad_norm": 2.105008602142334, + "learning_rate": 0.0001641308670467625, + "loss": 0.7252, + "step": 17296 + }, + { + "epoch": 1.2002636874609673, + "grad_norm": 3.092893600463867, + "learning_rate": 0.00016412655587745193, + "loss": 0.6387, + "step": 17297 + }, + { + "epoch": 1.200333078898064, + "grad_norm": 2.14448618888855, + "learning_rate": 0.00016412224450570127, + "loss": 0.6748, + "step": 17298 + }, + { + "epoch": 1.2004024703351606, + "grad_norm": 1.4753957986831665, + "learning_rate": 0.0001641179329315241, + "loss": 0.599, + "step": 17299 + }, + { + "epoch": 1.2004718617722574, + "grad_norm": 1.5749565362930298, + "learning_rate": 0.00016411362115493407, + "loss": 0.6416, + "step": 17300 + }, + { + "epoch": 1.200541253209354, + "grad_norm": 1.8223721981048584, + "learning_rate": 0.00016410930917594476, + "loss": 0.5219, + "step": 17301 + }, + { + "epoch": 1.2006106446464506, + "grad_norm": 2.395369529724121, + "learning_rate": 0.00016410499699456981, + "loss": 0.6987, + "step": 17302 + }, + { + "epoch": 1.2006800360835472, + "grad_norm": 1.5624938011169434, + "learning_rate": 0.00016410068461082284, + "loss": 0.5506, + "step": 17303 + }, + { + "epoch": 1.2007494275206438, + "grad_norm": 2.198352098464966, + "learning_rate": 0.00016409637202471742, + "loss": 0.6079, + "step": 17304 + }, + { + "epoch": 1.2008188189577407, + "grad_norm": 1.614347219467163, + "learning_rate": 0.0001640920592362672, + "loss": 0.4478, + "step": 17305 + }, + { + "epoch": 1.2008882103948373, + "grad_norm": 1.4643607139587402, + "learning_rate": 0.00016408774624548578, + "loss": 0.5813, + "step": 17306 + }, + { + "epoch": 1.200957601831934, + "grad_norm": 1.6551287174224854, + "learning_rate": 0.0001640834330523868, + "loss": 0.7364, + "step": 17307 + }, + { + "epoch": 1.2010269932690305, + "grad_norm": 2.2052597999572754, + "learning_rate": 0.00016407911965698378, + "loss": 0.627, + "step": 17308 + }, + { + "epoch": 1.2010963847061273, + "grad_norm": 1.894569993019104, + "learning_rate": 0.0001640748060592905, + "loss": 0.5627, + "step": 17309 + }, + { + "epoch": 1.201165776143224, + "grad_norm": 1.691179871559143, + "learning_rate": 0.00016407049225932046, + "loss": 0.5997, + "step": 17310 + }, + { + "epoch": 1.2012351675803206, + "grad_norm": 2.4183781147003174, + "learning_rate": 0.00016406617825708732, + "loss": 0.5986, + "step": 17311 + }, + { + "epoch": 1.2013045590174172, + "grad_norm": 2.8110594749450684, + "learning_rate": 0.00016406186405260465, + "loss": 0.7489, + "step": 17312 + }, + { + "epoch": 1.201373950454514, + "grad_norm": 2.7781074047088623, + "learning_rate": 0.00016405754964588616, + "loss": 0.7684, + "step": 17313 + }, + { + "epoch": 1.2014433418916106, + "grad_norm": 1.9185460805892944, + "learning_rate": 0.0001640532350369454, + "loss": 0.7735, + "step": 17314 + }, + { + "epoch": 1.2015127333287072, + "grad_norm": 1.7748830318450928, + "learning_rate": 0.00016404892022579598, + "loss": 0.6655, + "step": 17315 + }, + { + "epoch": 1.2015821247658038, + "grad_norm": 2.8727924823760986, + "learning_rate": 0.00016404460521245157, + "loss": 0.6855, + "step": 17316 + }, + { + "epoch": 1.2016515162029004, + "grad_norm": 1.6578178405761719, + "learning_rate": 0.00016404028999692581, + "loss": 0.8732, + "step": 17317 + }, + { + "epoch": 1.2017209076399973, + "grad_norm": 1.9008021354675293, + "learning_rate": 0.00016403597457923224, + "loss": 0.654, + "step": 17318 + }, + { + "epoch": 1.201790299077094, + "grad_norm": 2.0182223320007324, + "learning_rate": 0.00016403165895938454, + "loss": 0.9157, + "step": 17319 + }, + { + "epoch": 1.2018596905141905, + "grad_norm": 1.8925895690917969, + "learning_rate": 0.00016402734313739628, + "loss": 0.6235, + "step": 17320 + }, + { + "epoch": 1.2019290819512871, + "grad_norm": 2.2700231075286865, + "learning_rate": 0.00016402302711328122, + "loss": 0.783, + "step": 17321 + }, + { + "epoch": 1.201998473388384, + "grad_norm": 1.7709829807281494, + "learning_rate": 0.00016401871088705282, + "loss": 0.6313, + "step": 17322 + }, + { + "epoch": 1.2020678648254806, + "grad_norm": 1.6518168449401855, + "learning_rate": 0.0001640143944587248, + "loss": 0.6868, + "step": 17323 + }, + { + "epoch": 1.2021372562625772, + "grad_norm": 1.4609016180038452, + "learning_rate": 0.00016401007782831077, + "loss": 0.6121, + "step": 17324 + }, + { + "epoch": 1.2022066476996738, + "grad_norm": 1.3796961307525635, + "learning_rate": 0.00016400576099582433, + "loss": 0.6699, + "step": 17325 + }, + { + "epoch": 1.2022760391367706, + "grad_norm": 1.576703667640686, + "learning_rate": 0.00016400144396127913, + "loss": 0.5108, + "step": 17326 + }, + { + "epoch": 1.2023454305738672, + "grad_norm": 2.4631195068359375, + "learning_rate": 0.0001639971267246888, + "loss": 0.5716, + "step": 17327 + }, + { + "epoch": 1.2024148220109638, + "grad_norm": 1.2851358652114868, + "learning_rate": 0.000163992809286067, + "loss": 0.4998, + "step": 17328 + }, + { + "epoch": 1.2024842134480604, + "grad_norm": 1.5971477031707764, + "learning_rate": 0.00016398849164542728, + "loss": 0.7013, + "step": 17329 + }, + { + "epoch": 1.202553604885157, + "grad_norm": 2.4648876190185547, + "learning_rate": 0.00016398417380278333, + "loss": 0.7141, + "step": 17330 + }, + { + "epoch": 1.202622996322254, + "grad_norm": 2.1809639930725098, + "learning_rate": 0.00016397985575814877, + "loss": 0.8013, + "step": 17331 + }, + { + "epoch": 1.2026923877593505, + "grad_norm": 1.609556794166565, + "learning_rate": 0.00016397553751153725, + "loss": 0.5066, + "step": 17332 + }, + { + "epoch": 1.2027617791964471, + "grad_norm": 1.7289621829986572, + "learning_rate": 0.00016397121906296236, + "loss": 0.729, + "step": 17333 + }, + { + "epoch": 1.202831170633544, + "grad_norm": 2.0682971477508545, + "learning_rate": 0.00016396690041243773, + "loss": 0.7788, + "step": 17334 + }, + { + "epoch": 1.2029005620706406, + "grad_norm": 1.678466558456421, + "learning_rate": 0.00016396258155997707, + "loss": 0.6873, + "step": 17335 + }, + { + "epoch": 1.2029699535077372, + "grad_norm": 1.5675173997879028, + "learning_rate": 0.00016395826250559389, + "loss": 0.4608, + "step": 17336 + }, + { + "epoch": 1.2030393449448338, + "grad_norm": 2.2226614952087402, + "learning_rate": 0.00016395394324930193, + "loss": 0.8415, + "step": 17337 + }, + { + "epoch": 1.2031087363819304, + "grad_norm": 1.3733737468719482, + "learning_rate": 0.00016394962379111483, + "loss": 0.6068, + "step": 17338 + }, + { + "epoch": 1.2031781278190272, + "grad_norm": 1.7269127368927002, + "learning_rate": 0.00016394530413104614, + "loss": 0.8643, + "step": 17339 + }, + { + "epoch": 1.2032475192561238, + "grad_norm": 1.7962145805358887, + "learning_rate": 0.00016394098426910958, + "loss": 0.7764, + "step": 17340 + }, + { + "epoch": 1.2033169106932204, + "grad_norm": 1.7230019569396973, + "learning_rate": 0.00016393666420531873, + "loss": 0.7015, + "step": 17341 + }, + { + "epoch": 1.203386302130317, + "grad_norm": 1.1803227663040161, + "learning_rate": 0.00016393234393968724, + "loss": 0.3647, + "step": 17342 + }, + { + "epoch": 1.2034556935674137, + "grad_norm": 2.8981785774230957, + "learning_rate": 0.00016392802347222878, + "loss": 1.0292, + "step": 17343 + }, + { + "epoch": 1.2035250850045105, + "grad_norm": 2.5177834033966064, + "learning_rate": 0.00016392370280295695, + "loss": 0.6426, + "step": 17344 + }, + { + "epoch": 1.203594476441607, + "grad_norm": 2.420128583908081, + "learning_rate": 0.0001639193819318854, + "loss": 0.7127, + "step": 17345 + }, + { + "epoch": 1.2036638678787037, + "grad_norm": 1.8483364582061768, + "learning_rate": 0.00016391506085902777, + "loss": 0.5751, + "step": 17346 + }, + { + "epoch": 1.2037332593158006, + "grad_norm": 1.4313855171203613, + "learning_rate": 0.00016391073958439773, + "loss": 0.6042, + "step": 17347 + }, + { + "epoch": 1.2038026507528972, + "grad_norm": 1.497251033782959, + "learning_rate": 0.00016390641810800892, + "loss": 0.7412, + "step": 17348 + }, + { + "epoch": 1.2038720421899938, + "grad_norm": 1.6147977113723755, + "learning_rate": 0.00016390209642987494, + "loss": 0.7693, + "step": 17349 + }, + { + "epoch": 1.2039414336270904, + "grad_norm": 1.6841245889663696, + "learning_rate": 0.00016389777455000942, + "loss": 0.6298, + "step": 17350 + }, + { + "epoch": 1.204010825064187, + "grad_norm": 1.7231024503707886, + "learning_rate": 0.0001638934524684261, + "loss": 0.6453, + "step": 17351 + }, + { + "epoch": 1.2040802165012838, + "grad_norm": 1.6814932823181152, + "learning_rate": 0.00016388913018513854, + "loss": 0.581, + "step": 17352 + }, + { + "epoch": 1.2041496079383804, + "grad_norm": 1.7530608177185059, + "learning_rate": 0.0001638848077001604, + "loss": 0.5655, + "step": 17353 + }, + { + "epoch": 1.204218999375477, + "grad_norm": 2.0290918350219727, + "learning_rate": 0.00016388048501350533, + "loss": 0.7819, + "step": 17354 + }, + { + "epoch": 1.2042883908125737, + "grad_norm": 1.5393340587615967, + "learning_rate": 0.00016387616212518703, + "loss": 0.5656, + "step": 17355 + }, + { + "epoch": 1.2043577822496703, + "grad_norm": 1.8403764963150024, + "learning_rate": 0.00016387183903521902, + "loss": 0.5654, + "step": 17356 + }, + { + "epoch": 1.204427173686767, + "grad_norm": 4.068338871002197, + "learning_rate": 0.00016386751574361507, + "loss": 0.7514, + "step": 17357 + }, + { + "epoch": 1.2044965651238637, + "grad_norm": 2.210050582885742, + "learning_rate": 0.00016386319225038877, + "loss": 0.5644, + "step": 17358 + }, + { + "epoch": 1.2045659565609603, + "grad_norm": 1.4380064010620117, + "learning_rate": 0.0001638588685555538, + "loss": 0.5452, + "step": 17359 + }, + { + "epoch": 1.2046353479980572, + "grad_norm": 1.7323843240737915, + "learning_rate": 0.00016385454465912374, + "loss": 0.6649, + "step": 17360 + }, + { + "epoch": 1.2047047394351538, + "grad_norm": 1.5589582920074463, + "learning_rate": 0.00016385022056111234, + "loss": 0.6731, + "step": 17361 + }, + { + "epoch": 1.2047741308722504, + "grad_norm": 2.1797380447387695, + "learning_rate": 0.00016384589626153323, + "loss": 0.6339, + "step": 17362 + }, + { + "epoch": 1.204843522309347, + "grad_norm": 2.5016205310821533, + "learning_rate": 0.00016384157176039995, + "loss": 0.7266, + "step": 17363 + }, + { + "epoch": 1.2049129137464436, + "grad_norm": 1.8708441257476807, + "learning_rate": 0.00016383724705772628, + "loss": 0.7109, + "step": 17364 + }, + { + "epoch": 1.2049823051835404, + "grad_norm": 1.9425060749053955, + "learning_rate": 0.00016383292215352583, + "loss": 0.6204, + "step": 17365 + }, + { + "epoch": 1.205051696620637, + "grad_norm": 1.7055470943450928, + "learning_rate": 0.00016382859704781226, + "loss": 0.6682, + "step": 17366 + }, + { + "epoch": 1.2051210880577337, + "grad_norm": 1.7216657400131226, + "learning_rate": 0.0001638242717405992, + "loss": 0.7456, + "step": 17367 + }, + { + "epoch": 1.2051904794948303, + "grad_norm": 1.9286597967147827, + "learning_rate": 0.0001638199462319003, + "loss": 0.5231, + "step": 17368 + }, + { + "epoch": 1.2052598709319269, + "grad_norm": 1.6565265655517578, + "learning_rate": 0.00016381562052172929, + "loss": 0.6497, + "step": 17369 + }, + { + "epoch": 1.2053292623690237, + "grad_norm": 2.535421848297119, + "learning_rate": 0.00016381129461009972, + "loss": 0.7369, + "step": 17370 + }, + { + "epoch": 1.2053986538061203, + "grad_norm": 1.87631356716156, + "learning_rate": 0.0001638069684970253, + "loss": 0.5489, + "step": 17371 + }, + { + "epoch": 1.205468045243217, + "grad_norm": 1.6272408962249756, + "learning_rate": 0.0001638026421825197, + "loss": 0.9664, + "step": 17372 + }, + { + "epoch": 1.2055374366803138, + "grad_norm": 2.292912483215332, + "learning_rate": 0.00016379831566659656, + "loss": 0.7059, + "step": 17373 + }, + { + "epoch": 1.2056068281174104, + "grad_norm": 2.4195332527160645, + "learning_rate": 0.00016379398894926953, + "loss": 0.8105, + "step": 17374 + }, + { + "epoch": 1.205676219554507, + "grad_norm": 2.1145331859588623, + "learning_rate": 0.00016378966203055227, + "loss": 0.7971, + "step": 17375 + }, + { + "epoch": 1.2057456109916036, + "grad_norm": 1.9839916229248047, + "learning_rate": 0.00016378533491045847, + "loss": 0.7879, + "step": 17376 + }, + { + "epoch": 1.2058150024287002, + "grad_norm": 1.5396499633789062, + "learning_rate": 0.00016378100758900177, + "loss": 0.5721, + "step": 17377 + }, + { + "epoch": 1.205884393865797, + "grad_norm": 2.2878172397613525, + "learning_rate": 0.00016377668006619582, + "loss": 0.4871, + "step": 17378 + }, + { + "epoch": 1.2059537853028937, + "grad_norm": 2.5460667610168457, + "learning_rate": 0.00016377235234205429, + "loss": 0.9029, + "step": 17379 + }, + { + "epoch": 1.2060231767399903, + "grad_norm": 1.3837298154830933, + "learning_rate": 0.00016376802441659085, + "loss": 0.562, + "step": 17380 + }, + { + "epoch": 1.2060925681770869, + "grad_norm": 1.492263913154602, + "learning_rate": 0.0001637636962898191, + "loss": 0.3881, + "step": 17381 + }, + { + "epoch": 1.2061619596141835, + "grad_norm": 2.3739044666290283, + "learning_rate": 0.00016375936796175284, + "loss": 0.7606, + "step": 17382 + }, + { + "epoch": 1.2062313510512803, + "grad_norm": 1.7432492971420288, + "learning_rate": 0.00016375503943240563, + "loss": 0.6094, + "step": 17383 + }, + { + "epoch": 1.206300742488377, + "grad_norm": 1.6235883235931396, + "learning_rate": 0.00016375071070179114, + "loss": 0.6721, + "step": 17384 + }, + { + "epoch": 1.2063701339254735, + "grad_norm": 3.2218096256256104, + "learning_rate": 0.00016374638176992306, + "loss": 0.8628, + "step": 17385 + }, + { + "epoch": 1.2064395253625704, + "grad_norm": 1.713822364807129, + "learning_rate": 0.00016374205263681505, + "loss": 0.6594, + "step": 17386 + }, + { + "epoch": 1.206508916799667, + "grad_norm": 1.3695998191833496, + "learning_rate": 0.00016373772330248077, + "loss": 0.6051, + "step": 17387 + }, + { + "epoch": 1.2065783082367636, + "grad_norm": 1.3623788356781006, + "learning_rate": 0.00016373339376693385, + "loss": 0.5249, + "step": 17388 + }, + { + "epoch": 1.2066476996738602, + "grad_norm": 1.7311294078826904, + "learning_rate": 0.00016372906403018807, + "loss": 0.7899, + "step": 17389 + }, + { + "epoch": 1.2067170911109568, + "grad_norm": 2.091806411743164, + "learning_rate": 0.00016372473409225703, + "loss": 0.8018, + "step": 17390 + }, + { + "epoch": 1.2067864825480537, + "grad_norm": 1.6154042482376099, + "learning_rate": 0.00016372040395315436, + "loss": 0.6505, + "step": 17391 + }, + { + "epoch": 1.2068558739851503, + "grad_norm": 1.454484462738037, + "learning_rate": 0.00016371607361289375, + "loss": 0.639, + "step": 17392 + }, + { + "epoch": 1.2069252654222469, + "grad_norm": 1.869723916053772, + "learning_rate": 0.00016371174307148895, + "loss": 0.6091, + "step": 17393 + }, + { + "epoch": 1.2069946568593435, + "grad_norm": 3.563755989074707, + "learning_rate": 0.00016370741232895353, + "loss": 0.7522, + "step": 17394 + }, + { + "epoch": 1.20706404829644, + "grad_norm": 1.9326860904693604, + "learning_rate": 0.00016370308138530118, + "loss": 0.8158, + "step": 17395 + }, + { + "epoch": 1.207133439733537, + "grad_norm": 1.777464509010315, + "learning_rate": 0.00016369875024054565, + "loss": 0.5736, + "step": 17396 + }, + { + "epoch": 1.2072028311706335, + "grad_norm": 1.5772584676742554, + "learning_rate": 0.0001636944188947005, + "loss": 0.5468, + "step": 17397 + }, + { + "epoch": 1.2072722226077301, + "grad_norm": 2.859790325164795, + "learning_rate": 0.00016369008734777948, + "loss": 1.1065, + "step": 17398 + }, + { + "epoch": 1.207341614044827, + "grad_norm": 2.274930477142334, + "learning_rate": 0.00016368575559979628, + "loss": 0.9538, + "step": 17399 + }, + { + "epoch": 1.2074110054819236, + "grad_norm": 1.387427806854248, + "learning_rate": 0.0001636814236507645, + "loss": 0.4423, + "step": 17400 + }, + { + "epoch": 1.2074803969190202, + "grad_norm": 2.640042781829834, + "learning_rate": 0.00016367709150069785, + "loss": 0.8356, + "step": 17401 + }, + { + "epoch": 1.2075497883561168, + "grad_norm": 2.0224411487579346, + "learning_rate": 0.00016367275914961002, + "loss": 0.7345, + "step": 17402 + }, + { + "epoch": 1.2076191797932134, + "grad_norm": 1.7042368650436401, + "learning_rate": 0.0001636684265975147, + "loss": 0.6342, + "step": 17403 + }, + { + "epoch": 1.2076885712303103, + "grad_norm": 1.7640483379364014, + "learning_rate": 0.00016366409384442553, + "loss": 0.5443, + "step": 17404 + }, + { + "epoch": 1.2077579626674069, + "grad_norm": 1.5725531578063965, + "learning_rate": 0.0001636597608903562, + "loss": 0.5214, + "step": 17405 + }, + { + "epoch": 1.2078273541045035, + "grad_norm": 2.025101900100708, + "learning_rate": 0.00016365542773532035, + "loss": 0.4607, + "step": 17406 + }, + { + "epoch": 1.2078967455416, + "grad_norm": 2.437767267227173, + "learning_rate": 0.00016365109437933178, + "loss": 0.7155, + "step": 17407 + }, + { + "epoch": 1.207966136978697, + "grad_norm": 1.9183461666107178, + "learning_rate": 0.00016364676082240403, + "loss": 0.6217, + "step": 17408 + }, + { + "epoch": 1.2080355284157935, + "grad_norm": 1.9546343088150024, + "learning_rate": 0.00016364242706455085, + "loss": 0.5525, + "step": 17409 + }, + { + "epoch": 1.2081049198528901, + "grad_norm": 2.0155272483825684, + "learning_rate": 0.0001636380931057859, + "loss": 0.6215, + "step": 17410 + }, + { + "epoch": 1.2081743112899868, + "grad_norm": 1.8129501342773438, + "learning_rate": 0.0001636337589461229, + "loss": 0.6234, + "step": 17411 + }, + { + "epoch": 1.2082437027270836, + "grad_norm": 1.8193528652191162, + "learning_rate": 0.00016362942458557548, + "loss": 0.6992, + "step": 17412 + }, + { + "epoch": 1.2083130941641802, + "grad_norm": 1.9330185651779175, + "learning_rate": 0.00016362509002415738, + "loss": 0.6513, + "step": 17413 + }, + { + "epoch": 1.2083824856012768, + "grad_norm": 2.172538995742798, + "learning_rate": 0.00016362075526188226, + "loss": 0.779, + "step": 17414 + }, + { + "epoch": 1.2084518770383734, + "grad_norm": 2.618896722793579, + "learning_rate": 0.00016361642029876376, + "loss": 0.6031, + "step": 17415 + }, + { + "epoch": 1.20852126847547, + "grad_norm": 2.0593619346618652, + "learning_rate": 0.0001636120851348156, + "loss": 0.5915, + "step": 17416 + }, + { + "epoch": 1.2085906599125669, + "grad_norm": 1.7683618068695068, + "learning_rate": 0.0001636077497700515, + "loss": 0.5652, + "step": 17417 + }, + { + "epoch": 1.2086600513496635, + "grad_norm": 2.0732874870300293, + "learning_rate": 0.0001636034142044851, + "loss": 0.7809, + "step": 17418 + }, + { + "epoch": 1.20872944278676, + "grad_norm": 1.6970182657241821, + "learning_rate": 0.0001635990784381301, + "loss": 0.6265, + "step": 17419 + }, + { + "epoch": 1.2087988342238567, + "grad_norm": 1.7142438888549805, + "learning_rate": 0.00016359474247100016, + "loss": 0.712, + "step": 17420 + }, + { + "epoch": 1.2088682256609535, + "grad_norm": 1.9047398567199707, + "learning_rate": 0.00016359040630310907, + "loss": 0.9702, + "step": 17421 + }, + { + "epoch": 1.2089376170980501, + "grad_norm": 1.534883975982666, + "learning_rate": 0.0001635860699344704, + "loss": 0.4536, + "step": 17422 + }, + { + "epoch": 1.2090070085351468, + "grad_norm": 2.402186155319214, + "learning_rate": 0.0001635817333650979, + "loss": 0.8373, + "step": 17423 + }, + { + "epoch": 1.2090763999722434, + "grad_norm": 1.3487356901168823, + "learning_rate": 0.0001635773965950052, + "loss": 0.479, + "step": 17424 + }, + { + "epoch": 1.2091457914093402, + "grad_norm": 1.5157661437988281, + "learning_rate": 0.00016357305962420607, + "loss": 0.5509, + "step": 17425 + }, + { + "epoch": 1.2092151828464368, + "grad_norm": 1.8557096719741821, + "learning_rate": 0.00016356872245271414, + "loss": 0.4837, + "step": 17426 + }, + { + "epoch": 1.2092845742835334, + "grad_norm": 1.5759761333465576, + "learning_rate": 0.00016356438508054316, + "loss": 0.77, + "step": 17427 + }, + { + "epoch": 1.20935396572063, + "grad_norm": 1.9065159559249878, + "learning_rate": 0.00016356004750770677, + "loss": 0.7233, + "step": 17428 + }, + { + "epoch": 1.2094233571577266, + "grad_norm": 1.9525752067565918, + "learning_rate": 0.00016355570973421873, + "loss": 0.6209, + "step": 17429 + }, + { + "epoch": 1.2094927485948235, + "grad_norm": 2.0341732501983643, + "learning_rate": 0.00016355137176009263, + "loss": 0.6987, + "step": 17430 + }, + { + "epoch": 1.20956214003192, + "grad_norm": 1.7944034337997437, + "learning_rate": 0.00016354703358534224, + "loss": 0.6212, + "step": 17431 + }, + { + "epoch": 1.2096315314690167, + "grad_norm": 1.5252960920333862, + "learning_rate": 0.00016354269520998124, + "loss": 0.4163, + "step": 17432 + }, + { + "epoch": 1.2097009229061133, + "grad_norm": 2.451199769973755, + "learning_rate": 0.00016353835663402333, + "loss": 0.818, + "step": 17433 + }, + { + "epoch": 1.2097703143432101, + "grad_norm": 1.834149956703186, + "learning_rate": 0.00016353401785748219, + "loss": 0.6806, + "step": 17434 + }, + { + "epoch": 1.2098397057803068, + "grad_norm": 1.9857311248779297, + "learning_rate": 0.00016352967888037155, + "loss": 0.7688, + "step": 17435 + }, + { + "epoch": 1.2099090972174034, + "grad_norm": 2.873400926589966, + "learning_rate": 0.00016352533970270507, + "loss": 0.5649, + "step": 17436 + }, + { + "epoch": 1.2099784886545, + "grad_norm": 1.750522494316101, + "learning_rate": 0.00016352100032449647, + "loss": 0.608, + "step": 17437 + }, + { + "epoch": 1.2100478800915968, + "grad_norm": 2.1250579357147217, + "learning_rate": 0.00016351666074575946, + "loss": 0.5793, + "step": 17438 + }, + { + "epoch": 1.2101172715286934, + "grad_norm": 2.229538679122925, + "learning_rate": 0.00016351232096650765, + "loss": 0.9369, + "step": 17439 + }, + { + "epoch": 1.21018666296579, + "grad_norm": 1.5077769756317139, + "learning_rate": 0.00016350798098675485, + "loss": 0.7194, + "step": 17440 + }, + { + "epoch": 1.2102560544028866, + "grad_norm": 2.1440107822418213, + "learning_rate": 0.0001635036408065147, + "loss": 0.6715, + "step": 17441 + }, + { + "epoch": 1.2103254458399832, + "grad_norm": 2.454320192337036, + "learning_rate": 0.000163499300425801, + "loss": 0.7486, + "step": 17442 + }, + { + "epoch": 1.21039483727708, + "grad_norm": 1.8287947177886963, + "learning_rate": 0.00016349495984462734, + "loss": 0.6366, + "step": 17443 + }, + { + "epoch": 1.2104642287141767, + "grad_norm": 1.768294095993042, + "learning_rate": 0.00016349061906300742, + "loss": 0.7963, + "step": 17444 + }, + { + "epoch": 1.2105336201512733, + "grad_norm": 1.8261756896972656, + "learning_rate": 0.000163486278080955, + "loss": 0.7021, + "step": 17445 + }, + { + "epoch": 1.21060301158837, + "grad_norm": 1.9085639715194702, + "learning_rate": 0.00016348193689848378, + "loss": 0.6899, + "step": 17446 + }, + { + "epoch": 1.2106724030254667, + "grad_norm": 1.5881792306900024, + "learning_rate": 0.00016347759551560746, + "loss": 0.891, + "step": 17447 + }, + { + "epoch": 1.2107417944625634, + "grad_norm": 1.8945424556732178, + "learning_rate": 0.0001634732539323397, + "loss": 0.7476, + "step": 17448 + }, + { + "epoch": 1.21081118589966, + "grad_norm": 1.8268483877182007, + "learning_rate": 0.00016346891214869425, + "loss": 0.638, + "step": 17449 + }, + { + "epoch": 1.2108805773367566, + "grad_norm": 2.8384921550750732, + "learning_rate": 0.00016346457016468483, + "loss": 0.8826, + "step": 17450 + }, + { + "epoch": 1.2109499687738534, + "grad_norm": 1.9590733051300049, + "learning_rate": 0.0001634602279803251, + "loss": 0.5528, + "step": 17451 + }, + { + "epoch": 1.21101936021095, + "grad_norm": 2.272573709487915, + "learning_rate": 0.00016345588559562883, + "loss": 0.9028, + "step": 17452 + }, + { + "epoch": 1.2110887516480466, + "grad_norm": 1.7882471084594727, + "learning_rate": 0.00016345154301060965, + "loss": 0.7222, + "step": 17453 + }, + { + "epoch": 1.2111581430851432, + "grad_norm": 1.5944032669067383, + "learning_rate": 0.00016344720022528134, + "loss": 0.6239, + "step": 17454 + }, + { + "epoch": 1.2112275345222399, + "grad_norm": 2.2978930473327637, + "learning_rate": 0.00016344285723965757, + "loss": 0.8629, + "step": 17455 + }, + { + "epoch": 1.2112969259593367, + "grad_norm": 1.9708583354949951, + "learning_rate": 0.00016343851405375208, + "loss": 0.7557, + "step": 17456 + }, + { + "epoch": 1.2113663173964333, + "grad_norm": 2.606691360473633, + "learning_rate": 0.00016343417066757855, + "loss": 0.6627, + "step": 17457 + }, + { + "epoch": 1.21143570883353, + "grad_norm": 1.8571186065673828, + "learning_rate": 0.00016342982708115067, + "loss": 0.6161, + "step": 17458 + }, + { + "epoch": 1.2115051002706265, + "grad_norm": 2.5959393978118896, + "learning_rate": 0.00016342548329448224, + "loss": 0.699, + "step": 17459 + }, + { + "epoch": 1.2115744917077234, + "grad_norm": 1.2123289108276367, + "learning_rate": 0.00016342113930758686, + "loss": 0.4825, + "step": 17460 + }, + { + "epoch": 1.21164388314482, + "grad_norm": 3.403961420059204, + "learning_rate": 0.00016341679512047832, + "loss": 0.6959, + "step": 17461 + }, + { + "epoch": 1.2117132745819166, + "grad_norm": 1.8269546031951904, + "learning_rate": 0.00016341245073317034, + "loss": 0.8177, + "step": 17462 + }, + { + "epoch": 1.2117826660190132, + "grad_norm": 1.6742452383041382, + "learning_rate": 0.00016340810614567662, + "loss": 0.7046, + "step": 17463 + }, + { + "epoch": 1.21185205745611, + "grad_norm": 1.514987826347351, + "learning_rate": 0.00016340376135801082, + "loss": 0.6424, + "step": 17464 + }, + { + "epoch": 1.2119214488932066, + "grad_norm": 1.730650782585144, + "learning_rate": 0.00016339941637018673, + "loss": 0.5143, + "step": 17465 + }, + { + "epoch": 1.2119908403303032, + "grad_norm": 3.014744520187378, + "learning_rate": 0.00016339507118221804, + "loss": 0.8774, + "step": 17466 + }, + { + "epoch": 1.2120602317673999, + "grad_norm": 1.4166088104248047, + "learning_rate": 0.00016339072579411843, + "loss": 0.5455, + "step": 17467 + }, + { + "epoch": 1.2121296232044965, + "grad_norm": 2.2417967319488525, + "learning_rate": 0.00016338638020590169, + "loss": 0.8857, + "step": 17468 + }, + { + "epoch": 1.2121990146415933, + "grad_norm": 1.5689194202423096, + "learning_rate": 0.0001633820344175815, + "loss": 0.5894, + "step": 17469 + }, + { + "epoch": 1.21226840607869, + "grad_norm": 1.7199147939682007, + "learning_rate": 0.00016337768842917156, + "loss": 0.5802, + "step": 17470 + }, + { + "epoch": 1.2123377975157865, + "grad_norm": 1.631043791770935, + "learning_rate": 0.0001633733422406856, + "loss": 0.7136, + "step": 17471 + }, + { + "epoch": 1.2124071889528831, + "grad_norm": 1.9819339513778687, + "learning_rate": 0.00016336899585213738, + "loss": 0.608, + "step": 17472 + }, + { + "epoch": 1.21247658038998, + "grad_norm": 1.869123935699463, + "learning_rate": 0.0001633646492635406, + "loss": 0.684, + "step": 17473 + }, + { + "epoch": 1.2125459718270766, + "grad_norm": 1.8764137029647827, + "learning_rate": 0.00016336030247490898, + "loss": 0.7072, + "step": 17474 + }, + { + "epoch": 1.2126153632641732, + "grad_norm": 1.4786272048950195, + "learning_rate": 0.0001633559554862562, + "loss": 0.5594, + "step": 17475 + }, + { + "epoch": 1.2126847547012698, + "grad_norm": 1.8838304281234741, + "learning_rate": 0.00016335160829759605, + "loss": 0.552, + "step": 17476 + }, + { + "epoch": 1.2127541461383666, + "grad_norm": 1.6245232820510864, + "learning_rate": 0.0001633472609089422, + "loss": 0.6536, + "step": 17477 + }, + { + "epoch": 1.2128235375754632, + "grad_norm": 1.4124048948287964, + "learning_rate": 0.0001633429133203084, + "loss": 0.6043, + "step": 17478 + }, + { + "epoch": 1.2128929290125599, + "grad_norm": 1.5435329675674438, + "learning_rate": 0.0001633385655317084, + "loss": 0.6001, + "step": 17479 + }, + { + "epoch": 1.2129623204496565, + "grad_norm": 2.2380635738372803, + "learning_rate": 0.00016333421754315586, + "loss": 0.5333, + "step": 17480 + }, + { + "epoch": 1.213031711886753, + "grad_norm": 3.0300705432891846, + "learning_rate": 0.00016332986935466454, + "loss": 0.645, + "step": 17481 + }, + { + "epoch": 1.21310110332385, + "grad_norm": 2.3004045486450195, + "learning_rate": 0.00016332552096624817, + "loss": 0.8314, + "step": 17482 + }, + { + "epoch": 1.2131704947609465, + "grad_norm": 1.892543077468872, + "learning_rate": 0.00016332117237792052, + "loss": 0.7902, + "step": 17483 + }, + { + "epoch": 1.2132398861980431, + "grad_norm": 1.89604914188385, + "learning_rate": 0.00016331682358969527, + "loss": 0.6063, + "step": 17484 + }, + { + "epoch": 1.2133092776351397, + "grad_norm": 1.9071447849273682, + "learning_rate": 0.00016331247460158612, + "loss": 0.4985, + "step": 17485 + }, + { + "epoch": 1.2133786690722366, + "grad_norm": 2.0368528366088867, + "learning_rate": 0.00016330812541360686, + "loss": 0.7376, + "step": 17486 + }, + { + "epoch": 1.2134480605093332, + "grad_norm": 1.5487182140350342, + "learning_rate": 0.00016330377602577118, + "loss": 0.6068, + "step": 17487 + }, + { + "epoch": 1.2135174519464298, + "grad_norm": 1.5879350900650024, + "learning_rate": 0.0001632994264380928, + "loss": 0.6762, + "step": 17488 + }, + { + "epoch": 1.2135868433835264, + "grad_norm": 1.7914179563522339, + "learning_rate": 0.0001632950766505855, + "loss": 0.6373, + "step": 17489 + }, + { + "epoch": 1.2136562348206232, + "grad_norm": 1.3929702043533325, + "learning_rate": 0.00016329072666326303, + "loss": 0.6675, + "step": 17490 + }, + { + "epoch": 1.2137256262577198, + "grad_norm": 2.0035884380340576, + "learning_rate": 0.00016328637647613904, + "loss": 0.7135, + "step": 17491 + }, + { + "epoch": 1.2137950176948165, + "grad_norm": 2.2920937538146973, + "learning_rate": 0.00016328202608922724, + "loss": 0.7788, + "step": 17492 + }, + { + "epoch": 1.213864409131913, + "grad_norm": 2.4051456451416016, + "learning_rate": 0.00016327767550254152, + "loss": 0.8713, + "step": 17493 + }, + { + "epoch": 1.2139338005690097, + "grad_norm": 1.4221783876419067, + "learning_rate": 0.00016327332471609548, + "loss": 0.6891, + "step": 17494 + }, + { + "epoch": 1.2140031920061065, + "grad_norm": 1.9276187419891357, + "learning_rate": 0.00016326897372990288, + "loss": 0.6502, + "step": 17495 + }, + { + "epoch": 1.2140725834432031, + "grad_norm": 1.6612547636032104, + "learning_rate": 0.00016326462254397748, + "loss": 0.675, + "step": 17496 + }, + { + "epoch": 1.2141419748802997, + "grad_norm": 1.9394824504852295, + "learning_rate": 0.00016326027115833304, + "loss": 0.6076, + "step": 17497 + }, + { + "epoch": 1.2142113663173963, + "grad_norm": 1.9058810472488403, + "learning_rate": 0.00016325591957298325, + "loss": 0.7766, + "step": 17498 + }, + { + "epoch": 1.2142807577544932, + "grad_norm": 2.2743475437164307, + "learning_rate": 0.00016325156778794183, + "loss": 1.0133, + "step": 17499 + }, + { + "epoch": 1.2143501491915898, + "grad_norm": 3.3384649753570557, + "learning_rate": 0.00016324721580322259, + "loss": 0.9045, + "step": 17500 + }, + { + "epoch": 1.2144195406286864, + "grad_norm": 1.7314051389694214, + "learning_rate": 0.00016324286361883917, + "loss": 0.7368, + "step": 17501 + }, + { + "epoch": 1.214488932065783, + "grad_norm": 2.028818368911743, + "learning_rate": 0.00016323851123480542, + "loss": 0.7296, + "step": 17502 + }, + { + "epoch": 1.2145583235028798, + "grad_norm": 1.6288822889328003, + "learning_rate": 0.000163234158651135, + "loss": 0.5525, + "step": 17503 + }, + { + "epoch": 1.2146277149399765, + "grad_norm": 1.8800231218338013, + "learning_rate": 0.0001632298058678417, + "loss": 0.6113, + "step": 17504 + }, + { + "epoch": 1.214697106377073, + "grad_norm": 1.616384744644165, + "learning_rate": 0.0001632254528849392, + "loss": 0.6252, + "step": 17505 + }, + { + "epoch": 1.2147664978141697, + "grad_norm": 1.7060545682907104, + "learning_rate": 0.00016322109970244136, + "loss": 0.6551, + "step": 17506 + }, + { + "epoch": 1.2148358892512663, + "grad_norm": 2.455716609954834, + "learning_rate": 0.00016321674632036176, + "loss": 0.6058, + "step": 17507 + }, + { + "epoch": 1.2149052806883631, + "grad_norm": 1.764058232307434, + "learning_rate": 0.00016321239273871428, + "loss": 0.5231, + "step": 17508 + }, + { + "epoch": 1.2149746721254597, + "grad_norm": 1.4696669578552246, + "learning_rate": 0.0001632080389575126, + "loss": 0.6077, + "step": 17509 + }, + { + "epoch": 1.2150440635625563, + "grad_norm": 1.1218843460083008, + "learning_rate": 0.00016320368497677043, + "loss": 0.5017, + "step": 17510 + }, + { + "epoch": 1.215113454999653, + "grad_norm": 1.880536675453186, + "learning_rate": 0.0001631993307965016, + "loss": 0.5382, + "step": 17511 + }, + { + "epoch": 1.2151828464367498, + "grad_norm": 1.9183261394500732, + "learning_rate": 0.00016319497641671983, + "loss": 0.5245, + "step": 17512 + }, + { + "epoch": 1.2152522378738464, + "grad_norm": 1.7909235954284668, + "learning_rate": 0.0001631906218374388, + "loss": 0.6776, + "step": 17513 + }, + { + "epoch": 1.215321629310943, + "grad_norm": 1.9881268739700317, + "learning_rate": 0.00016318626705867236, + "loss": 0.8478, + "step": 17514 + }, + { + "epoch": 1.2153910207480396, + "grad_norm": 1.5149037837982178, + "learning_rate": 0.00016318191208043416, + "loss": 0.6531, + "step": 17515 + }, + { + "epoch": 1.2154604121851365, + "grad_norm": 1.72669517993927, + "learning_rate": 0.000163177556902738, + "loss": 0.5589, + "step": 17516 + }, + { + "epoch": 1.215529803622233, + "grad_norm": 2.2441253662109375, + "learning_rate": 0.00016317320152559766, + "loss": 0.6498, + "step": 17517 + }, + { + "epoch": 1.2155991950593297, + "grad_norm": 2.098416328430176, + "learning_rate": 0.00016316884594902683, + "loss": 0.9213, + "step": 17518 + }, + { + "epoch": 1.2156685864964263, + "grad_norm": 2.0838019847869873, + "learning_rate": 0.00016316449017303927, + "loss": 0.5291, + "step": 17519 + }, + { + "epoch": 1.215737977933523, + "grad_norm": 1.3356623649597168, + "learning_rate": 0.00016316013419764874, + "loss": 0.6263, + "step": 17520 + }, + { + "epoch": 1.2158073693706197, + "grad_norm": 2.274033308029175, + "learning_rate": 0.00016315577802286906, + "loss": 0.725, + "step": 17521 + }, + { + "epoch": 1.2158767608077163, + "grad_norm": 1.5601730346679688, + "learning_rate": 0.00016315142164871385, + "loss": 0.6867, + "step": 17522 + }, + { + "epoch": 1.215946152244813, + "grad_norm": 1.7987805604934692, + "learning_rate": 0.00016314706507519696, + "loss": 0.6723, + "step": 17523 + }, + { + "epoch": 1.2160155436819096, + "grad_norm": 1.4901437759399414, + "learning_rate": 0.00016314270830233206, + "loss": 0.7067, + "step": 17524 + }, + { + "epoch": 1.2160849351190064, + "grad_norm": 2.5544705390930176, + "learning_rate": 0.00016313835133013302, + "loss": 0.5458, + "step": 17525 + }, + { + "epoch": 1.216154326556103, + "grad_norm": 2.4945075511932373, + "learning_rate": 0.0001631339941586135, + "loss": 0.6217, + "step": 17526 + }, + { + "epoch": 1.2162237179931996, + "grad_norm": 1.8444303274154663, + "learning_rate": 0.0001631296367877873, + "loss": 0.7611, + "step": 17527 + }, + { + "epoch": 1.2162931094302962, + "grad_norm": 2.161299228668213, + "learning_rate": 0.00016312527921766816, + "loss": 0.7265, + "step": 17528 + }, + { + "epoch": 1.216362500867393, + "grad_norm": 2.358384370803833, + "learning_rate": 0.00016312092144826982, + "loss": 0.856, + "step": 17529 + }, + { + "epoch": 1.2164318923044897, + "grad_norm": 2.1451501846313477, + "learning_rate": 0.00016311656347960603, + "loss": 0.6695, + "step": 17530 + }, + { + "epoch": 1.2165012837415863, + "grad_norm": 1.7419275045394897, + "learning_rate": 0.00016311220531169064, + "loss": 0.6612, + "step": 17531 + }, + { + "epoch": 1.216570675178683, + "grad_norm": 1.8992054462432861, + "learning_rate": 0.0001631078469445373, + "loss": 0.6746, + "step": 17532 + }, + { + "epoch": 1.2166400666157795, + "grad_norm": 1.8080992698669434, + "learning_rate": 0.0001631034883781598, + "loss": 0.6759, + "step": 17533 + }, + { + "epoch": 1.2167094580528763, + "grad_norm": 1.5632271766662598, + "learning_rate": 0.00016309912961257191, + "loss": 0.5651, + "step": 17534 + }, + { + "epoch": 1.216778849489973, + "grad_norm": 1.9381022453308105, + "learning_rate": 0.00016309477064778741, + "loss": 0.581, + "step": 17535 + }, + { + "epoch": 1.2168482409270696, + "grad_norm": 2.132783889770508, + "learning_rate": 0.00016309041148382004, + "loss": 0.3599, + "step": 17536 + }, + { + "epoch": 1.2169176323641662, + "grad_norm": 1.6143718957901, + "learning_rate": 0.00016308605212068353, + "loss": 0.5992, + "step": 17537 + }, + { + "epoch": 1.216987023801263, + "grad_norm": 3.1878199577331543, + "learning_rate": 0.00016308169255839172, + "loss": 0.7804, + "step": 17538 + }, + { + "epoch": 1.2170564152383596, + "grad_norm": 1.6492164134979248, + "learning_rate": 0.00016307733279695828, + "loss": 0.8373, + "step": 17539 + }, + { + "epoch": 1.2171258066754562, + "grad_norm": 1.6527397632598877, + "learning_rate": 0.00016307297283639705, + "loss": 0.6793, + "step": 17540 + }, + { + "epoch": 1.2171951981125528, + "grad_norm": 2.1523168087005615, + "learning_rate": 0.00016306861267672177, + "loss": 0.59, + "step": 17541 + }, + { + "epoch": 1.2172645895496497, + "grad_norm": 1.89895498752594, + "learning_rate": 0.00016306425231794614, + "loss": 0.6027, + "step": 17542 + }, + { + "epoch": 1.2173339809867463, + "grad_norm": 1.6534806489944458, + "learning_rate": 0.00016305989176008403, + "loss": 0.5605, + "step": 17543 + }, + { + "epoch": 1.2174033724238429, + "grad_norm": 1.7408887147903442, + "learning_rate": 0.00016305553100314914, + "loss": 0.5952, + "step": 17544 + }, + { + "epoch": 1.2174727638609395, + "grad_norm": 2.082855224609375, + "learning_rate": 0.00016305117004715528, + "loss": 0.8124, + "step": 17545 + }, + { + "epoch": 1.217542155298036, + "grad_norm": 4.539830207824707, + "learning_rate": 0.00016304680889211615, + "loss": 0.7883, + "step": 17546 + }, + { + "epoch": 1.217611546735133, + "grad_norm": 2.139106035232544, + "learning_rate": 0.00016304244753804562, + "loss": 0.8966, + "step": 17547 + }, + { + "epoch": 1.2176809381722296, + "grad_norm": 2.0467400550842285, + "learning_rate": 0.00016303808598495733, + "loss": 0.745, + "step": 17548 + }, + { + "epoch": 1.2177503296093262, + "grad_norm": 2.745980978012085, + "learning_rate": 0.00016303372423286515, + "loss": 0.7354, + "step": 17549 + }, + { + "epoch": 1.2178197210464228, + "grad_norm": 2.134939193725586, + "learning_rate": 0.0001630293622817828, + "loss": 0.8061, + "step": 17550 + }, + { + "epoch": 1.2178891124835196, + "grad_norm": 2.9749667644500732, + "learning_rate": 0.0001630250001317241, + "loss": 0.5351, + "step": 17551 + }, + { + "epoch": 1.2179585039206162, + "grad_norm": 1.5787533521652222, + "learning_rate": 0.00016302063778270278, + "loss": 0.5593, + "step": 17552 + }, + { + "epoch": 1.2180278953577128, + "grad_norm": 1.5943729877471924, + "learning_rate": 0.0001630162752347326, + "loss": 0.4783, + "step": 17553 + }, + { + "epoch": 1.2180972867948094, + "grad_norm": 2.0695860385894775, + "learning_rate": 0.00016301191248782734, + "loss": 0.7419, + "step": 17554 + }, + { + "epoch": 1.2181666782319063, + "grad_norm": 2.22930645942688, + "learning_rate": 0.0001630075495420008, + "loss": 0.941, + "step": 17555 + }, + { + "epoch": 1.2182360696690029, + "grad_norm": 2.0138893127441406, + "learning_rate": 0.00016300318639726676, + "loss": 0.6871, + "step": 17556 + }, + { + "epoch": 1.2183054611060995, + "grad_norm": 2.6623735427856445, + "learning_rate": 0.00016299882305363892, + "loss": 0.6146, + "step": 17557 + }, + { + "epoch": 1.218374852543196, + "grad_norm": 1.650176763534546, + "learning_rate": 0.00016299445951113112, + "loss": 0.6982, + "step": 17558 + }, + { + "epoch": 1.2184442439802927, + "grad_norm": 2.8294692039489746, + "learning_rate": 0.00016299009576975712, + "loss": 0.8354, + "step": 17559 + }, + { + "epoch": 1.2185136354173896, + "grad_norm": 2.3294875621795654, + "learning_rate": 0.00016298573182953072, + "loss": 0.7559, + "step": 17560 + }, + { + "epoch": 1.2185830268544862, + "grad_norm": 1.7438632249832153, + "learning_rate": 0.00016298136769046567, + "loss": 0.6018, + "step": 17561 + }, + { + "epoch": 1.2186524182915828, + "grad_norm": 1.3804055452346802, + "learning_rate": 0.0001629770033525757, + "loss": 0.6176, + "step": 17562 + }, + { + "epoch": 1.2187218097286794, + "grad_norm": 2.1112449169158936, + "learning_rate": 0.0001629726388158747, + "loss": 0.7603, + "step": 17563 + }, + { + "epoch": 1.2187912011657762, + "grad_norm": 1.5712498426437378, + "learning_rate": 0.00016296827408037636, + "loss": 0.6848, + "step": 17564 + }, + { + "epoch": 1.2188605926028728, + "grad_norm": 1.4405041933059692, + "learning_rate": 0.00016296390914609447, + "loss": 0.6487, + "step": 17565 + }, + { + "epoch": 1.2189299840399694, + "grad_norm": 1.5601780414581299, + "learning_rate": 0.00016295954401304282, + "loss": 0.5989, + "step": 17566 + }, + { + "epoch": 1.218999375477066, + "grad_norm": 2.0665910243988037, + "learning_rate": 0.00016295517868123525, + "loss": 0.7077, + "step": 17567 + }, + { + "epoch": 1.2190687669141629, + "grad_norm": 2.362787961959839, + "learning_rate": 0.00016295081315068543, + "loss": 0.7957, + "step": 17568 + }, + { + "epoch": 1.2191381583512595, + "grad_norm": 2.951979637145996, + "learning_rate": 0.0001629464474214072, + "loss": 0.8343, + "step": 17569 + }, + { + "epoch": 1.219207549788356, + "grad_norm": 1.385589838027954, + "learning_rate": 0.00016294208149341438, + "loss": 0.57, + "step": 17570 + }, + { + "epoch": 1.2192769412254527, + "grad_norm": 1.3650349378585815, + "learning_rate": 0.00016293771536672066, + "loss": 0.4611, + "step": 17571 + }, + { + "epoch": 1.2193463326625493, + "grad_norm": 1.698762059211731, + "learning_rate": 0.0001629333490413399, + "loss": 0.5276, + "step": 17572 + }, + { + "epoch": 1.2194157240996462, + "grad_norm": 2.373089551925659, + "learning_rate": 0.00016292898251728586, + "loss": 0.7729, + "step": 17573 + }, + { + "epoch": 1.2194851155367428, + "grad_norm": 1.5998172760009766, + "learning_rate": 0.0001629246157945723, + "loss": 0.6891, + "step": 17574 + }, + { + "epoch": 1.2195545069738394, + "grad_norm": 1.6040980815887451, + "learning_rate": 0.00016292024887321307, + "loss": 0.5904, + "step": 17575 + }, + { + "epoch": 1.2196238984109362, + "grad_norm": 1.794416069984436, + "learning_rate": 0.00016291588175322186, + "loss": 0.6172, + "step": 17576 + }, + { + "epoch": 1.2196932898480328, + "grad_norm": 1.8395015001296997, + "learning_rate": 0.00016291151443461257, + "loss": 0.7094, + "step": 17577 + }, + { + "epoch": 1.2197626812851294, + "grad_norm": 1.6882779598236084, + "learning_rate": 0.0001629071469173989, + "loss": 0.6537, + "step": 17578 + }, + { + "epoch": 1.219832072722226, + "grad_norm": 1.5627702474594116, + "learning_rate": 0.00016290277920159467, + "loss": 0.5598, + "step": 17579 + }, + { + "epoch": 1.2199014641593227, + "grad_norm": 1.8456648588180542, + "learning_rate": 0.00016289841128721365, + "loss": 0.7429, + "step": 17580 + }, + { + "epoch": 1.2199708555964195, + "grad_norm": 1.6641650199890137, + "learning_rate": 0.00016289404317426965, + "loss": 0.685, + "step": 17581 + }, + { + "epoch": 1.220040247033516, + "grad_norm": 1.3982534408569336, + "learning_rate": 0.00016288967486277647, + "loss": 0.3981, + "step": 17582 + }, + { + "epoch": 1.2201096384706127, + "grad_norm": 2.1964316368103027, + "learning_rate": 0.00016288530635274787, + "loss": 0.7706, + "step": 17583 + }, + { + "epoch": 1.2201790299077093, + "grad_norm": 1.731704831123352, + "learning_rate": 0.00016288093764419765, + "loss": 0.6351, + "step": 17584 + }, + { + "epoch": 1.220248421344806, + "grad_norm": 1.6668537855148315, + "learning_rate": 0.0001628765687371396, + "loss": 0.6433, + "step": 17585 + }, + { + "epoch": 1.2203178127819028, + "grad_norm": 1.7062698602676392, + "learning_rate": 0.00016287219963158754, + "loss": 0.6614, + "step": 17586 + }, + { + "epoch": 1.2203872042189994, + "grad_norm": 1.5647528171539307, + "learning_rate": 0.00016286783032755525, + "loss": 0.6935, + "step": 17587 + }, + { + "epoch": 1.220456595656096, + "grad_norm": 1.464485764503479, + "learning_rate": 0.00016286346082505646, + "loss": 0.6908, + "step": 17588 + }, + { + "epoch": 1.2205259870931928, + "grad_norm": 1.5487995147705078, + "learning_rate": 0.00016285909112410507, + "loss": 0.6192, + "step": 17589 + }, + { + "epoch": 1.2205953785302894, + "grad_norm": 1.715000867843628, + "learning_rate": 0.00016285472122471478, + "loss": 0.923, + "step": 17590 + }, + { + "epoch": 1.220664769967386, + "grad_norm": 1.8179328441619873, + "learning_rate": 0.00016285035112689945, + "loss": 0.7578, + "step": 17591 + }, + { + "epoch": 1.2207341614044827, + "grad_norm": 1.7915587425231934, + "learning_rate": 0.00016284598083067287, + "loss": 0.6145, + "step": 17592 + }, + { + "epoch": 1.2208035528415793, + "grad_norm": 1.5275888442993164, + "learning_rate": 0.00016284161033604879, + "loss": 0.4689, + "step": 17593 + }, + { + "epoch": 1.220872944278676, + "grad_norm": 1.557023048400879, + "learning_rate": 0.00016283723964304105, + "loss": 0.693, + "step": 17594 + }, + { + "epoch": 1.2209423357157727, + "grad_norm": 1.8301690816879272, + "learning_rate": 0.00016283286875166342, + "loss": 0.5875, + "step": 17595 + }, + { + "epoch": 1.2210117271528693, + "grad_norm": 2.252185583114624, + "learning_rate": 0.00016282849766192974, + "loss": 0.684, + "step": 17596 + }, + { + "epoch": 1.221081118589966, + "grad_norm": 1.6207317113876343, + "learning_rate": 0.0001628241263738538, + "loss": 0.5762, + "step": 17597 + }, + { + "epoch": 1.2211505100270625, + "grad_norm": 2.0484859943389893, + "learning_rate": 0.00016281975488744932, + "loss": 0.7213, + "step": 17598 + }, + { + "epoch": 1.2212199014641594, + "grad_norm": 1.3168988227844238, + "learning_rate": 0.0001628153832027302, + "loss": 0.4807, + "step": 17599 + }, + { + "epoch": 1.221289292901256, + "grad_norm": 2.12066388130188, + "learning_rate": 0.00016281101131971018, + "loss": 0.8383, + "step": 17600 + }, + { + "epoch": 1.2213586843383526, + "grad_norm": 3.3912341594696045, + "learning_rate": 0.0001628066392384031, + "loss": 0.6149, + "step": 17601 + }, + { + "epoch": 1.2214280757754494, + "grad_norm": 1.9870572090148926, + "learning_rate": 0.00016280226695882278, + "loss": 0.7385, + "step": 17602 + }, + { + "epoch": 1.221497467212546, + "grad_norm": 1.8660426139831543, + "learning_rate": 0.00016279789448098291, + "loss": 0.7703, + "step": 17603 + }, + { + "epoch": 1.2215668586496427, + "grad_norm": 2.118605136871338, + "learning_rate": 0.00016279352180489744, + "loss": 0.9487, + "step": 17604 + }, + { + "epoch": 1.2216362500867393, + "grad_norm": 1.3757325410842896, + "learning_rate": 0.0001627891489305801, + "loss": 0.5168, + "step": 17605 + }, + { + "epoch": 1.2217056415238359, + "grad_norm": 2.165140151977539, + "learning_rate": 0.0001627847758580447, + "loss": 0.9817, + "step": 17606 + }, + { + "epoch": 1.2217750329609327, + "grad_norm": 2.7708780765533447, + "learning_rate": 0.00016278040258730503, + "loss": 1.0317, + "step": 17607 + }, + { + "epoch": 1.2218444243980293, + "grad_norm": 2.6004092693328857, + "learning_rate": 0.0001627760291183749, + "loss": 0.46, + "step": 17608 + }, + { + "epoch": 1.221913815835126, + "grad_norm": 1.6139872074127197, + "learning_rate": 0.00016277165545126816, + "loss": 0.6159, + "step": 17609 + }, + { + "epoch": 1.2219832072722225, + "grad_norm": 1.739380121231079, + "learning_rate": 0.0001627672815859986, + "loss": 0.5916, + "step": 17610 + }, + { + "epoch": 1.2220525987093191, + "grad_norm": 1.8355703353881836, + "learning_rate": 0.00016276290752257998, + "loss": 0.8137, + "step": 17611 + }, + { + "epoch": 1.222121990146416, + "grad_norm": 1.8471698760986328, + "learning_rate": 0.00016275853326102615, + "loss": 0.7926, + "step": 17612 + }, + { + "epoch": 1.2221913815835126, + "grad_norm": 1.4394062757492065, + "learning_rate": 0.0001627541588013509, + "loss": 0.3782, + "step": 17613 + }, + { + "epoch": 1.2222607730206092, + "grad_norm": 2.1102263927459717, + "learning_rate": 0.00016274978414356807, + "loss": 0.7151, + "step": 17614 + }, + { + "epoch": 1.222330164457706, + "grad_norm": 1.9827886819839478, + "learning_rate": 0.00016274540928769144, + "loss": 0.7623, + "step": 17615 + }, + { + "epoch": 1.2223995558948026, + "grad_norm": 1.8798519372940063, + "learning_rate": 0.00016274103423373484, + "loss": 0.6192, + "step": 17616 + }, + { + "epoch": 1.2224689473318993, + "grad_norm": 1.7257550954818726, + "learning_rate": 0.00016273665898171208, + "loss": 0.7086, + "step": 17617 + }, + { + "epoch": 1.2225383387689959, + "grad_norm": 1.8212440013885498, + "learning_rate": 0.00016273228353163698, + "loss": 0.5874, + "step": 17618 + }, + { + "epoch": 1.2226077302060925, + "grad_norm": 2.582768201828003, + "learning_rate": 0.0001627279078835233, + "loss": 1.2724, + "step": 17619 + }, + { + "epoch": 1.2226771216431893, + "grad_norm": 2.0162222385406494, + "learning_rate": 0.00016272353203738494, + "loss": 0.6123, + "step": 17620 + }, + { + "epoch": 1.222746513080286, + "grad_norm": 1.9763160943984985, + "learning_rate": 0.00016271915599323563, + "loss": 0.6799, + "step": 17621 + }, + { + "epoch": 1.2228159045173825, + "grad_norm": 1.862398624420166, + "learning_rate": 0.00016271477975108925, + "loss": 0.6629, + "step": 17622 + }, + { + "epoch": 1.2228852959544791, + "grad_norm": 1.5473662614822388, + "learning_rate": 0.00016271040331095955, + "loss": 0.5781, + "step": 17623 + }, + { + "epoch": 1.2229546873915758, + "grad_norm": 1.4900562763214111, + "learning_rate": 0.00016270602667286038, + "loss": 0.3847, + "step": 17624 + }, + { + "epoch": 1.2230240788286726, + "grad_norm": 1.9382866621017456, + "learning_rate": 0.0001627016498368056, + "loss": 0.6467, + "step": 17625 + }, + { + "epoch": 1.2230934702657692, + "grad_norm": 1.5246590375900269, + "learning_rate": 0.00016269727280280897, + "loss": 0.5157, + "step": 17626 + }, + { + "epoch": 1.2231628617028658, + "grad_norm": 1.743443250656128, + "learning_rate": 0.0001626928955708843, + "loss": 0.6919, + "step": 17627 + }, + { + "epoch": 1.2232322531399626, + "grad_norm": 2.5034525394439697, + "learning_rate": 0.00016268851814104548, + "loss": 0.5703, + "step": 17628 + }, + { + "epoch": 1.2233016445770593, + "grad_norm": 1.696709156036377, + "learning_rate": 0.00016268414051330622, + "loss": 0.7768, + "step": 17629 + }, + { + "epoch": 1.2233710360141559, + "grad_norm": 2.3390603065490723, + "learning_rate": 0.00016267976268768047, + "loss": 0.6416, + "step": 17630 + }, + { + "epoch": 1.2234404274512525, + "grad_norm": 1.8025226593017578, + "learning_rate": 0.00016267538466418192, + "loss": 0.8025, + "step": 17631 + }, + { + "epoch": 1.223509818888349, + "grad_norm": 1.5169968605041504, + "learning_rate": 0.0001626710064428245, + "loss": 0.5572, + "step": 17632 + }, + { + "epoch": 1.223579210325446, + "grad_norm": 1.6576030254364014, + "learning_rate": 0.00016266662802362193, + "loss": 0.5417, + "step": 17633 + }, + { + "epoch": 1.2236486017625425, + "grad_norm": 2.102454423904419, + "learning_rate": 0.00016266224940658815, + "loss": 0.7364, + "step": 17634 + }, + { + "epoch": 1.2237179931996391, + "grad_norm": 2.6727194786071777, + "learning_rate": 0.00016265787059173688, + "loss": 0.752, + "step": 17635 + }, + { + "epoch": 1.2237873846367358, + "grad_norm": 2.1323680877685547, + "learning_rate": 0.00016265349157908195, + "loss": 0.5009, + "step": 17636 + }, + { + "epoch": 1.2238567760738324, + "grad_norm": 1.8856558799743652, + "learning_rate": 0.0001626491123686373, + "loss": 0.6283, + "step": 17637 + }, + { + "epoch": 1.2239261675109292, + "grad_norm": 2.3828113079071045, + "learning_rate": 0.0001626447329604166, + "loss": 0.7249, + "step": 17638 + }, + { + "epoch": 1.2239955589480258, + "grad_norm": 2.272630453109741, + "learning_rate": 0.0001626403533544338, + "loss": 0.6793, + "step": 17639 + }, + { + "epoch": 1.2240649503851224, + "grad_norm": 2.287656545639038, + "learning_rate": 0.00016263597355070262, + "loss": 0.7942, + "step": 17640 + }, + { + "epoch": 1.2241343418222193, + "grad_norm": 1.5338239669799805, + "learning_rate": 0.00016263159354923696, + "loss": 0.5951, + "step": 17641 + }, + { + "epoch": 1.2242037332593159, + "grad_norm": 2.043283224105835, + "learning_rate": 0.00016262721335005064, + "loss": 0.6882, + "step": 17642 + }, + { + "epoch": 1.2242731246964125, + "grad_norm": 2.209110975265503, + "learning_rate": 0.00016262283295315745, + "loss": 0.5546, + "step": 17643 + }, + { + "epoch": 1.224342516133509, + "grad_norm": 1.725251317024231, + "learning_rate": 0.00016261845235857126, + "loss": 0.8445, + "step": 17644 + }, + { + "epoch": 1.2244119075706057, + "grad_norm": 1.6566123962402344, + "learning_rate": 0.00016261407156630589, + "loss": 0.6353, + "step": 17645 + }, + { + "epoch": 1.2244812990077025, + "grad_norm": 1.6356031894683838, + "learning_rate": 0.00016260969057637514, + "loss": 0.5872, + "step": 17646 + }, + { + "epoch": 1.2245506904447991, + "grad_norm": 2.2213051319122314, + "learning_rate": 0.00016260530938879287, + "loss": 0.6628, + "step": 17647 + }, + { + "epoch": 1.2246200818818958, + "grad_norm": 2.254262924194336, + "learning_rate": 0.00016260092800357288, + "loss": 0.6839, + "step": 17648 + }, + { + "epoch": 1.2246894733189924, + "grad_norm": 2.264181613922119, + "learning_rate": 0.00016259654642072904, + "loss": 0.6193, + "step": 17649 + }, + { + "epoch": 1.2247588647560892, + "grad_norm": 1.9741231203079224, + "learning_rate": 0.00016259216464027519, + "loss": 0.7662, + "step": 17650 + }, + { + "epoch": 1.2248282561931858, + "grad_norm": 1.6694822311401367, + "learning_rate": 0.0001625877826622251, + "loss": 0.4578, + "step": 17651 + }, + { + "epoch": 1.2248976476302824, + "grad_norm": 2.008899450302124, + "learning_rate": 0.00016258340048659267, + "loss": 0.6793, + "step": 17652 + }, + { + "epoch": 1.224967039067379, + "grad_norm": 1.8598694801330566, + "learning_rate": 0.0001625790181133917, + "loss": 0.6869, + "step": 17653 + }, + { + "epoch": 1.2250364305044759, + "grad_norm": 1.6664069890975952, + "learning_rate": 0.00016257463554263598, + "loss": 0.6123, + "step": 17654 + }, + { + "epoch": 1.2251058219415725, + "grad_norm": 1.6380964517593384, + "learning_rate": 0.00016257025277433947, + "loss": 0.563, + "step": 17655 + }, + { + "epoch": 1.225175213378669, + "grad_norm": 2.2776410579681396, + "learning_rate": 0.00016256586980851592, + "loss": 0.9999, + "step": 17656 + }, + { + "epoch": 1.2252446048157657, + "grad_norm": 1.7457695007324219, + "learning_rate": 0.00016256148664517913, + "loss": 0.7915, + "step": 17657 + }, + { + "epoch": 1.2253139962528623, + "grad_norm": 1.9162516593933105, + "learning_rate": 0.00016255710328434301, + "loss": 0.4962, + "step": 17658 + }, + { + "epoch": 1.2253833876899591, + "grad_norm": 1.8785189390182495, + "learning_rate": 0.0001625527197260214, + "loss": 0.5803, + "step": 17659 + }, + { + "epoch": 1.2254527791270557, + "grad_norm": 1.6514689922332764, + "learning_rate": 0.00016254833597022808, + "loss": 0.6065, + "step": 17660 + }, + { + "epoch": 1.2255221705641524, + "grad_norm": 2.1851017475128174, + "learning_rate": 0.0001625439520169769, + "loss": 1.0054, + "step": 17661 + }, + { + "epoch": 1.225591562001249, + "grad_norm": 2.0097198486328125, + "learning_rate": 0.00016253956786628179, + "loss": 0.8636, + "step": 17662 + }, + { + "epoch": 1.2256609534383458, + "grad_norm": 1.5614898204803467, + "learning_rate": 0.00016253518351815646, + "loss": 0.6421, + "step": 17663 + }, + { + "epoch": 1.2257303448754424, + "grad_norm": 1.2571914196014404, + "learning_rate": 0.00016253079897261482, + "loss": 0.4873, + "step": 17664 + }, + { + "epoch": 1.225799736312539, + "grad_norm": 2.501939296722412, + "learning_rate": 0.00016252641422967071, + "loss": 0.8092, + "step": 17665 + }, + { + "epoch": 1.2258691277496356, + "grad_norm": 1.5197901725769043, + "learning_rate": 0.00016252202928933798, + "loss": 0.5746, + "step": 17666 + }, + { + "epoch": 1.2259385191867325, + "grad_norm": 1.9821189641952515, + "learning_rate": 0.00016251764415163044, + "loss": 0.8774, + "step": 17667 + }, + { + "epoch": 1.226007910623829, + "grad_norm": 2.007049798965454, + "learning_rate": 0.00016251325881656197, + "loss": 0.8182, + "step": 17668 + }, + { + "epoch": 1.2260773020609257, + "grad_norm": 1.3559833765029907, + "learning_rate": 0.00016250887328414634, + "loss": 0.5502, + "step": 17669 + }, + { + "epoch": 1.2261466934980223, + "grad_norm": 1.403275966644287, + "learning_rate": 0.00016250448755439752, + "loss": 0.4616, + "step": 17670 + }, + { + "epoch": 1.226216084935119, + "grad_norm": 2.196929693222046, + "learning_rate": 0.00016250010162732924, + "loss": 0.6917, + "step": 17671 + }, + { + "epoch": 1.2262854763722157, + "grad_norm": 1.5942610502243042, + "learning_rate": 0.00016249571550295538, + "loss": 0.6388, + "step": 17672 + }, + { + "epoch": 1.2263548678093124, + "grad_norm": 1.9886950254440308, + "learning_rate": 0.00016249132918128985, + "loss": 0.4972, + "step": 17673 + }, + { + "epoch": 1.226424259246409, + "grad_norm": 2.2190418243408203, + "learning_rate": 0.00016248694266234642, + "loss": 0.5522, + "step": 17674 + }, + { + "epoch": 1.2264936506835056, + "grad_norm": 2.3550658226013184, + "learning_rate": 0.00016248255594613896, + "loss": 0.7138, + "step": 17675 + }, + { + "epoch": 1.2265630421206024, + "grad_norm": 2.077040195465088, + "learning_rate": 0.00016247816903268132, + "loss": 0.8177, + "step": 17676 + }, + { + "epoch": 1.226632433557699, + "grad_norm": 2.1206607818603516, + "learning_rate": 0.00016247378192198734, + "loss": 0.6351, + "step": 17677 + }, + { + "epoch": 1.2267018249947956, + "grad_norm": 2.208320379257202, + "learning_rate": 0.00016246939461407088, + "loss": 0.968, + "step": 17678 + }, + { + "epoch": 1.2267712164318922, + "grad_norm": 1.4212900400161743, + "learning_rate": 0.0001624650071089458, + "loss": 0.6946, + "step": 17679 + }, + { + "epoch": 1.226840607868989, + "grad_norm": 1.7441297769546509, + "learning_rate": 0.00016246061940662594, + "loss": 0.7766, + "step": 17680 + }, + { + "epoch": 1.2269099993060857, + "grad_norm": 1.1475721597671509, + "learning_rate": 0.00016245623150712513, + "loss": 0.4572, + "step": 17681 + }, + { + "epoch": 1.2269793907431823, + "grad_norm": 1.6534109115600586, + "learning_rate": 0.00016245184341045729, + "loss": 0.6848, + "step": 17682 + }, + { + "epoch": 1.227048782180279, + "grad_norm": 1.651078701019287, + "learning_rate": 0.00016244745511663616, + "loss": 0.6096, + "step": 17683 + }, + { + "epoch": 1.2271181736173755, + "grad_norm": 1.9152945280075073, + "learning_rate": 0.00016244306662567572, + "loss": 0.6675, + "step": 17684 + }, + { + "epoch": 1.2271875650544724, + "grad_norm": 1.5854477882385254, + "learning_rate": 0.0001624386779375897, + "loss": 0.9187, + "step": 17685 + }, + { + "epoch": 1.227256956491569, + "grad_norm": 1.6065949201583862, + "learning_rate": 0.00016243428905239209, + "loss": 0.5868, + "step": 17686 + }, + { + "epoch": 1.2273263479286656, + "grad_norm": 1.4108458757400513, + "learning_rate": 0.00016242989997009664, + "loss": 0.5241, + "step": 17687 + }, + { + "epoch": 1.2273957393657622, + "grad_norm": 1.4720041751861572, + "learning_rate": 0.00016242551069071723, + "loss": 0.6172, + "step": 17688 + }, + { + "epoch": 1.227465130802859, + "grad_norm": 1.8769917488098145, + "learning_rate": 0.0001624211212142677, + "loss": 0.5846, + "step": 17689 + }, + { + "epoch": 1.2275345222399556, + "grad_norm": 2.02758526802063, + "learning_rate": 0.000162416731540762, + "loss": 0.6259, + "step": 17690 + }, + { + "epoch": 1.2276039136770522, + "grad_norm": 1.742573857307434, + "learning_rate": 0.0001624123416702139, + "loss": 0.8066, + "step": 17691 + }, + { + "epoch": 1.2276733051141488, + "grad_norm": 1.7513587474822998, + "learning_rate": 0.00016240795160263723, + "loss": 0.5521, + "step": 17692 + }, + { + "epoch": 1.2277426965512457, + "grad_norm": 2.043896198272705, + "learning_rate": 0.00016240356133804596, + "loss": 0.7766, + "step": 17693 + }, + { + "epoch": 1.2278120879883423, + "grad_norm": 2.145179271697998, + "learning_rate": 0.00016239917087645383, + "loss": 0.4821, + "step": 17694 + }, + { + "epoch": 1.227881479425439, + "grad_norm": 1.3912047147750854, + "learning_rate": 0.00016239478021787482, + "loss": 0.5592, + "step": 17695 + }, + { + "epoch": 1.2279508708625355, + "grad_norm": 1.73405921459198, + "learning_rate": 0.00016239038936232268, + "loss": 0.7622, + "step": 17696 + }, + { + "epoch": 1.2280202622996321, + "grad_norm": 1.4403492212295532, + "learning_rate": 0.00016238599830981133, + "loss": 0.4077, + "step": 17697 + }, + { + "epoch": 1.228089653736729, + "grad_norm": 1.9768000841140747, + "learning_rate": 0.00016238160706035464, + "loss": 0.8781, + "step": 17698 + }, + { + "epoch": 1.2281590451738256, + "grad_norm": 1.5865917205810547, + "learning_rate": 0.0001623772156139664, + "loss": 0.6772, + "step": 17699 + }, + { + "epoch": 1.2282284366109222, + "grad_norm": 1.7011874914169312, + "learning_rate": 0.0001623728239706606, + "loss": 0.7181, + "step": 17700 + }, + { + "epoch": 1.2282978280480188, + "grad_norm": 1.9124668836593628, + "learning_rate": 0.000162368432130451, + "loss": 0.7321, + "step": 17701 + }, + { + "epoch": 1.2283672194851156, + "grad_norm": 2.114811420440674, + "learning_rate": 0.00016236404009335148, + "loss": 0.7458, + "step": 17702 + }, + { + "epoch": 1.2284366109222122, + "grad_norm": 1.9803426265716553, + "learning_rate": 0.00016235964785937595, + "loss": 0.6783, + "step": 17703 + }, + { + "epoch": 1.2285060023593088, + "grad_norm": 1.6899864673614502, + "learning_rate": 0.00016235525542853822, + "loss": 0.6078, + "step": 17704 + }, + { + "epoch": 1.2285753937964055, + "grad_norm": 1.1882919073104858, + "learning_rate": 0.00016235086280085218, + "loss": 0.3067, + "step": 17705 + }, + { + "epoch": 1.2286447852335023, + "grad_norm": 1.6949132680892944, + "learning_rate": 0.00016234646997633173, + "loss": 0.7413, + "step": 17706 + }, + { + "epoch": 1.228714176670599, + "grad_norm": 2.129438638687134, + "learning_rate": 0.0001623420769549907, + "loss": 0.7129, + "step": 17707 + }, + { + "epoch": 1.2287835681076955, + "grad_norm": 1.6398845911026, + "learning_rate": 0.00016233768373684294, + "loss": 0.675, + "step": 17708 + }, + { + "epoch": 1.2288529595447921, + "grad_norm": 2.2237963676452637, + "learning_rate": 0.00016233329032190236, + "loss": 0.6673, + "step": 17709 + }, + { + "epoch": 1.2289223509818887, + "grad_norm": 2.402001142501831, + "learning_rate": 0.00016232889671018283, + "loss": 0.8514, + "step": 17710 + }, + { + "epoch": 1.2289917424189856, + "grad_norm": 2.100785255432129, + "learning_rate": 0.00016232450290169822, + "loss": 0.6668, + "step": 17711 + }, + { + "epoch": 1.2290611338560822, + "grad_norm": 1.1642475128173828, + "learning_rate": 0.00016232010889646235, + "loss": 0.4625, + "step": 17712 + }, + { + "epoch": 1.2291305252931788, + "grad_norm": 1.5470566749572754, + "learning_rate": 0.00016231571469448912, + "loss": 0.5682, + "step": 17713 + }, + { + "epoch": 1.2291999167302754, + "grad_norm": 2.2488765716552734, + "learning_rate": 0.0001623113202957924, + "loss": 0.9747, + "step": 17714 + }, + { + "epoch": 1.2292693081673722, + "grad_norm": 1.9257481098175049, + "learning_rate": 0.0001623069257003861, + "loss": 0.5958, + "step": 17715 + }, + { + "epoch": 1.2293386996044688, + "grad_norm": 1.4547513723373413, + "learning_rate": 0.00016230253090828407, + "loss": 0.5732, + "step": 17716 + }, + { + "epoch": 1.2294080910415655, + "grad_norm": 2.4291985034942627, + "learning_rate": 0.00016229813591950017, + "loss": 0.6506, + "step": 17717 + }, + { + "epoch": 1.229477482478662, + "grad_norm": 1.5277349948883057, + "learning_rate": 0.00016229374073404828, + "loss": 0.5586, + "step": 17718 + }, + { + "epoch": 1.229546873915759, + "grad_norm": 1.6855539083480835, + "learning_rate": 0.00016228934535194228, + "loss": 0.6652, + "step": 17719 + }, + { + "epoch": 1.2296162653528555, + "grad_norm": 1.5761359930038452, + "learning_rate": 0.00016228494977319605, + "loss": 0.6777, + "step": 17720 + }, + { + "epoch": 1.2296856567899521, + "grad_norm": 1.7664945125579834, + "learning_rate": 0.00016228055399782347, + "loss": 0.7434, + "step": 17721 + }, + { + "epoch": 1.2297550482270487, + "grad_norm": 1.7398511171340942, + "learning_rate": 0.0001622761580258384, + "loss": 0.5461, + "step": 17722 + }, + { + "epoch": 1.2298244396641453, + "grad_norm": 1.7188947200775146, + "learning_rate": 0.00016227176185725468, + "loss": 0.6104, + "step": 17723 + }, + { + "epoch": 1.2298938311012422, + "grad_norm": 1.9271351099014282, + "learning_rate": 0.00016226736549208626, + "loss": 0.5424, + "step": 17724 + }, + { + "epoch": 1.2299632225383388, + "grad_norm": 2.2697625160217285, + "learning_rate": 0.000162262968930347, + "loss": 0.6081, + "step": 17725 + }, + { + "epoch": 1.2300326139754354, + "grad_norm": 1.9752025604248047, + "learning_rate": 0.0001622585721720508, + "loss": 0.8222, + "step": 17726 + }, + { + "epoch": 1.230102005412532, + "grad_norm": 1.7859182357788086, + "learning_rate": 0.00016225417521721145, + "loss": 0.7531, + "step": 17727 + }, + { + "epoch": 1.2301713968496288, + "grad_norm": 1.852062702178955, + "learning_rate": 0.00016224977806584293, + "loss": 0.671, + "step": 17728 + }, + { + "epoch": 1.2302407882867255, + "grad_norm": 2.494313955307007, + "learning_rate": 0.00016224538071795907, + "loss": 0.7675, + "step": 17729 + }, + { + "epoch": 1.230310179723822, + "grad_norm": 2.130610466003418, + "learning_rate": 0.00016224098317357374, + "loss": 0.7708, + "step": 17730 + }, + { + "epoch": 1.2303795711609187, + "grad_norm": 1.4458774328231812, + "learning_rate": 0.00016223658543270092, + "loss": 0.6375, + "step": 17731 + }, + { + "epoch": 1.2304489625980155, + "grad_norm": 1.8697566986083984, + "learning_rate": 0.00016223218749535437, + "loss": 0.5605, + "step": 17732 + }, + { + "epoch": 1.2305183540351121, + "grad_norm": 1.7588670253753662, + "learning_rate": 0.000162227789361548, + "loss": 0.4863, + "step": 17733 + }, + { + "epoch": 1.2305877454722087, + "grad_norm": 2.6475987434387207, + "learning_rate": 0.00016222339103129574, + "loss": 0.7013, + "step": 17734 + }, + { + "epoch": 1.2306571369093053, + "grad_norm": 1.3134753704071045, + "learning_rate": 0.0001622189925046115, + "loss": 0.5746, + "step": 17735 + }, + { + "epoch": 1.230726528346402, + "grad_norm": 2.0525736808776855, + "learning_rate": 0.00016221459378150909, + "loss": 0.7301, + "step": 17736 + }, + { + "epoch": 1.2307959197834988, + "grad_norm": 1.289599895477295, + "learning_rate": 0.0001622101948620024, + "loss": 0.6175, + "step": 17737 + }, + { + "epoch": 1.2308653112205954, + "grad_norm": 2.446413993835449, + "learning_rate": 0.00016220579574610536, + "loss": 0.6971, + "step": 17738 + }, + { + "epoch": 1.230934702657692, + "grad_norm": 1.7773451805114746, + "learning_rate": 0.00016220139643383186, + "loss": 0.6037, + "step": 17739 + }, + { + "epoch": 1.2310040940947886, + "grad_norm": 1.5042121410369873, + "learning_rate": 0.00016219699692519575, + "loss": 0.5864, + "step": 17740 + }, + { + "epoch": 1.2310734855318854, + "grad_norm": 1.7547883987426758, + "learning_rate": 0.00016219259722021092, + "loss": 0.7274, + "step": 17741 + }, + { + "epoch": 1.231142876968982, + "grad_norm": 1.5335330963134766, + "learning_rate": 0.0001621881973188913, + "loss": 0.646, + "step": 17742 + }, + { + "epoch": 1.2312122684060787, + "grad_norm": 1.4946376085281372, + "learning_rate": 0.00016218379722125075, + "loss": 0.6529, + "step": 17743 + }, + { + "epoch": 1.2312816598431753, + "grad_norm": 1.6263272762298584, + "learning_rate": 0.00016217939692730315, + "loss": 0.5739, + "step": 17744 + }, + { + "epoch": 1.2313510512802721, + "grad_norm": 2.757293224334717, + "learning_rate": 0.00016217499643706243, + "loss": 0.9572, + "step": 17745 + }, + { + "epoch": 1.2314204427173687, + "grad_norm": 1.49061119556427, + "learning_rate": 0.00016217059575054249, + "loss": 0.4862, + "step": 17746 + }, + { + "epoch": 1.2314898341544653, + "grad_norm": 1.8642151355743408, + "learning_rate": 0.00016216619486775712, + "loss": 0.6786, + "step": 17747 + }, + { + "epoch": 1.231559225591562, + "grad_norm": 1.7297677993774414, + "learning_rate": 0.00016216179378872032, + "loss": 0.6135, + "step": 17748 + }, + { + "epoch": 1.2316286170286586, + "grad_norm": 1.3301852941513062, + "learning_rate": 0.00016215739251344594, + "loss": 0.581, + "step": 17749 + }, + { + "epoch": 1.2316980084657554, + "grad_norm": 1.613600730895996, + "learning_rate": 0.00016215299104194793, + "loss": 0.5139, + "step": 17750 + }, + { + "epoch": 1.231767399902852, + "grad_norm": 1.6331547498703003, + "learning_rate": 0.00016214858937424007, + "loss": 0.4543, + "step": 17751 + }, + { + "epoch": 1.2318367913399486, + "grad_norm": 1.6863019466400146, + "learning_rate": 0.00016214418751033636, + "loss": 0.7034, + "step": 17752 + }, + { + "epoch": 1.2319061827770452, + "grad_norm": 2.005714178085327, + "learning_rate": 0.00016213978545025066, + "loss": 0.6971, + "step": 17753 + }, + { + "epoch": 1.231975574214142, + "grad_norm": 1.8137143850326538, + "learning_rate": 0.00016213538319399687, + "loss": 0.8561, + "step": 17754 + }, + { + "epoch": 1.2320449656512387, + "grad_norm": 2.0105700492858887, + "learning_rate": 0.00016213098074158885, + "loss": 0.6559, + "step": 17755 + }, + { + "epoch": 1.2321143570883353, + "grad_norm": 1.793421745300293, + "learning_rate": 0.0001621265780930406, + "loss": 0.6087, + "step": 17756 + }, + { + "epoch": 1.2321837485254319, + "grad_norm": 2.6089839935302734, + "learning_rate": 0.00016212217524836587, + "loss": 0.961, + "step": 17757 + }, + { + "epoch": 1.2322531399625287, + "grad_norm": 2.2635860443115234, + "learning_rate": 0.0001621177722075787, + "loss": 0.8571, + "step": 17758 + }, + { + "epoch": 1.2323225313996253, + "grad_norm": 1.423532247543335, + "learning_rate": 0.00016211336897069295, + "loss": 0.579, + "step": 17759 + }, + { + "epoch": 1.232391922836722, + "grad_norm": 1.663672685623169, + "learning_rate": 0.00016210896553772242, + "loss": 0.5533, + "step": 17760 + }, + { + "epoch": 1.2324613142738186, + "grad_norm": 2.1689157485961914, + "learning_rate": 0.00016210456190868112, + "loss": 0.6377, + "step": 17761 + }, + { + "epoch": 1.2325307057109152, + "grad_norm": 2.2581851482391357, + "learning_rate": 0.00016210015808358294, + "loss": 0.7731, + "step": 17762 + }, + { + "epoch": 1.232600097148012, + "grad_norm": 1.5617719888687134, + "learning_rate": 0.00016209575406244179, + "loss": 0.4907, + "step": 17763 + }, + { + "epoch": 1.2326694885851086, + "grad_norm": 2.1604347229003906, + "learning_rate": 0.0001620913498452715, + "loss": 0.6135, + "step": 17764 + }, + { + "epoch": 1.2327388800222052, + "grad_norm": 2.1330606937408447, + "learning_rate": 0.00016208694543208603, + "loss": 0.8797, + "step": 17765 + }, + { + "epoch": 1.2328082714593018, + "grad_norm": 1.4888808727264404, + "learning_rate": 0.00016208254082289932, + "loss": 0.4363, + "step": 17766 + }, + { + "epoch": 1.2328776628963987, + "grad_norm": 1.8220916986465454, + "learning_rate": 0.0001620781360177252, + "loss": 0.7132, + "step": 17767 + }, + { + "epoch": 1.2329470543334953, + "grad_norm": 1.7216663360595703, + "learning_rate": 0.00016207373101657758, + "loss": 0.6635, + "step": 17768 + }, + { + "epoch": 1.2330164457705919, + "grad_norm": 1.8902578353881836, + "learning_rate": 0.00016206932581947045, + "loss": 0.7394, + "step": 17769 + }, + { + "epoch": 1.2330858372076885, + "grad_norm": 1.333778738975525, + "learning_rate": 0.00016206492042641763, + "loss": 0.512, + "step": 17770 + }, + { + "epoch": 1.2331552286447853, + "grad_norm": 2.4042470455169678, + "learning_rate": 0.00016206051483743305, + "loss": 0.638, + "step": 17771 + }, + { + "epoch": 1.233224620081882, + "grad_norm": 1.8674342632293701, + "learning_rate": 0.00016205610905253065, + "loss": 0.7078, + "step": 17772 + }, + { + "epoch": 1.2332940115189786, + "grad_norm": 2.336679458618164, + "learning_rate": 0.0001620517030717243, + "loss": 0.8898, + "step": 17773 + }, + { + "epoch": 1.2333634029560752, + "grad_norm": 2.3122458457946777, + "learning_rate": 0.00016204729689502792, + "loss": 0.8769, + "step": 17774 + }, + { + "epoch": 1.2334327943931718, + "grad_norm": 3.0387532711029053, + "learning_rate": 0.0001620428905224554, + "loss": 0.869, + "step": 17775 + }, + { + "epoch": 1.2335021858302686, + "grad_norm": 2.05424165725708, + "learning_rate": 0.0001620384839540207, + "loss": 0.8515, + "step": 17776 + }, + { + "epoch": 1.2335715772673652, + "grad_norm": 1.8007253408432007, + "learning_rate": 0.00016203407718973773, + "loss": 0.4852, + "step": 17777 + }, + { + "epoch": 1.2336409687044618, + "grad_norm": 1.8245218992233276, + "learning_rate": 0.00016202967022962031, + "loss": 0.5621, + "step": 17778 + }, + { + "epoch": 1.2337103601415584, + "grad_norm": 1.828214168548584, + "learning_rate": 0.0001620252630736825, + "loss": 0.5735, + "step": 17779 + }, + { + "epoch": 1.2337797515786553, + "grad_norm": 1.5256584882736206, + "learning_rate": 0.00016202085572193805, + "loss": 0.674, + "step": 17780 + }, + { + "epoch": 1.2338491430157519, + "grad_norm": 1.8004612922668457, + "learning_rate": 0.000162016448174401, + "loss": 0.6853, + "step": 17781 + }, + { + "epoch": 1.2339185344528485, + "grad_norm": 1.653760313987732, + "learning_rate": 0.0001620120404310852, + "loss": 0.6036, + "step": 17782 + }, + { + "epoch": 1.233987925889945, + "grad_norm": 2.053457021713257, + "learning_rate": 0.00016200763249200458, + "loss": 0.975, + "step": 17783 + }, + { + "epoch": 1.234057317327042, + "grad_norm": 1.8250364065170288, + "learning_rate": 0.0001620032243571731, + "loss": 0.4587, + "step": 17784 + }, + { + "epoch": 1.2341267087641385, + "grad_norm": 2.4666197299957275, + "learning_rate": 0.00016199881602660458, + "loss": 0.7129, + "step": 17785 + }, + { + "epoch": 1.2341961002012352, + "grad_norm": 2.150702476501465, + "learning_rate": 0.000161994407500313, + "loss": 0.7397, + "step": 17786 + }, + { + "epoch": 1.2342654916383318, + "grad_norm": 1.544040560722351, + "learning_rate": 0.0001619899987783123, + "loss": 0.5417, + "step": 17787 + }, + { + "epoch": 1.2343348830754284, + "grad_norm": 2.2524585723876953, + "learning_rate": 0.00016198558986061637, + "loss": 0.7249, + "step": 17788 + }, + { + "epoch": 1.2344042745125252, + "grad_norm": 2.387263774871826, + "learning_rate": 0.00016198118074723907, + "loss": 0.8049, + "step": 17789 + }, + { + "epoch": 1.2344736659496218, + "grad_norm": 1.4871829748153687, + "learning_rate": 0.00016197677143819442, + "loss": 0.5726, + "step": 17790 + }, + { + "epoch": 1.2345430573867184, + "grad_norm": 1.7782649993896484, + "learning_rate": 0.0001619723619334963, + "loss": 0.5842, + "step": 17791 + }, + { + "epoch": 1.234612448823815, + "grad_norm": 1.7018994092941284, + "learning_rate": 0.00016196795223315857, + "loss": 0.6467, + "step": 17792 + }, + { + "epoch": 1.2346818402609119, + "grad_norm": 3.0170607566833496, + "learning_rate": 0.00016196354233719525, + "loss": 0.6532, + "step": 17793 + }, + { + "epoch": 1.2347512316980085, + "grad_norm": 1.74888277053833, + "learning_rate": 0.0001619591322456202, + "loss": 0.6906, + "step": 17794 + }, + { + "epoch": 1.234820623135105, + "grad_norm": 1.974989652633667, + "learning_rate": 0.00016195472195844736, + "loss": 0.7198, + "step": 17795 + }, + { + "epoch": 1.2348900145722017, + "grad_norm": 1.9668692350387573, + "learning_rate": 0.00016195031147569065, + "loss": 0.5888, + "step": 17796 + }, + { + "epoch": 1.2349594060092985, + "grad_norm": 1.7547905445098877, + "learning_rate": 0.000161945900797364, + "loss": 0.6836, + "step": 17797 + }, + { + "epoch": 1.2350287974463952, + "grad_norm": 2.130974531173706, + "learning_rate": 0.00016194148992348131, + "loss": 0.5244, + "step": 17798 + }, + { + "epoch": 1.2350981888834918, + "grad_norm": 1.7962733507156372, + "learning_rate": 0.00016193707885405652, + "loss": 0.7373, + "step": 17799 + }, + { + "epoch": 1.2351675803205884, + "grad_norm": 2.289344310760498, + "learning_rate": 0.00016193266758910362, + "loss": 0.8844, + "step": 17800 + }, + { + "epoch": 1.235236971757685, + "grad_norm": 3.1471660137176514, + "learning_rate": 0.00016192825612863643, + "loss": 0.7987, + "step": 17801 + }, + { + "epoch": 1.2353063631947818, + "grad_norm": 1.5301086902618408, + "learning_rate": 0.00016192384447266892, + "loss": 0.7098, + "step": 17802 + }, + { + "epoch": 1.2353757546318784, + "grad_norm": 2.4853804111480713, + "learning_rate": 0.000161919432621215, + "loss": 0.9664, + "step": 17803 + }, + { + "epoch": 1.235445146068975, + "grad_norm": 1.4721229076385498, + "learning_rate": 0.00016191502057428864, + "loss": 0.6996, + "step": 17804 + }, + { + "epoch": 1.2355145375060719, + "grad_norm": 1.6672366857528687, + "learning_rate": 0.00016191060833190376, + "loss": 0.6423, + "step": 17805 + }, + { + "epoch": 1.2355839289431685, + "grad_norm": 1.8627803325653076, + "learning_rate": 0.00016190619589407423, + "loss": 0.6952, + "step": 17806 + }, + { + "epoch": 1.235653320380265, + "grad_norm": 1.8342770338058472, + "learning_rate": 0.00016190178326081403, + "loss": 0.7603, + "step": 17807 + }, + { + "epoch": 1.2357227118173617, + "grad_norm": 1.8281930685043335, + "learning_rate": 0.0001618973704321371, + "loss": 0.6603, + "step": 17808 + }, + { + "epoch": 1.2357921032544583, + "grad_norm": 2.2192819118499756, + "learning_rate": 0.00016189295740805734, + "loss": 0.7662, + "step": 17809 + }, + { + "epoch": 1.2358614946915552, + "grad_norm": 1.383331537246704, + "learning_rate": 0.0001618885441885887, + "loss": 0.604, + "step": 17810 + }, + { + "epoch": 1.2359308861286518, + "grad_norm": 1.5596073865890503, + "learning_rate": 0.0001618841307737451, + "loss": 0.555, + "step": 17811 + }, + { + "epoch": 1.2360002775657484, + "grad_norm": 2.066530227661133, + "learning_rate": 0.0001618797171635405, + "loss": 0.6839, + "step": 17812 + }, + { + "epoch": 1.236069669002845, + "grad_norm": 2.128031015396118, + "learning_rate": 0.00016187530335798879, + "loss": 0.7539, + "step": 17813 + }, + { + "epoch": 1.2361390604399416, + "grad_norm": 1.4791399240493774, + "learning_rate": 0.00016187088935710393, + "loss": 0.5516, + "step": 17814 + }, + { + "epoch": 1.2362084518770384, + "grad_norm": 1.6872472763061523, + "learning_rate": 0.00016186647516089984, + "loss": 0.6891, + "step": 17815 + }, + { + "epoch": 1.236277843314135, + "grad_norm": 1.814329743385315, + "learning_rate": 0.0001618620607693905, + "loss": 0.652, + "step": 17816 + }, + { + "epoch": 1.2363472347512316, + "grad_norm": 1.9834462404251099, + "learning_rate": 0.00016185764618258978, + "loss": 0.8113, + "step": 17817 + }, + { + "epoch": 1.2364166261883285, + "grad_norm": 2.284519672393799, + "learning_rate": 0.00016185323140051168, + "loss": 0.8453, + "step": 17818 + }, + { + "epoch": 1.236486017625425, + "grad_norm": 1.609513521194458, + "learning_rate": 0.00016184881642317006, + "loss": 0.4976, + "step": 17819 + }, + { + "epoch": 1.2365554090625217, + "grad_norm": 1.928536057472229, + "learning_rate": 0.00016184440125057892, + "loss": 0.6144, + "step": 17820 + }, + { + "epoch": 1.2366248004996183, + "grad_norm": 2.687032461166382, + "learning_rate": 0.0001618399858827522, + "loss": 0.8838, + "step": 17821 + }, + { + "epoch": 1.236694191936715, + "grad_norm": 2.6948060989379883, + "learning_rate": 0.00016183557031970383, + "loss": 0.7389, + "step": 17822 + }, + { + "epoch": 1.2367635833738118, + "grad_norm": 2.2486214637756348, + "learning_rate": 0.0001618311545614477, + "loss": 0.7241, + "step": 17823 + }, + { + "epoch": 1.2368329748109084, + "grad_norm": 2.1387178897857666, + "learning_rate": 0.0001618267386079978, + "loss": 0.9073, + "step": 17824 + }, + { + "epoch": 1.236902366248005, + "grad_norm": 2.1295955181121826, + "learning_rate": 0.00016182232245936807, + "loss": 0.7274, + "step": 17825 + }, + { + "epoch": 1.2369717576851016, + "grad_norm": 2.0286009311676025, + "learning_rate": 0.00016181790611557244, + "loss": 0.7673, + "step": 17826 + }, + { + "epoch": 1.2370411491221982, + "grad_norm": 1.754207968711853, + "learning_rate": 0.00016181348957662484, + "loss": 0.6095, + "step": 17827 + }, + { + "epoch": 1.237110540559295, + "grad_norm": 1.780413269996643, + "learning_rate": 0.00016180907284253922, + "loss": 0.7343, + "step": 17828 + }, + { + "epoch": 1.2371799319963916, + "grad_norm": 1.7242003679275513, + "learning_rate": 0.00016180465591332955, + "loss": 0.6625, + "step": 17829 + }, + { + "epoch": 1.2372493234334883, + "grad_norm": 1.6299597024917603, + "learning_rate": 0.00016180023878900973, + "loss": 0.8448, + "step": 17830 + }, + { + "epoch": 1.237318714870585, + "grad_norm": 1.7329331636428833, + "learning_rate": 0.00016179582146959374, + "loss": 0.65, + "step": 17831 + }, + { + "epoch": 1.2373881063076817, + "grad_norm": 1.9305365085601807, + "learning_rate": 0.00016179140395509553, + "loss": 0.7874, + "step": 17832 + }, + { + "epoch": 1.2374574977447783, + "grad_norm": 2.0582478046417236, + "learning_rate": 0.000161786986245529, + "loss": 0.6553, + "step": 17833 + }, + { + "epoch": 1.237526889181875, + "grad_norm": 2.1247243881225586, + "learning_rate": 0.00016178256834090813, + "loss": 0.7015, + "step": 17834 + }, + { + "epoch": 1.2375962806189715, + "grad_norm": 1.6513330936431885, + "learning_rate": 0.00016177815024124686, + "loss": 0.6521, + "step": 17835 + }, + { + "epoch": 1.2376656720560684, + "grad_norm": 1.4222511053085327, + "learning_rate": 0.00016177373194655915, + "loss": 0.5351, + "step": 17836 + }, + { + "epoch": 1.237735063493165, + "grad_norm": 2.147554636001587, + "learning_rate": 0.00016176931345685891, + "loss": 0.9201, + "step": 17837 + }, + { + "epoch": 1.2378044549302616, + "grad_norm": 1.9110280275344849, + "learning_rate": 0.00016176489477216012, + "loss": 0.769, + "step": 17838 + }, + { + "epoch": 1.2378738463673582, + "grad_norm": 1.6840624809265137, + "learning_rate": 0.00016176047589247673, + "loss": 0.7219, + "step": 17839 + }, + { + "epoch": 1.2379432378044548, + "grad_norm": 1.7002406120300293, + "learning_rate": 0.00016175605681782268, + "loss": 0.6781, + "step": 17840 + }, + { + "epoch": 1.2380126292415516, + "grad_norm": 1.4507453441619873, + "learning_rate": 0.00016175163754821194, + "loss": 0.4321, + "step": 17841 + }, + { + "epoch": 1.2380820206786483, + "grad_norm": 1.836357831954956, + "learning_rate": 0.00016174721808365843, + "loss": 0.7091, + "step": 17842 + }, + { + "epoch": 1.2381514121157449, + "grad_norm": 3.320143699645996, + "learning_rate": 0.00016174279842417612, + "loss": 0.7248, + "step": 17843 + }, + { + "epoch": 1.2382208035528417, + "grad_norm": 1.6304264068603516, + "learning_rate": 0.00016173837856977893, + "loss": 0.6634, + "step": 17844 + }, + { + "epoch": 1.2382901949899383, + "grad_norm": 1.294197678565979, + "learning_rate": 0.00016173395852048087, + "loss": 0.589, + "step": 17845 + }, + { + "epoch": 1.238359586427035, + "grad_norm": 2.9044013023376465, + "learning_rate": 0.00016172953827629587, + "loss": 0.8175, + "step": 17846 + }, + { + "epoch": 1.2384289778641315, + "grad_norm": 2.112806797027588, + "learning_rate": 0.00016172511783723787, + "loss": 0.4905, + "step": 17847 + }, + { + "epoch": 1.2384983693012281, + "grad_norm": 1.7598886489868164, + "learning_rate": 0.00016172069720332084, + "loss": 0.7775, + "step": 17848 + }, + { + "epoch": 1.238567760738325, + "grad_norm": 1.6980316638946533, + "learning_rate": 0.00016171627637455872, + "loss": 0.5576, + "step": 17849 + }, + { + "epoch": 1.2386371521754216, + "grad_norm": 1.3923420906066895, + "learning_rate": 0.00016171185535096552, + "loss": 0.5326, + "step": 17850 + }, + { + "epoch": 1.2387065436125182, + "grad_norm": 2.1404078006744385, + "learning_rate": 0.0001617074341325551, + "loss": 0.8963, + "step": 17851 + }, + { + "epoch": 1.2387759350496148, + "grad_norm": 1.8368144035339355, + "learning_rate": 0.00016170301271934147, + "loss": 0.6921, + "step": 17852 + }, + { + "epoch": 1.2388453264867114, + "grad_norm": 1.667625904083252, + "learning_rate": 0.00016169859111133864, + "loss": 0.6851, + "step": 17853 + }, + { + "epoch": 1.2389147179238083, + "grad_norm": 1.6261770725250244, + "learning_rate": 0.00016169416930856046, + "loss": 0.4885, + "step": 17854 + }, + { + "epoch": 1.2389841093609049, + "grad_norm": 1.8402421474456787, + "learning_rate": 0.00016168974731102094, + "loss": 0.7875, + "step": 17855 + }, + { + "epoch": 1.2390535007980015, + "grad_norm": 1.7559863328933716, + "learning_rate": 0.00016168532511873408, + "loss": 0.6421, + "step": 17856 + }, + { + "epoch": 1.2391228922350983, + "grad_norm": 1.623030662536621, + "learning_rate": 0.0001616809027317138, + "loss": 0.6353, + "step": 17857 + }, + { + "epoch": 1.239192283672195, + "grad_norm": 1.946666955947876, + "learning_rate": 0.00016167648014997406, + "loss": 0.5867, + "step": 17858 + }, + { + "epoch": 1.2392616751092915, + "grad_norm": 2.472705364227295, + "learning_rate": 0.00016167205737352883, + "loss": 0.7795, + "step": 17859 + }, + { + "epoch": 1.2393310665463881, + "grad_norm": 1.9215102195739746, + "learning_rate": 0.00016166763440239208, + "loss": 0.7344, + "step": 17860 + }, + { + "epoch": 1.2394004579834847, + "grad_norm": 1.9274024963378906, + "learning_rate": 0.00016166321123657775, + "loss": 0.6886, + "step": 17861 + }, + { + "epoch": 1.2394698494205816, + "grad_norm": 1.4432417154312134, + "learning_rate": 0.0001616587878760998, + "loss": 0.6177, + "step": 17862 + }, + { + "epoch": 1.2395392408576782, + "grad_norm": 1.6276133060455322, + "learning_rate": 0.00016165436432097224, + "loss": 0.6647, + "step": 17863 + }, + { + "epoch": 1.2396086322947748, + "grad_norm": 1.5382347106933594, + "learning_rate": 0.00016164994057120898, + "loss": 0.5726, + "step": 17864 + }, + { + "epoch": 1.2396780237318714, + "grad_norm": 1.5334817171096802, + "learning_rate": 0.00016164551662682403, + "loss": 0.6827, + "step": 17865 + }, + { + "epoch": 1.239747415168968, + "grad_norm": 1.7131266593933105, + "learning_rate": 0.00016164109248783132, + "loss": 0.5794, + "step": 17866 + }, + { + "epoch": 1.2398168066060649, + "grad_norm": 1.7804197072982788, + "learning_rate": 0.00016163666815424484, + "loss": 0.6491, + "step": 17867 + }, + { + "epoch": 1.2398861980431615, + "grad_norm": 1.709470510482788, + "learning_rate": 0.00016163224362607858, + "loss": 0.7304, + "step": 17868 + }, + { + "epoch": 1.239955589480258, + "grad_norm": 1.5990381240844727, + "learning_rate": 0.00016162781890334644, + "loss": 0.5926, + "step": 17869 + }, + { + "epoch": 1.240024980917355, + "grad_norm": 1.5776348114013672, + "learning_rate": 0.00016162339398606245, + "loss": 0.7294, + "step": 17870 + }, + { + "epoch": 1.2400943723544515, + "grad_norm": 1.8615361452102661, + "learning_rate": 0.00016161896887424053, + "loss": 0.5895, + "step": 17871 + }, + { + "epoch": 1.2401637637915481, + "grad_norm": 1.3824098110198975, + "learning_rate": 0.0001616145435678947, + "loss": 0.5218, + "step": 17872 + }, + { + "epoch": 1.2402331552286447, + "grad_norm": 2.1087493896484375, + "learning_rate": 0.00016161011806703889, + "loss": 0.9049, + "step": 17873 + }, + { + "epoch": 1.2403025466657414, + "grad_norm": 2.53690242767334, + "learning_rate": 0.0001616056923716871, + "loss": 0.6322, + "step": 17874 + }, + { + "epoch": 1.2403719381028382, + "grad_norm": 2.0858399868011475, + "learning_rate": 0.00016160126648185325, + "loss": 0.8755, + "step": 17875 + }, + { + "epoch": 1.2404413295399348, + "grad_norm": 1.9683698415756226, + "learning_rate": 0.00016159684039755137, + "loss": 0.622, + "step": 17876 + }, + { + "epoch": 1.2405107209770314, + "grad_norm": 1.4901729822158813, + "learning_rate": 0.00016159241411879544, + "loss": 0.5998, + "step": 17877 + }, + { + "epoch": 1.240580112414128, + "grad_norm": 1.6143156290054321, + "learning_rate": 0.00016158798764559937, + "loss": 0.5386, + "step": 17878 + }, + { + "epoch": 1.2406495038512249, + "grad_norm": 2.0441348552703857, + "learning_rate": 0.00016158356097797718, + "loss": 0.6743, + "step": 17879 + }, + { + "epoch": 1.2407188952883215, + "grad_norm": 2.127755880355835, + "learning_rate": 0.00016157913411594284, + "loss": 0.9452, + "step": 17880 + }, + { + "epoch": 1.240788286725418, + "grad_norm": 1.5263193845748901, + "learning_rate": 0.0001615747070595103, + "loss": 0.6103, + "step": 17881 + }, + { + "epoch": 1.2408576781625147, + "grad_norm": 1.7587687969207764, + "learning_rate": 0.00016157027980869354, + "loss": 0.5571, + "step": 17882 + }, + { + "epoch": 1.2409270695996115, + "grad_norm": 1.7026525735855103, + "learning_rate": 0.0001615658523635066, + "loss": 0.4653, + "step": 17883 + }, + { + "epoch": 1.2409964610367081, + "grad_norm": 1.640243649482727, + "learning_rate": 0.00016156142472396336, + "loss": 0.8191, + "step": 17884 + }, + { + "epoch": 1.2410658524738047, + "grad_norm": 1.8816102743148804, + "learning_rate": 0.00016155699689007785, + "loss": 0.545, + "step": 17885 + }, + { + "epoch": 1.2411352439109014, + "grad_norm": 2.3014745712280273, + "learning_rate": 0.00016155256886186408, + "loss": 0.7251, + "step": 17886 + }, + { + "epoch": 1.241204635347998, + "grad_norm": 2.049184560775757, + "learning_rate": 0.00016154814063933597, + "loss": 0.6842, + "step": 17887 + }, + { + "epoch": 1.2412740267850948, + "grad_norm": 1.8068872690200806, + "learning_rate": 0.0001615437122225075, + "loss": 0.5207, + "step": 17888 + }, + { + "epoch": 1.2413434182221914, + "grad_norm": 1.591424584388733, + "learning_rate": 0.0001615392836113927, + "loss": 0.5259, + "step": 17889 + }, + { + "epoch": 1.241412809659288, + "grad_norm": 2.1250407695770264, + "learning_rate": 0.0001615348548060055, + "loss": 0.8475, + "step": 17890 + }, + { + "epoch": 1.2414822010963846, + "grad_norm": 2.2923789024353027, + "learning_rate": 0.0001615304258063599, + "loss": 0.7145, + "step": 17891 + }, + { + "epoch": 1.2415515925334815, + "grad_norm": 2.282773017883301, + "learning_rate": 0.0001615259966124699, + "loss": 0.6983, + "step": 17892 + }, + { + "epoch": 1.241620983970578, + "grad_norm": 1.7547643184661865, + "learning_rate": 0.00016152156722434947, + "loss": 0.6093, + "step": 17893 + }, + { + "epoch": 1.2416903754076747, + "grad_norm": 1.9105639457702637, + "learning_rate": 0.0001615171376420126, + "loss": 0.6156, + "step": 17894 + }, + { + "epoch": 1.2417597668447713, + "grad_norm": 2.0588982105255127, + "learning_rate": 0.00016151270786547324, + "loss": 0.6587, + "step": 17895 + }, + { + "epoch": 1.2418291582818681, + "grad_norm": 2.342627763748169, + "learning_rate": 0.0001615082778947454, + "loss": 0.7315, + "step": 17896 + }, + { + "epoch": 1.2418985497189647, + "grad_norm": 3.1153016090393066, + "learning_rate": 0.00016150384772984307, + "loss": 0.7649, + "step": 17897 + }, + { + "epoch": 1.2419679411560614, + "grad_norm": 1.7301056385040283, + "learning_rate": 0.00016149941737078025, + "loss": 0.5256, + "step": 17898 + }, + { + "epoch": 1.242037332593158, + "grad_norm": 1.4503661394119263, + "learning_rate": 0.00016149498681757086, + "loss": 0.5851, + "step": 17899 + }, + { + "epoch": 1.2421067240302546, + "grad_norm": 1.9643598794937134, + "learning_rate": 0.00016149055607022895, + "loss": 0.7131, + "step": 17900 + }, + { + "epoch": 1.2421761154673514, + "grad_norm": 2.843794584274292, + "learning_rate": 0.00016148612512876853, + "loss": 0.7964, + "step": 17901 + }, + { + "epoch": 1.242245506904448, + "grad_norm": 1.615548014640808, + "learning_rate": 0.00016148169399320347, + "loss": 0.7113, + "step": 17902 + }, + { + "epoch": 1.2423148983415446, + "grad_norm": 1.8021023273468018, + "learning_rate": 0.0001614772626635479, + "loss": 0.6323, + "step": 17903 + }, + { + "epoch": 1.2423842897786412, + "grad_norm": 1.70900297164917, + "learning_rate": 0.0001614728311398157, + "loss": 0.8276, + "step": 17904 + }, + { + "epoch": 1.242453681215738, + "grad_norm": 1.1267873048782349, + "learning_rate": 0.00016146839942202094, + "loss": 0.4703, + "step": 17905 + }, + { + "epoch": 1.2425230726528347, + "grad_norm": 1.5793728828430176, + "learning_rate": 0.00016146396751017756, + "loss": 0.6426, + "step": 17906 + }, + { + "epoch": 1.2425924640899313, + "grad_norm": 1.6416118144989014, + "learning_rate": 0.00016145953540429954, + "loss": 0.7657, + "step": 17907 + }, + { + "epoch": 1.242661855527028, + "grad_norm": 1.225581407546997, + "learning_rate": 0.00016145510310440096, + "loss": 0.5455, + "step": 17908 + }, + { + "epoch": 1.2427312469641247, + "grad_norm": 1.4283623695373535, + "learning_rate": 0.00016145067061049568, + "loss": 0.5224, + "step": 17909 + }, + { + "epoch": 1.2428006384012213, + "grad_norm": 1.765574336051941, + "learning_rate": 0.00016144623792259781, + "loss": 0.7632, + "step": 17910 + }, + { + "epoch": 1.242870029838318, + "grad_norm": 1.8962653875350952, + "learning_rate": 0.00016144180504072128, + "loss": 0.4303, + "step": 17911 + }, + { + "epoch": 1.2429394212754146, + "grad_norm": 1.406056523323059, + "learning_rate": 0.0001614373719648801, + "loss": 0.4787, + "step": 17912 + }, + { + "epoch": 1.2430088127125112, + "grad_norm": 1.881645679473877, + "learning_rate": 0.00016143293869508828, + "loss": 0.5871, + "step": 17913 + }, + { + "epoch": 1.243078204149608, + "grad_norm": 2.079895496368408, + "learning_rate": 0.00016142850523135976, + "loss": 0.8462, + "step": 17914 + }, + { + "epoch": 1.2431475955867046, + "grad_norm": 1.9578169584274292, + "learning_rate": 0.00016142407157370862, + "loss": 0.7059, + "step": 17915 + }, + { + "epoch": 1.2432169870238012, + "grad_norm": 2.285397529602051, + "learning_rate": 0.0001614196377221488, + "loss": 0.6807, + "step": 17916 + }, + { + "epoch": 1.2432863784608978, + "grad_norm": 1.5021699666976929, + "learning_rate": 0.00016141520367669428, + "loss": 0.6328, + "step": 17917 + }, + { + "epoch": 1.2433557698979947, + "grad_norm": 1.900294542312622, + "learning_rate": 0.00016141076943735914, + "loss": 0.6636, + "step": 17918 + }, + { + "epoch": 1.2434251613350913, + "grad_norm": 1.5144094228744507, + "learning_rate": 0.00016140633500415726, + "loss": 0.6037, + "step": 17919 + }, + { + "epoch": 1.243494552772188, + "grad_norm": 2.4988820552825928, + "learning_rate": 0.00016140190037710274, + "loss": 0.9283, + "step": 17920 + }, + { + "epoch": 1.2435639442092845, + "grad_norm": 2.745241403579712, + "learning_rate": 0.00016139746555620955, + "loss": 0.9, + "step": 17921 + }, + { + "epoch": 1.2436333356463813, + "grad_norm": 1.8265984058380127, + "learning_rate": 0.0001613930305414917, + "loss": 0.8393, + "step": 17922 + }, + { + "epoch": 1.243702727083478, + "grad_norm": 2.135833501815796, + "learning_rate": 0.00016138859533296315, + "loss": 0.7186, + "step": 17923 + }, + { + "epoch": 1.2437721185205746, + "grad_norm": 1.4963947534561157, + "learning_rate": 0.00016138415993063792, + "loss": 0.6761, + "step": 17924 + }, + { + "epoch": 1.2438415099576712, + "grad_norm": 2.2712414264678955, + "learning_rate": 0.00016137972433453003, + "loss": 0.6468, + "step": 17925 + }, + { + "epoch": 1.2439109013947678, + "grad_norm": 2.533686637878418, + "learning_rate": 0.00016137528854465348, + "loss": 0.8443, + "step": 17926 + }, + { + "epoch": 1.2439802928318646, + "grad_norm": 1.8672666549682617, + "learning_rate": 0.00016137085256102225, + "loss": 0.553, + "step": 17927 + }, + { + "epoch": 1.2440496842689612, + "grad_norm": 1.4417221546173096, + "learning_rate": 0.00016136641638365038, + "loss": 0.5001, + "step": 17928 + }, + { + "epoch": 1.2441190757060578, + "grad_norm": 1.4920705556869507, + "learning_rate": 0.00016136198001255183, + "loss": 0.6166, + "step": 17929 + }, + { + "epoch": 1.2441884671431545, + "grad_norm": 1.5818909406661987, + "learning_rate": 0.00016135754344774066, + "loss": 0.6115, + "step": 17930 + }, + { + "epoch": 1.2442578585802513, + "grad_norm": 1.3393160104751587, + "learning_rate": 0.0001613531066892308, + "loss": 0.4141, + "step": 17931 + }, + { + "epoch": 1.244327250017348, + "grad_norm": 2.231574773788452, + "learning_rate": 0.00016134866973703632, + "loss": 0.7527, + "step": 17932 + }, + { + "epoch": 1.2443966414544445, + "grad_norm": 1.0059984922409058, + "learning_rate": 0.00016134423259117124, + "loss": 0.4335, + "step": 17933 + }, + { + "epoch": 1.2444660328915411, + "grad_norm": 1.8430707454681396, + "learning_rate": 0.00016133979525164948, + "loss": 0.5469, + "step": 17934 + }, + { + "epoch": 1.244535424328638, + "grad_norm": 2.008075475692749, + "learning_rate": 0.00016133535771848512, + "loss": 0.6481, + "step": 17935 + }, + { + "epoch": 1.2446048157657346, + "grad_norm": 3.782680034637451, + "learning_rate": 0.00016133091999169218, + "loss": 0.8051, + "step": 17936 + }, + { + "epoch": 1.2446742072028312, + "grad_norm": 1.7190359830856323, + "learning_rate": 0.00016132648207128463, + "loss": 0.5899, + "step": 17937 + }, + { + "epoch": 1.2447435986399278, + "grad_norm": 2.2168102264404297, + "learning_rate": 0.00016132204395727647, + "loss": 0.7509, + "step": 17938 + }, + { + "epoch": 1.2448129900770244, + "grad_norm": 2.1727702617645264, + "learning_rate": 0.00016131760564968176, + "loss": 0.6842, + "step": 17939 + }, + { + "epoch": 1.2448823815141212, + "grad_norm": 2.2165122032165527, + "learning_rate": 0.00016131316714851446, + "loss": 0.7052, + "step": 17940 + }, + { + "epoch": 1.2449517729512178, + "grad_norm": 1.910067081451416, + "learning_rate": 0.0001613087284537886, + "loss": 0.7191, + "step": 17941 + }, + { + "epoch": 1.2450211643883144, + "grad_norm": 1.7570985555648804, + "learning_rate": 0.0001613042895655182, + "loss": 0.9183, + "step": 17942 + }, + { + "epoch": 1.245090555825411, + "grad_norm": 1.4160122871398926, + "learning_rate": 0.0001612998504837173, + "loss": 0.5056, + "step": 17943 + }, + { + "epoch": 1.245159947262508, + "grad_norm": 1.6158326864242554, + "learning_rate": 0.00016129541120839987, + "loss": 0.4844, + "step": 17944 + }, + { + "epoch": 1.2452293386996045, + "grad_norm": 1.9831854104995728, + "learning_rate": 0.00016129097173957994, + "loss": 0.762, + "step": 17945 + }, + { + "epoch": 1.2452987301367011, + "grad_norm": 1.4636924266815186, + "learning_rate": 0.00016128653207727152, + "loss": 0.4392, + "step": 17946 + }, + { + "epoch": 1.2453681215737977, + "grad_norm": 1.652005910873413, + "learning_rate": 0.0001612820922214886, + "loss": 0.5542, + "step": 17947 + }, + { + "epoch": 1.2454375130108946, + "grad_norm": 2.0675485134124756, + "learning_rate": 0.00016127765217224525, + "loss": 0.9284, + "step": 17948 + }, + { + "epoch": 1.2455069044479912, + "grad_norm": 2.0001325607299805, + "learning_rate": 0.0001612732119295555, + "loss": 0.6512, + "step": 17949 + }, + { + "epoch": 1.2455762958850878, + "grad_norm": 1.95497465133667, + "learning_rate": 0.00016126877149343327, + "loss": 0.7286, + "step": 17950 + }, + { + "epoch": 1.2456456873221844, + "grad_norm": 1.7017754316329956, + "learning_rate": 0.00016126433086389265, + "loss": 0.4176, + "step": 17951 + }, + { + "epoch": 1.245715078759281, + "grad_norm": 1.6737340688705444, + "learning_rate": 0.00016125989004094765, + "loss": 0.5508, + "step": 17952 + }, + { + "epoch": 1.2457844701963778, + "grad_norm": 1.7619913816452026, + "learning_rate": 0.00016125544902461232, + "loss": 0.5727, + "step": 17953 + }, + { + "epoch": 1.2458538616334744, + "grad_norm": 1.6938480138778687, + "learning_rate": 0.0001612510078149006, + "loss": 0.6827, + "step": 17954 + }, + { + "epoch": 1.245923253070571, + "grad_norm": 1.4514683485031128, + "learning_rate": 0.00016124656641182655, + "loss": 0.531, + "step": 17955 + }, + { + "epoch": 1.2459926445076677, + "grad_norm": 1.7559293508529663, + "learning_rate": 0.0001612421248154042, + "loss": 0.6142, + "step": 17956 + }, + { + "epoch": 1.2460620359447645, + "grad_norm": 2.1545066833496094, + "learning_rate": 0.0001612376830256476, + "loss": 0.8611, + "step": 17957 + }, + { + "epoch": 1.2461314273818611, + "grad_norm": 1.5698599815368652, + "learning_rate": 0.00016123324104257072, + "loss": 0.5534, + "step": 17958 + }, + { + "epoch": 1.2462008188189577, + "grad_norm": 1.7618900537490845, + "learning_rate": 0.00016122879886618758, + "loss": 0.5721, + "step": 17959 + }, + { + "epoch": 1.2462702102560543, + "grad_norm": 2.003366231918335, + "learning_rate": 0.00016122435649651226, + "loss": 0.6293, + "step": 17960 + }, + { + "epoch": 1.2463396016931512, + "grad_norm": 1.752489686012268, + "learning_rate": 0.00016121991393355876, + "loss": 0.8056, + "step": 17961 + }, + { + "epoch": 1.2464089931302478, + "grad_norm": 4.710575103759766, + "learning_rate": 0.00016121547117734105, + "loss": 0.6935, + "step": 17962 + }, + { + "epoch": 1.2464783845673444, + "grad_norm": 1.8768104314804077, + "learning_rate": 0.00016121102822787324, + "loss": 0.5881, + "step": 17963 + }, + { + "epoch": 1.246547776004441, + "grad_norm": 1.4586398601531982, + "learning_rate": 0.00016120658508516928, + "loss": 0.4901, + "step": 17964 + }, + { + "epoch": 1.2466171674415376, + "grad_norm": 1.7424331903457642, + "learning_rate": 0.00016120214174924328, + "loss": 0.8092, + "step": 17965 + }, + { + "epoch": 1.2466865588786344, + "grad_norm": 1.5518978834152222, + "learning_rate": 0.00016119769822010914, + "loss": 0.5404, + "step": 17966 + }, + { + "epoch": 1.246755950315731, + "grad_norm": 1.4747140407562256, + "learning_rate": 0.00016119325449778104, + "loss": 0.6137, + "step": 17967 + }, + { + "epoch": 1.2468253417528277, + "grad_norm": 1.755519986152649, + "learning_rate": 0.00016118881058227292, + "loss": 0.5852, + "step": 17968 + }, + { + "epoch": 1.2468947331899243, + "grad_norm": 2.1472623348236084, + "learning_rate": 0.0001611843664735988, + "loss": 0.7517, + "step": 17969 + }, + { + "epoch": 1.246964124627021, + "grad_norm": 1.9849976301193237, + "learning_rate": 0.0001611799221717728, + "loss": 0.6159, + "step": 17970 + }, + { + "epoch": 1.2470335160641177, + "grad_norm": 2.372103691101074, + "learning_rate": 0.0001611754776768088, + "loss": 0.6901, + "step": 17971 + }, + { + "epoch": 1.2471029075012143, + "grad_norm": 1.9887940883636475, + "learning_rate": 0.00016117103298872096, + "loss": 0.8028, + "step": 17972 + }, + { + "epoch": 1.247172298938311, + "grad_norm": 1.9468185901641846, + "learning_rate": 0.00016116658810752323, + "loss": 0.7275, + "step": 17973 + }, + { + "epoch": 1.2472416903754078, + "grad_norm": 2.5668556690216064, + "learning_rate": 0.00016116214303322972, + "loss": 0.832, + "step": 17974 + }, + { + "epoch": 1.2473110818125044, + "grad_norm": 2.397124767303467, + "learning_rate": 0.0001611576977658544, + "loss": 0.6281, + "step": 17975 + }, + { + "epoch": 1.247380473249601, + "grad_norm": 1.988083839416504, + "learning_rate": 0.00016115325230541133, + "loss": 0.5723, + "step": 17976 + }, + { + "epoch": 1.2474498646866976, + "grad_norm": 1.6288591623306274, + "learning_rate": 0.00016114880665191454, + "loss": 0.7532, + "step": 17977 + }, + { + "epoch": 1.2475192561237942, + "grad_norm": 1.4962592124938965, + "learning_rate": 0.00016114436080537803, + "loss": 0.7444, + "step": 17978 + }, + { + "epoch": 1.247588647560891, + "grad_norm": 2.2239537239074707, + "learning_rate": 0.0001611399147658159, + "loss": 0.923, + "step": 17979 + }, + { + "epoch": 1.2476580389979877, + "grad_norm": 1.6069097518920898, + "learning_rate": 0.00016113546853324215, + "loss": 0.5722, + "step": 17980 + }, + { + "epoch": 1.2477274304350843, + "grad_norm": 1.310163974761963, + "learning_rate": 0.00016113102210767082, + "loss": 0.3801, + "step": 17981 + }, + { + "epoch": 1.2477968218721809, + "grad_norm": 1.8864238262176514, + "learning_rate": 0.00016112657548911593, + "loss": 0.9127, + "step": 17982 + }, + { + "epoch": 1.2478662133092777, + "grad_norm": 1.9457710981369019, + "learning_rate": 0.00016112212867759157, + "loss": 0.5211, + "step": 17983 + }, + { + "epoch": 1.2479356047463743, + "grad_norm": 2.0318007469177246, + "learning_rate": 0.00016111768167311168, + "loss": 0.6434, + "step": 17984 + }, + { + "epoch": 1.248004996183471, + "grad_norm": 2.810952663421631, + "learning_rate": 0.0001611132344756904, + "loss": 0.9073, + "step": 17985 + }, + { + "epoch": 1.2480743876205675, + "grad_norm": 2.026995897293091, + "learning_rate": 0.00016110878708534173, + "loss": 0.913, + "step": 17986 + }, + { + "epoch": 1.2481437790576644, + "grad_norm": 2.1901488304138184, + "learning_rate": 0.0001611043395020797, + "loss": 0.8233, + "step": 17987 + }, + { + "epoch": 1.248213170494761, + "grad_norm": 1.548417329788208, + "learning_rate": 0.00016109989172591836, + "loss": 0.496, + "step": 17988 + }, + { + "epoch": 1.2482825619318576, + "grad_norm": 1.659671664237976, + "learning_rate": 0.00016109544375687176, + "loss": 0.5929, + "step": 17989 + }, + { + "epoch": 1.2483519533689542, + "grad_norm": 1.7569464445114136, + "learning_rate": 0.0001610909955949539, + "loss": 0.4622, + "step": 17990 + }, + { + "epoch": 1.2484213448060508, + "grad_norm": 1.4433259963989258, + "learning_rate": 0.00016108654724017888, + "loss": 0.6583, + "step": 17991 + }, + { + "epoch": 1.2484907362431477, + "grad_norm": 1.8789052963256836, + "learning_rate": 0.00016108209869256074, + "loss": 0.686, + "step": 17992 + }, + { + "epoch": 1.2485601276802443, + "grad_norm": 1.4253177642822266, + "learning_rate": 0.00016107764995211348, + "loss": 0.5231, + "step": 17993 + }, + { + "epoch": 1.2486295191173409, + "grad_norm": 1.9558823108673096, + "learning_rate": 0.00016107320101885116, + "loss": 0.708, + "step": 17994 + }, + { + "epoch": 1.2486989105544375, + "grad_norm": 1.5898113250732422, + "learning_rate": 0.00016106875189278786, + "loss": 0.5782, + "step": 17995 + }, + { + "epoch": 1.2487683019915343, + "grad_norm": 1.620502233505249, + "learning_rate": 0.00016106430257393755, + "loss": 0.6599, + "step": 17996 + }, + { + "epoch": 1.248837693428631, + "grad_norm": 2.020599126815796, + "learning_rate": 0.00016105985306231433, + "loss": 0.7119, + "step": 17997 + }, + { + "epoch": 1.2489070848657275, + "grad_norm": 1.641701579093933, + "learning_rate": 0.00016105540335793224, + "loss": 0.8399, + "step": 17998 + }, + { + "epoch": 1.2489764763028242, + "grad_norm": 2.0367016792297363, + "learning_rate": 0.00016105095346080536, + "loss": 0.8779, + "step": 17999 + }, + { + "epoch": 1.249045867739921, + "grad_norm": 1.675668478012085, + "learning_rate": 0.00016104650337094766, + "loss": 0.6251, + "step": 18000 + }, + { + "epoch": 1.2491152591770176, + "grad_norm": 1.807768702507019, + "learning_rate": 0.0001610420530883733, + "loss": 0.6197, + "step": 18001 + }, + { + "epoch": 1.2491846506141142, + "grad_norm": 1.4054288864135742, + "learning_rate": 0.0001610376026130962, + "loss": 0.5391, + "step": 18002 + }, + { + "epoch": 1.2492540420512108, + "grad_norm": 1.949294090270996, + "learning_rate": 0.00016103315194513046, + "loss": 0.7494, + "step": 18003 + }, + { + "epoch": 1.2493234334883074, + "grad_norm": 2.4440016746520996, + "learning_rate": 0.00016102870108449015, + "loss": 0.692, + "step": 18004 + }, + { + "epoch": 1.2493928249254043, + "grad_norm": 1.6342238187789917, + "learning_rate": 0.00016102425003118934, + "loss": 0.6539, + "step": 18005 + }, + { + "epoch": 1.2494622163625009, + "grad_norm": 2.2226006984710693, + "learning_rate": 0.00016101979878524204, + "loss": 0.558, + "step": 18006 + }, + { + "epoch": 1.2495316077995975, + "grad_norm": 1.4228906631469727, + "learning_rate": 0.0001610153473466623, + "loss": 0.5178, + "step": 18007 + }, + { + "epoch": 1.249600999236694, + "grad_norm": 1.2875404357910156, + "learning_rate": 0.0001610108957154642, + "loss": 0.6076, + "step": 18008 + }, + { + "epoch": 1.249670390673791, + "grad_norm": 1.4422527551651, + "learning_rate": 0.0001610064438916618, + "loss": 0.745, + "step": 18009 + }, + { + "epoch": 1.2497397821108875, + "grad_norm": 1.3408235311508179, + "learning_rate": 0.00016100199187526906, + "loss": 0.5926, + "step": 18010 + }, + { + "epoch": 1.2498091735479842, + "grad_norm": 1.5995148420333862, + "learning_rate": 0.0001609975396663002, + "loss": 0.7182, + "step": 18011 + }, + { + "epoch": 1.2498785649850808, + "grad_norm": 1.7111661434173584, + "learning_rate": 0.00016099308726476915, + "loss": 0.6273, + "step": 18012 + }, + { + "epoch": 1.2499479564221776, + "grad_norm": 2.0736098289489746, + "learning_rate": 0.00016098863467068998, + "loss": 0.7207, + "step": 18013 + }, + { + "epoch": 1.2500173478592742, + "grad_norm": 1.6934435367584229, + "learning_rate": 0.00016098418188407679, + "loss": 0.5569, + "step": 18014 + }, + { + "epoch": 1.2500867392963708, + "grad_norm": 2.199983835220337, + "learning_rate": 0.0001609797289049436, + "loss": 0.6459, + "step": 18015 + }, + { + "epoch": 1.2501561307334674, + "grad_norm": 1.4842702150344849, + "learning_rate": 0.00016097527573330448, + "loss": 0.6372, + "step": 18016 + }, + { + "epoch": 1.250225522170564, + "grad_norm": 1.869614601135254, + "learning_rate": 0.00016097082236917352, + "loss": 0.6659, + "step": 18017 + }, + { + "epoch": 1.2502949136076609, + "grad_norm": 1.8130247592926025, + "learning_rate": 0.0001609663688125647, + "loss": 0.6803, + "step": 18018 + }, + { + "epoch": 1.2503643050447575, + "grad_norm": 1.7079789638519287, + "learning_rate": 0.00016096191506349218, + "loss": 0.6358, + "step": 18019 + }, + { + "epoch": 1.250433696481854, + "grad_norm": 1.7095909118652344, + "learning_rate": 0.0001609574611219699, + "loss": 0.629, + "step": 18020 + }, + { + "epoch": 1.250503087918951, + "grad_norm": 1.9631248712539673, + "learning_rate": 0.00016095300698801203, + "loss": 0.6098, + "step": 18021 + }, + { + "epoch": 1.2505724793560475, + "grad_norm": 1.3579086065292358, + "learning_rate": 0.0001609485526616326, + "loss": 0.4264, + "step": 18022 + }, + { + "epoch": 1.2506418707931442, + "grad_norm": 1.7182894945144653, + "learning_rate": 0.00016094409814284565, + "loss": 0.7604, + "step": 18023 + }, + { + "epoch": 1.2507112622302408, + "grad_norm": 1.738000750541687, + "learning_rate": 0.00016093964343166523, + "loss": 0.8262, + "step": 18024 + }, + { + "epoch": 1.2507806536673374, + "grad_norm": 2.047062873840332, + "learning_rate": 0.00016093518852810544, + "loss": 0.4819, + "step": 18025 + }, + { + "epoch": 1.2508500451044342, + "grad_norm": 1.9962271451950073, + "learning_rate": 0.00016093073343218035, + "loss": 0.5697, + "step": 18026 + }, + { + "epoch": 1.2509194365415308, + "grad_norm": 1.960425853729248, + "learning_rate": 0.000160926278143904, + "loss": 0.9576, + "step": 18027 + }, + { + "epoch": 1.2509888279786274, + "grad_norm": 2.213087320327759, + "learning_rate": 0.00016092182266329042, + "loss": 0.6263, + "step": 18028 + }, + { + "epoch": 1.251058219415724, + "grad_norm": 1.8896340131759644, + "learning_rate": 0.00016091736699035377, + "loss": 0.5579, + "step": 18029 + }, + { + "epoch": 1.2511276108528206, + "grad_norm": 1.1221575736999512, + "learning_rate": 0.000160912911125108, + "loss": 0.4238, + "step": 18030 + }, + { + "epoch": 1.2511970022899175, + "grad_norm": 1.7008916139602661, + "learning_rate": 0.00016090845506756729, + "loss": 0.818, + "step": 18031 + }, + { + "epoch": 1.251266393727014, + "grad_norm": 1.6707212924957275, + "learning_rate": 0.00016090399881774566, + "loss": 0.7046, + "step": 18032 + }, + { + "epoch": 1.2513357851641107, + "grad_norm": 1.9588956832885742, + "learning_rate": 0.0001608995423756571, + "loss": 0.464, + "step": 18033 + }, + { + "epoch": 1.2514051766012075, + "grad_norm": 2.0279951095581055, + "learning_rate": 0.0001608950857413158, + "loss": 0.7373, + "step": 18034 + }, + { + "epoch": 1.2514745680383041, + "grad_norm": 2.159266471862793, + "learning_rate": 0.0001608906289147358, + "loss": 0.8444, + "step": 18035 + }, + { + "epoch": 1.2515439594754008, + "grad_norm": 1.2873799800872803, + "learning_rate": 0.0001608861718959311, + "loss": 0.4791, + "step": 18036 + }, + { + "epoch": 1.2516133509124974, + "grad_norm": 1.62180495262146, + "learning_rate": 0.00016088171468491583, + "loss": 0.4952, + "step": 18037 + }, + { + "epoch": 1.251682742349594, + "grad_norm": 1.6120235919952393, + "learning_rate": 0.00016087725728170405, + "loss": 0.6346, + "step": 18038 + }, + { + "epoch": 1.2517521337866908, + "grad_norm": 1.9576655626296997, + "learning_rate": 0.00016087279968630987, + "loss": 0.7429, + "step": 18039 + }, + { + "epoch": 1.2518215252237874, + "grad_norm": 1.474075198173523, + "learning_rate": 0.0001608683418987473, + "loss": 0.5414, + "step": 18040 + }, + { + "epoch": 1.251890916660884, + "grad_norm": 2.5584404468536377, + "learning_rate": 0.0001608638839190304, + "loss": 0.6816, + "step": 18041 + }, + { + "epoch": 1.2519603080979806, + "grad_norm": 1.9268736839294434, + "learning_rate": 0.00016085942574717334, + "loss": 0.6196, + "step": 18042 + }, + { + "epoch": 1.2520296995350773, + "grad_norm": 2.066261053085327, + "learning_rate": 0.0001608549673831901, + "loss": 0.7262, + "step": 18043 + }, + { + "epoch": 1.252099090972174, + "grad_norm": 1.6488126516342163, + "learning_rate": 0.0001608505088270948, + "loss": 0.6286, + "step": 18044 + }, + { + "epoch": 1.2521684824092707, + "grad_norm": 1.273197054862976, + "learning_rate": 0.0001608460500789015, + "loss": 0.4653, + "step": 18045 + }, + { + "epoch": 1.2522378738463673, + "grad_norm": 1.9478603601455688, + "learning_rate": 0.0001608415911386243, + "loss": 0.5706, + "step": 18046 + }, + { + "epoch": 1.2523072652834641, + "grad_norm": 1.851879358291626, + "learning_rate": 0.00016083713200627724, + "loss": 0.7905, + "step": 18047 + }, + { + "epoch": 1.2523766567205608, + "grad_norm": 1.8502732515335083, + "learning_rate": 0.0001608326726818744, + "loss": 0.6722, + "step": 18048 + }, + { + "epoch": 1.2524460481576574, + "grad_norm": 2.036578416824341, + "learning_rate": 0.00016082821316542987, + "loss": 0.6849, + "step": 18049 + }, + { + "epoch": 1.252515439594754, + "grad_norm": 2.141183614730835, + "learning_rate": 0.00016082375345695776, + "loss": 0.8215, + "step": 18050 + }, + { + "epoch": 1.2525848310318506, + "grad_norm": 1.5093967914581299, + "learning_rate": 0.00016081929355647208, + "loss": 0.5017, + "step": 18051 + }, + { + "epoch": 1.2526542224689474, + "grad_norm": 1.6141380071640015, + "learning_rate": 0.00016081483346398696, + "loss": 0.6447, + "step": 18052 + }, + { + "epoch": 1.252723613906044, + "grad_norm": 2.0760929584503174, + "learning_rate": 0.00016081037317951648, + "loss": 0.8648, + "step": 18053 + }, + { + "epoch": 1.2527930053431406, + "grad_norm": 3.0968000888824463, + "learning_rate": 0.0001608059127030747, + "loss": 0.89, + "step": 18054 + }, + { + "epoch": 1.2528623967802373, + "grad_norm": 2.3223085403442383, + "learning_rate": 0.00016080145203467566, + "loss": 0.8074, + "step": 18055 + }, + { + "epoch": 1.2529317882173339, + "grad_norm": 2.0309078693389893, + "learning_rate": 0.00016079699117433352, + "loss": 0.7305, + "step": 18056 + }, + { + "epoch": 1.2530011796544307, + "grad_norm": 1.6755565404891968, + "learning_rate": 0.0001607925301220624, + "loss": 0.5916, + "step": 18057 + }, + { + "epoch": 1.2530705710915273, + "grad_norm": 2.3687827587127686, + "learning_rate": 0.00016078806887787623, + "loss": 0.9549, + "step": 18058 + }, + { + "epoch": 1.253139962528624, + "grad_norm": 1.7112491130828857, + "learning_rate": 0.0001607836074417892, + "loss": 0.7114, + "step": 18059 + }, + { + "epoch": 1.2532093539657208, + "grad_norm": 1.3128082752227783, + "learning_rate": 0.0001607791458138154, + "loss": 0.5825, + "step": 18060 + }, + { + "epoch": 1.2532787454028174, + "grad_norm": 1.647361159324646, + "learning_rate": 0.00016077468399396886, + "loss": 0.5462, + "step": 18061 + }, + { + "epoch": 1.253348136839914, + "grad_norm": 1.5614818334579468, + "learning_rate": 0.00016077022198226373, + "loss": 0.5646, + "step": 18062 + }, + { + "epoch": 1.2534175282770106, + "grad_norm": 1.8650089502334595, + "learning_rate": 0.000160765759778714, + "loss": 0.6158, + "step": 18063 + }, + { + "epoch": 1.2534869197141072, + "grad_norm": 1.8957926034927368, + "learning_rate": 0.00016076129738333387, + "loss": 0.6421, + "step": 18064 + }, + { + "epoch": 1.253556311151204, + "grad_norm": 1.6499799489974976, + "learning_rate": 0.00016075683479613737, + "loss": 0.5278, + "step": 18065 + }, + { + "epoch": 1.2536257025883006, + "grad_norm": 1.7211805582046509, + "learning_rate": 0.00016075237201713857, + "loss": 0.7357, + "step": 18066 + }, + { + "epoch": 1.2536950940253972, + "grad_norm": 1.7273412942886353, + "learning_rate": 0.0001607479090463516, + "loss": 0.5956, + "step": 18067 + }, + { + "epoch": 1.2537644854624939, + "grad_norm": 1.3810508251190186, + "learning_rate": 0.0001607434458837905, + "loss": 0.5811, + "step": 18068 + }, + { + "epoch": 1.2538338768995905, + "grad_norm": 2.047081708908081, + "learning_rate": 0.00016073898252946938, + "loss": 0.8517, + "step": 18069 + }, + { + "epoch": 1.2539032683366873, + "grad_norm": 2.053525686264038, + "learning_rate": 0.0001607345189834024, + "loss": 0.8405, + "step": 18070 + }, + { + "epoch": 1.253972659773784, + "grad_norm": 1.6223182678222656, + "learning_rate": 0.00016073005524560357, + "loss": 0.7177, + "step": 18071 + }, + { + "epoch": 1.2540420512108805, + "grad_norm": 1.7048063278198242, + "learning_rate": 0.00016072559131608696, + "loss": 0.668, + "step": 18072 + }, + { + "epoch": 1.2541114426479774, + "grad_norm": 1.7507078647613525, + "learning_rate": 0.0001607211271948667, + "loss": 0.6116, + "step": 18073 + }, + { + "epoch": 1.254180834085074, + "grad_norm": 2.1046628952026367, + "learning_rate": 0.00016071666288195696, + "loss": 0.6572, + "step": 18074 + }, + { + "epoch": 1.2542502255221706, + "grad_norm": 1.3992023468017578, + "learning_rate": 0.00016071219837737171, + "loss": 0.573, + "step": 18075 + }, + { + "epoch": 1.2543196169592672, + "grad_norm": 1.5245033502578735, + "learning_rate": 0.00016070773368112508, + "loss": 0.5397, + "step": 18076 + }, + { + "epoch": 1.2543890083963638, + "grad_norm": 1.3223642110824585, + "learning_rate": 0.00016070326879323124, + "loss": 0.407, + "step": 18077 + }, + { + "epoch": 1.2544583998334606, + "grad_norm": 1.7423782348632812, + "learning_rate": 0.00016069880371370415, + "loss": 0.7809, + "step": 18078 + }, + { + "epoch": 1.2545277912705572, + "grad_norm": 1.5947165489196777, + "learning_rate": 0.000160694338442558, + "loss": 0.5918, + "step": 18079 + }, + { + "epoch": 1.2545971827076539, + "grad_norm": 1.3887581825256348, + "learning_rate": 0.0001606898729798069, + "loss": 0.4748, + "step": 18080 + }, + { + "epoch": 1.2546665741447505, + "grad_norm": 1.2865407466888428, + "learning_rate": 0.0001606854073254649, + "loss": 0.6147, + "step": 18081 + }, + { + "epoch": 1.254735965581847, + "grad_norm": 1.8465908765792847, + "learning_rate": 0.00016068094147954608, + "loss": 0.6242, + "step": 18082 + }, + { + "epoch": 1.254805357018944, + "grad_norm": 1.874404788017273, + "learning_rate": 0.0001606764754420646, + "loss": 0.5044, + "step": 18083 + }, + { + "epoch": 1.2548747484560405, + "grad_norm": 1.7329519987106323, + "learning_rate": 0.00016067200921303452, + "loss": 0.6057, + "step": 18084 + }, + { + "epoch": 1.2549441398931371, + "grad_norm": 2.1521894931793213, + "learning_rate": 0.0001606675427924699, + "loss": 0.8315, + "step": 18085 + }, + { + "epoch": 1.255013531330234, + "grad_norm": 1.7198702096939087, + "learning_rate": 0.00016066307618038493, + "loss": 0.4942, + "step": 18086 + }, + { + "epoch": 1.2550829227673306, + "grad_norm": 1.899653434753418, + "learning_rate": 0.00016065860937679367, + "loss": 0.6767, + "step": 18087 + }, + { + "epoch": 1.2551523142044272, + "grad_norm": 1.168166160583496, + "learning_rate": 0.00016065414238171025, + "loss": 0.4395, + "step": 18088 + }, + { + "epoch": 1.2552217056415238, + "grad_norm": 1.4306745529174805, + "learning_rate": 0.0001606496751951487, + "loss": 0.6318, + "step": 18089 + }, + { + "epoch": 1.2552910970786204, + "grad_norm": 1.8597060441970825, + "learning_rate": 0.00016064520781712313, + "loss": 0.5547, + "step": 18090 + }, + { + "epoch": 1.2553604885157172, + "grad_norm": 1.511344075202942, + "learning_rate": 0.0001606407402476477, + "loss": 0.4696, + "step": 18091 + }, + { + "epoch": 1.2554298799528139, + "grad_norm": 2.171083450317383, + "learning_rate": 0.00016063627248673651, + "loss": 0.7494, + "step": 18092 + }, + { + "epoch": 1.2554992713899105, + "grad_norm": 2.166731357574463, + "learning_rate": 0.00016063180453440362, + "loss": 0.6282, + "step": 18093 + }, + { + "epoch": 1.255568662827007, + "grad_norm": 2.530250310897827, + "learning_rate": 0.00016062733639066317, + "loss": 0.7132, + "step": 18094 + }, + { + "epoch": 1.2556380542641037, + "grad_norm": 2.0936455726623535, + "learning_rate": 0.00016062286805552925, + "loss": 0.9058, + "step": 18095 + }, + { + "epoch": 1.2557074457012005, + "grad_norm": 1.7089605331420898, + "learning_rate": 0.00016061839952901598, + "loss": 0.4291, + "step": 18096 + }, + { + "epoch": 1.2557768371382971, + "grad_norm": 2.0098414421081543, + "learning_rate": 0.00016061393081113744, + "loss": 0.5821, + "step": 18097 + }, + { + "epoch": 1.2558462285753937, + "grad_norm": 2.450417995452881, + "learning_rate": 0.00016060946190190776, + "loss": 0.6726, + "step": 18098 + }, + { + "epoch": 1.2559156200124906, + "grad_norm": 2.324639081954956, + "learning_rate": 0.00016060499280134104, + "loss": 0.758, + "step": 18099 + }, + { + "epoch": 1.2559850114495872, + "grad_norm": 1.7580373287200928, + "learning_rate": 0.0001606005235094514, + "loss": 0.7115, + "step": 18100 + }, + { + "epoch": 1.2560544028866838, + "grad_norm": 1.8800135850906372, + "learning_rate": 0.0001605960540262529, + "loss": 0.7796, + "step": 18101 + }, + { + "epoch": 1.2561237943237804, + "grad_norm": 1.2960947751998901, + "learning_rate": 0.0001605915843517597, + "loss": 0.4146, + "step": 18102 + }, + { + "epoch": 1.256193185760877, + "grad_norm": 2.107304573059082, + "learning_rate": 0.00016058711448598592, + "loss": 0.7907, + "step": 18103 + }, + { + "epoch": 1.2562625771979739, + "grad_norm": 2.063359022140503, + "learning_rate": 0.00016058264442894563, + "loss": 0.6941, + "step": 18104 + }, + { + "epoch": 1.2563319686350705, + "grad_norm": 3.189807176589966, + "learning_rate": 0.000160578174180653, + "loss": 0.9258, + "step": 18105 + }, + { + "epoch": 1.256401360072167, + "grad_norm": 2.0500309467315674, + "learning_rate": 0.00016057370374112205, + "loss": 0.622, + "step": 18106 + }, + { + "epoch": 1.2564707515092637, + "grad_norm": 2.944847345352173, + "learning_rate": 0.00016056923311036697, + "loss": 0.7287, + "step": 18107 + }, + { + "epoch": 1.2565401429463603, + "grad_norm": 1.4860217571258545, + "learning_rate": 0.00016056476228840184, + "loss": 0.5777, + "step": 18108 + }, + { + "epoch": 1.2566095343834571, + "grad_norm": 1.7286251783370972, + "learning_rate": 0.00016056029127524082, + "loss": 0.5489, + "step": 18109 + }, + { + "epoch": 1.2566789258205537, + "grad_norm": 1.6773632764816284, + "learning_rate": 0.00016055582007089793, + "loss": 0.6928, + "step": 18110 + }, + { + "epoch": 1.2567483172576503, + "grad_norm": 1.87760329246521, + "learning_rate": 0.00016055134867538738, + "loss": 0.6152, + "step": 18111 + }, + { + "epoch": 1.2568177086947472, + "grad_norm": 1.9609423875808716, + "learning_rate": 0.00016054687708872326, + "loss": 0.7561, + "step": 18112 + }, + { + "epoch": 1.2568871001318438, + "grad_norm": 2.033606767654419, + "learning_rate": 0.0001605424053109196, + "loss": 0.4814, + "step": 18113 + }, + { + "epoch": 1.2569564915689404, + "grad_norm": 1.6449054479599, + "learning_rate": 0.00016053793334199065, + "loss": 0.5593, + "step": 18114 + }, + { + "epoch": 1.257025883006037, + "grad_norm": 1.8035013675689697, + "learning_rate": 0.00016053346118195046, + "loss": 0.5378, + "step": 18115 + }, + { + "epoch": 1.2570952744431336, + "grad_norm": 2.1273179054260254, + "learning_rate": 0.00016052898883081316, + "loss": 0.7286, + "step": 18116 + }, + { + "epoch": 1.2571646658802305, + "grad_norm": 2.2614893913269043, + "learning_rate": 0.00016052451628859285, + "loss": 0.7022, + "step": 18117 + }, + { + "epoch": 1.257234057317327, + "grad_norm": 2.374966621398926, + "learning_rate": 0.00016052004355530367, + "loss": 0.6697, + "step": 18118 + }, + { + "epoch": 1.2573034487544237, + "grad_norm": 1.7327736616134644, + "learning_rate": 0.00016051557063095974, + "loss": 0.6373, + "step": 18119 + }, + { + "epoch": 1.2573728401915203, + "grad_norm": 1.0611774921417236, + "learning_rate": 0.00016051109751557516, + "loss": 0.4604, + "step": 18120 + }, + { + "epoch": 1.257442231628617, + "grad_norm": 1.4225385189056396, + "learning_rate": 0.0001605066242091641, + "loss": 0.4706, + "step": 18121 + }, + { + "epoch": 1.2575116230657137, + "grad_norm": 2.1824145317077637, + "learning_rate": 0.00016050215071174062, + "loss": 0.7264, + "step": 18122 + }, + { + "epoch": 1.2575810145028103, + "grad_norm": 1.9438982009887695, + "learning_rate": 0.0001604976770233189, + "loss": 0.6365, + "step": 18123 + }, + { + "epoch": 1.257650405939907, + "grad_norm": 1.365413784980774, + "learning_rate": 0.00016049320314391298, + "loss": 0.5325, + "step": 18124 + }, + { + "epoch": 1.2577197973770038, + "grad_norm": 2.0130391120910645, + "learning_rate": 0.00016048872907353708, + "loss": 0.6203, + "step": 18125 + }, + { + "epoch": 1.2577891888141004, + "grad_norm": 1.6097396612167358, + "learning_rate": 0.00016048425481220526, + "loss": 0.7176, + "step": 18126 + }, + { + "epoch": 1.257858580251197, + "grad_norm": 1.9816080331802368, + "learning_rate": 0.00016047978035993164, + "loss": 0.5552, + "step": 18127 + }, + { + "epoch": 1.2579279716882936, + "grad_norm": 2.8006529808044434, + "learning_rate": 0.0001604753057167304, + "loss": 0.7772, + "step": 18128 + }, + { + "epoch": 1.2579973631253902, + "grad_norm": 1.9171998500823975, + "learning_rate": 0.00016047083088261565, + "loss": 0.7878, + "step": 18129 + }, + { + "epoch": 1.258066754562487, + "grad_norm": 1.7927600145339966, + "learning_rate": 0.00016046635585760147, + "loss": 0.5774, + "step": 18130 + }, + { + "epoch": 1.2581361459995837, + "grad_norm": 2.638533353805542, + "learning_rate": 0.00016046188064170207, + "loss": 0.7147, + "step": 18131 + }, + { + "epoch": 1.2582055374366803, + "grad_norm": 1.9631861448287964, + "learning_rate": 0.00016045740523493147, + "loss": 0.7227, + "step": 18132 + }, + { + "epoch": 1.258274928873777, + "grad_norm": 1.4764866828918457, + "learning_rate": 0.0001604529296373039, + "loss": 0.662, + "step": 18133 + }, + { + "epoch": 1.2583443203108735, + "grad_norm": 1.922607660293579, + "learning_rate": 0.00016044845384883343, + "loss": 0.6089, + "step": 18134 + }, + { + "epoch": 1.2584137117479703, + "grad_norm": 1.8698925971984863, + "learning_rate": 0.00016044397786953418, + "loss": 0.6598, + "step": 18135 + }, + { + "epoch": 1.258483103185067, + "grad_norm": 1.893216848373413, + "learning_rate": 0.00016043950169942036, + "loss": 0.6595, + "step": 18136 + }, + { + "epoch": 1.2585524946221636, + "grad_norm": 1.8468403816223145, + "learning_rate": 0.00016043502533850598, + "loss": 0.906, + "step": 18137 + }, + { + "epoch": 1.2586218860592604, + "grad_norm": 1.6045095920562744, + "learning_rate": 0.00016043054878680529, + "loss": 0.4476, + "step": 18138 + }, + { + "epoch": 1.258691277496357, + "grad_norm": 1.4978183507919312, + "learning_rate": 0.00016042607204433233, + "loss": 0.5614, + "step": 18139 + }, + { + "epoch": 1.2587606689334536, + "grad_norm": 2.080897569656372, + "learning_rate": 0.0001604215951111013, + "loss": 0.6815, + "step": 18140 + }, + { + "epoch": 1.2588300603705502, + "grad_norm": 2.0436155796051025, + "learning_rate": 0.00016041711798712627, + "loss": 0.6377, + "step": 18141 + }, + { + "epoch": 1.2588994518076468, + "grad_norm": 1.7837809324264526, + "learning_rate": 0.00016041264067242144, + "loss": 0.7904, + "step": 18142 + }, + { + "epoch": 1.2589688432447437, + "grad_norm": 1.9583714008331299, + "learning_rate": 0.00016040816316700092, + "loss": 0.5485, + "step": 18143 + }, + { + "epoch": 1.2590382346818403, + "grad_norm": 2.441941022872925, + "learning_rate": 0.0001604036854708788, + "loss": 0.6186, + "step": 18144 + }, + { + "epoch": 1.259107626118937, + "grad_norm": 2.228447914123535, + "learning_rate": 0.00016039920758406927, + "loss": 0.7313, + "step": 18145 + }, + { + "epoch": 1.2591770175560335, + "grad_norm": 1.893997311592102, + "learning_rate": 0.0001603947295065865, + "loss": 0.5951, + "step": 18146 + }, + { + "epoch": 1.2592464089931301, + "grad_norm": 2.0801005363464355, + "learning_rate": 0.0001603902512384445, + "loss": 0.9016, + "step": 18147 + }, + { + "epoch": 1.259315800430227, + "grad_norm": 1.902347207069397, + "learning_rate": 0.00016038577277965752, + "loss": 0.5142, + "step": 18148 + }, + { + "epoch": 1.2593851918673236, + "grad_norm": 2.3165228366851807, + "learning_rate": 0.00016038129413023966, + "loss": 0.5244, + "step": 18149 + }, + { + "epoch": 1.2594545833044202, + "grad_norm": 1.723064661026001, + "learning_rate": 0.00016037681529020505, + "loss": 0.7661, + "step": 18150 + }, + { + "epoch": 1.259523974741517, + "grad_norm": 3.312047243118286, + "learning_rate": 0.00016037233625956784, + "loss": 0.5951, + "step": 18151 + }, + { + "epoch": 1.2595933661786136, + "grad_norm": 2.2431907653808594, + "learning_rate": 0.00016036785703834217, + "loss": 0.7138, + "step": 18152 + }, + { + "epoch": 1.2596627576157102, + "grad_norm": 1.7364473342895508, + "learning_rate": 0.0001603633776265422, + "loss": 0.5976, + "step": 18153 + }, + { + "epoch": 1.2597321490528068, + "grad_norm": 1.7351937294006348, + "learning_rate": 0.000160358898024182, + "loss": 0.5815, + "step": 18154 + }, + { + "epoch": 1.2598015404899034, + "grad_norm": 1.4223984479904175, + "learning_rate": 0.00016035441823127581, + "loss": 0.4786, + "step": 18155 + }, + { + "epoch": 1.2598709319270003, + "grad_norm": 1.9266953468322754, + "learning_rate": 0.00016034993824783772, + "loss": 0.5321, + "step": 18156 + }, + { + "epoch": 1.259940323364097, + "grad_norm": 2.233898878097534, + "learning_rate": 0.00016034545807388186, + "loss": 0.7835, + "step": 18157 + }, + { + "epoch": 1.2600097148011935, + "grad_norm": 1.6532518863677979, + "learning_rate": 0.00016034097770942237, + "loss": 0.7436, + "step": 18158 + }, + { + "epoch": 1.2600791062382901, + "grad_norm": 1.4985331296920776, + "learning_rate": 0.00016033649715447345, + "loss": 0.8928, + "step": 18159 + }, + { + "epoch": 1.2601484976753867, + "grad_norm": 1.9485936164855957, + "learning_rate": 0.0001603320164090492, + "loss": 0.6742, + "step": 18160 + }, + { + "epoch": 1.2602178891124836, + "grad_norm": 2.2339494228363037, + "learning_rate": 0.00016032753547316376, + "loss": 0.7726, + "step": 18161 + }, + { + "epoch": 1.2602872805495802, + "grad_norm": 2.309086799621582, + "learning_rate": 0.0001603230543468313, + "loss": 0.7294, + "step": 18162 + }, + { + "epoch": 1.2603566719866768, + "grad_norm": 1.7185291051864624, + "learning_rate": 0.00016031857303006595, + "loss": 0.6341, + "step": 18163 + }, + { + "epoch": 1.2604260634237736, + "grad_norm": 1.9033851623535156, + "learning_rate": 0.00016031409152288188, + "loss": 0.8597, + "step": 18164 + }, + { + "epoch": 1.2604954548608702, + "grad_norm": 2.08596134185791, + "learning_rate": 0.0001603096098252932, + "loss": 0.9149, + "step": 18165 + }, + { + "epoch": 1.2605648462979668, + "grad_norm": 2.087780714035034, + "learning_rate": 0.0001603051279373141, + "loss": 0.6703, + "step": 18166 + }, + { + "epoch": 1.2606342377350634, + "grad_norm": 1.8979744911193848, + "learning_rate": 0.0001603006458589587, + "loss": 0.6962, + "step": 18167 + }, + { + "epoch": 1.26070362917216, + "grad_norm": 2.8606460094451904, + "learning_rate": 0.00016029616359024116, + "loss": 0.744, + "step": 18168 + }, + { + "epoch": 1.260773020609257, + "grad_norm": 1.9066411256790161, + "learning_rate": 0.00016029168113117562, + "loss": 0.8914, + "step": 18169 + }, + { + "epoch": 1.2608424120463535, + "grad_norm": 1.5662084817886353, + "learning_rate": 0.00016028719848177624, + "loss": 0.6385, + "step": 18170 + }, + { + "epoch": 1.26091180348345, + "grad_norm": 1.8504555225372314, + "learning_rate": 0.00016028271564205718, + "loss": 0.6186, + "step": 18171 + }, + { + "epoch": 1.2609811949205467, + "grad_norm": 1.6876351833343506, + "learning_rate": 0.00016027823261203255, + "loss": 0.7135, + "step": 18172 + }, + { + "epoch": 1.2610505863576433, + "grad_norm": 2.599321126937866, + "learning_rate": 0.00016027374939171657, + "loss": 0.8661, + "step": 18173 + }, + { + "epoch": 1.2611199777947402, + "grad_norm": 2.414931058883667, + "learning_rate": 0.00016026926598112335, + "loss": 0.8575, + "step": 18174 + }, + { + "epoch": 1.2611893692318368, + "grad_norm": 1.889611005783081, + "learning_rate": 0.00016026478238026704, + "loss": 0.7338, + "step": 18175 + }, + { + "epoch": 1.2612587606689334, + "grad_norm": 2.0844151973724365, + "learning_rate": 0.0001602602985891618, + "loss": 0.7389, + "step": 18176 + }, + { + "epoch": 1.2613281521060302, + "grad_norm": 1.9811456203460693, + "learning_rate": 0.00016025581460782178, + "loss": 0.7832, + "step": 18177 + }, + { + "epoch": 1.2613975435431268, + "grad_norm": 1.7134521007537842, + "learning_rate": 0.00016025133043626118, + "loss": 0.6568, + "step": 18178 + }, + { + "epoch": 1.2614669349802234, + "grad_norm": 1.4756773710250854, + "learning_rate": 0.00016024684607449409, + "loss": 0.6035, + "step": 18179 + }, + { + "epoch": 1.26153632641732, + "grad_norm": 1.4581398963928223, + "learning_rate": 0.0001602423615225347, + "loss": 0.4692, + "step": 18180 + }, + { + "epoch": 1.2616057178544167, + "grad_norm": 1.8557285070419312, + "learning_rate": 0.00016023787678039718, + "loss": 0.5543, + "step": 18181 + }, + { + "epoch": 1.2616751092915135, + "grad_norm": 1.8804285526275635, + "learning_rate": 0.00016023339184809566, + "loss": 0.5574, + "step": 18182 + }, + { + "epoch": 1.26174450072861, + "grad_norm": 1.5485225915908813, + "learning_rate": 0.0001602289067256443, + "loss": 0.5675, + "step": 18183 + }, + { + "epoch": 1.2618138921657067, + "grad_norm": 2.4727985858917236, + "learning_rate": 0.00016022442141305727, + "loss": 0.5994, + "step": 18184 + }, + { + "epoch": 1.2618832836028036, + "grad_norm": 1.7969990968704224, + "learning_rate": 0.00016021993591034875, + "loss": 0.5924, + "step": 18185 + }, + { + "epoch": 1.2619526750399, + "grad_norm": 1.5945583581924438, + "learning_rate": 0.00016021545021753286, + "loss": 0.7113, + "step": 18186 + }, + { + "epoch": 1.2620220664769968, + "grad_norm": 1.929362177848816, + "learning_rate": 0.00016021096433462378, + "loss": 0.5494, + "step": 18187 + }, + { + "epoch": 1.2620914579140934, + "grad_norm": 1.5985480546951294, + "learning_rate": 0.0001602064782616357, + "loss": 0.6647, + "step": 18188 + }, + { + "epoch": 1.26216084935119, + "grad_norm": 1.9277594089508057, + "learning_rate": 0.00016020199199858273, + "loss": 0.9625, + "step": 18189 + }, + { + "epoch": 1.2622302407882868, + "grad_norm": 1.3225349187850952, + "learning_rate": 0.00016019750554547907, + "loss": 0.6149, + "step": 18190 + }, + { + "epoch": 1.2622996322253834, + "grad_norm": 1.6371333599090576, + "learning_rate": 0.00016019301890233883, + "loss": 0.5724, + "step": 18191 + }, + { + "epoch": 1.26236902366248, + "grad_norm": 1.7173526287078857, + "learning_rate": 0.00016018853206917623, + "loss": 0.56, + "step": 18192 + }, + { + "epoch": 1.2624384150995767, + "grad_norm": 1.615090250968933, + "learning_rate": 0.00016018404504600543, + "loss": 0.538, + "step": 18193 + }, + { + "epoch": 1.2625078065366733, + "grad_norm": 1.4479893445968628, + "learning_rate": 0.0001601795578328406, + "loss": 0.6551, + "step": 18194 + }, + { + "epoch": 1.26257719797377, + "grad_norm": 4.4534807205200195, + "learning_rate": 0.00016017507042969585, + "loss": 0.711, + "step": 18195 + }, + { + "epoch": 1.2626465894108667, + "grad_norm": 1.6877226829528809, + "learning_rate": 0.00016017058283658542, + "loss": 0.6914, + "step": 18196 + }, + { + "epoch": 1.2627159808479633, + "grad_norm": 1.8047723770141602, + "learning_rate": 0.0001601660950535234, + "loss": 0.6601, + "step": 18197 + }, + { + "epoch": 1.2627853722850602, + "grad_norm": 2.1329269409179688, + "learning_rate": 0.00016016160708052406, + "loss": 0.7674, + "step": 18198 + }, + { + "epoch": 1.2628547637221565, + "grad_norm": 1.935913324356079, + "learning_rate": 0.00016015711891760145, + "loss": 0.8291, + "step": 18199 + }, + { + "epoch": 1.2629241551592534, + "grad_norm": 1.5368369817733765, + "learning_rate": 0.0001601526305647698, + "loss": 0.5296, + "step": 18200 + }, + { + "epoch": 1.26299354659635, + "grad_norm": 1.5278384685516357, + "learning_rate": 0.00016014814202204328, + "loss": 0.5503, + "step": 18201 + }, + { + "epoch": 1.2630629380334466, + "grad_norm": 1.8522273302078247, + "learning_rate": 0.0001601436532894361, + "loss": 0.6631, + "step": 18202 + }, + { + "epoch": 1.2631323294705434, + "grad_norm": 2.2173373699188232, + "learning_rate": 0.00016013916436696235, + "loss": 0.5027, + "step": 18203 + }, + { + "epoch": 1.26320172090764, + "grad_norm": 1.6761517524719238, + "learning_rate": 0.0001601346752546362, + "loss": 0.6307, + "step": 18204 + }, + { + "epoch": 1.2632711123447367, + "grad_norm": 1.7572890520095825, + "learning_rate": 0.0001601301859524719, + "loss": 0.5798, + "step": 18205 + }, + { + "epoch": 1.2633405037818333, + "grad_norm": 2.584765672683716, + "learning_rate": 0.00016012569646048358, + "loss": 0.7775, + "step": 18206 + }, + { + "epoch": 1.2634098952189299, + "grad_norm": 2.1612706184387207, + "learning_rate": 0.0001601212067786854, + "loss": 0.6233, + "step": 18207 + }, + { + "epoch": 1.2634792866560267, + "grad_norm": 2.014272928237915, + "learning_rate": 0.00016011671690709152, + "loss": 0.6087, + "step": 18208 + }, + { + "epoch": 1.2635486780931233, + "grad_norm": 1.8285022974014282, + "learning_rate": 0.0001601122268457162, + "loss": 0.6734, + "step": 18209 + }, + { + "epoch": 1.26361806953022, + "grad_norm": 1.2402983903884888, + "learning_rate": 0.00016010773659457352, + "loss": 0.3893, + "step": 18210 + }, + { + "epoch": 1.2636874609673168, + "grad_norm": 1.9414877891540527, + "learning_rate": 0.0001601032461536777, + "loss": 0.6209, + "step": 18211 + }, + { + "epoch": 1.2637568524044134, + "grad_norm": 1.9332815408706665, + "learning_rate": 0.00016009875552304287, + "loss": 0.8713, + "step": 18212 + }, + { + "epoch": 1.26382624384151, + "grad_norm": 1.7089446783065796, + "learning_rate": 0.00016009426470268328, + "loss": 0.5342, + "step": 18213 + }, + { + "epoch": 1.2638956352786066, + "grad_norm": 3.475461006164551, + "learning_rate": 0.00016008977369261305, + "loss": 0.773, + "step": 18214 + }, + { + "epoch": 1.2639650267157032, + "grad_norm": 1.601523518562317, + "learning_rate": 0.00016008528249284636, + "loss": 0.5959, + "step": 18215 + }, + { + "epoch": 1.2640344181528, + "grad_norm": 2.3891448974609375, + "learning_rate": 0.00016008079110339744, + "loss": 0.8009, + "step": 18216 + }, + { + "epoch": 1.2641038095898967, + "grad_norm": 2.042355537414551, + "learning_rate": 0.00016007629952428042, + "loss": 0.7077, + "step": 18217 + }, + { + "epoch": 1.2641732010269933, + "grad_norm": 2.126451015472412, + "learning_rate": 0.0001600718077555095, + "loss": 0.6632, + "step": 18218 + }, + { + "epoch": 1.2642425924640899, + "grad_norm": 1.5158014297485352, + "learning_rate": 0.00016006731579709885, + "loss": 0.6404, + "step": 18219 + }, + { + "epoch": 1.2643119839011865, + "grad_norm": 1.9047513008117676, + "learning_rate": 0.00016006282364906262, + "loss": 0.7035, + "step": 18220 + }, + { + "epoch": 1.2643813753382833, + "grad_norm": 1.198936104774475, + "learning_rate": 0.00016005833131141505, + "loss": 0.3463, + "step": 18221 + }, + { + "epoch": 1.26445076677538, + "grad_norm": 2.2380430698394775, + "learning_rate": 0.0001600538387841703, + "loss": 0.7689, + "step": 18222 + }, + { + "epoch": 1.2645201582124765, + "grad_norm": 1.262662649154663, + "learning_rate": 0.00016004934606734253, + "loss": 0.4766, + "step": 18223 + }, + { + "epoch": 1.2645895496495734, + "grad_norm": 2.3835113048553467, + "learning_rate": 0.00016004485316094597, + "loss": 0.7091, + "step": 18224 + }, + { + "epoch": 1.26465894108667, + "grad_norm": 1.536553978919983, + "learning_rate": 0.00016004036006499477, + "loss": 0.5541, + "step": 18225 + }, + { + "epoch": 1.2647283325237666, + "grad_norm": 2.256356716156006, + "learning_rate": 0.00016003586677950312, + "loss": 0.768, + "step": 18226 + }, + { + "epoch": 1.2647977239608632, + "grad_norm": 2.1646413803100586, + "learning_rate": 0.00016003137330448518, + "loss": 0.6368, + "step": 18227 + }, + { + "epoch": 1.2648671153979598, + "grad_norm": 2.0655758380889893, + "learning_rate": 0.00016002687963995518, + "loss": 0.5654, + "step": 18228 + }, + { + "epoch": 1.2649365068350567, + "grad_norm": 2.130519151687622, + "learning_rate": 0.0001600223857859273, + "loss": 0.8848, + "step": 18229 + }, + { + "epoch": 1.2650058982721533, + "grad_norm": 1.5618276596069336, + "learning_rate": 0.00016001789174241568, + "loss": 0.5287, + "step": 18230 + }, + { + "epoch": 1.2650752897092499, + "grad_norm": 1.861830472946167, + "learning_rate": 0.00016001339750943455, + "loss": 0.7596, + "step": 18231 + }, + { + "epoch": 1.2651446811463465, + "grad_norm": 2.052067756652832, + "learning_rate": 0.0001600089030869981, + "loss": 0.5256, + "step": 18232 + }, + { + "epoch": 1.265214072583443, + "grad_norm": 1.1146798133850098, + "learning_rate": 0.00016000440847512048, + "loss": 0.5059, + "step": 18233 + }, + { + "epoch": 1.26528346402054, + "grad_norm": 2.026590585708618, + "learning_rate": 0.0001599999136738159, + "loss": 0.6014, + "step": 18234 + }, + { + "epoch": 1.2653528554576365, + "grad_norm": 2.2817211151123047, + "learning_rate": 0.00015999541868309857, + "loss": 0.693, + "step": 18235 + }, + { + "epoch": 1.2654222468947331, + "grad_norm": 1.980490803718567, + "learning_rate": 0.0001599909235029827, + "loss": 0.7048, + "step": 18236 + }, + { + "epoch": 1.26549163833183, + "grad_norm": 1.5677008628845215, + "learning_rate": 0.00015998642813348242, + "loss": 0.607, + "step": 18237 + }, + { + "epoch": 1.2655610297689266, + "grad_norm": 2.042006492614746, + "learning_rate": 0.00015998193257461192, + "loss": 0.5303, + "step": 18238 + }, + { + "epoch": 1.2656304212060232, + "grad_norm": 1.8593189716339111, + "learning_rate": 0.00015997743682638542, + "loss": 0.7609, + "step": 18239 + }, + { + "epoch": 1.2656998126431198, + "grad_norm": 1.6531929969787598, + "learning_rate": 0.00015997294088881714, + "loss": 0.6887, + "step": 18240 + }, + { + "epoch": 1.2657692040802164, + "grad_norm": 1.3932222127914429, + "learning_rate": 0.00015996844476192125, + "loss": 0.5299, + "step": 18241 + }, + { + "epoch": 1.2658385955173133, + "grad_norm": 1.7388828992843628, + "learning_rate": 0.0001599639484457119, + "loss": 0.5382, + "step": 18242 + }, + { + "epoch": 1.2659079869544099, + "grad_norm": 2.7689461708068848, + "learning_rate": 0.00015995945194020332, + "loss": 0.6444, + "step": 18243 + }, + { + "epoch": 1.2659773783915065, + "grad_norm": 1.9665052890777588, + "learning_rate": 0.00015995495524540973, + "loss": 0.7772, + "step": 18244 + }, + { + "epoch": 1.266046769828603, + "grad_norm": 1.4964375495910645, + "learning_rate": 0.0001599504583613453, + "loss": 0.5868, + "step": 18245 + }, + { + "epoch": 1.2661161612656997, + "grad_norm": 2.0433661937713623, + "learning_rate": 0.0001599459612880242, + "loss": 0.6293, + "step": 18246 + }, + { + "epoch": 1.2661855527027965, + "grad_norm": 1.4700039625167847, + "learning_rate": 0.00015994146402546072, + "loss": 0.5944, + "step": 18247 + }, + { + "epoch": 1.2662549441398931, + "grad_norm": 1.7470388412475586, + "learning_rate": 0.00015993696657366893, + "loss": 0.9128, + "step": 18248 + }, + { + "epoch": 1.2663243355769898, + "grad_norm": 2.6491687297821045, + "learning_rate": 0.00015993246893266313, + "loss": 0.8991, + "step": 18249 + }, + { + "epoch": 1.2663937270140866, + "grad_norm": 1.573773980140686, + "learning_rate": 0.00015992797110245746, + "loss": 0.5803, + "step": 18250 + }, + { + "epoch": 1.2664631184511832, + "grad_norm": 1.702839970588684, + "learning_rate": 0.00015992347308306615, + "loss": 0.6794, + "step": 18251 + }, + { + "epoch": 1.2665325098882798, + "grad_norm": 1.8408527374267578, + "learning_rate": 0.00015991897487450336, + "loss": 0.6158, + "step": 18252 + }, + { + "epoch": 1.2666019013253764, + "grad_norm": 1.902260661125183, + "learning_rate": 0.00015991447647678334, + "loss": 0.5914, + "step": 18253 + }, + { + "epoch": 1.266671292762473, + "grad_norm": 1.6690953969955444, + "learning_rate": 0.0001599099778899203, + "loss": 0.5788, + "step": 18254 + }, + { + "epoch": 1.2667406841995699, + "grad_norm": 1.7148836851119995, + "learning_rate": 0.00015990547911392834, + "loss": 0.8981, + "step": 18255 + }, + { + "epoch": 1.2668100756366665, + "grad_norm": 1.7474344968795776, + "learning_rate": 0.0001599009801488218, + "loss": 0.7106, + "step": 18256 + }, + { + "epoch": 1.266879467073763, + "grad_norm": 1.8950895071029663, + "learning_rate": 0.00015989648099461477, + "loss": 0.6289, + "step": 18257 + }, + { + "epoch": 1.2669488585108597, + "grad_norm": 2.265723466873169, + "learning_rate": 0.00015989198165132151, + "loss": 0.6384, + "step": 18258 + }, + { + "epoch": 1.2670182499479563, + "grad_norm": 1.7814093828201294, + "learning_rate": 0.00015988748211895622, + "loss": 0.6044, + "step": 18259 + }, + { + "epoch": 1.2670876413850531, + "grad_norm": 1.530847191810608, + "learning_rate": 0.00015988298239753309, + "loss": 0.5571, + "step": 18260 + }, + { + "epoch": 1.2671570328221498, + "grad_norm": 1.4233893156051636, + "learning_rate": 0.00015987848248706635, + "loss": 0.6109, + "step": 18261 + }, + { + "epoch": 1.2672264242592464, + "grad_norm": 1.8498790264129639, + "learning_rate": 0.00015987398238757016, + "loss": 0.9806, + "step": 18262 + }, + { + "epoch": 1.2672958156963432, + "grad_norm": 2.103412389755249, + "learning_rate": 0.00015986948209905876, + "loss": 0.6914, + "step": 18263 + }, + { + "epoch": 1.2673652071334398, + "grad_norm": 1.7461918592453003, + "learning_rate": 0.00015986498162154638, + "loss": 0.6929, + "step": 18264 + }, + { + "epoch": 1.2674345985705364, + "grad_norm": 1.210105299949646, + "learning_rate": 0.00015986048095504714, + "loss": 0.544, + "step": 18265 + }, + { + "epoch": 1.267503990007633, + "grad_norm": 1.8963823318481445, + "learning_rate": 0.00015985598009957537, + "loss": 0.509, + "step": 18266 + }, + { + "epoch": 1.2675733814447296, + "grad_norm": 2.1451833248138428, + "learning_rate": 0.00015985147905514521, + "loss": 0.8117, + "step": 18267 + }, + { + "epoch": 1.2676427728818265, + "grad_norm": 1.6099423170089722, + "learning_rate": 0.00015984697782177082, + "loss": 0.5377, + "step": 18268 + }, + { + "epoch": 1.267712164318923, + "grad_norm": 1.675623893737793, + "learning_rate": 0.0001598424763994665, + "loss": 0.4791, + "step": 18269 + }, + { + "epoch": 1.2677815557560197, + "grad_norm": 1.5757933855056763, + "learning_rate": 0.00015983797478824645, + "loss": 0.4904, + "step": 18270 + }, + { + "epoch": 1.2678509471931163, + "grad_norm": 1.5585750341415405, + "learning_rate": 0.00015983347298812484, + "loss": 0.4491, + "step": 18271 + }, + { + "epoch": 1.267920338630213, + "grad_norm": 1.7411890029907227, + "learning_rate": 0.00015982897099911587, + "loss": 0.5895, + "step": 18272 + }, + { + "epoch": 1.2679897300673098, + "grad_norm": 1.1968042850494385, + "learning_rate": 0.00015982446882123382, + "loss": 0.5245, + "step": 18273 + }, + { + "epoch": 1.2680591215044064, + "grad_norm": 1.8983957767486572, + "learning_rate": 0.00015981996645449283, + "loss": 0.6296, + "step": 18274 + }, + { + "epoch": 1.268128512941503, + "grad_norm": 2.276822328567505, + "learning_rate": 0.00015981546389890716, + "loss": 1.0462, + "step": 18275 + }, + { + "epoch": 1.2681979043785998, + "grad_norm": 2.6494390964508057, + "learning_rate": 0.000159810961154491, + "loss": 0.4812, + "step": 18276 + }, + { + "epoch": 1.2682672958156964, + "grad_norm": 1.5315790176391602, + "learning_rate": 0.00015980645822125857, + "loss": 0.4334, + "step": 18277 + }, + { + "epoch": 1.268336687252793, + "grad_norm": 1.5722719430923462, + "learning_rate": 0.0001598019550992241, + "loss": 0.7797, + "step": 18278 + }, + { + "epoch": 1.2684060786898896, + "grad_norm": 1.5030404329299927, + "learning_rate": 0.00015979745178840182, + "loss": 0.6365, + "step": 18279 + }, + { + "epoch": 1.2684754701269862, + "grad_norm": 1.8959881067276, + "learning_rate": 0.0001597929482888059, + "loss": 0.6044, + "step": 18280 + }, + { + "epoch": 1.268544861564083, + "grad_norm": 1.9987678527832031, + "learning_rate": 0.00015978844460045057, + "loss": 0.6612, + "step": 18281 + }, + { + "epoch": 1.2686142530011797, + "grad_norm": 2.198850393295288, + "learning_rate": 0.0001597839407233501, + "loss": 0.8204, + "step": 18282 + }, + { + "epoch": 1.2686836444382763, + "grad_norm": 1.8507111072540283, + "learning_rate": 0.00015977943665751863, + "loss": 0.7718, + "step": 18283 + }, + { + "epoch": 1.268753035875373, + "grad_norm": 1.8272948265075684, + "learning_rate": 0.00015977493240297038, + "loss": 0.7816, + "step": 18284 + }, + { + "epoch": 1.2688224273124695, + "grad_norm": 1.931828260421753, + "learning_rate": 0.00015977042795971966, + "loss": 0.7377, + "step": 18285 + }, + { + "epoch": 1.2688918187495664, + "grad_norm": 1.9402589797973633, + "learning_rate": 0.00015976592332778063, + "loss": 0.538, + "step": 18286 + }, + { + "epoch": 1.268961210186663, + "grad_norm": 2.220019817352295, + "learning_rate": 0.00015976141850716747, + "loss": 0.704, + "step": 18287 + }, + { + "epoch": 1.2690306016237596, + "grad_norm": 2.6880171298980713, + "learning_rate": 0.0001597569134978945, + "loss": 0.8219, + "step": 18288 + }, + { + "epoch": 1.2690999930608564, + "grad_norm": 1.5546886920928955, + "learning_rate": 0.00015975240829997584, + "loss": 0.7137, + "step": 18289 + }, + { + "epoch": 1.269169384497953, + "grad_norm": 1.6123876571655273, + "learning_rate": 0.0001597479029134258, + "loss": 0.5824, + "step": 18290 + }, + { + "epoch": 1.2692387759350496, + "grad_norm": 1.7731125354766846, + "learning_rate": 0.00015974339733825855, + "loss": 0.6853, + "step": 18291 + }, + { + "epoch": 1.2693081673721462, + "grad_norm": 1.8800045251846313, + "learning_rate": 0.00015973889157448832, + "loss": 0.8377, + "step": 18292 + }, + { + "epoch": 1.2693775588092429, + "grad_norm": 2.00993013381958, + "learning_rate": 0.00015973438562212933, + "loss": 0.7122, + "step": 18293 + }, + { + "epoch": 1.2694469502463397, + "grad_norm": 2.0898547172546387, + "learning_rate": 0.0001597298794811958, + "loss": 0.8265, + "step": 18294 + }, + { + "epoch": 1.2695163416834363, + "grad_norm": 1.9585801362991333, + "learning_rate": 0.00015972537315170202, + "loss": 0.7535, + "step": 18295 + }, + { + "epoch": 1.269585733120533, + "grad_norm": 1.8459290266036987, + "learning_rate": 0.00015972086663366213, + "loss": 0.6742, + "step": 18296 + }, + { + "epoch": 1.2696551245576295, + "grad_norm": 2.3001911640167236, + "learning_rate": 0.00015971635992709042, + "loss": 0.6935, + "step": 18297 + }, + { + "epoch": 1.2697245159947261, + "grad_norm": 1.851442813873291, + "learning_rate": 0.00015971185303200106, + "loss": 0.5891, + "step": 18298 + }, + { + "epoch": 1.269793907431823, + "grad_norm": 2.1611368656158447, + "learning_rate": 0.00015970734594840833, + "loss": 0.6151, + "step": 18299 + }, + { + "epoch": 1.2698632988689196, + "grad_norm": 2.7287657260894775, + "learning_rate": 0.00015970283867632642, + "loss": 0.742, + "step": 18300 + }, + { + "epoch": 1.2699326903060162, + "grad_norm": 2.2540500164031982, + "learning_rate": 0.0001596983312157696, + "loss": 0.7554, + "step": 18301 + }, + { + "epoch": 1.270002081743113, + "grad_norm": 2.168691396713257, + "learning_rate": 0.00015969382356675207, + "loss": 0.7296, + "step": 18302 + }, + { + "epoch": 1.2700714731802096, + "grad_norm": 1.7738473415374756, + "learning_rate": 0.00015968931572928803, + "loss": 0.6258, + "step": 18303 + }, + { + "epoch": 1.2701408646173062, + "grad_norm": 2.0190088748931885, + "learning_rate": 0.00015968480770339177, + "loss": 0.6015, + "step": 18304 + }, + { + "epoch": 1.2702102560544029, + "grad_norm": 2.0829224586486816, + "learning_rate": 0.0001596802994890775, + "loss": 0.7908, + "step": 18305 + }, + { + "epoch": 1.2702796474914995, + "grad_norm": 1.8760355710983276, + "learning_rate": 0.00015967579108635947, + "loss": 0.7738, + "step": 18306 + }, + { + "epoch": 1.2703490389285963, + "grad_norm": 1.6919658184051514, + "learning_rate": 0.00015967128249525184, + "loss": 0.7884, + "step": 18307 + }, + { + "epoch": 1.270418430365693, + "grad_norm": 1.941804051399231, + "learning_rate": 0.00015966677371576893, + "loss": 0.6278, + "step": 18308 + }, + { + "epoch": 1.2704878218027895, + "grad_norm": 1.6427892446517944, + "learning_rate": 0.00015966226474792495, + "loss": 0.584, + "step": 18309 + }, + { + "epoch": 1.2705572132398861, + "grad_norm": 2.0532054901123047, + "learning_rate": 0.00015965775559173407, + "loss": 0.7711, + "step": 18310 + }, + { + "epoch": 1.2706266046769827, + "grad_norm": 1.649863362312317, + "learning_rate": 0.0001596532462472106, + "loss": 0.5876, + "step": 18311 + }, + { + "epoch": 1.2706959961140796, + "grad_norm": 1.7627334594726562, + "learning_rate": 0.00015964873671436876, + "loss": 0.6533, + "step": 18312 + }, + { + "epoch": 1.2707653875511762, + "grad_norm": 2.4028873443603516, + "learning_rate": 0.0001596442269932228, + "loss": 0.6845, + "step": 18313 + }, + { + "epoch": 1.2708347789882728, + "grad_norm": 4.612305641174316, + "learning_rate": 0.00015963971708378687, + "loss": 1.0606, + "step": 18314 + }, + { + "epoch": 1.2709041704253696, + "grad_norm": 2.1785502433776855, + "learning_rate": 0.00015963520698607535, + "loss": 0.9056, + "step": 18315 + }, + { + "epoch": 1.2709735618624662, + "grad_norm": 1.697134256362915, + "learning_rate": 0.00015963069670010238, + "loss": 0.4936, + "step": 18316 + }, + { + "epoch": 1.2710429532995628, + "grad_norm": 1.2258868217468262, + "learning_rate": 0.0001596261862258822, + "loss": 0.5223, + "step": 18317 + }, + { + "epoch": 1.2711123447366595, + "grad_norm": 1.7194002866744995, + "learning_rate": 0.00015962167556342907, + "loss": 0.6088, + "step": 18318 + }, + { + "epoch": 1.271181736173756, + "grad_norm": 1.9900177717208862, + "learning_rate": 0.00015961716471275724, + "loss": 0.6058, + "step": 18319 + }, + { + "epoch": 1.271251127610853, + "grad_norm": 1.6968988180160522, + "learning_rate": 0.00015961265367388095, + "loss": 0.6684, + "step": 18320 + }, + { + "epoch": 1.2713205190479495, + "grad_norm": 2.286022186279297, + "learning_rate": 0.00015960814244681437, + "loss": 0.743, + "step": 18321 + }, + { + "epoch": 1.2713899104850461, + "grad_norm": 2.295696258544922, + "learning_rate": 0.00015960363103157183, + "loss": 0.8131, + "step": 18322 + }, + { + "epoch": 1.2714593019221427, + "grad_norm": 1.558897852897644, + "learning_rate": 0.0001595991194281676, + "loss": 0.4731, + "step": 18323 + }, + { + "epoch": 1.2715286933592393, + "grad_norm": 2.041201591491699, + "learning_rate": 0.0001595946076366158, + "loss": 0.5665, + "step": 18324 + }, + { + "epoch": 1.2715980847963362, + "grad_norm": 2.435535430908203, + "learning_rate": 0.00015959009565693076, + "loss": 0.8312, + "step": 18325 + }, + { + "epoch": 1.2716674762334328, + "grad_norm": 1.7155958414077759, + "learning_rate": 0.00015958558348912668, + "loss": 0.5702, + "step": 18326 + }, + { + "epoch": 1.2717368676705294, + "grad_norm": 2.123056650161743, + "learning_rate": 0.00015958107113321786, + "loss": 0.5933, + "step": 18327 + }, + { + "epoch": 1.2718062591076262, + "grad_norm": 1.6465810537338257, + "learning_rate": 0.00015957655858921848, + "loss": 0.6309, + "step": 18328 + }, + { + "epoch": 1.2718756505447228, + "grad_norm": 2.2055301666259766, + "learning_rate": 0.0001595720458571428, + "loss": 0.6628, + "step": 18329 + }, + { + "epoch": 1.2719450419818195, + "grad_norm": 2.4699175357818604, + "learning_rate": 0.00015956753293700513, + "loss": 0.7171, + "step": 18330 + }, + { + "epoch": 1.272014433418916, + "grad_norm": 1.7739166021347046, + "learning_rate": 0.00015956301982881965, + "loss": 0.9188, + "step": 18331 + }, + { + "epoch": 1.2720838248560127, + "grad_norm": 2.26497483253479, + "learning_rate": 0.00015955850653260063, + "loss": 0.8025, + "step": 18332 + }, + { + "epoch": 1.2721532162931095, + "grad_norm": 2.5433239936828613, + "learning_rate": 0.0001595539930483623, + "loss": 0.7753, + "step": 18333 + }, + { + "epoch": 1.2722226077302061, + "grad_norm": 1.7702635526657104, + "learning_rate": 0.00015954947937611897, + "loss": 0.6599, + "step": 18334 + }, + { + "epoch": 1.2722919991673027, + "grad_norm": 1.8832447528839111, + "learning_rate": 0.00015954496551588482, + "loss": 0.5466, + "step": 18335 + }, + { + "epoch": 1.2723613906043993, + "grad_norm": 1.9050103425979614, + "learning_rate": 0.00015954045146767412, + "loss": 0.7106, + "step": 18336 + }, + { + "epoch": 1.272430782041496, + "grad_norm": 1.7691402435302734, + "learning_rate": 0.0001595359372315011, + "loss": 0.691, + "step": 18337 + }, + { + "epoch": 1.2725001734785928, + "grad_norm": 2.2854602336883545, + "learning_rate": 0.00015953142280738004, + "loss": 0.7205, + "step": 18338 + }, + { + "epoch": 1.2725695649156894, + "grad_norm": 1.5813246965408325, + "learning_rate": 0.00015952690819532518, + "loss": 0.6708, + "step": 18339 + }, + { + "epoch": 1.272638956352786, + "grad_norm": 1.551544189453125, + "learning_rate": 0.00015952239339535082, + "loss": 0.6506, + "step": 18340 + }, + { + "epoch": 1.2727083477898828, + "grad_norm": 1.8544230461120605, + "learning_rate": 0.00015951787840747113, + "loss": 0.5848, + "step": 18341 + }, + { + "epoch": 1.2727777392269795, + "grad_norm": 2.5035653114318848, + "learning_rate": 0.0001595133632317004, + "loss": 0.644, + "step": 18342 + }, + { + "epoch": 1.272847130664076, + "grad_norm": 1.8818778991699219, + "learning_rate": 0.0001595088478680529, + "loss": 0.6138, + "step": 18343 + }, + { + "epoch": 1.2729165221011727, + "grad_norm": 2.0822505950927734, + "learning_rate": 0.0001595043323165429, + "loss": 0.8509, + "step": 18344 + }, + { + "epoch": 1.2729859135382693, + "grad_norm": 2.082841396331787, + "learning_rate": 0.00015949981657718457, + "loss": 0.6665, + "step": 18345 + }, + { + "epoch": 1.2730553049753661, + "grad_norm": 1.5415269136428833, + "learning_rate": 0.00015949530064999226, + "loss": 0.5823, + "step": 18346 + }, + { + "epoch": 1.2731246964124627, + "grad_norm": 1.7355109453201294, + "learning_rate": 0.0001594907845349802, + "loss": 0.6312, + "step": 18347 + }, + { + "epoch": 1.2731940878495593, + "grad_norm": 1.3355844020843506, + "learning_rate": 0.0001594862682321626, + "loss": 0.5817, + "step": 18348 + }, + { + "epoch": 1.273263479286656, + "grad_norm": 1.7411247491836548, + "learning_rate": 0.00015948175174155374, + "loss": 0.6498, + "step": 18349 + }, + { + "epoch": 1.2733328707237526, + "grad_norm": 1.5501958131790161, + "learning_rate": 0.00015947723506316795, + "loss": 0.6592, + "step": 18350 + }, + { + "epoch": 1.2734022621608494, + "grad_norm": 1.8116459846496582, + "learning_rate": 0.0001594727181970194, + "loss": 0.5813, + "step": 18351 + }, + { + "epoch": 1.273471653597946, + "grad_norm": 1.6086360216140747, + "learning_rate": 0.00015946820114312238, + "loss": 0.3922, + "step": 18352 + }, + { + "epoch": 1.2735410450350426, + "grad_norm": 1.3050225973129272, + "learning_rate": 0.00015946368390149114, + "loss": 0.6791, + "step": 18353 + }, + { + "epoch": 1.2736104364721395, + "grad_norm": 1.4399306774139404, + "learning_rate": 0.00015945916647213995, + "loss": 0.6794, + "step": 18354 + }, + { + "epoch": 1.273679827909236, + "grad_norm": 1.868776559829712, + "learning_rate": 0.0001594546488550831, + "loss": 0.6665, + "step": 18355 + }, + { + "epoch": 1.2737492193463327, + "grad_norm": 1.7894227504730225, + "learning_rate": 0.00015945013105033478, + "loss": 0.6421, + "step": 18356 + }, + { + "epoch": 1.2738186107834293, + "grad_norm": 1.2681511640548706, + "learning_rate": 0.00015944561305790932, + "loss": 0.5068, + "step": 18357 + }, + { + "epoch": 1.273888002220526, + "grad_norm": 1.8369665145874023, + "learning_rate": 0.00015944109487782093, + "loss": 0.5233, + "step": 18358 + }, + { + "epoch": 1.2739573936576227, + "grad_norm": 1.2079894542694092, + "learning_rate": 0.00015943657651008394, + "loss": 0.4981, + "step": 18359 + }, + { + "epoch": 1.2740267850947193, + "grad_norm": 1.5706706047058105, + "learning_rate": 0.00015943205795471252, + "loss": 0.4637, + "step": 18360 + }, + { + "epoch": 1.274096176531816, + "grad_norm": 1.4186798334121704, + "learning_rate": 0.00015942753921172105, + "loss": 0.4812, + "step": 18361 + }, + { + "epoch": 1.2741655679689126, + "grad_norm": 1.5461386442184448, + "learning_rate": 0.00015942302028112369, + "loss": 0.6887, + "step": 18362 + }, + { + "epoch": 1.2742349594060092, + "grad_norm": 1.7985020875930786, + "learning_rate": 0.00015941850116293478, + "loss": 0.6399, + "step": 18363 + }, + { + "epoch": 1.274304350843106, + "grad_norm": 1.67318594455719, + "learning_rate": 0.00015941398185716853, + "loss": 0.7552, + "step": 18364 + }, + { + "epoch": 1.2743737422802026, + "grad_norm": 1.9766710996627808, + "learning_rate": 0.00015940946236383922, + "loss": 0.7685, + "step": 18365 + }, + { + "epoch": 1.2744431337172992, + "grad_norm": 2.191533088684082, + "learning_rate": 0.00015940494268296119, + "loss": 0.6824, + "step": 18366 + }, + { + "epoch": 1.274512525154396, + "grad_norm": 1.6098366975784302, + "learning_rate": 0.0001594004228145486, + "loss": 0.5876, + "step": 18367 + }, + { + "epoch": 1.2745819165914927, + "grad_norm": 2.3897993564605713, + "learning_rate": 0.00015939590275861577, + "loss": 0.7914, + "step": 18368 + }, + { + "epoch": 1.2746513080285893, + "grad_norm": 1.9662504196166992, + "learning_rate": 0.00015939138251517697, + "loss": 0.5336, + "step": 18369 + }, + { + "epoch": 1.274720699465686, + "grad_norm": 1.5106298923492432, + "learning_rate": 0.00015938686208424646, + "loss": 0.8447, + "step": 18370 + }, + { + "epoch": 1.2747900909027825, + "grad_norm": 1.8626434803009033, + "learning_rate": 0.00015938234146583856, + "loss": 0.8152, + "step": 18371 + }, + { + "epoch": 1.2748594823398793, + "grad_norm": 2.1904337406158447, + "learning_rate": 0.00015937782065996743, + "loss": 1.0856, + "step": 18372 + }, + { + "epoch": 1.274928873776976, + "grad_norm": 1.6731386184692383, + "learning_rate": 0.00015937329966664743, + "loss": 0.6609, + "step": 18373 + }, + { + "epoch": 1.2749982652140726, + "grad_norm": 1.4405018091201782, + "learning_rate": 0.00015936877848589284, + "loss": 0.5873, + "step": 18374 + }, + { + "epoch": 1.2750676566511692, + "grad_norm": 1.5207526683807373, + "learning_rate": 0.0001593642571177179, + "loss": 0.5854, + "step": 18375 + }, + { + "epoch": 1.2751370480882658, + "grad_norm": 1.9598596096038818, + "learning_rate": 0.00015935973556213687, + "loss": 0.6601, + "step": 18376 + }, + { + "epoch": 1.2752064395253626, + "grad_norm": 2.063560962677002, + "learning_rate": 0.00015935521381916403, + "loss": 0.8988, + "step": 18377 + }, + { + "epoch": 1.2752758309624592, + "grad_norm": 1.5431745052337646, + "learning_rate": 0.00015935069188881368, + "loss": 0.7434, + "step": 18378 + }, + { + "epoch": 1.2753452223995558, + "grad_norm": 1.567090392112732, + "learning_rate": 0.0001593461697711001, + "loss": 0.5019, + "step": 18379 + }, + { + "epoch": 1.2754146138366527, + "grad_norm": 1.8663995265960693, + "learning_rate": 0.00015934164746603754, + "loss": 0.5162, + "step": 18380 + }, + { + "epoch": 1.2754840052737493, + "grad_norm": 1.6068373918533325, + "learning_rate": 0.00015933712497364025, + "loss": 0.5655, + "step": 18381 + }, + { + "epoch": 1.2755533967108459, + "grad_norm": 1.863309383392334, + "learning_rate": 0.0001593326022939226, + "loss": 0.6048, + "step": 18382 + }, + { + "epoch": 1.2756227881479425, + "grad_norm": 1.9931546449661255, + "learning_rate": 0.00015932807942689876, + "loss": 0.8034, + "step": 18383 + }, + { + "epoch": 1.275692179585039, + "grad_norm": 1.685265064239502, + "learning_rate": 0.00015932355637258307, + "loss": 0.6308, + "step": 18384 + }, + { + "epoch": 1.275761571022136, + "grad_norm": 2.07100248336792, + "learning_rate": 0.0001593190331309898, + "loss": 0.9153, + "step": 18385 + }, + { + "epoch": 1.2758309624592326, + "grad_norm": 1.9884527921676636, + "learning_rate": 0.00015931450970213323, + "loss": 0.5656, + "step": 18386 + }, + { + "epoch": 1.2759003538963292, + "grad_norm": 2.195704460144043, + "learning_rate": 0.0001593099860860276, + "loss": 0.8569, + "step": 18387 + }, + { + "epoch": 1.2759697453334258, + "grad_norm": 1.8821852207183838, + "learning_rate": 0.00015930546228268728, + "loss": 0.6192, + "step": 18388 + }, + { + "epoch": 1.2760391367705224, + "grad_norm": 1.6051225662231445, + "learning_rate": 0.00015930093829212647, + "loss": 0.5308, + "step": 18389 + }, + { + "epoch": 1.2761085282076192, + "grad_norm": 1.6821812391281128, + "learning_rate": 0.00015929641411435944, + "loss": 0.7217, + "step": 18390 + }, + { + "epoch": 1.2761779196447158, + "grad_norm": 1.7557324171066284, + "learning_rate": 0.00015929188974940056, + "loss": 0.6634, + "step": 18391 + }, + { + "epoch": 1.2762473110818124, + "grad_norm": 1.3477120399475098, + "learning_rate": 0.0001592873651972641, + "loss": 0.6667, + "step": 18392 + }, + { + "epoch": 1.2763167025189093, + "grad_norm": 1.6983078718185425, + "learning_rate": 0.00015928284045796422, + "loss": 0.6681, + "step": 18393 + }, + { + "epoch": 1.2763860939560059, + "grad_norm": 1.9512817859649658, + "learning_rate": 0.00015927831553151533, + "loss": 0.6039, + "step": 18394 + }, + { + "epoch": 1.2764554853931025, + "grad_norm": 1.9422017335891724, + "learning_rate": 0.00015927379041793169, + "loss": 0.7412, + "step": 18395 + }, + { + "epoch": 1.276524876830199, + "grad_norm": 1.5035935640335083, + "learning_rate": 0.00015926926511722754, + "loss": 0.6471, + "step": 18396 + }, + { + "epoch": 1.2765942682672957, + "grad_norm": 2.315786600112915, + "learning_rate": 0.00015926473962941723, + "loss": 1.1556, + "step": 18397 + }, + { + "epoch": 1.2766636597043926, + "grad_norm": 1.4383485317230225, + "learning_rate": 0.00015926021395451496, + "loss": 0.516, + "step": 18398 + }, + { + "epoch": 1.2767330511414892, + "grad_norm": 1.6986395120620728, + "learning_rate": 0.0001592556880925351, + "loss": 0.7013, + "step": 18399 + }, + { + "epoch": 1.2768024425785858, + "grad_norm": 1.9296208620071411, + "learning_rate": 0.00015925116204349192, + "loss": 0.5645, + "step": 18400 + }, + { + "epoch": 1.2768718340156824, + "grad_norm": 1.5313136577606201, + "learning_rate": 0.00015924663580739966, + "loss": 0.5335, + "step": 18401 + }, + { + "epoch": 1.276941225452779, + "grad_norm": 1.6933391094207764, + "learning_rate": 0.00015924210938427272, + "loss": 0.5164, + "step": 18402 + }, + { + "epoch": 1.2770106168898758, + "grad_norm": 1.4008675813674927, + "learning_rate": 0.00015923758277412527, + "loss": 0.4549, + "step": 18403 + }, + { + "epoch": 1.2770800083269724, + "grad_norm": 2.1881473064422607, + "learning_rate": 0.0001592330559769716, + "loss": 0.6108, + "step": 18404 + }, + { + "epoch": 1.277149399764069, + "grad_norm": 1.4983607530593872, + "learning_rate": 0.0001592285289928261, + "loss": 0.6224, + "step": 18405 + }, + { + "epoch": 1.2772187912011659, + "grad_norm": 1.9877479076385498, + "learning_rate": 0.00015922400182170302, + "loss": 0.7129, + "step": 18406 + }, + { + "epoch": 1.2772881826382625, + "grad_norm": 1.8027054071426392, + "learning_rate": 0.00015921947446361658, + "loss": 0.6192, + "step": 18407 + }, + { + "epoch": 1.277357574075359, + "grad_norm": 1.7903870344161987, + "learning_rate": 0.00015921494691858114, + "loss": 0.446, + "step": 18408 + }, + { + "epoch": 1.2774269655124557, + "grad_norm": 1.9454917907714844, + "learning_rate": 0.000159210419186611, + "loss": 0.8256, + "step": 18409 + }, + { + "epoch": 1.2774963569495523, + "grad_norm": 1.9742811918258667, + "learning_rate": 0.00015920589126772045, + "loss": 0.6362, + "step": 18410 + }, + { + "epoch": 1.2775657483866492, + "grad_norm": 1.8291994333267212, + "learning_rate": 0.00015920136316192375, + "loss": 0.7589, + "step": 18411 + }, + { + "epoch": 1.2776351398237458, + "grad_norm": 1.7978967428207397, + "learning_rate": 0.00015919683486923525, + "loss": 0.7114, + "step": 18412 + }, + { + "epoch": 1.2777045312608424, + "grad_norm": 1.9964492321014404, + "learning_rate": 0.00015919230638966915, + "loss": 0.7904, + "step": 18413 + }, + { + "epoch": 1.2777739226979392, + "grad_norm": 1.843389630317688, + "learning_rate": 0.00015918777772323985, + "loss": 0.7927, + "step": 18414 + }, + { + "epoch": 1.2778433141350356, + "grad_norm": 2.0055034160614014, + "learning_rate": 0.00015918324886996158, + "loss": 0.5398, + "step": 18415 + }, + { + "epoch": 1.2779127055721324, + "grad_norm": 1.7747678756713867, + "learning_rate": 0.0001591787198298487, + "loss": 0.7455, + "step": 18416 + }, + { + "epoch": 1.277982097009229, + "grad_norm": 1.6137964725494385, + "learning_rate": 0.00015917419060291544, + "loss": 0.4491, + "step": 18417 + }, + { + "epoch": 1.2780514884463257, + "grad_norm": 1.9056742191314697, + "learning_rate": 0.0001591696611891761, + "loss": 0.6186, + "step": 18418 + }, + { + "epoch": 1.2781208798834225, + "grad_norm": 5.624569892883301, + "learning_rate": 0.00015916513158864505, + "loss": 0.6464, + "step": 18419 + }, + { + "epoch": 1.278190271320519, + "grad_norm": 1.4946376085281372, + "learning_rate": 0.0001591606018013365, + "loss": 0.4218, + "step": 18420 + }, + { + "epoch": 1.2782596627576157, + "grad_norm": 3.977605104446411, + "learning_rate": 0.00015915607182726484, + "loss": 0.8877, + "step": 18421 + }, + { + "epoch": 1.2783290541947123, + "grad_norm": 1.4223287105560303, + "learning_rate": 0.0001591515416664443, + "loss": 0.712, + "step": 18422 + }, + { + "epoch": 1.278398445631809, + "grad_norm": 1.8115994930267334, + "learning_rate": 0.0001591470113188892, + "loss": 0.6433, + "step": 18423 + }, + { + "epoch": 1.2784678370689058, + "grad_norm": 1.6066213846206665, + "learning_rate": 0.00015914248078461384, + "loss": 0.5435, + "step": 18424 + }, + { + "epoch": 1.2785372285060024, + "grad_norm": 1.7034722566604614, + "learning_rate": 0.00015913795006363253, + "loss": 0.7324, + "step": 18425 + }, + { + "epoch": 1.278606619943099, + "grad_norm": 1.8265093564987183, + "learning_rate": 0.0001591334191559596, + "loss": 0.5864, + "step": 18426 + }, + { + "epoch": 1.2786760113801958, + "grad_norm": 2.0176198482513428, + "learning_rate": 0.00015912888806160933, + "loss": 0.7078, + "step": 18427 + }, + { + "epoch": 1.2787454028172922, + "grad_norm": 1.9329593181610107, + "learning_rate": 0.000159124356780596, + "loss": 0.7032, + "step": 18428 + }, + { + "epoch": 1.278814794254389, + "grad_norm": 2.418104648590088, + "learning_rate": 0.00015911982531293393, + "loss": 1.0366, + "step": 18429 + }, + { + "epoch": 1.2788841856914857, + "grad_norm": 2.052995204925537, + "learning_rate": 0.00015911529365863746, + "loss": 0.5721, + "step": 18430 + }, + { + "epoch": 1.2789535771285823, + "grad_norm": 1.742852807044983, + "learning_rate": 0.00015911076181772084, + "loss": 0.6662, + "step": 18431 + }, + { + "epoch": 1.279022968565679, + "grad_norm": 1.6423633098602295, + "learning_rate": 0.0001591062297901984, + "loss": 0.8372, + "step": 18432 + }, + { + "epoch": 1.2790923600027757, + "grad_norm": 2.924603223800659, + "learning_rate": 0.0001591016975760845, + "loss": 0.6307, + "step": 18433 + }, + { + "epoch": 1.2791617514398723, + "grad_norm": 1.6157681941986084, + "learning_rate": 0.00015909716517539333, + "loss": 0.4327, + "step": 18434 + }, + { + "epoch": 1.279231142876969, + "grad_norm": 1.6473039388656616, + "learning_rate": 0.0001590926325881393, + "loss": 0.5432, + "step": 18435 + }, + { + "epoch": 1.2793005343140655, + "grad_norm": 1.6550517082214355, + "learning_rate": 0.00015908809981433667, + "loss": 0.6888, + "step": 18436 + }, + { + "epoch": 1.2793699257511624, + "grad_norm": 1.6245262622833252, + "learning_rate": 0.0001590835668539998, + "loss": 0.6811, + "step": 18437 + }, + { + "epoch": 1.279439317188259, + "grad_norm": 1.9417253732681274, + "learning_rate": 0.0001590790337071429, + "loss": 0.6581, + "step": 18438 + }, + { + "epoch": 1.2795087086253556, + "grad_norm": 1.9296268224716187, + "learning_rate": 0.0001590745003737804, + "loss": 0.6427, + "step": 18439 + }, + { + "epoch": 1.2795781000624524, + "grad_norm": 1.6190831661224365, + "learning_rate": 0.00015906996685392656, + "loss": 0.5245, + "step": 18440 + }, + { + "epoch": 1.2796474914995488, + "grad_norm": 2.1048738956451416, + "learning_rate": 0.00015906543314759563, + "loss": 0.7462, + "step": 18441 + }, + { + "epoch": 1.2797168829366456, + "grad_norm": 1.3947093486785889, + "learning_rate": 0.00015906089925480204, + "loss": 0.5211, + "step": 18442 + }, + { + "epoch": 1.2797862743737423, + "grad_norm": 1.503280520439148, + "learning_rate": 0.00015905636517556001, + "loss": 0.6871, + "step": 18443 + }, + { + "epoch": 1.2798556658108389, + "grad_norm": 1.2349830865859985, + "learning_rate": 0.0001590518309098839, + "loss": 0.4882, + "step": 18444 + }, + { + "epoch": 1.2799250572479357, + "grad_norm": 1.968070149421692, + "learning_rate": 0.000159047296457788, + "loss": 0.6988, + "step": 18445 + }, + { + "epoch": 1.2799944486850323, + "grad_norm": 1.810112476348877, + "learning_rate": 0.00015904276181928666, + "loss": 0.6398, + "step": 18446 + }, + { + "epoch": 1.280063840122129, + "grad_norm": 1.3938027620315552, + "learning_rate": 0.00015903822699439414, + "loss": 0.5702, + "step": 18447 + }, + { + "epoch": 1.2801332315592255, + "grad_norm": 1.9335389137268066, + "learning_rate": 0.00015903369198312482, + "loss": 0.4701, + "step": 18448 + }, + { + "epoch": 1.2802026229963221, + "grad_norm": 2.2062201499938965, + "learning_rate": 0.00015902915678549294, + "loss": 0.5856, + "step": 18449 + }, + { + "epoch": 1.280272014433419, + "grad_norm": 1.5036704540252686, + "learning_rate": 0.00015902462140151294, + "loss": 0.5241, + "step": 18450 + }, + { + "epoch": 1.2803414058705156, + "grad_norm": 2.0240514278411865, + "learning_rate": 0.00015902008583119898, + "loss": 0.5217, + "step": 18451 + }, + { + "epoch": 1.2804107973076122, + "grad_norm": 1.7174596786499023, + "learning_rate": 0.00015901555007456547, + "loss": 0.5992, + "step": 18452 + }, + { + "epoch": 1.280480188744709, + "grad_norm": 1.6387648582458496, + "learning_rate": 0.00015901101413162672, + "loss": 0.551, + "step": 18453 + }, + { + "epoch": 1.2805495801818056, + "grad_norm": 1.810799479484558, + "learning_rate": 0.00015900647800239708, + "loss": 0.6067, + "step": 18454 + }, + { + "epoch": 1.2806189716189023, + "grad_norm": 1.8197194337844849, + "learning_rate": 0.0001590019416868908, + "loss": 0.7428, + "step": 18455 + }, + { + "epoch": 1.2806883630559989, + "grad_norm": 1.9417048692703247, + "learning_rate": 0.00015899740518512224, + "loss": 0.8102, + "step": 18456 + }, + { + "epoch": 1.2807577544930955, + "grad_norm": 2.7038745880126953, + "learning_rate": 0.00015899286849710572, + "loss": 0.9573, + "step": 18457 + }, + { + "epoch": 1.2808271459301923, + "grad_norm": 1.7403504848480225, + "learning_rate": 0.00015898833162285558, + "loss": 0.6268, + "step": 18458 + }, + { + "epoch": 1.280896537367289, + "grad_norm": 1.8501778841018677, + "learning_rate": 0.00015898379456238607, + "loss": 0.7297, + "step": 18459 + }, + { + "epoch": 1.2809659288043855, + "grad_norm": 2.424020767211914, + "learning_rate": 0.0001589792573157116, + "loss": 0.639, + "step": 18460 + }, + { + "epoch": 1.2810353202414821, + "grad_norm": 1.6705949306488037, + "learning_rate": 0.00015897471988284646, + "loss": 0.6254, + "step": 18461 + }, + { + "epoch": 1.2811047116785788, + "grad_norm": 1.6385830640792847, + "learning_rate": 0.00015897018226380496, + "loss": 0.5938, + "step": 18462 + }, + { + "epoch": 1.2811741031156756, + "grad_norm": 1.395358681678772, + "learning_rate": 0.00015896564445860145, + "loss": 0.5523, + "step": 18463 + }, + { + "epoch": 1.2812434945527722, + "grad_norm": 1.893692135810852, + "learning_rate": 0.00015896110646725022, + "loss": 0.7044, + "step": 18464 + }, + { + "epoch": 1.2813128859898688, + "grad_norm": 1.7497515678405762, + "learning_rate": 0.00015895656828976562, + "loss": 0.7752, + "step": 18465 + }, + { + "epoch": 1.2813822774269656, + "grad_norm": 2.6473448276519775, + "learning_rate": 0.000158952029926162, + "loss": 0.6815, + "step": 18466 + }, + { + "epoch": 1.2814516688640623, + "grad_norm": 1.8331481218338013, + "learning_rate": 0.00015894749137645363, + "loss": 0.6137, + "step": 18467 + }, + { + "epoch": 1.2815210603011589, + "grad_norm": 2.3133482933044434, + "learning_rate": 0.00015894295264065492, + "loss": 0.7388, + "step": 18468 + }, + { + "epoch": 1.2815904517382555, + "grad_norm": 1.7486315965652466, + "learning_rate": 0.00015893841371878006, + "loss": 0.5399, + "step": 18469 + }, + { + "epoch": 1.281659843175352, + "grad_norm": 2.017996072769165, + "learning_rate": 0.00015893387461084354, + "loss": 0.7514, + "step": 18470 + }, + { + "epoch": 1.281729234612449, + "grad_norm": 1.8980146646499634, + "learning_rate": 0.00015892933531685957, + "loss": 0.9252, + "step": 18471 + }, + { + "epoch": 1.2817986260495455, + "grad_norm": 1.7246721982955933, + "learning_rate": 0.00015892479583684256, + "loss": 0.7021, + "step": 18472 + }, + { + "epoch": 1.2818680174866421, + "grad_norm": 1.398397445678711, + "learning_rate": 0.0001589202561708068, + "loss": 0.5222, + "step": 18473 + }, + { + "epoch": 1.2819374089237388, + "grad_norm": 2.10068416595459, + "learning_rate": 0.00015891571631876662, + "loss": 0.705, + "step": 18474 + }, + { + "epoch": 1.2820068003608354, + "grad_norm": 1.7001832723617554, + "learning_rate": 0.00015891117628073637, + "loss": 0.6112, + "step": 18475 + }, + { + "epoch": 1.2820761917979322, + "grad_norm": 2.269477128982544, + "learning_rate": 0.00015890663605673034, + "loss": 0.7357, + "step": 18476 + }, + { + "epoch": 1.2821455832350288, + "grad_norm": 1.5741816759109497, + "learning_rate": 0.00015890209564676293, + "loss": 0.6486, + "step": 18477 + }, + { + "epoch": 1.2822149746721254, + "grad_norm": 1.8789366483688354, + "learning_rate": 0.00015889755505084843, + "loss": 0.7863, + "step": 18478 + }, + { + "epoch": 1.2822843661092223, + "grad_norm": 1.7003360986709595, + "learning_rate": 0.0001588930142690012, + "loss": 0.6447, + "step": 18479 + }, + { + "epoch": 1.2823537575463189, + "grad_norm": 1.5832451581954956, + "learning_rate": 0.0001588884733012355, + "loss": 0.6572, + "step": 18480 + }, + { + "epoch": 1.2824231489834155, + "grad_norm": 1.561470627784729, + "learning_rate": 0.00015888393214756575, + "loss": 0.5693, + "step": 18481 + }, + { + "epoch": 1.282492540420512, + "grad_norm": 1.8980388641357422, + "learning_rate": 0.00015887939080800627, + "loss": 0.6551, + "step": 18482 + }, + { + "epoch": 1.2825619318576087, + "grad_norm": 1.8213565349578857, + "learning_rate": 0.0001588748492825714, + "loss": 0.4933, + "step": 18483 + }, + { + "epoch": 1.2826313232947055, + "grad_norm": 1.762007713317871, + "learning_rate": 0.00015887030757127542, + "loss": 0.7316, + "step": 18484 + }, + { + "epoch": 1.2827007147318021, + "grad_norm": 2.026019811630249, + "learning_rate": 0.00015886576567413272, + "loss": 0.7519, + "step": 18485 + }, + { + "epoch": 1.2827701061688987, + "grad_norm": 1.8265222311019897, + "learning_rate": 0.00015886122359115766, + "loss": 0.6129, + "step": 18486 + }, + { + "epoch": 1.2828394976059954, + "grad_norm": 2.3248329162597656, + "learning_rate": 0.00015885668132236452, + "loss": 0.8945, + "step": 18487 + }, + { + "epoch": 1.282908889043092, + "grad_norm": 2.115983724594116, + "learning_rate": 0.00015885213886776764, + "loss": 0.9213, + "step": 18488 + }, + { + "epoch": 1.2829782804801888, + "grad_norm": 1.722561001777649, + "learning_rate": 0.00015884759622738142, + "loss": 0.6545, + "step": 18489 + }, + { + "epoch": 1.2830476719172854, + "grad_norm": 2.2824623584747314, + "learning_rate": 0.00015884305340122014, + "loss": 0.6832, + "step": 18490 + }, + { + "epoch": 1.283117063354382, + "grad_norm": 1.6222476959228516, + "learning_rate": 0.0001588385103892982, + "loss": 0.7408, + "step": 18491 + }, + { + "epoch": 1.2831864547914789, + "grad_norm": 1.6718873977661133, + "learning_rate": 0.00015883396719162992, + "loss": 0.6938, + "step": 18492 + }, + { + "epoch": 1.2832558462285755, + "grad_norm": 1.63229501247406, + "learning_rate": 0.0001588294238082296, + "loss": 0.5131, + "step": 18493 + }, + { + "epoch": 1.283325237665672, + "grad_norm": 2.289202928543091, + "learning_rate": 0.0001588248802391116, + "loss": 0.67, + "step": 18494 + }, + { + "epoch": 1.2833946291027687, + "grad_norm": 1.47176194190979, + "learning_rate": 0.0001588203364842903, + "loss": 0.5252, + "step": 18495 + }, + { + "epoch": 1.2834640205398653, + "grad_norm": 2.986335039138794, + "learning_rate": 0.00015881579254378004, + "loss": 0.918, + "step": 18496 + }, + { + "epoch": 1.2835334119769621, + "grad_norm": 2.789106607437134, + "learning_rate": 0.0001588112484175951, + "loss": 0.7073, + "step": 18497 + }, + { + "epoch": 1.2836028034140587, + "grad_norm": 1.5546263456344604, + "learning_rate": 0.0001588067041057499, + "loss": 0.4758, + "step": 18498 + }, + { + "epoch": 1.2836721948511554, + "grad_norm": 1.4599287509918213, + "learning_rate": 0.00015880215960825878, + "loss": 0.4358, + "step": 18499 + }, + { + "epoch": 1.283741586288252, + "grad_norm": 2.2166218757629395, + "learning_rate": 0.00015879761492513606, + "loss": 0.8534, + "step": 18500 + }, + { + "epoch": 1.2838109777253486, + "grad_norm": 1.7145682573318481, + "learning_rate": 0.00015879307005639605, + "loss": 0.5646, + "step": 18501 + }, + { + "epoch": 1.2838803691624454, + "grad_norm": 2.293766498565674, + "learning_rate": 0.00015878852500205318, + "loss": 0.8059, + "step": 18502 + }, + { + "epoch": 1.283949760599542, + "grad_norm": 1.7722594738006592, + "learning_rate": 0.00015878397976212174, + "loss": 0.593, + "step": 18503 + }, + { + "epoch": 1.2840191520366386, + "grad_norm": 2.105447292327881, + "learning_rate": 0.0001587794343366161, + "loss": 0.7381, + "step": 18504 + }, + { + "epoch": 1.2840885434737355, + "grad_norm": 1.762438178062439, + "learning_rate": 0.0001587748887255506, + "loss": 0.6718, + "step": 18505 + }, + { + "epoch": 1.284157934910832, + "grad_norm": 1.850932240486145, + "learning_rate": 0.00015877034292893958, + "loss": 0.8539, + "step": 18506 + }, + { + "epoch": 1.2842273263479287, + "grad_norm": 1.8198246955871582, + "learning_rate": 0.00015876579694679746, + "loss": 0.5389, + "step": 18507 + }, + { + "epoch": 1.2842967177850253, + "grad_norm": 1.4653834104537964, + "learning_rate": 0.00015876125077913847, + "loss": 0.5869, + "step": 18508 + }, + { + "epoch": 1.284366109222122, + "grad_norm": 1.6048810482025146, + "learning_rate": 0.00015875670442597707, + "loss": 0.623, + "step": 18509 + }, + { + "epoch": 1.2844355006592187, + "grad_norm": 2.2174251079559326, + "learning_rate": 0.00015875215788732758, + "loss": 0.6787, + "step": 18510 + }, + { + "epoch": 1.2845048920963154, + "grad_norm": 1.7316927909851074, + "learning_rate": 0.00015874761116320432, + "loss": 0.6231, + "step": 18511 + }, + { + "epoch": 1.284574283533412, + "grad_norm": 2.437617063522339, + "learning_rate": 0.00015874306425362166, + "loss": 0.4188, + "step": 18512 + }, + { + "epoch": 1.2846436749705086, + "grad_norm": 1.9147371053695679, + "learning_rate": 0.00015873851715859398, + "loss": 0.6219, + "step": 18513 + }, + { + "epoch": 1.2847130664076052, + "grad_norm": 1.3613554239273071, + "learning_rate": 0.0001587339698781356, + "loss": 0.4992, + "step": 18514 + }, + { + "epoch": 1.284782457844702, + "grad_norm": 2.305251121520996, + "learning_rate": 0.0001587294224122609, + "loss": 0.4011, + "step": 18515 + }, + { + "epoch": 1.2848518492817986, + "grad_norm": 1.5735067129135132, + "learning_rate": 0.00015872487476098422, + "loss": 0.4779, + "step": 18516 + }, + { + "epoch": 1.2849212407188952, + "grad_norm": 1.946276068687439, + "learning_rate": 0.00015872032692431994, + "loss": 0.7165, + "step": 18517 + }, + { + "epoch": 1.284990632155992, + "grad_norm": 1.4380035400390625, + "learning_rate": 0.0001587157789022824, + "loss": 0.6113, + "step": 18518 + }, + { + "epoch": 1.2850600235930887, + "grad_norm": 1.6048142910003662, + "learning_rate": 0.00015871123069488596, + "loss": 0.5592, + "step": 18519 + }, + { + "epoch": 1.2851294150301853, + "grad_norm": 1.6075164079666138, + "learning_rate": 0.00015870668230214498, + "loss": 0.5637, + "step": 18520 + }, + { + "epoch": 1.285198806467282, + "grad_norm": 2.0273797512054443, + "learning_rate": 0.00015870213372407382, + "loss": 0.5592, + "step": 18521 + }, + { + "epoch": 1.2852681979043785, + "grad_norm": 2.0210137367248535, + "learning_rate": 0.00015869758496068678, + "loss": 0.9282, + "step": 18522 + }, + { + "epoch": 1.2853375893414754, + "grad_norm": 1.3255913257598877, + "learning_rate": 0.00015869303601199835, + "loss": 0.5619, + "step": 18523 + }, + { + "epoch": 1.285406980778572, + "grad_norm": 1.895100712776184, + "learning_rate": 0.00015868848687802274, + "loss": 0.9083, + "step": 18524 + }, + { + "epoch": 1.2854763722156686, + "grad_norm": 1.8083077669143677, + "learning_rate": 0.00015868393755877444, + "loss": 0.8823, + "step": 18525 + }, + { + "epoch": 1.2855457636527652, + "grad_norm": 2.212083339691162, + "learning_rate": 0.00015867938805426773, + "loss": 0.5996, + "step": 18526 + }, + { + "epoch": 1.2856151550898618, + "grad_norm": 1.8691017627716064, + "learning_rate": 0.00015867483836451703, + "loss": 0.4757, + "step": 18527 + }, + { + "epoch": 1.2856845465269586, + "grad_norm": 1.35400390625, + "learning_rate": 0.00015867028848953665, + "loss": 0.5122, + "step": 18528 + }, + { + "epoch": 1.2857539379640552, + "grad_norm": 2.3600993156433105, + "learning_rate": 0.00015866573842934098, + "loss": 0.6847, + "step": 18529 + }, + { + "epoch": 1.2858233294011518, + "grad_norm": 1.9506257772445679, + "learning_rate": 0.00015866118818394438, + "loss": 0.5308, + "step": 18530 + }, + { + "epoch": 1.2858927208382487, + "grad_norm": 1.4946993589401245, + "learning_rate": 0.0001586566377533612, + "loss": 0.5675, + "step": 18531 + }, + { + "epoch": 1.2859621122753453, + "grad_norm": 1.2321255207061768, + "learning_rate": 0.00015865208713760584, + "loss": 0.424, + "step": 18532 + }, + { + "epoch": 1.286031503712442, + "grad_norm": 1.3028273582458496, + "learning_rate": 0.00015864753633669266, + "loss": 0.6096, + "step": 18533 + }, + { + "epoch": 1.2861008951495385, + "grad_norm": 2.7890195846557617, + "learning_rate": 0.00015864298535063598, + "loss": 0.8886, + "step": 18534 + }, + { + "epoch": 1.2861702865866351, + "grad_norm": 1.8546760082244873, + "learning_rate": 0.00015863843417945024, + "loss": 0.7158, + "step": 18535 + }, + { + "epoch": 1.286239678023732, + "grad_norm": 1.655853271484375, + "learning_rate": 0.00015863388282314972, + "loss": 0.7551, + "step": 18536 + }, + { + "epoch": 1.2863090694608286, + "grad_norm": 2.095188617706299, + "learning_rate": 0.0001586293312817489, + "loss": 0.6316, + "step": 18537 + }, + { + "epoch": 1.2863784608979252, + "grad_norm": 2.0125129222869873, + "learning_rate": 0.000158624779555262, + "loss": 0.6301, + "step": 18538 + }, + { + "epoch": 1.2864478523350218, + "grad_norm": 1.7834827899932861, + "learning_rate": 0.0001586202276437035, + "loss": 0.6459, + "step": 18539 + }, + { + "epoch": 1.2865172437721184, + "grad_norm": 2.1435835361480713, + "learning_rate": 0.00015861567554708778, + "loss": 0.8923, + "step": 18540 + }, + { + "epoch": 1.2865866352092152, + "grad_norm": 2.579660177230835, + "learning_rate": 0.00015861112326542913, + "loss": 0.566, + "step": 18541 + }, + { + "epoch": 1.2866560266463118, + "grad_norm": 3.8229095935821533, + "learning_rate": 0.00015860657079874198, + "loss": 0.9822, + "step": 18542 + }, + { + "epoch": 1.2867254180834085, + "grad_norm": 2.585817575454712, + "learning_rate": 0.00015860201814704068, + "loss": 0.9177, + "step": 18543 + }, + { + "epoch": 1.2867948095205053, + "grad_norm": 1.8837839365005493, + "learning_rate": 0.00015859746531033962, + "loss": 0.6631, + "step": 18544 + }, + { + "epoch": 1.286864200957602, + "grad_norm": 1.6845519542694092, + "learning_rate": 0.00015859291228865314, + "loss": 0.5799, + "step": 18545 + }, + { + "epoch": 1.2869335923946985, + "grad_norm": 1.603731632232666, + "learning_rate": 0.00015858835908199564, + "loss": 0.66, + "step": 18546 + }, + { + "epoch": 1.2870029838317951, + "grad_norm": 1.671710729598999, + "learning_rate": 0.00015858380569038153, + "loss": 0.6797, + "step": 18547 + }, + { + "epoch": 1.2870723752688917, + "grad_norm": 2.0672168731689453, + "learning_rate": 0.0001585792521138251, + "loss": 0.6849, + "step": 18548 + }, + { + "epoch": 1.2871417667059886, + "grad_norm": 2.1015279293060303, + "learning_rate": 0.00015857469835234075, + "loss": 0.7171, + "step": 18549 + }, + { + "epoch": 1.2872111581430852, + "grad_norm": 1.974705457687378, + "learning_rate": 0.0001585701444059429, + "loss": 0.7363, + "step": 18550 + }, + { + "epoch": 1.2872805495801818, + "grad_norm": 1.4871406555175781, + "learning_rate": 0.0001585655902746459, + "loss": 0.6079, + "step": 18551 + }, + { + "epoch": 1.2873499410172784, + "grad_norm": 1.542290449142456, + "learning_rate": 0.00015856103595846414, + "loss": 0.6146, + "step": 18552 + }, + { + "epoch": 1.287419332454375, + "grad_norm": 1.2522958517074585, + "learning_rate": 0.00015855648145741196, + "loss": 0.3857, + "step": 18553 + }, + { + "epoch": 1.2874887238914718, + "grad_norm": 1.6382672786712646, + "learning_rate": 0.0001585519267715038, + "loss": 0.7038, + "step": 18554 + }, + { + "epoch": 1.2875581153285685, + "grad_norm": 1.6376404762268066, + "learning_rate": 0.00015854737190075397, + "loss": 0.4887, + "step": 18555 + }, + { + "epoch": 1.287627506765665, + "grad_norm": 1.6967527866363525, + "learning_rate": 0.00015854281684517685, + "loss": 0.692, + "step": 18556 + }, + { + "epoch": 1.287696898202762, + "grad_norm": 1.7637394666671753, + "learning_rate": 0.00015853826160478688, + "loss": 0.7734, + "step": 18557 + }, + { + "epoch": 1.2877662896398585, + "grad_norm": 2.1727025508880615, + "learning_rate": 0.00015853370617959845, + "loss": 0.8713, + "step": 18558 + }, + { + "epoch": 1.2878356810769551, + "grad_norm": 2.317281484603882, + "learning_rate": 0.00015852915056962583, + "loss": 0.8813, + "step": 18559 + }, + { + "epoch": 1.2879050725140517, + "grad_norm": 1.8924692869186401, + "learning_rate": 0.0001585245947748835, + "loss": 0.6387, + "step": 18560 + }, + { + "epoch": 1.2879744639511483, + "grad_norm": 4.488071441650391, + "learning_rate": 0.00015852003879538584, + "loss": 0.6255, + "step": 18561 + }, + { + "epoch": 1.2880438553882452, + "grad_norm": 1.6559451818466187, + "learning_rate": 0.00015851548263114717, + "loss": 0.7246, + "step": 18562 + }, + { + "epoch": 1.2881132468253418, + "grad_norm": 1.6312788724899292, + "learning_rate": 0.00015851092628218192, + "loss": 0.6149, + "step": 18563 + }, + { + "epoch": 1.2881826382624384, + "grad_norm": 1.7857481241226196, + "learning_rate": 0.00015850636974850448, + "loss": 0.8343, + "step": 18564 + }, + { + "epoch": 1.288252029699535, + "grad_norm": 1.581064224243164, + "learning_rate": 0.00015850181303012922, + "loss": 0.4532, + "step": 18565 + }, + { + "epoch": 1.2883214211366316, + "grad_norm": 1.81459641456604, + "learning_rate": 0.0001584972561270705, + "loss": 0.4105, + "step": 18566 + }, + { + "epoch": 1.2883908125737285, + "grad_norm": 2.2160356044769287, + "learning_rate": 0.00015849269903934273, + "loss": 0.8103, + "step": 18567 + }, + { + "epoch": 1.288460204010825, + "grad_norm": 1.8908305168151855, + "learning_rate": 0.00015848814176696032, + "loss": 0.5636, + "step": 18568 + }, + { + "epoch": 1.2885295954479217, + "grad_norm": 1.6920169591903687, + "learning_rate": 0.00015848358430993763, + "loss": 0.8679, + "step": 18569 + }, + { + "epoch": 1.2885989868850185, + "grad_norm": 1.3651106357574463, + "learning_rate": 0.00015847902666828902, + "loss": 0.4855, + "step": 18570 + }, + { + "epoch": 1.2886683783221151, + "grad_norm": 2.1156516075134277, + "learning_rate": 0.00015847446884202895, + "loss": 0.6134, + "step": 18571 + }, + { + "epoch": 1.2887377697592117, + "grad_norm": 2.4363133907318115, + "learning_rate": 0.00015846991083117172, + "loss": 0.7504, + "step": 18572 + }, + { + "epoch": 1.2888071611963083, + "grad_norm": 1.9185237884521484, + "learning_rate": 0.00015846535263573178, + "loss": 0.7895, + "step": 18573 + }, + { + "epoch": 1.288876552633405, + "grad_norm": 1.8679012060165405, + "learning_rate": 0.0001584607942557235, + "loss": 0.6363, + "step": 18574 + }, + { + "epoch": 1.2889459440705018, + "grad_norm": 1.600464940071106, + "learning_rate": 0.00015845623569116132, + "loss": 0.4871, + "step": 18575 + }, + { + "epoch": 1.2890153355075984, + "grad_norm": 1.7176491022109985, + "learning_rate": 0.00015845167694205955, + "loss": 0.7484, + "step": 18576 + }, + { + "epoch": 1.289084726944695, + "grad_norm": 2.149595022201538, + "learning_rate": 0.0001584471180084326, + "loss": 0.5986, + "step": 18577 + }, + { + "epoch": 1.2891541183817916, + "grad_norm": 2.8262486457824707, + "learning_rate": 0.00015844255889029492, + "loss": 0.5162, + "step": 18578 + }, + { + "epoch": 1.2892235098188882, + "grad_norm": 1.319749116897583, + "learning_rate": 0.00015843799958766084, + "loss": 0.6705, + "step": 18579 + }, + { + "epoch": 1.289292901255985, + "grad_norm": 1.888264775276184, + "learning_rate": 0.00015843344010054477, + "loss": 0.5823, + "step": 18580 + }, + { + "epoch": 1.2893622926930817, + "grad_norm": 2.3741958141326904, + "learning_rate": 0.00015842888042896114, + "loss": 0.6215, + "step": 18581 + }, + { + "epoch": 1.2894316841301783, + "grad_norm": 2.0145137310028076, + "learning_rate": 0.0001584243205729243, + "loss": 0.8148, + "step": 18582 + }, + { + "epoch": 1.2895010755672751, + "grad_norm": 1.9453107118606567, + "learning_rate": 0.00015841976053244863, + "loss": 0.664, + "step": 18583 + }, + { + "epoch": 1.2895704670043717, + "grad_norm": 1.56158447265625, + "learning_rate": 0.00015841520030754856, + "loss": 0.5748, + "step": 18584 + }, + { + "epoch": 1.2896398584414683, + "grad_norm": 1.811829686164856, + "learning_rate": 0.00015841063989823853, + "loss": 0.5365, + "step": 18585 + }, + { + "epoch": 1.289709249878565, + "grad_norm": 1.6827425956726074, + "learning_rate": 0.00015840607930453282, + "loss": 0.6211, + "step": 18586 + }, + { + "epoch": 1.2897786413156616, + "grad_norm": 2.2741620540618896, + "learning_rate": 0.00015840151852644595, + "loss": 0.4859, + "step": 18587 + }, + { + "epoch": 1.2898480327527584, + "grad_norm": 1.7313542366027832, + "learning_rate": 0.00015839695756399223, + "loss": 0.7447, + "step": 18588 + }, + { + "epoch": 1.289917424189855, + "grad_norm": 1.745819091796875, + "learning_rate": 0.00015839239641718612, + "loss": 0.7338, + "step": 18589 + }, + { + "epoch": 1.2899868156269516, + "grad_norm": 1.5915921926498413, + "learning_rate": 0.00015838783508604194, + "loss": 0.5743, + "step": 18590 + }, + { + "epoch": 1.2900562070640482, + "grad_norm": 1.756345510482788, + "learning_rate": 0.00015838327357057413, + "loss": 0.5958, + "step": 18591 + }, + { + "epoch": 1.2901255985011448, + "grad_norm": 1.8174630403518677, + "learning_rate": 0.00015837871187079715, + "loss": 0.599, + "step": 18592 + }, + { + "epoch": 1.2901949899382417, + "grad_norm": 1.659735083580017, + "learning_rate": 0.00015837414998672532, + "loss": 0.6127, + "step": 18593 + }, + { + "epoch": 1.2902643813753383, + "grad_norm": 1.8110243082046509, + "learning_rate": 0.00015836958791837308, + "loss": 0.7879, + "step": 18594 + }, + { + "epoch": 1.2903337728124349, + "grad_norm": 1.1696301698684692, + "learning_rate": 0.00015836502566575486, + "loss": 0.5536, + "step": 18595 + }, + { + "epoch": 1.2904031642495317, + "grad_norm": 1.7650355100631714, + "learning_rate": 0.00015836046322888501, + "loss": 0.8324, + "step": 18596 + }, + { + "epoch": 1.2904725556866283, + "grad_norm": 1.4009523391723633, + "learning_rate": 0.00015835590060777789, + "loss": 0.5095, + "step": 18597 + }, + { + "epoch": 1.290541947123725, + "grad_norm": 2.190438747406006, + "learning_rate": 0.00015835133780244803, + "loss": 0.6104, + "step": 18598 + }, + { + "epoch": 1.2906113385608216, + "grad_norm": 1.666041374206543, + "learning_rate": 0.00015834677481290973, + "loss": 0.5377, + "step": 18599 + }, + { + "epoch": 1.2906807299979182, + "grad_norm": 1.3541648387908936, + "learning_rate": 0.00015834221163917747, + "loss": 0.5098, + "step": 18600 + }, + { + "epoch": 1.290750121435015, + "grad_norm": 1.7835429906845093, + "learning_rate": 0.00015833764828126557, + "loss": 0.6656, + "step": 18601 + }, + { + "epoch": 1.2908195128721116, + "grad_norm": 2.112678289413452, + "learning_rate": 0.00015833308473918852, + "loss": 0.7669, + "step": 18602 + }, + { + "epoch": 1.2908889043092082, + "grad_norm": 2.272507667541504, + "learning_rate": 0.00015832852101296065, + "loss": 0.9662, + "step": 18603 + }, + { + "epoch": 1.2909582957463048, + "grad_norm": 2.0317492485046387, + "learning_rate": 0.00015832395710259643, + "loss": 0.9608, + "step": 18604 + }, + { + "epoch": 1.2910276871834014, + "grad_norm": 1.4463070631027222, + "learning_rate": 0.00015831939300811024, + "loss": 0.4865, + "step": 18605 + }, + { + "epoch": 1.2910970786204983, + "grad_norm": 1.8988959789276123, + "learning_rate": 0.00015831482872951652, + "loss": 0.4519, + "step": 18606 + }, + { + "epoch": 1.2911664700575949, + "grad_norm": 1.7663772106170654, + "learning_rate": 0.0001583102642668296, + "loss": 0.5781, + "step": 18607 + }, + { + "epoch": 1.2912358614946915, + "grad_norm": 1.6516526937484741, + "learning_rate": 0.00015830569962006397, + "loss": 0.7216, + "step": 18608 + }, + { + "epoch": 1.2913052529317883, + "grad_norm": 1.927718997001648, + "learning_rate": 0.00015830113478923404, + "loss": 0.7504, + "step": 18609 + }, + { + "epoch": 1.291374644368885, + "grad_norm": 2.3376073837280273, + "learning_rate": 0.00015829656977435415, + "loss": 1.0588, + "step": 18610 + }, + { + "epoch": 1.2914440358059815, + "grad_norm": 1.448446273803711, + "learning_rate": 0.00015829200457543872, + "loss": 0.5327, + "step": 18611 + }, + { + "epoch": 1.2915134272430782, + "grad_norm": 1.9081532955169678, + "learning_rate": 0.00015828743919250223, + "loss": 0.4997, + "step": 18612 + }, + { + "epoch": 1.2915828186801748, + "grad_norm": 1.9158252477645874, + "learning_rate": 0.0001582828736255591, + "loss": 0.6058, + "step": 18613 + }, + { + "epoch": 1.2916522101172716, + "grad_norm": 1.3408225774765015, + "learning_rate": 0.00015827830787462367, + "loss": 0.475, + "step": 18614 + }, + { + "epoch": 1.2917216015543682, + "grad_norm": 1.6567491292953491, + "learning_rate": 0.00015827374193971036, + "loss": 0.5659, + "step": 18615 + }, + { + "epoch": 1.2917909929914648, + "grad_norm": 1.9307725429534912, + "learning_rate": 0.00015826917582083364, + "loss": 0.72, + "step": 18616 + }, + { + "epoch": 1.2918603844285614, + "grad_norm": 1.5989456176757812, + "learning_rate": 0.00015826460951800784, + "loss": 0.7707, + "step": 18617 + }, + { + "epoch": 1.291929775865658, + "grad_norm": 1.4554890394210815, + "learning_rate": 0.00015826004303124748, + "loss": 0.6259, + "step": 18618 + }, + { + "epoch": 1.2919991673027549, + "grad_norm": 3.1122653484344482, + "learning_rate": 0.0001582554763605669, + "loss": 0.7069, + "step": 18619 + }, + { + "epoch": 1.2920685587398515, + "grad_norm": 1.9208776950836182, + "learning_rate": 0.0001582509095059806, + "loss": 0.715, + "step": 18620 + }, + { + "epoch": 1.292137950176948, + "grad_norm": 1.7750434875488281, + "learning_rate": 0.00015824634246750286, + "loss": 0.4317, + "step": 18621 + }, + { + "epoch": 1.292207341614045, + "grad_norm": 2.0760600566864014, + "learning_rate": 0.0001582417752451482, + "loss": 0.7077, + "step": 18622 + }, + { + "epoch": 1.2922767330511415, + "grad_norm": 1.5426030158996582, + "learning_rate": 0.00015823720783893103, + "loss": 0.7309, + "step": 18623 + }, + { + "epoch": 1.2923461244882382, + "grad_norm": 1.5715712308883667, + "learning_rate": 0.00015823264024886574, + "loss": 0.513, + "step": 18624 + }, + { + "epoch": 1.2924155159253348, + "grad_norm": 4.744236946105957, + "learning_rate": 0.00015822807247496673, + "loss": 0.671, + "step": 18625 + }, + { + "epoch": 1.2924849073624314, + "grad_norm": 2.1341724395751953, + "learning_rate": 0.0001582235045172485, + "loss": 0.891, + "step": 18626 + }, + { + "epoch": 1.2925542987995282, + "grad_norm": 2.9470608234405518, + "learning_rate": 0.00015821893637572537, + "loss": 0.8452, + "step": 18627 + }, + { + "epoch": 1.2926236902366248, + "grad_norm": 1.99209725856781, + "learning_rate": 0.00015821436805041184, + "loss": 0.7721, + "step": 18628 + }, + { + "epoch": 1.2926930816737214, + "grad_norm": 1.8328725099563599, + "learning_rate": 0.0001582097995413223, + "loss": 0.5888, + "step": 18629 + }, + { + "epoch": 1.292762473110818, + "grad_norm": 2.1010971069335938, + "learning_rate": 0.00015820523084847122, + "loss": 0.8263, + "step": 18630 + }, + { + "epoch": 1.2928318645479147, + "grad_norm": 1.8336756229400635, + "learning_rate": 0.0001582006619718729, + "loss": 0.6338, + "step": 18631 + }, + { + "epoch": 1.2929012559850115, + "grad_norm": 1.945252776145935, + "learning_rate": 0.0001581960929115419, + "loss": 0.7059, + "step": 18632 + }, + { + "epoch": 1.292970647422108, + "grad_norm": 1.5296152830123901, + "learning_rate": 0.00015819152366749258, + "loss": 0.5948, + "step": 18633 + }, + { + "epoch": 1.2930400388592047, + "grad_norm": 1.839382529258728, + "learning_rate": 0.00015818695423973936, + "loss": 0.6425, + "step": 18634 + }, + { + "epoch": 1.2931094302963015, + "grad_norm": 1.6948248147964478, + "learning_rate": 0.0001581823846282967, + "loss": 0.6712, + "step": 18635 + }, + { + "epoch": 1.2931788217333982, + "grad_norm": 1.9250798225402832, + "learning_rate": 0.00015817781483317894, + "loss": 0.8665, + "step": 18636 + }, + { + "epoch": 1.2932482131704948, + "grad_norm": 1.5070335865020752, + "learning_rate": 0.00015817324485440062, + "loss": 0.6431, + "step": 18637 + }, + { + "epoch": 1.2933176046075914, + "grad_norm": 1.835192322731018, + "learning_rate": 0.00015816867469197612, + "loss": 0.766, + "step": 18638 + }, + { + "epoch": 1.293386996044688, + "grad_norm": 1.661874771118164, + "learning_rate": 0.00015816410434591984, + "loss": 0.7171, + "step": 18639 + }, + { + "epoch": 1.2934563874817848, + "grad_norm": 1.358620524406433, + "learning_rate": 0.00015815953381624625, + "loss": 0.5408, + "step": 18640 + }, + { + "epoch": 1.2935257789188814, + "grad_norm": 2.5709524154663086, + "learning_rate": 0.00015815496310296972, + "loss": 0.9253, + "step": 18641 + }, + { + "epoch": 1.293595170355978, + "grad_norm": 1.8473855257034302, + "learning_rate": 0.00015815039220610477, + "loss": 0.7073, + "step": 18642 + }, + { + "epoch": 1.2936645617930747, + "grad_norm": 2.468531370162964, + "learning_rate": 0.00015814582112566575, + "loss": 0.8704, + "step": 18643 + }, + { + "epoch": 1.2937339532301713, + "grad_norm": 1.5975260734558105, + "learning_rate": 0.00015814124986166713, + "loss": 0.5236, + "step": 18644 + }, + { + "epoch": 1.293803344667268, + "grad_norm": 1.7224838733673096, + "learning_rate": 0.0001581366784141233, + "loss": 0.5264, + "step": 18645 + }, + { + "epoch": 1.2938727361043647, + "grad_norm": 3.18335223197937, + "learning_rate": 0.00015813210678304875, + "loss": 0.6904, + "step": 18646 + }, + { + "epoch": 1.2939421275414613, + "grad_norm": 1.6628458499908447, + "learning_rate": 0.0001581275349684579, + "loss": 0.695, + "step": 18647 + }, + { + "epoch": 1.2940115189785582, + "grad_norm": 1.372363567352295, + "learning_rate": 0.0001581229629703651, + "loss": 0.4566, + "step": 18648 + }, + { + "epoch": 1.2940809104156548, + "grad_norm": 1.8415697813034058, + "learning_rate": 0.0001581183907887849, + "loss": 0.9233, + "step": 18649 + }, + { + "epoch": 1.2941503018527514, + "grad_norm": 1.5825711488723755, + "learning_rate": 0.00015811381842373168, + "loss": 0.6437, + "step": 18650 + }, + { + "epoch": 1.294219693289848, + "grad_norm": 1.9561131000518799, + "learning_rate": 0.0001581092458752199, + "loss": 1.0308, + "step": 18651 + }, + { + "epoch": 1.2942890847269446, + "grad_norm": 2.4952073097229004, + "learning_rate": 0.00015810467314326392, + "loss": 0.8137, + "step": 18652 + }, + { + "epoch": 1.2943584761640414, + "grad_norm": 2.0366013050079346, + "learning_rate": 0.00015810010022787826, + "loss": 0.7007, + "step": 18653 + }, + { + "epoch": 1.294427867601138, + "grad_norm": 1.9875046014785767, + "learning_rate": 0.00015809552712907734, + "loss": 0.8209, + "step": 18654 + }, + { + "epoch": 1.2944972590382346, + "grad_norm": 1.7511929273605347, + "learning_rate": 0.00015809095384687556, + "loss": 0.6462, + "step": 18655 + }, + { + "epoch": 1.2945666504753315, + "grad_norm": 2.021578073501587, + "learning_rate": 0.00015808638038128736, + "loss": 0.4451, + "step": 18656 + }, + { + "epoch": 1.2946360419124279, + "grad_norm": 2.1369500160217285, + "learning_rate": 0.00015808180673232723, + "loss": 0.6772, + "step": 18657 + }, + { + "epoch": 1.2947054333495247, + "grad_norm": 1.9756664037704468, + "learning_rate": 0.00015807723290000956, + "loss": 0.757, + "step": 18658 + }, + { + "epoch": 1.2947748247866213, + "grad_norm": 2.0263512134552, + "learning_rate": 0.00015807265888434881, + "loss": 0.6422, + "step": 18659 + }, + { + "epoch": 1.294844216223718, + "grad_norm": 1.6912155151367188, + "learning_rate": 0.0001580680846853594, + "loss": 0.5684, + "step": 18660 + }, + { + "epoch": 1.2949136076608148, + "grad_norm": 1.676611304283142, + "learning_rate": 0.0001580635103030558, + "loss": 0.7803, + "step": 18661 + }, + { + "epoch": 1.2949829990979114, + "grad_norm": 2.268803834915161, + "learning_rate": 0.00015805893573745243, + "loss": 0.7214, + "step": 18662 + }, + { + "epoch": 1.295052390535008, + "grad_norm": 1.1778104305267334, + "learning_rate": 0.00015805436098856372, + "loss": 0.5863, + "step": 18663 + }, + { + "epoch": 1.2951217819721046, + "grad_norm": 1.615459680557251, + "learning_rate": 0.00015804978605640413, + "loss": 0.617, + "step": 18664 + }, + { + "epoch": 1.2951911734092012, + "grad_norm": 2.000455617904663, + "learning_rate": 0.00015804521094098814, + "loss": 0.7148, + "step": 18665 + }, + { + "epoch": 1.295260564846298, + "grad_norm": 1.7615957260131836, + "learning_rate": 0.0001580406356423301, + "loss": 0.7296, + "step": 18666 + }, + { + "epoch": 1.2953299562833946, + "grad_norm": 1.8922175168991089, + "learning_rate": 0.00015803606016044454, + "loss": 0.6119, + "step": 18667 + }, + { + "epoch": 1.2953993477204913, + "grad_norm": 1.2354389429092407, + "learning_rate": 0.00015803148449534588, + "loss": 0.5269, + "step": 18668 + }, + { + "epoch": 1.295468739157588, + "grad_norm": 1.8422300815582275, + "learning_rate": 0.00015802690864704855, + "loss": 0.6521, + "step": 18669 + }, + { + "epoch": 1.2955381305946845, + "grad_norm": 1.861085295677185, + "learning_rate": 0.000158022332615567, + "loss": 0.7709, + "step": 18670 + }, + { + "epoch": 1.2956075220317813, + "grad_norm": 3.0065267086029053, + "learning_rate": 0.00015801775640091563, + "loss": 1.0266, + "step": 18671 + }, + { + "epoch": 1.295676913468878, + "grad_norm": 2.5632269382476807, + "learning_rate": 0.000158013180003109, + "loss": 0.6578, + "step": 18672 + }, + { + "epoch": 1.2957463049059745, + "grad_norm": 1.6550885438919067, + "learning_rate": 0.00015800860342216145, + "loss": 0.7363, + "step": 18673 + }, + { + "epoch": 1.2958156963430714, + "grad_norm": 1.444772720336914, + "learning_rate": 0.00015800402665808747, + "loss": 0.5134, + "step": 18674 + }, + { + "epoch": 1.295885087780168, + "grad_norm": 1.1349743604660034, + "learning_rate": 0.00015799944971090155, + "loss": 0.6118, + "step": 18675 + }, + { + "epoch": 1.2959544792172646, + "grad_norm": 1.6510239839553833, + "learning_rate": 0.00015799487258061804, + "loss": 0.5719, + "step": 18676 + }, + { + "epoch": 1.2960238706543612, + "grad_norm": 1.4580626487731934, + "learning_rate": 0.00015799029526725148, + "loss": 0.5723, + "step": 18677 + }, + { + "epoch": 1.2960932620914578, + "grad_norm": 1.5161361694335938, + "learning_rate": 0.0001579857177708163, + "loss": 0.5572, + "step": 18678 + }, + { + "epoch": 1.2961626535285546, + "grad_norm": 1.7775195837020874, + "learning_rate": 0.00015798114009132693, + "loss": 0.6104, + "step": 18679 + }, + { + "epoch": 1.2962320449656513, + "grad_norm": 1.7343270778656006, + "learning_rate": 0.00015797656222879777, + "loss": 0.7076, + "step": 18680 + }, + { + "epoch": 1.2963014364027479, + "grad_norm": 1.8542512655258179, + "learning_rate": 0.0001579719841832434, + "loss": 0.6576, + "step": 18681 + }, + { + "epoch": 1.2963708278398447, + "grad_norm": 1.8881182670593262, + "learning_rate": 0.00015796740595467817, + "loss": 0.4653, + "step": 18682 + }, + { + "epoch": 1.2964402192769413, + "grad_norm": 2.1014790534973145, + "learning_rate": 0.00015796282754311654, + "loss": 0.6379, + "step": 18683 + }, + { + "epoch": 1.296509610714038, + "grad_norm": 2.018436908721924, + "learning_rate": 0.00015795824894857305, + "loss": 0.6654, + "step": 18684 + }, + { + "epoch": 1.2965790021511345, + "grad_norm": 1.7922554016113281, + "learning_rate": 0.00015795367017106207, + "loss": 0.6426, + "step": 18685 + }, + { + "epoch": 1.2966483935882311, + "grad_norm": 1.8004041910171509, + "learning_rate": 0.00015794909121059808, + "loss": 0.7177, + "step": 18686 + }, + { + "epoch": 1.296717785025328, + "grad_norm": 1.931403398513794, + "learning_rate": 0.00015794451206719548, + "loss": 0.5178, + "step": 18687 + }, + { + "epoch": 1.2967871764624246, + "grad_norm": 1.8859457969665527, + "learning_rate": 0.00015793993274086882, + "loss": 0.6449, + "step": 18688 + }, + { + "epoch": 1.2968565678995212, + "grad_norm": 1.9593442678451538, + "learning_rate": 0.00015793535323163255, + "loss": 0.6476, + "step": 18689 + }, + { + "epoch": 1.2969259593366178, + "grad_norm": 1.9363384246826172, + "learning_rate": 0.00015793077353950104, + "loss": 0.7749, + "step": 18690 + }, + { + "epoch": 1.2969953507737144, + "grad_norm": 1.9216740131378174, + "learning_rate": 0.00015792619366448882, + "loss": 0.76, + "step": 18691 + }, + { + "epoch": 1.2970647422108113, + "grad_norm": 1.7208635807037354, + "learning_rate": 0.00015792161360661031, + "loss": 0.5096, + "step": 18692 + }, + { + "epoch": 1.2971341336479079, + "grad_norm": 1.89830482006073, + "learning_rate": 0.00015791703336588003, + "loss": 0.705, + "step": 18693 + }, + { + "epoch": 1.2972035250850045, + "grad_norm": 2.136338233947754, + "learning_rate": 0.00015791245294231233, + "loss": 0.7289, + "step": 18694 + }, + { + "epoch": 1.2972729165221013, + "grad_norm": 3.3486738204956055, + "learning_rate": 0.00015790787233592178, + "loss": 0.7814, + "step": 18695 + }, + { + "epoch": 1.297342307959198, + "grad_norm": 1.971979022026062, + "learning_rate": 0.00015790329154672278, + "loss": 0.6712, + "step": 18696 + }, + { + "epoch": 1.2974116993962945, + "grad_norm": 1.921808123588562, + "learning_rate": 0.00015789871057472982, + "loss": 0.7964, + "step": 18697 + }, + { + "epoch": 1.2974810908333911, + "grad_norm": 1.5672752857208252, + "learning_rate": 0.0001578941294199573, + "loss": 0.6441, + "step": 18698 + }, + { + "epoch": 1.2975504822704877, + "grad_norm": 2.1529855728149414, + "learning_rate": 0.00015788954808241981, + "loss": 0.6462, + "step": 18699 + }, + { + "epoch": 1.2976198737075846, + "grad_norm": 4.219674110412598, + "learning_rate": 0.00015788496656213166, + "loss": 0.7491, + "step": 18700 + }, + { + "epoch": 1.2976892651446812, + "grad_norm": 1.7702957391738892, + "learning_rate": 0.00015788038485910745, + "loss": 0.7067, + "step": 18701 + }, + { + "epoch": 1.2977586565817778, + "grad_norm": 1.9628502130508423, + "learning_rate": 0.00015787580297336152, + "loss": 0.7093, + "step": 18702 + }, + { + "epoch": 1.2978280480188744, + "grad_norm": 1.926043152809143, + "learning_rate": 0.00015787122090490848, + "loss": 0.8394, + "step": 18703 + }, + { + "epoch": 1.297897439455971, + "grad_norm": 1.533844232559204, + "learning_rate": 0.0001578666386537626, + "loss": 0.9056, + "step": 18704 + }, + { + "epoch": 1.2979668308930679, + "grad_norm": 1.6935583353042603, + "learning_rate": 0.00015786205621993852, + "loss": 0.7309, + "step": 18705 + }, + { + "epoch": 1.2980362223301645, + "grad_norm": 1.935384750366211, + "learning_rate": 0.00015785747360345064, + "loss": 0.5955, + "step": 18706 + }, + { + "epoch": 1.298105613767261, + "grad_norm": 2.106499671936035, + "learning_rate": 0.00015785289080431342, + "loss": 0.6659, + "step": 18707 + }, + { + "epoch": 1.298175005204358, + "grad_norm": 1.6541308164596558, + "learning_rate": 0.0001578483078225413, + "loss": 0.8184, + "step": 18708 + }, + { + "epoch": 1.2982443966414545, + "grad_norm": 2.1497998237609863, + "learning_rate": 0.00015784372465814885, + "loss": 0.676, + "step": 18709 + }, + { + "epoch": 1.2983137880785511, + "grad_norm": 1.932729959487915, + "learning_rate": 0.00015783914131115045, + "loss": 0.8057, + "step": 18710 + }, + { + "epoch": 1.2983831795156477, + "grad_norm": 3.356724500656128, + "learning_rate": 0.00015783455778156058, + "loss": 0.7086, + "step": 18711 + }, + { + "epoch": 1.2984525709527444, + "grad_norm": 2.124566078186035, + "learning_rate": 0.00015782997406939373, + "loss": 0.8458, + "step": 18712 + }, + { + "epoch": 1.2985219623898412, + "grad_norm": 1.9294055700302124, + "learning_rate": 0.0001578253901746644, + "loss": 0.6843, + "step": 18713 + }, + { + "epoch": 1.2985913538269378, + "grad_norm": 1.6671319007873535, + "learning_rate": 0.00015782080609738692, + "loss": 0.5642, + "step": 18714 + }, + { + "epoch": 1.2986607452640344, + "grad_norm": 1.409580945968628, + "learning_rate": 0.00015781622183757592, + "loss": 0.5897, + "step": 18715 + }, + { + "epoch": 1.298730136701131, + "grad_norm": 1.591317057609558, + "learning_rate": 0.00015781163739524585, + "loss": 0.6332, + "step": 18716 + }, + { + "epoch": 1.2987995281382276, + "grad_norm": 3.250105142593384, + "learning_rate": 0.0001578070527704111, + "loss": 0.7131, + "step": 18717 + }, + { + "epoch": 1.2988689195753245, + "grad_norm": 2.6863787174224854, + "learning_rate": 0.0001578024679630862, + "loss": 0.8843, + "step": 18718 + }, + { + "epoch": 1.298938311012421, + "grad_norm": 1.9760335683822632, + "learning_rate": 0.00015779788297328563, + "loss": 0.6441, + "step": 18719 + }, + { + "epoch": 1.2990077024495177, + "grad_norm": 3.1034297943115234, + "learning_rate": 0.00015779329780102384, + "loss": 0.6321, + "step": 18720 + }, + { + "epoch": 1.2990770938866145, + "grad_norm": 1.9796499013900757, + "learning_rate": 0.00015778871244631534, + "loss": 0.5905, + "step": 18721 + }, + { + "epoch": 1.2991464853237111, + "grad_norm": 1.7563892602920532, + "learning_rate": 0.00015778412690917453, + "loss": 0.5149, + "step": 18722 + }, + { + "epoch": 1.2992158767608077, + "grad_norm": 1.3255938291549683, + "learning_rate": 0.00015777954118961594, + "loss": 0.4309, + "step": 18723 + }, + { + "epoch": 1.2992852681979044, + "grad_norm": 1.80458402633667, + "learning_rate": 0.00015777495528765409, + "loss": 0.5305, + "step": 18724 + }, + { + "epoch": 1.299354659635001, + "grad_norm": 1.5118725299835205, + "learning_rate": 0.00015777036920330338, + "loss": 0.5443, + "step": 18725 + }, + { + "epoch": 1.2994240510720978, + "grad_norm": 2.0018439292907715, + "learning_rate": 0.0001577657829365783, + "loss": 0.754, + "step": 18726 + }, + { + "epoch": 1.2994934425091944, + "grad_norm": 2.334519386291504, + "learning_rate": 0.0001577611964874934, + "loss": 0.6076, + "step": 18727 + }, + { + "epoch": 1.299562833946291, + "grad_norm": 1.7780437469482422, + "learning_rate": 0.00015775660985606305, + "loss": 0.7554, + "step": 18728 + }, + { + "epoch": 1.2996322253833876, + "grad_norm": 1.7328613996505737, + "learning_rate": 0.00015775202304230178, + "loss": 0.5683, + "step": 18729 + }, + { + "epoch": 1.2997016168204842, + "grad_norm": 2.1365742683410645, + "learning_rate": 0.0001577474360462241, + "loss": 0.7322, + "step": 18730 + }, + { + "epoch": 1.299771008257581, + "grad_norm": 3.0357072353363037, + "learning_rate": 0.00015774284886784447, + "loss": 0.6577, + "step": 18731 + }, + { + "epoch": 1.2998403996946777, + "grad_norm": 3.1122937202453613, + "learning_rate": 0.00015773826150717736, + "loss": 1.0168, + "step": 18732 + }, + { + "epoch": 1.2999097911317743, + "grad_norm": 2.8975393772125244, + "learning_rate": 0.00015773367396423725, + "loss": 0.7606, + "step": 18733 + }, + { + "epoch": 1.2999791825688711, + "grad_norm": 2.133086681365967, + "learning_rate": 0.00015772908623903865, + "loss": 0.4884, + "step": 18734 + }, + { + "epoch": 1.3000485740059677, + "grad_norm": 1.6310381889343262, + "learning_rate": 0.00015772449833159597, + "loss": 0.6547, + "step": 18735 + }, + { + "epoch": 1.3001179654430643, + "grad_norm": 1.3669483661651611, + "learning_rate": 0.00015771991024192379, + "loss": 0.7546, + "step": 18736 + }, + { + "epoch": 1.300187356880161, + "grad_norm": 1.9975484609603882, + "learning_rate": 0.00015771532197003655, + "loss": 0.8145, + "step": 18737 + }, + { + "epoch": 1.3002567483172576, + "grad_norm": 2.3645529747009277, + "learning_rate": 0.00015771073351594873, + "loss": 0.9256, + "step": 18738 + }, + { + "epoch": 1.3003261397543544, + "grad_norm": 1.6354986429214478, + "learning_rate": 0.00015770614487967482, + "loss": 0.5037, + "step": 18739 + }, + { + "epoch": 1.300395531191451, + "grad_norm": 1.9129447937011719, + "learning_rate": 0.0001577015560612293, + "loss": 0.7451, + "step": 18740 + }, + { + "epoch": 1.3004649226285476, + "grad_norm": 1.8496427536010742, + "learning_rate": 0.00015769696706062668, + "loss": 0.6625, + "step": 18741 + }, + { + "epoch": 1.3005343140656442, + "grad_norm": 1.9156774282455444, + "learning_rate": 0.00015769237787788138, + "loss": 0.6321, + "step": 18742 + }, + { + "epoch": 1.3006037055027408, + "grad_norm": 1.509535789489746, + "learning_rate": 0.00015768778851300795, + "loss": 0.5071, + "step": 18743 + }, + { + "epoch": 1.3006730969398377, + "grad_norm": 2.1734135150909424, + "learning_rate": 0.00015768319896602092, + "loss": 0.8116, + "step": 18744 + }, + { + "epoch": 1.3007424883769343, + "grad_norm": 2.022226333618164, + "learning_rate": 0.00015767860923693468, + "loss": 0.7799, + "step": 18745 + }, + { + "epoch": 1.300811879814031, + "grad_norm": 1.5518454313278198, + "learning_rate": 0.0001576740193257638, + "loss": 0.781, + "step": 18746 + }, + { + "epoch": 1.3008812712511277, + "grad_norm": 1.8698725700378418, + "learning_rate": 0.00015766942923252266, + "loss": 0.6737, + "step": 18747 + }, + { + "epoch": 1.3009506626882243, + "grad_norm": 1.4987021684646606, + "learning_rate": 0.00015766483895722592, + "loss": 0.5617, + "step": 18748 + }, + { + "epoch": 1.301020054125321, + "grad_norm": 2.303877353668213, + "learning_rate": 0.00015766024849988788, + "loss": 0.7849, + "step": 18749 + }, + { + "epoch": 1.3010894455624176, + "grad_norm": 2.0210742950439453, + "learning_rate": 0.00015765565786052314, + "loss": 0.8211, + "step": 18750 + }, + { + "epoch": 1.3011588369995142, + "grad_norm": 2.587953567504883, + "learning_rate": 0.00015765106703914623, + "loss": 0.7368, + "step": 18751 + }, + { + "epoch": 1.301228228436611, + "grad_norm": 1.7217684984207153, + "learning_rate": 0.00015764647603577157, + "loss": 0.5849, + "step": 18752 + }, + { + "epoch": 1.3012976198737076, + "grad_norm": 1.8528330326080322, + "learning_rate": 0.00015764188485041365, + "loss": 0.6236, + "step": 18753 + }, + { + "epoch": 1.3013670113108042, + "grad_norm": 1.6690794229507446, + "learning_rate": 0.00015763729348308703, + "loss": 0.5767, + "step": 18754 + }, + { + "epoch": 1.3014364027479008, + "grad_norm": 1.8721052408218384, + "learning_rate": 0.0001576327019338061, + "loss": 0.58, + "step": 18755 + }, + { + "epoch": 1.3015057941849975, + "grad_norm": 1.8289889097213745, + "learning_rate": 0.0001576281102025855, + "loss": 0.9746, + "step": 18756 + }, + { + "epoch": 1.3015751856220943, + "grad_norm": 1.9992519617080688, + "learning_rate": 0.00015762351828943954, + "loss": 0.6823, + "step": 18757 + }, + { + "epoch": 1.301644577059191, + "grad_norm": 2.01584529876709, + "learning_rate": 0.0001576189261943829, + "loss": 0.825, + "step": 18758 + }, + { + "epoch": 1.3017139684962875, + "grad_norm": 2.190239906311035, + "learning_rate": 0.00015761433391742993, + "loss": 0.6539, + "step": 18759 + }, + { + "epoch": 1.3017833599333843, + "grad_norm": 2.3474442958831787, + "learning_rate": 0.00015760974145859527, + "loss": 0.8437, + "step": 18760 + }, + { + "epoch": 1.301852751370481, + "grad_norm": 1.468010663986206, + "learning_rate": 0.0001576051488178933, + "loss": 0.6645, + "step": 18761 + }, + { + "epoch": 1.3019221428075776, + "grad_norm": 1.843573808670044, + "learning_rate": 0.00015760055599533854, + "loss": 0.6087, + "step": 18762 + }, + { + "epoch": 1.3019915342446742, + "grad_norm": 2.7637579441070557, + "learning_rate": 0.00015759596299094551, + "loss": 0.6026, + "step": 18763 + }, + { + "epoch": 1.3020609256817708, + "grad_norm": 1.2555439472198486, + "learning_rate": 0.00015759136980472873, + "loss": 0.4593, + "step": 18764 + }, + { + "epoch": 1.3021303171188676, + "grad_norm": 1.9071519374847412, + "learning_rate": 0.00015758677643670267, + "loss": 0.6957, + "step": 18765 + }, + { + "epoch": 1.3021997085559642, + "grad_norm": 2.409113883972168, + "learning_rate": 0.00015758218288688185, + "loss": 0.6243, + "step": 18766 + }, + { + "epoch": 1.3022690999930608, + "grad_norm": 2.0961086750030518, + "learning_rate": 0.00015757758915528074, + "loss": 0.8851, + "step": 18767 + }, + { + "epoch": 1.3023384914301575, + "grad_norm": 1.6514782905578613, + "learning_rate": 0.00015757299524191387, + "loss": 0.7447, + "step": 18768 + }, + { + "epoch": 1.302407882867254, + "grad_norm": 2.1976234912872314, + "learning_rate": 0.0001575684011467957, + "loss": 0.6788, + "step": 18769 + }, + { + "epoch": 1.302477274304351, + "grad_norm": 1.677712321281433, + "learning_rate": 0.0001575638068699408, + "loss": 0.6386, + "step": 18770 + }, + { + "epoch": 1.3025466657414475, + "grad_norm": 1.810916781425476, + "learning_rate": 0.00015755921241136366, + "loss": 0.527, + "step": 18771 + }, + { + "epoch": 1.3026160571785441, + "grad_norm": 3.3503522872924805, + "learning_rate": 0.00015755461777107875, + "loss": 0.644, + "step": 18772 + }, + { + "epoch": 1.302685448615641, + "grad_norm": 1.7998086214065552, + "learning_rate": 0.0001575500229491006, + "loss": 0.5623, + "step": 18773 + }, + { + "epoch": 1.3027548400527376, + "grad_norm": 1.714936375617981, + "learning_rate": 0.00015754542794544366, + "loss": 0.4651, + "step": 18774 + }, + { + "epoch": 1.3028242314898342, + "grad_norm": 1.6827224493026733, + "learning_rate": 0.00015754083276012254, + "loss": 0.5105, + "step": 18775 + }, + { + "epoch": 1.3028936229269308, + "grad_norm": 1.5424546003341675, + "learning_rate": 0.00015753623739315164, + "loss": 0.6332, + "step": 18776 + }, + { + "epoch": 1.3029630143640274, + "grad_norm": 2.6445136070251465, + "learning_rate": 0.0001575316418445455, + "loss": 0.7884, + "step": 18777 + }, + { + "epoch": 1.3030324058011242, + "grad_norm": 2.7862887382507324, + "learning_rate": 0.0001575270461143187, + "loss": 0.7034, + "step": 18778 + }, + { + "epoch": 1.3031017972382208, + "grad_norm": 2.8129611015319824, + "learning_rate": 0.00015752245020248567, + "loss": 0.9584, + "step": 18779 + }, + { + "epoch": 1.3031711886753174, + "grad_norm": 2.3665077686309814, + "learning_rate": 0.0001575178541090609, + "loss": 0.7135, + "step": 18780 + }, + { + "epoch": 1.303240580112414, + "grad_norm": 2.4284842014312744, + "learning_rate": 0.00015751325783405897, + "loss": 0.5581, + "step": 18781 + }, + { + "epoch": 1.3033099715495107, + "grad_norm": 1.821467638015747, + "learning_rate": 0.00015750866137749435, + "loss": 0.6535, + "step": 18782 + }, + { + "epoch": 1.3033793629866075, + "grad_norm": 1.7785159349441528, + "learning_rate": 0.0001575040647393816, + "loss": 0.8239, + "step": 18783 + }, + { + "epoch": 1.3034487544237041, + "grad_norm": 1.8269973993301392, + "learning_rate": 0.00015749946791973512, + "loss": 0.6478, + "step": 18784 + }, + { + "epoch": 1.3035181458608007, + "grad_norm": 2.051403760910034, + "learning_rate": 0.00015749487091856954, + "loss": 0.5134, + "step": 18785 + }, + { + "epoch": 1.3035875372978976, + "grad_norm": 2.0330605506896973, + "learning_rate": 0.0001574902737358993, + "loss": 0.6037, + "step": 18786 + }, + { + "epoch": 1.3036569287349942, + "grad_norm": 1.7433027029037476, + "learning_rate": 0.00015748567637173893, + "loss": 0.8642, + "step": 18787 + }, + { + "epoch": 1.3037263201720908, + "grad_norm": 2.2598464488983154, + "learning_rate": 0.00015748107882610296, + "loss": 0.5133, + "step": 18788 + }, + { + "epoch": 1.3037957116091874, + "grad_norm": 1.3657811880111694, + "learning_rate": 0.0001574764810990059, + "loss": 0.5577, + "step": 18789 + }, + { + "epoch": 1.303865103046284, + "grad_norm": 1.4101845026016235, + "learning_rate": 0.00015747188319046226, + "loss": 0.668, + "step": 18790 + }, + { + "epoch": 1.3039344944833808, + "grad_norm": 2.030456304550171, + "learning_rate": 0.00015746728510048654, + "loss": 0.5749, + "step": 18791 + }, + { + "epoch": 1.3040038859204774, + "grad_norm": 1.2501286268234253, + "learning_rate": 0.00015746268682909328, + "loss": 0.4625, + "step": 18792 + }, + { + "epoch": 1.304073277357574, + "grad_norm": 1.5969971418380737, + "learning_rate": 0.000157458088376297, + "loss": 0.7355, + "step": 18793 + }, + { + "epoch": 1.3041426687946707, + "grad_norm": 1.8926063776016235, + "learning_rate": 0.00015745348974211215, + "loss": 0.642, + "step": 18794 + }, + { + "epoch": 1.3042120602317673, + "grad_norm": 2.3792037963867188, + "learning_rate": 0.00015744889092655332, + "loss": 0.7272, + "step": 18795 + }, + { + "epoch": 1.3042814516688641, + "grad_norm": 1.6136865615844727, + "learning_rate": 0.00015744429192963505, + "loss": 0.4808, + "step": 18796 + }, + { + "epoch": 1.3043508431059607, + "grad_norm": 1.9922704696655273, + "learning_rate": 0.00015743969275137175, + "loss": 0.7188, + "step": 18797 + }, + { + "epoch": 1.3044202345430573, + "grad_norm": 1.3986457586288452, + "learning_rate": 0.00015743509339177805, + "loss": 0.5789, + "step": 18798 + }, + { + "epoch": 1.3044896259801542, + "grad_norm": 2.1540706157684326, + "learning_rate": 0.00015743049385086842, + "loss": 0.9322, + "step": 18799 + }, + { + "epoch": 1.3045590174172508, + "grad_norm": 2.342406749725342, + "learning_rate": 0.00015742589412865736, + "loss": 0.7478, + "step": 18800 + }, + { + "epoch": 1.3046284088543474, + "grad_norm": 1.8224639892578125, + "learning_rate": 0.00015742129422515944, + "loss": 0.6209, + "step": 18801 + }, + { + "epoch": 1.304697800291444, + "grad_norm": 1.5552388429641724, + "learning_rate": 0.00015741669414038914, + "loss": 0.4977, + "step": 18802 + }, + { + "epoch": 1.3047671917285406, + "grad_norm": 1.6643739938735962, + "learning_rate": 0.000157412093874361, + "loss": 0.5731, + "step": 18803 + }, + { + "epoch": 1.3048365831656374, + "grad_norm": 1.8193129301071167, + "learning_rate": 0.00015740749342708953, + "loss": 0.6971, + "step": 18804 + }, + { + "epoch": 1.304905974602734, + "grad_norm": 1.7277547121047974, + "learning_rate": 0.0001574028927985893, + "loss": 0.5855, + "step": 18805 + }, + { + "epoch": 1.3049753660398307, + "grad_norm": 1.920870304107666, + "learning_rate": 0.00015739829198887476, + "loss": 0.6186, + "step": 18806 + }, + { + "epoch": 1.3050447574769273, + "grad_norm": 1.9470350742340088, + "learning_rate": 0.00015739369099796049, + "loss": 0.9329, + "step": 18807 + }, + { + "epoch": 1.3051141489140239, + "grad_norm": 1.985990047454834, + "learning_rate": 0.000157389089825861, + "loss": 0.7914, + "step": 18808 + }, + { + "epoch": 1.3051835403511207, + "grad_norm": 2.802184820175171, + "learning_rate": 0.0001573844884725908, + "loss": 1.0625, + "step": 18809 + }, + { + "epoch": 1.3052529317882173, + "grad_norm": 1.9340522289276123, + "learning_rate": 0.0001573798869381644, + "loss": 0.4959, + "step": 18810 + }, + { + "epoch": 1.305322323225314, + "grad_norm": 1.9751038551330566, + "learning_rate": 0.0001573752852225964, + "loss": 0.9691, + "step": 18811 + }, + { + "epoch": 1.3053917146624108, + "grad_norm": 2.361868381500244, + "learning_rate": 0.00015737068332590125, + "loss": 0.6987, + "step": 18812 + }, + { + "epoch": 1.3054611060995074, + "grad_norm": 1.925580620765686, + "learning_rate": 0.00015736608124809355, + "loss": 0.6463, + "step": 18813 + }, + { + "epoch": 1.305530497536604, + "grad_norm": 1.370038390159607, + "learning_rate": 0.00015736147898918776, + "loss": 0.5592, + "step": 18814 + }, + { + "epoch": 1.3055998889737006, + "grad_norm": 2.497178554534912, + "learning_rate": 0.00015735687654919845, + "loss": 0.8658, + "step": 18815 + }, + { + "epoch": 1.3056692804107972, + "grad_norm": 1.5668622255325317, + "learning_rate": 0.0001573522739281401, + "loss": 0.6111, + "step": 18816 + }, + { + "epoch": 1.305738671847894, + "grad_norm": 1.8606659173965454, + "learning_rate": 0.00015734767112602732, + "loss": 0.6151, + "step": 18817 + }, + { + "epoch": 1.3058080632849907, + "grad_norm": 2.654749631881714, + "learning_rate": 0.00015734306814287456, + "loss": 0.7457, + "step": 18818 + }, + { + "epoch": 1.3058774547220873, + "grad_norm": 1.3435786962509155, + "learning_rate": 0.00015733846497869642, + "loss": 0.3997, + "step": 18819 + }, + { + "epoch": 1.3059468461591839, + "grad_norm": 1.8829162120819092, + "learning_rate": 0.0001573338616335074, + "loss": 0.6493, + "step": 18820 + }, + { + "epoch": 1.3060162375962805, + "grad_norm": 1.7909858226776123, + "learning_rate": 0.00015732925810732197, + "loss": 0.7453, + "step": 18821 + }, + { + "epoch": 1.3060856290333773, + "grad_norm": 1.8506320714950562, + "learning_rate": 0.00015732465440015475, + "loss": 0.4945, + "step": 18822 + }, + { + "epoch": 1.306155020470474, + "grad_norm": 1.9347773790359497, + "learning_rate": 0.00015732005051202028, + "loss": 0.8115, + "step": 18823 + }, + { + "epoch": 1.3062244119075705, + "grad_norm": 1.4597071409225464, + "learning_rate": 0.00015731544644293304, + "loss": 0.4539, + "step": 18824 + }, + { + "epoch": 1.3062938033446674, + "grad_norm": 1.6657049655914307, + "learning_rate": 0.0001573108421929076, + "loss": 0.8068, + "step": 18825 + }, + { + "epoch": 1.306363194781764, + "grad_norm": 2.211554765701294, + "learning_rate": 0.00015730623776195847, + "loss": 0.7378, + "step": 18826 + }, + { + "epoch": 1.3064325862188606, + "grad_norm": 1.9333869218826294, + "learning_rate": 0.0001573016331501002, + "loss": 0.6693, + "step": 18827 + }, + { + "epoch": 1.3065019776559572, + "grad_norm": 1.8886754512786865, + "learning_rate": 0.0001572970283573473, + "loss": 0.6509, + "step": 18828 + }, + { + "epoch": 1.3065713690930538, + "grad_norm": 2.048638343811035, + "learning_rate": 0.00015729242338371436, + "loss": 0.6487, + "step": 18829 + }, + { + "epoch": 1.3066407605301507, + "grad_norm": 1.7830969095230103, + "learning_rate": 0.00015728781822921587, + "loss": 0.6442, + "step": 18830 + }, + { + "epoch": 1.3067101519672473, + "grad_norm": 1.7625139951705933, + "learning_rate": 0.00015728321289386639, + "loss": 0.6172, + "step": 18831 + }, + { + "epoch": 1.3067795434043439, + "grad_norm": 1.6438055038452148, + "learning_rate": 0.00015727860737768047, + "loss": 0.7225, + "step": 18832 + }, + { + "epoch": 1.3068489348414405, + "grad_norm": 1.6700518131256104, + "learning_rate": 0.00015727400168067258, + "loss": 0.531, + "step": 18833 + }, + { + "epoch": 1.306918326278537, + "grad_norm": 1.947726845741272, + "learning_rate": 0.00015726939580285737, + "loss": 0.6774, + "step": 18834 + }, + { + "epoch": 1.306987717715634, + "grad_norm": 2.369885206222534, + "learning_rate": 0.00015726478974424928, + "loss": 0.9867, + "step": 18835 + }, + { + "epoch": 1.3070571091527305, + "grad_norm": 1.8881001472473145, + "learning_rate": 0.0001572601835048629, + "loss": 0.6759, + "step": 18836 + }, + { + "epoch": 1.3071265005898272, + "grad_norm": 1.8579970598220825, + "learning_rate": 0.00015725557708471278, + "loss": 0.7777, + "step": 18837 + }, + { + "epoch": 1.307195892026924, + "grad_norm": 1.862892746925354, + "learning_rate": 0.00015725097048381343, + "loss": 0.9821, + "step": 18838 + }, + { + "epoch": 1.3072652834640206, + "grad_norm": 1.5399229526519775, + "learning_rate": 0.00015724636370217942, + "loss": 0.5478, + "step": 18839 + }, + { + "epoch": 1.3073346749011172, + "grad_norm": 1.8736234903335571, + "learning_rate": 0.00015724175673982526, + "loss": 0.4112, + "step": 18840 + }, + { + "epoch": 1.3074040663382138, + "grad_norm": 1.79910409450531, + "learning_rate": 0.00015723714959676557, + "loss": 0.8256, + "step": 18841 + }, + { + "epoch": 1.3074734577753104, + "grad_norm": 1.720637559890747, + "learning_rate": 0.0001572325422730148, + "loss": 0.6659, + "step": 18842 + }, + { + "epoch": 1.3075428492124073, + "grad_norm": 1.9834004640579224, + "learning_rate": 0.00015722793476858753, + "loss": 0.7981, + "step": 18843 + }, + { + "epoch": 1.3076122406495039, + "grad_norm": 1.6349058151245117, + "learning_rate": 0.00015722332708349832, + "loss": 0.5757, + "step": 18844 + }, + { + "epoch": 1.3076816320866005, + "grad_norm": 2.256641149520874, + "learning_rate": 0.0001572187192177617, + "loss": 0.4614, + "step": 18845 + }, + { + "epoch": 1.307751023523697, + "grad_norm": 1.9719583988189697, + "learning_rate": 0.00015721411117139224, + "loss": 0.6488, + "step": 18846 + }, + { + "epoch": 1.3078204149607937, + "grad_norm": 1.9668946266174316, + "learning_rate": 0.00015720950294440444, + "loss": 0.9087, + "step": 18847 + }, + { + "epoch": 1.3078898063978905, + "grad_norm": 1.6132599115371704, + "learning_rate": 0.00015720489453681288, + "loss": 0.7466, + "step": 18848 + }, + { + "epoch": 1.3079591978349872, + "grad_norm": 1.281765103340149, + "learning_rate": 0.00015720028594863212, + "loss": 0.4567, + "step": 18849 + }, + { + "epoch": 1.3080285892720838, + "grad_norm": 1.8959288597106934, + "learning_rate": 0.00015719567717987668, + "loss": 0.8339, + "step": 18850 + }, + { + "epoch": 1.3080979807091806, + "grad_norm": 1.8536006212234497, + "learning_rate": 0.00015719106823056117, + "loss": 0.7456, + "step": 18851 + }, + { + "epoch": 1.3081673721462772, + "grad_norm": 2.7373180389404297, + "learning_rate": 0.00015718645910070005, + "loss": 0.6089, + "step": 18852 + }, + { + "epoch": 1.3082367635833738, + "grad_norm": 1.8848509788513184, + "learning_rate": 0.00015718184979030792, + "loss": 0.4624, + "step": 18853 + }, + { + "epoch": 1.3083061550204704, + "grad_norm": 1.3878077268600464, + "learning_rate": 0.00015717724029939932, + "loss": 0.5867, + "step": 18854 + }, + { + "epoch": 1.308375546457567, + "grad_norm": 1.6091665029525757, + "learning_rate": 0.0001571726306279888, + "loss": 0.6822, + "step": 18855 + }, + { + "epoch": 1.3084449378946639, + "grad_norm": 1.6431905031204224, + "learning_rate": 0.00015716802077609093, + "loss": 0.7905, + "step": 18856 + }, + { + "epoch": 1.3085143293317605, + "grad_norm": 1.6416714191436768, + "learning_rate": 0.00015716341074372022, + "loss": 0.6301, + "step": 18857 + }, + { + "epoch": 1.308583720768857, + "grad_norm": 1.8924460411071777, + "learning_rate": 0.00015715880053089132, + "loss": 0.7164, + "step": 18858 + }, + { + "epoch": 1.3086531122059537, + "grad_norm": 1.4134840965270996, + "learning_rate": 0.00015715419013761868, + "loss": 0.6003, + "step": 18859 + }, + { + "epoch": 1.3087225036430503, + "grad_norm": 1.2746156454086304, + "learning_rate": 0.0001571495795639169, + "loss": 0.61, + "step": 18860 + }, + { + "epoch": 1.3087918950801471, + "grad_norm": 2.3567800521850586, + "learning_rate": 0.00015714496880980052, + "loss": 1.0028, + "step": 18861 + }, + { + "epoch": 1.3088612865172438, + "grad_norm": 1.7955057621002197, + "learning_rate": 0.0001571403578752841, + "loss": 0.8003, + "step": 18862 + }, + { + "epoch": 1.3089306779543404, + "grad_norm": 2.3285510540008545, + "learning_rate": 0.00015713574676038222, + "loss": 0.8, + "step": 18863 + }, + { + "epoch": 1.3090000693914372, + "grad_norm": 1.6999770402908325, + "learning_rate": 0.00015713113546510937, + "loss": 0.7616, + "step": 18864 + }, + { + "epoch": 1.3090694608285338, + "grad_norm": 1.5677413940429688, + "learning_rate": 0.00015712652398948022, + "loss": 0.5803, + "step": 18865 + }, + { + "epoch": 1.3091388522656304, + "grad_norm": 2.0802786350250244, + "learning_rate": 0.00015712191233350922, + "loss": 0.8468, + "step": 18866 + }, + { + "epoch": 1.309208243702727, + "grad_norm": 2.541149854660034, + "learning_rate": 0.00015711730049721096, + "loss": 0.7296, + "step": 18867 + }, + { + "epoch": 1.3092776351398236, + "grad_norm": 2.319997549057007, + "learning_rate": 0.00015711268848060003, + "loss": 0.9956, + "step": 18868 + }, + { + "epoch": 1.3093470265769205, + "grad_norm": 1.7429122924804688, + "learning_rate": 0.00015710807628369097, + "loss": 0.4361, + "step": 18869 + }, + { + "epoch": 1.309416418014017, + "grad_norm": 1.588259220123291, + "learning_rate": 0.00015710346390649833, + "loss": 0.5947, + "step": 18870 + }, + { + "epoch": 1.3094858094511137, + "grad_norm": 1.8115978240966797, + "learning_rate": 0.0001570988513490367, + "loss": 0.7787, + "step": 18871 + }, + { + "epoch": 1.3095552008882103, + "grad_norm": 1.6377958059310913, + "learning_rate": 0.0001570942386113206, + "loss": 0.5613, + "step": 18872 + }, + { + "epoch": 1.309624592325307, + "grad_norm": 2.1554274559020996, + "learning_rate": 0.00015708962569336462, + "loss": 1.0688, + "step": 18873 + }, + { + "epoch": 1.3096939837624038, + "grad_norm": 1.600151777267456, + "learning_rate": 0.0001570850125951833, + "loss": 0.6692, + "step": 18874 + }, + { + "epoch": 1.3097633751995004, + "grad_norm": 1.6661498546600342, + "learning_rate": 0.00015708039931679126, + "loss": 0.6529, + "step": 18875 + }, + { + "epoch": 1.309832766636597, + "grad_norm": 1.257789969444275, + "learning_rate": 0.00015707578585820298, + "loss": 0.5463, + "step": 18876 + }, + { + "epoch": 1.3099021580736938, + "grad_norm": 1.7098294496536255, + "learning_rate": 0.00015707117221943308, + "loss": 0.66, + "step": 18877 + }, + { + "epoch": 1.3099715495107904, + "grad_norm": 1.6880698204040527, + "learning_rate": 0.00015706655840049612, + "loss": 0.6016, + "step": 18878 + }, + { + "epoch": 1.310040940947887, + "grad_norm": 1.8530590534210205, + "learning_rate": 0.0001570619444014066, + "loss": 0.6575, + "step": 18879 + }, + { + "epoch": 1.3101103323849836, + "grad_norm": 1.7972638607025146, + "learning_rate": 0.00015705733022217924, + "loss": 0.7051, + "step": 18880 + }, + { + "epoch": 1.3101797238220803, + "grad_norm": 1.8725887537002563, + "learning_rate": 0.00015705271586282843, + "loss": 0.9128, + "step": 18881 + }, + { + "epoch": 1.310249115259177, + "grad_norm": 2.4368062019348145, + "learning_rate": 0.00015704810132336888, + "loss": 0.7832, + "step": 18882 + }, + { + "epoch": 1.3103185066962737, + "grad_norm": 2.5578441619873047, + "learning_rate": 0.00015704348660381502, + "loss": 0.8839, + "step": 18883 + }, + { + "epoch": 1.3103878981333703, + "grad_norm": 2.5958619117736816, + "learning_rate": 0.00015703887170418152, + "loss": 0.6644, + "step": 18884 + }, + { + "epoch": 1.3104572895704671, + "grad_norm": 1.8286378383636475, + "learning_rate": 0.00015703425662448294, + "loss": 0.6705, + "step": 18885 + }, + { + "epoch": 1.3105266810075635, + "grad_norm": 1.4662013053894043, + "learning_rate": 0.00015702964136473383, + "loss": 0.5611, + "step": 18886 + }, + { + "epoch": 1.3105960724446604, + "grad_norm": 1.4208580255508423, + "learning_rate": 0.00015702502592494875, + "loss": 0.5966, + "step": 18887 + }, + { + "epoch": 1.310665463881757, + "grad_norm": 2.0455498695373535, + "learning_rate": 0.00015702041030514225, + "loss": 0.6848, + "step": 18888 + }, + { + "epoch": 1.3107348553188536, + "grad_norm": 1.395711898803711, + "learning_rate": 0.00015701579450532897, + "loss": 0.4166, + "step": 18889 + }, + { + "epoch": 1.3108042467559504, + "grad_norm": 1.65997314453125, + "learning_rate": 0.00015701117852552342, + "loss": 0.7803, + "step": 18890 + }, + { + "epoch": 1.310873638193047, + "grad_norm": 1.3794811964035034, + "learning_rate": 0.0001570065623657402, + "loss": 0.6842, + "step": 18891 + }, + { + "epoch": 1.3109430296301436, + "grad_norm": 2.164517402648926, + "learning_rate": 0.00015700194602599385, + "loss": 0.9032, + "step": 18892 + }, + { + "epoch": 1.3110124210672403, + "grad_norm": 1.8194053173065186, + "learning_rate": 0.000156997329506299, + "loss": 0.6924, + "step": 18893 + }, + { + "epoch": 1.3110818125043369, + "grad_norm": 2.126396656036377, + "learning_rate": 0.0001569927128066702, + "loss": 0.5788, + "step": 18894 + }, + { + "epoch": 1.3111512039414337, + "grad_norm": 1.8493027687072754, + "learning_rate": 0.00015698809592712204, + "loss": 0.6167, + "step": 18895 + }, + { + "epoch": 1.3112205953785303, + "grad_norm": 2.302180767059326, + "learning_rate": 0.000156983478867669, + "loss": 0.9359, + "step": 18896 + }, + { + "epoch": 1.311289986815627, + "grad_norm": 1.7872917652130127, + "learning_rate": 0.0001569788616283258, + "loss": 0.7697, + "step": 18897 + }, + { + "epoch": 1.3113593782527238, + "grad_norm": 1.3102331161499023, + "learning_rate": 0.0001569742442091069, + "loss": 0.4493, + "step": 18898 + }, + { + "epoch": 1.3114287696898201, + "grad_norm": 2.314286470413208, + "learning_rate": 0.00015696962661002694, + "loss": 0.669, + "step": 18899 + }, + { + "epoch": 1.311498161126917, + "grad_norm": 4.426278114318848, + "learning_rate": 0.00015696500883110047, + "loss": 0.6752, + "step": 18900 + }, + { + "epoch": 1.3115675525640136, + "grad_norm": 1.7060322761535645, + "learning_rate": 0.00015696039087234208, + "loss": 0.612, + "step": 18901 + }, + { + "epoch": 1.3116369440011102, + "grad_norm": 1.9097039699554443, + "learning_rate": 0.00015695577273376634, + "loss": 0.5899, + "step": 18902 + }, + { + "epoch": 1.311706335438207, + "grad_norm": 1.8944978713989258, + "learning_rate": 0.00015695115441538785, + "loss": 0.5956, + "step": 18903 + }, + { + "epoch": 1.3117757268753036, + "grad_norm": 2.3473458290100098, + "learning_rate": 0.00015694653591722115, + "loss": 0.6748, + "step": 18904 + }, + { + "epoch": 1.3118451183124002, + "grad_norm": 1.3285624980926514, + "learning_rate": 0.00015694191723928085, + "loss": 0.5379, + "step": 18905 + }, + { + "epoch": 1.3119145097494969, + "grad_norm": 2.3757705688476562, + "learning_rate": 0.00015693729838158152, + "loss": 0.4677, + "step": 18906 + }, + { + "epoch": 1.3119839011865935, + "grad_norm": 1.9462673664093018, + "learning_rate": 0.00015693267934413778, + "loss": 0.8695, + "step": 18907 + }, + { + "epoch": 1.3120532926236903, + "grad_norm": 1.6454877853393555, + "learning_rate": 0.00015692806012696412, + "loss": 0.5532, + "step": 18908 + }, + { + "epoch": 1.312122684060787, + "grad_norm": 1.9510557651519775, + "learning_rate": 0.0001569234407300752, + "loss": 0.7276, + "step": 18909 + }, + { + "epoch": 1.3121920754978835, + "grad_norm": 1.8682684898376465, + "learning_rate": 0.00015691882115348565, + "loss": 0.7528, + "step": 18910 + }, + { + "epoch": 1.3122614669349804, + "grad_norm": 1.7242876291275024, + "learning_rate": 0.00015691420139720989, + "loss": 0.7553, + "step": 18911 + }, + { + "epoch": 1.3123308583720767, + "grad_norm": 1.8565586805343628, + "learning_rate": 0.00015690958146126263, + "loss": 0.5439, + "step": 18912 + }, + { + "epoch": 1.3124002498091736, + "grad_norm": 1.7220501899719238, + "learning_rate": 0.00015690496134565843, + "loss": 0.4595, + "step": 18913 + }, + { + "epoch": 1.3124696412462702, + "grad_norm": 1.606594204902649, + "learning_rate": 0.0001569003410504119, + "loss": 0.7094, + "step": 18914 + }, + { + "epoch": 1.3125390326833668, + "grad_norm": 2.8371360301971436, + "learning_rate": 0.00015689572057553755, + "loss": 0.7696, + "step": 18915 + }, + { + "epoch": 1.3126084241204636, + "grad_norm": 1.678305745124817, + "learning_rate": 0.00015689109992105, + "loss": 0.5052, + "step": 18916 + }, + { + "epoch": 1.3126778155575602, + "grad_norm": 2.252898693084717, + "learning_rate": 0.0001568864790869639, + "loss": 0.8368, + "step": 18917 + }, + { + "epoch": 1.3127472069946569, + "grad_norm": 2.0306358337402344, + "learning_rate": 0.00015688185807329375, + "loss": 0.6432, + "step": 18918 + }, + { + "epoch": 1.3128165984317535, + "grad_norm": 1.8273998498916626, + "learning_rate": 0.00015687723688005418, + "loss": 0.6617, + "step": 18919 + }, + { + "epoch": 1.31288598986885, + "grad_norm": 1.7141480445861816, + "learning_rate": 0.00015687261550725978, + "loss": 0.581, + "step": 18920 + }, + { + "epoch": 1.312955381305947, + "grad_norm": 1.526247262954712, + "learning_rate": 0.00015686799395492513, + "loss": 0.5669, + "step": 18921 + }, + { + "epoch": 1.3130247727430435, + "grad_norm": 1.7103468179702759, + "learning_rate": 0.0001568633722230648, + "loss": 0.5612, + "step": 18922 + }, + { + "epoch": 1.3130941641801401, + "grad_norm": 1.572701334953308, + "learning_rate": 0.00015685875031169343, + "loss": 0.5555, + "step": 18923 + }, + { + "epoch": 1.313163555617237, + "grad_norm": 1.6327465772628784, + "learning_rate": 0.00015685412822082556, + "loss": 0.5732, + "step": 18924 + }, + { + "epoch": 1.3132329470543336, + "grad_norm": 1.6812474727630615, + "learning_rate": 0.00015684950595047583, + "loss": 0.493, + "step": 18925 + }, + { + "epoch": 1.3133023384914302, + "grad_norm": 1.958485722541809, + "learning_rate": 0.00015684488350065879, + "loss": 0.6727, + "step": 18926 + }, + { + "epoch": 1.3133717299285268, + "grad_norm": 2.0872602462768555, + "learning_rate": 0.00015684026087138902, + "loss": 0.9295, + "step": 18927 + }, + { + "epoch": 1.3134411213656234, + "grad_norm": 2.344719648361206, + "learning_rate": 0.0001568356380626812, + "loss": 0.6737, + "step": 18928 + }, + { + "epoch": 1.3135105128027202, + "grad_norm": 1.4670106172561646, + "learning_rate": 0.00015683101507454983, + "loss": 0.531, + "step": 18929 + }, + { + "epoch": 1.3135799042398169, + "grad_norm": 1.6347284317016602, + "learning_rate": 0.00015682639190700952, + "loss": 0.5615, + "step": 18930 + }, + { + "epoch": 1.3136492956769135, + "grad_norm": 2.1170616149902344, + "learning_rate": 0.00015682176856007492, + "loss": 0.6698, + "step": 18931 + }, + { + "epoch": 1.31371868711401, + "grad_norm": 2.0540027618408203, + "learning_rate": 0.0001568171450337606, + "loss": 0.8106, + "step": 18932 + }, + { + "epoch": 1.3137880785511067, + "grad_norm": 1.7452659606933594, + "learning_rate": 0.00015681252132808112, + "loss": 0.61, + "step": 18933 + }, + { + "epoch": 1.3138574699882035, + "grad_norm": 2.5625836849212646, + "learning_rate": 0.0001568078974430511, + "loss": 0.7824, + "step": 18934 + }, + { + "epoch": 1.3139268614253001, + "grad_norm": 2.2901651859283447, + "learning_rate": 0.00015680327337868512, + "loss": 0.615, + "step": 18935 + }, + { + "epoch": 1.3139962528623967, + "grad_norm": 1.6212955713272095, + "learning_rate": 0.0001567986491349978, + "loss": 0.5408, + "step": 18936 + }, + { + "epoch": 1.3140656442994936, + "grad_norm": 1.8742955923080444, + "learning_rate": 0.00015679402471200374, + "loss": 0.4852, + "step": 18937 + }, + { + "epoch": 1.3141350357365902, + "grad_norm": 2.49322247505188, + "learning_rate": 0.00015678940010971755, + "loss": 0.6744, + "step": 18938 + }, + { + "epoch": 1.3142044271736868, + "grad_norm": 1.9224096536636353, + "learning_rate": 0.00015678477532815382, + "loss": 0.709, + "step": 18939 + }, + { + "epoch": 1.3142738186107834, + "grad_norm": 2.0179882049560547, + "learning_rate": 0.00015678015036732715, + "loss": 0.3767, + "step": 18940 + }, + { + "epoch": 1.31434321004788, + "grad_norm": 1.643602967262268, + "learning_rate": 0.00015677552522725213, + "loss": 0.6382, + "step": 18941 + }, + { + "epoch": 1.3144126014849769, + "grad_norm": 1.8619520664215088, + "learning_rate": 0.0001567708999079433, + "loss": 0.6004, + "step": 18942 + }, + { + "epoch": 1.3144819929220735, + "grad_norm": 1.6047401428222656, + "learning_rate": 0.00015676627440941538, + "loss": 0.5979, + "step": 18943 + }, + { + "epoch": 1.31455138435917, + "grad_norm": 1.4789658784866333, + "learning_rate": 0.00015676164873168293, + "loss": 0.6562, + "step": 18944 + }, + { + "epoch": 1.3146207757962667, + "grad_norm": 1.3599348068237305, + "learning_rate": 0.00015675702287476053, + "loss": 0.563, + "step": 18945 + }, + { + "epoch": 1.3146901672333633, + "grad_norm": 1.6244722604751587, + "learning_rate": 0.00015675239683866275, + "loss": 0.6776, + "step": 18946 + }, + { + "epoch": 1.3147595586704601, + "grad_norm": 1.7099019289016724, + "learning_rate": 0.00015674777062340428, + "loss": 0.5933, + "step": 18947 + }, + { + "epoch": 1.3148289501075567, + "grad_norm": 1.221557855606079, + "learning_rate": 0.0001567431442289997, + "loss": 0.5977, + "step": 18948 + }, + { + "epoch": 1.3148983415446533, + "grad_norm": 1.7394033670425415, + "learning_rate": 0.00015673851765546357, + "loss": 0.8142, + "step": 18949 + }, + { + "epoch": 1.3149677329817502, + "grad_norm": 1.6774077415466309, + "learning_rate": 0.00015673389090281053, + "loss": 0.6822, + "step": 18950 + }, + { + "epoch": 1.3150371244188468, + "grad_norm": 1.9609653949737549, + "learning_rate": 0.00015672926397105518, + "loss": 0.7294, + "step": 18951 + }, + { + "epoch": 1.3151065158559434, + "grad_norm": 1.4528703689575195, + "learning_rate": 0.0001567246368602121, + "loss": 0.6611, + "step": 18952 + }, + { + "epoch": 1.31517590729304, + "grad_norm": 1.8185077905654907, + "learning_rate": 0.00015672000957029592, + "loss": 0.9107, + "step": 18953 + }, + { + "epoch": 1.3152452987301366, + "grad_norm": 1.607125163078308, + "learning_rate": 0.0001567153821013213, + "loss": 0.7578, + "step": 18954 + }, + { + "epoch": 1.3153146901672335, + "grad_norm": 2.174243450164795, + "learning_rate": 0.00015671075445330276, + "loss": 0.5211, + "step": 18955 + }, + { + "epoch": 1.31538408160433, + "grad_norm": 1.518756628036499, + "learning_rate": 0.00015670612662625497, + "loss": 0.5364, + "step": 18956 + }, + { + "epoch": 1.3154534730414267, + "grad_norm": 2.6384694576263428, + "learning_rate": 0.0001567014986201925, + "loss": 0.8081, + "step": 18957 + }, + { + "epoch": 1.3155228644785233, + "grad_norm": 2.1228420734405518, + "learning_rate": 0.00015669687043512999, + "loss": 0.5586, + "step": 18958 + }, + { + "epoch": 1.31559225591562, + "grad_norm": 2.380600690841675, + "learning_rate": 0.00015669224207108203, + "loss": 0.8427, + "step": 18959 + }, + { + "epoch": 1.3156616473527167, + "grad_norm": 1.7144513130187988, + "learning_rate": 0.00015668761352806323, + "loss": 0.7266, + "step": 18960 + }, + { + "epoch": 1.3157310387898133, + "grad_norm": 1.5790965557098389, + "learning_rate": 0.00015668298480608824, + "loss": 0.6547, + "step": 18961 + }, + { + "epoch": 1.31580043022691, + "grad_norm": 1.5905660390853882, + "learning_rate": 0.00015667835590517162, + "loss": 0.5632, + "step": 18962 + }, + { + "epoch": 1.3158698216640068, + "grad_norm": 2.0849485397338867, + "learning_rate": 0.000156673726825328, + "loss": 0.8209, + "step": 18963 + }, + { + "epoch": 1.3159392131011034, + "grad_norm": 1.9692661762237549, + "learning_rate": 0.00015666909756657202, + "loss": 0.7045, + "step": 18964 + }, + { + "epoch": 1.3160086045382, + "grad_norm": 2.325230360031128, + "learning_rate": 0.00015666446812891827, + "loss": 0.7551, + "step": 18965 + }, + { + "epoch": 1.3160779959752966, + "grad_norm": 3.3425681591033936, + "learning_rate": 0.00015665983851238135, + "loss": 0.8632, + "step": 18966 + }, + { + "epoch": 1.3161473874123932, + "grad_norm": 1.459932804107666, + "learning_rate": 0.00015665520871697592, + "loss": 0.471, + "step": 18967 + }, + { + "epoch": 1.31621677884949, + "grad_norm": 1.4522682428359985, + "learning_rate": 0.00015665057874271652, + "loss": 0.5097, + "step": 18968 + }, + { + "epoch": 1.3162861702865867, + "grad_norm": 1.6912060976028442, + "learning_rate": 0.0001566459485896179, + "loss": 0.6257, + "step": 18969 + }, + { + "epoch": 1.3163555617236833, + "grad_norm": 1.6062391996383667, + "learning_rate": 0.00015664131825769454, + "loss": 0.6846, + "step": 18970 + }, + { + "epoch": 1.31642495316078, + "grad_norm": 1.6935789585113525, + "learning_rate": 0.00015663668774696106, + "loss": 0.5971, + "step": 18971 + }, + { + "epoch": 1.3164943445978765, + "grad_norm": 5.537757396697998, + "learning_rate": 0.0001566320570574322, + "loss": 0.6318, + "step": 18972 + }, + { + "epoch": 1.3165637360349733, + "grad_norm": 1.9474478960037231, + "learning_rate": 0.00015662742618912252, + "loss": 0.561, + "step": 18973 + }, + { + "epoch": 1.31663312747207, + "grad_norm": 1.4214282035827637, + "learning_rate": 0.00015662279514204659, + "loss": 0.4437, + "step": 18974 + }, + { + "epoch": 1.3167025189091666, + "grad_norm": 1.5399630069732666, + "learning_rate": 0.00015661816391621906, + "loss": 0.4643, + "step": 18975 + }, + { + "epoch": 1.3167719103462634, + "grad_norm": 2.2182843685150146, + "learning_rate": 0.0001566135325116546, + "loss": 0.7211, + "step": 18976 + }, + { + "epoch": 1.31684130178336, + "grad_norm": 1.5921955108642578, + "learning_rate": 0.00015660890092836772, + "loss": 0.705, + "step": 18977 + }, + { + "epoch": 1.3169106932204566, + "grad_norm": 1.9193960428237915, + "learning_rate": 0.00015660426916637313, + "loss": 0.6165, + "step": 18978 + }, + { + "epoch": 1.3169800846575532, + "grad_norm": 1.9945496320724487, + "learning_rate": 0.00015659963722568544, + "loss": 0.6792, + "step": 18979 + }, + { + "epoch": 1.3170494760946498, + "grad_norm": 1.9802706241607666, + "learning_rate": 0.00015659500510631927, + "loss": 0.9791, + "step": 18980 + }, + { + "epoch": 1.3171188675317467, + "grad_norm": 1.556711196899414, + "learning_rate": 0.00015659037280828922, + "loss": 0.8013, + "step": 18981 + }, + { + "epoch": 1.3171882589688433, + "grad_norm": 2.2196836471557617, + "learning_rate": 0.00015658574033160992, + "loss": 0.7002, + "step": 18982 + }, + { + "epoch": 1.31725765040594, + "grad_norm": 1.952659010887146, + "learning_rate": 0.00015658110767629603, + "loss": 0.6769, + "step": 18983 + }, + { + "epoch": 1.3173270418430365, + "grad_norm": 1.6764614582061768, + "learning_rate": 0.00015657647484236214, + "loss": 0.6838, + "step": 18984 + }, + { + "epoch": 1.3173964332801331, + "grad_norm": 2.012526035308838, + "learning_rate": 0.00015657184182982288, + "loss": 0.6675, + "step": 18985 + }, + { + "epoch": 1.31746582471723, + "grad_norm": 1.4494441747665405, + "learning_rate": 0.0001565672086386929, + "loss": 0.5769, + "step": 18986 + }, + { + "epoch": 1.3175352161543266, + "grad_norm": 2.1070730686187744, + "learning_rate": 0.00015656257526898675, + "loss": 0.4329, + "step": 18987 + }, + { + "epoch": 1.3176046075914232, + "grad_norm": 2.016287088394165, + "learning_rate": 0.00015655794172071916, + "loss": 0.5726, + "step": 18988 + }, + { + "epoch": 1.31767399902852, + "grad_norm": 2.3077478408813477, + "learning_rate": 0.0001565533079939047, + "loss": 0.9433, + "step": 18989 + }, + { + "epoch": 1.3177433904656166, + "grad_norm": 1.8427015542984009, + "learning_rate": 0.000156548674088558, + "loss": 0.5115, + "step": 18990 + }, + { + "epoch": 1.3178127819027132, + "grad_norm": 2.3983325958251953, + "learning_rate": 0.00015654404000469367, + "loss": 0.4852, + "step": 18991 + }, + { + "epoch": 1.3178821733398098, + "grad_norm": 1.953152060508728, + "learning_rate": 0.0001565394057423264, + "loss": 0.73, + "step": 18992 + }, + { + "epoch": 1.3179515647769064, + "grad_norm": 1.5475854873657227, + "learning_rate": 0.00015653477130147077, + "loss": 0.6135, + "step": 18993 + }, + { + "epoch": 1.3180209562140033, + "grad_norm": 2.073756217956543, + "learning_rate": 0.00015653013668214146, + "loss": 0.7781, + "step": 18994 + }, + { + "epoch": 1.3180903476511, + "grad_norm": 1.521407961845398, + "learning_rate": 0.00015652550188435303, + "loss": 0.5155, + "step": 18995 + }, + { + "epoch": 1.3181597390881965, + "grad_norm": 1.872129201889038, + "learning_rate": 0.00015652086690812019, + "loss": 0.6725, + "step": 18996 + }, + { + "epoch": 1.3182291305252931, + "grad_norm": 2.3420519828796387, + "learning_rate": 0.0001565162317534575, + "loss": 0.647, + "step": 18997 + }, + { + "epoch": 1.3182985219623897, + "grad_norm": 1.7892757654190063, + "learning_rate": 0.0001565115964203796, + "loss": 0.6652, + "step": 18998 + }, + { + "epoch": 1.3183679133994866, + "grad_norm": 2.0847461223602295, + "learning_rate": 0.0001565069609089012, + "loss": 0.6831, + "step": 18999 + }, + { + "epoch": 1.3184373048365832, + "grad_norm": 1.7054011821746826, + "learning_rate": 0.00015650232521903685, + "loss": 0.6121, + "step": 19000 + }, + { + "epoch": 1.3185066962736798, + "grad_norm": 2.0500500202178955, + "learning_rate": 0.00015649768935080123, + "loss": 0.7526, + "step": 19001 + }, + { + "epoch": 1.3185760877107766, + "grad_norm": 1.685196876525879, + "learning_rate": 0.00015649305330420893, + "loss": 0.818, + "step": 19002 + }, + { + "epoch": 1.3186454791478732, + "grad_norm": 1.7600542306900024, + "learning_rate": 0.00015648841707927464, + "loss": 0.5098, + "step": 19003 + }, + { + "epoch": 1.3187148705849698, + "grad_norm": 1.9301122426986694, + "learning_rate": 0.00015648378067601297, + "loss": 0.9398, + "step": 19004 + }, + { + "epoch": 1.3187842620220664, + "grad_norm": 1.5283793210983276, + "learning_rate": 0.00015647914409443856, + "loss": 0.4076, + "step": 19005 + }, + { + "epoch": 1.318853653459163, + "grad_norm": 1.7801138162612915, + "learning_rate": 0.00015647450733456602, + "loss": 0.6659, + "step": 19006 + }, + { + "epoch": 1.3189230448962599, + "grad_norm": 1.7145904302597046, + "learning_rate": 0.00015646987039641004, + "loss": 0.6195, + "step": 19007 + }, + { + "epoch": 1.3189924363333565, + "grad_norm": 1.3110426664352417, + "learning_rate": 0.0001564652332799852, + "loss": 0.359, + "step": 19008 + }, + { + "epoch": 1.319061827770453, + "grad_norm": 3.22969651222229, + "learning_rate": 0.00015646059598530622, + "loss": 0.6174, + "step": 19009 + }, + { + "epoch": 1.3191312192075497, + "grad_norm": 2.0395469665527344, + "learning_rate": 0.00015645595851238766, + "loss": 0.8609, + "step": 19010 + }, + { + "epoch": 1.3192006106446463, + "grad_norm": 1.74336838722229, + "learning_rate": 0.0001564513208612442, + "loss": 0.704, + "step": 19011 + }, + { + "epoch": 1.3192700020817432, + "grad_norm": 1.6969630718231201, + "learning_rate": 0.00015644668303189043, + "loss": 0.5843, + "step": 19012 + }, + { + "epoch": 1.3193393935188398, + "grad_norm": 1.992038369178772, + "learning_rate": 0.00015644204502434104, + "loss": 0.7778, + "step": 19013 + }, + { + "epoch": 1.3194087849559364, + "grad_norm": 1.5921154022216797, + "learning_rate": 0.0001564374068386107, + "loss": 0.5624, + "step": 19014 + }, + { + "epoch": 1.3194781763930332, + "grad_norm": 1.775877833366394, + "learning_rate": 0.000156432768474714, + "loss": 0.8331, + "step": 19015 + }, + { + "epoch": 1.3195475678301298, + "grad_norm": 1.9151599407196045, + "learning_rate": 0.00015642812993266558, + "loss": 0.6678, + "step": 19016 + }, + { + "epoch": 1.3196169592672264, + "grad_norm": 2.488590717315674, + "learning_rate": 0.00015642349121248013, + "loss": 0.7882, + "step": 19017 + }, + { + "epoch": 1.319686350704323, + "grad_norm": 2.613511085510254, + "learning_rate": 0.00015641885231417225, + "loss": 0.6487, + "step": 19018 + }, + { + "epoch": 1.3197557421414197, + "grad_norm": 1.9197957515716553, + "learning_rate": 0.0001564142132377566, + "loss": 0.6831, + "step": 19019 + }, + { + "epoch": 1.3198251335785165, + "grad_norm": 2.0542244911193848, + "learning_rate": 0.0001564095739832478, + "loss": 0.6836, + "step": 19020 + }, + { + "epoch": 1.319894525015613, + "grad_norm": 2.225620746612549, + "learning_rate": 0.0001564049345506606, + "loss": 0.667, + "step": 19021 + }, + { + "epoch": 1.3199639164527097, + "grad_norm": 2.095569133758545, + "learning_rate": 0.00015640029494000946, + "loss": 0.7009, + "step": 19022 + }, + { + "epoch": 1.3200333078898063, + "grad_norm": 1.8430960178375244, + "learning_rate": 0.0001563956551513092, + "loss": 0.8207, + "step": 19023 + }, + { + "epoch": 1.320102699326903, + "grad_norm": 2.0119130611419678, + "learning_rate": 0.00015639101518457438, + "loss": 0.6183, + "step": 19024 + }, + { + "epoch": 1.3201720907639998, + "grad_norm": 1.7466275691986084, + "learning_rate": 0.00015638637503981965, + "loss": 0.6668, + "step": 19025 + }, + { + "epoch": 1.3202414822010964, + "grad_norm": 2.261852502822876, + "learning_rate": 0.0001563817347170597, + "loss": 0.7565, + "step": 19026 + }, + { + "epoch": 1.320310873638193, + "grad_norm": 1.446597933769226, + "learning_rate": 0.00015637709421630915, + "loss": 0.5442, + "step": 19027 + }, + { + "epoch": 1.3203802650752898, + "grad_norm": 2.4868602752685547, + "learning_rate": 0.00015637245353758265, + "loss": 0.9742, + "step": 19028 + }, + { + "epoch": 1.3204496565123864, + "grad_norm": 3.024883508682251, + "learning_rate": 0.00015636781268089486, + "loss": 0.9301, + "step": 19029 + }, + { + "epoch": 1.320519047949483, + "grad_norm": 1.7322866916656494, + "learning_rate": 0.00015636317164626043, + "loss": 0.7065, + "step": 19030 + }, + { + "epoch": 1.3205884393865797, + "grad_norm": 1.9466854333877563, + "learning_rate": 0.000156358530433694, + "loss": 0.5674, + "step": 19031 + }, + { + "epoch": 1.3206578308236763, + "grad_norm": 2.1437625885009766, + "learning_rate": 0.00015635388904321022, + "loss": 0.6046, + "step": 19032 + }, + { + "epoch": 1.320727222260773, + "grad_norm": 1.6134555339813232, + "learning_rate": 0.00015634924747482375, + "loss": 0.5432, + "step": 19033 + }, + { + "epoch": 1.3207966136978697, + "grad_norm": 1.9261374473571777, + "learning_rate": 0.0001563446057285493, + "loss": 0.8107, + "step": 19034 + }, + { + "epoch": 1.3208660051349663, + "grad_norm": 2.2187271118164062, + "learning_rate": 0.00015633996380440135, + "loss": 0.7831, + "step": 19035 + }, + { + "epoch": 1.320935396572063, + "grad_norm": 1.329389214515686, + "learning_rate": 0.00015633532170239477, + "loss": 0.4746, + "step": 19036 + }, + { + "epoch": 1.3210047880091595, + "grad_norm": 2.0155274868011475, + "learning_rate": 0.0001563306794225441, + "loss": 0.7325, + "step": 19037 + }, + { + "epoch": 1.3210741794462564, + "grad_norm": 1.646958589553833, + "learning_rate": 0.00015632603696486398, + "loss": 0.5291, + "step": 19038 + }, + { + "epoch": 1.321143570883353, + "grad_norm": 2.5348925590515137, + "learning_rate": 0.0001563213943293691, + "loss": 0.9607, + "step": 19039 + }, + { + "epoch": 1.3212129623204496, + "grad_norm": 2.8378636837005615, + "learning_rate": 0.00015631675151607413, + "loss": 0.5999, + "step": 19040 + }, + { + "epoch": 1.3212823537575464, + "grad_norm": 2.774988889694214, + "learning_rate": 0.0001563121085249937, + "loss": 0.7319, + "step": 19041 + }, + { + "epoch": 1.321351745194643, + "grad_norm": 1.6740593910217285, + "learning_rate": 0.0001563074653561425, + "loss": 0.7175, + "step": 19042 + }, + { + "epoch": 1.3214211366317397, + "grad_norm": 1.6585423946380615, + "learning_rate": 0.00015630282200953516, + "loss": 0.5892, + "step": 19043 + }, + { + "epoch": 1.3214905280688363, + "grad_norm": 1.6822415590286255, + "learning_rate": 0.00015629817848518632, + "loss": 0.5188, + "step": 19044 + }, + { + "epoch": 1.3215599195059329, + "grad_norm": 1.7804181575775146, + "learning_rate": 0.00015629353478311067, + "loss": 0.5535, + "step": 19045 + }, + { + "epoch": 1.3216293109430297, + "grad_norm": 2.326795816421509, + "learning_rate": 0.00015628889090332285, + "loss": 1.0418, + "step": 19046 + }, + { + "epoch": 1.3216987023801263, + "grad_norm": 1.866465449333191, + "learning_rate": 0.00015628424684583755, + "loss": 0.5798, + "step": 19047 + }, + { + "epoch": 1.321768093817223, + "grad_norm": 1.8780717849731445, + "learning_rate": 0.00015627960261066943, + "loss": 0.7872, + "step": 19048 + }, + { + "epoch": 1.3218374852543195, + "grad_norm": 1.709996223449707, + "learning_rate": 0.0001562749581978331, + "loss": 0.7182, + "step": 19049 + }, + { + "epoch": 1.3219068766914162, + "grad_norm": 2.313845634460449, + "learning_rate": 0.00015627031360734327, + "loss": 1.2742, + "step": 19050 + }, + { + "epoch": 1.321976268128513, + "grad_norm": 2.0366392135620117, + "learning_rate": 0.00015626566883921456, + "loss": 0.9471, + "step": 19051 + }, + { + "epoch": 1.3220456595656096, + "grad_norm": 1.8512862920761108, + "learning_rate": 0.0001562610238934617, + "loss": 0.8163, + "step": 19052 + }, + { + "epoch": 1.3221150510027062, + "grad_norm": 1.4747735261917114, + "learning_rate": 0.00015625637877009929, + "loss": 0.7093, + "step": 19053 + }, + { + "epoch": 1.322184442439803, + "grad_norm": 1.9508001804351807, + "learning_rate": 0.00015625173346914205, + "loss": 0.6558, + "step": 19054 + }, + { + "epoch": 1.3222538338768997, + "grad_norm": 1.612737774848938, + "learning_rate": 0.0001562470879906046, + "loss": 0.5422, + "step": 19055 + }, + { + "epoch": 1.3223232253139963, + "grad_norm": 1.8445427417755127, + "learning_rate": 0.0001562424423345016, + "loss": 0.6859, + "step": 19056 + }, + { + "epoch": 1.3223926167510929, + "grad_norm": 1.7408281564712524, + "learning_rate": 0.00015623779650084774, + "loss": 0.6471, + "step": 19057 + }, + { + "epoch": 1.3224620081881895, + "grad_norm": 1.4967844486236572, + "learning_rate": 0.0001562331504896577, + "loss": 0.5927, + "step": 19058 + }, + { + "epoch": 1.3225313996252863, + "grad_norm": 2.0464060306549072, + "learning_rate": 0.00015622850430094615, + "loss": 0.6085, + "step": 19059 + }, + { + "epoch": 1.322600791062383, + "grad_norm": 1.5553680658340454, + "learning_rate": 0.0001562238579347277, + "loss": 0.6174, + "step": 19060 + }, + { + "epoch": 1.3226701824994795, + "grad_norm": 2.430678606033325, + "learning_rate": 0.00015621921139101703, + "loss": 0.7205, + "step": 19061 + }, + { + "epoch": 1.3227395739365762, + "grad_norm": 1.7545863389968872, + "learning_rate": 0.00015621456466982889, + "loss": 0.7775, + "step": 19062 + }, + { + "epoch": 1.3228089653736728, + "grad_norm": 2.9738099575042725, + "learning_rate": 0.00015620991777117786, + "loss": 0.7195, + "step": 19063 + }, + { + "epoch": 1.3228783568107696, + "grad_norm": 1.5244741439819336, + "learning_rate": 0.00015620527069507863, + "loss": 0.8125, + "step": 19064 + }, + { + "epoch": 1.3229477482478662, + "grad_norm": 1.6370313167572021, + "learning_rate": 0.00015620062344154592, + "loss": 0.4595, + "step": 19065 + }, + { + "epoch": 1.3230171396849628, + "grad_norm": 1.4782384634017944, + "learning_rate": 0.00015619597601059432, + "loss": 0.4299, + "step": 19066 + }, + { + "epoch": 1.3230865311220597, + "grad_norm": 2.22172212600708, + "learning_rate": 0.00015619132840223856, + "loss": 0.4926, + "step": 19067 + }, + { + "epoch": 1.3231559225591563, + "grad_norm": 1.4218804836273193, + "learning_rate": 0.00015618668061649328, + "loss": 0.5305, + "step": 19068 + }, + { + "epoch": 1.3232253139962529, + "grad_norm": 2.1094963550567627, + "learning_rate": 0.0001561820326533732, + "loss": 0.8259, + "step": 19069 + }, + { + "epoch": 1.3232947054333495, + "grad_norm": 2.083760976791382, + "learning_rate": 0.00015617738451289295, + "loss": 0.8565, + "step": 19070 + }, + { + "epoch": 1.323364096870446, + "grad_norm": 2.093595504760742, + "learning_rate": 0.0001561727361950672, + "loss": 0.6549, + "step": 19071 + }, + { + "epoch": 1.323433488307543, + "grad_norm": 1.6229015588760376, + "learning_rate": 0.00015616808769991066, + "loss": 0.5402, + "step": 19072 + }, + { + "epoch": 1.3235028797446395, + "grad_norm": 1.6997898817062378, + "learning_rate": 0.00015616343902743798, + "loss": 0.5101, + "step": 19073 + }, + { + "epoch": 1.3235722711817361, + "grad_norm": 2.2158730030059814, + "learning_rate": 0.00015615879017766382, + "loss": 0.7051, + "step": 19074 + }, + { + "epoch": 1.3236416626188328, + "grad_norm": 2.1394736766815186, + "learning_rate": 0.0001561541411506029, + "loss": 0.8531, + "step": 19075 + }, + { + "epoch": 1.3237110540559294, + "grad_norm": 2.160338878631592, + "learning_rate": 0.00015614949194626987, + "loss": 0.8546, + "step": 19076 + }, + { + "epoch": 1.3237804454930262, + "grad_norm": 1.9560567140579224, + "learning_rate": 0.0001561448425646794, + "loss": 0.7013, + "step": 19077 + }, + { + "epoch": 1.3238498369301228, + "grad_norm": 2.051518678665161, + "learning_rate": 0.00015614019300584615, + "loss": 0.5332, + "step": 19078 + }, + { + "epoch": 1.3239192283672194, + "grad_norm": 2.0278432369232178, + "learning_rate": 0.00015613554326978485, + "loss": 0.6205, + "step": 19079 + }, + { + "epoch": 1.3239886198043163, + "grad_norm": 2.230156183242798, + "learning_rate": 0.00015613089335651017, + "loss": 0.4382, + "step": 19080 + }, + { + "epoch": 1.3240580112414129, + "grad_norm": 1.412776231765747, + "learning_rate": 0.00015612624326603674, + "loss": 0.456, + "step": 19081 + }, + { + "epoch": 1.3241274026785095, + "grad_norm": 2.273249387741089, + "learning_rate": 0.0001561215929983793, + "loss": 0.8827, + "step": 19082 + }, + { + "epoch": 1.324196794115606, + "grad_norm": 2.7477834224700928, + "learning_rate": 0.00015611694255355247, + "loss": 0.857, + "step": 19083 + }, + { + "epoch": 1.3242661855527027, + "grad_norm": 1.8720802068710327, + "learning_rate": 0.00015611229193157098, + "loss": 0.7121, + "step": 19084 + }, + { + "epoch": 1.3243355769897995, + "grad_norm": 1.9765233993530273, + "learning_rate": 0.00015610764113244949, + "loss": 0.8076, + "step": 19085 + }, + { + "epoch": 1.3244049684268961, + "grad_norm": 1.9351701736450195, + "learning_rate": 0.00015610299015620272, + "loss": 0.6744, + "step": 19086 + }, + { + "epoch": 1.3244743598639928, + "grad_norm": 2.2521064281463623, + "learning_rate": 0.00015609833900284527, + "loss": 0.8661, + "step": 19087 + }, + { + "epoch": 1.3245437513010894, + "grad_norm": 1.6343345642089844, + "learning_rate": 0.00015609368767239187, + "loss": 0.608, + "step": 19088 + }, + { + "epoch": 1.324613142738186, + "grad_norm": 1.4121841192245483, + "learning_rate": 0.00015608903616485726, + "loss": 0.6343, + "step": 19089 + }, + { + "epoch": 1.3246825341752828, + "grad_norm": 1.8196755647659302, + "learning_rate": 0.00015608438448025603, + "loss": 0.7561, + "step": 19090 + }, + { + "epoch": 1.3247519256123794, + "grad_norm": 2.3205089569091797, + "learning_rate": 0.00015607973261860293, + "loss": 0.655, + "step": 19091 + }, + { + "epoch": 1.324821317049476, + "grad_norm": 1.8728581666946411, + "learning_rate": 0.00015607508057991257, + "loss": 0.5821, + "step": 19092 + }, + { + "epoch": 1.3248907084865729, + "grad_norm": 1.420910120010376, + "learning_rate": 0.00015607042836419976, + "loss": 0.5173, + "step": 19093 + }, + { + "epoch": 1.3249600999236695, + "grad_norm": 1.4579941034317017, + "learning_rate": 0.00015606577597147907, + "loss": 0.6206, + "step": 19094 + }, + { + "epoch": 1.325029491360766, + "grad_norm": 2.067448854446411, + "learning_rate": 0.00015606112340176522, + "loss": 0.9059, + "step": 19095 + }, + { + "epoch": 1.3250988827978627, + "grad_norm": 1.8682018518447876, + "learning_rate": 0.00015605647065507295, + "loss": 0.5762, + "step": 19096 + }, + { + "epoch": 1.3251682742349593, + "grad_norm": 2.072726249694824, + "learning_rate": 0.00015605181773141686, + "loss": 0.5664, + "step": 19097 + }, + { + "epoch": 1.3252376656720561, + "grad_norm": 1.9162768125534058, + "learning_rate": 0.0001560471646308117, + "loss": 0.6433, + "step": 19098 + }, + { + "epoch": 1.3253070571091528, + "grad_norm": 1.863916039466858, + "learning_rate": 0.00015604251135327215, + "loss": 0.5512, + "step": 19099 + }, + { + "epoch": 1.3253764485462494, + "grad_norm": 2.2399590015411377, + "learning_rate": 0.0001560378578988129, + "loss": 0.6314, + "step": 19100 + }, + { + "epoch": 1.325445839983346, + "grad_norm": 1.8787257671356201, + "learning_rate": 0.0001560332042674486, + "loss": 0.5725, + "step": 19101 + }, + { + "epoch": 1.3255152314204426, + "grad_norm": 1.937575101852417, + "learning_rate": 0.00015602855045919402, + "loss": 0.8334, + "step": 19102 + }, + { + "epoch": 1.3255846228575394, + "grad_norm": 1.7763020992279053, + "learning_rate": 0.0001560238964740638, + "loss": 0.9082, + "step": 19103 + }, + { + "epoch": 1.325654014294636, + "grad_norm": 2.009148359298706, + "learning_rate": 0.0001560192423120726, + "loss": 0.6265, + "step": 19104 + }, + { + "epoch": 1.3257234057317326, + "grad_norm": 1.5025628805160522, + "learning_rate": 0.00015601458797323518, + "loss": 0.6005, + "step": 19105 + }, + { + "epoch": 1.3257927971688295, + "grad_norm": 1.6184487342834473, + "learning_rate": 0.0001560099334575662, + "loss": 0.548, + "step": 19106 + }, + { + "epoch": 1.325862188605926, + "grad_norm": 1.5306079387664795, + "learning_rate": 0.00015600527876508038, + "loss": 0.7183, + "step": 19107 + }, + { + "epoch": 1.3259315800430227, + "grad_norm": 1.5922808647155762, + "learning_rate": 0.00015600062389579237, + "loss": 0.6327, + "step": 19108 + }, + { + "epoch": 1.3260009714801193, + "grad_norm": 1.7280133962631226, + "learning_rate": 0.00015599596884971687, + "loss": 0.6352, + "step": 19109 + }, + { + "epoch": 1.326070362917216, + "grad_norm": 2.549774646759033, + "learning_rate": 0.00015599131362686866, + "loss": 0.8729, + "step": 19110 + }, + { + "epoch": 1.3261397543543127, + "grad_norm": 1.7152751684188843, + "learning_rate": 0.0001559866582272623, + "loss": 0.7063, + "step": 19111 + }, + { + "epoch": 1.3262091457914094, + "grad_norm": 2.472695827484131, + "learning_rate": 0.0001559820026509126, + "loss": 0.8809, + "step": 19112 + }, + { + "epoch": 1.326278537228506, + "grad_norm": 1.9669405221939087, + "learning_rate": 0.00015597734689783415, + "loss": 0.6513, + "step": 19113 + }, + { + "epoch": 1.3263479286656026, + "grad_norm": 1.9835888147354126, + "learning_rate": 0.00015597269096804178, + "loss": 0.6395, + "step": 19114 + }, + { + "epoch": 1.3264173201026992, + "grad_norm": 1.848570466041565, + "learning_rate": 0.00015596803486155007, + "loss": 0.5599, + "step": 19115 + }, + { + "epoch": 1.326486711539796, + "grad_norm": 1.773750901222229, + "learning_rate": 0.0001559633785783738, + "loss": 0.5953, + "step": 19116 + }, + { + "epoch": 1.3265561029768926, + "grad_norm": 1.7557742595672607, + "learning_rate": 0.00015595872211852764, + "loss": 0.4795, + "step": 19117 + }, + { + "epoch": 1.3266254944139892, + "grad_norm": 1.531125545501709, + "learning_rate": 0.00015595406548202627, + "loss": 0.6364, + "step": 19118 + }, + { + "epoch": 1.326694885851086, + "grad_norm": 1.5978693962097168, + "learning_rate": 0.00015594940866888438, + "loss": 0.6173, + "step": 19119 + }, + { + "epoch": 1.3267642772881827, + "grad_norm": 3.1416895389556885, + "learning_rate": 0.00015594475167911672, + "loss": 0.6774, + "step": 19120 + }, + { + "epoch": 1.3268336687252793, + "grad_norm": 1.641964077949524, + "learning_rate": 0.00015594009451273802, + "loss": 0.6341, + "step": 19121 + }, + { + "epoch": 1.326903060162376, + "grad_norm": 1.7997337579727173, + "learning_rate": 0.00015593543716976288, + "loss": 0.7059, + "step": 19122 + }, + { + "epoch": 1.3269724515994725, + "grad_norm": 1.5503747463226318, + "learning_rate": 0.00015593077965020603, + "loss": 0.5474, + "step": 19123 + }, + { + "epoch": 1.3270418430365694, + "grad_norm": 1.7810330390930176, + "learning_rate": 0.00015592612195408225, + "loss": 0.5706, + "step": 19124 + }, + { + "epoch": 1.327111234473666, + "grad_norm": 2.465480327606201, + "learning_rate": 0.00015592146408140617, + "loss": 0.6485, + "step": 19125 + }, + { + "epoch": 1.3271806259107626, + "grad_norm": 2.338064193725586, + "learning_rate": 0.00015591680603219252, + "loss": 0.4856, + "step": 19126 + }, + { + "epoch": 1.3272500173478594, + "grad_norm": 2.326948642730713, + "learning_rate": 0.00015591214780645598, + "loss": 0.997, + "step": 19127 + }, + { + "epoch": 1.3273194087849558, + "grad_norm": 1.3847230672836304, + "learning_rate": 0.00015590748940421133, + "loss": 0.4799, + "step": 19128 + }, + { + "epoch": 1.3273888002220526, + "grad_norm": 2.2613351345062256, + "learning_rate": 0.00015590283082547317, + "loss": 0.9447, + "step": 19129 + }, + { + "epoch": 1.3274581916591492, + "grad_norm": 1.5054492950439453, + "learning_rate": 0.00015589817207025628, + "loss": 0.4558, + "step": 19130 + }, + { + "epoch": 1.3275275830962459, + "grad_norm": 1.988369107246399, + "learning_rate": 0.00015589351313857535, + "loss": 0.7784, + "step": 19131 + }, + { + "epoch": 1.3275969745333427, + "grad_norm": 1.695098876953125, + "learning_rate": 0.00015588885403044507, + "loss": 0.7165, + "step": 19132 + }, + { + "epoch": 1.3276663659704393, + "grad_norm": 1.3883681297302246, + "learning_rate": 0.0001558841947458802, + "loss": 0.515, + "step": 19133 + }, + { + "epoch": 1.327735757407536, + "grad_norm": 1.8685892820358276, + "learning_rate": 0.00015587953528489536, + "loss": 0.5587, + "step": 19134 + }, + { + "epoch": 1.3278051488446325, + "grad_norm": 2.0797367095947266, + "learning_rate": 0.00015587487564750534, + "loss": 0.6983, + "step": 19135 + }, + { + "epoch": 1.3278745402817291, + "grad_norm": 2.050727605819702, + "learning_rate": 0.00015587021583372482, + "loss": 0.6075, + "step": 19136 + }, + { + "epoch": 1.327943931718826, + "grad_norm": 1.9183720350265503, + "learning_rate": 0.00015586555584356848, + "loss": 0.4782, + "step": 19137 + }, + { + "epoch": 1.3280133231559226, + "grad_norm": 1.8232566118240356, + "learning_rate": 0.0001558608956770511, + "loss": 0.6948, + "step": 19138 + }, + { + "epoch": 1.3280827145930192, + "grad_norm": 2.852982759475708, + "learning_rate": 0.0001558562353341873, + "loss": 0.6238, + "step": 19139 + }, + { + "epoch": 1.328152106030116, + "grad_norm": 1.6096004247665405, + "learning_rate": 0.0001558515748149919, + "loss": 0.4988, + "step": 19140 + }, + { + "epoch": 1.3282214974672124, + "grad_norm": 2.2920444011688232, + "learning_rate": 0.00015584691411947956, + "loss": 0.7044, + "step": 19141 + }, + { + "epoch": 1.3282908889043092, + "grad_norm": 1.8726482391357422, + "learning_rate": 0.00015584225324766495, + "loss": 0.5923, + "step": 19142 + }, + { + "epoch": 1.3283602803414059, + "grad_norm": 2.656848192214966, + "learning_rate": 0.00015583759219956286, + "loss": 0.8945, + "step": 19143 + }, + { + "epoch": 1.3284296717785025, + "grad_norm": 1.4596827030181885, + "learning_rate": 0.00015583293097518795, + "loss": 0.5031, + "step": 19144 + }, + { + "epoch": 1.3284990632155993, + "grad_norm": 1.2706599235534668, + "learning_rate": 0.00015582826957455495, + "loss": 0.3988, + "step": 19145 + }, + { + "epoch": 1.328568454652696, + "grad_norm": 1.837755560874939, + "learning_rate": 0.0001558236079976786, + "loss": 0.6938, + "step": 19146 + }, + { + "epoch": 1.3286378460897925, + "grad_norm": 1.9564510583877563, + "learning_rate": 0.0001558189462445736, + "loss": 0.6831, + "step": 19147 + }, + { + "epoch": 1.3287072375268891, + "grad_norm": 2.7609267234802246, + "learning_rate": 0.00015581428431525463, + "loss": 0.6962, + "step": 19148 + }, + { + "epoch": 1.3287766289639857, + "grad_norm": 2.1024551391601562, + "learning_rate": 0.00015580962220973648, + "loss": 0.7887, + "step": 19149 + }, + { + "epoch": 1.3288460204010826, + "grad_norm": 1.9895490407943726, + "learning_rate": 0.00015580495992803382, + "loss": 0.5749, + "step": 19150 + }, + { + "epoch": 1.3289154118381792, + "grad_norm": 1.9200881719589233, + "learning_rate": 0.00015580029747016134, + "loss": 0.6442, + "step": 19151 + }, + { + "epoch": 1.3289848032752758, + "grad_norm": 1.743754506111145, + "learning_rate": 0.00015579563483613385, + "loss": 0.539, + "step": 19152 + }, + { + "epoch": 1.3290541947123726, + "grad_norm": 1.5481268167495728, + "learning_rate": 0.00015579097202596595, + "loss": 0.6314, + "step": 19153 + }, + { + "epoch": 1.3291235861494692, + "grad_norm": 1.735682725906372, + "learning_rate": 0.00015578630903967245, + "loss": 0.6673, + "step": 19154 + }, + { + "epoch": 1.3291929775865658, + "grad_norm": 1.7259455919265747, + "learning_rate": 0.00015578164587726808, + "loss": 0.7301, + "step": 19155 + }, + { + "epoch": 1.3292623690236625, + "grad_norm": 1.4696345329284668, + "learning_rate": 0.00015577698253876752, + "loss": 0.7333, + "step": 19156 + }, + { + "epoch": 1.329331760460759, + "grad_norm": 2.660426616668701, + "learning_rate": 0.00015577231902418544, + "loss": 0.8466, + "step": 19157 + }, + { + "epoch": 1.329401151897856, + "grad_norm": 1.8148542642593384, + "learning_rate": 0.00015576765533353666, + "loss": 0.5803, + "step": 19158 + }, + { + "epoch": 1.3294705433349525, + "grad_norm": 1.6485135555267334, + "learning_rate": 0.0001557629914668359, + "loss": 0.7154, + "step": 19159 + }, + { + "epoch": 1.3295399347720491, + "grad_norm": 1.6844581365585327, + "learning_rate": 0.00015575832742409777, + "loss": 0.6045, + "step": 19160 + }, + { + "epoch": 1.3296093262091457, + "grad_norm": 2.3017418384552, + "learning_rate": 0.0001557536632053371, + "loss": 0.7627, + "step": 19161 + }, + { + "epoch": 1.3296787176462423, + "grad_norm": 1.843209981918335, + "learning_rate": 0.0001557489988105686, + "loss": 0.6077, + "step": 19162 + }, + { + "epoch": 1.3297481090833392, + "grad_norm": 1.3715569972991943, + "learning_rate": 0.00015574433423980698, + "loss": 0.4933, + "step": 19163 + }, + { + "epoch": 1.3298175005204358, + "grad_norm": 2.0100677013397217, + "learning_rate": 0.00015573966949306692, + "loss": 0.9115, + "step": 19164 + }, + { + "epoch": 1.3298868919575324, + "grad_norm": 2.0411536693573, + "learning_rate": 0.00015573500457036324, + "loss": 0.7954, + "step": 19165 + }, + { + "epoch": 1.3299562833946292, + "grad_norm": 1.7345441579818726, + "learning_rate": 0.0001557303394717106, + "loss": 0.5026, + "step": 19166 + }, + { + "epoch": 1.3300256748317258, + "grad_norm": 2.1997766494750977, + "learning_rate": 0.00015572567419712375, + "loss": 0.539, + "step": 19167 + }, + { + "epoch": 1.3300950662688225, + "grad_norm": 1.5333119630813599, + "learning_rate": 0.00015572100874661739, + "loss": 0.5688, + "step": 19168 + }, + { + "epoch": 1.330164457705919, + "grad_norm": 1.8306400775909424, + "learning_rate": 0.0001557163431202063, + "loss": 0.7346, + "step": 19169 + }, + { + "epoch": 1.3302338491430157, + "grad_norm": 1.262782335281372, + "learning_rate": 0.00015571167731790516, + "loss": 0.6275, + "step": 19170 + }, + { + "epoch": 1.3303032405801125, + "grad_norm": 1.75770103931427, + "learning_rate": 0.0001557070113397287, + "loss": 0.5814, + "step": 19171 + }, + { + "epoch": 1.3303726320172091, + "grad_norm": 1.6169308423995972, + "learning_rate": 0.00015570234518569168, + "loss": 0.5055, + "step": 19172 + }, + { + "epoch": 1.3304420234543057, + "grad_norm": 1.9745934009552002, + "learning_rate": 0.00015569767885580886, + "loss": 0.6868, + "step": 19173 + }, + { + "epoch": 1.3305114148914023, + "grad_norm": 2.0655922889709473, + "learning_rate": 0.0001556930123500949, + "loss": 0.7313, + "step": 19174 + }, + { + "epoch": 1.330580806328499, + "grad_norm": 2.4019131660461426, + "learning_rate": 0.00015568834566856454, + "loss": 0.8588, + "step": 19175 + }, + { + "epoch": 1.3306501977655958, + "grad_norm": 2.062905788421631, + "learning_rate": 0.00015568367881123259, + "loss": 0.6768, + "step": 19176 + }, + { + "epoch": 1.3307195892026924, + "grad_norm": 1.7172141075134277, + "learning_rate": 0.00015567901177811368, + "loss": 0.6829, + "step": 19177 + }, + { + "epoch": 1.330788980639789, + "grad_norm": 1.4651457071304321, + "learning_rate": 0.0001556743445692226, + "loss": 0.4591, + "step": 19178 + }, + { + "epoch": 1.3308583720768858, + "grad_norm": 1.7463786602020264, + "learning_rate": 0.00015566967718457408, + "loss": 0.5987, + "step": 19179 + }, + { + "epoch": 1.3309277635139825, + "grad_norm": 1.7031131982803345, + "learning_rate": 0.00015566500962418283, + "loss": 0.786, + "step": 19180 + }, + { + "epoch": 1.330997154951079, + "grad_norm": 1.5875524282455444, + "learning_rate": 0.00015566034188806363, + "loss": 0.5633, + "step": 19181 + }, + { + "epoch": 1.3310665463881757, + "grad_norm": 2.1414685249328613, + "learning_rate": 0.00015565567397623115, + "loss": 0.6358, + "step": 19182 + }, + { + "epoch": 1.3311359378252723, + "grad_norm": 1.9181337356567383, + "learning_rate": 0.0001556510058887002, + "loss": 0.6512, + "step": 19183 + }, + { + "epoch": 1.3312053292623691, + "grad_norm": 2.6727187633514404, + "learning_rate": 0.0001556463376254855, + "loss": 0.5864, + "step": 19184 + }, + { + "epoch": 1.3312747206994657, + "grad_norm": 2.2241175174713135, + "learning_rate": 0.0001556416691866017, + "loss": 0.7605, + "step": 19185 + }, + { + "epoch": 1.3313441121365623, + "grad_norm": 1.4562257528305054, + "learning_rate": 0.00015563700057206366, + "loss": 0.5778, + "step": 19186 + }, + { + "epoch": 1.331413503573659, + "grad_norm": 1.9906846284866333, + "learning_rate": 0.00015563233178188607, + "loss": 0.6242, + "step": 19187 + }, + { + "epoch": 1.3314828950107556, + "grad_norm": 1.773474931716919, + "learning_rate": 0.00015562766281608363, + "loss": 0.5921, + "step": 19188 + }, + { + "epoch": 1.3315522864478524, + "grad_norm": 2.1541695594787598, + "learning_rate": 0.0001556229936746711, + "loss": 0.6865, + "step": 19189 + }, + { + "epoch": 1.331621677884949, + "grad_norm": 1.8405019044876099, + "learning_rate": 0.0001556183243576633, + "loss": 0.7221, + "step": 19190 + }, + { + "epoch": 1.3316910693220456, + "grad_norm": 2.286750316619873, + "learning_rate": 0.00015561365486507485, + "loss": 0.7465, + "step": 19191 + }, + { + "epoch": 1.3317604607591425, + "grad_norm": 2.6083579063415527, + "learning_rate": 0.00015560898519692052, + "loss": 0.6945, + "step": 19192 + }, + { + "epoch": 1.331829852196239, + "grad_norm": 1.950705885887146, + "learning_rate": 0.00015560431535321514, + "loss": 0.6145, + "step": 19193 + }, + { + "epoch": 1.3318992436333357, + "grad_norm": 1.4992272853851318, + "learning_rate": 0.00015559964533397332, + "loss": 0.5038, + "step": 19194 + }, + { + "epoch": 1.3319686350704323, + "grad_norm": 1.491416335105896, + "learning_rate": 0.00015559497513920994, + "loss": 0.6579, + "step": 19195 + }, + { + "epoch": 1.332038026507529, + "grad_norm": 1.8246790170669556, + "learning_rate": 0.00015559030476893962, + "loss": 0.7589, + "step": 19196 + }, + { + "epoch": 1.3321074179446257, + "grad_norm": 1.44532310962677, + "learning_rate": 0.0001555856342231772, + "loss": 0.8361, + "step": 19197 + }, + { + "epoch": 1.3321768093817223, + "grad_norm": 1.8402824401855469, + "learning_rate": 0.00015558096350193736, + "loss": 0.7897, + "step": 19198 + }, + { + "epoch": 1.332246200818819, + "grad_norm": 1.6867825984954834, + "learning_rate": 0.00015557629260523484, + "loss": 0.6951, + "step": 19199 + }, + { + "epoch": 1.3323155922559156, + "grad_norm": 2.0841028690338135, + "learning_rate": 0.00015557162153308446, + "loss": 0.7779, + "step": 19200 + }, + { + "epoch": 1.3323849836930122, + "grad_norm": 1.5893605947494507, + "learning_rate": 0.0001555669502855009, + "loss": 0.5101, + "step": 19201 + }, + { + "epoch": 1.332454375130109, + "grad_norm": 1.2963181734085083, + "learning_rate": 0.0001555622788624989, + "loss": 0.4671, + "step": 19202 + }, + { + "epoch": 1.3325237665672056, + "grad_norm": 2.4804091453552246, + "learning_rate": 0.00015555760726409327, + "loss": 0.6963, + "step": 19203 + }, + { + "epoch": 1.3325931580043022, + "grad_norm": 1.8465443849563599, + "learning_rate": 0.00015555293549029871, + "loss": 0.6467, + "step": 19204 + }, + { + "epoch": 1.332662549441399, + "grad_norm": 1.5714755058288574, + "learning_rate": 0.00015554826354112998, + "loss": 0.7613, + "step": 19205 + }, + { + "epoch": 1.3327319408784957, + "grad_norm": 1.7384172677993774, + "learning_rate": 0.00015554359141660182, + "loss": 0.7164, + "step": 19206 + }, + { + "epoch": 1.3328013323155923, + "grad_norm": 1.589766502380371, + "learning_rate": 0.000155538919116729, + "loss": 0.5636, + "step": 19207 + }, + { + "epoch": 1.332870723752689, + "grad_norm": 2.2689049243927, + "learning_rate": 0.0001555342466415262, + "loss": 0.5901, + "step": 19208 + }, + { + "epoch": 1.3329401151897855, + "grad_norm": 1.5927051305770874, + "learning_rate": 0.00015552957399100827, + "loss": 0.75, + "step": 19209 + }, + { + "epoch": 1.3330095066268823, + "grad_norm": 1.9579005241394043, + "learning_rate": 0.00015552490116518993, + "loss": 0.8195, + "step": 19210 + }, + { + "epoch": 1.333078898063979, + "grad_norm": 1.74929940700531, + "learning_rate": 0.0001555202281640859, + "loss": 0.6843, + "step": 19211 + }, + { + "epoch": 1.3331482895010756, + "grad_norm": 1.4847345352172852, + "learning_rate": 0.00015551555498771098, + "loss": 0.6188, + "step": 19212 + }, + { + "epoch": 1.3332176809381722, + "grad_norm": 1.481771469116211, + "learning_rate": 0.00015551088163607985, + "loss": 0.5872, + "step": 19213 + }, + { + "epoch": 1.3332870723752688, + "grad_norm": 1.8982659578323364, + "learning_rate": 0.00015550620810920736, + "loss": 0.5664, + "step": 19214 + }, + { + "epoch": 1.3333564638123656, + "grad_norm": 1.3698227405548096, + "learning_rate": 0.00015550153440710814, + "loss": 0.5317, + "step": 19215 + }, + { + "epoch": 1.3334258552494622, + "grad_norm": 1.950419545173645, + "learning_rate": 0.00015549686052979705, + "loss": 0.9958, + "step": 19216 + }, + { + "epoch": 1.3334952466865588, + "grad_norm": 1.8500815629959106, + "learning_rate": 0.00015549218647728882, + "loss": 0.8411, + "step": 19217 + }, + { + "epoch": 1.3335646381236557, + "grad_norm": 1.35706627368927, + "learning_rate": 0.00015548751224959818, + "loss": 0.5724, + "step": 19218 + }, + { + "epoch": 1.3336340295607523, + "grad_norm": 1.5004280805587769, + "learning_rate": 0.00015548283784673993, + "loss": 0.496, + "step": 19219 + }, + { + "epoch": 1.3337034209978489, + "grad_norm": 2.0584683418273926, + "learning_rate": 0.00015547816326872877, + "loss": 0.7092, + "step": 19220 + }, + { + "epoch": 1.3337728124349455, + "grad_norm": 1.9287315607070923, + "learning_rate": 0.0001554734885155795, + "loss": 0.713, + "step": 19221 + }, + { + "epoch": 1.333842203872042, + "grad_norm": 3.2194271087646484, + "learning_rate": 0.00015546881358730684, + "loss": 0.7029, + "step": 19222 + }, + { + "epoch": 1.333911595309139, + "grad_norm": 2.8324155807495117, + "learning_rate": 0.0001554641384839256, + "loss": 0.5673, + "step": 19223 + }, + { + "epoch": 1.3339809867462356, + "grad_norm": 1.7295591831207275, + "learning_rate": 0.0001554594632054505, + "loss": 0.6618, + "step": 19224 + }, + { + "epoch": 1.3340503781833322, + "grad_norm": 1.9632011651992798, + "learning_rate": 0.0001554547877518963, + "loss": 0.8622, + "step": 19225 + }, + { + "epoch": 1.3341197696204288, + "grad_norm": 1.7382524013519287, + "learning_rate": 0.00015545011212327777, + "loss": 0.6296, + "step": 19226 + }, + { + "epoch": 1.3341891610575254, + "grad_norm": 1.7659226655960083, + "learning_rate": 0.0001554454363196097, + "loss": 0.9005, + "step": 19227 + }, + { + "epoch": 1.3342585524946222, + "grad_norm": 1.4966552257537842, + "learning_rate": 0.0001554407603409068, + "loss": 0.6741, + "step": 19228 + }, + { + "epoch": 1.3343279439317188, + "grad_norm": 1.833400011062622, + "learning_rate": 0.00015543608418718382, + "loss": 0.6108, + "step": 19229 + }, + { + "epoch": 1.3343973353688154, + "grad_norm": 1.4071242809295654, + "learning_rate": 0.0001554314078584556, + "loss": 0.57, + "step": 19230 + }, + { + "epoch": 1.3344667268059123, + "grad_norm": 1.723198413848877, + "learning_rate": 0.00015542673135473685, + "loss": 0.6231, + "step": 19231 + }, + { + "epoch": 1.3345361182430089, + "grad_norm": 2.053314685821533, + "learning_rate": 0.00015542205467604232, + "loss": 0.5617, + "step": 19232 + }, + { + "epoch": 1.3346055096801055, + "grad_norm": 2.110154390335083, + "learning_rate": 0.0001554173778223868, + "loss": 0.4385, + "step": 19233 + }, + { + "epoch": 1.334674901117202, + "grad_norm": 2.077432155609131, + "learning_rate": 0.00015541270079378502, + "loss": 0.5935, + "step": 19234 + }, + { + "epoch": 1.3347442925542987, + "grad_norm": 1.6840258836746216, + "learning_rate": 0.00015540802359025185, + "loss": 0.6291, + "step": 19235 + }, + { + "epoch": 1.3348136839913955, + "grad_norm": 1.2122588157653809, + "learning_rate": 0.0001554033462118019, + "loss": 0.5303, + "step": 19236 + }, + { + "epoch": 1.3348830754284922, + "grad_norm": 1.4869962930679321, + "learning_rate": 0.00015539866865845007, + "loss": 0.4579, + "step": 19237 + }, + { + "epoch": 1.3349524668655888, + "grad_norm": 1.884350061416626, + "learning_rate": 0.00015539399093021102, + "loss": 0.7244, + "step": 19238 + }, + { + "epoch": 1.3350218583026854, + "grad_norm": 2.1796398162841797, + "learning_rate": 0.0001553893130270996, + "loss": 0.758, + "step": 19239 + }, + { + "epoch": 1.335091249739782, + "grad_norm": 1.5526601076126099, + "learning_rate": 0.00015538463494913055, + "loss": 0.5433, + "step": 19240 + }, + { + "epoch": 1.3351606411768788, + "grad_norm": 1.5998224020004272, + "learning_rate": 0.0001553799566963186, + "loss": 0.6968, + "step": 19241 + }, + { + "epoch": 1.3352300326139754, + "grad_norm": 1.540801763534546, + "learning_rate": 0.0001553752782686786, + "loss": 0.582, + "step": 19242 + }, + { + "epoch": 1.335299424051072, + "grad_norm": 2.0156712532043457, + "learning_rate": 0.00015537059966622523, + "loss": 0.731, + "step": 19243 + }, + { + "epoch": 1.3353688154881689, + "grad_norm": 2.0211822986602783, + "learning_rate": 0.00015536592088897329, + "loss": 0.6973, + "step": 19244 + }, + { + "epoch": 1.3354382069252655, + "grad_norm": 1.711868405342102, + "learning_rate": 0.0001553612419369376, + "loss": 0.7891, + "step": 19245 + }, + { + "epoch": 1.335507598362362, + "grad_norm": 2.0443003177642822, + "learning_rate": 0.00015535656281013286, + "loss": 0.8706, + "step": 19246 + }, + { + "epoch": 1.3355769897994587, + "grad_norm": 1.4855413436889648, + "learning_rate": 0.00015535188350857392, + "loss": 0.7271, + "step": 19247 + }, + { + "epoch": 1.3356463812365553, + "grad_norm": 1.488874077796936, + "learning_rate": 0.00015534720403227547, + "loss": 0.5808, + "step": 19248 + }, + { + "epoch": 1.3357157726736522, + "grad_norm": 1.3642743825912476, + "learning_rate": 0.00015534252438125233, + "loss": 0.4759, + "step": 19249 + }, + { + "epoch": 1.3357851641107488, + "grad_norm": 1.5971280336380005, + "learning_rate": 0.00015533784455551926, + "loss": 0.4701, + "step": 19250 + }, + { + "epoch": 1.3358545555478454, + "grad_norm": 1.4442973136901855, + "learning_rate": 0.00015533316455509103, + "loss": 0.6047, + "step": 19251 + }, + { + "epoch": 1.335923946984942, + "grad_norm": 1.6481611728668213, + "learning_rate": 0.00015532848437998243, + "loss": 0.7685, + "step": 19252 + }, + { + "epoch": 1.3359933384220386, + "grad_norm": 1.556579351425171, + "learning_rate": 0.0001553238040302082, + "loss": 0.644, + "step": 19253 + }, + { + "epoch": 1.3360627298591354, + "grad_norm": 1.5579766035079956, + "learning_rate": 0.00015531912350578314, + "loss": 0.5925, + "step": 19254 + }, + { + "epoch": 1.336132121296232, + "grad_norm": 1.6455206871032715, + "learning_rate": 0.00015531444280672205, + "loss": 0.7867, + "step": 19255 + }, + { + "epoch": 1.3362015127333287, + "grad_norm": 1.4748080968856812, + "learning_rate": 0.00015530976193303968, + "loss": 0.6091, + "step": 19256 + }, + { + "epoch": 1.3362709041704255, + "grad_norm": 2.331742286682129, + "learning_rate": 0.0001553050808847508, + "loss": 0.8013, + "step": 19257 + }, + { + "epoch": 1.336340295607522, + "grad_norm": 1.5034900903701782, + "learning_rate": 0.0001553003996618702, + "loss": 0.63, + "step": 19258 + }, + { + "epoch": 1.3364096870446187, + "grad_norm": 4.817315578460693, + "learning_rate": 0.00015529571826441266, + "loss": 0.6094, + "step": 19259 + }, + { + "epoch": 1.3364790784817153, + "grad_norm": 1.3402833938598633, + "learning_rate": 0.00015529103669239294, + "loss": 0.5936, + "step": 19260 + }, + { + "epoch": 1.336548469918812, + "grad_norm": 1.8171497583389282, + "learning_rate": 0.00015528635494582583, + "loss": 0.7673, + "step": 19261 + }, + { + "epoch": 1.3366178613559088, + "grad_norm": 1.7948296070098877, + "learning_rate": 0.00015528167302472613, + "loss": 0.616, + "step": 19262 + }, + { + "epoch": 1.3366872527930054, + "grad_norm": 1.3732025623321533, + "learning_rate": 0.00015527699092910858, + "loss": 0.3466, + "step": 19263 + }, + { + "epoch": 1.336756644230102, + "grad_norm": 1.5746451616287231, + "learning_rate": 0.00015527230865898802, + "loss": 0.6829, + "step": 19264 + }, + { + "epoch": 1.3368260356671986, + "grad_norm": 1.733426809310913, + "learning_rate": 0.00015526762621437916, + "loss": 0.6906, + "step": 19265 + }, + { + "epoch": 1.3368954271042952, + "grad_norm": 1.667182445526123, + "learning_rate": 0.0001552629435952968, + "loss": 0.6219, + "step": 19266 + }, + { + "epoch": 1.336964818541392, + "grad_norm": 2.11893367767334, + "learning_rate": 0.00015525826080175578, + "loss": 0.7353, + "step": 19267 + }, + { + "epoch": 1.3370342099784887, + "grad_norm": 2.184157371520996, + "learning_rate": 0.00015525357783377082, + "loss": 0.8195, + "step": 19268 + }, + { + "epoch": 1.3371036014155853, + "grad_norm": 1.6204733848571777, + "learning_rate": 0.00015524889469135676, + "loss": 0.5014, + "step": 19269 + }, + { + "epoch": 1.337172992852682, + "grad_norm": 1.3794333934783936, + "learning_rate": 0.00015524421137452828, + "loss": 0.489, + "step": 19270 + }, + { + "epoch": 1.3372423842897787, + "grad_norm": 1.769325852394104, + "learning_rate": 0.00015523952788330027, + "loss": 0.547, + "step": 19271 + }, + { + "epoch": 1.3373117757268753, + "grad_norm": 1.467291235923767, + "learning_rate": 0.0001552348442176875, + "loss": 0.7487, + "step": 19272 + }, + { + "epoch": 1.337381167163972, + "grad_norm": 1.6781363487243652, + "learning_rate": 0.0001552301603777047, + "loss": 0.6721, + "step": 19273 + }, + { + "epoch": 1.3374505586010685, + "grad_norm": 2.060277223587036, + "learning_rate": 0.00015522547636336674, + "loss": 0.8337, + "step": 19274 + }, + { + "epoch": 1.3375199500381654, + "grad_norm": 2.064476728439331, + "learning_rate": 0.0001552207921746883, + "loss": 0.8825, + "step": 19275 + }, + { + "epoch": 1.337589341475262, + "grad_norm": 1.7308279275894165, + "learning_rate": 0.00015521610781168426, + "loss": 0.619, + "step": 19276 + }, + { + "epoch": 1.3376587329123586, + "grad_norm": 1.495762586593628, + "learning_rate": 0.00015521142327436937, + "loss": 0.5152, + "step": 19277 + }, + { + "epoch": 1.3377281243494552, + "grad_norm": 1.604895830154419, + "learning_rate": 0.00015520673856275841, + "loss": 0.7267, + "step": 19278 + }, + { + "epoch": 1.3377975157865518, + "grad_norm": 1.9745622873306274, + "learning_rate": 0.0001552020536768662, + "loss": 0.8742, + "step": 19279 + }, + { + "epoch": 1.3378669072236486, + "grad_norm": 1.715224027633667, + "learning_rate": 0.0001551973686167075, + "loss": 0.6401, + "step": 19280 + }, + { + "epoch": 1.3379362986607453, + "grad_norm": 1.761698603630066, + "learning_rate": 0.00015519268338229714, + "loss": 0.6628, + "step": 19281 + }, + { + "epoch": 1.3380056900978419, + "grad_norm": 1.9339836835861206, + "learning_rate": 0.00015518799797364984, + "loss": 0.761, + "step": 19282 + }, + { + "epoch": 1.3380750815349387, + "grad_norm": 2.922729969024658, + "learning_rate": 0.0001551833123907805, + "loss": 0.6989, + "step": 19283 + }, + { + "epoch": 1.3381444729720353, + "grad_norm": 1.9118399620056152, + "learning_rate": 0.00015517862663370376, + "loss": 0.917, + "step": 19284 + }, + { + "epoch": 1.338213864409132, + "grad_norm": 1.6159915924072266, + "learning_rate": 0.00015517394070243453, + "loss": 0.6311, + "step": 19285 + }, + { + "epoch": 1.3382832558462285, + "grad_norm": 1.6040127277374268, + "learning_rate": 0.00015516925459698757, + "loss": 0.6118, + "step": 19286 + }, + { + "epoch": 1.3383526472833251, + "grad_norm": 1.7973488569259644, + "learning_rate": 0.0001551645683173777, + "loss": 0.7329, + "step": 19287 + }, + { + "epoch": 1.338422038720422, + "grad_norm": 2.1089842319488525, + "learning_rate": 0.00015515988186361965, + "loss": 0.442, + "step": 19288 + }, + { + "epoch": 1.3384914301575186, + "grad_norm": 1.6563376188278198, + "learning_rate": 0.00015515519523572825, + "loss": 0.7287, + "step": 19289 + }, + { + "epoch": 1.3385608215946152, + "grad_norm": 1.3939110040664673, + "learning_rate": 0.00015515050843371833, + "loss": 0.6937, + "step": 19290 + }, + { + "epoch": 1.3386302130317118, + "grad_norm": 1.2675600051879883, + "learning_rate": 0.00015514582145760466, + "loss": 0.5755, + "step": 19291 + }, + { + "epoch": 1.3386996044688084, + "grad_norm": 1.6704938411712646, + "learning_rate": 0.00015514113430740198, + "loss": 0.5681, + "step": 19292 + }, + { + "epoch": 1.3387689959059053, + "grad_norm": 1.775806188583374, + "learning_rate": 0.00015513644698312518, + "loss": 0.604, + "step": 19293 + }, + { + "epoch": 1.3388383873430019, + "grad_norm": 1.792062520980835, + "learning_rate": 0.00015513175948478903, + "loss": 0.8534, + "step": 19294 + }, + { + "epoch": 1.3389077787800985, + "grad_norm": 1.2254400253295898, + "learning_rate": 0.00015512707181240825, + "loss": 0.4268, + "step": 19295 + }, + { + "epoch": 1.3389771702171953, + "grad_norm": 1.7562097311019897, + "learning_rate": 0.00015512238396599775, + "loss": 0.7932, + "step": 19296 + }, + { + "epoch": 1.339046561654292, + "grad_norm": 1.3119065761566162, + "learning_rate": 0.00015511769594557225, + "loss": 0.4698, + "step": 19297 + }, + { + "epoch": 1.3391159530913885, + "grad_norm": 1.9126100540161133, + "learning_rate": 0.00015511300775114658, + "loss": 0.7446, + "step": 19298 + }, + { + "epoch": 1.3391853445284851, + "grad_norm": 1.554699182510376, + "learning_rate": 0.00015510831938273556, + "loss": 0.7381, + "step": 19299 + }, + { + "epoch": 1.3392547359655818, + "grad_norm": 2.109156370162964, + "learning_rate": 0.00015510363084035396, + "loss": 0.5231, + "step": 19300 + }, + { + "epoch": 1.3393241274026786, + "grad_norm": 2.0324392318725586, + "learning_rate": 0.00015509894212401658, + "loss": 0.6206, + "step": 19301 + }, + { + "epoch": 1.3393935188397752, + "grad_norm": 1.9912875890731812, + "learning_rate": 0.00015509425323373825, + "loss": 0.6948, + "step": 19302 + }, + { + "epoch": 1.3394629102768718, + "grad_norm": 1.6754306554794312, + "learning_rate": 0.00015508956416953375, + "loss": 0.6344, + "step": 19303 + }, + { + "epoch": 1.3395323017139684, + "grad_norm": 1.6140410900115967, + "learning_rate": 0.0001550848749314179, + "loss": 0.4157, + "step": 19304 + }, + { + "epoch": 1.339601693151065, + "grad_norm": 3.0001637935638428, + "learning_rate": 0.00015508018551940546, + "loss": 0.8073, + "step": 19305 + }, + { + "epoch": 1.3396710845881619, + "grad_norm": 1.540703535079956, + "learning_rate": 0.00015507549593351126, + "loss": 0.6308, + "step": 19306 + }, + { + "epoch": 1.3397404760252585, + "grad_norm": 1.944834589958191, + "learning_rate": 0.00015507080617375017, + "loss": 0.7083, + "step": 19307 + }, + { + "epoch": 1.339809867462355, + "grad_norm": 1.380335807800293, + "learning_rate": 0.00015506611624013688, + "loss": 0.5937, + "step": 19308 + }, + { + "epoch": 1.339879258899452, + "grad_norm": 2.1967930793762207, + "learning_rate": 0.00015506142613268625, + "loss": 0.8491, + "step": 19309 + }, + { + "epoch": 1.3399486503365485, + "grad_norm": 1.9502506256103516, + "learning_rate": 0.0001550567358514131, + "loss": 0.599, + "step": 19310 + }, + { + "epoch": 1.3400180417736451, + "grad_norm": 1.7438825368881226, + "learning_rate": 0.00015505204539633223, + "loss": 0.7269, + "step": 19311 + }, + { + "epoch": 1.3400874332107418, + "grad_norm": 2.0501585006713867, + "learning_rate": 0.00015504735476745842, + "loss": 0.7547, + "step": 19312 + }, + { + "epoch": 1.3401568246478384, + "grad_norm": 1.948089599609375, + "learning_rate": 0.0001550426639648065, + "loss": 0.8482, + "step": 19313 + }, + { + "epoch": 1.3402262160849352, + "grad_norm": 2.117161989212036, + "learning_rate": 0.00015503797298839129, + "loss": 0.5458, + "step": 19314 + }, + { + "epoch": 1.3402956075220318, + "grad_norm": 1.957937479019165, + "learning_rate": 0.0001550332818382276, + "loss": 0.6597, + "step": 19315 + }, + { + "epoch": 1.3403649989591284, + "grad_norm": 1.8056703805923462, + "learning_rate": 0.00015502859051433018, + "loss": 0.803, + "step": 19316 + }, + { + "epoch": 1.340434390396225, + "grad_norm": 1.7389860153198242, + "learning_rate": 0.00015502389901671392, + "loss": 0.545, + "step": 19317 + }, + { + "epoch": 1.3405037818333216, + "grad_norm": 2.0301709175109863, + "learning_rate": 0.0001550192073453936, + "loss": 0.6635, + "step": 19318 + }, + { + "epoch": 1.3405731732704185, + "grad_norm": 1.6923160552978516, + "learning_rate": 0.000155014515500384, + "loss": 0.6341, + "step": 19319 + }, + { + "epoch": 1.340642564707515, + "grad_norm": 2.0715365409851074, + "learning_rate": 0.00015500982348169994, + "loss": 0.7569, + "step": 19320 + }, + { + "epoch": 1.3407119561446117, + "grad_norm": 1.570729374885559, + "learning_rate": 0.0001550051312893563, + "loss": 0.3756, + "step": 19321 + }, + { + "epoch": 1.3407813475817085, + "grad_norm": 1.4525902271270752, + "learning_rate": 0.0001550004389233678, + "loss": 0.4603, + "step": 19322 + }, + { + "epoch": 1.3408507390188051, + "grad_norm": 1.6868425607681274, + "learning_rate": 0.00015499574638374932, + "loss": 0.6655, + "step": 19323 + }, + { + "epoch": 1.3409201304559017, + "grad_norm": 1.5877898931503296, + "learning_rate": 0.00015499105367051564, + "loss": 0.4685, + "step": 19324 + }, + { + "epoch": 1.3409895218929984, + "grad_norm": 1.408666729927063, + "learning_rate": 0.00015498636078368157, + "loss": 0.5386, + "step": 19325 + }, + { + "epoch": 1.341058913330095, + "grad_norm": 2.0359625816345215, + "learning_rate": 0.00015498166772326196, + "loss": 0.6788, + "step": 19326 + }, + { + "epoch": 1.3411283047671918, + "grad_norm": 1.2391669750213623, + "learning_rate": 0.0001549769744892716, + "loss": 0.508, + "step": 19327 + }, + { + "epoch": 1.3411976962042884, + "grad_norm": 1.8739501237869263, + "learning_rate": 0.00015497228108172535, + "loss": 0.696, + "step": 19328 + }, + { + "epoch": 1.341267087641385, + "grad_norm": 1.6311767101287842, + "learning_rate": 0.00015496758750063792, + "loss": 0.7984, + "step": 19329 + }, + { + "epoch": 1.3413364790784816, + "grad_norm": 1.5914374589920044, + "learning_rate": 0.0001549628937460242, + "loss": 0.5031, + "step": 19330 + }, + { + "epoch": 1.3414058705155782, + "grad_norm": 1.605855941772461, + "learning_rate": 0.00015495819981789906, + "loss": 0.5435, + "step": 19331 + }, + { + "epoch": 1.341475261952675, + "grad_norm": 1.7513532638549805, + "learning_rate": 0.0001549535057162772, + "loss": 0.956, + "step": 19332 + }, + { + "epoch": 1.3415446533897717, + "grad_norm": 1.7672648429870605, + "learning_rate": 0.00015494881144117353, + "loss": 0.7163, + "step": 19333 + }, + { + "epoch": 1.3416140448268683, + "grad_norm": 1.655411958694458, + "learning_rate": 0.00015494411699260282, + "loss": 0.7619, + "step": 19334 + }, + { + "epoch": 1.3416834362639651, + "grad_norm": 2.007425308227539, + "learning_rate": 0.00015493942237057993, + "loss": 0.7875, + "step": 19335 + }, + { + "epoch": 1.3417528277010617, + "grad_norm": 1.6863564252853394, + "learning_rate": 0.0001549347275751196, + "loss": 0.6893, + "step": 19336 + }, + { + "epoch": 1.3418222191381584, + "grad_norm": 2.354907274246216, + "learning_rate": 0.00015493003260623678, + "loss": 0.5994, + "step": 19337 + }, + { + "epoch": 1.341891610575255, + "grad_norm": 1.999159336090088, + "learning_rate": 0.0001549253374639462, + "loss": 0.5558, + "step": 19338 + }, + { + "epoch": 1.3419610020123516, + "grad_norm": 1.7498451471328735, + "learning_rate": 0.0001549206421482627, + "loss": 0.5487, + "step": 19339 + }, + { + "epoch": 1.3420303934494484, + "grad_norm": 1.7437067031860352, + "learning_rate": 0.00015491594665920108, + "loss": 0.8914, + "step": 19340 + }, + { + "epoch": 1.342099784886545, + "grad_norm": 3.3511524200439453, + "learning_rate": 0.0001549112509967762, + "loss": 0.9358, + "step": 19341 + }, + { + "epoch": 1.3421691763236416, + "grad_norm": 1.3287080526351929, + "learning_rate": 0.0001549065551610029, + "loss": 0.5754, + "step": 19342 + }, + { + "epoch": 1.3422385677607382, + "grad_norm": 2.000277042388916, + "learning_rate": 0.00015490185915189593, + "loss": 0.741, + "step": 19343 + }, + { + "epoch": 1.3423079591978349, + "grad_norm": 2.00504994392395, + "learning_rate": 0.0001548971629694702, + "loss": 0.812, + "step": 19344 + }, + { + "epoch": 1.3423773506349317, + "grad_norm": 1.4524641036987305, + "learning_rate": 0.0001548924666137405, + "loss": 0.5132, + "step": 19345 + }, + { + "epoch": 1.3424467420720283, + "grad_norm": 1.7240349054336548, + "learning_rate": 0.00015488777008472159, + "loss": 0.5402, + "step": 19346 + }, + { + "epoch": 1.342516133509125, + "grad_norm": 1.6372482776641846, + "learning_rate": 0.0001548830733824284, + "loss": 0.8301, + "step": 19347 + }, + { + "epoch": 1.3425855249462217, + "grad_norm": 1.509171724319458, + "learning_rate": 0.0001548783765068757, + "loss": 0.551, + "step": 19348 + }, + { + "epoch": 1.3426549163833184, + "grad_norm": 1.7556076049804688, + "learning_rate": 0.00015487367945807838, + "loss": 0.6772, + "step": 19349 + }, + { + "epoch": 1.342724307820415, + "grad_norm": 1.54200279712677, + "learning_rate": 0.00015486898223605118, + "loss": 0.5667, + "step": 19350 + }, + { + "epoch": 1.3427936992575116, + "grad_norm": 2.436413049697876, + "learning_rate": 0.00015486428484080898, + "loss": 0.7946, + "step": 19351 + }, + { + "epoch": 1.3428630906946082, + "grad_norm": 1.7353708744049072, + "learning_rate": 0.0001548595872723666, + "loss": 0.76, + "step": 19352 + }, + { + "epoch": 1.342932482131705, + "grad_norm": 1.4724904298782349, + "learning_rate": 0.00015485488953073884, + "loss": 0.602, + "step": 19353 + }, + { + "epoch": 1.3430018735688016, + "grad_norm": 2.507298231124878, + "learning_rate": 0.00015485019161594056, + "loss": 0.9862, + "step": 19354 + }, + { + "epoch": 1.3430712650058982, + "grad_norm": 1.5289478302001953, + "learning_rate": 0.0001548454935279866, + "loss": 0.6658, + "step": 19355 + }, + { + "epoch": 1.3431406564429949, + "grad_norm": 1.4083492755889893, + "learning_rate": 0.0001548407952668918, + "loss": 0.4862, + "step": 19356 + }, + { + "epoch": 1.3432100478800915, + "grad_norm": 2.0368032455444336, + "learning_rate": 0.00015483609683267097, + "loss": 0.6303, + "step": 19357 + }, + { + "epoch": 1.3432794393171883, + "grad_norm": 2.3811962604522705, + "learning_rate": 0.00015483139822533893, + "loss": 0.7341, + "step": 19358 + }, + { + "epoch": 1.343348830754285, + "grad_norm": 1.9824459552764893, + "learning_rate": 0.00015482669944491054, + "loss": 0.8168, + "step": 19359 + }, + { + "epoch": 1.3434182221913815, + "grad_norm": 1.7495397329330444, + "learning_rate": 0.0001548220004914006, + "loss": 0.6395, + "step": 19360 + }, + { + "epoch": 1.3434876136284783, + "grad_norm": 2.1728708744049072, + "learning_rate": 0.00015481730136482398, + "loss": 0.805, + "step": 19361 + }, + { + "epoch": 1.343557005065575, + "grad_norm": 2.3137400150299072, + "learning_rate": 0.00015481260206519552, + "loss": 0.6683, + "step": 19362 + }, + { + "epoch": 1.3436263965026716, + "grad_norm": 1.712731122970581, + "learning_rate": 0.00015480790259253, + "loss": 0.6461, + "step": 19363 + }, + { + "epoch": 1.3436957879397682, + "grad_norm": 2.1629955768585205, + "learning_rate": 0.0001548032029468423, + "loss": 0.5892, + "step": 19364 + }, + { + "epoch": 1.3437651793768648, + "grad_norm": 2.2253570556640625, + "learning_rate": 0.00015479850312814726, + "loss": 1.1128, + "step": 19365 + }, + { + "epoch": 1.3438345708139616, + "grad_norm": 1.4025059938430786, + "learning_rate": 0.0001547938031364597, + "loss": 0.5166, + "step": 19366 + }, + { + "epoch": 1.3439039622510582, + "grad_norm": 1.5644820928573608, + "learning_rate": 0.00015478910297179446, + "loss": 0.5828, + "step": 19367 + }, + { + "epoch": 1.3439733536881548, + "grad_norm": 2.4456188678741455, + "learning_rate": 0.00015478440263416636, + "loss": 0.6648, + "step": 19368 + }, + { + "epoch": 1.3440427451252517, + "grad_norm": 1.6156965494155884, + "learning_rate": 0.00015477970212359024, + "loss": 0.7261, + "step": 19369 + }, + { + "epoch": 1.344112136562348, + "grad_norm": 1.730194091796875, + "learning_rate": 0.000154775001440081, + "loss": 0.6126, + "step": 19370 + }, + { + "epoch": 1.344181527999445, + "grad_norm": 1.9495495557785034, + "learning_rate": 0.00015477030058365344, + "loss": 0.6828, + "step": 19371 + }, + { + "epoch": 1.3442509194365415, + "grad_norm": 1.818607211112976, + "learning_rate": 0.00015476559955432238, + "loss": 0.4801, + "step": 19372 + }, + { + "epoch": 1.3443203108736381, + "grad_norm": 1.0900081396102905, + "learning_rate": 0.00015476089835210268, + "loss": 0.5398, + "step": 19373 + }, + { + "epoch": 1.344389702310735, + "grad_norm": 1.828843116760254, + "learning_rate": 0.00015475619697700915, + "loss": 0.5887, + "step": 19374 + }, + { + "epoch": 1.3444590937478316, + "grad_norm": 1.8002710342407227, + "learning_rate": 0.00015475149542905668, + "loss": 0.5674, + "step": 19375 + }, + { + "epoch": 1.3445284851849282, + "grad_norm": 1.6473373174667358, + "learning_rate": 0.0001547467937082601, + "loss": 0.851, + "step": 19376 + }, + { + "epoch": 1.3445978766220248, + "grad_norm": 1.98256516456604, + "learning_rate": 0.00015474209181463424, + "loss": 0.552, + "step": 19377 + }, + { + "epoch": 1.3446672680591214, + "grad_norm": 2.1644396781921387, + "learning_rate": 0.00015473738974819393, + "loss": 0.7615, + "step": 19378 + }, + { + "epoch": 1.3447366594962182, + "grad_norm": 2.865816354751587, + "learning_rate": 0.00015473268750895405, + "loss": 0.8, + "step": 19379 + }, + { + "epoch": 1.3448060509333148, + "grad_norm": 2.2392473220825195, + "learning_rate": 0.00015472798509692941, + "loss": 0.7322, + "step": 19380 + }, + { + "epoch": 1.3448754423704115, + "grad_norm": 1.5485953092575073, + "learning_rate": 0.0001547232825121349, + "loss": 0.6654, + "step": 19381 + }, + { + "epoch": 1.3449448338075083, + "grad_norm": 2.289372444152832, + "learning_rate": 0.00015471857975458527, + "loss": 0.6521, + "step": 19382 + }, + { + "epoch": 1.3450142252446047, + "grad_norm": 1.5197155475616455, + "learning_rate": 0.0001547138768242955, + "loss": 0.5661, + "step": 19383 + }, + { + "epoch": 1.3450836166817015, + "grad_norm": 2.988741397857666, + "learning_rate": 0.00015470917372128036, + "loss": 0.6066, + "step": 19384 + }, + { + "epoch": 1.3451530081187981, + "grad_norm": 1.1864887475967407, + "learning_rate": 0.00015470447044555467, + "loss": 0.4507, + "step": 19385 + }, + { + "epoch": 1.3452223995558947, + "grad_norm": 2.050124168395996, + "learning_rate": 0.00015469976699713334, + "loss": 0.7779, + "step": 19386 + }, + { + "epoch": 1.3452917909929916, + "grad_norm": 1.2901017665863037, + "learning_rate": 0.00015469506337603122, + "loss": 0.53, + "step": 19387 + }, + { + "epoch": 1.3453611824300882, + "grad_norm": 1.5895733833312988, + "learning_rate": 0.00015469035958226308, + "loss": 0.5878, + "step": 19388 + }, + { + "epoch": 1.3454305738671848, + "grad_norm": 2.2700769901275635, + "learning_rate": 0.00015468565561584381, + "loss": 0.7117, + "step": 19389 + }, + { + "epoch": 1.3454999653042814, + "grad_norm": 1.6965289115905762, + "learning_rate": 0.0001546809514767883, + "loss": 0.8072, + "step": 19390 + }, + { + "epoch": 1.345569356741378, + "grad_norm": 1.9313613176345825, + "learning_rate": 0.0001546762471651114, + "loss": 0.8524, + "step": 19391 + }, + { + "epoch": 1.3456387481784748, + "grad_norm": 2.1417858600616455, + "learning_rate": 0.0001546715426808279, + "loss": 0.6631, + "step": 19392 + }, + { + "epoch": 1.3457081396155715, + "grad_norm": 2.232856512069702, + "learning_rate": 0.00015466683802395267, + "loss": 0.826, + "step": 19393 + }, + { + "epoch": 1.345777531052668, + "grad_norm": 3.2781481742858887, + "learning_rate": 0.0001546621331945006, + "loss": 0.9191, + "step": 19394 + }, + { + "epoch": 1.345846922489765, + "grad_norm": 1.7317954301834106, + "learning_rate": 0.0001546574281924865, + "loss": 0.6917, + "step": 19395 + }, + { + "epoch": 1.3459163139268615, + "grad_norm": 1.9985297918319702, + "learning_rate": 0.00015465272301792523, + "loss": 0.82, + "step": 19396 + }, + { + "epoch": 1.3459857053639581, + "grad_norm": 1.8262494802474976, + "learning_rate": 0.0001546480176708317, + "loss": 0.4683, + "step": 19397 + }, + { + "epoch": 1.3460550968010547, + "grad_norm": 1.600968599319458, + "learning_rate": 0.0001546433121512207, + "loss": 0.5251, + "step": 19398 + }, + { + "epoch": 1.3461244882381513, + "grad_norm": 1.6083934307098389, + "learning_rate": 0.00015463860645910707, + "loss": 0.5971, + "step": 19399 + }, + { + "epoch": 1.3461938796752482, + "grad_norm": 2.3153154850006104, + "learning_rate": 0.0001546339005945057, + "loss": 0.6514, + "step": 19400 + }, + { + "epoch": 1.3462632711123448, + "grad_norm": 1.187263011932373, + "learning_rate": 0.0001546291945574315, + "loss": 0.5876, + "step": 19401 + }, + { + "epoch": 1.3463326625494414, + "grad_norm": 1.831898808479309, + "learning_rate": 0.00015462448834789922, + "loss": 0.6491, + "step": 19402 + }, + { + "epoch": 1.346402053986538, + "grad_norm": 1.9067384004592896, + "learning_rate": 0.00015461978196592378, + "loss": 0.7069, + "step": 19403 + }, + { + "epoch": 1.3464714454236346, + "grad_norm": 1.4407583475112915, + "learning_rate": 0.00015461507541152006, + "loss": 0.6646, + "step": 19404 + }, + { + "epoch": 1.3465408368607314, + "grad_norm": 1.9224135875701904, + "learning_rate": 0.00015461036868470282, + "loss": 0.7881, + "step": 19405 + }, + { + "epoch": 1.346610228297828, + "grad_norm": 1.7560153007507324, + "learning_rate": 0.00015460566178548702, + "loss": 0.6762, + "step": 19406 + }, + { + "epoch": 1.3466796197349247, + "grad_norm": 2.0365068912506104, + "learning_rate": 0.0001546009547138875, + "loss": 0.8456, + "step": 19407 + }, + { + "epoch": 1.3467490111720215, + "grad_norm": 2.1140477657318115, + "learning_rate": 0.00015459624746991908, + "loss": 0.5977, + "step": 19408 + }, + { + "epoch": 1.3468184026091181, + "grad_norm": 1.6084144115447998, + "learning_rate": 0.0001545915400535966, + "loss": 0.7085, + "step": 19409 + }, + { + "epoch": 1.3468877940462147, + "grad_norm": 1.4850267171859741, + "learning_rate": 0.00015458683246493503, + "loss": 0.635, + "step": 19410 + }, + { + "epoch": 1.3469571854833113, + "grad_norm": 1.7882994413375854, + "learning_rate": 0.00015458212470394913, + "loss": 0.7104, + "step": 19411 + }, + { + "epoch": 1.347026576920408, + "grad_norm": 2.3467421531677246, + "learning_rate": 0.00015457741677065378, + "loss": 0.7661, + "step": 19412 + }, + { + "epoch": 1.3470959683575048, + "grad_norm": 1.5656254291534424, + "learning_rate": 0.00015457270866506387, + "loss": 0.6697, + "step": 19413 + }, + { + "epoch": 1.3471653597946014, + "grad_norm": 1.8999425172805786, + "learning_rate": 0.00015456800038719427, + "loss": 0.6885, + "step": 19414 + }, + { + "epoch": 1.347234751231698, + "grad_norm": 2.425583839416504, + "learning_rate": 0.00015456329193705982, + "loss": 0.6604, + "step": 19415 + }, + { + "epoch": 1.3473041426687946, + "grad_norm": 1.2300440073013306, + "learning_rate": 0.00015455858331467538, + "loss": 0.6153, + "step": 19416 + }, + { + "epoch": 1.3473735341058912, + "grad_norm": 1.6331048011779785, + "learning_rate": 0.00015455387452005583, + "loss": 0.7892, + "step": 19417 + }, + { + "epoch": 1.347442925542988, + "grad_norm": 1.3231210708618164, + "learning_rate": 0.00015454916555321602, + "loss": 0.4848, + "step": 19418 + }, + { + "epoch": 1.3475123169800847, + "grad_norm": 2.0316081047058105, + "learning_rate": 0.0001545444564141708, + "loss": 0.6296, + "step": 19419 + }, + { + "epoch": 1.3475817084171813, + "grad_norm": 1.5236793756484985, + "learning_rate": 0.00015453974710293508, + "loss": 0.5214, + "step": 19420 + }, + { + "epoch": 1.3476510998542781, + "grad_norm": 1.6214532852172852, + "learning_rate": 0.00015453503761952373, + "loss": 0.7204, + "step": 19421 + }, + { + "epoch": 1.3477204912913747, + "grad_norm": 2.1017305850982666, + "learning_rate": 0.00015453032796395157, + "loss": 0.63, + "step": 19422 + }, + { + "epoch": 1.3477898827284713, + "grad_norm": 2.827223777770996, + "learning_rate": 0.00015452561813623347, + "loss": 0.5136, + "step": 19423 + }, + { + "epoch": 1.347859274165568, + "grad_norm": 1.8455065488815308, + "learning_rate": 0.00015452090813638435, + "loss": 0.6051, + "step": 19424 + }, + { + "epoch": 1.3479286656026646, + "grad_norm": 1.7996364831924438, + "learning_rate": 0.00015451619796441905, + "loss": 0.7952, + "step": 19425 + }, + { + "epoch": 1.3479980570397614, + "grad_norm": 2.9511797428131104, + "learning_rate": 0.00015451148762035243, + "loss": 0.8128, + "step": 19426 + }, + { + "epoch": 1.348067448476858, + "grad_norm": 2.0411205291748047, + "learning_rate": 0.00015450677710419935, + "loss": 0.6996, + "step": 19427 + }, + { + "epoch": 1.3481368399139546, + "grad_norm": 1.8749116659164429, + "learning_rate": 0.00015450206641597474, + "loss": 0.7275, + "step": 19428 + }, + { + "epoch": 1.3482062313510512, + "grad_norm": 1.8082376718521118, + "learning_rate": 0.0001544973555556934, + "loss": 0.5746, + "step": 19429 + }, + { + "epoch": 1.3482756227881478, + "grad_norm": 1.3573931455612183, + "learning_rate": 0.00015449264452337022, + "loss": 0.6109, + "step": 19430 + }, + { + "epoch": 1.3483450142252447, + "grad_norm": 2.1964621543884277, + "learning_rate": 0.0001544879333190201, + "loss": 0.6563, + "step": 19431 + }, + { + "epoch": 1.3484144056623413, + "grad_norm": 1.7605910301208496, + "learning_rate": 0.0001544832219426579, + "loss": 0.6562, + "step": 19432 + }, + { + "epoch": 1.3484837970994379, + "grad_norm": 1.5403262376785278, + "learning_rate": 0.0001544785103942985, + "loss": 0.719, + "step": 19433 + }, + { + "epoch": 1.3485531885365347, + "grad_norm": 1.5203344821929932, + "learning_rate": 0.00015447379867395672, + "loss": 0.5237, + "step": 19434 + }, + { + "epoch": 1.3486225799736313, + "grad_norm": 2.336707830429077, + "learning_rate": 0.00015446908678164753, + "loss": 0.7693, + "step": 19435 + }, + { + "epoch": 1.348691971410728, + "grad_norm": 2.1833412647247314, + "learning_rate": 0.00015446437471738572, + "loss": 0.805, + "step": 19436 + }, + { + "epoch": 1.3487613628478246, + "grad_norm": 1.9399827718734741, + "learning_rate": 0.0001544596624811862, + "loss": 0.6249, + "step": 19437 + }, + { + "epoch": 1.3488307542849212, + "grad_norm": 2.1851468086242676, + "learning_rate": 0.00015445495007306384, + "loss": 0.6956, + "step": 19438 + }, + { + "epoch": 1.348900145722018, + "grad_norm": 2.140702486038208, + "learning_rate": 0.00015445023749303353, + "loss": 0.7894, + "step": 19439 + }, + { + "epoch": 1.3489695371591146, + "grad_norm": 3.005040168762207, + "learning_rate": 0.00015444552474111017, + "loss": 0.9446, + "step": 19440 + }, + { + "epoch": 1.3490389285962112, + "grad_norm": 2.167117118835449, + "learning_rate": 0.00015444081181730857, + "loss": 0.8149, + "step": 19441 + }, + { + "epoch": 1.3491083200333078, + "grad_norm": 2.686293840408325, + "learning_rate": 0.00015443609872164364, + "loss": 0.6803, + "step": 19442 + }, + { + "epoch": 1.3491777114704044, + "grad_norm": 1.7399667501449585, + "learning_rate": 0.00015443138545413027, + "loss": 0.7674, + "step": 19443 + }, + { + "epoch": 1.3492471029075013, + "grad_norm": 1.7403863668441772, + "learning_rate": 0.00015442667201478333, + "loss": 0.6942, + "step": 19444 + }, + { + "epoch": 1.3493164943445979, + "grad_norm": 1.918662428855896, + "learning_rate": 0.00015442195840361772, + "loss": 0.7016, + "step": 19445 + }, + { + "epoch": 1.3493858857816945, + "grad_norm": 2.1636266708374023, + "learning_rate": 0.00015441724462064826, + "loss": 1.0144, + "step": 19446 + }, + { + "epoch": 1.3494552772187913, + "grad_norm": 3.0924606323242188, + "learning_rate": 0.0001544125306658899, + "loss": 0.712, + "step": 19447 + }, + { + "epoch": 1.349524668655888, + "grad_norm": 1.8245010375976562, + "learning_rate": 0.0001544078165393575, + "loss": 0.7527, + "step": 19448 + }, + { + "epoch": 1.3495940600929845, + "grad_norm": 1.8082945346832275, + "learning_rate": 0.0001544031022410659, + "loss": 0.5572, + "step": 19449 + }, + { + "epoch": 1.3496634515300812, + "grad_norm": 9.9871187210083, + "learning_rate": 0.00015439838777103008, + "loss": 0.5106, + "step": 19450 + }, + { + "epoch": 1.3497328429671778, + "grad_norm": 2.754678964614868, + "learning_rate": 0.0001543936731292648, + "loss": 0.7469, + "step": 19451 + }, + { + "epoch": 1.3498022344042746, + "grad_norm": 2.049619436264038, + "learning_rate": 0.00015438895831578506, + "loss": 0.6335, + "step": 19452 + }, + { + "epoch": 1.3498716258413712, + "grad_norm": 2.05002760887146, + "learning_rate": 0.00015438424333060567, + "loss": 0.8266, + "step": 19453 + }, + { + "epoch": 1.3499410172784678, + "grad_norm": 2.0244946479797363, + "learning_rate": 0.0001543795281737415, + "loss": 0.657, + "step": 19454 + }, + { + "epoch": 1.3500104087155644, + "grad_norm": 2.0641283988952637, + "learning_rate": 0.00015437481284520747, + "loss": 0.6083, + "step": 19455 + }, + { + "epoch": 1.350079800152661, + "grad_norm": 5.328135967254639, + "learning_rate": 0.0001543700973450185, + "loss": 0.5821, + "step": 19456 + }, + { + "epoch": 1.3501491915897579, + "grad_norm": 1.9290095567703247, + "learning_rate": 0.00015436538167318942, + "loss": 0.7248, + "step": 19457 + }, + { + "epoch": 1.3502185830268545, + "grad_norm": 2.169212818145752, + "learning_rate": 0.00015436066582973513, + "loss": 0.5868, + "step": 19458 + }, + { + "epoch": 1.350287974463951, + "grad_norm": 1.7662550210952759, + "learning_rate": 0.00015435594981467055, + "loss": 0.6394, + "step": 19459 + }, + { + "epoch": 1.350357365901048, + "grad_norm": 1.741206407546997, + "learning_rate": 0.00015435123362801055, + "loss": 0.5996, + "step": 19460 + }, + { + "epoch": 1.3504267573381445, + "grad_norm": 2.1940536499023438, + "learning_rate": 0.00015434651726977, + "loss": 0.727, + "step": 19461 + }, + { + "epoch": 1.3504961487752412, + "grad_norm": 2.4720406532287598, + "learning_rate": 0.0001543418007399638, + "loss": 0.8167, + "step": 19462 + }, + { + "epoch": 1.3505655402123378, + "grad_norm": 1.7581510543823242, + "learning_rate": 0.00015433708403860683, + "loss": 0.7305, + "step": 19463 + }, + { + "epoch": 1.3506349316494344, + "grad_norm": 2.1446962356567383, + "learning_rate": 0.00015433236716571398, + "loss": 0.5608, + "step": 19464 + }, + { + "epoch": 1.3507043230865312, + "grad_norm": 1.7852485179901123, + "learning_rate": 0.00015432765012130014, + "loss": 0.6625, + "step": 19465 + }, + { + "epoch": 1.3507737145236278, + "grad_norm": 2.0653092861175537, + "learning_rate": 0.00015432293290538026, + "loss": 0.679, + "step": 19466 + }, + { + "epoch": 1.3508431059607244, + "grad_norm": 1.777023196220398, + "learning_rate": 0.00015431821551796913, + "loss": 0.8088, + "step": 19467 + }, + { + "epoch": 1.350912497397821, + "grad_norm": 1.7123595476150513, + "learning_rate": 0.00015431349795908174, + "loss": 0.6801, + "step": 19468 + }, + { + "epoch": 1.3509818888349177, + "grad_norm": 1.9751256704330444, + "learning_rate": 0.0001543087802287329, + "loss": 0.5556, + "step": 19469 + }, + { + "epoch": 1.3510512802720145, + "grad_norm": 1.5205105543136597, + "learning_rate": 0.0001543040623269376, + "loss": 0.6326, + "step": 19470 + }, + { + "epoch": 1.351120671709111, + "grad_norm": 1.8116945028305054, + "learning_rate": 0.00015429934425371063, + "loss": 0.5768, + "step": 19471 + }, + { + "epoch": 1.3511900631462077, + "grad_norm": 1.7738779783248901, + "learning_rate": 0.00015429462600906696, + "loss": 0.525, + "step": 19472 + }, + { + "epoch": 1.3512594545833045, + "grad_norm": 1.639035940170288, + "learning_rate": 0.00015428990759302143, + "loss": 0.6053, + "step": 19473 + }, + { + "epoch": 1.3513288460204012, + "grad_norm": 2.206428050994873, + "learning_rate": 0.00015428518900558898, + "loss": 0.672, + "step": 19474 + }, + { + "epoch": 1.3513982374574978, + "grad_norm": 1.9019479751586914, + "learning_rate": 0.00015428047024678446, + "loss": 0.5773, + "step": 19475 + }, + { + "epoch": 1.3514676288945944, + "grad_norm": 2.1754536628723145, + "learning_rate": 0.0001542757513166228, + "loss": 0.6283, + "step": 19476 + }, + { + "epoch": 1.351537020331691, + "grad_norm": 2.5698816776275635, + "learning_rate": 0.0001542710322151189, + "loss": 0.576, + "step": 19477 + }, + { + "epoch": 1.3516064117687878, + "grad_norm": 2.4201278686523438, + "learning_rate": 0.00015426631294228763, + "loss": 0.7936, + "step": 19478 + }, + { + "epoch": 1.3516758032058844, + "grad_norm": 2.4099526405334473, + "learning_rate": 0.0001542615934981439, + "loss": 0.6895, + "step": 19479 + }, + { + "epoch": 1.351745194642981, + "grad_norm": 1.8362224102020264, + "learning_rate": 0.00015425687388270266, + "loss": 0.7342, + "step": 19480 + }, + { + "epoch": 1.3518145860800777, + "grad_norm": 1.497679591178894, + "learning_rate": 0.00015425215409597875, + "loss": 0.3536, + "step": 19481 + }, + { + "epoch": 1.3518839775171743, + "grad_norm": 2.631514310836792, + "learning_rate": 0.00015424743413798705, + "loss": 0.8293, + "step": 19482 + }, + { + "epoch": 1.351953368954271, + "grad_norm": 1.8721072673797607, + "learning_rate": 0.00015424271400874253, + "loss": 0.5474, + "step": 19483 + }, + { + "epoch": 1.3520227603913677, + "grad_norm": 1.826735496520996, + "learning_rate": 0.00015423799370826004, + "loss": 0.5273, + "step": 19484 + }, + { + "epoch": 1.3520921518284643, + "grad_norm": 1.7384834289550781, + "learning_rate": 0.00015423327323655452, + "loss": 0.53, + "step": 19485 + }, + { + "epoch": 1.3521615432655611, + "grad_norm": 1.8625314235687256, + "learning_rate": 0.00015422855259364078, + "loss": 0.8065, + "step": 19486 + }, + { + "epoch": 1.3522309347026578, + "grad_norm": 2.575774669647217, + "learning_rate": 0.00015422383177953388, + "loss": 0.7306, + "step": 19487 + }, + { + "epoch": 1.3523003261397544, + "grad_norm": 1.6191667318344116, + "learning_rate": 0.0001542191107942486, + "loss": 0.6895, + "step": 19488 + }, + { + "epoch": 1.352369717576851, + "grad_norm": 2.2010881900787354, + "learning_rate": 0.00015421438963779985, + "loss": 0.7335, + "step": 19489 + }, + { + "epoch": 1.3524391090139476, + "grad_norm": 1.8537429571151733, + "learning_rate": 0.0001542096683102026, + "loss": 0.7311, + "step": 19490 + }, + { + "epoch": 1.3525085004510444, + "grad_norm": 2.8329269886016846, + "learning_rate": 0.0001542049468114717, + "loss": 0.7149, + "step": 19491 + }, + { + "epoch": 1.352577891888141, + "grad_norm": 1.7434815168380737, + "learning_rate": 0.00015420022514162204, + "loss": 0.7256, + "step": 19492 + }, + { + "epoch": 1.3526472833252376, + "grad_norm": 2.419623374938965, + "learning_rate": 0.00015419550330066862, + "loss": 0.8129, + "step": 19493 + }, + { + "epoch": 1.3527166747623343, + "grad_norm": 1.8757399320602417, + "learning_rate": 0.00015419078128862626, + "loss": 0.5086, + "step": 19494 + }, + { + "epoch": 1.3527860661994309, + "grad_norm": 1.4104399681091309, + "learning_rate": 0.00015418605910550988, + "loss": 0.5092, + "step": 19495 + }, + { + "epoch": 1.3528554576365277, + "grad_norm": 2.0354113578796387, + "learning_rate": 0.00015418133675133437, + "loss": 0.5674, + "step": 19496 + }, + { + "epoch": 1.3529248490736243, + "grad_norm": 3.223029613494873, + "learning_rate": 0.00015417661422611472, + "loss": 0.8216, + "step": 19497 + }, + { + "epoch": 1.352994240510721, + "grad_norm": 1.6954680681228638, + "learning_rate": 0.00015417189152986579, + "loss": 0.5535, + "step": 19498 + }, + { + "epoch": 1.3530636319478178, + "grad_norm": 1.8113596439361572, + "learning_rate": 0.00015416716866260242, + "loss": 0.6507, + "step": 19499 + }, + { + "epoch": 1.3531330233849144, + "grad_norm": 1.7353205680847168, + "learning_rate": 0.00015416244562433963, + "loss": 0.5396, + "step": 19500 + }, + { + "epoch": 1.353202414822011, + "grad_norm": 2.0030956268310547, + "learning_rate": 0.00015415772241509225, + "loss": 0.5979, + "step": 19501 + }, + { + "epoch": 1.3532718062591076, + "grad_norm": 1.7207144498825073, + "learning_rate": 0.00015415299903487526, + "loss": 0.7661, + "step": 19502 + }, + { + "epoch": 1.3533411976962042, + "grad_norm": 1.5457788705825806, + "learning_rate": 0.00015414827548370353, + "loss": 0.5702, + "step": 19503 + }, + { + "epoch": 1.353410589133301, + "grad_norm": 2.1639292240142822, + "learning_rate": 0.000154143551761592, + "loss": 0.7443, + "step": 19504 + }, + { + "epoch": 1.3534799805703976, + "grad_norm": 1.7558174133300781, + "learning_rate": 0.00015413882786855548, + "loss": 0.5403, + "step": 19505 + }, + { + "epoch": 1.3535493720074943, + "grad_norm": 1.3843226432800293, + "learning_rate": 0.00015413410380460903, + "loss": 0.5022, + "step": 19506 + }, + { + "epoch": 1.3536187634445909, + "grad_norm": 1.7288998365402222, + "learning_rate": 0.00015412937956976746, + "loss": 0.5941, + "step": 19507 + }, + { + "epoch": 1.3536881548816875, + "grad_norm": 1.6325836181640625, + "learning_rate": 0.00015412465516404574, + "loss": 0.4563, + "step": 19508 + }, + { + "epoch": 1.3537575463187843, + "grad_norm": 2.4683303833007812, + "learning_rate": 0.00015411993058745874, + "loss": 0.7854, + "step": 19509 + }, + { + "epoch": 1.353826937755881, + "grad_norm": 2.822537899017334, + "learning_rate": 0.00015411520584002143, + "loss": 0.9514, + "step": 19510 + }, + { + "epoch": 1.3538963291929775, + "grad_norm": 1.9601463079452515, + "learning_rate": 0.00015411048092174866, + "loss": 0.8933, + "step": 19511 + }, + { + "epoch": 1.3539657206300744, + "grad_norm": 2.752291679382324, + "learning_rate": 0.00015410575583265542, + "loss": 1.3771, + "step": 19512 + }, + { + "epoch": 1.354035112067171, + "grad_norm": 1.7716008424758911, + "learning_rate": 0.00015410103057275656, + "loss": 0.6127, + "step": 19513 + }, + { + "epoch": 1.3541045035042676, + "grad_norm": 1.8975000381469727, + "learning_rate": 0.00015409630514206702, + "loss": 0.7653, + "step": 19514 + }, + { + "epoch": 1.3541738949413642, + "grad_norm": 1.7134732007980347, + "learning_rate": 0.0001540915795406017, + "loss": 0.5672, + "step": 19515 + }, + { + "epoch": 1.3542432863784608, + "grad_norm": 1.8993922472000122, + "learning_rate": 0.0001540868537683756, + "loss": 0.6218, + "step": 19516 + }, + { + "epoch": 1.3543126778155576, + "grad_norm": 1.667570948600769, + "learning_rate": 0.00015408212782540352, + "loss": 0.6836, + "step": 19517 + }, + { + "epoch": 1.3543820692526543, + "grad_norm": 1.7219444513320923, + "learning_rate": 0.0001540774017117005, + "loss": 0.696, + "step": 19518 + }, + { + "epoch": 1.3544514606897509, + "grad_norm": 1.708685278892517, + "learning_rate": 0.00015407267542728135, + "loss": 0.6012, + "step": 19519 + }, + { + "epoch": 1.3545208521268475, + "grad_norm": 1.920346736907959, + "learning_rate": 0.00015406794897216103, + "loss": 0.6914, + "step": 19520 + }, + { + "epoch": 1.354590243563944, + "grad_norm": 1.7986714839935303, + "learning_rate": 0.0001540632223463545, + "loss": 0.6744, + "step": 19521 + }, + { + "epoch": 1.354659635001041, + "grad_norm": 1.678605556488037, + "learning_rate": 0.00015405849554987663, + "loss": 0.5575, + "step": 19522 + }, + { + "epoch": 1.3547290264381375, + "grad_norm": 1.8066010475158691, + "learning_rate": 0.00015405376858274237, + "loss": 0.5052, + "step": 19523 + }, + { + "epoch": 1.3547984178752341, + "grad_norm": 1.9284861087799072, + "learning_rate": 0.00015404904144496661, + "loss": 0.5781, + "step": 19524 + }, + { + "epoch": 1.354867809312331, + "grad_norm": 2.2741434574127197, + "learning_rate": 0.00015404431413656433, + "loss": 0.7719, + "step": 19525 + }, + { + "epoch": 1.3549372007494276, + "grad_norm": 2.028329610824585, + "learning_rate": 0.0001540395866575504, + "loss": 0.7039, + "step": 19526 + }, + { + "epoch": 1.3550065921865242, + "grad_norm": 1.5910894870758057, + "learning_rate": 0.00015403485900793978, + "loss": 0.6196, + "step": 19527 + }, + { + "epoch": 1.3550759836236208, + "grad_norm": 1.8856406211853027, + "learning_rate": 0.00015403013118774737, + "loss": 0.8194, + "step": 19528 + }, + { + "epoch": 1.3551453750607174, + "grad_norm": 1.8764653205871582, + "learning_rate": 0.0001540254031969881, + "loss": 0.7243, + "step": 19529 + }, + { + "epoch": 1.3552147664978142, + "grad_norm": 1.5652559995651245, + "learning_rate": 0.0001540206750356769, + "loss": 0.5745, + "step": 19530 + }, + { + "epoch": 1.3552841579349109, + "grad_norm": 2.0925393104553223, + "learning_rate": 0.0001540159467038287, + "loss": 0.6218, + "step": 19531 + }, + { + "epoch": 1.3553535493720075, + "grad_norm": 1.5265953540802002, + "learning_rate": 0.00015401121820145843, + "loss": 0.5693, + "step": 19532 + }, + { + "epoch": 1.355422940809104, + "grad_norm": 1.785506010055542, + "learning_rate": 0.00015400648952858102, + "loss": 0.8066, + "step": 19533 + }, + { + "epoch": 1.3554923322462007, + "grad_norm": 1.6880018711090088, + "learning_rate": 0.00015400176068521137, + "loss": 0.6371, + "step": 19534 + }, + { + "epoch": 1.3555617236832975, + "grad_norm": 1.6169805526733398, + "learning_rate": 0.00015399703167136446, + "loss": 0.7312, + "step": 19535 + }, + { + "epoch": 1.3556311151203941, + "grad_norm": 2.2068259716033936, + "learning_rate": 0.00015399230248705514, + "loss": 0.7133, + "step": 19536 + }, + { + "epoch": 1.3557005065574907, + "grad_norm": 2.138939142227173, + "learning_rate": 0.0001539875731322984, + "loss": 0.959, + "step": 19537 + }, + { + "epoch": 1.3557698979945876, + "grad_norm": 1.6631356477737427, + "learning_rate": 0.00015398284360710913, + "loss": 0.6704, + "step": 19538 + }, + { + "epoch": 1.3558392894316842, + "grad_norm": 1.4422788619995117, + "learning_rate": 0.0001539781139115023, + "loss": 0.518, + "step": 19539 + }, + { + "epoch": 1.3559086808687808, + "grad_norm": 1.9516855478286743, + "learning_rate": 0.00015397338404549287, + "loss": 0.5808, + "step": 19540 + }, + { + "epoch": 1.3559780723058774, + "grad_norm": 1.9919945001602173, + "learning_rate": 0.0001539686540090957, + "loss": 0.5081, + "step": 19541 + }, + { + "epoch": 1.356047463742974, + "grad_norm": 1.6377816200256348, + "learning_rate": 0.00015396392380232575, + "loss": 0.5878, + "step": 19542 + }, + { + "epoch": 1.3561168551800709, + "grad_norm": 1.4871985912322998, + "learning_rate": 0.00015395919342519796, + "loss": 0.5976, + "step": 19543 + }, + { + "epoch": 1.3561862466171675, + "grad_norm": 1.5799816846847534, + "learning_rate": 0.00015395446287772723, + "loss": 0.8527, + "step": 19544 + }, + { + "epoch": 1.356255638054264, + "grad_norm": 1.846648097038269, + "learning_rate": 0.00015394973215992854, + "loss": 0.5174, + "step": 19545 + }, + { + "epoch": 1.3563250294913607, + "grad_norm": 1.4261584281921387, + "learning_rate": 0.0001539450012718168, + "loss": 0.5412, + "step": 19546 + }, + { + "epoch": 1.3563944209284573, + "grad_norm": 1.8294308185577393, + "learning_rate": 0.00015394027021340697, + "loss": 0.5977, + "step": 19547 + }, + { + "epoch": 1.3564638123655541, + "grad_norm": 1.8020957708358765, + "learning_rate": 0.00015393553898471397, + "loss": 0.5797, + "step": 19548 + }, + { + "epoch": 1.3565332038026507, + "grad_norm": 1.8035686016082764, + "learning_rate": 0.00015393080758575268, + "loss": 0.6022, + "step": 19549 + }, + { + "epoch": 1.3566025952397474, + "grad_norm": 1.3912540674209595, + "learning_rate": 0.00015392607601653812, + "loss": 0.4542, + "step": 19550 + }, + { + "epoch": 1.3566719866768442, + "grad_norm": 1.7344632148742676, + "learning_rate": 0.00015392134427708523, + "loss": 0.7718, + "step": 19551 + }, + { + "epoch": 1.3567413781139408, + "grad_norm": 1.830007791519165, + "learning_rate": 0.00015391661236740885, + "loss": 0.6748, + "step": 19552 + }, + { + "epoch": 1.3568107695510374, + "grad_norm": 1.9441800117492676, + "learning_rate": 0.00015391188028752404, + "loss": 0.4911, + "step": 19553 + }, + { + "epoch": 1.356880160988134, + "grad_norm": 1.7606239318847656, + "learning_rate": 0.0001539071480374456, + "loss": 0.5685, + "step": 19554 + }, + { + "epoch": 1.3569495524252306, + "grad_norm": 1.6504216194152832, + "learning_rate": 0.00015390241561718862, + "loss": 0.6197, + "step": 19555 + }, + { + "epoch": 1.3570189438623275, + "grad_norm": 1.696355938911438, + "learning_rate": 0.00015389768302676794, + "loss": 0.5626, + "step": 19556 + }, + { + "epoch": 1.357088335299424, + "grad_norm": 1.937628984451294, + "learning_rate": 0.00015389295026619853, + "loss": 0.6778, + "step": 19557 + }, + { + "epoch": 1.3571577267365207, + "grad_norm": 1.6037448644638062, + "learning_rate": 0.00015388821733549533, + "loss": 0.4895, + "step": 19558 + }, + { + "epoch": 1.3572271181736173, + "grad_norm": 1.6771165132522583, + "learning_rate": 0.0001538834842346733, + "loss": 0.5893, + "step": 19559 + }, + { + "epoch": 1.357296509610714, + "grad_norm": 1.8848819732666016, + "learning_rate": 0.0001538787509637473, + "loss": 0.5935, + "step": 19560 + }, + { + "epoch": 1.3573659010478107, + "grad_norm": 1.713982105255127, + "learning_rate": 0.00015387401752273237, + "loss": 0.5474, + "step": 19561 + }, + { + "epoch": 1.3574352924849074, + "grad_norm": 1.565346360206604, + "learning_rate": 0.00015386928391164344, + "loss": 0.6133, + "step": 19562 + }, + { + "epoch": 1.357504683922004, + "grad_norm": 1.0968296527862549, + "learning_rate": 0.00015386455013049541, + "loss": 0.4136, + "step": 19563 + }, + { + "epoch": 1.3575740753591008, + "grad_norm": 1.9434865713119507, + "learning_rate": 0.00015385981617930324, + "loss": 0.755, + "step": 19564 + }, + { + "epoch": 1.3576434667961974, + "grad_norm": 1.1929739713668823, + "learning_rate": 0.00015385508205808187, + "loss": 0.4865, + "step": 19565 + }, + { + "epoch": 1.357712858233294, + "grad_norm": 1.6088519096374512, + "learning_rate": 0.0001538503477668463, + "loss": 0.8322, + "step": 19566 + }, + { + "epoch": 1.3577822496703906, + "grad_norm": 1.6052894592285156, + "learning_rate": 0.00015384561330561138, + "loss": 0.5941, + "step": 19567 + }, + { + "epoch": 1.3578516411074872, + "grad_norm": 3.0749399662017822, + "learning_rate": 0.00015384087867439215, + "loss": 0.543, + "step": 19568 + }, + { + "epoch": 1.357921032544584, + "grad_norm": 4.015931606292725, + "learning_rate": 0.0001538361438732035, + "loss": 0.7738, + "step": 19569 + }, + { + "epoch": 1.3579904239816807, + "grad_norm": 1.8698534965515137, + "learning_rate": 0.00015383140890206035, + "loss": 0.778, + "step": 19570 + }, + { + "epoch": 1.3580598154187773, + "grad_norm": 1.4835054874420166, + "learning_rate": 0.00015382667376097773, + "loss": 0.5602, + "step": 19571 + }, + { + "epoch": 1.358129206855874, + "grad_norm": 1.3199493885040283, + "learning_rate": 0.00015382193844997051, + "loss": 0.4182, + "step": 19572 + }, + { + "epoch": 1.3581985982929705, + "grad_norm": 1.8585463762283325, + "learning_rate": 0.0001538172029690537, + "loss": 0.6061, + "step": 19573 + }, + { + "epoch": 1.3582679897300673, + "grad_norm": 1.9901660680770874, + "learning_rate": 0.00015381246731824223, + "loss": 0.81, + "step": 19574 + }, + { + "epoch": 1.358337381167164, + "grad_norm": 1.1372368335723877, + "learning_rate": 0.00015380773149755104, + "loss": 0.4613, + "step": 19575 + }, + { + "epoch": 1.3584067726042606, + "grad_norm": 1.3622266054153442, + "learning_rate": 0.00015380299550699506, + "loss": 0.3409, + "step": 19576 + }, + { + "epoch": 1.3584761640413574, + "grad_norm": 1.663332223892212, + "learning_rate": 0.00015379825934658928, + "loss": 0.6562, + "step": 19577 + }, + { + "epoch": 1.358545555478454, + "grad_norm": 1.6287556886672974, + "learning_rate": 0.00015379352301634865, + "loss": 0.5984, + "step": 19578 + }, + { + "epoch": 1.3586149469155506, + "grad_norm": 1.7969509363174438, + "learning_rate": 0.0001537887865162881, + "loss": 0.6087, + "step": 19579 + }, + { + "epoch": 1.3586843383526472, + "grad_norm": 2.5160250663757324, + "learning_rate": 0.0001537840498464226, + "loss": 0.9116, + "step": 19580 + }, + { + "epoch": 1.3587537297897438, + "grad_norm": 1.619519829750061, + "learning_rate": 0.0001537793130067671, + "loss": 0.6083, + "step": 19581 + }, + { + "epoch": 1.3588231212268407, + "grad_norm": 1.8065840005874634, + "learning_rate": 0.00015377457599733655, + "loss": 0.6956, + "step": 19582 + }, + { + "epoch": 1.3588925126639373, + "grad_norm": 2.2910947799682617, + "learning_rate": 0.00015376983881814585, + "loss": 0.5995, + "step": 19583 + }, + { + "epoch": 1.358961904101034, + "grad_norm": 1.8751256465911865, + "learning_rate": 0.00015376510146921008, + "loss": 0.5272, + "step": 19584 + }, + { + "epoch": 1.3590312955381305, + "grad_norm": 2.2024126052856445, + "learning_rate": 0.00015376036395054406, + "loss": 0.7733, + "step": 19585 + }, + { + "epoch": 1.3591006869752271, + "grad_norm": 1.5105233192443848, + "learning_rate": 0.00015375562626216284, + "loss": 0.4941, + "step": 19586 + }, + { + "epoch": 1.359170078412324, + "grad_norm": 1.4746917486190796, + "learning_rate": 0.00015375088840408134, + "loss": 0.462, + "step": 19587 + }, + { + "epoch": 1.3592394698494206, + "grad_norm": 1.7487775087356567, + "learning_rate": 0.00015374615037631455, + "loss": 0.6024, + "step": 19588 + }, + { + "epoch": 1.3593088612865172, + "grad_norm": 2.119109630584717, + "learning_rate": 0.00015374141217887736, + "loss": 0.5662, + "step": 19589 + }, + { + "epoch": 1.359378252723614, + "grad_norm": 1.5879052877426147, + "learning_rate": 0.0001537366738117848, + "loss": 0.7266, + "step": 19590 + }, + { + "epoch": 1.3594476441607106, + "grad_norm": 1.5708791017532349, + "learning_rate": 0.00015373193527505176, + "loss": 0.5542, + "step": 19591 + }, + { + "epoch": 1.3595170355978072, + "grad_norm": 2.019824743270874, + "learning_rate": 0.00015372719656869322, + "loss": 0.5793, + "step": 19592 + }, + { + "epoch": 1.3595864270349038, + "grad_norm": 1.7521402835845947, + "learning_rate": 0.00015372245769272423, + "loss": 0.6233, + "step": 19593 + }, + { + "epoch": 1.3596558184720005, + "grad_norm": 1.3294665813446045, + "learning_rate": 0.00015371771864715962, + "loss": 0.3848, + "step": 19594 + }, + { + "epoch": 1.3597252099090973, + "grad_norm": 2.103548049926758, + "learning_rate": 0.0001537129794320144, + "loss": 0.6548, + "step": 19595 + }, + { + "epoch": 1.359794601346194, + "grad_norm": 1.6701737642288208, + "learning_rate": 0.00015370824004730355, + "loss": 0.6106, + "step": 19596 + }, + { + "epoch": 1.3598639927832905, + "grad_norm": 1.3918849229812622, + "learning_rate": 0.00015370350049304202, + "loss": 0.6534, + "step": 19597 + }, + { + "epoch": 1.3599333842203873, + "grad_norm": 1.6358911991119385, + "learning_rate": 0.00015369876076924474, + "loss": 0.5586, + "step": 19598 + }, + { + "epoch": 1.3600027756574837, + "grad_norm": 1.7964959144592285, + "learning_rate": 0.00015369402087592673, + "loss": 0.8157, + "step": 19599 + }, + { + "epoch": 1.3600721670945806, + "grad_norm": 2.3590972423553467, + "learning_rate": 0.00015368928081310293, + "loss": 0.5386, + "step": 19600 + }, + { + "epoch": 1.3601415585316772, + "grad_norm": 1.9695827960968018, + "learning_rate": 0.0001536845405807883, + "loss": 0.8317, + "step": 19601 + }, + { + "epoch": 1.3602109499687738, + "grad_norm": 1.6997787952423096, + "learning_rate": 0.00015367980017899779, + "loss": 0.6325, + "step": 19602 + }, + { + "epoch": 1.3602803414058706, + "grad_norm": 1.7124215364456177, + "learning_rate": 0.0001536750596077464, + "loss": 0.6045, + "step": 19603 + }, + { + "epoch": 1.3603497328429672, + "grad_norm": 3.3015782833099365, + "learning_rate": 0.00015367031886704907, + "loss": 0.6853, + "step": 19604 + }, + { + "epoch": 1.3604191242800638, + "grad_norm": 2.2525839805603027, + "learning_rate": 0.00015366557795692075, + "loss": 0.7496, + "step": 19605 + }, + { + "epoch": 1.3604885157171605, + "grad_norm": 2.1702706813812256, + "learning_rate": 0.00015366083687737641, + "loss": 0.7325, + "step": 19606 + }, + { + "epoch": 1.360557907154257, + "grad_norm": 2.1155412197113037, + "learning_rate": 0.00015365609562843107, + "loss": 0.5475, + "step": 19607 + }, + { + "epoch": 1.360627298591354, + "grad_norm": 1.4959806203842163, + "learning_rate": 0.00015365135421009966, + "loss": 0.5438, + "step": 19608 + }, + { + "epoch": 1.3606966900284505, + "grad_norm": 1.2002543210983276, + "learning_rate": 0.00015364661262239713, + "loss": 0.6932, + "step": 19609 + }, + { + "epoch": 1.3607660814655471, + "grad_norm": 1.874493956565857, + "learning_rate": 0.0001536418708653385, + "loss": 0.8126, + "step": 19610 + }, + { + "epoch": 1.360835472902644, + "grad_norm": 1.970669150352478, + "learning_rate": 0.00015363712893893866, + "loss": 0.7383, + "step": 19611 + }, + { + "epoch": 1.3609048643397403, + "grad_norm": 1.9695565700531006, + "learning_rate": 0.00015363238684321266, + "loss": 0.629, + "step": 19612 + }, + { + "epoch": 1.3609742557768372, + "grad_norm": 2.2417099475860596, + "learning_rate": 0.00015362764457817543, + "loss": 0.8935, + "step": 19613 + }, + { + "epoch": 1.3610436472139338, + "grad_norm": 1.8474085330963135, + "learning_rate": 0.00015362290214384196, + "loss": 0.5544, + "step": 19614 + }, + { + "epoch": 1.3611130386510304, + "grad_norm": 2.158234119415283, + "learning_rate": 0.0001536181595402272, + "loss": 0.7889, + "step": 19615 + }, + { + "epoch": 1.3611824300881272, + "grad_norm": 1.6004054546356201, + "learning_rate": 0.00015361341676734612, + "loss": 0.6871, + "step": 19616 + }, + { + "epoch": 1.3612518215252238, + "grad_norm": 1.9720381498336792, + "learning_rate": 0.0001536086738252137, + "loss": 0.6914, + "step": 19617 + }, + { + "epoch": 1.3613212129623204, + "grad_norm": 1.8616251945495605, + "learning_rate": 0.00015360393071384496, + "loss": 0.4604, + "step": 19618 + }, + { + "epoch": 1.361390604399417, + "grad_norm": 1.3825658559799194, + "learning_rate": 0.0001535991874332548, + "loss": 0.4733, + "step": 19619 + }, + { + "epoch": 1.3614599958365137, + "grad_norm": 1.6582595109939575, + "learning_rate": 0.0001535944439834582, + "loss": 0.5429, + "step": 19620 + }, + { + "epoch": 1.3615293872736105, + "grad_norm": 1.371316909790039, + "learning_rate": 0.00015358970036447018, + "loss": 0.563, + "step": 19621 + }, + { + "epoch": 1.3615987787107071, + "grad_norm": 1.785049319267273, + "learning_rate": 0.00015358495657630573, + "loss": 0.572, + "step": 19622 + }, + { + "epoch": 1.3616681701478037, + "grad_norm": 2.211195707321167, + "learning_rate": 0.00015358021261897972, + "loss": 0.5827, + "step": 19623 + }, + { + "epoch": 1.3617375615849006, + "grad_norm": 1.6678375005722046, + "learning_rate": 0.00015357546849250725, + "loss": 0.6232, + "step": 19624 + }, + { + "epoch": 1.3618069530219972, + "grad_norm": 1.705150842666626, + "learning_rate": 0.00015357072419690325, + "loss": 0.5891, + "step": 19625 + }, + { + "epoch": 1.3618763444590938, + "grad_norm": 1.6505680084228516, + "learning_rate": 0.00015356597973218263, + "loss": 0.5678, + "step": 19626 + }, + { + "epoch": 1.3619457358961904, + "grad_norm": 1.5557518005371094, + "learning_rate": 0.00015356123509836043, + "loss": 0.6529, + "step": 19627 + }, + { + "epoch": 1.362015127333287, + "grad_norm": 2.1064488887786865, + "learning_rate": 0.00015355649029545168, + "loss": 0.6441, + "step": 19628 + }, + { + "epoch": 1.3620845187703838, + "grad_norm": 1.9494692087173462, + "learning_rate": 0.00015355174532347127, + "loss": 0.6657, + "step": 19629 + }, + { + "epoch": 1.3621539102074804, + "grad_norm": 1.8019448518753052, + "learning_rate": 0.00015354700018243423, + "loss": 0.6695, + "step": 19630 + }, + { + "epoch": 1.362223301644577, + "grad_norm": 2.8001012802124023, + "learning_rate": 0.0001535422548723555, + "loss": 0.9087, + "step": 19631 + }, + { + "epoch": 1.3622926930816737, + "grad_norm": 1.7477506399154663, + "learning_rate": 0.0001535375093932501, + "loss": 0.6327, + "step": 19632 + }, + { + "epoch": 1.3623620845187703, + "grad_norm": 1.9378368854522705, + "learning_rate": 0.00015353276374513298, + "loss": 0.8387, + "step": 19633 + }, + { + "epoch": 1.362431475955867, + "grad_norm": 1.8833738565444946, + "learning_rate": 0.00015352801792801915, + "loss": 0.6928, + "step": 19634 + }, + { + "epoch": 1.3625008673929637, + "grad_norm": 1.499712347984314, + "learning_rate": 0.0001535232719419236, + "loss": 0.622, + "step": 19635 + }, + { + "epoch": 1.3625702588300603, + "grad_norm": 1.473600149154663, + "learning_rate": 0.00015351852578686126, + "loss": 0.4446, + "step": 19636 + }, + { + "epoch": 1.3626396502671572, + "grad_norm": 2.5717177391052246, + "learning_rate": 0.00015351377946284714, + "loss": 0.7614, + "step": 19637 + }, + { + "epoch": 1.3627090417042538, + "grad_norm": 1.577411413192749, + "learning_rate": 0.00015350903296989623, + "loss": 0.4889, + "step": 19638 + }, + { + "epoch": 1.3627784331413504, + "grad_norm": 2.01663875579834, + "learning_rate": 0.0001535042863080235, + "loss": 0.5898, + "step": 19639 + }, + { + "epoch": 1.362847824578447, + "grad_norm": 2.630337715148926, + "learning_rate": 0.000153499539477244, + "loss": 0.9389, + "step": 19640 + }, + { + "epoch": 1.3629172160155436, + "grad_norm": 1.653836965560913, + "learning_rate": 0.00015349479247757258, + "loss": 0.622, + "step": 19641 + }, + { + "epoch": 1.3629866074526404, + "grad_norm": 1.5178632736206055, + "learning_rate": 0.00015349004530902438, + "loss": 0.7712, + "step": 19642 + }, + { + "epoch": 1.363055998889737, + "grad_norm": 1.431081771850586, + "learning_rate": 0.00015348529797161429, + "loss": 0.5576, + "step": 19643 + }, + { + "epoch": 1.3631253903268337, + "grad_norm": 2.6081490516662598, + "learning_rate": 0.00015348055046535727, + "loss": 0.7595, + "step": 19644 + }, + { + "epoch": 1.3631947817639303, + "grad_norm": 2.2887959480285645, + "learning_rate": 0.0001534758027902684, + "loss": 0.853, + "step": 19645 + }, + { + "epoch": 1.3632641732010269, + "grad_norm": 1.5659117698669434, + "learning_rate": 0.00015347105494636264, + "loss": 0.5138, + "step": 19646 + }, + { + "epoch": 1.3633335646381237, + "grad_norm": 1.333042860031128, + "learning_rate": 0.00015346630693365495, + "loss": 0.5379, + "step": 19647 + }, + { + "epoch": 1.3634029560752203, + "grad_norm": 1.9575282335281372, + "learning_rate": 0.00015346155875216032, + "loss": 0.6513, + "step": 19648 + }, + { + "epoch": 1.363472347512317, + "grad_norm": 1.8544645309448242, + "learning_rate": 0.00015345681040189376, + "loss": 0.7876, + "step": 19649 + }, + { + "epoch": 1.3635417389494138, + "grad_norm": 1.7718031406402588, + "learning_rate": 0.00015345206188287027, + "loss": 0.4388, + "step": 19650 + }, + { + "epoch": 1.3636111303865104, + "grad_norm": 2.20448637008667, + "learning_rate": 0.0001534473131951048, + "loss": 0.6168, + "step": 19651 + }, + { + "epoch": 1.363680521823607, + "grad_norm": 2.123370885848999, + "learning_rate": 0.00015344256433861235, + "loss": 0.8889, + "step": 19652 + }, + { + "epoch": 1.3637499132607036, + "grad_norm": 1.913679838180542, + "learning_rate": 0.00015343781531340796, + "loss": 0.6605, + "step": 19653 + }, + { + "epoch": 1.3638193046978002, + "grad_norm": 1.9610966444015503, + "learning_rate": 0.00015343306611950656, + "loss": 0.6447, + "step": 19654 + }, + { + "epoch": 1.363888696134897, + "grad_norm": 2.606961727142334, + "learning_rate": 0.0001534283167569232, + "loss": 0.6006, + "step": 19655 + }, + { + "epoch": 1.3639580875719937, + "grad_norm": 1.6347885131835938, + "learning_rate": 0.0001534235672256728, + "loss": 0.5597, + "step": 19656 + }, + { + "epoch": 1.3640274790090903, + "grad_norm": 1.5554907321929932, + "learning_rate": 0.00015341881752577044, + "loss": 0.645, + "step": 19657 + }, + { + "epoch": 1.3640968704461869, + "grad_norm": 2.7543294429779053, + "learning_rate": 0.00015341406765723105, + "loss": 0.5677, + "step": 19658 + }, + { + "epoch": 1.3641662618832835, + "grad_norm": 1.6518033742904663, + "learning_rate": 0.00015340931762006962, + "loss": 0.6355, + "step": 19659 + }, + { + "epoch": 1.3642356533203803, + "grad_norm": 2.512925386428833, + "learning_rate": 0.0001534045674143012, + "loss": 0.6682, + "step": 19660 + }, + { + "epoch": 1.364305044757477, + "grad_norm": 1.7485716342926025, + "learning_rate": 0.00015339981703994075, + "loss": 0.746, + "step": 19661 + }, + { + "epoch": 1.3643744361945735, + "grad_norm": 1.9173052310943604, + "learning_rate": 0.0001533950664970033, + "loss": 0.6937, + "step": 19662 + }, + { + "epoch": 1.3644438276316704, + "grad_norm": 2.3577303886413574, + "learning_rate": 0.0001533903157855038, + "loss": 0.7989, + "step": 19663 + }, + { + "epoch": 1.364513219068767, + "grad_norm": 1.7350393533706665, + "learning_rate": 0.00015338556490545723, + "loss": 0.7869, + "step": 19664 + }, + { + "epoch": 1.3645826105058636, + "grad_norm": 2.481004238128662, + "learning_rate": 0.00015338081385687867, + "loss": 0.7029, + "step": 19665 + }, + { + "epoch": 1.3646520019429602, + "grad_norm": 1.997665524482727, + "learning_rate": 0.00015337606263978305, + "loss": 0.8515, + "step": 19666 + }, + { + "epoch": 1.3647213933800568, + "grad_norm": 1.9242427349090576, + "learning_rate": 0.00015337131125418542, + "loss": 0.8684, + "step": 19667 + }, + { + "epoch": 1.3647907848171537, + "grad_norm": 1.9580167531967163, + "learning_rate": 0.00015336655970010074, + "loss": 0.5599, + "step": 19668 + }, + { + "epoch": 1.3648601762542503, + "grad_norm": 1.8187143802642822, + "learning_rate": 0.00015336180797754403, + "loss": 0.6729, + "step": 19669 + }, + { + "epoch": 1.3649295676913469, + "grad_norm": 1.960936427116394, + "learning_rate": 0.0001533570560865303, + "loss": 0.6655, + "step": 19670 + }, + { + "epoch": 1.3649989591284435, + "grad_norm": 1.7264882326126099, + "learning_rate": 0.00015335230402707448, + "loss": 0.7052, + "step": 19671 + }, + { + "epoch": 1.36506835056554, + "grad_norm": 1.6130033731460571, + "learning_rate": 0.00015334755179919165, + "loss": 0.599, + "step": 19672 + }, + { + "epoch": 1.365137742002637, + "grad_norm": 2.2026617527008057, + "learning_rate": 0.0001533427994028968, + "loss": 0.865, + "step": 19673 + }, + { + "epoch": 1.3652071334397335, + "grad_norm": 2.031435251235962, + "learning_rate": 0.0001533380468382049, + "loss": 0.5476, + "step": 19674 + }, + { + "epoch": 1.3652765248768302, + "grad_norm": 2.4418327808380127, + "learning_rate": 0.000153333294105131, + "loss": 0.6922, + "step": 19675 + }, + { + "epoch": 1.365345916313927, + "grad_norm": 1.8658344745635986, + "learning_rate": 0.0001533285412036901, + "loss": 0.6015, + "step": 19676 + }, + { + "epoch": 1.3654153077510236, + "grad_norm": 2.245553970336914, + "learning_rate": 0.00015332378813389714, + "loss": 0.6531, + "step": 19677 + }, + { + "epoch": 1.3654846991881202, + "grad_norm": 2.051929473876953, + "learning_rate": 0.00015331903489576714, + "loss": 0.6397, + "step": 19678 + }, + { + "epoch": 1.3655540906252168, + "grad_norm": 1.9225407838821411, + "learning_rate": 0.0001533142814893152, + "loss": 0.4917, + "step": 19679 + }, + { + "epoch": 1.3656234820623134, + "grad_norm": 2.794374704360962, + "learning_rate": 0.00015330952791455625, + "loss": 0.7867, + "step": 19680 + }, + { + "epoch": 1.3656928734994103, + "grad_norm": 1.6209956407546997, + "learning_rate": 0.00015330477417150527, + "loss": 0.5903, + "step": 19681 + }, + { + "epoch": 1.3657622649365069, + "grad_norm": 1.7590430974960327, + "learning_rate": 0.0001533000202601773, + "loss": 0.5631, + "step": 19682 + }, + { + "epoch": 1.3658316563736035, + "grad_norm": 2.091355562210083, + "learning_rate": 0.00015329526618058735, + "loss": 0.5683, + "step": 19683 + }, + { + "epoch": 1.3659010478107, + "grad_norm": 1.8777085542678833, + "learning_rate": 0.00015329051193275047, + "loss": 0.6325, + "step": 19684 + }, + { + "epoch": 1.3659704392477967, + "grad_norm": 2.586135149002075, + "learning_rate": 0.0001532857575166816, + "loss": 0.5882, + "step": 19685 + }, + { + "epoch": 1.3660398306848935, + "grad_norm": 1.7449915409088135, + "learning_rate": 0.00015328100293239577, + "loss": 0.6139, + "step": 19686 + }, + { + "epoch": 1.3661092221219902, + "grad_norm": 3.3519160747528076, + "learning_rate": 0.000153276248179908, + "loss": 0.658, + "step": 19687 + }, + { + "epoch": 1.3661786135590868, + "grad_norm": 2.226396083831787, + "learning_rate": 0.0001532714932592333, + "loss": 0.6224, + "step": 19688 + }, + { + "epoch": 1.3662480049961836, + "grad_norm": 2.23101806640625, + "learning_rate": 0.00015326673817038668, + "loss": 0.5658, + "step": 19689 + }, + { + "epoch": 1.3663173964332802, + "grad_norm": 1.4938874244689941, + "learning_rate": 0.00015326198291338317, + "loss": 0.5609, + "step": 19690 + }, + { + "epoch": 1.3663867878703768, + "grad_norm": 2.5168018341064453, + "learning_rate": 0.00015325722748823773, + "loss": 0.7274, + "step": 19691 + }, + { + "epoch": 1.3664561793074734, + "grad_norm": 1.5523078441619873, + "learning_rate": 0.00015325247189496542, + "loss": 0.7453, + "step": 19692 + }, + { + "epoch": 1.36652557074457, + "grad_norm": 1.611393690109253, + "learning_rate": 0.0001532477161335812, + "loss": 0.6792, + "step": 19693 + }, + { + "epoch": 1.3665949621816669, + "grad_norm": 1.8334835767745972, + "learning_rate": 0.00015324296020410017, + "loss": 0.7041, + "step": 19694 + }, + { + "epoch": 1.3666643536187635, + "grad_norm": 2.3821535110473633, + "learning_rate": 0.0001532382041065372, + "loss": 0.6858, + "step": 19695 + }, + { + "epoch": 1.36673374505586, + "grad_norm": 2.23612380027771, + "learning_rate": 0.00015323344784090747, + "loss": 0.6968, + "step": 19696 + }, + { + "epoch": 1.3668031364929567, + "grad_norm": 1.507827639579773, + "learning_rate": 0.00015322869140722592, + "loss": 0.5511, + "step": 19697 + }, + { + "epoch": 1.3668725279300533, + "grad_norm": 1.7595926523208618, + "learning_rate": 0.00015322393480550757, + "loss": 0.5696, + "step": 19698 + }, + { + "epoch": 1.3669419193671501, + "grad_norm": 3.15875244140625, + "learning_rate": 0.0001532191780357674, + "loss": 0.8291, + "step": 19699 + }, + { + "epoch": 1.3670113108042468, + "grad_norm": 1.4692022800445557, + "learning_rate": 0.00015321442109802045, + "loss": 0.613, + "step": 19700 + }, + { + "epoch": 1.3670807022413434, + "grad_norm": 2.3282783031463623, + "learning_rate": 0.0001532096639922818, + "loss": 0.651, + "step": 19701 + }, + { + "epoch": 1.3671500936784402, + "grad_norm": 2.074068069458008, + "learning_rate": 0.0001532049067185664, + "loss": 0.5476, + "step": 19702 + }, + { + "epoch": 1.3672194851155368, + "grad_norm": 1.6374390125274658, + "learning_rate": 0.00015320014927688925, + "loss": 0.6476, + "step": 19703 + }, + { + "epoch": 1.3672888765526334, + "grad_norm": 2.765270948410034, + "learning_rate": 0.00015319539166726542, + "loss": 0.5671, + "step": 19704 + }, + { + "epoch": 1.36735826798973, + "grad_norm": 2.9122121334075928, + "learning_rate": 0.00015319063388970992, + "loss": 0.8059, + "step": 19705 + }, + { + "epoch": 1.3674276594268266, + "grad_norm": 1.513602375984192, + "learning_rate": 0.00015318587594423773, + "loss": 0.7351, + "step": 19706 + }, + { + "epoch": 1.3674970508639235, + "grad_norm": 2.6931235790252686, + "learning_rate": 0.00015318111783086392, + "loss": 0.8272, + "step": 19707 + }, + { + "epoch": 1.36756644230102, + "grad_norm": 1.4507083892822266, + "learning_rate": 0.0001531763595496035, + "loss": 0.6566, + "step": 19708 + }, + { + "epoch": 1.3676358337381167, + "grad_norm": 1.6895898580551147, + "learning_rate": 0.0001531716011004715, + "loss": 0.5735, + "step": 19709 + }, + { + "epoch": 1.3677052251752133, + "grad_norm": 1.790918231010437, + "learning_rate": 0.0001531668424834829, + "loss": 0.5933, + "step": 19710 + }, + { + "epoch": 1.36777461661231, + "grad_norm": 1.8201947212219238, + "learning_rate": 0.00015316208369865275, + "loss": 0.7583, + "step": 19711 + }, + { + "epoch": 1.3678440080494068, + "grad_norm": 2.7246975898742676, + "learning_rate": 0.00015315732474599606, + "loss": 0.7569, + "step": 19712 + }, + { + "epoch": 1.3679133994865034, + "grad_norm": 2.302372694015503, + "learning_rate": 0.00015315256562552787, + "loss": 0.9105, + "step": 19713 + }, + { + "epoch": 1.3679827909236, + "grad_norm": 2.2958428859710693, + "learning_rate": 0.0001531478063372632, + "loss": 0.8609, + "step": 19714 + }, + { + "epoch": 1.3680521823606968, + "grad_norm": 1.9485857486724854, + "learning_rate": 0.00015314304688121707, + "loss": 0.714, + "step": 19715 + }, + { + "epoch": 1.3681215737977934, + "grad_norm": 1.629878282546997, + "learning_rate": 0.00015313828725740448, + "loss": 0.5977, + "step": 19716 + }, + { + "epoch": 1.36819096523489, + "grad_norm": 3.336463689804077, + "learning_rate": 0.00015313352746584052, + "loss": 0.9545, + "step": 19717 + }, + { + "epoch": 1.3682603566719866, + "grad_norm": 2.203185558319092, + "learning_rate": 0.00015312876750654016, + "loss": 0.7472, + "step": 19718 + }, + { + "epoch": 1.3683297481090833, + "grad_norm": 3.026435136795044, + "learning_rate": 0.00015312400737951847, + "loss": 0.6693, + "step": 19719 + }, + { + "epoch": 1.36839913954618, + "grad_norm": 1.5156060457229614, + "learning_rate": 0.0001531192470847904, + "loss": 0.5415, + "step": 19720 + }, + { + "epoch": 1.3684685309832767, + "grad_norm": 2.0555999279022217, + "learning_rate": 0.0001531144866223711, + "loss": 0.691, + "step": 19721 + }, + { + "epoch": 1.3685379224203733, + "grad_norm": 1.7558271884918213, + "learning_rate": 0.00015310972599227548, + "loss": 0.7467, + "step": 19722 + }, + { + "epoch": 1.36860731385747, + "grad_norm": 2.635512351989746, + "learning_rate": 0.00015310496519451862, + "loss": 0.6935, + "step": 19723 + }, + { + "epoch": 1.3686767052945665, + "grad_norm": 1.7452688217163086, + "learning_rate": 0.00015310020422911556, + "loss": 0.5818, + "step": 19724 + }, + { + "epoch": 1.3687460967316634, + "grad_norm": 2.037384033203125, + "learning_rate": 0.00015309544309608132, + "loss": 0.6072, + "step": 19725 + }, + { + "epoch": 1.36881548816876, + "grad_norm": 2.3712551593780518, + "learning_rate": 0.00015309068179543092, + "loss": 0.6709, + "step": 19726 + }, + { + "epoch": 1.3688848796058566, + "grad_norm": 2.168133497238159, + "learning_rate": 0.0001530859203271794, + "loss": 1.0817, + "step": 19727 + }, + { + "epoch": 1.3689542710429534, + "grad_norm": 1.7331734895706177, + "learning_rate": 0.00015308115869134176, + "loss": 0.6754, + "step": 19728 + }, + { + "epoch": 1.36902366248005, + "grad_norm": 1.6439088582992554, + "learning_rate": 0.0001530763968879331, + "loss": 0.5313, + "step": 19729 + }, + { + "epoch": 1.3690930539171466, + "grad_norm": 1.4557065963745117, + "learning_rate": 0.00015307163491696838, + "loss": 0.685, + "step": 19730 + }, + { + "epoch": 1.3691624453542433, + "grad_norm": 1.684359073638916, + "learning_rate": 0.0001530668727784627, + "loss": 0.9146, + "step": 19731 + }, + { + "epoch": 1.3692318367913399, + "grad_norm": 1.906079649925232, + "learning_rate": 0.00015306211047243104, + "loss": 0.4796, + "step": 19732 + }, + { + "epoch": 1.3693012282284367, + "grad_norm": 1.8089115619659424, + "learning_rate": 0.00015305734799888845, + "loss": 0.5582, + "step": 19733 + }, + { + "epoch": 1.3693706196655333, + "grad_norm": 1.8978378772735596, + "learning_rate": 0.00015305258535784996, + "loss": 0.7507, + "step": 19734 + }, + { + "epoch": 1.36944001110263, + "grad_norm": 2.0048158168792725, + "learning_rate": 0.00015304782254933064, + "loss": 0.7686, + "step": 19735 + }, + { + "epoch": 1.3695094025397265, + "grad_norm": 2.017665386199951, + "learning_rate": 0.00015304305957334548, + "loss": 0.6188, + "step": 19736 + }, + { + "epoch": 1.3695787939768231, + "grad_norm": 2.153029441833496, + "learning_rate": 0.00015303829642990954, + "loss": 0.9181, + "step": 19737 + }, + { + "epoch": 1.36964818541392, + "grad_norm": 2.2013368606567383, + "learning_rate": 0.00015303353311903783, + "loss": 0.8584, + "step": 19738 + }, + { + "epoch": 1.3697175768510166, + "grad_norm": 1.252684473991394, + "learning_rate": 0.00015302876964074546, + "loss": 0.5536, + "step": 19739 + }, + { + "epoch": 1.3697869682881132, + "grad_norm": 1.9808895587921143, + "learning_rate": 0.00015302400599504737, + "loss": 0.4681, + "step": 19740 + }, + { + "epoch": 1.36985635972521, + "grad_norm": 2.106135845184326, + "learning_rate": 0.00015301924218195866, + "loss": 0.6245, + "step": 19741 + }, + { + "epoch": 1.3699257511623066, + "grad_norm": 1.6401435136795044, + "learning_rate": 0.00015301447820149436, + "loss": 0.6091, + "step": 19742 + }, + { + "epoch": 1.3699951425994032, + "grad_norm": 1.7980217933654785, + "learning_rate": 0.0001530097140536695, + "loss": 0.7249, + "step": 19743 + }, + { + "epoch": 1.3700645340364999, + "grad_norm": 4.378880023956299, + "learning_rate": 0.00015300494973849915, + "loss": 0.927, + "step": 19744 + }, + { + "epoch": 1.3701339254735965, + "grad_norm": 2.5044825077056885, + "learning_rate": 0.0001530001852559983, + "loss": 0.7407, + "step": 19745 + }, + { + "epoch": 1.3702033169106933, + "grad_norm": 2.2658438682556152, + "learning_rate": 0.00015299542060618203, + "loss": 0.8382, + "step": 19746 + }, + { + "epoch": 1.37027270834779, + "grad_norm": 2.5342423915863037, + "learning_rate": 0.00015299065578906534, + "loss": 1.0151, + "step": 19747 + }, + { + "epoch": 1.3703420997848865, + "grad_norm": 2.2035927772521973, + "learning_rate": 0.00015298589080466328, + "loss": 0.6761, + "step": 19748 + }, + { + "epoch": 1.3704114912219831, + "grad_norm": 1.5153473615646362, + "learning_rate": 0.00015298112565299095, + "loss": 0.5539, + "step": 19749 + }, + { + "epoch": 1.3704808826590797, + "grad_norm": 1.8138235807418823, + "learning_rate": 0.00015297636033406332, + "loss": 0.5661, + "step": 19750 + }, + { + "epoch": 1.3705502740961766, + "grad_norm": 1.9773972034454346, + "learning_rate": 0.0001529715948478955, + "loss": 0.7835, + "step": 19751 + }, + { + "epoch": 1.3706196655332732, + "grad_norm": 1.9595811367034912, + "learning_rate": 0.00015296682919450246, + "loss": 0.6524, + "step": 19752 + }, + { + "epoch": 1.3706890569703698, + "grad_norm": 1.7624177932739258, + "learning_rate": 0.00015296206337389934, + "loss": 0.7177, + "step": 19753 + }, + { + "epoch": 1.3707584484074666, + "grad_norm": 2.681218385696411, + "learning_rate": 0.0001529572973861011, + "loss": 0.7762, + "step": 19754 + }, + { + "epoch": 1.3708278398445632, + "grad_norm": 1.8381056785583496, + "learning_rate": 0.00015295253123112282, + "loss": 0.9684, + "step": 19755 + }, + { + "epoch": 1.3708972312816599, + "grad_norm": 2.100341558456421, + "learning_rate": 0.00015294776490897954, + "loss": 0.9769, + "step": 19756 + }, + { + "epoch": 1.3709666227187565, + "grad_norm": 1.6368085145950317, + "learning_rate": 0.0001529429984196863, + "loss": 0.7163, + "step": 19757 + }, + { + "epoch": 1.371036014155853, + "grad_norm": 1.8410530090332031, + "learning_rate": 0.00015293823176325817, + "loss": 0.9014, + "step": 19758 + }, + { + "epoch": 1.37110540559295, + "grad_norm": 1.8627318143844604, + "learning_rate": 0.00015293346493971014, + "loss": 0.9198, + "step": 19759 + }, + { + "epoch": 1.3711747970300465, + "grad_norm": 1.7466727495193481, + "learning_rate": 0.00015292869794905736, + "loss": 0.6373, + "step": 19760 + }, + { + "epoch": 1.3712441884671431, + "grad_norm": 1.7991949319839478, + "learning_rate": 0.0001529239307913148, + "loss": 0.5341, + "step": 19761 + }, + { + "epoch": 1.3713135799042397, + "grad_norm": 2.237780809402466, + "learning_rate": 0.00015291916346649759, + "loss": 0.4237, + "step": 19762 + }, + { + "epoch": 1.3713829713413364, + "grad_norm": 1.7362743616104126, + "learning_rate": 0.00015291439597462064, + "loss": 0.6656, + "step": 19763 + }, + { + "epoch": 1.3714523627784332, + "grad_norm": 1.4097148180007935, + "learning_rate": 0.0001529096283156991, + "loss": 0.5175, + "step": 19764 + }, + { + "epoch": 1.3715217542155298, + "grad_norm": 2.368133544921875, + "learning_rate": 0.00015290486048974802, + "loss": 0.789, + "step": 19765 + }, + { + "epoch": 1.3715911456526264, + "grad_norm": 1.670597791671753, + "learning_rate": 0.0001529000924967824, + "loss": 0.6658, + "step": 19766 + }, + { + "epoch": 1.3716605370897232, + "grad_norm": 1.4750921726226807, + "learning_rate": 0.00015289532433681736, + "loss": 0.5944, + "step": 19767 + }, + { + "epoch": 1.3717299285268199, + "grad_norm": 1.6333539485931396, + "learning_rate": 0.0001528905560098679, + "loss": 0.6293, + "step": 19768 + }, + { + "epoch": 1.3717993199639165, + "grad_norm": 1.6810107231140137, + "learning_rate": 0.0001528857875159491, + "loss": 0.7283, + "step": 19769 + }, + { + "epoch": 1.371868711401013, + "grad_norm": 1.3322275876998901, + "learning_rate": 0.000152881018855076, + "loss": 0.5173, + "step": 19770 + }, + { + "epoch": 1.3719381028381097, + "grad_norm": 1.8728199005126953, + "learning_rate": 0.00015287625002726365, + "loss": 0.7507, + "step": 19771 + }, + { + "epoch": 1.3720074942752065, + "grad_norm": 1.4928568601608276, + "learning_rate": 0.00015287148103252713, + "loss": 0.6235, + "step": 19772 + }, + { + "epoch": 1.3720768857123031, + "grad_norm": 1.8000274896621704, + "learning_rate": 0.00015286671187088148, + "loss": 0.7522, + "step": 19773 + }, + { + "epoch": 1.3721462771493997, + "grad_norm": 2.1507675647735596, + "learning_rate": 0.00015286194254234174, + "loss": 0.9006, + "step": 19774 + }, + { + "epoch": 1.3722156685864964, + "grad_norm": 2.0665276050567627, + "learning_rate": 0.00015285717304692299, + "loss": 0.725, + "step": 19775 + }, + { + "epoch": 1.372285060023593, + "grad_norm": 1.448595643043518, + "learning_rate": 0.00015285240338464026, + "loss": 0.5465, + "step": 19776 + }, + { + "epoch": 1.3723544514606898, + "grad_norm": 1.9165394306182861, + "learning_rate": 0.00015284763355550866, + "loss": 0.6213, + "step": 19777 + }, + { + "epoch": 1.3724238428977864, + "grad_norm": 1.6921032667160034, + "learning_rate": 0.0001528428635595432, + "loss": 0.8097, + "step": 19778 + }, + { + "epoch": 1.372493234334883, + "grad_norm": 2.672849416732788, + "learning_rate": 0.00015283809339675894, + "loss": 0.9193, + "step": 19779 + }, + { + "epoch": 1.3725626257719798, + "grad_norm": 1.7292941808700562, + "learning_rate": 0.00015283332306717095, + "loss": 0.7115, + "step": 19780 + }, + { + "epoch": 1.3726320172090765, + "grad_norm": 1.7513716220855713, + "learning_rate": 0.0001528285525707943, + "loss": 0.9211, + "step": 19781 + }, + { + "epoch": 1.372701408646173, + "grad_norm": 2.096886157989502, + "learning_rate": 0.00015282378190764403, + "loss": 0.7621, + "step": 19782 + }, + { + "epoch": 1.3727708000832697, + "grad_norm": 1.798027753829956, + "learning_rate": 0.0001528190110777352, + "loss": 0.7819, + "step": 19783 + }, + { + "epoch": 1.3728401915203663, + "grad_norm": 1.390677571296692, + "learning_rate": 0.0001528142400810829, + "loss": 0.5944, + "step": 19784 + }, + { + "epoch": 1.3729095829574631, + "grad_norm": 2.5819408893585205, + "learning_rate": 0.00015280946891770215, + "loss": 0.7424, + "step": 19785 + }, + { + "epoch": 1.3729789743945597, + "grad_norm": 2.073347330093384, + "learning_rate": 0.00015280469758760805, + "loss": 0.8701, + "step": 19786 + }, + { + "epoch": 1.3730483658316563, + "grad_norm": 1.7172977924346924, + "learning_rate": 0.00015279992609081565, + "loss": 0.7048, + "step": 19787 + }, + { + "epoch": 1.373117757268753, + "grad_norm": 2.051457405090332, + "learning_rate": 0.00015279515442734, + "loss": 0.7395, + "step": 19788 + }, + { + "epoch": 1.3731871487058496, + "grad_norm": 2.181319236755371, + "learning_rate": 0.00015279038259719615, + "loss": 0.8185, + "step": 19789 + }, + { + "epoch": 1.3732565401429464, + "grad_norm": 1.926101565361023, + "learning_rate": 0.00015278561060039922, + "loss": 0.7581, + "step": 19790 + }, + { + "epoch": 1.373325931580043, + "grad_norm": 1.3749785423278809, + "learning_rate": 0.00015278083843696424, + "loss": 0.5421, + "step": 19791 + }, + { + "epoch": 1.3733953230171396, + "grad_norm": 1.8132034540176392, + "learning_rate": 0.00015277606610690625, + "loss": 0.4538, + "step": 19792 + }, + { + "epoch": 1.3734647144542365, + "grad_norm": 1.3004674911499023, + "learning_rate": 0.00015277129361024037, + "loss": 0.4465, + "step": 19793 + }, + { + "epoch": 1.373534105891333, + "grad_norm": 1.9587957859039307, + "learning_rate": 0.00015276652094698163, + "loss": 0.8639, + "step": 19794 + }, + { + "epoch": 1.3736034973284297, + "grad_norm": 2.846195697784424, + "learning_rate": 0.0001527617481171451, + "loss": 0.694, + "step": 19795 + }, + { + "epoch": 1.3736728887655263, + "grad_norm": 1.73430335521698, + "learning_rate": 0.00015275697512074585, + "loss": 0.6686, + "step": 19796 + }, + { + "epoch": 1.373742280202623, + "grad_norm": 2.3736732006073, + "learning_rate": 0.00015275220195779899, + "loss": 0.6115, + "step": 19797 + }, + { + "epoch": 1.3738116716397197, + "grad_norm": 1.4939829111099243, + "learning_rate": 0.00015274742862831953, + "loss": 0.6846, + "step": 19798 + }, + { + "epoch": 1.3738810630768163, + "grad_norm": 2.053577184677124, + "learning_rate": 0.0001527426551323225, + "loss": 0.7346, + "step": 19799 + }, + { + "epoch": 1.373950454513913, + "grad_norm": 2.3320043087005615, + "learning_rate": 0.0001527378814698231, + "loss": 0.7958, + "step": 19800 + }, + { + "epoch": 1.3740198459510096, + "grad_norm": 2.0683207511901855, + "learning_rate": 0.0001527331076408363, + "loss": 0.5949, + "step": 19801 + }, + { + "epoch": 1.3740892373881062, + "grad_norm": 1.4353852272033691, + "learning_rate": 0.00015272833364537717, + "loss": 0.4118, + "step": 19802 + }, + { + "epoch": 1.374158628825203, + "grad_norm": 1.3606374263763428, + "learning_rate": 0.00015272355948346084, + "loss": 0.5624, + "step": 19803 + }, + { + "epoch": 1.3742280202622996, + "grad_norm": 1.8204160928726196, + "learning_rate": 0.00015271878515510235, + "loss": 0.7707, + "step": 19804 + }, + { + "epoch": 1.3742974116993962, + "grad_norm": 1.9071663618087769, + "learning_rate": 0.00015271401066031676, + "loss": 0.7409, + "step": 19805 + }, + { + "epoch": 1.374366803136493, + "grad_norm": 1.9409641027450562, + "learning_rate": 0.00015270923599911918, + "loss": 0.534, + "step": 19806 + }, + { + "epoch": 1.3744361945735897, + "grad_norm": 1.543241262435913, + "learning_rate": 0.00015270446117152461, + "loss": 0.5451, + "step": 19807 + }, + { + "epoch": 1.3745055860106863, + "grad_norm": 1.6462024450302124, + "learning_rate": 0.00015269968617754822, + "loss": 0.6481, + "step": 19808 + }, + { + "epoch": 1.374574977447783, + "grad_norm": 1.8400824069976807, + "learning_rate": 0.000152694911017205, + "loss": 0.6547, + "step": 19809 + }, + { + "epoch": 1.3746443688848795, + "grad_norm": 1.8101942539215088, + "learning_rate": 0.00015269013569051007, + "loss": 0.5176, + "step": 19810 + }, + { + "epoch": 1.3747137603219763, + "grad_norm": 1.3759804964065552, + "learning_rate": 0.00015268536019747849, + "loss": 0.4561, + "step": 19811 + }, + { + "epoch": 1.374783151759073, + "grad_norm": 1.795461654663086, + "learning_rate": 0.00015268058453812535, + "loss": 0.6004, + "step": 19812 + }, + { + "epoch": 1.3748525431961696, + "grad_norm": 2.1625900268554688, + "learning_rate": 0.0001526758087124657, + "loss": 0.6218, + "step": 19813 + }, + { + "epoch": 1.3749219346332662, + "grad_norm": 1.718964695930481, + "learning_rate": 0.00015267103272051467, + "loss": 0.6015, + "step": 19814 + }, + { + "epoch": 1.3749913260703628, + "grad_norm": 1.6982004642486572, + "learning_rate": 0.0001526662565622873, + "loss": 0.5738, + "step": 19815 + }, + { + "epoch": 1.3750607175074596, + "grad_norm": 1.806675910949707, + "learning_rate": 0.00015266148023779862, + "loss": 0.584, + "step": 19816 + }, + { + "epoch": 1.3751301089445562, + "grad_norm": 1.823927879333496, + "learning_rate": 0.00015265670374706378, + "loss": 0.8503, + "step": 19817 + }, + { + "epoch": 1.3751995003816528, + "grad_norm": 1.790899634361267, + "learning_rate": 0.00015265192709009785, + "loss": 0.5164, + "step": 19818 + }, + { + "epoch": 1.3752688918187497, + "grad_norm": 1.508486270904541, + "learning_rate": 0.00015264715026691588, + "loss": 0.6042, + "step": 19819 + }, + { + "epoch": 1.3753382832558463, + "grad_norm": 2.0897364616394043, + "learning_rate": 0.00015264237327753297, + "loss": 0.7007, + "step": 19820 + }, + { + "epoch": 1.375407674692943, + "grad_norm": 1.7322534322738647, + "learning_rate": 0.0001526375961219642, + "loss": 0.6445, + "step": 19821 + }, + { + "epoch": 1.3754770661300395, + "grad_norm": 1.6648106575012207, + "learning_rate": 0.00015263281880022466, + "loss": 0.5706, + "step": 19822 + }, + { + "epoch": 1.3755464575671361, + "grad_norm": 2.1111881732940674, + "learning_rate": 0.00015262804131232937, + "loss": 0.6162, + "step": 19823 + }, + { + "epoch": 1.375615849004233, + "grad_norm": 1.8004413843154907, + "learning_rate": 0.0001526232636582935, + "loss": 0.6717, + "step": 19824 + }, + { + "epoch": 1.3756852404413296, + "grad_norm": 1.4115779399871826, + "learning_rate": 0.00015261848583813205, + "loss": 0.5625, + "step": 19825 + }, + { + "epoch": 1.3757546318784262, + "grad_norm": 1.496530532836914, + "learning_rate": 0.0001526137078518602, + "loss": 0.6087, + "step": 19826 + }, + { + "epoch": 1.3758240233155228, + "grad_norm": 1.4722528457641602, + "learning_rate": 0.00015260892969949294, + "loss": 0.5641, + "step": 19827 + }, + { + "epoch": 1.3758934147526194, + "grad_norm": 2.2313878536224365, + "learning_rate": 0.0001526041513810454, + "loss": 0.8592, + "step": 19828 + }, + { + "epoch": 1.3759628061897162, + "grad_norm": 1.9145119190216064, + "learning_rate": 0.0001525993728965327, + "loss": 0.7323, + "step": 19829 + }, + { + "epoch": 1.3760321976268128, + "grad_norm": 1.6475569009780884, + "learning_rate": 0.00015259459424596982, + "loss": 0.6345, + "step": 19830 + }, + { + "epoch": 1.3761015890639094, + "grad_norm": 1.5048843622207642, + "learning_rate": 0.00015258981542937193, + "loss": 0.6715, + "step": 19831 + }, + { + "epoch": 1.3761709805010063, + "grad_norm": 2.7430574893951416, + "learning_rate": 0.00015258503644675412, + "loss": 0.8734, + "step": 19832 + }, + { + "epoch": 1.376240371938103, + "grad_norm": 1.4363820552825928, + "learning_rate": 0.00015258025729813143, + "loss": 0.6343, + "step": 19833 + }, + { + "epoch": 1.3763097633751995, + "grad_norm": 1.4863975048065186, + "learning_rate": 0.00015257547798351896, + "loss": 0.4959, + "step": 19834 + }, + { + "epoch": 1.3763791548122961, + "grad_norm": 1.294142723083496, + "learning_rate": 0.00015257069850293183, + "loss": 0.6454, + "step": 19835 + }, + { + "epoch": 1.3764485462493927, + "grad_norm": 1.945261001586914, + "learning_rate": 0.0001525659188563851, + "loss": 0.6052, + "step": 19836 + }, + { + "epoch": 1.3765179376864896, + "grad_norm": 2.0784175395965576, + "learning_rate": 0.0001525611390438938, + "loss": 0.7034, + "step": 19837 + }, + { + "epoch": 1.3765873291235862, + "grad_norm": 1.994410753250122, + "learning_rate": 0.00015255635906547314, + "loss": 0.7004, + "step": 19838 + }, + { + "epoch": 1.3766567205606828, + "grad_norm": 2.290411949157715, + "learning_rate": 0.00015255157892113817, + "loss": 0.6267, + "step": 19839 + }, + { + "epoch": 1.3767261119977796, + "grad_norm": 1.504020094871521, + "learning_rate": 0.00015254679861090393, + "loss": 0.5438, + "step": 19840 + }, + { + "epoch": 1.376795503434876, + "grad_norm": 1.5530904531478882, + "learning_rate": 0.00015254201813478553, + "loss": 0.4722, + "step": 19841 + }, + { + "epoch": 1.3768648948719728, + "grad_norm": 1.4665977954864502, + "learning_rate": 0.0001525372374927981, + "loss": 0.6826, + "step": 19842 + }, + { + "epoch": 1.3769342863090694, + "grad_norm": 1.632465124130249, + "learning_rate": 0.00015253245668495673, + "loss": 0.7147, + "step": 19843 + }, + { + "epoch": 1.377003677746166, + "grad_norm": 3.5518572330474854, + "learning_rate": 0.00015252767571127644, + "loss": 0.6407, + "step": 19844 + }, + { + "epoch": 1.3770730691832629, + "grad_norm": 1.6756718158721924, + "learning_rate": 0.0001525228945717724, + "loss": 0.6282, + "step": 19845 + }, + { + "epoch": 1.3771424606203595, + "grad_norm": 1.8539475202560425, + "learning_rate": 0.00015251811326645967, + "loss": 0.8592, + "step": 19846 + }, + { + "epoch": 1.377211852057456, + "grad_norm": 2.5330779552459717, + "learning_rate": 0.00015251333179535335, + "loss": 0.7025, + "step": 19847 + }, + { + "epoch": 1.3772812434945527, + "grad_norm": 2.4330785274505615, + "learning_rate": 0.00015250855015846852, + "loss": 0.6596, + "step": 19848 + }, + { + "epoch": 1.3773506349316493, + "grad_norm": 1.790954351425171, + "learning_rate": 0.0001525037683558203, + "loss": 0.4401, + "step": 19849 + }, + { + "epoch": 1.3774200263687462, + "grad_norm": 2.0362398624420166, + "learning_rate": 0.0001524989863874238, + "loss": 0.9921, + "step": 19850 + }, + { + "epoch": 1.3774894178058428, + "grad_norm": 1.1518093347549438, + "learning_rate": 0.00015249420425329405, + "loss": 0.3848, + "step": 19851 + }, + { + "epoch": 1.3775588092429394, + "grad_norm": 2.2477800846099854, + "learning_rate": 0.00015248942195344623, + "loss": 0.6729, + "step": 19852 + }, + { + "epoch": 1.3776282006800362, + "grad_norm": 2.852031707763672, + "learning_rate": 0.00015248463948789538, + "loss": 0.7374, + "step": 19853 + }, + { + "epoch": 1.3776975921171326, + "grad_norm": 2.1874094009399414, + "learning_rate": 0.0001524798568566566, + "loss": 0.4956, + "step": 19854 + }, + { + "epoch": 1.3777669835542294, + "grad_norm": 2.432887315750122, + "learning_rate": 0.00015247507405974497, + "loss": 0.7727, + "step": 19855 + }, + { + "epoch": 1.377836374991326, + "grad_norm": 1.4327771663665771, + "learning_rate": 0.00015247029109717564, + "loss": 0.7396, + "step": 19856 + }, + { + "epoch": 1.3779057664284227, + "grad_norm": 3.113584041595459, + "learning_rate": 0.00015246550796896372, + "loss": 0.428, + "step": 19857 + }, + { + "epoch": 1.3779751578655195, + "grad_norm": 2.4792943000793457, + "learning_rate": 0.00015246072467512426, + "loss": 0.6733, + "step": 19858 + }, + { + "epoch": 1.378044549302616, + "grad_norm": 1.749353051185608, + "learning_rate": 0.0001524559412156724, + "loss": 0.5975, + "step": 19859 + }, + { + "epoch": 1.3781139407397127, + "grad_norm": 1.6943076848983765, + "learning_rate": 0.0001524511575906232, + "loss": 0.6805, + "step": 19860 + }, + { + "epoch": 1.3781833321768093, + "grad_norm": 1.4618314504623413, + "learning_rate": 0.00015244637379999176, + "loss": 0.5669, + "step": 19861 + }, + { + "epoch": 1.378252723613906, + "grad_norm": 1.5835975408554077, + "learning_rate": 0.00015244158984379323, + "loss": 0.4685, + "step": 19862 + }, + { + "epoch": 1.3783221150510028, + "grad_norm": 1.8724608421325684, + "learning_rate": 0.00015243680572204268, + "loss": 0.5616, + "step": 19863 + }, + { + "epoch": 1.3783915064880994, + "grad_norm": 2.0099940299987793, + "learning_rate": 0.0001524320214347552, + "loss": 0.6975, + "step": 19864 + }, + { + "epoch": 1.378460897925196, + "grad_norm": 2.610523223876953, + "learning_rate": 0.00015242723698194593, + "loss": 0.8423, + "step": 19865 + }, + { + "epoch": 1.3785302893622928, + "grad_norm": 1.9153711795806885, + "learning_rate": 0.00015242245236362997, + "loss": 0.6046, + "step": 19866 + }, + { + "epoch": 1.3785996807993894, + "grad_norm": 1.3026624917984009, + "learning_rate": 0.00015241766757982238, + "loss": 0.3934, + "step": 19867 + }, + { + "epoch": 1.378669072236486, + "grad_norm": 1.4269919395446777, + "learning_rate": 0.00015241288263053834, + "loss": 0.4962, + "step": 19868 + }, + { + "epoch": 1.3787384636735827, + "grad_norm": 2.435642957687378, + "learning_rate": 0.00015240809751579284, + "loss": 0.5681, + "step": 19869 + }, + { + "epoch": 1.3788078551106793, + "grad_norm": 2.235938549041748, + "learning_rate": 0.00015240331223560113, + "loss": 0.7172, + "step": 19870 + }, + { + "epoch": 1.378877246547776, + "grad_norm": 2.659695625305176, + "learning_rate": 0.0001523985267899782, + "loss": 0.8412, + "step": 19871 + }, + { + "epoch": 1.3789466379848727, + "grad_norm": 2.050136089324951, + "learning_rate": 0.0001523937411789392, + "loss": 0.6177, + "step": 19872 + }, + { + "epoch": 1.3790160294219693, + "grad_norm": 2.0754072666168213, + "learning_rate": 0.0001523889554024992, + "loss": 0.7525, + "step": 19873 + }, + { + "epoch": 1.379085420859066, + "grad_norm": 2.4760613441467285, + "learning_rate": 0.00015238416946067345, + "loss": 0.736, + "step": 19874 + }, + { + "epoch": 1.3791548122961625, + "grad_norm": 1.6558855772018433, + "learning_rate": 0.00015237938335347688, + "loss": 0.5013, + "step": 19875 + }, + { + "epoch": 1.3792242037332594, + "grad_norm": 2.071045160293579, + "learning_rate": 0.00015237459708092468, + "loss": 0.5935, + "step": 19876 + }, + { + "epoch": 1.379293595170356, + "grad_norm": 1.785677433013916, + "learning_rate": 0.00015236981064303197, + "loss": 0.6933, + "step": 19877 + }, + { + "epoch": 1.3793629866074526, + "grad_norm": 1.4772632122039795, + "learning_rate": 0.00015236502403981382, + "loss": 0.5869, + "step": 19878 + }, + { + "epoch": 1.3794323780445494, + "grad_norm": 1.7454228401184082, + "learning_rate": 0.0001523602372712854, + "loss": 0.5874, + "step": 19879 + }, + { + "epoch": 1.379501769481646, + "grad_norm": 2.2284188270568848, + "learning_rate": 0.00015235545033746175, + "loss": 0.7241, + "step": 19880 + }, + { + "epoch": 1.3795711609187427, + "grad_norm": 1.6176177263259888, + "learning_rate": 0.00015235066323835805, + "loss": 0.6929, + "step": 19881 + }, + { + "epoch": 1.3796405523558393, + "grad_norm": 1.8390496969223022, + "learning_rate": 0.00015234587597398938, + "loss": 0.5583, + "step": 19882 + }, + { + "epoch": 1.3797099437929359, + "grad_norm": 1.5788687467575073, + "learning_rate": 0.00015234108854437085, + "loss": 0.6718, + "step": 19883 + }, + { + "epoch": 1.3797793352300327, + "grad_norm": 2.3087873458862305, + "learning_rate": 0.00015233630094951757, + "loss": 0.55, + "step": 19884 + }, + { + "epoch": 1.3798487266671293, + "grad_norm": 2.207357168197632, + "learning_rate": 0.00015233151318944462, + "loss": 0.7196, + "step": 19885 + }, + { + "epoch": 1.379918118104226, + "grad_norm": 1.9066879749298096, + "learning_rate": 0.00015232672526416724, + "loss": 0.7023, + "step": 19886 + }, + { + "epoch": 1.3799875095413225, + "grad_norm": 2.3977906703948975, + "learning_rate": 0.00015232193717370035, + "loss": 0.5076, + "step": 19887 + }, + { + "epoch": 1.3800569009784192, + "grad_norm": 1.5366098880767822, + "learning_rate": 0.00015231714891805927, + "loss": 0.8155, + "step": 19888 + }, + { + "epoch": 1.380126292415516, + "grad_norm": 1.844801902770996, + "learning_rate": 0.00015231236049725899, + "loss": 0.6975, + "step": 19889 + }, + { + "epoch": 1.3801956838526126, + "grad_norm": 1.3900270462036133, + "learning_rate": 0.00015230757191131463, + "loss": 0.5971, + "step": 19890 + }, + { + "epoch": 1.3802650752897092, + "grad_norm": 1.6569983959197998, + "learning_rate": 0.00015230278316024135, + "loss": 0.6688, + "step": 19891 + }, + { + "epoch": 1.380334466726806, + "grad_norm": 2.088639736175537, + "learning_rate": 0.0001522979942440543, + "loss": 0.541, + "step": 19892 + }, + { + "epoch": 1.3804038581639027, + "grad_norm": 1.9553245306015015, + "learning_rate": 0.0001522932051627685, + "loss": 0.5043, + "step": 19893 + }, + { + "epoch": 1.3804732496009993, + "grad_norm": 2.0868070125579834, + "learning_rate": 0.00015228841591639908, + "loss": 0.685, + "step": 19894 + }, + { + "epoch": 1.3805426410380959, + "grad_norm": 1.3054579496383667, + "learning_rate": 0.0001522836265049613, + "loss": 0.606, + "step": 19895 + }, + { + "epoch": 1.3806120324751925, + "grad_norm": 1.8565970659255981, + "learning_rate": 0.0001522788369284701, + "loss": 0.6692, + "step": 19896 + }, + { + "epoch": 1.3806814239122893, + "grad_norm": 1.6599053144454956, + "learning_rate": 0.00015227404718694067, + "loss": 0.4744, + "step": 19897 + }, + { + "epoch": 1.380750815349386, + "grad_norm": 3.028510570526123, + "learning_rate": 0.0001522692572803882, + "loss": 0.6042, + "step": 19898 + }, + { + "epoch": 1.3808202067864825, + "grad_norm": 2.6769261360168457, + "learning_rate": 0.0001522644672088277, + "loss": 0.8202, + "step": 19899 + }, + { + "epoch": 1.3808895982235792, + "grad_norm": 2.0178372859954834, + "learning_rate": 0.00015225967697227435, + "loss": 0.8118, + "step": 19900 + }, + { + "epoch": 1.3809589896606758, + "grad_norm": 1.985137939453125, + "learning_rate": 0.00015225488657074327, + "loss": 0.8791, + "step": 19901 + }, + { + "epoch": 1.3810283810977726, + "grad_norm": 1.5783339738845825, + "learning_rate": 0.00015225009600424957, + "loss": 0.5006, + "step": 19902 + }, + { + "epoch": 1.3810977725348692, + "grad_norm": 1.9079352617263794, + "learning_rate": 0.0001522453052728084, + "loss": 0.7042, + "step": 19903 + }, + { + "epoch": 1.3811671639719658, + "grad_norm": 2.1682288646698, + "learning_rate": 0.00015224051437643486, + "loss": 0.8295, + "step": 19904 + }, + { + "epoch": 1.3812365554090626, + "grad_norm": 1.5122435092926025, + "learning_rate": 0.00015223572331514405, + "loss": 0.5397, + "step": 19905 + }, + { + "epoch": 1.3813059468461593, + "grad_norm": 2.157073736190796, + "learning_rate": 0.00015223093208895113, + "loss": 0.5543, + "step": 19906 + }, + { + "epoch": 1.3813753382832559, + "grad_norm": 1.442600131034851, + "learning_rate": 0.00015222614069787123, + "loss": 0.4172, + "step": 19907 + }, + { + "epoch": 1.3814447297203525, + "grad_norm": 1.5129376649856567, + "learning_rate": 0.00015222134914191947, + "loss": 0.4169, + "step": 19908 + }, + { + "epoch": 1.381514121157449, + "grad_norm": 1.9837884902954102, + "learning_rate": 0.00015221655742111095, + "loss": 0.9099, + "step": 19909 + }, + { + "epoch": 1.381583512594546, + "grad_norm": 1.934269666671753, + "learning_rate": 0.0001522117655354608, + "loss": 0.6064, + "step": 19910 + }, + { + "epoch": 1.3816529040316425, + "grad_norm": 1.7289543151855469, + "learning_rate": 0.0001522069734849842, + "loss": 0.5907, + "step": 19911 + }, + { + "epoch": 1.3817222954687391, + "grad_norm": 2.3082120418548584, + "learning_rate": 0.00015220218126969624, + "loss": 0.8885, + "step": 19912 + }, + { + "epoch": 1.3817916869058358, + "grad_norm": 2.0596630573272705, + "learning_rate": 0.00015219738888961205, + "loss": 0.6837, + "step": 19913 + }, + { + "epoch": 1.3818610783429324, + "grad_norm": 1.7923029661178589, + "learning_rate": 0.00015219259634474672, + "loss": 0.6385, + "step": 19914 + }, + { + "epoch": 1.3819304697800292, + "grad_norm": 1.57412588596344, + "learning_rate": 0.00015218780363511546, + "loss": 0.6289, + "step": 19915 + }, + { + "epoch": 1.3819998612171258, + "grad_norm": 1.872856616973877, + "learning_rate": 0.00015218301076073333, + "loss": 0.595, + "step": 19916 + }, + { + "epoch": 1.3820692526542224, + "grad_norm": 2.20391845703125, + "learning_rate": 0.00015217821772161552, + "loss": 0.7972, + "step": 19917 + }, + { + "epoch": 1.3821386440913193, + "grad_norm": 1.5839194059371948, + "learning_rate": 0.0001521734245177771, + "loss": 0.5662, + "step": 19918 + }, + { + "epoch": 1.3822080355284159, + "grad_norm": 1.9679710865020752, + "learning_rate": 0.00015216863114923324, + "loss": 0.6727, + "step": 19919 + }, + { + "epoch": 1.3822774269655125, + "grad_norm": 1.3044185638427734, + "learning_rate": 0.0001521638376159991, + "loss": 0.4919, + "step": 19920 + }, + { + "epoch": 1.382346818402609, + "grad_norm": 1.8827451467514038, + "learning_rate": 0.00015215904391808972, + "loss": 0.4559, + "step": 19921 + }, + { + "epoch": 1.3824162098397057, + "grad_norm": 1.4623510837554932, + "learning_rate": 0.00015215425005552034, + "loss": 0.677, + "step": 19922 + }, + { + "epoch": 1.3824856012768025, + "grad_norm": 1.5690209865570068, + "learning_rate": 0.00015214945602830602, + "loss": 0.7453, + "step": 19923 + }, + { + "epoch": 1.3825549927138991, + "grad_norm": 1.6360770463943481, + "learning_rate": 0.0001521446618364619, + "loss": 0.5563, + "step": 19924 + }, + { + "epoch": 1.3826243841509958, + "grad_norm": 1.514268398284912, + "learning_rate": 0.00015213986748000314, + "loss": 0.7558, + "step": 19925 + }, + { + "epoch": 1.3826937755880924, + "grad_norm": 1.7505453824996948, + "learning_rate": 0.00015213507295894492, + "loss": 0.5726, + "step": 19926 + }, + { + "epoch": 1.382763167025189, + "grad_norm": 2.20174503326416, + "learning_rate": 0.00015213027827330226, + "loss": 0.9258, + "step": 19927 + }, + { + "epoch": 1.3828325584622858, + "grad_norm": 1.7121379375457764, + "learning_rate": 0.00015212548342309037, + "loss": 0.513, + "step": 19928 + }, + { + "epoch": 1.3829019498993824, + "grad_norm": 2.1258692741394043, + "learning_rate": 0.00015212068840832443, + "loss": 0.7136, + "step": 19929 + }, + { + "epoch": 1.382971341336479, + "grad_norm": 3.234347343444824, + "learning_rate": 0.00015211589322901948, + "loss": 0.7902, + "step": 19930 + }, + { + "epoch": 1.3830407327735759, + "grad_norm": 2.296189069747925, + "learning_rate": 0.0001521110978851907, + "loss": 0.6557, + "step": 19931 + }, + { + "epoch": 1.3831101242106725, + "grad_norm": 1.4233869314193726, + "learning_rate": 0.00015210630237685318, + "loss": 0.7896, + "step": 19932 + }, + { + "epoch": 1.383179515647769, + "grad_norm": 2.2644035816192627, + "learning_rate": 0.00015210150670402222, + "loss": 0.7048, + "step": 19933 + }, + { + "epoch": 1.3832489070848657, + "grad_norm": 1.7975239753723145, + "learning_rate": 0.00015209671086671275, + "loss": 0.4882, + "step": 19934 + }, + { + "epoch": 1.3833182985219623, + "grad_norm": 1.554343342781067, + "learning_rate": 0.00015209191486494005, + "loss": 0.4731, + "step": 19935 + }, + { + "epoch": 1.3833876899590591, + "grad_norm": 1.6191520690917969, + "learning_rate": 0.00015208711869871925, + "loss": 0.5225, + "step": 19936 + }, + { + "epoch": 1.3834570813961558, + "grad_norm": 1.7813472747802734, + "learning_rate": 0.0001520823223680654, + "loss": 0.5589, + "step": 19937 + }, + { + "epoch": 1.3835264728332524, + "grad_norm": 2.2647886276245117, + "learning_rate": 0.00015207752587299375, + "loss": 0.7676, + "step": 19938 + }, + { + "epoch": 1.383595864270349, + "grad_norm": 1.4234071969985962, + "learning_rate": 0.00015207272921351938, + "loss": 0.4987, + "step": 19939 + }, + { + "epoch": 1.3836652557074456, + "grad_norm": 1.5432853698730469, + "learning_rate": 0.00015206793238965743, + "loss": 0.6341, + "step": 19940 + }, + { + "epoch": 1.3837346471445424, + "grad_norm": 1.4965909719467163, + "learning_rate": 0.00015206313540142306, + "loss": 0.5294, + "step": 19941 + }, + { + "epoch": 1.383804038581639, + "grad_norm": 1.8196585178375244, + "learning_rate": 0.00015205833824883142, + "loss": 0.776, + "step": 19942 + }, + { + "epoch": 1.3838734300187356, + "grad_norm": 1.798231840133667, + "learning_rate": 0.00015205354093189767, + "loss": 0.5428, + "step": 19943 + }, + { + "epoch": 1.3839428214558325, + "grad_norm": 1.642849087715149, + "learning_rate": 0.0001520487434506369, + "loss": 0.6566, + "step": 19944 + }, + { + "epoch": 1.384012212892929, + "grad_norm": 1.6223844289779663, + "learning_rate": 0.00015204394580506427, + "loss": 0.6869, + "step": 19945 + }, + { + "epoch": 1.3840816043300257, + "grad_norm": 1.7404690980911255, + "learning_rate": 0.00015203914799519498, + "loss": 0.7138, + "step": 19946 + }, + { + "epoch": 1.3841509957671223, + "grad_norm": 1.2418842315673828, + "learning_rate": 0.0001520343500210441, + "loss": 0.5215, + "step": 19947 + }, + { + "epoch": 1.384220387204219, + "grad_norm": 2.091412305831909, + "learning_rate": 0.00015202955188262681, + "loss": 0.8579, + "step": 19948 + }, + { + "epoch": 1.3842897786413157, + "grad_norm": 1.7788059711456299, + "learning_rate": 0.0001520247535799583, + "loss": 0.619, + "step": 19949 + }, + { + "epoch": 1.3843591700784124, + "grad_norm": 1.9572488069534302, + "learning_rate": 0.00015201995511305368, + "loss": 0.7183, + "step": 19950 + }, + { + "epoch": 1.384428561515509, + "grad_norm": 1.8760020732879639, + "learning_rate": 0.00015201515648192806, + "loss": 0.5602, + "step": 19951 + }, + { + "epoch": 1.3844979529526056, + "grad_norm": 1.69828200340271, + "learning_rate": 0.00015201035768659663, + "loss": 0.5116, + "step": 19952 + }, + { + "epoch": 1.3845673443897022, + "grad_norm": 2.0070018768310547, + "learning_rate": 0.0001520055587270746, + "loss": 0.6706, + "step": 19953 + }, + { + "epoch": 1.384636735826799, + "grad_norm": 1.4064973592758179, + "learning_rate": 0.00015200075960337698, + "loss": 0.4606, + "step": 19954 + }, + { + "epoch": 1.3847061272638956, + "grad_norm": 1.7995022535324097, + "learning_rate": 0.00015199596031551904, + "loss": 0.6728, + "step": 19955 + }, + { + "epoch": 1.3847755187009922, + "grad_norm": 1.6588822603225708, + "learning_rate": 0.00015199116086351587, + "loss": 0.663, + "step": 19956 + }, + { + "epoch": 1.384844910138089, + "grad_norm": 2.1293787956237793, + "learning_rate": 0.00015198636124738263, + "loss": 0.6886, + "step": 19957 + }, + { + "epoch": 1.3849143015751857, + "grad_norm": 1.7741408348083496, + "learning_rate": 0.00015198156146713448, + "loss": 0.6335, + "step": 19958 + }, + { + "epoch": 1.3849836930122823, + "grad_norm": 1.847408413887024, + "learning_rate": 0.00015197676152278657, + "loss": 0.8069, + "step": 19959 + }, + { + "epoch": 1.385053084449379, + "grad_norm": 2.5995728969573975, + "learning_rate": 0.0001519719614143541, + "loss": 0.7031, + "step": 19960 + }, + { + "epoch": 1.3851224758864755, + "grad_norm": 1.7675731182098389, + "learning_rate": 0.0001519671611418521, + "loss": 0.69, + "step": 19961 + }, + { + "epoch": 1.3851918673235724, + "grad_norm": 1.7555068731307983, + "learning_rate": 0.00015196236070529583, + "loss": 0.6564, + "step": 19962 + }, + { + "epoch": 1.385261258760669, + "grad_norm": 1.9666733741760254, + "learning_rate": 0.00015195756010470045, + "loss": 0.8162, + "step": 19963 + }, + { + "epoch": 1.3853306501977656, + "grad_norm": 1.5873175859451294, + "learning_rate": 0.00015195275934008103, + "loss": 0.5526, + "step": 19964 + }, + { + "epoch": 1.3854000416348622, + "grad_norm": 1.880584955215454, + "learning_rate": 0.0001519479584114528, + "loss": 0.5453, + "step": 19965 + }, + { + "epoch": 1.3854694330719588, + "grad_norm": 2.1084864139556885, + "learning_rate": 0.00015194315731883087, + "loss": 0.605, + "step": 19966 + }, + { + "epoch": 1.3855388245090556, + "grad_norm": 1.2287436723709106, + "learning_rate": 0.0001519383560622305, + "loss": 0.4477, + "step": 19967 + }, + { + "epoch": 1.3856082159461522, + "grad_norm": 1.9096227884292603, + "learning_rate": 0.0001519335546416667, + "loss": 0.6837, + "step": 19968 + }, + { + "epoch": 1.3856776073832489, + "grad_norm": 2.963939666748047, + "learning_rate": 0.00015192875305715468, + "loss": 0.6844, + "step": 19969 + }, + { + "epoch": 1.3857469988203457, + "grad_norm": 1.938314437866211, + "learning_rate": 0.0001519239513087096, + "loss": 0.5584, + "step": 19970 + }, + { + "epoch": 1.3858163902574423, + "grad_norm": 1.432527780532837, + "learning_rate": 0.0001519191493963467, + "loss": 0.5173, + "step": 19971 + }, + { + "epoch": 1.385885781694539, + "grad_norm": 1.9098783731460571, + "learning_rate": 0.00015191434732008102, + "loss": 0.9068, + "step": 19972 + }, + { + "epoch": 1.3859551731316355, + "grad_norm": 2.327971935272217, + "learning_rate": 0.0001519095450799278, + "loss": 0.7808, + "step": 19973 + }, + { + "epoch": 1.3860245645687321, + "grad_norm": 1.7136919498443604, + "learning_rate": 0.00015190474267590216, + "loss": 0.6335, + "step": 19974 + }, + { + "epoch": 1.386093956005829, + "grad_norm": 2.03422474861145, + "learning_rate": 0.00015189994010801925, + "loss": 0.625, + "step": 19975 + }, + { + "epoch": 1.3861633474429256, + "grad_norm": 1.6770991086959839, + "learning_rate": 0.00015189513737629427, + "loss": 0.7218, + "step": 19976 + }, + { + "epoch": 1.3862327388800222, + "grad_norm": 2.132046937942505, + "learning_rate": 0.00015189033448074235, + "loss": 0.5053, + "step": 19977 + }, + { + "epoch": 1.3863021303171188, + "grad_norm": 2.05466890335083, + "learning_rate": 0.00015188553142137868, + "loss": 0.8805, + "step": 19978 + }, + { + "epoch": 1.3863715217542154, + "grad_norm": 3.5390799045562744, + "learning_rate": 0.00015188072819821838, + "loss": 1.0235, + "step": 19979 + }, + { + "epoch": 1.3864409131913122, + "grad_norm": 1.5529236793518066, + "learning_rate": 0.00015187592481127665, + "loss": 0.4913, + "step": 19980 + }, + { + "epoch": 1.3865103046284089, + "grad_norm": 2.0571374893188477, + "learning_rate": 0.00015187112126056865, + "loss": 0.7837, + "step": 19981 + }, + { + "epoch": 1.3865796960655055, + "grad_norm": 1.7857779264450073, + "learning_rate": 0.00015186631754610954, + "loss": 0.6945, + "step": 19982 + }, + { + "epoch": 1.3866490875026023, + "grad_norm": 1.5251542329788208, + "learning_rate": 0.00015186151366791445, + "loss": 0.5592, + "step": 19983 + }, + { + "epoch": 1.386718478939699, + "grad_norm": 1.634055733680725, + "learning_rate": 0.00015185670962599863, + "loss": 0.6066, + "step": 19984 + }, + { + "epoch": 1.3867878703767955, + "grad_norm": 1.472957730293274, + "learning_rate": 0.00015185190542037718, + "loss": 0.5243, + "step": 19985 + }, + { + "epoch": 1.3868572618138921, + "grad_norm": 1.5028256177902222, + "learning_rate": 0.0001518471010510653, + "loss": 0.6099, + "step": 19986 + }, + { + "epoch": 1.3869266532509887, + "grad_norm": 2.4907140731811523, + "learning_rate": 0.00015184229651807806, + "loss": 0.8033, + "step": 19987 + }, + { + "epoch": 1.3869960446880856, + "grad_norm": 1.845404028892517, + "learning_rate": 0.0001518374918214308, + "loss": 0.7286, + "step": 19988 + }, + { + "epoch": 1.3870654361251822, + "grad_norm": 2.123868942260742, + "learning_rate": 0.00015183268696113852, + "loss": 0.7331, + "step": 19989 + }, + { + "epoch": 1.3871348275622788, + "grad_norm": 2.0835204124450684, + "learning_rate": 0.00015182788193721648, + "loss": 0.8103, + "step": 19990 + }, + { + "epoch": 1.3872042189993754, + "grad_norm": 3.2108898162841797, + "learning_rate": 0.00015182307674967987, + "loss": 1.1097, + "step": 19991 + }, + { + "epoch": 1.387273610436472, + "grad_norm": 1.506867527961731, + "learning_rate": 0.0001518182713985438, + "loss": 0.5503, + "step": 19992 + }, + { + "epoch": 1.3873430018735688, + "grad_norm": 1.5323138236999512, + "learning_rate": 0.00015181346588382343, + "loss": 0.5944, + "step": 19993 + }, + { + "epoch": 1.3874123933106655, + "grad_norm": 2.068242311477661, + "learning_rate": 0.00015180866020553398, + "loss": 0.6404, + "step": 19994 + }, + { + "epoch": 1.387481784747762, + "grad_norm": 1.2576993703842163, + "learning_rate": 0.0001518038543636906, + "loss": 0.6963, + "step": 19995 + }, + { + "epoch": 1.387551176184859, + "grad_norm": 1.8561289310455322, + "learning_rate": 0.00015179904835830845, + "loss": 0.5084, + "step": 19996 + }, + { + "epoch": 1.3876205676219555, + "grad_norm": 1.4982258081436157, + "learning_rate": 0.0001517942421894027, + "loss": 0.6052, + "step": 19997 + }, + { + "epoch": 1.3876899590590521, + "grad_norm": 2.334562063217163, + "learning_rate": 0.0001517894358569886, + "loss": 0.751, + "step": 19998 + }, + { + "epoch": 1.3877593504961487, + "grad_norm": 1.7423399686813354, + "learning_rate": 0.0001517846293610812, + "loss": 0.6944, + "step": 19999 + }, + { + "epoch": 1.3878287419332453, + "grad_norm": 1.8921380043029785, + "learning_rate": 0.00015177982270169575, + "loss": 0.5567, + "step": 20000 + }, + { + "epoch": 1.3878287419332453, + "eval_loss": 0.6905989050865173, + "eval_runtime": 3971.9327, + "eval_samples_per_second": 6.111, + "eval_steps_per_second": 0.764, + "step": 20000 + }, + { + "epoch": 1.3878981333703422, + "grad_norm": 2.2236664295196533, + "learning_rate": 0.00015177501587884738, + "loss": 0.6136, + "step": 20001 + }, + { + "epoch": 1.3879675248074388, + "grad_norm": 1.5039520263671875, + "learning_rate": 0.00015177020889255136, + "loss": 0.6541, + "step": 20002 + }, + { + "epoch": 1.3880369162445354, + "grad_norm": 2.1006877422332764, + "learning_rate": 0.00015176540174282273, + "loss": 0.5654, + "step": 20003 + }, + { + "epoch": 1.388106307681632, + "grad_norm": 2.223113536834717, + "learning_rate": 0.00015176059442967676, + "loss": 0.7045, + "step": 20004 + }, + { + "epoch": 1.3881756991187286, + "grad_norm": 2.027308940887451, + "learning_rate": 0.0001517557869531286, + "loss": 0.4783, + "step": 20005 + }, + { + "epoch": 1.3882450905558255, + "grad_norm": 2.447641611099243, + "learning_rate": 0.00015175097931319345, + "loss": 0.5385, + "step": 20006 + }, + { + "epoch": 1.388314481992922, + "grad_norm": 2.0745110511779785, + "learning_rate": 0.00015174617150988645, + "loss": 0.6453, + "step": 20007 + }, + { + "epoch": 1.3883838734300187, + "grad_norm": 2.2945775985717773, + "learning_rate": 0.00015174136354322276, + "loss": 0.6775, + "step": 20008 + }, + { + "epoch": 1.3884532648671155, + "grad_norm": 1.7815593481063843, + "learning_rate": 0.00015173655541321764, + "loss": 0.6353, + "step": 20009 + }, + { + "epoch": 1.3885226563042121, + "grad_norm": 1.5167397260665894, + "learning_rate": 0.00015173174711988617, + "loss": 0.4831, + "step": 20010 + }, + { + "epoch": 1.3885920477413087, + "grad_norm": 2.0055792331695557, + "learning_rate": 0.0001517269386632436, + "loss": 0.5483, + "step": 20011 + }, + { + "epoch": 1.3886614391784053, + "grad_norm": 1.3376095294952393, + "learning_rate": 0.00015172213004330506, + "loss": 0.4585, + "step": 20012 + }, + { + "epoch": 1.388730830615502, + "grad_norm": 1.7774107456207275, + "learning_rate": 0.00015171732126008578, + "loss": 0.7558, + "step": 20013 + }, + { + "epoch": 1.3888002220525988, + "grad_norm": 1.4333137273788452, + "learning_rate": 0.0001517125123136009, + "loss": 0.6061, + "step": 20014 + }, + { + "epoch": 1.3888696134896954, + "grad_norm": 2.467658042907715, + "learning_rate": 0.00015170770320386564, + "loss": 0.9444, + "step": 20015 + }, + { + "epoch": 1.388939004926792, + "grad_norm": 1.7936842441558838, + "learning_rate": 0.00015170289393089513, + "loss": 0.6379, + "step": 20016 + }, + { + "epoch": 1.3890083963638886, + "grad_norm": 1.3731040954589844, + "learning_rate": 0.0001516980844947046, + "loss": 0.6051, + "step": 20017 + }, + { + "epoch": 1.3890777878009852, + "grad_norm": 2.9842989444732666, + "learning_rate": 0.00015169327489530923, + "loss": 0.6933, + "step": 20018 + }, + { + "epoch": 1.389147179238082, + "grad_norm": 1.483363389968872, + "learning_rate": 0.00015168846513272417, + "loss": 0.5163, + "step": 20019 + }, + { + "epoch": 1.3892165706751787, + "grad_norm": 1.8009753227233887, + "learning_rate": 0.00015168365520696464, + "loss": 0.4597, + "step": 20020 + }, + { + "epoch": 1.3892859621122753, + "grad_norm": 1.3600918054580688, + "learning_rate": 0.0001516788451180458, + "loss": 0.6097, + "step": 20021 + }, + { + "epoch": 1.3893553535493721, + "grad_norm": 1.5341992378234863, + "learning_rate": 0.00015167403486598282, + "loss": 0.5755, + "step": 20022 + }, + { + "epoch": 1.3894247449864687, + "grad_norm": 1.593080759048462, + "learning_rate": 0.00015166922445079093, + "loss": 0.5092, + "step": 20023 + }, + { + "epoch": 1.3894941364235653, + "grad_norm": 1.8900272846221924, + "learning_rate": 0.00015166441387248528, + "loss": 0.7446, + "step": 20024 + }, + { + "epoch": 1.389563527860662, + "grad_norm": 3.5921542644500732, + "learning_rate": 0.00015165960313108108, + "loss": 0.9008, + "step": 20025 + }, + { + "epoch": 1.3896329192977586, + "grad_norm": 2.0042006969451904, + "learning_rate": 0.0001516547922265935, + "loss": 0.7553, + "step": 20026 + }, + { + "epoch": 1.3897023107348554, + "grad_norm": 1.518573522567749, + "learning_rate": 0.00015164998115903776, + "loss": 0.5747, + "step": 20027 + }, + { + "epoch": 1.389771702171952, + "grad_norm": 2.5576188564300537, + "learning_rate": 0.000151645169928429, + "loss": 0.7429, + "step": 20028 + }, + { + "epoch": 1.3898410936090486, + "grad_norm": 1.7242966890335083, + "learning_rate": 0.00015164035853478243, + "loss": 0.4818, + "step": 20029 + }, + { + "epoch": 1.3899104850461452, + "grad_norm": 1.7673348188400269, + "learning_rate": 0.00015163554697811323, + "loss": 0.737, + "step": 20030 + }, + { + "epoch": 1.3899798764832418, + "grad_norm": 2.437782049179077, + "learning_rate": 0.0001516307352584366, + "loss": 0.6144, + "step": 20031 + }, + { + "epoch": 1.3900492679203387, + "grad_norm": 2.190185546875, + "learning_rate": 0.00015162592337576773, + "loss": 0.6726, + "step": 20032 + }, + { + "epoch": 1.3901186593574353, + "grad_norm": 1.5942507982254028, + "learning_rate": 0.00015162111133012186, + "loss": 0.4946, + "step": 20033 + }, + { + "epoch": 1.390188050794532, + "grad_norm": 1.846614122390747, + "learning_rate": 0.00015161629912151408, + "loss": 0.6317, + "step": 20034 + }, + { + "epoch": 1.3902574422316287, + "grad_norm": 2.1343915462493896, + "learning_rate": 0.00015161148674995963, + "loss": 0.6329, + "step": 20035 + }, + { + "epoch": 1.3903268336687253, + "grad_norm": 1.4797275066375732, + "learning_rate": 0.00015160667421547372, + "loss": 0.4448, + "step": 20036 + }, + { + "epoch": 1.390396225105822, + "grad_norm": 2.463381290435791, + "learning_rate": 0.0001516018615180715, + "loss": 0.6848, + "step": 20037 + }, + { + "epoch": 1.3904656165429186, + "grad_norm": 2.0178959369659424, + "learning_rate": 0.0001515970486577682, + "loss": 0.6521, + "step": 20038 + }, + { + "epoch": 1.3905350079800152, + "grad_norm": 1.784626841545105, + "learning_rate": 0.000151592235634579, + "loss": 0.9162, + "step": 20039 + }, + { + "epoch": 1.390604399417112, + "grad_norm": 2.5569143295288086, + "learning_rate": 0.00015158742244851916, + "loss": 0.6219, + "step": 20040 + }, + { + "epoch": 1.3906737908542086, + "grad_norm": 1.6973671913146973, + "learning_rate": 0.00015158260909960375, + "loss": 0.5933, + "step": 20041 + }, + { + "epoch": 1.3907431822913052, + "grad_norm": 1.60431706905365, + "learning_rate": 0.00015157779558784803, + "loss": 0.6628, + "step": 20042 + }, + { + "epoch": 1.3908125737284018, + "grad_norm": 1.69297194480896, + "learning_rate": 0.00015157298191326724, + "loss": 0.7904, + "step": 20043 + }, + { + "epoch": 1.3908819651654984, + "grad_norm": 1.0965538024902344, + "learning_rate": 0.00015156816807587647, + "loss": 0.5314, + "step": 20044 + }, + { + "epoch": 1.3909513566025953, + "grad_norm": 2.44045090675354, + "learning_rate": 0.00015156335407569103, + "loss": 0.8299, + "step": 20045 + }, + { + "epoch": 1.391020748039692, + "grad_norm": 1.65873384475708, + "learning_rate": 0.000151558539912726, + "loss": 0.6436, + "step": 20046 + }, + { + "epoch": 1.3910901394767885, + "grad_norm": 7.172749042510986, + "learning_rate": 0.0001515537255869967, + "loss": 0.9894, + "step": 20047 + }, + { + "epoch": 1.3911595309138853, + "grad_norm": 1.9360206127166748, + "learning_rate": 0.00015154891109851827, + "loss": 0.7235, + "step": 20048 + }, + { + "epoch": 1.391228922350982, + "grad_norm": 1.940015196800232, + "learning_rate": 0.00015154409644730586, + "loss": 0.6502, + "step": 20049 + }, + { + "epoch": 1.3912983137880786, + "grad_norm": 2.1305646896362305, + "learning_rate": 0.00015153928163337476, + "loss": 0.6887, + "step": 20050 + }, + { + "epoch": 1.3913677052251752, + "grad_norm": 1.7704174518585205, + "learning_rate": 0.0001515344666567401, + "loss": 0.528, + "step": 20051 + }, + { + "epoch": 1.3914370966622718, + "grad_norm": 1.8005857467651367, + "learning_rate": 0.0001515296515174171, + "loss": 0.6782, + "step": 20052 + }, + { + "epoch": 1.3915064880993686, + "grad_norm": 1.7543805837631226, + "learning_rate": 0.00015152483621542099, + "loss": 0.6295, + "step": 20053 + }, + { + "epoch": 1.3915758795364652, + "grad_norm": 1.6723241806030273, + "learning_rate": 0.00015152002075076695, + "loss": 0.4877, + "step": 20054 + }, + { + "epoch": 1.3916452709735618, + "grad_norm": 1.5858873128890991, + "learning_rate": 0.0001515152051234702, + "loss": 0.5912, + "step": 20055 + }, + { + "epoch": 1.3917146624106584, + "grad_norm": 2.0883467197418213, + "learning_rate": 0.0001515103893335459, + "loss": 0.8605, + "step": 20056 + }, + { + "epoch": 1.391784053847755, + "grad_norm": 1.6133369207382202, + "learning_rate": 0.00015150557338100928, + "loss": 0.4595, + "step": 20057 + }, + { + "epoch": 1.3918534452848519, + "grad_norm": 1.746343731880188, + "learning_rate": 0.00015150075726587554, + "loss": 0.6673, + "step": 20058 + }, + { + "epoch": 1.3919228367219485, + "grad_norm": 2.794940233230591, + "learning_rate": 0.00015149594098815987, + "loss": 0.7033, + "step": 20059 + }, + { + "epoch": 1.391992228159045, + "grad_norm": 1.8341714143753052, + "learning_rate": 0.00015149112454787754, + "loss": 0.6541, + "step": 20060 + }, + { + "epoch": 1.392061619596142, + "grad_norm": 1.4373979568481445, + "learning_rate": 0.00015148630794504364, + "loss": 0.537, + "step": 20061 + }, + { + "epoch": 1.3921310110332386, + "grad_norm": 1.4108608961105347, + "learning_rate": 0.0001514814911796735, + "loss": 0.506, + "step": 20062 + }, + { + "epoch": 1.3922004024703352, + "grad_norm": 2.4382553100585938, + "learning_rate": 0.00015147667425178223, + "loss": 0.723, + "step": 20063 + }, + { + "epoch": 1.3922697939074318, + "grad_norm": 2.004329204559326, + "learning_rate": 0.00015147185716138508, + "loss": 0.7309, + "step": 20064 + }, + { + "epoch": 1.3923391853445284, + "grad_norm": 1.626484751701355, + "learning_rate": 0.00015146703990849723, + "loss": 0.649, + "step": 20065 + }, + { + "epoch": 1.3924085767816252, + "grad_norm": 2.2279608249664307, + "learning_rate": 0.0001514622224931339, + "loss": 0.6557, + "step": 20066 + }, + { + "epoch": 1.3924779682187218, + "grad_norm": 1.6167430877685547, + "learning_rate": 0.00015145740491531034, + "loss": 0.6809, + "step": 20067 + }, + { + "epoch": 1.3925473596558184, + "grad_norm": 2.00787353515625, + "learning_rate": 0.00015145258717504172, + "loss": 0.5631, + "step": 20068 + }, + { + "epoch": 1.3926167510929153, + "grad_norm": 2.0661845207214355, + "learning_rate": 0.00015144776927234323, + "loss": 0.6923, + "step": 20069 + }, + { + "epoch": 1.3926861425300117, + "grad_norm": 1.7449198961257935, + "learning_rate": 0.0001514429512072301, + "loss": 0.7067, + "step": 20070 + }, + { + "epoch": 1.3927555339671085, + "grad_norm": 1.9213694334030151, + "learning_rate": 0.00015143813297971757, + "loss": 0.7675, + "step": 20071 + }, + { + "epoch": 1.392824925404205, + "grad_norm": 1.781704306602478, + "learning_rate": 0.0001514333145898208, + "loss": 0.5888, + "step": 20072 + }, + { + "epoch": 1.3928943168413017, + "grad_norm": 1.5743886232376099, + "learning_rate": 0.00015142849603755503, + "loss": 0.5192, + "step": 20073 + }, + { + "epoch": 1.3929637082783985, + "grad_norm": 1.4661123752593994, + "learning_rate": 0.00015142367732293548, + "loss": 0.5695, + "step": 20074 + }, + { + "epoch": 1.3930330997154952, + "grad_norm": 1.5771527290344238, + "learning_rate": 0.0001514188584459773, + "loss": 0.7895, + "step": 20075 + }, + { + "epoch": 1.3931024911525918, + "grad_norm": 1.5715539455413818, + "learning_rate": 0.00015141403940669576, + "loss": 0.5689, + "step": 20076 + }, + { + "epoch": 1.3931718825896884, + "grad_norm": 1.4666435718536377, + "learning_rate": 0.0001514092202051061, + "loss": 0.6911, + "step": 20077 + }, + { + "epoch": 1.393241274026785, + "grad_norm": 2.076648235321045, + "learning_rate": 0.00015140440084122347, + "loss": 0.5891, + "step": 20078 + }, + { + "epoch": 1.3933106654638818, + "grad_norm": 1.6244298219680786, + "learning_rate": 0.00015139958131506312, + "loss": 0.744, + "step": 20079 + }, + { + "epoch": 1.3933800569009784, + "grad_norm": 1.8290222883224487, + "learning_rate": 0.0001513947616266403, + "loss": 0.5225, + "step": 20080 + }, + { + "epoch": 1.393449448338075, + "grad_norm": 2.533283233642578, + "learning_rate": 0.00015138994177597013, + "loss": 0.8113, + "step": 20081 + }, + { + "epoch": 1.3935188397751719, + "grad_norm": 2.0052807331085205, + "learning_rate": 0.00015138512176306786, + "loss": 0.5387, + "step": 20082 + }, + { + "epoch": 1.3935882312122683, + "grad_norm": 1.938892126083374, + "learning_rate": 0.00015138030158794874, + "loss": 0.4538, + "step": 20083 + }, + { + "epoch": 1.393657622649365, + "grad_norm": 1.9110347032546997, + "learning_rate": 0.00015137548125062798, + "loss": 0.5928, + "step": 20084 + }, + { + "epoch": 1.3937270140864617, + "grad_norm": 1.7112592458724976, + "learning_rate": 0.00015137066075112077, + "loss": 0.5823, + "step": 20085 + }, + { + "epoch": 1.3937964055235583, + "grad_norm": 3.069269895553589, + "learning_rate": 0.00015136584008944236, + "loss": 0.6051, + "step": 20086 + }, + { + "epoch": 1.3938657969606552, + "grad_norm": 2.132067918777466, + "learning_rate": 0.00015136101926560794, + "loss": 0.6611, + "step": 20087 + }, + { + "epoch": 1.3939351883977518, + "grad_norm": 1.5642513036727905, + "learning_rate": 0.00015135619827963276, + "loss": 0.542, + "step": 20088 + }, + { + "epoch": 1.3940045798348484, + "grad_norm": 1.6794238090515137, + "learning_rate": 0.00015135137713153199, + "loss": 0.5431, + "step": 20089 + }, + { + "epoch": 1.394073971271945, + "grad_norm": 1.4033488035202026, + "learning_rate": 0.0001513465558213209, + "loss": 0.5982, + "step": 20090 + }, + { + "epoch": 1.3941433627090416, + "grad_norm": 1.6812134981155396, + "learning_rate": 0.00015134173434901466, + "loss": 0.6216, + "step": 20091 + }, + { + "epoch": 1.3942127541461384, + "grad_norm": 1.4187743663787842, + "learning_rate": 0.00015133691271462858, + "loss": 0.3984, + "step": 20092 + }, + { + "epoch": 1.394282145583235, + "grad_norm": 2.484875202178955, + "learning_rate": 0.00015133209091817778, + "loss": 0.8616, + "step": 20093 + }, + { + "epoch": 1.3943515370203317, + "grad_norm": 1.7310380935668945, + "learning_rate": 0.00015132726895967754, + "loss": 0.7068, + "step": 20094 + }, + { + "epoch": 1.3944209284574285, + "grad_norm": 1.6968597173690796, + "learning_rate": 0.00015132244683914306, + "loss": 0.4239, + "step": 20095 + }, + { + "epoch": 1.394490319894525, + "grad_norm": 2.4916603565216064, + "learning_rate": 0.0001513176245565896, + "loss": 0.7534, + "step": 20096 + }, + { + "epoch": 1.3945597113316217, + "grad_norm": 1.5796918869018555, + "learning_rate": 0.0001513128021120323, + "loss": 0.7737, + "step": 20097 + }, + { + "epoch": 1.3946291027687183, + "grad_norm": 1.5819140672683716, + "learning_rate": 0.0001513079795054865, + "loss": 0.5102, + "step": 20098 + }, + { + "epoch": 1.394698494205815, + "grad_norm": 1.7643029689788818, + "learning_rate": 0.0001513031567369673, + "loss": 0.7763, + "step": 20099 + }, + { + "epoch": 1.3947678856429118, + "grad_norm": 1.9810080528259277, + "learning_rate": 0.00015129833380649, + "loss": 0.7876, + "step": 20100 + }, + { + "epoch": 1.3948372770800084, + "grad_norm": 2.252601146697998, + "learning_rate": 0.00015129351071406982, + "loss": 0.8057, + "step": 20101 + }, + { + "epoch": 1.394906668517105, + "grad_norm": 1.6914968490600586, + "learning_rate": 0.000151288687459722, + "loss": 0.6601, + "step": 20102 + }, + { + "epoch": 1.3949760599542016, + "grad_norm": 2.011092185974121, + "learning_rate": 0.00015128386404346172, + "loss": 0.8794, + "step": 20103 + }, + { + "epoch": 1.3950454513912982, + "grad_norm": 2.46403431892395, + "learning_rate": 0.00015127904046530424, + "loss": 0.5941, + "step": 20104 + }, + { + "epoch": 1.395114842828395, + "grad_norm": 1.64536452293396, + "learning_rate": 0.00015127421672526478, + "loss": 0.6808, + "step": 20105 + }, + { + "epoch": 1.3951842342654917, + "grad_norm": 1.6441481113433838, + "learning_rate": 0.00015126939282335858, + "loss": 0.5041, + "step": 20106 + }, + { + "epoch": 1.3952536257025883, + "grad_norm": 1.4611455202102661, + "learning_rate": 0.00015126456875960083, + "loss": 0.593, + "step": 20107 + }, + { + "epoch": 1.395323017139685, + "grad_norm": 1.717397689819336, + "learning_rate": 0.00015125974453400676, + "loss": 0.5377, + "step": 20108 + }, + { + "epoch": 1.3953924085767817, + "grad_norm": 2.5298397541046143, + "learning_rate": 0.00015125492014659168, + "loss": 0.7404, + "step": 20109 + }, + { + "epoch": 1.3954618000138783, + "grad_norm": 1.819663405418396, + "learning_rate": 0.00015125009559737072, + "loss": 0.6835, + "step": 20110 + }, + { + "epoch": 1.395531191450975, + "grad_norm": 1.484110713005066, + "learning_rate": 0.00015124527088635915, + "loss": 0.545, + "step": 20111 + }, + { + "epoch": 1.3956005828880715, + "grad_norm": 1.4064496755599976, + "learning_rate": 0.00015124044601357226, + "loss": 0.6128, + "step": 20112 + }, + { + "epoch": 1.3956699743251684, + "grad_norm": 1.8214410543441772, + "learning_rate": 0.0001512356209790252, + "loss": 0.6681, + "step": 20113 + }, + { + "epoch": 1.395739365762265, + "grad_norm": 1.6650261878967285, + "learning_rate": 0.00015123079578273322, + "loss": 0.6504, + "step": 20114 + }, + { + "epoch": 1.3958087571993616, + "grad_norm": 1.5899176597595215, + "learning_rate": 0.00015122597042471157, + "loss": 0.7555, + "step": 20115 + }, + { + "epoch": 1.3958781486364582, + "grad_norm": 2.1010117530822754, + "learning_rate": 0.0001512211449049755, + "loss": 0.6413, + "step": 20116 + }, + { + "epoch": 1.3959475400735548, + "grad_norm": 1.869015097618103, + "learning_rate": 0.00015121631922354015, + "loss": 0.5701, + "step": 20117 + }, + { + "epoch": 1.3960169315106516, + "grad_norm": 1.316972017288208, + "learning_rate": 0.00015121149338042085, + "loss": 0.4149, + "step": 20118 + }, + { + "epoch": 1.3960863229477483, + "grad_norm": 2.060431957244873, + "learning_rate": 0.00015120666737563284, + "loss": 0.7611, + "step": 20119 + }, + { + "epoch": 1.3961557143848449, + "grad_norm": 2.1273441314697266, + "learning_rate": 0.00015120184120919128, + "loss": 0.6424, + "step": 20120 + }, + { + "epoch": 1.3962251058219417, + "grad_norm": 2.561317205429077, + "learning_rate": 0.00015119701488111148, + "loss": 0.6693, + "step": 20121 + }, + { + "epoch": 1.3962944972590383, + "grad_norm": 1.356608271598816, + "learning_rate": 0.00015119218839140865, + "loss": 0.5768, + "step": 20122 + }, + { + "epoch": 1.396363888696135, + "grad_norm": 2.428978443145752, + "learning_rate": 0.000151187361740098, + "loss": 0.567, + "step": 20123 + }, + { + "epoch": 1.3964332801332315, + "grad_norm": 1.6849626302719116, + "learning_rate": 0.00015118253492719477, + "loss": 0.5902, + "step": 20124 + }, + { + "epoch": 1.3965026715703281, + "grad_norm": 1.9477758407592773, + "learning_rate": 0.00015117770795271425, + "loss": 0.8411, + "step": 20125 + }, + { + "epoch": 1.396572063007425, + "grad_norm": 2.0636720657348633, + "learning_rate": 0.00015117288081667164, + "loss": 0.5759, + "step": 20126 + }, + { + "epoch": 1.3966414544445216, + "grad_norm": 2.157674789428711, + "learning_rate": 0.00015116805351908215, + "loss": 0.636, + "step": 20127 + }, + { + "epoch": 1.3967108458816182, + "grad_norm": 1.5980422496795654, + "learning_rate": 0.0001511632260599611, + "loss": 0.6065, + "step": 20128 + }, + { + "epoch": 1.3967802373187148, + "grad_norm": 1.6983764171600342, + "learning_rate": 0.00015115839843932367, + "loss": 0.5736, + "step": 20129 + }, + { + "epoch": 1.3968496287558114, + "grad_norm": 2.462905168533325, + "learning_rate": 0.00015115357065718506, + "loss": 0.5422, + "step": 20130 + }, + { + "epoch": 1.3969190201929083, + "grad_norm": 1.4954044818878174, + "learning_rate": 0.00015114874271356064, + "loss": 0.4405, + "step": 20131 + }, + { + "epoch": 1.3969884116300049, + "grad_norm": 1.9919536113739014, + "learning_rate": 0.0001511439146084655, + "loss": 0.7821, + "step": 20132 + }, + { + "epoch": 1.3970578030671015, + "grad_norm": 2.4486355781555176, + "learning_rate": 0.000151139086341915, + "loss": 0.5651, + "step": 20133 + }, + { + "epoch": 1.3971271945041983, + "grad_norm": 2.1364691257476807, + "learning_rate": 0.00015113425791392433, + "loss": 0.8471, + "step": 20134 + }, + { + "epoch": 1.397196585941295, + "grad_norm": 1.7198799848556519, + "learning_rate": 0.00015112942932450873, + "loss": 0.6754, + "step": 20135 + }, + { + "epoch": 1.3972659773783915, + "grad_norm": 2.075786590576172, + "learning_rate": 0.00015112460057368346, + "loss": 0.7138, + "step": 20136 + }, + { + "epoch": 1.3973353688154881, + "grad_norm": 1.5855520963668823, + "learning_rate": 0.00015111977166146377, + "loss": 0.5556, + "step": 20137 + }, + { + "epoch": 1.3974047602525848, + "grad_norm": 2.3152666091918945, + "learning_rate": 0.00015111494258786487, + "loss": 0.7776, + "step": 20138 + }, + { + "epoch": 1.3974741516896816, + "grad_norm": 2.0223636627197266, + "learning_rate": 0.00015111011335290205, + "loss": 0.7782, + "step": 20139 + }, + { + "epoch": 1.3975435431267782, + "grad_norm": 1.9531787633895874, + "learning_rate": 0.0001511052839565905, + "loss": 0.6456, + "step": 20140 + }, + { + "epoch": 1.3976129345638748, + "grad_norm": 1.8362993001937866, + "learning_rate": 0.0001511004543989455, + "loss": 0.6054, + "step": 20141 + }, + { + "epoch": 1.3976823260009714, + "grad_norm": 2.644238233566284, + "learning_rate": 0.00015109562467998232, + "loss": 0.9141, + "step": 20142 + }, + { + "epoch": 1.397751717438068, + "grad_norm": 1.9944370985031128, + "learning_rate": 0.00015109079479971619, + "loss": 0.6142, + "step": 20143 + }, + { + "epoch": 1.3978211088751649, + "grad_norm": 1.2802256345748901, + "learning_rate": 0.00015108596475816232, + "loss": 0.4047, + "step": 20144 + }, + { + "epoch": 1.3978905003122615, + "grad_norm": 1.4505040645599365, + "learning_rate": 0.00015108113455533602, + "loss": 0.476, + "step": 20145 + }, + { + "epoch": 1.397959891749358, + "grad_norm": 1.8343000411987305, + "learning_rate": 0.00015107630419125244, + "loss": 0.7401, + "step": 20146 + }, + { + "epoch": 1.398029283186455, + "grad_norm": 1.4250092506408691, + "learning_rate": 0.00015107147366592695, + "loss": 0.4894, + "step": 20147 + }, + { + "epoch": 1.3980986746235515, + "grad_norm": 1.7688188552856445, + "learning_rate": 0.00015106664297937473, + "loss": 0.6059, + "step": 20148 + }, + { + "epoch": 1.3981680660606481, + "grad_norm": 1.737086296081543, + "learning_rate": 0.00015106181213161106, + "loss": 0.6857, + "step": 20149 + }, + { + "epoch": 1.3982374574977448, + "grad_norm": 3.2883646488189697, + "learning_rate": 0.00015105698112265117, + "loss": 0.9052, + "step": 20150 + }, + { + "epoch": 1.3983068489348414, + "grad_norm": 2.8261210918426514, + "learning_rate": 0.00015105214995251028, + "loss": 0.5987, + "step": 20151 + }, + { + "epoch": 1.3983762403719382, + "grad_norm": 1.9115831851959229, + "learning_rate": 0.00015104731862120369, + "loss": 0.6266, + "step": 20152 + }, + { + "epoch": 1.3984456318090348, + "grad_norm": 1.5334254503250122, + "learning_rate": 0.00015104248712874662, + "loss": 0.5328, + "step": 20153 + }, + { + "epoch": 1.3985150232461314, + "grad_norm": 2.9630889892578125, + "learning_rate": 0.00015103765547515438, + "loss": 0.5902, + "step": 20154 + }, + { + "epoch": 1.398584414683228, + "grad_norm": 1.769497275352478, + "learning_rate": 0.00015103282366044218, + "loss": 0.588, + "step": 20155 + }, + { + "epoch": 1.3986538061203246, + "grad_norm": 2.0165233612060547, + "learning_rate": 0.00015102799168462524, + "loss": 0.6252, + "step": 20156 + }, + { + "epoch": 1.3987231975574215, + "grad_norm": 1.5400837659835815, + "learning_rate": 0.00015102315954771892, + "loss": 0.6416, + "step": 20157 + }, + { + "epoch": 1.398792588994518, + "grad_norm": 1.7502071857452393, + "learning_rate": 0.00015101832724973832, + "loss": 0.5943, + "step": 20158 + }, + { + "epoch": 1.3988619804316147, + "grad_norm": 1.4043816328048706, + "learning_rate": 0.0001510134947906988, + "loss": 0.4715, + "step": 20159 + }, + { + "epoch": 1.3989313718687115, + "grad_norm": 1.4236624240875244, + "learning_rate": 0.00015100866217061562, + "loss": 0.5041, + "step": 20160 + }, + { + "epoch": 1.3990007633058081, + "grad_norm": 1.981632947921753, + "learning_rate": 0.00015100382938950402, + "loss": 0.598, + "step": 20161 + }, + { + "epoch": 1.3990701547429047, + "grad_norm": 1.6318975687026978, + "learning_rate": 0.00015099899644737923, + "loss": 0.6313, + "step": 20162 + }, + { + "epoch": 1.3991395461800014, + "grad_norm": 1.5823898315429688, + "learning_rate": 0.00015099416334425652, + "loss": 0.7102, + "step": 20163 + }, + { + "epoch": 1.399208937617098, + "grad_norm": 1.520118236541748, + "learning_rate": 0.00015098933008015117, + "loss": 0.6754, + "step": 20164 + }, + { + "epoch": 1.3992783290541948, + "grad_norm": 1.3794828653335571, + "learning_rate": 0.00015098449665507842, + "loss": 0.5638, + "step": 20165 + }, + { + "epoch": 1.3993477204912914, + "grad_norm": 1.6711454391479492, + "learning_rate": 0.0001509796630690535, + "loss": 0.6512, + "step": 20166 + }, + { + "epoch": 1.399417111928388, + "grad_norm": 1.8751680850982666, + "learning_rate": 0.00015097482932209172, + "loss": 0.8459, + "step": 20167 + }, + { + "epoch": 1.3994865033654846, + "grad_norm": 1.3659957647323608, + "learning_rate": 0.00015096999541420835, + "loss": 0.5961, + "step": 20168 + }, + { + "epoch": 1.3995558948025812, + "grad_norm": 2.834521532058716, + "learning_rate": 0.00015096516134541855, + "loss": 0.7554, + "step": 20169 + }, + { + "epoch": 1.399625286239678, + "grad_norm": 1.49254310131073, + "learning_rate": 0.0001509603271157377, + "loss": 0.6297, + "step": 20170 + }, + { + "epoch": 1.3996946776767747, + "grad_norm": 1.7510061264038086, + "learning_rate": 0.000150955492725181, + "loss": 0.816, + "step": 20171 + }, + { + "epoch": 1.3997640691138713, + "grad_norm": 1.3551373481750488, + "learning_rate": 0.00015095065817376374, + "loss": 0.3677, + "step": 20172 + }, + { + "epoch": 1.3998334605509681, + "grad_norm": 1.5381660461425781, + "learning_rate": 0.0001509458234615011, + "loss": 0.5487, + "step": 20173 + }, + { + "epoch": 1.3999028519880647, + "grad_norm": 1.8891515731811523, + "learning_rate": 0.00015094098858840847, + "loss": 0.7466, + "step": 20174 + }, + { + "epoch": 1.3999722434251614, + "grad_norm": 1.4345849752426147, + "learning_rate": 0.00015093615355450102, + "loss": 0.6861, + "step": 20175 + }, + { + "epoch": 1.400041634862258, + "grad_norm": 1.691623330116272, + "learning_rate": 0.00015093131835979405, + "loss": 0.655, + "step": 20176 + }, + { + "epoch": 1.4001110262993546, + "grad_norm": 2.1733131408691406, + "learning_rate": 0.0001509264830043028, + "loss": 0.6493, + "step": 20177 + }, + { + "epoch": 1.4001804177364514, + "grad_norm": 1.8962656259536743, + "learning_rate": 0.00015092164748804255, + "loss": 0.6438, + "step": 20178 + }, + { + "epoch": 1.400249809173548, + "grad_norm": 3.5212039947509766, + "learning_rate": 0.0001509168118110286, + "loss": 0.7831, + "step": 20179 + }, + { + "epoch": 1.4003192006106446, + "grad_norm": 1.5237635374069214, + "learning_rate": 0.00015091197597327615, + "loss": 0.6528, + "step": 20180 + }, + { + "epoch": 1.4003885920477412, + "grad_norm": 1.8148196935653687, + "learning_rate": 0.00015090713997480056, + "loss": 0.6067, + "step": 20181 + }, + { + "epoch": 1.4004579834848379, + "grad_norm": 2.0118508338928223, + "learning_rate": 0.00015090230381561694, + "loss": 0.6089, + "step": 20182 + }, + { + "epoch": 1.4005273749219347, + "grad_norm": 1.4682834148406982, + "learning_rate": 0.00015089746749574073, + "loss": 0.673, + "step": 20183 + }, + { + "epoch": 1.4005967663590313, + "grad_norm": 1.8518251180648804, + "learning_rate": 0.0001508926310151871, + "loss": 0.7171, + "step": 20184 + }, + { + "epoch": 1.400666157796128, + "grad_norm": 1.8573102951049805, + "learning_rate": 0.00015088779437397133, + "loss": 0.7613, + "step": 20185 + }, + { + "epoch": 1.4007355492332247, + "grad_norm": 1.5355781316757202, + "learning_rate": 0.00015088295757210868, + "loss": 0.5304, + "step": 20186 + }, + { + "epoch": 1.4008049406703214, + "grad_norm": 2.107583522796631, + "learning_rate": 0.00015087812060961446, + "loss": 0.834, + "step": 20187 + }, + { + "epoch": 1.400874332107418, + "grad_norm": 2.9907636642456055, + "learning_rate": 0.0001508732834865039, + "loss": 0.8591, + "step": 20188 + }, + { + "epoch": 1.4009437235445146, + "grad_norm": 2.0495107173919678, + "learning_rate": 0.00015086844620279232, + "loss": 0.8966, + "step": 20189 + }, + { + "epoch": 1.4010131149816112, + "grad_norm": 1.8805756568908691, + "learning_rate": 0.00015086360875849492, + "loss": 0.4729, + "step": 20190 + }, + { + "epoch": 1.401082506418708, + "grad_norm": 1.2418525218963623, + "learning_rate": 0.00015085877115362704, + "loss": 0.5382, + "step": 20191 + }, + { + "epoch": 1.4011518978558046, + "grad_norm": 1.9250026941299438, + "learning_rate": 0.00015085393338820389, + "loss": 0.8228, + "step": 20192 + }, + { + "epoch": 1.4012212892929012, + "grad_norm": 1.9235559701919556, + "learning_rate": 0.0001508490954622408, + "loss": 0.4874, + "step": 20193 + }, + { + "epoch": 1.4012906807299979, + "grad_norm": 1.5431361198425293, + "learning_rate": 0.00015084425737575297, + "loss": 0.5517, + "step": 20194 + }, + { + "epoch": 1.4013600721670945, + "grad_norm": 1.4918551445007324, + "learning_rate": 0.0001508394191287558, + "loss": 0.4245, + "step": 20195 + }, + { + "epoch": 1.4014294636041913, + "grad_norm": 1.5266422033309937, + "learning_rate": 0.0001508345807212644, + "loss": 0.6185, + "step": 20196 + }, + { + "epoch": 1.401498855041288, + "grad_norm": 2.2316713333129883, + "learning_rate": 0.00015082974215329414, + "loss": 0.9519, + "step": 20197 + }, + { + "epoch": 1.4015682464783845, + "grad_norm": 2.458966016769409, + "learning_rate": 0.0001508249034248603, + "loss": 0.6694, + "step": 20198 + }, + { + "epoch": 1.4016376379154813, + "grad_norm": 1.4816107749938965, + "learning_rate": 0.00015082006453597817, + "loss": 0.391, + "step": 20199 + }, + { + "epoch": 1.401707029352578, + "grad_norm": 1.4257906675338745, + "learning_rate": 0.00015081522548666295, + "loss": 0.3645, + "step": 20200 + }, + { + "epoch": 1.4017764207896746, + "grad_norm": 1.9965741634368896, + "learning_rate": 0.00015081038627692996, + "loss": 0.6562, + "step": 20201 + }, + { + "epoch": 1.4018458122267712, + "grad_norm": 1.6364994049072266, + "learning_rate": 0.00015080554690679446, + "loss": 0.5673, + "step": 20202 + }, + { + "epoch": 1.4019152036638678, + "grad_norm": 1.898818016052246, + "learning_rate": 0.0001508007073762718, + "loss": 0.5574, + "step": 20203 + }, + { + "epoch": 1.4019845951009646, + "grad_norm": 1.9054205417633057, + "learning_rate": 0.00015079586768537717, + "loss": 0.8319, + "step": 20204 + }, + { + "epoch": 1.4020539865380612, + "grad_norm": 1.6741055250167847, + "learning_rate": 0.00015079102783412587, + "loss": 0.6534, + "step": 20205 + }, + { + "epoch": 1.4021233779751578, + "grad_norm": 1.7223566770553589, + "learning_rate": 0.0001507861878225332, + "loss": 0.602, + "step": 20206 + }, + { + "epoch": 1.4021927694122545, + "grad_norm": 1.8779217004776, + "learning_rate": 0.0001507813476506144, + "loss": 0.6422, + "step": 20207 + }, + { + "epoch": 1.402262160849351, + "grad_norm": 2.5830953121185303, + "learning_rate": 0.0001507765073183848, + "loss": 0.7296, + "step": 20208 + }, + { + "epoch": 1.402331552286448, + "grad_norm": 1.5783941745758057, + "learning_rate": 0.00015077166682585968, + "loss": 0.4679, + "step": 20209 + }, + { + "epoch": 1.4024009437235445, + "grad_norm": 1.6375203132629395, + "learning_rate": 0.00015076682617305427, + "loss": 0.5328, + "step": 20210 + }, + { + "epoch": 1.4024703351606411, + "grad_norm": 1.811993956565857, + "learning_rate": 0.00015076198535998386, + "loss": 0.6063, + "step": 20211 + }, + { + "epoch": 1.402539726597738, + "grad_norm": 1.6702425479888916, + "learning_rate": 0.0001507571443866638, + "loss": 0.522, + "step": 20212 + }, + { + "epoch": 1.4026091180348346, + "grad_norm": 2.597508192062378, + "learning_rate": 0.0001507523032531093, + "loss": 0.8322, + "step": 20213 + }, + { + "epoch": 1.4026785094719312, + "grad_norm": 1.4423869848251343, + "learning_rate": 0.00015074746195933565, + "loss": 0.8166, + "step": 20214 + }, + { + "epoch": 1.4027479009090278, + "grad_norm": 1.3970928192138672, + "learning_rate": 0.00015074262050535816, + "loss": 0.4555, + "step": 20215 + }, + { + "epoch": 1.4028172923461244, + "grad_norm": 2.1387417316436768, + "learning_rate": 0.00015073777889119212, + "loss": 0.7712, + "step": 20216 + }, + { + "epoch": 1.4028866837832212, + "grad_norm": 2.2383713722229004, + "learning_rate": 0.00015073293711685278, + "loss": 0.8326, + "step": 20217 + }, + { + "epoch": 1.4029560752203178, + "grad_norm": 1.4370636940002441, + "learning_rate": 0.0001507280951823555, + "loss": 0.6451, + "step": 20218 + }, + { + "epoch": 1.4030254666574145, + "grad_norm": 1.5995241403579712, + "learning_rate": 0.00015072325308771543, + "loss": 0.535, + "step": 20219 + }, + { + "epoch": 1.403094858094511, + "grad_norm": 2.5049357414245605, + "learning_rate": 0.00015071841083294797, + "loss": 0.6644, + "step": 20220 + }, + { + "epoch": 1.4031642495316077, + "grad_norm": 1.9169986248016357, + "learning_rate": 0.00015071356841806836, + "loss": 0.7305, + "step": 20221 + }, + { + "epoch": 1.4032336409687045, + "grad_norm": 2.3040783405303955, + "learning_rate": 0.00015070872584309192, + "loss": 0.6629, + "step": 20222 + }, + { + "epoch": 1.4033030324058011, + "grad_norm": 1.5268166065216064, + "learning_rate": 0.0001507038831080339, + "loss": 0.4185, + "step": 20223 + }, + { + "epoch": 1.4033724238428977, + "grad_norm": 1.9626104831695557, + "learning_rate": 0.00015069904021290962, + "loss": 0.6695, + "step": 20224 + }, + { + "epoch": 1.4034418152799946, + "grad_norm": 1.6744613647460938, + "learning_rate": 0.00015069419715773433, + "loss": 0.524, + "step": 20225 + }, + { + "epoch": 1.4035112067170912, + "grad_norm": 1.478836178779602, + "learning_rate": 0.00015068935394252337, + "loss": 0.4996, + "step": 20226 + }, + { + "epoch": 1.4035805981541878, + "grad_norm": 1.8346471786499023, + "learning_rate": 0.00015068451056729195, + "loss": 0.7525, + "step": 20227 + }, + { + "epoch": 1.4036499895912844, + "grad_norm": 1.740748643875122, + "learning_rate": 0.00015067966703205548, + "loss": 0.6714, + "step": 20228 + }, + { + "epoch": 1.403719381028381, + "grad_norm": 1.6944944858551025, + "learning_rate": 0.0001506748233368291, + "loss": 0.758, + "step": 20229 + }, + { + "epoch": 1.4037887724654778, + "grad_norm": 1.6456636190414429, + "learning_rate": 0.00015066997948162824, + "loss": 0.7108, + "step": 20230 + }, + { + "epoch": 1.4038581639025745, + "grad_norm": 1.9621810913085938, + "learning_rate": 0.00015066513546646812, + "loss": 0.6033, + "step": 20231 + }, + { + "epoch": 1.403927555339671, + "grad_norm": 2.662498950958252, + "learning_rate": 0.00015066029129136402, + "loss": 0.4558, + "step": 20232 + }, + { + "epoch": 1.4039969467767677, + "grad_norm": 1.0623304843902588, + "learning_rate": 0.00015065544695633132, + "loss": 0.5116, + "step": 20233 + }, + { + "epoch": 1.4040663382138643, + "grad_norm": 2.1144840717315674, + "learning_rate": 0.0001506506024613852, + "loss": 0.8181, + "step": 20234 + }, + { + "epoch": 1.4041357296509611, + "grad_norm": 2.357577323913574, + "learning_rate": 0.00015064575780654102, + "loss": 0.823, + "step": 20235 + }, + { + "epoch": 1.4042051210880577, + "grad_norm": 1.8963539600372314, + "learning_rate": 0.00015064091299181408, + "loss": 0.5996, + "step": 20236 + }, + { + "epoch": 1.4042745125251543, + "grad_norm": 1.5305366516113281, + "learning_rate": 0.00015063606801721962, + "loss": 0.5614, + "step": 20237 + }, + { + "epoch": 1.4043439039622512, + "grad_norm": 1.547203779220581, + "learning_rate": 0.00015063122288277298, + "loss": 0.4933, + "step": 20238 + }, + { + "epoch": 1.4044132953993478, + "grad_norm": 2.2011513710021973, + "learning_rate": 0.00015062637758848942, + "loss": 0.8785, + "step": 20239 + }, + { + "epoch": 1.4044826868364444, + "grad_norm": 1.3209887742996216, + "learning_rate": 0.00015062153213438433, + "loss": 0.4437, + "step": 20240 + }, + { + "epoch": 1.404552078273541, + "grad_norm": 1.3783103227615356, + "learning_rate": 0.00015061668652047289, + "loss": 0.567, + "step": 20241 + }, + { + "epoch": 1.4046214697106376, + "grad_norm": 1.5263170003890991, + "learning_rate": 0.0001506118407467704, + "loss": 0.4891, + "step": 20242 + }, + { + "epoch": 1.4046908611477344, + "grad_norm": 1.416528582572937, + "learning_rate": 0.00015060699481329225, + "loss": 0.592, + "step": 20243 + }, + { + "epoch": 1.404760252584831, + "grad_norm": 2.3271894454956055, + "learning_rate": 0.0001506021487200537, + "loss": 0.7107, + "step": 20244 + }, + { + "epoch": 1.4048296440219277, + "grad_norm": 1.827679991722107, + "learning_rate": 0.00015059730246707, + "loss": 0.7034, + "step": 20245 + }, + { + "epoch": 1.4048990354590243, + "grad_norm": 1.5967191457748413, + "learning_rate": 0.00015059245605435648, + "loss": 0.6349, + "step": 20246 + }, + { + "epoch": 1.404968426896121, + "grad_norm": 1.7585400342941284, + "learning_rate": 0.00015058760948192848, + "loss": 0.8966, + "step": 20247 + }, + { + "epoch": 1.4050378183332177, + "grad_norm": 1.3326239585876465, + "learning_rate": 0.00015058276274980127, + "loss": 0.5968, + "step": 20248 + }, + { + "epoch": 1.4051072097703143, + "grad_norm": 2.078768730163574, + "learning_rate": 0.00015057791585799013, + "loss": 0.6381, + "step": 20249 + }, + { + "epoch": 1.405176601207411, + "grad_norm": 1.7593469619750977, + "learning_rate": 0.00015057306880651039, + "loss": 0.5464, + "step": 20250 + }, + { + "epoch": 1.4052459926445078, + "grad_norm": 1.640249252319336, + "learning_rate": 0.00015056822159537734, + "loss": 0.541, + "step": 20251 + }, + { + "epoch": 1.4053153840816044, + "grad_norm": 1.7161005735397339, + "learning_rate": 0.00015056337422460623, + "loss": 0.5613, + "step": 20252 + }, + { + "epoch": 1.405384775518701, + "grad_norm": 1.2929878234863281, + "learning_rate": 0.00015055852669421249, + "loss": 0.5983, + "step": 20253 + }, + { + "epoch": 1.4054541669557976, + "grad_norm": 2.131805896759033, + "learning_rate": 0.0001505536790042113, + "loss": 0.6963, + "step": 20254 + }, + { + "epoch": 1.4055235583928942, + "grad_norm": 2.561140775680542, + "learning_rate": 0.00015054883115461805, + "loss": 0.6041, + "step": 20255 + }, + { + "epoch": 1.405592949829991, + "grad_norm": 2.4247336387634277, + "learning_rate": 0.00015054398314544794, + "loss": 0.5882, + "step": 20256 + }, + { + "epoch": 1.4056623412670877, + "grad_norm": 1.485290765762329, + "learning_rate": 0.0001505391349767164, + "loss": 0.4489, + "step": 20257 + }, + { + "epoch": 1.4057317327041843, + "grad_norm": 1.690400242805481, + "learning_rate": 0.00015053428664843867, + "loss": 0.5544, + "step": 20258 + }, + { + "epoch": 1.4058011241412809, + "grad_norm": 1.9057263135910034, + "learning_rate": 0.00015052943816063003, + "loss": 0.6595, + "step": 20259 + }, + { + "epoch": 1.4058705155783775, + "grad_norm": 2.281769275665283, + "learning_rate": 0.00015052458951330584, + "loss": 0.5446, + "step": 20260 + }, + { + "epoch": 1.4059399070154743, + "grad_norm": 1.8158584833145142, + "learning_rate": 0.00015051974070648137, + "loss": 0.6921, + "step": 20261 + }, + { + "epoch": 1.406009298452571, + "grad_norm": 2.181522846221924, + "learning_rate": 0.00015051489174017191, + "loss": 0.866, + "step": 20262 + }, + { + "epoch": 1.4060786898896676, + "grad_norm": 2.386354446411133, + "learning_rate": 0.00015051004261439283, + "loss": 0.9929, + "step": 20263 + }, + { + "epoch": 1.4061480813267644, + "grad_norm": 1.7562414407730103, + "learning_rate": 0.00015050519332915943, + "loss": 0.6652, + "step": 20264 + }, + { + "epoch": 1.406217472763861, + "grad_norm": 1.8053922653198242, + "learning_rate": 0.00015050034388448697, + "loss": 0.5322, + "step": 20265 + }, + { + "epoch": 1.4062868642009576, + "grad_norm": 1.883225917816162, + "learning_rate": 0.00015049549428039076, + "loss": 0.6246, + "step": 20266 + }, + { + "epoch": 1.4063562556380542, + "grad_norm": 1.5429009199142456, + "learning_rate": 0.00015049064451688615, + "loss": 0.6908, + "step": 20267 + }, + { + "epoch": 1.4064256470751508, + "grad_norm": 1.3906065225601196, + "learning_rate": 0.00015048579459398846, + "loss": 0.6909, + "step": 20268 + }, + { + "epoch": 1.4064950385122477, + "grad_norm": 1.718207597732544, + "learning_rate": 0.00015048094451171296, + "loss": 0.5779, + "step": 20269 + }, + { + "epoch": 1.4065644299493443, + "grad_norm": 2.3735873699188232, + "learning_rate": 0.00015047609427007496, + "loss": 0.4947, + "step": 20270 + }, + { + "epoch": 1.4066338213864409, + "grad_norm": 1.9307746887207031, + "learning_rate": 0.0001504712438690898, + "loss": 0.6639, + "step": 20271 + }, + { + "epoch": 1.4067032128235375, + "grad_norm": 1.8605281114578247, + "learning_rate": 0.0001504663933087728, + "loss": 0.5523, + "step": 20272 + }, + { + "epoch": 1.406772604260634, + "grad_norm": 1.8977535963058472, + "learning_rate": 0.0001504615425891392, + "loss": 0.6312, + "step": 20273 + }, + { + "epoch": 1.406841995697731, + "grad_norm": 3.47961163520813, + "learning_rate": 0.00015045669171020437, + "loss": 0.628, + "step": 20274 + }, + { + "epoch": 1.4069113871348276, + "grad_norm": 1.973327875137329, + "learning_rate": 0.00015045184067198365, + "loss": 0.7134, + "step": 20275 + }, + { + "epoch": 1.4069807785719242, + "grad_norm": 1.7937029600143433, + "learning_rate": 0.00015044698947449232, + "loss": 0.6714, + "step": 20276 + }, + { + "epoch": 1.407050170009021, + "grad_norm": 1.687002420425415, + "learning_rate": 0.00015044213811774568, + "loss": 0.4473, + "step": 20277 + }, + { + "epoch": 1.4071195614461176, + "grad_norm": 1.3915778398513794, + "learning_rate": 0.0001504372866017591, + "loss": 0.5423, + "step": 20278 + }, + { + "epoch": 1.4071889528832142, + "grad_norm": 1.903671145439148, + "learning_rate": 0.0001504324349265478, + "loss": 0.5417, + "step": 20279 + }, + { + "epoch": 1.4072583443203108, + "grad_norm": 1.655938982963562, + "learning_rate": 0.0001504275830921272, + "loss": 0.5913, + "step": 20280 + }, + { + "epoch": 1.4073277357574074, + "grad_norm": 1.9049469232559204, + "learning_rate": 0.00015042273109851253, + "loss": 0.6553, + "step": 20281 + }, + { + "epoch": 1.4073971271945043, + "grad_norm": 1.6135571002960205, + "learning_rate": 0.00015041787894571918, + "loss": 0.6429, + "step": 20282 + }, + { + "epoch": 1.4074665186316009, + "grad_norm": 1.7023564577102661, + "learning_rate": 0.00015041302663376243, + "loss": 0.5186, + "step": 20283 + }, + { + "epoch": 1.4075359100686975, + "grad_norm": 1.9927105903625488, + "learning_rate": 0.0001504081741626576, + "loss": 0.5608, + "step": 20284 + }, + { + "epoch": 1.407605301505794, + "grad_norm": 1.3539091348648071, + "learning_rate": 0.00015040332153242, + "loss": 0.4841, + "step": 20285 + }, + { + "epoch": 1.4076746929428907, + "grad_norm": 1.8247451782226562, + "learning_rate": 0.00015039846874306497, + "loss": 0.7181, + "step": 20286 + }, + { + "epoch": 1.4077440843799875, + "grad_norm": 2.3855998516082764, + "learning_rate": 0.00015039361579460785, + "loss": 0.7157, + "step": 20287 + }, + { + "epoch": 1.4078134758170842, + "grad_norm": 1.8606854677200317, + "learning_rate": 0.00015038876268706392, + "loss": 0.5577, + "step": 20288 + }, + { + "epoch": 1.4078828672541808, + "grad_norm": 1.586734652519226, + "learning_rate": 0.00015038390942044848, + "loss": 0.6618, + "step": 20289 + }, + { + "epoch": 1.4079522586912776, + "grad_norm": 1.8339827060699463, + "learning_rate": 0.0001503790559947769, + "loss": 0.625, + "step": 20290 + }, + { + "epoch": 1.4080216501283742, + "grad_norm": 1.7386987209320068, + "learning_rate": 0.0001503742024100645, + "loss": 0.4679, + "step": 20291 + }, + { + "epoch": 1.4080910415654708, + "grad_norm": 2.612896203994751, + "learning_rate": 0.0001503693486663266, + "loss": 0.6521, + "step": 20292 + }, + { + "epoch": 1.4081604330025674, + "grad_norm": 2.2942683696746826, + "learning_rate": 0.00015036449476357847, + "loss": 0.8884, + "step": 20293 + }, + { + "epoch": 1.408229824439664, + "grad_norm": 1.7427639961242676, + "learning_rate": 0.0001503596407018355, + "loss": 0.6355, + "step": 20294 + }, + { + "epoch": 1.4082992158767609, + "grad_norm": 1.715379238128662, + "learning_rate": 0.00015035478648111296, + "loss": 0.7395, + "step": 20295 + }, + { + "epoch": 1.4083686073138575, + "grad_norm": 1.4597702026367188, + "learning_rate": 0.00015034993210142621, + "loss": 0.4906, + "step": 20296 + }, + { + "epoch": 1.408437998750954, + "grad_norm": 1.5509777069091797, + "learning_rate": 0.00015034507756279058, + "loss": 0.6032, + "step": 20297 + }, + { + "epoch": 1.4085073901880507, + "grad_norm": 2.725245952606201, + "learning_rate": 0.00015034022286522137, + "loss": 0.6937, + "step": 20298 + }, + { + "epoch": 1.4085767816251473, + "grad_norm": 2.222395896911621, + "learning_rate": 0.00015033536800873395, + "loss": 0.8408, + "step": 20299 + }, + { + "epoch": 1.4086461730622442, + "grad_norm": 1.912100076675415, + "learning_rate": 0.00015033051299334358, + "loss": 0.804, + "step": 20300 + }, + { + "epoch": 1.4087155644993408, + "grad_norm": 1.2396929264068604, + "learning_rate": 0.0001503256578190656, + "loss": 0.5195, + "step": 20301 + }, + { + "epoch": 1.4087849559364374, + "grad_norm": 1.9474705457687378, + "learning_rate": 0.00015032080248591543, + "loss": 0.7765, + "step": 20302 + }, + { + "epoch": 1.4088543473735342, + "grad_norm": 1.502404808998108, + "learning_rate": 0.00015031594699390824, + "loss": 0.731, + "step": 20303 + }, + { + "epoch": 1.4089237388106308, + "grad_norm": 1.593798279762268, + "learning_rate": 0.00015031109134305948, + "loss": 0.513, + "step": 20304 + }, + { + "epoch": 1.4089931302477274, + "grad_norm": 2.1392714977264404, + "learning_rate": 0.00015030623553338444, + "loss": 0.6876, + "step": 20305 + }, + { + "epoch": 1.409062521684824, + "grad_norm": 1.7180202007293701, + "learning_rate": 0.00015030137956489848, + "loss": 0.6532, + "step": 20306 + }, + { + "epoch": 1.4091319131219207, + "grad_norm": 1.723012089729309, + "learning_rate": 0.00015029652343761689, + "loss": 0.5463, + "step": 20307 + }, + { + "epoch": 1.4092013045590175, + "grad_norm": 2.1753735542297363, + "learning_rate": 0.000150291667151555, + "loss": 0.777, + "step": 20308 + }, + { + "epoch": 1.409270695996114, + "grad_norm": 1.574678897857666, + "learning_rate": 0.00015028681070672814, + "loss": 0.5704, + "step": 20309 + }, + { + "epoch": 1.4093400874332107, + "grad_norm": 2.1047868728637695, + "learning_rate": 0.00015028195410315168, + "loss": 0.639, + "step": 20310 + }, + { + "epoch": 1.4094094788703075, + "grad_norm": 1.352887511253357, + "learning_rate": 0.00015027709734084088, + "loss": 0.5977, + "step": 20311 + }, + { + "epoch": 1.409478870307404, + "grad_norm": 1.9096567630767822, + "learning_rate": 0.00015027224041981114, + "loss": 0.6438, + "step": 20312 + }, + { + "epoch": 1.4095482617445008, + "grad_norm": 1.917803168296814, + "learning_rate": 0.00015026738334007777, + "loss": 0.7078, + "step": 20313 + }, + { + "epoch": 1.4096176531815974, + "grad_norm": 1.475817322731018, + "learning_rate": 0.0001502625261016561, + "loss": 0.7025, + "step": 20314 + }, + { + "epoch": 1.409687044618694, + "grad_norm": 2.1502909660339355, + "learning_rate": 0.00015025766870456147, + "loss": 0.928, + "step": 20315 + }, + { + "epoch": 1.4097564360557908, + "grad_norm": 1.4975148439407349, + "learning_rate": 0.00015025281114880922, + "loss": 0.6565, + "step": 20316 + }, + { + "epoch": 1.4098258274928874, + "grad_norm": 1.482061505317688, + "learning_rate": 0.00015024795343441467, + "loss": 0.5153, + "step": 20317 + }, + { + "epoch": 1.409895218929984, + "grad_norm": 1.1506738662719727, + "learning_rate": 0.00015024309556139315, + "loss": 0.5417, + "step": 20318 + }, + { + "epoch": 1.4099646103670807, + "grad_norm": 1.5915720462799072, + "learning_rate": 0.00015023823752976001, + "loss": 0.4931, + "step": 20319 + }, + { + "epoch": 1.4100340018041773, + "grad_norm": 1.7100939750671387, + "learning_rate": 0.00015023337933953062, + "loss": 0.7253, + "step": 20320 + }, + { + "epoch": 1.410103393241274, + "grad_norm": 1.9617316722869873, + "learning_rate": 0.0001502285209907202, + "loss": 0.8684, + "step": 20321 + }, + { + "epoch": 1.4101727846783707, + "grad_norm": 2.16872239112854, + "learning_rate": 0.00015022366248334423, + "loss": 0.6932, + "step": 20322 + }, + { + "epoch": 1.4102421761154673, + "grad_norm": 2.6603071689605713, + "learning_rate": 0.00015021880381741799, + "loss": 0.6496, + "step": 20323 + }, + { + "epoch": 1.4103115675525641, + "grad_norm": 1.266218662261963, + "learning_rate": 0.00015021394499295678, + "loss": 0.4909, + "step": 20324 + }, + { + "epoch": 1.4103809589896605, + "grad_norm": 1.6764472723007202, + "learning_rate": 0.000150209086009976, + "loss": 0.5858, + "step": 20325 + }, + { + "epoch": 1.4104503504267574, + "grad_norm": 1.5034884214401245, + "learning_rate": 0.00015020422686849093, + "loss": 0.7007, + "step": 20326 + }, + { + "epoch": 1.410519741863854, + "grad_norm": 2.5656819343566895, + "learning_rate": 0.00015019936756851697, + "loss": 0.756, + "step": 20327 + }, + { + "epoch": 1.4105891333009506, + "grad_norm": 1.694890022277832, + "learning_rate": 0.0001501945081100694, + "loss": 0.5507, + "step": 20328 + }, + { + "epoch": 1.4106585247380474, + "grad_norm": 1.803073763847351, + "learning_rate": 0.0001501896484931636, + "loss": 0.6814, + "step": 20329 + }, + { + "epoch": 1.410727916175144, + "grad_norm": 2.5303781032562256, + "learning_rate": 0.00015018478871781495, + "loss": 0.7772, + "step": 20330 + }, + { + "epoch": 1.4107973076122406, + "grad_norm": 2.0100948810577393, + "learning_rate": 0.00015017992878403868, + "loss": 0.7893, + "step": 20331 + }, + { + "epoch": 1.4108666990493373, + "grad_norm": 1.250588297843933, + "learning_rate": 0.0001501750686918502, + "loss": 0.7049, + "step": 20332 + }, + { + "epoch": 1.4109360904864339, + "grad_norm": 2.270949602127075, + "learning_rate": 0.00015017020844126488, + "loss": 0.7822, + "step": 20333 + }, + { + "epoch": 1.4110054819235307, + "grad_norm": 1.500404953956604, + "learning_rate": 0.00015016534803229803, + "loss": 0.6632, + "step": 20334 + }, + { + "epoch": 1.4110748733606273, + "grad_norm": 1.3960356712341309, + "learning_rate": 0.00015016048746496497, + "loss": 0.6842, + "step": 20335 + }, + { + "epoch": 1.411144264797724, + "grad_norm": 1.6569973230361938, + "learning_rate": 0.0001501556267392811, + "loss": 0.6668, + "step": 20336 + }, + { + "epoch": 1.4112136562348208, + "grad_norm": 1.6857465505599976, + "learning_rate": 0.00015015076585526177, + "loss": 0.5024, + "step": 20337 + }, + { + "epoch": 1.4112830476719174, + "grad_norm": 1.45856773853302, + "learning_rate": 0.0001501459048129222, + "loss": 0.4653, + "step": 20338 + }, + { + "epoch": 1.411352439109014, + "grad_norm": 1.6954740285873413, + "learning_rate": 0.00015014104361227786, + "loss": 0.6975, + "step": 20339 + }, + { + "epoch": 1.4114218305461106, + "grad_norm": 2.764193534851074, + "learning_rate": 0.0001501361822533441, + "loss": 0.8965, + "step": 20340 + }, + { + "epoch": 1.4114912219832072, + "grad_norm": 1.980275273323059, + "learning_rate": 0.0001501313207361362, + "loss": 0.656, + "step": 20341 + }, + { + "epoch": 1.411560613420304, + "grad_norm": 1.7851508855819702, + "learning_rate": 0.00015012645906066954, + "loss": 0.7849, + "step": 20342 + }, + { + "epoch": 1.4116300048574006, + "grad_norm": 1.7000179290771484, + "learning_rate": 0.00015012159722695948, + "loss": 0.5124, + "step": 20343 + }, + { + "epoch": 1.4116993962944973, + "grad_norm": 2.8212873935699463, + "learning_rate": 0.00015011673523502135, + "loss": 0.8037, + "step": 20344 + }, + { + "epoch": 1.4117687877315939, + "grad_norm": 1.5520814657211304, + "learning_rate": 0.0001501118730848705, + "loss": 0.5899, + "step": 20345 + }, + { + "epoch": 1.4118381791686905, + "grad_norm": 1.6851588487625122, + "learning_rate": 0.00015010701077652224, + "loss": 0.7653, + "step": 20346 + }, + { + "epoch": 1.4119075706057873, + "grad_norm": 1.4812631607055664, + "learning_rate": 0.00015010214830999202, + "loss": 0.5994, + "step": 20347 + }, + { + "epoch": 1.411976962042884, + "grad_norm": 1.976517677307129, + "learning_rate": 0.00015009728568529508, + "loss": 0.6587, + "step": 20348 + }, + { + "epoch": 1.4120463534799805, + "grad_norm": 1.664754033088684, + "learning_rate": 0.00015009242290244687, + "loss": 0.7022, + "step": 20349 + }, + { + "epoch": 1.4121157449170774, + "grad_norm": 1.5661051273345947, + "learning_rate": 0.00015008755996146267, + "loss": 0.6017, + "step": 20350 + }, + { + "epoch": 1.412185136354174, + "grad_norm": 1.4879133701324463, + "learning_rate": 0.00015008269686235783, + "loss": 0.5581, + "step": 20351 + }, + { + "epoch": 1.4122545277912706, + "grad_norm": 1.6583584547042847, + "learning_rate": 0.00015007783360514775, + "loss": 0.678, + "step": 20352 + }, + { + "epoch": 1.4123239192283672, + "grad_norm": 1.3106896877288818, + "learning_rate": 0.00015007297018984777, + "loss": 0.6161, + "step": 20353 + }, + { + "epoch": 1.4123933106654638, + "grad_norm": 2.176028251647949, + "learning_rate": 0.00015006810661647322, + "loss": 0.7846, + "step": 20354 + }, + { + "epoch": 1.4124627021025606, + "grad_norm": 1.8181401491165161, + "learning_rate": 0.00015006324288503948, + "loss": 0.6007, + "step": 20355 + }, + { + "epoch": 1.4125320935396573, + "grad_norm": 2.052394390106201, + "learning_rate": 0.00015005837899556185, + "loss": 0.7234, + "step": 20356 + }, + { + "epoch": 1.4126014849767539, + "grad_norm": 1.5226004123687744, + "learning_rate": 0.00015005351494805575, + "loss": 0.7896, + "step": 20357 + }, + { + "epoch": 1.4126708764138505, + "grad_norm": 2.2619876861572266, + "learning_rate": 0.00015004865074253653, + "loss": 0.8892, + "step": 20358 + }, + { + "epoch": 1.412740267850947, + "grad_norm": 2.0084445476531982, + "learning_rate": 0.00015004378637901953, + "loss": 0.7823, + "step": 20359 + }, + { + "epoch": 1.412809659288044, + "grad_norm": 1.9176908731460571, + "learning_rate": 0.00015003892185752008, + "loss": 0.6709, + "step": 20360 + }, + { + "epoch": 1.4128790507251405, + "grad_norm": 1.3755470514297485, + "learning_rate": 0.00015003405717805356, + "loss": 0.5474, + "step": 20361 + }, + { + "epoch": 1.4129484421622371, + "grad_norm": 1.9025076627731323, + "learning_rate": 0.00015002919234063532, + "loss": 0.6357, + "step": 20362 + }, + { + "epoch": 1.413017833599334, + "grad_norm": 1.526358723640442, + "learning_rate": 0.00015002432734528072, + "loss": 0.5414, + "step": 20363 + }, + { + "epoch": 1.4130872250364306, + "grad_norm": 1.5297828912734985, + "learning_rate": 0.00015001946219200514, + "loss": 0.6534, + "step": 20364 + }, + { + "epoch": 1.4131566164735272, + "grad_norm": 1.8403171300888062, + "learning_rate": 0.00015001459688082395, + "loss": 0.7554, + "step": 20365 + }, + { + "epoch": 1.4132260079106238, + "grad_norm": 1.7757048606872559, + "learning_rate": 0.00015000973141175242, + "loss": 0.6418, + "step": 20366 + }, + { + "epoch": 1.4132953993477204, + "grad_norm": 9.2096586227417, + "learning_rate": 0.00015000486578480599, + "loss": 0.506, + "step": 20367 + }, + { + "epoch": 1.4133647907848172, + "grad_norm": 1.3506590127944946, + "learning_rate": 0.00015000000000000001, + "loss": 0.5209, + "step": 20368 + }, + { + "epoch": 1.4134341822219139, + "grad_norm": 1.642455816268921, + "learning_rate": 0.00014999513405734982, + "loss": 0.5416, + "step": 20369 + }, + { + "epoch": 1.4135035736590105, + "grad_norm": 1.3027043342590332, + "learning_rate": 0.00014999026795687082, + "loss": 0.4536, + "step": 20370 + }, + { + "epoch": 1.413572965096107, + "grad_norm": 1.5861124992370605, + "learning_rate": 0.0001499854016985783, + "loss": 0.5815, + "step": 20371 + }, + { + "epoch": 1.4136423565332037, + "grad_norm": 1.7404762506484985, + "learning_rate": 0.0001499805352824877, + "loss": 0.7487, + "step": 20372 + }, + { + "epoch": 1.4137117479703005, + "grad_norm": 1.2529809474945068, + "learning_rate": 0.00014997566870861433, + "loss": 0.4817, + "step": 20373 + }, + { + "epoch": 1.4137811394073971, + "grad_norm": 2.00626540184021, + "learning_rate": 0.00014997080197697357, + "loss": 0.6716, + "step": 20374 + }, + { + "epoch": 1.4138505308444937, + "grad_norm": 1.7646929025650024, + "learning_rate": 0.00014996593508758083, + "loss": 0.6113, + "step": 20375 + }, + { + "epoch": 1.4139199222815906, + "grad_norm": 1.465341329574585, + "learning_rate": 0.0001499610680404514, + "loss": 0.6691, + "step": 20376 + }, + { + "epoch": 1.4139893137186872, + "grad_norm": 2.2194104194641113, + "learning_rate": 0.00014995620083560067, + "loss": 0.7272, + "step": 20377 + }, + { + "epoch": 1.4140587051557838, + "grad_norm": 2.576822519302368, + "learning_rate": 0.000149951333473044, + "loss": 0.752, + "step": 20378 + }, + { + "epoch": 1.4141280965928804, + "grad_norm": 1.5861412286758423, + "learning_rate": 0.0001499464659527968, + "loss": 0.7405, + "step": 20379 + }, + { + "epoch": 1.414197488029977, + "grad_norm": 1.6651026010513306, + "learning_rate": 0.00014994159827487436, + "loss": 0.5986, + "step": 20380 + }, + { + "epoch": 1.4142668794670739, + "grad_norm": 1.8040506839752197, + "learning_rate": 0.0001499367304392921, + "loss": 0.7137, + "step": 20381 + }, + { + "epoch": 1.4143362709041705, + "grad_norm": 1.8156414031982422, + "learning_rate": 0.0001499318624460654, + "loss": 0.6789, + "step": 20382 + }, + { + "epoch": 1.414405662341267, + "grad_norm": 1.7936389446258545, + "learning_rate": 0.0001499269942952096, + "loss": 0.7511, + "step": 20383 + }, + { + "epoch": 1.4144750537783637, + "grad_norm": 1.7702103853225708, + "learning_rate": 0.00014992212598674003, + "loss": 0.4884, + "step": 20384 + }, + { + "epoch": 1.4145444452154603, + "grad_norm": 1.9677457809448242, + "learning_rate": 0.00014991725752067217, + "loss": 0.7618, + "step": 20385 + }, + { + "epoch": 1.4146138366525571, + "grad_norm": 1.532073974609375, + "learning_rate": 0.00014991238889702126, + "loss": 0.5198, + "step": 20386 + }, + { + "epoch": 1.4146832280896537, + "grad_norm": 2.155799150466919, + "learning_rate": 0.00014990752011580274, + "loss": 0.8955, + "step": 20387 + }, + { + "epoch": 1.4147526195267504, + "grad_norm": 1.7223149538040161, + "learning_rate": 0.00014990265117703196, + "loss": 0.5033, + "step": 20388 + }, + { + "epoch": 1.4148220109638472, + "grad_norm": 1.6438134908676147, + "learning_rate": 0.00014989778208072434, + "loss": 0.5939, + "step": 20389 + }, + { + "epoch": 1.4148914024009438, + "grad_norm": 1.505699872970581, + "learning_rate": 0.0001498929128268952, + "loss": 0.5908, + "step": 20390 + }, + { + "epoch": 1.4149607938380404, + "grad_norm": 1.7339017391204834, + "learning_rate": 0.0001498880434155599, + "loss": 0.5938, + "step": 20391 + }, + { + "epoch": 1.415030185275137, + "grad_norm": 1.9827091693878174, + "learning_rate": 0.00014988317384673384, + "loss": 0.5343, + "step": 20392 + }, + { + "epoch": 1.4150995767122336, + "grad_norm": 1.674938678741455, + "learning_rate": 0.0001498783041204324, + "loss": 0.6016, + "step": 20393 + }, + { + "epoch": 1.4151689681493305, + "grad_norm": 1.6272310018539429, + "learning_rate": 0.00014987343423667095, + "loss": 0.6201, + "step": 20394 + }, + { + "epoch": 1.415238359586427, + "grad_norm": 1.6833269596099854, + "learning_rate": 0.0001498685641954648, + "loss": 0.6546, + "step": 20395 + }, + { + "epoch": 1.4153077510235237, + "grad_norm": 1.2504475116729736, + "learning_rate": 0.00014986369399682946, + "loss": 0.3703, + "step": 20396 + }, + { + "epoch": 1.4153771424606203, + "grad_norm": 1.4009844064712524, + "learning_rate": 0.00014985882364078017, + "loss": 0.5793, + "step": 20397 + }, + { + "epoch": 1.415446533897717, + "grad_norm": 2.2345027923583984, + "learning_rate": 0.00014985395312733238, + "loss": 0.9449, + "step": 20398 + }, + { + "epoch": 1.4155159253348137, + "grad_norm": 1.8719550371170044, + "learning_rate": 0.00014984908245650145, + "loss": 0.7182, + "step": 20399 + }, + { + "epoch": 1.4155853167719104, + "grad_norm": 1.6583490371704102, + "learning_rate": 0.00014984421162830272, + "loss": 0.6244, + "step": 20400 + }, + { + "epoch": 1.415654708209007, + "grad_norm": 1.512515902519226, + "learning_rate": 0.0001498393406427516, + "loss": 0.4515, + "step": 20401 + }, + { + "epoch": 1.4157240996461038, + "grad_norm": 2.3209550380706787, + "learning_rate": 0.0001498344694998635, + "loss": 0.8369, + "step": 20402 + }, + { + "epoch": 1.4157934910832004, + "grad_norm": 1.8666775226593018, + "learning_rate": 0.00014982959819965372, + "loss": 0.6814, + "step": 20403 + }, + { + "epoch": 1.415862882520297, + "grad_norm": 2.208193302154541, + "learning_rate": 0.00014982472674213767, + "loss": 0.7434, + "step": 20404 + }, + { + "epoch": 1.4159322739573936, + "grad_norm": 1.9503370523452759, + "learning_rate": 0.00014981985512733077, + "loss": 0.8365, + "step": 20405 + }, + { + "epoch": 1.4160016653944902, + "grad_norm": 1.879099726676941, + "learning_rate": 0.00014981498335524837, + "loss": 0.6196, + "step": 20406 + }, + { + "epoch": 1.416071056831587, + "grad_norm": 1.6988649368286133, + "learning_rate": 0.00014981011142590583, + "loss": 0.6506, + "step": 20407 + }, + { + "epoch": 1.4161404482686837, + "grad_norm": 1.923526644706726, + "learning_rate": 0.00014980523933931855, + "loss": 0.6758, + "step": 20408 + }, + { + "epoch": 1.4162098397057803, + "grad_norm": 2.9926533699035645, + "learning_rate": 0.00014980036709550193, + "loss": 0.9117, + "step": 20409 + }, + { + "epoch": 1.416279231142877, + "grad_norm": 2.2349281311035156, + "learning_rate": 0.00014979549469447133, + "loss": 0.8388, + "step": 20410 + }, + { + "epoch": 1.4163486225799735, + "grad_norm": 1.999472737312317, + "learning_rate": 0.0001497906221362421, + "loss": 0.717, + "step": 20411 + }, + { + "epoch": 1.4164180140170703, + "grad_norm": 2.431652307510376, + "learning_rate": 0.0001497857494208297, + "loss": 0.8737, + "step": 20412 + }, + { + "epoch": 1.416487405454167, + "grad_norm": 1.7952560186386108, + "learning_rate": 0.0001497808765482494, + "loss": 0.6164, + "step": 20413 + }, + { + "epoch": 1.4165567968912636, + "grad_norm": 1.7489678859710693, + "learning_rate": 0.00014977600351851668, + "loss": 0.6556, + "step": 20414 + }, + { + "epoch": 1.4166261883283604, + "grad_norm": 2.4441754817962646, + "learning_rate": 0.0001497711303316469, + "loss": 0.6225, + "step": 20415 + }, + { + "epoch": 1.416695579765457, + "grad_norm": 1.7246211767196655, + "learning_rate": 0.00014976625698765546, + "loss": 0.463, + "step": 20416 + }, + { + "epoch": 1.4167649712025536, + "grad_norm": 2.1352968215942383, + "learning_rate": 0.00014976138348655766, + "loss": 0.883, + "step": 20417 + }, + { + "epoch": 1.4168343626396502, + "grad_norm": 1.4694275856018066, + "learning_rate": 0.00014975650982836897, + "loss": 0.6053, + "step": 20418 + }, + { + "epoch": 1.4169037540767468, + "grad_norm": 1.3916444778442383, + "learning_rate": 0.00014975163601310475, + "loss": 0.6164, + "step": 20419 + }, + { + "epoch": 1.4169731455138437, + "grad_norm": 1.8621975183486938, + "learning_rate": 0.0001497467620407804, + "loss": 0.6027, + "step": 20420 + }, + { + "epoch": 1.4170425369509403, + "grad_norm": 1.6911084651947021, + "learning_rate": 0.00014974188791141128, + "loss": 0.6768, + "step": 20421 + }, + { + "epoch": 1.417111928388037, + "grad_norm": 2.2947113513946533, + "learning_rate": 0.0001497370136250128, + "loss": 0.5296, + "step": 20422 + }, + { + "epoch": 1.4171813198251335, + "grad_norm": 1.517127275466919, + "learning_rate": 0.00014973213918160035, + "loss": 0.5283, + "step": 20423 + }, + { + "epoch": 1.4172507112622301, + "grad_norm": 1.2505793571472168, + "learning_rate": 0.0001497272645811893, + "loss": 0.4773, + "step": 20424 + }, + { + "epoch": 1.417320102699327, + "grad_norm": 1.842326283454895, + "learning_rate": 0.00014972238982379502, + "loss": 0.721, + "step": 20425 + }, + { + "epoch": 1.4173894941364236, + "grad_norm": 2.1406021118164062, + "learning_rate": 0.00014971751490943294, + "loss": 0.7276, + "step": 20426 + }, + { + "epoch": 1.4174588855735202, + "grad_norm": 1.3560923337936401, + "learning_rate": 0.00014971263983811845, + "loss": 0.3896, + "step": 20427 + }, + { + "epoch": 1.417528277010617, + "grad_norm": 1.559923529624939, + "learning_rate": 0.0001497077646098669, + "loss": 0.5105, + "step": 20428 + }, + { + "epoch": 1.4175976684477136, + "grad_norm": 1.918311595916748, + "learning_rate": 0.00014970288922469373, + "loss": 0.661, + "step": 20429 + }, + { + "epoch": 1.4176670598848102, + "grad_norm": 2.0853347778320312, + "learning_rate": 0.0001496980136826143, + "loss": 0.5384, + "step": 20430 + }, + { + "epoch": 1.4177364513219068, + "grad_norm": 1.312351942062378, + "learning_rate": 0.000149693137983644, + "loss": 0.4908, + "step": 20431 + }, + { + "epoch": 1.4178058427590035, + "grad_norm": 2.4280407428741455, + "learning_rate": 0.00014968826212779823, + "loss": 0.7408, + "step": 20432 + }, + { + "epoch": 1.4178752341961003, + "grad_norm": 2.0816125869750977, + "learning_rate": 0.00014968338611509236, + "loss": 0.6216, + "step": 20433 + }, + { + "epoch": 1.417944625633197, + "grad_norm": 1.5753941535949707, + "learning_rate": 0.00014967850994554183, + "loss": 0.5597, + "step": 20434 + }, + { + "epoch": 1.4180140170702935, + "grad_norm": 2.400649309158325, + "learning_rate": 0.000149673633619162, + "loss": 0.966, + "step": 20435 + }, + { + "epoch": 1.4180834085073901, + "grad_norm": 1.8584544658660889, + "learning_rate": 0.00014966875713596825, + "loss": 0.5328, + "step": 20436 + }, + { + "epoch": 1.4181527999444867, + "grad_norm": 1.593842625617981, + "learning_rate": 0.00014966388049597606, + "loss": 0.8049, + "step": 20437 + }, + { + "epoch": 1.4182221913815836, + "grad_norm": 2.257460832595825, + "learning_rate": 0.0001496590036992007, + "loss": 0.7327, + "step": 20438 + }, + { + "epoch": 1.4182915828186802, + "grad_norm": 2.267150402069092, + "learning_rate": 0.00014965412674565763, + "loss": 0.6489, + "step": 20439 + }, + { + "epoch": 1.4183609742557768, + "grad_norm": 1.505815863609314, + "learning_rate": 0.00014964924963536228, + "loss": 0.6571, + "step": 20440 + }, + { + "epoch": 1.4184303656928736, + "grad_norm": 1.8605694770812988, + "learning_rate": 0.00014964437236832998, + "loss": 0.6777, + "step": 20441 + }, + { + "epoch": 1.4184997571299702, + "grad_norm": 1.837669014930725, + "learning_rate": 0.00014963949494457618, + "loss": 0.7165, + "step": 20442 + }, + { + "epoch": 1.4185691485670668, + "grad_norm": 1.8408722877502441, + "learning_rate": 0.0001496346173641162, + "loss": 0.7657, + "step": 20443 + }, + { + "epoch": 1.4186385400041635, + "grad_norm": 1.8133666515350342, + "learning_rate": 0.00014962973962696555, + "loss": 0.5922, + "step": 20444 + }, + { + "epoch": 1.41870793144126, + "grad_norm": 1.3497234582901, + "learning_rate": 0.00014962486173313955, + "loss": 0.6023, + "step": 20445 + }, + { + "epoch": 1.418777322878357, + "grad_norm": 1.6148122549057007, + "learning_rate": 0.00014961998368265362, + "loss": 0.6412, + "step": 20446 + }, + { + "epoch": 1.4188467143154535, + "grad_norm": 2.100649356842041, + "learning_rate": 0.00014961510547552317, + "loss": 0.9386, + "step": 20447 + }, + { + "epoch": 1.4189161057525501, + "grad_norm": 1.143001914024353, + "learning_rate": 0.00014961022711176359, + "loss": 0.399, + "step": 20448 + }, + { + "epoch": 1.4189854971896467, + "grad_norm": 2.1849467754364014, + "learning_rate": 0.00014960534859139024, + "loss": 0.7027, + "step": 20449 + }, + { + "epoch": 1.4190548886267433, + "grad_norm": 2.0887720584869385, + "learning_rate": 0.00014960046991441858, + "loss": 0.5333, + "step": 20450 + }, + { + "epoch": 1.4191242800638402, + "grad_norm": 1.7781271934509277, + "learning_rate": 0.000149595591080864, + "loss": 0.6916, + "step": 20451 + }, + { + "epoch": 1.4191936715009368, + "grad_norm": 2.0120365619659424, + "learning_rate": 0.0001495907120907419, + "loss": 0.8036, + "step": 20452 + }, + { + "epoch": 1.4192630629380334, + "grad_norm": 1.6387178897857666, + "learning_rate": 0.00014958583294406768, + "loss": 0.7761, + "step": 20453 + }, + { + "epoch": 1.4193324543751302, + "grad_norm": 1.5514743328094482, + "learning_rate": 0.00014958095364085673, + "loss": 0.6039, + "step": 20454 + }, + { + "epoch": 1.4194018458122268, + "grad_norm": 2.474914312362671, + "learning_rate": 0.00014957607418112444, + "loss": 0.5152, + "step": 20455 + }, + { + "epoch": 1.4194712372493234, + "grad_norm": 1.6659135818481445, + "learning_rate": 0.00014957119456488624, + "loss": 0.5901, + "step": 20456 + }, + { + "epoch": 1.41954062868642, + "grad_norm": 2.1669039726257324, + "learning_rate": 0.00014956631479215752, + "loss": 0.6361, + "step": 20457 + }, + { + "epoch": 1.4196100201235167, + "grad_norm": 1.5423280000686646, + "learning_rate": 0.00014956143486295374, + "loss": 0.5917, + "step": 20458 + }, + { + "epoch": 1.4196794115606135, + "grad_norm": 1.6988040208816528, + "learning_rate": 0.00014955655477729023, + "loss": 0.7951, + "step": 20459 + }, + { + "epoch": 1.4197488029977101, + "grad_norm": 1.962564468383789, + "learning_rate": 0.00014955167453518242, + "loss": 0.694, + "step": 20460 + }, + { + "epoch": 1.4198181944348067, + "grad_norm": 2.0968849658966064, + "learning_rate": 0.00014954679413664573, + "loss": 0.7482, + "step": 20461 + }, + { + "epoch": 1.4198875858719033, + "grad_norm": 1.7587199211120605, + "learning_rate": 0.0001495419135816956, + "loss": 0.6933, + "step": 20462 + }, + { + "epoch": 1.419956977309, + "grad_norm": 1.7598820924758911, + "learning_rate": 0.00014953703287034733, + "loss": 0.6567, + "step": 20463 + }, + { + "epoch": 1.4200263687460968, + "grad_norm": 1.641195297241211, + "learning_rate": 0.00014953215200261646, + "loss": 0.3514, + "step": 20464 + }, + { + "epoch": 1.4200957601831934, + "grad_norm": 1.4198616743087769, + "learning_rate": 0.00014952727097851826, + "loss": 0.4935, + "step": 20465 + }, + { + "epoch": 1.42016515162029, + "grad_norm": 1.4953110218048096, + "learning_rate": 0.00014952238979806828, + "loss": 0.4211, + "step": 20466 + }, + { + "epoch": 1.4202345430573868, + "grad_norm": 2.4273595809936523, + "learning_rate": 0.00014951750846128184, + "loss": 0.6997, + "step": 20467 + }, + { + "epoch": 1.4203039344944834, + "grad_norm": 1.0732682943344116, + "learning_rate": 0.00014951262696817437, + "loss": 0.3706, + "step": 20468 + }, + { + "epoch": 1.42037332593158, + "grad_norm": 2.557706594467163, + "learning_rate": 0.00014950774531876125, + "loss": 0.6329, + "step": 20469 + }, + { + "epoch": 1.4204427173686767, + "grad_norm": 1.4442521333694458, + "learning_rate": 0.00014950286351305794, + "loss": 0.3876, + "step": 20470 + }, + { + "epoch": 1.4205121088057733, + "grad_norm": 1.617301344871521, + "learning_rate": 0.00014949798155107984, + "loss": 0.6586, + "step": 20471 + }, + { + "epoch": 1.42058150024287, + "grad_norm": 1.2914748191833496, + "learning_rate": 0.00014949309943284236, + "loss": 0.5137, + "step": 20472 + }, + { + "epoch": 1.4206508916799667, + "grad_norm": 1.9267196655273438, + "learning_rate": 0.0001494882171583609, + "loss": 0.5673, + "step": 20473 + }, + { + "epoch": 1.4207202831170633, + "grad_norm": 1.3756183385849, + "learning_rate": 0.00014948333472765088, + "loss": 0.5438, + "step": 20474 + }, + { + "epoch": 1.42078967455416, + "grad_norm": 1.5073928833007812, + "learning_rate": 0.0001494784521407277, + "loss": 0.6041, + "step": 20475 + }, + { + "epoch": 1.4208590659912566, + "grad_norm": 1.7691365480422974, + "learning_rate": 0.00014947356939760682, + "loss": 0.5706, + "step": 20476 + }, + { + "epoch": 1.4209284574283534, + "grad_norm": 1.9156705141067505, + "learning_rate": 0.00014946868649830358, + "loss": 0.838, + "step": 20477 + }, + { + "epoch": 1.42099784886545, + "grad_norm": 2.215014934539795, + "learning_rate": 0.0001494638034428335, + "loss": 0.7348, + "step": 20478 + }, + { + "epoch": 1.4210672403025466, + "grad_norm": 1.8464592695236206, + "learning_rate": 0.00014945892023121186, + "loss": 0.6099, + "step": 20479 + }, + { + "epoch": 1.4211366317396434, + "grad_norm": 2.1292545795440674, + "learning_rate": 0.0001494540368634542, + "loss": 0.8634, + "step": 20480 + }, + { + "epoch": 1.42120602317674, + "grad_norm": 1.7533445358276367, + "learning_rate": 0.00014944915333957586, + "loss": 0.7843, + "step": 20481 + }, + { + "epoch": 1.4212754146138367, + "grad_norm": 1.442874789237976, + "learning_rate": 0.00014944426965959225, + "loss": 0.5346, + "step": 20482 + }, + { + "epoch": 1.4213448060509333, + "grad_norm": 1.8744878768920898, + "learning_rate": 0.00014943938582351888, + "loss": 0.7275, + "step": 20483 + }, + { + "epoch": 1.4214141974880299, + "grad_norm": 1.435461163520813, + "learning_rate": 0.00014943450183137107, + "loss": 0.533, + "step": 20484 + }, + { + "epoch": 1.4214835889251267, + "grad_norm": 1.9679820537567139, + "learning_rate": 0.00014942961768316427, + "loss": 0.6976, + "step": 20485 + }, + { + "epoch": 1.4215529803622233, + "grad_norm": 1.6571027040481567, + "learning_rate": 0.0001494247333789139, + "loss": 0.5225, + "step": 20486 + }, + { + "epoch": 1.42162237179932, + "grad_norm": 2.0219507217407227, + "learning_rate": 0.00014941984891863542, + "loss": 0.945, + "step": 20487 + }, + { + "epoch": 1.4216917632364166, + "grad_norm": 1.7399251461029053, + "learning_rate": 0.00014941496430234414, + "loss": 0.6012, + "step": 20488 + }, + { + "epoch": 1.4217611546735132, + "grad_norm": 1.6415457725524902, + "learning_rate": 0.0001494100795300556, + "loss": 0.4297, + "step": 20489 + }, + { + "epoch": 1.42183054611061, + "grad_norm": 2.012669086456299, + "learning_rate": 0.00014940519460178518, + "loss": 0.8236, + "step": 20490 + }, + { + "epoch": 1.4218999375477066, + "grad_norm": 1.8627948760986328, + "learning_rate": 0.00014940030951754826, + "loss": 0.7629, + "step": 20491 + }, + { + "epoch": 1.4219693289848032, + "grad_norm": 2.9821908473968506, + "learning_rate": 0.00014939542427736032, + "loss": 0.5926, + "step": 20492 + }, + { + "epoch": 1.4220387204219, + "grad_norm": 1.8320677280426025, + "learning_rate": 0.00014939053888123673, + "loss": 0.6097, + "step": 20493 + }, + { + "epoch": 1.4221081118589967, + "grad_norm": 1.7090364694595337, + "learning_rate": 0.00014938565332919297, + "loss": 0.6618, + "step": 20494 + }, + { + "epoch": 1.4221775032960933, + "grad_norm": 1.812712550163269, + "learning_rate": 0.0001493807676212444, + "loss": 0.5213, + "step": 20495 + }, + { + "epoch": 1.4222468947331899, + "grad_norm": 1.914439082145691, + "learning_rate": 0.00014937588175740652, + "loss": 0.6305, + "step": 20496 + }, + { + "epoch": 1.4223162861702865, + "grad_norm": 1.7915583848953247, + "learning_rate": 0.00014937099573769465, + "loss": 0.6351, + "step": 20497 + }, + { + "epoch": 1.4223856776073833, + "grad_norm": 2.597134828567505, + "learning_rate": 0.0001493661095621243, + "loss": 0.5362, + "step": 20498 + }, + { + "epoch": 1.42245506904448, + "grad_norm": 2.164350986480713, + "learning_rate": 0.0001493612232307109, + "loss": 0.6235, + "step": 20499 + }, + { + "epoch": 1.4225244604815765, + "grad_norm": 1.4510186910629272, + "learning_rate": 0.00014935633674346982, + "loss": 0.7259, + "step": 20500 + }, + { + "epoch": 1.4225938519186732, + "grad_norm": 1.569140911102295, + "learning_rate": 0.0001493514501004165, + "loss": 0.6737, + "step": 20501 + }, + { + "epoch": 1.4226632433557698, + "grad_norm": 1.9205247163772583, + "learning_rate": 0.0001493465633015664, + "loss": 0.7258, + "step": 20502 + }, + { + "epoch": 1.4227326347928666, + "grad_norm": 1.5491842031478882, + "learning_rate": 0.00014934167634693494, + "loss": 0.7686, + "step": 20503 + }, + { + "epoch": 1.4228020262299632, + "grad_norm": 1.8001788854599, + "learning_rate": 0.0001493367892365375, + "loss": 0.8636, + "step": 20504 + }, + { + "epoch": 1.4228714176670598, + "grad_norm": 1.7829484939575195, + "learning_rate": 0.00014933190197038954, + "loss": 0.6644, + "step": 20505 + }, + { + "epoch": 1.4229408091041567, + "grad_norm": 1.3639426231384277, + "learning_rate": 0.00014932701454850654, + "loss": 0.4965, + "step": 20506 + }, + { + "epoch": 1.4230102005412533, + "grad_norm": 1.70647132396698, + "learning_rate": 0.0001493221269709038, + "loss": 0.7068, + "step": 20507 + }, + { + "epoch": 1.4230795919783499, + "grad_norm": 1.6233124732971191, + "learning_rate": 0.00014931723923759685, + "loss": 0.5942, + "step": 20508 + }, + { + "epoch": 1.4231489834154465, + "grad_norm": 1.7303630113601685, + "learning_rate": 0.00014931235134860114, + "loss": 0.5026, + "step": 20509 + }, + { + "epoch": 1.423218374852543, + "grad_norm": 2.0492570400238037, + "learning_rate": 0.00014930746330393203, + "loss": 0.7503, + "step": 20510 + }, + { + "epoch": 1.42328776628964, + "grad_norm": 2.55942964553833, + "learning_rate": 0.00014930257510360496, + "loss": 0.8094, + "step": 20511 + }, + { + "epoch": 1.4233571577267365, + "grad_norm": 1.8121181726455688, + "learning_rate": 0.0001492976867476354, + "loss": 0.7747, + "step": 20512 + }, + { + "epoch": 1.4234265491638332, + "grad_norm": 1.6313947439193726, + "learning_rate": 0.00014929279823603877, + "loss": 0.7613, + "step": 20513 + }, + { + "epoch": 1.4234959406009298, + "grad_norm": 1.6988877058029175, + "learning_rate": 0.00014928790956883047, + "loss": 0.5959, + "step": 20514 + }, + { + "epoch": 1.4235653320380264, + "grad_norm": 2.185222864151001, + "learning_rate": 0.000149283020746026, + "loss": 0.8366, + "step": 20515 + }, + { + "epoch": 1.4236347234751232, + "grad_norm": 1.7712041139602661, + "learning_rate": 0.00014927813176764075, + "loss": 0.8242, + "step": 20516 + }, + { + "epoch": 1.4237041149122198, + "grad_norm": 1.9471830129623413, + "learning_rate": 0.0001492732426336901, + "loss": 0.6589, + "step": 20517 + }, + { + "epoch": 1.4237735063493164, + "grad_norm": 1.2224098443984985, + "learning_rate": 0.00014926835334418958, + "loss": 0.4958, + "step": 20518 + }, + { + "epoch": 1.4238428977864133, + "grad_norm": 2.406777858734131, + "learning_rate": 0.0001492634638991546, + "loss": 0.845, + "step": 20519 + }, + { + "epoch": 1.4239122892235099, + "grad_norm": 1.9999346733093262, + "learning_rate": 0.00014925857429860055, + "loss": 0.6524, + "step": 20520 + }, + { + "epoch": 1.4239816806606065, + "grad_norm": 1.4046427011489868, + "learning_rate": 0.0001492536845425429, + "loss": 0.4608, + "step": 20521 + }, + { + "epoch": 1.424051072097703, + "grad_norm": 1.702226161956787, + "learning_rate": 0.00014924879463099706, + "loss": 0.5343, + "step": 20522 + }, + { + "epoch": 1.4241204635347997, + "grad_norm": 1.8322491645812988, + "learning_rate": 0.00014924390456397854, + "loss": 0.4773, + "step": 20523 + }, + { + "epoch": 1.4241898549718965, + "grad_norm": 1.7804559469223022, + "learning_rate": 0.0001492390143415027, + "loss": 0.6176, + "step": 20524 + }, + { + "epoch": 1.4242592464089932, + "grad_norm": 1.4140591621398926, + "learning_rate": 0.000149234123963585, + "loss": 0.5682, + "step": 20525 + }, + { + "epoch": 1.4243286378460898, + "grad_norm": 2.3540568351745605, + "learning_rate": 0.00014922923343024087, + "loss": 0.5337, + "step": 20526 + }, + { + "epoch": 1.4243980292831864, + "grad_norm": 2.474544048309326, + "learning_rate": 0.0001492243427414858, + "loss": 0.7317, + "step": 20527 + }, + { + "epoch": 1.424467420720283, + "grad_norm": 1.9113160371780396, + "learning_rate": 0.00014921945189733514, + "loss": 0.6528, + "step": 20528 + }, + { + "epoch": 1.4245368121573798, + "grad_norm": 2.5526206493377686, + "learning_rate": 0.00014921456089780444, + "loss": 0.8288, + "step": 20529 + }, + { + "epoch": 1.4246062035944764, + "grad_norm": 1.577906847000122, + "learning_rate": 0.00014920966974290905, + "loss": 0.4969, + "step": 20530 + }, + { + "epoch": 1.424675595031573, + "grad_norm": 2.255030393600464, + "learning_rate": 0.00014920477843266444, + "loss": 0.6339, + "step": 20531 + }, + { + "epoch": 1.4247449864686699, + "grad_norm": 2.002122640609741, + "learning_rate": 0.00014919988696708605, + "loss": 0.6339, + "step": 20532 + }, + { + "epoch": 1.4248143779057665, + "grad_norm": 1.713012456893921, + "learning_rate": 0.00014919499534618932, + "loss": 0.7184, + "step": 20533 + }, + { + "epoch": 1.424883769342863, + "grad_norm": 2.2127718925476074, + "learning_rate": 0.00014919010356998972, + "loss": 0.6606, + "step": 20534 + }, + { + "epoch": 1.4249531607799597, + "grad_norm": 1.5433781147003174, + "learning_rate": 0.00014918521163850268, + "loss": 0.698, + "step": 20535 + }, + { + "epoch": 1.4250225522170563, + "grad_norm": 3.045503854751587, + "learning_rate": 0.0001491803195517436, + "loss": 0.5934, + "step": 20536 + }, + { + "epoch": 1.4250919436541531, + "grad_norm": 1.788246989250183, + "learning_rate": 0.000149175427309728, + "loss": 0.5188, + "step": 20537 + }, + { + "epoch": 1.4251613350912498, + "grad_norm": 2.7538082599639893, + "learning_rate": 0.00014917053491247123, + "loss": 0.7047, + "step": 20538 + }, + { + "epoch": 1.4252307265283464, + "grad_norm": 1.311635136604309, + "learning_rate": 0.0001491656423599888, + "loss": 0.7573, + "step": 20539 + }, + { + "epoch": 1.4253001179654432, + "grad_norm": 1.612791895866394, + "learning_rate": 0.00014916074965229617, + "loss": 0.5435, + "step": 20540 + }, + { + "epoch": 1.4253695094025396, + "grad_norm": 1.7806286811828613, + "learning_rate": 0.00014915585678940873, + "loss": 0.4679, + "step": 20541 + }, + { + "epoch": 1.4254389008396364, + "grad_norm": 1.8833481073379517, + "learning_rate": 0.00014915096377134197, + "loss": 0.7384, + "step": 20542 + }, + { + "epoch": 1.425508292276733, + "grad_norm": 1.653560996055603, + "learning_rate": 0.0001491460705981113, + "loss": 0.6922, + "step": 20543 + }, + { + "epoch": 1.4255776837138296, + "grad_norm": 2.065459966659546, + "learning_rate": 0.0001491411772697322, + "loss": 0.785, + "step": 20544 + }, + { + "epoch": 1.4256470751509265, + "grad_norm": 1.9640611410140991, + "learning_rate": 0.0001491362837862201, + "loss": 0.5104, + "step": 20545 + }, + { + "epoch": 1.425716466588023, + "grad_norm": 1.4722903966903687, + "learning_rate": 0.00014913139014759044, + "loss": 0.4691, + "step": 20546 + }, + { + "epoch": 1.4257858580251197, + "grad_norm": 2.964365243911743, + "learning_rate": 0.0001491264963538587, + "loss": 0.7548, + "step": 20547 + }, + { + "epoch": 1.4258552494622163, + "grad_norm": 1.655368685722351, + "learning_rate": 0.00014912160240504032, + "loss": 0.5637, + "step": 20548 + }, + { + "epoch": 1.425924640899313, + "grad_norm": 2.041095733642578, + "learning_rate": 0.00014911670830115072, + "loss": 0.5355, + "step": 20549 + }, + { + "epoch": 1.4259940323364098, + "grad_norm": 1.6658340692520142, + "learning_rate": 0.0001491118140422054, + "loss": 0.5901, + "step": 20550 + }, + { + "epoch": 1.4260634237735064, + "grad_norm": 1.3665087223052979, + "learning_rate": 0.00014910691962821974, + "loss": 0.5588, + "step": 20551 + }, + { + "epoch": 1.426132815210603, + "grad_norm": 1.5976780652999878, + "learning_rate": 0.00014910202505920926, + "loss": 0.6061, + "step": 20552 + }, + { + "epoch": 1.4262022066476998, + "grad_norm": 2.1933045387268066, + "learning_rate": 0.00014909713033518936, + "loss": 0.5512, + "step": 20553 + }, + { + "epoch": 1.4262715980847962, + "grad_norm": 1.6950474977493286, + "learning_rate": 0.00014909223545617555, + "loss": 0.7618, + "step": 20554 + }, + { + "epoch": 1.426340989521893, + "grad_norm": 1.8022276163101196, + "learning_rate": 0.00014908734042218323, + "loss": 0.6777, + "step": 20555 + }, + { + "epoch": 1.4264103809589896, + "grad_norm": 1.4977060556411743, + "learning_rate": 0.00014908244523322787, + "loss": 0.5148, + "step": 20556 + }, + { + "epoch": 1.4264797723960863, + "grad_norm": 2.2506866455078125, + "learning_rate": 0.00014907754988932492, + "loss": 0.5541, + "step": 20557 + }, + { + "epoch": 1.426549163833183, + "grad_norm": 1.8253711462020874, + "learning_rate": 0.00014907265439048986, + "loss": 0.5599, + "step": 20558 + }, + { + "epoch": 1.4266185552702797, + "grad_norm": 1.5493996143341064, + "learning_rate": 0.0001490677587367381, + "loss": 0.5588, + "step": 20559 + }, + { + "epoch": 1.4266879467073763, + "grad_norm": 1.525198221206665, + "learning_rate": 0.00014906286292808512, + "loss": 0.5645, + "step": 20560 + }, + { + "epoch": 1.426757338144473, + "grad_norm": 1.7043967247009277, + "learning_rate": 0.0001490579669645464, + "loss": 0.6555, + "step": 20561 + }, + { + "epoch": 1.4268267295815695, + "grad_norm": 2.16796875, + "learning_rate": 0.00014905307084613734, + "loss": 0.8665, + "step": 20562 + }, + { + "epoch": 1.4268961210186664, + "grad_norm": 2.3178775310516357, + "learning_rate": 0.00014904817457287342, + "loss": 0.8695, + "step": 20563 + }, + { + "epoch": 1.426965512455763, + "grad_norm": 2.1994686126708984, + "learning_rate": 0.00014904327814477012, + "loss": 0.6345, + "step": 20564 + }, + { + "epoch": 1.4270349038928596, + "grad_norm": 2.0586960315704346, + "learning_rate": 0.0001490383815618429, + "loss": 0.6743, + "step": 20565 + }, + { + "epoch": 1.4271042953299564, + "grad_norm": 2.827244520187378, + "learning_rate": 0.0001490334848241072, + "loss": 0.9, + "step": 20566 + }, + { + "epoch": 1.4271736867670528, + "grad_norm": 1.46988844871521, + "learning_rate": 0.00014902858793157845, + "loss": 0.6063, + "step": 20567 + }, + { + "epoch": 1.4272430782041496, + "grad_norm": 2.1126949787139893, + "learning_rate": 0.00014902369088427215, + "loss": 0.673, + "step": 20568 + }, + { + "epoch": 1.4273124696412463, + "grad_norm": 1.764966607093811, + "learning_rate": 0.00014901879368220373, + "loss": 0.7987, + "step": 20569 + }, + { + "epoch": 1.4273818610783429, + "grad_norm": 1.839228868484497, + "learning_rate": 0.0001490138963253887, + "loss": 0.7587, + "step": 20570 + }, + { + "epoch": 1.4274512525154397, + "grad_norm": 1.931031584739685, + "learning_rate": 0.00014900899881384246, + "loss": 0.8213, + "step": 20571 + }, + { + "epoch": 1.4275206439525363, + "grad_norm": 1.6935932636260986, + "learning_rate": 0.0001490041011475805, + "loss": 0.6514, + "step": 20572 + }, + { + "epoch": 1.427590035389633, + "grad_norm": 1.784183144569397, + "learning_rate": 0.00014899920332661827, + "loss": 0.7245, + "step": 20573 + }, + { + "epoch": 1.4276594268267295, + "grad_norm": 1.8345805406570435, + "learning_rate": 0.00014899430535097124, + "loss": 0.6531, + "step": 20574 + }, + { + "epoch": 1.4277288182638261, + "grad_norm": 2.3260059356689453, + "learning_rate": 0.00014898940722065488, + "loss": 0.8257, + "step": 20575 + }, + { + "epoch": 1.427798209700923, + "grad_norm": 2.211399793624878, + "learning_rate": 0.00014898450893568465, + "loss": 0.699, + "step": 20576 + }, + { + "epoch": 1.4278676011380196, + "grad_norm": 1.4518181085586548, + "learning_rate": 0.00014897961049607599, + "loss": 0.4414, + "step": 20577 + }, + { + "epoch": 1.4279369925751162, + "grad_norm": 1.3401950597763062, + "learning_rate": 0.0001489747119018444, + "loss": 0.5438, + "step": 20578 + }, + { + "epoch": 1.428006384012213, + "grad_norm": 1.9856537580490112, + "learning_rate": 0.00014896981315300534, + "loss": 0.6873, + "step": 20579 + }, + { + "epoch": 1.4280757754493096, + "grad_norm": 1.970186471939087, + "learning_rate": 0.00014896491424957423, + "loss": 0.6485, + "step": 20580 + }, + { + "epoch": 1.4281451668864062, + "grad_norm": 1.5439625978469849, + "learning_rate": 0.00014896001519156658, + "loss": 0.6537, + "step": 20581 + }, + { + "epoch": 1.4282145583235029, + "grad_norm": 1.6898796558380127, + "learning_rate": 0.00014895511597899785, + "loss": 0.7545, + "step": 20582 + }, + { + "epoch": 1.4282839497605995, + "grad_norm": 2.3265888690948486, + "learning_rate": 0.00014895021661188347, + "loss": 0.734, + "step": 20583 + }, + { + "epoch": 1.4283533411976963, + "grad_norm": 1.7098183631896973, + "learning_rate": 0.00014894531709023899, + "loss": 0.6373, + "step": 20584 + }, + { + "epoch": 1.428422732634793, + "grad_norm": 2.1531381607055664, + "learning_rate": 0.00014894041741407975, + "loss": 0.6692, + "step": 20585 + }, + { + "epoch": 1.4284921240718895, + "grad_norm": 2.033342123031616, + "learning_rate": 0.00014893551758342135, + "loss": 0.7272, + "step": 20586 + }, + { + "epoch": 1.4285615155089861, + "grad_norm": 1.5202043056488037, + "learning_rate": 0.00014893061759827918, + "loss": 0.5167, + "step": 20587 + }, + { + "epoch": 1.4286309069460827, + "grad_norm": 1.7963110208511353, + "learning_rate": 0.00014892571745866873, + "loss": 0.824, + "step": 20588 + }, + { + "epoch": 1.4287002983831796, + "grad_norm": 1.7563362121582031, + "learning_rate": 0.00014892081716460546, + "loss": 0.6289, + "step": 20589 + }, + { + "epoch": 1.4287696898202762, + "grad_norm": 1.6092771291732788, + "learning_rate": 0.00014891591671610486, + "loss": 0.7574, + "step": 20590 + }, + { + "epoch": 1.4288390812573728, + "grad_norm": 1.6154319047927856, + "learning_rate": 0.00014891101611318237, + "loss": 0.5006, + "step": 20591 + }, + { + "epoch": 1.4289084726944696, + "grad_norm": 1.5419517755508423, + "learning_rate": 0.00014890611535585348, + "loss": 0.664, + "step": 20592 + }, + { + "epoch": 1.4289778641315662, + "grad_norm": 2.671098470687866, + "learning_rate": 0.00014890121444413368, + "loss": 0.8586, + "step": 20593 + }, + { + "epoch": 1.4290472555686629, + "grad_norm": 1.4603002071380615, + "learning_rate": 0.00014889631337803837, + "loss": 0.657, + "step": 20594 + }, + { + "epoch": 1.4291166470057595, + "grad_norm": 1.9867197275161743, + "learning_rate": 0.0001488914121575831, + "loss": 0.5732, + "step": 20595 + }, + { + "epoch": 1.429186038442856, + "grad_norm": 1.425847053527832, + "learning_rate": 0.00014888651078278332, + "loss": 0.6116, + "step": 20596 + }, + { + "epoch": 1.429255429879953, + "grad_norm": 1.1566168069839478, + "learning_rate": 0.0001488816092536545, + "loss": 0.3719, + "step": 20597 + }, + { + "epoch": 1.4293248213170495, + "grad_norm": 1.2967112064361572, + "learning_rate": 0.0001488767075702121, + "loss": 0.5681, + "step": 20598 + }, + { + "epoch": 1.4293942127541461, + "grad_norm": 1.8910526037216187, + "learning_rate": 0.00014887180573247162, + "loss": 0.6971, + "step": 20599 + }, + { + "epoch": 1.4294636041912427, + "grad_norm": 1.849965214729309, + "learning_rate": 0.0001488669037404485, + "loss": 0.6845, + "step": 20600 + }, + { + "epoch": 1.4295329956283394, + "grad_norm": 1.7610307931900024, + "learning_rate": 0.00014886200159415825, + "loss": 0.6686, + "step": 20601 + }, + { + "epoch": 1.4296023870654362, + "grad_norm": 1.5730950832366943, + "learning_rate": 0.0001488570992936163, + "loss": 0.8201, + "step": 20602 + }, + { + "epoch": 1.4296717785025328, + "grad_norm": 1.345844030380249, + "learning_rate": 0.0001488521968388382, + "loss": 0.548, + "step": 20603 + }, + { + "epoch": 1.4297411699396294, + "grad_norm": 2.2646446228027344, + "learning_rate": 0.00014884729422983935, + "loss": 0.5659, + "step": 20604 + }, + { + "epoch": 1.4298105613767262, + "grad_norm": 1.7637441158294678, + "learning_rate": 0.00014884239146663528, + "loss": 0.651, + "step": 20605 + }, + { + "epoch": 1.4298799528138229, + "grad_norm": 1.9081743955612183, + "learning_rate": 0.00014883748854924145, + "loss": 0.4641, + "step": 20606 + }, + { + "epoch": 1.4299493442509195, + "grad_norm": 1.874038577079773, + "learning_rate": 0.00014883258547767332, + "loss": 0.532, + "step": 20607 + }, + { + "epoch": 1.430018735688016, + "grad_norm": 2.358583927154541, + "learning_rate": 0.00014882768225194637, + "loss": 0.653, + "step": 20608 + }, + { + "epoch": 1.4300881271251127, + "grad_norm": 1.5338730812072754, + "learning_rate": 0.0001488227788720761, + "loss": 0.6212, + "step": 20609 + }, + { + "epoch": 1.4301575185622095, + "grad_norm": 1.4669458866119385, + "learning_rate": 0.00014881787533807802, + "loss": 0.4798, + "step": 20610 + }, + { + "epoch": 1.4302269099993061, + "grad_norm": 1.8450593948364258, + "learning_rate": 0.00014881297164996753, + "loss": 0.7913, + "step": 20611 + }, + { + "epoch": 1.4302963014364027, + "grad_norm": 1.521964192390442, + "learning_rate": 0.00014880806780776014, + "loss": 0.7356, + "step": 20612 + }, + { + "epoch": 1.4303656928734994, + "grad_norm": 1.5501015186309814, + "learning_rate": 0.0001488031638114714, + "loss": 0.5788, + "step": 20613 + }, + { + "epoch": 1.430435084310596, + "grad_norm": 1.7869884967803955, + "learning_rate": 0.0001487982596611167, + "loss": 0.7286, + "step": 20614 + }, + { + "epoch": 1.4305044757476928, + "grad_norm": 2.6391592025756836, + "learning_rate": 0.00014879335535671151, + "loss": 0.7621, + "step": 20615 + }, + { + "epoch": 1.4305738671847894, + "grad_norm": 2.4213483333587646, + "learning_rate": 0.0001487884508982714, + "loss": 0.7891, + "step": 20616 + }, + { + "epoch": 1.430643258621886, + "grad_norm": 1.3467072248458862, + "learning_rate": 0.00014878354628581185, + "loss": 0.4956, + "step": 20617 + }, + { + "epoch": 1.4307126500589828, + "grad_norm": 1.8125687837600708, + "learning_rate": 0.00014877864151934827, + "loss": 0.7008, + "step": 20618 + }, + { + "epoch": 1.4307820414960795, + "grad_norm": 2.50500750541687, + "learning_rate": 0.00014877373659889616, + "loss": 0.7256, + "step": 20619 + }, + { + "epoch": 1.430851432933176, + "grad_norm": 1.6600427627563477, + "learning_rate": 0.00014876883152447105, + "loss": 0.8264, + "step": 20620 + }, + { + "epoch": 1.4309208243702727, + "grad_norm": 1.4134114980697632, + "learning_rate": 0.00014876392629608836, + "loss": 0.5093, + "step": 20621 + }, + { + "epoch": 1.4309902158073693, + "grad_norm": 1.5996097326278687, + "learning_rate": 0.00014875902091376366, + "loss": 0.5608, + "step": 20622 + }, + { + "epoch": 1.4310596072444661, + "grad_norm": 2.0456137657165527, + "learning_rate": 0.00014875411537751236, + "loss": 0.6164, + "step": 20623 + }, + { + "epoch": 1.4311289986815627, + "grad_norm": 1.865213394165039, + "learning_rate": 0.00014874920968735, + "loss": 0.937, + "step": 20624 + }, + { + "epoch": 1.4311983901186593, + "grad_norm": 1.468688726425171, + "learning_rate": 0.00014874430384329202, + "loss": 0.591, + "step": 20625 + }, + { + "epoch": 1.431267781555756, + "grad_norm": 1.4397130012512207, + "learning_rate": 0.0001487393978453539, + "loss": 0.6051, + "step": 20626 + }, + { + "epoch": 1.4313371729928526, + "grad_norm": 1.5114067792892456, + "learning_rate": 0.00014873449169355122, + "loss": 0.4743, + "step": 20627 + }, + { + "epoch": 1.4314065644299494, + "grad_norm": 1.5966057777404785, + "learning_rate": 0.00014872958538789937, + "loss": 0.4712, + "step": 20628 + }, + { + "epoch": 1.431475955867046, + "grad_norm": 1.2330573797225952, + "learning_rate": 0.00014872467892841388, + "loss": 0.4664, + "step": 20629 + }, + { + "epoch": 1.4315453473041426, + "grad_norm": 1.8500947952270508, + "learning_rate": 0.00014871977231511023, + "loss": 0.6561, + "step": 20630 + }, + { + "epoch": 1.4316147387412395, + "grad_norm": 1.9507008790969849, + "learning_rate": 0.0001487148655480039, + "loss": 0.6971, + "step": 20631 + }, + { + "epoch": 1.431684130178336, + "grad_norm": 1.3520333766937256, + "learning_rate": 0.00014870995862711038, + "loss": 0.6643, + "step": 20632 + }, + { + "epoch": 1.4317535216154327, + "grad_norm": 1.571871280670166, + "learning_rate": 0.00014870505155244518, + "loss": 0.4499, + "step": 20633 + }, + { + "epoch": 1.4318229130525293, + "grad_norm": 1.474856972694397, + "learning_rate": 0.00014870014432402382, + "loss": 0.542, + "step": 20634 + }, + { + "epoch": 1.431892304489626, + "grad_norm": 2.3302664756774902, + "learning_rate": 0.0001486952369418617, + "loss": 0.7468, + "step": 20635 + }, + { + "epoch": 1.4319616959267227, + "grad_norm": 1.538962483406067, + "learning_rate": 0.0001486903294059744, + "loss": 0.5256, + "step": 20636 + }, + { + "epoch": 1.4320310873638193, + "grad_norm": 1.4727224111557007, + "learning_rate": 0.00014868542171637739, + "loss": 0.5072, + "step": 20637 + }, + { + "epoch": 1.432100478800916, + "grad_norm": 1.6972050666809082, + "learning_rate": 0.0001486805138730861, + "loss": 0.4949, + "step": 20638 + }, + { + "epoch": 1.4321698702380126, + "grad_norm": 1.7323992252349854, + "learning_rate": 0.00014867560587611612, + "loss": 0.7177, + "step": 20639 + }, + { + "epoch": 1.4322392616751092, + "grad_norm": 1.7590746879577637, + "learning_rate": 0.0001486706977254829, + "loss": 0.7316, + "step": 20640 + }, + { + "epoch": 1.432308653112206, + "grad_norm": 1.7072728872299194, + "learning_rate": 0.00014866578942120194, + "loss": 0.7028, + "step": 20641 + }, + { + "epoch": 1.4323780445493026, + "grad_norm": 1.473257303237915, + "learning_rate": 0.00014866088096328873, + "loss": 0.4886, + "step": 20642 + }, + { + "epoch": 1.4324474359863992, + "grad_norm": 1.5596498250961304, + "learning_rate": 0.0001486559723517587, + "loss": 0.6393, + "step": 20643 + }, + { + "epoch": 1.432516827423496, + "grad_norm": 1.4880192279815674, + "learning_rate": 0.00014865106358662747, + "loss": 0.5407, + "step": 20644 + }, + { + "epoch": 1.4325862188605927, + "grad_norm": 1.5388325452804565, + "learning_rate": 0.00014864615466791048, + "loss": 0.6263, + "step": 20645 + }, + { + "epoch": 1.4326556102976893, + "grad_norm": 2.0462002754211426, + "learning_rate": 0.0001486412455956232, + "loss": 0.6298, + "step": 20646 + }, + { + "epoch": 1.432725001734786, + "grad_norm": 2.253143548965454, + "learning_rate": 0.00014863633636978115, + "loss": 0.8387, + "step": 20647 + }, + { + "epoch": 1.4327943931718825, + "grad_norm": 2.1192798614501953, + "learning_rate": 0.00014863142699039986, + "loss": 0.5538, + "step": 20648 + }, + { + "epoch": 1.4328637846089793, + "grad_norm": 1.6039246320724487, + "learning_rate": 0.00014862651745749475, + "loss": 0.4941, + "step": 20649 + }, + { + "epoch": 1.432933176046076, + "grad_norm": 1.8784202337265015, + "learning_rate": 0.0001486216077710814, + "loss": 0.6557, + "step": 20650 + }, + { + "epoch": 1.4330025674831726, + "grad_norm": 1.8275123834609985, + "learning_rate": 0.0001486166979311753, + "loss": 0.6108, + "step": 20651 + }, + { + "epoch": 1.4330719589202692, + "grad_norm": 1.775838017463684, + "learning_rate": 0.00014861178793779188, + "loss": 0.6384, + "step": 20652 + }, + { + "epoch": 1.4331413503573658, + "grad_norm": 1.577883243560791, + "learning_rate": 0.0001486068777909467, + "loss": 0.589, + "step": 20653 + }, + { + "epoch": 1.4332107417944626, + "grad_norm": 1.6231567859649658, + "learning_rate": 0.0001486019674906552, + "loss": 0.6284, + "step": 20654 + }, + { + "epoch": 1.4332801332315592, + "grad_norm": 2.014240026473999, + "learning_rate": 0.000148597057036933, + "loss": 0.6545, + "step": 20655 + }, + { + "epoch": 1.4333495246686558, + "grad_norm": 1.5787243843078613, + "learning_rate": 0.0001485921464297955, + "loss": 0.5305, + "step": 20656 + }, + { + "epoch": 1.4334189161057527, + "grad_norm": 2.2493736743927, + "learning_rate": 0.00014858723566925824, + "loss": 0.824, + "step": 20657 + }, + { + "epoch": 1.4334883075428493, + "grad_norm": 1.7517602443695068, + "learning_rate": 0.0001485823247553367, + "loss": 0.5923, + "step": 20658 + }, + { + "epoch": 1.433557698979946, + "grad_norm": 1.3517673015594482, + "learning_rate": 0.00014857741368804644, + "loss": 0.5414, + "step": 20659 + }, + { + "epoch": 1.4336270904170425, + "grad_norm": 1.4079550504684448, + "learning_rate": 0.00014857250246740286, + "loss": 0.4697, + "step": 20660 + }, + { + "epoch": 1.4336964818541391, + "grad_norm": 2.3170320987701416, + "learning_rate": 0.00014856759109342155, + "loss": 0.5975, + "step": 20661 + }, + { + "epoch": 1.433765873291236, + "grad_norm": 1.8697925806045532, + "learning_rate": 0.000148562679566118, + "loss": 0.8058, + "step": 20662 + }, + { + "epoch": 1.4338352647283326, + "grad_norm": 1.967331051826477, + "learning_rate": 0.00014855776788550768, + "loss": 0.9076, + "step": 20663 + }, + { + "epoch": 1.4339046561654292, + "grad_norm": 1.8500210046768188, + "learning_rate": 0.00014855285605160613, + "loss": 0.6884, + "step": 20664 + }, + { + "epoch": 1.4339740476025258, + "grad_norm": 1.7283403873443604, + "learning_rate": 0.00014854794406442889, + "loss": 0.5852, + "step": 20665 + }, + { + "epoch": 1.4340434390396224, + "grad_norm": 1.8306595087051392, + "learning_rate": 0.00014854303192399137, + "loss": 0.6539, + "step": 20666 + }, + { + "epoch": 1.4341128304767192, + "grad_norm": 1.4917035102844238, + "learning_rate": 0.00014853811963030914, + "loss": 0.485, + "step": 20667 + }, + { + "epoch": 1.4341822219138158, + "grad_norm": 1.5259003639221191, + "learning_rate": 0.00014853320718339773, + "loss": 0.5395, + "step": 20668 + }, + { + "epoch": 1.4342516133509124, + "grad_norm": 1.6666016578674316, + "learning_rate": 0.00014852829458327258, + "loss": 0.5373, + "step": 20669 + }, + { + "epoch": 1.4343210047880093, + "grad_norm": 1.3997882604599, + "learning_rate": 0.00014852338182994924, + "loss": 0.5658, + "step": 20670 + }, + { + "epoch": 1.434390396225106, + "grad_norm": 2.287806272506714, + "learning_rate": 0.0001485184689234432, + "loss": 0.7259, + "step": 20671 + }, + { + "epoch": 1.4344597876622025, + "grad_norm": 1.9185835123062134, + "learning_rate": 0.00014851355586377002, + "loss": 0.5953, + "step": 20672 + }, + { + "epoch": 1.4345291790992991, + "grad_norm": 1.9474774599075317, + "learning_rate": 0.00014850864265094512, + "loss": 0.5034, + "step": 20673 + }, + { + "epoch": 1.4345985705363957, + "grad_norm": 1.737578272819519, + "learning_rate": 0.0001485037292849841, + "loss": 0.7585, + "step": 20674 + }, + { + "epoch": 1.4346679619734926, + "grad_norm": 1.772007703781128, + "learning_rate": 0.00014849881576590245, + "loss": 0.7095, + "step": 20675 + }, + { + "epoch": 1.4347373534105892, + "grad_norm": 1.7266238927841187, + "learning_rate": 0.00014849390209371562, + "loss": 0.5672, + "step": 20676 + }, + { + "epoch": 1.4348067448476858, + "grad_norm": 1.7768093347549438, + "learning_rate": 0.00014848898826843923, + "loss": 0.6873, + "step": 20677 + }, + { + "epoch": 1.4348761362847824, + "grad_norm": 1.7705905437469482, + "learning_rate": 0.00014848407429008867, + "loss": 0.6929, + "step": 20678 + }, + { + "epoch": 1.434945527721879, + "grad_norm": 1.8423744440078735, + "learning_rate": 0.00014847916015867956, + "loss": 0.6033, + "step": 20679 + }, + { + "epoch": 1.4350149191589758, + "grad_norm": 2.4145240783691406, + "learning_rate": 0.00014847424587422734, + "loss": 0.7421, + "step": 20680 + }, + { + "epoch": 1.4350843105960724, + "grad_norm": 1.8548147678375244, + "learning_rate": 0.00014846933143674752, + "loss": 0.9796, + "step": 20681 + }, + { + "epoch": 1.435153702033169, + "grad_norm": 1.887965440750122, + "learning_rate": 0.00014846441684625568, + "loss": 0.7622, + "step": 20682 + }, + { + "epoch": 1.4352230934702659, + "grad_norm": 1.7829025983810425, + "learning_rate": 0.0001484595021027673, + "loss": 0.6325, + "step": 20683 + }, + { + "epoch": 1.4352924849073625, + "grad_norm": 1.7235761880874634, + "learning_rate": 0.0001484545872062979, + "loss": 0.4862, + "step": 20684 + }, + { + "epoch": 1.435361876344459, + "grad_norm": 1.6099992990493774, + "learning_rate": 0.00014844967215686295, + "loss": 0.8228, + "step": 20685 + }, + { + "epoch": 1.4354312677815557, + "grad_norm": 1.3845272064208984, + "learning_rate": 0.00014844475695447807, + "loss": 0.6743, + "step": 20686 + }, + { + "epoch": 1.4355006592186523, + "grad_norm": 1.8764485120773315, + "learning_rate": 0.00014843984159915866, + "loss": 0.6615, + "step": 20687 + }, + { + "epoch": 1.4355700506557492, + "grad_norm": 1.2413671016693115, + "learning_rate": 0.0001484349260909203, + "loss": 0.4873, + "step": 20688 + }, + { + "epoch": 1.4356394420928458, + "grad_norm": 2.0142931938171387, + "learning_rate": 0.00014843001042977853, + "loss": 0.4702, + "step": 20689 + }, + { + "epoch": 1.4357088335299424, + "grad_norm": 1.2058616876602173, + "learning_rate": 0.0001484250946157488, + "loss": 0.5376, + "step": 20690 + }, + { + "epoch": 1.435778224967039, + "grad_norm": 2.2049221992492676, + "learning_rate": 0.00014842017864884666, + "loss": 0.6077, + "step": 20691 + }, + { + "epoch": 1.4358476164041356, + "grad_norm": 1.6715179681777954, + "learning_rate": 0.00014841526252908766, + "loss": 0.6155, + "step": 20692 + }, + { + "epoch": 1.4359170078412324, + "grad_norm": 1.6699352264404297, + "learning_rate": 0.0001484103462564873, + "loss": 0.7014, + "step": 20693 + }, + { + "epoch": 1.435986399278329, + "grad_norm": 1.6923327445983887, + "learning_rate": 0.00014840542983106107, + "loss": 0.6065, + "step": 20694 + }, + { + "epoch": 1.4360557907154257, + "grad_norm": 1.2182116508483887, + "learning_rate": 0.00014840051325282452, + "loss": 0.4937, + "step": 20695 + }, + { + "epoch": 1.4361251821525225, + "grad_norm": 1.4715561866760254, + "learning_rate": 0.00014839559652179317, + "loss": 0.4325, + "step": 20696 + }, + { + "epoch": 1.436194573589619, + "grad_norm": 1.3744410276412964, + "learning_rate": 0.00014839067963798254, + "loss": 0.4398, + "step": 20697 + }, + { + "epoch": 1.4362639650267157, + "grad_norm": 1.973371982574463, + "learning_rate": 0.00014838576260140814, + "loss": 0.9211, + "step": 20698 + }, + { + "epoch": 1.4363333564638123, + "grad_norm": 2.9111580848693848, + "learning_rate": 0.0001483808454120855, + "loss": 0.6083, + "step": 20699 + }, + { + "epoch": 1.436402747900909, + "grad_norm": 2.8190419673919678, + "learning_rate": 0.00014837592807003013, + "loss": 0.8641, + "step": 20700 + }, + { + "epoch": 1.4364721393380058, + "grad_norm": 1.609382152557373, + "learning_rate": 0.0001483710105752576, + "loss": 0.5263, + "step": 20701 + }, + { + "epoch": 1.4365415307751024, + "grad_norm": 1.759642243385315, + "learning_rate": 0.0001483660929277834, + "loss": 0.5548, + "step": 20702 + }, + { + "epoch": 1.436610922212199, + "grad_norm": 1.2526049613952637, + "learning_rate": 0.00014836117512762305, + "loss": 0.3561, + "step": 20703 + }, + { + "epoch": 1.4366803136492956, + "grad_norm": 1.7183462381362915, + "learning_rate": 0.00014835625717479206, + "loss": 0.7675, + "step": 20704 + }, + { + "epoch": 1.4367497050863922, + "grad_norm": 2.191589117050171, + "learning_rate": 0.00014835133906930596, + "loss": 0.7663, + "step": 20705 + }, + { + "epoch": 1.436819096523489, + "grad_norm": 1.4605978727340698, + "learning_rate": 0.00014834642081118033, + "loss": 0.5316, + "step": 20706 + }, + { + "epoch": 1.4368884879605857, + "grad_norm": 1.7805757522583008, + "learning_rate": 0.00014834150240043066, + "loss": 0.6293, + "step": 20707 + }, + { + "epoch": 1.4369578793976823, + "grad_norm": 2.3122494220733643, + "learning_rate": 0.00014833658383707246, + "loss": 0.6622, + "step": 20708 + }, + { + "epoch": 1.437027270834779, + "grad_norm": 1.5587843656539917, + "learning_rate": 0.0001483316651211213, + "loss": 0.7082, + "step": 20709 + }, + { + "epoch": 1.4370966622718757, + "grad_norm": 2.8051064014434814, + "learning_rate": 0.00014832674625259265, + "loss": 0.744, + "step": 20710 + }, + { + "epoch": 1.4371660537089723, + "grad_norm": 1.7460490465164185, + "learning_rate": 0.00014832182723150208, + "loss": 0.4848, + "step": 20711 + }, + { + "epoch": 1.437235445146069, + "grad_norm": 1.4780994653701782, + "learning_rate": 0.0001483169080578651, + "loss": 0.3922, + "step": 20712 + }, + { + "epoch": 1.4373048365831655, + "grad_norm": 1.5094103813171387, + "learning_rate": 0.00014831198873169727, + "loss": 0.6123, + "step": 20713 + }, + { + "epoch": 1.4373742280202624, + "grad_norm": 1.3361436128616333, + "learning_rate": 0.0001483070692530141, + "loss": 0.493, + "step": 20714 + }, + { + "epoch": 1.437443619457359, + "grad_norm": 1.4191726446151733, + "learning_rate": 0.0001483021496218311, + "loss": 0.7693, + "step": 20715 + }, + { + "epoch": 1.4375130108944556, + "grad_norm": 1.5921616554260254, + "learning_rate": 0.00014829722983816378, + "loss": 0.6911, + "step": 20716 + }, + { + "epoch": 1.4375824023315522, + "grad_norm": 1.7809622287750244, + "learning_rate": 0.00014829230990202777, + "loss": 0.6407, + "step": 20717 + }, + { + "epoch": 1.4376517937686488, + "grad_norm": 1.88105309009552, + "learning_rate": 0.0001482873898134385, + "loss": 0.5992, + "step": 20718 + }, + { + "epoch": 1.4377211852057457, + "grad_norm": 1.977306842803955, + "learning_rate": 0.00014828246957241157, + "loss": 0.6871, + "step": 20719 + }, + { + "epoch": 1.4377905766428423, + "grad_norm": 1.7206133604049683, + "learning_rate": 0.0001482775491789625, + "loss": 0.6921, + "step": 20720 + }, + { + "epoch": 1.4378599680799389, + "grad_norm": 2.0093576908111572, + "learning_rate": 0.00014827262863310674, + "loss": 0.5467, + "step": 20721 + }, + { + "epoch": 1.4379293595170357, + "grad_norm": 2.205767869949341, + "learning_rate": 0.00014826770793485993, + "loss": 0.8386, + "step": 20722 + }, + { + "epoch": 1.4379987509541323, + "grad_norm": 2.3821332454681396, + "learning_rate": 0.00014826278708423757, + "loss": 0.9017, + "step": 20723 + }, + { + "epoch": 1.438068142391229, + "grad_norm": 1.6830017566680908, + "learning_rate": 0.00014825786608125522, + "loss": 0.6623, + "step": 20724 + }, + { + "epoch": 1.4381375338283255, + "grad_norm": 1.5413602590560913, + "learning_rate": 0.00014825294492592833, + "loss": 0.5453, + "step": 20725 + }, + { + "epoch": 1.4382069252654222, + "grad_norm": 1.6914499998092651, + "learning_rate": 0.00014824802361827253, + "loss": 0.5842, + "step": 20726 + }, + { + "epoch": 1.438276316702519, + "grad_norm": 1.6382325887680054, + "learning_rate": 0.0001482431021583033, + "loss": 0.6577, + "step": 20727 + }, + { + "epoch": 1.4383457081396156, + "grad_norm": 1.8592596054077148, + "learning_rate": 0.0001482381805460362, + "loss": 0.646, + "step": 20728 + }, + { + "epoch": 1.4384150995767122, + "grad_norm": 1.9513288736343384, + "learning_rate": 0.00014823325878148677, + "loss": 0.6144, + "step": 20729 + }, + { + "epoch": 1.4384844910138088, + "grad_norm": 1.9316437244415283, + "learning_rate": 0.0001482283368646705, + "loss": 0.5914, + "step": 20730 + }, + { + "epoch": 1.4385538824509054, + "grad_norm": 1.8269007205963135, + "learning_rate": 0.000148223414795603, + "loss": 0.6158, + "step": 20731 + }, + { + "epoch": 1.4386232738880023, + "grad_norm": 1.5535210371017456, + "learning_rate": 0.00014821849257429977, + "loss": 0.6616, + "step": 20732 + }, + { + "epoch": 1.4386926653250989, + "grad_norm": 2.0845465660095215, + "learning_rate": 0.00014821357020077636, + "loss": 0.5384, + "step": 20733 + }, + { + "epoch": 1.4387620567621955, + "grad_norm": 1.5809837579727173, + "learning_rate": 0.0001482086476750483, + "loss": 0.5712, + "step": 20734 + }, + { + "epoch": 1.4388314481992923, + "grad_norm": 1.519981026649475, + "learning_rate": 0.00014820372499713114, + "loss": 0.5043, + "step": 20735 + }, + { + "epoch": 1.438900839636389, + "grad_norm": 1.5793555974960327, + "learning_rate": 0.00014819880216704039, + "loss": 0.6196, + "step": 20736 + }, + { + "epoch": 1.4389702310734855, + "grad_norm": 2.4012768268585205, + "learning_rate": 0.00014819387918479161, + "loss": 0.657, + "step": 20737 + }, + { + "epoch": 1.4390396225105822, + "grad_norm": 2.097745656967163, + "learning_rate": 0.00014818895605040038, + "loss": 0.4886, + "step": 20738 + }, + { + "epoch": 1.4391090139476788, + "grad_norm": 2.522904396057129, + "learning_rate": 0.00014818403276388215, + "loss": 0.6931, + "step": 20739 + }, + { + "epoch": 1.4391784053847756, + "grad_norm": 1.8405183553695679, + "learning_rate": 0.00014817910932525256, + "loss": 0.7276, + "step": 20740 + }, + { + "epoch": 1.4392477968218722, + "grad_norm": 1.8806885480880737, + "learning_rate": 0.00014817418573452712, + "loss": 0.6212, + "step": 20741 + }, + { + "epoch": 1.4393171882589688, + "grad_norm": 2.5714781284332275, + "learning_rate": 0.00014816926199172134, + "loss": 0.8885, + "step": 20742 + }, + { + "epoch": 1.4393865796960654, + "grad_norm": 1.9996079206466675, + "learning_rate": 0.0001481643380968508, + "loss": 0.6799, + "step": 20743 + }, + { + "epoch": 1.439455971133162, + "grad_norm": 1.7749496698379517, + "learning_rate": 0.00014815941404993104, + "loss": 0.7029, + "step": 20744 + }, + { + "epoch": 1.4395253625702589, + "grad_norm": 1.67693030834198, + "learning_rate": 0.0001481544898509776, + "loss": 0.6265, + "step": 20745 + }, + { + "epoch": 1.4395947540073555, + "grad_norm": 2.365955114364624, + "learning_rate": 0.00014814956550000601, + "loss": 0.8052, + "step": 20746 + }, + { + "epoch": 1.439664145444452, + "grad_norm": 1.9547572135925293, + "learning_rate": 0.00014814464099703183, + "loss": 0.8269, + "step": 20747 + }, + { + "epoch": 1.439733536881549, + "grad_norm": 2.1664652824401855, + "learning_rate": 0.00014813971634207063, + "loss": 0.6058, + "step": 20748 + }, + { + "epoch": 1.4398029283186455, + "grad_norm": 1.4424915313720703, + "learning_rate": 0.00014813479153513788, + "loss": 0.4964, + "step": 20749 + }, + { + "epoch": 1.4398723197557421, + "grad_norm": 1.7059129476547241, + "learning_rate": 0.0001481298665762492, + "loss": 0.7179, + "step": 20750 + }, + { + "epoch": 1.4399417111928388, + "grad_norm": 1.8887360095977783, + "learning_rate": 0.00014812494146542013, + "loss": 0.7487, + "step": 20751 + }, + { + "epoch": 1.4400111026299354, + "grad_norm": 1.1966211795806885, + "learning_rate": 0.00014812001620266617, + "loss": 0.4833, + "step": 20752 + }, + { + "epoch": 1.4400804940670322, + "grad_norm": 1.4297858476638794, + "learning_rate": 0.00014811509078800292, + "loss": 0.5987, + "step": 20753 + }, + { + "epoch": 1.4401498855041288, + "grad_norm": 1.8885895013809204, + "learning_rate": 0.00014811016522144592, + "loss": 0.7392, + "step": 20754 + }, + { + "epoch": 1.4402192769412254, + "grad_norm": 1.8408441543579102, + "learning_rate": 0.0001481052395030107, + "loss": 0.649, + "step": 20755 + }, + { + "epoch": 1.440288668378322, + "grad_norm": 1.591897964477539, + "learning_rate": 0.0001481003136327128, + "loss": 0.6307, + "step": 20756 + }, + { + "epoch": 1.4403580598154186, + "grad_norm": 1.2675732374191284, + "learning_rate": 0.00014809538761056783, + "loss": 0.4509, + "step": 20757 + }, + { + "epoch": 1.4404274512525155, + "grad_norm": 1.584384799003601, + "learning_rate": 0.0001480904614365913, + "loss": 0.7285, + "step": 20758 + }, + { + "epoch": 1.440496842689612, + "grad_norm": 1.68545401096344, + "learning_rate": 0.00014808553511079875, + "loss": 0.575, + "step": 20759 + }, + { + "epoch": 1.4405662341267087, + "grad_norm": 1.138473629951477, + "learning_rate": 0.00014808060863320574, + "loss": 0.453, + "step": 20760 + }, + { + "epoch": 1.4406356255638055, + "grad_norm": 1.4128810167312622, + "learning_rate": 0.00014807568200382782, + "loss": 0.5971, + "step": 20761 + }, + { + "epoch": 1.4407050170009021, + "grad_norm": 1.5296610593795776, + "learning_rate": 0.00014807075522268057, + "loss": 0.6213, + "step": 20762 + }, + { + "epoch": 1.4407744084379988, + "grad_norm": 2.152625560760498, + "learning_rate": 0.0001480658282897795, + "loss": 0.6626, + "step": 20763 + }, + { + "epoch": 1.4408437998750954, + "grad_norm": 1.7964814901351929, + "learning_rate": 0.0001480609012051402, + "loss": 0.5156, + "step": 20764 + }, + { + "epoch": 1.440913191312192, + "grad_norm": 2.356208324432373, + "learning_rate": 0.0001480559739687782, + "loss": 0.8352, + "step": 20765 + }, + { + "epoch": 1.4409825827492888, + "grad_norm": 1.5356446504592896, + "learning_rate": 0.0001480510465807091, + "loss": 0.6955, + "step": 20766 + }, + { + "epoch": 1.4410519741863854, + "grad_norm": 1.7411028146743774, + "learning_rate": 0.00014804611904094842, + "loss": 0.7894, + "step": 20767 + }, + { + "epoch": 1.441121365623482, + "grad_norm": 1.7164273262023926, + "learning_rate": 0.00014804119134951166, + "loss": 0.6555, + "step": 20768 + }, + { + "epoch": 1.4411907570605786, + "grad_norm": 1.618096947669983, + "learning_rate": 0.0001480362635064145, + "loss": 0.5865, + "step": 20769 + }, + { + "epoch": 1.4412601484976753, + "grad_norm": 1.7357652187347412, + "learning_rate": 0.0001480313355116724, + "loss": 0.6928, + "step": 20770 + }, + { + "epoch": 1.441329539934772, + "grad_norm": 1.8442111015319824, + "learning_rate": 0.0001480264073653009, + "loss": 0.5588, + "step": 20771 + }, + { + "epoch": 1.4413989313718687, + "grad_norm": 1.7344164848327637, + "learning_rate": 0.00014802147906731568, + "loss": 0.4764, + "step": 20772 + }, + { + "epoch": 1.4414683228089653, + "grad_norm": 1.7926639318466187, + "learning_rate": 0.00014801655061773215, + "loss": 0.5139, + "step": 20773 + }, + { + "epoch": 1.4415377142460621, + "grad_norm": 2.478928565979004, + "learning_rate": 0.000148011622016566, + "loss": 0.7249, + "step": 20774 + }, + { + "epoch": 1.4416071056831588, + "grad_norm": 2.309136390686035, + "learning_rate": 0.00014800669326383269, + "loss": 0.7856, + "step": 20775 + }, + { + "epoch": 1.4416764971202554, + "grad_norm": 2.017192840576172, + "learning_rate": 0.00014800176435954783, + "loss": 0.727, + "step": 20776 + }, + { + "epoch": 1.441745888557352, + "grad_norm": 1.5597907304763794, + "learning_rate": 0.00014799683530372702, + "loss": 0.6262, + "step": 20777 + }, + { + "epoch": 1.4418152799944486, + "grad_norm": 1.9745301008224487, + "learning_rate": 0.00014799190609638568, + "loss": 0.7508, + "step": 20778 + }, + { + "epoch": 1.4418846714315454, + "grad_norm": 2.012883186340332, + "learning_rate": 0.00014798697673753955, + "loss": 0.6334, + "step": 20779 + }, + { + "epoch": 1.441954062868642, + "grad_norm": 1.3949607610702515, + "learning_rate": 0.00014798204722720403, + "loss": 0.6108, + "step": 20780 + }, + { + "epoch": 1.4420234543057386, + "grad_norm": 1.4648370742797852, + "learning_rate": 0.00014797711756539482, + "loss": 0.6183, + "step": 20781 + }, + { + "epoch": 1.4420928457428355, + "grad_norm": 1.597528338432312, + "learning_rate": 0.00014797218775212736, + "loss": 0.5491, + "step": 20782 + }, + { + "epoch": 1.4421622371799319, + "grad_norm": 2.1976587772369385, + "learning_rate": 0.0001479672577874173, + "loss": 0.8903, + "step": 20783 + }, + { + "epoch": 1.4422316286170287, + "grad_norm": 2.21012282371521, + "learning_rate": 0.00014796232767128015, + "loss": 0.6513, + "step": 20784 + }, + { + "epoch": 1.4423010200541253, + "grad_norm": 1.6207334995269775, + "learning_rate": 0.00014795739740373152, + "loss": 0.606, + "step": 20785 + }, + { + "epoch": 1.442370411491222, + "grad_norm": 1.611396312713623, + "learning_rate": 0.00014795246698478694, + "loss": 0.7263, + "step": 20786 + }, + { + "epoch": 1.4424398029283187, + "grad_norm": 1.7459428310394287, + "learning_rate": 0.00014794753641446197, + "loss": 0.5823, + "step": 20787 + }, + { + "epoch": 1.4425091943654154, + "grad_norm": 1.0159270763397217, + "learning_rate": 0.0001479426056927722, + "loss": 0.4603, + "step": 20788 + }, + { + "epoch": 1.442578585802512, + "grad_norm": 1.3504202365875244, + "learning_rate": 0.0001479376748197332, + "loss": 0.5893, + "step": 20789 + }, + { + "epoch": 1.4426479772396086, + "grad_norm": 1.6776562929153442, + "learning_rate": 0.00014793274379536051, + "loss": 0.4566, + "step": 20790 + }, + { + "epoch": 1.4427173686767052, + "grad_norm": 1.8500981330871582, + "learning_rate": 0.0001479278126196697, + "loss": 0.7928, + "step": 20791 + }, + { + "epoch": 1.442786760113802, + "grad_norm": 2.039348602294922, + "learning_rate": 0.00014792288129267635, + "loss": 0.6337, + "step": 20792 + }, + { + "epoch": 1.4428561515508986, + "grad_norm": 1.5287489891052246, + "learning_rate": 0.00014791794981439604, + "loss": 0.6327, + "step": 20793 + }, + { + "epoch": 1.4429255429879952, + "grad_norm": 2.163388252258301, + "learning_rate": 0.00014791301818484431, + "loss": 0.478, + "step": 20794 + }, + { + "epoch": 1.442994934425092, + "grad_norm": 2.8473339080810547, + "learning_rate": 0.00014790808640403674, + "loss": 0.9648, + "step": 20795 + }, + { + "epoch": 1.4430643258621885, + "grad_norm": 1.5942144393920898, + "learning_rate": 0.00014790315447198892, + "loss": 0.5533, + "step": 20796 + }, + { + "epoch": 1.4431337172992853, + "grad_norm": 1.8143033981323242, + "learning_rate": 0.00014789822238871637, + "loss": 0.6456, + "step": 20797 + }, + { + "epoch": 1.443203108736382, + "grad_norm": 1.6853899955749512, + "learning_rate": 0.0001478932901542347, + "loss": 0.6178, + "step": 20798 + }, + { + "epoch": 1.4432725001734785, + "grad_norm": 1.7565709352493286, + "learning_rate": 0.00014788835776855948, + "loss": 0.5584, + "step": 20799 + }, + { + "epoch": 1.4433418916105754, + "grad_norm": 1.772334098815918, + "learning_rate": 0.00014788342523170626, + "loss": 0.5227, + "step": 20800 + }, + { + "epoch": 1.443411283047672, + "grad_norm": 2.2416977882385254, + "learning_rate": 0.00014787849254369064, + "loss": 0.579, + "step": 20801 + }, + { + "epoch": 1.4434806744847686, + "grad_norm": 2.0365149974823, + "learning_rate": 0.00014787355970452814, + "loss": 0.7495, + "step": 20802 + }, + { + "epoch": 1.4435500659218652, + "grad_norm": 2.1891727447509766, + "learning_rate": 0.0001478686267142344, + "loss": 0.6769, + "step": 20803 + }, + { + "epoch": 1.4436194573589618, + "grad_norm": 1.770903468132019, + "learning_rate": 0.00014786369357282494, + "loss": 0.7804, + "step": 20804 + }, + { + "epoch": 1.4436888487960586, + "grad_norm": 1.5418859720230103, + "learning_rate": 0.00014785876028031533, + "loss": 0.6351, + "step": 20805 + }, + { + "epoch": 1.4437582402331552, + "grad_norm": 1.5334426164627075, + "learning_rate": 0.0001478538268367212, + "loss": 0.5743, + "step": 20806 + }, + { + "epoch": 1.4438276316702519, + "grad_norm": 2.2766377925872803, + "learning_rate": 0.00014784889324205808, + "loss": 0.7018, + "step": 20807 + }, + { + "epoch": 1.4438970231073487, + "grad_norm": 1.6794100999832153, + "learning_rate": 0.00014784395949634158, + "loss": 0.4308, + "step": 20808 + }, + { + "epoch": 1.4439664145444453, + "grad_norm": 1.8274736404418945, + "learning_rate": 0.0001478390255995872, + "loss": 0.5201, + "step": 20809 + }, + { + "epoch": 1.444035805981542, + "grad_norm": 1.864288330078125, + "learning_rate": 0.0001478340915518106, + "loss": 0.6388, + "step": 20810 + }, + { + "epoch": 1.4441051974186385, + "grad_norm": 1.4747474193572998, + "learning_rate": 0.00014782915735302732, + "loss": 0.6596, + "step": 20811 + }, + { + "epoch": 1.4441745888557351, + "grad_norm": 1.6404350996017456, + "learning_rate": 0.00014782422300325295, + "loss": 0.5944, + "step": 20812 + }, + { + "epoch": 1.444243980292832, + "grad_norm": 1.8810073137283325, + "learning_rate": 0.00014781928850250302, + "loss": 0.7016, + "step": 20813 + }, + { + "epoch": 1.4443133717299286, + "grad_norm": 1.5728552341461182, + "learning_rate": 0.0001478143538507932, + "loss": 0.4774, + "step": 20814 + }, + { + "epoch": 1.4443827631670252, + "grad_norm": 2.199887752532959, + "learning_rate": 0.00014780941904813896, + "loss": 0.578, + "step": 20815 + }, + { + "epoch": 1.4444521546041218, + "grad_norm": 1.8019137382507324, + "learning_rate": 0.00014780448409455597, + "loss": 0.6412, + "step": 20816 + }, + { + "epoch": 1.4445215460412184, + "grad_norm": 3.7817771434783936, + "learning_rate": 0.00014779954899005974, + "loss": 0.6102, + "step": 20817 + }, + { + "epoch": 1.4445909374783152, + "grad_norm": 1.4740337133407593, + "learning_rate": 0.00014779461373466593, + "loss": 0.6064, + "step": 20818 + }, + { + "epoch": 1.4446603289154119, + "grad_norm": 1.1936495304107666, + "learning_rate": 0.00014778967832839, + "loss": 0.4479, + "step": 20819 + }, + { + "epoch": 1.4447297203525085, + "grad_norm": 1.4125909805297852, + "learning_rate": 0.00014778474277124762, + "loss": 0.4742, + "step": 20820 + }, + { + "epoch": 1.4447991117896053, + "grad_norm": 1.5957902669906616, + "learning_rate": 0.00014777980706325437, + "loss": 0.7883, + "step": 20821 + }, + { + "epoch": 1.444868503226702, + "grad_norm": 2.4353067874908447, + "learning_rate": 0.0001477748712044258, + "loss": 0.6092, + "step": 20822 + }, + { + "epoch": 1.4449378946637985, + "grad_norm": 1.633218765258789, + "learning_rate": 0.0001477699351947775, + "loss": 0.5751, + "step": 20823 + }, + { + "epoch": 1.4450072861008951, + "grad_norm": 1.548179268836975, + "learning_rate": 0.0001477649990343251, + "loss": 0.6118, + "step": 20824 + }, + { + "epoch": 1.4450766775379917, + "grad_norm": 1.7187961339950562, + "learning_rate": 0.00014776006272308409, + "loss": 0.7622, + "step": 20825 + }, + { + "epoch": 1.4451460689750886, + "grad_norm": 1.5621395111083984, + "learning_rate": 0.00014775512626107013, + "loss": 0.7377, + "step": 20826 + }, + { + "epoch": 1.4452154604121852, + "grad_norm": 3.0441458225250244, + "learning_rate": 0.00014775018964829876, + "loss": 0.8039, + "step": 20827 + }, + { + "epoch": 1.4452848518492818, + "grad_norm": 1.7979857921600342, + "learning_rate": 0.00014774525288478558, + "loss": 0.5925, + "step": 20828 + }, + { + "epoch": 1.4453542432863784, + "grad_norm": 1.4800471067428589, + "learning_rate": 0.0001477403159705462, + "loss": 0.5172, + "step": 20829 + }, + { + "epoch": 1.445423634723475, + "grad_norm": 2.357807159423828, + "learning_rate": 0.00014773537890559616, + "loss": 0.6678, + "step": 20830 + }, + { + "epoch": 1.4454930261605718, + "grad_norm": 2.161458730697632, + "learning_rate": 0.00014773044168995108, + "loss": 0.6692, + "step": 20831 + }, + { + "epoch": 1.4455624175976685, + "grad_norm": 1.7706186771392822, + "learning_rate": 0.0001477255043236265, + "loss": 0.6016, + "step": 20832 + }, + { + "epoch": 1.445631809034765, + "grad_norm": 1.3921700716018677, + "learning_rate": 0.0001477205668066381, + "loss": 0.6575, + "step": 20833 + }, + { + "epoch": 1.445701200471862, + "grad_norm": 1.5338330268859863, + "learning_rate": 0.00014771562913900137, + "loss": 0.6172, + "step": 20834 + }, + { + "epoch": 1.4457705919089585, + "grad_norm": 2.8063879013061523, + "learning_rate": 0.00014771069132073195, + "loss": 0.599, + "step": 20835 + }, + { + "epoch": 1.4458399833460551, + "grad_norm": 3.99605131149292, + "learning_rate": 0.00014770575335184538, + "loss": 0.6522, + "step": 20836 + }, + { + "epoch": 1.4459093747831517, + "grad_norm": 1.895042896270752, + "learning_rate": 0.0001477008152323573, + "loss": 0.6276, + "step": 20837 + }, + { + "epoch": 1.4459787662202483, + "grad_norm": 3.402468204498291, + "learning_rate": 0.0001476958769622833, + "loss": 0.5882, + "step": 20838 + }, + { + "epoch": 1.4460481576573452, + "grad_norm": 1.5136831998825073, + "learning_rate": 0.0001476909385416389, + "loss": 0.7377, + "step": 20839 + }, + { + "epoch": 1.4461175490944418, + "grad_norm": 1.6454330682754517, + "learning_rate": 0.0001476859999704398, + "loss": 0.6516, + "step": 20840 + }, + { + "epoch": 1.4461869405315384, + "grad_norm": 1.902898907661438, + "learning_rate": 0.00014768106124870152, + "loss": 0.6541, + "step": 20841 + }, + { + "epoch": 1.446256331968635, + "grad_norm": 1.8195130825042725, + "learning_rate": 0.00014767612237643962, + "loss": 0.6543, + "step": 20842 + }, + { + "epoch": 1.4463257234057316, + "grad_norm": 1.8865047693252563, + "learning_rate": 0.00014767118335366977, + "loss": 0.7544, + "step": 20843 + }, + { + "epoch": 1.4463951148428285, + "grad_norm": 1.7306370735168457, + "learning_rate": 0.0001476662441804075, + "loss": 0.5516, + "step": 20844 + }, + { + "epoch": 1.446464506279925, + "grad_norm": 1.9258171319961548, + "learning_rate": 0.00014766130485666844, + "loss": 0.76, + "step": 20845 + }, + { + "epoch": 1.4465338977170217, + "grad_norm": 1.859739899635315, + "learning_rate": 0.00014765636538246818, + "loss": 0.6344, + "step": 20846 + }, + { + "epoch": 1.4466032891541185, + "grad_norm": 2.688542604446411, + "learning_rate": 0.00014765142575782228, + "loss": 0.877, + "step": 20847 + }, + { + "epoch": 1.4466726805912151, + "grad_norm": 1.6099597215652466, + "learning_rate": 0.0001476464859827464, + "loss": 0.4667, + "step": 20848 + }, + { + "epoch": 1.4467420720283117, + "grad_norm": 1.8603228330612183, + "learning_rate": 0.00014764154605725603, + "loss": 0.7614, + "step": 20849 + }, + { + "epoch": 1.4468114634654083, + "grad_norm": 1.7333788871765137, + "learning_rate": 0.00014763660598136687, + "loss": 0.6602, + "step": 20850 + }, + { + "epoch": 1.446880854902505, + "grad_norm": 1.7939797639846802, + "learning_rate": 0.00014763166575509448, + "loss": 0.4438, + "step": 20851 + }, + { + "epoch": 1.4469502463396018, + "grad_norm": 1.5256537199020386, + "learning_rate": 0.00014762672537845445, + "loss": 0.6408, + "step": 20852 + }, + { + "epoch": 1.4470196377766984, + "grad_norm": 1.4563946723937988, + "learning_rate": 0.00014762178485146235, + "loss": 0.6258, + "step": 20853 + }, + { + "epoch": 1.447089029213795, + "grad_norm": 1.521040678024292, + "learning_rate": 0.0001476168441741338, + "loss": 0.6233, + "step": 20854 + }, + { + "epoch": 1.4471584206508916, + "grad_norm": 1.3444627523422241, + "learning_rate": 0.00014761190334648441, + "loss": 0.3162, + "step": 20855 + }, + { + "epoch": 1.4472278120879882, + "grad_norm": 1.9474048614501953, + "learning_rate": 0.00014760696236852975, + "loss": 0.5625, + "step": 20856 + }, + { + "epoch": 1.447297203525085, + "grad_norm": 1.4608092308044434, + "learning_rate": 0.00014760202124028545, + "loss": 0.5119, + "step": 20857 + }, + { + "epoch": 1.4473665949621817, + "grad_norm": 1.9864864349365234, + "learning_rate": 0.00014759707996176707, + "loss": 0.6572, + "step": 20858 + }, + { + "epoch": 1.4474359863992783, + "grad_norm": 1.891736388206482, + "learning_rate": 0.00014759213853299022, + "loss": 0.5638, + "step": 20859 + }, + { + "epoch": 1.4475053778363751, + "grad_norm": 1.9618862867355347, + "learning_rate": 0.00014758719695397057, + "loss": 0.9012, + "step": 20860 + }, + { + "epoch": 1.4475747692734717, + "grad_norm": 1.7447261810302734, + "learning_rate": 0.0001475822552247236, + "loss": 0.8177, + "step": 20861 + }, + { + "epoch": 1.4476441607105683, + "grad_norm": 1.5832319259643555, + "learning_rate": 0.00014757731334526502, + "loss": 0.5878, + "step": 20862 + }, + { + "epoch": 1.447713552147665, + "grad_norm": 1.5964899063110352, + "learning_rate": 0.00014757237131561034, + "loss": 0.7087, + "step": 20863 + }, + { + "epoch": 1.4477829435847616, + "grad_norm": 1.4241491556167603, + "learning_rate": 0.0001475674291357752, + "loss": 0.51, + "step": 20864 + }, + { + "epoch": 1.4478523350218584, + "grad_norm": 1.7354240417480469, + "learning_rate": 0.00014756248680577525, + "loss": 0.761, + "step": 20865 + }, + { + "epoch": 1.447921726458955, + "grad_norm": 1.8471505641937256, + "learning_rate": 0.000147557544325626, + "loss": 0.519, + "step": 20866 + }, + { + "epoch": 1.4479911178960516, + "grad_norm": 2.219447374343872, + "learning_rate": 0.00014755260169534314, + "loss": 0.7522, + "step": 20867 + }, + { + "epoch": 1.4480605093331482, + "grad_norm": 1.6430906057357788, + "learning_rate": 0.0001475476589149422, + "loss": 0.6353, + "step": 20868 + }, + { + "epoch": 1.4481299007702448, + "grad_norm": 2.1980373859405518, + "learning_rate": 0.00014754271598443886, + "loss": 0.6822, + "step": 20869 + }, + { + "epoch": 1.4481992922073417, + "grad_norm": 2.6251749992370605, + "learning_rate": 0.00014753777290384864, + "loss": 0.5727, + "step": 20870 + }, + { + "epoch": 1.4482686836444383, + "grad_norm": 3.3548147678375244, + "learning_rate": 0.0001475328296731872, + "loss": 0.6565, + "step": 20871 + }, + { + "epoch": 1.448338075081535, + "grad_norm": 1.985350251197815, + "learning_rate": 0.00014752788629247013, + "loss": 0.6114, + "step": 20872 + }, + { + "epoch": 1.4484074665186317, + "grad_norm": 1.383495569229126, + "learning_rate": 0.00014752294276171304, + "loss": 0.5904, + "step": 20873 + }, + { + "epoch": 1.4484768579557283, + "grad_norm": 1.815982699394226, + "learning_rate": 0.00014751799908093154, + "loss": 0.6621, + "step": 20874 + }, + { + "epoch": 1.448546249392825, + "grad_norm": 2.14467716217041, + "learning_rate": 0.00014751305525014124, + "loss": 1.0754, + "step": 20875 + }, + { + "epoch": 1.4486156408299216, + "grad_norm": 1.3223425149917603, + "learning_rate": 0.0001475081112693577, + "loss": 0.6565, + "step": 20876 + }, + { + "epoch": 1.4486850322670182, + "grad_norm": 2.218418598175049, + "learning_rate": 0.0001475031671385966, + "loss": 0.6105, + "step": 20877 + }, + { + "epoch": 1.448754423704115, + "grad_norm": 2.3032643795013428, + "learning_rate": 0.0001474982228578735, + "loss": 0.6047, + "step": 20878 + }, + { + "epoch": 1.4488238151412116, + "grad_norm": 1.4959893226623535, + "learning_rate": 0.00014749327842720401, + "loss": 0.4937, + "step": 20879 + }, + { + "epoch": 1.4488932065783082, + "grad_norm": 3.1564674377441406, + "learning_rate": 0.00014748833384660377, + "loss": 0.6583, + "step": 20880 + }, + { + "epoch": 1.4489625980154048, + "grad_norm": 2.1346287727355957, + "learning_rate": 0.00014748338911608835, + "loss": 0.8831, + "step": 20881 + }, + { + "epoch": 1.4490319894525014, + "grad_norm": 2.013685941696167, + "learning_rate": 0.00014747844423567335, + "loss": 0.6247, + "step": 20882 + }, + { + "epoch": 1.4491013808895983, + "grad_norm": 1.9278556108474731, + "learning_rate": 0.00014747349920537448, + "loss": 0.4492, + "step": 20883 + }, + { + "epoch": 1.449170772326695, + "grad_norm": 1.4734418392181396, + "learning_rate": 0.00014746855402520724, + "loss": 0.6297, + "step": 20884 + }, + { + "epoch": 1.4492401637637915, + "grad_norm": 1.792027235031128, + "learning_rate": 0.0001474636086951873, + "loss": 0.4911, + "step": 20885 + }, + { + "epoch": 1.4493095552008883, + "grad_norm": 1.8268694877624512, + "learning_rate": 0.00014745866321533025, + "loss": 0.6708, + "step": 20886 + }, + { + "epoch": 1.449378946637985, + "grad_norm": 1.8647501468658447, + "learning_rate": 0.0001474537175856517, + "loss": 0.7357, + "step": 20887 + }, + { + "epoch": 1.4494483380750816, + "grad_norm": 1.6032599210739136, + "learning_rate": 0.00014744877180616726, + "loss": 0.6023, + "step": 20888 + }, + { + "epoch": 1.4495177295121782, + "grad_norm": 1.544019341468811, + "learning_rate": 0.0001474438258768926, + "loss": 0.6307, + "step": 20889 + }, + { + "epoch": 1.4495871209492748, + "grad_norm": 1.6285045146942139, + "learning_rate": 0.00014743887979784326, + "loss": 0.7616, + "step": 20890 + }, + { + "epoch": 1.4496565123863716, + "grad_norm": 1.3612512350082397, + "learning_rate": 0.00014743393356903486, + "loss": 0.6277, + "step": 20891 + }, + { + "epoch": 1.4497259038234682, + "grad_norm": 1.446960210800171, + "learning_rate": 0.00014742898719048304, + "loss": 0.5421, + "step": 20892 + }, + { + "epoch": 1.4497952952605648, + "grad_norm": 2.2777791023254395, + "learning_rate": 0.00014742404066220343, + "loss": 0.6051, + "step": 20893 + }, + { + "epoch": 1.4498646866976614, + "grad_norm": 2.4014675617218018, + "learning_rate": 0.0001474190939842116, + "loss": 0.6372, + "step": 20894 + }, + { + "epoch": 1.449934078134758, + "grad_norm": 2.2238736152648926, + "learning_rate": 0.0001474141471565232, + "loss": 0.7756, + "step": 20895 + }, + { + "epoch": 1.4500034695718549, + "grad_norm": 1.6042757034301758, + "learning_rate": 0.00014740920017915384, + "loss": 0.5228, + "step": 20896 + }, + { + "epoch": 1.4500728610089515, + "grad_norm": 1.5134893655776978, + "learning_rate": 0.00014740425305211912, + "loss": 0.4845, + "step": 20897 + }, + { + "epoch": 1.450142252446048, + "grad_norm": 1.6530154943466187, + "learning_rate": 0.0001473993057754347, + "loss": 0.7498, + "step": 20898 + }, + { + "epoch": 1.450211643883145, + "grad_norm": 1.499556303024292, + "learning_rate": 0.00014739435834911616, + "loss": 0.3562, + "step": 20899 + }, + { + "epoch": 1.4502810353202416, + "grad_norm": 1.676912546157837, + "learning_rate": 0.00014738941077317916, + "loss": 0.4594, + "step": 20900 + }, + { + "epoch": 1.4503504267573382, + "grad_norm": 2.4733381271362305, + "learning_rate": 0.00014738446304763924, + "loss": 0.478, + "step": 20901 + }, + { + "epoch": 1.4504198181944348, + "grad_norm": 1.6225337982177734, + "learning_rate": 0.0001473795151725121, + "loss": 0.5488, + "step": 20902 + }, + { + "epoch": 1.4504892096315314, + "grad_norm": 2.3541247844696045, + "learning_rate": 0.00014737456714781333, + "loss": 0.7327, + "step": 20903 + }, + { + "epoch": 1.4505586010686282, + "grad_norm": 2.521716594696045, + "learning_rate": 0.0001473696189735585, + "loss": 0.8586, + "step": 20904 + }, + { + "epoch": 1.4506279925057248, + "grad_norm": 2.424668312072754, + "learning_rate": 0.00014736467064976332, + "loss": 0.6809, + "step": 20905 + }, + { + "epoch": 1.4506973839428214, + "grad_norm": 2.2287232875823975, + "learning_rate": 0.00014735972217644338, + "loss": 0.5939, + "step": 20906 + }, + { + "epoch": 1.450766775379918, + "grad_norm": 1.7509337663650513, + "learning_rate": 0.00014735477355361427, + "loss": 0.5384, + "step": 20907 + }, + { + "epoch": 1.4508361668170147, + "grad_norm": 1.3918042182922363, + "learning_rate": 0.00014734982478129165, + "loss": 0.5926, + "step": 20908 + }, + { + "epoch": 1.4509055582541115, + "grad_norm": 1.73238205909729, + "learning_rate": 0.0001473448758594911, + "loss": 0.6557, + "step": 20909 + }, + { + "epoch": 1.450974949691208, + "grad_norm": 2.0206263065338135, + "learning_rate": 0.0001473399267882283, + "loss": 0.613, + "step": 20910 + }, + { + "epoch": 1.4510443411283047, + "grad_norm": 1.1667662858963013, + "learning_rate": 0.00014733497756751882, + "loss": 0.3811, + "step": 20911 + }, + { + "epoch": 1.4511137325654015, + "grad_norm": 1.9021611213684082, + "learning_rate": 0.0001473300281973783, + "loss": 0.8071, + "step": 20912 + }, + { + "epoch": 1.4511831240024982, + "grad_norm": 1.4774693250656128, + "learning_rate": 0.00014732507867782238, + "loss": 0.6228, + "step": 20913 + }, + { + "epoch": 1.4512525154395948, + "grad_norm": 2.3910329341888428, + "learning_rate": 0.0001473201290088667, + "loss": 0.8431, + "step": 20914 + }, + { + "epoch": 1.4513219068766914, + "grad_norm": 2.300584077835083, + "learning_rate": 0.00014731517919052683, + "loss": 0.6966, + "step": 20915 + }, + { + "epoch": 1.451391298313788, + "grad_norm": 1.3898732662200928, + "learning_rate": 0.00014731022922281846, + "loss": 0.3679, + "step": 20916 + }, + { + "epoch": 1.4514606897508848, + "grad_norm": 2.001339912414551, + "learning_rate": 0.0001473052791057572, + "loss": 0.6815, + "step": 20917 + }, + { + "epoch": 1.4515300811879814, + "grad_norm": 2.6764895915985107, + "learning_rate": 0.00014730032883935863, + "loss": 0.7723, + "step": 20918 + }, + { + "epoch": 1.451599472625078, + "grad_norm": 1.6894668340682983, + "learning_rate": 0.00014729537842363838, + "loss": 0.459, + "step": 20919 + }, + { + "epoch": 1.4516688640621747, + "grad_norm": 1.9960899353027344, + "learning_rate": 0.00014729042785861215, + "loss": 0.7459, + "step": 20920 + }, + { + "epoch": 1.4517382554992713, + "grad_norm": 1.9780941009521484, + "learning_rate": 0.00014728547714429552, + "loss": 0.731, + "step": 20921 + }, + { + "epoch": 1.451807646936368, + "grad_norm": 1.3159048557281494, + "learning_rate": 0.00014728052628070414, + "loss": 0.622, + "step": 20922 + }, + { + "epoch": 1.4518770383734647, + "grad_norm": 1.3898509740829468, + "learning_rate": 0.0001472755752678536, + "loss": 0.4464, + "step": 20923 + }, + { + "epoch": 1.4519464298105613, + "grad_norm": 1.7722480297088623, + "learning_rate": 0.00014727062410575958, + "loss": 0.6855, + "step": 20924 + }, + { + "epoch": 1.4520158212476582, + "grad_norm": 1.676762342453003, + "learning_rate": 0.00014726567279443767, + "loss": 0.6467, + "step": 20925 + }, + { + "epoch": 1.4520852126847548, + "grad_norm": 2.2835559844970703, + "learning_rate": 0.0001472607213339035, + "loss": 0.7805, + "step": 20926 + }, + { + "epoch": 1.4521546041218514, + "grad_norm": 2.214165210723877, + "learning_rate": 0.00014725576972417272, + "loss": 0.7917, + "step": 20927 + }, + { + "epoch": 1.452223995558948, + "grad_norm": 1.9133192300796509, + "learning_rate": 0.00014725081796526096, + "loss": 0.7337, + "step": 20928 + }, + { + "epoch": 1.4522933869960446, + "grad_norm": 2.5874195098876953, + "learning_rate": 0.00014724586605718387, + "loss": 0.4608, + "step": 20929 + }, + { + "epoch": 1.4523627784331414, + "grad_norm": 2.650904417037964, + "learning_rate": 0.00014724091399995702, + "loss": 0.6801, + "step": 20930 + }, + { + "epoch": 1.452432169870238, + "grad_norm": 1.563271403312683, + "learning_rate": 0.00014723596179359614, + "loss": 0.6587, + "step": 20931 + }, + { + "epoch": 1.4525015613073347, + "grad_norm": 1.3832197189331055, + "learning_rate": 0.00014723100943811677, + "loss": 0.4512, + "step": 20932 + }, + { + "epoch": 1.4525709527444313, + "grad_norm": 1.7529103755950928, + "learning_rate": 0.0001472260569335346, + "loss": 0.6936, + "step": 20933 + }, + { + "epoch": 1.4526403441815279, + "grad_norm": 1.4027819633483887, + "learning_rate": 0.0001472211042798652, + "loss": 0.4462, + "step": 20934 + }, + { + "epoch": 1.4527097356186247, + "grad_norm": 1.3999689817428589, + "learning_rate": 0.00014721615147712434, + "loss": 0.5852, + "step": 20935 + }, + { + "epoch": 1.4527791270557213, + "grad_norm": 1.8154727220535278, + "learning_rate": 0.0001472111985253275, + "loss": 0.8165, + "step": 20936 + }, + { + "epoch": 1.452848518492818, + "grad_norm": 1.6944594383239746, + "learning_rate": 0.0001472062454244904, + "loss": 0.5464, + "step": 20937 + }, + { + "epoch": 1.4529179099299148, + "grad_norm": 1.9738978147506714, + "learning_rate": 0.0001472012921746287, + "loss": 0.7768, + "step": 20938 + }, + { + "epoch": 1.4529873013670114, + "grad_norm": 1.434037446975708, + "learning_rate": 0.00014719633877575793, + "loss": 0.6453, + "step": 20939 + }, + { + "epoch": 1.453056692804108, + "grad_norm": 1.3823434114456177, + "learning_rate": 0.00014719138522789385, + "loss": 0.5451, + "step": 20940 + }, + { + "epoch": 1.4531260842412046, + "grad_norm": 2.542970657348633, + "learning_rate": 0.00014718643153105202, + "loss": 0.6985, + "step": 20941 + }, + { + "epoch": 1.4531954756783012, + "grad_norm": 1.812201976776123, + "learning_rate": 0.00014718147768524807, + "loss": 0.6953, + "step": 20942 + }, + { + "epoch": 1.453264867115398, + "grad_norm": 1.9208924770355225, + "learning_rate": 0.0001471765236904977, + "loss": 0.5973, + "step": 20943 + }, + { + "epoch": 1.4533342585524947, + "grad_norm": 1.8209843635559082, + "learning_rate": 0.00014717156954681653, + "loss": 0.6562, + "step": 20944 + }, + { + "epoch": 1.4534036499895913, + "grad_norm": 3.5132155418395996, + "learning_rate": 0.00014716661525422016, + "loss": 0.58, + "step": 20945 + }, + { + "epoch": 1.4534730414266879, + "grad_norm": 1.8336753845214844, + "learning_rate": 0.0001471616608127243, + "loss": 0.6849, + "step": 20946 + }, + { + "epoch": 1.4535424328637845, + "grad_norm": 1.9435651302337646, + "learning_rate": 0.0001471567062223445, + "loss": 0.7758, + "step": 20947 + }, + { + "epoch": 1.4536118243008813, + "grad_norm": 1.2893344163894653, + "learning_rate": 0.00014715175148309647, + "loss": 0.4959, + "step": 20948 + }, + { + "epoch": 1.453681215737978, + "grad_norm": 1.217354416847229, + "learning_rate": 0.00014714679659499585, + "loss": 0.435, + "step": 20949 + }, + { + "epoch": 1.4537506071750745, + "grad_norm": 2.6080198287963867, + "learning_rate": 0.00014714184155805823, + "loss": 0.8437, + "step": 20950 + }, + { + "epoch": 1.4538199986121714, + "grad_norm": 1.520395040512085, + "learning_rate": 0.00014713688637229928, + "loss": 0.474, + "step": 20951 + }, + { + "epoch": 1.453889390049268, + "grad_norm": 1.8402239084243774, + "learning_rate": 0.00014713193103773467, + "loss": 0.7135, + "step": 20952 + }, + { + "epoch": 1.4539587814863646, + "grad_norm": 1.3180670738220215, + "learning_rate": 0.00014712697555438004, + "loss": 0.5217, + "step": 20953 + }, + { + "epoch": 1.4540281729234612, + "grad_norm": 1.9976435899734497, + "learning_rate": 0.000147122019922251, + "loss": 0.5692, + "step": 20954 + }, + { + "epoch": 1.4540975643605578, + "grad_norm": 1.4611855745315552, + "learning_rate": 0.0001471170641413632, + "loss": 0.3896, + "step": 20955 + }, + { + "epoch": 1.4541669557976546, + "grad_norm": 4.056459903717041, + "learning_rate": 0.0001471121082117323, + "loss": 0.7179, + "step": 20956 + }, + { + "epoch": 1.4542363472347513, + "grad_norm": 1.9000725746154785, + "learning_rate": 0.00014710715213337396, + "loss": 0.4832, + "step": 20957 + }, + { + "epoch": 1.4543057386718479, + "grad_norm": 2.297262668609619, + "learning_rate": 0.0001471021959063038, + "loss": 0.6518, + "step": 20958 + }, + { + "epoch": 1.4543751301089445, + "grad_norm": 1.3400295972824097, + "learning_rate": 0.00014709723953053744, + "loss": 0.3832, + "step": 20959 + }, + { + "epoch": 1.454444521546041, + "grad_norm": 1.6563836336135864, + "learning_rate": 0.0001470922830060906, + "loss": 0.5696, + "step": 20960 + }, + { + "epoch": 1.454513912983138, + "grad_norm": 1.72965669631958, + "learning_rate": 0.00014708732633297882, + "loss": 0.4686, + "step": 20961 + }, + { + "epoch": 1.4545833044202345, + "grad_norm": 1.708404302597046, + "learning_rate": 0.0001470823695112179, + "loss": 0.4559, + "step": 20962 + }, + { + "epoch": 1.4546526958573311, + "grad_norm": 1.548078179359436, + "learning_rate": 0.00014707741254082338, + "loss": 0.6219, + "step": 20963 + }, + { + "epoch": 1.454722087294428, + "grad_norm": 1.5115067958831787, + "learning_rate": 0.0001470724554218109, + "loss": 0.535, + "step": 20964 + }, + { + "epoch": 1.4547914787315246, + "grad_norm": 1.6742911338806152, + "learning_rate": 0.00014706749815419614, + "loss": 0.6698, + "step": 20965 + }, + { + "epoch": 1.4548608701686212, + "grad_norm": 1.8825825452804565, + "learning_rate": 0.00014706254073799477, + "loss": 0.7381, + "step": 20966 + }, + { + "epoch": 1.4549302616057178, + "grad_norm": 1.768850326538086, + "learning_rate": 0.0001470575831732224, + "loss": 0.6321, + "step": 20967 + }, + { + "epoch": 1.4549996530428144, + "grad_norm": 2.6719555854797363, + "learning_rate": 0.00014705262545989468, + "loss": 0.7114, + "step": 20968 + }, + { + "epoch": 1.4550690444799113, + "grad_norm": 2.256276845932007, + "learning_rate": 0.00014704766759802735, + "loss": 0.8134, + "step": 20969 + }, + { + "epoch": 1.4551384359170079, + "grad_norm": 2.20719313621521, + "learning_rate": 0.00014704270958763592, + "loss": 0.7206, + "step": 20970 + }, + { + "epoch": 1.4552078273541045, + "grad_norm": 1.8181865215301514, + "learning_rate": 0.00014703775142873616, + "loss": 0.6736, + "step": 20971 + }, + { + "epoch": 1.455277218791201, + "grad_norm": 2.0252318382263184, + "learning_rate": 0.00014703279312134365, + "loss": 0.8028, + "step": 20972 + }, + { + "epoch": 1.4553466102282977, + "grad_norm": 1.8546617031097412, + "learning_rate": 0.0001470278346654741, + "loss": 0.7634, + "step": 20973 + }, + { + "epoch": 1.4554160016653945, + "grad_norm": 1.608742117881775, + "learning_rate": 0.0001470228760611431, + "loss": 0.5535, + "step": 20974 + }, + { + "epoch": 1.4554853931024911, + "grad_norm": 2.181490421295166, + "learning_rate": 0.00014701791730836638, + "loss": 0.59, + "step": 20975 + }, + { + "epoch": 1.4555547845395878, + "grad_norm": 1.4255057573318481, + "learning_rate": 0.0001470129584071595, + "loss": 0.5209, + "step": 20976 + }, + { + "epoch": 1.4556241759766846, + "grad_norm": 2.354731559753418, + "learning_rate": 0.0001470079993575382, + "loss": 0.8076, + "step": 20977 + }, + { + "epoch": 1.4556935674137812, + "grad_norm": 1.810673475265503, + "learning_rate": 0.00014700304015951806, + "loss": 0.5365, + "step": 20978 + }, + { + "epoch": 1.4557629588508778, + "grad_norm": 1.585456371307373, + "learning_rate": 0.00014699808081311484, + "loss": 0.7189, + "step": 20979 + }, + { + "epoch": 1.4558323502879744, + "grad_norm": 2.3081161975860596, + "learning_rate": 0.0001469931213183441, + "loss": 0.7227, + "step": 20980 + }, + { + "epoch": 1.455901741725071, + "grad_norm": 1.6436123847961426, + "learning_rate": 0.0001469881616752215, + "loss": 0.6043, + "step": 20981 + }, + { + "epoch": 1.4559711331621679, + "grad_norm": 1.723504900932312, + "learning_rate": 0.00014698320188376274, + "loss": 0.5763, + "step": 20982 + }, + { + "epoch": 1.4560405245992645, + "grad_norm": 3.5066750049591064, + "learning_rate": 0.00014697824194398349, + "loss": 0.4695, + "step": 20983 + }, + { + "epoch": 1.456109916036361, + "grad_norm": 1.5958271026611328, + "learning_rate": 0.00014697328185589933, + "loss": 0.569, + "step": 20984 + }, + { + "epoch": 1.4561793074734577, + "grad_norm": 1.385988712310791, + "learning_rate": 0.000146968321619526, + "loss": 0.5536, + "step": 20985 + }, + { + "epoch": 1.4562486989105543, + "grad_norm": 2.4941375255584717, + "learning_rate": 0.00014696336123487912, + "loss": 0.7437, + "step": 20986 + }, + { + "epoch": 1.4563180903476511, + "grad_norm": 1.8339979648590088, + "learning_rate": 0.00014695840070197438, + "loss": 0.5692, + "step": 20987 + }, + { + "epoch": 1.4563874817847478, + "grad_norm": 1.4139988422393799, + "learning_rate": 0.00014695344002082738, + "loss": 0.5222, + "step": 20988 + }, + { + "epoch": 1.4564568732218444, + "grad_norm": 2.2068047523498535, + "learning_rate": 0.00014694847919145384, + "loss": 0.58, + "step": 20989 + }, + { + "epoch": 1.4565262646589412, + "grad_norm": 1.5123291015625, + "learning_rate": 0.0001469435182138694, + "loss": 0.4095, + "step": 20990 + }, + { + "epoch": 1.4565956560960378, + "grad_norm": 1.8539233207702637, + "learning_rate": 0.0001469385570880897, + "loss": 0.5874, + "step": 20991 + }, + { + "epoch": 1.4566650475331344, + "grad_norm": 2.88285231590271, + "learning_rate": 0.00014693359581413045, + "loss": 0.6002, + "step": 20992 + }, + { + "epoch": 1.456734438970231, + "grad_norm": 2.2324025630950928, + "learning_rate": 0.00014692863439200727, + "loss": 0.8379, + "step": 20993 + }, + { + "epoch": 1.4568038304073276, + "grad_norm": 1.745849370956421, + "learning_rate": 0.00014692367282173586, + "loss": 0.6677, + "step": 20994 + }, + { + "epoch": 1.4568732218444245, + "grad_norm": 1.8434120416641235, + "learning_rate": 0.00014691871110333185, + "loss": 0.7741, + "step": 20995 + }, + { + "epoch": 1.456942613281521, + "grad_norm": 1.5546486377716064, + "learning_rate": 0.00014691374923681087, + "loss": 0.5681, + "step": 20996 + }, + { + "epoch": 1.4570120047186177, + "grad_norm": 1.7063617706298828, + "learning_rate": 0.00014690878722218865, + "loss": 0.5306, + "step": 20997 + }, + { + "epoch": 1.4570813961557143, + "grad_norm": 2.0528788566589355, + "learning_rate": 0.00014690382505948085, + "loss": 0.7092, + "step": 20998 + }, + { + "epoch": 1.457150787592811, + "grad_norm": 1.8930741548538208, + "learning_rate": 0.0001468988627487031, + "loss": 0.9339, + "step": 20999 + }, + { + "epoch": 1.4572201790299077, + "grad_norm": 1.658867597579956, + "learning_rate": 0.0001468939002898711, + "loss": 0.5035, + "step": 21000 + }, + { + "epoch": 1.4572895704670044, + "grad_norm": 1.4549745321273804, + "learning_rate": 0.00014688893768300044, + "loss": 0.5374, + "step": 21001 + }, + { + "epoch": 1.457358961904101, + "grad_norm": 1.9677495956420898, + "learning_rate": 0.0001468839749281069, + "loss": 0.6767, + "step": 21002 + }, + { + "epoch": 1.4574283533411978, + "grad_norm": 1.7912108898162842, + "learning_rate": 0.00014687901202520607, + "loss": 0.5537, + "step": 21003 + }, + { + "epoch": 1.4574977447782944, + "grad_norm": 2.043231248855591, + "learning_rate": 0.00014687404897431366, + "loss": 0.7715, + "step": 21004 + }, + { + "epoch": 1.457567136215391, + "grad_norm": 2.082512855529785, + "learning_rate": 0.00014686908577544528, + "loss": 0.8955, + "step": 21005 + }, + { + "epoch": 1.4576365276524876, + "grad_norm": 1.8094061613082886, + "learning_rate": 0.00014686412242861664, + "loss": 0.6127, + "step": 21006 + }, + { + "epoch": 1.4577059190895842, + "grad_norm": 1.459897518157959, + "learning_rate": 0.00014685915893384346, + "loss": 0.5487, + "step": 21007 + }, + { + "epoch": 1.457775310526681, + "grad_norm": 2.1013736724853516, + "learning_rate": 0.00014685419529114128, + "loss": 0.7508, + "step": 21008 + }, + { + "epoch": 1.4578447019637777, + "grad_norm": 2.040786027908325, + "learning_rate": 0.00014684923150052586, + "loss": 0.5166, + "step": 21009 + }, + { + "epoch": 1.4579140934008743, + "grad_norm": 1.3600860834121704, + "learning_rate": 0.0001468442675620129, + "loss": 0.6037, + "step": 21010 + }, + { + "epoch": 1.457983484837971, + "grad_norm": 1.8339827060699463, + "learning_rate": 0.000146839303475618, + "loss": 0.6607, + "step": 21011 + }, + { + "epoch": 1.4580528762750675, + "grad_norm": 1.6504374742507935, + "learning_rate": 0.0001468343392413568, + "loss": 0.6976, + "step": 21012 + }, + { + "epoch": 1.4581222677121644, + "grad_norm": 1.9364691972732544, + "learning_rate": 0.00014682937485924506, + "loss": 0.8197, + "step": 21013 + }, + { + "epoch": 1.458191659149261, + "grad_norm": 1.5623563528060913, + "learning_rate": 0.00014682441032929844, + "loss": 0.5821, + "step": 21014 + }, + { + "epoch": 1.4582610505863576, + "grad_norm": 1.8476046323776245, + "learning_rate": 0.00014681944565153257, + "loss": 0.7714, + "step": 21015 + }, + { + "epoch": 1.4583304420234544, + "grad_norm": 2.250939130783081, + "learning_rate": 0.00014681448082596313, + "loss": 0.5458, + "step": 21016 + }, + { + "epoch": 1.458399833460551, + "grad_norm": 1.669986367225647, + "learning_rate": 0.0001468095158526058, + "loss": 0.5974, + "step": 21017 + }, + { + "epoch": 1.4584692248976476, + "grad_norm": 1.5658841133117676, + "learning_rate": 0.00014680455073147626, + "loss": 0.7471, + "step": 21018 + }, + { + "epoch": 1.4585386163347442, + "grad_norm": 1.810194730758667, + "learning_rate": 0.00014679958546259022, + "loss": 0.5202, + "step": 21019 + }, + { + "epoch": 1.4586080077718409, + "grad_norm": 2.933014154434204, + "learning_rate": 0.00014679462004596328, + "loss": 0.959, + "step": 21020 + }, + { + "epoch": 1.4586773992089377, + "grad_norm": 1.5072942972183228, + "learning_rate": 0.0001467896544816112, + "loss": 0.4952, + "step": 21021 + }, + { + "epoch": 1.4587467906460343, + "grad_norm": 1.9737989902496338, + "learning_rate": 0.00014678468876954957, + "loss": 0.6418, + "step": 21022 + }, + { + "epoch": 1.458816182083131, + "grad_norm": 2.0309319496154785, + "learning_rate": 0.0001467797229097941, + "loss": 0.6041, + "step": 21023 + }, + { + "epoch": 1.4588855735202277, + "grad_norm": 1.6692309379577637, + "learning_rate": 0.0001467747569023605, + "loss": 0.4743, + "step": 21024 + }, + { + "epoch": 1.4589549649573241, + "grad_norm": 1.951597809791565, + "learning_rate": 0.0001467697907472644, + "loss": 0.6198, + "step": 21025 + }, + { + "epoch": 1.459024356394421, + "grad_norm": 1.696624517440796, + "learning_rate": 0.0001467648244445215, + "loss": 0.5193, + "step": 21026 + }, + { + "epoch": 1.4590937478315176, + "grad_norm": 1.5446563959121704, + "learning_rate": 0.00014675985799414749, + "loss": 0.6225, + "step": 21027 + }, + { + "epoch": 1.4591631392686142, + "grad_norm": 1.838517189025879, + "learning_rate": 0.000146754891396158, + "loss": 0.5666, + "step": 21028 + }, + { + "epoch": 1.459232530705711, + "grad_norm": 2.034209966659546, + "learning_rate": 0.0001467499246505688, + "loss": 0.4658, + "step": 21029 + }, + { + "epoch": 1.4593019221428076, + "grad_norm": 1.3443818092346191, + "learning_rate": 0.00014674495775739548, + "loss": 0.5314, + "step": 21030 + }, + { + "epoch": 1.4593713135799042, + "grad_norm": 2.7189528942108154, + "learning_rate": 0.00014673999071665374, + "loss": 0.6683, + "step": 21031 + }, + { + "epoch": 1.4594407050170008, + "grad_norm": 2.0864970684051514, + "learning_rate": 0.00014673502352835927, + "loss": 0.5897, + "step": 21032 + }, + { + "epoch": 1.4595100964540975, + "grad_norm": 1.4097888469696045, + "learning_rate": 0.00014673005619252776, + "loss": 0.4976, + "step": 21033 + }, + { + "epoch": 1.4595794878911943, + "grad_norm": 1.9133305549621582, + "learning_rate": 0.0001467250887091749, + "loss": 0.7186, + "step": 21034 + }, + { + "epoch": 1.459648879328291, + "grad_norm": 2.015803575515747, + "learning_rate": 0.00014672012107831635, + "loss": 0.5881, + "step": 21035 + }, + { + "epoch": 1.4597182707653875, + "grad_norm": 2.323425769805908, + "learning_rate": 0.00014671515329996778, + "loss": 0.5653, + "step": 21036 + }, + { + "epoch": 1.4597876622024843, + "grad_norm": 1.6214319467544556, + "learning_rate": 0.0001467101853741449, + "loss": 0.6145, + "step": 21037 + }, + { + "epoch": 1.4598570536395807, + "grad_norm": 3.1335456371307373, + "learning_rate": 0.0001467052173008634, + "loss": 0.4603, + "step": 21038 + }, + { + "epoch": 1.4599264450766776, + "grad_norm": 2.0322937965393066, + "learning_rate": 0.00014670024908013893, + "loss": 0.9714, + "step": 21039 + }, + { + "epoch": 1.4599958365137742, + "grad_norm": 1.859729290008545, + "learning_rate": 0.0001466952807119872, + "loss": 0.6773, + "step": 21040 + }, + { + "epoch": 1.4600652279508708, + "grad_norm": 1.6911060810089111, + "learning_rate": 0.00014669031219642387, + "loss": 0.6201, + "step": 21041 + }, + { + "epoch": 1.4601346193879676, + "grad_norm": 2.2418296337127686, + "learning_rate": 0.00014668534353346466, + "loss": 0.6356, + "step": 21042 + }, + { + "epoch": 1.4602040108250642, + "grad_norm": 1.7586948871612549, + "learning_rate": 0.00014668037472312522, + "loss": 0.5762, + "step": 21043 + }, + { + "epoch": 1.4602734022621608, + "grad_norm": 1.882353663444519, + "learning_rate": 0.0001466754057654213, + "loss": 0.6047, + "step": 21044 + }, + { + "epoch": 1.4603427936992575, + "grad_norm": 1.8662500381469727, + "learning_rate": 0.00014667043666036848, + "loss": 0.7137, + "step": 21045 + }, + { + "epoch": 1.460412185136354, + "grad_norm": 2.857419490814209, + "learning_rate": 0.00014666546740798255, + "loss": 0.5555, + "step": 21046 + }, + { + "epoch": 1.460481576573451, + "grad_norm": 1.9715486764907837, + "learning_rate": 0.00014666049800827913, + "loss": 0.655, + "step": 21047 + }, + { + "epoch": 1.4605509680105475, + "grad_norm": 2.1778016090393066, + "learning_rate": 0.00014665552846127394, + "loss": 0.6832, + "step": 21048 + }, + { + "epoch": 1.4606203594476441, + "grad_norm": 1.5229835510253906, + "learning_rate": 0.00014665055876698265, + "loss": 0.521, + "step": 21049 + }, + { + "epoch": 1.460689750884741, + "grad_norm": 1.4659682512283325, + "learning_rate": 0.00014664558892542095, + "loss": 0.4653, + "step": 21050 + }, + { + "epoch": 1.4607591423218376, + "grad_norm": 1.456419587135315, + "learning_rate": 0.00014664061893660454, + "loss": 0.5214, + "step": 21051 + }, + { + "epoch": 1.4608285337589342, + "grad_norm": 2.030789375305176, + "learning_rate": 0.00014663564880054914, + "loss": 0.5027, + "step": 21052 + }, + { + "epoch": 1.4608979251960308, + "grad_norm": 2.4683024883270264, + "learning_rate": 0.00014663067851727036, + "loss": 0.81, + "step": 21053 + }, + { + "epoch": 1.4609673166331274, + "grad_norm": 2.4122300148010254, + "learning_rate": 0.00014662570808678398, + "loss": 0.6553, + "step": 21054 + }, + { + "epoch": 1.4610367080702242, + "grad_norm": 1.8264352083206177, + "learning_rate": 0.00014662073750910563, + "loss": 0.6071, + "step": 21055 + }, + { + "epoch": 1.4611060995073208, + "grad_norm": 1.6301164627075195, + "learning_rate": 0.000146615766784251, + "loss": 0.5489, + "step": 21056 + }, + { + "epoch": 1.4611754909444175, + "grad_norm": 1.9530341625213623, + "learning_rate": 0.00014661079591223587, + "loss": 0.6297, + "step": 21057 + }, + { + "epoch": 1.461244882381514, + "grad_norm": 1.648749828338623, + "learning_rate": 0.00014660582489307579, + "loss": 0.6146, + "step": 21058 + }, + { + "epoch": 1.4613142738186107, + "grad_norm": 1.9582839012145996, + "learning_rate": 0.00014660085372678657, + "loss": 0.7223, + "step": 21059 + }, + { + "epoch": 1.4613836652557075, + "grad_norm": 2.078392744064331, + "learning_rate": 0.00014659588241338383, + "loss": 0.4643, + "step": 21060 + }, + { + "epoch": 1.4614530566928041, + "grad_norm": 1.3571299314498901, + "learning_rate": 0.00014659091095288332, + "loss": 0.414, + "step": 21061 + }, + { + "epoch": 1.4615224481299007, + "grad_norm": 1.5594968795776367, + "learning_rate": 0.00014658593934530073, + "loss": 0.48, + "step": 21062 + }, + { + "epoch": 1.4615918395669976, + "grad_norm": 1.7803726196289062, + "learning_rate": 0.0001465809675906517, + "loss": 0.5048, + "step": 21063 + }, + { + "epoch": 1.4616612310040942, + "grad_norm": 2.6602110862731934, + "learning_rate": 0.000146575995688952, + "loss": 0.8577, + "step": 21064 + }, + { + "epoch": 1.4617306224411908, + "grad_norm": 1.635736346244812, + "learning_rate": 0.00014657102364021721, + "loss": 0.6881, + "step": 21065 + }, + { + "epoch": 1.4618000138782874, + "grad_norm": 2.3727798461914062, + "learning_rate": 0.00014656605144446318, + "loss": 0.6365, + "step": 21066 + }, + { + "epoch": 1.461869405315384, + "grad_norm": 2.2161426544189453, + "learning_rate": 0.0001465610791017055, + "loss": 0.7461, + "step": 21067 + }, + { + "epoch": 1.4619387967524808, + "grad_norm": 1.8853111267089844, + "learning_rate": 0.0001465561066119599, + "loss": 0.7113, + "step": 21068 + }, + { + "epoch": 1.4620081881895775, + "grad_norm": 3.4068477153778076, + "learning_rate": 0.00014655113397524206, + "loss": 0.8552, + "step": 21069 + }, + { + "epoch": 1.462077579626674, + "grad_norm": 2.4226083755493164, + "learning_rate": 0.00014654616119156766, + "loss": 0.7129, + "step": 21070 + }, + { + "epoch": 1.4621469710637707, + "grad_norm": 1.7253518104553223, + "learning_rate": 0.00014654118826095246, + "loss": 0.8016, + "step": 21071 + }, + { + "epoch": 1.4622163625008673, + "grad_norm": 1.470850944519043, + "learning_rate": 0.00014653621518341214, + "loss": 0.6468, + "step": 21072 + }, + { + "epoch": 1.4622857539379641, + "grad_norm": 2.1735146045684814, + "learning_rate": 0.0001465312419589624, + "loss": 0.6597, + "step": 21073 + }, + { + "epoch": 1.4623551453750607, + "grad_norm": 1.7169520854949951, + "learning_rate": 0.00014652626858761887, + "loss": 0.5229, + "step": 21074 + }, + { + "epoch": 1.4624245368121573, + "grad_norm": 1.8345242738723755, + "learning_rate": 0.00014652129506939734, + "loss": 0.7879, + "step": 21075 + }, + { + "epoch": 1.4624939282492542, + "grad_norm": 2.8404555320739746, + "learning_rate": 0.00014651632140431353, + "loss": 0.8011, + "step": 21076 + }, + { + "epoch": 1.4625633196863508, + "grad_norm": 2.0966148376464844, + "learning_rate": 0.00014651134759238304, + "loss": 0.9136, + "step": 21077 + }, + { + "epoch": 1.4626327111234474, + "grad_norm": 2.136206865310669, + "learning_rate": 0.00014650637363362162, + "loss": 0.8728, + "step": 21078 + }, + { + "epoch": 1.462702102560544, + "grad_norm": 1.7209134101867676, + "learning_rate": 0.000146501399528045, + "loss": 0.5796, + "step": 21079 + }, + { + "epoch": 1.4627714939976406, + "grad_norm": 1.889438271522522, + "learning_rate": 0.00014649642527566882, + "loss": 0.5301, + "step": 21080 + }, + { + "epoch": 1.4628408854347374, + "grad_norm": 1.5778486728668213, + "learning_rate": 0.00014649145087650888, + "loss": 0.7112, + "step": 21081 + }, + { + "epoch": 1.462910276871834, + "grad_norm": 1.920558214187622, + "learning_rate": 0.00014648647633058075, + "loss": 0.5372, + "step": 21082 + }, + { + "epoch": 1.4629796683089307, + "grad_norm": 2.096233367919922, + "learning_rate": 0.00014648150163790025, + "loss": 0.7031, + "step": 21083 + }, + { + "epoch": 1.4630490597460273, + "grad_norm": 1.3468515872955322, + "learning_rate": 0.00014647652679848302, + "loss": 0.5689, + "step": 21084 + }, + { + "epoch": 1.463118451183124, + "grad_norm": 1.3313066959381104, + "learning_rate": 0.0001464715518123448, + "loss": 0.5354, + "step": 21085 + }, + { + "epoch": 1.4631878426202207, + "grad_norm": 1.7723809480667114, + "learning_rate": 0.0001464665766795013, + "loss": 0.7109, + "step": 21086 + }, + { + "epoch": 1.4632572340573173, + "grad_norm": 1.746073603630066, + "learning_rate": 0.00014646160139996817, + "loss": 0.6323, + "step": 21087 + }, + { + "epoch": 1.463326625494414, + "grad_norm": 3.370293378829956, + "learning_rate": 0.00014645662597376118, + "loss": 0.6189, + "step": 21088 + }, + { + "epoch": 1.4633960169315108, + "grad_norm": 1.6236799955368042, + "learning_rate": 0.000146451650400896, + "loss": 0.5511, + "step": 21089 + }, + { + "epoch": 1.4634654083686074, + "grad_norm": 1.5415773391723633, + "learning_rate": 0.00014644667468138837, + "loss": 0.5424, + "step": 21090 + }, + { + "epoch": 1.463534799805704, + "grad_norm": 1.4222209453582764, + "learning_rate": 0.00014644169881525396, + "loss": 0.5311, + "step": 21091 + }, + { + "epoch": 1.4636041912428006, + "grad_norm": 2.670297861099243, + "learning_rate": 0.0001464367228025085, + "loss": 0.8012, + "step": 21092 + }, + { + "epoch": 1.4636735826798972, + "grad_norm": 2.5165226459503174, + "learning_rate": 0.0001464317466431677, + "loss": 0.6867, + "step": 21093 + }, + { + "epoch": 1.463742974116994, + "grad_norm": 2.1637520790100098, + "learning_rate": 0.00014642677033724724, + "loss": 0.6012, + "step": 21094 + }, + { + "epoch": 1.4638123655540907, + "grad_norm": 2.231226682662964, + "learning_rate": 0.00014642179388476287, + "loss": 0.6509, + "step": 21095 + }, + { + "epoch": 1.4638817569911873, + "grad_norm": 2.8815014362335205, + "learning_rate": 0.00014641681728573026, + "loss": 0.5399, + "step": 21096 + }, + { + "epoch": 1.4639511484282839, + "grad_norm": 2.1722354888916016, + "learning_rate": 0.00014641184054016518, + "loss": 0.817, + "step": 21097 + }, + { + "epoch": 1.4640205398653805, + "grad_norm": 1.3011789321899414, + "learning_rate": 0.00014640686364808333, + "loss": 0.3832, + "step": 21098 + }, + { + "epoch": 1.4640899313024773, + "grad_norm": 1.6216846704483032, + "learning_rate": 0.0001464018866095003, + "loss": 0.5362, + "step": 21099 + }, + { + "epoch": 1.464159322739574, + "grad_norm": 1.4632558822631836, + "learning_rate": 0.00014639690942443197, + "loss": 0.63, + "step": 21100 + }, + { + "epoch": 1.4642287141766706, + "grad_norm": 2.020350456237793, + "learning_rate": 0.00014639193209289397, + "loss": 0.6191, + "step": 21101 + }, + { + "epoch": 1.4642981056137674, + "grad_norm": 3.0569729804992676, + "learning_rate": 0.00014638695461490202, + "loss": 0.7973, + "step": 21102 + }, + { + "epoch": 1.464367497050864, + "grad_norm": 2.4370217323303223, + "learning_rate": 0.0001463819769904718, + "loss": 0.8763, + "step": 21103 + }, + { + "epoch": 1.4644368884879606, + "grad_norm": 1.7286674976348877, + "learning_rate": 0.0001463769992196191, + "loss": 0.6004, + "step": 21104 + }, + { + "epoch": 1.4645062799250572, + "grad_norm": 2.0348806381225586, + "learning_rate": 0.0001463720213023596, + "loss": 0.7605, + "step": 21105 + }, + { + "epoch": 1.4645756713621538, + "grad_norm": 1.7818772792816162, + "learning_rate": 0.00014636704323870898, + "loss": 0.6505, + "step": 21106 + }, + { + "epoch": 1.4646450627992507, + "grad_norm": 1.948950171470642, + "learning_rate": 0.000146362065028683, + "loss": 0.6214, + "step": 21107 + }, + { + "epoch": 1.4647144542363473, + "grad_norm": 4.702037811279297, + "learning_rate": 0.00014635708667229736, + "loss": 0.7429, + "step": 21108 + }, + { + "epoch": 1.4647838456734439, + "grad_norm": 1.5345677137374878, + "learning_rate": 0.00014635210816956774, + "loss": 0.457, + "step": 21109 + }, + { + "epoch": 1.4648532371105405, + "grad_norm": 1.9144296646118164, + "learning_rate": 0.00014634712952050992, + "loss": 0.6683, + "step": 21110 + }, + { + "epoch": 1.464922628547637, + "grad_norm": 2.589405059814453, + "learning_rate": 0.0001463421507251396, + "loss": 0.9058, + "step": 21111 + }, + { + "epoch": 1.464992019984734, + "grad_norm": 1.7148610353469849, + "learning_rate": 0.0001463371717834725, + "loss": 0.4239, + "step": 21112 + }, + { + "epoch": 1.4650614114218306, + "grad_norm": 1.52851140499115, + "learning_rate": 0.0001463321926955243, + "loss": 0.4751, + "step": 21113 + }, + { + "epoch": 1.4651308028589272, + "grad_norm": 1.7186251878738403, + "learning_rate": 0.00014632721346131074, + "loss": 0.5413, + "step": 21114 + }, + { + "epoch": 1.465200194296024, + "grad_norm": 1.399124026298523, + "learning_rate": 0.00014632223408084756, + "loss": 0.4803, + "step": 21115 + }, + { + "epoch": 1.4652695857331206, + "grad_norm": 1.7641468048095703, + "learning_rate": 0.00014631725455415047, + "loss": 0.701, + "step": 21116 + }, + { + "epoch": 1.4653389771702172, + "grad_norm": 1.9621623754501343, + "learning_rate": 0.00014631227488123516, + "loss": 0.7553, + "step": 21117 + }, + { + "epoch": 1.4654083686073138, + "grad_norm": 1.615972876548767, + "learning_rate": 0.00014630729506211738, + "loss": 0.6146, + "step": 21118 + }, + { + "epoch": 1.4654777600444104, + "grad_norm": 2.3743245601654053, + "learning_rate": 0.00014630231509681283, + "loss": 0.6538, + "step": 21119 + }, + { + "epoch": 1.4655471514815073, + "grad_norm": 1.96566641330719, + "learning_rate": 0.00014629733498533725, + "loss": 0.5223, + "step": 21120 + }, + { + "epoch": 1.4656165429186039, + "grad_norm": 1.589331030845642, + "learning_rate": 0.00014629235472770637, + "loss": 0.5246, + "step": 21121 + }, + { + "epoch": 1.4656859343557005, + "grad_norm": 2.2041280269622803, + "learning_rate": 0.00014628737432393588, + "loss": 0.817, + "step": 21122 + }, + { + "epoch": 1.465755325792797, + "grad_norm": 2.463226556777954, + "learning_rate": 0.0001462823937740415, + "loss": 0.7171, + "step": 21123 + }, + { + "epoch": 1.4658247172298937, + "grad_norm": 1.1629301309585571, + "learning_rate": 0.000146277413078039, + "loss": 0.6274, + "step": 21124 + }, + { + "epoch": 1.4658941086669905, + "grad_norm": 2.1756675243377686, + "learning_rate": 0.00014627243223594406, + "loss": 0.8299, + "step": 21125 + }, + { + "epoch": 1.4659635001040872, + "grad_norm": 1.8305375576019287, + "learning_rate": 0.00014626745124777244, + "loss": 0.5642, + "step": 21126 + }, + { + "epoch": 1.4660328915411838, + "grad_norm": 2.677316665649414, + "learning_rate": 0.00014626247011353982, + "loss": 0.8696, + "step": 21127 + }, + { + "epoch": 1.4661022829782806, + "grad_norm": 1.8127408027648926, + "learning_rate": 0.000146257488833262, + "loss": 0.7798, + "step": 21128 + }, + { + "epoch": 1.4661716744153772, + "grad_norm": 1.6076507568359375, + "learning_rate": 0.0001462525074069546, + "loss": 0.756, + "step": 21129 + }, + { + "epoch": 1.4662410658524738, + "grad_norm": 2.2558233737945557, + "learning_rate": 0.00014624752583463342, + "loss": 0.9521, + "step": 21130 + }, + { + "epoch": 1.4663104572895704, + "grad_norm": 1.7207081317901611, + "learning_rate": 0.00014624254411631417, + "loss": 0.5165, + "step": 21131 + }, + { + "epoch": 1.466379848726667, + "grad_norm": 2.022799015045166, + "learning_rate": 0.00014623756225201254, + "loss": 0.6296, + "step": 21132 + }, + { + "epoch": 1.4664492401637639, + "grad_norm": 2.4907631874084473, + "learning_rate": 0.00014623258024174435, + "loss": 0.7105, + "step": 21133 + }, + { + "epoch": 1.4665186316008605, + "grad_norm": 1.9703171253204346, + "learning_rate": 0.00014622759808552522, + "loss": 0.7402, + "step": 21134 + }, + { + "epoch": 1.466588023037957, + "grad_norm": 2.0319268703460693, + "learning_rate": 0.000146222615783371, + "loss": 0.7094, + "step": 21135 + }, + { + "epoch": 1.4666574144750537, + "grad_norm": 1.8537250757217407, + "learning_rate": 0.00014621763333529726, + "loss": 0.5782, + "step": 21136 + }, + { + "epoch": 1.4667268059121503, + "grad_norm": 1.7034649848937988, + "learning_rate": 0.00014621265074131983, + "loss": 0.4487, + "step": 21137 + }, + { + "epoch": 1.4667961973492472, + "grad_norm": 2.0522167682647705, + "learning_rate": 0.00014620766800145445, + "loss": 0.6513, + "step": 21138 + }, + { + "epoch": 1.4668655887863438, + "grad_norm": 2.1305837631225586, + "learning_rate": 0.0001462026851157168, + "loss": 0.6539, + "step": 21139 + }, + { + "epoch": 1.4669349802234404, + "grad_norm": 5.474754333496094, + "learning_rate": 0.00014619770208412262, + "loss": 0.8324, + "step": 21140 + }, + { + "epoch": 1.4670043716605372, + "grad_norm": 2.0691468715667725, + "learning_rate": 0.0001461927189066877, + "loss": 0.6878, + "step": 21141 + }, + { + "epoch": 1.4670737630976338, + "grad_norm": 2.2244739532470703, + "learning_rate": 0.0001461877355834277, + "loss": 0.6251, + "step": 21142 + }, + { + "epoch": 1.4671431545347304, + "grad_norm": 3.462613582611084, + "learning_rate": 0.0001461827521143584, + "loss": 0.931, + "step": 21143 + }, + { + "epoch": 1.467212545971827, + "grad_norm": 1.9471451044082642, + "learning_rate": 0.00014617776849949548, + "loss": 0.7767, + "step": 21144 + }, + { + "epoch": 1.4672819374089237, + "grad_norm": 4.3993072509765625, + "learning_rate": 0.00014617278473885473, + "loss": 0.6609, + "step": 21145 + }, + { + "epoch": 1.4673513288460205, + "grad_norm": 2.3401172161102295, + "learning_rate": 0.0001461678008324518, + "loss": 0.6657, + "step": 21146 + }, + { + "epoch": 1.467420720283117, + "grad_norm": 2.1998910903930664, + "learning_rate": 0.00014616281678030252, + "loss": 0.6913, + "step": 21147 + }, + { + "epoch": 1.4674901117202137, + "grad_norm": 3.0609006881713867, + "learning_rate": 0.00014615783258242258, + "loss": 0.534, + "step": 21148 + }, + { + "epoch": 1.4675595031573103, + "grad_norm": 2.223700523376465, + "learning_rate": 0.00014615284823882774, + "loss": 0.6997, + "step": 21149 + }, + { + "epoch": 1.467628894594407, + "grad_norm": 2.087763786315918, + "learning_rate": 0.00014614786374953366, + "loss": 0.8141, + "step": 21150 + }, + { + "epoch": 1.4676982860315038, + "grad_norm": 2.3861870765686035, + "learning_rate": 0.00014614287911455617, + "loss": 0.8157, + "step": 21151 + }, + { + "epoch": 1.4677676774686004, + "grad_norm": 1.8853331804275513, + "learning_rate": 0.00014613789433391098, + "loss": 0.551, + "step": 21152 + }, + { + "epoch": 1.467837068905697, + "grad_norm": 1.877318263053894, + "learning_rate": 0.00014613290940761376, + "loss": 0.7557, + "step": 21153 + }, + { + "epoch": 1.4679064603427938, + "grad_norm": 1.7712669372558594, + "learning_rate": 0.00014612792433568034, + "loss": 0.5689, + "step": 21154 + }, + { + "epoch": 1.4679758517798904, + "grad_norm": 1.835761308670044, + "learning_rate": 0.00014612293911812638, + "loss": 0.5547, + "step": 21155 + }, + { + "epoch": 1.468045243216987, + "grad_norm": 2.2340283393859863, + "learning_rate": 0.0001461179537549677, + "loss": 0.7059, + "step": 21156 + }, + { + "epoch": 1.4681146346540837, + "grad_norm": 1.6961473226547241, + "learning_rate": 0.00014611296824621994, + "loss": 0.655, + "step": 21157 + }, + { + "epoch": 1.4681840260911803, + "grad_norm": 2.342046022415161, + "learning_rate": 0.0001461079825918989, + "loss": 0.7564, + "step": 21158 + }, + { + "epoch": 1.468253417528277, + "grad_norm": 1.947138786315918, + "learning_rate": 0.00014610299679202032, + "loss": 0.639, + "step": 21159 + }, + { + "epoch": 1.4683228089653737, + "grad_norm": 3.445906162261963, + "learning_rate": 0.00014609801084659993, + "loss": 0.6575, + "step": 21160 + }, + { + "epoch": 1.4683922004024703, + "grad_norm": 2.8740158081054688, + "learning_rate": 0.00014609302475565344, + "loss": 0.6975, + "step": 21161 + }, + { + "epoch": 1.468461591839567, + "grad_norm": 1.9309930801391602, + "learning_rate": 0.00014608803851919665, + "loss": 0.7743, + "step": 21162 + }, + { + "epoch": 1.4685309832766635, + "grad_norm": 1.7072633504867554, + "learning_rate": 0.0001460830521372453, + "loss": 0.6266, + "step": 21163 + }, + { + "epoch": 1.4686003747137604, + "grad_norm": 2.121976375579834, + "learning_rate": 0.00014607806560981503, + "loss": 0.5535, + "step": 21164 + }, + { + "epoch": 1.468669766150857, + "grad_norm": 2.3683130741119385, + "learning_rate": 0.00014607307893692168, + "loss": 0.764, + "step": 21165 + }, + { + "epoch": 1.4687391575879536, + "grad_norm": 3.534557342529297, + "learning_rate": 0.00014606809211858098, + "loss": 0.5976, + "step": 21166 + }, + { + "epoch": 1.4688085490250504, + "grad_norm": 1.5502530336380005, + "learning_rate": 0.00014606310515480865, + "loss": 0.5449, + "step": 21167 + }, + { + "epoch": 1.468877940462147, + "grad_norm": 2.7573416233062744, + "learning_rate": 0.00014605811804562047, + "loss": 0.6924, + "step": 21168 + }, + { + "epoch": 1.4689473318992436, + "grad_norm": 12.733675003051758, + "learning_rate": 0.0001460531307910321, + "loss": 0.6435, + "step": 21169 + }, + { + "epoch": 1.4690167233363403, + "grad_norm": 1.2592922449111938, + "learning_rate": 0.00014604814339105938, + "loss": 0.5069, + "step": 21170 + }, + { + "epoch": 1.4690861147734369, + "grad_norm": 1.4280967712402344, + "learning_rate": 0.000146043155845718, + "loss": 0.4499, + "step": 21171 + }, + { + "epoch": 1.4691555062105337, + "grad_norm": 2.417968511581421, + "learning_rate": 0.0001460381681550237, + "loss": 0.5711, + "step": 21172 + }, + { + "epoch": 1.4692248976476303, + "grad_norm": 1.8637893199920654, + "learning_rate": 0.00014603318031899227, + "loss": 0.7325, + "step": 21173 + }, + { + "epoch": 1.469294289084727, + "grad_norm": 2.2861170768737793, + "learning_rate": 0.00014602819233763942, + "loss": 0.9121, + "step": 21174 + }, + { + "epoch": 1.4693636805218235, + "grad_norm": 2.6283271312713623, + "learning_rate": 0.0001460232042109809, + "loss": 0.7446, + "step": 21175 + }, + { + "epoch": 1.4694330719589201, + "grad_norm": 2.014394998550415, + "learning_rate": 0.0001460182159390325, + "loss": 0.7588, + "step": 21176 + }, + { + "epoch": 1.469502463396017, + "grad_norm": 2.0727427005767822, + "learning_rate": 0.0001460132275218099, + "loss": 0.6209, + "step": 21177 + }, + { + "epoch": 1.4695718548331136, + "grad_norm": 2.471041440963745, + "learning_rate": 0.00014600823895932888, + "loss": 0.6555, + "step": 21178 + }, + { + "epoch": 1.4696412462702102, + "grad_norm": 1.9065682888031006, + "learning_rate": 0.00014600325025160518, + "loss": 0.6616, + "step": 21179 + }, + { + "epoch": 1.469710637707307, + "grad_norm": 1.7811589241027832, + "learning_rate": 0.0001459982613986546, + "loss": 0.7719, + "step": 21180 + }, + { + "epoch": 1.4697800291444036, + "grad_norm": 1.9942412376403809, + "learning_rate": 0.00014599327240049278, + "loss": 0.7684, + "step": 21181 + }, + { + "epoch": 1.4698494205815003, + "grad_norm": 1.5125675201416016, + "learning_rate": 0.00014598828325713557, + "loss": 0.6105, + "step": 21182 + }, + { + "epoch": 1.4699188120185969, + "grad_norm": 1.3607733249664307, + "learning_rate": 0.00014598329396859868, + "loss": 0.6447, + "step": 21183 + }, + { + "epoch": 1.4699882034556935, + "grad_norm": 2.4806067943573, + "learning_rate": 0.00014597830453489787, + "loss": 0.8222, + "step": 21184 + }, + { + "epoch": 1.4700575948927903, + "grad_norm": 2.3245644569396973, + "learning_rate": 0.00014597331495604888, + "loss": 0.6447, + "step": 21185 + }, + { + "epoch": 1.470126986329887, + "grad_norm": 2.8204400539398193, + "learning_rate": 0.00014596832523206752, + "loss": 0.7428, + "step": 21186 + }, + { + "epoch": 1.4701963777669835, + "grad_norm": 1.6275250911712646, + "learning_rate": 0.00014596333536296945, + "loss": 0.8703, + "step": 21187 + }, + { + "epoch": 1.4702657692040801, + "grad_norm": 2.286815643310547, + "learning_rate": 0.00014595834534877046, + "loss": 0.7717, + "step": 21188 + }, + { + "epoch": 1.4703351606411768, + "grad_norm": 1.9057471752166748, + "learning_rate": 0.00014595335518948627, + "loss": 0.5553, + "step": 21189 + }, + { + "epoch": 1.4704045520782736, + "grad_norm": 1.9123499393463135, + "learning_rate": 0.00014594836488513273, + "loss": 0.672, + "step": 21190 + }, + { + "epoch": 1.4704739435153702, + "grad_norm": 2.2682812213897705, + "learning_rate": 0.0001459433744357255, + "loss": 0.6927, + "step": 21191 + }, + { + "epoch": 1.4705433349524668, + "grad_norm": 2.190863847732544, + "learning_rate": 0.00014593838384128037, + "loss": 0.6777, + "step": 21192 + }, + { + "epoch": 1.4706127263895636, + "grad_norm": 1.410244345664978, + "learning_rate": 0.0001459333931018131, + "loss": 0.5909, + "step": 21193 + }, + { + "epoch": 1.4706821178266603, + "grad_norm": 1.980581521987915, + "learning_rate": 0.00014592840221733943, + "loss": 0.8133, + "step": 21194 + }, + { + "epoch": 1.4707515092637569, + "grad_norm": 2.0255990028381348, + "learning_rate": 0.00014592341118787515, + "loss": 0.5766, + "step": 21195 + }, + { + "epoch": 1.4708209007008535, + "grad_norm": 1.8695170879364014, + "learning_rate": 0.00014591842001343593, + "loss": 0.6758, + "step": 21196 + }, + { + "epoch": 1.47089029213795, + "grad_norm": 1.834130883216858, + "learning_rate": 0.00014591342869403766, + "loss": 0.8375, + "step": 21197 + }, + { + "epoch": 1.470959683575047, + "grad_norm": 1.9083927869796753, + "learning_rate": 0.00014590843722969596, + "loss": 0.6179, + "step": 21198 + }, + { + "epoch": 1.4710290750121435, + "grad_norm": 1.8581178188323975, + "learning_rate": 0.00014590344562042667, + "loss": 0.7158, + "step": 21199 + }, + { + "epoch": 1.4710984664492401, + "grad_norm": 1.4763031005859375, + "learning_rate": 0.00014589845386624552, + "loss": 0.4907, + "step": 21200 + }, + { + "epoch": 1.4711678578863367, + "grad_norm": 1.5653594732284546, + "learning_rate": 0.00014589346196716832, + "loss": 0.7981, + "step": 21201 + }, + { + "epoch": 1.4712372493234334, + "grad_norm": 1.7266231775283813, + "learning_rate": 0.00014588846992321072, + "loss": 0.6561, + "step": 21202 + }, + { + "epoch": 1.4713066407605302, + "grad_norm": 1.4061572551727295, + "learning_rate": 0.00014588347773438861, + "loss": 0.491, + "step": 21203 + }, + { + "epoch": 1.4713760321976268, + "grad_norm": 1.9159637689590454, + "learning_rate": 0.00014587848540071764, + "loss": 0.692, + "step": 21204 + }, + { + "epoch": 1.4714454236347234, + "grad_norm": 1.3739290237426758, + "learning_rate": 0.00014587349292221363, + "loss": 0.3563, + "step": 21205 + }, + { + "epoch": 1.4715148150718202, + "grad_norm": 1.411228895187378, + "learning_rate": 0.00014586850029889235, + "loss": 0.4272, + "step": 21206 + }, + { + "epoch": 1.4715842065089169, + "grad_norm": 2.0443170070648193, + "learning_rate": 0.00014586350753076948, + "loss": 0.8225, + "step": 21207 + }, + { + "epoch": 1.4716535979460135, + "grad_norm": 2.1140193939208984, + "learning_rate": 0.0001458585146178609, + "loss": 0.8964, + "step": 21208 + }, + { + "epoch": 1.47172298938311, + "grad_norm": 1.2188340425491333, + "learning_rate": 0.00014585352156018226, + "loss": 0.4928, + "step": 21209 + }, + { + "epoch": 1.4717923808202067, + "grad_norm": 1.4601435661315918, + "learning_rate": 0.0001458485283577494, + "loss": 0.5909, + "step": 21210 + }, + { + "epoch": 1.4718617722573035, + "grad_norm": 3.4297118186950684, + "learning_rate": 0.00014584353501057804, + "loss": 0.8563, + "step": 21211 + }, + { + "epoch": 1.4719311636944001, + "grad_norm": 1.4640998840332031, + "learning_rate": 0.00014583854151868395, + "loss": 0.5721, + "step": 21212 + }, + { + "epoch": 1.4720005551314967, + "grad_norm": 1.4547483921051025, + "learning_rate": 0.00014583354788208293, + "loss": 0.5611, + "step": 21213 + }, + { + "epoch": 1.4720699465685934, + "grad_norm": 1.6980609893798828, + "learning_rate": 0.0001458285541007907, + "loss": 0.7416, + "step": 21214 + }, + { + "epoch": 1.47213933800569, + "grad_norm": 1.757510781288147, + "learning_rate": 0.00014582356017482303, + "loss": 0.6376, + "step": 21215 + }, + { + "epoch": 1.4722087294427868, + "grad_norm": 1.500664234161377, + "learning_rate": 0.0001458185661041957, + "loss": 0.7021, + "step": 21216 + }, + { + "epoch": 1.4722781208798834, + "grad_norm": 1.9165626764297485, + "learning_rate": 0.0001458135718889245, + "loss": 0.5873, + "step": 21217 + }, + { + "epoch": 1.47234751231698, + "grad_norm": 1.2741039991378784, + "learning_rate": 0.00014580857752902516, + "loss": 0.4696, + "step": 21218 + }, + { + "epoch": 1.4724169037540769, + "grad_norm": 1.8129507303237915, + "learning_rate": 0.00014580358302451342, + "loss": 0.5548, + "step": 21219 + }, + { + "epoch": 1.4724862951911735, + "grad_norm": 1.941640853881836, + "learning_rate": 0.00014579858837540514, + "loss": 0.6718, + "step": 21220 + }, + { + "epoch": 1.47255568662827, + "grad_norm": 2.0212411880493164, + "learning_rate": 0.00014579359358171597, + "loss": 0.6051, + "step": 21221 + }, + { + "epoch": 1.4726250780653667, + "grad_norm": 1.7411242723464966, + "learning_rate": 0.0001457885986434618, + "loss": 0.6641, + "step": 21222 + }, + { + "epoch": 1.4726944695024633, + "grad_norm": 2.2946813106536865, + "learning_rate": 0.0001457836035606583, + "loss": 0.5586, + "step": 21223 + }, + { + "epoch": 1.4727638609395601, + "grad_norm": 1.8899465799331665, + "learning_rate": 0.0001457786083333213, + "loss": 0.7575, + "step": 21224 + }, + { + "epoch": 1.4728332523766567, + "grad_norm": 1.6718502044677734, + "learning_rate": 0.00014577361296146652, + "loss": 0.4755, + "step": 21225 + }, + { + "epoch": 1.4729026438137534, + "grad_norm": 2.181318998336792, + "learning_rate": 0.00014576861744510975, + "loss": 0.751, + "step": 21226 + }, + { + "epoch": 1.47297203525085, + "grad_norm": 1.3975005149841309, + "learning_rate": 0.00014576362178426677, + "loss": 0.8036, + "step": 21227 + }, + { + "epoch": 1.4730414266879466, + "grad_norm": 1.5784722566604614, + "learning_rate": 0.0001457586259789534, + "loss": 0.7, + "step": 21228 + }, + { + "epoch": 1.4731108181250434, + "grad_norm": 1.5462380647659302, + "learning_rate": 0.00014575363002918527, + "loss": 0.5455, + "step": 21229 + }, + { + "epoch": 1.47318020956214, + "grad_norm": 1.6451846361160278, + "learning_rate": 0.00014574863393497828, + "loss": 0.5766, + "step": 21230 + }, + { + "epoch": 1.4732496009992366, + "grad_norm": 1.8787599802017212, + "learning_rate": 0.00014574363769634814, + "loss": 0.641, + "step": 21231 + }, + { + "epoch": 1.4733189924363335, + "grad_norm": 1.8663579225540161, + "learning_rate": 0.00014573864131331066, + "loss": 0.5518, + "step": 21232 + }, + { + "epoch": 1.47338838387343, + "grad_norm": 1.4147323369979858, + "learning_rate": 0.0001457336447858816, + "loss": 0.7279, + "step": 21233 + }, + { + "epoch": 1.4734577753105267, + "grad_norm": 2.849006175994873, + "learning_rate": 0.00014572864811407673, + "loss": 0.5806, + "step": 21234 + }, + { + "epoch": 1.4735271667476233, + "grad_norm": 1.7720286846160889, + "learning_rate": 0.00014572365129791184, + "loss": 0.6489, + "step": 21235 + }, + { + "epoch": 1.47359655818472, + "grad_norm": 1.7007476091384888, + "learning_rate": 0.00014571865433740265, + "loss": 0.4727, + "step": 21236 + }, + { + "epoch": 1.4736659496218167, + "grad_norm": 1.4859354496002197, + "learning_rate": 0.00014571365723256498, + "loss": 0.5124, + "step": 21237 + }, + { + "epoch": 1.4737353410589134, + "grad_norm": 1.5912636518478394, + "learning_rate": 0.00014570865998341463, + "loss": 0.6203, + "step": 21238 + }, + { + "epoch": 1.47380473249601, + "grad_norm": 2.4276540279388428, + "learning_rate": 0.0001457036625899673, + "loss": 0.5987, + "step": 21239 + }, + { + "epoch": 1.4738741239331066, + "grad_norm": 1.6191705465316772, + "learning_rate": 0.00014569866505223885, + "loss": 0.7039, + "step": 21240 + }, + { + "epoch": 1.4739435153702032, + "grad_norm": 2.136319398880005, + "learning_rate": 0.000145693667370245, + "loss": 0.6616, + "step": 21241 + }, + { + "epoch": 1.4740129068073, + "grad_norm": 3.1498355865478516, + "learning_rate": 0.00014568866954400153, + "loss": 0.8499, + "step": 21242 + }, + { + "epoch": 1.4740822982443966, + "grad_norm": 1.8601012229919434, + "learning_rate": 0.00014568367157352424, + "loss": 0.7723, + "step": 21243 + }, + { + "epoch": 1.4741516896814932, + "grad_norm": 2.0754384994506836, + "learning_rate": 0.0001456786734588289, + "loss": 0.6347, + "step": 21244 + }, + { + "epoch": 1.47422108111859, + "grad_norm": 1.4626803398132324, + "learning_rate": 0.00014567367519993128, + "loss": 0.6357, + "step": 21245 + }, + { + "epoch": 1.4742904725556867, + "grad_norm": 1.5673162937164307, + "learning_rate": 0.0001456686767968472, + "loss": 0.5024, + "step": 21246 + }, + { + "epoch": 1.4743598639927833, + "grad_norm": 2.1714508533477783, + "learning_rate": 0.00014566367824959237, + "loss": 0.5525, + "step": 21247 + }, + { + "epoch": 1.47442925542988, + "grad_norm": 1.4858161211013794, + "learning_rate": 0.0001456586795581826, + "loss": 0.5118, + "step": 21248 + }, + { + "epoch": 1.4744986468669765, + "grad_norm": 1.8261606693267822, + "learning_rate": 0.00014565368072263373, + "loss": 0.6325, + "step": 21249 + }, + { + "epoch": 1.4745680383040733, + "grad_norm": 2.361642837524414, + "learning_rate": 0.0001456486817429614, + "loss": 0.7225, + "step": 21250 + }, + { + "epoch": 1.47463742974117, + "grad_norm": 2.8649277687072754, + "learning_rate": 0.00014564368261918155, + "loss": 0.7587, + "step": 21251 + }, + { + "epoch": 1.4747068211782666, + "grad_norm": 1.7956912517547607, + "learning_rate": 0.00014563868335130988, + "loss": 0.6583, + "step": 21252 + }, + { + "epoch": 1.4747762126153634, + "grad_norm": 2.0609171390533447, + "learning_rate": 0.00014563368393936217, + "loss": 0.6431, + "step": 21253 + }, + { + "epoch": 1.4748456040524598, + "grad_norm": 2.158334970474243, + "learning_rate": 0.0001456286843833542, + "loss": 0.8212, + "step": 21254 + }, + { + "epoch": 1.4749149954895566, + "grad_norm": 2.5419375896453857, + "learning_rate": 0.00014562368468330178, + "loss": 0.5499, + "step": 21255 + }, + { + "epoch": 1.4749843869266532, + "grad_norm": 1.9936597347259521, + "learning_rate": 0.0001456186848392207, + "loss": 0.7069, + "step": 21256 + }, + { + "epoch": 1.4750537783637498, + "grad_norm": 2.3474647998809814, + "learning_rate": 0.00014561368485112669, + "loss": 0.7687, + "step": 21257 + }, + { + "epoch": 1.4751231698008467, + "grad_norm": 1.6847035884857178, + "learning_rate": 0.0001456086847190356, + "loss": 0.555, + "step": 21258 + }, + { + "epoch": 1.4751925612379433, + "grad_norm": 1.9605754613876343, + "learning_rate": 0.00014560368444296317, + "loss": 0.5851, + "step": 21259 + }, + { + "epoch": 1.47526195267504, + "grad_norm": 2.579418420791626, + "learning_rate": 0.0001455986840229252, + "loss": 1.0732, + "step": 21260 + }, + { + "epoch": 1.4753313441121365, + "grad_norm": 1.4552099704742432, + "learning_rate": 0.00014559368345893746, + "loss": 0.4955, + "step": 21261 + }, + { + "epoch": 1.4754007355492331, + "grad_norm": 2.6800341606140137, + "learning_rate": 0.00014558868275101576, + "loss": 0.5659, + "step": 21262 + }, + { + "epoch": 1.47547012698633, + "grad_norm": 1.313935399055481, + "learning_rate": 0.0001455836818991759, + "loss": 0.5181, + "step": 21263 + }, + { + "epoch": 1.4755395184234266, + "grad_norm": 1.6171766519546509, + "learning_rate": 0.00014557868090343362, + "loss": 0.4935, + "step": 21264 + }, + { + "epoch": 1.4756089098605232, + "grad_norm": 1.3350253105163574, + "learning_rate": 0.00014557367976380473, + "loss": 0.5609, + "step": 21265 + }, + { + "epoch": 1.47567830129762, + "grad_norm": 1.7881628274917603, + "learning_rate": 0.00014556867848030503, + "loss": 0.8425, + "step": 21266 + }, + { + "epoch": 1.4757476927347164, + "grad_norm": 1.8801891803741455, + "learning_rate": 0.0001455636770529503, + "loss": 0.4426, + "step": 21267 + }, + { + "epoch": 1.4758170841718132, + "grad_norm": 2.384267807006836, + "learning_rate": 0.00014555867548175634, + "loss": 0.6914, + "step": 21268 + }, + { + "epoch": 1.4758864756089098, + "grad_norm": 1.966555118560791, + "learning_rate": 0.0001455536737667389, + "loss": 0.7714, + "step": 21269 + }, + { + "epoch": 1.4759558670460065, + "grad_norm": 1.7630774974822998, + "learning_rate": 0.0001455486719079138, + "loss": 0.7118, + "step": 21270 + }, + { + "epoch": 1.4760252584831033, + "grad_norm": 1.4490896463394165, + "learning_rate": 0.00014554366990529686, + "loss": 0.4896, + "step": 21271 + }, + { + "epoch": 1.4760946499202, + "grad_norm": 2.1746227741241455, + "learning_rate": 0.00014553866775890382, + "loss": 0.8052, + "step": 21272 + }, + { + "epoch": 1.4761640413572965, + "grad_norm": 1.851007103919983, + "learning_rate": 0.00014553366546875046, + "loss": 0.5761, + "step": 21273 + }, + { + "epoch": 1.4762334327943931, + "grad_norm": 1.9062888622283936, + "learning_rate": 0.00014552866303485265, + "loss": 0.6262, + "step": 21274 + }, + { + "epoch": 1.4763028242314897, + "grad_norm": 1.5060065984725952, + "learning_rate": 0.00014552366045722612, + "loss": 0.5985, + "step": 21275 + }, + { + "epoch": 1.4763722156685866, + "grad_norm": 2.0586750507354736, + "learning_rate": 0.00014551865773588664, + "loss": 0.5877, + "step": 21276 + }, + { + "epoch": 1.4764416071056832, + "grad_norm": 1.5178602933883667, + "learning_rate": 0.0001455136548708501, + "loss": 0.7087, + "step": 21277 + }, + { + "epoch": 1.4765109985427798, + "grad_norm": 2.23559832572937, + "learning_rate": 0.00014550865186213217, + "loss": 0.5754, + "step": 21278 + }, + { + "epoch": 1.4765803899798766, + "grad_norm": 2.5031898021698, + "learning_rate": 0.00014550364870974874, + "loss": 0.8901, + "step": 21279 + }, + { + "epoch": 1.4766497814169732, + "grad_norm": 2.0259909629821777, + "learning_rate": 0.0001454986454137156, + "loss": 0.782, + "step": 21280 + }, + { + "epoch": 1.4767191728540698, + "grad_norm": 2.0820188522338867, + "learning_rate": 0.00014549364197404845, + "loss": 1.019, + "step": 21281 + }, + { + "epoch": 1.4767885642911665, + "grad_norm": 1.4237160682678223, + "learning_rate": 0.00014548863839076317, + "loss": 0.5297, + "step": 21282 + }, + { + "epoch": 1.476857955728263, + "grad_norm": 3.125248908996582, + "learning_rate": 0.00014548363466387556, + "loss": 0.7571, + "step": 21283 + }, + { + "epoch": 1.47692734716536, + "grad_norm": 1.8624398708343506, + "learning_rate": 0.0001454786307934014, + "loss": 0.5786, + "step": 21284 + }, + { + "epoch": 1.4769967386024565, + "grad_norm": 1.1703468561172485, + "learning_rate": 0.00014547362677935644, + "loss": 0.37, + "step": 21285 + }, + { + "epoch": 1.4770661300395531, + "grad_norm": 1.6712747812271118, + "learning_rate": 0.0001454686226217565, + "loss": 0.8409, + "step": 21286 + }, + { + "epoch": 1.4771355214766497, + "grad_norm": 1.7546107769012451, + "learning_rate": 0.00014546361832061744, + "loss": 0.7264, + "step": 21287 + }, + { + "epoch": 1.4772049129137463, + "grad_norm": 1.754343867301941, + "learning_rate": 0.000145458613875955, + "loss": 0.4656, + "step": 21288 + }, + { + "epoch": 1.4772743043508432, + "grad_norm": 2.2032718658447266, + "learning_rate": 0.00014545360928778497, + "loss": 0.5188, + "step": 21289 + }, + { + "epoch": 1.4773436957879398, + "grad_norm": 2.222824811935425, + "learning_rate": 0.00014544860455612321, + "loss": 0.7911, + "step": 21290 + }, + { + "epoch": 1.4774130872250364, + "grad_norm": 2.307544231414795, + "learning_rate": 0.00014544359968098543, + "loss": 0.7875, + "step": 21291 + }, + { + "epoch": 1.4774824786621332, + "grad_norm": 2.7036852836608887, + "learning_rate": 0.00014543859466238754, + "loss": 0.6734, + "step": 21292 + }, + { + "epoch": 1.4775518700992298, + "grad_norm": 1.5940907001495361, + "learning_rate": 0.00014543358950034523, + "loss": 0.6217, + "step": 21293 + }, + { + "epoch": 1.4776212615363264, + "grad_norm": 1.6480902433395386, + "learning_rate": 0.00014542858419487438, + "loss": 0.4213, + "step": 21294 + }, + { + "epoch": 1.477690652973423, + "grad_norm": 1.5591156482696533, + "learning_rate": 0.00014542357874599074, + "loss": 0.4487, + "step": 21295 + }, + { + "epoch": 1.4777600444105197, + "grad_norm": 2.175593376159668, + "learning_rate": 0.00014541857315371012, + "loss": 0.4636, + "step": 21296 + }, + { + "epoch": 1.4778294358476165, + "grad_norm": 1.6742719411849976, + "learning_rate": 0.00014541356741804836, + "loss": 0.4973, + "step": 21297 + }, + { + "epoch": 1.4778988272847131, + "grad_norm": 1.0470893383026123, + "learning_rate": 0.00014540856153902123, + "loss": 0.3489, + "step": 21298 + }, + { + "epoch": 1.4779682187218097, + "grad_norm": 1.846759557723999, + "learning_rate": 0.00014540355551664454, + "loss": 0.5161, + "step": 21299 + }, + { + "epoch": 1.4780376101589063, + "grad_norm": 1.944606900215149, + "learning_rate": 0.00014539854935093405, + "loss": 0.5712, + "step": 21300 + }, + { + "epoch": 1.478107001596003, + "grad_norm": 2.4233908653259277, + "learning_rate": 0.0001453935430419057, + "loss": 0.5497, + "step": 21301 + }, + { + "epoch": 1.4781763930330998, + "grad_norm": 1.6528791189193726, + "learning_rate": 0.0001453885365895751, + "loss": 0.5243, + "step": 21302 + }, + { + "epoch": 1.4782457844701964, + "grad_norm": 1.755560278892517, + "learning_rate": 0.0001453835299939582, + "loss": 0.6917, + "step": 21303 + }, + { + "epoch": 1.478315175907293, + "grad_norm": 2.5551671981811523, + "learning_rate": 0.0001453785232550708, + "loss": 0.8106, + "step": 21304 + }, + { + "epoch": 1.4783845673443898, + "grad_norm": 2.050121545791626, + "learning_rate": 0.00014537351637292862, + "loss": 0.6418, + "step": 21305 + }, + { + "epoch": 1.4784539587814864, + "grad_norm": 2.047767400741577, + "learning_rate": 0.0001453685093475475, + "loss": 0.6166, + "step": 21306 + }, + { + "epoch": 1.478523350218583, + "grad_norm": 2.156475067138672, + "learning_rate": 0.0001453635021789433, + "loss": 0.7538, + "step": 21307 + }, + { + "epoch": 1.4785927416556797, + "grad_norm": 1.6087027788162231, + "learning_rate": 0.00014535849486713178, + "loss": 0.555, + "step": 21308 + }, + { + "epoch": 1.4786621330927763, + "grad_norm": 2.0558269023895264, + "learning_rate": 0.00014535348741212875, + "loss": 0.5672, + "step": 21309 + }, + { + "epoch": 1.478731524529873, + "grad_norm": 1.7662498950958252, + "learning_rate": 0.00014534847981395002, + "loss": 0.7628, + "step": 21310 + }, + { + "epoch": 1.4788009159669697, + "grad_norm": 1.990609049797058, + "learning_rate": 0.0001453434720726114, + "loss": 0.5475, + "step": 21311 + }, + { + "epoch": 1.4788703074040663, + "grad_norm": 2.676636219024658, + "learning_rate": 0.0001453384641881287, + "loss": 0.8334, + "step": 21312 + }, + { + "epoch": 1.478939698841163, + "grad_norm": 1.5255494117736816, + "learning_rate": 0.0001453334561605177, + "loss": 0.5627, + "step": 21313 + }, + { + "epoch": 1.4790090902782596, + "grad_norm": 1.847171425819397, + "learning_rate": 0.00014532844798979426, + "loss": 0.5803, + "step": 21314 + }, + { + "epoch": 1.4790784817153564, + "grad_norm": 1.721801996231079, + "learning_rate": 0.0001453234396759742, + "loss": 0.5963, + "step": 21315 + }, + { + "epoch": 1.479147873152453, + "grad_norm": 1.7512654066085815, + "learning_rate": 0.00014531843121907325, + "loss": 0.545, + "step": 21316 + }, + { + "epoch": 1.4792172645895496, + "grad_norm": 14.583597183227539, + "learning_rate": 0.0001453134226191073, + "loss": 0.6364, + "step": 21317 + }, + { + "epoch": 1.4792866560266464, + "grad_norm": 1.9270702600479126, + "learning_rate": 0.00014530841387609214, + "loss": 0.9386, + "step": 21318 + }, + { + "epoch": 1.479356047463743, + "grad_norm": 1.6650373935699463, + "learning_rate": 0.00014530340499004354, + "loss": 0.5654, + "step": 21319 + }, + { + "epoch": 1.4794254389008397, + "grad_norm": 1.607571005821228, + "learning_rate": 0.00014529839596097736, + "loss": 0.38, + "step": 21320 + }, + { + "epoch": 1.4794948303379363, + "grad_norm": 1.4656556844711304, + "learning_rate": 0.00014529338678890942, + "loss": 0.6315, + "step": 21321 + }, + { + "epoch": 1.4795642217750329, + "grad_norm": 1.5610989332199097, + "learning_rate": 0.00014528837747385549, + "loss": 0.5834, + "step": 21322 + }, + { + "epoch": 1.4796336132121297, + "grad_norm": 2.10042667388916, + "learning_rate": 0.0001452833680158314, + "loss": 0.7278, + "step": 21323 + }, + { + "epoch": 1.4797030046492263, + "grad_norm": 1.7882885932922363, + "learning_rate": 0.000145278358414853, + "loss": 0.6328, + "step": 21324 + }, + { + "epoch": 1.479772396086323, + "grad_norm": 1.2642508745193481, + "learning_rate": 0.00014527334867093608, + "loss": 0.6079, + "step": 21325 + }, + { + "epoch": 1.4798417875234195, + "grad_norm": 1.9585611820220947, + "learning_rate": 0.0001452683387840964, + "loss": 0.7225, + "step": 21326 + }, + { + "epoch": 1.4799111789605162, + "grad_norm": 2.159299850463867, + "learning_rate": 0.00014526332875434988, + "loss": 0.6654, + "step": 21327 + }, + { + "epoch": 1.479980570397613, + "grad_norm": 2.622438907623291, + "learning_rate": 0.00014525831858171225, + "loss": 0.8282, + "step": 21328 + }, + { + "epoch": 1.4800499618347096, + "grad_norm": 2.0386011600494385, + "learning_rate": 0.00014525330826619938, + "loss": 0.5541, + "step": 21329 + }, + { + "epoch": 1.4801193532718062, + "grad_norm": 2.0469913482666016, + "learning_rate": 0.00014524829780782704, + "loss": 0.5756, + "step": 21330 + }, + { + "epoch": 1.480188744708903, + "grad_norm": 2.613060474395752, + "learning_rate": 0.0001452432872066111, + "loss": 0.6926, + "step": 21331 + }, + { + "epoch": 1.4802581361459997, + "grad_norm": 1.6882710456848145, + "learning_rate": 0.00014523827646256736, + "loss": 0.6069, + "step": 21332 + }, + { + "epoch": 1.4803275275830963, + "grad_norm": 2.686190366744995, + "learning_rate": 0.0001452332655757116, + "loss": 0.8602, + "step": 21333 + }, + { + "epoch": 1.4803969190201929, + "grad_norm": 1.7393615245819092, + "learning_rate": 0.00014522825454605967, + "loss": 0.7838, + "step": 21334 + }, + { + "epoch": 1.4804663104572895, + "grad_norm": 1.9702446460723877, + "learning_rate": 0.0001452232433736274, + "loss": 0.7704, + "step": 21335 + }, + { + "epoch": 1.4805357018943863, + "grad_norm": 1.60978102684021, + "learning_rate": 0.0001452182320584306, + "loss": 0.5763, + "step": 21336 + }, + { + "epoch": 1.480605093331483, + "grad_norm": 1.6512174606323242, + "learning_rate": 0.00014521322060048508, + "loss": 0.5333, + "step": 21337 + }, + { + "epoch": 1.4806744847685795, + "grad_norm": 1.5972744226455688, + "learning_rate": 0.00014520820899980666, + "loss": 0.5653, + "step": 21338 + }, + { + "epoch": 1.4807438762056762, + "grad_norm": 1.8946990966796875, + "learning_rate": 0.0001452031972564112, + "loss": 0.7874, + "step": 21339 + }, + { + "epoch": 1.4808132676427728, + "grad_norm": 2.3079404830932617, + "learning_rate": 0.00014519818537031448, + "loss": 1.0215, + "step": 21340 + }, + { + "epoch": 1.4808826590798696, + "grad_norm": 1.6268703937530518, + "learning_rate": 0.0001451931733415323, + "loss": 0.5645, + "step": 21341 + }, + { + "epoch": 1.4809520505169662, + "grad_norm": 1.6807804107666016, + "learning_rate": 0.00014518816117008057, + "loss": 0.6108, + "step": 21342 + }, + { + "epoch": 1.4810214419540628, + "grad_norm": 2.7916364669799805, + "learning_rate": 0.00014518314885597503, + "loss": 0.9129, + "step": 21343 + }, + { + "epoch": 1.4810908333911597, + "grad_norm": 1.5317583084106445, + "learning_rate": 0.00014517813639923152, + "loss": 0.5069, + "step": 21344 + }, + { + "epoch": 1.4811602248282563, + "grad_norm": 2.181804895401001, + "learning_rate": 0.00014517312379986588, + "loss": 0.5093, + "step": 21345 + }, + { + "epoch": 1.4812296162653529, + "grad_norm": 2.2848927974700928, + "learning_rate": 0.00014516811105789394, + "loss": 0.6746, + "step": 21346 + }, + { + "epoch": 1.4812990077024495, + "grad_norm": 3.0954339504241943, + "learning_rate": 0.0001451630981733315, + "loss": 0.7221, + "step": 21347 + }, + { + "epoch": 1.481368399139546, + "grad_norm": 2.3566365242004395, + "learning_rate": 0.0001451580851461944, + "loss": 0.7813, + "step": 21348 + }, + { + "epoch": 1.481437790576643, + "grad_norm": 2.1833629608154297, + "learning_rate": 0.0001451530719764985, + "loss": 0.6535, + "step": 21349 + }, + { + "epoch": 1.4815071820137395, + "grad_norm": 2.02308988571167, + "learning_rate": 0.00014514805866425954, + "loss": 0.6647, + "step": 21350 + }, + { + "epoch": 1.4815765734508362, + "grad_norm": 1.6905632019042969, + "learning_rate": 0.00014514304520949343, + "loss": 0.5727, + "step": 21351 + }, + { + "epoch": 1.4816459648879328, + "grad_norm": 2.5051231384277344, + "learning_rate": 0.00014513803161221594, + "loss": 0.6818, + "step": 21352 + }, + { + "epoch": 1.4817153563250294, + "grad_norm": 1.4934372901916504, + "learning_rate": 0.00014513301787244293, + "loss": 0.5377, + "step": 21353 + }, + { + "epoch": 1.4817847477621262, + "grad_norm": 1.5002942085266113, + "learning_rate": 0.00014512800399019021, + "loss": 0.53, + "step": 21354 + }, + { + "epoch": 1.4818541391992228, + "grad_norm": 1.8811073303222656, + "learning_rate": 0.00014512298996547363, + "loss": 0.867, + "step": 21355 + }, + { + "epoch": 1.4819235306363194, + "grad_norm": 1.7821518182754517, + "learning_rate": 0.00014511797579830902, + "loss": 0.6615, + "step": 21356 + }, + { + "epoch": 1.4819929220734163, + "grad_norm": 1.775634527206421, + "learning_rate": 0.00014511296148871216, + "loss": 0.9612, + "step": 21357 + }, + { + "epoch": 1.4820623135105129, + "grad_norm": 2.094346523284912, + "learning_rate": 0.00014510794703669893, + "loss": 0.5457, + "step": 21358 + }, + { + "epoch": 1.4821317049476095, + "grad_norm": 1.9114766120910645, + "learning_rate": 0.00014510293244228518, + "loss": 0.5952, + "step": 21359 + }, + { + "epoch": 1.482201096384706, + "grad_norm": 1.5268125534057617, + "learning_rate": 0.00014509791770548667, + "loss": 0.6353, + "step": 21360 + }, + { + "epoch": 1.4822704878218027, + "grad_norm": 1.828139305114746, + "learning_rate": 0.00014509290282631928, + "loss": 0.5173, + "step": 21361 + }, + { + "epoch": 1.4823398792588995, + "grad_norm": 1.8014256954193115, + "learning_rate": 0.0001450878878047988, + "loss": 0.7382, + "step": 21362 + }, + { + "epoch": 1.4824092706959962, + "grad_norm": 2.2032318115234375, + "learning_rate": 0.00014508287264094113, + "loss": 0.8053, + "step": 21363 + }, + { + "epoch": 1.4824786621330928, + "grad_norm": 2.80696177482605, + "learning_rate": 0.00014507785733476201, + "loss": 0.7934, + "step": 21364 + }, + { + "epoch": 1.4825480535701894, + "grad_norm": 1.8297744989395142, + "learning_rate": 0.00014507284188627738, + "loss": 0.7725, + "step": 21365 + }, + { + "epoch": 1.482617445007286, + "grad_norm": 1.3444457054138184, + "learning_rate": 0.000145067826295503, + "loss": 0.4573, + "step": 21366 + }, + { + "epoch": 1.4826868364443828, + "grad_norm": 1.861237645149231, + "learning_rate": 0.0001450628105624547, + "loss": 0.7269, + "step": 21367 + }, + { + "epoch": 1.4827562278814794, + "grad_norm": 1.4543594121932983, + "learning_rate": 0.00014505779468714834, + "loss": 0.4523, + "step": 21368 + }, + { + "epoch": 1.482825619318576, + "grad_norm": 1.9073712825775146, + "learning_rate": 0.00014505277866959975, + "loss": 0.4973, + "step": 21369 + }, + { + "epoch": 1.4828950107556729, + "grad_norm": 2.185006856918335, + "learning_rate": 0.0001450477625098248, + "loss": 0.7159, + "step": 21370 + }, + { + "epoch": 1.4829644021927695, + "grad_norm": 3.1860697269439697, + "learning_rate": 0.00014504274620783924, + "loss": 0.8759, + "step": 21371 + }, + { + "epoch": 1.483033793629866, + "grad_norm": 1.8780800104141235, + "learning_rate": 0.00014503772976365897, + "loss": 0.6357, + "step": 21372 + }, + { + "epoch": 1.4831031850669627, + "grad_norm": 2.0166494846343994, + "learning_rate": 0.00014503271317729984, + "loss": 0.7751, + "step": 21373 + }, + { + "epoch": 1.4831725765040593, + "grad_norm": 1.395114541053772, + "learning_rate": 0.00014502769644877766, + "loss": 0.6131, + "step": 21374 + }, + { + "epoch": 1.4832419679411561, + "grad_norm": 1.7133554220199585, + "learning_rate": 0.00014502267957810822, + "loss": 0.5366, + "step": 21375 + }, + { + "epoch": 1.4833113593782528, + "grad_norm": 1.672522783279419, + "learning_rate": 0.0001450176625653074, + "loss": 0.6756, + "step": 21376 + }, + { + "epoch": 1.4833807508153494, + "grad_norm": 1.638616681098938, + "learning_rate": 0.0001450126454103911, + "loss": 0.7521, + "step": 21377 + }, + { + "epoch": 1.483450142252446, + "grad_norm": 5.124837875366211, + "learning_rate": 0.00014500762811337506, + "loss": 0.9106, + "step": 21378 + }, + { + "epoch": 1.4835195336895426, + "grad_norm": 2.4745094776153564, + "learning_rate": 0.0001450026106742752, + "loss": 0.5942, + "step": 21379 + }, + { + "epoch": 1.4835889251266394, + "grad_norm": 1.4479854106903076, + "learning_rate": 0.0001449975930931073, + "loss": 0.6443, + "step": 21380 + }, + { + "epoch": 1.483658316563736, + "grad_norm": 1.650774359703064, + "learning_rate": 0.00014499257536988722, + "loss": 0.6817, + "step": 21381 + }, + { + "epoch": 1.4837277080008326, + "grad_norm": 2.351097822189331, + "learning_rate": 0.00014498755750463077, + "loss": 0.6891, + "step": 21382 + }, + { + "epoch": 1.4837970994379295, + "grad_norm": 1.5392134189605713, + "learning_rate": 0.00014498253949735384, + "loss": 0.6977, + "step": 21383 + }, + { + "epoch": 1.483866490875026, + "grad_norm": 1.7928109169006348, + "learning_rate": 0.0001449775213480723, + "loss": 0.6921, + "step": 21384 + }, + { + "epoch": 1.4839358823121227, + "grad_norm": 1.6480727195739746, + "learning_rate": 0.00014497250305680188, + "loss": 0.6488, + "step": 21385 + }, + { + "epoch": 1.4840052737492193, + "grad_norm": 2.5433740615844727, + "learning_rate": 0.00014496748462355848, + "loss": 0.8304, + "step": 21386 + }, + { + "epoch": 1.484074665186316, + "grad_norm": 2.007749557495117, + "learning_rate": 0.000144962466048358, + "loss": 0.728, + "step": 21387 + }, + { + "epoch": 1.4841440566234128, + "grad_norm": 2.0512490272521973, + "learning_rate": 0.00014495744733121618, + "loss": 0.7054, + "step": 21388 + }, + { + "epoch": 1.4842134480605094, + "grad_norm": 2.1177587509155273, + "learning_rate": 0.00014495242847214893, + "loss": 0.6222, + "step": 21389 + }, + { + "epoch": 1.484282839497606, + "grad_norm": 1.9084446430206299, + "learning_rate": 0.0001449474094711721, + "loss": 0.5213, + "step": 21390 + }, + { + "epoch": 1.4843522309347026, + "grad_norm": 2.2311484813690186, + "learning_rate": 0.0001449423903283015, + "loss": 0.9093, + "step": 21391 + }, + { + "epoch": 1.4844216223717992, + "grad_norm": 1.8410251140594482, + "learning_rate": 0.000144937371043553, + "loss": 0.7231, + "step": 21392 + }, + { + "epoch": 1.484491013808896, + "grad_norm": 2.143747329711914, + "learning_rate": 0.00014493235161694242, + "loss": 0.5871, + "step": 21393 + }, + { + "epoch": 1.4845604052459926, + "grad_norm": 2.790018320083618, + "learning_rate": 0.00014492733204848562, + "loss": 0.7052, + "step": 21394 + }, + { + "epoch": 1.4846297966830893, + "grad_norm": 2.0036818981170654, + "learning_rate": 0.00014492231233819843, + "loss": 0.7618, + "step": 21395 + }, + { + "epoch": 1.484699188120186, + "grad_norm": 2.1983182430267334, + "learning_rate": 0.00014491729248609675, + "loss": 0.6168, + "step": 21396 + }, + { + "epoch": 1.4847685795572827, + "grad_norm": 2.7499091625213623, + "learning_rate": 0.00014491227249219637, + "loss": 0.7501, + "step": 21397 + }, + { + "epoch": 1.4848379709943793, + "grad_norm": 2.0626471042633057, + "learning_rate": 0.00014490725235651315, + "loss": 0.5547, + "step": 21398 + }, + { + "epoch": 1.484907362431476, + "grad_norm": 2.3159902095794678, + "learning_rate": 0.00014490223207906295, + "loss": 0.6758, + "step": 21399 + }, + { + "epoch": 1.4849767538685725, + "grad_norm": 1.8731842041015625, + "learning_rate": 0.0001448972116598616, + "loss": 0.6322, + "step": 21400 + }, + { + "epoch": 1.4850461453056694, + "grad_norm": 2.185405969619751, + "learning_rate": 0.00014489219109892497, + "loss": 0.7516, + "step": 21401 + }, + { + "epoch": 1.485115536742766, + "grad_norm": 2.6031086444854736, + "learning_rate": 0.0001448871703962689, + "loss": 0.8885, + "step": 21402 + }, + { + "epoch": 1.4851849281798626, + "grad_norm": 2.585355043411255, + "learning_rate": 0.00014488214955190925, + "loss": 0.914, + "step": 21403 + }, + { + "epoch": 1.4852543196169592, + "grad_norm": 1.6917132139205933, + "learning_rate": 0.00014487712856586186, + "loss": 0.65, + "step": 21404 + }, + { + "epoch": 1.4853237110540558, + "grad_norm": 1.4744186401367188, + "learning_rate": 0.0001448721074381426, + "loss": 0.4745, + "step": 21405 + }, + { + "epoch": 1.4853931024911526, + "grad_norm": 1.8334070444107056, + "learning_rate": 0.00014486708616876727, + "loss": 0.7048, + "step": 21406 + }, + { + "epoch": 1.4854624939282493, + "grad_norm": 1.3877464532852173, + "learning_rate": 0.00014486206475775176, + "loss": 0.4364, + "step": 21407 + }, + { + "epoch": 1.4855318853653459, + "grad_norm": 1.402289867401123, + "learning_rate": 0.0001448570432051119, + "loss": 0.5187, + "step": 21408 + }, + { + "epoch": 1.4856012768024427, + "grad_norm": 1.6844488382339478, + "learning_rate": 0.0001448520215108636, + "loss": 0.5525, + "step": 21409 + }, + { + "epoch": 1.4856706682395393, + "grad_norm": 1.9442132711410522, + "learning_rate": 0.00014484699967502265, + "loss": 0.6241, + "step": 21410 + }, + { + "epoch": 1.485740059676636, + "grad_norm": 1.4113849401474, + "learning_rate": 0.00014484197769760491, + "loss": 0.5848, + "step": 21411 + }, + { + "epoch": 1.4858094511137325, + "grad_norm": 2.1535091400146484, + "learning_rate": 0.00014483695557862628, + "loss": 0.4906, + "step": 21412 + }, + { + "epoch": 1.4858788425508291, + "grad_norm": 1.9821083545684814, + "learning_rate": 0.00014483193331810257, + "loss": 0.7304, + "step": 21413 + }, + { + "epoch": 1.485948233987926, + "grad_norm": 2.1876380443573, + "learning_rate": 0.00014482691091604965, + "loss": 0.661, + "step": 21414 + }, + { + "epoch": 1.4860176254250226, + "grad_norm": 1.6772183179855347, + "learning_rate": 0.00014482188837248335, + "loss": 0.6646, + "step": 21415 + }, + { + "epoch": 1.4860870168621192, + "grad_norm": 2.1860883235931396, + "learning_rate": 0.00014481686568741962, + "loss": 0.4533, + "step": 21416 + }, + { + "epoch": 1.4861564082992158, + "grad_norm": 2.5397324562072754, + "learning_rate": 0.00014481184286087417, + "loss": 0.8755, + "step": 21417 + }, + { + "epoch": 1.4862257997363124, + "grad_norm": 2.8915231227874756, + "learning_rate": 0.00014480681989286298, + "loss": 0.746, + "step": 21418 + }, + { + "epoch": 1.4862951911734092, + "grad_norm": 1.7413421869277954, + "learning_rate": 0.00014480179678340183, + "loss": 0.6713, + "step": 21419 + }, + { + "epoch": 1.4863645826105059, + "grad_norm": 1.9240182638168335, + "learning_rate": 0.0001447967735325066, + "loss": 0.6812, + "step": 21420 + }, + { + "epoch": 1.4864339740476025, + "grad_norm": 1.2920176982879639, + "learning_rate": 0.00014479175014019314, + "loss": 0.5377, + "step": 21421 + }, + { + "epoch": 1.4865033654846993, + "grad_norm": 1.9057555198669434, + "learning_rate": 0.00014478672660647737, + "loss": 0.6329, + "step": 21422 + }, + { + "epoch": 1.486572756921796, + "grad_norm": 1.8623855113983154, + "learning_rate": 0.00014478170293137508, + "loss": 0.8003, + "step": 21423 + }, + { + "epoch": 1.4866421483588925, + "grad_norm": 2.0543601512908936, + "learning_rate": 0.00014477667911490211, + "loss": 0.8001, + "step": 21424 + }, + { + "epoch": 1.4867115397959891, + "grad_norm": 2.3879964351654053, + "learning_rate": 0.00014477165515707443, + "loss": 0.7297, + "step": 21425 + }, + { + "epoch": 1.4867809312330857, + "grad_norm": 2.029433012008667, + "learning_rate": 0.00014476663105790778, + "loss": 0.51, + "step": 21426 + }, + { + "epoch": 1.4868503226701826, + "grad_norm": 1.7456328868865967, + "learning_rate": 0.00014476160681741807, + "loss": 0.6701, + "step": 21427 + }, + { + "epoch": 1.4869197141072792, + "grad_norm": 2.3118114471435547, + "learning_rate": 0.00014475658243562117, + "loss": 0.7628, + "step": 21428 + }, + { + "epoch": 1.4869891055443758, + "grad_norm": 2.3163464069366455, + "learning_rate": 0.00014475155791253294, + "loss": 0.5288, + "step": 21429 + }, + { + "epoch": 1.4870584969814724, + "grad_norm": 1.4761651754379272, + "learning_rate": 0.0001447465332481692, + "loss": 0.6598, + "step": 21430 + }, + { + "epoch": 1.487127888418569, + "grad_norm": 1.8633878231048584, + "learning_rate": 0.0001447415084425459, + "loss": 0.5304, + "step": 21431 + }, + { + "epoch": 1.4871972798556659, + "grad_norm": 1.895725965499878, + "learning_rate": 0.00014473648349567884, + "loss": 0.7076, + "step": 21432 + }, + { + "epoch": 1.4872666712927625, + "grad_norm": 1.6329065561294556, + "learning_rate": 0.00014473145840758389, + "loss": 0.5429, + "step": 21433 + }, + { + "epoch": 1.487336062729859, + "grad_norm": 1.7349162101745605, + "learning_rate": 0.0001447264331782769, + "loss": 0.6928, + "step": 21434 + }, + { + "epoch": 1.487405454166956, + "grad_norm": 1.7634638547897339, + "learning_rate": 0.00014472140780777374, + "loss": 0.6581, + "step": 21435 + }, + { + "epoch": 1.4874748456040525, + "grad_norm": 2.149050235748291, + "learning_rate": 0.0001447163822960903, + "loss": 0.4873, + "step": 21436 + }, + { + "epoch": 1.4875442370411491, + "grad_norm": 1.94026517868042, + "learning_rate": 0.00014471135664324242, + "loss": 0.6017, + "step": 21437 + }, + { + "epoch": 1.4876136284782457, + "grad_norm": 1.7433489561080933, + "learning_rate": 0.00014470633084924596, + "loss": 0.6251, + "step": 21438 + }, + { + "epoch": 1.4876830199153424, + "grad_norm": 1.85026216506958, + "learning_rate": 0.00014470130491411683, + "loss": 0.6962, + "step": 21439 + }, + { + "epoch": 1.4877524113524392, + "grad_norm": 1.7312126159667969, + "learning_rate": 0.00014469627883787088, + "loss": 0.5919, + "step": 21440 + }, + { + "epoch": 1.4878218027895358, + "grad_norm": 1.6592164039611816, + "learning_rate": 0.00014469125262052393, + "loss": 0.4124, + "step": 21441 + }, + { + "epoch": 1.4878911942266324, + "grad_norm": 1.571110725402832, + "learning_rate": 0.00014468622626209187, + "loss": 0.601, + "step": 21442 + }, + { + "epoch": 1.487960585663729, + "grad_norm": 2.795074939727783, + "learning_rate": 0.0001446811997625906, + "loss": 0.7503, + "step": 21443 + }, + { + "epoch": 1.4880299771008256, + "grad_norm": 1.4445141553878784, + "learning_rate": 0.00014467617312203596, + "loss": 0.4914, + "step": 21444 + }, + { + "epoch": 1.4880993685379225, + "grad_norm": 1.665200114250183, + "learning_rate": 0.00014467114634044383, + "loss": 0.685, + "step": 21445 + }, + { + "epoch": 1.488168759975019, + "grad_norm": 1.222834587097168, + "learning_rate": 0.0001446661194178301, + "loss": 0.485, + "step": 21446 + }, + { + "epoch": 1.4882381514121157, + "grad_norm": 1.516342282295227, + "learning_rate": 0.00014466109235421057, + "loss": 0.5187, + "step": 21447 + }, + { + "epoch": 1.4883075428492125, + "grad_norm": 1.7944697141647339, + "learning_rate": 0.00014465606514960116, + "loss": 0.4493, + "step": 21448 + }, + { + "epoch": 1.4883769342863091, + "grad_norm": 1.5958163738250732, + "learning_rate": 0.00014465103780401774, + "loss": 0.494, + "step": 21449 + }, + { + "epoch": 1.4884463257234057, + "grad_norm": 1.4581636190414429, + "learning_rate": 0.00014464601031747619, + "loss": 0.6248, + "step": 21450 + }, + { + "epoch": 1.4885157171605023, + "grad_norm": 1.3880395889282227, + "learning_rate": 0.00014464098268999237, + "loss": 0.5544, + "step": 21451 + }, + { + "epoch": 1.488585108597599, + "grad_norm": 2.321613073348999, + "learning_rate": 0.0001446359549215821, + "loss": 0.5191, + "step": 21452 + }, + { + "epoch": 1.4886545000346958, + "grad_norm": 2.334660530090332, + "learning_rate": 0.00014463092701226136, + "loss": 0.7988, + "step": 21453 + }, + { + "epoch": 1.4887238914717924, + "grad_norm": 2.1200296878814697, + "learning_rate": 0.00014462589896204592, + "loss": 0.8895, + "step": 21454 + }, + { + "epoch": 1.488793282908889, + "grad_norm": 1.8260799646377563, + "learning_rate": 0.0001446208707709517, + "loss": 0.5048, + "step": 21455 + }, + { + "epoch": 1.4888626743459856, + "grad_norm": 1.5306191444396973, + "learning_rate": 0.00014461584243899458, + "loss": 0.7709, + "step": 21456 + }, + { + "epoch": 1.4889320657830822, + "grad_norm": 2.2248568534851074, + "learning_rate": 0.00014461081396619042, + "loss": 0.6462, + "step": 21457 + }, + { + "epoch": 1.489001457220179, + "grad_norm": 2.3688089847564697, + "learning_rate": 0.00014460578535255506, + "loss": 0.7487, + "step": 21458 + }, + { + "epoch": 1.4890708486572757, + "grad_norm": 2.31272029876709, + "learning_rate": 0.00014460075659810447, + "loss": 0.856, + "step": 21459 + }, + { + "epoch": 1.4891402400943723, + "grad_norm": 2.402841567993164, + "learning_rate": 0.00014459572770285443, + "loss": 0.7974, + "step": 21460 + }, + { + "epoch": 1.4892096315314691, + "grad_norm": 1.762424349784851, + "learning_rate": 0.00014459069866682085, + "loss": 0.5566, + "step": 21461 + }, + { + "epoch": 1.4892790229685657, + "grad_norm": 1.2160156965255737, + "learning_rate": 0.00014458566949001963, + "loss": 0.4609, + "step": 21462 + }, + { + "epoch": 1.4893484144056623, + "grad_norm": 1.859452247619629, + "learning_rate": 0.0001445806401724666, + "loss": 0.7609, + "step": 21463 + }, + { + "epoch": 1.489417805842759, + "grad_norm": 2.2707502841949463, + "learning_rate": 0.00014457561071417767, + "loss": 0.6657, + "step": 21464 + }, + { + "epoch": 1.4894871972798556, + "grad_norm": 2.1479644775390625, + "learning_rate": 0.00014457058111516871, + "loss": 0.9079, + "step": 21465 + }, + { + "epoch": 1.4895565887169524, + "grad_norm": 1.9759204387664795, + "learning_rate": 0.0001445655513754556, + "loss": 0.6524, + "step": 21466 + }, + { + "epoch": 1.489625980154049, + "grad_norm": 2.2327351570129395, + "learning_rate": 0.0001445605214950542, + "loss": 0.6555, + "step": 21467 + }, + { + "epoch": 1.4896953715911456, + "grad_norm": 1.6677062511444092, + "learning_rate": 0.00014455549147398045, + "loss": 0.598, + "step": 21468 + }, + { + "epoch": 1.4897647630282422, + "grad_norm": 1.915425181388855, + "learning_rate": 0.00014455046131225014, + "loss": 0.4726, + "step": 21469 + }, + { + "epoch": 1.4898341544653388, + "grad_norm": 2.4864611625671387, + "learning_rate": 0.0001445454310098792, + "loss": 0.9442, + "step": 21470 + }, + { + "epoch": 1.4899035459024357, + "grad_norm": 1.8145734071731567, + "learning_rate": 0.0001445404005668835, + "loss": 0.5656, + "step": 21471 + }, + { + "epoch": 1.4899729373395323, + "grad_norm": 1.5202354192733765, + "learning_rate": 0.00014453536998327893, + "loss": 0.6563, + "step": 21472 + }, + { + "epoch": 1.490042328776629, + "grad_norm": 1.5201040506362915, + "learning_rate": 0.00014453033925908136, + "loss": 0.5006, + "step": 21473 + }, + { + "epoch": 1.4901117202137257, + "grad_norm": 1.8944107294082642, + "learning_rate": 0.0001445253083943067, + "loss": 0.7754, + "step": 21474 + }, + { + "epoch": 1.4901811116508223, + "grad_norm": 2.5598785877227783, + "learning_rate": 0.00014452027738897077, + "loss": 0.6433, + "step": 21475 + }, + { + "epoch": 1.490250503087919, + "grad_norm": 2.096503496170044, + "learning_rate": 0.00014451524624308948, + "loss": 0.7655, + "step": 21476 + }, + { + "epoch": 1.4903198945250156, + "grad_norm": 2.3571414947509766, + "learning_rate": 0.00014451021495667877, + "loss": 0.7378, + "step": 21477 + }, + { + "epoch": 1.4903892859621122, + "grad_norm": 1.1863616704940796, + "learning_rate": 0.00014450518352975445, + "loss": 0.5257, + "step": 21478 + }, + { + "epoch": 1.490458677399209, + "grad_norm": 2.435732841491699, + "learning_rate": 0.0001445001519623324, + "loss": 0.9613, + "step": 21479 + }, + { + "epoch": 1.4905280688363056, + "grad_norm": 1.7114900350570679, + "learning_rate": 0.00014449512025442859, + "loss": 0.7497, + "step": 21480 + }, + { + "epoch": 1.4905974602734022, + "grad_norm": 1.3788182735443115, + "learning_rate": 0.0001444900884060588, + "loss": 0.6155, + "step": 21481 + }, + { + "epoch": 1.4906668517104988, + "grad_norm": 1.293872356414795, + "learning_rate": 0.000144485056417239, + "loss": 0.5604, + "step": 21482 + }, + { + "epoch": 1.4907362431475955, + "grad_norm": 1.1833746433258057, + "learning_rate": 0.000144480024287985, + "loss": 0.5699, + "step": 21483 + }, + { + "epoch": 1.4908056345846923, + "grad_norm": 1.8078325986862183, + "learning_rate": 0.00014447499201831275, + "loss": 0.7604, + "step": 21484 + }, + { + "epoch": 1.490875026021789, + "grad_norm": 2.235736608505249, + "learning_rate": 0.00014446995960823813, + "loss": 0.6944, + "step": 21485 + }, + { + "epoch": 1.4909444174588855, + "grad_norm": 1.642537236213684, + "learning_rate": 0.000144464927057777, + "loss": 0.5424, + "step": 21486 + }, + { + "epoch": 1.4910138088959823, + "grad_norm": 1.5154345035552979, + "learning_rate": 0.00014445989436694522, + "loss": 0.6386, + "step": 21487 + }, + { + "epoch": 1.491083200333079, + "grad_norm": 2.0731403827667236, + "learning_rate": 0.00014445486153575875, + "loss": 0.8356, + "step": 21488 + }, + { + "epoch": 1.4911525917701756, + "grad_norm": 1.8542550802230835, + "learning_rate": 0.0001444498285642334, + "loss": 0.5844, + "step": 21489 + }, + { + "epoch": 1.4912219832072722, + "grad_norm": 3.8777379989624023, + "learning_rate": 0.0001444447954523851, + "loss": 0.7621, + "step": 21490 + }, + { + "epoch": 1.4912913746443688, + "grad_norm": 1.9275838136672974, + "learning_rate": 0.00014443976220022977, + "loss": 0.7226, + "step": 21491 + }, + { + "epoch": 1.4913607660814656, + "grad_norm": 1.7961921691894531, + "learning_rate": 0.00014443472880778323, + "loss": 0.8235, + "step": 21492 + }, + { + "epoch": 1.4914301575185622, + "grad_norm": 1.651086449623108, + "learning_rate": 0.00014442969527506144, + "loss": 0.502, + "step": 21493 + }, + { + "epoch": 1.4914995489556588, + "grad_norm": 1.7663747072219849, + "learning_rate": 0.00014442466160208026, + "loss": 0.6128, + "step": 21494 + }, + { + "epoch": 1.4915689403927557, + "grad_norm": 2.734718084335327, + "learning_rate": 0.00014441962778885556, + "loss": 0.6295, + "step": 21495 + }, + { + "epoch": 1.491638331829852, + "grad_norm": 1.319105625152588, + "learning_rate": 0.00014441459383540323, + "loss": 0.4886, + "step": 21496 + }, + { + "epoch": 1.491707723266949, + "grad_norm": 2.004223108291626, + "learning_rate": 0.0001444095597417392, + "loss": 0.6017, + "step": 21497 + }, + { + "epoch": 1.4917771147040455, + "grad_norm": 1.6612664461135864, + "learning_rate": 0.00014440452550787937, + "loss": 0.6487, + "step": 21498 + }, + { + "epoch": 1.4918465061411421, + "grad_norm": 1.557169795036316, + "learning_rate": 0.00014439949113383957, + "loss": 0.5636, + "step": 21499 + }, + { + "epoch": 1.491915897578239, + "grad_norm": 1.498873233795166, + "learning_rate": 0.00014439445661963574, + "loss": 0.8245, + "step": 21500 + }, + { + "epoch": 1.4919852890153356, + "grad_norm": 1.412842035293579, + "learning_rate": 0.00014438942196528376, + "loss": 0.5181, + "step": 21501 + }, + { + "epoch": 1.4920546804524322, + "grad_norm": 1.6974844932556152, + "learning_rate": 0.00014438438717079952, + "loss": 0.6373, + "step": 21502 + }, + { + "epoch": 1.4921240718895288, + "grad_norm": 2.1848440170288086, + "learning_rate": 0.00014437935223619893, + "loss": 0.6917, + "step": 21503 + }, + { + "epoch": 1.4921934633266254, + "grad_norm": 1.6239994764328003, + "learning_rate": 0.00014437431716149784, + "loss": 0.5266, + "step": 21504 + }, + { + "epoch": 1.4922628547637222, + "grad_norm": 1.5732189416885376, + "learning_rate": 0.00014436928194671222, + "loss": 0.4687, + "step": 21505 + }, + { + "epoch": 1.4923322462008188, + "grad_norm": 1.8913929462432861, + "learning_rate": 0.00014436424659185792, + "loss": 0.7666, + "step": 21506 + }, + { + "epoch": 1.4924016376379154, + "grad_norm": 1.745680809020996, + "learning_rate": 0.0001443592110969508, + "loss": 0.5549, + "step": 21507 + }, + { + "epoch": 1.4924710290750123, + "grad_norm": 1.376230239868164, + "learning_rate": 0.00014435417546200684, + "loss": 0.4983, + "step": 21508 + }, + { + "epoch": 1.4925404205121087, + "grad_norm": 1.2771127223968506, + "learning_rate": 0.00014434913968704188, + "loss": 0.5426, + "step": 21509 + }, + { + "epoch": 1.4926098119492055, + "grad_norm": 1.6408939361572266, + "learning_rate": 0.00014434410377207183, + "loss": 0.6087, + "step": 21510 + }, + { + "epoch": 1.4926792033863021, + "grad_norm": 1.8158568143844604, + "learning_rate": 0.0001443390677171126, + "loss": 0.52, + "step": 21511 + }, + { + "epoch": 1.4927485948233987, + "grad_norm": 1.881463885307312, + "learning_rate": 0.00014433403152218008, + "loss": 0.8899, + "step": 21512 + }, + { + "epoch": 1.4928179862604956, + "grad_norm": 2.080449342727661, + "learning_rate": 0.00014432899518729015, + "loss": 0.7844, + "step": 21513 + }, + { + "epoch": 1.4928873776975922, + "grad_norm": 1.9753191471099854, + "learning_rate": 0.00014432395871245873, + "loss": 0.628, + "step": 21514 + }, + { + "epoch": 1.4929567691346888, + "grad_norm": 1.9389209747314453, + "learning_rate": 0.00014431892209770174, + "loss": 0.9042, + "step": 21515 + }, + { + "epoch": 1.4930261605717854, + "grad_norm": 1.784575343132019, + "learning_rate": 0.000144313885343035, + "loss": 0.4077, + "step": 21516 + }, + { + "epoch": 1.493095552008882, + "grad_norm": 1.4776734113693237, + "learning_rate": 0.0001443088484484745, + "loss": 0.6186, + "step": 21517 + }, + { + "epoch": 1.4931649434459788, + "grad_norm": 1.271274209022522, + "learning_rate": 0.00014430381141403612, + "loss": 0.5087, + "step": 21518 + }, + { + "epoch": 1.4932343348830754, + "grad_norm": 1.6365957260131836, + "learning_rate": 0.00014429877423973573, + "loss": 0.5317, + "step": 21519 + }, + { + "epoch": 1.493303726320172, + "grad_norm": 1.34060537815094, + "learning_rate": 0.00014429373692558924, + "loss": 0.4003, + "step": 21520 + }, + { + "epoch": 1.4933731177572689, + "grad_norm": 1.5584545135498047, + "learning_rate": 0.00014428869947161256, + "loss": 0.6085, + "step": 21521 + }, + { + "epoch": 1.4934425091943655, + "grad_norm": 2.2244455814361572, + "learning_rate": 0.00014428366187782163, + "loss": 0.7477, + "step": 21522 + }, + { + "epoch": 1.493511900631462, + "grad_norm": 2.4118876457214355, + "learning_rate": 0.0001442786241442323, + "loss": 0.5455, + "step": 21523 + }, + { + "epoch": 1.4935812920685587, + "grad_norm": 2.093097448348999, + "learning_rate": 0.00014427358627086044, + "loss": 0.8328, + "step": 21524 + }, + { + "epoch": 1.4936506835056553, + "grad_norm": 1.5904541015625, + "learning_rate": 0.0001442685482577221, + "loss": 0.5476, + "step": 21525 + }, + { + "epoch": 1.4937200749427522, + "grad_norm": 1.375937581062317, + "learning_rate": 0.00014426351010483302, + "loss": 0.6461, + "step": 21526 + }, + { + "epoch": 1.4937894663798488, + "grad_norm": 1.7858386039733887, + "learning_rate": 0.00014425847181220918, + "loss": 0.5278, + "step": 21527 + }, + { + "epoch": 1.4938588578169454, + "grad_norm": 2.0867202281951904, + "learning_rate": 0.0001442534333798665, + "loss": 0.7354, + "step": 21528 + }, + { + "epoch": 1.493928249254042, + "grad_norm": 1.990688443183899, + "learning_rate": 0.00014424839480782085, + "loss": 0.776, + "step": 21529 + }, + { + "epoch": 1.4939976406911386, + "grad_norm": 2.6871531009674072, + "learning_rate": 0.00014424335609608817, + "loss": 0.7959, + "step": 21530 + }, + { + "epoch": 1.4940670321282354, + "grad_norm": 1.474277138710022, + "learning_rate": 0.0001442383172446843, + "loss": 0.4928, + "step": 21531 + }, + { + "epoch": 1.494136423565332, + "grad_norm": 1.8111701011657715, + "learning_rate": 0.00014423327825362526, + "loss": 0.6736, + "step": 21532 + }, + { + "epoch": 1.4942058150024287, + "grad_norm": 2.722353219985962, + "learning_rate": 0.00014422823912292688, + "loss": 0.525, + "step": 21533 + }, + { + "epoch": 1.4942752064395255, + "grad_norm": 2.2988643646240234, + "learning_rate": 0.00014422319985260505, + "loss": 0.6126, + "step": 21534 + }, + { + "epoch": 1.494344597876622, + "grad_norm": 1.434909462928772, + "learning_rate": 0.00014421816044267569, + "loss": 0.5296, + "step": 21535 + }, + { + "epoch": 1.4944139893137187, + "grad_norm": 1.5003066062927246, + "learning_rate": 0.00014421312089315478, + "loss": 0.4946, + "step": 21536 + }, + { + "epoch": 1.4944833807508153, + "grad_norm": 1.620145320892334, + "learning_rate": 0.00014420808120405815, + "loss": 0.5028, + "step": 21537 + }, + { + "epoch": 1.494552772187912, + "grad_norm": 1.1346585750579834, + "learning_rate": 0.00014420304137540178, + "loss": 0.336, + "step": 21538 + }, + { + "epoch": 1.4946221636250088, + "grad_norm": 1.9528359174728394, + "learning_rate": 0.0001441980014072015, + "loss": 0.6039, + "step": 21539 + }, + { + "epoch": 1.4946915550621054, + "grad_norm": 1.5536901950836182, + "learning_rate": 0.00014419296129947323, + "loss": 0.6014, + "step": 21540 + }, + { + "epoch": 1.494760946499202, + "grad_norm": 1.6759032011032104, + "learning_rate": 0.00014418792105223296, + "loss": 0.5365, + "step": 21541 + }, + { + "epoch": 1.4948303379362986, + "grad_norm": 1.7482330799102783, + "learning_rate": 0.0001441828806654965, + "loss": 0.5415, + "step": 21542 + }, + { + "epoch": 1.4948997293733952, + "grad_norm": 1.65962815284729, + "learning_rate": 0.00014417784013927987, + "loss": 0.5358, + "step": 21543 + }, + { + "epoch": 1.494969120810492, + "grad_norm": 1.6708070039749146, + "learning_rate": 0.00014417279947359887, + "loss": 0.4964, + "step": 21544 + }, + { + "epoch": 1.4950385122475887, + "grad_norm": 1.562293291091919, + "learning_rate": 0.0001441677586684695, + "loss": 0.5258, + "step": 21545 + }, + { + "epoch": 1.4951079036846853, + "grad_norm": 1.9865902662277222, + "learning_rate": 0.00014416271772390763, + "loss": 0.675, + "step": 21546 + }, + { + "epoch": 1.495177295121782, + "grad_norm": 1.3617240190505981, + "learning_rate": 0.00014415767663992918, + "loss": 0.4678, + "step": 21547 + }, + { + "epoch": 1.4952466865588787, + "grad_norm": 2.186650514602661, + "learning_rate": 0.00014415263541655006, + "loss": 0.6939, + "step": 21548 + }, + { + "epoch": 1.4953160779959753, + "grad_norm": 1.5540108680725098, + "learning_rate": 0.0001441475940537862, + "loss": 0.6065, + "step": 21549 + }, + { + "epoch": 1.495385469433072, + "grad_norm": 1.4963117837905884, + "learning_rate": 0.00014414255255165354, + "loss": 0.6375, + "step": 21550 + }, + { + "epoch": 1.4954548608701685, + "grad_norm": 1.4373592138290405, + "learning_rate": 0.00014413751091016792, + "loss": 0.5099, + "step": 21551 + }, + { + "epoch": 1.4955242523072654, + "grad_norm": 1.3264485597610474, + "learning_rate": 0.0001441324691293453, + "loss": 0.5646, + "step": 21552 + }, + { + "epoch": 1.495593643744362, + "grad_norm": 2.0890417098999023, + "learning_rate": 0.00014412742720920163, + "loss": 0.7649, + "step": 21553 + }, + { + "epoch": 1.4956630351814586, + "grad_norm": 1.404135823249817, + "learning_rate": 0.00014412238514975275, + "loss": 0.4153, + "step": 21554 + }, + { + "epoch": 1.4957324266185552, + "grad_norm": 2.5554025173187256, + "learning_rate": 0.00014411734295101468, + "loss": 0.7935, + "step": 21555 + }, + { + "epoch": 1.4958018180556518, + "grad_norm": 1.54981529712677, + "learning_rate": 0.00014411230061300322, + "loss": 0.5275, + "step": 21556 + }, + { + "epoch": 1.4958712094927487, + "grad_norm": 1.3246475458145142, + "learning_rate": 0.00014410725813573437, + "loss": 0.4786, + "step": 21557 + }, + { + "epoch": 1.4959406009298453, + "grad_norm": 1.61319100856781, + "learning_rate": 0.00014410221551922403, + "loss": 0.3883, + "step": 21558 + }, + { + "epoch": 1.4960099923669419, + "grad_norm": 1.8607767820358276, + "learning_rate": 0.0001440971727634881, + "loss": 0.6853, + "step": 21559 + }, + { + "epoch": 1.4960793838040387, + "grad_norm": 1.7840864658355713, + "learning_rate": 0.0001440921298685425, + "loss": 0.7199, + "step": 21560 + }, + { + "epoch": 1.4961487752411353, + "grad_norm": 2.1114001274108887, + "learning_rate": 0.00014408708683440318, + "loss": 0.8399, + "step": 21561 + }, + { + "epoch": 1.496218166678232, + "grad_norm": 1.592462420463562, + "learning_rate": 0.00014408204366108603, + "loss": 0.6548, + "step": 21562 + }, + { + "epoch": 1.4962875581153285, + "grad_norm": 1.3694807291030884, + "learning_rate": 0.000144077000348607, + "loss": 0.5498, + "step": 21563 + }, + { + "epoch": 1.4963569495524252, + "grad_norm": 2.6799159049987793, + "learning_rate": 0.00014407195689698196, + "loss": 0.7148, + "step": 21564 + }, + { + "epoch": 1.496426340989522, + "grad_norm": 1.6293106079101562, + "learning_rate": 0.00014406691330622688, + "loss": 0.5982, + "step": 21565 + }, + { + "epoch": 1.4964957324266186, + "grad_norm": 1.7054094076156616, + "learning_rate": 0.00014406186957635769, + "loss": 0.6494, + "step": 21566 + }, + { + "epoch": 1.4965651238637152, + "grad_norm": 1.5399073362350464, + "learning_rate": 0.00014405682570739026, + "loss": 0.5768, + "step": 21567 + }, + { + "epoch": 1.4966345153008118, + "grad_norm": 2.8798646926879883, + "learning_rate": 0.00014405178169934056, + "loss": 0.6137, + "step": 21568 + }, + { + "epoch": 1.4967039067379084, + "grad_norm": 2.003666400909424, + "learning_rate": 0.00014404673755222448, + "loss": 0.5747, + "step": 21569 + }, + { + "epoch": 1.4967732981750053, + "grad_norm": 2.0889699459075928, + "learning_rate": 0.00014404169326605798, + "loss": 0.6757, + "step": 21570 + }, + { + "epoch": 1.4968426896121019, + "grad_norm": 1.4459483623504639, + "learning_rate": 0.00014403664884085694, + "loss": 0.5371, + "step": 21571 + }, + { + "epoch": 1.4969120810491985, + "grad_norm": 2.338245153427124, + "learning_rate": 0.00014403160427663736, + "loss": 0.7004, + "step": 21572 + }, + { + "epoch": 1.4969814724862953, + "grad_norm": 1.7155420780181885, + "learning_rate": 0.00014402655957341506, + "loss": 0.6201, + "step": 21573 + }, + { + "epoch": 1.497050863923392, + "grad_norm": 2.5676429271698, + "learning_rate": 0.00014402151473120607, + "loss": 0.6074, + "step": 21574 + }, + { + "epoch": 1.4971202553604885, + "grad_norm": 2.0576298236846924, + "learning_rate": 0.00014401646975002622, + "loss": 0.7367, + "step": 21575 + }, + { + "epoch": 1.4971896467975851, + "grad_norm": 12.809576988220215, + "learning_rate": 0.00014401142462989148, + "loss": 0.8342, + "step": 21576 + }, + { + "epoch": 1.4972590382346818, + "grad_norm": 1.9080079793930054, + "learning_rate": 0.00014400637937081782, + "loss": 0.6346, + "step": 21577 + }, + { + "epoch": 1.4973284296717786, + "grad_norm": 2.392756462097168, + "learning_rate": 0.0001440013339728211, + "loss": 0.6355, + "step": 21578 + }, + { + "epoch": 1.4973978211088752, + "grad_norm": 1.9168888330459595, + "learning_rate": 0.00014399628843591726, + "loss": 0.6736, + "step": 21579 + }, + { + "epoch": 1.4974672125459718, + "grad_norm": 1.6161305904388428, + "learning_rate": 0.00014399124276012222, + "loss": 0.6631, + "step": 21580 + }, + { + "epoch": 1.4975366039830684, + "grad_norm": 1.2731930017471313, + "learning_rate": 0.000143986196945452, + "loss": 0.5409, + "step": 21581 + }, + { + "epoch": 1.497605995420165, + "grad_norm": 1.7844594717025757, + "learning_rate": 0.00014398115099192242, + "loss": 0.5972, + "step": 21582 + }, + { + "epoch": 1.4976753868572619, + "grad_norm": 2.5568196773529053, + "learning_rate": 0.00014397610489954945, + "loss": 0.8025, + "step": 21583 + }, + { + "epoch": 1.4977447782943585, + "grad_norm": 2.2833735942840576, + "learning_rate": 0.00014397105866834902, + "loss": 0.7963, + "step": 21584 + }, + { + "epoch": 1.497814169731455, + "grad_norm": 2.099468231201172, + "learning_rate": 0.00014396601229833706, + "loss": 0.8121, + "step": 21585 + }, + { + "epoch": 1.497883561168552, + "grad_norm": 2.4698057174682617, + "learning_rate": 0.0001439609657895295, + "loss": 0.7462, + "step": 21586 + }, + { + "epoch": 1.4979529526056485, + "grad_norm": 1.4367097616195679, + "learning_rate": 0.0001439559191419423, + "loss": 0.5517, + "step": 21587 + }, + { + "epoch": 1.4980223440427451, + "grad_norm": 1.76886785030365, + "learning_rate": 0.00014395087235559133, + "loss": 0.6845, + "step": 21588 + }, + { + "epoch": 1.4980917354798418, + "grad_norm": 1.6878485679626465, + "learning_rate": 0.00014394582543049254, + "loss": 0.5031, + "step": 21589 + }, + { + "epoch": 1.4981611269169384, + "grad_norm": 1.610556721687317, + "learning_rate": 0.00014394077836666193, + "loss": 0.6374, + "step": 21590 + }, + { + "epoch": 1.4982305183540352, + "grad_norm": 1.5284411907196045, + "learning_rate": 0.00014393573116411537, + "loss": 0.5487, + "step": 21591 + }, + { + "epoch": 1.4982999097911318, + "grad_norm": 1.750566840171814, + "learning_rate": 0.00014393068382286879, + "loss": 0.7983, + "step": 21592 + }, + { + "epoch": 1.4983693012282284, + "grad_norm": 2.0130600929260254, + "learning_rate": 0.00014392563634293813, + "loss": 0.7998, + "step": 21593 + }, + { + "epoch": 1.498438692665325, + "grad_norm": 1.621565818786621, + "learning_rate": 0.00014392058872433934, + "loss": 0.5698, + "step": 21594 + }, + { + "epoch": 1.4985080841024216, + "grad_norm": 1.9985147714614868, + "learning_rate": 0.00014391554096708834, + "loss": 0.6816, + "step": 21595 + }, + { + "epoch": 1.4985774755395185, + "grad_norm": 1.7285873889923096, + "learning_rate": 0.00014391049307120105, + "loss": 0.5447, + "step": 21596 + }, + { + "epoch": 1.498646866976615, + "grad_norm": 3.0701894760131836, + "learning_rate": 0.00014390544503669345, + "loss": 0.854, + "step": 21597 + }, + { + "epoch": 1.4987162584137117, + "grad_norm": 1.6125460863113403, + "learning_rate": 0.00014390039686358147, + "loss": 0.7444, + "step": 21598 + }, + { + "epoch": 1.4987856498508085, + "grad_norm": 1.519440770149231, + "learning_rate": 0.00014389534855188102, + "loss": 0.549, + "step": 21599 + }, + { + "epoch": 1.4988550412879051, + "grad_norm": 1.43959379196167, + "learning_rate": 0.00014389030010160804, + "loss": 0.6455, + "step": 21600 + }, + { + "epoch": 1.4989244327250018, + "grad_norm": 1.7805547714233398, + "learning_rate": 0.0001438852515127785, + "loss": 0.7126, + "step": 21601 + }, + { + "epoch": 1.4989938241620984, + "grad_norm": 1.9031500816345215, + "learning_rate": 0.00014388020278540827, + "loss": 0.7232, + "step": 21602 + }, + { + "epoch": 1.499063215599195, + "grad_norm": 1.7542740106582642, + "learning_rate": 0.00014387515391951335, + "loss": 0.7157, + "step": 21603 + }, + { + "epoch": 1.4991326070362918, + "grad_norm": 1.032836675643921, + "learning_rate": 0.00014387010491510966, + "loss": 0.3694, + "step": 21604 + }, + { + "epoch": 1.4992019984733884, + "grad_norm": 1.7838352918624878, + "learning_rate": 0.00014386505577221317, + "loss": 0.7649, + "step": 21605 + }, + { + "epoch": 1.499271389910485, + "grad_norm": 2.2128188610076904, + "learning_rate": 0.00014386000649083976, + "loss": 0.6433, + "step": 21606 + }, + { + "epoch": 1.4993407813475816, + "grad_norm": 2.81473970413208, + "learning_rate": 0.0001438549570710054, + "loss": 0.6739, + "step": 21607 + }, + { + "epoch": 1.4994101727846783, + "grad_norm": 1.979280948638916, + "learning_rate": 0.00014384990751272603, + "loss": 0.5404, + "step": 21608 + }, + { + "epoch": 1.499479564221775, + "grad_norm": 1.4961063861846924, + "learning_rate": 0.0001438448578160176, + "loss": 0.5338, + "step": 21609 + }, + { + "epoch": 1.4995489556588717, + "grad_norm": 1.3818762302398682, + "learning_rate": 0.00014383980798089602, + "loss": 0.6734, + "step": 21610 + }, + { + "epoch": 1.4996183470959683, + "grad_norm": 1.2646446228027344, + "learning_rate": 0.00014383475800737724, + "loss": 0.4852, + "step": 21611 + }, + { + "epoch": 1.4996877385330651, + "grad_norm": 1.9528590440750122, + "learning_rate": 0.00014382970789547726, + "loss": 0.8465, + "step": 21612 + }, + { + "epoch": 1.4997571299701618, + "grad_norm": 1.530206561088562, + "learning_rate": 0.00014382465764521194, + "loss": 0.5118, + "step": 21613 + }, + { + "epoch": 1.4998265214072584, + "grad_norm": 1.60640549659729, + "learning_rate": 0.00014381960725659727, + "loss": 0.5661, + "step": 21614 + }, + { + "epoch": 1.499895912844355, + "grad_norm": 1.5851181745529175, + "learning_rate": 0.0001438145567296492, + "loss": 0.5714, + "step": 21615 + }, + { + "epoch": 1.4999653042814516, + "grad_norm": 1.7120698690414429, + "learning_rate": 0.00014380950606438363, + "loss": 0.601, + "step": 21616 + }, + { + "epoch": 1.5000346957185484, + "grad_norm": 2.2047817707061768, + "learning_rate": 0.00014380445526081657, + "loss": 0.5569, + "step": 21617 + }, + { + "epoch": 1.500104087155645, + "grad_norm": 2.192833185195923, + "learning_rate": 0.00014379940431896387, + "loss": 0.6956, + "step": 21618 + }, + { + "epoch": 1.5001734785927416, + "grad_norm": 2.0373332500457764, + "learning_rate": 0.0001437943532388416, + "loss": 0.6228, + "step": 21619 + }, + { + "epoch": 1.5002428700298385, + "grad_norm": 1.8974565267562866, + "learning_rate": 0.0001437893020204656, + "loss": 0.704, + "step": 21620 + }, + { + "epoch": 1.5003122614669349, + "grad_norm": 1.219956398010254, + "learning_rate": 0.00014378425066385183, + "loss": 0.5212, + "step": 21621 + }, + { + "epoch": 1.5003816529040317, + "grad_norm": 1.7240782976150513, + "learning_rate": 0.0001437791991690163, + "loss": 0.5743, + "step": 21622 + }, + { + "epoch": 1.5004510443411283, + "grad_norm": 2.8750970363616943, + "learning_rate": 0.00014377414753597492, + "loss": 0.6594, + "step": 21623 + }, + { + "epoch": 1.500520435778225, + "grad_norm": 1.8677735328674316, + "learning_rate": 0.0001437690957647436, + "loss": 0.7463, + "step": 21624 + }, + { + "epoch": 1.5005898272153217, + "grad_norm": 1.414811134338379, + "learning_rate": 0.00014376404385533836, + "loss": 0.6597, + "step": 21625 + }, + { + "epoch": 1.5006592186524181, + "grad_norm": 1.374969482421875, + "learning_rate": 0.0001437589918077751, + "loss": 0.5301, + "step": 21626 + }, + { + "epoch": 1.500728610089515, + "grad_norm": 1.761082410812378, + "learning_rate": 0.00014375393962206978, + "loss": 0.6902, + "step": 21627 + }, + { + "epoch": 1.5007980015266116, + "grad_norm": 1.8022816181182861, + "learning_rate": 0.00014374888729823835, + "loss": 0.9215, + "step": 21628 + }, + { + "epoch": 1.5008673929637082, + "grad_norm": 1.4290612936019897, + "learning_rate": 0.00014374383483629674, + "loss": 0.5252, + "step": 21629 + }, + { + "epoch": 1.500936784400805, + "grad_norm": 1.7376810312271118, + "learning_rate": 0.00014373878223626093, + "loss": 0.7018, + "step": 21630 + }, + { + "epoch": 1.5010061758379016, + "grad_norm": 1.516444444656372, + "learning_rate": 0.00014373372949814684, + "loss": 0.5599, + "step": 21631 + }, + { + "epoch": 1.5010755672749982, + "grad_norm": 2.7627294063568115, + "learning_rate": 0.0001437286766219705, + "loss": 0.647, + "step": 21632 + }, + { + "epoch": 1.501144958712095, + "grad_norm": 1.972124695777893, + "learning_rate": 0.00014372362360774775, + "loss": 0.6425, + "step": 21633 + }, + { + "epoch": 1.5012143501491915, + "grad_norm": 1.8929387331008911, + "learning_rate": 0.0001437185704554946, + "loss": 0.67, + "step": 21634 + }, + { + "epoch": 1.5012837415862883, + "grad_norm": 1.4380054473876953, + "learning_rate": 0.000143713517165227, + "loss": 0.4917, + "step": 21635 + }, + { + "epoch": 1.501353133023385, + "grad_norm": 1.3690729141235352, + "learning_rate": 0.0001437084637369609, + "loss": 0.5044, + "step": 21636 + }, + { + "epoch": 1.5014225244604815, + "grad_norm": 2.147122621536255, + "learning_rate": 0.00014370341017071224, + "loss": 0.789, + "step": 21637 + }, + { + "epoch": 1.5014919158975784, + "grad_norm": 2.147616386413574, + "learning_rate": 0.000143698356466497, + "loss": 0.7655, + "step": 21638 + }, + { + "epoch": 1.5015613073346747, + "grad_norm": 1.892918348312378, + "learning_rate": 0.0001436933026243311, + "loss": 0.5162, + "step": 21639 + }, + { + "epoch": 1.5016306987717716, + "grad_norm": 1.3665196895599365, + "learning_rate": 0.00014368824864423053, + "loss": 0.4433, + "step": 21640 + }, + { + "epoch": 1.5017000902088682, + "grad_norm": 2.1689984798431396, + "learning_rate": 0.00014368319452621122, + "loss": 0.5241, + "step": 21641 + }, + { + "epoch": 1.5017694816459648, + "grad_norm": 1.3389707803726196, + "learning_rate": 0.00014367814027028915, + "loss": 0.5744, + "step": 21642 + }, + { + "epoch": 1.5018388730830616, + "grad_norm": 1.6569920778274536, + "learning_rate": 0.00014367308587648022, + "loss": 0.7244, + "step": 21643 + }, + { + "epoch": 1.5019082645201582, + "grad_norm": 1.8767603635787964, + "learning_rate": 0.0001436680313448005, + "loss": 0.6115, + "step": 21644 + }, + { + "epoch": 1.5019776559572549, + "grad_norm": 1.9572651386260986, + "learning_rate": 0.0001436629766752658, + "loss": 0.7318, + "step": 21645 + }, + { + "epoch": 1.5020470473943517, + "grad_norm": 1.6967718601226807, + "learning_rate": 0.0001436579218678922, + "loss": 0.5656, + "step": 21646 + }, + { + "epoch": 1.502116438831448, + "grad_norm": 1.6906155347824097, + "learning_rate": 0.00014365286692269555, + "loss": 0.668, + "step": 21647 + }, + { + "epoch": 1.502185830268545, + "grad_norm": 2.0370707511901855, + "learning_rate": 0.00014364781183969188, + "loss": 0.5389, + "step": 21648 + }, + { + "epoch": 1.5022552217056415, + "grad_norm": 1.7932754755020142, + "learning_rate": 0.00014364275661889715, + "loss": 0.8453, + "step": 21649 + }, + { + "epoch": 1.5023246131427381, + "grad_norm": 1.670940637588501, + "learning_rate": 0.0001436377012603273, + "loss": 0.6934, + "step": 21650 + }, + { + "epoch": 1.502394004579835, + "grad_norm": 7.251010417938232, + "learning_rate": 0.00014363264576399827, + "loss": 0.7735, + "step": 21651 + }, + { + "epoch": 1.5024633960169314, + "grad_norm": 1.762721300125122, + "learning_rate": 0.00014362759012992603, + "loss": 0.7581, + "step": 21652 + }, + { + "epoch": 1.5025327874540282, + "grad_norm": 2.0525453090667725, + "learning_rate": 0.0001436225343581266, + "loss": 0.6657, + "step": 21653 + }, + { + "epoch": 1.5026021788911248, + "grad_norm": 1.9581350088119507, + "learning_rate": 0.00014361747844861587, + "loss": 0.6512, + "step": 21654 + }, + { + "epoch": 1.5026715703282214, + "grad_norm": 1.7311880588531494, + "learning_rate": 0.0001436124224014098, + "loss": 0.7661, + "step": 21655 + }, + { + "epoch": 1.5027409617653182, + "grad_norm": 1.5738286972045898, + "learning_rate": 0.0001436073662165244, + "loss": 0.5193, + "step": 21656 + }, + { + "epoch": 1.5028103532024149, + "grad_norm": 1.202493667602539, + "learning_rate": 0.00014360230989397558, + "loss": 0.4459, + "step": 21657 + }, + { + "epoch": 1.5028797446395115, + "grad_norm": 2.0310988426208496, + "learning_rate": 0.00014359725343377933, + "loss": 0.715, + "step": 21658 + }, + { + "epoch": 1.5029491360766083, + "grad_norm": 2.4090723991394043, + "learning_rate": 0.00014359219683595162, + "loss": 0.6836, + "step": 21659 + }, + { + "epoch": 1.5030185275137047, + "grad_norm": 2.0706794261932373, + "learning_rate": 0.0001435871401005084, + "loss": 0.7076, + "step": 21660 + }, + { + "epoch": 1.5030879189508015, + "grad_norm": 1.6453709602355957, + "learning_rate": 0.00014358208322746565, + "loss": 0.6046, + "step": 21661 + }, + { + "epoch": 1.5031573103878981, + "grad_norm": 2.2533249855041504, + "learning_rate": 0.00014357702621683928, + "loss": 0.7196, + "step": 21662 + }, + { + "epoch": 1.5032267018249947, + "grad_norm": 1.8803118467330933, + "learning_rate": 0.00014357196906864531, + "loss": 0.5745, + "step": 21663 + }, + { + "epoch": 1.5032960932620916, + "grad_norm": 1.8689682483673096, + "learning_rate": 0.00014356691178289973, + "loss": 0.5682, + "step": 21664 + }, + { + "epoch": 1.503365484699188, + "grad_norm": 1.6266382932662964, + "learning_rate": 0.00014356185435961843, + "loss": 0.7064, + "step": 21665 + }, + { + "epoch": 1.5034348761362848, + "grad_norm": 1.8416738510131836, + "learning_rate": 0.0001435567967988174, + "loss": 0.8389, + "step": 21666 + }, + { + "epoch": 1.5035042675733814, + "grad_norm": 1.5751982927322388, + "learning_rate": 0.00014355173910051263, + "loss": 0.6033, + "step": 21667 + }, + { + "epoch": 1.503573659010478, + "grad_norm": 1.4718551635742188, + "learning_rate": 0.00014354668126472005, + "loss": 0.7536, + "step": 21668 + }, + { + "epoch": 1.5036430504475748, + "grad_norm": 1.558807134628296, + "learning_rate": 0.00014354162329145565, + "loss": 0.7262, + "step": 21669 + }, + { + "epoch": 1.5037124418846715, + "grad_norm": 2.3462002277374268, + "learning_rate": 0.00014353656518073545, + "loss": 0.6549, + "step": 21670 + }, + { + "epoch": 1.503781833321768, + "grad_norm": 2.0516648292541504, + "learning_rate": 0.00014353150693257534, + "loss": 0.6235, + "step": 21671 + }, + { + "epoch": 1.503851224758865, + "grad_norm": 2.6079654693603516, + "learning_rate": 0.0001435264485469913, + "loss": 0.7838, + "step": 21672 + }, + { + "epoch": 1.5039206161959613, + "grad_norm": 2.2278692722320557, + "learning_rate": 0.0001435213900239993, + "loss": 0.6616, + "step": 21673 + }, + { + "epoch": 1.5039900076330581, + "grad_norm": 2.0754787921905518, + "learning_rate": 0.00014351633136361534, + "loss": 0.8135, + "step": 21674 + }, + { + "epoch": 1.5040593990701547, + "grad_norm": 1.3770020008087158, + "learning_rate": 0.00014351127256585536, + "loss": 0.525, + "step": 21675 + }, + { + "epoch": 1.5041287905072513, + "grad_norm": 1.5656530857086182, + "learning_rate": 0.00014350621363073537, + "loss": 0.5655, + "step": 21676 + }, + { + "epoch": 1.5041981819443482, + "grad_norm": 2.0707285404205322, + "learning_rate": 0.00014350115455827127, + "loss": 0.7579, + "step": 21677 + }, + { + "epoch": 1.5042675733814446, + "grad_norm": 1.769800066947937, + "learning_rate": 0.0001434960953484791, + "loss": 0.5748, + "step": 21678 + }, + { + "epoch": 1.5043369648185414, + "grad_norm": 1.4385532140731812, + "learning_rate": 0.0001434910360013748, + "loss": 0.5594, + "step": 21679 + }, + { + "epoch": 1.504406356255638, + "grad_norm": 2.5263044834136963, + "learning_rate": 0.00014348597651697432, + "loss": 0.5497, + "step": 21680 + }, + { + "epoch": 1.5044757476927346, + "grad_norm": 4.810978889465332, + "learning_rate": 0.00014348091689529371, + "loss": 0.5983, + "step": 21681 + }, + { + "epoch": 1.5045451391298315, + "grad_norm": 1.617701530456543, + "learning_rate": 0.00014347585713634885, + "loss": 0.6907, + "step": 21682 + }, + { + "epoch": 1.504614530566928, + "grad_norm": 1.8913309574127197, + "learning_rate": 0.00014347079724015576, + "loss": 0.5077, + "step": 21683 + }, + { + "epoch": 1.5046839220040247, + "grad_norm": 3.230116844177246, + "learning_rate": 0.00014346573720673043, + "loss": 0.7733, + "step": 21684 + }, + { + "epoch": 1.5047533134411215, + "grad_norm": 2.722574472427368, + "learning_rate": 0.00014346067703608877, + "loss": 0.873, + "step": 21685 + }, + { + "epoch": 1.504822704878218, + "grad_norm": 2.5791099071502686, + "learning_rate": 0.0001434556167282468, + "loss": 1.0692, + "step": 21686 + }, + { + "epoch": 1.5048920963153147, + "grad_norm": 1.3676984310150146, + "learning_rate": 0.00014345055628322051, + "loss": 0.5551, + "step": 21687 + }, + { + "epoch": 1.5049614877524113, + "grad_norm": 1.8500525951385498, + "learning_rate": 0.00014344549570102585, + "loss": 0.5611, + "step": 21688 + }, + { + "epoch": 1.505030879189508, + "grad_norm": 2.053114175796509, + "learning_rate": 0.0001434404349816788, + "loss": 0.6734, + "step": 21689 + }, + { + "epoch": 1.5051002706266048, + "grad_norm": 1.9265644550323486, + "learning_rate": 0.0001434353741251953, + "loss": 0.6744, + "step": 21690 + }, + { + "epoch": 1.5051696620637012, + "grad_norm": 1.436086654663086, + "learning_rate": 0.00014343031313159143, + "loss": 0.7112, + "step": 21691 + }, + { + "epoch": 1.505239053500798, + "grad_norm": 2.5073723793029785, + "learning_rate": 0.00014342525200088305, + "loss": 0.8012, + "step": 21692 + }, + { + "epoch": 1.5053084449378946, + "grad_norm": 1.4963030815124512, + "learning_rate": 0.0001434201907330862, + "loss": 0.5103, + "step": 21693 + }, + { + "epoch": 1.5053778363749912, + "grad_norm": 2.1781673431396484, + "learning_rate": 0.00014341512932821687, + "loss": 0.6605, + "step": 21694 + }, + { + "epoch": 1.505447227812088, + "grad_norm": 1.6888164281845093, + "learning_rate": 0.000143410067786291, + "loss": 0.7506, + "step": 21695 + }, + { + "epoch": 1.5055166192491847, + "grad_norm": 1.6818103790283203, + "learning_rate": 0.0001434050061073246, + "loss": 0.4829, + "step": 21696 + }, + { + "epoch": 1.5055860106862813, + "grad_norm": 2.4614155292510986, + "learning_rate": 0.00014339994429133358, + "loss": 0.7196, + "step": 21697 + }, + { + "epoch": 1.5056554021233781, + "grad_norm": 2.047008514404297, + "learning_rate": 0.00014339488233833402, + "loss": 0.7053, + "step": 21698 + }, + { + "epoch": 1.5057247935604745, + "grad_norm": 1.9498485326766968, + "learning_rate": 0.00014338982024834186, + "loss": 0.7864, + "step": 21699 + }, + { + "epoch": 1.5057941849975713, + "grad_norm": 1.9278080463409424, + "learning_rate": 0.000143384758021373, + "loss": 0.6981, + "step": 21700 + }, + { + "epoch": 1.505863576434668, + "grad_norm": 1.3957948684692383, + "learning_rate": 0.00014337969565744353, + "loss": 0.5857, + "step": 21701 + }, + { + "epoch": 1.5059329678717646, + "grad_norm": 1.5276650190353394, + "learning_rate": 0.00014337463315656942, + "loss": 0.6414, + "step": 21702 + }, + { + "epoch": 1.5060023593088614, + "grad_norm": 1.7654577493667603, + "learning_rate": 0.0001433695705187666, + "loss": 0.6919, + "step": 21703 + }, + { + "epoch": 1.5060717507459578, + "grad_norm": 1.7229927778244019, + "learning_rate": 0.0001433645077440511, + "loss": 0.5377, + "step": 21704 + }, + { + "epoch": 1.5061411421830546, + "grad_norm": 2.1467645168304443, + "learning_rate": 0.00014335944483243886, + "loss": 0.6238, + "step": 21705 + }, + { + "epoch": 1.5062105336201512, + "grad_norm": 1.910959005355835, + "learning_rate": 0.0001433543817839459, + "loss": 0.541, + "step": 21706 + }, + { + "epoch": 1.5062799250572478, + "grad_norm": 1.867449402809143, + "learning_rate": 0.00014334931859858816, + "loss": 0.5472, + "step": 21707 + }, + { + "epoch": 1.5063493164943447, + "grad_norm": 1.68704354763031, + "learning_rate": 0.00014334425527638166, + "loss": 0.6242, + "step": 21708 + }, + { + "epoch": 1.5064187079314413, + "grad_norm": 2.0761632919311523, + "learning_rate": 0.0001433391918173424, + "loss": 0.6011, + "step": 21709 + }, + { + "epoch": 1.506488099368538, + "grad_norm": 2.187734365463257, + "learning_rate": 0.0001433341282214863, + "loss": 0.6509, + "step": 21710 + }, + { + "epoch": 1.5065574908056347, + "grad_norm": 1.6814271211624146, + "learning_rate": 0.00014332906448882938, + "loss": 0.7528, + "step": 21711 + }, + { + "epoch": 1.5066268822427311, + "grad_norm": 1.9605560302734375, + "learning_rate": 0.0001433240006193877, + "loss": 0.5905, + "step": 21712 + }, + { + "epoch": 1.506696273679828, + "grad_norm": 2.332017660140991, + "learning_rate": 0.00014331893661317712, + "loss": 0.8405, + "step": 21713 + }, + { + "epoch": 1.5067656651169246, + "grad_norm": 2.2360010147094727, + "learning_rate": 0.0001433138724702137, + "loss": 0.8635, + "step": 21714 + }, + { + "epoch": 1.5068350565540212, + "grad_norm": 2.0051140785217285, + "learning_rate": 0.00014330880819051343, + "loss": 0.7058, + "step": 21715 + }, + { + "epoch": 1.506904447991118, + "grad_norm": 1.9079053401947021, + "learning_rate": 0.00014330374377409224, + "loss": 0.6791, + "step": 21716 + }, + { + "epoch": 1.5069738394282146, + "grad_norm": 1.8421969413757324, + "learning_rate": 0.00014329867922096616, + "loss": 0.6887, + "step": 21717 + }, + { + "epoch": 1.5070432308653112, + "grad_norm": 1.359639048576355, + "learning_rate": 0.00014329361453115116, + "loss": 0.523, + "step": 21718 + }, + { + "epoch": 1.5071126223024078, + "grad_norm": 1.365610957145691, + "learning_rate": 0.00014328854970466329, + "loss": 0.565, + "step": 21719 + }, + { + "epoch": 1.5071820137395044, + "grad_norm": 1.660911202430725, + "learning_rate": 0.00014328348474151845, + "loss": 0.5226, + "step": 21720 + }, + { + "epoch": 1.5072514051766013, + "grad_norm": 2.9203052520751953, + "learning_rate": 0.00014327841964173268, + "loss": 0.9136, + "step": 21721 + }, + { + "epoch": 1.5073207966136979, + "grad_norm": 2.2258145809173584, + "learning_rate": 0.00014327335440532198, + "loss": 0.8387, + "step": 21722 + }, + { + "epoch": 1.5073901880507945, + "grad_norm": 2.048062801361084, + "learning_rate": 0.0001432682890323023, + "loss": 0.7221, + "step": 21723 + }, + { + "epoch": 1.5074595794878913, + "grad_norm": 1.4337263107299805, + "learning_rate": 0.00014326322352268964, + "loss": 0.5456, + "step": 21724 + }, + { + "epoch": 1.5075289709249877, + "grad_norm": 2.0149950981140137, + "learning_rate": 0.00014325815787650002, + "loss": 0.5915, + "step": 21725 + }, + { + "epoch": 1.5075983623620846, + "grad_norm": 1.7345894575119019, + "learning_rate": 0.0001432530920937494, + "loss": 0.6315, + "step": 21726 + }, + { + "epoch": 1.5076677537991812, + "grad_norm": 1.8458921909332275, + "learning_rate": 0.0001432480261744538, + "loss": 0.5997, + "step": 21727 + }, + { + "epoch": 1.5077371452362778, + "grad_norm": 1.6655524969100952, + "learning_rate": 0.0001432429601186292, + "loss": 0.8834, + "step": 21728 + }, + { + "epoch": 1.5078065366733746, + "grad_norm": 1.5870624780654907, + "learning_rate": 0.0001432378939262916, + "loss": 0.8571, + "step": 21729 + }, + { + "epoch": 1.5078759281104712, + "grad_norm": 2.1642651557922363, + "learning_rate": 0.00014323282759745698, + "loss": 0.5751, + "step": 21730 + }, + { + "epoch": 1.5079453195475678, + "grad_norm": 1.83030366897583, + "learning_rate": 0.00014322776113214137, + "loss": 0.5211, + "step": 21731 + }, + { + "epoch": 1.5080147109846644, + "grad_norm": 1.3195598125457764, + "learning_rate": 0.00014322269453036066, + "loss": 0.6323, + "step": 21732 + }, + { + "epoch": 1.508084102421761, + "grad_norm": 1.5458993911743164, + "learning_rate": 0.00014321762779213098, + "loss": 0.6534, + "step": 21733 + }, + { + "epoch": 1.5081534938588579, + "grad_norm": 2.454035520553589, + "learning_rate": 0.00014321256091746825, + "loss": 0.6749, + "step": 21734 + }, + { + "epoch": 1.5082228852959545, + "grad_norm": 2.006166458129883, + "learning_rate": 0.00014320749390638846, + "loss": 0.6609, + "step": 21735 + }, + { + "epoch": 1.508292276733051, + "grad_norm": 1.466680645942688, + "learning_rate": 0.00014320242675890763, + "loss": 0.4871, + "step": 21736 + }, + { + "epoch": 1.508361668170148, + "grad_norm": 2.0667028427124023, + "learning_rate": 0.00014319735947504177, + "loss": 0.513, + "step": 21737 + }, + { + "epoch": 1.5084310596072443, + "grad_norm": 1.5327147245407104, + "learning_rate": 0.00014319229205480682, + "loss": 0.7274, + "step": 21738 + }, + { + "epoch": 1.5085004510443412, + "grad_norm": 1.5133769512176514, + "learning_rate": 0.0001431872244982188, + "loss": 0.7341, + "step": 21739 + }, + { + "epoch": 1.5085698424814378, + "grad_norm": 1.9014747142791748, + "learning_rate": 0.00014318215680529382, + "loss": 0.6785, + "step": 21740 + }, + { + "epoch": 1.5086392339185344, + "grad_norm": 2.212815999984741, + "learning_rate": 0.00014317708897604772, + "loss": 0.715, + "step": 21741 + }, + { + "epoch": 1.5087086253556312, + "grad_norm": 1.5191149711608887, + "learning_rate": 0.00014317202101049654, + "loss": 0.6086, + "step": 21742 + }, + { + "epoch": 1.5087780167927278, + "grad_norm": 2.0726044178009033, + "learning_rate": 0.00014316695290865635, + "loss": 0.8333, + "step": 21743 + }, + { + "epoch": 1.5088474082298244, + "grad_norm": 1.5869747400283813, + "learning_rate": 0.00014316188467054303, + "loss": 0.6207, + "step": 21744 + }, + { + "epoch": 1.508916799666921, + "grad_norm": 1.5532718896865845, + "learning_rate": 0.0001431568162961727, + "loss": 0.6784, + "step": 21745 + }, + { + "epoch": 1.5089861911040177, + "grad_norm": 1.7909471988677979, + "learning_rate": 0.00014315174778556128, + "loss": 0.8565, + "step": 21746 + }, + { + "epoch": 1.5090555825411145, + "grad_norm": 1.2360981702804565, + "learning_rate": 0.0001431466791387248, + "loss": 0.4482, + "step": 21747 + }, + { + "epoch": 1.509124973978211, + "grad_norm": 1.39632248878479, + "learning_rate": 0.0001431416103556793, + "loss": 0.574, + "step": 21748 + }, + { + "epoch": 1.5091943654153077, + "grad_norm": 1.861269474029541, + "learning_rate": 0.00014313654143644072, + "loss": 0.6877, + "step": 21749 + }, + { + "epoch": 1.5092637568524045, + "grad_norm": 1.7332532405853271, + "learning_rate": 0.0001431314723810251, + "loss": 0.6771, + "step": 21750 + }, + { + "epoch": 1.509333148289501, + "grad_norm": 2.0465402603149414, + "learning_rate": 0.0001431264031894484, + "loss": 0.6873, + "step": 21751 + }, + { + "epoch": 1.5094025397265978, + "grad_norm": 1.322620153427124, + "learning_rate": 0.00014312133386172664, + "loss": 0.6025, + "step": 21752 + }, + { + "epoch": 1.5094719311636944, + "grad_norm": 1.5823904275894165, + "learning_rate": 0.00014311626439787584, + "loss": 0.7788, + "step": 21753 + }, + { + "epoch": 1.509541322600791, + "grad_norm": 3.2029881477355957, + "learning_rate": 0.000143111194797912, + "loss": 0.5531, + "step": 21754 + }, + { + "epoch": 1.5096107140378878, + "grad_norm": 2.1654248237609863, + "learning_rate": 0.00014310612506185112, + "loss": 0.7878, + "step": 21755 + }, + { + "epoch": 1.5096801054749844, + "grad_norm": 1.4962166547775269, + "learning_rate": 0.0001431010551897092, + "loss": 0.4654, + "step": 21756 + }, + { + "epoch": 1.509749496912081, + "grad_norm": 1.978833794593811, + "learning_rate": 0.00014309598518150228, + "loss": 0.7115, + "step": 21757 + }, + { + "epoch": 1.5098188883491777, + "grad_norm": 2.038930892944336, + "learning_rate": 0.00014309091503724626, + "loss": 0.6308, + "step": 21758 + }, + { + "epoch": 1.5098882797862743, + "grad_norm": 1.7461212873458862, + "learning_rate": 0.00014308584475695727, + "loss": 0.652, + "step": 21759 + }, + { + "epoch": 1.509957671223371, + "grad_norm": 1.3694632053375244, + "learning_rate": 0.0001430807743406513, + "loss": 0.5742, + "step": 21760 + }, + { + "epoch": 1.5100270626604677, + "grad_norm": 1.8815916776657104, + "learning_rate": 0.00014307570378834428, + "loss": 0.8286, + "step": 21761 + }, + { + "epoch": 1.5100964540975643, + "grad_norm": 1.3129414319992065, + "learning_rate": 0.00014307063310005225, + "loss": 0.4796, + "step": 21762 + }, + { + "epoch": 1.5101658455346612, + "grad_norm": 1.6287285089492798, + "learning_rate": 0.00014306556227579128, + "loss": 0.5538, + "step": 21763 + }, + { + "epoch": 1.5102352369717575, + "grad_norm": 2.3472304344177246, + "learning_rate": 0.00014306049131557728, + "loss": 0.6958, + "step": 21764 + }, + { + "epoch": 1.5103046284088544, + "grad_norm": 1.934877634048462, + "learning_rate": 0.00014305542021942632, + "loss": 0.5952, + "step": 21765 + }, + { + "epoch": 1.510374019845951, + "grad_norm": 1.278637409210205, + "learning_rate": 0.0001430503489873544, + "loss": 0.3965, + "step": 21766 + }, + { + "epoch": 1.5104434112830476, + "grad_norm": 1.8921887874603271, + "learning_rate": 0.0001430452776193775, + "loss": 0.6815, + "step": 21767 + }, + { + "epoch": 1.5105128027201444, + "grad_norm": 2.5507547855377197, + "learning_rate": 0.0001430402061155117, + "loss": 0.599, + "step": 21768 + }, + { + "epoch": 1.510582194157241, + "grad_norm": 1.6267127990722656, + "learning_rate": 0.00014303513447577293, + "loss": 0.6792, + "step": 21769 + }, + { + "epoch": 1.5106515855943377, + "grad_norm": 1.642425537109375, + "learning_rate": 0.00014303006270017724, + "loss": 0.5682, + "step": 21770 + }, + { + "epoch": 1.5107209770314345, + "grad_norm": 2.6604552268981934, + "learning_rate": 0.00014302499078874063, + "loss": 0.4978, + "step": 21771 + }, + { + "epoch": 1.5107903684685309, + "grad_norm": 1.995314598083496, + "learning_rate": 0.00014301991874147912, + "loss": 0.7438, + "step": 21772 + }, + { + "epoch": 1.5108597599056277, + "grad_norm": 3.6258316040039062, + "learning_rate": 0.0001430148465584087, + "loss": 0.8883, + "step": 21773 + }, + { + "epoch": 1.5109291513427243, + "grad_norm": 1.7310727834701538, + "learning_rate": 0.00014300977423954543, + "loss": 0.6615, + "step": 21774 + }, + { + "epoch": 1.510998542779821, + "grad_norm": 1.5109134912490845, + "learning_rate": 0.00014300470178490527, + "loss": 0.6322, + "step": 21775 + }, + { + "epoch": 1.5110679342169178, + "grad_norm": 1.9883416891098022, + "learning_rate": 0.00014299962919450424, + "loss": 0.6438, + "step": 21776 + }, + { + "epoch": 1.5111373256540142, + "grad_norm": 1.3933929204940796, + "learning_rate": 0.0001429945564683584, + "loss": 0.4564, + "step": 21777 + }, + { + "epoch": 1.511206717091111, + "grad_norm": 1.6577454805374146, + "learning_rate": 0.00014298948360648374, + "loss": 0.5531, + "step": 21778 + }, + { + "epoch": 1.5112761085282076, + "grad_norm": 1.9401732683181763, + "learning_rate": 0.00014298441060889625, + "loss": 0.5343, + "step": 21779 + }, + { + "epoch": 1.5113454999653042, + "grad_norm": 1.5600858926773071, + "learning_rate": 0.00014297933747561196, + "loss": 0.6456, + "step": 21780 + }, + { + "epoch": 1.511414891402401, + "grad_norm": 5.367219924926758, + "learning_rate": 0.0001429742642066469, + "loss": 0.7442, + "step": 21781 + }, + { + "epoch": 1.5114842828394977, + "grad_norm": 1.8233174085617065, + "learning_rate": 0.00014296919080201707, + "loss": 0.8398, + "step": 21782 + }, + { + "epoch": 1.5115536742765943, + "grad_norm": 1.9721218347549438, + "learning_rate": 0.0001429641172617385, + "loss": 0.7168, + "step": 21783 + }, + { + "epoch": 1.511623065713691, + "grad_norm": 1.962866187095642, + "learning_rate": 0.00014295904358582718, + "loss": 0.7554, + "step": 21784 + }, + { + "epoch": 1.5116924571507875, + "grad_norm": 1.7973169088363647, + "learning_rate": 0.00014295396977429917, + "loss": 0.6397, + "step": 21785 + }, + { + "epoch": 1.5117618485878843, + "grad_norm": 2.6822755336761475, + "learning_rate": 0.00014294889582717045, + "loss": 0.767, + "step": 21786 + }, + { + "epoch": 1.511831240024981, + "grad_norm": 2.758112668991089, + "learning_rate": 0.00014294382174445702, + "loss": 0.5404, + "step": 21787 + }, + { + "epoch": 1.5119006314620775, + "grad_norm": 1.8776891231536865, + "learning_rate": 0.00014293874752617495, + "loss": 0.7328, + "step": 21788 + }, + { + "epoch": 1.5119700228991744, + "grad_norm": 2.240302801132202, + "learning_rate": 0.00014293367317234023, + "loss": 0.7089, + "step": 21789 + }, + { + "epoch": 1.5120394143362708, + "grad_norm": 2.3581602573394775, + "learning_rate": 0.00014292859868296888, + "loss": 0.812, + "step": 21790 + }, + { + "epoch": 1.5121088057733676, + "grad_norm": 1.4759223461151123, + "learning_rate": 0.00014292352405807695, + "loss": 0.5869, + "step": 21791 + }, + { + "epoch": 1.5121781972104642, + "grad_norm": 1.8118468523025513, + "learning_rate": 0.0001429184492976804, + "loss": 0.535, + "step": 21792 + }, + { + "epoch": 1.5122475886475608, + "grad_norm": 1.9143420457839966, + "learning_rate": 0.00014291337440179528, + "loss": 0.5809, + "step": 21793 + }, + { + "epoch": 1.5123169800846576, + "grad_norm": 1.9046521186828613, + "learning_rate": 0.00014290829937043764, + "loss": 0.6602, + "step": 21794 + }, + { + "epoch": 1.5123863715217543, + "grad_norm": 1.8371587991714478, + "learning_rate": 0.0001429032242036235, + "loss": 0.5738, + "step": 21795 + }, + { + "epoch": 1.5124557629588509, + "grad_norm": 1.7989386320114136, + "learning_rate": 0.0001428981489013688, + "loss": 0.6718, + "step": 21796 + }, + { + "epoch": 1.5125251543959477, + "grad_norm": 1.7721706628799438, + "learning_rate": 0.00014289307346368967, + "loss": 0.6193, + "step": 21797 + }, + { + "epoch": 1.512594545833044, + "grad_norm": 1.4734501838684082, + "learning_rate": 0.0001428879978906021, + "loss": 0.4858, + "step": 21798 + }, + { + "epoch": 1.512663937270141, + "grad_norm": 1.5181715488433838, + "learning_rate": 0.00014288292218212205, + "loss": 0.544, + "step": 21799 + }, + { + "epoch": 1.5127333287072375, + "grad_norm": 1.624587059020996, + "learning_rate": 0.00014287784633826558, + "loss": 0.7365, + "step": 21800 + }, + { + "epoch": 1.5128027201443341, + "grad_norm": 2.0807113647460938, + "learning_rate": 0.00014287277035904877, + "loss": 0.7227, + "step": 21801 + }, + { + "epoch": 1.512872111581431, + "grad_norm": 1.7886430025100708, + "learning_rate": 0.00014286769424448755, + "loss": 0.5544, + "step": 21802 + }, + { + "epoch": 1.5129415030185274, + "grad_norm": 1.9617929458618164, + "learning_rate": 0.00014286261799459804, + "loss": 0.56, + "step": 21803 + }, + { + "epoch": 1.5130108944556242, + "grad_norm": 1.8730618953704834, + "learning_rate": 0.0001428575416093962, + "loss": 0.8351, + "step": 21804 + }, + { + "epoch": 1.5130802858927208, + "grad_norm": 2.0131914615631104, + "learning_rate": 0.0001428524650888981, + "loss": 0.6937, + "step": 21805 + }, + { + "epoch": 1.5131496773298174, + "grad_norm": 1.7377474308013916, + "learning_rate": 0.0001428473884331197, + "loss": 0.4701, + "step": 21806 + }, + { + "epoch": 1.5132190687669143, + "grad_norm": 1.5510826110839844, + "learning_rate": 0.00014284231164207707, + "loss": 0.5325, + "step": 21807 + }, + { + "epoch": 1.5132884602040109, + "grad_norm": 2.3128271102905273, + "learning_rate": 0.00014283723471578624, + "loss": 0.8497, + "step": 21808 + }, + { + "epoch": 1.5133578516411075, + "grad_norm": 1.2424792051315308, + "learning_rate": 0.00014283215765426322, + "loss": 0.4486, + "step": 21809 + }, + { + "epoch": 1.5134272430782043, + "grad_norm": 1.8957256078720093, + "learning_rate": 0.00014282708045752407, + "loss": 0.6328, + "step": 21810 + }, + { + "epoch": 1.5134966345153007, + "grad_norm": 1.2652723789215088, + "learning_rate": 0.00014282200312558475, + "loss": 0.5194, + "step": 21811 + }, + { + "epoch": 1.5135660259523975, + "grad_norm": 1.8549320697784424, + "learning_rate": 0.00014281692565846138, + "loss": 0.5463, + "step": 21812 + }, + { + "epoch": 1.5136354173894941, + "grad_norm": 1.6003271341323853, + "learning_rate": 0.00014281184805616995, + "loss": 0.4068, + "step": 21813 + }, + { + "epoch": 1.5137048088265908, + "grad_norm": 2.0186092853546143, + "learning_rate": 0.00014280677031872644, + "loss": 0.6914, + "step": 21814 + }, + { + "epoch": 1.5137742002636876, + "grad_norm": 1.7293803691864014, + "learning_rate": 0.00014280169244614693, + "loss": 0.7216, + "step": 21815 + }, + { + "epoch": 1.513843591700784, + "grad_norm": 2.469416379928589, + "learning_rate": 0.00014279661443844746, + "loss": 0.6864, + "step": 21816 + }, + { + "epoch": 1.5139129831378808, + "grad_norm": 2.1846539974212646, + "learning_rate": 0.00014279153629564407, + "loss": 0.667, + "step": 21817 + }, + { + "epoch": 1.5139823745749774, + "grad_norm": 1.7058258056640625, + "learning_rate": 0.00014278645801775274, + "loss": 0.5553, + "step": 21818 + }, + { + "epoch": 1.514051766012074, + "grad_norm": 1.6066350936889648, + "learning_rate": 0.00014278137960478952, + "loss": 0.5172, + "step": 21819 + }, + { + "epoch": 1.5141211574491709, + "grad_norm": 1.5627787113189697, + "learning_rate": 0.00014277630105677047, + "loss": 0.6833, + "step": 21820 + }, + { + "epoch": 1.5141905488862675, + "grad_norm": 2.32299542427063, + "learning_rate": 0.00014277122237371157, + "loss": 0.8071, + "step": 21821 + }, + { + "epoch": 1.514259940323364, + "grad_norm": 1.6738200187683105, + "learning_rate": 0.0001427661435556289, + "loss": 0.6685, + "step": 21822 + }, + { + "epoch": 1.514329331760461, + "grad_norm": 2.381608486175537, + "learning_rate": 0.00014276106460253847, + "loss": 0.6809, + "step": 21823 + }, + { + "epoch": 1.5143987231975573, + "grad_norm": 1.9447425603866577, + "learning_rate": 0.00014275598551445634, + "loss": 0.7555, + "step": 21824 + }, + { + "epoch": 1.5144681146346541, + "grad_norm": 4.789464473724365, + "learning_rate": 0.00014275090629139846, + "loss": 0.5166, + "step": 21825 + }, + { + "epoch": 1.5145375060717507, + "grad_norm": 1.620834469795227, + "learning_rate": 0.000142745826933381, + "loss": 0.499, + "step": 21826 + }, + { + "epoch": 1.5146068975088474, + "grad_norm": 2.1059274673461914, + "learning_rate": 0.0001427407474404199, + "loss": 0.9491, + "step": 21827 + }, + { + "epoch": 1.5146762889459442, + "grad_norm": 1.6303942203521729, + "learning_rate": 0.0001427356678125312, + "loss": 0.6877, + "step": 21828 + }, + { + "epoch": 1.5147456803830406, + "grad_norm": 2.0767931938171387, + "learning_rate": 0.000142730588049731, + "loss": 0.6327, + "step": 21829 + }, + { + "epoch": 1.5148150718201374, + "grad_norm": 2.616499662399292, + "learning_rate": 0.00014272550815203525, + "loss": 0.7023, + "step": 21830 + }, + { + "epoch": 1.514884463257234, + "grad_norm": 2.001469135284424, + "learning_rate": 0.00014272042811946004, + "loss": 0.5977, + "step": 21831 + }, + { + "epoch": 1.5149538546943306, + "grad_norm": 2.68157958984375, + "learning_rate": 0.00014271534795202138, + "loss": 0.6836, + "step": 21832 + }, + { + "epoch": 1.5150232461314275, + "grad_norm": 1.958469271659851, + "learning_rate": 0.00014271026764973535, + "loss": 0.7517, + "step": 21833 + }, + { + "epoch": 1.515092637568524, + "grad_norm": 2.434147357940674, + "learning_rate": 0.00014270518721261794, + "loss": 0.5791, + "step": 21834 + }, + { + "epoch": 1.5151620290056207, + "grad_norm": 1.6860853433609009, + "learning_rate": 0.00014270010664068521, + "loss": 0.646, + "step": 21835 + }, + { + "epoch": 1.5152314204427175, + "grad_norm": 1.929362416267395, + "learning_rate": 0.00014269502593395323, + "loss": 0.5563, + "step": 21836 + }, + { + "epoch": 1.515300811879814, + "grad_norm": 1.3940300941467285, + "learning_rate": 0.000142689945092438, + "loss": 0.5313, + "step": 21837 + }, + { + "epoch": 1.5153702033169107, + "grad_norm": 4.197970867156982, + "learning_rate": 0.00014268486411615555, + "loss": 0.6554, + "step": 21838 + }, + { + "epoch": 1.5154395947540074, + "grad_norm": 1.5269601345062256, + "learning_rate": 0.00014267978300512192, + "loss": 0.4894, + "step": 21839 + }, + { + "epoch": 1.515508986191104, + "grad_norm": 1.591634750366211, + "learning_rate": 0.0001426747017593532, + "loss": 0.5812, + "step": 21840 + }, + { + "epoch": 1.5155783776282008, + "grad_norm": 2.444242477416992, + "learning_rate": 0.00014266962037886538, + "loss": 0.6702, + "step": 21841 + }, + { + "epoch": 1.5156477690652972, + "grad_norm": 1.4064728021621704, + "learning_rate": 0.0001426645388636745, + "loss": 0.7358, + "step": 21842 + }, + { + "epoch": 1.515717160502394, + "grad_norm": 2.877415895462036, + "learning_rate": 0.00014265945721379667, + "loss": 1.0714, + "step": 21843 + }, + { + "epoch": 1.5157865519394906, + "grad_norm": 2.358773946762085, + "learning_rate": 0.0001426543754292478, + "loss": 0.6074, + "step": 21844 + }, + { + "epoch": 1.5158559433765872, + "grad_norm": 1.8623714447021484, + "learning_rate": 0.00014264929351004408, + "loss": 0.5614, + "step": 21845 + }, + { + "epoch": 1.515925334813684, + "grad_norm": 2.0652434825897217, + "learning_rate": 0.0001426442114562015, + "loss": 0.7588, + "step": 21846 + }, + { + "epoch": 1.5159947262507807, + "grad_norm": 1.9475667476654053, + "learning_rate": 0.00014263912926773607, + "loss": 0.5724, + "step": 21847 + }, + { + "epoch": 1.5160641176878773, + "grad_norm": 1.3850712776184082, + "learning_rate": 0.00014263404694466386, + "loss": 0.499, + "step": 21848 + }, + { + "epoch": 1.5161335091249741, + "grad_norm": 1.803321361541748, + "learning_rate": 0.00014262896448700088, + "loss": 0.4087, + "step": 21849 + }, + { + "epoch": 1.5162029005620705, + "grad_norm": 1.9088020324707031, + "learning_rate": 0.00014262388189476325, + "loss": 0.6204, + "step": 21850 + }, + { + "epoch": 1.5162722919991674, + "grad_norm": 1.6228182315826416, + "learning_rate": 0.00014261879916796695, + "loss": 0.4065, + "step": 21851 + }, + { + "epoch": 1.516341683436264, + "grad_norm": 1.530465006828308, + "learning_rate": 0.00014261371630662804, + "loss": 0.6325, + "step": 21852 + }, + { + "epoch": 1.5164110748733606, + "grad_norm": 1.8438183069229126, + "learning_rate": 0.00014260863331076257, + "loss": 0.5705, + "step": 21853 + }, + { + "epoch": 1.5164804663104574, + "grad_norm": 1.4555644989013672, + "learning_rate": 0.0001426035501803866, + "loss": 0.7026, + "step": 21854 + }, + { + "epoch": 1.5165498577475538, + "grad_norm": 1.9154149293899536, + "learning_rate": 0.00014259846691551617, + "loss": 0.6942, + "step": 21855 + }, + { + "epoch": 1.5166192491846506, + "grad_norm": 1.6020344495773315, + "learning_rate": 0.0001425933835161673, + "loss": 0.5444, + "step": 21856 + }, + { + "epoch": 1.5166886406217472, + "grad_norm": 1.9399205446243286, + "learning_rate": 0.0001425882999823561, + "loss": 0.6054, + "step": 21857 + }, + { + "epoch": 1.5167580320588439, + "grad_norm": 4.130819320678711, + "learning_rate": 0.00014258321631409855, + "loss": 0.8781, + "step": 21858 + }, + { + "epoch": 1.5168274234959407, + "grad_norm": 1.55586576461792, + "learning_rate": 0.0001425781325114107, + "loss": 0.6005, + "step": 21859 + }, + { + "epoch": 1.5168968149330373, + "grad_norm": 1.5466934442520142, + "learning_rate": 0.0001425730485743087, + "loss": 0.6049, + "step": 21860 + }, + { + "epoch": 1.516966206370134, + "grad_norm": 1.5271308422088623, + "learning_rate": 0.00014256796450280845, + "loss": 0.5316, + "step": 21861 + }, + { + "epoch": 1.5170355978072307, + "grad_norm": 1.8661353588104248, + "learning_rate": 0.00014256288029692611, + "loss": 0.6311, + "step": 21862 + }, + { + "epoch": 1.5171049892443271, + "grad_norm": 2.838265895843506, + "learning_rate": 0.00014255779595667765, + "loss": 0.7991, + "step": 21863 + }, + { + "epoch": 1.517174380681424, + "grad_norm": 1.5411336421966553, + "learning_rate": 0.00014255271148207923, + "loss": 0.5963, + "step": 21864 + }, + { + "epoch": 1.5172437721185206, + "grad_norm": 2.464157819747925, + "learning_rate": 0.0001425476268731468, + "loss": 0.8473, + "step": 21865 + }, + { + "epoch": 1.5173131635556172, + "grad_norm": 2.4951670169830322, + "learning_rate": 0.00014254254212989645, + "loss": 0.4513, + "step": 21866 + }, + { + "epoch": 1.517382554992714, + "grad_norm": 1.578331708908081, + "learning_rate": 0.00014253745725234426, + "loss": 0.4868, + "step": 21867 + }, + { + "epoch": 1.5174519464298104, + "grad_norm": 1.6952775716781616, + "learning_rate": 0.00014253237224050625, + "loss": 0.5864, + "step": 21868 + }, + { + "epoch": 1.5175213378669072, + "grad_norm": 1.654994010925293, + "learning_rate": 0.00014252728709439845, + "loss": 0.7306, + "step": 21869 + }, + { + "epoch": 1.5175907293040038, + "grad_norm": 3.092836380004883, + "learning_rate": 0.00014252220181403696, + "loss": 0.7781, + "step": 21870 + }, + { + "epoch": 1.5176601207411005, + "grad_norm": 1.6362327337265015, + "learning_rate": 0.0001425171163994378, + "loss": 0.6314, + "step": 21871 + }, + { + "epoch": 1.5177295121781973, + "grad_norm": 2.4983816146850586, + "learning_rate": 0.00014251203085061703, + "loss": 0.6872, + "step": 21872 + }, + { + "epoch": 1.517798903615294, + "grad_norm": 1.517519474029541, + "learning_rate": 0.0001425069451675907, + "loss": 0.642, + "step": 21873 + }, + { + "epoch": 1.5178682950523905, + "grad_norm": 1.8018862009048462, + "learning_rate": 0.0001425018593503749, + "loss": 0.7369, + "step": 21874 + }, + { + "epoch": 1.5179376864894873, + "grad_norm": 1.6166905164718628, + "learning_rate": 0.00014249677339898565, + "loss": 0.5887, + "step": 21875 + }, + { + "epoch": 1.5180070779265837, + "grad_norm": 1.296906590461731, + "learning_rate": 0.00014249168731343903, + "loss": 0.5696, + "step": 21876 + }, + { + "epoch": 1.5180764693636806, + "grad_norm": 1.4905259609222412, + "learning_rate": 0.00014248660109375107, + "loss": 0.5279, + "step": 21877 + }, + { + "epoch": 1.5181458608007772, + "grad_norm": 1.6244796514511108, + "learning_rate": 0.00014248151473993784, + "loss": 0.7325, + "step": 21878 + }, + { + "epoch": 1.5182152522378738, + "grad_norm": 1.6175041198730469, + "learning_rate": 0.00014247642825201542, + "loss": 0.7077, + "step": 21879 + }, + { + "epoch": 1.5182846436749706, + "grad_norm": 2.009868621826172, + "learning_rate": 0.0001424713416299998, + "loss": 0.7792, + "step": 21880 + }, + { + "epoch": 1.518354035112067, + "grad_norm": 1.651928186416626, + "learning_rate": 0.00014246625487390714, + "loss": 0.6268, + "step": 21881 + }, + { + "epoch": 1.5184234265491638, + "grad_norm": 2.866072416305542, + "learning_rate": 0.00014246116798375338, + "loss": 0.4874, + "step": 21882 + }, + { + "epoch": 1.5184928179862605, + "grad_norm": 1.9470865726470947, + "learning_rate": 0.00014245608095955466, + "loss": 0.6782, + "step": 21883 + }, + { + "epoch": 1.518562209423357, + "grad_norm": 1.5692336559295654, + "learning_rate": 0.00014245099380132703, + "loss": 0.8265, + "step": 21884 + }, + { + "epoch": 1.518631600860454, + "grad_norm": 1.5770986080169678, + "learning_rate": 0.00014244590650908654, + "loss": 0.5525, + "step": 21885 + }, + { + "epoch": 1.5187009922975505, + "grad_norm": 2.384540557861328, + "learning_rate": 0.00014244081908284923, + "loss": 0.735, + "step": 21886 + }, + { + "epoch": 1.5187703837346471, + "grad_norm": 1.8825262784957886, + "learning_rate": 0.00014243573152263118, + "loss": 0.608, + "step": 21887 + }, + { + "epoch": 1.518839775171744, + "grad_norm": 1.3173983097076416, + "learning_rate": 0.00014243064382844846, + "loss": 0.6377, + "step": 21888 + }, + { + "epoch": 1.5189091666088403, + "grad_norm": 1.541064739227295, + "learning_rate": 0.0001424255560003171, + "loss": 0.661, + "step": 21889 + }, + { + "epoch": 1.5189785580459372, + "grad_norm": 1.8675042390823364, + "learning_rate": 0.0001424204680382532, + "loss": 0.5048, + "step": 21890 + }, + { + "epoch": 1.5190479494830338, + "grad_norm": 1.930405855178833, + "learning_rate": 0.00014241537994227283, + "loss": 0.6181, + "step": 21891 + }, + { + "epoch": 1.5191173409201304, + "grad_norm": 1.4970190525054932, + "learning_rate": 0.000142410291712392, + "loss": 0.5552, + "step": 21892 + }, + { + "epoch": 1.5191867323572272, + "grad_norm": 1.4584760665893555, + "learning_rate": 0.00014240520334862683, + "loss": 0.418, + "step": 21893 + }, + { + "epoch": 1.5192561237943236, + "grad_norm": 2.0549612045288086, + "learning_rate": 0.0001424001148509933, + "loss": 0.6253, + "step": 21894 + }, + { + "epoch": 1.5193255152314205, + "grad_norm": 1.6900794506072998, + "learning_rate": 0.00014239502621950754, + "loss": 0.6652, + "step": 21895 + }, + { + "epoch": 1.519394906668517, + "grad_norm": 2.5206873416900635, + "learning_rate": 0.0001423899374541856, + "loss": 0.6353, + "step": 21896 + }, + { + "epoch": 1.5194642981056137, + "grad_norm": 2.051223039627075, + "learning_rate": 0.00014238484855504357, + "loss": 0.7219, + "step": 21897 + }, + { + "epoch": 1.5195336895427105, + "grad_norm": 1.4993149042129517, + "learning_rate": 0.00014237975952209746, + "loss": 0.3479, + "step": 21898 + }, + { + "epoch": 1.5196030809798071, + "grad_norm": 2.3298237323760986, + "learning_rate": 0.0001423746703553634, + "loss": 0.481, + "step": 21899 + }, + { + "epoch": 1.5196724724169037, + "grad_norm": 3.147763967514038, + "learning_rate": 0.0001423695810548574, + "loss": 0.646, + "step": 21900 + }, + { + "epoch": 1.5197418638540006, + "grad_norm": 1.938828468322754, + "learning_rate": 0.00014236449162059555, + "loss": 0.8053, + "step": 21901 + }, + { + "epoch": 1.519811255291097, + "grad_norm": 1.5793300867080688, + "learning_rate": 0.00014235940205259393, + "loss": 0.5844, + "step": 21902 + }, + { + "epoch": 1.5198806467281938, + "grad_norm": 2.941305637359619, + "learning_rate": 0.00014235431235086856, + "loss": 0.7904, + "step": 21903 + }, + { + "epoch": 1.5199500381652904, + "grad_norm": 2.06540584564209, + "learning_rate": 0.00014234922251543558, + "loss": 0.7898, + "step": 21904 + }, + { + "epoch": 1.520019429602387, + "grad_norm": 1.5444822311401367, + "learning_rate": 0.00014234413254631098, + "loss": 0.4936, + "step": 21905 + }, + { + "epoch": 1.5200888210394838, + "grad_norm": 1.3412882089614868, + "learning_rate": 0.00014233904244351087, + "loss": 0.4764, + "step": 21906 + }, + { + "epoch": 1.5201582124765802, + "grad_norm": 2.482327938079834, + "learning_rate": 0.00014233395220705137, + "loss": 0.9267, + "step": 21907 + }, + { + "epoch": 1.520227603913677, + "grad_norm": 2.044069528579712, + "learning_rate": 0.00014232886183694846, + "loss": 0.5425, + "step": 21908 + }, + { + "epoch": 1.5202969953507737, + "grad_norm": 2.04587459564209, + "learning_rate": 0.00014232377133321824, + "loss": 0.721, + "step": 21909 + }, + { + "epoch": 1.5203663867878703, + "grad_norm": 1.9787579774856567, + "learning_rate": 0.00014231868069587678, + "loss": 0.6683, + "step": 21910 + }, + { + "epoch": 1.5204357782249671, + "grad_norm": 1.8425405025482178, + "learning_rate": 0.00014231358992494015, + "loss": 0.6074, + "step": 21911 + }, + { + "epoch": 1.5205051696620637, + "grad_norm": 1.8880105018615723, + "learning_rate": 0.00014230849902042444, + "loss": 0.5794, + "step": 21912 + }, + { + "epoch": 1.5205745610991603, + "grad_norm": 1.5642234086990356, + "learning_rate": 0.0001423034079823457, + "loss": 0.5691, + "step": 21913 + }, + { + "epoch": 1.5206439525362572, + "grad_norm": 1.718988299369812, + "learning_rate": 0.00014229831681071998, + "loss": 0.6749, + "step": 21914 + }, + { + "epoch": 1.5207133439733536, + "grad_norm": 3.6997742652893066, + "learning_rate": 0.00014229322550556343, + "loss": 0.4774, + "step": 21915 + }, + { + "epoch": 1.5207827354104504, + "grad_norm": 1.7775120735168457, + "learning_rate": 0.00014228813406689205, + "loss": 0.7289, + "step": 21916 + }, + { + "epoch": 1.520852126847547, + "grad_norm": 2.405416250228882, + "learning_rate": 0.00014228304249472195, + "loss": 0.7219, + "step": 21917 + }, + { + "epoch": 1.5209215182846436, + "grad_norm": 1.7376132011413574, + "learning_rate": 0.00014227795078906917, + "loss": 0.6387, + "step": 21918 + }, + { + "epoch": 1.5209909097217404, + "grad_norm": 2.05314302444458, + "learning_rate": 0.00014227285894994983, + "loss": 0.6783, + "step": 21919 + }, + { + "epoch": 1.5210603011588368, + "grad_norm": 1.945576548576355, + "learning_rate": 0.00014226776697737995, + "loss": 0.6693, + "step": 21920 + }, + { + "epoch": 1.5211296925959337, + "grad_norm": 1.8342688083648682, + "learning_rate": 0.00014226267487137565, + "loss": 0.662, + "step": 21921 + }, + { + "epoch": 1.5211990840330303, + "grad_norm": 2.403244733810425, + "learning_rate": 0.00014225758263195298, + "loss": 0.6835, + "step": 21922 + }, + { + "epoch": 1.521268475470127, + "grad_norm": 1.9636458158493042, + "learning_rate": 0.00014225249025912805, + "loss": 0.5868, + "step": 21923 + }, + { + "epoch": 1.5213378669072237, + "grad_norm": 1.2133090496063232, + "learning_rate": 0.00014224739775291692, + "loss": 0.5518, + "step": 21924 + }, + { + "epoch": 1.5214072583443203, + "grad_norm": 1.7908116579055786, + "learning_rate": 0.0001422423051133356, + "loss": 0.5885, + "step": 21925 + }, + { + "epoch": 1.521476649781417, + "grad_norm": 11.285310745239258, + "learning_rate": 0.00014223721234040028, + "loss": 0.5387, + "step": 21926 + }, + { + "epoch": 1.5215460412185138, + "grad_norm": 1.9939647912979126, + "learning_rate": 0.00014223211943412695, + "loss": 0.8566, + "step": 21927 + }, + { + "epoch": 1.5216154326556102, + "grad_norm": 1.5176283121109009, + "learning_rate": 0.00014222702639453175, + "loss": 0.6542, + "step": 21928 + }, + { + "epoch": 1.521684824092707, + "grad_norm": 1.67366361618042, + "learning_rate": 0.00014222193322163066, + "loss": 0.6421, + "step": 21929 + }, + { + "epoch": 1.5217542155298036, + "grad_norm": 1.261275053024292, + "learning_rate": 0.00014221683991543987, + "loss": 0.5802, + "step": 21930 + }, + { + "epoch": 1.5218236069669002, + "grad_norm": 1.8147978782653809, + "learning_rate": 0.0001422117464759754, + "loss": 0.6242, + "step": 21931 + }, + { + "epoch": 1.521892998403997, + "grad_norm": 2.1130154132843018, + "learning_rate": 0.00014220665290325338, + "loss": 0.5056, + "step": 21932 + }, + { + "epoch": 1.5219623898410934, + "grad_norm": 1.626439094543457, + "learning_rate": 0.00014220155919728985, + "loss": 0.7385, + "step": 21933 + }, + { + "epoch": 1.5220317812781903, + "grad_norm": 1.9796082973480225, + "learning_rate": 0.00014219646535810085, + "loss": 0.7368, + "step": 21934 + }, + { + "epoch": 1.5221011727152869, + "grad_norm": 1.8011205196380615, + "learning_rate": 0.0001421913713857025, + "loss": 0.6711, + "step": 21935 + }, + { + "epoch": 1.5221705641523835, + "grad_norm": 1.5962283611297607, + "learning_rate": 0.00014218627728011094, + "loss": 0.6559, + "step": 21936 + }, + { + "epoch": 1.5222399555894803, + "grad_norm": 1.791020154953003, + "learning_rate": 0.00014218118304134216, + "loss": 0.5609, + "step": 21937 + }, + { + "epoch": 1.522309347026577, + "grad_norm": 2.480926990509033, + "learning_rate": 0.0001421760886694123, + "loss": 0.8734, + "step": 21938 + }, + { + "epoch": 1.5223787384636736, + "grad_norm": 2.345752000808716, + "learning_rate": 0.0001421709941643374, + "loss": 0.8304, + "step": 21939 + }, + { + "epoch": 1.5224481299007704, + "grad_norm": 1.7842371463775635, + "learning_rate": 0.0001421658995261336, + "loss": 0.6717, + "step": 21940 + }, + { + "epoch": 1.5225175213378668, + "grad_norm": 1.9187589883804321, + "learning_rate": 0.0001421608047548169, + "loss": 0.7907, + "step": 21941 + }, + { + "epoch": 1.5225869127749636, + "grad_norm": 1.5074204206466675, + "learning_rate": 0.00014215570985040348, + "loss": 0.4955, + "step": 21942 + }, + { + "epoch": 1.5226563042120602, + "grad_norm": 1.6354678869247437, + "learning_rate": 0.0001421506148129093, + "loss": 0.5474, + "step": 21943 + }, + { + "epoch": 1.5227256956491568, + "grad_norm": 1.7281792163848877, + "learning_rate": 0.00014214551964235062, + "loss": 0.4413, + "step": 21944 + }, + { + "epoch": 1.5227950870862537, + "grad_norm": 1.380313515663147, + "learning_rate": 0.00014214042433874336, + "loss": 0.4986, + "step": 21945 + }, + { + "epoch": 1.5228644785233503, + "grad_norm": 2.274646282196045, + "learning_rate": 0.00014213532890210365, + "loss": 0.6431, + "step": 21946 + }, + { + "epoch": 1.5229338699604469, + "grad_norm": 2.0500872135162354, + "learning_rate": 0.00014213023333244765, + "loss": 0.6825, + "step": 21947 + }, + { + "epoch": 1.5230032613975435, + "grad_norm": 1.9747310876846313, + "learning_rate": 0.00014212513762979135, + "loss": 0.7083, + "step": 21948 + }, + { + "epoch": 1.52307265283464, + "grad_norm": 1.656905174255371, + "learning_rate": 0.0001421200417941509, + "loss": 0.57, + "step": 21949 + }, + { + "epoch": 1.523142044271737, + "grad_norm": 2.225616693496704, + "learning_rate": 0.00014211494582554236, + "loss": 0.7047, + "step": 21950 + }, + { + "epoch": 1.5232114357088335, + "grad_norm": 1.5801277160644531, + "learning_rate": 0.00014210984972398184, + "loss": 0.5382, + "step": 21951 + }, + { + "epoch": 1.5232808271459302, + "grad_norm": 2.8779401779174805, + "learning_rate": 0.00014210475348948537, + "loss": 0.6605, + "step": 21952 + }, + { + "epoch": 1.523350218583027, + "grad_norm": 2.4502930641174316, + "learning_rate": 0.0001420996571220691, + "loss": 0.74, + "step": 21953 + }, + { + "epoch": 1.5234196100201234, + "grad_norm": 1.999325156211853, + "learning_rate": 0.0001420945606217491, + "loss": 0.4665, + "step": 21954 + }, + { + "epoch": 1.5234890014572202, + "grad_norm": 2.855647325515747, + "learning_rate": 0.00014208946398854146, + "loss": 0.7512, + "step": 21955 + }, + { + "epoch": 1.5235583928943168, + "grad_norm": 1.6509652137756348, + "learning_rate": 0.00014208436722246227, + "loss": 0.622, + "step": 21956 + }, + { + "epoch": 1.5236277843314134, + "grad_norm": 1.56626558303833, + "learning_rate": 0.00014207927032352762, + "loss": 0.6651, + "step": 21957 + }, + { + "epoch": 1.5236971757685103, + "grad_norm": 1.8424268960952759, + "learning_rate": 0.00014207417329175357, + "loss": 0.7577, + "step": 21958 + }, + { + "epoch": 1.5237665672056069, + "grad_norm": 2.6697728633880615, + "learning_rate": 0.00014206907612715624, + "loss": 0.6256, + "step": 21959 + }, + { + "epoch": 1.5238359586427035, + "grad_norm": 2.4016737937927246, + "learning_rate": 0.00014206397882975173, + "loss": 0.8987, + "step": 21960 + }, + { + "epoch": 1.5239053500798, + "grad_norm": 1.5554747581481934, + "learning_rate": 0.00014205888139955612, + "loss": 0.5832, + "step": 21961 + }, + { + "epoch": 1.5239747415168967, + "grad_norm": 1.442121982574463, + "learning_rate": 0.00014205378383658549, + "loss": 0.5429, + "step": 21962 + }, + { + "epoch": 1.5240441329539935, + "grad_norm": 2.0002293586730957, + "learning_rate": 0.00014204868614085597, + "loss": 0.6317, + "step": 21963 + }, + { + "epoch": 1.5241135243910902, + "grad_norm": 1.6844356060028076, + "learning_rate": 0.0001420435883123836, + "loss": 0.6413, + "step": 21964 + }, + { + "epoch": 1.5241829158281868, + "grad_norm": 1.5980503559112549, + "learning_rate": 0.00014203849035118447, + "loss": 0.6827, + "step": 21965 + }, + { + "epoch": 1.5242523072652836, + "grad_norm": 1.47346830368042, + "learning_rate": 0.00014203339225727474, + "loss": 0.5882, + "step": 21966 + }, + { + "epoch": 1.52432169870238, + "grad_norm": 2.2792270183563232, + "learning_rate": 0.0001420282940306705, + "loss": 0.6988, + "step": 21967 + }, + { + "epoch": 1.5243910901394768, + "grad_norm": 2.0175163745880127, + "learning_rate": 0.00014202319567138778, + "loss": 0.6735, + "step": 21968 + }, + { + "epoch": 1.5244604815765734, + "grad_norm": 1.33473539352417, + "learning_rate": 0.00014201809717944268, + "loss": 0.512, + "step": 21969 + }, + { + "epoch": 1.52452987301367, + "grad_norm": 1.8950241804122925, + "learning_rate": 0.00014201299855485134, + "loss": 0.7464, + "step": 21970 + }, + { + "epoch": 1.5245992644507669, + "grad_norm": 1.799226999282837, + "learning_rate": 0.00014200789979762986, + "loss": 0.5239, + "step": 21971 + }, + { + "epoch": 1.5246686558878635, + "grad_norm": 2.53678560256958, + "learning_rate": 0.0001420028009077943, + "loss": 0.7372, + "step": 21972 + }, + { + "epoch": 1.52473804732496, + "grad_norm": 1.7007997035980225, + "learning_rate": 0.00014199770188536075, + "loss": 0.548, + "step": 21973 + }, + { + "epoch": 1.5248074387620567, + "grad_norm": 2.338400363922119, + "learning_rate": 0.00014199260273034536, + "loss": 0.8506, + "step": 21974 + }, + { + "epoch": 1.5248768301991533, + "grad_norm": 1.8895716667175293, + "learning_rate": 0.0001419875034427642, + "loss": 0.8125, + "step": 21975 + }, + { + "epoch": 1.5249462216362502, + "grad_norm": 1.7080577611923218, + "learning_rate": 0.00014198240402263333, + "loss": 0.5879, + "step": 21976 + }, + { + "epoch": 1.5250156130733468, + "grad_norm": 2.3511908054351807, + "learning_rate": 0.0001419773044699689, + "loss": 0.7015, + "step": 21977 + }, + { + "epoch": 1.5250850045104434, + "grad_norm": 1.7917951345443726, + "learning_rate": 0.00014197220478478698, + "loss": 0.5391, + "step": 21978 + }, + { + "epoch": 1.5251543959475402, + "grad_norm": 1.7181577682495117, + "learning_rate": 0.0001419671049671037, + "loss": 0.5909, + "step": 21979 + }, + { + "epoch": 1.5252237873846366, + "grad_norm": 1.921303629875183, + "learning_rate": 0.00014196200501693512, + "loss": 0.7948, + "step": 21980 + }, + { + "epoch": 1.5252931788217334, + "grad_norm": 2.017948865890503, + "learning_rate": 0.00014195690493429739, + "loss": 0.6851, + "step": 21981 + }, + { + "epoch": 1.52536257025883, + "grad_norm": 1.5423957109451294, + "learning_rate": 0.00014195180471920652, + "loss": 0.5211, + "step": 21982 + }, + { + "epoch": 1.5254319616959267, + "grad_norm": 1.549202561378479, + "learning_rate": 0.00014194670437167869, + "loss": 0.4777, + "step": 21983 + }, + { + "epoch": 1.5255013531330235, + "grad_norm": 1.6575870513916016, + "learning_rate": 0.00014194160389173, + "loss": 0.5209, + "step": 21984 + }, + { + "epoch": 1.52557074457012, + "grad_norm": 1.5202975273132324, + "learning_rate": 0.00014193650327937657, + "loss": 0.6741, + "step": 21985 + }, + { + "epoch": 1.5256401360072167, + "grad_norm": 2.2232131958007812, + "learning_rate": 0.00014193140253463438, + "loss": 0.683, + "step": 21986 + }, + { + "epoch": 1.5257095274443133, + "grad_norm": 1.738411784172058, + "learning_rate": 0.00014192630165751967, + "loss": 0.6533, + "step": 21987 + }, + { + "epoch": 1.52577891888141, + "grad_norm": 1.8786561489105225, + "learning_rate": 0.00014192120064804848, + "loss": 0.4772, + "step": 21988 + }, + { + "epoch": 1.5258483103185068, + "grad_norm": 1.9459530115127563, + "learning_rate": 0.00014191609950623692, + "loss": 0.5855, + "step": 21989 + }, + { + "epoch": 1.5259177017556034, + "grad_norm": 1.71768057346344, + "learning_rate": 0.00014191099823210107, + "loss": 0.7207, + "step": 21990 + }, + { + "epoch": 1.5259870931927, + "grad_norm": 2.3091073036193848, + "learning_rate": 0.0001419058968256571, + "loss": 0.6112, + "step": 21991 + }, + { + "epoch": 1.5260564846297968, + "grad_norm": 1.9255784749984741, + "learning_rate": 0.0001419007952869211, + "loss": 0.6977, + "step": 21992 + }, + { + "epoch": 1.5261258760668932, + "grad_norm": 1.9521501064300537, + "learning_rate": 0.00014189569361590908, + "loss": 0.6424, + "step": 21993 + }, + { + "epoch": 1.52619526750399, + "grad_norm": 1.705527424812317, + "learning_rate": 0.00014189059181263727, + "loss": 0.71, + "step": 21994 + }, + { + "epoch": 1.5262646589410866, + "grad_norm": 1.5144091844558716, + "learning_rate": 0.00014188548987712168, + "loss": 0.5969, + "step": 21995 + }, + { + "epoch": 1.5263340503781833, + "grad_norm": 1.5597717761993408, + "learning_rate": 0.0001418803878093785, + "loss": 0.5855, + "step": 21996 + }, + { + "epoch": 1.52640344181528, + "grad_norm": 2.0379903316497803, + "learning_rate": 0.00014187528560942377, + "loss": 0.7429, + "step": 21997 + }, + { + "epoch": 1.5264728332523767, + "grad_norm": 1.5292433500289917, + "learning_rate": 0.00014187018327727362, + "loss": 0.5279, + "step": 21998 + }, + { + "epoch": 1.5265422246894733, + "grad_norm": 1.839338779449463, + "learning_rate": 0.0001418650808129442, + "loss": 0.7486, + "step": 21999 + }, + { + "epoch": 1.52661161612657, + "grad_norm": 2.309736967086792, + "learning_rate": 0.00014185997821645154, + "loss": 0.6428, + "step": 22000 + }, + { + "epoch": 1.5266810075636665, + "grad_norm": 1.9658974409103394, + "learning_rate": 0.0001418548754878118, + "loss": 0.6038, + "step": 22001 + }, + { + "epoch": 1.5267503990007634, + "grad_norm": 1.7197483777999878, + "learning_rate": 0.00014184977262704106, + "loss": 0.6753, + "step": 22002 + }, + { + "epoch": 1.52681979043786, + "grad_norm": 1.7001211643218994, + "learning_rate": 0.0001418446696341554, + "loss": 0.5149, + "step": 22003 + }, + { + "epoch": 1.5268891818749566, + "grad_norm": 1.6503340005874634, + "learning_rate": 0.00014183956650917106, + "loss": 0.5357, + "step": 22004 + }, + { + "epoch": 1.5269585733120534, + "grad_norm": 2.2778780460357666, + "learning_rate": 0.000141834463252104, + "loss": 0.5737, + "step": 22005 + }, + { + "epoch": 1.5270279647491498, + "grad_norm": 1.5013717412948608, + "learning_rate": 0.00014182935986297042, + "loss": 0.6878, + "step": 22006 + }, + { + "epoch": 1.5270973561862466, + "grad_norm": 1.5633516311645508, + "learning_rate": 0.0001418242563417864, + "loss": 0.5836, + "step": 22007 + }, + { + "epoch": 1.5271667476233433, + "grad_norm": 1.4602724313735962, + "learning_rate": 0.00014181915268856804, + "loss": 0.7453, + "step": 22008 + }, + { + "epoch": 1.5272361390604399, + "grad_norm": 1.9186804294586182, + "learning_rate": 0.0001418140489033315, + "loss": 0.6171, + "step": 22009 + }, + { + "epoch": 1.5273055304975367, + "grad_norm": 1.4468278884887695, + "learning_rate": 0.00014180894498609282, + "loss": 0.5327, + "step": 22010 + }, + { + "epoch": 1.5273749219346333, + "grad_norm": 2.136284112930298, + "learning_rate": 0.00014180384093686818, + "loss": 0.6824, + "step": 22011 + }, + { + "epoch": 1.52744431337173, + "grad_norm": 2.1399683952331543, + "learning_rate": 0.00014179873675567367, + "loss": 0.7559, + "step": 22012 + }, + { + "epoch": 1.5275137048088268, + "grad_norm": 2.0453603267669678, + "learning_rate": 0.00014179363244252538, + "loss": 0.5128, + "step": 22013 + }, + { + "epoch": 1.5275830962459231, + "grad_norm": 1.8264765739440918, + "learning_rate": 0.00014178852799743948, + "loss": 0.6173, + "step": 22014 + }, + { + "epoch": 1.52765248768302, + "grad_norm": 1.3881787061691284, + "learning_rate": 0.00014178342342043197, + "loss": 0.4503, + "step": 22015 + }, + { + "epoch": 1.5277218791201166, + "grad_norm": 2.4503934383392334, + "learning_rate": 0.0001417783187115191, + "loss": 0.7442, + "step": 22016 + }, + { + "epoch": 1.5277912705572132, + "grad_norm": 2.3613016605377197, + "learning_rate": 0.0001417732138707169, + "loss": 0.6629, + "step": 22017 + }, + { + "epoch": 1.52786066199431, + "grad_norm": 2.517131805419922, + "learning_rate": 0.0001417681088980415, + "loss": 0.7753, + "step": 22018 + }, + { + "epoch": 1.5279300534314064, + "grad_norm": 1.5583395957946777, + "learning_rate": 0.00014176300379350906, + "loss": 0.8907, + "step": 22019 + }, + { + "epoch": 1.5279994448685033, + "grad_norm": 1.4740749597549438, + "learning_rate": 0.00014175789855713566, + "loss": 0.5974, + "step": 22020 + }, + { + "epoch": 1.5280688363055999, + "grad_norm": 1.5630465745925903, + "learning_rate": 0.0001417527931889374, + "loss": 0.7743, + "step": 22021 + }, + { + "epoch": 1.5281382277426965, + "grad_norm": 1.2459735870361328, + "learning_rate": 0.00014174768768893041, + "loss": 0.5175, + "step": 22022 + }, + { + "epoch": 1.5282076191797933, + "grad_norm": 2.5566623210906982, + "learning_rate": 0.00014174258205713086, + "loss": 0.6121, + "step": 22023 + }, + { + "epoch": 1.52827701061689, + "grad_norm": 1.5181185007095337, + "learning_rate": 0.00014173747629355477, + "loss": 0.6044, + "step": 22024 + }, + { + "epoch": 1.5283464020539865, + "grad_norm": 1.5413986444473267, + "learning_rate": 0.0001417323703982183, + "loss": 0.5078, + "step": 22025 + }, + { + "epoch": 1.5284157934910834, + "grad_norm": 2.1161282062530518, + "learning_rate": 0.0001417272643711376, + "loss": 0.4979, + "step": 22026 + }, + { + "epoch": 1.5284851849281798, + "grad_norm": 1.3330442905426025, + "learning_rate": 0.00014172215821232879, + "loss": 0.4191, + "step": 22027 + }, + { + "epoch": 1.5285545763652766, + "grad_norm": 1.5776381492614746, + "learning_rate": 0.00014171705192180794, + "loss": 0.6064, + "step": 22028 + }, + { + "epoch": 1.5286239678023732, + "grad_norm": 1.7377296686172485, + "learning_rate": 0.0001417119454995912, + "loss": 0.4978, + "step": 22029 + }, + { + "epoch": 1.5286933592394698, + "grad_norm": 2.0352444648742676, + "learning_rate": 0.0001417068389456947, + "loss": 0.4553, + "step": 22030 + }, + { + "epoch": 1.5287627506765666, + "grad_norm": 2.3849058151245117, + "learning_rate": 0.00014170173226013457, + "loss": 0.8779, + "step": 22031 + }, + { + "epoch": 1.528832142113663, + "grad_norm": 2.408949375152588, + "learning_rate": 0.00014169662544292685, + "loss": 0.755, + "step": 22032 + }, + { + "epoch": 1.5289015335507599, + "grad_norm": 2.01057767868042, + "learning_rate": 0.00014169151849408778, + "loss": 0.7412, + "step": 22033 + }, + { + "epoch": 1.5289709249878565, + "grad_norm": 1.6031346321105957, + "learning_rate": 0.00014168641141363338, + "loss": 0.609, + "step": 22034 + }, + { + "epoch": 1.529040316424953, + "grad_norm": 1.9271774291992188, + "learning_rate": 0.00014168130420157983, + "loss": 0.7115, + "step": 22035 + }, + { + "epoch": 1.52910970786205, + "grad_norm": 2.2424235343933105, + "learning_rate": 0.00014167619685794325, + "loss": 0.7053, + "step": 22036 + }, + { + "epoch": 1.5291790992991465, + "grad_norm": 1.634131669998169, + "learning_rate": 0.00014167108938273974, + "loss": 0.5976, + "step": 22037 + }, + { + "epoch": 1.5292484907362431, + "grad_norm": 1.6110455989837646, + "learning_rate": 0.00014166598177598543, + "loss": 0.7288, + "step": 22038 + }, + { + "epoch": 1.52931788217334, + "grad_norm": 1.401528000831604, + "learning_rate": 0.00014166087403769646, + "loss": 0.6405, + "step": 22039 + }, + { + "epoch": 1.5293872736104364, + "grad_norm": 2.2431304454803467, + "learning_rate": 0.00014165576616788895, + "loss": 0.7211, + "step": 22040 + }, + { + "epoch": 1.5294566650475332, + "grad_norm": 1.9095046520233154, + "learning_rate": 0.000141650658166579, + "loss": 0.7207, + "step": 22041 + }, + { + "epoch": 1.5295260564846298, + "grad_norm": 1.5568218231201172, + "learning_rate": 0.00014164555003378277, + "loss": 0.7964, + "step": 22042 + }, + { + "epoch": 1.5295954479217264, + "grad_norm": 1.701638102531433, + "learning_rate": 0.00014164044176951635, + "loss": 0.5077, + "step": 22043 + }, + { + "epoch": 1.5296648393588232, + "grad_norm": 2.8683135509490967, + "learning_rate": 0.00014163533337379592, + "loss": 0.7766, + "step": 22044 + }, + { + "epoch": 1.5297342307959196, + "grad_norm": 2.8350942134857178, + "learning_rate": 0.00014163022484663755, + "loss": 0.7487, + "step": 22045 + }, + { + "epoch": 1.5298036222330165, + "grad_norm": 1.2694878578186035, + "learning_rate": 0.0001416251161880574, + "loss": 0.4881, + "step": 22046 + }, + { + "epoch": 1.529873013670113, + "grad_norm": 2.2399473190307617, + "learning_rate": 0.00014162000739807157, + "loss": 0.6381, + "step": 22047 + }, + { + "epoch": 1.5299424051072097, + "grad_norm": 1.9900223016738892, + "learning_rate": 0.00014161489847669626, + "loss": 0.7226, + "step": 22048 + }, + { + "epoch": 1.5300117965443065, + "grad_norm": 1.4944936037063599, + "learning_rate": 0.00014160978942394747, + "loss": 0.8042, + "step": 22049 + }, + { + "epoch": 1.5300811879814031, + "grad_norm": 1.6524136066436768, + "learning_rate": 0.00014160468023984146, + "loss": 0.5272, + "step": 22050 + }, + { + "epoch": 1.5301505794184997, + "grad_norm": 1.3719077110290527, + "learning_rate": 0.00014159957092439425, + "loss": 0.5643, + "step": 22051 + }, + { + "epoch": 1.5302199708555966, + "grad_norm": 1.3345061540603638, + "learning_rate": 0.00014159446147762205, + "loss": 0.5069, + "step": 22052 + }, + { + "epoch": 1.530289362292693, + "grad_norm": 2.1865789890289307, + "learning_rate": 0.00014158935189954094, + "loss": 0.5376, + "step": 22053 + }, + { + "epoch": 1.5303587537297898, + "grad_norm": 1.5254656076431274, + "learning_rate": 0.0001415842421901671, + "loss": 0.785, + "step": 22054 + }, + { + "epoch": 1.5304281451668864, + "grad_norm": 1.4760223627090454, + "learning_rate": 0.00014157913234951662, + "loss": 0.5858, + "step": 22055 + }, + { + "epoch": 1.530497536603983, + "grad_norm": 1.6133112907409668, + "learning_rate": 0.00014157402237760565, + "loss": 0.4766, + "step": 22056 + }, + { + "epoch": 1.5305669280410799, + "grad_norm": 2.4065701961517334, + "learning_rate": 0.0001415689122744503, + "loss": 0.7082, + "step": 22057 + }, + { + "epoch": 1.5306363194781762, + "grad_norm": 1.8655214309692383, + "learning_rate": 0.00014156380204006675, + "loss": 0.7085, + "step": 22058 + }, + { + "epoch": 1.530705710915273, + "grad_norm": 2.0139217376708984, + "learning_rate": 0.00014155869167447108, + "loss": 0.8018, + "step": 22059 + }, + { + "epoch": 1.5307751023523697, + "grad_norm": 1.7668468952178955, + "learning_rate": 0.0001415535811776794, + "loss": 0.5679, + "step": 22060 + }, + { + "epoch": 1.5308444937894663, + "grad_norm": 1.834032654762268, + "learning_rate": 0.00014154847054970794, + "loss": 0.6212, + "step": 22061 + }, + { + "epoch": 1.5309138852265631, + "grad_norm": 1.5767340660095215, + "learning_rate": 0.00014154335979057277, + "loss": 0.658, + "step": 22062 + }, + { + "epoch": 1.5309832766636597, + "grad_norm": 1.4643315076828003, + "learning_rate": 0.00014153824890029003, + "loss": 0.5133, + "step": 22063 + }, + { + "epoch": 1.5310526681007564, + "grad_norm": 1.722136378288269, + "learning_rate": 0.00014153313787887583, + "loss": 0.6184, + "step": 22064 + }, + { + "epoch": 1.5311220595378532, + "grad_norm": 1.575280785560608, + "learning_rate": 0.00014152802672634635, + "loss": 0.5895, + "step": 22065 + }, + { + "epoch": 1.5311914509749496, + "grad_norm": 1.287507176399231, + "learning_rate": 0.00014152291544271774, + "loss": 0.5371, + "step": 22066 + }, + { + "epoch": 1.5312608424120464, + "grad_norm": 1.1065179109573364, + "learning_rate": 0.00014151780402800606, + "loss": 0.4302, + "step": 22067 + }, + { + "epoch": 1.531330233849143, + "grad_norm": 1.7617002725601196, + "learning_rate": 0.0001415126924822275, + "loss": 0.5939, + "step": 22068 + }, + { + "epoch": 1.5313996252862396, + "grad_norm": 1.9468543529510498, + "learning_rate": 0.00014150758080539821, + "loss": 0.7621, + "step": 22069 + }, + { + "epoch": 1.5314690167233365, + "grad_norm": 1.6623339653015137, + "learning_rate": 0.00014150246899753428, + "loss": 0.4474, + "step": 22070 + }, + { + "epoch": 1.5315384081604329, + "grad_norm": 1.555401086807251, + "learning_rate": 0.00014149735705865188, + "loss": 0.4615, + "step": 22071 + }, + { + "epoch": 1.5316077995975297, + "grad_norm": 2.6193642616271973, + "learning_rate": 0.00014149224498876715, + "loss": 0.6634, + "step": 22072 + }, + { + "epoch": 1.5316771910346263, + "grad_norm": 2.194457769393921, + "learning_rate": 0.0001414871327878962, + "loss": 0.844, + "step": 22073 + }, + { + "epoch": 1.531746582471723, + "grad_norm": 1.7049745321273804, + "learning_rate": 0.00014148202045605516, + "loss": 0.6874, + "step": 22074 + }, + { + "epoch": 1.5318159739088197, + "grad_norm": 1.7515249252319336, + "learning_rate": 0.00014147690799326022, + "loss": 0.5971, + "step": 22075 + }, + { + "epoch": 1.5318853653459163, + "grad_norm": 1.856995701789856, + "learning_rate": 0.0001414717953995275, + "loss": 0.8126, + "step": 22076 + }, + { + "epoch": 1.531954756783013, + "grad_norm": 3.008993625640869, + "learning_rate": 0.00014146668267487312, + "loss": 0.5523, + "step": 22077 + }, + { + "epoch": 1.5320241482201098, + "grad_norm": 1.4478998184204102, + "learning_rate": 0.00014146156981931327, + "loss": 0.5293, + "step": 22078 + }, + { + "epoch": 1.5320935396572062, + "grad_norm": 1.703850269317627, + "learning_rate": 0.00014145645683286403, + "loss": 0.5687, + "step": 22079 + }, + { + "epoch": 1.532162931094303, + "grad_norm": 1.2857152223587036, + "learning_rate": 0.00014145134371554153, + "loss": 0.4555, + "step": 22080 + }, + { + "epoch": 1.5322323225313996, + "grad_norm": 1.6007024049758911, + "learning_rate": 0.00014144623046736198, + "loss": 0.7704, + "step": 22081 + }, + { + "epoch": 1.5323017139684962, + "grad_norm": 1.6702524423599243, + "learning_rate": 0.00014144111708834148, + "loss": 0.6226, + "step": 22082 + }, + { + "epoch": 1.532371105405593, + "grad_norm": 1.9273653030395508, + "learning_rate": 0.0001414360035784962, + "loss": 0.58, + "step": 22083 + }, + { + "epoch": 1.5324404968426895, + "grad_norm": 1.6157945394515991, + "learning_rate": 0.00014143088993784226, + "loss": 0.6907, + "step": 22084 + }, + { + "epoch": 1.5325098882797863, + "grad_norm": 1.6029229164123535, + "learning_rate": 0.0001414257761663958, + "loss": 0.5879, + "step": 22085 + }, + { + "epoch": 1.532579279716883, + "grad_norm": 4.727777481079102, + "learning_rate": 0.00014142066226417298, + "loss": 0.557, + "step": 22086 + }, + { + "epoch": 1.5326486711539795, + "grad_norm": 2.0080792903900146, + "learning_rate": 0.00014141554823118992, + "loss": 0.7387, + "step": 22087 + }, + { + "epoch": 1.5327180625910763, + "grad_norm": 1.7954524755477905, + "learning_rate": 0.0001414104340674628, + "loss": 0.6131, + "step": 22088 + }, + { + "epoch": 1.532787454028173, + "grad_norm": 1.5086551904678345, + "learning_rate": 0.00014140531977300773, + "loss": 0.6182, + "step": 22089 + }, + { + "epoch": 1.5328568454652696, + "grad_norm": 1.490478754043579, + "learning_rate": 0.00014140020534784083, + "loss": 0.5592, + "step": 22090 + }, + { + "epoch": 1.5329262369023664, + "grad_norm": 2.2953202724456787, + "learning_rate": 0.00014139509079197834, + "loss": 0.7775, + "step": 22091 + }, + { + "epoch": 1.5329956283394628, + "grad_norm": 2.2083730697631836, + "learning_rate": 0.00014138997610543634, + "loss": 0.9158, + "step": 22092 + }, + { + "epoch": 1.5330650197765596, + "grad_norm": 1.9564898014068604, + "learning_rate": 0.000141384861288231, + "loss": 0.5704, + "step": 22093 + }, + { + "epoch": 1.5331344112136562, + "grad_norm": 1.6792699098587036, + "learning_rate": 0.0001413797463403784, + "loss": 0.6199, + "step": 22094 + }, + { + "epoch": 1.5332038026507528, + "grad_norm": 6.4352803230285645, + "learning_rate": 0.0001413746312618948, + "loss": 0.6478, + "step": 22095 + }, + { + "epoch": 1.5332731940878497, + "grad_norm": 1.8524225950241089, + "learning_rate": 0.00014136951605279627, + "loss": 0.7352, + "step": 22096 + }, + { + "epoch": 1.533342585524946, + "grad_norm": 1.9669193029403687, + "learning_rate": 0.00014136440071309898, + "loss": 0.7069, + "step": 22097 + }, + { + "epoch": 1.533411976962043, + "grad_norm": 1.5637001991271973, + "learning_rate": 0.00014135928524281906, + "loss": 0.5913, + "step": 22098 + }, + { + "epoch": 1.5334813683991395, + "grad_norm": 1.6672230958938599, + "learning_rate": 0.0001413541696419727, + "loss": 0.5685, + "step": 22099 + }, + { + "epoch": 1.5335507598362361, + "grad_norm": 1.3477534055709839, + "learning_rate": 0.00014134905391057601, + "loss": 0.4544, + "step": 22100 + }, + { + "epoch": 1.533620151273333, + "grad_norm": 3.3146913051605225, + "learning_rate": 0.00014134393804864515, + "loss": 0.8035, + "step": 22101 + }, + { + "epoch": 1.5336895427104296, + "grad_norm": 1.5571317672729492, + "learning_rate": 0.0001413388220561963, + "loss": 0.5455, + "step": 22102 + }, + { + "epoch": 1.5337589341475262, + "grad_norm": 1.5741115808486938, + "learning_rate": 0.00014133370593324556, + "loss": 0.7262, + "step": 22103 + }, + { + "epoch": 1.533828325584623, + "grad_norm": 2.1844375133514404, + "learning_rate": 0.0001413285896798091, + "loss": 0.9354, + "step": 22104 + }, + { + "epoch": 1.5338977170217194, + "grad_norm": 1.901394009590149, + "learning_rate": 0.0001413234732959031, + "loss": 0.6783, + "step": 22105 + }, + { + "epoch": 1.5339671084588162, + "grad_norm": 2.2638094425201416, + "learning_rate": 0.00014131835678154369, + "loss": 0.8628, + "step": 22106 + }, + { + "epoch": 1.5340364998959128, + "grad_norm": 1.9417129755020142, + "learning_rate": 0.00014131324013674697, + "loss": 0.6573, + "step": 22107 + }, + { + "epoch": 1.5341058913330095, + "grad_norm": 1.6053228378295898, + "learning_rate": 0.0001413081233615292, + "loss": 0.6566, + "step": 22108 + }, + { + "epoch": 1.5341752827701063, + "grad_norm": 2.2307379245758057, + "learning_rate": 0.00014130300645590646, + "loss": 0.3448, + "step": 22109 + }, + { + "epoch": 1.5342446742072027, + "grad_norm": 2.001896619796753, + "learning_rate": 0.0001412978894198949, + "loss": 0.6122, + "step": 22110 + }, + { + "epoch": 1.5343140656442995, + "grad_norm": 1.594063639640808, + "learning_rate": 0.0001412927722535107, + "loss": 0.6431, + "step": 22111 + }, + { + "epoch": 1.5343834570813961, + "grad_norm": 2.1976659297943115, + "learning_rate": 0.00014128765495677006, + "loss": 0.7243, + "step": 22112 + }, + { + "epoch": 1.5344528485184927, + "grad_norm": 2.1426424980163574, + "learning_rate": 0.00014128253752968906, + "loss": 0.7605, + "step": 22113 + }, + { + "epoch": 1.5345222399555896, + "grad_norm": 1.5245475769042969, + "learning_rate": 0.00014127741997228385, + "loss": 0.8102, + "step": 22114 + }, + { + "epoch": 1.5345916313926862, + "grad_norm": 1.806379795074463, + "learning_rate": 0.00014127230228457062, + "loss": 0.7178, + "step": 22115 + }, + { + "epoch": 1.5346610228297828, + "grad_norm": 2.0196564197540283, + "learning_rate": 0.00014126718446656555, + "loss": 0.9138, + "step": 22116 + }, + { + "epoch": 1.5347304142668796, + "grad_norm": 5.185642719268799, + "learning_rate": 0.00014126206651828474, + "loss": 0.8247, + "step": 22117 + }, + { + "epoch": 1.534799805703976, + "grad_norm": 1.526566505432129, + "learning_rate": 0.0001412569484397444, + "loss": 0.282, + "step": 22118 + }, + { + "epoch": 1.5348691971410728, + "grad_norm": 1.665457010269165, + "learning_rate": 0.00014125183023096062, + "loss": 0.8123, + "step": 22119 + }, + { + "epoch": 1.5349385885781694, + "grad_norm": 1.3542418479919434, + "learning_rate": 0.00014124671189194964, + "loss": 0.4051, + "step": 22120 + }, + { + "epoch": 1.535007980015266, + "grad_norm": 1.6500182151794434, + "learning_rate": 0.00014124159342272753, + "loss": 0.5723, + "step": 22121 + }, + { + "epoch": 1.535077371452363, + "grad_norm": 1.955481767654419, + "learning_rate": 0.00014123647482331054, + "loss": 0.7014, + "step": 22122 + }, + { + "epoch": 1.5351467628894593, + "grad_norm": 1.493597388267517, + "learning_rate": 0.00014123135609371475, + "loss": 0.5164, + "step": 22123 + }, + { + "epoch": 1.5352161543265561, + "grad_norm": 1.560842514038086, + "learning_rate": 0.00014122623723395638, + "loss": 0.6061, + "step": 22124 + }, + { + "epoch": 1.5352855457636527, + "grad_norm": 1.7085142135620117, + "learning_rate": 0.00014122111824405154, + "loss": 0.6474, + "step": 22125 + }, + { + "epoch": 1.5353549372007493, + "grad_norm": 1.3656651973724365, + "learning_rate": 0.00014121599912401644, + "loss": 0.4896, + "step": 22126 + }, + { + "epoch": 1.5354243286378462, + "grad_norm": 1.7594376802444458, + "learning_rate": 0.0001412108798738672, + "loss": 0.8074, + "step": 22127 + }, + { + "epoch": 1.5354937200749428, + "grad_norm": 1.520195722579956, + "learning_rate": 0.00014120576049361994, + "loss": 0.4591, + "step": 22128 + }, + { + "epoch": 1.5355631115120394, + "grad_norm": 1.7554377317428589, + "learning_rate": 0.00014120064098329093, + "loss": 0.6883, + "step": 22129 + }, + { + "epoch": 1.5356325029491362, + "grad_norm": 1.8980294466018677, + "learning_rate": 0.00014119552134289628, + "loss": 0.7436, + "step": 22130 + }, + { + "epoch": 1.5357018943862326, + "grad_norm": 1.9752590656280518, + "learning_rate": 0.00014119040157245213, + "loss": 0.6011, + "step": 22131 + }, + { + "epoch": 1.5357712858233294, + "grad_norm": 2.214641809463501, + "learning_rate": 0.00014118528167197467, + "loss": 0.8806, + "step": 22132 + }, + { + "epoch": 1.535840677260426, + "grad_norm": 2.1036927700042725, + "learning_rate": 0.00014118016164148007, + "loss": 0.4247, + "step": 22133 + }, + { + "epoch": 1.5359100686975227, + "grad_norm": 1.9719924926757812, + "learning_rate": 0.00014117504148098445, + "loss": 0.6704, + "step": 22134 + }, + { + "epoch": 1.5359794601346195, + "grad_norm": 1.2891603708267212, + "learning_rate": 0.00014116992119050402, + "loss": 0.5458, + "step": 22135 + }, + { + "epoch": 1.536048851571716, + "grad_norm": 1.2558715343475342, + "learning_rate": 0.0001411648007700549, + "loss": 0.519, + "step": 22136 + }, + { + "epoch": 1.5361182430088127, + "grad_norm": 1.8452398777008057, + "learning_rate": 0.0001411596802196533, + "loss": 0.7957, + "step": 22137 + }, + { + "epoch": 1.5361876344459093, + "grad_norm": 2.169315814971924, + "learning_rate": 0.00014115455953931538, + "loss": 0.6281, + "step": 22138 + }, + { + "epoch": 1.536257025883006, + "grad_norm": 2.1673450469970703, + "learning_rate": 0.00014114943872905726, + "loss": 0.7622, + "step": 22139 + }, + { + "epoch": 1.5363264173201028, + "grad_norm": 1.982295274734497, + "learning_rate": 0.00014114431778889514, + "loss": 0.666, + "step": 22140 + }, + { + "epoch": 1.5363958087571994, + "grad_norm": 1.9746241569519043, + "learning_rate": 0.00014113919671884518, + "loss": 0.4338, + "step": 22141 + }, + { + "epoch": 1.536465200194296, + "grad_norm": 1.7350980043411255, + "learning_rate": 0.00014113407551892356, + "loss": 0.586, + "step": 22142 + }, + { + "epoch": 1.5365345916313928, + "grad_norm": 2.0026779174804688, + "learning_rate": 0.00014112895418914645, + "loss": 0.6572, + "step": 22143 + }, + { + "epoch": 1.5366039830684892, + "grad_norm": 1.8265186548233032, + "learning_rate": 0.00014112383272952997, + "loss": 0.4069, + "step": 22144 + }, + { + "epoch": 1.536673374505586, + "grad_norm": 1.951187252998352, + "learning_rate": 0.00014111871114009036, + "loss": 0.7279, + "step": 22145 + }, + { + "epoch": 1.5367427659426827, + "grad_norm": 1.2141143083572388, + "learning_rate": 0.0001411135894208437, + "loss": 0.4661, + "step": 22146 + }, + { + "epoch": 1.5368121573797793, + "grad_norm": 1.7178657054901123, + "learning_rate": 0.00014110846757180626, + "loss": 0.5526, + "step": 22147 + }, + { + "epoch": 1.536881548816876, + "grad_norm": 1.7210211753845215, + "learning_rate": 0.00014110334559299414, + "loss": 0.4857, + "step": 22148 + }, + { + "epoch": 1.5369509402539725, + "grad_norm": 1.8903422355651855, + "learning_rate": 0.0001410982234844235, + "loss": 0.6108, + "step": 22149 + }, + { + "epoch": 1.5370203316910693, + "grad_norm": 1.5478945970535278, + "learning_rate": 0.00014109310124611057, + "loss": 0.7745, + "step": 22150 + }, + { + "epoch": 1.537089723128166, + "grad_norm": 1.6311248540878296, + "learning_rate": 0.00014108797887807148, + "loss": 0.6838, + "step": 22151 + }, + { + "epoch": 1.5371591145652626, + "grad_norm": 1.435826301574707, + "learning_rate": 0.0001410828563803224, + "loss": 0.6106, + "step": 22152 + }, + { + "epoch": 1.5372285060023594, + "grad_norm": 1.5674341917037964, + "learning_rate": 0.0001410777337528795, + "loss": 0.6848, + "step": 22153 + }, + { + "epoch": 1.537297897439456, + "grad_norm": 1.635141372680664, + "learning_rate": 0.00014107261099575899, + "loss": 0.6534, + "step": 22154 + }, + { + "epoch": 1.5373672888765526, + "grad_norm": 2.049684524536133, + "learning_rate": 0.00014106748810897702, + "loss": 0.7202, + "step": 22155 + }, + { + "epoch": 1.5374366803136494, + "grad_norm": 2.143332004547119, + "learning_rate": 0.0001410623650925497, + "loss": 0.5944, + "step": 22156 + }, + { + "epoch": 1.5375060717507458, + "grad_norm": 1.7434000968933105, + "learning_rate": 0.0001410572419464933, + "loss": 0.7526, + "step": 22157 + }, + { + "epoch": 1.5375754631878427, + "grad_norm": 1.807361364364624, + "learning_rate": 0.00014105211867082392, + "loss": 0.7305, + "step": 22158 + }, + { + "epoch": 1.5376448546249393, + "grad_norm": 1.8927282094955444, + "learning_rate": 0.0001410469952655578, + "loss": 0.7044, + "step": 22159 + }, + { + "epoch": 1.5377142460620359, + "grad_norm": 1.4807699918746948, + "learning_rate": 0.00014104187173071106, + "loss": 0.636, + "step": 22160 + }, + { + "epoch": 1.5377836374991327, + "grad_norm": 1.8201922178268433, + "learning_rate": 0.0001410367480662999, + "loss": 0.7857, + "step": 22161 + }, + { + "epoch": 1.537853028936229, + "grad_norm": 2.1262757778167725, + "learning_rate": 0.00014103162427234046, + "loss": 0.3841, + "step": 22162 + }, + { + "epoch": 1.537922420373326, + "grad_norm": 2.394146680831909, + "learning_rate": 0.00014102650034884896, + "loss": 0.6566, + "step": 22163 + }, + { + "epoch": 1.5379918118104225, + "grad_norm": 2.7827517986297607, + "learning_rate": 0.0001410213762958416, + "loss": 0.882, + "step": 22164 + }, + { + "epoch": 1.5380612032475192, + "grad_norm": 1.4172745943069458, + "learning_rate": 0.00014101625211333443, + "loss": 0.4892, + "step": 22165 + }, + { + "epoch": 1.538130594684616, + "grad_norm": 1.3376383781433105, + "learning_rate": 0.00014101112780134378, + "loss": 0.5767, + "step": 22166 + }, + { + "epoch": 1.5381999861217126, + "grad_norm": 2.1016147136688232, + "learning_rate": 0.00014100600335988572, + "loss": 0.5524, + "step": 22167 + }, + { + "epoch": 1.5382693775588092, + "grad_norm": 1.9379088878631592, + "learning_rate": 0.0001410008787889765, + "loss": 0.6435, + "step": 22168 + }, + { + "epoch": 1.538338768995906, + "grad_norm": 1.8396433591842651, + "learning_rate": 0.00014099575408863224, + "loss": 0.7884, + "step": 22169 + }, + { + "epoch": 1.5384081604330024, + "grad_norm": 1.7151763439178467, + "learning_rate": 0.00014099062925886915, + "loss": 0.3999, + "step": 22170 + }, + { + "epoch": 1.5384775518700993, + "grad_norm": 1.929962396621704, + "learning_rate": 0.0001409855042997034, + "loss": 0.6298, + "step": 22171 + }, + { + "epoch": 1.5385469433071959, + "grad_norm": 1.8312098979949951, + "learning_rate": 0.00014098037921115114, + "loss": 0.7168, + "step": 22172 + }, + { + "epoch": 1.5386163347442925, + "grad_norm": 1.411097526550293, + "learning_rate": 0.0001409752539932286, + "loss": 0.519, + "step": 22173 + }, + { + "epoch": 1.5386857261813893, + "grad_norm": 2.0363399982452393, + "learning_rate": 0.00014097012864595194, + "loss": 0.7758, + "step": 22174 + }, + { + "epoch": 1.5387551176184857, + "grad_norm": 1.8917686939239502, + "learning_rate": 0.00014096500316933733, + "loss": 0.9479, + "step": 22175 + }, + { + "epoch": 1.5388245090555825, + "grad_norm": 1.3161340951919556, + "learning_rate": 0.00014095987756340098, + "loss": 0.6534, + "step": 22176 + }, + { + "epoch": 1.5388939004926792, + "grad_norm": 2.0544018745422363, + "learning_rate": 0.000140954751828159, + "loss": 0.7131, + "step": 22177 + }, + { + "epoch": 1.5389632919297758, + "grad_norm": 1.3364564180374146, + "learning_rate": 0.00014094962596362766, + "loss": 0.6846, + "step": 22178 + }, + { + "epoch": 1.5390326833668726, + "grad_norm": 2.1307084560394287, + "learning_rate": 0.00014094449996982308, + "loss": 0.9106, + "step": 22179 + }, + { + "epoch": 1.5391020748039692, + "grad_norm": 1.943743348121643, + "learning_rate": 0.0001409393738467615, + "loss": 0.735, + "step": 22180 + }, + { + "epoch": 1.5391714662410658, + "grad_norm": 2.1674818992614746, + "learning_rate": 0.00014093424759445902, + "loss": 0.7042, + "step": 22181 + }, + { + "epoch": 1.5392408576781627, + "grad_norm": 1.437727928161621, + "learning_rate": 0.0001409291212129319, + "loss": 0.4449, + "step": 22182 + }, + { + "epoch": 1.539310249115259, + "grad_norm": 1.6221941709518433, + "learning_rate": 0.00014092399470219625, + "loss": 0.5989, + "step": 22183 + }, + { + "epoch": 1.5393796405523559, + "grad_norm": 1.3724725246429443, + "learning_rate": 0.00014091886806226836, + "loss": 0.5933, + "step": 22184 + }, + { + "epoch": 1.5394490319894525, + "grad_norm": 1.3661718368530273, + "learning_rate": 0.00014091374129316433, + "loss": 0.6277, + "step": 22185 + }, + { + "epoch": 1.539518423426549, + "grad_norm": 1.5465208292007446, + "learning_rate": 0.00014090861439490033, + "loss": 0.7491, + "step": 22186 + }, + { + "epoch": 1.539587814863646, + "grad_norm": 1.9314504861831665, + "learning_rate": 0.0001409034873674926, + "loss": 0.7295, + "step": 22187 + }, + { + "epoch": 1.5396572063007425, + "grad_norm": 1.7620258331298828, + "learning_rate": 0.00014089836021095734, + "loss": 0.6173, + "step": 22188 + }, + { + "epoch": 1.5397265977378392, + "grad_norm": 1.7760941982269287, + "learning_rate": 0.0001408932329253107, + "loss": 0.5928, + "step": 22189 + }, + { + "epoch": 1.5397959891749358, + "grad_norm": 2.0403103828430176, + "learning_rate": 0.00014088810551056883, + "loss": 0.7495, + "step": 22190 + }, + { + "epoch": 1.5398653806120324, + "grad_norm": 1.8829388618469238, + "learning_rate": 0.00014088297796674797, + "loss": 0.7784, + "step": 22191 + }, + { + "epoch": 1.5399347720491292, + "grad_norm": 1.383949637413025, + "learning_rate": 0.0001408778502938643, + "loss": 0.4837, + "step": 22192 + }, + { + "epoch": 1.5400041634862258, + "grad_norm": 1.8541761636734009, + "learning_rate": 0.00014087272249193399, + "loss": 0.64, + "step": 22193 + }, + { + "epoch": 1.5400735549233224, + "grad_norm": 1.9388890266418457, + "learning_rate": 0.00014086759456097321, + "loss": 0.8162, + "step": 22194 + }, + { + "epoch": 1.5401429463604193, + "grad_norm": 1.6051605939865112, + "learning_rate": 0.00014086246650099818, + "loss": 0.7206, + "step": 22195 + }, + { + "epoch": 1.5402123377975157, + "grad_norm": 1.5918818712234497, + "learning_rate": 0.00014085733831202513, + "loss": 0.6196, + "step": 22196 + }, + { + "epoch": 1.5402817292346125, + "grad_norm": 1.7601206302642822, + "learning_rate": 0.0001408522099940702, + "loss": 0.6917, + "step": 22197 + }, + { + "epoch": 1.540351120671709, + "grad_norm": 1.6413941383361816, + "learning_rate": 0.00014084708154714953, + "loss": 0.6622, + "step": 22198 + }, + { + "epoch": 1.5404205121088057, + "grad_norm": 1.7657686471939087, + "learning_rate": 0.00014084195297127942, + "loss": 0.7249, + "step": 22199 + }, + { + "epoch": 1.5404899035459025, + "grad_norm": 1.5782499313354492, + "learning_rate": 0.000140836824266476, + "loss": 0.6467, + "step": 22200 + }, + { + "epoch": 1.5405592949829992, + "grad_norm": 1.312617540359497, + "learning_rate": 0.00014083169543275542, + "loss": 0.521, + "step": 22201 + }, + { + "epoch": 1.5406286864200958, + "grad_norm": 1.9863982200622559, + "learning_rate": 0.00014082656647013394, + "loss": 0.4505, + "step": 22202 + }, + { + "epoch": 1.5406980778571924, + "grad_norm": 2.040287971496582, + "learning_rate": 0.0001408214373786277, + "loss": 0.5658, + "step": 22203 + }, + { + "epoch": 1.540767469294289, + "grad_norm": 1.1875760555267334, + "learning_rate": 0.00014081630815825295, + "loss": 0.5508, + "step": 22204 + }, + { + "epoch": 1.5408368607313858, + "grad_norm": 1.396004319190979, + "learning_rate": 0.00014081117880902584, + "loss": 0.5627, + "step": 22205 + }, + { + "epoch": 1.5409062521684824, + "grad_norm": 1.9434425830841064, + "learning_rate": 0.0001408060493309626, + "loss": 0.5586, + "step": 22206 + }, + { + "epoch": 1.540975643605579, + "grad_norm": 1.6113576889038086, + "learning_rate": 0.00014080091972407937, + "loss": 0.5054, + "step": 22207 + }, + { + "epoch": 1.5410450350426759, + "grad_norm": 1.4798132181167603, + "learning_rate": 0.00014079578998839237, + "loss": 0.5419, + "step": 22208 + }, + { + "epoch": 1.5411144264797723, + "grad_norm": 1.305991768836975, + "learning_rate": 0.0001407906601239178, + "loss": 0.5384, + "step": 22209 + }, + { + "epoch": 1.541183817916869, + "grad_norm": 1.356768250465393, + "learning_rate": 0.00014078553013067185, + "loss": 0.5578, + "step": 22210 + }, + { + "epoch": 1.5412532093539657, + "grad_norm": 1.8447966575622559, + "learning_rate": 0.0001407804000086707, + "loss": 0.7018, + "step": 22211 + }, + { + "epoch": 1.5413226007910623, + "grad_norm": 1.6591070890426636, + "learning_rate": 0.00014077526975793056, + "loss": 0.7931, + "step": 22212 + }, + { + "epoch": 1.5413919922281591, + "grad_norm": 1.8666354417800903, + "learning_rate": 0.00014077013937846767, + "loss": 0.6131, + "step": 22213 + }, + { + "epoch": 1.5414613836652558, + "grad_norm": 1.7938944101333618, + "learning_rate": 0.0001407650088702981, + "loss": 0.6633, + "step": 22214 + }, + { + "epoch": 1.5415307751023524, + "grad_norm": 2.620875358581543, + "learning_rate": 0.0001407598782334382, + "loss": 0.7748, + "step": 22215 + }, + { + "epoch": 1.541600166539449, + "grad_norm": 1.4276732206344604, + "learning_rate": 0.00014075474746790407, + "loss": 0.6023, + "step": 22216 + }, + { + "epoch": 1.5416695579765456, + "grad_norm": 1.751685619354248, + "learning_rate": 0.00014074961657371192, + "loss": 0.5342, + "step": 22217 + }, + { + "epoch": 1.5417389494136424, + "grad_norm": 1.6283104419708252, + "learning_rate": 0.00014074448555087793, + "loss": 0.6775, + "step": 22218 + }, + { + "epoch": 1.541808340850739, + "grad_norm": 1.3537659645080566, + "learning_rate": 0.00014073935439941834, + "loss": 0.4291, + "step": 22219 + }, + { + "epoch": 1.5418777322878356, + "grad_norm": 1.8132160902023315, + "learning_rate": 0.00014073422311934938, + "loss": 0.5595, + "step": 22220 + }, + { + "epoch": 1.5419471237249325, + "grad_norm": 1.7967159748077393, + "learning_rate": 0.00014072909171068717, + "loss": 0.7284, + "step": 22221 + }, + { + "epoch": 1.5420165151620289, + "grad_norm": 1.3273087739944458, + "learning_rate": 0.00014072396017344798, + "loss": 0.4414, + "step": 22222 + }, + { + "epoch": 1.5420859065991257, + "grad_norm": 1.9456249475479126, + "learning_rate": 0.00014071882850764792, + "loss": 0.9898, + "step": 22223 + }, + { + "epoch": 1.5421552980362223, + "grad_norm": 1.7916803359985352, + "learning_rate": 0.00014071369671330328, + "loss": 0.572, + "step": 22224 + }, + { + "epoch": 1.542224689473319, + "grad_norm": 1.7885571718215942, + "learning_rate": 0.00014070856479043019, + "loss": 0.638, + "step": 22225 + }, + { + "epoch": 1.5422940809104158, + "grad_norm": 2.0349884033203125, + "learning_rate": 0.0001407034327390449, + "loss": 0.9042, + "step": 22226 + }, + { + "epoch": 1.5423634723475124, + "grad_norm": 1.100748896598816, + "learning_rate": 0.00014069830055916363, + "loss": 0.3594, + "step": 22227 + }, + { + "epoch": 1.542432863784609, + "grad_norm": 1.666340947151184, + "learning_rate": 0.0001406931682508025, + "loss": 0.5561, + "step": 22228 + }, + { + "epoch": 1.5425022552217056, + "grad_norm": 1.8834184408187866, + "learning_rate": 0.00014068803581397777, + "loss": 0.7958, + "step": 22229 + }, + { + "epoch": 1.5425716466588022, + "grad_norm": 1.5905307531356812, + "learning_rate": 0.00014068290324870564, + "loss": 0.5385, + "step": 22230 + }, + { + "epoch": 1.542641038095899, + "grad_norm": 1.3196831941604614, + "learning_rate": 0.0001406777705550023, + "loss": 0.562, + "step": 22231 + }, + { + "epoch": 1.5427104295329956, + "grad_norm": 2.422348976135254, + "learning_rate": 0.00014067263773288391, + "loss": 0.7613, + "step": 22232 + }, + { + "epoch": 1.5427798209700923, + "grad_norm": 1.5423895120620728, + "learning_rate": 0.00014066750478236678, + "loss": 0.6088, + "step": 22233 + }, + { + "epoch": 1.542849212407189, + "grad_norm": 1.959360122680664, + "learning_rate": 0.00014066237170346705, + "loss": 0.8305, + "step": 22234 + }, + { + "epoch": 1.5429186038442855, + "grad_norm": 2.012486696243286, + "learning_rate": 0.00014065723849620093, + "loss": 0.7535, + "step": 22235 + }, + { + "epoch": 1.5429879952813823, + "grad_norm": 1.5921261310577393, + "learning_rate": 0.00014065210516058462, + "loss": 0.4179, + "step": 22236 + }, + { + "epoch": 1.543057386718479, + "grad_norm": 1.6480793952941895, + "learning_rate": 0.00014064697169663434, + "loss": 0.5642, + "step": 22237 + }, + { + "epoch": 1.5431267781555755, + "grad_norm": 1.1961692571640015, + "learning_rate": 0.00014064183810436626, + "loss": 0.4834, + "step": 22238 + }, + { + "epoch": 1.5431961695926724, + "grad_norm": 1.952702522277832, + "learning_rate": 0.00014063670438379662, + "loss": 0.5525, + "step": 22239 + }, + { + "epoch": 1.543265561029769, + "grad_norm": 1.7064273357391357, + "learning_rate": 0.00014063157053494164, + "loss": 0.6394, + "step": 22240 + }, + { + "epoch": 1.5433349524668656, + "grad_norm": 1.6188032627105713, + "learning_rate": 0.00014062643655781752, + "loss": 0.4999, + "step": 22241 + }, + { + "epoch": 1.5434043439039624, + "grad_norm": 1.570229411125183, + "learning_rate": 0.0001406213024524404, + "loss": 0.7261, + "step": 22242 + }, + { + "epoch": 1.5434737353410588, + "grad_norm": 1.4025591611862183, + "learning_rate": 0.0001406161682188266, + "loss": 0.561, + "step": 22243 + }, + { + "epoch": 1.5435431267781556, + "grad_norm": 1.5386178493499756, + "learning_rate": 0.00014061103385699223, + "loss": 0.5462, + "step": 22244 + }, + { + "epoch": 1.5436125182152522, + "grad_norm": 1.8058056831359863, + "learning_rate": 0.00014060589936695355, + "loss": 0.7176, + "step": 22245 + }, + { + "epoch": 1.5436819096523489, + "grad_norm": 1.6260401010513306, + "learning_rate": 0.00014060076474872675, + "loss": 0.6628, + "step": 22246 + }, + { + "epoch": 1.5437513010894457, + "grad_norm": 1.458409309387207, + "learning_rate": 0.00014059563000232806, + "loss": 0.4931, + "step": 22247 + }, + { + "epoch": 1.543820692526542, + "grad_norm": 1.639043927192688, + "learning_rate": 0.00014059049512777364, + "loss": 0.5061, + "step": 22248 + }, + { + "epoch": 1.543890083963639, + "grad_norm": 1.6260638236999512, + "learning_rate": 0.00014058536012507978, + "loss": 0.5651, + "step": 22249 + }, + { + "epoch": 1.5439594754007355, + "grad_norm": 1.7696789503097534, + "learning_rate": 0.0001405802249942626, + "loss": 0.4832, + "step": 22250 + }, + { + "epoch": 1.5440288668378321, + "grad_norm": 1.5760389566421509, + "learning_rate": 0.0001405750897353384, + "loss": 0.6424, + "step": 22251 + }, + { + "epoch": 1.544098258274929, + "grad_norm": 1.942494511604309, + "learning_rate": 0.00014056995434832335, + "loss": 0.5618, + "step": 22252 + }, + { + "epoch": 1.5441676497120256, + "grad_norm": 2.4120914936065674, + "learning_rate": 0.00014056481883323362, + "loss": 0.7482, + "step": 22253 + }, + { + "epoch": 1.5442370411491222, + "grad_norm": 2.0792179107666016, + "learning_rate": 0.0001405596831900855, + "loss": 1.0608, + "step": 22254 + }, + { + "epoch": 1.544306432586219, + "grad_norm": 2.9190514087677, + "learning_rate": 0.00014055454741889516, + "loss": 0.7978, + "step": 22255 + }, + { + "epoch": 1.5443758240233154, + "grad_norm": 1.6833256483078003, + "learning_rate": 0.0001405494115196788, + "loss": 0.5302, + "step": 22256 + }, + { + "epoch": 1.5444452154604122, + "grad_norm": 1.854509949684143, + "learning_rate": 0.00014054427549245268, + "loss": 0.5355, + "step": 22257 + }, + { + "epoch": 1.5445146068975089, + "grad_norm": 1.9765617847442627, + "learning_rate": 0.00014053913933723299, + "loss": 0.8062, + "step": 22258 + }, + { + "epoch": 1.5445839983346055, + "grad_norm": 1.4901490211486816, + "learning_rate": 0.0001405340030540359, + "loss": 0.6625, + "step": 22259 + }, + { + "epoch": 1.5446533897717023, + "grad_norm": 2.1852798461914062, + "learning_rate": 0.00014052886664287767, + "loss": 0.7118, + "step": 22260 + }, + { + "epoch": 1.5447227812087987, + "grad_norm": 1.4392409324645996, + "learning_rate": 0.00014052373010377457, + "loss": 0.4186, + "step": 22261 + }, + { + "epoch": 1.5447921726458955, + "grad_norm": 2.0073037147521973, + "learning_rate": 0.00014051859343674268, + "loss": 0.6027, + "step": 22262 + }, + { + "epoch": 1.5448615640829921, + "grad_norm": 1.8044931888580322, + "learning_rate": 0.00014051345664179834, + "loss": 0.6079, + "step": 22263 + }, + { + "epoch": 1.5449309555200887, + "grad_norm": 1.629504680633545, + "learning_rate": 0.0001405083197189577, + "loss": 0.5624, + "step": 22264 + }, + { + "epoch": 1.5450003469571856, + "grad_norm": 1.8968689441680908, + "learning_rate": 0.000140503182668237, + "loss": 0.5422, + "step": 22265 + }, + { + "epoch": 1.5450697383942822, + "grad_norm": 1.766559362411499, + "learning_rate": 0.00014049804548965246, + "loss": 0.7497, + "step": 22266 + }, + { + "epoch": 1.5451391298313788, + "grad_norm": 1.559870958328247, + "learning_rate": 0.00014049290818322026, + "loss": 0.6285, + "step": 22267 + }, + { + "epoch": 1.5452085212684756, + "grad_norm": 2.345937490463257, + "learning_rate": 0.0001404877707489567, + "loss": 0.7968, + "step": 22268 + }, + { + "epoch": 1.545277912705572, + "grad_norm": 2.294449806213379, + "learning_rate": 0.0001404826331868779, + "loss": 0.7176, + "step": 22269 + }, + { + "epoch": 1.5453473041426689, + "grad_norm": 1.3800878524780273, + "learning_rate": 0.0001404774954970001, + "loss": 0.5694, + "step": 22270 + }, + { + "epoch": 1.5454166955797655, + "grad_norm": 1.5187410116195679, + "learning_rate": 0.0001404723576793396, + "loss": 0.4884, + "step": 22271 + }, + { + "epoch": 1.545486087016862, + "grad_norm": 2.422933578491211, + "learning_rate": 0.00014046721973391254, + "loss": 0.703, + "step": 22272 + }, + { + "epoch": 1.545555478453959, + "grad_norm": 1.7321232557296753, + "learning_rate": 0.00014046208166073517, + "loss": 0.5541, + "step": 22273 + }, + { + "epoch": 1.5456248698910553, + "grad_norm": 1.1486459970474243, + "learning_rate": 0.00014045694345982368, + "loss": 0.5079, + "step": 22274 + }, + { + "epoch": 1.5456942613281521, + "grad_norm": 1.4464129209518433, + "learning_rate": 0.00014045180513119434, + "loss": 0.5611, + "step": 22275 + }, + { + "epoch": 1.5457636527652487, + "grad_norm": 1.9313385486602783, + "learning_rate": 0.00014044666667486333, + "loss": 0.7099, + "step": 22276 + }, + { + "epoch": 1.5458330442023454, + "grad_norm": 1.7166999578475952, + "learning_rate": 0.0001404415280908469, + "loss": 0.5173, + "step": 22277 + }, + { + "epoch": 1.5459024356394422, + "grad_norm": 1.5456162691116333, + "learning_rate": 0.00014043638937916126, + "loss": 0.6256, + "step": 22278 + }, + { + "epoch": 1.5459718270765388, + "grad_norm": 1.977553367614746, + "learning_rate": 0.0001404312505398226, + "loss": 0.8961, + "step": 22279 + }, + { + "epoch": 1.5460412185136354, + "grad_norm": 1.2026361227035522, + "learning_rate": 0.0001404261115728472, + "loss": 0.4471, + "step": 22280 + }, + { + "epoch": 1.5461106099507322, + "grad_norm": 1.1243396997451782, + "learning_rate": 0.00014042097247825123, + "loss": 0.4773, + "step": 22281 + }, + { + "epoch": 1.5461800013878286, + "grad_norm": 1.6217429637908936, + "learning_rate": 0.00014041583325605095, + "loss": 0.5432, + "step": 22282 + }, + { + "epoch": 1.5462493928249255, + "grad_norm": 1.756330132484436, + "learning_rate": 0.0001404106939062626, + "loss": 0.6248, + "step": 22283 + }, + { + "epoch": 1.546318784262022, + "grad_norm": 1.412993311882019, + "learning_rate": 0.00014040555442890233, + "loss": 0.4468, + "step": 22284 + }, + { + "epoch": 1.5463881756991187, + "grad_norm": 1.4157609939575195, + "learning_rate": 0.00014040041482398646, + "loss": 0.5698, + "step": 22285 + }, + { + "epoch": 1.5464575671362155, + "grad_norm": 1.643589735031128, + "learning_rate": 0.00014039527509153115, + "loss": 0.5535, + "step": 22286 + }, + { + "epoch": 1.546526958573312, + "grad_norm": 1.588789463043213, + "learning_rate": 0.00014039013523155266, + "loss": 0.603, + "step": 22287 + }, + { + "epoch": 1.5465963500104087, + "grad_norm": 1.1473546028137207, + "learning_rate": 0.00014038499524406713, + "loss": 0.4041, + "step": 22288 + }, + { + "epoch": 1.5466657414475053, + "grad_norm": 1.2396011352539062, + "learning_rate": 0.00014037985512909093, + "loss": 0.5585, + "step": 22289 + }, + { + "epoch": 1.546735132884602, + "grad_norm": 2.336775779724121, + "learning_rate": 0.00014037471488664017, + "loss": 0.7718, + "step": 22290 + }, + { + "epoch": 1.5468045243216988, + "grad_norm": 2.439343214035034, + "learning_rate": 0.00014036957451673114, + "loss": 0.8848, + "step": 22291 + }, + { + "epoch": 1.5468739157587954, + "grad_norm": 1.572983741760254, + "learning_rate": 0.00014036443401938006, + "loss": 0.5677, + "step": 22292 + }, + { + "epoch": 1.546943307195892, + "grad_norm": 1.3081564903259277, + "learning_rate": 0.00014035929339460313, + "loss": 0.5728, + "step": 22293 + }, + { + "epoch": 1.5470126986329888, + "grad_norm": 1.4848710298538208, + "learning_rate": 0.00014035415264241662, + "loss": 0.741, + "step": 22294 + }, + { + "epoch": 1.5470820900700852, + "grad_norm": 1.9596178531646729, + "learning_rate": 0.00014034901176283667, + "loss": 0.6655, + "step": 22295 + }, + { + "epoch": 1.547151481507182, + "grad_norm": 1.8072341680526733, + "learning_rate": 0.0001403438707558796, + "loss": 0.6365, + "step": 22296 + }, + { + "epoch": 1.5472208729442787, + "grad_norm": 2.363163948059082, + "learning_rate": 0.00014033872962156165, + "loss": 0.5593, + "step": 22297 + }, + { + "epoch": 1.5472902643813753, + "grad_norm": 4.287585735321045, + "learning_rate": 0.00014033358835989897, + "loss": 0.8059, + "step": 22298 + }, + { + "epoch": 1.5473596558184721, + "grad_norm": 1.5717726945877075, + "learning_rate": 0.00014032844697090782, + "loss": 0.7216, + "step": 22299 + }, + { + "epoch": 1.5474290472555685, + "grad_norm": 1.6332271099090576, + "learning_rate": 0.00014032330545460449, + "loss": 0.4901, + "step": 22300 + }, + { + "epoch": 1.5474984386926653, + "grad_norm": 2.8749005794525146, + "learning_rate": 0.0001403181638110051, + "loss": 0.7197, + "step": 22301 + }, + { + "epoch": 1.547567830129762, + "grad_norm": 1.5793884992599487, + "learning_rate": 0.00014031302204012597, + "loss": 0.6011, + "step": 22302 + }, + { + "epoch": 1.5476372215668586, + "grad_norm": 2.0959525108337402, + "learning_rate": 0.00014030788014198335, + "loss": 0.8087, + "step": 22303 + }, + { + "epoch": 1.5477066130039554, + "grad_norm": 1.6645132303237915, + "learning_rate": 0.00014030273811659337, + "loss": 0.7291, + "step": 22304 + }, + { + "epoch": 1.547776004441052, + "grad_norm": 1.564610481262207, + "learning_rate": 0.00014029759596397235, + "loss": 0.6651, + "step": 22305 + }, + { + "epoch": 1.5478453958781486, + "grad_norm": 1.2255715131759644, + "learning_rate": 0.00014029245368413651, + "loss": 0.3967, + "step": 22306 + }, + { + "epoch": 1.5479147873152455, + "grad_norm": 1.282315731048584, + "learning_rate": 0.00014028731127710203, + "loss": 0.5265, + "step": 22307 + }, + { + "epoch": 1.5479841787523418, + "grad_norm": 2.7147319316864014, + "learning_rate": 0.0001402821687428852, + "loss": 0.5758, + "step": 22308 + }, + { + "epoch": 1.5480535701894387, + "grad_norm": 2.0113656520843506, + "learning_rate": 0.00014027702608150223, + "loss": 0.6906, + "step": 22309 + }, + { + "epoch": 1.5481229616265353, + "grad_norm": 1.4451171159744263, + "learning_rate": 0.0001402718832929694, + "loss": 0.641, + "step": 22310 + }, + { + "epoch": 1.548192353063632, + "grad_norm": 2.5789215564727783, + "learning_rate": 0.00014026674037730287, + "loss": 0.5786, + "step": 22311 + }, + { + "epoch": 1.5482617445007287, + "grad_norm": 1.4455727338790894, + "learning_rate": 0.0001402615973345189, + "loss": 0.5576, + "step": 22312 + }, + { + "epoch": 1.5483311359378251, + "grad_norm": 1.7391127347946167, + "learning_rate": 0.00014025645416463378, + "loss": 0.5808, + "step": 22313 + }, + { + "epoch": 1.548400527374922, + "grad_norm": 1.5744794607162476, + "learning_rate": 0.00014025131086766368, + "loss": 0.5552, + "step": 22314 + }, + { + "epoch": 1.5484699188120186, + "grad_norm": 1.511738657951355, + "learning_rate": 0.00014024616744362484, + "loss": 0.449, + "step": 22315 + }, + { + "epoch": 1.5485393102491152, + "grad_norm": 1.632355809211731, + "learning_rate": 0.00014024102389253358, + "loss": 0.7555, + "step": 22316 + }, + { + "epoch": 1.548608701686212, + "grad_norm": 1.97630774974823, + "learning_rate": 0.000140235880214406, + "loss": 0.701, + "step": 22317 + }, + { + "epoch": 1.5486780931233086, + "grad_norm": 1.5543702840805054, + "learning_rate": 0.00014023073640925847, + "loss": 0.6683, + "step": 22318 + }, + { + "epoch": 1.5487474845604052, + "grad_norm": 1.9478720426559448, + "learning_rate": 0.00014022559247710715, + "loss": 0.9037, + "step": 22319 + }, + { + "epoch": 1.548816875997502, + "grad_norm": 1.3125368356704712, + "learning_rate": 0.00014022044841796832, + "loss": 0.397, + "step": 22320 + }, + { + "epoch": 1.5488862674345985, + "grad_norm": 2.4183294773101807, + "learning_rate": 0.00014021530423185817, + "loss": 0.6552, + "step": 22321 + }, + { + "epoch": 1.5489556588716953, + "grad_norm": 1.1346696615219116, + "learning_rate": 0.000140210159918793, + "loss": 0.4579, + "step": 22322 + }, + { + "epoch": 1.549025050308792, + "grad_norm": 2.0192625522613525, + "learning_rate": 0.000140205015478789, + "loss": 0.6302, + "step": 22323 + }, + { + "epoch": 1.5490944417458885, + "grad_norm": 1.5396620035171509, + "learning_rate": 0.00014019987091186246, + "loss": 0.5879, + "step": 22324 + }, + { + "epoch": 1.5491638331829853, + "grad_norm": 3.625077962875366, + "learning_rate": 0.00014019472621802957, + "loss": 0.8012, + "step": 22325 + }, + { + "epoch": 1.5492332246200817, + "grad_norm": 1.7606513500213623, + "learning_rate": 0.00014018958139730658, + "loss": 0.6538, + "step": 22326 + }, + { + "epoch": 1.5493026160571786, + "grad_norm": 1.856825828552246, + "learning_rate": 0.00014018443644970975, + "loss": 0.7038, + "step": 22327 + }, + { + "epoch": 1.5493720074942752, + "grad_norm": 1.7581663131713867, + "learning_rate": 0.00014017929137525535, + "loss": 0.6973, + "step": 22328 + }, + { + "epoch": 1.5494413989313718, + "grad_norm": 5.469581604003906, + "learning_rate": 0.00014017414617395956, + "loss": 0.7264, + "step": 22329 + }, + { + "epoch": 1.5495107903684686, + "grad_norm": 1.4515892267227173, + "learning_rate": 0.00014016900084583864, + "loss": 0.6123, + "step": 22330 + }, + { + "epoch": 1.5495801818055652, + "grad_norm": 1.4201287031173706, + "learning_rate": 0.00014016385539090888, + "loss": 0.7121, + "step": 22331 + }, + { + "epoch": 1.5496495732426618, + "grad_norm": 1.7345519065856934, + "learning_rate": 0.00014015870980918645, + "loss": 0.8166, + "step": 22332 + }, + { + "epoch": 1.5497189646797587, + "grad_norm": 1.9077227115631104, + "learning_rate": 0.00014015356410068764, + "loss": 0.8534, + "step": 22333 + }, + { + "epoch": 1.549788356116855, + "grad_norm": 1.5688594579696655, + "learning_rate": 0.00014014841826542874, + "loss": 0.571, + "step": 22334 + }, + { + "epoch": 1.549857747553952, + "grad_norm": 2.260509490966797, + "learning_rate": 0.0001401432723034259, + "loss": 0.611, + "step": 22335 + }, + { + "epoch": 1.5499271389910485, + "grad_norm": 1.2877957820892334, + "learning_rate": 0.00014013812621469537, + "loss": 0.5308, + "step": 22336 + }, + { + "epoch": 1.5499965304281451, + "grad_norm": 1.250525712966919, + "learning_rate": 0.0001401329799992535, + "loss": 0.5906, + "step": 22337 + }, + { + "epoch": 1.550065921865242, + "grad_norm": 1.5854415893554688, + "learning_rate": 0.00014012783365711644, + "loss": 0.5418, + "step": 22338 + }, + { + "epoch": 1.5501353133023383, + "grad_norm": 2.6319358348846436, + "learning_rate": 0.00014012268718830044, + "loss": 0.9792, + "step": 22339 + }, + { + "epoch": 1.5502047047394352, + "grad_norm": 1.9380475282669067, + "learning_rate": 0.0001401175405928218, + "loss": 0.8721, + "step": 22340 + }, + { + "epoch": 1.5502740961765318, + "grad_norm": 2.5772786140441895, + "learning_rate": 0.00014011239387069673, + "loss": 0.7569, + "step": 22341 + }, + { + "epoch": 1.5503434876136284, + "grad_norm": 2.763547897338867, + "learning_rate": 0.0001401072470219415, + "loss": 0.8913, + "step": 22342 + }, + { + "epoch": 1.5504128790507252, + "grad_norm": 2.2919304370880127, + "learning_rate": 0.0001401021000465723, + "loss": 0.5063, + "step": 22343 + }, + { + "epoch": 1.5504822704878218, + "grad_norm": 3.216257333755493, + "learning_rate": 0.0001400969529446055, + "loss": 0.8021, + "step": 22344 + }, + { + "epoch": 1.5505516619249184, + "grad_norm": 2.025432586669922, + "learning_rate": 0.0001400918057160572, + "loss": 0.7542, + "step": 22345 + }, + { + "epoch": 1.5506210533620153, + "grad_norm": 1.296686053276062, + "learning_rate": 0.00014008665836094376, + "loss": 0.5574, + "step": 22346 + }, + { + "epoch": 1.5506904447991117, + "grad_norm": 1.7145330905914307, + "learning_rate": 0.0001400815108792814, + "loss": 0.6662, + "step": 22347 + }, + { + "epoch": 1.5507598362362085, + "grad_norm": 2.2190327644348145, + "learning_rate": 0.00014007636327108633, + "loss": 0.6638, + "step": 22348 + }, + { + "epoch": 1.550829227673305, + "grad_norm": 1.7145991325378418, + "learning_rate": 0.00014007121553637482, + "loss": 0.5759, + "step": 22349 + }, + { + "epoch": 1.5508986191104017, + "grad_norm": 2.66457462310791, + "learning_rate": 0.00014006606767516316, + "loss": 0.7187, + "step": 22350 + }, + { + "epoch": 1.5509680105474986, + "grad_norm": 2.1166205406188965, + "learning_rate": 0.0001400609196874676, + "loss": 0.6108, + "step": 22351 + }, + { + "epoch": 1.551037401984595, + "grad_norm": 1.2829203605651855, + "learning_rate": 0.0001400557715733043, + "loss": 0.4663, + "step": 22352 + }, + { + "epoch": 1.5511067934216918, + "grad_norm": 1.8796509504318237, + "learning_rate": 0.00014005062333268963, + "loss": 0.6778, + "step": 22353 + }, + { + "epoch": 1.5511761848587884, + "grad_norm": 1.47195565700531, + "learning_rate": 0.00014004547496563975, + "loss": 0.5728, + "step": 22354 + }, + { + "epoch": 1.551245576295885, + "grad_norm": 2.3155195713043213, + "learning_rate": 0.000140040326472171, + "loss": 0.7258, + "step": 22355 + }, + { + "epoch": 1.5513149677329818, + "grad_norm": 2.7844271659851074, + "learning_rate": 0.0001400351778522995, + "loss": 0.6312, + "step": 22356 + }, + { + "epoch": 1.5513843591700784, + "grad_norm": 1.537264347076416, + "learning_rate": 0.00014003002910604163, + "loss": 0.6817, + "step": 22357 + }, + { + "epoch": 1.551453750607175, + "grad_norm": 1.7864536046981812, + "learning_rate": 0.00014002488023341365, + "loss": 0.6573, + "step": 22358 + }, + { + "epoch": 1.5515231420442719, + "grad_norm": 2.0481514930725098, + "learning_rate": 0.00014001973123443172, + "loss": 0.7907, + "step": 22359 + }, + { + "epoch": 1.5515925334813683, + "grad_norm": 1.655166506767273, + "learning_rate": 0.00014001458210911212, + "loss": 0.4083, + "step": 22360 + }, + { + "epoch": 1.551661924918465, + "grad_norm": 1.9478322267532349, + "learning_rate": 0.0001400094328574712, + "loss": 0.8399, + "step": 22361 + }, + { + "epoch": 1.5517313163555617, + "grad_norm": 1.8939785957336426, + "learning_rate": 0.0001400042834795251, + "loss": 0.9082, + "step": 22362 + }, + { + "epoch": 1.5518007077926583, + "grad_norm": 2.444946050643921, + "learning_rate": 0.0001399991339752901, + "loss": 0.6816, + "step": 22363 + }, + { + "epoch": 1.5518700992297552, + "grad_norm": 1.904017448425293, + "learning_rate": 0.0001399939843447825, + "loss": 0.6653, + "step": 22364 + }, + { + "epoch": 1.5519394906668516, + "grad_norm": 1.295562982559204, + "learning_rate": 0.00013998883458801856, + "loss": 0.4138, + "step": 22365 + }, + { + "epoch": 1.5520088821039484, + "grad_norm": 1.4564441442489624, + "learning_rate": 0.00013998368470501444, + "loss": 0.6616, + "step": 22366 + }, + { + "epoch": 1.552078273541045, + "grad_norm": 1.5834436416625977, + "learning_rate": 0.0001399785346957865, + "loss": 0.4905, + "step": 22367 + }, + { + "epoch": 1.5521476649781416, + "grad_norm": 1.859691858291626, + "learning_rate": 0.00013997338456035098, + "loss": 0.7347, + "step": 22368 + }, + { + "epoch": 1.5522170564152384, + "grad_norm": 1.410359263420105, + "learning_rate": 0.0001399682342987241, + "loss": 0.707, + "step": 22369 + }, + { + "epoch": 1.552286447852335, + "grad_norm": 1.3056113719940186, + "learning_rate": 0.00013996308391092213, + "loss": 0.5191, + "step": 22370 + }, + { + "epoch": 1.5523558392894317, + "grad_norm": 1.4066026210784912, + "learning_rate": 0.00013995793339696138, + "loss": 0.52, + "step": 22371 + }, + { + "epoch": 1.5524252307265285, + "grad_norm": 1.707316279411316, + "learning_rate": 0.00013995278275685805, + "loss": 0.6822, + "step": 22372 + }, + { + "epoch": 1.5524946221636249, + "grad_norm": 1.6099144220352173, + "learning_rate": 0.00013994763199062842, + "loss": 0.6424, + "step": 22373 + }, + { + "epoch": 1.5525640136007217, + "grad_norm": 1.6840828657150269, + "learning_rate": 0.00013994248109828873, + "loss": 0.5212, + "step": 22374 + }, + { + "epoch": 1.5526334050378183, + "grad_norm": 1.796893835067749, + "learning_rate": 0.0001399373300798553, + "loss": 0.5842, + "step": 22375 + }, + { + "epoch": 1.552702796474915, + "grad_norm": 1.1866059303283691, + "learning_rate": 0.00013993217893534433, + "loss": 0.4779, + "step": 22376 + }, + { + "epoch": 1.5527721879120118, + "grad_norm": 1.448988676071167, + "learning_rate": 0.00013992702766477212, + "loss": 0.4999, + "step": 22377 + }, + { + "epoch": 1.5528415793491082, + "grad_norm": 1.599812626838684, + "learning_rate": 0.00013992187626815487, + "loss": 0.769, + "step": 22378 + }, + { + "epoch": 1.552910970786205, + "grad_norm": 1.3847976922988892, + "learning_rate": 0.00013991672474550893, + "loss": 0.5248, + "step": 22379 + }, + { + "epoch": 1.5529803622233016, + "grad_norm": 1.8972907066345215, + "learning_rate": 0.00013991157309685052, + "loss": 0.5879, + "step": 22380 + }, + { + "epoch": 1.5530497536603982, + "grad_norm": 1.9293103218078613, + "learning_rate": 0.0001399064213221959, + "loss": 0.8099, + "step": 22381 + }, + { + "epoch": 1.553119145097495, + "grad_norm": 1.208803415298462, + "learning_rate": 0.00013990126942156134, + "loss": 0.4924, + "step": 22382 + }, + { + "epoch": 1.5531885365345917, + "grad_norm": 1.5490645170211792, + "learning_rate": 0.0001398961173949631, + "loss": 0.5597, + "step": 22383 + }, + { + "epoch": 1.5532579279716883, + "grad_norm": 1.409266471862793, + "learning_rate": 0.00013989096524241745, + "loss": 0.5762, + "step": 22384 + }, + { + "epoch": 1.553327319408785, + "grad_norm": 1.8575170040130615, + "learning_rate": 0.00013988581296394063, + "loss": 0.5884, + "step": 22385 + }, + { + "epoch": 1.5533967108458815, + "grad_norm": 2.007908582687378, + "learning_rate": 0.00013988066055954895, + "loss": 0.6447, + "step": 22386 + }, + { + "epoch": 1.5534661022829783, + "grad_norm": 1.508756160736084, + "learning_rate": 0.00013987550802925863, + "loss": 0.6553, + "step": 22387 + }, + { + "epoch": 1.553535493720075, + "grad_norm": 1.5602173805236816, + "learning_rate": 0.000139870355373086, + "loss": 0.6751, + "step": 22388 + }, + { + "epoch": 1.5536048851571715, + "grad_norm": 1.6363364458084106, + "learning_rate": 0.00013986520259104724, + "loss": 0.5575, + "step": 22389 + }, + { + "epoch": 1.5536742765942684, + "grad_norm": 2.438798427581787, + "learning_rate": 0.0001398600496831587, + "loss": 0.6654, + "step": 22390 + }, + { + "epoch": 1.5537436680313648, + "grad_norm": 1.8723342418670654, + "learning_rate": 0.00013985489664943655, + "loss": 0.6318, + "step": 22391 + }, + { + "epoch": 1.5538130594684616, + "grad_norm": 1.8461381196975708, + "learning_rate": 0.00013984974348989717, + "loss": 0.6946, + "step": 22392 + }, + { + "epoch": 1.5538824509055582, + "grad_norm": 3.373342990875244, + "learning_rate": 0.00013984459020455675, + "loss": 0.7987, + "step": 22393 + }, + { + "epoch": 1.5539518423426548, + "grad_norm": 1.7421228885650635, + "learning_rate": 0.00013983943679343156, + "loss": 0.6806, + "step": 22394 + }, + { + "epoch": 1.5540212337797517, + "grad_norm": 1.4061259031295776, + "learning_rate": 0.0001398342832565379, + "loss": 0.5917, + "step": 22395 + }, + { + "epoch": 1.5540906252168483, + "grad_norm": 1.7690287828445435, + "learning_rate": 0.00013982912959389207, + "loss": 0.6223, + "step": 22396 + }, + { + "epoch": 1.5541600166539449, + "grad_norm": 2.194744825363159, + "learning_rate": 0.00013982397580551025, + "loss": 0.7816, + "step": 22397 + }, + { + "epoch": 1.5542294080910417, + "grad_norm": 1.2176663875579834, + "learning_rate": 0.0001398188218914088, + "loss": 0.4983, + "step": 22398 + }, + { + "epoch": 1.554298799528138, + "grad_norm": 1.4735299348831177, + "learning_rate": 0.00013981366785160396, + "loss": 0.5461, + "step": 22399 + }, + { + "epoch": 1.554368190965235, + "grad_norm": 1.6072266101837158, + "learning_rate": 0.00013980851368611193, + "loss": 0.4472, + "step": 22400 + }, + { + "epoch": 1.5544375824023315, + "grad_norm": 1.5446504354476929, + "learning_rate": 0.00013980335939494908, + "loss": 0.7821, + "step": 22401 + }, + { + "epoch": 1.5545069738394282, + "grad_norm": 1.765095829963684, + "learning_rate": 0.00013979820497813164, + "loss": 0.614, + "step": 22402 + }, + { + "epoch": 1.554576365276525, + "grad_norm": 1.814786672592163, + "learning_rate": 0.00013979305043567588, + "loss": 0.6856, + "step": 22403 + }, + { + "epoch": 1.5546457567136214, + "grad_norm": 1.54603910446167, + "learning_rate": 0.00013978789576759809, + "loss": 0.5778, + "step": 22404 + }, + { + "epoch": 1.5547151481507182, + "grad_norm": 1.1846535205841064, + "learning_rate": 0.0001397827409739145, + "loss": 0.4699, + "step": 22405 + }, + { + "epoch": 1.5547845395878148, + "grad_norm": 1.8928052186965942, + "learning_rate": 0.00013977758605464142, + "loss": 0.6707, + "step": 22406 + }, + { + "epoch": 1.5548539310249114, + "grad_norm": 1.6600160598754883, + "learning_rate": 0.0001397724310097951, + "loss": 0.6399, + "step": 22407 + }, + { + "epoch": 1.5549233224620083, + "grad_norm": 1.6367260217666626, + "learning_rate": 0.00013976727583939186, + "loss": 0.4681, + "step": 22408 + }, + { + "epoch": 1.5549927138991049, + "grad_norm": 1.474575161933899, + "learning_rate": 0.00013976212054344793, + "loss": 0.6831, + "step": 22409 + }, + { + "epoch": 1.5550621053362015, + "grad_norm": 1.4626646041870117, + "learning_rate": 0.00013975696512197962, + "loss": 0.6231, + "step": 22410 + }, + { + "epoch": 1.5551314967732983, + "grad_norm": 1.8697408437728882, + "learning_rate": 0.00013975180957500314, + "loss": 0.7008, + "step": 22411 + }, + { + "epoch": 1.5552008882103947, + "grad_norm": 1.6487398147583008, + "learning_rate": 0.00013974665390253485, + "loss": 0.5183, + "step": 22412 + }, + { + "epoch": 1.5552702796474915, + "grad_norm": 1.7742643356323242, + "learning_rate": 0.000139741498104591, + "loss": 0.468, + "step": 22413 + }, + { + "epoch": 1.5553396710845881, + "grad_norm": 1.4777462482452393, + "learning_rate": 0.0001397363421811878, + "loss": 0.6887, + "step": 22414 + }, + { + "epoch": 1.5554090625216848, + "grad_norm": 1.6819045543670654, + "learning_rate": 0.0001397311861323416, + "loss": 0.573, + "step": 22415 + }, + { + "epoch": 1.5554784539587816, + "grad_norm": 1.425065517425537, + "learning_rate": 0.00013972602995806867, + "loss": 0.565, + "step": 22416 + }, + { + "epoch": 1.5555478453958782, + "grad_norm": 1.378156065940857, + "learning_rate": 0.00013972087365838524, + "loss": 0.5945, + "step": 22417 + }, + { + "epoch": 1.5556172368329748, + "grad_norm": 1.4866191148757935, + "learning_rate": 0.00013971571723330766, + "loss": 0.6302, + "step": 22418 + }, + { + "epoch": 1.5556866282700714, + "grad_norm": 1.5118359327316284, + "learning_rate": 0.00013971056068285212, + "loss": 0.4487, + "step": 22419 + }, + { + "epoch": 1.555756019707168, + "grad_norm": 1.5285128355026245, + "learning_rate": 0.000139705404007035, + "loss": 0.6294, + "step": 22420 + }, + { + "epoch": 1.5558254111442649, + "grad_norm": 1.32316255569458, + "learning_rate": 0.00013970024720587248, + "loss": 0.5652, + "step": 22421 + }, + { + "epoch": 1.5558948025813615, + "grad_norm": 1.5760743618011475, + "learning_rate": 0.00013969509027938093, + "loss": 0.5932, + "step": 22422 + }, + { + "epoch": 1.555964194018458, + "grad_norm": 1.4699068069458008, + "learning_rate": 0.00013968993322757657, + "loss": 0.4561, + "step": 22423 + }, + { + "epoch": 1.556033585455555, + "grad_norm": 1.6656056642532349, + "learning_rate": 0.00013968477605047567, + "loss": 0.6071, + "step": 22424 + }, + { + "epoch": 1.5561029768926513, + "grad_norm": 1.4333264827728271, + "learning_rate": 0.00013967961874809457, + "loss": 0.5593, + "step": 22425 + }, + { + "epoch": 1.5561723683297481, + "grad_norm": 1.903123140335083, + "learning_rate": 0.00013967446132044948, + "loss": 0.6425, + "step": 22426 + }, + { + "epoch": 1.5562417597668448, + "grad_norm": 1.910994291305542, + "learning_rate": 0.00013966930376755677, + "loss": 0.5646, + "step": 22427 + }, + { + "epoch": 1.5563111512039414, + "grad_norm": 1.7435842752456665, + "learning_rate": 0.0001396641460894326, + "loss": 0.5045, + "step": 22428 + }, + { + "epoch": 1.5563805426410382, + "grad_norm": 1.6511458158493042, + "learning_rate": 0.0001396589882860934, + "loss": 0.5656, + "step": 22429 + }, + { + "epoch": 1.5564499340781348, + "grad_norm": 1.280936598777771, + "learning_rate": 0.00013965383035755534, + "loss": 0.4816, + "step": 22430 + }, + { + "epoch": 1.5565193255152314, + "grad_norm": 1.2795485258102417, + "learning_rate": 0.0001396486723038347, + "loss": 0.4336, + "step": 22431 + }, + { + "epoch": 1.556588716952328, + "grad_norm": 1.3689419031143188, + "learning_rate": 0.00013964351412494786, + "loss": 0.4576, + "step": 22432 + }, + { + "epoch": 1.5566581083894246, + "grad_norm": 2.1136577129364014, + "learning_rate": 0.00013963835582091102, + "loss": 0.9008, + "step": 22433 + }, + { + "epoch": 1.5567274998265215, + "grad_norm": 3.147510528564453, + "learning_rate": 0.0001396331973917405, + "loss": 0.5964, + "step": 22434 + }, + { + "epoch": 1.556796891263618, + "grad_norm": 2.0740790367126465, + "learning_rate": 0.00013962803883745258, + "loss": 0.671, + "step": 22435 + }, + { + "epoch": 1.5568662827007147, + "grad_norm": 1.7854046821594238, + "learning_rate": 0.0001396228801580635, + "loss": 0.5061, + "step": 22436 + }, + { + "epoch": 1.5569356741378115, + "grad_norm": 2.161555767059326, + "learning_rate": 0.00013961772135358963, + "loss": 0.4781, + "step": 22437 + }, + { + "epoch": 1.557005065574908, + "grad_norm": 2.927363395690918, + "learning_rate": 0.0001396125624240472, + "loss": 0.8823, + "step": 22438 + }, + { + "epoch": 1.5570744570120048, + "grad_norm": 1.563965082168579, + "learning_rate": 0.00013960740336945248, + "loss": 0.7057, + "step": 22439 + }, + { + "epoch": 1.5571438484491014, + "grad_norm": 1.6350016593933105, + "learning_rate": 0.0001396022441898218, + "loss": 0.5511, + "step": 22440 + }, + { + "epoch": 1.557213239886198, + "grad_norm": 1.8452320098876953, + "learning_rate": 0.00013959708488517144, + "loss": 0.6045, + "step": 22441 + }, + { + "epoch": 1.5572826313232948, + "grad_norm": 1.5830554962158203, + "learning_rate": 0.00013959192545551767, + "loss": 0.5419, + "step": 22442 + }, + { + "epoch": 1.5573520227603914, + "grad_norm": 1.860139012336731, + "learning_rate": 0.00013958676590087677, + "loss": 0.5009, + "step": 22443 + }, + { + "epoch": 1.557421414197488, + "grad_norm": 1.5285115242004395, + "learning_rate": 0.00013958160622126506, + "loss": 0.6233, + "step": 22444 + }, + { + "epoch": 1.5574908056345846, + "grad_norm": 2.7204513549804688, + "learning_rate": 0.0001395764464166988, + "loss": 0.791, + "step": 22445 + }, + { + "epoch": 1.5575601970716813, + "grad_norm": 1.3805986642837524, + "learning_rate": 0.0001395712864871943, + "loss": 0.6475, + "step": 22446 + }, + { + "epoch": 1.557629588508778, + "grad_norm": 1.5785460472106934, + "learning_rate": 0.00013956612643276783, + "loss": 0.6644, + "step": 22447 + }, + { + "epoch": 1.5576989799458747, + "grad_norm": 1.7551321983337402, + "learning_rate": 0.00013956096625343571, + "loss": 0.44, + "step": 22448 + }, + { + "epoch": 1.5577683713829713, + "grad_norm": 1.5132375955581665, + "learning_rate": 0.0001395558059492142, + "loss": 0.6357, + "step": 22449 + }, + { + "epoch": 1.5578377628200681, + "grad_norm": 1.4698896408081055, + "learning_rate": 0.00013955064552011957, + "loss": 0.4957, + "step": 22450 + }, + { + "epoch": 1.5579071542571645, + "grad_norm": 1.3671971559524536, + "learning_rate": 0.0001395454849661682, + "loss": 0.5713, + "step": 22451 + }, + { + "epoch": 1.5579765456942614, + "grad_norm": 1.5645594596862793, + "learning_rate": 0.00013954032428737626, + "loss": 0.5032, + "step": 22452 + }, + { + "epoch": 1.558045937131358, + "grad_norm": 2.3858206272125244, + "learning_rate": 0.00013953516348376015, + "loss": 0.8185, + "step": 22453 + }, + { + "epoch": 1.5581153285684546, + "grad_norm": 1.711440920829773, + "learning_rate": 0.0001395300025553361, + "loss": 0.4918, + "step": 22454 + }, + { + "epoch": 1.5581847200055514, + "grad_norm": 2.0397884845733643, + "learning_rate": 0.00013952484150212043, + "loss": 0.6386, + "step": 22455 + }, + { + "epoch": 1.558254111442648, + "grad_norm": 1.7586145401000977, + "learning_rate": 0.0001395196803241294, + "loss": 0.4825, + "step": 22456 + }, + { + "epoch": 1.5583235028797446, + "grad_norm": 1.6579062938690186, + "learning_rate": 0.00013951451902137932, + "loss": 0.5348, + "step": 22457 + }, + { + "epoch": 1.5583928943168412, + "grad_norm": 1.5468509197235107, + "learning_rate": 0.0001395093575938865, + "loss": 0.6849, + "step": 22458 + }, + { + "epoch": 1.5584622857539379, + "grad_norm": 1.6082181930541992, + "learning_rate": 0.00013950419604166723, + "loss": 0.6403, + "step": 22459 + }, + { + "epoch": 1.5585316771910347, + "grad_norm": 3.021516799926758, + "learning_rate": 0.00013949903436473777, + "loss": 0.7389, + "step": 22460 + }, + { + "epoch": 1.5586010686281313, + "grad_norm": 1.4452543258666992, + "learning_rate": 0.00013949387256311446, + "loss": 0.4975, + "step": 22461 + }, + { + "epoch": 1.558670460065228, + "grad_norm": 3.2809157371520996, + "learning_rate": 0.00013948871063681359, + "loss": 0.7278, + "step": 22462 + }, + { + "epoch": 1.5587398515023247, + "grad_norm": 1.9229934215545654, + "learning_rate": 0.00013948354858585142, + "loss": 0.4312, + "step": 22463 + }, + { + "epoch": 1.5588092429394211, + "grad_norm": 1.7055623531341553, + "learning_rate": 0.00013947838641024428, + "loss": 0.6377, + "step": 22464 + }, + { + "epoch": 1.558878634376518, + "grad_norm": 2.2039453983306885, + "learning_rate": 0.00013947322411000845, + "loss": 0.8049, + "step": 22465 + }, + { + "epoch": 1.5589480258136146, + "grad_norm": 1.6493960618972778, + "learning_rate": 0.00013946806168516022, + "loss": 0.6577, + "step": 22466 + }, + { + "epoch": 1.5590174172507112, + "grad_norm": 2.098193407058716, + "learning_rate": 0.00013946289913571591, + "loss": 0.5686, + "step": 22467 + }, + { + "epoch": 1.559086808687808, + "grad_norm": 1.4927723407745361, + "learning_rate": 0.00013945773646169182, + "loss": 0.6224, + "step": 22468 + }, + { + "epoch": 1.5591562001249046, + "grad_norm": 1.4653159379959106, + "learning_rate": 0.00013945257366310422, + "loss": 0.5759, + "step": 22469 + }, + { + "epoch": 1.5592255915620012, + "grad_norm": 1.5418682098388672, + "learning_rate": 0.00013944741073996944, + "loss": 0.7032, + "step": 22470 + }, + { + "epoch": 1.5592949829990979, + "grad_norm": 1.3144117593765259, + "learning_rate": 0.00013944224769230372, + "loss": 0.5157, + "step": 22471 + }, + { + "epoch": 1.5593643744361945, + "grad_norm": 1.832535982131958, + "learning_rate": 0.00013943708452012344, + "loss": 0.5418, + "step": 22472 + }, + { + "epoch": 1.5594337658732913, + "grad_norm": 1.7994505167007446, + "learning_rate": 0.00013943192122344483, + "loss": 0.4581, + "step": 22473 + }, + { + "epoch": 1.559503157310388, + "grad_norm": 1.9340626001358032, + "learning_rate": 0.00013942675780228422, + "loss": 0.6729, + "step": 22474 + }, + { + "epoch": 1.5595725487474845, + "grad_norm": 1.794053077697754, + "learning_rate": 0.00013942159425665794, + "loss": 0.7355, + "step": 22475 + }, + { + "epoch": 1.5596419401845814, + "grad_norm": 2.1035728454589844, + "learning_rate": 0.00013941643058658224, + "loss": 0.7034, + "step": 22476 + }, + { + "epoch": 1.5597113316216777, + "grad_norm": 1.504089593887329, + "learning_rate": 0.00013941126679207347, + "loss": 0.4969, + "step": 22477 + }, + { + "epoch": 1.5597807230587746, + "grad_norm": 1.6568080186843872, + "learning_rate": 0.00013940610287314784, + "loss": 0.5467, + "step": 22478 + }, + { + "epoch": 1.5598501144958712, + "grad_norm": 1.864745020866394, + "learning_rate": 0.00013940093882982177, + "loss": 0.6191, + "step": 22479 + }, + { + "epoch": 1.5599195059329678, + "grad_norm": 1.9753549098968506, + "learning_rate": 0.00013939577466211148, + "loss": 0.4883, + "step": 22480 + }, + { + "epoch": 1.5599888973700646, + "grad_norm": 3.820375442504883, + "learning_rate": 0.0001393906103700333, + "loss": 0.7528, + "step": 22481 + }, + { + "epoch": 1.5600582888071612, + "grad_norm": 2.0303144454956055, + "learning_rate": 0.00013938544595360356, + "loss": 0.6992, + "step": 22482 + }, + { + "epoch": 1.5601276802442579, + "grad_norm": 1.6395808458328247, + "learning_rate": 0.00013938028141283852, + "loss": 0.6882, + "step": 22483 + }, + { + "epoch": 1.5601970716813547, + "grad_norm": 1.3928382396697998, + "learning_rate": 0.0001393751167477545, + "loss": 0.6092, + "step": 22484 + }, + { + "epoch": 1.560266463118451, + "grad_norm": 1.2463557720184326, + "learning_rate": 0.00013936995195836782, + "loss": 0.5504, + "step": 22485 + }, + { + "epoch": 1.560335854555548, + "grad_norm": 1.4539427757263184, + "learning_rate": 0.00013936478704469475, + "loss": 0.7278, + "step": 22486 + }, + { + "epoch": 1.5604052459926445, + "grad_norm": 1.6975171566009521, + "learning_rate": 0.00013935962200675164, + "loss": 0.6577, + "step": 22487 + }, + { + "epoch": 1.5604746374297411, + "grad_norm": 1.079408884048462, + "learning_rate": 0.00013935445684455476, + "loss": 0.4543, + "step": 22488 + }, + { + "epoch": 1.560544028866838, + "grad_norm": 2.359447956085205, + "learning_rate": 0.00013934929155812043, + "loss": 0.8644, + "step": 22489 + }, + { + "epoch": 1.5606134203039344, + "grad_norm": 1.487871527671814, + "learning_rate": 0.00013934412614746494, + "loss": 0.6019, + "step": 22490 + }, + { + "epoch": 1.5606828117410312, + "grad_norm": 1.9146283864974976, + "learning_rate": 0.0001393389606126046, + "loss": 0.556, + "step": 22491 + }, + { + "epoch": 1.5607522031781278, + "grad_norm": 1.4101085662841797, + "learning_rate": 0.0001393337949535557, + "loss": 0.7215, + "step": 22492 + }, + { + "epoch": 1.5608215946152244, + "grad_norm": 1.5796728134155273, + "learning_rate": 0.00013932862917033462, + "loss": 0.6422, + "step": 22493 + }, + { + "epoch": 1.5608909860523212, + "grad_norm": 2.229461669921875, + "learning_rate": 0.0001393234632629576, + "loss": 0.6672, + "step": 22494 + }, + { + "epoch": 1.5609603774894178, + "grad_norm": 2.0210955142974854, + "learning_rate": 0.00013931829723144097, + "loss": 0.6896, + "step": 22495 + }, + { + "epoch": 1.5610297689265145, + "grad_norm": 1.4546746015548706, + "learning_rate": 0.00013931313107580105, + "loss": 0.588, + "step": 22496 + }, + { + "epoch": 1.5610991603636113, + "grad_norm": 1.9571607112884521, + "learning_rate": 0.00013930796479605414, + "loss": 0.7165, + "step": 22497 + }, + { + "epoch": 1.5611685518007077, + "grad_norm": 2.3657925128936768, + "learning_rate": 0.00013930279839221652, + "loss": 0.8801, + "step": 22498 + }, + { + "epoch": 1.5612379432378045, + "grad_norm": 2.5848968029022217, + "learning_rate": 0.00013929763186430454, + "loss": 0.6669, + "step": 22499 + }, + { + "epoch": 1.5613073346749011, + "grad_norm": 2.4383034706115723, + "learning_rate": 0.0001392924652123345, + "loss": 0.6223, + "step": 22500 + }, + { + "epoch": 1.5613767261119977, + "grad_norm": 1.6412948369979858, + "learning_rate": 0.00013928729843632267, + "loss": 0.5656, + "step": 22501 + }, + { + "epoch": 1.5614461175490946, + "grad_norm": 1.3985685110092163, + "learning_rate": 0.00013928213153628543, + "loss": 0.5621, + "step": 22502 + }, + { + "epoch": 1.561515508986191, + "grad_norm": 1.5881601572036743, + "learning_rate": 0.0001392769645122391, + "loss": 0.6615, + "step": 22503 + }, + { + "epoch": 1.5615849004232878, + "grad_norm": 1.2527086734771729, + "learning_rate": 0.0001392717973641999, + "loss": 0.5755, + "step": 22504 + }, + { + "epoch": 1.5616542918603844, + "grad_norm": 1.586552381515503, + "learning_rate": 0.0001392666300921842, + "loss": 0.4637, + "step": 22505 + }, + { + "epoch": 1.561723683297481, + "grad_norm": 1.5397547483444214, + "learning_rate": 0.00013926146269620829, + "loss": 0.4543, + "step": 22506 + }, + { + "epoch": 1.5617930747345778, + "grad_norm": 2.35431170463562, + "learning_rate": 0.0001392562951762885, + "loss": 0.746, + "step": 22507 + }, + { + "epoch": 1.5618624661716745, + "grad_norm": 1.6414018869400024, + "learning_rate": 0.00013925112753244116, + "loss": 0.8546, + "step": 22508 + }, + { + "epoch": 1.561931857608771, + "grad_norm": 1.7297618389129639, + "learning_rate": 0.00013924595976468255, + "loss": 0.6567, + "step": 22509 + }, + { + "epoch": 1.562001249045868, + "grad_norm": 1.8139832019805908, + "learning_rate": 0.000139240791873029, + "loss": 0.7589, + "step": 22510 + }, + { + "epoch": 1.5620706404829643, + "grad_norm": 2.5234649181365967, + "learning_rate": 0.00013923562385749684, + "loss": 0.6903, + "step": 22511 + }, + { + "epoch": 1.5621400319200611, + "grad_norm": 1.9043282270431519, + "learning_rate": 0.00013923045571810234, + "loss": 0.7785, + "step": 22512 + }, + { + "epoch": 1.5622094233571577, + "grad_norm": 1.4699469804763794, + "learning_rate": 0.00013922528745486187, + "loss": 0.671, + "step": 22513 + }, + { + "epoch": 1.5622788147942543, + "grad_norm": 1.5765385627746582, + "learning_rate": 0.00013922011906779172, + "loss": 0.7269, + "step": 22514 + }, + { + "epoch": 1.5623482062313512, + "grad_norm": 2.355219841003418, + "learning_rate": 0.00013921495055690817, + "loss": 0.4477, + "step": 22515 + }, + { + "epoch": 1.5624175976684476, + "grad_norm": 1.8889678716659546, + "learning_rate": 0.00013920978192222757, + "loss": 0.6364, + "step": 22516 + }, + { + "epoch": 1.5624869891055444, + "grad_norm": 1.4279190301895142, + "learning_rate": 0.00013920461316376627, + "loss": 0.6374, + "step": 22517 + }, + { + "epoch": 1.562556380542641, + "grad_norm": 1.3086037635803223, + "learning_rate": 0.00013919944428154055, + "loss": 0.4129, + "step": 22518 + }, + { + "epoch": 1.5626257719797376, + "grad_norm": 1.7258672714233398, + "learning_rate": 0.00013919427527556673, + "loss": 0.694, + "step": 22519 + }, + { + "epoch": 1.5626951634168345, + "grad_norm": 1.7306803464889526, + "learning_rate": 0.00013918910614586113, + "loss": 0.5065, + "step": 22520 + }, + { + "epoch": 1.562764554853931, + "grad_norm": 2.12272572517395, + "learning_rate": 0.00013918393689244004, + "loss": 0.7048, + "step": 22521 + }, + { + "epoch": 1.5628339462910277, + "grad_norm": 1.7807676792144775, + "learning_rate": 0.00013917876751531986, + "loss": 0.5267, + "step": 22522 + }, + { + "epoch": 1.5629033377281245, + "grad_norm": 1.8851568698883057, + "learning_rate": 0.00013917359801451684, + "loss": 0.5988, + "step": 22523 + }, + { + "epoch": 1.562972729165221, + "grad_norm": 1.1722509860992432, + "learning_rate": 0.00013916842839004732, + "loss": 0.4373, + "step": 22524 + }, + { + "epoch": 1.5630421206023177, + "grad_norm": 2.067361354827881, + "learning_rate": 0.00013916325864192757, + "loss": 0.7462, + "step": 22525 + }, + { + "epoch": 1.5631115120394143, + "grad_norm": 1.8662185668945312, + "learning_rate": 0.000139158088770174, + "loss": 0.6552, + "step": 22526 + }, + { + "epoch": 1.563180903476511, + "grad_norm": 1.5688058137893677, + "learning_rate": 0.00013915291877480287, + "loss": 0.4816, + "step": 22527 + }, + { + "epoch": 1.5632502949136078, + "grad_norm": 1.8331282138824463, + "learning_rate": 0.00013914774865583053, + "loss": 0.6619, + "step": 22528 + }, + { + "epoch": 1.5633196863507042, + "grad_norm": 1.8726508617401123, + "learning_rate": 0.00013914257841327328, + "loss": 0.3865, + "step": 22529 + }, + { + "epoch": 1.563389077787801, + "grad_norm": 1.9516292810440063, + "learning_rate": 0.00013913740804714743, + "loss": 0.6359, + "step": 22530 + }, + { + "epoch": 1.5634584692248976, + "grad_norm": 1.5544887781143188, + "learning_rate": 0.00013913223755746936, + "loss": 0.5055, + "step": 22531 + }, + { + "epoch": 1.5635278606619942, + "grad_norm": 1.6586668491363525, + "learning_rate": 0.00013912706694425535, + "loss": 0.382, + "step": 22532 + }, + { + "epoch": 1.563597252099091, + "grad_norm": 2.0305707454681396, + "learning_rate": 0.0001391218962075217, + "loss": 0.659, + "step": 22533 + }, + { + "epoch": 1.5636666435361877, + "grad_norm": 1.4538836479187012, + "learning_rate": 0.0001391167253472848, + "loss": 0.5316, + "step": 22534 + }, + { + "epoch": 1.5637360349732843, + "grad_norm": 1.841181755065918, + "learning_rate": 0.00013911155436356092, + "loss": 0.7345, + "step": 22535 + }, + { + "epoch": 1.5638054264103811, + "grad_norm": 1.3302661180496216, + "learning_rate": 0.0001391063832563664, + "loss": 0.4567, + "step": 22536 + }, + { + "epoch": 1.5638748178474775, + "grad_norm": 1.725564956665039, + "learning_rate": 0.00013910121202571755, + "loss": 0.6579, + "step": 22537 + }, + { + "epoch": 1.5639442092845743, + "grad_norm": 2.0980982780456543, + "learning_rate": 0.00013909604067163076, + "loss": 0.9519, + "step": 22538 + }, + { + "epoch": 1.564013600721671, + "grad_norm": 2.143106698989868, + "learning_rate": 0.00013909086919412227, + "loss": 0.7, + "step": 22539 + }, + { + "epoch": 1.5640829921587676, + "grad_norm": 2.375981092453003, + "learning_rate": 0.00013908569759320843, + "loss": 0.8492, + "step": 22540 + }, + { + "epoch": 1.5641523835958644, + "grad_norm": 1.294427752494812, + "learning_rate": 0.0001390805258689056, + "loss": 0.5299, + "step": 22541 + }, + { + "epoch": 1.5642217750329608, + "grad_norm": 2.023543119430542, + "learning_rate": 0.0001390753540212301, + "loss": 0.564, + "step": 22542 + }, + { + "epoch": 1.5642911664700576, + "grad_norm": 2.5266990661621094, + "learning_rate": 0.0001390701820501982, + "loss": 0.6259, + "step": 22543 + }, + { + "epoch": 1.5643605579071542, + "grad_norm": 1.5606540441513062, + "learning_rate": 0.00013906500995582627, + "loss": 0.5227, + "step": 22544 + }, + { + "epoch": 1.5644299493442508, + "grad_norm": 1.3053741455078125, + "learning_rate": 0.00013905983773813068, + "loss": 0.5011, + "step": 22545 + }, + { + "epoch": 1.5644993407813477, + "grad_norm": 1.2947498559951782, + "learning_rate": 0.00013905466539712767, + "loss": 0.6866, + "step": 22546 + }, + { + "epoch": 1.5645687322184443, + "grad_norm": 1.8968024253845215, + "learning_rate": 0.0001390494929328336, + "loss": 0.8208, + "step": 22547 + }, + { + "epoch": 1.564638123655541, + "grad_norm": 1.1751818656921387, + "learning_rate": 0.00013904432034526486, + "loss": 0.5058, + "step": 22548 + }, + { + "epoch": 1.5647075150926377, + "grad_norm": 1.7578418254852295, + "learning_rate": 0.0001390391476344377, + "loss": 0.5891, + "step": 22549 + }, + { + "epoch": 1.5647769065297341, + "grad_norm": 1.914774775505066, + "learning_rate": 0.00013903397480036849, + "loss": 0.6628, + "step": 22550 + }, + { + "epoch": 1.564846297966831, + "grad_norm": 1.5525058507919312, + "learning_rate": 0.00013902880184307357, + "loss": 0.6443, + "step": 22551 + }, + { + "epoch": 1.5649156894039276, + "grad_norm": 1.531387448310852, + "learning_rate": 0.00013902362876256922, + "loss": 0.6285, + "step": 22552 + }, + { + "epoch": 1.5649850808410242, + "grad_norm": 1.4078289270401, + "learning_rate": 0.0001390184555588718, + "loss": 0.4923, + "step": 22553 + }, + { + "epoch": 1.565054472278121, + "grad_norm": 1.430504322052002, + "learning_rate": 0.00013901328223199765, + "loss": 0.5907, + "step": 22554 + }, + { + "epoch": 1.5651238637152174, + "grad_norm": 1.419756531715393, + "learning_rate": 0.0001390081087819631, + "loss": 0.6439, + "step": 22555 + }, + { + "epoch": 1.5651932551523142, + "grad_norm": 2.6341371536254883, + "learning_rate": 0.00013900293520878449, + "loss": 0.6102, + "step": 22556 + }, + { + "epoch": 1.5652626465894108, + "grad_norm": 2.0108413696289062, + "learning_rate": 0.00013899776151247815, + "loss": 0.7874, + "step": 22557 + }, + { + "epoch": 1.5653320380265074, + "grad_norm": 1.4223695993423462, + "learning_rate": 0.00013899258769306037, + "loss": 0.681, + "step": 22558 + }, + { + "epoch": 1.5654014294636043, + "grad_norm": 1.2572710514068604, + "learning_rate": 0.00013898741375054752, + "loss": 0.4087, + "step": 22559 + }, + { + "epoch": 1.5654708209007009, + "grad_norm": 2.2856028079986572, + "learning_rate": 0.00013898223968495592, + "loss": 0.9575, + "step": 22560 + }, + { + "epoch": 1.5655402123377975, + "grad_norm": 2.2109737396240234, + "learning_rate": 0.0001389770654963019, + "loss": 0.5333, + "step": 22561 + }, + { + "epoch": 1.5656096037748943, + "grad_norm": 1.5864344835281372, + "learning_rate": 0.00013897189118460186, + "loss": 0.4939, + "step": 22562 + }, + { + "epoch": 1.5656789952119907, + "grad_norm": 1.7853981256484985, + "learning_rate": 0.00013896671674987202, + "loss": 0.6834, + "step": 22563 + }, + { + "epoch": 1.5657483866490876, + "grad_norm": 2.736762046813965, + "learning_rate": 0.0001389615421921288, + "loss": 0.6542, + "step": 22564 + }, + { + "epoch": 1.5658177780861842, + "grad_norm": 2.027345657348633, + "learning_rate": 0.00013895636751138854, + "loss": 0.4922, + "step": 22565 + }, + { + "epoch": 1.5658871695232808, + "grad_norm": 1.3051437139511108, + "learning_rate": 0.0001389511927076675, + "loss": 0.5576, + "step": 22566 + }, + { + "epoch": 1.5659565609603776, + "grad_norm": 2.234358787536621, + "learning_rate": 0.0001389460177809821, + "loss": 0.8735, + "step": 22567 + }, + { + "epoch": 1.566025952397474, + "grad_norm": 1.2736531496047974, + "learning_rate": 0.00013894084273134862, + "loss": 0.4641, + "step": 22568 + }, + { + "epoch": 1.5660953438345708, + "grad_norm": 2.5435914993286133, + "learning_rate": 0.00013893566755878345, + "loss": 0.5444, + "step": 22569 + }, + { + "epoch": 1.5661647352716674, + "grad_norm": 2.450685977935791, + "learning_rate": 0.00013893049226330285, + "loss": 0.7487, + "step": 22570 + }, + { + "epoch": 1.566234126708764, + "grad_norm": 1.98538339138031, + "learning_rate": 0.00013892531684492324, + "loss": 0.486, + "step": 22571 + }, + { + "epoch": 1.5663035181458609, + "grad_norm": 2.6720495223999023, + "learning_rate": 0.0001389201413036609, + "loss": 0.755, + "step": 22572 + }, + { + "epoch": 1.5663729095829575, + "grad_norm": 2.36118483543396, + "learning_rate": 0.0001389149656395322, + "loss": 0.7758, + "step": 22573 + }, + { + "epoch": 1.566442301020054, + "grad_norm": 1.687578558921814, + "learning_rate": 0.00013890978985255345, + "loss": 0.7004, + "step": 22574 + }, + { + "epoch": 1.566511692457151, + "grad_norm": 1.945406198501587, + "learning_rate": 0.00013890461394274102, + "loss": 0.6675, + "step": 22575 + }, + { + "epoch": 1.5665810838942473, + "grad_norm": 1.4881162643432617, + "learning_rate": 0.00013889943791011128, + "loss": 0.5312, + "step": 22576 + }, + { + "epoch": 1.5666504753313442, + "grad_norm": 3.2286291122436523, + "learning_rate": 0.00013889426175468046, + "loss": 0.5547, + "step": 22577 + }, + { + "epoch": 1.5667198667684408, + "grad_norm": 1.7718082666397095, + "learning_rate": 0.00013888908547646497, + "loss": 0.5622, + "step": 22578 + }, + { + "epoch": 1.5667892582055374, + "grad_norm": 1.8260103464126587, + "learning_rate": 0.0001388839090754812, + "loss": 0.7026, + "step": 22579 + }, + { + "epoch": 1.5668586496426342, + "grad_norm": 1.260262131690979, + "learning_rate": 0.00013887873255174539, + "loss": 0.5876, + "step": 22580 + }, + { + "epoch": 1.5669280410797306, + "grad_norm": 1.699733018875122, + "learning_rate": 0.00013887355590527396, + "loss": 0.517, + "step": 22581 + }, + { + "epoch": 1.5669974325168274, + "grad_norm": 2.2961158752441406, + "learning_rate": 0.0001388683791360832, + "loss": 0.8892, + "step": 22582 + }, + { + "epoch": 1.567066823953924, + "grad_norm": 2.09692120552063, + "learning_rate": 0.00013886320224418952, + "loss": 0.5027, + "step": 22583 + }, + { + "epoch": 1.5671362153910207, + "grad_norm": 1.7829303741455078, + "learning_rate": 0.00013885802522960918, + "loss": 0.5503, + "step": 22584 + }, + { + "epoch": 1.5672056068281175, + "grad_norm": 1.7921311855316162, + "learning_rate": 0.0001388528480923586, + "loss": 0.764, + "step": 22585 + }, + { + "epoch": 1.567274998265214, + "grad_norm": 2.2407572269439697, + "learning_rate": 0.00013884767083245406, + "loss": 0.6413, + "step": 22586 + }, + { + "epoch": 1.5673443897023107, + "grad_norm": 1.7153394222259521, + "learning_rate": 0.00013884249344991192, + "loss": 0.7004, + "step": 22587 + }, + { + "epoch": 1.5674137811394075, + "grad_norm": 1.7725845575332642, + "learning_rate": 0.00013883731594474855, + "loss": 0.7571, + "step": 22588 + }, + { + "epoch": 1.567483172576504, + "grad_norm": 2.287907600402832, + "learning_rate": 0.0001388321383169803, + "loss": 0.5827, + "step": 22589 + }, + { + "epoch": 1.5675525640136008, + "grad_norm": 2.157857894897461, + "learning_rate": 0.00013882696056662346, + "loss": 0.8309, + "step": 22590 + }, + { + "epoch": 1.5676219554506974, + "grad_norm": 1.7364495992660522, + "learning_rate": 0.0001388217826936944, + "loss": 0.6071, + "step": 22591 + }, + { + "epoch": 1.567691346887794, + "grad_norm": 1.7151978015899658, + "learning_rate": 0.00013881660469820952, + "loss": 0.6143, + "step": 22592 + }, + { + "epoch": 1.5677607383248908, + "grad_norm": 2.1623642444610596, + "learning_rate": 0.0001388114265801851, + "loss": 0.5911, + "step": 22593 + }, + { + "epoch": 1.5678301297619872, + "grad_norm": 1.5609666109085083, + "learning_rate": 0.00013880624833963754, + "loss": 0.4614, + "step": 22594 + }, + { + "epoch": 1.567899521199084, + "grad_norm": 1.6479965448379517, + "learning_rate": 0.0001388010699765831, + "loss": 0.6392, + "step": 22595 + }, + { + "epoch": 1.5679689126361807, + "grad_norm": 1.3975543975830078, + "learning_rate": 0.00013879589149103823, + "loss": 0.5416, + "step": 22596 + }, + { + "epoch": 1.5680383040732773, + "grad_norm": 1.342066764831543, + "learning_rate": 0.0001387907128830192, + "loss": 0.4936, + "step": 22597 + }, + { + "epoch": 1.568107695510374, + "grad_norm": 1.6666604280471802, + "learning_rate": 0.0001387855341525424, + "loss": 0.6284, + "step": 22598 + }, + { + "epoch": 1.5681770869474707, + "grad_norm": 1.5682629346847534, + "learning_rate": 0.00013878035529962416, + "loss": 0.6451, + "step": 22599 + }, + { + "epoch": 1.5682464783845673, + "grad_norm": 2.4030721187591553, + "learning_rate": 0.00013877517632428089, + "loss": 0.9724, + "step": 22600 + }, + { + "epoch": 1.5683158698216642, + "grad_norm": 1.6922341585159302, + "learning_rate": 0.00013876999722652882, + "loss": 0.5394, + "step": 22601 + }, + { + "epoch": 1.5683852612587605, + "grad_norm": 3.203172206878662, + "learning_rate": 0.00013876481800638438, + "loss": 0.583, + "step": 22602 + }, + { + "epoch": 1.5684546526958574, + "grad_norm": 2.004991292953491, + "learning_rate": 0.00013875963866386395, + "loss": 0.7087, + "step": 22603 + }, + { + "epoch": 1.568524044132954, + "grad_norm": 1.401818037033081, + "learning_rate": 0.0001387544591989838, + "loss": 0.5277, + "step": 22604 + }, + { + "epoch": 1.5685934355700506, + "grad_norm": 2.5292000770568848, + "learning_rate": 0.0001387492796117603, + "loss": 0.5716, + "step": 22605 + }, + { + "epoch": 1.5686628270071474, + "grad_norm": 1.4249016046524048, + "learning_rate": 0.00013874409990220988, + "loss": 0.4862, + "step": 22606 + }, + { + "epoch": 1.5687322184442438, + "grad_norm": 1.570291519165039, + "learning_rate": 0.00013873892007034878, + "loss": 0.5321, + "step": 22607 + }, + { + "epoch": 1.5688016098813407, + "grad_norm": 1.354781150817871, + "learning_rate": 0.00013873374011619344, + "loss": 0.5561, + "step": 22608 + }, + { + "epoch": 1.5688710013184373, + "grad_norm": 3.2584524154663086, + "learning_rate": 0.00013872856003976016, + "loss": 0.6253, + "step": 22609 + }, + { + "epoch": 1.5689403927555339, + "grad_norm": 1.8823671340942383, + "learning_rate": 0.0001387233798410653, + "loss": 0.8958, + "step": 22610 + }, + { + "epoch": 1.5690097841926307, + "grad_norm": 2.5147039890289307, + "learning_rate": 0.00013871819952012527, + "loss": 0.6665, + "step": 22611 + }, + { + "epoch": 1.5690791756297273, + "grad_norm": 1.7896353006362915, + "learning_rate": 0.0001387130190769563, + "loss": 0.6893, + "step": 22612 + }, + { + "epoch": 1.569148567066824, + "grad_norm": 1.6121121644973755, + "learning_rate": 0.00013870783851157486, + "loss": 0.5735, + "step": 22613 + }, + { + "epoch": 1.5692179585039208, + "grad_norm": 2.0750279426574707, + "learning_rate": 0.00013870265782399727, + "loss": 0.6036, + "step": 22614 + }, + { + "epoch": 1.5692873499410172, + "grad_norm": 1.2827427387237549, + "learning_rate": 0.00013869747701423987, + "loss": 0.6078, + "step": 22615 + }, + { + "epoch": 1.569356741378114, + "grad_norm": 2.291982412338257, + "learning_rate": 0.00013869229608231903, + "loss": 0.9276, + "step": 22616 + }, + { + "epoch": 1.5694261328152106, + "grad_norm": 1.3860948085784912, + "learning_rate": 0.0001386871150282511, + "loss": 0.5391, + "step": 22617 + }, + { + "epoch": 1.5694955242523072, + "grad_norm": 1.7665550708770752, + "learning_rate": 0.00013868193385205245, + "loss": 0.9573, + "step": 22618 + }, + { + "epoch": 1.569564915689404, + "grad_norm": 1.9714441299438477, + "learning_rate": 0.00013867675255373936, + "loss": 0.6547, + "step": 22619 + }, + { + "epoch": 1.5696343071265004, + "grad_norm": 2.1135621070861816, + "learning_rate": 0.0001386715711333283, + "loss": 0.7072, + "step": 22620 + }, + { + "epoch": 1.5697036985635973, + "grad_norm": 1.6184444427490234, + "learning_rate": 0.0001386663895908356, + "loss": 0.5021, + "step": 22621 + }, + { + "epoch": 1.5697730900006939, + "grad_norm": 1.4072331190109253, + "learning_rate": 0.00013866120792627756, + "loss": 0.5543, + "step": 22622 + }, + { + "epoch": 1.5698424814377905, + "grad_norm": 1.525979995727539, + "learning_rate": 0.00013865602613967058, + "loss": 0.5905, + "step": 22623 + }, + { + "epoch": 1.5699118728748873, + "grad_norm": 1.558453917503357, + "learning_rate": 0.000138650844231031, + "loss": 0.8357, + "step": 22624 + }, + { + "epoch": 1.569981264311984, + "grad_norm": 2.0295042991638184, + "learning_rate": 0.0001386456622003752, + "loss": 0.8267, + "step": 22625 + }, + { + "epoch": 1.5700506557490805, + "grad_norm": 1.6110119819641113, + "learning_rate": 0.00013864048004771953, + "loss": 0.5396, + "step": 22626 + }, + { + "epoch": 1.5701200471861774, + "grad_norm": 1.7236968278884888, + "learning_rate": 0.00013863529777308035, + "loss": 0.6821, + "step": 22627 + }, + { + "epoch": 1.5701894386232738, + "grad_norm": 1.715753436088562, + "learning_rate": 0.00013863011537647402, + "loss": 0.5597, + "step": 22628 + }, + { + "epoch": 1.5702588300603706, + "grad_norm": 1.676435112953186, + "learning_rate": 0.0001386249328579169, + "loss": 0.6198, + "step": 22629 + }, + { + "epoch": 1.5703282214974672, + "grad_norm": 1.5870987176895142, + "learning_rate": 0.00013861975021742533, + "loss": 0.6325, + "step": 22630 + }, + { + "epoch": 1.5703976129345638, + "grad_norm": 2.219729423522949, + "learning_rate": 0.00013861456745501571, + "loss": 0.7388, + "step": 22631 + }, + { + "epoch": 1.5704670043716606, + "grad_norm": 1.3110517263412476, + "learning_rate": 0.00013860938457070437, + "loss": 0.5747, + "step": 22632 + }, + { + "epoch": 1.570536395808757, + "grad_norm": 1.5805249214172363, + "learning_rate": 0.00013860420156450767, + "loss": 0.7337, + "step": 22633 + }, + { + "epoch": 1.5706057872458539, + "grad_norm": 1.6460298299789429, + "learning_rate": 0.00013859901843644198, + "loss": 0.4059, + "step": 22634 + }, + { + "epoch": 1.5706751786829505, + "grad_norm": 1.5788456201553345, + "learning_rate": 0.00013859383518652368, + "loss": 0.3583, + "step": 22635 + }, + { + "epoch": 1.570744570120047, + "grad_norm": 2.0849316120147705, + "learning_rate": 0.00013858865181476913, + "loss": 0.8286, + "step": 22636 + }, + { + "epoch": 1.570813961557144, + "grad_norm": 2.0431413650512695, + "learning_rate": 0.00013858346832119466, + "loss": 0.7594, + "step": 22637 + }, + { + "epoch": 1.5708833529942405, + "grad_norm": 1.308890461921692, + "learning_rate": 0.0001385782847058167, + "loss": 0.5193, + "step": 22638 + }, + { + "epoch": 1.5709527444313371, + "grad_norm": 1.3815711736679077, + "learning_rate": 0.00013857310096865153, + "loss": 0.5145, + "step": 22639 + }, + { + "epoch": 1.571022135868434, + "grad_norm": 1.6201647520065308, + "learning_rate": 0.00013856791710971557, + "loss": 0.8852, + "step": 22640 + }, + { + "epoch": 1.5710915273055304, + "grad_norm": 2.0976133346557617, + "learning_rate": 0.00013856273312902516, + "loss": 0.5599, + "step": 22641 + }, + { + "epoch": 1.5711609187426272, + "grad_norm": 1.4318832159042358, + "learning_rate": 0.00013855754902659668, + "loss": 0.6933, + "step": 22642 + }, + { + "epoch": 1.5712303101797238, + "grad_norm": 1.4394861459732056, + "learning_rate": 0.00013855236480244648, + "loss": 0.6205, + "step": 22643 + }, + { + "epoch": 1.5712997016168204, + "grad_norm": 1.4347374439239502, + "learning_rate": 0.00013854718045659098, + "loss": 0.5156, + "step": 22644 + }, + { + "epoch": 1.5713690930539173, + "grad_norm": 1.6443568468093872, + "learning_rate": 0.00013854199598904647, + "loss": 0.5572, + "step": 22645 + }, + { + "epoch": 1.5714384844910136, + "grad_norm": 1.7713027000427246, + "learning_rate": 0.00013853681139982938, + "loss": 0.553, + "step": 22646 + }, + { + "epoch": 1.5715078759281105, + "grad_norm": 2.4151854515075684, + "learning_rate": 0.00013853162668895602, + "loss": 0.5416, + "step": 22647 + }, + { + "epoch": 1.571577267365207, + "grad_norm": 7.010601043701172, + "learning_rate": 0.0001385264418564428, + "loss": 0.7732, + "step": 22648 + }, + { + "epoch": 1.5716466588023037, + "grad_norm": 1.8428469896316528, + "learning_rate": 0.00013852125690230605, + "loss": 0.6901, + "step": 22649 + }, + { + "epoch": 1.5717160502394005, + "grad_norm": 1.5711385011672974, + "learning_rate": 0.00013851607182656215, + "loss": 0.5951, + "step": 22650 + }, + { + "epoch": 1.5717854416764971, + "grad_norm": 1.9660698175430298, + "learning_rate": 0.0001385108866292275, + "loss": 0.7895, + "step": 22651 + }, + { + "epoch": 1.5718548331135938, + "grad_norm": 1.6868996620178223, + "learning_rate": 0.00013850570131031847, + "loss": 0.6033, + "step": 22652 + }, + { + "epoch": 1.5719242245506906, + "grad_norm": 1.7704774141311646, + "learning_rate": 0.0001385005158698514, + "loss": 0.6577, + "step": 22653 + }, + { + "epoch": 1.571993615987787, + "grad_norm": 1.7682820558547974, + "learning_rate": 0.00013849533030784266, + "loss": 0.6266, + "step": 22654 + }, + { + "epoch": 1.5720630074248838, + "grad_norm": 1.7901313304901123, + "learning_rate": 0.00013849014462430863, + "loss": 0.6504, + "step": 22655 + }, + { + "epoch": 1.5721323988619804, + "grad_norm": 1.6424570083618164, + "learning_rate": 0.00013848495881926568, + "loss": 0.5282, + "step": 22656 + }, + { + "epoch": 1.572201790299077, + "grad_norm": 1.5482736825942993, + "learning_rate": 0.00013847977289273015, + "loss": 0.5739, + "step": 22657 + }, + { + "epoch": 1.5722711817361739, + "grad_norm": 1.8391541242599487, + "learning_rate": 0.00013847458684471847, + "loss": 0.6163, + "step": 22658 + }, + { + "epoch": 1.5723405731732705, + "grad_norm": 2.410658836364746, + "learning_rate": 0.000138469400675247, + "loss": 0.6149, + "step": 22659 + }, + { + "epoch": 1.572409964610367, + "grad_norm": 2.074789524078369, + "learning_rate": 0.00013846421438433207, + "loss": 0.4894, + "step": 22660 + }, + { + "epoch": 1.5724793560474637, + "grad_norm": 1.5146112442016602, + "learning_rate": 0.00013845902797199008, + "loss": 0.4873, + "step": 22661 + }, + { + "epoch": 1.5725487474845603, + "grad_norm": 2.0652995109558105, + "learning_rate": 0.00013845384143823744, + "loss": 0.6118, + "step": 22662 + }, + { + "epoch": 1.5726181389216571, + "grad_norm": 2.4719486236572266, + "learning_rate": 0.00013844865478309046, + "loss": 0.7204, + "step": 22663 + }, + { + "epoch": 1.5726875303587537, + "grad_norm": 1.6236083507537842, + "learning_rate": 0.00013844346800656552, + "loss": 0.4832, + "step": 22664 + }, + { + "epoch": 1.5727569217958504, + "grad_norm": 1.6549092531204224, + "learning_rate": 0.00013843828110867904, + "loss": 0.4681, + "step": 22665 + }, + { + "epoch": 1.5728263132329472, + "grad_norm": 1.6745707988739014, + "learning_rate": 0.00013843309408944731, + "loss": 0.6025, + "step": 22666 + }, + { + "epoch": 1.5728957046700436, + "grad_norm": 1.2497340440750122, + "learning_rate": 0.00013842790694888678, + "loss": 0.4218, + "step": 22667 + }, + { + "epoch": 1.5729650961071404, + "grad_norm": 1.6490060091018677, + "learning_rate": 0.00013842271968701384, + "loss": 0.6838, + "step": 22668 + }, + { + "epoch": 1.573034487544237, + "grad_norm": 1.7035771608352661, + "learning_rate": 0.00013841753230384483, + "loss": 0.6638, + "step": 22669 + }, + { + "epoch": 1.5731038789813336, + "grad_norm": 1.6858341693878174, + "learning_rate": 0.0001384123447993961, + "loss": 0.5483, + "step": 22670 + }, + { + "epoch": 1.5731732704184305, + "grad_norm": 1.4528899192810059, + "learning_rate": 0.00013840715717368407, + "loss": 0.5917, + "step": 22671 + }, + { + "epoch": 1.573242661855527, + "grad_norm": 2.670842170715332, + "learning_rate": 0.00013840196942672508, + "loss": 0.6749, + "step": 22672 + }, + { + "epoch": 1.5733120532926237, + "grad_norm": 1.7591018676757812, + "learning_rate": 0.00013839678155853555, + "loss": 0.547, + "step": 22673 + }, + { + "epoch": 1.5733814447297203, + "grad_norm": 1.655158519744873, + "learning_rate": 0.00013839159356913183, + "loss": 0.5466, + "step": 22674 + }, + { + "epoch": 1.573450836166817, + "grad_norm": 1.5053564310073853, + "learning_rate": 0.0001383864054585303, + "loss": 0.4042, + "step": 22675 + }, + { + "epoch": 1.5735202276039137, + "grad_norm": 1.7738752365112305, + "learning_rate": 0.00013838121722674736, + "loss": 0.6353, + "step": 22676 + }, + { + "epoch": 1.5735896190410104, + "grad_norm": 1.2436559200286865, + "learning_rate": 0.00013837602887379935, + "loss": 0.4511, + "step": 22677 + }, + { + "epoch": 1.573659010478107, + "grad_norm": 2.2427897453308105, + "learning_rate": 0.00013837084039970267, + "loss": 0.6935, + "step": 22678 + }, + { + "epoch": 1.5737284019152038, + "grad_norm": 1.4776676893234253, + "learning_rate": 0.00013836565180447372, + "loss": 0.5989, + "step": 22679 + }, + { + "epoch": 1.5737977933523002, + "grad_norm": 1.7924325466156006, + "learning_rate": 0.00013836046308812882, + "loss": 0.8487, + "step": 22680 + }, + { + "epoch": 1.573867184789397, + "grad_norm": 1.833733081817627, + "learning_rate": 0.00013835527425068444, + "loss": 0.9519, + "step": 22681 + }, + { + "epoch": 1.5739365762264936, + "grad_norm": 1.533184289932251, + "learning_rate": 0.00013835008529215688, + "loss": 0.514, + "step": 22682 + }, + { + "epoch": 1.5740059676635902, + "grad_norm": 1.6208412647247314, + "learning_rate": 0.00013834489621256254, + "loss": 0.5812, + "step": 22683 + }, + { + "epoch": 1.574075359100687, + "grad_norm": 1.5090805292129517, + "learning_rate": 0.00013833970701191783, + "loss": 0.4896, + "step": 22684 + }, + { + "epoch": 1.5741447505377837, + "grad_norm": 1.8267822265625, + "learning_rate": 0.00013833451769023911, + "loss": 0.5957, + "step": 22685 + }, + { + "epoch": 1.5742141419748803, + "grad_norm": 1.6837610006332397, + "learning_rate": 0.00013832932824754278, + "loss": 0.5949, + "step": 22686 + }, + { + "epoch": 1.574283533411977, + "grad_norm": 2.486721992492676, + "learning_rate": 0.00013832413868384517, + "loss": 0.7671, + "step": 22687 + }, + { + "epoch": 1.5743529248490735, + "grad_norm": 1.6462880373001099, + "learning_rate": 0.0001383189489991627, + "loss": 0.5976, + "step": 22688 + }, + { + "epoch": 1.5744223162861704, + "grad_norm": 2.110288143157959, + "learning_rate": 0.00013831375919351178, + "loss": 0.7389, + "step": 22689 + }, + { + "epoch": 1.574491707723267, + "grad_norm": 1.6135530471801758, + "learning_rate": 0.00013830856926690878, + "loss": 0.6727, + "step": 22690 + }, + { + "epoch": 1.5745610991603636, + "grad_norm": 1.3766772747039795, + "learning_rate": 0.00013830337921937, + "loss": 0.5549, + "step": 22691 + }, + { + "epoch": 1.5746304905974604, + "grad_norm": 1.8287451267242432, + "learning_rate": 0.000138298189050912, + "loss": 0.6952, + "step": 22692 + }, + { + "epoch": 1.5746998820345568, + "grad_norm": 1.7417280673980713, + "learning_rate": 0.00013829299876155099, + "loss": 0.3639, + "step": 22693 + }, + { + "epoch": 1.5747692734716536, + "grad_norm": 1.2718342542648315, + "learning_rate": 0.00013828780835130344, + "loss": 0.4781, + "step": 22694 + }, + { + "epoch": 1.5748386649087502, + "grad_norm": 1.395475149154663, + "learning_rate": 0.00013828261782018573, + "loss": 0.6478, + "step": 22695 + }, + { + "epoch": 1.5749080563458469, + "grad_norm": 1.8217999935150146, + "learning_rate": 0.00013827742716821423, + "loss": 0.693, + "step": 22696 + }, + { + "epoch": 1.5749774477829437, + "grad_norm": 1.817655086517334, + "learning_rate": 0.00013827223639540533, + "loss": 0.6002, + "step": 22697 + }, + { + "epoch": 1.5750468392200403, + "grad_norm": 1.3077837228775024, + "learning_rate": 0.00013826704550177544, + "loss": 0.3634, + "step": 22698 + }, + { + "epoch": 1.575116230657137, + "grad_norm": 1.4952154159545898, + "learning_rate": 0.0001382618544873409, + "loss": 0.5451, + "step": 22699 + }, + { + "epoch": 1.5751856220942335, + "grad_norm": 1.572182059288025, + "learning_rate": 0.00013825666335211815, + "loss": 0.595, + "step": 22700 + }, + { + "epoch": 1.5752550135313301, + "grad_norm": 1.1382068395614624, + "learning_rate": 0.00013825147209612355, + "loss": 0.3246, + "step": 22701 + }, + { + "epoch": 1.575324404968427, + "grad_norm": 1.8277438879013062, + "learning_rate": 0.00013824628071937346, + "loss": 0.4727, + "step": 22702 + }, + { + "epoch": 1.5753937964055236, + "grad_norm": 2.3886375427246094, + "learning_rate": 0.00013824108922188435, + "loss": 0.6214, + "step": 22703 + }, + { + "epoch": 1.5754631878426202, + "grad_norm": 1.2992726564407349, + "learning_rate": 0.00013823589760367252, + "loss": 0.6611, + "step": 22704 + }, + { + "epoch": 1.575532579279717, + "grad_norm": 1.5886633396148682, + "learning_rate": 0.0001382307058647544, + "loss": 0.539, + "step": 22705 + }, + { + "epoch": 1.5756019707168134, + "grad_norm": 1.4029977321624756, + "learning_rate": 0.00013822551400514637, + "loss": 0.6096, + "step": 22706 + }, + { + "epoch": 1.5756713621539102, + "grad_norm": 1.2750712633132935, + "learning_rate": 0.00013822032202486485, + "loss": 0.6929, + "step": 22707 + }, + { + "epoch": 1.5757407535910068, + "grad_norm": 1.2255496978759766, + "learning_rate": 0.0001382151299239262, + "loss": 0.5089, + "step": 22708 + }, + { + "epoch": 1.5758101450281035, + "grad_norm": 3.498879909515381, + "learning_rate": 0.0001382099377023468, + "loss": 0.668, + "step": 22709 + }, + { + "epoch": 1.5758795364652003, + "grad_norm": 1.634962558746338, + "learning_rate": 0.0001382047453601431, + "loss": 0.7329, + "step": 22710 + }, + { + "epoch": 1.575948927902297, + "grad_norm": 2.3802032470703125, + "learning_rate": 0.00013819955289733142, + "loss": 0.7446, + "step": 22711 + }, + { + "epoch": 1.5760183193393935, + "grad_norm": 1.3604261875152588, + "learning_rate": 0.00013819436031392817, + "loss": 0.5397, + "step": 22712 + }, + { + "epoch": 1.5760877107764903, + "grad_norm": 1.54496431350708, + "learning_rate": 0.00013818916760994978, + "loss": 0.72, + "step": 22713 + }, + { + "epoch": 1.5761571022135867, + "grad_norm": 1.7732089757919312, + "learning_rate": 0.00013818397478541263, + "loss": 0.4645, + "step": 22714 + }, + { + "epoch": 1.5762264936506836, + "grad_norm": 3.13175106048584, + "learning_rate": 0.0001381787818403331, + "loss": 0.733, + "step": 22715 + }, + { + "epoch": 1.5762958850877802, + "grad_norm": 1.6762677431106567, + "learning_rate": 0.00013817358877472757, + "loss": 0.5874, + "step": 22716 + }, + { + "epoch": 1.5763652765248768, + "grad_norm": 2.3443734645843506, + "learning_rate": 0.00013816839558861244, + "loss": 0.6656, + "step": 22717 + }, + { + "epoch": 1.5764346679619736, + "grad_norm": 1.9311912059783936, + "learning_rate": 0.00013816320228200412, + "loss": 0.7388, + "step": 22718 + }, + { + "epoch": 1.57650405939907, + "grad_norm": 1.6149476766586304, + "learning_rate": 0.000138158008854919, + "loss": 0.4097, + "step": 22719 + }, + { + "epoch": 1.5765734508361668, + "grad_norm": 1.5509172677993774, + "learning_rate": 0.0001381528153073735, + "loss": 0.581, + "step": 22720 + }, + { + "epoch": 1.5766428422732635, + "grad_norm": 1.4042030572891235, + "learning_rate": 0.00013814762163938396, + "loss": 0.5052, + "step": 22721 + }, + { + "epoch": 1.57671223371036, + "grad_norm": 2.1646182537078857, + "learning_rate": 0.0001381424278509668, + "loss": 0.7172, + "step": 22722 + }, + { + "epoch": 1.576781625147457, + "grad_norm": 3.3164713382720947, + "learning_rate": 0.00013813723394213843, + "loss": 0.5781, + "step": 22723 + }, + { + "epoch": 1.5768510165845535, + "grad_norm": 1.6733720302581787, + "learning_rate": 0.00013813203991291527, + "loss": 0.479, + "step": 22724 + }, + { + "epoch": 1.5769204080216501, + "grad_norm": 1.607609510421753, + "learning_rate": 0.00013812684576331363, + "loss": 0.7009, + "step": 22725 + }, + { + "epoch": 1.576989799458747, + "grad_norm": 1.5006033182144165, + "learning_rate": 0.00013812165149334996, + "loss": 0.5509, + "step": 22726 + }, + { + "epoch": 1.5770591908958433, + "grad_norm": 1.435890793800354, + "learning_rate": 0.0001381164571030407, + "loss": 0.4843, + "step": 22727 + }, + { + "epoch": 1.5771285823329402, + "grad_norm": 1.9194906949996948, + "learning_rate": 0.0001381112625924022, + "loss": 0.5783, + "step": 22728 + }, + { + "epoch": 1.5771979737700368, + "grad_norm": 1.347439169883728, + "learning_rate": 0.00013810606796145085, + "loss": 0.699, + "step": 22729 + }, + { + "epoch": 1.5772673652071334, + "grad_norm": 1.4298545122146606, + "learning_rate": 0.00013810087321020303, + "loss": 0.5571, + "step": 22730 + }, + { + "epoch": 1.5773367566442302, + "grad_norm": 1.4830173254013062, + "learning_rate": 0.00013809567833867524, + "loss": 0.3327, + "step": 22731 + }, + { + "epoch": 1.5774061480813266, + "grad_norm": 1.5295249223709106, + "learning_rate": 0.00013809048334688379, + "loss": 0.4383, + "step": 22732 + }, + { + "epoch": 1.5774755395184235, + "grad_norm": 1.9575011730194092, + "learning_rate": 0.0001380852882348451, + "loss": 0.6659, + "step": 22733 + }, + { + "epoch": 1.57754493095552, + "grad_norm": 2.5943257808685303, + "learning_rate": 0.0001380800930025756, + "loss": 0.641, + "step": 22734 + }, + { + "epoch": 1.5776143223926167, + "grad_norm": 1.6308696269989014, + "learning_rate": 0.00013807489765009162, + "loss": 0.4674, + "step": 22735 + }, + { + "epoch": 1.5776837138297135, + "grad_norm": 1.9700487852096558, + "learning_rate": 0.0001380697021774096, + "loss": 0.5813, + "step": 22736 + }, + { + "epoch": 1.5777531052668101, + "grad_norm": 1.7628906965255737, + "learning_rate": 0.00013806450658454598, + "loss": 0.8609, + "step": 22737 + }, + { + "epoch": 1.5778224967039067, + "grad_norm": 1.3883646726608276, + "learning_rate": 0.00013805931087151714, + "loss": 0.6966, + "step": 22738 + }, + { + "epoch": 1.5778918881410036, + "grad_norm": 2.0202131271362305, + "learning_rate": 0.00013805411503833945, + "loss": 0.6178, + "step": 22739 + }, + { + "epoch": 1.5779612795781, + "grad_norm": 1.2241665124893188, + "learning_rate": 0.00013804891908502932, + "loss": 0.5121, + "step": 22740 + }, + { + "epoch": 1.5780306710151968, + "grad_norm": 1.5410939455032349, + "learning_rate": 0.0001380437230116032, + "loss": 0.5947, + "step": 22741 + }, + { + "epoch": 1.5781000624522934, + "grad_norm": 1.4607529640197754, + "learning_rate": 0.00013803852681807744, + "loss": 0.6856, + "step": 22742 + }, + { + "epoch": 1.57816945388939, + "grad_norm": 1.4944539070129395, + "learning_rate": 0.00013803333050446846, + "loss": 0.5815, + "step": 22743 + }, + { + "epoch": 1.5782388453264868, + "grad_norm": 2.4695520401000977, + "learning_rate": 0.00013802813407079266, + "loss": 0.6659, + "step": 22744 + }, + { + "epoch": 1.5783082367635832, + "grad_norm": 1.2632781267166138, + "learning_rate": 0.00013802293751706646, + "loss": 0.5686, + "step": 22745 + }, + { + "epoch": 1.57837762820068, + "grad_norm": 1.9656509160995483, + "learning_rate": 0.0001380177408433063, + "loss": 0.7237, + "step": 22746 + }, + { + "epoch": 1.5784470196377767, + "grad_norm": 1.4944514036178589, + "learning_rate": 0.00013801254404952848, + "loss": 0.5561, + "step": 22747 + }, + { + "epoch": 1.5785164110748733, + "grad_norm": 1.9296661615371704, + "learning_rate": 0.00013800734713574953, + "loss": 0.8415, + "step": 22748 + }, + { + "epoch": 1.5785858025119701, + "grad_norm": 1.8211603164672852, + "learning_rate": 0.00013800215010198574, + "loss": 0.847, + "step": 22749 + }, + { + "epoch": 1.5786551939490667, + "grad_norm": 1.6799631118774414, + "learning_rate": 0.00013799695294825358, + "loss": 0.6144, + "step": 22750 + }, + { + "epoch": 1.5787245853861633, + "grad_norm": 1.8549984693527222, + "learning_rate": 0.00013799175567456948, + "loss": 0.6649, + "step": 22751 + }, + { + "epoch": 1.5787939768232602, + "grad_norm": 1.5417871475219727, + "learning_rate": 0.0001379865582809498, + "loss": 0.4631, + "step": 22752 + }, + { + "epoch": 1.5788633682603566, + "grad_norm": 1.5120645761489868, + "learning_rate": 0.00013798136076741097, + "loss": 0.4923, + "step": 22753 + }, + { + "epoch": 1.5789327596974534, + "grad_norm": 1.9470282793045044, + "learning_rate": 0.00013797616313396938, + "loss": 0.7955, + "step": 22754 + }, + { + "epoch": 1.57900215113455, + "grad_norm": 1.3414714336395264, + "learning_rate": 0.00013797096538064146, + "loss": 0.4811, + "step": 22755 + }, + { + "epoch": 1.5790715425716466, + "grad_norm": 1.7923822402954102, + "learning_rate": 0.0001379657675074436, + "loss": 0.5191, + "step": 22756 + }, + { + "epoch": 1.5791409340087434, + "grad_norm": 1.1513580083847046, + "learning_rate": 0.00013796056951439218, + "loss": 0.4292, + "step": 22757 + }, + { + "epoch": 1.5792103254458398, + "grad_norm": 1.4530287981033325, + "learning_rate": 0.0001379553714015037, + "loss": 0.6084, + "step": 22758 + }, + { + "epoch": 1.5792797168829367, + "grad_norm": 2.1165144443511963, + "learning_rate": 0.00013795017316879448, + "loss": 0.766, + "step": 22759 + }, + { + "epoch": 1.5793491083200333, + "grad_norm": 1.5844573974609375, + "learning_rate": 0.00013794497481628098, + "loss": 0.5026, + "step": 22760 + }, + { + "epoch": 1.57941849975713, + "grad_norm": 2.250788927078247, + "learning_rate": 0.0001379397763439796, + "loss": 0.7014, + "step": 22761 + }, + { + "epoch": 1.5794878911942267, + "grad_norm": 1.5057169198989868, + "learning_rate": 0.00013793457775190677, + "loss": 0.5896, + "step": 22762 + }, + { + "epoch": 1.5795572826313233, + "grad_norm": 1.9654686450958252, + "learning_rate": 0.00013792937904007885, + "loss": 0.5934, + "step": 22763 + }, + { + "epoch": 1.57962667406842, + "grad_norm": 2.1438069343566895, + "learning_rate": 0.00013792418020851227, + "loss": 0.6017, + "step": 22764 + }, + { + "epoch": 1.5796960655055168, + "grad_norm": 1.3780078887939453, + "learning_rate": 0.00013791898125722346, + "loss": 0.4466, + "step": 22765 + }, + { + "epoch": 1.5797654569426132, + "grad_norm": 2.9262032508850098, + "learning_rate": 0.00013791378218622885, + "loss": 0.6622, + "step": 22766 + }, + { + "epoch": 1.57983484837971, + "grad_norm": 1.7709946632385254, + "learning_rate": 0.0001379085829955448, + "loss": 0.3857, + "step": 22767 + }, + { + "epoch": 1.5799042398168066, + "grad_norm": 1.2596323490142822, + "learning_rate": 0.00013790338368518777, + "loss": 0.5573, + "step": 22768 + }, + { + "epoch": 1.5799736312539032, + "grad_norm": 2.2769551277160645, + "learning_rate": 0.00013789818425517417, + "loss": 0.9863, + "step": 22769 + }, + { + "epoch": 1.580043022691, + "grad_norm": 1.5438822507858276, + "learning_rate": 0.00013789298470552038, + "loss": 0.5433, + "step": 22770 + }, + { + "epoch": 1.5801124141280964, + "grad_norm": 1.640572428703308, + "learning_rate": 0.00013788778503624286, + "loss": 0.6298, + "step": 22771 + }, + { + "epoch": 1.5801818055651933, + "grad_norm": 3.0272316932678223, + "learning_rate": 0.00013788258524735796, + "loss": 0.7234, + "step": 22772 + }, + { + "epoch": 1.5802511970022899, + "grad_norm": 1.3331561088562012, + "learning_rate": 0.00013787738533888215, + "loss": 0.4968, + "step": 22773 + }, + { + "epoch": 1.5803205884393865, + "grad_norm": 1.7411352396011353, + "learning_rate": 0.00013787218531083182, + "loss": 0.6331, + "step": 22774 + }, + { + "epoch": 1.5803899798764833, + "grad_norm": 1.5646580457687378, + "learning_rate": 0.00013786698516322338, + "loss": 0.5993, + "step": 22775 + }, + { + "epoch": 1.58045937131358, + "grad_norm": 1.514456868171692, + "learning_rate": 0.00013786178489607332, + "loss": 0.876, + "step": 22776 + }, + { + "epoch": 1.5805287627506766, + "grad_norm": 1.8397959470748901, + "learning_rate": 0.00013785658450939794, + "loss": 0.561, + "step": 22777 + }, + { + "epoch": 1.5805981541877734, + "grad_norm": 1.8556115627288818, + "learning_rate": 0.0001378513840032137, + "loss": 0.626, + "step": 22778 + }, + { + "epoch": 1.5806675456248698, + "grad_norm": 1.5396395921707153, + "learning_rate": 0.00013784618337753712, + "loss": 0.4845, + "step": 22779 + }, + { + "epoch": 1.5807369370619666, + "grad_norm": 1.8185099363327026, + "learning_rate": 0.00013784098263238445, + "loss": 0.4826, + "step": 22780 + }, + { + "epoch": 1.5808063284990632, + "grad_norm": 1.5269407033920288, + "learning_rate": 0.0001378357817677722, + "loss": 0.4937, + "step": 22781 + }, + { + "epoch": 1.5808757199361598, + "grad_norm": 1.4565966129302979, + "learning_rate": 0.0001378305807837168, + "loss": 0.555, + "step": 22782 + }, + { + "epoch": 1.5809451113732567, + "grad_norm": 1.708241581916809, + "learning_rate": 0.00013782537968023466, + "loss": 0.7196, + "step": 22783 + }, + { + "epoch": 1.581014502810353, + "grad_norm": 1.9747283458709717, + "learning_rate": 0.00013782017845734214, + "loss": 0.7721, + "step": 22784 + }, + { + "epoch": 1.5810838942474499, + "grad_norm": 1.3623206615447998, + "learning_rate": 0.00013781497711505574, + "loss": 0.6624, + "step": 22785 + }, + { + "epoch": 1.5811532856845465, + "grad_norm": 2.0151777267456055, + "learning_rate": 0.00013780977565339183, + "loss": 0.9366, + "step": 22786 + }, + { + "epoch": 1.581222677121643, + "grad_norm": 2.0979413986206055, + "learning_rate": 0.00013780457407236687, + "loss": 0.5175, + "step": 22787 + }, + { + "epoch": 1.58129206855874, + "grad_norm": 1.6890698671340942, + "learning_rate": 0.0001377993723719972, + "loss": 0.7004, + "step": 22788 + }, + { + "epoch": 1.5813614599958365, + "grad_norm": 2.013120412826538, + "learning_rate": 0.00013779417055229934, + "loss": 0.76, + "step": 22789 + }, + { + "epoch": 1.5814308514329332, + "grad_norm": 1.367562174797058, + "learning_rate": 0.00013778896861328966, + "loss": 0.4947, + "step": 22790 + }, + { + "epoch": 1.58150024287003, + "grad_norm": 1.340962529182434, + "learning_rate": 0.0001377837665549846, + "loss": 0.4441, + "step": 22791 + }, + { + "epoch": 1.5815696343071264, + "grad_norm": 1.7819254398345947, + "learning_rate": 0.00013777856437740055, + "loss": 0.7787, + "step": 22792 + }, + { + "epoch": 1.5816390257442232, + "grad_norm": 1.7610212564468384, + "learning_rate": 0.00013777336208055399, + "loss": 0.5752, + "step": 22793 + }, + { + "epoch": 1.5817084171813198, + "grad_norm": 1.7815945148468018, + "learning_rate": 0.00013776815966446126, + "loss": 0.6274, + "step": 22794 + }, + { + "epoch": 1.5817778086184164, + "grad_norm": 1.1686080694198608, + "learning_rate": 0.00013776295712913884, + "loss": 0.396, + "step": 22795 + }, + { + "epoch": 1.5818472000555133, + "grad_norm": 1.695518970489502, + "learning_rate": 0.00013775775447460317, + "loss": 0.5741, + "step": 22796 + }, + { + "epoch": 1.5819165914926097, + "grad_norm": 2.334348201751709, + "learning_rate": 0.00013775255170087063, + "loss": 0.7821, + "step": 22797 + }, + { + "epoch": 1.5819859829297065, + "grad_norm": 2.1821303367614746, + "learning_rate": 0.0001377473488079577, + "loss": 0.9024, + "step": 22798 + }, + { + "epoch": 1.582055374366803, + "grad_norm": 1.8119935989379883, + "learning_rate": 0.00013774214579588074, + "loss": 0.6338, + "step": 22799 + }, + { + "epoch": 1.5821247658038997, + "grad_norm": 1.5856884717941284, + "learning_rate": 0.00013773694266465622, + "loss": 0.7191, + "step": 22800 + }, + { + "epoch": 1.5821941572409965, + "grad_norm": 1.696885585784912, + "learning_rate": 0.00013773173941430053, + "loss": 0.8574, + "step": 22801 + }, + { + "epoch": 1.5822635486780932, + "grad_norm": 1.6733840703964233, + "learning_rate": 0.00013772653604483012, + "loss": 0.5794, + "step": 22802 + }, + { + "epoch": 1.5823329401151898, + "grad_norm": 1.3760557174682617, + "learning_rate": 0.00013772133255626144, + "loss": 0.6107, + "step": 22803 + }, + { + "epoch": 1.5824023315522866, + "grad_norm": 1.5271238088607788, + "learning_rate": 0.00013771612894861084, + "loss": 0.7276, + "step": 22804 + }, + { + "epoch": 1.582471722989383, + "grad_norm": 1.3744877576828003, + "learning_rate": 0.00013771092522189485, + "loss": 0.621, + "step": 22805 + }, + { + "epoch": 1.5825411144264798, + "grad_norm": 1.4401623010635376, + "learning_rate": 0.00013770572137612982, + "loss": 0.699, + "step": 22806 + }, + { + "epoch": 1.5826105058635764, + "grad_norm": 1.4995954036712646, + "learning_rate": 0.00013770051741133222, + "loss": 0.6307, + "step": 22807 + }, + { + "epoch": 1.582679897300673, + "grad_norm": 1.8195873498916626, + "learning_rate": 0.00013769531332751843, + "loss": 0.8727, + "step": 22808 + }, + { + "epoch": 1.5827492887377699, + "grad_norm": 1.6464561223983765, + "learning_rate": 0.00013769010912470492, + "loss": 0.7409, + "step": 22809 + }, + { + "epoch": 1.5828186801748663, + "grad_norm": 1.8889859914779663, + "learning_rate": 0.0001376849048029081, + "loss": 0.5446, + "step": 22810 + }, + { + "epoch": 1.582888071611963, + "grad_norm": 1.3224432468414307, + "learning_rate": 0.00013767970036214443, + "loss": 0.499, + "step": 22811 + }, + { + "epoch": 1.5829574630490597, + "grad_norm": 2.6011815071105957, + "learning_rate": 0.0001376744958024303, + "loss": 0.6214, + "step": 22812 + }, + { + "epoch": 1.5830268544861563, + "grad_norm": 1.4774373769760132, + "learning_rate": 0.00013766929112378216, + "loss": 0.4982, + "step": 22813 + }, + { + "epoch": 1.5830962459232532, + "grad_norm": 1.5272120237350464, + "learning_rate": 0.00013766408632621645, + "loss": 0.6723, + "step": 22814 + }, + { + "epoch": 1.5831656373603498, + "grad_norm": 1.489487648010254, + "learning_rate": 0.00013765888140974958, + "loss": 0.5523, + "step": 22815 + }, + { + "epoch": 1.5832350287974464, + "grad_norm": 2.5547404289245605, + "learning_rate": 0.00013765367637439798, + "loss": 0.5772, + "step": 22816 + }, + { + "epoch": 1.5833044202345432, + "grad_norm": 2.285067558288574, + "learning_rate": 0.00013764847122017812, + "loss": 0.6683, + "step": 22817 + }, + { + "epoch": 1.5833738116716396, + "grad_norm": 1.6659669876098633, + "learning_rate": 0.0001376432659471064, + "loss": 0.5044, + "step": 22818 + }, + { + "epoch": 1.5834432031087364, + "grad_norm": 1.3473273515701294, + "learning_rate": 0.00013763806055519924, + "loss": 0.3036, + "step": 22819 + }, + { + "epoch": 1.583512594545833, + "grad_norm": 1.5525319576263428, + "learning_rate": 0.0001376328550444731, + "loss": 0.4422, + "step": 22820 + }, + { + "epoch": 1.5835819859829297, + "grad_norm": 1.5543256998062134, + "learning_rate": 0.0001376276494149444, + "loss": 0.5232, + "step": 22821 + }, + { + "epoch": 1.5836513774200265, + "grad_norm": 1.7901583909988403, + "learning_rate": 0.00013762244366662956, + "loss": 0.8306, + "step": 22822 + }, + { + "epoch": 1.5837207688571229, + "grad_norm": 1.5724973678588867, + "learning_rate": 0.00013761723779954506, + "loss": 0.7666, + "step": 22823 + }, + { + "epoch": 1.5837901602942197, + "grad_norm": 1.333107352256775, + "learning_rate": 0.0001376120318137073, + "loss": 0.4983, + "step": 22824 + }, + { + "epoch": 1.5838595517313163, + "grad_norm": 3.128655433654785, + "learning_rate": 0.0001376068257091327, + "loss": 0.6387, + "step": 22825 + }, + { + "epoch": 1.583928943168413, + "grad_norm": 2.516268253326416, + "learning_rate": 0.0001376016194858377, + "loss": 0.5528, + "step": 22826 + }, + { + "epoch": 1.5839983346055098, + "grad_norm": 1.695046067237854, + "learning_rate": 0.0001375964131438388, + "loss": 0.4956, + "step": 22827 + }, + { + "epoch": 1.5840677260426064, + "grad_norm": 1.9086347818374634, + "learning_rate": 0.00013759120668315234, + "loss": 0.7336, + "step": 22828 + }, + { + "epoch": 1.584137117479703, + "grad_norm": 1.9409456253051758, + "learning_rate": 0.00013758600010379483, + "loss": 0.6647, + "step": 22829 + }, + { + "epoch": 1.5842065089167998, + "grad_norm": 2.2768490314483643, + "learning_rate": 0.00013758079340578266, + "loss": 0.774, + "step": 22830 + }, + { + "epoch": 1.5842759003538962, + "grad_norm": 1.2758897542953491, + "learning_rate": 0.00013757558658913233, + "loss": 0.4105, + "step": 22831 + }, + { + "epoch": 1.584345291790993, + "grad_norm": 1.8875916004180908, + "learning_rate": 0.0001375703796538602, + "loss": 0.6099, + "step": 22832 + }, + { + "epoch": 1.5844146832280896, + "grad_norm": 1.7916641235351562, + "learning_rate": 0.00013756517259998273, + "loss": 0.6855, + "step": 22833 + }, + { + "epoch": 1.5844840746651863, + "grad_norm": 1.87177574634552, + "learning_rate": 0.00013755996542751635, + "loss": 0.9353, + "step": 22834 + }, + { + "epoch": 1.584553466102283, + "grad_norm": 2.35117506980896, + "learning_rate": 0.00013755475813647755, + "loss": 0.6587, + "step": 22835 + }, + { + "epoch": 1.5846228575393795, + "grad_norm": 1.7602794170379639, + "learning_rate": 0.00013754955072688274, + "loss": 0.5801, + "step": 22836 + }, + { + "epoch": 1.5846922489764763, + "grad_norm": 1.4000290632247925, + "learning_rate": 0.00013754434319874833, + "loss": 0.5532, + "step": 22837 + }, + { + "epoch": 1.584761640413573, + "grad_norm": 1.9211708307266235, + "learning_rate": 0.00013753913555209082, + "loss": 0.6033, + "step": 22838 + }, + { + "epoch": 1.5848310318506695, + "grad_norm": 2.19164776802063, + "learning_rate": 0.0001375339277869266, + "loss": 0.7324, + "step": 22839 + }, + { + "epoch": 1.5849004232877664, + "grad_norm": 1.35867178440094, + "learning_rate": 0.00013752871990327213, + "loss": 0.6298, + "step": 22840 + }, + { + "epoch": 1.584969814724863, + "grad_norm": 1.5962519645690918, + "learning_rate": 0.0001375235119011438, + "loss": 0.6099, + "step": 22841 + }, + { + "epoch": 1.5850392061619596, + "grad_norm": 1.4914329051971436, + "learning_rate": 0.00013751830378055816, + "loss": 0.521, + "step": 22842 + }, + { + "epoch": 1.5851085975990564, + "grad_norm": 1.175559639930725, + "learning_rate": 0.00013751309554153152, + "loss": 0.4503, + "step": 22843 + }, + { + "epoch": 1.5851779890361528, + "grad_norm": 1.441549301147461, + "learning_rate": 0.00013750788718408044, + "loss": 0.6915, + "step": 22844 + }, + { + "epoch": 1.5852473804732496, + "grad_norm": 2.201246500015259, + "learning_rate": 0.0001375026787082213, + "loss": 0.6292, + "step": 22845 + }, + { + "epoch": 1.5853167719103463, + "grad_norm": 1.7218623161315918, + "learning_rate": 0.00013749747011397055, + "loss": 0.5918, + "step": 22846 + }, + { + "epoch": 1.5853861633474429, + "grad_norm": 1.7639936208724976, + "learning_rate": 0.00013749226140134465, + "loss": 0.6241, + "step": 22847 + }, + { + "epoch": 1.5854555547845397, + "grad_norm": 1.9657865762710571, + "learning_rate": 0.00013748705257036, + "loss": 0.7337, + "step": 22848 + }, + { + "epoch": 1.585524946221636, + "grad_norm": 1.4671679735183716, + "learning_rate": 0.00013748184362103315, + "loss": 0.6173, + "step": 22849 + }, + { + "epoch": 1.585594337658733, + "grad_norm": 1.7786952257156372, + "learning_rate": 0.0001374766345533804, + "loss": 0.6625, + "step": 22850 + }, + { + "epoch": 1.5856637290958295, + "grad_norm": 1.7830605506896973, + "learning_rate": 0.00013747142536741825, + "loss": 0.7829, + "step": 22851 + }, + { + "epoch": 1.5857331205329261, + "grad_norm": 1.754504919052124, + "learning_rate": 0.0001374662160631632, + "loss": 0.811, + "step": 22852 + }, + { + "epoch": 1.585802511970023, + "grad_norm": 3.501016616821289, + "learning_rate": 0.0001374610066406316, + "loss": 0.7046, + "step": 22853 + }, + { + "epoch": 1.5858719034071196, + "grad_norm": 1.3591995239257812, + "learning_rate": 0.00013745579709984002, + "loss": 0.5075, + "step": 22854 + }, + { + "epoch": 1.5859412948442162, + "grad_norm": 1.9495242834091187, + "learning_rate": 0.0001374505874408048, + "loss": 0.6074, + "step": 22855 + }, + { + "epoch": 1.586010686281313, + "grad_norm": 1.8690598011016846, + "learning_rate": 0.00013744537766354243, + "loss": 0.4421, + "step": 22856 + }, + { + "epoch": 1.5860800777184094, + "grad_norm": 2.057671308517456, + "learning_rate": 0.00013744016776806933, + "loss": 0.6485, + "step": 22857 + }, + { + "epoch": 1.5861494691555063, + "grad_norm": 2.5825750827789307, + "learning_rate": 0.000137434957754402, + "loss": 0.6726, + "step": 22858 + }, + { + "epoch": 1.5862188605926029, + "grad_norm": 1.543465495109558, + "learning_rate": 0.0001374297476225568, + "loss": 0.7808, + "step": 22859 + }, + { + "epoch": 1.5862882520296995, + "grad_norm": 1.795109510421753, + "learning_rate": 0.00013742453737255024, + "loss": 0.6976, + "step": 22860 + }, + { + "epoch": 1.5863576434667963, + "grad_norm": 2.000675916671753, + "learning_rate": 0.00013741932700439877, + "loss": 0.6659, + "step": 22861 + }, + { + "epoch": 1.5864270349038927, + "grad_norm": 2.269453525543213, + "learning_rate": 0.00013741411651811887, + "loss": 0.5215, + "step": 22862 + }, + { + "epoch": 1.5864964263409895, + "grad_norm": 2.7729172706604004, + "learning_rate": 0.00013740890591372687, + "loss": 0.6589, + "step": 22863 + }, + { + "epoch": 1.5865658177780861, + "grad_norm": 2.512849807739258, + "learning_rate": 0.00013740369519123932, + "loss": 0.5658, + "step": 22864 + }, + { + "epoch": 1.5866352092151828, + "grad_norm": 1.7270172834396362, + "learning_rate": 0.00013739848435067264, + "loss": 0.5308, + "step": 22865 + }, + { + "epoch": 1.5867046006522796, + "grad_norm": 1.7824560403823853, + "learning_rate": 0.00013739327339204333, + "loss": 0.6333, + "step": 22866 + }, + { + "epoch": 1.5867739920893762, + "grad_norm": 1.2960039377212524, + "learning_rate": 0.00013738806231536774, + "loss": 0.7207, + "step": 22867 + }, + { + "epoch": 1.5868433835264728, + "grad_norm": 2.305307626724243, + "learning_rate": 0.00013738285112066239, + "loss": 0.5538, + "step": 22868 + }, + { + "epoch": 1.5869127749635696, + "grad_norm": 1.7074087858200073, + "learning_rate": 0.00013737763980794374, + "loss": 0.433, + "step": 22869 + }, + { + "epoch": 1.586982166400666, + "grad_norm": 1.6402720212936401, + "learning_rate": 0.0001373724283772282, + "loss": 0.6588, + "step": 22870 + }, + { + "epoch": 1.5870515578377629, + "grad_norm": 1.3585662841796875, + "learning_rate": 0.0001373672168285322, + "loss": 0.412, + "step": 22871 + }, + { + "epoch": 1.5871209492748595, + "grad_norm": 1.8265247344970703, + "learning_rate": 0.0001373620051618723, + "loss": 0.4882, + "step": 22872 + }, + { + "epoch": 1.587190340711956, + "grad_norm": 1.7280408143997192, + "learning_rate": 0.00013735679337726484, + "loss": 0.6145, + "step": 22873 + }, + { + "epoch": 1.587259732149053, + "grad_norm": 1.7682894468307495, + "learning_rate": 0.00013735158147472633, + "loss": 0.5997, + "step": 22874 + }, + { + "epoch": 1.5873291235861493, + "grad_norm": 2.015228509902954, + "learning_rate": 0.00013734636945427321, + "loss": 0.7817, + "step": 22875 + }, + { + "epoch": 1.5873985150232461, + "grad_norm": 1.775712013244629, + "learning_rate": 0.00013734115731592195, + "loss": 0.8922, + "step": 22876 + }, + { + "epoch": 1.5874679064603427, + "grad_norm": 1.7744890451431274, + "learning_rate": 0.000137335945059689, + "loss": 0.7061, + "step": 22877 + }, + { + "epoch": 1.5875372978974394, + "grad_norm": 1.9835405349731445, + "learning_rate": 0.00013733073268559073, + "loss": 0.6139, + "step": 22878 + }, + { + "epoch": 1.5876066893345362, + "grad_norm": 1.3989498615264893, + "learning_rate": 0.00013732552019364374, + "loss": 0.4832, + "step": 22879 + }, + { + "epoch": 1.5876760807716328, + "grad_norm": 1.6520826816558838, + "learning_rate": 0.00013732030758386438, + "loss": 0.693, + "step": 22880 + }, + { + "epoch": 1.5877454722087294, + "grad_norm": 1.858420968055725, + "learning_rate": 0.00013731509485626916, + "loss": 0.6982, + "step": 22881 + }, + { + "epoch": 1.5878148636458262, + "grad_norm": 2.2529098987579346, + "learning_rate": 0.0001373098820108745, + "loss": 0.5543, + "step": 22882 + }, + { + "epoch": 1.5878842550829226, + "grad_norm": 1.7003674507141113, + "learning_rate": 0.00013730466904769687, + "loss": 0.4475, + "step": 22883 + }, + { + "epoch": 1.5879536465200195, + "grad_norm": 1.4729050397872925, + "learning_rate": 0.00013729945596675273, + "loss": 0.6698, + "step": 22884 + }, + { + "epoch": 1.588023037957116, + "grad_norm": 1.8526153564453125, + "learning_rate": 0.00013729424276805852, + "loss": 0.8877, + "step": 22885 + }, + { + "epoch": 1.5880924293942127, + "grad_norm": 1.9137707948684692, + "learning_rate": 0.00013728902945163074, + "loss": 0.6388, + "step": 22886 + }, + { + "epoch": 1.5881618208313095, + "grad_norm": 1.3733807802200317, + "learning_rate": 0.00013728381601748582, + "loss": 0.5291, + "step": 22887 + }, + { + "epoch": 1.5882312122684061, + "grad_norm": 1.579300880432129, + "learning_rate": 0.00013727860246564018, + "loss": 0.5486, + "step": 22888 + }, + { + "epoch": 1.5883006037055027, + "grad_norm": 1.26077401638031, + "learning_rate": 0.00013727338879611036, + "loss": 0.5444, + "step": 22889 + }, + { + "epoch": 1.5883699951425994, + "grad_norm": 1.658134937286377, + "learning_rate": 0.00013726817500891279, + "loss": 0.498, + "step": 22890 + }, + { + "epoch": 1.588439386579696, + "grad_norm": 2.7997348308563232, + "learning_rate": 0.00013726296110406388, + "loss": 0.6568, + "step": 22891 + }, + { + "epoch": 1.5885087780167928, + "grad_norm": 2.0208277702331543, + "learning_rate": 0.00013725774708158012, + "loss": 0.484, + "step": 22892 + }, + { + "epoch": 1.5885781694538894, + "grad_norm": 1.4678270816802979, + "learning_rate": 0.000137252532941478, + "loss": 0.5187, + "step": 22893 + }, + { + "epoch": 1.588647560890986, + "grad_norm": 2.2782366275787354, + "learning_rate": 0.00013724731868377396, + "loss": 0.787, + "step": 22894 + }, + { + "epoch": 1.5887169523280829, + "grad_norm": 1.978822112083435, + "learning_rate": 0.00013724210430848444, + "loss": 0.7809, + "step": 22895 + }, + { + "epoch": 1.5887863437651792, + "grad_norm": 1.5858300924301147, + "learning_rate": 0.0001372368898156259, + "loss": 0.745, + "step": 22896 + }, + { + "epoch": 1.588855735202276, + "grad_norm": 1.5259582996368408, + "learning_rate": 0.00013723167520521485, + "loss": 0.4946, + "step": 22897 + }, + { + "epoch": 1.5889251266393727, + "grad_norm": 1.6527478694915771, + "learning_rate": 0.0001372264604772677, + "loss": 0.7729, + "step": 22898 + }, + { + "epoch": 1.5889945180764693, + "grad_norm": 1.5054268836975098, + "learning_rate": 0.00013722124563180093, + "loss": 0.6725, + "step": 22899 + }, + { + "epoch": 1.5890639095135661, + "grad_norm": 1.8611605167388916, + "learning_rate": 0.00013721603066883104, + "loss": 0.6099, + "step": 22900 + }, + { + "epoch": 1.5891333009506627, + "grad_norm": 1.543164610862732, + "learning_rate": 0.00013721081558837442, + "loss": 0.4041, + "step": 22901 + }, + { + "epoch": 1.5892026923877594, + "grad_norm": 1.6668823957443237, + "learning_rate": 0.0001372056003904476, + "loss": 0.4991, + "step": 22902 + }, + { + "epoch": 1.589272083824856, + "grad_norm": 2.193953514099121, + "learning_rate": 0.000137200385075067, + "loss": 0.6953, + "step": 22903 + }, + { + "epoch": 1.5893414752619526, + "grad_norm": 2.2817022800445557, + "learning_rate": 0.0001371951696422491, + "loss": 0.7505, + "step": 22904 + }, + { + "epoch": 1.5894108666990494, + "grad_norm": 2.309921979904175, + "learning_rate": 0.00013718995409201036, + "loss": 0.6417, + "step": 22905 + }, + { + "epoch": 1.589480258136146, + "grad_norm": 2.7333285808563232, + "learning_rate": 0.00013718473842436725, + "loss": 0.5451, + "step": 22906 + }, + { + "epoch": 1.5895496495732426, + "grad_norm": 1.916327714920044, + "learning_rate": 0.00013717952263933626, + "loss": 0.4521, + "step": 22907 + }, + { + "epoch": 1.5896190410103395, + "grad_norm": 1.809524655342102, + "learning_rate": 0.0001371743067369338, + "loss": 0.6294, + "step": 22908 + }, + { + "epoch": 1.5896884324474359, + "grad_norm": 1.8525077104568481, + "learning_rate": 0.00013716909071717642, + "loss": 0.7072, + "step": 22909 + }, + { + "epoch": 1.5897578238845327, + "grad_norm": 2.3460967540740967, + "learning_rate": 0.0001371638745800805, + "loss": 0.5327, + "step": 22910 + }, + { + "epoch": 1.5898272153216293, + "grad_norm": 1.803574562072754, + "learning_rate": 0.00013715865832566253, + "loss": 0.653, + "step": 22911 + }, + { + "epoch": 1.589896606758726, + "grad_norm": 1.637646198272705, + "learning_rate": 0.000137153441953939, + "loss": 0.5562, + "step": 22912 + }, + { + "epoch": 1.5899659981958227, + "grad_norm": 2.1185951232910156, + "learning_rate": 0.00013714822546492635, + "loss": 0.756, + "step": 22913 + }, + { + "epoch": 1.5900353896329193, + "grad_norm": 2.2395851612091064, + "learning_rate": 0.00013714300885864106, + "loss": 0.619, + "step": 22914 + }, + { + "epoch": 1.590104781070016, + "grad_norm": 1.5416226387023926, + "learning_rate": 0.0001371377921350996, + "loss": 0.5066, + "step": 22915 + }, + { + "epoch": 1.5901741725071126, + "grad_norm": 2.07092547416687, + "learning_rate": 0.00013713257529431844, + "loss": 0.4957, + "step": 22916 + }, + { + "epoch": 1.5902435639442092, + "grad_norm": 1.810107707977295, + "learning_rate": 0.00013712735833631406, + "loss": 0.6866, + "step": 22917 + }, + { + "epoch": 1.590312955381306, + "grad_norm": 1.7508338689804077, + "learning_rate": 0.00013712214126110293, + "loss": 0.4823, + "step": 22918 + }, + { + "epoch": 1.5903823468184026, + "grad_norm": 1.8730567693710327, + "learning_rate": 0.00013711692406870147, + "loss": 0.4649, + "step": 22919 + }, + { + "epoch": 1.5904517382554992, + "grad_norm": 1.8236011266708374, + "learning_rate": 0.00013711170675912622, + "loss": 0.8373, + "step": 22920 + }, + { + "epoch": 1.590521129692596, + "grad_norm": 1.8822412490844727, + "learning_rate": 0.0001371064893323936, + "loss": 0.7442, + "step": 22921 + }, + { + "epoch": 1.5905905211296925, + "grad_norm": 1.7875808477401733, + "learning_rate": 0.0001371012717885201, + "loss": 0.5961, + "step": 22922 + }, + { + "epoch": 1.5906599125667893, + "grad_norm": 1.6260149478912354, + "learning_rate": 0.00013709605412752217, + "loss": 0.4864, + "step": 22923 + }, + { + "epoch": 1.590729304003886, + "grad_norm": 1.367664098739624, + "learning_rate": 0.00013709083634941632, + "loss": 0.5546, + "step": 22924 + }, + { + "epoch": 1.5907986954409825, + "grad_norm": 1.2593936920166016, + "learning_rate": 0.00013708561845421902, + "loss": 0.4352, + "step": 22925 + }, + { + "epoch": 1.5908680868780793, + "grad_norm": 1.8804210424423218, + "learning_rate": 0.00013708040044194673, + "loss": 0.6155, + "step": 22926 + }, + { + "epoch": 1.590937478315176, + "grad_norm": 2.016019344329834, + "learning_rate": 0.00013707518231261592, + "loss": 0.8277, + "step": 22927 + }, + { + "epoch": 1.5910068697522726, + "grad_norm": 2.538940906524658, + "learning_rate": 0.00013706996406624304, + "loss": 0.6533, + "step": 22928 + }, + { + "epoch": 1.5910762611893692, + "grad_norm": 1.7431614398956299, + "learning_rate": 0.00013706474570284458, + "loss": 0.6074, + "step": 22929 + }, + { + "epoch": 1.5911456526264658, + "grad_norm": 1.839585542678833, + "learning_rate": 0.000137059527222437, + "loss": 0.5898, + "step": 22930 + }, + { + "epoch": 1.5912150440635626, + "grad_norm": 1.8034623861312866, + "learning_rate": 0.00013705430862503683, + "loss": 0.5521, + "step": 22931 + }, + { + "epoch": 1.5912844355006592, + "grad_norm": 1.5331519842147827, + "learning_rate": 0.0001370490899106605, + "loss": 0.5513, + "step": 22932 + }, + { + "epoch": 1.5913538269377558, + "grad_norm": 1.7582190036773682, + "learning_rate": 0.00013704387107932447, + "loss": 0.6784, + "step": 22933 + }, + { + "epoch": 1.5914232183748527, + "grad_norm": 2.505082607269287, + "learning_rate": 0.00013703865213104526, + "loss": 0.6773, + "step": 22934 + }, + { + "epoch": 1.591492609811949, + "grad_norm": 2.0677261352539062, + "learning_rate": 0.00013703343306583932, + "loss": 0.7944, + "step": 22935 + }, + { + "epoch": 1.591562001249046, + "grad_norm": 1.6734603643417358, + "learning_rate": 0.0001370282138837231, + "loss": 0.7455, + "step": 22936 + }, + { + "epoch": 1.5916313926861425, + "grad_norm": 1.456794023513794, + "learning_rate": 0.00013702299458471313, + "loss": 0.7382, + "step": 22937 + }, + { + "epoch": 1.5917007841232391, + "grad_norm": 1.1894810199737549, + "learning_rate": 0.00013701777516882588, + "loss": 0.3649, + "step": 22938 + }, + { + "epoch": 1.591770175560336, + "grad_norm": 1.8690533638000488, + "learning_rate": 0.00013701255563607778, + "loss": 0.6301, + "step": 22939 + }, + { + "epoch": 1.5918395669974326, + "grad_norm": 2.494978666305542, + "learning_rate": 0.00013700733598648532, + "loss": 0.8124, + "step": 22940 + }, + { + "epoch": 1.5919089584345292, + "grad_norm": 1.8362699747085571, + "learning_rate": 0.000137002116220065, + "loss": 0.7024, + "step": 22941 + }, + { + "epoch": 1.5919783498716258, + "grad_norm": 2.2065508365631104, + "learning_rate": 0.00013699689633683333, + "loss": 0.5958, + "step": 22942 + }, + { + "epoch": 1.5920477413087224, + "grad_norm": 2.1011292934417725, + "learning_rate": 0.00013699167633680675, + "loss": 1.0326, + "step": 22943 + }, + { + "epoch": 1.5921171327458192, + "grad_norm": 1.6154309511184692, + "learning_rate": 0.0001369864562200017, + "loss": 0.559, + "step": 22944 + }, + { + "epoch": 1.5921865241829158, + "grad_norm": 1.760011076927185, + "learning_rate": 0.00013698123598643473, + "loss": 0.5967, + "step": 22945 + }, + { + "epoch": 1.5922559156200125, + "grad_norm": 2.0005218982696533, + "learning_rate": 0.00013697601563612228, + "loss": 0.6348, + "step": 22946 + }, + { + "epoch": 1.5923253070571093, + "grad_norm": 2.603431463241577, + "learning_rate": 0.00013697079516908085, + "loss": 0.4853, + "step": 22947 + }, + { + "epoch": 1.5923946984942057, + "grad_norm": 2.475682497024536, + "learning_rate": 0.0001369655745853269, + "loss": 0.697, + "step": 22948 + }, + { + "epoch": 1.5924640899313025, + "grad_norm": 1.4490082263946533, + "learning_rate": 0.00013696035388487688, + "loss": 0.5577, + "step": 22949 + }, + { + "epoch": 1.5925334813683991, + "grad_norm": 1.496437430381775, + "learning_rate": 0.00013695513306774735, + "loss": 0.4664, + "step": 22950 + }, + { + "epoch": 1.5926028728054957, + "grad_norm": 1.519460916519165, + "learning_rate": 0.00013694991213395477, + "loss": 0.6472, + "step": 22951 + }, + { + "epoch": 1.5926722642425926, + "grad_norm": 1.3575361967086792, + "learning_rate": 0.00013694469108351556, + "loss": 0.2944, + "step": 22952 + }, + { + "epoch": 1.5927416556796892, + "grad_norm": 2.2053136825561523, + "learning_rate": 0.0001369394699164463, + "loss": 0.7266, + "step": 22953 + }, + { + "epoch": 1.5928110471167858, + "grad_norm": 1.4303518533706665, + "learning_rate": 0.00013693424863276337, + "loss": 0.5239, + "step": 22954 + }, + { + "epoch": 1.5928804385538826, + "grad_norm": 1.400506854057312, + "learning_rate": 0.0001369290272324833, + "loss": 0.6908, + "step": 22955 + }, + { + "epoch": 1.592949829990979, + "grad_norm": 1.3682295083999634, + "learning_rate": 0.00013692380571562261, + "loss": 0.3658, + "step": 22956 + }, + { + "epoch": 1.5930192214280758, + "grad_norm": 2.2970099449157715, + "learning_rate": 0.00013691858408219773, + "loss": 0.7278, + "step": 22957 + }, + { + "epoch": 1.5930886128651724, + "grad_norm": 1.1125351190567017, + "learning_rate": 0.00013691336233222516, + "loss": 0.4334, + "step": 22958 + }, + { + "epoch": 1.593158004302269, + "grad_norm": 1.6133395433425903, + "learning_rate": 0.0001369081404657214, + "loss": 0.649, + "step": 22959 + }, + { + "epoch": 1.593227395739366, + "grad_norm": 1.5078004598617554, + "learning_rate": 0.00013690291848270292, + "loss": 0.567, + "step": 22960 + }, + { + "epoch": 1.5932967871764623, + "grad_norm": 1.7413170337677002, + "learning_rate": 0.00013689769638318622, + "loss": 0.6503, + "step": 22961 + }, + { + "epoch": 1.5933661786135591, + "grad_norm": 1.8355498313903809, + "learning_rate": 0.00013689247416718775, + "loss": 0.5017, + "step": 22962 + }, + { + "epoch": 1.5934355700506557, + "grad_norm": 1.7153724431991577, + "learning_rate": 0.00013688725183472403, + "loss": 0.4909, + "step": 22963 + }, + { + "epoch": 1.5935049614877523, + "grad_norm": 1.6177642345428467, + "learning_rate": 0.00013688202938581156, + "loss": 0.6009, + "step": 22964 + }, + { + "epoch": 1.5935743529248492, + "grad_norm": 2.1416828632354736, + "learning_rate": 0.00013687680682046674, + "loss": 0.6298, + "step": 22965 + }, + { + "epoch": 1.5936437443619458, + "grad_norm": 1.6080496311187744, + "learning_rate": 0.00013687158413870618, + "loss": 0.5237, + "step": 22966 + }, + { + "epoch": 1.5937131357990424, + "grad_norm": 2.0994725227355957, + "learning_rate": 0.0001368663613405463, + "loss": 0.4435, + "step": 22967 + }, + { + "epoch": 1.5937825272361392, + "grad_norm": 1.7123808860778809, + "learning_rate": 0.00013686113842600357, + "loss": 0.5557, + "step": 22968 + }, + { + "epoch": 1.5938519186732356, + "grad_norm": 1.8631553649902344, + "learning_rate": 0.0001368559153950945, + "loss": 0.6984, + "step": 22969 + }, + { + "epoch": 1.5939213101103324, + "grad_norm": 1.603476643562317, + "learning_rate": 0.0001368506922478356, + "loss": 0.5206, + "step": 22970 + }, + { + "epoch": 1.593990701547429, + "grad_norm": 1.4357060194015503, + "learning_rate": 0.00013684546898424334, + "loss": 0.5411, + "step": 22971 + }, + { + "epoch": 1.5940600929845257, + "grad_norm": 1.8395861387252808, + "learning_rate": 0.00013684024560433418, + "loss": 0.657, + "step": 22972 + }, + { + "epoch": 1.5941294844216225, + "grad_norm": 1.8136909008026123, + "learning_rate": 0.00013683502210812466, + "loss": 0.6714, + "step": 22973 + }, + { + "epoch": 1.594198875858719, + "grad_norm": 1.5709282159805298, + "learning_rate": 0.00013682979849563126, + "loss": 0.449, + "step": 22974 + }, + { + "epoch": 1.5942682672958157, + "grad_norm": 1.6983665227890015, + "learning_rate": 0.00013682457476687043, + "loss": 0.702, + "step": 22975 + }, + { + "epoch": 1.5943376587329123, + "grad_norm": 1.6967054605484009, + "learning_rate": 0.00013681935092185873, + "loss": 0.5498, + "step": 22976 + }, + { + "epoch": 1.594407050170009, + "grad_norm": 1.643446922302246, + "learning_rate": 0.00013681412696061256, + "loss": 0.4759, + "step": 22977 + }, + { + "epoch": 1.5944764416071058, + "grad_norm": 2.118473768234253, + "learning_rate": 0.00013680890288314852, + "loss": 0.5857, + "step": 22978 + }, + { + "epoch": 1.5945458330442024, + "grad_norm": 1.9222092628479004, + "learning_rate": 0.00013680367868948302, + "loss": 0.6679, + "step": 22979 + }, + { + "epoch": 1.594615224481299, + "grad_norm": 1.8432934284210205, + "learning_rate": 0.00013679845437963257, + "loss": 0.6408, + "step": 22980 + }, + { + "epoch": 1.5946846159183958, + "grad_norm": 1.655666708946228, + "learning_rate": 0.00013679322995361368, + "loss": 0.6335, + "step": 22981 + }, + { + "epoch": 1.5947540073554922, + "grad_norm": 1.6715679168701172, + "learning_rate": 0.0001367880054114428, + "loss": 0.6219, + "step": 22982 + }, + { + "epoch": 1.594823398792589, + "grad_norm": 2.2407920360565186, + "learning_rate": 0.00013678278075313652, + "loss": 0.8214, + "step": 22983 + }, + { + "epoch": 1.5948927902296857, + "grad_norm": 1.5963040590286255, + "learning_rate": 0.0001367775559787112, + "loss": 0.5852, + "step": 22984 + }, + { + "epoch": 1.5949621816667823, + "grad_norm": 1.6951708793640137, + "learning_rate": 0.00013677233108818344, + "loss": 0.9779, + "step": 22985 + }, + { + "epoch": 1.595031573103879, + "grad_norm": 1.5719093084335327, + "learning_rate": 0.00013676710608156965, + "loss": 0.6462, + "step": 22986 + }, + { + "epoch": 1.5951009645409755, + "grad_norm": 1.4328184127807617, + "learning_rate": 0.00013676188095888643, + "loss": 0.4735, + "step": 22987 + }, + { + "epoch": 1.5951703559780723, + "grad_norm": 2.089601993560791, + "learning_rate": 0.0001367566557201502, + "loss": 0.6373, + "step": 22988 + }, + { + "epoch": 1.595239747415169, + "grad_norm": 1.4754520654678345, + "learning_rate": 0.00013675143036537746, + "loss": 0.4501, + "step": 22989 + }, + { + "epoch": 1.5953091388522656, + "grad_norm": 1.361376404762268, + "learning_rate": 0.00013674620489458473, + "loss": 0.4743, + "step": 22990 + }, + { + "epoch": 1.5953785302893624, + "grad_norm": 2.555969476699829, + "learning_rate": 0.0001367409793077885, + "loss": 0.5311, + "step": 22991 + }, + { + "epoch": 1.595447921726459, + "grad_norm": 1.7128328084945679, + "learning_rate": 0.00013673575360500525, + "loss": 0.5461, + "step": 22992 + }, + { + "epoch": 1.5955173131635556, + "grad_norm": 1.8009141683578491, + "learning_rate": 0.00013673052778625147, + "loss": 0.7225, + "step": 22993 + }, + { + "epoch": 1.5955867046006524, + "grad_norm": 3.0328187942504883, + "learning_rate": 0.0001367253018515437, + "loss": 0.9622, + "step": 22994 + }, + { + "epoch": 1.5956560960377488, + "grad_norm": 1.7714866399765015, + "learning_rate": 0.00013672007580089841, + "loss": 0.5025, + "step": 22995 + }, + { + "epoch": 1.5957254874748457, + "grad_norm": 1.6107326745986938, + "learning_rate": 0.00013671484963433212, + "loss": 0.4511, + "step": 22996 + }, + { + "epoch": 1.5957948789119423, + "grad_norm": 1.6066384315490723, + "learning_rate": 0.00013670962335186128, + "loss": 0.5692, + "step": 22997 + }, + { + "epoch": 1.5958642703490389, + "grad_norm": 1.5278245210647583, + "learning_rate": 0.00013670439695350246, + "loss": 0.701, + "step": 22998 + }, + { + "epoch": 1.5959336617861357, + "grad_norm": 1.5278563499450684, + "learning_rate": 0.00013669917043927207, + "loss": 0.7856, + "step": 22999 + }, + { + "epoch": 1.596003053223232, + "grad_norm": 1.7268699407577515, + "learning_rate": 0.00013669394380918666, + "loss": 0.7248, + "step": 23000 + }, + { + "epoch": 1.596072444660329, + "grad_norm": 1.752643346786499, + "learning_rate": 0.00013668871706326273, + "loss": 0.5533, + "step": 23001 + }, + { + "epoch": 1.5961418360974255, + "grad_norm": 2.0005156993865967, + "learning_rate": 0.0001366834902015168, + "loss": 0.6498, + "step": 23002 + }, + { + "epoch": 1.5962112275345222, + "grad_norm": 1.3127824068069458, + "learning_rate": 0.00013667826322396532, + "loss": 0.4578, + "step": 23003 + }, + { + "epoch": 1.596280618971619, + "grad_norm": 1.3043715953826904, + "learning_rate": 0.00013667303613062486, + "loss": 0.4464, + "step": 23004 + }, + { + "epoch": 1.5963500104087156, + "grad_norm": 1.6653941869735718, + "learning_rate": 0.00013666780892151182, + "loss": 0.4698, + "step": 23005 + }, + { + "epoch": 1.5964194018458122, + "grad_norm": 1.3644264936447144, + "learning_rate": 0.0001366625815966428, + "loss": 0.4831, + "step": 23006 + }, + { + "epoch": 1.596488793282909, + "grad_norm": 1.8125520944595337, + "learning_rate": 0.00013665735415603427, + "loss": 0.7783, + "step": 23007 + }, + { + "epoch": 1.5965581847200054, + "grad_norm": 1.9395912885665894, + "learning_rate": 0.0001366521265997027, + "loss": 0.5637, + "step": 23008 + }, + { + "epoch": 1.5966275761571023, + "grad_norm": 1.0902905464172363, + "learning_rate": 0.00013664689892766462, + "loss": 0.4716, + "step": 23009 + }, + { + "epoch": 1.5966969675941989, + "grad_norm": 1.9057239294052124, + "learning_rate": 0.00013664167113993652, + "loss": 0.6156, + "step": 23010 + }, + { + "epoch": 1.5967663590312955, + "grad_norm": 1.9159901142120361, + "learning_rate": 0.00013663644323653497, + "loss": 0.5556, + "step": 23011 + }, + { + "epoch": 1.5968357504683923, + "grad_norm": 1.2814034223556519, + "learning_rate": 0.00013663121521747637, + "loss": 0.4934, + "step": 23012 + }, + { + "epoch": 1.5969051419054887, + "grad_norm": 1.4375869035720825, + "learning_rate": 0.0001366259870827773, + "loss": 0.4787, + "step": 23013 + }, + { + "epoch": 1.5969745333425855, + "grad_norm": 1.5991781949996948, + "learning_rate": 0.00013662075883245423, + "loss": 0.6128, + "step": 23014 + }, + { + "epoch": 1.5970439247796822, + "grad_norm": 1.098879337310791, + "learning_rate": 0.00013661553046652369, + "loss": 0.449, + "step": 23015 + }, + { + "epoch": 1.5971133162167788, + "grad_norm": 1.307460904121399, + "learning_rate": 0.00013661030198500214, + "loss": 0.4315, + "step": 23016 + }, + { + "epoch": 1.5971827076538756, + "grad_norm": 2.791832685470581, + "learning_rate": 0.00013660507338790611, + "loss": 0.7983, + "step": 23017 + }, + { + "epoch": 1.5972520990909722, + "grad_norm": 1.9626723527908325, + "learning_rate": 0.00013659984467525216, + "loss": 0.7635, + "step": 23018 + }, + { + "epoch": 1.5973214905280688, + "grad_norm": 1.8457920551300049, + "learning_rate": 0.0001365946158470567, + "loss": 0.7274, + "step": 23019 + }, + { + "epoch": 1.5973908819651657, + "grad_norm": 1.9599361419677734, + "learning_rate": 0.0001365893869033363, + "loss": 0.5624, + "step": 23020 + }, + { + "epoch": 1.597460273402262, + "grad_norm": 1.3394426107406616, + "learning_rate": 0.00013658415784410745, + "loss": 0.4972, + "step": 23021 + }, + { + "epoch": 1.5975296648393589, + "grad_norm": 1.5497034788131714, + "learning_rate": 0.00013657892866938664, + "loss": 0.7045, + "step": 23022 + }, + { + "epoch": 1.5975990562764555, + "grad_norm": 2.069584608078003, + "learning_rate": 0.0001365736993791904, + "loss": 0.7084, + "step": 23023 + }, + { + "epoch": 1.597668447713552, + "grad_norm": 1.4533933401107788, + "learning_rate": 0.00013656846997353524, + "loss": 0.5008, + "step": 23024 + }, + { + "epoch": 1.597737839150649, + "grad_norm": 1.2552698850631714, + "learning_rate": 0.0001365632404524377, + "loss": 0.5488, + "step": 23025 + }, + { + "epoch": 1.5978072305877453, + "grad_norm": 1.6823416948318481, + "learning_rate": 0.0001365580108159142, + "loss": 0.5761, + "step": 23026 + }, + { + "epoch": 1.5978766220248422, + "grad_norm": 1.5290950536727905, + "learning_rate": 0.0001365527810639813, + "loss": 0.4889, + "step": 23027 + }, + { + "epoch": 1.5979460134619388, + "grad_norm": 2.0147459506988525, + "learning_rate": 0.00013654755119665556, + "loss": 0.9121, + "step": 23028 + }, + { + "epoch": 1.5980154048990354, + "grad_norm": 1.4009515047073364, + "learning_rate": 0.0001365423212139534, + "loss": 0.5553, + "step": 23029 + }, + { + "epoch": 1.5980847963361322, + "grad_norm": 1.4391978979110718, + "learning_rate": 0.00013653709111589136, + "loss": 0.5688, + "step": 23030 + }, + { + "epoch": 1.5981541877732288, + "grad_norm": 1.1825120449066162, + "learning_rate": 0.00013653186090248603, + "loss": 0.5976, + "step": 23031 + }, + { + "epoch": 1.5982235792103254, + "grad_norm": 1.2506128549575806, + "learning_rate": 0.0001365266305737538, + "loss": 0.7168, + "step": 23032 + }, + { + "epoch": 1.5982929706474223, + "grad_norm": 1.208385944366455, + "learning_rate": 0.00013652140012971125, + "loss": 0.53, + "step": 23033 + }, + { + "epoch": 1.5983623620845187, + "grad_norm": 1.556627631187439, + "learning_rate": 0.00013651616957037488, + "loss": 0.543, + "step": 23034 + }, + { + "epoch": 1.5984317535216155, + "grad_norm": 1.7004501819610596, + "learning_rate": 0.0001365109388957612, + "loss": 0.7713, + "step": 23035 + }, + { + "epoch": 1.598501144958712, + "grad_norm": 3.8526031970977783, + "learning_rate": 0.0001365057081058867, + "loss": 0.8468, + "step": 23036 + }, + { + "epoch": 1.5985705363958087, + "grad_norm": 1.6808491945266724, + "learning_rate": 0.00013650047720076793, + "loss": 0.5351, + "step": 23037 + }, + { + "epoch": 1.5986399278329055, + "grad_norm": 1.639513373374939, + "learning_rate": 0.00013649524618042139, + "loss": 0.7172, + "step": 23038 + }, + { + "epoch": 1.598709319270002, + "grad_norm": 1.1792513132095337, + "learning_rate": 0.0001364900150448636, + "loss": 0.5215, + "step": 23039 + }, + { + "epoch": 1.5987787107070988, + "grad_norm": 1.6683202981948853, + "learning_rate": 0.00013648478379411105, + "loss": 0.6431, + "step": 23040 + }, + { + "epoch": 1.5988481021441954, + "grad_norm": 1.1041765213012695, + "learning_rate": 0.00013647955242818027, + "loss": 0.4551, + "step": 23041 + }, + { + "epoch": 1.598917493581292, + "grad_norm": 1.8906948566436768, + "learning_rate": 0.0001364743209470878, + "loss": 0.8556, + "step": 23042 + }, + { + "epoch": 1.5989868850183888, + "grad_norm": 1.5374146699905396, + "learning_rate": 0.0001364690893508501, + "loss": 0.5482, + "step": 23043 + }, + { + "epoch": 1.5990562764554854, + "grad_norm": 2.1894302368164062, + "learning_rate": 0.0001364638576394837, + "loss": 0.6097, + "step": 23044 + }, + { + "epoch": 1.599125667892582, + "grad_norm": 1.5214085578918457, + "learning_rate": 0.0001364586258130052, + "loss": 0.6238, + "step": 23045 + }, + { + "epoch": 1.5991950593296789, + "grad_norm": 2.0040743350982666, + "learning_rate": 0.000136453393871431, + "loss": 0.6196, + "step": 23046 + }, + { + "epoch": 1.5992644507667753, + "grad_norm": 1.1162031888961792, + "learning_rate": 0.00013644816181477767, + "loss": 0.4446, + "step": 23047 + }, + { + "epoch": 1.599333842203872, + "grad_norm": 1.783729910850525, + "learning_rate": 0.00013644292964306174, + "loss": 0.6928, + "step": 23048 + }, + { + "epoch": 1.5994032336409687, + "grad_norm": 1.9448555707931519, + "learning_rate": 0.00013643769735629968, + "loss": 0.6185, + "step": 23049 + }, + { + "epoch": 1.5994726250780653, + "grad_norm": 1.4822680950164795, + "learning_rate": 0.00013643246495450805, + "loss": 0.7195, + "step": 23050 + }, + { + "epoch": 1.5995420165151621, + "grad_norm": 1.3063640594482422, + "learning_rate": 0.00013642723243770336, + "loss": 0.4656, + "step": 23051 + }, + { + "epoch": 1.5996114079522585, + "grad_norm": 1.8223286867141724, + "learning_rate": 0.00013642199980590215, + "loss": 0.6894, + "step": 23052 + }, + { + "epoch": 1.5996807993893554, + "grad_norm": 2.0236711502075195, + "learning_rate": 0.00013641676705912085, + "loss": 0.6764, + "step": 23053 + }, + { + "epoch": 1.599750190826452, + "grad_norm": 1.4907597303390503, + "learning_rate": 0.00013641153419737608, + "loss": 0.6566, + "step": 23054 + }, + { + "epoch": 1.5998195822635486, + "grad_norm": 1.8417813777923584, + "learning_rate": 0.0001364063012206843, + "loss": 0.6788, + "step": 23055 + }, + { + "epoch": 1.5998889737006454, + "grad_norm": 2.0337226390838623, + "learning_rate": 0.0001364010681290621, + "loss": 0.4599, + "step": 23056 + }, + { + "epoch": 1.599958365137742, + "grad_norm": 1.4147804975509644, + "learning_rate": 0.0001363958349225259, + "loss": 0.6571, + "step": 23057 + }, + { + "epoch": 1.6000277565748386, + "grad_norm": 2.7372279167175293, + "learning_rate": 0.0001363906016010923, + "loss": 0.6303, + "step": 23058 + }, + { + "epoch": 1.6000971480119355, + "grad_norm": 1.5517308712005615, + "learning_rate": 0.0001363853681647778, + "loss": 0.4689, + "step": 23059 + }, + { + "epoch": 1.6001665394490319, + "grad_norm": 1.4189863204956055, + "learning_rate": 0.00013638013461359885, + "loss": 0.5028, + "step": 23060 + }, + { + "epoch": 1.6002359308861287, + "grad_norm": 1.499143123626709, + "learning_rate": 0.0001363749009475721, + "loss": 0.5957, + "step": 23061 + }, + { + "epoch": 1.6003053223232253, + "grad_norm": 1.145057201385498, + "learning_rate": 0.00013636966716671397, + "loss": 0.3644, + "step": 23062 + }, + { + "epoch": 1.600374713760322, + "grad_norm": 1.4967466592788696, + "learning_rate": 0.00013636443327104107, + "loss": 0.6879, + "step": 23063 + }, + { + "epoch": 1.6004441051974188, + "grad_norm": 1.6696698665618896, + "learning_rate": 0.00013635919926056984, + "loss": 0.7178, + "step": 23064 + }, + { + "epoch": 1.6005134966345151, + "grad_norm": 1.8471870422363281, + "learning_rate": 0.00013635396513531683, + "loss": 0.6471, + "step": 23065 + }, + { + "epoch": 1.600582888071612, + "grad_norm": 1.566108226776123, + "learning_rate": 0.00013634873089529858, + "loss": 0.6147, + "step": 23066 + }, + { + "epoch": 1.6006522795087086, + "grad_norm": 1.1990681886672974, + "learning_rate": 0.0001363434965405316, + "loss": 0.5027, + "step": 23067 + }, + { + "epoch": 1.6007216709458052, + "grad_norm": 1.2468421459197998, + "learning_rate": 0.00013633826207103245, + "loss": 0.5575, + "step": 23068 + }, + { + "epoch": 1.600791062382902, + "grad_norm": 2.2951595783233643, + "learning_rate": 0.00013633302748681757, + "loss": 0.5015, + "step": 23069 + }, + { + "epoch": 1.6008604538199986, + "grad_norm": 1.7908985614776611, + "learning_rate": 0.0001363277927879036, + "loss": 0.8934, + "step": 23070 + }, + { + "epoch": 1.6009298452570953, + "grad_norm": 1.7564278841018677, + "learning_rate": 0.00013632255797430695, + "loss": 0.7478, + "step": 23071 + }, + { + "epoch": 1.600999236694192, + "grad_norm": 1.6072170734405518, + "learning_rate": 0.0001363173230460442, + "loss": 0.6811, + "step": 23072 + }, + { + "epoch": 1.6010686281312885, + "grad_norm": 1.3609893321990967, + "learning_rate": 0.00013631208800313191, + "loss": 0.5109, + "step": 23073 + }, + { + "epoch": 1.6011380195683853, + "grad_norm": 1.8218889236450195, + "learning_rate": 0.00013630685284558654, + "loss": 0.5735, + "step": 23074 + }, + { + "epoch": 1.601207411005482, + "grad_norm": 1.525052785873413, + "learning_rate": 0.00013630161757342466, + "loss": 0.665, + "step": 23075 + }, + { + "epoch": 1.6012768024425785, + "grad_norm": 1.6015528440475464, + "learning_rate": 0.00013629638218666277, + "loss": 0.6979, + "step": 23076 + }, + { + "epoch": 1.6013461938796754, + "grad_norm": 2.6032094955444336, + "learning_rate": 0.00013629114668531743, + "loss": 0.6517, + "step": 23077 + }, + { + "epoch": 1.6014155853167718, + "grad_norm": 1.5702927112579346, + "learning_rate": 0.00013628591106940515, + "loss": 0.644, + "step": 23078 + }, + { + "epoch": 1.6014849767538686, + "grad_norm": 1.8855117559432983, + "learning_rate": 0.00013628067533894245, + "loss": 0.5699, + "step": 23079 + }, + { + "epoch": 1.6015543681909652, + "grad_norm": 2.1992361545562744, + "learning_rate": 0.00013627543949394587, + "loss": 0.727, + "step": 23080 + }, + { + "epoch": 1.6016237596280618, + "grad_norm": 1.8706982135772705, + "learning_rate": 0.00013627020353443195, + "loss": 0.5583, + "step": 23081 + }, + { + "epoch": 1.6016931510651586, + "grad_norm": 2.2125205993652344, + "learning_rate": 0.00013626496746041718, + "loss": 0.7961, + "step": 23082 + }, + { + "epoch": 1.6017625425022552, + "grad_norm": 1.9701206684112549, + "learning_rate": 0.00013625973127191814, + "loss": 0.603, + "step": 23083 + }, + { + "epoch": 1.6018319339393519, + "grad_norm": 1.8123785257339478, + "learning_rate": 0.0001362544949689513, + "loss": 0.5209, + "step": 23084 + }, + { + "epoch": 1.6019013253764487, + "grad_norm": 2.5225589275360107, + "learning_rate": 0.00013624925855153328, + "loss": 0.6546, + "step": 23085 + }, + { + "epoch": 1.601970716813545, + "grad_norm": 2.7612709999084473, + "learning_rate": 0.0001362440220196805, + "loss": 0.7455, + "step": 23086 + }, + { + "epoch": 1.602040108250642, + "grad_norm": 1.9761559963226318, + "learning_rate": 0.00013623878537340962, + "loss": 0.4347, + "step": 23087 + }, + { + "epoch": 1.6021094996877385, + "grad_norm": 4.021450996398926, + "learning_rate": 0.00013623354861273703, + "loss": 0.7933, + "step": 23088 + }, + { + "epoch": 1.6021788911248351, + "grad_norm": 1.417052984237671, + "learning_rate": 0.00013622831173767936, + "loss": 0.4416, + "step": 23089 + }, + { + "epoch": 1.602248282561932, + "grad_norm": 1.6634395122528076, + "learning_rate": 0.00013622307474825312, + "loss": 0.5729, + "step": 23090 + }, + { + "epoch": 1.6023176739990284, + "grad_norm": 1.9224070310592651, + "learning_rate": 0.0001362178376444748, + "loss": 0.6396, + "step": 23091 + }, + { + "epoch": 1.6023870654361252, + "grad_norm": 1.1450430154800415, + "learning_rate": 0.000136212600426361, + "loss": 0.3755, + "step": 23092 + }, + { + "epoch": 1.6024564568732218, + "grad_norm": 1.418516993522644, + "learning_rate": 0.00013620736309392822, + "loss": 0.5224, + "step": 23093 + }, + { + "epoch": 1.6025258483103184, + "grad_norm": 3.70929217338562, + "learning_rate": 0.000136202125647193, + "loss": 0.9942, + "step": 23094 + }, + { + "epoch": 1.6025952397474152, + "grad_norm": 1.8143378496170044, + "learning_rate": 0.00013619688808617184, + "loss": 0.6012, + "step": 23095 + }, + { + "epoch": 1.6026646311845119, + "grad_norm": 1.95394766330719, + "learning_rate": 0.00013619165041088132, + "loss": 0.5477, + "step": 23096 + }, + { + "epoch": 1.6027340226216085, + "grad_norm": 1.646095633506775, + "learning_rate": 0.00013618641262133795, + "loss": 0.5657, + "step": 23097 + }, + { + "epoch": 1.6028034140587053, + "grad_norm": 1.8746521472930908, + "learning_rate": 0.0001361811747175583, + "loss": 0.5711, + "step": 23098 + }, + { + "epoch": 1.6028728054958017, + "grad_norm": 2.0764074325561523, + "learning_rate": 0.00013617593669955886, + "loss": 0.5235, + "step": 23099 + }, + { + "epoch": 1.6029421969328985, + "grad_norm": 1.8132133483886719, + "learning_rate": 0.0001361706985673562, + "loss": 0.5602, + "step": 23100 + }, + { + "epoch": 1.6030115883699951, + "grad_norm": 1.435089349746704, + "learning_rate": 0.00013616546032096683, + "loss": 0.5731, + "step": 23101 + }, + { + "epoch": 1.6030809798070917, + "grad_norm": 2.019339084625244, + "learning_rate": 0.00013616022196040727, + "loss": 0.8215, + "step": 23102 + }, + { + "epoch": 1.6031503712441886, + "grad_norm": 1.6560556888580322, + "learning_rate": 0.00013615498348569413, + "loss": 0.619, + "step": 23103 + }, + { + "epoch": 1.603219762681285, + "grad_norm": 2.3301265239715576, + "learning_rate": 0.0001361497448968439, + "loss": 0.7112, + "step": 23104 + }, + { + "epoch": 1.6032891541183818, + "grad_norm": 1.3446892499923706, + "learning_rate": 0.0001361445061938731, + "loss": 0.517, + "step": 23105 + }, + { + "epoch": 1.6033585455554784, + "grad_norm": 1.6043102741241455, + "learning_rate": 0.00013613926737679827, + "loss": 0.5319, + "step": 23106 + }, + { + "epoch": 1.603427936992575, + "grad_norm": 1.4775289297103882, + "learning_rate": 0.000136134028445636, + "loss": 0.3989, + "step": 23107 + }, + { + "epoch": 1.6034973284296719, + "grad_norm": 1.7696795463562012, + "learning_rate": 0.0001361287894004028, + "loss": 0.6142, + "step": 23108 + }, + { + "epoch": 1.6035667198667685, + "grad_norm": 1.4457508325576782, + "learning_rate": 0.00013612355024111516, + "loss": 0.5766, + "step": 23109 + }, + { + "epoch": 1.603636111303865, + "grad_norm": 2.0801162719726562, + "learning_rate": 0.0001361183109677897, + "loss": 0.5795, + "step": 23110 + }, + { + "epoch": 1.603705502740962, + "grad_norm": 2.1641392707824707, + "learning_rate": 0.00013611307158044292, + "loss": 0.6064, + "step": 23111 + }, + { + "epoch": 1.6037748941780583, + "grad_norm": 1.617098331451416, + "learning_rate": 0.00013610783207909134, + "loss": 0.6154, + "step": 23112 + }, + { + "epoch": 1.6038442856151551, + "grad_norm": 1.5883972644805908, + "learning_rate": 0.00013610259246375154, + "loss": 0.6955, + "step": 23113 + }, + { + "epoch": 1.6039136770522517, + "grad_norm": 2.5888662338256836, + "learning_rate": 0.00013609735273444003, + "loss": 0.6641, + "step": 23114 + }, + { + "epoch": 1.6039830684893484, + "grad_norm": 4.256608009338379, + "learning_rate": 0.0001360921128911734, + "loss": 0.5358, + "step": 23115 + }, + { + "epoch": 1.6040524599264452, + "grad_norm": 1.9134596586227417, + "learning_rate": 0.0001360868729339681, + "loss": 0.5686, + "step": 23116 + }, + { + "epoch": 1.6041218513635416, + "grad_norm": 1.4730700254440308, + "learning_rate": 0.00013608163286284077, + "loss": 0.5672, + "step": 23117 + }, + { + "epoch": 1.6041912428006384, + "grad_norm": 1.891469955444336, + "learning_rate": 0.00013607639267780792, + "loss": 0.6594, + "step": 23118 + }, + { + "epoch": 1.604260634237735, + "grad_norm": 1.6764041185379028, + "learning_rate": 0.00013607115237888604, + "loss": 0.6951, + "step": 23119 + }, + { + "epoch": 1.6043300256748316, + "grad_norm": 2.3669180870056152, + "learning_rate": 0.00013606591196609177, + "loss": 0.6906, + "step": 23120 + }, + { + "epoch": 1.6043994171119285, + "grad_norm": 1.7568801641464233, + "learning_rate": 0.00013606067143944158, + "loss": 0.5339, + "step": 23121 + }, + { + "epoch": 1.604468808549025, + "grad_norm": 1.1223379373550415, + "learning_rate": 0.00013605543079895204, + "loss": 0.4813, + "step": 23122 + }, + { + "epoch": 1.6045381999861217, + "grad_norm": 1.3607494831085205, + "learning_rate": 0.0001360501900446397, + "loss": 0.5162, + "step": 23123 + }, + { + "epoch": 1.6046075914232185, + "grad_norm": 1.9300682544708252, + "learning_rate": 0.00013604494917652103, + "loss": 0.7093, + "step": 23124 + }, + { + "epoch": 1.604676982860315, + "grad_norm": 2.138164520263672, + "learning_rate": 0.00013603970819461272, + "loss": 0.509, + "step": 23125 + }, + { + "epoch": 1.6047463742974117, + "grad_norm": 2.192197799682617, + "learning_rate": 0.00013603446709893117, + "loss": 0.7079, + "step": 23126 + }, + { + "epoch": 1.6048157657345083, + "grad_norm": 1.8972848653793335, + "learning_rate": 0.000136029225889493, + "loss": 0.5949, + "step": 23127 + }, + { + "epoch": 1.604885157171605, + "grad_norm": 1.6053837537765503, + "learning_rate": 0.00013602398456631477, + "loss": 0.5124, + "step": 23128 + }, + { + "epoch": 1.6049545486087018, + "grad_norm": 1.7146087884902954, + "learning_rate": 0.000136018743129413, + "loss": 0.7192, + "step": 23129 + }, + { + "epoch": 1.6050239400457984, + "grad_norm": 1.3032492399215698, + "learning_rate": 0.0001360135015788042, + "loss": 0.5346, + "step": 23130 + }, + { + "epoch": 1.605093331482895, + "grad_norm": 1.506927728652954, + "learning_rate": 0.00013600825991450497, + "loss": 0.6973, + "step": 23131 + }, + { + "epoch": 1.6051627229199916, + "grad_norm": 1.934618353843689, + "learning_rate": 0.00013600301813653185, + "loss": 0.6409, + "step": 23132 + }, + { + "epoch": 1.6052321143570882, + "grad_norm": 1.4051834344863892, + "learning_rate": 0.00013599777624490134, + "loss": 0.4525, + "step": 23133 + }, + { + "epoch": 1.605301505794185, + "grad_norm": 2.159296751022339, + "learning_rate": 0.00013599253423963008, + "loss": 0.8669, + "step": 23134 + }, + { + "epoch": 1.6053708972312817, + "grad_norm": 1.532110333442688, + "learning_rate": 0.00013598729212073457, + "loss": 0.6516, + "step": 23135 + }, + { + "epoch": 1.6054402886683783, + "grad_norm": 1.5764901638031006, + "learning_rate": 0.0001359820498882313, + "loss": 0.5861, + "step": 23136 + }, + { + "epoch": 1.6055096801054751, + "grad_norm": 1.9200626611709595, + "learning_rate": 0.00013597680754213692, + "loss": 0.709, + "step": 23137 + }, + { + "epoch": 1.6055790715425715, + "grad_norm": 2.6239264011383057, + "learning_rate": 0.0001359715650824679, + "loss": 0.7858, + "step": 23138 + }, + { + "epoch": 1.6056484629796683, + "grad_norm": 2.0355870723724365, + "learning_rate": 0.00013596632250924086, + "loss": 0.6596, + "step": 23139 + }, + { + "epoch": 1.605717854416765, + "grad_norm": 1.9275619983673096, + "learning_rate": 0.0001359610798224723, + "loss": 0.7156, + "step": 23140 + }, + { + "epoch": 1.6057872458538616, + "grad_norm": 1.5829594135284424, + "learning_rate": 0.00013595583702217875, + "loss": 0.5299, + "step": 23141 + }, + { + "epoch": 1.6058566372909584, + "grad_norm": 1.3262346982955933, + "learning_rate": 0.00013595059410837686, + "loss": 0.4967, + "step": 23142 + }, + { + "epoch": 1.605926028728055, + "grad_norm": 1.5011457204818726, + "learning_rate": 0.00013594535108108307, + "loss": 0.5489, + "step": 23143 + }, + { + "epoch": 1.6059954201651516, + "grad_norm": 1.1673680543899536, + "learning_rate": 0.00013594010794031394, + "loss": 0.5603, + "step": 23144 + }, + { + "epoch": 1.6060648116022482, + "grad_norm": 2.0134849548339844, + "learning_rate": 0.00013593486468608612, + "loss": 0.7694, + "step": 23145 + }, + { + "epoch": 1.6061342030393448, + "grad_norm": 1.5287209749221802, + "learning_rate": 0.0001359296213184161, + "loss": 0.4951, + "step": 23146 + }, + { + "epoch": 1.6062035944764417, + "grad_norm": 1.4251163005828857, + "learning_rate": 0.0001359243778373204, + "loss": 0.5189, + "step": 23147 + }, + { + "epoch": 1.6062729859135383, + "grad_norm": 1.4427739381790161, + "learning_rate": 0.00013591913424281562, + "loss": 0.4046, + "step": 23148 + }, + { + "epoch": 1.606342377350635, + "grad_norm": 1.8196914196014404, + "learning_rate": 0.0001359138905349183, + "loss": 0.7884, + "step": 23149 + }, + { + "epoch": 1.6064117687877317, + "grad_norm": 1.5840003490447998, + "learning_rate": 0.000135908646713645, + "loss": 0.4882, + "step": 23150 + }, + { + "epoch": 1.6064811602248281, + "grad_norm": 1.2658944129943848, + "learning_rate": 0.00013590340277901225, + "loss": 0.5316, + "step": 23151 + }, + { + "epoch": 1.606550551661925, + "grad_norm": 1.6331067085266113, + "learning_rate": 0.00013589815873103663, + "loss": 0.7327, + "step": 23152 + }, + { + "epoch": 1.6066199430990216, + "grad_norm": 1.20637047290802, + "learning_rate": 0.00013589291456973473, + "loss": 0.4941, + "step": 23153 + }, + { + "epoch": 1.6066893345361182, + "grad_norm": 2.0682213306427, + "learning_rate": 0.00013588767029512302, + "loss": 0.6956, + "step": 23154 + }, + { + "epoch": 1.606758725973215, + "grad_norm": 1.8472496271133423, + "learning_rate": 0.00013588242590721811, + "loss": 0.7608, + "step": 23155 + }, + { + "epoch": 1.6068281174103116, + "grad_norm": 1.4964102506637573, + "learning_rate": 0.00013587718140603653, + "loss": 0.5797, + "step": 23156 + }, + { + "epoch": 1.6068975088474082, + "grad_norm": 1.221511960029602, + "learning_rate": 0.0001358719367915949, + "loss": 0.4649, + "step": 23157 + }, + { + "epoch": 1.6069669002845048, + "grad_norm": 2.2010955810546875, + "learning_rate": 0.00013586669206390968, + "loss": 0.7242, + "step": 23158 + }, + { + "epoch": 1.6070362917216015, + "grad_norm": 1.3640828132629395, + "learning_rate": 0.0001358614472229975, + "loss": 0.5657, + "step": 23159 + }, + { + "epoch": 1.6071056831586983, + "grad_norm": 1.7253416776657104, + "learning_rate": 0.0001358562022688749, + "loss": 0.6359, + "step": 23160 + }, + { + "epoch": 1.607175074595795, + "grad_norm": 1.176404356956482, + "learning_rate": 0.00013585095720155838, + "loss": 0.4296, + "step": 23161 + }, + { + "epoch": 1.6072444660328915, + "grad_norm": 1.6436951160430908, + "learning_rate": 0.00013584571202106456, + "loss": 0.751, + "step": 23162 + }, + { + "epoch": 1.6073138574699883, + "grad_norm": 1.7020294666290283, + "learning_rate": 0.00013584046672741003, + "loss": 0.6231, + "step": 23163 + }, + { + "epoch": 1.6073832489070847, + "grad_norm": 1.5581508874893188, + "learning_rate": 0.00013583522132061125, + "loss": 0.6147, + "step": 23164 + }, + { + "epoch": 1.6074526403441816, + "grad_norm": 1.569456696510315, + "learning_rate": 0.00013582997580068487, + "loss": 0.8135, + "step": 23165 + }, + { + "epoch": 1.6075220317812782, + "grad_norm": 1.523551106452942, + "learning_rate": 0.00013582473016764742, + "loss": 0.5658, + "step": 23166 + }, + { + "epoch": 1.6075914232183748, + "grad_norm": 1.9932206869125366, + "learning_rate": 0.00013581948442151543, + "loss": 0.6937, + "step": 23167 + }, + { + "epoch": 1.6076608146554716, + "grad_norm": 2.2070982456207275, + "learning_rate": 0.0001358142385623055, + "loss": 0.8617, + "step": 23168 + }, + { + "epoch": 1.6077302060925682, + "grad_norm": 1.7607543468475342, + "learning_rate": 0.00013580899259003413, + "loss": 0.6135, + "step": 23169 + }, + { + "epoch": 1.6077995975296648, + "grad_norm": 1.524053692817688, + "learning_rate": 0.00013580374650471797, + "loss": 0.5916, + "step": 23170 + }, + { + "epoch": 1.6078689889667614, + "grad_norm": 1.5362591743469238, + "learning_rate": 0.00013579850030637352, + "loss": 0.6213, + "step": 23171 + }, + { + "epoch": 1.607938380403858, + "grad_norm": 1.4606815576553345, + "learning_rate": 0.0001357932539950174, + "loss": 0.5937, + "step": 23172 + }, + { + "epoch": 1.608007771840955, + "grad_norm": 1.695791482925415, + "learning_rate": 0.00013578800757066607, + "loss": 0.5452, + "step": 23173 + }, + { + "epoch": 1.6080771632780515, + "grad_norm": 1.4919780492782593, + "learning_rate": 0.00013578276103333618, + "loss": 0.637, + "step": 23174 + }, + { + "epoch": 1.6081465547151481, + "grad_norm": 1.5284675359725952, + "learning_rate": 0.00013577751438304425, + "loss": 0.7126, + "step": 23175 + }, + { + "epoch": 1.608215946152245, + "grad_norm": 1.3578063249588013, + "learning_rate": 0.00013577226761980686, + "loss": 0.3814, + "step": 23176 + }, + { + "epoch": 1.6082853375893413, + "grad_norm": 1.2154165506362915, + "learning_rate": 0.0001357670207436406, + "loss": 0.5756, + "step": 23177 + }, + { + "epoch": 1.6083547290264382, + "grad_norm": 1.693916916847229, + "learning_rate": 0.00013576177375456197, + "loss": 0.5344, + "step": 23178 + }, + { + "epoch": 1.6084241204635348, + "grad_norm": 1.6063992977142334, + "learning_rate": 0.00013575652665258757, + "loss": 0.6299, + "step": 23179 + }, + { + "epoch": 1.6084935119006314, + "grad_norm": 1.6691902875900269, + "learning_rate": 0.00013575127943773403, + "loss": 0.5199, + "step": 23180 + }, + { + "epoch": 1.6085629033377282, + "grad_norm": 1.7318706512451172, + "learning_rate": 0.00013574603211001778, + "loss": 0.7826, + "step": 23181 + }, + { + "epoch": 1.6086322947748248, + "grad_norm": 1.731642723083496, + "learning_rate": 0.00013574078466945547, + "loss": 0.7365, + "step": 23182 + }, + { + "epoch": 1.6087016862119214, + "grad_norm": 1.9589859247207642, + "learning_rate": 0.00013573553711606365, + "loss": 0.7648, + "step": 23183 + }, + { + "epoch": 1.6087710776490183, + "grad_norm": 2.5722715854644775, + "learning_rate": 0.0001357302894498589, + "loss": 0.8331, + "step": 23184 + }, + { + "epoch": 1.6088404690861147, + "grad_norm": 1.8196147680282593, + "learning_rate": 0.00013572504167085774, + "loss": 0.6056, + "step": 23185 + }, + { + "epoch": 1.6089098605232115, + "grad_norm": 1.4956634044647217, + "learning_rate": 0.00013571979377907678, + "loss": 0.503, + "step": 23186 + }, + { + "epoch": 1.608979251960308, + "grad_norm": 2.574075937271118, + "learning_rate": 0.0001357145457745326, + "loss": 0.5781, + "step": 23187 + }, + { + "epoch": 1.6090486433974047, + "grad_norm": 1.4613783359527588, + "learning_rate": 0.0001357092976572417, + "loss": 0.6959, + "step": 23188 + }, + { + "epoch": 1.6091180348345016, + "grad_norm": 1.7890625, + "learning_rate": 0.00013570404942722073, + "loss": 0.7086, + "step": 23189 + }, + { + "epoch": 1.609187426271598, + "grad_norm": 2.0883779525756836, + "learning_rate": 0.0001356988010844862, + "loss": 0.6748, + "step": 23190 + }, + { + "epoch": 1.6092568177086948, + "grad_norm": 2.127380609512329, + "learning_rate": 0.00013569355262905473, + "loss": 0.5652, + "step": 23191 + }, + { + "epoch": 1.6093262091457914, + "grad_norm": 1.8123445510864258, + "learning_rate": 0.0001356883040609428, + "loss": 0.7282, + "step": 23192 + }, + { + "epoch": 1.609395600582888, + "grad_norm": 1.3694725036621094, + "learning_rate": 0.00013568305538016705, + "loss": 0.5909, + "step": 23193 + }, + { + "epoch": 1.6094649920199848, + "grad_norm": 1.9717676639556885, + "learning_rate": 0.00013567780658674407, + "loss": 0.8953, + "step": 23194 + }, + { + "epoch": 1.6095343834570814, + "grad_norm": 2.629986524581909, + "learning_rate": 0.00013567255768069035, + "loss": 0.8864, + "step": 23195 + }, + { + "epoch": 1.609603774894178, + "grad_norm": 2.6024417877197266, + "learning_rate": 0.00013566730866202252, + "loss": 0.6909, + "step": 23196 + }, + { + "epoch": 1.6096731663312749, + "grad_norm": 1.3272978067398071, + "learning_rate": 0.00013566205953075716, + "loss": 0.4798, + "step": 23197 + }, + { + "epoch": 1.6097425577683713, + "grad_norm": 1.6660164594650269, + "learning_rate": 0.0001356568102869108, + "loss": 0.5279, + "step": 23198 + }, + { + "epoch": 1.609811949205468, + "grad_norm": 1.0993760824203491, + "learning_rate": 0.0001356515609305, + "loss": 0.4139, + "step": 23199 + }, + { + "epoch": 1.6098813406425647, + "grad_norm": 1.972936987876892, + "learning_rate": 0.00013564631146154135, + "loss": 0.7635, + "step": 23200 + }, + { + "epoch": 1.6099507320796613, + "grad_norm": 1.5943368673324585, + "learning_rate": 0.00013564106188005147, + "loss": 0.433, + "step": 23201 + }, + { + "epoch": 1.6100201235167582, + "grad_norm": 2.029844284057617, + "learning_rate": 0.0001356358121860469, + "loss": 0.6636, + "step": 23202 + }, + { + "epoch": 1.6100895149538546, + "grad_norm": 1.5694142580032349, + "learning_rate": 0.00013563056237954417, + "loss": 0.7268, + "step": 23203 + }, + { + "epoch": 1.6101589063909514, + "grad_norm": 2.435725450515747, + "learning_rate": 0.00013562531246055992, + "loss": 0.726, + "step": 23204 + }, + { + "epoch": 1.610228297828048, + "grad_norm": 1.3375427722930908, + "learning_rate": 0.00013562006242911066, + "loss": 0.5433, + "step": 23205 + }, + { + "epoch": 1.6102976892651446, + "grad_norm": 1.649266242980957, + "learning_rate": 0.00013561481228521302, + "loss": 0.5058, + "step": 23206 + }, + { + "epoch": 1.6103670807022414, + "grad_norm": 2.0529377460479736, + "learning_rate": 0.00013560956202888355, + "loss": 0.7771, + "step": 23207 + }, + { + "epoch": 1.610436472139338, + "grad_norm": 2.1642818450927734, + "learning_rate": 0.00013560431166013883, + "loss": 0.5405, + "step": 23208 + }, + { + "epoch": 1.6105058635764347, + "grad_norm": 1.6220773458480835, + "learning_rate": 0.0001355990611789954, + "loss": 0.5496, + "step": 23209 + }, + { + "epoch": 1.6105752550135315, + "grad_norm": 1.552699089050293, + "learning_rate": 0.00013559381058546988, + "loss": 0.4781, + "step": 23210 + }, + { + "epoch": 1.6106446464506279, + "grad_norm": 1.277591347694397, + "learning_rate": 0.00013558855987957886, + "loss": 0.5101, + "step": 23211 + }, + { + "epoch": 1.6107140378877247, + "grad_norm": 1.4558203220367432, + "learning_rate": 0.00013558330906133885, + "loss": 0.6459, + "step": 23212 + }, + { + "epoch": 1.6107834293248213, + "grad_norm": 1.7641627788543701, + "learning_rate": 0.00013557805813076644, + "loss": 0.7237, + "step": 23213 + }, + { + "epoch": 1.610852820761918, + "grad_norm": 1.9989320039749146, + "learning_rate": 0.00013557280708787826, + "loss": 0.6129, + "step": 23214 + }, + { + "epoch": 1.6109222121990148, + "grad_norm": 1.7840783596038818, + "learning_rate": 0.00013556755593269087, + "loss": 0.5245, + "step": 23215 + }, + { + "epoch": 1.6109916036361112, + "grad_norm": 1.2934105396270752, + "learning_rate": 0.0001355623046652208, + "loss": 0.3855, + "step": 23216 + }, + { + "epoch": 1.611060995073208, + "grad_norm": 1.4074146747589111, + "learning_rate": 0.00013555705328548466, + "loss": 0.4372, + "step": 23217 + }, + { + "epoch": 1.6111303865103046, + "grad_norm": 1.7316043376922607, + "learning_rate": 0.00013555180179349906, + "loss": 0.7259, + "step": 23218 + }, + { + "epoch": 1.6111997779474012, + "grad_norm": 2.665864944458008, + "learning_rate": 0.00013554655018928053, + "loss": 0.6241, + "step": 23219 + }, + { + "epoch": 1.611269169384498, + "grad_norm": 2.098806142807007, + "learning_rate": 0.00013554129847284565, + "loss": 0.8015, + "step": 23220 + }, + { + "epoch": 1.6113385608215947, + "grad_norm": 2.2071738243103027, + "learning_rate": 0.00013553604664421102, + "loss": 0.6537, + "step": 23221 + }, + { + "epoch": 1.6114079522586913, + "grad_norm": 2.0368826389312744, + "learning_rate": 0.00013553079470339323, + "loss": 0.6729, + "step": 23222 + }, + { + "epoch": 1.611477343695788, + "grad_norm": 1.7977396249771118, + "learning_rate": 0.00013552554265040882, + "loss": 0.6224, + "step": 23223 + }, + { + "epoch": 1.6115467351328845, + "grad_norm": 2.5949490070343018, + "learning_rate": 0.0001355202904852744, + "loss": 0.7688, + "step": 23224 + }, + { + "epoch": 1.6116161265699813, + "grad_norm": 2.4984169006347656, + "learning_rate": 0.00013551503820800656, + "loss": 0.8255, + "step": 23225 + }, + { + "epoch": 1.611685518007078, + "grad_norm": 1.7856537103652954, + "learning_rate": 0.0001355097858186219, + "loss": 0.72, + "step": 23226 + }, + { + "epoch": 1.6117549094441745, + "grad_norm": 2.7375850677490234, + "learning_rate": 0.0001355045333171369, + "loss": 0.6805, + "step": 23227 + }, + { + "epoch": 1.6118243008812714, + "grad_norm": 1.811913251876831, + "learning_rate": 0.00013549928070356824, + "loss": 0.6358, + "step": 23228 + }, + { + "epoch": 1.6118936923183678, + "grad_norm": 1.8236184120178223, + "learning_rate": 0.00013549402797793247, + "loss": 0.6558, + "step": 23229 + }, + { + "epoch": 1.6119630837554646, + "grad_norm": 1.569371223449707, + "learning_rate": 0.00013548877514024614, + "loss": 0.671, + "step": 23230 + }, + { + "epoch": 1.6120324751925612, + "grad_norm": 1.7531166076660156, + "learning_rate": 0.0001354835221905259, + "loss": 0.5257, + "step": 23231 + }, + { + "epoch": 1.6121018666296578, + "grad_norm": 1.6838876008987427, + "learning_rate": 0.00013547826912878826, + "loss": 0.7754, + "step": 23232 + }, + { + "epoch": 1.6121712580667547, + "grad_norm": 2.034984588623047, + "learning_rate": 0.00013547301595504986, + "loss": 0.445, + "step": 23233 + }, + { + "epoch": 1.6122406495038513, + "grad_norm": 1.7922550439834595, + "learning_rate": 0.00013546776266932728, + "loss": 0.6313, + "step": 23234 + }, + { + "epoch": 1.6123100409409479, + "grad_norm": 1.7589147090911865, + "learning_rate": 0.00013546250927163708, + "loss": 0.5603, + "step": 23235 + }, + { + "epoch": 1.6123794323780447, + "grad_norm": 1.6014819145202637, + "learning_rate": 0.00013545725576199583, + "loss": 0.52, + "step": 23236 + }, + { + "epoch": 1.612448823815141, + "grad_norm": 1.6287134885787964, + "learning_rate": 0.00013545200214042018, + "loss": 0.4564, + "step": 23237 + }, + { + "epoch": 1.612518215252238, + "grad_norm": 1.593997597694397, + "learning_rate": 0.00013544674840692663, + "loss": 0.5123, + "step": 23238 + }, + { + "epoch": 1.6125876066893345, + "grad_norm": 2.015026569366455, + "learning_rate": 0.00013544149456153184, + "loss": 0.5894, + "step": 23239 + }, + { + "epoch": 1.6126569981264312, + "grad_norm": 2.9850504398345947, + "learning_rate": 0.00013543624060425235, + "loss": 0.6617, + "step": 23240 + }, + { + "epoch": 1.612726389563528, + "grad_norm": 2.7741312980651855, + "learning_rate": 0.00013543098653510474, + "loss": 0.9759, + "step": 23241 + }, + { + "epoch": 1.6127957810006244, + "grad_norm": 1.181217074394226, + "learning_rate": 0.00013542573235410566, + "loss": 0.3572, + "step": 23242 + }, + { + "epoch": 1.6128651724377212, + "grad_norm": 1.26838219165802, + "learning_rate": 0.00013542047806127161, + "loss": 0.4661, + "step": 23243 + }, + { + "epoch": 1.6129345638748178, + "grad_norm": 1.679858922958374, + "learning_rate": 0.00013541522365661927, + "loss": 0.8408, + "step": 23244 + }, + { + "epoch": 1.6130039553119144, + "grad_norm": 2.155668258666992, + "learning_rate": 0.00013540996914016514, + "loss": 0.6953, + "step": 23245 + }, + { + "epoch": 1.6130733467490113, + "grad_norm": 1.9482940435409546, + "learning_rate": 0.00013540471451192586, + "loss": 0.6314, + "step": 23246 + }, + { + "epoch": 1.6131427381861079, + "grad_norm": 8.128630638122559, + "learning_rate": 0.000135399459771918, + "loss": 0.5563, + "step": 23247 + }, + { + "epoch": 1.6132121296232045, + "grad_norm": 2.1923794746398926, + "learning_rate": 0.00013539420492015812, + "loss": 0.7031, + "step": 23248 + }, + { + "epoch": 1.6132815210603013, + "grad_norm": 2.0807926654815674, + "learning_rate": 0.0001353889499566629, + "loss": 0.6559, + "step": 23249 + }, + { + "epoch": 1.6133509124973977, + "grad_norm": 2.2505264282226562, + "learning_rate": 0.00013538369488144884, + "loss": 0.6464, + "step": 23250 + }, + { + "epoch": 1.6134203039344945, + "grad_norm": 1.7061607837677002, + "learning_rate": 0.00013537843969453256, + "loss": 0.7088, + "step": 23251 + }, + { + "epoch": 1.6134896953715911, + "grad_norm": 1.5981796979904175, + "learning_rate": 0.00013537318439593062, + "loss": 0.5215, + "step": 23252 + }, + { + "epoch": 1.6135590868086878, + "grad_norm": 2.3678829669952393, + "learning_rate": 0.0001353679289856597, + "loss": 0.6771, + "step": 23253 + }, + { + "epoch": 1.6136284782457846, + "grad_norm": 2.1656417846679688, + "learning_rate": 0.0001353626734637363, + "loss": 0.6935, + "step": 23254 + }, + { + "epoch": 1.613697869682881, + "grad_norm": 3.3288686275482178, + "learning_rate": 0.00013535741783017705, + "loss": 0.687, + "step": 23255 + }, + { + "epoch": 1.6137672611199778, + "grad_norm": 3.523667335510254, + "learning_rate": 0.00013535216208499854, + "loss": 0.8073, + "step": 23256 + }, + { + "epoch": 1.6138366525570744, + "grad_norm": 3.2953011989593506, + "learning_rate": 0.00013534690622821736, + "loss": 0.4821, + "step": 23257 + }, + { + "epoch": 1.613906043994171, + "grad_norm": 1.8404396772384644, + "learning_rate": 0.00013534165025985007, + "loss": 0.5861, + "step": 23258 + }, + { + "epoch": 1.6139754354312679, + "grad_norm": 1.3803735971450806, + "learning_rate": 0.00013533639417991333, + "loss": 0.4566, + "step": 23259 + }, + { + "epoch": 1.6140448268683645, + "grad_norm": 2.689934492111206, + "learning_rate": 0.00013533113798842365, + "loss": 0.4929, + "step": 23260 + }, + { + "epoch": 1.614114218305461, + "grad_norm": 1.9924451112747192, + "learning_rate": 0.0001353258816853977, + "loss": 0.8854, + "step": 23261 + }, + { + "epoch": 1.614183609742558, + "grad_norm": 3.4283363819122314, + "learning_rate": 0.00013532062527085202, + "loss": 0.5746, + "step": 23262 + }, + { + "epoch": 1.6142530011796543, + "grad_norm": 1.5601955652236938, + "learning_rate": 0.00013531536874480327, + "loss": 0.5181, + "step": 23263 + }, + { + "epoch": 1.6143223926167511, + "grad_norm": 2.316925287246704, + "learning_rate": 0.00013531011210726794, + "loss": 0.7009, + "step": 23264 + }, + { + "epoch": 1.6143917840538478, + "grad_norm": 1.8170957565307617, + "learning_rate": 0.0001353048553582627, + "loss": 0.4825, + "step": 23265 + }, + { + "epoch": 1.6144611754909444, + "grad_norm": 1.629549503326416, + "learning_rate": 0.00013529959849780413, + "loss": 0.5046, + "step": 23266 + }, + { + "epoch": 1.6145305669280412, + "grad_norm": 1.618106484413147, + "learning_rate": 0.00013529434152590884, + "loss": 0.794, + "step": 23267 + }, + { + "epoch": 1.6145999583651376, + "grad_norm": 2.7941362857818604, + "learning_rate": 0.00013528908444259335, + "loss": 0.7302, + "step": 23268 + }, + { + "epoch": 1.6146693498022344, + "grad_norm": 2.2537825107574463, + "learning_rate": 0.00013528382724787435, + "loss": 0.5528, + "step": 23269 + }, + { + "epoch": 1.614738741239331, + "grad_norm": 1.396978735923767, + "learning_rate": 0.00013527856994176844, + "loss": 0.416, + "step": 23270 + }, + { + "epoch": 1.6148081326764276, + "grad_norm": 1.5155121088027954, + "learning_rate": 0.00013527331252429213, + "loss": 0.5741, + "step": 23271 + }, + { + "epoch": 1.6148775241135245, + "grad_norm": 2.791170597076416, + "learning_rate": 0.0001352680549954621, + "loss": 0.6372, + "step": 23272 + }, + { + "epoch": 1.614946915550621, + "grad_norm": 1.598612904548645, + "learning_rate": 0.0001352627973552949, + "loss": 0.452, + "step": 23273 + }, + { + "epoch": 1.6150163069877177, + "grad_norm": 1.900024652481079, + "learning_rate": 0.00013525753960380712, + "loss": 0.6306, + "step": 23274 + }, + { + "epoch": 1.6150856984248145, + "grad_norm": 2.307786464691162, + "learning_rate": 0.00013525228174101538, + "loss": 0.7196, + "step": 23275 + }, + { + "epoch": 1.615155089861911, + "grad_norm": 1.6379696130752563, + "learning_rate": 0.00013524702376693627, + "loss": 0.5854, + "step": 23276 + }, + { + "epoch": 1.6152244812990078, + "grad_norm": 1.5923022031784058, + "learning_rate": 0.00013524176568158644, + "loss": 0.4903, + "step": 23277 + }, + { + "epoch": 1.6152938727361044, + "grad_norm": 1.9261082410812378, + "learning_rate": 0.0001352365074849824, + "loss": 0.7251, + "step": 23278 + }, + { + "epoch": 1.615363264173201, + "grad_norm": 1.4099702835083008, + "learning_rate": 0.00013523124917714084, + "loss": 0.5634, + "step": 23279 + }, + { + "epoch": 1.6154326556102978, + "grad_norm": 1.329439640045166, + "learning_rate": 0.00013522599075807827, + "loss": 0.4798, + "step": 23280 + }, + { + "epoch": 1.6155020470473942, + "grad_norm": 1.866212248802185, + "learning_rate": 0.00013522073222781137, + "loss": 0.506, + "step": 23281 + }, + { + "epoch": 1.615571438484491, + "grad_norm": 1.9615837335586548, + "learning_rate": 0.00013521547358635668, + "loss": 0.6757, + "step": 23282 + }, + { + "epoch": 1.6156408299215876, + "grad_norm": 1.942999005317688, + "learning_rate": 0.0001352102148337308, + "loss": 0.4813, + "step": 23283 + }, + { + "epoch": 1.6157102213586843, + "grad_norm": 1.5745280981063843, + "learning_rate": 0.0001352049559699504, + "loss": 0.5369, + "step": 23284 + }, + { + "epoch": 1.615779612795781, + "grad_norm": 1.5922703742980957, + "learning_rate": 0.00013519969699503202, + "loss": 0.6097, + "step": 23285 + }, + { + "epoch": 1.6158490042328777, + "grad_norm": 1.3648765087127686, + "learning_rate": 0.0001351944379089923, + "loss": 0.438, + "step": 23286 + }, + { + "epoch": 1.6159183956699743, + "grad_norm": 2.2104361057281494, + "learning_rate": 0.0001351891787118478, + "loss": 0.661, + "step": 23287 + }, + { + "epoch": 1.6159877871070711, + "grad_norm": 2.3613767623901367, + "learning_rate": 0.00013518391940361515, + "loss": 0.6143, + "step": 23288 + }, + { + "epoch": 1.6160571785441675, + "grad_norm": 1.8182339668273926, + "learning_rate": 0.00013517865998431095, + "loss": 0.6413, + "step": 23289 + }, + { + "epoch": 1.6161265699812644, + "grad_norm": 1.4528181552886963, + "learning_rate": 0.00013517340045395178, + "loss": 0.4012, + "step": 23290 + }, + { + "epoch": 1.616195961418361, + "grad_norm": 1.6671921014785767, + "learning_rate": 0.00013516814081255432, + "loss": 0.4372, + "step": 23291 + }, + { + "epoch": 1.6162653528554576, + "grad_norm": 1.8140606880187988, + "learning_rate": 0.0001351628810601351, + "loss": 0.4995, + "step": 23292 + }, + { + "epoch": 1.6163347442925544, + "grad_norm": 2.083742618560791, + "learning_rate": 0.00013515762119671072, + "loss": 0.77, + "step": 23293 + }, + { + "epoch": 1.6164041357296508, + "grad_norm": 1.4659556150436401, + "learning_rate": 0.00013515236122229782, + "loss": 0.5672, + "step": 23294 + }, + { + "epoch": 1.6164735271667476, + "grad_norm": 2.2282583713531494, + "learning_rate": 0.00013514710113691298, + "loss": 0.6535, + "step": 23295 + }, + { + "epoch": 1.6165429186038442, + "grad_norm": 1.475689172744751, + "learning_rate": 0.00013514184094057287, + "loss": 0.4914, + "step": 23296 + }, + { + "epoch": 1.6166123100409409, + "grad_norm": 2.0105321407318115, + "learning_rate": 0.000135136580633294, + "loss": 0.5392, + "step": 23297 + }, + { + "epoch": 1.6166817014780377, + "grad_norm": 2.1668200492858887, + "learning_rate": 0.00013513132021509305, + "loss": 0.7446, + "step": 23298 + }, + { + "epoch": 1.6167510929151343, + "grad_norm": 1.6727222204208374, + "learning_rate": 0.00013512605968598657, + "loss": 0.4455, + "step": 23299 + }, + { + "epoch": 1.616820484352231, + "grad_norm": 1.7619221210479736, + "learning_rate": 0.0001351207990459912, + "loss": 0.5271, + "step": 23300 + }, + { + "epoch": 1.6168898757893277, + "grad_norm": 2.7315971851348877, + "learning_rate": 0.00013511553829512356, + "loss": 0.845, + "step": 23301 + }, + { + "epoch": 1.6169592672264241, + "grad_norm": 1.2524892091751099, + "learning_rate": 0.00013511027743340023, + "loss": 0.5301, + "step": 23302 + }, + { + "epoch": 1.617028658663521, + "grad_norm": 1.5824840068817139, + "learning_rate": 0.0001351050164608378, + "loss": 0.7462, + "step": 23303 + }, + { + "epoch": 1.6170980501006176, + "grad_norm": 1.6430747509002686, + "learning_rate": 0.00013509975537745293, + "loss": 0.6895, + "step": 23304 + }, + { + "epoch": 1.6171674415377142, + "grad_norm": 1.540815830230713, + "learning_rate": 0.00013509449418326223, + "loss": 0.5993, + "step": 23305 + }, + { + "epoch": 1.617236832974811, + "grad_norm": 2.1624467372894287, + "learning_rate": 0.00013508923287828224, + "loss": 0.8479, + "step": 23306 + }, + { + "epoch": 1.6173062244119074, + "grad_norm": 2.0995595455169678, + "learning_rate": 0.00013508397146252965, + "loss": 0.6325, + "step": 23307 + }, + { + "epoch": 1.6173756158490042, + "grad_norm": 1.7288998365402222, + "learning_rate": 0.000135078709936021, + "loss": 0.5688, + "step": 23308 + }, + { + "epoch": 1.6174450072861009, + "grad_norm": 3.658163547515869, + "learning_rate": 0.00013507344829877295, + "loss": 0.679, + "step": 23309 + }, + { + "epoch": 1.6175143987231975, + "grad_norm": 1.8495454788208008, + "learning_rate": 0.00013506818655080208, + "loss": 0.5483, + "step": 23310 + }, + { + "epoch": 1.6175837901602943, + "grad_norm": 2.4847819805145264, + "learning_rate": 0.00013506292469212504, + "loss": 0.5677, + "step": 23311 + }, + { + "epoch": 1.617653181597391, + "grad_norm": 1.7538880109786987, + "learning_rate": 0.0001350576627227584, + "loss": 0.5582, + "step": 23312 + }, + { + "epoch": 1.6177225730344875, + "grad_norm": 1.9138094186782837, + "learning_rate": 0.0001350524006427188, + "loss": 0.5652, + "step": 23313 + }, + { + "epoch": 1.6177919644715844, + "grad_norm": 2.0352284908294678, + "learning_rate": 0.0001350471384520228, + "loss": 0.6422, + "step": 23314 + }, + { + "epoch": 1.6178613559086807, + "grad_norm": 2.307725429534912, + "learning_rate": 0.0001350418761506871, + "loss": 0.7822, + "step": 23315 + }, + { + "epoch": 1.6179307473457776, + "grad_norm": 2.0278327465057373, + "learning_rate": 0.00013503661373872821, + "loss": 0.55, + "step": 23316 + }, + { + "epoch": 1.6180001387828742, + "grad_norm": 2.354100227355957, + "learning_rate": 0.00013503135121616284, + "loss": 0.6372, + "step": 23317 + }, + { + "epoch": 1.6180695302199708, + "grad_norm": 2.3526082038879395, + "learning_rate": 0.00013502608858300755, + "loss": 0.5181, + "step": 23318 + }, + { + "epoch": 1.6181389216570676, + "grad_norm": 1.451820969581604, + "learning_rate": 0.00013502082583927895, + "loss": 0.7524, + "step": 23319 + }, + { + "epoch": 1.618208313094164, + "grad_norm": 3.365752696990967, + "learning_rate": 0.00013501556298499366, + "loss": 0.5601, + "step": 23320 + }, + { + "epoch": 1.6182777045312609, + "grad_norm": 2.109388589859009, + "learning_rate": 0.0001350103000201683, + "loss": 0.7322, + "step": 23321 + }, + { + "epoch": 1.6183470959683575, + "grad_norm": 3.419877052307129, + "learning_rate": 0.00013500503694481952, + "loss": 0.69, + "step": 23322 + }, + { + "epoch": 1.618416487405454, + "grad_norm": 1.5185189247131348, + "learning_rate": 0.00013499977375896384, + "loss": 0.5423, + "step": 23323 + }, + { + "epoch": 1.618485878842551, + "grad_norm": 1.3945540189743042, + "learning_rate": 0.00013499451046261796, + "loss": 0.5203, + "step": 23324 + }, + { + "epoch": 1.6185552702796475, + "grad_norm": 1.8591721057891846, + "learning_rate": 0.00013498924705579853, + "loss": 0.6607, + "step": 23325 + }, + { + "epoch": 1.6186246617167441, + "grad_norm": 1.9230254888534546, + "learning_rate": 0.00013498398353852203, + "loss": 0.7509, + "step": 23326 + }, + { + "epoch": 1.618694053153841, + "grad_norm": 1.78435218334198, + "learning_rate": 0.00013497871991080518, + "loss": 0.8404, + "step": 23327 + }, + { + "epoch": 1.6187634445909374, + "grad_norm": 2.287510633468628, + "learning_rate": 0.00013497345617266454, + "loss": 0.6171, + "step": 23328 + }, + { + "epoch": 1.6188328360280342, + "grad_norm": 1.4260022640228271, + "learning_rate": 0.00013496819232411682, + "loss": 0.5042, + "step": 23329 + }, + { + "epoch": 1.6189022274651308, + "grad_norm": 2.3594441413879395, + "learning_rate": 0.00013496292836517851, + "loss": 0.6519, + "step": 23330 + }, + { + "epoch": 1.6189716189022274, + "grad_norm": 2.1465625762939453, + "learning_rate": 0.00013495766429586634, + "loss": 0.6066, + "step": 23331 + }, + { + "epoch": 1.6190410103393242, + "grad_norm": 2.632741689682007, + "learning_rate": 0.00013495240011619685, + "loss": 0.7666, + "step": 23332 + }, + { + "epoch": 1.6191104017764206, + "grad_norm": 1.4822423458099365, + "learning_rate": 0.00013494713582618673, + "loss": 0.486, + "step": 23333 + }, + { + "epoch": 1.6191797932135175, + "grad_norm": 1.884340524673462, + "learning_rate": 0.0001349418714258525, + "loss": 0.5618, + "step": 23334 + }, + { + "epoch": 1.619249184650614, + "grad_norm": 1.7264931201934814, + "learning_rate": 0.00013493660691521087, + "loss": 0.6797, + "step": 23335 + }, + { + "epoch": 1.6193185760877107, + "grad_norm": 1.8214410543441772, + "learning_rate": 0.00013493134229427842, + "loss": 0.8029, + "step": 23336 + }, + { + "epoch": 1.6193879675248075, + "grad_norm": 1.815924048423767, + "learning_rate": 0.00013492607756307177, + "loss": 0.553, + "step": 23337 + }, + { + "epoch": 1.6194573589619041, + "grad_norm": 1.6446478366851807, + "learning_rate": 0.00013492081272160753, + "loss": 0.6107, + "step": 23338 + }, + { + "epoch": 1.6195267503990007, + "grad_norm": 1.6243302822113037, + "learning_rate": 0.00013491554776990237, + "loss": 0.5726, + "step": 23339 + }, + { + "epoch": 1.6195961418360976, + "grad_norm": 1.7684762477874756, + "learning_rate": 0.00013491028270797286, + "loss": 0.7087, + "step": 23340 + }, + { + "epoch": 1.619665533273194, + "grad_norm": 1.7498773336410522, + "learning_rate": 0.00013490501753583563, + "loss": 0.648, + "step": 23341 + }, + { + "epoch": 1.6197349247102908, + "grad_norm": 1.6981480121612549, + "learning_rate": 0.00013489975225350727, + "loss": 0.7109, + "step": 23342 + }, + { + "epoch": 1.6198043161473874, + "grad_norm": 1.8971501588821411, + "learning_rate": 0.00013489448686100451, + "loss": 0.5479, + "step": 23343 + }, + { + "epoch": 1.619873707584484, + "grad_norm": 1.598398208618164, + "learning_rate": 0.0001348892213583439, + "loss": 0.7238, + "step": 23344 + }, + { + "epoch": 1.6199430990215808, + "grad_norm": 1.3332902193069458, + "learning_rate": 0.000134883955745542, + "loss": 0.5564, + "step": 23345 + }, + { + "epoch": 1.6200124904586772, + "grad_norm": 2.028188943862915, + "learning_rate": 0.0001348786900226156, + "loss": 0.7006, + "step": 23346 + }, + { + "epoch": 1.620081881895774, + "grad_norm": 1.8680864572525024, + "learning_rate": 0.00013487342418958115, + "loss": 0.7606, + "step": 23347 + }, + { + "epoch": 1.6201512733328707, + "grad_norm": 2.5665283203125, + "learning_rate": 0.00013486815824645537, + "loss": 0.6313, + "step": 23348 + }, + { + "epoch": 1.6202206647699673, + "grad_norm": 2.009312152862549, + "learning_rate": 0.00013486289219325485, + "loss": 0.6221, + "step": 23349 + }, + { + "epoch": 1.6202900562070641, + "grad_norm": 2.052711009979248, + "learning_rate": 0.00013485762602999626, + "loss": 0.6546, + "step": 23350 + }, + { + "epoch": 1.6203594476441607, + "grad_norm": 1.1473431587219238, + "learning_rate": 0.00013485235975669614, + "loss": 0.3453, + "step": 23351 + }, + { + "epoch": 1.6204288390812573, + "grad_norm": 2.0199360847473145, + "learning_rate": 0.00013484709337337116, + "loss": 0.7734, + "step": 23352 + }, + { + "epoch": 1.6204982305183542, + "grad_norm": 1.0622377395629883, + "learning_rate": 0.00013484182688003798, + "loss": 0.4557, + "step": 23353 + }, + { + "epoch": 1.6205676219554506, + "grad_norm": 2.532794952392578, + "learning_rate": 0.00013483656027671317, + "loss": 0.6294, + "step": 23354 + }, + { + "epoch": 1.6206370133925474, + "grad_norm": 2.4030163288116455, + "learning_rate": 0.00013483129356341343, + "loss": 0.5394, + "step": 23355 + }, + { + "epoch": 1.620706404829644, + "grad_norm": 2.1167399883270264, + "learning_rate": 0.0001348260267401553, + "loss": 0.5905, + "step": 23356 + }, + { + "epoch": 1.6207757962667406, + "grad_norm": 1.7985033988952637, + "learning_rate": 0.00013482075980695543, + "loss": 0.4599, + "step": 23357 + }, + { + "epoch": 1.6208451877038375, + "grad_norm": 1.6299718618392944, + "learning_rate": 0.0001348154927638305, + "loss": 0.6122, + "step": 23358 + }, + { + "epoch": 1.620914579140934, + "grad_norm": 1.9246718883514404, + "learning_rate": 0.00013481022561079707, + "loss": 0.7607, + "step": 23359 + }, + { + "epoch": 1.6209839705780307, + "grad_norm": 2.150825023651123, + "learning_rate": 0.0001348049583478718, + "loss": 0.7132, + "step": 23360 + }, + { + "epoch": 1.6210533620151273, + "grad_norm": 1.7214100360870361, + "learning_rate": 0.0001347996909750713, + "loss": 0.7077, + "step": 23361 + }, + { + "epoch": 1.621122753452224, + "grad_norm": 1.4943524599075317, + "learning_rate": 0.00013479442349241224, + "loss": 0.5514, + "step": 23362 + }, + { + "epoch": 1.6211921448893207, + "grad_norm": 1.8894673585891724, + "learning_rate": 0.0001347891558999112, + "loss": 0.5657, + "step": 23363 + }, + { + "epoch": 1.6212615363264173, + "grad_norm": 1.3807326555252075, + "learning_rate": 0.00013478388819758484, + "loss": 0.5753, + "step": 23364 + }, + { + "epoch": 1.621330927763514, + "grad_norm": 1.804715871810913, + "learning_rate": 0.0001347786203854498, + "loss": 0.6579, + "step": 23365 + }, + { + "epoch": 1.6214003192006108, + "grad_norm": 1.4789314270019531, + "learning_rate": 0.00013477335246352267, + "loss": 0.449, + "step": 23366 + }, + { + "epoch": 1.6214697106377072, + "grad_norm": 1.7672789096832275, + "learning_rate": 0.00013476808443182013, + "loss": 0.75, + "step": 23367 + }, + { + "epoch": 1.621539102074804, + "grad_norm": 1.7805556058883667, + "learning_rate": 0.00013476281629035876, + "loss": 0.6122, + "step": 23368 + }, + { + "epoch": 1.6216084935119006, + "grad_norm": 1.391942024230957, + "learning_rate": 0.00013475754803915518, + "loss": 0.3822, + "step": 23369 + }, + { + "epoch": 1.6216778849489972, + "grad_norm": 1.8327692747116089, + "learning_rate": 0.00013475227967822613, + "loss": 0.9588, + "step": 23370 + }, + { + "epoch": 1.621747276386094, + "grad_norm": 1.237421989440918, + "learning_rate": 0.0001347470112075881, + "loss": 0.4349, + "step": 23371 + }, + { + "epoch": 1.6218166678231907, + "grad_norm": 1.8131097555160522, + "learning_rate": 0.00013474174262725777, + "loss": 0.7439, + "step": 23372 + }, + { + "epoch": 1.6218860592602873, + "grad_norm": 1.6538617610931396, + "learning_rate": 0.00013473647393725185, + "loss": 0.6736, + "step": 23373 + }, + { + "epoch": 1.621955450697384, + "grad_norm": 1.9102013111114502, + "learning_rate": 0.0001347312051375869, + "loss": 0.6403, + "step": 23374 + }, + { + "epoch": 1.6220248421344805, + "grad_norm": 2.170654058456421, + "learning_rate": 0.00013472593622827954, + "loss": 0.8308, + "step": 23375 + }, + { + "epoch": 1.6220942335715773, + "grad_norm": 2.026120185852051, + "learning_rate": 0.00013472066720934643, + "loss": 0.6476, + "step": 23376 + }, + { + "epoch": 1.622163625008674, + "grad_norm": 1.6096450090408325, + "learning_rate": 0.00013471539808080423, + "loss": 0.5059, + "step": 23377 + }, + { + "epoch": 1.6222330164457706, + "grad_norm": 1.217847228050232, + "learning_rate": 0.00013471012884266951, + "loss": 0.3168, + "step": 23378 + }, + { + "epoch": 1.6223024078828674, + "grad_norm": 1.8215135335922241, + "learning_rate": 0.00013470485949495896, + "loss": 0.7355, + "step": 23379 + }, + { + "epoch": 1.6223717993199638, + "grad_norm": 1.8842675685882568, + "learning_rate": 0.0001346995900376892, + "loss": 0.6855, + "step": 23380 + }, + { + "epoch": 1.6224411907570606, + "grad_norm": 1.937721610069275, + "learning_rate": 0.00013469432047087687, + "loss": 0.7574, + "step": 23381 + }, + { + "epoch": 1.6225105821941572, + "grad_norm": 1.3452800512313843, + "learning_rate": 0.00013468905079453856, + "loss": 0.4272, + "step": 23382 + }, + { + "epoch": 1.6225799736312538, + "grad_norm": 2.4963603019714355, + "learning_rate": 0.00013468378100869096, + "loss": 0.5755, + "step": 23383 + }, + { + "epoch": 1.6226493650683507, + "grad_norm": 1.6176356077194214, + "learning_rate": 0.0001346785111133507, + "loss": 0.589, + "step": 23384 + }, + { + "epoch": 1.6227187565054473, + "grad_norm": 3.0825366973876953, + "learning_rate": 0.0001346732411085344, + "loss": 1.0255, + "step": 23385 + }, + { + "epoch": 1.622788147942544, + "grad_norm": 1.682318925857544, + "learning_rate": 0.00013466797099425867, + "loss": 0.6218, + "step": 23386 + }, + { + "epoch": 1.6228575393796405, + "grad_norm": 1.4479658603668213, + "learning_rate": 0.0001346627007705402, + "loss": 0.4656, + "step": 23387 + }, + { + "epoch": 1.6229269308167371, + "grad_norm": 1.5942895412445068, + "learning_rate": 0.00013465743043739563, + "loss": 0.5199, + "step": 23388 + }, + { + "epoch": 1.622996322253834, + "grad_norm": 1.5800259113311768, + "learning_rate": 0.00013465215999484154, + "loss": 0.502, + "step": 23389 + }, + { + "epoch": 1.6230657136909306, + "grad_norm": 1.9868640899658203, + "learning_rate": 0.0001346468894428946, + "loss": 0.5966, + "step": 23390 + }, + { + "epoch": 1.6231351051280272, + "grad_norm": 0.8649921417236328, + "learning_rate": 0.00013464161878157146, + "loss": 0.4163, + "step": 23391 + }, + { + "epoch": 1.623204496565124, + "grad_norm": 1.6683638095855713, + "learning_rate": 0.00013463634801088876, + "loss": 0.6076, + "step": 23392 + }, + { + "epoch": 1.6232738880022204, + "grad_norm": 3.9108142852783203, + "learning_rate": 0.0001346310771308631, + "loss": 0.7306, + "step": 23393 + }, + { + "epoch": 1.6233432794393172, + "grad_norm": 1.3646628856658936, + "learning_rate": 0.0001346258061415112, + "loss": 0.4013, + "step": 23394 + }, + { + "epoch": 1.6234126708764138, + "grad_norm": 2.012990713119507, + "learning_rate": 0.0001346205350428496, + "loss": 0.9046, + "step": 23395 + }, + { + "epoch": 1.6234820623135104, + "grad_norm": 1.659745216369629, + "learning_rate": 0.00013461526383489497, + "loss": 0.5738, + "step": 23396 + }, + { + "epoch": 1.6235514537506073, + "grad_norm": 1.6018047332763672, + "learning_rate": 0.000134609992517664, + "loss": 0.6951, + "step": 23397 + }, + { + "epoch": 1.6236208451877039, + "grad_norm": 1.1356526613235474, + "learning_rate": 0.0001346047210911733, + "loss": 0.4833, + "step": 23398 + }, + { + "epoch": 1.6236902366248005, + "grad_norm": 2.9643328189849854, + "learning_rate": 0.0001345994495554395, + "loss": 0.4839, + "step": 23399 + }, + { + "epoch": 1.623759628061897, + "grad_norm": 1.855145812034607, + "learning_rate": 0.00013459417791047927, + "loss": 0.7773, + "step": 23400 + }, + { + "epoch": 1.6238290194989937, + "grad_norm": 1.3628897666931152, + "learning_rate": 0.00013458890615630922, + "loss": 0.543, + "step": 23401 + }, + { + "epoch": 1.6238984109360906, + "grad_norm": 2.296753406524658, + "learning_rate": 0.00013458363429294602, + "loss": 0.7799, + "step": 23402 + }, + { + "epoch": 1.6239678023731872, + "grad_norm": 1.9182207584381104, + "learning_rate": 0.0001345783623204063, + "loss": 0.7431, + "step": 23403 + }, + { + "epoch": 1.6240371938102838, + "grad_norm": 1.7307524681091309, + "learning_rate": 0.00013457309023870664, + "loss": 0.672, + "step": 23404 + }, + { + "epoch": 1.6241065852473806, + "grad_norm": 2.027308225631714, + "learning_rate": 0.00013456781804786384, + "loss": 0.5612, + "step": 23405 + }, + { + "epoch": 1.624175976684477, + "grad_norm": 1.679264783859253, + "learning_rate": 0.0001345625457478944, + "loss": 0.4671, + "step": 23406 + }, + { + "epoch": 1.6242453681215738, + "grad_norm": 2.3555989265441895, + "learning_rate": 0.000134557273338815, + "loss": 0.5162, + "step": 23407 + }, + { + "epoch": 1.6243147595586704, + "grad_norm": 1.7935818433761597, + "learning_rate": 0.00013455200082064232, + "loss": 0.7527, + "step": 23408 + }, + { + "epoch": 1.624384150995767, + "grad_norm": 2.3462753295898438, + "learning_rate": 0.000134546728193393, + "loss": 1.0969, + "step": 23409 + }, + { + "epoch": 1.6244535424328639, + "grad_norm": 2.5166404247283936, + "learning_rate": 0.00013454145545708363, + "loss": 0.5611, + "step": 23410 + }, + { + "epoch": 1.6245229338699605, + "grad_norm": 1.8909363746643066, + "learning_rate": 0.0001345361826117309, + "loss": 0.798, + "step": 23411 + }, + { + "epoch": 1.624592325307057, + "grad_norm": 1.5026845932006836, + "learning_rate": 0.00013453090965735145, + "loss": 0.5819, + "step": 23412 + }, + { + "epoch": 1.6246617167441537, + "grad_norm": 1.4323437213897705, + "learning_rate": 0.00013452563659396193, + "loss": 0.4605, + "step": 23413 + }, + { + "epoch": 1.6247311081812503, + "grad_norm": 1.8601590394973755, + "learning_rate": 0.00013452036342157897, + "loss": 0.7661, + "step": 23414 + }, + { + "epoch": 1.6248004996183472, + "grad_norm": 1.5016247034072876, + "learning_rate": 0.00013451509014021925, + "loss": 0.5966, + "step": 23415 + }, + { + "epoch": 1.6248698910554438, + "grad_norm": 1.4757111072540283, + "learning_rate": 0.00013450981674989937, + "loss": 0.5982, + "step": 23416 + }, + { + "epoch": 1.6249392824925404, + "grad_norm": 1.9443717002868652, + "learning_rate": 0.00013450454325063602, + "loss": 0.4307, + "step": 23417 + }, + { + "epoch": 1.6250086739296372, + "grad_norm": 1.5917490720748901, + "learning_rate": 0.00013449926964244582, + "loss": 0.6286, + "step": 23418 + }, + { + "epoch": 1.6250780653667336, + "grad_norm": 1.79570734500885, + "learning_rate": 0.00013449399592534544, + "loss": 0.7087, + "step": 23419 + }, + { + "epoch": 1.6251474568038304, + "grad_norm": 2.0341663360595703, + "learning_rate": 0.00013448872209935152, + "loss": 0.6612, + "step": 23420 + }, + { + "epoch": 1.625216848240927, + "grad_norm": 3.7109546661376953, + "learning_rate": 0.0001344834481644807, + "loss": 0.6734, + "step": 23421 + }, + { + "epoch": 1.6252862396780237, + "grad_norm": 1.6807609796524048, + "learning_rate": 0.00013447817412074966, + "loss": 0.6482, + "step": 23422 + }, + { + "epoch": 1.6253556311151205, + "grad_norm": 2.124939203262329, + "learning_rate": 0.00013447289996817497, + "loss": 0.5205, + "step": 23423 + }, + { + "epoch": 1.625425022552217, + "grad_norm": 1.471373438835144, + "learning_rate": 0.00013446762570677335, + "loss": 0.6564, + "step": 23424 + }, + { + "epoch": 1.6254944139893137, + "grad_norm": 2.110719919204712, + "learning_rate": 0.00013446235133656145, + "loss": 0.401, + "step": 23425 + }, + { + "epoch": 1.6255638054264105, + "grad_norm": 2.144761800765991, + "learning_rate": 0.0001344570768575559, + "loss": 0.6148, + "step": 23426 + }, + { + "epoch": 1.625633196863507, + "grad_norm": 1.3281311988830566, + "learning_rate": 0.00013445180226977335, + "loss": 0.4026, + "step": 23427 + }, + { + "epoch": 1.6257025883006038, + "grad_norm": 1.494923710823059, + "learning_rate": 0.00013444652757323047, + "loss": 0.5891, + "step": 23428 + }, + { + "epoch": 1.6257719797377004, + "grad_norm": 1.873335599899292, + "learning_rate": 0.0001344412527679439, + "loss": 0.7058, + "step": 23429 + }, + { + "epoch": 1.625841371174797, + "grad_norm": 1.4660269021987915, + "learning_rate": 0.00013443597785393028, + "loss": 0.4737, + "step": 23430 + }, + { + "epoch": 1.6259107626118938, + "grad_norm": 1.893796682357788, + "learning_rate": 0.00013443070283120626, + "loss": 0.8291, + "step": 23431 + }, + { + "epoch": 1.6259801540489902, + "grad_norm": 1.731006145477295, + "learning_rate": 0.00013442542769978854, + "loss": 0.5778, + "step": 23432 + }, + { + "epoch": 1.626049545486087, + "grad_norm": 1.2878350019454956, + "learning_rate": 0.00013442015245969372, + "loss": 0.4571, + "step": 23433 + }, + { + "epoch": 1.6261189369231837, + "grad_norm": 1.9220709800720215, + "learning_rate": 0.00013441487711093848, + "loss": 0.6997, + "step": 23434 + }, + { + "epoch": 1.6261883283602803, + "grad_norm": 2.2999837398529053, + "learning_rate": 0.00013440960165353946, + "loss": 0.6391, + "step": 23435 + }, + { + "epoch": 1.626257719797377, + "grad_norm": 1.9539803266525269, + "learning_rate": 0.00013440432608751332, + "loss": 0.8628, + "step": 23436 + }, + { + "epoch": 1.6263271112344737, + "grad_norm": 1.5581934452056885, + "learning_rate": 0.00013439905041287673, + "loss": 0.5928, + "step": 23437 + }, + { + "epoch": 1.6263965026715703, + "grad_norm": 1.863205909729004, + "learning_rate": 0.0001343937746296463, + "loss": 0.5442, + "step": 23438 + }, + { + "epoch": 1.6264658941086672, + "grad_norm": 2.0420713424682617, + "learning_rate": 0.00013438849873783875, + "loss": 0.7388, + "step": 23439 + }, + { + "epoch": 1.6265352855457635, + "grad_norm": 1.6309407949447632, + "learning_rate": 0.00013438322273747066, + "loss": 0.5981, + "step": 23440 + }, + { + "epoch": 1.6266046769828604, + "grad_norm": 1.6940808296203613, + "learning_rate": 0.00013437794662855875, + "loss": 0.5681, + "step": 23441 + }, + { + "epoch": 1.626674068419957, + "grad_norm": 1.9550061225891113, + "learning_rate": 0.00013437267041111964, + "loss": 0.7023, + "step": 23442 + }, + { + "epoch": 1.6267434598570536, + "grad_norm": 1.5778056383132935, + "learning_rate": 0.00013436739408517, + "loss": 0.5926, + "step": 23443 + }, + { + "epoch": 1.6268128512941504, + "grad_norm": 1.9773770570755005, + "learning_rate": 0.0001343621176507265, + "loss": 0.7682, + "step": 23444 + }, + { + "epoch": 1.6268822427312468, + "grad_norm": 2.029568672180176, + "learning_rate": 0.00013435684110780572, + "loss": 0.7764, + "step": 23445 + }, + { + "epoch": 1.6269516341683437, + "grad_norm": 1.7406530380249023, + "learning_rate": 0.00013435156445642445, + "loss": 0.4668, + "step": 23446 + }, + { + "epoch": 1.6270210256054403, + "grad_norm": 1.76614511013031, + "learning_rate": 0.00013434628769659924, + "loss": 0.5602, + "step": 23447 + }, + { + "epoch": 1.6270904170425369, + "grad_norm": 2.255873680114746, + "learning_rate": 0.0001343410108283468, + "loss": 0.6062, + "step": 23448 + }, + { + "epoch": 1.6271598084796337, + "grad_norm": 1.3499536514282227, + "learning_rate": 0.00013433573385168373, + "loss": 0.467, + "step": 23449 + }, + { + "epoch": 1.6272291999167303, + "grad_norm": 2.154521942138672, + "learning_rate": 0.00013433045676662682, + "loss": 0.6419, + "step": 23450 + }, + { + "epoch": 1.627298591353827, + "grad_norm": 1.5466338396072388, + "learning_rate": 0.00013432517957319258, + "loss": 0.7144, + "step": 23451 + }, + { + "epoch": 1.6273679827909238, + "grad_norm": 2.096611738204956, + "learning_rate": 0.00013431990227139772, + "loss": 0.7182, + "step": 23452 + }, + { + "epoch": 1.6274373742280202, + "grad_norm": 1.5308873653411865, + "learning_rate": 0.00013431462486125897, + "loss": 0.5427, + "step": 23453 + }, + { + "epoch": 1.627506765665117, + "grad_norm": 1.5060397386550903, + "learning_rate": 0.00013430934734279287, + "loss": 0.6969, + "step": 23454 + }, + { + "epoch": 1.6275761571022136, + "grad_norm": 1.6505945920944214, + "learning_rate": 0.00013430406971601618, + "loss": 0.4904, + "step": 23455 + }, + { + "epoch": 1.6276455485393102, + "grad_norm": 1.612939476966858, + "learning_rate": 0.00013429879198094549, + "loss": 0.4699, + "step": 23456 + }, + { + "epoch": 1.627714939976407, + "grad_norm": 1.9778203964233398, + "learning_rate": 0.00013429351413759752, + "loss": 0.7096, + "step": 23457 + }, + { + "epoch": 1.6277843314135034, + "grad_norm": 1.2018204927444458, + "learning_rate": 0.00013428823618598888, + "loss": 0.4226, + "step": 23458 + }, + { + "epoch": 1.6278537228506003, + "grad_norm": 1.7866623401641846, + "learning_rate": 0.00013428295812613626, + "loss": 0.5607, + "step": 23459 + }, + { + "epoch": 1.6279231142876969, + "grad_norm": 1.8011516332626343, + "learning_rate": 0.00013427767995805636, + "loss": 0.4959, + "step": 23460 + }, + { + "epoch": 1.6279925057247935, + "grad_norm": 1.7663730382919312, + "learning_rate": 0.00013427240168176576, + "loss": 0.5853, + "step": 23461 + }, + { + "epoch": 1.6280618971618903, + "grad_norm": 2.1607210636138916, + "learning_rate": 0.00013426712329728116, + "loss": 0.651, + "step": 23462 + }, + { + "epoch": 1.628131288598987, + "grad_norm": 2.1626105308532715, + "learning_rate": 0.00013426184480461924, + "loss": 0.4988, + "step": 23463 + }, + { + "epoch": 1.6282006800360835, + "grad_norm": 1.4929487705230713, + "learning_rate": 0.00013425656620379668, + "loss": 0.5298, + "step": 23464 + }, + { + "epoch": 1.6282700714731804, + "grad_norm": 2.126295804977417, + "learning_rate": 0.00013425128749483008, + "loss": 0.5044, + "step": 23465 + }, + { + "epoch": 1.6283394629102768, + "grad_norm": 2.2466936111450195, + "learning_rate": 0.00013424600867773613, + "loss": 0.6179, + "step": 23466 + }, + { + "epoch": 1.6284088543473736, + "grad_norm": 3.2101593017578125, + "learning_rate": 0.00013424072975253153, + "loss": 0.7871, + "step": 23467 + }, + { + "epoch": 1.6284782457844702, + "grad_norm": 2.1784517765045166, + "learning_rate": 0.00013423545071923292, + "loss": 0.5505, + "step": 23468 + }, + { + "epoch": 1.6285476372215668, + "grad_norm": 1.2886499166488647, + "learning_rate": 0.00013423017157785692, + "loss": 0.6524, + "step": 23469 + }, + { + "epoch": 1.6286170286586636, + "grad_norm": 1.3313549757003784, + "learning_rate": 0.0001342248923284203, + "loss": 0.5931, + "step": 23470 + }, + { + "epoch": 1.62868642009576, + "grad_norm": 1.819820523262024, + "learning_rate": 0.00013421961297093964, + "loss": 0.6569, + "step": 23471 + }, + { + "epoch": 1.6287558115328569, + "grad_norm": 2.258298397064209, + "learning_rate": 0.00013421433350543164, + "loss": 0.6005, + "step": 23472 + }, + { + "epoch": 1.6288252029699535, + "grad_norm": 1.4500449895858765, + "learning_rate": 0.00013420905393191293, + "loss": 0.6059, + "step": 23473 + }, + { + "epoch": 1.62889459440705, + "grad_norm": 1.9699501991271973, + "learning_rate": 0.00013420377425040026, + "loss": 0.5732, + "step": 23474 + }, + { + "epoch": 1.628963985844147, + "grad_norm": 1.4764074087142944, + "learning_rate": 0.00013419849446091018, + "loss": 0.6473, + "step": 23475 + }, + { + "epoch": 1.6290333772812435, + "grad_norm": 2.160984516143799, + "learning_rate": 0.00013419321456345946, + "loss": 0.7676, + "step": 23476 + }, + { + "epoch": 1.6291027687183401, + "grad_norm": 1.882732629776001, + "learning_rate": 0.00013418793455806472, + "loss": 0.7123, + "step": 23477 + }, + { + "epoch": 1.629172160155437, + "grad_norm": 1.558178424835205, + "learning_rate": 0.00013418265444474264, + "loss": 0.5524, + "step": 23478 + }, + { + "epoch": 1.6292415515925334, + "grad_norm": 1.7141532897949219, + "learning_rate": 0.00013417737422350987, + "loss": 0.597, + "step": 23479 + }, + { + "epoch": 1.6293109430296302, + "grad_norm": 1.236363172531128, + "learning_rate": 0.0001341720938943831, + "loss": 0.4558, + "step": 23480 + }, + { + "epoch": 1.6293803344667268, + "grad_norm": 1.4136583805084229, + "learning_rate": 0.00013416681345737903, + "loss": 0.5982, + "step": 23481 + }, + { + "epoch": 1.6294497259038234, + "grad_norm": 1.7184669971466064, + "learning_rate": 0.00013416153291251427, + "loss": 0.5875, + "step": 23482 + }, + { + "epoch": 1.6295191173409203, + "grad_norm": 1.906515121459961, + "learning_rate": 0.0001341562522598055, + "loss": 0.7712, + "step": 23483 + }, + { + "epoch": 1.6295885087780166, + "grad_norm": 1.8985542058944702, + "learning_rate": 0.0001341509714992694, + "loss": 0.5617, + "step": 23484 + }, + { + "epoch": 1.6296579002151135, + "grad_norm": 1.468146800994873, + "learning_rate": 0.00013414569063092266, + "loss": 0.6123, + "step": 23485 + }, + { + "epoch": 1.62972729165221, + "grad_norm": 1.4737728834152222, + "learning_rate": 0.00013414040965478193, + "loss": 0.6065, + "step": 23486 + }, + { + "epoch": 1.6297966830893067, + "grad_norm": 1.7089513540267944, + "learning_rate": 0.00013413512857086388, + "loss": 0.79, + "step": 23487 + }, + { + "epoch": 1.6298660745264035, + "grad_norm": 2.0073165893554688, + "learning_rate": 0.00013412984737918522, + "loss": 0.5116, + "step": 23488 + }, + { + "epoch": 1.6299354659635001, + "grad_norm": 1.8674547672271729, + "learning_rate": 0.00013412456607976258, + "loss": 0.4632, + "step": 23489 + }, + { + "epoch": 1.6300048574005968, + "grad_norm": 1.6065568923950195, + "learning_rate": 0.00013411928467261264, + "loss": 0.5785, + "step": 23490 + }, + { + "epoch": 1.6300742488376936, + "grad_norm": 1.4471256732940674, + "learning_rate": 0.0001341140031577521, + "loss": 0.411, + "step": 23491 + }, + { + "epoch": 1.63014364027479, + "grad_norm": 1.6263033151626587, + "learning_rate": 0.00013410872153519755, + "loss": 0.8205, + "step": 23492 + }, + { + "epoch": 1.6302130317118868, + "grad_norm": 1.4500672817230225, + "learning_rate": 0.00013410343980496575, + "loss": 0.501, + "step": 23493 + }, + { + "epoch": 1.6302824231489834, + "grad_norm": 1.5792665481567383, + "learning_rate": 0.00013409815796707334, + "loss": 0.5489, + "step": 23494 + }, + { + "epoch": 1.63035181458608, + "grad_norm": 2.2703559398651123, + "learning_rate": 0.00013409287602153707, + "loss": 0.8092, + "step": 23495 + }, + { + "epoch": 1.6304212060231769, + "grad_norm": 1.71962308883667, + "learning_rate": 0.00013408759396837347, + "loss": 0.6646, + "step": 23496 + }, + { + "epoch": 1.6304905974602732, + "grad_norm": 1.5724226236343384, + "learning_rate": 0.0001340823118075993, + "loss": 0.551, + "step": 23497 + }, + { + "epoch": 1.63055998889737, + "grad_norm": 3.0628621578216553, + "learning_rate": 0.00013407702953923125, + "loss": 0.8504, + "step": 23498 + }, + { + "epoch": 1.6306293803344667, + "grad_norm": 1.7076668739318848, + "learning_rate": 0.00013407174716328598, + "loss": 0.6661, + "step": 23499 + }, + { + "epoch": 1.6306987717715633, + "grad_norm": 1.0512690544128418, + "learning_rate": 0.0001340664646797801, + "loss": 0.3829, + "step": 23500 + }, + { + "epoch": 1.6307681632086601, + "grad_norm": 1.8246279954910278, + "learning_rate": 0.00013406118208873037, + "loss": 0.6438, + "step": 23501 + }, + { + "epoch": 1.6308375546457567, + "grad_norm": 1.6982656717300415, + "learning_rate": 0.00013405589939015347, + "loss": 0.6418, + "step": 23502 + }, + { + "epoch": 1.6309069460828534, + "grad_norm": 2.1827588081359863, + "learning_rate": 0.00013405061658406604, + "loss": 0.5556, + "step": 23503 + }, + { + "epoch": 1.6309763375199502, + "grad_norm": 1.893457055091858, + "learning_rate": 0.00013404533367048472, + "loss": 0.6755, + "step": 23504 + }, + { + "epoch": 1.6310457289570466, + "grad_norm": 1.4017143249511719, + "learning_rate": 0.0001340400506494263, + "loss": 0.4091, + "step": 23505 + }, + { + "epoch": 1.6311151203941434, + "grad_norm": 1.6291840076446533, + "learning_rate": 0.00013403476752090734, + "loss": 0.4367, + "step": 23506 + }, + { + "epoch": 1.63118451183124, + "grad_norm": 1.8182650804519653, + "learning_rate": 0.00013402948428494458, + "loss": 0.6306, + "step": 23507 + }, + { + "epoch": 1.6312539032683366, + "grad_norm": 2.162907600402832, + "learning_rate": 0.0001340242009415547, + "loss": 0.7343, + "step": 23508 + }, + { + "epoch": 1.6313232947054335, + "grad_norm": 2.9270384311676025, + "learning_rate": 0.00013401891749075437, + "loss": 0.8807, + "step": 23509 + }, + { + "epoch": 1.6313926861425299, + "grad_norm": 1.1678314208984375, + "learning_rate": 0.00013401363393256025, + "loss": 0.4727, + "step": 23510 + }, + { + "epoch": 1.6314620775796267, + "grad_norm": 1.5307272672653198, + "learning_rate": 0.00013400835026698902, + "loss": 0.5228, + "step": 23511 + }, + { + "epoch": 1.6315314690167233, + "grad_norm": 2.905832290649414, + "learning_rate": 0.00013400306649405738, + "loss": 0.8425, + "step": 23512 + }, + { + "epoch": 1.63160086045382, + "grad_norm": 1.8886330127716064, + "learning_rate": 0.000133997782613782, + "loss": 0.5974, + "step": 23513 + }, + { + "epoch": 1.6316702518909167, + "grad_norm": 1.9819034337997437, + "learning_rate": 0.00013399249862617959, + "loss": 0.6191, + "step": 23514 + }, + { + "epoch": 1.6317396433280134, + "grad_norm": 1.3343254327774048, + "learning_rate": 0.0001339872145312668, + "loss": 0.4809, + "step": 23515 + }, + { + "epoch": 1.63180903476511, + "grad_norm": 1.7877837419509888, + "learning_rate": 0.00013398193032906027, + "loss": 0.7999, + "step": 23516 + }, + { + "epoch": 1.6318784262022068, + "grad_norm": 1.331049919128418, + "learning_rate": 0.0001339766460195768, + "loss": 0.4866, + "step": 23517 + }, + { + "epoch": 1.6319478176393032, + "grad_norm": 1.467116117477417, + "learning_rate": 0.00013397136160283292, + "loss": 0.6696, + "step": 23518 + }, + { + "epoch": 1.6320172090764, + "grad_norm": 1.5588477849960327, + "learning_rate": 0.00013396607707884546, + "loss": 0.6404, + "step": 23519 + }, + { + "epoch": 1.6320866005134966, + "grad_norm": 1.7696634531021118, + "learning_rate": 0.000133960792447631, + "loss": 0.5579, + "step": 23520 + }, + { + "epoch": 1.6321559919505932, + "grad_norm": 1.489577054977417, + "learning_rate": 0.00013395550770920627, + "loss": 0.6113, + "step": 23521 + }, + { + "epoch": 1.63222538338769, + "grad_norm": 1.589468002319336, + "learning_rate": 0.00013395022286358797, + "loss": 0.6966, + "step": 23522 + }, + { + "epoch": 1.6322947748247865, + "grad_norm": 1.3483794927597046, + "learning_rate": 0.0001339449379107927, + "loss": 0.5787, + "step": 23523 + }, + { + "epoch": 1.6323641662618833, + "grad_norm": 2.5922598838806152, + "learning_rate": 0.00013393965285083724, + "loss": 0.6142, + "step": 23524 + }, + { + "epoch": 1.63243355769898, + "grad_norm": 1.5617748498916626, + "learning_rate": 0.00013393436768373821, + "loss": 0.5186, + "step": 23525 + }, + { + "epoch": 1.6325029491360765, + "grad_norm": 1.9229776859283447, + "learning_rate": 0.00013392908240951237, + "loss": 0.7161, + "step": 23526 + }, + { + "epoch": 1.6325723405731734, + "grad_norm": 1.7133715152740479, + "learning_rate": 0.0001339237970281763, + "loss": 0.4864, + "step": 23527 + }, + { + "epoch": 1.63264173201027, + "grad_norm": 2.039769411087036, + "learning_rate": 0.00013391851153974675, + "loss": 0.684, + "step": 23528 + }, + { + "epoch": 1.6327111234473666, + "grad_norm": 1.7236812114715576, + "learning_rate": 0.0001339132259442404, + "loss": 0.5315, + "step": 23529 + }, + { + "epoch": 1.6327805148844634, + "grad_norm": 1.2350354194641113, + "learning_rate": 0.0001339079402416739, + "loss": 0.4539, + "step": 23530 + }, + { + "epoch": 1.6328499063215598, + "grad_norm": 3.574082374572754, + "learning_rate": 0.00013390265443206402, + "loss": 0.7085, + "step": 23531 + }, + { + "epoch": 1.6329192977586566, + "grad_norm": 1.9785362482070923, + "learning_rate": 0.00013389736851542736, + "loss": 0.8429, + "step": 23532 + }, + { + "epoch": 1.6329886891957532, + "grad_norm": 1.5087920427322388, + "learning_rate": 0.00013389208249178064, + "loss": 0.5441, + "step": 23533 + }, + { + "epoch": 1.6330580806328499, + "grad_norm": 2.5039615631103516, + "learning_rate": 0.00013388679636114058, + "loss": 0.5783, + "step": 23534 + }, + { + "epoch": 1.6331274720699467, + "grad_norm": 1.7545850276947021, + "learning_rate": 0.0001338815101235238, + "loss": 0.7393, + "step": 23535 + }, + { + "epoch": 1.633196863507043, + "grad_norm": 1.2820461988449097, + "learning_rate": 0.00013387622377894704, + "loss": 0.5029, + "step": 23536 + }, + { + "epoch": 1.63326625494414, + "grad_norm": 1.7542004585266113, + "learning_rate": 0.00013387093732742696, + "loss": 0.7349, + "step": 23537 + }, + { + "epoch": 1.6333356463812365, + "grad_norm": 1.6853094100952148, + "learning_rate": 0.00013386565076898025, + "loss": 0.6195, + "step": 23538 + }, + { + "epoch": 1.6334050378183331, + "grad_norm": 1.6925678253173828, + "learning_rate": 0.0001338603641036236, + "loss": 0.7078, + "step": 23539 + }, + { + "epoch": 1.63347442925543, + "grad_norm": 1.194434404373169, + "learning_rate": 0.00013385507733137378, + "loss": 0.383, + "step": 23540 + }, + { + "epoch": 1.6335438206925266, + "grad_norm": 1.7977015972137451, + "learning_rate": 0.00013384979045224734, + "loss": 0.807, + "step": 23541 + }, + { + "epoch": 1.6336132121296232, + "grad_norm": 1.307870864868164, + "learning_rate": 0.0001338445034662611, + "loss": 0.6003, + "step": 23542 + }, + { + "epoch": 1.63368260356672, + "grad_norm": 2.80116605758667, + "learning_rate": 0.00013383921637343164, + "loss": 0.6031, + "step": 23543 + }, + { + "epoch": 1.6337519950038164, + "grad_norm": 1.8723628520965576, + "learning_rate": 0.00013383392917377574, + "loss": 0.7242, + "step": 23544 + }, + { + "epoch": 1.6338213864409132, + "grad_norm": 1.3024173974990845, + "learning_rate": 0.00013382864186731, + "loss": 0.4923, + "step": 23545 + }, + { + "epoch": 1.6338907778780098, + "grad_norm": 1.8845010995864868, + "learning_rate": 0.00013382335445405122, + "loss": 0.451, + "step": 23546 + }, + { + "epoch": 1.6339601693151065, + "grad_norm": 2.0588619709014893, + "learning_rate": 0.00013381806693401599, + "loss": 0.7112, + "step": 23547 + }, + { + "epoch": 1.6340295607522033, + "grad_norm": 1.846339225769043, + "learning_rate": 0.00013381277930722102, + "loss": 0.728, + "step": 23548 + }, + { + "epoch": 1.6340989521892997, + "grad_norm": 1.6098206043243408, + "learning_rate": 0.00013380749157368307, + "loss": 0.5334, + "step": 23549 + }, + { + "epoch": 1.6341683436263965, + "grad_norm": 2.3919224739074707, + "learning_rate": 0.0001338022037334188, + "loss": 0.6187, + "step": 23550 + }, + { + "epoch": 1.6342377350634931, + "grad_norm": 1.7119401693344116, + "learning_rate": 0.00013379691578644488, + "loss": 0.5921, + "step": 23551 + }, + { + "epoch": 1.6343071265005897, + "grad_norm": 1.323711633682251, + "learning_rate": 0.000133791627732778, + "loss": 0.6216, + "step": 23552 + }, + { + "epoch": 1.6343765179376866, + "grad_norm": 2.0586650371551514, + "learning_rate": 0.00013378633957243492, + "loss": 0.5076, + "step": 23553 + }, + { + "epoch": 1.6344459093747832, + "grad_norm": 1.2641781568527222, + "learning_rate": 0.00013378105130543223, + "loss": 0.417, + "step": 23554 + }, + { + "epoch": 1.6345153008118798, + "grad_norm": 1.7932043075561523, + "learning_rate": 0.00013377576293178672, + "loss": 0.5144, + "step": 23555 + }, + { + "epoch": 1.6345846922489766, + "grad_norm": 1.4886537790298462, + "learning_rate": 0.00013377047445151504, + "loss": 0.5989, + "step": 23556 + }, + { + "epoch": 1.634654083686073, + "grad_norm": 2.2688117027282715, + "learning_rate": 0.00013376518586463387, + "loss": 0.6159, + "step": 23557 + }, + { + "epoch": 1.6347234751231698, + "grad_norm": 1.5067598819732666, + "learning_rate": 0.00013375989717115993, + "loss": 0.4506, + "step": 23558 + }, + { + "epoch": 1.6347928665602665, + "grad_norm": 1.3742388486862183, + "learning_rate": 0.00013375460837110995, + "loss": 0.5152, + "step": 23559 + }, + { + "epoch": 1.634862257997363, + "grad_norm": 2.3319103717803955, + "learning_rate": 0.00013374931946450054, + "loss": 0.4804, + "step": 23560 + }, + { + "epoch": 1.63493164943446, + "grad_norm": 1.9651345014572144, + "learning_rate": 0.00013374403045134846, + "loss": 0.5942, + "step": 23561 + }, + { + "epoch": 1.6350010408715563, + "grad_norm": 2.1350600719451904, + "learning_rate": 0.00013373874133167042, + "loss": 0.7902, + "step": 23562 + }, + { + "epoch": 1.6350704323086531, + "grad_norm": 1.8401206731796265, + "learning_rate": 0.00013373345210548304, + "loss": 0.7056, + "step": 23563 + }, + { + "epoch": 1.6351398237457497, + "grad_norm": 2.107311248779297, + "learning_rate": 0.00013372816277280308, + "loss": 0.6761, + "step": 23564 + }, + { + "epoch": 1.6352092151828463, + "grad_norm": 1.5575785636901855, + "learning_rate": 0.00013372287333364723, + "loss": 0.5567, + "step": 23565 + }, + { + "epoch": 1.6352786066199432, + "grad_norm": 1.605299472808838, + "learning_rate": 0.00013371758378803218, + "loss": 0.427, + "step": 23566 + }, + { + "epoch": 1.6353479980570398, + "grad_norm": 1.6220935583114624, + "learning_rate": 0.00013371229413597466, + "loss": 0.7302, + "step": 23567 + }, + { + "epoch": 1.6354173894941364, + "grad_norm": 1.16502845287323, + "learning_rate": 0.0001337070043774913, + "loss": 0.4094, + "step": 23568 + }, + { + "epoch": 1.6354867809312332, + "grad_norm": 1.514840006828308, + "learning_rate": 0.00013370171451259884, + "loss": 0.7558, + "step": 23569 + }, + { + "epoch": 1.6355561723683296, + "grad_norm": 1.2835736274719238, + "learning_rate": 0.00013369642454131396, + "loss": 0.5293, + "step": 23570 + }, + { + "epoch": 1.6356255638054265, + "grad_norm": 1.7251425981521606, + "learning_rate": 0.00013369113446365342, + "loss": 0.6022, + "step": 23571 + }, + { + "epoch": 1.635694955242523, + "grad_norm": 1.6596288681030273, + "learning_rate": 0.00013368584427963387, + "loss": 0.4382, + "step": 23572 + }, + { + "epoch": 1.6357643466796197, + "grad_norm": 1.9128659963607788, + "learning_rate": 0.000133680553989272, + "loss": 0.5572, + "step": 23573 + }, + { + "epoch": 1.6358337381167165, + "grad_norm": 1.5391336679458618, + "learning_rate": 0.00013367526359258456, + "loss": 0.5445, + "step": 23574 + }, + { + "epoch": 1.635903129553813, + "grad_norm": 1.6818805932998657, + "learning_rate": 0.00013366997308958818, + "loss": 0.5308, + "step": 23575 + }, + { + "epoch": 1.6359725209909097, + "grad_norm": 1.2181849479675293, + "learning_rate": 0.0001336646824802996, + "loss": 0.452, + "step": 23576 + }, + { + "epoch": 1.6360419124280063, + "grad_norm": 1.4593552350997925, + "learning_rate": 0.00013365939176473556, + "loss": 0.588, + "step": 23577 + }, + { + "epoch": 1.636111303865103, + "grad_norm": 2.068666458129883, + "learning_rate": 0.00013365410094291272, + "loss": 0.6806, + "step": 23578 + }, + { + "epoch": 1.6361806953021998, + "grad_norm": 1.4998719692230225, + "learning_rate": 0.00013364881001484778, + "loss": 0.5852, + "step": 23579 + }, + { + "epoch": 1.6362500867392964, + "grad_norm": 1.6971497535705566, + "learning_rate": 0.00013364351898055742, + "loss": 0.5171, + "step": 23580 + }, + { + "epoch": 1.636319478176393, + "grad_norm": 1.9211134910583496, + "learning_rate": 0.00013363822784005842, + "loss": 0.553, + "step": 23581 + }, + { + "epoch": 1.6363888696134898, + "grad_norm": 1.415550947189331, + "learning_rate": 0.0001336329365933674, + "loss": 0.372, + "step": 23582 + }, + { + "epoch": 1.6364582610505862, + "grad_norm": 1.9349560737609863, + "learning_rate": 0.0001336276452405011, + "loss": 0.7539, + "step": 23583 + }, + { + "epoch": 1.636527652487683, + "grad_norm": 1.4928396940231323, + "learning_rate": 0.00013362235378147627, + "loss": 0.5104, + "step": 23584 + }, + { + "epoch": 1.6365970439247797, + "grad_norm": 1.5643390417099, + "learning_rate": 0.00013361706221630954, + "loss": 0.5054, + "step": 23585 + }, + { + "epoch": 1.6366664353618763, + "grad_norm": 1.7369450330734253, + "learning_rate": 0.00013361177054501764, + "loss": 0.7242, + "step": 23586 + }, + { + "epoch": 1.6367358267989731, + "grad_norm": 1.7302650213241577, + "learning_rate": 0.00013360647876761727, + "loss": 0.5939, + "step": 23587 + }, + { + "epoch": 1.6368052182360695, + "grad_norm": 2.245701313018799, + "learning_rate": 0.00013360118688412518, + "loss": 0.8144, + "step": 23588 + }, + { + "epoch": 1.6368746096731663, + "grad_norm": 1.5229566097259521, + "learning_rate": 0.00013359589489455803, + "loss": 0.5019, + "step": 23589 + }, + { + "epoch": 1.636944001110263, + "grad_norm": 1.5338072776794434, + "learning_rate": 0.00013359060279893251, + "loss": 0.5909, + "step": 23590 + }, + { + "epoch": 1.6370133925473596, + "grad_norm": 1.6547608375549316, + "learning_rate": 0.0001335853105972654, + "loss": 0.6571, + "step": 23591 + }, + { + "epoch": 1.6370827839844564, + "grad_norm": 1.544039249420166, + "learning_rate": 0.00013358001828957334, + "loss": 0.5168, + "step": 23592 + }, + { + "epoch": 1.637152175421553, + "grad_norm": 1.7975647449493408, + "learning_rate": 0.00013357472587587305, + "loss": 0.7126, + "step": 23593 + }, + { + "epoch": 1.6372215668586496, + "grad_norm": 2.1611385345458984, + "learning_rate": 0.00013356943335618126, + "loss": 0.7377, + "step": 23594 + }, + { + "epoch": 1.6372909582957464, + "grad_norm": 1.724172830581665, + "learning_rate": 0.00013356414073051464, + "loss": 0.6855, + "step": 23595 + }, + { + "epoch": 1.6373603497328428, + "grad_norm": 2.071603298187256, + "learning_rate": 0.00013355884799888995, + "loss": 0.5452, + "step": 23596 + }, + { + "epoch": 1.6374297411699397, + "grad_norm": 1.8626738786697388, + "learning_rate": 0.00013355355516132385, + "loss": 0.6465, + "step": 23597 + }, + { + "epoch": 1.6374991326070363, + "grad_norm": 1.2800787687301636, + "learning_rate": 0.0001335482622178331, + "loss": 0.5285, + "step": 23598 + }, + { + "epoch": 1.637568524044133, + "grad_norm": 1.9374701976776123, + "learning_rate": 0.00013354296916843437, + "loss": 0.7166, + "step": 23599 + }, + { + "epoch": 1.6376379154812297, + "grad_norm": 1.5928303003311157, + "learning_rate": 0.00013353767601314434, + "loss": 0.62, + "step": 23600 + }, + { + "epoch": 1.6377073069183263, + "grad_norm": 1.7698436975479126, + "learning_rate": 0.00013353238275197978, + "loss": 0.8705, + "step": 23601 + }, + { + "epoch": 1.637776698355423, + "grad_norm": 1.9686156511306763, + "learning_rate": 0.0001335270893849574, + "loss": 0.5765, + "step": 23602 + }, + { + "epoch": 1.6378460897925196, + "grad_norm": 1.1370353698730469, + "learning_rate": 0.00013352179591209385, + "loss": 0.5324, + "step": 23603 + }, + { + "epoch": 1.6379154812296162, + "grad_norm": 2.0698740482330322, + "learning_rate": 0.00013351650233340588, + "loss": 0.537, + "step": 23604 + }, + { + "epoch": 1.637984872666713, + "grad_norm": 2.2669129371643066, + "learning_rate": 0.00013351120864891025, + "loss": 0.6211, + "step": 23605 + }, + { + "epoch": 1.6380542641038096, + "grad_norm": 1.6076709032058716, + "learning_rate": 0.00013350591485862358, + "loss": 0.7124, + "step": 23606 + }, + { + "epoch": 1.6381236555409062, + "grad_norm": 1.5095903873443604, + "learning_rate": 0.00013350062096256264, + "loss": 0.4785, + "step": 23607 + }, + { + "epoch": 1.638193046978003, + "grad_norm": 1.2675246000289917, + "learning_rate": 0.00013349532696074412, + "loss": 0.467, + "step": 23608 + }, + { + "epoch": 1.6382624384150994, + "grad_norm": 1.7302836179733276, + "learning_rate": 0.00013349003285318474, + "loss": 0.5555, + "step": 23609 + }, + { + "epoch": 1.6383318298521963, + "grad_norm": 1.5970383882522583, + "learning_rate": 0.0001334847386399012, + "loss": 0.6351, + "step": 23610 + }, + { + "epoch": 1.6384012212892929, + "grad_norm": 3.0368764400482178, + "learning_rate": 0.00013347944432091027, + "loss": 0.9576, + "step": 23611 + }, + { + "epoch": 1.6384706127263895, + "grad_norm": 3.0422229766845703, + "learning_rate": 0.00013347414989622858, + "loss": 0.9362, + "step": 23612 + }, + { + "epoch": 1.6385400041634863, + "grad_norm": 1.738789439201355, + "learning_rate": 0.00013346885536587292, + "loss": 0.8509, + "step": 23613 + }, + { + "epoch": 1.638609395600583, + "grad_norm": 1.345281958580017, + "learning_rate": 0.0001334635607298599, + "loss": 0.5181, + "step": 23614 + }, + { + "epoch": 1.6386787870376796, + "grad_norm": 1.2431427240371704, + "learning_rate": 0.00013345826598820635, + "loss": 0.4889, + "step": 23615 + }, + { + "epoch": 1.6387481784747762, + "grad_norm": 2.328583240509033, + "learning_rate": 0.00013345297114092893, + "loss": 0.7718, + "step": 23616 + }, + { + "epoch": 1.6388175699118728, + "grad_norm": 1.582891821861267, + "learning_rate": 0.00013344767618804433, + "loss": 0.5827, + "step": 23617 + }, + { + "epoch": 1.6388869613489696, + "grad_norm": 1.6576110124588013, + "learning_rate": 0.00013344238112956934, + "loss": 0.8011, + "step": 23618 + }, + { + "epoch": 1.6389563527860662, + "grad_norm": 2.25498104095459, + "learning_rate": 0.0001334370859655206, + "loss": 0.6392, + "step": 23619 + }, + { + "epoch": 1.6390257442231628, + "grad_norm": 1.7311673164367676, + "learning_rate": 0.00013343179069591487, + "loss": 0.6025, + "step": 23620 + }, + { + "epoch": 1.6390951356602597, + "grad_norm": 2.029972791671753, + "learning_rate": 0.00013342649532076885, + "loss": 0.7315, + "step": 23621 + }, + { + "epoch": 1.639164527097356, + "grad_norm": 1.475338101387024, + "learning_rate": 0.0001334211998400993, + "loss": 0.5427, + "step": 23622 + }, + { + "epoch": 1.6392339185344529, + "grad_norm": 1.3708982467651367, + "learning_rate": 0.00013341590425392286, + "loss": 0.7531, + "step": 23623 + }, + { + "epoch": 1.6393033099715495, + "grad_norm": 1.590425968170166, + "learning_rate": 0.0001334106085622563, + "loss": 0.6243, + "step": 23624 + }, + { + "epoch": 1.639372701408646, + "grad_norm": 2.374607563018799, + "learning_rate": 0.00013340531276511628, + "loss": 0.4936, + "step": 23625 + }, + { + "epoch": 1.639442092845743, + "grad_norm": 1.294339656829834, + "learning_rate": 0.00013340001686251964, + "loss": 0.514, + "step": 23626 + }, + { + "epoch": 1.6395114842828395, + "grad_norm": 2.3296902179718018, + "learning_rate": 0.000133394720854483, + "loss": 0.9432, + "step": 23627 + }, + { + "epoch": 1.6395808757199362, + "grad_norm": 2.5258407592773438, + "learning_rate": 0.00013338942474102307, + "loss": 0.6417, + "step": 23628 + }, + { + "epoch": 1.6396502671570328, + "grad_norm": 1.477604627609253, + "learning_rate": 0.00013338412852215663, + "loss": 0.6232, + "step": 23629 + }, + { + "epoch": 1.6397196585941294, + "grad_norm": 1.4282697439193726, + "learning_rate": 0.00013337883219790035, + "loss": 0.5926, + "step": 23630 + }, + { + "epoch": 1.6397890500312262, + "grad_norm": 3.423341989517212, + "learning_rate": 0.000133373535768271, + "loss": 0.5462, + "step": 23631 + }, + { + "epoch": 1.6398584414683228, + "grad_norm": 1.2566092014312744, + "learning_rate": 0.00013336823923328526, + "loss": 0.5453, + "step": 23632 + }, + { + "epoch": 1.6399278329054194, + "grad_norm": 1.623766303062439, + "learning_rate": 0.00013336294259295986, + "loss": 0.5461, + "step": 23633 + }, + { + "epoch": 1.6399972243425163, + "grad_norm": 2.626044750213623, + "learning_rate": 0.0001333576458473115, + "loss": 0.8376, + "step": 23634 + }, + { + "epoch": 1.6400666157796127, + "grad_norm": 2.1702828407287598, + "learning_rate": 0.00013335234899635693, + "loss": 0.6056, + "step": 23635 + }, + { + "epoch": 1.6401360072167095, + "grad_norm": 2.3790442943573, + "learning_rate": 0.00013334705204011286, + "loss": 0.8433, + "step": 23636 + }, + { + "epoch": 1.640205398653806, + "grad_norm": 1.2757656574249268, + "learning_rate": 0.00013334175497859603, + "loss": 0.5203, + "step": 23637 + }, + { + "epoch": 1.6402747900909027, + "grad_norm": 1.6558129787445068, + "learning_rate": 0.0001333364578118231, + "loss": 0.4863, + "step": 23638 + }, + { + "epoch": 1.6403441815279995, + "grad_norm": 2.6498022079467773, + "learning_rate": 0.00013333116053981087, + "loss": 0.859, + "step": 23639 + }, + { + "epoch": 1.6404135729650962, + "grad_norm": 2.415659189224243, + "learning_rate": 0.00013332586316257606, + "loss": 0.7, + "step": 23640 + }, + { + "epoch": 1.6404829644021928, + "grad_norm": 1.435288667678833, + "learning_rate": 0.00013332056568013536, + "loss": 0.5772, + "step": 23641 + }, + { + "epoch": 1.6405523558392894, + "grad_norm": 1.8098894357681274, + "learning_rate": 0.00013331526809250548, + "loss": 0.742, + "step": 23642 + }, + { + "epoch": 1.640621747276386, + "grad_norm": 1.6328738927841187, + "learning_rate": 0.00013330997039970319, + "loss": 0.5584, + "step": 23643 + }, + { + "epoch": 1.6406911387134828, + "grad_norm": 2.102989912033081, + "learning_rate": 0.00013330467260174516, + "loss": 0.8408, + "step": 23644 + }, + { + "epoch": 1.6407605301505794, + "grad_norm": 2.5540754795074463, + "learning_rate": 0.00013329937469864814, + "loss": 0.5693, + "step": 23645 + }, + { + "epoch": 1.640829921587676, + "grad_norm": 1.3541003465652466, + "learning_rate": 0.00013329407669042888, + "loss": 0.6391, + "step": 23646 + }, + { + "epoch": 1.6408993130247729, + "grad_norm": 1.8499979972839355, + "learning_rate": 0.00013328877857710406, + "loss": 0.7878, + "step": 23647 + }, + { + "epoch": 1.6409687044618693, + "grad_norm": 1.685048222541809, + "learning_rate": 0.00013328348035869048, + "loss": 0.59, + "step": 23648 + }, + { + "epoch": 1.641038095898966, + "grad_norm": 2.091756582260132, + "learning_rate": 0.00013327818203520477, + "loss": 0.657, + "step": 23649 + }, + { + "epoch": 1.6411074873360627, + "grad_norm": 1.3500996828079224, + "learning_rate": 0.00013327288360666373, + "loss": 0.4037, + "step": 23650 + }, + { + "epoch": 1.6411768787731593, + "grad_norm": 0.9875463843345642, + "learning_rate": 0.000133267585073084, + "loss": 0.5051, + "step": 23651 + }, + { + "epoch": 1.6412462702102562, + "grad_norm": 1.0924105644226074, + "learning_rate": 0.0001332622864344824, + "loss": 0.3621, + "step": 23652 + }, + { + "epoch": 1.6413156616473528, + "grad_norm": 1.3495820760726929, + "learning_rate": 0.00013325698769087562, + "loss": 0.5282, + "step": 23653 + }, + { + "epoch": 1.6413850530844494, + "grad_norm": 1.582168459892273, + "learning_rate": 0.0001332516888422804, + "loss": 0.585, + "step": 23654 + }, + { + "epoch": 1.6414544445215462, + "grad_norm": 1.8521312475204468, + "learning_rate": 0.00013324638988871344, + "loss": 0.674, + "step": 23655 + }, + { + "epoch": 1.6415238359586426, + "grad_norm": 2.362135410308838, + "learning_rate": 0.00013324109083019147, + "loss": 0.5201, + "step": 23656 + }, + { + "epoch": 1.6415932273957394, + "grad_norm": 2.4483559131622314, + "learning_rate": 0.00013323579166673127, + "loss": 0.5623, + "step": 23657 + }, + { + "epoch": 1.641662618832836, + "grad_norm": 1.9384992122650146, + "learning_rate": 0.0001332304923983495, + "loss": 0.6413, + "step": 23658 + }, + { + "epoch": 1.6417320102699327, + "grad_norm": 1.6404863595962524, + "learning_rate": 0.00013322519302506295, + "loss": 0.7759, + "step": 23659 + }, + { + "epoch": 1.6418014017070295, + "grad_norm": 1.7902400493621826, + "learning_rate": 0.0001332198935468883, + "loss": 0.6603, + "step": 23660 + }, + { + "epoch": 1.6418707931441259, + "grad_norm": 2.371317148208618, + "learning_rate": 0.0001332145939638423, + "loss": 0.5274, + "step": 23661 + }, + { + "epoch": 1.6419401845812227, + "grad_norm": 1.7573769092559814, + "learning_rate": 0.00013320929427594166, + "loss": 0.6045, + "step": 23662 + }, + { + "epoch": 1.6420095760183193, + "grad_norm": 1.684506893157959, + "learning_rate": 0.00013320399448320315, + "loss": 0.5865, + "step": 23663 + }, + { + "epoch": 1.642078967455416, + "grad_norm": 1.2220230102539062, + "learning_rate": 0.0001331986945856435, + "loss": 0.4107, + "step": 23664 + }, + { + "epoch": 1.6421483588925128, + "grad_norm": 2.0615310668945312, + "learning_rate": 0.00013319339458327942, + "loss": 0.5132, + "step": 23665 + }, + { + "epoch": 1.6422177503296094, + "grad_norm": 1.1708264350891113, + "learning_rate": 0.00013318809447612764, + "loss": 0.3484, + "step": 23666 + }, + { + "epoch": 1.642287141766706, + "grad_norm": 1.8468894958496094, + "learning_rate": 0.0001331827942642049, + "loss": 0.7647, + "step": 23667 + }, + { + "epoch": 1.6423565332038028, + "grad_norm": 3.375159978866577, + "learning_rate": 0.0001331774939475279, + "loss": 0.644, + "step": 23668 + }, + { + "epoch": 1.6424259246408992, + "grad_norm": 1.7742936611175537, + "learning_rate": 0.00013317219352611343, + "loss": 0.739, + "step": 23669 + }, + { + "epoch": 1.642495316077996, + "grad_norm": 1.4958568811416626, + "learning_rate": 0.00013316689299997818, + "loss": 0.6561, + "step": 23670 + }, + { + "epoch": 1.6425647075150926, + "grad_norm": 1.6323765516281128, + "learning_rate": 0.0001331615923691389, + "loss": 0.5841, + "step": 23671 + }, + { + "epoch": 1.6426340989521893, + "grad_norm": 1.6581146717071533, + "learning_rate": 0.00013315629163361233, + "loss": 0.5684, + "step": 23672 + }, + { + "epoch": 1.642703490389286, + "grad_norm": 1.4101357460021973, + "learning_rate": 0.00013315099079341515, + "loss": 0.5467, + "step": 23673 + }, + { + "epoch": 1.6427728818263825, + "grad_norm": 1.7520909309387207, + "learning_rate": 0.0001331456898485642, + "loss": 0.7347, + "step": 23674 + }, + { + "epoch": 1.6428422732634793, + "grad_norm": 1.7963206768035889, + "learning_rate": 0.00013314038879907612, + "loss": 0.7452, + "step": 23675 + }, + { + "epoch": 1.642911664700576, + "grad_norm": 1.3777422904968262, + "learning_rate": 0.00013313508764496766, + "loss": 0.4858, + "step": 23676 + }, + { + "epoch": 1.6429810561376725, + "grad_norm": 1.34181809425354, + "learning_rate": 0.0001331297863862556, + "loss": 0.466, + "step": 23677 + }, + { + "epoch": 1.6430504475747694, + "grad_norm": 2.277789831161499, + "learning_rate": 0.00013312448502295667, + "loss": 0.9491, + "step": 23678 + }, + { + "epoch": 1.643119839011866, + "grad_norm": 1.7640669345855713, + "learning_rate": 0.00013311918355508754, + "loss": 0.8131, + "step": 23679 + }, + { + "epoch": 1.6431892304489626, + "grad_norm": 1.7412543296813965, + "learning_rate": 0.000133113881982665, + "loss": 0.548, + "step": 23680 + }, + { + "epoch": 1.6432586218860594, + "grad_norm": 1.5296872854232788, + "learning_rate": 0.0001331085803057058, + "loss": 0.6354, + "step": 23681 + }, + { + "epoch": 1.6433280133231558, + "grad_norm": 1.4201916456222534, + "learning_rate": 0.00013310327852422663, + "loss": 0.5743, + "step": 23682 + }, + { + "epoch": 1.6433974047602526, + "grad_norm": 1.563968539237976, + "learning_rate": 0.00013309797663824428, + "loss": 0.6632, + "step": 23683 + }, + { + "epoch": 1.6434667961973493, + "grad_norm": 1.6847119331359863, + "learning_rate": 0.0001330926746477754, + "loss": 0.6024, + "step": 23684 + }, + { + "epoch": 1.6435361876344459, + "grad_norm": 2.1177966594696045, + "learning_rate": 0.00013308737255283686, + "loss": 0.999, + "step": 23685 + }, + { + "epoch": 1.6436055790715427, + "grad_norm": 1.6817044019699097, + "learning_rate": 0.0001330820703534453, + "loss": 0.6142, + "step": 23686 + }, + { + "epoch": 1.643674970508639, + "grad_norm": 1.7452938556671143, + "learning_rate": 0.00013307676804961743, + "loss": 0.7361, + "step": 23687 + }, + { + "epoch": 1.643744361945736, + "grad_norm": 2.0268495082855225, + "learning_rate": 0.0001330714656413701, + "loss": 0.7095, + "step": 23688 + }, + { + "epoch": 1.6438137533828325, + "grad_norm": 1.7123723030090332, + "learning_rate": 0.00013306616312871997, + "loss": 0.5252, + "step": 23689 + }, + { + "epoch": 1.6438831448199291, + "grad_norm": 1.8267430067062378, + "learning_rate": 0.00013306086051168378, + "loss": 0.4445, + "step": 23690 + }, + { + "epoch": 1.643952536257026, + "grad_norm": 1.4252893924713135, + "learning_rate": 0.0001330555577902783, + "loss": 0.5712, + "step": 23691 + }, + { + "epoch": 1.6440219276941226, + "grad_norm": 2.198425531387329, + "learning_rate": 0.0001330502549645203, + "loss": 0.5548, + "step": 23692 + }, + { + "epoch": 1.6440913191312192, + "grad_norm": 2.216125726699829, + "learning_rate": 0.00013304495203442641, + "loss": 0.7875, + "step": 23693 + }, + { + "epoch": 1.644160710568316, + "grad_norm": 2.229553461074829, + "learning_rate": 0.00013303964900001348, + "loss": 0.562, + "step": 23694 + }, + { + "epoch": 1.6442301020054124, + "grad_norm": 1.2944785356521606, + "learning_rate": 0.00013303434586129823, + "loss": 0.6646, + "step": 23695 + }, + { + "epoch": 1.6442994934425093, + "grad_norm": 1.6776549816131592, + "learning_rate": 0.00013302904261829735, + "loss": 0.5992, + "step": 23696 + }, + { + "epoch": 1.6443688848796059, + "grad_norm": 2.1397712230682373, + "learning_rate": 0.00013302373927102763, + "loss": 0.6148, + "step": 23697 + }, + { + "epoch": 1.6444382763167025, + "grad_norm": 1.6475478410720825, + "learning_rate": 0.0001330184358195058, + "loss": 0.5483, + "step": 23698 + }, + { + "epoch": 1.6445076677537993, + "grad_norm": 2.065358877182007, + "learning_rate": 0.0001330131322637486, + "loss": 0.6389, + "step": 23699 + }, + { + "epoch": 1.6445770591908957, + "grad_norm": 1.9807809591293335, + "learning_rate": 0.00013300782860377274, + "loss": 0.545, + "step": 23700 + }, + { + "epoch": 1.6446464506279925, + "grad_norm": 1.2985846996307373, + "learning_rate": 0.000133002524839595, + "loss": 0.431, + "step": 23701 + }, + { + "epoch": 1.6447158420650891, + "grad_norm": 1.6868114471435547, + "learning_rate": 0.00013299722097123218, + "loss": 0.6893, + "step": 23702 + }, + { + "epoch": 1.6447852335021858, + "grad_norm": 2.01151180267334, + "learning_rate": 0.0001329919169987009, + "loss": 0.6865, + "step": 23703 + }, + { + "epoch": 1.6448546249392826, + "grad_norm": 1.1663693189620972, + "learning_rate": 0.00013298661292201797, + "loss": 0.5032, + "step": 23704 + }, + { + "epoch": 1.6449240163763792, + "grad_norm": 1.7075773477554321, + "learning_rate": 0.00013298130874120017, + "loss": 0.6815, + "step": 23705 + }, + { + "epoch": 1.6449934078134758, + "grad_norm": 1.7529295682907104, + "learning_rate": 0.00013297600445626417, + "loss": 0.6848, + "step": 23706 + }, + { + "epoch": 1.6450627992505726, + "grad_norm": 2.0295963287353516, + "learning_rate": 0.00013297070006722673, + "loss": 0.6925, + "step": 23707 + }, + { + "epoch": 1.645132190687669, + "grad_norm": 2.494293451309204, + "learning_rate": 0.00013296539557410467, + "loss": 0.9632, + "step": 23708 + }, + { + "epoch": 1.6452015821247659, + "grad_norm": 2.709901809692383, + "learning_rate": 0.00013296009097691464, + "loss": 0.52, + "step": 23709 + }, + { + "epoch": 1.6452709735618625, + "grad_norm": 1.699501395225525, + "learning_rate": 0.00013295478627567344, + "loss": 0.5874, + "step": 23710 + }, + { + "epoch": 1.645340364998959, + "grad_norm": 1.3807002305984497, + "learning_rate": 0.0001329494814703978, + "loss": 0.4603, + "step": 23711 + }, + { + "epoch": 1.645409756436056, + "grad_norm": 2.0201797485351562, + "learning_rate": 0.0001329441765611045, + "loss": 0.7756, + "step": 23712 + }, + { + "epoch": 1.6454791478731523, + "grad_norm": 1.97903311252594, + "learning_rate": 0.0001329388715478102, + "loss": 0.5537, + "step": 23713 + }, + { + "epoch": 1.6455485393102491, + "grad_norm": 1.4444085359573364, + "learning_rate": 0.00013293356643053175, + "loss": 0.5623, + "step": 23714 + }, + { + "epoch": 1.6456179307473457, + "grad_norm": 2.097867965698242, + "learning_rate": 0.00013292826120928582, + "loss": 0.6605, + "step": 23715 + }, + { + "epoch": 1.6456873221844424, + "grad_norm": 1.5957176685333252, + "learning_rate": 0.0001329229558840892, + "loss": 0.4162, + "step": 23716 + }, + { + "epoch": 1.6457567136215392, + "grad_norm": 1.5969481468200684, + "learning_rate": 0.00013291765045495864, + "loss": 0.5307, + "step": 23717 + }, + { + "epoch": 1.6458261050586358, + "grad_norm": 1.650812029838562, + "learning_rate": 0.0001329123449219109, + "loss": 0.6648, + "step": 23718 + }, + { + "epoch": 1.6458954964957324, + "grad_norm": 1.8815113306045532, + "learning_rate": 0.00013290703928496263, + "loss": 0.6658, + "step": 23719 + }, + { + "epoch": 1.6459648879328292, + "grad_norm": 1.4620158672332764, + "learning_rate": 0.00013290173354413073, + "loss": 0.6009, + "step": 23720 + }, + { + "epoch": 1.6460342793699256, + "grad_norm": 1.9910428524017334, + "learning_rate": 0.00013289642769943186, + "loss": 0.7014, + "step": 23721 + }, + { + "epoch": 1.6461036708070225, + "grad_norm": 1.5706846714019775, + "learning_rate": 0.00013289112175088276, + "loss": 0.5088, + "step": 23722 + }, + { + "epoch": 1.646173062244119, + "grad_norm": 1.363664150238037, + "learning_rate": 0.00013288581569850024, + "loss": 0.4756, + "step": 23723 + }, + { + "epoch": 1.6462424536812157, + "grad_norm": 1.4005707502365112, + "learning_rate": 0.00013288050954230095, + "loss": 0.5893, + "step": 23724 + }, + { + "epoch": 1.6463118451183125, + "grad_norm": 1.4987505674362183, + "learning_rate": 0.00013287520328230174, + "loss": 0.5757, + "step": 23725 + }, + { + "epoch": 1.646381236555409, + "grad_norm": 1.3234584331512451, + "learning_rate": 0.00013286989691851936, + "loss": 0.4182, + "step": 23726 + }, + { + "epoch": 1.6464506279925057, + "grad_norm": 1.385455846786499, + "learning_rate": 0.00013286459045097049, + "loss": 0.4617, + "step": 23727 + }, + { + "epoch": 1.6465200194296024, + "grad_norm": 1.9271644353866577, + "learning_rate": 0.00013285928387967192, + "loss": 0.6701, + "step": 23728 + }, + { + "epoch": 1.646589410866699, + "grad_norm": 1.6604474782943726, + "learning_rate": 0.00013285397720464045, + "loss": 0.7183, + "step": 23729 + }, + { + "epoch": 1.6466588023037958, + "grad_norm": 1.3766366243362427, + "learning_rate": 0.00013284867042589275, + "loss": 0.5207, + "step": 23730 + }, + { + "epoch": 1.6467281937408924, + "grad_norm": 2.1081783771514893, + "learning_rate": 0.0001328433635434456, + "loss": 0.712, + "step": 23731 + }, + { + "epoch": 1.646797585177989, + "grad_norm": 1.7087900638580322, + "learning_rate": 0.00013283805655731575, + "loss": 0.5701, + "step": 23732 + }, + { + "epoch": 1.6468669766150859, + "grad_norm": 1.4098891019821167, + "learning_rate": 0.00013283274946752002, + "loss": 0.5709, + "step": 23733 + }, + { + "epoch": 1.6469363680521822, + "grad_norm": 1.2601011991500854, + "learning_rate": 0.00013282744227407508, + "loss": 0.582, + "step": 23734 + }, + { + "epoch": 1.647005759489279, + "grad_norm": 2.1709482669830322, + "learning_rate": 0.0001328221349769977, + "loss": 0.6457, + "step": 23735 + }, + { + "epoch": 1.6470751509263757, + "grad_norm": 2.0014548301696777, + "learning_rate": 0.00013281682757630464, + "loss": 0.456, + "step": 23736 + }, + { + "epoch": 1.6471445423634723, + "grad_norm": 1.8725837469100952, + "learning_rate": 0.00013281152007201273, + "loss": 0.6138, + "step": 23737 + }, + { + "epoch": 1.6472139338005691, + "grad_norm": 1.8646504878997803, + "learning_rate": 0.00013280621246413858, + "loss": 0.6671, + "step": 23738 + }, + { + "epoch": 1.6472833252376655, + "grad_norm": 1.6937737464904785, + "learning_rate": 0.00013280090475269906, + "loss": 0.6174, + "step": 23739 + }, + { + "epoch": 1.6473527166747624, + "grad_norm": 1.2383984327316284, + "learning_rate": 0.00013279559693771088, + "loss": 0.6341, + "step": 23740 + }, + { + "epoch": 1.647422108111859, + "grad_norm": 1.887049674987793, + "learning_rate": 0.0001327902890191908, + "loss": 0.6299, + "step": 23741 + }, + { + "epoch": 1.6474914995489556, + "grad_norm": 1.572691798210144, + "learning_rate": 0.0001327849809971556, + "loss": 0.6517, + "step": 23742 + }, + { + "epoch": 1.6475608909860524, + "grad_norm": 1.6509859561920166, + "learning_rate": 0.000132779672871622, + "loss": 0.6404, + "step": 23743 + }, + { + "epoch": 1.647630282423149, + "grad_norm": 3.947401285171509, + "learning_rate": 0.0001327743646426068, + "loss": 0.4422, + "step": 23744 + }, + { + "epoch": 1.6476996738602456, + "grad_norm": 1.8237138986587524, + "learning_rate": 0.0001327690563101267, + "loss": 0.6411, + "step": 23745 + }, + { + "epoch": 1.6477690652973425, + "grad_norm": 1.5466396808624268, + "learning_rate": 0.00013276374787419849, + "loss": 0.4783, + "step": 23746 + }, + { + "epoch": 1.6478384567344389, + "grad_norm": 1.781254768371582, + "learning_rate": 0.00013275843933483898, + "loss": 0.461, + "step": 23747 + }, + { + "epoch": 1.6479078481715357, + "grad_norm": 1.4462640285491943, + "learning_rate": 0.00013275313069206486, + "loss": 0.4816, + "step": 23748 + }, + { + "epoch": 1.6479772396086323, + "grad_norm": 1.3955843448638916, + "learning_rate": 0.00013274782194589287, + "loss": 0.6369, + "step": 23749 + }, + { + "epoch": 1.648046631045729, + "grad_norm": 1.4463515281677246, + "learning_rate": 0.00013274251309633981, + "loss": 0.4863, + "step": 23750 + }, + { + "epoch": 1.6481160224828257, + "grad_norm": 1.886420726776123, + "learning_rate": 0.00013273720414342248, + "loss": 0.7818, + "step": 23751 + }, + { + "epoch": 1.6481854139199221, + "grad_norm": 1.5992789268493652, + "learning_rate": 0.00013273189508715755, + "loss": 0.4361, + "step": 23752 + }, + { + "epoch": 1.648254805357019, + "grad_norm": 1.487708330154419, + "learning_rate": 0.00013272658592756187, + "loss": 0.6248, + "step": 23753 + }, + { + "epoch": 1.6483241967941156, + "grad_norm": 1.754732608795166, + "learning_rate": 0.00013272127666465213, + "loss": 0.4162, + "step": 23754 + }, + { + "epoch": 1.6483935882312122, + "grad_norm": 1.5113343000411987, + "learning_rate": 0.0001327159672984451, + "loss": 0.4354, + "step": 23755 + }, + { + "epoch": 1.648462979668309, + "grad_norm": 1.818299651145935, + "learning_rate": 0.00013271065782895756, + "loss": 0.5246, + "step": 23756 + }, + { + "epoch": 1.6485323711054056, + "grad_norm": 2.4535727500915527, + "learning_rate": 0.00013270534825620631, + "loss": 0.6225, + "step": 23757 + }, + { + "epoch": 1.6486017625425022, + "grad_norm": 1.8833138942718506, + "learning_rate": 0.00013270003858020804, + "loss": 0.6757, + "step": 23758 + }, + { + "epoch": 1.648671153979599, + "grad_norm": 1.4360601902008057, + "learning_rate": 0.00013269472880097954, + "loss": 0.5938, + "step": 23759 + }, + { + "epoch": 1.6487405454166955, + "grad_norm": 1.5958294868469238, + "learning_rate": 0.00013268941891853757, + "loss": 0.6958, + "step": 23760 + }, + { + "epoch": 1.6488099368537923, + "grad_norm": 1.970930576324463, + "learning_rate": 0.0001326841089328989, + "loss": 0.6719, + "step": 23761 + }, + { + "epoch": 1.648879328290889, + "grad_norm": 1.841846227645874, + "learning_rate": 0.0001326787988440803, + "loss": 0.5256, + "step": 23762 + }, + { + "epoch": 1.6489487197279855, + "grad_norm": 1.6526238918304443, + "learning_rate": 0.0001326734886520985, + "loss": 0.4424, + "step": 23763 + }, + { + "epoch": 1.6490181111650823, + "grad_norm": 1.8014577627182007, + "learning_rate": 0.00013266817835697032, + "loss": 0.4441, + "step": 23764 + }, + { + "epoch": 1.6490875026021787, + "grad_norm": 1.3727118968963623, + "learning_rate": 0.00013266286795871247, + "loss": 0.4014, + "step": 23765 + }, + { + "epoch": 1.6491568940392756, + "grad_norm": 1.4222558736801147, + "learning_rate": 0.0001326575574573417, + "loss": 0.4162, + "step": 23766 + }, + { + "epoch": 1.6492262854763722, + "grad_norm": 1.8559075593948364, + "learning_rate": 0.00013265224685287487, + "loss": 0.7281, + "step": 23767 + }, + { + "epoch": 1.6492956769134688, + "grad_norm": 1.3594496250152588, + "learning_rate": 0.00013264693614532867, + "loss": 0.464, + "step": 23768 + }, + { + "epoch": 1.6493650683505656, + "grad_norm": 1.3379734754562378, + "learning_rate": 0.0001326416253347199, + "loss": 0.3811, + "step": 23769 + }, + { + "epoch": 1.6494344597876622, + "grad_norm": 2.163691997528076, + "learning_rate": 0.00013263631442106527, + "loss": 0.4125, + "step": 23770 + }, + { + "epoch": 1.6495038512247588, + "grad_norm": 0.9978824257850647, + "learning_rate": 0.0001326310034043816, + "loss": 0.3277, + "step": 23771 + }, + { + "epoch": 1.6495732426618557, + "grad_norm": 1.425901174545288, + "learning_rate": 0.00013262569228468564, + "loss": 0.5782, + "step": 23772 + }, + { + "epoch": 1.649642634098952, + "grad_norm": 1.906529188156128, + "learning_rate": 0.00013262038106199415, + "loss": 0.6786, + "step": 23773 + }, + { + "epoch": 1.649712025536049, + "grad_norm": 1.8478964567184448, + "learning_rate": 0.00013261506973632388, + "loss": 0.6461, + "step": 23774 + }, + { + "epoch": 1.6497814169731455, + "grad_norm": 1.548516035079956, + "learning_rate": 0.00013260975830769165, + "loss": 0.3914, + "step": 23775 + }, + { + "epoch": 1.6498508084102421, + "grad_norm": 1.6230427026748657, + "learning_rate": 0.00013260444677611419, + "loss": 0.5964, + "step": 23776 + }, + { + "epoch": 1.649920199847339, + "grad_norm": 1.4592682123184204, + "learning_rate": 0.00013259913514160825, + "loss": 0.5907, + "step": 23777 + }, + { + "epoch": 1.6499895912844353, + "grad_norm": 2.3593947887420654, + "learning_rate": 0.00013259382340419067, + "loss": 0.7034, + "step": 23778 + }, + { + "epoch": 1.6500589827215322, + "grad_norm": 1.7138781547546387, + "learning_rate": 0.00013258851156387812, + "loss": 0.5541, + "step": 23779 + }, + { + "epoch": 1.6501283741586288, + "grad_norm": 1.6733957529067993, + "learning_rate": 0.00013258319962068744, + "loss": 0.5611, + "step": 23780 + }, + { + "epoch": 1.6501977655957254, + "grad_norm": 1.2321100234985352, + "learning_rate": 0.00013257788757463539, + "loss": 0.4292, + "step": 23781 + }, + { + "epoch": 1.6502671570328222, + "grad_norm": 6.373149394989014, + "learning_rate": 0.0001325725754257387, + "loss": 0.8006, + "step": 23782 + }, + { + "epoch": 1.6503365484699188, + "grad_norm": 1.4346134662628174, + "learning_rate": 0.0001325672631740142, + "loss": 0.4376, + "step": 23783 + }, + { + "epoch": 1.6504059399070155, + "grad_norm": 2.4258337020874023, + "learning_rate": 0.0001325619508194786, + "loss": 0.4835, + "step": 23784 + }, + { + "epoch": 1.6504753313441123, + "grad_norm": 1.5529050827026367, + "learning_rate": 0.00013255663836214874, + "loss": 0.51, + "step": 23785 + }, + { + "epoch": 1.6505447227812087, + "grad_norm": 1.9793829917907715, + "learning_rate": 0.0001325513258020413, + "loss": 0.7814, + "step": 23786 + }, + { + "epoch": 1.6506141142183055, + "grad_norm": 1.7314478158950806, + "learning_rate": 0.00013254601313917314, + "loss": 0.5388, + "step": 23787 + }, + { + "epoch": 1.6506835056554021, + "grad_norm": 2.0976264476776123, + "learning_rate": 0.000132540700373561, + "loss": 0.6329, + "step": 23788 + }, + { + "epoch": 1.6507528970924987, + "grad_norm": 1.7474833726882935, + "learning_rate": 0.00013253538750522163, + "loss": 0.4114, + "step": 23789 + }, + { + "epoch": 1.6508222885295956, + "grad_norm": 2.1133816242218018, + "learning_rate": 0.00013253007453417178, + "loss": 0.7155, + "step": 23790 + }, + { + "epoch": 1.650891679966692, + "grad_norm": 1.4324995279312134, + "learning_rate": 0.0001325247614604283, + "loss": 0.5087, + "step": 23791 + }, + { + "epoch": 1.6509610714037888, + "grad_norm": 1.8224561214447021, + "learning_rate": 0.00013251944828400792, + "loss": 0.6444, + "step": 23792 + }, + { + "epoch": 1.6510304628408854, + "grad_norm": 2.028215169906616, + "learning_rate": 0.0001325141350049274, + "loss": 0.818, + "step": 23793 + }, + { + "epoch": 1.651099854277982, + "grad_norm": 1.6549283266067505, + "learning_rate": 0.00013250882162320354, + "loss": 0.64, + "step": 23794 + }, + { + "epoch": 1.6511692457150788, + "grad_norm": 1.2372370958328247, + "learning_rate": 0.00013250350813885309, + "loss": 0.3435, + "step": 23795 + }, + { + "epoch": 1.6512386371521754, + "grad_norm": 2.2501559257507324, + "learning_rate": 0.00013249819455189282, + "loss": 0.7645, + "step": 23796 + }, + { + "epoch": 1.651308028589272, + "grad_norm": 1.5015957355499268, + "learning_rate": 0.00013249288086233956, + "loss": 0.5313, + "step": 23797 + }, + { + "epoch": 1.651377420026369, + "grad_norm": 1.7719100713729858, + "learning_rate": 0.00013248756707021002, + "loss": 0.7678, + "step": 23798 + }, + { + "epoch": 1.6514468114634653, + "grad_norm": 1.5716923475265503, + "learning_rate": 0.000132482253175521, + "loss": 0.4972, + "step": 23799 + }, + { + "epoch": 1.6515162029005621, + "grad_norm": 1.8218060731887817, + "learning_rate": 0.00013247693917828928, + "loss": 0.7455, + "step": 23800 + }, + { + "epoch": 1.6515855943376587, + "grad_norm": 2.073258399963379, + "learning_rate": 0.00013247162507853163, + "loss": 0.6007, + "step": 23801 + }, + { + "epoch": 1.6516549857747553, + "grad_norm": 1.9128667116165161, + "learning_rate": 0.00013246631087626484, + "loss": 0.6233, + "step": 23802 + }, + { + "epoch": 1.6517243772118522, + "grad_norm": 2.023937225341797, + "learning_rate": 0.00013246099657150565, + "loss": 0.6173, + "step": 23803 + }, + { + "epoch": 1.6517937686489486, + "grad_norm": 1.4393447637557983, + "learning_rate": 0.0001324556821642709, + "loss": 0.5255, + "step": 23804 + }, + { + "epoch": 1.6518631600860454, + "grad_norm": 2.042616844177246, + "learning_rate": 0.00013245036765457727, + "loss": 0.8191, + "step": 23805 + }, + { + "epoch": 1.651932551523142, + "grad_norm": 1.7451659440994263, + "learning_rate": 0.00013244505304244164, + "loss": 0.5556, + "step": 23806 + }, + { + "epoch": 1.6520019429602386, + "grad_norm": 2.101402521133423, + "learning_rate": 0.00013243973832788075, + "loss": 0.9107, + "step": 23807 + }, + { + "epoch": 1.6520713343973354, + "grad_norm": 1.4426310062408447, + "learning_rate": 0.00013243442351091133, + "loss": 0.4017, + "step": 23808 + }, + { + "epoch": 1.652140725834432, + "grad_norm": 1.3946493864059448, + "learning_rate": 0.0001324291085915502, + "loss": 0.5398, + "step": 23809 + }, + { + "epoch": 1.6522101172715287, + "grad_norm": 1.7952152490615845, + "learning_rate": 0.0001324237935698142, + "loss": 0.6207, + "step": 23810 + }, + { + "epoch": 1.6522795087086255, + "grad_norm": 1.5983861684799194, + "learning_rate": 0.00013241847844571995, + "loss": 0.6302, + "step": 23811 + }, + { + "epoch": 1.6523489001457219, + "grad_norm": 1.7696807384490967, + "learning_rate": 0.0001324131632192844, + "loss": 0.5997, + "step": 23812 + }, + { + "epoch": 1.6524182915828187, + "grad_norm": 2.3088598251342773, + "learning_rate": 0.0001324078478905242, + "loss": 0.6807, + "step": 23813 + }, + { + "epoch": 1.6524876830199153, + "grad_norm": 1.5406198501586914, + "learning_rate": 0.00013240253245945622, + "loss": 0.5353, + "step": 23814 + }, + { + "epoch": 1.652557074457012, + "grad_norm": 1.9688552618026733, + "learning_rate": 0.0001323972169260972, + "loss": 0.7514, + "step": 23815 + }, + { + "epoch": 1.6526264658941088, + "grad_norm": 1.813569188117981, + "learning_rate": 0.00013239190129046393, + "loss": 0.5508, + "step": 23816 + }, + { + "epoch": 1.6526958573312052, + "grad_norm": 1.9540339708328247, + "learning_rate": 0.00013238658555257316, + "loss": 0.5625, + "step": 23817 + }, + { + "epoch": 1.652765248768302, + "grad_norm": 1.945834994316101, + "learning_rate": 0.00013238126971244172, + "loss": 0.533, + "step": 23818 + }, + { + "epoch": 1.6528346402053986, + "grad_norm": 1.779441237449646, + "learning_rate": 0.00013237595377008635, + "loss": 0.6069, + "step": 23819 + }, + { + "epoch": 1.6529040316424952, + "grad_norm": 1.7537338733673096, + "learning_rate": 0.00013237063772552387, + "loss": 0.6481, + "step": 23820 + }, + { + "epoch": 1.652973423079592, + "grad_norm": 2.1212875843048096, + "learning_rate": 0.00013236532157877103, + "loss": 0.8653, + "step": 23821 + }, + { + "epoch": 1.6530428145166887, + "grad_norm": 2.3980462551116943, + "learning_rate": 0.00013236000532984462, + "loss": 0.7871, + "step": 23822 + }, + { + "epoch": 1.6531122059537853, + "grad_norm": 1.4729622602462769, + "learning_rate": 0.0001323546889787614, + "loss": 0.5043, + "step": 23823 + }, + { + "epoch": 1.653181597390882, + "grad_norm": 1.1514679193496704, + "learning_rate": 0.00013234937252553825, + "loss": 0.4148, + "step": 23824 + }, + { + "epoch": 1.6532509888279785, + "grad_norm": 2.1204540729522705, + "learning_rate": 0.00013234405597019187, + "loss": 0.7935, + "step": 23825 + }, + { + "epoch": 1.6533203802650753, + "grad_norm": 1.4545564651489258, + "learning_rate": 0.00013233873931273903, + "loss": 0.5778, + "step": 23826 + }, + { + "epoch": 1.653389771702172, + "grad_norm": 1.7093992233276367, + "learning_rate": 0.00013233342255319656, + "loss": 0.6566, + "step": 23827 + }, + { + "epoch": 1.6534591631392686, + "grad_norm": 1.9777553081512451, + "learning_rate": 0.00013232810569158122, + "loss": 0.7327, + "step": 23828 + }, + { + "epoch": 1.6535285545763654, + "grad_norm": 1.5636389255523682, + "learning_rate": 0.00013232278872790976, + "loss": 0.6977, + "step": 23829 + }, + { + "epoch": 1.6535979460134618, + "grad_norm": 1.5702658891677856, + "learning_rate": 0.00013231747166219908, + "loss": 0.5499, + "step": 23830 + }, + { + "epoch": 1.6536673374505586, + "grad_norm": 1.4423537254333496, + "learning_rate": 0.00013231215449446582, + "loss": 0.6326, + "step": 23831 + }, + { + "epoch": 1.6537367288876552, + "grad_norm": 1.6726691722869873, + "learning_rate": 0.0001323068372247269, + "loss": 0.5814, + "step": 23832 + }, + { + "epoch": 1.6538061203247518, + "grad_norm": 1.5016441345214844, + "learning_rate": 0.000132301519852999, + "loss": 0.5232, + "step": 23833 + }, + { + "epoch": 1.6538755117618487, + "grad_norm": 1.8929109573364258, + "learning_rate": 0.00013229620237929898, + "loss": 0.6531, + "step": 23834 + }, + { + "epoch": 1.6539449031989453, + "grad_norm": 1.708269476890564, + "learning_rate": 0.00013229088480364357, + "loss": 0.576, + "step": 23835 + }, + { + "epoch": 1.6540142946360419, + "grad_norm": 1.554236650466919, + "learning_rate": 0.00013228556712604957, + "loss": 0.5534, + "step": 23836 + }, + { + "epoch": 1.6540836860731387, + "grad_norm": 1.692251443862915, + "learning_rate": 0.00013228024934653382, + "loss": 0.7388, + "step": 23837 + }, + { + "epoch": 1.654153077510235, + "grad_norm": 1.6772472858428955, + "learning_rate": 0.00013227493146511306, + "loss": 0.6218, + "step": 23838 + }, + { + "epoch": 1.654222468947332, + "grad_norm": 2.1549901962280273, + "learning_rate": 0.00013226961348180406, + "loss": 0.8313, + "step": 23839 + }, + { + "epoch": 1.6542918603844285, + "grad_norm": 1.4675592184066772, + "learning_rate": 0.00013226429539662369, + "loss": 0.5989, + "step": 23840 + }, + { + "epoch": 1.6543612518215252, + "grad_norm": 1.7777743339538574, + "learning_rate": 0.00013225897720958865, + "loss": 0.6249, + "step": 23841 + }, + { + "epoch": 1.654430643258622, + "grad_norm": 1.1917788982391357, + "learning_rate": 0.00013225365892071574, + "loss": 0.5428, + "step": 23842 + }, + { + "epoch": 1.6545000346957186, + "grad_norm": 1.9772104024887085, + "learning_rate": 0.0001322483405300218, + "loss": 0.6901, + "step": 23843 + }, + { + "epoch": 1.6545694261328152, + "grad_norm": 1.7322767972946167, + "learning_rate": 0.0001322430220375236, + "loss": 0.5162, + "step": 23844 + }, + { + "epoch": 1.6546388175699118, + "grad_norm": 1.3641941547393799, + "learning_rate": 0.00013223770344323788, + "loss": 0.4774, + "step": 23845 + }, + { + "epoch": 1.6547082090070084, + "grad_norm": 1.6829450130462646, + "learning_rate": 0.00013223238474718147, + "loss": 0.5255, + "step": 23846 + }, + { + "epoch": 1.6547776004441053, + "grad_norm": 1.695811152458191, + "learning_rate": 0.0001322270659493712, + "loss": 0.5718, + "step": 23847 + }, + { + "epoch": 1.6548469918812019, + "grad_norm": 1.509883999824524, + "learning_rate": 0.0001322217470498238, + "loss": 0.7749, + "step": 23848 + }, + { + "epoch": 1.6549163833182985, + "grad_norm": 2.1044657230377197, + "learning_rate": 0.00013221642804855607, + "loss": 0.7112, + "step": 23849 + }, + { + "epoch": 1.6549857747553953, + "grad_norm": 1.5656399726867676, + "learning_rate": 0.00013221110894558487, + "loss": 0.6539, + "step": 23850 + }, + { + "epoch": 1.6550551661924917, + "grad_norm": 1.4656776189804077, + "learning_rate": 0.00013220578974092688, + "loss": 0.519, + "step": 23851 + }, + { + "epoch": 1.6551245576295885, + "grad_norm": 1.4013644456863403, + "learning_rate": 0.00013220047043459892, + "loss": 0.4652, + "step": 23852 + }, + { + "epoch": 1.6551939490666852, + "grad_norm": 1.280583143234253, + "learning_rate": 0.00013219515102661788, + "loss": 0.4818, + "step": 23853 + }, + { + "epoch": 1.6552633405037818, + "grad_norm": 2.118868112564087, + "learning_rate": 0.00013218983151700047, + "loss": 0.5946, + "step": 23854 + }, + { + "epoch": 1.6553327319408786, + "grad_norm": 1.6367932558059692, + "learning_rate": 0.00013218451190576347, + "loss": 0.5484, + "step": 23855 + }, + { + "epoch": 1.6554021233779752, + "grad_norm": 1.8219952583312988, + "learning_rate": 0.00013217919219292372, + "loss": 0.7368, + "step": 23856 + }, + { + "epoch": 1.6554715148150718, + "grad_norm": 1.8886069059371948, + "learning_rate": 0.000132173872378498, + "loss": 0.6001, + "step": 23857 + }, + { + "epoch": 1.6555409062521684, + "grad_norm": 1.6623520851135254, + "learning_rate": 0.00013216855246250307, + "loss": 0.6834, + "step": 23858 + }, + { + "epoch": 1.655610297689265, + "grad_norm": 1.7433459758758545, + "learning_rate": 0.00013216323244495578, + "loss": 0.5976, + "step": 23859 + }, + { + "epoch": 1.6556796891263619, + "grad_norm": 1.4793651103973389, + "learning_rate": 0.00013215791232587288, + "loss": 0.675, + "step": 23860 + }, + { + "epoch": 1.6557490805634585, + "grad_norm": 1.221074104309082, + "learning_rate": 0.00013215259210527122, + "loss": 0.4652, + "step": 23861 + }, + { + "epoch": 1.655818472000555, + "grad_norm": 1.432253360748291, + "learning_rate": 0.0001321472717831675, + "loss": 0.4732, + "step": 23862 + }, + { + "epoch": 1.655887863437652, + "grad_norm": 1.3323076963424683, + "learning_rate": 0.0001321419513595786, + "loss": 0.5886, + "step": 23863 + }, + { + "epoch": 1.6559572548747483, + "grad_norm": 1.6635867357254028, + "learning_rate": 0.0001321366308345213, + "loss": 0.6055, + "step": 23864 + }, + { + "epoch": 1.6560266463118452, + "grad_norm": 2.60129976272583, + "learning_rate": 0.00013213131020801234, + "loss": 0.7169, + "step": 23865 + }, + { + "epoch": 1.6560960377489418, + "grad_norm": 1.6061029434204102, + "learning_rate": 0.0001321259894800686, + "loss": 0.5746, + "step": 23866 + }, + { + "epoch": 1.6561654291860384, + "grad_norm": 1.3862429857254028, + "learning_rate": 0.00013212066865070683, + "loss": 0.5723, + "step": 23867 + }, + { + "epoch": 1.6562348206231352, + "grad_norm": 1.7060405015945435, + "learning_rate": 0.00013211534771994385, + "loss": 0.7532, + "step": 23868 + }, + { + "epoch": 1.6563042120602318, + "grad_norm": 1.2327195405960083, + "learning_rate": 0.00013211002668779643, + "loss": 0.5566, + "step": 23869 + }, + { + "epoch": 1.6563736034973284, + "grad_norm": 1.2944694757461548, + "learning_rate": 0.00013210470555428136, + "loss": 0.5284, + "step": 23870 + }, + { + "epoch": 1.656442994934425, + "grad_norm": 1.5303080081939697, + "learning_rate": 0.0001320993843194155, + "loss": 0.5913, + "step": 23871 + }, + { + "epoch": 1.6565123863715217, + "grad_norm": 1.5670852661132812, + "learning_rate": 0.0001320940629832156, + "loss": 0.6948, + "step": 23872 + }, + { + "epoch": 1.6565817778086185, + "grad_norm": 1.6460881233215332, + "learning_rate": 0.00013208874154569845, + "loss": 0.6598, + "step": 23873 + }, + { + "epoch": 1.656651169245715, + "grad_norm": 1.721426010131836, + "learning_rate": 0.00013208342000688085, + "loss": 0.461, + "step": 23874 + }, + { + "epoch": 1.6567205606828117, + "grad_norm": 1.7694733142852783, + "learning_rate": 0.00013207809836677964, + "loss": 0.706, + "step": 23875 + }, + { + "epoch": 1.6567899521199085, + "grad_norm": 1.6760562658309937, + "learning_rate": 0.0001320727766254116, + "loss": 0.8044, + "step": 23876 + }, + { + "epoch": 1.656859343557005, + "grad_norm": 1.7753732204437256, + "learning_rate": 0.00013206745478279354, + "loss": 0.5705, + "step": 23877 + }, + { + "epoch": 1.6569287349941018, + "grad_norm": 1.3488985300064087, + "learning_rate": 0.00013206213283894222, + "loss": 0.561, + "step": 23878 + }, + { + "epoch": 1.6569981264311984, + "grad_norm": 1.9091801643371582, + "learning_rate": 0.00013205681079387447, + "loss": 0.6722, + "step": 23879 + }, + { + "epoch": 1.657067517868295, + "grad_norm": 1.3249449729919434, + "learning_rate": 0.00013205148864760708, + "loss": 0.5069, + "step": 23880 + }, + { + "epoch": 1.6571369093053918, + "grad_norm": 1.4021048545837402, + "learning_rate": 0.00013204616640015687, + "loss": 0.5864, + "step": 23881 + }, + { + "epoch": 1.6572063007424884, + "grad_norm": 2.032705307006836, + "learning_rate": 0.00013204084405154062, + "loss": 0.5619, + "step": 23882 + }, + { + "epoch": 1.657275692179585, + "grad_norm": 1.577326774597168, + "learning_rate": 0.00013203552160177514, + "loss": 0.5121, + "step": 23883 + }, + { + "epoch": 1.6573450836166816, + "grad_norm": 1.6632620096206665, + "learning_rate": 0.00013203019905087726, + "loss": 0.7798, + "step": 23884 + }, + { + "epoch": 1.6574144750537783, + "grad_norm": 1.861167073249817, + "learning_rate": 0.00013202487639886373, + "loss": 0.6055, + "step": 23885 + }, + { + "epoch": 1.657483866490875, + "grad_norm": 1.6549487113952637, + "learning_rate": 0.0001320195536457514, + "loss": 0.4956, + "step": 23886 + }, + { + "epoch": 1.6575532579279717, + "grad_norm": 1.5397361516952515, + "learning_rate": 0.00013201423079155705, + "loss": 0.6092, + "step": 23887 + }, + { + "epoch": 1.6576226493650683, + "grad_norm": 1.5721442699432373, + "learning_rate": 0.00013200890783629748, + "loss": 0.4286, + "step": 23888 + }, + { + "epoch": 1.6576920408021651, + "grad_norm": 1.6804792881011963, + "learning_rate": 0.00013200358477998952, + "loss": 0.5642, + "step": 23889 + }, + { + "epoch": 1.6577614322392615, + "grad_norm": 1.2572561502456665, + "learning_rate": 0.00013199826162264993, + "loss": 0.4926, + "step": 23890 + }, + { + "epoch": 1.6578308236763584, + "grad_norm": 1.8660155534744263, + "learning_rate": 0.00013199293836429555, + "loss": 0.5731, + "step": 23891 + }, + { + "epoch": 1.657900215113455, + "grad_norm": 1.7242519855499268, + "learning_rate": 0.00013198761500494318, + "loss": 0.4686, + "step": 23892 + }, + { + "epoch": 1.6579696065505516, + "grad_norm": 1.7337547540664673, + "learning_rate": 0.0001319822915446096, + "loss": 0.479, + "step": 23893 + }, + { + "epoch": 1.6580389979876484, + "grad_norm": 1.8897260427474976, + "learning_rate": 0.00013197696798331168, + "loss": 0.6672, + "step": 23894 + }, + { + "epoch": 1.658108389424745, + "grad_norm": 2.4131886959075928, + "learning_rate": 0.00013197164432106616, + "loss": 0.533, + "step": 23895 + }, + { + "epoch": 1.6581777808618416, + "grad_norm": 1.4848527908325195, + "learning_rate": 0.0001319663205578899, + "loss": 0.6126, + "step": 23896 + }, + { + "epoch": 1.6582471722989385, + "grad_norm": 2.027613401412964, + "learning_rate": 0.00013196099669379963, + "loss": 0.627, + "step": 23897 + }, + { + "epoch": 1.6583165637360349, + "grad_norm": 1.3672935962677002, + "learning_rate": 0.0001319556727288122, + "loss": 0.5392, + "step": 23898 + }, + { + "epoch": 1.6583859551731317, + "grad_norm": 2.325809955596924, + "learning_rate": 0.00013195034866294445, + "loss": 0.59, + "step": 23899 + }, + { + "epoch": 1.6584553466102283, + "grad_norm": 1.7370928525924683, + "learning_rate": 0.00013194502449621312, + "loss": 0.6427, + "step": 23900 + }, + { + "epoch": 1.658524738047325, + "grad_norm": 1.600807547569275, + "learning_rate": 0.00013193970022863506, + "loss": 0.613, + "step": 23901 + }, + { + "epoch": 1.6585941294844218, + "grad_norm": 1.5080262422561646, + "learning_rate": 0.0001319343758602271, + "loss": 0.6182, + "step": 23902 + }, + { + "epoch": 1.6586635209215181, + "grad_norm": 1.3367079496383667, + "learning_rate": 0.000131929051391006, + "loss": 0.4376, + "step": 23903 + }, + { + "epoch": 1.658732912358615, + "grad_norm": 2.3429250717163086, + "learning_rate": 0.00013192372682098858, + "loss": 0.7085, + "step": 23904 + }, + { + "epoch": 1.6588023037957116, + "grad_norm": 1.2716894149780273, + "learning_rate": 0.00013191840215019167, + "loss": 0.4774, + "step": 23905 + }, + { + "epoch": 1.6588716952328082, + "grad_norm": 1.5021332502365112, + "learning_rate": 0.0001319130773786321, + "loss": 0.6412, + "step": 23906 + }, + { + "epoch": 1.658941086669905, + "grad_norm": 1.175902247428894, + "learning_rate": 0.00013190775250632657, + "loss": 0.459, + "step": 23907 + }, + { + "epoch": 1.6590104781070016, + "grad_norm": 1.7548106908798218, + "learning_rate": 0.00013190242753329204, + "loss": 0.6806, + "step": 23908 + }, + { + "epoch": 1.6590798695440983, + "grad_norm": 1.7919052839279175, + "learning_rate": 0.00013189710245954523, + "loss": 0.7327, + "step": 23909 + }, + { + "epoch": 1.659149260981195, + "grad_norm": 1.4928432703018188, + "learning_rate": 0.00013189177728510293, + "loss": 0.3892, + "step": 23910 + }, + { + "epoch": 1.6592186524182915, + "grad_norm": 2.569017171859741, + "learning_rate": 0.00013188645200998205, + "loss": 1.0437, + "step": 23911 + }, + { + "epoch": 1.6592880438553883, + "grad_norm": 2.289827346801758, + "learning_rate": 0.0001318811266341993, + "loss": 0.6356, + "step": 23912 + }, + { + "epoch": 1.659357435292485, + "grad_norm": 2.0369627475738525, + "learning_rate": 0.00013187580115777157, + "loss": 0.7073, + "step": 23913 + }, + { + "epoch": 1.6594268267295815, + "grad_norm": 1.198941946029663, + "learning_rate": 0.0001318704755807156, + "loss": 0.4441, + "step": 23914 + }, + { + "epoch": 1.6594962181666784, + "grad_norm": 1.2555902004241943, + "learning_rate": 0.00013186514990304823, + "loss": 0.6601, + "step": 23915 + }, + { + "epoch": 1.6595656096037747, + "grad_norm": 1.813205361366272, + "learning_rate": 0.0001318598241247863, + "loss": 0.8792, + "step": 23916 + }, + { + "epoch": 1.6596350010408716, + "grad_norm": 1.7112349271774292, + "learning_rate": 0.00013185449824594658, + "loss": 0.7849, + "step": 23917 + }, + { + "epoch": 1.6597043924779682, + "grad_norm": 1.1793146133422852, + "learning_rate": 0.0001318491722665459, + "loss": 0.5309, + "step": 23918 + }, + { + "epoch": 1.6597737839150648, + "grad_norm": 1.4072993993759155, + "learning_rate": 0.0001318438461866011, + "loss": 0.4159, + "step": 23919 + }, + { + "epoch": 1.6598431753521616, + "grad_norm": 1.3151568174362183, + "learning_rate": 0.00013183852000612898, + "loss": 0.4883, + "step": 23920 + }, + { + "epoch": 1.6599125667892582, + "grad_norm": 1.7973356246948242, + "learning_rate": 0.00013183319372514634, + "loss": 0.6396, + "step": 23921 + }, + { + "epoch": 1.6599819582263549, + "grad_norm": 1.6288028955459595, + "learning_rate": 0.00013182786734366997, + "loss": 0.6356, + "step": 23922 + }, + { + "epoch": 1.6600513496634517, + "grad_norm": 2.091205358505249, + "learning_rate": 0.00013182254086171677, + "loss": 0.5779, + "step": 23923 + }, + { + "epoch": 1.660120741100548, + "grad_norm": 1.40376615524292, + "learning_rate": 0.00013181721427930345, + "loss": 0.59, + "step": 23924 + }, + { + "epoch": 1.660190132537645, + "grad_norm": 1.6046738624572754, + "learning_rate": 0.0001318118875964469, + "loss": 0.6726, + "step": 23925 + }, + { + "epoch": 1.6602595239747415, + "grad_norm": 1.2665884494781494, + "learning_rate": 0.0001318065608131639, + "loss": 0.4978, + "step": 23926 + }, + { + "epoch": 1.6603289154118381, + "grad_norm": 1.9615137577056885, + "learning_rate": 0.0001318012339294713, + "loss": 0.6893, + "step": 23927 + }, + { + "epoch": 1.660398306848935, + "grad_norm": 2.4338018894195557, + "learning_rate": 0.00013179590694538588, + "loss": 0.9682, + "step": 23928 + }, + { + "epoch": 1.6604676982860314, + "grad_norm": 2.1372721195220947, + "learning_rate": 0.00013179057986092448, + "loss": 0.6173, + "step": 23929 + }, + { + "epoch": 1.6605370897231282, + "grad_norm": 1.4268819093704224, + "learning_rate": 0.0001317852526761039, + "loss": 0.508, + "step": 23930 + }, + { + "epoch": 1.6606064811602248, + "grad_norm": 1.8022407293319702, + "learning_rate": 0.00013177992539094099, + "loss": 0.5664, + "step": 23931 + }, + { + "epoch": 1.6606758725973214, + "grad_norm": 1.883552074432373, + "learning_rate": 0.00013177459800545248, + "loss": 0.7584, + "step": 23932 + }, + { + "epoch": 1.6607452640344182, + "grad_norm": 1.1281852722167969, + "learning_rate": 0.00013176927051965532, + "loss": 0.4562, + "step": 23933 + }, + { + "epoch": 1.6608146554715149, + "grad_norm": 1.4508132934570312, + "learning_rate": 0.0001317639429335662, + "loss": 0.6658, + "step": 23934 + }, + { + "epoch": 1.6608840469086115, + "grad_norm": 1.3085966110229492, + "learning_rate": 0.00013175861524720203, + "loss": 0.4237, + "step": 23935 + }, + { + "epoch": 1.6609534383457083, + "grad_norm": 2.0232479572296143, + "learning_rate": 0.0001317532874605796, + "loss": 0.7643, + "step": 23936 + }, + { + "epoch": 1.6610228297828047, + "grad_norm": 2.4602346420288086, + "learning_rate": 0.00013174795957371572, + "loss": 0.614, + "step": 23937 + }, + { + "epoch": 1.6610922212199015, + "grad_norm": 2.1105973720550537, + "learning_rate": 0.0001317426315866272, + "loss": 0.8719, + "step": 23938 + }, + { + "epoch": 1.6611616126569981, + "grad_norm": 2.2748143672943115, + "learning_rate": 0.00013173730349933085, + "loss": 0.4559, + "step": 23939 + }, + { + "epoch": 1.6612310040940947, + "grad_norm": 1.500013828277588, + "learning_rate": 0.00013173197531184358, + "loss": 0.5806, + "step": 23940 + }, + { + "epoch": 1.6613003955311916, + "grad_norm": 1.6853938102722168, + "learning_rate": 0.00013172664702418212, + "loss": 0.8077, + "step": 23941 + }, + { + "epoch": 1.661369786968288, + "grad_norm": 1.5149465799331665, + "learning_rate": 0.0001317213186363633, + "loss": 0.5176, + "step": 23942 + }, + { + "epoch": 1.6614391784053848, + "grad_norm": 1.654923677444458, + "learning_rate": 0.00013171599014840396, + "loss": 0.6524, + "step": 23943 + }, + { + "epoch": 1.6615085698424814, + "grad_norm": 1.6145482063293457, + "learning_rate": 0.00013171066156032095, + "loss": 0.4832, + "step": 23944 + }, + { + "epoch": 1.661577961279578, + "grad_norm": 1.6385763883590698, + "learning_rate": 0.00013170533287213104, + "loss": 0.6414, + "step": 23945 + }, + { + "epoch": 1.6616473527166749, + "grad_norm": 1.9601794481277466, + "learning_rate": 0.00013170000408385107, + "loss": 0.7238, + "step": 23946 + }, + { + "epoch": 1.6617167441537715, + "grad_norm": 1.6044628620147705, + "learning_rate": 0.00013169467519549786, + "loss": 0.8469, + "step": 23947 + }, + { + "epoch": 1.661786135590868, + "grad_norm": 1.613390326499939, + "learning_rate": 0.00013168934620708827, + "loss": 0.5936, + "step": 23948 + }, + { + "epoch": 1.661855527027965, + "grad_norm": 1.3006658554077148, + "learning_rate": 0.00013168401711863905, + "loss": 0.3638, + "step": 23949 + }, + { + "epoch": 1.6619249184650613, + "grad_norm": 1.49314284324646, + "learning_rate": 0.00013167868793016708, + "loss": 0.564, + "step": 23950 + }, + { + "epoch": 1.6619943099021581, + "grad_norm": 1.9236671924591064, + "learning_rate": 0.00013167335864168917, + "loss": 0.582, + "step": 23951 + }, + { + "epoch": 1.6620637013392547, + "grad_norm": 1.6501151323318481, + "learning_rate": 0.00013166802925322212, + "loss": 0.5053, + "step": 23952 + }, + { + "epoch": 1.6621330927763514, + "grad_norm": 2.1600799560546875, + "learning_rate": 0.0001316626997647828, + "loss": 0.6448, + "step": 23953 + }, + { + "epoch": 1.6622024842134482, + "grad_norm": 1.1068226099014282, + "learning_rate": 0.00013165737017638803, + "loss": 0.3706, + "step": 23954 + }, + { + "epoch": 1.6622718756505446, + "grad_norm": 1.5897234678268433, + "learning_rate": 0.00013165204048805457, + "loss": 0.5958, + "step": 23955 + }, + { + "epoch": 1.6623412670876414, + "grad_norm": 2.074031114578247, + "learning_rate": 0.0001316467106997993, + "loss": 0.4843, + "step": 23956 + }, + { + "epoch": 1.662410658524738, + "grad_norm": 1.4392709732055664, + "learning_rate": 0.00013164138081163903, + "loss": 0.6105, + "step": 23957 + }, + { + "epoch": 1.6624800499618346, + "grad_norm": 1.5074050426483154, + "learning_rate": 0.00013163605082359064, + "loss": 0.6162, + "step": 23958 + }, + { + "epoch": 1.6625494413989315, + "grad_norm": 2.134265899658203, + "learning_rate": 0.00013163072073567086, + "loss": 0.3698, + "step": 23959 + }, + { + "epoch": 1.662618832836028, + "grad_norm": 1.5978400707244873, + "learning_rate": 0.00013162539054789658, + "loss": 0.4702, + "step": 23960 + }, + { + "epoch": 1.6626882242731247, + "grad_norm": 1.4997491836547852, + "learning_rate": 0.00013162006026028463, + "loss": 0.5367, + "step": 23961 + }, + { + "epoch": 1.6627576157102215, + "grad_norm": 1.6165965795516968, + "learning_rate": 0.0001316147298728518, + "loss": 0.6732, + "step": 23962 + }, + { + "epoch": 1.662827007147318, + "grad_norm": 1.182722568511963, + "learning_rate": 0.00013160939938561495, + "loss": 0.512, + "step": 23963 + }, + { + "epoch": 1.6628963985844147, + "grad_norm": 1.0686107873916626, + "learning_rate": 0.00013160406879859086, + "loss": 0.4202, + "step": 23964 + }, + { + "epoch": 1.6629657900215113, + "grad_norm": 2.0378289222717285, + "learning_rate": 0.00013159873811179642, + "loss": 0.5072, + "step": 23965 + }, + { + "epoch": 1.663035181458608, + "grad_norm": 1.9847372770309448, + "learning_rate": 0.0001315934073252484, + "loss": 0.5569, + "step": 23966 + }, + { + "epoch": 1.6631045728957048, + "grad_norm": 1.5681496858596802, + "learning_rate": 0.0001315880764389637, + "loss": 0.5673, + "step": 23967 + }, + { + "epoch": 1.6631739643328012, + "grad_norm": 1.5706590414047241, + "learning_rate": 0.0001315827454529591, + "loss": 0.6127, + "step": 23968 + }, + { + "epoch": 1.663243355769898, + "grad_norm": 1.9060627222061157, + "learning_rate": 0.0001315774143672514, + "loss": 0.7088, + "step": 23969 + }, + { + "epoch": 1.6633127472069946, + "grad_norm": 1.6098979711532593, + "learning_rate": 0.0001315720831818575, + "loss": 0.4976, + "step": 23970 + }, + { + "epoch": 1.6633821386440912, + "grad_norm": 1.7309646606445312, + "learning_rate": 0.00013156675189679418, + "loss": 0.6246, + "step": 23971 + }, + { + "epoch": 1.663451530081188, + "grad_norm": 1.4280864000320435, + "learning_rate": 0.00013156142051207832, + "loss": 0.5779, + "step": 23972 + }, + { + "epoch": 1.6635209215182847, + "grad_norm": 1.6584789752960205, + "learning_rate": 0.00013155608902772666, + "loss": 0.7311, + "step": 23973 + }, + { + "epoch": 1.6635903129553813, + "grad_norm": 1.6972678899765015, + "learning_rate": 0.00013155075744375613, + "loss": 0.5874, + "step": 23974 + }, + { + "epoch": 1.6636597043924781, + "grad_norm": 1.6075596809387207, + "learning_rate": 0.0001315454257601835, + "loss": 0.4744, + "step": 23975 + }, + { + "epoch": 1.6637290958295745, + "grad_norm": 1.5666829347610474, + "learning_rate": 0.00013154009397702563, + "loss": 0.8194, + "step": 23976 + }, + { + "epoch": 1.6637984872666713, + "grad_norm": 1.7193483114242554, + "learning_rate": 0.00013153476209429933, + "loss": 0.7438, + "step": 23977 + }, + { + "epoch": 1.663867878703768, + "grad_norm": 1.7388755083084106, + "learning_rate": 0.00013152943011202149, + "loss": 0.6276, + "step": 23978 + }, + { + "epoch": 1.6639372701408646, + "grad_norm": 1.580510139465332, + "learning_rate": 0.00013152409803020884, + "loss": 0.4694, + "step": 23979 + }, + { + "epoch": 1.6640066615779614, + "grad_norm": 1.4395864009857178, + "learning_rate": 0.0001315187658488783, + "loss": 0.6943, + "step": 23980 + }, + { + "epoch": 1.6640760530150578, + "grad_norm": 1.3888063430786133, + "learning_rate": 0.00013151343356804669, + "loss": 0.5469, + "step": 23981 + }, + { + "epoch": 1.6641454444521546, + "grad_norm": 1.5476189851760864, + "learning_rate": 0.0001315081011877308, + "loss": 0.7262, + "step": 23982 + }, + { + "epoch": 1.6642148358892512, + "grad_norm": 2.21681547164917, + "learning_rate": 0.00013150276870794748, + "loss": 0.488, + "step": 23983 + }, + { + "epoch": 1.6642842273263478, + "grad_norm": 1.5032901763916016, + "learning_rate": 0.00013149743612871358, + "loss": 0.5844, + "step": 23984 + }, + { + "epoch": 1.6643536187634447, + "grad_norm": 1.4123094081878662, + "learning_rate": 0.00013149210345004595, + "loss": 0.5575, + "step": 23985 + }, + { + "epoch": 1.6644230102005413, + "grad_norm": 1.6859577894210815, + "learning_rate": 0.0001314867706719614, + "loss": 0.5941, + "step": 23986 + }, + { + "epoch": 1.664492401637638, + "grad_norm": 1.822213888168335, + "learning_rate": 0.00013148143779447675, + "loss": 0.6306, + "step": 23987 + }, + { + "epoch": 1.6645617930747347, + "grad_norm": 1.4421125650405884, + "learning_rate": 0.00013147610481760886, + "loss": 0.5523, + "step": 23988 + }, + { + "epoch": 1.6646311845118311, + "grad_norm": 1.1575112342834473, + "learning_rate": 0.00013147077174137457, + "loss": 0.4815, + "step": 23989 + }, + { + "epoch": 1.664700575948928, + "grad_norm": 1.304139494895935, + "learning_rate": 0.0001314654385657907, + "loss": 0.4844, + "step": 23990 + }, + { + "epoch": 1.6647699673860246, + "grad_norm": 1.7775615453720093, + "learning_rate": 0.00013146010529087407, + "loss": 0.7422, + "step": 23991 + }, + { + "epoch": 1.6648393588231212, + "grad_norm": 1.6688625812530518, + "learning_rate": 0.00013145477191664156, + "loss": 0.6628, + "step": 23992 + }, + { + "epoch": 1.664908750260218, + "grad_norm": 1.5822045803070068, + "learning_rate": 0.00013144943844310998, + "loss": 0.5819, + "step": 23993 + }, + { + "epoch": 1.6649781416973144, + "grad_norm": 1.6070393323898315, + "learning_rate": 0.00013144410487029616, + "loss": 0.595, + "step": 23994 + }, + { + "epoch": 1.6650475331344112, + "grad_norm": 1.7782307863235474, + "learning_rate": 0.00013143877119821694, + "loss": 0.6698, + "step": 23995 + }, + { + "epoch": 1.6651169245715078, + "grad_norm": 1.3435612916946411, + "learning_rate": 0.0001314334374268892, + "loss": 0.5892, + "step": 23996 + }, + { + "epoch": 1.6651863160086045, + "grad_norm": 1.0569747686386108, + "learning_rate": 0.00013142810355632972, + "loss": 0.497, + "step": 23997 + }, + { + "epoch": 1.6652557074457013, + "grad_norm": 1.3311872482299805, + "learning_rate": 0.0001314227695865554, + "loss": 0.5552, + "step": 23998 + }, + { + "epoch": 1.665325098882798, + "grad_norm": 1.3228915929794312, + "learning_rate": 0.000131417435517583, + "loss": 0.5625, + "step": 23999 + }, + { + "epoch": 1.6653944903198945, + "grad_norm": 1.9159287214279175, + "learning_rate": 0.00013141210134942944, + "loss": 0.4911, + "step": 24000 + }, + { + "epoch": 1.6653944903198945, + "eval_loss": 0.6305449604988098, + "eval_runtime": 3859.8069, + "eval_samples_per_second": 6.288, + "eval_steps_per_second": 0.786, + "step": 24000 + }, + { + "epoch": 1.6654638817569913, + "grad_norm": 1.1250481605529785, + "learning_rate": 0.0001314067670821115, + "loss": 0.5008, + "step": 24001 + }, + { + "epoch": 1.6655332731940877, + "grad_norm": 1.4461610317230225, + "learning_rate": 0.00013140143271564604, + "loss": 0.5735, + "step": 24002 + }, + { + "epoch": 1.6656026646311846, + "grad_norm": 1.4844099283218384, + "learning_rate": 0.0001313960982500499, + "loss": 0.56, + "step": 24003 + }, + { + "epoch": 1.6656720560682812, + "grad_norm": 2.286714792251587, + "learning_rate": 0.0001313907636853399, + "loss": 0.8385, + "step": 24004 + }, + { + "epoch": 1.6657414475053778, + "grad_norm": 1.5499378442764282, + "learning_rate": 0.00013138542902153293, + "loss": 0.4556, + "step": 24005 + }, + { + "epoch": 1.6658108389424746, + "grad_norm": 2.251716136932373, + "learning_rate": 0.0001313800942586458, + "loss": 0.7551, + "step": 24006 + }, + { + "epoch": 1.665880230379571, + "grad_norm": 2.164405107498169, + "learning_rate": 0.0001313747593966953, + "loss": 0.7787, + "step": 24007 + }, + { + "epoch": 1.6659496218166678, + "grad_norm": 3.0741798877716064, + "learning_rate": 0.0001313694244356984, + "loss": 0.7206, + "step": 24008 + }, + { + "epoch": 1.6660190132537644, + "grad_norm": 1.442979097366333, + "learning_rate": 0.00013136408937567185, + "loss": 0.4643, + "step": 24009 + }, + { + "epoch": 1.666088404690861, + "grad_norm": 1.8041884899139404, + "learning_rate": 0.0001313587542166325, + "loss": 0.6354, + "step": 24010 + }, + { + "epoch": 1.666157796127958, + "grad_norm": 1.6555569171905518, + "learning_rate": 0.00013135341895859716, + "loss": 0.5699, + "step": 24011 + }, + { + "epoch": 1.6662271875650545, + "grad_norm": 1.342340350151062, + "learning_rate": 0.00013134808360158276, + "loss": 0.4622, + "step": 24012 + }, + { + "epoch": 1.6662965790021511, + "grad_norm": 1.6791410446166992, + "learning_rate": 0.0001313427481456061, + "loss": 0.4416, + "step": 24013 + }, + { + "epoch": 1.666365970439248, + "grad_norm": 1.7002917528152466, + "learning_rate": 0.00013133741259068398, + "loss": 0.4406, + "step": 24014 + }, + { + "epoch": 1.6664353618763443, + "grad_norm": 1.2580209970474243, + "learning_rate": 0.00013133207693683332, + "loss": 0.5426, + "step": 24015 + }, + { + "epoch": 1.6665047533134412, + "grad_norm": 1.4569000005722046, + "learning_rate": 0.00013132674118407093, + "loss": 0.5624, + "step": 24016 + }, + { + "epoch": 1.6665741447505378, + "grad_norm": 2.1199991703033447, + "learning_rate": 0.00013132140533241363, + "loss": 0.8168, + "step": 24017 + }, + { + "epoch": 1.6666435361876344, + "grad_norm": 1.4677141904830933, + "learning_rate": 0.00013131606938187834, + "loss": 0.5667, + "step": 24018 + }, + { + "epoch": 1.6667129276247312, + "grad_norm": 1.0506556034088135, + "learning_rate": 0.0001313107333324818, + "loss": 0.601, + "step": 24019 + }, + { + "epoch": 1.6667823190618276, + "grad_norm": 2.0937936305999756, + "learning_rate": 0.00013130539718424092, + "loss": 0.5897, + "step": 24020 + }, + { + "epoch": 1.6668517104989244, + "grad_norm": 1.5823194980621338, + "learning_rate": 0.00013130006093717254, + "loss": 0.6454, + "step": 24021 + }, + { + "epoch": 1.666921101936021, + "grad_norm": 1.5117202997207642, + "learning_rate": 0.0001312947245912935, + "loss": 0.7598, + "step": 24022 + }, + { + "epoch": 1.6669904933731177, + "grad_norm": 1.860458493232727, + "learning_rate": 0.00013128938814662065, + "loss": 0.76, + "step": 24023 + }, + { + "epoch": 1.6670598848102145, + "grad_norm": 1.873517632484436, + "learning_rate": 0.0001312840516031708, + "loss": 0.6753, + "step": 24024 + }, + { + "epoch": 1.667129276247311, + "grad_norm": 1.8622655868530273, + "learning_rate": 0.00013127871496096085, + "loss": 0.7076, + "step": 24025 + }, + { + "epoch": 1.6671986676844077, + "grad_norm": 1.8819472789764404, + "learning_rate": 0.00013127337822000762, + "loss": 0.7913, + "step": 24026 + }, + { + "epoch": 1.6672680591215046, + "grad_norm": 2.002460241317749, + "learning_rate": 0.000131268041380328, + "loss": 0.6587, + "step": 24027 + }, + { + "epoch": 1.667337450558601, + "grad_norm": 2.389209508895874, + "learning_rate": 0.00013126270444193875, + "loss": 0.5882, + "step": 24028 + }, + { + "epoch": 1.6674068419956978, + "grad_norm": 1.8357990980148315, + "learning_rate": 0.00013125736740485678, + "loss": 0.6072, + "step": 24029 + }, + { + "epoch": 1.6674762334327944, + "grad_norm": 1.808931827545166, + "learning_rate": 0.00013125203026909894, + "loss": 0.7788, + "step": 24030 + }, + { + "epoch": 1.667545624869891, + "grad_norm": 1.5242334604263306, + "learning_rate": 0.00013124669303468203, + "loss": 0.5371, + "step": 24031 + }, + { + "epoch": 1.6676150163069878, + "grad_norm": 1.507633924484253, + "learning_rate": 0.00013124135570162298, + "loss": 0.6204, + "step": 24032 + }, + { + "epoch": 1.6676844077440842, + "grad_norm": 1.585376501083374, + "learning_rate": 0.00013123601826993858, + "loss": 0.6246, + "step": 24033 + }, + { + "epoch": 1.667753799181181, + "grad_norm": 1.6913766860961914, + "learning_rate": 0.0001312306807396457, + "loss": 0.6754, + "step": 24034 + }, + { + "epoch": 1.6678231906182777, + "grad_norm": 2.214993953704834, + "learning_rate": 0.0001312253431107612, + "loss": 0.7997, + "step": 24035 + }, + { + "epoch": 1.6678925820553743, + "grad_norm": 1.3793717622756958, + "learning_rate": 0.00013122000538330188, + "loss": 0.5687, + "step": 24036 + }, + { + "epoch": 1.667961973492471, + "grad_norm": 1.956911325454712, + "learning_rate": 0.00013121466755728467, + "loss": 0.5833, + "step": 24037 + }, + { + "epoch": 1.6680313649295677, + "grad_norm": 1.7793283462524414, + "learning_rate": 0.00013120932963272633, + "loss": 0.5432, + "step": 24038 + }, + { + "epoch": 1.6681007563666643, + "grad_norm": 1.5932127237319946, + "learning_rate": 0.00013120399160964377, + "loss": 0.5958, + "step": 24039 + }, + { + "epoch": 1.6681701478037612, + "grad_norm": 1.7866666316986084, + "learning_rate": 0.00013119865348805386, + "loss": 0.6047, + "step": 24040 + }, + { + "epoch": 1.6682395392408575, + "grad_norm": 1.538727045059204, + "learning_rate": 0.0001311933152679734, + "loss": 0.7105, + "step": 24041 + }, + { + "epoch": 1.6683089306779544, + "grad_norm": 1.709892988204956, + "learning_rate": 0.00013118797694941923, + "loss": 0.6033, + "step": 24042 + }, + { + "epoch": 1.668378322115051, + "grad_norm": 2.029003620147705, + "learning_rate": 0.00013118263853240824, + "loss": 0.4131, + "step": 24043 + }, + { + "epoch": 1.6684477135521476, + "grad_norm": 6.827780723571777, + "learning_rate": 0.0001311773000169573, + "loss": 0.6917, + "step": 24044 + }, + { + "epoch": 1.6685171049892444, + "grad_norm": 1.6214317083358765, + "learning_rate": 0.00013117196140308325, + "loss": 0.5249, + "step": 24045 + }, + { + "epoch": 1.6685864964263408, + "grad_norm": 1.771946907043457, + "learning_rate": 0.00013116662269080294, + "loss": 0.7137, + "step": 24046 + }, + { + "epoch": 1.6686558878634377, + "grad_norm": 1.3038965463638306, + "learning_rate": 0.0001311612838801332, + "loss": 0.4972, + "step": 24047 + }, + { + "epoch": 1.6687252793005343, + "grad_norm": 1.4294164180755615, + "learning_rate": 0.0001311559449710909, + "loss": 0.7267, + "step": 24048 + }, + { + "epoch": 1.6687946707376309, + "grad_norm": 2.4057562351226807, + "learning_rate": 0.0001311506059636929, + "loss": 0.6223, + "step": 24049 + }, + { + "epoch": 1.6688640621747277, + "grad_norm": 2.108579397201538, + "learning_rate": 0.00013114526685795605, + "loss": 0.6793, + "step": 24050 + }, + { + "epoch": 1.6689334536118243, + "grad_norm": 1.8752152919769287, + "learning_rate": 0.00013113992765389723, + "loss": 0.5637, + "step": 24051 + }, + { + "epoch": 1.669002845048921, + "grad_norm": 3.430388927459717, + "learning_rate": 0.00013113458835153325, + "loss": 0.7981, + "step": 24052 + }, + { + "epoch": 1.6690722364860178, + "grad_norm": 1.921799898147583, + "learning_rate": 0.000131129248950881, + "loss": 0.6524, + "step": 24053 + }, + { + "epoch": 1.6691416279231142, + "grad_norm": 2.661619186401367, + "learning_rate": 0.00013112390945195732, + "loss": 0.8293, + "step": 24054 + }, + { + "epoch": 1.669211019360211, + "grad_norm": 1.858812928199768, + "learning_rate": 0.00013111856985477903, + "loss": 0.6468, + "step": 24055 + }, + { + "epoch": 1.6692804107973076, + "grad_norm": 1.5627561807632446, + "learning_rate": 0.00013111323015936306, + "loss": 0.5902, + "step": 24056 + }, + { + "epoch": 1.6693498022344042, + "grad_norm": 1.620646595954895, + "learning_rate": 0.0001311078903657262, + "loss": 0.7219, + "step": 24057 + }, + { + "epoch": 1.669419193671501, + "grad_norm": 1.8835521936416626, + "learning_rate": 0.0001311025504738854, + "loss": 0.8391, + "step": 24058 + }, + { + "epoch": 1.6694885851085974, + "grad_norm": 1.541430950164795, + "learning_rate": 0.00013109721048385742, + "loss": 0.5008, + "step": 24059 + }, + { + "epoch": 1.6695579765456943, + "grad_norm": 1.4954341650009155, + "learning_rate": 0.00013109187039565912, + "loss": 0.5288, + "step": 24060 + }, + { + "epoch": 1.6696273679827909, + "grad_norm": 1.3368459939956665, + "learning_rate": 0.00013108653020930746, + "loss": 0.6187, + "step": 24061 + }, + { + "epoch": 1.6696967594198875, + "grad_norm": 2.0507659912109375, + "learning_rate": 0.00013108118992481919, + "loss": 0.661, + "step": 24062 + }, + { + "epoch": 1.6697661508569843, + "grad_norm": 2.1774749755859375, + "learning_rate": 0.0001310758495422112, + "loss": 0.6779, + "step": 24063 + }, + { + "epoch": 1.669835542294081, + "grad_norm": 1.6242144107818604, + "learning_rate": 0.0001310705090615004, + "loss": 0.5715, + "step": 24064 + }, + { + "epoch": 1.6699049337311775, + "grad_norm": 1.6801679134368896, + "learning_rate": 0.0001310651684827036, + "loss": 0.4228, + "step": 24065 + }, + { + "epoch": 1.6699743251682744, + "grad_norm": 1.6699634790420532, + "learning_rate": 0.00013105982780583762, + "loss": 0.6683, + "step": 24066 + }, + { + "epoch": 1.6700437166053708, + "grad_norm": 1.3460074663162231, + "learning_rate": 0.00013105448703091941, + "loss": 0.5975, + "step": 24067 + }, + { + "epoch": 1.6701131080424676, + "grad_norm": 1.7260022163391113, + "learning_rate": 0.0001310491461579658, + "loss": 0.5588, + "step": 24068 + }, + { + "epoch": 1.6701824994795642, + "grad_norm": 1.1195745468139648, + "learning_rate": 0.0001310438051869936, + "loss": 0.3184, + "step": 24069 + }, + { + "epoch": 1.6702518909166608, + "grad_norm": 1.6511305570602417, + "learning_rate": 0.00013103846411801975, + "loss": 0.4435, + "step": 24070 + }, + { + "epoch": 1.6703212823537577, + "grad_norm": 2.1274006366729736, + "learning_rate": 0.00013103312295106103, + "loss": 0.606, + "step": 24071 + }, + { + "epoch": 1.6703906737908543, + "grad_norm": 1.8201777935028076, + "learning_rate": 0.00013102778168613436, + "loss": 0.626, + "step": 24072 + }, + { + "epoch": 1.6704600652279509, + "grad_norm": 1.3844703435897827, + "learning_rate": 0.0001310224403232566, + "loss": 0.5185, + "step": 24073 + }, + { + "epoch": 1.6705294566650475, + "grad_norm": 1.8426058292388916, + "learning_rate": 0.00013101709886244455, + "loss": 0.6515, + "step": 24074 + }, + { + "epoch": 1.670598848102144, + "grad_norm": 2.33183217048645, + "learning_rate": 0.00013101175730371515, + "loss": 0.7225, + "step": 24075 + }, + { + "epoch": 1.670668239539241, + "grad_norm": 1.2838743925094604, + "learning_rate": 0.00013100641564708522, + "loss": 0.5354, + "step": 24076 + }, + { + "epoch": 1.6707376309763375, + "grad_norm": 1.299728512763977, + "learning_rate": 0.00013100107389257164, + "loss": 0.4566, + "step": 24077 + }, + { + "epoch": 1.6708070224134342, + "grad_norm": 1.640634298324585, + "learning_rate": 0.00013099573204019126, + "loss": 0.4926, + "step": 24078 + }, + { + "epoch": 1.670876413850531, + "grad_norm": 1.610254168510437, + "learning_rate": 0.00013099039008996095, + "loss": 0.5245, + "step": 24079 + }, + { + "epoch": 1.6709458052876274, + "grad_norm": 3.0186448097229004, + "learning_rate": 0.00013098504804189758, + "loss": 0.6109, + "step": 24080 + }, + { + "epoch": 1.6710151967247242, + "grad_norm": 1.948990821838379, + "learning_rate": 0.00013097970589601797, + "loss": 0.5739, + "step": 24081 + }, + { + "epoch": 1.6710845881618208, + "grad_norm": 1.5887596607208252, + "learning_rate": 0.00013097436365233907, + "loss": 0.5245, + "step": 24082 + }, + { + "epoch": 1.6711539795989174, + "grad_norm": 1.214834451675415, + "learning_rate": 0.0001309690213108777, + "loss": 0.5051, + "step": 24083 + }, + { + "epoch": 1.6712233710360143, + "grad_norm": 1.6015650033950806, + "learning_rate": 0.0001309636788716507, + "loss": 0.4872, + "step": 24084 + }, + { + "epoch": 1.6712927624731109, + "grad_norm": 1.213624358177185, + "learning_rate": 0.00013095833633467498, + "loss": 0.417, + "step": 24085 + }, + { + "epoch": 1.6713621539102075, + "grad_norm": 1.7926956415176392, + "learning_rate": 0.00013095299369996735, + "loss": 0.5464, + "step": 24086 + }, + { + "epoch": 1.671431545347304, + "grad_norm": 2.1062417030334473, + "learning_rate": 0.00013094765096754475, + "loss": 0.8748, + "step": 24087 + }, + { + "epoch": 1.6715009367844007, + "grad_norm": 2.1017138957977295, + "learning_rate": 0.000130942308137424, + "loss": 0.4882, + "step": 24088 + }, + { + "epoch": 1.6715703282214975, + "grad_norm": 1.6319522857666016, + "learning_rate": 0.00013093696520962195, + "loss": 0.3775, + "step": 24089 + }, + { + "epoch": 1.6716397196585941, + "grad_norm": 1.7734544277191162, + "learning_rate": 0.0001309316221841555, + "loss": 0.5147, + "step": 24090 + }, + { + "epoch": 1.6717091110956908, + "grad_norm": 4.488664150238037, + "learning_rate": 0.0001309262790610415, + "loss": 0.7834, + "step": 24091 + }, + { + "epoch": 1.6717785025327876, + "grad_norm": 1.7787127494812012, + "learning_rate": 0.00013092093584029683, + "loss": 0.732, + "step": 24092 + }, + { + "epoch": 1.671847893969884, + "grad_norm": 1.5314595699310303, + "learning_rate": 0.00013091559252193836, + "loss": 0.5345, + "step": 24093 + }, + { + "epoch": 1.6719172854069808, + "grad_norm": 1.5617525577545166, + "learning_rate": 0.00013091024910598294, + "loss": 0.57, + "step": 24094 + }, + { + "epoch": 1.6719866768440774, + "grad_norm": 1.512999415397644, + "learning_rate": 0.00013090490559244746, + "loss": 0.4496, + "step": 24095 + }, + { + "epoch": 1.672056068281174, + "grad_norm": 2.2330541610717773, + "learning_rate": 0.0001308995619813488, + "loss": 0.6879, + "step": 24096 + }, + { + "epoch": 1.6721254597182709, + "grad_norm": 1.5794994831085205, + "learning_rate": 0.00013089421827270376, + "loss": 0.7215, + "step": 24097 + }, + { + "epoch": 1.6721948511553675, + "grad_norm": 1.9612596035003662, + "learning_rate": 0.00013088887446652926, + "loss": 0.6518, + "step": 24098 + }, + { + "epoch": 1.672264242592464, + "grad_norm": 1.7499524354934692, + "learning_rate": 0.0001308835305628422, + "loss": 0.682, + "step": 24099 + }, + { + "epoch": 1.6723336340295607, + "grad_norm": 1.3808066844940186, + "learning_rate": 0.00013087818656165942, + "loss": 0.5284, + "step": 24100 + }, + { + "epoch": 1.6724030254666573, + "grad_norm": 1.7541332244873047, + "learning_rate": 0.00013087284246299775, + "loss": 0.6531, + "step": 24101 + }, + { + "epoch": 1.6724724169037541, + "grad_norm": 2.0895798206329346, + "learning_rate": 0.0001308674982668741, + "loss": 0.9339, + "step": 24102 + }, + { + "epoch": 1.6725418083408508, + "grad_norm": 1.5381017923355103, + "learning_rate": 0.0001308621539733054, + "loss": 0.5707, + "step": 24103 + }, + { + "epoch": 1.6726111997779474, + "grad_norm": 1.554079294204712, + "learning_rate": 0.0001308568095823084, + "loss": 0.5743, + "step": 24104 + }, + { + "epoch": 1.6726805912150442, + "grad_norm": 2.228020191192627, + "learning_rate": 0.00013085146509390008, + "loss": 0.5341, + "step": 24105 + }, + { + "epoch": 1.6727499826521406, + "grad_norm": 1.7543411254882812, + "learning_rate": 0.00013084612050809724, + "loss": 0.5808, + "step": 24106 + }, + { + "epoch": 1.6728193740892374, + "grad_norm": 4.009032726287842, + "learning_rate": 0.00013084077582491675, + "loss": 0.7322, + "step": 24107 + }, + { + "epoch": 1.672888765526334, + "grad_norm": 1.0434646606445312, + "learning_rate": 0.00013083543104437556, + "loss": 0.4831, + "step": 24108 + }, + { + "epoch": 1.6729581569634306, + "grad_norm": 1.616480827331543, + "learning_rate": 0.00013083008616649044, + "loss": 0.6279, + "step": 24109 + }, + { + "epoch": 1.6730275484005275, + "grad_norm": 2.3430910110473633, + "learning_rate": 0.00013082474119127834, + "loss": 0.6583, + "step": 24110 + }, + { + "epoch": 1.673096939837624, + "grad_norm": 1.6135225296020508, + "learning_rate": 0.0001308193961187561, + "loss": 0.5429, + "step": 24111 + }, + { + "epoch": 1.6731663312747207, + "grad_norm": 2.267299175262451, + "learning_rate": 0.0001308140509489406, + "loss": 0.6763, + "step": 24112 + }, + { + "epoch": 1.6732357227118173, + "grad_norm": 1.3006230592727661, + "learning_rate": 0.00013080870568184876, + "loss": 0.5727, + "step": 24113 + }, + { + "epoch": 1.673305114148914, + "grad_norm": 1.1248224973678589, + "learning_rate": 0.00013080336031749736, + "loss": 0.5117, + "step": 24114 + }, + { + "epoch": 1.6733745055860108, + "grad_norm": 1.676609992980957, + "learning_rate": 0.00013079801485590332, + "loss": 0.7115, + "step": 24115 + }, + { + "epoch": 1.6734438970231074, + "grad_norm": 1.6546714305877686, + "learning_rate": 0.00013079266929708357, + "loss": 0.6433, + "step": 24116 + }, + { + "epoch": 1.673513288460204, + "grad_norm": 2.0644145011901855, + "learning_rate": 0.0001307873236410549, + "loss": 0.7842, + "step": 24117 + }, + { + "epoch": 1.6735826798973008, + "grad_norm": 1.345750331878662, + "learning_rate": 0.00013078197788783422, + "loss": 0.6121, + "step": 24118 + }, + { + "epoch": 1.6736520713343972, + "grad_norm": 1.265360951423645, + "learning_rate": 0.00013077663203743843, + "loss": 0.3291, + "step": 24119 + }, + { + "epoch": 1.673721462771494, + "grad_norm": 1.7050327062606812, + "learning_rate": 0.00013077128608988435, + "loss": 0.5783, + "step": 24120 + }, + { + "epoch": 1.6737908542085906, + "grad_norm": 1.6821889877319336, + "learning_rate": 0.0001307659400451889, + "loss": 0.7206, + "step": 24121 + }, + { + "epoch": 1.6738602456456873, + "grad_norm": 2.3046228885650635, + "learning_rate": 0.00013076059390336898, + "loss": 0.7974, + "step": 24122 + }, + { + "epoch": 1.673929637082784, + "grad_norm": 1.4958082437515259, + "learning_rate": 0.0001307552476644414, + "loss": 0.6358, + "step": 24123 + }, + { + "epoch": 1.6739990285198807, + "grad_norm": 2.965958833694458, + "learning_rate": 0.0001307499013284231, + "loss": 0.7477, + "step": 24124 + }, + { + "epoch": 1.6740684199569773, + "grad_norm": 1.9744330644607544, + "learning_rate": 0.0001307445548953309, + "loss": 0.8603, + "step": 24125 + }, + { + "epoch": 1.674137811394074, + "grad_norm": 1.5023224353790283, + "learning_rate": 0.0001307392083651817, + "loss": 0.627, + "step": 24126 + }, + { + "epoch": 1.6742072028311705, + "grad_norm": 1.3515764474868774, + "learning_rate": 0.00013073386173799242, + "loss": 0.4899, + "step": 24127 + }, + { + "epoch": 1.6742765942682674, + "grad_norm": 1.1832300424575806, + "learning_rate": 0.0001307285150137799, + "loss": 0.3543, + "step": 24128 + }, + { + "epoch": 1.674345985705364, + "grad_norm": 3.0058441162109375, + "learning_rate": 0.000130723168192561, + "loss": 0.5779, + "step": 24129 + }, + { + "epoch": 1.6744153771424606, + "grad_norm": 1.5317882299423218, + "learning_rate": 0.00013071782127435264, + "loss": 0.5694, + "step": 24130 + }, + { + "epoch": 1.6744847685795574, + "grad_norm": 1.6364529132843018, + "learning_rate": 0.00013071247425917168, + "loss": 0.6088, + "step": 24131 + }, + { + "epoch": 1.6745541600166538, + "grad_norm": 1.9838664531707764, + "learning_rate": 0.000130707127147035, + "loss": 0.7323, + "step": 24132 + }, + { + "epoch": 1.6746235514537506, + "grad_norm": 2.1000022888183594, + "learning_rate": 0.00013070177993795948, + "loss": 0.6722, + "step": 24133 + }, + { + "epoch": 1.6746929428908472, + "grad_norm": 2.16896390914917, + "learning_rate": 0.00013069643263196202, + "loss": 0.6692, + "step": 24134 + }, + { + "epoch": 1.6747623343279439, + "grad_norm": 1.2227263450622559, + "learning_rate": 0.00013069108522905943, + "loss": 0.5482, + "step": 24135 + }, + { + "epoch": 1.6748317257650407, + "grad_norm": 1.7403990030288696, + "learning_rate": 0.0001306857377292687, + "loss": 0.7376, + "step": 24136 + }, + { + "epoch": 1.6749011172021373, + "grad_norm": 1.6256166696548462, + "learning_rate": 0.00013068039013260663, + "loss": 0.7417, + "step": 24137 + }, + { + "epoch": 1.674970508639234, + "grad_norm": 1.7850574254989624, + "learning_rate": 0.00013067504243909013, + "loss": 0.6278, + "step": 24138 + }, + { + "epoch": 1.6750399000763307, + "grad_norm": 1.9028297662734985, + "learning_rate": 0.0001306696946487361, + "loss": 0.6503, + "step": 24139 + }, + { + "epoch": 1.6751092915134271, + "grad_norm": 0.9835538268089294, + "learning_rate": 0.0001306643467615614, + "loss": 0.5226, + "step": 24140 + }, + { + "epoch": 1.675178682950524, + "grad_norm": 1.0147309303283691, + "learning_rate": 0.0001306589987775829, + "loss": 0.5473, + "step": 24141 + }, + { + "epoch": 1.6752480743876206, + "grad_norm": 1.7808905839920044, + "learning_rate": 0.0001306536506968175, + "loss": 0.5879, + "step": 24142 + }, + { + "epoch": 1.6753174658247172, + "grad_norm": 1.560417652130127, + "learning_rate": 0.00013064830251928206, + "loss": 0.567, + "step": 24143 + }, + { + "epoch": 1.675386857261814, + "grad_norm": 2.589519739151001, + "learning_rate": 0.00013064295424499353, + "loss": 0.5368, + "step": 24144 + }, + { + "epoch": 1.6754562486989104, + "grad_norm": 1.9702017307281494, + "learning_rate": 0.00013063760587396871, + "loss": 0.6781, + "step": 24145 + }, + { + "epoch": 1.6755256401360072, + "grad_norm": 1.4091312885284424, + "learning_rate": 0.00013063225740622453, + "loss": 0.589, + "step": 24146 + }, + { + "epoch": 1.6755950315731039, + "grad_norm": 2.0859336853027344, + "learning_rate": 0.00013062690884177786, + "loss": 0.7177, + "step": 24147 + }, + { + "epoch": 1.6756644230102005, + "grad_norm": 1.141921043395996, + "learning_rate": 0.0001306215601806456, + "loss": 0.4049, + "step": 24148 + }, + { + "epoch": 1.6757338144472973, + "grad_norm": 1.8449630737304688, + "learning_rate": 0.00013061621142284466, + "loss": 0.5371, + "step": 24149 + }, + { + "epoch": 1.675803205884394, + "grad_norm": 1.7954447269439697, + "learning_rate": 0.00013061086256839185, + "loss": 0.7939, + "step": 24150 + }, + { + "epoch": 1.6758725973214905, + "grad_norm": 2.144557237625122, + "learning_rate": 0.00013060551361730412, + "loss": 0.6514, + "step": 24151 + }, + { + "epoch": 1.6759419887585874, + "grad_norm": 1.5300531387329102, + "learning_rate": 0.00013060016456959835, + "loss": 0.7061, + "step": 24152 + }, + { + "epoch": 1.6760113801956837, + "grad_norm": 2.034083604812622, + "learning_rate": 0.00013059481542529138, + "loss": 0.6518, + "step": 24153 + }, + { + "epoch": 1.6760807716327806, + "grad_norm": 1.3166152238845825, + "learning_rate": 0.00013058946618440016, + "loss": 0.4791, + "step": 24154 + }, + { + "epoch": 1.6761501630698772, + "grad_norm": 1.4955767393112183, + "learning_rate": 0.0001305841168469415, + "loss": 0.6141, + "step": 24155 + }, + { + "epoch": 1.6762195545069738, + "grad_norm": 1.6081974506378174, + "learning_rate": 0.00013057876741293236, + "loss": 0.5574, + "step": 24156 + }, + { + "epoch": 1.6762889459440706, + "grad_norm": 1.900011658668518, + "learning_rate": 0.0001305734178823896, + "loss": 0.7251, + "step": 24157 + }, + { + "epoch": 1.676358337381167, + "grad_norm": 2.040095567703247, + "learning_rate": 0.00013056806825533009, + "loss": 0.6715, + "step": 24158 + }, + { + "epoch": 1.6764277288182639, + "grad_norm": 1.3839815855026245, + "learning_rate": 0.00013056271853177076, + "loss": 0.5265, + "step": 24159 + }, + { + "epoch": 1.6764971202553605, + "grad_norm": 1.9314950704574585, + "learning_rate": 0.00013055736871172844, + "loss": 0.6418, + "step": 24160 + }, + { + "epoch": 1.676566511692457, + "grad_norm": 1.9513919353485107, + "learning_rate": 0.0001305520187952201, + "loss": 0.6424, + "step": 24161 + }, + { + "epoch": 1.676635903129554, + "grad_norm": 1.6550315618515015, + "learning_rate": 0.00013054666878226253, + "loss": 0.58, + "step": 24162 + }, + { + "epoch": 1.6767052945666505, + "grad_norm": 1.8797701597213745, + "learning_rate": 0.0001305413186728727, + "loss": 0.5894, + "step": 24163 + }, + { + "epoch": 1.6767746860037471, + "grad_norm": 1.9332693815231323, + "learning_rate": 0.00013053596846706746, + "loss": 0.6883, + "step": 24164 + }, + { + "epoch": 1.676844077440844, + "grad_norm": 1.6199288368225098, + "learning_rate": 0.0001305306181648637, + "loss": 0.482, + "step": 24165 + }, + { + "epoch": 1.6769134688779403, + "grad_norm": 1.414988398551941, + "learning_rate": 0.00013052526776627835, + "loss": 0.4392, + "step": 24166 + }, + { + "epoch": 1.6769828603150372, + "grad_norm": 1.708893895149231, + "learning_rate": 0.00013051991727132823, + "loss": 0.6228, + "step": 24167 + }, + { + "epoch": 1.6770522517521338, + "grad_norm": 2.0778400897979736, + "learning_rate": 0.00013051456668003032, + "loss": 0.6625, + "step": 24168 + }, + { + "epoch": 1.6771216431892304, + "grad_norm": 1.6661087274551392, + "learning_rate": 0.00013050921599240143, + "loss": 0.5519, + "step": 24169 + }, + { + "epoch": 1.6771910346263272, + "grad_norm": 1.9301146268844604, + "learning_rate": 0.0001305038652084585, + "loss": 0.5092, + "step": 24170 + }, + { + "epoch": 1.6772604260634236, + "grad_norm": 1.5516984462738037, + "learning_rate": 0.0001304985143282184, + "loss": 0.7508, + "step": 24171 + }, + { + "epoch": 1.6773298175005205, + "grad_norm": 1.8475687503814697, + "learning_rate": 0.00013049316335169805, + "loss": 0.6878, + "step": 24172 + }, + { + "epoch": 1.677399208937617, + "grad_norm": 1.585796594619751, + "learning_rate": 0.00013048781227891428, + "loss": 0.5761, + "step": 24173 + }, + { + "epoch": 1.6774686003747137, + "grad_norm": 1.8718661069869995, + "learning_rate": 0.00013048246110988407, + "loss": 0.8935, + "step": 24174 + }, + { + "epoch": 1.6775379918118105, + "grad_norm": 1.7483378648757935, + "learning_rate": 0.00013047710984462425, + "loss": 0.6833, + "step": 24175 + }, + { + "epoch": 1.6776073832489071, + "grad_norm": 1.5860570669174194, + "learning_rate": 0.00013047175848315172, + "loss": 0.4793, + "step": 24176 + }, + { + "epoch": 1.6776767746860037, + "grad_norm": 2.0124099254608154, + "learning_rate": 0.0001304664070254834, + "loss": 0.6418, + "step": 24177 + }, + { + "epoch": 1.6777461661231006, + "grad_norm": 2.3112545013427734, + "learning_rate": 0.00013046105547163614, + "loss": 0.6057, + "step": 24178 + }, + { + "epoch": 1.677815557560197, + "grad_norm": 2.014646053314209, + "learning_rate": 0.00013045570382162687, + "loss": 0.5504, + "step": 24179 + }, + { + "epoch": 1.6778849489972938, + "grad_norm": 1.459521770477295, + "learning_rate": 0.00013045035207547248, + "loss": 0.5364, + "step": 24180 + }, + { + "epoch": 1.6779543404343904, + "grad_norm": 1.4936414957046509, + "learning_rate": 0.00013044500023318988, + "loss": 0.6836, + "step": 24181 + }, + { + "epoch": 1.678023731871487, + "grad_norm": 1.9772554636001587, + "learning_rate": 0.00013043964829479593, + "loss": 0.7006, + "step": 24182 + }, + { + "epoch": 1.6780931233085838, + "grad_norm": 2.1181955337524414, + "learning_rate": 0.00013043429626030754, + "loss": 0.8243, + "step": 24183 + }, + { + "epoch": 1.6781625147456802, + "grad_norm": 1.6815403699874878, + "learning_rate": 0.0001304289441297416, + "loss": 0.992, + "step": 24184 + }, + { + "epoch": 1.678231906182777, + "grad_norm": 1.4309583902359009, + "learning_rate": 0.00013042359190311502, + "loss": 0.6258, + "step": 24185 + }, + { + "epoch": 1.6783012976198737, + "grad_norm": 1.3835995197296143, + "learning_rate": 0.00013041823958044471, + "loss": 0.5258, + "step": 24186 + }, + { + "epoch": 1.6783706890569703, + "grad_norm": 3.010798215866089, + "learning_rate": 0.00013041288716174755, + "loss": 0.5126, + "step": 24187 + }, + { + "epoch": 1.6784400804940671, + "grad_norm": 1.5419241189956665, + "learning_rate": 0.0001304075346470404, + "loss": 0.5878, + "step": 24188 + }, + { + "epoch": 1.6785094719311637, + "grad_norm": 1.670186161994934, + "learning_rate": 0.00013040218203634021, + "loss": 0.4641, + "step": 24189 + }, + { + "epoch": 1.6785788633682603, + "grad_norm": 1.686379075050354, + "learning_rate": 0.00013039682932966385, + "loss": 0.7858, + "step": 24190 + }, + { + "epoch": 1.6786482548053572, + "grad_norm": 1.9167238473892212, + "learning_rate": 0.00013039147652702822, + "loss": 0.6233, + "step": 24191 + }, + { + "epoch": 1.6787176462424536, + "grad_norm": 1.365656852722168, + "learning_rate": 0.00013038612362845027, + "loss": 0.4268, + "step": 24192 + }, + { + "epoch": 1.6787870376795504, + "grad_norm": 2.1550228595733643, + "learning_rate": 0.0001303807706339468, + "loss": 0.3305, + "step": 24193 + }, + { + "epoch": 1.678856429116647, + "grad_norm": 1.368327021598816, + "learning_rate": 0.0001303754175435348, + "loss": 0.6321, + "step": 24194 + }, + { + "epoch": 1.6789258205537436, + "grad_norm": 2.173576831817627, + "learning_rate": 0.00013037006435723112, + "loss": 0.6623, + "step": 24195 + }, + { + "epoch": 1.6789952119908405, + "grad_norm": 1.5601943731307983, + "learning_rate": 0.00013036471107505265, + "loss": 0.6281, + "step": 24196 + }, + { + "epoch": 1.6790646034279368, + "grad_norm": 1.7668112516403198, + "learning_rate": 0.00013035935769701632, + "loss": 0.6277, + "step": 24197 + }, + { + "epoch": 1.6791339948650337, + "grad_norm": 1.7521899938583374, + "learning_rate": 0.00013035400422313902, + "loss": 0.6357, + "step": 24198 + }, + { + "epoch": 1.6792033863021303, + "grad_norm": 1.72926664352417, + "learning_rate": 0.00013034865065343767, + "loss": 0.4464, + "step": 24199 + }, + { + "epoch": 1.679272777739227, + "grad_norm": 2.326519012451172, + "learning_rate": 0.00013034329698792912, + "loss": 0.7064, + "step": 24200 + }, + { + "epoch": 1.6793421691763237, + "grad_norm": 1.6433384418487549, + "learning_rate": 0.0001303379432266303, + "loss": 0.5468, + "step": 24201 + }, + { + "epoch": 1.6794115606134203, + "grad_norm": 1.484316110610962, + "learning_rate": 0.00013033258936955813, + "loss": 0.4756, + "step": 24202 + }, + { + "epoch": 1.679480952050517, + "grad_norm": 1.5353416204452515, + "learning_rate": 0.00013032723541672953, + "loss": 0.6984, + "step": 24203 + }, + { + "epoch": 1.6795503434876138, + "grad_norm": 1.4572511911392212, + "learning_rate": 0.0001303218813681613, + "loss": 0.6959, + "step": 24204 + }, + { + "epoch": 1.6796197349247102, + "grad_norm": 1.9291396141052246, + "learning_rate": 0.00013031652722387044, + "loss": 0.5293, + "step": 24205 + }, + { + "epoch": 1.679689126361807, + "grad_norm": 1.915462613105774, + "learning_rate": 0.00013031117298387384, + "loss": 0.8976, + "step": 24206 + }, + { + "epoch": 1.6797585177989036, + "grad_norm": 1.599625825881958, + "learning_rate": 0.00013030581864818835, + "loss": 0.6312, + "step": 24207 + }, + { + "epoch": 1.6798279092360002, + "grad_norm": 1.6158204078674316, + "learning_rate": 0.00013030046421683093, + "loss": 0.4612, + "step": 24208 + }, + { + "epoch": 1.679897300673097, + "grad_norm": 1.8105924129486084, + "learning_rate": 0.00013029510968981845, + "loss": 0.6139, + "step": 24209 + }, + { + "epoch": 1.6799666921101934, + "grad_norm": 1.4509321451187134, + "learning_rate": 0.00013028975506716782, + "loss": 0.5216, + "step": 24210 + }, + { + "epoch": 1.6800360835472903, + "grad_norm": 1.5783745050430298, + "learning_rate": 0.00013028440034889598, + "loss": 0.5758, + "step": 24211 + }, + { + "epoch": 1.680105474984387, + "grad_norm": 2.2594480514526367, + "learning_rate": 0.00013027904553501976, + "loss": 0.4785, + "step": 24212 + }, + { + "epoch": 1.6801748664214835, + "grad_norm": 1.4880423545837402, + "learning_rate": 0.00013027369062555612, + "loss": 0.5287, + "step": 24213 + }, + { + "epoch": 1.6802442578585803, + "grad_norm": 2.2036514282226562, + "learning_rate": 0.00013026833562052197, + "loss": 0.5897, + "step": 24214 + }, + { + "epoch": 1.680313649295677, + "grad_norm": 1.4573034048080444, + "learning_rate": 0.00013026298051993415, + "loss": 0.5453, + "step": 24215 + }, + { + "epoch": 1.6803830407327736, + "grad_norm": 1.5557547807693481, + "learning_rate": 0.00013025762532380963, + "loss": 0.4869, + "step": 24216 + }, + { + "epoch": 1.6804524321698704, + "grad_norm": 1.6008033752441406, + "learning_rate": 0.00013025227003216533, + "loss": 0.5451, + "step": 24217 + }, + { + "epoch": 1.6805218236069668, + "grad_norm": 1.6623964309692383, + "learning_rate": 0.00013024691464501808, + "loss": 0.5168, + "step": 24218 + }, + { + "epoch": 1.6805912150440636, + "grad_norm": 1.781006932258606, + "learning_rate": 0.00013024155916238484, + "loss": 0.421, + "step": 24219 + }, + { + "epoch": 1.6806606064811602, + "grad_norm": 1.7876405715942383, + "learning_rate": 0.00013023620358428253, + "loss": 0.6678, + "step": 24220 + }, + { + "epoch": 1.6807299979182568, + "grad_norm": 1.590355396270752, + "learning_rate": 0.00013023084791072802, + "loss": 0.483, + "step": 24221 + }, + { + "epoch": 1.6807993893553537, + "grad_norm": 1.5893114805221558, + "learning_rate": 0.00013022549214173823, + "loss": 0.63, + "step": 24222 + }, + { + "epoch": 1.68086878079245, + "grad_norm": 1.547316312789917, + "learning_rate": 0.00013022013627733006, + "loss": 0.4664, + "step": 24223 + }, + { + "epoch": 1.680938172229547, + "grad_norm": 2.203160047531128, + "learning_rate": 0.00013021478031752045, + "loss": 0.5318, + "step": 24224 + }, + { + "epoch": 1.6810075636666435, + "grad_norm": 2.2998673915863037, + "learning_rate": 0.00013020942426232625, + "loss": 0.752, + "step": 24225 + }, + { + "epoch": 1.6810769551037401, + "grad_norm": 1.327335238456726, + "learning_rate": 0.0001302040681117644, + "loss": 0.4691, + "step": 24226 + }, + { + "epoch": 1.681146346540837, + "grad_norm": 1.1559984683990479, + "learning_rate": 0.00013019871186585187, + "loss": 0.4998, + "step": 24227 + }, + { + "epoch": 1.6812157379779336, + "grad_norm": 2.069688558578491, + "learning_rate": 0.0001301933555246055, + "loss": 0.8198, + "step": 24228 + }, + { + "epoch": 1.6812851294150302, + "grad_norm": 2.0017764568328857, + "learning_rate": 0.00013018799908804215, + "loss": 0.7999, + "step": 24229 + }, + { + "epoch": 1.681354520852127, + "grad_norm": 1.8092161417007446, + "learning_rate": 0.00013018264255617882, + "loss": 0.6572, + "step": 24230 + }, + { + "epoch": 1.6814239122892234, + "grad_norm": 1.315791368484497, + "learning_rate": 0.0001301772859290324, + "loss": 0.4338, + "step": 24231 + }, + { + "epoch": 1.6814933037263202, + "grad_norm": 1.60601007938385, + "learning_rate": 0.00013017192920661977, + "loss": 0.6071, + "step": 24232 + }, + { + "epoch": 1.6815626951634168, + "grad_norm": 1.3683921098709106, + "learning_rate": 0.00013016657238895787, + "loss": 0.5537, + "step": 24233 + }, + { + "epoch": 1.6816320866005134, + "grad_norm": 1.9172987937927246, + "learning_rate": 0.00013016121547606362, + "loss": 0.869, + "step": 24234 + }, + { + "epoch": 1.6817014780376103, + "grad_norm": 1.5884348154067993, + "learning_rate": 0.00013015585846795386, + "loss": 0.6439, + "step": 24235 + }, + { + "epoch": 1.6817708694747067, + "grad_norm": 1.798487901687622, + "learning_rate": 0.0001301505013646456, + "loss": 0.5179, + "step": 24236 + }, + { + "epoch": 1.6818402609118035, + "grad_norm": 1.4169955253601074, + "learning_rate": 0.00013014514416615568, + "loss": 0.6941, + "step": 24237 + }, + { + "epoch": 1.6819096523489, + "grad_norm": 1.7684612274169922, + "learning_rate": 0.00013013978687250106, + "loss": 0.4764, + "step": 24238 + }, + { + "epoch": 1.6819790437859967, + "grad_norm": 2.7817063331604004, + "learning_rate": 0.00013013442948369862, + "loss": 0.8559, + "step": 24239 + }, + { + "epoch": 1.6820484352230936, + "grad_norm": 1.6090989112854004, + "learning_rate": 0.00013012907199976527, + "loss": 0.6164, + "step": 24240 + }, + { + "epoch": 1.6821178266601902, + "grad_norm": 3.1524322032928467, + "learning_rate": 0.00013012371442071795, + "loss": 0.7682, + "step": 24241 + }, + { + "epoch": 1.6821872180972868, + "grad_norm": 2.20493483543396, + "learning_rate": 0.00013011835674657356, + "loss": 0.7324, + "step": 24242 + }, + { + "epoch": 1.6822566095343836, + "grad_norm": 1.736346960067749, + "learning_rate": 0.000130112998977349, + "loss": 0.6369, + "step": 24243 + }, + { + "epoch": 1.68232600097148, + "grad_norm": 1.8199275732040405, + "learning_rate": 0.0001301076411130612, + "loss": 0.8837, + "step": 24244 + }, + { + "epoch": 1.6823953924085768, + "grad_norm": 1.5384414196014404, + "learning_rate": 0.00013010228315372708, + "loss": 0.5305, + "step": 24245 + }, + { + "epoch": 1.6824647838456734, + "grad_norm": 2.4889941215515137, + "learning_rate": 0.00013009692509936355, + "loss": 0.6021, + "step": 24246 + }, + { + "epoch": 1.68253417528277, + "grad_norm": 1.392497181892395, + "learning_rate": 0.00013009156694998748, + "loss": 0.6066, + "step": 24247 + }, + { + "epoch": 1.6826035667198669, + "grad_norm": 1.528563141822815, + "learning_rate": 0.00013008620870561587, + "loss": 0.5592, + "step": 24248 + }, + { + "epoch": 1.6826729581569633, + "grad_norm": 1.5487427711486816, + "learning_rate": 0.00013008085036626558, + "loss": 0.6818, + "step": 24249 + }, + { + "epoch": 1.68274234959406, + "grad_norm": 2.54909086227417, + "learning_rate": 0.0001300754919319535, + "loss": 0.8722, + "step": 24250 + }, + { + "epoch": 1.6828117410311567, + "grad_norm": 1.5398813486099243, + "learning_rate": 0.00013007013340269664, + "loss": 0.6568, + "step": 24251 + }, + { + "epoch": 1.6828811324682533, + "grad_norm": 2.591817617416382, + "learning_rate": 0.00013006477477851182, + "loss": 0.8047, + "step": 24252 + }, + { + "epoch": 1.6829505239053502, + "grad_norm": 1.3008084297180176, + "learning_rate": 0.000130059416059416, + "loss": 0.5836, + "step": 24253 + }, + { + "epoch": 1.6830199153424468, + "grad_norm": 2.3446462154388428, + "learning_rate": 0.00013005405724542606, + "loss": 0.8005, + "step": 24254 + }, + { + "epoch": 1.6830893067795434, + "grad_norm": 1.513707160949707, + "learning_rate": 0.00013004869833655903, + "loss": 0.5911, + "step": 24255 + }, + { + "epoch": 1.6831586982166402, + "grad_norm": 1.2458837032318115, + "learning_rate": 0.00013004333933283165, + "loss": 0.6296, + "step": 24256 + }, + { + "epoch": 1.6832280896537366, + "grad_norm": 2.110170841217041, + "learning_rate": 0.000130037980234261, + "loss": 0.6472, + "step": 24257 + }, + { + "epoch": 1.6832974810908334, + "grad_norm": 1.8617349863052368, + "learning_rate": 0.00013003262104086392, + "loss": 0.6696, + "step": 24258 + }, + { + "epoch": 1.68336687252793, + "grad_norm": 1.1301807165145874, + "learning_rate": 0.00013002726175265732, + "loss": 0.5324, + "step": 24259 + }, + { + "epoch": 1.6834362639650267, + "grad_norm": 2.0736196041107178, + "learning_rate": 0.00013002190236965816, + "loss": 0.7368, + "step": 24260 + }, + { + "epoch": 1.6835056554021235, + "grad_norm": 1.1239393949508667, + "learning_rate": 0.00013001654289188333, + "loss": 0.5245, + "step": 24261 + }, + { + "epoch": 1.6835750468392199, + "grad_norm": 1.823171854019165, + "learning_rate": 0.00013001118331934974, + "loss": 0.4497, + "step": 24262 + }, + { + "epoch": 1.6836444382763167, + "grad_norm": 1.4940316677093506, + "learning_rate": 0.00013000582365207435, + "loss": 0.4187, + "step": 24263 + }, + { + "epoch": 1.6837138297134133, + "grad_norm": 1.995030403137207, + "learning_rate": 0.00013000046389007402, + "loss": 0.4098, + "step": 24264 + }, + { + "epoch": 1.68378322115051, + "grad_norm": 1.492631196975708, + "learning_rate": 0.00012999510403336573, + "loss": 0.4873, + "step": 24265 + }, + { + "epoch": 1.6838526125876068, + "grad_norm": 1.5805796384811401, + "learning_rate": 0.0001299897440819664, + "loss": 0.683, + "step": 24266 + }, + { + "epoch": 1.6839220040247034, + "grad_norm": 1.7755954265594482, + "learning_rate": 0.00012998438403589288, + "loss": 0.576, + "step": 24267 + }, + { + "epoch": 1.6839913954618, + "grad_norm": 1.9144757986068726, + "learning_rate": 0.0001299790238951622, + "loss": 0.7834, + "step": 24268 + }, + { + "epoch": 1.6840607868988968, + "grad_norm": 1.6382967233657837, + "learning_rate": 0.00012997366365979115, + "loss": 0.5321, + "step": 24269 + }, + { + "epoch": 1.6841301783359932, + "grad_norm": 1.4557114839553833, + "learning_rate": 0.00012996830332979677, + "loss": 0.6836, + "step": 24270 + }, + { + "epoch": 1.68419956977309, + "grad_norm": 1.7314789295196533, + "learning_rate": 0.0001299629429051959, + "loss": 0.5841, + "step": 24271 + }, + { + "epoch": 1.6842689612101867, + "grad_norm": 6.494709014892578, + "learning_rate": 0.00012995758238600554, + "loss": 0.5918, + "step": 24272 + }, + { + "epoch": 1.6843383526472833, + "grad_norm": 1.5435879230499268, + "learning_rate": 0.00012995222177224252, + "loss": 0.7781, + "step": 24273 + }, + { + "epoch": 1.68440774408438, + "grad_norm": 1.9308885335922241, + "learning_rate": 0.00012994686106392384, + "loss": 0.6576, + "step": 24274 + }, + { + "epoch": 1.6844771355214765, + "grad_norm": 1.7049683332443237, + "learning_rate": 0.00012994150026106639, + "loss": 0.7477, + "step": 24275 + }, + { + "epoch": 1.6845465269585733, + "grad_norm": 1.7374085187911987, + "learning_rate": 0.0001299361393636871, + "loss": 0.6789, + "step": 24276 + }, + { + "epoch": 1.68461591839567, + "grad_norm": 2.1758923530578613, + "learning_rate": 0.00012993077837180286, + "loss": 0.5325, + "step": 24277 + }, + { + "epoch": 1.6846853098327665, + "grad_norm": 1.7669074535369873, + "learning_rate": 0.00012992541728543063, + "loss": 0.5127, + "step": 24278 + }, + { + "epoch": 1.6847547012698634, + "grad_norm": 1.8167582750320435, + "learning_rate": 0.00012992005610458737, + "loss": 0.4752, + "step": 24279 + }, + { + "epoch": 1.68482409270696, + "grad_norm": 1.6695654392242432, + "learning_rate": 0.00012991469482928996, + "loss": 0.5381, + "step": 24280 + }, + { + "epoch": 1.6848934841440566, + "grad_norm": 1.8029654026031494, + "learning_rate": 0.00012990933345955532, + "loss": 0.7055, + "step": 24281 + }, + { + "epoch": 1.6849628755811534, + "grad_norm": 2.164760112762451, + "learning_rate": 0.00012990397199540038, + "loss": 0.7317, + "step": 24282 + }, + { + "epoch": 1.6850322670182498, + "grad_norm": 1.6427723169326782, + "learning_rate": 0.00012989861043684206, + "loss": 0.7133, + "step": 24283 + }, + { + "epoch": 1.6851016584553467, + "grad_norm": 1.5855931043624878, + "learning_rate": 0.0001298932487838973, + "loss": 0.5406, + "step": 24284 + }, + { + "epoch": 1.6851710498924433, + "grad_norm": 1.6068477630615234, + "learning_rate": 0.00012988788703658305, + "loss": 0.6313, + "step": 24285 + }, + { + "epoch": 1.6852404413295399, + "grad_norm": 1.8116192817687988, + "learning_rate": 0.00012988252519491618, + "loss": 0.6427, + "step": 24286 + }, + { + "epoch": 1.6853098327666367, + "grad_norm": 1.498313546180725, + "learning_rate": 0.00012987716325891363, + "loss": 0.6614, + "step": 24287 + }, + { + "epoch": 1.685379224203733, + "grad_norm": 1.7183879613876343, + "learning_rate": 0.00012987180122859236, + "loss": 0.7077, + "step": 24288 + }, + { + "epoch": 1.68544861564083, + "grad_norm": 1.5246446132659912, + "learning_rate": 0.0001298664391039693, + "loss": 0.4888, + "step": 24289 + }, + { + "epoch": 1.6855180070779265, + "grad_norm": 1.7697234153747559, + "learning_rate": 0.0001298610768850613, + "loss": 0.6098, + "step": 24290 + }, + { + "epoch": 1.6855873985150231, + "grad_norm": 2.0569653511047363, + "learning_rate": 0.0001298557145718854, + "loss": 0.7921, + "step": 24291 + }, + { + "epoch": 1.68565678995212, + "grad_norm": 1.7332489490509033, + "learning_rate": 0.00012985035216445844, + "loss": 0.6627, + "step": 24292 + }, + { + "epoch": 1.6857261813892166, + "grad_norm": 1.3051782846450806, + "learning_rate": 0.0001298449896627974, + "loss": 0.6596, + "step": 24293 + }, + { + "epoch": 1.6857955728263132, + "grad_norm": 2.420175790786743, + "learning_rate": 0.0001298396270669192, + "loss": 0.4882, + "step": 24294 + }, + { + "epoch": 1.68586496426341, + "grad_norm": 2.3359742164611816, + "learning_rate": 0.0001298342643768407, + "loss": 0.7451, + "step": 24295 + }, + { + "epoch": 1.6859343557005064, + "grad_norm": 2.0444328784942627, + "learning_rate": 0.00012982890159257896, + "loss": 0.7683, + "step": 24296 + }, + { + "epoch": 1.6860037471376033, + "grad_norm": 1.1372166872024536, + "learning_rate": 0.00012982353871415084, + "loss": 0.3286, + "step": 24297 + }, + { + "epoch": 1.6860731385746999, + "grad_norm": 1.3089948892593384, + "learning_rate": 0.00012981817574157325, + "loss": 0.5328, + "step": 24298 + }, + { + "epoch": 1.6861425300117965, + "grad_norm": 1.9452615976333618, + "learning_rate": 0.00012981281267486312, + "loss": 0.6935, + "step": 24299 + }, + { + "epoch": 1.6862119214488933, + "grad_norm": 2.396327018737793, + "learning_rate": 0.00012980744951403741, + "loss": 0.638, + "step": 24300 + }, + { + "epoch": 1.6862813128859897, + "grad_norm": 1.567166805267334, + "learning_rate": 0.00012980208625911304, + "loss": 0.6129, + "step": 24301 + }, + { + "epoch": 1.6863507043230865, + "grad_norm": 1.9095344543457031, + "learning_rate": 0.00012979672291010691, + "loss": 0.833, + "step": 24302 + }, + { + "epoch": 1.6864200957601831, + "grad_norm": 1.9358038902282715, + "learning_rate": 0.00012979135946703606, + "loss": 0.5526, + "step": 24303 + }, + { + "epoch": 1.6864894871972798, + "grad_norm": 1.5063613653182983, + "learning_rate": 0.0001297859959299173, + "loss": 0.5195, + "step": 24304 + }, + { + "epoch": 1.6865588786343766, + "grad_norm": 1.5146323442459106, + "learning_rate": 0.0001297806322987676, + "loss": 0.612, + "step": 24305 + }, + { + "epoch": 1.6866282700714732, + "grad_norm": 1.58100426197052, + "learning_rate": 0.00012977526857360386, + "loss": 0.5571, + "step": 24306 + }, + { + "epoch": 1.6866976615085698, + "grad_norm": 1.3575204610824585, + "learning_rate": 0.00012976990475444312, + "loss": 0.5827, + "step": 24307 + }, + { + "epoch": 1.6867670529456666, + "grad_norm": 1.9822266101837158, + "learning_rate": 0.00012976454084130223, + "loss": 0.6221, + "step": 24308 + }, + { + "epoch": 1.686836444382763, + "grad_norm": 1.6411409378051758, + "learning_rate": 0.00012975917683419813, + "loss": 0.7022, + "step": 24309 + }, + { + "epoch": 1.6869058358198599, + "grad_norm": 1.3503711223602295, + "learning_rate": 0.00012975381273314774, + "loss": 0.4843, + "step": 24310 + }, + { + "epoch": 1.6869752272569565, + "grad_norm": 2.0312817096710205, + "learning_rate": 0.00012974844853816804, + "loss": 0.4952, + "step": 24311 + }, + { + "epoch": 1.687044618694053, + "grad_norm": 1.1816811561584473, + "learning_rate": 0.00012974308424927594, + "loss": 0.5673, + "step": 24312 + }, + { + "epoch": 1.68711401013115, + "grad_norm": 1.234460473060608, + "learning_rate": 0.00012973771986648838, + "loss": 0.2751, + "step": 24313 + }, + { + "epoch": 1.6871834015682465, + "grad_norm": 1.4631831645965576, + "learning_rate": 0.00012973235538982225, + "loss": 0.5836, + "step": 24314 + }, + { + "epoch": 1.6872527930053431, + "grad_norm": 1.3872963190078735, + "learning_rate": 0.00012972699081929457, + "loss": 0.621, + "step": 24315 + }, + { + "epoch": 1.6873221844424398, + "grad_norm": 1.8342915773391724, + "learning_rate": 0.0001297216261549222, + "loss": 0.5791, + "step": 24316 + }, + { + "epoch": 1.6873915758795364, + "grad_norm": 5.4007673263549805, + "learning_rate": 0.00012971626139672213, + "loss": 0.6301, + "step": 24317 + }, + { + "epoch": 1.6874609673166332, + "grad_norm": 1.2173981666564941, + "learning_rate": 0.00012971089654471125, + "loss": 0.5155, + "step": 24318 + }, + { + "epoch": 1.6875303587537298, + "grad_norm": 4.8567352294921875, + "learning_rate": 0.0001297055315989065, + "loss": 0.6516, + "step": 24319 + }, + { + "epoch": 1.6875997501908264, + "grad_norm": 1.4326015710830688, + "learning_rate": 0.00012970016655932488, + "loss": 0.4774, + "step": 24320 + }, + { + "epoch": 1.6876691416279233, + "grad_norm": 2.9457952976226807, + "learning_rate": 0.00012969480142598324, + "loss": 0.5983, + "step": 24321 + }, + { + "epoch": 1.6877385330650196, + "grad_norm": 1.698500394821167, + "learning_rate": 0.00012968943619889856, + "loss": 0.6753, + "step": 24322 + }, + { + "epoch": 1.6878079245021165, + "grad_norm": 2.621652603149414, + "learning_rate": 0.0001296840708780878, + "loss": 0.4831, + "step": 24323 + }, + { + "epoch": 1.687877315939213, + "grad_norm": 1.7793283462524414, + "learning_rate": 0.00012967870546356787, + "loss": 0.5273, + "step": 24324 + }, + { + "epoch": 1.6879467073763097, + "grad_norm": 1.7639641761779785, + "learning_rate": 0.00012967333995535569, + "loss": 0.6562, + "step": 24325 + }, + { + "epoch": 1.6880160988134065, + "grad_norm": 2.519608736038208, + "learning_rate": 0.00012966797435346822, + "loss": 0.5009, + "step": 24326 + }, + { + "epoch": 1.6880854902505031, + "grad_norm": 1.7235959768295288, + "learning_rate": 0.0001296626086579224, + "loss": 0.4776, + "step": 24327 + }, + { + "epoch": 1.6881548816875998, + "grad_norm": 1.7466425895690918, + "learning_rate": 0.00012965724286873517, + "loss": 0.5784, + "step": 24328 + }, + { + "epoch": 1.6882242731246964, + "grad_norm": 1.7427297830581665, + "learning_rate": 0.00012965187698592347, + "loss": 0.6516, + "step": 24329 + }, + { + "epoch": 1.688293664561793, + "grad_norm": 2.0938034057617188, + "learning_rate": 0.00012964651100950421, + "loss": 0.339, + "step": 24330 + }, + { + "epoch": 1.6883630559988898, + "grad_norm": 2.6448750495910645, + "learning_rate": 0.0001296411449394944, + "loss": 0.6985, + "step": 24331 + }, + { + "epoch": 1.6884324474359864, + "grad_norm": 1.9580988883972168, + "learning_rate": 0.0001296357787759109, + "loss": 0.6743, + "step": 24332 + }, + { + "epoch": 1.688501838873083, + "grad_norm": 1.610191822052002, + "learning_rate": 0.0001296304125187707, + "loss": 0.685, + "step": 24333 + }, + { + "epoch": 1.6885712303101799, + "grad_norm": 1.8529691696166992, + "learning_rate": 0.00012962504616809072, + "loss": 0.5863, + "step": 24334 + }, + { + "epoch": 1.6886406217472762, + "grad_norm": 1.6505540609359741, + "learning_rate": 0.00012961967972388791, + "loss": 0.7384, + "step": 24335 + }, + { + "epoch": 1.688710013184373, + "grad_norm": 1.759591817855835, + "learning_rate": 0.0001296143131861792, + "loss": 0.4251, + "step": 24336 + }, + { + "epoch": 1.6887794046214697, + "grad_norm": 3.159728527069092, + "learning_rate": 0.00012960894655498155, + "loss": 0.8596, + "step": 24337 + }, + { + "epoch": 1.6888487960585663, + "grad_norm": 1.9368995428085327, + "learning_rate": 0.00012960357983031188, + "loss": 0.5429, + "step": 24338 + }, + { + "epoch": 1.6889181874956631, + "grad_norm": 5.268118381500244, + "learning_rate": 0.00012959821301218716, + "loss": 0.6963, + "step": 24339 + }, + { + "epoch": 1.6889875789327597, + "grad_norm": 1.2909849882125854, + "learning_rate": 0.00012959284610062428, + "loss": 0.455, + "step": 24340 + }, + { + "epoch": 1.6890569703698564, + "grad_norm": 1.6206650733947754, + "learning_rate": 0.00012958747909564027, + "loss": 0.5504, + "step": 24341 + }, + { + "epoch": 1.689126361806953, + "grad_norm": 1.3879426717758179, + "learning_rate": 0.00012958211199725196, + "loss": 0.4919, + "step": 24342 + }, + { + "epoch": 1.6891957532440496, + "grad_norm": 1.7987109422683716, + "learning_rate": 0.0001295767448054764, + "loss": 0.5844, + "step": 24343 + }, + { + "epoch": 1.6892651446811464, + "grad_norm": 2.572561740875244, + "learning_rate": 0.00012957137752033046, + "loss": 0.678, + "step": 24344 + }, + { + "epoch": 1.689334536118243, + "grad_norm": 5.292041301727295, + "learning_rate": 0.00012956601014183113, + "loss": 0.5815, + "step": 24345 + }, + { + "epoch": 1.6894039275553396, + "grad_norm": 1.9733866453170776, + "learning_rate": 0.00012956064266999532, + "loss": 0.679, + "step": 24346 + }, + { + "epoch": 1.6894733189924365, + "grad_norm": 1.8473433256149292, + "learning_rate": 0.00012955527510484, + "loss": 0.8747, + "step": 24347 + }, + { + "epoch": 1.6895427104295329, + "grad_norm": 2.3378896713256836, + "learning_rate": 0.00012954990744638213, + "loss": 0.8017, + "step": 24348 + }, + { + "epoch": 1.6896121018666297, + "grad_norm": 1.7724390029907227, + "learning_rate": 0.0001295445396946386, + "loss": 0.68, + "step": 24349 + }, + { + "epoch": 1.6896814933037263, + "grad_norm": 1.5071088075637817, + "learning_rate": 0.0001295391718496264, + "loss": 0.6838, + "step": 24350 + }, + { + "epoch": 1.689750884740823, + "grad_norm": 1.9800547361373901, + "learning_rate": 0.00012953380391136245, + "loss": 0.6599, + "step": 24351 + }, + { + "epoch": 1.6898202761779197, + "grad_norm": 1.4737547636032104, + "learning_rate": 0.00012952843587986373, + "loss": 0.4945, + "step": 24352 + }, + { + "epoch": 1.6898896676150164, + "grad_norm": 2.093879461288452, + "learning_rate": 0.00012952306775514716, + "loss": 0.8725, + "step": 24353 + }, + { + "epoch": 1.689959059052113, + "grad_norm": 1.432090401649475, + "learning_rate": 0.00012951769953722968, + "loss": 0.5318, + "step": 24354 + }, + { + "epoch": 1.6900284504892096, + "grad_norm": 1.9624955654144287, + "learning_rate": 0.00012951233122612826, + "loss": 0.5098, + "step": 24355 + }, + { + "epoch": 1.6900978419263062, + "grad_norm": 2.3675050735473633, + "learning_rate": 0.0001295069628218598, + "loss": 0.5423, + "step": 24356 + }, + { + "epoch": 1.690167233363403, + "grad_norm": 1.6846750974655151, + "learning_rate": 0.0001295015943244413, + "loss": 0.6067, + "step": 24357 + }, + { + "epoch": 1.6902366248004996, + "grad_norm": 1.8294416666030884, + "learning_rate": 0.0001294962257338897, + "loss": 0.6863, + "step": 24358 + }, + { + "epoch": 1.6903060162375962, + "grad_norm": 1.3959118127822876, + "learning_rate": 0.00012949085705022192, + "loss": 0.6624, + "step": 24359 + }, + { + "epoch": 1.690375407674693, + "grad_norm": 1.8778318166732788, + "learning_rate": 0.00012948548827345494, + "loss": 0.7357, + "step": 24360 + }, + { + "epoch": 1.6904447991117895, + "grad_norm": 2.1412031650543213, + "learning_rate": 0.00012948011940360567, + "loss": 0.7873, + "step": 24361 + }, + { + "epoch": 1.6905141905488863, + "grad_norm": 1.8234504461288452, + "learning_rate": 0.00012947475044069112, + "loss": 0.5772, + "step": 24362 + }, + { + "epoch": 1.690583581985983, + "grad_norm": 1.6519696712493896, + "learning_rate": 0.00012946938138472816, + "loss": 0.6007, + "step": 24363 + }, + { + "epoch": 1.6906529734230795, + "grad_norm": 1.9578092098236084, + "learning_rate": 0.00012946401223573384, + "loss": 0.6181, + "step": 24364 + }, + { + "epoch": 1.6907223648601764, + "grad_norm": 2.012669801712036, + "learning_rate": 0.00012945864299372502, + "loss": 0.5463, + "step": 24365 + }, + { + "epoch": 1.690791756297273, + "grad_norm": 2.671938419342041, + "learning_rate": 0.00012945327365871867, + "loss": 0.3709, + "step": 24366 + }, + { + "epoch": 1.6908611477343696, + "grad_norm": 1.4548982381820679, + "learning_rate": 0.00012944790423073178, + "loss": 0.6626, + "step": 24367 + }, + { + "epoch": 1.6909305391714664, + "grad_norm": 1.5578070878982544, + "learning_rate": 0.00012944253470978124, + "loss": 0.6217, + "step": 24368 + }, + { + "epoch": 1.6909999306085628, + "grad_norm": 1.6034703254699707, + "learning_rate": 0.00012943716509588405, + "loss": 0.5368, + "step": 24369 + }, + { + "epoch": 1.6910693220456596, + "grad_norm": 1.7154912948608398, + "learning_rate": 0.00012943179538905716, + "loss": 0.7665, + "step": 24370 + }, + { + "epoch": 1.6911387134827562, + "grad_norm": 1.4415696859359741, + "learning_rate": 0.00012942642558931748, + "loss": 0.5023, + "step": 24371 + }, + { + "epoch": 1.6912081049198529, + "grad_norm": 1.594553828239441, + "learning_rate": 0.00012942105569668203, + "loss": 0.4869, + "step": 24372 + }, + { + "epoch": 1.6912774963569497, + "grad_norm": 1.4496209621429443, + "learning_rate": 0.00012941568571116766, + "loss": 0.4907, + "step": 24373 + }, + { + "epoch": 1.691346887794046, + "grad_norm": 3.1556200981140137, + "learning_rate": 0.0001294103156327914, + "loss": 0.7544, + "step": 24374 + }, + { + "epoch": 1.691416279231143, + "grad_norm": 1.392149806022644, + "learning_rate": 0.00012940494546157024, + "loss": 0.6821, + "step": 24375 + }, + { + "epoch": 1.6914856706682395, + "grad_norm": 1.4109876155853271, + "learning_rate": 0.00012939957519752103, + "loss": 0.6201, + "step": 24376 + }, + { + "epoch": 1.6915550621053361, + "grad_norm": 1.9565767049789429, + "learning_rate": 0.00012939420484066075, + "loss": 0.5409, + "step": 24377 + }, + { + "epoch": 1.691624453542433, + "grad_norm": 1.5649594068527222, + "learning_rate": 0.00012938883439100643, + "loss": 0.5759, + "step": 24378 + }, + { + "epoch": 1.6916938449795296, + "grad_norm": 1.5161776542663574, + "learning_rate": 0.00012938346384857495, + "loss": 0.5328, + "step": 24379 + }, + { + "epoch": 1.6917632364166262, + "grad_norm": 4.221161365509033, + "learning_rate": 0.00012937809321338328, + "loss": 0.6265, + "step": 24380 + }, + { + "epoch": 1.691832627853723, + "grad_norm": 2.336066484451294, + "learning_rate": 0.00012937272248544836, + "loss": 0.8601, + "step": 24381 + }, + { + "epoch": 1.6919020192908194, + "grad_norm": 1.2820099592208862, + "learning_rate": 0.00012936735166478718, + "loss": 0.5881, + "step": 24382 + }, + { + "epoch": 1.6919714107279162, + "grad_norm": 1.9269418716430664, + "learning_rate": 0.0001293619807514167, + "loss": 0.5607, + "step": 24383 + }, + { + "epoch": 1.6920408021650128, + "grad_norm": 1.1603782176971436, + "learning_rate": 0.00012935660974535383, + "loss": 0.5438, + "step": 24384 + }, + { + "epoch": 1.6921101936021095, + "grad_norm": 1.6200542449951172, + "learning_rate": 0.00012935123864661556, + "loss": 0.679, + "step": 24385 + }, + { + "epoch": 1.6921795850392063, + "grad_norm": 1.7407019138336182, + "learning_rate": 0.00012934586745521882, + "loss": 0.5853, + "step": 24386 + }, + { + "epoch": 1.6922489764763027, + "grad_norm": 1.8168790340423584, + "learning_rate": 0.0001293404961711806, + "loss": 0.6267, + "step": 24387 + }, + { + "epoch": 1.6923183679133995, + "grad_norm": 1.6106340885162354, + "learning_rate": 0.00012933512479451783, + "loss": 0.7081, + "step": 24388 + }, + { + "epoch": 1.6923877593504961, + "grad_norm": 2.139497995376587, + "learning_rate": 0.00012932975332524747, + "loss": 0.6551, + "step": 24389 + }, + { + "epoch": 1.6924571507875927, + "grad_norm": 1.3776516914367676, + "learning_rate": 0.00012932438176338647, + "loss": 0.5198, + "step": 24390 + }, + { + "epoch": 1.6925265422246896, + "grad_norm": 1.6552788019180298, + "learning_rate": 0.00012931901010895182, + "loss": 0.8743, + "step": 24391 + }, + { + "epoch": 1.6925959336617862, + "grad_norm": 1.7813419103622437, + "learning_rate": 0.00012931363836196042, + "loss": 0.5746, + "step": 24392 + }, + { + "epoch": 1.6926653250988828, + "grad_norm": 1.497096300125122, + "learning_rate": 0.0001293082665224293, + "loss": 0.4604, + "step": 24393 + }, + { + "epoch": 1.6927347165359796, + "grad_norm": 1.2705782651901245, + "learning_rate": 0.00012930289459037539, + "loss": 0.4519, + "step": 24394 + }, + { + "epoch": 1.692804107973076, + "grad_norm": 1.8646163940429688, + "learning_rate": 0.00012929752256581563, + "loss": 0.4941, + "step": 24395 + }, + { + "epoch": 1.6928734994101728, + "grad_norm": 2.080188512802124, + "learning_rate": 0.00012929215044876698, + "loss": 0.6192, + "step": 24396 + }, + { + "epoch": 1.6929428908472695, + "grad_norm": 1.5054477453231812, + "learning_rate": 0.00012928677823924642, + "loss": 0.6246, + "step": 24397 + }, + { + "epoch": 1.693012282284366, + "grad_norm": 1.8037657737731934, + "learning_rate": 0.00012928140593727088, + "loss": 0.8496, + "step": 24398 + }, + { + "epoch": 1.693081673721463, + "grad_norm": 1.7515007257461548, + "learning_rate": 0.00012927603354285735, + "loss": 0.5786, + "step": 24399 + }, + { + "epoch": 1.6931510651585593, + "grad_norm": 1.536539912223816, + "learning_rate": 0.00012927066105602282, + "loss": 0.6515, + "step": 24400 + }, + { + "epoch": 1.6932204565956561, + "grad_norm": 1.4759385585784912, + "learning_rate": 0.00012926528847678414, + "loss": 0.5935, + "step": 24401 + }, + { + "epoch": 1.6932898480327527, + "grad_norm": 1.3902437686920166, + "learning_rate": 0.00012925991580515838, + "loss": 0.6156, + "step": 24402 + }, + { + "epoch": 1.6933592394698493, + "grad_norm": 1.757087230682373, + "learning_rate": 0.00012925454304116247, + "loss": 0.5431, + "step": 24403 + }, + { + "epoch": 1.6934286309069462, + "grad_norm": 1.6073534488677979, + "learning_rate": 0.00012924917018481335, + "loss": 0.6828, + "step": 24404 + }, + { + "epoch": 1.6934980223440428, + "grad_norm": 1.368058204650879, + "learning_rate": 0.000129243797236128, + "loss": 0.5112, + "step": 24405 + }, + { + "epoch": 1.6935674137811394, + "grad_norm": 1.9010859727859497, + "learning_rate": 0.00012923842419512333, + "loss": 0.72, + "step": 24406 + }, + { + "epoch": 1.6936368052182362, + "grad_norm": 2.074688673019409, + "learning_rate": 0.0001292330510618164, + "loss": 0.6692, + "step": 24407 + }, + { + "epoch": 1.6937061966553326, + "grad_norm": 1.5374939441680908, + "learning_rate": 0.0001292276778362241, + "loss": 0.5938, + "step": 24408 + }, + { + "epoch": 1.6937755880924295, + "grad_norm": 2.2271459102630615, + "learning_rate": 0.0001292223045183634, + "loss": 0.5945, + "step": 24409 + }, + { + "epoch": 1.693844979529526, + "grad_norm": 1.6975778341293335, + "learning_rate": 0.0001292169311082513, + "loss": 0.6177, + "step": 24410 + }, + { + "epoch": 1.6939143709666227, + "grad_norm": 2.0179684162139893, + "learning_rate": 0.0001292115576059047, + "loss": 0.6122, + "step": 24411 + }, + { + "epoch": 1.6939837624037195, + "grad_norm": 1.5216089487075806, + "learning_rate": 0.00012920618401134063, + "loss": 0.6707, + "step": 24412 + }, + { + "epoch": 1.694053153840816, + "grad_norm": 2.2414984703063965, + "learning_rate": 0.00012920081032457602, + "loss": 0.7056, + "step": 24413 + }, + { + "epoch": 1.6941225452779127, + "grad_norm": 1.6795250177383423, + "learning_rate": 0.00012919543654562785, + "loss": 0.811, + "step": 24414 + }, + { + "epoch": 1.6941919367150093, + "grad_norm": 1.571135401725769, + "learning_rate": 0.00012919006267451307, + "loss": 0.4138, + "step": 24415 + }, + { + "epoch": 1.694261328152106, + "grad_norm": 1.3580985069274902, + "learning_rate": 0.00012918468871124864, + "loss": 0.4797, + "step": 24416 + }, + { + "epoch": 1.6943307195892028, + "grad_norm": 1.7861047983169556, + "learning_rate": 0.00012917931465585152, + "loss": 0.7811, + "step": 24417 + }, + { + "epoch": 1.6944001110262994, + "grad_norm": 1.440488338470459, + "learning_rate": 0.00012917394050833872, + "loss": 0.6045, + "step": 24418 + }, + { + "epoch": 1.694469502463396, + "grad_norm": 1.8706861734390259, + "learning_rate": 0.00012916856626872715, + "loss": 0.6977, + "step": 24419 + }, + { + "epoch": 1.6945388939004928, + "grad_norm": 1.418397307395935, + "learning_rate": 0.00012916319193703382, + "loss": 0.3831, + "step": 24420 + }, + { + "epoch": 1.6946082853375892, + "grad_norm": 1.4500371217727661, + "learning_rate": 0.00012915781751327566, + "loss": 0.5251, + "step": 24421 + }, + { + "epoch": 1.694677676774686, + "grad_norm": 1.6478351354599, + "learning_rate": 0.00012915244299746968, + "loss": 0.6374, + "step": 24422 + }, + { + "epoch": 1.6947470682117827, + "grad_norm": 1.4127496480941772, + "learning_rate": 0.0001291470683896328, + "loss": 0.4868, + "step": 24423 + }, + { + "epoch": 1.6948164596488793, + "grad_norm": 1.9490913152694702, + "learning_rate": 0.000129141693689782, + "loss": 0.6795, + "step": 24424 + }, + { + "epoch": 1.6948858510859761, + "grad_norm": 1.404525637626648, + "learning_rate": 0.00012913631889793428, + "loss": 0.5833, + "step": 24425 + }, + { + "epoch": 1.6949552425230725, + "grad_norm": 1.9942867755889893, + "learning_rate": 0.00012913094401410658, + "loss": 0.5686, + "step": 24426 + }, + { + "epoch": 1.6950246339601693, + "grad_norm": 1.9127652645111084, + "learning_rate": 0.00012912556903831585, + "loss": 0.5802, + "step": 24427 + }, + { + "epoch": 1.695094025397266, + "grad_norm": 1.9285383224487305, + "learning_rate": 0.00012912019397057908, + "loss": 0.5224, + "step": 24428 + }, + { + "epoch": 1.6951634168343626, + "grad_norm": 1.3611398935317993, + "learning_rate": 0.00012911481881091322, + "loss": 0.5762, + "step": 24429 + }, + { + "epoch": 1.6952328082714594, + "grad_norm": 2.1786367893218994, + "learning_rate": 0.00012910944355933528, + "loss": 0.7172, + "step": 24430 + }, + { + "epoch": 1.695302199708556, + "grad_norm": 1.9581685066223145, + "learning_rate": 0.00012910406821586222, + "loss": 0.6741, + "step": 24431 + }, + { + "epoch": 1.6953715911456526, + "grad_norm": 1.8116581439971924, + "learning_rate": 0.00012909869278051095, + "loss": 0.7004, + "step": 24432 + }, + { + "epoch": 1.6954409825827494, + "grad_norm": 1.5036287307739258, + "learning_rate": 0.0001290933172532985, + "loss": 0.48, + "step": 24433 + }, + { + "epoch": 1.6955103740198458, + "grad_norm": 1.260255217552185, + "learning_rate": 0.00012908794163424183, + "loss": 0.4472, + "step": 24434 + }, + { + "epoch": 1.6955797654569427, + "grad_norm": 2.899259090423584, + "learning_rate": 0.0001290825659233579, + "loss": 0.6494, + "step": 24435 + }, + { + "epoch": 1.6956491568940393, + "grad_norm": 4.12876558303833, + "learning_rate": 0.0001290771901206637, + "loss": 0.7321, + "step": 24436 + }, + { + "epoch": 1.6957185483311359, + "grad_norm": 1.5403525829315186, + "learning_rate": 0.00012907181422617615, + "loss": 0.6615, + "step": 24437 + }, + { + "epoch": 1.6957879397682327, + "grad_norm": 1.4016292095184326, + "learning_rate": 0.00012906643823991227, + "loss": 0.5174, + "step": 24438 + }, + { + "epoch": 1.695857331205329, + "grad_norm": 1.9760067462921143, + "learning_rate": 0.00012906106216188903, + "loss": 0.4979, + "step": 24439 + }, + { + "epoch": 1.695926722642426, + "grad_norm": 1.523073673248291, + "learning_rate": 0.0001290556859921234, + "loss": 0.5885, + "step": 24440 + }, + { + "epoch": 1.6959961140795226, + "grad_norm": 1.8327947854995728, + "learning_rate": 0.0001290503097306323, + "loss": 0.6361, + "step": 24441 + }, + { + "epoch": 1.6960655055166192, + "grad_norm": 2.043403387069702, + "learning_rate": 0.00012904493337743275, + "loss": 0.7259, + "step": 24442 + }, + { + "epoch": 1.696134896953716, + "grad_norm": 2.1375808715820312, + "learning_rate": 0.00012903955693254173, + "loss": 0.776, + "step": 24443 + }, + { + "epoch": 1.6962042883908126, + "grad_norm": 1.553098201751709, + "learning_rate": 0.0001290341803959762, + "loss": 0.6369, + "step": 24444 + }, + { + "epoch": 1.6962736798279092, + "grad_norm": 1.413266897201538, + "learning_rate": 0.00012902880376775312, + "loss": 0.5392, + "step": 24445 + }, + { + "epoch": 1.696343071265006, + "grad_norm": 1.3671143054962158, + "learning_rate": 0.00012902342704788948, + "loss": 0.3549, + "step": 24446 + }, + { + "epoch": 1.6964124627021024, + "grad_norm": 1.4836660623550415, + "learning_rate": 0.00012901805023640223, + "loss": 0.5272, + "step": 24447 + }, + { + "epoch": 1.6964818541391993, + "grad_norm": 1.6036953926086426, + "learning_rate": 0.0001290126733333084, + "loss": 0.6117, + "step": 24448 + }, + { + "epoch": 1.6965512455762959, + "grad_norm": 2.348181962966919, + "learning_rate": 0.00012900729633862488, + "loss": 0.5324, + "step": 24449 + }, + { + "epoch": 1.6966206370133925, + "grad_norm": 1.577909231185913, + "learning_rate": 0.00012900191925236873, + "loss": 0.5366, + "step": 24450 + }, + { + "epoch": 1.6966900284504893, + "grad_norm": 1.3868142366409302, + "learning_rate": 0.00012899654207455688, + "loss": 0.4549, + "step": 24451 + }, + { + "epoch": 1.6967594198875857, + "grad_norm": 2.0390241146087646, + "learning_rate": 0.00012899116480520633, + "loss": 0.6939, + "step": 24452 + }, + { + "epoch": 1.6968288113246826, + "grad_norm": 1.782833456993103, + "learning_rate": 0.000128985787444334, + "loss": 0.6134, + "step": 24453 + }, + { + "epoch": 1.6968982027617792, + "grad_norm": 2.1706087589263916, + "learning_rate": 0.0001289804099919569, + "loss": 0.4844, + "step": 24454 + }, + { + "epoch": 1.6969675941988758, + "grad_norm": 1.8159966468811035, + "learning_rate": 0.00012897503244809201, + "loss": 0.4723, + "step": 24455 + }, + { + "epoch": 1.6970369856359726, + "grad_norm": 2.4959475994110107, + "learning_rate": 0.0001289696548127563, + "loss": 0.6901, + "step": 24456 + }, + { + "epoch": 1.6971063770730692, + "grad_norm": 1.8534232378005981, + "learning_rate": 0.0001289642770859668, + "loss": 0.5947, + "step": 24457 + }, + { + "epoch": 1.6971757685101658, + "grad_norm": 1.5034934282302856, + "learning_rate": 0.0001289588992677404, + "loss": 0.6769, + "step": 24458 + }, + { + "epoch": 1.6972451599472627, + "grad_norm": 2.815639019012451, + "learning_rate": 0.00012895352135809413, + "loss": 0.5903, + "step": 24459 + }, + { + "epoch": 1.697314551384359, + "grad_norm": 1.8035566806793213, + "learning_rate": 0.00012894814335704492, + "loss": 0.7366, + "step": 24460 + }, + { + "epoch": 1.6973839428214559, + "grad_norm": 1.8082823753356934, + "learning_rate": 0.0001289427652646098, + "loss": 0.5614, + "step": 24461 + }, + { + "epoch": 1.6974533342585525, + "grad_norm": 1.2651281356811523, + "learning_rate": 0.00012893738708080575, + "loss": 0.4443, + "step": 24462 + }, + { + "epoch": 1.697522725695649, + "grad_norm": 1.8453847169876099, + "learning_rate": 0.00012893200880564968, + "loss": 0.7873, + "step": 24463 + }, + { + "epoch": 1.697592117132746, + "grad_norm": 1.670535922050476, + "learning_rate": 0.00012892663043915863, + "loss": 0.4372, + "step": 24464 + }, + { + "epoch": 1.6976615085698423, + "grad_norm": 2.6618452072143555, + "learning_rate": 0.0001289212519813496, + "loss": 0.8443, + "step": 24465 + }, + { + "epoch": 1.6977309000069392, + "grad_norm": 1.8050260543823242, + "learning_rate": 0.0001289158734322395, + "loss": 0.5662, + "step": 24466 + }, + { + "epoch": 1.6978002914440358, + "grad_norm": 2.1253609657287598, + "learning_rate": 0.00012891049479184536, + "loss": 0.4113, + "step": 24467 + }, + { + "epoch": 1.6978696828811324, + "grad_norm": 1.8060497045516968, + "learning_rate": 0.00012890511606018413, + "loss": 0.7308, + "step": 24468 + }, + { + "epoch": 1.6979390743182292, + "grad_norm": 1.5272947549819946, + "learning_rate": 0.0001288997372372728, + "loss": 0.6389, + "step": 24469 + }, + { + "epoch": 1.6980084657553258, + "grad_norm": 1.302390217781067, + "learning_rate": 0.00012889435832312836, + "loss": 0.5158, + "step": 24470 + }, + { + "epoch": 1.6980778571924224, + "grad_norm": 1.0585689544677734, + "learning_rate": 0.00012888897931776777, + "loss": 0.3671, + "step": 24471 + }, + { + "epoch": 1.6981472486295193, + "grad_norm": 1.1809581518173218, + "learning_rate": 0.00012888360022120801, + "loss": 0.4902, + "step": 24472 + }, + { + "epoch": 1.6982166400666157, + "grad_norm": 1.7523473501205444, + "learning_rate": 0.00012887822103346613, + "loss": 0.4866, + "step": 24473 + }, + { + "epoch": 1.6982860315037125, + "grad_norm": 1.7592432498931885, + "learning_rate": 0.00012887284175455902, + "loss": 0.5556, + "step": 24474 + }, + { + "epoch": 1.698355422940809, + "grad_norm": 1.416932463645935, + "learning_rate": 0.0001288674623845037, + "loss": 0.6697, + "step": 24475 + }, + { + "epoch": 1.6984248143779057, + "grad_norm": 1.7882881164550781, + "learning_rate": 0.00012886208292331718, + "loss": 0.6417, + "step": 24476 + }, + { + "epoch": 1.6984942058150025, + "grad_norm": 1.8700389862060547, + "learning_rate": 0.00012885670337101635, + "loss": 0.5913, + "step": 24477 + }, + { + "epoch": 1.698563597252099, + "grad_norm": 1.6091225147247314, + "learning_rate": 0.0001288513237276183, + "loss": 0.5413, + "step": 24478 + }, + { + "epoch": 1.6986329886891958, + "grad_norm": 1.6948052644729614, + "learning_rate": 0.00012884594399314, + "loss": 0.5325, + "step": 24479 + }, + { + "epoch": 1.6987023801262924, + "grad_norm": 1.3403372764587402, + "learning_rate": 0.00012884056416759832, + "loss": 0.3523, + "step": 24480 + }, + { + "epoch": 1.698771771563389, + "grad_norm": 1.5500400066375732, + "learning_rate": 0.00012883518425101039, + "loss": 0.5673, + "step": 24481 + }, + { + "epoch": 1.6988411630004858, + "grad_norm": 1.8085658550262451, + "learning_rate": 0.00012882980424339308, + "loss": 0.5039, + "step": 24482 + }, + { + "epoch": 1.6989105544375824, + "grad_norm": 1.5013993978500366, + "learning_rate": 0.00012882442414476346, + "loss": 0.5429, + "step": 24483 + }, + { + "epoch": 1.698979945874679, + "grad_norm": 2.007925510406494, + "learning_rate": 0.00012881904395513845, + "loss": 0.6352, + "step": 24484 + }, + { + "epoch": 1.6990493373117759, + "grad_norm": 2.0263121128082275, + "learning_rate": 0.0001288136636745351, + "loss": 0.7725, + "step": 24485 + }, + { + "epoch": 1.6991187287488723, + "grad_norm": 1.7820792198181152, + "learning_rate": 0.00012880828330297033, + "loss": 0.7545, + "step": 24486 + }, + { + "epoch": 1.699188120185969, + "grad_norm": 1.8445245027542114, + "learning_rate": 0.00012880290284046116, + "loss": 0.7354, + "step": 24487 + }, + { + "epoch": 1.6992575116230657, + "grad_norm": 1.9025968313217163, + "learning_rate": 0.00012879752228702455, + "loss": 0.6453, + "step": 24488 + }, + { + "epoch": 1.6993269030601623, + "grad_norm": 1.6099646091461182, + "learning_rate": 0.0001287921416426775, + "loss": 0.541, + "step": 24489 + }, + { + "epoch": 1.6993962944972592, + "grad_norm": 1.9491461515426636, + "learning_rate": 0.00012878676090743703, + "loss": 0.6294, + "step": 24490 + }, + { + "epoch": 1.6994656859343555, + "grad_norm": 2.130850315093994, + "learning_rate": 0.0001287813800813201, + "loss": 0.5929, + "step": 24491 + }, + { + "epoch": 1.6995350773714524, + "grad_norm": 1.4043453931808472, + "learning_rate": 0.00012877599916434368, + "loss": 0.5159, + "step": 24492 + }, + { + "epoch": 1.699604468808549, + "grad_norm": 1.3375635147094727, + "learning_rate": 0.00012877061815652478, + "loss": 0.5241, + "step": 24493 + }, + { + "epoch": 1.6996738602456456, + "grad_norm": 1.4544748067855835, + "learning_rate": 0.00012876523705788036, + "loss": 0.534, + "step": 24494 + }, + { + "epoch": 1.6997432516827424, + "grad_norm": 1.0459680557250977, + "learning_rate": 0.0001287598558684274, + "loss": 0.3411, + "step": 24495 + }, + { + "epoch": 1.699812643119839, + "grad_norm": 1.4668760299682617, + "learning_rate": 0.00012875447458818293, + "loss": 0.5319, + "step": 24496 + }, + { + "epoch": 1.6998820345569357, + "grad_norm": 1.5023990869522095, + "learning_rate": 0.00012874909321716396, + "loss": 0.5493, + "step": 24497 + }, + { + "epoch": 1.6999514259940325, + "grad_norm": 1.4148976802825928, + "learning_rate": 0.00012874371175538739, + "loss": 0.5078, + "step": 24498 + }, + { + "epoch": 1.7000208174311289, + "grad_norm": 1.352620244026184, + "learning_rate": 0.00012873833020287027, + "loss": 0.593, + "step": 24499 + }, + { + "epoch": 1.7000902088682257, + "grad_norm": 1.9060008525848389, + "learning_rate": 0.00012873294855962958, + "loss": 0.4639, + "step": 24500 + }, + { + "epoch": 1.7001596003053223, + "grad_norm": 1.6903212070465088, + "learning_rate": 0.0001287275668256823, + "loss": 0.6994, + "step": 24501 + }, + { + "epoch": 1.700228991742419, + "grad_norm": 1.272194266319275, + "learning_rate": 0.0001287221850010454, + "loss": 0.5256, + "step": 24502 + }, + { + "epoch": 1.7002983831795158, + "grad_norm": 2.061840057373047, + "learning_rate": 0.00012871680308573595, + "loss": 0.613, + "step": 24503 + }, + { + "epoch": 1.7003677746166121, + "grad_norm": 1.5743194818496704, + "learning_rate": 0.00012871142107977085, + "loss": 0.6437, + "step": 24504 + }, + { + "epoch": 1.700437166053709, + "grad_norm": 1.70174241065979, + "learning_rate": 0.00012870603898316714, + "loss": 0.3536, + "step": 24505 + }, + { + "epoch": 1.7005065574908056, + "grad_norm": 2.075834274291992, + "learning_rate": 0.00012870065679594176, + "loss": 0.7526, + "step": 24506 + }, + { + "epoch": 1.7005759489279022, + "grad_norm": 1.6449397802352905, + "learning_rate": 0.0001286952745181118, + "loss": 0.482, + "step": 24507 + }, + { + "epoch": 1.700645340364999, + "grad_norm": 1.5537513494491577, + "learning_rate": 0.00012868989214969413, + "loss": 0.6862, + "step": 24508 + }, + { + "epoch": 1.7007147318020956, + "grad_norm": 1.943603277206421, + "learning_rate": 0.00012868450969070582, + "loss": 0.445, + "step": 24509 + }, + { + "epoch": 1.7007841232391923, + "grad_norm": 1.4709175825119019, + "learning_rate": 0.00012867912714116384, + "loss": 0.5931, + "step": 24510 + }, + { + "epoch": 1.700853514676289, + "grad_norm": 2.147307872772217, + "learning_rate": 0.0001286737445010852, + "loss": 0.669, + "step": 24511 + }, + { + "epoch": 1.7009229061133855, + "grad_norm": 2.0086286067962646, + "learning_rate": 0.00012866836177048685, + "loss": 0.6149, + "step": 24512 + }, + { + "epoch": 1.7009922975504823, + "grad_norm": 1.955094337463379, + "learning_rate": 0.00012866297894938578, + "loss": 0.4839, + "step": 24513 + }, + { + "epoch": 1.701061688987579, + "grad_norm": 2.82958722114563, + "learning_rate": 0.0001286575960377991, + "loss": 0.6712, + "step": 24514 + }, + { + "epoch": 1.7011310804246755, + "grad_norm": 1.4555920362472534, + "learning_rate": 0.00012865221303574363, + "loss": 0.4048, + "step": 24515 + }, + { + "epoch": 1.7012004718617724, + "grad_norm": 1.759928584098816, + "learning_rate": 0.00012864682994323648, + "loss": 0.6168, + "step": 24516 + }, + { + "epoch": 1.7012698632988688, + "grad_norm": 2.091020345687866, + "learning_rate": 0.00012864144676029464, + "loss": 0.6798, + "step": 24517 + }, + { + "epoch": 1.7013392547359656, + "grad_norm": 1.645820140838623, + "learning_rate": 0.000128636063486935, + "loss": 0.5975, + "step": 24518 + }, + { + "epoch": 1.7014086461730622, + "grad_norm": 1.1472731828689575, + "learning_rate": 0.0001286306801231747, + "loss": 0.4136, + "step": 24519 + }, + { + "epoch": 1.7014780376101588, + "grad_norm": 1.486130714416504, + "learning_rate": 0.00012862529666903062, + "loss": 0.508, + "step": 24520 + }, + { + "epoch": 1.7015474290472556, + "grad_norm": 1.5516542196273804, + "learning_rate": 0.00012861991312451985, + "loss": 0.7808, + "step": 24521 + }, + { + "epoch": 1.7016168204843523, + "grad_norm": 2.6149728298187256, + "learning_rate": 0.0001286145294896593, + "loss": 0.7749, + "step": 24522 + }, + { + "epoch": 1.7016862119214489, + "grad_norm": 1.1632860898971558, + "learning_rate": 0.000128609145764466, + "loss": 0.4177, + "step": 24523 + }, + { + "epoch": 1.7017556033585457, + "grad_norm": 1.6059924364089966, + "learning_rate": 0.00012860376194895695, + "loss": 0.6173, + "step": 24524 + }, + { + "epoch": 1.701824994795642, + "grad_norm": 1.8493642807006836, + "learning_rate": 0.00012859837804314914, + "loss": 0.5866, + "step": 24525 + }, + { + "epoch": 1.701894386232739, + "grad_norm": 1.999571681022644, + "learning_rate": 0.0001285929940470596, + "loss": 0.6715, + "step": 24526 + }, + { + "epoch": 1.7019637776698355, + "grad_norm": 1.5792714357376099, + "learning_rate": 0.00012858760996070526, + "loss": 0.7209, + "step": 24527 + }, + { + "epoch": 1.7020331691069321, + "grad_norm": 1.6236273050308228, + "learning_rate": 0.00012858222578410316, + "loss": 0.5941, + "step": 24528 + }, + { + "epoch": 1.702102560544029, + "grad_norm": 2.4643163681030273, + "learning_rate": 0.0001285768415172703, + "loss": 0.5844, + "step": 24529 + }, + { + "epoch": 1.7021719519811254, + "grad_norm": 2.000105619430542, + "learning_rate": 0.00012857145716022362, + "loss": 0.5962, + "step": 24530 + }, + { + "epoch": 1.7022413434182222, + "grad_norm": 3.756840705871582, + "learning_rate": 0.0001285660727129802, + "loss": 0.6961, + "step": 24531 + }, + { + "epoch": 1.7023107348553188, + "grad_norm": 2.1050753593444824, + "learning_rate": 0.00012856068817555702, + "loss": 0.7902, + "step": 24532 + }, + { + "epoch": 1.7023801262924154, + "grad_norm": 1.3635296821594238, + "learning_rate": 0.00012855530354797105, + "loss": 0.6019, + "step": 24533 + }, + { + "epoch": 1.7024495177295123, + "grad_norm": 1.9242666959762573, + "learning_rate": 0.0001285499188302393, + "loss": 0.6575, + "step": 24534 + }, + { + "epoch": 1.7025189091666089, + "grad_norm": 1.854631781578064, + "learning_rate": 0.00012854453402237875, + "loss": 0.9397, + "step": 24535 + }, + { + "epoch": 1.7025883006037055, + "grad_norm": 1.6614800691604614, + "learning_rate": 0.00012853914912440642, + "loss": 0.5301, + "step": 24536 + }, + { + "epoch": 1.7026576920408023, + "grad_norm": 1.7105653285980225, + "learning_rate": 0.00012853376413633934, + "loss": 0.5499, + "step": 24537 + }, + { + "epoch": 1.7027270834778987, + "grad_norm": 1.6294891834259033, + "learning_rate": 0.00012852837905819446, + "loss": 0.4582, + "step": 24538 + }, + { + "epoch": 1.7027964749149955, + "grad_norm": 1.3911737203598022, + "learning_rate": 0.00012852299388998878, + "loss": 0.5154, + "step": 24539 + }, + { + "epoch": 1.7028658663520921, + "grad_norm": 1.9637733697891235, + "learning_rate": 0.00012851760863173933, + "loss": 0.8121, + "step": 24540 + }, + { + "epoch": 1.7029352577891887, + "grad_norm": 1.9191657304763794, + "learning_rate": 0.0001285122232834631, + "loss": 0.8503, + "step": 24541 + }, + { + "epoch": 1.7030046492262856, + "grad_norm": 1.819748044013977, + "learning_rate": 0.00012850683784517707, + "loss": 0.5512, + "step": 24542 + }, + { + "epoch": 1.7030740406633822, + "grad_norm": 1.3593220710754395, + "learning_rate": 0.0001285014523168983, + "loss": 0.6051, + "step": 24543 + }, + { + "epoch": 1.7031434321004788, + "grad_norm": 1.2579805850982666, + "learning_rate": 0.0001284960666986437, + "loss": 0.4478, + "step": 24544 + }, + { + "epoch": 1.7032128235375754, + "grad_norm": 1.372563123703003, + "learning_rate": 0.00012849068099043037, + "loss": 0.6339, + "step": 24545 + }, + { + "epoch": 1.703282214974672, + "grad_norm": 2.0148892402648926, + "learning_rate": 0.00012848529519227526, + "loss": 0.624, + "step": 24546 + }, + { + "epoch": 1.7033516064117689, + "grad_norm": 2.0769829750061035, + "learning_rate": 0.0001284799093041954, + "loss": 0.6533, + "step": 24547 + }, + { + "epoch": 1.7034209978488655, + "grad_norm": 2.065154790878296, + "learning_rate": 0.00012847452332620773, + "loss": 0.8354, + "step": 24548 + }, + { + "epoch": 1.703490389285962, + "grad_norm": 1.9468995332717896, + "learning_rate": 0.00012846913725832927, + "loss": 0.4595, + "step": 24549 + }, + { + "epoch": 1.703559780723059, + "grad_norm": 1.5793551206588745, + "learning_rate": 0.00012846375110057707, + "loss": 0.5565, + "step": 24550 + }, + { + "epoch": 1.7036291721601553, + "grad_norm": 1.7699260711669922, + "learning_rate": 0.00012845836485296812, + "loss": 0.5993, + "step": 24551 + }, + { + "epoch": 1.7036985635972521, + "grad_norm": 1.2390598058700562, + "learning_rate": 0.00012845297851551944, + "loss": 0.4945, + "step": 24552 + }, + { + "epoch": 1.7037679550343487, + "grad_norm": 2.0780861377716064, + "learning_rate": 0.00012844759208824796, + "loss": 0.7851, + "step": 24553 + }, + { + "epoch": 1.7038373464714454, + "grad_norm": 2.0240840911865234, + "learning_rate": 0.00012844220557117075, + "loss": 0.4988, + "step": 24554 + }, + { + "epoch": 1.7039067379085422, + "grad_norm": 1.4275106191635132, + "learning_rate": 0.00012843681896430482, + "loss": 0.5264, + "step": 24555 + }, + { + "epoch": 1.7039761293456388, + "grad_norm": 1.7407923936843872, + "learning_rate": 0.0001284314322676671, + "loss": 0.4623, + "step": 24556 + }, + { + "epoch": 1.7040455207827354, + "grad_norm": 1.3952757120132446, + "learning_rate": 0.00012842604548127468, + "loss": 0.5891, + "step": 24557 + }, + { + "epoch": 1.704114912219832, + "grad_norm": 2.1339914798736572, + "learning_rate": 0.0001284206586051445, + "loss": 0.7754, + "step": 24558 + }, + { + "epoch": 1.7041843036569286, + "grad_norm": 1.5008548498153687, + "learning_rate": 0.00012841527163929368, + "loss": 0.5805, + "step": 24559 + }, + { + "epoch": 1.7042536950940255, + "grad_norm": 1.0252561569213867, + "learning_rate": 0.00012840988458373906, + "loss": 0.4697, + "step": 24560 + }, + { + "epoch": 1.704323086531122, + "grad_norm": 1.172619342803955, + "learning_rate": 0.0001284044974384978, + "loss": 0.5196, + "step": 24561 + }, + { + "epoch": 1.7043924779682187, + "grad_norm": 1.2515701055526733, + "learning_rate": 0.00012839911020358678, + "loss": 0.3831, + "step": 24562 + }, + { + "epoch": 1.7044618694053155, + "grad_norm": 1.238840937614441, + "learning_rate": 0.00012839372287902308, + "loss": 0.5077, + "step": 24563 + }, + { + "epoch": 1.704531260842412, + "grad_norm": 1.4488723278045654, + "learning_rate": 0.0001283883354648237, + "loss": 0.6575, + "step": 24564 + }, + { + "epoch": 1.7046006522795087, + "grad_norm": 2.159351110458374, + "learning_rate": 0.00012838294796100564, + "loss": 0.5205, + "step": 24565 + }, + { + "epoch": 1.7046700437166054, + "grad_norm": 1.4704139232635498, + "learning_rate": 0.0001283775603675859, + "loss": 0.5489, + "step": 24566 + }, + { + "epoch": 1.704739435153702, + "grad_norm": 1.9685465097427368, + "learning_rate": 0.0001283721726845815, + "loss": 0.7794, + "step": 24567 + }, + { + "epoch": 1.7048088265907988, + "grad_norm": 1.6685298681259155, + "learning_rate": 0.00012836678491200942, + "loss": 0.5359, + "step": 24568 + }, + { + "epoch": 1.7048782180278954, + "grad_norm": 1.948519229888916, + "learning_rate": 0.00012836139704988672, + "loss": 0.5642, + "step": 24569 + }, + { + "epoch": 1.704947609464992, + "grad_norm": 1.3571054935455322, + "learning_rate": 0.00012835600909823035, + "loss": 0.5746, + "step": 24570 + }, + { + "epoch": 1.7050170009020886, + "grad_norm": 1.3928581476211548, + "learning_rate": 0.00012835062105705737, + "loss": 0.5204, + "step": 24571 + }, + { + "epoch": 1.7050863923391852, + "grad_norm": 2.293588876724243, + "learning_rate": 0.00012834523292638476, + "loss": 0.6127, + "step": 24572 + }, + { + "epoch": 1.705155783776282, + "grad_norm": 1.6539108753204346, + "learning_rate": 0.00012833984470622957, + "loss": 0.7542, + "step": 24573 + }, + { + "epoch": 1.7052251752133787, + "grad_norm": 1.8020211458206177, + "learning_rate": 0.00012833445639660876, + "loss": 0.6303, + "step": 24574 + }, + { + "epoch": 1.7052945666504753, + "grad_norm": 1.5682530403137207, + "learning_rate": 0.00012832906799753932, + "loss": 0.4891, + "step": 24575 + }, + { + "epoch": 1.7053639580875721, + "grad_norm": 1.4210158586502075, + "learning_rate": 0.00012832367950903835, + "loss": 0.5833, + "step": 24576 + }, + { + "epoch": 1.7054333495246685, + "grad_norm": 3.079815149307251, + "learning_rate": 0.00012831829093112277, + "loss": 0.5659, + "step": 24577 + }, + { + "epoch": 1.7055027409617654, + "grad_norm": 3.0652084350585938, + "learning_rate": 0.00012831290226380965, + "loss": 1.0902, + "step": 24578 + }, + { + "epoch": 1.705572132398862, + "grad_norm": 1.4474765062332153, + "learning_rate": 0.000128307513507116, + "loss": 0.6113, + "step": 24579 + }, + { + "epoch": 1.7056415238359586, + "grad_norm": 1.552261233329773, + "learning_rate": 0.00012830212466105883, + "loss": 0.6667, + "step": 24580 + }, + { + "epoch": 1.7057109152730554, + "grad_norm": 1.5122593641281128, + "learning_rate": 0.0001282967357256551, + "loss": 0.5333, + "step": 24581 + }, + { + "epoch": 1.705780306710152, + "grad_norm": 2.43400502204895, + "learning_rate": 0.00012829134670092185, + "loss": 0.6348, + "step": 24582 + }, + { + "epoch": 1.7058496981472486, + "grad_norm": 1.7191498279571533, + "learning_rate": 0.00012828595758687612, + "loss": 0.5345, + "step": 24583 + }, + { + "epoch": 1.7059190895843452, + "grad_norm": 1.2215187549591064, + "learning_rate": 0.00012828056838353488, + "loss": 0.664, + "step": 24584 + }, + { + "epoch": 1.7059884810214418, + "grad_norm": 2.5117299556732178, + "learning_rate": 0.00012827517909091518, + "loss": 0.5181, + "step": 24585 + }, + { + "epoch": 1.7060578724585387, + "grad_norm": 1.4782545566558838, + "learning_rate": 0.00012826978970903406, + "loss": 0.5674, + "step": 24586 + }, + { + "epoch": 1.7061272638956353, + "grad_norm": 1.9195096492767334, + "learning_rate": 0.00012826440023790844, + "loss": 0.6081, + "step": 24587 + }, + { + "epoch": 1.706196655332732, + "grad_norm": 3.028987407684326, + "learning_rate": 0.0001282590106775554, + "loss": 0.7326, + "step": 24588 + }, + { + "epoch": 1.7062660467698287, + "grad_norm": 1.9604357481002808, + "learning_rate": 0.0001282536210279919, + "loss": 0.5395, + "step": 24589 + }, + { + "epoch": 1.7063354382069251, + "grad_norm": 1.6305317878723145, + "learning_rate": 0.00012824823128923507, + "loss": 0.4578, + "step": 24590 + }, + { + "epoch": 1.706404829644022, + "grad_norm": 1.6176011562347412, + "learning_rate": 0.00012824284146130182, + "loss": 0.7609, + "step": 24591 + }, + { + "epoch": 1.7064742210811186, + "grad_norm": 2.28548264503479, + "learning_rate": 0.0001282374515442092, + "loss": 0.5685, + "step": 24592 + }, + { + "epoch": 1.7065436125182152, + "grad_norm": 1.3914564847946167, + "learning_rate": 0.00012823206153797425, + "loss": 0.4165, + "step": 24593 + }, + { + "epoch": 1.706613003955312, + "grad_norm": 1.2022067308425903, + "learning_rate": 0.0001282266714426139, + "loss": 0.4211, + "step": 24594 + }, + { + "epoch": 1.7066823953924086, + "grad_norm": 2.5411384105682373, + "learning_rate": 0.00012822128125814528, + "loss": 0.8254, + "step": 24595 + }, + { + "epoch": 1.7067517868295052, + "grad_norm": 1.880111813545227, + "learning_rate": 0.00012821589098458534, + "loss": 0.6505, + "step": 24596 + }, + { + "epoch": 1.7068211782666018, + "grad_norm": 2.4785053730010986, + "learning_rate": 0.0001282105006219511, + "loss": 0.5809, + "step": 24597 + }, + { + "epoch": 1.7068905697036985, + "grad_norm": 1.4724013805389404, + "learning_rate": 0.00012820511017025955, + "loss": 0.5522, + "step": 24598 + }, + { + "epoch": 1.7069599611407953, + "grad_norm": 1.557199478149414, + "learning_rate": 0.00012819971962952774, + "loss": 0.4199, + "step": 24599 + }, + { + "epoch": 1.707029352577892, + "grad_norm": 1.709197998046875, + "learning_rate": 0.00012819432899977272, + "loss": 0.5348, + "step": 24600 + }, + { + "epoch": 1.7070987440149885, + "grad_norm": 1.4978957176208496, + "learning_rate": 0.00012818893828101148, + "loss": 0.6256, + "step": 24601 + }, + { + "epoch": 1.7071681354520853, + "grad_norm": 2.362450361251831, + "learning_rate": 0.000128183547473261, + "loss": 0.7209, + "step": 24602 + }, + { + "epoch": 1.7072375268891817, + "grad_norm": 1.1533949375152588, + "learning_rate": 0.00012817815657653833, + "loss": 0.3963, + "step": 24603 + }, + { + "epoch": 1.7073069183262786, + "grad_norm": 1.7254637479782104, + "learning_rate": 0.00012817276559086053, + "loss": 0.6124, + "step": 24604 + }, + { + "epoch": 1.7073763097633752, + "grad_norm": 1.4576996564865112, + "learning_rate": 0.00012816737451624454, + "loss": 0.5041, + "step": 24605 + }, + { + "epoch": 1.7074457012004718, + "grad_norm": 2.150477647781372, + "learning_rate": 0.0001281619833527074, + "loss": 0.7784, + "step": 24606 + }, + { + "epoch": 1.7075150926375686, + "grad_norm": 1.8308978080749512, + "learning_rate": 0.0001281565921002662, + "loss": 0.53, + "step": 24607 + }, + { + "epoch": 1.7075844840746652, + "grad_norm": 1.5892974138259888, + "learning_rate": 0.00012815120075893786, + "loss": 0.7966, + "step": 24608 + }, + { + "epoch": 1.7076538755117618, + "grad_norm": 1.5302951335906982, + "learning_rate": 0.00012814580932873947, + "loss": 0.4515, + "step": 24609 + }, + { + "epoch": 1.7077232669488587, + "grad_norm": 1.5777721405029297, + "learning_rate": 0.00012814041780968803, + "loss": 0.5018, + "step": 24610 + }, + { + "epoch": 1.707792658385955, + "grad_norm": 1.488208293914795, + "learning_rate": 0.00012813502620180055, + "loss": 0.6835, + "step": 24611 + }, + { + "epoch": 1.707862049823052, + "grad_norm": 1.8292975425720215, + "learning_rate": 0.00012812963450509405, + "loss": 0.68, + "step": 24612 + }, + { + "epoch": 1.7079314412601485, + "grad_norm": 3.2813100814819336, + "learning_rate": 0.00012812424271958558, + "loss": 0.6454, + "step": 24613 + }, + { + "epoch": 1.7080008326972451, + "grad_norm": 1.687295913696289, + "learning_rate": 0.00012811885084529213, + "loss": 0.668, + "step": 24614 + }, + { + "epoch": 1.708070224134342, + "grad_norm": 1.2304641008377075, + "learning_rate": 0.00012811345888223073, + "loss": 0.3714, + "step": 24615 + }, + { + "epoch": 1.7081396155714383, + "grad_norm": 2.804443836212158, + "learning_rate": 0.00012810806683041836, + "loss": 0.5107, + "step": 24616 + }, + { + "epoch": 1.7082090070085352, + "grad_norm": 1.4249204397201538, + "learning_rate": 0.00012810267468987214, + "loss": 0.6274, + "step": 24617 + }, + { + "epoch": 1.7082783984456318, + "grad_norm": 1.5167707204818726, + "learning_rate": 0.000128097282460609, + "loss": 0.6285, + "step": 24618 + }, + { + "epoch": 1.7083477898827284, + "grad_norm": 1.8259385824203491, + "learning_rate": 0.000128091890142646, + "loss": 0.62, + "step": 24619 + }, + { + "epoch": 1.7084171813198252, + "grad_norm": 1.9082971811294556, + "learning_rate": 0.00012808649773600017, + "loss": 0.6945, + "step": 24620 + }, + { + "epoch": 1.7084865727569218, + "grad_norm": 1.9244879484176636, + "learning_rate": 0.00012808110524068856, + "loss": 0.6015, + "step": 24621 + }, + { + "epoch": 1.7085559641940185, + "grad_norm": 2.494408130645752, + "learning_rate": 0.0001280757126567281, + "loss": 0.8487, + "step": 24622 + }, + { + "epoch": 1.7086253556311153, + "grad_norm": 1.482715129852295, + "learning_rate": 0.0001280703199841359, + "loss": 0.4529, + "step": 24623 + }, + { + "epoch": 1.7086947470682117, + "grad_norm": 1.6529431343078613, + "learning_rate": 0.00012806492722292897, + "loss": 0.6184, + "step": 24624 + }, + { + "epoch": 1.7087641385053085, + "grad_norm": 2.564470052719116, + "learning_rate": 0.0001280595343731243, + "loss": 1.0101, + "step": 24625 + }, + { + "epoch": 1.7088335299424051, + "grad_norm": 2.053619623184204, + "learning_rate": 0.00012805414143473892, + "loss": 0.4542, + "step": 24626 + }, + { + "epoch": 1.7089029213795017, + "grad_norm": 1.9468324184417725, + "learning_rate": 0.00012804874840778992, + "loss": 0.6463, + "step": 24627 + }, + { + "epoch": 1.7089723128165986, + "grad_norm": 1.9014718532562256, + "learning_rate": 0.00012804335529229425, + "loss": 0.5886, + "step": 24628 + }, + { + "epoch": 1.709041704253695, + "grad_norm": 1.2408273220062256, + "learning_rate": 0.00012803796208826895, + "loss": 0.4956, + "step": 24629 + }, + { + "epoch": 1.7091110956907918, + "grad_norm": 1.759870171546936, + "learning_rate": 0.00012803256879573107, + "loss": 0.8243, + "step": 24630 + }, + { + "epoch": 1.7091804871278884, + "grad_norm": 1.950333833694458, + "learning_rate": 0.00012802717541469762, + "loss": 0.6582, + "step": 24631 + }, + { + "epoch": 1.709249878564985, + "grad_norm": 1.2553364038467407, + "learning_rate": 0.0001280217819451856, + "loss": 0.4764, + "step": 24632 + }, + { + "epoch": 1.7093192700020818, + "grad_norm": 1.515109896659851, + "learning_rate": 0.00012801638838721212, + "loss": 0.6147, + "step": 24633 + }, + { + "epoch": 1.7093886614391784, + "grad_norm": 1.7105258703231812, + "learning_rate": 0.00012801099474079412, + "loss": 0.742, + "step": 24634 + }, + { + "epoch": 1.709458052876275, + "grad_norm": 1.9567302465438843, + "learning_rate": 0.0001280056010059487, + "loss": 0.7753, + "step": 24635 + }, + { + "epoch": 1.709527444313372, + "grad_norm": 2.4760525226593018, + "learning_rate": 0.0001280002071826928, + "loss": 0.6428, + "step": 24636 + }, + { + "epoch": 1.7095968357504683, + "grad_norm": 1.78591787815094, + "learning_rate": 0.00012799481327104348, + "loss": 0.7359, + "step": 24637 + }, + { + "epoch": 1.7096662271875651, + "grad_norm": 2.291454315185547, + "learning_rate": 0.00012798941927101784, + "loss": 0.6432, + "step": 24638 + }, + { + "epoch": 1.7097356186246617, + "grad_norm": 2.3140554428100586, + "learning_rate": 0.00012798402518263282, + "loss": 0.6405, + "step": 24639 + }, + { + "epoch": 1.7098050100617583, + "grad_norm": 1.6310056447982788, + "learning_rate": 0.00012797863100590546, + "loss": 0.4921, + "step": 24640 + }, + { + "epoch": 1.7098744014988552, + "grad_norm": 1.7265582084655762, + "learning_rate": 0.0001279732367408528, + "loss": 0.4543, + "step": 24641 + }, + { + "epoch": 1.7099437929359516, + "grad_norm": 1.7797554731369019, + "learning_rate": 0.00012796784238749195, + "loss": 0.4382, + "step": 24642 + }, + { + "epoch": 1.7100131843730484, + "grad_norm": 1.33844792842865, + "learning_rate": 0.0001279624479458398, + "loss": 0.4363, + "step": 24643 + }, + { + "epoch": 1.710082575810145, + "grad_norm": 2.4401633739471436, + "learning_rate": 0.00012795705341591345, + "loss": 0.4503, + "step": 24644 + }, + { + "epoch": 1.7101519672472416, + "grad_norm": 2.5648341178894043, + "learning_rate": 0.00012795165879772997, + "loss": 0.6587, + "step": 24645 + }, + { + "epoch": 1.7102213586843384, + "grad_norm": 1.7547019720077515, + "learning_rate": 0.00012794626409130632, + "loss": 0.5567, + "step": 24646 + }, + { + "epoch": 1.710290750121435, + "grad_norm": 1.6706569194793701, + "learning_rate": 0.00012794086929665953, + "loss": 0.478, + "step": 24647 + }, + { + "epoch": 1.7103601415585317, + "grad_norm": 1.4720113277435303, + "learning_rate": 0.00012793547441380672, + "loss": 0.5543, + "step": 24648 + }, + { + "epoch": 1.7104295329956285, + "grad_norm": 1.6738029718399048, + "learning_rate": 0.00012793007944276482, + "loss": 0.4911, + "step": 24649 + }, + { + "epoch": 1.7104989244327249, + "grad_norm": 1.84152352809906, + "learning_rate": 0.0001279246843835509, + "loss": 0.6421, + "step": 24650 + }, + { + "epoch": 1.7105683158698217, + "grad_norm": 3.976331949234009, + "learning_rate": 0.00012791928923618198, + "loss": 0.6607, + "step": 24651 + }, + { + "epoch": 1.7106377073069183, + "grad_norm": 1.551031231880188, + "learning_rate": 0.00012791389400067512, + "loss": 0.5949, + "step": 24652 + }, + { + "epoch": 1.710707098744015, + "grad_norm": 1.4356669187545776, + "learning_rate": 0.00012790849867704734, + "loss": 0.4931, + "step": 24653 + }, + { + "epoch": 1.7107764901811118, + "grad_norm": 2.3498177528381348, + "learning_rate": 0.00012790310326531566, + "loss": 0.6997, + "step": 24654 + }, + { + "epoch": 1.7108458816182082, + "grad_norm": 2.11478853225708, + "learning_rate": 0.00012789770776549711, + "loss": 0.502, + "step": 24655 + }, + { + "epoch": 1.710915273055305, + "grad_norm": 1.8173149824142456, + "learning_rate": 0.00012789231217760874, + "loss": 0.4535, + "step": 24656 + }, + { + "epoch": 1.7109846644924016, + "grad_norm": 1.7868187427520752, + "learning_rate": 0.00012788691650166757, + "loss": 0.5398, + "step": 24657 + }, + { + "epoch": 1.7110540559294982, + "grad_norm": 1.8025989532470703, + "learning_rate": 0.00012788152073769067, + "loss": 0.6111, + "step": 24658 + }, + { + "epoch": 1.711123447366595, + "grad_norm": 2.0111782550811768, + "learning_rate": 0.000127876124885695, + "loss": 0.5415, + "step": 24659 + }, + { + "epoch": 1.7111928388036917, + "grad_norm": 2.740102767944336, + "learning_rate": 0.00012787072894569767, + "loss": 0.4896, + "step": 24660 + }, + { + "epoch": 1.7112622302407883, + "grad_norm": 2.5269856452941895, + "learning_rate": 0.00012786533291771568, + "loss": 0.6246, + "step": 24661 + }, + { + "epoch": 1.711331621677885, + "grad_norm": 2.751408338546753, + "learning_rate": 0.00012785993680176607, + "loss": 0.4601, + "step": 24662 + }, + { + "epoch": 1.7114010131149815, + "grad_norm": 2.442880153656006, + "learning_rate": 0.00012785454059786587, + "loss": 0.7398, + "step": 24663 + }, + { + "epoch": 1.7114704045520783, + "grad_norm": 7.065618991851807, + "learning_rate": 0.00012784914430603208, + "loss": 0.7521, + "step": 24664 + }, + { + "epoch": 1.711539795989175, + "grad_norm": 2.251814603805542, + "learning_rate": 0.0001278437479262818, + "loss": 0.7085, + "step": 24665 + }, + { + "epoch": 1.7116091874262715, + "grad_norm": 1.681715726852417, + "learning_rate": 0.00012783835145863205, + "loss": 0.6443, + "step": 24666 + }, + { + "epoch": 1.7116785788633684, + "grad_norm": 1.9164836406707764, + "learning_rate": 0.00012783295490309982, + "loss": 0.4869, + "step": 24667 + }, + { + "epoch": 1.7117479703004648, + "grad_norm": 1.9118163585662842, + "learning_rate": 0.00012782755825970223, + "loss": 0.7449, + "step": 24668 + }, + { + "epoch": 1.7118173617375616, + "grad_norm": 1.307438850402832, + "learning_rate": 0.00012782216152845628, + "loss": 0.4272, + "step": 24669 + }, + { + "epoch": 1.7118867531746582, + "grad_norm": 1.4307868480682373, + "learning_rate": 0.00012781676470937893, + "loss": 0.3438, + "step": 24670 + }, + { + "epoch": 1.7119561446117548, + "grad_norm": 2.236523151397705, + "learning_rate": 0.0001278113678024873, + "loss": 0.5214, + "step": 24671 + }, + { + "epoch": 1.7120255360488517, + "grad_norm": 2.6433818340301514, + "learning_rate": 0.00012780597080779842, + "loss": 0.8333, + "step": 24672 + }, + { + "epoch": 1.7120949274859483, + "grad_norm": 2.1086549758911133, + "learning_rate": 0.00012780057372532933, + "loss": 0.6089, + "step": 24673 + }, + { + "epoch": 1.7121643189230449, + "grad_norm": 1.8242651224136353, + "learning_rate": 0.000127795176555097, + "loss": 0.6095, + "step": 24674 + }, + { + "epoch": 1.7122337103601417, + "grad_norm": 1.9524625539779663, + "learning_rate": 0.00012778977929711857, + "loss": 0.5626, + "step": 24675 + }, + { + "epoch": 1.712303101797238, + "grad_norm": 1.6406053304672241, + "learning_rate": 0.00012778438195141102, + "loss": 0.5271, + "step": 24676 + }, + { + "epoch": 1.712372493234335, + "grad_norm": 1.413239598274231, + "learning_rate": 0.0001277789845179914, + "loss": 0.5635, + "step": 24677 + }, + { + "epoch": 1.7124418846714315, + "grad_norm": 1.8279383182525635, + "learning_rate": 0.00012777358699687675, + "loss": 0.6948, + "step": 24678 + }, + { + "epoch": 1.7125112761085282, + "grad_norm": 1.8827391862869263, + "learning_rate": 0.00012776818938808407, + "loss": 1.2204, + "step": 24679 + }, + { + "epoch": 1.712580667545625, + "grad_norm": 2.395446538925171, + "learning_rate": 0.0001277627916916305, + "loss": 0.8634, + "step": 24680 + }, + { + "epoch": 1.7126500589827214, + "grad_norm": 1.8007242679595947, + "learning_rate": 0.000127757393907533, + "loss": 0.6064, + "step": 24681 + }, + { + "epoch": 1.7127194504198182, + "grad_norm": 2.1159703731536865, + "learning_rate": 0.0001277519960358086, + "loss": 0.7904, + "step": 24682 + }, + { + "epoch": 1.7127888418569148, + "grad_norm": 3.5411524772644043, + "learning_rate": 0.0001277465980764744, + "loss": 0.6247, + "step": 24683 + }, + { + "epoch": 1.7128582332940114, + "grad_norm": 3.872474431991577, + "learning_rate": 0.0001277412000295474, + "loss": 0.4167, + "step": 24684 + }, + { + "epoch": 1.7129276247311083, + "grad_norm": 1.9734801054000854, + "learning_rate": 0.00012773580189504467, + "loss": 0.5804, + "step": 24685 + }, + { + "epoch": 1.7129970161682049, + "grad_norm": 2.0014679431915283, + "learning_rate": 0.0001277304036729832, + "loss": 0.5112, + "step": 24686 + }, + { + "epoch": 1.7130664076053015, + "grad_norm": 1.7267179489135742, + "learning_rate": 0.00012772500536338007, + "loss": 0.5425, + "step": 24687 + }, + { + "epoch": 1.7131357990423983, + "grad_norm": 2.253218650817871, + "learning_rate": 0.0001277196069662523, + "loss": 0.7183, + "step": 24688 + }, + { + "epoch": 1.7132051904794947, + "grad_norm": 2.4344043731689453, + "learning_rate": 0.00012771420848161697, + "loss": 0.7659, + "step": 24689 + }, + { + "epoch": 1.7132745819165915, + "grad_norm": 2.178410291671753, + "learning_rate": 0.0001277088099094911, + "loss": 0.6749, + "step": 24690 + }, + { + "epoch": 1.7133439733536882, + "grad_norm": 1.8032357692718506, + "learning_rate": 0.00012770341124989173, + "loss": 0.741, + "step": 24691 + }, + { + "epoch": 1.7134133647907848, + "grad_norm": 1.8032746315002441, + "learning_rate": 0.00012769801250283588, + "loss": 0.6665, + "step": 24692 + }, + { + "epoch": 1.7134827562278816, + "grad_norm": 1.6688669919967651, + "learning_rate": 0.00012769261366834067, + "loss": 0.6401, + "step": 24693 + }, + { + "epoch": 1.713552147664978, + "grad_norm": 2.149846315383911, + "learning_rate": 0.00012768721474642304, + "loss": 0.7555, + "step": 24694 + }, + { + "epoch": 1.7136215391020748, + "grad_norm": 2.1069090366363525, + "learning_rate": 0.0001276818157371001, + "loss": 0.7593, + "step": 24695 + }, + { + "epoch": 1.7136909305391714, + "grad_norm": 3.3225977420806885, + "learning_rate": 0.00012767641664038887, + "loss": 0.4131, + "step": 24696 + }, + { + "epoch": 1.713760321976268, + "grad_norm": 1.7061748504638672, + "learning_rate": 0.00012767101745630644, + "loss": 0.633, + "step": 24697 + }, + { + "epoch": 1.7138297134133649, + "grad_norm": 1.3591192960739136, + "learning_rate": 0.0001276656181848698, + "loss": 0.6046, + "step": 24698 + }, + { + "epoch": 1.7138991048504615, + "grad_norm": 1.6746442317962646, + "learning_rate": 0.00012766021882609598, + "loss": 0.5198, + "step": 24699 + }, + { + "epoch": 1.713968496287558, + "grad_norm": 1.5782195329666138, + "learning_rate": 0.0001276548193800021, + "loss": 0.5879, + "step": 24700 + }, + { + "epoch": 1.714037887724655, + "grad_norm": 1.7455281019210815, + "learning_rate": 0.00012764941984660515, + "loss": 0.6224, + "step": 24701 + }, + { + "epoch": 1.7141072791617513, + "grad_norm": 2.283681869506836, + "learning_rate": 0.0001276440202259222, + "loss": 0.6925, + "step": 24702 + }, + { + "epoch": 1.7141766705988482, + "grad_norm": 1.786350131034851, + "learning_rate": 0.00012763862051797027, + "loss": 0.7082, + "step": 24703 + }, + { + "epoch": 1.7142460620359448, + "grad_norm": 1.925075650215149, + "learning_rate": 0.00012763322072276644, + "loss": 0.6534, + "step": 24704 + }, + { + "epoch": 1.7143154534730414, + "grad_norm": 1.3836032152175903, + "learning_rate": 0.00012762782084032773, + "loss": 0.5604, + "step": 24705 + }, + { + "epoch": 1.7143848449101382, + "grad_norm": 1.956191062927246, + "learning_rate": 0.00012762242087067116, + "loss": 0.5804, + "step": 24706 + }, + { + "epoch": 1.7144542363472346, + "grad_norm": 1.703157901763916, + "learning_rate": 0.00012761702081381387, + "loss": 0.6624, + "step": 24707 + }, + { + "epoch": 1.7145236277843314, + "grad_norm": 1.9245648384094238, + "learning_rate": 0.00012761162066977284, + "loss": 0.8063, + "step": 24708 + }, + { + "epoch": 1.714593019221428, + "grad_norm": 1.8780486583709717, + "learning_rate": 0.0001276062204385651, + "loss": 0.4715, + "step": 24709 + }, + { + "epoch": 1.7146624106585246, + "grad_norm": 1.7458503246307373, + "learning_rate": 0.0001276008201202077, + "loss": 0.6306, + "step": 24710 + }, + { + "epoch": 1.7147318020956215, + "grad_norm": 2.1922736167907715, + "learning_rate": 0.00012759541971471777, + "loss": 0.6736, + "step": 24711 + }, + { + "epoch": 1.714801193532718, + "grad_norm": 2.2500553131103516, + "learning_rate": 0.00012759001922211225, + "loss": 0.7688, + "step": 24712 + }, + { + "epoch": 1.7148705849698147, + "grad_norm": 2.491816759109497, + "learning_rate": 0.0001275846186424083, + "loss": 0.6848, + "step": 24713 + }, + { + "epoch": 1.7149399764069115, + "grad_norm": 2.008615493774414, + "learning_rate": 0.00012757921797562288, + "loss": 0.6322, + "step": 24714 + }, + { + "epoch": 1.715009367844008, + "grad_norm": 2.488124370574951, + "learning_rate": 0.00012757381722177307, + "loss": 0.7069, + "step": 24715 + }, + { + "epoch": 1.7150787592811048, + "grad_norm": 1.7056102752685547, + "learning_rate": 0.0001275684163808759, + "loss": 0.7123, + "step": 24716 + }, + { + "epoch": 1.7151481507182014, + "grad_norm": 1.671138882637024, + "learning_rate": 0.00012756301545294844, + "loss": 0.4977, + "step": 24717 + }, + { + "epoch": 1.715217542155298, + "grad_norm": 2.3611462116241455, + "learning_rate": 0.00012755761443800778, + "loss": 0.7292, + "step": 24718 + }, + { + "epoch": 1.7152869335923948, + "grad_norm": 1.6116955280303955, + "learning_rate": 0.00012755221333607088, + "loss": 0.7067, + "step": 24719 + }, + { + "epoch": 1.7153563250294912, + "grad_norm": 1.8395212888717651, + "learning_rate": 0.0001275468121471549, + "loss": 0.58, + "step": 24720 + }, + { + "epoch": 1.715425716466588, + "grad_norm": 2.1600594520568848, + "learning_rate": 0.00012754141087127678, + "loss": 0.6463, + "step": 24721 + }, + { + "epoch": 1.7154951079036846, + "grad_norm": 1.8572033643722534, + "learning_rate": 0.00012753600950845363, + "loss": 0.7188, + "step": 24722 + }, + { + "epoch": 1.7155644993407813, + "grad_norm": 2.3420677185058594, + "learning_rate": 0.0001275306080587025, + "loss": 0.8498, + "step": 24723 + }, + { + "epoch": 1.715633890777878, + "grad_norm": 1.6631933450698853, + "learning_rate": 0.0001275252065220404, + "loss": 0.5334, + "step": 24724 + }, + { + "epoch": 1.7157032822149747, + "grad_norm": 4.4520583152771, + "learning_rate": 0.00012751980489848446, + "loss": 0.747, + "step": 24725 + }, + { + "epoch": 1.7157726736520713, + "grad_norm": 1.4794119596481323, + "learning_rate": 0.00012751440318805167, + "loss": 0.6177, + "step": 24726 + }, + { + "epoch": 1.7158420650891681, + "grad_norm": 2.224946975708008, + "learning_rate": 0.0001275090013907591, + "loss": 0.7294, + "step": 24727 + }, + { + "epoch": 1.7159114565262645, + "grad_norm": 1.5042827129364014, + "learning_rate": 0.00012750359950662382, + "loss": 0.4972, + "step": 24728 + }, + { + "epoch": 1.7159808479633614, + "grad_norm": 2.4671337604522705, + "learning_rate": 0.00012749819753566285, + "loss": 0.5313, + "step": 24729 + }, + { + "epoch": 1.716050239400458, + "grad_norm": 1.4725453853607178, + "learning_rate": 0.00012749279547789326, + "loss": 0.5998, + "step": 24730 + }, + { + "epoch": 1.7161196308375546, + "grad_norm": 1.4608228206634521, + "learning_rate": 0.00012748739333333213, + "loss": 0.6303, + "step": 24731 + }, + { + "epoch": 1.7161890222746514, + "grad_norm": 1.4061789512634277, + "learning_rate": 0.00012748199110199646, + "loss": 0.5165, + "step": 24732 + }, + { + "epoch": 1.7162584137117478, + "grad_norm": 1.3811759948730469, + "learning_rate": 0.00012747658878390334, + "loss": 0.3871, + "step": 24733 + }, + { + "epoch": 1.7163278051488446, + "grad_norm": 1.4656548500061035, + "learning_rate": 0.0001274711863790698, + "loss": 0.4944, + "step": 24734 + }, + { + "epoch": 1.7163971965859413, + "grad_norm": 1.624692440032959, + "learning_rate": 0.00012746578388751295, + "loss": 0.6588, + "step": 24735 + }, + { + "epoch": 1.7164665880230379, + "grad_norm": 2.1208345890045166, + "learning_rate": 0.00012746038130924978, + "loss": 0.5799, + "step": 24736 + }, + { + "epoch": 1.7165359794601347, + "grad_norm": 1.8026254177093506, + "learning_rate": 0.0001274549786442974, + "loss": 0.458, + "step": 24737 + }, + { + "epoch": 1.7166053708972313, + "grad_norm": 1.688747763633728, + "learning_rate": 0.00012744957589267282, + "loss": 0.6447, + "step": 24738 + }, + { + "epoch": 1.716674762334328, + "grad_norm": 1.568185567855835, + "learning_rate": 0.0001274441730543931, + "loss": 0.5308, + "step": 24739 + }, + { + "epoch": 1.7167441537714248, + "grad_norm": 1.9513412714004517, + "learning_rate": 0.00012743877012947534, + "loss": 0.6197, + "step": 24740 + }, + { + "epoch": 1.7168135452085211, + "grad_norm": 1.694015622138977, + "learning_rate": 0.00012743336711793654, + "loss": 0.5605, + "step": 24741 + }, + { + "epoch": 1.716882936645618, + "grad_norm": 1.7087385654449463, + "learning_rate": 0.00012742796401979379, + "loss": 0.4967, + "step": 24742 + }, + { + "epoch": 1.7169523280827146, + "grad_norm": 2.2458431720733643, + "learning_rate": 0.00012742256083506415, + "loss": 0.9082, + "step": 24743 + }, + { + "epoch": 1.7170217195198112, + "grad_norm": 1.9155659675598145, + "learning_rate": 0.00012741715756376464, + "loss": 0.913, + "step": 24744 + }, + { + "epoch": 1.717091110956908, + "grad_norm": 1.8820034265518188, + "learning_rate": 0.0001274117542059124, + "loss": 0.9214, + "step": 24745 + }, + { + "epoch": 1.7171605023940044, + "grad_norm": 1.9635379314422607, + "learning_rate": 0.00012740635076152437, + "loss": 0.665, + "step": 24746 + }, + { + "epoch": 1.7172298938311013, + "grad_norm": 1.4209554195404053, + "learning_rate": 0.00012740094723061768, + "loss": 0.5487, + "step": 24747 + }, + { + "epoch": 1.7172992852681979, + "grad_norm": 1.443388819694519, + "learning_rate": 0.0001273955436132094, + "loss": 0.5107, + "step": 24748 + }, + { + "epoch": 1.7173686767052945, + "grad_norm": 2.1000213623046875, + "learning_rate": 0.00012739013990931657, + "loss": 0.5362, + "step": 24749 + }, + { + "epoch": 1.7174380681423913, + "grad_norm": 1.8772526979446411, + "learning_rate": 0.00012738473611895624, + "loss": 0.5087, + "step": 24750 + }, + { + "epoch": 1.717507459579488, + "grad_norm": 1.6008695363998413, + "learning_rate": 0.00012737933224214545, + "loss": 0.4646, + "step": 24751 + }, + { + "epoch": 1.7175768510165845, + "grad_norm": 1.629729986190796, + "learning_rate": 0.00012737392827890133, + "loss": 0.636, + "step": 24752 + }, + { + "epoch": 1.7176462424536814, + "grad_norm": 1.6845967769622803, + "learning_rate": 0.00012736852422924087, + "loss": 0.7022, + "step": 24753 + }, + { + "epoch": 1.7177156338907777, + "grad_norm": 1.616368055343628, + "learning_rate": 0.00012736312009318113, + "loss": 0.6153, + "step": 24754 + }, + { + "epoch": 1.7177850253278746, + "grad_norm": 1.149376630783081, + "learning_rate": 0.00012735771587073927, + "loss": 0.4958, + "step": 24755 + }, + { + "epoch": 1.7178544167649712, + "grad_norm": 1.6168289184570312, + "learning_rate": 0.00012735231156193222, + "loss": 0.6551, + "step": 24756 + }, + { + "epoch": 1.7179238082020678, + "grad_norm": 1.4826411008834839, + "learning_rate": 0.00012734690716677708, + "loss": 0.667, + "step": 24757 + }, + { + "epoch": 1.7179931996391646, + "grad_norm": 1.871275782585144, + "learning_rate": 0.00012734150268529093, + "loss": 0.7527, + "step": 24758 + }, + { + "epoch": 1.718062591076261, + "grad_norm": 2.359861373901367, + "learning_rate": 0.00012733609811749089, + "loss": 0.6297, + "step": 24759 + }, + { + "epoch": 1.7181319825133579, + "grad_norm": 1.4405616521835327, + "learning_rate": 0.0001273306934633939, + "loss": 0.5728, + "step": 24760 + }, + { + "epoch": 1.7182013739504545, + "grad_norm": 1.2301579713821411, + "learning_rate": 0.0001273252887230171, + "loss": 0.4988, + "step": 24761 + }, + { + "epoch": 1.718270765387551, + "grad_norm": 1.5685840845108032, + "learning_rate": 0.00012731988389637754, + "loss": 0.6746, + "step": 24762 + }, + { + "epoch": 1.718340156824648, + "grad_norm": 1.7211650609970093, + "learning_rate": 0.00012731447898349228, + "loss": 0.6155, + "step": 24763 + }, + { + "epoch": 1.7184095482617445, + "grad_norm": 1.9717234373092651, + "learning_rate": 0.00012730907398437834, + "loss": 0.7331, + "step": 24764 + }, + { + "epoch": 1.7184789396988411, + "grad_norm": 1.8180960416793823, + "learning_rate": 0.00012730366889905284, + "loss": 0.5173, + "step": 24765 + }, + { + "epoch": 1.718548331135938, + "grad_norm": 1.8095468282699585, + "learning_rate": 0.00012729826372753284, + "loss": 0.6182, + "step": 24766 + }, + { + "epoch": 1.7186177225730344, + "grad_norm": 1.8329088687896729, + "learning_rate": 0.00012729285846983538, + "loss": 0.5181, + "step": 24767 + }, + { + "epoch": 1.7186871140101312, + "grad_norm": 1.653051733970642, + "learning_rate": 0.00012728745312597754, + "loss": 0.5628, + "step": 24768 + }, + { + "epoch": 1.7187565054472278, + "grad_norm": 2.266188383102417, + "learning_rate": 0.0001272820476959764, + "loss": 0.5161, + "step": 24769 + }, + { + "epoch": 1.7188258968843244, + "grad_norm": 1.4010306596755981, + "learning_rate": 0.00012727664217984894, + "loss": 0.5356, + "step": 24770 + }, + { + "epoch": 1.7188952883214212, + "grad_norm": 2.098524808883667, + "learning_rate": 0.0001272712365776123, + "loss": 0.6965, + "step": 24771 + }, + { + "epoch": 1.7189646797585176, + "grad_norm": 2.1027047634124756, + "learning_rate": 0.00012726583088928356, + "loss": 0.7857, + "step": 24772 + }, + { + "epoch": 1.7190340711956145, + "grad_norm": 2.877364158630371, + "learning_rate": 0.00012726042511487974, + "loss": 0.741, + "step": 24773 + }, + { + "epoch": 1.719103462632711, + "grad_norm": 1.958807349205017, + "learning_rate": 0.00012725501925441797, + "loss": 0.7249, + "step": 24774 + }, + { + "epoch": 1.7191728540698077, + "grad_norm": 2.028099536895752, + "learning_rate": 0.0001272496133079152, + "loss": 0.6379, + "step": 24775 + }, + { + "epoch": 1.7192422455069045, + "grad_norm": 1.6381020545959473, + "learning_rate": 0.0001272442072753886, + "loss": 0.4847, + "step": 24776 + }, + { + "epoch": 1.7193116369440011, + "grad_norm": 1.2990180253982544, + "learning_rate": 0.00012723880115685516, + "loss": 0.4203, + "step": 24777 + }, + { + "epoch": 1.7193810283810977, + "grad_norm": 1.2230337858200073, + "learning_rate": 0.000127233394952332, + "loss": 0.3544, + "step": 24778 + }, + { + "epoch": 1.7194504198181946, + "grad_norm": 1.9744795560836792, + "learning_rate": 0.00012722798866183615, + "loss": 0.5432, + "step": 24779 + }, + { + "epoch": 1.719519811255291, + "grad_norm": 1.6030304431915283, + "learning_rate": 0.00012722258228538475, + "loss": 0.3935, + "step": 24780 + }, + { + "epoch": 1.7195892026923878, + "grad_norm": 2.438845157623291, + "learning_rate": 0.0001272171758229948, + "loss": 0.6488, + "step": 24781 + }, + { + "epoch": 1.7196585941294844, + "grad_norm": 2.353919744491577, + "learning_rate": 0.00012721176927468335, + "loss": 0.6814, + "step": 24782 + }, + { + "epoch": 1.719727985566581, + "grad_norm": 3.1619138717651367, + "learning_rate": 0.00012720636264046756, + "loss": 0.7677, + "step": 24783 + }, + { + "epoch": 1.7197973770036779, + "grad_norm": 1.6711136102676392, + "learning_rate": 0.0001272009559203644, + "loss": 0.5487, + "step": 24784 + }, + { + "epoch": 1.7198667684407745, + "grad_norm": 1.7235380411148071, + "learning_rate": 0.00012719554911439098, + "loss": 0.6416, + "step": 24785 + }, + { + "epoch": 1.719936159877871, + "grad_norm": 2.623652458190918, + "learning_rate": 0.00012719014222256435, + "loss": 0.6177, + "step": 24786 + }, + { + "epoch": 1.7200055513149677, + "grad_norm": 1.6967228651046753, + "learning_rate": 0.00012718473524490163, + "loss": 0.4926, + "step": 24787 + }, + { + "epoch": 1.7200749427520643, + "grad_norm": 1.7840700149536133, + "learning_rate": 0.00012717932818141983, + "loss": 0.5725, + "step": 24788 + }, + { + "epoch": 1.7201443341891611, + "grad_norm": 1.673601746559143, + "learning_rate": 0.0001271739210321361, + "loss": 0.4928, + "step": 24789 + }, + { + "epoch": 1.7202137256262577, + "grad_norm": 1.6023743152618408, + "learning_rate": 0.0001271685137970674, + "loss": 0.4217, + "step": 24790 + }, + { + "epoch": 1.7202831170633544, + "grad_norm": 1.7324155569076538, + "learning_rate": 0.00012716310647623087, + "loss": 0.6011, + "step": 24791 + }, + { + "epoch": 1.7203525085004512, + "grad_norm": 2.248401165008545, + "learning_rate": 0.00012715769906964357, + "loss": 0.7583, + "step": 24792 + }, + { + "epoch": 1.7204218999375476, + "grad_norm": 2.108269214630127, + "learning_rate": 0.00012715229157732254, + "loss": 0.6797, + "step": 24793 + }, + { + "epoch": 1.7204912913746444, + "grad_norm": 1.6945756673812866, + "learning_rate": 0.0001271468839992849, + "loss": 0.5041, + "step": 24794 + }, + { + "epoch": 1.720560682811741, + "grad_norm": 1.6994297504425049, + "learning_rate": 0.00012714147633554766, + "loss": 0.6083, + "step": 24795 + }, + { + "epoch": 1.7206300742488376, + "grad_norm": 1.4726661443710327, + "learning_rate": 0.00012713606858612796, + "loss": 0.646, + "step": 24796 + }, + { + "epoch": 1.7206994656859345, + "grad_norm": 2.24623966217041, + "learning_rate": 0.00012713066075104284, + "loss": 0.587, + "step": 24797 + }, + { + "epoch": 1.720768857123031, + "grad_norm": 1.7675501108169556, + "learning_rate": 0.00012712525283030938, + "loss": 0.8197, + "step": 24798 + }, + { + "epoch": 1.7208382485601277, + "grad_norm": 1.724840521812439, + "learning_rate": 0.0001271198448239446, + "loss": 0.4841, + "step": 24799 + }, + { + "epoch": 1.7209076399972243, + "grad_norm": 1.1975890398025513, + "learning_rate": 0.00012711443673196564, + "loss": 0.4188, + "step": 24800 + }, + { + "epoch": 1.720977031434321, + "grad_norm": 2.121265172958374, + "learning_rate": 0.0001271090285543896, + "loss": 0.7128, + "step": 24801 + }, + { + "epoch": 1.7210464228714177, + "grad_norm": 1.8262510299682617, + "learning_rate": 0.00012710362029123344, + "loss": 0.6, + "step": 24802 + }, + { + "epoch": 1.7211158143085143, + "grad_norm": 2.131857395172119, + "learning_rate": 0.00012709821194251432, + "loss": 0.5655, + "step": 24803 + }, + { + "epoch": 1.721185205745611, + "grad_norm": 1.7694107294082642, + "learning_rate": 0.00012709280350824928, + "loss": 0.6128, + "step": 24804 + }, + { + "epoch": 1.7212545971827078, + "grad_norm": 1.8095693588256836, + "learning_rate": 0.00012708739498845542, + "loss": 0.6429, + "step": 24805 + }, + { + "epoch": 1.7213239886198042, + "grad_norm": 1.4454131126403809, + "learning_rate": 0.00012708198638314977, + "loss": 0.6591, + "step": 24806 + }, + { + "epoch": 1.721393380056901, + "grad_norm": 1.9626497030258179, + "learning_rate": 0.00012707657769234944, + "loss": 0.579, + "step": 24807 + }, + { + "epoch": 1.7214627714939976, + "grad_norm": 2.173130512237549, + "learning_rate": 0.0001270711689160715, + "loss": 0.6941, + "step": 24808 + }, + { + "epoch": 1.7215321629310942, + "grad_norm": 1.5530012845993042, + "learning_rate": 0.00012706576005433305, + "loss": 0.6162, + "step": 24809 + }, + { + "epoch": 1.721601554368191, + "grad_norm": 1.5365657806396484, + "learning_rate": 0.00012706035110715108, + "loss": 0.5129, + "step": 24810 + }, + { + "epoch": 1.7216709458052877, + "grad_norm": 1.3666678667068481, + "learning_rate": 0.00012705494207454276, + "loss": 0.5897, + "step": 24811 + }, + { + "epoch": 1.7217403372423843, + "grad_norm": 2.157428503036499, + "learning_rate": 0.0001270495329565251, + "loss": 0.6818, + "step": 24812 + }, + { + "epoch": 1.721809728679481, + "grad_norm": 1.5017911195755005, + "learning_rate": 0.00012704412375311522, + "loss": 0.5591, + "step": 24813 + }, + { + "epoch": 1.7218791201165775, + "grad_norm": 1.17620050907135, + "learning_rate": 0.00012703871446433017, + "loss": 0.4206, + "step": 24814 + }, + { + "epoch": 1.7219485115536743, + "grad_norm": 1.65938138961792, + "learning_rate": 0.00012703330509018706, + "loss": 0.449, + "step": 24815 + }, + { + "epoch": 1.722017902990771, + "grad_norm": 2.767526626586914, + "learning_rate": 0.0001270278956307029, + "loss": 0.5388, + "step": 24816 + }, + { + "epoch": 1.7220872944278676, + "grad_norm": 1.6230989694595337, + "learning_rate": 0.00012702248608589483, + "loss": 0.4614, + "step": 24817 + }, + { + "epoch": 1.7221566858649644, + "grad_norm": 1.6798685789108276, + "learning_rate": 0.0001270170764557799, + "loss": 0.6687, + "step": 24818 + }, + { + "epoch": 1.7222260773020608, + "grad_norm": 1.8118996620178223, + "learning_rate": 0.0001270116667403752, + "loss": 0.7113, + "step": 24819 + }, + { + "epoch": 1.7222954687391576, + "grad_norm": 1.5255228281021118, + "learning_rate": 0.0001270062569396978, + "loss": 0.5542, + "step": 24820 + }, + { + "epoch": 1.7223648601762542, + "grad_norm": 1.5378854274749756, + "learning_rate": 0.00012700084705376483, + "loss": 0.5921, + "step": 24821 + }, + { + "epoch": 1.7224342516133508, + "grad_norm": 1.5640144348144531, + "learning_rate": 0.00012699543708259326, + "loss": 0.4956, + "step": 24822 + }, + { + "epoch": 1.7225036430504477, + "grad_norm": 1.6526163816452026, + "learning_rate": 0.0001269900270262002, + "loss": 0.6994, + "step": 24823 + }, + { + "epoch": 1.7225730344875443, + "grad_norm": 1.6497589349746704, + "learning_rate": 0.00012698461688460282, + "loss": 0.5462, + "step": 24824 + }, + { + "epoch": 1.722642425924641, + "grad_norm": 1.833306074142456, + "learning_rate": 0.00012697920665781813, + "loss": 0.5117, + "step": 24825 + }, + { + "epoch": 1.7227118173617375, + "grad_norm": 1.456369400024414, + "learning_rate": 0.00012697379634586323, + "loss": 0.6411, + "step": 24826 + }, + { + "epoch": 1.7227812087988341, + "grad_norm": 6.387732982635498, + "learning_rate": 0.00012696838594875511, + "loss": 0.5227, + "step": 24827 + }, + { + "epoch": 1.722850600235931, + "grad_norm": 1.4029911756515503, + "learning_rate": 0.000126962975466511, + "loss": 0.5329, + "step": 24828 + }, + { + "epoch": 1.7229199916730276, + "grad_norm": 1.9964522123336792, + "learning_rate": 0.00012695756489914786, + "loss": 0.7448, + "step": 24829 + }, + { + "epoch": 1.7229893831101242, + "grad_norm": 1.5957317352294922, + "learning_rate": 0.0001269521542466828, + "loss": 0.617, + "step": 24830 + }, + { + "epoch": 1.723058774547221, + "grad_norm": 1.4211390018463135, + "learning_rate": 0.00012694674350913294, + "loss": 0.455, + "step": 24831 + }, + { + "epoch": 1.7231281659843174, + "grad_norm": 1.6134923696517944, + "learning_rate": 0.00012694133268651537, + "loss": 0.5727, + "step": 24832 + }, + { + "epoch": 1.7231975574214142, + "grad_norm": 1.5948625802993774, + "learning_rate": 0.0001269359217788471, + "loss": 0.6034, + "step": 24833 + }, + { + "epoch": 1.7232669488585108, + "grad_norm": 1.5092575550079346, + "learning_rate": 0.00012693051078614524, + "loss": 0.717, + "step": 24834 + }, + { + "epoch": 1.7233363402956074, + "grad_norm": 1.538747787475586, + "learning_rate": 0.0001269250997084269, + "loss": 0.4898, + "step": 24835 + }, + { + "epoch": 1.7234057317327043, + "grad_norm": 1.2610623836517334, + "learning_rate": 0.00012691968854570916, + "loss": 0.4426, + "step": 24836 + }, + { + "epoch": 1.723475123169801, + "grad_norm": 1.1969971656799316, + "learning_rate": 0.00012691427729800904, + "loss": 0.3458, + "step": 24837 + }, + { + "epoch": 1.7235445146068975, + "grad_norm": 1.657304286956787, + "learning_rate": 0.00012690886596534372, + "loss": 0.4821, + "step": 24838 + }, + { + "epoch": 1.7236139060439943, + "grad_norm": 1.8307628631591797, + "learning_rate": 0.0001269034545477302, + "loss": 0.6543, + "step": 24839 + }, + { + "epoch": 1.7236832974810907, + "grad_norm": 1.6329090595245361, + "learning_rate": 0.0001268980430451856, + "loss": 0.7209, + "step": 24840 + }, + { + "epoch": 1.7237526889181876, + "grad_norm": 1.8819377422332764, + "learning_rate": 0.000126892631457727, + "loss": 0.536, + "step": 24841 + }, + { + "epoch": 1.7238220803552842, + "grad_norm": 1.5316401720046997, + "learning_rate": 0.0001268872197853715, + "loss": 0.4924, + "step": 24842 + }, + { + "epoch": 1.7238914717923808, + "grad_norm": 1.4137409925460815, + "learning_rate": 0.00012688180802813618, + "loss": 0.6169, + "step": 24843 + }, + { + "epoch": 1.7239608632294776, + "grad_norm": 1.9512170553207397, + "learning_rate": 0.00012687639618603807, + "loss": 0.5347, + "step": 24844 + }, + { + "epoch": 1.724030254666574, + "grad_norm": 1.7337735891342163, + "learning_rate": 0.00012687098425909435, + "loss": 0.545, + "step": 24845 + }, + { + "epoch": 1.7240996461036708, + "grad_norm": 1.3508448600769043, + "learning_rate": 0.000126865572247322, + "loss": 0.4803, + "step": 24846 + }, + { + "epoch": 1.7241690375407674, + "grad_norm": 1.5415304899215698, + "learning_rate": 0.00012686016015073817, + "loss": 0.4294, + "step": 24847 + }, + { + "epoch": 1.724238428977864, + "grad_norm": 2.223846912384033, + "learning_rate": 0.0001268547479693599, + "loss": 0.628, + "step": 24848 + }, + { + "epoch": 1.724307820414961, + "grad_norm": 1.411400318145752, + "learning_rate": 0.00012684933570320433, + "loss": 0.5718, + "step": 24849 + }, + { + "epoch": 1.7243772118520575, + "grad_norm": 2.2508952617645264, + "learning_rate": 0.00012684392335228853, + "loss": 0.6504, + "step": 24850 + }, + { + "epoch": 1.7244466032891541, + "grad_norm": 2.044301748275757, + "learning_rate": 0.00012683851091662958, + "loss": 0.5236, + "step": 24851 + }, + { + "epoch": 1.724515994726251, + "grad_norm": 1.8162341117858887, + "learning_rate": 0.00012683309839624455, + "loss": 0.7514, + "step": 24852 + }, + { + "epoch": 1.7245853861633473, + "grad_norm": 2.1113388538360596, + "learning_rate": 0.00012682768579115058, + "loss": 0.5026, + "step": 24853 + }, + { + "epoch": 1.7246547776004442, + "grad_norm": 1.4783034324645996, + "learning_rate": 0.00012682227310136467, + "loss": 0.5755, + "step": 24854 + }, + { + "epoch": 1.7247241690375408, + "grad_norm": 1.8222627639770508, + "learning_rate": 0.00012681686032690396, + "loss": 0.6056, + "step": 24855 + }, + { + "epoch": 1.7247935604746374, + "grad_norm": 2.1118004322052, + "learning_rate": 0.00012681144746778557, + "loss": 0.685, + "step": 24856 + }, + { + "epoch": 1.7248629519117342, + "grad_norm": 1.9002447128295898, + "learning_rate": 0.00012680603452402652, + "loss": 0.546, + "step": 24857 + }, + { + "epoch": 1.7249323433488306, + "grad_norm": 1.7952120304107666, + "learning_rate": 0.00012680062149564392, + "loss": 0.5228, + "step": 24858 + }, + { + "epoch": 1.7250017347859274, + "grad_norm": 2.060777187347412, + "learning_rate": 0.0001267952083826549, + "loss": 0.6861, + "step": 24859 + }, + { + "epoch": 1.725071126223024, + "grad_norm": 1.9743095636367798, + "learning_rate": 0.00012678979518507649, + "loss": 0.5961, + "step": 24860 + }, + { + "epoch": 1.7251405176601207, + "grad_norm": 1.7894389629364014, + "learning_rate": 0.00012678438190292585, + "loss": 0.7445, + "step": 24861 + }, + { + "epoch": 1.7252099090972175, + "grad_norm": 1.8675907850265503, + "learning_rate": 0.00012677896853621997, + "loss": 0.4469, + "step": 24862 + }, + { + "epoch": 1.725279300534314, + "grad_norm": 1.7498468160629272, + "learning_rate": 0.00012677355508497603, + "loss": 0.5259, + "step": 24863 + }, + { + "epoch": 1.7253486919714107, + "grad_norm": 1.4190868139266968, + "learning_rate": 0.00012676814154921105, + "loss": 0.6735, + "step": 24864 + }, + { + "epoch": 1.7254180834085076, + "grad_norm": 2.2929515838623047, + "learning_rate": 0.00012676272792894218, + "loss": 0.5166, + "step": 24865 + }, + { + "epoch": 1.725487474845604, + "grad_norm": 1.4840235710144043, + "learning_rate": 0.00012675731422418649, + "loss": 0.5304, + "step": 24866 + }, + { + "epoch": 1.7255568662827008, + "grad_norm": 1.7856274843215942, + "learning_rate": 0.00012675190043496103, + "loss": 0.636, + "step": 24867 + }, + { + "epoch": 1.7256262577197974, + "grad_norm": 2.0458695888519287, + "learning_rate": 0.00012674648656128295, + "loss": 0.5454, + "step": 24868 + }, + { + "epoch": 1.725695649156894, + "grad_norm": 1.9174147844314575, + "learning_rate": 0.0001267410726031693, + "loss": 0.5756, + "step": 24869 + }, + { + "epoch": 1.7257650405939908, + "grad_norm": 1.5457608699798584, + "learning_rate": 0.0001267356585606372, + "loss": 0.6565, + "step": 24870 + }, + { + "epoch": 1.7258344320310872, + "grad_norm": 4.359904766082764, + "learning_rate": 0.00012673024443370372, + "loss": 0.5778, + "step": 24871 + }, + { + "epoch": 1.725903823468184, + "grad_norm": 2.9114506244659424, + "learning_rate": 0.00012672483022238594, + "loss": 0.8117, + "step": 24872 + }, + { + "epoch": 1.7259732149052807, + "grad_norm": 1.8767791986465454, + "learning_rate": 0.00012671941592670103, + "loss": 0.7461, + "step": 24873 + }, + { + "epoch": 1.7260426063423773, + "grad_norm": 1.8438255786895752, + "learning_rate": 0.00012671400154666595, + "loss": 0.6017, + "step": 24874 + }, + { + "epoch": 1.726111997779474, + "grad_norm": 2.04756236076355, + "learning_rate": 0.0001267085870822979, + "loss": 0.7791, + "step": 24875 + }, + { + "epoch": 1.7261813892165707, + "grad_norm": 1.737747311592102, + "learning_rate": 0.00012670317253361395, + "loss": 0.4905, + "step": 24876 + }, + { + "epoch": 1.7262507806536673, + "grad_norm": 1.7117526531219482, + "learning_rate": 0.00012669775790063117, + "loss": 0.5628, + "step": 24877 + }, + { + "epoch": 1.7263201720907642, + "grad_norm": 2.1766397953033447, + "learning_rate": 0.00012669234318336668, + "loss": 0.7408, + "step": 24878 + }, + { + "epoch": 1.7263895635278605, + "grad_norm": 2.498495578765869, + "learning_rate": 0.00012668692838183757, + "loss": 0.8813, + "step": 24879 + }, + { + "epoch": 1.7264589549649574, + "grad_norm": 1.8578498363494873, + "learning_rate": 0.0001266815134960609, + "loss": 0.5975, + "step": 24880 + }, + { + "epoch": 1.726528346402054, + "grad_norm": 1.7947999238967896, + "learning_rate": 0.0001266760985260538, + "loss": 0.4965, + "step": 24881 + }, + { + "epoch": 1.7265977378391506, + "grad_norm": 2.0757150650024414, + "learning_rate": 0.0001266706834718333, + "loss": 0.5257, + "step": 24882 + }, + { + "epoch": 1.7266671292762474, + "grad_norm": 1.946494698524475, + "learning_rate": 0.0001266652683334166, + "loss": 0.5159, + "step": 24883 + }, + { + "epoch": 1.7267365207133438, + "grad_norm": 3.2391488552093506, + "learning_rate": 0.00012665985311082072, + "loss": 0.7405, + "step": 24884 + }, + { + "epoch": 1.7268059121504407, + "grad_norm": 1.3933500051498413, + "learning_rate": 0.00012665443780406278, + "loss": 0.4403, + "step": 24885 + }, + { + "epoch": 1.7268753035875373, + "grad_norm": 1.6620715856552124, + "learning_rate": 0.00012664902241315987, + "loss": 0.781, + "step": 24886 + }, + { + "epoch": 1.7269446950246339, + "grad_norm": 2.3671140670776367, + "learning_rate": 0.00012664360693812914, + "loss": 0.8262, + "step": 24887 + }, + { + "epoch": 1.7270140864617307, + "grad_norm": 2.3122358322143555, + "learning_rate": 0.00012663819137898757, + "loss": 0.6849, + "step": 24888 + }, + { + "epoch": 1.7270834778988273, + "grad_norm": 1.6341315507888794, + "learning_rate": 0.00012663277573575233, + "loss": 0.546, + "step": 24889 + }, + { + "epoch": 1.727152869335924, + "grad_norm": 2.226215124130249, + "learning_rate": 0.00012662736000844054, + "loss": 0.7259, + "step": 24890 + }, + { + "epoch": 1.7272222607730208, + "grad_norm": 2.3762142658233643, + "learning_rate": 0.00012662194419706924, + "loss": 0.8643, + "step": 24891 + }, + { + "epoch": 1.7272916522101172, + "grad_norm": 1.3880431652069092, + "learning_rate": 0.00012661652830165556, + "loss": 0.5107, + "step": 24892 + }, + { + "epoch": 1.727361043647214, + "grad_norm": 2.0979726314544678, + "learning_rate": 0.00012661111232221658, + "loss": 0.566, + "step": 24893 + }, + { + "epoch": 1.7274304350843106, + "grad_norm": 3.261115789413452, + "learning_rate": 0.0001266056962587694, + "loss": 0.6998, + "step": 24894 + }, + { + "epoch": 1.7274998265214072, + "grad_norm": 2.4106624126434326, + "learning_rate": 0.00012660028011133112, + "loss": 0.662, + "step": 24895 + }, + { + "epoch": 1.727569217958504, + "grad_norm": 1.9137276411056519, + "learning_rate": 0.00012659486387991888, + "loss": 0.5945, + "step": 24896 + }, + { + "epoch": 1.7276386093956004, + "grad_norm": 1.8714557886123657, + "learning_rate": 0.0001265894475645497, + "loss": 0.5146, + "step": 24897 + }, + { + "epoch": 1.7277080008326973, + "grad_norm": 1.6696343421936035, + "learning_rate": 0.00012658403116524072, + "loss": 0.6851, + "step": 24898 + }, + { + "epoch": 1.7277773922697939, + "grad_norm": 1.7239466905593872, + "learning_rate": 0.00012657861468200906, + "loss": 0.706, + "step": 24899 + }, + { + "epoch": 1.7278467837068905, + "grad_norm": 2.67720103263855, + "learning_rate": 0.00012657319811487178, + "loss": 0.7585, + "step": 24900 + }, + { + "epoch": 1.7279161751439873, + "grad_norm": 1.7446304559707642, + "learning_rate": 0.00012656778146384604, + "loss": 0.7741, + "step": 24901 + }, + { + "epoch": 1.727985566581084, + "grad_norm": 1.4256923198699951, + "learning_rate": 0.00012656236472894885, + "loss": 0.6436, + "step": 24902 + }, + { + "epoch": 1.7280549580181805, + "grad_norm": 2.0148942470550537, + "learning_rate": 0.00012655694791019736, + "loss": 0.6502, + "step": 24903 + }, + { + "epoch": 1.7281243494552774, + "grad_norm": 1.2371807098388672, + "learning_rate": 0.00012655153100760868, + "loss": 0.4369, + "step": 24904 + }, + { + "epoch": 1.7281937408923738, + "grad_norm": 2.7863221168518066, + "learning_rate": 0.00012654611402119988, + "loss": 0.6023, + "step": 24905 + }, + { + "epoch": 1.7282631323294706, + "grad_norm": 1.6745433807373047, + "learning_rate": 0.0001265406969509881, + "loss": 0.5507, + "step": 24906 + }, + { + "epoch": 1.7283325237665672, + "grad_norm": 1.736214518547058, + "learning_rate": 0.0001265352797969904, + "loss": 0.6473, + "step": 24907 + }, + { + "epoch": 1.7284019152036638, + "grad_norm": 2.5831422805786133, + "learning_rate": 0.0001265298625592239, + "loss": 0.7754, + "step": 24908 + }, + { + "epoch": 1.7284713066407607, + "grad_norm": 1.84392511844635, + "learning_rate": 0.0001265244452377057, + "loss": 0.4587, + "step": 24909 + }, + { + "epoch": 1.728540698077857, + "grad_norm": 2.041816234588623, + "learning_rate": 0.0001265190278324529, + "loss": 0.6429, + "step": 24910 + }, + { + "epoch": 1.7286100895149539, + "grad_norm": 1.6202181577682495, + "learning_rate": 0.00012651361034348265, + "loss": 0.5519, + "step": 24911 + }, + { + "epoch": 1.7286794809520505, + "grad_norm": 2.0897932052612305, + "learning_rate": 0.00012650819277081196, + "loss": 0.5257, + "step": 24912 + }, + { + "epoch": 1.728748872389147, + "grad_norm": 1.7692126035690308, + "learning_rate": 0.00012650277511445804, + "loss": 0.434, + "step": 24913 + }, + { + "epoch": 1.728818263826244, + "grad_norm": 1.6092097759246826, + "learning_rate": 0.00012649735737443787, + "loss": 0.6147, + "step": 24914 + }, + { + "epoch": 1.7288876552633405, + "grad_norm": 1.8308792114257812, + "learning_rate": 0.00012649193955076865, + "loss": 0.5861, + "step": 24915 + }, + { + "epoch": 1.7289570467004372, + "grad_norm": 2.351189136505127, + "learning_rate": 0.00012648652164346744, + "loss": 0.4947, + "step": 24916 + }, + { + "epoch": 1.729026438137534, + "grad_norm": 1.651146411895752, + "learning_rate": 0.00012648110365255132, + "loss": 0.6248, + "step": 24917 + }, + { + "epoch": 1.7290958295746304, + "grad_norm": 1.415754795074463, + "learning_rate": 0.00012647568557803748, + "loss": 0.6752, + "step": 24918 + }, + { + "epoch": 1.7291652210117272, + "grad_norm": 1.9926815032958984, + "learning_rate": 0.00012647026741994296, + "loss": 0.5123, + "step": 24919 + }, + { + "epoch": 1.7292346124488238, + "grad_norm": 2.7558186054229736, + "learning_rate": 0.00012646484917828486, + "loss": 0.9057, + "step": 24920 + }, + { + "epoch": 1.7293040038859204, + "grad_norm": 1.8758654594421387, + "learning_rate": 0.00012645943085308034, + "loss": 0.6293, + "step": 24921 + }, + { + "epoch": 1.7293733953230173, + "grad_norm": 1.6466399431228638, + "learning_rate": 0.00012645401244434644, + "loss": 0.3719, + "step": 24922 + }, + { + "epoch": 1.7294427867601136, + "grad_norm": 2.2078967094421387, + "learning_rate": 0.00012644859395210026, + "loss": 0.4829, + "step": 24923 + }, + { + "epoch": 1.7295121781972105, + "grad_norm": 1.1624212265014648, + "learning_rate": 0.00012644317537635898, + "loss": 0.3334, + "step": 24924 + }, + { + "epoch": 1.729581569634307, + "grad_norm": 1.4193071126937866, + "learning_rate": 0.00012643775671713966, + "loss": 0.4889, + "step": 24925 + }, + { + "epoch": 1.7296509610714037, + "grad_norm": 1.6827198266983032, + "learning_rate": 0.0001264323379744594, + "loss": 0.6662, + "step": 24926 + }, + { + "epoch": 1.7297203525085005, + "grad_norm": 1.8730134963989258, + "learning_rate": 0.0001264269191483353, + "loss": 0.5955, + "step": 24927 + }, + { + "epoch": 1.7297897439455971, + "grad_norm": 1.7079371213912964, + "learning_rate": 0.00012642150023878452, + "loss": 0.4936, + "step": 24928 + }, + { + "epoch": 1.7298591353826938, + "grad_norm": 2.1102874279022217, + "learning_rate": 0.0001264160812458241, + "loss": 0.8868, + "step": 24929 + }, + { + "epoch": 1.7299285268197906, + "grad_norm": 1.9739131927490234, + "learning_rate": 0.00012641066216947117, + "loss": 0.4424, + "step": 24930 + }, + { + "epoch": 1.729997918256887, + "grad_norm": 1.3748666048049927, + "learning_rate": 0.00012640524300974287, + "loss": 0.507, + "step": 24931 + }, + { + "epoch": 1.7300673096939838, + "grad_norm": 1.7878947257995605, + "learning_rate": 0.00012639982376665628, + "loss": 0.5736, + "step": 24932 + }, + { + "epoch": 1.7301367011310804, + "grad_norm": 1.5648328065872192, + "learning_rate": 0.0001263944044402285, + "loss": 0.595, + "step": 24933 + }, + { + "epoch": 1.730206092568177, + "grad_norm": 1.7629294395446777, + "learning_rate": 0.00012638898503047666, + "loss": 0.5984, + "step": 24934 + }, + { + "epoch": 1.7302754840052739, + "grad_norm": 1.812645673751831, + "learning_rate": 0.00012638356553741785, + "loss": 0.4599, + "step": 24935 + }, + { + "epoch": 1.7303448754423703, + "grad_norm": 2.201165199279785, + "learning_rate": 0.00012637814596106917, + "loss": 0.7654, + "step": 24936 + }, + { + "epoch": 1.730414266879467, + "grad_norm": 1.8797236680984497, + "learning_rate": 0.00012637272630144775, + "loss": 0.7091, + "step": 24937 + }, + { + "epoch": 1.7304836583165637, + "grad_norm": 2.075561046600342, + "learning_rate": 0.0001263673065585707, + "loss": 0.8352, + "step": 24938 + }, + { + "epoch": 1.7305530497536603, + "grad_norm": 4.913623332977295, + "learning_rate": 0.00012636188673245511, + "loss": 0.6433, + "step": 24939 + }, + { + "epoch": 1.7306224411907571, + "grad_norm": 1.446603775024414, + "learning_rate": 0.0001263564668231181, + "loss": 0.5463, + "step": 24940 + }, + { + "epoch": 1.7306918326278538, + "grad_norm": 1.9613134860992432, + "learning_rate": 0.0001263510468305768, + "loss": 0.6487, + "step": 24941 + }, + { + "epoch": 1.7307612240649504, + "grad_norm": 2.2710483074188232, + "learning_rate": 0.0001263456267548483, + "loss": 0.6273, + "step": 24942 + }, + { + "epoch": 1.7308306155020472, + "grad_norm": 1.6587488651275635, + "learning_rate": 0.0001263402065959497, + "loss": 0.7136, + "step": 24943 + }, + { + "epoch": 1.7309000069391436, + "grad_norm": 3.4858169555664062, + "learning_rate": 0.00012633478635389812, + "loss": 0.473, + "step": 24944 + }, + { + "epoch": 1.7309693983762404, + "grad_norm": 1.813029408454895, + "learning_rate": 0.00012632936602871068, + "loss": 0.6032, + "step": 24945 + }, + { + "epoch": 1.731038789813337, + "grad_norm": 1.7443383932113647, + "learning_rate": 0.0001263239456204045, + "loss": 0.6756, + "step": 24946 + }, + { + "epoch": 1.7311081812504336, + "grad_norm": 2.027214527130127, + "learning_rate": 0.00012631852512899668, + "loss": 0.588, + "step": 24947 + }, + { + "epoch": 1.7311775726875305, + "grad_norm": 1.8245775699615479, + "learning_rate": 0.00012631310455450435, + "loss": 0.7154, + "step": 24948 + }, + { + "epoch": 1.7312469641246269, + "grad_norm": 1.6674845218658447, + "learning_rate": 0.00012630768389694454, + "loss": 0.5977, + "step": 24949 + }, + { + "epoch": 1.7313163555617237, + "grad_norm": 2.2275502681732178, + "learning_rate": 0.00012630226315633446, + "loss": 0.4622, + "step": 24950 + }, + { + "epoch": 1.7313857469988203, + "grad_norm": 1.8500239849090576, + "learning_rate": 0.0001262968423326912, + "loss": 0.6177, + "step": 24951 + }, + { + "epoch": 1.731455138435917, + "grad_norm": 1.509137749671936, + "learning_rate": 0.00012629142142603182, + "loss": 0.5336, + "step": 24952 + }, + { + "epoch": 1.7315245298730138, + "grad_norm": 2.0269670486450195, + "learning_rate": 0.00012628600043637352, + "loss": 0.7607, + "step": 24953 + }, + { + "epoch": 1.7315939213101104, + "grad_norm": 1.9359875917434692, + "learning_rate": 0.00012628057936373333, + "loss": 0.8113, + "step": 24954 + }, + { + "epoch": 1.731663312747207, + "grad_norm": 2.365988254547119, + "learning_rate": 0.0001262751582081284, + "loss": 0.5888, + "step": 24955 + }, + { + "epoch": 1.7317327041843038, + "grad_norm": 1.394945502281189, + "learning_rate": 0.00012626973696957587, + "loss": 0.4644, + "step": 24956 + }, + { + "epoch": 1.7318020956214002, + "grad_norm": 2.0442142486572266, + "learning_rate": 0.0001262643156480928, + "loss": 0.5303, + "step": 24957 + }, + { + "epoch": 1.731871487058497, + "grad_norm": 1.8911123275756836, + "learning_rate": 0.00012625889424369634, + "loss": 0.674, + "step": 24958 + }, + { + "epoch": 1.7319408784955936, + "grad_norm": 2.5487539768218994, + "learning_rate": 0.00012625347275640363, + "loss": 0.7331, + "step": 24959 + }, + { + "epoch": 1.7320102699326902, + "grad_norm": 1.7172735929489136, + "learning_rate": 0.00012624805118623171, + "loss": 0.5154, + "step": 24960 + }, + { + "epoch": 1.732079661369787, + "grad_norm": 1.5637147426605225, + "learning_rate": 0.00012624262953319775, + "loss": 0.5727, + "step": 24961 + }, + { + "epoch": 1.7321490528068835, + "grad_norm": 2.060063362121582, + "learning_rate": 0.00012623720779731884, + "loss": 0.5831, + "step": 24962 + }, + { + "epoch": 1.7322184442439803, + "grad_norm": 1.1891634464263916, + "learning_rate": 0.00012623178597861214, + "loss": 0.3682, + "step": 24963 + }, + { + "epoch": 1.732287835681077, + "grad_norm": 1.9859914779663086, + "learning_rate": 0.0001262263640770947, + "loss": 0.5317, + "step": 24964 + }, + { + "epoch": 1.7323572271181735, + "grad_norm": 1.6778545379638672, + "learning_rate": 0.0001262209420927837, + "loss": 0.7102, + "step": 24965 + }, + { + "epoch": 1.7324266185552704, + "grad_norm": 1.966796875, + "learning_rate": 0.00012621552002569622, + "loss": 0.6618, + "step": 24966 + }, + { + "epoch": 1.732496009992367, + "grad_norm": 2.1263203620910645, + "learning_rate": 0.0001262100978758494, + "loss": 0.5919, + "step": 24967 + }, + { + "epoch": 1.7325654014294636, + "grad_norm": 1.790092945098877, + "learning_rate": 0.00012620467564326035, + "loss": 0.8227, + "step": 24968 + }, + { + "epoch": 1.7326347928665604, + "grad_norm": 1.5126681327819824, + "learning_rate": 0.00012619925332794612, + "loss": 0.4905, + "step": 24969 + }, + { + "epoch": 1.7327041843036568, + "grad_norm": 1.753670334815979, + "learning_rate": 0.00012619383092992396, + "loss": 0.6635, + "step": 24970 + }, + { + "epoch": 1.7327735757407536, + "grad_norm": 1.3537607192993164, + "learning_rate": 0.00012618840844921087, + "loss": 0.524, + "step": 24971 + }, + { + "epoch": 1.7328429671778502, + "grad_norm": 1.4004020690917969, + "learning_rate": 0.00012618298588582403, + "loss": 0.515, + "step": 24972 + }, + { + "epoch": 1.7329123586149469, + "grad_norm": 2.718471050262451, + "learning_rate": 0.00012617756323978055, + "loss": 0.5985, + "step": 24973 + }, + { + "epoch": 1.7329817500520437, + "grad_norm": 1.7577913999557495, + "learning_rate": 0.0001261721405110975, + "loss": 0.9325, + "step": 24974 + }, + { + "epoch": 1.73305114148914, + "grad_norm": 2.3739805221557617, + "learning_rate": 0.00012616671769979205, + "loss": 0.7557, + "step": 24975 + }, + { + "epoch": 1.733120532926237, + "grad_norm": 1.1228350400924683, + "learning_rate": 0.00012616129480588133, + "loss": 0.4536, + "step": 24976 + }, + { + "epoch": 1.7331899243633335, + "grad_norm": 1.3431686162948608, + "learning_rate": 0.00012615587182938242, + "loss": 0.5845, + "step": 24977 + }, + { + "epoch": 1.7332593158004301, + "grad_norm": 1.5238078832626343, + "learning_rate": 0.00012615044877031245, + "loss": 0.6953, + "step": 24978 + }, + { + "epoch": 1.733328707237527, + "grad_norm": 1.6377854347229004, + "learning_rate": 0.00012614502562868858, + "loss": 0.5347, + "step": 24979 + }, + { + "epoch": 1.7333980986746236, + "grad_norm": 1.5658512115478516, + "learning_rate": 0.00012613960240452786, + "loss": 0.5916, + "step": 24980 + }, + { + "epoch": 1.7334674901117202, + "grad_norm": 2.346859931945801, + "learning_rate": 0.00012613417909784747, + "loss": 0.7119, + "step": 24981 + }, + { + "epoch": 1.733536881548817, + "grad_norm": 1.3105653524398804, + "learning_rate": 0.00012612875570866452, + "loss": 0.6387, + "step": 24982 + }, + { + "epoch": 1.7336062729859134, + "grad_norm": 1.9326814413070679, + "learning_rate": 0.00012612333223699612, + "loss": 0.4391, + "step": 24983 + }, + { + "epoch": 1.7336756644230102, + "grad_norm": 1.4755496978759766, + "learning_rate": 0.00012611790868285934, + "loss": 0.6463, + "step": 24984 + }, + { + "epoch": 1.7337450558601069, + "grad_norm": 1.3072612285614014, + "learning_rate": 0.00012611248504627143, + "loss": 0.509, + "step": 24985 + }, + { + "epoch": 1.7338144472972035, + "grad_norm": 1.6619856357574463, + "learning_rate": 0.00012610706132724938, + "loss": 0.7273, + "step": 24986 + }, + { + "epoch": 1.7338838387343003, + "grad_norm": 1.4256830215454102, + "learning_rate": 0.0001261016375258104, + "loss": 0.4035, + "step": 24987 + }, + { + "epoch": 1.7339532301713967, + "grad_norm": 2.045344829559326, + "learning_rate": 0.00012609621364197154, + "loss": 0.7147, + "step": 24988 + }, + { + "epoch": 1.7340226216084935, + "grad_norm": 3.2597944736480713, + "learning_rate": 0.00012609078967574998, + "loss": 0.706, + "step": 24989 + }, + { + "epoch": 1.7340920130455901, + "grad_norm": 2.009282112121582, + "learning_rate": 0.00012608536562716283, + "loss": 0.6317, + "step": 24990 + }, + { + "epoch": 1.7341614044826867, + "grad_norm": 2.205379009246826, + "learning_rate": 0.0001260799414962272, + "loss": 0.7814, + "step": 24991 + }, + { + "epoch": 1.7342307959197836, + "grad_norm": 1.9711397886276245, + "learning_rate": 0.00012607451728296023, + "loss": 0.5486, + "step": 24992 + }, + { + "epoch": 1.7343001873568802, + "grad_norm": 1.4703574180603027, + "learning_rate": 0.000126069092987379, + "loss": 0.6051, + "step": 24993 + }, + { + "epoch": 1.7343695787939768, + "grad_norm": 1.156347632408142, + "learning_rate": 0.0001260636686095007, + "loss": 0.5693, + "step": 24994 + }, + { + "epoch": 1.7344389702310736, + "grad_norm": 3.000200033187866, + "learning_rate": 0.00012605824414934242, + "loss": 0.55, + "step": 24995 + }, + { + "epoch": 1.73450836166817, + "grad_norm": 1.6989591121673584, + "learning_rate": 0.0001260528196069213, + "loss": 0.5343, + "step": 24996 + }, + { + "epoch": 1.7345777531052669, + "grad_norm": 1.3075624704360962, + "learning_rate": 0.00012604739498225443, + "loss": 0.4973, + "step": 24997 + }, + { + "epoch": 1.7346471445423635, + "grad_norm": 1.9403643608093262, + "learning_rate": 0.00012604197027535894, + "loss": 0.8306, + "step": 24998 + }, + { + "epoch": 1.73471653597946, + "grad_norm": 2.2071855068206787, + "learning_rate": 0.00012603654548625198, + "loss": 0.8493, + "step": 24999 + }, + { + "epoch": 1.734785927416557, + "grad_norm": 1.4590075016021729, + "learning_rate": 0.0001260311206149507, + "loss": 0.6352, + "step": 25000 + }, + { + "epoch": 1.7348553188536533, + "grad_norm": 1.6656874418258667, + "learning_rate": 0.00012602569566147218, + "loss": 0.7116, + "step": 25001 + }, + { + "epoch": 1.7349247102907501, + "grad_norm": 1.0854090452194214, + "learning_rate": 0.00012602027062583355, + "loss": 0.5486, + "step": 25002 + }, + { + "epoch": 1.7349941017278467, + "grad_norm": 1.5321449041366577, + "learning_rate": 0.00012601484550805195, + "loss": 0.5834, + "step": 25003 + }, + { + "epoch": 1.7350634931649433, + "grad_norm": 1.8157541751861572, + "learning_rate": 0.0001260094203081445, + "loss": 0.8201, + "step": 25004 + }, + { + "epoch": 1.7351328846020402, + "grad_norm": 1.3124029636383057, + "learning_rate": 0.0001260039950261283, + "loss": 0.3256, + "step": 25005 + }, + { + "epoch": 1.7352022760391368, + "grad_norm": 1.8991179466247559, + "learning_rate": 0.00012599856966202054, + "loss": 0.6576, + "step": 25006 + }, + { + "epoch": 1.7352716674762334, + "grad_norm": 1.0581165552139282, + "learning_rate": 0.00012599314421583827, + "loss": 0.5317, + "step": 25007 + }, + { + "epoch": 1.7353410589133302, + "grad_norm": 1.572882890701294, + "learning_rate": 0.0001259877186875987, + "loss": 0.544, + "step": 25008 + }, + { + "epoch": 1.7354104503504266, + "grad_norm": 2.0351109504699707, + "learning_rate": 0.0001259822930773189, + "loss": 0.816, + "step": 25009 + }, + { + "epoch": 1.7354798417875235, + "grad_norm": 1.4682648181915283, + "learning_rate": 0.00012597686738501603, + "loss": 0.4581, + "step": 25010 + }, + { + "epoch": 1.73554923322462, + "grad_norm": 1.5733996629714966, + "learning_rate": 0.0001259714416107072, + "loss": 0.6869, + "step": 25011 + }, + { + "epoch": 1.7356186246617167, + "grad_norm": 1.7776178121566772, + "learning_rate": 0.00012596601575440953, + "loss": 0.6286, + "step": 25012 + }, + { + "epoch": 1.7356880160988135, + "grad_norm": 1.724827527999878, + "learning_rate": 0.00012596058981614015, + "loss": 0.4324, + "step": 25013 + }, + { + "epoch": 1.7357574075359101, + "grad_norm": 1.237383484840393, + "learning_rate": 0.0001259551637959162, + "loss": 0.5262, + "step": 25014 + }, + { + "epoch": 1.7358267989730067, + "grad_norm": 1.8108388185501099, + "learning_rate": 0.00012594973769375486, + "loss": 0.4652, + "step": 25015 + }, + { + "epoch": 1.7358961904101033, + "grad_norm": 1.5106801986694336, + "learning_rate": 0.00012594431150967316, + "loss": 0.5625, + "step": 25016 + }, + { + "epoch": 1.7359655818472, + "grad_norm": 2.26226806640625, + "learning_rate": 0.00012593888524368828, + "loss": 0.4706, + "step": 25017 + }, + { + "epoch": 1.7360349732842968, + "grad_norm": 1.8612821102142334, + "learning_rate": 0.00012593345889581737, + "loss": 0.5474, + "step": 25018 + }, + { + "epoch": 1.7361043647213934, + "grad_norm": 2.099095106124878, + "learning_rate": 0.00012592803246607755, + "loss": 0.5869, + "step": 25019 + }, + { + "epoch": 1.73617375615849, + "grad_norm": 1.5161099433898926, + "learning_rate": 0.00012592260595448593, + "loss": 0.6754, + "step": 25020 + }, + { + "epoch": 1.7362431475955868, + "grad_norm": 1.3598363399505615, + "learning_rate": 0.00012591717936105964, + "loss": 0.4879, + "step": 25021 + }, + { + "epoch": 1.7363125390326832, + "grad_norm": 1.7676423788070679, + "learning_rate": 0.00012591175268581584, + "loss": 0.6371, + "step": 25022 + }, + { + "epoch": 1.73638193046978, + "grad_norm": 1.4603524208068848, + "learning_rate": 0.0001259063259287716, + "loss": 0.5557, + "step": 25023 + }, + { + "epoch": 1.7364513219068767, + "grad_norm": 1.4557316303253174, + "learning_rate": 0.00012590089908994415, + "loss": 0.5448, + "step": 25024 + }, + { + "epoch": 1.7365207133439733, + "grad_norm": 1.8616693019866943, + "learning_rate": 0.00012589547216935055, + "loss": 0.6391, + "step": 25025 + }, + { + "epoch": 1.7365901047810701, + "grad_norm": 1.7998734712600708, + "learning_rate": 0.0001258900451670079, + "loss": 0.7337, + "step": 25026 + }, + { + "epoch": 1.7366594962181667, + "grad_norm": 1.6429259777069092, + "learning_rate": 0.0001258846180829334, + "loss": 0.5534, + "step": 25027 + }, + { + "epoch": 1.7367288876552633, + "grad_norm": 1.5027263164520264, + "learning_rate": 0.0001258791909171442, + "loss": 0.826, + "step": 25028 + }, + { + "epoch": 1.73679827909236, + "grad_norm": 1.7599921226501465, + "learning_rate": 0.00012587376366965738, + "loss": 0.6728, + "step": 25029 + }, + { + "epoch": 1.7368676705294566, + "grad_norm": 1.4804906845092773, + "learning_rate": 0.00012586833634049008, + "loss": 0.5323, + "step": 25030 + }, + { + "epoch": 1.7369370619665534, + "grad_norm": 1.7238452434539795, + "learning_rate": 0.00012586290892965945, + "loss": 0.6113, + "step": 25031 + }, + { + "epoch": 1.73700645340365, + "grad_norm": 1.2519712448120117, + "learning_rate": 0.00012585748143718263, + "loss": 0.6048, + "step": 25032 + }, + { + "epoch": 1.7370758448407466, + "grad_norm": 1.4068498611450195, + "learning_rate": 0.00012585205386307676, + "loss": 0.4677, + "step": 25033 + }, + { + "epoch": 1.7371452362778435, + "grad_norm": 2.8749799728393555, + "learning_rate": 0.0001258466262073589, + "loss": 0.9295, + "step": 25034 + }, + { + "epoch": 1.7372146277149398, + "grad_norm": 1.3580036163330078, + "learning_rate": 0.00012584119847004627, + "loss": 0.6725, + "step": 25035 + }, + { + "epoch": 1.7372840191520367, + "grad_norm": 1.7872059345245361, + "learning_rate": 0.00012583577065115596, + "loss": 0.6567, + "step": 25036 + }, + { + "epoch": 1.7373534105891333, + "grad_norm": 1.518401026725769, + "learning_rate": 0.00012583034275070516, + "loss": 0.5432, + "step": 25037 + }, + { + "epoch": 1.73742280202623, + "grad_norm": 1.4874629974365234, + "learning_rate": 0.00012582491476871091, + "loss": 0.6939, + "step": 25038 + }, + { + "epoch": 1.7374921934633267, + "grad_norm": 1.7810810804367065, + "learning_rate": 0.00012581948670519047, + "loss": 0.5664, + "step": 25039 + }, + { + "epoch": 1.7375615849004233, + "grad_norm": 1.37873113155365, + "learning_rate": 0.00012581405856016082, + "loss": 0.416, + "step": 25040 + }, + { + "epoch": 1.73763097633752, + "grad_norm": 2.270721673965454, + "learning_rate": 0.00012580863033363922, + "loss": 0.8317, + "step": 25041 + }, + { + "epoch": 1.7377003677746166, + "grad_norm": 1.4237655401229858, + "learning_rate": 0.00012580320202564278, + "loss": 0.4933, + "step": 25042 + }, + { + "epoch": 1.7377697592117132, + "grad_norm": 1.6603262424468994, + "learning_rate": 0.00012579777363618861, + "loss": 0.6765, + "step": 25043 + }, + { + "epoch": 1.73783915064881, + "grad_norm": 1.3135654926300049, + "learning_rate": 0.00012579234516529387, + "loss": 0.5198, + "step": 25044 + }, + { + "epoch": 1.7379085420859066, + "grad_norm": 1.4063552618026733, + "learning_rate": 0.00012578691661297567, + "loss": 0.3895, + "step": 25045 + }, + { + "epoch": 1.7379779335230032, + "grad_norm": 1.3318718671798706, + "learning_rate": 0.0001257814879792512, + "loss": 0.5166, + "step": 25046 + }, + { + "epoch": 1.7380473249601, + "grad_norm": 1.3159575462341309, + "learning_rate": 0.00012577605926413754, + "loss": 0.7387, + "step": 25047 + }, + { + "epoch": 1.7381167163971964, + "grad_norm": 1.4028910398483276, + "learning_rate": 0.00012577063046765186, + "loss": 0.5246, + "step": 25048 + }, + { + "epoch": 1.7381861078342933, + "grad_norm": 1.398466944694519, + "learning_rate": 0.0001257652015898113, + "loss": 0.5184, + "step": 25049 + }, + { + "epoch": 1.73825549927139, + "grad_norm": 2.069956064224243, + "learning_rate": 0.00012575977263063298, + "loss": 0.7105, + "step": 25050 + }, + { + "epoch": 1.7383248907084865, + "grad_norm": 1.7973461151123047, + "learning_rate": 0.00012575434359013402, + "loss": 0.8145, + "step": 25051 + }, + { + "epoch": 1.7383942821455833, + "grad_norm": 1.363571286201477, + "learning_rate": 0.00012574891446833161, + "loss": 0.5903, + "step": 25052 + }, + { + "epoch": 1.73846367358268, + "grad_norm": 1.436282992362976, + "learning_rate": 0.00012574348526524286, + "loss": 0.55, + "step": 25053 + }, + { + "epoch": 1.7385330650197766, + "grad_norm": 1.4701982736587524, + "learning_rate": 0.00012573805598088496, + "loss": 0.6914, + "step": 25054 + }, + { + "epoch": 1.7386024564568732, + "grad_norm": 1.1896754503250122, + "learning_rate": 0.00012573262661527495, + "loss": 0.4215, + "step": 25055 + }, + { + "epoch": 1.7386718478939698, + "grad_norm": 1.6073799133300781, + "learning_rate": 0.00012572719716843003, + "loss": 0.4959, + "step": 25056 + }, + { + "epoch": 1.7387412393310666, + "grad_norm": 1.0727317333221436, + "learning_rate": 0.00012572176764036734, + "loss": 0.5281, + "step": 25057 + }, + { + "epoch": 1.7388106307681632, + "grad_norm": 1.3456718921661377, + "learning_rate": 0.00012571633803110398, + "loss": 0.4151, + "step": 25058 + }, + { + "epoch": 1.7388800222052598, + "grad_norm": 1.3788682222366333, + "learning_rate": 0.00012571090834065715, + "loss": 0.6057, + "step": 25059 + }, + { + "epoch": 1.7389494136423567, + "grad_norm": 1.5243149995803833, + "learning_rate": 0.000125705478569044, + "loss": 0.6152, + "step": 25060 + }, + { + "epoch": 1.739018805079453, + "grad_norm": 1.4838346242904663, + "learning_rate": 0.0001257000487162816, + "loss": 0.5472, + "step": 25061 + }, + { + "epoch": 1.73908819651655, + "grad_norm": 1.5853875875473022, + "learning_rate": 0.00012569461878238713, + "loss": 0.5833, + "step": 25062 + }, + { + "epoch": 1.7391575879536465, + "grad_norm": 1.8658745288848877, + "learning_rate": 0.00012568918876737774, + "loss": 0.5963, + "step": 25063 + }, + { + "epoch": 1.739226979390743, + "grad_norm": 1.7846050262451172, + "learning_rate": 0.00012568375867127055, + "loss": 0.6386, + "step": 25064 + }, + { + "epoch": 1.73929637082784, + "grad_norm": 1.6588960886001587, + "learning_rate": 0.0001256783284940827, + "loss": 0.6404, + "step": 25065 + }, + { + "epoch": 1.7393657622649366, + "grad_norm": 1.6554749011993408, + "learning_rate": 0.00012567289823583138, + "loss": 0.5201, + "step": 25066 + }, + { + "epoch": 1.7394351537020332, + "grad_norm": 1.2955322265625, + "learning_rate": 0.00012566746789653368, + "loss": 0.3582, + "step": 25067 + }, + { + "epoch": 1.7395045451391298, + "grad_norm": 1.75235116481781, + "learning_rate": 0.00012566203747620675, + "loss": 0.5683, + "step": 25068 + }, + { + "epoch": 1.7395739365762264, + "grad_norm": 1.3427988290786743, + "learning_rate": 0.00012565660697486775, + "loss": 0.4976, + "step": 25069 + }, + { + "epoch": 1.7396433280133232, + "grad_norm": 1.8069689273834229, + "learning_rate": 0.00012565117639253386, + "loss": 0.7875, + "step": 25070 + }, + { + "epoch": 1.7397127194504198, + "grad_norm": 1.8070400953292847, + "learning_rate": 0.00012564574572922213, + "loss": 0.5669, + "step": 25071 + }, + { + "epoch": 1.7397821108875164, + "grad_norm": 1.0957280397415161, + "learning_rate": 0.00012564031498494982, + "loss": 0.4422, + "step": 25072 + }, + { + "epoch": 1.7398515023246133, + "grad_norm": 1.6583433151245117, + "learning_rate": 0.00012563488415973394, + "loss": 0.5942, + "step": 25073 + }, + { + "epoch": 1.7399208937617097, + "grad_norm": 1.4006590843200684, + "learning_rate": 0.00012562945325359176, + "loss": 0.5083, + "step": 25074 + }, + { + "epoch": 1.7399902851988065, + "grad_norm": 2.2067179679870605, + "learning_rate": 0.00012562402226654033, + "loss": 0.5111, + "step": 25075 + }, + { + "epoch": 1.740059676635903, + "grad_norm": 2.032975435256958, + "learning_rate": 0.00012561859119859683, + "loss": 0.6145, + "step": 25076 + }, + { + "epoch": 1.7401290680729997, + "grad_norm": 1.481447458267212, + "learning_rate": 0.00012561316004977845, + "loss": 0.6397, + "step": 25077 + }, + { + "epoch": 1.7401984595100966, + "grad_norm": 1.2030564546585083, + "learning_rate": 0.00012560772882010224, + "loss": 0.3876, + "step": 25078 + }, + { + "epoch": 1.7402678509471932, + "grad_norm": 0.9836276769638062, + "learning_rate": 0.00012560229750958546, + "loss": 0.4028, + "step": 25079 + }, + { + "epoch": 1.7403372423842898, + "grad_norm": 2.365459442138672, + "learning_rate": 0.00012559686611824517, + "loss": 0.5335, + "step": 25080 + }, + { + "epoch": 1.7404066338213866, + "grad_norm": 1.394032597541809, + "learning_rate": 0.00012559143464609856, + "loss": 0.5211, + "step": 25081 + }, + { + "epoch": 1.740476025258483, + "grad_norm": 1.903566837310791, + "learning_rate": 0.00012558600309316273, + "loss": 0.551, + "step": 25082 + }, + { + "epoch": 1.7405454166955798, + "grad_norm": 1.3152055740356445, + "learning_rate": 0.00012558057145945484, + "loss": 0.5004, + "step": 25083 + }, + { + "epoch": 1.7406148081326764, + "grad_norm": 1.4765788316726685, + "learning_rate": 0.0001255751397449921, + "loss": 0.581, + "step": 25084 + }, + { + "epoch": 1.740684199569773, + "grad_norm": 2.342165470123291, + "learning_rate": 0.0001255697079497916, + "loss": 0.6609, + "step": 25085 + }, + { + "epoch": 1.7407535910068699, + "grad_norm": 2.018383741378784, + "learning_rate": 0.0001255642760738705, + "loss": 0.6854, + "step": 25086 + }, + { + "epoch": 1.7408229824439663, + "grad_norm": 1.389981746673584, + "learning_rate": 0.00012555884411724594, + "loss": 0.6561, + "step": 25087 + }, + { + "epoch": 1.740892373881063, + "grad_norm": 1.7891294956207275, + "learning_rate": 0.00012555341207993507, + "loss": 0.6994, + "step": 25088 + }, + { + "epoch": 1.7409617653181597, + "grad_norm": 1.7908014059066772, + "learning_rate": 0.00012554797996195508, + "loss": 0.5208, + "step": 25089 + }, + { + "epoch": 1.7410311567552563, + "grad_norm": 1.7909436225891113, + "learning_rate": 0.00012554254776332303, + "loss": 0.8854, + "step": 25090 + }, + { + "epoch": 1.7411005481923532, + "grad_norm": 1.3585597276687622, + "learning_rate": 0.00012553711548405616, + "loss": 0.5312, + "step": 25091 + }, + { + "epoch": 1.7411699396294498, + "grad_norm": 1.7538310289382935, + "learning_rate": 0.00012553168312417156, + "loss": 0.5837, + "step": 25092 + }, + { + "epoch": 1.7412393310665464, + "grad_norm": 2.0311217308044434, + "learning_rate": 0.00012552625068368636, + "loss": 0.7188, + "step": 25093 + }, + { + "epoch": 1.7413087225036432, + "grad_norm": 2.1901791095733643, + "learning_rate": 0.00012552081816261782, + "loss": 0.6156, + "step": 25094 + }, + { + "epoch": 1.7413781139407396, + "grad_norm": 1.2307130098342896, + "learning_rate": 0.00012551538556098297, + "loss": 0.324, + "step": 25095 + }, + { + "epoch": 1.7414475053778364, + "grad_norm": 1.2154258489608765, + "learning_rate": 0.00012550995287879903, + "loss": 0.5973, + "step": 25096 + }, + { + "epoch": 1.741516896814933, + "grad_norm": 1.2857393026351929, + "learning_rate": 0.00012550452011608312, + "loss": 0.5224, + "step": 25097 + }, + { + "epoch": 1.7415862882520297, + "grad_norm": 1.3669159412384033, + "learning_rate": 0.0001254990872728524, + "loss": 0.4567, + "step": 25098 + }, + { + "epoch": 1.7416556796891265, + "grad_norm": 1.5351831912994385, + "learning_rate": 0.00012549365434912402, + "loss": 0.8524, + "step": 25099 + }, + { + "epoch": 1.7417250711262229, + "grad_norm": 1.5780636072158813, + "learning_rate": 0.00012548822134491512, + "loss": 0.6825, + "step": 25100 + }, + { + "epoch": 1.7417944625633197, + "grad_norm": 1.8513723611831665, + "learning_rate": 0.0001254827882602429, + "loss": 0.6297, + "step": 25101 + }, + { + "epoch": 1.7418638540004163, + "grad_norm": 2.0239875316619873, + "learning_rate": 0.00012547735509512443, + "loss": 0.6122, + "step": 25102 + }, + { + "epoch": 1.741933245437513, + "grad_norm": 1.2608461380004883, + "learning_rate": 0.00012547192184957692, + "loss": 0.6903, + "step": 25103 + }, + { + "epoch": 1.7420026368746098, + "grad_norm": 1.639480471611023, + "learning_rate": 0.00012546648852361752, + "loss": 0.6912, + "step": 25104 + }, + { + "epoch": 1.7420720283117064, + "grad_norm": 1.9816906452178955, + "learning_rate": 0.00012546105511726337, + "loss": 0.8444, + "step": 25105 + }, + { + "epoch": 1.742141419748803, + "grad_norm": 2.08353328704834, + "learning_rate": 0.0001254556216305316, + "loss": 0.6544, + "step": 25106 + }, + { + "epoch": 1.7422108111858998, + "grad_norm": 1.3104737997055054, + "learning_rate": 0.0001254501880634394, + "loss": 0.4699, + "step": 25107 + }, + { + "epoch": 1.7422802026229962, + "grad_norm": 1.5716928243637085, + "learning_rate": 0.00012544475441600392, + "loss": 0.6974, + "step": 25108 + }, + { + "epoch": 1.742349594060093, + "grad_norm": 1.6492418050765991, + "learning_rate": 0.0001254393206882423, + "loss": 0.7254, + "step": 25109 + }, + { + "epoch": 1.7424189854971897, + "grad_norm": 1.9350982904434204, + "learning_rate": 0.00012543388688017167, + "loss": 0.624, + "step": 25110 + }, + { + "epoch": 1.7424883769342863, + "grad_norm": 1.2797490358352661, + "learning_rate": 0.00012542845299180926, + "loss": 0.4506, + "step": 25111 + }, + { + "epoch": 1.742557768371383, + "grad_norm": 1.6499125957489014, + "learning_rate": 0.00012542301902317214, + "loss": 0.5372, + "step": 25112 + }, + { + "epoch": 1.7426271598084795, + "grad_norm": 1.67772376537323, + "learning_rate": 0.0001254175849742775, + "loss": 0.4503, + "step": 25113 + }, + { + "epoch": 1.7426965512455763, + "grad_norm": 2.3139941692352295, + "learning_rate": 0.00012541215084514246, + "loss": 0.6012, + "step": 25114 + }, + { + "epoch": 1.742765942682673, + "grad_norm": 1.3599040508270264, + "learning_rate": 0.00012540671663578427, + "loss": 0.5486, + "step": 25115 + }, + { + "epoch": 1.7428353341197695, + "grad_norm": 1.457380771636963, + "learning_rate": 0.00012540128234621997, + "loss": 0.5688, + "step": 25116 + }, + { + "epoch": 1.7429047255568664, + "grad_norm": 1.4546252489089966, + "learning_rate": 0.00012539584797646678, + "loss": 0.5636, + "step": 25117 + }, + { + "epoch": 1.742974116993963, + "grad_norm": 2.2687125205993652, + "learning_rate": 0.00012539041352654188, + "loss": 0.7183, + "step": 25118 + }, + { + "epoch": 1.7430435084310596, + "grad_norm": 1.7522369623184204, + "learning_rate": 0.00012538497899646238, + "loss": 0.6284, + "step": 25119 + }, + { + "epoch": 1.7431128998681564, + "grad_norm": 1.7777267694473267, + "learning_rate": 0.00012537954438624543, + "loss": 0.608, + "step": 25120 + }, + { + "epoch": 1.7431822913052528, + "grad_norm": 1.3787368535995483, + "learning_rate": 0.00012537410969590818, + "loss": 0.6243, + "step": 25121 + }, + { + "epoch": 1.7432516827423497, + "grad_norm": 1.5507055521011353, + "learning_rate": 0.00012536867492546786, + "loss": 0.6165, + "step": 25122 + }, + { + "epoch": 1.7433210741794463, + "grad_norm": 1.451672077178955, + "learning_rate": 0.00012536324007494156, + "loss": 0.58, + "step": 25123 + }, + { + "epoch": 1.7433904656165429, + "grad_norm": 1.5894668102264404, + "learning_rate": 0.00012535780514434644, + "loss": 0.4574, + "step": 25124 + }, + { + "epoch": 1.7434598570536397, + "grad_norm": 1.4311729669570923, + "learning_rate": 0.0001253523701336997, + "loss": 0.6322, + "step": 25125 + }, + { + "epoch": 1.743529248490736, + "grad_norm": 1.572933554649353, + "learning_rate": 0.00012534693504301844, + "loss": 0.4585, + "step": 25126 + }, + { + "epoch": 1.743598639927833, + "grad_norm": 1.3523504734039307, + "learning_rate": 0.00012534149987231987, + "loss": 0.6254, + "step": 25127 + }, + { + "epoch": 1.7436680313649295, + "grad_norm": 1.5207982063293457, + "learning_rate": 0.00012533606462162108, + "loss": 0.6813, + "step": 25128 + }, + { + "epoch": 1.7437374228020261, + "grad_norm": 1.6004586219787598, + "learning_rate": 0.00012533062929093934, + "loss": 0.48, + "step": 25129 + }, + { + "epoch": 1.743806814239123, + "grad_norm": 1.5022913217544556, + "learning_rate": 0.0001253251938802917, + "loss": 0.4601, + "step": 25130 + }, + { + "epoch": 1.7438762056762196, + "grad_norm": 2.468842029571533, + "learning_rate": 0.00012531975838969535, + "loss": 0.7745, + "step": 25131 + }, + { + "epoch": 1.7439455971133162, + "grad_norm": 1.553346872329712, + "learning_rate": 0.0001253143228191675, + "loss": 0.696, + "step": 25132 + }, + { + "epoch": 1.744014988550413, + "grad_norm": 1.3384382724761963, + "learning_rate": 0.00012530888716872523, + "loss": 0.3672, + "step": 25133 + }, + { + "epoch": 1.7440843799875094, + "grad_norm": 1.5121785402297974, + "learning_rate": 0.00012530345143838575, + "loss": 0.5783, + "step": 25134 + }, + { + "epoch": 1.7441537714246063, + "grad_norm": 2.7060022354125977, + "learning_rate": 0.00012529801562816622, + "loss": 0.6121, + "step": 25135 + }, + { + "epoch": 1.7442231628617029, + "grad_norm": 1.7290951013565063, + "learning_rate": 0.00012529257973808378, + "loss": 0.6809, + "step": 25136 + }, + { + "epoch": 1.7442925542987995, + "grad_norm": 1.390098214149475, + "learning_rate": 0.0001252871437681556, + "loss": 0.6262, + "step": 25137 + }, + { + "epoch": 1.7443619457358963, + "grad_norm": 1.5103733539581299, + "learning_rate": 0.00012528170771839884, + "loss": 0.6076, + "step": 25138 + }, + { + "epoch": 1.7444313371729927, + "grad_norm": 1.378481388092041, + "learning_rate": 0.00012527627158883067, + "loss": 0.6018, + "step": 25139 + }, + { + "epoch": 1.7445007286100895, + "grad_norm": 1.987129807472229, + "learning_rate": 0.00012527083537946826, + "loss": 0.8215, + "step": 25140 + }, + { + "epoch": 1.7445701200471861, + "grad_norm": 1.1802922487258911, + "learning_rate": 0.00012526539909032874, + "loss": 0.4483, + "step": 25141 + }, + { + "epoch": 1.7446395114842828, + "grad_norm": 1.4602112770080566, + "learning_rate": 0.00012525996272142928, + "loss": 0.4415, + "step": 25142 + }, + { + "epoch": 1.7447089029213796, + "grad_norm": 1.5929014682769775, + "learning_rate": 0.00012525452627278706, + "loss": 0.4542, + "step": 25143 + }, + { + "epoch": 1.7447782943584762, + "grad_norm": 1.2508504390716553, + "learning_rate": 0.00012524908974441922, + "loss": 0.5625, + "step": 25144 + }, + { + "epoch": 1.7448476857955728, + "grad_norm": 2.0709214210510254, + "learning_rate": 0.00012524365313634295, + "loss": 0.7049, + "step": 25145 + }, + { + "epoch": 1.7449170772326696, + "grad_norm": 1.5578153133392334, + "learning_rate": 0.00012523821644857538, + "loss": 0.6214, + "step": 25146 + }, + { + "epoch": 1.744986468669766, + "grad_norm": 1.8455191850662231, + "learning_rate": 0.0001252327796811337, + "loss": 0.6569, + "step": 25147 + }, + { + "epoch": 1.7450558601068629, + "grad_norm": 1.3586652278900146, + "learning_rate": 0.00012522734283403503, + "loss": 0.5249, + "step": 25148 + }, + { + "epoch": 1.7451252515439595, + "grad_norm": 1.3117451667785645, + "learning_rate": 0.00012522190590729662, + "loss": 0.6565, + "step": 25149 + }, + { + "epoch": 1.745194642981056, + "grad_norm": 1.6188669204711914, + "learning_rate": 0.00012521646890093554, + "loss": 0.536, + "step": 25150 + }, + { + "epoch": 1.745264034418153, + "grad_norm": 1.8388607501983643, + "learning_rate": 0.00012521103181496897, + "loss": 0.7661, + "step": 25151 + }, + { + "epoch": 1.7453334258552493, + "grad_norm": 1.2208647727966309, + "learning_rate": 0.00012520559464941413, + "loss": 0.3529, + "step": 25152 + }, + { + "epoch": 1.7454028172923461, + "grad_norm": 1.3054683208465576, + "learning_rate": 0.00012520015740428818, + "loss": 0.6805, + "step": 25153 + }, + { + "epoch": 1.7454722087294428, + "grad_norm": 1.8316919803619385, + "learning_rate": 0.00012519472007960819, + "loss": 0.8198, + "step": 25154 + }, + { + "epoch": 1.7455416001665394, + "grad_norm": 1.8706094026565552, + "learning_rate": 0.00012518928267539145, + "loss": 0.7085, + "step": 25155 + }, + { + "epoch": 1.7456109916036362, + "grad_norm": 1.6624131202697754, + "learning_rate": 0.00012518384519165504, + "loss": 0.479, + "step": 25156 + }, + { + "epoch": 1.7456803830407328, + "grad_norm": 1.5242276191711426, + "learning_rate": 0.00012517840762841614, + "loss": 0.6121, + "step": 25157 + }, + { + "epoch": 1.7457497744778294, + "grad_norm": 1.6658811569213867, + "learning_rate": 0.00012517296998569197, + "loss": 0.536, + "step": 25158 + }, + { + "epoch": 1.7458191659149263, + "grad_norm": 1.5749188661575317, + "learning_rate": 0.00012516753226349964, + "loss": 0.5103, + "step": 25159 + }, + { + "epoch": 1.7458885573520226, + "grad_norm": 1.5066306591033936, + "learning_rate": 0.0001251620944618563, + "loss": 0.4883, + "step": 25160 + }, + { + "epoch": 1.7459579487891195, + "grad_norm": 1.4869465827941895, + "learning_rate": 0.00012515665658077919, + "loss": 0.6451, + "step": 25161 + }, + { + "epoch": 1.746027340226216, + "grad_norm": 1.9528173208236694, + "learning_rate": 0.00012515121862028542, + "loss": 0.4121, + "step": 25162 + }, + { + "epoch": 1.7460967316633127, + "grad_norm": 1.6450746059417725, + "learning_rate": 0.00012514578058039216, + "loss": 0.4563, + "step": 25163 + }, + { + "epoch": 1.7461661231004095, + "grad_norm": 1.6452295780181885, + "learning_rate": 0.0001251403424611166, + "loss": 0.749, + "step": 25164 + }, + { + "epoch": 1.746235514537506, + "grad_norm": 1.9609158039093018, + "learning_rate": 0.00012513490426247588, + "loss": 0.7487, + "step": 25165 + }, + { + "epoch": 1.7463049059746028, + "grad_norm": 1.7907609939575195, + "learning_rate": 0.0001251294659844872, + "loss": 0.8413, + "step": 25166 + }, + { + "epoch": 1.7463742974116994, + "grad_norm": 1.8642290830612183, + "learning_rate": 0.0001251240276271677, + "loss": 0.6014, + "step": 25167 + }, + { + "epoch": 1.746443688848796, + "grad_norm": 1.588813066482544, + "learning_rate": 0.00012511858919053457, + "loss": 0.4605, + "step": 25168 + }, + { + "epoch": 1.7465130802858928, + "grad_norm": 1.4439468383789062, + "learning_rate": 0.00012511315067460495, + "loss": 0.5475, + "step": 25169 + }, + { + "epoch": 1.7465824717229894, + "grad_norm": 1.6654138565063477, + "learning_rate": 0.00012510771207939605, + "loss": 0.5353, + "step": 25170 + }, + { + "epoch": 1.746651863160086, + "grad_norm": 1.108212947845459, + "learning_rate": 0.00012510227340492498, + "loss": 0.4327, + "step": 25171 + }, + { + "epoch": 1.7467212545971829, + "grad_norm": 1.3645497560501099, + "learning_rate": 0.00012509683465120896, + "loss": 0.5632, + "step": 25172 + }, + { + "epoch": 1.7467906460342792, + "grad_norm": 2.235978126525879, + "learning_rate": 0.00012509139581826518, + "loss": 0.6083, + "step": 25173 + }, + { + "epoch": 1.746860037471376, + "grad_norm": 1.7132503986358643, + "learning_rate": 0.00012508595690611073, + "loss": 0.4782, + "step": 25174 + }, + { + "epoch": 1.7469294289084727, + "grad_norm": 1.6112291812896729, + "learning_rate": 0.00012508051791476285, + "loss": 0.5708, + "step": 25175 + }, + { + "epoch": 1.7469988203455693, + "grad_norm": 1.2569849491119385, + "learning_rate": 0.0001250750788442387, + "loss": 0.4963, + "step": 25176 + }, + { + "epoch": 1.7470682117826661, + "grad_norm": 1.665120244026184, + "learning_rate": 0.0001250696396945554, + "loss": 0.5323, + "step": 25177 + }, + { + "epoch": 1.7471376032197625, + "grad_norm": 1.346197485923767, + "learning_rate": 0.00012506420046573018, + "loss": 0.4143, + "step": 25178 + }, + { + "epoch": 1.7472069946568594, + "grad_norm": 1.5486724376678467, + "learning_rate": 0.00012505876115778016, + "loss": 0.5719, + "step": 25179 + }, + { + "epoch": 1.747276386093956, + "grad_norm": 1.8797208070755005, + "learning_rate": 0.00012505332177072256, + "loss": 0.8985, + "step": 25180 + }, + { + "epoch": 1.7473457775310526, + "grad_norm": 1.3355469703674316, + "learning_rate": 0.00012504788230457455, + "loss": 0.5484, + "step": 25181 + }, + { + "epoch": 1.7474151689681494, + "grad_norm": 1.560898780822754, + "learning_rate": 0.00012504244275935323, + "loss": 0.5792, + "step": 25182 + }, + { + "epoch": 1.747484560405246, + "grad_norm": 2.281325340270996, + "learning_rate": 0.00012503700313507585, + "loss": 0.5456, + "step": 25183 + }, + { + "epoch": 1.7475539518423426, + "grad_norm": 1.4899965524673462, + "learning_rate": 0.00012503156343175956, + "loss": 0.5726, + "step": 25184 + }, + { + "epoch": 1.7476233432794395, + "grad_norm": 1.290927529335022, + "learning_rate": 0.00012502612364942152, + "loss": 0.4627, + "step": 25185 + }, + { + "epoch": 1.7476927347165359, + "grad_norm": 1.7905726432800293, + "learning_rate": 0.0001250206837880789, + "loss": 0.5831, + "step": 25186 + }, + { + "epoch": 1.7477621261536327, + "grad_norm": 1.7707167863845825, + "learning_rate": 0.0001250152438477489, + "loss": 0.6435, + "step": 25187 + }, + { + "epoch": 1.7478315175907293, + "grad_norm": 1.3155840635299683, + "learning_rate": 0.0001250098038284487, + "loss": 0.4255, + "step": 25188 + }, + { + "epoch": 1.747900909027826, + "grad_norm": 1.4619311094284058, + "learning_rate": 0.00012500436373019542, + "loss": 0.6047, + "step": 25189 + }, + { + "epoch": 1.7479703004649227, + "grad_norm": 2.1486141681671143, + "learning_rate": 0.00012499892355300626, + "loss": 0.6019, + "step": 25190 + }, + { + "epoch": 1.7480396919020191, + "grad_norm": 1.4805964231491089, + "learning_rate": 0.00012499348329689843, + "loss": 0.4927, + "step": 25191 + }, + { + "epoch": 1.748109083339116, + "grad_norm": 1.3698678016662598, + "learning_rate": 0.00012498804296188904, + "loss": 0.4565, + "step": 25192 + }, + { + "epoch": 1.7481784747762126, + "grad_norm": 1.6587884426116943, + "learning_rate": 0.00012498260254799532, + "loss": 0.6698, + "step": 25193 + }, + { + "epoch": 1.7482478662133092, + "grad_norm": 1.6809048652648926, + "learning_rate": 0.00012497716205523443, + "loss": 0.6424, + "step": 25194 + }, + { + "epoch": 1.748317257650406, + "grad_norm": 1.3306082487106323, + "learning_rate": 0.0001249717214836235, + "loss": 0.3853, + "step": 25195 + }, + { + "epoch": 1.7483866490875026, + "grad_norm": 2.170109510421753, + "learning_rate": 0.00012496628083317977, + "loss": 0.6616, + "step": 25196 + }, + { + "epoch": 1.7484560405245992, + "grad_norm": 1.1635844707489014, + "learning_rate": 0.0001249608401039204, + "loss": 0.363, + "step": 25197 + }, + { + "epoch": 1.748525431961696, + "grad_norm": 1.3944634199142456, + "learning_rate": 0.00012495539929586254, + "loss": 0.6405, + "step": 25198 + }, + { + "epoch": 1.7485948233987925, + "grad_norm": 1.4239065647125244, + "learning_rate": 0.00012494995840902338, + "loss": 0.5667, + "step": 25199 + }, + { + "epoch": 1.7486642148358893, + "grad_norm": 1.6007159948349, + "learning_rate": 0.00012494451744342008, + "loss": 0.6459, + "step": 25200 + }, + { + "epoch": 1.748733606272986, + "grad_norm": 2.5084078311920166, + "learning_rate": 0.00012493907639906984, + "loss": 0.8267, + "step": 25201 + }, + { + "epoch": 1.7488029977100825, + "grad_norm": 2.3860247135162354, + "learning_rate": 0.00012493363527598986, + "loss": 0.4816, + "step": 25202 + }, + { + "epoch": 1.7488723891471794, + "grad_norm": 1.1445808410644531, + "learning_rate": 0.00012492819407419723, + "loss": 0.4691, + "step": 25203 + }, + { + "epoch": 1.7489417805842757, + "grad_norm": 1.293816328048706, + "learning_rate": 0.00012492275279370923, + "loss": 0.4634, + "step": 25204 + }, + { + "epoch": 1.7490111720213726, + "grad_norm": 1.8741576671600342, + "learning_rate": 0.00012491731143454297, + "loss": 0.7295, + "step": 25205 + }, + { + "epoch": 1.7490805634584692, + "grad_norm": 1.7196937799453735, + "learning_rate": 0.00012491186999671566, + "loss": 0.6527, + "step": 25206 + }, + { + "epoch": 1.7491499548955658, + "grad_norm": 1.186072587966919, + "learning_rate": 0.00012490642848024444, + "loss": 0.3645, + "step": 25207 + }, + { + "epoch": 1.7492193463326626, + "grad_norm": 1.648606538772583, + "learning_rate": 0.00012490098688514655, + "loss": 0.5981, + "step": 25208 + }, + { + "epoch": 1.7492887377697592, + "grad_norm": 1.7479426860809326, + "learning_rate": 0.00012489554521143912, + "loss": 0.4424, + "step": 25209 + }, + { + "epoch": 1.7493581292068558, + "grad_norm": 1.5445352792739868, + "learning_rate": 0.00012489010345913932, + "loss": 0.5305, + "step": 25210 + }, + { + "epoch": 1.7494275206439527, + "grad_norm": 1.334710717201233, + "learning_rate": 0.0001248846616282644, + "loss": 0.48, + "step": 25211 + }, + { + "epoch": 1.749496912081049, + "grad_norm": 1.3957245349884033, + "learning_rate": 0.00012487921971883145, + "loss": 0.5463, + "step": 25212 + }, + { + "epoch": 1.749566303518146, + "grad_norm": 1.3324815034866333, + "learning_rate": 0.00012487377773085772, + "loss": 0.4514, + "step": 25213 + }, + { + "epoch": 1.7496356949552425, + "grad_norm": 1.8813984394073486, + "learning_rate": 0.00012486833566436033, + "loss": 0.6024, + "step": 25214 + }, + { + "epoch": 1.7497050863923391, + "grad_norm": 1.3843079805374146, + "learning_rate": 0.0001248628935193565, + "loss": 0.6013, + "step": 25215 + }, + { + "epoch": 1.749774477829436, + "grad_norm": 1.5162464380264282, + "learning_rate": 0.00012485745129586342, + "loss": 0.5312, + "step": 25216 + }, + { + "epoch": 1.7498438692665323, + "grad_norm": 1.2571666240692139, + "learning_rate": 0.0001248520089938982, + "loss": 0.4361, + "step": 25217 + }, + { + "epoch": 1.7499132607036292, + "grad_norm": 1.3624608516693115, + "learning_rate": 0.0001248465666134781, + "loss": 0.5037, + "step": 25218 + }, + { + "epoch": 1.7499826521407258, + "grad_norm": 1.436144232749939, + "learning_rate": 0.0001248411241546203, + "loss": 0.612, + "step": 25219 + }, + { + "epoch": 1.7500520435778224, + "grad_norm": 1.476334810256958, + "learning_rate": 0.00012483568161734192, + "loss": 0.5683, + "step": 25220 + }, + { + "epoch": 1.7501214350149192, + "grad_norm": 1.510809063911438, + "learning_rate": 0.0001248302390016602, + "loss": 0.7374, + "step": 25221 + }, + { + "epoch": 1.7501908264520158, + "grad_norm": 1.3007137775421143, + "learning_rate": 0.00012482479630759226, + "loss": 0.6012, + "step": 25222 + }, + { + "epoch": 1.7502602178891125, + "grad_norm": 1.4844621419906616, + "learning_rate": 0.00012481935353515533, + "loss": 0.4646, + "step": 25223 + }, + { + "epoch": 1.7503296093262093, + "grad_norm": 1.4182533025741577, + "learning_rate": 0.00012481391068436658, + "loss": 0.5331, + "step": 25224 + }, + { + "epoch": 1.7503990007633057, + "grad_norm": 1.1640108823776245, + "learning_rate": 0.0001248084677552432, + "loss": 0.5257, + "step": 25225 + }, + { + "epoch": 1.7504683922004025, + "grad_norm": 0.9383459091186523, + "learning_rate": 0.00012480302474780238, + "loss": 0.3627, + "step": 25226 + }, + { + "epoch": 1.7505377836374991, + "grad_norm": 1.734283208847046, + "learning_rate": 0.00012479758166206127, + "loss": 0.6314, + "step": 25227 + }, + { + "epoch": 1.7506071750745957, + "grad_norm": 2.279757499694824, + "learning_rate": 0.00012479213849803705, + "loss": 0.6903, + "step": 25228 + }, + { + "epoch": 1.7506765665116926, + "grad_norm": 1.1944705247879028, + "learning_rate": 0.00012478669525574698, + "loss": 0.4854, + "step": 25229 + }, + { + "epoch": 1.750745957948789, + "grad_norm": 2.0934996604919434, + "learning_rate": 0.00012478125193520818, + "loss": 0.7289, + "step": 25230 + }, + { + "epoch": 1.7508153493858858, + "grad_norm": 1.3160381317138672, + "learning_rate": 0.0001247758085364378, + "loss": 0.4437, + "step": 25231 + }, + { + "epoch": 1.7508847408229824, + "grad_norm": 1.6127080917358398, + "learning_rate": 0.0001247703650594531, + "loss": 0.5714, + "step": 25232 + }, + { + "epoch": 1.750954132260079, + "grad_norm": 1.6574279069900513, + "learning_rate": 0.00012476492150427122, + "loss": 0.5952, + "step": 25233 + }, + { + "epoch": 1.7510235236971758, + "grad_norm": 1.9206714630126953, + "learning_rate": 0.00012475947787090935, + "loss": 0.6379, + "step": 25234 + }, + { + "epoch": 1.7510929151342725, + "grad_norm": 1.5627989768981934, + "learning_rate": 0.0001247540341593847, + "loss": 0.5412, + "step": 25235 + }, + { + "epoch": 1.751162306571369, + "grad_norm": 1.2777702808380127, + "learning_rate": 0.00012474859036971442, + "loss": 0.3991, + "step": 25236 + }, + { + "epoch": 1.751231698008466, + "grad_norm": 1.3788517713546753, + "learning_rate": 0.0001247431465019157, + "loss": 0.5373, + "step": 25237 + }, + { + "epoch": 1.7513010894455623, + "grad_norm": 1.5013419389724731, + "learning_rate": 0.00012473770255600576, + "loss": 0.7191, + "step": 25238 + }, + { + "epoch": 1.7513704808826591, + "grad_norm": 1.5878033638000488, + "learning_rate": 0.00012473225853200175, + "loss": 0.5331, + "step": 25239 + }, + { + "epoch": 1.7514398723197557, + "grad_norm": 2.1680963039398193, + "learning_rate": 0.00012472681442992088, + "loss": 0.5733, + "step": 25240 + }, + { + "epoch": 1.7515092637568523, + "grad_norm": 1.7866886854171753, + "learning_rate": 0.00012472137024978033, + "loss": 0.7539, + "step": 25241 + }, + { + "epoch": 1.7515786551939492, + "grad_norm": 1.7465816736221313, + "learning_rate": 0.00012471592599159727, + "loss": 0.7106, + "step": 25242 + }, + { + "epoch": 1.7516480466310456, + "grad_norm": 1.6396663188934326, + "learning_rate": 0.0001247104816553889, + "loss": 0.5567, + "step": 25243 + }, + { + "epoch": 1.7517174380681424, + "grad_norm": 1.8479161262512207, + "learning_rate": 0.0001247050372411724, + "loss": 0.4822, + "step": 25244 + }, + { + "epoch": 1.751786829505239, + "grad_norm": 1.6291022300720215, + "learning_rate": 0.00012469959274896496, + "loss": 0.6413, + "step": 25245 + }, + { + "epoch": 1.7518562209423356, + "grad_norm": 1.6599429845809937, + "learning_rate": 0.00012469414817878381, + "loss": 0.5443, + "step": 25246 + }, + { + "epoch": 1.7519256123794325, + "grad_norm": 2.003059148788452, + "learning_rate": 0.0001246887035306461, + "loss": 0.4601, + "step": 25247 + }, + { + "epoch": 1.751995003816529, + "grad_norm": 1.554176688194275, + "learning_rate": 0.00012468325880456897, + "loss": 0.4775, + "step": 25248 + }, + { + "epoch": 1.7520643952536257, + "grad_norm": 1.4118975400924683, + "learning_rate": 0.00012467781400056968, + "loss": 0.5365, + "step": 25249 + }, + { + "epoch": 1.7521337866907225, + "grad_norm": 1.6680831909179688, + "learning_rate": 0.0001246723691186654, + "loss": 0.6173, + "step": 25250 + }, + { + "epoch": 1.752203178127819, + "grad_norm": 1.633911371231079, + "learning_rate": 0.0001246669241588733, + "loss": 0.5789, + "step": 25251 + }, + { + "epoch": 1.7522725695649157, + "grad_norm": 1.0165064334869385, + "learning_rate": 0.00012466147912121057, + "loss": 0.463, + "step": 25252 + }, + { + "epoch": 1.7523419610020123, + "grad_norm": 1.4433237314224243, + "learning_rate": 0.00012465603400569445, + "loss": 0.5565, + "step": 25253 + }, + { + "epoch": 1.752411352439109, + "grad_norm": 2.417576789855957, + "learning_rate": 0.00012465058881234206, + "loss": 0.7746, + "step": 25254 + }, + { + "epoch": 1.7524807438762058, + "grad_norm": 1.750977635383606, + "learning_rate": 0.00012464514354117062, + "loss": 0.5695, + "step": 25255 + }, + { + "epoch": 1.7525501353133024, + "grad_norm": 1.394423484802246, + "learning_rate": 0.00012463969819219735, + "loss": 0.4922, + "step": 25256 + }, + { + "epoch": 1.752619526750399, + "grad_norm": 1.2350157499313354, + "learning_rate": 0.0001246342527654394, + "loss": 0.4833, + "step": 25257 + }, + { + "epoch": 1.7526889181874956, + "grad_norm": 1.4933897256851196, + "learning_rate": 0.00012462880726091396, + "loss": 0.683, + "step": 25258 + }, + { + "epoch": 1.7527583096245922, + "grad_norm": 1.7847158908843994, + "learning_rate": 0.0001246233616786382, + "loss": 0.639, + "step": 25259 + }, + { + "epoch": 1.752827701061689, + "grad_norm": 2.775160074234009, + "learning_rate": 0.00012461791601862943, + "loss": 0.6493, + "step": 25260 + }, + { + "epoch": 1.7528970924987857, + "grad_norm": 1.3248454332351685, + "learning_rate": 0.00012461247028090468, + "loss": 0.5463, + "step": 25261 + }, + { + "epoch": 1.7529664839358823, + "grad_norm": 1.4300544261932373, + "learning_rate": 0.00012460702446548126, + "loss": 0.4968, + "step": 25262 + }, + { + "epoch": 1.7530358753729791, + "grad_norm": 1.7720279693603516, + "learning_rate": 0.0001246015785723763, + "loss": 0.6024, + "step": 25263 + }, + { + "epoch": 1.7531052668100755, + "grad_norm": 1.875666856765747, + "learning_rate": 0.000124596132601607, + "loss": 0.5735, + "step": 25264 + }, + { + "epoch": 1.7531746582471723, + "grad_norm": 1.495361566543579, + "learning_rate": 0.0001245906865531906, + "loss": 0.4106, + "step": 25265 + }, + { + "epoch": 1.753244049684269, + "grad_norm": 1.7342448234558105, + "learning_rate": 0.00012458524042714423, + "loss": 0.4046, + "step": 25266 + }, + { + "epoch": 1.7533134411213656, + "grad_norm": 2.2032899856567383, + "learning_rate": 0.00012457979422348512, + "loss": 0.5875, + "step": 25267 + }, + { + "epoch": 1.7533828325584624, + "grad_norm": 1.1817800998687744, + "learning_rate": 0.00012457434794223044, + "loss": 0.4925, + "step": 25268 + }, + { + "epoch": 1.753452223995559, + "grad_norm": 1.6136438846588135, + "learning_rate": 0.00012456890158339738, + "loss": 0.5093, + "step": 25269 + }, + { + "epoch": 1.7535216154326556, + "grad_norm": 1.4567351341247559, + "learning_rate": 0.00012456345514700318, + "loss": 0.6203, + "step": 25270 + }, + { + "epoch": 1.7535910068697522, + "grad_norm": 1.4746874570846558, + "learning_rate": 0.000124558008633065, + "loss": 0.4532, + "step": 25271 + }, + { + "epoch": 1.7536603983068488, + "grad_norm": 2.5200347900390625, + "learning_rate": 0.0001245525620416, + "loss": 0.6245, + "step": 25272 + }, + { + "epoch": 1.7537297897439457, + "grad_norm": 1.1144070625305176, + "learning_rate": 0.00012454711537262543, + "loss": 0.3932, + "step": 25273 + }, + { + "epoch": 1.7537991811810423, + "grad_norm": 1.4169775247573853, + "learning_rate": 0.0001245416686261585, + "loss": 0.6798, + "step": 25274 + }, + { + "epoch": 1.7538685726181389, + "grad_norm": 1.1194568872451782, + "learning_rate": 0.0001245362218022163, + "loss": 0.3267, + "step": 25275 + }, + { + "epoch": 1.7539379640552357, + "grad_norm": 1.4917699098587036, + "learning_rate": 0.00012453077490081612, + "loss": 0.574, + "step": 25276 + }, + { + "epoch": 1.754007355492332, + "grad_norm": 1.5355452299118042, + "learning_rate": 0.00012452532792197517, + "loss": 0.4857, + "step": 25277 + }, + { + "epoch": 1.754076746929429, + "grad_norm": 4.30681037902832, + "learning_rate": 0.00012451988086571058, + "loss": 0.5616, + "step": 25278 + }, + { + "epoch": 1.7541461383665256, + "grad_norm": 1.4847365617752075, + "learning_rate": 0.00012451443373203954, + "loss": 0.5699, + "step": 25279 + }, + { + "epoch": 1.7542155298036222, + "grad_norm": 1.4392701387405396, + "learning_rate": 0.0001245089865209793, + "loss": 0.4985, + "step": 25280 + }, + { + "epoch": 1.754284921240719, + "grad_norm": 1.9224010705947876, + "learning_rate": 0.00012450353923254708, + "loss": 0.6228, + "step": 25281 + }, + { + "epoch": 1.7543543126778156, + "grad_norm": 1.3323678970336914, + "learning_rate": 0.00012449809186675995, + "loss": 0.5303, + "step": 25282 + }, + { + "epoch": 1.7544237041149122, + "grad_norm": 2.0457870960235596, + "learning_rate": 0.00012449264442363525, + "loss": 0.719, + "step": 25283 + }, + { + "epoch": 1.7544930955520088, + "grad_norm": 1.3691174983978271, + "learning_rate": 0.00012448719690319007, + "loss": 0.4431, + "step": 25284 + }, + { + "epoch": 1.7545624869891054, + "grad_norm": 1.3392126560211182, + "learning_rate": 0.00012448174930544169, + "loss": 0.4731, + "step": 25285 + }, + { + "epoch": 1.7546318784262023, + "grad_norm": 1.5780359506607056, + "learning_rate": 0.00012447630163040725, + "loss": 0.5879, + "step": 25286 + }, + { + "epoch": 1.7547012698632989, + "grad_norm": 1.61165189743042, + "learning_rate": 0.00012447085387810394, + "loss": 0.3857, + "step": 25287 + }, + { + "epoch": 1.7547706613003955, + "grad_norm": 1.308813452720642, + "learning_rate": 0.00012446540604854904, + "loss": 0.3787, + "step": 25288 + }, + { + "epoch": 1.7548400527374923, + "grad_norm": 1.4109197854995728, + "learning_rate": 0.00012445995814175964, + "loss": 0.4835, + "step": 25289 + }, + { + "epoch": 1.7549094441745887, + "grad_norm": 1.6322122812271118, + "learning_rate": 0.000124454510157753, + "loss": 0.8043, + "step": 25290 + }, + { + "epoch": 1.7549788356116856, + "grad_norm": 2.378835916519165, + "learning_rate": 0.00012444906209654633, + "loss": 0.6146, + "step": 25291 + }, + { + "epoch": 1.7550482270487822, + "grad_norm": 1.6155897378921509, + "learning_rate": 0.0001244436139581568, + "loss": 0.5637, + "step": 25292 + }, + { + "epoch": 1.7551176184858788, + "grad_norm": 1.9789625406265259, + "learning_rate": 0.0001244381657426016, + "loss": 0.6468, + "step": 25293 + }, + { + "epoch": 1.7551870099229756, + "grad_norm": 1.3581352233886719, + "learning_rate": 0.00012443271744989792, + "loss": 0.5461, + "step": 25294 + }, + { + "epoch": 1.7552564013600722, + "grad_norm": 1.5187044143676758, + "learning_rate": 0.00012442726908006306, + "loss": 0.5291, + "step": 25295 + }, + { + "epoch": 1.7553257927971688, + "grad_norm": 1.7261143922805786, + "learning_rate": 0.00012442182063311408, + "loss": 0.479, + "step": 25296 + }, + { + "epoch": 1.7553951842342654, + "grad_norm": 1.279412865638733, + "learning_rate": 0.00012441637210906827, + "loss": 0.6241, + "step": 25297 + }, + { + "epoch": 1.755464575671362, + "grad_norm": 1.4536339044570923, + "learning_rate": 0.00012441092350794284, + "loss": 0.4624, + "step": 25298 + }, + { + "epoch": 1.7555339671084589, + "grad_norm": 1.331878900527954, + "learning_rate": 0.0001244054748297549, + "loss": 0.4396, + "step": 25299 + }, + { + "epoch": 1.7556033585455555, + "grad_norm": 1.6266050338745117, + "learning_rate": 0.00012440002607452173, + "loss": 0.5568, + "step": 25300 + }, + { + "epoch": 1.755672749982652, + "grad_norm": 1.7478787899017334, + "learning_rate": 0.00012439457724226053, + "loss": 0.753, + "step": 25301 + }, + { + "epoch": 1.755742141419749, + "grad_norm": 1.1894935369491577, + "learning_rate": 0.00012438912833298847, + "loss": 0.4719, + "step": 25302 + }, + { + "epoch": 1.7558115328568453, + "grad_norm": 1.5655931234359741, + "learning_rate": 0.00012438367934672275, + "loss": 0.6009, + "step": 25303 + }, + { + "epoch": 1.7558809242939422, + "grad_norm": 1.405173420906067, + "learning_rate": 0.00012437823028348057, + "loss": 0.4965, + "step": 25304 + }, + { + "epoch": 1.7559503157310388, + "grad_norm": 1.6882028579711914, + "learning_rate": 0.0001243727811432792, + "loss": 0.4369, + "step": 25305 + }, + { + "epoch": 1.7560197071681354, + "grad_norm": 1.998009204864502, + "learning_rate": 0.00012436733192613573, + "loss": 0.7312, + "step": 25306 + }, + { + "epoch": 1.7560890986052322, + "grad_norm": 1.4857856035232544, + "learning_rate": 0.00012436188263206742, + "loss": 0.4126, + "step": 25307 + }, + { + "epoch": 1.7561584900423288, + "grad_norm": 1.798397183418274, + "learning_rate": 0.00012435643326109153, + "loss": 0.5388, + "step": 25308 + }, + { + "epoch": 1.7562278814794254, + "grad_norm": 1.2943716049194336, + "learning_rate": 0.00012435098381322515, + "loss": 0.5368, + "step": 25309 + }, + { + "epoch": 1.7562972729165223, + "grad_norm": 1.4353344440460205, + "learning_rate": 0.00012434553428848554, + "loss": 0.6891, + "step": 25310 + }, + { + "epoch": 1.7563666643536187, + "grad_norm": 1.535099744796753, + "learning_rate": 0.00012434008468688993, + "loss": 0.5608, + "step": 25311 + }, + { + "epoch": 1.7564360557907155, + "grad_norm": 1.6074368953704834, + "learning_rate": 0.0001243346350084555, + "loss": 0.6325, + "step": 25312 + }, + { + "epoch": 1.756505447227812, + "grad_norm": 1.4607346057891846, + "learning_rate": 0.00012432918525319943, + "loss": 0.5367, + "step": 25313 + }, + { + "epoch": 1.7565748386649087, + "grad_norm": 1.8658862113952637, + "learning_rate": 0.00012432373542113894, + "loss": 0.5015, + "step": 25314 + }, + { + "epoch": 1.7566442301020055, + "grad_norm": 1.9620320796966553, + "learning_rate": 0.00012431828551229127, + "loss": 0.6741, + "step": 25315 + }, + { + "epoch": 1.756713621539102, + "grad_norm": 1.0318526029586792, + "learning_rate": 0.00012431283552667358, + "loss": 0.4053, + "step": 25316 + }, + { + "epoch": 1.7567830129761988, + "grad_norm": 7.189537048339844, + "learning_rate": 0.00012430738546430307, + "loss": 0.4486, + "step": 25317 + }, + { + "epoch": 1.7568524044132954, + "grad_norm": 1.5804855823516846, + "learning_rate": 0.000124301935325197, + "loss": 0.656, + "step": 25318 + }, + { + "epoch": 1.756921795850392, + "grad_norm": 1.7236982583999634, + "learning_rate": 0.00012429648510937252, + "loss": 0.5537, + "step": 25319 + }, + { + "epoch": 1.7569911872874888, + "grad_norm": 1.8084295988082886, + "learning_rate": 0.00012429103481684684, + "loss": 0.6138, + "step": 25320 + }, + { + "epoch": 1.7570605787245854, + "grad_norm": 1.4842036962509155, + "learning_rate": 0.0001242855844476372, + "loss": 0.6349, + "step": 25321 + }, + { + "epoch": 1.757129970161682, + "grad_norm": 2.5489985942840576, + "learning_rate": 0.0001242801340017608, + "loss": 0.9494, + "step": 25322 + }, + { + "epoch": 1.7571993615987789, + "grad_norm": 1.4549189805984497, + "learning_rate": 0.0001242746834792348, + "loss": 0.5269, + "step": 25323 + }, + { + "epoch": 1.7572687530358753, + "grad_norm": 1.3996886014938354, + "learning_rate": 0.00012426923288007645, + "loss": 0.3741, + "step": 25324 + }, + { + "epoch": 1.757338144472972, + "grad_norm": 1.269357442855835, + "learning_rate": 0.00012426378220430298, + "loss": 0.5961, + "step": 25325 + }, + { + "epoch": 1.7574075359100687, + "grad_norm": 1.5714377164840698, + "learning_rate": 0.00012425833145193152, + "loss": 0.5186, + "step": 25326 + }, + { + "epoch": 1.7574769273471653, + "grad_norm": 1.9355239868164062, + "learning_rate": 0.00012425288062297934, + "loss": 0.5294, + "step": 25327 + }, + { + "epoch": 1.7575463187842622, + "grad_norm": 1.3736110925674438, + "learning_rate": 0.00012424742971746364, + "loss": 0.682, + "step": 25328 + }, + { + "epoch": 1.7576157102213585, + "grad_norm": 1.7294689416885376, + "learning_rate": 0.00012424197873540162, + "loss": 0.5721, + "step": 25329 + }, + { + "epoch": 1.7576851016584554, + "grad_norm": 1.4088411331176758, + "learning_rate": 0.00012423652767681046, + "loss": 0.5402, + "step": 25330 + }, + { + "epoch": 1.757754493095552, + "grad_norm": 1.578539490699768, + "learning_rate": 0.0001242310765417074, + "loss": 0.6598, + "step": 25331 + }, + { + "epoch": 1.7578238845326486, + "grad_norm": 1.6623592376708984, + "learning_rate": 0.00012422562533010966, + "loss": 0.4488, + "step": 25332 + }, + { + "epoch": 1.7578932759697454, + "grad_norm": 1.5098180770874023, + "learning_rate": 0.00012422017404203442, + "loss": 0.4916, + "step": 25333 + }, + { + "epoch": 1.757962667406842, + "grad_norm": 1.6723353862762451, + "learning_rate": 0.0001242147226774989, + "loss": 0.4507, + "step": 25334 + }, + { + "epoch": 1.7580320588439386, + "grad_norm": 1.8168915510177612, + "learning_rate": 0.00012420927123652035, + "loss": 0.5487, + "step": 25335 + }, + { + "epoch": 1.7581014502810355, + "grad_norm": 1.649466872215271, + "learning_rate": 0.0001242038197191159, + "loss": 0.5729, + "step": 25336 + }, + { + "epoch": 1.7581708417181319, + "grad_norm": 1.7988770008087158, + "learning_rate": 0.00012419836812530282, + "loss": 0.9354, + "step": 25337 + }, + { + "epoch": 1.7582402331552287, + "grad_norm": 1.9478899240493774, + "learning_rate": 0.00012419291645509827, + "loss": 0.8667, + "step": 25338 + }, + { + "epoch": 1.7583096245923253, + "grad_norm": 1.6202762126922607, + "learning_rate": 0.0001241874647085195, + "loss": 0.556, + "step": 25339 + }, + { + "epoch": 1.758379016029422, + "grad_norm": 1.718288540840149, + "learning_rate": 0.0001241820128855837, + "loss": 0.5176, + "step": 25340 + }, + { + "epoch": 1.7584484074665188, + "grad_norm": 1.3211479187011719, + "learning_rate": 0.00012417656098630812, + "loss": 0.4731, + "step": 25341 + }, + { + "epoch": 1.7585177989036151, + "grad_norm": 1.2853820323944092, + "learning_rate": 0.00012417110901070992, + "loss": 0.4795, + "step": 25342 + }, + { + "epoch": 1.758587190340712, + "grad_norm": 1.4436204433441162, + "learning_rate": 0.00012416565695880634, + "loss": 0.5466, + "step": 25343 + }, + { + "epoch": 1.7586565817778086, + "grad_norm": 1.8112046718597412, + "learning_rate": 0.00012416020483061458, + "loss": 0.6023, + "step": 25344 + }, + { + "epoch": 1.7587259732149052, + "grad_norm": 1.8038954734802246, + "learning_rate": 0.00012415475262615185, + "loss": 0.6673, + "step": 25345 + }, + { + "epoch": 1.758795364652002, + "grad_norm": 1.8309845924377441, + "learning_rate": 0.0001241493003454354, + "loss": 0.4868, + "step": 25346 + }, + { + "epoch": 1.7588647560890986, + "grad_norm": 1.4566668272018433, + "learning_rate": 0.00012414384798848236, + "loss": 0.6559, + "step": 25347 + }, + { + "epoch": 1.7589341475261953, + "grad_norm": 1.04926598072052, + "learning_rate": 0.00012413839555531003, + "loss": 0.3378, + "step": 25348 + }, + { + "epoch": 1.759003538963292, + "grad_norm": 1.2210133075714111, + "learning_rate": 0.00012413294304593556, + "loss": 0.3424, + "step": 25349 + }, + { + "epoch": 1.7590729304003885, + "grad_norm": 1.4622303247451782, + "learning_rate": 0.00012412749046037625, + "loss": 0.5191, + "step": 25350 + }, + { + "epoch": 1.7591423218374853, + "grad_norm": 1.5650676488876343, + "learning_rate": 0.0001241220377986492, + "loss": 0.5364, + "step": 25351 + }, + { + "epoch": 1.759211713274582, + "grad_norm": 1.1867153644561768, + "learning_rate": 0.00012411658506077168, + "loss": 0.5363, + "step": 25352 + }, + { + "epoch": 1.7592811047116785, + "grad_norm": 1.8442339897155762, + "learning_rate": 0.0001241111322467609, + "loss": 0.6815, + "step": 25353 + }, + { + "epoch": 1.7593504961487754, + "grad_norm": 1.4570201635360718, + "learning_rate": 0.00012410567935663408, + "loss": 0.5228, + "step": 25354 + }, + { + "epoch": 1.7594198875858718, + "grad_norm": 1.5177850723266602, + "learning_rate": 0.00012410022639040844, + "loss": 0.5476, + "step": 25355 + }, + { + "epoch": 1.7594892790229686, + "grad_norm": 1.5631403923034668, + "learning_rate": 0.00012409477334810113, + "loss": 0.6848, + "step": 25356 + }, + { + "epoch": 1.7595586704600652, + "grad_norm": 1.4990746974945068, + "learning_rate": 0.00012408932022972948, + "loss": 0.4619, + "step": 25357 + }, + { + "epoch": 1.7596280618971618, + "grad_norm": 1.539575219154358, + "learning_rate": 0.0001240838670353106, + "loss": 0.621, + "step": 25358 + }, + { + "epoch": 1.7596974533342586, + "grad_norm": 1.6614032983779907, + "learning_rate": 0.00012407841376486177, + "loss": 0.5767, + "step": 25359 + }, + { + "epoch": 1.7597668447713553, + "grad_norm": 1.7275195121765137, + "learning_rate": 0.00012407296041840016, + "loss": 0.773, + "step": 25360 + }, + { + "epoch": 1.7598362362084519, + "grad_norm": 1.3419569730758667, + "learning_rate": 0.000124067506995943, + "loss": 0.4823, + "step": 25361 + }, + { + "epoch": 1.7599056276455487, + "grad_norm": 1.080998420715332, + "learning_rate": 0.00012406205349750754, + "loss": 0.3755, + "step": 25362 + }, + { + "epoch": 1.759975019082645, + "grad_norm": 2.0729126930236816, + "learning_rate": 0.00012405659992311094, + "loss": 0.5294, + "step": 25363 + }, + { + "epoch": 1.760044410519742, + "grad_norm": 1.457204818725586, + "learning_rate": 0.00012405114627277048, + "loss": 0.4334, + "step": 25364 + }, + { + "epoch": 1.7601138019568385, + "grad_norm": 1.295912504196167, + "learning_rate": 0.0001240456925465033, + "loss": 0.6434, + "step": 25365 + }, + { + "epoch": 1.7601831933939351, + "grad_norm": 2.207026720046997, + "learning_rate": 0.00012404023874432667, + "loss": 0.5572, + "step": 25366 + }, + { + "epoch": 1.760252584831032, + "grad_norm": 1.2417179346084595, + "learning_rate": 0.00012403478486625784, + "loss": 0.586, + "step": 25367 + }, + { + "epoch": 1.7603219762681284, + "grad_norm": 1.4538272619247437, + "learning_rate": 0.00012402933091231394, + "loss": 0.5141, + "step": 25368 + }, + { + "epoch": 1.7603913677052252, + "grad_norm": 1.637961506843567, + "learning_rate": 0.00012402387688251226, + "loss": 0.8852, + "step": 25369 + }, + { + "epoch": 1.7604607591423218, + "grad_norm": 1.1675275564193726, + "learning_rate": 0.00012401842277686998, + "loss": 0.4507, + "step": 25370 + }, + { + "epoch": 1.7605301505794184, + "grad_norm": 1.6439048051834106, + "learning_rate": 0.0001240129685954043, + "loss": 0.6245, + "step": 25371 + }, + { + "epoch": 1.7605995420165153, + "grad_norm": 1.5363993644714355, + "learning_rate": 0.0001240075143381325, + "loss": 0.4627, + "step": 25372 + }, + { + "epoch": 1.7606689334536119, + "grad_norm": 1.5413469076156616, + "learning_rate": 0.00012400206000507172, + "loss": 0.6557, + "step": 25373 + }, + { + "epoch": 1.7607383248907085, + "grad_norm": 1.2105659246444702, + "learning_rate": 0.00012399660559623928, + "loss": 0.3621, + "step": 25374 + }, + { + "epoch": 1.7608077163278053, + "grad_norm": 1.2939534187316895, + "learning_rate": 0.00012399115111165227, + "loss": 0.4524, + "step": 25375 + }, + { + "epoch": 1.7608771077649017, + "grad_norm": 1.5055633783340454, + "learning_rate": 0.00012398569655132802, + "loss": 0.5732, + "step": 25376 + }, + { + "epoch": 1.7609464992019985, + "grad_norm": 1.6199227571487427, + "learning_rate": 0.0001239802419152837, + "loss": 0.617, + "step": 25377 + }, + { + "epoch": 1.7610158906390951, + "grad_norm": 1.7867287397384644, + "learning_rate": 0.00012397478720353654, + "loss": 0.6237, + "step": 25378 + }, + { + "epoch": 1.7610852820761917, + "grad_norm": 1.6172571182250977, + "learning_rate": 0.00012396933241610376, + "loss": 0.7693, + "step": 25379 + }, + { + "epoch": 1.7611546735132886, + "grad_norm": 1.4533635377883911, + "learning_rate": 0.00012396387755300256, + "loss": 0.5939, + "step": 25380 + }, + { + "epoch": 1.761224064950385, + "grad_norm": 1.5369716882705688, + "learning_rate": 0.0001239584226142502, + "loss": 0.719, + "step": 25381 + }, + { + "epoch": 1.7612934563874818, + "grad_norm": 1.76707124710083, + "learning_rate": 0.00012395296759986387, + "loss": 0.5704, + "step": 25382 + }, + { + "epoch": 1.7613628478245784, + "grad_norm": 1.6615560054779053, + "learning_rate": 0.00012394751250986081, + "loss": 0.6108, + "step": 25383 + }, + { + "epoch": 1.761432239261675, + "grad_norm": 1.5478525161743164, + "learning_rate": 0.00012394205734425823, + "loss": 0.4457, + "step": 25384 + }, + { + "epoch": 1.7615016306987719, + "grad_norm": 0.8641141653060913, + "learning_rate": 0.00012393660210307337, + "loss": 0.4134, + "step": 25385 + }, + { + "epoch": 1.7615710221358685, + "grad_norm": 2.2566118240356445, + "learning_rate": 0.0001239311467863234, + "loss": 0.6464, + "step": 25386 + }, + { + "epoch": 1.761640413572965, + "grad_norm": 1.2557342052459717, + "learning_rate": 0.0001239256913940256, + "loss": 0.425, + "step": 25387 + }, + { + "epoch": 1.761709805010062, + "grad_norm": 1.307309865951538, + "learning_rate": 0.00012392023592619714, + "loss": 0.4852, + "step": 25388 + }, + { + "epoch": 1.7617791964471583, + "grad_norm": 2.1672205924987793, + "learning_rate": 0.00012391478038285532, + "loss": 0.8713, + "step": 25389 + }, + { + "epoch": 1.7618485878842551, + "grad_norm": 1.6156840324401855, + "learning_rate": 0.00012390932476401726, + "loss": 0.6561, + "step": 25390 + }, + { + "epoch": 1.7619179793213517, + "grad_norm": 1.457549810409546, + "learning_rate": 0.00012390386906970026, + "loss": 0.5917, + "step": 25391 + }, + { + "epoch": 1.7619873707584484, + "grad_norm": 1.3431978225708008, + "learning_rate": 0.00012389841329992152, + "loss": 0.5319, + "step": 25392 + }, + { + "epoch": 1.7620567621955452, + "grad_norm": 1.298502802848816, + "learning_rate": 0.00012389295745469827, + "loss": 0.4875, + "step": 25393 + }, + { + "epoch": 1.7621261536326416, + "grad_norm": 1.421234369277954, + "learning_rate": 0.00012388750153404768, + "loss": 0.5301, + "step": 25394 + }, + { + "epoch": 1.7621955450697384, + "grad_norm": 1.1668119430541992, + "learning_rate": 0.00012388204553798708, + "loss": 0.3998, + "step": 25395 + }, + { + "epoch": 1.762264936506835, + "grad_norm": 1.861578106880188, + "learning_rate": 0.00012387658946653359, + "loss": 0.7504, + "step": 25396 + }, + { + "epoch": 1.7623343279439316, + "grad_norm": 2.1221537590026855, + "learning_rate": 0.0001238711333197045, + "loss": 0.5653, + "step": 25397 + }, + { + "epoch": 1.7624037193810285, + "grad_norm": 1.4513137340545654, + "learning_rate": 0.00012386567709751702, + "loss": 0.5911, + "step": 25398 + }, + { + "epoch": 1.762473110818125, + "grad_norm": 2.109724521636963, + "learning_rate": 0.00012386022079998834, + "loss": 0.477, + "step": 25399 + }, + { + "epoch": 1.7625425022552217, + "grad_norm": 1.6985818147659302, + "learning_rate": 0.0001238547644271357, + "loss": 0.3984, + "step": 25400 + }, + { + "epoch": 1.7626118936923185, + "grad_norm": 1.5344116687774658, + "learning_rate": 0.00012384930797897635, + "loss": 0.6329, + "step": 25401 + }, + { + "epoch": 1.762681285129415, + "grad_norm": 1.9588370323181152, + "learning_rate": 0.00012384385145552753, + "loss": 0.7103, + "step": 25402 + }, + { + "epoch": 1.7627506765665117, + "grad_norm": 1.808577537536621, + "learning_rate": 0.0001238383948568064, + "loss": 0.7132, + "step": 25403 + }, + { + "epoch": 1.7628200680036084, + "grad_norm": 1.6283304691314697, + "learning_rate": 0.00012383293818283027, + "loss": 0.6389, + "step": 25404 + }, + { + "epoch": 1.762889459440705, + "grad_norm": 1.4464111328125, + "learning_rate": 0.0001238274814336163, + "loss": 0.4611, + "step": 25405 + }, + { + "epoch": 1.7629588508778018, + "grad_norm": 1.6477601528167725, + "learning_rate": 0.00012382202460918174, + "loss": 0.6215, + "step": 25406 + }, + { + "epoch": 1.7630282423148982, + "grad_norm": 1.8296364545822144, + "learning_rate": 0.00012381656770954377, + "loss": 0.6587, + "step": 25407 + }, + { + "epoch": 1.763097633751995, + "grad_norm": 2.0542595386505127, + "learning_rate": 0.0001238111107347197, + "loss": 0.5442, + "step": 25408 + }, + { + "epoch": 1.7631670251890916, + "grad_norm": 1.1776058673858643, + "learning_rate": 0.0001238056536847267, + "loss": 0.3441, + "step": 25409 + }, + { + "epoch": 1.7632364166261882, + "grad_norm": 1.5176371335983276, + "learning_rate": 0.000123800196559582, + "loss": 0.6179, + "step": 25410 + }, + { + "epoch": 1.763305808063285, + "grad_norm": 1.943915843963623, + "learning_rate": 0.00012379473935930285, + "loss": 0.4923, + "step": 25411 + }, + { + "epoch": 1.7633751995003817, + "grad_norm": 1.4103769063949585, + "learning_rate": 0.0001237892820839065, + "loss": 0.7448, + "step": 25412 + }, + { + "epoch": 1.7634445909374783, + "grad_norm": 1.7341227531433105, + "learning_rate": 0.00012378382473341012, + "loss": 0.6367, + "step": 25413 + }, + { + "epoch": 1.7635139823745751, + "grad_norm": 1.5823969841003418, + "learning_rate": 0.00012377836730783094, + "loss": 0.6094, + "step": 25414 + }, + { + "epoch": 1.7635833738116715, + "grad_norm": 1.3434207439422607, + "learning_rate": 0.00012377290980718626, + "loss": 0.4822, + "step": 25415 + }, + { + "epoch": 1.7636527652487684, + "grad_norm": 1.5300291776657104, + "learning_rate": 0.00012376745223149324, + "loss": 0.5459, + "step": 25416 + }, + { + "epoch": 1.763722156685865, + "grad_norm": 1.5154447555541992, + "learning_rate": 0.00012376199458076913, + "loss": 0.5122, + "step": 25417 + }, + { + "epoch": 1.7637915481229616, + "grad_norm": 1.9736876487731934, + "learning_rate": 0.00012375653685503117, + "loss": 0.7656, + "step": 25418 + }, + { + "epoch": 1.7638609395600584, + "grad_norm": 1.9586436748504639, + "learning_rate": 0.00012375107905429657, + "loss": 0.6252, + "step": 25419 + }, + { + "epoch": 1.7639303309971548, + "grad_norm": 1.3128374814987183, + "learning_rate": 0.00012374562117858256, + "loss": 0.5056, + "step": 25420 + }, + { + "epoch": 1.7639997224342516, + "grad_norm": 1.6102988719940186, + "learning_rate": 0.0001237401632279064, + "loss": 0.6636, + "step": 25421 + }, + { + "epoch": 1.7640691138713482, + "grad_norm": 1.6610798835754395, + "learning_rate": 0.00012373470520228528, + "loss": 0.5525, + "step": 25422 + }, + { + "epoch": 1.7641385053084448, + "grad_norm": 1.3094412088394165, + "learning_rate": 0.00012372924710173647, + "loss": 0.606, + "step": 25423 + }, + { + "epoch": 1.7642078967455417, + "grad_norm": 1.1883269548416138, + "learning_rate": 0.00012372378892627718, + "loss": 0.6662, + "step": 25424 + }, + { + "epoch": 1.7642772881826383, + "grad_norm": 1.2234992980957031, + "learning_rate": 0.00012371833067592461, + "loss": 0.5097, + "step": 25425 + }, + { + "epoch": 1.764346679619735, + "grad_norm": 1.1899927854537964, + "learning_rate": 0.00012371287235069607, + "loss": 0.3731, + "step": 25426 + }, + { + "epoch": 1.7644160710568317, + "grad_norm": 1.995797872543335, + "learning_rate": 0.0001237074139506087, + "loss": 0.7148, + "step": 25427 + }, + { + "epoch": 1.7644854624939281, + "grad_norm": 1.401867151260376, + "learning_rate": 0.0001237019554756798, + "loss": 0.5258, + "step": 25428 + }, + { + "epoch": 1.764554853931025, + "grad_norm": 1.5830583572387695, + "learning_rate": 0.00012369649692592658, + "loss": 0.5751, + "step": 25429 + }, + { + "epoch": 1.7646242453681216, + "grad_norm": 1.9489378929138184, + "learning_rate": 0.00012369103830136625, + "loss": 0.7317, + "step": 25430 + }, + { + "epoch": 1.7646936368052182, + "grad_norm": 1.8765801191329956, + "learning_rate": 0.00012368557960201605, + "loss": 0.7265, + "step": 25431 + }, + { + "epoch": 1.764763028242315, + "grad_norm": 1.4903411865234375, + "learning_rate": 0.00012368012082789322, + "loss": 0.6091, + "step": 25432 + }, + { + "epoch": 1.7648324196794114, + "grad_norm": 1.4850516319274902, + "learning_rate": 0.00012367466197901503, + "loss": 0.4598, + "step": 25433 + }, + { + "epoch": 1.7649018111165082, + "grad_norm": 3.1649975776672363, + "learning_rate": 0.00012366920305539867, + "loss": 0.5341, + "step": 25434 + }, + { + "epoch": 1.7649712025536048, + "grad_norm": 1.764390468597412, + "learning_rate": 0.00012366374405706136, + "loss": 0.5434, + "step": 25435 + }, + { + "epoch": 1.7650405939907015, + "grad_norm": 1.4684524536132812, + "learning_rate": 0.0001236582849840204, + "loss": 0.5601, + "step": 25436 + }, + { + "epoch": 1.7651099854277983, + "grad_norm": 1.21327805519104, + "learning_rate": 0.0001236528258362929, + "loss": 0.4525, + "step": 25437 + }, + { + "epoch": 1.765179376864895, + "grad_norm": 2.895200729370117, + "learning_rate": 0.00012364736661389625, + "loss": 0.642, + "step": 25438 + }, + { + "epoch": 1.7652487683019915, + "grad_norm": 1.1503663063049316, + "learning_rate": 0.00012364190731684757, + "loss": 0.4667, + "step": 25439 + }, + { + "epoch": 1.7653181597390883, + "grad_norm": 2.129603147506714, + "learning_rate": 0.00012363644794516416, + "loss": 0.4522, + "step": 25440 + }, + { + "epoch": 1.7653875511761847, + "grad_norm": 1.163205623626709, + "learning_rate": 0.0001236309884988632, + "loss": 0.517, + "step": 25441 + }, + { + "epoch": 1.7654569426132816, + "grad_norm": 1.782641887664795, + "learning_rate": 0.00012362552897796195, + "loss": 0.5624, + "step": 25442 + }, + { + "epoch": 1.7655263340503782, + "grad_norm": 1.8745126724243164, + "learning_rate": 0.00012362006938247767, + "loss": 0.6393, + "step": 25443 + }, + { + "epoch": 1.7655957254874748, + "grad_norm": 1.4343469142913818, + "learning_rate": 0.00012361460971242755, + "loss": 0.4508, + "step": 25444 + }, + { + "epoch": 1.7656651169245716, + "grad_norm": 1.8081092834472656, + "learning_rate": 0.00012360914996782887, + "loss": 0.7753, + "step": 25445 + }, + { + "epoch": 1.765734508361668, + "grad_norm": 2.049093008041382, + "learning_rate": 0.0001236036901486988, + "loss": 0.7423, + "step": 25446 + }, + { + "epoch": 1.7658038997987648, + "grad_norm": 1.5778515338897705, + "learning_rate": 0.00012359823025505465, + "loss": 0.5899, + "step": 25447 + }, + { + "epoch": 1.7658732912358615, + "grad_norm": 1.4137026071548462, + "learning_rate": 0.00012359277028691363, + "loss": 0.4035, + "step": 25448 + }, + { + "epoch": 1.765942682672958, + "grad_norm": 2.266751289367676, + "learning_rate": 0.00012358731024429296, + "loss": 0.4581, + "step": 25449 + }, + { + "epoch": 1.766012074110055, + "grad_norm": 1.4041491746902466, + "learning_rate": 0.00012358185012720992, + "loss": 0.5348, + "step": 25450 + }, + { + "epoch": 1.7660814655471515, + "grad_norm": 1.4009733200073242, + "learning_rate": 0.00012357638993568167, + "loss": 0.3794, + "step": 25451 + }, + { + "epoch": 1.7661508569842481, + "grad_norm": 2.3715248107910156, + "learning_rate": 0.00012357092966972546, + "loss": 0.6478, + "step": 25452 + }, + { + "epoch": 1.766220248421345, + "grad_norm": 1.709430456161499, + "learning_rate": 0.00012356546932935862, + "loss": 0.6977, + "step": 25453 + }, + { + "epoch": 1.7662896398584413, + "grad_norm": 1.2034108638763428, + "learning_rate": 0.00012356000891459832, + "loss": 0.4597, + "step": 25454 + }, + { + "epoch": 1.7663590312955382, + "grad_norm": 1.5803250074386597, + "learning_rate": 0.0001235545484254618, + "loss": 0.4711, + "step": 25455 + }, + { + "epoch": 1.7664284227326348, + "grad_norm": 1.530903935432434, + "learning_rate": 0.0001235490878619663, + "loss": 0.4904, + "step": 25456 + }, + { + "epoch": 1.7664978141697314, + "grad_norm": 2.4025824069976807, + "learning_rate": 0.0001235436272241291, + "loss": 0.711, + "step": 25457 + }, + { + "epoch": 1.7665672056068282, + "grad_norm": 1.841667890548706, + "learning_rate": 0.00012353816651196735, + "loss": 0.653, + "step": 25458 + }, + { + "epoch": 1.7666365970439246, + "grad_norm": 1.7603741884231567, + "learning_rate": 0.00012353270572549837, + "loss": 0.5241, + "step": 25459 + }, + { + "epoch": 1.7667059884810214, + "grad_norm": 1.469537377357483, + "learning_rate": 0.0001235272448647393, + "loss": 0.6258, + "step": 25460 + }, + { + "epoch": 1.766775379918118, + "grad_norm": 2.48980450630188, + "learning_rate": 0.00012352178392970755, + "loss": 0.5799, + "step": 25461 + }, + { + "epoch": 1.7668447713552147, + "grad_norm": 1.331573486328125, + "learning_rate": 0.0001235163229204202, + "loss": 0.428, + "step": 25462 + }, + { + "epoch": 1.7669141627923115, + "grad_norm": 1.8276923894882202, + "learning_rate": 0.00012351086183689456, + "loss": 0.3536, + "step": 25463 + }, + { + "epoch": 1.7669835542294081, + "grad_norm": 1.9502315521240234, + "learning_rate": 0.00012350540067914784, + "loss": 0.4764, + "step": 25464 + }, + { + "epoch": 1.7670529456665047, + "grad_norm": 2.070615768432617, + "learning_rate": 0.0001234999394471973, + "loss": 0.8236, + "step": 25465 + }, + { + "epoch": 1.7671223371036016, + "grad_norm": 1.578588604927063, + "learning_rate": 0.0001234944781410602, + "loss": 0.6193, + "step": 25466 + }, + { + "epoch": 1.767191728540698, + "grad_norm": 2.251178503036499, + "learning_rate": 0.00012348901676075374, + "loss": 0.4306, + "step": 25467 + }, + { + "epoch": 1.7672611199777948, + "grad_norm": 1.5819398164749146, + "learning_rate": 0.0001234835553062952, + "loss": 0.5714, + "step": 25468 + }, + { + "epoch": 1.7673305114148914, + "grad_norm": 1.7338366508483887, + "learning_rate": 0.00012347809377770178, + "loss": 0.6539, + "step": 25469 + }, + { + "epoch": 1.767399902851988, + "grad_norm": 1.5810651779174805, + "learning_rate": 0.00012347263217499073, + "loss": 0.6032, + "step": 25470 + }, + { + "epoch": 1.7674692942890848, + "grad_norm": 1.8552587032318115, + "learning_rate": 0.00012346717049817934, + "loss": 0.5257, + "step": 25471 + }, + { + "epoch": 1.7675386857261812, + "grad_norm": 1.9700227975845337, + "learning_rate": 0.00012346170874728477, + "loss": 0.536, + "step": 25472 + }, + { + "epoch": 1.767608077163278, + "grad_norm": 1.4155982732772827, + "learning_rate": 0.00012345624692232434, + "loss": 0.8019, + "step": 25473 + }, + { + "epoch": 1.7676774686003747, + "grad_norm": 1.7286839485168457, + "learning_rate": 0.00012345078502331525, + "loss": 0.7372, + "step": 25474 + }, + { + "epoch": 1.7677468600374713, + "grad_norm": 1.968288779258728, + "learning_rate": 0.00012344532305027476, + "loss": 0.6306, + "step": 25475 + }, + { + "epoch": 1.7678162514745681, + "grad_norm": 1.603670358657837, + "learning_rate": 0.00012343986100322012, + "loss": 0.6881, + "step": 25476 + }, + { + "epoch": 1.7678856429116647, + "grad_norm": 1.5424108505249023, + "learning_rate": 0.00012343439888216852, + "loss": 0.6597, + "step": 25477 + }, + { + "epoch": 1.7679550343487613, + "grad_norm": 1.9473305940628052, + "learning_rate": 0.00012342893668713728, + "loss": 0.532, + "step": 25478 + }, + { + "epoch": 1.7680244257858582, + "grad_norm": 1.2532761096954346, + "learning_rate": 0.00012342347441814358, + "loss": 0.3329, + "step": 25479 + }, + { + "epoch": 1.7680938172229546, + "grad_norm": 2.197307825088501, + "learning_rate": 0.0001234180120752047, + "loss": 0.6644, + "step": 25480 + }, + { + "epoch": 1.7681632086600514, + "grad_norm": 1.6689423322677612, + "learning_rate": 0.00012341254965833785, + "loss": 0.5182, + "step": 25481 + }, + { + "epoch": 1.768232600097148, + "grad_norm": 1.312035083770752, + "learning_rate": 0.00012340708716756033, + "loss": 0.7969, + "step": 25482 + }, + { + "epoch": 1.7683019915342446, + "grad_norm": 1.822278618812561, + "learning_rate": 0.00012340162460288933, + "loss": 0.3817, + "step": 25483 + }, + { + "epoch": 1.7683713829713414, + "grad_norm": 4.766312122344971, + "learning_rate": 0.00012339616196434212, + "loss": 0.6739, + "step": 25484 + }, + { + "epoch": 1.7684407744084378, + "grad_norm": 1.8794219493865967, + "learning_rate": 0.00012339069925193595, + "loss": 0.4591, + "step": 25485 + }, + { + "epoch": 1.7685101658455347, + "grad_norm": 2.185606002807617, + "learning_rate": 0.00012338523646568804, + "loss": 0.647, + "step": 25486 + }, + { + "epoch": 1.7685795572826313, + "grad_norm": 2.200657844543457, + "learning_rate": 0.00012337977360561567, + "loss": 0.5653, + "step": 25487 + }, + { + "epoch": 1.7686489487197279, + "grad_norm": 1.428438425064087, + "learning_rate": 0.00012337431067173605, + "loss": 0.5398, + "step": 25488 + }, + { + "epoch": 1.7687183401568247, + "grad_norm": 1.8048067092895508, + "learning_rate": 0.00012336884766406644, + "loss": 0.6993, + "step": 25489 + }, + { + "epoch": 1.7687877315939213, + "grad_norm": 1.4136528968811035, + "learning_rate": 0.0001233633845826241, + "loss": 0.4238, + "step": 25490 + }, + { + "epoch": 1.768857123031018, + "grad_norm": 1.7860939502716064, + "learning_rate": 0.00012335792142742627, + "loss": 0.5324, + "step": 25491 + }, + { + "epoch": 1.7689265144681148, + "grad_norm": 1.1960842609405518, + "learning_rate": 0.00012335245819849018, + "loss": 0.3196, + "step": 25492 + }, + { + "epoch": 1.7689959059052112, + "grad_norm": 1.6187869310379028, + "learning_rate": 0.0001233469948958331, + "loss": 0.4839, + "step": 25493 + }, + { + "epoch": 1.769065297342308, + "grad_norm": 1.735094666481018, + "learning_rate": 0.00012334153151947227, + "loss": 0.7544, + "step": 25494 + }, + { + "epoch": 1.7691346887794046, + "grad_norm": 1.6028156280517578, + "learning_rate": 0.00012333606806942495, + "loss": 0.603, + "step": 25495 + }, + { + "epoch": 1.7692040802165012, + "grad_norm": 1.7447447776794434, + "learning_rate": 0.00012333060454570836, + "loss": 0.4214, + "step": 25496 + }, + { + "epoch": 1.769273471653598, + "grad_norm": 1.4310458898544312, + "learning_rate": 0.00012332514094833973, + "loss": 0.4416, + "step": 25497 + }, + { + "epoch": 1.7693428630906947, + "grad_norm": 1.5338201522827148, + "learning_rate": 0.00012331967727733636, + "loss": 0.3499, + "step": 25498 + }, + { + "epoch": 1.7694122545277913, + "grad_norm": 1.899861454963684, + "learning_rate": 0.00012331421353271545, + "loss": 0.4654, + "step": 25499 + }, + { + "epoch": 1.7694816459648879, + "grad_norm": 1.2805243730545044, + "learning_rate": 0.0001233087497144943, + "loss": 0.3915, + "step": 25500 + }, + { + "epoch": 1.7695510374019845, + "grad_norm": 3.7854318618774414, + "learning_rate": 0.0001233032858226901, + "loss": 0.9986, + "step": 25501 + }, + { + "epoch": 1.7696204288390813, + "grad_norm": 2.3342092037200928, + "learning_rate": 0.00012329782185732018, + "loss": 0.5182, + "step": 25502 + }, + { + "epoch": 1.769689820276178, + "grad_norm": 1.9709573984146118, + "learning_rate": 0.0001232923578184017, + "loss": 0.5267, + "step": 25503 + }, + { + "epoch": 1.7697592117132745, + "grad_norm": 2.0799026489257812, + "learning_rate": 0.00012328689370595195, + "loss": 0.5767, + "step": 25504 + }, + { + "epoch": 1.7698286031503714, + "grad_norm": 1.7733794450759888, + "learning_rate": 0.0001232814295199882, + "loss": 0.66, + "step": 25505 + }, + { + "epoch": 1.7698979945874678, + "grad_norm": 1.687516212463379, + "learning_rate": 0.00012327596526052765, + "loss": 0.6425, + "step": 25506 + }, + { + "epoch": 1.7699673860245646, + "grad_norm": 2.529205560684204, + "learning_rate": 0.00012327050092758758, + "loss": 0.6789, + "step": 25507 + }, + { + "epoch": 1.7700367774616612, + "grad_norm": 2.268451690673828, + "learning_rate": 0.00012326503652118525, + "loss": 0.6208, + "step": 25508 + }, + { + "epoch": 1.7701061688987578, + "grad_norm": 2.0665199756622314, + "learning_rate": 0.0001232595720413379, + "loss": 0.6755, + "step": 25509 + }, + { + "epoch": 1.7701755603358547, + "grad_norm": 0.8403061032295227, + "learning_rate": 0.0001232541074880628, + "loss": 0.3786, + "step": 25510 + }, + { + "epoch": 1.7702449517729513, + "grad_norm": 1.6797351837158203, + "learning_rate": 0.00012324864286137715, + "loss": 0.4773, + "step": 25511 + }, + { + "epoch": 1.7703143432100479, + "grad_norm": 1.684373378753662, + "learning_rate": 0.00012324317816129825, + "loss": 0.5877, + "step": 25512 + }, + { + "epoch": 1.7703837346471445, + "grad_norm": 1.4526296854019165, + "learning_rate": 0.00012323771338784335, + "loss": 0.491, + "step": 25513 + }, + { + "epoch": 1.770453126084241, + "grad_norm": 1.532791256904602, + "learning_rate": 0.00012323224854102966, + "loss": 0.5474, + "step": 25514 + }, + { + "epoch": 1.770522517521338, + "grad_norm": 1.8577866554260254, + "learning_rate": 0.00012322678362087442, + "loss": 0.6472, + "step": 25515 + }, + { + "epoch": 1.7705919089584345, + "grad_norm": 2.6921579837799072, + "learning_rate": 0.000123221318627395, + "loss": 0.8851, + "step": 25516 + }, + { + "epoch": 1.7706613003955312, + "grad_norm": 2.0839364528656006, + "learning_rate": 0.0001232158535606085, + "loss": 0.6486, + "step": 25517 + }, + { + "epoch": 1.770730691832628, + "grad_norm": 1.4541687965393066, + "learning_rate": 0.00012321038842053226, + "loss": 0.646, + "step": 25518 + }, + { + "epoch": 1.7708000832697244, + "grad_norm": 2.2330052852630615, + "learning_rate": 0.00012320492320718355, + "loss": 0.6296, + "step": 25519 + }, + { + "epoch": 1.7708694747068212, + "grad_norm": 1.4337778091430664, + "learning_rate": 0.00012319945792057957, + "loss": 0.486, + "step": 25520 + }, + { + "epoch": 1.7709388661439178, + "grad_norm": 2.256047010421753, + "learning_rate": 0.00012319399256073758, + "loss": 0.6082, + "step": 25521 + }, + { + "epoch": 1.7710082575810144, + "grad_norm": 1.5231400728225708, + "learning_rate": 0.00012318852712767486, + "loss": 0.5378, + "step": 25522 + }, + { + "epoch": 1.7710776490181113, + "grad_norm": 1.8528268337249756, + "learning_rate": 0.00012318306162140868, + "loss": 0.5025, + "step": 25523 + }, + { + "epoch": 1.7711470404552079, + "grad_norm": 1.1713008880615234, + "learning_rate": 0.0001231775960419562, + "loss": 0.5095, + "step": 25524 + }, + { + "epoch": 1.7712164318923045, + "grad_norm": 1.3308377265930176, + "learning_rate": 0.00012317213038933478, + "loss": 0.4337, + "step": 25525 + }, + { + "epoch": 1.771285823329401, + "grad_norm": 3.3527557849884033, + "learning_rate": 0.00012316666466356165, + "loss": 0.4925, + "step": 25526 + }, + { + "epoch": 1.7713552147664977, + "grad_norm": 1.4889205694198608, + "learning_rate": 0.000123161198864654, + "loss": 0.5114, + "step": 25527 + }, + { + "epoch": 1.7714246062035945, + "grad_norm": 1.3453766107559204, + "learning_rate": 0.00012315573299262917, + "loss": 0.5137, + "step": 25528 + }, + { + "epoch": 1.7714939976406912, + "grad_norm": 1.5279699563980103, + "learning_rate": 0.00012315026704750435, + "loss": 0.8072, + "step": 25529 + }, + { + "epoch": 1.7715633890777878, + "grad_norm": 1.2967355251312256, + "learning_rate": 0.00012314480102929685, + "loss": 0.4867, + "step": 25530 + }, + { + "epoch": 1.7716327805148846, + "grad_norm": 1.7357097864151, + "learning_rate": 0.00012313933493802388, + "loss": 0.5877, + "step": 25531 + }, + { + "epoch": 1.771702171951981, + "grad_norm": 1.605130672454834, + "learning_rate": 0.0001231338687737027, + "loss": 0.6303, + "step": 25532 + }, + { + "epoch": 1.7717715633890778, + "grad_norm": 1.6035372018814087, + "learning_rate": 0.0001231284025363506, + "loss": 0.7203, + "step": 25533 + }, + { + "epoch": 1.7718409548261744, + "grad_norm": 1.8287057876586914, + "learning_rate": 0.00012312293622598482, + "loss": 0.3973, + "step": 25534 + }, + { + "epoch": 1.771910346263271, + "grad_norm": 1.8373650312423706, + "learning_rate": 0.0001231174698426226, + "loss": 0.6761, + "step": 25535 + }, + { + "epoch": 1.7719797377003679, + "grad_norm": 1.2877060174942017, + "learning_rate": 0.00012311200338628123, + "loss": 0.491, + "step": 25536 + }, + { + "epoch": 1.7720491291374645, + "grad_norm": 2.478976011276245, + "learning_rate": 0.00012310653685697792, + "loss": 0.7803, + "step": 25537 + }, + { + "epoch": 1.772118520574561, + "grad_norm": 1.6232273578643799, + "learning_rate": 0.00012310107025472996, + "loss": 0.5287, + "step": 25538 + }, + { + "epoch": 1.7721879120116577, + "grad_norm": 1.9055233001708984, + "learning_rate": 0.0001230956035795546, + "loss": 0.7203, + "step": 25539 + }, + { + "epoch": 1.7722573034487543, + "grad_norm": 1.5885143280029297, + "learning_rate": 0.00012309013683146913, + "loss": 0.6903, + "step": 25540 + }, + { + "epoch": 1.7723266948858512, + "grad_norm": 1.6375060081481934, + "learning_rate": 0.00012308467001049072, + "loss": 0.6067, + "step": 25541 + }, + { + "epoch": 1.7723960863229478, + "grad_norm": 1.3968664407730103, + "learning_rate": 0.00012307920311663674, + "loss": 0.5211, + "step": 25542 + }, + { + "epoch": 1.7724654777600444, + "grad_norm": 1.731865644454956, + "learning_rate": 0.0001230737361499244, + "loss": 0.4514, + "step": 25543 + }, + { + "epoch": 1.7725348691971412, + "grad_norm": 1.6499390602111816, + "learning_rate": 0.0001230682691103709, + "loss": 0.7985, + "step": 25544 + }, + { + "epoch": 1.7726042606342376, + "grad_norm": 1.0432729721069336, + "learning_rate": 0.00012306280199799355, + "loss": 0.4843, + "step": 25545 + }, + { + "epoch": 1.7726736520713344, + "grad_norm": 1.4543421268463135, + "learning_rate": 0.00012305733481280968, + "loss": 0.5585, + "step": 25546 + }, + { + "epoch": 1.772743043508431, + "grad_norm": 1.4067682027816772, + "learning_rate": 0.00012305186755483643, + "loss": 0.4915, + "step": 25547 + }, + { + "epoch": 1.7728124349455276, + "grad_norm": 2.3803937435150146, + "learning_rate": 0.00012304640022409112, + "loss": 0.9499, + "step": 25548 + }, + { + "epoch": 1.7728818263826245, + "grad_norm": 1.8222546577453613, + "learning_rate": 0.00012304093282059097, + "loss": 0.559, + "step": 25549 + }, + { + "epoch": 1.772951217819721, + "grad_norm": 2.3821680545806885, + "learning_rate": 0.00012303546534435332, + "loss": 0.7515, + "step": 25550 + }, + { + "epoch": 1.7730206092568177, + "grad_norm": 1.608243465423584, + "learning_rate": 0.00012302999779539536, + "loss": 0.5471, + "step": 25551 + }, + { + "epoch": 1.7730900006939145, + "grad_norm": 1.5184310674667358, + "learning_rate": 0.00012302453017373432, + "loss": 0.7331, + "step": 25552 + }, + { + "epoch": 1.773159392131011, + "grad_norm": 1.1693097352981567, + "learning_rate": 0.00012301906247938756, + "loss": 0.472, + "step": 25553 + }, + { + "epoch": 1.7732287835681078, + "grad_norm": 1.423690676689148, + "learning_rate": 0.00012301359471237228, + "loss": 0.7736, + "step": 25554 + }, + { + "epoch": 1.7732981750052044, + "grad_norm": 1.5428763628005981, + "learning_rate": 0.00012300812687270573, + "loss": 0.6003, + "step": 25555 + }, + { + "epoch": 1.773367566442301, + "grad_norm": 1.7751634120941162, + "learning_rate": 0.00012300265896040522, + "loss": 0.4176, + "step": 25556 + }, + { + "epoch": 1.7734369578793978, + "grad_norm": 1.7915184497833252, + "learning_rate": 0.000122997190975488, + "loss": 0.5269, + "step": 25557 + }, + { + "epoch": 1.7735063493164942, + "grad_norm": 1.8020435571670532, + "learning_rate": 0.00012299172291797128, + "loss": 0.5579, + "step": 25558 + }, + { + "epoch": 1.773575740753591, + "grad_norm": 1.6952227354049683, + "learning_rate": 0.00012298625478787236, + "loss": 0.7215, + "step": 25559 + }, + { + "epoch": 1.7736451321906876, + "grad_norm": 1.0982120037078857, + "learning_rate": 0.0001229807865852085, + "loss": 0.4144, + "step": 25560 + }, + { + "epoch": 1.7737145236277843, + "grad_norm": 1.8081650733947754, + "learning_rate": 0.000122975318309997, + "loss": 0.745, + "step": 25561 + }, + { + "epoch": 1.773783915064881, + "grad_norm": 1.1907379627227783, + "learning_rate": 0.00012296984996225505, + "loss": 0.5762, + "step": 25562 + }, + { + "epoch": 1.7738533065019777, + "grad_norm": 1.447100281715393, + "learning_rate": 0.00012296438154199996, + "loss": 0.6407, + "step": 25563 + }, + { + "epoch": 1.7739226979390743, + "grad_norm": 3.3888087272644043, + "learning_rate": 0.00012295891304924897, + "loss": 0.5029, + "step": 25564 + }, + { + "epoch": 1.7739920893761711, + "grad_norm": 2.0182855129241943, + "learning_rate": 0.0001229534444840194, + "loss": 0.7433, + "step": 25565 + }, + { + "epoch": 1.7740614808132675, + "grad_norm": 1.254611849784851, + "learning_rate": 0.00012294797584632843, + "loss": 0.5711, + "step": 25566 + }, + { + "epoch": 1.7741308722503644, + "grad_norm": 1.7259736061096191, + "learning_rate": 0.00012294250713619335, + "loss": 0.5683, + "step": 25567 + }, + { + "epoch": 1.774200263687461, + "grad_norm": 1.2821364402770996, + "learning_rate": 0.00012293703835363146, + "loss": 0.6379, + "step": 25568 + }, + { + "epoch": 1.7742696551245576, + "grad_norm": 1.771788239479065, + "learning_rate": 0.00012293156949866001, + "loss": 0.6067, + "step": 25569 + }, + { + "epoch": 1.7743390465616544, + "grad_norm": 1.9492074251174927, + "learning_rate": 0.00012292610057129623, + "loss": 0.4972, + "step": 25570 + }, + { + "epoch": 1.7744084379987508, + "grad_norm": 2.2161922454833984, + "learning_rate": 0.00012292063157155742, + "loss": 0.4672, + "step": 25571 + }, + { + "epoch": 1.7744778294358476, + "grad_norm": 1.6729233264923096, + "learning_rate": 0.00012291516249946085, + "loss": 0.609, + "step": 25572 + }, + { + "epoch": 1.7745472208729443, + "grad_norm": 1.669848918914795, + "learning_rate": 0.00012290969335502374, + "loss": 0.6486, + "step": 25573 + }, + { + "epoch": 1.7746166123100409, + "grad_norm": 1.3355575799942017, + "learning_rate": 0.0001229042241382634, + "loss": 0.5184, + "step": 25574 + }, + { + "epoch": 1.7746860037471377, + "grad_norm": 1.740831732749939, + "learning_rate": 0.0001228987548491971, + "loss": 0.7531, + "step": 25575 + }, + { + "epoch": 1.7747553951842343, + "grad_norm": 1.57390296459198, + "learning_rate": 0.00012289328548784206, + "loss": 0.4624, + "step": 25576 + }, + { + "epoch": 1.774824786621331, + "grad_norm": 1.6745805740356445, + "learning_rate": 0.00012288781605421557, + "loss": 0.4369, + "step": 25577 + }, + { + "epoch": 1.7748941780584278, + "grad_norm": 1.3377134799957275, + "learning_rate": 0.00012288234654833494, + "loss": 0.5078, + "step": 25578 + }, + { + "epoch": 1.7749635694955241, + "grad_norm": 1.751548171043396, + "learning_rate": 0.00012287687697021737, + "loss": 0.6304, + "step": 25579 + }, + { + "epoch": 1.775032960932621, + "grad_norm": 2.3625576496124268, + "learning_rate": 0.00012287140731988017, + "loss": 0.62, + "step": 25580 + }, + { + "epoch": 1.7751023523697176, + "grad_norm": 1.4484305381774902, + "learning_rate": 0.00012286593759734058, + "loss": 0.5839, + "step": 25581 + }, + { + "epoch": 1.7751717438068142, + "grad_norm": 1.7229509353637695, + "learning_rate": 0.00012286046780261587, + "loss": 0.712, + "step": 25582 + }, + { + "epoch": 1.775241135243911, + "grad_norm": 1.5193023681640625, + "learning_rate": 0.00012285499793572333, + "loss": 0.5958, + "step": 25583 + }, + { + "epoch": 1.7753105266810074, + "grad_norm": 1.2909942865371704, + "learning_rate": 0.0001228495279966802, + "loss": 0.586, + "step": 25584 + }, + { + "epoch": 1.7753799181181042, + "grad_norm": 1.1659305095672607, + "learning_rate": 0.0001228440579855038, + "loss": 0.5977, + "step": 25585 + }, + { + "epoch": 1.7754493095552009, + "grad_norm": 1.806894063949585, + "learning_rate": 0.0001228385879022113, + "loss": 0.6976, + "step": 25586 + }, + { + "epoch": 1.7755187009922975, + "grad_norm": 2.492703676223755, + "learning_rate": 0.00012283311774682006, + "loss": 0.6696, + "step": 25587 + }, + { + "epoch": 1.7755880924293943, + "grad_norm": 1.6044450998306274, + "learning_rate": 0.0001228276475193473, + "loss": 0.6307, + "step": 25588 + }, + { + "epoch": 1.775657483866491, + "grad_norm": 1.6917154788970947, + "learning_rate": 0.0001228221772198103, + "loss": 0.6193, + "step": 25589 + }, + { + "epoch": 1.7757268753035875, + "grad_norm": 1.247231364250183, + "learning_rate": 0.00012281670684822634, + "loss": 0.3106, + "step": 25590 + }, + { + "epoch": 1.7757962667406844, + "grad_norm": 1.3727890253067017, + "learning_rate": 0.0001228112364046127, + "loss": 0.5433, + "step": 25591 + }, + { + "epoch": 1.7758656581777807, + "grad_norm": 1.6096618175506592, + "learning_rate": 0.00012280576588898665, + "loss": 0.4607, + "step": 25592 + }, + { + "epoch": 1.7759350496148776, + "grad_norm": 1.2940309047698975, + "learning_rate": 0.00012280029530136542, + "loss": 0.5232, + "step": 25593 + }, + { + "epoch": 1.7760044410519742, + "grad_norm": 1.9665336608886719, + "learning_rate": 0.0001227948246417663, + "loss": 0.7238, + "step": 25594 + }, + { + "epoch": 1.7760738324890708, + "grad_norm": 1.6239995956420898, + "learning_rate": 0.00012278935391020655, + "loss": 0.5388, + "step": 25595 + }, + { + "epoch": 1.7761432239261676, + "grad_norm": 2.0242114067077637, + "learning_rate": 0.00012278388310670347, + "loss": 0.6817, + "step": 25596 + }, + { + "epoch": 1.776212615363264, + "grad_norm": 2.359574556350708, + "learning_rate": 0.0001227784122312743, + "loss": 0.6222, + "step": 25597 + }, + { + "epoch": 1.7762820068003609, + "grad_norm": 1.4098814725875854, + "learning_rate": 0.00012277294128393637, + "loss": 0.369, + "step": 25598 + }, + { + "epoch": 1.7763513982374575, + "grad_norm": 1.3147987127304077, + "learning_rate": 0.00012276747026470684, + "loss": 0.4579, + "step": 25599 + }, + { + "epoch": 1.776420789674554, + "grad_norm": 1.222232460975647, + "learning_rate": 0.0001227619991736031, + "loss": 0.588, + "step": 25600 + }, + { + "epoch": 1.776490181111651, + "grad_norm": 1.165778398513794, + "learning_rate": 0.00012275652801064236, + "loss": 0.4892, + "step": 25601 + }, + { + "epoch": 1.7765595725487475, + "grad_norm": 1.8265926837921143, + "learning_rate": 0.0001227510567758419, + "loss": 0.5258, + "step": 25602 + }, + { + "epoch": 1.7766289639858441, + "grad_norm": 1.5207539796829224, + "learning_rate": 0.00012274558546921898, + "loss": 0.7825, + "step": 25603 + }, + { + "epoch": 1.776698355422941, + "grad_norm": 1.350525975227356, + "learning_rate": 0.00012274011409079089, + "loss": 0.5607, + "step": 25604 + }, + { + "epoch": 1.7767677468600374, + "grad_norm": 1.3974626064300537, + "learning_rate": 0.0001227346426405749, + "loss": 0.5322, + "step": 25605 + }, + { + "epoch": 1.7768371382971342, + "grad_norm": 1.5694334506988525, + "learning_rate": 0.00012272917111858832, + "loss": 0.5116, + "step": 25606 + }, + { + "epoch": 1.7769065297342308, + "grad_norm": 1.9409072399139404, + "learning_rate": 0.00012272369952484832, + "loss": 0.5237, + "step": 25607 + }, + { + "epoch": 1.7769759211713274, + "grad_norm": 2.5587759017944336, + "learning_rate": 0.00012271822785937225, + "loss": 0.4377, + "step": 25608 + }, + { + "epoch": 1.7770453126084242, + "grad_norm": 1.781933069229126, + "learning_rate": 0.0001227127561221774, + "loss": 0.698, + "step": 25609 + }, + { + "epoch": 1.7771147040455206, + "grad_norm": 1.189469337463379, + "learning_rate": 0.000122707284313281, + "loss": 0.4106, + "step": 25610 + }, + { + "epoch": 1.7771840954826175, + "grad_norm": 1.3284876346588135, + "learning_rate": 0.00012270181243270034, + "loss": 0.5834, + "step": 25611 + }, + { + "epoch": 1.777253486919714, + "grad_norm": 1.6969877481460571, + "learning_rate": 0.0001226963404804527, + "loss": 0.4441, + "step": 25612 + }, + { + "epoch": 1.7773228783568107, + "grad_norm": 1.3922539949417114, + "learning_rate": 0.00012269086845655532, + "loss": 0.4547, + "step": 25613 + }, + { + "epoch": 1.7773922697939075, + "grad_norm": 1.695383906364441, + "learning_rate": 0.00012268539636102553, + "loss": 0.5056, + "step": 25614 + }, + { + "epoch": 1.7774616612310041, + "grad_norm": 1.9714726209640503, + "learning_rate": 0.00012267992419388052, + "loss": 0.7686, + "step": 25615 + }, + { + "epoch": 1.7775310526681007, + "grad_norm": 2.327592611312866, + "learning_rate": 0.0001226744519551377, + "loss": 0.5362, + "step": 25616 + }, + { + "epoch": 1.7776004441051976, + "grad_norm": 1.089847207069397, + "learning_rate": 0.00012266897964481423, + "loss": 0.4297, + "step": 25617 + }, + { + "epoch": 1.777669835542294, + "grad_norm": 1.8255854845046997, + "learning_rate": 0.00012266350726292742, + "loss": 0.669, + "step": 25618 + }, + { + "epoch": 1.7777392269793908, + "grad_norm": 1.2073308229446411, + "learning_rate": 0.00012265803480949455, + "loss": 0.4718, + "step": 25619 + }, + { + "epoch": 1.7778086184164874, + "grad_norm": 1.7898916006088257, + "learning_rate": 0.00012265256228453288, + "loss": 0.8855, + "step": 25620 + }, + { + "epoch": 1.777878009853584, + "grad_norm": 1.4630717039108276, + "learning_rate": 0.00012264708968805972, + "loss": 0.6347, + "step": 25621 + }, + { + "epoch": 1.7779474012906809, + "grad_norm": 1.274287223815918, + "learning_rate": 0.0001226416170200923, + "loss": 0.4798, + "step": 25622 + }, + { + "epoch": 1.7780167927277772, + "grad_norm": 1.4420958757400513, + "learning_rate": 0.00012263614428064794, + "loss": 0.5789, + "step": 25623 + }, + { + "epoch": 1.778086184164874, + "grad_norm": 1.7304179668426514, + "learning_rate": 0.00012263067146974388, + "loss": 0.7287, + "step": 25624 + }, + { + "epoch": 1.7781555756019707, + "grad_norm": 1.9668281078338623, + "learning_rate": 0.00012262519858739743, + "loss": 0.8454, + "step": 25625 + }, + { + "epoch": 1.7782249670390673, + "grad_norm": 1.682045340538025, + "learning_rate": 0.00012261972563362586, + "loss": 0.5811, + "step": 25626 + }, + { + "epoch": 1.7782943584761641, + "grad_norm": 1.9085267782211304, + "learning_rate": 0.00012261425260844642, + "loss": 0.5684, + "step": 25627 + }, + { + "epoch": 1.7783637499132607, + "grad_norm": 1.3226020336151123, + "learning_rate": 0.00012260877951187645, + "loss": 0.3574, + "step": 25628 + }, + { + "epoch": 1.7784331413503573, + "grad_norm": 2.107506036758423, + "learning_rate": 0.00012260330634393312, + "loss": 0.5998, + "step": 25629 + }, + { + "epoch": 1.7785025327874542, + "grad_norm": 1.5517456531524658, + "learning_rate": 0.00012259783310463384, + "loss": 0.6972, + "step": 25630 + }, + { + "epoch": 1.7785719242245506, + "grad_norm": 2.062941789627075, + "learning_rate": 0.0001225923597939958, + "loss": 0.6542, + "step": 25631 + }, + { + "epoch": 1.7786413156616474, + "grad_norm": 1.3823870420455933, + "learning_rate": 0.00012258688641203627, + "loss": 0.5984, + "step": 25632 + }, + { + "epoch": 1.778710707098744, + "grad_norm": 2.1182289123535156, + "learning_rate": 0.00012258141295877262, + "loss": 0.6337, + "step": 25633 + }, + { + "epoch": 1.7787800985358406, + "grad_norm": 1.530236005783081, + "learning_rate": 0.00012257593943422202, + "loss": 0.4104, + "step": 25634 + }, + { + "epoch": 1.7788494899729375, + "grad_norm": 2.25435471534729, + "learning_rate": 0.0001225704658384018, + "loss": 0.7832, + "step": 25635 + }, + { + "epoch": 1.7789188814100338, + "grad_norm": 1.3875826597213745, + "learning_rate": 0.00012256499217132925, + "loss": 0.4578, + "step": 25636 + }, + { + "epoch": 1.7789882728471307, + "grad_norm": 1.6312180757522583, + "learning_rate": 0.00012255951843302167, + "loss": 0.5861, + "step": 25637 + }, + { + "epoch": 1.7790576642842273, + "grad_norm": 1.6378291845321655, + "learning_rate": 0.00012255404462349626, + "loss": 0.69, + "step": 25638 + }, + { + "epoch": 1.779127055721324, + "grad_norm": 1.2639257907867432, + "learning_rate": 0.00012254857074277037, + "loss": 0.502, + "step": 25639 + }, + { + "epoch": 1.7791964471584207, + "grad_norm": 1.0938663482666016, + "learning_rate": 0.00012254309679086126, + "loss": 0.4624, + "step": 25640 + }, + { + "epoch": 1.7792658385955173, + "grad_norm": 1.936745524406433, + "learning_rate": 0.0001225376227677862, + "loss": 0.4568, + "step": 25641 + }, + { + "epoch": 1.779335230032614, + "grad_norm": 1.5890456438064575, + "learning_rate": 0.0001225321486735625, + "loss": 0.6073, + "step": 25642 + }, + { + "epoch": 1.7794046214697108, + "grad_norm": 1.407665729522705, + "learning_rate": 0.0001225266745082074, + "loss": 0.4531, + "step": 25643 + }, + { + "epoch": 1.7794740129068072, + "grad_norm": 2.133976936340332, + "learning_rate": 0.0001225212002717382, + "loss": 0.6719, + "step": 25644 + }, + { + "epoch": 1.779543404343904, + "grad_norm": 2.109982490539551, + "learning_rate": 0.00012251572596417217, + "loss": 0.6184, + "step": 25645 + }, + { + "epoch": 1.7796127957810006, + "grad_norm": 1.6504614353179932, + "learning_rate": 0.00012251025158552663, + "loss": 0.5479, + "step": 25646 + }, + { + "epoch": 1.7796821872180972, + "grad_norm": 1.7652161121368408, + "learning_rate": 0.00012250477713581884, + "loss": 0.664, + "step": 25647 + }, + { + "epoch": 1.779751578655194, + "grad_norm": 1.8221017122268677, + "learning_rate": 0.00012249930261506607, + "loss": 0.6221, + "step": 25648 + }, + { + "epoch": 1.7798209700922905, + "grad_norm": 1.600188970565796, + "learning_rate": 0.0001224938280232856, + "loss": 0.5273, + "step": 25649 + }, + { + "epoch": 1.7798903615293873, + "grad_norm": 2.1707210540771484, + "learning_rate": 0.00012248835336049475, + "loss": 0.5996, + "step": 25650 + }, + { + "epoch": 1.779959752966484, + "grad_norm": 1.4801363945007324, + "learning_rate": 0.00012248287862671079, + "loss": 0.5657, + "step": 25651 + }, + { + "epoch": 1.7800291444035805, + "grad_norm": 1.8219435214996338, + "learning_rate": 0.00012247740382195096, + "loss": 0.7628, + "step": 25652 + }, + { + "epoch": 1.7800985358406773, + "grad_norm": 1.5709788799285889, + "learning_rate": 0.00012247192894623257, + "loss": 0.5321, + "step": 25653 + }, + { + "epoch": 1.780167927277774, + "grad_norm": 1.473758578300476, + "learning_rate": 0.00012246645399957294, + "loss": 0.5655, + "step": 25654 + }, + { + "epoch": 1.7802373187148706, + "grad_norm": 1.8917934894561768, + "learning_rate": 0.0001224609789819893, + "loss": 0.5404, + "step": 25655 + }, + { + "epoch": 1.7803067101519674, + "grad_norm": 1.7605350017547607, + "learning_rate": 0.00012245550389349894, + "loss": 0.5776, + "step": 25656 + }, + { + "epoch": 1.7803761015890638, + "grad_norm": 1.4917653799057007, + "learning_rate": 0.00012245002873411916, + "loss": 0.4272, + "step": 25657 + }, + { + "epoch": 1.7804454930261606, + "grad_norm": 1.4119274616241455, + "learning_rate": 0.0001224445535038673, + "loss": 0.5964, + "step": 25658 + }, + { + "epoch": 1.7805148844632572, + "grad_norm": 1.756996512413025, + "learning_rate": 0.0001224390782027605, + "loss": 0.9688, + "step": 25659 + }, + { + "epoch": 1.7805842759003538, + "grad_norm": 1.6549277305603027, + "learning_rate": 0.00012243360283081616, + "loss": 0.6161, + "step": 25660 + }, + { + "epoch": 1.7806536673374507, + "grad_norm": 2.7369329929351807, + "learning_rate": 0.0001224281273880516, + "loss": 0.7513, + "step": 25661 + }, + { + "epoch": 1.780723058774547, + "grad_norm": 1.8431994915008545, + "learning_rate": 0.00012242265187448399, + "loss": 0.5737, + "step": 25662 + }, + { + "epoch": 1.780792450211644, + "grad_norm": 1.4469447135925293, + "learning_rate": 0.00012241717629013068, + "loss": 0.52, + "step": 25663 + }, + { + "epoch": 1.7808618416487405, + "grad_norm": 1.4951589107513428, + "learning_rate": 0.00012241170063500894, + "loss": 0.4118, + "step": 25664 + }, + { + "epoch": 1.7809312330858371, + "grad_norm": 1.507014513015747, + "learning_rate": 0.00012240622490913608, + "loss": 0.5816, + "step": 25665 + }, + { + "epoch": 1.781000624522934, + "grad_norm": 2.9065475463867188, + "learning_rate": 0.00012240074911252938, + "loss": 0.8935, + "step": 25666 + }, + { + "epoch": 1.7810700159600306, + "grad_norm": 1.8474704027175903, + "learning_rate": 0.00012239527324520605, + "loss": 0.7205, + "step": 25667 + }, + { + "epoch": 1.7811394073971272, + "grad_norm": 1.4018133878707886, + "learning_rate": 0.00012238979730718353, + "loss": 0.5258, + "step": 25668 + }, + { + "epoch": 1.781208798834224, + "grad_norm": 1.697820782661438, + "learning_rate": 0.00012238432129847894, + "loss": 0.6143, + "step": 25669 + }, + { + "epoch": 1.7812781902713204, + "grad_norm": 1.3400466442108154, + "learning_rate": 0.0001223788452191097, + "loss": 0.5367, + "step": 25670 + }, + { + "epoch": 1.7813475817084172, + "grad_norm": 1.7021217346191406, + "learning_rate": 0.00012237336906909302, + "loss": 0.7591, + "step": 25671 + }, + { + "epoch": 1.7814169731455138, + "grad_norm": 1.9406648874282837, + "learning_rate": 0.00012236789284844618, + "loss": 0.7142, + "step": 25672 + }, + { + "epoch": 1.7814863645826104, + "grad_norm": 1.851243495941162, + "learning_rate": 0.00012236241655718652, + "loss": 0.6787, + "step": 25673 + }, + { + "epoch": 1.7815557560197073, + "grad_norm": 1.2955602407455444, + "learning_rate": 0.00012235694019533132, + "loss": 0.5513, + "step": 25674 + }, + { + "epoch": 1.7816251474568037, + "grad_norm": 1.8635051250457764, + "learning_rate": 0.00012235146376289786, + "loss": 0.6459, + "step": 25675 + }, + { + "epoch": 1.7816945388939005, + "grad_norm": 1.1895731687545776, + "learning_rate": 0.00012234598725990338, + "loss": 0.4153, + "step": 25676 + }, + { + "epoch": 1.7817639303309971, + "grad_norm": 1.6230992078781128, + "learning_rate": 0.00012234051068636524, + "loss": 0.5891, + "step": 25677 + }, + { + "epoch": 1.7818333217680937, + "grad_norm": 1.834468126296997, + "learning_rate": 0.00012233503404230072, + "loss": 0.5668, + "step": 25678 + }, + { + "epoch": 1.7819027132051906, + "grad_norm": 1.8469421863555908, + "learning_rate": 0.00012232955732772703, + "loss": 0.5394, + "step": 25679 + }, + { + "epoch": 1.7819721046422872, + "grad_norm": 1.5944491624832153, + "learning_rate": 0.00012232408054266156, + "loss": 0.6881, + "step": 25680 + }, + { + "epoch": 1.7820414960793838, + "grad_norm": 1.690974473953247, + "learning_rate": 0.00012231860368712154, + "loss": 0.6774, + "step": 25681 + }, + { + "epoch": 1.7821108875164806, + "grad_norm": 1.8262877464294434, + "learning_rate": 0.0001223131267611243, + "loss": 0.532, + "step": 25682 + }, + { + "epoch": 1.782180278953577, + "grad_norm": 1.9330074787139893, + "learning_rate": 0.0001223076497646871, + "loss": 0.5239, + "step": 25683 + }, + { + "epoch": 1.7822496703906738, + "grad_norm": 1.8490078449249268, + "learning_rate": 0.00012230217269782723, + "loss": 0.6775, + "step": 25684 + }, + { + "epoch": 1.7823190618277704, + "grad_norm": 1.5100369453430176, + "learning_rate": 0.00012229669556056198, + "loss": 0.493, + "step": 25685 + }, + { + "epoch": 1.782388453264867, + "grad_norm": 1.9002928733825684, + "learning_rate": 0.00012229121835290866, + "loss": 0.6933, + "step": 25686 + }, + { + "epoch": 1.782457844701964, + "grad_norm": 1.3131942749023438, + "learning_rate": 0.00012228574107488456, + "loss": 0.5442, + "step": 25687 + }, + { + "epoch": 1.7825272361390603, + "grad_norm": 4.381040096282959, + "learning_rate": 0.00012228026372650697, + "loss": 0.6039, + "step": 25688 + }, + { + "epoch": 1.7825966275761571, + "grad_norm": 2.1977009773254395, + "learning_rate": 0.00012227478630779314, + "loss": 0.671, + "step": 25689 + }, + { + "epoch": 1.7826660190132537, + "grad_norm": 1.9897469282150269, + "learning_rate": 0.00012226930881876043, + "loss": 0.6907, + "step": 25690 + }, + { + "epoch": 1.7827354104503503, + "grad_norm": 1.6165200471878052, + "learning_rate": 0.00012226383125942604, + "loss": 0.6102, + "step": 25691 + }, + { + "epoch": 1.7828048018874472, + "grad_norm": 1.3697446584701538, + "learning_rate": 0.00012225835362980738, + "loss": 0.4442, + "step": 25692 + }, + { + "epoch": 1.7828741933245438, + "grad_norm": 1.5286190509796143, + "learning_rate": 0.00012225287592992163, + "loss": 0.458, + "step": 25693 + }, + { + "epoch": 1.7829435847616404, + "grad_norm": 1.5289838314056396, + "learning_rate": 0.00012224739815978617, + "loss": 0.7338, + "step": 25694 + }, + { + "epoch": 1.7830129761987372, + "grad_norm": 2.023855209350586, + "learning_rate": 0.00012224192031941823, + "loss": 0.5517, + "step": 25695 + }, + { + "epoch": 1.7830823676358336, + "grad_norm": 2.3943824768066406, + "learning_rate": 0.00012223644240883514, + "loss": 0.6799, + "step": 25696 + }, + { + "epoch": 1.7831517590729304, + "grad_norm": 1.8786261081695557, + "learning_rate": 0.00012223096442805418, + "loss": 0.7112, + "step": 25697 + }, + { + "epoch": 1.783221150510027, + "grad_norm": 1.9335709810256958, + "learning_rate": 0.00012222548637709263, + "loss": 0.8742, + "step": 25698 + }, + { + "epoch": 1.7832905419471237, + "grad_norm": 1.1243839263916016, + "learning_rate": 0.00012222000825596783, + "loss": 0.3741, + "step": 25699 + }, + { + "epoch": 1.7833599333842205, + "grad_norm": 1.9207956790924072, + "learning_rate": 0.00012221453006469702, + "loss": 0.6176, + "step": 25700 + }, + { + "epoch": 1.7834293248213169, + "grad_norm": 1.6917554140090942, + "learning_rate": 0.00012220905180329749, + "loss": 0.7167, + "step": 25701 + }, + { + "epoch": 1.7834987162584137, + "grad_norm": 1.5241789817810059, + "learning_rate": 0.00012220357347178661, + "loss": 0.482, + "step": 25702 + }, + { + "epoch": 1.7835681076955103, + "grad_norm": 1.7352784872055054, + "learning_rate": 0.00012219809507018158, + "loss": 0.5215, + "step": 25703 + }, + { + "epoch": 1.783637499132607, + "grad_norm": 2.093756914138794, + "learning_rate": 0.0001221926165984998, + "loss": 0.7294, + "step": 25704 + }, + { + "epoch": 1.7837068905697038, + "grad_norm": 1.7979035377502441, + "learning_rate": 0.00012218713805675846, + "loss": 0.6118, + "step": 25705 + }, + { + "epoch": 1.7837762820068004, + "grad_norm": 1.4767842292785645, + "learning_rate": 0.0001221816594449749, + "loss": 0.4692, + "step": 25706 + }, + { + "epoch": 1.783845673443897, + "grad_norm": 1.6415594816207886, + "learning_rate": 0.00012217618076316644, + "loss": 0.5194, + "step": 25707 + }, + { + "epoch": 1.7839150648809938, + "grad_norm": 2.1284353733062744, + "learning_rate": 0.00012217070201135031, + "loss": 0.9154, + "step": 25708 + }, + { + "epoch": 1.7839844563180902, + "grad_norm": 1.5306283235549927, + "learning_rate": 0.0001221652231895439, + "loss": 0.5182, + "step": 25709 + }, + { + "epoch": 1.784053847755187, + "grad_norm": 0.8663111329078674, + "learning_rate": 0.0001221597442977644, + "loss": 0.3661, + "step": 25710 + }, + { + "epoch": 1.7841232391922837, + "grad_norm": 1.3653843402862549, + "learning_rate": 0.00012215426533602916, + "loss": 0.5383, + "step": 25711 + }, + { + "epoch": 1.7841926306293803, + "grad_norm": 1.4548667669296265, + "learning_rate": 0.00012214878630435549, + "loss": 0.4473, + "step": 25712 + }, + { + "epoch": 1.784262022066477, + "grad_norm": 2.143789768218994, + "learning_rate": 0.0001221433072027607, + "loss": 0.7221, + "step": 25713 + }, + { + "epoch": 1.7843314135035735, + "grad_norm": 2.0356392860412598, + "learning_rate": 0.00012213782803126202, + "loss": 0.4986, + "step": 25714 + }, + { + "epoch": 1.7844008049406703, + "grad_norm": 1.5271097421646118, + "learning_rate": 0.00012213234878987678, + "loss": 0.4881, + "step": 25715 + }, + { + "epoch": 1.784470196377767, + "grad_norm": 1.9699755907058716, + "learning_rate": 0.00012212686947862234, + "loss": 0.7082, + "step": 25716 + }, + { + "epoch": 1.7845395878148635, + "grad_norm": 1.6044130325317383, + "learning_rate": 0.00012212139009751586, + "loss": 0.5777, + "step": 25717 + }, + { + "epoch": 1.7846089792519604, + "grad_norm": 1.7579219341278076, + "learning_rate": 0.00012211591064657478, + "loss": 0.4879, + "step": 25718 + }, + { + "epoch": 1.784678370689057, + "grad_norm": 1.9112433195114136, + "learning_rate": 0.0001221104311258163, + "loss": 0.5841, + "step": 25719 + }, + { + "epoch": 1.7847477621261536, + "grad_norm": 1.1525331735610962, + "learning_rate": 0.00012210495153525778, + "loss": 0.6316, + "step": 25720 + }, + { + "epoch": 1.7848171535632504, + "grad_norm": 1.5813021659851074, + "learning_rate": 0.00012209947187491647, + "loss": 0.5948, + "step": 25721 + }, + { + "epoch": 1.7848865450003468, + "grad_norm": 1.5114455223083496, + "learning_rate": 0.00012209399214480973, + "loss": 0.6242, + "step": 25722 + }, + { + "epoch": 1.7849559364374437, + "grad_norm": 1.5856318473815918, + "learning_rate": 0.00012208851234495478, + "loss": 0.7444, + "step": 25723 + }, + { + "epoch": 1.7850253278745403, + "grad_norm": 1.939516305923462, + "learning_rate": 0.00012208303247536898, + "loss": 0.5889, + "step": 25724 + }, + { + "epoch": 1.7850947193116369, + "grad_norm": 1.6683505773544312, + "learning_rate": 0.00012207755253606958, + "loss": 0.6587, + "step": 25725 + }, + { + "epoch": 1.7851641107487337, + "grad_norm": 1.1752984523773193, + "learning_rate": 0.00012207207252707395, + "loss": 0.5836, + "step": 25726 + }, + { + "epoch": 1.7852335021858303, + "grad_norm": 1.6381394863128662, + "learning_rate": 0.00012206659244839933, + "loss": 0.4885, + "step": 25727 + }, + { + "epoch": 1.785302893622927, + "grad_norm": 1.218394160270691, + "learning_rate": 0.00012206111230006303, + "loss": 0.4434, + "step": 25728 + }, + { + "epoch": 1.7853722850600235, + "grad_norm": 1.65910005569458, + "learning_rate": 0.00012205563208208235, + "loss": 0.5407, + "step": 25729 + }, + { + "epoch": 1.7854416764971202, + "grad_norm": 2.0342869758605957, + "learning_rate": 0.00012205015179447463, + "loss": 0.6613, + "step": 25730 + }, + { + "epoch": 1.785511067934217, + "grad_norm": 1.5072163343429565, + "learning_rate": 0.00012204467143725711, + "loss": 0.5233, + "step": 25731 + }, + { + "epoch": 1.7855804593713136, + "grad_norm": 1.8041698932647705, + "learning_rate": 0.00012203919101044711, + "loss": 0.6287, + "step": 25732 + }, + { + "epoch": 1.7856498508084102, + "grad_norm": 2.0314202308654785, + "learning_rate": 0.00012203371051406199, + "loss": 0.541, + "step": 25733 + }, + { + "epoch": 1.785719242245507, + "grad_norm": 2.1324117183685303, + "learning_rate": 0.00012202822994811895, + "loss": 0.6967, + "step": 25734 + }, + { + "epoch": 1.7857886336826034, + "grad_norm": 1.9925904273986816, + "learning_rate": 0.00012202274931263538, + "loss": 0.6345, + "step": 25735 + }, + { + "epoch": 1.7858580251197003, + "grad_norm": 1.6959209442138672, + "learning_rate": 0.00012201726860762852, + "loss": 0.6311, + "step": 25736 + }, + { + "epoch": 1.7859274165567969, + "grad_norm": 1.3096965551376343, + "learning_rate": 0.00012201178783311573, + "loss": 0.522, + "step": 25737 + }, + { + "epoch": 1.7859968079938935, + "grad_norm": 1.6932533979415894, + "learning_rate": 0.00012200630698911423, + "loss": 0.6154, + "step": 25738 + }, + { + "epoch": 1.7860661994309903, + "grad_norm": 1.4296565055847168, + "learning_rate": 0.00012200082607564143, + "loss": 0.5756, + "step": 25739 + }, + { + "epoch": 1.786135590868087, + "grad_norm": 1.6856549978256226, + "learning_rate": 0.00012199534509271454, + "loss": 0.5194, + "step": 25740 + }, + { + "epoch": 1.7862049823051835, + "grad_norm": 1.550944447517395, + "learning_rate": 0.0001219898640403509, + "loss": 0.6964, + "step": 25741 + }, + { + "epoch": 1.7862743737422802, + "grad_norm": 1.3554972410202026, + "learning_rate": 0.00012198438291856782, + "loss": 0.3829, + "step": 25742 + }, + { + "epoch": 1.7863437651793768, + "grad_norm": 1.4731111526489258, + "learning_rate": 0.00012197890172738257, + "loss": 0.5944, + "step": 25743 + }, + { + "epoch": 1.7864131566164736, + "grad_norm": 1.7391868829727173, + "learning_rate": 0.00012197342046681252, + "loss": 0.6263, + "step": 25744 + }, + { + "epoch": 1.7864825480535702, + "grad_norm": 1.0671895742416382, + "learning_rate": 0.0001219679391368749, + "loss": 0.4124, + "step": 25745 + }, + { + "epoch": 1.7865519394906668, + "grad_norm": 2.4104301929473877, + "learning_rate": 0.00012196245773758706, + "loss": 0.6864, + "step": 25746 + }, + { + "epoch": 1.7866213309277637, + "grad_norm": 1.3337336778640747, + "learning_rate": 0.0001219569762689663, + "loss": 0.5018, + "step": 25747 + }, + { + "epoch": 1.78669072236486, + "grad_norm": 1.452012538909912, + "learning_rate": 0.00012195149473102986, + "loss": 0.6169, + "step": 25748 + }, + { + "epoch": 1.7867601138019569, + "grad_norm": 2.2631895542144775, + "learning_rate": 0.00012194601312379514, + "loss": 0.6508, + "step": 25749 + }, + { + "epoch": 1.7868295052390535, + "grad_norm": 1.852865219116211, + "learning_rate": 0.00012194053144727937, + "loss": 0.5698, + "step": 25750 + }, + { + "epoch": 1.78689889667615, + "grad_norm": 1.2385950088500977, + "learning_rate": 0.00012193504970149993, + "loss": 0.4987, + "step": 25751 + }, + { + "epoch": 1.786968288113247, + "grad_norm": 1.7627737522125244, + "learning_rate": 0.00012192956788647407, + "loss": 0.5341, + "step": 25752 + }, + { + "epoch": 1.7870376795503435, + "grad_norm": 2.1310806274414062, + "learning_rate": 0.00012192408600221909, + "loss": 0.5242, + "step": 25753 + }, + { + "epoch": 1.7871070709874401, + "grad_norm": 1.5689022541046143, + "learning_rate": 0.00012191860404875235, + "loss": 0.5303, + "step": 25754 + }, + { + "epoch": 1.7871764624245368, + "grad_norm": 1.9375100135803223, + "learning_rate": 0.00012191312202609109, + "loss": 0.5917, + "step": 25755 + }, + { + "epoch": 1.7872458538616334, + "grad_norm": 1.6194136142730713, + "learning_rate": 0.00012190763993425265, + "loss": 0.7763, + "step": 25756 + }, + { + "epoch": 1.7873152452987302, + "grad_norm": 1.7226238250732422, + "learning_rate": 0.00012190215777325433, + "loss": 0.5915, + "step": 25757 + }, + { + "epoch": 1.7873846367358268, + "grad_norm": 1.7546337842941284, + "learning_rate": 0.00012189667554311346, + "loss": 0.6784, + "step": 25758 + }, + { + "epoch": 1.7874540281729234, + "grad_norm": 1.6488139629364014, + "learning_rate": 0.00012189119324384732, + "loss": 0.7593, + "step": 25759 + }, + { + "epoch": 1.7875234196100203, + "grad_norm": 1.8357362747192383, + "learning_rate": 0.00012188571087547321, + "loss": 0.4475, + "step": 25760 + }, + { + "epoch": 1.7875928110471166, + "grad_norm": 1.2555865049362183, + "learning_rate": 0.00012188022843800846, + "loss": 0.5035, + "step": 25761 + }, + { + "epoch": 1.7876622024842135, + "grad_norm": 1.5846152305603027, + "learning_rate": 0.00012187474593147037, + "loss": 0.5953, + "step": 25762 + }, + { + "epoch": 1.78773159392131, + "grad_norm": 1.7475054264068604, + "learning_rate": 0.00012186926335587623, + "loss": 0.8394, + "step": 25763 + }, + { + "epoch": 1.7878009853584067, + "grad_norm": 0.9942208528518677, + "learning_rate": 0.00012186378071124338, + "loss": 0.519, + "step": 25764 + }, + { + "epoch": 1.7878703767955035, + "grad_norm": 1.5596327781677246, + "learning_rate": 0.00012185829799758907, + "loss": 0.5401, + "step": 25765 + }, + { + "epoch": 1.7879397682326001, + "grad_norm": 2.0210764408111572, + "learning_rate": 0.00012185281521493069, + "loss": 0.9512, + "step": 25766 + }, + { + "epoch": 1.7880091596696968, + "grad_norm": 1.3025375604629517, + "learning_rate": 0.0001218473323632855, + "loss": 0.5, + "step": 25767 + }, + { + "epoch": 1.7880785511067934, + "grad_norm": 1.9459279775619507, + "learning_rate": 0.00012184184944267083, + "loss": 0.5132, + "step": 25768 + }, + { + "epoch": 1.78814794254389, + "grad_norm": 1.5995744466781616, + "learning_rate": 0.00012183636645310395, + "loss": 0.5327, + "step": 25769 + }, + { + "epoch": 1.7882173339809868, + "grad_norm": 1.4360555410385132, + "learning_rate": 0.00012183088339460218, + "loss": 0.4815, + "step": 25770 + }, + { + "epoch": 1.7882867254180834, + "grad_norm": 1.1825742721557617, + "learning_rate": 0.0001218254002671829, + "loss": 0.3881, + "step": 25771 + }, + { + "epoch": 1.78835611685518, + "grad_norm": 2.763740301132202, + "learning_rate": 0.00012181991707086331, + "loss": 0.6343, + "step": 25772 + }, + { + "epoch": 1.7884255082922769, + "grad_norm": 2.0221879482269287, + "learning_rate": 0.0001218144338056608, + "loss": 0.7601, + "step": 25773 + }, + { + "epoch": 1.7884948997293733, + "grad_norm": 1.368933081626892, + "learning_rate": 0.00012180895047159266, + "loss": 0.6367, + "step": 25774 + }, + { + "epoch": 1.78856429116647, + "grad_norm": 1.6697977781295776, + "learning_rate": 0.00012180346706867616, + "loss": 0.6319, + "step": 25775 + }, + { + "epoch": 1.7886336826035667, + "grad_norm": 1.5743857622146606, + "learning_rate": 0.00012179798359692869, + "loss": 0.5208, + "step": 25776 + }, + { + "epoch": 1.7887030740406633, + "grad_norm": 2.3347785472869873, + "learning_rate": 0.00012179250005636749, + "loss": 0.8154, + "step": 25777 + }, + { + "epoch": 1.7887724654777601, + "grad_norm": 1.5701789855957031, + "learning_rate": 0.0001217870164470099, + "loss": 0.6109, + "step": 25778 + }, + { + "epoch": 1.7888418569148568, + "grad_norm": 1.5475127696990967, + "learning_rate": 0.00012178153276887323, + "loss": 0.5698, + "step": 25779 + }, + { + "epoch": 1.7889112483519534, + "grad_norm": 1.8758606910705566, + "learning_rate": 0.00012177604902197477, + "loss": 0.579, + "step": 25780 + }, + { + "epoch": 1.7889806397890502, + "grad_norm": 1.259189486503601, + "learning_rate": 0.00012177056520633186, + "loss": 0.3349, + "step": 25781 + }, + { + "epoch": 1.7890500312261466, + "grad_norm": 2.1866772174835205, + "learning_rate": 0.0001217650813219618, + "loss": 0.709, + "step": 25782 + }, + { + "epoch": 1.7891194226632434, + "grad_norm": 1.488670825958252, + "learning_rate": 0.00012175959736888191, + "loss": 0.5884, + "step": 25783 + }, + { + "epoch": 1.78918881410034, + "grad_norm": 1.1823477745056152, + "learning_rate": 0.00012175411334710949, + "loss": 0.4602, + "step": 25784 + }, + { + "epoch": 1.7892582055374366, + "grad_norm": 1.5313396453857422, + "learning_rate": 0.00012174862925666186, + "loss": 0.5562, + "step": 25785 + }, + { + "epoch": 1.7893275969745335, + "grad_norm": 1.23040771484375, + "learning_rate": 0.00012174314509755633, + "loss": 0.5207, + "step": 25786 + }, + { + "epoch": 1.7893969884116299, + "grad_norm": 1.979196548461914, + "learning_rate": 0.0001217376608698102, + "loss": 0.7329, + "step": 25787 + }, + { + "epoch": 1.7894663798487267, + "grad_norm": 1.787619948387146, + "learning_rate": 0.0001217321765734408, + "loss": 0.5511, + "step": 25788 + }, + { + "epoch": 1.7895357712858233, + "grad_norm": 3.597292900085449, + "learning_rate": 0.00012172669220846548, + "loss": 0.7303, + "step": 25789 + }, + { + "epoch": 1.78960516272292, + "grad_norm": 1.524781584739685, + "learning_rate": 0.00012172120777490149, + "loss": 0.659, + "step": 25790 + }, + { + "epoch": 1.7896745541600168, + "grad_norm": 1.9472088813781738, + "learning_rate": 0.00012171572327276615, + "loss": 0.6196, + "step": 25791 + }, + { + "epoch": 1.7897439455971134, + "grad_norm": 1.565375804901123, + "learning_rate": 0.00012171023870207681, + "loss": 0.4166, + "step": 25792 + }, + { + "epoch": 1.78981333703421, + "grad_norm": 1.7225807905197144, + "learning_rate": 0.00012170475406285077, + "loss": 0.7284, + "step": 25793 + }, + { + "epoch": 1.7898827284713068, + "grad_norm": 1.6476633548736572, + "learning_rate": 0.00012169926935510532, + "loss": 0.582, + "step": 25794 + }, + { + "epoch": 1.7899521199084032, + "grad_norm": 2.6319808959960938, + "learning_rate": 0.0001216937845788578, + "loss": 0.5191, + "step": 25795 + }, + { + "epoch": 1.7900215113455, + "grad_norm": 1.3725273609161377, + "learning_rate": 0.00012168829973412554, + "loss": 0.4553, + "step": 25796 + }, + { + "epoch": 1.7900909027825966, + "grad_norm": 2.3591816425323486, + "learning_rate": 0.00012168281482092581, + "loss": 0.9546, + "step": 25797 + }, + { + "epoch": 1.7901602942196932, + "grad_norm": 1.87462317943573, + "learning_rate": 0.00012167732983927594, + "loss": 0.7007, + "step": 25798 + }, + { + "epoch": 1.79022968565679, + "grad_norm": 1.3445489406585693, + "learning_rate": 0.00012167184478919329, + "loss": 0.5845, + "step": 25799 + }, + { + "epoch": 1.7902990770938865, + "grad_norm": 1.62014639377594, + "learning_rate": 0.0001216663596706951, + "loss": 0.5222, + "step": 25800 + }, + { + "epoch": 1.7903684685309833, + "grad_norm": 1.7167071104049683, + "learning_rate": 0.00012166087448379875, + "loss": 0.5069, + "step": 25801 + }, + { + "epoch": 1.79043785996808, + "grad_norm": 1.5126665830612183, + "learning_rate": 0.00012165538922852154, + "loss": 0.545, + "step": 25802 + }, + { + "epoch": 1.7905072514051765, + "grad_norm": 1.682905673980713, + "learning_rate": 0.00012164990390488075, + "loss": 0.7307, + "step": 25803 + }, + { + "epoch": 1.7905766428422734, + "grad_norm": 1.2239621877670288, + "learning_rate": 0.00012164441851289374, + "loss": 0.4813, + "step": 25804 + }, + { + "epoch": 1.79064603427937, + "grad_norm": 1.2061643600463867, + "learning_rate": 0.0001216389330525778, + "loss": 0.629, + "step": 25805 + }, + { + "epoch": 1.7907154257164666, + "grad_norm": 1.4200607538223267, + "learning_rate": 0.00012163344752395029, + "loss": 0.658, + "step": 25806 + }, + { + "epoch": 1.7907848171535634, + "grad_norm": 1.935913324356079, + "learning_rate": 0.00012162796192702848, + "loss": 0.4453, + "step": 25807 + }, + { + "epoch": 1.7908542085906598, + "grad_norm": 1.8939018249511719, + "learning_rate": 0.00012162247626182969, + "loss": 0.4235, + "step": 25808 + }, + { + "epoch": 1.7909236000277566, + "grad_norm": 1.217688798904419, + "learning_rate": 0.00012161699052837128, + "loss": 0.3666, + "step": 25809 + }, + { + "epoch": 1.7909929914648532, + "grad_norm": 1.0924546718597412, + "learning_rate": 0.0001216115047266705, + "loss": 0.4632, + "step": 25810 + }, + { + "epoch": 1.7910623829019499, + "grad_norm": 1.8534200191497803, + "learning_rate": 0.00012160601885674477, + "loss": 0.5384, + "step": 25811 + }, + { + "epoch": 1.7911317743390467, + "grad_norm": 1.6156253814697266, + "learning_rate": 0.00012160053291861129, + "loss": 0.5226, + "step": 25812 + }, + { + "epoch": 1.791201165776143, + "grad_norm": 1.9718542098999023, + "learning_rate": 0.00012159504691228747, + "loss": 0.8484, + "step": 25813 + }, + { + "epoch": 1.79127055721324, + "grad_norm": 1.9605183601379395, + "learning_rate": 0.00012158956083779056, + "loss": 0.6618, + "step": 25814 + }, + { + "epoch": 1.7913399486503365, + "grad_norm": 1.6720900535583496, + "learning_rate": 0.00012158407469513791, + "loss": 0.7482, + "step": 25815 + }, + { + "epoch": 1.7914093400874331, + "grad_norm": 2.1844911575317383, + "learning_rate": 0.00012157858848434687, + "loss": 0.604, + "step": 25816 + }, + { + "epoch": 1.79147873152453, + "grad_norm": 1.5063047409057617, + "learning_rate": 0.00012157310220543469, + "loss": 0.6195, + "step": 25817 + }, + { + "epoch": 1.7915481229616266, + "grad_norm": 1.6159217357635498, + "learning_rate": 0.00012156761585841876, + "loss": 0.7919, + "step": 25818 + }, + { + "epoch": 1.7916175143987232, + "grad_norm": 1.3441333770751953, + "learning_rate": 0.00012156212944331635, + "loss": 0.4232, + "step": 25819 + }, + { + "epoch": 1.79168690583582, + "grad_norm": 1.8142668008804321, + "learning_rate": 0.00012155664296014481, + "loss": 0.7564, + "step": 25820 + }, + { + "epoch": 1.7917562972729164, + "grad_norm": 1.7433096170425415, + "learning_rate": 0.00012155115640892145, + "loss": 0.5531, + "step": 25821 + }, + { + "epoch": 1.7918256887100132, + "grad_norm": 1.5577017068862915, + "learning_rate": 0.0001215456697896636, + "loss": 0.5877, + "step": 25822 + }, + { + "epoch": 1.7918950801471099, + "grad_norm": 1.2487293481826782, + "learning_rate": 0.00012154018310238858, + "loss": 0.6538, + "step": 25823 + }, + { + "epoch": 1.7919644715842065, + "grad_norm": 1.6216185092926025, + "learning_rate": 0.00012153469634711366, + "loss": 0.7039, + "step": 25824 + }, + { + "epoch": 1.7920338630213033, + "grad_norm": 1.292345404624939, + "learning_rate": 0.0001215292095238562, + "loss": 0.4906, + "step": 25825 + }, + { + "epoch": 1.7921032544583997, + "grad_norm": 3.1742241382598877, + "learning_rate": 0.00012152372263263356, + "loss": 0.4062, + "step": 25826 + }, + { + "epoch": 1.7921726458954965, + "grad_norm": 2.0127851963043213, + "learning_rate": 0.00012151823567346302, + "loss": 0.7114, + "step": 25827 + }, + { + "epoch": 1.7922420373325931, + "grad_norm": 2.2319371700286865, + "learning_rate": 0.0001215127486463619, + "loss": 0.6717, + "step": 25828 + }, + { + "epoch": 1.7923114287696897, + "grad_norm": 1.287582278251648, + "learning_rate": 0.00012150726155134753, + "loss": 0.6763, + "step": 25829 + }, + { + "epoch": 1.7923808202067866, + "grad_norm": 1.753739833831787, + "learning_rate": 0.00012150177438843725, + "loss": 0.6558, + "step": 25830 + }, + { + "epoch": 1.7924502116438832, + "grad_norm": 1.157359004020691, + "learning_rate": 0.00012149628715764832, + "loss": 0.3889, + "step": 25831 + }, + { + "epoch": 1.7925196030809798, + "grad_norm": 1.9613598585128784, + "learning_rate": 0.00012149079985899812, + "loss": 0.5584, + "step": 25832 + }, + { + "epoch": 1.7925889945180766, + "grad_norm": 1.4530209302902222, + "learning_rate": 0.00012148531249250394, + "loss": 0.6635, + "step": 25833 + }, + { + "epoch": 1.792658385955173, + "grad_norm": 2.3669536113739014, + "learning_rate": 0.00012147982505818316, + "loss": 0.7547, + "step": 25834 + }, + { + "epoch": 1.7927277773922699, + "grad_norm": 1.5583637952804565, + "learning_rate": 0.00012147433755605304, + "loss": 0.5808, + "step": 25835 + }, + { + "epoch": 1.7927971688293665, + "grad_norm": 1.6403697729110718, + "learning_rate": 0.00012146884998613093, + "loss": 0.629, + "step": 25836 + }, + { + "epoch": 1.792866560266463, + "grad_norm": 3.378488779067993, + "learning_rate": 0.00012146336234843419, + "loss": 0.5481, + "step": 25837 + }, + { + "epoch": 1.79293595170356, + "grad_norm": 1.6264293193817139, + "learning_rate": 0.00012145787464298006, + "loss": 0.5547, + "step": 25838 + }, + { + "epoch": 1.7930053431406563, + "grad_norm": 2.4446845054626465, + "learning_rate": 0.00012145238686978591, + "loss": 0.7979, + "step": 25839 + }, + { + "epoch": 1.7930747345777531, + "grad_norm": 1.6376334428787231, + "learning_rate": 0.0001214468990288691, + "loss": 0.6002, + "step": 25840 + }, + { + "epoch": 1.7931441260148497, + "grad_norm": 1.2516424655914307, + "learning_rate": 0.00012144141112024687, + "loss": 0.3749, + "step": 25841 + }, + { + "epoch": 1.7932135174519463, + "grad_norm": 1.4006378650665283, + "learning_rate": 0.00012143592314393661, + "loss": 0.4358, + "step": 25842 + }, + { + "epoch": 1.7932829088890432, + "grad_norm": 1.0248206853866577, + "learning_rate": 0.00012143043509995564, + "loss": 0.295, + "step": 25843 + }, + { + "epoch": 1.7933523003261398, + "grad_norm": 1.4652965068817139, + "learning_rate": 0.00012142494698832128, + "loss": 0.6293, + "step": 25844 + }, + { + "epoch": 1.7934216917632364, + "grad_norm": 2.320300579071045, + "learning_rate": 0.00012141945880905083, + "loss": 0.8216, + "step": 25845 + }, + { + "epoch": 1.7934910832003332, + "grad_norm": 0.9435873627662659, + "learning_rate": 0.00012141397056216164, + "loss": 0.4261, + "step": 25846 + }, + { + "epoch": 1.7935604746374296, + "grad_norm": 1.6729178428649902, + "learning_rate": 0.00012140848224767105, + "loss": 0.6524, + "step": 25847 + }, + { + "epoch": 1.7936298660745265, + "grad_norm": 1.7297343015670776, + "learning_rate": 0.00012140299386559632, + "loss": 0.7538, + "step": 25848 + }, + { + "epoch": 1.793699257511623, + "grad_norm": 1.7074931859970093, + "learning_rate": 0.00012139750541595485, + "loss": 0.5789, + "step": 25849 + }, + { + "epoch": 1.7937686489487197, + "grad_norm": 1.7544769048690796, + "learning_rate": 0.00012139201689876394, + "loss": 0.5837, + "step": 25850 + }, + { + "epoch": 1.7938380403858165, + "grad_norm": 1.7499438524246216, + "learning_rate": 0.00012138652831404092, + "loss": 0.4891, + "step": 25851 + }, + { + "epoch": 1.793907431822913, + "grad_norm": 1.8626995086669922, + "learning_rate": 0.00012138103966180308, + "loss": 0.7312, + "step": 25852 + }, + { + "epoch": 1.7939768232600097, + "grad_norm": 1.8579614162445068, + "learning_rate": 0.00012137555094206779, + "loss": 0.6396, + "step": 25853 + }, + { + "epoch": 1.7940462146971063, + "grad_norm": 1.3688544034957886, + "learning_rate": 0.00012137006215485238, + "loss": 0.6254, + "step": 25854 + }, + { + "epoch": 1.794115606134203, + "grad_norm": 1.883475661277771, + "learning_rate": 0.00012136457330017416, + "loss": 0.6641, + "step": 25855 + }, + { + "epoch": 1.7941849975712998, + "grad_norm": 1.9409470558166504, + "learning_rate": 0.00012135908437805046, + "loss": 0.6257, + "step": 25856 + }, + { + "epoch": 1.7942543890083964, + "grad_norm": 2.0849685668945312, + "learning_rate": 0.0001213535953884986, + "loss": 0.5051, + "step": 25857 + }, + { + "epoch": 1.794323780445493, + "grad_norm": 1.2399506568908691, + "learning_rate": 0.00012134810633153593, + "loss": 0.441, + "step": 25858 + }, + { + "epoch": 1.7943931718825898, + "grad_norm": 1.6344740390777588, + "learning_rate": 0.00012134261720717974, + "loss": 0.6437, + "step": 25859 + }, + { + "epoch": 1.7944625633196862, + "grad_norm": 1.3048977851867676, + "learning_rate": 0.00012133712801544741, + "loss": 0.4672, + "step": 25860 + }, + { + "epoch": 1.794531954756783, + "grad_norm": 1.566235065460205, + "learning_rate": 0.00012133163875635625, + "loss": 0.5571, + "step": 25861 + }, + { + "epoch": 1.7946013461938797, + "grad_norm": 1.2398681640625, + "learning_rate": 0.00012132614942992355, + "loss": 0.4701, + "step": 25862 + }, + { + "epoch": 1.7946707376309763, + "grad_norm": 1.3381953239440918, + "learning_rate": 0.00012132066003616668, + "loss": 0.4788, + "step": 25863 + }, + { + "epoch": 1.7947401290680731, + "grad_norm": 2.2225341796875, + "learning_rate": 0.00012131517057510298, + "loss": 0.795, + "step": 25864 + }, + { + "epoch": 1.7948095205051695, + "grad_norm": 1.650523066520691, + "learning_rate": 0.00012130968104674976, + "loss": 0.6158, + "step": 25865 + }, + { + "epoch": 1.7948789119422663, + "grad_norm": 1.7731239795684814, + "learning_rate": 0.00012130419145112433, + "loss": 0.6262, + "step": 25866 + }, + { + "epoch": 1.794948303379363, + "grad_norm": 1.619067668914795, + "learning_rate": 0.00012129870178824403, + "loss": 0.5565, + "step": 25867 + }, + { + "epoch": 1.7950176948164596, + "grad_norm": 1.6178104877471924, + "learning_rate": 0.00012129321205812623, + "loss": 0.4895, + "step": 25868 + }, + { + "epoch": 1.7950870862535564, + "grad_norm": 1.8667405843734741, + "learning_rate": 0.00012128772226078821, + "loss": 0.6079, + "step": 25869 + }, + { + "epoch": 1.795156477690653, + "grad_norm": 3.5904345512390137, + "learning_rate": 0.00012128223239624732, + "loss": 0.4694, + "step": 25870 + }, + { + "epoch": 1.7952258691277496, + "grad_norm": 2.348721742630005, + "learning_rate": 0.00012127674246452088, + "loss": 0.8271, + "step": 25871 + }, + { + "epoch": 1.7952952605648465, + "grad_norm": 1.7108794450759888, + "learning_rate": 0.00012127125246562627, + "loss": 0.4598, + "step": 25872 + }, + { + "epoch": 1.7953646520019428, + "grad_norm": 1.8423770666122437, + "learning_rate": 0.00012126576239958074, + "loss": 0.5487, + "step": 25873 + }, + { + "epoch": 1.7954340434390397, + "grad_norm": 1.6015079021453857, + "learning_rate": 0.00012126027226640169, + "loss": 0.6844, + "step": 25874 + }, + { + "epoch": 1.7955034348761363, + "grad_norm": 1.7600778341293335, + "learning_rate": 0.00012125478206610643, + "loss": 0.4246, + "step": 25875 + }, + { + "epoch": 1.795572826313233, + "grad_norm": 1.7501516342163086, + "learning_rate": 0.0001212492917987123, + "loss": 0.777, + "step": 25876 + }, + { + "epoch": 1.7956422177503297, + "grad_norm": 1.674100637435913, + "learning_rate": 0.00012124380146423658, + "loss": 0.7267, + "step": 25877 + }, + { + "epoch": 1.7957116091874261, + "grad_norm": 1.4568300247192383, + "learning_rate": 0.00012123831106269667, + "loss": 0.522, + "step": 25878 + }, + { + "epoch": 1.795781000624523, + "grad_norm": 1.4076056480407715, + "learning_rate": 0.00012123282059410987, + "loss": 0.4837, + "step": 25879 + }, + { + "epoch": 1.7958503920616196, + "grad_norm": 1.492196798324585, + "learning_rate": 0.00012122733005849355, + "loss": 0.5476, + "step": 25880 + }, + { + "epoch": 1.7959197834987162, + "grad_norm": 1.4373093843460083, + "learning_rate": 0.00012122183945586498, + "loss": 0.6384, + "step": 25881 + }, + { + "epoch": 1.795989174935813, + "grad_norm": 1.4886776208877563, + "learning_rate": 0.00012121634878624155, + "loss": 0.6192, + "step": 25882 + }, + { + "epoch": 1.7960585663729096, + "grad_norm": 1.951303482055664, + "learning_rate": 0.00012121085804964054, + "loss": 0.6611, + "step": 25883 + }, + { + "epoch": 1.7961279578100062, + "grad_norm": 1.6011667251586914, + "learning_rate": 0.00012120536724607932, + "loss": 0.5893, + "step": 25884 + }, + { + "epoch": 1.796197349247103, + "grad_norm": 2.5745551586151123, + "learning_rate": 0.00012119987637557522, + "loss": 0.628, + "step": 25885 + }, + { + "epoch": 1.7962667406841994, + "grad_norm": 1.5012927055358887, + "learning_rate": 0.00012119438543814556, + "loss": 0.6521, + "step": 25886 + }, + { + "epoch": 1.7963361321212963, + "grad_norm": 3.930189371109009, + "learning_rate": 0.00012118889443380768, + "loss": 0.3989, + "step": 25887 + }, + { + "epoch": 1.796405523558393, + "grad_norm": 2.0452423095703125, + "learning_rate": 0.00012118340336257892, + "loss": 0.5936, + "step": 25888 + }, + { + "epoch": 1.7964749149954895, + "grad_norm": 1.821447491645813, + "learning_rate": 0.00012117791222447663, + "loss": 0.6314, + "step": 25889 + }, + { + "epoch": 1.7965443064325863, + "grad_norm": 1.710146188735962, + "learning_rate": 0.00012117242101951813, + "loss": 0.5752, + "step": 25890 + }, + { + "epoch": 1.7966136978696827, + "grad_norm": 2.2582156658172607, + "learning_rate": 0.00012116692974772072, + "loss": 0.7387, + "step": 25891 + }, + { + "epoch": 1.7966830893067796, + "grad_norm": 1.48561429977417, + "learning_rate": 0.00012116143840910182, + "loss": 0.4286, + "step": 25892 + }, + { + "epoch": 1.7967524807438762, + "grad_norm": 2.2658607959747314, + "learning_rate": 0.00012115594700367867, + "loss": 0.6963, + "step": 25893 + }, + { + "epoch": 1.7968218721809728, + "grad_norm": 2.1717569828033447, + "learning_rate": 0.00012115045553146867, + "loss": 0.9102, + "step": 25894 + }, + { + "epoch": 1.7968912636180696, + "grad_norm": 1.9261438846588135, + "learning_rate": 0.00012114496399248914, + "loss": 0.5136, + "step": 25895 + }, + { + "epoch": 1.7969606550551662, + "grad_norm": 1.9215010404586792, + "learning_rate": 0.0001211394723867574, + "loss": 0.5093, + "step": 25896 + }, + { + "epoch": 1.7970300464922628, + "grad_norm": 1.5289961099624634, + "learning_rate": 0.0001211339807142908, + "loss": 0.6128, + "step": 25897 + }, + { + "epoch": 1.7970994379293597, + "grad_norm": 2.407241106033325, + "learning_rate": 0.00012112848897510668, + "loss": 0.6509, + "step": 25898 + }, + { + "epoch": 1.797168829366456, + "grad_norm": 1.3529853820800781, + "learning_rate": 0.00012112299716922238, + "loss": 0.5632, + "step": 25899 + }, + { + "epoch": 1.7972382208035529, + "grad_norm": 2.1546318531036377, + "learning_rate": 0.00012111750529665522, + "loss": 0.7276, + "step": 25900 + }, + { + "epoch": 1.7973076122406495, + "grad_norm": 1.1813864707946777, + "learning_rate": 0.00012111201335742255, + "loss": 0.5597, + "step": 25901 + }, + { + "epoch": 1.797377003677746, + "grad_norm": 1.235625982284546, + "learning_rate": 0.00012110652135154168, + "loss": 0.5193, + "step": 25902 + }, + { + "epoch": 1.797446395114843, + "grad_norm": 1.4896520376205444, + "learning_rate": 0.00012110102927903, + "loss": 0.4822, + "step": 25903 + }, + { + "epoch": 1.7975157865519393, + "grad_norm": 1.6170843839645386, + "learning_rate": 0.0001210955371399048, + "loss": 0.6199, + "step": 25904 + }, + { + "epoch": 1.7975851779890362, + "grad_norm": 2.232971668243408, + "learning_rate": 0.00012109004493418343, + "loss": 0.7057, + "step": 25905 + }, + { + "epoch": 1.7976545694261328, + "grad_norm": 1.594744324684143, + "learning_rate": 0.00012108455266188328, + "loss": 0.7543, + "step": 25906 + }, + { + "epoch": 1.7977239608632294, + "grad_norm": 1.6383734941482544, + "learning_rate": 0.00012107906032302159, + "loss": 0.5169, + "step": 25907 + }, + { + "epoch": 1.7977933523003262, + "grad_norm": 1.7900633811950684, + "learning_rate": 0.00012107356791761576, + "loss": 0.6414, + "step": 25908 + }, + { + "epoch": 1.7978627437374228, + "grad_norm": 1.1390564441680908, + "learning_rate": 0.00012106807544568315, + "loss": 0.562, + "step": 25909 + }, + { + "epoch": 1.7979321351745194, + "grad_norm": 1.3824070692062378, + "learning_rate": 0.00012106258290724106, + "loss": 0.541, + "step": 25910 + }, + { + "epoch": 1.7980015266116163, + "grad_norm": 1.6399693489074707, + "learning_rate": 0.00012105709030230683, + "loss": 0.4822, + "step": 25911 + }, + { + "epoch": 1.7980709180487127, + "grad_norm": 1.5650677680969238, + "learning_rate": 0.00012105159763089782, + "loss": 0.524, + "step": 25912 + }, + { + "epoch": 1.7981403094858095, + "grad_norm": 1.7551337480545044, + "learning_rate": 0.00012104610489303136, + "loss": 0.6804, + "step": 25913 + }, + { + "epoch": 1.798209700922906, + "grad_norm": 1.332903504371643, + "learning_rate": 0.00012104061208872476, + "loss": 0.6486, + "step": 25914 + }, + { + "epoch": 1.7982790923600027, + "grad_norm": 1.0556210279464722, + "learning_rate": 0.00012103511921799544, + "loss": 0.3929, + "step": 25915 + }, + { + "epoch": 1.7983484837970996, + "grad_norm": 2.325392484664917, + "learning_rate": 0.00012102962628086065, + "loss": 0.5908, + "step": 25916 + }, + { + "epoch": 1.798417875234196, + "grad_norm": 1.6700270175933838, + "learning_rate": 0.00012102413327733779, + "loss": 0.5847, + "step": 25917 + }, + { + "epoch": 1.7984872666712928, + "grad_norm": 1.5205730199813843, + "learning_rate": 0.00012101864020744416, + "loss": 0.5122, + "step": 25918 + }, + { + "epoch": 1.7985566581083894, + "grad_norm": 1.2468472719192505, + "learning_rate": 0.00012101314707119713, + "loss": 0.4526, + "step": 25919 + }, + { + "epoch": 1.798626049545486, + "grad_norm": 2.0317041873931885, + "learning_rate": 0.00012100765386861404, + "loss": 0.6333, + "step": 25920 + }, + { + "epoch": 1.7986954409825828, + "grad_norm": 1.7734607458114624, + "learning_rate": 0.00012100216059971222, + "loss": 0.8238, + "step": 25921 + }, + { + "epoch": 1.7987648324196794, + "grad_norm": 1.4436395168304443, + "learning_rate": 0.000120996667264509, + "loss": 0.4512, + "step": 25922 + }, + { + "epoch": 1.798834223856776, + "grad_norm": 1.7949936389923096, + "learning_rate": 0.00012099117386302177, + "loss": 0.6393, + "step": 25923 + }, + { + "epoch": 1.7989036152938729, + "grad_norm": 1.3531427383422852, + "learning_rate": 0.0001209856803952678, + "loss": 0.5317, + "step": 25924 + }, + { + "epoch": 1.7989730067309693, + "grad_norm": 1.7279736995697021, + "learning_rate": 0.00012098018686126448, + "loss": 0.47, + "step": 25925 + }, + { + "epoch": 1.799042398168066, + "grad_norm": 1.6765036582946777, + "learning_rate": 0.00012097469326102917, + "loss": 0.5765, + "step": 25926 + }, + { + "epoch": 1.7991117896051627, + "grad_norm": 1.4851576089859009, + "learning_rate": 0.00012096919959457917, + "loss": 0.5866, + "step": 25927 + }, + { + "epoch": 1.7991811810422593, + "grad_norm": 1.1033154726028442, + "learning_rate": 0.00012096370586193184, + "loss": 0.367, + "step": 25928 + }, + { + "epoch": 1.7992505724793562, + "grad_norm": 1.492592453956604, + "learning_rate": 0.00012095821206310453, + "loss": 0.7086, + "step": 25929 + }, + { + "epoch": 1.7993199639164525, + "grad_norm": 1.7967498302459717, + "learning_rate": 0.00012095271819811458, + "loss": 0.5765, + "step": 25930 + }, + { + "epoch": 1.7993893553535494, + "grad_norm": 1.4872746467590332, + "learning_rate": 0.00012094722426697932, + "loss": 0.6841, + "step": 25931 + }, + { + "epoch": 1.799458746790646, + "grad_norm": 1.447988748550415, + "learning_rate": 0.0001209417302697161, + "loss": 0.5792, + "step": 25932 + }, + { + "epoch": 1.7995281382277426, + "grad_norm": 2.210503339767456, + "learning_rate": 0.0001209362362063423, + "loss": 0.616, + "step": 25933 + }, + { + "epoch": 1.7995975296648394, + "grad_norm": 1.1553101539611816, + "learning_rate": 0.00012093074207687518, + "loss": 0.4711, + "step": 25934 + }, + { + "epoch": 1.799666921101936, + "grad_norm": 1.3107402324676514, + "learning_rate": 0.00012092524788133217, + "loss": 0.5146, + "step": 25935 + }, + { + "epoch": 1.7997363125390327, + "grad_norm": 2.2753560543060303, + "learning_rate": 0.00012091975361973056, + "loss": 0.5538, + "step": 25936 + }, + { + "epoch": 1.7998057039761295, + "grad_norm": 1.768818974494934, + "learning_rate": 0.00012091425929208775, + "loss": 0.5846, + "step": 25937 + }, + { + "epoch": 1.7998750954132259, + "grad_norm": 1.457952618598938, + "learning_rate": 0.00012090876489842103, + "loss": 0.6127, + "step": 25938 + }, + { + "epoch": 1.7999444868503227, + "grad_norm": 1.7161614894866943, + "learning_rate": 0.00012090327043874773, + "loss": 0.612, + "step": 25939 + }, + { + "epoch": 1.8000138782874193, + "grad_norm": 1.3549737930297852, + "learning_rate": 0.00012089777591308527, + "loss": 0.4763, + "step": 25940 + }, + { + "epoch": 1.800083269724516, + "grad_norm": 1.4047093391418457, + "learning_rate": 0.00012089228132145094, + "loss": 0.4907, + "step": 25941 + }, + { + "epoch": 1.8001526611616128, + "grad_norm": 1.65049147605896, + "learning_rate": 0.00012088678666386211, + "loss": 0.6715, + "step": 25942 + }, + { + "epoch": 1.8002220525987092, + "grad_norm": 1.4567844867706299, + "learning_rate": 0.0001208812919403361, + "loss": 0.6145, + "step": 25943 + }, + { + "epoch": 1.800291444035806, + "grad_norm": 1.730517029762268, + "learning_rate": 0.0001208757971508903, + "loss": 0.6358, + "step": 25944 + }, + { + "epoch": 1.8003608354729026, + "grad_norm": 1.6784274578094482, + "learning_rate": 0.00012087030229554202, + "loss": 0.6014, + "step": 25945 + }, + { + "epoch": 1.8004302269099992, + "grad_norm": 1.7115092277526855, + "learning_rate": 0.0001208648073743086, + "loss": 0.7576, + "step": 25946 + }, + { + "epoch": 1.800499618347096, + "grad_norm": 1.5873539447784424, + "learning_rate": 0.00012085931238720743, + "loss": 0.6885, + "step": 25947 + }, + { + "epoch": 1.8005690097841927, + "grad_norm": 2.2281577587127686, + "learning_rate": 0.00012085381733425582, + "loss": 0.8066, + "step": 25948 + }, + { + "epoch": 1.8006384012212893, + "grad_norm": 1.4532668590545654, + "learning_rate": 0.00012084832221547113, + "loss": 0.3428, + "step": 25949 + }, + { + "epoch": 1.800707792658386, + "grad_norm": 1.6867622137069702, + "learning_rate": 0.00012084282703087073, + "loss": 0.6731, + "step": 25950 + }, + { + "epoch": 1.8007771840954825, + "grad_norm": 1.8359731435775757, + "learning_rate": 0.00012083733178047189, + "loss": 0.4866, + "step": 25951 + }, + { + "epoch": 1.8008465755325793, + "grad_norm": 1.3602325916290283, + "learning_rate": 0.00012083183646429209, + "loss": 0.6392, + "step": 25952 + }, + { + "epoch": 1.800915966969676, + "grad_norm": 1.7532691955566406, + "learning_rate": 0.00012082634108234852, + "loss": 0.6338, + "step": 25953 + }, + { + "epoch": 1.8009853584067725, + "grad_norm": 1.5444250106811523, + "learning_rate": 0.00012082084563465865, + "loss": 0.6455, + "step": 25954 + }, + { + "epoch": 1.8010547498438694, + "grad_norm": 1.99277663230896, + "learning_rate": 0.00012081535012123977, + "loss": 0.6258, + "step": 25955 + }, + { + "epoch": 1.8011241412809658, + "grad_norm": 1.4192121028900146, + "learning_rate": 0.00012080985454210925, + "loss": 0.5109, + "step": 25956 + }, + { + "epoch": 1.8011935327180626, + "grad_norm": 2.363708972930908, + "learning_rate": 0.00012080435889728441, + "loss": 0.5981, + "step": 25957 + }, + { + "epoch": 1.8012629241551592, + "grad_norm": 1.6024945974349976, + "learning_rate": 0.00012079886318678266, + "loss": 0.7158, + "step": 25958 + }, + { + "epoch": 1.8013323155922558, + "grad_norm": 1.4588614702224731, + "learning_rate": 0.00012079336741062129, + "loss": 0.622, + "step": 25959 + }, + { + "epoch": 1.8014017070293527, + "grad_norm": 1.3873120546340942, + "learning_rate": 0.00012078787156881768, + "loss": 0.5275, + "step": 25960 + }, + { + "epoch": 1.8014710984664493, + "grad_norm": 1.1026499271392822, + "learning_rate": 0.00012078237566138917, + "loss": 0.462, + "step": 25961 + }, + { + "epoch": 1.8015404899035459, + "grad_norm": 1.7312676906585693, + "learning_rate": 0.00012077687968835312, + "loss": 0.748, + "step": 25962 + }, + { + "epoch": 1.8016098813406427, + "grad_norm": 1.6557350158691406, + "learning_rate": 0.00012077138364972684, + "loss": 0.5356, + "step": 25963 + }, + { + "epoch": 1.801679272777739, + "grad_norm": 2.3335044384002686, + "learning_rate": 0.00012076588754552773, + "loss": 0.7082, + "step": 25964 + }, + { + "epoch": 1.801748664214836, + "grad_norm": 1.9032858610153198, + "learning_rate": 0.00012076039137577313, + "loss": 0.57, + "step": 25965 + }, + { + "epoch": 1.8018180556519325, + "grad_norm": 1.5657336711883545, + "learning_rate": 0.00012075489514048038, + "loss": 0.6338, + "step": 25966 + }, + { + "epoch": 1.8018874470890291, + "grad_norm": 2.3987934589385986, + "learning_rate": 0.00012074939883966682, + "loss": 0.6467, + "step": 25967 + }, + { + "epoch": 1.801956838526126, + "grad_norm": 1.8040815591812134, + "learning_rate": 0.00012074390247334985, + "loss": 0.5471, + "step": 25968 + }, + { + "epoch": 1.8020262299632226, + "grad_norm": 1.2177457809448242, + "learning_rate": 0.00012073840604154678, + "loss": 0.5513, + "step": 25969 + }, + { + "epoch": 1.8020956214003192, + "grad_norm": 1.817055106163025, + "learning_rate": 0.00012073290954427495, + "loss": 0.6649, + "step": 25970 + }, + { + "epoch": 1.8021650128374158, + "grad_norm": 1.2345560789108276, + "learning_rate": 0.00012072741298155172, + "loss": 0.5252, + "step": 25971 + }, + { + "epoch": 1.8022344042745124, + "grad_norm": 1.9614259004592896, + "learning_rate": 0.00012072191635339448, + "loss": 0.5193, + "step": 25972 + }, + { + "epoch": 1.8023037957116093, + "grad_norm": 2.4751811027526855, + "learning_rate": 0.00012071641965982053, + "loss": 0.7677, + "step": 25973 + }, + { + "epoch": 1.8023731871487059, + "grad_norm": 1.1961116790771484, + "learning_rate": 0.00012071092290084724, + "loss": 0.4335, + "step": 25974 + }, + { + "epoch": 1.8024425785858025, + "grad_norm": 2.034421443939209, + "learning_rate": 0.000120705426076492, + "loss": 0.4744, + "step": 25975 + }, + { + "epoch": 1.8025119700228993, + "grad_norm": 1.5845614671707153, + "learning_rate": 0.00012069992918677211, + "loss": 0.5037, + "step": 25976 + }, + { + "epoch": 1.8025813614599957, + "grad_norm": 1.5316365957260132, + "learning_rate": 0.00012069443223170496, + "loss": 0.5637, + "step": 25977 + }, + { + "epoch": 1.8026507528970925, + "grad_norm": 1.4343979358673096, + "learning_rate": 0.00012068893521130786, + "loss": 0.5089, + "step": 25978 + }, + { + "epoch": 1.8027201443341891, + "grad_norm": 1.41852605342865, + "learning_rate": 0.00012068343812559823, + "loss": 0.5423, + "step": 25979 + }, + { + "epoch": 1.8027895357712858, + "grad_norm": 1.378730297088623, + "learning_rate": 0.00012067794097459338, + "loss": 0.4756, + "step": 25980 + }, + { + "epoch": 1.8028589272083826, + "grad_norm": 2.225829839706421, + "learning_rate": 0.00012067244375831064, + "loss": 0.7351, + "step": 25981 + }, + { + "epoch": 1.8029283186454792, + "grad_norm": 1.9836130142211914, + "learning_rate": 0.00012066694647676742, + "loss": 0.6352, + "step": 25982 + }, + { + "epoch": 1.8029977100825758, + "grad_norm": 1.9983528852462769, + "learning_rate": 0.00012066144912998102, + "loss": 0.747, + "step": 25983 + }, + { + "epoch": 1.8030671015196724, + "grad_norm": 2.0334651470184326, + "learning_rate": 0.00012065595171796884, + "loss": 0.7561, + "step": 25984 + }, + { + "epoch": 1.803136492956769, + "grad_norm": 1.3540010452270508, + "learning_rate": 0.00012065045424074824, + "loss": 0.4599, + "step": 25985 + }, + { + "epoch": 1.8032058843938659, + "grad_norm": 1.4874781370162964, + "learning_rate": 0.00012064495669833652, + "loss": 0.6873, + "step": 25986 + }, + { + "epoch": 1.8032752758309625, + "grad_norm": 1.332566261291504, + "learning_rate": 0.0001206394590907511, + "loss": 0.4014, + "step": 25987 + }, + { + "epoch": 1.803344667268059, + "grad_norm": 2.0874836444854736, + "learning_rate": 0.00012063396141800927, + "loss": 0.5594, + "step": 25988 + }, + { + "epoch": 1.803414058705156, + "grad_norm": 1.6191959381103516, + "learning_rate": 0.00012062846368012843, + "loss": 0.5471, + "step": 25989 + }, + { + "epoch": 1.8034834501422523, + "grad_norm": 0.9456624388694763, + "learning_rate": 0.00012062296587712592, + "loss": 0.5053, + "step": 25990 + }, + { + "epoch": 1.8035528415793491, + "grad_norm": 1.204764723777771, + "learning_rate": 0.0001206174680090191, + "loss": 0.3425, + "step": 25991 + }, + { + "epoch": 1.8036222330164458, + "grad_norm": 1.2944279909133911, + "learning_rate": 0.00012061197007582535, + "loss": 0.4233, + "step": 25992 + }, + { + "epoch": 1.8036916244535424, + "grad_norm": 1.591565489768982, + "learning_rate": 0.00012060647207756195, + "loss": 0.7578, + "step": 25993 + }, + { + "epoch": 1.8037610158906392, + "grad_norm": 1.4530596733093262, + "learning_rate": 0.00012060097401424632, + "loss": 0.3941, + "step": 25994 + }, + { + "epoch": 1.8038304073277358, + "grad_norm": 1.6230659484863281, + "learning_rate": 0.00012059547588589581, + "loss": 0.5647, + "step": 25995 + }, + { + "epoch": 1.8038997987648324, + "grad_norm": 1.6832334995269775, + "learning_rate": 0.00012058997769252783, + "loss": 0.6199, + "step": 25996 + }, + { + "epoch": 1.803969190201929, + "grad_norm": 2.043544054031372, + "learning_rate": 0.0001205844794341596, + "loss": 0.6431, + "step": 25997 + }, + { + "epoch": 1.8040385816390256, + "grad_norm": 1.2817325592041016, + "learning_rate": 0.00012057898111080859, + "loss": 0.495, + "step": 25998 + }, + { + "epoch": 1.8041079730761225, + "grad_norm": 2.182116746902466, + "learning_rate": 0.00012057348272249213, + "loss": 0.7085, + "step": 25999 + }, + { + "epoch": 1.804177364513219, + "grad_norm": 1.6559484004974365, + "learning_rate": 0.00012056798426922758, + "loss": 0.566, + "step": 26000 + }, + { + "epoch": 1.8042467559503157, + "grad_norm": 1.6086633205413818, + "learning_rate": 0.00012056248575103227, + "loss": 0.67, + "step": 26001 + }, + { + "epoch": 1.8043161473874125, + "grad_norm": 1.3599693775177002, + "learning_rate": 0.00012055698716792358, + "loss": 0.5825, + "step": 26002 + }, + { + "epoch": 1.804385538824509, + "grad_norm": 1.449005365371704, + "learning_rate": 0.00012055148851991886, + "loss": 0.4084, + "step": 26003 + }, + { + "epoch": 1.8044549302616057, + "grad_norm": 1.145786166191101, + "learning_rate": 0.0001205459898070355, + "loss": 0.4456, + "step": 26004 + }, + { + "epoch": 1.8045243216987024, + "grad_norm": 1.708178997039795, + "learning_rate": 0.00012054049102929082, + "loss": 0.5763, + "step": 26005 + }, + { + "epoch": 1.804593713135799, + "grad_norm": 2.357462167739868, + "learning_rate": 0.00012053499218670222, + "loss": 0.7109, + "step": 26006 + }, + { + "epoch": 1.8046631045728958, + "grad_norm": 1.855222463607788, + "learning_rate": 0.00012052949327928698, + "loss": 0.5214, + "step": 26007 + }, + { + "epoch": 1.8047324960099924, + "grad_norm": 2.339198350906372, + "learning_rate": 0.00012052399430706253, + "loss": 0.7965, + "step": 26008 + }, + { + "epoch": 1.804801887447089, + "grad_norm": 1.6995302438735962, + "learning_rate": 0.00012051849527004622, + "loss": 0.621, + "step": 26009 + }, + { + "epoch": 1.8048712788841856, + "grad_norm": 1.7408252954483032, + "learning_rate": 0.0001205129961682554, + "loss": 0.4537, + "step": 26010 + }, + { + "epoch": 1.8049406703212822, + "grad_norm": 2.43125319480896, + "learning_rate": 0.00012050749700170742, + "loss": 0.6168, + "step": 26011 + }, + { + "epoch": 1.805010061758379, + "grad_norm": 2.0530447959899902, + "learning_rate": 0.00012050199777041966, + "loss": 0.4993, + "step": 26012 + }, + { + "epoch": 1.8050794531954757, + "grad_norm": 1.6582566499710083, + "learning_rate": 0.00012049649847440949, + "loss": 0.7183, + "step": 26013 + }, + { + "epoch": 1.8051488446325723, + "grad_norm": 1.6337097883224487, + "learning_rate": 0.00012049099911369421, + "loss": 0.5028, + "step": 26014 + }, + { + "epoch": 1.8052182360696691, + "grad_norm": 1.3228360414505005, + "learning_rate": 0.00012048549968829126, + "loss": 0.4963, + "step": 26015 + }, + { + "epoch": 1.8052876275067655, + "grad_norm": 2.148648262023926, + "learning_rate": 0.00012048000019821795, + "loss": 0.7898, + "step": 26016 + }, + { + "epoch": 1.8053570189438624, + "grad_norm": 1.9421865940093994, + "learning_rate": 0.00012047450064349166, + "loss": 0.573, + "step": 26017 + }, + { + "epoch": 1.805426410380959, + "grad_norm": 1.1523866653442383, + "learning_rate": 0.00012046900102412974, + "loss": 0.3649, + "step": 26018 + }, + { + "epoch": 1.8054958018180556, + "grad_norm": 1.3883507251739502, + "learning_rate": 0.00012046350134014956, + "loss": 0.4696, + "step": 26019 + }, + { + "epoch": 1.8055651932551524, + "grad_norm": 1.4110143184661865, + "learning_rate": 0.00012045800159156847, + "loss": 0.4653, + "step": 26020 + }, + { + "epoch": 1.805634584692249, + "grad_norm": 1.7093440294265747, + "learning_rate": 0.00012045250177840385, + "loss": 0.6728, + "step": 26021 + }, + { + "epoch": 1.8057039761293456, + "grad_norm": 1.7926980257034302, + "learning_rate": 0.00012044700190067311, + "loss": 0.5441, + "step": 26022 + }, + { + "epoch": 1.8057733675664425, + "grad_norm": 1.7581042051315308, + "learning_rate": 0.0001204415019583935, + "loss": 0.7164, + "step": 26023 + }, + { + "epoch": 1.8058427590035389, + "grad_norm": 1.2421082258224487, + "learning_rate": 0.00012043600195158245, + "loss": 0.4826, + "step": 26024 + }, + { + "epoch": 1.8059121504406357, + "grad_norm": 2.0543293952941895, + "learning_rate": 0.00012043050188025731, + "loss": 0.6414, + "step": 26025 + }, + { + "epoch": 1.8059815418777323, + "grad_norm": 1.7974460124969482, + "learning_rate": 0.00012042500174443543, + "loss": 0.4471, + "step": 26026 + }, + { + "epoch": 1.806050933314829, + "grad_norm": 1.6152952909469604, + "learning_rate": 0.00012041950154413421, + "loss": 0.5457, + "step": 26027 + }, + { + "epoch": 1.8061203247519257, + "grad_norm": 2.4720675945281982, + "learning_rate": 0.000120414001279371, + "loss": 0.5155, + "step": 26028 + }, + { + "epoch": 1.8061897161890221, + "grad_norm": 1.4455093145370483, + "learning_rate": 0.00012040850095016314, + "loss": 0.7563, + "step": 26029 + }, + { + "epoch": 1.806259107626119, + "grad_norm": 1.303459882736206, + "learning_rate": 0.00012040300055652803, + "loss": 0.4406, + "step": 26030 + }, + { + "epoch": 1.8063284990632156, + "grad_norm": 1.0430086851119995, + "learning_rate": 0.00012039750009848299, + "loss": 0.2986, + "step": 26031 + }, + { + "epoch": 1.8063978905003122, + "grad_norm": 1.8593417406082153, + "learning_rate": 0.00012039199957604541, + "loss": 0.551, + "step": 26032 + }, + { + "epoch": 1.806467281937409, + "grad_norm": 1.7918410301208496, + "learning_rate": 0.00012038649898923266, + "loss": 0.5452, + "step": 26033 + }, + { + "epoch": 1.8065366733745056, + "grad_norm": 1.2804030179977417, + "learning_rate": 0.00012038099833806212, + "loss": 0.443, + "step": 26034 + }, + { + "epoch": 1.8066060648116022, + "grad_norm": 1.8184806108474731, + "learning_rate": 0.00012037549762255109, + "loss": 0.5694, + "step": 26035 + }, + { + "epoch": 1.806675456248699, + "grad_norm": 1.825477123260498, + "learning_rate": 0.00012036999684271699, + "loss": 0.6637, + "step": 26036 + }, + { + "epoch": 1.8067448476857955, + "grad_norm": 1.3137439489364624, + "learning_rate": 0.0001203644959985772, + "loss": 0.6243, + "step": 26037 + }, + { + "epoch": 1.8068142391228923, + "grad_norm": 1.304944396018982, + "learning_rate": 0.00012035899509014905, + "loss": 0.3105, + "step": 26038 + }, + { + "epoch": 1.806883630559989, + "grad_norm": 1.4570926427841187, + "learning_rate": 0.00012035349411744992, + "loss": 0.4929, + "step": 26039 + }, + { + "epoch": 1.8069530219970855, + "grad_norm": 1.6524213552474976, + "learning_rate": 0.00012034799308049715, + "loss": 0.7195, + "step": 26040 + }, + { + "epoch": 1.8070224134341824, + "grad_norm": 1.5115368366241455, + "learning_rate": 0.00012034249197930813, + "loss": 0.6896, + "step": 26041 + }, + { + "epoch": 1.8070918048712787, + "grad_norm": 1.752201795578003, + "learning_rate": 0.00012033699081390024, + "loss": 0.6303, + "step": 26042 + }, + { + "epoch": 1.8071611963083756, + "grad_norm": 2.3079662322998047, + "learning_rate": 0.00012033148958429082, + "loss": 0.7389, + "step": 26043 + }, + { + "epoch": 1.8072305877454722, + "grad_norm": 1.3472769260406494, + "learning_rate": 0.00012032598829049723, + "loss": 0.4813, + "step": 26044 + }, + { + "epoch": 1.8072999791825688, + "grad_norm": 2.5934231281280518, + "learning_rate": 0.00012032048693253688, + "loss": 0.9952, + "step": 26045 + }, + { + "epoch": 1.8073693706196656, + "grad_norm": 1.402470350265503, + "learning_rate": 0.00012031498551042708, + "loss": 0.4406, + "step": 26046 + }, + { + "epoch": 1.8074387620567622, + "grad_norm": 1.9591145515441895, + "learning_rate": 0.00012030948402418525, + "loss": 0.6582, + "step": 26047 + }, + { + "epoch": 1.8075081534938588, + "grad_norm": 3.9177465438842773, + "learning_rate": 0.00012030398247382875, + "loss": 0.6862, + "step": 26048 + }, + { + "epoch": 1.8075775449309557, + "grad_norm": 1.8445007801055908, + "learning_rate": 0.0001202984808593749, + "loss": 0.4799, + "step": 26049 + }, + { + "epoch": 1.807646936368052, + "grad_norm": 1.853616714477539, + "learning_rate": 0.0001202929791808411, + "loss": 0.4055, + "step": 26050 + }, + { + "epoch": 1.807716327805149, + "grad_norm": 1.5747950077056885, + "learning_rate": 0.00012028747743824475, + "loss": 0.5938, + "step": 26051 + }, + { + "epoch": 1.8077857192422455, + "grad_norm": 1.8703893423080444, + "learning_rate": 0.00012028197563160314, + "loss": 0.5931, + "step": 26052 + }, + { + "epoch": 1.8078551106793421, + "grad_norm": 1.4297380447387695, + "learning_rate": 0.00012027647376093373, + "loss": 0.7439, + "step": 26053 + }, + { + "epoch": 1.807924502116439, + "grad_norm": 1.8367677927017212, + "learning_rate": 0.00012027097182625381, + "loss": 0.6368, + "step": 26054 + }, + { + "epoch": 1.8079938935535353, + "grad_norm": 1.780444622039795, + "learning_rate": 0.00012026546982758084, + "loss": 0.679, + "step": 26055 + }, + { + "epoch": 1.8080632849906322, + "grad_norm": 1.9971046447753906, + "learning_rate": 0.00012025996776493209, + "loss": 0.6901, + "step": 26056 + }, + { + "epoch": 1.8081326764277288, + "grad_norm": 2.336324453353882, + "learning_rate": 0.00012025446563832498, + "loss": 0.6929, + "step": 26057 + }, + { + "epoch": 1.8082020678648254, + "grad_norm": 2.3841869831085205, + "learning_rate": 0.0001202489634477769, + "loss": 1.1043, + "step": 26058 + }, + { + "epoch": 1.8082714593019222, + "grad_norm": 1.5178251266479492, + "learning_rate": 0.00012024346119330516, + "loss": 0.6724, + "step": 26059 + }, + { + "epoch": 1.8083408507390188, + "grad_norm": 2.077535629272461, + "learning_rate": 0.00012023795887492715, + "loss": 0.6256, + "step": 26060 + }, + { + "epoch": 1.8084102421761155, + "grad_norm": 1.3218508958816528, + "learning_rate": 0.00012023245649266027, + "loss": 0.5575, + "step": 26061 + }, + { + "epoch": 1.8084796336132123, + "grad_norm": 1.4954144954681396, + "learning_rate": 0.00012022695404652189, + "loss": 0.5679, + "step": 26062 + }, + { + "epoch": 1.8085490250503087, + "grad_norm": 1.3806463479995728, + "learning_rate": 0.00012022145153652933, + "loss": 0.5091, + "step": 26063 + }, + { + "epoch": 1.8086184164874055, + "grad_norm": 1.8184314966201782, + "learning_rate": 0.00012021594896269999, + "loss": 0.6744, + "step": 26064 + }, + { + "epoch": 1.8086878079245021, + "grad_norm": 1.3862743377685547, + "learning_rate": 0.0001202104463250513, + "loss": 0.555, + "step": 26065 + }, + { + "epoch": 1.8087571993615987, + "grad_norm": 1.5199331045150757, + "learning_rate": 0.00012020494362360051, + "loss": 0.7831, + "step": 26066 + }, + { + "epoch": 1.8088265907986956, + "grad_norm": 1.5146902799606323, + "learning_rate": 0.00012019944085836508, + "loss": 0.4702, + "step": 26067 + }, + { + "epoch": 1.808895982235792, + "grad_norm": 0.9590409398078918, + "learning_rate": 0.00012019393802936239, + "loss": 0.3906, + "step": 26068 + }, + { + "epoch": 1.8089653736728888, + "grad_norm": 1.9245188236236572, + "learning_rate": 0.00012018843513660974, + "loss": 0.607, + "step": 26069 + }, + { + "epoch": 1.8090347651099854, + "grad_norm": 1.0712355375289917, + "learning_rate": 0.00012018293218012455, + "loss": 0.4888, + "step": 26070 + }, + { + "epoch": 1.809104156547082, + "grad_norm": 1.3542886972427368, + "learning_rate": 0.00012017742915992418, + "loss": 0.5461, + "step": 26071 + }, + { + "epoch": 1.8091735479841788, + "grad_norm": 2.536832094192505, + "learning_rate": 0.00012017192607602603, + "loss": 0.7229, + "step": 26072 + }, + { + "epoch": 1.8092429394212755, + "grad_norm": 1.2814992666244507, + "learning_rate": 0.00012016642292844745, + "loss": 0.6533, + "step": 26073 + }, + { + "epoch": 1.809312330858372, + "grad_norm": 1.1236677169799805, + "learning_rate": 0.0001201609197172058, + "loss": 0.4599, + "step": 26074 + }, + { + "epoch": 1.809381722295469, + "grad_norm": 1.8051835298538208, + "learning_rate": 0.00012015541644231847, + "loss": 0.549, + "step": 26075 + }, + { + "epoch": 1.8094511137325653, + "grad_norm": 1.7295048236846924, + "learning_rate": 0.00012014991310380283, + "loss": 0.541, + "step": 26076 + }, + { + "epoch": 1.8095205051696621, + "grad_norm": 1.6125192642211914, + "learning_rate": 0.00012014440970167624, + "loss": 0.6325, + "step": 26077 + }, + { + "epoch": 1.8095898966067587, + "grad_norm": 1.1120021343231201, + "learning_rate": 0.00012013890623595608, + "loss": 0.4139, + "step": 26078 + }, + { + "epoch": 1.8096592880438553, + "grad_norm": 1.726705551147461, + "learning_rate": 0.00012013340270665976, + "loss": 0.6209, + "step": 26079 + }, + { + "epoch": 1.8097286794809522, + "grad_norm": 1.2438921928405762, + "learning_rate": 0.00012012789911380459, + "loss": 0.4389, + "step": 26080 + }, + { + "epoch": 1.8097980709180486, + "grad_norm": 1.3863134384155273, + "learning_rate": 0.000120122395457408, + "loss": 0.5858, + "step": 26081 + }, + { + "epoch": 1.8098674623551454, + "grad_norm": 1.2290291786193848, + "learning_rate": 0.00012011689173748735, + "loss": 0.5261, + "step": 26082 + }, + { + "epoch": 1.809936853792242, + "grad_norm": 1.2320928573608398, + "learning_rate": 0.00012011138795405998, + "loss": 0.4586, + "step": 26083 + }, + { + "epoch": 1.8100062452293386, + "grad_norm": 1.8063621520996094, + "learning_rate": 0.00012010588410714329, + "loss": 0.6646, + "step": 26084 + }, + { + "epoch": 1.8100756366664355, + "grad_norm": 1.4227888584136963, + "learning_rate": 0.00012010038019675467, + "loss": 0.5132, + "step": 26085 + }, + { + "epoch": 1.810145028103532, + "grad_norm": 1.0228296518325806, + "learning_rate": 0.0001200948762229115, + "loss": 0.4286, + "step": 26086 + }, + { + "epoch": 1.8102144195406287, + "grad_norm": 1.592087984085083, + "learning_rate": 0.0001200893721856311, + "loss": 0.7212, + "step": 26087 + }, + { + "epoch": 1.8102838109777255, + "grad_norm": 2.2239620685577393, + "learning_rate": 0.0001200838680849309, + "loss": 0.4981, + "step": 26088 + }, + { + "epoch": 1.810353202414822, + "grad_norm": 1.7137237787246704, + "learning_rate": 0.00012007836392082827, + "loss": 0.6766, + "step": 26089 + }, + { + "epoch": 1.8104225938519187, + "grad_norm": 1.6096608638763428, + "learning_rate": 0.00012007285969334056, + "loss": 0.7328, + "step": 26090 + }, + { + "epoch": 1.8104919852890153, + "grad_norm": 1.6801248788833618, + "learning_rate": 0.00012006735540248519, + "loss": 0.8769, + "step": 26091 + }, + { + "epoch": 1.810561376726112, + "grad_norm": 1.2801676988601685, + "learning_rate": 0.00012006185104827947, + "loss": 0.4522, + "step": 26092 + }, + { + "epoch": 1.8106307681632088, + "grad_norm": 1.7474693059921265, + "learning_rate": 0.00012005634663074085, + "loss": 0.6207, + "step": 26093 + }, + { + "epoch": 1.8107001596003052, + "grad_norm": 1.7633954286575317, + "learning_rate": 0.00012005084214988664, + "loss": 0.4707, + "step": 26094 + }, + { + "epoch": 1.810769551037402, + "grad_norm": 1.2181121110916138, + "learning_rate": 0.00012004533760573424, + "loss": 0.6779, + "step": 26095 + }, + { + "epoch": 1.8108389424744986, + "grad_norm": 1.717673897743225, + "learning_rate": 0.00012003983299830108, + "loss": 0.7682, + "step": 26096 + }, + { + "epoch": 1.8109083339115952, + "grad_norm": 1.5492987632751465, + "learning_rate": 0.00012003432832760447, + "loss": 0.5231, + "step": 26097 + }, + { + "epoch": 1.810977725348692, + "grad_norm": 2.46429181098938, + "learning_rate": 0.00012002882359366179, + "loss": 0.5885, + "step": 26098 + }, + { + "epoch": 1.8110471167857887, + "grad_norm": 1.9889473915100098, + "learning_rate": 0.00012002331879649043, + "loss": 0.7337, + "step": 26099 + }, + { + "epoch": 1.8111165082228853, + "grad_norm": 1.382429838180542, + "learning_rate": 0.0001200178139361078, + "loss": 0.6961, + "step": 26100 + }, + { + "epoch": 1.8111858996599821, + "grad_norm": 1.6768251657485962, + "learning_rate": 0.00012001230901253126, + "loss": 0.5663, + "step": 26101 + }, + { + "epoch": 1.8112552910970785, + "grad_norm": 1.6703970432281494, + "learning_rate": 0.00012000680402577816, + "loss": 0.5846, + "step": 26102 + }, + { + "epoch": 1.8113246825341753, + "grad_norm": 1.6546683311462402, + "learning_rate": 0.00012000129897586593, + "loss": 0.5992, + "step": 26103 + }, + { + "epoch": 1.811394073971272, + "grad_norm": 2.1861934661865234, + "learning_rate": 0.0001199957938628119, + "loss": 0.789, + "step": 26104 + }, + { + "epoch": 1.8114634654083686, + "grad_norm": 1.86986243724823, + "learning_rate": 0.00011999028868663347, + "loss": 0.6021, + "step": 26105 + }, + { + "epoch": 1.8115328568454654, + "grad_norm": 2.0595312118530273, + "learning_rate": 0.00011998478344734801, + "loss": 0.4946, + "step": 26106 + }, + { + "epoch": 1.8116022482825618, + "grad_norm": 1.6477794647216797, + "learning_rate": 0.00011997927814497291, + "loss": 0.5316, + "step": 26107 + }, + { + "epoch": 1.8116716397196586, + "grad_norm": 1.9101680517196655, + "learning_rate": 0.00011997377277952555, + "loss": 0.5247, + "step": 26108 + }, + { + "epoch": 1.8117410311567552, + "grad_norm": 1.6426771879196167, + "learning_rate": 0.00011996826735102333, + "loss": 0.5988, + "step": 26109 + }, + { + "epoch": 1.8118104225938518, + "grad_norm": 1.5742205381393433, + "learning_rate": 0.00011996276185948358, + "loss": 0.7808, + "step": 26110 + }, + { + "epoch": 1.8118798140309487, + "grad_norm": 1.703606367111206, + "learning_rate": 0.00011995725630492374, + "loss": 0.4863, + "step": 26111 + }, + { + "epoch": 1.8119492054680453, + "grad_norm": 1.5083976984024048, + "learning_rate": 0.00011995175068736113, + "loss": 0.5961, + "step": 26112 + }, + { + "epoch": 1.8120185969051419, + "grad_norm": 1.6973705291748047, + "learning_rate": 0.00011994624500681317, + "loss": 0.7224, + "step": 26113 + }, + { + "epoch": 1.8120879883422387, + "grad_norm": 2.4442145824432373, + "learning_rate": 0.0001199407392632972, + "loss": 0.884, + "step": 26114 + }, + { + "epoch": 1.812157379779335, + "grad_norm": 1.323957920074463, + "learning_rate": 0.00011993523345683065, + "loss": 0.6294, + "step": 26115 + }, + { + "epoch": 1.812226771216432, + "grad_norm": 1.4188311100006104, + "learning_rate": 0.00011992972758743087, + "loss": 0.5693, + "step": 26116 + }, + { + "epoch": 1.8122961626535286, + "grad_norm": 1.1209557056427002, + "learning_rate": 0.00011992422165511528, + "loss": 0.5509, + "step": 26117 + }, + { + "epoch": 1.8123655540906252, + "grad_norm": 1.7214165925979614, + "learning_rate": 0.0001199187156599012, + "loss": 0.6907, + "step": 26118 + }, + { + "epoch": 1.812434945527722, + "grad_norm": 1.5569201707839966, + "learning_rate": 0.00011991320960180605, + "loss": 0.4264, + "step": 26119 + }, + { + "epoch": 1.8125043369648184, + "grad_norm": 1.68424391746521, + "learning_rate": 0.00011990770348084723, + "loss": 0.6049, + "step": 26120 + }, + { + "epoch": 1.8125737284019152, + "grad_norm": 1.5438529253005981, + "learning_rate": 0.00011990219729704209, + "loss": 0.5985, + "step": 26121 + }, + { + "epoch": 1.8126431198390118, + "grad_norm": 2.6070632934570312, + "learning_rate": 0.00011989669105040801, + "loss": 0.4852, + "step": 26122 + }, + { + "epoch": 1.8127125112761084, + "grad_norm": 1.9292795658111572, + "learning_rate": 0.00011989118474096239, + "loss": 0.7301, + "step": 26123 + }, + { + "epoch": 1.8127819027132053, + "grad_norm": 1.947664499282837, + "learning_rate": 0.00011988567836872264, + "loss": 0.7023, + "step": 26124 + }, + { + "epoch": 1.8128512941503019, + "grad_norm": 1.8346694707870483, + "learning_rate": 0.00011988017193370607, + "loss": 0.6125, + "step": 26125 + }, + { + "epoch": 1.8129206855873985, + "grad_norm": 1.8433706760406494, + "learning_rate": 0.00011987466543593014, + "loss": 0.7681, + "step": 26126 + }, + { + "epoch": 1.8129900770244953, + "grad_norm": 1.634030818939209, + "learning_rate": 0.00011986915887541216, + "loss": 0.7037, + "step": 26127 + }, + { + "epoch": 1.8130594684615917, + "grad_norm": 1.6177512407302856, + "learning_rate": 0.00011986365225216958, + "loss": 0.4995, + "step": 26128 + }, + { + "epoch": 1.8131288598986885, + "grad_norm": 2.455434799194336, + "learning_rate": 0.00011985814556621973, + "loss": 0.6744, + "step": 26129 + }, + { + "epoch": 1.8131982513357852, + "grad_norm": 1.2948105335235596, + "learning_rate": 0.00011985263881758001, + "loss": 0.5543, + "step": 26130 + }, + { + "epoch": 1.8132676427728818, + "grad_norm": 1.8622791767120361, + "learning_rate": 0.00011984713200626785, + "loss": 0.7323, + "step": 26131 + }, + { + "epoch": 1.8133370342099786, + "grad_norm": 1.6491705179214478, + "learning_rate": 0.00011984162513230057, + "loss": 0.5108, + "step": 26132 + }, + { + "epoch": 1.813406425647075, + "grad_norm": 1.3829253911972046, + "learning_rate": 0.00011983611819569558, + "loss": 0.5272, + "step": 26133 + }, + { + "epoch": 1.8134758170841718, + "grad_norm": 1.6525055170059204, + "learning_rate": 0.0001198306111964703, + "loss": 0.4905, + "step": 26134 + }, + { + "epoch": 1.8135452085212684, + "grad_norm": 1.4568589925765991, + "learning_rate": 0.00011982510413464206, + "loss": 0.6617, + "step": 26135 + }, + { + "epoch": 1.813614599958365, + "grad_norm": 2.142639636993408, + "learning_rate": 0.00011981959701022824, + "loss": 0.6345, + "step": 26136 + }, + { + "epoch": 1.8136839913954619, + "grad_norm": 1.3471630811691284, + "learning_rate": 0.00011981408982324627, + "loss": 0.7024, + "step": 26137 + }, + { + "epoch": 1.8137533828325585, + "grad_norm": 1.7600722312927246, + "learning_rate": 0.00011980858257371354, + "loss": 0.6286, + "step": 26138 + }, + { + "epoch": 1.813822774269655, + "grad_norm": 1.0327086448669434, + "learning_rate": 0.00011980307526164738, + "loss": 0.5088, + "step": 26139 + }, + { + "epoch": 1.813892165706752, + "grad_norm": 1.2214330434799194, + "learning_rate": 0.00011979756788706522, + "loss": 0.4658, + "step": 26140 + }, + { + "epoch": 1.8139615571438483, + "grad_norm": 1.6259015798568726, + "learning_rate": 0.00011979206044998445, + "loss": 0.6066, + "step": 26141 + }, + { + "epoch": 1.8140309485809452, + "grad_norm": 1.5476064682006836, + "learning_rate": 0.00011978655295042242, + "loss": 0.4987, + "step": 26142 + }, + { + "epoch": 1.8141003400180418, + "grad_norm": 1.319495677947998, + "learning_rate": 0.00011978104538839654, + "loss": 0.6439, + "step": 26143 + }, + { + "epoch": 1.8141697314551384, + "grad_norm": 2.176652669906616, + "learning_rate": 0.00011977553776392422, + "loss": 0.6786, + "step": 26144 + }, + { + "epoch": 1.8142391228922352, + "grad_norm": 1.3520288467407227, + "learning_rate": 0.00011977003007702284, + "loss": 0.5137, + "step": 26145 + }, + { + "epoch": 1.8143085143293316, + "grad_norm": 1.5874210596084595, + "learning_rate": 0.00011976452232770972, + "loss": 0.6073, + "step": 26146 + }, + { + "epoch": 1.8143779057664284, + "grad_norm": 1.4125105142593384, + "learning_rate": 0.00011975901451600228, + "loss": 0.6035, + "step": 26147 + }, + { + "epoch": 1.814447297203525, + "grad_norm": 1.559430480003357, + "learning_rate": 0.00011975350664191798, + "loss": 0.592, + "step": 26148 + }, + { + "epoch": 1.8145166886406217, + "grad_norm": 1.5652714967727661, + "learning_rate": 0.00011974799870547413, + "loss": 0.5755, + "step": 26149 + }, + { + "epoch": 1.8145860800777185, + "grad_norm": 1.4378366470336914, + "learning_rate": 0.00011974249070668813, + "loss": 0.4009, + "step": 26150 + }, + { + "epoch": 1.814655471514815, + "grad_norm": 1.683479905128479, + "learning_rate": 0.00011973698264557738, + "loss": 0.664, + "step": 26151 + }, + { + "epoch": 1.8147248629519117, + "grad_norm": 1.1876460313796997, + "learning_rate": 0.00011973147452215927, + "loss": 0.4949, + "step": 26152 + }, + { + "epoch": 1.8147942543890085, + "grad_norm": 1.944859266281128, + "learning_rate": 0.00011972596633645116, + "loss": 0.7084, + "step": 26153 + }, + { + "epoch": 1.814863645826105, + "grad_norm": 2.48956036567688, + "learning_rate": 0.00011972045808847048, + "loss": 0.6249, + "step": 26154 + }, + { + "epoch": 1.8149330372632018, + "grad_norm": 1.5208183526992798, + "learning_rate": 0.00011971494977823463, + "loss": 0.414, + "step": 26155 + }, + { + "epoch": 1.8150024287002984, + "grad_norm": 1.555768370628357, + "learning_rate": 0.00011970944140576094, + "loss": 0.5505, + "step": 26156 + }, + { + "epoch": 1.815071820137395, + "grad_norm": 2.4422833919525146, + "learning_rate": 0.00011970393297106683, + "loss": 0.6972, + "step": 26157 + }, + { + "epoch": 1.8151412115744918, + "grad_norm": 1.7754019498825073, + "learning_rate": 0.0001196984244741697, + "loss": 0.8104, + "step": 26158 + }, + { + "epoch": 1.8152106030115882, + "grad_norm": 1.6265885829925537, + "learning_rate": 0.00011969291591508693, + "loss": 0.526, + "step": 26159 + }, + { + "epoch": 1.815279994448685, + "grad_norm": 1.3538517951965332, + "learning_rate": 0.0001196874072938359, + "loss": 0.4204, + "step": 26160 + }, + { + "epoch": 1.8153493858857817, + "grad_norm": 1.7382065057754517, + "learning_rate": 0.00011968189861043402, + "loss": 0.6107, + "step": 26161 + }, + { + "epoch": 1.8154187773228783, + "grad_norm": 1.4943886995315552, + "learning_rate": 0.00011967638986489866, + "loss": 0.5159, + "step": 26162 + }, + { + "epoch": 1.815488168759975, + "grad_norm": 1.5936357975006104, + "learning_rate": 0.00011967088105724724, + "loss": 0.5279, + "step": 26163 + }, + { + "epoch": 1.8155575601970717, + "grad_norm": 1.2823879718780518, + "learning_rate": 0.00011966537218749712, + "loss": 0.4919, + "step": 26164 + }, + { + "epoch": 1.8156269516341683, + "grad_norm": 1.667467713356018, + "learning_rate": 0.0001196598632556657, + "loss": 0.615, + "step": 26165 + }, + { + "epoch": 1.8156963430712652, + "grad_norm": 1.3207954168319702, + "learning_rate": 0.00011965435426177037, + "loss": 0.4335, + "step": 26166 + }, + { + "epoch": 1.8157657345083615, + "grad_norm": 1.5936590433120728, + "learning_rate": 0.0001196488452058285, + "loss": 0.4354, + "step": 26167 + }, + { + "epoch": 1.8158351259454584, + "grad_norm": 1.4814763069152832, + "learning_rate": 0.00011964333608785754, + "loss": 0.7683, + "step": 26168 + }, + { + "epoch": 1.815904517382555, + "grad_norm": 1.9291001558303833, + "learning_rate": 0.00011963782690787484, + "loss": 0.5487, + "step": 26169 + }, + { + "epoch": 1.8159739088196516, + "grad_norm": 2.243771553039551, + "learning_rate": 0.0001196323176658978, + "loss": 0.7929, + "step": 26170 + }, + { + "epoch": 1.8160433002567484, + "grad_norm": 1.5214775800704956, + "learning_rate": 0.0001196268083619438, + "loss": 0.4848, + "step": 26171 + }, + { + "epoch": 1.8161126916938448, + "grad_norm": 1.4196301698684692, + "learning_rate": 0.00011962129899603024, + "loss": 0.412, + "step": 26172 + }, + { + "epoch": 1.8161820831309416, + "grad_norm": 1.2986586093902588, + "learning_rate": 0.00011961578956817451, + "loss": 0.4969, + "step": 26173 + }, + { + "epoch": 1.8162514745680383, + "grad_norm": 1.3013770580291748, + "learning_rate": 0.00011961028007839403, + "loss": 0.4691, + "step": 26174 + }, + { + "epoch": 1.8163208660051349, + "grad_norm": 1.7966750860214233, + "learning_rate": 0.00011960477052670616, + "loss": 0.6444, + "step": 26175 + }, + { + "epoch": 1.8163902574422317, + "grad_norm": 2.1656382083892822, + "learning_rate": 0.00011959926091312832, + "loss": 0.8137, + "step": 26176 + }, + { + "epoch": 1.8164596488793283, + "grad_norm": 1.2674376964569092, + "learning_rate": 0.00011959375123767786, + "loss": 0.387, + "step": 26177 + }, + { + "epoch": 1.816529040316425, + "grad_norm": 1.2298884391784668, + "learning_rate": 0.00011958824150037224, + "loss": 0.4341, + "step": 26178 + }, + { + "epoch": 1.8165984317535218, + "grad_norm": 2.203261613845825, + "learning_rate": 0.00011958273170122878, + "loss": 0.6215, + "step": 26179 + }, + { + "epoch": 1.8166678231906181, + "grad_norm": 1.6759101152420044, + "learning_rate": 0.00011957722184026495, + "loss": 0.7095, + "step": 26180 + }, + { + "epoch": 1.816737214627715, + "grad_norm": 1.3523149490356445, + "learning_rate": 0.00011957171191749805, + "loss": 0.5303, + "step": 26181 + }, + { + "epoch": 1.8168066060648116, + "grad_norm": 1.5370444059371948, + "learning_rate": 0.00011956620193294554, + "loss": 0.5208, + "step": 26182 + }, + { + "epoch": 1.8168759975019082, + "grad_norm": 1.6183291673660278, + "learning_rate": 0.00011956069188662482, + "loss": 0.4549, + "step": 26183 + }, + { + "epoch": 1.816945388939005, + "grad_norm": 1.6608879566192627, + "learning_rate": 0.00011955518177855325, + "loss": 0.7587, + "step": 26184 + }, + { + "epoch": 1.8170147803761014, + "grad_norm": 1.976927638053894, + "learning_rate": 0.00011954967160874822, + "loss": 0.6531, + "step": 26185 + }, + { + "epoch": 1.8170841718131983, + "grad_norm": 2.1710257530212402, + "learning_rate": 0.0001195441613772272, + "loss": 0.548, + "step": 26186 + }, + { + "epoch": 1.8171535632502949, + "grad_norm": 1.7301337718963623, + "learning_rate": 0.00011953865108400747, + "loss": 0.8321, + "step": 26187 + }, + { + "epoch": 1.8172229546873915, + "grad_norm": 1.5459965467453003, + "learning_rate": 0.0001195331407291065, + "loss": 0.5439, + "step": 26188 + }, + { + "epoch": 1.8172923461244883, + "grad_norm": 1.668841004371643, + "learning_rate": 0.0001195276303125417, + "loss": 0.6104, + "step": 26189 + }, + { + "epoch": 1.817361737561585, + "grad_norm": 1.7250334024429321, + "learning_rate": 0.00011952211983433042, + "loss": 0.4547, + "step": 26190 + }, + { + "epoch": 1.8174311289986815, + "grad_norm": 2.1926112174987793, + "learning_rate": 0.00011951660929449006, + "loss": 0.9213, + "step": 26191 + }, + { + "epoch": 1.8175005204357784, + "grad_norm": 1.7347196340560913, + "learning_rate": 0.00011951109869303805, + "loss": 0.5489, + "step": 26192 + }, + { + "epoch": 1.8175699118728748, + "grad_norm": 1.0498199462890625, + "learning_rate": 0.00011950558802999177, + "loss": 0.4967, + "step": 26193 + }, + { + "epoch": 1.8176393033099716, + "grad_norm": 1.5344067811965942, + "learning_rate": 0.00011950007730536859, + "loss": 0.6599, + "step": 26194 + }, + { + "epoch": 1.8177086947470682, + "grad_norm": 1.1332165002822876, + "learning_rate": 0.00011949456651918591, + "loss": 0.4912, + "step": 26195 + }, + { + "epoch": 1.8177780861841648, + "grad_norm": 1.2294237613677979, + "learning_rate": 0.00011948905567146121, + "loss": 0.4083, + "step": 26196 + }, + { + "epoch": 1.8178474776212616, + "grad_norm": 2.1463589668273926, + "learning_rate": 0.00011948354476221175, + "loss": 0.7663, + "step": 26197 + }, + { + "epoch": 1.8179168690583583, + "grad_norm": 1.9208357334136963, + "learning_rate": 0.00011947803379145506, + "loss": 0.535, + "step": 26198 + }, + { + "epoch": 1.8179862604954549, + "grad_norm": 2.007793426513672, + "learning_rate": 0.00011947252275920844, + "loss": 0.5335, + "step": 26199 + }, + { + "epoch": 1.8180556519325515, + "grad_norm": 1.2665587663650513, + "learning_rate": 0.00011946701166548935, + "loss": 0.5208, + "step": 26200 + }, + { + "epoch": 1.818125043369648, + "grad_norm": 1.283340334892273, + "learning_rate": 0.00011946150051031515, + "loss": 0.4365, + "step": 26201 + }, + { + "epoch": 1.818194434806745, + "grad_norm": 1.8669068813323975, + "learning_rate": 0.00011945598929370323, + "loss": 0.5332, + "step": 26202 + }, + { + "epoch": 1.8182638262438415, + "grad_norm": 2.084712505340576, + "learning_rate": 0.00011945047801567106, + "loss": 0.5382, + "step": 26203 + }, + { + "epoch": 1.8183332176809381, + "grad_norm": 1.3871841430664062, + "learning_rate": 0.00011944496667623594, + "loss": 0.4476, + "step": 26204 + }, + { + "epoch": 1.818402609118035, + "grad_norm": 2.085070848464966, + "learning_rate": 0.00011943945527541533, + "loss": 0.5786, + "step": 26205 + }, + { + "epoch": 1.8184720005551314, + "grad_norm": 1.612406611442566, + "learning_rate": 0.00011943394381322661, + "loss": 0.5596, + "step": 26206 + }, + { + "epoch": 1.8185413919922282, + "grad_norm": 1.800331950187683, + "learning_rate": 0.00011942843228968721, + "loss": 0.4664, + "step": 26207 + }, + { + "epoch": 1.8186107834293248, + "grad_norm": 1.499463677406311, + "learning_rate": 0.00011942292070481447, + "loss": 0.4782, + "step": 26208 + }, + { + "epoch": 1.8186801748664214, + "grad_norm": 1.382474422454834, + "learning_rate": 0.00011941740905862584, + "loss": 0.5437, + "step": 26209 + }, + { + "epoch": 1.8187495663035183, + "grad_norm": 1.5535070896148682, + "learning_rate": 0.00011941189735113871, + "loss": 0.6907, + "step": 26210 + }, + { + "epoch": 1.8188189577406149, + "grad_norm": 1.6571838855743408, + "learning_rate": 0.00011940638558237045, + "loss": 0.4436, + "step": 26211 + }, + { + "epoch": 1.8188883491777115, + "grad_norm": 1.1991822719573975, + "learning_rate": 0.00011940087375233848, + "loss": 0.5239, + "step": 26212 + }, + { + "epoch": 1.818957740614808, + "grad_norm": 1.9554197788238525, + "learning_rate": 0.00011939536186106025, + "loss": 0.6729, + "step": 26213 + }, + { + "epoch": 1.8190271320519047, + "grad_norm": 1.9796273708343506, + "learning_rate": 0.00011938984990855306, + "loss": 0.7441, + "step": 26214 + }, + { + "epoch": 1.8190965234890015, + "grad_norm": 2.525294303894043, + "learning_rate": 0.00011938433789483441, + "loss": 0.667, + "step": 26215 + }, + { + "epoch": 1.8191659149260981, + "grad_norm": 1.3785256147384644, + "learning_rate": 0.00011937882581992161, + "loss": 0.4189, + "step": 26216 + }, + { + "epoch": 1.8192353063631947, + "grad_norm": 1.4392971992492676, + "learning_rate": 0.00011937331368383213, + "loss": 0.535, + "step": 26217 + }, + { + "epoch": 1.8193046978002916, + "grad_norm": 1.20177161693573, + "learning_rate": 0.00011936780148658336, + "loss": 0.3864, + "step": 26218 + }, + { + "epoch": 1.819374089237388, + "grad_norm": 3.265146255493164, + "learning_rate": 0.00011936228922819264, + "loss": 0.7109, + "step": 26219 + }, + { + "epoch": 1.8194434806744848, + "grad_norm": 1.932992696762085, + "learning_rate": 0.00011935677690867743, + "loss": 0.6515, + "step": 26220 + }, + { + "epoch": 1.8195128721115814, + "grad_norm": 1.6219666004180908, + "learning_rate": 0.00011935126452805516, + "loss": 0.5869, + "step": 26221 + }, + { + "epoch": 1.819582263548678, + "grad_norm": 1.4988608360290527, + "learning_rate": 0.00011934575208634317, + "loss": 0.5678, + "step": 26222 + }, + { + "epoch": 1.8196516549857749, + "grad_norm": 2.1110458374023438, + "learning_rate": 0.00011934023958355889, + "loss": 0.7376, + "step": 26223 + }, + { + "epoch": 1.8197210464228715, + "grad_norm": 1.580115795135498, + "learning_rate": 0.00011933472701971973, + "loss": 0.6285, + "step": 26224 + }, + { + "epoch": 1.819790437859968, + "grad_norm": 1.3095190525054932, + "learning_rate": 0.00011932921439484303, + "loss": 0.5104, + "step": 26225 + }, + { + "epoch": 1.8198598292970647, + "grad_norm": 1.4284024238586426, + "learning_rate": 0.00011932370170894629, + "loss": 0.301, + "step": 26226 + }, + { + "epoch": 1.8199292207341613, + "grad_norm": 1.6080191135406494, + "learning_rate": 0.00011931818896204685, + "loss": 0.5271, + "step": 26227 + }, + { + "epoch": 1.8199986121712581, + "grad_norm": 1.8558038473129272, + "learning_rate": 0.00011931267615416213, + "loss": 0.5905, + "step": 26228 + }, + { + "epoch": 1.8200680036083547, + "grad_norm": 1.668009877204895, + "learning_rate": 0.00011930716328530954, + "loss": 0.5435, + "step": 26229 + }, + { + "epoch": 1.8201373950454514, + "grad_norm": 1.111889362335205, + "learning_rate": 0.00011930165035550644, + "loss": 0.4501, + "step": 26230 + }, + { + "epoch": 1.8202067864825482, + "grad_norm": 1.8892871141433716, + "learning_rate": 0.00011929613736477032, + "loss": 0.5641, + "step": 26231 + }, + { + "epoch": 1.8202761779196446, + "grad_norm": 1.6165666580200195, + "learning_rate": 0.0001192906243131185, + "loss": 0.5798, + "step": 26232 + }, + { + "epoch": 1.8203455693567414, + "grad_norm": 1.2257003784179688, + "learning_rate": 0.00011928511120056844, + "loss": 0.4354, + "step": 26233 + }, + { + "epoch": 1.820414960793838, + "grad_norm": 1.2381854057312012, + "learning_rate": 0.00011927959802713753, + "loss": 0.4883, + "step": 26234 + }, + { + "epoch": 1.8204843522309346, + "grad_norm": 1.0832973718643188, + "learning_rate": 0.00011927408479284314, + "loss": 0.3407, + "step": 26235 + }, + { + "epoch": 1.8205537436680315, + "grad_norm": 1.6939141750335693, + "learning_rate": 0.0001192685714977027, + "loss": 0.5686, + "step": 26236 + }, + { + "epoch": 1.820623135105128, + "grad_norm": 1.2843358516693115, + "learning_rate": 0.00011926305814173361, + "loss": 0.5591, + "step": 26237 + }, + { + "epoch": 1.8206925265422247, + "grad_norm": 2.15704345703125, + "learning_rate": 0.00011925754472495332, + "loss": 0.5787, + "step": 26238 + }, + { + "epoch": 1.8207619179793213, + "grad_norm": 1.4802711009979248, + "learning_rate": 0.00011925203124737913, + "loss": 0.6595, + "step": 26239 + }, + { + "epoch": 1.820831309416418, + "grad_norm": 1.3168280124664307, + "learning_rate": 0.00011924651770902855, + "loss": 0.4178, + "step": 26240 + }, + { + "epoch": 1.8209007008535147, + "grad_norm": 1.3166937828063965, + "learning_rate": 0.00011924100410991897, + "loss": 0.3714, + "step": 26241 + }, + { + "epoch": 1.8209700922906114, + "grad_norm": 1.7994712591171265, + "learning_rate": 0.00011923549045006774, + "loss": 0.729, + "step": 26242 + }, + { + "epoch": 1.821039483727708, + "grad_norm": 1.6039727926254272, + "learning_rate": 0.0001192299767294923, + "loss": 0.6972, + "step": 26243 + }, + { + "epoch": 1.8211088751648048, + "grad_norm": 1.706931710243225, + "learning_rate": 0.00011922446294821006, + "loss": 0.6071, + "step": 26244 + }, + { + "epoch": 1.8211782666019012, + "grad_norm": 1.6428483724594116, + "learning_rate": 0.00011921894910623844, + "loss": 0.5996, + "step": 26245 + }, + { + "epoch": 1.821247658038998, + "grad_norm": 2.1724743843078613, + "learning_rate": 0.00011921343520359479, + "loss": 0.5151, + "step": 26246 + }, + { + "epoch": 1.8213170494760946, + "grad_norm": 1.0626155138015747, + "learning_rate": 0.00011920792124029656, + "loss": 0.3262, + "step": 26247 + }, + { + "epoch": 1.8213864409131912, + "grad_norm": 2.1570217609405518, + "learning_rate": 0.00011920240721636118, + "loss": 0.7772, + "step": 26248 + }, + { + "epoch": 1.821455832350288, + "grad_norm": 2.55780029296875, + "learning_rate": 0.000119196893131806, + "loss": 0.6414, + "step": 26249 + }, + { + "epoch": 1.8215252237873847, + "grad_norm": 1.9773794412612915, + "learning_rate": 0.00011919137898664849, + "loss": 0.6803, + "step": 26250 + }, + { + "epoch": 1.8215946152244813, + "grad_norm": 1.3792166709899902, + "learning_rate": 0.00011918586478090599, + "loss": 0.5752, + "step": 26251 + }, + { + "epoch": 1.821664006661578, + "grad_norm": 1.6233543157577515, + "learning_rate": 0.00011918035051459598, + "loss": 0.5659, + "step": 26252 + }, + { + "epoch": 1.8217333980986745, + "grad_norm": 1.9150729179382324, + "learning_rate": 0.00011917483618773577, + "loss": 0.7214, + "step": 26253 + }, + { + "epoch": 1.8218027895357713, + "grad_norm": 1.2450182437896729, + "learning_rate": 0.00011916932180034285, + "loss": 0.4983, + "step": 26254 + }, + { + "epoch": 1.821872180972868, + "grad_norm": 1.4800525903701782, + "learning_rate": 0.00011916380735243465, + "loss": 0.6296, + "step": 26255 + }, + { + "epoch": 1.8219415724099646, + "grad_norm": 1.4089359045028687, + "learning_rate": 0.00011915829284402849, + "loss": 0.6376, + "step": 26256 + }, + { + "epoch": 1.8220109638470614, + "grad_norm": 1.276757836341858, + "learning_rate": 0.00011915277827514182, + "loss": 0.4429, + "step": 26257 + }, + { + "epoch": 1.8220803552841578, + "grad_norm": 1.2744656801223755, + "learning_rate": 0.00011914726364579202, + "loss": 0.4907, + "step": 26258 + }, + { + "epoch": 1.8221497467212546, + "grad_norm": 1.2447094917297363, + "learning_rate": 0.00011914174895599662, + "loss": 0.3318, + "step": 26259 + }, + { + "epoch": 1.8222191381583512, + "grad_norm": 1.9321448802947998, + "learning_rate": 0.00011913623420577286, + "loss": 0.6198, + "step": 26260 + }, + { + "epoch": 1.8222885295954478, + "grad_norm": 1.2955740690231323, + "learning_rate": 0.00011913071939513827, + "loss": 0.5086, + "step": 26261 + }, + { + "epoch": 1.8223579210325447, + "grad_norm": 1.62236750125885, + "learning_rate": 0.0001191252045241102, + "loss": 0.8186, + "step": 26262 + }, + { + "epoch": 1.8224273124696413, + "grad_norm": 1.5780061483383179, + "learning_rate": 0.00011911968959270607, + "loss": 0.5117, + "step": 26263 + }, + { + "epoch": 1.822496703906738, + "grad_norm": 1.4337377548217773, + "learning_rate": 0.00011911417460094332, + "loss": 0.5237, + "step": 26264 + }, + { + "epoch": 1.8225660953438347, + "grad_norm": 1.74016535282135, + "learning_rate": 0.00011910865954883934, + "loss": 0.6224, + "step": 26265 + }, + { + "epoch": 1.8226354867809311, + "grad_norm": 2.53798246383667, + "learning_rate": 0.00011910314443641154, + "loss": 0.5701, + "step": 26266 + }, + { + "epoch": 1.822704878218028, + "grad_norm": 1.4597957134246826, + "learning_rate": 0.00011909762926367731, + "loss": 0.5994, + "step": 26267 + }, + { + "epoch": 1.8227742696551246, + "grad_norm": 1.576431155204773, + "learning_rate": 0.0001190921140306541, + "loss": 0.544, + "step": 26268 + }, + { + "epoch": 1.8228436610922212, + "grad_norm": 1.5781488418579102, + "learning_rate": 0.0001190865987373593, + "loss": 0.4445, + "step": 26269 + }, + { + "epoch": 1.822913052529318, + "grad_norm": 1.329730749130249, + "learning_rate": 0.0001190810833838103, + "loss": 0.6054, + "step": 26270 + }, + { + "epoch": 1.8229824439664144, + "grad_norm": 2.06199049949646, + "learning_rate": 0.00011907556797002455, + "loss": 0.5338, + "step": 26271 + }, + { + "epoch": 1.8230518354035112, + "grad_norm": 1.067585825920105, + "learning_rate": 0.00011907005249601946, + "loss": 0.4726, + "step": 26272 + }, + { + "epoch": 1.8231212268406078, + "grad_norm": 1.9631296396255493, + "learning_rate": 0.00011906453696181239, + "loss": 0.7362, + "step": 26273 + }, + { + "epoch": 1.8231906182777045, + "grad_norm": 1.8132967948913574, + "learning_rate": 0.0001190590213674208, + "loss": 0.5227, + "step": 26274 + }, + { + "epoch": 1.8232600097148013, + "grad_norm": 1.308119535446167, + "learning_rate": 0.00011905350571286209, + "loss": 0.4984, + "step": 26275 + }, + { + "epoch": 1.823329401151898, + "grad_norm": 2.0241520404815674, + "learning_rate": 0.0001190479899981537, + "loss": 0.8048, + "step": 26276 + }, + { + "epoch": 1.8233987925889945, + "grad_norm": 1.7107830047607422, + "learning_rate": 0.00011904247422331298, + "loss": 0.9717, + "step": 26277 + }, + { + "epoch": 1.8234681840260913, + "grad_norm": 1.5040777921676636, + "learning_rate": 0.0001190369583883574, + "loss": 0.5696, + "step": 26278 + }, + { + "epoch": 1.8235375754631877, + "grad_norm": 2.2616515159606934, + "learning_rate": 0.00011903144249330436, + "loss": 0.6526, + "step": 26279 + }, + { + "epoch": 1.8236069669002846, + "grad_norm": 1.8183650970458984, + "learning_rate": 0.00011902592653817124, + "loss": 0.785, + "step": 26280 + }, + { + "epoch": 1.8236763583373812, + "grad_norm": 1.7003693580627441, + "learning_rate": 0.00011902041052297547, + "loss": 0.6557, + "step": 26281 + }, + { + "epoch": 1.8237457497744778, + "grad_norm": 1.7846410274505615, + "learning_rate": 0.0001190148944477345, + "loss": 0.5559, + "step": 26282 + }, + { + "epoch": 1.8238151412115746, + "grad_norm": 1.706945538520813, + "learning_rate": 0.0001190093783124657, + "loss": 0.5464, + "step": 26283 + }, + { + "epoch": 1.823884532648671, + "grad_norm": 1.6064378023147583, + "learning_rate": 0.0001190038621171865, + "loss": 0.4819, + "step": 26284 + }, + { + "epoch": 1.8239539240857678, + "grad_norm": 1.4482073783874512, + "learning_rate": 0.00011899834586191435, + "loss": 0.5057, + "step": 26285 + }, + { + "epoch": 1.8240233155228645, + "grad_norm": 1.891282081604004, + "learning_rate": 0.00011899282954666655, + "loss": 0.6897, + "step": 26286 + }, + { + "epoch": 1.824092706959961, + "grad_norm": 1.3177884817123413, + "learning_rate": 0.00011898731317146065, + "loss": 0.5339, + "step": 26287 + }, + { + "epoch": 1.824162098397058, + "grad_norm": 1.3377540111541748, + "learning_rate": 0.00011898179673631397, + "loss": 0.4655, + "step": 26288 + }, + { + "epoch": 1.8242314898341545, + "grad_norm": 2.195754051208496, + "learning_rate": 0.00011897628024124398, + "loss": 0.55, + "step": 26289 + }, + { + "epoch": 1.8243008812712511, + "grad_norm": 1.6493041515350342, + "learning_rate": 0.0001189707636862681, + "loss": 0.4589, + "step": 26290 + }, + { + "epoch": 1.824370272708348, + "grad_norm": 1.537462592124939, + "learning_rate": 0.00011896524707140366, + "loss": 0.6437, + "step": 26291 + }, + { + "epoch": 1.8244396641454443, + "grad_norm": 2.116786241531372, + "learning_rate": 0.00011895973039666817, + "loss": 0.8225, + "step": 26292 + }, + { + "epoch": 1.8245090555825412, + "grad_norm": 1.7830811738967896, + "learning_rate": 0.00011895421366207902, + "loss": 0.7344, + "step": 26293 + }, + { + "epoch": 1.8245784470196378, + "grad_norm": 1.7123799324035645, + "learning_rate": 0.0001189486968676536, + "loss": 0.5763, + "step": 26294 + }, + { + "epoch": 1.8246478384567344, + "grad_norm": 1.4545283317565918, + "learning_rate": 0.00011894318001340933, + "loss": 0.5705, + "step": 26295 + }, + { + "epoch": 1.8247172298938312, + "grad_norm": 1.0512135028839111, + "learning_rate": 0.00011893766309936366, + "loss": 0.5489, + "step": 26296 + }, + { + "epoch": 1.8247866213309276, + "grad_norm": 1.6422995328903198, + "learning_rate": 0.00011893214612553399, + "loss": 0.6027, + "step": 26297 + }, + { + "epoch": 1.8248560127680244, + "grad_norm": 1.2085049152374268, + "learning_rate": 0.0001189266290919377, + "loss": 0.5902, + "step": 26298 + }, + { + "epoch": 1.824925404205121, + "grad_norm": 1.2077910900115967, + "learning_rate": 0.00011892111199859227, + "loss": 0.4593, + "step": 26299 + }, + { + "epoch": 1.8249947956422177, + "grad_norm": 1.6079131364822388, + "learning_rate": 0.00011891559484551507, + "loss": 0.5918, + "step": 26300 + }, + { + "epoch": 1.8250641870793145, + "grad_norm": 1.5501818656921387, + "learning_rate": 0.00011891007763272353, + "loss": 0.5099, + "step": 26301 + }, + { + "epoch": 1.8251335785164111, + "grad_norm": 1.1586780548095703, + "learning_rate": 0.0001189045603602351, + "loss": 0.4773, + "step": 26302 + }, + { + "epoch": 1.8252029699535077, + "grad_norm": 2.294156074523926, + "learning_rate": 0.00011889904302806711, + "loss": 0.8565, + "step": 26303 + }, + { + "epoch": 1.8252723613906046, + "grad_norm": 2.225724697113037, + "learning_rate": 0.00011889352563623708, + "loss": 0.7945, + "step": 26304 + }, + { + "epoch": 1.825341752827701, + "grad_norm": 1.9600601196289062, + "learning_rate": 0.00011888800818476236, + "loss": 0.733, + "step": 26305 + }, + { + "epoch": 1.8254111442647978, + "grad_norm": 1.1910765171051025, + "learning_rate": 0.0001188824906736604, + "loss": 0.4733, + "step": 26306 + }, + { + "epoch": 1.8254805357018944, + "grad_norm": 1.5019075870513916, + "learning_rate": 0.00011887697310294861, + "loss": 0.5337, + "step": 26307 + }, + { + "epoch": 1.825549927138991, + "grad_norm": 1.6536273956298828, + "learning_rate": 0.00011887145547264438, + "loss": 0.7273, + "step": 26308 + }, + { + "epoch": 1.8256193185760878, + "grad_norm": 1.4282433986663818, + "learning_rate": 0.00011886593778276517, + "loss": 0.4982, + "step": 26309 + }, + { + "epoch": 1.8256887100131842, + "grad_norm": 1.1262813806533813, + "learning_rate": 0.0001188604200333284, + "loss": 0.5023, + "step": 26310 + }, + { + "epoch": 1.825758101450281, + "grad_norm": 2.2247323989868164, + "learning_rate": 0.00011885490222435145, + "loss": 0.8895, + "step": 26311 + }, + { + "epoch": 1.8258274928873777, + "grad_norm": 1.6689623594284058, + "learning_rate": 0.00011884938435585175, + "loss": 0.5742, + "step": 26312 + }, + { + "epoch": 1.8258968843244743, + "grad_norm": 1.4082856178283691, + "learning_rate": 0.00011884386642784675, + "loss": 0.4971, + "step": 26313 + }, + { + "epoch": 1.8259662757615711, + "grad_norm": 1.9134644269943237, + "learning_rate": 0.00011883834844035386, + "loss": 0.6099, + "step": 26314 + }, + { + "epoch": 1.8260356671986677, + "grad_norm": 1.2498109340667725, + "learning_rate": 0.00011883283039339049, + "loss": 0.6592, + "step": 26315 + }, + { + "epoch": 1.8261050586357643, + "grad_norm": 1.6976792812347412, + "learning_rate": 0.00011882731228697404, + "loss": 0.6266, + "step": 26316 + }, + { + "epoch": 1.8261744500728612, + "grad_norm": 1.9496104717254639, + "learning_rate": 0.00011882179412112198, + "loss": 0.8487, + "step": 26317 + }, + { + "epoch": 1.8262438415099576, + "grad_norm": 1.6607482433319092, + "learning_rate": 0.00011881627589585167, + "loss": 0.6184, + "step": 26318 + }, + { + "epoch": 1.8263132329470544, + "grad_norm": 1.9293038845062256, + "learning_rate": 0.00011881075761118055, + "loss": 0.5542, + "step": 26319 + }, + { + "epoch": 1.826382624384151, + "grad_norm": 1.090752363204956, + "learning_rate": 0.00011880523926712608, + "loss": 0.4525, + "step": 26320 + }, + { + "epoch": 1.8264520158212476, + "grad_norm": 1.6998437643051147, + "learning_rate": 0.00011879972086370565, + "loss": 0.8676, + "step": 26321 + }, + { + "epoch": 1.8265214072583444, + "grad_norm": 2.4091503620147705, + "learning_rate": 0.00011879420240093668, + "loss": 0.5773, + "step": 26322 + }, + { + "epoch": 1.8265907986954408, + "grad_norm": 1.3966314792633057, + "learning_rate": 0.00011878868387883656, + "loss": 0.592, + "step": 26323 + }, + { + "epoch": 1.8266601901325377, + "grad_norm": 1.738827109336853, + "learning_rate": 0.0001187831652974228, + "loss": 0.6081, + "step": 26324 + }, + { + "epoch": 1.8267295815696343, + "grad_norm": 1.4717905521392822, + "learning_rate": 0.00011877764665671273, + "loss": 0.5717, + "step": 26325 + }, + { + "epoch": 1.8267989730067309, + "grad_norm": 1.625024676322937, + "learning_rate": 0.00011877212795672381, + "loss": 0.7263, + "step": 26326 + }, + { + "epoch": 1.8268683644438277, + "grad_norm": 1.3456075191497803, + "learning_rate": 0.00011876660919747347, + "loss": 0.5196, + "step": 26327 + }, + { + "epoch": 1.8269377558809243, + "grad_norm": 1.4024056196212769, + "learning_rate": 0.00011876109037897912, + "loss": 0.4833, + "step": 26328 + }, + { + "epoch": 1.827007147318021, + "grad_norm": 1.6914129257202148, + "learning_rate": 0.00011875557150125818, + "loss": 0.4792, + "step": 26329 + }, + { + "epoch": 1.8270765387551178, + "grad_norm": 2.669126510620117, + "learning_rate": 0.00011875005256432805, + "loss": 0.5635, + "step": 26330 + }, + { + "epoch": 1.8271459301922142, + "grad_norm": 1.7222105264663696, + "learning_rate": 0.00011874453356820623, + "loss": 0.4296, + "step": 26331 + }, + { + "epoch": 1.827215321629311, + "grad_norm": 1.8021522760391235, + "learning_rate": 0.00011873901451291007, + "loss": 0.6844, + "step": 26332 + }, + { + "epoch": 1.8272847130664076, + "grad_norm": 1.7513201236724854, + "learning_rate": 0.00011873349539845701, + "loss": 0.829, + "step": 26333 + }, + { + "epoch": 1.8273541045035042, + "grad_norm": 1.1841151714324951, + "learning_rate": 0.00011872797622486447, + "loss": 0.4415, + "step": 26334 + }, + { + "epoch": 1.827423495940601, + "grad_norm": 2.102039098739624, + "learning_rate": 0.00011872245699214993, + "loss": 0.4872, + "step": 26335 + }, + { + "epoch": 1.8274928873776974, + "grad_norm": 1.4916592836380005, + "learning_rate": 0.00011871693770033072, + "loss": 0.58, + "step": 26336 + }, + { + "epoch": 1.8275622788147943, + "grad_norm": 1.339314579963684, + "learning_rate": 0.00011871141834942433, + "loss": 0.595, + "step": 26337 + }, + { + "epoch": 1.8276316702518909, + "grad_norm": 2.0031678676605225, + "learning_rate": 0.00011870589893944814, + "loss": 0.6254, + "step": 26338 + }, + { + "epoch": 1.8277010616889875, + "grad_norm": 2.291049003601074, + "learning_rate": 0.00011870037947041963, + "loss": 0.7959, + "step": 26339 + }, + { + "epoch": 1.8277704531260843, + "grad_norm": 1.4528366327285767, + "learning_rate": 0.00011869485994235616, + "loss": 0.5996, + "step": 26340 + }, + { + "epoch": 1.827839844563181, + "grad_norm": 1.6009689569473267, + "learning_rate": 0.00011868934035527522, + "loss": 0.5485, + "step": 26341 + }, + { + "epoch": 1.8279092360002775, + "grad_norm": 1.4532264471054077, + "learning_rate": 0.00011868382070919417, + "loss": 0.6623, + "step": 26342 + }, + { + "epoch": 1.8279786274373744, + "grad_norm": 1.6294194459915161, + "learning_rate": 0.00011867830100413047, + "loss": 0.5416, + "step": 26343 + }, + { + "epoch": 1.8280480188744708, + "grad_norm": 1.3932009935379028, + "learning_rate": 0.00011867278124010155, + "loss": 0.5515, + "step": 26344 + }, + { + "epoch": 1.8281174103115676, + "grad_norm": 2.389801502227783, + "learning_rate": 0.00011866726141712483, + "loss": 0.5904, + "step": 26345 + }, + { + "epoch": 1.8281868017486642, + "grad_norm": 2.1769375801086426, + "learning_rate": 0.00011866174153521772, + "loss": 0.6236, + "step": 26346 + }, + { + "epoch": 1.8282561931857608, + "grad_norm": 2.1241748332977295, + "learning_rate": 0.00011865622159439764, + "loss": 0.5177, + "step": 26347 + }, + { + "epoch": 1.8283255846228577, + "grad_norm": 1.454065203666687, + "learning_rate": 0.00011865070159468207, + "loss": 0.5466, + "step": 26348 + }, + { + "epoch": 1.828394976059954, + "grad_norm": 1.6568169593811035, + "learning_rate": 0.0001186451815360884, + "loss": 0.576, + "step": 26349 + }, + { + "epoch": 1.8284643674970509, + "grad_norm": 1.1424304246902466, + "learning_rate": 0.00011863966141863403, + "loss": 0.4012, + "step": 26350 + }, + { + "epoch": 1.8285337589341475, + "grad_norm": 1.5615229606628418, + "learning_rate": 0.0001186341412423364, + "loss": 0.6106, + "step": 26351 + }, + { + "epoch": 1.828603150371244, + "grad_norm": 2.4140117168426514, + "learning_rate": 0.000118628621007213, + "loss": 0.58, + "step": 26352 + }, + { + "epoch": 1.828672541808341, + "grad_norm": 1.3951081037521362, + "learning_rate": 0.00011862310071328116, + "loss": 0.5413, + "step": 26353 + }, + { + "epoch": 1.8287419332454375, + "grad_norm": 1.1565735340118408, + "learning_rate": 0.00011861758036055836, + "loss": 0.4453, + "step": 26354 + }, + { + "epoch": 1.8288113246825342, + "grad_norm": 1.3888365030288696, + "learning_rate": 0.00011861205994906201, + "loss": 0.7109, + "step": 26355 + }, + { + "epoch": 1.828880716119631, + "grad_norm": 1.2359732389450073, + "learning_rate": 0.0001186065394788096, + "loss": 0.559, + "step": 26356 + }, + { + "epoch": 1.8289501075567274, + "grad_norm": 5.330831527709961, + "learning_rate": 0.00011860101894981845, + "loss": 0.9353, + "step": 26357 + }, + { + "epoch": 1.8290194989938242, + "grad_norm": 1.600558876991272, + "learning_rate": 0.00011859549836210604, + "loss": 0.8146, + "step": 26358 + }, + { + "epoch": 1.8290888904309208, + "grad_norm": 2.0071465969085693, + "learning_rate": 0.00011858997771568982, + "loss": 0.5219, + "step": 26359 + }, + { + "epoch": 1.8291582818680174, + "grad_norm": 1.8293898105621338, + "learning_rate": 0.0001185844570105872, + "loss": 0.6211, + "step": 26360 + }, + { + "epoch": 1.8292276733051143, + "grad_norm": 1.4003633260726929, + "learning_rate": 0.00011857893624681558, + "loss": 0.5795, + "step": 26361 + }, + { + "epoch": 1.8292970647422107, + "grad_norm": 1.8382912874221802, + "learning_rate": 0.00011857341542439244, + "loss": 0.5887, + "step": 26362 + }, + { + "epoch": 1.8293664561793075, + "grad_norm": 1.9018268585205078, + "learning_rate": 0.00011856789454333516, + "loss": 0.7609, + "step": 26363 + }, + { + "epoch": 1.829435847616404, + "grad_norm": 1.3999170064926147, + "learning_rate": 0.0001185623736036612, + "loss": 0.5461, + "step": 26364 + }, + { + "epoch": 1.8295052390535007, + "grad_norm": 1.5046619176864624, + "learning_rate": 0.00011855685260538798, + "loss": 0.5844, + "step": 26365 + }, + { + "epoch": 1.8295746304905975, + "grad_norm": 1.6996245384216309, + "learning_rate": 0.00011855133154853296, + "loss": 0.7193, + "step": 26366 + }, + { + "epoch": 1.8296440219276942, + "grad_norm": 1.6924453973770142, + "learning_rate": 0.0001185458104331135, + "loss": 0.6069, + "step": 26367 + }, + { + "epoch": 1.8297134133647908, + "grad_norm": 1.7874027490615845, + "learning_rate": 0.00011854028925914706, + "loss": 0.4797, + "step": 26368 + }, + { + "epoch": 1.8297828048018876, + "grad_norm": 1.7190792560577393, + "learning_rate": 0.00011853476802665113, + "loss": 0.4149, + "step": 26369 + }, + { + "epoch": 1.829852196238984, + "grad_norm": 1.4446431398391724, + "learning_rate": 0.00011852924673564306, + "loss": 0.5024, + "step": 26370 + }, + { + "epoch": 1.8299215876760808, + "grad_norm": 1.8115893602371216, + "learning_rate": 0.00011852372538614028, + "loss": 0.7355, + "step": 26371 + }, + { + "epoch": 1.8299909791131774, + "grad_norm": 2.7035458087921143, + "learning_rate": 0.0001185182039781603, + "loss": 0.6497, + "step": 26372 + }, + { + "epoch": 1.830060370550274, + "grad_norm": 1.3962244987487793, + "learning_rate": 0.00011851268251172045, + "loss": 0.4123, + "step": 26373 + }, + { + "epoch": 1.8301297619873709, + "grad_norm": 1.4327139854431152, + "learning_rate": 0.00011850716098683825, + "loss": 0.5822, + "step": 26374 + }, + { + "epoch": 1.8301991534244673, + "grad_norm": 1.301762580871582, + "learning_rate": 0.00011850163940353105, + "loss": 0.5503, + "step": 26375 + }, + { + "epoch": 1.830268544861564, + "grad_norm": 1.601704478263855, + "learning_rate": 0.00011849611776181637, + "loss": 0.6029, + "step": 26376 + }, + { + "epoch": 1.8303379362986607, + "grad_norm": 2.2342724800109863, + "learning_rate": 0.00011849059606171153, + "loss": 0.6459, + "step": 26377 + }, + { + "epoch": 1.8304073277357573, + "grad_norm": 1.8369660377502441, + "learning_rate": 0.00011848507430323406, + "loss": 0.4273, + "step": 26378 + }, + { + "epoch": 1.8304767191728541, + "grad_norm": 1.4136114120483398, + "learning_rate": 0.00011847955248640137, + "loss": 0.5055, + "step": 26379 + }, + { + "epoch": 1.8305461106099508, + "grad_norm": 1.4368830919265747, + "learning_rate": 0.00011847403061123085, + "loss": 0.6086, + "step": 26380 + }, + { + "epoch": 1.8306155020470474, + "grad_norm": 1.8661855459213257, + "learning_rate": 0.00011846850867773995, + "loss": 0.7528, + "step": 26381 + }, + { + "epoch": 1.8306848934841442, + "grad_norm": 1.3919615745544434, + "learning_rate": 0.00011846298668594613, + "loss": 0.3838, + "step": 26382 + }, + { + "epoch": 1.8307542849212406, + "grad_norm": 1.6252799034118652, + "learning_rate": 0.00011845746463586682, + "loss": 0.6285, + "step": 26383 + }, + { + "epoch": 1.8308236763583374, + "grad_norm": 1.7283294200897217, + "learning_rate": 0.0001184519425275194, + "loss": 0.4909, + "step": 26384 + }, + { + "epoch": 1.830893067795434, + "grad_norm": 1.625260591506958, + "learning_rate": 0.00011844642036092135, + "loss": 0.638, + "step": 26385 + }, + { + "epoch": 1.8309624592325306, + "grad_norm": 1.5821545124053955, + "learning_rate": 0.00011844089813609011, + "loss": 0.5068, + "step": 26386 + }, + { + "epoch": 1.8310318506696275, + "grad_norm": 1.578779697418213, + "learning_rate": 0.00011843537585304308, + "loss": 0.5136, + "step": 26387 + }, + { + "epoch": 1.8311012421067239, + "grad_norm": 1.5289685726165771, + "learning_rate": 0.0001184298535117977, + "loss": 0.8648, + "step": 26388 + }, + { + "epoch": 1.8311706335438207, + "grad_norm": 1.3357868194580078, + "learning_rate": 0.00011842433111237144, + "loss": 0.5146, + "step": 26389 + }, + { + "epoch": 1.8312400249809173, + "grad_norm": 1.3684792518615723, + "learning_rate": 0.00011841880865478165, + "loss": 0.5644, + "step": 26390 + }, + { + "epoch": 1.831309416418014, + "grad_norm": 2.7282845973968506, + "learning_rate": 0.00011841328613904588, + "loss": 0.8418, + "step": 26391 + }, + { + "epoch": 1.8313788078551108, + "grad_norm": 1.7479236125946045, + "learning_rate": 0.00011840776356518147, + "loss": 0.7103, + "step": 26392 + }, + { + "epoch": 1.8314481992922074, + "grad_norm": 1.2101472616195679, + "learning_rate": 0.00011840224093320589, + "loss": 0.4901, + "step": 26393 + }, + { + "epoch": 1.831517590729304, + "grad_norm": 1.5853326320648193, + "learning_rate": 0.00011839671824313657, + "loss": 0.6023, + "step": 26394 + }, + { + "epoch": 1.8315869821664008, + "grad_norm": 1.9132159948349, + "learning_rate": 0.00011839119549499094, + "loss": 0.7258, + "step": 26395 + }, + { + "epoch": 1.8316563736034972, + "grad_norm": 2.340251922607422, + "learning_rate": 0.00011838567268878641, + "loss": 0.6467, + "step": 26396 + }, + { + "epoch": 1.831725765040594, + "grad_norm": 1.6739646196365356, + "learning_rate": 0.00011838014982454051, + "loss": 0.5662, + "step": 26397 + }, + { + "epoch": 1.8317951564776906, + "grad_norm": 1.0880463123321533, + "learning_rate": 0.00011837462690227058, + "loss": 0.5031, + "step": 26398 + }, + { + "epoch": 1.8318645479147873, + "grad_norm": 1.3274215459823608, + "learning_rate": 0.00011836910392199408, + "loss": 0.5177, + "step": 26399 + }, + { + "epoch": 1.831933939351884, + "grad_norm": 1.5915827751159668, + "learning_rate": 0.00011836358088372845, + "loss": 0.5353, + "step": 26400 + }, + { + "epoch": 1.8320033307889805, + "grad_norm": 1.8319358825683594, + "learning_rate": 0.00011835805778749112, + "loss": 0.7454, + "step": 26401 + }, + { + "epoch": 1.8320727222260773, + "grad_norm": 1.2969359159469604, + "learning_rate": 0.00011835253463329954, + "loss": 0.4599, + "step": 26402 + }, + { + "epoch": 1.832142113663174, + "grad_norm": 1.5699257850646973, + "learning_rate": 0.00011834701142117112, + "loss": 0.5943, + "step": 26403 + }, + { + "epoch": 1.8322115051002705, + "grad_norm": 2.1499762535095215, + "learning_rate": 0.00011834148815112336, + "loss": 0.3952, + "step": 26404 + }, + { + "epoch": 1.8322808965373674, + "grad_norm": 1.3624663352966309, + "learning_rate": 0.0001183359648231736, + "loss": 0.5183, + "step": 26405 + }, + { + "epoch": 1.832350287974464, + "grad_norm": 1.874550223350525, + "learning_rate": 0.00011833044143733933, + "loss": 0.7037, + "step": 26406 + }, + { + "epoch": 1.8324196794115606, + "grad_norm": 1.7000229358673096, + "learning_rate": 0.00011832491799363803, + "loss": 0.4063, + "step": 26407 + }, + { + "epoch": 1.8324890708486574, + "grad_norm": 1.395063042640686, + "learning_rate": 0.00011831939449208704, + "loss": 0.4068, + "step": 26408 + }, + { + "epoch": 1.8325584622857538, + "grad_norm": 1.4305791854858398, + "learning_rate": 0.0001183138709327039, + "loss": 0.5432, + "step": 26409 + }, + { + "epoch": 1.8326278537228506, + "grad_norm": 1.8491376638412476, + "learning_rate": 0.00011830834731550595, + "loss": 0.5955, + "step": 26410 + }, + { + "epoch": 1.8326972451599473, + "grad_norm": 1.9213383197784424, + "learning_rate": 0.00011830282364051069, + "loss": 0.5443, + "step": 26411 + }, + { + "epoch": 1.8327666365970439, + "grad_norm": 1.9847614765167236, + "learning_rate": 0.00011829729990773552, + "loss": 0.5847, + "step": 26412 + }, + { + "epoch": 1.8328360280341407, + "grad_norm": 1.5564417839050293, + "learning_rate": 0.0001182917761171979, + "loss": 0.6274, + "step": 26413 + }, + { + "epoch": 1.832905419471237, + "grad_norm": 1.4441988468170166, + "learning_rate": 0.0001182862522689153, + "loss": 0.4184, + "step": 26414 + }, + { + "epoch": 1.832974810908334, + "grad_norm": 1.4158657789230347, + "learning_rate": 0.00011828072836290509, + "loss": 0.7899, + "step": 26415 + }, + { + "epoch": 1.8330442023454305, + "grad_norm": 1.8296998739242554, + "learning_rate": 0.00011827520439918474, + "loss": 0.7245, + "step": 26416 + }, + { + "epoch": 1.8331135937825271, + "grad_norm": 1.8192529678344727, + "learning_rate": 0.0001182696803777717, + "loss": 0.7073, + "step": 26417 + }, + { + "epoch": 1.833182985219624, + "grad_norm": 2.0190868377685547, + "learning_rate": 0.00011826415629868339, + "loss": 0.5691, + "step": 26418 + }, + { + "epoch": 1.8332523766567206, + "grad_norm": 1.463835597038269, + "learning_rate": 0.00011825863216193726, + "loss": 0.3518, + "step": 26419 + }, + { + "epoch": 1.8333217680938172, + "grad_norm": 1.3066470623016357, + "learning_rate": 0.00011825310796755076, + "loss": 0.4802, + "step": 26420 + }, + { + "epoch": 1.833391159530914, + "grad_norm": 2.58027982711792, + "learning_rate": 0.00011824758371554132, + "loss": 0.5951, + "step": 26421 + }, + { + "epoch": 1.8334605509680104, + "grad_norm": 2.067443370819092, + "learning_rate": 0.00011824205940592637, + "loss": 0.7584, + "step": 26422 + }, + { + "epoch": 1.8335299424051072, + "grad_norm": 1.3170528411865234, + "learning_rate": 0.00011823653503872331, + "loss": 0.42, + "step": 26423 + }, + { + "epoch": 1.8335993338422039, + "grad_norm": 1.3245017528533936, + "learning_rate": 0.0001182310106139497, + "loss": 0.5836, + "step": 26424 + }, + { + "epoch": 1.8336687252793005, + "grad_norm": 1.2553287744522095, + "learning_rate": 0.00011822548613162286, + "loss": 0.4245, + "step": 26425 + }, + { + "epoch": 1.8337381167163973, + "grad_norm": 1.6086255311965942, + "learning_rate": 0.0001182199615917603, + "loss": 0.6109, + "step": 26426 + }, + { + "epoch": 1.8338075081534937, + "grad_norm": 1.3277502059936523, + "learning_rate": 0.00011821443699437939, + "loss": 0.4968, + "step": 26427 + }, + { + "epoch": 1.8338768995905905, + "grad_norm": 1.3661584854125977, + "learning_rate": 0.00011820891233949767, + "loss": 0.5645, + "step": 26428 + }, + { + "epoch": 1.8339462910276871, + "grad_norm": 1.2319496870040894, + "learning_rate": 0.0001182033876271325, + "loss": 0.4804, + "step": 26429 + }, + { + "epoch": 1.8340156824647837, + "grad_norm": 1.214240550994873, + "learning_rate": 0.00011819786285730132, + "loss": 0.4409, + "step": 26430 + }, + { + "epoch": 1.8340850739018806, + "grad_norm": 1.8901447057724, + "learning_rate": 0.00011819233803002165, + "loss": 0.5581, + "step": 26431 + }, + { + "epoch": 1.8341544653389772, + "grad_norm": 1.3933846950531006, + "learning_rate": 0.00011818681314531083, + "loss": 0.5393, + "step": 26432 + }, + { + "epoch": 1.8342238567760738, + "grad_norm": 1.0622469186782837, + "learning_rate": 0.00011818128820318639, + "loss": 0.4813, + "step": 26433 + }, + { + "epoch": 1.8342932482131706, + "grad_norm": 1.444359302520752, + "learning_rate": 0.0001181757632036657, + "loss": 0.5427, + "step": 26434 + }, + { + "epoch": 1.834362639650267, + "grad_norm": 1.673293948173523, + "learning_rate": 0.00011817023814676627, + "loss": 0.4861, + "step": 26435 + }, + { + "epoch": 1.8344320310873639, + "grad_norm": 1.5255025625228882, + "learning_rate": 0.00011816471303250549, + "loss": 0.6077, + "step": 26436 + }, + { + "epoch": 1.8345014225244605, + "grad_norm": 1.5963258743286133, + "learning_rate": 0.0001181591878609008, + "loss": 0.561, + "step": 26437 + }, + { + "epoch": 1.834570813961557, + "grad_norm": 1.8712974786758423, + "learning_rate": 0.0001181536626319697, + "loss": 0.6761, + "step": 26438 + }, + { + "epoch": 1.834640205398654, + "grad_norm": 1.339341640472412, + "learning_rate": 0.00011814813734572956, + "loss": 0.3848, + "step": 26439 + }, + { + "epoch": 1.8347095968357505, + "grad_norm": 2.309244155883789, + "learning_rate": 0.00011814261200219787, + "loss": 0.7545, + "step": 26440 + }, + { + "epoch": 1.8347789882728471, + "grad_norm": 2.329322099685669, + "learning_rate": 0.00011813708660139204, + "loss": 0.708, + "step": 26441 + }, + { + "epoch": 1.8348483797099437, + "grad_norm": 1.4908360242843628, + "learning_rate": 0.00011813156114332957, + "loss": 0.4748, + "step": 26442 + }, + { + "epoch": 1.8349177711470404, + "grad_norm": 1.086230754852295, + "learning_rate": 0.00011812603562802784, + "loss": 0.441, + "step": 26443 + }, + { + "epoch": 1.8349871625841372, + "grad_norm": 2.1315231323242188, + "learning_rate": 0.00011812051005550433, + "loss": 0.4929, + "step": 26444 + }, + { + "epoch": 1.8350565540212338, + "grad_norm": 1.198751449584961, + "learning_rate": 0.00011811498442577647, + "loss": 0.5593, + "step": 26445 + }, + { + "epoch": 1.8351259454583304, + "grad_norm": 1.5429015159606934, + "learning_rate": 0.0001181094587388617, + "loss": 0.4749, + "step": 26446 + }, + { + "epoch": 1.8351953368954272, + "grad_norm": 1.7900334596633911, + "learning_rate": 0.00011810393299477747, + "loss": 0.6732, + "step": 26447 + }, + { + "epoch": 1.8352647283325236, + "grad_norm": 1.3874455690383911, + "learning_rate": 0.0001180984071935412, + "loss": 0.4823, + "step": 26448 + }, + { + "epoch": 1.8353341197696205, + "grad_norm": 1.898866891860962, + "learning_rate": 0.0001180928813351704, + "loss": 0.7968, + "step": 26449 + }, + { + "epoch": 1.835403511206717, + "grad_norm": 1.0864920616149902, + "learning_rate": 0.00011808735541968245, + "loss": 0.482, + "step": 26450 + }, + { + "epoch": 1.8354729026438137, + "grad_norm": 1.5602329969406128, + "learning_rate": 0.0001180818294470948, + "loss": 0.4395, + "step": 26451 + }, + { + "epoch": 1.8355422940809105, + "grad_norm": 1.5763216018676758, + "learning_rate": 0.00011807630341742495, + "loss": 0.6306, + "step": 26452 + }, + { + "epoch": 1.8356116855180071, + "grad_norm": 1.6359202861785889, + "learning_rate": 0.0001180707773306903, + "loss": 0.5747, + "step": 26453 + }, + { + "epoch": 1.8356810769551037, + "grad_norm": 1.5182360410690308, + "learning_rate": 0.00011806525118690827, + "loss": 0.4663, + "step": 26454 + }, + { + "epoch": 1.8357504683922004, + "grad_norm": 1.5426031351089478, + "learning_rate": 0.00011805972498609639, + "loss": 0.5518, + "step": 26455 + }, + { + "epoch": 1.835819859829297, + "grad_norm": 3.3632466793060303, + "learning_rate": 0.00011805419872827202, + "loss": 0.7627, + "step": 26456 + }, + { + "epoch": 1.8358892512663938, + "grad_norm": 1.987576961517334, + "learning_rate": 0.00011804867241345263, + "loss": 0.7374, + "step": 26457 + }, + { + "epoch": 1.8359586427034904, + "grad_norm": 1.8378963470458984, + "learning_rate": 0.00011804314604165569, + "loss": 0.4685, + "step": 26458 + }, + { + "epoch": 1.836028034140587, + "grad_norm": 1.2090859413146973, + "learning_rate": 0.00011803761961289865, + "loss": 0.421, + "step": 26459 + }, + { + "epoch": 1.8360974255776839, + "grad_norm": 1.6606045961380005, + "learning_rate": 0.00011803209312719889, + "loss": 0.5789, + "step": 26460 + }, + { + "epoch": 1.8361668170147802, + "grad_norm": 1.4127699136734009, + "learning_rate": 0.00011802656658457396, + "loss": 0.5471, + "step": 26461 + }, + { + "epoch": 1.836236208451877, + "grad_norm": 1.6996365785598755, + "learning_rate": 0.00011802103998504121, + "loss": 0.6484, + "step": 26462 + }, + { + "epoch": 1.8363055998889737, + "grad_norm": 1.3906677961349487, + "learning_rate": 0.00011801551332861818, + "loss": 0.3733, + "step": 26463 + }, + { + "epoch": 1.8363749913260703, + "grad_norm": 2.0799031257629395, + "learning_rate": 0.00011800998661532221, + "loss": 0.7812, + "step": 26464 + }, + { + "epoch": 1.8364443827631671, + "grad_norm": 1.4417511224746704, + "learning_rate": 0.00011800445984517082, + "loss": 0.4888, + "step": 26465 + }, + { + "epoch": 1.8365137742002637, + "grad_norm": 1.7452203035354614, + "learning_rate": 0.00011799893301818144, + "loss": 0.5675, + "step": 26466 + }, + { + "epoch": 1.8365831656373603, + "grad_norm": 2.41618013381958, + "learning_rate": 0.00011799340613437152, + "loss": 0.6488, + "step": 26467 + }, + { + "epoch": 1.836652557074457, + "grad_norm": 0.9494455456733704, + "learning_rate": 0.0001179878791937585, + "loss": 0.3253, + "step": 26468 + }, + { + "epoch": 1.8367219485115536, + "grad_norm": 1.5050922632217407, + "learning_rate": 0.00011798235219635983, + "loss": 0.6958, + "step": 26469 + }, + { + "epoch": 1.8367913399486504, + "grad_norm": 1.348105788230896, + "learning_rate": 0.00011797682514219295, + "loss": 0.5678, + "step": 26470 + }, + { + "epoch": 1.836860731385747, + "grad_norm": 1.8550549745559692, + "learning_rate": 0.00011797129803127532, + "loss": 0.7142, + "step": 26471 + }, + { + "epoch": 1.8369301228228436, + "grad_norm": 1.3900272846221924, + "learning_rate": 0.0001179657708636244, + "loss": 0.5132, + "step": 26472 + }, + { + "epoch": 1.8369995142599405, + "grad_norm": 1.4874134063720703, + "learning_rate": 0.00011796024363925762, + "loss": 0.5749, + "step": 26473 + }, + { + "epoch": 1.8370689056970368, + "grad_norm": 1.4201009273529053, + "learning_rate": 0.00011795471635819244, + "loss": 0.4455, + "step": 26474 + }, + { + "epoch": 1.8371382971341337, + "grad_norm": 1.6612420082092285, + "learning_rate": 0.0001179491890204463, + "loss": 0.8065, + "step": 26475 + }, + { + "epoch": 1.8372076885712303, + "grad_norm": 1.3779304027557373, + "learning_rate": 0.00011794366162603665, + "loss": 0.4721, + "step": 26476 + }, + { + "epoch": 1.837277080008327, + "grad_norm": 1.2667272090911865, + "learning_rate": 0.00011793813417498093, + "loss": 0.4838, + "step": 26477 + }, + { + "epoch": 1.8373464714454237, + "grad_norm": 2.033750295639038, + "learning_rate": 0.00011793260666729663, + "loss": 0.4998, + "step": 26478 + }, + { + "epoch": 1.8374158628825203, + "grad_norm": 1.2880197763442993, + "learning_rate": 0.00011792707910300118, + "loss": 0.5232, + "step": 26479 + }, + { + "epoch": 1.837485254319617, + "grad_norm": 1.2880889177322388, + "learning_rate": 0.00011792155148211201, + "loss": 0.5015, + "step": 26480 + }, + { + "epoch": 1.8375546457567136, + "grad_norm": 1.378530502319336, + "learning_rate": 0.00011791602380464657, + "loss": 0.6299, + "step": 26481 + }, + { + "epoch": 1.8376240371938102, + "grad_norm": 2.4737343788146973, + "learning_rate": 0.0001179104960706223, + "loss": 0.6897, + "step": 26482 + }, + { + "epoch": 1.837693428630907, + "grad_norm": 1.2765111923217773, + "learning_rate": 0.00011790496828005672, + "loss": 0.4772, + "step": 26483 + }, + { + "epoch": 1.8377628200680036, + "grad_norm": 1.416549801826477, + "learning_rate": 0.00011789944043296719, + "loss": 0.5279, + "step": 26484 + }, + { + "epoch": 1.8378322115051002, + "grad_norm": 1.2446650266647339, + "learning_rate": 0.0001178939125293712, + "loss": 0.5859, + "step": 26485 + }, + { + "epoch": 1.837901602942197, + "grad_norm": 1.402205467224121, + "learning_rate": 0.00011788838456928624, + "loss": 0.4521, + "step": 26486 + }, + { + "epoch": 1.8379709943792935, + "grad_norm": 1.6717877388000488, + "learning_rate": 0.00011788285655272973, + "loss": 0.5735, + "step": 26487 + }, + { + "epoch": 1.8380403858163903, + "grad_norm": 1.4869835376739502, + "learning_rate": 0.00011787732847971909, + "loss": 0.4452, + "step": 26488 + }, + { + "epoch": 1.838109777253487, + "grad_norm": 1.5451877117156982, + "learning_rate": 0.00011787180035027178, + "loss": 0.5223, + "step": 26489 + }, + { + "epoch": 1.8381791686905835, + "grad_norm": 1.18710458278656, + "learning_rate": 0.00011786627216440531, + "loss": 0.3179, + "step": 26490 + }, + { + "epoch": 1.8382485601276803, + "grad_norm": 1.209546685218811, + "learning_rate": 0.00011786074392213708, + "loss": 0.7957, + "step": 26491 + }, + { + "epoch": 1.838317951564777, + "grad_norm": 2.028561592102051, + "learning_rate": 0.00011785521562348454, + "loss": 0.8511, + "step": 26492 + }, + { + "epoch": 1.8383873430018736, + "grad_norm": 1.4422932863235474, + "learning_rate": 0.00011784968726846516, + "loss": 0.4859, + "step": 26493 + }, + { + "epoch": 1.8384567344389704, + "grad_norm": 1.8228384256362915, + "learning_rate": 0.00011784415885709641, + "loss": 0.6898, + "step": 26494 + }, + { + "epoch": 1.8385261258760668, + "grad_norm": 1.612877607345581, + "learning_rate": 0.00011783863038939569, + "loss": 0.6018, + "step": 26495 + }, + { + "epoch": 1.8385955173131636, + "grad_norm": 1.315978765487671, + "learning_rate": 0.00011783310186538052, + "loss": 0.5284, + "step": 26496 + }, + { + "epoch": 1.8386649087502602, + "grad_norm": 1.8641400337219238, + "learning_rate": 0.00011782757328506831, + "loss": 0.8657, + "step": 26497 + }, + { + "epoch": 1.8387343001873568, + "grad_norm": 1.5107245445251465, + "learning_rate": 0.00011782204464847651, + "loss": 0.5004, + "step": 26498 + }, + { + "epoch": 1.8388036916244537, + "grad_norm": 1.5460988283157349, + "learning_rate": 0.00011781651595562256, + "loss": 0.3618, + "step": 26499 + }, + { + "epoch": 1.83887308306155, + "grad_norm": 1.2963279485702515, + "learning_rate": 0.00011781098720652396, + "loss": 0.5975, + "step": 26500 + }, + { + "epoch": 1.838942474498647, + "grad_norm": 1.3525570631027222, + "learning_rate": 0.00011780545840119812, + "loss": 0.4928, + "step": 26501 + }, + { + "epoch": 1.8390118659357435, + "grad_norm": 1.6232510805130005, + "learning_rate": 0.00011779992953966253, + "loss": 0.6246, + "step": 26502 + }, + { + "epoch": 1.8390812573728401, + "grad_norm": 1.2948417663574219, + "learning_rate": 0.00011779440062193461, + "loss": 0.4041, + "step": 26503 + }, + { + "epoch": 1.839150648809937, + "grad_norm": 1.6975198984146118, + "learning_rate": 0.00011778887164803187, + "loss": 0.6246, + "step": 26504 + }, + { + "epoch": 1.8392200402470336, + "grad_norm": 1.172488808631897, + "learning_rate": 0.00011778334261797171, + "loss": 0.4477, + "step": 26505 + }, + { + "epoch": 1.8392894316841302, + "grad_norm": 1.5760639905929565, + "learning_rate": 0.00011777781353177158, + "loss": 0.5496, + "step": 26506 + }, + { + "epoch": 1.839358823121227, + "grad_norm": 1.4832868576049805, + "learning_rate": 0.00011777228438944896, + "loss": 0.5297, + "step": 26507 + }, + { + "epoch": 1.8394282145583234, + "grad_norm": 2.200152635574341, + "learning_rate": 0.00011776675519102131, + "loss": 0.6234, + "step": 26508 + }, + { + "epoch": 1.8394976059954202, + "grad_norm": 1.4974877834320068, + "learning_rate": 0.00011776122593650608, + "loss": 0.6034, + "step": 26509 + }, + { + "epoch": 1.8395669974325168, + "grad_norm": 1.814011812210083, + "learning_rate": 0.0001177556966259207, + "loss": 0.6541, + "step": 26510 + }, + { + "epoch": 1.8396363888696134, + "grad_norm": 1.99647855758667, + "learning_rate": 0.00011775016725928268, + "loss": 0.5763, + "step": 26511 + }, + { + "epoch": 1.8397057803067103, + "grad_norm": 1.4987579584121704, + "learning_rate": 0.0001177446378366094, + "loss": 0.5855, + "step": 26512 + }, + { + "epoch": 1.8397751717438067, + "grad_norm": 1.3895167112350464, + "learning_rate": 0.00011773910835791839, + "loss": 0.556, + "step": 26513 + }, + { + "epoch": 1.8398445631809035, + "grad_norm": 1.2456566095352173, + "learning_rate": 0.00011773357882322707, + "loss": 0.5825, + "step": 26514 + }, + { + "epoch": 1.8399139546180001, + "grad_norm": 2.290006160736084, + "learning_rate": 0.0001177280492325529, + "loss": 0.5428, + "step": 26515 + }, + { + "epoch": 1.8399833460550967, + "grad_norm": 1.7090046405792236, + "learning_rate": 0.00011772251958591331, + "loss": 0.5283, + "step": 26516 + }, + { + "epoch": 1.8400527374921936, + "grad_norm": 1.2026941776275635, + "learning_rate": 0.0001177169898833258, + "loss": 0.3472, + "step": 26517 + }, + { + "epoch": 1.8401221289292902, + "grad_norm": 1.596773386001587, + "learning_rate": 0.0001177114601248078, + "loss": 0.6826, + "step": 26518 + }, + { + "epoch": 1.8401915203663868, + "grad_norm": 1.8585693836212158, + "learning_rate": 0.0001177059303103768, + "loss": 0.616, + "step": 26519 + }, + { + "epoch": 1.8402609118034836, + "grad_norm": 1.8322882652282715, + "learning_rate": 0.00011770040044005019, + "loss": 0.6255, + "step": 26520 + }, + { + "epoch": 1.84033030324058, + "grad_norm": 1.9356611967086792, + "learning_rate": 0.00011769487051384551, + "loss": 0.612, + "step": 26521 + }, + { + "epoch": 1.8403996946776768, + "grad_norm": 1.3782525062561035, + "learning_rate": 0.00011768934053178016, + "loss": 0.5286, + "step": 26522 + }, + { + "epoch": 1.8404690861147734, + "grad_norm": 1.0670921802520752, + "learning_rate": 0.0001176838104938716, + "loss": 0.4126, + "step": 26523 + }, + { + "epoch": 1.84053847755187, + "grad_norm": 1.217563509941101, + "learning_rate": 0.00011767828040013731, + "loss": 0.4991, + "step": 26524 + }, + { + "epoch": 1.8406078689889669, + "grad_norm": 1.5485464334487915, + "learning_rate": 0.00011767275025059477, + "loss": 0.4654, + "step": 26525 + }, + { + "epoch": 1.8406772604260633, + "grad_norm": 3.1545629501342773, + "learning_rate": 0.00011766722004526136, + "loss": 0.5523, + "step": 26526 + }, + { + "epoch": 1.84074665186316, + "grad_norm": 1.5529004335403442, + "learning_rate": 0.00011766168978415461, + "loss": 0.5567, + "step": 26527 + }, + { + "epoch": 1.8408160433002567, + "grad_norm": 1.3408600091934204, + "learning_rate": 0.00011765615946729198, + "loss": 0.4123, + "step": 26528 + }, + { + "epoch": 1.8408854347373533, + "grad_norm": 1.6081639528274536, + "learning_rate": 0.00011765062909469086, + "loss": 0.6185, + "step": 26529 + }, + { + "epoch": 1.8409548261744502, + "grad_norm": 1.260459542274475, + "learning_rate": 0.00011764509866636877, + "loss": 0.4757, + "step": 26530 + }, + { + "epoch": 1.8410242176115468, + "grad_norm": 1.9785314798355103, + "learning_rate": 0.00011763956818234316, + "loss": 0.6421, + "step": 26531 + }, + { + "epoch": 1.8410936090486434, + "grad_norm": 1.9209364652633667, + "learning_rate": 0.00011763403764263148, + "loss": 0.6605, + "step": 26532 + }, + { + "epoch": 1.8411630004857402, + "grad_norm": 1.5550085306167603, + "learning_rate": 0.00011762850704725117, + "loss": 0.5557, + "step": 26533 + }, + { + "epoch": 1.8412323919228366, + "grad_norm": 1.1659883260726929, + "learning_rate": 0.00011762297639621971, + "loss": 0.3906, + "step": 26534 + }, + { + "epoch": 1.8413017833599334, + "grad_norm": 1.5526679754257202, + "learning_rate": 0.0001176174456895546, + "loss": 0.5935, + "step": 26535 + }, + { + "epoch": 1.84137117479703, + "grad_norm": 1.5687508583068848, + "learning_rate": 0.0001176119149272732, + "loss": 0.5201, + "step": 26536 + }, + { + "epoch": 1.8414405662341267, + "grad_norm": 1.2502312660217285, + "learning_rate": 0.00011760638410939304, + "loss": 0.3598, + "step": 26537 + }, + { + "epoch": 1.8415099576712235, + "grad_norm": 1.2158489227294922, + "learning_rate": 0.0001176008532359316, + "loss": 0.4766, + "step": 26538 + }, + { + "epoch": 1.8415793491083199, + "grad_norm": 1.636702537536621, + "learning_rate": 0.00011759532230690629, + "loss": 0.6169, + "step": 26539 + }, + { + "epoch": 1.8416487405454167, + "grad_norm": 1.1189262866973877, + "learning_rate": 0.00011758979132233457, + "loss": 0.4382, + "step": 26540 + }, + { + "epoch": 1.8417181319825133, + "grad_norm": 1.7332640886306763, + "learning_rate": 0.00011758426028223392, + "loss": 0.6977, + "step": 26541 + }, + { + "epoch": 1.84178752341961, + "grad_norm": 1.901024341583252, + "learning_rate": 0.00011757872918662186, + "loss": 0.5321, + "step": 26542 + }, + { + "epoch": 1.8418569148567068, + "grad_norm": 1.6855621337890625, + "learning_rate": 0.00011757319803551571, + "loss": 0.4634, + "step": 26543 + }, + { + "epoch": 1.8419263062938034, + "grad_norm": 1.1849979162216187, + "learning_rate": 0.00011756766682893305, + "loss": 0.5152, + "step": 26544 + }, + { + "epoch": 1.8419956977309, + "grad_norm": 1.857638955116272, + "learning_rate": 0.00011756213556689133, + "loss": 0.5118, + "step": 26545 + }, + { + "epoch": 1.8420650891679968, + "grad_norm": 1.6724128723144531, + "learning_rate": 0.00011755660424940795, + "loss": 0.5393, + "step": 26546 + }, + { + "epoch": 1.8421344806050932, + "grad_norm": 1.631389856338501, + "learning_rate": 0.00011755107287650042, + "loss": 0.5932, + "step": 26547 + }, + { + "epoch": 1.84220387204219, + "grad_norm": 1.3781956434249878, + "learning_rate": 0.00011754554144818619, + "loss": 0.6161, + "step": 26548 + }, + { + "epoch": 1.8422732634792867, + "grad_norm": 1.6104258298873901, + "learning_rate": 0.0001175400099644827, + "loss": 0.6253, + "step": 26549 + }, + { + "epoch": 1.8423426549163833, + "grad_norm": 1.5570542812347412, + "learning_rate": 0.00011753447842540748, + "loss": 0.6531, + "step": 26550 + }, + { + "epoch": 1.84241204635348, + "grad_norm": 2.641545295715332, + "learning_rate": 0.00011752894683097792, + "loss": 0.5423, + "step": 26551 + }, + { + "epoch": 1.8424814377905765, + "grad_norm": 1.3714293241500854, + "learning_rate": 0.0001175234151812115, + "loss": 0.5552, + "step": 26552 + }, + { + "epoch": 1.8425508292276733, + "grad_norm": 1.2609717845916748, + "learning_rate": 0.0001175178834761257, + "loss": 0.5071, + "step": 26553 + }, + { + "epoch": 1.84262022066477, + "grad_norm": 1.5723038911819458, + "learning_rate": 0.00011751235171573798, + "loss": 0.7533, + "step": 26554 + }, + { + "epoch": 1.8426896121018665, + "grad_norm": 1.5500595569610596, + "learning_rate": 0.00011750681990006578, + "loss": 0.4368, + "step": 26555 + }, + { + "epoch": 1.8427590035389634, + "grad_norm": 1.2920665740966797, + "learning_rate": 0.00011750128802912662, + "loss": 0.4714, + "step": 26556 + }, + { + "epoch": 1.84282839497606, + "grad_norm": 1.0722295045852661, + "learning_rate": 0.00011749575610293785, + "loss": 0.3944, + "step": 26557 + }, + { + "epoch": 1.8428977864131566, + "grad_norm": 1.567149043083191, + "learning_rate": 0.00011749022412151707, + "loss": 0.7018, + "step": 26558 + }, + { + "epoch": 1.8429671778502534, + "grad_norm": 1.3168761730194092, + "learning_rate": 0.00011748469208488166, + "loss": 0.6316, + "step": 26559 + }, + { + "epoch": 1.8430365692873498, + "grad_norm": 1.6088454723358154, + "learning_rate": 0.00011747915999304913, + "loss": 0.6076, + "step": 26560 + }, + { + "epoch": 1.8431059607244467, + "grad_norm": 1.7025048732757568, + "learning_rate": 0.00011747362784603688, + "loss": 0.7365, + "step": 26561 + }, + { + "epoch": 1.8431753521615433, + "grad_norm": 1.7878718376159668, + "learning_rate": 0.00011746809564386243, + "loss": 0.6888, + "step": 26562 + }, + { + "epoch": 1.8432447435986399, + "grad_norm": 1.3150070905685425, + "learning_rate": 0.00011746256338654325, + "loss": 0.4445, + "step": 26563 + }, + { + "epoch": 1.8433141350357367, + "grad_norm": 1.8108021020889282, + "learning_rate": 0.00011745703107409676, + "loss": 0.5956, + "step": 26564 + }, + { + "epoch": 1.843383526472833, + "grad_norm": 1.7446579933166504, + "learning_rate": 0.00011745149870654046, + "loss": 0.4785, + "step": 26565 + }, + { + "epoch": 1.84345291790993, + "grad_norm": 1.8659241199493408, + "learning_rate": 0.0001174459662838918, + "loss": 0.7008, + "step": 26566 + }, + { + "epoch": 1.8435223093470265, + "grad_norm": 1.591285228729248, + "learning_rate": 0.00011744043380616825, + "loss": 0.6231, + "step": 26567 + }, + { + "epoch": 1.8435917007841232, + "grad_norm": 1.5222233533859253, + "learning_rate": 0.00011743490127338728, + "loss": 0.5847, + "step": 26568 + }, + { + "epoch": 1.84366109222122, + "grad_norm": 3.9137470722198486, + "learning_rate": 0.00011742936868556636, + "loss": 0.836, + "step": 26569 + }, + { + "epoch": 1.8437304836583166, + "grad_norm": 1.3433692455291748, + "learning_rate": 0.00011742383604272289, + "loss": 0.445, + "step": 26570 + }, + { + "epoch": 1.8437998750954132, + "grad_norm": 1.6504344940185547, + "learning_rate": 0.00011741830334487443, + "loss": 0.5547, + "step": 26571 + }, + { + "epoch": 1.84386926653251, + "grad_norm": 1.583978533744812, + "learning_rate": 0.00011741277059203839, + "loss": 0.6421, + "step": 26572 + }, + { + "epoch": 1.8439386579696064, + "grad_norm": 2.1075994968414307, + "learning_rate": 0.00011740723778423228, + "loss": 0.661, + "step": 26573 + }, + { + "epoch": 1.8440080494067033, + "grad_norm": 1.7688180208206177, + "learning_rate": 0.00011740170492147354, + "loss": 0.6282, + "step": 26574 + }, + { + "epoch": 1.8440774408437999, + "grad_norm": 1.1761279106140137, + "learning_rate": 0.00011739617200377959, + "loss": 0.4631, + "step": 26575 + }, + { + "epoch": 1.8441468322808965, + "grad_norm": 2.548215627670288, + "learning_rate": 0.000117390639031168, + "loss": 0.6348, + "step": 26576 + }, + { + "epoch": 1.8442162237179933, + "grad_norm": 1.7182338237762451, + "learning_rate": 0.00011738510600365616, + "loss": 0.6742, + "step": 26577 + }, + { + "epoch": 1.8442856151550897, + "grad_norm": 1.593705177307129, + "learning_rate": 0.00011737957292126154, + "loss": 0.5315, + "step": 26578 + }, + { + "epoch": 1.8443550065921865, + "grad_norm": 2.93662166595459, + "learning_rate": 0.00011737403978400163, + "loss": 0.7446, + "step": 26579 + }, + { + "epoch": 1.8444243980292832, + "grad_norm": 1.3100476264953613, + "learning_rate": 0.00011736850659189391, + "loss": 0.4996, + "step": 26580 + }, + { + "epoch": 1.8444937894663798, + "grad_norm": 1.1760931015014648, + "learning_rate": 0.00011736297334495583, + "loss": 0.3698, + "step": 26581 + }, + { + "epoch": 1.8445631809034766, + "grad_norm": 1.6633657217025757, + "learning_rate": 0.00011735744004320483, + "loss": 0.5541, + "step": 26582 + }, + { + "epoch": 1.8446325723405732, + "grad_norm": 1.5174537897109985, + "learning_rate": 0.00011735190668665845, + "loss": 0.7733, + "step": 26583 + }, + { + "epoch": 1.8447019637776698, + "grad_norm": 1.8884917497634888, + "learning_rate": 0.00011734637327533406, + "loss": 0.4064, + "step": 26584 + }, + { + "epoch": 1.8447713552147667, + "grad_norm": 1.7276535034179688, + "learning_rate": 0.00011734083980924924, + "loss": 0.8737, + "step": 26585 + }, + { + "epoch": 1.844840746651863, + "grad_norm": 1.637215256690979, + "learning_rate": 0.00011733530628842136, + "loss": 0.7061, + "step": 26586 + }, + { + "epoch": 1.8449101380889599, + "grad_norm": 1.414908528327942, + "learning_rate": 0.00011732977271286796, + "loss": 0.4451, + "step": 26587 + }, + { + "epoch": 1.8449795295260565, + "grad_norm": 2.195645809173584, + "learning_rate": 0.00011732423908260646, + "loss": 0.8975, + "step": 26588 + }, + { + "epoch": 1.845048920963153, + "grad_norm": 1.805095911026001, + "learning_rate": 0.00011731870539765433, + "loss": 0.6396, + "step": 26589 + }, + { + "epoch": 1.84511831240025, + "grad_norm": 1.4859861135482788, + "learning_rate": 0.00011731317165802909, + "loss": 0.491, + "step": 26590 + }, + { + "epoch": 1.8451877038373463, + "grad_norm": 1.2415554523468018, + "learning_rate": 0.00011730763786374814, + "loss": 0.5211, + "step": 26591 + }, + { + "epoch": 1.8452570952744431, + "grad_norm": 1.4220824241638184, + "learning_rate": 0.000117302104014829, + "loss": 0.4613, + "step": 26592 + }, + { + "epoch": 1.8453264867115398, + "grad_norm": 1.9644454717636108, + "learning_rate": 0.00011729657011128912, + "loss": 0.5289, + "step": 26593 + }, + { + "epoch": 1.8453958781486364, + "grad_norm": 1.3247967958450317, + "learning_rate": 0.000117291036153146, + "loss": 0.6791, + "step": 26594 + }, + { + "epoch": 1.8454652695857332, + "grad_norm": 1.2094823122024536, + "learning_rate": 0.00011728550214041707, + "loss": 0.5418, + "step": 26595 + }, + { + "epoch": 1.8455346610228298, + "grad_norm": 2.059100866317749, + "learning_rate": 0.00011727996807311981, + "loss": 0.6425, + "step": 26596 + }, + { + "epoch": 1.8456040524599264, + "grad_norm": 1.54485285282135, + "learning_rate": 0.00011727443395127172, + "loss": 0.7182, + "step": 26597 + }, + { + "epoch": 1.8456734438970233, + "grad_norm": 2.0585038661956787, + "learning_rate": 0.00011726889977489019, + "loss": 0.8348, + "step": 26598 + }, + { + "epoch": 1.8457428353341196, + "grad_norm": 1.985856056213379, + "learning_rate": 0.0001172633655439928, + "loss": 0.6165, + "step": 26599 + }, + { + "epoch": 1.8458122267712165, + "grad_norm": 3.028109312057495, + "learning_rate": 0.00011725783125859696, + "loss": 0.8569, + "step": 26600 + }, + { + "epoch": 1.845881618208313, + "grad_norm": 1.468531608581543, + "learning_rate": 0.00011725229691872013, + "loss": 0.4573, + "step": 26601 + }, + { + "epoch": 1.8459510096454097, + "grad_norm": 1.3577836751937866, + "learning_rate": 0.00011724676252437983, + "loss": 0.4709, + "step": 26602 + }, + { + "epoch": 1.8460204010825065, + "grad_norm": 1.699925184249878, + "learning_rate": 0.00011724122807559346, + "loss": 0.6012, + "step": 26603 + }, + { + "epoch": 1.846089792519603, + "grad_norm": 1.2617335319519043, + "learning_rate": 0.00011723569357237858, + "loss": 0.4197, + "step": 26604 + }, + { + "epoch": 1.8461591839566998, + "grad_norm": 1.4526230096817017, + "learning_rate": 0.00011723015901475257, + "loss": 0.6397, + "step": 26605 + }, + { + "epoch": 1.8462285753937964, + "grad_norm": 1.6385259628295898, + "learning_rate": 0.00011722462440273297, + "loss": 0.4887, + "step": 26606 + }, + { + "epoch": 1.846297966830893, + "grad_norm": 1.2490463256835938, + "learning_rate": 0.00011721908973633723, + "loss": 0.4071, + "step": 26607 + }, + { + "epoch": 1.8463673582679898, + "grad_norm": 1.6619852781295776, + "learning_rate": 0.00011721355501558283, + "loss": 0.425, + "step": 26608 + }, + { + "epoch": 1.8464367497050864, + "grad_norm": 1.898040533065796, + "learning_rate": 0.0001172080202404872, + "loss": 0.5006, + "step": 26609 + }, + { + "epoch": 1.846506141142183, + "grad_norm": 1.6776875257492065, + "learning_rate": 0.00011720248541106786, + "loss": 0.5115, + "step": 26610 + }, + { + "epoch": 1.8465755325792799, + "grad_norm": 1.8331449031829834, + "learning_rate": 0.0001171969505273423, + "loss": 0.742, + "step": 26611 + }, + { + "epoch": 1.8466449240163763, + "grad_norm": 2.2979118824005127, + "learning_rate": 0.00011719141558932791, + "loss": 0.9895, + "step": 26612 + }, + { + "epoch": 1.846714315453473, + "grad_norm": 1.6144304275512695, + "learning_rate": 0.00011718588059704225, + "loss": 0.4619, + "step": 26613 + }, + { + "epoch": 1.8467837068905697, + "grad_norm": 2.3684511184692383, + "learning_rate": 0.00011718034555050276, + "loss": 0.8067, + "step": 26614 + }, + { + "epoch": 1.8468530983276663, + "grad_norm": 1.5864061117172241, + "learning_rate": 0.00011717481044972691, + "loss": 0.529, + "step": 26615 + }, + { + "epoch": 1.8469224897647631, + "grad_norm": 1.6336069107055664, + "learning_rate": 0.00011716927529473216, + "loss": 0.487, + "step": 26616 + }, + { + "epoch": 1.8469918812018595, + "grad_norm": 1.455028772354126, + "learning_rate": 0.00011716374008553602, + "loss": 0.4424, + "step": 26617 + }, + { + "epoch": 1.8470612726389564, + "grad_norm": 1.7828298807144165, + "learning_rate": 0.00011715820482215593, + "loss": 0.5566, + "step": 26618 + }, + { + "epoch": 1.847130664076053, + "grad_norm": 1.7542521953582764, + "learning_rate": 0.00011715266950460942, + "loss": 0.6365, + "step": 26619 + }, + { + "epoch": 1.8472000555131496, + "grad_norm": 1.787001132965088, + "learning_rate": 0.00011714713413291388, + "loss": 0.6504, + "step": 26620 + }, + { + "epoch": 1.8472694469502464, + "grad_norm": 1.429474115371704, + "learning_rate": 0.00011714159870708683, + "loss": 0.5981, + "step": 26621 + }, + { + "epoch": 1.847338838387343, + "grad_norm": 1.2279949188232422, + "learning_rate": 0.00011713606322714574, + "loss": 0.5349, + "step": 26622 + }, + { + "epoch": 1.8474082298244396, + "grad_norm": 1.6688741445541382, + "learning_rate": 0.00011713052769310808, + "loss": 0.6859, + "step": 26623 + }, + { + "epoch": 1.8474776212615365, + "grad_norm": 0.9355474710464478, + "learning_rate": 0.00011712499210499135, + "loss": 0.4571, + "step": 26624 + }, + { + "epoch": 1.8475470126986329, + "grad_norm": 1.4863337278366089, + "learning_rate": 0.00011711945646281301, + "loss": 0.5593, + "step": 26625 + }, + { + "epoch": 1.8476164041357297, + "grad_norm": 1.34451425075531, + "learning_rate": 0.00011711392076659051, + "loss": 0.489, + "step": 26626 + }, + { + "epoch": 1.8476857955728263, + "grad_norm": 1.5205172300338745, + "learning_rate": 0.00011710838501634136, + "loss": 0.5642, + "step": 26627 + }, + { + "epoch": 1.847755187009923, + "grad_norm": 1.2220306396484375, + "learning_rate": 0.00011710284921208304, + "loss": 0.4884, + "step": 26628 + }, + { + "epoch": 1.8478245784470197, + "grad_norm": 1.5862092971801758, + "learning_rate": 0.00011709731335383299, + "loss": 0.5938, + "step": 26629 + }, + { + "epoch": 1.8478939698841161, + "grad_norm": 1.7568213939666748, + "learning_rate": 0.00011709177744160868, + "loss": 0.6261, + "step": 26630 + }, + { + "epoch": 1.847963361321213, + "grad_norm": 1.3229023218154907, + "learning_rate": 0.00011708624147542764, + "loss": 0.4771, + "step": 26631 + }, + { + "epoch": 1.8480327527583096, + "grad_norm": 3.7279574871063232, + "learning_rate": 0.00011708070545530734, + "loss": 0.6929, + "step": 26632 + }, + { + "epoch": 1.8481021441954062, + "grad_norm": 1.9270458221435547, + "learning_rate": 0.0001170751693812652, + "loss": 0.6138, + "step": 26633 + }, + { + "epoch": 1.848171535632503, + "grad_norm": 1.2329615354537964, + "learning_rate": 0.00011706963325331875, + "loss": 0.5326, + "step": 26634 + }, + { + "epoch": 1.8482409270695996, + "grad_norm": 1.8594074249267578, + "learning_rate": 0.00011706409707148544, + "loss": 0.5213, + "step": 26635 + }, + { + "epoch": 1.8483103185066962, + "grad_norm": 2.308730363845825, + "learning_rate": 0.00011705856083578277, + "loss": 0.8285, + "step": 26636 + }, + { + "epoch": 1.848379709943793, + "grad_norm": 1.690053939819336, + "learning_rate": 0.0001170530245462282, + "loss": 0.5077, + "step": 26637 + }, + { + "epoch": 1.8484491013808895, + "grad_norm": 1.250970482826233, + "learning_rate": 0.00011704748820283919, + "loss": 0.5146, + "step": 26638 + }, + { + "epoch": 1.8485184928179863, + "grad_norm": 1.2936713695526123, + "learning_rate": 0.00011704195180563327, + "loss": 0.5749, + "step": 26639 + }, + { + "epoch": 1.848587884255083, + "grad_norm": 1.3185123205184937, + "learning_rate": 0.00011703641535462784, + "loss": 0.457, + "step": 26640 + }, + { + "epoch": 1.8486572756921795, + "grad_norm": 1.9601781368255615, + "learning_rate": 0.00011703087884984046, + "loss": 1.0526, + "step": 26641 + }, + { + "epoch": 1.8487266671292764, + "grad_norm": 1.5916242599487305, + "learning_rate": 0.00011702534229128856, + "loss": 0.4043, + "step": 26642 + }, + { + "epoch": 1.8487960585663727, + "grad_norm": 1.86936354637146, + "learning_rate": 0.00011701980567898963, + "loss": 0.5331, + "step": 26643 + }, + { + "epoch": 1.8488654500034696, + "grad_norm": 1.923013687133789, + "learning_rate": 0.00011701426901296113, + "loss": 0.7205, + "step": 26644 + }, + { + "epoch": 1.8489348414405662, + "grad_norm": 3.6780314445495605, + "learning_rate": 0.00011700873229322058, + "loss": 0.4713, + "step": 26645 + }, + { + "epoch": 1.8490042328776628, + "grad_norm": 1.4813222885131836, + "learning_rate": 0.00011700319551978543, + "loss": 0.4611, + "step": 26646 + }, + { + "epoch": 1.8490736243147596, + "grad_norm": 1.9497742652893066, + "learning_rate": 0.00011699765869267314, + "loss": 0.5347, + "step": 26647 + }, + { + "epoch": 1.8491430157518562, + "grad_norm": 1.5663542747497559, + "learning_rate": 0.00011699212181190124, + "loss": 0.6007, + "step": 26648 + }, + { + "epoch": 1.8492124071889529, + "grad_norm": 2.0826237201690674, + "learning_rate": 0.00011698658487748719, + "loss": 0.56, + "step": 26649 + }, + { + "epoch": 1.8492817986260497, + "grad_norm": 2.811911106109619, + "learning_rate": 0.00011698104788944845, + "loss": 0.5842, + "step": 26650 + }, + { + "epoch": 1.849351190063146, + "grad_norm": 1.5934251546859741, + "learning_rate": 0.00011697551084780252, + "loss": 0.6134, + "step": 26651 + }, + { + "epoch": 1.849420581500243, + "grad_norm": 2.0443837642669678, + "learning_rate": 0.00011696997375256686, + "loss": 0.7756, + "step": 26652 + }, + { + "epoch": 1.8494899729373395, + "grad_norm": 1.9846911430358887, + "learning_rate": 0.00011696443660375895, + "loss": 0.6158, + "step": 26653 + }, + { + "epoch": 1.8495593643744361, + "grad_norm": 1.8365466594696045, + "learning_rate": 0.00011695889940139632, + "loss": 0.7099, + "step": 26654 + }, + { + "epoch": 1.849628755811533, + "grad_norm": 1.2027692794799805, + "learning_rate": 0.00011695336214549639, + "loss": 0.3283, + "step": 26655 + }, + { + "epoch": 1.8496981472486294, + "grad_norm": 2.157419443130493, + "learning_rate": 0.0001169478248360767, + "loss": 0.6343, + "step": 26656 + }, + { + "epoch": 1.8497675386857262, + "grad_norm": 1.6345499753952026, + "learning_rate": 0.00011694228747315464, + "loss": 0.6644, + "step": 26657 + }, + { + "epoch": 1.8498369301228228, + "grad_norm": 1.2061927318572998, + "learning_rate": 0.00011693675005674776, + "loss": 0.302, + "step": 26658 + }, + { + "epoch": 1.8499063215599194, + "grad_norm": 1.601681113243103, + "learning_rate": 0.00011693121258687352, + "loss": 0.6668, + "step": 26659 + }, + { + "epoch": 1.8499757129970162, + "grad_norm": 1.3331217765808105, + "learning_rate": 0.00011692567506354941, + "loss": 0.5088, + "step": 26660 + }, + { + "epoch": 1.8500451044341129, + "grad_norm": 1.5217469930648804, + "learning_rate": 0.00011692013748679289, + "loss": 0.5522, + "step": 26661 + }, + { + "epoch": 1.8501144958712095, + "grad_norm": 1.5148470401763916, + "learning_rate": 0.00011691459985662147, + "loss": 0.4764, + "step": 26662 + }, + { + "epoch": 1.8501838873083063, + "grad_norm": 3.216634511947632, + "learning_rate": 0.00011690906217305266, + "loss": 0.5384, + "step": 26663 + }, + { + "epoch": 1.8502532787454027, + "grad_norm": 1.4700158834457397, + "learning_rate": 0.00011690352443610387, + "loss": 0.4119, + "step": 26664 + }, + { + "epoch": 1.8503226701824995, + "grad_norm": 1.336395263671875, + "learning_rate": 0.00011689798664579262, + "loss": 0.4863, + "step": 26665 + }, + { + "epoch": 1.8503920616195961, + "grad_norm": 2.269747495651245, + "learning_rate": 0.00011689244880213638, + "loss": 0.8811, + "step": 26666 + }, + { + "epoch": 1.8504614530566927, + "grad_norm": 1.1122106313705444, + "learning_rate": 0.00011688691090515265, + "loss": 0.3797, + "step": 26667 + }, + { + "epoch": 1.8505308444937896, + "grad_norm": 1.5207444429397583, + "learning_rate": 0.00011688137295485888, + "loss": 0.6058, + "step": 26668 + }, + { + "epoch": 1.8506002359308862, + "grad_norm": 1.6265441179275513, + "learning_rate": 0.00011687583495127257, + "loss": 0.6337, + "step": 26669 + }, + { + "epoch": 1.8506696273679828, + "grad_norm": 2.222482442855835, + "learning_rate": 0.00011687029689441125, + "loss": 0.5327, + "step": 26670 + }, + { + "epoch": 1.8507390188050794, + "grad_norm": 2.1051881313323975, + "learning_rate": 0.00011686475878429233, + "loss": 0.7693, + "step": 26671 + }, + { + "epoch": 1.850808410242176, + "grad_norm": 1.5727038383483887, + "learning_rate": 0.00011685922062093333, + "loss": 0.4818, + "step": 26672 + }, + { + "epoch": 1.8508778016792728, + "grad_norm": 1.5951508283615112, + "learning_rate": 0.00011685368240435175, + "loss": 0.5117, + "step": 26673 + }, + { + "epoch": 1.8509471931163695, + "grad_norm": 2.2699615955352783, + "learning_rate": 0.00011684814413456502, + "loss": 0.6911, + "step": 26674 + }, + { + "epoch": 1.851016584553466, + "grad_norm": 2.3335623741149902, + "learning_rate": 0.00011684260581159066, + "loss": 0.514, + "step": 26675 + }, + { + "epoch": 1.851085975990563, + "grad_norm": 2.1457159519195557, + "learning_rate": 0.00011683706743544614, + "loss": 0.6549, + "step": 26676 + }, + { + "epoch": 1.8511553674276593, + "grad_norm": 1.6215487718582153, + "learning_rate": 0.00011683152900614895, + "loss": 0.66, + "step": 26677 + }, + { + "epoch": 1.8512247588647561, + "grad_norm": 1.5221526622772217, + "learning_rate": 0.00011682599052371659, + "loss": 0.4088, + "step": 26678 + }, + { + "epoch": 1.8512941503018527, + "grad_norm": 1.3408890962600708, + "learning_rate": 0.00011682045198816651, + "loss": 0.5861, + "step": 26679 + }, + { + "epoch": 1.8513635417389493, + "grad_norm": 1.3940314054489136, + "learning_rate": 0.00011681491339951624, + "loss": 0.5331, + "step": 26680 + }, + { + "epoch": 1.8514329331760462, + "grad_norm": 0.928605854511261, + "learning_rate": 0.00011680937475778322, + "loss": 0.4356, + "step": 26681 + }, + { + "epoch": 1.8515023246131428, + "grad_norm": 1.590961217880249, + "learning_rate": 0.00011680383606298496, + "loss": 0.7912, + "step": 26682 + }, + { + "epoch": 1.8515717160502394, + "grad_norm": 1.6925102472305298, + "learning_rate": 0.00011679829731513895, + "loss": 0.4929, + "step": 26683 + }, + { + "epoch": 1.851641107487336, + "grad_norm": 1.9451910257339478, + "learning_rate": 0.00011679275851426266, + "loss": 0.5862, + "step": 26684 + }, + { + "epoch": 1.8517104989244326, + "grad_norm": 1.573625087738037, + "learning_rate": 0.00011678721966037355, + "loss": 0.6704, + "step": 26685 + }, + { + "epoch": 1.8517798903615295, + "grad_norm": 1.1818422079086304, + "learning_rate": 0.00011678168075348916, + "loss": 0.4137, + "step": 26686 + }, + { + "epoch": 1.851849281798626, + "grad_norm": 1.6588706970214844, + "learning_rate": 0.00011677614179362696, + "loss": 0.4853, + "step": 26687 + }, + { + "epoch": 1.8519186732357227, + "grad_norm": 2.4942731857299805, + "learning_rate": 0.00011677060278080443, + "loss": 0.8821, + "step": 26688 + }, + { + "epoch": 1.8519880646728195, + "grad_norm": 1.6013314723968506, + "learning_rate": 0.00011676506371503903, + "loss": 0.6584, + "step": 26689 + }, + { + "epoch": 1.852057456109916, + "grad_norm": 1.640591025352478, + "learning_rate": 0.00011675952459634829, + "loss": 0.5457, + "step": 26690 + }, + { + "epoch": 1.8521268475470127, + "grad_norm": 2.337419033050537, + "learning_rate": 0.00011675398542474969, + "loss": 0.5448, + "step": 26691 + }, + { + "epoch": 1.8521962389841093, + "grad_norm": 8.01895809173584, + "learning_rate": 0.00011674844620026066, + "loss": 0.6722, + "step": 26692 + }, + { + "epoch": 1.852265630421206, + "grad_norm": 1.2430572509765625, + "learning_rate": 0.00011674290692289873, + "loss": 0.5319, + "step": 26693 + }, + { + "epoch": 1.8523350218583028, + "grad_norm": 1.058141827583313, + "learning_rate": 0.00011673736759268141, + "loss": 0.299, + "step": 26694 + }, + { + "epoch": 1.8524044132953994, + "grad_norm": 1.462307333946228, + "learning_rate": 0.00011673182820962616, + "loss": 0.4682, + "step": 26695 + }, + { + "epoch": 1.852473804732496, + "grad_norm": 1.4368683099746704, + "learning_rate": 0.00011672628877375044, + "loss": 0.5561, + "step": 26696 + }, + { + "epoch": 1.8525431961695926, + "grad_norm": 1.6053462028503418, + "learning_rate": 0.00011672074928507182, + "loss": 0.5515, + "step": 26697 + }, + { + "epoch": 1.8526125876066892, + "grad_norm": 1.736706018447876, + "learning_rate": 0.0001167152097436077, + "loss": 0.6604, + "step": 26698 + }, + { + "epoch": 1.852681979043786, + "grad_norm": 1.546852707862854, + "learning_rate": 0.00011670967014937559, + "loss": 0.6278, + "step": 26699 + }, + { + "epoch": 1.8527513704808827, + "grad_norm": 1.6773220300674438, + "learning_rate": 0.00011670413050239299, + "loss": 0.4791, + "step": 26700 + }, + { + "epoch": 1.8528207619179793, + "grad_norm": 5.7856669425964355, + "learning_rate": 0.00011669859080267742, + "loss": 0.6111, + "step": 26701 + }, + { + "epoch": 1.8528901533550761, + "grad_norm": 1.46244215965271, + "learning_rate": 0.00011669305105024632, + "loss": 0.3246, + "step": 26702 + }, + { + "epoch": 1.8529595447921725, + "grad_norm": 1.6774085760116577, + "learning_rate": 0.0001166875112451172, + "loss": 0.6382, + "step": 26703 + }, + { + "epoch": 1.8530289362292693, + "grad_norm": 1.9717751741409302, + "learning_rate": 0.00011668197138730755, + "loss": 0.3832, + "step": 26704 + }, + { + "epoch": 1.853098327666366, + "grad_norm": 1.5738030672073364, + "learning_rate": 0.00011667643147683483, + "loss": 0.6161, + "step": 26705 + }, + { + "epoch": 1.8531677191034626, + "grad_norm": 1.6838834285736084, + "learning_rate": 0.00011667089151371655, + "loss": 0.6941, + "step": 26706 + }, + { + "epoch": 1.8532371105405594, + "grad_norm": 2.252528190612793, + "learning_rate": 0.00011666535149797022, + "loss": 0.5421, + "step": 26707 + }, + { + "epoch": 1.853306501977656, + "grad_norm": 1.8309544324874878, + "learning_rate": 0.0001166598114296133, + "loss": 0.4955, + "step": 26708 + }, + { + "epoch": 1.8533758934147526, + "grad_norm": 2.30187726020813, + "learning_rate": 0.0001166542713086633, + "loss": 0.6451, + "step": 26709 + }, + { + "epoch": 1.8534452848518492, + "grad_norm": 2.3113865852355957, + "learning_rate": 0.00011664873113513767, + "loss": 0.8912, + "step": 26710 + }, + { + "epoch": 1.8535146762889458, + "grad_norm": 1.6466881036758423, + "learning_rate": 0.00011664319090905396, + "loss": 0.6406, + "step": 26711 + }, + { + "epoch": 1.8535840677260427, + "grad_norm": 2.255643129348755, + "learning_rate": 0.0001166376506304296, + "loss": 0.5626, + "step": 26712 + }, + { + "epoch": 1.8536534591631393, + "grad_norm": 1.498197317123413, + "learning_rate": 0.00011663211029928212, + "loss": 0.6712, + "step": 26713 + }, + { + "epoch": 1.853722850600236, + "grad_norm": 1.257367730140686, + "learning_rate": 0.00011662656991562899, + "loss": 0.5092, + "step": 26714 + }, + { + "epoch": 1.8537922420373327, + "grad_norm": 1.838079810142517, + "learning_rate": 0.00011662102947948772, + "loss": 0.5671, + "step": 26715 + }, + { + "epoch": 1.8538616334744291, + "grad_norm": 1.9873683452606201, + "learning_rate": 0.00011661548899087578, + "loss": 0.6887, + "step": 26716 + }, + { + "epoch": 1.853931024911526, + "grad_norm": 2.0232748985290527, + "learning_rate": 0.00011660994844981065, + "loss": 0.6401, + "step": 26717 + }, + { + "epoch": 1.8540004163486226, + "grad_norm": 1.8058099746704102, + "learning_rate": 0.00011660440785630988, + "loss": 0.7099, + "step": 26718 + }, + { + "epoch": 1.8540698077857192, + "grad_norm": 1.4759924411773682, + "learning_rate": 0.00011659886721039091, + "loss": 0.502, + "step": 26719 + }, + { + "epoch": 1.854139199222816, + "grad_norm": 1.189131259918213, + "learning_rate": 0.0001165933265120712, + "loss": 0.4241, + "step": 26720 + }, + { + "epoch": 1.8542085906599126, + "grad_norm": 1.8423746824264526, + "learning_rate": 0.00011658778576136833, + "loss": 0.6139, + "step": 26721 + }, + { + "epoch": 1.8542779820970092, + "grad_norm": 2.2624621391296387, + "learning_rate": 0.00011658224495829974, + "loss": 0.7314, + "step": 26722 + }, + { + "epoch": 1.8543473735341058, + "grad_norm": 2.095581531524658, + "learning_rate": 0.00011657670410288291, + "loss": 0.7211, + "step": 26723 + }, + { + "epoch": 1.8544167649712024, + "grad_norm": 2.3924190998077393, + "learning_rate": 0.00011657116319513536, + "loss": 0.6445, + "step": 26724 + }, + { + "epoch": 1.8544861564082993, + "grad_norm": 3.5890052318573, + "learning_rate": 0.00011656562223507456, + "loss": 0.5821, + "step": 26725 + }, + { + "epoch": 1.854555547845396, + "grad_norm": 1.7529144287109375, + "learning_rate": 0.00011656008122271804, + "loss": 0.5605, + "step": 26726 + }, + { + "epoch": 1.8546249392824925, + "grad_norm": 1.6142042875289917, + "learning_rate": 0.00011655454015808323, + "loss": 0.6747, + "step": 26727 + }, + { + "epoch": 1.8546943307195893, + "grad_norm": 1.8153090476989746, + "learning_rate": 0.0001165489990411877, + "loss": 0.5153, + "step": 26728 + }, + { + "epoch": 1.8547637221566857, + "grad_norm": 1.547369122505188, + "learning_rate": 0.00011654345787204885, + "loss": 0.5111, + "step": 26729 + }, + { + "epoch": 1.8548331135937826, + "grad_norm": 1.5596048831939697, + "learning_rate": 0.00011653791665068423, + "loss": 0.5029, + "step": 26730 + }, + { + "epoch": 1.8549025050308792, + "grad_norm": 1.702444076538086, + "learning_rate": 0.00011653237537711134, + "loss": 0.6641, + "step": 26731 + }, + { + "epoch": 1.8549718964679758, + "grad_norm": 1.6285021305084229, + "learning_rate": 0.00011652683405134768, + "loss": 0.544, + "step": 26732 + }, + { + "epoch": 1.8550412879050726, + "grad_norm": 1.4283190965652466, + "learning_rate": 0.0001165212926734107, + "loss": 0.5227, + "step": 26733 + }, + { + "epoch": 1.8551106793421692, + "grad_norm": 1.7928638458251953, + "learning_rate": 0.00011651575124331791, + "loss": 0.5834, + "step": 26734 + }, + { + "epoch": 1.8551800707792658, + "grad_norm": 2.113050937652588, + "learning_rate": 0.00011651020976108682, + "loss": 0.8106, + "step": 26735 + }, + { + "epoch": 1.8552494622163627, + "grad_norm": 1.6162021160125732, + "learning_rate": 0.00011650466822673491, + "loss": 0.6226, + "step": 26736 + }, + { + "epoch": 1.855318853653459, + "grad_norm": 1.74959397315979, + "learning_rate": 0.00011649912664027967, + "loss": 0.5898, + "step": 26737 + }, + { + "epoch": 1.8553882450905559, + "grad_norm": 1.5364882946014404, + "learning_rate": 0.00011649358500173859, + "loss": 0.6046, + "step": 26738 + }, + { + "epoch": 1.8554576365276525, + "grad_norm": 1.159549593925476, + "learning_rate": 0.00011648804331112922, + "loss": 0.4985, + "step": 26739 + }, + { + "epoch": 1.855527027964749, + "grad_norm": 1.301254391670227, + "learning_rate": 0.00011648250156846896, + "loss": 0.4906, + "step": 26740 + }, + { + "epoch": 1.855596419401846, + "grad_norm": 3.0077240467071533, + "learning_rate": 0.00011647695977377542, + "loss": 0.874, + "step": 26741 + }, + { + "epoch": 1.8556658108389423, + "grad_norm": 0.9461413025856018, + "learning_rate": 0.00011647141792706598, + "loss": 0.3255, + "step": 26742 + }, + { + "epoch": 1.8557352022760392, + "grad_norm": 1.4562283754348755, + "learning_rate": 0.00011646587602835819, + "loss": 0.5407, + "step": 26743 + }, + { + "epoch": 1.8558045937131358, + "grad_norm": 1.795627474784851, + "learning_rate": 0.00011646033407766955, + "loss": 0.6601, + "step": 26744 + }, + { + "epoch": 1.8558739851502324, + "grad_norm": 1.980362057685852, + "learning_rate": 0.00011645479207501752, + "loss": 0.7163, + "step": 26745 + }, + { + "epoch": 1.8559433765873292, + "grad_norm": 1.6650737524032593, + "learning_rate": 0.00011644925002041966, + "loss": 0.3754, + "step": 26746 + }, + { + "epoch": 1.8560127680244258, + "grad_norm": 1.4092938899993896, + "learning_rate": 0.00011644370791389341, + "loss": 0.4779, + "step": 26747 + }, + { + "epoch": 1.8560821594615224, + "grad_norm": 1.2977681159973145, + "learning_rate": 0.00011643816575545625, + "loss": 0.5744, + "step": 26748 + }, + { + "epoch": 1.8561515508986193, + "grad_norm": 1.8476017713546753, + "learning_rate": 0.00011643262354512576, + "loss": 0.5249, + "step": 26749 + }, + { + "epoch": 1.8562209423357157, + "grad_norm": 2.094909191131592, + "learning_rate": 0.00011642708128291935, + "loss": 0.8871, + "step": 26750 + }, + { + "epoch": 1.8562903337728125, + "grad_norm": 2.290123701095581, + "learning_rate": 0.00011642153896885455, + "loss": 0.9691, + "step": 26751 + }, + { + "epoch": 1.856359725209909, + "grad_norm": 1.324892520904541, + "learning_rate": 0.00011641599660294886, + "loss": 0.5106, + "step": 26752 + }, + { + "epoch": 1.8564291166470057, + "grad_norm": 1.770308494567871, + "learning_rate": 0.00011641045418521978, + "loss": 0.6949, + "step": 26753 + }, + { + "epoch": 1.8564985080841025, + "grad_norm": 1.629908800125122, + "learning_rate": 0.00011640491171568482, + "loss": 0.7444, + "step": 26754 + }, + { + "epoch": 1.856567899521199, + "grad_norm": 1.3946932554244995, + "learning_rate": 0.00011639936919436141, + "loss": 0.4357, + "step": 26755 + }, + { + "epoch": 1.8566372909582958, + "grad_norm": 1.942138671875, + "learning_rate": 0.00011639382662126714, + "loss": 0.6432, + "step": 26756 + }, + { + "epoch": 1.8567066823953924, + "grad_norm": 1.8396447896957397, + "learning_rate": 0.00011638828399641944, + "loss": 0.7635, + "step": 26757 + }, + { + "epoch": 1.856776073832489, + "grad_norm": 1.9662868976593018, + "learning_rate": 0.00011638274131983582, + "loss": 0.5906, + "step": 26758 + }, + { + "epoch": 1.8568454652695858, + "grad_norm": 1.6331979036331177, + "learning_rate": 0.00011637719859153382, + "loss": 0.6691, + "step": 26759 + }, + { + "epoch": 1.8569148567066824, + "grad_norm": 1.522177815437317, + "learning_rate": 0.00011637165581153087, + "loss": 0.5745, + "step": 26760 + }, + { + "epoch": 1.856984248143779, + "grad_norm": 4.476472854614258, + "learning_rate": 0.00011636611297984452, + "loss": 0.5208, + "step": 26761 + }, + { + "epoch": 1.8570536395808759, + "grad_norm": 1.4999394416809082, + "learning_rate": 0.00011636057009649223, + "loss": 0.5281, + "step": 26762 + }, + { + "epoch": 1.8571230310179723, + "grad_norm": 5.304169654846191, + "learning_rate": 0.00011635502716149155, + "loss": 0.4938, + "step": 26763 + }, + { + "epoch": 1.857192422455069, + "grad_norm": 1.6378207206726074, + "learning_rate": 0.00011634948417485993, + "loss": 0.6241, + "step": 26764 + }, + { + "epoch": 1.8572618138921657, + "grad_norm": 1.4747411012649536, + "learning_rate": 0.00011634394113661488, + "loss": 0.492, + "step": 26765 + }, + { + "epoch": 1.8573312053292623, + "grad_norm": 2.312387704849243, + "learning_rate": 0.00011633839804677391, + "loss": 0.5252, + "step": 26766 + }, + { + "epoch": 1.8574005967663592, + "grad_norm": 1.474291205406189, + "learning_rate": 0.00011633285490535453, + "loss": 0.4949, + "step": 26767 + }, + { + "epoch": 1.8574699882034555, + "grad_norm": 1.7543355226516724, + "learning_rate": 0.00011632731171237419, + "loss": 0.4842, + "step": 26768 + }, + { + "epoch": 1.8575393796405524, + "grad_norm": 2.0778584480285645, + "learning_rate": 0.00011632176846785042, + "loss": 0.6527, + "step": 26769 + }, + { + "epoch": 1.857608771077649, + "grad_norm": 2.1068906784057617, + "learning_rate": 0.00011631622517180074, + "loss": 0.7722, + "step": 26770 + }, + { + "epoch": 1.8576781625147456, + "grad_norm": 1.6916992664337158, + "learning_rate": 0.00011631068182424263, + "loss": 0.5965, + "step": 26771 + }, + { + "epoch": 1.8577475539518424, + "grad_norm": 1.6407816410064697, + "learning_rate": 0.00011630513842519357, + "loss": 0.7714, + "step": 26772 + }, + { + "epoch": 1.857816945388939, + "grad_norm": 1.188715934753418, + "learning_rate": 0.00011629959497467111, + "loss": 0.5911, + "step": 26773 + }, + { + "epoch": 1.8578863368260357, + "grad_norm": 1.383219599723816, + "learning_rate": 0.0001162940514726927, + "loss": 0.4793, + "step": 26774 + }, + { + "epoch": 1.8579557282631325, + "grad_norm": 3.452852725982666, + "learning_rate": 0.00011628850791927586, + "loss": 0.7628, + "step": 26775 + }, + { + "epoch": 1.8580251197002289, + "grad_norm": 1.363048791885376, + "learning_rate": 0.0001162829643144381, + "loss": 0.5602, + "step": 26776 + }, + { + "epoch": 1.8580945111373257, + "grad_norm": 1.9249604940414429, + "learning_rate": 0.00011627742065819691, + "loss": 0.6208, + "step": 26777 + }, + { + "epoch": 1.8581639025744223, + "grad_norm": 1.4608440399169922, + "learning_rate": 0.0001162718769505698, + "loss": 0.6714, + "step": 26778 + }, + { + "epoch": 1.858233294011519, + "grad_norm": 2.158057689666748, + "learning_rate": 0.00011626633319157425, + "loss": 0.696, + "step": 26779 + }, + { + "epoch": 1.8583026854486158, + "grad_norm": 1.7116034030914307, + "learning_rate": 0.00011626078938122777, + "loss": 0.548, + "step": 26780 + }, + { + "epoch": 1.8583720768857122, + "grad_norm": 4.897891998291016, + "learning_rate": 0.00011625524551954788, + "loss": 0.9185, + "step": 26781 + }, + { + "epoch": 1.858441468322809, + "grad_norm": 1.5069005489349365, + "learning_rate": 0.00011624970160655204, + "loss": 0.4512, + "step": 26782 + }, + { + "epoch": 1.8585108597599056, + "grad_norm": 1.2405825853347778, + "learning_rate": 0.00011624415764225778, + "loss": 0.4778, + "step": 26783 + }, + { + "epoch": 1.8585802511970022, + "grad_norm": 1.8218185901641846, + "learning_rate": 0.00011623861362668262, + "loss": 0.576, + "step": 26784 + }, + { + "epoch": 1.858649642634099, + "grad_norm": 2.234740972518921, + "learning_rate": 0.00011623306955984403, + "loss": 0.7207, + "step": 26785 + }, + { + "epoch": 1.8587190340711957, + "grad_norm": 1.6525847911834717, + "learning_rate": 0.00011622752544175952, + "loss": 0.5942, + "step": 26786 + }, + { + "epoch": 1.8587884255082923, + "grad_norm": 2.5703678131103516, + "learning_rate": 0.00011622198127244661, + "loss": 0.7099, + "step": 26787 + }, + { + "epoch": 1.858857816945389, + "grad_norm": 1.8882360458374023, + "learning_rate": 0.00011621643705192278, + "loss": 0.5787, + "step": 26788 + }, + { + "epoch": 1.8589272083824855, + "grad_norm": 2.1368069648742676, + "learning_rate": 0.00011621089278020552, + "loss": 0.7196, + "step": 26789 + }, + { + "epoch": 1.8589965998195823, + "grad_norm": 2.0525434017181396, + "learning_rate": 0.00011620534845731236, + "loss": 0.6639, + "step": 26790 + }, + { + "epoch": 1.859065991256679, + "grad_norm": 1.4539798498153687, + "learning_rate": 0.00011619980408326081, + "loss": 0.444, + "step": 26791 + }, + { + "epoch": 1.8591353826937755, + "grad_norm": 1.6577880382537842, + "learning_rate": 0.00011619425965806834, + "loss": 0.5323, + "step": 26792 + }, + { + "epoch": 1.8592047741308724, + "grad_norm": 1.954363465309143, + "learning_rate": 0.00011618871518175249, + "loss": 0.6668, + "step": 26793 + }, + { + "epoch": 1.8592741655679688, + "grad_norm": 1.7640247344970703, + "learning_rate": 0.00011618317065433074, + "loss": 0.6058, + "step": 26794 + }, + { + "epoch": 1.8593435570050656, + "grad_norm": 1.6434636116027832, + "learning_rate": 0.00011617762607582062, + "loss": 0.6089, + "step": 26795 + }, + { + "epoch": 1.8594129484421622, + "grad_norm": 1.5967960357666016, + "learning_rate": 0.00011617208144623956, + "loss": 0.4823, + "step": 26796 + }, + { + "epoch": 1.8594823398792588, + "grad_norm": 1.061346411705017, + "learning_rate": 0.00011616653676560516, + "loss": 0.3352, + "step": 26797 + }, + { + "epoch": 1.8595517313163556, + "grad_norm": 1.2371559143066406, + "learning_rate": 0.00011616099203393487, + "loss": 0.4279, + "step": 26798 + }, + { + "epoch": 1.8596211227534523, + "grad_norm": 1.4401007890701294, + "learning_rate": 0.00011615544725124619, + "loss": 0.7677, + "step": 26799 + }, + { + "epoch": 1.8596905141905489, + "grad_norm": 1.3832318782806396, + "learning_rate": 0.00011614990241755662, + "loss": 0.5323, + "step": 26800 + }, + { + "epoch": 1.8597599056276457, + "grad_norm": 2.5580108165740967, + "learning_rate": 0.0001161443575328837, + "loss": 0.6158, + "step": 26801 + }, + { + "epoch": 1.859829297064742, + "grad_norm": 1.3766118288040161, + "learning_rate": 0.00011613881259724492, + "loss": 0.5732, + "step": 26802 + }, + { + "epoch": 1.859898688501839, + "grad_norm": 1.655400276184082, + "learning_rate": 0.00011613326761065778, + "loss": 0.5327, + "step": 26803 + }, + { + "epoch": 1.8599680799389355, + "grad_norm": 1.552209496498108, + "learning_rate": 0.0001161277225731398, + "loss": 0.4634, + "step": 26804 + }, + { + "epoch": 1.8600374713760321, + "grad_norm": 1.9484468698501587, + "learning_rate": 0.00011612217748470843, + "loss": 0.5221, + "step": 26805 + }, + { + "epoch": 1.860106862813129, + "grad_norm": 1.249404788017273, + "learning_rate": 0.00011611663234538126, + "loss": 0.4775, + "step": 26806 + }, + { + "epoch": 1.8601762542502254, + "grad_norm": 1.382330298423767, + "learning_rate": 0.00011611108715517571, + "loss": 0.53, + "step": 26807 + }, + { + "epoch": 1.8602456456873222, + "grad_norm": 1.69806706905365, + "learning_rate": 0.00011610554191410939, + "loss": 0.6623, + "step": 26808 + }, + { + "epoch": 1.8603150371244188, + "grad_norm": 1.3396141529083252, + "learning_rate": 0.00011609999662219969, + "loss": 0.4784, + "step": 26809 + }, + { + "epoch": 1.8603844285615154, + "grad_norm": 1.4340943098068237, + "learning_rate": 0.00011609445127946417, + "loss": 0.5531, + "step": 26810 + }, + { + "epoch": 1.8604538199986123, + "grad_norm": 1.4442646503448486, + "learning_rate": 0.00011608890588592036, + "loss": 0.585, + "step": 26811 + }, + { + "epoch": 1.8605232114357089, + "grad_norm": 2.717843532562256, + "learning_rate": 0.00011608336044158572, + "loss": 0.6471, + "step": 26812 + }, + { + "epoch": 1.8605926028728055, + "grad_norm": 1.7045258283615112, + "learning_rate": 0.00011607781494647781, + "loss": 0.6234, + "step": 26813 + }, + { + "epoch": 1.8606619943099023, + "grad_norm": 1.9052151441574097, + "learning_rate": 0.00011607226940061406, + "loss": 0.67, + "step": 26814 + }, + { + "epoch": 1.8607313857469987, + "grad_norm": 2.3220107555389404, + "learning_rate": 0.00011606672380401207, + "loss": 0.7522, + "step": 26815 + }, + { + "epoch": 1.8608007771840955, + "grad_norm": 1.637265920639038, + "learning_rate": 0.00011606117815668924, + "loss": 0.6471, + "step": 26816 + }, + { + "epoch": 1.8608701686211921, + "grad_norm": 1.8563337326049805, + "learning_rate": 0.00011605563245866316, + "loss": 0.562, + "step": 26817 + }, + { + "epoch": 1.8609395600582888, + "grad_norm": 1.6983915567398071, + "learning_rate": 0.00011605008670995133, + "loss": 0.5735, + "step": 26818 + }, + { + "epoch": 1.8610089514953856, + "grad_norm": 1.8389261960983276, + "learning_rate": 0.00011604454091057122, + "loss": 0.6376, + "step": 26819 + }, + { + "epoch": 1.861078342932482, + "grad_norm": 1.6733940839767456, + "learning_rate": 0.00011603899506054034, + "loss": 0.5928, + "step": 26820 + }, + { + "epoch": 1.8611477343695788, + "grad_norm": 1.7673283815383911, + "learning_rate": 0.00011603344915987622, + "loss": 0.4716, + "step": 26821 + }, + { + "epoch": 1.8612171258066754, + "grad_norm": 1.213800072669983, + "learning_rate": 0.00011602790320859638, + "loss": 0.4115, + "step": 26822 + }, + { + "epoch": 1.861286517243772, + "grad_norm": 1.3020596504211426, + "learning_rate": 0.0001160223572067183, + "loss": 0.4986, + "step": 26823 + }, + { + "epoch": 1.8613559086808689, + "grad_norm": 2.402360677719116, + "learning_rate": 0.00011601681115425949, + "loss": 0.651, + "step": 26824 + }, + { + "epoch": 1.8614253001179655, + "grad_norm": 1.5144195556640625, + "learning_rate": 0.00011601126505123747, + "loss": 0.4501, + "step": 26825 + }, + { + "epoch": 1.861494691555062, + "grad_norm": 1.0625211000442505, + "learning_rate": 0.00011600571889766976, + "loss": 0.4358, + "step": 26826 + }, + { + "epoch": 1.861564082992159, + "grad_norm": 1.3860901594161987, + "learning_rate": 0.00011600017269357381, + "loss": 0.3661, + "step": 26827 + }, + { + "epoch": 1.8616334744292553, + "grad_norm": 1.2839903831481934, + "learning_rate": 0.00011599462643896722, + "loss": 0.4445, + "step": 26828 + }, + { + "epoch": 1.8617028658663521, + "grad_norm": 2.3126327991485596, + "learning_rate": 0.00011598908013386743, + "loss": 0.6544, + "step": 26829 + }, + { + "epoch": 1.8617722573034488, + "grad_norm": 1.4498587846755981, + "learning_rate": 0.00011598353377829198, + "loss": 0.5145, + "step": 26830 + }, + { + "epoch": 1.8618416487405454, + "grad_norm": 1.1283942461013794, + "learning_rate": 0.00011597798737225835, + "loss": 0.4628, + "step": 26831 + }, + { + "epoch": 1.8619110401776422, + "grad_norm": 1.2155640125274658, + "learning_rate": 0.0001159724409157841, + "loss": 0.4923, + "step": 26832 + }, + { + "epoch": 1.8619804316147386, + "grad_norm": 1.1233845949172974, + "learning_rate": 0.00011596689440888666, + "loss": 0.3538, + "step": 26833 + }, + { + "epoch": 1.8620498230518354, + "grad_norm": 1.2174293994903564, + "learning_rate": 0.00011596134785158361, + "loss": 0.407, + "step": 26834 + }, + { + "epoch": 1.862119214488932, + "grad_norm": 2.168396234512329, + "learning_rate": 0.00011595580124389245, + "loss": 0.5564, + "step": 26835 + }, + { + "epoch": 1.8621886059260286, + "grad_norm": 2.3641107082366943, + "learning_rate": 0.00011595025458583065, + "loss": 0.5724, + "step": 26836 + }, + { + "epoch": 1.8622579973631255, + "grad_norm": 1.3637372255325317, + "learning_rate": 0.00011594470787741576, + "loss": 0.565, + "step": 26837 + }, + { + "epoch": 1.862327388800222, + "grad_norm": 1.264691948890686, + "learning_rate": 0.00011593916111866527, + "loss": 0.4112, + "step": 26838 + }, + { + "epoch": 1.8623967802373187, + "grad_norm": 1.4016022682189941, + "learning_rate": 0.00011593361430959671, + "loss": 0.4106, + "step": 26839 + }, + { + "epoch": 1.8624661716744155, + "grad_norm": 1.3821520805358887, + "learning_rate": 0.00011592806745022758, + "loss": 0.492, + "step": 26840 + }, + { + "epoch": 1.862535563111512, + "grad_norm": 1.822183609008789, + "learning_rate": 0.00011592252054057536, + "loss": 0.4785, + "step": 26841 + }, + { + "epoch": 1.8626049545486087, + "grad_norm": 1.6429768800735474, + "learning_rate": 0.00011591697358065763, + "loss": 0.5123, + "step": 26842 + }, + { + "epoch": 1.8626743459857054, + "grad_norm": 1.3798929452896118, + "learning_rate": 0.00011591142657049183, + "loss": 0.5271, + "step": 26843 + }, + { + "epoch": 1.862743737422802, + "grad_norm": 2.940831422805786, + "learning_rate": 0.0001159058795100955, + "loss": 0.4726, + "step": 26844 + }, + { + "epoch": 1.8628131288598988, + "grad_norm": 1.486984372138977, + "learning_rate": 0.00011590033239948616, + "loss": 0.6516, + "step": 26845 + }, + { + "epoch": 1.8628825202969952, + "grad_norm": 1.4446251392364502, + "learning_rate": 0.00011589478523868135, + "loss": 0.6557, + "step": 26846 + }, + { + "epoch": 1.862951911734092, + "grad_norm": 1.786402702331543, + "learning_rate": 0.0001158892380276985, + "loss": 0.6155, + "step": 26847 + }, + { + "epoch": 1.8630213031711886, + "grad_norm": 2.151132822036743, + "learning_rate": 0.0001158836907665552, + "loss": 0.6502, + "step": 26848 + }, + { + "epoch": 1.8630906946082852, + "grad_norm": 1.7973688840866089, + "learning_rate": 0.00011587814345526892, + "loss": 0.5419, + "step": 26849 + }, + { + "epoch": 1.863160086045382, + "grad_norm": 1.5398404598236084, + "learning_rate": 0.00011587259609385721, + "loss": 0.7085, + "step": 26850 + }, + { + "epoch": 1.8632294774824787, + "grad_norm": 1.9910322427749634, + "learning_rate": 0.00011586704868233753, + "loss": 0.5927, + "step": 26851 + }, + { + "epoch": 1.8632988689195753, + "grad_norm": 1.7209383249282837, + "learning_rate": 0.0001158615012207274, + "loss": 0.5257, + "step": 26852 + }, + { + "epoch": 1.8633682603566721, + "grad_norm": 1.7941124439239502, + "learning_rate": 0.00011585595370904439, + "loss": 0.5893, + "step": 26853 + }, + { + "epoch": 1.8634376517937685, + "grad_norm": 1.6176481246948242, + "learning_rate": 0.00011585040614730593, + "loss": 0.5829, + "step": 26854 + }, + { + "epoch": 1.8635070432308654, + "grad_norm": 1.640733003616333, + "learning_rate": 0.0001158448585355296, + "loss": 0.5328, + "step": 26855 + }, + { + "epoch": 1.863576434667962, + "grad_norm": 1.9786278009414673, + "learning_rate": 0.00011583931087373292, + "loss": 0.5141, + "step": 26856 + }, + { + "epoch": 1.8636458261050586, + "grad_norm": 2.165137767791748, + "learning_rate": 0.00011583376316193333, + "loss": 0.7113, + "step": 26857 + }, + { + "epoch": 1.8637152175421554, + "grad_norm": 1.5870035886764526, + "learning_rate": 0.00011582821540014839, + "loss": 0.3752, + "step": 26858 + }, + { + "epoch": 1.8637846089792518, + "grad_norm": 1.63341224193573, + "learning_rate": 0.00011582266758839562, + "loss": 0.5646, + "step": 26859 + }, + { + "epoch": 1.8638540004163486, + "grad_norm": 1.624585509300232, + "learning_rate": 0.00011581711972669254, + "loss": 0.553, + "step": 26860 + }, + { + "epoch": 1.8639233918534452, + "grad_norm": 1.9558289051055908, + "learning_rate": 0.00011581157181505661, + "loss": 0.4776, + "step": 26861 + }, + { + "epoch": 1.8639927832905419, + "grad_norm": 2.679807424545288, + "learning_rate": 0.0001158060238535054, + "loss": 0.5405, + "step": 26862 + }, + { + "epoch": 1.8640621747276387, + "grad_norm": 1.9345464706420898, + "learning_rate": 0.00011580047584205643, + "loss": 0.6172, + "step": 26863 + }, + { + "epoch": 1.8641315661647353, + "grad_norm": 1.9396032094955444, + "learning_rate": 0.00011579492778072716, + "loss": 0.5663, + "step": 26864 + }, + { + "epoch": 1.864200957601832, + "grad_norm": 2.217465877532959, + "learning_rate": 0.00011578937966953516, + "loss": 0.6617, + "step": 26865 + }, + { + "epoch": 1.8642703490389287, + "grad_norm": 1.7376213073730469, + "learning_rate": 0.00011578383150849789, + "loss": 0.6435, + "step": 26866 + }, + { + "epoch": 1.8643397404760251, + "grad_norm": 1.503321886062622, + "learning_rate": 0.00011577828329763294, + "loss": 0.4712, + "step": 26867 + }, + { + "epoch": 1.864409131913122, + "grad_norm": 1.5875142812728882, + "learning_rate": 0.00011577273503695776, + "loss": 0.7052, + "step": 26868 + }, + { + "epoch": 1.8644785233502186, + "grad_norm": 1.767134666442871, + "learning_rate": 0.00011576718672648988, + "loss": 0.59, + "step": 26869 + }, + { + "epoch": 1.8645479147873152, + "grad_norm": 1.8796554803848267, + "learning_rate": 0.0001157616383662468, + "loss": 0.505, + "step": 26870 + }, + { + "epoch": 1.864617306224412, + "grad_norm": 1.4856449365615845, + "learning_rate": 0.0001157560899562461, + "loss": 0.6048, + "step": 26871 + }, + { + "epoch": 1.8646866976615084, + "grad_norm": 1.7741565704345703, + "learning_rate": 0.00011575054149650522, + "loss": 0.625, + "step": 26872 + }, + { + "epoch": 1.8647560890986052, + "grad_norm": 2.331692934036255, + "learning_rate": 0.00011574499298704171, + "loss": 0.7064, + "step": 26873 + }, + { + "epoch": 1.8648254805357019, + "grad_norm": 1.831995964050293, + "learning_rate": 0.0001157394444278731, + "loss": 0.6059, + "step": 26874 + }, + { + "epoch": 1.8648948719727985, + "grad_norm": 1.5081427097320557, + "learning_rate": 0.00011573389581901685, + "loss": 0.6101, + "step": 26875 + }, + { + "epoch": 1.8649642634098953, + "grad_norm": 1.7055608034133911, + "learning_rate": 0.00011572834716049054, + "loss": 0.5795, + "step": 26876 + }, + { + "epoch": 1.865033654846992, + "grad_norm": 1.375541090965271, + "learning_rate": 0.00011572279845231171, + "loss": 0.5032, + "step": 26877 + }, + { + "epoch": 1.8651030462840885, + "grad_norm": 1.6484644412994385, + "learning_rate": 0.00011571724969449777, + "loss": 0.3876, + "step": 26878 + }, + { + "epoch": 1.8651724377211854, + "grad_norm": 1.574116587638855, + "learning_rate": 0.00011571170088706631, + "loss": 0.3504, + "step": 26879 + }, + { + "epoch": 1.8652418291582817, + "grad_norm": 1.5434473752975464, + "learning_rate": 0.00011570615203003485, + "loss": 0.4051, + "step": 26880 + }, + { + "epoch": 1.8653112205953786, + "grad_norm": 2.7673733234405518, + "learning_rate": 0.00011570060312342086, + "loss": 0.7888, + "step": 26881 + }, + { + "epoch": 1.8653806120324752, + "grad_norm": 1.8352491855621338, + "learning_rate": 0.00011569505416724191, + "loss": 0.7571, + "step": 26882 + }, + { + "epoch": 1.8654500034695718, + "grad_norm": 2.236114501953125, + "learning_rate": 0.0001156895051615155, + "loss": 0.7145, + "step": 26883 + }, + { + "epoch": 1.8655193949066686, + "grad_norm": 1.5205217599868774, + "learning_rate": 0.00011568395610625916, + "loss": 0.7762, + "step": 26884 + }, + { + "epoch": 1.865588786343765, + "grad_norm": 1.5924623012542725, + "learning_rate": 0.00011567840700149036, + "loss": 0.5818, + "step": 26885 + }, + { + "epoch": 1.8656581777808618, + "grad_norm": 1.5809046030044556, + "learning_rate": 0.00011567285784722663, + "loss": 0.5973, + "step": 26886 + }, + { + "epoch": 1.8657275692179585, + "grad_norm": 1.3417209386825562, + "learning_rate": 0.00011566730864348554, + "loss": 0.509, + "step": 26887 + }, + { + "epoch": 1.865796960655055, + "grad_norm": 1.5150065422058105, + "learning_rate": 0.00011566175939028456, + "loss": 0.6149, + "step": 26888 + }, + { + "epoch": 1.865866352092152, + "grad_norm": 1.4098474979400635, + "learning_rate": 0.00011565621008764123, + "loss": 0.6072, + "step": 26889 + }, + { + "epoch": 1.8659357435292485, + "grad_norm": 1.3396399021148682, + "learning_rate": 0.00011565066073557304, + "loss": 0.5005, + "step": 26890 + }, + { + "epoch": 1.8660051349663451, + "grad_norm": 1.6065043210983276, + "learning_rate": 0.00011564511133409756, + "loss": 0.6605, + "step": 26891 + }, + { + "epoch": 1.866074526403442, + "grad_norm": 1.486178994178772, + "learning_rate": 0.00011563956188323227, + "loss": 0.5476, + "step": 26892 + }, + { + "epoch": 1.8661439178405383, + "grad_norm": 1.6678444147109985, + "learning_rate": 0.00011563401238299467, + "loss": 0.8104, + "step": 26893 + }, + { + "epoch": 1.8662133092776352, + "grad_norm": 2.2450499534606934, + "learning_rate": 0.00011562846283340234, + "loss": 0.5057, + "step": 26894 + }, + { + "epoch": 1.8662827007147318, + "grad_norm": 1.2411631345748901, + "learning_rate": 0.00011562291323447276, + "loss": 0.3969, + "step": 26895 + }, + { + "epoch": 1.8663520921518284, + "grad_norm": 1.6144789457321167, + "learning_rate": 0.00011561736358622344, + "loss": 0.7184, + "step": 26896 + }, + { + "epoch": 1.8664214835889252, + "grad_norm": 1.729364275932312, + "learning_rate": 0.00011561181388867193, + "loss": 0.4743, + "step": 26897 + }, + { + "epoch": 1.8664908750260216, + "grad_norm": 1.4666624069213867, + "learning_rate": 0.00011560626414183573, + "loss": 0.6814, + "step": 26898 + }, + { + "epoch": 1.8665602664631185, + "grad_norm": 1.8949674367904663, + "learning_rate": 0.00011560071434573236, + "loss": 0.5042, + "step": 26899 + }, + { + "epoch": 1.866629657900215, + "grad_norm": 1.173822045326233, + "learning_rate": 0.00011559516450037938, + "loss": 0.4847, + "step": 26900 + }, + { + "epoch": 1.8666990493373117, + "grad_norm": 1.6133021116256714, + "learning_rate": 0.00011558961460579423, + "loss": 0.4891, + "step": 26901 + }, + { + "epoch": 1.8667684407744085, + "grad_norm": 1.541717290878296, + "learning_rate": 0.00011558406466199449, + "loss": 0.5854, + "step": 26902 + }, + { + "epoch": 1.8668378322115051, + "grad_norm": 1.1181234121322632, + "learning_rate": 0.00011557851466899767, + "loss": 0.5469, + "step": 26903 + }, + { + "epoch": 1.8669072236486017, + "grad_norm": 1.4292426109313965, + "learning_rate": 0.00011557296462682128, + "loss": 0.5384, + "step": 26904 + }, + { + "epoch": 1.8669766150856986, + "grad_norm": 1.6341315507888794, + "learning_rate": 0.00011556741453548286, + "loss": 0.857, + "step": 26905 + }, + { + "epoch": 1.867046006522795, + "grad_norm": 1.9572222232818604, + "learning_rate": 0.0001155618643949999, + "loss": 0.6103, + "step": 26906 + }, + { + "epoch": 1.8671153979598918, + "grad_norm": 1.8382667303085327, + "learning_rate": 0.00011555631420538995, + "loss": 0.5985, + "step": 26907 + }, + { + "epoch": 1.8671847893969884, + "grad_norm": 1.5453888177871704, + "learning_rate": 0.00011555076396667053, + "loss": 0.58, + "step": 26908 + }, + { + "epoch": 1.867254180834085, + "grad_norm": 1.6435911655426025, + "learning_rate": 0.00011554521367885913, + "loss": 0.6394, + "step": 26909 + }, + { + "epoch": 1.8673235722711818, + "grad_norm": 1.2775218486785889, + "learning_rate": 0.0001155396633419733, + "loss": 0.4712, + "step": 26910 + }, + { + "epoch": 1.8673929637082785, + "grad_norm": 1.413673758506775, + "learning_rate": 0.00011553411295603058, + "loss": 0.6581, + "step": 26911 + }, + { + "epoch": 1.867462355145375, + "grad_norm": 1.7601815462112427, + "learning_rate": 0.00011552856252104844, + "loss": 0.5411, + "step": 26912 + }, + { + "epoch": 1.8675317465824717, + "grad_norm": 1.3703296184539795, + "learning_rate": 0.00011552301203704443, + "loss": 0.4809, + "step": 26913 + }, + { + "epoch": 1.8676011380195683, + "grad_norm": 1.3722517490386963, + "learning_rate": 0.00011551746150403609, + "loss": 0.4741, + "step": 26914 + }, + { + "epoch": 1.8676705294566651, + "grad_norm": 1.4074832201004028, + "learning_rate": 0.00011551191092204092, + "loss": 0.5278, + "step": 26915 + }, + { + "epoch": 1.8677399208937617, + "grad_norm": 1.6430585384368896, + "learning_rate": 0.00011550636029107645, + "loss": 0.4778, + "step": 26916 + }, + { + "epoch": 1.8678093123308583, + "grad_norm": 1.284704327583313, + "learning_rate": 0.00011550080961116019, + "loss": 0.4427, + "step": 26917 + }, + { + "epoch": 1.8678787037679552, + "grad_norm": 1.6251381635665894, + "learning_rate": 0.00011549525888230968, + "loss": 0.5597, + "step": 26918 + }, + { + "epoch": 1.8679480952050516, + "grad_norm": 1.5185471773147583, + "learning_rate": 0.00011548970810454245, + "loss": 0.5814, + "step": 26919 + }, + { + "epoch": 1.8680174866421484, + "grad_norm": 1.6234501600265503, + "learning_rate": 0.00011548415727787598, + "loss": 0.6854, + "step": 26920 + }, + { + "epoch": 1.868086878079245, + "grad_norm": 3.893650531768799, + "learning_rate": 0.00011547860640232781, + "loss": 0.5134, + "step": 26921 + }, + { + "epoch": 1.8681562695163416, + "grad_norm": 1.7172452211380005, + "learning_rate": 0.00011547305547791551, + "loss": 0.65, + "step": 26922 + }, + { + "epoch": 1.8682256609534384, + "grad_norm": 1.4484854936599731, + "learning_rate": 0.00011546750450465655, + "loss": 0.6265, + "step": 26923 + }, + { + "epoch": 1.868295052390535, + "grad_norm": 1.6072204113006592, + "learning_rate": 0.00011546195348256848, + "loss": 0.6029, + "step": 26924 + }, + { + "epoch": 1.8683644438276317, + "grad_norm": 1.912437081336975, + "learning_rate": 0.00011545640241166882, + "loss": 0.6107, + "step": 26925 + }, + { + "epoch": 1.8684338352647283, + "grad_norm": 1.394391655921936, + "learning_rate": 0.00011545085129197508, + "loss": 0.5812, + "step": 26926 + }, + { + "epoch": 1.868503226701825, + "grad_norm": 1.825885534286499, + "learning_rate": 0.0001154453001235048, + "loss": 0.6464, + "step": 26927 + }, + { + "epoch": 1.8685726181389217, + "grad_norm": 1.6319825649261475, + "learning_rate": 0.00011543974890627549, + "loss": 0.622, + "step": 26928 + }, + { + "epoch": 1.8686420095760183, + "grad_norm": 1.6745747327804565, + "learning_rate": 0.00011543419764030472, + "loss": 0.4812, + "step": 26929 + }, + { + "epoch": 1.868711401013115, + "grad_norm": 1.8871138095855713, + "learning_rate": 0.00011542864632560994, + "loss": 0.7006, + "step": 26930 + }, + { + "epoch": 1.8687807924502118, + "grad_norm": 1.6647756099700928, + "learning_rate": 0.0001154230949622087, + "loss": 0.5682, + "step": 26931 + }, + { + "epoch": 1.8688501838873082, + "grad_norm": 1.6074126958847046, + "learning_rate": 0.00011541754355011861, + "loss": 0.5764, + "step": 26932 + }, + { + "epoch": 1.868919575324405, + "grad_norm": 1.3178648948669434, + "learning_rate": 0.00011541199208935707, + "loss": 0.5251, + "step": 26933 + }, + { + "epoch": 1.8689889667615016, + "grad_norm": 1.2363531589508057, + "learning_rate": 0.00011540644057994166, + "loss": 0.5246, + "step": 26934 + }, + { + "epoch": 1.8690583581985982, + "grad_norm": 2.58958101272583, + "learning_rate": 0.00011540088902188992, + "loss": 0.6147, + "step": 26935 + }, + { + "epoch": 1.869127749635695, + "grad_norm": 2.0794501304626465, + "learning_rate": 0.00011539533741521936, + "loss": 0.5689, + "step": 26936 + }, + { + "epoch": 1.8691971410727917, + "grad_norm": 1.1906852722167969, + "learning_rate": 0.0001153897857599475, + "loss": 0.5282, + "step": 26937 + }, + { + "epoch": 1.8692665325098883, + "grad_norm": 1.9987890720367432, + "learning_rate": 0.00011538423405609187, + "loss": 0.606, + "step": 26938 + }, + { + "epoch": 1.869335923946985, + "grad_norm": 1.0887309312820435, + "learning_rate": 0.00011537868230367002, + "loss": 0.4206, + "step": 26939 + }, + { + "epoch": 1.8694053153840815, + "grad_norm": 1.4226523637771606, + "learning_rate": 0.00011537313050269942, + "loss": 0.6202, + "step": 26940 + }, + { + "epoch": 1.8694747068211783, + "grad_norm": 1.4967433214187622, + "learning_rate": 0.00011536757865319763, + "loss": 0.5348, + "step": 26941 + }, + { + "epoch": 1.869544098258275, + "grad_norm": 1.2895432710647583, + "learning_rate": 0.0001153620267551822, + "loss": 0.3569, + "step": 26942 + }, + { + "epoch": 1.8696134896953716, + "grad_norm": 2.297852039337158, + "learning_rate": 0.00011535647480867062, + "loss": 0.7979, + "step": 26943 + }, + { + "epoch": 1.8696828811324684, + "grad_norm": 1.6580760478973389, + "learning_rate": 0.00011535092281368042, + "loss": 0.561, + "step": 26944 + }, + { + "epoch": 1.8697522725695648, + "grad_norm": 1.8857693672180176, + "learning_rate": 0.00011534537077022915, + "loss": 0.5927, + "step": 26945 + }, + { + "epoch": 1.8698216640066616, + "grad_norm": 1.4539426565170288, + "learning_rate": 0.00011533981867833437, + "loss": 0.5218, + "step": 26946 + }, + { + "epoch": 1.8698910554437582, + "grad_norm": 1.467353105545044, + "learning_rate": 0.00011533426653801349, + "loss": 0.5984, + "step": 26947 + }, + { + "epoch": 1.8699604468808548, + "grad_norm": 2.120427370071411, + "learning_rate": 0.00011532871434928414, + "loss": 0.6937, + "step": 26948 + }, + { + "epoch": 1.8700298383179517, + "grad_norm": 1.905296802520752, + "learning_rate": 0.00011532316211216383, + "loss": 0.5161, + "step": 26949 + }, + { + "epoch": 1.8700992297550483, + "grad_norm": 1.7324110269546509, + "learning_rate": 0.00011531760982667006, + "loss": 0.4905, + "step": 26950 + }, + { + "epoch": 1.8701686211921449, + "grad_norm": 1.645552635192871, + "learning_rate": 0.00011531205749282037, + "loss": 0.3458, + "step": 26951 + }, + { + "epoch": 1.8702380126292415, + "grad_norm": 1.435165524482727, + "learning_rate": 0.0001153065051106323, + "loss": 0.4849, + "step": 26952 + }, + { + "epoch": 1.870307404066338, + "grad_norm": 1.2400413751602173, + "learning_rate": 0.00011530095268012337, + "loss": 0.4161, + "step": 26953 + }, + { + "epoch": 1.870376795503435, + "grad_norm": 2.033813238143921, + "learning_rate": 0.00011529540020131111, + "loss": 0.9268, + "step": 26954 + }, + { + "epoch": 1.8704461869405316, + "grad_norm": 1.293705701828003, + "learning_rate": 0.00011528984767421302, + "loss": 0.4657, + "step": 26955 + }, + { + "epoch": 1.8705155783776282, + "grad_norm": 2.141664981842041, + "learning_rate": 0.0001152842950988467, + "loss": 0.841, + "step": 26956 + }, + { + "epoch": 1.870584969814725, + "grad_norm": 1.509031891822815, + "learning_rate": 0.0001152787424752296, + "loss": 0.5608, + "step": 26957 + }, + { + "epoch": 1.8706543612518214, + "grad_norm": 1.7478582859039307, + "learning_rate": 0.00011527318980337929, + "loss": 0.4829, + "step": 26958 + }, + { + "epoch": 1.8707237526889182, + "grad_norm": 2.360736846923828, + "learning_rate": 0.00011526763708331327, + "loss": 0.5369, + "step": 26959 + }, + { + "epoch": 1.8707931441260148, + "grad_norm": 1.2061519622802734, + "learning_rate": 0.00011526208431504915, + "loss": 0.5604, + "step": 26960 + }, + { + "epoch": 1.8708625355631114, + "grad_norm": 1.4863535165786743, + "learning_rate": 0.00011525653149860436, + "loss": 0.6307, + "step": 26961 + }, + { + "epoch": 1.8709319270002083, + "grad_norm": 1.5986032485961914, + "learning_rate": 0.00011525097863399645, + "loss": 0.6316, + "step": 26962 + }, + { + "epoch": 1.8710013184373049, + "grad_norm": 1.2562090158462524, + "learning_rate": 0.00011524542572124302, + "loss": 0.5025, + "step": 26963 + }, + { + "epoch": 1.8710707098744015, + "grad_norm": 1.6487212181091309, + "learning_rate": 0.00011523987276036153, + "loss": 0.5999, + "step": 26964 + }, + { + "epoch": 1.8711401013114983, + "grad_norm": 1.2898106575012207, + "learning_rate": 0.0001152343197513695, + "loss": 0.4929, + "step": 26965 + }, + { + "epoch": 1.8712094927485947, + "grad_norm": 1.6830353736877441, + "learning_rate": 0.0001152287666942845, + "loss": 0.6604, + "step": 26966 + }, + { + "epoch": 1.8712788841856915, + "grad_norm": 1.5819673538208008, + "learning_rate": 0.0001152232135891241, + "loss": 0.6156, + "step": 26967 + }, + { + "epoch": 1.8713482756227882, + "grad_norm": 2.423880100250244, + "learning_rate": 0.00011521766043590572, + "loss": 0.6969, + "step": 26968 + }, + { + "epoch": 1.8714176670598848, + "grad_norm": 1.6182384490966797, + "learning_rate": 0.00011521210723464696, + "loss": 0.7707, + "step": 26969 + }, + { + "epoch": 1.8714870584969816, + "grad_norm": 1.3142361640930176, + "learning_rate": 0.0001152065539853654, + "loss": 0.4455, + "step": 26970 + }, + { + "epoch": 1.871556449934078, + "grad_norm": 1.752894639968872, + "learning_rate": 0.00011520100068807847, + "loss": 0.4591, + "step": 26971 + }, + { + "epoch": 1.8716258413711748, + "grad_norm": 1.313173770904541, + "learning_rate": 0.00011519544734280375, + "loss": 0.4415, + "step": 26972 + }, + { + "epoch": 1.8716952328082714, + "grad_norm": 1.9165972471237183, + "learning_rate": 0.00011518989394955874, + "loss": 0.5723, + "step": 26973 + }, + { + "epoch": 1.871764624245368, + "grad_norm": 1.6311839818954468, + "learning_rate": 0.00011518434050836105, + "loss": 0.5008, + "step": 26974 + }, + { + "epoch": 1.8718340156824649, + "grad_norm": 1.2875796556472778, + "learning_rate": 0.00011517878701922811, + "loss": 0.5117, + "step": 26975 + }, + { + "epoch": 1.8719034071195615, + "grad_norm": 1.1704884767532349, + "learning_rate": 0.00011517323348217752, + "loss": 0.48, + "step": 26976 + }, + { + "epoch": 1.871972798556658, + "grad_norm": 1.3109259605407715, + "learning_rate": 0.0001151676798972268, + "loss": 0.4686, + "step": 26977 + }, + { + "epoch": 1.872042189993755, + "grad_norm": 1.3724339008331299, + "learning_rate": 0.00011516212626439346, + "loss": 0.5891, + "step": 26978 + }, + { + "epoch": 1.8721115814308513, + "grad_norm": 1.7068589925765991, + "learning_rate": 0.00011515657258369505, + "loss": 0.697, + "step": 26979 + }, + { + "epoch": 1.8721809728679482, + "grad_norm": 1.547316551208496, + "learning_rate": 0.0001151510188551491, + "loss": 0.5922, + "step": 26980 + }, + { + "epoch": 1.8722503643050448, + "grad_norm": 1.3019107580184937, + "learning_rate": 0.00011514546507877315, + "loss": 0.4205, + "step": 26981 + }, + { + "epoch": 1.8723197557421414, + "grad_norm": 1.312319278717041, + "learning_rate": 0.0001151399112545847, + "loss": 0.5583, + "step": 26982 + }, + { + "epoch": 1.8723891471792382, + "grad_norm": 1.5897730588912964, + "learning_rate": 0.00011513435738260132, + "loss": 0.5901, + "step": 26983 + }, + { + "epoch": 1.8724585386163346, + "grad_norm": 1.210067868232727, + "learning_rate": 0.00011512880346284056, + "loss": 0.5905, + "step": 26984 + }, + { + "epoch": 1.8725279300534314, + "grad_norm": 1.4710592031478882, + "learning_rate": 0.00011512324949531987, + "loss": 0.6871, + "step": 26985 + }, + { + "epoch": 1.872597321490528, + "grad_norm": 2.1903109550476074, + "learning_rate": 0.00011511769548005686, + "loss": 0.6359, + "step": 26986 + }, + { + "epoch": 1.8726667129276247, + "grad_norm": 1.5142791271209717, + "learning_rate": 0.00011511214141706905, + "loss": 0.6102, + "step": 26987 + }, + { + "epoch": 1.8727361043647215, + "grad_norm": 1.2582452297210693, + "learning_rate": 0.00011510658730637396, + "loss": 0.426, + "step": 26988 + }, + { + "epoch": 1.872805495801818, + "grad_norm": 1.672755479812622, + "learning_rate": 0.00011510103314798912, + "loss": 0.5476, + "step": 26989 + }, + { + "epoch": 1.8728748872389147, + "grad_norm": 1.4998530149459839, + "learning_rate": 0.00011509547894193206, + "loss": 0.6068, + "step": 26990 + }, + { + "epoch": 1.8729442786760115, + "grad_norm": 1.8892494440078735, + "learning_rate": 0.00011508992468822036, + "loss": 0.462, + "step": 26991 + }, + { + "epoch": 1.873013670113108, + "grad_norm": 1.395640254020691, + "learning_rate": 0.0001150843703868715, + "loss": 0.365, + "step": 26992 + }, + { + "epoch": 1.8730830615502048, + "grad_norm": 1.9971575736999512, + "learning_rate": 0.00011507881603790301, + "loss": 0.6953, + "step": 26993 + }, + { + "epoch": 1.8731524529873014, + "grad_norm": 1.5233519077301025, + "learning_rate": 0.00011507326164133247, + "loss": 0.4456, + "step": 26994 + }, + { + "epoch": 1.873221844424398, + "grad_norm": 1.1753320693969727, + "learning_rate": 0.0001150677071971774, + "loss": 0.6438, + "step": 26995 + }, + { + "epoch": 1.8732912358614948, + "grad_norm": 1.747014045715332, + "learning_rate": 0.00011506215270545529, + "loss": 0.4016, + "step": 26996 + }, + { + "epoch": 1.8733606272985912, + "grad_norm": 1.2899631261825562, + "learning_rate": 0.00011505659816618373, + "loss": 0.5098, + "step": 26997 + }, + { + "epoch": 1.873430018735688, + "grad_norm": 1.6451425552368164, + "learning_rate": 0.00011505104357938026, + "loss": 0.6137, + "step": 26998 + }, + { + "epoch": 1.8734994101727847, + "grad_norm": 1.2145284414291382, + "learning_rate": 0.00011504548894506236, + "loss": 0.4811, + "step": 26999 + }, + { + "epoch": 1.8735688016098813, + "grad_norm": 1.5308866500854492, + "learning_rate": 0.0001150399342632476, + "loss": 0.656, + "step": 27000 + }, + { + "epoch": 1.873638193046978, + "grad_norm": 1.7632091045379639, + "learning_rate": 0.00011503437953395352, + "loss": 0.6877, + "step": 27001 + }, + { + "epoch": 1.8737075844840747, + "grad_norm": 1.4905219078063965, + "learning_rate": 0.00011502882475719765, + "loss": 0.5363, + "step": 27002 + }, + { + "epoch": 1.8737769759211713, + "grad_norm": 1.3639793395996094, + "learning_rate": 0.00011502326993299753, + "loss": 0.5269, + "step": 27003 + }, + { + "epoch": 1.8738463673582682, + "grad_norm": 1.7315682172775269, + "learning_rate": 0.00011501771506137067, + "loss": 0.5106, + "step": 27004 + }, + { + "epoch": 1.8739157587953645, + "grad_norm": 1.1550780534744263, + "learning_rate": 0.00011501216014233464, + "loss": 0.4172, + "step": 27005 + }, + { + "epoch": 1.8739851502324614, + "grad_norm": 1.494346261024475, + "learning_rate": 0.00011500660517590698, + "loss": 0.6074, + "step": 27006 + }, + { + "epoch": 1.874054541669558, + "grad_norm": 1.4169220924377441, + "learning_rate": 0.00011500105016210519, + "loss": 0.6267, + "step": 27007 + }, + { + "epoch": 1.8741239331066546, + "grad_norm": 1.476222276687622, + "learning_rate": 0.00011499549510094683, + "loss": 0.6046, + "step": 27008 + }, + { + "epoch": 1.8741933245437514, + "grad_norm": 2.0565412044525146, + "learning_rate": 0.00011498993999244943, + "loss": 0.5079, + "step": 27009 + }, + { + "epoch": 1.8742627159808478, + "grad_norm": 1.4652595520019531, + "learning_rate": 0.00011498438483663052, + "loss": 0.6508, + "step": 27010 + }, + { + "epoch": 1.8743321074179446, + "grad_norm": 3.304668664932251, + "learning_rate": 0.00011497882963350764, + "loss": 0.6262, + "step": 27011 + }, + { + "epoch": 1.8744014988550413, + "grad_norm": 1.253269910812378, + "learning_rate": 0.00011497327438309835, + "loss": 0.5384, + "step": 27012 + }, + { + "epoch": 1.8744708902921379, + "grad_norm": 1.9096159934997559, + "learning_rate": 0.00011496771908542017, + "loss": 0.9096, + "step": 27013 + }, + { + "epoch": 1.8745402817292347, + "grad_norm": 2.1173534393310547, + "learning_rate": 0.0001149621637404906, + "loss": 0.8592, + "step": 27014 + }, + { + "epoch": 1.8746096731663313, + "grad_norm": 1.1670268774032593, + "learning_rate": 0.00011495660834832727, + "loss": 0.5457, + "step": 27015 + }, + { + "epoch": 1.874679064603428, + "grad_norm": 1.6221644878387451, + "learning_rate": 0.00011495105290894763, + "loss": 0.7123, + "step": 27016 + }, + { + "epoch": 1.8747484560405248, + "grad_norm": 1.508751630783081, + "learning_rate": 0.00011494549742236926, + "loss": 0.4375, + "step": 27017 + }, + { + "epoch": 1.8748178474776211, + "grad_norm": 1.4773918390274048, + "learning_rate": 0.00011493994188860969, + "loss": 0.5396, + "step": 27018 + }, + { + "epoch": 1.874887238914718, + "grad_norm": 1.3590244054794312, + "learning_rate": 0.00011493438630768645, + "loss": 0.4884, + "step": 27019 + }, + { + "epoch": 1.8749566303518146, + "grad_norm": 1.4939885139465332, + "learning_rate": 0.00011492883067961709, + "loss": 0.6399, + "step": 27020 + }, + { + "epoch": 1.8750260217889112, + "grad_norm": 1.2066724300384521, + "learning_rate": 0.00011492327500441912, + "loss": 0.48, + "step": 27021 + }, + { + "epoch": 1.875095413226008, + "grad_norm": 2.151228427886963, + "learning_rate": 0.00011491771928211015, + "loss": 0.8665, + "step": 27022 + }, + { + "epoch": 1.8751648046631044, + "grad_norm": 1.3661084175109863, + "learning_rate": 0.00011491216351270763, + "loss": 0.5547, + "step": 27023 + }, + { + "epoch": 1.8752341961002013, + "grad_norm": 1.501739263534546, + "learning_rate": 0.00011490660769622918, + "loss": 0.7486, + "step": 27024 + }, + { + "epoch": 1.8753035875372979, + "grad_norm": 1.2083078622817993, + "learning_rate": 0.00011490105183269229, + "loss": 0.4376, + "step": 27025 + }, + { + "epoch": 1.8753729789743945, + "grad_norm": 1.594436526298523, + "learning_rate": 0.00011489549592211451, + "loss": 0.4741, + "step": 27026 + }, + { + "epoch": 1.8754423704114913, + "grad_norm": 1.2048083543777466, + "learning_rate": 0.00011488993996451333, + "loss": 0.5403, + "step": 27027 + }, + { + "epoch": 1.875511761848588, + "grad_norm": 2.2228691577911377, + "learning_rate": 0.00011488438395990637, + "loss": 0.7729, + "step": 27028 + }, + { + "epoch": 1.8755811532856845, + "grad_norm": 1.410815715789795, + "learning_rate": 0.00011487882790831114, + "loss": 0.7859, + "step": 27029 + }, + { + "epoch": 1.8756505447227814, + "grad_norm": 1.315637469291687, + "learning_rate": 0.00011487327180974516, + "loss": 0.4426, + "step": 27030 + }, + { + "epoch": 1.8757199361598778, + "grad_norm": 1.6515189409255981, + "learning_rate": 0.00011486771566422601, + "loss": 0.7283, + "step": 27031 + }, + { + "epoch": 1.8757893275969746, + "grad_norm": 1.4320282936096191, + "learning_rate": 0.0001148621594717712, + "loss": 0.5356, + "step": 27032 + }, + { + "epoch": 1.8758587190340712, + "grad_norm": 1.7502470016479492, + "learning_rate": 0.00011485660323239826, + "loss": 0.7697, + "step": 27033 + }, + { + "epoch": 1.8759281104711678, + "grad_norm": 1.6094082593917847, + "learning_rate": 0.00011485104694612476, + "loss": 0.5837, + "step": 27034 + }, + { + "epoch": 1.8759975019082646, + "grad_norm": 1.1750510931015015, + "learning_rate": 0.00011484549061296821, + "loss": 0.404, + "step": 27035 + }, + { + "epoch": 1.876066893345361, + "grad_norm": 2.092888116836548, + "learning_rate": 0.00011483993423294621, + "loss": 0.509, + "step": 27036 + }, + { + "epoch": 1.8761362847824579, + "grad_norm": 2.1731603145599365, + "learning_rate": 0.00011483437780607622, + "loss": 0.5964, + "step": 27037 + }, + { + "epoch": 1.8762056762195545, + "grad_norm": 1.6963236331939697, + "learning_rate": 0.00011482882133237583, + "loss": 0.6504, + "step": 27038 + }, + { + "epoch": 1.876275067656651, + "grad_norm": 1.9580906629562378, + "learning_rate": 0.00011482326481186259, + "loss": 0.5645, + "step": 27039 + }, + { + "epoch": 1.876344459093748, + "grad_norm": 1.58641517162323, + "learning_rate": 0.000114817708244554, + "loss": 0.3783, + "step": 27040 + }, + { + "epoch": 1.8764138505308445, + "grad_norm": 1.0468239784240723, + "learning_rate": 0.00011481215163046766, + "loss": 0.423, + "step": 27041 + }, + { + "epoch": 1.8764832419679411, + "grad_norm": 1.8039395809173584, + "learning_rate": 0.00011480659496962103, + "loss": 0.7414, + "step": 27042 + }, + { + "epoch": 1.876552633405038, + "grad_norm": 1.257078766822815, + "learning_rate": 0.00011480103826203174, + "loss": 0.5785, + "step": 27043 + }, + { + "epoch": 1.8766220248421344, + "grad_norm": 1.8606886863708496, + "learning_rate": 0.00011479548150771726, + "loss": 0.5457, + "step": 27044 + }, + { + "epoch": 1.8766914162792312, + "grad_norm": 2.950422525405884, + "learning_rate": 0.00011478992470669516, + "loss": 0.6773, + "step": 27045 + }, + { + "epoch": 1.8767608077163278, + "grad_norm": 1.2065266370773315, + "learning_rate": 0.00011478436785898301, + "loss": 0.3565, + "step": 27046 + }, + { + "epoch": 1.8768301991534244, + "grad_norm": 1.4086856842041016, + "learning_rate": 0.00011477881096459828, + "loss": 0.4256, + "step": 27047 + }, + { + "epoch": 1.8768995905905212, + "grad_norm": 2.0691580772399902, + "learning_rate": 0.00011477325402355858, + "loss": 0.7367, + "step": 27048 + }, + { + "epoch": 1.8769689820276176, + "grad_norm": 1.8661168813705444, + "learning_rate": 0.00011476769703588144, + "loss": 0.4984, + "step": 27049 + }, + { + "epoch": 1.8770383734647145, + "grad_norm": 1.4809015989303589, + "learning_rate": 0.0001147621400015844, + "loss": 0.5758, + "step": 27050 + }, + { + "epoch": 1.877107764901811, + "grad_norm": 1.0997021198272705, + "learning_rate": 0.00011475658292068499, + "loss": 0.5928, + "step": 27051 + }, + { + "epoch": 1.8771771563389077, + "grad_norm": 1.42896568775177, + "learning_rate": 0.00011475102579320075, + "loss": 0.5397, + "step": 27052 + }, + { + "epoch": 1.8772465477760045, + "grad_norm": 2.109111785888672, + "learning_rate": 0.00011474546861914925, + "loss": 0.5955, + "step": 27053 + }, + { + "epoch": 1.8773159392131011, + "grad_norm": 1.4455451965332031, + "learning_rate": 0.00011473991139854801, + "loss": 0.4379, + "step": 27054 + }, + { + "epoch": 1.8773853306501977, + "grad_norm": 1.7803599834442139, + "learning_rate": 0.00011473435413141456, + "loss": 0.766, + "step": 27055 + }, + { + "epoch": 1.8774547220872946, + "grad_norm": 1.4428677558898926, + "learning_rate": 0.00011472879681776648, + "loss": 0.5934, + "step": 27056 + }, + { + "epoch": 1.877524113524391, + "grad_norm": 1.1060268878936768, + "learning_rate": 0.00011472323945762131, + "loss": 0.3482, + "step": 27057 + }, + { + "epoch": 1.8775935049614878, + "grad_norm": 2.1864919662475586, + "learning_rate": 0.00011471768205099655, + "loss": 0.6696, + "step": 27058 + }, + { + "epoch": 1.8776628963985844, + "grad_norm": 1.1766852140426636, + "learning_rate": 0.00011471212459790981, + "loss": 0.4422, + "step": 27059 + }, + { + "epoch": 1.877732287835681, + "grad_norm": 1.3613910675048828, + "learning_rate": 0.00011470656709837859, + "loss": 0.564, + "step": 27060 + }, + { + "epoch": 1.8778016792727779, + "grad_norm": 1.5180381536483765, + "learning_rate": 0.00011470100955242044, + "loss": 0.5509, + "step": 27061 + }, + { + "epoch": 1.8778710707098742, + "grad_norm": 1.3645554780960083, + "learning_rate": 0.00011469545196005289, + "loss": 0.6544, + "step": 27062 + }, + { + "epoch": 1.877940462146971, + "grad_norm": 1.1011945009231567, + "learning_rate": 0.00011468989432129353, + "loss": 0.3773, + "step": 27063 + }, + { + "epoch": 1.8780098535840677, + "grad_norm": 1.6738039255142212, + "learning_rate": 0.00011468433663615987, + "loss": 0.5458, + "step": 27064 + }, + { + "epoch": 1.8780792450211643, + "grad_norm": 1.5101991891860962, + "learning_rate": 0.00011467877890466944, + "loss": 0.5647, + "step": 27065 + }, + { + "epoch": 1.8781486364582611, + "grad_norm": 1.90489661693573, + "learning_rate": 0.00011467322112683985, + "loss": 0.9041, + "step": 27066 + }, + { + "epoch": 1.8782180278953577, + "grad_norm": 1.260510802268982, + "learning_rate": 0.00011466766330268859, + "loss": 0.4748, + "step": 27067 + }, + { + "epoch": 1.8782874193324544, + "grad_norm": 1.4064708948135376, + "learning_rate": 0.0001146621054322332, + "loss": 0.4268, + "step": 27068 + }, + { + "epoch": 1.8783568107695512, + "grad_norm": 1.758375883102417, + "learning_rate": 0.00011465654751549125, + "loss": 0.7117, + "step": 27069 + }, + { + "epoch": 1.8784262022066476, + "grad_norm": 1.552359938621521, + "learning_rate": 0.0001146509895524803, + "loss": 0.6649, + "step": 27070 + }, + { + "epoch": 1.8784955936437444, + "grad_norm": 1.321242094039917, + "learning_rate": 0.00011464543154321785, + "loss": 0.4354, + "step": 27071 + }, + { + "epoch": 1.878564985080841, + "grad_norm": 1.9316188097000122, + "learning_rate": 0.00011463987348772147, + "loss": 0.684, + "step": 27072 + }, + { + "epoch": 1.8786343765179376, + "grad_norm": 1.5717591047286987, + "learning_rate": 0.0001146343153860087, + "loss": 0.6896, + "step": 27073 + }, + { + "epoch": 1.8787037679550345, + "grad_norm": 1.3501142263412476, + "learning_rate": 0.00011462875723809714, + "loss": 0.4656, + "step": 27074 + }, + { + "epoch": 1.8787731593921309, + "grad_norm": 1.075613021850586, + "learning_rate": 0.00011462319904400426, + "loss": 0.5287, + "step": 27075 + }, + { + "epoch": 1.8788425508292277, + "grad_norm": 2.2284059524536133, + "learning_rate": 0.00011461764080374766, + "loss": 0.9403, + "step": 27076 + }, + { + "epoch": 1.8789119422663243, + "grad_norm": 1.5986491441726685, + "learning_rate": 0.00011461208251734484, + "loss": 0.4896, + "step": 27077 + }, + { + "epoch": 1.878981333703421, + "grad_norm": 1.7689462900161743, + "learning_rate": 0.00011460652418481339, + "loss": 0.4795, + "step": 27078 + }, + { + "epoch": 1.8790507251405177, + "grad_norm": 1.5682424306869507, + "learning_rate": 0.00011460096580617082, + "loss": 0.4393, + "step": 27079 + }, + { + "epoch": 1.8791201165776144, + "grad_norm": 1.224554419517517, + "learning_rate": 0.00011459540738143468, + "loss": 0.4514, + "step": 27080 + }, + { + "epoch": 1.879189508014711, + "grad_norm": 1.4232441186904907, + "learning_rate": 0.00011458984891062257, + "loss": 0.4931, + "step": 27081 + }, + { + "epoch": 1.8792588994518078, + "grad_norm": 1.5166983604431152, + "learning_rate": 0.00011458429039375199, + "loss": 0.6871, + "step": 27082 + }, + { + "epoch": 1.8793282908889042, + "grad_norm": 1.3265653848648071, + "learning_rate": 0.00011457873183084048, + "loss": 0.4246, + "step": 27083 + }, + { + "epoch": 1.879397682326001, + "grad_norm": 1.937994360923767, + "learning_rate": 0.00011457317322190561, + "loss": 0.7285, + "step": 27084 + }, + { + "epoch": 1.8794670737630976, + "grad_norm": 1.350217580795288, + "learning_rate": 0.00011456761456696492, + "loss": 0.4019, + "step": 27085 + }, + { + "epoch": 1.8795364652001942, + "grad_norm": 1.6970787048339844, + "learning_rate": 0.00011456205586603596, + "loss": 0.5065, + "step": 27086 + }, + { + "epoch": 1.879605856637291, + "grad_norm": 1.9417394399642944, + "learning_rate": 0.0001145564971191363, + "loss": 0.616, + "step": 27087 + }, + { + "epoch": 1.8796752480743875, + "grad_norm": 2.233135223388672, + "learning_rate": 0.00011455093832628345, + "loss": 0.7605, + "step": 27088 + }, + { + "epoch": 1.8797446395114843, + "grad_norm": 1.682513952255249, + "learning_rate": 0.00011454537948749497, + "loss": 0.5377, + "step": 27089 + }, + { + "epoch": 1.879814030948581, + "grad_norm": 1.6644073724746704, + "learning_rate": 0.00011453982060278843, + "loss": 0.5468, + "step": 27090 + }, + { + "epoch": 1.8798834223856775, + "grad_norm": 1.2903106212615967, + "learning_rate": 0.00011453426167218137, + "loss": 0.5089, + "step": 27091 + }, + { + "epoch": 1.8799528138227743, + "grad_norm": 1.3052759170532227, + "learning_rate": 0.00011452870269569132, + "loss": 0.4575, + "step": 27092 + }, + { + "epoch": 1.880022205259871, + "grad_norm": 1.0567842721939087, + "learning_rate": 0.00011452314367333586, + "loss": 0.488, + "step": 27093 + }, + { + "epoch": 1.8800915966969676, + "grad_norm": 1.924118161201477, + "learning_rate": 0.00011451758460513249, + "loss": 0.5365, + "step": 27094 + }, + { + "epoch": 1.8801609881340644, + "grad_norm": 1.1436067819595337, + "learning_rate": 0.0001145120254910988, + "loss": 0.5528, + "step": 27095 + }, + { + "epoch": 1.8802303795711608, + "grad_norm": 1.277494192123413, + "learning_rate": 0.00011450646633125233, + "loss": 0.5681, + "step": 27096 + }, + { + "epoch": 1.8802997710082576, + "grad_norm": 1.4253491163253784, + "learning_rate": 0.00011450090712561063, + "loss": 0.5193, + "step": 27097 + }, + { + "epoch": 1.8803691624453542, + "grad_norm": 1.5698699951171875, + "learning_rate": 0.00011449534787419125, + "loss": 0.6254, + "step": 27098 + }, + { + "epoch": 1.8804385538824508, + "grad_norm": 2.163681983947754, + "learning_rate": 0.00011448978857701174, + "loss": 0.5225, + "step": 27099 + }, + { + "epoch": 1.8805079453195477, + "grad_norm": 1.189383864402771, + "learning_rate": 0.00011448422923408963, + "loss": 0.5396, + "step": 27100 + }, + { + "epoch": 1.880577336756644, + "grad_norm": 1.5145139694213867, + "learning_rate": 0.00011447866984544251, + "loss": 0.4435, + "step": 27101 + }, + { + "epoch": 1.880646728193741, + "grad_norm": 1.904976725578308, + "learning_rate": 0.00011447311041108792, + "loss": 0.675, + "step": 27102 + }, + { + "epoch": 1.8807161196308375, + "grad_norm": 1.918602705001831, + "learning_rate": 0.00011446755093104338, + "loss": 0.6915, + "step": 27103 + }, + { + "epoch": 1.8807855110679341, + "grad_norm": 1.4086309671401978, + "learning_rate": 0.00011446199140532645, + "loss": 0.5511, + "step": 27104 + }, + { + "epoch": 1.880854902505031, + "grad_norm": 1.4119162559509277, + "learning_rate": 0.00011445643183395472, + "loss": 0.4384, + "step": 27105 + }, + { + "epoch": 1.8809242939421276, + "grad_norm": 1.7339451313018799, + "learning_rate": 0.00011445087221694569, + "loss": 0.6088, + "step": 27106 + }, + { + "epoch": 1.8809936853792242, + "grad_norm": 1.6532059907913208, + "learning_rate": 0.00011444531255431692, + "loss": 0.642, + "step": 27107 + }, + { + "epoch": 1.881063076816321, + "grad_norm": 1.6963523626327515, + "learning_rate": 0.00011443975284608602, + "loss": 0.52, + "step": 27108 + }, + { + "epoch": 1.8811324682534174, + "grad_norm": 2.4037904739379883, + "learning_rate": 0.00011443419309227045, + "loss": 0.4023, + "step": 27109 + }, + { + "epoch": 1.8812018596905142, + "grad_norm": 2.725071907043457, + "learning_rate": 0.00011442863329288783, + "loss": 0.6396, + "step": 27110 + }, + { + "epoch": 1.8812712511276108, + "grad_norm": 1.6539251804351807, + "learning_rate": 0.00011442307344795568, + "loss": 0.6336, + "step": 27111 + }, + { + "epoch": 1.8813406425647075, + "grad_norm": 1.0870810747146606, + "learning_rate": 0.00011441751355749158, + "loss": 0.4821, + "step": 27112 + }, + { + "epoch": 1.8814100340018043, + "grad_norm": 1.2605265378952026, + "learning_rate": 0.00011441195362151304, + "loss": 0.4713, + "step": 27113 + }, + { + "epoch": 1.8814794254389007, + "grad_norm": 1.2396854162216187, + "learning_rate": 0.00011440639364003764, + "loss": 0.5982, + "step": 27114 + }, + { + "epoch": 1.8815488168759975, + "grad_norm": 1.6901530027389526, + "learning_rate": 0.00011440083361308296, + "loss": 0.544, + "step": 27115 + }, + { + "epoch": 1.8816182083130941, + "grad_norm": 1.3767976760864258, + "learning_rate": 0.00011439527354066648, + "loss": 0.4864, + "step": 27116 + }, + { + "epoch": 1.8816875997501907, + "grad_norm": 1.9132498502731323, + "learning_rate": 0.0001143897134228058, + "loss": 0.7118, + "step": 27117 + }, + { + "epoch": 1.8817569911872876, + "grad_norm": 1.5422157049179077, + "learning_rate": 0.00011438415325951846, + "loss": 0.5144, + "step": 27118 + }, + { + "epoch": 1.8818263826243842, + "grad_norm": 1.532369613647461, + "learning_rate": 0.00011437859305082201, + "loss": 0.4753, + "step": 27119 + }, + { + "epoch": 1.8818957740614808, + "grad_norm": 1.4083101749420166, + "learning_rate": 0.00011437303279673401, + "loss": 0.5685, + "step": 27120 + }, + { + "epoch": 1.8819651654985776, + "grad_norm": 1.0612425804138184, + "learning_rate": 0.00011436747249727205, + "loss": 0.4602, + "step": 27121 + }, + { + "epoch": 1.882034556935674, + "grad_norm": 1.8311818838119507, + "learning_rate": 0.00011436191215245362, + "loss": 0.7583, + "step": 27122 + }, + { + "epoch": 1.8821039483727708, + "grad_norm": 1.4091061353683472, + "learning_rate": 0.0001143563517622963, + "loss": 0.6218, + "step": 27123 + }, + { + "epoch": 1.8821733398098675, + "grad_norm": 1.438303828239441, + "learning_rate": 0.00011435079132681763, + "loss": 0.4928, + "step": 27124 + }, + { + "epoch": 1.882242731246964, + "grad_norm": 1.535671353340149, + "learning_rate": 0.00011434523084603519, + "loss": 0.5719, + "step": 27125 + }, + { + "epoch": 1.882312122684061, + "grad_norm": 1.4390223026275635, + "learning_rate": 0.00011433967031996653, + "loss": 0.5444, + "step": 27126 + }, + { + "epoch": 1.8823815141211573, + "grad_norm": 1.2140705585479736, + "learning_rate": 0.00011433410974862916, + "loss": 0.4201, + "step": 27127 + }, + { + "epoch": 1.8824509055582541, + "grad_norm": 3.199066638946533, + "learning_rate": 0.00011432854913204072, + "loss": 0.9007, + "step": 27128 + }, + { + "epoch": 1.8825202969953507, + "grad_norm": 1.700296401977539, + "learning_rate": 0.00011432298847021868, + "loss": 0.5741, + "step": 27129 + }, + { + "epoch": 1.8825896884324473, + "grad_norm": 1.5459924936294556, + "learning_rate": 0.00011431742776318064, + "loss": 0.576, + "step": 27130 + }, + { + "epoch": 1.8826590798695442, + "grad_norm": 1.074575424194336, + "learning_rate": 0.00011431186701094414, + "loss": 0.4629, + "step": 27131 + }, + { + "epoch": 1.8827284713066408, + "grad_norm": 1.3503916263580322, + "learning_rate": 0.00011430630621352672, + "loss": 0.6063, + "step": 27132 + }, + { + "epoch": 1.8827978627437374, + "grad_norm": 1.6488769054412842, + "learning_rate": 0.00011430074537094597, + "loss": 0.4751, + "step": 27133 + }, + { + "epoch": 1.8828672541808342, + "grad_norm": 2.308997631072998, + "learning_rate": 0.00011429518448321941, + "loss": 0.6306, + "step": 27134 + }, + { + "epoch": 1.8829366456179306, + "grad_norm": 1.485434889793396, + "learning_rate": 0.00011428962355036464, + "loss": 0.4802, + "step": 27135 + }, + { + "epoch": 1.8830060370550274, + "grad_norm": 1.1625710725784302, + "learning_rate": 0.00011428406257239916, + "loss": 0.5121, + "step": 27136 + }, + { + "epoch": 1.883075428492124, + "grad_norm": 1.4307258129119873, + "learning_rate": 0.00011427850154934057, + "loss": 0.5702, + "step": 27137 + }, + { + "epoch": 1.8831448199292207, + "grad_norm": 1.1698353290557861, + "learning_rate": 0.00011427294048120637, + "loss": 0.4271, + "step": 27138 + }, + { + "epoch": 1.8832142113663175, + "grad_norm": 1.9036728143692017, + "learning_rate": 0.00011426737936801417, + "loss": 0.7647, + "step": 27139 + }, + { + "epoch": 1.8832836028034141, + "grad_norm": 1.7080345153808594, + "learning_rate": 0.00011426181820978153, + "loss": 0.5936, + "step": 27140 + }, + { + "epoch": 1.8833529942405107, + "grad_norm": 1.947922945022583, + "learning_rate": 0.00011425625700652598, + "loss": 0.6193, + "step": 27141 + }, + { + "epoch": 1.8834223856776073, + "grad_norm": 0.8736234903335571, + "learning_rate": 0.00011425069575826506, + "loss": 0.3795, + "step": 27142 + }, + { + "epoch": 1.883491777114704, + "grad_norm": 1.571768045425415, + "learning_rate": 0.00011424513446501637, + "loss": 0.7081, + "step": 27143 + }, + { + "epoch": 1.8835611685518008, + "grad_norm": 1.878232717514038, + "learning_rate": 0.00011423957312679743, + "loss": 0.6293, + "step": 27144 + }, + { + "epoch": 1.8836305599888974, + "grad_norm": 1.4409741163253784, + "learning_rate": 0.0001142340117436258, + "loss": 0.4892, + "step": 27145 + }, + { + "epoch": 1.883699951425994, + "grad_norm": 1.7483679056167603, + "learning_rate": 0.00011422845031551908, + "loss": 0.6669, + "step": 27146 + }, + { + "epoch": 1.8837693428630908, + "grad_norm": 2.4238128662109375, + "learning_rate": 0.00011422288884249477, + "loss": 0.7062, + "step": 27147 + }, + { + "epoch": 1.8838387343001872, + "grad_norm": 0.9855535626411438, + "learning_rate": 0.00011421732732457045, + "loss": 0.4575, + "step": 27148 + }, + { + "epoch": 1.883908125737284, + "grad_norm": 1.6472517251968384, + "learning_rate": 0.00011421176576176367, + "loss": 0.6314, + "step": 27149 + }, + { + "epoch": 1.8839775171743807, + "grad_norm": 1.5216515064239502, + "learning_rate": 0.00011420620415409203, + "loss": 0.4833, + "step": 27150 + }, + { + "epoch": 1.8840469086114773, + "grad_norm": 1.6917531490325928, + "learning_rate": 0.000114200642501573, + "loss": 0.7478, + "step": 27151 + }, + { + "epoch": 1.884116300048574, + "grad_norm": 2.1246771812438965, + "learning_rate": 0.0001141950808042242, + "loss": 0.7404, + "step": 27152 + }, + { + "epoch": 1.8841856914856707, + "grad_norm": 1.3801600933074951, + "learning_rate": 0.00011418951906206322, + "loss": 0.4611, + "step": 27153 + }, + { + "epoch": 1.8842550829227673, + "grad_norm": 1.17328941822052, + "learning_rate": 0.00011418395727510753, + "loss": 0.5185, + "step": 27154 + }, + { + "epoch": 1.884324474359864, + "grad_norm": 1.4018949270248413, + "learning_rate": 0.00011417839544337475, + "loss": 0.6787, + "step": 27155 + }, + { + "epoch": 1.8843938657969606, + "grad_norm": 1.5787444114685059, + "learning_rate": 0.00011417283356688239, + "loss": 0.5911, + "step": 27156 + }, + { + "epoch": 1.8844632572340574, + "grad_norm": 1.219927430152893, + "learning_rate": 0.00011416727164564811, + "loss": 0.5301, + "step": 27157 + }, + { + "epoch": 1.884532648671154, + "grad_norm": 1.8254609107971191, + "learning_rate": 0.00011416170967968933, + "loss": 0.5742, + "step": 27158 + }, + { + "epoch": 1.8846020401082506, + "grad_norm": 1.423095464706421, + "learning_rate": 0.00011415614766902369, + "loss": 0.5754, + "step": 27159 + }, + { + "epoch": 1.8846714315453474, + "grad_norm": 1.4961419105529785, + "learning_rate": 0.00011415058561366879, + "loss": 0.5663, + "step": 27160 + }, + { + "epoch": 1.8847408229824438, + "grad_norm": 2.0180935859680176, + "learning_rate": 0.00011414502351364204, + "loss": 0.8904, + "step": 27161 + }, + { + "epoch": 1.8848102144195407, + "grad_norm": 1.3240572214126587, + "learning_rate": 0.00011413946136896114, + "loss": 0.508, + "step": 27162 + }, + { + "epoch": 1.8848796058566373, + "grad_norm": 1.4648085832595825, + "learning_rate": 0.00011413389917964364, + "loss": 0.6448, + "step": 27163 + }, + { + "epoch": 1.8849489972937339, + "grad_norm": 2.084869384765625, + "learning_rate": 0.00011412833694570699, + "loss": 0.7848, + "step": 27164 + }, + { + "epoch": 1.8850183887308307, + "grad_norm": 1.3680858612060547, + "learning_rate": 0.00011412277466716888, + "loss": 0.742, + "step": 27165 + }, + { + "epoch": 1.8850877801679273, + "grad_norm": 2.3403329849243164, + "learning_rate": 0.00011411721234404677, + "loss": 0.5451, + "step": 27166 + }, + { + "epoch": 1.885157171605024, + "grad_norm": 1.4398112297058105, + "learning_rate": 0.00011411164997635828, + "loss": 0.6024, + "step": 27167 + }, + { + "epoch": 1.8852265630421206, + "grad_norm": 1.6419419050216675, + "learning_rate": 0.00011410608756412093, + "loss": 0.6013, + "step": 27168 + }, + { + "epoch": 1.8852959544792172, + "grad_norm": 1.7046787738800049, + "learning_rate": 0.00011410052510735229, + "loss": 0.5937, + "step": 27169 + }, + { + "epoch": 1.885365345916314, + "grad_norm": 1.6984962224960327, + "learning_rate": 0.00011409496260606994, + "loss": 0.3815, + "step": 27170 + }, + { + "epoch": 1.8854347373534106, + "grad_norm": 1.57987642288208, + "learning_rate": 0.00011408940006029145, + "loss": 0.5488, + "step": 27171 + }, + { + "epoch": 1.8855041287905072, + "grad_norm": 1.4157501459121704, + "learning_rate": 0.00011408383747003433, + "loss": 0.5924, + "step": 27172 + }, + { + "epoch": 1.885573520227604, + "grad_norm": 2.1665337085723877, + "learning_rate": 0.00011407827483531617, + "loss": 0.6922, + "step": 27173 + }, + { + "epoch": 1.8856429116647004, + "grad_norm": 1.2003793716430664, + "learning_rate": 0.00011407271215615454, + "loss": 0.3956, + "step": 27174 + }, + { + "epoch": 1.8857123031017973, + "grad_norm": 1.6310944557189941, + "learning_rate": 0.00011406714943256698, + "loss": 0.5411, + "step": 27175 + }, + { + "epoch": 1.8857816945388939, + "grad_norm": 2.3433029651641846, + "learning_rate": 0.00011406158666457108, + "loss": 0.8731, + "step": 27176 + }, + { + "epoch": 1.8858510859759905, + "grad_norm": 2.1094133853912354, + "learning_rate": 0.00011405602385218434, + "loss": 0.6428, + "step": 27177 + }, + { + "epoch": 1.8859204774130873, + "grad_norm": 1.058802843093872, + "learning_rate": 0.00011405046099542442, + "loss": 0.4107, + "step": 27178 + }, + { + "epoch": 1.885989868850184, + "grad_norm": 1.372524619102478, + "learning_rate": 0.00011404489809430878, + "loss": 0.5397, + "step": 27179 + }, + { + "epoch": 1.8860592602872805, + "grad_norm": 1.9798331260681152, + "learning_rate": 0.00011403933514885505, + "loss": 0.5383, + "step": 27180 + }, + { + "epoch": 1.8861286517243772, + "grad_norm": 1.8885024785995483, + "learning_rate": 0.00011403377215908077, + "loss": 0.8738, + "step": 27181 + }, + { + "epoch": 1.8861980431614738, + "grad_norm": 1.6491905450820923, + "learning_rate": 0.00011402820912500351, + "loss": 0.4913, + "step": 27182 + }, + { + "epoch": 1.8862674345985706, + "grad_norm": 1.581772804260254, + "learning_rate": 0.00011402264604664081, + "loss": 0.5063, + "step": 27183 + }, + { + "epoch": 1.8863368260356672, + "grad_norm": 2.103327751159668, + "learning_rate": 0.00011401708292401026, + "loss": 0.5852, + "step": 27184 + }, + { + "epoch": 1.8864062174727638, + "grad_norm": 1.8436241149902344, + "learning_rate": 0.00011401151975712936, + "loss": 0.492, + "step": 27185 + }, + { + "epoch": 1.8864756089098607, + "grad_norm": 1.4326844215393066, + "learning_rate": 0.00011400595654601574, + "loss": 0.4463, + "step": 27186 + }, + { + "epoch": 1.886545000346957, + "grad_norm": 2.5372962951660156, + "learning_rate": 0.00011400039329068693, + "loss": 0.4955, + "step": 27187 + }, + { + "epoch": 1.8866143917840539, + "grad_norm": 1.2747013568878174, + "learning_rate": 0.00011399482999116054, + "loss": 0.4531, + "step": 27188 + }, + { + "epoch": 1.8866837832211505, + "grad_norm": 1.2753148078918457, + "learning_rate": 0.00011398926664745406, + "loss": 0.633, + "step": 27189 + }, + { + "epoch": 1.886753174658247, + "grad_norm": 1.4980833530426025, + "learning_rate": 0.0001139837032595851, + "loss": 0.4497, + "step": 27190 + }, + { + "epoch": 1.886822566095344, + "grad_norm": 1.5646352767944336, + "learning_rate": 0.00011397813982757122, + "loss": 0.7717, + "step": 27191 + }, + { + "epoch": 1.8868919575324405, + "grad_norm": 1.3813585042953491, + "learning_rate": 0.00011397257635142995, + "loss": 0.3095, + "step": 27192 + }, + { + "epoch": 1.8869613489695372, + "grad_norm": 1.281683087348938, + "learning_rate": 0.00011396701283117887, + "loss": 0.4714, + "step": 27193 + }, + { + "epoch": 1.8870307404066338, + "grad_norm": 1.5952801704406738, + "learning_rate": 0.00011396144926683558, + "loss": 0.5358, + "step": 27194 + }, + { + "epoch": 1.8871001318437304, + "grad_norm": 1.123768925666809, + "learning_rate": 0.00011395588565841762, + "loss": 0.3028, + "step": 27195 + }, + { + "epoch": 1.8871695232808272, + "grad_norm": 1.843537449836731, + "learning_rate": 0.00011395032200594253, + "loss": 0.7033, + "step": 27196 + }, + { + "epoch": 1.8872389147179238, + "grad_norm": 2.2992401123046875, + "learning_rate": 0.00011394475830942788, + "loss": 0.614, + "step": 27197 + }, + { + "epoch": 1.8873083061550204, + "grad_norm": 2.0309457778930664, + "learning_rate": 0.00011393919456889128, + "loss": 0.8754, + "step": 27198 + }, + { + "epoch": 1.8873776975921173, + "grad_norm": 2.0093472003936768, + "learning_rate": 0.00011393363078435024, + "loss": 0.6912, + "step": 27199 + }, + { + "epoch": 1.8874470890292137, + "grad_norm": 3.1727700233459473, + "learning_rate": 0.00011392806695582237, + "loss": 0.5256, + "step": 27200 + }, + { + "epoch": 1.8875164804663105, + "grad_norm": 2.5907211303710938, + "learning_rate": 0.00011392250308332517, + "loss": 0.7941, + "step": 27201 + }, + { + "epoch": 1.887585871903407, + "grad_norm": 2.282250165939331, + "learning_rate": 0.00011391693916687627, + "loss": 0.7356, + "step": 27202 + }, + { + "epoch": 1.8876552633405037, + "grad_norm": 1.1885862350463867, + "learning_rate": 0.0001139113752064932, + "loss": 0.4631, + "step": 27203 + }, + { + "epoch": 1.8877246547776005, + "grad_norm": 1.9999711513519287, + "learning_rate": 0.00011390581120219353, + "loss": 0.4578, + "step": 27204 + }, + { + "epoch": 1.8877940462146972, + "grad_norm": 1.5815279483795166, + "learning_rate": 0.00011390024715399484, + "loss": 0.6237, + "step": 27205 + }, + { + "epoch": 1.8878634376517938, + "grad_norm": 1.517505168914795, + "learning_rate": 0.00011389468306191464, + "loss": 0.531, + "step": 27206 + }, + { + "epoch": 1.8879328290888906, + "grad_norm": 1.420846939086914, + "learning_rate": 0.00011388911892597056, + "loss": 0.4853, + "step": 27207 + }, + { + "epoch": 1.888002220525987, + "grad_norm": 1.457947015762329, + "learning_rate": 0.00011388355474618015, + "loss": 0.5146, + "step": 27208 + }, + { + "epoch": 1.8880716119630838, + "grad_norm": 1.1793475151062012, + "learning_rate": 0.00011387799052256098, + "loss": 0.4874, + "step": 27209 + }, + { + "epoch": 1.8881410034001804, + "grad_norm": 1.329214096069336, + "learning_rate": 0.0001138724262551306, + "loss": 0.4641, + "step": 27210 + }, + { + "epoch": 1.888210394837277, + "grad_norm": 1.4680891036987305, + "learning_rate": 0.00011386686194390655, + "loss": 0.5014, + "step": 27211 + }, + { + "epoch": 1.8882797862743739, + "grad_norm": 1.7122677564620972, + "learning_rate": 0.00011386129758890647, + "loss": 0.4648, + "step": 27212 + }, + { + "epoch": 1.8883491777114703, + "grad_norm": 1.994659185409546, + "learning_rate": 0.00011385573319014784, + "loss": 0.8323, + "step": 27213 + }, + { + "epoch": 1.888418569148567, + "grad_norm": 1.7495118379592896, + "learning_rate": 0.00011385016874764829, + "loss": 0.4967, + "step": 27214 + }, + { + "epoch": 1.8884879605856637, + "grad_norm": 1.6031185388565063, + "learning_rate": 0.00011384460426142536, + "loss": 0.4835, + "step": 27215 + }, + { + "epoch": 1.8885573520227603, + "grad_norm": 1.5205979347229004, + "learning_rate": 0.00011383903973149661, + "loss": 0.448, + "step": 27216 + }, + { + "epoch": 1.8886267434598571, + "grad_norm": 2.0381429195404053, + "learning_rate": 0.00011383347515787966, + "loss": 0.5884, + "step": 27217 + }, + { + "epoch": 1.8886961348969538, + "grad_norm": 1.2515209913253784, + "learning_rate": 0.00011382791054059199, + "loss": 0.5357, + "step": 27218 + }, + { + "epoch": 1.8887655263340504, + "grad_norm": 1.8526464700698853, + "learning_rate": 0.00011382234587965124, + "loss": 0.6289, + "step": 27219 + }, + { + "epoch": 1.8888349177711472, + "grad_norm": 1.7865055799484253, + "learning_rate": 0.00011381678117507493, + "loss": 0.5591, + "step": 27220 + }, + { + "epoch": 1.8889043092082436, + "grad_norm": 1.6747978925704956, + "learning_rate": 0.00011381121642688063, + "loss": 0.5837, + "step": 27221 + }, + { + "epoch": 1.8889737006453404, + "grad_norm": 1.9169373512268066, + "learning_rate": 0.00011380565163508596, + "loss": 0.9252, + "step": 27222 + }, + { + "epoch": 1.889043092082437, + "grad_norm": 0.9820898771286011, + "learning_rate": 0.00011380008679970841, + "loss": 0.2835, + "step": 27223 + }, + { + "epoch": 1.8891124835195336, + "grad_norm": 2.0352559089660645, + "learning_rate": 0.0001137945219207656, + "loss": 0.5726, + "step": 27224 + }, + { + "epoch": 1.8891818749566305, + "grad_norm": 1.2880159616470337, + "learning_rate": 0.00011378895699827508, + "loss": 0.5898, + "step": 27225 + }, + { + "epoch": 1.8892512663937269, + "grad_norm": 1.6024823188781738, + "learning_rate": 0.00011378339203225446, + "loss": 0.4925, + "step": 27226 + }, + { + "epoch": 1.8893206578308237, + "grad_norm": 1.4610599279403687, + "learning_rate": 0.00011377782702272121, + "loss": 0.5265, + "step": 27227 + }, + { + "epoch": 1.8893900492679203, + "grad_norm": 1.286354422569275, + "learning_rate": 0.00011377226196969299, + "loss": 0.5574, + "step": 27228 + }, + { + "epoch": 1.889459440705017, + "grad_norm": 1.5576852560043335, + "learning_rate": 0.00011376669687318738, + "loss": 0.4525, + "step": 27229 + }, + { + "epoch": 1.8895288321421138, + "grad_norm": 1.544062852859497, + "learning_rate": 0.00011376113173322185, + "loss": 0.4965, + "step": 27230 + }, + { + "epoch": 1.8895982235792104, + "grad_norm": 2.299812078475952, + "learning_rate": 0.00011375556654981403, + "loss": 0.489, + "step": 27231 + }, + { + "epoch": 1.889667615016307, + "grad_norm": 2.083591938018799, + "learning_rate": 0.00011375000132298148, + "loss": 0.6198, + "step": 27232 + }, + { + "epoch": 1.8897370064534038, + "grad_norm": 1.1586369276046753, + "learning_rate": 0.0001137444360527418, + "loss": 0.3466, + "step": 27233 + }, + { + "epoch": 1.8898063978905002, + "grad_norm": 1.561800479888916, + "learning_rate": 0.00011373887073911249, + "loss": 0.6021, + "step": 27234 + }, + { + "epoch": 1.889875789327597, + "grad_norm": 1.2527227401733398, + "learning_rate": 0.0001137333053821112, + "loss": 0.4305, + "step": 27235 + }, + { + "epoch": 1.8899451807646936, + "grad_norm": 1.6802947521209717, + "learning_rate": 0.00011372773998175545, + "loss": 0.5027, + "step": 27236 + }, + { + "epoch": 1.8900145722017903, + "grad_norm": 1.987906575202942, + "learning_rate": 0.0001137221745380628, + "loss": 0.5542, + "step": 27237 + }, + { + "epoch": 1.890083963638887, + "grad_norm": 1.5922200679779053, + "learning_rate": 0.00011371660905105084, + "loss": 0.5374, + "step": 27238 + }, + { + "epoch": 1.8901533550759835, + "grad_norm": 2.053352117538452, + "learning_rate": 0.00011371104352073714, + "loss": 0.4811, + "step": 27239 + }, + { + "epoch": 1.8902227465130803, + "grad_norm": 1.5809874534606934, + "learning_rate": 0.00011370547794713926, + "loss": 0.4658, + "step": 27240 + }, + { + "epoch": 1.890292137950177, + "grad_norm": 2.7028048038482666, + "learning_rate": 0.0001136999123302748, + "loss": 0.7486, + "step": 27241 + }, + { + "epoch": 1.8903615293872735, + "grad_norm": 1.4408528804779053, + "learning_rate": 0.00011369434667016126, + "loss": 0.4061, + "step": 27242 + }, + { + "epoch": 1.8904309208243704, + "grad_norm": 1.7247138023376465, + "learning_rate": 0.0001136887809668163, + "loss": 0.544, + "step": 27243 + }, + { + "epoch": 1.890500312261467, + "grad_norm": 1.2487813234329224, + "learning_rate": 0.00011368321522025742, + "loss": 0.5762, + "step": 27244 + }, + { + "epoch": 1.8905697036985636, + "grad_norm": 1.3633288145065308, + "learning_rate": 0.00011367764943050223, + "loss": 0.5209, + "step": 27245 + }, + { + "epoch": 1.8906390951356604, + "grad_norm": 1.372920274734497, + "learning_rate": 0.00011367208359756827, + "loss": 0.4262, + "step": 27246 + }, + { + "epoch": 1.8907084865727568, + "grad_norm": 1.3873692750930786, + "learning_rate": 0.00011366651772147316, + "loss": 0.5689, + "step": 27247 + }, + { + "epoch": 1.8907778780098536, + "grad_norm": 2.8363192081451416, + "learning_rate": 0.00011366095180223442, + "loss": 0.4954, + "step": 27248 + }, + { + "epoch": 1.8908472694469503, + "grad_norm": 1.9182167053222656, + "learning_rate": 0.00011365538583986964, + "loss": 0.7107, + "step": 27249 + }, + { + "epoch": 1.8909166608840469, + "grad_norm": 1.4932876825332642, + "learning_rate": 0.0001136498198343964, + "loss": 0.6024, + "step": 27250 + }, + { + "epoch": 1.8909860523211437, + "grad_norm": 1.9628568887710571, + "learning_rate": 0.00011364425378583226, + "loss": 0.5343, + "step": 27251 + }, + { + "epoch": 1.89105544375824, + "grad_norm": 1.7485277652740479, + "learning_rate": 0.00011363868769419481, + "loss": 0.5559, + "step": 27252 + }, + { + "epoch": 1.891124835195337, + "grad_norm": 1.4683072566986084, + "learning_rate": 0.00011363312155950156, + "loss": 0.6044, + "step": 27253 + }, + { + "epoch": 1.8911942266324335, + "grad_norm": 2.0618035793304443, + "learning_rate": 0.00011362755538177018, + "loss": 0.5588, + "step": 27254 + }, + { + "epoch": 1.8912636180695301, + "grad_norm": 1.2192039489746094, + "learning_rate": 0.00011362198916101817, + "loss": 0.4989, + "step": 27255 + }, + { + "epoch": 1.891333009506627, + "grad_norm": 1.4809589385986328, + "learning_rate": 0.00011361642289726308, + "loss": 0.5734, + "step": 27256 + }, + { + "epoch": 1.8914024009437236, + "grad_norm": 1.4664617776870728, + "learning_rate": 0.00011361085659052257, + "loss": 0.7109, + "step": 27257 + }, + { + "epoch": 1.8914717923808202, + "grad_norm": 1.4779833555221558, + "learning_rate": 0.00011360529024081414, + "loss": 0.7232, + "step": 27258 + }, + { + "epoch": 1.891541183817917, + "grad_norm": 1.558871865272522, + "learning_rate": 0.0001135997238481554, + "loss": 0.5277, + "step": 27259 + }, + { + "epoch": 1.8916105752550134, + "grad_norm": 1.4869095087051392, + "learning_rate": 0.00011359415741256394, + "loss": 0.513, + "step": 27260 + }, + { + "epoch": 1.8916799666921102, + "grad_norm": 1.7618683576583862, + "learning_rate": 0.00011358859093405725, + "loss": 0.5536, + "step": 27261 + }, + { + "epoch": 1.8917493581292069, + "grad_norm": 1.4091745615005493, + "learning_rate": 0.00011358302441265296, + "loss": 0.6812, + "step": 27262 + }, + { + "epoch": 1.8918187495663035, + "grad_norm": 1.238601803779602, + "learning_rate": 0.00011357745784836866, + "loss": 0.4052, + "step": 27263 + }, + { + "epoch": 1.8918881410034003, + "grad_norm": 1.5439242124557495, + "learning_rate": 0.0001135718912412219, + "loss": 0.5151, + "step": 27264 + }, + { + "epoch": 1.8919575324404967, + "grad_norm": 1.756496787071228, + "learning_rate": 0.00011356632459123024, + "loss": 0.5405, + "step": 27265 + }, + { + "epoch": 1.8920269238775935, + "grad_norm": 1.8351413011550903, + "learning_rate": 0.00011356075789841127, + "loss": 0.5752, + "step": 27266 + }, + { + "epoch": 1.8920963153146901, + "grad_norm": 1.34707772731781, + "learning_rate": 0.00011355519116278258, + "loss": 0.5795, + "step": 27267 + }, + { + "epoch": 1.8921657067517867, + "grad_norm": 1.8247805833816528, + "learning_rate": 0.0001135496243843617, + "loss": 0.7259, + "step": 27268 + }, + { + "epoch": 1.8922350981888836, + "grad_norm": 2.3874149322509766, + "learning_rate": 0.00011354405756316627, + "loss": 0.538, + "step": 27269 + }, + { + "epoch": 1.8923044896259802, + "grad_norm": 1.3327101469039917, + "learning_rate": 0.00011353849069921376, + "loss": 0.491, + "step": 27270 + }, + { + "epoch": 1.8923738810630768, + "grad_norm": 1.2660490274429321, + "learning_rate": 0.00011353292379252185, + "loss": 0.532, + "step": 27271 + }, + { + "epoch": 1.8924432725001736, + "grad_norm": 1.6732169389724731, + "learning_rate": 0.00011352735684310804, + "loss": 0.5819, + "step": 27272 + }, + { + "epoch": 1.89251266393727, + "grad_norm": 1.5753268003463745, + "learning_rate": 0.00011352178985098995, + "loss": 0.4992, + "step": 27273 + }, + { + "epoch": 1.8925820553743669, + "grad_norm": 1.0880192518234253, + "learning_rate": 0.00011351622281618514, + "loss": 0.3314, + "step": 27274 + }, + { + "epoch": 1.8926514468114635, + "grad_norm": 1.2662628889083862, + "learning_rate": 0.00011351065573871118, + "loss": 0.5553, + "step": 27275 + }, + { + "epoch": 1.89272083824856, + "grad_norm": 1.6745587587356567, + "learning_rate": 0.00011350508861858565, + "loss": 0.5631, + "step": 27276 + }, + { + "epoch": 1.892790229685657, + "grad_norm": 1.5659031867980957, + "learning_rate": 0.0001134995214558261, + "loss": 0.5592, + "step": 27277 + }, + { + "epoch": 1.8928596211227533, + "grad_norm": 1.4372638463974, + "learning_rate": 0.00011349395425045018, + "loss": 0.5267, + "step": 27278 + }, + { + "epoch": 1.8929290125598501, + "grad_norm": 0.9956067204475403, + "learning_rate": 0.00011348838700247535, + "loss": 0.4138, + "step": 27279 + }, + { + "epoch": 1.8929984039969467, + "grad_norm": 1.2586264610290527, + "learning_rate": 0.00011348281971191927, + "loss": 0.4563, + "step": 27280 + }, + { + "epoch": 1.8930677954340434, + "grad_norm": 1.5658527612686157, + "learning_rate": 0.00011347725237879951, + "loss": 0.5697, + "step": 27281 + }, + { + "epoch": 1.8931371868711402, + "grad_norm": 3.5307514667510986, + "learning_rate": 0.00011347168500313361, + "loss": 0.5238, + "step": 27282 + }, + { + "epoch": 1.8932065783082368, + "grad_norm": 1.2790583372116089, + "learning_rate": 0.00011346611758493917, + "loss": 0.454, + "step": 27283 + }, + { + "epoch": 1.8932759697453334, + "grad_norm": 1.303675651550293, + "learning_rate": 0.00011346055012423375, + "loss": 0.5704, + "step": 27284 + }, + { + "epoch": 1.8933453611824302, + "grad_norm": 1.516852855682373, + "learning_rate": 0.00011345498262103496, + "loss": 0.5155, + "step": 27285 + }, + { + "epoch": 1.8934147526195266, + "grad_norm": 1.7055768966674805, + "learning_rate": 0.0001134494150753603, + "loss": 0.626, + "step": 27286 + }, + { + "epoch": 1.8934841440566235, + "grad_norm": 1.0937144756317139, + "learning_rate": 0.00011344384748722747, + "loss": 0.6696, + "step": 27287 + }, + { + "epoch": 1.89355353549372, + "grad_norm": 1.3456709384918213, + "learning_rate": 0.00011343827985665392, + "loss": 0.4657, + "step": 27288 + }, + { + "epoch": 1.8936229269308167, + "grad_norm": 1.7992969751358032, + "learning_rate": 0.00011343271218365732, + "loss": 0.6857, + "step": 27289 + }, + { + "epoch": 1.8936923183679135, + "grad_norm": 1.6074230670928955, + "learning_rate": 0.00011342714446825516, + "loss": 0.4644, + "step": 27290 + }, + { + "epoch": 1.89376170980501, + "grad_norm": 1.3691984415054321, + "learning_rate": 0.00011342157671046507, + "loss": 0.7323, + "step": 27291 + }, + { + "epoch": 1.8938311012421067, + "grad_norm": 1.2590080499649048, + "learning_rate": 0.00011341600891030466, + "loss": 0.5521, + "step": 27292 + }, + { + "epoch": 1.8939004926792034, + "grad_norm": 1.4332417249679565, + "learning_rate": 0.00011341044106779142, + "loss": 0.4596, + "step": 27293 + }, + { + "epoch": 1.8939698841163, + "grad_norm": 1.2792463302612305, + "learning_rate": 0.000113404873182943, + "loss": 0.5566, + "step": 27294 + }, + { + "epoch": 1.8940392755533968, + "grad_norm": 1.4796801805496216, + "learning_rate": 0.00011339930525577696, + "loss": 0.7041, + "step": 27295 + }, + { + "epoch": 1.8941086669904934, + "grad_norm": 1.1861584186553955, + "learning_rate": 0.00011339373728631084, + "loss": 0.3779, + "step": 27296 + }, + { + "epoch": 1.89417805842759, + "grad_norm": 1.6610195636749268, + "learning_rate": 0.00011338816927456224, + "loss": 0.7749, + "step": 27297 + }, + { + "epoch": 1.8942474498646868, + "grad_norm": 1.8444485664367676, + "learning_rate": 0.00011338260122054878, + "loss": 0.6389, + "step": 27298 + }, + { + "epoch": 1.8943168413017832, + "grad_norm": 1.68149733543396, + "learning_rate": 0.00011337703312428796, + "loss": 0.4764, + "step": 27299 + }, + { + "epoch": 1.89438623273888, + "grad_norm": 1.7667996883392334, + "learning_rate": 0.00011337146498579741, + "loss": 0.7813, + "step": 27300 + }, + { + "epoch": 1.8944556241759767, + "grad_norm": 1.3482978343963623, + "learning_rate": 0.0001133658968050947, + "loss": 0.6515, + "step": 27301 + }, + { + "epoch": 1.8945250156130733, + "grad_norm": 2.1255335807800293, + "learning_rate": 0.00011336032858219743, + "loss": 0.6862, + "step": 27302 + }, + { + "epoch": 1.8945944070501701, + "grad_norm": 1.6603171825408936, + "learning_rate": 0.00011335476031712312, + "loss": 0.6473, + "step": 27303 + }, + { + "epoch": 1.8946637984872665, + "grad_norm": 1.318311095237732, + "learning_rate": 0.00011334919200988941, + "loss": 0.5324, + "step": 27304 + }, + { + "epoch": 1.8947331899243633, + "grad_norm": 2.2099246978759766, + "learning_rate": 0.00011334362366051383, + "loss": 0.6295, + "step": 27305 + }, + { + "epoch": 1.89480258136146, + "grad_norm": 1.6864361763000488, + "learning_rate": 0.00011333805526901398, + "loss": 0.5361, + "step": 27306 + }, + { + "epoch": 1.8948719727985566, + "grad_norm": 1.745215892791748, + "learning_rate": 0.00011333248683540744, + "loss": 0.5408, + "step": 27307 + }, + { + "epoch": 1.8949413642356534, + "grad_norm": 1.5227577686309814, + "learning_rate": 0.00011332691835971177, + "loss": 0.3848, + "step": 27308 + }, + { + "epoch": 1.89501075567275, + "grad_norm": 1.4013633728027344, + "learning_rate": 0.0001133213498419446, + "loss": 0.5029, + "step": 27309 + }, + { + "epoch": 1.8950801471098466, + "grad_norm": 1.7959065437316895, + "learning_rate": 0.00011331578128212344, + "loss": 0.5511, + "step": 27310 + }, + { + "epoch": 1.8951495385469435, + "grad_norm": 1.194070816040039, + "learning_rate": 0.00011331021268026592, + "loss": 0.504, + "step": 27311 + }, + { + "epoch": 1.8952189299840398, + "grad_norm": 1.4552003145217896, + "learning_rate": 0.0001133046440363896, + "loss": 0.4986, + "step": 27312 + }, + { + "epoch": 1.8952883214211367, + "grad_norm": 1.2947378158569336, + "learning_rate": 0.00011329907535051206, + "loss": 0.4714, + "step": 27313 + }, + { + "epoch": 1.8953577128582333, + "grad_norm": 1.9804673194885254, + "learning_rate": 0.00011329350662265088, + "loss": 0.6412, + "step": 27314 + }, + { + "epoch": 1.89542710429533, + "grad_norm": 1.2240517139434814, + "learning_rate": 0.00011328793785282363, + "loss": 0.3826, + "step": 27315 + }, + { + "epoch": 1.8954964957324267, + "grad_norm": 1.7581912279129028, + "learning_rate": 0.00011328236904104791, + "loss": 0.482, + "step": 27316 + }, + { + "epoch": 1.8955658871695231, + "grad_norm": 0.9309616684913635, + "learning_rate": 0.0001132768001873413, + "loss": 0.32, + "step": 27317 + }, + { + "epoch": 1.89563527860662, + "grad_norm": 1.8258212804794312, + "learning_rate": 0.00011327123129172136, + "loss": 0.5447, + "step": 27318 + }, + { + "epoch": 1.8957046700437166, + "grad_norm": 2.2009201049804688, + "learning_rate": 0.00011326566235420572, + "loss": 0.7195, + "step": 27319 + }, + { + "epoch": 1.8957740614808132, + "grad_norm": 1.7676899433135986, + "learning_rate": 0.0001132600933748119, + "loss": 0.4842, + "step": 27320 + }, + { + "epoch": 1.89584345291791, + "grad_norm": 1.0627835988998413, + "learning_rate": 0.00011325452435355748, + "loss": 0.5372, + "step": 27321 + }, + { + "epoch": 1.8959128443550066, + "grad_norm": 1.3386033773422241, + "learning_rate": 0.00011324895529046011, + "loss": 0.556, + "step": 27322 + }, + { + "epoch": 1.8959822357921032, + "grad_norm": 1.5700205564498901, + "learning_rate": 0.0001132433861855373, + "loss": 0.6672, + "step": 27323 + }, + { + "epoch": 1.8960516272292, + "grad_norm": 1.216740608215332, + "learning_rate": 0.00011323781703880665, + "loss": 0.5163, + "step": 27324 + }, + { + "epoch": 1.8961210186662965, + "grad_norm": 1.6491402387619019, + "learning_rate": 0.00011323224785028576, + "loss": 0.6213, + "step": 27325 + }, + { + "epoch": 1.8961904101033933, + "grad_norm": 1.4666004180908203, + "learning_rate": 0.00011322667861999221, + "loss": 0.4969, + "step": 27326 + }, + { + "epoch": 1.89625980154049, + "grad_norm": 1.5923254489898682, + "learning_rate": 0.00011322110934794355, + "loss": 0.5713, + "step": 27327 + }, + { + "epoch": 1.8963291929775865, + "grad_norm": 1.424355387687683, + "learning_rate": 0.00011321554003415737, + "loss": 0.4389, + "step": 27328 + }, + { + "epoch": 1.8963985844146833, + "grad_norm": 1.511742353439331, + "learning_rate": 0.00011320997067865126, + "loss": 0.5585, + "step": 27329 + }, + { + "epoch": 1.8964679758517797, + "grad_norm": 1.4596302509307861, + "learning_rate": 0.00011320440128144286, + "loss": 0.4333, + "step": 27330 + }, + { + "epoch": 1.8965373672888766, + "grad_norm": 1.836081862449646, + "learning_rate": 0.00011319883184254965, + "loss": 0.6781, + "step": 27331 + }, + { + "epoch": 1.8966067587259732, + "grad_norm": 1.4778480529785156, + "learning_rate": 0.00011319326236198926, + "loss": 0.474, + "step": 27332 + }, + { + "epoch": 1.8966761501630698, + "grad_norm": 1.7460061311721802, + "learning_rate": 0.00011318769283977928, + "loss": 0.6059, + "step": 27333 + }, + { + "epoch": 1.8967455416001666, + "grad_norm": 2.511951446533203, + "learning_rate": 0.00011318212327593728, + "loss": 0.7004, + "step": 27334 + }, + { + "epoch": 1.8968149330372632, + "grad_norm": 2.014509916305542, + "learning_rate": 0.00011317655367048086, + "loss": 0.599, + "step": 27335 + }, + { + "epoch": 1.8968843244743598, + "grad_norm": 2.2413370609283447, + "learning_rate": 0.00011317098402342759, + "loss": 0.6409, + "step": 27336 + }, + { + "epoch": 1.8969537159114567, + "grad_norm": 1.467834234237671, + "learning_rate": 0.00011316541433479502, + "loss": 0.466, + "step": 27337 + }, + { + "epoch": 1.897023107348553, + "grad_norm": 1.1673640012741089, + "learning_rate": 0.00011315984460460078, + "loss": 0.4853, + "step": 27338 + }, + { + "epoch": 1.89709249878565, + "grad_norm": 1.5176447629928589, + "learning_rate": 0.00011315427483286248, + "loss": 0.4747, + "step": 27339 + }, + { + "epoch": 1.8971618902227465, + "grad_norm": 2.405411720275879, + "learning_rate": 0.0001131487050195976, + "loss": 0.6249, + "step": 27340 + }, + { + "epoch": 1.8972312816598431, + "grad_norm": 1.24778413772583, + "learning_rate": 0.00011314313516482382, + "loss": 0.514, + "step": 27341 + }, + { + "epoch": 1.89730067309694, + "grad_norm": 1.6105222702026367, + "learning_rate": 0.00011313756526855867, + "loss": 0.6031, + "step": 27342 + }, + { + "epoch": 1.8973700645340363, + "grad_norm": 1.4935556650161743, + "learning_rate": 0.00011313199533081977, + "loss": 0.4839, + "step": 27343 + }, + { + "epoch": 1.8974394559711332, + "grad_norm": 1.4861687421798706, + "learning_rate": 0.00011312642535162465, + "loss": 0.5461, + "step": 27344 + }, + { + "epoch": 1.8975088474082298, + "grad_norm": 1.1994742155075073, + "learning_rate": 0.00011312085533099094, + "loss": 0.5347, + "step": 27345 + }, + { + "epoch": 1.8975782388453264, + "grad_norm": 1.4476211071014404, + "learning_rate": 0.00011311528526893622, + "loss": 0.6385, + "step": 27346 + }, + { + "epoch": 1.8976476302824232, + "grad_norm": 1.5504133701324463, + "learning_rate": 0.00011310971516547808, + "loss": 0.5377, + "step": 27347 + }, + { + "epoch": 1.8977170217195198, + "grad_norm": 1.8432979583740234, + "learning_rate": 0.00011310414502063407, + "loss": 0.5081, + "step": 27348 + }, + { + "epoch": 1.8977864131566164, + "grad_norm": 1.4438316822052002, + "learning_rate": 0.00011309857483442178, + "loss": 0.6326, + "step": 27349 + }, + { + "epoch": 1.8978558045937133, + "grad_norm": 1.7130515575408936, + "learning_rate": 0.00011309300460685885, + "loss": 0.584, + "step": 27350 + }, + { + "epoch": 1.8979251960308097, + "grad_norm": 1.2065625190734863, + "learning_rate": 0.00011308743433796281, + "loss": 0.4298, + "step": 27351 + }, + { + "epoch": 1.8979945874679065, + "grad_norm": 1.5484946966171265, + "learning_rate": 0.00011308186402775125, + "loss": 0.5866, + "step": 27352 + }, + { + "epoch": 1.8980639789050031, + "grad_norm": 1.8685873746871948, + "learning_rate": 0.00011307629367624174, + "loss": 0.5165, + "step": 27353 + }, + { + "epoch": 1.8981333703420997, + "grad_norm": 1.7437525987625122, + "learning_rate": 0.00011307072328345194, + "loss": 0.6231, + "step": 27354 + }, + { + "epoch": 1.8982027617791966, + "grad_norm": 1.7590230703353882, + "learning_rate": 0.00011306515284939933, + "loss": 0.5772, + "step": 27355 + }, + { + "epoch": 1.898272153216293, + "grad_norm": 1.4955555200576782, + "learning_rate": 0.0001130595823741016, + "loss": 0.4625, + "step": 27356 + }, + { + "epoch": 1.8983415446533898, + "grad_norm": 1.3471503257751465, + "learning_rate": 0.00011305401185757625, + "loss": 0.6046, + "step": 27357 + }, + { + "epoch": 1.8984109360904864, + "grad_norm": 2.4771568775177, + "learning_rate": 0.00011304844129984094, + "loss": 0.4588, + "step": 27358 + }, + { + "epoch": 1.898480327527583, + "grad_norm": 1.6547044515609741, + "learning_rate": 0.00011304287070091315, + "loss": 0.61, + "step": 27359 + }, + { + "epoch": 1.8985497189646798, + "grad_norm": 1.828500509262085, + "learning_rate": 0.00011303730006081057, + "loss": 0.6497, + "step": 27360 + }, + { + "epoch": 1.8986191104017764, + "grad_norm": 1.5453041791915894, + "learning_rate": 0.00011303172937955076, + "loss": 0.6014, + "step": 27361 + }, + { + "epoch": 1.898688501838873, + "grad_norm": 1.6712167263031006, + "learning_rate": 0.00011302615865715124, + "loss": 0.575, + "step": 27362 + }, + { + "epoch": 1.8987578932759699, + "grad_norm": 1.5200005769729614, + "learning_rate": 0.00011302058789362968, + "loss": 0.5671, + "step": 27363 + }, + { + "epoch": 1.8988272847130663, + "grad_norm": 1.4448249340057373, + "learning_rate": 0.00011301501708900363, + "loss": 0.3815, + "step": 27364 + }, + { + "epoch": 1.898896676150163, + "grad_norm": 1.4230408668518066, + "learning_rate": 0.00011300944624329068, + "loss": 0.5216, + "step": 27365 + }, + { + "epoch": 1.8989660675872597, + "grad_norm": 1.4544992446899414, + "learning_rate": 0.00011300387535650843, + "loss": 0.4399, + "step": 27366 + }, + { + "epoch": 1.8990354590243563, + "grad_norm": 1.7698158025741577, + "learning_rate": 0.00011299830442867443, + "loss": 0.4938, + "step": 27367 + }, + { + "epoch": 1.8991048504614532, + "grad_norm": 1.3891549110412598, + "learning_rate": 0.00011299273345980635, + "loss": 0.6447, + "step": 27368 + }, + { + "epoch": 1.8991742418985496, + "grad_norm": 1.9711326360702515, + "learning_rate": 0.00011298716244992165, + "loss": 0.4196, + "step": 27369 + }, + { + "epoch": 1.8992436333356464, + "grad_norm": 1.5080578327178955, + "learning_rate": 0.00011298159139903801, + "loss": 0.6132, + "step": 27370 + }, + { + "epoch": 1.899313024772743, + "grad_norm": 1.8720375299453735, + "learning_rate": 0.000112976020307173, + "loss": 0.5286, + "step": 27371 + }, + { + "epoch": 1.8993824162098396, + "grad_norm": 10.970367431640625, + "learning_rate": 0.00011297044917434419, + "loss": 0.51, + "step": 27372 + }, + { + "epoch": 1.8994518076469364, + "grad_norm": 1.300367832183838, + "learning_rate": 0.00011296487800056917, + "loss": 0.5765, + "step": 27373 + }, + { + "epoch": 1.899521199084033, + "grad_norm": 1.9726979732513428, + "learning_rate": 0.00011295930678586555, + "loss": 0.7173, + "step": 27374 + }, + { + "epoch": 1.8995905905211297, + "grad_norm": 0.9963580369949341, + "learning_rate": 0.00011295373553025088, + "loss": 0.4483, + "step": 27375 + }, + { + "epoch": 1.8996599819582265, + "grad_norm": 1.0208011865615845, + "learning_rate": 0.00011294816423374282, + "loss": 0.293, + "step": 27376 + }, + { + "epoch": 1.8997293733953229, + "grad_norm": 2.744009256362915, + "learning_rate": 0.00011294259289635884, + "loss": 0.7175, + "step": 27377 + }, + { + "epoch": 1.8997987648324197, + "grad_norm": 1.5440105199813843, + "learning_rate": 0.00011293702151811665, + "loss": 0.5157, + "step": 27378 + }, + { + "epoch": 1.8998681562695163, + "grad_norm": 1.5271265506744385, + "learning_rate": 0.00011293145009903376, + "loss": 0.4345, + "step": 27379 + }, + { + "epoch": 1.899937547706613, + "grad_norm": 1.5308068990707397, + "learning_rate": 0.00011292587863912778, + "loss": 0.6277, + "step": 27380 + }, + { + "epoch": 1.9000069391437098, + "grad_norm": 1.1199692487716675, + "learning_rate": 0.00011292030713841633, + "loss": 0.4749, + "step": 27381 + }, + { + "epoch": 1.9000763305808064, + "grad_norm": 1.4009758234024048, + "learning_rate": 0.00011291473559691693, + "loss": 0.5357, + "step": 27382 + }, + { + "epoch": 1.900145722017903, + "grad_norm": 1.313703179359436, + "learning_rate": 0.00011290916401464721, + "loss": 0.4575, + "step": 27383 + }, + { + "epoch": 1.9002151134549996, + "grad_norm": 1.5376496315002441, + "learning_rate": 0.00011290359239162475, + "loss": 0.4822, + "step": 27384 + }, + { + "epoch": 1.9002845048920962, + "grad_norm": 2.1487371921539307, + "learning_rate": 0.00011289802072786719, + "loss": 0.5635, + "step": 27385 + }, + { + "epoch": 1.900353896329193, + "grad_norm": 2.2088377475738525, + "learning_rate": 0.00011289244902339204, + "loss": 0.508, + "step": 27386 + }, + { + "epoch": 1.9004232877662897, + "grad_norm": 0.8680053949356079, + "learning_rate": 0.00011288687727821694, + "loss": 0.3719, + "step": 27387 + }, + { + "epoch": 1.9004926792033863, + "grad_norm": 1.6887189149856567, + "learning_rate": 0.00011288130549235947, + "loss": 0.5228, + "step": 27388 + }, + { + "epoch": 1.900562070640483, + "grad_norm": 1.4363662004470825, + "learning_rate": 0.0001128757336658372, + "loss": 0.5673, + "step": 27389 + }, + { + "epoch": 1.9006314620775795, + "grad_norm": 2.1822478771209717, + "learning_rate": 0.00011287016179866772, + "loss": 1.0675, + "step": 27390 + }, + { + "epoch": 1.9007008535146763, + "grad_norm": 2.0150153636932373, + "learning_rate": 0.00011286458989086868, + "loss": 0.6593, + "step": 27391 + }, + { + "epoch": 1.900770244951773, + "grad_norm": 1.9426480531692505, + "learning_rate": 0.00011285901794245756, + "loss": 0.7861, + "step": 27392 + }, + { + "epoch": 1.9008396363888695, + "grad_norm": 1.4404278993606567, + "learning_rate": 0.00011285344595345206, + "loss": 0.5286, + "step": 27393 + }, + { + "epoch": 1.9009090278259664, + "grad_norm": 2.2352240085601807, + "learning_rate": 0.00011284787392386971, + "loss": 0.7777, + "step": 27394 + }, + { + "epoch": 1.900978419263063, + "grad_norm": 1.3586091995239258, + "learning_rate": 0.0001128423018537281, + "loss": 0.6574, + "step": 27395 + }, + { + "epoch": 1.9010478107001596, + "grad_norm": 1.8862274885177612, + "learning_rate": 0.00011283672974304484, + "loss": 0.6004, + "step": 27396 + }, + { + "epoch": 1.9011172021372562, + "grad_norm": 3.159087657928467, + "learning_rate": 0.0001128311575918375, + "loss": 0.5654, + "step": 27397 + }, + { + "epoch": 1.9011865935743528, + "grad_norm": 1.7067394256591797, + "learning_rate": 0.00011282558540012369, + "loss": 0.6864, + "step": 27398 + }, + { + "epoch": 1.9012559850114497, + "grad_norm": 1.749598503112793, + "learning_rate": 0.00011282001316792104, + "loss": 0.5647, + "step": 27399 + }, + { + "epoch": 1.9013253764485463, + "grad_norm": 1.9636385440826416, + "learning_rate": 0.00011281444089524704, + "loss": 0.6009, + "step": 27400 + }, + { + "epoch": 1.9013947678856429, + "grad_norm": 2.610103130340576, + "learning_rate": 0.00011280886858211934, + "loss": 1.0337, + "step": 27401 + }, + { + "epoch": 1.9014641593227397, + "grad_norm": 1.5299872159957886, + "learning_rate": 0.00011280329622855557, + "loss": 0.6895, + "step": 27402 + }, + { + "epoch": 1.901533550759836, + "grad_norm": 1.3705699443817139, + "learning_rate": 0.00011279772383457324, + "loss": 0.4723, + "step": 27403 + }, + { + "epoch": 1.901602942196933, + "grad_norm": 1.5863597393035889, + "learning_rate": 0.00011279215140018999, + "loss": 0.5574, + "step": 27404 + }, + { + "epoch": 1.9016723336340295, + "grad_norm": 1.3593446016311646, + "learning_rate": 0.00011278657892542339, + "loss": 0.5671, + "step": 27405 + }, + { + "epoch": 1.9017417250711262, + "grad_norm": 1.458744764328003, + "learning_rate": 0.00011278100641029108, + "loss": 0.7371, + "step": 27406 + }, + { + "epoch": 1.901811116508223, + "grad_norm": 1.4269880056381226, + "learning_rate": 0.00011277543385481057, + "loss": 0.5398, + "step": 27407 + }, + { + "epoch": 1.9018805079453196, + "grad_norm": 1.6955876350402832, + "learning_rate": 0.00011276986125899953, + "loss": 0.5265, + "step": 27408 + }, + { + "epoch": 1.9019498993824162, + "grad_norm": 1.3568223714828491, + "learning_rate": 0.00011276428862287551, + "loss": 0.5259, + "step": 27409 + }, + { + "epoch": 1.9020192908195128, + "grad_norm": 1.4367177486419678, + "learning_rate": 0.00011275871594645613, + "loss": 0.4744, + "step": 27410 + }, + { + "epoch": 1.9020886822566094, + "grad_norm": 1.229997158050537, + "learning_rate": 0.00011275314322975897, + "loss": 0.364, + "step": 27411 + }, + { + "epoch": 1.9021580736937063, + "grad_norm": 1.417952299118042, + "learning_rate": 0.0001127475704728016, + "loss": 0.4225, + "step": 27412 + }, + { + "epoch": 1.9022274651308029, + "grad_norm": 2.2335281372070312, + "learning_rate": 0.00011274199767560161, + "loss": 0.4888, + "step": 27413 + }, + { + "epoch": 1.9022968565678995, + "grad_norm": 1.506917953491211, + "learning_rate": 0.00011273642483817661, + "loss": 0.5613, + "step": 27414 + }, + { + "epoch": 1.9023662480049963, + "grad_norm": 1.2448357343673706, + "learning_rate": 0.0001127308519605442, + "loss": 0.4368, + "step": 27415 + }, + { + "epoch": 1.9024356394420927, + "grad_norm": 1.2652339935302734, + "learning_rate": 0.00011272527904272198, + "loss": 0.4862, + "step": 27416 + }, + { + "epoch": 1.9025050308791895, + "grad_norm": 2.089580535888672, + "learning_rate": 0.00011271970608472753, + "loss": 0.8107, + "step": 27417 + }, + { + "epoch": 1.9025744223162862, + "grad_norm": 1.6791495084762573, + "learning_rate": 0.00011271413308657844, + "loss": 0.6113, + "step": 27418 + }, + { + "epoch": 1.9026438137533828, + "grad_norm": 1.7218992710113525, + "learning_rate": 0.00011270856004829231, + "loss": 0.6723, + "step": 27419 + }, + { + "epoch": 1.9027132051904796, + "grad_norm": 1.5303990840911865, + "learning_rate": 0.00011270298696988672, + "loss": 0.5436, + "step": 27420 + }, + { + "epoch": 1.9027825966275762, + "grad_norm": 5.144526958465576, + "learning_rate": 0.00011269741385137928, + "loss": 0.6683, + "step": 27421 + }, + { + "epoch": 1.9028519880646728, + "grad_norm": 1.594449520111084, + "learning_rate": 0.00011269184069278757, + "loss": 0.4947, + "step": 27422 + }, + { + "epoch": 1.9029213795017694, + "grad_norm": 1.4898128509521484, + "learning_rate": 0.00011268626749412922, + "loss": 0.4767, + "step": 27423 + }, + { + "epoch": 1.902990770938866, + "grad_norm": 2.056793451309204, + "learning_rate": 0.00011268069425542176, + "loss": 0.7188, + "step": 27424 + }, + { + "epoch": 1.9030601623759629, + "grad_norm": 1.2431635856628418, + "learning_rate": 0.00011267512097668284, + "loss": 0.4684, + "step": 27425 + }, + { + "epoch": 1.9031295538130595, + "grad_norm": 2.411616086959839, + "learning_rate": 0.00011266954765793005, + "loss": 0.6509, + "step": 27426 + }, + { + "epoch": 1.903198945250156, + "grad_norm": 2.272876739501953, + "learning_rate": 0.00011266397429918094, + "loss": 0.6119, + "step": 27427 + }, + { + "epoch": 1.903268336687253, + "grad_norm": 1.4080322980880737, + "learning_rate": 0.00011265840090045314, + "loss": 0.5271, + "step": 27428 + }, + { + "epoch": 1.9033377281243493, + "grad_norm": 1.4190093278884888, + "learning_rate": 0.00011265282746176425, + "loss": 0.446, + "step": 27429 + }, + { + "epoch": 1.9034071195614461, + "grad_norm": 1.5019861459732056, + "learning_rate": 0.00011264725398313186, + "loss": 0.5719, + "step": 27430 + }, + { + "epoch": 1.9034765109985428, + "grad_norm": 1.0679244995117188, + "learning_rate": 0.0001126416804645735, + "loss": 0.5564, + "step": 27431 + }, + { + "epoch": 1.9035459024356394, + "grad_norm": 1.5388351678848267, + "learning_rate": 0.00011263610690610686, + "loss": 0.5467, + "step": 27432 + }, + { + "epoch": 1.9036152938727362, + "grad_norm": 2.032895088195801, + "learning_rate": 0.00011263053330774951, + "loss": 0.5957, + "step": 27433 + }, + { + "epoch": 1.9036846853098328, + "grad_norm": 1.926756739616394, + "learning_rate": 0.00011262495966951901, + "loss": 0.8968, + "step": 27434 + }, + { + "epoch": 1.9037540767469294, + "grad_norm": 1.6050899028778076, + "learning_rate": 0.00011261938599143298, + "loss": 0.6554, + "step": 27435 + }, + { + "epoch": 1.9038234681840263, + "grad_norm": 1.7395082712173462, + "learning_rate": 0.00011261381227350901, + "loss": 0.7574, + "step": 27436 + }, + { + "epoch": 1.9038928596211226, + "grad_norm": 1.7387301921844482, + "learning_rate": 0.00011260823851576469, + "loss": 0.7293, + "step": 27437 + }, + { + "epoch": 1.9039622510582195, + "grad_norm": 1.4105054140090942, + "learning_rate": 0.00011260266471821763, + "loss": 0.4213, + "step": 27438 + }, + { + "epoch": 1.904031642495316, + "grad_norm": 1.6847702264785767, + "learning_rate": 0.00011259709088088544, + "loss": 0.5471, + "step": 27439 + }, + { + "epoch": 1.9041010339324127, + "grad_norm": 1.3606219291687012, + "learning_rate": 0.0001125915170037857, + "loss": 0.5493, + "step": 27440 + }, + { + "epoch": 1.9041704253695095, + "grad_norm": 1.3763847351074219, + "learning_rate": 0.00011258594308693596, + "loss": 0.5671, + "step": 27441 + }, + { + "epoch": 1.904239816806606, + "grad_norm": 1.5244911909103394, + "learning_rate": 0.00011258036913035388, + "loss": 0.9195, + "step": 27442 + }, + { + "epoch": 1.9043092082437028, + "grad_norm": 1.822812795639038, + "learning_rate": 0.00011257479513405702, + "loss": 0.7883, + "step": 27443 + }, + { + "epoch": 1.9043785996807994, + "grad_norm": 1.9152705669403076, + "learning_rate": 0.00011256922109806302, + "loss": 0.6763, + "step": 27444 + }, + { + "epoch": 1.904447991117896, + "grad_norm": 2.344338893890381, + "learning_rate": 0.00011256364702238947, + "loss": 0.7393, + "step": 27445 + }, + { + "epoch": 1.9045173825549928, + "grad_norm": 1.2028348445892334, + "learning_rate": 0.00011255807290705389, + "loss": 0.4336, + "step": 27446 + }, + { + "epoch": 1.9045867739920894, + "grad_norm": 1.6742452383041382, + "learning_rate": 0.00011255249875207396, + "loss": 0.4649, + "step": 27447 + }, + { + "epoch": 1.904656165429186, + "grad_norm": 1.3572338819503784, + "learning_rate": 0.00011254692455746722, + "loss": 0.6582, + "step": 27448 + }, + { + "epoch": 1.9047255568662829, + "grad_norm": 1.7824031114578247, + "learning_rate": 0.0001125413503232513, + "loss": 0.6059, + "step": 27449 + }, + { + "epoch": 1.9047949483033793, + "grad_norm": 1.1692379713058472, + "learning_rate": 0.00011253577604944383, + "loss": 0.5047, + "step": 27450 + }, + { + "epoch": 1.904864339740476, + "grad_norm": 1.5522090196609497, + "learning_rate": 0.00011253020173606233, + "loss": 0.5886, + "step": 27451 + }, + { + "epoch": 1.9049337311775727, + "grad_norm": 1.6223891973495483, + "learning_rate": 0.00011252462738312446, + "loss": 0.6639, + "step": 27452 + }, + { + "epoch": 1.9050031226146693, + "grad_norm": 1.4971849918365479, + "learning_rate": 0.00011251905299064776, + "loss": 0.4998, + "step": 27453 + }, + { + "epoch": 1.9050725140517661, + "grad_norm": 1.9416329860687256, + "learning_rate": 0.00011251347855864991, + "loss": 0.5283, + "step": 27454 + }, + { + "epoch": 1.9051419054888625, + "grad_norm": 1.6134830713272095, + "learning_rate": 0.00011250790408714842, + "loss": 0.5151, + "step": 27455 + }, + { + "epoch": 1.9052112969259594, + "grad_norm": 3.051847457885742, + "learning_rate": 0.00011250232957616093, + "loss": 0.574, + "step": 27456 + }, + { + "epoch": 1.905280688363056, + "grad_norm": 2.144840717315674, + "learning_rate": 0.00011249675502570507, + "loss": 0.5434, + "step": 27457 + }, + { + "epoch": 1.9053500798001526, + "grad_norm": 1.2780708074569702, + "learning_rate": 0.00011249118043579839, + "loss": 0.4925, + "step": 27458 + }, + { + "epoch": 1.9054194712372494, + "grad_norm": 1.242377519607544, + "learning_rate": 0.00011248560580645848, + "loss": 0.5623, + "step": 27459 + }, + { + "epoch": 1.905488862674346, + "grad_norm": 1.3274320363998413, + "learning_rate": 0.00011248003113770298, + "loss": 0.3889, + "step": 27460 + }, + { + "epoch": 1.9055582541114426, + "grad_norm": 1.498252272605896, + "learning_rate": 0.00011247445642954948, + "loss": 0.6054, + "step": 27461 + }, + { + "epoch": 1.9056276455485395, + "grad_norm": 1.4579309225082397, + "learning_rate": 0.00011246888168201554, + "loss": 0.3544, + "step": 27462 + }, + { + "epoch": 1.9056970369856359, + "grad_norm": 1.5395491123199463, + "learning_rate": 0.00011246330689511881, + "loss": 0.6007, + "step": 27463 + }, + { + "epoch": 1.9057664284227327, + "grad_norm": 1.1940118074417114, + "learning_rate": 0.00011245773206887687, + "loss": 0.5517, + "step": 27464 + }, + { + "epoch": 1.9058358198598293, + "grad_norm": 1.517006278038025, + "learning_rate": 0.00011245215720330731, + "loss": 0.591, + "step": 27465 + }, + { + "epoch": 1.905905211296926, + "grad_norm": 1.503396987915039, + "learning_rate": 0.00011244658229842772, + "loss": 0.5581, + "step": 27466 + }, + { + "epoch": 1.9059746027340227, + "grad_norm": 1.5817135572433472, + "learning_rate": 0.00011244100735425573, + "loss": 0.4315, + "step": 27467 + }, + { + "epoch": 1.9060439941711191, + "grad_norm": 1.2251218557357788, + "learning_rate": 0.00011243543237080893, + "loss": 0.4899, + "step": 27468 + }, + { + "epoch": 1.906113385608216, + "grad_norm": 1.5299896001815796, + "learning_rate": 0.0001124298573481049, + "loss": 0.6419, + "step": 27469 + }, + { + "epoch": 1.9061827770453126, + "grad_norm": 2.815300464630127, + "learning_rate": 0.00011242428228616124, + "loss": 0.8453, + "step": 27470 + }, + { + "epoch": 1.9062521684824092, + "grad_norm": 1.7504686117172241, + "learning_rate": 0.00011241870718499559, + "loss": 0.7072, + "step": 27471 + }, + { + "epoch": 1.906321559919506, + "grad_norm": 1.4609442949295044, + "learning_rate": 0.00011241313204462549, + "loss": 0.5086, + "step": 27472 + }, + { + "epoch": 1.9063909513566026, + "grad_norm": 2.1484975814819336, + "learning_rate": 0.00011240755686506858, + "loss": 0.7922, + "step": 27473 + }, + { + "epoch": 1.9064603427936992, + "grad_norm": 1.615202784538269, + "learning_rate": 0.00011240198164634246, + "loss": 0.6727, + "step": 27474 + }, + { + "epoch": 1.906529734230796, + "grad_norm": 1.6160590648651123, + "learning_rate": 0.00011239640638846474, + "loss": 0.4397, + "step": 27475 + }, + { + "epoch": 1.9065991256678925, + "grad_norm": 2.246054172515869, + "learning_rate": 0.00011239083109145297, + "loss": 0.5266, + "step": 27476 + }, + { + "epoch": 1.9066685171049893, + "grad_norm": 2.0659193992614746, + "learning_rate": 0.0001123852557553248, + "loss": 0.7152, + "step": 27477 + }, + { + "epoch": 1.906737908542086, + "grad_norm": 1.848048448562622, + "learning_rate": 0.00011237968038009783, + "loss": 0.5138, + "step": 27478 + }, + { + "epoch": 1.9068072999791825, + "grad_norm": 1.606529712677002, + "learning_rate": 0.00011237410496578962, + "loss": 0.5275, + "step": 27479 + }, + { + "epoch": 1.9068766914162794, + "grad_norm": 1.8266440629959106, + "learning_rate": 0.00011236852951241783, + "loss": 0.6454, + "step": 27480 + }, + { + "epoch": 1.9069460828533757, + "grad_norm": 1.4931117296218872, + "learning_rate": 0.00011236295402, + "loss": 0.4444, + "step": 27481 + }, + { + "epoch": 1.9070154742904726, + "grad_norm": 1.262686014175415, + "learning_rate": 0.00011235737848855377, + "loss": 0.4883, + "step": 27482 + }, + { + "epoch": 1.9070848657275692, + "grad_norm": 1.3606619834899902, + "learning_rate": 0.00011235180291809671, + "loss": 0.4586, + "step": 27483 + }, + { + "epoch": 1.9071542571646658, + "grad_norm": 1.5105952024459839, + "learning_rate": 0.00011234622730864646, + "loss": 0.6223, + "step": 27484 + }, + { + "epoch": 1.9072236486017626, + "grad_norm": 1.5895888805389404, + "learning_rate": 0.0001123406516602206, + "loss": 0.5999, + "step": 27485 + }, + { + "epoch": 1.9072930400388592, + "grad_norm": 1.343737006187439, + "learning_rate": 0.00011233507597283672, + "loss": 0.6703, + "step": 27486 + }, + { + "epoch": 1.9073624314759559, + "grad_norm": 2.433537721633911, + "learning_rate": 0.00011232950024651246, + "loss": 0.6216, + "step": 27487 + }, + { + "epoch": 1.9074318229130527, + "grad_norm": 1.337338924407959, + "learning_rate": 0.00011232392448126539, + "loss": 0.4777, + "step": 27488 + }, + { + "epoch": 1.907501214350149, + "grad_norm": 1.6726254224777222, + "learning_rate": 0.0001123183486771131, + "loss": 0.6503, + "step": 27489 + }, + { + "epoch": 1.907570605787246, + "grad_norm": 1.297639012336731, + "learning_rate": 0.00011231277283407322, + "loss": 0.5447, + "step": 27490 + }, + { + "epoch": 1.9076399972243425, + "grad_norm": 1.924308180809021, + "learning_rate": 0.00011230719695216334, + "loss": 0.7897, + "step": 27491 + }, + { + "epoch": 1.9077093886614391, + "grad_norm": 1.75028395652771, + "learning_rate": 0.00011230162103140109, + "loss": 0.5137, + "step": 27492 + }, + { + "epoch": 1.907778780098536, + "grad_norm": 1.3871746063232422, + "learning_rate": 0.00011229604507180405, + "loss": 0.6202, + "step": 27493 + }, + { + "epoch": 1.9078481715356324, + "grad_norm": 1.8768125772476196, + "learning_rate": 0.0001122904690733898, + "loss": 0.7364, + "step": 27494 + }, + { + "epoch": 1.9079175629727292, + "grad_norm": 1.6024913787841797, + "learning_rate": 0.000112284893036176, + "loss": 0.7927, + "step": 27495 + }, + { + "epoch": 1.9079869544098258, + "grad_norm": 1.3533927202224731, + "learning_rate": 0.00011227931696018019, + "loss": 0.6401, + "step": 27496 + }, + { + "epoch": 1.9080563458469224, + "grad_norm": 1.748398780822754, + "learning_rate": 0.00011227374084541999, + "loss": 0.5395, + "step": 27497 + }, + { + "epoch": 1.9081257372840192, + "grad_norm": 1.588916540145874, + "learning_rate": 0.00011226816469191306, + "loss": 0.6272, + "step": 27498 + }, + { + "epoch": 1.9081951287211159, + "grad_norm": 1.9388484954833984, + "learning_rate": 0.0001122625884996769, + "loss": 0.4731, + "step": 27499 + }, + { + "epoch": 1.9082645201582125, + "grad_norm": 1.4055163860321045, + "learning_rate": 0.00011225701226872922, + "loss": 0.5137, + "step": 27500 + }, + { + "epoch": 1.9083339115953093, + "grad_norm": 2.11731219291687, + "learning_rate": 0.00011225143599908753, + "loss": 0.4394, + "step": 27501 + }, + { + "epoch": 1.9084033030324057, + "grad_norm": 1.3371089696884155, + "learning_rate": 0.00011224585969076953, + "loss": 0.5179, + "step": 27502 + }, + { + "epoch": 1.9084726944695025, + "grad_norm": 1.4781523942947388, + "learning_rate": 0.00011224028334379273, + "loss": 0.4432, + "step": 27503 + }, + { + "epoch": 1.9085420859065991, + "grad_norm": 1.7886202335357666, + "learning_rate": 0.00011223470695817479, + "loss": 0.656, + "step": 27504 + }, + { + "epoch": 1.9086114773436957, + "grad_norm": 1.489011526107788, + "learning_rate": 0.00011222913053393328, + "loss": 0.5876, + "step": 27505 + }, + { + "epoch": 1.9086808687807926, + "grad_norm": 1.6266884803771973, + "learning_rate": 0.00011222355407108587, + "loss": 0.6015, + "step": 27506 + }, + { + "epoch": 1.908750260217889, + "grad_norm": 1.9681676626205444, + "learning_rate": 0.0001122179775696501, + "loss": 0.6395, + "step": 27507 + }, + { + "epoch": 1.9088196516549858, + "grad_norm": 1.1206613779067993, + "learning_rate": 0.00011221240102964356, + "loss": 0.4102, + "step": 27508 + }, + { + "epoch": 1.9088890430920824, + "grad_norm": 1.5987533330917358, + "learning_rate": 0.00011220682445108393, + "loss": 0.7026, + "step": 27509 + }, + { + "epoch": 1.908958434529179, + "grad_norm": 1.961652398109436, + "learning_rate": 0.00011220124783398873, + "loss": 0.7011, + "step": 27510 + }, + { + "epoch": 1.9090278259662758, + "grad_norm": 1.8428179025650024, + "learning_rate": 0.00011219567117837562, + "loss": 0.6013, + "step": 27511 + }, + { + "epoch": 1.9090972174033725, + "grad_norm": 1.4610052108764648, + "learning_rate": 0.00011219009448426219, + "loss": 0.6001, + "step": 27512 + }, + { + "epoch": 1.909166608840469, + "grad_norm": 1.0788400173187256, + "learning_rate": 0.00011218451775166608, + "loss": 0.51, + "step": 27513 + }, + { + "epoch": 1.909236000277566, + "grad_norm": 1.799314260482788, + "learning_rate": 0.00011217894098060482, + "loss": 0.6296, + "step": 27514 + }, + { + "epoch": 1.9093053917146623, + "grad_norm": 0.8382961750030518, + "learning_rate": 0.00011217336417109607, + "loss": 0.3651, + "step": 27515 + }, + { + "epoch": 1.9093747831517591, + "grad_norm": 2.4811573028564453, + "learning_rate": 0.00011216778732315742, + "loss": 0.6215, + "step": 27516 + }, + { + "epoch": 1.9094441745888557, + "grad_norm": 2.5737056732177734, + "learning_rate": 0.0001121622104368065, + "loss": 0.7519, + "step": 27517 + }, + { + "epoch": 1.9095135660259523, + "grad_norm": 1.1674715280532837, + "learning_rate": 0.00011215663351206087, + "loss": 0.4893, + "step": 27518 + }, + { + "epoch": 1.9095829574630492, + "grad_norm": 1.4812334775924683, + "learning_rate": 0.00011215105654893816, + "loss": 0.5903, + "step": 27519 + }, + { + "epoch": 1.9096523489001456, + "grad_norm": 1.3140356540679932, + "learning_rate": 0.00011214547954745598, + "loss": 0.6167, + "step": 27520 + }, + { + "epoch": 1.9097217403372424, + "grad_norm": 1.4810899496078491, + "learning_rate": 0.00011213990250763193, + "loss": 0.5446, + "step": 27521 + }, + { + "epoch": 1.909791131774339, + "grad_norm": 1.491164207458496, + "learning_rate": 0.0001121343254294836, + "loss": 0.5137, + "step": 27522 + }, + { + "epoch": 1.9098605232114356, + "grad_norm": 1.2365325689315796, + "learning_rate": 0.00011212874831302864, + "loss": 0.5773, + "step": 27523 + }, + { + "epoch": 1.9099299146485325, + "grad_norm": 1.5489836931228638, + "learning_rate": 0.00011212317115828462, + "loss": 0.6758, + "step": 27524 + }, + { + "epoch": 1.909999306085629, + "grad_norm": 1.8655967712402344, + "learning_rate": 0.00011211759396526912, + "loss": 0.4796, + "step": 27525 + }, + { + "epoch": 1.9100686975227257, + "grad_norm": 1.0785269737243652, + "learning_rate": 0.00011211201673399982, + "loss": 0.3654, + "step": 27526 + }, + { + "epoch": 1.9101380889598225, + "grad_norm": 1.8359819650650024, + "learning_rate": 0.00011210643946449428, + "loss": 0.5987, + "step": 27527 + }, + { + "epoch": 1.910207480396919, + "grad_norm": 1.680998682975769, + "learning_rate": 0.00011210086215677012, + "loss": 0.437, + "step": 27528 + }, + { + "epoch": 1.9102768718340157, + "grad_norm": 1.3006254434585571, + "learning_rate": 0.00011209528481084493, + "loss": 0.4513, + "step": 27529 + }, + { + "epoch": 1.9103462632711123, + "grad_norm": 1.793158769607544, + "learning_rate": 0.00011208970742673635, + "loss": 0.5066, + "step": 27530 + }, + { + "epoch": 1.910415654708209, + "grad_norm": 2.1324360370635986, + "learning_rate": 0.00011208413000446194, + "loss": 0.6335, + "step": 27531 + }, + { + "epoch": 1.9104850461453058, + "grad_norm": 1.554970383644104, + "learning_rate": 0.00011207855254403931, + "loss": 0.5269, + "step": 27532 + }, + { + "epoch": 1.9105544375824022, + "grad_norm": 1.9423162937164307, + "learning_rate": 0.00011207297504548614, + "loss": 0.5314, + "step": 27533 + }, + { + "epoch": 1.910623829019499, + "grad_norm": 5.043851852416992, + "learning_rate": 0.00011206739750882, + "loss": 0.6557, + "step": 27534 + }, + { + "epoch": 1.9106932204565956, + "grad_norm": 2.6735341548919678, + "learning_rate": 0.00011206181993405847, + "loss": 0.8717, + "step": 27535 + }, + { + "epoch": 1.9107626118936922, + "grad_norm": 1.506767749786377, + "learning_rate": 0.00011205624232121915, + "loss": 0.4226, + "step": 27536 + }, + { + "epoch": 1.910832003330789, + "grad_norm": 1.2559067010879517, + "learning_rate": 0.0001120506646703197, + "loss": 0.5427, + "step": 27537 + }, + { + "epoch": 1.9109013947678857, + "grad_norm": 1.8680529594421387, + "learning_rate": 0.00011204508698137768, + "loss": 0.4529, + "step": 27538 + }, + { + "epoch": 1.9109707862049823, + "grad_norm": 2.008513927459717, + "learning_rate": 0.00011203950925441071, + "loss": 0.6669, + "step": 27539 + }, + { + "epoch": 1.9110401776420791, + "grad_norm": 1.7137900590896606, + "learning_rate": 0.00011203393148943644, + "loss": 0.5481, + "step": 27540 + }, + { + "epoch": 1.9111095690791755, + "grad_norm": 1.7378395795822144, + "learning_rate": 0.00011202835368647242, + "loss": 0.5247, + "step": 27541 + }, + { + "epoch": 1.9111789605162723, + "grad_norm": 1.3228269815444946, + "learning_rate": 0.00011202277584553629, + "loss": 0.5114, + "step": 27542 + }, + { + "epoch": 1.911248351953369, + "grad_norm": 1.5183478593826294, + "learning_rate": 0.00011201719796664562, + "loss": 0.6245, + "step": 27543 + }, + { + "epoch": 1.9113177433904656, + "grad_norm": 1.9739850759506226, + "learning_rate": 0.00011201162004981811, + "loss": 0.8009, + "step": 27544 + }, + { + "epoch": 1.9113871348275624, + "grad_norm": 1.3134177923202515, + "learning_rate": 0.00011200604209507126, + "loss": 0.6871, + "step": 27545 + }, + { + "epoch": 1.9114565262646588, + "grad_norm": 1.7767224311828613, + "learning_rate": 0.00011200046410242275, + "loss": 0.8942, + "step": 27546 + }, + { + "epoch": 1.9115259177017556, + "grad_norm": 1.0852867364883423, + "learning_rate": 0.00011199488607189018, + "loss": 0.609, + "step": 27547 + }, + { + "epoch": 1.9115953091388522, + "grad_norm": 1.818468689918518, + "learning_rate": 0.00011198930800349112, + "loss": 0.581, + "step": 27548 + }, + { + "epoch": 1.9116647005759488, + "grad_norm": 3.4799914360046387, + "learning_rate": 0.00011198372989724319, + "loss": 0.7162, + "step": 27549 + }, + { + "epoch": 1.9117340920130457, + "grad_norm": 1.3032989501953125, + "learning_rate": 0.00011197815175316406, + "loss": 0.5184, + "step": 27550 + }, + { + "epoch": 1.9118034834501423, + "grad_norm": 1.288039207458496, + "learning_rate": 0.0001119725735712713, + "loss": 0.5717, + "step": 27551 + }, + { + "epoch": 1.911872874887239, + "grad_norm": 1.6122756004333496, + "learning_rate": 0.00011196699535158248, + "loss": 0.5717, + "step": 27552 + }, + { + "epoch": 1.9119422663243357, + "grad_norm": 1.8324909210205078, + "learning_rate": 0.00011196141709411525, + "loss": 0.6999, + "step": 27553 + }, + { + "epoch": 1.9120116577614321, + "grad_norm": 1.3078328371047974, + "learning_rate": 0.00011195583879888722, + "loss": 0.3624, + "step": 27554 + }, + { + "epoch": 1.912081049198529, + "grad_norm": 2.096938371658325, + "learning_rate": 0.00011195026046591599, + "loss": 0.7616, + "step": 27555 + }, + { + "epoch": 1.9121504406356256, + "grad_norm": 1.794953465461731, + "learning_rate": 0.00011194468209521917, + "loss": 0.6342, + "step": 27556 + }, + { + "epoch": 1.9122198320727222, + "grad_norm": 2.096534490585327, + "learning_rate": 0.00011193910368681437, + "loss": 0.6804, + "step": 27557 + }, + { + "epoch": 1.912289223509819, + "grad_norm": 1.3505641222000122, + "learning_rate": 0.00011193352524071923, + "loss": 0.5986, + "step": 27558 + }, + { + "epoch": 1.9123586149469154, + "grad_norm": 1.8440693616867065, + "learning_rate": 0.00011192794675695133, + "loss": 0.735, + "step": 27559 + }, + { + "epoch": 1.9124280063840122, + "grad_norm": 1.6812422275543213, + "learning_rate": 0.00011192236823552827, + "loss": 0.5649, + "step": 27560 + }, + { + "epoch": 1.9124973978211088, + "grad_norm": 1.4843881130218506, + "learning_rate": 0.00011191678967646768, + "loss": 0.5015, + "step": 27561 + }, + { + "epoch": 1.9125667892582054, + "grad_norm": 1.7028270959854126, + "learning_rate": 0.00011191121107978717, + "loss": 0.5052, + "step": 27562 + }, + { + "epoch": 1.9126361806953023, + "grad_norm": 0.7734923958778381, + "learning_rate": 0.00011190563244550434, + "loss": 0.2817, + "step": 27563 + }, + { + "epoch": 1.912705572132399, + "grad_norm": 2.010035753250122, + "learning_rate": 0.00011190005377363684, + "loss": 0.7726, + "step": 27564 + }, + { + "epoch": 1.9127749635694955, + "grad_norm": 1.8656501770019531, + "learning_rate": 0.00011189447506420223, + "loss": 0.6361, + "step": 27565 + }, + { + "epoch": 1.9128443550065923, + "grad_norm": 1.3988198041915894, + "learning_rate": 0.00011188889631721813, + "loss": 0.5439, + "step": 27566 + }, + { + "epoch": 1.9129137464436887, + "grad_norm": 1.2441353797912598, + "learning_rate": 0.00011188331753270218, + "loss": 0.4598, + "step": 27567 + }, + { + "epoch": 1.9129831378807856, + "grad_norm": 1.1194665431976318, + "learning_rate": 0.00011187773871067198, + "loss": 0.4943, + "step": 27568 + }, + { + "epoch": 1.9130525293178822, + "grad_norm": 1.9979526996612549, + "learning_rate": 0.00011187215985114514, + "loss": 0.383, + "step": 27569 + }, + { + "epoch": 1.9131219207549788, + "grad_norm": 1.0529385805130005, + "learning_rate": 0.00011186658095413925, + "loss": 0.4197, + "step": 27570 + }, + { + "epoch": 1.9131913121920756, + "grad_norm": 1.3671103715896606, + "learning_rate": 0.00011186100201967195, + "loss": 0.557, + "step": 27571 + }, + { + "epoch": 1.913260703629172, + "grad_norm": 2.1436758041381836, + "learning_rate": 0.00011185542304776085, + "loss": 0.5527, + "step": 27572 + }, + { + "epoch": 1.9133300950662688, + "grad_norm": 1.3170719146728516, + "learning_rate": 0.00011184984403842354, + "loss": 0.4464, + "step": 27573 + }, + { + "epoch": 1.9133994865033654, + "grad_norm": 1.1659369468688965, + "learning_rate": 0.00011184426499167764, + "loss": 0.6753, + "step": 27574 + }, + { + "epoch": 1.913468877940462, + "grad_norm": 1.1393969058990479, + "learning_rate": 0.00011183868590754079, + "loss": 0.4276, + "step": 27575 + }, + { + "epoch": 1.9135382693775589, + "grad_norm": 1.1465880870819092, + "learning_rate": 0.00011183310678603056, + "loss": 0.3824, + "step": 27576 + }, + { + "epoch": 1.9136076608146555, + "grad_norm": 1.7372195720672607, + "learning_rate": 0.0001118275276271646, + "loss": 0.4248, + "step": 27577 + }, + { + "epoch": 1.913677052251752, + "grad_norm": 1.6155041456222534, + "learning_rate": 0.00011182194843096052, + "loss": 0.5853, + "step": 27578 + }, + { + "epoch": 1.913746443688849, + "grad_norm": 1.2001488208770752, + "learning_rate": 0.0001118163691974359, + "loss": 0.5295, + "step": 27579 + }, + { + "epoch": 1.9138158351259453, + "grad_norm": 1.271536111831665, + "learning_rate": 0.00011181078992660838, + "loss": 0.5662, + "step": 27580 + }, + { + "epoch": 1.9138852265630422, + "grad_norm": 2.3923866748809814, + "learning_rate": 0.00011180521061849556, + "loss": 0.6603, + "step": 27581 + }, + { + "epoch": 1.9139546180001388, + "grad_norm": 1.3491203784942627, + "learning_rate": 0.00011179963127311509, + "loss": 0.503, + "step": 27582 + }, + { + "epoch": 1.9140240094372354, + "grad_norm": 1.8824843168258667, + "learning_rate": 0.00011179405189048451, + "loss": 0.6963, + "step": 27583 + }, + { + "epoch": 1.9140934008743322, + "grad_norm": 1.3945282697677612, + "learning_rate": 0.0001117884724706215, + "loss": 0.4684, + "step": 27584 + }, + { + "epoch": 1.9141627923114286, + "grad_norm": 1.2409826517105103, + "learning_rate": 0.00011178289301354365, + "loss": 0.6197, + "step": 27585 + }, + { + "epoch": 1.9142321837485254, + "grad_norm": 1.317510962486267, + "learning_rate": 0.00011177731351926857, + "loss": 0.5406, + "step": 27586 + }, + { + "epoch": 1.914301575185622, + "grad_norm": 1.5986348390579224, + "learning_rate": 0.00011177173398781388, + "loss": 0.5418, + "step": 27587 + }, + { + "epoch": 1.9143709666227187, + "grad_norm": 1.5554436445236206, + "learning_rate": 0.0001117661544191972, + "loss": 0.5465, + "step": 27588 + }, + { + "epoch": 1.9144403580598155, + "grad_norm": 1.6990001201629639, + "learning_rate": 0.00011176057481343613, + "loss": 0.6296, + "step": 27589 + }, + { + "epoch": 1.914509749496912, + "grad_norm": 1.5203522443771362, + "learning_rate": 0.00011175499517054828, + "loss": 0.5726, + "step": 27590 + }, + { + "epoch": 1.9145791409340087, + "grad_norm": 1.6352263689041138, + "learning_rate": 0.00011174941549055127, + "loss": 0.7294, + "step": 27591 + }, + { + "epoch": 1.9146485323711055, + "grad_norm": 1.2988394498825073, + "learning_rate": 0.00011174383577346274, + "loss": 0.5025, + "step": 27592 + }, + { + "epoch": 1.914717923808202, + "grad_norm": 1.6619197130203247, + "learning_rate": 0.00011173825601930026, + "loss": 0.6006, + "step": 27593 + }, + { + "epoch": 1.9147873152452988, + "grad_norm": 1.4788157939910889, + "learning_rate": 0.00011173267622808147, + "loss": 0.4545, + "step": 27594 + }, + { + "epoch": 1.9148567066823954, + "grad_norm": 1.3436591625213623, + "learning_rate": 0.00011172709639982399, + "loss": 0.4243, + "step": 27595 + }, + { + "epoch": 1.914926098119492, + "grad_norm": 2.357980489730835, + "learning_rate": 0.00011172151653454544, + "loss": 0.6723, + "step": 27596 + }, + { + "epoch": 1.9149954895565888, + "grad_norm": 1.2994630336761475, + "learning_rate": 0.0001117159366322634, + "loss": 0.542, + "step": 27597 + }, + { + "epoch": 1.9150648809936852, + "grad_norm": 1.3967385292053223, + "learning_rate": 0.00011171035669299552, + "loss": 0.5399, + "step": 27598 + }, + { + "epoch": 1.915134272430782, + "grad_norm": 1.6018126010894775, + "learning_rate": 0.0001117047767167594, + "loss": 0.5787, + "step": 27599 + }, + { + "epoch": 1.9152036638678787, + "grad_norm": 1.119402527809143, + "learning_rate": 0.00011169919670357268, + "loss": 0.5264, + "step": 27600 + }, + { + "epoch": 1.9152730553049753, + "grad_norm": 1.1633694171905518, + "learning_rate": 0.0001116936166534529, + "loss": 0.5977, + "step": 27601 + }, + { + "epoch": 1.915342446742072, + "grad_norm": 1.6421440839767456, + "learning_rate": 0.00011168803656641777, + "loss": 0.7462, + "step": 27602 + }, + { + "epoch": 1.9154118381791687, + "grad_norm": 1.4386675357818604, + "learning_rate": 0.00011168245644248485, + "loss": 0.7466, + "step": 27603 + }, + { + "epoch": 1.9154812296162653, + "grad_norm": 1.1715353727340698, + "learning_rate": 0.00011167687628167181, + "loss": 0.5961, + "step": 27604 + }, + { + "epoch": 1.9155506210533622, + "grad_norm": 2.007932424545288, + "learning_rate": 0.00011167129608399618, + "loss": 0.6216, + "step": 27605 + }, + { + "epoch": 1.9156200124904585, + "grad_norm": 1.5390429496765137, + "learning_rate": 0.00011166571584947565, + "loss": 0.6586, + "step": 27606 + }, + { + "epoch": 1.9156894039275554, + "grad_norm": 1.9342107772827148, + "learning_rate": 0.00011166013557812778, + "loss": 0.7262, + "step": 27607 + }, + { + "epoch": 1.915758795364652, + "grad_norm": 1.4018141031265259, + "learning_rate": 0.00011165455526997022, + "loss": 0.3867, + "step": 27608 + }, + { + "epoch": 1.9158281868017486, + "grad_norm": 2.15202260017395, + "learning_rate": 0.0001116489749250206, + "loss": 0.508, + "step": 27609 + }, + { + "epoch": 1.9158975782388454, + "grad_norm": 1.5436619520187378, + "learning_rate": 0.00011164339454329651, + "loss": 0.4925, + "step": 27610 + }, + { + "epoch": 1.9159669696759418, + "grad_norm": 1.491524338722229, + "learning_rate": 0.00011163781412481555, + "loss": 0.6095, + "step": 27611 + }, + { + "epoch": 1.9160363611130387, + "grad_norm": 1.6086887121200562, + "learning_rate": 0.00011163223366959538, + "loss": 0.5283, + "step": 27612 + }, + { + "epoch": 1.9161057525501353, + "grad_norm": 1.6773210763931274, + "learning_rate": 0.00011162665317765361, + "loss": 0.5998, + "step": 27613 + }, + { + "epoch": 1.9161751439872319, + "grad_norm": 1.505770206451416, + "learning_rate": 0.00011162107264900783, + "loss": 0.4117, + "step": 27614 + }, + { + "epoch": 1.9162445354243287, + "grad_norm": 1.682250738143921, + "learning_rate": 0.00011161549208367568, + "loss": 0.6514, + "step": 27615 + }, + { + "epoch": 1.9163139268614253, + "grad_norm": 1.1271648406982422, + "learning_rate": 0.00011160991148167477, + "loss": 0.4016, + "step": 27616 + }, + { + "epoch": 1.916383318298522, + "grad_norm": 1.3072786331176758, + "learning_rate": 0.00011160433084302271, + "loss": 0.443, + "step": 27617 + }, + { + "epoch": 1.9164527097356188, + "grad_norm": 1.572873830795288, + "learning_rate": 0.0001115987501677371, + "loss": 0.6318, + "step": 27618 + }, + { + "epoch": 1.9165221011727152, + "grad_norm": 1.2046477794647217, + "learning_rate": 0.0001115931694558356, + "loss": 0.4955, + "step": 27619 + }, + { + "epoch": 1.916591492609812, + "grad_norm": 1.3136694431304932, + "learning_rate": 0.00011158758870733584, + "loss": 0.564, + "step": 27620 + }, + { + "epoch": 1.9166608840469086, + "grad_norm": 1.1481168270111084, + "learning_rate": 0.0001115820079222554, + "loss": 0.3687, + "step": 27621 + }, + { + "epoch": 1.9167302754840052, + "grad_norm": 1.399690866470337, + "learning_rate": 0.00011157642710061189, + "loss": 0.4748, + "step": 27622 + }, + { + "epoch": 1.916799666921102, + "grad_norm": 1.397171139717102, + "learning_rate": 0.00011157084624242296, + "loss": 0.7443, + "step": 27623 + }, + { + "epoch": 1.9168690583581987, + "grad_norm": 1.5236213207244873, + "learning_rate": 0.00011156526534770618, + "loss": 0.5158, + "step": 27624 + }, + { + "epoch": 1.9169384497952953, + "grad_norm": 1.473326563835144, + "learning_rate": 0.0001115596844164792, + "loss": 0.6167, + "step": 27625 + }, + { + "epoch": 1.9170078412323919, + "grad_norm": 1.8731383085250854, + "learning_rate": 0.00011155410344875963, + "loss": 0.6904, + "step": 27626 + }, + { + "epoch": 1.9170772326694885, + "grad_norm": 2.131422519683838, + "learning_rate": 0.00011154852244456513, + "loss": 0.4342, + "step": 27627 + }, + { + "epoch": 1.9171466241065853, + "grad_norm": 1.1028990745544434, + "learning_rate": 0.00011154294140391326, + "loss": 0.467, + "step": 27628 + }, + { + "epoch": 1.917216015543682, + "grad_norm": 1.7401477098464966, + "learning_rate": 0.00011153736032682167, + "loss": 0.7089, + "step": 27629 + }, + { + "epoch": 1.9172854069807785, + "grad_norm": 1.704095721244812, + "learning_rate": 0.000111531779213308, + "loss": 0.5389, + "step": 27630 + }, + { + "epoch": 1.9173547984178754, + "grad_norm": 1.3852180242538452, + "learning_rate": 0.0001115261980633898, + "loss": 0.5701, + "step": 27631 + }, + { + "epoch": 1.9174241898549718, + "grad_norm": 1.4013679027557373, + "learning_rate": 0.00011152061687708472, + "loss": 0.5217, + "step": 27632 + }, + { + "epoch": 1.9174935812920686, + "grad_norm": 1.5047743320465088, + "learning_rate": 0.00011151503565441042, + "loss": 0.5022, + "step": 27633 + }, + { + "epoch": 1.9175629727291652, + "grad_norm": 1.470076560974121, + "learning_rate": 0.0001115094543953845, + "loss": 0.5721, + "step": 27634 + }, + { + "epoch": 1.9176323641662618, + "grad_norm": 1.391966700553894, + "learning_rate": 0.00011150387310002455, + "loss": 0.4239, + "step": 27635 + }, + { + "epoch": 1.9177017556033586, + "grad_norm": 1.4578157663345337, + "learning_rate": 0.00011149829176834818, + "loss": 0.5455, + "step": 27636 + }, + { + "epoch": 1.9177711470404553, + "grad_norm": 1.2320476770401, + "learning_rate": 0.00011149271040037309, + "loss": 0.6195, + "step": 27637 + }, + { + "epoch": 1.9178405384775519, + "grad_norm": 1.8338335752487183, + "learning_rate": 0.00011148712899611682, + "loss": 0.7874, + "step": 27638 + }, + { + "epoch": 1.9179099299146485, + "grad_norm": 1.3948490619659424, + "learning_rate": 0.00011148154755559703, + "loss": 0.5258, + "step": 27639 + }, + { + "epoch": 1.917979321351745, + "grad_norm": 1.8076434135437012, + "learning_rate": 0.00011147596607883131, + "loss": 0.5852, + "step": 27640 + }, + { + "epoch": 1.918048712788842, + "grad_norm": 1.1112803220748901, + "learning_rate": 0.00011147038456583732, + "loss": 0.2712, + "step": 27641 + }, + { + "epoch": 1.9181181042259385, + "grad_norm": 3.106785535812378, + "learning_rate": 0.00011146480301663263, + "loss": 0.6539, + "step": 27642 + }, + { + "epoch": 1.9181874956630351, + "grad_norm": 1.4996269941329956, + "learning_rate": 0.00011145922143123488, + "loss": 0.7101, + "step": 27643 + }, + { + "epoch": 1.918256887100132, + "grad_norm": 1.6567163467407227, + "learning_rate": 0.00011145363980966173, + "loss": 0.7457, + "step": 27644 + }, + { + "epoch": 1.9183262785372284, + "grad_norm": 1.321925401687622, + "learning_rate": 0.00011144805815193073, + "loss": 0.4286, + "step": 27645 + }, + { + "epoch": 1.9183956699743252, + "grad_norm": 1.450534701347351, + "learning_rate": 0.00011144247645805956, + "loss": 0.5049, + "step": 27646 + }, + { + "epoch": 1.9184650614114218, + "grad_norm": 1.7217644453048706, + "learning_rate": 0.00011143689472806582, + "loss": 0.691, + "step": 27647 + }, + { + "epoch": 1.9185344528485184, + "grad_norm": 1.9489399194717407, + "learning_rate": 0.00011143131296196714, + "loss": 0.526, + "step": 27648 + }, + { + "epoch": 1.9186038442856153, + "grad_norm": 1.42020583152771, + "learning_rate": 0.0001114257311597811, + "loss": 0.4687, + "step": 27649 + }, + { + "epoch": 1.9186732357227119, + "grad_norm": 1.9390956163406372, + "learning_rate": 0.00011142014932152536, + "loss": 0.5699, + "step": 27650 + }, + { + "epoch": 1.9187426271598085, + "grad_norm": 1.5232970714569092, + "learning_rate": 0.00011141456744721757, + "loss": 0.4968, + "step": 27651 + }, + { + "epoch": 1.918812018596905, + "grad_norm": 1.2723711729049683, + "learning_rate": 0.00011140898553687528, + "loss": 0.4444, + "step": 27652 + }, + { + "epoch": 1.9188814100340017, + "grad_norm": 2.3800578117370605, + "learning_rate": 0.00011140340359051615, + "loss": 0.5709, + "step": 27653 + }, + { + "epoch": 1.9189508014710985, + "grad_norm": 1.563329815864563, + "learning_rate": 0.00011139782160815781, + "loss": 0.6448, + "step": 27654 + }, + { + "epoch": 1.9190201929081951, + "grad_norm": 2.518559455871582, + "learning_rate": 0.00011139223958981785, + "loss": 0.6846, + "step": 27655 + }, + { + "epoch": 1.9190895843452918, + "grad_norm": 2.236929178237915, + "learning_rate": 0.00011138665753551396, + "loss": 0.6048, + "step": 27656 + }, + { + "epoch": 1.9191589757823886, + "grad_norm": 1.5211296081542969, + "learning_rate": 0.00011138107544526367, + "loss": 0.5161, + "step": 27657 + }, + { + "epoch": 1.919228367219485, + "grad_norm": 2.2189929485321045, + "learning_rate": 0.00011137549331908466, + "loss": 0.6947, + "step": 27658 + }, + { + "epoch": 1.9192977586565818, + "grad_norm": 1.6575251817703247, + "learning_rate": 0.00011136991115699454, + "loss": 0.5586, + "step": 27659 + }, + { + "epoch": 1.9193671500936784, + "grad_norm": 1.3175101280212402, + "learning_rate": 0.0001113643289590109, + "loss": 0.3574, + "step": 27660 + }, + { + "epoch": 1.919436541530775, + "grad_norm": 3.1072986125946045, + "learning_rate": 0.00011135874672515143, + "loss": 0.8815, + "step": 27661 + }, + { + "epoch": 1.9195059329678719, + "grad_norm": 1.8367273807525635, + "learning_rate": 0.00011135316445543369, + "loss": 0.7649, + "step": 27662 + }, + { + "epoch": 1.9195753244049685, + "grad_norm": 1.5635572671890259, + "learning_rate": 0.00011134758214987533, + "loss": 0.7116, + "step": 27663 + }, + { + "epoch": 1.919644715842065, + "grad_norm": 1.7439146041870117, + "learning_rate": 0.00011134199980849396, + "loss": 0.5072, + "step": 27664 + }, + { + "epoch": 1.9197141072791617, + "grad_norm": 1.8820470571517944, + "learning_rate": 0.00011133641743130725, + "loss": 0.7495, + "step": 27665 + }, + { + "epoch": 1.9197834987162583, + "grad_norm": 1.3980265855789185, + "learning_rate": 0.00011133083501833275, + "loss": 0.4535, + "step": 27666 + }, + { + "epoch": 1.9198528901533551, + "grad_norm": 1.5518009662628174, + "learning_rate": 0.00011132525256958812, + "loss": 0.6116, + "step": 27667 + }, + { + "epoch": 1.9199222815904518, + "grad_norm": 2.3280246257781982, + "learning_rate": 0.00011131967008509101, + "loss": 0.8525, + "step": 27668 + }, + { + "epoch": 1.9199916730275484, + "grad_norm": 1.9179213047027588, + "learning_rate": 0.000111314087564859, + "loss": 0.6943, + "step": 27669 + }, + { + "epoch": 1.9200610644646452, + "grad_norm": 1.896269679069519, + "learning_rate": 0.0001113085050089097, + "loss": 0.7412, + "step": 27670 + }, + { + "epoch": 1.9201304559017416, + "grad_norm": 1.1623294353485107, + "learning_rate": 0.00011130292241726079, + "loss": 0.4441, + "step": 27671 + }, + { + "epoch": 1.9201998473388384, + "grad_norm": 1.1080759763717651, + "learning_rate": 0.00011129733978992986, + "loss": 0.4794, + "step": 27672 + }, + { + "epoch": 1.920269238775935, + "grad_norm": 1.5730186700820923, + "learning_rate": 0.00011129175712693455, + "loss": 0.5067, + "step": 27673 + }, + { + "epoch": 1.9203386302130316, + "grad_norm": 1.7290380001068115, + "learning_rate": 0.00011128617442829248, + "loss": 0.4177, + "step": 27674 + }, + { + "epoch": 1.9204080216501285, + "grad_norm": 1.9451736211776733, + "learning_rate": 0.00011128059169402126, + "loss": 0.8174, + "step": 27675 + }, + { + "epoch": 1.920477413087225, + "grad_norm": 1.1100205183029175, + "learning_rate": 0.00011127500892413851, + "loss": 0.5451, + "step": 27676 + }, + { + "epoch": 1.9205468045243217, + "grad_norm": 1.5668015480041504, + "learning_rate": 0.00011126942611866186, + "loss": 0.7336, + "step": 27677 + }, + { + "epoch": 1.9206161959614185, + "grad_norm": 1.6943737268447876, + "learning_rate": 0.00011126384327760896, + "loss": 0.5867, + "step": 27678 + }, + { + "epoch": 1.920685587398515, + "grad_norm": 0.9996331334114075, + "learning_rate": 0.00011125826040099738, + "loss": 0.5117, + "step": 27679 + }, + { + "epoch": 1.9207549788356117, + "grad_norm": 1.320014238357544, + "learning_rate": 0.00011125267748884481, + "loss": 0.4055, + "step": 27680 + }, + { + "epoch": 1.9208243702727084, + "grad_norm": 1.8048189878463745, + "learning_rate": 0.00011124709454116881, + "loss": 0.6804, + "step": 27681 + }, + { + "epoch": 1.920893761709805, + "grad_norm": 1.302391767501831, + "learning_rate": 0.00011124151155798708, + "loss": 0.4213, + "step": 27682 + }, + { + "epoch": 1.9209631531469018, + "grad_norm": 1.6113570928573608, + "learning_rate": 0.00011123592853931717, + "loss": 0.7566, + "step": 27683 + }, + { + "epoch": 1.9210325445839982, + "grad_norm": 1.301422119140625, + "learning_rate": 0.00011123034548517674, + "loss": 0.4271, + "step": 27684 + }, + { + "epoch": 1.921101936021095, + "grad_norm": 1.6228677034378052, + "learning_rate": 0.00011122476239558344, + "loss": 0.7456, + "step": 27685 + }, + { + "epoch": 1.9211713274581916, + "grad_norm": 1.3115317821502686, + "learning_rate": 0.00011121917927055483, + "loss": 0.5318, + "step": 27686 + }, + { + "epoch": 1.9212407188952882, + "grad_norm": 1.5898933410644531, + "learning_rate": 0.0001112135961101086, + "loss": 0.5086, + "step": 27687 + }, + { + "epoch": 1.921310110332385, + "grad_norm": 1.9261292219161987, + "learning_rate": 0.00011120801291426232, + "loss": 0.7243, + "step": 27688 + }, + { + "epoch": 1.9213795017694817, + "grad_norm": 1.4226348400115967, + "learning_rate": 0.00011120242968303367, + "loss": 0.6469, + "step": 27689 + }, + { + "epoch": 1.9214488932065783, + "grad_norm": 1.2903043031692505, + "learning_rate": 0.00011119684641644024, + "loss": 0.6887, + "step": 27690 + }, + { + "epoch": 1.9215182846436751, + "grad_norm": 1.4409799575805664, + "learning_rate": 0.00011119126311449968, + "loss": 0.5947, + "step": 27691 + }, + { + "epoch": 1.9215876760807715, + "grad_norm": 1.5120402574539185, + "learning_rate": 0.00011118567977722958, + "loss": 0.5402, + "step": 27692 + }, + { + "epoch": 1.9216570675178684, + "grad_norm": 1.0656602382659912, + "learning_rate": 0.00011118009640464761, + "loss": 0.4261, + "step": 27693 + }, + { + "epoch": 1.921726458954965, + "grad_norm": 3.1094725131988525, + "learning_rate": 0.00011117451299677135, + "loss": 0.6167, + "step": 27694 + }, + { + "epoch": 1.9217958503920616, + "grad_norm": 1.5450342893600464, + "learning_rate": 0.00011116892955361844, + "loss": 0.5834, + "step": 27695 + }, + { + "epoch": 1.9218652418291584, + "grad_norm": 1.324501872062683, + "learning_rate": 0.00011116334607520656, + "loss": 0.5157, + "step": 27696 + }, + { + "epoch": 1.9219346332662548, + "grad_norm": 1.4244792461395264, + "learning_rate": 0.00011115776256155323, + "loss": 0.5794, + "step": 27697 + }, + { + "epoch": 1.9220040247033516, + "grad_norm": 1.2553136348724365, + "learning_rate": 0.00011115217901267618, + "loss": 0.5741, + "step": 27698 + }, + { + "epoch": 1.9220734161404482, + "grad_norm": 1.2411915063858032, + "learning_rate": 0.00011114659542859301, + "loss": 0.5366, + "step": 27699 + }, + { + "epoch": 1.9221428075775449, + "grad_norm": 1.1437091827392578, + "learning_rate": 0.00011114101180932129, + "loss": 0.529, + "step": 27700 + }, + { + "epoch": 1.9222121990146417, + "grad_norm": 1.9127308130264282, + "learning_rate": 0.0001111354281548787, + "loss": 0.5817, + "step": 27701 + }, + { + "epoch": 1.9222815904517383, + "grad_norm": 1.3143894672393799, + "learning_rate": 0.00011112984446528286, + "loss": 0.5794, + "step": 27702 + }, + { + "epoch": 1.922350981888835, + "grad_norm": 1.7228091955184937, + "learning_rate": 0.00011112426074055137, + "loss": 0.6371, + "step": 27703 + }, + { + "epoch": 1.9224203733259317, + "grad_norm": 1.410858392715454, + "learning_rate": 0.00011111867698070191, + "loss": 0.5896, + "step": 27704 + }, + { + "epoch": 1.9224897647630281, + "grad_norm": 1.7377485036849976, + "learning_rate": 0.00011111309318575207, + "loss": 0.6246, + "step": 27705 + }, + { + "epoch": 1.922559156200125, + "grad_norm": 1.021048903465271, + "learning_rate": 0.0001111075093557195, + "loss": 0.5604, + "step": 27706 + }, + { + "epoch": 1.9226285476372216, + "grad_norm": 1.6169161796569824, + "learning_rate": 0.00011110192549062177, + "loss": 0.6923, + "step": 27707 + }, + { + "epoch": 1.9226979390743182, + "grad_norm": 1.8252731561660767, + "learning_rate": 0.00011109634159047658, + "loss": 0.7355, + "step": 27708 + }, + { + "epoch": 1.922767330511415, + "grad_norm": 1.5885523557662964, + "learning_rate": 0.00011109075765530153, + "loss": 0.7454, + "step": 27709 + }, + { + "epoch": 1.9228367219485114, + "grad_norm": 1.596540927886963, + "learning_rate": 0.00011108517368511424, + "loss": 0.6379, + "step": 27710 + }, + { + "epoch": 1.9229061133856082, + "grad_norm": 1.6827993392944336, + "learning_rate": 0.00011107958967993235, + "loss": 0.5767, + "step": 27711 + }, + { + "epoch": 1.9229755048227049, + "grad_norm": 1.2691689729690552, + "learning_rate": 0.00011107400563977347, + "loss": 0.3959, + "step": 27712 + }, + { + "epoch": 1.9230448962598015, + "grad_norm": 1.2385685443878174, + "learning_rate": 0.00011106842156465524, + "loss": 0.508, + "step": 27713 + }, + { + "epoch": 1.9231142876968983, + "grad_norm": 1.9339313507080078, + "learning_rate": 0.00011106283745459527, + "loss": 0.8174, + "step": 27714 + }, + { + "epoch": 1.923183679133995, + "grad_norm": 3.463113784790039, + "learning_rate": 0.0001110572533096112, + "loss": 0.7866, + "step": 27715 + }, + { + "epoch": 1.9232530705710915, + "grad_norm": 1.1295502185821533, + "learning_rate": 0.00011105166912972071, + "loss": 0.4555, + "step": 27716 + }, + { + "epoch": 1.9233224620081883, + "grad_norm": 1.577605128288269, + "learning_rate": 0.00011104608491494135, + "loss": 0.4477, + "step": 27717 + }, + { + "epoch": 1.9233918534452847, + "grad_norm": 1.4224106073379517, + "learning_rate": 0.0001110405006652908, + "loss": 0.4631, + "step": 27718 + }, + { + "epoch": 1.9234612448823816, + "grad_norm": 1.4131062030792236, + "learning_rate": 0.00011103491638078665, + "loss": 0.6869, + "step": 27719 + }, + { + "epoch": 1.9235306363194782, + "grad_norm": 1.3779218196868896, + "learning_rate": 0.00011102933206144656, + "loss": 0.6401, + "step": 27720 + }, + { + "epoch": 1.9236000277565748, + "grad_norm": 2.050809383392334, + "learning_rate": 0.00011102374770728814, + "loss": 0.5945, + "step": 27721 + }, + { + "epoch": 1.9236694191936716, + "grad_norm": 1.246573567390442, + "learning_rate": 0.00011101816331832903, + "loss": 0.5401, + "step": 27722 + }, + { + "epoch": 1.923738810630768, + "grad_norm": 1.969213843345642, + "learning_rate": 0.00011101257889458688, + "loss": 0.5533, + "step": 27723 + }, + { + "epoch": 1.9238082020678648, + "grad_norm": 1.3276253938674927, + "learning_rate": 0.00011100699443607927, + "loss": 0.5417, + "step": 27724 + }, + { + "epoch": 1.9238775935049615, + "grad_norm": 1.8943378925323486, + "learning_rate": 0.00011100140994282386, + "loss": 0.507, + "step": 27725 + }, + { + "epoch": 1.923946984942058, + "grad_norm": 1.9184849262237549, + "learning_rate": 0.00011099582541483829, + "loss": 0.5739, + "step": 27726 + }, + { + "epoch": 1.924016376379155, + "grad_norm": 1.5261826515197754, + "learning_rate": 0.00011099024085214016, + "loss": 0.7148, + "step": 27727 + }, + { + "epoch": 1.9240857678162515, + "grad_norm": 2.0548975467681885, + "learning_rate": 0.00011098465625474713, + "loss": 0.5039, + "step": 27728 + }, + { + "epoch": 1.9241551592533481, + "grad_norm": 1.6173065900802612, + "learning_rate": 0.0001109790716226768, + "loss": 0.6514, + "step": 27729 + }, + { + "epoch": 1.924224550690445, + "grad_norm": 2.038611888885498, + "learning_rate": 0.00011097348695594682, + "loss": 0.823, + "step": 27730 + }, + { + "epoch": 1.9242939421275413, + "grad_norm": 1.3074848651885986, + "learning_rate": 0.0001109679022545748, + "loss": 0.457, + "step": 27731 + }, + { + "epoch": 1.9243633335646382, + "grad_norm": 1.6131336688995361, + "learning_rate": 0.00011096231751857839, + "loss": 0.6071, + "step": 27732 + }, + { + "epoch": 1.9244327250017348, + "grad_norm": 1.4491277933120728, + "learning_rate": 0.00011095673274797522, + "loss": 0.5024, + "step": 27733 + }, + { + "epoch": 1.9245021164388314, + "grad_norm": 1.9166710376739502, + "learning_rate": 0.00011095114794278294, + "loss": 0.6642, + "step": 27734 + }, + { + "epoch": 1.9245715078759282, + "grad_norm": 1.5630707740783691, + "learning_rate": 0.00011094556310301913, + "loss": 0.6252, + "step": 27735 + }, + { + "epoch": 1.9246408993130246, + "grad_norm": 1.4543557167053223, + "learning_rate": 0.00011093997822870144, + "loss": 0.5861, + "step": 27736 + }, + { + "epoch": 1.9247102907501215, + "grad_norm": 0.9397665858268738, + "learning_rate": 0.00011093439331984754, + "loss": 0.3983, + "step": 27737 + }, + { + "epoch": 1.924779682187218, + "grad_norm": 1.5822725296020508, + "learning_rate": 0.00011092880837647499, + "loss": 0.6559, + "step": 27738 + }, + { + "epoch": 1.9248490736243147, + "grad_norm": 1.6077661514282227, + "learning_rate": 0.00011092322339860148, + "loss": 0.6397, + "step": 27739 + }, + { + "epoch": 1.9249184650614115, + "grad_norm": 1.2580530643463135, + "learning_rate": 0.0001109176383862446, + "loss": 0.4344, + "step": 27740 + }, + { + "epoch": 1.9249878564985081, + "grad_norm": 1.4230457544326782, + "learning_rate": 0.00011091205333942206, + "loss": 0.4855, + "step": 27741 + }, + { + "epoch": 1.9250572479356047, + "grad_norm": 1.7787196636199951, + "learning_rate": 0.00011090646825815138, + "loss": 0.5428, + "step": 27742 + }, + { + "epoch": 1.9251266393727016, + "grad_norm": 1.5240064859390259, + "learning_rate": 0.00011090088314245028, + "loss": 0.4805, + "step": 27743 + }, + { + "epoch": 1.925196030809798, + "grad_norm": 1.4392962455749512, + "learning_rate": 0.00011089529799233633, + "loss": 0.3951, + "step": 27744 + }, + { + "epoch": 1.9252654222468948, + "grad_norm": 2.0663747787475586, + "learning_rate": 0.00011088971280782721, + "loss": 0.6574, + "step": 27745 + }, + { + "epoch": 1.9253348136839914, + "grad_norm": 1.4268368482589722, + "learning_rate": 0.00011088412758894052, + "loss": 0.5462, + "step": 27746 + }, + { + "epoch": 1.925404205121088, + "grad_norm": 1.4034912586212158, + "learning_rate": 0.00011087854233569388, + "loss": 0.4094, + "step": 27747 + }, + { + "epoch": 1.9254735965581848, + "grad_norm": 2.123687505722046, + "learning_rate": 0.00011087295704810498, + "loss": 0.5918, + "step": 27748 + }, + { + "epoch": 1.9255429879952812, + "grad_norm": 1.9704957008361816, + "learning_rate": 0.0001108673717261914, + "loss": 0.5545, + "step": 27749 + }, + { + "epoch": 1.925612379432378, + "grad_norm": 1.5232082605361938, + "learning_rate": 0.00011086178636997078, + "loss": 0.5296, + "step": 27750 + }, + { + "epoch": 1.9256817708694747, + "grad_norm": 1.014711856842041, + "learning_rate": 0.00011085620097946077, + "loss": 0.4741, + "step": 27751 + }, + { + "epoch": 1.9257511623065713, + "grad_norm": 1.6531338691711426, + "learning_rate": 0.00011085061555467899, + "loss": 0.5296, + "step": 27752 + }, + { + "epoch": 1.9258205537436681, + "grad_norm": 1.6650640964508057, + "learning_rate": 0.00011084503009564306, + "loss": 0.5483, + "step": 27753 + }, + { + "epoch": 1.9258899451807647, + "grad_norm": 1.5323432683944702, + "learning_rate": 0.00011083944460237065, + "loss": 0.6571, + "step": 27754 + }, + { + "epoch": 1.9259593366178613, + "grad_norm": 1.5628252029418945, + "learning_rate": 0.00011083385907487938, + "loss": 0.5023, + "step": 27755 + }, + { + "epoch": 1.9260287280549582, + "grad_norm": 1.5857837200164795, + "learning_rate": 0.00011082827351318685, + "loss": 0.5403, + "step": 27756 + }, + { + "epoch": 1.9260981194920546, + "grad_norm": 1.6500450372695923, + "learning_rate": 0.00011082268791731073, + "loss": 0.4766, + "step": 27757 + }, + { + "epoch": 1.9261675109291514, + "grad_norm": 2.504441261291504, + "learning_rate": 0.00011081710228726863, + "loss": 0.5472, + "step": 27758 + }, + { + "epoch": 1.926236902366248, + "grad_norm": 1.943048119544983, + "learning_rate": 0.0001108115166230782, + "loss": 0.6912, + "step": 27759 + }, + { + "epoch": 1.9263062938033446, + "grad_norm": 1.2157412767410278, + "learning_rate": 0.00011080593092475706, + "loss": 0.5314, + "step": 27760 + }, + { + "epoch": 1.9263756852404414, + "grad_norm": 2.3408851623535156, + "learning_rate": 0.00011080034519232287, + "loss": 0.8567, + "step": 27761 + }, + { + "epoch": 1.9264450766775378, + "grad_norm": 1.4670054912567139, + "learning_rate": 0.0001107947594257932, + "loss": 0.5827, + "step": 27762 + }, + { + "epoch": 1.9265144681146347, + "grad_norm": 1.439712405204773, + "learning_rate": 0.00011078917362518579, + "loss": 0.7557, + "step": 27763 + }, + { + "epoch": 1.9265838595517313, + "grad_norm": 1.8322404623031616, + "learning_rate": 0.00011078358779051816, + "loss": 0.6996, + "step": 27764 + }, + { + "epoch": 1.926653250988828, + "grad_norm": 1.1890205144882202, + "learning_rate": 0.00011077800192180802, + "loss": 0.4671, + "step": 27765 + }, + { + "epoch": 1.9267226424259247, + "grad_norm": 1.676169753074646, + "learning_rate": 0.00011077241601907298, + "loss": 0.5493, + "step": 27766 + }, + { + "epoch": 1.9267920338630213, + "grad_norm": 1.2430305480957031, + "learning_rate": 0.00011076683008233065, + "loss": 0.4011, + "step": 27767 + }, + { + "epoch": 1.926861425300118, + "grad_norm": 2.1801466941833496, + "learning_rate": 0.00011076124411159871, + "loss": 0.5406, + "step": 27768 + }, + { + "epoch": 1.9269308167372148, + "grad_norm": 2.2818732261657715, + "learning_rate": 0.00011075565810689475, + "loss": 0.4043, + "step": 27769 + }, + { + "epoch": 1.9270002081743112, + "grad_norm": 1.7127227783203125, + "learning_rate": 0.00011075007206823642, + "loss": 0.6125, + "step": 27770 + }, + { + "epoch": 1.927069599611408, + "grad_norm": 2.3075830936431885, + "learning_rate": 0.00011074448599564138, + "loss": 0.6852, + "step": 27771 + }, + { + "epoch": 1.9271389910485046, + "grad_norm": 2.8618757724761963, + "learning_rate": 0.00011073889988912724, + "loss": 0.6435, + "step": 27772 + }, + { + "epoch": 1.9272083824856012, + "grad_norm": 1.3947128057479858, + "learning_rate": 0.00011073331374871163, + "loss": 0.5331, + "step": 27773 + }, + { + "epoch": 1.927277773922698, + "grad_norm": 1.3604357242584229, + "learning_rate": 0.00011072772757441222, + "loss": 0.5559, + "step": 27774 + }, + { + "epoch": 1.9273471653597944, + "grad_norm": 1.692846655845642, + "learning_rate": 0.00011072214136624659, + "loss": 0.7456, + "step": 27775 + }, + { + "epoch": 1.9274165567968913, + "grad_norm": 1.63394296169281, + "learning_rate": 0.0001107165551242324, + "loss": 0.5081, + "step": 27776 + }, + { + "epoch": 1.927485948233988, + "grad_norm": 1.5410875082015991, + "learning_rate": 0.0001107109688483873, + "loss": 0.5471, + "step": 27777 + }, + { + "epoch": 1.9275553396710845, + "grad_norm": 1.8533685207366943, + "learning_rate": 0.00011070538253872893, + "loss": 0.5318, + "step": 27778 + }, + { + "epoch": 1.9276247311081813, + "grad_norm": 1.7428308725357056, + "learning_rate": 0.00011069979619527489, + "loss": 0.6234, + "step": 27779 + }, + { + "epoch": 1.927694122545278, + "grad_norm": 1.65537428855896, + "learning_rate": 0.00011069420981804286, + "loss": 0.6353, + "step": 27780 + }, + { + "epoch": 1.9277635139823746, + "grad_norm": 1.624923586845398, + "learning_rate": 0.00011068862340705042, + "loss": 0.5928, + "step": 27781 + }, + { + "epoch": 1.9278329054194714, + "grad_norm": 1.9324339628219604, + "learning_rate": 0.00011068303696231526, + "loss": 0.4666, + "step": 27782 + }, + { + "epoch": 1.9279022968565678, + "grad_norm": 1.7187788486480713, + "learning_rate": 0.00011067745048385498, + "loss": 0.5328, + "step": 27783 + }, + { + "epoch": 1.9279716882936646, + "grad_norm": 2.0716817378997803, + "learning_rate": 0.00011067186397168723, + "loss": 0.7794, + "step": 27784 + }, + { + "epoch": 1.9280410797307612, + "grad_norm": 2.351790189743042, + "learning_rate": 0.00011066627742582962, + "loss": 0.713, + "step": 27785 + }, + { + "epoch": 1.9281104711678578, + "grad_norm": 1.1803784370422363, + "learning_rate": 0.00011066069084629985, + "loss": 0.2415, + "step": 27786 + }, + { + "epoch": 1.9281798626049547, + "grad_norm": 1.4414459466934204, + "learning_rate": 0.0001106551042331155, + "loss": 0.7198, + "step": 27787 + }, + { + "epoch": 1.928249254042051, + "grad_norm": 1.4290266036987305, + "learning_rate": 0.00011064951758629418, + "loss": 0.5561, + "step": 27788 + }, + { + "epoch": 1.9283186454791479, + "grad_norm": 2.243595838546753, + "learning_rate": 0.00011064393090585363, + "loss": 0.6333, + "step": 27789 + }, + { + "epoch": 1.9283880369162445, + "grad_norm": 1.5551470518112183, + "learning_rate": 0.00011063834419181141, + "loss": 0.8046, + "step": 27790 + }, + { + "epoch": 1.928457428353341, + "grad_norm": 5.754045486450195, + "learning_rate": 0.00011063275744418515, + "loss": 0.5448, + "step": 27791 + }, + { + "epoch": 1.928526819790438, + "grad_norm": 1.4407365322113037, + "learning_rate": 0.00011062717066299252, + "loss": 0.5337, + "step": 27792 + }, + { + "epoch": 1.9285962112275346, + "grad_norm": 1.3093971014022827, + "learning_rate": 0.00011062158384825114, + "loss": 0.3494, + "step": 27793 + }, + { + "epoch": 1.9286656026646312, + "grad_norm": 1.5026919841766357, + "learning_rate": 0.00011061599699997867, + "loss": 0.4476, + "step": 27794 + }, + { + "epoch": 1.928734994101728, + "grad_norm": 1.2432022094726562, + "learning_rate": 0.00011061041011819272, + "loss": 0.6251, + "step": 27795 + }, + { + "epoch": 1.9288043855388244, + "grad_norm": 2.0316390991210938, + "learning_rate": 0.00011060482320291092, + "loss": 0.5011, + "step": 27796 + }, + { + "epoch": 1.9288737769759212, + "grad_norm": 1.5813230276107788, + "learning_rate": 0.00011059923625415094, + "loss": 0.4467, + "step": 27797 + }, + { + "epoch": 1.9289431684130178, + "grad_norm": 1.4567456245422363, + "learning_rate": 0.00011059364927193041, + "loss": 0.4699, + "step": 27798 + }, + { + "epoch": 1.9290125598501144, + "grad_norm": 2.249929666519165, + "learning_rate": 0.00011058806225626696, + "loss": 0.7975, + "step": 27799 + }, + { + "epoch": 1.9290819512872113, + "grad_norm": 1.6473898887634277, + "learning_rate": 0.00011058247520717822, + "loss": 0.5765, + "step": 27800 + }, + { + "epoch": 1.9291513427243077, + "grad_norm": 2.1772677898406982, + "learning_rate": 0.00011057688812468181, + "loss": 0.5861, + "step": 27801 + }, + { + "epoch": 1.9292207341614045, + "grad_norm": 1.2814387083053589, + "learning_rate": 0.0001105713010087954, + "loss": 0.4712, + "step": 27802 + }, + { + "epoch": 1.929290125598501, + "grad_norm": 1.3734447956085205, + "learning_rate": 0.00011056571385953665, + "loss": 0.5919, + "step": 27803 + }, + { + "epoch": 1.9293595170355977, + "grad_norm": 2.010342597961426, + "learning_rate": 0.00011056012667692313, + "loss": 0.6388, + "step": 27804 + }, + { + "epoch": 1.9294289084726945, + "grad_norm": 1.0540919303894043, + "learning_rate": 0.00011055453946097251, + "loss": 0.3223, + "step": 27805 + }, + { + "epoch": 1.9294982999097912, + "grad_norm": 1.4615211486816406, + "learning_rate": 0.00011054895221170247, + "loss": 0.5504, + "step": 27806 + }, + { + "epoch": 1.9295676913468878, + "grad_norm": 1.388541340827942, + "learning_rate": 0.0001105433649291306, + "loss": 0.4898, + "step": 27807 + }, + { + "epoch": 1.9296370827839846, + "grad_norm": 1.4749360084533691, + "learning_rate": 0.00011053777761327453, + "loss": 0.5362, + "step": 27808 + }, + { + "epoch": 1.929706474221081, + "grad_norm": 1.600319504737854, + "learning_rate": 0.00011053219026415193, + "loss": 0.5202, + "step": 27809 + }, + { + "epoch": 1.9297758656581778, + "grad_norm": 1.4108564853668213, + "learning_rate": 0.00011052660288178045, + "loss": 0.5192, + "step": 27810 + }, + { + "epoch": 1.9298452570952744, + "grad_norm": 2.432410478591919, + "learning_rate": 0.00011052101546617768, + "loss": 0.7305, + "step": 27811 + }, + { + "epoch": 1.929914648532371, + "grad_norm": 1.7261000871658325, + "learning_rate": 0.00011051542801736128, + "loss": 0.7114, + "step": 27812 + }, + { + "epoch": 1.9299840399694679, + "grad_norm": 2.2274084091186523, + "learning_rate": 0.00011050984053534892, + "loss": 0.711, + "step": 27813 + }, + { + "epoch": 1.9300534314065643, + "grad_norm": 1.3850185871124268, + "learning_rate": 0.0001105042530201582, + "loss": 0.5065, + "step": 27814 + }, + { + "epoch": 1.930122822843661, + "grad_norm": 2.2071988582611084, + "learning_rate": 0.00011049866547180678, + "loss": 0.5781, + "step": 27815 + }, + { + "epoch": 1.9301922142807577, + "grad_norm": 1.7111287117004395, + "learning_rate": 0.00011049307789031227, + "loss": 0.5189, + "step": 27816 + }, + { + "epoch": 1.9302616057178543, + "grad_norm": 1.35882568359375, + "learning_rate": 0.00011048749027569236, + "loss": 0.5325, + "step": 27817 + }, + { + "epoch": 1.9303309971549512, + "grad_norm": 1.6341722011566162, + "learning_rate": 0.00011048190262796463, + "loss": 0.6075, + "step": 27818 + }, + { + "epoch": 1.9304003885920478, + "grad_norm": 1.707092046737671, + "learning_rate": 0.00011047631494714674, + "loss": 0.5877, + "step": 27819 + }, + { + "epoch": 1.9304697800291444, + "grad_norm": 1.3729618787765503, + "learning_rate": 0.0001104707272332564, + "loss": 0.4421, + "step": 27820 + }, + { + "epoch": 1.9305391714662412, + "grad_norm": 2.088477373123169, + "learning_rate": 0.00011046513948631113, + "loss": 0.7404, + "step": 27821 + }, + { + "epoch": 1.9306085629033376, + "grad_norm": 1.7252227067947388, + "learning_rate": 0.00011045955170632865, + "loss": 0.4896, + "step": 27822 + }, + { + "epoch": 1.9306779543404344, + "grad_norm": 1.780632734298706, + "learning_rate": 0.00011045396389332656, + "loss": 0.4904, + "step": 27823 + }, + { + "epoch": 1.930747345777531, + "grad_norm": 1.5008971691131592, + "learning_rate": 0.00011044837604732254, + "loss": 0.5262, + "step": 27824 + }, + { + "epoch": 1.9308167372146277, + "grad_norm": 1.674634337425232, + "learning_rate": 0.0001104427881683342, + "loss": 0.5827, + "step": 27825 + }, + { + "epoch": 1.9308861286517245, + "grad_norm": 7.518560886383057, + "learning_rate": 0.0001104372002563792, + "loss": 0.5048, + "step": 27826 + }, + { + "epoch": 1.9309555200888209, + "grad_norm": 1.5403549671173096, + "learning_rate": 0.00011043161231147519, + "loss": 0.6735, + "step": 27827 + }, + { + "epoch": 1.9310249115259177, + "grad_norm": 1.4377951622009277, + "learning_rate": 0.00011042602433363973, + "loss": 0.5741, + "step": 27828 + }, + { + "epoch": 1.9310943029630143, + "grad_norm": 1.9151793718338013, + "learning_rate": 0.00011042043632289057, + "loss": 0.5886, + "step": 27829 + }, + { + "epoch": 1.931163694400111, + "grad_norm": 1.4794975519180298, + "learning_rate": 0.00011041484827924529, + "loss": 0.6447, + "step": 27830 + }, + { + "epoch": 1.9312330858372078, + "grad_norm": 1.3158825635910034, + "learning_rate": 0.00011040926020272153, + "loss": 0.5365, + "step": 27831 + }, + { + "epoch": 1.9313024772743044, + "grad_norm": 1.3348461389541626, + "learning_rate": 0.00011040367209333698, + "loss": 0.5032, + "step": 27832 + }, + { + "epoch": 1.931371868711401, + "grad_norm": 2.016411542892456, + "learning_rate": 0.0001103980839511092, + "loss": 0.54, + "step": 27833 + }, + { + "epoch": 1.9314412601484978, + "grad_norm": 1.6843235492706299, + "learning_rate": 0.0001103924957760559, + "loss": 0.7243, + "step": 27834 + }, + { + "epoch": 1.9315106515855942, + "grad_norm": 2.4250333309173584, + "learning_rate": 0.00011038690756819469, + "loss": 0.8885, + "step": 27835 + }, + { + "epoch": 1.931580043022691, + "grad_norm": 1.7886637449264526, + "learning_rate": 0.00011038131932754321, + "loss": 0.3836, + "step": 27836 + }, + { + "epoch": 1.9316494344597877, + "grad_norm": 1.6984885931015015, + "learning_rate": 0.00011037573105411914, + "loss": 0.6824, + "step": 27837 + }, + { + "epoch": 1.9317188258968843, + "grad_norm": 1.7131588459014893, + "learning_rate": 0.00011037014274794004, + "loss": 0.7683, + "step": 27838 + }, + { + "epoch": 1.931788217333981, + "grad_norm": 1.9897738695144653, + "learning_rate": 0.00011036455440902363, + "loss": 0.7513, + "step": 27839 + }, + { + "epoch": 1.9318576087710775, + "grad_norm": 1.4745899438858032, + "learning_rate": 0.00011035896603738751, + "loss": 0.675, + "step": 27840 + }, + { + "epoch": 1.9319270002081743, + "grad_norm": 2.3492071628570557, + "learning_rate": 0.00011035337763304936, + "loss": 0.8112, + "step": 27841 + }, + { + "epoch": 1.931996391645271, + "grad_norm": 1.7798055410385132, + "learning_rate": 0.0001103477891960268, + "loss": 0.5731, + "step": 27842 + }, + { + "epoch": 1.9320657830823675, + "grad_norm": 1.6310144662857056, + "learning_rate": 0.00011034220072633743, + "loss": 0.5951, + "step": 27843 + }, + { + "epoch": 1.9321351745194644, + "grad_norm": 1.915094017982483, + "learning_rate": 0.00011033661222399898, + "loss": 0.6764, + "step": 27844 + }, + { + "epoch": 1.932204565956561, + "grad_norm": 1.3835619688034058, + "learning_rate": 0.00011033102368902901, + "loss": 0.5855, + "step": 27845 + }, + { + "epoch": 1.9322739573936576, + "grad_norm": 1.5764119625091553, + "learning_rate": 0.00011032543512144522, + "loss": 0.7092, + "step": 27846 + }, + { + "epoch": 1.9323433488307544, + "grad_norm": 2.7178916931152344, + "learning_rate": 0.00011031984652126521, + "loss": 0.9072, + "step": 27847 + }, + { + "epoch": 1.9324127402678508, + "grad_norm": 1.415383219718933, + "learning_rate": 0.00011031425788850667, + "loss": 0.508, + "step": 27848 + }, + { + "epoch": 1.9324821317049476, + "grad_norm": 3.6783928871154785, + "learning_rate": 0.00011030866922318719, + "loss": 0.5229, + "step": 27849 + }, + { + "epoch": 1.9325515231420443, + "grad_norm": 1.5050325393676758, + "learning_rate": 0.00011030308052532445, + "loss": 0.6552, + "step": 27850 + }, + { + "epoch": 1.9326209145791409, + "grad_norm": 1.2793694734573364, + "learning_rate": 0.00011029749179493606, + "loss": 0.5378, + "step": 27851 + }, + { + "epoch": 1.9326903060162377, + "grad_norm": 1.4711209535598755, + "learning_rate": 0.00011029190303203971, + "loss": 0.5569, + "step": 27852 + }, + { + "epoch": 1.9327596974533343, + "grad_norm": 1.4978419542312622, + "learning_rate": 0.00011028631423665301, + "loss": 0.6533, + "step": 27853 + }, + { + "epoch": 1.932829088890431, + "grad_norm": 1.4061769247055054, + "learning_rate": 0.00011028072540879359, + "loss": 0.5342, + "step": 27854 + }, + { + "epoch": 1.9328984803275275, + "grad_norm": 2.243966579437256, + "learning_rate": 0.00011027513654847915, + "loss": 0.7063, + "step": 27855 + }, + { + "epoch": 1.9329678717646241, + "grad_norm": 1.6523011922836304, + "learning_rate": 0.00011026954765572726, + "loss": 0.4492, + "step": 27856 + }, + { + "epoch": 1.933037263201721, + "grad_norm": 0.9819595813751221, + "learning_rate": 0.0001102639587305556, + "loss": 0.4176, + "step": 27857 + }, + { + "epoch": 1.9331066546388176, + "grad_norm": 2.6797196865081787, + "learning_rate": 0.00011025836977298183, + "loss": 0.5096, + "step": 27858 + }, + { + "epoch": 1.9331760460759142, + "grad_norm": 1.3793216943740845, + "learning_rate": 0.00011025278078302356, + "loss": 0.6322, + "step": 27859 + }, + { + "epoch": 1.933245437513011, + "grad_norm": 1.533010482788086, + "learning_rate": 0.00011024719176069847, + "loss": 0.6079, + "step": 27860 + }, + { + "epoch": 1.9333148289501074, + "grad_norm": 1.3778599500656128, + "learning_rate": 0.00011024160270602417, + "loss": 0.4955, + "step": 27861 + }, + { + "epoch": 1.9333842203872043, + "grad_norm": 1.6249232292175293, + "learning_rate": 0.00011023601361901834, + "loss": 0.4943, + "step": 27862 + }, + { + "epoch": 1.9334536118243009, + "grad_norm": 1.7662596702575684, + "learning_rate": 0.00011023042449969858, + "loss": 0.5712, + "step": 27863 + }, + { + "epoch": 1.9335230032613975, + "grad_norm": 2.5004196166992188, + "learning_rate": 0.00011022483534808254, + "loss": 0.6284, + "step": 27864 + }, + { + "epoch": 1.9335923946984943, + "grad_norm": 1.5179871320724487, + "learning_rate": 0.00011021924616418793, + "loss": 0.4948, + "step": 27865 + }, + { + "epoch": 1.933661786135591, + "grad_norm": 2.0199599266052246, + "learning_rate": 0.00011021365694803232, + "loss": 0.6267, + "step": 27866 + }, + { + "epoch": 1.9337311775726875, + "grad_norm": 1.8070086240768433, + "learning_rate": 0.00011020806769963339, + "loss": 0.4083, + "step": 27867 + }, + { + "epoch": 1.9338005690097841, + "grad_norm": 1.4110442399978638, + "learning_rate": 0.00011020247841900875, + "loss": 0.4518, + "step": 27868 + }, + { + "epoch": 1.9338699604468808, + "grad_norm": 1.5618928670883179, + "learning_rate": 0.00011019688910617612, + "loss": 0.708, + "step": 27869 + }, + { + "epoch": 1.9339393518839776, + "grad_norm": 1.3430285453796387, + "learning_rate": 0.00011019129976115304, + "loss": 0.4013, + "step": 27870 + }, + { + "epoch": 1.9340087433210742, + "grad_norm": 1.7303247451782227, + "learning_rate": 0.00011018571038395721, + "loss": 0.7565, + "step": 27871 + }, + { + "epoch": 1.9340781347581708, + "grad_norm": 1.5107479095458984, + "learning_rate": 0.00011018012097460633, + "loss": 0.5696, + "step": 27872 + }, + { + "epoch": 1.9341475261952676, + "grad_norm": 1.7554117441177368, + "learning_rate": 0.00011017453153311791, + "loss": 0.4345, + "step": 27873 + }, + { + "epoch": 1.934216917632364, + "grad_norm": 1.2914750576019287, + "learning_rate": 0.00011016894205950974, + "loss": 0.5508, + "step": 27874 + }, + { + "epoch": 1.9342863090694609, + "grad_norm": 1.4660736322402954, + "learning_rate": 0.00011016335255379937, + "loss": 0.5491, + "step": 27875 + }, + { + "epoch": 1.9343557005065575, + "grad_norm": 2.0268547534942627, + "learning_rate": 0.00011015776301600449, + "loss": 0.7495, + "step": 27876 + }, + { + "epoch": 1.934425091943654, + "grad_norm": 1.3443807363510132, + "learning_rate": 0.0001101521734461427, + "loss": 0.5074, + "step": 27877 + }, + { + "epoch": 1.934494483380751, + "grad_norm": 1.466284155845642, + "learning_rate": 0.0001101465838442317, + "loss": 0.6203, + "step": 27878 + }, + { + "epoch": 1.9345638748178475, + "grad_norm": 1.7139225006103516, + "learning_rate": 0.00011014099421028911, + "loss": 0.5063, + "step": 27879 + }, + { + "epoch": 1.9346332662549441, + "grad_norm": 1.426743745803833, + "learning_rate": 0.00011013540454433258, + "loss": 0.6713, + "step": 27880 + }, + { + "epoch": 1.9347026576920408, + "grad_norm": 1.6256635189056396, + "learning_rate": 0.00011012981484637974, + "loss": 0.5132, + "step": 27881 + }, + { + "epoch": 1.9347720491291374, + "grad_norm": 1.2256202697753906, + "learning_rate": 0.00011012422511644829, + "loss": 0.4307, + "step": 27882 + }, + { + "epoch": 1.9348414405662342, + "grad_norm": 2.100052833557129, + "learning_rate": 0.0001101186353545558, + "loss": 0.5539, + "step": 27883 + }, + { + "epoch": 1.9349108320033308, + "grad_norm": 1.7949914932250977, + "learning_rate": 0.00011011304556071993, + "loss": 0.8875, + "step": 27884 + }, + { + "epoch": 1.9349802234404274, + "grad_norm": 1.3530693054199219, + "learning_rate": 0.00011010745573495839, + "loss": 0.5997, + "step": 27885 + }, + { + "epoch": 1.9350496148775242, + "grad_norm": 1.1418622732162476, + "learning_rate": 0.0001101018658772888, + "loss": 0.4768, + "step": 27886 + }, + { + "epoch": 1.9351190063146206, + "grad_norm": 1.8149404525756836, + "learning_rate": 0.00011009627598772876, + "loss": 0.6801, + "step": 27887 + }, + { + "epoch": 1.9351883977517175, + "grad_norm": 1.8265001773834229, + "learning_rate": 0.00011009068606629594, + "loss": 0.7694, + "step": 27888 + }, + { + "epoch": 1.935257789188814, + "grad_norm": 1.2984731197357178, + "learning_rate": 0.00011008509611300802, + "loss": 0.4259, + "step": 27889 + }, + { + "epoch": 1.9353271806259107, + "grad_norm": 1.3106005191802979, + "learning_rate": 0.00011007950612788261, + "loss": 0.4376, + "step": 27890 + }, + { + "epoch": 1.9353965720630075, + "grad_norm": 2.0524842739105225, + "learning_rate": 0.00011007391611093736, + "loss": 0.5794, + "step": 27891 + }, + { + "epoch": 1.9354659635001041, + "grad_norm": 1.6401269435882568, + "learning_rate": 0.0001100683260621899, + "loss": 0.594, + "step": 27892 + }, + { + "epoch": 1.9355353549372007, + "grad_norm": 1.6990082263946533, + "learning_rate": 0.00011006273598165797, + "loss": 0.3972, + "step": 27893 + }, + { + "epoch": 1.9356047463742974, + "grad_norm": 1.624245524406433, + "learning_rate": 0.0001100571458693591, + "loss": 0.5331, + "step": 27894 + }, + { + "epoch": 1.935674137811394, + "grad_norm": 1.4552561044692993, + "learning_rate": 0.000110051555725311, + "loss": 0.5595, + "step": 27895 + }, + { + "epoch": 1.9357435292484908, + "grad_norm": 1.2189955711364746, + "learning_rate": 0.00011004596554953133, + "loss": 0.3792, + "step": 27896 + }, + { + "epoch": 1.9358129206855874, + "grad_norm": 2.2354044914245605, + "learning_rate": 0.00011004037534203768, + "loss": 0.7046, + "step": 27897 + }, + { + "epoch": 1.935882312122684, + "grad_norm": 1.6899501085281372, + "learning_rate": 0.00011003478510284772, + "loss": 0.4856, + "step": 27898 + }, + { + "epoch": 1.9359517035597809, + "grad_norm": 1.8943947553634644, + "learning_rate": 0.00011002919483197914, + "loss": 0.6758, + "step": 27899 + }, + { + "epoch": 1.9360210949968772, + "grad_norm": 1.5280687808990479, + "learning_rate": 0.00011002360452944956, + "loss": 0.5048, + "step": 27900 + }, + { + "epoch": 1.936090486433974, + "grad_norm": 2.046823501586914, + "learning_rate": 0.0001100180141952766, + "loss": 0.5773, + "step": 27901 + }, + { + "epoch": 1.9361598778710707, + "grad_norm": 1.3431100845336914, + "learning_rate": 0.00011001242382947795, + "loss": 0.3831, + "step": 27902 + }, + { + "epoch": 1.9362292693081673, + "grad_norm": 2.08754825592041, + "learning_rate": 0.0001100068334320712, + "loss": 0.5879, + "step": 27903 + }, + { + "epoch": 1.9362986607452641, + "grad_norm": 2.0290541648864746, + "learning_rate": 0.00011000124300307409, + "loss": 0.6303, + "step": 27904 + }, + { + "epoch": 1.9363680521823607, + "grad_norm": 2.6569111347198486, + "learning_rate": 0.0001099956525425042, + "loss": 0.7054, + "step": 27905 + }, + { + "epoch": 1.9364374436194574, + "grad_norm": 1.234817624092102, + "learning_rate": 0.00010999006205037917, + "loss": 0.4507, + "step": 27906 + }, + { + "epoch": 1.9365068350565542, + "grad_norm": 1.783172369003296, + "learning_rate": 0.00010998447152671669, + "loss": 0.4556, + "step": 27907 + }, + { + "epoch": 1.9365762264936506, + "grad_norm": 1.6979174613952637, + "learning_rate": 0.0001099788809715344, + "loss": 0.5157, + "step": 27908 + }, + { + "epoch": 1.9366456179307474, + "grad_norm": 1.8567240238189697, + "learning_rate": 0.00010997329038484991, + "loss": 0.6681, + "step": 27909 + }, + { + "epoch": 1.936715009367844, + "grad_norm": 1.9843770265579224, + "learning_rate": 0.00010996769976668093, + "loss": 0.5916, + "step": 27910 + }, + { + "epoch": 1.9367844008049406, + "grad_norm": 2.0433402061462402, + "learning_rate": 0.00010996210911704507, + "loss": 0.5831, + "step": 27911 + }, + { + "epoch": 1.9368537922420375, + "grad_norm": 1.9557075500488281, + "learning_rate": 0.00010995651843595998, + "loss": 0.5091, + "step": 27912 + }, + { + "epoch": 1.9369231836791339, + "grad_norm": 1.6967355012893677, + "learning_rate": 0.00010995092772344333, + "loss": 0.591, + "step": 27913 + }, + { + "epoch": 1.9369925751162307, + "grad_norm": 1.4298198223114014, + "learning_rate": 0.00010994533697951275, + "loss": 0.6709, + "step": 27914 + }, + { + "epoch": 1.9370619665533273, + "grad_norm": 1.7372814416885376, + "learning_rate": 0.00010993974620418588, + "loss": 0.5411, + "step": 27915 + }, + { + "epoch": 1.937131357990424, + "grad_norm": 1.8858287334442139, + "learning_rate": 0.0001099341553974804, + "loss": 0.4931, + "step": 27916 + }, + { + "epoch": 1.9372007494275207, + "grad_norm": 1.4332631826400757, + "learning_rate": 0.00010992856455941396, + "loss": 0.4675, + "step": 27917 + }, + { + "epoch": 1.9372701408646174, + "grad_norm": 1.500624179840088, + "learning_rate": 0.00010992297369000418, + "loss": 0.4803, + "step": 27918 + }, + { + "epoch": 1.937339532301714, + "grad_norm": 1.7311346530914307, + "learning_rate": 0.00010991738278926873, + "loss": 0.5671, + "step": 27919 + }, + { + "epoch": 1.9374089237388108, + "grad_norm": 1.4639496803283691, + "learning_rate": 0.00010991179185722522, + "loss": 0.3527, + "step": 27920 + }, + { + "epoch": 1.9374783151759072, + "grad_norm": 2.0655646324157715, + "learning_rate": 0.00010990620089389137, + "loss": 0.6409, + "step": 27921 + }, + { + "epoch": 1.937547706613004, + "grad_norm": 1.6600289344787598, + "learning_rate": 0.00010990060989928476, + "loss": 0.815, + "step": 27922 + }, + { + "epoch": 1.9376170980501006, + "grad_norm": 2.129060983657837, + "learning_rate": 0.0001098950188734231, + "loss": 0.6467, + "step": 27923 + }, + { + "epoch": 1.9376864894871972, + "grad_norm": 2.2721598148345947, + "learning_rate": 0.00010988942781632402, + "loss": 0.633, + "step": 27924 + }, + { + "epoch": 1.937755880924294, + "grad_norm": 1.5918502807617188, + "learning_rate": 0.00010988383672800515, + "loss": 0.5566, + "step": 27925 + }, + { + "epoch": 1.9378252723613905, + "grad_norm": 1.927267074584961, + "learning_rate": 0.00010987824560848415, + "loss": 0.6664, + "step": 27926 + }, + { + "epoch": 1.9378946637984873, + "grad_norm": 1.6835534572601318, + "learning_rate": 0.00010987265445777868, + "loss": 0.6198, + "step": 27927 + }, + { + "epoch": 1.937964055235584, + "grad_norm": 1.5803483724594116, + "learning_rate": 0.00010986706327590636, + "loss": 0.4915, + "step": 27928 + }, + { + "epoch": 1.9380334466726805, + "grad_norm": 1.6335029602050781, + "learning_rate": 0.00010986147206288491, + "loss": 0.5521, + "step": 27929 + }, + { + "epoch": 1.9381028381097773, + "grad_norm": 1.274368166923523, + "learning_rate": 0.0001098558808187319, + "loss": 0.3945, + "step": 27930 + }, + { + "epoch": 1.938172229546874, + "grad_norm": 1.534406065940857, + "learning_rate": 0.00010985028954346505, + "loss": 0.5637, + "step": 27931 + }, + { + "epoch": 1.9382416209839706, + "grad_norm": 3.9168646335601807, + "learning_rate": 0.00010984469823710195, + "loss": 0.6189, + "step": 27932 + }, + { + "epoch": 1.9383110124210674, + "grad_norm": 1.5049315690994263, + "learning_rate": 0.00010983910689966028, + "loss": 0.4384, + "step": 27933 + }, + { + "epoch": 1.9383804038581638, + "grad_norm": 1.1896265745162964, + "learning_rate": 0.00010983351553115772, + "loss": 0.3385, + "step": 27934 + }, + { + "epoch": 1.9384497952952606, + "grad_norm": 1.4518892765045166, + "learning_rate": 0.00010982792413161187, + "loss": 0.4282, + "step": 27935 + }, + { + "epoch": 1.9385191867323572, + "grad_norm": 1.65552818775177, + "learning_rate": 0.0001098223327010404, + "loss": 0.722, + "step": 27936 + }, + { + "epoch": 1.9385885781694538, + "grad_norm": 1.464772343635559, + "learning_rate": 0.000109816741239461, + "loss": 0.4872, + "step": 27937 + }, + { + "epoch": 1.9386579696065507, + "grad_norm": 1.587453842163086, + "learning_rate": 0.00010981114974689126, + "loss": 0.7235, + "step": 27938 + }, + { + "epoch": 1.938727361043647, + "grad_norm": 1.9041789770126343, + "learning_rate": 0.00010980555822334886, + "loss": 0.6555, + "step": 27939 + }, + { + "epoch": 1.938796752480744, + "grad_norm": 1.220838189125061, + "learning_rate": 0.00010979996666885146, + "loss": 0.5067, + "step": 27940 + }, + { + "epoch": 1.9388661439178405, + "grad_norm": 2.331143617630005, + "learning_rate": 0.0001097943750834167, + "loss": 0.4679, + "step": 27941 + }, + { + "epoch": 1.9389355353549371, + "grad_norm": 2.806858777999878, + "learning_rate": 0.00010978878346706222, + "loss": 0.8849, + "step": 27942 + }, + { + "epoch": 1.939004926792034, + "grad_norm": 1.2762030363082886, + "learning_rate": 0.0001097831918198057, + "loss": 0.5668, + "step": 27943 + }, + { + "epoch": 1.9390743182291306, + "grad_norm": 1.464495062828064, + "learning_rate": 0.00010977760014166476, + "loss": 0.544, + "step": 27944 + }, + { + "epoch": 1.9391437096662272, + "grad_norm": 1.558761477470398, + "learning_rate": 0.0001097720084326571, + "loss": 0.5489, + "step": 27945 + }, + { + "epoch": 1.939213101103324, + "grad_norm": 1.5654047727584839, + "learning_rate": 0.0001097664166928003, + "loss": 0.5849, + "step": 27946 + }, + { + "epoch": 1.9392824925404204, + "grad_norm": 1.6913613080978394, + "learning_rate": 0.00010976082492211208, + "loss": 0.6156, + "step": 27947 + }, + { + "epoch": 1.9393518839775172, + "grad_norm": 1.2592458724975586, + "learning_rate": 0.00010975523312061009, + "loss": 0.5316, + "step": 27948 + }, + { + "epoch": 1.9394212754146138, + "grad_norm": 1.516851782798767, + "learning_rate": 0.00010974964128831193, + "loss": 0.633, + "step": 27949 + }, + { + "epoch": 1.9394906668517105, + "grad_norm": 1.9369401931762695, + "learning_rate": 0.00010974404942523528, + "loss": 0.5898, + "step": 27950 + }, + { + "epoch": 1.9395600582888073, + "grad_norm": 1.5137181282043457, + "learning_rate": 0.00010973845753139784, + "loss": 0.6009, + "step": 27951 + }, + { + "epoch": 1.9396294497259037, + "grad_norm": 1.6507925987243652, + "learning_rate": 0.00010973286560681717, + "loss": 0.5788, + "step": 27952 + }, + { + "epoch": 1.9396988411630005, + "grad_norm": 1.3040549755096436, + "learning_rate": 0.00010972727365151099, + "loss": 0.4369, + "step": 27953 + }, + { + "epoch": 1.9397682326000971, + "grad_norm": 1.2019630670547485, + "learning_rate": 0.00010972168166549695, + "loss": 0.7121, + "step": 27954 + }, + { + "epoch": 1.9398376240371937, + "grad_norm": 1.265655517578125, + "learning_rate": 0.00010971608964879268, + "loss": 0.3992, + "step": 27955 + }, + { + "epoch": 1.9399070154742906, + "grad_norm": 2.343358039855957, + "learning_rate": 0.00010971049760141585, + "loss": 0.651, + "step": 27956 + }, + { + "epoch": 1.9399764069113872, + "grad_norm": 1.3913283348083496, + "learning_rate": 0.00010970490552338408, + "loss": 0.5053, + "step": 27957 + }, + { + "epoch": 1.9400457983484838, + "grad_norm": 1.312822937965393, + "learning_rate": 0.00010969931341471509, + "loss": 0.5005, + "step": 27958 + }, + { + "epoch": 1.9401151897855806, + "grad_norm": 1.3780231475830078, + "learning_rate": 0.00010969372127542645, + "loss": 0.5641, + "step": 27959 + }, + { + "epoch": 1.940184581222677, + "grad_norm": 1.3065881729125977, + "learning_rate": 0.00010968812910553587, + "loss": 0.6041, + "step": 27960 + }, + { + "epoch": 1.9402539726597738, + "grad_norm": 1.3530939817428589, + "learning_rate": 0.00010968253690506099, + "loss": 0.4976, + "step": 27961 + }, + { + "epoch": 1.9403233640968705, + "grad_norm": 1.6446378231048584, + "learning_rate": 0.00010967694467401948, + "loss": 0.6017, + "step": 27962 + }, + { + "epoch": 1.940392755533967, + "grad_norm": 2.132667303085327, + "learning_rate": 0.00010967135241242897, + "loss": 0.5608, + "step": 27963 + }, + { + "epoch": 1.940462146971064, + "grad_norm": 1.9880366325378418, + "learning_rate": 0.00010966576012030711, + "loss": 0.6715, + "step": 27964 + }, + { + "epoch": 1.9405315384081603, + "grad_norm": 1.2146581411361694, + "learning_rate": 0.00010966016779767158, + "loss": 0.5161, + "step": 27965 + }, + { + "epoch": 1.9406009298452571, + "grad_norm": 1.746273398399353, + "learning_rate": 0.00010965457544454, + "loss": 0.4846, + "step": 27966 + }, + { + "epoch": 1.9406703212823537, + "grad_norm": 1.8766088485717773, + "learning_rate": 0.00010964898306093008, + "loss": 0.6587, + "step": 27967 + }, + { + "epoch": 1.9407397127194503, + "grad_norm": 1.6580557823181152, + "learning_rate": 0.0001096433906468594, + "loss": 0.5881, + "step": 27968 + }, + { + "epoch": 1.9408091041565472, + "grad_norm": 2.0482256412506104, + "learning_rate": 0.0001096377982023457, + "loss": 0.6623, + "step": 27969 + }, + { + "epoch": 1.9408784955936438, + "grad_norm": 1.1397403478622437, + "learning_rate": 0.00010963220572740656, + "loss": 0.6803, + "step": 27970 + }, + { + "epoch": 1.9409478870307404, + "grad_norm": 1.7656004428863525, + "learning_rate": 0.00010962661322205968, + "loss": 0.7233, + "step": 27971 + }, + { + "epoch": 1.9410172784678372, + "grad_norm": 1.3792319297790527, + "learning_rate": 0.00010962102068632269, + "loss": 0.6662, + "step": 27972 + }, + { + "epoch": 1.9410866699049336, + "grad_norm": 1.6507583856582642, + "learning_rate": 0.00010961542812021323, + "loss": 0.5556, + "step": 27973 + }, + { + "epoch": 1.9411560613420304, + "grad_norm": 1.4670066833496094, + "learning_rate": 0.00010960983552374904, + "loss": 0.4294, + "step": 27974 + }, + { + "epoch": 1.941225452779127, + "grad_norm": 1.9112545251846313, + "learning_rate": 0.00010960424289694766, + "loss": 0.7086, + "step": 27975 + }, + { + "epoch": 1.9412948442162237, + "grad_norm": 1.5943586826324463, + "learning_rate": 0.00010959865023982683, + "loss": 0.5086, + "step": 27976 + }, + { + "epoch": 1.9413642356533205, + "grad_norm": 1.5920151472091675, + "learning_rate": 0.00010959305755240415, + "loss": 0.561, + "step": 27977 + }, + { + "epoch": 1.941433627090417, + "grad_norm": 1.7114795446395874, + "learning_rate": 0.00010958746483469731, + "loss": 0.5936, + "step": 27978 + }, + { + "epoch": 1.9415030185275137, + "grad_norm": 1.6736013889312744, + "learning_rate": 0.00010958187208672397, + "loss": 0.6128, + "step": 27979 + }, + { + "epoch": 1.9415724099646103, + "grad_norm": 1.6029341220855713, + "learning_rate": 0.00010957627930850177, + "loss": 0.5336, + "step": 27980 + }, + { + "epoch": 1.941641801401707, + "grad_norm": 1.5825501680374146, + "learning_rate": 0.00010957068650004833, + "loss": 0.5621, + "step": 27981 + }, + { + "epoch": 1.9417111928388038, + "grad_norm": 1.4989114999771118, + "learning_rate": 0.00010956509366138137, + "loss": 0.6916, + "step": 27982 + }, + { + "epoch": 1.9417805842759004, + "grad_norm": 1.499913215637207, + "learning_rate": 0.00010955950079251854, + "loss": 0.6756, + "step": 27983 + }, + { + "epoch": 1.941849975712997, + "grad_norm": 1.6743676662445068, + "learning_rate": 0.00010955390789347744, + "loss": 0.6285, + "step": 27984 + }, + { + "epoch": 1.9419193671500938, + "grad_norm": 1.1922581195831299, + "learning_rate": 0.00010954831496427576, + "loss": 0.5229, + "step": 27985 + }, + { + "epoch": 1.9419887585871902, + "grad_norm": 1.850078821182251, + "learning_rate": 0.00010954272200493119, + "loss": 0.5918, + "step": 27986 + }, + { + "epoch": 1.942058150024287, + "grad_norm": 1.292722463607788, + "learning_rate": 0.00010953712901546134, + "loss": 0.3938, + "step": 27987 + }, + { + "epoch": 1.9421275414613837, + "grad_norm": 1.2005506753921509, + "learning_rate": 0.00010953153599588386, + "loss": 0.4292, + "step": 27988 + }, + { + "epoch": 1.9421969328984803, + "grad_norm": 1.6468462944030762, + "learning_rate": 0.00010952594294621645, + "loss": 0.4897, + "step": 27989 + }, + { + "epoch": 1.942266324335577, + "grad_norm": 1.3521885871887207, + "learning_rate": 0.00010952034986647673, + "loss": 0.5794, + "step": 27990 + }, + { + "epoch": 1.9423357157726735, + "grad_norm": 1.8321008682250977, + "learning_rate": 0.00010951475675668239, + "loss": 0.7608, + "step": 27991 + }, + { + "epoch": 1.9424051072097703, + "grad_norm": 1.248361349105835, + "learning_rate": 0.00010950916361685104, + "loss": 0.631, + "step": 27992 + }, + { + "epoch": 1.942474498646867, + "grad_norm": 1.0474990606307983, + "learning_rate": 0.0001095035704470004, + "loss": 0.623, + "step": 27993 + }, + { + "epoch": 1.9425438900839636, + "grad_norm": 1.509479284286499, + "learning_rate": 0.00010949797724714806, + "loss": 0.6, + "step": 27994 + }, + { + "epoch": 1.9426132815210604, + "grad_norm": 1.6718113422393799, + "learning_rate": 0.0001094923840173117, + "loss": 0.4254, + "step": 27995 + }, + { + "epoch": 1.942682672958157, + "grad_norm": 1.4849858283996582, + "learning_rate": 0.00010948679075750902, + "loss": 0.6597, + "step": 27996 + }, + { + "epoch": 1.9427520643952536, + "grad_norm": 1.7243481874465942, + "learning_rate": 0.00010948119746775762, + "loss": 0.5058, + "step": 27997 + }, + { + "epoch": 1.9428214558323504, + "grad_norm": 1.9055345058441162, + "learning_rate": 0.00010947560414807519, + "loss": 0.7791, + "step": 27998 + }, + { + "epoch": 1.9428908472694468, + "grad_norm": 1.1570682525634766, + "learning_rate": 0.00010947001079847933, + "loss": 0.4029, + "step": 27999 + }, + { + "epoch": 1.9429602387065437, + "grad_norm": 1.1404385566711426, + "learning_rate": 0.00010946441741898781, + "loss": 0.3996, + "step": 28000 + }, + { + "epoch": 1.9429602387065437, + "eval_loss": 0.5921171307563782, + "eval_runtime": 3855.7664, + "eval_samples_per_second": 6.295, + "eval_steps_per_second": 0.787, + "step": 28000 + }, + { + "epoch": 1.9430296301436403, + "grad_norm": 1.3631230592727661, + "learning_rate": 0.0001094588240096182, + "loss": 0.504, + "step": 28001 + }, + { + "epoch": 1.9430990215807369, + "grad_norm": 1.4287221431732178, + "learning_rate": 0.00010945323057038816, + "loss": 0.5408, + "step": 28002 + }, + { + "epoch": 1.9431684130178337, + "grad_norm": 2.378075361251831, + "learning_rate": 0.0001094476371013154, + "loss": 0.5682, + "step": 28003 + }, + { + "epoch": 1.94323780445493, + "grad_norm": 1.8568707704544067, + "learning_rate": 0.00010944204360241752, + "loss": 0.7074, + "step": 28004 + }, + { + "epoch": 1.943307195892027, + "grad_norm": 1.3430918455123901, + "learning_rate": 0.00010943645007371221, + "loss": 0.765, + "step": 28005 + }, + { + "epoch": 1.9433765873291236, + "grad_norm": 1.751417875289917, + "learning_rate": 0.00010943085651521712, + "loss": 0.63, + "step": 28006 + }, + { + "epoch": 1.9434459787662202, + "grad_norm": 1.2263739109039307, + "learning_rate": 0.00010942526292694992, + "loss": 0.5503, + "step": 28007 + }, + { + "epoch": 1.943515370203317, + "grad_norm": 2.0567541122436523, + "learning_rate": 0.00010941966930892828, + "loss": 0.6867, + "step": 28008 + }, + { + "epoch": 1.9435847616404136, + "grad_norm": 1.3508069515228271, + "learning_rate": 0.00010941407566116981, + "loss": 0.4831, + "step": 28009 + }, + { + "epoch": 1.9436541530775102, + "grad_norm": 2.0746994018554688, + "learning_rate": 0.0001094084819836922, + "loss": 0.6026, + "step": 28010 + }, + { + "epoch": 1.943723544514607, + "grad_norm": 2.0072031021118164, + "learning_rate": 0.0001094028882765131, + "loss": 0.7462, + "step": 28011 + }, + { + "epoch": 1.9437929359517034, + "grad_norm": 1.2113724946975708, + "learning_rate": 0.00010939729453965015, + "loss": 0.4015, + "step": 28012 + }, + { + "epoch": 1.9438623273888003, + "grad_norm": 1.3829330205917358, + "learning_rate": 0.00010939170077312106, + "loss": 0.507, + "step": 28013 + }, + { + "epoch": 1.9439317188258969, + "grad_norm": 1.8308939933776855, + "learning_rate": 0.00010938610697694347, + "loss": 0.6567, + "step": 28014 + }, + { + "epoch": 1.9440011102629935, + "grad_norm": 1.939402461051941, + "learning_rate": 0.000109380513151135, + "loss": 0.6241, + "step": 28015 + }, + { + "epoch": 1.9440705017000903, + "grad_norm": 1.9025636911392212, + "learning_rate": 0.00010937491929571335, + "loss": 0.7508, + "step": 28016 + }, + { + "epoch": 1.9441398931371867, + "grad_norm": 2.3942177295684814, + "learning_rate": 0.00010936932541069618, + "loss": 0.4996, + "step": 28017 + }, + { + "epoch": 1.9442092845742835, + "grad_norm": 2.291302442550659, + "learning_rate": 0.00010936373149610114, + "loss": 0.6182, + "step": 28018 + }, + { + "epoch": 1.9442786760113802, + "grad_norm": 1.4062474966049194, + "learning_rate": 0.00010935813755194586, + "loss": 0.6953, + "step": 28019 + }, + { + "epoch": 1.9443480674484768, + "grad_norm": 1.8895550966262817, + "learning_rate": 0.00010935254357824803, + "loss": 0.7183, + "step": 28020 + }, + { + "epoch": 1.9444174588855736, + "grad_norm": 1.2958850860595703, + "learning_rate": 0.00010934694957502532, + "loss": 0.3846, + "step": 28021 + }, + { + "epoch": 1.9444868503226702, + "grad_norm": 2.011106491088867, + "learning_rate": 0.00010934135554229536, + "loss": 0.8002, + "step": 28022 + }, + { + "epoch": 1.9445562417597668, + "grad_norm": 1.105746865272522, + "learning_rate": 0.00010933576148007582, + "loss": 0.5481, + "step": 28023 + }, + { + "epoch": 1.9446256331968637, + "grad_norm": 1.3108407258987427, + "learning_rate": 0.0001093301673883844, + "loss": 0.5631, + "step": 28024 + }, + { + "epoch": 1.94469502463396, + "grad_norm": 2.2328317165374756, + "learning_rate": 0.00010932457326723868, + "loss": 0.723, + "step": 28025 + }, + { + "epoch": 1.9447644160710569, + "grad_norm": 1.2992900609970093, + "learning_rate": 0.0001093189791166564, + "loss": 0.4578, + "step": 28026 + }, + { + "epoch": 1.9448338075081535, + "grad_norm": 1.1641665697097778, + "learning_rate": 0.00010931338493665517, + "loss": 0.3734, + "step": 28027 + }, + { + "epoch": 1.94490319894525, + "grad_norm": 1.8382900953292847, + "learning_rate": 0.00010930779072725268, + "loss": 0.5636, + "step": 28028 + }, + { + "epoch": 1.944972590382347, + "grad_norm": 1.2545562982559204, + "learning_rate": 0.00010930219648846656, + "loss": 0.5724, + "step": 28029 + }, + { + "epoch": 1.9450419818194433, + "grad_norm": 1.7736889123916626, + "learning_rate": 0.00010929660222031446, + "loss": 0.7398, + "step": 28030 + }, + { + "epoch": 1.9451113732565402, + "grad_norm": 1.9293783903121948, + "learning_rate": 0.0001092910079228141, + "loss": 0.695, + "step": 28031 + }, + { + "epoch": 1.9451807646936368, + "grad_norm": 1.4599907398223877, + "learning_rate": 0.0001092854135959831, + "loss": 0.5493, + "step": 28032 + }, + { + "epoch": 1.9452501561307334, + "grad_norm": 1.8079062700271606, + "learning_rate": 0.0001092798192398391, + "loss": 0.8243, + "step": 28033 + }, + { + "epoch": 1.9453195475678302, + "grad_norm": 1.534628987312317, + "learning_rate": 0.00010927422485439983, + "loss": 0.6618, + "step": 28034 + }, + { + "epoch": 1.9453889390049268, + "grad_norm": 1.895742416381836, + "learning_rate": 0.00010926863043968288, + "loss": 0.6749, + "step": 28035 + }, + { + "epoch": 1.9454583304420234, + "grad_norm": 1.5635515451431274, + "learning_rate": 0.00010926303599570593, + "loss": 0.575, + "step": 28036 + }, + { + "epoch": 1.9455277218791203, + "grad_norm": 1.3528642654418945, + "learning_rate": 0.00010925744152248666, + "loss": 0.442, + "step": 28037 + }, + { + "epoch": 1.9455971133162167, + "grad_norm": 1.2516343593597412, + "learning_rate": 0.00010925184702004274, + "loss": 0.5164, + "step": 28038 + }, + { + "epoch": 1.9456665047533135, + "grad_norm": 2.091270923614502, + "learning_rate": 0.0001092462524883918, + "loss": 0.6901, + "step": 28039 + }, + { + "epoch": 1.94573589619041, + "grad_norm": 1.5487955808639526, + "learning_rate": 0.00010924065792755149, + "loss": 0.3668, + "step": 28040 + }, + { + "epoch": 1.9458052876275067, + "grad_norm": 1.394322156906128, + "learning_rate": 0.00010923506333753954, + "loss": 0.5272, + "step": 28041 + }, + { + "epoch": 1.9458746790646035, + "grad_norm": 1.3800804615020752, + "learning_rate": 0.00010922946871837354, + "loss": 0.3946, + "step": 28042 + }, + { + "epoch": 1.9459440705017, + "grad_norm": 1.964178442955017, + "learning_rate": 0.0001092238740700712, + "loss": 0.6112, + "step": 28043 + }, + { + "epoch": 1.9460134619387968, + "grad_norm": 1.5529899597167969, + "learning_rate": 0.00010921827939265013, + "loss": 0.7103, + "step": 28044 + }, + { + "epoch": 1.9460828533758934, + "grad_norm": 1.8282058238983154, + "learning_rate": 0.00010921268468612804, + "loss": 0.5119, + "step": 28045 + }, + { + "epoch": 1.94615224481299, + "grad_norm": 1.2550204992294312, + "learning_rate": 0.00010920708995052256, + "loss": 0.3993, + "step": 28046 + }, + { + "epoch": 1.9462216362500868, + "grad_norm": 1.2559722661972046, + "learning_rate": 0.00010920149518585136, + "loss": 0.527, + "step": 28047 + }, + { + "epoch": 1.9462910276871834, + "grad_norm": 2.1370861530303955, + "learning_rate": 0.00010919590039213215, + "loss": 0.8674, + "step": 28048 + }, + { + "epoch": 1.94636041912428, + "grad_norm": 1.2717175483703613, + "learning_rate": 0.0001091903055693825, + "loss": 0.3109, + "step": 28049 + }, + { + "epoch": 1.9464298105613769, + "grad_norm": 1.625821590423584, + "learning_rate": 0.00010918471071762014, + "loss": 0.613, + "step": 28050 + }, + { + "epoch": 1.9464992019984733, + "grad_norm": 1.8938970565795898, + "learning_rate": 0.0001091791158368627, + "loss": 0.6398, + "step": 28051 + }, + { + "epoch": 1.94656859343557, + "grad_norm": 1.7993361949920654, + "learning_rate": 0.0001091735209271279, + "loss": 0.885, + "step": 28052 + }, + { + "epoch": 1.9466379848726667, + "grad_norm": 1.6936614513397217, + "learning_rate": 0.0001091679259884333, + "loss": 0.5427, + "step": 28053 + }, + { + "epoch": 1.9467073763097633, + "grad_norm": 1.3714302778244019, + "learning_rate": 0.00010916233102079665, + "loss": 0.5629, + "step": 28054 + }, + { + "epoch": 1.9467767677468601, + "grad_norm": 1.8912838697433472, + "learning_rate": 0.0001091567360242356, + "loss": 0.5921, + "step": 28055 + }, + { + "epoch": 1.9468461591839565, + "grad_norm": 1.3612165451049805, + "learning_rate": 0.00010915114099876778, + "loss": 0.5493, + "step": 28056 + }, + { + "epoch": 1.9469155506210534, + "grad_norm": 1.4060428142547607, + "learning_rate": 0.00010914554594441087, + "loss": 0.5394, + "step": 28057 + }, + { + "epoch": 1.94698494205815, + "grad_norm": 1.234294056892395, + "learning_rate": 0.00010913995086118252, + "loss": 0.435, + "step": 28058 + }, + { + "epoch": 1.9470543334952466, + "grad_norm": 1.6955978870391846, + "learning_rate": 0.00010913435574910043, + "loss": 0.4588, + "step": 28059 + }, + { + "epoch": 1.9471237249323434, + "grad_norm": 1.675599217414856, + "learning_rate": 0.00010912876060818222, + "loss": 0.6078, + "step": 28060 + }, + { + "epoch": 1.94719311636944, + "grad_norm": 1.6018478870391846, + "learning_rate": 0.00010912316543844558, + "loss": 0.5428, + "step": 28061 + }, + { + "epoch": 1.9472625078065366, + "grad_norm": 1.9364145994186401, + "learning_rate": 0.00010911757023990818, + "loss": 0.6493, + "step": 28062 + }, + { + "epoch": 1.9473318992436335, + "grad_norm": 2.1311192512512207, + "learning_rate": 0.00010911197501258765, + "loss": 0.551, + "step": 28063 + }, + { + "epoch": 1.9474012906807299, + "grad_norm": 1.6626170873641968, + "learning_rate": 0.00010910637975650166, + "loss": 0.654, + "step": 28064 + }, + { + "epoch": 1.9474706821178267, + "grad_norm": 2.401779890060425, + "learning_rate": 0.00010910078447166792, + "loss": 0.5221, + "step": 28065 + }, + { + "epoch": 1.9475400735549233, + "grad_norm": 1.5420349836349487, + "learning_rate": 0.00010909518915810404, + "loss": 0.6015, + "step": 28066 + }, + { + "epoch": 1.94760946499202, + "grad_norm": 1.8297609090805054, + "learning_rate": 0.00010908959381582771, + "loss": 0.5002, + "step": 28067 + }, + { + "epoch": 1.9476788564291168, + "grad_norm": 2.2926478385925293, + "learning_rate": 0.00010908399844485655, + "loss": 0.5807, + "step": 28068 + }, + { + "epoch": 1.9477482478662131, + "grad_norm": 1.4602439403533936, + "learning_rate": 0.00010907840304520831, + "loss": 0.5842, + "step": 28069 + }, + { + "epoch": 1.94781763930331, + "grad_norm": 5.756504058837891, + "learning_rate": 0.00010907280761690058, + "loss": 0.6763, + "step": 28070 + }, + { + "epoch": 1.9478870307404066, + "grad_norm": 1.8355686664581299, + "learning_rate": 0.00010906721215995106, + "loss": 0.6165, + "step": 28071 + }, + { + "epoch": 1.9479564221775032, + "grad_norm": 1.8875259160995483, + "learning_rate": 0.0001090616166743774, + "loss": 0.622, + "step": 28072 + }, + { + "epoch": 1.9480258136146, + "grad_norm": 1.560139775276184, + "learning_rate": 0.00010905602116019726, + "loss": 0.5884, + "step": 28073 + }, + { + "epoch": 1.9480952050516966, + "grad_norm": 1.5173835754394531, + "learning_rate": 0.00010905042561742831, + "loss": 0.661, + "step": 28074 + }, + { + "epoch": 1.9481645964887933, + "grad_norm": 1.0180283784866333, + "learning_rate": 0.00010904483004608821, + "loss": 0.3286, + "step": 28075 + }, + { + "epoch": 1.94823398792589, + "grad_norm": 1.8529562950134277, + "learning_rate": 0.00010903923444619466, + "loss": 0.5861, + "step": 28076 + }, + { + "epoch": 1.9483033793629865, + "grad_norm": 1.4288954734802246, + "learning_rate": 0.00010903363881776527, + "loss": 0.567, + "step": 28077 + }, + { + "epoch": 1.9483727708000833, + "grad_norm": 2.005200147628784, + "learning_rate": 0.00010902804316081776, + "loss": 0.5128, + "step": 28078 + }, + { + "epoch": 1.94844216223718, + "grad_norm": 1.222337245941162, + "learning_rate": 0.00010902244747536973, + "loss": 0.4567, + "step": 28079 + }, + { + "epoch": 1.9485115536742765, + "grad_norm": 1.8538297414779663, + "learning_rate": 0.00010901685176143892, + "loss": 0.6617, + "step": 28080 + }, + { + "epoch": 1.9485809451113734, + "grad_norm": 1.7933595180511475, + "learning_rate": 0.00010901125601904292, + "loss": 0.5144, + "step": 28081 + }, + { + "epoch": 1.9486503365484698, + "grad_norm": 1.5259932279586792, + "learning_rate": 0.00010900566024819945, + "loss": 0.3852, + "step": 28082 + }, + { + "epoch": 1.9487197279855666, + "grad_norm": 2.4774794578552246, + "learning_rate": 0.00010900006444892614, + "loss": 0.7844, + "step": 28083 + }, + { + "epoch": 1.9487891194226632, + "grad_norm": 1.2383805513381958, + "learning_rate": 0.00010899446862124067, + "loss": 0.5275, + "step": 28084 + }, + { + "epoch": 1.9488585108597598, + "grad_norm": 1.567516803741455, + "learning_rate": 0.00010898887276516073, + "loss": 0.6279, + "step": 28085 + }, + { + "epoch": 1.9489279022968566, + "grad_norm": 1.350079894065857, + "learning_rate": 0.00010898327688070395, + "loss": 0.6153, + "step": 28086 + }, + { + "epoch": 1.9489972937339533, + "grad_norm": 2.2834813594818115, + "learning_rate": 0.00010897768096788799, + "loss": 0.6899, + "step": 28087 + }, + { + "epoch": 1.9490666851710499, + "grad_norm": 1.5999958515167236, + "learning_rate": 0.00010897208502673053, + "loss": 0.4641, + "step": 28088 + }, + { + "epoch": 1.9491360766081467, + "grad_norm": 2.221696138381958, + "learning_rate": 0.00010896648905724925, + "loss": 0.8018, + "step": 28089 + }, + { + "epoch": 1.949205468045243, + "grad_norm": 1.3776462078094482, + "learning_rate": 0.00010896089305946182, + "loss": 0.3789, + "step": 28090 + }, + { + "epoch": 1.94927485948234, + "grad_norm": 1.4756748676300049, + "learning_rate": 0.00010895529703338587, + "loss": 0.4873, + "step": 28091 + }, + { + "epoch": 1.9493442509194365, + "grad_norm": 1.8130173683166504, + "learning_rate": 0.0001089497009790391, + "loss": 0.6455, + "step": 28092 + }, + { + "epoch": 1.9494136423565331, + "grad_norm": 1.2186509370803833, + "learning_rate": 0.00010894410489643916, + "loss": 0.5745, + "step": 28093 + }, + { + "epoch": 1.94948303379363, + "grad_norm": 1.6888729333877563, + "learning_rate": 0.00010893850878560371, + "loss": 0.3395, + "step": 28094 + }, + { + "epoch": 1.9495524252307266, + "grad_norm": 1.4763257503509521, + "learning_rate": 0.00010893291264655046, + "loss": 0.4568, + "step": 28095 + }, + { + "epoch": 1.9496218166678232, + "grad_norm": 1.9804182052612305, + "learning_rate": 0.00010892731647929699, + "loss": 0.7679, + "step": 28096 + }, + { + "epoch": 1.9496912081049198, + "grad_norm": 1.8380206823349, + "learning_rate": 0.00010892172028386105, + "loss": 0.5957, + "step": 28097 + }, + { + "epoch": 1.9497605995420164, + "grad_norm": 1.334096074104309, + "learning_rate": 0.00010891612406026026, + "loss": 0.5875, + "step": 28098 + }, + { + "epoch": 1.9498299909791132, + "grad_norm": 1.2703744173049927, + "learning_rate": 0.0001089105278085123, + "loss": 0.5471, + "step": 28099 + }, + { + "epoch": 1.9498993824162099, + "grad_norm": 1.4206620454788208, + "learning_rate": 0.00010890493152863485, + "loss": 0.5434, + "step": 28100 + }, + { + "epoch": 1.9499687738533065, + "grad_norm": 1.4868557453155518, + "learning_rate": 0.00010889933522064554, + "loss": 0.5374, + "step": 28101 + }, + { + "epoch": 1.9500381652904033, + "grad_norm": 1.4703952074050903, + "learning_rate": 0.00010889373888456208, + "loss": 0.4997, + "step": 28102 + }, + { + "epoch": 1.9501075567274997, + "grad_norm": 1.8122026920318604, + "learning_rate": 0.00010888814252040214, + "loss": 0.5237, + "step": 28103 + }, + { + "epoch": 1.9501769481645965, + "grad_norm": 1.9636995792388916, + "learning_rate": 0.00010888254612818332, + "loss": 0.6642, + "step": 28104 + }, + { + "epoch": 1.9502463396016931, + "grad_norm": 1.6827515363693237, + "learning_rate": 0.00010887694970792335, + "loss": 0.6228, + "step": 28105 + }, + { + "epoch": 1.9503157310387897, + "grad_norm": 1.7485301494598389, + "learning_rate": 0.00010887135325963989, + "loss": 0.6718, + "step": 28106 + }, + { + "epoch": 1.9503851224758866, + "grad_norm": 1.434821367263794, + "learning_rate": 0.00010886575678335058, + "loss": 0.5982, + "step": 28107 + }, + { + "epoch": 1.9504545139129832, + "grad_norm": 1.27090322971344, + "learning_rate": 0.0001088601602790731, + "loss": 0.3775, + "step": 28108 + }, + { + "epoch": 1.9505239053500798, + "grad_norm": 2.2015254497528076, + "learning_rate": 0.00010885456374682514, + "loss": 0.7558, + "step": 28109 + }, + { + "epoch": 1.9505932967871764, + "grad_norm": 1.5848397016525269, + "learning_rate": 0.00010884896718662437, + "loss": 0.4925, + "step": 28110 + }, + { + "epoch": 1.950662688224273, + "grad_norm": 1.9210124015808105, + "learning_rate": 0.0001088433705984884, + "loss": 0.5996, + "step": 28111 + }, + { + "epoch": 1.9507320796613699, + "grad_norm": 1.8027950525283813, + "learning_rate": 0.00010883777398243493, + "loss": 0.6481, + "step": 28112 + }, + { + "epoch": 1.9508014710984665, + "grad_norm": 1.718153476715088, + "learning_rate": 0.00010883217733848167, + "loss": 0.4555, + "step": 28113 + }, + { + "epoch": 1.950870862535563, + "grad_norm": 1.277629017829895, + "learning_rate": 0.00010882658066664624, + "loss": 0.4559, + "step": 28114 + }, + { + "epoch": 1.95094025397266, + "grad_norm": 1.5369116067886353, + "learning_rate": 0.00010882098396694631, + "loss": 0.7077, + "step": 28115 + }, + { + "epoch": 1.9510096454097563, + "grad_norm": 1.5202982425689697, + "learning_rate": 0.00010881538723939954, + "loss": 0.6424, + "step": 28116 + }, + { + "epoch": 1.9510790368468531, + "grad_norm": 1.1268088817596436, + "learning_rate": 0.00010880979048402366, + "loss": 0.4439, + "step": 28117 + }, + { + "epoch": 1.9511484282839497, + "grad_norm": 1.5635119676589966, + "learning_rate": 0.00010880419370083625, + "loss": 0.7523, + "step": 28118 + }, + { + "epoch": 1.9512178197210464, + "grad_norm": 1.2136297225952148, + "learning_rate": 0.00010879859688985506, + "loss": 0.4359, + "step": 28119 + }, + { + "epoch": 1.9512872111581432, + "grad_norm": 2.824223279953003, + "learning_rate": 0.00010879300005109768, + "loss": 0.7001, + "step": 28120 + }, + { + "epoch": 1.9513566025952398, + "grad_norm": 1.6707956790924072, + "learning_rate": 0.00010878740318458184, + "loss": 0.746, + "step": 28121 + }, + { + "epoch": 1.9514259940323364, + "grad_norm": 2.2287063598632812, + "learning_rate": 0.0001087818062903252, + "loss": 0.9456, + "step": 28122 + }, + { + "epoch": 1.951495385469433, + "grad_norm": 1.3450756072998047, + "learning_rate": 0.00010877620936834538, + "loss": 0.6345, + "step": 28123 + }, + { + "epoch": 1.9515647769065296, + "grad_norm": 1.4226303100585938, + "learning_rate": 0.00010877061241866014, + "loss": 0.5755, + "step": 28124 + }, + { + "epoch": 1.9516341683436265, + "grad_norm": 1.6883065700531006, + "learning_rate": 0.00010876501544128704, + "loss": 0.4391, + "step": 28125 + }, + { + "epoch": 1.951703559780723, + "grad_norm": 1.4020758867263794, + "learning_rate": 0.00010875941843624382, + "loss": 0.6451, + "step": 28126 + }, + { + "epoch": 1.9517729512178197, + "grad_norm": 1.096627116203308, + "learning_rate": 0.00010875382140354813, + "loss": 0.4915, + "step": 28127 + }, + { + "epoch": 1.9518423426549165, + "grad_norm": 1.2064038515090942, + "learning_rate": 0.00010874822434321765, + "loss": 0.5298, + "step": 28128 + }, + { + "epoch": 1.951911734092013, + "grad_norm": 1.6723474264144897, + "learning_rate": 0.00010874262725527005, + "loss": 0.7477, + "step": 28129 + }, + { + "epoch": 1.9519811255291097, + "grad_norm": 1.0002073049545288, + "learning_rate": 0.00010873703013972298, + "loss": 0.445, + "step": 28130 + }, + { + "epoch": 1.9520505169662064, + "grad_norm": 1.2908707857131958, + "learning_rate": 0.00010873143299659412, + "loss": 0.5176, + "step": 28131 + }, + { + "epoch": 1.952119908403303, + "grad_norm": 1.446280598640442, + "learning_rate": 0.00010872583582590114, + "loss": 0.4594, + "step": 28132 + }, + { + "epoch": 1.9521892998403998, + "grad_norm": 1.418805718421936, + "learning_rate": 0.0001087202386276617, + "loss": 0.4436, + "step": 28133 + }, + { + "epoch": 1.9522586912774964, + "grad_norm": 1.3938632011413574, + "learning_rate": 0.00010871464140189349, + "loss": 0.8096, + "step": 28134 + }, + { + "epoch": 1.952328082714593, + "grad_norm": 1.880682110786438, + "learning_rate": 0.00010870904414861419, + "loss": 0.5391, + "step": 28135 + }, + { + "epoch": 1.9523974741516896, + "grad_norm": 1.4631147384643555, + "learning_rate": 0.0001087034468678414, + "loss": 0.8392, + "step": 28136 + }, + { + "epoch": 1.9524668655887862, + "grad_norm": 1.4442075490951538, + "learning_rate": 0.00010869784955959286, + "loss": 0.6286, + "step": 28137 + }, + { + "epoch": 1.952536257025883, + "grad_norm": 1.316939353942871, + "learning_rate": 0.00010869225222388621, + "loss": 0.4172, + "step": 28138 + }, + { + "epoch": 1.9526056484629797, + "grad_norm": 2.4551429748535156, + "learning_rate": 0.00010868665486073914, + "loss": 0.4462, + "step": 28139 + }, + { + "epoch": 1.9526750399000763, + "grad_norm": 1.4343596696853638, + "learning_rate": 0.0001086810574701693, + "loss": 0.5675, + "step": 28140 + }, + { + "epoch": 1.9527444313371731, + "grad_norm": 0.8451719284057617, + "learning_rate": 0.00010867546005219438, + "loss": 0.3577, + "step": 28141 + }, + { + "epoch": 1.9528138227742695, + "grad_norm": 1.510632872581482, + "learning_rate": 0.00010866986260683203, + "loss": 0.52, + "step": 28142 + }, + { + "epoch": 1.9528832142113663, + "grad_norm": 1.6202285289764404, + "learning_rate": 0.00010866426513409993, + "loss": 0.4864, + "step": 28143 + }, + { + "epoch": 1.952952605648463, + "grad_norm": 2.0242919921875, + "learning_rate": 0.00010865866763401573, + "loss": 0.5819, + "step": 28144 + }, + { + "epoch": 1.9530219970855596, + "grad_norm": 1.3512578010559082, + "learning_rate": 0.00010865307010659715, + "loss": 0.4972, + "step": 28145 + }, + { + "epoch": 1.9530913885226564, + "grad_norm": 1.6848148107528687, + "learning_rate": 0.00010864747255186182, + "loss": 0.5131, + "step": 28146 + }, + { + "epoch": 1.953160779959753, + "grad_norm": 1.453736662864685, + "learning_rate": 0.00010864187496982743, + "loss": 0.5398, + "step": 28147 + }, + { + "epoch": 1.9532301713968496, + "grad_norm": 1.9326502084732056, + "learning_rate": 0.00010863627736051165, + "loss": 0.4556, + "step": 28148 + }, + { + "epoch": 1.9532995628339465, + "grad_norm": 1.5563219785690308, + "learning_rate": 0.00010863067972393212, + "loss": 0.5, + "step": 28149 + }, + { + "epoch": 1.9533689542710428, + "grad_norm": 1.4090216159820557, + "learning_rate": 0.00010862508206010658, + "loss": 0.6655, + "step": 28150 + }, + { + "epoch": 1.9534383457081397, + "grad_norm": 1.4398819208145142, + "learning_rate": 0.00010861948436905261, + "loss": 0.5328, + "step": 28151 + }, + { + "epoch": 1.9535077371452363, + "grad_norm": 1.7474758625030518, + "learning_rate": 0.00010861388665078796, + "loss": 0.4864, + "step": 28152 + }, + { + "epoch": 1.953577128582333, + "grad_norm": 2.341188907623291, + "learning_rate": 0.00010860828890533023, + "loss": 0.6429, + "step": 28153 + }, + { + "epoch": 1.9536465200194297, + "grad_norm": 1.4741069078445435, + "learning_rate": 0.00010860269113269716, + "loss": 0.4649, + "step": 28154 + }, + { + "epoch": 1.9537159114565261, + "grad_norm": 1.5465764999389648, + "learning_rate": 0.0001085970933329064, + "loss": 0.5076, + "step": 28155 + }, + { + "epoch": 1.953785302893623, + "grad_norm": 1.1974120140075684, + "learning_rate": 0.00010859149550597561, + "loss": 0.3333, + "step": 28156 + }, + { + "epoch": 1.9538546943307196, + "grad_norm": 1.875087857246399, + "learning_rate": 0.00010858589765192244, + "loss": 0.5492, + "step": 28157 + }, + { + "epoch": 1.9539240857678162, + "grad_norm": 1.6425269842147827, + "learning_rate": 0.00010858029977076459, + "loss": 0.6159, + "step": 28158 + }, + { + "epoch": 1.953993477204913, + "grad_norm": 1.6627293825149536, + "learning_rate": 0.00010857470186251977, + "loss": 0.4232, + "step": 28159 + }, + { + "epoch": 1.9540628686420096, + "grad_norm": 1.560394525527954, + "learning_rate": 0.00010856910392720556, + "loss": 0.9363, + "step": 28160 + }, + { + "epoch": 1.9541322600791062, + "grad_norm": 1.5712107419967651, + "learning_rate": 0.00010856350596483971, + "loss": 0.4828, + "step": 28161 + }, + { + "epoch": 1.954201651516203, + "grad_norm": 1.1645270586013794, + "learning_rate": 0.00010855790797543988, + "loss": 0.3751, + "step": 28162 + }, + { + "epoch": 1.9542710429532995, + "grad_norm": 1.9787837266921997, + "learning_rate": 0.00010855230995902368, + "loss": 0.6697, + "step": 28163 + }, + { + "epoch": 1.9543404343903963, + "grad_norm": 1.5381394624710083, + "learning_rate": 0.00010854671191560886, + "loss": 0.5828, + "step": 28164 + }, + { + "epoch": 1.954409825827493, + "grad_norm": 1.400877594947815, + "learning_rate": 0.00010854111384521309, + "loss": 0.528, + "step": 28165 + }, + { + "epoch": 1.9544792172645895, + "grad_norm": 1.6621774435043335, + "learning_rate": 0.00010853551574785396, + "loss": 0.486, + "step": 28166 + }, + { + "epoch": 1.9545486087016863, + "grad_norm": 1.5742028951644897, + "learning_rate": 0.00010852991762354925, + "loss": 0.5673, + "step": 28167 + }, + { + "epoch": 1.9546180001387827, + "grad_norm": 1.5640709400177002, + "learning_rate": 0.00010852431947231654, + "loss": 0.4708, + "step": 28168 + }, + { + "epoch": 1.9546873915758796, + "grad_norm": 1.3863385915756226, + "learning_rate": 0.00010851872129417358, + "loss": 0.4319, + "step": 28169 + }, + { + "epoch": 1.9547567830129762, + "grad_norm": 2.5139286518096924, + "learning_rate": 0.00010851312308913795, + "loss": 0.681, + "step": 28170 + }, + { + "epoch": 1.9548261744500728, + "grad_norm": 1.280490756034851, + "learning_rate": 0.00010850752485722742, + "loss": 0.6349, + "step": 28171 + }, + { + "epoch": 1.9548955658871696, + "grad_norm": 2.34256911277771, + "learning_rate": 0.00010850192659845959, + "loss": 0.6753, + "step": 28172 + }, + { + "epoch": 1.9549649573242662, + "grad_norm": 2.112637758255005, + "learning_rate": 0.00010849632831285219, + "loss": 0.6219, + "step": 28173 + }, + { + "epoch": 1.9550343487613628, + "grad_norm": 1.721814513206482, + "learning_rate": 0.00010849073000042286, + "loss": 0.6267, + "step": 28174 + }, + { + "epoch": 1.9551037401984597, + "grad_norm": 1.7066441774368286, + "learning_rate": 0.00010848513166118928, + "loss": 0.78, + "step": 28175 + }, + { + "epoch": 1.955173131635556, + "grad_norm": 1.0669301748275757, + "learning_rate": 0.00010847953329516914, + "loss": 0.3036, + "step": 28176 + }, + { + "epoch": 1.955242523072653, + "grad_norm": 1.606278657913208, + "learning_rate": 0.00010847393490238007, + "loss": 0.6513, + "step": 28177 + }, + { + "epoch": 1.9553119145097495, + "grad_norm": 1.3193283081054688, + "learning_rate": 0.00010846833648283977, + "loss": 0.5254, + "step": 28178 + }, + { + "epoch": 1.9553813059468461, + "grad_norm": 1.9564800262451172, + "learning_rate": 0.00010846273803656595, + "loss": 0.5519, + "step": 28179 + }, + { + "epoch": 1.955450697383943, + "grad_norm": 1.2364331483840942, + "learning_rate": 0.00010845713956357621, + "loss": 0.5155, + "step": 28180 + }, + { + "epoch": 1.9555200888210393, + "grad_norm": 2.082993507385254, + "learning_rate": 0.00010845154106388827, + "loss": 0.7027, + "step": 28181 + }, + { + "epoch": 1.9555894802581362, + "grad_norm": 2.844231605529785, + "learning_rate": 0.00010844594253751979, + "loss": 0.6181, + "step": 28182 + }, + { + "epoch": 1.9556588716952328, + "grad_norm": 0.7932032346725464, + "learning_rate": 0.00010844034398448847, + "loss": 0.2611, + "step": 28183 + }, + { + "epoch": 1.9557282631323294, + "grad_norm": 1.2593777179718018, + "learning_rate": 0.00010843474540481197, + "loss": 0.456, + "step": 28184 + }, + { + "epoch": 1.9557976545694262, + "grad_norm": 1.5184067487716675, + "learning_rate": 0.00010842914679850795, + "loss": 0.515, + "step": 28185 + }, + { + "epoch": 1.9558670460065228, + "grad_norm": 1.1935349702835083, + "learning_rate": 0.00010842354816559408, + "loss": 0.4446, + "step": 28186 + }, + { + "epoch": 1.9559364374436194, + "grad_norm": 1.7060861587524414, + "learning_rate": 0.00010841794950608806, + "loss": 0.6083, + "step": 28187 + }, + { + "epoch": 1.9560058288807163, + "grad_norm": 1.6432538032531738, + "learning_rate": 0.00010841235082000752, + "loss": 0.4927, + "step": 28188 + }, + { + "epoch": 1.9560752203178127, + "grad_norm": 1.7636590003967285, + "learning_rate": 0.00010840675210737016, + "loss": 0.486, + "step": 28189 + }, + { + "epoch": 1.9561446117549095, + "grad_norm": 1.8533345460891724, + "learning_rate": 0.00010840115336819372, + "loss": 0.7813, + "step": 28190 + }, + { + "epoch": 1.9562140031920061, + "grad_norm": 1.2809141874313354, + "learning_rate": 0.00010839555460249577, + "loss": 0.5821, + "step": 28191 + }, + { + "epoch": 1.9562833946291027, + "grad_norm": 1.4530457258224487, + "learning_rate": 0.00010838995581029401, + "loss": 0.7121, + "step": 28192 + }, + { + "epoch": 1.9563527860661996, + "grad_norm": 1.6537600755691528, + "learning_rate": 0.00010838435699160618, + "loss": 0.4327, + "step": 28193 + }, + { + "epoch": 1.956422177503296, + "grad_norm": 1.2493176460266113, + "learning_rate": 0.00010837875814644989, + "loss": 0.4867, + "step": 28194 + }, + { + "epoch": 1.9564915689403928, + "grad_norm": 1.68645441532135, + "learning_rate": 0.00010837315927484281, + "loss": 0.5676, + "step": 28195 + }, + { + "epoch": 1.9565609603774894, + "grad_norm": 1.7574634552001953, + "learning_rate": 0.00010836756037680265, + "loss": 0.4226, + "step": 28196 + }, + { + "epoch": 1.956630351814586, + "grad_norm": 2.2547104358673096, + "learning_rate": 0.0001083619614523471, + "loss": 0.5241, + "step": 28197 + }, + { + "epoch": 1.9566997432516828, + "grad_norm": 1.932137131690979, + "learning_rate": 0.00010835636250149377, + "loss": 0.5624, + "step": 28198 + }, + { + "epoch": 1.9567691346887794, + "grad_norm": 1.3518747091293335, + "learning_rate": 0.00010835076352426037, + "loss": 0.3815, + "step": 28199 + }, + { + "epoch": 1.956838526125876, + "grad_norm": 1.522734522819519, + "learning_rate": 0.00010834516452066461, + "loss": 0.5521, + "step": 28200 + }, + { + "epoch": 1.9569079175629729, + "grad_norm": 1.6547197103500366, + "learning_rate": 0.00010833956549072413, + "loss": 0.5421, + "step": 28201 + }, + { + "epoch": 1.9569773090000693, + "grad_norm": 2.4097237586975098, + "learning_rate": 0.00010833396643445661, + "loss": 0.6749, + "step": 28202 + }, + { + "epoch": 1.957046700437166, + "grad_norm": 1.7196025848388672, + "learning_rate": 0.0001083283673518797, + "loss": 0.5766, + "step": 28203 + }, + { + "epoch": 1.9571160918742627, + "grad_norm": 1.4737906455993652, + "learning_rate": 0.00010832276824301113, + "loss": 0.7206, + "step": 28204 + }, + { + "epoch": 1.9571854833113593, + "grad_norm": 2.0792088508605957, + "learning_rate": 0.00010831716910786854, + "loss": 0.6122, + "step": 28205 + }, + { + "epoch": 1.9572548747484562, + "grad_norm": 2.0215630531311035, + "learning_rate": 0.00010831156994646958, + "loss": 0.6472, + "step": 28206 + }, + { + "epoch": 1.9573242661855526, + "grad_norm": 2.7328102588653564, + "learning_rate": 0.000108305970758832, + "loss": 0.5043, + "step": 28207 + }, + { + "epoch": 1.9573936576226494, + "grad_norm": 1.7368746995925903, + "learning_rate": 0.00010830037154497341, + "loss": 0.5605, + "step": 28208 + }, + { + "epoch": 1.957463049059746, + "grad_norm": 1.6778316497802734, + "learning_rate": 0.00010829477230491151, + "loss": 0.4958, + "step": 28209 + }, + { + "epoch": 1.9575324404968426, + "grad_norm": 1.7559937238693237, + "learning_rate": 0.00010828917303866397, + "loss": 0.7444, + "step": 28210 + }, + { + "epoch": 1.9576018319339394, + "grad_norm": 1.3775782585144043, + "learning_rate": 0.00010828357374624852, + "loss": 0.562, + "step": 28211 + }, + { + "epoch": 1.957671223371036, + "grad_norm": 1.3420565128326416, + "learning_rate": 0.00010827797442768273, + "loss": 0.5286, + "step": 28212 + }, + { + "epoch": 1.9577406148081327, + "grad_norm": 2.0952823162078857, + "learning_rate": 0.00010827237508298436, + "loss": 0.6102, + "step": 28213 + }, + { + "epoch": 1.9578100062452295, + "grad_norm": 1.7760664224624634, + "learning_rate": 0.00010826677571217108, + "loss": 0.6037, + "step": 28214 + }, + { + "epoch": 1.9578793976823259, + "grad_norm": 2.0984420776367188, + "learning_rate": 0.00010826117631526051, + "loss": 0.6501, + "step": 28215 + }, + { + "epoch": 1.9579487891194227, + "grad_norm": 1.1589618921279907, + "learning_rate": 0.00010825557689227038, + "loss": 0.4883, + "step": 28216 + }, + { + "epoch": 1.9580181805565193, + "grad_norm": 1.6561919450759888, + "learning_rate": 0.00010824997744321836, + "loss": 0.4453, + "step": 28217 + }, + { + "epoch": 1.958087571993616, + "grad_norm": 2.0169780254364014, + "learning_rate": 0.00010824437796812211, + "loss": 0.5506, + "step": 28218 + }, + { + "epoch": 1.9581569634307128, + "grad_norm": 1.4109562635421753, + "learning_rate": 0.00010823877846699932, + "loss": 0.5137, + "step": 28219 + }, + { + "epoch": 1.9582263548678092, + "grad_norm": 1.2850905656814575, + "learning_rate": 0.00010823317893986766, + "loss": 0.4846, + "step": 28220 + }, + { + "epoch": 1.958295746304906, + "grad_norm": 1.1160298585891724, + "learning_rate": 0.00010822757938674483, + "loss": 0.4486, + "step": 28221 + }, + { + "epoch": 1.9583651377420026, + "grad_norm": 1.7401418685913086, + "learning_rate": 0.00010822197980764846, + "loss": 0.5046, + "step": 28222 + }, + { + "epoch": 1.9584345291790992, + "grad_norm": 1.4929907321929932, + "learning_rate": 0.00010821638020259626, + "loss": 0.4939, + "step": 28223 + }, + { + "epoch": 1.958503920616196, + "grad_norm": 1.5835100412368774, + "learning_rate": 0.00010821078057160592, + "loss": 0.6354, + "step": 28224 + }, + { + "epoch": 1.9585733120532927, + "grad_norm": 2.2264790534973145, + "learning_rate": 0.00010820518091469507, + "loss": 0.4763, + "step": 28225 + }, + { + "epoch": 1.9586427034903893, + "grad_norm": 1.6317611932754517, + "learning_rate": 0.0001081995812318814, + "loss": 0.6086, + "step": 28226 + }, + { + "epoch": 1.958712094927486, + "grad_norm": 4.174581527709961, + "learning_rate": 0.00010819398152318264, + "loss": 0.604, + "step": 28227 + }, + { + "epoch": 1.9587814863645825, + "grad_norm": 1.4780558347702026, + "learning_rate": 0.00010818838178861643, + "loss": 0.5149, + "step": 28228 + }, + { + "epoch": 1.9588508778016793, + "grad_norm": 1.442467212677002, + "learning_rate": 0.00010818278202820042, + "loss": 0.7241, + "step": 28229 + }, + { + "epoch": 1.958920269238776, + "grad_norm": 1.2633284330368042, + "learning_rate": 0.00010817718224195233, + "loss": 0.4842, + "step": 28230 + }, + { + "epoch": 1.9589896606758725, + "grad_norm": 1.6444242000579834, + "learning_rate": 0.00010817158242988984, + "loss": 0.5979, + "step": 28231 + }, + { + "epoch": 1.9590590521129694, + "grad_norm": 1.4081323146820068, + "learning_rate": 0.00010816598259203061, + "loss": 0.4794, + "step": 28232 + }, + { + "epoch": 1.9591284435500658, + "grad_norm": 1.705851435661316, + "learning_rate": 0.00010816038272839231, + "loss": 0.5784, + "step": 28233 + }, + { + "epoch": 1.9591978349871626, + "grad_norm": 1.7196773290634155, + "learning_rate": 0.00010815478283899261, + "loss": 0.5536, + "step": 28234 + }, + { + "epoch": 1.9592672264242592, + "grad_norm": 2.9738636016845703, + "learning_rate": 0.00010814918292384923, + "loss": 0.5982, + "step": 28235 + }, + { + "epoch": 1.9593366178613558, + "grad_norm": 1.409872055053711, + "learning_rate": 0.00010814358298297981, + "loss": 0.4752, + "step": 28236 + }, + { + "epoch": 1.9594060092984527, + "grad_norm": 1.9036694765090942, + "learning_rate": 0.00010813798301640208, + "loss": 0.4557, + "step": 28237 + }, + { + "epoch": 1.9594754007355493, + "grad_norm": 1.8892487287521362, + "learning_rate": 0.00010813238302413369, + "loss": 0.6226, + "step": 28238 + }, + { + "epoch": 1.9595447921726459, + "grad_norm": 1.4406676292419434, + "learning_rate": 0.00010812678300619226, + "loss": 0.4648, + "step": 28239 + }, + { + "epoch": 1.9596141836097427, + "grad_norm": 1.9730498790740967, + "learning_rate": 0.00010812118296259553, + "loss": 0.7078, + "step": 28240 + }, + { + "epoch": 1.959683575046839, + "grad_norm": 1.322622537612915, + "learning_rate": 0.00010811558289336117, + "loss": 0.4764, + "step": 28241 + }, + { + "epoch": 1.959752966483936, + "grad_norm": 1.5038955211639404, + "learning_rate": 0.00010810998279850689, + "loss": 0.602, + "step": 28242 + }, + { + "epoch": 1.9598223579210325, + "grad_norm": 1.44461190700531, + "learning_rate": 0.00010810438267805028, + "loss": 0.6123, + "step": 28243 + }, + { + "epoch": 1.9598917493581292, + "grad_norm": 1.5541198253631592, + "learning_rate": 0.00010809878253200909, + "loss": 0.4401, + "step": 28244 + }, + { + "epoch": 1.959961140795226, + "grad_norm": 1.7941230535507202, + "learning_rate": 0.000108093182360401, + "loss": 0.5805, + "step": 28245 + }, + { + "epoch": 1.9600305322323224, + "grad_norm": 1.5310593843460083, + "learning_rate": 0.00010808758216324366, + "loss": 0.5143, + "step": 28246 + }, + { + "epoch": 1.9600999236694192, + "grad_norm": 1.3715167045593262, + "learning_rate": 0.00010808198194055477, + "loss": 0.5508, + "step": 28247 + }, + { + "epoch": 1.9601693151065158, + "grad_norm": 1.4674830436706543, + "learning_rate": 0.00010807638169235199, + "loss": 0.431, + "step": 28248 + }, + { + "epoch": 1.9602387065436124, + "grad_norm": 1.5802594423294067, + "learning_rate": 0.00010807078141865304, + "loss": 0.8884, + "step": 28249 + }, + { + "epoch": 1.9603080979807093, + "grad_norm": 1.3519139289855957, + "learning_rate": 0.00010806518111947553, + "loss": 0.4404, + "step": 28250 + }, + { + "epoch": 1.9603774894178059, + "grad_norm": 1.3880435228347778, + "learning_rate": 0.0001080595807948372, + "loss": 0.772, + "step": 28251 + }, + { + "epoch": 1.9604468808549025, + "grad_norm": 1.6238081455230713, + "learning_rate": 0.00010805398044475571, + "loss": 0.3528, + "step": 28252 + }, + { + "epoch": 1.9605162722919993, + "grad_norm": 1.169903039932251, + "learning_rate": 0.00010804838006924873, + "loss": 0.4391, + "step": 28253 + }, + { + "epoch": 1.9605856637290957, + "grad_norm": 2.0444109439849854, + "learning_rate": 0.00010804277966833396, + "loss": 0.6386, + "step": 28254 + }, + { + "epoch": 1.9606550551661925, + "grad_norm": 2.469194173812866, + "learning_rate": 0.00010803717924202905, + "loss": 0.6333, + "step": 28255 + }, + { + "epoch": 1.9607244466032892, + "grad_norm": 2.1196672916412354, + "learning_rate": 0.00010803157879035173, + "loss": 0.6602, + "step": 28256 + }, + { + "epoch": 1.9607938380403858, + "grad_norm": 1.342544674873352, + "learning_rate": 0.00010802597831331959, + "loss": 0.4615, + "step": 28257 + }, + { + "epoch": 1.9608632294774826, + "grad_norm": 2.1001689434051514, + "learning_rate": 0.0001080203778109504, + "loss": 0.4299, + "step": 28258 + }, + { + "epoch": 1.960932620914579, + "grad_norm": 1.5338600873947144, + "learning_rate": 0.00010801477728326184, + "loss": 0.4684, + "step": 28259 + }, + { + "epoch": 1.9610020123516758, + "grad_norm": 1.7866283655166626, + "learning_rate": 0.0001080091767302715, + "loss": 0.7778, + "step": 28260 + }, + { + "epoch": 1.9610714037887724, + "grad_norm": 2.351966381072998, + "learning_rate": 0.00010800357615199717, + "loss": 0.7095, + "step": 28261 + }, + { + "epoch": 1.961140795225869, + "grad_norm": 1.1443969011306763, + "learning_rate": 0.00010799797554845646, + "loss": 0.5215, + "step": 28262 + }, + { + "epoch": 1.9612101866629659, + "grad_norm": 1.6892739534378052, + "learning_rate": 0.00010799237491966705, + "loss": 0.494, + "step": 28263 + }, + { + "epoch": 1.9612795781000625, + "grad_norm": 1.9152395725250244, + "learning_rate": 0.00010798677426564665, + "loss": 0.4592, + "step": 28264 + }, + { + "epoch": 1.961348969537159, + "grad_norm": 1.7860801219940186, + "learning_rate": 0.00010798117358641291, + "loss": 0.6356, + "step": 28265 + }, + { + "epoch": 1.961418360974256, + "grad_norm": 2.0837740898132324, + "learning_rate": 0.00010797557288198359, + "loss": 0.5205, + "step": 28266 + }, + { + "epoch": 1.9614877524113523, + "grad_norm": 1.5384517908096313, + "learning_rate": 0.00010796997215237625, + "loss": 0.4939, + "step": 28267 + }, + { + "epoch": 1.9615571438484491, + "grad_norm": 1.461395502090454, + "learning_rate": 0.00010796437139760867, + "loss": 0.5963, + "step": 28268 + }, + { + "epoch": 1.9616265352855458, + "grad_norm": 1.849226713180542, + "learning_rate": 0.00010795877061769848, + "loss": 0.6548, + "step": 28269 + }, + { + "epoch": 1.9616959267226424, + "grad_norm": 1.1361066102981567, + "learning_rate": 0.00010795316981266336, + "loss": 0.3453, + "step": 28270 + }, + { + "epoch": 1.9617653181597392, + "grad_norm": 1.3783551454544067, + "learning_rate": 0.00010794756898252103, + "loss": 0.687, + "step": 28271 + }, + { + "epoch": 1.9618347095968356, + "grad_norm": 1.86902916431427, + "learning_rate": 0.00010794196812728917, + "loss": 0.745, + "step": 28272 + }, + { + "epoch": 1.9619041010339324, + "grad_norm": 1.2621300220489502, + "learning_rate": 0.00010793636724698542, + "loss": 0.4943, + "step": 28273 + }, + { + "epoch": 1.961973492471029, + "grad_norm": 1.3500893115997314, + "learning_rate": 0.00010793076634162744, + "loss": 0.5132, + "step": 28274 + }, + { + "epoch": 1.9620428839081256, + "grad_norm": 1.25644850730896, + "learning_rate": 0.00010792516541123296, + "loss": 0.4424, + "step": 28275 + }, + { + "epoch": 1.9621122753452225, + "grad_norm": 1.5464612245559692, + "learning_rate": 0.0001079195644558197, + "loss": 0.4505, + "step": 28276 + }, + { + "epoch": 1.962181666782319, + "grad_norm": 1.7909921407699585, + "learning_rate": 0.00010791396347540526, + "loss": 0.8113, + "step": 28277 + }, + { + "epoch": 1.9622510582194157, + "grad_norm": 1.31807541847229, + "learning_rate": 0.00010790836247000735, + "loss": 0.5394, + "step": 28278 + }, + { + "epoch": 1.9623204496565125, + "grad_norm": 1.4491386413574219, + "learning_rate": 0.00010790276143964366, + "loss": 0.5141, + "step": 28279 + }, + { + "epoch": 1.962389841093609, + "grad_norm": 1.4937680959701538, + "learning_rate": 0.00010789716038433189, + "loss": 0.5832, + "step": 28280 + }, + { + "epoch": 1.9624592325307058, + "grad_norm": 1.2678736448287964, + "learning_rate": 0.00010789155930408967, + "loss": 0.4806, + "step": 28281 + }, + { + "epoch": 1.9625286239678024, + "grad_norm": 1.489017367362976, + "learning_rate": 0.00010788595819893472, + "loss": 0.4625, + "step": 28282 + }, + { + "epoch": 1.962598015404899, + "grad_norm": 1.9464551210403442, + "learning_rate": 0.00010788035706888476, + "loss": 0.5229, + "step": 28283 + }, + { + "epoch": 1.9626674068419958, + "grad_norm": 1.95199453830719, + "learning_rate": 0.00010787475591395737, + "loss": 0.5935, + "step": 28284 + }, + { + "epoch": 1.9627367982790922, + "grad_norm": 1.4430947303771973, + "learning_rate": 0.0001078691547341703, + "loss": 0.6511, + "step": 28285 + }, + { + "epoch": 1.962806189716189, + "grad_norm": 1.319189190864563, + "learning_rate": 0.00010786355352954123, + "loss": 0.5452, + "step": 28286 + }, + { + "epoch": 1.9628755811532856, + "grad_norm": 1.543495535850525, + "learning_rate": 0.00010785795230008783, + "loss": 0.5639, + "step": 28287 + }, + { + "epoch": 1.9629449725903823, + "grad_norm": 1.0490190982818604, + "learning_rate": 0.0001078523510458278, + "loss": 0.5355, + "step": 28288 + }, + { + "epoch": 1.963014364027479, + "grad_norm": 1.5416946411132812, + "learning_rate": 0.00010784674976677882, + "loss": 0.4926, + "step": 28289 + }, + { + "epoch": 1.9630837554645757, + "grad_norm": 1.7771308422088623, + "learning_rate": 0.00010784114846295852, + "loss": 0.5869, + "step": 28290 + }, + { + "epoch": 1.9631531469016723, + "grad_norm": 1.4840692281723022, + "learning_rate": 0.00010783554713438466, + "loss": 0.6568, + "step": 28291 + }, + { + "epoch": 1.9632225383387691, + "grad_norm": 1.4207122325897217, + "learning_rate": 0.00010782994578107488, + "loss": 0.3392, + "step": 28292 + }, + { + "epoch": 1.9632919297758655, + "grad_norm": 1.6067309379577637, + "learning_rate": 0.00010782434440304687, + "loss": 0.5471, + "step": 28293 + }, + { + "epoch": 1.9633613212129624, + "grad_norm": 1.6874604225158691, + "learning_rate": 0.00010781874300031828, + "loss": 0.5872, + "step": 28294 + }, + { + "epoch": 1.963430712650059, + "grad_norm": 2.083203077316284, + "learning_rate": 0.00010781314157290685, + "loss": 0.5471, + "step": 28295 + }, + { + "epoch": 1.9635001040871556, + "grad_norm": 1.9306777715682983, + "learning_rate": 0.00010780754012083026, + "loss": 0.3861, + "step": 28296 + }, + { + "epoch": 1.9635694955242524, + "grad_norm": 1.287301778793335, + "learning_rate": 0.00010780193864410615, + "loss": 0.4297, + "step": 28297 + }, + { + "epoch": 1.9636388869613488, + "grad_norm": 1.3889455795288086, + "learning_rate": 0.00010779633714275222, + "loss": 0.4244, + "step": 28298 + }, + { + "epoch": 1.9637082783984456, + "grad_norm": 1.5434929132461548, + "learning_rate": 0.00010779073561678614, + "loss": 0.3766, + "step": 28299 + }, + { + "epoch": 1.9637776698355422, + "grad_norm": 1.8343783617019653, + "learning_rate": 0.00010778513406622566, + "loss": 0.6491, + "step": 28300 + }, + { + "epoch": 1.9638470612726389, + "grad_norm": 1.7880549430847168, + "learning_rate": 0.00010777953249108839, + "loss": 0.56, + "step": 28301 + }, + { + "epoch": 1.9639164527097357, + "grad_norm": 1.9907573461532593, + "learning_rate": 0.00010777393089139203, + "loss": 0.6862, + "step": 28302 + }, + { + "epoch": 1.9639858441468323, + "grad_norm": 1.1856966018676758, + "learning_rate": 0.00010776832926715428, + "loss": 0.3987, + "step": 28303 + }, + { + "epoch": 1.964055235583929, + "grad_norm": 1.4272620677947998, + "learning_rate": 0.00010776272761839281, + "loss": 0.4601, + "step": 28304 + }, + { + "epoch": 1.9641246270210257, + "grad_norm": 1.6991512775421143, + "learning_rate": 0.00010775712594512532, + "loss": 0.4379, + "step": 28305 + }, + { + "epoch": 1.9641940184581221, + "grad_norm": 1.540060043334961, + "learning_rate": 0.00010775152424736949, + "loss": 0.5888, + "step": 28306 + }, + { + "epoch": 1.964263409895219, + "grad_norm": 1.346970796585083, + "learning_rate": 0.00010774592252514297, + "loss": 0.4708, + "step": 28307 + }, + { + "epoch": 1.9643328013323156, + "grad_norm": 1.833644151687622, + "learning_rate": 0.00010774032077846352, + "loss": 0.5379, + "step": 28308 + }, + { + "epoch": 1.9644021927694122, + "grad_norm": 2.201946973800659, + "learning_rate": 0.00010773471900734873, + "loss": 0.6407, + "step": 28309 + }, + { + "epoch": 1.964471584206509, + "grad_norm": 1.7145028114318848, + "learning_rate": 0.00010772911721181633, + "loss": 0.8194, + "step": 28310 + }, + { + "epoch": 1.9645409756436054, + "grad_norm": 1.7422035932540894, + "learning_rate": 0.00010772351539188404, + "loss": 0.4668, + "step": 28311 + }, + { + "epoch": 1.9646103670807022, + "grad_norm": 1.457556962966919, + "learning_rate": 0.00010771791354756949, + "loss": 0.6503, + "step": 28312 + }, + { + "epoch": 1.9646797585177989, + "grad_norm": 1.5668503046035767, + "learning_rate": 0.00010771231167889035, + "loss": 0.5609, + "step": 28313 + }, + { + "epoch": 1.9647491499548955, + "grad_norm": 1.4150892496109009, + "learning_rate": 0.00010770670978586439, + "loss": 0.5592, + "step": 28314 + }, + { + "epoch": 1.9648185413919923, + "grad_norm": 1.4649852514266968, + "learning_rate": 0.0001077011078685092, + "loss": 0.5521, + "step": 28315 + }, + { + "epoch": 1.964887932829089, + "grad_norm": 1.3669160604476929, + "learning_rate": 0.00010769550592684253, + "loss": 0.6637, + "step": 28316 + }, + { + "epoch": 1.9649573242661855, + "grad_norm": 1.8258647918701172, + "learning_rate": 0.00010768990396088201, + "loss": 0.564, + "step": 28317 + }, + { + "epoch": 1.9650267157032824, + "grad_norm": 1.9737271070480347, + "learning_rate": 0.00010768430197064538, + "loss": 0.5791, + "step": 28318 + }, + { + "epoch": 1.9650961071403787, + "grad_norm": 1.2412179708480835, + "learning_rate": 0.0001076786999561503, + "loss": 0.4246, + "step": 28319 + }, + { + "epoch": 1.9651654985774756, + "grad_norm": 1.8789657354354858, + "learning_rate": 0.00010767309791741445, + "loss": 0.6783, + "step": 28320 + }, + { + "epoch": 1.9652348900145722, + "grad_norm": 1.376668095588684, + "learning_rate": 0.00010766749585445554, + "loss": 0.6731, + "step": 28321 + }, + { + "epoch": 1.9653042814516688, + "grad_norm": 1.6721453666687012, + "learning_rate": 0.00010766189376729121, + "loss": 0.5758, + "step": 28322 + }, + { + "epoch": 1.9653736728887656, + "grad_norm": 1.5374621152877808, + "learning_rate": 0.00010765629165593916, + "loss": 0.589, + "step": 28323 + }, + { + "epoch": 1.9654430643258622, + "grad_norm": 1.1399987936019897, + "learning_rate": 0.00010765068952041712, + "loss": 0.5654, + "step": 28324 + }, + { + "epoch": 1.9655124557629589, + "grad_norm": 0.9253231883049011, + "learning_rate": 0.00010764508736074272, + "loss": 0.3596, + "step": 28325 + }, + { + "epoch": 1.9655818472000555, + "grad_norm": 1.3429752588272095, + "learning_rate": 0.0001076394851769337, + "loss": 0.6499, + "step": 28326 + }, + { + "epoch": 1.965651238637152, + "grad_norm": 1.6405234336853027, + "learning_rate": 0.00010763388296900767, + "loss": 0.6813, + "step": 28327 + }, + { + "epoch": 1.965720630074249, + "grad_norm": 1.7231594324111938, + "learning_rate": 0.00010762828073698237, + "loss": 0.7754, + "step": 28328 + }, + { + "epoch": 1.9657900215113455, + "grad_norm": 1.8380028009414673, + "learning_rate": 0.00010762267848087547, + "loss": 0.602, + "step": 28329 + }, + { + "epoch": 1.9658594129484421, + "grad_norm": 1.5926593542099, + "learning_rate": 0.00010761707620070465, + "loss": 0.5415, + "step": 28330 + }, + { + "epoch": 1.965928804385539, + "grad_norm": 1.4211686849594116, + "learning_rate": 0.00010761147389648763, + "loss": 0.5221, + "step": 28331 + }, + { + "epoch": 1.9659981958226354, + "grad_norm": 1.4908205270767212, + "learning_rate": 0.00010760587156824204, + "loss": 0.7459, + "step": 28332 + }, + { + "epoch": 1.9660675872597322, + "grad_norm": 1.8893423080444336, + "learning_rate": 0.0001076002692159856, + "loss": 0.6403, + "step": 28333 + }, + { + "epoch": 1.9661369786968288, + "grad_norm": 1.5385438203811646, + "learning_rate": 0.000107594666839736, + "loss": 0.4835, + "step": 28334 + }, + { + "epoch": 1.9662063701339254, + "grad_norm": 1.5426452159881592, + "learning_rate": 0.00010758906443951095, + "loss": 0.5201, + "step": 28335 + }, + { + "epoch": 1.9662757615710222, + "grad_norm": 1.369040608406067, + "learning_rate": 0.00010758346201532807, + "loss": 0.657, + "step": 28336 + }, + { + "epoch": 1.9663451530081189, + "grad_norm": 1.379473328590393, + "learning_rate": 0.00010757785956720507, + "loss": 0.5469, + "step": 28337 + }, + { + "epoch": 1.9664145444452155, + "grad_norm": 1.4516453742980957, + "learning_rate": 0.00010757225709515967, + "loss": 0.6459, + "step": 28338 + }, + { + "epoch": 1.966483935882312, + "grad_norm": 1.3426774740219116, + "learning_rate": 0.00010756665459920952, + "loss": 0.4426, + "step": 28339 + }, + { + "epoch": 1.9665533273194087, + "grad_norm": 1.4410138130187988, + "learning_rate": 0.00010756105207937233, + "loss": 0.5459, + "step": 28340 + }, + { + "epoch": 1.9666227187565055, + "grad_norm": 0.9658097624778748, + "learning_rate": 0.00010755544953566578, + "loss": 0.5187, + "step": 28341 + }, + { + "epoch": 1.9666921101936021, + "grad_norm": 1.587548017501831, + "learning_rate": 0.00010754984696810753, + "loss": 0.4606, + "step": 28342 + }, + { + "epoch": 1.9667615016306987, + "grad_norm": 1.5539844036102295, + "learning_rate": 0.00010754424437671531, + "loss": 0.4856, + "step": 28343 + }, + { + "epoch": 1.9668308930677956, + "grad_norm": 1.1935830116271973, + "learning_rate": 0.00010753864176150677, + "loss": 0.5177, + "step": 28344 + }, + { + "epoch": 1.966900284504892, + "grad_norm": 1.9745818376541138, + "learning_rate": 0.00010753303912249963, + "loss": 0.5394, + "step": 28345 + }, + { + "epoch": 1.9669696759419888, + "grad_norm": 1.4828729629516602, + "learning_rate": 0.00010752743645971156, + "loss": 0.4098, + "step": 28346 + }, + { + "epoch": 1.9670390673790854, + "grad_norm": 1.5436819791793823, + "learning_rate": 0.0001075218337731602, + "loss": 0.622, + "step": 28347 + }, + { + "epoch": 1.967108458816182, + "grad_norm": 1.3239827156066895, + "learning_rate": 0.00010751623106286332, + "loss": 0.5707, + "step": 28348 + }, + { + "epoch": 1.9671778502532788, + "grad_norm": 1.3689857721328735, + "learning_rate": 0.0001075106283288386, + "loss": 0.6104, + "step": 28349 + }, + { + "epoch": 1.9672472416903755, + "grad_norm": 2.090991258621216, + "learning_rate": 0.00010750502557110365, + "loss": 0.5544, + "step": 28350 + }, + { + "epoch": 1.967316633127472, + "grad_norm": 1.5540720224380493, + "learning_rate": 0.00010749942278967622, + "loss": 0.6958, + "step": 28351 + }, + { + "epoch": 1.9673860245645687, + "grad_norm": 1.5815495252609253, + "learning_rate": 0.000107493819984574, + "loss": 0.5012, + "step": 28352 + }, + { + "epoch": 1.9674554160016653, + "grad_norm": 1.6129196882247925, + "learning_rate": 0.00010748821715581464, + "loss": 0.4946, + "step": 28353 + }, + { + "epoch": 1.9675248074387621, + "grad_norm": 1.5161526203155518, + "learning_rate": 0.00010748261430341585, + "loss": 0.613, + "step": 28354 + }, + { + "epoch": 1.9675941988758587, + "grad_norm": 1.831059455871582, + "learning_rate": 0.0001074770114273953, + "loss": 0.6583, + "step": 28355 + }, + { + "epoch": 1.9676635903129553, + "grad_norm": 1.9488781690597534, + "learning_rate": 0.00010747140852777071, + "loss": 0.6005, + "step": 28356 + }, + { + "epoch": 1.9677329817500522, + "grad_norm": 1.1935415267944336, + "learning_rate": 0.00010746580560455975, + "loss": 0.6869, + "step": 28357 + }, + { + "epoch": 1.9678023731871486, + "grad_norm": 1.8311163187026978, + "learning_rate": 0.0001074602026577801, + "loss": 0.5876, + "step": 28358 + }, + { + "epoch": 1.9678717646242454, + "grad_norm": 1.1026033163070679, + "learning_rate": 0.00010745459968744948, + "loss": 0.3835, + "step": 28359 + }, + { + "epoch": 1.967941156061342, + "grad_norm": 1.3435083627700806, + "learning_rate": 0.00010744899669358555, + "loss": 0.4502, + "step": 28360 + }, + { + "epoch": 1.9680105474984386, + "grad_norm": 1.9254714250564575, + "learning_rate": 0.00010744339367620597, + "loss": 0.7809, + "step": 28361 + }, + { + "epoch": 1.9680799389355355, + "grad_norm": 1.8301762342453003, + "learning_rate": 0.00010743779063532848, + "loss": 0.6827, + "step": 28362 + }, + { + "epoch": 1.968149330372632, + "grad_norm": 1.7918726205825806, + "learning_rate": 0.00010743218757097076, + "loss": 0.8144, + "step": 28363 + }, + { + "epoch": 1.9682187218097287, + "grad_norm": 1.9351106882095337, + "learning_rate": 0.00010742658448315046, + "loss": 0.6664, + "step": 28364 + }, + { + "epoch": 1.9682881132468253, + "grad_norm": 1.9263367652893066, + "learning_rate": 0.0001074209813718853, + "loss": 0.5664, + "step": 28365 + }, + { + "epoch": 1.968357504683922, + "grad_norm": 1.3346848487854004, + "learning_rate": 0.00010741537823719299, + "loss": 0.3867, + "step": 28366 + }, + { + "epoch": 1.9684268961210187, + "grad_norm": 1.4846339225769043, + "learning_rate": 0.00010740977507909116, + "loss": 0.5676, + "step": 28367 + }, + { + "epoch": 1.9684962875581153, + "grad_norm": 2.1239490509033203, + "learning_rate": 0.00010740417189759753, + "loss": 0.5562, + "step": 28368 + }, + { + "epoch": 1.968565678995212, + "grad_norm": 1.9992822408676147, + "learning_rate": 0.00010739856869272982, + "loss": 0.5734, + "step": 28369 + }, + { + "epoch": 1.9686350704323088, + "grad_norm": 1.298896312713623, + "learning_rate": 0.00010739296546450566, + "loss": 0.3302, + "step": 28370 + }, + { + "epoch": 1.9687044618694052, + "grad_norm": 1.7646784782409668, + "learning_rate": 0.00010738736221294277, + "loss": 0.5004, + "step": 28371 + }, + { + "epoch": 1.968773853306502, + "grad_norm": 1.327444314956665, + "learning_rate": 0.00010738175893805882, + "loss": 0.4793, + "step": 28372 + }, + { + "epoch": 1.9688432447435986, + "grad_norm": 1.475546956062317, + "learning_rate": 0.00010737615563987154, + "loss": 0.6126, + "step": 28373 + }, + { + "epoch": 1.9689126361806952, + "grad_norm": 1.1581594944000244, + "learning_rate": 0.00010737055231839856, + "loss": 0.5936, + "step": 28374 + }, + { + "epoch": 1.968982027617792, + "grad_norm": 1.0081573724746704, + "learning_rate": 0.00010736494897365763, + "loss": 0.462, + "step": 28375 + }, + { + "epoch": 1.9690514190548887, + "grad_norm": 2.12658953666687, + "learning_rate": 0.0001073593456056664, + "loss": 0.6123, + "step": 28376 + }, + { + "epoch": 1.9691208104919853, + "grad_norm": 1.6585966348648071, + "learning_rate": 0.00010735374221444259, + "loss": 0.6804, + "step": 28377 + }, + { + "epoch": 1.969190201929082, + "grad_norm": 1.7091028690338135, + "learning_rate": 0.00010734813880000386, + "loss": 0.5643, + "step": 28378 + }, + { + "epoch": 1.9692595933661785, + "grad_norm": 1.4029004573822021, + "learning_rate": 0.00010734253536236788, + "loss": 0.668, + "step": 28379 + }, + { + "epoch": 1.9693289848032753, + "grad_norm": 1.1474714279174805, + "learning_rate": 0.0001073369319015524, + "loss": 0.4881, + "step": 28380 + }, + { + "epoch": 1.969398376240372, + "grad_norm": 1.5655908584594727, + "learning_rate": 0.00010733132841757507, + "loss": 0.5488, + "step": 28381 + }, + { + "epoch": 1.9694677676774686, + "grad_norm": 1.2050200700759888, + "learning_rate": 0.00010732572491045358, + "loss": 0.5269, + "step": 28382 + }, + { + "epoch": 1.9695371591145654, + "grad_norm": 1.6002026796340942, + "learning_rate": 0.00010732012138020565, + "loss": 0.5586, + "step": 28383 + }, + { + "epoch": 1.9696065505516618, + "grad_norm": 1.653245449066162, + "learning_rate": 0.00010731451782684891, + "loss": 0.5223, + "step": 28384 + }, + { + "epoch": 1.9696759419887586, + "grad_norm": 1.8369836807250977, + "learning_rate": 0.00010730891425040108, + "loss": 0.6961, + "step": 28385 + }, + { + "epoch": 1.9697453334258552, + "grad_norm": 2.796971559524536, + "learning_rate": 0.00010730331065087988, + "loss": 0.587, + "step": 28386 + }, + { + "epoch": 1.9698147248629518, + "grad_norm": 1.361938714981079, + "learning_rate": 0.000107297707028303, + "loss": 0.6298, + "step": 28387 + }, + { + "epoch": 1.9698841163000487, + "grad_norm": 1.557465672492981, + "learning_rate": 0.00010729210338268807, + "loss": 0.5663, + "step": 28388 + }, + { + "epoch": 1.9699535077371453, + "grad_norm": 1.5124696493148804, + "learning_rate": 0.0001072864997140528, + "loss": 0.6979, + "step": 28389 + }, + { + "epoch": 1.970022899174242, + "grad_norm": 1.8565808534622192, + "learning_rate": 0.00010728089602241493, + "loss": 0.518, + "step": 28390 + }, + { + "epoch": 1.9700922906113387, + "grad_norm": 1.089015007019043, + "learning_rate": 0.00010727529230779211, + "loss": 0.538, + "step": 28391 + }, + { + "epoch": 1.9701616820484351, + "grad_norm": 1.9287054538726807, + "learning_rate": 0.00010726968857020203, + "loss": 0.5264, + "step": 28392 + }, + { + "epoch": 1.970231073485532, + "grad_norm": 1.2252808809280396, + "learning_rate": 0.0001072640848096624, + "loss": 0.4939, + "step": 28393 + }, + { + "epoch": 1.9703004649226286, + "grad_norm": 1.3816677331924438, + "learning_rate": 0.00010725848102619087, + "loss": 0.4892, + "step": 28394 + }, + { + "epoch": 1.9703698563597252, + "grad_norm": 1.1327805519104004, + "learning_rate": 0.00010725287721980519, + "loss": 0.5188, + "step": 28395 + }, + { + "epoch": 1.970439247796822, + "grad_norm": 1.317704200744629, + "learning_rate": 0.00010724727339052299, + "loss": 0.4977, + "step": 28396 + }, + { + "epoch": 1.9705086392339184, + "grad_norm": 1.8930894136428833, + "learning_rate": 0.00010724166953836201, + "loss": 0.5439, + "step": 28397 + }, + { + "epoch": 1.9705780306710152, + "grad_norm": 1.0244097709655762, + "learning_rate": 0.0001072360656633399, + "loss": 0.4753, + "step": 28398 + }, + { + "epoch": 1.9706474221081118, + "grad_norm": 1.187177062034607, + "learning_rate": 0.00010723046176547438, + "loss": 0.4741, + "step": 28399 + }, + { + "epoch": 1.9707168135452084, + "grad_norm": 1.3898983001708984, + "learning_rate": 0.00010722485784478314, + "loss": 0.5112, + "step": 28400 + }, + { + "epoch": 1.9707862049823053, + "grad_norm": 1.579363226890564, + "learning_rate": 0.00010721925390128385, + "loss": 0.482, + "step": 28401 + }, + { + "epoch": 1.970855596419402, + "grad_norm": 1.5865147113800049, + "learning_rate": 0.00010721364993499422, + "loss": 0.5609, + "step": 28402 + }, + { + "epoch": 1.9709249878564985, + "grad_norm": 1.5419788360595703, + "learning_rate": 0.00010720804594593192, + "loss": 0.5436, + "step": 28403 + }, + { + "epoch": 1.9709943792935953, + "grad_norm": 1.9083000421524048, + "learning_rate": 0.00010720244193411466, + "loss": 0.772, + "step": 28404 + }, + { + "epoch": 1.9710637707306917, + "grad_norm": 1.7969012260437012, + "learning_rate": 0.00010719683789956014, + "loss": 0.6687, + "step": 28405 + }, + { + "epoch": 1.9711331621677886, + "grad_norm": 1.5043201446533203, + "learning_rate": 0.00010719123384228603, + "loss": 0.6373, + "step": 28406 + }, + { + "epoch": 1.9712025536048852, + "grad_norm": 1.7273980379104614, + "learning_rate": 0.00010718562976231003, + "loss": 0.4355, + "step": 28407 + }, + { + "epoch": 1.9712719450419818, + "grad_norm": 2.7085988521575928, + "learning_rate": 0.00010718002565964982, + "loss": 0.81, + "step": 28408 + }, + { + "epoch": 1.9713413364790786, + "grad_norm": 1.3190001249313354, + "learning_rate": 0.00010717442153432311, + "loss": 0.5771, + "step": 28409 + }, + { + "epoch": 1.971410727916175, + "grad_norm": 1.2028452157974243, + "learning_rate": 0.00010716881738634756, + "loss": 0.508, + "step": 28410 + }, + { + "epoch": 1.9714801193532718, + "grad_norm": 1.5834306478500366, + "learning_rate": 0.00010716321321574094, + "loss": 0.5533, + "step": 28411 + }, + { + "epoch": 1.9715495107903684, + "grad_norm": 1.0353870391845703, + "learning_rate": 0.00010715760902252083, + "loss": 0.4098, + "step": 28412 + }, + { + "epoch": 1.971618902227465, + "grad_norm": 1.2200604677200317, + "learning_rate": 0.00010715200480670504, + "loss": 0.3968, + "step": 28413 + }, + { + "epoch": 1.9716882936645619, + "grad_norm": 1.977462887763977, + "learning_rate": 0.00010714640056831118, + "loss": 0.4401, + "step": 28414 + }, + { + "epoch": 1.9717576851016585, + "grad_norm": 1.5147180557250977, + "learning_rate": 0.00010714079630735693, + "loss": 0.5823, + "step": 28415 + }, + { + "epoch": 1.971827076538755, + "grad_norm": 1.7660942077636719, + "learning_rate": 0.00010713519202386003, + "loss": 0.6269, + "step": 28416 + }, + { + "epoch": 1.971896467975852, + "grad_norm": 1.2548012733459473, + "learning_rate": 0.00010712958771783814, + "loss": 0.5851, + "step": 28417 + }, + { + "epoch": 1.9719658594129483, + "grad_norm": 1.5604678392410278, + "learning_rate": 0.00010712398338930903, + "loss": 0.3811, + "step": 28418 + }, + { + "epoch": 1.9720352508500452, + "grad_norm": 1.3302191495895386, + "learning_rate": 0.00010711837903829027, + "loss": 0.5066, + "step": 28419 + }, + { + "epoch": 1.9721046422871418, + "grad_norm": 1.2661056518554688, + "learning_rate": 0.00010711277466479963, + "loss": 0.4147, + "step": 28420 + }, + { + "epoch": 1.9721740337242384, + "grad_norm": 1.705749750137329, + "learning_rate": 0.00010710717026885479, + "loss": 0.7079, + "step": 28421 + }, + { + "epoch": 1.9722434251613352, + "grad_norm": 1.740889072418213, + "learning_rate": 0.00010710156585047344, + "loss": 0.529, + "step": 28422 + }, + { + "epoch": 1.9723128165984316, + "grad_norm": 1.886319875717163, + "learning_rate": 0.00010709596140967326, + "loss": 0.5621, + "step": 28423 + }, + { + "epoch": 1.9723822080355284, + "grad_norm": 1.8227121829986572, + "learning_rate": 0.00010709035694647194, + "loss": 0.726, + "step": 28424 + }, + { + "epoch": 1.972451599472625, + "grad_norm": 1.3116117715835571, + "learning_rate": 0.00010708475246088723, + "loss": 0.3842, + "step": 28425 + }, + { + "epoch": 1.9725209909097217, + "grad_norm": 1.3659569025039673, + "learning_rate": 0.00010707914795293674, + "loss": 0.4972, + "step": 28426 + }, + { + "epoch": 1.9725903823468185, + "grad_norm": 1.5013396739959717, + "learning_rate": 0.0001070735434226382, + "loss": 0.6189, + "step": 28427 + }, + { + "epoch": 1.972659773783915, + "grad_norm": 1.0450878143310547, + "learning_rate": 0.00010706793887000935, + "loss": 0.3482, + "step": 28428 + }, + { + "epoch": 1.9727291652210117, + "grad_norm": 0.9615880250930786, + "learning_rate": 0.0001070623342950678, + "loss": 0.3639, + "step": 28429 + }, + { + "epoch": 1.9727985566581085, + "grad_norm": 1.5144562721252441, + "learning_rate": 0.00010705672969783129, + "loss": 0.4668, + "step": 28430 + }, + { + "epoch": 1.972867948095205, + "grad_norm": 1.6207643747329712, + "learning_rate": 0.0001070511250783175, + "loss": 0.6183, + "step": 28431 + }, + { + "epoch": 1.9729373395323018, + "grad_norm": 1.644383192062378, + "learning_rate": 0.00010704552043654415, + "loss": 0.6072, + "step": 28432 + }, + { + "epoch": 1.9730067309693984, + "grad_norm": 1.1814590692520142, + "learning_rate": 0.00010703991577252886, + "loss": 0.4112, + "step": 28433 + }, + { + "epoch": 1.973076122406495, + "grad_norm": 1.2436033487319946, + "learning_rate": 0.00010703431108628938, + "loss": 0.5581, + "step": 28434 + }, + { + "epoch": 1.9731455138435918, + "grad_norm": 1.3600348234176636, + "learning_rate": 0.00010702870637784343, + "loss": 0.6748, + "step": 28435 + }, + { + "epoch": 1.9732149052806882, + "grad_norm": 1.512279987335205, + "learning_rate": 0.00010702310164720866, + "loss": 0.5462, + "step": 28436 + }, + { + "epoch": 1.973284296717785, + "grad_norm": 1.545222282409668, + "learning_rate": 0.00010701749689440275, + "loss": 0.6604, + "step": 28437 + }, + { + "epoch": 1.9733536881548817, + "grad_norm": 1.5719484090805054, + "learning_rate": 0.00010701189211944342, + "loss": 0.5674, + "step": 28438 + }, + { + "epoch": 1.9734230795919783, + "grad_norm": 1.544870376586914, + "learning_rate": 0.00010700628732234841, + "loss": 0.6127, + "step": 28439 + }, + { + "epoch": 1.973492471029075, + "grad_norm": 1.0893810987472534, + "learning_rate": 0.0001070006825031353, + "loss": 0.4562, + "step": 28440 + }, + { + "epoch": 1.9735618624661717, + "grad_norm": 1.643193244934082, + "learning_rate": 0.00010699507766182188, + "loss": 0.7313, + "step": 28441 + }, + { + "epoch": 1.9736312539032683, + "grad_norm": 1.6885390281677246, + "learning_rate": 0.00010698947279842582, + "loss": 0.4235, + "step": 28442 + }, + { + "epoch": 1.9737006453403652, + "grad_norm": 1.4808170795440674, + "learning_rate": 0.00010698386791296478, + "loss": 0.5617, + "step": 28443 + }, + { + "epoch": 1.9737700367774615, + "grad_norm": 1.109549880027771, + "learning_rate": 0.00010697826300545648, + "loss": 0.4704, + "step": 28444 + }, + { + "epoch": 1.9738394282145584, + "grad_norm": 1.546413779258728, + "learning_rate": 0.00010697265807591863, + "loss": 0.5079, + "step": 28445 + }, + { + "epoch": 1.973908819651655, + "grad_norm": 1.2553789615631104, + "learning_rate": 0.0001069670531243689, + "loss": 0.5528, + "step": 28446 + }, + { + "epoch": 1.9739782110887516, + "grad_norm": 0.9933515191078186, + "learning_rate": 0.00010696144815082498, + "loss": 0.3425, + "step": 28447 + }, + { + "epoch": 1.9740476025258484, + "grad_norm": 1.8627434968948364, + "learning_rate": 0.0001069558431553046, + "loss": 0.4249, + "step": 28448 + }, + { + "epoch": 1.9741169939629448, + "grad_norm": 1.727878212928772, + "learning_rate": 0.00010695023813782546, + "loss": 0.5762, + "step": 28449 + }, + { + "epoch": 1.9741863854000417, + "grad_norm": 1.3073413372039795, + "learning_rate": 0.00010694463309840517, + "loss": 0.5276, + "step": 28450 + }, + { + "epoch": 1.9742557768371383, + "grad_norm": 1.3974947929382324, + "learning_rate": 0.00010693902803706148, + "loss": 0.3944, + "step": 28451 + }, + { + "epoch": 1.9743251682742349, + "grad_norm": 2.001047372817993, + "learning_rate": 0.0001069334229538121, + "loss": 0.4374, + "step": 28452 + }, + { + "epoch": 1.9743945597113317, + "grad_norm": 1.7701557874679565, + "learning_rate": 0.00010692781784867469, + "loss": 0.6223, + "step": 28453 + }, + { + "epoch": 1.9744639511484283, + "grad_norm": 1.194091796875, + "learning_rate": 0.00010692221272166699, + "loss": 0.4905, + "step": 28454 + }, + { + "epoch": 1.974533342585525, + "grad_norm": 1.8900476694107056, + "learning_rate": 0.00010691660757280663, + "loss": 0.7416, + "step": 28455 + }, + { + "epoch": 1.9746027340226218, + "grad_norm": 1.7116122245788574, + "learning_rate": 0.00010691100240211138, + "loss": 0.5368, + "step": 28456 + }, + { + "epoch": 1.9746721254597182, + "grad_norm": 1.2966684103012085, + "learning_rate": 0.00010690539720959889, + "loss": 0.5202, + "step": 28457 + }, + { + "epoch": 1.974741516896815, + "grad_norm": 1.249450922012329, + "learning_rate": 0.00010689979199528684, + "loss": 0.4171, + "step": 28458 + }, + { + "epoch": 1.9748109083339116, + "grad_norm": 1.66975998878479, + "learning_rate": 0.00010689418675919299, + "loss": 0.6134, + "step": 28459 + }, + { + "epoch": 1.9748802997710082, + "grad_norm": 1.4028160572052002, + "learning_rate": 0.00010688858150133496, + "loss": 0.6407, + "step": 28460 + }, + { + "epoch": 1.974949691208105, + "grad_norm": 1.1557623147964478, + "learning_rate": 0.00010688297622173047, + "loss": 0.4643, + "step": 28461 + }, + { + "epoch": 1.9750190826452014, + "grad_norm": 1.6653156280517578, + "learning_rate": 0.00010687737092039722, + "loss": 0.5587, + "step": 28462 + }, + { + "epoch": 1.9750884740822983, + "grad_norm": 1.9254071712493896, + "learning_rate": 0.00010687176559735294, + "loss": 0.4485, + "step": 28463 + }, + { + "epoch": 1.9751578655193949, + "grad_norm": 2.7193851470947266, + "learning_rate": 0.00010686616025261527, + "loss": 0.764, + "step": 28464 + }, + { + "epoch": 1.9752272569564915, + "grad_norm": 2.448439598083496, + "learning_rate": 0.00010686055488620195, + "loss": 0.8196, + "step": 28465 + }, + { + "epoch": 1.9752966483935883, + "grad_norm": 1.8193674087524414, + "learning_rate": 0.00010685494949813064, + "loss": 0.6374, + "step": 28466 + }, + { + "epoch": 1.975366039830685, + "grad_norm": 1.9683868885040283, + "learning_rate": 0.00010684934408841907, + "loss": 0.7064, + "step": 28467 + }, + { + "epoch": 1.9754354312677815, + "grad_norm": 2.0318922996520996, + "learning_rate": 0.00010684373865708489, + "loss": 0.7304, + "step": 28468 + }, + { + "epoch": 1.9755048227048784, + "grad_norm": 1.3231992721557617, + "learning_rate": 0.0001068381332041458, + "loss": 0.484, + "step": 28469 + }, + { + "epoch": 1.9755742141419748, + "grad_norm": 1.613550066947937, + "learning_rate": 0.00010683252772961956, + "loss": 0.5644, + "step": 28470 + }, + { + "epoch": 1.9756436055790716, + "grad_norm": 1.1655248403549194, + "learning_rate": 0.0001068269222335238, + "loss": 0.4611, + "step": 28471 + }, + { + "epoch": 1.9757129970161682, + "grad_norm": 1.3960614204406738, + "learning_rate": 0.00010682131671587623, + "loss": 0.5258, + "step": 28472 + }, + { + "epoch": 1.9757823884532648, + "grad_norm": 1.4653394222259521, + "learning_rate": 0.0001068157111766946, + "loss": 0.521, + "step": 28473 + }, + { + "epoch": 1.9758517798903616, + "grad_norm": 1.7588955163955688, + "learning_rate": 0.00010681010561599651, + "loss": 0.6687, + "step": 28474 + }, + { + "epoch": 1.975921171327458, + "grad_norm": 1.6344493627548218, + "learning_rate": 0.00010680450003379972, + "loss": 0.5902, + "step": 28475 + }, + { + "epoch": 1.9759905627645549, + "grad_norm": 1.2663453817367554, + "learning_rate": 0.00010679889443012192, + "loss": 0.3887, + "step": 28476 + }, + { + "epoch": 1.9760599542016515, + "grad_norm": 1.5574734210968018, + "learning_rate": 0.0001067932888049808, + "loss": 0.6864, + "step": 28477 + }, + { + "epoch": 1.976129345638748, + "grad_norm": 1.740624189376831, + "learning_rate": 0.00010678768315839405, + "loss": 0.5772, + "step": 28478 + }, + { + "epoch": 1.976198737075845, + "grad_norm": 1.2783890962600708, + "learning_rate": 0.00010678207749037936, + "loss": 0.4501, + "step": 28479 + }, + { + "epoch": 1.9762681285129415, + "grad_norm": 1.3858907222747803, + "learning_rate": 0.00010677647180095447, + "loss": 0.6301, + "step": 28480 + }, + { + "epoch": 1.9763375199500381, + "grad_norm": 1.5542457103729248, + "learning_rate": 0.00010677086609013702, + "loss": 0.5487, + "step": 28481 + }, + { + "epoch": 1.976406911387135, + "grad_norm": 1.9993393421173096, + "learning_rate": 0.00010676526035794477, + "loss": 0.6882, + "step": 28482 + }, + { + "epoch": 1.9764763028242314, + "grad_norm": 1.0801678895950317, + "learning_rate": 0.00010675965460439532, + "loss": 0.4084, + "step": 28483 + }, + { + "epoch": 1.9765456942613282, + "grad_norm": 1.4965629577636719, + "learning_rate": 0.00010675404882950648, + "loss": 0.5903, + "step": 28484 + }, + { + "epoch": 1.9766150856984248, + "grad_norm": 1.4924077987670898, + "learning_rate": 0.00010674844303329586, + "loss": 0.5193, + "step": 28485 + }, + { + "epoch": 1.9766844771355214, + "grad_norm": 1.411219596862793, + "learning_rate": 0.00010674283721578117, + "loss": 0.6107, + "step": 28486 + }, + { + "epoch": 1.9767538685726183, + "grad_norm": 1.1917215585708618, + "learning_rate": 0.00010673723137698018, + "loss": 0.4513, + "step": 28487 + }, + { + "epoch": 1.9768232600097146, + "grad_norm": 1.5189337730407715, + "learning_rate": 0.00010673162551691048, + "loss": 0.5556, + "step": 28488 + }, + { + "epoch": 1.9768926514468115, + "grad_norm": 1.7676880359649658, + "learning_rate": 0.00010672601963558983, + "loss": 0.7798, + "step": 28489 + }, + { + "epoch": 1.976962042883908, + "grad_norm": 1.634340524673462, + "learning_rate": 0.00010672041373303594, + "loss": 0.4462, + "step": 28490 + }, + { + "epoch": 1.9770314343210047, + "grad_norm": 1.3138916492462158, + "learning_rate": 0.00010671480780926645, + "loss": 0.6708, + "step": 28491 + }, + { + "epoch": 1.9771008257581015, + "grad_norm": 1.6743813753128052, + "learning_rate": 0.00010670920186429912, + "loss": 0.6837, + "step": 28492 + }, + { + "epoch": 1.9771702171951981, + "grad_norm": 1.264713168144226, + "learning_rate": 0.0001067035958981516, + "loss": 0.5518, + "step": 28493 + }, + { + "epoch": 1.9772396086322948, + "grad_norm": 2.0280611515045166, + "learning_rate": 0.00010669798991084161, + "loss": 0.7355, + "step": 28494 + }, + { + "epoch": 1.9773090000693916, + "grad_norm": 1.4149279594421387, + "learning_rate": 0.00010669238390238685, + "loss": 0.5852, + "step": 28495 + }, + { + "epoch": 1.977378391506488, + "grad_norm": 1.5153745412826538, + "learning_rate": 0.00010668677787280499, + "loss": 0.5723, + "step": 28496 + }, + { + "epoch": 1.9774477829435848, + "grad_norm": 1.5517042875289917, + "learning_rate": 0.00010668117182211379, + "loss": 0.6086, + "step": 28497 + }, + { + "epoch": 1.9775171743806814, + "grad_norm": 1.8050209283828735, + "learning_rate": 0.00010667556575033086, + "loss": 0.6585, + "step": 28498 + }, + { + "epoch": 1.977586565817778, + "grad_norm": 1.2471281290054321, + "learning_rate": 0.00010666995965747397, + "loss": 0.4122, + "step": 28499 + }, + { + "epoch": 1.9776559572548749, + "grad_norm": 1.4783838987350464, + "learning_rate": 0.0001066643535435608, + "loss": 0.5693, + "step": 28500 + }, + { + "epoch": 1.9777253486919713, + "grad_norm": 1.1917335987091064, + "learning_rate": 0.00010665874740860901, + "loss": 0.5765, + "step": 28501 + }, + { + "epoch": 1.977794740129068, + "grad_norm": 1.1664499044418335, + "learning_rate": 0.00010665314125263635, + "loss": 0.5386, + "step": 28502 + }, + { + "epoch": 1.9778641315661647, + "grad_norm": 1.5914151668548584, + "learning_rate": 0.00010664753507566048, + "loss": 0.6627, + "step": 28503 + }, + { + "epoch": 1.9779335230032613, + "grad_norm": 2.4703407287597656, + "learning_rate": 0.00010664192887769915, + "loss": 0.5495, + "step": 28504 + }, + { + "epoch": 1.9780029144403581, + "grad_norm": 1.624300479888916, + "learning_rate": 0.00010663632265876997, + "loss": 0.7459, + "step": 28505 + }, + { + "epoch": 1.9780723058774548, + "grad_norm": 1.3507696390151978, + "learning_rate": 0.00010663071641889071, + "loss": 0.7793, + "step": 28506 + }, + { + "epoch": 1.9781416973145514, + "grad_norm": 1.3350791931152344, + "learning_rate": 0.00010662511015807903, + "loss": 0.4199, + "step": 28507 + }, + { + "epoch": 1.9782110887516482, + "grad_norm": 1.419891357421875, + "learning_rate": 0.00010661950387635266, + "loss": 0.7633, + "step": 28508 + }, + { + "epoch": 1.9782804801887446, + "grad_norm": 1.4939696788787842, + "learning_rate": 0.00010661389757372931, + "loss": 0.4875, + "step": 28509 + }, + { + "epoch": 1.9783498716258414, + "grad_norm": 1.9802690744400024, + "learning_rate": 0.00010660829125022659, + "loss": 0.743, + "step": 28510 + }, + { + "epoch": 1.978419263062938, + "grad_norm": 1.7778565883636475, + "learning_rate": 0.00010660268490586234, + "loss": 0.527, + "step": 28511 + }, + { + "epoch": 1.9784886545000346, + "grad_norm": 1.2922639846801758, + "learning_rate": 0.00010659707854065413, + "loss": 0.4107, + "step": 28512 + }, + { + "epoch": 1.9785580459371315, + "grad_norm": 1.6758278608322144, + "learning_rate": 0.0001065914721546197, + "loss": 0.5304, + "step": 28513 + }, + { + "epoch": 1.9786274373742279, + "grad_norm": 2.0992462635040283, + "learning_rate": 0.00010658586574777678, + "loss": 0.7075, + "step": 28514 + }, + { + "epoch": 1.9786968288113247, + "grad_norm": 1.8627030849456787, + "learning_rate": 0.00010658025932014306, + "loss": 0.4734, + "step": 28515 + }, + { + "epoch": 1.9787662202484213, + "grad_norm": 1.522568702697754, + "learning_rate": 0.0001065746528717362, + "loss": 0.4757, + "step": 28516 + }, + { + "epoch": 1.978835611685518, + "grad_norm": 1.7659205198287964, + "learning_rate": 0.00010656904640257395, + "loss": 0.6279, + "step": 28517 + }, + { + "epoch": 1.9789050031226147, + "grad_norm": 1.46174955368042, + "learning_rate": 0.00010656343991267395, + "loss": 0.5842, + "step": 28518 + }, + { + "epoch": 1.9789743945597114, + "grad_norm": 1.7040220499038696, + "learning_rate": 0.00010655783340205397, + "loss": 0.4542, + "step": 28519 + }, + { + "epoch": 1.979043785996808, + "grad_norm": 1.8923392295837402, + "learning_rate": 0.00010655222687073163, + "loss": 0.8752, + "step": 28520 + }, + { + "epoch": 1.9791131774339048, + "grad_norm": 0.9946019649505615, + "learning_rate": 0.00010654662031872468, + "loss": 0.3757, + "step": 28521 + }, + { + "epoch": 1.9791825688710012, + "grad_norm": 1.2026947736740112, + "learning_rate": 0.00010654101374605083, + "loss": 0.3979, + "step": 28522 + }, + { + "epoch": 1.979251960308098, + "grad_norm": 2.1630446910858154, + "learning_rate": 0.00010653540715272772, + "loss": 0.7865, + "step": 28523 + }, + { + "epoch": 1.9793213517451946, + "grad_norm": 1.4967201948165894, + "learning_rate": 0.00010652980053877311, + "loss": 0.6612, + "step": 28524 + }, + { + "epoch": 1.9793907431822912, + "grad_norm": 1.0669491291046143, + "learning_rate": 0.00010652419390420469, + "loss": 0.4776, + "step": 28525 + }, + { + "epoch": 1.979460134619388, + "grad_norm": 1.1794713735580444, + "learning_rate": 0.00010651858724904014, + "loss": 0.5829, + "step": 28526 + }, + { + "epoch": 1.9795295260564845, + "grad_norm": 1.252151608467102, + "learning_rate": 0.00010651298057329715, + "loss": 0.3065, + "step": 28527 + }, + { + "epoch": 1.9795989174935813, + "grad_norm": 1.3999543190002441, + "learning_rate": 0.00010650737387699345, + "loss": 0.5881, + "step": 28528 + }, + { + "epoch": 1.979668308930678, + "grad_norm": 1.1485873460769653, + "learning_rate": 0.00010650176716014673, + "loss": 0.3699, + "step": 28529 + }, + { + "epoch": 1.9797377003677745, + "grad_norm": 1.3691494464874268, + "learning_rate": 0.00010649616042277467, + "loss": 0.5774, + "step": 28530 + }, + { + "epoch": 1.9798070918048714, + "grad_norm": 1.8828144073486328, + "learning_rate": 0.00010649055366489498, + "loss": 0.6646, + "step": 28531 + }, + { + "epoch": 1.979876483241968, + "grad_norm": 1.9213706254959106, + "learning_rate": 0.0001064849468865254, + "loss": 0.6936, + "step": 28532 + }, + { + "epoch": 1.9799458746790646, + "grad_norm": 1.2447988986968994, + "learning_rate": 0.00010647934008768357, + "loss": 0.4333, + "step": 28533 + }, + { + "epoch": 1.9800152661161614, + "grad_norm": 1.2643307447433472, + "learning_rate": 0.00010647373326838721, + "loss": 0.5153, + "step": 28534 + }, + { + "epoch": 1.9800846575532578, + "grad_norm": 1.350130319595337, + "learning_rate": 0.00010646812642865404, + "loss": 0.4136, + "step": 28535 + }, + { + "epoch": 1.9801540489903546, + "grad_norm": 1.8878670930862427, + "learning_rate": 0.00010646251956850175, + "loss": 0.5647, + "step": 28536 + }, + { + "epoch": 1.9802234404274512, + "grad_norm": 1.9563956260681152, + "learning_rate": 0.00010645691268794802, + "loss": 0.6403, + "step": 28537 + }, + { + "epoch": 1.9802928318645479, + "grad_norm": 1.5599863529205322, + "learning_rate": 0.00010645130578701057, + "loss": 0.6671, + "step": 28538 + }, + { + "epoch": 1.9803622233016447, + "grad_norm": 1.6232539415359497, + "learning_rate": 0.00010644569886570713, + "loss": 0.5499, + "step": 28539 + }, + { + "epoch": 1.980431614738741, + "grad_norm": 1.6936190128326416, + "learning_rate": 0.00010644009192405531, + "loss": 0.5183, + "step": 28540 + }, + { + "epoch": 1.980501006175838, + "grad_norm": 1.4379807710647583, + "learning_rate": 0.00010643448496207289, + "loss": 0.4867, + "step": 28541 + }, + { + "epoch": 1.9805703976129345, + "grad_norm": 1.2096199989318848, + "learning_rate": 0.00010642887797977756, + "loss": 0.5028, + "step": 28542 + }, + { + "epoch": 1.9806397890500311, + "grad_norm": 1.1553350687026978, + "learning_rate": 0.00010642327097718699, + "loss": 0.3464, + "step": 28543 + }, + { + "epoch": 1.980709180487128, + "grad_norm": 1.870216965675354, + "learning_rate": 0.00010641766395431889, + "loss": 0.7039, + "step": 28544 + }, + { + "epoch": 1.9807785719242246, + "grad_norm": 2.2465767860412598, + "learning_rate": 0.00010641205691119098, + "loss": 0.5356, + "step": 28545 + }, + { + "epoch": 1.9808479633613212, + "grad_norm": 1.4486767053604126, + "learning_rate": 0.00010640644984782098, + "loss": 0.5746, + "step": 28546 + }, + { + "epoch": 1.980917354798418, + "grad_norm": 1.5248126983642578, + "learning_rate": 0.00010640084276422653, + "loss": 0.4878, + "step": 28547 + }, + { + "epoch": 1.9809867462355144, + "grad_norm": 1.743385910987854, + "learning_rate": 0.00010639523566042535, + "loss": 0.4816, + "step": 28548 + }, + { + "epoch": 1.9810561376726112, + "grad_norm": 2.339970111846924, + "learning_rate": 0.0001063896285364352, + "loss": 0.7273, + "step": 28549 + }, + { + "epoch": 1.9811255291097079, + "grad_norm": 1.9399113655090332, + "learning_rate": 0.00010638402139227368, + "loss": 0.5403, + "step": 28550 + }, + { + "epoch": 1.9811949205468045, + "grad_norm": 1.9564765691757202, + "learning_rate": 0.00010637841422795855, + "loss": 0.5827, + "step": 28551 + }, + { + "epoch": 1.9812643119839013, + "grad_norm": 1.3098489046096802, + "learning_rate": 0.00010637280704350756, + "loss": 0.4358, + "step": 28552 + }, + { + "epoch": 1.9813337034209977, + "grad_norm": 1.9563482999801636, + "learning_rate": 0.0001063671998389383, + "loss": 0.4422, + "step": 28553 + }, + { + "epoch": 1.9814030948580945, + "grad_norm": 1.4380773305892944, + "learning_rate": 0.00010636159261426859, + "loss": 0.5135, + "step": 28554 + }, + { + "epoch": 1.9814724862951911, + "grad_norm": 1.2629531621932983, + "learning_rate": 0.00010635598536951601, + "loss": 0.4773, + "step": 28555 + }, + { + "epoch": 1.9815418777322877, + "grad_norm": 2.19337797164917, + "learning_rate": 0.00010635037810469837, + "loss": 0.8192, + "step": 28556 + }, + { + "epoch": 1.9816112691693846, + "grad_norm": 1.0443158149719238, + "learning_rate": 0.00010634477081983329, + "loss": 0.3375, + "step": 28557 + }, + { + "epoch": 1.9816806606064812, + "grad_norm": 1.6465773582458496, + "learning_rate": 0.00010633916351493851, + "loss": 0.4214, + "step": 28558 + }, + { + "epoch": 1.9817500520435778, + "grad_norm": 1.1357042789459229, + "learning_rate": 0.00010633355619003173, + "loss": 0.5051, + "step": 28559 + }, + { + "epoch": 1.9818194434806746, + "grad_norm": 2.065211057662964, + "learning_rate": 0.00010632794884513064, + "loss": 0.6377, + "step": 28560 + }, + { + "epoch": 1.981888834917771, + "grad_norm": 1.406003713607788, + "learning_rate": 0.00010632234148025295, + "loss": 0.5786, + "step": 28561 + }, + { + "epoch": 1.9819582263548678, + "grad_norm": 1.9219348430633545, + "learning_rate": 0.00010631673409541636, + "loss": 0.8258, + "step": 28562 + }, + { + "epoch": 1.9820276177919645, + "grad_norm": 1.5133332014083862, + "learning_rate": 0.0001063111266906386, + "loss": 0.6251, + "step": 28563 + }, + { + "epoch": 1.982097009229061, + "grad_norm": 2.1834232807159424, + "learning_rate": 0.00010630551926593733, + "loss": 0.6279, + "step": 28564 + }, + { + "epoch": 1.982166400666158, + "grad_norm": 1.5523180961608887, + "learning_rate": 0.00010629991182133024, + "loss": 0.6235, + "step": 28565 + }, + { + "epoch": 1.9822357921032545, + "grad_norm": 1.3223373889923096, + "learning_rate": 0.0001062943043568351, + "loss": 0.4906, + "step": 28566 + }, + { + "epoch": 1.9823051835403511, + "grad_norm": 1.8956118822097778, + "learning_rate": 0.00010628869687246954, + "loss": 0.7431, + "step": 28567 + }, + { + "epoch": 1.9823745749774477, + "grad_norm": 1.7813831567764282, + "learning_rate": 0.00010628308936825131, + "loss": 0.6526, + "step": 28568 + }, + { + "epoch": 1.9824439664145443, + "grad_norm": 1.6342363357543945, + "learning_rate": 0.0001062774818441981, + "loss": 0.4887, + "step": 28569 + }, + { + "epoch": 1.9825133578516412, + "grad_norm": 1.4954724311828613, + "learning_rate": 0.00010627187430032761, + "loss": 0.5027, + "step": 28570 + }, + { + "epoch": 1.9825827492887378, + "grad_norm": 1.1504836082458496, + "learning_rate": 0.00010626626673665754, + "loss": 0.7589, + "step": 28571 + }, + { + "epoch": 1.9826521407258344, + "grad_norm": 2.003978967666626, + "learning_rate": 0.00010626065915320557, + "loss": 0.6917, + "step": 28572 + }, + { + "epoch": 1.9827215321629312, + "grad_norm": 1.6475547552108765, + "learning_rate": 0.00010625505154998947, + "loss": 0.5154, + "step": 28573 + }, + { + "epoch": 1.9827909236000276, + "grad_norm": 1.3184257745742798, + "learning_rate": 0.00010624944392702684, + "loss": 0.489, + "step": 28574 + }, + { + "epoch": 1.9828603150371245, + "grad_norm": 2.2367494106292725, + "learning_rate": 0.00010624383628433547, + "loss": 0.9592, + "step": 28575 + }, + { + "epoch": 1.982929706474221, + "grad_norm": 1.6820080280303955, + "learning_rate": 0.00010623822862193303, + "loss": 0.7577, + "step": 28576 + }, + { + "epoch": 1.9829990979113177, + "grad_norm": 2.5649008750915527, + "learning_rate": 0.00010623262093983726, + "loss": 0.6039, + "step": 28577 + }, + { + "epoch": 1.9830684893484145, + "grad_norm": 1.48207426071167, + "learning_rate": 0.00010622701323806579, + "loss": 0.6591, + "step": 28578 + }, + { + "epoch": 1.9831378807855111, + "grad_norm": 1.582681655883789, + "learning_rate": 0.00010622140551663635, + "loss": 0.4814, + "step": 28579 + }, + { + "epoch": 1.9832072722226077, + "grad_norm": 1.2064656019210815, + "learning_rate": 0.00010621579777556669, + "loss": 0.3367, + "step": 28580 + }, + { + "epoch": 1.9832766636597043, + "grad_norm": 1.7974138259887695, + "learning_rate": 0.00010621019001487444, + "loss": 0.4477, + "step": 28581 + }, + { + "epoch": 1.983346055096801, + "grad_norm": 1.9598159790039062, + "learning_rate": 0.00010620458223457737, + "loss": 0.5864, + "step": 28582 + }, + { + "epoch": 1.9834154465338978, + "grad_norm": 1.6164453029632568, + "learning_rate": 0.00010619897443469313, + "loss": 0.4465, + "step": 28583 + }, + { + "epoch": 1.9834848379709944, + "grad_norm": 1.3361773490905762, + "learning_rate": 0.00010619336661523949, + "loss": 0.5958, + "step": 28584 + }, + { + "epoch": 1.983554229408091, + "grad_norm": 1.4551868438720703, + "learning_rate": 0.00010618775877623406, + "loss": 0.4837, + "step": 28585 + }, + { + "epoch": 1.9836236208451878, + "grad_norm": 1.339618444442749, + "learning_rate": 0.00010618215091769461, + "loss": 0.4473, + "step": 28586 + }, + { + "epoch": 1.9836930122822842, + "grad_norm": 1.4673376083374023, + "learning_rate": 0.00010617654303963884, + "loss": 0.4348, + "step": 28587 + }, + { + "epoch": 1.983762403719381, + "grad_norm": 1.5667924880981445, + "learning_rate": 0.00010617093514208445, + "loss": 0.6789, + "step": 28588 + }, + { + "epoch": 1.9838317951564777, + "grad_norm": 0.9620230793952942, + "learning_rate": 0.00010616532722504912, + "loss": 0.4236, + "step": 28589 + }, + { + "epoch": 1.9839011865935743, + "grad_norm": 1.3506855964660645, + "learning_rate": 0.00010615971928855055, + "loss": 0.5892, + "step": 28590 + }, + { + "epoch": 1.9839705780306711, + "grad_norm": 1.5561927556991577, + "learning_rate": 0.00010615411133260649, + "loss": 0.6034, + "step": 28591 + }, + { + "epoch": 1.9840399694677677, + "grad_norm": 1.5303934812545776, + "learning_rate": 0.00010614850335723461, + "loss": 0.5545, + "step": 28592 + }, + { + "epoch": 1.9841093609048643, + "grad_norm": 1.5381697416305542, + "learning_rate": 0.00010614289536245258, + "loss": 0.4362, + "step": 28593 + }, + { + "epoch": 1.984178752341961, + "grad_norm": 1.3987829685211182, + "learning_rate": 0.00010613728734827819, + "loss": 0.7263, + "step": 28594 + }, + { + "epoch": 1.9842481437790576, + "grad_norm": 2.1050832271575928, + "learning_rate": 0.00010613167931472907, + "loss": 0.6944, + "step": 28595 + }, + { + "epoch": 1.9843175352161544, + "grad_norm": 1.3701668977737427, + "learning_rate": 0.00010612607126182295, + "loss": 0.6628, + "step": 28596 + }, + { + "epoch": 1.984386926653251, + "grad_norm": 1.2034152746200562, + "learning_rate": 0.00010612046318957756, + "loss": 0.5116, + "step": 28597 + }, + { + "epoch": 1.9844563180903476, + "grad_norm": 1.352134346961975, + "learning_rate": 0.00010611485509801054, + "loss": 0.4882, + "step": 28598 + }, + { + "epoch": 1.9845257095274444, + "grad_norm": 1.8831008672714233, + "learning_rate": 0.00010610924698713964, + "loss": 0.6755, + "step": 28599 + }, + { + "epoch": 1.9845951009645408, + "grad_norm": 1.5071587562561035, + "learning_rate": 0.00010610363885698256, + "loss": 0.4954, + "step": 28600 + }, + { + "epoch": 1.9846644924016377, + "grad_norm": 1.2552834749221802, + "learning_rate": 0.00010609803070755702, + "loss": 0.6616, + "step": 28601 + }, + { + "epoch": 1.9847338838387343, + "grad_norm": 1.10537850856781, + "learning_rate": 0.0001060924225388807, + "loss": 0.4111, + "step": 28602 + }, + { + "epoch": 1.984803275275831, + "grad_norm": 1.1642441749572754, + "learning_rate": 0.00010608681435097127, + "loss": 0.3406, + "step": 28603 + }, + { + "epoch": 1.9848726667129277, + "grad_norm": 1.3348658084869385, + "learning_rate": 0.00010608120614384653, + "loss": 0.4513, + "step": 28604 + }, + { + "epoch": 1.9849420581500243, + "grad_norm": 1.4562697410583496, + "learning_rate": 0.00010607559791752409, + "loss": 0.6095, + "step": 28605 + }, + { + "epoch": 1.985011449587121, + "grad_norm": 1.3870199918746948, + "learning_rate": 0.0001060699896720217, + "loss": 0.4731, + "step": 28606 + }, + { + "epoch": 1.9850808410242176, + "grad_norm": 1.8215292692184448, + "learning_rate": 0.00010606438140735707, + "loss": 0.774, + "step": 28607 + }, + { + "epoch": 1.9851502324613142, + "grad_norm": 1.276366114616394, + "learning_rate": 0.00010605877312354791, + "loss": 0.4516, + "step": 28608 + }, + { + "epoch": 1.985219623898411, + "grad_norm": 1.2095612287521362, + "learning_rate": 0.00010605316482061185, + "loss": 0.4459, + "step": 28609 + }, + { + "epoch": 1.9852890153355076, + "grad_norm": 1.3105186223983765, + "learning_rate": 0.00010604755649856668, + "loss": 0.573, + "step": 28610 + }, + { + "epoch": 1.9853584067726042, + "grad_norm": 3.794114112854004, + "learning_rate": 0.00010604194815743009, + "loss": 0.7735, + "step": 28611 + }, + { + "epoch": 1.985427798209701, + "grad_norm": 1.2737622261047363, + "learning_rate": 0.00010603633979721974, + "loss": 0.3967, + "step": 28612 + }, + { + "epoch": 1.9854971896467974, + "grad_norm": 1.759628415107727, + "learning_rate": 0.00010603073141795337, + "loss": 0.7353, + "step": 28613 + }, + { + "epoch": 1.9855665810838943, + "grad_norm": 1.852262258529663, + "learning_rate": 0.00010602512301964868, + "loss": 0.6944, + "step": 28614 + }, + { + "epoch": 1.9856359725209909, + "grad_norm": 1.2050862312316895, + "learning_rate": 0.0001060195146023234, + "loss": 0.4833, + "step": 28615 + }, + { + "epoch": 1.9857053639580875, + "grad_norm": 1.6388252973556519, + "learning_rate": 0.0001060139061659952, + "loss": 0.6792, + "step": 28616 + }, + { + "epoch": 1.9857747553951843, + "grad_norm": 1.5978364944458008, + "learning_rate": 0.0001060082977106818, + "loss": 0.5486, + "step": 28617 + }, + { + "epoch": 1.985844146832281, + "grad_norm": 2.0452778339385986, + "learning_rate": 0.00010600268923640088, + "loss": 0.6682, + "step": 28618 + }, + { + "epoch": 1.9859135382693776, + "grad_norm": 2.514843702316284, + "learning_rate": 0.00010599708074317019, + "loss": 0.5501, + "step": 28619 + }, + { + "epoch": 1.9859829297064744, + "grad_norm": 1.1805944442749023, + "learning_rate": 0.00010599147223100739, + "loss": 0.607, + "step": 28620 + }, + { + "epoch": 1.9860523211435708, + "grad_norm": 1.8412001132965088, + "learning_rate": 0.00010598586369993022, + "loss": 0.5997, + "step": 28621 + }, + { + "epoch": 1.9861217125806676, + "grad_norm": 1.6635847091674805, + "learning_rate": 0.00010598025514995638, + "loss": 0.6609, + "step": 28622 + }, + { + "epoch": 1.9861911040177642, + "grad_norm": 1.5467156171798706, + "learning_rate": 0.00010597464658110354, + "loss": 0.4535, + "step": 28623 + }, + { + "epoch": 1.9862604954548608, + "grad_norm": 1.359511375427246, + "learning_rate": 0.00010596903799338948, + "loss": 0.53, + "step": 28624 + }, + { + "epoch": 1.9863298868919577, + "grad_norm": 1.4868395328521729, + "learning_rate": 0.00010596342938683186, + "loss": 0.6326, + "step": 28625 + }, + { + "epoch": 1.986399278329054, + "grad_norm": 2.0666866302490234, + "learning_rate": 0.00010595782076144834, + "loss": 0.4929, + "step": 28626 + }, + { + "epoch": 1.9864686697661509, + "grad_norm": 1.5439624786376953, + "learning_rate": 0.00010595221211725668, + "loss": 0.4913, + "step": 28627 + }, + { + "epoch": 1.9865380612032475, + "grad_norm": 1.3267971277236938, + "learning_rate": 0.00010594660345427458, + "loss": 0.4647, + "step": 28628 + }, + { + "epoch": 1.986607452640344, + "grad_norm": 1.0278149843215942, + "learning_rate": 0.00010594099477251978, + "loss": 0.5508, + "step": 28629 + }, + { + "epoch": 1.986676844077441, + "grad_norm": 1.2291394472122192, + "learning_rate": 0.0001059353860720099, + "loss": 0.556, + "step": 28630 + }, + { + "epoch": 1.9867462355145376, + "grad_norm": 1.2036199569702148, + "learning_rate": 0.00010592977735276271, + "loss": 0.4424, + "step": 28631 + }, + { + "epoch": 1.9868156269516342, + "grad_norm": 1.57524836063385, + "learning_rate": 0.00010592416861479592, + "loss": 0.8776, + "step": 28632 + }, + { + "epoch": 1.986885018388731, + "grad_norm": 1.5718978643417358, + "learning_rate": 0.0001059185598581272, + "loss": 0.6266, + "step": 28633 + }, + { + "epoch": 1.9869544098258274, + "grad_norm": 1.3702688217163086, + "learning_rate": 0.00010591295108277426, + "loss": 0.6237, + "step": 28634 + }, + { + "epoch": 1.9870238012629242, + "grad_norm": 1.4111170768737793, + "learning_rate": 0.00010590734228875485, + "loss": 0.5282, + "step": 28635 + }, + { + "epoch": 1.9870931927000208, + "grad_norm": 1.4950776100158691, + "learning_rate": 0.00010590173347608663, + "loss": 0.6304, + "step": 28636 + }, + { + "epoch": 1.9871625841371174, + "grad_norm": 1.5934245586395264, + "learning_rate": 0.00010589612464478731, + "loss": 0.533, + "step": 28637 + }, + { + "epoch": 1.9872319755742143, + "grad_norm": 1.2333850860595703, + "learning_rate": 0.00010589051579487462, + "loss": 0.5383, + "step": 28638 + }, + { + "epoch": 1.9873013670113107, + "grad_norm": 2.623605966567993, + "learning_rate": 0.00010588490692636627, + "loss": 0.5777, + "step": 28639 + }, + { + "epoch": 1.9873707584484075, + "grad_norm": 1.523634433746338, + "learning_rate": 0.00010587929803927994, + "loss": 0.7353, + "step": 28640 + }, + { + "epoch": 1.987440149885504, + "grad_norm": 2.2360994815826416, + "learning_rate": 0.00010587368913363336, + "loss": 0.6075, + "step": 28641 + }, + { + "epoch": 1.9875095413226007, + "grad_norm": 1.9721462726593018, + "learning_rate": 0.00010586808020944419, + "loss": 0.6092, + "step": 28642 + }, + { + "epoch": 1.9875789327596975, + "grad_norm": 1.3597434759140015, + "learning_rate": 0.00010586247126673021, + "loss": 0.6422, + "step": 28643 + }, + { + "epoch": 1.9876483241967942, + "grad_norm": 1.117524266242981, + "learning_rate": 0.00010585686230550906, + "loss": 0.3079, + "step": 28644 + }, + { + "epoch": 1.9877177156338908, + "grad_norm": 1.4174280166625977, + "learning_rate": 0.00010585125332579849, + "loss": 0.6214, + "step": 28645 + }, + { + "epoch": 1.9877871070709876, + "grad_norm": 1.3136216402053833, + "learning_rate": 0.00010584564432761619, + "loss": 0.4741, + "step": 28646 + }, + { + "epoch": 1.987856498508084, + "grad_norm": 1.6990506649017334, + "learning_rate": 0.00010584003531097985, + "loss": 0.8121, + "step": 28647 + }, + { + "epoch": 1.9879258899451808, + "grad_norm": 2.0845022201538086, + "learning_rate": 0.00010583442627590722, + "loss": 0.7001, + "step": 28648 + }, + { + "epoch": 1.9879952813822774, + "grad_norm": 1.4395759105682373, + "learning_rate": 0.000105828817222416, + "loss": 0.6599, + "step": 28649 + }, + { + "epoch": 1.988064672819374, + "grad_norm": 1.3010497093200684, + "learning_rate": 0.00010582320815052384, + "loss": 0.463, + "step": 28650 + }, + { + "epoch": 1.9881340642564709, + "grad_norm": 2.415764570236206, + "learning_rate": 0.0001058175990602485, + "loss": 1.0049, + "step": 28651 + }, + { + "epoch": 1.9882034556935673, + "grad_norm": 1.313132882118225, + "learning_rate": 0.00010581198995160769, + "loss": 0.5116, + "step": 28652 + }, + { + "epoch": 1.988272847130664, + "grad_norm": 1.687726378440857, + "learning_rate": 0.00010580638082461908, + "loss": 0.7522, + "step": 28653 + }, + { + "epoch": 1.9883422385677607, + "grad_norm": 1.6950595378875732, + "learning_rate": 0.0001058007716793004, + "loss": 0.5453, + "step": 28654 + }, + { + "epoch": 1.9884116300048573, + "grad_norm": 1.3191334009170532, + "learning_rate": 0.00010579516251566936, + "loss": 0.4404, + "step": 28655 + }, + { + "epoch": 1.9884810214419542, + "grad_norm": 1.6116057634353638, + "learning_rate": 0.00010578955333374369, + "loss": 0.4211, + "step": 28656 + }, + { + "epoch": 1.9885504128790508, + "grad_norm": 1.4153039455413818, + "learning_rate": 0.00010578394413354104, + "loss": 0.575, + "step": 28657 + }, + { + "epoch": 1.9886198043161474, + "grad_norm": 1.5037305355072021, + "learning_rate": 0.0001057783349150792, + "loss": 0.6407, + "step": 28658 + }, + { + "epoch": 1.9886891957532442, + "grad_norm": 1.6474813222885132, + "learning_rate": 0.00010577272567837578, + "loss": 0.3939, + "step": 28659 + }, + { + "epoch": 1.9887585871903406, + "grad_norm": 1.1246141195297241, + "learning_rate": 0.00010576711642344856, + "loss": 0.5599, + "step": 28660 + }, + { + "epoch": 1.9888279786274374, + "grad_norm": 1.213690996170044, + "learning_rate": 0.00010576150715031518, + "loss": 0.3969, + "step": 28661 + }, + { + "epoch": 1.988897370064534, + "grad_norm": 1.4475443363189697, + "learning_rate": 0.00010575589785899342, + "loss": 0.3757, + "step": 28662 + }, + { + "epoch": 1.9889667615016307, + "grad_norm": 2.000467300415039, + "learning_rate": 0.00010575028854950098, + "loss": 0.5481, + "step": 28663 + }, + { + "epoch": 1.9890361529387275, + "grad_norm": 1.4434689283370972, + "learning_rate": 0.00010574467922185551, + "loss": 0.4332, + "step": 28664 + }, + { + "epoch": 1.9891055443758239, + "grad_norm": 1.5952646732330322, + "learning_rate": 0.00010573906987607476, + "loss": 0.544, + "step": 28665 + }, + { + "epoch": 1.9891749358129207, + "grad_norm": 2.1730570793151855, + "learning_rate": 0.00010573346051217643, + "loss": 0.7413, + "step": 28666 + }, + { + "epoch": 1.9892443272500173, + "grad_norm": 1.7226990461349487, + "learning_rate": 0.00010572785113017824, + "loss": 0.4946, + "step": 28667 + }, + { + "epoch": 1.989313718687114, + "grad_norm": 1.088474988937378, + "learning_rate": 0.00010572224173009791, + "loss": 0.4971, + "step": 28668 + }, + { + "epoch": 1.9893831101242108, + "grad_norm": 1.5925230979919434, + "learning_rate": 0.00010571663231195308, + "loss": 0.6353, + "step": 28669 + }, + { + "epoch": 1.9894525015613074, + "grad_norm": 1.2525393962860107, + "learning_rate": 0.00010571102287576155, + "loss": 0.5994, + "step": 28670 + }, + { + "epoch": 1.989521892998404, + "grad_norm": 1.1553504467010498, + "learning_rate": 0.00010570541342154094, + "loss": 0.4535, + "step": 28671 + }, + { + "epoch": 1.9895912844355008, + "grad_norm": 1.3387832641601562, + "learning_rate": 0.00010569980394930903, + "loss": 0.3855, + "step": 28672 + }, + { + "epoch": 1.9896606758725972, + "grad_norm": 7.017827987670898, + "learning_rate": 0.0001056941944590835, + "loss": 0.4858, + "step": 28673 + }, + { + "epoch": 1.989730067309694, + "grad_norm": 1.310163974761963, + "learning_rate": 0.00010568858495088205, + "loss": 0.4963, + "step": 28674 + }, + { + "epoch": 1.9897994587467907, + "grad_norm": 1.3639814853668213, + "learning_rate": 0.00010568297542472241, + "loss": 0.5635, + "step": 28675 + }, + { + "epoch": 1.9898688501838873, + "grad_norm": 3.106365442276001, + "learning_rate": 0.00010567736588062227, + "loss": 0.7424, + "step": 28676 + }, + { + "epoch": 1.989938241620984, + "grad_norm": 1.267905831336975, + "learning_rate": 0.00010567175631859934, + "loss": 0.6105, + "step": 28677 + }, + { + "epoch": 1.9900076330580805, + "grad_norm": 1.2742977142333984, + "learning_rate": 0.00010566614673867135, + "loss": 0.4392, + "step": 28678 + }, + { + "epoch": 1.9900770244951773, + "grad_norm": 1.1844723224639893, + "learning_rate": 0.00010566053714085598, + "loss": 0.4465, + "step": 28679 + }, + { + "epoch": 1.990146415932274, + "grad_norm": 1.2483023405075073, + "learning_rate": 0.00010565492752517094, + "loss": 0.5639, + "step": 28680 + }, + { + "epoch": 1.9902158073693705, + "grad_norm": 1.6721413135528564, + "learning_rate": 0.00010564931789163397, + "loss": 0.4623, + "step": 28681 + }, + { + "epoch": 1.9902851988064674, + "grad_norm": 1.3204017877578735, + "learning_rate": 0.00010564370824026273, + "loss": 0.6022, + "step": 28682 + }, + { + "epoch": 1.990354590243564, + "grad_norm": 1.5487200021743774, + "learning_rate": 0.00010563809857107496, + "loss": 0.5873, + "step": 28683 + }, + { + "epoch": 1.9904239816806606, + "grad_norm": 1.1821246147155762, + "learning_rate": 0.00010563248888408841, + "loss": 0.4503, + "step": 28684 + }, + { + "epoch": 1.9904933731177574, + "grad_norm": 1.5058010816574097, + "learning_rate": 0.0001056268791793207, + "loss": 0.4617, + "step": 28685 + }, + { + "epoch": 1.9905627645548538, + "grad_norm": 1.6279529333114624, + "learning_rate": 0.0001056212694567896, + "loss": 0.5246, + "step": 28686 + }, + { + "epoch": 1.9906321559919506, + "grad_norm": 1.5805542469024658, + "learning_rate": 0.00010561565971651282, + "loss": 0.7518, + "step": 28687 + }, + { + "epoch": 1.9907015474290473, + "grad_norm": 2.0689361095428467, + "learning_rate": 0.00010561004995850803, + "loss": 0.7546, + "step": 28688 + }, + { + "epoch": 1.9907709388661439, + "grad_norm": 1.714417815208435, + "learning_rate": 0.00010560444018279297, + "loss": 0.5905, + "step": 28689 + }, + { + "epoch": 1.9908403303032407, + "grad_norm": 1.097120761871338, + "learning_rate": 0.00010559883038938533, + "loss": 0.2936, + "step": 28690 + }, + { + "epoch": 1.990909721740337, + "grad_norm": 1.3666313886642456, + "learning_rate": 0.00010559322057830287, + "loss": 0.6417, + "step": 28691 + }, + { + "epoch": 1.990979113177434, + "grad_norm": 1.1584614515304565, + "learning_rate": 0.00010558761074956325, + "loss": 0.5191, + "step": 28692 + }, + { + "epoch": 1.9910485046145305, + "grad_norm": 1.6407933235168457, + "learning_rate": 0.00010558200090318418, + "loss": 0.6097, + "step": 28693 + }, + { + "epoch": 1.9911178960516271, + "grad_norm": 3.5667824745178223, + "learning_rate": 0.00010557639103918338, + "loss": 0.7617, + "step": 28694 + }, + { + "epoch": 1.991187287488724, + "grad_norm": 1.6799906492233276, + "learning_rate": 0.00010557078115757857, + "loss": 0.5599, + "step": 28695 + }, + { + "epoch": 1.9912566789258206, + "grad_norm": 1.3174703121185303, + "learning_rate": 0.00010556517125838745, + "loss": 0.4265, + "step": 28696 + }, + { + "epoch": 1.9913260703629172, + "grad_norm": 1.2022473812103271, + "learning_rate": 0.0001055595613416277, + "loss": 0.441, + "step": 28697 + }, + { + "epoch": 1.991395461800014, + "grad_norm": 1.75735342502594, + "learning_rate": 0.0001055539514073171, + "loss": 0.712, + "step": 28698 + }, + { + "epoch": 1.9914648532371104, + "grad_norm": 1.453547477722168, + "learning_rate": 0.0001055483414554733, + "loss": 0.6461, + "step": 28699 + }, + { + "epoch": 1.9915342446742073, + "grad_norm": 1.597657561302185, + "learning_rate": 0.00010554273148611402, + "loss": 0.4868, + "step": 28700 + }, + { + "epoch": 1.9916036361113039, + "grad_norm": 1.7849918603897095, + "learning_rate": 0.000105537121499257, + "loss": 0.6886, + "step": 28701 + }, + { + "epoch": 1.9916730275484005, + "grad_norm": 2.266552448272705, + "learning_rate": 0.00010553151149491994, + "loss": 0.5999, + "step": 28702 + }, + { + "epoch": 1.9917424189854973, + "grad_norm": 2.1686172485351562, + "learning_rate": 0.00010552590147312052, + "loss": 0.6432, + "step": 28703 + }, + { + "epoch": 1.9918118104225937, + "grad_norm": 1.1932626962661743, + "learning_rate": 0.00010552029143387645, + "loss": 0.5971, + "step": 28704 + }, + { + "epoch": 1.9918812018596905, + "grad_norm": 2.173745632171631, + "learning_rate": 0.0001055146813772055, + "loss": 0.7426, + "step": 28705 + }, + { + "epoch": 1.9919505932967871, + "grad_norm": 1.4898548126220703, + "learning_rate": 0.0001055090713031253, + "loss": 0.412, + "step": 28706 + }, + { + "epoch": 1.9920199847338838, + "grad_norm": 1.7103191614151, + "learning_rate": 0.00010550346121165365, + "loss": 0.8155, + "step": 28707 + }, + { + "epoch": 1.9920893761709806, + "grad_norm": 1.303999900817871, + "learning_rate": 0.00010549785110280818, + "loss": 0.4181, + "step": 28708 + }, + { + "epoch": 1.9921587676080772, + "grad_norm": 1.5561274290084839, + "learning_rate": 0.00010549224097660664, + "loss": 0.4764, + "step": 28709 + }, + { + "epoch": 1.9922281590451738, + "grad_norm": 1.6993266344070435, + "learning_rate": 0.00010548663083306672, + "loss": 0.7102, + "step": 28710 + }, + { + "epoch": 1.9922975504822706, + "grad_norm": 1.76482093334198, + "learning_rate": 0.00010548102067220615, + "loss": 0.7403, + "step": 28711 + }, + { + "epoch": 1.992366941919367, + "grad_norm": 1.9742332696914673, + "learning_rate": 0.00010547541049404266, + "loss": 0.6635, + "step": 28712 + }, + { + "epoch": 1.9924363333564639, + "grad_norm": 1.504778504371643, + "learning_rate": 0.0001054698002985939, + "loss": 0.4129, + "step": 28713 + }, + { + "epoch": 1.9925057247935605, + "grad_norm": 1.4906513690948486, + "learning_rate": 0.00010546419008587763, + "loss": 0.4564, + "step": 28714 + }, + { + "epoch": 1.992575116230657, + "grad_norm": 0.9778075814247131, + "learning_rate": 0.00010545857985591157, + "loss": 0.4065, + "step": 28715 + }, + { + "epoch": 1.992644507667754, + "grad_norm": 1.5007495880126953, + "learning_rate": 0.00010545296960871335, + "loss": 0.349, + "step": 28716 + }, + { + "epoch": 1.9927138991048503, + "grad_norm": 1.540010929107666, + "learning_rate": 0.00010544735934430077, + "loss": 0.5317, + "step": 28717 + }, + { + "epoch": 1.9927832905419471, + "grad_norm": 1.7910432815551758, + "learning_rate": 0.00010544174906269151, + "loss": 0.5591, + "step": 28718 + }, + { + "epoch": 1.9928526819790437, + "grad_norm": 2.0626354217529297, + "learning_rate": 0.00010543613876390328, + "loss": 0.5012, + "step": 28719 + }, + { + "epoch": 1.9929220734161404, + "grad_norm": 1.5799816846847534, + "learning_rate": 0.00010543052844795378, + "loss": 0.4452, + "step": 28720 + }, + { + "epoch": 1.9929914648532372, + "grad_norm": 1.539365530014038, + "learning_rate": 0.00010542491811486072, + "loss": 0.6408, + "step": 28721 + }, + { + "epoch": 1.9930608562903338, + "grad_norm": 1.2773149013519287, + "learning_rate": 0.00010541930776464186, + "loss": 0.4562, + "step": 28722 + }, + { + "epoch": 1.9931302477274304, + "grad_norm": 1.810287594795227, + "learning_rate": 0.00010541369739731484, + "loss": 0.5399, + "step": 28723 + }, + { + "epoch": 1.9931996391645272, + "grad_norm": 1.8538329601287842, + "learning_rate": 0.00010540808701289742, + "loss": 0.6542, + "step": 28724 + }, + { + "epoch": 1.9932690306016236, + "grad_norm": 1.5006086826324463, + "learning_rate": 0.00010540247661140731, + "loss": 0.4364, + "step": 28725 + }, + { + "epoch": 1.9933384220387205, + "grad_norm": 1.5188616514205933, + "learning_rate": 0.00010539686619286219, + "loss": 0.6013, + "step": 28726 + }, + { + "epoch": 1.993407813475817, + "grad_norm": 1.586509346961975, + "learning_rate": 0.0001053912557572798, + "loss": 0.55, + "step": 28727 + }, + { + "epoch": 1.9934772049129137, + "grad_norm": 1.8144086599349976, + "learning_rate": 0.00010538564530467782, + "loss": 0.5742, + "step": 28728 + }, + { + "epoch": 1.9935465963500105, + "grad_norm": 1.620875358581543, + "learning_rate": 0.00010538003483507402, + "loss": 0.4478, + "step": 28729 + }, + { + "epoch": 1.993615987787107, + "grad_norm": 1.5689302682876587, + "learning_rate": 0.00010537442434848606, + "loss": 0.4697, + "step": 28730 + }, + { + "epoch": 1.9936853792242037, + "grad_norm": 1.2968908548355103, + "learning_rate": 0.00010536881384493164, + "loss": 0.4315, + "step": 28731 + }, + { + "epoch": 1.9937547706613004, + "grad_norm": 1.4260252714157104, + "learning_rate": 0.00010536320332442853, + "loss": 0.577, + "step": 28732 + }, + { + "epoch": 1.993824162098397, + "grad_norm": 1.1217557191848755, + "learning_rate": 0.00010535759278699439, + "loss": 0.3513, + "step": 28733 + }, + { + "epoch": 1.9938935535354938, + "grad_norm": 1.1807323694229126, + "learning_rate": 0.00010535198223264694, + "loss": 0.4521, + "step": 28734 + }, + { + "epoch": 1.9939629449725904, + "grad_norm": 1.312474250793457, + "learning_rate": 0.00010534637166140393, + "loss": 0.6157, + "step": 28735 + }, + { + "epoch": 1.994032336409687, + "grad_norm": 1.4903769493103027, + "learning_rate": 0.00010534076107328304, + "loss": 0.4361, + "step": 28736 + }, + { + "epoch": 1.9941017278467839, + "grad_norm": 1.3459042310714722, + "learning_rate": 0.00010533515046830198, + "loss": 0.5845, + "step": 28737 + }, + { + "epoch": 1.9941711192838802, + "grad_norm": 1.1514019966125488, + "learning_rate": 0.00010532953984647847, + "loss": 0.3808, + "step": 28738 + }, + { + "epoch": 1.994240510720977, + "grad_norm": 1.7247997522354126, + "learning_rate": 0.00010532392920783024, + "loss": 0.5084, + "step": 28739 + }, + { + "epoch": 1.9943099021580737, + "grad_norm": 1.4229859113693237, + "learning_rate": 0.00010531831855237495, + "loss": 0.7081, + "step": 28740 + }, + { + "epoch": 1.9943792935951703, + "grad_norm": 1.3195115327835083, + "learning_rate": 0.00010531270788013035, + "loss": 0.523, + "step": 28741 + }, + { + "epoch": 1.9944486850322671, + "grad_norm": 1.1756799221038818, + "learning_rate": 0.00010530709719111414, + "loss": 0.3677, + "step": 28742 + }, + { + "epoch": 1.9945180764693635, + "grad_norm": 1.3732762336730957, + "learning_rate": 0.00010530148648534407, + "loss": 0.5739, + "step": 28743 + }, + { + "epoch": 1.9945874679064604, + "grad_norm": 1.447336196899414, + "learning_rate": 0.00010529587576283782, + "loss": 0.6612, + "step": 28744 + }, + { + "epoch": 1.994656859343557, + "grad_norm": 1.4720731973648071, + "learning_rate": 0.00010529026502361307, + "loss": 0.6143, + "step": 28745 + }, + { + "epoch": 1.9947262507806536, + "grad_norm": 1.9524298906326294, + "learning_rate": 0.00010528465426768761, + "loss": 0.8426, + "step": 28746 + }, + { + "epoch": 1.9947956422177504, + "grad_norm": 1.497719407081604, + "learning_rate": 0.00010527904349507911, + "loss": 0.6495, + "step": 28747 + }, + { + "epoch": 1.994865033654847, + "grad_norm": 2.321305513381958, + "learning_rate": 0.00010527343270580525, + "loss": 0.4634, + "step": 28748 + }, + { + "epoch": 1.9949344250919436, + "grad_norm": 2.150644063949585, + "learning_rate": 0.0001052678218998838, + "loss": 0.4801, + "step": 28749 + }, + { + "epoch": 1.9950038165290405, + "grad_norm": 1.198905110359192, + "learning_rate": 0.00010526221107733243, + "loss": 0.5029, + "step": 28750 + }, + { + "epoch": 1.9950732079661369, + "grad_norm": 1.3786207437515259, + "learning_rate": 0.00010525660023816888, + "loss": 0.6878, + "step": 28751 + }, + { + "epoch": 1.9951425994032337, + "grad_norm": 1.3438694477081299, + "learning_rate": 0.00010525098938241082, + "loss": 0.3988, + "step": 28752 + }, + { + "epoch": 1.9952119908403303, + "grad_norm": 1.4318819046020508, + "learning_rate": 0.00010524537851007605, + "loss": 0.5987, + "step": 28753 + }, + { + "epoch": 1.995281382277427, + "grad_norm": 1.8271764516830444, + "learning_rate": 0.0001052397676211822, + "loss": 0.705, + "step": 28754 + }, + { + "epoch": 1.9953507737145237, + "grad_norm": 1.5042637586593628, + "learning_rate": 0.00010523415671574699, + "loss": 0.6263, + "step": 28755 + }, + { + "epoch": 1.9954201651516201, + "grad_norm": 1.9146015644073486, + "learning_rate": 0.0001052285457937882, + "loss": 0.5973, + "step": 28756 + }, + { + "epoch": 1.995489556588717, + "grad_norm": 1.161239743232727, + "learning_rate": 0.00010522293485532346, + "loss": 0.4663, + "step": 28757 + }, + { + "epoch": 1.9955589480258136, + "grad_norm": 1.5733060836791992, + "learning_rate": 0.00010521732390037053, + "loss": 0.7647, + "step": 28758 + }, + { + "epoch": 1.9956283394629102, + "grad_norm": 1.3033431768417358, + "learning_rate": 0.00010521171292894711, + "loss": 0.4933, + "step": 28759 + }, + { + "epoch": 1.995697730900007, + "grad_norm": 1.5791455507278442, + "learning_rate": 0.00010520610194107096, + "loss": 0.4239, + "step": 28760 + }, + { + "epoch": 1.9957671223371036, + "grad_norm": 1.6621959209442139, + "learning_rate": 0.0001052004909367597, + "loss": 0.6678, + "step": 28761 + }, + { + "epoch": 1.9958365137742002, + "grad_norm": 2.132086753845215, + "learning_rate": 0.00010519487991603112, + "loss": 0.7457, + "step": 28762 + }, + { + "epoch": 1.995905905211297, + "grad_norm": 1.9792035818099976, + "learning_rate": 0.0001051892688789029, + "loss": 0.4872, + "step": 28763 + }, + { + "epoch": 1.9959752966483935, + "grad_norm": 1.334894061088562, + "learning_rate": 0.00010518365782539275, + "loss": 0.4601, + "step": 28764 + }, + { + "epoch": 1.9960446880854903, + "grad_norm": 1.08052396774292, + "learning_rate": 0.00010517804675551842, + "loss": 0.321, + "step": 28765 + }, + { + "epoch": 1.996114079522587, + "grad_norm": 1.521896481513977, + "learning_rate": 0.00010517243566929757, + "loss": 0.4056, + "step": 28766 + }, + { + "epoch": 1.9961834709596835, + "grad_norm": 1.7054873704910278, + "learning_rate": 0.00010516682456674796, + "loss": 0.5572, + "step": 28767 + }, + { + "epoch": 1.9962528623967803, + "grad_norm": 1.2104543447494507, + "learning_rate": 0.00010516121344788727, + "loss": 0.4778, + "step": 28768 + }, + { + "epoch": 1.9963222538338767, + "grad_norm": 1.3496711254119873, + "learning_rate": 0.00010515560231273322, + "loss": 0.4688, + "step": 28769 + }, + { + "epoch": 1.9963916452709736, + "grad_norm": 1.4748705625534058, + "learning_rate": 0.00010514999116130358, + "loss": 0.6486, + "step": 28770 + }, + { + "epoch": 1.9964610367080702, + "grad_norm": 2.016430377960205, + "learning_rate": 0.00010514437999361595, + "loss": 0.3942, + "step": 28771 + }, + { + "epoch": 1.9965304281451668, + "grad_norm": 2.075594663619995, + "learning_rate": 0.00010513876880968813, + "loss": 0.625, + "step": 28772 + }, + { + "epoch": 1.9965998195822636, + "grad_norm": 1.3651362657546997, + "learning_rate": 0.00010513315760953783, + "loss": 0.5388, + "step": 28773 + }, + { + "epoch": 1.9966692110193602, + "grad_norm": 1.2860136032104492, + "learning_rate": 0.00010512754639318275, + "loss": 0.5838, + "step": 28774 + }, + { + "epoch": 1.9967386024564568, + "grad_norm": 1.0614304542541504, + "learning_rate": 0.00010512193516064057, + "loss": 0.4426, + "step": 28775 + }, + { + "epoch": 1.9968079938935537, + "grad_norm": 1.8222553730010986, + "learning_rate": 0.00010511632391192905, + "loss": 0.7977, + "step": 28776 + }, + { + "epoch": 1.99687738533065, + "grad_norm": 1.5834853649139404, + "learning_rate": 0.00010511071264706591, + "loss": 0.498, + "step": 28777 + }, + { + "epoch": 1.996946776767747, + "grad_norm": 1.5770257711410522, + "learning_rate": 0.00010510510136606881, + "loss": 0.5556, + "step": 28778 + }, + { + "epoch": 1.9970161682048435, + "grad_norm": 1.2589472532272339, + "learning_rate": 0.00010509949006895551, + "loss": 0.5121, + "step": 28779 + }, + { + "epoch": 1.9970855596419401, + "grad_norm": 1.5925301313400269, + "learning_rate": 0.00010509387875574374, + "loss": 0.5306, + "step": 28780 + }, + { + "epoch": 1.997154951079037, + "grad_norm": 1.280758023262024, + "learning_rate": 0.00010508826742645115, + "loss": 0.4135, + "step": 28781 + }, + { + "epoch": 1.9972243425161333, + "grad_norm": 1.698853850364685, + "learning_rate": 0.00010508265608109553, + "loss": 0.5068, + "step": 28782 + }, + { + "epoch": 1.9972937339532302, + "grad_norm": 1.179672122001648, + "learning_rate": 0.00010507704471969451, + "loss": 0.5001, + "step": 28783 + }, + { + "epoch": 1.9973631253903268, + "grad_norm": 1.207527995109558, + "learning_rate": 0.00010507143334226588, + "loss": 0.4479, + "step": 28784 + }, + { + "epoch": 1.9974325168274234, + "grad_norm": 1.381813645362854, + "learning_rate": 0.0001050658219488273, + "loss": 0.468, + "step": 28785 + }, + { + "epoch": 1.9975019082645202, + "grad_norm": 3.793532609939575, + "learning_rate": 0.00010506021053939649, + "loss": 0.8222, + "step": 28786 + }, + { + "epoch": 1.9975712997016168, + "grad_norm": 1.6643985509872437, + "learning_rate": 0.00010505459911399122, + "loss": 0.5765, + "step": 28787 + }, + { + "epoch": 1.9976406911387135, + "grad_norm": 1.4922146797180176, + "learning_rate": 0.00010504898767262916, + "loss": 0.5498, + "step": 28788 + }, + { + "epoch": 1.9977100825758103, + "grad_norm": 1.66867995262146, + "learning_rate": 0.00010504337621532803, + "loss": 0.4573, + "step": 28789 + }, + { + "epoch": 1.9977794740129067, + "grad_norm": 1.4389921426773071, + "learning_rate": 0.00010503776474210554, + "loss": 0.6122, + "step": 28790 + }, + { + "epoch": 1.9978488654500035, + "grad_norm": 1.6462424993515015, + "learning_rate": 0.00010503215325297942, + "loss": 0.6158, + "step": 28791 + }, + { + "epoch": 1.9979182568871001, + "grad_norm": 1.6425843238830566, + "learning_rate": 0.00010502654174796735, + "loss": 0.4611, + "step": 28792 + }, + { + "epoch": 1.9979876483241967, + "grad_norm": 1.4698121547698975, + "learning_rate": 0.0001050209302270871, + "loss": 0.6144, + "step": 28793 + }, + { + "epoch": 1.9980570397612936, + "grad_norm": 2.756589651107788, + "learning_rate": 0.00010501531869035637, + "loss": 0.6443, + "step": 28794 + }, + { + "epoch": 1.9981264311983902, + "grad_norm": 1.472197413444519, + "learning_rate": 0.00010500970713779282, + "loss": 0.6916, + "step": 28795 + }, + { + "epoch": 1.9981958226354868, + "grad_norm": 2.994056224822998, + "learning_rate": 0.00010500409556941421, + "loss": 0.5398, + "step": 28796 + }, + { + "epoch": 1.9982652140725834, + "grad_norm": 1.3785362243652344, + "learning_rate": 0.00010499848398523827, + "loss": 0.4958, + "step": 28797 + }, + { + "epoch": 1.99833460550968, + "grad_norm": 1.2795307636260986, + "learning_rate": 0.0001049928723852827, + "loss": 0.6924, + "step": 28798 + }, + { + "epoch": 1.9984039969467768, + "grad_norm": 1.8309577703475952, + "learning_rate": 0.0001049872607695652, + "loss": 0.6913, + "step": 28799 + }, + { + "epoch": 1.9984733883838735, + "grad_norm": 1.924533724784851, + "learning_rate": 0.00010498164913810352, + "loss": 0.6558, + "step": 28800 + }, + { + "epoch": 1.99854277982097, + "grad_norm": 2.7632648944854736, + "learning_rate": 0.00010497603749091534, + "loss": 0.4643, + "step": 28801 + }, + { + "epoch": 1.998612171258067, + "grad_norm": 2.9997122287750244, + "learning_rate": 0.00010497042582801837, + "loss": 0.5476, + "step": 28802 + }, + { + "epoch": 1.9986815626951633, + "grad_norm": 1.795455813407898, + "learning_rate": 0.00010496481414943033, + "loss": 0.7167, + "step": 28803 + }, + { + "epoch": 1.9987509541322601, + "grad_norm": 2.0721914768218994, + "learning_rate": 0.00010495920245516897, + "loss": 0.4909, + "step": 28804 + }, + { + "epoch": 1.9988203455693567, + "grad_norm": 1.304821252822876, + "learning_rate": 0.000104953590745252, + "loss": 0.4841, + "step": 28805 + }, + { + "epoch": 1.9988897370064533, + "grad_norm": 1.328071117401123, + "learning_rate": 0.00010494797901969708, + "loss": 0.4892, + "step": 28806 + }, + { + "epoch": 1.9989591284435502, + "grad_norm": 1.5557645559310913, + "learning_rate": 0.00010494236727852198, + "loss": 0.5843, + "step": 28807 + }, + { + "epoch": 1.9990285198806468, + "grad_norm": 1.0541276931762695, + "learning_rate": 0.00010493675552174442, + "loss": 0.4287, + "step": 28808 + }, + { + "epoch": 1.9990979113177434, + "grad_norm": 1.305662751197815, + "learning_rate": 0.00010493114374938207, + "loss": 0.6318, + "step": 28809 + }, + { + "epoch": 1.99916730275484, + "grad_norm": 1.6533631086349487, + "learning_rate": 0.00010492553196145267, + "loss": 0.7295, + "step": 28810 + }, + { + "epoch": 1.9992366941919366, + "grad_norm": 1.6809892654418945, + "learning_rate": 0.00010491992015797393, + "loss": 0.6977, + "step": 28811 + }, + { + "epoch": 1.9993060856290334, + "grad_norm": 1.4048376083374023, + "learning_rate": 0.0001049143083389636, + "loss": 0.5245, + "step": 28812 + }, + { + "epoch": 1.99937547706613, + "grad_norm": 1.6611186265945435, + "learning_rate": 0.00010490869650443935, + "loss": 0.5483, + "step": 28813 + }, + { + "epoch": 1.9994448685032267, + "grad_norm": 1.5951131582260132, + "learning_rate": 0.00010490308465441894, + "loss": 0.4524, + "step": 28814 + }, + { + "epoch": 1.9995142599403235, + "grad_norm": 1.367564082145691, + "learning_rate": 0.00010489747278892004, + "loss": 0.4963, + "step": 28815 + }, + { + "epoch": 1.99958365137742, + "grad_norm": 2.3562216758728027, + "learning_rate": 0.0001048918609079604, + "loss": 0.4295, + "step": 28816 + }, + { + "epoch": 1.9996530428145167, + "grad_norm": 1.7068079710006714, + "learning_rate": 0.00010488624901155772, + "loss": 0.4458, + "step": 28817 + }, + { + "epoch": 1.9997224342516133, + "grad_norm": 1.2482243776321411, + "learning_rate": 0.0001048806370997297, + "loss": 0.4926, + "step": 28818 + }, + { + "epoch": 1.99979182568871, + "grad_norm": 1.068454623222351, + "learning_rate": 0.0001048750251724941, + "loss": 0.4712, + "step": 28819 + }, + { + "epoch": 1.9998612171258068, + "grad_norm": 1.569770336151123, + "learning_rate": 0.00010486941322986858, + "loss": 0.6732, + "step": 28820 + }, + { + "epoch": 1.9999306085629034, + "grad_norm": 1.3513970375061035, + "learning_rate": 0.00010486380127187092, + "loss": 0.501, + "step": 28821 + }, + { + "epoch": 2.0, + "grad_norm": 1.321709394454956, + "learning_rate": 0.0001048581892985188, + "loss": 0.568, + "step": 28822 + }, + { + "epoch": 2.000069391437097, + "grad_norm": 1.4868488311767578, + "learning_rate": 0.00010485257730982991, + "loss": 0.3809, + "step": 28823 + }, + { + "epoch": 2.000138782874193, + "grad_norm": 1.1138412952423096, + "learning_rate": 0.000104846965305822, + "loss": 0.4608, + "step": 28824 + }, + { + "epoch": 2.00020817431129, + "grad_norm": 1.582426905632019, + "learning_rate": 0.0001048413532865128, + "loss": 0.3292, + "step": 28825 + }, + { + "epoch": 2.0002775657483864, + "grad_norm": 1.3628352880477905, + "learning_rate": 0.00010483574125192004, + "loss": 0.4188, + "step": 28826 + }, + { + "epoch": 2.0003469571854833, + "grad_norm": 1.2047468423843384, + "learning_rate": 0.00010483012920206135, + "loss": 0.3795, + "step": 28827 + }, + { + "epoch": 2.00041634862258, + "grad_norm": 1.824715256690979, + "learning_rate": 0.00010482451713695451, + "loss": 0.5063, + "step": 28828 + }, + { + "epoch": 2.0004857400596765, + "grad_norm": 1.2264853715896606, + "learning_rate": 0.00010481890505661727, + "loss": 0.5031, + "step": 28829 + }, + { + "epoch": 2.0005551314967733, + "grad_norm": 0.924426257610321, + "learning_rate": 0.00010481329296106727, + "loss": 0.4008, + "step": 28830 + }, + { + "epoch": 2.00062452293387, + "grad_norm": 1.3168601989746094, + "learning_rate": 0.00010480768085032226, + "loss": 0.3414, + "step": 28831 + }, + { + "epoch": 2.0006939143709666, + "grad_norm": 1.4673974514007568, + "learning_rate": 0.0001048020687244, + "loss": 0.5211, + "step": 28832 + }, + { + "epoch": 2.0007633058080634, + "grad_norm": 1.3745251893997192, + "learning_rate": 0.00010479645658331814, + "loss": 0.4977, + "step": 28833 + }, + { + "epoch": 2.0008326972451598, + "grad_norm": 1.3470697402954102, + "learning_rate": 0.00010479084442709444, + "loss": 0.4721, + "step": 28834 + }, + { + "epoch": 2.0009020886822566, + "grad_norm": 1.624735713005066, + "learning_rate": 0.00010478523225574657, + "loss": 0.5669, + "step": 28835 + }, + { + "epoch": 2.0009714801193534, + "grad_norm": 1.3935812711715698, + "learning_rate": 0.00010477962006929229, + "loss": 0.3831, + "step": 28836 + }, + { + "epoch": 2.00104087155645, + "grad_norm": 1.7471579313278198, + "learning_rate": 0.00010477400786774931, + "loss": 0.7098, + "step": 28837 + }, + { + "epoch": 2.0011102629935467, + "grad_norm": 1.3758116960525513, + "learning_rate": 0.00010476839565113532, + "loss": 0.508, + "step": 28838 + }, + { + "epoch": 2.001179654430643, + "grad_norm": 1.5204707384109497, + "learning_rate": 0.00010476278341946808, + "loss": 0.5138, + "step": 28839 + }, + { + "epoch": 2.00124904586774, + "grad_norm": 2.2605857849121094, + "learning_rate": 0.00010475717117276529, + "loss": 0.4055, + "step": 28840 + }, + { + "epoch": 2.0013184373048367, + "grad_norm": 1.1889203786849976, + "learning_rate": 0.00010475155891104464, + "loss": 0.3563, + "step": 28841 + }, + { + "epoch": 2.001387828741933, + "grad_norm": 1.691994071006775, + "learning_rate": 0.00010474594663432387, + "loss": 0.5217, + "step": 28842 + }, + { + "epoch": 2.00145722017903, + "grad_norm": 0.8203446865081787, + "learning_rate": 0.00010474033434262072, + "loss": 0.304, + "step": 28843 + }, + { + "epoch": 2.0015266116161268, + "grad_norm": 1.5253639221191406, + "learning_rate": 0.00010473472203595287, + "loss": 0.3545, + "step": 28844 + }, + { + "epoch": 2.001596003053223, + "grad_norm": 1.6181780099868774, + "learning_rate": 0.00010472910971433804, + "loss": 0.4498, + "step": 28845 + }, + { + "epoch": 2.00166539449032, + "grad_norm": 1.9419314861297607, + "learning_rate": 0.00010472349737779399, + "loss": 0.4472, + "step": 28846 + }, + { + "epoch": 2.0017347859274164, + "grad_norm": 1.3158948421478271, + "learning_rate": 0.00010471788502633837, + "loss": 0.5547, + "step": 28847 + }, + { + "epoch": 2.001804177364513, + "grad_norm": 0.8892105221748352, + "learning_rate": 0.00010471227265998896, + "loss": 0.3424, + "step": 28848 + }, + { + "epoch": 2.00187356880161, + "grad_norm": 1.3059606552124023, + "learning_rate": 0.00010470666027876344, + "loss": 0.6165, + "step": 28849 + }, + { + "epoch": 2.0019429602387064, + "grad_norm": 1.19566011428833, + "learning_rate": 0.00010470104788267956, + "loss": 0.3955, + "step": 28850 + }, + { + "epoch": 2.0020123516758033, + "grad_norm": 1.0628693103790283, + "learning_rate": 0.000104695435471755, + "loss": 0.4617, + "step": 28851 + }, + { + "epoch": 2.0020817431128997, + "grad_norm": 1.8479043245315552, + "learning_rate": 0.0001046898230460075, + "loss": 0.4343, + "step": 28852 + }, + { + "epoch": 2.0021511345499965, + "grad_norm": 1.5240387916564941, + "learning_rate": 0.00010468421060545476, + "loss": 0.5563, + "step": 28853 + }, + { + "epoch": 2.0022205259870933, + "grad_norm": 1.1721299886703491, + "learning_rate": 0.00010467859815011454, + "loss": 0.4473, + "step": 28854 + }, + { + "epoch": 2.0022899174241897, + "grad_norm": 1.8794927597045898, + "learning_rate": 0.0001046729856800045, + "loss": 0.4612, + "step": 28855 + }, + { + "epoch": 2.0023593088612865, + "grad_norm": 1.0862653255462646, + "learning_rate": 0.0001046673731951424, + "loss": 0.3655, + "step": 28856 + }, + { + "epoch": 2.0024287002983834, + "grad_norm": 1.0261024236679077, + "learning_rate": 0.00010466176069554593, + "loss": 0.3831, + "step": 28857 + }, + { + "epoch": 2.0024980917354798, + "grad_norm": 0.9922997951507568, + "learning_rate": 0.00010465614818123284, + "loss": 0.2773, + "step": 28858 + }, + { + "epoch": 2.0025674831725766, + "grad_norm": 1.4932862520217896, + "learning_rate": 0.00010465053565222083, + "loss": 0.4561, + "step": 28859 + }, + { + "epoch": 2.002636874609673, + "grad_norm": 1.3025431632995605, + "learning_rate": 0.00010464492310852763, + "loss": 0.5776, + "step": 28860 + }, + { + "epoch": 2.00270626604677, + "grad_norm": 1.115041971206665, + "learning_rate": 0.0001046393105501709, + "loss": 0.2689, + "step": 28861 + }, + { + "epoch": 2.0027756574838667, + "grad_norm": 1.303514838218689, + "learning_rate": 0.00010463369797716844, + "loss": 0.5806, + "step": 28862 + }, + { + "epoch": 2.002845048920963, + "grad_norm": 1.7186760902404785, + "learning_rate": 0.00010462808538953792, + "loss": 0.494, + "step": 28863 + }, + { + "epoch": 2.00291444035806, + "grad_norm": 1.5190035104751587, + "learning_rate": 0.0001046224727872971, + "loss": 0.5157, + "step": 28864 + }, + { + "epoch": 2.0029838317951563, + "grad_norm": 1.721813678741455, + "learning_rate": 0.00010461686017046361, + "loss": 0.3302, + "step": 28865 + }, + { + "epoch": 2.003053223232253, + "grad_norm": 1.227729320526123, + "learning_rate": 0.00010461124753905527, + "loss": 0.569, + "step": 28866 + }, + { + "epoch": 2.00312261466935, + "grad_norm": 1.6291753053665161, + "learning_rate": 0.00010460563489308977, + "loss": 0.4398, + "step": 28867 + }, + { + "epoch": 2.0031920061064463, + "grad_norm": 1.2160645723342896, + "learning_rate": 0.00010460002223258479, + "loss": 0.3136, + "step": 28868 + }, + { + "epoch": 2.003261397543543, + "grad_norm": 1.436157464981079, + "learning_rate": 0.00010459440955755808, + "loss": 0.4533, + "step": 28869 + }, + { + "epoch": 2.00333078898064, + "grad_norm": 2.278916358947754, + "learning_rate": 0.00010458879686802736, + "loss": 0.5608, + "step": 28870 + }, + { + "epoch": 2.0034001804177364, + "grad_norm": 1.5068448781967163, + "learning_rate": 0.00010458318416401033, + "loss": 0.3941, + "step": 28871 + }, + { + "epoch": 2.003469571854833, + "grad_norm": 1.219077229499817, + "learning_rate": 0.00010457757144552474, + "loss": 0.359, + "step": 28872 + }, + { + "epoch": 2.0035389632919296, + "grad_norm": 1.9217016696929932, + "learning_rate": 0.00010457195871258825, + "loss": 0.6491, + "step": 28873 + }, + { + "epoch": 2.0036083547290264, + "grad_norm": 1.235845685005188, + "learning_rate": 0.00010456634596521866, + "loss": 0.4341, + "step": 28874 + }, + { + "epoch": 2.0036777461661233, + "grad_norm": 1.1391935348510742, + "learning_rate": 0.0001045607332034336, + "loss": 0.3769, + "step": 28875 + }, + { + "epoch": 2.0037471376032197, + "grad_norm": 1.2487049102783203, + "learning_rate": 0.00010455512042725086, + "loss": 0.4865, + "step": 28876 + }, + { + "epoch": 2.0038165290403165, + "grad_norm": 1.3913065195083618, + "learning_rate": 0.00010454950763668814, + "loss": 0.4288, + "step": 28877 + }, + { + "epoch": 2.003885920477413, + "grad_norm": 0.7932944893836975, + "learning_rate": 0.00010454389483176315, + "loss": 0.2441, + "step": 28878 + }, + { + "epoch": 2.0039553119145097, + "grad_norm": 1.4085693359375, + "learning_rate": 0.00010453828201249358, + "loss": 0.5946, + "step": 28879 + }, + { + "epoch": 2.0040247033516065, + "grad_norm": 1.0906342267990112, + "learning_rate": 0.00010453266917889721, + "loss": 0.4168, + "step": 28880 + }, + { + "epoch": 2.004094094788703, + "grad_norm": 1.0125484466552734, + "learning_rate": 0.00010452705633099173, + "loss": 0.3743, + "step": 28881 + }, + { + "epoch": 2.0041634862257998, + "grad_norm": 1.625593662261963, + "learning_rate": 0.00010452144346879485, + "loss": 0.4972, + "step": 28882 + }, + { + "epoch": 2.0042328776628966, + "grad_norm": 1.2872525453567505, + "learning_rate": 0.00010451583059232429, + "loss": 0.4786, + "step": 28883 + }, + { + "epoch": 2.004302269099993, + "grad_norm": 1.778637409210205, + "learning_rate": 0.00010451021770159781, + "loss": 0.4849, + "step": 28884 + }, + { + "epoch": 2.00437166053709, + "grad_norm": 1.243674635887146, + "learning_rate": 0.00010450460479663306, + "loss": 0.4699, + "step": 28885 + }, + { + "epoch": 2.004441051974186, + "grad_norm": 1.7807725667953491, + "learning_rate": 0.0001044989918774478, + "loss": 0.6752, + "step": 28886 + }, + { + "epoch": 2.004510443411283, + "grad_norm": 1.625490427017212, + "learning_rate": 0.00010449337894405977, + "loss": 0.4876, + "step": 28887 + }, + { + "epoch": 2.00457983484838, + "grad_norm": 1.1712936162948608, + "learning_rate": 0.00010448776599648666, + "loss": 0.4206, + "step": 28888 + }, + { + "epoch": 2.0046492262854763, + "grad_norm": 1.3705837726593018, + "learning_rate": 0.00010448215303474618, + "loss": 0.3817, + "step": 28889 + }, + { + "epoch": 2.004718617722573, + "grad_norm": 0.9435433149337769, + "learning_rate": 0.00010447654005885605, + "loss": 0.4707, + "step": 28890 + }, + { + "epoch": 2.0047880091596695, + "grad_norm": 1.5212494134902954, + "learning_rate": 0.00010447092706883403, + "loss": 0.4576, + "step": 28891 + }, + { + "epoch": 2.0048574005967663, + "grad_norm": 1.244970679283142, + "learning_rate": 0.00010446531406469779, + "loss": 0.4414, + "step": 28892 + }, + { + "epoch": 2.004926792033863, + "grad_norm": 1.5355304479599, + "learning_rate": 0.00010445970104646507, + "loss": 0.4614, + "step": 28893 + }, + { + "epoch": 2.0049961834709595, + "grad_norm": 0.9762346744537354, + "learning_rate": 0.0001044540880141536, + "loss": 0.3051, + "step": 28894 + }, + { + "epoch": 2.0050655749080564, + "grad_norm": 1.048706293106079, + "learning_rate": 0.0001044484749677811, + "loss": 0.3269, + "step": 28895 + }, + { + "epoch": 2.005134966345153, + "grad_norm": 1.410786509513855, + "learning_rate": 0.00010444286190736525, + "loss": 0.4076, + "step": 28896 + }, + { + "epoch": 2.0052043577822496, + "grad_norm": 1.9898185729980469, + "learning_rate": 0.00010443724883292383, + "loss": 0.5763, + "step": 28897 + }, + { + "epoch": 2.0052737492193464, + "grad_norm": 1.4030863046646118, + "learning_rate": 0.00010443163574447452, + "loss": 0.4913, + "step": 28898 + }, + { + "epoch": 2.005343140656443, + "grad_norm": 1.245185136795044, + "learning_rate": 0.00010442602264203505, + "loss": 0.451, + "step": 28899 + }, + { + "epoch": 2.0054125320935396, + "grad_norm": 0.9568861126899719, + "learning_rate": 0.00010442040952562314, + "loss": 0.3944, + "step": 28900 + }, + { + "epoch": 2.0054819235306365, + "grad_norm": 0.9689924120903015, + "learning_rate": 0.00010441479639525651, + "loss": 0.3233, + "step": 28901 + }, + { + "epoch": 2.005551314967733, + "grad_norm": 0.9289880394935608, + "learning_rate": 0.00010440918325095287, + "loss": 0.2942, + "step": 28902 + }, + { + "epoch": 2.0056207064048297, + "grad_norm": 1.021094560623169, + "learning_rate": 0.00010440357009272996, + "loss": 0.3324, + "step": 28903 + }, + { + "epoch": 2.005690097841926, + "grad_norm": 2.009598970413208, + "learning_rate": 0.00010439795692060552, + "loss": 0.7375, + "step": 28904 + }, + { + "epoch": 2.005759489279023, + "grad_norm": 1.2458248138427734, + "learning_rate": 0.00010439234373459721, + "loss": 0.3877, + "step": 28905 + }, + { + "epoch": 2.0058288807161198, + "grad_norm": 0.8576017022132874, + "learning_rate": 0.00010438673053472279, + "loss": 0.248, + "step": 28906 + }, + { + "epoch": 2.005898272153216, + "grad_norm": 1.1253336668014526, + "learning_rate": 0.00010438111732099995, + "loss": 0.3595, + "step": 28907 + }, + { + "epoch": 2.005967663590313, + "grad_norm": 1.45293128490448, + "learning_rate": 0.00010437550409344647, + "loss": 0.4392, + "step": 28908 + }, + { + "epoch": 2.00603705502741, + "grad_norm": 1.4820809364318848, + "learning_rate": 0.00010436989085208, + "loss": 0.4636, + "step": 28909 + }, + { + "epoch": 2.006106446464506, + "grad_norm": 2.119270086288452, + "learning_rate": 0.00010436427759691829, + "loss": 0.4044, + "step": 28910 + }, + { + "epoch": 2.006175837901603, + "grad_norm": 1.0244144201278687, + "learning_rate": 0.00010435866432797905, + "loss": 0.3297, + "step": 28911 + }, + { + "epoch": 2.0062452293386994, + "grad_norm": 1.5681287050247192, + "learning_rate": 0.00010435305104528006, + "loss": 0.3754, + "step": 28912 + }, + { + "epoch": 2.0063146207757963, + "grad_norm": 1.0713874101638794, + "learning_rate": 0.00010434743774883895, + "loss": 0.3668, + "step": 28913 + }, + { + "epoch": 2.006384012212893, + "grad_norm": 2.1269843578338623, + "learning_rate": 0.00010434182443867348, + "loss": 0.5234, + "step": 28914 + }, + { + "epoch": 2.0064534036499895, + "grad_norm": 1.1166961193084717, + "learning_rate": 0.00010433621111480143, + "loss": 0.3357, + "step": 28915 + }, + { + "epoch": 2.0065227950870863, + "grad_norm": 1.2665969133377075, + "learning_rate": 0.00010433059777724041, + "loss": 0.4163, + "step": 28916 + }, + { + "epoch": 2.0065921865241827, + "grad_norm": 0.9627382755279541, + "learning_rate": 0.0001043249844260082, + "loss": 0.2949, + "step": 28917 + }, + { + "epoch": 2.0066615779612795, + "grad_norm": 1.2545244693756104, + "learning_rate": 0.00010431937106112253, + "loss": 0.3762, + "step": 28918 + }, + { + "epoch": 2.0067309693983764, + "grad_norm": 1.3429956436157227, + "learning_rate": 0.00010431375768260112, + "loss": 0.4964, + "step": 28919 + }, + { + "epoch": 2.0068003608354728, + "grad_norm": 1.5575850009918213, + "learning_rate": 0.00010430814429046165, + "loss": 0.4437, + "step": 28920 + }, + { + "epoch": 2.0068697522725696, + "grad_norm": 1.1069179773330688, + "learning_rate": 0.00010430253088472186, + "loss": 0.3456, + "step": 28921 + }, + { + "epoch": 2.0069391437096664, + "grad_norm": 0.710713803768158, + "learning_rate": 0.00010429691746539953, + "loss": 0.3221, + "step": 28922 + }, + { + "epoch": 2.007008535146763, + "grad_norm": 1.5519781112670898, + "learning_rate": 0.00010429130403251231, + "loss": 0.5138, + "step": 28923 + }, + { + "epoch": 2.0070779265838596, + "grad_norm": 1.2946796417236328, + "learning_rate": 0.00010428569058607793, + "loss": 0.4873, + "step": 28924 + }, + { + "epoch": 2.007147318020956, + "grad_norm": 1.44992196559906, + "learning_rate": 0.00010428007712611411, + "loss": 0.4843, + "step": 28925 + }, + { + "epoch": 2.007216709458053, + "grad_norm": 1.6927852630615234, + "learning_rate": 0.00010427446365263862, + "loss": 0.5229, + "step": 28926 + }, + { + "epoch": 2.0072861008951497, + "grad_norm": 0.9838063716888428, + "learning_rate": 0.00010426885016566909, + "loss": 0.3695, + "step": 28927 + }, + { + "epoch": 2.007355492332246, + "grad_norm": 1.9896701574325562, + "learning_rate": 0.00010426323666522334, + "loss": 0.5393, + "step": 28928 + }, + { + "epoch": 2.007424883769343, + "grad_norm": 1.3886117935180664, + "learning_rate": 0.00010425762315131904, + "loss": 0.3919, + "step": 28929 + }, + { + "epoch": 2.0074942752064393, + "grad_norm": 1.0396872758865356, + "learning_rate": 0.0001042520096239739, + "loss": 0.4245, + "step": 28930 + }, + { + "epoch": 2.007563666643536, + "grad_norm": 1.3165291547775269, + "learning_rate": 0.00010424639608320565, + "loss": 0.4191, + "step": 28931 + }, + { + "epoch": 2.007633058080633, + "grad_norm": 1.552569031715393, + "learning_rate": 0.00010424078252903204, + "loss": 0.4429, + "step": 28932 + }, + { + "epoch": 2.0077024495177294, + "grad_norm": 1.1979174613952637, + "learning_rate": 0.00010423516896147078, + "loss": 0.4484, + "step": 28933 + }, + { + "epoch": 2.007771840954826, + "grad_norm": 1.8449934720993042, + "learning_rate": 0.00010422955538053958, + "loss": 0.47, + "step": 28934 + }, + { + "epoch": 2.007841232391923, + "grad_norm": 3.328707218170166, + "learning_rate": 0.00010422394178625614, + "loss": 0.6315, + "step": 28935 + }, + { + "epoch": 2.0079106238290194, + "grad_norm": 1.1166136264801025, + "learning_rate": 0.00010421832817863823, + "loss": 0.2708, + "step": 28936 + }, + { + "epoch": 2.0079800152661162, + "grad_norm": 1.6685047149658203, + "learning_rate": 0.00010421271455770355, + "loss": 0.5156, + "step": 28937 + }, + { + "epoch": 2.0080494067032126, + "grad_norm": 1.1528247594833374, + "learning_rate": 0.00010420710092346979, + "loss": 0.4332, + "step": 28938 + }, + { + "epoch": 2.0081187981403095, + "grad_norm": 1.3251302242279053, + "learning_rate": 0.00010420148727595473, + "loss": 0.4948, + "step": 28939 + }, + { + "epoch": 2.0081881895774063, + "grad_norm": 1.1563537120819092, + "learning_rate": 0.00010419587361517606, + "loss": 0.3235, + "step": 28940 + }, + { + "epoch": 2.0082575810145027, + "grad_norm": 1.0434038639068604, + "learning_rate": 0.0001041902599411515, + "loss": 0.3878, + "step": 28941 + }, + { + "epoch": 2.0083269724515995, + "grad_norm": 1.6952924728393555, + "learning_rate": 0.00010418464625389876, + "loss": 0.5011, + "step": 28942 + }, + { + "epoch": 2.0083963638886964, + "grad_norm": 1.5986778736114502, + "learning_rate": 0.00010417903255343562, + "loss": 0.3955, + "step": 28943 + }, + { + "epoch": 2.0084657553257927, + "grad_norm": 1.8765207529067993, + "learning_rate": 0.0001041734188397797, + "loss": 0.5399, + "step": 28944 + }, + { + "epoch": 2.0085351467628896, + "grad_norm": 1.3282077312469482, + "learning_rate": 0.00010416780511294881, + "loss": 0.3955, + "step": 28945 + }, + { + "epoch": 2.008604538199986, + "grad_norm": 1.4041697978973389, + "learning_rate": 0.00010416219137296066, + "loss": 0.427, + "step": 28946 + }, + { + "epoch": 2.008673929637083, + "grad_norm": 1.3516706228256226, + "learning_rate": 0.00010415657761983293, + "loss": 0.4233, + "step": 28947 + }, + { + "epoch": 2.0087433210741796, + "grad_norm": 1.3162469863891602, + "learning_rate": 0.00010415096385358336, + "loss": 0.3783, + "step": 28948 + }, + { + "epoch": 2.008812712511276, + "grad_norm": 1.4136677980422974, + "learning_rate": 0.0001041453500742297, + "loss": 0.6223, + "step": 28949 + }, + { + "epoch": 2.008882103948373, + "grad_norm": 1.4009853601455688, + "learning_rate": 0.00010413973628178966, + "loss": 0.3254, + "step": 28950 + }, + { + "epoch": 2.0089514953854692, + "grad_norm": 1.3690147399902344, + "learning_rate": 0.00010413412247628091, + "loss": 0.4011, + "step": 28951 + }, + { + "epoch": 2.009020886822566, + "grad_norm": 1.026168704032898, + "learning_rate": 0.00010412850865772124, + "loss": 0.3214, + "step": 28952 + }, + { + "epoch": 2.009090278259663, + "grad_norm": 1.483748435974121, + "learning_rate": 0.00010412289482612836, + "loss": 0.5234, + "step": 28953 + }, + { + "epoch": 2.0091596696967593, + "grad_norm": 0.9279558658599854, + "learning_rate": 0.00010411728098151996, + "loss": 0.3832, + "step": 28954 + }, + { + "epoch": 2.009229061133856, + "grad_norm": 1.8377714157104492, + "learning_rate": 0.00010411166712391378, + "loss": 0.5701, + "step": 28955 + }, + { + "epoch": 2.009298452570953, + "grad_norm": 1.3573322296142578, + "learning_rate": 0.00010410605325332757, + "loss": 0.3809, + "step": 28956 + }, + { + "epoch": 2.0093678440080494, + "grad_norm": 1.6893450021743774, + "learning_rate": 0.000104100439369779, + "loss": 0.5729, + "step": 28957 + }, + { + "epoch": 2.009437235445146, + "grad_norm": 1.5681074857711792, + "learning_rate": 0.00010409482547328585, + "loss": 0.3662, + "step": 28958 + }, + { + "epoch": 2.0095066268822426, + "grad_norm": 1.0715930461883545, + "learning_rate": 0.00010408921156386578, + "loss": 0.2815, + "step": 28959 + }, + { + "epoch": 2.0095760183193394, + "grad_norm": 1.4562032222747803, + "learning_rate": 0.00010408359764153657, + "loss": 0.4162, + "step": 28960 + }, + { + "epoch": 2.0096454097564362, + "grad_norm": 0.9524075388908386, + "learning_rate": 0.0001040779837063159, + "loss": 0.3802, + "step": 28961 + }, + { + "epoch": 2.0097148011935326, + "grad_norm": 1.3965694904327393, + "learning_rate": 0.00010407236975822148, + "loss": 0.4149, + "step": 28962 + }, + { + "epoch": 2.0097841926306295, + "grad_norm": 0.8522751331329346, + "learning_rate": 0.00010406675579727107, + "loss": 0.3391, + "step": 28963 + }, + { + "epoch": 2.009853584067726, + "grad_norm": 1.3007701635360718, + "learning_rate": 0.00010406114182348243, + "loss": 0.3994, + "step": 28964 + }, + { + "epoch": 2.0099229755048227, + "grad_norm": 1.1621744632720947, + "learning_rate": 0.0001040555278368732, + "loss": 0.2739, + "step": 28965 + }, + { + "epoch": 2.0099923669419195, + "grad_norm": 1.5184173583984375, + "learning_rate": 0.00010404991383746114, + "loss": 0.3419, + "step": 28966 + }, + { + "epoch": 2.010061758379016, + "grad_norm": 2.9443910121917725, + "learning_rate": 0.000104044299825264, + "loss": 0.427, + "step": 28967 + }, + { + "epoch": 2.0101311498161127, + "grad_norm": 1.251407504081726, + "learning_rate": 0.00010403868580029945, + "loss": 0.3432, + "step": 28968 + }, + { + "epoch": 2.0102005412532096, + "grad_norm": 1.8042653799057007, + "learning_rate": 0.00010403307176258524, + "loss": 0.4676, + "step": 28969 + }, + { + "epoch": 2.010269932690306, + "grad_norm": 1.1019480228424072, + "learning_rate": 0.00010402745771213906, + "loss": 0.3273, + "step": 28970 + }, + { + "epoch": 2.010339324127403, + "grad_norm": 1.5974787473678589, + "learning_rate": 0.0001040218436489787, + "loss": 0.6026, + "step": 28971 + }, + { + "epoch": 2.010408715564499, + "grad_norm": 1.4901567697525024, + "learning_rate": 0.00010401622957312185, + "loss": 0.6141, + "step": 28972 + }, + { + "epoch": 2.010478107001596, + "grad_norm": 1.0609912872314453, + "learning_rate": 0.00010401061548458622, + "loss": 0.4031, + "step": 28973 + }, + { + "epoch": 2.010547498438693, + "grad_norm": 1.2251336574554443, + "learning_rate": 0.00010400500138338956, + "loss": 0.2697, + "step": 28974 + }, + { + "epoch": 2.0106168898757892, + "grad_norm": 1.0741498470306396, + "learning_rate": 0.00010399938726954954, + "loss": 0.3197, + "step": 28975 + }, + { + "epoch": 2.010686281312886, + "grad_norm": 1.0841056108474731, + "learning_rate": 0.00010399377314308396, + "loss": 0.3828, + "step": 28976 + }, + { + "epoch": 2.0107556727499825, + "grad_norm": 1.3708850145339966, + "learning_rate": 0.00010398815900401046, + "loss": 0.5695, + "step": 28977 + }, + { + "epoch": 2.0108250641870793, + "grad_norm": 1.6342991590499878, + "learning_rate": 0.00010398254485234684, + "loss": 0.4482, + "step": 28978 + }, + { + "epoch": 2.010894455624176, + "grad_norm": 0.8842546343803406, + "learning_rate": 0.00010397693068811076, + "loss": 0.3585, + "step": 28979 + }, + { + "epoch": 2.0109638470612725, + "grad_norm": 1.83793306350708, + "learning_rate": 0.00010397131651132, + "loss": 0.5131, + "step": 28980 + }, + { + "epoch": 2.0110332384983693, + "grad_norm": 1.879325270652771, + "learning_rate": 0.00010396570232199223, + "loss": 0.4191, + "step": 28981 + }, + { + "epoch": 2.011102629935466, + "grad_norm": 1.28347909450531, + "learning_rate": 0.00010396008812014519, + "loss": 0.4439, + "step": 28982 + }, + { + "epoch": 2.0111720213725626, + "grad_norm": 1.5476524829864502, + "learning_rate": 0.00010395447390579662, + "loss": 0.4472, + "step": 28983 + }, + { + "epoch": 2.0112414128096594, + "grad_norm": 1.1753010749816895, + "learning_rate": 0.00010394885967896426, + "loss": 0.2818, + "step": 28984 + }, + { + "epoch": 2.011310804246756, + "grad_norm": 1.4334442615509033, + "learning_rate": 0.00010394324543966579, + "loss": 0.3183, + "step": 28985 + }, + { + "epoch": 2.0113801956838526, + "grad_norm": 1.4114245176315308, + "learning_rate": 0.00010393763118791895, + "loss": 0.5204, + "step": 28986 + }, + { + "epoch": 2.0114495871209495, + "grad_norm": 1.1538121700286865, + "learning_rate": 0.00010393201692374144, + "loss": 0.4475, + "step": 28987 + }, + { + "epoch": 2.011518978558046, + "grad_norm": 1.5049455165863037, + "learning_rate": 0.00010392640264715106, + "loss": 0.4034, + "step": 28988 + }, + { + "epoch": 2.0115883699951427, + "grad_norm": 2.138376474380493, + "learning_rate": 0.00010392078835816546, + "loss": 0.4617, + "step": 28989 + }, + { + "epoch": 2.011657761432239, + "grad_norm": 1.373773217201233, + "learning_rate": 0.00010391517405680236, + "loss": 0.4125, + "step": 28990 + }, + { + "epoch": 2.011727152869336, + "grad_norm": 1.2625269889831543, + "learning_rate": 0.00010390955974307955, + "loss": 0.5212, + "step": 28991 + }, + { + "epoch": 2.0117965443064327, + "grad_norm": 0.9533843398094177, + "learning_rate": 0.00010390394541701469, + "loss": 0.3684, + "step": 28992 + }, + { + "epoch": 2.011865935743529, + "grad_norm": 1.4717408418655396, + "learning_rate": 0.00010389833107862555, + "loss": 0.4672, + "step": 28993 + }, + { + "epoch": 2.011935327180626, + "grad_norm": 1.5703620910644531, + "learning_rate": 0.0001038927167279298, + "loss": 0.4211, + "step": 28994 + }, + { + "epoch": 2.012004718617723, + "grad_norm": 1.3123432397842407, + "learning_rate": 0.00010388710236494523, + "loss": 0.4999, + "step": 28995 + }, + { + "epoch": 2.012074110054819, + "grad_norm": 1.7041230201721191, + "learning_rate": 0.00010388148798968949, + "loss": 0.5017, + "step": 28996 + }, + { + "epoch": 2.012143501491916, + "grad_norm": 1.614388346672058, + "learning_rate": 0.00010387587360218036, + "loss": 0.5559, + "step": 28997 + }, + { + "epoch": 2.0122128929290124, + "grad_norm": 1.6287351846694946, + "learning_rate": 0.00010387025920243556, + "loss": 0.3903, + "step": 28998 + }, + { + "epoch": 2.0122822843661092, + "grad_norm": 1.2412244081497192, + "learning_rate": 0.00010386464479047277, + "loss": 0.4329, + "step": 28999 + }, + { + "epoch": 2.012351675803206, + "grad_norm": 2.351236343383789, + "learning_rate": 0.00010385903036630978, + "loss": 0.3194, + "step": 29000 + }, + { + "epoch": 2.0124210672403025, + "grad_norm": 1.387172818183899, + "learning_rate": 0.00010385341592996425, + "loss": 0.3718, + "step": 29001 + }, + { + "epoch": 2.0124904586773993, + "grad_norm": 2.21818470954895, + "learning_rate": 0.00010384780148145394, + "loss": 0.4964, + "step": 29002 + }, + { + "epoch": 2.0125598501144957, + "grad_norm": 1.1485159397125244, + "learning_rate": 0.00010384218702079658, + "loss": 0.5519, + "step": 29003 + }, + { + "epoch": 2.0126292415515925, + "grad_norm": 1.1400463581085205, + "learning_rate": 0.00010383657254800984, + "loss": 0.4082, + "step": 29004 + }, + { + "epoch": 2.0126986329886893, + "grad_norm": 1.779792308807373, + "learning_rate": 0.00010383095806311157, + "loss": 0.5663, + "step": 29005 + }, + { + "epoch": 2.0127680244257857, + "grad_norm": 1.4452334642410278, + "learning_rate": 0.00010382534356611932, + "loss": 0.5259, + "step": 29006 + }, + { + "epoch": 2.0128374158628826, + "grad_norm": 1.2566906213760376, + "learning_rate": 0.00010381972905705095, + "loss": 0.4097, + "step": 29007 + }, + { + "epoch": 2.0129068072999794, + "grad_norm": 1.0077308416366577, + "learning_rate": 0.00010381411453592411, + "loss": 0.3117, + "step": 29008 + }, + { + "epoch": 2.012976198737076, + "grad_norm": 1.0326534509658813, + "learning_rate": 0.0001038085000027566, + "loss": 0.372, + "step": 29009 + }, + { + "epoch": 2.0130455901741726, + "grad_norm": 1.4354829788208008, + "learning_rate": 0.0001038028854575661, + "loss": 0.4112, + "step": 29010 + }, + { + "epoch": 2.013114981611269, + "grad_norm": 1.5145183801651, + "learning_rate": 0.00010379727090037027, + "loss": 0.4152, + "step": 29011 + }, + { + "epoch": 2.013184373048366, + "grad_norm": 1.5749680995941162, + "learning_rate": 0.00010379165633118695, + "loss": 0.5976, + "step": 29012 + }, + { + "epoch": 2.0132537644854627, + "grad_norm": 1.5019314289093018, + "learning_rate": 0.0001037860417500338, + "loss": 0.4237, + "step": 29013 + }, + { + "epoch": 2.013323155922559, + "grad_norm": 1.405910611152649, + "learning_rate": 0.00010378042715692853, + "loss": 0.4071, + "step": 29014 + }, + { + "epoch": 2.013392547359656, + "grad_norm": 1.4204180240631104, + "learning_rate": 0.0001037748125518889, + "loss": 0.3778, + "step": 29015 + }, + { + "epoch": 2.0134619387967523, + "grad_norm": 1.1813420057296753, + "learning_rate": 0.00010376919793493265, + "loss": 0.3209, + "step": 29016 + }, + { + "epoch": 2.013531330233849, + "grad_norm": 1.1680750846862793, + "learning_rate": 0.00010376358330607745, + "loss": 0.4413, + "step": 29017 + }, + { + "epoch": 2.013600721670946, + "grad_norm": 1.677817463874817, + "learning_rate": 0.00010375796866534104, + "loss": 0.5388, + "step": 29018 + }, + { + "epoch": 2.0136701131080423, + "grad_norm": 1.0773849487304688, + "learning_rate": 0.00010375235401274122, + "loss": 0.3077, + "step": 29019 + }, + { + "epoch": 2.013739504545139, + "grad_norm": 1.2687867879867554, + "learning_rate": 0.00010374673934829559, + "loss": 0.4304, + "step": 29020 + }, + { + "epoch": 2.013808895982236, + "grad_norm": 1.5952004194259644, + "learning_rate": 0.00010374112467202197, + "loss": 0.4203, + "step": 29021 + }, + { + "epoch": 2.0138782874193324, + "grad_norm": 0.9540979862213135, + "learning_rate": 0.00010373550998393806, + "loss": 0.4143, + "step": 29022 + }, + { + "epoch": 2.0139476788564292, + "grad_norm": 1.4233078956604004, + "learning_rate": 0.00010372989528406156, + "loss": 0.4953, + "step": 29023 + }, + { + "epoch": 2.0140170702935256, + "grad_norm": 1.348086953163147, + "learning_rate": 0.00010372428057241022, + "loss": 0.5959, + "step": 29024 + }, + { + "epoch": 2.0140864617306224, + "grad_norm": 1.178460955619812, + "learning_rate": 0.00010371866584900173, + "loss": 0.4141, + "step": 29025 + }, + { + "epoch": 2.0141558531677193, + "grad_norm": 1.6398320198059082, + "learning_rate": 0.0001037130511138539, + "loss": 0.59, + "step": 29026 + }, + { + "epoch": 2.0142252446048157, + "grad_norm": 1.0768522024154663, + "learning_rate": 0.00010370743636698436, + "loss": 0.487, + "step": 29027 + }, + { + "epoch": 2.0142946360419125, + "grad_norm": 1.6099591255187988, + "learning_rate": 0.0001037018216084109, + "loss": 0.3668, + "step": 29028 + }, + { + "epoch": 2.014364027479009, + "grad_norm": 1.2259994745254517, + "learning_rate": 0.0001036962068381512, + "loss": 0.4638, + "step": 29029 + }, + { + "epoch": 2.0144334189161057, + "grad_norm": 1.3821649551391602, + "learning_rate": 0.00010369059205622302, + "loss": 0.5049, + "step": 29030 + }, + { + "epoch": 2.0145028103532026, + "grad_norm": 0.977729856967926, + "learning_rate": 0.00010368497726264405, + "loss": 0.3224, + "step": 29031 + }, + { + "epoch": 2.014572201790299, + "grad_norm": 1.476975440979004, + "learning_rate": 0.00010367936245743204, + "loss": 0.493, + "step": 29032 + }, + { + "epoch": 2.0146415932273958, + "grad_norm": 0.8857207298278809, + "learning_rate": 0.00010367374764060471, + "loss": 0.3198, + "step": 29033 + }, + { + "epoch": 2.0147109846644926, + "grad_norm": 1.279450535774231, + "learning_rate": 0.0001036681328121798, + "loss": 0.3981, + "step": 29034 + }, + { + "epoch": 2.014780376101589, + "grad_norm": 1.5159258842468262, + "learning_rate": 0.00010366251797217498, + "loss": 0.4622, + "step": 29035 + }, + { + "epoch": 2.014849767538686, + "grad_norm": 0.8902837634086609, + "learning_rate": 0.00010365690312060807, + "loss": 0.3502, + "step": 29036 + }, + { + "epoch": 2.014919158975782, + "grad_norm": 1.36227548122406, + "learning_rate": 0.0001036512882574967, + "loss": 0.5552, + "step": 29037 + }, + { + "epoch": 2.014988550412879, + "grad_norm": 1.353514552116394, + "learning_rate": 0.00010364567338285864, + "loss": 0.4324, + "step": 29038 + }, + { + "epoch": 2.015057941849976, + "grad_norm": 1.038263201713562, + "learning_rate": 0.00010364005849671162, + "loss": 0.4679, + "step": 29039 + }, + { + "epoch": 2.0151273332870723, + "grad_norm": 1.2234004735946655, + "learning_rate": 0.00010363444359907337, + "loss": 0.4231, + "step": 29040 + }, + { + "epoch": 2.015196724724169, + "grad_norm": 1.1590269804000854, + "learning_rate": 0.00010362882868996157, + "loss": 0.4239, + "step": 29041 + }, + { + "epoch": 2.0152661161612655, + "grad_norm": 1.844512939453125, + "learning_rate": 0.00010362321376939401, + "loss": 0.5882, + "step": 29042 + }, + { + "epoch": 2.0153355075983623, + "grad_norm": 1.1207106113433838, + "learning_rate": 0.00010361759883738838, + "loss": 0.4439, + "step": 29043 + }, + { + "epoch": 2.015404899035459, + "grad_norm": 1.4768526554107666, + "learning_rate": 0.0001036119838939624, + "loss": 0.6879, + "step": 29044 + }, + { + "epoch": 2.0154742904725556, + "grad_norm": 1.4139167070388794, + "learning_rate": 0.00010360636893913382, + "loss": 0.2768, + "step": 29045 + }, + { + "epoch": 2.0155436819096524, + "grad_norm": 1.7328537702560425, + "learning_rate": 0.00010360075397292035, + "loss": 0.467, + "step": 29046 + }, + { + "epoch": 2.015613073346749, + "grad_norm": 1.4231997728347778, + "learning_rate": 0.00010359513899533972, + "loss": 0.4518, + "step": 29047 + }, + { + "epoch": 2.0156824647838456, + "grad_norm": 1.0892480611801147, + "learning_rate": 0.00010358952400640965, + "loss": 0.4399, + "step": 29048 + }, + { + "epoch": 2.0157518562209424, + "grad_norm": 2.801530599594116, + "learning_rate": 0.00010358390900614785, + "loss": 0.486, + "step": 29049 + }, + { + "epoch": 2.015821247658039, + "grad_norm": 1.0601727962493896, + "learning_rate": 0.00010357829399457209, + "loss": 0.2884, + "step": 29050 + }, + { + "epoch": 2.0158906390951357, + "grad_norm": 1.075250267982483, + "learning_rate": 0.00010357267897170005, + "loss": 0.539, + "step": 29051 + }, + { + "epoch": 2.0159600305322325, + "grad_norm": 1.4566644430160522, + "learning_rate": 0.00010356706393754948, + "loss": 0.4248, + "step": 29052 + }, + { + "epoch": 2.016029421969329, + "grad_norm": 1.0858076810836792, + "learning_rate": 0.00010356144889213811, + "loss": 0.2981, + "step": 29053 + }, + { + "epoch": 2.0160988134064257, + "grad_norm": 1.6503373384475708, + "learning_rate": 0.00010355583383548366, + "loss": 0.4354, + "step": 29054 + }, + { + "epoch": 2.016168204843522, + "grad_norm": 1.5388318300247192, + "learning_rate": 0.00010355021876760384, + "loss": 0.3124, + "step": 29055 + }, + { + "epoch": 2.016237596280619, + "grad_norm": 1.3908981084823608, + "learning_rate": 0.0001035446036885164, + "loss": 0.4698, + "step": 29056 + }, + { + "epoch": 2.0163069877177158, + "grad_norm": 2.001164674758911, + "learning_rate": 0.00010353898859823908, + "loss": 0.5486, + "step": 29057 + }, + { + "epoch": 2.016376379154812, + "grad_norm": 1.5204068422317505, + "learning_rate": 0.00010353337349678955, + "loss": 0.4919, + "step": 29058 + }, + { + "epoch": 2.016445770591909, + "grad_norm": 1.0485870838165283, + "learning_rate": 0.00010352775838418558, + "loss": 0.4262, + "step": 29059 + }, + { + "epoch": 2.016515162029006, + "grad_norm": 0.9817293882369995, + "learning_rate": 0.00010352214326044492, + "loss": 0.3425, + "step": 29060 + }, + { + "epoch": 2.016584553466102, + "grad_norm": 1.004327654838562, + "learning_rate": 0.0001035165281255852, + "loss": 0.4235, + "step": 29061 + }, + { + "epoch": 2.016653944903199, + "grad_norm": 1.1236191987991333, + "learning_rate": 0.00010351091297962425, + "loss": 0.3948, + "step": 29062 + }, + { + "epoch": 2.0167233363402954, + "grad_norm": 1.0277307033538818, + "learning_rate": 0.00010350529782257975, + "loss": 0.4339, + "step": 29063 + }, + { + "epoch": 2.0167927277773923, + "grad_norm": 1.690388560295105, + "learning_rate": 0.00010349968265446943, + "loss": 0.3729, + "step": 29064 + }, + { + "epoch": 2.016862119214489, + "grad_norm": 1.3542991876602173, + "learning_rate": 0.00010349406747531103, + "loss": 0.4764, + "step": 29065 + }, + { + "epoch": 2.0169315106515855, + "grad_norm": 1.2628250122070312, + "learning_rate": 0.00010348845228512225, + "loss": 0.3884, + "step": 29066 + }, + { + "epoch": 2.0170009020886823, + "grad_norm": 1.3890509605407715, + "learning_rate": 0.00010348283708392084, + "loss": 0.3578, + "step": 29067 + }, + { + "epoch": 2.0170702935257787, + "grad_norm": 1.1993564367294312, + "learning_rate": 0.00010347722187172451, + "loss": 0.3938, + "step": 29068 + }, + { + "epoch": 2.0171396849628755, + "grad_norm": 1.0946528911590576, + "learning_rate": 0.00010347160664855097, + "loss": 0.3603, + "step": 29069 + }, + { + "epoch": 2.0172090763999724, + "grad_norm": 1.3673986196517944, + "learning_rate": 0.000103465991414418, + "loss": 0.3939, + "step": 29070 + }, + { + "epoch": 2.0172784678370688, + "grad_norm": 1.1598533391952515, + "learning_rate": 0.0001034603761693433, + "loss": 0.5196, + "step": 29071 + }, + { + "epoch": 2.0173478592741656, + "grad_norm": 1.6337745189666748, + "learning_rate": 0.00010345476091334458, + "loss": 0.3614, + "step": 29072 + }, + { + "epoch": 2.0174172507112624, + "grad_norm": 1.1087501049041748, + "learning_rate": 0.00010344914564643958, + "loss": 0.315, + "step": 29073 + }, + { + "epoch": 2.017486642148359, + "grad_norm": 1.0985941886901855, + "learning_rate": 0.00010344353036864605, + "loss": 0.3003, + "step": 29074 + }, + { + "epoch": 2.0175560335854557, + "grad_norm": 1.3476942777633667, + "learning_rate": 0.00010343791507998166, + "loss": 0.4429, + "step": 29075 + }, + { + "epoch": 2.017625425022552, + "grad_norm": 1.2599713802337646, + "learning_rate": 0.00010343229978046419, + "loss": 0.3796, + "step": 29076 + }, + { + "epoch": 2.017694816459649, + "grad_norm": 0.9446254968643188, + "learning_rate": 0.00010342668447011132, + "loss": 0.2991, + "step": 29077 + }, + { + "epoch": 2.0177642078967457, + "grad_norm": 1.4625937938690186, + "learning_rate": 0.00010342106914894086, + "loss": 0.3022, + "step": 29078 + }, + { + "epoch": 2.017833599333842, + "grad_norm": 1.1866929531097412, + "learning_rate": 0.00010341545381697044, + "loss": 0.3768, + "step": 29079 + }, + { + "epoch": 2.017902990770939, + "grad_norm": 0.9583152532577515, + "learning_rate": 0.00010340983847421786, + "loss": 0.3401, + "step": 29080 + }, + { + "epoch": 2.0179723822080353, + "grad_norm": 1.2312397956848145, + "learning_rate": 0.00010340422312070079, + "loss": 0.4833, + "step": 29081 + }, + { + "epoch": 2.018041773645132, + "grad_norm": 1.0702085494995117, + "learning_rate": 0.00010339860775643699, + "loss": 0.4255, + "step": 29082 + }, + { + "epoch": 2.018111165082229, + "grad_norm": 1.1202783584594727, + "learning_rate": 0.00010339299238144418, + "loss": 0.2258, + "step": 29083 + }, + { + "epoch": 2.0181805565193254, + "grad_norm": 1.3303831815719604, + "learning_rate": 0.00010338737699574008, + "loss": 0.7376, + "step": 29084 + }, + { + "epoch": 2.018249947956422, + "grad_norm": 1.3083733320236206, + "learning_rate": 0.00010338176159934245, + "loss": 0.3973, + "step": 29085 + }, + { + "epoch": 2.018319339393519, + "grad_norm": 1.6957703828811646, + "learning_rate": 0.00010337614619226896, + "loss": 0.467, + "step": 29086 + }, + { + "epoch": 2.0183887308306154, + "grad_norm": 1.4556471109390259, + "learning_rate": 0.00010337053077453738, + "loss": 0.4409, + "step": 29087 + }, + { + "epoch": 2.0184581222677123, + "grad_norm": 1.2504479885101318, + "learning_rate": 0.00010336491534616544, + "loss": 0.4324, + "step": 29088 + }, + { + "epoch": 2.0185275137048087, + "grad_norm": 1.558319330215454, + "learning_rate": 0.00010335929990717082, + "loss": 0.5557, + "step": 29089 + }, + { + "epoch": 2.0185969051419055, + "grad_norm": 3.189551830291748, + "learning_rate": 0.00010335368445757128, + "loss": 0.3649, + "step": 29090 + }, + { + "epoch": 2.0186662965790023, + "grad_norm": 1.3965890407562256, + "learning_rate": 0.00010334806899738456, + "loss": 0.4506, + "step": 29091 + }, + { + "epoch": 2.0187356880160987, + "grad_norm": 1.5524593591690063, + "learning_rate": 0.00010334245352662839, + "loss": 0.4118, + "step": 29092 + }, + { + "epoch": 2.0188050794531955, + "grad_norm": 1.1687037944793701, + "learning_rate": 0.00010333683804532046, + "loss": 0.3766, + "step": 29093 + }, + { + "epoch": 2.018874470890292, + "grad_norm": 1.8143577575683594, + "learning_rate": 0.00010333122255347854, + "loss": 0.5665, + "step": 29094 + }, + { + "epoch": 2.0189438623273888, + "grad_norm": 1.2467963695526123, + "learning_rate": 0.00010332560705112033, + "loss": 0.413, + "step": 29095 + }, + { + "epoch": 2.0190132537644856, + "grad_norm": 1.5475718975067139, + "learning_rate": 0.00010331999153826355, + "loss": 0.4344, + "step": 29096 + }, + { + "epoch": 2.019082645201582, + "grad_norm": 2.5081114768981934, + "learning_rate": 0.00010331437601492595, + "loss": 0.5316, + "step": 29097 + }, + { + "epoch": 2.019152036638679, + "grad_norm": 1.2565090656280518, + "learning_rate": 0.00010330876048112528, + "loss": 0.3931, + "step": 29098 + }, + { + "epoch": 2.0192214280757756, + "grad_norm": 1.6205899715423584, + "learning_rate": 0.0001033031449368792, + "loss": 0.4605, + "step": 29099 + }, + { + "epoch": 2.019290819512872, + "grad_norm": 1.4398819208145142, + "learning_rate": 0.0001032975293822055, + "loss": 0.4132, + "step": 29100 + }, + { + "epoch": 2.019360210949969, + "grad_norm": 1.5698599815368652, + "learning_rate": 0.00010329191381712184, + "loss": 0.5869, + "step": 29101 + }, + { + "epoch": 2.0194296023870653, + "grad_norm": 1.2896865606307983, + "learning_rate": 0.00010328629824164603, + "loss": 0.615, + "step": 29102 + }, + { + "epoch": 2.019498993824162, + "grad_norm": 1.277548909187317, + "learning_rate": 0.00010328068265579573, + "loss": 0.5591, + "step": 29103 + }, + { + "epoch": 2.019568385261259, + "grad_norm": 1.1608803272247314, + "learning_rate": 0.00010327506705958872, + "loss": 0.4303, + "step": 29104 + }, + { + "epoch": 2.0196377766983553, + "grad_norm": 1.4771959781646729, + "learning_rate": 0.0001032694514530427, + "loss": 0.5, + "step": 29105 + }, + { + "epoch": 2.019707168135452, + "grad_norm": 1.3541394472122192, + "learning_rate": 0.00010326383583617538, + "loss": 0.4317, + "step": 29106 + }, + { + "epoch": 2.0197765595725485, + "grad_norm": 0.9532462358474731, + "learning_rate": 0.00010325822020900453, + "loss": 0.3963, + "step": 29107 + }, + { + "epoch": 2.0198459510096454, + "grad_norm": 1.0359376668930054, + "learning_rate": 0.00010325260457154783, + "loss": 0.3634, + "step": 29108 + }, + { + "epoch": 2.019915342446742, + "grad_norm": 1.6530894041061401, + "learning_rate": 0.00010324698892382307, + "loss": 0.5234, + "step": 29109 + }, + { + "epoch": 2.0199847338838386, + "grad_norm": 1.4057360887527466, + "learning_rate": 0.00010324137326584793, + "loss": 0.4314, + "step": 29110 + }, + { + "epoch": 2.0200541253209354, + "grad_norm": 1.1640115976333618, + "learning_rate": 0.00010323575759764015, + "loss": 0.4365, + "step": 29111 + }, + { + "epoch": 2.0201235167580323, + "grad_norm": 1.100735068321228, + "learning_rate": 0.00010323014191921745, + "loss": 0.3006, + "step": 29112 + }, + { + "epoch": 2.0201929081951286, + "grad_norm": 2.3349838256835938, + "learning_rate": 0.00010322452623059758, + "loss": 0.4973, + "step": 29113 + }, + { + "epoch": 2.0202622996322255, + "grad_norm": 1.0845221281051636, + "learning_rate": 0.00010321891053179823, + "loss": 0.3264, + "step": 29114 + }, + { + "epoch": 2.020331691069322, + "grad_norm": 1.4728893041610718, + "learning_rate": 0.0001032132948228372, + "loss": 0.365, + "step": 29115 + }, + { + "epoch": 2.0204010825064187, + "grad_norm": 1.2366955280303955, + "learning_rate": 0.00010320767910373213, + "loss": 0.6426, + "step": 29116 + }, + { + "epoch": 2.0204704739435155, + "grad_norm": 1.7466946840286255, + "learning_rate": 0.0001032020633745008, + "loss": 0.6234, + "step": 29117 + }, + { + "epoch": 2.020539865380612, + "grad_norm": 1.1207716464996338, + "learning_rate": 0.00010319644763516095, + "loss": 0.3667, + "step": 29118 + }, + { + "epoch": 2.0206092568177088, + "grad_norm": 1.5601789951324463, + "learning_rate": 0.00010319083188573027, + "loss": 0.4656, + "step": 29119 + }, + { + "epoch": 2.020678648254805, + "grad_norm": 1.4537324905395508, + "learning_rate": 0.00010318521612622649, + "loss": 0.4425, + "step": 29120 + }, + { + "epoch": 2.020748039691902, + "grad_norm": 1.1907802820205688, + "learning_rate": 0.00010317960035666734, + "loss": 0.408, + "step": 29121 + }, + { + "epoch": 2.020817431128999, + "grad_norm": 1.2037672996520996, + "learning_rate": 0.00010317398457707057, + "loss": 0.3062, + "step": 29122 + }, + { + "epoch": 2.020886822566095, + "grad_norm": 1.6487568616867065, + "learning_rate": 0.00010316836878745393, + "loss": 0.5447, + "step": 29123 + }, + { + "epoch": 2.020956214003192, + "grad_norm": 1.29842209815979, + "learning_rate": 0.00010316275298783509, + "loss": 0.3423, + "step": 29124 + }, + { + "epoch": 2.021025605440289, + "grad_norm": 1.3795524835586548, + "learning_rate": 0.00010315713717823182, + "loss": 0.3969, + "step": 29125 + }, + { + "epoch": 2.0210949968773853, + "grad_norm": 1.2024478912353516, + "learning_rate": 0.00010315152135866184, + "loss": 0.4742, + "step": 29126 + }, + { + "epoch": 2.021164388314482, + "grad_norm": 2.055111885070801, + "learning_rate": 0.00010314590552914286, + "loss": 0.3732, + "step": 29127 + }, + { + "epoch": 2.0212337797515785, + "grad_norm": 1.4351539611816406, + "learning_rate": 0.00010314028968969261, + "loss": 0.4941, + "step": 29128 + }, + { + "epoch": 2.0213031711886753, + "grad_norm": 1.1725798845291138, + "learning_rate": 0.00010313467384032882, + "loss": 0.3937, + "step": 29129 + }, + { + "epoch": 2.021372562625772, + "grad_norm": 0.7777882218360901, + "learning_rate": 0.00010312905798106927, + "loss": 0.2662, + "step": 29130 + }, + { + "epoch": 2.0214419540628685, + "grad_norm": 1.1306536197662354, + "learning_rate": 0.00010312344211193161, + "loss": 0.599, + "step": 29131 + }, + { + "epoch": 2.0215113454999654, + "grad_norm": 1.1903129816055298, + "learning_rate": 0.00010311782623293364, + "loss": 0.4285, + "step": 29132 + }, + { + "epoch": 2.0215807369370618, + "grad_norm": 1.678035855293274, + "learning_rate": 0.00010311221034409304, + "loss": 0.7113, + "step": 29133 + }, + { + "epoch": 2.0216501283741586, + "grad_norm": 1.516343593597412, + "learning_rate": 0.00010310659444542756, + "loss": 0.5074, + "step": 29134 + }, + { + "epoch": 2.0217195198112554, + "grad_norm": 1.01144540309906, + "learning_rate": 0.0001031009785369549, + "loss": 0.2999, + "step": 29135 + }, + { + "epoch": 2.021788911248352, + "grad_norm": 1.0807065963745117, + "learning_rate": 0.00010309536261869283, + "loss": 0.4308, + "step": 29136 + }, + { + "epoch": 2.0218583026854486, + "grad_norm": 1.6848621368408203, + "learning_rate": 0.00010308974669065905, + "loss": 0.6296, + "step": 29137 + }, + { + "epoch": 2.0219276941225455, + "grad_norm": 1.1355245113372803, + "learning_rate": 0.00010308413075287129, + "loss": 0.4024, + "step": 29138 + }, + { + "epoch": 2.021997085559642, + "grad_norm": 1.0517991781234741, + "learning_rate": 0.0001030785148053473, + "loss": 0.3045, + "step": 29139 + }, + { + "epoch": 2.0220664769967387, + "grad_norm": 0.9701635837554932, + "learning_rate": 0.0001030728988481048, + "loss": 0.4039, + "step": 29140 + }, + { + "epoch": 2.022135868433835, + "grad_norm": 1.0584055185317993, + "learning_rate": 0.0001030672828811615, + "loss": 0.3076, + "step": 29141 + }, + { + "epoch": 2.022205259870932, + "grad_norm": 1.0530340671539307, + "learning_rate": 0.00010306166690453513, + "loss": 0.3732, + "step": 29142 + }, + { + "epoch": 2.0222746513080287, + "grad_norm": 1.145523190498352, + "learning_rate": 0.00010305605091824348, + "loss": 0.4065, + "step": 29143 + }, + { + "epoch": 2.022344042745125, + "grad_norm": 1.543013334274292, + "learning_rate": 0.0001030504349223042, + "loss": 0.6486, + "step": 29144 + }, + { + "epoch": 2.022413434182222, + "grad_norm": 1.0473092794418335, + "learning_rate": 0.00010304481891673505, + "loss": 0.3489, + "step": 29145 + }, + { + "epoch": 2.0224828256193184, + "grad_norm": 1.5898256301879883, + "learning_rate": 0.00010303920290155375, + "loss": 0.4425, + "step": 29146 + }, + { + "epoch": 2.022552217056415, + "grad_norm": 1.1240465641021729, + "learning_rate": 0.00010303358687677807, + "loss": 0.4944, + "step": 29147 + }, + { + "epoch": 2.022621608493512, + "grad_norm": 1.2589521408081055, + "learning_rate": 0.00010302797084242569, + "loss": 0.4606, + "step": 29148 + }, + { + "epoch": 2.0226909999306084, + "grad_norm": 1.735456109046936, + "learning_rate": 0.00010302235479851434, + "loss": 0.4958, + "step": 29149 + }, + { + "epoch": 2.0227603913677052, + "grad_norm": 2.0324459075927734, + "learning_rate": 0.00010301673874506182, + "loss": 0.5065, + "step": 29150 + }, + { + "epoch": 2.022829782804802, + "grad_norm": 1.7325429916381836, + "learning_rate": 0.00010301112268208577, + "loss": 0.5353, + "step": 29151 + }, + { + "epoch": 2.0228991742418985, + "grad_norm": 1.5483322143554688, + "learning_rate": 0.00010300550660960398, + "loss": 0.3878, + "step": 29152 + }, + { + "epoch": 2.0229685656789953, + "grad_norm": 1.23611319065094, + "learning_rate": 0.0001029998905276341, + "loss": 0.3379, + "step": 29153 + }, + { + "epoch": 2.0230379571160917, + "grad_norm": 1.7924860715866089, + "learning_rate": 0.00010299427443619395, + "loss": 0.4899, + "step": 29154 + }, + { + "epoch": 2.0231073485531885, + "grad_norm": 2.5728368759155273, + "learning_rate": 0.00010298865833530123, + "loss": 0.4339, + "step": 29155 + }, + { + "epoch": 2.0231767399902854, + "grad_norm": 1.6086580753326416, + "learning_rate": 0.00010298304222497364, + "loss": 0.3806, + "step": 29156 + }, + { + "epoch": 2.0232461314273817, + "grad_norm": 1.3376681804656982, + "learning_rate": 0.00010297742610522896, + "loss": 0.2876, + "step": 29157 + }, + { + "epoch": 2.0233155228644786, + "grad_norm": 1.3092650175094604, + "learning_rate": 0.00010297180997608486, + "loss": 0.5697, + "step": 29158 + }, + { + "epoch": 2.023384914301575, + "grad_norm": 0.968689501285553, + "learning_rate": 0.00010296619383755913, + "loss": 0.3427, + "step": 29159 + }, + { + "epoch": 2.023454305738672, + "grad_norm": 1.2879482507705688, + "learning_rate": 0.00010296057768966945, + "loss": 0.3511, + "step": 29160 + }, + { + "epoch": 2.0235236971757686, + "grad_norm": 1.3038369417190552, + "learning_rate": 0.00010295496153243358, + "loss": 0.5608, + "step": 29161 + }, + { + "epoch": 2.023593088612865, + "grad_norm": 1.1698415279388428, + "learning_rate": 0.00010294934536586922, + "loss": 0.4097, + "step": 29162 + }, + { + "epoch": 2.023662480049962, + "grad_norm": 1.6950857639312744, + "learning_rate": 0.00010294372918999413, + "loss": 0.7807, + "step": 29163 + }, + { + "epoch": 2.0237318714870587, + "grad_norm": 1.4788411855697632, + "learning_rate": 0.00010293811300482605, + "loss": 0.5068, + "step": 29164 + }, + { + "epoch": 2.023801262924155, + "grad_norm": 1.1634526252746582, + "learning_rate": 0.00010293249681038267, + "loss": 0.2669, + "step": 29165 + }, + { + "epoch": 2.023870654361252, + "grad_norm": 1.4369453191757202, + "learning_rate": 0.00010292688060668171, + "loss": 0.4774, + "step": 29166 + }, + { + "epoch": 2.0239400457983483, + "grad_norm": 1.4317641258239746, + "learning_rate": 0.00010292126439374099, + "loss": 0.4366, + "step": 29167 + }, + { + "epoch": 2.024009437235445, + "grad_norm": 1.859649658203125, + "learning_rate": 0.00010291564817157812, + "loss": 0.5712, + "step": 29168 + }, + { + "epoch": 2.024078828672542, + "grad_norm": 1.6777288913726807, + "learning_rate": 0.00010291003194021094, + "loss": 0.5186, + "step": 29169 + }, + { + "epoch": 2.0241482201096384, + "grad_norm": 1.0527527332305908, + "learning_rate": 0.00010290441569965707, + "loss": 0.4101, + "step": 29170 + }, + { + "epoch": 2.024217611546735, + "grad_norm": 0.9132850766181946, + "learning_rate": 0.00010289879944993434, + "loss": 0.2526, + "step": 29171 + }, + { + "epoch": 2.024287002983832, + "grad_norm": 1.4971882104873657, + "learning_rate": 0.00010289318319106043, + "loss": 0.5097, + "step": 29172 + }, + { + "epoch": 2.0243563944209284, + "grad_norm": 1.1108537912368774, + "learning_rate": 0.00010288756692305306, + "loss": 0.3981, + "step": 29173 + }, + { + "epoch": 2.0244257858580252, + "grad_norm": 1.768435001373291, + "learning_rate": 0.00010288195064592998, + "loss": 0.4522, + "step": 29174 + }, + { + "epoch": 2.0244951772951216, + "grad_norm": 1.381365418434143, + "learning_rate": 0.00010287633435970892, + "loss": 0.4011, + "step": 29175 + }, + { + "epoch": 2.0245645687322185, + "grad_norm": 1.2422175407409668, + "learning_rate": 0.0001028707180644076, + "loss": 0.2566, + "step": 29176 + }, + { + "epoch": 2.0246339601693153, + "grad_norm": 0.8110491633415222, + "learning_rate": 0.00010286510176004374, + "loss": 0.2015, + "step": 29177 + }, + { + "epoch": 2.0247033516064117, + "grad_norm": 1.2323391437530518, + "learning_rate": 0.00010285948544663513, + "loss": 0.3917, + "step": 29178 + }, + { + "epoch": 2.0247727430435085, + "grad_norm": 0.9120848774909973, + "learning_rate": 0.00010285386912419943, + "loss": 0.3958, + "step": 29179 + }, + { + "epoch": 2.024842134480605, + "grad_norm": 0.7255436778068542, + "learning_rate": 0.00010284825279275438, + "loss": 0.2635, + "step": 29180 + }, + { + "epoch": 2.0249115259177017, + "grad_norm": 1.7871503829956055, + "learning_rate": 0.00010284263645231777, + "loss": 0.5207, + "step": 29181 + }, + { + "epoch": 2.0249809173547986, + "grad_norm": 0.8996536731719971, + "learning_rate": 0.00010283702010290726, + "loss": 0.3461, + "step": 29182 + }, + { + "epoch": 2.025050308791895, + "grad_norm": 1.2964400053024292, + "learning_rate": 0.00010283140374454059, + "loss": 0.432, + "step": 29183 + }, + { + "epoch": 2.025119700228992, + "grad_norm": 1.4167017936706543, + "learning_rate": 0.00010282578737723552, + "loss": 0.3306, + "step": 29184 + }, + { + "epoch": 2.025189091666088, + "grad_norm": 1.2142508029937744, + "learning_rate": 0.00010282017100100977, + "loss": 0.3512, + "step": 29185 + }, + { + "epoch": 2.025258483103185, + "grad_norm": 0.8915568590164185, + "learning_rate": 0.0001028145546158811, + "loss": 0.362, + "step": 29186 + }, + { + "epoch": 2.025327874540282, + "grad_norm": 1.391727089881897, + "learning_rate": 0.00010280893822186716, + "loss": 0.5332, + "step": 29187 + }, + { + "epoch": 2.0253972659773782, + "grad_norm": 1.3320517539978027, + "learning_rate": 0.00010280332181898578, + "loss": 0.432, + "step": 29188 + }, + { + "epoch": 2.025466657414475, + "grad_norm": 1.2894690036773682, + "learning_rate": 0.0001027977054072546, + "loss": 0.2817, + "step": 29189 + }, + { + "epoch": 2.025536048851572, + "grad_norm": 1.5357439517974854, + "learning_rate": 0.00010279208898669138, + "loss": 0.3535, + "step": 29190 + }, + { + "epoch": 2.0256054402886683, + "grad_norm": 1.101513385772705, + "learning_rate": 0.00010278647255731387, + "loss": 0.3087, + "step": 29191 + }, + { + "epoch": 2.025674831725765, + "grad_norm": 1.1273620128631592, + "learning_rate": 0.00010278085611913979, + "loss": 0.3968, + "step": 29192 + }, + { + "epoch": 2.0257442231628615, + "grad_norm": 1.5541645288467407, + "learning_rate": 0.00010277523967218687, + "loss": 0.5547, + "step": 29193 + }, + { + "epoch": 2.0258136145999583, + "grad_norm": 1.3485517501831055, + "learning_rate": 0.00010276962321647283, + "loss": 0.5289, + "step": 29194 + }, + { + "epoch": 2.025883006037055, + "grad_norm": 1.4463354349136353, + "learning_rate": 0.00010276400675201543, + "loss": 0.529, + "step": 29195 + }, + { + "epoch": 2.0259523974741516, + "grad_norm": 1.2955914735794067, + "learning_rate": 0.00010275839027883239, + "loss": 0.4693, + "step": 29196 + }, + { + "epoch": 2.0260217889112484, + "grad_norm": 1.3326586484909058, + "learning_rate": 0.0001027527737969414, + "loss": 0.3704, + "step": 29197 + }, + { + "epoch": 2.0260911803483452, + "grad_norm": 1.1398357152938843, + "learning_rate": 0.00010274715730636022, + "loss": 0.3576, + "step": 29198 + }, + { + "epoch": 2.0261605717854416, + "grad_norm": 1.2063533067703247, + "learning_rate": 0.00010274154080710663, + "loss": 0.6023, + "step": 29199 + }, + { + "epoch": 2.0262299632225385, + "grad_norm": 1.093679428100586, + "learning_rate": 0.00010273592429919827, + "loss": 0.3873, + "step": 29200 + }, + { + "epoch": 2.026299354659635, + "grad_norm": 1.0640969276428223, + "learning_rate": 0.00010273030778265291, + "loss": 0.3687, + "step": 29201 + }, + { + "epoch": 2.0263687460967317, + "grad_norm": 1.5016367435455322, + "learning_rate": 0.00010272469125748833, + "loss": 0.509, + "step": 29202 + }, + { + "epoch": 2.0264381375338285, + "grad_norm": 1.656153917312622, + "learning_rate": 0.00010271907472372218, + "loss": 0.3684, + "step": 29203 + }, + { + "epoch": 2.026507528970925, + "grad_norm": 1.4399986267089844, + "learning_rate": 0.00010271345818137225, + "loss": 0.6053, + "step": 29204 + }, + { + "epoch": 2.0265769204080217, + "grad_norm": 1.4431359767913818, + "learning_rate": 0.00010270784163045624, + "loss": 0.4038, + "step": 29205 + }, + { + "epoch": 2.026646311845118, + "grad_norm": 1.6564903259277344, + "learning_rate": 0.00010270222507099188, + "loss": 0.4822, + "step": 29206 + }, + { + "epoch": 2.026715703282215, + "grad_norm": 1.1097339391708374, + "learning_rate": 0.00010269660850299691, + "loss": 0.4605, + "step": 29207 + }, + { + "epoch": 2.026785094719312, + "grad_norm": 1.4942125082015991, + "learning_rate": 0.00010269099192648906, + "loss": 0.482, + "step": 29208 + }, + { + "epoch": 2.026854486156408, + "grad_norm": 1.4391835927963257, + "learning_rate": 0.00010268537534148605, + "loss": 0.3787, + "step": 29209 + }, + { + "epoch": 2.026923877593505, + "grad_norm": 0.8835957646369934, + "learning_rate": 0.00010267975874800564, + "loss": 0.3906, + "step": 29210 + }, + { + "epoch": 2.026993269030602, + "grad_norm": 1.7088837623596191, + "learning_rate": 0.00010267414214606552, + "loss": 0.4183, + "step": 29211 + }, + { + "epoch": 2.0270626604676982, + "grad_norm": 1.2422401905059814, + "learning_rate": 0.0001026685255356835, + "loss": 0.4489, + "step": 29212 + }, + { + "epoch": 2.027132051904795, + "grad_norm": 0.9437703490257263, + "learning_rate": 0.0001026629089168772, + "loss": 0.3878, + "step": 29213 + }, + { + "epoch": 2.0272014433418915, + "grad_norm": 1.4465900659561157, + "learning_rate": 0.00010265729228966441, + "loss": 0.3966, + "step": 29214 + }, + { + "epoch": 2.0272708347789883, + "grad_norm": 1.0436310768127441, + "learning_rate": 0.00010265167565406287, + "loss": 0.5201, + "step": 29215 + }, + { + "epoch": 2.027340226216085, + "grad_norm": 1.033200979232788, + "learning_rate": 0.0001026460590100903, + "loss": 0.4404, + "step": 29216 + }, + { + "epoch": 2.0274096176531815, + "grad_norm": 1.543034315109253, + "learning_rate": 0.0001026404423577644, + "loss": 0.4046, + "step": 29217 + }, + { + "epoch": 2.0274790090902783, + "grad_norm": 1.1763713359832764, + "learning_rate": 0.00010263482569710294, + "loss": 0.427, + "step": 29218 + }, + { + "epoch": 2.0275484005273747, + "grad_norm": 1.942618727684021, + "learning_rate": 0.00010262920902812368, + "loss": 0.5496, + "step": 29219 + }, + { + "epoch": 2.0276177919644716, + "grad_norm": 1.355000376701355, + "learning_rate": 0.00010262359235084426, + "loss": 0.4836, + "step": 29220 + }, + { + "epoch": 2.0276871834015684, + "grad_norm": 1.5325229167938232, + "learning_rate": 0.00010261797566528253, + "loss": 0.4525, + "step": 29221 + }, + { + "epoch": 2.027756574838665, + "grad_norm": 1.2259891033172607, + "learning_rate": 0.00010261235897145609, + "loss": 0.384, + "step": 29222 + }, + { + "epoch": 2.0278259662757616, + "grad_norm": 1.3029407262802124, + "learning_rate": 0.00010260674226938279, + "loss": 0.4399, + "step": 29223 + }, + { + "epoch": 2.0278953577128584, + "grad_norm": 1.190334439277649, + "learning_rate": 0.00010260112555908025, + "loss": 0.4068, + "step": 29224 + }, + { + "epoch": 2.027964749149955, + "grad_norm": 1.265439510345459, + "learning_rate": 0.00010259550884056628, + "loss": 0.436, + "step": 29225 + }, + { + "epoch": 2.0280341405870517, + "grad_norm": 1.781752586364746, + "learning_rate": 0.0001025898921138586, + "loss": 0.7491, + "step": 29226 + }, + { + "epoch": 2.028103532024148, + "grad_norm": 1.6037440299987793, + "learning_rate": 0.00010258427537897493, + "loss": 0.4383, + "step": 29227 + }, + { + "epoch": 2.028172923461245, + "grad_norm": 1.3001580238342285, + "learning_rate": 0.00010257865863593299, + "loss": 0.5336, + "step": 29228 + }, + { + "epoch": 2.0282423148983417, + "grad_norm": 1.2675952911376953, + "learning_rate": 0.00010257304188475052, + "loss": 0.468, + "step": 29229 + }, + { + "epoch": 2.028311706335438, + "grad_norm": 1.1194065809249878, + "learning_rate": 0.00010256742512544529, + "loss": 0.4213, + "step": 29230 + }, + { + "epoch": 2.028381097772535, + "grad_norm": 1.5592793226242065, + "learning_rate": 0.00010256180835803495, + "loss": 0.4569, + "step": 29231 + }, + { + "epoch": 2.0284504892096313, + "grad_norm": 1.1657848358154297, + "learning_rate": 0.00010255619158253731, + "loss": 0.4565, + "step": 29232 + }, + { + "epoch": 2.028519880646728, + "grad_norm": 1.4661550521850586, + "learning_rate": 0.00010255057479897008, + "loss": 0.5165, + "step": 29233 + }, + { + "epoch": 2.028589272083825, + "grad_norm": 1.7348586320877075, + "learning_rate": 0.00010254495800735095, + "loss": 0.3851, + "step": 29234 + }, + { + "epoch": 2.0286586635209214, + "grad_norm": 1.2034080028533936, + "learning_rate": 0.00010253934120769768, + "loss": 0.3507, + "step": 29235 + }, + { + "epoch": 2.0287280549580182, + "grad_norm": 1.404524326324463, + "learning_rate": 0.00010253372440002802, + "loss": 0.5408, + "step": 29236 + }, + { + "epoch": 2.028797446395115, + "grad_norm": 0.9388622641563416, + "learning_rate": 0.00010252810758435969, + "loss": 0.3186, + "step": 29237 + }, + { + "epoch": 2.0288668378322114, + "grad_norm": 1.1473339796066284, + "learning_rate": 0.00010252249076071041, + "loss": 0.3333, + "step": 29238 + }, + { + "epoch": 2.0289362292693083, + "grad_norm": 0.9475210905075073, + "learning_rate": 0.00010251687392909794, + "loss": 0.251, + "step": 29239 + }, + { + "epoch": 2.0290056207064047, + "grad_norm": 1.0816220045089722, + "learning_rate": 0.00010251125708953997, + "loss": 0.6131, + "step": 29240 + }, + { + "epoch": 2.0290750121435015, + "grad_norm": 1.4409880638122559, + "learning_rate": 0.00010250564024205427, + "loss": 0.4172, + "step": 29241 + }, + { + "epoch": 2.0291444035805983, + "grad_norm": 1.8649206161499023, + "learning_rate": 0.00010250002338665853, + "loss": 0.4852, + "step": 29242 + }, + { + "epoch": 2.0292137950176947, + "grad_norm": 1.0352833271026611, + "learning_rate": 0.00010249440652337052, + "loss": 0.4203, + "step": 29243 + }, + { + "epoch": 2.0292831864547916, + "grad_norm": 1.5336225032806396, + "learning_rate": 0.00010248878965220796, + "loss": 0.5938, + "step": 29244 + }, + { + "epoch": 2.029352577891888, + "grad_norm": 2.518153429031372, + "learning_rate": 0.00010248317277318858, + "loss": 0.5799, + "step": 29245 + }, + { + "epoch": 2.0294219693289848, + "grad_norm": 1.5245062112808228, + "learning_rate": 0.0001024775558863301, + "loss": 0.4376, + "step": 29246 + }, + { + "epoch": 2.0294913607660816, + "grad_norm": 1.1616605520248413, + "learning_rate": 0.00010247193899165028, + "loss": 0.5009, + "step": 29247 + }, + { + "epoch": 2.029560752203178, + "grad_norm": 1.4437556266784668, + "learning_rate": 0.00010246632208916682, + "loss": 0.477, + "step": 29248 + }, + { + "epoch": 2.029630143640275, + "grad_norm": 1.097146987915039, + "learning_rate": 0.00010246070517889746, + "loss": 0.4335, + "step": 29249 + }, + { + "epoch": 2.0296995350773717, + "grad_norm": 1.3202991485595703, + "learning_rate": 0.00010245508826085997, + "loss": 0.4699, + "step": 29250 + }, + { + "epoch": 2.029768926514468, + "grad_norm": 1.117323398590088, + "learning_rate": 0.00010244947133507203, + "loss": 0.4776, + "step": 29251 + }, + { + "epoch": 2.029838317951565, + "grad_norm": 1.1530309915542603, + "learning_rate": 0.00010244385440155138, + "loss": 0.3803, + "step": 29252 + }, + { + "epoch": 2.0299077093886613, + "grad_norm": 1.0811904668807983, + "learning_rate": 0.00010243823746031577, + "loss": 0.3779, + "step": 29253 + }, + { + "epoch": 2.029977100825758, + "grad_norm": 1.4701929092407227, + "learning_rate": 0.00010243262051138296, + "loss": 0.4159, + "step": 29254 + }, + { + "epoch": 2.030046492262855, + "grad_norm": 1.1516021490097046, + "learning_rate": 0.00010242700355477061, + "loss": 0.396, + "step": 29255 + }, + { + "epoch": 2.0301158836999513, + "grad_norm": 1.137852430343628, + "learning_rate": 0.00010242138659049654, + "loss": 0.4035, + "step": 29256 + }, + { + "epoch": 2.030185275137048, + "grad_norm": 1.298030972480774, + "learning_rate": 0.0001024157696185784, + "loss": 0.6177, + "step": 29257 + }, + { + "epoch": 2.0302546665741446, + "grad_norm": 1.40691077709198, + "learning_rate": 0.00010241015263903397, + "loss": 0.5359, + "step": 29258 + }, + { + "epoch": 2.0303240580112414, + "grad_norm": 1.776984691619873, + "learning_rate": 0.00010240453565188095, + "loss": 0.4197, + "step": 29259 + }, + { + "epoch": 2.030393449448338, + "grad_norm": 1.0873957872390747, + "learning_rate": 0.00010239891865713709, + "loss": 0.4701, + "step": 29260 + }, + { + "epoch": 2.0304628408854346, + "grad_norm": 1.7906602621078491, + "learning_rate": 0.00010239330165482014, + "loss": 0.4287, + "step": 29261 + }, + { + "epoch": 2.0305322323225314, + "grad_norm": 1.2582173347473145, + "learning_rate": 0.00010238768464494778, + "loss": 0.4565, + "step": 29262 + }, + { + "epoch": 2.0306016237596283, + "grad_norm": 1.0294495820999146, + "learning_rate": 0.00010238206762753782, + "loss": 0.3406, + "step": 29263 + }, + { + "epoch": 2.0306710151967247, + "grad_norm": 1.897547960281372, + "learning_rate": 0.00010237645060260792, + "loss": 0.6966, + "step": 29264 + }, + { + "epoch": 2.0307404066338215, + "grad_norm": 1.225661277770996, + "learning_rate": 0.00010237083357017585, + "loss": 0.3863, + "step": 29265 + }, + { + "epoch": 2.030809798070918, + "grad_norm": 1.2862273454666138, + "learning_rate": 0.00010236521653025933, + "loss": 0.5254, + "step": 29266 + }, + { + "epoch": 2.0308791895080147, + "grad_norm": 1.346550703048706, + "learning_rate": 0.00010235959948287607, + "loss": 0.4679, + "step": 29267 + }, + { + "epoch": 2.0309485809451115, + "grad_norm": 1.1170400381088257, + "learning_rate": 0.00010235398242804388, + "loss": 0.378, + "step": 29268 + }, + { + "epoch": 2.031017972382208, + "grad_norm": 1.031684398651123, + "learning_rate": 0.00010234836536578043, + "loss": 0.444, + "step": 29269 + }, + { + "epoch": 2.0310873638193048, + "grad_norm": 1.8636538982391357, + "learning_rate": 0.00010234274829610344, + "loss": 0.6393, + "step": 29270 + }, + { + "epoch": 2.031156755256401, + "grad_norm": 1.1147551536560059, + "learning_rate": 0.00010233713121903068, + "loss": 0.333, + "step": 29271 + }, + { + "epoch": 2.031226146693498, + "grad_norm": 1.4738037586212158, + "learning_rate": 0.00010233151413457986, + "loss": 0.6788, + "step": 29272 + }, + { + "epoch": 2.031295538130595, + "grad_norm": 1.2655508518218994, + "learning_rate": 0.00010232589704276871, + "loss": 0.376, + "step": 29273 + }, + { + "epoch": 2.031364929567691, + "grad_norm": 1.1708194017410278, + "learning_rate": 0.00010232027994361501, + "loss": 0.5072, + "step": 29274 + }, + { + "epoch": 2.031434321004788, + "grad_norm": 3.044995069503784, + "learning_rate": 0.00010231466283713644, + "loss": 0.6238, + "step": 29275 + }, + { + "epoch": 2.031503712441885, + "grad_norm": 1.2211576700210571, + "learning_rate": 0.00010230904572335074, + "loss": 0.3873, + "step": 29276 + }, + { + "epoch": 2.0315731038789813, + "grad_norm": 1.7956064939498901, + "learning_rate": 0.00010230342860227565, + "loss": 0.5109, + "step": 29277 + }, + { + "epoch": 2.031642495316078, + "grad_norm": 1.3706731796264648, + "learning_rate": 0.00010229781147392892, + "loss": 0.5422, + "step": 29278 + }, + { + "epoch": 2.0317118867531745, + "grad_norm": 1.320607304573059, + "learning_rate": 0.00010229219433832824, + "loss": 0.4813, + "step": 29279 + }, + { + "epoch": 2.0317812781902713, + "grad_norm": 0.8822451233863831, + "learning_rate": 0.00010228657719549137, + "loss": 0.2939, + "step": 29280 + }, + { + "epoch": 2.031850669627368, + "grad_norm": 1.203540325164795, + "learning_rate": 0.00010228096004543604, + "loss": 0.4219, + "step": 29281 + }, + { + "epoch": 2.0319200610644645, + "grad_norm": 1.2625155448913574, + "learning_rate": 0.00010227534288818001, + "loss": 0.4718, + "step": 29282 + }, + { + "epoch": 2.0319894525015614, + "grad_norm": 1.2975800037384033, + "learning_rate": 0.00010226972572374096, + "loss": 0.5331, + "step": 29283 + }, + { + "epoch": 2.0320588439386578, + "grad_norm": 1.3088170289993286, + "learning_rate": 0.00010226410855213666, + "loss": 0.6131, + "step": 29284 + }, + { + "epoch": 2.0321282353757546, + "grad_norm": 0.940000593662262, + "learning_rate": 0.00010225849137338483, + "loss": 0.2791, + "step": 29285 + }, + { + "epoch": 2.0321976268128514, + "grad_norm": 1.1847567558288574, + "learning_rate": 0.0001022528741875032, + "loss": 0.394, + "step": 29286 + }, + { + "epoch": 2.032267018249948, + "grad_norm": 1.2507060766220093, + "learning_rate": 0.00010224725699450952, + "loss": 0.3913, + "step": 29287 + }, + { + "epoch": 2.0323364096870447, + "grad_norm": 1.2311774492263794, + "learning_rate": 0.00010224163979442152, + "loss": 0.4631, + "step": 29288 + }, + { + "epoch": 2.0324058011241415, + "grad_norm": 1.1225510835647583, + "learning_rate": 0.00010223602258725688, + "loss": 0.3644, + "step": 29289 + }, + { + "epoch": 2.032475192561238, + "grad_norm": 1.5201971530914307, + "learning_rate": 0.00010223040537303341, + "loss": 0.5606, + "step": 29290 + }, + { + "epoch": 2.0325445839983347, + "grad_norm": 0.9976015090942383, + "learning_rate": 0.00010222478815176881, + "loss": 0.2737, + "step": 29291 + }, + { + "epoch": 2.032613975435431, + "grad_norm": 1.1375209093093872, + "learning_rate": 0.0001022191709234808, + "loss": 0.5662, + "step": 29292 + }, + { + "epoch": 2.032683366872528, + "grad_norm": 1.0592690706253052, + "learning_rate": 0.00010221355368818714, + "loss": 0.3893, + "step": 29293 + }, + { + "epoch": 2.0327527583096248, + "grad_norm": 1.0299572944641113, + "learning_rate": 0.00010220793644590554, + "loss": 0.4126, + "step": 29294 + }, + { + "epoch": 2.032822149746721, + "grad_norm": 1.2283458709716797, + "learning_rate": 0.00010220231919665375, + "loss": 0.5428, + "step": 29295 + }, + { + "epoch": 2.032891541183818, + "grad_norm": 1.3721345663070679, + "learning_rate": 0.00010219670194044947, + "loss": 0.5695, + "step": 29296 + }, + { + "epoch": 2.0329609326209144, + "grad_norm": 0.980994701385498, + "learning_rate": 0.00010219108467731046, + "loss": 0.3734, + "step": 29297 + }, + { + "epoch": 2.033030324058011, + "grad_norm": 2.0600600242614746, + "learning_rate": 0.00010218546740725446, + "loss": 0.5278, + "step": 29298 + }, + { + "epoch": 2.033099715495108, + "grad_norm": 1.0818769931793213, + "learning_rate": 0.0001021798501302992, + "loss": 0.4876, + "step": 29299 + }, + { + "epoch": 2.0331691069322044, + "grad_norm": 1.221487045288086, + "learning_rate": 0.0001021742328464624, + "loss": 0.4802, + "step": 29300 + }, + { + "epoch": 2.0332384983693013, + "grad_norm": 1.1908501386642456, + "learning_rate": 0.00010216861555576179, + "loss": 0.4509, + "step": 29301 + }, + { + "epoch": 2.033307889806398, + "grad_norm": 1.679365634918213, + "learning_rate": 0.00010216299825821512, + "loss": 0.5901, + "step": 29302 + }, + { + "epoch": 2.0333772812434945, + "grad_norm": 1.426599383354187, + "learning_rate": 0.0001021573809538401, + "loss": 0.5433, + "step": 29303 + }, + { + "epoch": 2.0334466726805913, + "grad_norm": 1.4455941915512085, + "learning_rate": 0.00010215176364265451, + "loss": 0.5248, + "step": 29304 + }, + { + "epoch": 2.0335160641176877, + "grad_norm": 1.2547190189361572, + "learning_rate": 0.000102146146324676, + "loss": 0.4456, + "step": 29305 + }, + { + "epoch": 2.0335854555547845, + "grad_norm": 1.685389518737793, + "learning_rate": 0.00010214052899992242, + "loss": 0.534, + "step": 29306 + }, + { + "epoch": 2.0336548469918814, + "grad_norm": 1.7466963529586792, + "learning_rate": 0.0001021349116684114, + "loss": 0.5017, + "step": 29307 + }, + { + "epoch": 2.0337242384289778, + "grad_norm": 1.6245157718658447, + "learning_rate": 0.00010212929433016075, + "loss": 0.2753, + "step": 29308 + }, + { + "epoch": 2.0337936298660746, + "grad_norm": 1.2399005889892578, + "learning_rate": 0.00010212367698518812, + "loss": 0.3663, + "step": 29309 + }, + { + "epoch": 2.033863021303171, + "grad_norm": 0.8671343326568604, + "learning_rate": 0.0001021180596335113, + "loss": 0.3685, + "step": 29310 + }, + { + "epoch": 2.033932412740268, + "grad_norm": 1.3140356540679932, + "learning_rate": 0.00010211244227514802, + "loss": 0.3521, + "step": 29311 + }, + { + "epoch": 2.0340018041773646, + "grad_norm": 1.54664146900177, + "learning_rate": 0.00010210682491011598, + "loss": 0.4659, + "step": 29312 + }, + { + "epoch": 2.034071195614461, + "grad_norm": 1.1831880807876587, + "learning_rate": 0.00010210120753843298, + "loss": 0.4469, + "step": 29313 + }, + { + "epoch": 2.034140587051558, + "grad_norm": 1.350964903831482, + "learning_rate": 0.00010209559016011668, + "loss": 0.2998, + "step": 29314 + }, + { + "epoch": 2.0342099784886547, + "grad_norm": 1.7793190479278564, + "learning_rate": 0.00010208997277518484, + "loss": 0.4922, + "step": 29315 + }, + { + "epoch": 2.034279369925751, + "grad_norm": 1.5642403364181519, + "learning_rate": 0.00010208435538365524, + "loss": 0.4434, + "step": 29316 + }, + { + "epoch": 2.034348761362848, + "grad_norm": 1.2590744495391846, + "learning_rate": 0.00010207873798554552, + "loss": 0.4248, + "step": 29317 + }, + { + "epoch": 2.0344181527999443, + "grad_norm": 1.7751706838607788, + "learning_rate": 0.00010207312058087349, + "loss": 0.5025, + "step": 29318 + }, + { + "epoch": 2.034487544237041, + "grad_norm": 1.2173597812652588, + "learning_rate": 0.00010206750316965686, + "loss": 0.2568, + "step": 29319 + }, + { + "epoch": 2.034556935674138, + "grad_norm": 1.2473139762878418, + "learning_rate": 0.00010206188575191338, + "loss": 0.4516, + "step": 29320 + }, + { + "epoch": 2.0346263271112344, + "grad_norm": 1.2900364398956299, + "learning_rate": 0.00010205626832766074, + "loss": 0.3438, + "step": 29321 + }, + { + "epoch": 2.034695718548331, + "grad_norm": 1.2927701473236084, + "learning_rate": 0.00010205065089691667, + "loss": 0.389, + "step": 29322 + }, + { + "epoch": 2.0347651099854276, + "grad_norm": 1.5925602912902832, + "learning_rate": 0.000102045033459699, + "loss": 0.5826, + "step": 29323 + }, + { + "epoch": 2.0348345014225244, + "grad_norm": 1.4103924036026, + "learning_rate": 0.00010203941601602536, + "loss": 0.4577, + "step": 29324 + }, + { + "epoch": 2.0349038928596213, + "grad_norm": 0.9751808643341064, + "learning_rate": 0.00010203379856591351, + "loss": 0.3887, + "step": 29325 + }, + { + "epoch": 2.0349732842967176, + "grad_norm": 1.2542140483856201, + "learning_rate": 0.00010202818110938124, + "loss": 0.5153, + "step": 29326 + }, + { + "epoch": 2.0350426757338145, + "grad_norm": 1.7294059991836548, + "learning_rate": 0.0001020225636464462, + "loss": 0.6381, + "step": 29327 + }, + { + "epoch": 2.0351120671709113, + "grad_norm": 1.0461504459381104, + "learning_rate": 0.00010201694617712618, + "loss": 0.3467, + "step": 29328 + }, + { + "epoch": 2.0351814586080077, + "grad_norm": 1.2292059659957886, + "learning_rate": 0.00010201132870143888, + "loss": 0.3438, + "step": 29329 + }, + { + "epoch": 2.0352508500451045, + "grad_norm": 1.5251206159591675, + "learning_rate": 0.00010200571121940205, + "loss": 0.5805, + "step": 29330 + }, + { + "epoch": 2.035320241482201, + "grad_norm": 1.2413556575775146, + "learning_rate": 0.00010200009373103343, + "loss": 0.509, + "step": 29331 + }, + { + "epoch": 2.0353896329192978, + "grad_norm": 1.1996169090270996, + "learning_rate": 0.00010199447623635073, + "loss": 0.4699, + "step": 29332 + }, + { + "epoch": 2.0354590243563946, + "grad_norm": 1.1255766153335571, + "learning_rate": 0.00010198885873537171, + "loss": 0.4029, + "step": 29333 + }, + { + "epoch": 2.035528415793491, + "grad_norm": 1.1604433059692383, + "learning_rate": 0.00010198324122811409, + "loss": 0.3909, + "step": 29334 + }, + { + "epoch": 2.035597807230588, + "grad_norm": 1.057076096534729, + "learning_rate": 0.0001019776237145956, + "loss": 0.3235, + "step": 29335 + }, + { + "epoch": 2.035667198667684, + "grad_norm": 1.3371872901916504, + "learning_rate": 0.000101972006194834, + "loss": 0.5326, + "step": 29336 + }, + { + "epoch": 2.035736590104781, + "grad_norm": 1.6834115982055664, + "learning_rate": 0.00010196638866884702, + "loss": 0.4622, + "step": 29337 + }, + { + "epoch": 2.035805981541878, + "grad_norm": 1.6119999885559082, + "learning_rate": 0.00010196077113665234, + "loss": 0.6271, + "step": 29338 + }, + { + "epoch": 2.0358753729789743, + "grad_norm": 1.0614458322525024, + "learning_rate": 0.00010195515359826775, + "loss": 0.3551, + "step": 29339 + }, + { + "epoch": 2.035944764416071, + "grad_norm": 1.7023955583572388, + "learning_rate": 0.00010194953605371098, + "loss": 0.4548, + "step": 29340 + }, + { + "epoch": 2.036014155853168, + "grad_norm": 1.1696289777755737, + "learning_rate": 0.00010194391850299974, + "loss": 0.3773, + "step": 29341 + }, + { + "epoch": 2.0360835472902643, + "grad_norm": 1.818565011024475, + "learning_rate": 0.00010193830094615176, + "loss": 0.5158, + "step": 29342 + }, + { + "epoch": 2.036152938727361, + "grad_norm": 1.122241497039795, + "learning_rate": 0.0001019326833831848, + "loss": 0.3985, + "step": 29343 + }, + { + "epoch": 2.0362223301644575, + "grad_norm": 1.2916927337646484, + "learning_rate": 0.0001019270658141166, + "loss": 0.5502, + "step": 29344 + }, + { + "epoch": 2.0362917216015544, + "grad_norm": 0.970413327217102, + "learning_rate": 0.00010192144823896485, + "loss": 0.3895, + "step": 29345 + }, + { + "epoch": 2.036361113038651, + "grad_norm": 0.8490826487541199, + "learning_rate": 0.00010191583065774733, + "loss": 0.2118, + "step": 29346 + }, + { + "epoch": 2.0364305044757476, + "grad_norm": 1.2856534719467163, + "learning_rate": 0.00010191021307048177, + "loss": 0.4416, + "step": 29347 + }, + { + "epoch": 2.0364998959128444, + "grad_norm": 1.4387937784194946, + "learning_rate": 0.00010190459547718585, + "loss": 0.3749, + "step": 29348 + }, + { + "epoch": 2.036569287349941, + "grad_norm": 1.1588387489318848, + "learning_rate": 0.00010189897787787734, + "loss": 0.3939, + "step": 29349 + }, + { + "epoch": 2.0366386787870376, + "grad_norm": 1.2982722520828247, + "learning_rate": 0.00010189336027257402, + "loss": 0.4109, + "step": 29350 + }, + { + "epoch": 2.0367080702241345, + "grad_norm": 3.196352481842041, + "learning_rate": 0.00010188774266129356, + "loss": 0.6229, + "step": 29351 + }, + { + "epoch": 2.036777461661231, + "grad_norm": 1.0136693716049194, + "learning_rate": 0.0001018821250440537, + "loss": 0.5597, + "step": 29352 + }, + { + "epoch": 2.0368468530983277, + "grad_norm": 1.3718647956848145, + "learning_rate": 0.0001018765074208722, + "loss": 0.3464, + "step": 29353 + }, + { + "epoch": 2.0369162445354245, + "grad_norm": 1.0926471948623657, + "learning_rate": 0.0001018708897917668, + "loss": 0.4501, + "step": 29354 + }, + { + "epoch": 2.036985635972521, + "grad_norm": 1.3082876205444336, + "learning_rate": 0.00010186527215675518, + "loss": 0.4385, + "step": 29355 + }, + { + "epoch": 2.0370550274096177, + "grad_norm": 1.2814264297485352, + "learning_rate": 0.00010185965451585516, + "loss": 0.5211, + "step": 29356 + }, + { + "epoch": 2.037124418846714, + "grad_norm": 1.6006489992141724, + "learning_rate": 0.0001018540368690844, + "loss": 0.4746, + "step": 29357 + }, + { + "epoch": 2.037193810283811, + "grad_norm": 1.4095746278762817, + "learning_rate": 0.00010184841921646066, + "loss": 0.4207, + "step": 29358 + }, + { + "epoch": 2.037263201720908, + "grad_norm": 1.5044043064117432, + "learning_rate": 0.00010184280155800167, + "loss": 0.4262, + "step": 29359 + }, + { + "epoch": 2.037332593158004, + "grad_norm": 1.1585181951522827, + "learning_rate": 0.00010183718389372517, + "loss": 0.3101, + "step": 29360 + }, + { + "epoch": 2.037401984595101, + "grad_norm": 1.4677942991256714, + "learning_rate": 0.00010183156622364891, + "loss": 0.4529, + "step": 29361 + }, + { + "epoch": 2.0374713760321974, + "grad_norm": 1.3693654537200928, + "learning_rate": 0.00010182594854779061, + "loss": 0.4565, + "step": 29362 + }, + { + "epoch": 2.0375407674692942, + "grad_norm": 1.7415430545806885, + "learning_rate": 0.000101820330866168, + "loss": 0.5531, + "step": 29363 + }, + { + "epoch": 2.037610158906391, + "grad_norm": 1.2411012649536133, + "learning_rate": 0.00010181471317879882, + "loss": 0.429, + "step": 29364 + }, + { + "epoch": 2.0376795503434875, + "grad_norm": 1.3353103399276733, + "learning_rate": 0.0001018090954857008, + "loss": 0.3331, + "step": 29365 + }, + { + "epoch": 2.0377489417805843, + "grad_norm": 1.2994887828826904, + "learning_rate": 0.00010180347778689164, + "loss": 0.4067, + "step": 29366 + }, + { + "epoch": 2.037818333217681, + "grad_norm": 2.1383745670318604, + "learning_rate": 0.00010179786008238914, + "loss": 0.4976, + "step": 29367 + }, + { + "epoch": 2.0378877246547775, + "grad_norm": 3.8205337524414062, + "learning_rate": 0.00010179224237221101, + "loss": 0.6139, + "step": 29368 + }, + { + "epoch": 2.0379571160918744, + "grad_norm": 1.280743956565857, + "learning_rate": 0.00010178662465637496, + "loss": 0.3496, + "step": 29369 + }, + { + "epoch": 2.0380265075289707, + "grad_norm": 1.2718582153320312, + "learning_rate": 0.00010178100693489873, + "loss": 0.3228, + "step": 29370 + }, + { + "epoch": 2.0380958989660676, + "grad_norm": 1.1995288133621216, + "learning_rate": 0.00010177538920780012, + "loss": 0.446, + "step": 29371 + }, + { + "epoch": 2.0381652904031644, + "grad_norm": 0.8854115605354309, + "learning_rate": 0.00010176977147509676, + "loss": 0.235, + "step": 29372 + }, + { + "epoch": 2.038234681840261, + "grad_norm": 1.7425967454910278, + "learning_rate": 0.00010176415373680646, + "loss": 0.4819, + "step": 29373 + }, + { + "epoch": 2.0383040732773576, + "grad_norm": 1.1237671375274658, + "learning_rate": 0.00010175853599294691, + "loss": 0.4013, + "step": 29374 + }, + { + "epoch": 2.038373464714454, + "grad_norm": 1.6556824445724487, + "learning_rate": 0.00010175291824353589, + "loss": 0.3782, + "step": 29375 + }, + { + "epoch": 2.038442856151551, + "grad_norm": 0.9964264631271362, + "learning_rate": 0.00010174730048859111, + "loss": 0.2406, + "step": 29376 + }, + { + "epoch": 2.0385122475886477, + "grad_norm": 1.471255898475647, + "learning_rate": 0.0001017416827281303, + "loss": 0.3743, + "step": 29377 + }, + { + "epoch": 2.038581639025744, + "grad_norm": 1.374603033065796, + "learning_rate": 0.00010173606496217121, + "loss": 0.3909, + "step": 29378 + }, + { + "epoch": 2.038651030462841, + "grad_norm": 1.799710988998413, + "learning_rate": 0.00010173044719073155, + "loss": 0.5375, + "step": 29379 + }, + { + "epoch": 2.0387204218999377, + "grad_norm": 1.1471177339553833, + "learning_rate": 0.0001017248294138291, + "loss": 0.3816, + "step": 29380 + }, + { + "epoch": 2.038789813337034, + "grad_norm": 1.2795430421829224, + "learning_rate": 0.00010171921163148152, + "loss": 0.4232, + "step": 29381 + }, + { + "epoch": 2.038859204774131, + "grad_norm": 1.5883612632751465, + "learning_rate": 0.00010171359384370662, + "loss": 0.558, + "step": 29382 + }, + { + "epoch": 2.0389285962112274, + "grad_norm": 1.5390870571136475, + "learning_rate": 0.0001017079760505221, + "loss": 0.5994, + "step": 29383 + }, + { + "epoch": 2.038997987648324, + "grad_norm": 1.112267017364502, + "learning_rate": 0.00010170235825194567, + "loss": 0.4441, + "step": 29384 + }, + { + "epoch": 2.039067379085421, + "grad_norm": 1.3958090543746948, + "learning_rate": 0.00010169674044799511, + "loss": 0.3665, + "step": 29385 + }, + { + "epoch": 2.0391367705225174, + "grad_norm": 1.8185945749282837, + "learning_rate": 0.00010169112263868814, + "loss": 0.3853, + "step": 29386 + }, + { + "epoch": 2.0392061619596142, + "grad_norm": 1.1075719594955444, + "learning_rate": 0.00010168550482404249, + "loss": 0.4277, + "step": 29387 + }, + { + "epoch": 2.0392755533967106, + "grad_norm": 1.6402170658111572, + "learning_rate": 0.00010167988700407587, + "loss": 0.6274, + "step": 29388 + }, + { + "epoch": 2.0393449448338075, + "grad_norm": 0.8672887682914734, + "learning_rate": 0.00010167426917880609, + "loss": 0.2846, + "step": 29389 + }, + { + "epoch": 2.0394143362709043, + "grad_norm": 1.4038796424865723, + "learning_rate": 0.00010166865134825081, + "loss": 0.4346, + "step": 29390 + }, + { + "epoch": 2.0394837277080007, + "grad_norm": 1.0268172025680542, + "learning_rate": 0.00010166303351242778, + "loss": 0.3966, + "step": 29391 + }, + { + "epoch": 2.0395531191450975, + "grad_norm": 0.9968161582946777, + "learning_rate": 0.00010165741567135478, + "loss": 0.1948, + "step": 29392 + }, + { + "epoch": 2.0396225105821943, + "grad_norm": 1.3589824438095093, + "learning_rate": 0.00010165179782504948, + "loss": 0.571, + "step": 29393 + }, + { + "epoch": 2.0396919020192907, + "grad_norm": 1.3952171802520752, + "learning_rate": 0.00010164617997352966, + "loss": 0.4054, + "step": 29394 + }, + { + "epoch": 2.0397612934563876, + "grad_norm": 1.5130387544631958, + "learning_rate": 0.00010164056211681303, + "loss": 0.6348, + "step": 29395 + }, + { + "epoch": 2.039830684893484, + "grad_norm": 1.2217319011688232, + "learning_rate": 0.00010163494425491735, + "loss": 0.4716, + "step": 29396 + }, + { + "epoch": 2.039900076330581, + "grad_norm": 1.2678109407424927, + "learning_rate": 0.00010162932638786035, + "loss": 0.5098, + "step": 29397 + }, + { + "epoch": 2.0399694677676776, + "grad_norm": 1.0817803144454956, + "learning_rate": 0.00010162370851565974, + "loss": 0.307, + "step": 29398 + }, + { + "epoch": 2.040038859204774, + "grad_norm": 1.4511094093322754, + "learning_rate": 0.00010161809063833328, + "loss": 0.6384, + "step": 29399 + }, + { + "epoch": 2.040108250641871, + "grad_norm": 1.261601209640503, + "learning_rate": 0.0001016124727558987, + "loss": 0.4007, + "step": 29400 + }, + { + "epoch": 2.0401776420789677, + "grad_norm": 1.0013433694839478, + "learning_rate": 0.0001016068548683737, + "loss": 0.3623, + "step": 29401 + }, + { + "epoch": 2.040247033516064, + "grad_norm": 1.1935023069381714, + "learning_rate": 0.00010160123697577608, + "loss": 0.3229, + "step": 29402 + }, + { + "epoch": 2.040316424953161, + "grad_norm": 1.4285050630569458, + "learning_rate": 0.00010159561907812351, + "loss": 0.3471, + "step": 29403 + }, + { + "epoch": 2.0403858163902573, + "grad_norm": 1.1548049449920654, + "learning_rate": 0.00010159000117543375, + "loss": 0.4173, + "step": 29404 + }, + { + "epoch": 2.040455207827354, + "grad_norm": 1.60816490650177, + "learning_rate": 0.00010158438326772457, + "loss": 0.6417, + "step": 29405 + }, + { + "epoch": 2.040524599264451, + "grad_norm": 2.0403518676757812, + "learning_rate": 0.00010157876535501367, + "loss": 0.489, + "step": 29406 + }, + { + "epoch": 2.0405939907015473, + "grad_norm": 1.7784825563430786, + "learning_rate": 0.00010157314743731878, + "loss": 0.4275, + "step": 29407 + }, + { + "epoch": 2.040663382138644, + "grad_norm": 1.0741220712661743, + "learning_rate": 0.00010156752951465764, + "loss": 0.3693, + "step": 29408 + }, + { + "epoch": 2.0407327735757406, + "grad_norm": 1.0744993686676025, + "learning_rate": 0.00010156191158704802, + "loss": 0.5441, + "step": 29409 + }, + { + "epoch": 2.0408021650128374, + "grad_norm": 1.1781522035598755, + "learning_rate": 0.0001015562936545076, + "loss": 0.3948, + "step": 29410 + }, + { + "epoch": 2.0408715564499342, + "grad_norm": 1.3412470817565918, + "learning_rate": 0.00010155067571705417, + "loss": 0.4568, + "step": 29411 + }, + { + "epoch": 2.0409409478870306, + "grad_norm": 1.3921451568603516, + "learning_rate": 0.0001015450577747054, + "loss": 0.6145, + "step": 29412 + }, + { + "epoch": 2.0410103393241275, + "grad_norm": 1.6726365089416504, + "learning_rate": 0.00010153943982747909, + "loss": 0.342, + "step": 29413 + }, + { + "epoch": 2.041079730761224, + "grad_norm": 1.4092130661010742, + "learning_rate": 0.00010153382187539293, + "loss": 0.6034, + "step": 29414 + }, + { + "epoch": 2.0411491221983207, + "grad_norm": 1.561846375465393, + "learning_rate": 0.0001015282039184647, + "loss": 0.4478, + "step": 29415 + }, + { + "epoch": 2.0412185136354175, + "grad_norm": 1.1843364238739014, + "learning_rate": 0.00010152258595671209, + "loss": 0.3734, + "step": 29416 + }, + { + "epoch": 2.041287905072514, + "grad_norm": 1.1982711553573608, + "learning_rate": 0.00010151696799015285, + "loss": 0.5517, + "step": 29417 + }, + { + "epoch": 2.0413572965096107, + "grad_norm": 1.1564524173736572, + "learning_rate": 0.00010151135001880472, + "loss": 0.4705, + "step": 29418 + }, + { + "epoch": 2.0414266879467076, + "grad_norm": 1.0632102489471436, + "learning_rate": 0.00010150573204268543, + "loss": 0.3443, + "step": 29419 + }, + { + "epoch": 2.041496079383804, + "grad_norm": 1.1514801979064941, + "learning_rate": 0.00010150011406181273, + "loss": 0.4868, + "step": 29420 + }, + { + "epoch": 2.041565470820901, + "grad_norm": 1.398762822151184, + "learning_rate": 0.00010149449607620431, + "loss": 0.3539, + "step": 29421 + }, + { + "epoch": 2.041634862257997, + "grad_norm": 1.3253408670425415, + "learning_rate": 0.00010148887808587796, + "loss": 0.5586, + "step": 29422 + }, + { + "epoch": 2.041704253695094, + "grad_norm": 1.0428481101989746, + "learning_rate": 0.00010148326009085143, + "loss": 0.4735, + "step": 29423 + }, + { + "epoch": 2.041773645132191, + "grad_norm": 1.2461646795272827, + "learning_rate": 0.00010147764209114237, + "loss": 0.5206, + "step": 29424 + }, + { + "epoch": 2.0418430365692872, + "grad_norm": 1.060458779335022, + "learning_rate": 0.00010147202408676857, + "loss": 0.3349, + "step": 29425 + }, + { + "epoch": 2.041912428006384, + "grad_norm": 1.144743800163269, + "learning_rate": 0.00010146640607774778, + "loss": 0.4116, + "step": 29426 + }, + { + "epoch": 2.041981819443481, + "grad_norm": 0.7429307699203491, + "learning_rate": 0.00010146078806409771, + "loss": 0.2619, + "step": 29427 + }, + { + "epoch": 2.0420512108805773, + "grad_norm": 0.8857275247573853, + "learning_rate": 0.0001014551700458361, + "loss": 0.3756, + "step": 29428 + }, + { + "epoch": 2.042120602317674, + "grad_norm": 1.1635715961456299, + "learning_rate": 0.0001014495520229807, + "loss": 0.3488, + "step": 29429 + }, + { + "epoch": 2.0421899937547705, + "grad_norm": 1.3685780763626099, + "learning_rate": 0.00010144393399554923, + "loss": 0.5646, + "step": 29430 + }, + { + "epoch": 2.0422593851918673, + "grad_norm": 1.598116159439087, + "learning_rate": 0.00010143831596355942, + "loss": 0.4575, + "step": 29431 + }, + { + "epoch": 2.042328776628964, + "grad_norm": 1.1846638917922974, + "learning_rate": 0.00010143269792702901, + "loss": 0.3999, + "step": 29432 + }, + { + "epoch": 2.0423981680660606, + "grad_norm": 1.7285680770874023, + "learning_rate": 0.00010142707988597573, + "loss": 0.6838, + "step": 29433 + }, + { + "epoch": 2.0424675595031574, + "grad_norm": 1.0171663761138916, + "learning_rate": 0.00010142146184041737, + "loss": 0.4293, + "step": 29434 + }, + { + "epoch": 2.042536950940254, + "grad_norm": 1.0161117315292358, + "learning_rate": 0.00010141584379037156, + "loss": 0.3652, + "step": 29435 + }, + { + "epoch": 2.0426063423773506, + "grad_norm": 1.9896557331085205, + "learning_rate": 0.00010141022573585612, + "loss": 0.6054, + "step": 29436 + }, + { + "epoch": 2.0426757338144474, + "grad_norm": 1.3975831270217896, + "learning_rate": 0.00010140460767688877, + "loss": 0.4994, + "step": 29437 + }, + { + "epoch": 2.042745125251544, + "grad_norm": 1.2385462522506714, + "learning_rate": 0.00010139898961348723, + "loss": 0.485, + "step": 29438 + }, + { + "epoch": 2.0428145166886407, + "grad_norm": 1.0520670413970947, + "learning_rate": 0.00010139337154566924, + "loss": 0.454, + "step": 29439 + }, + { + "epoch": 2.0428839081257375, + "grad_norm": 1.260748028755188, + "learning_rate": 0.00010138775347345255, + "loss": 0.4115, + "step": 29440 + }, + { + "epoch": 2.042953299562834, + "grad_norm": 1.230807900428772, + "learning_rate": 0.00010138213539685486, + "loss": 0.5197, + "step": 29441 + }, + { + "epoch": 2.0430226909999307, + "grad_norm": 1.5606539249420166, + "learning_rate": 0.00010137651731589392, + "loss": 0.6014, + "step": 29442 + }, + { + "epoch": 2.043092082437027, + "grad_norm": 1.187427043914795, + "learning_rate": 0.00010137089923058751, + "loss": 0.3399, + "step": 29443 + }, + { + "epoch": 2.043161473874124, + "grad_norm": 1.404617428779602, + "learning_rate": 0.00010136528114095332, + "loss": 0.5755, + "step": 29444 + }, + { + "epoch": 2.043230865311221, + "grad_norm": 1.4884623289108276, + "learning_rate": 0.00010135966304700908, + "loss": 0.6466, + "step": 29445 + }, + { + "epoch": 2.043300256748317, + "grad_norm": 1.5953408479690552, + "learning_rate": 0.00010135404494877255, + "loss": 0.372, + "step": 29446 + }, + { + "epoch": 2.043369648185414, + "grad_norm": 0.9188346266746521, + "learning_rate": 0.00010134842684626147, + "loss": 0.372, + "step": 29447 + }, + { + "epoch": 2.0434390396225104, + "grad_norm": 1.2062538862228394, + "learning_rate": 0.00010134280873949355, + "loss": 0.4241, + "step": 29448 + }, + { + "epoch": 2.0435084310596072, + "grad_norm": 1.9390779733657837, + "learning_rate": 0.00010133719062848654, + "loss": 0.6077, + "step": 29449 + }, + { + "epoch": 2.043577822496704, + "grad_norm": 1.895939588546753, + "learning_rate": 0.00010133157251325818, + "loss": 0.5332, + "step": 29450 + }, + { + "epoch": 2.0436472139338004, + "grad_norm": 1.2545217275619507, + "learning_rate": 0.00010132595439382621, + "loss": 0.4821, + "step": 29451 + }, + { + "epoch": 2.0437166053708973, + "grad_norm": 1.1343960762023926, + "learning_rate": 0.00010132033627020833, + "loss": 0.314, + "step": 29452 + }, + { + "epoch": 2.043785996807994, + "grad_norm": 1.4694498777389526, + "learning_rate": 0.00010131471814242233, + "loss": 0.4649, + "step": 29453 + }, + { + "epoch": 2.0438553882450905, + "grad_norm": 1.602099061012268, + "learning_rate": 0.0001013091000104859, + "loss": 0.5575, + "step": 29454 + }, + { + "epoch": 2.0439247796821873, + "grad_norm": 1.7264875173568726, + "learning_rate": 0.00010130348187441678, + "loss": 0.4878, + "step": 29455 + }, + { + "epoch": 2.0439941711192837, + "grad_norm": 1.1326682567596436, + "learning_rate": 0.00010129786373423274, + "loss": 0.2154, + "step": 29456 + }, + { + "epoch": 2.0440635625563806, + "grad_norm": 0.9388313293457031, + "learning_rate": 0.00010129224558995147, + "loss": 0.3341, + "step": 29457 + }, + { + "epoch": 2.0441329539934774, + "grad_norm": 1.4802488088607788, + "learning_rate": 0.00010128662744159077, + "loss": 0.5281, + "step": 29458 + }, + { + "epoch": 2.0442023454305738, + "grad_norm": 1.0331510305404663, + "learning_rate": 0.0001012810092891683, + "loss": 0.3191, + "step": 29459 + }, + { + "epoch": 2.0442717368676706, + "grad_norm": 0.8336971402168274, + "learning_rate": 0.00010127539113270185, + "loss": 0.297, + "step": 29460 + }, + { + "epoch": 2.044341128304767, + "grad_norm": 1.3506163358688354, + "learning_rate": 0.00010126977297220914, + "loss": 0.496, + "step": 29461 + }, + { + "epoch": 2.044410519741864, + "grad_norm": 2.332209587097168, + "learning_rate": 0.0001012641548077079, + "loss": 0.6943, + "step": 29462 + }, + { + "epoch": 2.0444799111789607, + "grad_norm": 1.3728656768798828, + "learning_rate": 0.00010125853663921585, + "loss": 0.5496, + "step": 29463 + }, + { + "epoch": 2.044549302616057, + "grad_norm": 1.5194777250289917, + "learning_rate": 0.00010125291846675076, + "loss": 0.525, + "step": 29464 + }, + { + "epoch": 2.044618694053154, + "grad_norm": 1.1667560338974, + "learning_rate": 0.00010124730029033038, + "loss": 0.337, + "step": 29465 + }, + { + "epoch": 2.0446880854902507, + "grad_norm": 1.3690320253372192, + "learning_rate": 0.00010124168210997238, + "loss": 0.5737, + "step": 29466 + }, + { + "epoch": 2.044757476927347, + "grad_norm": 1.3388934135437012, + "learning_rate": 0.00010123606392569458, + "loss": 0.4119, + "step": 29467 + }, + { + "epoch": 2.044826868364444, + "grad_norm": 1.1420495510101318, + "learning_rate": 0.00010123044573751461, + "loss": 0.351, + "step": 29468 + }, + { + "epoch": 2.0448962598015403, + "grad_norm": 1.0972644090652466, + "learning_rate": 0.0001012248275454503, + "loss": 0.4246, + "step": 29469 + }, + { + "epoch": 2.044965651238637, + "grad_norm": 1.0008244514465332, + "learning_rate": 0.00010121920934951936, + "loss": 0.2606, + "step": 29470 + }, + { + "epoch": 2.045035042675734, + "grad_norm": 1.384663462638855, + "learning_rate": 0.00010121359114973949, + "loss": 0.3983, + "step": 29471 + }, + { + "epoch": 2.0451044341128304, + "grad_norm": 1.1792209148406982, + "learning_rate": 0.00010120797294612847, + "loss": 0.4568, + "step": 29472 + }, + { + "epoch": 2.045173825549927, + "grad_norm": 2.4245080947875977, + "learning_rate": 0.00010120235473870401, + "loss": 0.414, + "step": 29473 + }, + { + "epoch": 2.0452432169870236, + "grad_norm": 1.5861070156097412, + "learning_rate": 0.00010119673652748387, + "loss": 0.6221, + "step": 29474 + }, + { + "epoch": 2.0453126084241204, + "grad_norm": 1.7286001443862915, + "learning_rate": 0.00010119111831248577, + "loss": 0.6241, + "step": 29475 + }, + { + "epoch": 2.0453819998612173, + "grad_norm": 2.0772838592529297, + "learning_rate": 0.00010118550009372743, + "loss": 0.5398, + "step": 29476 + }, + { + "epoch": 2.0454513912983137, + "grad_norm": 1.1594711542129517, + "learning_rate": 0.0001011798818712266, + "loss": 0.4013, + "step": 29477 + }, + { + "epoch": 2.0455207827354105, + "grad_norm": 1.4840912818908691, + "learning_rate": 0.00010117426364500107, + "loss": 0.5021, + "step": 29478 + }, + { + "epoch": 2.0455901741725073, + "grad_norm": 1.3635625839233398, + "learning_rate": 0.00010116864541506848, + "loss": 0.3595, + "step": 29479 + }, + { + "epoch": 2.0456595656096037, + "grad_norm": 1.3889118432998657, + "learning_rate": 0.00010116302718144661, + "loss": 0.5296, + "step": 29480 + }, + { + "epoch": 2.0457289570467005, + "grad_norm": 1.4577242136001587, + "learning_rate": 0.00010115740894415319, + "loss": 0.5237, + "step": 29481 + }, + { + "epoch": 2.045798348483797, + "grad_norm": 1.2220778465270996, + "learning_rate": 0.00010115179070320602, + "loss": 0.5047, + "step": 29482 + }, + { + "epoch": 2.0458677399208938, + "grad_norm": 1.3829619884490967, + "learning_rate": 0.00010114617245862272, + "loss": 0.492, + "step": 29483 + }, + { + "epoch": 2.0459371313579906, + "grad_norm": 1.1461620330810547, + "learning_rate": 0.00010114055421042112, + "loss": 0.3993, + "step": 29484 + }, + { + "epoch": 2.046006522795087, + "grad_norm": 1.3044308423995972, + "learning_rate": 0.00010113493595861891, + "loss": 0.527, + "step": 29485 + }, + { + "epoch": 2.046075914232184, + "grad_norm": 1.4011414051055908, + "learning_rate": 0.00010112931770323387, + "loss": 0.4381, + "step": 29486 + }, + { + "epoch": 2.04614530566928, + "grad_norm": 1.1703189611434937, + "learning_rate": 0.00010112369944428366, + "loss": 0.3333, + "step": 29487 + }, + { + "epoch": 2.046214697106377, + "grad_norm": 1.3092467784881592, + "learning_rate": 0.00010111808118178607, + "loss": 0.4977, + "step": 29488 + }, + { + "epoch": 2.046284088543474, + "grad_norm": 0.8831027746200562, + "learning_rate": 0.00010111246291575887, + "loss": 0.2322, + "step": 29489 + }, + { + "epoch": 2.0463534799805703, + "grad_norm": 1.3946452140808105, + "learning_rate": 0.00010110684464621968, + "loss": 0.3497, + "step": 29490 + }, + { + "epoch": 2.046422871417667, + "grad_norm": 1.1489193439483643, + "learning_rate": 0.00010110122637318636, + "loss": 0.3047, + "step": 29491 + }, + { + "epoch": 2.046492262854764, + "grad_norm": 1.1613951921463013, + "learning_rate": 0.0001010956080966766, + "loss": 0.3663, + "step": 29492 + }, + { + "epoch": 2.0465616542918603, + "grad_norm": 1.208536982536316, + "learning_rate": 0.00010108998981670812, + "loss": 0.3595, + "step": 29493 + }, + { + "epoch": 2.046631045728957, + "grad_norm": 1.1400507688522339, + "learning_rate": 0.00010108437153329866, + "loss": 0.3777, + "step": 29494 + }, + { + "epoch": 2.0467004371660535, + "grad_norm": 1.2000590562820435, + "learning_rate": 0.00010107875324646596, + "loss": 0.4611, + "step": 29495 + }, + { + "epoch": 2.0467698286031504, + "grad_norm": 1.6433485746383667, + "learning_rate": 0.00010107313495622779, + "loss": 0.4989, + "step": 29496 + }, + { + "epoch": 2.046839220040247, + "grad_norm": 1.181986689567566, + "learning_rate": 0.00010106751666260183, + "loss": 0.388, + "step": 29497 + }, + { + "epoch": 2.0469086114773436, + "grad_norm": 1.0398184061050415, + "learning_rate": 0.00010106189836560585, + "loss": 0.3848, + "step": 29498 + }, + { + "epoch": 2.0469780029144404, + "grad_norm": 1.9424917697906494, + "learning_rate": 0.00010105628006525761, + "loss": 0.4805, + "step": 29499 + }, + { + "epoch": 2.047047394351537, + "grad_norm": 2.279158592224121, + "learning_rate": 0.00010105066176157475, + "loss": 0.68, + "step": 29500 + }, + { + "epoch": 2.0471167857886337, + "grad_norm": 1.1151326894760132, + "learning_rate": 0.00010104504345457512, + "loss": 0.3741, + "step": 29501 + }, + { + "epoch": 2.0471861772257305, + "grad_norm": 1.1528394222259521, + "learning_rate": 0.00010103942514427642, + "loss": 0.3537, + "step": 29502 + }, + { + "epoch": 2.047255568662827, + "grad_norm": 1.1594394445419312, + "learning_rate": 0.00010103380683069634, + "loss": 0.3002, + "step": 29503 + }, + { + "epoch": 2.0473249600999237, + "grad_norm": 1.1774630546569824, + "learning_rate": 0.00010102818851385269, + "loss": 0.4094, + "step": 29504 + }, + { + "epoch": 2.0473943515370205, + "grad_norm": 1.9495604038238525, + "learning_rate": 0.00010102257019376316, + "loss": 0.6708, + "step": 29505 + }, + { + "epoch": 2.047463742974117, + "grad_norm": 1.3668463230133057, + "learning_rate": 0.00010101695187044548, + "loss": 0.513, + "step": 29506 + }, + { + "epoch": 2.0475331344112138, + "grad_norm": 0.941816508769989, + "learning_rate": 0.0001010113335439174, + "loss": 0.2533, + "step": 29507 + }, + { + "epoch": 2.04760252584831, + "grad_norm": 1.1001900434494019, + "learning_rate": 0.00010100571521419666, + "loss": 0.3445, + "step": 29508 + }, + { + "epoch": 2.047671917285407, + "grad_norm": 1.5378772020339966, + "learning_rate": 0.00010100009688130098, + "loss": 0.5925, + "step": 29509 + }, + { + "epoch": 2.047741308722504, + "grad_norm": 1.4102661609649658, + "learning_rate": 0.00010099447854524812, + "loss": 0.5281, + "step": 29510 + }, + { + "epoch": 2.0478107001596, + "grad_norm": 1.7146129608154297, + "learning_rate": 0.00010098886020605583, + "loss": 0.4539, + "step": 29511 + }, + { + "epoch": 2.047880091596697, + "grad_norm": 1.6543070077896118, + "learning_rate": 0.0001009832418637418, + "loss": 0.56, + "step": 29512 + }, + { + "epoch": 2.0479494830337934, + "grad_norm": 1.3795549869537354, + "learning_rate": 0.00010097762351832381, + "loss": 0.5068, + "step": 29513 + }, + { + "epoch": 2.0480188744708903, + "grad_norm": 0.9927282333374023, + "learning_rate": 0.00010097200516981952, + "loss": 0.3612, + "step": 29514 + }, + { + "epoch": 2.048088265907987, + "grad_norm": 1.8838016986846924, + "learning_rate": 0.00010096638681824677, + "loss": 0.3492, + "step": 29515 + }, + { + "epoch": 2.0481576573450835, + "grad_norm": 1.2412341833114624, + "learning_rate": 0.00010096076846362326, + "loss": 0.4326, + "step": 29516 + }, + { + "epoch": 2.0482270487821803, + "grad_norm": 1.3421502113342285, + "learning_rate": 0.00010095515010596668, + "loss": 0.4531, + "step": 29517 + }, + { + "epoch": 2.048296440219277, + "grad_norm": 1.0900564193725586, + "learning_rate": 0.00010094953174529482, + "loss": 0.2812, + "step": 29518 + }, + { + "epoch": 2.0483658316563735, + "grad_norm": 1.304819941520691, + "learning_rate": 0.0001009439133816254, + "loss": 0.5219, + "step": 29519 + }, + { + "epoch": 2.0484352230934704, + "grad_norm": 0.9922022223472595, + "learning_rate": 0.00010093829501497615, + "loss": 0.2747, + "step": 29520 + }, + { + "epoch": 2.0485046145305668, + "grad_norm": 1.7868125438690186, + "learning_rate": 0.00010093267664536484, + "loss": 0.4349, + "step": 29521 + }, + { + "epoch": 2.0485740059676636, + "grad_norm": 1.3353545665740967, + "learning_rate": 0.00010092705827280912, + "loss": 0.5583, + "step": 29522 + }, + { + "epoch": 2.0486433974047604, + "grad_norm": 1.1982409954071045, + "learning_rate": 0.00010092143989732685, + "loss": 0.4556, + "step": 29523 + }, + { + "epoch": 2.048712788841857, + "grad_norm": 1.2281908988952637, + "learning_rate": 0.00010091582151893566, + "loss": 0.3147, + "step": 29524 + }, + { + "epoch": 2.0487821802789536, + "grad_norm": 1.3976547718048096, + "learning_rate": 0.00010091020313765331, + "loss": 0.5351, + "step": 29525 + }, + { + "epoch": 2.04885157171605, + "grad_norm": 1.1130963563919067, + "learning_rate": 0.00010090458475349757, + "loss": 0.4505, + "step": 29526 + }, + { + "epoch": 2.048920963153147, + "grad_norm": 1.8051832914352417, + "learning_rate": 0.00010089896636648619, + "loss": 0.468, + "step": 29527 + }, + { + "epoch": 2.0489903545902437, + "grad_norm": 1.7358125448226929, + "learning_rate": 0.00010089334797663686, + "loss": 0.3842, + "step": 29528 + }, + { + "epoch": 2.04905974602734, + "grad_norm": 1.2589423656463623, + "learning_rate": 0.00010088772958396732, + "loss": 0.4237, + "step": 29529 + }, + { + "epoch": 2.049129137464437, + "grad_norm": 1.7954020500183105, + "learning_rate": 0.00010088211118849536, + "loss": 0.4453, + "step": 29530 + }, + { + "epoch": 2.0491985289015338, + "grad_norm": 1.2352948188781738, + "learning_rate": 0.00010087649279023863, + "loss": 0.4537, + "step": 29531 + }, + { + "epoch": 2.04926792033863, + "grad_norm": 1.1219159364700317, + "learning_rate": 0.00010087087438921492, + "loss": 0.4046, + "step": 29532 + }, + { + "epoch": 2.049337311775727, + "grad_norm": 1.3453078269958496, + "learning_rate": 0.00010086525598544198, + "loss": 0.3518, + "step": 29533 + }, + { + "epoch": 2.0494067032128234, + "grad_norm": 1.1404705047607422, + "learning_rate": 0.00010085963757893753, + "loss": 0.5001, + "step": 29534 + }, + { + "epoch": 2.04947609464992, + "grad_norm": 1.3222243785858154, + "learning_rate": 0.00010085401916971929, + "loss": 0.5225, + "step": 29535 + }, + { + "epoch": 2.049545486087017, + "grad_norm": 1.594138741493225, + "learning_rate": 0.00010084840075780501, + "loss": 0.3711, + "step": 29536 + }, + { + "epoch": 2.0496148775241134, + "grad_norm": 1.3119150400161743, + "learning_rate": 0.00010084278234321245, + "loss": 0.3458, + "step": 29537 + }, + { + "epoch": 2.0496842689612103, + "grad_norm": 1.0508975982666016, + "learning_rate": 0.0001008371639259593, + "loss": 0.408, + "step": 29538 + }, + { + "epoch": 2.0497536603983066, + "grad_norm": 4.422329902648926, + "learning_rate": 0.00010083154550606333, + "loss": 0.6891, + "step": 29539 + }, + { + "epoch": 2.0498230518354035, + "grad_norm": 1.7264950275421143, + "learning_rate": 0.00010082592708354226, + "loss": 0.4305, + "step": 29540 + }, + { + "epoch": 2.0498924432725003, + "grad_norm": 0.8943764567375183, + "learning_rate": 0.00010082030865841385, + "loss": 0.2052, + "step": 29541 + }, + { + "epoch": 2.0499618347095967, + "grad_norm": 2.4665448665618896, + "learning_rate": 0.0001008146902306958, + "loss": 0.3487, + "step": 29542 + }, + { + "epoch": 2.0500312261466935, + "grad_norm": 1.7461093664169312, + "learning_rate": 0.00010080907180040589, + "loss": 0.3811, + "step": 29543 + }, + { + "epoch": 2.0501006175837904, + "grad_norm": 1.4127923250198364, + "learning_rate": 0.00010080345336756183, + "loss": 0.4966, + "step": 29544 + }, + { + "epoch": 2.0501700090208868, + "grad_norm": 1.1560806035995483, + "learning_rate": 0.00010079783493218134, + "loss": 0.4132, + "step": 29545 + }, + { + "epoch": 2.0502394004579836, + "grad_norm": 1.0346919298171997, + "learning_rate": 0.00010079221649428218, + "loss": 0.3749, + "step": 29546 + }, + { + "epoch": 2.05030879189508, + "grad_norm": 1.240074872970581, + "learning_rate": 0.00010078659805388209, + "loss": 0.4327, + "step": 29547 + }, + { + "epoch": 2.050378183332177, + "grad_norm": 1.462221622467041, + "learning_rate": 0.00010078097961099881, + "loss": 0.3855, + "step": 29548 + }, + { + "epoch": 2.0504475747692736, + "grad_norm": 1.8513447046279907, + "learning_rate": 0.00010077536116565008, + "loss": 0.4641, + "step": 29549 + }, + { + "epoch": 2.05051696620637, + "grad_norm": 1.6209278106689453, + "learning_rate": 0.00010076974271785359, + "loss": 0.6193, + "step": 29550 + }, + { + "epoch": 2.050586357643467, + "grad_norm": 1.2200846672058105, + "learning_rate": 0.00010076412426762714, + "loss": 0.3218, + "step": 29551 + }, + { + "epoch": 2.0506557490805633, + "grad_norm": 1.518493890762329, + "learning_rate": 0.00010075850581498842, + "loss": 0.4619, + "step": 29552 + }, + { + "epoch": 2.05072514051766, + "grad_norm": 1.1854058504104614, + "learning_rate": 0.00010075288735995518, + "loss": 0.3441, + "step": 29553 + }, + { + "epoch": 2.050794531954757, + "grad_norm": 1.056356430053711, + "learning_rate": 0.0001007472689025452, + "loss": 0.3267, + "step": 29554 + }, + { + "epoch": 2.0508639233918533, + "grad_norm": 1.0456187725067139, + "learning_rate": 0.00010074165044277615, + "loss": 0.4419, + "step": 29555 + }, + { + "epoch": 2.05093331482895, + "grad_norm": 1.9065957069396973, + "learning_rate": 0.00010073603198066582, + "loss": 0.387, + "step": 29556 + }, + { + "epoch": 2.051002706266047, + "grad_norm": 1.9807287454605103, + "learning_rate": 0.0001007304135162319, + "loss": 0.3779, + "step": 29557 + }, + { + "epoch": 2.0510720977031434, + "grad_norm": 1.4799911975860596, + "learning_rate": 0.00010072479504949215, + "loss": 0.5144, + "step": 29558 + }, + { + "epoch": 2.05114148914024, + "grad_norm": 1.3607664108276367, + "learning_rate": 0.0001007191765804643, + "loss": 0.4632, + "step": 29559 + }, + { + "epoch": 2.0512108805773366, + "grad_norm": 1.3955525159835815, + "learning_rate": 0.0001007135581091661, + "loss": 0.5298, + "step": 29560 + }, + { + "epoch": 2.0512802720144334, + "grad_norm": 2.020512819290161, + "learning_rate": 0.00010070793963561528, + "loss": 0.5278, + "step": 29561 + }, + { + "epoch": 2.0513496634515302, + "grad_norm": 1.2705129384994507, + "learning_rate": 0.00010070232115982958, + "loss": 0.4213, + "step": 29562 + }, + { + "epoch": 2.0514190548886266, + "grad_norm": 1.1245403289794922, + "learning_rate": 0.00010069670268182673, + "loss": 0.311, + "step": 29563 + }, + { + "epoch": 2.0514884463257235, + "grad_norm": 1.021382451057434, + "learning_rate": 0.00010069108420162445, + "loss": 0.3754, + "step": 29564 + }, + { + "epoch": 2.05155783776282, + "grad_norm": 1.347537875175476, + "learning_rate": 0.00010068546571924053, + "loss": 0.3849, + "step": 29565 + }, + { + "epoch": 2.0516272291999167, + "grad_norm": 1.931353211402893, + "learning_rate": 0.00010067984723469266, + "loss": 0.624, + "step": 29566 + }, + { + "epoch": 2.0516966206370135, + "grad_norm": 1.2072654962539673, + "learning_rate": 0.00010067422874799859, + "loss": 0.4143, + "step": 29567 + }, + { + "epoch": 2.05176601207411, + "grad_norm": 2.005220651626587, + "learning_rate": 0.00010066861025917608, + "loss": 0.426, + "step": 29568 + }, + { + "epoch": 2.0518354035112067, + "grad_norm": 1.6124634742736816, + "learning_rate": 0.00010066299176824283, + "loss": 0.4583, + "step": 29569 + }, + { + "epoch": 2.0519047949483036, + "grad_norm": 1.1503461599349976, + "learning_rate": 0.00010065737327521659, + "loss": 0.5108, + "step": 29570 + }, + { + "epoch": 2.0519741863854, + "grad_norm": 0.9679075479507446, + "learning_rate": 0.00010065175478011509, + "loss": 0.3205, + "step": 29571 + }, + { + "epoch": 2.052043577822497, + "grad_norm": 1.109431505203247, + "learning_rate": 0.00010064613628295607, + "loss": 0.3471, + "step": 29572 + }, + { + "epoch": 2.052112969259593, + "grad_norm": 1.929701328277588, + "learning_rate": 0.00010064051778375732, + "loss": 0.5794, + "step": 29573 + }, + { + "epoch": 2.05218236069669, + "grad_norm": 1.1706701517105103, + "learning_rate": 0.00010063489928253648, + "loss": 0.5278, + "step": 29574 + }, + { + "epoch": 2.052251752133787, + "grad_norm": 1.2889013290405273, + "learning_rate": 0.00010062928077931137, + "loss": 0.4873, + "step": 29575 + }, + { + "epoch": 2.0523211435708832, + "grad_norm": 1.0732535123825073, + "learning_rate": 0.00010062366227409967, + "loss": 0.4655, + "step": 29576 + }, + { + "epoch": 2.05239053500798, + "grad_norm": 1.1784031391143799, + "learning_rate": 0.00010061804376691914, + "loss": 0.3214, + "step": 29577 + }, + { + "epoch": 2.0524599264450765, + "grad_norm": 1.0846635103225708, + "learning_rate": 0.00010061242525778751, + "loss": 0.3062, + "step": 29578 + }, + { + "epoch": 2.0525293178821733, + "grad_norm": 1.2409253120422363, + "learning_rate": 0.00010060680674672256, + "loss": 0.4579, + "step": 29579 + }, + { + "epoch": 2.05259870931927, + "grad_norm": 1.0034979581832886, + "learning_rate": 0.00010060118823374196, + "loss": 0.3893, + "step": 29580 + }, + { + "epoch": 2.0526681007563665, + "grad_norm": 2.188187837600708, + "learning_rate": 0.00010059556971886348, + "loss": 0.6601, + "step": 29581 + }, + { + "epoch": 2.0527374921934634, + "grad_norm": 1.2233279943466187, + "learning_rate": 0.00010058995120210488, + "loss": 0.4366, + "step": 29582 + }, + { + "epoch": 2.05280688363056, + "grad_norm": 1.4774705171585083, + "learning_rate": 0.00010058433268348384, + "loss": 0.4432, + "step": 29583 + }, + { + "epoch": 2.0528762750676566, + "grad_norm": 1.5722780227661133, + "learning_rate": 0.00010057871416301814, + "loss": 0.4153, + "step": 29584 + }, + { + "epoch": 2.0529456665047534, + "grad_norm": 1.34284508228302, + "learning_rate": 0.00010057309564072552, + "loss": 0.4334, + "step": 29585 + }, + { + "epoch": 2.05301505794185, + "grad_norm": 2.037268877029419, + "learning_rate": 0.0001005674771166237, + "loss": 0.5487, + "step": 29586 + }, + { + "epoch": 2.0530844493789466, + "grad_norm": 1.5969189405441284, + "learning_rate": 0.00010056185859073042, + "loss": 0.4205, + "step": 29587 + }, + { + "epoch": 2.0531538408160435, + "grad_norm": 1.0808442831039429, + "learning_rate": 0.00010055624006306341, + "loss": 0.3762, + "step": 29588 + }, + { + "epoch": 2.05322323225314, + "grad_norm": 1.2361717224121094, + "learning_rate": 0.00010055062153364044, + "loss": 0.5489, + "step": 29589 + }, + { + "epoch": 2.0532926236902367, + "grad_norm": 1.366955280303955, + "learning_rate": 0.00010054500300247917, + "loss": 0.4466, + "step": 29590 + }, + { + "epoch": 2.053362015127333, + "grad_norm": 1.0997562408447266, + "learning_rate": 0.00010053938446959746, + "loss": 0.2431, + "step": 29591 + }, + { + "epoch": 2.05343140656443, + "grad_norm": 2.068636894226074, + "learning_rate": 0.00010053376593501292, + "loss": 0.6321, + "step": 29592 + }, + { + "epoch": 2.0535007980015267, + "grad_norm": 1.4799244403839111, + "learning_rate": 0.00010052814739874337, + "loss": 0.5786, + "step": 29593 + }, + { + "epoch": 2.053570189438623, + "grad_norm": 1.014973759651184, + "learning_rate": 0.00010052252886080649, + "loss": 0.3993, + "step": 29594 + }, + { + "epoch": 2.05363958087572, + "grad_norm": 1.4186509847640991, + "learning_rate": 0.00010051691032122005, + "loss": 0.4991, + "step": 29595 + }, + { + "epoch": 2.053708972312817, + "grad_norm": 1.2385159730911255, + "learning_rate": 0.00010051129178000183, + "loss": 0.5138, + "step": 29596 + }, + { + "epoch": 2.053778363749913, + "grad_norm": 1.3726869821548462, + "learning_rate": 0.00010050567323716948, + "loss": 0.5337, + "step": 29597 + }, + { + "epoch": 2.05384775518701, + "grad_norm": 1.3419930934906006, + "learning_rate": 0.00010050005469274078, + "loss": 0.3816, + "step": 29598 + }, + { + "epoch": 2.0539171466241064, + "grad_norm": 1.5936561822891235, + "learning_rate": 0.00010049443614673348, + "loss": 0.6286, + "step": 29599 + }, + { + "epoch": 2.0539865380612032, + "grad_norm": 1.1022602319717407, + "learning_rate": 0.00010048881759916531, + "loss": 0.5257, + "step": 29600 + }, + { + "epoch": 2.0540559294983, + "grad_norm": 1.2670269012451172, + "learning_rate": 0.00010048319905005397, + "loss": 0.4124, + "step": 29601 + }, + { + "epoch": 2.0541253209353965, + "grad_norm": 2.018732786178589, + "learning_rate": 0.00010047758049941722, + "loss": 0.4858, + "step": 29602 + }, + { + "epoch": 2.0541947123724933, + "grad_norm": 1.9945682287216187, + "learning_rate": 0.00010047196194727286, + "loss": 0.4008, + "step": 29603 + }, + { + "epoch": 2.0542641038095897, + "grad_norm": 1.3599905967712402, + "learning_rate": 0.00010046634339363851, + "loss": 0.3711, + "step": 29604 + }, + { + "epoch": 2.0543334952466865, + "grad_norm": 1.644348382949829, + "learning_rate": 0.000100460724838532, + "loss": 0.6078, + "step": 29605 + }, + { + "epoch": 2.0544028866837833, + "grad_norm": 1.3018447160720825, + "learning_rate": 0.00010045510628197106, + "loss": 0.4946, + "step": 29606 + }, + { + "epoch": 2.0544722781208797, + "grad_norm": 1.0278735160827637, + "learning_rate": 0.00010044948772397335, + "loss": 0.3137, + "step": 29607 + }, + { + "epoch": 2.0545416695579766, + "grad_norm": 1.2173771858215332, + "learning_rate": 0.0001004438691645567, + "loss": 0.3477, + "step": 29608 + }, + { + "epoch": 2.0546110609950734, + "grad_norm": 1.1342182159423828, + "learning_rate": 0.00010043825060373877, + "loss": 0.5289, + "step": 29609 + }, + { + "epoch": 2.05468045243217, + "grad_norm": 1.1483724117279053, + "learning_rate": 0.00010043263204153738, + "loss": 0.52, + "step": 29610 + }, + { + "epoch": 2.0547498438692666, + "grad_norm": 1.115148901939392, + "learning_rate": 0.00010042701347797018, + "loss": 0.426, + "step": 29611 + }, + { + "epoch": 2.054819235306363, + "grad_norm": 1.3272247314453125, + "learning_rate": 0.00010042139491305495, + "loss": 0.492, + "step": 29612 + }, + { + "epoch": 2.05488862674346, + "grad_norm": 1.1972285509109497, + "learning_rate": 0.00010041577634680945, + "loss": 0.2751, + "step": 29613 + }, + { + "epoch": 2.0549580181805567, + "grad_norm": 1.8926342725753784, + "learning_rate": 0.00010041015777925138, + "loss": 0.2916, + "step": 29614 + }, + { + "epoch": 2.055027409617653, + "grad_norm": 1.1723966598510742, + "learning_rate": 0.00010040453921039847, + "loss": 0.3255, + "step": 29615 + }, + { + "epoch": 2.05509680105475, + "grad_norm": 1.0726677179336548, + "learning_rate": 0.0001003989206402685, + "loss": 0.3423, + "step": 29616 + }, + { + "epoch": 2.0551661924918463, + "grad_norm": 1.0776973962783813, + "learning_rate": 0.0001003933020688792, + "loss": 0.4254, + "step": 29617 + }, + { + "epoch": 2.055235583928943, + "grad_norm": 2.181532144546509, + "learning_rate": 0.00010038768349624825, + "loss": 0.4023, + "step": 29618 + }, + { + "epoch": 2.05530497536604, + "grad_norm": 1.627112627029419, + "learning_rate": 0.00010038206492239343, + "loss": 0.3509, + "step": 29619 + }, + { + "epoch": 2.0553743668031363, + "grad_norm": 1.0384490489959717, + "learning_rate": 0.00010037644634733252, + "loss": 0.4145, + "step": 29620 + }, + { + "epoch": 2.055443758240233, + "grad_norm": 1.2312408685684204, + "learning_rate": 0.00010037082777108318, + "loss": 0.4334, + "step": 29621 + }, + { + "epoch": 2.05551314967733, + "grad_norm": 1.3707844018936157, + "learning_rate": 0.00010036520919366318, + "loss": 0.4724, + "step": 29622 + }, + { + "epoch": 2.0555825411144264, + "grad_norm": 1.4526094198226929, + "learning_rate": 0.00010035959061509026, + "loss": 0.4747, + "step": 29623 + }, + { + "epoch": 2.0556519325515232, + "grad_norm": 1.0561256408691406, + "learning_rate": 0.00010035397203538218, + "loss": 0.3178, + "step": 29624 + }, + { + "epoch": 2.0557213239886196, + "grad_norm": 1.3835536241531372, + "learning_rate": 0.0001003483534545566, + "loss": 0.363, + "step": 29625 + }, + { + "epoch": 2.0557907154257165, + "grad_norm": 1.4525583982467651, + "learning_rate": 0.00010034273487263134, + "loss": 0.5311, + "step": 29626 + }, + { + "epoch": 2.0558601068628133, + "grad_norm": 1.1193479299545288, + "learning_rate": 0.00010033711628962414, + "loss": 0.4752, + "step": 29627 + }, + { + "epoch": 2.0559294982999097, + "grad_norm": 1.3590883016586304, + "learning_rate": 0.00010033149770555266, + "loss": 0.3411, + "step": 29628 + }, + { + "epoch": 2.0559988897370065, + "grad_norm": 1.2130389213562012, + "learning_rate": 0.00010032587912043468, + "loss": 0.3511, + "step": 29629 + }, + { + "epoch": 2.0560682811741033, + "grad_norm": 1.101815938949585, + "learning_rate": 0.00010032026053428795, + "loss": 0.3602, + "step": 29630 + }, + { + "epoch": 2.0561376726111997, + "grad_norm": 1.3488622903823853, + "learning_rate": 0.00010031464194713018, + "loss": 0.4067, + "step": 29631 + }, + { + "epoch": 2.0562070640482966, + "grad_norm": 1.7729209661483765, + "learning_rate": 0.00010030902335897912, + "loss": 0.5279, + "step": 29632 + }, + { + "epoch": 2.056276455485393, + "grad_norm": 1.015715479850769, + "learning_rate": 0.00010030340476985251, + "loss": 0.347, + "step": 29633 + }, + { + "epoch": 2.05634584692249, + "grad_norm": 1.2106839418411255, + "learning_rate": 0.0001002977861797681, + "loss": 0.3899, + "step": 29634 + }, + { + "epoch": 2.0564152383595866, + "grad_norm": 1.1103423833847046, + "learning_rate": 0.00010029216758874362, + "loss": 0.4373, + "step": 29635 + }, + { + "epoch": 2.056484629796683, + "grad_norm": 1.3211828470230103, + "learning_rate": 0.00010028654899679678, + "loss": 0.4072, + "step": 29636 + }, + { + "epoch": 2.05655402123378, + "grad_norm": 1.2392184734344482, + "learning_rate": 0.00010028093040394536, + "loss": 0.3596, + "step": 29637 + }, + { + "epoch": 2.0566234126708762, + "grad_norm": 1.4294449090957642, + "learning_rate": 0.00010027531181020706, + "loss": 0.4359, + "step": 29638 + }, + { + "epoch": 2.056692804107973, + "grad_norm": 1.4484350681304932, + "learning_rate": 0.00010026969321559963, + "loss": 0.72, + "step": 29639 + }, + { + "epoch": 2.05676219554507, + "grad_norm": 1.4278696775436401, + "learning_rate": 0.0001002640746201408, + "loss": 0.4198, + "step": 29640 + }, + { + "epoch": 2.0568315869821663, + "grad_norm": 1.3369873762130737, + "learning_rate": 0.00010025845602384836, + "loss": 0.3565, + "step": 29641 + }, + { + "epoch": 2.056900978419263, + "grad_norm": 1.335413932800293, + "learning_rate": 0.00010025283742673996, + "loss": 0.3376, + "step": 29642 + }, + { + "epoch": 2.0569703698563595, + "grad_norm": 1.9399666786193848, + "learning_rate": 0.00010024721882883341, + "loss": 0.6702, + "step": 29643 + }, + { + "epoch": 2.0570397612934563, + "grad_norm": 1.7277944087982178, + "learning_rate": 0.00010024160023014639, + "loss": 0.3151, + "step": 29644 + }, + { + "epoch": 2.057109152730553, + "grad_norm": 1.3607509136199951, + "learning_rate": 0.0001002359816306967, + "loss": 0.4105, + "step": 29645 + }, + { + "epoch": 2.0571785441676496, + "grad_norm": 1.7548383474349976, + "learning_rate": 0.00010023036303050202, + "loss": 0.4851, + "step": 29646 + }, + { + "epoch": 2.0572479356047464, + "grad_norm": 1.1124966144561768, + "learning_rate": 0.00010022474442958012, + "loss": 0.3721, + "step": 29647 + }, + { + "epoch": 2.0573173270418432, + "grad_norm": 1.1545240879058838, + "learning_rate": 0.00010021912582794872, + "loss": 0.2788, + "step": 29648 + }, + { + "epoch": 2.0573867184789396, + "grad_norm": 1.3543291091918945, + "learning_rate": 0.00010021350722562557, + "loss": 0.5085, + "step": 29649 + }, + { + "epoch": 2.0574561099160364, + "grad_norm": 1.5299708843231201, + "learning_rate": 0.0001002078886226284, + "loss": 0.6017, + "step": 29650 + }, + { + "epoch": 2.057525501353133, + "grad_norm": 1.0690444707870483, + "learning_rate": 0.00010020227001897498, + "loss": 0.3813, + "step": 29651 + }, + { + "epoch": 2.0575948927902297, + "grad_norm": 1.2622932195663452, + "learning_rate": 0.00010019665141468298, + "loss": 0.3398, + "step": 29652 + }, + { + "epoch": 2.0576642842273265, + "grad_norm": 1.139072060585022, + "learning_rate": 0.00010019103280977016, + "loss": 0.3715, + "step": 29653 + }, + { + "epoch": 2.057733675664423, + "grad_norm": 1.560715675354004, + "learning_rate": 0.00010018541420425429, + "loss": 0.4039, + "step": 29654 + }, + { + "epoch": 2.0578030671015197, + "grad_norm": 1.2613561153411865, + "learning_rate": 0.00010017979559815312, + "loss": 0.4922, + "step": 29655 + }, + { + "epoch": 2.0578724585386166, + "grad_norm": 1.6165508031845093, + "learning_rate": 0.00010017417699148433, + "loss": 0.5322, + "step": 29656 + }, + { + "epoch": 2.057941849975713, + "grad_norm": 1.2530581951141357, + "learning_rate": 0.00010016855838426566, + "loss": 0.4555, + "step": 29657 + }, + { + "epoch": 2.0580112414128098, + "grad_norm": 1.1466057300567627, + "learning_rate": 0.00010016293977651492, + "loss": 0.4488, + "step": 29658 + }, + { + "epoch": 2.058080632849906, + "grad_norm": 0.735421895980835, + "learning_rate": 0.00010015732116824976, + "loss": 0.2959, + "step": 29659 + }, + { + "epoch": 2.058150024287003, + "grad_norm": 1.2634867429733276, + "learning_rate": 0.00010015170255948799, + "loss": 0.3669, + "step": 29660 + }, + { + "epoch": 2.0582194157241, + "grad_norm": 1.0055702924728394, + "learning_rate": 0.00010014608395024727, + "loss": 0.2959, + "step": 29661 + }, + { + "epoch": 2.058288807161196, + "grad_norm": 1.1539089679718018, + "learning_rate": 0.00010014046534054542, + "loss": 0.6278, + "step": 29662 + }, + { + "epoch": 2.058358198598293, + "grad_norm": 0.9027893543243408, + "learning_rate": 0.0001001348467304001, + "loss": 0.2966, + "step": 29663 + }, + { + "epoch": 2.0584275900353894, + "grad_norm": 1.4530647993087769, + "learning_rate": 0.0001001292281198291, + "loss": 0.4023, + "step": 29664 + }, + { + "epoch": 2.0584969814724863, + "grad_norm": 1.4522432088851929, + "learning_rate": 0.00010012360950885016, + "loss": 0.3012, + "step": 29665 + }, + { + "epoch": 2.058566372909583, + "grad_norm": 1.1873531341552734, + "learning_rate": 0.00010011799089748097, + "loss": 0.3633, + "step": 29666 + }, + { + "epoch": 2.0586357643466795, + "grad_norm": 1.1967461109161377, + "learning_rate": 0.00010011237228573931, + "loss": 0.5546, + "step": 29667 + }, + { + "epoch": 2.0587051557837763, + "grad_norm": 1.584303379058838, + "learning_rate": 0.00010010675367364292, + "loss": 0.3911, + "step": 29668 + }, + { + "epoch": 2.0587745472208727, + "grad_norm": 1.5956110954284668, + "learning_rate": 0.00010010113506120949, + "loss": 0.4339, + "step": 29669 + }, + { + "epoch": 2.0588439386579696, + "grad_norm": 1.3249820470809937, + "learning_rate": 0.0001000955164484568, + "loss": 0.5107, + "step": 29670 + }, + { + "epoch": 2.0589133300950664, + "grad_norm": 1.2404999732971191, + "learning_rate": 0.00010008989783540257, + "loss": 0.4933, + "step": 29671 + }, + { + "epoch": 2.0589827215321628, + "grad_norm": 1.2406420707702637, + "learning_rate": 0.00010008427922206456, + "loss": 0.5494, + "step": 29672 + }, + { + "epoch": 2.0590521129692596, + "grad_norm": 1.526923656463623, + "learning_rate": 0.00010007866060846047, + "loss": 0.4589, + "step": 29673 + }, + { + "epoch": 2.0591215044063564, + "grad_norm": 1.1358344554901123, + "learning_rate": 0.00010007304199460807, + "loss": 0.4205, + "step": 29674 + }, + { + "epoch": 2.059190895843453, + "grad_norm": 1.4304907321929932, + "learning_rate": 0.0001000674233805251, + "loss": 0.4265, + "step": 29675 + }, + { + "epoch": 2.0592602872805497, + "grad_norm": 1.4555964469909668, + "learning_rate": 0.00010006180476622926, + "loss": 0.412, + "step": 29676 + }, + { + "epoch": 2.059329678717646, + "grad_norm": 1.9511299133300781, + "learning_rate": 0.0001000561861517383, + "loss": 0.5019, + "step": 29677 + }, + { + "epoch": 2.059399070154743, + "grad_norm": 1.3743579387664795, + "learning_rate": 0.00010005056753707001, + "loss": 0.3218, + "step": 29678 + }, + { + "epoch": 2.0594684615918397, + "grad_norm": 1.0689833164215088, + "learning_rate": 0.00010004494892224203, + "loss": 0.3782, + "step": 29679 + }, + { + "epoch": 2.059537853028936, + "grad_norm": 1.2291444540023804, + "learning_rate": 0.00010003933030727221, + "loss": 0.4954, + "step": 29680 + }, + { + "epoch": 2.059607244466033, + "grad_norm": 1.4761799573898315, + "learning_rate": 0.0001000337116921782, + "loss": 0.7268, + "step": 29681 + }, + { + "epoch": 2.0596766359031298, + "grad_norm": 1.305253028869629, + "learning_rate": 0.00010002809307697776, + "loss": 0.4312, + "step": 29682 + }, + { + "epoch": 2.059746027340226, + "grad_norm": 0.8396056294441223, + "learning_rate": 0.00010002247446168865, + "loss": 0.3658, + "step": 29683 + }, + { + "epoch": 2.059815418777323, + "grad_norm": 1.6089227199554443, + "learning_rate": 0.00010001685584632857, + "loss": 0.4946, + "step": 29684 + }, + { + "epoch": 2.0598848102144194, + "grad_norm": 1.3470492362976074, + "learning_rate": 0.00010001123723091525, + "loss": 0.4657, + "step": 29685 + }, + { + "epoch": 2.059954201651516, + "grad_norm": 1.2032666206359863, + "learning_rate": 0.00010000561861546651, + "loss": 0.4202, + "step": 29686 + }, + { + "epoch": 2.060023593088613, + "grad_norm": 1.5754365921020508, + "learning_rate": 0.0001, + "loss": 0.4733, + "step": 29687 + }, + { + "epoch": 2.0600929845257094, + "grad_norm": 1.142340064048767, + "learning_rate": 9.999438138453351e-05, + "loss": 0.4294, + "step": 29688 + }, + { + "epoch": 2.0601623759628063, + "grad_norm": 1.8499782085418701, + "learning_rate": 9.998876276908474e-05, + "loss": 0.5561, + "step": 29689 + }, + { + "epoch": 2.0602317673999027, + "grad_norm": 1.1557234525680542, + "learning_rate": 9.998314415367144e-05, + "loss": 0.4799, + "step": 29690 + }, + { + "epoch": 2.0603011588369995, + "grad_norm": 1.4756821393966675, + "learning_rate": 9.997752553831142e-05, + "loss": 0.3895, + "step": 29691 + }, + { + "epoch": 2.0603705502740963, + "grad_norm": 1.3296160697937012, + "learning_rate": 9.997190692302229e-05, + "loss": 0.5627, + "step": 29692 + }, + { + "epoch": 2.0604399417111927, + "grad_norm": 1.6017906665802002, + "learning_rate": 9.996628830782183e-05, + "loss": 0.5451, + "step": 29693 + }, + { + "epoch": 2.0605093331482895, + "grad_norm": 1.327296495437622, + "learning_rate": 9.996066969272781e-05, + "loss": 0.4462, + "step": 29694 + }, + { + "epoch": 2.0605787245853864, + "grad_norm": 1.4061336517333984, + "learning_rate": 9.995505107775798e-05, + "loss": 0.392, + "step": 29695 + }, + { + "epoch": 2.0606481160224828, + "grad_norm": 1.0811002254486084, + "learning_rate": 9.994943246293001e-05, + "loss": 0.3938, + "step": 29696 + }, + { + "epoch": 2.0607175074595796, + "grad_norm": 1.7281619310379028, + "learning_rate": 9.994381384826169e-05, + "loss": 0.3358, + "step": 29697 + }, + { + "epoch": 2.060786898896676, + "grad_norm": 1.1462039947509766, + "learning_rate": 9.993819523377076e-05, + "loss": 0.3004, + "step": 29698 + }, + { + "epoch": 2.060856290333773, + "grad_norm": 1.2619630098342896, + "learning_rate": 9.993257661947495e-05, + "loss": 0.4409, + "step": 29699 + }, + { + "epoch": 2.0609256817708697, + "grad_norm": 0.9738430380821228, + "learning_rate": 9.992695800539194e-05, + "loss": 0.4196, + "step": 29700 + }, + { + "epoch": 2.060995073207966, + "grad_norm": 1.0263538360595703, + "learning_rate": 9.992133939153954e-05, + "loss": 0.4194, + "step": 29701 + }, + { + "epoch": 2.061064464645063, + "grad_norm": 0.9172302484512329, + "learning_rate": 9.991572077793545e-05, + "loss": 0.3277, + "step": 29702 + }, + { + "epoch": 2.0611338560821593, + "grad_norm": 1.72574782371521, + "learning_rate": 9.991010216459743e-05, + "loss": 0.5727, + "step": 29703 + }, + { + "epoch": 2.061203247519256, + "grad_norm": 1.090675950050354, + "learning_rate": 9.99044835515432e-05, + "loss": 0.4292, + "step": 29704 + }, + { + "epoch": 2.061272638956353, + "grad_norm": 1.4833500385284424, + "learning_rate": 9.989886493879055e-05, + "loss": 0.3932, + "step": 29705 + }, + { + "epoch": 2.0613420303934493, + "grad_norm": 1.5636423826217651, + "learning_rate": 9.989324632635713e-05, + "loss": 0.4431, + "step": 29706 + }, + { + "epoch": 2.061411421830546, + "grad_norm": 1.0151331424713135, + "learning_rate": 9.988762771426071e-05, + "loss": 0.5735, + "step": 29707 + }, + { + "epoch": 2.061480813267643, + "grad_norm": 1.7806771993637085, + "learning_rate": 9.988200910251905e-05, + "loss": 0.5567, + "step": 29708 + }, + { + "epoch": 2.0615502047047394, + "grad_norm": 1.2598152160644531, + "learning_rate": 9.987639049114986e-05, + "loss": 0.3546, + "step": 29709 + }, + { + "epoch": 2.061619596141836, + "grad_norm": 1.3867090940475464, + "learning_rate": 9.98707718801709e-05, + "loss": 0.404, + "step": 29710 + }, + { + "epoch": 2.0616889875789326, + "grad_norm": 1.2062320709228516, + "learning_rate": 9.986515326959991e-05, + "loss": 0.4338, + "step": 29711 + }, + { + "epoch": 2.0617583790160294, + "grad_norm": 1.4938756227493286, + "learning_rate": 9.985953465945463e-05, + "loss": 0.3859, + "step": 29712 + }, + { + "epoch": 2.0618277704531263, + "grad_norm": 1.1970973014831543, + "learning_rate": 9.985391604975274e-05, + "loss": 0.4795, + "step": 29713 + }, + { + "epoch": 2.0618971618902227, + "grad_norm": 1.3479207754135132, + "learning_rate": 9.984829744051204e-05, + "loss": 0.4953, + "step": 29714 + }, + { + "epoch": 2.0619665533273195, + "grad_norm": 1.2881019115447998, + "learning_rate": 9.984267883175028e-05, + "loss": 0.4199, + "step": 29715 + }, + { + "epoch": 2.062035944764416, + "grad_norm": 2.0405731201171875, + "learning_rate": 9.983706022348511e-05, + "loss": 0.4617, + "step": 29716 + }, + { + "epoch": 2.0621053362015127, + "grad_norm": 1.2196805477142334, + "learning_rate": 9.983144161573434e-05, + "loss": 0.6104, + "step": 29717 + }, + { + "epoch": 2.0621747276386095, + "grad_norm": 1.0491056442260742, + "learning_rate": 9.982582300851568e-05, + "loss": 0.3202, + "step": 29718 + }, + { + "epoch": 2.062244119075706, + "grad_norm": 1.2203088998794556, + "learning_rate": 9.982020440184693e-05, + "loss": 0.4529, + "step": 29719 + }, + { + "epoch": 2.0623135105128028, + "grad_norm": 1.2448498010635376, + "learning_rate": 9.981458579574574e-05, + "loss": 0.4834, + "step": 29720 + }, + { + "epoch": 2.0623829019498996, + "grad_norm": 1.039320707321167, + "learning_rate": 9.980896719022985e-05, + "loss": 0.3164, + "step": 29721 + }, + { + "epoch": 2.062452293386996, + "grad_norm": 1.8326789140701294, + "learning_rate": 9.980334858531704e-05, + "loss": 0.6833, + "step": 29722 + }, + { + "epoch": 2.062521684824093, + "grad_norm": 1.5938246250152588, + "learning_rate": 9.979772998102506e-05, + "loss": 0.4534, + "step": 29723 + }, + { + "epoch": 2.062591076261189, + "grad_norm": 1.649309515953064, + "learning_rate": 9.97921113773716e-05, + "loss": 0.5953, + "step": 29724 + }, + { + "epoch": 2.062660467698286, + "grad_norm": 1.4078783988952637, + "learning_rate": 9.978649277437442e-05, + "loss": 0.5983, + "step": 29725 + }, + { + "epoch": 2.062729859135383, + "grad_norm": 1.2169651985168457, + "learning_rate": 9.978087417205131e-05, + "loss": 0.4332, + "step": 29726 + }, + { + "epoch": 2.0627992505724793, + "grad_norm": 0.9545867443084717, + "learning_rate": 9.977525557041992e-05, + "loss": 0.4695, + "step": 29727 + }, + { + "epoch": 2.062868642009576, + "grad_norm": 1.1795403957366943, + "learning_rate": 9.9769636969498e-05, + "loss": 0.4367, + "step": 29728 + }, + { + "epoch": 2.0629380334466725, + "grad_norm": 1.3026337623596191, + "learning_rate": 9.976401836930331e-05, + "loss": 0.4014, + "step": 29729 + }, + { + "epoch": 2.0630074248837693, + "grad_norm": 1.400017261505127, + "learning_rate": 9.975839976985363e-05, + "loss": 0.4695, + "step": 29730 + }, + { + "epoch": 2.063076816320866, + "grad_norm": 1.0896670818328857, + "learning_rate": 9.975278117116661e-05, + "loss": 0.3698, + "step": 29731 + }, + { + "epoch": 2.0631462077579625, + "grad_norm": 1.2660235166549683, + "learning_rate": 9.974716257326006e-05, + "loss": 0.3435, + "step": 29732 + }, + { + "epoch": 2.0632155991950594, + "grad_norm": 1.3988456726074219, + "learning_rate": 9.974154397615167e-05, + "loss": 0.3953, + "step": 29733 + }, + { + "epoch": 2.063284990632156, + "grad_norm": 1.1035619974136353, + "learning_rate": 9.973592537985921e-05, + "loss": 0.4281, + "step": 29734 + }, + { + "epoch": 2.0633543820692526, + "grad_norm": 1.4441677331924438, + "learning_rate": 9.97303067844004e-05, + "loss": 0.399, + "step": 29735 + }, + { + "epoch": 2.0634237735063494, + "grad_norm": 1.204957127571106, + "learning_rate": 9.972468818979298e-05, + "loss": 0.3229, + "step": 29736 + }, + { + "epoch": 2.063493164943446, + "grad_norm": 1.3139036893844604, + "learning_rate": 9.971906959605465e-05, + "loss": 0.3687, + "step": 29737 + }, + { + "epoch": 2.0635625563805426, + "grad_norm": 1.1078728437423706, + "learning_rate": 9.971345100320322e-05, + "loss": 0.361, + "step": 29738 + }, + { + "epoch": 2.0636319478176395, + "grad_norm": 0.913166344165802, + "learning_rate": 9.970783241125643e-05, + "loss": 0.447, + "step": 29739 + }, + { + "epoch": 2.063701339254736, + "grad_norm": 1.6202161312103271, + "learning_rate": 9.970221382023192e-05, + "loss": 0.4352, + "step": 29740 + }, + { + "epoch": 2.0637707306918327, + "grad_norm": 0.9493832588195801, + "learning_rate": 9.969659523014751e-05, + "loss": 0.4684, + "step": 29741 + }, + { + "epoch": 2.063840122128929, + "grad_norm": 1.2160941362380981, + "learning_rate": 9.969097664102089e-05, + "loss": 0.4348, + "step": 29742 + }, + { + "epoch": 2.063909513566026, + "grad_norm": 2.0398318767547607, + "learning_rate": 9.968535805286984e-05, + "loss": 0.5607, + "step": 29743 + }, + { + "epoch": 2.0639789050031228, + "grad_norm": 1.2510095834732056, + "learning_rate": 9.967973946571207e-05, + "loss": 0.3342, + "step": 29744 + }, + { + "epoch": 2.064048296440219, + "grad_norm": 1.356126308441162, + "learning_rate": 9.967412087956533e-05, + "loss": 0.431, + "step": 29745 + }, + { + "epoch": 2.064117687877316, + "grad_norm": 1.3028534650802612, + "learning_rate": 9.966850229444739e-05, + "loss": 0.4184, + "step": 29746 + }, + { + "epoch": 2.064187079314413, + "grad_norm": 1.2740411758422852, + "learning_rate": 9.966288371037591e-05, + "loss": 0.2723, + "step": 29747 + }, + { + "epoch": 2.064256470751509, + "grad_norm": 1.5716705322265625, + "learning_rate": 9.965726512736867e-05, + "loss": 0.4436, + "step": 29748 + }, + { + "epoch": 2.064325862188606, + "grad_norm": 1.0010679960250854, + "learning_rate": 9.965164654544342e-05, + "loss": 0.4386, + "step": 29749 + }, + { + "epoch": 2.0643952536257024, + "grad_norm": 1.5289818048477173, + "learning_rate": 9.964602796461785e-05, + "loss": 0.4669, + "step": 29750 + }, + { + "epoch": 2.0644646450627993, + "grad_norm": 1.083097219467163, + "learning_rate": 9.964040938490974e-05, + "loss": 0.3653, + "step": 29751 + }, + { + "epoch": 2.064534036499896, + "grad_norm": 0.8629450798034668, + "learning_rate": 9.963479080633683e-05, + "loss": 0.254, + "step": 29752 + }, + { + "epoch": 2.0646034279369925, + "grad_norm": 1.1699116230010986, + "learning_rate": 9.962917222891686e-05, + "loss": 0.3736, + "step": 29753 + }, + { + "epoch": 2.0646728193740893, + "grad_norm": 0.9651243090629578, + "learning_rate": 9.962355365266753e-05, + "loss": 0.2563, + "step": 29754 + }, + { + "epoch": 2.0647422108111857, + "grad_norm": 1.1664985418319702, + "learning_rate": 9.961793507760658e-05, + "loss": 0.4266, + "step": 29755 + }, + { + "epoch": 2.0648116022482825, + "grad_norm": 1.429197907447815, + "learning_rate": 9.961231650375176e-05, + "loss": 0.3883, + "step": 29756 + }, + { + "epoch": 2.0648809936853794, + "grad_norm": 1.633571743965149, + "learning_rate": 9.960669793112082e-05, + "loss": 0.4163, + "step": 29757 + }, + { + "epoch": 2.0649503851224758, + "grad_norm": 1.304451823234558, + "learning_rate": 9.96010793597315e-05, + "loss": 0.5954, + "step": 29758 + }, + { + "epoch": 2.0650197765595726, + "grad_norm": 1.105311393737793, + "learning_rate": 9.959546078960151e-05, + "loss": 0.4657, + "step": 29759 + }, + { + "epoch": 2.0650891679966694, + "grad_norm": 1.2880655527114868, + "learning_rate": 9.958984222074867e-05, + "loss": 0.5949, + "step": 29760 + }, + { + "epoch": 2.065158559433766, + "grad_norm": 1.2590644359588623, + "learning_rate": 9.958422365319059e-05, + "loss": 0.3123, + "step": 29761 + }, + { + "epoch": 2.0652279508708626, + "grad_norm": 1.3075119256973267, + "learning_rate": 9.957860508694507e-05, + "loss": 0.4956, + "step": 29762 + }, + { + "epoch": 2.065297342307959, + "grad_norm": 1.317564606666565, + "learning_rate": 9.957298652202983e-05, + "loss": 0.4801, + "step": 29763 + }, + { + "epoch": 2.065366733745056, + "grad_norm": 1.2967708110809326, + "learning_rate": 9.956736795846264e-05, + "loss": 0.3825, + "step": 29764 + }, + { + "epoch": 2.0654361251821527, + "grad_norm": 2.309727668762207, + "learning_rate": 9.956174939626124e-05, + "loss": 0.5549, + "step": 29765 + }, + { + "epoch": 2.065505516619249, + "grad_norm": 1.4291177988052368, + "learning_rate": 9.955613083544332e-05, + "loss": 0.5713, + "step": 29766 + }, + { + "epoch": 2.065574908056346, + "grad_norm": 1.5507478713989258, + "learning_rate": 9.955051227602666e-05, + "loss": 0.4496, + "step": 29767 + }, + { + "epoch": 2.0656442994934423, + "grad_norm": 1.2200146913528442, + "learning_rate": 9.954489371802898e-05, + "loss": 0.5175, + "step": 29768 + }, + { + "epoch": 2.065713690930539, + "grad_norm": 1.0664124488830566, + "learning_rate": 9.953927516146802e-05, + "loss": 0.4352, + "step": 29769 + }, + { + "epoch": 2.065783082367636, + "grad_norm": 1.0180407762527466, + "learning_rate": 9.95336566063615e-05, + "loss": 0.3059, + "step": 29770 + }, + { + "epoch": 2.0658524738047324, + "grad_norm": 2.5827462673187256, + "learning_rate": 9.952803805272717e-05, + "loss": 0.4538, + "step": 29771 + }, + { + "epoch": 2.065921865241829, + "grad_norm": 1.2823964357376099, + "learning_rate": 9.952241950058276e-05, + "loss": 0.4475, + "step": 29772 + }, + { + "epoch": 2.065991256678926, + "grad_norm": 1.19797945022583, + "learning_rate": 9.951680094994603e-05, + "loss": 0.2976, + "step": 29773 + }, + { + "epoch": 2.0660606481160224, + "grad_norm": 1.6250109672546387, + "learning_rate": 9.951118240083474e-05, + "loss": 0.4251, + "step": 29774 + }, + { + "epoch": 2.0661300395531192, + "grad_norm": 1.4161369800567627, + "learning_rate": 9.950556385326656e-05, + "loss": 0.6094, + "step": 29775 + }, + { + "epoch": 2.0661994309902156, + "grad_norm": 1.811155915260315, + "learning_rate": 9.949994530725925e-05, + "loss": 0.6733, + "step": 29776 + }, + { + "epoch": 2.0662688224273125, + "grad_norm": 1.6153922080993652, + "learning_rate": 9.949432676283055e-05, + "loss": 0.6064, + "step": 29777 + }, + { + "epoch": 2.0663382138644093, + "grad_norm": 1.1812382936477661, + "learning_rate": 9.94887082199982e-05, + "loss": 0.4691, + "step": 29778 + }, + { + "epoch": 2.0664076053015057, + "grad_norm": 1.3161875009536743, + "learning_rate": 9.948308967877994e-05, + "loss": 0.4218, + "step": 29779 + }, + { + "epoch": 2.0664769967386025, + "grad_norm": 1.0146764516830444, + "learning_rate": 9.94774711391935e-05, + "loss": 0.335, + "step": 29780 + }, + { + "epoch": 2.066546388175699, + "grad_norm": 0.9512661695480347, + "learning_rate": 9.947185260125668e-05, + "loss": 0.3541, + "step": 29781 + }, + { + "epoch": 2.0666157796127957, + "grad_norm": 1.7558605670928955, + "learning_rate": 9.94662340649871e-05, + "loss": 0.4965, + "step": 29782 + }, + { + "epoch": 2.0666851710498926, + "grad_norm": 0.9945314526557922, + "learning_rate": 9.946061553040258e-05, + "loss": 0.3737, + "step": 29783 + }, + { + "epoch": 2.066754562486989, + "grad_norm": 1.5794504880905151, + "learning_rate": 9.945499699752085e-05, + "loss": 0.4722, + "step": 29784 + }, + { + "epoch": 2.066823953924086, + "grad_norm": 1.5407301187515259, + "learning_rate": 9.94493784663596e-05, + "loss": 0.5155, + "step": 29785 + }, + { + "epoch": 2.0668933453611826, + "grad_norm": 1.1635788679122925, + "learning_rate": 9.94437599369366e-05, + "loss": 0.4487, + "step": 29786 + }, + { + "epoch": 2.066962736798279, + "grad_norm": 1.3046131134033203, + "learning_rate": 9.943814140926958e-05, + "loss": 0.5225, + "step": 29787 + }, + { + "epoch": 2.067032128235376, + "grad_norm": 1.8020638227462769, + "learning_rate": 9.943252288337635e-05, + "loss": 0.5813, + "step": 29788 + }, + { + "epoch": 2.0671015196724722, + "grad_norm": 1.6092318296432495, + "learning_rate": 9.942690435927451e-05, + "loss": 0.4839, + "step": 29789 + }, + { + "epoch": 2.067170911109569, + "grad_norm": 0.9296863675117493, + "learning_rate": 9.942128583698187e-05, + "loss": 0.3328, + "step": 29790 + }, + { + "epoch": 2.067240302546666, + "grad_norm": 1.046053171157837, + "learning_rate": 9.941566731651617e-05, + "loss": 0.3212, + "step": 29791 + }, + { + "epoch": 2.0673096939837623, + "grad_norm": 1.3584178686141968, + "learning_rate": 9.941004879789514e-05, + "loss": 0.4357, + "step": 29792 + }, + { + "epoch": 2.067379085420859, + "grad_norm": 0.9674185514450073, + "learning_rate": 9.940443028113651e-05, + "loss": 0.3778, + "step": 29793 + }, + { + "epoch": 2.0674484768579555, + "grad_norm": 1.1464124917984009, + "learning_rate": 9.939881176625804e-05, + "loss": 0.3484, + "step": 29794 + }, + { + "epoch": 2.0675178682950524, + "grad_norm": 1.0686384439468384, + "learning_rate": 9.939319325327748e-05, + "loss": 0.4697, + "step": 29795 + }, + { + "epoch": 2.067587259732149, + "grad_norm": 1.3987466096878052, + "learning_rate": 9.938757474221251e-05, + "loss": 0.4005, + "step": 29796 + }, + { + "epoch": 2.0676566511692456, + "grad_norm": 1.1234198808670044, + "learning_rate": 9.938195623308088e-05, + "loss": 0.359, + "step": 29797 + }, + { + "epoch": 2.0677260426063424, + "grad_norm": 0.8820004463195801, + "learning_rate": 9.937633772590037e-05, + "loss": 0.263, + "step": 29798 + }, + { + "epoch": 2.0677954340434392, + "grad_norm": 1.4039437770843506, + "learning_rate": 9.937071922068865e-05, + "loss": 0.5092, + "step": 29799 + }, + { + "epoch": 2.0678648254805356, + "grad_norm": 1.5575107336044312, + "learning_rate": 9.936510071746354e-05, + "loss": 0.4885, + "step": 29800 + }, + { + "epoch": 2.0679342169176325, + "grad_norm": 1.2484313249588013, + "learning_rate": 9.935948221624271e-05, + "loss": 0.43, + "step": 29801 + }, + { + "epoch": 2.068003608354729, + "grad_norm": 1.4822007417678833, + "learning_rate": 9.935386371704394e-05, + "loss": 0.4278, + "step": 29802 + }, + { + "epoch": 2.0680729997918257, + "grad_norm": 1.2928509712219238, + "learning_rate": 9.934824521988495e-05, + "loss": 0.4408, + "step": 29803 + }, + { + "epoch": 2.0681423912289225, + "grad_norm": 1.4005056619644165, + "learning_rate": 9.934262672478345e-05, + "loss": 0.4564, + "step": 29804 + }, + { + "epoch": 2.068211782666019, + "grad_norm": 0.998462975025177, + "learning_rate": 9.93370082317572e-05, + "loss": 0.2922, + "step": 29805 + }, + { + "epoch": 2.0682811741031157, + "grad_norm": 1.2608462572097778, + "learning_rate": 9.933138974082394e-05, + "loss": 0.5598, + "step": 29806 + }, + { + "epoch": 2.068350565540212, + "grad_norm": 1.099226713180542, + "learning_rate": 9.932577125200139e-05, + "loss": 0.4906, + "step": 29807 + }, + { + "epoch": 2.068419956977309, + "grad_norm": 1.1888351440429688, + "learning_rate": 9.932015276530737e-05, + "loss": 0.3514, + "step": 29808 + }, + { + "epoch": 2.068489348414406, + "grad_norm": 1.3951736688613892, + "learning_rate": 9.931453428075949e-05, + "loss": 0.4231, + "step": 29809 + }, + { + "epoch": 2.068558739851502, + "grad_norm": 1.6464598178863525, + "learning_rate": 9.930891579837556e-05, + "loss": 0.6328, + "step": 29810 + }, + { + "epoch": 2.068628131288599, + "grad_norm": 1.32315993309021, + "learning_rate": 9.930329731817329e-05, + "loss": 0.4755, + "step": 29811 + }, + { + "epoch": 2.068697522725696, + "grad_norm": 0.9729449152946472, + "learning_rate": 9.929767884017045e-05, + "loss": 0.3385, + "step": 29812 + }, + { + "epoch": 2.0687669141627922, + "grad_norm": 0.8558481931686401, + "learning_rate": 9.929206036438473e-05, + "loss": 0.3538, + "step": 29813 + }, + { + "epoch": 2.068836305599889, + "grad_norm": 1.0271079540252686, + "learning_rate": 9.92864418908339e-05, + "loss": 0.4452, + "step": 29814 + }, + { + "epoch": 2.0689056970369855, + "grad_norm": 0.8673293590545654, + "learning_rate": 9.928082341953573e-05, + "loss": 0.3324, + "step": 29815 + }, + { + "epoch": 2.0689750884740823, + "grad_norm": 1.183410882949829, + "learning_rate": 9.927520495050789e-05, + "loss": 0.4073, + "step": 29816 + }, + { + "epoch": 2.069044479911179, + "grad_norm": 1.4372644424438477, + "learning_rate": 9.926958648376813e-05, + "loss": 0.7046, + "step": 29817 + }, + { + "epoch": 2.0691138713482755, + "grad_norm": 1.0802005529403687, + "learning_rate": 9.926396801933421e-05, + "loss": 0.4313, + "step": 29818 + }, + { + "epoch": 2.0691832627853723, + "grad_norm": 1.03073251247406, + "learning_rate": 9.925834955722387e-05, + "loss": 0.3532, + "step": 29819 + }, + { + "epoch": 2.0692526542224687, + "grad_norm": 1.0773931741714478, + "learning_rate": 9.925273109745482e-05, + "loss": 0.3925, + "step": 29820 + }, + { + "epoch": 2.0693220456595656, + "grad_norm": 1.3813204765319824, + "learning_rate": 9.92471126400448e-05, + "loss": 0.3908, + "step": 29821 + }, + { + "epoch": 2.0693914370966624, + "grad_norm": 1.4676884412765503, + "learning_rate": 9.924149418501159e-05, + "loss": 0.4598, + "step": 29822 + }, + { + "epoch": 2.069460828533759, + "grad_norm": 1.3778525590896606, + "learning_rate": 9.92358757323729e-05, + "loss": 0.4396, + "step": 29823 + }, + { + "epoch": 2.0695302199708556, + "grad_norm": 1.1965712308883667, + "learning_rate": 9.923025728214643e-05, + "loss": 0.5504, + "step": 29824 + }, + { + "epoch": 2.0695996114079525, + "grad_norm": 1.6231070756912231, + "learning_rate": 9.922463883434995e-05, + "loss": 0.4751, + "step": 29825 + }, + { + "epoch": 2.069669002845049, + "grad_norm": 1.7571078538894653, + "learning_rate": 9.92190203890012e-05, + "loss": 0.5184, + "step": 29826 + }, + { + "epoch": 2.0697383942821457, + "grad_norm": 1.6742738485336304, + "learning_rate": 9.92134019461179e-05, + "loss": 0.6129, + "step": 29827 + }, + { + "epoch": 2.069807785719242, + "grad_norm": 1.1959776878356934, + "learning_rate": 9.920778350571781e-05, + "loss": 0.3622, + "step": 29828 + }, + { + "epoch": 2.069877177156339, + "grad_norm": 0.9013462662696838, + "learning_rate": 9.92021650678187e-05, + "loss": 0.2792, + "step": 29829 + }, + { + "epoch": 2.0699465685934357, + "grad_norm": 0.9933355450630188, + "learning_rate": 9.919654663243821e-05, + "loss": 0.4323, + "step": 29830 + }, + { + "epoch": 2.070015960030532, + "grad_norm": 0.9514181017875671, + "learning_rate": 9.919092819959413e-05, + "loss": 0.3879, + "step": 29831 + }, + { + "epoch": 2.070085351467629, + "grad_norm": 1.034648060798645, + "learning_rate": 9.918530976930421e-05, + "loss": 0.4546, + "step": 29832 + }, + { + "epoch": 2.0701547429047253, + "grad_norm": 1.0698765516281128, + "learning_rate": 9.917969134158616e-05, + "loss": 0.4032, + "step": 29833 + }, + { + "epoch": 2.070224134341822, + "grad_norm": 1.2448742389678955, + "learning_rate": 9.917407291645773e-05, + "loss": 0.4816, + "step": 29834 + }, + { + "epoch": 2.070293525778919, + "grad_norm": 1.3470878601074219, + "learning_rate": 9.916845449393668e-05, + "loss": 0.4812, + "step": 29835 + }, + { + "epoch": 2.0703629172160154, + "grad_norm": 1.3303368091583252, + "learning_rate": 9.916283607404073e-05, + "loss": 0.5116, + "step": 29836 + }, + { + "epoch": 2.0704323086531122, + "grad_norm": 1.3037759065628052, + "learning_rate": 9.915721765678757e-05, + "loss": 0.3058, + "step": 29837 + }, + { + "epoch": 2.070501700090209, + "grad_norm": 1.193500280380249, + "learning_rate": 9.915159924219501e-05, + "loss": 0.3687, + "step": 29838 + }, + { + "epoch": 2.0705710915273055, + "grad_norm": 1.0065841674804688, + "learning_rate": 9.914598083028072e-05, + "loss": 0.4192, + "step": 29839 + }, + { + "epoch": 2.0706404829644023, + "grad_norm": 1.1269259452819824, + "learning_rate": 9.914036242106248e-05, + "loss": 0.3816, + "step": 29840 + }, + { + "epoch": 2.0707098744014987, + "grad_norm": 1.0414628982543945, + "learning_rate": 9.913474401455803e-05, + "loss": 0.4054, + "step": 29841 + }, + { + "epoch": 2.0707792658385955, + "grad_norm": 1.35110604763031, + "learning_rate": 9.912912561078507e-05, + "loss": 0.6241, + "step": 29842 + }, + { + "epoch": 2.0708486572756923, + "grad_norm": 1.5032320022583008, + "learning_rate": 9.912350720976141e-05, + "loss": 0.5739, + "step": 29843 + }, + { + "epoch": 2.0709180487127887, + "grad_norm": 1.5083283185958862, + "learning_rate": 9.91178888115047e-05, + "loss": 0.4216, + "step": 29844 + }, + { + "epoch": 2.0709874401498856, + "grad_norm": 1.3313794136047363, + "learning_rate": 9.911227041603269e-05, + "loss": 0.5074, + "step": 29845 + }, + { + "epoch": 2.071056831586982, + "grad_norm": 1.4653598070144653, + "learning_rate": 9.910665202336315e-05, + "loss": 0.4662, + "step": 29846 + }, + { + "epoch": 2.071126223024079, + "grad_norm": 0.9017951488494873, + "learning_rate": 9.910103363351382e-05, + "loss": 0.3026, + "step": 29847 + }, + { + "epoch": 2.0711956144611756, + "grad_norm": 1.541914701461792, + "learning_rate": 9.909541524650241e-05, + "loss": 0.4757, + "step": 29848 + }, + { + "epoch": 2.071265005898272, + "grad_norm": 1.2122204303741455, + "learning_rate": 9.908979686234667e-05, + "loss": 0.402, + "step": 29849 + }, + { + "epoch": 2.071334397335369, + "grad_norm": 1.5400140285491943, + "learning_rate": 9.908417848106439e-05, + "loss": 0.7123, + "step": 29850 + }, + { + "epoch": 2.0714037887724657, + "grad_norm": 1.9946659803390503, + "learning_rate": 9.90785601026732e-05, + "loss": 0.4431, + "step": 29851 + }, + { + "epoch": 2.071473180209562, + "grad_norm": 1.2892504930496216, + "learning_rate": 9.907294172719089e-05, + "loss": 0.3293, + "step": 29852 + }, + { + "epoch": 2.071542571646659, + "grad_norm": 1.080317497253418, + "learning_rate": 9.906732335463519e-05, + "loss": 0.4421, + "step": 29853 + }, + { + "epoch": 2.0716119630837553, + "grad_norm": 1.200553059577942, + "learning_rate": 9.906170498502386e-05, + "loss": 0.3691, + "step": 29854 + }, + { + "epoch": 2.071681354520852, + "grad_norm": 1.2450252771377563, + "learning_rate": 9.90560866183746e-05, + "loss": 0.3705, + "step": 29855 + }, + { + "epoch": 2.071750745957949, + "grad_norm": 1.2471904754638672, + "learning_rate": 9.905046825470517e-05, + "loss": 0.5181, + "step": 29856 + }, + { + "epoch": 2.0718201373950453, + "grad_norm": 1.5965536832809448, + "learning_rate": 9.904484989403333e-05, + "loss": 0.5006, + "step": 29857 + }, + { + "epoch": 2.071889528832142, + "grad_norm": 1.4121729135513306, + "learning_rate": 9.903923153637678e-05, + "loss": 0.5084, + "step": 29858 + }, + { + "epoch": 2.071958920269239, + "grad_norm": 1.39894437789917, + "learning_rate": 9.903361318175324e-05, + "loss": 0.5522, + "step": 29859 + }, + { + "epoch": 2.0720283117063354, + "grad_norm": 1.0704379081726074, + "learning_rate": 9.902799483018049e-05, + "loss": 0.3117, + "step": 29860 + }, + { + "epoch": 2.0720977031434322, + "grad_norm": 1.2580533027648926, + "learning_rate": 9.902237648167623e-05, + "loss": 0.3144, + "step": 29861 + }, + { + "epoch": 2.0721670945805286, + "grad_norm": 1.251321792602539, + "learning_rate": 9.901675813625821e-05, + "loss": 0.55, + "step": 29862 + }, + { + "epoch": 2.0722364860176254, + "grad_norm": 1.8048646450042725, + "learning_rate": 9.901113979394418e-05, + "loss": 0.371, + "step": 29863 + }, + { + "epoch": 2.0723058774547223, + "grad_norm": 1.5854190587997437, + "learning_rate": 9.90055214547519e-05, + "loss": 0.516, + "step": 29864 + }, + { + "epoch": 2.0723752688918187, + "grad_norm": 1.6541067361831665, + "learning_rate": 9.899990311869904e-05, + "loss": 0.4465, + "step": 29865 + }, + { + "epoch": 2.0724446603289155, + "grad_norm": 1.318044900894165, + "learning_rate": 9.899428478580337e-05, + "loss": 0.462, + "step": 29866 + }, + { + "epoch": 2.072514051766012, + "grad_norm": 1.1658174991607666, + "learning_rate": 9.898866645608262e-05, + "loss": 0.4603, + "step": 29867 + }, + { + "epoch": 2.0725834432031087, + "grad_norm": 1.1424638032913208, + "learning_rate": 9.898304812955455e-05, + "loss": 0.3986, + "step": 29868 + }, + { + "epoch": 2.0726528346402056, + "grad_norm": 1.351547122001648, + "learning_rate": 9.897742980623685e-05, + "loss": 0.4855, + "step": 29869 + }, + { + "epoch": 2.072722226077302, + "grad_norm": 1.1166949272155762, + "learning_rate": 9.897181148614732e-05, + "loss": 0.4089, + "step": 29870 + }, + { + "epoch": 2.0727916175143988, + "grad_norm": 1.2025721073150635, + "learning_rate": 9.896619316930367e-05, + "loss": 0.4176, + "step": 29871 + }, + { + "epoch": 2.072861008951495, + "grad_norm": 1.2069091796875, + "learning_rate": 9.89605748557236e-05, + "loss": 0.5062, + "step": 29872 + }, + { + "epoch": 2.072930400388592, + "grad_norm": 1.3379414081573486, + "learning_rate": 9.89549565454249e-05, + "loss": 0.521, + "step": 29873 + }, + { + "epoch": 2.072999791825689, + "grad_norm": 1.045518159866333, + "learning_rate": 9.894933823842526e-05, + "loss": 0.3626, + "step": 29874 + }, + { + "epoch": 2.073069183262785, + "grad_norm": 1.038894772529602, + "learning_rate": 9.894371993474243e-05, + "loss": 0.3994, + "step": 29875 + }, + { + "epoch": 2.073138574699882, + "grad_norm": 1.0875790119171143, + "learning_rate": 9.893810163439416e-05, + "loss": 0.4442, + "step": 29876 + }, + { + "epoch": 2.073207966136979, + "grad_norm": 1.2676892280578613, + "learning_rate": 9.893248333739818e-05, + "loss": 0.5117, + "step": 29877 + }, + { + "epoch": 2.0732773575740753, + "grad_norm": 1.9719414710998535, + "learning_rate": 9.892686504377226e-05, + "loss": 0.5043, + "step": 29878 + }, + { + "epoch": 2.073346749011172, + "grad_norm": 1.9765381813049316, + "learning_rate": 9.892124675353407e-05, + "loss": 0.5087, + "step": 29879 + }, + { + "epoch": 2.0734161404482685, + "grad_norm": 1.1790480613708496, + "learning_rate": 9.891562846670137e-05, + "loss": 0.3564, + "step": 29880 + }, + { + "epoch": 2.0734855318853653, + "grad_norm": 1.2708845138549805, + "learning_rate": 9.891001018329191e-05, + "loss": 0.3157, + "step": 29881 + }, + { + "epoch": 2.073554923322462, + "grad_norm": 1.3599185943603516, + "learning_rate": 9.890439190332342e-05, + "loss": 0.3398, + "step": 29882 + }, + { + "epoch": 2.0736243147595586, + "grad_norm": 1.1834741830825806, + "learning_rate": 9.889877362681364e-05, + "loss": 0.3313, + "step": 29883 + }, + { + "epoch": 2.0736937061966554, + "grad_norm": 1.874881625175476, + "learning_rate": 9.889315535378034e-05, + "loss": 0.4822, + "step": 29884 + }, + { + "epoch": 2.073763097633752, + "grad_norm": 1.52498197555542, + "learning_rate": 9.88875370842412e-05, + "loss": 0.4174, + "step": 29885 + }, + { + "epoch": 2.0738324890708486, + "grad_norm": 1.1348495483398438, + "learning_rate": 9.888191881821395e-05, + "loss": 0.3917, + "step": 29886 + }, + { + "epoch": 2.0739018805079454, + "grad_norm": 1.4670329093933105, + "learning_rate": 9.887630055571636e-05, + "loss": 0.4005, + "step": 29887 + }, + { + "epoch": 2.073971271945042, + "grad_norm": 1.1013044118881226, + "learning_rate": 9.887068229676616e-05, + "loss": 0.3455, + "step": 29888 + }, + { + "epoch": 2.0740406633821387, + "grad_norm": 1.1195471286773682, + "learning_rate": 9.886506404138111e-05, + "loss": 0.3912, + "step": 29889 + }, + { + "epoch": 2.0741100548192355, + "grad_norm": 1.8762750625610352, + "learning_rate": 9.88594457895789e-05, + "loss": 0.6957, + "step": 29890 + }, + { + "epoch": 2.074179446256332, + "grad_norm": 1.4580658674240112, + "learning_rate": 9.88538275413773e-05, + "loss": 0.4338, + "step": 29891 + }, + { + "epoch": 2.0742488376934287, + "grad_norm": 1.9640281200408936, + "learning_rate": 9.884820929679403e-05, + "loss": 0.3396, + "step": 29892 + }, + { + "epoch": 2.074318229130525, + "grad_norm": 1.0755897760391235, + "learning_rate": 9.884259105584682e-05, + "loss": 0.3964, + "step": 29893 + }, + { + "epoch": 2.074387620567622, + "grad_norm": 0.9815100431442261, + "learning_rate": 9.883697281855342e-05, + "loss": 0.3218, + "step": 29894 + }, + { + "epoch": 2.0744570120047188, + "grad_norm": 1.811519742012024, + "learning_rate": 9.883135458493156e-05, + "loss": 0.6327, + "step": 29895 + }, + { + "epoch": 2.074526403441815, + "grad_norm": 1.5824859142303467, + "learning_rate": 9.882573635499897e-05, + "loss": 0.4242, + "step": 29896 + }, + { + "epoch": 2.074595794878912, + "grad_norm": 2.2068779468536377, + "learning_rate": 9.882011812877338e-05, + "loss": 0.3038, + "step": 29897 + }, + { + "epoch": 2.0746651863160084, + "grad_norm": 1.3383784294128418, + "learning_rate": 9.88144999062726e-05, + "loss": 0.4969, + "step": 29898 + }, + { + "epoch": 2.074734577753105, + "grad_norm": 1.7027286291122437, + "learning_rate": 9.880888168751428e-05, + "loss": 0.5346, + "step": 29899 + }, + { + "epoch": 2.074803969190202, + "grad_norm": 1.5733281373977661, + "learning_rate": 9.880326347251617e-05, + "loss": 0.6025, + "step": 29900 + }, + { + "epoch": 2.0748733606272984, + "grad_norm": 1.4498460292816162, + "learning_rate": 9.8797645261296e-05, + "loss": 0.4794, + "step": 29901 + }, + { + "epoch": 2.0749427520643953, + "grad_norm": 1.0846889019012451, + "learning_rate": 9.879202705387154e-05, + "loss": 0.4412, + "step": 29902 + }, + { + "epoch": 2.075012143501492, + "grad_norm": 1.374497890472412, + "learning_rate": 9.878640885026052e-05, + "loss": 0.4311, + "step": 29903 + }, + { + "epoch": 2.0750815349385885, + "grad_norm": 0.850318968296051, + "learning_rate": 9.878079065048067e-05, + "loss": 0.275, + "step": 29904 + }, + { + "epoch": 2.0751509263756853, + "grad_norm": 2.093480348587036, + "learning_rate": 9.877517245454972e-05, + "loss": 0.5848, + "step": 29905 + }, + { + "epoch": 2.0752203178127817, + "grad_norm": 1.994759202003479, + "learning_rate": 9.87695542624854e-05, + "loss": 0.6627, + "step": 29906 + }, + { + "epoch": 2.0752897092498785, + "grad_norm": 1.6866776943206787, + "learning_rate": 9.876393607430545e-05, + "loss": 0.5241, + "step": 29907 + }, + { + "epoch": 2.0753591006869754, + "grad_norm": 1.234466314315796, + "learning_rate": 9.875831789002763e-05, + "loss": 0.3645, + "step": 29908 + }, + { + "epoch": 2.0754284921240718, + "grad_norm": 1.9056763648986816, + "learning_rate": 9.875269970966963e-05, + "loss": 0.571, + "step": 29909 + }, + { + "epoch": 2.0754978835611686, + "grad_norm": 1.0522396564483643, + "learning_rate": 9.874708153324924e-05, + "loss": 0.3425, + "step": 29910 + }, + { + "epoch": 2.0755672749982654, + "grad_norm": 1.1073658466339111, + "learning_rate": 9.874146336078413e-05, + "loss": 0.4192, + "step": 29911 + }, + { + "epoch": 2.075636666435362, + "grad_norm": 1.1299183368682861, + "learning_rate": 9.873584519229214e-05, + "loss": 0.4213, + "step": 29912 + }, + { + "epoch": 2.0757060578724587, + "grad_norm": 1.5732065439224243, + "learning_rate": 9.87302270277909e-05, + "loss": 0.8204, + "step": 29913 + }, + { + "epoch": 2.075775449309555, + "grad_norm": 1.6750555038452148, + "learning_rate": 9.872460886729817e-05, + "loss": 0.7128, + "step": 29914 + }, + { + "epoch": 2.075844840746652, + "grad_norm": 1.1616315841674805, + "learning_rate": 9.871899071083172e-05, + "loss": 0.3601, + "step": 29915 + }, + { + "epoch": 2.0759142321837487, + "grad_norm": 1.6783099174499512, + "learning_rate": 9.871337255840925e-05, + "loss": 0.3831, + "step": 29916 + }, + { + "epoch": 2.075983623620845, + "grad_norm": 1.4172037839889526, + "learning_rate": 9.870775441004853e-05, + "loss": 0.4059, + "step": 29917 + }, + { + "epoch": 2.076053015057942, + "grad_norm": 1.5864841938018799, + "learning_rate": 9.870213626576727e-05, + "loss": 0.3867, + "step": 29918 + }, + { + "epoch": 2.0761224064950383, + "grad_norm": 1.296764850616455, + "learning_rate": 9.869651812558325e-05, + "loss": 0.3717, + "step": 29919 + }, + { + "epoch": 2.076191797932135, + "grad_norm": 1.300022006034851, + "learning_rate": 9.869089998951414e-05, + "loss": 0.5309, + "step": 29920 + }, + { + "epoch": 2.076261189369232, + "grad_norm": 1.285704493522644, + "learning_rate": 9.86852818575777e-05, + "loss": 0.4008, + "step": 29921 + }, + { + "epoch": 2.0763305808063284, + "grad_norm": 1.4614231586456299, + "learning_rate": 9.867966372979168e-05, + "loss": 0.5321, + "step": 29922 + }, + { + "epoch": 2.076399972243425, + "grad_norm": 1.234366536140442, + "learning_rate": 9.867404560617381e-05, + "loss": 0.4544, + "step": 29923 + }, + { + "epoch": 2.0764693636805216, + "grad_norm": 1.0587016344070435, + "learning_rate": 9.866842748674183e-05, + "loss": 0.3952, + "step": 29924 + }, + { + "epoch": 2.0765387551176184, + "grad_norm": 1.1931991577148438, + "learning_rate": 9.866280937151347e-05, + "loss": 0.3431, + "step": 29925 + }, + { + "epoch": 2.0766081465547153, + "grad_norm": 1.1606318950653076, + "learning_rate": 9.865719126050647e-05, + "loss": 0.3286, + "step": 29926 + }, + { + "epoch": 2.0766775379918117, + "grad_norm": 1.4832682609558105, + "learning_rate": 9.865157315373857e-05, + "loss": 0.559, + "step": 29927 + }, + { + "epoch": 2.0767469294289085, + "grad_norm": 1.1894042491912842, + "learning_rate": 9.864595505122746e-05, + "loss": 0.3743, + "step": 29928 + }, + { + "epoch": 2.0768163208660053, + "grad_norm": 1.7016468048095703, + "learning_rate": 9.864033695299093e-05, + "loss": 0.9246, + "step": 29929 + }, + { + "epoch": 2.0768857123031017, + "grad_norm": 1.4685635566711426, + "learning_rate": 9.863471885904669e-05, + "loss": 0.3616, + "step": 29930 + }, + { + "epoch": 2.0769551037401985, + "grad_norm": 0.9572318196296692, + "learning_rate": 9.86291007694125e-05, + "loss": 0.2988, + "step": 29931 + }, + { + "epoch": 2.077024495177295, + "grad_norm": 1.373633623123169, + "learning_rate": 9.862348268410606e-05, + "loss": 0.4627, + "step": 29932 + }, + { + "epoch": 2.0770938866143918, + "grad_norm": 1.0096153020858765, + "learning_rate": 9.861786460314518e-05, + "loss": 0.4006, + "step": 29933 + }, + { + "epoch": 2.0771632780514886, + "grad_norm": 0.9318418502807617, + "learning_rate": 9.86122465265475e-05, + "loss": 0.2629, + "step": 29934 + }, + { + "epoch": 2.077232669488585, + "grad_norm": 1.595103144645691, + "learning_rate": 9.860662845433078e-05, + "loss": 0.5206, + "step": 29935 + }, + { + "epoch": 2.077302060925682, + "grad_norm": 1.2740318775177002, + "learning_rate": 9.860101038651279e-05, + "loss": 0.3577, + "step": 29936 + }, + { + "epoch": 2.0773714523627786, + "grad_norm": 1.6505489349365234, + "learning_rate": 9.859539232311124e-05, + "loss": 0.5225, + "step": 29937 + }, + { + "epoch": 2.077440843799875, + "grad_norm": 2.262768268585205, + "learning_rate": 9.858977426414388e-05, + "loss": 0.4925, + "step": 29938 + }, + { + "epoch": 2.077510235236972, + "grad_norm": 1.3825668096542358, + "learning_rate": 9.858415620962845e-05, + "loss": 0.4157, + "step": 29939 + }, + { + "epoch": 2.0775796266740683, + "grad_norm": 1.5326420068740845, + "learning_rate": 9.857853815958268e-05, + "loss": 0.4837, + "step": 29940 + }, + { + "epoch": 2.077649018111165, + "grad_norm": 1.2733744382858276, + "learning_rate": 9.857292011402428e-05, + "loss": 0.4239, + "step": 29941 + }, + { + "epoch": 2.077718409548262, + "grad_norm": 1.8640857934951782, + "learning_rate": 9.856730207297101e-05, + "loss": 0.3897, + "step": 29942 + }, + { + "epoch": 2.0777878009853583, + "grad_norm": 1.273960828781128, + "learning_rate": 9.856168403644062e-05, + "loss": 0.5214, + "step": 29943 + }, + { + "epoch": 2.077857192422455, + "grad_norm": 1.0564466714859009, + "learning_rate": 9.85560660044508e-05, + "loss": 0.4412, + "step": 29944 + }, + { + "epoch": 2.0779265838595515, + "grad_norm": 0.9996270537376404, + "learning_rate": 9.855044797701931e-05, + "loss": 0.3493, + "step": 29945 + }, + { + "epoch": 2.0779959752966484, + "grad_norm": 1.6259548664093018, + "learning_rate": 9.85448299541639e-05, + "loss": 0.5775, + "step": 29946 + }, + { + "epoch": 2.078065366733745, + "grad_norm": 1.2550129890441895, + "learning_rate": 9.853921193590233e-05, + "loss": 0.3312, + "step": 29947 + }, + { + "epoch": 2.0781347581708416, + "grad_norm": 1.2068564891815186, + "learning_rate": 9.853359392225224e-05, + "loss": 0.4371, + "step": 29948 + }, + { + "epoch": 2.0782041496079384, + "grad_norm": 1.9699565172195435, + "learning_rate": 9.852797591323144e-05, + "loss": 0.5244, + "step": 29949 + }, + { + "epoch": 2.0782735410450353, + "grad_norm": 1.3405742645263672, + "learning_rate": 9.852235790885764e-05, + "loss": 0.4576, + "step": 29950 + }, + { + "epoch": 2.0783429324821316, + "grad_norm": 1.4500353336334229, + "learning_rate": 9.85167399091486e-05, + "loss": 0.3572, + "step": 29951 + }, + { + "epoch": 2.0784123239192285, + "grad_norm": 1.381399154663086, + "learning_rate": 9.851112191412203e-05, + "loss": 0.5426, + "step": 29952 + }, + { + "epoch": 2.078481715356325, + "grad_norm": 1.6443746089935303, + "learning_rate": 9.850550392379567e-05, + "loss": 0.3946, + "step": 29953 + }, + { + "epoch": 2.0785511067934217, + "grad_norm": 1.0630995035171509, + "learning_rate": 9.849988593818732e-05, + "loss": 0.4248, + "step": 29954 + }, + { + "epoch": 2.0786204982305185, + "grad_norm": 1.149280071258545, + "learning_rate": 9.84942679573146e-05, + "loss": 0.3329, + "step": 29955 + }, + { + "epoch": 2.078689889667615, + "grad_norm": 1.7053065299987793, + "learning_rate": 9.84886499811953e-05, + "loss": 0.4173, + "step": 29956 + }, + { + "epoch": 2.0787592811047118, + "grad_norm": 1.5092822313308716, + "learning_rate": 9.848303200984716e-05, + "loss": 0.5111, + "step": 29957 + }, + { + "epoch": 2.078828672541808, + "grad_norm": 2.3236258029937744, + "learning_rate": 9.847741404328794e-05, + "loss": 0.4771, + "step": 29958 + }, + { + "epoch": 2.078898063978905, + "grad_norm": 1.3277390003204346, + "learning_rate": 9.847179608153531e-05, + "loss": 0.2959, + "step": 29959 + }, + { + "epoch": 2.078967455416002, + "grad_norm": 1.2898544073104858, + "learning_rate": 9.84661781246071e-05, + "loss": 0.5328, + "step": 29960 + }, + { + "epoch": 2.079036846853098, + "grad_norm": 1.8547313213348389, + "learning_rate": 9.846056017252092e-05, + "loss": 0.5219, + "step": 29961 + }, + { + "epoch": 2.079106238290195, + "grad_norm": 1.2775802612304688, + "learning_rate": 9.845494222529462e-05, + "loss": 0.4101, + "step": 29962 + }, + { + "epoch": 2.079175629727292, + "grad_norm": 1.2267367839813232, + "learning_rate": 9.844932428294587e-05, + "loss": 0.4916, + "step": 29963 + }, + { + "epoch": 2.0792450211643883, + "grad_norm": 1.2117987871170044, + "learning_rate": 9.844370634549241e-05, + "loss": 0.4029, + "step": 29964 + }, + { + "epoch": 2.079314412601485, + "grad_norm": 1.9684468507766724, + "learning_rate": 9.843808841295199e-05, + "loss": 0.4554, + "step": 29965 + }, + { + "epoch": 2.0793838040385815, + "grad_norm": 1.3553749322891235, + "learning_rate": 9.843247048534236e-05, + "loss": 0.5052, + "step": 29966 + }, + { + "epoch": 2.0794531954756783, + "grad_norm": 1.2982444763183594, + "learning_rate": 9.842685256268126e-05, + "loss": 0.4479, + "step": 29967 + }, + { + "epoch": 2.079522586912775, + "grad_norm": 1.3139030933380127, + "learning_rate": 9.842123464498637e-05, + "loss": 0.3657, + "step": 29968 + }, + { + "epoch": 2.0795919783498715, + "grad_norm": 1.3135558366775513, + "learning_rate": 9.841561673227546e-05, + "loss": 0.3053, + "step": 29969 + }, + { + "epoch": 2.0796613697869684, + "grad_norm": 0.8902036547660828, + "learning_rate": 9.840999882456627e-05, + "loss": 0.3091, + "step": 29970 + }, + { + "epoch": 2.0797307612240648, + "grad_norm": 1.0917326211929321, + "learning_rate": 9.840438092187652e-05, + "loss": 0.5532, + "step": 29971 + }, + { + "epoch": 2.0798001526611616, + "grad_norm": 1.5856151580810547, + "learning_rate": 9.839876302422395e-05, + "loss": 0.271, + "step": 29972 + }, + { + "epoch": 2.0798695440982584, + "grad_norm": 1.4167670011520386, + "learning_rate": 9.839314513162631e-05, + "loss": 0.543, + "step": 29973 + }, + { + "epoch": 2.079938935535355, + "grad_norm": 1.2395018339157104, + "learning_rate": 9.838752724410136e-05, + "loss": 0.5262, + "step": 29974 + }, + { + "epoch": 2.0800083269724516, + "grad_norm": 0.8825464248657227, + "learning_rate": 9.838190936166676e-05, + "loss": 0.3366, + "step": 29975 + }, + { + "epoch": 2.0800777184095485, + "grad_norm": 1.1936067342758179, + "learning_rate": 9.83762914843403e-05, + "loss": 0.3406, + "step": 29976 + }, + { + "epoch": 2.080147109846645, + "grad_norm": 1.3617491722106934, + "learning_rate": 9.837067361213967e-05, + "loss": 0.4322, + "step": 29977 + }, + { + "epoch": 2.0802165012837417, + "grad_norm": 1.4618158340454102, + "learning_rate": 9.836505574508267e-05, + "loss": 0.4773, + "step": 29978 + }, + { + "epoch": 2.080285892720838, + "grad_norm": 1.2476606369018555, + "learning_rate": 9.835943788318698e-05, + "loss": 0.6531, + "step": 29979 + }, + { + "epoch": 2.080355284157935, + "grad_norm": 1.4819846153259277, + "learning_rate": 9.835382002647035e-05, + "loss": 0.4824, + "step": 29980 + }, + { + "epoch": 2.0804246755950317, + "grad_norm": 1.1074137687683105, + "learning_rate": 9.834820217495056e-05, + "loss": 0.406, + "step": 29981 + }, + { + "epoch": 2.080494067032128, + "grad_norm": 1.4022871255874634, + "learning_rate": 9.834258432864526e-05, + "loss": 0.4076, + "step": 29982 + }, + { + "epoch": 2.080563458469225, + "grad_norm": 1.265766978263855, + "learning_rate": 9.833696648757224e-05, + "loss": 0.5019, + "step": 29983 + }, + { + "epoch": 2.0806328499063214, + "grad_norm": 1.6392399072647095, + "learning_rate": 9.833134865174921e-05, + "loss": 0.3442, + "step": 29984 + }, + { + "epoch": 2.080702241343418, + "grad_norm": 1.300714135169983, + "learning_rate": 9.832573082119393e-05, + "loss": 0.5125, + "step": 29985 + }, + { + "epoch": 2.080771632780515, + "grad_norm": 1.443249225616455, + "learning_rate": 9.832011299592412e-05, + "loss": 0.3775, + "step": 29986 + }, + { + "epoch": 2.0808410242176114, + "grad_norm": 1.5209410190582275, + "learning_rate": 9.831449517595752e-05, + "loss": 0.4378, + "step": 29987 + }, + { + "epoch": 2.0809104156547082, + "grad_norm": 1.3962645530700684, + "learning_rate": 9.83088773613119e-05, + "loss": 0.468, + "step": 29988 + }, + { + "epoch": 2.080979807091805, + "grad_norm": 1.4301596879959106, + "learning_rate": 9.830325955200493e-05, + "loss": 0.5215, + "step": 29989 + }, + { + "epoch": 2.0810491985289015, + "grad_norm": 1.2217812538146973, + "learning_rate": 9.829764174805435e-05, + "loss": 0.4901, + "step": 29990 + }, + { + "epoch": 2.0811185899659983, + "grad_norm": 1.2211440801620483, + "learning_rate": 9.829202394947794e-05, + "loss": 0.5606, + "step": 29991 + }, + { + "epoch": 2.0811879814030947, + "grad_norm": 2.205949306488037, + "learning_rate": 9.82864061562934e-05, + "loss": 0.6481, + "step": 29992 + }, + { + "epoch": 2.0812573728401915, + "grad_norm": 1.862080693244934, + "learning_rate": 9.82807883685185e-05, + "loss": 0.4814, + "step": 29993 + }, + { + "epoch": 2.0813267642772884, + "grad_norm": 1.0451645851135254, + "learning_rate": 9.827517058617092e-05, + "loss": 0.3719, + "step": 29994 + }, + { + "epoch": 2.0813961557143847, + "grad_norm": 0.9380237460136414, + "learning_rate": 9.826955280926846e-05, + "loss": 0.3054, + "step": 29995 + }, + { + "epoch": 2.0814655471514816, + "grad_norm": 1.3744343519210815, + "learning_rate": 9.82639350378288e-05, + "loss": 0.4114, + "step": 29996 + }, + { + "epoch": 2.081534938588578, + "grad_norm": 1.9808449745178223, + "learning_rate": 9.825831727186971e-05, + "loss": 0.471, + "step": 29997 + }, + { + "epoch": 2.081604330025675, + "grad_norm": 0.8881341814994812, + "learning_rate": 9.825269951140891e-05, + "loss": 0.3181, + "step": 29998 + }, + { + "epoch": 2.0816737214627716, + "grad_norm": 1.526633620262146, + "learning_rate": 9.824708175646412e-05, + "loss": 0.6126, + "step": 29999 + }, + { + "epoch": 2.081743112899868, + "grad_norm": 1.2047356367111206, + "learning_rate": 9.824146400705307e-05, + "loss": 0.3836, + "step": 30000 + }, + { + "epoch": 2.081812504336965, + "grad_norm": 1.4035520553588867, + "learning_rate": 9.823584626319354e-05, + "loss": 0.4521, + "step": 30001 + }, + { + "epoch": 2.0818818957740617, + "grad_norm": 1.361632227897644, + "learning_rate": 9.823022852490328e-05, + "loss": 0.4737, + "step": 30002 + }, + { + "epoch": 2.081951287211158, + "grad_norm": 1.442171573638916, + "learning_rate": 9.822461079219994e-05, + "loss": 0.4542, + "step": 30003 + }, + { + "epoch": 2.082020678648255, + "grad_norm": 0.8679547905921936, + "learning_rate": 9.821899306510129e-05, + "loss": 0.441, + "step": 30004 + }, + { + "epoch": 2.0820900700853513, + "grad_norm": 1.3614259958267212, + "learning_rate": 9.821337534362507e-05, + "loss": 0.5821, + "step": 30005 + }, + { + "epoch": 2.082159461522448, + "grad_norm": 2.226093053817749, + "learning_rate": 9.820775762778903e-05, + "loss": 0.5142, + "step": 30006 + }, + { + "epoch": 2.082228852959545, + "grad_norm": 1.205101728439331, + "learning_rate": 9.820213991761087e-05, + "loss": 0.3287, + "step": 30007 + }, + { + "epoch": 2.0822982443966414, + "grad_norm": 1.2093803882598877, + "learning_rate": 9.819652221310836e-05, + "loss": 0.465, + "step": 30008 + }, + { + "epoch": 2.082367635833738, + "grad_norm": 1.2445436716079712, + "learning_rate": 9.819090451429927e-05, + "loss": 0.4108, + "step": 30009 + }, + { + "epoch": 2.0824370272708346, + "grad_norm": 3.463550329208374, + "learning_rate": 9.818528682120121e-05, + "loss": 0.6255, + "step": 30010 + }, + { + "epoch": 2.0825064187079314, + "grad_norm": 0.9710832834243774, + "learning_rate": 9.817966913383203e-05, + "loss": 0.3392, + "step": 30011 + }, + { + "epoch": 2.0825758101450282, + "grad_norm": 1.4370354413986206, + "learning_rate": 9.81740514522094e-05, + "loss": 0.5053, + "step": 30012 + }, + { + "epoch": 2.0826452015821246, + "grad_norm": 1.1375385522842407, + "learning_rate": 9.81684337763511e-05, + "loss": 0.422, + "step": 30013 + }, + { + "epoch": 2.0827145930192215, + "grad_norm": 1.4445229768753052, + "learning_rate": 9.816281610627483e-05, + "loss": 0.5036, + "step": 30014 + }, + { + "epoch": 2.0827839844563183, + "grad_norm": 1.2722686529159546, + "learning_rate": 9.815719844199834e-05, + "loss": 0.3648, + "step": 30015 + }, + { + "epoch": 2.0828533758934147, + "grad_norm": 1.2452406883239746, + "learning_rate": 9.815158078353939e-05, + "loss": 0.4519, + "step": 30016 + }, + { + "epoch": 2.0829227673305115, + "grad_norm": 0.8615474700927734, + "learning_rate": 9.814596313091564e-05, + "loss": 0.3637, + "step": 30017 + }, + { + "epoch": 2.082992158767608, + "grad_norm": 1.8900375366210938, + "learning_rate": 9.814034548414488e-05, + "loss": 0.9427, + "step": 30018 + }, + { + "epoch": 2.0830615502047047, + "grad_norm": 1.393748164176941, + "learning_rate": 9.813472784324483e-05, + "loss": 0.4783, + "step": 30019 + }, + { + "epoch": 2.0831309416418016, + "grad_norm": 1.294984221458435, + "learning_rate": 9.812911020823323e-05, + "loss": 0.3282, + "step": 30020 + }, + { + "epoch": 2.083200333078898, + "grad_norm": 1.4665278196334839, + "learning_rate": 9.81234925791278e-05, + "loss": 0.4764, + "step": 30021 + }, + { + "epoch": 2.083269724515995, + "grad_norm": 1.3183404207229614, + "learning_rate": 9.81178749559463e-05, + "loss": 0.4986, + "step": 30022 + }, + { + "epoch": 2.083339115953091, + "grad_norm": 1.4815864562988281, + "learning_rate": 9.81122573387065e-05, + "loss": 0.4856, + "step": 30023 + }, + { + "epoch": 2.083408507390188, + "grad_norm": 1.353101134300232, + "learning_rate": 9.810663972742602e-05, + "loss": 0.3297, + "step": 30024 + }, + { + "epoch": 2.083477898827285, + "grad_norm": 1.4121801853179932, + "learning_rate": 9.810102212212267e-05, + "loss": 0.3563, + "step": 30025 + }, + { + "epoch": 2.0835472902643812, + "grad_norm": 1.2479674816131592, + "learning_rate": 9.809540452281417e-05, + "loss": 0.4112, + "step": 30026 + }, + { + "epoch": 2.083616681701478, + "grad_norm": 1.334893822669983, + "learning_rate": 9.808978692951825e-05, + "loss": 0.3387, + "step": 30027 + }, + { + "epoch": 2.083686073138575, + "grad_norm": 1.4412720203399658, + "learning_rate": 9.808416934225268e-05, + "loss": 0.6003, + "step": 30028 + }, + { + "epoch": 2.0837554645756713, + "grad_norm": 1.5772855281829834, + "learning_rate": 9.807855176103515e-05, + "loss": 0.5266, + "step": 30029 + }, + { + "epoch": 2.083824856012768, + "grad_norm": 1.5798773765563965, + "learning_rate": 9.807293418588344e-05, + "loss": 0.5371, + "step": 30030 + }, + { + "epoch": 2.0838942474498645, + "grad_norm": 1.4859628677368164, + "learning_rate": 9.806731661681521e-05, + "loss": 0.471, + "step": 30031 + }, + { + "epoch": 2.0839636388869613, + "grad_norm": 1.105532169342041, + "learning_rate": 9.806169905384826e-05, + "loss": 0.4852, + "step": 30032 + }, + { + "epoch": 2.084033030324058, + "grad_norm": 1.0960437059402466, + "learning_rate": 9.80560814970003e-05, + "loss": 0.3478, + "step": 30033 + }, + { + "epoch": 2.0841024217611546, + "grad_norm": 0.9572561383247375, + "learning_rate": 9.805046394628905e-05, + "loss": 0.3419, + "step": 30034 + }, + { + "epoch": 2.0841718131982514, + "grad_norm": 0.7462886571884155, + "learning_rate": 9.804484640173224e-05, + "loss": 0.253, + "step": 30035 + }, + { + "epoch": 2.084241204635348, + "grad_norm": 1.3208345174789429, + "learning_rate": 9.80392288633477e-05, + "loss": 0.4485, + "step": 30036 + }, + { + "epoch": 2.0843105960724446, + "grad_norm": 1.3594331741333008, + "learning_rate": 9.803361133115303e-05, + "loss": 0.489, + "step": 30037 + }, + { + "epoch": 2.0843799875095415, + "grad_norm": 1.2238157987594604, + "learning_rate": 9.802799380516602e-05, + "loss": 0.4498, + "step": 30038 + }, + { + "epoch": 2.084449378946638, + "grad_norm": 1.3524919748306274, + "learning_rate": 9.802237628540441e-05, + "loss": 0.4906, + "step": 30039 + }, + { + "epoch": 2.0845187703837347, + "grad_norm": 1.5787972211837769, + "learning_rate": 9.801675877188592e-05, + "loss": 0.4003, + "step": 30040 + }, + { + "epoch": 2.0845881618208315, + "grad_norm": 0.951831579208374, + "learning_rate": 9.80111412646283e-05, + "loss": 0.27, + "step": 30041 + }, + { + "epoch": 2.084657553257928, + "grad_norm": 1.282487154006958, + "learning_rate": 9.800552376364928e-05, + "loss": 0.3542, + "step": 30042 + }, + { + "epoch": 2.0847269446950247, + "grad_norm": 1.3969823122024536, + "learning_rate": 9.799990626896662e-05, + "loss": 0.53, + "step": 30043 + }, + { + "epoch": 2.084796336132121, + "grad_norm": 1.0030035972595215, + "learning_rate": 9.799428878059798e-05, + "loss": 0.3137, + "step": 30044 + }, + { + "epoch": 2.084865727569218, + "grad_norm": 2.085925340652466, + "learning_rate": 9.798867129856116e-05, + "loss": 0.2628, + "step": 30045 + }, + { + "epoch": 2.084935119006315, + "grad_norm": 2.3298304080963135, + "learning_rate": 9.798305382287385e-05, + "loss": 0.5, + "step": 30046 + }, + { + "epoch": 2.085004510443411, + "grad_norm": 1.0181827545166016, + "learning_rate": 9.797743635355384e-05, + "loss": 0.3861, + "step": 30047 + }, + { + "epoch": 2.085073901880508, + "grad_norm": 1.348345160484314, + "learning_rate": 9.79718188906188e-05, + "loss": 0.349, + "step": 30048 + }, + { + "epoch": 2.0851432933176044, + "grad_norm": 1.7443952560424805, + "learning_rate": 9.796620143408648e-05, + "loss": 0.5885, + "step": 30049 + }, + { + "epoch": 2.0852126847547012, + "grad_norm": 1.4412742853164673, + "learning_rate": 9.796058398397467e-05, + "loss": 0.4771, + "step": 30050 + }, + { + "epoch": 2.085282076191798, + "grad_norm": 1.1781898736953735, + "learning_rate": 9.795496654030105e-05, + "loss": 0.4311, + "step": 30051 + }, + { + "epoch": 2.0853514676288945, + "grad_norm": 1.3230071067810059, + "learning_rate": 9.794934910308334e-05, + "loss": 0.4423, + "step": 30052 + }, + { + "epoch": 2.0854208590659913, + "grad_norm": 1.642909288406372, + "learning_rate": 9.794373167233929e-05, + "loss": 0.6045, + "step": 30053 + }, + { + "epoch": 2.085490250503088, + "grad_norm": 1.5893481969833374, + "learning_rate": 9.793811424808665e-05, + "loss": 0.306, + "step": 30054 + }, + { + "epoch": 2.0855596419401845, + "grad_norm": 1.3307141065597534, + "learning_rate": 9.793249683034314e-05, + "loss": 0.3511, + "step": 30055 + }, + { + "epoch": 2.0856290333772813, + "grad_norm": 1.8222764730453491, + "learning_rate": 9.792687941912651e-05, + "loss": 0.5763, + "step": 30056 + }, + { + "epoch": 2.0856984248143777, + "grad_norm": 1.2294626235961914, + "learning_rate": 9.792126201445451e-05, + "loss": 0.3743, + "step": 30057 + }, + { + "epoch": 2.0857678162514746, + "grad_norm": 1.518074631690979, + "learning_rate": 9.791564461634481e-05, + "loss": 0.4752, + "step": 30058 + }, + { + "epoch": 2.0858372076885714, + "grad_norm": 1.0336813926696777, + "learning_rate": 9.791002722481516e-05, + "loss": 0.4242, + "step": 30059 + }, + { + "epoch": 2.085906599125668, + "grad_norm": 1.5055255889892578, + "learning_rate": 9.790440983988333e-05, + "loss": 0.4167, + "step": 30060 + }, + { + "epoch": 2.0859759905627646, + "grad_norm": 1.1461889743804932, + "learning_rate": 9.789879246156705e-05, + "loss": 0.3402, + "step": 30061 + }, + { + "epoch": 2.086045381999861, + "grad_norm": 1.0643543004989624, + "learning_rate": 9.789317508988402e-05, + "loss": 0.3782, + "step": 30062 + }, + { + "epoch": 2.086114773436958, + "grad_norm": 1.1877738237380981, + "learning_rate": 9.788755772485201e-05, + "loss": 0.356, + "step": 30063 + }, + { + "epoch": 2.0861841648740547, + "grad_norm": 1.1467094421386719, + "learning_rate": 9.788194036648873e-05, + "loss": 0.341, + "step": 30064 + }, + { + "epoch": 2.086253556311151, + "grad_norm": 1.5520319938659668, + "learning_rate": 9.787632301481192e-05, + "loss": 0.3008, + "step": 30065 + }, + { + "epoch": 2.086322947748248, + "grad_norm": 1.053045392036438, + "learning_rate": 9.787070566983929e-05, + "loss": 0.3945, + "step": 30066 + }, + { + "epoch": 2.0863923391853447, + "grad_norm": 1.0521252155303955, + "learning_rate": 9.786508833158861e-05, + "loss": 0.3126, + "step": 30067 + }, + { + "epoch": 2.086461730622441, + "grad_norm": 1.3691147565841675, + "learning_rate": 9.78594710000776e-05, + "loss": 0.3707, + "step": 30068 + }, + { + "epoch": 2.086531122059538, + "grad_norm": 1.1671472787857056, + "learning_rate": 9.785385367532398e-05, + "loss": 0.4532, + "step": 30069 + }, + { + "epoch": 2.0866005134966343, + "grad_norm": 1.4755158424377441, + "learning_rate": 9.78482363573455e-05, + "loss": 0.5972, + "step": 30070 + }, + { + "epoch": 2.086669904933731, + "grad_norm": 1.248420238494873, + "learning_rate": 9.784261904615993e-05, + "loss": 0.4271, + "step": 30071 + }, + { + "epoch": 2.086739296370828, + "grad_norm": 0.9588969349861145, + "learning_rate": 9.783700174178492e-05, + "loss": 0.3166, + "step": 30072 + }, + { + "epoch": 2.0868086878079244, + "grad_norm": 1.2722978591918945, + "learning_rate": 9.783138444423824e-05, + "loss": 0.4783, + "step": 30073 + }, + { + "epoch": 2.0868780792450212, + "grad_norm": 0.9960372447967529, + "learning_rate": 9.782576715353763e-05, + "loss": 0.342, + "step": 30074 + }, + { + "epoch": 2.0869474706821176, + "grad_norm": 0.9674511551856995, + "learning_rate": 9.782014986970082e-05, + "loss": 0.4404, + "step": 30075 + }, + { + "epoch": 2.0870168621192144, + "grad_norm": 1.2990326881408691, + "learning_rate": 9.781453259274555e-05, + "loss": 0.4628, + "step": 30076 + }, + { + "epoch": 2.0870862535563113, + "grad_norm": 1.5380505323410034, + "learning_rate": 9.780891532268953e-05, + "loss": 0.4949, + "step": 30077 + }, + { + "epoch": 2.0871556449934077, + "grad_norm": 1.4300605058670044, + "learning_rate": 9.780329805955058e-05, + "loss": 0.4144, + "step": 30078 + }, + { + "epoch": 2.0872250364305045, + "grad_norm": 1.6394283771514893, + "learning_rate": 9.779768080334629e-05, + "loss": 0.4643, + "step": 30079 + }, + { + "epoch": 2.0872944278676013, + "grad_norm": 1.4914859533309937, + "learning_rate": 9.779206355409448e-05, + "loss": 0.6348, + "step": 30080 + }, + { + "epoch": 2.0873638193046977, + "grad_norm": 1.1234759092330933, + "learning_rate": 9.778644631181287e-05, + "loss": 0.4266, + "step": 30081 + }, + { + "epoch": 2.0874332107417946, + "grad_norm": 1.3001400232315063, + "learning_rate": 9.778082907651922e-05, + "loss": 0.4022, + "step": 30082 + }, + { + "epoch": 2.087502602178891, + "grad_norm": 1.3290176391601562, + "learning_rate": 9.777521184823121e-05, + "loss": 0.4879, + "step": 30083 + }, + { + "epoch": 2.0875719936159878, + "grad_norm": 1.3190335035324097, + "learning_rate": 9.776959462696658e-05, + "loss": 0.393, + "step": 30084 + }, + { + "epoch": 2.0876413850530846, + "grad_norm": 1.2700244188308716, + "learning_rate": 9.776397741274313e-05, + "loss": 0.3737, + "step": 30085 + }, + { + "epoch": 2.087710776490181, + "grad_norm": 1.198657512664795, + "learning_rate": 9.775836020557852e-05, + "loss": 0.4307, + "step": 30086 + }, + { + "epoch": 2.087780167927278, + "grad_norm": 0.9151464700698853, + "learning_rate": 9.775274300549052e-05, + "loss": 0.3632, + "step": 30087 + }, + { + "epoch": 2.0878495593643747, + "grad_norm": 1.1409612894058228, + "learning_rate": 9.774712581249681e-05, + "loss": 0.3587, + "step": 30088 + }, + { + "epoch": 2.087918950801471, + "grad_norm": 0.8971983790397644, + "learning_rate": 9.774150862661518e-05, + "loss": 0.2835, + "step": 30089 + }, + { + "epoch": 2.087988342238568, + "grad_norm": 1.29948091506958, + "learning_rate": 9.773589144786335e-05, + "loss": 0.5601, + "step": 30090 + }, + { + "epoch": 2.0880577336756643, + "grad_norm": 1.593304991722107, + "learning_rate": 9.773027427625904e-05, + "loss": 0.46, + "step": 30091 + }, + { + "epoch": 2.088127125112761, + "grad_norm": 2.1930599212646484, + "learning_rate": 9.772465711182002e-05, + "loss": 0.5582, + "step": 30092 + }, + { + "epoch": 2.088196516549858, + "grad_norm": 1.1952438354492188, + "learning_rate": 9.771903995456398e-05, + "loss": 0.3494, + "step": 30093 + }, + { + "epoch": 2.0882659079869543, + "grad_norm": 1.5071347951889038, + "learning_rate": 9.771342280450865e-05, + "loss": 0.567, + "step": 30094 + }, + { + "epoch": 2.088335299424051, + "grad_norm": 1.3961217403411865, + "learning_rate": 9.770780566167178e-05, + "loss": 0.3501, + "step": 30095 + }, + { + "epoch": 2.0884046908611476, + "grad_norm": 1.1028670072555542, + "learning_rate": 9.77021885260711e-05, + "loss": 0.3841, + "step": 30096 + }, + { + "epoch": 2.0884740822982444, + "grad_norm": 1.6653895378112793, + "learning_rate": 9.769657139772435e-05, + "loss": 0.4731, + "step": 30097 + }, + { + "epoch": 2.088543473735341, + "grad_norm": 4.294885635375977, + "learning_rate": 9.769095427664929e-05, + "loss": 0.4682, + "step": 30098 + }, + { + "epoch": 2.0886128651724376, + "grad_norm": 1.3322829008102417, + "learning_rate": 9.76853371628636e-05, + "loss": 0.4924, + "step": 30099 + }, + { + "epoch": 2.0886822566095344, + "grad_norm": 1.3219331502914429, + "learning_rate": 9.7679720056385e-05, + "loss": 0.4132, + "step": 30100 + }, + { + "epoch": 2.088751648046631, + "grad_norm": 1.2869820594787598, + "learning_rate": 9.767410295723131e-05, + "loss": 0.4548, + "step": 30101 + }, + { + "epoch": 2.0888210394837277, + "grad_norm": 1.3374501466751099, + "learning_rate": 9.766848586542015e-05, + "loss": 0.5855, + "step": 30102 + }, + { + "epoch": 2.0888904309208245, + "grad_norm": 1.0762943029403687, + "learning_rate": 9.766286878096934e-05, + "loss": 0.3674, + "step": 30103 + }, + { + "epoch": 2.088959822357921, + "grad_norm": 1.6504631042480469, + "learning_rate": 9.765725170389657e-05, + "loss": 0.3958, + "step": 30104 + }, + { + "epoch": 2.0890292137950177, + "grad_norm": 1.9431158304214478, + "learning_rate": 9.765163463421958e-05, + "loss": 0.6455, + "step": 30105 + }, + { + "epoch": 2.0890986052321145, + "grad_norm": 1.3600343465805054, + "learning_rate": 9.764601757195614e-05, + "loss": 0.3884, + "step": 30106 + }, + { + "epoch": 2.089167996669211, + "grad_norm": 1.5456997156143188, + "learning_rate": 9.764040051712394e-05, + "loss": 0.445, + "step": 30107 + }, + { + "epoch": 2.0892373881063078, + "grad_norm": 1.2585047483444214, + "learning_rate": 9.76347834697407e-05, + "loss": 0.4367, + "step": 30108 + }, + { + "epoch": 2.089306779543404, + "grad_norm": 1.2860357761383057, + "learning_rate": 9.762916642982417e-05, + "loss": 0.4758, + "step": 30109 + }, + { + "epoch": 2.089376170980501, + "grad_norm": 1.256404161453247, + "learning_rate": 9.762354939739209e-05, + "loss": 0.4299, + "step": 30110 + }, + { + "epoch": 2.089445562417598, + "grad_norm": 0.9396397471427917, + "learning_rate": 9.761793237246219e-05, + "loss": 0.452, + "step": 30111 + }, + { + "epoch": 2.089514953854694, + "grad_norm": 1.1137641668319702, + "learning_rate": 9.761231535505225e-05, + "loss": 0.4189, + "step": 30112 + }, + { + "epoch": 2.089584345291791, + "grad_norm": 0.9264672994613647, + "learning_rate": 9.76066983451799e-05, + "loss": 0.341, + "step": 30113 + }, + { + "epoch": 2.089653736728888, + "grad_norm": 1.4291220903396606, + "learning_rate": 9.760108134286294e-05, + "loss": 0.4966, + "step": 30114 + }, + { + "epoch": 2.0897231281659843, + "grad_norm": 1.4390602111816406, + "learning_rate": 9.759546434811908e-05, + "loss": 0.6237, + "step": 30115 + }, + { + "epoch": 2.089792519603081, + "grad_norm": 1.0547972917556763, + "learning_rate": 9.758984736096605e-05, + "loss": 0.3396, + "step": 30116 + }, + { + "epoch": 2.0898619110401775, + "grad_norm": 1.3866453170776367, + "learning_rate": 9.758423038142163e-05, + "loss": 0.5635, + "step": 30117 + }, + { + "epoch": 2.0899313024772743, + "grad_norm": 1.3597577810287476, + "learning_rate": 9.757861340950348e-05, + "loss": 0.5767, + "step": 30118 + }, + { + "epoch": 2.090000693914371, + "grad_norm": 1.2222791910171509, + "learning_rate": 9.757299644522941e-05, + "loss": 0.5055, + "step": 30119 + }, + { + "epoch": 2.0900700853514675, + "grad_norm": 1.0239921808242798, + "learning_rate": 9.756737948861707e-05, + "loss": 0.3165, + "step": 30120 + }, + { + "epoch": 2.0901394767885644, + "grad_norm": 1.4998074769973755, + "learning_rate": 9.756176253968424e-05, + "loss": 0.4849, + "step": 30121 + }, + { + "epoch": 2.0902088682256608, + "grad_norm": 1.6486482620239258, + "learning_rate": 9.755614559844863e-05, + "loss": 0.3487, + "step": 30122 + }, + { + "epoch": 2.0902782596627576, + "grad_norm": 1.6930465698242188, + "learning_rate": 9.7550528664928e-05, + "loss": 0.5269, + "step": 30123 + }, + { + "epoch": 2.0903476510998544, + "grad_norm": 1.6669580936431885, + "learning_rate": 9.754491173914005e-05, + "loss": 0.4161, + "step": 30124 + }, + { + "epoch": 2.090417042536951, + "grad_norm": 1.4986329078674316, + "learning_rate": 9.753929482110255e-05, + "loss": 0.4315, + "step": 30125 + }, + { + "epoch": 2.0904864339740477, + "grad_norm": 1.2466191053390503, + "learning_rate": 9.753367791083324e-05, + "loss": 0.432, + "step": 30126 + }, + { + "epoch": 2.090555825411144, + "grad_norm": 1.0857822895050049, + "learning_rate": 9.752806100834977e-05, + "loss": 0.4588, + "step": 30127 + }, + { + "epoch": 2.090625216848241, + "grad_norm": 1.1155955791473389, + "learning_rate": 9.752244411366993e-05, + "loss": 0.2302, + "step": 30128 + }, + { + "epoch": 2.0906946082853377, + "grad_norm": 1.3784470558166504, + "learning_rate": 9.751682722681144e-05, + "loss": 0.4937, + "step": 30129 + }, + { + "epoch": 2.090763999722434, + "grad_norm": 1.689375400543213, + "learning_rate": 9.751121034779206e-05, + "loss": 0.5036, + "step": 30130 + }, + { + "epoch": 2.090833391159531, + "grad_norm": 1.0757256746292114, + "learning_rate": 9.750559347662949e-05, + "loss": 0.4573, + "step": 30131 + }, + { + "epoch": 2.0909027825966278, + "grad_norm": 1.3268324136734009, + "learning_rate": 9.749997661334148e-05, + "loss": 0.5528, + "step": 30132 + }, + { + "epoch": 2.090972174033724, + "grad_norm": 1.1794167757034302, + "learning_rate": 9.749435975794578e-05, + "loss": 0.4931, + "step": 30133 + }, + { + "epoch": 2.091041565470821, + "grad_norm": 1.163497805595398, + "learning_rate": 9.748874291046005e-05, + "loss": 0.4822, + "step": 30134 + }, + { + "epoch": 2.0911109569079174, + "grad_norm": 1.2436110973358154, + "learning_rate": 9.748312607090208e-05, + "loss": 0.3783, + "step": 30135 + }, + { + "epoch": 2.091180348345014, + "grad_norm": 1.2061057090759277, + "learning_rate": 9.747750923928961e-05, + "loss": 0.5197, + "step": 30136 + }, + { + "epoch": 2.091249739782111, + "grad_norm": 0.9600421190261841, + "learning_rate": 9.747189241564032e-05, + "loss": 0.3462, + "step": 30137 + }, + { + "epoch": 2.0913191312192074, + "grad_norm": 1.6790239810943604, + "learning_rate": 9.746627559997198e-05, + "loss": 0.3695, + "step": 30138 + }, + { + "epoch": 2.0913885226563043, + "grad_norm": 1.2263412475585938, + "learning_rate": 9.746065879230232e-05, + "loss": 0.5923, + "step": 30139 + }, + { + "epoch": 2.091457914093401, + "grad_norm": 1.1703071594238281, + "learning_rate": 9.74550419926491e-05, + "loss": 0.4006, + "step": 30140 + }, + { + "epoch": 2.0915273055304975, + "grad_norm": 1.7203086614608765, + "learning_rate": 9.744942520102998e-05, + "loss": 0.4069, + "step": 30141 + }, + { + "epoch": 2.0915966969675943, + "grad_norm": 1.476571798324585, + "learning_rate": 9.744380841746273e-05, + "loss": 0.4956, + "step": 30142 + }, + { + "epoch": 2.0916660884046907, + "grad_norm": 1.1901330947875977, + "learning_rate": 9.743819164196506e-05, + "loss": 0.4377, + "step": 30143 + }, + { + "epoch": 2.0917354798417875, + "grad_norm": 1.1728055477142334, + "learning_rate": 9.743257487455474e-05, + "loss": 0.3395, + "step": 30144 + }, + { + "epoch": 2.0918048712788844, + "grad_norm": 2.6351661682128906, + "learning_rate": 9.742695811524948e-05, + "loss": 0.4063, + "step": 30145 + }, + { + "epoch": 2.0918742627159808, + "grad_norm": 1.289207935333252, + "learning_rate": 9.7421341364067e-05, + "loss": 0.3809, + "step": 30146 + }, + { + "epoch": 2.0919436541530776, + "grad_norm": 1.1128751039505005, + "learning_rate": 9.741572462102512e-05, + "loss": 0.4025, + "step": 30147 + }, + { + "epoch": 2.092013045590174, + "grad_norm": 0.9013368487358093, + "learning_rate": 9.741010788614143e-05, + "loss": 0.4101, + "step": 30148 + }, + { + "epoch": 2.092082437027271, + "grad_norm": 1.4280376434326172, + "learning_rate": 9.740449115943374e-05, + "loss": 0.2204, + "step": 30149 + }, + { + "epoch": 2.0921518284643676, + "grad_norm": 1.6957861185073853, + "learning_rate": 9.739887444091976e-05, + "loss": 0.3548, + "step": 30150 + }, + { + "epoch": 2.092221219901464, + "grad_norm": 1.8825163841247559, + "learning_rate": 9.739325773061725e-05, + "loss": 0.4493, + "step": 30151 + }, + { + "epoch": 2.092290611338561, + "grad_norm": 1.3694995641708374, + "learning_rate": 9.738764102854393e-05, + "loss": 0.5063, + "step": 30152 + }, + { + "epoch": 2.0923600027756573, + "grad_norm": 1.389936089515686, + "learning_rate": 9.73820243347175e-05, + "loss": 0.4808, + "step": 30153 + }, + { + "epoch": 2.092429394212754, + "grad_norm": 1.4032400846481323, + "learning_rate": 9.737640764915575e-05, + "loss": 0.4452, + "step": 30154 + }, + { + "epoch": 2.092498785649851, + "grad_norm": 1.338302493095398, + "learning_rate": 9.737079097187637e-05, + "loss": 0.3433, + "step": 30155 + }, + { + "epoch": 2.0925681770869473, + "grad_norm": 1.1331969499588013, + "learning_rate": 9.736517430289708e-05, + "loss": 0.5476, + "step": 30156 + }, + { + "epoch": 2.092637568524044, + "grad_norm": 1.9780555963516235, + "learning_rate": 9.735955764223562e-05, + "loss": 0.6869, + "step": 30157 + }, + { + "epoch": 2.092706959961141, + "grad_norm": 1.283984661102295, + "learning_rate": 9.735394098990973e-05, + "loss": 0.4884, + "step": 30158 + }, + { + "epoch": 2.0927763513982374, + "grad_norm": 0.6814709305763245, + "learning_rate": 9.734832434593714e-05, + "loss": 0.2927, + "step": 30159 + }, + { + "epoch": 2.092845742835334, + "grad_norm": 0.8507387638092041, + "learning_rate": 9.73427077103356e-05, + "loss": 0.4503, + "step": 30160 + }, + { + "epoch": 2.0929151342724306, + "grad_norm": 1.3861271142959595, + "learning_rate": 9.733709108312286e-05, + "loss": 0.4695, + "step": 30161 + }, + { + "epoch": 2.0929845257095274, + "grad_norm": 1.4401301145553589, + "learning_rate": 9.733147446431656e-05, + "loss": 0.655, + "step": 30162 + }, + { + "epoch": 2.0930539171466243, + "grad_norm": 1.3934152126312256, + "learning_rate": 9.732585785393449e-05, + "loss": 0.5007, + "step": 30163 + }, + { + "epoch": 2.0931233085837206, + "grad_norm": 1.2154948711395264, + "learning_rate": 9.732024125199438e-05, + "loss": 0.5104, + "step": 30164 + }, + { + "epoch": 2.0931927000208175, + "grad_norm": 1.092466950416565, + "learning_rate": 9.731462465851396e-05, + "loss": 0.3698, + "step": 30165 + }, + { + "epoch": 2.0932620914579143, + "grad_norm": 1.2932201623916626, + "learning_rate": 9.730900807351095e-05, + "loss": 0.4547, + "step": 30166 + }, + { + "epoch": 2.0933314828950107, + "grad_norm": 1.0486875772476196, + "learning_rate": 9.73033914970031e-05, + "loss": 0.4846, + "step": 30167 + }, + { + "epoch": 2.0934008743321075, + "grad_norm": 1.5333534479141235, + "learning_rate": 9.729777492900816e-05, + "loss": 0.4449, + "step": 30168 + }, + { + "epoch": 2.093470265769204, + "grad_norm": 1.7216899394989014, + "learning_rate": 9.72921583695438e-05, + "loss": 0.5073, + "step": 30169 + }, + { + "epoch": 2.0935396572063008, + "grad_norm": 1.4134244918823242, + "learning_rate": 9.728654181862777e-05, + "loss": 0.5735, + "step": 30170 + }, + { + "epoch": 2.0936090486433976, + "grad_norm": 1.0347776412963867, + "learning_rate": 9.728092527627784e-05, + "loss": 0.3399, + "step": 30171 + }, + { + "epoch": 2.093678440080494, + "grad_norm": 1.0964378118515015, + "learning_rate": 9.727530874251169e-05, + "loss": 0.458, + "step": 30172 + }, + { + "epoch": 2.093747831517591, + "grad_norm": 1.413980484008789, + "learning_rate": 9.726969221734709e-05, + "loss": 0.3509, + "step": 30173 + }, + { + "epoch": 2.093817222954687, + "grad_norm": 1.5654573440551758, + "learning_rate": 9.726407570080174e-05, + "loss": 0.6255, + "step": 30174 + }, + { + "epoch": 2.093886614391784, + "grad_norm": 1.0635123252868652, + "learning_rate": 9.725845919289342e-05, + "loss": 0.337, + "step": 30175 + }, + { + "epoch": 2.093956005828881, + "grad_norm": 1.3863054513931274, + "learning_rate": 9.72528426936398e-05, + "loss": 0.2853, + "step": 30176 + }, + { + "epoch": 2.0940253972659773, + "grad_norm": 1.320102572441101, + "learning_rate": 9.724722620305863e-05, + "loss": 0.329, + "step": 30177 + }, + { + "epoch": 2.094094788703074, + "grad_norm": 1.527604341506958, + "learning_rate": 9.724160972116765e-05, + "loss": 0.4302, + "step": 30178 + }, + { + "epoch": 2.094164180140171, + "grad_norm": 1.3608275651931763, + "learning_rate": 9.723599324798458e-05, + "loss": 0.3515, + "step": 30179 + }, + { + "epoch": 2.0942335715772673, + "grad_norm": 1.0965577363967896, + "learning_rate": 9.723037678352716e-05, + "loss": 0.388, + "step": 30180 + }, + { + "epoch": 2.094302963014364, + "grad_norm": 1.2612446546554565, + "learning_rate": 9.722476032781314e-05, + "loss": 0.5593, + "step": 30181 + }, + { + "epoch": 2.0943723544514605, + "grad_norm": 1.710481882095337, + "learning_rate": 9.721914388086025e-05, + "loss": 0.5013, + "step": 30182 + }, + { + "epoch": 2.0944417458885574, + "grad_norm": 1.1270289421081543, + "learning_rate": 9.721352744268616e-05, + "loss": 0.4358, + "step": 30183 + }, + { + "epoch": 2.094511137325654, + "grad_norm": 0.9884874820709229, + "learning_rate": 9.720791101330865e-05, + "loss": 0.2885, + "step": 30184 + }, + { + "epoch": 2.0945805287627506, + "grad_norm": 1.4151188135147095, + "learning_rate": 9.720229459274543e-05, + "loss": 0.424, + "step": 30185 + }, + { + "epoch": 2.0946499201998474, + "grad_norm": 0.8440718650817871, + "learning_rate": 9.719667818101424e-05, + "loss": 0.3364, + "step": 30186 + }, + { + "epoch": 2.094719311636944, + "grad_norm": 2.920496702194214, + "learning_rate": 9.719106177813285e-05, + "loss": 0.5362, + "step": 30187 + }, + { + "epoch": 2.0947887030740406, + "grad_norm": 1.0374730825424194, + "learning_rate": 9.718544538411895e-05, + "loss": 0.3369, + "step": 30188 + }, + { + "epoch": 2.0948580945111375, + "grad_norm": 2.1513261795043945, + "learning_rate": 9.717982899899024e-05, + "loss": 0.4681, + "step": 30189 + }, + { + "epoch": 2.094927485948234, + "grad_norm": 1.8093949556350708, + "learning_rate": 9.71742126227645e-05, + "loss": 0.2941, + "step": 30190 + }, + { + "epoch": 2.0949968773853307, + "grad_norm": 1.3746007680892944, + "learning_rate": 9.716859625545944e-05, + "loss": 0.4279, + "step": 30191 + }, + { + "epoch": 2.0950662688224275, + "grad_norm": 0.9121811985969543, + "learning_rate": 9.716297989709278e-05, + "loss": 0.3552, + "step": 30192 + }, + { + "epoch": 2.095135660259524, + "grad_norm": 1.9873573780059814, + "learning_rate": 9.715736354768225e-05, + "loss": 0.5369, + "step": 30193 + }, + { + "epoch": 2.0952050516966207, + "grad_norm": 1.6013778448104858, + "learning_rate": 9.715174720724561e-05, + "loss": 0.5519, + "step": 30194 + }, + { + "epoch": 2.095274443133717, + "grad_norm": 1.2153915166854858, + "learning_rate": 9.714613087580063e-05, + "loss": 0.4602, + "step": 30195 + }, + { + "epoch": 2.095343834570814, + "grad_norm": 1.3994855880737305, + "learning_rate": 9.714051455336492e-05, + "loss": 0.5148, + "step": 30196 + }, + { + "epoch": 2.095413226007911, + "grad_norm": 1.6772894859313965, + "learning_rate": 9.713489823995627e-05, + "loss": 0.516, + "step": 30197 + }, + { + "epoch": 2.095482617445007, + "grad_norm": 1.3660762310028076, + "learning_rate": 9.712928193559244e-05, + "loss": 0.4448, + "step": 30198 + }, + { + "epoch": 2.095552008882104, + "grad_norm": 1.129697561264038, + "learning_rate": 9.712366564029111e-05, + "loss": 0.3653, + "step": 30199 + }, + { + "epoch": 2.0956214003192004, + "grad_norm": 1.0672931671142578, + "learning_rate": 9.711804935407004e-05, + "loss": 0.524, + "step": 30200 + }, + { + "epoch": 2.0956907917562972, + "grad_norm": 1.418650507926941, + "learning_rate": 9.711243307694695e-05, + "loss": 0.5783, + "step": 30201 + }, + { + "epoch": 2.095760183193394, + "grad_norm": 2.168241500854492, + "learning_rate": 9.710681680893963e-05, + "loss": 0.4609, + "step": 30202 + }, + { + "epoch": 2.0958295746304905, + "grad_norm": 1.5995819568634033, + "learning_rate": 9.71012005500657e-05, + "loss": 0.5787, + "step": 30203 + }, + { + "epoch": 2.0958989660675873, + "grad_norm": 1.5458883047103882, + "learning_rate": 9.709558430034295e-05, + "loss": 0.3822, + "step": 30204 + }, + { + "epoch": 2.095968357504684, + "grad_norm": 0.9140687584877014, + "learning_rate": 9.708996805978909e-05, + "loss": 0.4511, + "step": 30205 + }, + { + "epoch": 2.0960377489417805, + "grad_norm": 1.3833808898925781, + "learning_rate": 9.70843518284219e-05, + "loss": 0.4822, + "step": 30206 + }, + { + "epoch": 2.0961071403788774, + "grad_norm": 1.4577339887619019, + "learning_rate": 9.707873560625905e-05, + "loss": 0.3842, + "step": 30207 + }, + { + "epoch": 2.0961765318159737, + "grad_norm": 1.3081543445587158, + "learning_rate": 9.707311939331828e-05, + "loss": 0.4528, + "step": 30208 + }, + { + "epoch": 2.0962459232530706, + "grad_norm": 1.335853934288025, + "learning_rate": 9.706750318961736e-05, + "loss": 0.4321, + "step": 30209 + }, + { + "epoch": 2.0963153146901674, + "grad_norm": 1.6710236072540283, + "learning_rate": 9.7061886995174e-05, + "loss": 0.54, + "step": 30210 + }, + { + "epoch": 2.096384706127264, + "grad_norm": 1.1944080591201782, + "learning_rate": 9.705627081000589e-05, + "loss": 0.4245, + "step": 30211 + }, + { + "epoch": 2.0964540975643606, + "grad_norm": 1.1731069087982178, + "learning_rate": 9.70506546341308e-05, + "loss": 0.586, + "step": 30212 + }, + { + "epoch": 2.096523489001457, + "grad_norm": 1.372529149055481, + "learning_rate": 9.704503846756644e-05, + "loss": 0.3486, + "step": 30213 + }, + { + "epoch": 2.096592880438554, + "grad_norm": 1.143597960472107, + "learning_rate": 9.703942231033056e-05, + "loss": 0.4592, + "step": 30214 + }, + { + "epoch": 2.0966622718756507, + "grad_norm": 1.3988587856292725, + "learning_rate": 9.703380616244087e-05, + "loss": 0.4477, + "step": 30215 + }, + { + "epoch": 2.096731663312747, + "grad_norm": 1.6111699342727661, + "learning_rate": 9.702819002391516e-05, + "loss": 0.6754, + "step": 30216 + }, + { + "epoch": 2.096801054749844, + "grad_norm": 0.998439610004425, + "learning_rate": 9.702257389477107e-05, + "loss": 0.5196, + "step": 30217 + }, + { + "epoch": 2.0968704461869407, + "grad_norm": 1.6377032995224, + "learning_rate": 9.701695777502637e-05, + "loss": 0.3994, + "step": 30218 + }, + { + "epoch": 2.096939837624037, + "grad_norm": 1.3655518293380737, + "learning_rate": 9.701134166469878e-05, + "loss": 0.3694, + "step": 30219 + }, + { + "epoch": 2.097009229061134, + "grad_norm": 1.5204811096191406, + "learning_rate": 9.700572556380606e-05, + "loss": 0.5742, + "step": 30220 + }, + { + "epoch": 2.0970786204982304, + "grad_norm": 1.0628772974014282, + "learning_rate": 9.700010947236589e-05, + "loss": 0.5947, + "step": 30221 + }, + { + "epoch": 2.097148011935327, + "grad_norm": 1.9421879053115845, + "learning_rate": 9.699449339039606e-05, + "loss": 0.6573, + "step": 30222 + }, + { + "epoch": 2.097217403372424, + "grad_norm": 1.1016219854354858, + "learning_rate": 9.698887731791427e-05, + "loss": 0.3782, + "step": 30223 + }, + { + "epoch": 2.0972867948095204, + "grad_norm": 1.5957568883895874, + "learning_rate": 9.69832612549382e-05, + "loss": 0.541, + "step": 30224 + }, + { + "epoch": 2.0973561862466172, + "grad_norm": 1.333282470703125, + "learning_rate": 9.697764520148567e-05, + "loss": 0.4432, + "step": 30225 + }, + { + "epoch": 2.0974255776837136, + "grad_norm": 1.3278883695602417, + "learning_rate": 9.697202915757432e-05, + "loss": 0.3168, + "step": 30226 + }, + { + "epoch": 2.0974949691208105, + "grad_norm": 1.6823666095733643, + "learning_rate": 9.696641312322194e-05, + "loss": 0.4487, + "step": 30227 + }, + { + "epoch": 2.0975643605579073, + "grad_norm": 1.160291314125061, + "learning_rate": 9.696079709844624e-05, + "loss": 0.349, + "step": 30228 + }, + { + "epoch": 2.0976337519950037, + "grad_norm": 1.1929055452346802, + "learning_rate": 9.695518108326496e-05, + "loss": 0.4273, + "step": 30229 + }, + { + "epoch": 2.0977031434321005, + "grad_norm": 1.039244532585144, + "learning_rate": 9.694956507769585e-05, + "loss": 0.3956, + "step": 30230 + }, + { + "epoch": 2.0977725348691973, + "grad_norm": 1.6011650562286377, + "learning_rate": 9.694394908175657e-05, + "loss": 0.4514, + "step": 30231 + }, + { + "epoch": 2.0978419263062937, + "grad_norm": 1.3109320402145386, + "learning_rate": 9.693833309546488e-05, + "loss": 0.3594, + "step": 30232 + }, + { + "epoch": 2.0979113177433906, + "grad_norm": 1.5938456058502197, + "learning_rate": 9.693271711883853e-05, + "loss": 0.3845, + "step": 30233 + }, + { + "epoch": 2.097980709180487, + "grad_norm": 0.7880846261978149, + "learning_rate": 9.692710115189523e-05, + "loss": 0.2084, + "step": 30234 + }, + { + "epoch": 2.098050100617584, + "grad_norm": 1.0059633255004883, + "learning_rate": 9.69214851946527e-05, + "loss": 0.3651, + "step": 30235 + }, + { + "epoch": 2.0981194920546806, + "grad_norm": 1.2337138652801514, + "learning_rate": 9.69158692471287e-05, + "loss": 0.3131, + "step": 30236 + }, + { + "epoch": 2.098188883491777, + "grad_norm": 1.1905535459518433, + "learning_rate": 9.6910253309341e-05, + "loss": 0.3816, + "step": 30237 + }, + { + "epoch": 2.098258274928874, + "grad_norm": 0.8812781572341919, + "learning_rate": 9.69046373813072e-05, + "loss": 0.3149, + "step": 30238 + }, + { + "epoch": 2.0983276663659702, + "grad_norm": 1.5126137733459473, + "learning_rate": 9.689902146304512e-05, + "loss": 0.66, + "step": 30239 + }, + { + "epoch": 2.098397057803067, + "grad_norm": 0.90171217918396, + "learning_rate": 9.689340555457246e-05, + "loss": 0.4207, + "step": 30240 + }, + { + "epoch": 2.098466449240164, + "grad_norm": 1.0395228862762451, + "learning_rate": 9.6887789655907e-05, + "loss": 0.4257, + "step": 30241 + }, + { + "epoch": 2.0985358406772603, + "grad_norm": 1.4516112804412842, + "learning_rate": 9.688217376706638e-05, + "loss": 0.4352, + "step": 30242 + }, + { + "epoch": 2.098605232114357, + "grad_norm": 0.9152401685714722, + "learning_rate": 9.687655788806838e-05, + "loss": 0.3173, + "step": 30243 + }, + { + "epoch": 2.098674623551454, + "grad_norm": 1.2964354753494263, + "learning_rate": 9.687094201893078e-05, + "loss": 0.3945, + "step": 30244 + }, + { + "epoch": 2.0987440149885503, + "grad_norm": 1.5607837438583374, + "learning_rate": 9.68653261596712e-05, + "loss": 0.5061, + "step": 30245 + }, + { + "epoch": 2.098813406425647, + "grad_norm": 1.9465398788452148, + "learning_rate": 9.685971031030743e-05, + "loss": 0.486, + "step": 30246 + }, + { + "epoch": 2.0988827978627436, + "grad_norm": 1.0757042169570923, + "learning_rate": 9.685409447085718e-05, + "loss": 0.3815, + "step": 30247 + }, + { + "epoch": 2.0989521892998404, + "grad_norm": 1.2109369039535522, + "learning_rate": 9.68484786413382e-05, + "loss": 0.3015, + "step": 30248 + }, + { + "epoch": 2.0990215807369372, + "grad_norm": 0.9843244552612305, + "learning_rate": 9.68428628217682e-05, + "loss": 0.3388, + "step": 30249 + }, + { + "epoch": 2.0990909721740336, + "grad_norm": 0.8000674843788147, + "learning_rate": 9.683724701216491e-05, + "loss": 0.333, + "step": 30250 + }, + { + "epoch": 2.0991603636111305, + "grad_norm": 2.143383502960205, + "learning_rate": 9.683163121254612e-05, + "loss": 0.5351, + "step": 30251 + }, + { + "epoch": 2.099229755048227, + "grad_norm": 1.116452693939209, + "learning_rate": 9.682601542292944e-05, + "loss": 0.4123, + "step": 30252 + }, + { + "epoch": 2.0992991464853237, + "grad_norm": 1.3682359457015991, + "learning_rate": 9.682039964333267e-05, + "loss": 0.3685, + "step": 30253 + }, + { + "epoch": 2.0993685379224205, + "grad_norm": 1.686751127243042, + "learning_rate": 9.681478387377353e-05, + "loss": 0.5357, + "step": 30254 + }, + { + "epoch": 2.099437929359517, + "grad_norm": 1.1842868328094482, + "learning_rate": 9.680916811426977e-05, + "loss": 0.3732, + "step": 30255 + }, + { + "epoch": 2.0995073207966137, + "grad_norm": 0.8073277473449707, + "learning_rate": 9.680355236483909e-05, + "loss": 0.3021, + "step": 30256 + }, + { + "epoch": 2.0995767122337106, + "grad_norm": 1.3660439252853394, + "learning_rate": 9.679793662549921e-05, + "loss": 0.4022, + "step": 30257 + }, + { + "epoch": 2.099646103670807, + "grad_norm": 1.2534232139587402, + "learning_rate": 9.67923208962679e-05, + "loss": 0.5628, + "step": 30258 + }, + { + "epoch": 2.099715495107904, + "grad_norm": 1.203792691230774, + "learning_rate": 9.678670517716283e-05, + "loss": 0.4181, + "step": 30259 + }, + { + "epoch": 2.099784886545, + "grad_norm": 1.264856219291687, + "learning_rate": 9.678108946820179e-05, + "loss": 0.2344, + "step": 30260 + }, + { + "epoch": 2.099854277982097, + "grad_norm": 1.2884364128112793, + "learning_rate": 9.677547376940245e-05, + "loss": 0.377, + "step": 30261 + }, + { + "epoch": 2.099923669419194, + "grad_norm": 0.919234573841095, + "learning_rate": 9.676985808078256e-05, + "loss": 0.3379, + "step": 30262 + }, + { + "epoch": 2.0999930608562902, + "grad_norm": 0.8978764414787292, + "learning_rate": 9.676424240235986e-05, + "loss": 0.3269, + "step": 30263 + }, + { + "epoch": 2.100062452293387, + "grad_norm": 1.4071943759918213, + "learning_rate": 9.675862673415211e-05, + "loss": 0.4594, + "step": 30264 + }, + { + "epoch": 2.1001318437304834, + "grad_norm": 1.0683965682983398, + "learning_rate": 9.675301107617696e-05, + "loss": 0.4771, + "step": 30265 + }, + { + "epoch": 2.1002012351675803, + "grad_norm": 1.9835455417633057, + "learning_rate": 9.674739542845218e-05, + "loss": 0.4353, + "step": 30266 + }, + { + "epoch": 2.100270626604677, + "grad_norm": 0.9363054633140564, + "learning_rate": 9.674177979099549e-05, + "loss": 0.4496, + "step": 30267 + }, + { + "epoch": 2.1003400180417735, + "grad_norm": 1.4025473594665527, + "learning_rate": 9.673616416382463e-05, + "loss": 0.3875, + "step": 30268 + }, + { + "epoch": 2.1004094094788703, + "grad_norm": 1.7090083360671997, + "learning_rate": 9.673054854695732e-05, + "loss": 0.5645, + "step": 30269 + }, + { + "epoch": 2.100478800915967, + "grad_norm": 1.0990164279937744, + "learning_rate": 9.672493294041129e-05, + "loss": 0.3148, + "step": 30270 + }, + { + "epoch": 2.1005481923530636, + "grad_norm": 1.4105604887008667, + "learning_rate": 9.67193173442043e-05, + "loss": 0.4513, + "step": 30271 + }, + { + "epoch": 2.1006175837901604, + "grad_norm": 1.4460293054580688, + "learning_rate": 9.6713701758354e-05, + "loss": 0.4555, + "step": 30272 + }, + { + "epoch": 2.100686975227257, + "grad_norm": 1.228796362876892, + "learning_rate": 9.670808618287817e-05, + "loss": 0.4487, + "step": 30273 + }, + { + "epoch": 2.1007563666643536, + "grad_norm": 1.0721863508224487, + "learning_rate": 9.670247061779454e-05, + "loss": 0.3177, + "step": 30274 + }, + { + "epoch": 2.1008257581014504, + "grad_norm": 3.934948205947876, + "learning_rate": 9.669685506312081e-05, + "loss": 0.406, + "step": 30275 + }, + { + "epoch": 2.100895149538547, + "grad_norm": 1.1658865213394165, + "learning_rate": 9.669123951887476e-05, + "loss": 0.4211, + "step": 30276 + }, + { + "epoch": 2.1009645409756437, + "grad_norm": 1.8664220571517944, + "learning_rate": 9.668562398507405e-05, + "loss": 0.4848, + "step": 30277 + }, + { + "epoch": 2.10103393241274, + "grad_norm": 1.3958765268325806, + "learning_rate": 9.668000846173647e-05, + "loss": 0.3905, + "step": 30278 + }, + { + "epoch": 2.101103323849837, + "grad_norm": 1.1890113353729248, + "learning_rate": 9.667439294887971e-05, + "loss": 0.3741, + "step": 30279 + }, + { + "epoch": 2.1011727152869337, + "grad_norm": 1.6170748472213745, + "learning_rate": 9.66687774465215e-05, + "loss": 0.4551, + "step": 30280 + }, + { + "epoch": 2.10124210672403, + "grad_norm": 1.6668329238891602, + "learning_rate": 9.666316195467955e-05, + "loss": 0.4882, + "step": 30281 + }, + { + "epoch": 2.101311498161127, + "grad_norm": 0.9678457975387573, + "learning_rate": 9.665754647337163e-05, + "loss": 0.3347, + "step": 30282 + }, + { + "epoch": 2.1013808895982238, + "grad_norm": 1.455877661705017, + "learning_rate": 9.665193100261543e-05, + "loss": 0.4987, + "step": 30283 + }, + { + "epoch": 2.10145028103532, + "grad_norm": 1.2839621305465698, + "learning_rate": 9.664631554242872e-05, + "loss": 0.6043, + "step": 30284 + }, + { + "epoch": 2.101519672472417, + "grad_norm": 1.127826452255249, + "learning_rate": 9.664070009282923e-05, + "loss": 0.3848, + "step": 30285 + }, + { + "epoch": 2.1015890639095134, + "grad_norm": 1.5705310106277466, + "learning_rate": 9.663508465383461e-05, + "loss": 0.5051, + "step": 30286 + }, + { + "epoch": 2.10165845534661, + "grad_norm": 1.3026021718978882, + "learning_rate": 9.662946922546266e-05, + "loss": 0.3204, + "step": 30287 + }, + { + "epoch": 2.101727846783707, + "grad_norm": 1.7692835330963135, + "learning_rate": 9.662385380773107e-05, + "loss": 0.4733, + "step": 30288 + }, + { + "epoch": 2.1017972382208034, + "grad_norm": 1.322917103767395, + "learning_rate": 9.661823840065758e-05, + "loss": 0.4445, + "step": 30289 + }, + { + "epoch": 2.1018666296579003, + "grad_norm": 1.6728259325027466, + "learning_rate": 9.661262300425992e-05, + "loss": 0.6841, + "step": 30290 + }, + { + "epoch": 2.1019360210949967, + "grad_norm": 1.612533450126648, + "learning_rate": 9.660700761855583e-05, + "loss": 0.3235, + "step": 30291 + }, + { + "epoch": 2.1020054125320935, + "grad_norm": 1.27902090549469, + "learning_rate": 9.660139224356303e-05, + "loss": 0.4491, + "step": 30292 + }, + { + "epoch": 2.1020748039691903, + "grad_norm": 1.7807896137237549, + "learning_rate": 9.659577687929923e-05, + "loss": 0.378, + "step": 30293 + }, + { + "epoch": 2.1021441954062867, + "grad_norm": 1.2196217775344849, + "learning_rate": 9.659016152578217e-05, + "loss": 0.5896, + "step": 30294 + }, + { + "epoch": 2.1022135868433836, + "grad_norm": 1.7636399269104004, + "learning_rate": 9.658454618302957e-05, + "loss": 0.6512, + "step": 30295 + }, + { + "epoch": 2.1022829782804804, + "grad_norm": 1.1993920803070068, + "learning_rate": 9.657893085105915e-05, + "loss": 0.4439, + "step": 30296 + }, + { + "epoch": 2.1023523697175768, + "grad_norm": 2.1155216693878174, + "learning_rate": 9.657331552988866e-05, + "loss": 0.5258, + "step": 30297 + }, + { + "epoch": 2.1024217611546736, + "grad_norm": 1.3992726802825928, + "learning_rate": 9.656770021953582e-05, + "loss": 0.7061, + "step": 30298 + }, + { + "epoch": 2.10249115259177, + "grad_norm": 1.5195512771606445, + "learning_rate": 9.656208492001838e-05, + "loss": 0.3637, + "step": 30299 + }, + { + "epoch": 2.102560544028867, + "grad_norm": 1.3614513874053955, + "learning_rate": 9.655646963135399e-05, + "loss": 0.5747, + "step": 30300 + }, + { + "epoch": 2.1026299354659637, + "grad_norm": 1.1020569801330566, + "learning_rate": 9.655085435356043e-05, + "loss": 0.3513, + "step": 30301 + }, + { + "epoch": 2.10269932690306, + "grad_norm": 1.4564054012298584, + "learning_rate": 9.654523908665545e-05, + "loss": 0.6151, + "step": 30302 + }, + { + "epoch": 2.102768718340157, + "grad_norm": 1.8133071660995483, + "learning_rate": 9.653962383065672e-05, + "loss": 0.6206, + "step": 30303 + }, + { + "epoch": 2.1028381097772533, + "grad_norm": 1.3647691011428833, + "learning_rate": 9.6534008585582e-05, + "loss": 0.5354, + "step": 30304 + }, + { + "epoch": 2.10290750121435, + "grad_norm": 1.5485483407974243, + "learning_rate": 9.652839335144902e-05, + "loss": 0.4281, + "step": 30305 + }, + { + "epoch": 2.102976892651447, + "grad_norm": 1.0175490379333496, + "learning_rate": 9.652277812827554e-05, + "loss": 0.3146, + "step": 30306 + }, + { + "epoch": 2.1030462840885433, + "grad_norm": 1.5193374156951904, + "learning_rate": 9.651716291607921e-05, + "loss": 0.4685, + "step": 30307 + }, + { + "epoch": 2.10311567552564, + "grad_norm": 1.3697983026504517, + "learning_rate": 9.651154771487778e-05, + "loss": 0.633, + "step": 30308 + }, + { + "epoch": 2.103185066962737, + "grad_norm": 1.3294761180877686, + "learning_rate": 9.650593252468899e-05, + "loss": 0.3755, + "step": 30309 + }, + { + "epoch": 2.1032544583998334, + "grad_norm": 1.3998762369155884, + "learning_rate": 9.650031734553059e-05, + "loss": 0.3215, + "step": 30310 + }, + { + "epoch": 2.10332384983693, + "grad_norm": 1.5276360511779785, + "learning_rate": 9.649470217742025e-05, + "loss": 0.4945, + "step": 30311 + }, + { + "epoch": 2.1033932412740266, + "grad_norm": 1.4831502437591553, + "learning_rate": 9.648908702037576e-05, + "loss": 0.4575, + "step": 30312 + }, + { + "epoch": 2.1034626327111234, + "grad_norm": 1.5388489961624146, + "learning_rate": 9.64834718744148e-05, + "loss": 0.4832, + "step": 30313 + }, + { + "epoch": 2.1035320241482203, + "grad_norm": 0.9650977849960327, + "learning_rate": 9.647785673955515e-05, + "loss": 0.2473, + "step": 30314 + }, + { + "epoch": 2.1036014155853167, + "grad_norm": 1.7019606828689575, + "learning_rate": 9.647224161581444e-05, + "loss": 0.8638, + "step": 30315 + }, + { + "epoch": 2.1036708070224135, + "grad_norm": 2.555575370788574, + "learning_rate": 9.646662650321047e-05, + "loss": 0.3723, + "step": 30316 + }, + { + "epoch": 2.1037401984595103, + "grad_norm": 1.0395394563674927, + "learning_rate": 9.646101140176095e-05, + "loss": 0.309, + "step": 30317 + }, + { + "epoch": 2.1038095898966067, + "grad_norm": 1.4866658449172974, + "learning_rate": 9.64553963114836e-05, + "loss": 0.4348, + "step": 30318 + }, + { + "epoch": 2.1038789813337035, + "grad_norm": 1.5793066024780273, + "learning_rate": 9.644978123239617e-05, + "loss": 0.4984, + "step": 30319 + }, + { + "epoch": 2.1039483727708, + "grad_norm": 1.2676445245742798, + "learning_rate": 9.644416616451639e-05, + "loss": 0.4728, + "step": 30320 + }, + { + "epoch": 2.1040177642078968, + "grad_norm": 1.6902272701263428, + "learning_rate": 9.643855110786193e-05, + "loss": 0.3798, + "step": 30321 + }, + { + "epoch": 2.1040871556449936, + "grad_norm": 1.4223626852035522, + "learning_rate": 9.643293606245055e-05, + "loss": 0.4749, + "step": 30322 + }, + { + "epoch": 2.10415654708209, + "grad_norm": 3.029484748840332, + "learning_rate": 9.642732102829998e-05, + "loss": 0.4705, + "step": 30323 + }, + { + "epoch": 2.104225938519187, + "grad_norm": 1.2203052043914795, + "learning_rate": 9.642170600542794e-05, + "loss": 0.4255, + "step": 30324 + }, + { + "epoch": 2.104295329956283, + "grad_norm": 1.7585315704345703, + "learning_rate": 9.641609099385216e-05, + "loss": 0.5156, + "step": 30325 + }, + { + "epoch": 2.10436472139338, + "grad_norm": 1.4785728454589844, + "learning_rate": 9.641047599359039e-05, + "loss": 0.2954, + "step": 30326 + }, + { + "epoch": 2.104434112830477, + "grad_norm": 1.2250862121582031, + "learning_rate": 9.640486100466033e-05, + "loss": 0.4813, + "step": 30327 + }, + { + "epoch": 2.1045035042675733, + "grad_norm": 0.9719887375831604, + "learning_rate": 9.639924602707969e-05, + "loss": 0.3267, + "step": 30328 + }, + { + "epoch": 2.10457289570467, + "grad_norm": 1.4085787534713745, + "learning_rate": 9.63936310608662e-05, + "loss": 0.2984, + "step": 30329 + }, + { + "epoch": 2.1046422871417665, + "grad_norm": 1.5782407522201538, + "learning_rate": 9.638801610603763e-05, + "loss": 0.4598, + "step": 30330 + }, + { + "epoch": 2.1047116785788633, + "grad_norm": 1.2528252601623535, + "learning_rate": 9.638240116261163e-05, + "loss": 0.317, + "step": 30331 + }, + { + "epoch": 2.10478107001596, + "grad_norm": 2.382167339324951, + "learning_rate": 9.6376786230606e-05, + "loss": 0.4833, + "step": 30332 + }, + { + "epoch": 2.1048504614530565, + "grad_norm": 1.4772671461105347, + "learning_rate": 9.637117131003842e-05, + "loss": 0.3589, + "step": 30333 + }, + { + "epoch": 2.1049198528901534, + "grad_norm": 1.6751466989517212, + "learning_rate": 9.636555640092668e-05, + "loss": 0.4215, + "step": 30334 + }, + { + "epoch": 2.10498924432725, + "grad_norm": 1.3951387405395508, + "learning_rate": 9.635994150328841e-05, + "loss": 0.3567, + "step": 30335 + }, + { + "epoch": 2.1050586357643466, + "grad_norm": 0.9968182444572449, + "learning_rate": 9.635432661714138e-05, + "loss": 0.3881, + "step": 30336 + }, + { + "epoch": 2.1051280272014434, + "grad_norm": 1.2877004146575928, + "learning_rate": 9.634871174250333e-05, + "loss": 0.5225, + "step": 30337 + }, + { + "epoch": 2.10519741863854, + "grad_norm": 1.5707454681396484, + "learning_rate": 9.634309687939196e-05, + "loss": 0.6956, + "step": 30338 + }, + { + "epoch": 2.1052668100756367, + "grad_norm": 1.1731452941894531, + "learning_rate": 9.633748202782501e-05, + "loss": 0.5192, + "step": 30339 + }, + { + "epoch": 2.1053362015127335, + "grad_norm": 1.4613507986068726, + "learning_rate": 9.633186718782026e-05, + "loss": 0.3284, + "step": 30340 + }, + { + "epoch": 2.10540559294983, + "grad_norm": 2.2229654788970947, + "learning_rate": 9.632625235939532e-05, + "loss": 0.5137, + "step": 30341 + }, + { + "epoch": 2.1054749843869267, + "grad_norm": 1.406882643699646, + "learning_rate": 9.632063754256799e-05, + "loss": 0.4972, + "step": 30342 + }, + { + "epoch": 2.1055443758240235, + "grad_norm": 1.3864307403564453, + "learning_rate": 9.631502273735596e-05, + "loss": 0.4684, + "step": 30343 + }, + { + "epoch": 2.10561376726112, + "grad_norm": 1.6634992361068726, + "learning_rate": 9.6309407943777e-05, + "loss": 0.5019, + "step": 30344 + }, + { + "epoch": 2.1056831586982168, + "grad_norm": 1.2574940919876099, + "learning_rate": 9.630379316184882e-05, + "loss": 0.3636, + "step": 30345 + }, + { + "epoch": 2.105752550135313, + "grad_norm": 1.4795522689819336, + "learning_rate": 9.629817839158912e-05, + "loss": 0.3745, + "step": 30346 + }, + { + "epoch": 2.10582194157241, + "grad_norm": 1.0841100215911865, + "learning_rate": 9.629256363301566e-05, + "loss": 0.2624, + "step": 30347 + }, + { + "epoch": 2.105891333009507, + "grad_norm": 1.366081953048706, + "learning_rate": 9.628694888614611e-05, + "loss": 0.4443, + "step": 30348 + }, + { + "epoch": 2.105960724446603, + "grad_norm": 2.5219242572784424, + "learning_rate": 9.628133415099828e-05, + "loss": 0.5623, + "step": 30349 + }, + { + "epoch": 2.1060301158837, + "grad_norm": 1.1199122667312622, + "learning_rate": 9.62757194275898e-05, + "loss": 0.3657, + "step": 30350 + }, + { + "epoch": 2.1060995073207964, + "grad_norm": 1.1800377368927002, + "learning_rate": 9.627010471593846e-05, + "loss": 0.4098, + "step": 30351 + }, + { + "epoch": 2.1061688987578933, + "grad_norm": 1.3972208499908447, + "learning_rate": 9.626449001606196e-05, + "loss": 0.3823, + "step": 30352 + }, + { + "epoch": 2.10623829019499, + "grad_norm": 1.3110113143920898, + "learning_rate": 9.625887532797803e-05, + "loss": 0.5081, + "step": 30353 + }, + { + "epoch": 2.1063076816320865, + "grad_norm": 1.0228976011276245, + "learning_rate": 9.625326065170445e-05, + "loss": 0.4443, + "step": 30354 + }, + { + "epoch": 2.1063770730691833, + "grad_norm": 1.455538034439087, + "learning_rate": 9.624764598725883e-05, + "loss": 0.6746, + "step": 30355 + }, + { + "epoch": 2.1064464645062797, + "grad_norm": 1.2233730554580688, + "learning_rate": 9.624203133465897e-05, + "loss": 0.4849, + "step": 30356 + }, + { + "epoch": 2.1065158559433765, + "grad_norm": 1.2262986898422241, + "learning_rate": 9.623641669392258e-05, + "loss": 0.3449, + "step": 30357 + }, + { + "epoch": 2.1065852473804734, + "grad_norm": 1.5695791244506836, + "learning_rate": 9.623080206506738e-05, + "loss": 0.4131, + "step": 30358 + }, + { + "epoch": 2.1066546388175698, + "grad_norm": 1.1878597736358643, + "learning_rate": 9.62251874481111e-05, + "loss": 0.4039, + "step": 30359 + }, + { + "epoch": 2.1067240302546666, + "grad_norm": 1.3422297239303589, + "learning_rate": 9.621957284307148e-05, + "loss": 0.4159, + "step": 30360 + }, + { + "epoch": 2.1067934216917634, + "grad_norm": 1.0051766633987427, + "learning_rate": 9.621395824996626e-05, + "loss": 0.3592, + "step": 30361 + }, + { + "epoch": 2.10686281312886, + "grad_norm": 1.134962558746338, + "learning_rate": 9.620834366881308e-05, + "loss": 0.4447, + "step": 30362 + }, + { + "epoch": 2.1069322045659566, + "grad_norm": 1.4848722219467163, + "learning_rate": 9.620272909962974e-05, + "loss": 0.4616, + "step": 30363 + }, + { + "epoch": 2.107001596003053, + "grad_norm": 1.614741563796997, + "learning_rate": 9.619711454243393e-05, + "loss": 0.4821, + "step": 30364 + }, + { + "epoch": 2.10707098744015, + "grad_norm": 1.2046252489089966, + "learning_rate": 9.619149999724341e-05, + "loss": 0.4356, + "step": 30365 + }, + { + "epoch": 2.1071403788772467, + "grad_norm": 1.6149784326553345, + "learning_rate": 9.618588546407587e-05, + "loss": 0.6232, + "step": 30366 + }, + { + "epoch": 2.107209770314343, + "grad_norm": 1.076128363609314, + "learning_rate": 9.618027094294906e-05, + "loss": 0.3778, + "step": 30367 + }, + { + "epoch": 2.10727916175144, + "grad_norm": 1.665002703666687, + "learning_rate": 9.617465643388071e-05, + "loss": 0.4529, + "step": 30368 + }, + { + "epoch": 2.1073485531885368, + "grad_norm": 1.2600599527359009, + "learning_rate": 9.61690419368885e-05, + "loss": 0.4844, + "step": 30369 + }, + { + "epoch": 2.107417944625633, + "grad_norm": 1.5955193042755127, + "learning_rate": 9.616342745199017e-05, + "loss": 0.5849, + "step": 30370 + }, + { + "epoch": 2.10748733606273, + "grad_norm": 1.4029862880706787, + "learning_rate": 9.615781297920346e-05, + "loss": 0.3336, + "step": 30371 + }, + { + "epoch": 2.1075567274998264, + "grad_norm": 1.344583511352539, + "learning_rate": 9.615219851854608e-05, + "loss": 0.6021, + "step": 30372 + }, + { + "epoch": 2.107626118936923, + "grad_norm": 1.1116597652435303, + "learning_rate": 9.614658407003576e-05, + "loss": 0.3827, + "step": 30373 + }, + { + "epoch": 2.10769551037402, + "grad_norm": 2.5587899684906006, + "learning_rate": 9.614096963369023e-05, + "loss": 0.4247, + "step": 30374 + }, + { + "epoch": 2.1077649018111164, + "grad_norm": 1.1589267253875732, + "learning_rate": 9.613535520952727e-05, + "loss": 0.3497, + "step": 30375 + }, + { + "epoch": 2.1078342932482133, + "grad_norm": 1.419926643371582, + "learning_rate": 9.612974079756449e-05, + "loss": 0.5154, + "step": 30376 + }, + { + "epoch": 2.1079036846853096, + "grad_norm": 1.5923643112182617, + "learning_rate": 9.612412639781966e-05, + "loss": 0.4179, + "step": 30377 + }, + { + "epoch": 2.1079730761224065, + "grad_norm": 1.579192042350769, + "learning_rate": 9.611851201031052e-05, + "loss": 0.4637, + "step": 30378 + }, + { + "epoch": 2.1080424675595033, + "grad_norm": 1.3180850744247437, + "learning_rate": 9.61128976350548e-05, + "loss": 0.4577, + "step": 30379 + }, + { + "epoch": 2.1081118589965997, + "grad_norm": 1.0102826356887817, + "learning_rate": 9.610728327207022e-05, + "loss": 0.2239, + "step": 30380 + }, + { + "epoch": 2.1081812504336965, + "grad_norm": 1.1326984167099, + "learning_rate": 9.610166892137448e-05, + "loss": 0.3472, + "step": 30381 + }, + { + "epoch": 2.108250641870793, + "grad_norm": 2.1854028701782227, + "learning_rate": 9.609605458298534e-05, + "loss": 0.5685, + "step": 30382 + }, + { + "epoch": 2.1083200333078898, + "grad_norm": 1.2211062908172607, + "learning_rate": 9.609044025692048e-05, + "loss": 0.3836, + "step": 30383 + }, + { + "epoch": 2.1083894247449866, + "grad_norm": 1.4427040815353394, + "learning_rate": 9.608482594319766e-05, + "loss": 0.5116, + "step": 30384 + }, + { + "epoch": 2.108458816182083, + "grad_norm": 1.342729926109314, + "learning_rate": 9.607921164183458e-05, + "loss": 0.4277, + "step": 30385 + }, + { + "epoch": 2.10852820761918, + "grad_norm": 1.1084156036376953, + "learning_rate": 9.607359735284896e-05, + "loss": 0.4047, + "step": 30386 + }, + { + "epoch": 2.1085975990562766, + "grad_norm": 1.9124019145965576, + "learning_rate": 9.606798307625856e-05, + "loss": 0.4352, + "step": 30387 + }, + { + "epoch": 2.108666990493373, + "grad_norm": 2.0358428955078125, + "learning_rate": 9.606236881208106e-05, + "loss": 0.5339, + "step": 30388 + }, + { + "epoch": 2.10873638193047, + "grad_norm": 0.8418405055999756, + "learning_rate": 9.605675456033426e-05, + "loss": 0.275, + "step": 30389 + }, + { + "epoch": 2.1088057733675662, + "grad_norm": 1.056457757949829, + "learning_rate": 9.605114032103578e-05, + "loss": 0.4645, + "step": 30390 + }, + { + "epoch": 2.108875164804663, + "grad_norm": 0.9194480776786804, + "learning_rate": 9.604552609420339e-05, + "loss": 0.2851, + "step": 30391 + }, + { + "epoch": 2.10894455624176, + "grad_norm": 1.3224831819534302, + "learning_rate": 9.603991187985482e-05, + "loss": 0.3578, + "step": 30392 + }, + { + "epoch": 2.1090139476788563, + "grad_norm": 0.9396175742149353, + "learning_rate": 9.60342976780078e-05, + "loss": 0.4256, + "step": 30393 + }, + { + "epoch": 2.109083339115953, + "grad_norm": 2.1501283645629883, + "learning_rate": 9.602868348868001e-05, + "loss": 0.3259, + "step": 30394 + }, + { + "epoch": 2.10915273055305, + "grad_norm": 1.2899730205535889, + "learning_rate": 9.602306931188922e-05, + "loss": 0.4463, + "step": 30395 + }, + { + "epoch": 2.1092221219901464, + "grad_norm": 1.6543800830841064, + "learning_rate": 9.60174551476532e-05, + "loss": 0.5078, + "step": 30396 + }, + { + "epoch": 2.109291513427243, + "grad_norm": 0.9436617493629456, + "learning_rate": 9.601184099598955e-05, + "loss": 0.3097, + "step": 30397 + }, + { + "epoch": 2.1093609048643396, + "grad_norm": 1.2421023845672607, + "learning_rate": 9.600622685691605e-05, + "loss": 0.4288, + "step": 30398 + }, + { + "epoch": 2.1094302963014364, + "grad_norm": 0.9495114088058472, + "learning_rate": 9.600061273045047e-05, + "loss": 0.3532, + "step": 30399 + }, + { + "epoch": 2.1094996877385332, + "grad_norm": 1.3052729368209839, + "learning_rate": 9.599499861661047e-05, + "loss": 0.3743, + "step": 30400 + }, + { + "epoch": 2.1095690791756296, + "grad_norm": 1.0769635438919067, + "learning_rate": 9.598938451541378e-05, + "loss": 0.382, + "step": 30401 + }, + { + "epoch": 2.1096384706127265, + "grad_norm": 1.232349157333374, + "learning_rate": 9.598377042687814e-05, + "loss": 0.6342, + "step": 30402 + }, + { + "epoch": 2.109707862049823, + "grad_norm": 1.7104823589324951, + "learning_rate": 9.597815635102132e-05, + "loss": 0.4757, + "step": 30403 + }, + { + "epoch": 2.1097772534869197, + "grad_norm": 1.388197660446167, + "learning_rate": 9.597254228786095e-05, + "loss": 0.5232, + "step": 30404 + }, + { + "epoch": 2.1098466449240165, + "grad_norm": 1.7242329120635986, + "learning_rate": 9.59669282374148e-05, + "loss": 0.6317, + "step": 30405 + }, + { + "epoch": 2.109916036361113, + "grad_norm": 1.1704093217849731, + "learning_rate": 9.596131419970058e-05, + "loss": 0.3506, + "step": 30406 + }, + { + "epoch": 2.1099854277982097, + "grad_norm": 2.2752041816711426, + "learning_rate": 9.595570017473603e-05, + "loss": 0.4291, + "step": 30407 + }, + { + "epoch": 2.1100548192353066, + "grad_norm": 1.221140742301941, + "learning_rate": 9.595008616253886e-05, + "loss": 0.3854, + "step": 30408 + }, + { + "epoch": 2.110124210672403, + "grad_norm": 1.0827804803848267, + "learning_rate": 9.59444721631268e-05, + "loss": 0.4368, + "step": 30409 + }, + { + "epoch": 2.1101936021095, + "grad_norm": 0.9708098769187927, + "learning_rate": 9.593885817651762e-05, + "loss": 0.2928, + "step": 30410 + }, + { + "epoch": 2.110262993546596, + "grad_norm": 1.292245864868164, + "learning_rate": 9.593324420272894e-05, + "loss": 0.4668, + "step": 30411 + }, + { + "epoch": 2.110332384983693, + "grad_norm": 1.3454318046569824, + "learning_rate": 9.592763024177853e-05, + "loss": 0.3173, + "step": 30412 + }, + { + "epoch": 2.11040177642079, + "grad_norm": 1.089777946472168, + "learning_rate": 9.592201629368414e-05, + "loss": 0.4728, + "step": 30413 + }, + { + "epoch": 2.1104711678578862, + "grad_norm": 1.075679898262024, + "learning_rate": 9.591640235846345e-05, + "loss": 0.4665, + "step": 30414 + }, + { + "epoch": 2.110540559294983, + "grad_norm": 1.2660101652145386, + "learning_rate": 9.591078843613425e-05, + "loss": 0.405, + "step": 30415 + }, + { + "epoch": 2.1106099507320795, + "grad_norm": 1.3265752792358398, + "learning_rate": 9.590517452671419e-05, + "loss": 0.3856, + "step": 30416 + }, + { + "epoch": 2.1106793421691763, + "grad_norm": 1.4401627779006958, + "learning_rate": 9.589956063022102e-05, + "loss": 0.2824, + "step": 30417 + }, + { + "epoch": 2.110748733606273, + "grad_norm": 0.9547753930091858, + "learning_rate": 9.589394674667244e-05, + "loss": 0.3074, + "step": 30418 + }, + { + "epoch": 2.1108181250433695, + "grad_norm": 0.9901387691497803, + "learning_rate": 9.588833287608623e-05, + "loss": 0.4205, + "step": 30419 + }, + { + "epoch": 2.1108875164804664, + "grad_norm": 1.1611051559448242, + "learning_rate": 9.588271901848006e-05, + "loss": 0.3885, + "step": 30420 + }, + { + "epoch": 2.110956907917563, + "grad_norm": 1.1406396627426147, + "learning_rate": 9.587710517387166e-05, + "loss": 0.2751, + "step": 30421 + }, + { + "epoch": 2.1110262993546596, + "grad_norm": 1.1040422916412354, + "learning_rate": 9.587149134227876e-05, + "loss": 0.4057, + "step": 30422 + }, + { + "epoch": 2.1110956907917564, + "grad_norm": 1.7135429382324219, + "learning_rate": 9.586587752371913e-05, + "loss": 0.6207, + "step": 30423 + }, + { + "epoch": 2.111165082228853, + "grad_norm": 1.1231791973114014, + "learning_rate": 9.58602637182104e-05, + "loss": 0.3548, + "step": 30424 + }, + { + "epoch": 2.1112344736659496, + "grad_norm": 1.8311645984649658, + "learning_rate": 9.585464992577033e-05, + "loss": 0.3723, + "step": 30425 + }, + { + "epoch": 2.1113038651030465, + "grad_norm": 1.580915093421936, + "learning_rate": 9.584903614641665e-05, + "loss": 0.46, + "step": 30426 + }, + { + "epoch": 2.111373256540143, + "grad_norm": 1.304455041885376, + "learning_rate": 9.58434223801671e-05, + "loss": 0.3425, + "step": 30427 + }, + { + "epoch": 2.1114426479772397, + "grad_norm": 1.488835096359253, + "learning_rate": 9.583780862703936e-05, + "loss": 0.4173, + "step": 30428 + }, + { + "epoch": 2.111512039414336, + "grad_norm": 1.3075346946716309, + "learning_rate": 9.583219488705119e-05, + "loss": 0.316, + "step": 30429 + }, + { + "epoch": 2.111581430851433, + "grad_norm": 1.430762767791748, + "learning_rate": 9.582658116022033e-05, + "loss": 0.4584, + "step": 30430 + }, + { + "epoch": 2.1116508222885297, + "grad_norm": 1.1714006662368774, + "learning_rate": 9.582096744656443e-05, + "loss": 0.3838, + "step": 30431 + }, + { + "epoch": 2.111720213725626, + "grad_norm": 1.2503838539123535, + "learning_rate": 9.581535374610125e-05, + "loss": 0.4848, + "step": 30432 + }, + { + "epoch": 2.111789605162723, + "grad_norm": 1.0882807970046997, + "learning_rate": 9.580974005884852e-05, + "loss": 0.344, + "step": 30433 + }, + { + "epoch": 2.11185899659982, + "grad_norm": 1.2178065776824951, + "learning_rate": 9.580412638482397e-05, + "loss": 0.5001, + "step": 30434 + }, + { + "epoch": 2.111928388036916, + "grad_norm": 1.111873745918274, + "learning_rate": 9.579851272404527e-05, + "loss": 0.3993, + "step": 30435 + }, + { + "epoch": 2.111997779474013, + "grad_norm": 2.420414924621582, + "learning_rate": 9.57928990765302e-05, + "loss": 0.7961, + "step": 30436 + }, + { + "epoch": 2.1120671709111094, + "grad_norm": 1.2785478830337524, + "learning_rate": 9.578728544229648e-05, + "loss": 0.4435, + "step": 30437 + }, + { + "epoch": 2.1121365623482062, + "grad_norm": 1.0024068355560303, + "learning_rate": 9.57816718213618e-05, + "loss": 0.3598, + "step": 30438 + }, + { + "epoch": 2.112205953785303, + "grad_norm": 1.55971097946167, + "learning_rate": 9.577605821374388e-05, + "loss": 0.4988, + "step": 30439 + }, + { + "epoch": 2.1122753452223995, + "grad_norm": 1.4591859579086304, + "learning_rate": 9.577044461946046e-05, + "loss": 0.3556, + "step": 30440 + }, + { + "epoch": 2.1123447366594963, + "grad_norm": 1.3921170234680176, + "learning_rate": 9.576483103852924e-05, + "loss": 0.3843, + "step": 30441 + }, + { + "epoch": 2.1124141280965927, + "grad_norm": 1.1151527166366577, + "learning_rate": 9.575921747096795e-05, + "loss": 0.3309, + "step": 30442 + }, + { + "epoch": 2.1124835195336895, + "grad_norm": 1.094149112701416, + "learning_rate": 9.575360391679434e-05, + "loss": 0.4261, + "step": 30443 + }, + { + "epoch": 2.1125529109707863, + "grad_norm": 1.4355378150939941, + "learning_rate": 9.574799037602615e-05, + "loss": 0.5071, + "step": 30444 + }, + { + "epoch": 2.1126223024078827, + "grad_norm": 1.2976855039596558, + "learning_rate": 9.574237684868101e-05, + "loss": 0.4611, + "step": 30445 + }, + { + "epoch": 2.1126916938449796, + "grad_norm": 1.053110957145691, + "learning_rate": 9.573676333477669e-05, + "loss": 0.2968, + "step": 30446 + }, + { + "epoch": 2.1127610852820764, + "grad_norm": 1.458308219909668, + "learning_rate": 9.573114983433092e-05, + "loss": 0.5111, + "step": 30447 + }, + { + "epoch": 2.112830476719173, + "grad_norm": 1.3246170282363892, + "learning_rate": 9.572553634736142e-05, + "loss": 0.4463, + "step": 30448 + }, + { + "epoch": 2.1128998681562696, + "grad_norm": 1.6837812662124634, + "learning_rate": 9.57199228738859e-05, + "loss": 0.3777, + "step": 30449 + }, + { + "epoch": 2.112969259593366, + "grad_norm": 1.2977687120437622, + "learning_rate": 9.57143094139221e-05, + "loss": 0.4614, + "step": 30450 + }, + { + "epoch": 2.113038651030463, + "grad_norm": 1.1980137825012207, + "learning_rate": 9.570869596748774e-05, + "loss": 0.4602, + "step": 30451 + }, + { + "epoch": 2.1131080424675597, + "grad_norm": 1.0944641828536987, + "learning_rate": 9.57030825346005e-05, + "loss": 0.3439, + "step": 30452 + }, + { + "epoch": 2.113177433904656, + "grad_norm": 1.6397472620010376, + "learning_rate": 9.569746911527815e-05, + "loss": 0.4808, + "step": 30453 + }, + { + "epoch": 2.113246825341753, + "grad_norm": 1.116735577583313, + "learning_rate": 9.569185570953837e-05, + "loss": 0.3189, + "step": 30454 + }, + { + "epoch": 2.1133162167788493, + "grad_norm": 1.1936392784118652, + "learning_rate": 9.568624231739891e-05, + "loss": 0.3735, + "step": 30455 + }, + { + "epoch": 2.113385608215946, + "grad_norm": 0.7336491346359253, + "learning_rate": 9.568062893887746e-05, + "loss": 0.3147, + "step": 30456 + }, + { + "epoch": 2.113454999653043, + "grad_norm": 0.9312344789505005, + "learning_rate": 9.56750155739918e-05, + "loss": 0.252, + "step": 30457 + }, + { + "epoch": 2.1135243910901393, + "grad_norm": 1.2937520742416382, + "learning_rate": 9.566940222275962e-05, + "loss": 0.4129, + "step": 30458 + }, + { + "epoch": 2.113593782527236, + "grad_norm": 1.3727126121520996, + "learning_rate": 9.566378888519863e-05, + "loss": 0.5909, + "step": 30459 + }, + { + "epoch": 2.113663173964333, + "grad_norm": 2.3948850631713867, + "learning_rate": 9.565817556132653e-05, + "loss": 0.5781, + "step": 30460 + }, + { + "epoch": 2.1137325654014294, + "grad_norm": 0.864486038684845, + "learning_rate": 9.565256225116107e-05, + "loss": 0.2943, + "step": 30461 + }, + { + "epoch": 2.1138019568385262, + "grad_norm": 1.174772024154663, + "learning_rate": 9.564694895471996e-05, + "loss": 0.3665, + "step": 30462 + }, + { + "epoch": 2.1138713482756226, + "grad_norm": 2.193912982940674, + "learning_rate": 9.564133567202094e-05, + "loss": 0.4147, + "step": 30463 + }, + { + "epoch": 2.1139407397127195, + "grad_norm": 1.582560658454895, + "learning_rate": 9.563572240308172e-05, + "loss": 0.68, + "step": 30464 + }, + { + "epoch": 2.1140101311498163, + "grad_norm": 1.5628701448440552, + "learning_rate": 9.563010914792005e-05, + "loss": 0.3711, + "step": 30465 + }, + { + "epoch": 2.1140795225869127, + "grad_norm": 1.8544498682022095, + "learning_rate": 9.562449590655358e-05, + "loss": 0.4387, + "step": 30466 + }, + { + "epoch": 2.1141489140240095, + "grad_norm": 1.6753045320510864, + "learning_rate": 9.561888267900006e-05, + "loss": 0.5434, + "step": 30467 + }, + { + "epoch": 2.114218305461106, + "grad_norm": 1.875027060508728, + "learning_rate": 9.561326946527723e-05, + "loss": 0.6828, + "step": 30468 + }, + { + "epoch": 2.1142876968982027, + "grad_norm": 1.187253713607788, + "learning_rate": 9.560765626540281e-05, + "loss": 0.4304, + "step": 30469 + }, + { + "epoch": 2.1143570883352996, + "grad_norm": 1.0968756675720215, + "learning_rate": 9.56020430793945e-05, + "loss": 0.3701, + "step": 30470 + }, + { + "epoch": 2.114426479772396, + "grad_norm": 1.427374243736267, + "learning_rate": 9.559642990727002e-05, + "loss": 0.5029, + "step": 30471 + }, + { + "epoch": 2.114495871209493, + "grad_norm": 1.1893121004104614, + "learning_rate": 9.559081674904714e-05, + "loss": 0.4599, + "step": 30472 + }, + { + "epoch": 2.1145652626465896, + "grad_norm": 1.1730129718780518, + "learning_rate": 9.558520360474353e-05, + "loss": 0.4032, + "step": 30473 + }, + { + "epoch": 2.114634654083686, + "grad_norm": 1.2149996757507324, + "learning_rate": 9.557959047437689e-05, + "loss": 0.3929, + "step": 30474 + }, + { + "epoch": 2.114704045520783, + "grad_norm": 1.1726994514465332, + "learning_rate": 9.557397735796497e-05, + "loss": 0.2717, + "step": 30475 + }, + { + "epoch": 2.1147734369578792, + "grad_norm": 1.2229149341583252, + "learning_rate": 9.556836425552549e-05, + "loss": 0.4294, + "step": 30476 + }, + { + "epoch": 2.114842828394976, + "grad_norm": 1.2219513654708862, + "learning_rate": 9.556275116707618e-05, + "loss": 0.3381, + "step": 30477 + }, + { + "epoch": 2.114912219832073, + "grad_norm": 1.2324559688568115, + "learning_rate": 9.555713809263475e-05, + "loss": 0.4202, + "step": 30478 + }, + { + "epoch": 2.1149816112691693, + "grad_norm": 1.7144317626953125, + "learning_rate": 9.555152503221895e-05, + "loss": 0.5418, + "step": 30479 + }, + { + "epoch": 2.115051002706266, + "grad_norm": 1.7739537954330444, + "learning_rate": 9.554591198584644e-05, + "loss": 0.4172, + "step": 30480 + }, + { + "epoch": 2.1151203941433625, + "grad_norm": 0.9546814560890198, + "learning_rate": 9.554029895353495e-05, + "loss": 0.2348, + "step": 30481 + }, + { + "epoch": 2.1151897855804593, + "grad_norm": 1.2413357496261597, + "learning_rate": 9.553468593530223e-05, + "loss": 0.4521, + "step": 30482 + }, + { + "epoch": 2.115259177017556, + "grad_norm": 1.0967540740966797, + "learning_rate": 9.5529072931166e-05, + "loss": 0.4442, + "step": 30483 + }, + { + "epoch": 2.1153285684546526, + "grad_norm": 1.0744311809539795, + "learning_rate": 9.552345994114395e-05, + "loss": 0.4058, + "step": 30484 + }, + { + "epoch": 2.1153979598917494, + "grad_norm": 1.3659038543701172, + "learning_rate": 9.551784696525386e-05, + "loss": 0.4654, + "step": 30485 + }, + { + "epoch": 2.1154673513288462, + "grad_norm": 1.1950538158416748, + "learning_rate": 9.551223400351339e-05, + "loss": 0.382, + "step": 30486 + }, + { + "epoch": 2.1155367427659426, + "grad_norm": 1.1569123268127441, + "learning_rate": 9.550662105594025e-05, + "loss": 0.4655, + "step": 30487 + }, + { + "epoch": 2.1156061342030394, + "grad_norm": 0.8015210628509521, + "learning_rate": 9.550100812255222e-05, + "loss": 0.2727, + "step": 30488 + }, + { + "epoch": 2.115675525640136, + "grad_norm": 0.9245842099189758, + "learning_rate": 9.549539520336697e-05, + "loss": 0.2745, + "step": 30489 + }, + { + "epoch": 2.1157449170772327, + "grad_norm": 1.8461350202560425, + "learning_rate": 9.548978229840223e-05, + "loss": 0.571, + "step": 30490 + }, + { + "epoch": 2.1158143085143295, + "grad_norm": 1.2201825380325317, + "learning_rate": 9.548416940767571e-05, + "loss": 0.3796, + "step": 30491 + }, + { + "epoch": 2.115883699951426, + "grad_norm": 1.3406741619110107, + "learning_rate": 9.54785565312052e-05, + "loss": 0.5461, + "step": 30492 + }, + { + "epoch": 2.1159530913885227, + "grad_norm": 1.778027892112732, + "learning_rate": 9.54729436690083e-05, + "loss": 0.4965, + "step": 30493 + }, + { + "epoch": 2.116022482825619, + "grad_norm": 1.5929806232452393, + "learning_rate": 9.546733082110282e-05, + "loss": 0.4009, + "step": 30494 + }, + { + "epoch": 2.116091874262716, + "grad_norm": 1.5544935464859009, + "learning_rate": 9.546171798750643e-05, + "loss": 0.4231, + "step": 30495 + }, + { + "epoch": 2.1161612656998128, + "grad_norm": 1.1417186260223389, + "learning_rate": 9.545610516823687e-05, + "loss": 0.3854, + "step": 30496 + }, + { + "epoch": 2.116230657136909, + "grad_norm": 1.2955087423324585, + "learning_rate": 9.545049236331187e-05, + "loss": 0.5056, + "step": 30497 + }, + { + "epoch": 2.116300048574006, + "grad_norm": 0.824533224105835, + "learning_rate": 9.544487957274914e-05, + "loss": 0.2446, + "step": 30498 + }, + { + "epoch": 2.116369440011103, + "grad_norm": 1.1384055614471436, + "learning_rate": 9.543926679656643e-05, + "loss": 0.424, + "step": 30499 + }, + { + "epoch": 2.116438831448199, + "grad_norm": 1.5238900184631348, + "learning_rate": 9.543365403478138e-05, + "loss": 0.4334, + "step": 30500 + }, + { + "epoch": 2.116508222885296, + "grad_norm": 2.2160685062408447, + "learning_rate": 9.542804128741176e-05, + "loss": 0.8064, + "step": 30501 + }, + { + "epoch": 2.1165776143223924, + "grad_norm": 1.331322431564331, + "learning_rate": 9.54224285544753e-05, + "loss": 0.3267, + "step": 30502 + }, + { + "epoch": 2.1166470057594893, + "grad_norm": 1.6290079355239868, + "learning_rate": 9.541681583598968e-05, + "loss": 0.3975, + "step": 30503 + }, + { + "epoch": 2.116716397196586, + "grad_norm": 1.3037949800491333, + "learning_rate": 9.541120313197265e-05, + "loss": 0.4001, + "step": 30504 + }, + { + "epoch": 2.1167857886336825, + "grad_norm": 1.3775771856307983, + "learning_rate": 9.540559044244193e-05, + "loss": 0.5509, + "step": 30505 + }, + { + "epoch": 2.1168551800707793, + "grad_norm": 1.4197174310684204, + "learning_rate": 9.539997776741524e-05, + "loss": 0.4578, + "step": 30506 + }, + { + "epoch": 2.1169245715078757, + "grad_norm": 1.4693409204483032, + "learning_rate": 9.539436510691026e-05, + "loss": 0.2424, + "step": 30507 + }, + { + "epoch": 2.1169939629449726, + "grad_norm": 1.5172829627990723, + "learning_rate": 9.538875246094475e-05, + "loss": 0.4061, + "step": 30508 + }, + { + "epoch": 2.1170633543820694, + "grad_norm": 1.2462682723999023, + "learning_rate": 9.53831398295364e-05, + "loss": 0.4672, + "step": 30509 + }, + { + "epoch": 2.1171327458191658, + "grad_norm": 1.261549472808838, + "learning_rate": 9.537752721270293e-05, + "loss": 0.4133, + "step": 30510 + }, + { + "epoch": 2.1172021372562626, + "grad_norm": 1.1554814577102661, + "learning_rate": 9.537191461046209e-05, + "loss": 0.3715, + "step": 30511 + }, + { + "epoch": 2.1172715286933594, + "grad_norm": 1.5517051219940186, + "learning_rate": 9.536630202283157e-05, + "loss": 0.4986, + "step": 30512 + }, + { + "epoch": 2.117340920130456, + "grad_norm": 1.3258084058761597, + "learning_rate": 9.536068944982913e-05, + "loss": 0.545, + "step": 30513 + }, + { + "epoch": 2.1174103115675527, + "grad_norm": 1.3830538988113403, + "learning_rate": 9.535507689147242e-05, + "loss": 0.3683, + "step": 30514 + }, + { + "epoch": 2.117479703004649, + "grad_norm": 0.9105071425437927, + "learning_rate": 9.53494643477792e-05, + "loss": 0.3677, + "step": 30515 + }, + { + "epoch": 2.117549094441746, + "grad_norm": 1.5454249382019043, + "learning_rate": 9.534385181876719e-05, + "loss": 0.4699, + "step": 30516 + }, + { + "epoch": 2.1176184858788427, + "grad_norm": 1.1693400144577026, + "learning_rate": 9.533823930445407e-05, + "loss": 0.4819, + "step": 30517 + }, + { + "epoch": 2.117687877315939, + "grad_norm": 1.2011319398880005, + "learning_rate": 9.533262680485761e-05, + "loss": 0.441, + "step": 30518 + }, + { + "epoch": 2.117757268753036, + "grad_norm": 0.9462348222732544, + "learning_rate": 9.532701431999548e-05, + "loss": 0.359, + "step": 30519 + }, + { + "epoch": 2.1178266601901323, + "grad_norm": 1.39228093624115, + "learning_rate": 9.532140184988551e-05, + "loss": 0.4529, + "step": 30520 + }, + { + "epoch": 2.117896051627229, + "grad_norm": 1.180921196937561, + "learning_rate": 9.531578939454525e-05, + "loss": 0.3543, + "step": 30521 + }, + { + "epoch": 2.117965443064326, + "grad_norm": 1.5495461225509644, + "learning_rate": 9.531017695399251e-05, + "loss": 0.5145, + "step": 30522 + }, + { + "epoch": 2.1180348345014224, + "grad_norm": 1.052189826965332, + "learning_rate": 9.530456452824503e-05, + "loss": 0.4515, + "step": 30523 + }, + { + "epoch": 2.118104225938519, + "grad_norm": 1.0411921739578247, + "learning_rate": 9.529895211732046e-05, + "loss": 0.4003, + "step": 30524 + }, + { + "epoch": 2.118173617375616, + "grad_norm": 1.5184097290039062, + "learning_rate": 9.529333972123656e-05, + "loss": 0.5521, + "step": 30525 + }, + { + "epoch": 2.1182430088127124, + "grad_norm": 1.4855467081069946, + "learning_rate": 9.528772734001105e-05, + "loss": 0.4012, + "step": 30526 + }, + { + "epoch": 2.1183124002498093, + "grad_norm": 1.5467166900634766, + "learning_rate": 9.528211497366167e-05, + "loss": 0.3561, + "step": 30527 + }, + { + "epoch": 2.1183817916869057, + "grad_norm": 0.8044387698173523, + "learning_rate": 9.527650262220605e-05, + "loss": 0.3674, + "step": 30528 + }, + { + "epoch": 2.1184511831240025, + "grad_norm": 2.0236287117004395, + "learning_rate": 9.527089028566198e-05, + "loss": 0.4803, + "step": 30529 + }, + { + "epoch": 2.1185205745610993, + "grad_norm": 1.0542654991149902, + "learning_rate": 9.526527796404716e-05, + "loss": 0.3509, + "step": 30530 + }, + { + "epoch": 2.1185899659981957, + "grad_norm": 1.6527721881866455, + "learning_rate": 9.52596656573793e-05, + "loss": 0.3939, + "step": 30531 + }, + { + "epoch": 2.1186593574352925, + "grad_norm": 1.279563069343567, + "learning_rate": 9.525405336567613e-05, + "loss": 0.4675, + "step": 30532 + }, + { + "epoch": 2.118728748872389, + "grad_norm": 1.0452629327774048, + "learning_rate": 9.524844108895537e-05, + "loss": 0.4118, + "step": 30533 + }, + { + "epoch": 2.1187981403094858, + "grad_norm": 1.3756048679351807, + "learning_rate": 9.524282882723477e-05, + "loss": 0.5031, + "step": 30534 + }, + { + "epoch": 2.1188675317465826, + "grad_norm": 1.4098289012908936, + "learning_rate": 9.523721658053195e-05, + "loss": 0.3838, + "step": 30535 + }, + { + "epoch": 2.118936923183679, + "grad_norm": 1.3265825510025024, + "learning_rate": 9.52316043488647e-05, + "loss": 0.5978, + "step": 30536 + }, + { + "epoch": 2.119006314620776, + "grad_norm": 1.3205870389938354, + "learning_rate": 9.522599213225071e-05, + "loss": 0.4353, + "step": 30537 + }, + { + "epoch": 2.1190757060578727, + "grad_norm": 1.4244017601013184, + "learning_rate": 9.522037993070773e-05, + "loss": 0.4068, + "step": 30538 + }, + { + "epoch": 2.119145097494969, + "grad_norm": 1.4743833541870117, + "learning_rate": 9.521476774425345e-05, + "loss": 0.382, + "step": 30539 + }, + { + "epoch": 2.119214488932066, + "grad_norm": 0.9735744595527649, + "learning_rate": 9.52091555729056e-05, + "loss": 0.2982, + "step": 30540 + }, + { + "epoch": 2.1192838803691623, + "grad_norm": 1.6083259582519531, + "learning_rate": 9.520354341668188e-05, + "loss": 0.4745, + "step": 30541 + }, + { + "epoch": 2.119353271806259, + "grad_norm": 1.6817835569381714, + "learning_rate": 9.519793127560004e-05, + "loss": 0.4823, + "step": 30542 + }, + { + "epoch": 2.119422663243356, + "grad_norm": 1.3898024559020996, + "learning_rate": 9.519231914967775e-05, + "loss": 0.3441, + "step": 30543 + }, + { + "epoch": 2.1194920546804523, + "grad_norm": 1.2250566482543945, + "learning_rate": 9.518670703893274e-05, + "loss": 0.3614, + "step": 30544 + }, + { + "epoch": 2.119561446117549, + "grad_norm": 1.057992935180664, + "learning_rate": 9.518109494338274e-05, + "loss": 0.3442, + "step": 30545 + }, + { + "epoch": 2.119630837554646, + "grad_norm": 1.6888320446014404, + "learning_rate": 9.517548286304547e-05, + "loss": 0.4094, + "step": 30546 + }, + { + "epoch": 2.1197002289917424, + "grad_norm": 1.090700387954712, + "learning_rate": 9.516987079793865e-05, + "loss": 0.3211, + "step": 30547 + }, + { + "epoch": 2.119769620428839, + "grad_norm": 1.1776543855667114, + "learning_rate": 9.516425874808001e-05, + "loss": 0.3253, + "step": 30548 + }, + { + "epoch": 2.1198390118659356, + "grad_norm": 1.3219833374023438, + "learning_rate": 9.515864671348722e-05, + "loss": 0.5065, + "step": 30549 + }, + { + "epoch": 2.1199084033030324, + "grad_norm": 1.0258351564407349, + "learning_rate": 9.5153034694178e-05, + "loss": 0.3508, + "step": 30550 + }, + { + "epoch": 2.1199777947401293, + "grad_norm": 1.3056330680847168, + "learning_rate": 9.514742269017011e-05, + "loss": 0.3126, + "step": 30551 + }, + { + "epoch": 2.1200471861772257, + "grad_norm": 1.5104351043701172, + "learning_rate": 9.514181070148122e-05, + "loss": 0.5763, + "step": 30552 + }, + { + "epoch": 2.1201165776143225, + "grad_norm": 1.1839269399642944, + "learning_rate": 9.513619872812909e-05, + "loss": 0.4817, + "step": 30553 + }, + { + "epoch": 2.120185969051419, + "grad_norm": 1.5135397911071777, + "learning_rate": 9.513058677013143e-05, + "loss": 0.3738, + "step": 30554 + }, + { + "epoch": 2.1202553604885157, + "grad_norm": 1.4342637062072754, + "learning_rate": 9.512497482750594e-05, + "loss": 0.7646, + "step": 30555 + }, + { + "epoch": 2.1203247519256125, + "grad_norm": 1.0917338132858276, + "learning_rate": 9.511936290027032e-05, + "loss": 0.3335, + "step": 30556 + }, + { + "epoch": 2.120394143362709, + "grad_norm": 1.340877890586853, + "learning_rate": 9.511375098844231e-05, + "loss": 0.4129, + "step": 30557 + }, + { + "epoch": 2.1204635347998058, + "grad_norm": 1.3482125997543335, + "learning_rate": 9.510813909203964e-05, + "loss": 0.4898, + "step": 30558 + }, + { + "epoch": 2.120532926236902, + "grad_norm": 1.4208056926727295, + "learning_rate": 9.510252721107997e-05, + "loss": 0.4235, + "step": 30559 + }, + { + "epoch": 2.120602317673999, + "grad_norm": 1.3592746257781982, + "learning_rate": 9.509691534558107e-05, + "loss": 0.4566, + "step": 30560 + }, + { + "epoch": 2.120671709111096, + "grad_norm": 1.1105583906173706, + "learning_rate": 9.509130349556063e-05, + "loss": 0.4854, + "step": 30561 + }, + { + "epoch": 2.120741100548192, + "grad_norm": 1.1615660190582275, + "learning_rate": 9.508569166103643e-05, + "loss": 0.3534, + "step": 30562 + }, + { + "epoch": 2.120810491985289, + "grad_norm": 1.3994554281234741, + "learning_rate": 9.508007984202609e-05, + "loss": 0.359, + "step": 30563 + }, + { + "epoch": 2.120879883422386, + "grad_norm": 1.1109657287597656, + "learning_rate": 9.507446803854736e-05, + "loss": 0.4146, + "step": 30564 + }, + { + "epoch": 2.1209492748594823, + "grad_norm": 1.1150553226470947, + "learning_rate": 9.506885625061795e-05, + "loss": 0.3749, + "step": 30565 + }, + { + "epoch": 2.121018666296579, + "grad_norm": 1.2725046873092651, + "learning_rate": 9.506324447825561e-05, + "loss": 0.4965, + "step": 30566 + }, + { + "epoch": 2.1210880577336755, + "grad_norm": 0.8863480091094971, + "learning_rate": 9.505763272147803e-05, + "loss": 0.2573, + "step": 30567 + }, + { + "epoch": 2.1211574491707723, + "grad_norm": 1.1944854259490967, + "learning_rate": 9.505202098030297e-05, + "loss": 0.4525, + "step": 30568 + }, + { + "epoch": 2.121226840607869, + "grad_norm": 1.8075356483459473, + "learning_rate": 9.504640925474806e-05, + "loss": 0.7328, + "step": 30569 + }, + { + "epoch": 2.1212962320449655, + "grad_norm": 1.5409451723098755, + "learning_rate": 9.504079754483106e-05, + "loss": 0.3716, + "step": 30570 + }, + { + "epoch": 2.1213656234820624, + "grad_norm": 1.4577020406723022, + "learning_rate": 9.503518585056969e-05, + "loss": 0.4619, + "step": 30571 + }, + { + "epoch": 2.121435014919159, + "grad_norm": 1.3457331657409668, + "learning_rate": 9.502957417198166e-05, + "loss": 0.4598, + "step": 30572 + }, + { + "epoch": 2.1215044063562556, + "grad_norm": 1.702864646911621, + "learning_rate": 9.502396250908468e-05, + "loss": 0.5263, + "step": 30573 + }, + { + "epoch": 2.1215737977933524, + "grad_norm": 1.176858901977539, + "learning_rate": 9.50183508618965e-05, + "loss": 0.3815, + "step": 30574 + }, + { + "epoch": 2.121643189230449, + "grad_norm": 1.431075930595398, + "learning_rate": 9.501273923043483e-05, + "loss": 0.4348, + "step": 30575 + }, + { + "epoch": 2.1217125806675456, + "grad_norm": 1.4348293542861938, + "learning_rate": 9.500712761471732e-05, + "loss": 0.3618, + "step": 30576 + }, + { + "epoch": 2.1217819721046425, + "grad_norm": 1.1845580339431763, + "learning_rate": 9.500151601476176e-05, + "loss": 0.3329, + "step": 30577 + }, + { + "epoch": 2.121851363541739, + "grad_norm": 1.1301631927490234, + "learning_rate": 9.49959044305858e-05, + "loss": 0.4026, + "step": 30578 + }, + { + "epoch": 2.1219207549788357, + "grad_norm": 0.978385865688324, + "learning_rate": 9.499029286220719e-05, + "loss": 0.2592, + "step": 30579 + }, + { + "epoch": 2.121990146415932, + "grad_norm": 1.396755337715149, + "learning_rate": 9.498468130964367e-05, + "loss": 0.4951, + "step": 30580 + }, + { + "epoch": 2.122059537853029, + "grad_norm": 1.190539836883545, + "learning_rate": 9.49790697729129e-05, + "loss": 0.465, + "step": 30581 + }, + { + "epoch": 2.1221289292901258, + "grad_norm": 2.1231460571289062, + "learning_rate": 9.497345825203267e-05, + "loss": 0.547, + "step": 30582 + }, + { + "epoch": 2.122198320727222, + "grad_norm": 1.6423715353012085, + "learning_rate": 9.496784674702062e-05, + "loss": 0.62, + "step": 30583 + }, + { + "epoch": 2.122267712164319, + "grad_norm": 1.2797201871871948, + "learning_rate": 9.49622352578945e-05, + "loss": 0.3447, + "step": 30584 + }, + { + "epoch": 2.1223371036014154, + "grad_norm": 1.287554144859314, + "learning_rate": 9.495662378467199e-05, + "loss": 0.3937, + "step": 30585 + }, + { + "epoch": 2.122406495038512, + "grad_norm": 1.5492788553237915, + "learning_rate": 9.495101232737086e-05, + "loss": 0.5254, + "step": 30586 + }, + { + "epoch": 2.122475886475609, + "grad_norm": 1.5986415147781372, + "learning_rate": 9.494540088600879e-05, + "loss": 0.6305, + "step": 30587 + }, + { + "epoch": 2.1225452779127054, + "grad_norm": 1.1141867637634277, + "learning_rate": 9.493978946060349e-05, + "loss": 0.452, + "step": 30588 + }, + { + "epoch": 2.1226146693498023, + "grad_norm": 1.4935846328735352, + "learning_rate": 9.493417805117275e-05, + "loss": 0.4843, + "step": 30589 + }, + { + "epoch": 2.122684060786899, + "grad_norm": 1.4203888177871704, + "learning_rate": 9.492856665773416e-05, + "loss": 0.2831, + "step": 30590 + }, + { + "epoch": 2.1227534522239955, + "grad_norm": 1.410009741783142, + "learning_rate": 9.492295528030551e-05, + "loss": 0.5089, + "step": 30591 + }, + { + "epoch": 2.1228228436610923, + "grad_norm": 1.121528148651123, + "learning_rate": 9.49173439189045e-05, + "loss": 0.4722, + "step": 30592 + }, + { + "epoch": 2.1228922350981887, + "grad_norm": 1.8475711345672607, + "learning_rate": 9.491173257354886e-05, + "loss": 0.5743, + "step": 30593 + }, + { + "epoch": 2.1229616265352855, + "grad_norm": 1.4119998216629028, + "learning_rate": 9.490612124425628e-05, + "loss": 0.4711, + "step": 30594 + }, + { + "epoch": 2.1230310179723824, + "grad_norm": 1.4221407175064087, + "learning_rate": 9.490050993104448e-05, + "loss": 0.3936, + "step": 30595 + }, + { + "epoch": 2.1231004094094788, + "grad_norm": 1.176054835319519, + "learning_rate": 9.489489863393121e-05, + "loss": 0.442, + "step": 30596 + }, + { + "epoch": 2.1231698008465756, + "grad_norm": 1.7167755365371704, + "learning_rate": 9.488928735293414e-05, + "loss": 0.398, + "step": 30597 + }, + { + "epoch": 2.1232391922836724, + "grad_norm": 1.5145102739334106, + "learning_rate": 9.488367608807096e-05, + "loss": 0.3717, + "step": 30598 + }, + { + "epoch": 2.123308583720769, + "grad_norm": 1.393190622329712, + "learning_rate": 9.487806483935944e-05, + "loss": 0.4494, + "step": 30599 + }, + { + "epoch": 2.1233779751578656, + "grad_norm": 1.2980859279632568, + "learning_rate": 9.487245360681726e-05, + "loss": 0.4082, + "step": 30600 + }, + { + "epoch": 2.123447366594962, + "grad_norm": 1.427463412284851, + "learning_rate": 9.486684239046218e-05, + "loss": 0.3769, + "step": 30601 + }, + { + "epoch": 2.123516758032059, + "grad_norm": 1.5526728630065918, + "learning_rate": 9.486123119031186e-05, + "loss": 0.5022, + "step": 30602 + }, + { + "epoch": 2.1235861494691557, + "grad_norm": 1.8838543891906738, + "learning_rate": 9.485562000638408e-05, + "loss": 0.4899, + "step": 30603 + }, + { + "epoch": 2.123655540906252, + "grad_norm": 1.3379677534103394, + "learning_rate": 9.485000883869649e-05, + "loss": 0.4475, + "step": 30604 + }, + { + "epoch": 2.123724932343349, + "grad_norm": 1.3626734018325806, + "learning_rate": 9.48443976872668e-05, + "loss": 0.4034, + "step": 30605 + }, + { + "epoch": 2.1237943237804453, + "grad_norm": 0.8707955479621887, + "learning_rate": 9.483878655211275e-05, + "loss": 0.2775, + "step": 30606 + }, + { + "epoch": 2.123863715217542, + "grad_norm": 1.599665880203247, + "learning_rate": 9.483317543325205e-05, + "loss": 0.4245, + "step": 30607 + }, + { + "epoch": 2.123933106654639, + "grad_norm": 1.4839831590652466, + "learning_rate": 9.482756433070244e-05, + "loss": 0.4954, + "step": 30608 + }, + { + "epoch": 2.1240024980917354, + "grad_norm": 1.3698644638061523, + "learning_rate": 9.482195324448159e-05, + "loss": 0.4581, + "step": 30609 + }, + { + "epoch": 2.124071889528832, + "grad_norm": 1.192555546760559, + "learning_rate": 9.481634217460726e-05, + "loss": 0.4475, + "step": 30610 + }, + { + "epoch": 2.1241412809659286, + "grad_norm": 1.5034528970718384, + "learning_rate": 9.481073112109712e-05, + "loss": 0.4328, + "step": 30611 + }, + { + "epoch": 2.1242106724030254, + "grad_norm": 0.9814884066581726, + "learning_rate": 9.480512008396892e-05, + "loss": 0.3418, + "step": 30612 + }, + { + "epoch": 2.1242800638401222, + "grad_norm": 1.122562289237976, + "learning_rate": 9.47995090632403e-05, + "loss": 0.3634, + "step": 30613 + }, + { + "epoch": 2.1243494552772186, + "grad_norm": 1.2264991998672485, + "learning_rate": 9.479389805892906e-05, + "loss": 0.5326, + "step": 30614 + }, + { + "epoch": 2.1244188467143155, + "grad_norm": 1.0807082653045654, + "learning_rate": 9.478828707105287e-05, + "loss": 0.4677, + "step": 30615 + }, + { + "epoch": 2.1244882381514123, + "grad_norm": 1.7715070247650146, + "learning_rate": 9.478267609962945e-05, + "loss": 0.4658, + "step": 30616 + }, + { + "epoch": 2.1245576295885087, + "grad_norm": 1.305924415588379, + "learning_rate": 9.477706514467656e-05, + "loss": 0.4806, + "step": 30617 + }, + { + "epoch": 2.1246270210256055, + "grad_norm": 1.1874498128890991, + "learning_rate": 9.477145420621184e-05, + "loss": 0.5126, + "step": 30618 + }, + { + "epoch": 2.124696412462702, + "grad_norm": 1.4193446636199951, + "learning_rate": 9.476584328425302e-05, + "loss": 0.4082, + "step": 30619 + }, + { + "epoch": 2.1247658038997987, + "grad_norm": 1.2919259071350098, + "learning_rate": 9.476023237881783e-05, + "loss": 0.6387, + "step": 30620 + }, + { + "epoch": 2.1248351953368956, + "grad_norm": 1.584975242614746, + "learning_rate": 9.475462148992397e-05, + "loss": 0.5049, + "step": 30621 + }, + { + "epoch": 2.124904586773992, + "grad_norm": 2.1638717651367188, + "learning_rate": 9.474901061758918e-05, + "loss": 0.5567, + "step": 30622 + }, + { + "epoch": 2.124973978211089, + "grad_norm": 1.6924670934677124, + "learning_rate": 9.474339976183113e-05, + "loss": 0.4689, + "step": 30623 + }, + { + "epoch": 2.1250433696481856, + "grad_norm": 1.0286649465560913, + "learning_rate": 9.473778892266762e-05, + "loss": 0.4655, + "step": 30624 + }, + { + "epoch": 2.125112761085282, + "grad_norm": 1.4640098810195923, + "learning_rate": 9.473217810011624e-05, + "loss": 0.4329, + "step": 30625 + }, + { + "epoch": 2.125182152522379, + "grad_norm": 1.2699977159500122, + "learning_rate": 9.472656729419476e-05, + "loss": 0.5152, + "step": 30626 + }, + { + "epoch": 2.1252515439594752, + "grad_norm": 1.0865834951400757, + "learning_rate": 9.472095650492092e-05, + "loss": 0.3872, + "step": 30627 + }, + { + "epoch": 2.125320935396572, + "grad_norm": 1.7222503423690796, + "learning_rate": 9.471534573231241e-05, + "loss": 0.5556, + "step": 30628 + }, + { + "epoch": 2.125390326833669, + "grad_norm": 1.1995872259140015, + "learning_rate": 9.470973497638694e-05, + "loss": 0.3557, + "step": 30629 + }, + { + "epoch": 2.1254597182707653, + "grad_norm": 1.4415515661239624, + "learning_rate": 9.470412423716219e-05, + "loss": 0.5951, + "step": 30630 + }, + { + "epoch": 2.125529109707862, + "grad_norm": 1.1008554697036743, + "learning_rate": 9.469851351465598e-05, + "loss": 0.3311, + "step": 30631 + }, + { + "epoch": 2.1255985011449585, + "grad_norm": 1.290783166885376, + "learning_rate": 9.469290280888587e-05, + "loss": 0.5, + "step": 30632 + }, + { + "epoch": 2.1256678925820554, + "grad_norm": 1.2983940839767456, + "learning_rate": 9.468729211986967e-05, + "loss": 0.5428, + "step": 30633 + }, + { + "epoch": 2.125737284019152, + "grad_norm": 1.1728355884552002, + "learning_rate": 9.468168144762508e-05, + "loss": 0.4466, + "step": 30634 + }, + { + "epoch": 2.1258066754562486, + "grad_norm": 1.5269488096237183, + "learning_rate": 9.46760707921698e-05, + "loss": 0.6195, + "step": 30635 + }, + { + "epoch": 2.1258760668933454, + "grad_norm": 1.1398383378982544, + "learning_rate": 9.467046015352154e-05, + "loss": 0.4729, + "step": 30636 + }, + { + "epoch": 2.125945458330442, + "grad_norm": 1.1550915241241455, + "learning_rate": 9.466484953169803e-05, + "loss": 0.4122, + "step": 30637 + }, + { + "epoch": 2.1260148497675386, + "grad_norm": 1.20350182056427, + "learning_rate": 9.465923892671701e-05, + "loss": 0.2935, + "step": 30638 + }, + { + "epoch": 2.1260842412046355, + "grad_norm": 0.8475693464279175, + "learning_rate": 9.46536283385961e-05, + "loss": 0.2781, + "step": 30639 + }, + { + "epoch": 2.126153632641732, + "grad_norm": 1.1819140911102295, + "learning_rate": 9.464801776735307e-05, + "loss": 0.4255, + "step": 30640 + }, + { + "epoch": 2.1262230240788287, + "grad_norm": 0.8295304775238037, + "learning_rate": 9.464240721300563e-05, + "loss": 0.3256, + "step": 30641 + }, + { + "epoch": 2.1262924155159255, + "grad_norm": 1.6274688243865967, + "learning_rate": 9.46367966755715e-05, + "loss": 0.5279, + "step": 30642 + }, + { + "epoch": 2.126361806953022, + "grad_norm": 1.3679029941558838, + "learning_rate": 9.463118615506838e-05, + "loss": 0.5759, + "step": 30643 + }, + { + "epoch": 2.1264311983901187, + "grad_norm": 1.29278564453125, + "learning_rate": 9.462557565151396e-05, + "loss": 0.5665, + "step": 30644 + }, + { + "epoch": 2.126500589827215, + "grad_norm": 1.1113932132720947, + "learning_rate": 9.461996516492602e-05, + "loss": 0.3891, + "step": 30645 + }, + { + "epoch": 2.126569981264312, + "grad_norm": 0.9393777251243591, + "learning_rate": 9.461435469532218e-05, + "loss": 0.299, + "step": 30646 + }, + { + "epoch": 2.126639372701409, + "grad_norm": 1.1733063459396362, + "learning_rate": 9.460874424272024e-05, + "loss": 0.3209, + "step": 30647 + }, + { + "epoch": 2.126708764138505, + "grad_norm": 1.078351616859436, + "learning_rate": 9.460313380713783e-05, + "loss": 0.4128, + "step": 30648 + }, + { + "epoch": 2.126778155575602, + "grad_norm": 1.5556360483169556, + "learning_rate": 9.459752338859271e-05, + "loss": 0.3858, + "step": 30649 + }, + { + "epoch": 2.126847547012699, + "grad_norm": 1.2257601022720337, + "learning_rate": 9.459191298710259e-05, + "loss": 0.3419, + "step": 30650 + }, + { + "epoch": 2.1269169384497952, + "grad_norm": 1.092769742012024, + "learning_rate": 9.45863026026852e-05, + "loss": 0.2911, + "step": 30651 + }, + { + "epoch": 2.126986329886892, + "grad_norm": 1.7110673189163208, + "learning_rate": 9.458069223535818e-05, + "loss": 0.3832, + "step": 30652 + }, + { + "epoch": 2.1270557213239885, + "grad_norm": 0.8277124762535095, + "learning_rate": 9.457508188513929e-05, + "loss": 0.366, + "step": 30653 + }, + { + "epoch": 2.1271251127610853, + "grad_norm": 1.3758211135864258, + "learning_rate": 9.456947155204625e-05, + "loss": 0.4448, + "step": 30654 + }, + { + "epoch": 2.127194504198182, + "grad_norm": 2.2362780570983887, + "learning_rate": 9.456386123609674e-05, + "loss": 0.6679, + "step": 30655 + }, + { + "epoch": 2.1272638956352785, + "grad_norm": 1.7707713842391968, + "learning_rate": 9.45582509373085e-05, + "loss": 0.6848, + "step": 30656 + }, + { + "epoch": 2.1273332870723753, + "grad_norm": 1.324841022491455, + "learning_rate": 9.455264065569923e-05, + "loss": 0.3932, + "step": 30657 + }, + { + "epoch": 2.1274026785094717, + "grad_norm": 1.3898937702178955, + "learning_rate": 9.454703039128668e-05, + "loss": 0.3538, + "step": 30658 + }, + { + "epoch": 2.1274720699465686, + "grad_norm": 1.659550428390503, + "learning_rate": 9.45414201440885e-05, + "loss": 0.4742, + "step": 30659 + }, + { + "epoch": 2.1275414613836654, + "grad_norm": 1.8611056804656982, + "learning_rate": 9.453580991412238e-05, + "loss": 0.4274, + "step": 30660 + }, + { + "epoch": 2.127610852820762, + "grad_norm": 1.269333004951477, + "learning_rate": 9.45301997014061e-05, + "loss": 0.5477, + "step": 30661 + }, + { + "epoch": 2.1276802442578586, + "grad_norm": 1.2211191654205322, + "learning_rate": 9.452458950595735e-05, + "loss": 0.4252, + "step": 30662 + }, + { + "epoch": 2.127749635694955, + "grad_norm": 1.4217299222946167, + "learning_rate": 9.451897932779386e-05, + "loss": 0.3593, + "step": 30663 + }, + { + "epoch": 2.127819027132052, + "grad_norm": 1.180715560913086, + "learning_rate": 9.451336916693329e-05, + "loss": 0.4182, + "step": 30664 + }, + { + "epoch": 2.1278884185691487, + "grad_norm": 1.1226699352264404, + "learning_rate": 9.45077590233934e-05, + "loss": 0.3093, + "step": 30665 + }, + { + "epoch": 2.127957810006245, + "grad_norm": 1.165877342224121, + "learning_rate": 9.450214889719187e-05, + "loss": 0.4779, + "step": 30666 + }, + { + "epoch": 2.128027201443342, + "grad_norm": 1.211106538772583, + "learning_rate": 9.449653878834639e-05, + "loss": 0.3337, + "step": 30667 + }, + { + "epoch": 2.1280965928804387, + "grad_norm": 1.330114722251892, + "learning_rate": 9.44909286968747e-05, + "loss": 0.4029, + "step": 30668 + }, + { + "epoch": 2.128165984317535, + "grad_norm": 1.3878300189971924, + "learning_rate": 9.448531862279453e-05, + "loss": 0.4415, + "step": 30669 + }, + { + "epoch": 2.128235375754632, + "grad_norm": 1.7983680963516235, + "learning_rate": 9.447970856612354e-05, + "loss": 0.4771, + "step": 30670 + }, + { + "epoch": 2.1283047671917283, + "grad_norm": 1.007087230682373, + "learning_rate": 9.447409852687949e-05, + "loss": 0.3367, + "step": 30671 + }, + { + "epoch": 2.128374158628825, + "grad_norm": 1.0713756084442139, + "learning_rate": 9.446848850508011e-05, + "loss": 0.405, + "step": 30672 + }, + { + "epoch": 2.128443550065922, + "grad_norm": 1.4474658966064453, + "learning_rate": 9.446287850074302e-05, + "loss": 0.5469, + "step": 30673 + }, + { + "epoch": 2.1285129415030184, + "grad_norm": 1.4420167207717896, + "learning_rate": 9.445726851388599e-05, + "loss": 0.6569, + "step": 30674 + }, + { + "epoch": 2.1285823329401152, + "grad_norm": 1.1501243114471436, + "learning_rate": 9.445165854452673e-05, + "loss": 0.4129, + "step": 30675 + }, + { + "epoch": 2.128651724377212, + "grad_norm": 1.18562912940979, + "learning_rate": 9.444604859268291e-05, + "loss": 0.3855, + "step": 30676 + }, + { + "epoch": 2.1287211158143085, + "grad_norm": 1.3581171035766602, + "learning_rate": 9.444043865837229e-05, + "loss": 0.3187, + "step": 30677 + }, + { + "epoch": 2.1287905072514053, + "grad_norm": 2.148458480834961, + "learning_rate": 9.443482874161258e-05, + "loss": 0.4456, + "step": 30678 + }, + { + "epoch": 2.1288598986885017, + "grad_norm": 1.8941638469696045, + "learning_rate": 9.442921884242147e-05, + "loss": 0.4229, + "step": 30679 + }, + { + "epoch": 2.1289292901255985, + "grad_norm": 1.1654144525527954, + "learning_rate": 9.442360896081664e-05, + "loss": 0.3591, + "step": 30680 + }, + { + "epoch": 2.1289986815626953, + "grad_norm": 1.037611722946167, + "learning_rate": 9.441799909681583e-05, + "loss": 0.2968, + "step": 30681 + }, + { + "epoch": 2.1290680729997917, + "grad_norm": 1.1423118114471436, + "learning_rate": 9.441238925043678e-05, + "loss": 0.426, + "step": 30682 + }, + { + "epoch": 2.1291374644368886, + "grad_norm": 1.748112440109253, + "learning_rate": 9.440677942169715e-05, + "loss": 0.5886, + "step": 30683 + }, + { + "epoch": 2.129206855873985, + "grad_norm": 1.8529154062271118, + "learning_rate": 9.440116961061466e-05, + "loss": 0.4829, + "step": 30684 + }, + { + "epoch": 2.129276247311082, + "grad_norm": 1.7701141834259033, + "learning_rate": 9.439555981720703e-05, + "loss": 0.5528, + "step": 30685 + }, + { + "epoch": 2.1293456387481786, + "grad_norm": 1.1493377685546875, + "learning_rate": 9.4389950041492e-05, + "loss": 0.5218, + "step": 30686 + }, + { + "epoch": 2.129415030185275, + "grad_norm": 0.9529268741607666, + "learning_rate": 9.438434028348723e-05, + "loss": 0.3511, + "step": 30687 + }, + { + "epoch": 2.129484421622372, + "grad_norm": 1.3265517950057983, + "learning_rate": 9.437873054321042e-05, + "loss": 0.3548, + "step": 30688 + }, + { + "epoch": 2.1295538130594682, + "grad_norm": 1.3802491426467896, + "learning_rate": 9.437312082067932e-05, + "loss": 0.4611, + "step": 30689 + }, + { + "epoch": 2.129623204496565, + "grad_norm": 1.7789170742034912, + "learning_rate": 9.436751111591162e-05, + "loss": 0.4903, + "step": 30690 + }, + { + "epoch": 2.129692595933662, + "grad_norm": 0.9342666268348694, + "learning_rate": 9.436190142892503e-05, + "loss": 0.3247, + "step": 30691 + }, + { + "epoch": 2.1297619873707583, + "grad_norm": 1.61837899684906, + "learning_rate": 9.435629175973728e-05, + "loss": 0.4826, + "step": 30692 + }, + { + "epoch": 2.129831378807855, + "grad_norm": 1.96987783908844, + "learning_rate": 9.435068210836609e-05, + "loss": 0.4173, + "step": 30693 + }, + { + "epoch": 2.129900770244952, + "grad_norm": 1.1468555927276611, + "learning_rate": 9.43450724748291e-05, + "loss": 0.4033, + "step": 30694 + }, + { + "epoch": 2.1299701616820483, + "grad_norm": 1.6551718711853027, + "learning_rate": 9.433946285914406e-05, + "loss": 0.4677, + "step": 30695 + }, + { + "epoch": 2.130039553119145, + "grad_norm": 1.0473788976669312, + "learning_rate": 9.433385326132868e-05, + "loss": 0.2795, + "step": 30696 + }, + { + "epoch": 2.1301089445562416, + "grad_norm": 1.2630846500396729, + "learning_rate": 9.432824368140068e-05, + "loss": 0.2928, + "step": 30697 + }, + { + "epoch": 2.1301783359933384, + "grad_norm": 5.713863849639893, + "learning_rate": 9.432263411937776e-05, + "loss": 0.6349, + "step": 30698 + }, + { + "epoch": 2.1302477274304352, + "grad_norm": 1.5957789421081543, + "learning_rate": 9.43170245752776e-05, + "loss": 0.4351, + "step": 30699 + }, + { + "epoch": 2.1303171188675316, + "grad_norm": 1.5973560810089111, + "learning_rate": 9.431141504911797e-05, + "loss": 0.4865, + "step": 30700 + }, + { + "epoch": 2.1303865103046284, + "grad_norm": 1.4559119939804077, + "learning_rate": 9.430580554091654e-05, + "loss": 0.5112, + "step": 30701 + }, + { + "epoch": 2.1304559017417253, + "grad_norm": 1.1375266313552856, + "learning_rate": 9.4300196050691e-05, + "loss": 0.3905, + "step": 30702 + }, + { + "epoch": 2.1305252931788217, + "grad_norm": 0.9967392086982727, + "learning_rate": 9.429458657845907e-05, + "loss": 0.3232, + "step": 30703 + }, + { + "epoch": 2.1305946846159185, + "grad_norm": 1.6347495317459106, + "learning_rate": 9.428897712423847e-05, + "loss": 0.5591, + "step": 30704 + }, + { + "epoch": 2.130664076053015, + "grad_norm": 1.1828570365905762, + "learning_rate": 9.428336768804693e-05, + "loss": 0.5169, + "step": 30705 + }, + { + "epoch": 2.1307334674901117, + "grad_norm": 1.152005672454834, + "learning_rate": 9.42777582699021e-05, + "loss": 0.3381, + "step": 30706 + }, + { + "epoch": 2.1308028589272086, + "grad_norm": 1.4811533689498901, + "learning_rate": 9.427214886982179e-05, + "loss": 0.7081, + "step": 30707 + }, + { + "epoch": 2.130872250364305, + "grad_norm": 1.12933349609375, + "learning_rate": 9.426653948782359e-05, + "loss": 0.3705, + "step": 30708 + }, + { + "epoch": 2.1309416418014018, + "grad_norm": 1.405829906463623, + "learning_rate": 9.426093012392526e-05, + "loss": 0.5204, + "step": 30709 + }, + { + "epoch": 2.131011033238498, + "grad_norm": 1.4858136177062988, + "learning_rate": 9.425532077814451e-05, + "loss": 0.5249, + "step": 30710 + }, + { + "epoch": 2.131080424675595, + "grad_norm": 1.4034894704818726, + "learning_rate": 9.424971145049904e-05, + "loss": 0.5913, + "step": 30711 + }, + { + "epoch": 2.131149816112692, + "grad_norm": 1.6481198072433472, + "learning_rate": 9.424410214100658e-05, + "loss": 0.6153, + "step": 30712 + }, + { + "epoch": 2.131219207549788, + "grad_norm": 1.754535436630249, + "learning_rate": 9.423849284968482e-05, + "loss": 0.4459, + "step": 30713 + }, + { + "epoch": 2.131288598986885, + "grad_norm": 1.9609766006469727, + "learning_rate": 9.423288357655149e-05, + "loss": 0.5827, + "step": 30714 + }, + { + "epoch": 2.131357990423982, + "grad_norm": 1.0657356977462769, + "learning_rate": 9.422727432162424e-05, + "loss": 0.2607, + "step": 30715 + }, + { + "epoch": 2.1314273818610783, + "grad_norm": 1.911313772201538, + "learning_rate": 9.422166508492083e-05, + "loss": 0.3471, + "step": 30716 + }, + { + "epoch": 2.131496773298175, + "grad_norm": 1.264618992805481, + "learning_rate": 9.421605586645897e-05, + "loss": 0.5041, + "step": 30717 + }, + { + "epoch": 2.1315661647352715, + "grad_norm": 1.0474214553833008, + "learning_rate": 9.421044666625633e-05, + "loss": 0.3867, + "step": 30718 + }, + { + "epoch": 2.1316355561723683, + "grad_norm": 1.5755892992019653, + "learning_rate": 9.420483748433062e-05, + "loss": 0.4981, + "step": 30719 + }, + { + "epoch": 2.131704947609465, + "grad_norm": 1.1977699995040894, + "learning_rate": 9.419922832069958e-05, + "loss": 0.5094, + "step": 30720 + }, + { + "epoch": 2.1317743390465616, + "grad_norm": 1.1729943752288818, + "learning_rate": 9.419361917538095e-05, + "loss": 0.3991, + "step": 30721 + }, + { + "epoch": 2.1318437304836584, + "grad_norm": 1.1169220209121704, + "learning_rate": 9.418801004839235e-05, + "loss": 0.4626, + "step": 30722 + }, + { + "epoch": 2.1319131219207548, + "grad_norm": 1.228375792503357, + "learning_rate": 9.418240093975152e-05, + "loss": 0.4804, + "step": 30723 + }, + { + "epoch": 2.1319825133578516, + "grad_norm": 1.2967699766159058, + "learning_rate": 9.417679184947618e-05, + "loss": 0.3634, + "step": 30724 + }, + { + "epoch": 2.1320519047949484, + "grad_norm": 1.4077421426773071, + "learning_rate": 9.417118277758403e-05, + "loss": 0.4882, + "step": 30725 + }, + { + "epoch": 2.132121296232045, + "grad_norm": 1.5114892721176147, + "learning_rate": 9.416557372409279e-05, + "loss": 0.4957, + "step": 30726 + }, + { + "epoch": 2.1321906876691417, + "grad_norm": 1.440903663635254, + "learning_rate": 9.415996468902017e-05, + "loss": 0.3897, + "step": 30727 + }, + { + "epoch": 2.1322600791062385, + "grad_norm": 1.3400295972824097, + "learning_rate": 9.415435567238385e-05, + "loss": 0.3241, + "step": 30728 + }, + { + "epoch": 2.132329470543335, + "grad_norm": 1.2693166732788086, + "learning_rate": 9.414874667420153e-05, + "loss": 0.3492, + "step": 30729 + }, + { + "epoch": 2.1323988619804317, + "grad_norm": 1.1268097162246704, + "learning_rate": 9.414313769449095e-05, + "loss": 0.4192, + "step": 30730 + }, + { + "epoch": 2.132468253417528, + "grad_norm": 1.2975516319274902, + "learning_rate": 9.413752873326981e-05, + "loss": 0.4647, + "step": 30731 + }, + { + "epoch": 2.132537644854625, + "grad_norm": 1.2282016277313232, + "learning_rate": 9.413191979055582e-05, + "loss": 0.3238, + "step": 30732 + }, + { + "epoch": 2.1326070362917218, + "grad_norm": 1.1278250217437744, + "learning_rate": 9.412631086636668e-05, + "loss": 0.4654, + "step": 30733 + }, + { + "epoch": 2.132676427728818, + "grad_norm": 1.3205652236938477, + "learning_rate": 9.412070196072009e-05, + "loss": 0.4754, + "step": 30734 + }, + { + "epoch": 2.132745819165915, + "grad_norm": 1.2791600227355957, + "learning_rate": 9.411509307363375e-05, + "loss": 0.4474, + "step": 30735 + }, + { + "epoch": 2.1328152106030114, + "grad_norm": 1.417728066444397, + "learning_rate": 9.41094842051254e-05, + "loss": 0.4451, + "step": 30736 + }, + { + "epoch": 2.132884602040108, + "grad_norm": 1.0185351371765137, + "learning_rate": 9.410387535521272e-05, + "loss": 0.5076, + "step": 30737 + }, + { + "epoch": 2.132953993477205, + "grad_norm": 1.4931154251098633, + "learning_rate": 9.409826652391339e-05, + "loss": 0.412, + "step": 30738 + }, + { + "epoch": 2.1330233849143014, + "grad_norm": 1.1634000539779663, + "learning_rate": 9.409265771124517e-05, + "loss": 0.415, + "step": 30739 + }, + { + "epoch": 2.1330927763513983, + "grad_norm": 1.4635170698165894, + "learning_rate": 9.408704891722573e-05, + "loss": 0.4697, + "step": 30740 + }, + { + "epoch": 2.133162167788495, + "grad_norm": 1.1639600992202759, + "learning_rate": 9.408144014187284e-05, + "loss": 0.3773, + "step": 30741 + }, + { + "epoch": 2.1332315592255915, + "grad_norm": 1.4157524108886719, + "learning_rate": 9.407583138520413e-05, + "loss": 0.5479, + "step": 30742 + }, + { + "epoch": 2.1333009506626883, + "grad_norm": 1.336846947669983, + "learning_rate": 9.40702226472373e-05, + "loss": 0.4685, + "step": 30743 + }, + { + "epoch": 2.1333703420997847, + "grad_norm": 1.514332890510559, + "learning_rate": 9.406461392799012e-05, + "loss": 0.3818, + "step": 30744 + }, + { + "epoch": 2.1334397335368815, + "grad_norm": 1.1528605222702026, + "learning_rate": 9.405900522748025e-05, + "loss": 0.2563, + "step": 30745 + }, + { + "epoch": 2.1335091249739784, + "grad_norm": 1.6022605895996094, + "learning_rate": 9.405339654572542e-05, + "loss": 0.6705, + "step": 30746 + }, + { + "epoch": 2.1335785164110748, + "grad_norm": 1.2236545085906982, + "learning_rate": 9.404778788274332e-05, + "loss": 0.3188, + "step": 30747 + }, + { + "epoch": 2.1336479078481716, + "grad_norm": 1.655619502067566, + "learning_rate": 9.40421792385517e-05, + "loss": 0.4131, + "step": 30748 + }, + { + "epoch": 2.1337172992852684, + "grad_norm": 1.2396185398101807, + "learning_rate": 9.40365706131682e-05, + "loss": 0.3399, + "step": 30749 + }, + { + "epoch": 2.133786690722365, + "grad_norm": 1.2241852283477783, + "learning_rate": 9.403096200661054e-05, + "loss": 0.4917, + "step": 30750 + }, + { + "epoch": 2.1338560821594617, + "grad_norm": 0.9098857641220093, + "learning_rate": 9.402535341889647e-05, + "loss": 0.2938, + "step": 30751 + }, + { + "epoch": 2.133925473596558, + "grad_norm": 2.001798629760742, + "learning_rate": 9.401974485004365e-05, + "loss": 0.4902, + "step": 30752 + }, + { + "epoch": 2.133994865033655, + "grad_norm": 1.0388755798339844, + "learning_rate": 9.40141363000698e-05, + "loss": 0.3127, + "step": 30753 + }, + { + "epoch": 2.1340642564707517, + "grad_norm": 1.266022801399231, + "learning_rate": 9.400852776899261e-05, + "loss": 0.3929, + "step": 30754 + }, + { + "epoch": 2.134133647907848, + "grad_norm": 0.9040820002555847, + "learning_rate": 9.400291925682986e-05, + "loss": 0.3733, + "step": 30755 + }, + { + "epoch": 2.134203039344945, + "grad_norm": 1.290415644645691, + "learning_rate": 9.399731076359915e-05, + "loss": 0.5098, + "step": 30756 + }, + { + "epoch": 2.1342724307820413, + "grad_norm": 1.176538348197937, + "learning_rate": 9.399170228931824e-05, + "loss": 0.3175, + "step": 30757 + }, + { + "epoch": 2.134341822219138, + "grad_norm": 1.1777830123901367, + "learning_rate": 9.398609383400483e-05, + "loss": 0.4077, + "step": 30758 + }, + { + "epoch": 2.134411213656235, + "grad_norm": 1.0106865167617798, + "learning_rate": 9.398048539767661e-05, + "loss": 0.4194, + "step": 30759 + }, + { + "epoch": 2.1344806050933314, + "grad_norm": 1.249631643295288, + "learning_rate": 9.397487698035131e-05, + "loss": 0.5399, + "step": 30760 + }, + { + "epoch": 2.134549996530428, + "grad_norm": 1.4858112335205078, + "learning_rate": 9.396926858204664e-05, + "loss": 0.4811, + "step": 30761 + }, + { + "epoch": 2.1346193879675246, + "grad_norm": 1.2520326375961304, + "learning_rate": 9.39636602027803e-05, + "loss": 0.2904, + "step": 30762 + }, + { + "epoch": 2.1346887794046214, + "grad_norm": 1.1533267498016357, + "learning_rate": 9.395805184256996e-05, + "loss": 0.3603, + "step": 30763 + }, + { + "epoch": 2.1347581708417183, + "grad_norm": 1.1859681606292725, + "learning_rate": 9.395244350143336e-05, + "loss": 0.3887, + "step": 30764 + }, + { + "epoch": 2.1348275622788146, + "grad_norm": 0.9785109162330627, + "learning_rate": 9.394683517938817e-05, + "loss": 0.2582, + "step": 30765 + }, + { + "epoch": 2.1348969537159115, + "grad_norm": 1.2598752975463867, + "learning_rate": 9.394122687645213e-05, + "loss": 0.3422, + "step": 30766 + }, + { + "epoch": 2.1349663451530083, + "grad_norm": 1.061959147453308, + "learning_rate": 9.393561859264294e-05, + "loss": 0.3891, + "step": 30767 + }, + { + "epoch": 2.1350357365901047, + "grad_norm": 0.9928296804428101, + "learning_rate": 9.39300103279783e-05, + "loss": 0.4603, + "step": 30768 + }, + { + "epoch": 2.1351051280272015, + "grad_norm": 1.4908967018127441, + "learning_rate": 9.392440208247592e-05, + "loss": 0.5151, + "step": 30769 + }, + { + "epoch": 2.135174519464298, + "grad_norm": 1.0956289768218994, + "learning_rate": 9.39187938561535e-05, + "loss": 0.3823, + "step": 30770 + }, + { + "epoch": 2.1352439109013948, + "grad_norm": 1.300658106803894, + "learning_rate": 9.391318564902875e-05, + "loss": 0.389, + "step": 30771 + }, + { + "epoch": 2.1353133023384916, + "grad_norm": 1.0667482614517212, + "learning_rate": 9.390757746111933e-05, + "loss": 0.3985, + "step": 30772 + }, + { + "epoch": 2.135382693775588, + "grad_norm": 0.9042049050331116, + "learning_rate": 9.390196929244299e-05, + "loss": 0.2927, + "step": 30773 + }, + { + "epoch": 2.135452085212685, + "grad_norm": 1.510793685913086, + "learning_rate": 9.389636114301742e-05, + "loss": 0.5634, + "step": 30774 + }, + { + "epoch": 2.1355214766497816, + "grad_norm": 1.547908902168274, + "learning_rate": 9.389075301286036e-05, + "loss": 0.5015, + "step": 30775 + }, + { + "epoch": 2.135590868086878, + "grad_norm": 1.4843478202819824, + "learning_rate": 9.38851449019895e-05, + "loss": 0.3807, + "step": 30776 + }, + { + "epoch": 2.135660259523975, + "grad_norm": 1.3480874300003052, + "learning_rate": 9.387953681042249e-05, + "loss": 0.4061, + "step": 30777 + }, + { + "epoch": 2.1357296509610713, + "grad_norm": 1.2596986293792725, + "learning_rate": 9.387392873817708e-05, + "loss": 0.451, + "step": 30778 + }, + { + "epoch": 2.135799042398168, + "grad_norm": 1.7805166244506836, + "learning_rate": 9.386832068527094e-05, + "loss": 0.6294, + "step": 30779 + }, + { + "epoch": 2.135868433835265, + "grad_norm": 1.3079841136932373, + "learning_rate": 9.386271265172183e-05, + "loss": 0.5696, + "step": 30780 + }, + { + "epoch": 2.1359378252723613, + "grad_norm": 1.322579026222229, + "learning_rate": 9.385710463754741e-05, + "loss": 0.4315, + "step": 30781 + }, + { + "epoch": 2.136007216709458, + "grad_norm": 1.4741101264953613, + "learning_rate": 9.38514966427654e-05, + "loss": 0.5022, + "step": 30782 + }, + { + "epoch": 2.1360766081465545, + "grad_norm": 1.258455753326416, + "learning_rate": 9.384588866739355e-05, + "loss": 0.4076, + "step": 30783 + }, + { + "epoch": 2.1361459995836514, + "grad_norm": 1.2099467515945435, + "learning_rate": 9.384028071144947e-05, + "loss": 0.4367, + "step": 30784 + }, + { + "epoch": 2.136215391020748, + "grad_norm": 1.0688072443008423, + "learning_rate": 9.38346727749509e-05, + "loss": 0.3704, + "step": 30785 + }, + { + "epoch": 2.1362847824578446, + "grad_norm": 1.0757358074188232, + "learning_rate": 9.382906485791559e-05, + "loss": 0.3875, + "step": 30786 + }, + { + "epoch": 2.1363541738949414, + "grad_norm": 1.0410600900650024, + "learning_rate": 9.382345696036116e-05, + "loss": 0.41, + "step": 30787 + }, + { + "epoch": 2.136423565332038, + "grad_norm": 1.2811774015426636, + "learning_rate": 9.381784908230539e-05, + "loss": 0.3133, + "step": 30788 + }, + { + "epoch": 2.1364929567691346, + "grad_norm": 1.7732397317886353, + "learning_rate": 9.381224122376593e-05, + "loss": 0.5621, + "step": 30789 + }, + { + "epoch": 2.1365623482062315, + "grad_norm": 1.0307414531707764, + "learning_rate": 9.380663338476056e-05, + "loss": 0.3866, + "step": 30790 + }, + { + "epoch": 2.136631739643328, + "grad_norm": 1.145896553993225, + "learning_rate": 9.380102556530689e-05, + "loss": 0.455, + "step": 30791 + }, + { + "epoch": 2.1367011310804247, + "grad_norm": 0.9836928248405457, + "learning_rate": 9.379541776542266e-05, + "loss": 0.4183, + "step": 30792 + }, + { + "epoch": 2.1367705225175215, + "grad_norm": 0.807759702205658, + "learning_rate": 9.378980998512558e-05, + "loss": 0.3219, + "step": 30793 + }, + { + "epoch": 2.136839913954618, + "grad_norm": 1.3121099472045898, + "learning_rate": 9.378420222443333e-05, + "loss": 0.4934, + "step": 30794 + }, + { + "epoch": 2.1369093053917148, + "grad_norm": 1.5657719373703003, + "learning_rate": 9.377859448336364e-05, + "loss": 0.4192, + "step": 30795 + }, + { + "epoch": 2.136978696828811, + "grad_norm": 1.236527681350708, + "learning_rate": 9.377298676193422e-05, + "loss": 0.4242, + "step": 30796 + }, + { + "epoch": 2.137048088265908, + "grad_norm": 1.0930019617080688, + "learning_rate": 9.37673790601628e-05, + "loss": 0.3926, + "step": 30797 + }, + { + "epoch": 2.137117479703005, + "grad_norm": 0.982574462890625, + "learning_rate": 9.376177137806698e-05, + "loss": 0.3106, + "step": 30798 + }, + { + "epoch": 2.137186871140101, + "grad_norm": 1.2037163972854614, + "learning_rate": 9.375616371566455e-05, + "loss": 0.3162, + "step": 30799 + }, + { + "epoch": 2.137256262577198, + "grad_norm": 1.4545496702194214, + "learning_rate": 9.375055607297317e-05, + "loss": 0.475, + "step": 30800 + }, + { + "epoch": 2.137325654014295, + "grad_norm": 1.4225964546203613, + "learning_rate": 9.374494845001056e-05, + "loss": 0.307, + "step": 30801 + }, + { + "epoch": 2.1373950454513913, + "grad_norm": 1.1822247505187988, + "learning_rate": 9.373934084679444e-05, + "loss": 0.4219, + "step": 30802 + }, + { + "epoch": 2.137464436888488, + "grad_norm": 1.458850622177124, + "learning_rate": 9.37337332633425e-05, + "loss": 0.4609, + "step": 30803 + }, + { + "epoch": 2.1375338283255845, + "grad_norm": 1.3295040130615234, + "learning_rate": 9.372812569967243e-05, + "loss": 0.3147, + "step": 30804 + }, + { + "epoch": 2.1376032197626813, + "grad_norm": 1.3080027103424072, + "learning_rate": 9.372251815580191e-05, + "loss": 0.4394, + "step": 30805 + }, + { + "epoch": 2.137672611199778, + "grad_norm": 1.411167860031128, + "learning_rate": 9.371691063174871e-05, + "loss": 0.3578, + "step": 30806 + }, + { + "epoch": 2.1377420026368745, + "grad_norm": 1.0467679500579834, + "learning_rate": 9.371130312753047e-05, + "loss": 0.3191, + "step": 30807 + }, + { + "epoch": 2.1378113940739714, + "grad_norm": 1.264601230621338, + "learning_rate": 9.370569564316491e-05, + "loss": 0.3603, + "step": 30808 + }, + { + "epoch": 2.1378807855110677, + "grad_norm": 1.0320050716400146, + "learning_rate": 9.370008817866976e-05, + "loss": 0.4197, + "step": 30809 + }, + { + "epoch": 2.1379501769481646, + "grad_norm": 1.5110145807266235, + "learning_rate": 9.369448073406273e-05, + "loss": 0.499, + "step": 30810 + }, + { + "epoch": 2.1380195683852614, + "grad_norm": 1.5071111917495728, + "learning_rate": 9.368887330936145e-05, + "loss": 0.6153, + "step": 30811 + }, + { + "epoch": 2.138088959822358, + "grad_norm": 1.6840909719467163, + "learning_rate": 9.368326590458366e-05, + "loss": 0.528, + "step": 30812 + }, + { + "epoch": 2.1381583512594546, + "grad_norm": 1.4048449993133545, + "learning_rate": 9.367765851974706e-05, + "loss": 0.4044, + "step": 30813 + }, + { + "epoch": 2.138227742696551, + "grad_norm": 0.8618572354316711, + "learning_rate": 9.367205115486938e-05, + "loss": 0.3827, + "step": 30814 + }, + { + "epoch": 2.138297134133648, + "grad_norm": 0.9004795551300049, + "learning_rate": 9.366644380996829e-05, + "loss": 0.3189, + "step": 30815 + }, + { + "epoch": 2.1383665255707447, + "grad_norm": 1.3605421781539917, + "learning_rate": 9.36608364850615e-05, + "loss": 0.643, + "step": 30816 + }, + { + "epoch": 2.138435917007841, + "grad_norm": 1.4413936138153076, + "learning_rate": 9.365522918016676e-05, + "loss": 0.385, + "step": 30817 + }, + { + "epoch": 2.138505308444938, + "grad_norm": 1.2964377403259277, + "learning_rate": 9.364962189530168e-05, + "loss": 0.4528, + "step": 30818 + }, + { + "epoch": 2.1385746998820347, + "grad_norm": 1.56414794921875, + "learning_rate": 9.3644014630484e-05, + "loss": 0.7893, + "step": 30819 + }, + { + "epoch": 2.138644091319131, + "grad_norm": 1.3823814392089844, + "learning_rate": 9.363840738573145e-05, + "loss": 0.4822, + "step": 30820 + }, + { + "epoch": 2.138713482756228, + "grad_norm": 1.3876659870147705, + "learning_rate": 9.36328001610617e-05, + "loss": 0.4676, + "step": 30821 + }, + { + "epoch": 2.1387828741933244, + "grad_norm": 1.6310595273971558, + "learning_rate": 9.362719295649248e-05, + "loss": 0.4211, + "step": 30822 + }, + { + "epoch": 2.138852265630421, + "grad_norm": 1.433838963508606, + "learning_rate": 9.362158577204143e-05, + "loss": 0.423, + "step": 30823 + }, + { + "epoch": 2.138921657067518, + "grad_norm": 1.2119500637054443, + "learning_rate": 9.361597860772634e-05, + "loss": 0.3815, + "step": 30824 + }, + { + "epoch": 2.1389910485046144, + "grad_norm": 2.1421282291412354, + "learning_rate": 9.361037146356486e-05, + "loss": 0.3703, + "step": 30825 + }, + { + "epoch": 2.1390604399417112, + "grad_norm": 1.3249015808105469, + "learning_rate": 9.360476433957467e-05, + "loss": 0.3693, + "step": 30826 + }, + { + "epoch": 2.139129831378808, + "grad_norm": 1.6090362071990967, + "learning_rate": 9.359915723577351e-05, + "loss": 0.3783, + "step": 30827 + }, + { + "epoch": 2.1391992228159045, + "grad_norm": 1.041283130645752, + "learning_rate": 9.359355015217906e-05, + "loss": 0.4096, + "step": 30828 + }, + { + "epoch": 2.1392686142530013, + "grad_norm": 1.456138014793396, + "learning_rate": 9.358794308880902e-05, + "loss": 0.4344, + "step": 30829 + }, + { + "epoch": 2.1393380056900977, + "grad_norm": 0.9291871190071106, + "learning_rate": 9.358233604568111e-05, + "loss": 0.3288, + "step": 30830 + }, + { + "epoch": 2.1394073971271945, + "grad_norm": 1.1211520433425903, + "learning_rate": 9.357672902281306e-05, + "loss": 0.3554, + "step": 30831 + }, + { + "epoch": 2.1394767885642914, + "grad_norm": 1.692278504371643, + "learning_rate": 9.357112202022249e-05, + "loss": 0.5201, + "step": 30832 + }, + { + "epoch": 2.1395461800013877, + "grad_norm": 1.0434931516647339, + "learning_rate": 9.356551503792715e-05, + "loss": 0.2309, + "step": 30833 + }, + { + "epoch": 2.1396155714384846, + "grad_norm": 1.4999347925186157, + "learning_rate": 9.355990807594471e-05, + "loss": 0.449, + "step": 30834 + }, + { + "epoch": 2.139684962875581, + "grad_norm": 1.2442909479141235, + "learning_rate": 9.355430113429291e-05, + "loss": 0.3774, + "step": 30835 + }, + { + "epoch": 2.139754354312678, + "grad_norm": 1.12522554397583, + "learning_rate": 9.354869421298943e-05, + "loss": 0.3825, + "step": 30836 + }, + { + "epoch": 2.1398237457497746, + "grad_norm": 1.723965048789978, + "learning_rate": 9.354308731205199e-05, + "loss": 0.5389, + "step": 30837 + }, + { + "epoch": 2.139893137186871, + "grad_norm": 1.0525959730148315, + "learning_rate": 9.353748043149828e-05, + "loss": 0.3572, + "step": 30838 + }, + { + "epoch": 2.139962528623968, + "grad_norm": 2.4643263816833496, + "learning_rate": 9.353187357134598e-05, + "loss": 0.3631, + "step": 30839 + }, + { + "epoch": 2.1400319200610642, + "grad_norm": 1.8835384845733643, + "learning_rate": 9.352626673161282e-05, + "loss": 0.3696, + "step": 30840 + }, + { + "epoch": 2.140101311498161, + "grad_norm": 1.348966360092163, + "learning_rate": 9.352065991231645e-05, + "loss": 0.3809, + "step": 30841 + }, + { + "epoch": 2.140170702935258, + "grad_norm": 1.0476490259170532, + "learning_rate": 9.351505311347463e-05, + "loss": 0.3127, + "step": 30842 + }, + { + "epoch": 2.1402400943723543, + "grad_norm": 1.1880450248718262, + "learning_rate": 9.350944633510501e-05, + "loss": 0.4247, + "step": 30843 + }, + { + "epoch": 2.140309485809451, + "grad_norm": 1.2008411884307861, + "learning_rate": 9.350383957722533e-05, + "loss": 0.3428, + "step": 30844 + }, + { + "epoch": 2.140378877246548, + "grad_norm": 1.4467862844467163, + "learning_rate": 9.349823283985332e-05, + "loss": 0.6437, + "step": 30845 + }, + { + "epoch": 2.1404482686836444, + "grad_norm": 1.723488211631775, + "learning_rate": 9.349262612300658e-05, + "loss": 0.3658, + "step": 30846 + }, + { + "epoch": 2.140517660120741, + "grad_norm": 1.3523966073989868, + "learning_rate": 9.348701942670288e-05, + "loss": 0.4532, + "step": 30847 + }, + { + "epoch": 2.1405870515578376, + "grad_norm": 1.061440348625183, + "learning_rate": 9.348141275095988e-05, + "loss": 0.4125, + "step": 30848 + }, + { + "epoch": 2.1406564429949344, + "grad_norm": 1.7956150770187378, + "learning_rate": 9.347580609579532e-05, + "loss": 0.4802, + "step": 30849 + }, + { + "epoch": 2.1407258344320312, + "grad_norm": 1.0812915563583374, + "learning_rate": 9.347019946122688e-05, + "loss": 0.4501, + "step": 30850 + }, + { + "epoch": 2.1407952258691276, + "grad_norm": 0.8067030906677246, + "learning_rate": 9.346459284727228e-05, + "loss": 0.2605, + "step": 30851 + }, + { + "epoch": 2.1408646173062245, + "grad_norm": 0.9555844068527222, + "learning_rate": 9.345898625394922e-05, + "loss": 0.4149, + "step": 30852 + }, + { + "epoch": 2.1409340087433213, + "grad_norm": 1.1435637474060059, + "learning_rate": 9.345337968127534e-05, + "loss": 0.3947, + "step": 30853 + }, + { + "epoch": 2.1410034001804177, + "grad_norm": 1.0513818264007568, + "learning_rate": 9.34477731292684e-05, + "loss": 0.3, + "step": 30854 + }, + { + "epoch": 2.1410727916175145, + "grad_norm": 1.5535322427749634, + "learning_rate": 9.344216659794607e-05, + "loss": 0.3356, + "step": 30855 + }, + { + "epoch": 2.141142183054611, + "grad_norm": 1.1853922605514526, + "learning_rate": 9.343656008732608e-05, + "loss": 0.3895, + "step": 30856 + }, + { + "epoch": 2.1412115744917077, + "grad_norm": 1.2637953758239746, + "learning_rate": 9.343095359742609e-05, + "loss": 0.4823, + "step": 30857 + }, + { + "epoch": 2.1412809659288046, + "grad_norm": 1.368909239768982, + "learning_rate": 9.34253471282638e-05, + "loss": 0.4214, + "step": 30858 + }, + { + "epoch": 2.141350357365901, + "grad_norm": 1.2428443431854248, + "learning_rate": 9.341974067985697e-05, + "loss": 0.4344, + "step": 30859 + }, + { + "epoch": 2.141419748802998, + "grad_norm": 0.962273895740509, + "learning_rate": 9.341413425222324e-05, + "loss": 0.2736, + "step": 30860 + }, + { + "epoch": 2.141489140240094, + "grad_norm": 0.8807049989700317, + "learning_rate": 9.340852784538032e-05, + "loss": 0.3512, + "step": 30861 + }, + { + "epoch": 2.141558531677191, + "grad_norm": 1.0951716899871826, + "learning_rate": 9.340292145934589e-05, + "loss": 0.2847, + "step": 30862 + }, + { + "epoch": 2.141627923114288, + "grad_norm": 1.5299620628356934, + "learning_rate": 9.339731509413769e-05, + "loss": 0.4211, + "step": 30863 + }, + { + "epoch": 2.1416973145513842, + "grad_norm": 0.8759989142417908, + "learning_rate": 9.339170874977339e-05, + "loss": 0.2645, + "step": 30864 + }, + { + "epoch": 2.141766705988481, + "grad_norm": 1.9421312808990479, + "learning_rate": 9.33861024262707e-05, + "loss": 0.3699, + "step": 30865 + }, + { + "epoch": 2.1418360974255775, + "grad_norm": 1.1719478368759155, + "learning_rate": 9.338049612364736e-05, + "loss": 0.4364, + "step": 30866 + }, + { + "epoch": 2.1419054888626743, + "grad_norm": 1.2589514255523682, + "learning_rate": 9.337488984192099e-05, + "loss": 0.374, + "step": 30867 + }, + { + "epoch": 2.141974880299771, + "grad_norm": 1.0673723220825195, + "learning_rate": 9.336928358110933e-05, + "loss": 0.347, + "step": 30868 + }, + { + "epoch": 2.1420442717368675, + "grad_norm": 0.9826440811157227, + "learning_rate": 9.336367734123005e-05, + "loss": 0.4196, + "step": 30869 + }, + { + "epoch": 2.1421136631739643, + "grad_norm": 1.289212942123413, + "learning_rate": 9.335807112230089e-05, + "loss": 0.4591, + "step": 30870 + }, + { + "epoch": 2.142183054611061, + "grad_norm": 1.0189820528030396, + "learning_rate": 9.335246492433952e-05, + "loss": 0.3023, + "step": 30871 + }, + { + "epoch": 2.1422524460481576, + "grad_norm": 1.7231254577636719, + "learning_rate": 9.334685874736366e-05, + "loss": 0.4253, + "step": 30872 + }, + { + "epoch": 2.1423218374852544, + "grad_norm": 1.2558767795562744, + "learning_rate": 9.334125259139101e-05, + "loss": 0.2934, + "step": 30873 + }, + { + "epoch": 2.142391228922351, + "grad_norm": 1.3346991539001465, + "learning_rate": 9.333564645643923e-05, + "loss": 0.5394, + "step": 30874 + }, + { + "epoch": 2.1424606203594476, + "grad_norm": 1.1296204328536987, + "learning_rate": 9.333004034252606e-05, + "loss": 0.4499, + "step": 30875 + }, + { + "epoch": 2.1425300117965445, + "grad_norm": 1.2247745990753174, + "learning_rate": 9.332443424966916e-05, + "loss": 0.3917, + "step": 30876 + }, + { + "epoch": 2.142599403233641, + "grad_norm": 1.5262303352355957, + "learning_rate": 9.331882817788624e-05, + "loss": 0.4446, + "step": 30877 + }, + { + "epoch": 2.1426687946707377, + "grad_norm": 1.087205171585083, + "learning_rate": 9.3313222127195e-05, + "loss": 0.4712, + "step": 30878 + }, + { + "epoch": 2.1427381861078345, + "grad_norm": 1.0895490646362305, + "learning_rate": 9.33076160976132e-05, + "loss": 0.3729, + "step": 30879 + }, + { + "epoch": 2.142807577544931, + "grad_norm": 1.5006273984909058, + "learning_rate": 9.330201008915843e-05, + "loss": 0.4326, + "step": 30880 + }, + { + "epoch": 2.1428769689820277, + "grad_norm": 1.3446335792541504, + "learning_rate": 9.329640410184843e-05, + "loss": 0.465, + "step": 30881 + }, + { + "epoch": 2.142946360419124, + "grad_norm": 1.0225577354431152, + "learning_rate": 9.32907981357009e-05, + "loss": 0.2003, + "step": 30882 + }, + { + "epoch": 2.143015751856221, + "grad_norm": 1.6170027256011963, + "learning_rate": 9.328519219073355e-05, + "loss": 0.3543, + "step": 30883 + }, + { + "epoch": 2.143085143293318, + "grad_norm": 1.402198076248169, + "learning_rate": 9.327958626696408e-05, + "loss": 0.4558, + "step": 30884 + }, + { + "epoch": 2.143154534730414, + "grad_norm": 1.1862515211105347, + "learning_rate": 9.327398036441018e-05, + "loss": 0.3378, + "step": 30885 + }, + { + "epoch": 2.143223926167511, + "grad_norm": 1.3186384439468384, + "learning_rate": 9.326837448308956e-05, + "loss": 0.509, + "step": 30886 + }, + { + "epoch": 2.1432933176046074, + "grad_norm": 1.0086590051651, + "learning_rate": 9.326276862301988e-05, + "loss": 0.4041, + "step": 30887 + }, + { + "epoch": 2.1433627090417042, + "grad_norm": 1.1685587167739868, + "learning_rate": 9.325716278421885e-05, + "loss": 0.3153, + "step": 30888 + }, + { + "epoch": 2.143432100478801, + "grad_norm": 1.0941821336746216, + "learning_rate": 9.325155696670417e-05, + "loss": 0.3229, + "step": 30889 + }, + { + "epoch": 2.1435014919158975, + "grad_norm": 1.4890259504318237, + "learning_rate": 9.324595117049356e-05, + "loss": 0.3876, + "step": 30890 + }, + { + "epoch": 2.1435708833529943, + "grad_norm": 0.9038133025169373, + "learning_rate": 9.324034539560469e-05, + "loss": 0.3268, + "step": 30891 + }, + { + "epoch": 2.1436402747900907, + "grad_norm": 1.2796975374221802, + "learning_rate": 9.323473964205525e-05, + "loss": 0.4792, + "step": 30892 + }, + { + "epoch": 2.1437096662271875, + "grad_norm": 1.344618320465088, + "learning_rate": 9.3229133909863e-05, + "loss": 0.3384, + "step": 30893 + }, + { + "epoch": 2.1437790576642843, + "grad_norm": 1.5067299604415894, + "learning_rate": 9.322352819904558e-05, + "loss": 0.386, + "step": 30894 + }, + { + "epoch": 2.1438484491013807, + "grad_norm": 1.5204896926879883, + "learning_rate": 9.321792250962065e-05, + "loss": 0.5829, + "step": 30895 + }, + { + "epoch": 2.1439178405384776, + "grad_norm": 1.135918378829956, + "learning_rate": 9.321231684160597e-05, + "loss": 0.4312, + "step": 30896 + }, + { + "epoch": 2.1439872319755744, + "grad_norm": 1.5384578704833984, + "learning_rate": 9.320671119501922e-05, + "loss": 0.3867, + "step": 30897 + }, + { + "epoch": 2.144056623412671, + "grad_norm": 1.521366000175476, + "learning_rate": 9.32011055698781e-05, + "loss": 0.4585, + "step": 30898 + }, + { + "epoch": 2.1441260148497676, + "grad_norm": 1.8060767650604248, + "learning_rate": 9.319549996620028e-05, + "loss": 0.6014, + "step": 30899 + }, + { + "epoch": 2.144195406286864, + "grad_norm": 1.4561269283294678, + "learning_rate": 9.318989438400352e-05, + "loss": 0.4805, + "step": 30900 + }, + { + "epoch": 2.144264797723961, + "grad_norm": 1.4929085969924927, + "learning_rate": 9.318428882330545e-05, + "loss": 0.4423, + "step": 30901 + }, + { + "epoch": 2.1443341891610577, + "grad_norm": 1.2168817520141602, + "learning_rate": 9.317868328412377e-05, + "loss": 0.5055, + "step": 30902 + }, + { + "epoch": 2.144403580598154, + "grad_norm": 1.558768630027771, + "learning_rate": 9.317307776647623e-05, + "loss": 0.449, + "step": 30903 + }, + { + "epoch": 2.144472972035251, + "grad_norm": 1.4007208347320557, + "learning_rate": 9.316747227038046e-05, + "loss": 0.5207, + "step": 30904 + }, + { + "epoch": 2.1445423634723477, + "grad_norm": 1.1924116611480713, + "learning_rate": 9.31618667958542e-05, + "loss": 0.3205, + "step": 30905 + }, + { + "epoch": 2.144611754909444, + "grad_norm": 0.9184662699699402, + "learning_rate": 9.315626134291515e-05, + "loss": 0.3552, + "step": 30906 + }, + { + "epoch": 2.144681146346541, + "grad_norm": 2.187772035598755, + "learning_rate": 9.315065591158099e-05, + "loss": 0.3791, + "step": 30907 + }, + { + "epoch": 2.1447505377836373, + "grad_norm": 1.4235187768936157, + "learning_rate": 9.314505050186939e-05, + "loss": 0.4226, + "step": 30908 + }, + { + "epoch": 2.144819929220734, + "grad_norm": 1.4912071228027344, + "learning_rate": 9.313944511379808e-05, + "loss": 0.4397, + "step": 30909 + }, + { + "epoch": 2.144889320657831, + "grad_norm": 2.7622997760772705, + "learning_rate": 9.313383974738475e-05, + "loss": 0.4365, + "step": 30910 + }, + { + "epoch": 2.1449587120949274, + "grad_norm": 1.310124158859253, + "learning_rate": 9.312823440264708e-05, + "loss": 0.3356, + "step": 30911 + }, + { + "epoch": 2.145028103532024, + "grad_norm": 1.0945273637771606, + "learning_rate": 9.312262907960278e-05, + "loss": 0.3506, + "step": 30912 + }, + { + "epoch": 2.1450974949691206, + "grad_norm": 1.483491063117981, + "learning_rate": 9.311702377826952e-05, + "loss": 0.2992, + "step": 30913 + }, + { + "epoch": 2.1451668864062174, + "grad_norm": 0.9886041283607483, + "learning_rate": 9.31114184986651e-05, + "loss": 0.273, + "step": 30914 + }, + { + "epoch": 2.1452362778433143, + "grad_norm": 1.4878240823745728, + "learning_rate": 9.310581324080706e-05, + "loss": 0.4114, + "step": 30915 + }, + { + "epoch": 2.1453056692804107, + "grad_norm": 1.0507240295410156, + "learning_rate": 9.310020800471317e-05, + "loss": 0.4236, + "step": 30916 + }, + { + "epoch": 2.1453750607175075, + "grad_norm": 1.5379117727279663, + "learning_rate": 9.309460279040114e-05, + "loss": 0.3087, + "step": 30917 + }, + { + "epoch": 2.145444452154604, + "grad_norm": 1.3555890321731567, + "learning_rate": 9.308899759788863e-05, + "loss": 0.3796, + "step": 30918 + }, + { + "epoch": 2.1455138435917007, + "grad_norm": 1.0489195585250854, + "learning_rate": 9.308339242719336e-05, + "loss": 0.5314, + "step": 30919 + }, + { + "epoch": 2.1455832350287976, + "grad_norm": 1.1812033653259277, + "learning_rate": 9.307778727833302e-05, + "loss": 0.315, + "step": 30920 + }, + { + "epoch": 2.145652626465894, + "grad_norm": 1.2479150295257568, + "learning_rate": 9.307218215132533e-05, + "loss": 0.5069, + "step": 30921 + }, + { + "epoch": 2.1457220179029908, + "grad_norm": 1.6986440420150757, + "learning_rate": 9.306657704618794e-05, + "loss": 0.4223, + "step": 30922 + }, + { + "epoch": 2.1457914093400876, + "grad_norm": 1.3679232597351074, + "learning_rate": 9.306097196293855e-05, + "loss": 0.474, + "step": 30923 + }, + { + "epoch": 2.145860800777184, + "grad_norm": 2.9512829780578613, + "learning_rate": 9.305536690159486e-05, + "loss": 0.5334, + "step": 30924 + }, + { + "epoch": 2.145930192214281, + "grad_norm": 1.1537688970565796, + "learning_rate": 9.304976186217458e-05, + "loss": 0.5311, + "step": 30925 + }, + { + "epoch": 2.145999583651377, + "grad_norm": 1.2545617818832397, + "learning_rate": 9.30441568446954e-05, + "loss": 0.3667, + "step": 30926 + }, + { + "epoch": 2.146068975088474, + "grad_norm": 1.3427085876464844, + "learning_rate": 9.303855184917501e-05, + "loss": 0.4085, + "step": 30927 + }, + { + "epoch": 2.146138366525571, + "grad_norm": 1.065122365951538, + "learning_rate": 9.303294687563112e-05, + "loss": 0.2755, + "step": 30928 + }, + { + "epoch": 2.1462077579626673, + "grad_norm": 1.5082365274429321, + "learning_rate": 9.30273419240814e-05, + "loss": 0.4911, + "step": 30929 + }, + { + "epoch": 2.146277149399764, + "grad_norm": 1.3686726093292236, + "learning_rate": 9.302173699454353e-05, + "loss": 0.5042, + "step": 30930 + }, + { + "epoch": 2.146346540836861, + "grad_norm": 1.0558393001556396, + "learning_rate": 9.301613208703524e-05, + "loss": 0.3686, + "step": 30931 + }, + { + "epoch": 2.1464159322739573, + "grad_norm": 1.226975679397583, + "learning_rate": 9.30105272015742e-05, + "loss": 0.4936, + "step": 30932 + }, + { + "epoch": 2.146485323711054, + "grad_norm": 1.7008219957351685, + "learning_rate": 9.300492233817813e-05, + "loss": 0.3991, + "step": 30933 + }, + { + "epoch": 2.1465547151481505, + "grad_norm": 1.4774059057235718, + "learning_rate": 9.299931749686469e-05, + "loss": 0.3049, + "step": 30934 + }, + { + "epoch": 2.1466241065852474, + "grad_norm": 1.1785918474197388, + "learning_rate": 9.299371267765164e-05, + "loss": 0.5145, + "step": 30935 + }, + { + "epoch": 2.146693498022344, + "grad_norm": 1.8703302145004272, + "learning_rate": 9.298810788055659e-05, + "loss": 0.5124, + "step": 30936 + }, + { + "epoch": 2.1467628894594406, + "grad_norm": 1.4236705303192139, + "learning_rate": 9.298250310559726e-05, + "loss": 0.6541, + "step": 30937 + }, + { + "epoch": 2.1468322808965374, + "grad_norm": 1.5130689144134521, + "learning_rate": 9.297689835279137e-05, + "loss": 0.3787, + "step": 30938 + }, + { + "epoch": 2.146901672333634, + "grad_norm": 1.2174919843673706, + "learning_rate": 9.297129362215658e-05, + "loss": 0.3525, + "step": 30939 + }, + { + "epoch": 2.1469710637707307, + "grad_norm": 1.1358412504196167, + "learning_rate": 9.29656889137106e-05, + "loss": 0.346, + "step": 30940 + }, + { + "epoch": 2.1470404552078275, + "grad_norm": 1.2370307445526123, + "learning_rate": 9.296008422747115e-05, + "loss": 0.429, + "step": 30941 + }, + { + "epoch": 2.147109846644924, + "grad_norm": 1.0575597286224365, + "learning_rate": 9.295447956345591e-05, + "loss": 0.4446, + "step": 30942 + }, + { + "epoch": 2.1471792380820207, + "grad_norm": 1.5842888355255127, + "learning_rate": 9.294887492168253e-05, + "loss": 0.372, + "step": 30943 + }, + { + "epoch": 2.1472486295191175, + "grad_norm": 1.2402678728103638, + "learning_rate": 9.294327030216873e-05, + "loss": 0.378, + "step": 30944 + }, + { + "epoch": 2.147318020956214, + "grad_norm": 1.4159495830535889, + "learning_rate": 9.293766570493223e-05, + "loss": 0.4066, + "step": 30945 + }, + { + "epoch": 2.1473874123933108, + "grad_norm": 1.0833059549331665, + "learning_rate": 9.293206112999067e-05, + "loss": 0.4012, + "step": 30946 + }, + { + "epoch": 2.147456803830407, + "grad_norm": 0.9859259128570557, + "learning_rate": 9.292645657736179e-05, + "loss": 0.3715, + "step": 30947 + }, + { + "epoch": 2.147526195267504, + "grad_norm": 1.4335421323776245, + "learning_rate": 9.292085204706326e-05, + "loss": 0.3991, + "step": 30948 + }, + { + "epoch": 2.147595586704601, + "grad_norm": 1.2385331392288208, + "learning_rate": 9.291524753911281e-05, + "loss": 0.5235, + "step": 30949 + }, + { + "epoch": 2.147664978141697, + "grad_norm": 1.6792300939559937, + "learning_rate": 9.290964305352807e-05, + "loss": 0.5128, + "step": 30950 + }, + { + "epoch": 2.147734369578794, + "grad_norm": 1.253461480140686, + "learning_rate": 9.290403859032677e-05, + "loss": 0.4219, + "step": 30951 + }, + { + "epoch": 2.1478037610158904, + "grad_norm": 1.5485906600952148, + "learning_rate": 9.28984341495266e-05, + "loss": 0.486, + "step": 30952 + }, + { + "epoch": 2.1478731524529873, + "grad_norm": 1.2240386009216309, + "learning_rate": 9.289282973114523e-05, + "loss": 0.3002, + "step": 30953 + }, + { + "epoch": 2.147942543890084, + "grad_norm": 1.934558391571045, + "learning_rate": 9.288722533520038e-05, + "loss": 0.4858, + "step": 30954 + }, + { + "epoch": 2.1480119353271805, + "grad_norm": 1.4839838743209839, + "learning_rate": 9.288162096170978e-05, + "loss": 0.4879, + "step": 30955 + }, + { + "epoch": 2.1480813267642773, + "grad_norm": 1.3130067586898804, + "learning_rate": 9.287601661069104e-05, + "loss": 0.5218, + "step": 30956 + }, + { + "epoch": 2.148150718201374, + "grad_norm": 1.5323238372802734, + "learning_rate": 9.287041228216187e-05, + "loss": 0.6538, + "step": 30957 + }, + { + "epoch": 2.1482201096384705, + "grad_norm": 1.333310604095459, + "learning_rate": 9.286480797613998e-05, + "loss": 0.3509, + "step": 30958 + }, + { + "epoch": 2.1482895010755674, + "grad_norm": 1.6811869144439697, + "learning_rate": 9.285920369264309e-05, + "loss": 0.4771, + "step": 30959 + }, + { + "epoch": 2.1483588925126638, + "grad_norm": 1.1202242374420166, + "learning_rate": 9.285359943168886e-05, + "loss": 0.3298, + "step": 30960 + }, + { + "epoch": 2.1484282839497606, + "grad_norm": 1.3727656602859497, + "learning_rate": 9.284799519329499e-05, + "loss": 0.4145, + "step": 30961 + }, + { + "epoch": 2.1484976753868574, + "grad_norm": 0.8920046091079712, + "learning_rate": 9.284239097747918e-05, + "loss": 0.2809, + "step": 30962 + }, + { + "epoch": 2.148567066823954, + "grad_norm": 1.4418636560440063, + "learning_rate": 9.283678678425909e-05, + "loss": 0.2975, + "step": 30963 + }, + { + "epoch": 2.1486364582610507, + "grad_norm": 1.4590245485305786, + "learning_rate": 9.283118261365245e-05, + "loss": 0.5231, + "step": 30964 + }, + { + "epoch": 2.148705849698147, + "grad_norm": 1.4687076807022095, + "learning_rate": 9.282557846567692e-05, + "loss": 0.4441, + "step": 30965 + }, + { + "epoch": 2.148775241135244, + "grad_norm": 1.35951566696167, + "learning_rate": 9.28199743403502e-05, + "loss": 0.4409, + "step": 30966 + }, + { + "epoch": 2.1488446325723407, + "grad_norm": 1.349181890487671, + "learning_rate": 9.281437023768999e-05, + "loss": 0.4426, + "step": 30967 + }, + { + "epoch": 2.148914024009437, + "grad_norm": 1.146058440208435, + "learning_rate": 9.280876615771398e-05, + "loss": 0.3609, + "step": 30968 + }, + { + "epoch": 2.148983415446534, + "grad_norm": 1.2603249549865723, + "learning_rate": 9.280316210043991e-05, + "loss": 0.335, + "step": 30969 + }, + { + "epoch": 2.1490528068836308, + "grad_norm": 2.845708131790161, + "learning_rate": 9.279755806588537e-05, + "loss": 0.5197, + "step": 30970 + }, + { + "epoch": 2.149122198320727, + "grad_norm": 1.2406494617462158, + "learning_rate": 9.279195405406811e-05, + "loss": 0.4416, + "step": 30971 + }, + { + "epoch": 2.149191589757824, + "grad_norm": 0.9563839435577393, + "learning_rate": 9.278635006500581e-05, + "loss": 0.3553, + "step": 30972 + }, + { + "epoch": 2.1492609811949204, + "grad_norm": 1.5215648412704468, + "learning_rate": 9.278074609871616e-05, + "loss": 0.6632, + "step": 30973 + }, + { + "epoch": 2.149330372632017, + "grad_norm": 1.1952561140060425, + "learning_rate": 9.277514215521687e-05, + "loss": 0.5463, + "step": 30974 + }, + { + "epoch": 2.149399764069114, + "grad_norm": 1.0861421823501587, + "learning_rate": 9.276953823452563e-05, + "loss": 0.4352, + "step": 30975 + }, + { + "epoch": 2.1494691555062104, + "grad_norm": 1.662699580192566, + "learning_rate": 9.276393433666013e-05, + "loss": 0.5397, + "step": 30976 + }, + { + "epoch": 2.1495385469433073, + "grad_norm": 1.0165152549743652, + "learning_rate": 9.275833046163802e-05, + "loss": 0.3488, + "step": 30977 + }, + { + "epoch": 2.149607938380404, + "grad_norm": 2.0446534156799316, + "learning_rate": 9.275272660947702e-05, + "loss": 0.4069, + "step": 30978 + }, + { + "epoch": 2.1496773298175005, + "grad_norm": 1.8032612800598145, + "learning_rate": 9.274712278019482e-05, + "loss": 0.476, + "step": 30979 + }, + { + "epoch": 2.1497467212545973, + "grad_norm": 0.9845882058143616, + "learning_rate": 9.274151897380914e-05, + "loss": 0.3532, + "step": 30980 + }, + { + "epoch": 2.1498161126916937, + "grad_norm": 1.0339232683181763, + "learning_rate": 9.273591519033763e-05, + "loss": 0.3331, + "step": 30981 + }, + { + "epoch": 2.1498855041287905, + "grad_norm": 1.2479742765426636, + "learning_rate": 9.273031142979798e-05, + "loss": 0.4793, + "step": 30982 + }, + { + "epoch": 2.1499548955658874, + "grad_norm": 1.619126319885254, + "learning_rate": 9.272470769220794e-05, + "loss": 0.3761, + "step": 30983 + }, + { + "epoch": 2.1500242870029838, + "grad_norm": 1.2356882095336914, + "learning_rate": 9.271910397758509e-05, + "loss": 0.4544, + "step": 30984 + }, + { + "epoch": 2.1500936784400806, + "grad_norm": 1.2752659320831299, + "learning_rate": 9.271350028594722e-05, + "loss": 0.456, + "step": 30985 + }, + { + "epoch": 2.150163069877177, + "grad_norm": 1.784808874130249, + "learning_rate": 9.270789661731196e-05, + "loss": 0.5128, + "step": 30986 + }, + { + "epoch": 2.150232461314274, + "grad_norm": 1.0642750263214111, + "learning_rate": 9.270229297169704e-05, + "loss": 0.4859, + "step": 30987 + }, + { + "epoch": 2.1503018527513706, + "grad_norm": 1.383298397064209, + "learning_rate": 9.269668934912011e-05, + "loss": 0.3297, + "step": 30988 + }, + { + "epoch": 2.150371244188467, + "grad_norm": 1.2327123880386353, + "learning_rate": 9.26910857495989e-05, + "loss": 0.6399, + "step": 30989 + }, + { + "epoch": 2.150440635625564, + "grad_norm": 1.3096530437469482, + "learning_rate": 9.268548217315114e-05, + "loss": 0.433, + "step": 30990 + }, + { + "epoch": 2.1505100270626603, + "grad_norm": 1.2364981174468994, + "learning_rate": 9.26798786197944e-05, + "loss": 0.4715, + "step": 30991 + }, + { + "epoch": 2.150579418499757, + "grad_norm": 1.6634113788604736, + "learning_rate": 9.267427508954645e-05, + "loss": 0.4501, + "step": 30992 + }, + { + "epoch": 2.150648809936854, + "grad_norm": 1.0107108354568481, + "learning_rate": 9.266867158242495e-05, + "loss": 0.4167, + "step": 30993 + }, + { + "epoch": 2.1507182013739503, + "grad_norm": 1.2297883033752441, + "learning_rate": 9.26630680984476e-05, + "loss": 0.3237, + "step": 30994 + }, + { + "epoch": 2.150787592811047, + "grad_norm": 1.6108919382095337, + "learning_rate": 9.265746463763213e-05, + "loss": 0.5751, + "step": 30995 + }, + { + "epoch": 2.150856984248144, + "grad_norm": 1.9965672492980957, + "learning_rate": 9.265186119999615e-05, + "loss": 0.5596, + "step": 30996 + }, + { + "epoch": 2.1509263756852404, + "grad_norm": 1.757446050643921, + "learning_rate": 9.264625778555745e-05, + "loss": 0.4392, + "step": 30997 + }, + { + "epoch": 2.150995767122337, + "grad_norm": 1.0893551111221313, + "learning_rate": 9.264065439433361e-05, + "loss": 0.3744, + "step": 30998 + }, + { + "epoch": 2.1510651585594336, + "grad_norm": 1.1434457302093506, + "learning_rate": 9.26350510263424e-05, + "loss": 0.396, + "step": 30999 + }, + { + "epoch": 2.1511345499965304, + "grad_norm": 1.0506120920181274, + "learning_rate": 9.262944768160145e-05, + "loss": 0.2948, + "step": 31000 + }, + { + "epoch": 2.1512039414336273, + "grad_norm": 0.9273906350135803, + "learning_rate": 9.262384436012848e-05, + "loss": 0.4346, + "step": 31001 + }, + { + "epoch": 2.1512733328707236, + "grad_norm": 1.28305983543396, + "learning_rate": 9.261824106194119e-05, + "loss": 0.4164, + "step": 31002 + }, + { + "epoch": 2.1513427243078205, + "grad_norm": 1.1073260307312012, + "learning_rate": 9.261263778705724e-05, + "loss": 0.3863, + "step": 31003 + }, + { + "epoch": 2.1514121157449173, + "grad_norm": 1.0790060758590698, + "learning_rate": 9.26070345354944e-05, + "loss": 0.3596, + "step": 31004 + }, + { + "epoch": 2.1514815071820137, + "grad_norm": 1.6107330322265625, + "learning_rate": 9.260143130727023e-05, + "loss": 0.537, + "step": 31005 + }, + { + "epoch": 2.1515508986191105, + "grad_norm": 0.9997148513793945, + "learning_rate": 9.259582810240249e-05, + "loss": 0.4293, + "step": 31006 + }, + { + "epoch": 2.151620290056207, + "grad_norm": 1.5544459819793701, + "learning_rate": 9.259022492090886e-05, + "loss": 0.4033, + "step": 31007 + }, + { + "epoch": 2.1516896814933038, + "grad_norm": 1.304246187210083, + "learning_rate": 9.258462176280703e-05, + "loss": 0.4793, + "step": 31008 + }, + { + "epoch": 2.1517590729304006, + "grad_norm": 1.2408250570297241, + "learning_rate": 9.25790186281147e-05, + "loss": 0.3826, + "step": 31009 + }, + { + "epoch": 2.151828464367497, + "grad_norm": 1.432374119758606, + "learning_rate": 9.257341551684953e-05, + "loss": 0.4259, + "step": 31010 + }, + { + "epoch": 2.151897855804594, + "grad_norm": 1.9970914125442505, + "learning_rate": 9.25678124290293e-05, + "loss": 0.7101, + "step": 31011 + }, + { + "epoch": 2.15196724724169, + "grad_norm": 1.1977205276489258, + "learning_rate": 9.256220936467154e-05, + "loss": 0.3826, + "step": 31012 + }, + { + "epoch": 2.152036638678787, + "grad_norm": 1.8639520406723022, + "learning_rate": 9.255660632379404e-05, + "loss": 0.4607, + "step": 31013 + }, + { + "epoch": 2.152106030115884, + "grad_norm": 1.2275058031082153, + "learning_rate": 9.255100330641447e-05, + "loss": 0.4052, + "step": 31014 + }, + { + "epoch": 2.1521754215529803, + "grad_norm": 1.4211236238479614, + "learning_rate": 9.254540031255054e-05, + "loss": 0.497, + "step": 31015 + }, + { + "epoch": 2.152244812990077, + "grad_norm": 1.0981357097625732, + "learning_rate": 9.253979734221991e-05, + "loss": 0.2993, + "step": 31016 + }, + { + "epoch": 2.1523142044271735, + "grad_norm": 1.1723699569702148, + "learning_rate": 9.253419439544025e-05, + "loss": 0.5314, + "step": 31017 + }, + { + "epoch": 2.1523835958642703, + "grad_norm": 1.084976315498352, + "learning_rate": 9.252859147222931e-05, + "loss": 0.3326, + "step": 31018 + }, + { + "epoch": 2.152452987301367, + "grad_norm": 1.058988094329834, + "learning_rate": 9.252298857260472e-05, + "loss": 0.3691, + "step": 31019 + }, + { + "epoch": 2.1525223787384635, + "grad_norm": 1.1538728475570679, + "learning_rate": 9.251738569658418e-05, + "loss": 0.391, + "step": 31020 + }, + { + "epoch": 2.1525917701755604, + "grad_norm": 1.4984080791473389, + "learning_rate": 9.251178284418539e-05, + "loss": 0.4668, + "step": 31021 + }, + { + "epoch": 2.152661161612657, + "grad_norm": 1.2669950723648071, + "learning_rate": 9.250618001542603e-05, + "loss": 0.3775, + "step": 31022 + }, + { + "epoch": 2.1527305530497536, + "grad_norm": 1.4561052322387695, + "learning_rate": 9.250057721032378e-05, + "loss": 0.5006, + "step": 31023 + }, + { + "epoch": 2.1527999444868504, + "grad_norm": 1.032434344291687, + "learning_rate": 9.249497442889634e-05, + "loss": 0.3283, + "step": 31024 + }, + { + "epoch": 2.152869335923947, + "grad_norm": 1.4692906141281128, + "learning_rate": 9.248937167116145e-05, + "loss": 0.519, + "step": 31025 + }, + { + "epoch": 2.1529387273610436, + "grad_norm": 1.2473877668380737, + "learning_rate": 9.248376893713669e-05, + "loss": 0.4483, + "step": 31026 + }, + { + "epoch": 2.1530081187981405, + "grad_norm": 1.6862328052520752, + "learning_rate": 9.24781662268398e-05, + "loss": 0.4732, + "step": 31027 + }, + { + "epoch": 2.153077510235237, + "grad_norm": 1.1614900827407837, + "learning_rate": 9.247256354028847e-05, + "loss": 0.3435, + "step": 31028 + }, + { + "epoch": 2.1531469016723337, + "grad_norm": 1.3006412982940674, + "learning_rate": 9.246696087750039e-05, + "loss": 0.4298, + "step": 31029 + }, + { + "epoch": 2.1532162931094305, + "grad_norm": 1.4461055994033813, + "learning_rate": 9.246135823849324e-05, + "loss": 0.44, + "step": 31030 + }, + { + "epoch": 2.153285684546527, + "grad_norm": 1.2530173063278198, + "learning_rate": 9.245575562328472e-05, + "loss": 0.3892, + "step": 31031 + }, + { + "epoch": 2.1533550759836237, + "grad_norm": 0.9267441034317017, + "learning_rate": 9.24501530318925e-05, + "loss": 0.4043, + "step": 31032 + }, + { + "epoch": 2.15342446742072, + "grad_norm": 0.995049238204956, + "learning_rate": 9.244455046433424e-05, + "loss": 0.447, + "step": 31033 + }, + { + "epoch": 2.153493858857817, + "grad_norm": 1.6901181936264038, + "learning_rate": 9.24389479206277e-05, + "loss": 0.4711, + "step": 31034 + }, + { + "epoch": 2.153563250294914, + "grad_norm": 1.1286019086837769, + "learning_rate": 9.243334540079049e-05, + "loss": 0.402, + "step": 31035 + }, + { + "epoch": 2.15363264173201, + "grad_norm": 1.1728333234786987, + "learning_rate": 9.242774290484035e-05, + "loss": 0.4429, + "step": 31036 + }, + { + "epoch": 2.153702033169107, + "grad_norm": 1.049885869026184, + "learning_rate": 9.242214043279494e-05, + "loss": 0.3646, + "step": 31037 + }, + { + "epoch": 2.1537714246062034, + "grad_norm": 1.0153944492340088, + "learning_rate": 9.241653798467198e-05, + "loss": 0.4224, + "step": 31038 + }, + { + "epoch": 2.1538408160433002, + "grad_norm": 1.499998927116394, + "learning_rate": 9.24109355604891e-05, + "loss": 0.5167, + "step": 31039 + }, + { + "epoch": 2.153910207480397, + "grad_norm": 1.3738319873809814, + "learning_rate": 9.2405333160264e-05, + "loss": 0.5252, + "step": 31040 + }, + { + "epoch": 2.1539795989174935, + "grad_norm": 1.5069868564605713, + "learning_rate": 9.23997307840144e-05, + "loss": 0.4229, + "step": 31041 + }, + { + "epoch": 2.1540489903545903, + "grad_norm": 1.3529678583145142, + "learning_rate": 9.239412843175797e-05, + "loss": 0.3473, + "step": 31042 + }, + { + "epoch": 2.1541183817916867, + "grad_norm": 1.1884177923202515, + "learning_rate": 9.238852610351238e-05, + "loss": 0.4004, + "step": 31043 + }, + { + "epoch": 2.1541877732287835, + "grad_norm": 1.609096884727478, + "learning_rate": 9.238292379929535e-05, + "loss": 0.414, + "step": 31044 + }, + { + "epoch": 2.1542571646658804, + "grad_norm": 1.020477056503296, + "learning_rate": 9.237732151912457e-05, + "loss": 0.4115, + "step": 31045 + }, + { + "epoch": 2.1543265561029767, + "grad_norm": 1.1805713176727295, + "learning_rate": 9.237171926301766e-05, + "loss": 0.3644, + "step": 31046 + }, + { + "epoch": 2.1543959475400736, + "grad_norm": 1.4935517311096191, + "learning_rate": 9.236611703099236e-05, + "loss": 0.4033, + "step": 31047 + }, + { + "epoch": 2.1544653389771704, + "grad_norm": 1.3016287088394165, + "learning_rate": 9.236051482306633e-05, + "loss": 0.3113, + "step": 31048 + }, + { + "epoch": 2.154534730414267, + "grad_norm": 1.215519905090332, + "learning_rate": 9.23549126392573e-05, + "loss": 0.3696, + "step": 31049 + }, + { + "epoch": 2.1546041218513636, + "grad_norm": 1.157323956489563, + "learning_rate": 9.23493104795829e-05, + "loss": 0.3301, + "step": 31050 + }, + { + "epoch": 2.15467351328846, + "grad_norm": 1.2333414554595947, + "learning_rate": 9.234370834406084e-05, + "loss": 0.4157, + "step": 31051 + }, + { + "epoch": 2.154742904725557, + "grad_norm": 1.0804702043533325, + "learning_rate": 9.233810623270882e-05, + "loss": 0.4019, + "step": 31052 + }, + { + "epoch": 2.1548122961626537, + "grad_norm": 1.790221095085144, + "learning_rate": 9.233250414554451e-05, + "loss": 0.485, + "step": 31053 + }, + { + "epoch": 2.15488168759975, + "grad_norm": 0.9446962475776672, + "learning_rate": 9.232690208258557e-05, + "loss": 0.2735, + "step": 31054 + }, + { + "epoch": 2.154951079036847, + "grad_norm": 1.2476991415023804, + "learning_rate": 9.232130004384972e-05, + "loss": 0.5086, + "step": 31055 + }, + { + "epoch": 2.1550204704739437, + "grad_norm": 1.1929727792739868, + "learning_rate": 9.231569802935463e-05, + "loss": 0.4128, + "step": 31056 + }, + { + "epoch": 2.15508986191104, + "grad_norm": 1.1462504863739014, + "learning_rate": 9.231009603911799e-05, + "loss": 0.3102, + "step": 31057 + }, + { + "epoch": 2.155159253348137, + "grad_norm": 1.1686536073684692, + "learning_rate": 9.230449407315748e-05, + "loss": 0.4807, + "step": 31058 + }, + { + "epoch": 2.1552286447852333, + "grad_norm": 2.0661191940307617, + "learning_rate": 9.229889213149085e-05, + "loss": 0.7025, + "step": 31059 + }, + { + "epoch": 2.15529803622233, + "grad_norm": 1.8211878538131714, + "learning_rate": 9.229329021413566e-05, + "loss": 0.4312, + "step": 31060 + }, + { + "epoch": 2.155367427659427, + "grad_norm": 1.1491756439208984, + "learning_rate": 9.228768832110967e-05, + "loss": 0.3484, + "step": 31061 + }, + { + "epoch": 2.1554368190965234, + "grad_norm": 1.526801586151123, + "learning_rate": 9.228208645243055e-05, + "loss": 0.4821, + "step": 31062 + }, + { + "epoch": 2.1555062105336202, + "grad_norm": 0.9785160422325134, + "learning_rate": 9.227648460811597e-05, + "loss": 0.2342, + "step": 31063 + }, + { + "epoch": 2.1555756019707166, + "grad_norm": 0.8411625027656555, + "learning_rate": 9.227088278818367e-05, + "loss": 0.3537, + "step": 31064 + }, + { + "epoch": 2.1556449934078135, + "grad_norm": 1.4611061811447144, + "learning_rate": 9.226528099265129e-05, + "loss": 0.3996, + "step": 31065 + }, + { + "epoch": 2.1557143848449103, + "grad_norm": 0.8727956414222717, + "learning_rate": 9.225967922153653e-05, + "loss": 0.3267, + "step": 31066 + }, + { + "epoch": 2.1557837762820067, + "grad_norm": 0.9626240730285645, + "learning_rate": 9.225407747485704e-05, + "loss": 0.3477, + "step": 31067 + }, + { + "epoch": 2.1558531677191035, + "grad_norm": 1.2618908882141113, + "learning_rate": 9.224847575263053e-05, + "loss": 0.3948, + "step": 31068 + }, + { + "epoch": 2.1559225591562, + "grad_norm": 1.3758854866027832, + "learning_rate": 9.22428740548747e-05, + "loss": 0.4418, + "step": 31069 + }, + { + "epoch": 2.1559919505932967, + "grad_norm": 2.1887593269348145, + "learning_rate": 9.22372723816072e-05, + "loss": 0.5381, + "step": 31070 + }, + { + "epoch": 2.1560613420303936, + "grad_norm": 1.022631287574768, + "learning_rate": 9.223167073284573e-05, + "loss": 0.4653, + "step": 31071 + }, + { + "epoch": 2.15613073346749, + "grad_norm": 1.2159065008163452, + "learning_rate": 9.222606910860798e-05, + "loss": 0.5366, + "step": 31072 + }, + { + "epoch": 2.156200124904587, + "grad_norm": 1.1141594648361206, + "learning_rate": 9.222046750891165e-05, + "loss": 0.5075, + "step": 31073 + }, + { + "epoch": 2.1562695163416836, + "grad_norm": 1.0675503015518188, + "learning_rate": 9.221486593377439e-05, + "loss": 0.4765, + "step": 31074 + }, + { + "epoch": 2.15633890777878, + "grad_norm": 1.3955051898956299, + "learning_rate": 9.220926438321387e-05, + "loss": 0.4074, + "step": 31075 + }, + { + "epoch": 2.156408299215877, + "grad_norm": 1.2596840858459473, + "learning_rate": 9.220366285724781e-05, + "loss": 0.373, + "step": 31076 + }, + { + "epoch": 2.1564776906529732, + "grad_norm": 1.75222909450531, + "learning_rate": 9.219806135589388e-05, + "loss": 0.4198, + "step": 31077 + }, + { + "epoch": 2.15654708209007, + "grad_norm": 1.7201992273330688, + "learning_rate": 9.219245987916975e-05, + "loss": 0.5573, + "step": 31078 + }, + { + "epoch": 2.156616473527167, + "grad_norm": 1.287231683731079, + "learning_rate": 9.218685842709314e-05, + "loss": 0.38, + "step": 31079 + }, + { + "epoch": 2.1566858649642633, + "grad_norm": 1.2367240190505981, + "learning_rate": 9.218125699968174e-05, + "loss": 0.5671, + "step": 31080 + }, + { + "epoch": 2.15675525640136, + "grad_norm": 1.5864101648330688, + "learning_rate": 9.217565559695317e-05, + "loss": 0.4998, + "step": 31081 + }, + { + "epoch": 2.156824647838457, + "grad_norm": 1.2693647146224976, + "learning_rate": 9.217005421892515e-05, + "loss": 0.4197, + "step": 31082 + }, + { + "epoch": 2.1568940392755533, + "grad_norm": 1.1539801359176636, + "learning_rate": 9.216445286561535e-05, + "loss": 0.3997, + "step": 31083 + }, + { + "epoch": 2.15696343071265, + "grad_norm": 1.3017451763153076, + "learning_rate": 9.215885153704149e-05, + "loss": 0.3543, + "step": 31084 + }, + { + "epoch": 2.1570328221497466, + "grad_norm": 1.641596794128418, + "learning_rate": 9.215325023322121e-05, + "loss": 0.4259, + "step": 31085 + }, + { + "epoch": 2.1571022135868434, + "grad_norm": 1.3070794343948364, + "learning_rate": 9.21476489541722e-05, + "loss": 0.4058, + "step": 31086 + }, + { + "epoch": 2.1571716050239402, + "grad_norm": 1.8668097257614136, + "learning_rate": 9.214204769991218e-05, + "loss": 0.6729, + "step": 31087 + }, + { + "epoch": 2.1572409964610366, + "grad_norm": 1.1964030265808105, + "learning_rate": 9.21364464704588e-05, + "loss": 0.4201, + "step": 31088 + }, + { + "epoch": 2.1573103878981335, + "grad_norm": 1.248749852180481, + "learning_rate": 9.213084526582973e-05, + "loss": 0.4468, + "step": 31089 + }, + { + "epoch": 2.15737977933523, + "grad_norm": 1.832057237625122, + "learning_rate": 9.212524408604265e-05, + "loss": 0.4891, + "step": 31090 + }, + { + "epoch": 2.1574491707723267, + "grad_norm": 3.28005313873291, + "learning_rate": 9.211964293111528e-05, + "loss": 0.4562, + "step": 31091 + }, + { + "epoch": 2.1575185622094235, + "grad_norm": 1.1818121671676636, + "learning_rate": 9.211404180106527e-05, + "loss": 0.3181, + "step": 31092 + }, + { + "epoch": 2.15758795364652, + "grad_norm": 1.0198242664337158, + "learning_rate": 9.210844069591033e-05, + "loss": 0.2338, + "step": 31093 + }, + { + "epoch": 2.1576573450836167, + "grad_norm": 1.3548541069030762, + "learning_rate": 9.210283961566816e-05, + "loss": 0.4681, + "step": 31094 + }, + { + "epoch": 2.157726736520713, + "grad_norm": 1.2775835990905762, + "learning_rate": 9.209723856035637e-05, + "loss": 0.4492, + "step": 31095 + }, + { + "epoch": 2.15779612795781, + "grad_norm": 1.2271089553833008, + "learning_rate": 9.209163752999267e-05, + "loss": 0.3307, + "step": 31096 + }, + { + "epoch": 2.157865519394907, + "grad_norm": 1.3698698282241821, + "learning_rate": 9.208603652459476e-05, + "loss": 0.4831, + "step": 31097 + }, + { + "epoch": 2.157934910832003, + "grad_norm": 1.2445576190948486, + "learning_rate": 9.208043554418032e-05, + "loss": 0.4392, + "step": 31098 + }, + { + "epoch": 2.1580043022691, + "grad_norm": 0.9761453866958618, + "learning_rate": 9.207483458876702e-05, + "loss": 0.3718, + "step": 31099 + }, + { + "epoch": 2.158073693706197, + "grad_norm": 1.3672149181365967, + "learning_rate": 9.206923365837257e-05, + "loss": 0.5199, + "step": 31100 + }, + { + "epoch": 2.1581430851432932, + "grad_norm": 1.2914719581604004, + "learning_rate": 9.206363275301463e-05, + "loss": 0.3905, + "step": 31101 + }, + { + "epoch": 2.15821247658039, + "grad_norm": 1.3909878730773926, + "learning_rate": 9.205803187271087e-05, + "loss": 0.6647, + "step": 31102 + }, + { + "epoch": 2.1582818680174864, + "grad_norm": 1.0929412841796875, + "learning_rate": 9.2052431017479e-05, + "loss": 0.3159, + "step": 31103 + }, + { + "epoch": 2.1583512594545833, + "grad_norm": 1.5620880126953125, + "learning_rate": 9.204683018733665e-05, + "loss": 0.3384, + "step": 31104 + }, + { + "epoch": 2.15842065089168, + "grad_norm": 1.1194546222686768, + "learning_rate": 9.204122938230153e-05, + "loss": 0.3061, + "step": 31105 + }, + { + "epoch": 2.1584900423287765, + "grad_norm": 0.9564115405082703, + "learning_rate": 9.203562860239134e-05, + "loss": 0.3134, + "step": 31106 + }, + { + "epoch": 2.1585594337658733, + "grad_norm": 1.6536015272140503, + "learning_rate": 9.20300278476238e-05, + "loss": 0.4705, + "step": 31107 + }, + { + "epoch": 2.15862882520297, + "grad_norm": 1.0102179050445557, + "learning_rate": 9.202442711801646e-05, + "loss": 0.3957, + "step": 31108 + }, + { + "epoch": 2.1586982166400666, + "grad_norm": 1.1697869300842285, + "learning_rate": 9.20188264135871e-05, + "loss": 0.3768, + "step": 31109 + }, + { + "epoch": 2.1587676080771634, + "grad_norm": 1.2360838651657104, + "learning_rate": 9.201322573435338e-05, + "loss": 0.4308, + "step": 31110 + }, + { + "epoch": 2.15883699951426, + "grad_norm": 1.537156581878662, + "learning_rate": 9.200762508033297e-05, + "loss": 0.6691, + "step": 31111 + }, + { + "epoch": 2.1589063909513566, + "grad_norm": 1.2824363708496094, + "learning_rate": 9.200202445154356e-05, + "loss": 0.2854, + "step": 31112 + }, + { + "epoch": 2.1589757823884534, + "grad_norm": 1.7030267715454102, + "learning_rate": 9.199642384800284e-05, + "loss": 0.5158, + "step": 31113 + }, + { + "epoch": 2.15904517382555, + "grad_norm": 1.6732443571090698, + "learning_rate": 9.199082326972853e-05, + "loss": 0.5901, + "step": 31114 + }, + { + "epoch": 2.1591145652626467, + "grad_norm": 1.2600973844528198, + "learning_rate": 9.198522271673821e-05, + "loss": 0.417, + "step": 31115 + }, + { + "epoch": 2.159183956699743, + "grad_norm": 0.981351912021637, + "learning_rate": 9.197962218904961e-05, + "loss": 0.2882, + "step": 31116 + }, + { + "epoch": 2.15925334813684, + "grad_norm": 1.8735699653625488, + "learning_rate": 9.197402168668042e-05, + "loss": 0.4348, + "step": 31117 + }, + { + "epoch": 2.1593227395739367, + "grad_norm": 1.533492922782898, + "learning_rate": 9.196842120964831e-05, + "loss": 0.4933, + "step": 31118 + }, + { + "epoch": 2.159392131011033, + "grad_norm": 1.2219749689102173, + "learning_rate": 9.196282075797097e-05, + "loss": 0.4173, + "step": 31119 + }, + { + "epoch": 2.15946152244813, + "grad_norm": 1.1908466815948486, + "learning_rate": 9.195722033166607e-05, + "loss": 0.3954, + "step": 31120 + }, + { + "epoch": 2.1595309138852263, + "grad_norm": 1.5802606344223022, + "learning_rate": 9.195161993075129e-05, + "loss": 0.3854, + "step": 31121 + }, + { + "epoch": 2.159600305322323, + "grad_norm": 0.9982832670211792, + "learning_rate": 9.194601955524431e-05, + "loss": 0.3458, + "step": 31122 + }, + { + "epoch": 2.15966969675942, + "grad_norm": 1.4699666500091553, + "learning_rate": 9.194041920516283e-05, + "loss": 0.4506, + "step": 31123 + }, + { + "epoch": 2.1597390881965164, + "grad_norm": 1.5983479022979736, + "learning_rate": 9.193481888052448e-05, + "loss": 0.3801, + "step": 31124 + }, + { + "epoch": 2.159808479633613, + "grad_norm": 1.2461172342300415, + "learning_rate": 9.192921858134699e-05, + "loss": 0.4372, + "step": 31125 + }, + { + "epoch": 2.15987787107071, + "grad_norm": 1.269440770149231, + "learning_rate": 9.192361830764801e-05, + "loss": 0.3169, + "step": 31126 + }, + { + "epoch": 2.1599472625078064, + "grad_norm": 1.0719581842422485, + "learning_rate": 9.191801805944524e-05, + "loss": 0.3926, + "step": 31127 + }, + { + "epoch": 2.1600166539449033, + "grad_norm": 1.142151117324829, + "learning_rate": 9.191241783675638e-05, + "loss": 0.42, + "step": 31128 + }, + { + "epoch": 2.1600860453819997, + "grad_norm": 1.4041694402694702, + "learning_rate": 9.190681763959904e-05, + "loss": 0.5189, + "step": 31129 + }, + { + "epoch": 2.1601554368190965, + "grad_norm": 1.8534914255142212, + "learning_rate": 9.190121746799093e-05, + "loss": 0.4693, + "step": 31130 + }, + { + "epoch": 2.1602248282561933, + "grad_norm": 1.7052900791168213, + "learning_rate": 9.189561732194974e-05, + "loss": 0.5414, + "step": 31131 + }, + { + "epoch": 2.1602942196932897, + "grad_norm": 1.543976068496704, + "learning_rate": 9.189001720149315e-05, + "loss": 0.406, + "step": 31132 + }, + { + "epoch": 2.1603636111303866, + "grad_norm": 1.0322366952896118, + "learning_rate": 9.188441710663884e-05, + "loss": 0.3407, + "step": 31133 + }, + { + "epoch": 2.1604330025674834, + "grad_norm": 1.2363002300262451, + "learning_rate": 9.187881703740448e-05, + "loss": 0.4122, + "step": 31134 + }, + { + "epoch": 2.1605023940045798, + "grad_norm": 1.476051688194275, + "learning_rate": 9.187321699380779e-05, + "loss": 0.3524, + "step": 31135 + }, + { + "epoch": 2.1605717854416766, + "grad_norm": 0.9391993880271912, + "learning_rate": 9.186761697586637e-05, + "loss": 0.5145, + "step": 31136 + }, + { + "epoch": 2.160641176878773, + "grad_norm": 1.4403104782104492, + "learning_rate": 9.186201698359793e-05, + "loss": 0.3384, + "step": 31137 + }, + { + "epoch": 2.16071056831587, + "grad_norm": 1.2138078212738037, + "learning_rate": 9.18564170170202e-05, + "loss": 0.4346, + "step": 31138 + }, + { + "epoch": 2.1607799597529667, + "grad_norm": 1.5139890909194946, + "learning_rate": 9.185081707615078e-05, + "loss": 0.3477, + "step": 31139 + }, + { + "epoch": 2.160849351190063, + "grad_norm": 1.4987199306488037, + "learning_rate": 9.18452171610074e-05, + "loss": 0.4654, + "step": 31140 + }, + { + "epoch": 2.16091874262716, + "grad_norm": 1.1691416501998901, + "learning_rate": 9.18396172716077e-05, + "loss": 0.5136, + "step": 31141 + }, + { + "epoch": 2.1609881340642563, + "grad_norm": 1.5281577110290527, + "learning_rate": 9.183401740796944e-05, + "loss": 0.4455, + "step": 31142 + }, + { + "epoch": 2.161057525501353, + "grad_norm": 1.325868010520935, + "learning_rate": 9.182841757011019e-05, + "loss": 0.4133, + "step": 31143 + }, + { + "epoch": 2.16112691693845, + "grad_norm": 0.9137415885925293, + "learning_rate": 9.182281775804768e-05, + "loss": 0.3568, + "step": 31144 + }, + { + "epoch": 2.1611963083755463, + "grad_norm": 1.5943583250045776, + "learning_rate": 9.181721797179959e-05, + "loss": 0.5263, + "step": 31145 + }, + { + "epoch": 2.161265699812643, + "grad_norm": 1.6457233428955078, + "learning_rate": 9.18116182113836e-05, + "loss": 0.3169, + "step": 31146 + }, + { + "epoch": 2.1613350912497395, + "grad_norm": 1.1896969079971313, + "learning_rate": 9.180601847681737e-05, + "loss": 0.2661, + "step": 31147 + }, + { + "epoch": 2.1614044826868364, + "grad_norm": 1.102378487586975, + "learning_rate": 9.180041876811859e-05, + "loss": 0.4074, + "step": 31148 + }, + { + "epoch": 2.161473874123933, + "grad_norm": 1.0046900510787964, + "learning_rate": 9.179481908530499e-05, + "loss": 0.3691, + "step": 31149 + }, + { + "epoch": 2.1615432655610296, + "grad_norm": 1.0501197576522827, + "learning_rate": 9.178921942839414e-05, + "loss": 0.3823, + "step": 31150 + }, + { + "epoch": 2.1616126569981264, + "grad_norm": 1.1926836967468262, + "learning_rate": 9.178361979740376e-05, + "loss": 0.4941, + "step": 31151 + }, + { + "epoch": 2.1616820484352233, + "grad_norm": 1.34445059299469, + "learning_rate": 9.177802019235156e-05, + "loss": 0.3694, + "step": 31152 + }, + { + "epoch": 2.1617514398723197, + "grad_norm": 1.006881594657898, + "learning_rate": 9.177242061325519e-05, + "loss": 0.3864, + "step": 31153 + }, + { + "epoch": 2.1618208313094165, + "grad_norm": 1.4115961790084839, + "learning_rate": 9.176682106013236e-05, + "loss": 0.487, + "step": 31154 + }, + { + "epoch": 2.161890222746513, + "grad_norm": 1.2775110006332397, + "learning_rate": 9.176122153300069e-05, + "loss": 0.3945, + "step": 31155 + }, + { + "epoch": 2.1619596141836097, + "grad_norm": 1.5030492544174194, + "learning_rate": 9.17556220318779e-05, + "loss": 0.4635, + "step": 31156 + }, + { + "epoch": 2.1620290056207065, + "grad_norm": 1.3829749822616577, + "learning_rate": 9.175002255678166e-05, + "loss": 0.4045, + "step": 31157 + }, + { + "epoch": 2.162098397057803, + "grad_norm": 1.6374258995056152, + "learning_rate": 9.174442310772965e-05, + "loss": 0.545, + "step": 31158 + }, + { + "epoch": 2.1621677884948998, + "grad_norm": 1.452744722366333, + "learning_rate": 9.173882368473952e-05, + "loss": 0.3955, + "step": 31159 + }, + { + "epoch": 2.1622371799319966, + "grad_norm": 3.216975212097168, + "learning_rate": 9.173322428782896e-05, + "loss": 0.4987, + "step": 31160 + }, + { + "epoch": 2.162306571369093, + "grad_norm": 0.9251525402069092, + "learning_rate": 9.172762491701565e-05, + "loss": 0.3517, + "step": 31161 + }, + { + "epoch": 2.16237596280619, + "grad_norm": 0.973861038684845, + "learning_rate": 9.172202557231727e-05, + "loss": 0.3389, + "step": 31162 + }, + { + "epoch": 2.162445354243286, + "grad_norm": 1.569814682006836, + "learning_rate": 9.171642625375155e-05, + "loss": 0.3946, + "step": 31163 + }, + { + "epoch": 2.162514745680383, + "grad_norm": 1.2801337242126465, + "learning_rate": 9.171082696133605e-05, + "loss": 0.4073, + "step": 31164 + }, + { + "epoch": 2.16258413711748, + "grad_norm": 1.2079792022705078, + "learning_rate": 9.170522769508851e-05, + "loss": 0.3878, + "step": 31165 + }, + { + "epoch": 2.1626535285545763, + "grad_norm": 1.3536262512207031, + "learning_rate": 9.16996284550266e-05, + "loss": 0.4995, + "step": 31166 + }, + { + "epoch": 2.162722919991673, + "grad_norm": 1.9189811944961548, + "learning_rate": 9.169402924116802e-05, + "loss": 0.5609, + "step": 31167 + }, + { + "epoch": 2.1627923114287695, + "grad_norm": 0.9625144600868225, + "learning_rate": 9.168843005353042e-05, + "loss": 0.3294, + "step": 31168 + }, + { + "epoch": 2.1628617028658663, + "grad_norm": 1.2812820672988892, + "learning_rate": 9.168283089213146e-05, + "loss": 0.4633, + "step": 31169 + }, + { + "epoch": 2.162931094302963, + "grad_norm": 1.368248701095581, + "learning_rate": 9.167723175698891e-05, + "loss": 0.4932, + "step": 31170 + }, + { + "epoch": 2.1630004857400595, + "grad_norm": 1.425634503364563, + "learning_rate": 9.167163264812032e-05, + "loss": 0.5078, + "step": 31171 + }, + { + "epoch": 2.1630698771771564, + "grad_norm": 1.6951504945755005, + "learning_rate": 9.166603356554342e-05, + "loss": 0.4701, + "step": 31172 + }, + { + "epoch": 2.163139268614253, + "grad_norm": 0.9104652404785156, + "learning_rate": 9.166043450927589e-05, + "loss": 0.4116, + "step": 31173 + }, + { + "epoch": 2.1632086600513496, + "grad_norm": 1.0971615314483643, + "learning_rate": 9.16548354793354e-05, + "loss": 0.4324, + "step": 31174 + }, + { + "epoch": 2.1632780514884464, + "grad_norm": 1.0032844543457031, + "learning_rate": 9.164923647573961e-05, + "loss": 0.2824, + "step": 31175 + }, + { + "epoch": 2.163347442925543, + "grad_norm": 1.3250916004180908, + "learning_rate": 9.164363749850623e-05, + "loss": 0.4178, + "step": 31176 + }, + { + "epoch": 2.1634168343626397, + "grad_norm": 1.007547378540039, + "learning_rate": 9.163803854765296e-05, + "loss": 0.3177, + "step": 31177 + }, + { + "epoch": 2.1634862257997365, + "grad_norm": 1.795814037322998, + "learning_rate": 9.163243962319738e-05, + "loss": 0.4164, + "step": 31178 + }, + { + "epoch": 2.163555617236833, + "grad_norm": 1.4312984943389893, + "learning_rate": 9.162684072515721e-05, + "loss": 0.4309, + "step": 31179 + }, + { + "epoch": 2.1636250086739297, + "grad_norm": 2.5500640869140625, + "learning_rate": 9.162124185355015e-05, + "loss": 0.4078, + "step": 31180 + }, + { + "epoch": 2.163694400111026, + "grad_norm": 1.812931776046753, + "learning_rate": 9.161564300839384e-05, + "loss": 0.5772, + "step": 31181 + }, + { + "epoch": 2.163763791548123, + "grad_norm": 1.2872180938720703, + "learning_rate": 9.161004418970598e-05, + "loss": 0.4133, + "step": 31182 + }, + { + "epoch": 2.1638331829852198, + "grad_norm": 1.0941979885101318, + "learning_rate": 9.160444539750428e-05, + "loss": 0.3673, + "step": 31183 + }, + { + "epoch": 2.163902574422316, + "grad_norm": 1.1766952276229858, + "learning_rate": 9.159884663180633e-05, + "loss": 0.4813, + "step": 31184 + }, + { + "epoch": 2.163971965859413, + "grad_norm": 1.3527781963348389, + "learning_rate": 9.159324789262985e-05, + "loss": 0.4232, + "step": 31185 + }, + { + "epoch": 2.16404135729651, + "grad_norm": 1.8063355684280396, + "learning_rate": 9.15876491799925e-05, + "loss": 0.4555, + "step": 31186 + }, + { + "epoch": 2.164110748733606, + "grad_norm": 2.3219118118286133, + "learning_rate": 9.158205049391198e-05, + "loss": 0.4809, + "step": 31187 + }, + { + "epoch": 2.164180140170703, + "grad_norm": 1.450730323791504, + "learning_rate": 9.157645183440593e-05, + "loss": 0.3836, + "step": 31188 + }, + { + "epoch": 2.1642495316077994, + "grad_norm": 1.77505362033844, + "learning_rate": 9.157085320149207e-05, + "loss": 0.4462, + "step": 31189 + }, + { + "epoch": 2.1643189230448963, + "grad_norm": 1.4401406049728394, + "learning_rate": 9.156525459518807e-05, + "loss": 0.421, + "step": 31190 + }, + { + "epoch": 2.164388314481993, + "grad_norm": 1.956093192100525, + "learning_rate": 9.155965601551153e-05, + "loss": 0.5777, + "step": 31191 + }, + { + "epoch": 2.1644577059190895, + "grad_norm": 3.6270365715026855, + "learning_rate": 9.155405746248022e-05, + "loss": 0.5862, + "step": 31192 + }, + { + "epoch": 2.1645270973561863, + "grad_norm": 1.1378525495529175, + "learning_rate": 9.154845893611175e-05, + "loss": 0.4558, + "step": 31193 + }, + { + "epoch": 2.1645964887932827, + "grad_norm": 1.0166261196136475, + "learning_rate": 9.154286043642381e-05, + "loss": 0.3397, + "step": 31194 + }, + { + "epoch": 2.1646658802303795, + "grad_norm": 1.925047516822815, + "learning_rate": 9.153726196343409e-05, + "loss": 0.3697, + "step": 31195 + }, + { + "epoch": 2.1647352716674764, + "grad_norm": 0.9238433837890625, + "learning_rate": 9.153166351716023e-05, + "loss": 0.3408, + "step": 31196 + }, + { + "epoch": 2.1648046631045728, + "grad_norm": 1.2603729963302612, + "learning_rate": 9.152606509761998e-05, + "loss": 0.4957, + "step": 31197 + }, + { + "epoch": 2.1648740545416696, + "grad_norm": 1.2773633003234863, + "learning_rate": 9.152046670483091e-05, + "loss": 0.5001, + "step": 31198 + }, + { + "epoch": 2.1649434459787664, + "grad_norm": 1.0258996486663818, + "learning_rate": 9.151486833881076e-05, + "loss": 0.4674, + "step": 31199 + }, + { + "epoch": 2.165012837415863, + "grad_norm": 1.3284904956817627, + "learning_rate": 9.150926999957716e-05, + "loss": 0.5138, + "step": 31200 + }, + { + "epoch": 2.1650822288529596, + "grad_norm": 1.0073282718658447, + "learning_rate": 9.150367168714783e-05, + "loss": 0.354, + "step": 31201 + }, + { + "epoch": 2.165151620290056, + "grad_norm": 1.3549985885620117, + "learning_rate": 9.14980734015404e-05, + "loss": 0.6547, + "step": 31202 + }, + { + "epoch": 2.165221011727153, + "grad_norm": 1.1505966186523438, + "learning_rate": 9.149247514277259e-05, + "loss": 0.4305, + "step": 31203 + }, + { + "epoch": 2.1652904031642497, + "grad_norm": 1.5305167436599731, + "learning_rate": 9.148687691086209e-05, + "loss": 0.6588, + "step": 31204 + }, + { + "epoch": 2.165359794601346, + "grad_norm": 1.5078532695770264, + "learning_rate": 9.148127870582648e-05, + "loss": 0.5983, + "step": 31205 + }, + { + "epoch": 2.165429186038443, + "grad_norm": 1.2393014430999756, + "learning_rate": 9.147568052768348e-05, + "loss": 0.4967, + "step": 31206 + }, + { + "epoch": 2.1654985774755398, + "grad_norm": 1.0252281427383423, + "learning_rate": 9.147008237645078e-05, + "loss": 0.5472, + "step": 31207 + }, + { + "epoch": 2.165567968912636, + "grad_norm": 1.2930757999420166, + "learning_rate": 9.146448425214605e-05, + "loss": 0.4515, + "step": 31208 + }, + { + "epoch": 2.165637360349733, + "grad_norm": 1.929614543914795, + "learning_rate": 9.145888615478695e-05, + "loss": 0.6214, + "step": 31209 + }, + { + "epoch": 2.1657067517868294, + "grad_norm": 1.3270570039749146, + "learning_rate": 9.145328808439115e-05, + "loss": 0.4182, + "step": 31210 + }, + { + "epoch": 2.165776143223926, + "grad_norm": 1.2071254253387451, + "learning_rate": 9.144769004097632e-05, + "loss": 0.3934, + "step": 31211 + }, + { + "epoch": 2.165845534661023, + "grad_norm": 1.7603254318237305, + "learning_rate": 9.144209202456018e-05, + "loss": 0.493, + "step": 31212 + }, + { + "epoch": 2.1659149260981194, + "grad_norm": 0.8885806798934937, + "learning_rate": 9.143649403516033e-05, + "loss": 0.3827, + "step": 31213 + }, + { + "epoch": 2.1659843175352163, + "grad_norm": 0.8752229809761047, + "learning_rate": 9.143089607279445e-05, + "loss": 0.3222, + "step": 31214 + }, + { + "epoch": 2.1660537089723126, + "grad_norm": 1.567085862159729, + "learning_rate": 9.142529813748027e-05, + "loss": 0.5388, + "step": 31215 + }, + { + "epoch": 2.1661231004094095, + "grad_norm": 1.29578697681427, + "learning_rate": 9.141970022923541e-05, + "loss": 0.3643, + "step": 31216 + }, + { + "epoch": 2.1661924918465063, + "grad_norm": 1.036797046661377, + "learning_rate": 9.141410234807757e-05, + "loss": 0.2716, + "step": 31217 + }, + { + "epoch": 2.1662618832836027, + "grad_norm": 1.2642219066619873, + "learning_rate": 9.140850449402445e-05, + "loss": 0.526, + "step": 31218 + }, + { + "epoch": 2.1663312747206995, + "grad_norm": 1.457806944847107, + "learning_rate": 9.140290666709363e-05, + "loss": 0.4509, + "step": 31219 + }, + { + "epoch": 2.166400666157796, + "grad_norm": 1.5174100399017334, + "learning_rate": 9.139730886730286e-05, + "loss": 0.442, + "step": 31220 + }, + { + "epoch": 2.1664700575948928, + "grad_norm": 1.1691874265670776, + "learning_rate": 9.139171109466978e-05, + "loss": 0.3512, + "step": 31221 + }, + { + "epoch": 2.1665394490319896, + "grad_norm": 2.2414588928222656, + "learning_rate": 9.138611334921206e-05, + "loss": 0.6502, + "step": 31222 + }, + { + "epoch": 2.166608840469086, + "grad_norm": 1.8910508155822754, + "learning_rate": 9.138051563094739e-05, + "loss": 0.4496, + "step": 31223 + }, + { + "epoch": 2.166678231906183, + "grad_norm": 1.019460678100586, + "learning_rate": 9.137491793989345e-05, + "loss": 0.3465, + "step": 31224 + }, + { + "epoch": 2.1667476233432796, + "grad_norm": 1.2924617528915405, + "learning_rate": 9.13693202760679e-05, + "loss": 0.5684, + "step": 31225 + }, + { + "epoch": 2.166817014780376, + "grad_norm": 1.3058966398239136, + "learning_rate": 9.136372263948837e-05, + "loss": 0.4332, + "step": 31226 + }, + { + "epoch": 2.166886406217473, + "grad_norm": 1.0214134454727173, + "learning_rate": 9.135812503017259e-05, + "loss": 0.2234, + "step": 31227 + }, + { + "epoch": 2.1669557976545692, + "grad_norm": 1.040507435798645, + "learning_rate": 9.135252744813819e-05, + "loss": 0.3274, + "step": 31228 + }, + { + "epoch": 2.167025189091666, + "grad_norm": 1.278128981590271, + "learning_rate": 9.134692989340286e-05, + "loss": 0.3522, + "step": 31229 + }, + { + "epoch": 2.167094580528763, + "grad_norm": 1.5440024137496948, + "learning_rate": 9.134133236598427e-05, + "loss": 0.618, + "step": 31230 + }, + { + "epoch": 2.1671639719658593, + "grad_norm": 1.207556128501892, + "learning_rate": 9.133573486590008e-05, + "loss": 0.3583, + "step": 31231 + }, + { + "epoch": 2.167233363402956, + "grad_norm": 0.9183173179626465, + "learning_rate": 9.133013739316802e-05, + "loss": 0.2975, + "step": 31232 + }, + { + "epoch": 2.167302754840053, + "grad_norm": 0.810581624507904, + "learning_rate": 9.132453994780566e-05, + "loss": 0.2827, + "step": 31233 + }, + { + "epoch": 2.1673721462771494, + "grad_norm": 1.317558765411377, + "learning_rate": 9.131894252983074e-05, + "loss": 0.4807, + "step": 31234 + }, + { + "epoch": 2.167441537714246, + "grad_norm": 0.9595603346824646, + "learning_rate": 9.131334513926089e-05, + "loss": 0.3822, + "step": 31235 + }, + { + "epoch": 2.1675109291513426, + "grad_norm": 1.8497816324234009, + "learning_rate": 9.13077477761138e-05, + "loss": 0.4954, + "step": 31236 + }, + { + "epoch": 2.1675803205884394, + "grad_norm": 1.1193997859954834, + "learning_rate": 9.130215044040715e-05, + "loss": 0.4549, + "step": 31237 + }, + { + "epoch": 2.1676497120255362, + "grad_norm": 1.4178991317749023, + "learning_rate": 9.129655313215861e-05, + "loss": 0.4398, + "step": 31238 + }, + { + "epoch": 2.1677191034626326, + "grad_norm": 1.6668974161148071, + "learning_rate": 9.129095585138588e-05, + "loss": 0.4944, + "step": 31239 + }, + { + "epoch": 2.1677884948997295, + "grad_norm": 1.2373725175857544, + "learning_rate": 9.128535859810654e-05, + "loss": 0.3736, + "step": 31240 + }, + { + "epoch": 2.167857886336826, + "grad_norm": 1.5284030437469482, + "learning_rate": 9.127976137233831e-05, + "loss": 0.5618, + "step": 31241 + }, + { + "epoch": 2.1679272777739227, + "grad_norm": 1.41618812084198, + "learning_rate": 9.127416417409887e-05, + "loss": 0.3617, + "step": 31242 + }, + { + "epoch": 2.1679966692110195, + "grad_norm": 1.2629305124282837, + "learning_rate": 9.12685670034059e-05, + "loss": 0.4253, + "step": 31243 + }, + { + "epoch": 2.168066060648116, + "grad_norm": 1.3061083555221558, + "learning_rate": 9.126296986027703e-05, + "loss": 0.3456, + "step": 31244 + }, + { + "epoch": 2.1681354520852127, + "grad_norm": 1.5200973749160767, + "learning_rate": 9.125737274472996e-05, + "loss": 0.3879, + "step": 31245 + }, + { + "epoch": 2.168204843522309, + "grad_norm": 2.123068332672119, + "learning_rate": 9.125177565678239e-05, + "loss": 0.5003, + "step": 31246 + }, + { + "epoch": 2.168274234959406, + "grad_norm": 1.139946460723877, + "learning_rate": 9.124617859645189e-05, + "loss": 0.4471, + "step": 31247 + }, + { + "epoch": 2.168343626396503, + "grad_norm": 1.1072076559066772, + "learning_rate": 9.12405815637562e-05, + "loss": 0.5231, + "step": 31248 + }, + { + "epoch": 2.168413017833599, + "grad_norm": 1.6398652791976929, + "learning_rate": 9.123498455871297e-05, + "loss": 0.7226, + "step": 31249 + }, + { + "epoch": 2.168482409270696, + "grad_norm": 0.9847334027290344, + "learning_rate": 9.12293875813399e-05, + "loss": 0.4518, + "step": 31250 + }, + { + "epoch": 2.168551800707793, + "grad_norm": 1.2229020595550537, + "learning_rate": 9.122379063165462e-05, + "loss": 0.3978, + "step": 31251 + }, + { + "epoch": 2.1686211921448892, + "grad_norm": 1.9029085636138916, + "learning_rate": 9.121819370967481e-05, + "loss": 0.4877, + "step": 31252 + }, + { + "epoch": 2.168690583581986, + "grad_norm": 1.4008959531784058, + "learning_rate": 9.121259681541819e-05, + "loss": 0.4143, + "step": 31253 + }, + { + "epoch": 2.1687599750190825, + "grad_norm": 1.1477710008621216, + "learning_rate": 9.120699994890233e-05, + "loss": 0.5812, + "step": 31254 + }, + { + "epoch": 2.1688293664561793, + "grad_norm": 1.1030954122543335, + "learning_rate": 9.120140311014498e-05, + "loss": 0.5101, + "step": 31255 + }, + { + "epoch": 2.168898757893276, + "grad_norm": 1.1483116149902344, + "learning_rate": 9.119580629916376e-05, + "loss": 0.3754, + "step": 31256 + }, + { + "epoch": 2.1689681493303725, + "grad_norm": 1.5579253435134888, + "learning_rate": 9.119020951597636e-05, + "loss": 0.4088, + "step": 31257 + }, + { + "epoch": 2.1690375407674694, + "grad_norm": 1.730054497718811, + "learning_rate": 9.118461276060047e-05, + "loss": 0.4932, + "step": 31258 + }, + { + "epoch": 2.169106932204566, + "grad_norm": 1.5831563472747803, + "learning_rate": 9.117901603305373e-05, + "loss": 0.4066, + "step": 31259 + }, + { + "epoch": 2.1691763236416626, + "grad_norm": 1.3456224203109741, + "learning_rate": 9.11734193333538e-05, + "loss": 0.494, + "step": 31260 + }, + { + "epoch": 2.1692457150787594, + "grad_norm": 1.3774423599243164, + "learning_rate": 9.116782266151834e-05, + "loss": 0.4765, + "step": 31261 + }, + { + "epoch": 2.169315106515856, + "grad_norm": 1.3784663677215576, + "learning_rate": 9.116222601756508e-05, + "loss": 0.4021, + "step": 31262 + }, + { + "epoch": 2.1693844979529526, + "grad_norm": 1.0005766153335571, + "learning_rate": 9.115662940151161e-05, + "loss": 0.3134, + "step": 31263 + }, + { + "epoch": 2.1694538893900495, + "grad_norm": 1.3504060506820679, + "learning_rate": 9.115103281337566e-05, + "loss": 0.4887, + "step": 31264 + }, + { + "epoch": 2.169523280827146, + "grad_norm": 1.1998100280761719, + "learning_rate": 9.114543625317486e-05, + "loss": 0.4257, + "step": 31265 + }, + { + "epoch": 2.1695926722642427, + "grad_norm": 1.212206482887268, + "learning_rate": 9.113983972092692e-05, + "loss": 0.3854, + "step": 31266 + }, + { + "epoch": 2.169662063701339, + "grad_norm": 1.334667444229126, + "learning_rate": 9.113424321664946e-05, + "loss": 0.6383, + "step": 31267 + }, + { + "epoch": 2.169731455138436, + "grad_norm": 0.8371138572692871, + "learning_rate": 9.112864674036014e-05, + "loss": 0.321, + "step": 31268 + }, + { + "epoch": 2.1698008465755327, + "grad_norm": 1.2361767292022705, + "learning_rate": 9.112305029207667e-05, + "loss": 0.4035, + "step": 31269 + }, + { + "epoch": 2.169870238012629, + "grad_norm": 1.2326217889785767, + "learning_rate": 9.111745387181669e-05, + "loss": 0.3678, + "step": 31270 + }, + { + "epoch": 2.169939629449726, + "grad_norm": 1.0631190538406372, + "learning_rate": 9.11118574795979e-05, + "loss": 0.3132, + "step": 31271 + }, + { + "epoch": 2.1700090208868223, + "grad_norm": 1.1995705366134644, + "learning_rate": 9.110626111543792e-05, + "loss": 0.3283, + "step": 31272 + }, + { + "epoch": 2.170078412323919, + "grad_norm": 1.5008279085159302, + "learning_rate": 9.11006647793545e-05, + "loss": 0.5775, + "step": 31273 + }, + { + "epoch": 2.170147803761016, + "grad_norm": 1.4446533918380737, + "learning_rate": 9.109506847136518e-05, + "loss": 0.5905, + "step": 31274 + }, + { + "epoch": 2.1702171951981124, + "grad_norm": 1.2739660739898682, + "learning_rate": 9.108947219148773e-05, + "loss": 0.5476, + "step": 31275 + }, + { + "epoch": 2.1702865866352092, + "grad_norm": 1.0725088119506836, + "learning_rate": 9.108387593973976e-05, + "loss": 0.2796, + "step": 31276 + }, + { + "epoch": 2.170355978072306, + "grad_norm": 0.9454468488693237, + "learning_rate": 9.107827971613897e-05, + "loss": 0.3473, + "step": 31277 + }, + { + "epoch": 2.1704253695094025, + "grad_norm": 1.4506874084472656, + "learning_rate": 9.107268352070303e-05, + "loss": 0.4681, + "step": 31278 + }, + { + "epoch": 2.1704947609464993, + "grad_norm": 1.7790629863739014, + "learning_rate": 9.106708735344957e-05, + "loss": 0.4978, + "step": 31279 + }, + { + "epoch": 2.1705641523835957, + "grad_norm": 1.3201367855072021, + "learning_rate": 9.106149121439631e-05, + "loss": 0.4884, + "step": 31280 + }, + { + "epoch": 2.1706335438206925, + "grad_norm": 1.728901982307434, + "learning_rate": 9.105589510356087e-05, + "loss": 0.4603, + "step": 31281 + }, + { + "epoch": 2.1707029352577893, + "grad_norm": 1.5546317100524902, + "learning_rate": 9.105029902096092e-05, + "loss": 0.5118, + "step": 31282 + }, + { + "epoch": 2.1707723266948857, + "grad_norm": 2.1824426651000977, + "learning_rate": 9.104470296661415e-05, + "loss": 0.6499, + "step": 31283 + }, + { + "epoch": 2.1708417181319826, + "grad_norm": 1.3075283765792847, + "learning_rate": 9.10391069405382e-05, + "loss": 0.3894, + "step": 31284 + }, + { + "epoch": 2.1709111095690794, + "grad_norm": 1.3012588024139404, + "learning_rate": 9.103351094275075e-05, + "loss": 0.3727, + "step": 31285 + }, + { + "epoch": 2.170980501006176, + "grad_norm": 1.2067010402679443, + "learning_rate": 9.102791497326945e-05, + "loss": 0.3162, + "step": 31286 + }, + { + "epoch": 2.1710498924432726, + "grad_norm": 1.2298862934112549, + "learning_rate": 9.102231903211206e-05, + "loss": 0.4836, + "step": 31287 + }, + { + "epoch": 2.171119283880369, + "grad_norm": 1.0352158546447754, + "learning_rate": 9.10167231192961e-05, + "loss": 0.3453, + "step": 31288 + }, + { + "epoch": 2.171188675317466, + "grad_norm": 1.1624690294265747, + "learning_rate": 9.101112723483931e-05, + "loss": 0.2906, + "step": 31289 + }, + { + "epoch": 2.1712580667545627, + "grad_norm": 1.4495073556900024, + "learning_rate": 9.100553137875934e-05, + "loss": 0.4718, + "step": 31290 + }, + { + "epoch": 2.171327458191659, + "grad_norm": 1.2320542335510254, + "learning_rate": 9.099993555107387e-05, + "loss": 0.2975, + "step": 31291 + }, + { + "epoch": 2.171396849628756, + "grad_norm": 1.1647385358810425, + "learning_rate": 9.099433975180056e-05, + "loss": 0.4494, + "step": 31292 + }, + { + "epoch": 2.1714662410658523, + "grad_norm": 1.0970442295074463, + "learning_rate": 9.098874398095709e-05, + "loss": 0.3354, + "step": 31293 + }, + { + "epoch": 2.171535632502949, + "grad_norm": 1.107611894607544, + "learning_rate": 9.098314823856112e-05, + "loss": 0.2846, + "step": 31294 + }, + { + "epoch": 2.171605023940046, + "grad_norm": 1.6445343494415283, + "learning_rate": 9.097755252463029e-05, + "loss": 0.6732, + "step": 31295 + }, + { + "epoch": 2.1716744153771423, + "grad_norm": 1.1094862222671509, + "learning_rate": 9.097195683918226e-05, + "loss": 0.505, + "step": 31296 + }, + { + "epoch": 2.171743806814239, + "grad_norm": 1.212406873703003, + "learning_rate": 9.096636118223475e-05, + "loss": 0.4044, + "step": 31297 + }, + { + "epoch": 2.1718131982513356, + "grad_norm": 1.2362393140792847, + "learning_rate": 9.096076555380536e-05, + "loss": 0.4954, + "step": 31298 + }, + { + "epoch": 2.1718825896884324, + "grad_norm": 1.4775640964508057, + "learning_rate": 9.09551699539118e-05, + "loss": 0.6074, + "step": 31299 + }, + { + "epoch": 2.1719519811255292, + "grad_norm": 1.7780550718307495, + "learning_rate": 9.094957438257168e-05, + "loss": 0.4675, + "step": 31300 + }, + { + "epoch": 2.1720213725626256, + "grad_norm": 1.0160725116729736, + "learning_rate": 9.094397883980279e-05, + "loss": 0.3319, + "step": 31301 + }, + { + "epoch": 2.1720907639997225, + "grad_norm": 1.2713980674743652, + "learning_rate": 9.093838332562264e-05, + "loss": 0.3851, + "step": 31302 + }, + { + "epoch": 2.1721601554368193, + "grad_norm": 1.413114070892334, + "learning_rate": 9.093278784004897e-05, + "loss": 0.6036, + "step": 31303 + }, + { + "epoch": 2.1722295468739157, + "grad_norm": 0.9771046042442322, + "learning_rate": 9.092719238309943e-05, + "loss": 0.3921, + "step": 31304 + }, + { + "epoch": 2.1722989383110125, + "grad_norm": 1.4490262269973755, + "learning_rate": 9.092159695479172e-05, + "loss": 0.3322, + "step": 31305 + }, + { + "epoch": 2.172368329748109, + "grad_norm": 1.5869137048721313, + "learning_rate": 9.091600155514343e-05, + "loss": 0.4747, + "step": 31306 + }, + { + "epoch": 2.1724377211852057, + "grad_norm": 1.4150968790054321, + "learning_rate": 9.09104061841723e-05, + "loss": 0.4827, + "step": 31307 + }, + { + "epoch": 2.1725071126223026, + "grad_norm": 1.1332378387451172, + "learning_rate": 9.090481084189601e-05, + "loss": 0.3483, + "step": 31308 + }, + { + "epoch": 2.172576504059399, + "grad_norm": 1.1814239025115967, + "learning_rate": 9.089921552833212e-05, + "loss": 0.429, + "step": 31309 + }, + { + "epoch": 2.172645895496496, + "grad_norm": 1.4007455110549927, + "learning_rate": 9.089362024349835e-05, + "loss": 0.2942, + "step": 31310 + }, + { + "epoch": 2.1727152869335926, + "grad_norm": 0.9710192084312439, + "learning_rate": 9.088802498741237e-05, + "loss": 0.2992, + "step": 31311 + }, + { + "epoch": 2.172784678370689, + "grad_norm": 1.2792383432388306, + "learning_rate": 9.088242976009184e-05, + "loss": 0.4275, + "step": 31312 + }, + { + "epoch": 2.172854069807786, + "grad_norm": 1.2403836250305176, + "learning_rate": 9.087683456155444e-05, + "loss": 0.4733, + "step": 31313 + }, + { + "epoch": 2.1729234612448822, + "grad_norm": 1.8190996646881104, + "learning_rate": 9.087123939181779e-05, + "loss": 0.3997, + "step": 31314 + }, + { + "epoch": 2.172992852681979, + "grad_norm": 0.9990788698196411, + "learning_rate": 9.08656442508996e-05, + "loss": 0.2812, + "step": 31315 + }, + { + "epoch": 2.173062244119076, + "grad_norm": 1.6409409046173096, + "learning_rate": 9.086004913881752e-05, + "loss": 0.5566, + "step": 31316 + }, + { + "epoch": 2.1731316355561723, + "grad_norm": 1.8367491960525513, + "learning_rate": 9.085445405558917e-05, + "loss": 0.5615, + "step": 31317 + }, + { + "epoch": 2.173201026993269, + "grad_norm": 1.0059701204299927, + "learning_rate": 9.084885900123224e-05, + "loss": 0.338, + "step": 31318 + }, + { + "epoch": 2.1732704184303655, + "grad_norm": 1.3615802526474, + "learning_rate": 9.084326397576442e-05, + "loss": 0.3843, + "step": 31319 + }, + { + "epoch": 2.1733398098674623, + "grad_norm": 1.2832062244415283, + "learning_rate": 9.083766897920334e-05, + "loss": 0.4188, + "step": 31320 + }, + { + "epoch": 2.173409201304559, + "grad_norm": 1.4152350425720215, + "learning_rate": 9.083207401156669e-05, + "loss": 0.4809, + "step": 31321 + }, + { + "epoch": 2.1734785927416556, + "grad_norm": 1.396790862083435, + "learning_rate": 9.082647907287217e-05, + "loss": 0.3726, + "step": 31322 + }, + { + "epoch": 2.1735479841787524, + "grad_norm": 1.0107781887054443, + "learning_rate": 9.082088416313732e-05, + "loss": 0.3777, + "step": 31323 + }, + { + "epoch": 2.173617375615849, + "grad_norm": 1.68502938747406, + "learning_rate": 9.081528928237989e-05, + "loss": 0.5513, + "step": 31324 + }, + { + "epoch": 2.1736867670529456, + "grad_norm": 1.2739460468292236, + "learning_rate": 9.080969443061752e-05, + "loss": 0.5242, + "step": 31325 + }, + { + "epoch": 2.1737561584900424, + "grad_norm": 1.312580943107605, + "learning_rate": 9.080409960786787e-05, + "loss": 0.4635, + "step": 31326 + }, + { + "epoch": 2.173825549927139, + "grad_norm": 1.4938100576400757, + "learning_rate": 9.079850481414863e-05, + "loss": 0.4709, + "step": 31327 + }, + { + "epoch": 2.1738949413642357, + "grad_norm": 1.277543306350708, + "learning_rate": 9.079291004947745e-05, + "loss": 0.4095, + "step": 31328 + }, + { + "epoch": 2.1739643328013325, + "grad_norm": 2.074784755706787, + "learning_rate": 9.0787315313872e-05, + "loss": 0.5764, + "step": 31329 + }, + { + "epoch": 2.174033724238429, + "grad_norm": 1.7357932329177856, + "learning_rate": 9.07817206073499e-05, + "loss": 0.7433, + "step": 31330 + }, + { + "epoch": 2.1741031156755257, + "grad_norm": 1.1879312992095947, + "learning_rate": 9.077612592992882e-05, + "loss": 0.3297, + "step": 31331 + }, + { + "epoch": 2.174172507112622, + "grad_norm": 1.0162442922592163, + "learning_rate": 9.077053128162648e-05, + "loss": 0.4189, + "step": 31332 + }, + { + "epoch": 2.174241898549719, + "grad_norm": 1.3312491178512573, + "learning_rate": 9.076493666246048e-05, + "loss": 0.5296, + "step": 31333 + }, + { + "epoch": 2.1743112899868158, + "grad_norm": 1.1281976699829102, + "learning_rate": 9.075934207244851e-05, + "loss": 0.3706, + "step": 31334 + }, + { + "epoch": 2.174380681423912, + "grad_norm": 0.9620879292488098, + "learning_rate": 9.075374751160825e-05, + "loss": 0.3033, + "step": 31335 + }, + { + "epoch": 2.174450072861009, + "grad_norm": 1.1723687648773193, + "learning_rate": 9.074815297995731e-05, + "loss": 0.4814, + "step": 31336 + }, + { + "epoch": 2.174519464298106, + "grad_norm": 1.086909294128418, + "learning_rate": 9.074255847751336e-05, + "loss": 0.4635, + "step": 31337 + }, + { + "epoch": 2.174588855735202, + "grad_norm": 1.1455241441726685, + "learning_rate": 9.07369640042941e-05, + "loss": 0.5041, + "step": 31338 + }, + { + "epoch": 2.174658247172299, + "grad_norm": 1.605603575706482, + "learning_rate": 9.073136956031714e-05, + "loss": 0.4821, + "step": 31339 + }, + { + "epoch": 2.1747276386093954, + "grad_norm": 1.2098777294158936, + "learning_rate": 9.072577514560018e-05, + "loss": 0.3867, + "step": 31340 + }, + { + "epoch": 2.1747970300464923, + "grad_norm": 1.830230951309204, + "learning_rate": 9.07201807601609e-05, + "loss": 0.633, + "step": 31341 + }, + { + "epoch": 2.174866421483589, + "grad_norm": 1.2403424978256226, + "learning_rate": 9.071458640401695e-05, + "loss": 0.4954, + "step": 31342 + }, + { + "epoch": 2.1749358129206855, + "grad_norm": 1.2950725555419922, + "learning_rate": 9.070899207718594e-05, + "loss": 0.3836, + "step": 31343 + }, + { + "epoch": 2.1750052043577823, + "grad_norm": 1.3771485090255737, + "learning_rate": 9.070339777968556e-05, + "loss": 0.487, + "step": 31344 + }, + { + "epoch": 2.1750745957948787, + "grad_norm": 1.1312824487686157, + "learning_rate": 9.069780351153348e-05, + "loss": 0.3042, + "step": 31345 + }, + { + "epoch": 2.1751439872319756, + "grad_norm": 1.744369626045227, + "learning_rate": 9.069220927274734e-05, + "loss": 0.5065, + "step": 31346 + }, + { + "epoch": 2.1752133786690724, + "grad_norm": 1.1002614498138428, + "learning_rate": 9.068661506334485e-05, + "loss": 0.3672, + "step": 31347 + }, + { + "epoch": 2.1752827701061688, + "grad_norm": 1.3524985313415527, + "learning_rate": 9.068102088334362e-05, + "loss": 0.5407, + "step": 31348 + }, + { + "epoch": 2.1753521615432656, + "grad_norm": 1.1778053045272827, + "learning_rate": 9.067542673276133e-05, + "loss": 0.4024, + "step": 31349 + }, + { + "epoch": 2.175421552980362, + "grad_norm": 1.2823680639266968, + "learning_rate": 9.066983261161562e-05, + "loss": 0.4451, + "step": 31350 + }, + { + "epoch": 2.175490944417459, + "grad_norm": 1.0069942474365234, + "learning_rate": 9.066423851992419e-05, + "loss": 0.4519, + "step": 31351 + }, + { + "epoch": 2.1755603358545557, + "grad_norm": 1.1654177904129028, + "learning_rate": 9.065864445770466e-05, + "loss": 0.4577, + "step": 31352 + }, + { + "epoch": 2.175629727291652, + "grad_norm": 1.4061814546585083, + "learning_rate": 9.06530504249747e-05, + "loss": 0.4542, + "step": 31353 + }, + { + "epoch": 2.175699118728749, + "grad_norm": 1.7900124788284302, + "learning_rate": 9.064745642175197e-05, + "loss": 0.431, + "step": 31354 + }, + { + "epoch": 2.1757685101658457, + "grad_norm": 1.097760558128357, + "learning_rate": 9.064186244805415e-05, + "loss": 0.3726, + "step": 31355 + }, + { + "epoch": 2.175837901602942, + "grad_norm": 1.937077283859253, + "learning_rate": 9.063626850389893e-05, + "loss": 0.5122, + "step": 31356 + }, + { + "epoch": 2.175907293040039, + "grad_norm": 0.9965130090713501, + "learning_rate": 9.063067458930386e-05, + "loss": 0.3144, + "step": 31357 + }, + { + "epoch": 2.1759766844771353, + "grad_norm": 1.3990728855133057, + "learning_rate": 9.062508070428667e-05, + "loss": 0.4803, + "step": 31358 + }, + { + "epoch": 2.176046075914232, + "grad_norm": 1.5470335483551025, + "learning_rate": 9.061948684886501e-05, + "loss": 0.5006, + "step": 31359 + }, + { + "epoch": 2.176115467351329, + "grad_norm": 1.4704327583312988, + "learning_rate": 9.061389302305655e-05, + "loss": 0.4243, + "step": 31360 + }, + { + "epoch": 2.1761848587884254, + "grad_norm": 1.0947755575180054, + "learning_rate": 9.060829922687895e-05, + "loss": 0.3923, + "step": 31361 + }, + { + "epoch": 2.176254250225522, + "grad_norm": 1.3123347759246826, + "learning_rate": 9.060270546034983e-05, + "loss": 0.4311, + "step": 31362 + }, + { + "epoch": 2.176323641662619, + "grad_norm": 1.5296754837036133, + "learning_rate": 9.059711172348695e-05, + "loss": 0.3399, + "step": 31363 + }, + { + "epoch": 2.1763930330997154, + "grad_norm": 1.282678484916687, + "learning_rate": 9.059151801630784e-05, + "loss": 0.4239, + "step": 31364 + }, + { + "epoch": 2.1764624245368123, + "grad_norm": 1.2345867156982422, + "learning_rate": 9.058592433883022e-05, + "loss": 0.411, + "step": 31365 + }, + { + "epoch": 2.1765318159739087, + "grad_norm": 1.2364997863769531, + "learning_rate": 9.058033069107175e-05, + "loss": 0.4597, + "step": 31366 + }, + { + "epoch": 2.1766012074110055, + "grad_norm": 1.5551577806472778, + "learning_rate": 9.057473707305009e-05, + "loss": 0.5625, + "step": 31367 + }, + { + "epoch": 2.1766705988481023, + "grad_norm": 1.124509572982788, + "learning_rate": 9.056914348478289e-05, + "loss": 0.3446, + "step": 31368 + }, + { + "epoch": 2.1767399902851987, + "grad_norm": 1.0407283306121826, + "learning_rate": 9.05635499262878e-05, + "loss": 0.2247, + "step": 31369 + }, + { + "epoch": 2.1768093817222955, + "grad_norm": 1.8702765703201294, + "learning_rate": 9.055795639758252e-05, + "loss": 0.5832, + "step": 31370 + }, + { + "epoch": 2.176878773159392, + "grad_norm": 1.4868155717849731, + "learning_rate": 9.055236289868464e-05, + "loss": 0.4261, + "step": 31371 + }, + { + "epoch": 2.1769481645964888, + "grad_norm": 1.3900465965270996, + "learning_rate": 9.054676942961186e-05, + "loss": 0.6336, + "step": 31372 + }, + { + "epoch": 2.1770175560335856, + "grad_norm": 1.2442641258239746, + "learning_rate": 9.054117599038183e-05, + "loss": 0.4002, + "step": 31373 + }, + { + "epoch": 2.177086947470682, + "grad_norm": 1.2343319654464722, + "learning_rate": 9.05355825810122e-05, + "loss": 0.5082, + "step": 31374 + }, + { + "epoch": 2.177156338907779, + "grad_norm": 1.8898552656173706, + "learning_rate": 9.052998920152065e-05, + "loss": 0.4726, + "step": 31375 + }, + { + "epoch": 2.177225730344875, + "grad_norm": 1.6270101070404053, + "learning_rate": 9.052439585192482e-05, + "loss": 0.4225, + "step": 31376 + }, + { + "epoch": 2.177295121781972, + "grad_norm": 1.5607153177261353, + "learning_rate": 9.051880253224243e-05, + "loss": 0.4067, + "step": 31377 + }, + { + "epoch": 2.177364513219069, + "grad_norm": 1.6139618158340454, + "learning_rate": 9.051320924249103e-05, + "loss": 0.4124, + "step": 31378 + }, + { + "epoch": 2.1774339046561653, + "grad_norm": 1.2791978120803833, + "learning_rate": 9.050761598268831e-05, + "loss": 0.459, + "step": 31379 + }, + { + "epoch": 2.177503296093262, + "grad_norm": 1.0642112493515015, + "learning_rate": 9.050202275285195e-05, + "loss": 0.4843, + "step": 31380 + }, + { + "epoch": 2.177572687530359, + "grad_norm": 1.2166502475738525, + "learning_rate": 9.049642955299961e-05, + "loss": 0.4333, + "step": 31381 + }, + { + "epoch": 2.1776420789674553, + "grad_norm": 3.581246852874756, + "learning_rate": 9.049083638314897e-05, + "loss": 0.3984, + "step": 31382 + }, + { + "epoch": 2.177711470404552, + "grad_norm": 0.973895251750946, + "learning_rate": 9.048524324331763e-05, + "loss": 0.3746, + "step": 31383 + }, + { + "epoch": 2.1777808618416485, + "grad_norm": 1.5495340824127197, + "learning_rate": 9.047965013352328e-05, + "loss": 0.4538, + "step": 31384 + }, + { + "epoch": 2.1778502532787454, + "grad_norm": 2.034071922302246, + "learning_rate": 9.047405705378357e-05, + "loss": 0.3776, + "step": 31385 + }, + { + "epoch": 2.177919644715842, + "grad_norm": 1.0020966529846191, + "learning_rate": 9.046846400411618e-05, + "loss": 0.291, + "step": 31386 + }, + { + "epoch": 2.1779890361529386, + "grad_norm": 1.0411332845687866, + "learning_rate": 9.04628709845387e-05, + "loss": 0.3257, + "step": 31387 + }, + { + "epoch": 2.1780584275900354, + "grad_norm": 1.3968158960342407, + "learning_rate": 9.045727799506884e-05, + "loss": 0.4522, + "step": 31388 + }, + { + "epoch": 2.1781278190271323, + "grad_norm": 1.3055373430252075, + "learning_rate": 9.045168503572424e-05, + "loss": 0.4239, + "step": 31389 + }, + { + "epoch": 2.1781972104642287, + "grad_norm": 1.1793959140777588, + "learning_rate": 9.044609210652256e-05, + "loss": 0.3742, + "step": 31390 + }, + { + "epoch": 2.1782666019013255, + "grad_norm": 1.025048851966858, + "learning_rate": 9.044049920748153e-05, + "loss": 0.2879, + "step": 31391 + }, + { + "epoch": 2.178335993338422, + "grad_norm": 1.6135687828063965, + "learning_rate": 9.043490633861867e-05, + "loss": 0.402, + "step": 31392 + }, + { + "epoch": 2.1784053847755187, + "grad_norm": 1.2728571891784668, + "learning_rate": 9.042931349995169e-05, + "loss": 0.3184, + "step": 31393 + }, + { + "epoch": 2.1784747762126155, + "grad_norm": 1.076973795890808, + "learning_rate": 9.042372069149827e-05, + "loss": 0.2364, + "step": 31394 + }, + { + "epoch": 2.178544167649712, + "grad_norm": 1.1504632234573364, + "learning_rate": 9.041812791327605e-05, + "loss": 0.5491, + "step": 31395 + }, + { + "epoch": 2.1786135590868088, + "grad_norm": 1.5603944063186646, + "learning_rate": 9.041253516530269e-05, + "loss": 0.4515, + "step": 31396 + }, + { + "epoch": 2.178682950523905, + "grad_norm": 1.53835129737854, + "learning_rate": 9.040694244759585e-05, + "loss": 0.4647, + "step": 31397 + }, + { + "epoch": 2.178752341961002, + "grad_norm": 1.2519075870513916, + "learning_rate": 9.040134976017322e-05, + "loss": 0.3017, + "step": 31398 + }, + { + "epoch": 2.178821733398099, + "grad_norm": 1.2062757015228271, + "learning_rate": 9.039575710305236e-05, + "loss": 0.4976, + "step": 31399 + }, + { + "epoch": 2.178891124835195, + "grad_norm": 1.4613630771636963, + "learning_rate": 9.0390164476251e-05, + "loss": 0.4296, + "step": 31400 + }, + { + "epoch": 2.178960516272292, + "grad_norm": 1.0647560358047485, + "learning_rate": 9.038457187978678e-05, + "loss": 0.4261, + "step": 31401 + }, + { + "epoch": 2.179029907709389, + "grad_norm": 1.3478254079818726, + "learning_rate": 9.037897931367733e-05, + "loss": 0.4994, + "step": 31402 + }, + { + "epoch": 2.1790992991464853, + "grad_norm": 1.5743566751480103, + "learning_rate": 9.037338677794033e-05, + "loss": 0.5565, + "step": 31403 + }, + { + "epoch": 2.179168690583582, + "grad_norm": 1.1652628183364868, + "learning_rate": 9.036779427259344e-05, + "loss": 0.3074, + "step": 31404 + }, + { + "epoch": 2.1792380820206785, + "grad_norm": 0.928229808807373, + "learning_rate": 9.036220179765434e-05, + "loss": 0.2658, + "step": 31405 + }, + { + "epoch": 2.1793074734577753, + "grad_norm": 1.1431597471237183, + "learning_rate": 9.03566093531406e-05, + "loss": 0.3237, + "step": 31406 + }, + { + "epoch": 2.179376864894872, + "grad_norm": 1.4474807977676392, + "learning_rate": 9.035101693906995e-05, + "loss": 0.5845, + "step": 31407 + }, + { + "epoch": 2.1794462563319685, + "grad_norm": 1.9421601295471191, + "learning_rate": 9.034542455546e-05, + "loss": 0.5362, + "step": 31408 + }, + { + "epoch": 2.1795156477690654, + "grad_norm": 1.2674678564071655, + "learning_rate": 9.033983220232843e-05, + "loss": 0.4875, + "step": 31409 + }, + { + "epoch": 2.1795850392061618, + "grad_norm": 1.1656590700149536, + "learning_rate": 9.03342398796929e-05, + "loss": 0.3927, + "step": 31410 + }, + { + "epoch": 2.1796544306432586, + "grad_norm": 1.047182559967041, + "learning_rate": 9.032864758757109e-05, + "loss": 0.4276, + "step": 31411 + }, + { + "epoch": 2.1797238220803554, + "grad_norm": 1.0789713859558105, + "learning_rate": 9.032305532598057e-05, + "loss": 0.3228, + "step": 31412 + }, + { + "epoch": 2.179793213517452, + "grad_norm": 1.313731074333191, + "learning_rate": 9.031746309493903e-05, + "loss": 0.4522, + "step": 31413 + }, + { + "epoch": 2.1798626049545486, + "grad_norm": 1.129883050918579, + "learning_rate": 9.031187089446414e-05, + "loss": 0.3815, + "step": 31414 + }, + { + "epoch": 2.1799319963916455, + "grad_norm": 1.3519130945205688, + "learning_rate": 9.030627872457357e-05, + "loss": 0.4746, + "step": 31415 + }, + { + "epoch": 2.180001387828742, + "grad_norm": 1.1861542463302612, + "learning_rate": 9.030068658528494e-05, + "loss": 0.5698, + "step": 31416 + }, + { + "epoch": 2.1800707792658387, + "grad_norm": 1.4403325319290161, + "learning_rate": 9.029509447661593e-05, + "loss": 0.4279, + "step": 31417 + }, + { + "epoch": 2.180140170702935, + "grad_norm": 1.67463219165802, + "learning_rate": 9.028950239858419e-05, + "loss": 0.4722, + "step": 31418 + }, + { + "epoch": 2.180209562140032, + "grad_norm": 0.9409431219100952, + "learning_rate": 9.028391035120734e-05, + "loss": 0.3705, + "step": 31419 + }, + { + "epoch": 2.1802789535771288, + "grad_norm": 1.1504806280136108, + "learning_rate": 9.027831833450306e-05, + "loss": 0.4385, + "step": 31420 + }, + { + "epoch": 2.180348345014225, + "grad_norm": 1.335270881652832, + "learning_rate": 9.027272634848904e-05, + "loss": 0.351, + "step": 31421 + }, + { + "epoch": 2.180417736451322, + "grad_norm": 1.6536905765533447, + "learning_rate": 9.026713439318286e-05, + "loss": 0.5285, + "step": 31422 + }, + { + "epoch": 2.1804871278884184, + "grad_norm": 1.7852667570114136, + "learning_rate": 9.026154246860219e-05, + "loss": 0.3495, + "step": 31423 + }, + { + "epoch": 2.180556519325515, + "grad_norm": 1.446803092956543, + "learning_rate": 9.025595057476472e-05, + "loss": 0.4371, + "step": 31424 + }, + { + "epoch": 2.180625910762612, + "grad_norm": 1.6671679019927979, + "learning_rate": 9.02503587116881e-05, + "loss": 0.5929, + "step": 31425 + }, + { + "epoch": 2.1806953021997084, + "grad_norm": 1.4437915086746216, + "learning_rate": 9.024476687938996e-05, + "loss": 0.4568, + "step": 31426 + }, + { + "epoch": 2.1807646936368053, + "grad_norm": 1.3271323442459106, + "learning_rate": 9.023917507788794e-05, + "loss": 0.4274, + "step": 31427 + }, + { + "epoch": 2.180834085073902, + "grad_norm": 1.1382722854614258, + "learning_rate": 9.02335833071997e-05, + "loss": 0.374, + "step": 31428 + }, + { + "epoch": 2.1809034765109985, + "grad_norm": 1.0508222579956055, + "learning_rate": 9.022799156734293e-05, + "loss": 0.3432, + "step": 31429 + }, + { + "epoch": 2.1809728679480953, + "grad_norm": 1.258834719657898, + "learning_rate": 9.022239985833525e-05, + "loss": 0.404, + "step": 31430 + }, + { + "epoch": 2.1810422593851917, + "grad_norm": 1.153047800064087, + "learning_rate": 9.021680818019431e-05, + "loss": 0.4901, + "step": 31431 + }, + { + "epoch": 2.1811116508222885, + "grad_norm": 1.299680233001709, + "learning_rate": 9.021121653293783e-05, + "loss": 0.4684, + "step": 31432 + }, + { + "epoch": 2.1811810422593854, + "grad_norm": 1.7369372844696045, + "learning_rate": 9.020562491658335e-05, + "loss": 0.5674, + "step": 31433 + }, + { + "epoch": 2.1812504336964817, + "grad_norm": 1.196308970451355, + "learning_rate": 9.020003333114857e-05, + "loss": 0.4993, + "step": 31434 + }, + { + "epoch": 2.1813198251335786, + "grad_norm": 1.1056207418441772, + "learning_rate": 9.019444177665115e-05, + "loss": 0.4125, + "step": 31435 + }, + { + "epoch": 2.181389216570675, + "grad_norm": 1.030193567276001, + "learning_rate": 9.018885025310878e-05, + "loss": 0.2783, + "step": 31436 + }, + { + "epoch": 2.181458608007772, + "grad_norm": 1.2090847492218018, + "learning_rate": 9.018325876053902e-05, + "loss": 0.4217, + "step": 31437 + }, + { + "epoch": 2.1815279994448686, + "grad_norm": 1.47293221950531, + "learning_rate": 9.017766729895959e-05, + "loss": 0.3283, + "step": 31438 + }, + { + "epoch": 2.181597390881965, + "grad_norm": 1.1871957778930664, + "learning_rate": 9.017207586838814e-05, + "loss": 0.3465, + "step": 31439 + }, + { + "epoch": 2.181666782319062, + "grad_norm": 0.9311440587043762, + "learning_rate": 9.016648446884231e-05, + "loss": 0.2492, + "step": 31440 + }, + { + "epoch": 2.1817361737561587, + "grad_norm": 1.8670282363891602, + "learning_rate": 9.016089310033973e-05, + "loss": 0.3347, + "step": 31441 + }, + { + "epoch": 2.181805565193255, + "grad_norm": 1.1020398139953613, + "learning_rate": 9.015530176289807e-05, + "loss": 0.4339, + "step": 31442 + }, + { + "epoch": 2.181874956630352, + "grad_norm": 1.0287200212478638, + "learning_rate": 9.014971045653497e-05, + "loss": 0.2643, + "step": 31443 + }, + { + "epoch": 2.1819443480674483, + "grad_norm": 1.1094156503677368, + "learning_rate": 9.014411918126811e-05, + "loss": 0.3505, + "step": 31444 + }, + { + "epoch": 2.182013739504545, + "grad_norm": 1.2774163484573364, + "learning_rate": 9.01385279371151e-05, + "loss": 0.4179, + "step": 31445 + }, + { + "epoch": 2.182083130941642, + "grad_norm": 2.1650390625, + "learning_rate": 9.013293672409366e-05, + "loss": 0.364, + "step": 31446 + }, + { + "epoch": 2.1821525223787384, + "grad_norm": 1.2331737279891968, + "learning_rate": 9.012734554222137e-05, + "loss": 0.3888, + "step": 31447 + }, + { + "epoch": 2.182221913815835, + "grad_norm": 1.866647720336914, + "learning_rate": 9.012175439151588e-05, + "loss": 0.6197, + "step": 31448 + }, + { + "epoch": 2.1822913052529316, + "grad_norm": 1.4455885887145996, + "learning_rate": 9.011616327199488e-05, + "loss": 0.4007, + "step": 31449 + }, + { + "epoch": 2.1823606966900284, + "grad_norm": 1.1479007005691528, + "learning_rate": 9.0110572183676e-05, + "loss": 0.4227, + "step": 31450 + }, + { + "epoch": 2.1824300881271252, + "grad_norm": 0.8575451374053955, + "learning_rate": 9.010498112657689e-05, + "loss": 0.2959, + "step": 31451 + }, + { + "epoch": 2.1824994795642216, + "grad_norm": 1.2604633569717407, + "learning_rate": 9.009939010071525e-05, + "loss": 0.4287, + "step": 31452 + }, + { + "epoch": 2.1825688710013185, + "grad_norm": 1.1581085920333862, + "learning_rate": 9.009379910610867e-05, + "loss": 0.3599, + "step": 31453 + }, + { + "epoch": 2.1826382624384153, + "grad_norm": 1.4284108877182007, + "learning_rate": 9.00882081427748e-05, + "loss": 0.4151, + "step": 31454 + }, + { + "epoch": 2.1827076538755117, + "grad_norm": 1.3523865938186646, + "learning_rate": 9.00826172107313e-05, + "loss": 0.3038, + "step": 31455 + }, + { + "epoch": 2.1827770453126085, + "grad_norm": 1.487734079360962, + "learning_rate": 9.007702630999586e-05, + "loss": 0.6017, + "step": 31456 + }, + { + "epoch": 2.182846436749705, + "grad_norm": 1.2972749471664429, + "learning_rate": 9.007143544058607e-05, + "loss": 0.5029, + "step": 31457 + }, + { + "epoch": 2.1829158281868017, + "grad_norm": 0.9151338934898376, + "learning_rate": 9.00658446025196e-05, + "loss": 0.3975, + "step": 31458 + }, + { + "epoch": 2.1829852196238986, + "grad_norm": 1.1790542602539062, + "learning_rate": 9.00602537958141e-05, + "loss": 0.3555, + "step": 31459 + }, + { + "epoch": 2.183054611060995, + "grad_norm": 1.3553849458694458, + "learning_rate": 9.00546630204873e-05, + "loss": 0.4533, + "step": 31460 + }, + { + "epoch": 2.183124002498092, + "grad_norm": 1.4186758995056152, + "learning_rate": 9.00490722765567e-05, + "loss": 0.4365, + "step": 31461 + }, + { + "epoch": 2.1831933939351886, + "grad_norm": 1.1157089471817017, + "learning_rate": 9.004348156404003e-05, + "loss": 0.3272, + "step": 31462 + }, + { + "epoch": 2.183262785372285, + "grad_norm": 1.048336148262024, + "learning_rate": 9.003789088295495e-05, + "loss": 0.4274, + "step": 31463 + }, + { + "epoch": 2.183332176809382, + "grad_norm": 1.02488112449646, + "learning_rate": 9.003230023331908e-05, + "loss": 0.3472, + "step": 31464 + }, + { + "epoch": 2.1834015682464782, + "grad_norm": 1.0679184198379517, + "learning_rate": 9.002670961515008e-05, + "loss": 0.4005, + "step": 31465 + }, + { + "epoch": 2.183470959683575, + "grad_norm": 2.0479326248168945, + "learning_rate": 9.00211190284656e-05, + "loss": 0.5062, + "step": 31466 + }, + { + "epoch": 2.183540351120672, + "grad_norm": 1.2483432292938232, + "learning_rate": 9.001552847328335e-05, + "loss": 0.4235, + "step": 31467 + }, + { + "epoch": 2.1836097425577683, + "grad_norm": 1.251599669456482, + "learning_rate": 9.000993794962086e-05, + "loss": 0.4823, + "step": 31468 + }, + { + "epoch": 2.183679133994865, + "grad_norm": 1.350548267364502, + "learning_rate": 9.000434745749584e-05, + "loss": 0.3507, + "step": 31469 + }, + { + "epoch": 2.1837485254319615, + "grad_norm": 2.581362009048462, + "learning_rate": 8.999875699692592e-05, + "loss": 0.447, + "step": 31470 + }, + { + "epoch": 2.1838179168690584, + "grad_norm": 1.451354742050171, + "learning_rate": 8.99931665679288e-05, + "loss": 0.4147, + "step": 31471 + }, + { + "epoch": 2.183887308306155, + "grad_norm": 1.5353668928146362, + "learning_rate": 8.998757617052208e-05, + "loss": 0.5161, + "step": 31472 + }, + { + "epoch": 2.1839566997432516, + "grad_norm": 1.4893728494644165, + "learning_rate": 8.99819858047234e-05, + "loss": 0.4433, + "step": 31473 + }, + { + "epoch": 2.1840260911803484, + "grad_norm": 1.3924484252929688, + "learning_rate": 8.997639547055048e-05, + "loss": 0.4277, + "step": 31474 + }, + { + "epoch": 2.184095482617445, + "grad_norm": 1.2365280389785767, + "learning_rate": 8.997080516802089e-05, + "loss": 0.4525, + "step": 31475 + }, + { + "epoch": 2.1841648740545416, + "grad_norm": 1.1128911972045898, + "learning_rate": 8.996521489715229e-05, + "loss": 0.3658, + "step": 31476 + }, + { + "epoch": 2.1842342654916385, + "grad_norm": 0.8288919925689697, + "learning_rate": 8.995962465796234e-05, + "loss": 0.5023, + "step": 31477 + }, + { + "epoch": 2.184303656928735, + "grad_norm": 1.2095222473144531, + "learning_rate": 8.99540344504687e-05, + "loss": 0.3027, + "step": 31478 + }, + { + "epoch": 2.1843730483658317, + "grad_norm": 1.1891415119171143, + "learning_rate": 8.9948444274689e-05, + "loss": 0.381, + "step": 31479 + }, + { + "epoch": 2.1844424398029285, + "grad_norm": 1.2790364027023315, + "learning_rate": 8.99428541306409e-05, + "loss": 0.3548, + "step": 31480 + }, + { + "epoch": 2.184511831240025, + "grad_norm": 2.096949815750122, + "learning_rate": 8.993726401834208e-05, + "loss": 0.5741, + "step": 31481 + }, + { + "epoch": 2.1845812226771217, + "grad_norm": 1.387017846107483, + "learning_rate": 8.99316739378101e-05, + "loss": 0.4189, + "step": 31482 + }, + { + "epoch": 2.184650614114218, + "grad_norm": 1.1596652269363403, + "learning_rate": 8.992608388906267e-05, + "loss": 0.3863, + "step": 31483 + }, + { + "epoch": 2.184720005551315, + "grad_norm": 1.4824374914169312, + "learning_rate": 8.992049387211742e-05, + "loss": 0.5728, + "step": 31484 + }, + { + "epoch": 2.184789396988412, + "grad_norm": 1.6516400575637817, + "learning_rate": 8.9914903886992e-05, + "loss": 0.4471, + "step": 31485 + }, + { + "epoch": 2.184858788425508, + "grad_norm": 1.5740327835083008, + "learning_rate": 8.990931393370405e-05, + "loss": 0.4034, + "step": 31486 + }, + { + "epoch": 2.184928179862605, + "grad_norm": 2.12282395362854, + "learning_rate": 8.990372401227129e-05, + "loss": 0.5181, + "step": 31487 + }, + { + "epoch": 2.184997571299702, + "grad_norm": 1.006012201309204, + "learning_rate": 8.989813412271124e-05, + "loss": 0.3249, + "step": 31488 + }, + { + "epoch": 2.1850669627367982, + "grad_norm": 1.1536837816238403, + "learning_rate": 8.989254426504162e-05, + "loss": 0.4083, + "step": 31489 + }, + { + "epoch": 2.185136354173895, + "grad_norm": 1.3109418153762817, + "learning_rate": 8.988695443928008e-05, + "loss": 0.4521, + "step": 31490 + }, + { + "epoch": 2.1852057456109915, + "grad_norm": 1.4447449445724487, + "learning_rate": 8.988136464544423e-05, + "loss": 0.478, + "step": 31491 + }, + { + "epoch": 2.1852751370480883, + "grad_norm": 1.57283353805542, + "learning_rate": 8.987577488355175e-05, + "loss": 0.5859, + "step": 31492 + }, + { + "epoch": 2.185344528485185, + "grad_norm": 1.2226402759552002, + "learning_rate": 8.987018515362025e-05, + "loss": 0.4346, + "step": 31493 + }, + { + "epoch": 2.1854139199222815, + "grad_norm": 1.407124638557434, + "learning_rate": 8.986459545566747e-05, + "loss": 0.4649, + "step": 31494 + }, + { + "epoch": 2.1854833113593783, + "grad_norm": 1.2451475858688354, + "learning_rate": 8.985900578971092e-05, + "loss": 0.468, + "step": 31495 + }, + { + "epoch": 2.1855527027964747, + "grad_norm": 1.3307089805603027, + "learning_rate": 8.985341615576832e-05, + "loss": 0.4125, + "step": 31496 + }, + { + "epoch": 2.1856220942335716, + "grad_norm": 1.1545459032058716, + "learning_rate": 8.984782655385732e-05, + "loss": 0.3871, + "step": 31497 + }, + { + "epoch": 2.1856914856706684, + "grad_norm": 1.393884301185608, + "learning_rate": 8.984223698399554e-05, + "loss": 0.4754, + "step": 31498 + }, + { + "epoch": 2.185760877107765, + "grad_norm": 1.6577918529510498, + "learning_rate": 8.983664744620064e-05, + "loss": 0.5759, + "step": 31499 + }, + { + "epoch": 2.1858302685448616, + "grad_norm": 1.589274525642395, + "learning_rate": 8.983105794049027e-05, + "loss": 0.4184, + "step": 31500 + }, + { + "epoch": 2.185899659981958, + "grad_norm": 1.4558550119400024, + "learning_rate": 8.982546846688211e-05, + "loss": 0.4099, + "step": 31501 + }, + { + "epoch": 2.185969051419055, + "grad_norm": 3.437467336654663, + "learning_rate": 8.981987902539372e-05, + "loss": 0.4241, + "step": 31502 + }, + { + "epoch": 2.1860384428561517, + "grad_norm": 1.3049405813217163, + "learning_rate": 8.98142896160428e-05, + "loss": 0.5085, + "step": 31503 + }, + { + "epoch": 2.186107834293248, + "grad_norm": 1.6002763509750366, + "learning_rate": 8.980870023884699e-05, + "loss": 0.4315, + "step": 31504 + }, + { + "epoch": 2.186177225730345, + "grad_norm": 0.9995367527008057, + "learning_rate": 8.980311089382392e-05, + "loss": 0.359, + "step": 31505 + }, + { + "epoch": 2.1862466171674417, + "grad_norm": 1.3299434185028076, + "learning_rate": 8.979752158099126e-05, + "loss": 0.3791, + "step": 31506 + }, + { + "epoch": 2.186316008604538, + "grad_norm": 1.426027774810791, + "learning_rate": 8.979193230036664e-05, + "loss": 0.4534, + "step": 31507 + }, + { + "epoch": 2.186385400041635, + "grad_norm": 0.7328210473060608, + "learning_rate": 8.97863430519677e-05, + "loss": 0.2215, + "step": 31508 + }, + { + "epoch": 2.1864547914787313, + "grad_norm": 1.4595952033996582, + "learning_rate": 8.978075383581209e-05, + "loss": 0.3909, + "step": 31509 + }, + { + "epoch": 2.186524182915828, + "grad_norm": 1.1911836862564087, + "learning_rate": 8.977516465191747e-05, + "loss": 0.4444, + "step": 31510 + }, + { + "epoch": 2.186593574352925, + "grad_norm": 0.9427695274353027, + "learning_rate": 8.976957550030145e-05, + "loss": 0.3642, + "step": 31511 + }, + { + "epoch": 2.1866629657900214, + "grad_norm": 1.1835205554962158, + "learning_rate": 8.97639863809817e-05, + "loss": 0.4533, + "step": 31512 + }, + { + "epoch": 2.1867323572271182, + "grad_norm": 1.0068382024765015, + "learning_rate": 8.975839729397584e-05, + "loss": 0.3493, + "step": 31513 + }, + { + "epoch": 2.186801748664215, + "grad_norm": 1.1431293487548828, + "learning_rate": 8.975280823930154e-05, + "loss": 0.5215, + "step": 31514 + }, + { + "epoch": 2.1868711401013115, + "grad_norm": 1.0600045919418335, + "learning_rate": 8.974721921697647e-05, + "loss": 0.3401, + "step": 31515 + }, + { + "epoch": 2.1869405315384083, + "grad_norm": 1.171040654182434, + "learning_rate": 8.97416302270182e-05, + "loss": 0.3735, + "step": 31516 + }, + { + "epoch": 2.1870099229755047, + "grad_norm": 1.5935413837432861, + "learning_rate": 8.973604126944443e-05, + "loss": 0.4794, + "step": 31517 + }, + { + "epoch": 2.1870793144126015, + "grad_norm": 1.3181736469268799, + "learning_rate": 8.973045234427277e-05, + "loss": 0.3468, + "step": 31518 + }, + { + "epoch": 2.1871487058496983, + "grad_norm": 0.9469603300094604, + "learning_rate": 8.972486345152088e-05, + "loss": 0.3571, + "step": 31519 + }, + { + "epoch": 2.1872180972867947, + "grad_norm": 1.3727121353149414, + "learning_rate": 8.971927459120642e-05, + "loss": 0.3433, + "step": 31520 + }, + { + "epoch": 2.1872874887238916, + "grad_norm": 1.5469714403152466, + "learning_rate": 8.9713685763347e-05, + "loss": 0.3742, + "step": 31521 + }, + { + "epoch": 2.187356880160988, + "grad_norm": 1.305974006652832, + "learning_rate": 8.970809696796032e-05, + "loss": 0.4127, + "step": 31522 + }, + { + "epoch": 2.187426271598085, + "grad_norm": 1.3960963487625122, + "learning_rate": 8.970250820506394e-05, + "loss": 0.3998, + "step": 31523 + }, + { + "epoch": 2.1874956630351816, + "grad_norm": 1.4296807050704956, + "learning_rate": 8.969691947467557e-05, + "loss": 0.4476, + "step": 31524 + }, + { + "epoch": 2.187565054472278, + "grad_norm": 2.1322765350341797, + "learning_rate": 8.969133077681285e-05, + "loss": 0.4274, + "step": 31525 + }, + { + "epoch": 2.187634445909375, + "grad_norm": 1.7574567794799805, + "learning_rate": 8.968574211149337e-05, + "loss": 0.3224, + "step": 31526 + }, + { + "epoch": 2.1877038373464712, + "grad_norm": 1.5207030773162842, + "learning_rate": 8.96801534787348e-05, + "loss": 0.3553, + "step": 31527 + }, + { + "epoch": 2.187773228783568, + "grad_norm": 1.163319706916809, + "learning_rate": 8.967456487855479e-05, + "loss": 0.4018, + "step": 31528 + }, + { + "epoch": 2.187842620220665, + "grad_norm": 1.3800297975540161, + "learning_rate": 8.966897631097102e-05, + "loss": 0.4478, + "step": 31529 + }, + { + "epoch": 2.1879120116577613, + "grad_norm": 0.9844465851783752, + "learning_rate": 8.966338777600106e-05, + "loss": 0.3235, + "step": 31530 + }, + { + "epoch": 2.187981403094858, + "grad_norm": 1.1759203672409058, + "learning_rate": 8.965779927366258e-05, + "loss": 0.4857, + "step": 31531 + }, + { + "epoch": 2.188050794531955, + "grad_norm": 1.368510127067566, + "learning_rate": 8.965221080397324e-05, + "loss": 0.3284, + "step": 31532 + }, + { + "epoch": 2.1881201859690513, + "grad_norm": 1.274407982826233, + "learning_rate": 8.964662236695065e-05, + "loss": 0.3256, + "step": 31533 + }, + { + "epoch": 2.188189577406148, + "grad_norm": 1.3565287590026855, + "learning_rate": 8.964103396261248e-05, + "loss": 0.4304, + "step": 31534 + }, + { + "epoch": 2.1882589688432446, + "grad_norm": 1.5020281076431274, + "learning_rate": 8.963544559097638e-05, + "loss": 0.3777, + "step": 31535 + }, + { + "epoch": 2.1883283602803414, + "grad_norm": 1.319763422012329, + "learning_rate": 8.962985725206e-05, + "loss": 0.4871, + "step": 31536 + }, + { + "epoch": 2.1883977517174382, + "grad_norm": 3.2912232875823975, + "learning_rate": 8.962426894588091e-05, + "loss": 0.3294, + "step": 31537 + }, + { + "epoch": 2.1884671431545346, + "grad_norm": 1.0965828895568848, + "learning_rate": 8.961868067245682e-05, + "loss": 0.4312, + "step": 31538 + }, + { + "epoch": 2.1885365345916314, + "grad_norm": 1.3656699657440186, + "learning_rate": 8.961309243180533e-05, + "loss": 0.3027, + "step": 31539 + }, + { + "epoch": 2.1886059260287283, + "grad_norm": 1.4058067798614502, + "learning_rate": 8.960750422394412e-05, + "loss": 0.4036, + "step": 31540 + }, + { + "epoch": 2.1886753174658247, + "grad_norm": 1.6781997680664062, + "learning_rate": 8.960191604889082e-05, + "loss": 0.4381, + "step": 31541 + }, + { + "epoch": 2.1887447089029215, + "grad_norm": 1.1407139301300049, + "learning_rate": 8.959632790666306e-05, + "loss": 0.4403, + "step": 31542 + }, + { + "epoch": 2.188814100340018, + "grad_norm": 1.8765465021133423, + "learning_rate": 8.959073979727849e-05, + "loss": 0.5954, + "step": 31543 + }, + { + "epoch": 2.1888834917771147, + "grad_norm": 1.0644718408584595, + "learning_rate": 8.958515172075475e-05, + "loss": 0.301, + "step": 31544 + }, + { + "epoch": 2.1889528832142116, + "grad_norm": 1.3403942584991455, + "learning_rate": 8.957956367710946e-05, + "loss": 0.5151, + "step": 31545 + }, + { + "epoch": 2.189022274651308, + "grad_norm": 1.0567116737365723, + "learning_rate": 8.957397566636028e-05, + "loss": 0.3582, + "step": 31546 + }, + { + "epoch": 2.1890916660884048, + "grad_norm": 1.4501712322235107, + "learning_rate": 8.956838768852485e-05, + "loss": 0.4284, + "step": 31547 + }, + { + "epoch": 2.189161057525501, + "grad_norm": 2.1575138568878174, + "learning_rate": 8.956279974362081e-05, + "loss": 0.487, + "step": 31548 + }, + { + "epoch": 2.189230448962598, + "grad_norm": 1.2928293943405151, + "learning_rate": 8.955721183166578e-05, + "loss": 0.4183, + "step": 31549 + }, + { + "epoch": 2.189299840399695, + "grad_norm": 2.6615355014801025, + "learning_rate": 8.955162395267749e-05, + "loss": 0.7763, + "step": 31550 + }, + { + "epoch": 2.189369231836791, + "grad_norm": 1.6577702760696411, + "learning_rate": 8.954603610667346e-05, + "loss": 0.4377, + "step": 31551 + }, + { + "epoch": 2.189438623273888, + "grad_norm": 1.0272172689437866, + "learning_rate": 8.954044829367139e-05, + "loss": 0.3361, + "step": 31552 + }, + { + "epoch": 2.1895080147109844, + "grad_norm": 1.0804176330566406, + "learning_rate": 8.953486051368888e-05, + "loss": 0.3907, + "step": 31553 + }, + { + "epoch": 2.1895774061480813, + "grad_norm": 1.3068397045135498, + "learning_rate": 8.952927276674363e-05, + "loss": 0.4124, + "step": 31554 + }, + { + "epoch": 2.189646797585178, + "grad_norm": 1.801028847694397, + "learning_rate": 8.952368505285325e-05, + "loss": 0.6457, + "step": 31555 + }, + { + "epoch": 2.1897161890222745, + "grad_norm": 1.3669087886810303, + "learning_rate": 8.95180973720354e-05, + "loss": 0.4756, + "step": 31556 + }, + { + "epoch": 2.1897855804593713, + "grad_norm": 1.2460989952087402, + "learning_rate": 8.951250972430769e-05, + "loss": 0.3928, + "step": 31557 + }, + { + "epoch": 2.189854971896468, + "grad_norm": 1.686761736869812, + "learning_rate": 8.950692210968775e-05, + "loss": 0.435, + "step": 31558 + }, + { + "epoch": 2.1899243633335645, + "grad_norm": 1.0676510334014893, + "learning_rate": 8.950133452819324e-05, + "loss": 0.3406, + "step": 31559 + }, + { + "epoch": 2.1899937547706614, + "grad_norm": 2.5470638275146484, + "learning_rate": 8.949574697984183e-05, + "loss": 0.5634, + "step": 31560 + }, + { + "epoch": 2.1900631462077578, + "grad_norm": 1.4465323686599731, + "learning_rate": 8.94901594646511e-05, + "loss": 0.4214, + "step": 31561 + }, + { + "epoch": 2.1901325376448546, + "grad_norm": 1.3466403484344482, + "learning_rate": 8.948457198263872e-05, + "loss": 0.3952, + "step": 31562 + }, + { + "epoch": 2.1902019290819514, + "grad_norm": 1.1594244241714478, + "learning_rate": 8.947898453382233e-05, + "loss": 0.44, + "step": 31563 + }, + { + "epoch": 2.190271320519048, + "grad_norm": 1.7275137901306152, + "learning_rate": 8.94733971182196e-05, + "loss": 0.397, + "step": 31564 + }, + { + "epoch": 2.1903407119561447, + "grad_norm": 1.2115427255630493, + "learning_rate": 8.946780973584809e-05, + "loss": 0.572, + "step": 31565 + }, + { + "epoch": 2.1904101033932415, + "grad_norm": 1.676469087600708, + "learning_rate": 8.94622223867255e-05, + "loss": 0.6188, + "step": 31566 + }, + { + "epoch": 2.190479494830338, + "grad_norm": 1.5944441556930542, + "learning_rate": 8.945663507086943e-05, + "loss": 0.533, + "step": 31567 + }, + { + "epoch": 2.1905488862674347, + "grad_norm": 1.163773536682129, + "learning_rate": 8.945104778829755e-05, + "loss": 0.4297, + "step": 31568 + }, + { + "epoch": 2.190618277704531, + "grad_norm": 1.547195553779602, + "learning_rate": 8.944546053902748e-05, + "loss": 0.3189, + "step": 31569 + }, + { + "epoch": 2.190687669141628, + "grad_norm": 1.351800799369812, + "learning_rate": 8.943987332307692e-05, + "loss": 0.4159, + "step": 31570 + }, + { + "epoch": 2.1907570605787248, + "grad_norm": 1.6722052097320557, + "learning_rate": 8.943428614046341e-05, + "loss": 0.3947, + "step": 31571 + }, + { + "epoch": 2.190826452015821, + "grad_norm": 1.446089744567871, + "learning_rate": 8.942869899120462e-05, + "loss": 0.4697, + "step": 31572 + }, + { + "epoch": 2.190895843452918, + "grad_norm": 1.623352289199829, + "learning_rate": 8.942311187531821e-05, + "loss": 0.514, + "step": 31573 + }, + { + "epoch": 2.1909652348900144, + "grad_norm": 1.3498939275741577, + "learning_rate": 8.941752479282182e-05, + "loss": 0.4533, + "step": 31574 + }, + { + "epoch": 2.191034626327111, + "grad_norm": 1.6022707223892212, + "learning_rate": 8.941193774373305e-05, + "loss": 0.3349, + "step": 31575 + }, + { + "epoch": 2.191104017764208, + "grad_norm": 1.3889966011047363, + "learning_rate": 8.940635072806961e-05, + "loss": 0.443, + "step": 31576 + }, + { + "epoch": 2.1911734092013044, + "grad_norm": 0.9441969990730286, + "learning_rate": 8.940076374584908e-05, + "loss": 0.2963, + "step": 31577 + }, + { + "epoch": 2.1912428006384013, + "grad_norm": 1.1412456035614014, + "learning_rate": 8.939517679708909e-05, + "loss": 0.4403, + "step": 31578 + }, + { + "epoch": 2.1913121920754977, + "grad_norm": 1.8768028020858765, + "learning_rate": 8.938958988180732e-05, + "loss": 0.6356, + "step": 31579 + }, + { + "epoch": 2.1913815835125945, + "grad_norm": 1.2906172275543213, + "learning_rate": 8.938400300002137e-05, + "loss": 0.4346, + "step": 31580 + }, + { + "epoch": 2.1914509749496913, + "grad_norm": 0.9635930061340332, + "learning_rate": 8.937841615174887e-05, + "loss": 0.3031, + "step": 31581 + }, + { + "epoch": 2.1915203663867877, + "grad_norm": 1.130603551864624, + "learning_rate": 8.937282933700749e-05, + "loss": 0.4328, + "step": 31582 + }, + { + "epoch": 2.1915897578238845, + "grad_norm": 1.0893875360488892, + "learning_rate": 8.936724255581486e-05, + "loss": 0.3091, + "step": 31583 + }, + { + "epoch": 2.1916591492609814, + "grad_norm": 1.2940701246261597, + "learning_rate": 8.936165580818865e-05, + "loss": 0.4791, + "step": 31584 + }, + { + "epoch": 2.1917285406980778, + "grad_norm": 1.009712815284729, + "learning_rate": 8.935606909414641e-05, + "loss": 0.3182, + "step": 31585 + }, + { + "epoch": 2.1917979321351746, + "grad_norm": 1.4084309339523315, + "learning_rate": 8.935048241370583e-05, + "loss": 0.4519, + "step": 31586 + }, + { + "epoch": 2.191867323572271, + "grad_norm": 1.9093914031982422, + "learning_rate": 8.934489576688454e-05, + "loss": 0.3595, + "step": 31587 + }, + { + "epoch": 2.191936715009368, + "grad_norm": 1.167741060256958, + "learning_rate": 8.933930915370018e-05, + "loss": 0.246, + "step": 31588 + }, + { + "epoch": 2.1920061064464647, + "grad_norm": 1.4045032262802124, + "learning_rate": 8.933372257417037e-05, + "loss": 0.6161, + "step": 31589 + }, + { + "epoch": 2.192075497883561, + "grad_norm": 1.7695187330245972, + "learning_rate": 8.932813602831278e-05, + "loss": 0.7007, + "step": 31590 + }, + { + "epoch": 2.192144889320658, + "grad_norm": 1.3966857194900513, + "learning_rate": 8.932254951614507e-05, + "loss": 0.4196, + "step": 31591 + }, + { + "epoch": 2.1922142807577547, + "grad_norm": 1.2278636693954468, + "learning_rate": 8.931696303768478e-05, + "loss": 0.4681, + "step": 31592 + }, + { + "epoch": 2.192283672194851, + "grad_norm": 1.6929633617401123, + "learning_rate": 8.931137659294959e-05, + "loss": 0.4207, + "step": 31593 + }, + { + "epoch": 2.192353063631948, + "grad_norm": 1.1193766593933105, + "learning_rate": 8.930579018195717e-05, + "loss": 0.6094, + "step": 31594 + }, + { + "epoch": 2.1924224550690443, + "grad_norm": 1.2136435508728027, + "learning_rate": 8.930020380472512e-05, + "loss": 0.4372, + "step": 31595 + }, + { + "epoch": 2.192491846506141, + "grad_norm": 1.5139374732971191, + "learning_rate": 8.929461746127108e-05, + "loss": 0.428, + "step": 31596 + }, + { + "epoch": 2.192561237943238, + "grad_norm": 1.3546987771987915, + "learning_rate": 8.92890311516127e-05, + "loss": 0.5281, + "step": 31597 + }, + { + "epoch": 2.1926306293803344, + "grad_norm": 1.2679901123046875, + "learning_rate": 8.92834448757676e-05, + "loss": 0.3987, + "step": 31598 + }, + { + "epoch": 2.192700020817431, + "grad_norm": 1.2111552953720093, + "learning_rate": 8.927785863375346e-05, + "loss": 0.3717, + "step": 31599 + }, + { + "epoch": 2.1927694122545276, + "grad_norm": 1.9435559511184692, + "learning_rate": 8.927227242558782e-05, + "loss": 0.4694, + "step": 31600 + }, + { + "epoch": 2.1928388036916244, + "grad_norm": 1.2661831378936768, + "learning_rate": 8.926668625128838e-05, + "loss": 0.5199, + "step": 31601 + }, + { + "epoch": 2.1929081951287213, + "grad_norm": 1.1459710597991943, + "learning_rate": 8.926110011087277e-05, + "loss": 0.3375, + "step": 31602 + }, + { + "epoch": 2.1929775865658176, + "grad_norm": 2.603055477142334, + "learning_rate": 8.925551400435863e-05, + "loss": 0.4306, + "step": 31603 + }, + { + "epoch": 2.1930469780029145, + "grad_norm": 1.3873155117034912, + "learning_rate": 8.924992793176357e-05, + "loss": 0.4042, + "step": 31604 + }, + { + "epoch": 2.193116369440011, + "grad_norm": 1.4164273738861084, + "learning_rate": 8.924434189310529e-05, + "loss": 0.4391, + "step": 31605 + }, + { + "epoch": 2.1931857608771077, + "grad_norm": 1.7937397956848145, + "learning_rate": 8.923875588840134e-05, + "loss": 0.6345, + "step": 31606 + }, + { + "epoch": 2.1932551523142045, + "grad_norm": 1.2860618829727173, + "learning_rate": 8.923316991766938e-05, + "loss": 0.4495, + "step": 31607 + }, + { + "epoch": 2.193324543751301, + "grad_norm": 1.8089423179626465, + "learning_rate": 8.922758398092706e-05, + "loss": 0.4373, + "step": 31608 + }, + { + "epoch": 2.1933939351883978, + "grad_norm": 3.787997007369995, + "learning_rate": 8.9221998078192e-05, + "loss": 0.631, + "step": 31609 + }, + { + "epoch": 2.1934633266254946, + "grad_norm": 1.2881897687911987, + "learning_rate": 8.921641220948183e-05, + "loss": 0.5246, + "step": 31610 + }, + { + "epoch": 2.193532718062591, + "grad_norm": 1.3746544122695923, + "learning_rate": 8.921082637481423e-05, + "loss": 0.3963, + "step": 31611 + }, + { + "epoch": 2.193602109499688, + "grad_norm": 1.2361557483673096, + "learning_rate": 8.92052405742068e-05, + "loss": 0.4314, + "step": 31612 + }, + { + "epoch": 2.193671500936784, + "grad_norm": 1.3199249505996704, + "learning_rate": 8.919965480767715e-05, + "loss": 0.3915, + "step": 31613 + }, + { + "epoch": 2.193740892373881, + "grad_norm": 1.4835984706878662, + "learning_rate": 8.919406907524298e-05, + "loss": 0.4009, + "step": 31614 + }, + { + "epoch": 2.193810283810978, + "grad_norm": 1.431180715560913, + "learning_rate": 8.918848337692182e-05, + "loss": 0.5158, + "step": 31615 + }, + { + "epoch": 2.1938796752480743, + "grad_norm": 1.3532440662384033, + "learning_rate": 8.918289771273138e-05, + "loss": 0.5279, + "step": 31616 + }, + { + "epoch": 2.193949066685171, + "grad_norm": 1.0708597898483276, + "learning_rate": 8.917731208268928e-05, + "loss": 0.3594, + "step": 31617 + }, + { + "epoch": 2.194018458122268, + "grad_norm": 1.2197444438934326, + "learning_rate": 8.917172648681316e-05, + "loss": 0.432, + "step": 31618 + }, + { + "epoch": 2.1940878495593643, + "grad_norm": 1.3433496952056885, + "learning_rate": 8.916614092512067e-05, + "loss": 0.4692, + "step": 31619 + }, + { + "epoch": 2.194157240996461, + "grad_norm": 1.23085355758667, + "learning_rate": 8.916055539762937e-05, + "loss": 0.4164, + "step": 31620 + }, + { + "epoch": 2.1942266324335575, + "grad_norm": 1.6573262214660645, + "learning_rate": 8.915496990435695e-05, + "loss": 0.6709, + "step": 31621 + }, + { + "epoch": 2.1942960238706544, + "grad_norm": 1.4175103902816772, + "learning_rate": 8.914938444532103e-05, + "loss": 0.4025, + "step": 31622 + }, + { + "epoch": 2.194365415307751, + "grad_norm": 1.3505792617797852, + "learning_rate": 8.914379902053925e-05, + "loss": 0.4991, + "step": 31623 + }, + { + "epoch": 2.1944348067448476, + "grad_norm": 1.9644837379455566, + "learning_rate": 8.913821363002923e-05, + "loss": 0.5553, + "step": 31624 + }, + { + "epoch": 2.1945041981819444, + "grad_norm": 0.9165959358215332, + "learning_rate": 8.91326282738086e-05, + "loss": 0.3266, + "step": 31625 + }, + { + "epoch": 2.194573589619041, + "grad_norm": 2.024580717086792, + "learning_rate": 8.912704295189507e-05, + "loss": 0.5154, + "step": 31626 + }, + { + "epoch": 2.1946429810561376, + "grad_norm": 1.4123173952102661, + "learning_rate": 8.912145766430614e-05, + "loss": 0.3332, + "step": 31627 + }, + { + "epoch": 2.1947123724932345, + "grad_norm": 1.2058279514312744, + "learning_rate": 8.911587241105951e-05, + "loss": 0.3352, + "step": 31628 + }, + { + "epoch": 2.194781763930331, + "grad_norm": 1.2609995603561401, + "learning_rate": 8.911028719217281e-05, + "loss": 0.559, + "step": 31629 + }, + { + "epoch": 2.1948511553674277, + "grad_norm": 1.4775334596633911, + "learning_rate": 8.910470200766369e-05, + "loss": 0.705, + "step": 31630 + }, + { + "epoch": 2.194920546804524, + "grad_norm": 1.0370272397994995, + "learning_rate": 8.909911685754974e-05, + "loss": 0.3654, + "step": 31631 + }, + { + "epoch": 2.194989938241621, + "grad_norm": 2.9807796478271484, + "learning_rate": 8.909353174184861e-05, + "loss": 0.6364, + "step": 31632 + }, + { + "epoch": 2.1950593296787178, + "grad_norm": 1.3916774988174438, + "learning_rate": 8.9087946660578e-05, + "loss": 0.4711, + "step": 31633 + }, + { + "epoch": 2.195128721115814, + "grad_norm": 1.467641830444336, + "learning_rate": 8.90823616137554e-05, + "loss": 0.472, + "step": 31634 + }, + { + "epoch": 2.195198112552911, + "grad_norm": 1.0656436681747437, + "learning_rate": 8.907677660139853e-05, + "loss": 0.4253, + "step": 31635 + }, + { + "epoch": 2.195267503990008, + "grad_norm": 1.251880168914795, + "learning_rate": 8.907119162352502e-05, + "loss": 0.4419, + "step": 31636 + }, + { + "epoch": 2.195336895427104, + "grad_norm": 1.208026647567749, + "learning_rate": 8.906560668015247e-05, + "loss": 0.3764, + "step": 31637 + }, + { + "epoch": 2.195406286864201, + "grad_norm": 1.1468887329101562, + "learning_rate": 8.906002177129855e-05, + "loss": 0.4367, + "step": 31638 + }, + { + "epoch": 2.1954756783012974, + "grad_norm": 2.7961502075195312, + "learning_rate": 8.905443689698088e-05, + "loss": 0.425, + "step": 31639 + }, + { + "epoch": 2.1955450697383943, + "grad_norm": 1.0352416038513184, + "learning_rate": 8.90488520572171e-05, + "loss": 0.3726, + "step": 31640 + }, + { + "epoch": 2.195614461175491, + "grad_norm": 1.1539913415908813, + "learning_rate": 8.90432672520248e-05, + "loss": 0.3818, + "step": 31641 + }, + { + "epoch": 2.1956838526125875, + "grad_norm": 2.1969802379608154, + "learning_rate": 8.903768248142162e-05, + "loss": 0.512, + "step": 31642 + }, + { + "epoch": 2.1957532440496843, + "grad_norm": 1.1617823839187622, + "learning_rate": 8.903209774542521e-05, + "loss": 0.4361, + "step": 31643 + }, + { + "epoch": 2.195822635486781, + "grad_norm": 1.320627212524414, + "learning_rate": 8.902651304405319e-05, + "loss": 0.5001, + "step": 31644 + }, + { + "epoch": 2.1958920269238775, + "grad_norm": 1.2256596088409424, + "learning_rate": 8.902092837732323e-05, + "loss": 0.5621, + "step": 31645 + }, + { + "epoch": 2.1959614183609744, + "grad_norm": 1.07349693775177, + "learning_rate": 8.901534374525291e-05, + "loss": 0.2805, + "step": 31646 + }, + { + "epoch": 2.1960308097980707, + "grad_norm": 1.5105301141738892, + "learning_rate": 8.900975914785988e-05, + "loss": 0.4988, + "step": 31647 + }, + { + "epoch": 2.1961002012351676, + "grad_norm": 1.3653976917266846, + "learning_rate": 8.900417458516173e-05, + "loss": 0.4611, + "step": 31648 + }, + { + "epoch": 2.1961695926722644, + "grad_norm": 1.0967044830322266, + "learning_rate": 8.899859005717617e-05, + "loss": 0.255, + "step": 31649 + }, + { + "epoch": 2.196238984109361, + "grad_norm": 1.133219838142395, + "learning_rate": 8.899300556392075e-05, + "loss": 0.4374, + "step": 31650 + }, + { + "epoch": 2.1963083755464576, + "grad_norm": 1.507526159286499, + "learning_rate": 8.898742110541313e-05, + "loss": 0.3622, + "step": 31651 + }, + { + "epoch": 2.196377766983554, + "grad_norm": 1.3725676536560059, + "learning_rate": 8.898183668167097e-05, + "loss": 0.4875, + "step": 31652 + }, + { + "epoch": 2.196447158420651, + "grad_norm": 1.7014449834823608, + "learning_rate": 8.897625229271189e-05, + "loss": 0.5318, + "step": 31653 + }, + { + "epoch": 2.1965165498577477, + "grad_norm": 0.996874988079071, + "learning_rate": 8.897066793855348e-05, + "loss": 0.4469, + "step": 31654 + }, + { + "epoch": 2.196585941294844, + "grad_norm": 1.1163644790649414, + "learning_rate": 8.896508361921339e-05, + "loss": 0.3495, + "step": 31655 + }, + { + "epoch": 2.196655332731941, + "grad_norm": 0.9592838287353516, + "learning_rate": 8.895949933470923e-05, + "loss": 0.3015, + "step": 31656 + }, + { + "epoch": 2.1967247241690377, + "grad_norm": 1.3663880825042725, + "learning_rate": 8.895391508505866e-05, + "loss": 0.6086, + "step": 31657 + }, + { + "epoch": 2.196794115606134, + "grad_norm": 1.0312217473983765, + "learning_rate": 8.894833087027931e-05, + "loss": 0.3404, + "step": 31658 + }, + { + "epoch": 2.196863507043231, + "grad_norm": 0.9972070455551147, + "learning_rate": 8.894274669038878e-05, + "loss": 0.3194, + "step": 31659 + }, + { + "epoch": 2.1969328984803274, + "grad_norm": 1.2512013912200928, + "learning_rate": 8.893716254540477e-05, + "loss": 0.4205, + "step": 31660 + }, + { + "epoch": 2.197002289917424, + "grad_norm": 1.3439140319824219, + "learning_rate": 8.893157843534481e-05, + "loss": 0.5395, + "step": 31661 + }, + { + "epoch": 2.197071681354521, + "grad_norm": 1.3045462369918823, + "learning_rate": 8.892599436022657e-05, + "loss": 0.3805, + "step": 31662 + }, + { + "epoch": 2.1971410727916174, + "grad_norm": 1.2031238079071045, + "learning_rate": 8.892041032006769e-05, + "loss": 0.3241, + "step": 31663 + }, + { + "epoch": 2.1972104642287142, + "grad_norm": 1.467081904411316, + "learning_rate": 8.891482631488577e-05, + "loss": 0.5568, + "step": 31664 + }, + { + "epoch": 2.1972798556658106, + "grad_norm": 1.5555853843688965, + "learning_rate": 8.89092423446985e-05, + "loss": 0.4252, + "step": 31665 + }, + { + "epoch": 2.1973492471029075, + "grad_norm": 1.1658298969268799, + "learning_rate": 8.890365840952342e-05, + "loss": 0.3758, + "step": 31666 + }, + { + "epoch": 2.1974186385400043, + "grad_norm": 1.2230241298675537, + "learning_rate": 8.889807450937824e-05, + "loss": 0.3914, + "step": 31667 + }, + { + "epoch": 2.1974880299771007, + "grad_norm": 1.5633453130722046, + "learning_rate": 8.889249064428056e-05, + "loss": 0.5114, + "step": 31668 + }, + { + "epoch": 2.1975574214141975, + "grad_norm": 0.9428788423538208, + "learning_rate": 8.888690681424796e-05, + "loss": 0.3708, + "step": 31669 + }, + { + "epoch": 2.1976268128512944, + "grad_norm": 1.245862603187561, + "learning_rate": 8.888132301929811e-05, + "loss": 0.4353, + "step": 31670 + }, + { + "epoch": 2.1976962042883907, + "grad_norm": 1.6566606760025024, + "learning_rate": 8.887573925944864e-05, + "loss": 0.5284, + "step": 31671 + }, + { + "epoch": 2.1977655957254876, + "grad_norm": 1.1058202981948853, + "learning_rate": 8.887015553471715e-05, + "loss": 0.3131, + "step": 31672 + }, + { + "epoch": 2.197834987162584, + "grad_norm": 1.1939985752105713, + "learning_rate": 8.886457184512131e-05, + "loss": 0.4478, + "step": 31673 + }, + { + "epoch": 2.197904378599681, + "grad_norm": 1.6520682573318481, + "learning_rate": 8.885898819067876e-05, + "loss": 0.5111, + "step": 31674 + }, + { + "epoch": 2.1979737700367776, + "grad_norm": 1.0961792469024658, + "learning_rate": 8.885340457140705e-05, + "loss": 0.3508, + "step": 31675 + }, + { + "epoch": 2.198043161473874, + "grad_norm": 1.1808924674987793, + "learning_rate": 8.884782098732384e-05, + "loss": 0.3175, + "step": 31676 + }, + { + "epoch": 2.198112552910971, + "grad_norm": 1.2770558595657349, + "learning_rate": 8.884223743844678e-05, + "loss": 0.3945, + "step": 31677 + }, + { + "epoch": 2.1981819443480672, + "grad_norm": 1.870046615600586, + "learning_rate": 8.883665392479348e-05, + "loss": 0.4322, + "step": 31678 + }, + { + "epoch": 2.198251335785164, + "grad_norm": 1.559995412826538, + "learning_rate": 8.883107044638156e-05, + "loss": 0.7551, + "step": 31679 + }, + { + "epoch": 2.198320727222261, + "grad_norm": 0.7735399007797241, + "learning_rate": 8.882548700322868e-05, + "loss": 0.2855, + "step": 31680 + }, + { + "epoch": 2.1983901186593573, + "grad_norm": 1.5257161855697632, + "learning_rate": 8.881990359535244e-05, + "loss": 0.5267, + "step": 31681 + }, + { + "epoch": 2.198459510096454, + "grad_norm": 1.0378347635269165, + "learning_rate": 8.881432022277044e-05, + "loss": 0.3325, + "step": 31682 + }, + { + "epoch": 2.198528901533551, + "grad_norm": 1.3361235857009888, + "learning_rate": 8.880873688550034e-05, + "loss": 0.5655, + "step": 31683 + }, + { + "epoch": 2.1985982929706473, + "grad_norm": 1.3114581108093262, + "learning_rate": 8.880315358355978e-05, + "loss": 0.4784, + "step": 31684 + }, + { + "epoch": 2.198667684407744, + "grad_norm": 1.813162088394165, + "learning_rate": 8.879757031696635e-05, + "loss": 0.7599, + "step": 31685 + }, + { + "epoch": 2.1987370758448406, + "grad_norm": 1.3642961978912354, + "learning_rate": 8.879198708573769e-05, + "loss": 0.4958, + "step": 31686 + }, + { + "epoch": 2.1988064672819374, + "grad_norm": 1.3792009353637695, + "learning_rate": 8.878640388989141e-05, + "loss": 0.3106, + "step": 31687 + }, + { + "epoch": 2.1988758587190342, + "grad_norm": 1.2258062362670898, + "learning_rate": 8.878082072944521e-05, + "loss": 0.3903, + "step": 31688 + }, + { + "epoch": 2.1989452501561306, + "grad_norm": 1.1297446489334106, + "learning_rate": 8.87752376044166e-05, + "loss": 0.4328, + "step": 31689 + }, + { + "epoch": 2.1990146415932275, + "grad_norm": 1.3221487998962402, + "learning_rate": 8.876965451482327e-05, + "loss": 0.3742, + "step": 31690 + }, + { + "epoch": 2.1990840330303243, + "grad_norm": 1.7832003831863403, + "learning_rate": 8.876407146068285e-05, + "loss": 0.593, + "step": 31691 + }, + { + "epoch": 2.1991534244674207, + "grad_norm": 0.9875884056091309, + "learning_rate": 8.875848844201294e-05, + "loss": 0.4096, + "step": 31692 + }, + { + "epoch": 2.1992228159045175, + "grad_norm": 1.2632766962051392, + "learning_rate": 8.875290545883118e-05, + "loss": 0.4592, + "step": 31693 + }, + { + "epoch": 2.199292207341614, + "grad_norm": 1.4264390468597412, + "learning_rate": 8.87473225111552e-05, + "loss": 0.5754, + "step": 31694 + }, + { + "epoch": 2.1993615987787107, + "grad_norm": 1.2025010585784912, + "learning_rate": 8.874173959900264e-05, + "loss": 0.3812, + "step": 31695 + }, + { + "epoch": 2.1994309902158076, + "grad_norm": 1.428938388824463, + "learning_rate": 8.873615672239109e-05, + "loss": 0.3342, + "step": 31696 + }, + { + "epoch": 2.199500381652904, + "grad_norm": 1.7685987949371338, + "learning_rate": 8.873057388133816e-05, + "loss": 0.5896, + "step": 31697 + }, + { + "epoch": 2.199569773090001, + "grad_norm": 1.2035163640975952, + "learning_rate": 8.87249910758615e-05, + "loss": 0.4629, + "step": 31698 + }, + { + "epoch": 2.199639164527097, + "grad_norm": 1.5989325046539307, + "learning_rate": 8.871940830597878e-05, + "loss": 0.6549, + "step": 31699 + }, + { + "epoch": 2.199708555964194, + "grad_norm": 1.1974910497665405, + "learning_rate": 8.871382557170754e-05, + "loss": 0.3775, + "step": 31700 + }, + { + "epoch": 2.199777947401291, + "grad_norm": 2.357193946838379, + "learning_rate": 8.870824287306545e-05, + "loss": 0.4961, + "step": 31701 + }, + { + "epoch": 2.1998473388383872, + "grad_norm": 1.1044813394546509, + "learning_rate": 8.870266021007015e-05, + "loss": 0.2803, + "step": 31702 + }, + { + "epoch": 2.199916730275484, + "grad_norm": 1.051427960395813, + "learning_rate": 8.869707758273923e-05, + "loss": 0.3948, + "step": 31703 + }, + { + "epoch": 2.1999861217125805, + "grad_norm": 1.508939266204834, + "learning_rate": 8.869149499109031e-05, + "loss": 0.5204, + "step": 31704 + }, + { + "epoch": 2.2000555131496773, + "grad_norm": 1.3270745277404785, + "learning_rate": 8.868591243514103e-05, + "loss": 0.3949, + "step": 31705 + }, + { + "epoch": 2.200124904586774, + "grad_norm": 1.032907247543335, + "learning_rate": 8.868032991490902e-05, + "loss": 0.4108, + "step": 31706 + }, + { + "epoch": 2.2001942960238705, + "grad_norm": 1.3804713487625122, + "learning_rate": 8.867474743041187e-05, + "loss": 0.4914, + "step": 31707 + }, + { + "epoch": 2.2002636874609673, + "grad_norm": 1.1954060792922974, + "learning_rate": 8.866916498166726e-05, + "loss": 0.3492, + "step": 31708 + }, + { + "epoch": 2.200333078898064, + "grad_norm": 1.3797805309295654, + "learning_rate": 8.866358256869279e-05, + "loss": 0.357, + "step": 31709 + }, + { + "epoch": 2.2004024703351606, + "grad_norm": 1.307831883430481, + "learning_rate": 8.865800019150605e-05, + "loss": 0.4852, + "step": 31710 + }, + { + "epoch": 2.2004718617722574, + "grad_norm": 1.4490442276000977, + "learning_rate": 8.865241785012469e-05, + "loss": 0.4013, + "step": 31711 + }, + { + "epoch": 2.200541253209354, + "grad_norm": 1.7528307437896729, + "learning_rate": 8.864683554456633e-05, + "loss": 0.5025, + "step": 31712 + }, + { + "epoch": 2.2006106446464506, + "grad_norm": 1.2475723028182983, + "learning_rate": 8.864125327484858e-05, + "loss": 0.5401, + "step": 31713 + }, + { + "epoch": 2.2006800360835475, + "grad_norm": 1.117064356803894, + "learning_rate": 8.86356710409891e-05, + "loss": 0.5466, + "step": 31714 + }, + { + "epoch": 2.200749427520644, + "grad_norm": 1.2996783256530762, + "learning_rate": 8.863008884300549e-05, + "loss": 0.3506, + "step": 31715 + }, + { + "epoch": 2.2008188189577407, + "grad_norm": 1.4771702289581299, + "learning_rate": 8.862450668091537e-05, + "loss": 0.3773, + "step": 31716 + }, + { + "epoch": 2.2008882103948375, + "grad_norm": 1.0948208570480347, + "learning_rate": 8.861892455473634e-05, + "loss": 0.3337, + "step": 31717 + }, + { + "epoch": 2.200957601831934, + "grad_norm": 1.0522533655166626, + "learning_rate": 8.861334246448608e-05, + "loss": 0.3094, + "step": 31718 + }, + { + "epoch": 2.2010269932690307, + "grad_norm": 0.829657256603241, + "learning_rate": 8.860776041018216e-05, + "loss": 0.3564, + "step": 31719 + }, + { + "epoch": 2.201096384706127, + "grad_norm": 1.3909317255020142, + "learning_rate": 8.860217839184221e-05, + "loss": 0.4357, + "step": 31720 + }, + { + "epoch": 2.201165776143224, + "grad_norm": 1.4747081995010376, + "learning_rate": 8.859659640948386e-05, + "loss": 0.4579, + "step": 31721 + }, + { + "epoch": 2.201235167580321, + "grad_norm": 1.3913087844848633, + "learning_rate": 8.859101446312477e-05, + "loss": 0.2426, + "step": 31722 + }, + { + "epoch": 2.201304559017417, + "grad_norm": 0.8937862515449524, + "learning_rate": 8.858543255278247e-05, + "loss": 0.4271, + "step": 31723 + }, + { + "epoch": 2.201373950454514, + "grad_norm": 1.4060677289962769, + "learning_rate": 8.857985067847465e-05, + "loss": 0.4541, + "step": 31724 + }, + { + "epoch": 2.2014433418916104, + "grad_norm": 1.6066261529922485, + "learning_rate": 8.857426884021892e-05, + "loss": 0.4301, + "step": 31725 + }, + { + "epoch": 2.2015127333287072, + "grad_norm": 1.1754955053329468, + "learning_rate": 8.856868703803289e-05, + "loss": 0.3005, + "step": 31726 + }, + { + "epoch": 2.201582124765804, + "grad_norm": 1.4505363702774048, + "learning_rate": 8.85631052719342e-05, + "loss": 0.4662, + "step": 31727 + }, + { + "epoch": 2.2016515162029004, + "grad_norm": 0.9324424266815186, + "learning_rate": 8.855752354194044e-05, + "loss": 0.3609, + "step": 31728 + }, + { + "epoch": 2.2017209076399973, + "grad_norm": 1.4782854318618774, + "learning_rate": 8.85519418480693e-05, + "loss": 0.4636, + "step": 31729 + }, + { + "epoch": 2.2017902990770937, + "grad_norm": 1.805310845375061, + "learning_rate": 8.854636019033832e-05, + "loss": 0.6184, + "step": 31730 + }, + { + "epoch": 2.2018596905141905, + "grad_norm": 1.5917856693267822, + "learning_rate": 8.854077856876515e-05, + "loss": 0.429, + "step": 31731 + }, + { + "epoch": 2.2019290819512873, + "grad_norm": 1.4018951654434204, + "learning_rate": 8.853519698336741e-05, + "loss": 0.5553, + "step": 31732 + }, + { + "epoch": 2.2019984733883837, + "grad_norm": 1.2463912963867188, + "learning_rate": 8.852961543416272e-05, + "loss": 0.2757, + "step": 31733 + }, + { + "epoch": 2.2020678648254806, + "grad_norm": 1.2212475538253784, + "learning_rate": 8.852403392116871e-05, + "loss": 0.3906, + "step": 31734 + }, + { + "epoch": 2.2021372562625774, + "grad_norm": 1.565329670906067, + "learning_rate": 8.8518452444403e-05, + "loss": 0.4232, + "step": 31735 + }, + { + "epoch": 2.202206647699674, + "grad_norm": 1.072800636291504, + "learning_rate": 8.851287100388322e-05, + "loss": 0.3745, + "step": 31736 + }, + { + "epoch": 2.2022760391367706, + "grad_norm": 1.2633832693099976, + "learning_rate": 8.850728959962692e-05, + "loss": 0.484, + "step": 31737 + }, + { + "epoch": 2.202345430573867, + "grad_norm": 1.6611406803131104, + "learning_rate": 8.850170823165183e-05, + "loss": 0.5069, + "step": 31738 + }, + { + "epoch": 2.202414822010964, + "grad_norm": 1.686201810836792, + "learning_rate": 8.849612689997549e-05, + "loss": 0.5206, + "step": 31739 + }, + { + "epoch": 2.2024842134480607, + "grad_norm": 1.4080188274383545, + "learning_rate": 8.849054560461553e-05, + "loss": 0.3866, + "step": 31740 + }, + { + "epoch": 2.202553604885157, + "grad_norm": 1.3350706100463867, + "learning_rate": 8.848496434558958e-05, + "loss": 0.6007, + "step": 31741 + }, + { + "epoch": 2.202622996322254, + "grad_norm": 1.3242305517196655, + "learning_rate": 8.847938312291526e-05, + "loss": 0.4404, + "step": 31742 + }, + { + "epoch": 2.2026923877593507, + "grad_norm": 1.3108333349227905, + "learning_rate": 8.847380193661025e-05, + "loss": 0.2998, + "step": 31743 + }, + { + "epoch": 2.202761779196447, + "grad_norm": 1.0606218576431274, + "learning_rate": 8.846822078669206e-05, + "loss": 0.3701, + "step": 31744 + }, + { + "epoch": 2.202831170633544, + "grad_norm": 1.1765669584274292, + "learning_rate": 8.846263967317835e-05, + "loss": 0.2571, + "step": 31745 + }, + { + "epoch": 2.2029005620706403, + "grad_norm": 1.515329360961914, + "learning_rate": 8.845705859608676e-05, + "loss": 0.5116, + "step": 31746 + }, + { + "epoch": 2.202969953507737, + "grad_norm": 1.5682804584503174, + "learning_rate": 8.845147755543489e-05, + "loss": 0.3843, + "step": 31747 + }, + { + "epoch": 2.203039344944834, + "grad_norm": 1.2549076080322266, + "learning_rate": 8.844589655124036e-05, + "loss": 0.3462, + "step": 31748 + }, + { + "epoch": 2.2031087363819304, + "grad_norm": 1.3665677309036255, + "learning_rate": 8.844031558352081e-05, + "loss": 0.2679, + "step": 31749 + }, + { + "epoch": 2.203178127819027, + "grad_norm": 1.6558113098144531, + "learning_rate": 8.843473465229387e-05, + "loss": 0.5096, + "step": 31750 + }, + { + "epoch": 2.2032475192561236, + "grad_norm": 1.0540907382965088, + "learning_rate": 8.84291537575771e-05, + "loss": 0.4412, + "step": 31751 + }, + { + "epoch": 2.2033169106932204, + "grad_norm": 0.9974960088729858, + "learning_rate": 8.842357289938814e-05, + "loss": 0.3258, + "step": 31752 + }, + { + "epoch": 2.2033863021303173, + "grad_norm": 1.2626923322677612, + "learning_rate": 8.841799207774461e-05, + "loss": 0.4924, + "step": 31753 + }, + { + "epoch": 2.2034556935674137, + "grad_norm": 1.520054817199707, + "learning_rate": 8.841241129266417e-05, + "loss": 0.4902, + "step": 31754 + }, + { + "epoch": 2.2035250850045105, + "grad_norm": 1.704481601715088, + "learning_rate": 8.840683054416439e-05, + "loss": 0.5052, + "step": 31755 + }, + { + "epoch": 2.203594476441607, + "grad_norm": 1.620399832725525, + "learning_rate": 8.840124983226289e-05, + "loss": 0.4747, + "step": 31756 + }, + { + "epoch": 2.2036638678787037, + "grad_norm": 1.1836377382278442, + "learning_rate": 8.839566915697734e-05, + "loss": 0.4703, + "step": 31757 + }, + { + "epoch": 2.2037332593158006, + "grad_norm": 1.23200261592865, + "learning_rate": 8.839008851832528e-05, + "loss": 0.3855, + "step": 31758 + }, + { + "epoch": 2.203802650752897, + "grad_norm": 1.4080232381820679, + "learning_rate": 8.838450791632436e-05, + "loss": 0.3476, + "step": 31759 + }, + { + "epoch": 2.2038720421899938, + "grad_norm": 1.5651334524154663, + "learning_rate": 8.83789273509922e-05, + "loss": 0.4018, + "step": 31760 + }, + { + "epoch": 2.2039414336270906, + "grad_norm": 1.3401398658752441, + "learning_rate": 8.837334682234641e-05, + "loss": 0.3958, + "step": 31761 + }, + { + "epoch": 2.204010825064187, + "grad_norm": 1.0741422176361084, + "learning_rate": 8.836776633040463e-05, + "loss": 0.387, + "step": 31762 + }, + { + "epoch": 2.204080216501284, + "grad_norm": 1.2252275943756104, + "learning_rate": 8.836218587518444e-05, + "loss": 0.2775, + "step": 31763 + }, + { + "epoch": 2.20414960793838, + "grad_norm": 1.7190978527069092, + "learning_rate": 8.835660545670355e-05, + "loss": 0.5519, + "step": 31764 + }, + { + "epoch": 2.204218999375477, + "grad_norm": 1.3933063745498657, + "learning_rate": 8.835102507497944e-05, + "loss": 0.4337, + "step": 31765 + }, + { + "epoch": 2.204288390812574, + "grad_norm": 1.3492168188095093, + "learning_rate": 8.83454447300298e-05, + "loss": 0.4926, + "step": 31766 + }, + { + "epoch": 2.2043577822496703, + "grad_norm": 1.2411699295043945, + "learning_rate": 8.833986442187224e-05, + "loss": 0.3993, + "step": 31767 + }, + { + "epoch": 2.204427173686767, + "grad_norm": 1.533025860786438, + "learning_rate": 8.833428415052438e-05, + "loss": 0.5712, + "step": 31768 + }, + { + "epoch": 2.204496565123864, + "grad_norm": 1.1062294244766235, + "learning_rate": 8.832870391600385e-05, + "loss": 0.5545, + "step": 31769 + }, + { + "epoch": 2.2045659565609603, + "grad_norm": 1.5375425815582275, + "learning_rate": 8.832312371832823e-05, + "loss": 0.4917, + "step": 31770 + }, + { + "epoch": 2.204635347998057, + "grad_norm": 1.7776031494140625, + "learning_rate": 8.831754355751516e-05, + "loss": 0.4033, + "step": 31771 + }, + { + "epoch": 2.2047047394351535, + "grad_norm": 1.8148590326309204, + "learning_rate": 8.831196343358225e-05, + "loss": 0.4756, + "step": 31772 + }, + { + "epoch": 2.2047741308722504, + "grad_norm": 1.782007098197937, + "learning_rate": 8.830638334654711e-05, + "loss": 0.5079, + "step": 31773 + }, + { + "epoch": 2.204843522309347, + "grad_norm": 1.8073738813400269, + "learning_rate": 8.830080329642736e-05, + "loss": 0.5015, + "step": 31774 + }, + { + "epoch": 2.2049129137464436, + "grad_norm": 1.335374116897583, + "learning_rate": 8.829522328324062e-05, + "loss": 0.3856, + "step": 31775 + }, + { + "epoch": 2.2049823051835404, + "grad_norm": 1.4944671392440796, + "learning_rate": 8.828964330700449e-05, + "loss": 0.6536, + "step": 31776 + }, + { + "epoch": 2.205051696620637, + "grad_norm": 1.174066424369812, + "learning_rate": 8.828406336773659e-05, + "loss": 0.387, + "step": 31777 + }, + { + "epoch": 2.2051210880577337, + "grad_norm": 1.3755879402160645, + "learning_rate": 8.82784834654546e-05, + "loss": 0.3839, + "step": 31778 + }, + { + "epoch": 2.2051904794948305, + "grad_norm": 1.2961972951889038, + "learning_rate": 8.827290360017603e-05, + "loss": 0.3552, + "step": 31779 + }, + { + "epoch": 2.205259870931927, + "grad_norm": 1.6422854661941528, + "learning_rate": 8.826732377191854e-05, + "loss": 0.4166, + "step": 31780 + }, + { + "epoch": 2.2053292623690237, + "grad_norm": 1.390190601348877, + "learning_rate": 8.826174398069975e-05, + "loss": 0.2574, + "step": 31781 + }, + { + "epoch": 2.20539865380612, + "grad_norm": 3.0143330097198486, + "learning_rate": 8.825616422653728e-05, + "loss": 0.5067, + "step": 31782 + }, + { + "epoch": 2.205468045243217, + "grad_norm": 1.9142680168151855, + "learning_rate": 8.825058450944873e-05, + "loss": 0.4897, + "step": 31783 + }, + { + "epoch": 2.2055374366803138, + "grad_norm": 2.127380132675171, + "learning_rate": 8.824500482945173e-05, + "loss": 0.7492, + "step": 31784 + }, + { + "epoch": 2.20560682811741, + "grad_norm": 1.0946375131607056, + "learning_rate": 8.823942518656392e-05, + "loss": 0.2436, + "step": 31785 + }, + { + "epoch": 2.205676219554507, + "grad_norm": 1.3846098184585571, + "learning_rate": 8.823384558080283e-05, + "loss": 0.3459, + "step": 31786 + }, + { + "epoch": 2.205745610991604, + "grad_norm": 1.551680564880371, + "learning_rate": 8.822826601218614e-05, + "loss": 0.4749, + "step": 31787 + }, + { + "epoch": 2.2058150024287, + "grad_norm": 1.7359718084335327, + "learning_rate": 8.822268648073146e-05, + "loss": 0.4481, + "step": 31788 + }, + { + "epoch": 2.205884393865797, + "grad_norm": 1.0733649730682373, + "learning_rate": 8.821710698645637e-05, + "loss": 0.2456, + "step": 31789 + }, + { + "epoch": 2.2059537853028934, + "grad_norm": 1.8071974515914917, + "learning_rate": 8.821152752937851e-05, + "loss": 0.4503, + "step": 31790 + }, + { + "epoch": 2.2060231767399903, + "grad_norm": 1.1063587665557861, + "learning_rate": 8.820594810951548e-05, + "loss": 0.4438, + "step": 31791 + }, + { + "epoch": 2.206092568177087, + "grad_norm": 1.2723946571350098, + "learning_rate": 8.820036872688497e-05, + "loss": 0.5536, + "step": 31792 + }, + { + "epoch": 2.2061619596141835, + "grad_norm": 1.3498342037200928, + "learning_rate": 8.819478938150446e-05, + "loss": 0.5057, + "step": 31793 + }, + { + "epoch": 2.2062313510512803, + "grad_norm": 1.8705601692199707, + "learning_rate": 8.818921007339164e-05, + "loss": 0.359, + "step": 31794 + }, + { + "epoch": 2.206300742488377, + "grad_norm": 0.9588918089866638, + "learning_rate": 8.818363080256412e-05, + "loss": 0.2649, + "step": 31795 + }, + { + "epoch": 2.2063701339254735, + "grad_norm": 1.352196216583252, + "learning_rate": 8.817805156903949e-05, + "loss": 0.4392, + "step": 31796 + }, + { + "epoch": 2.2064395253625704, + "grad_norm": 1.293831467628479, + "learning_rate": 8.81724723728354e-05, + "loss": 0.4232, + "step": 31797 + }, + { + "epoch": 2.2065089167996668, + "grad_norm": 1.3471430540084839, + "learning_rate": 8.816689321396947e-05, + "loss": 0.3774, + "step": 31798 + }, + { + "epoch": 2.2065783082367636, + "grad_norm": 1.3815016746520996, + "learning_rate": 8.816131409245925e-05, + "loss": 0.4052, + "step": 31799 + }, + { + "epoch": 2.2066476996738604, + "grad_norm": 1.5703623294830322, + "learning_rate": 8.815573500832238e-05, + "loss": 0.7108, + "step": 31800 + }, + { + "epoch": 2.206717091110957, + "grad_norm": 1.273685097694397, + "learning_rate": 8.81501559615765e-05, + "loss": 0.2865, + "step": 31801 + }, + { + "epoch": 2.2067864825480537, + "grad_norm": 1.3362665176391602, + "learning_rate": 8.814457695223917e-05, + "loss": 0.4838, + "step": 31802 + }, + { + "epoch": 2.20685587398515, + "grad_norm": 1.9996638298034668, + "learning_rate": 8.813899798032806e-05, + "loss": 0.3898, + "step": 31803 + }, + { + "epoch": 2.206925265422247, + "grad_norm": 1.0931416749954224, + "learning_rate": 8.813341904586077e-05, + "loss": 0.4342, + "step": 31804 + }, + { + "epoch": 2.2069946568593437, + "grad_norm": 1.6243946552276611, + "learning_rate": 8.812784014885491e-05, + "loss": 0.452, + "step": 31805 + }, + { + "epoch": 2.20706404829644, + "grad_norm": 0.9088832139968872, + "learning_rate": 8.812226128932805e-05, + "loss": 0.3245, + "step": 31806 + }, + { + "epoch": 2.207133439733537, + "grad_norm": 1.7919687032699585, + "learning_rate": 8.811668246729782e-05, + "loss": 0.388, + "step": 31807 + }, + { + "epoch": 2.2072028311706333, + "grad_norm": 1.8719758987426758, + "learning_rate": 8.81111036827819e-05, + "loss": 0.4898, + "step": 31808 + }, + { + "epoch": 2.20727222260773, + "grad_norm": 1.9328100681304932, + "learning_rate": 8.81055249357978e-05, + "loss": 0.5634, + "step": 31809 + }, + { + "epoch": 2.207341614044827, + "grad_norm": 1.2351661920547485, + "learning_rate": 8.809994622636318e-05, + "loss": 0.5173, + "step": 31810 + }, + { + "epoch": 2.2074110054819234, + "grad_norm": 1.7013905048370361, + "learning_rate": 8.809436755449564e-05, + "loss": 0.4307, + "step": 31811 + }, + { + "epoch": 2.20748039691902, + "grad_norm": 1.7681082487106323, + "learning_rate": 8.808878892021288e-05, + "loss": 0.7028, + "step": 31812 + }, + { + "epoch": 2.207549788356117, + "grad_norm": 0.9020578861236572, + "learning_rate": 8.808321032353236e-05, + "loss": 0.2126, + "step": 31813 + }, + { + "epoch": 2.2076191797932134, + "grad_norm": 1.494079828262329, + "learning_rate": 8.807763176447175e-05, + "loss": 0.4454, + "step": 31814 + }, + { + "epoch": 2.2076885712303103, + "grad_norm": 1.6092222929000854, + "learning_rate": 8.80720532430487e-05, + "loss": 0.2845, + "step": 31815 + }, + { + "epoch": 2.2077579626674066, + "grad_norm": 1.163926124572754, + "learning_rate": 8.806647475928078e-05, + "loss": 0.4784, + "step": 31816 + }, + { + "epoch": 2.2078273541045035, + "grad_norm": 1.0807743072509766, + "learning_rate": 8.806089631318563e-05, + "loss": 0.3737, + "step": 31817 + }, + { + "epoch": 2.2078967455416003, + "grad_norm": 1.459625482559204, + "learning_rate": 8.805531790478082e-05, + "loss": 0.5038, + "step": 31818 + }, + { + "epoch": 2.2079661369786967, + "grad_norm": 0.7307684421539307, + "learning_rate": 8.804973953408405e-05, + "loss": 0.2364, + "step": 31819 + }, + { + "epoch": 2.2080355284157935, + "grad_norm": 1.2692409753799438, + "learning_rate": 8.80441612011128e-05, + "loss": 0.3317, + "step": 31820 + }, + { + "epoch": 2.2081049198528904, + "grad_norm": 1.3860962390899658, + "learning_rate": 8.803858290588477e-05, + "loss": 0.6223, + "step": 31821 + }, + { + "epoch": 2.2081743112899868, + "grad_norm": 1.6893508434295654, + "learning_rate": 8.803300464841754e-05, + "loss": 0.5682, + "step": 31822 + }, + { + "epoch": 2.2082437027270836, + "grad_norm": 0.9247175455093384, + "learning_rate": 8.802742642872874e-05, + "loss": 0.3768, + "step": 31823 + }, + { + "epoch": 2.20831309416418, + "grad_norm": 1.4644408226013184, + "learning_rate": 8.802184824683595e-05, + "loss": 0.3637, + "step": 31824 + }, + { + "epoch": 2.208382485601277, + "grad_norm": 1.064605474472046, + "learning_rate": 8.801627010275679e-05, + "loss": 0.3644, + "step": 31825 + }, + { + "epoch": 2.2084518770383736, + "grad_norm": 1.216998815536499, + "learning_rate": 8.801069199650892e-05, + "loss": 0.456, + "step": 31826 + }, + { + "epoch": 2.20852126847547, + "grad_norm": 1.2734936475753784, + "learning_rate": 8.800511392810987e-05, + "loss": 0.4405, + "step": 31827 + }, + { + "epoch": 2.208590659912567, + "grad_norm": 0.8760127425193787, + "learning_rate": 8.799953589757727e-05, + "loss": 0.4234, + "step": 31828 + }, + { + "epoch": 2.2086600513496633, + "grad_norm": 1.011373519897461, + "learning_rate": 8.799395790492875e-05, + "loss": 0.3432, + "step": 31829 + }, + { + "epoch": 2.20872944278676, + "grad_norm": 2.3305115699768066, + "learning_rate": 8.798837995018192e-05, + "loss": 0.6838, + "step": 31830 + }, + { + "epoch": 2.208798834223857, + "grad_norm": 1.012394905090332, + "learning_rate": 8.798280203335437e-05, + "loss": 0.3688, + "step": 31831 + }, + { + "epoch": 2.2088682256609533, + "grad_norm": 1.9246861934661865, + "learning_rate": 8.797722415446372e-05, + "loss": 0.5828, + "step": 31832 + }, + { + "epoch": 2.20893761709805, + "grad_norm": 1.6044563055038452, + "learning_rate": 8.797164631352763e-05, + "loss": 0.6303, + "step": 31833 + }, + { + "epoch": 2.2090070085351465, + "grad_norm": 1.347172737121582, + "learning_rate": 8.79660685105636e-05, + "loss": 0.3542, + "step": 31834 + }, + { + "epoch": 2.2090763999722434, + "grad_norm": 1.3597415685653687, + "learning_rate": 8.796049074558931e-05, + "loss": 0.4534, + "step": 31835 + }, + { + "epoch": 2.20914579140934, + "grad_norm": 1.2544852495193481, + "learning_rate": 8.795491301862234e-05, + "loss": 0.3837, + "step": 31836 + }, + { + "epoch": 2.2092151828464366, + "grad_norm": 1.555424690246582, + "learning_rate": 8.794933532968033e-05, + "loss": 0.4017, + "step": 31837 + }, + { + "epoch": 2.2092845742835334, + "grad_norm": 1.50709068775177, + "learning_rate": 8.794375767878085e-05, + "loss": 0.4666, + "step": 31838 + }, + { + "epoch": 2.2093539657206303, + "grad_norm": 1.608648419380188, + "learning_rate": 8.793818006594155e-05, + "loss": 0.5218, + "step": 31839 + }, + { + "epoch": 2.2094233571577266, + "grad_norm": 1.3845158815383911, + "learning_rate": 8.793260249118004e-05, + "loss": 0.4393, + "step": 31840 + }, + { + "epoch": 2.2094927485948235, + "grad_norm": 1.2852733135223389, + "learning_rate": 8.792702495451387e-05, + "loss": 0.3164, + "step": 31841 + }, + { + "epoch": 2.20956214003192, + "grad_norm": 1.3621461391448975, + "learning_rate": 8.79214474559607e-05, + "loss": 0.494, + "step": 31842 + }, + { + "epoch": 2.2096315314690167, + "grad_norm": 1.107617735862732, + "learning_rate": 8.79158699955381e-05, + "loss": 0.2661, + "step": 31843 + }, + { + "epoch": 2.2097009229061135, + "grad_norm": 1.6297110319137573, + "learning_rate": 8.791029257326369e-05, + "loss": 0.3252, + "step": 31844 + }, + { + "epoch": 2.20977031434321, + "grad_norm": 1.2804126739501953, + "learning_rate": 8.790471518915508e-05, + "loss": 0.3835, + "step": 31845 + }, + { + "epoch": 2.2098397057803068, + "grad_norm": 1.2132667303085327, + "learning_rate": 8.789913784322989e-05, + "loss": 0.4692, + "step": 31846 + }, + { + "epoch": 2.2099090972174036, + "grad_norm": 2.3088278770446777, + "learning_rate": 8.789356053550575e-05, + "loss": 0.32, + "step": 31847 + }, + { + "epoch": 2.2099784886545, + "grad_norm": 1.3155980110168457, + "learning_rate": 8.788798326600022e-05, + "loss": 0.3361, + "step": 31848 + }, + { + "epoch": 2.210047880091597, + "grad_norm": 1.090214729309082, + "learning_rate": 8.78824060347309e-05, + "loss": 0.4417, + "step": 31849 + }, + { + "epoch": 2.210117271528693, + "grad_norm": 1.0718985795974731, + "learning_rate": 8.787682884171542e-05, + "loss": 0.4002, + "step": 31850 + }, + { + "epoch": 2.21018666296579, + "grad_norm": 2.6994707584381104, + "learning_rate": 8.78712516869714e-05, + "loss": 0.3261, + "step": 31851 + }, + { + "epoch": 2.210256054402887, + "grad_norm": 1.5162967443466187, + "learning_rate": 8.78656745705164e-05, + "loss": 0.5698, + "step": 31852 + }, + { + "epoch": 2.2103254458399832, + "grad_norm": 1.0123059749603271, + "learning_rate": 8.786009749236808e-05, + "loss": 0.3285, + "step": 31853 + }, + { + "epoch": 2.21039483727708, + "grad_norm": 1.8081059455871582, + "learning_rate": 8.785452045254405e-05, + "loss": 0.4681, + "step": 31854 + }, + { + "epoch": 2.2104642287141765, + "grad_norm": 1.2742481231689453, + "learning_rate": 8.784894345106187e-05, + "loss": 0.3617, + "step": 31855 + }, + { + "epoch": 2.2105336201512733, + "grad_norm": 2.085247755050659, + "learning_rate": 8.784336648793915e-05, + "loss": 0.5313, + "step": 31856 + }, + { + "epoch": 2.21060301158837, + "grad_norm": 1.1104562282562256, + "learning_rate": 8.783778956319352e-05, + "loss": 0.4585, + "step": 31857 + }, + { + "epoch": 2.2106724030254665, + "grad_norm": 1.2692992687225342, + "learning_rate": 8.783221267684259e-05, + "loss": 0.3868, + "step": 31858 + }, + { + "epoch": 2.2107417944625634, + "grad_norm": 1.334486722946167, + "learning_rate": 8.782663582890394e-05, + "loss": 0.482, + "step": 31859 + }, + { + "epoch": 2.2108111858996597, + "grad_norm": 0.9998845458030701, + "learning_rate": 8.782105901939518e-05, + "loss": 0.2761, + "step": 31860 + }, + { + "epoch": 2.2108805773367566, + "grad_norm": 1.752555251121521, + "learning_rate": 8.781548224833394e-05, + "loss": 0.527, + "step": 31861 + }, + { + "epoch": 2.2109499687738534, + "grad_norm": 1.4566446542739868, + "learning_rate": 8.780990551573783e-05, + "loss": 0.41, + "step": 31862 + }, + { + "epoch": 2.21101936021095, + "grad_norm": 2.0350446701049805, + "learning_rate": 8.78043288216244e-05, + "loss": 0.6229, + "step": 31863 + }, + { + "epoch": 2.2110887516480466, + "grad_norm": 1.3545093536376953, + "learning_rate": 8.77987521660113e-05, + "loss": 0.5875, + "step": 31864 + }, + { + "epoch": 2.2111581430851435, + "grad_norm": 1.197719693183899, + "learning_rate": 8.77931755489161e-05, + "loss": 0.3372, + "step": 31865 + }, + { + "epoch": 2.21122753452224, + "grad_norm": 1.4212677478790283, + "learning_rate": 8.778759897035645e-05, + "loss": 0.543, + "step": 31866 + }, + { + "epoch": 2.2112969259593367, + "grad_norm": 1.470365285873413, + "learning_rate": 8.778202243034992e-05, + "loss": 0.4941, + "step": 31867 + }, + { + "epoch": 2.211366317396433, + "grad_norm": 1.1003000736236572, + "learning_rate": 8.777644592891418e-05, + "loss": 0.3864, + "step": 31868 + }, + { + "epoch": 2.21143570883353, + "grad_norm": 1.3793141841888428, + "learning_rate": 8.777086946606673e-05, + "loss": 0.5339, + "step": 31869 + }, + { + "epoch": 2.2115051002706267, + "grad_norm": 2.12835955619812, + "learning_rate": 8.776529304182523e-05, + "loss": 0.7371, + "step": 31870 + }, + { + "epoch": 2.211574491707723, + "grad_norm": 1.3810462951660156, + "learning_rate": 8.775971665620728e-05, + "loss": 0.4634, + "step": 31871 + }, + { + "epoch": 2.21164388314482, + "grad_norm": 1.3451100587844849, + "learning_rate": 8.775414030923049e-05, + "loss": 0.3503, + "step": 31872 + }, + { + "epoch": 2.211713274581917, + "grad_norm": 0.9600117802619934, + "learning_rate": 8.774856400091245e-05, + "loss": 0.395, + "step": 31873 + }, + { + "epoch": 2.211782666019013, + "grad_norm": 1.4272561073303223, + "learning_rate": 8.774298773127083e-05, + "loss": 0.4111, + "step": 31874 + }, + { + "epoch": 2.21185205745611, + "grad_norm": 1.2218209505081177, + "learning_rate": 8.77374115003231e-05, + "loss": 0.3826, + "step": 31875 + }, + { + "epoch": 2.2119214488932064, + "grad_norm": 1.205299973487854, + "learning_rate": 8.773183530808697e-05, + "loss": 0.3125, + "step": 31876 + }, + { + "epoch": 2.2119908403303032, + "grad_norm": 1.670498013496399, + "learning_rate": 8.772625915458003e-05, + "loss": 0.5903, + "step": 31877 + }, + { + "epoch": 2.2120602317674, + "grad_norm": 0.9511279463768005, + "learning_rate": 8.772068303981983e-05, + "loss": 0.26, + "step": 31878 + }, + { + "epoch": 2.2121296232044965, + "grad_norm": 1.5949864387512207, + "learning_rate": 8.771510696382402e-05, + "loss": 0.4481, + "step": 31879 + }, + { + "epoch": 2.2121990146415933, + "grad_norm": 1.402753472328186, + "learning_rate": 8.77095309266102e-05, + "loss": 0.5642, + "step": 31880 + }, + { + "epoch": 2.2122684060786897, + "grad_norm": 1.2311512231826782, + "learning_rate": 8.7703954928196e-05, + "loss": 0.3854, + "step": 31881 + }, + { + "epoch": 2.2123377975157865, + "grad_norm": 1.8829089403152466, + "learning_rate": 8.769837896859893e-05, + "loss": 0.3343, + "step": 31882 + }, + { + "epoch": 2.2124071889528834, + "grad_norm": 1.5803747177124023, + "learning_rate": 8.769280304783667e-05, + "loss": 0.403, + "step": 31883 + }, + { + "epoch": 2.2124765803899797, + "grad_norm": 1.4944792985916138, + "learning_rate": 8.76872271659268e-05, + "loss": 0.4585, + "step": 31884 + }, + { + "epoch": 2.2125459718270766, + "grad_norm": 1.090357780456543, + "learning_rate": 8.768165132288692e-05, + "loss": 0.2631, + "step": 31885 + }, + { + "epoch": 2.2126153632641734, + "grad_norm": 1.5197023153305054, + "learning_rate": 8.767607551873464e-05, + "loss": 0.3741, + "step": 31886 + }, + { + "epoch": 2.21268475470127, + "grad_norm": 1.4536051750183105, + "learning_rate": 8.767049975348755e-05, + "loss": 0.4214, + "step": 31887 + }, + { + "epoch": 2.2127541461383666, + "grad_norm": 1.157022476196289, + "learning_rate": 8.766492402716332e-05, + "loss": 0.4166, + "step": 31888 + }, + { + "epoch": 2.212823537575463, + "grad_norm": 1.8130158185958862, + "learning_rate": 8.765934833977945e-05, + "loss": 0.3717, + "step": 31889 + }, + { + "epoch": 2.21289292901256, + "grad_norm": 1.1971023082733154, + "learning_rate": 8.765377269135358e-05, + "loss": 0.3609, + "step": 31890 + }, + { + "epoch": 2.2129623204496567, + "grad_norm": 1.323392391204834, + "learning_rate": 8.76481970819033e-05, + "loss": 0.4679, + "step": 31891 + }, + { + "epoch": 2.213031711886753, + "grad_norm": 1.9602988958358765, + "learning_rate": 8.764262151144626e-05, + "loss": 0.4628, + "step": 31892 + }, + { + "epoch": 2.21310110332385, + "grad_norm": 1.3910183906555176, + "learning_rate": 8.763704598000003e-05, + "loss": 0.4773, + "step": 31893 + }, + { + "epoch": 2.2131704947609463, + "grad_norm": 1.12703537940979, + "learning_rate": 8.76314704875822e-05, + "loss": 0.4417, + "step": 31894 + }, + { + "epoch": 2.213239886198043, + "grad_norm": 1.6188944578170776, + "learning_rate": 8.76258950342104e-05, + "loss": 0.4131, + "step": 31895 + }, + { + "epoch": 2.21330927763514, + "grad_norm": 1.2171071767807007, + "learning_rate": 8.762031961990222e-05, + "loss": 0.3744, + "step": 31896 + }, + { + "epoch": 2.2133786690722363, + "grad_norm": 1.164249062538147, + "learning_rate": 8.761474424467522e-05, + "loss": 0.4758, + "step": 31897 + }, + { + "epoch": 2.213448060509333, + "grad_norm": 2.417996644973755, + "learning_rate": 8.760916890854704e-05, + "loss": 0.5147, + "step": 31898 + }, + { + "epoch": 2.21351745194643, + "grad_norm": 2.2732324600219727, + "learning_rate": 8.76035936115353e-05, + "loss": 0.3722, + "step": 31899 + }, + { + "epoch": 2.2135868433835264, + "grad_norm": 1.3711386919021606, + "learning_rate": 8.759801835365755e-05, + "loss": 0.518, + "step": 31900 + }, + { + "epoch": 2.2136562348206232, + "grad_norm": 2.418715715408325, + "learning_rate": 8.759244313493143e-05, + "loss": 0.554, + "step": 31901 + }, + { + "epoch": 2.2137256262577196, + "grad_norm": 1.4844720363616943, + "learning_rate": 8.758686795537455e-05, + "loss": 0.5533, + "step": 31902 + }, + { + "epoch": 2.2137950176948165, + "grad_norm": 0.929597020149231, + "learning_rate": 8.758129281500446e-05, + "loss": 0.3096, + "step": 31903 + }, + { + "epoch": 2.2138644091319133, + "grad_norm": 1.6917294263839722, + "learning_rate": 8.757571771383879e-05, + "loss": 0.6932, + "step": 31904 + }, + { + "epoch": 2.2139338005690097, + "grad_norm": 1.872865915298462, + "learning_rate": 8.757014265189514e-05, + "loss": 0.3683, + "step": 31905 + }, + { + "epoch": 2.2140031920061065, + "grad_norm": 1.5554931163787842, + "learning_rate": 8.756456762919109e-05, + "loss": 0.5142, + "step": 31906 + }, + { + "epoch": 2.214072583443203, + "grad_norm": 0.921294093132019, + "learning_rate": 8.755899264574428e-05, + "loss": 0.4266, + "step": 31907 + }, + { + "epoch": 2.2141419748802997, + "grad_norm": 2.4592649936676025, + "learning_rate": 8.755341770157227e-05, + "loss": 0.5506, + "step": 31908 + }, + { + "epoch": 2.2142113663173966, + "grad_norm": 1.336582064628601, + "learning_rate": 8.754784279669273e-05, + "loss": 0.5033, + "step": 31909 + }, + { + "epoch": 2.214280757754493, + "grad_norm": 1.288715124130249, + "learning_rate": 8.754226793112315e-05, + "loss": 0.4042, + "step": 31910 + }, + { + "epoch": 2.21435014919159, + "grad_norm": 1.9704029560089111, + "learning_rate": 8.75366931048812e-05, + "loss": 0.4227, + "step": 31911 + }, + { + "epoch": 2.2144195406286866, + "grad_norm": 1.370176911354065, + "learning_rate": 8.753111831798448e-05, + "loss": 0.4788, + "step": 31912 + }, + { + "epoch": 2.214488932065783, + "grad_norm": 1.426589846611023, + "learning_rate": 8.752554357045055e-05, + "loss": 0.2984, + "step": 31913 + }, + { + "epoch": 2.21455832350288, + "grad_norm": 1.148285984992981, + "learning_rate": 8.751996886229703e-05, + "loss": 0.4204, + "step": 31914 + }, + { + "epoch": 2.2146277149399762, + "grad_norm": 1.1368634700775146, + "learning_rate": 8.751439419354151e-05, + "loss": 0.4023, + "step": 31915 + }, + { + "epoch": 2.214697106377073, + "grad_norm": 1.1559146642684937, + "learning_rate": 8.750881956420166e-05, + "loss": 0.331, + "step": 31916 + }, + { + "epoch": 2.21476649781417, + "grad_norm": 1.6957428455352783, + "learning_rate": 8.750324497429496e-05, + "loss": 0.5328, + "step": 31917 + }, + { + "epoch": 2.2148358892512663, + "grad_norm": 1.046989917755127, + "learning_rate": 8.749767042383908e-05, + "loss": 0.3803, + "step": 31918 + }, + { + "epoch": 2.214905280688363, + "grad_norm": 0.8304651379585266, + "learning_rate": 8.749209591285159e-05, + "loss": 0.3638, + "step": 31919 + }, + { + "epoch": 2.21497467212546, + "grad_norm": 0.9609668254852295, + "learning_rate": 8.748652144135012e-05, + "loss": 0.3928, + "step": 31920 + }, + { + "epoch": 2.2150440635625563, + "grad_norm": 1.3609628677368164, + "learning_rate": 8.748094700935223e-05, + "loss": 0.328, + "step": 31921 + }, + { + "epoch": 2.215113454999653, + "grad_norm": 1.344374418258667, + "learning_rate": 8.747537261687554e-05, + "loss": 0.3531, + "step": 31922 + }, + { + "epoch": 2.2151828464367496, + "grad_norm": 0.9328548312187195, + "learning_rate": 8.746979826393771e-05, + "loss": 0.34, + "step": 31923 + }, + { + "epoch": 2.2152522378738464, + "grad_norm": 1.2075631618499756, + "learning_rate": 8.746422395055621e-05, + "loss": 0.4331, + "step": 31924 + }, + { + "epoch": 2.2153216293109432, + "grad_norm": 1.9874659776687622, + "learning_rate": 8.745864967674871e-05, + "loss": 0.4408, + "step": 31925 + }, + { + "epoch": 2.2153910207480396, + "grad_norm": 1.9690375328063965, + "learning_rate": 8.74530754425328e-05, + "loss": 0.6562, + "step": 31926 + }, + { + "epoch": 2.2154604121851365, + "grad_norm": 1.8279123306274414, + "learning_rate": 8.744750124792606e-05, + "loss": 0.4688, + "step": 31927 + }, + { + "epoch": 2.215529803622233, + "grad_norm": 1.5008419752120972, + "learning_rate": 8.744192709294615e-05, + "loss": 0.3322, + "step": 31928 + }, + { + "epoch": 2.2155991950593297, + "grad_norm": 1.0483707189559937, + "learning_rate": 8.743635297761057e-05, + "loss": 0.2719, + "step": 31929 + }, + { + "epoch": 2.2156685864964265, + "grad_norm": 1.8997328281402588, + "learning_rate": 8.743077890193699e-05, + "loss": 0.4375, + "step": 31930 + }, + { + "epoch": 2.215737977933523, + "grad_norm": 1.1230666637420654, + "learning_rate": 8.742520486594299e-05, + "loss": 0.419, + "step": 31931 + }, + { + "epoch": 2.2158073693706197, + "grad_norm": 1.388747215270996, + "learning_rate": 8.741963086964614e-05, + "loss": 0.4114, + "step": 31932 + }, + { + "epoch": 2.215876760807716, + "grad_norm": 1.2259329557418823, + "learning_rate": 8.741405691306405e-05, + "loss": 0.3477, + "step": 31933 + }, + { + "epoch": 2.215946152244813, + "grad_norm": 1.4718478918075562, + "learning_rate": 8.740848299621434e-05, + "loss": 0.4105, + "step": 31934 + }, + { + "epoch": 2.21601554368191, + "grad_norm": 1.4339569807052612, + "learning_rate": 8.740290911911457e-05, + "loss": 0.4844, + "step": 31935 + }, + { + "epoch": 2.216084935119006, + "grad_norm": 1.3070818185806274, + "learning_rate": 8.739733528178235e-05, + "loss": 0.4735, + "step": 31936 + }, + { + "epoch": 2.216154326556103, + "grad_norm": 1.1845966577529907, + "learning_rate": 8.739176148423533e-05, + "loss": 0.3798, + "step": 31937 + }, + { + "epoch": 2.2162237179932, + "grad_norm": 1.7816998958587646, + "learning_rate": 8.738618772649103e-05, + "loss": 0.4206, + "step": 31938 + }, + { + "epoch": 2.2162931094302962, + "grad_norm": 1.3123832941055298, + "learning_rate": 8.738061400856705e-05, + "loss": 0.2813, + "step": 31939 + }, + { + "epoch": 2.216362500867393, + "grad_norm": 1.0375359058380127, + "learning_rate": 8.737504033048102e-05, + "loss": 0.3396, + "step": 31940 + }, + { + "epoch": 2.2164318923044894, + "grad_norm": 1.4261106252670288, + "learning_rate": 8.736946669225052e-05, + "loss": 0.4017, + "step": 31941 + }, + { + "epoch": 2.2165012837415863, + "grad_norm": 1.1859561204910278, + "learning_rate": 8.736389309389315e-05, + "loss": 0.3572, + "step": 31942 + }, + { + "epoch": 2.216570675178683, + "grad_norm": 1.2132306098937988, + "learning_rate": 8.735831953542652e-05, + "loss": 0.3754, + "step": 31943 + }, + { + "epoch": 2.2166400666157795, + "grad_norm": 0.9282236099243164, + "learning_rate": 8.735274601686821e-05, + "loss": 0.2979, + "step": 31944 + }, + { + "epoch": 2.2167094580528763, + "grad_norm": 1.9942810535430908, + "learning_rate": 8.734717253823579e-05, + "loss": 0.5802, + "step": 31945 + }, + { + "epoch": 2.216778849489973, + "grad_norm": 1.519432544708252, + "learning_rate": 8.734159909954687e-05, + "loss": 0.5061, + "step": 31946 + }, + { + "epoch": 2.2168482409270696, + "grad_norm": 0.9708229899406433, + "learning_rate": 8.73360257008191e-05, + "loss": 0.3089, + "step": 31947 + }, + { + "epoch": 2.2169176323641664, + "grad_norm": 1.1926000118255615, + "learning_rate": 8.733045234206998e-05, + "loss": 0.5007, + "step": 31948 + }, + { + "epoch": 2.216987023801263, + "grad_norm": 1.5964549779891968, + "learning_rate": 8.732487902331717e-05, + "loss": 0.5105, + "step": 31949 + }, + { + "epoch": 2.2170564152383596, + "grad_norm": 1.4016544818878174, + "learning_rate": 8.731930574457825e-05, + "loss": 0.508, + "step": 31950 + }, + { + "epoch": 2.2171258066754564, + "grad_norm": 1.151551604270935, + "learning_rate": 8.731373250587083e-05, + "loss": 0.3611, + "step": 31951 + }, + { + "epoch": 2.217195198112553, + "grad_norm": 1.3027070760726929, + "learning_rate": 8.730815930721244e-05, + "loss": 0.2631, + "step": 31952 + }, + { + "epoch": 2.2172645895496497, + "grad_norm": 1.3718050718307495, + "learning_rate": 8.730258614862074e-05, + "loss": 0.6805, + "step": 31953 + }, + { + "epoch": 2.217333980986746, + "grad_norm": 0.954171895980835, + "learning_rate": 8.72970130301133e-05, + "loss": 0.3399, + "step": 31954 + }, + { + "epoch": 2.217403372423843, + "grad_norm": 1.76689875125885, + "learning_rate": 8.72914399517077e-05, + "loss": 0.5206, + "step": 31955 + }, + { + "epoch": 2.2174727638609397, + "grad_norm": 1.2880375385284424, + "learning_rate": 8.728586691342157e-05, + "loss": 0.41, + "step": 31956 + }, + { + "epoch": 2.217542155298036, + "grad_norm": 1.3649786710739136, + "learning_rate": 8.728029391527251e-05, + "loss": 0.37, + "step": 31957 + }, + { + "epoch": 2.217611546735133, + "grad_norm": 1.4553076028823853, + "learning_rate": 8.727472095727804e-05, + "loss": 0.4925, + "step": 31958 + }, + { + "epoch": 2.2176809381722293, + "grad_norm": 0.9967132806777954, + "learning_rate": 8.726914803945582e-05, + "loss": 0.321, + "step": 31959 + }, + { + "epoch": 2.217750329609326, + "grad_norm": 1.0244507789611816, + "learning_rate": 8.726357516182341e-05, + "loss": 0.4317, + "step": 31960 + }, + { + "epoch": 2.217819721046423, + "grad_norm": 1.6116702556610107, + "learning_rate": 8.725800232439843e-05, + "loss": 0.4998, + "step": 31961 + }, + { + "epoch": 2.2178891124835194, + "grad_norm": 1.5283876657485962, + "learning_rate": 8.725242952719843e-05, + "loss": 0.6359, + "step": 31962 + }, + { + "epoch": 2.217958503920616, + "grad_norm": 1.6695308685302734, + "learning_rate": 8.724685677024107e-05, + "loss": 0.5846, + "step": 31963 + }, + { + "epoch": 2.218027895357713, + "grad_norm": 1.1869510412216187, + "learning_rate": 8.72412840535439e-05, + "loss": 0.4284, + "step": 31964 + }, + { + "epoch": 2.2180972867948094, + "grad_norm": 1.0345388650894165, + "learning_rate": 8.72357113771245e-05, + "loss": 0.2767, + "step": 31965 + }, + { + "epoch": 2.2181666782319063, + "grad_norm": 1.7462986707687378, + "learning_rate": 8.72301387410005e-05, + "loss": 0.3409, + "step": 31966 + }, + { + "epoch": 2.2182360696690027, + "grad_norm": 1.1059443950653076, + "learning_rate": 8.722456614518944e-05, + "loss": 0.339, + "step": 31967 + }, + { + "epoch": 2.2183054611060995, + "grad_norm": 2.1851954460144043, + "learning_rate": 8.721899358970895e-05, + "loss": 0.5138, + "step": 31968 + }, + { + "epoch": 2.2183748525431963, + "grad_norm": 1.823148488998413, + "learning_rate": 8.721342107457661e-05, + "loss": 0.409, + "step": 31969 + }, + { + "epoch": 2.2184442439802927, + "grad_norm": 1.171716570854187, + "learning_rate": 8.720784859981002e-05, + "loss": 0.3346, + "step": 31970 + }, + { + "epoch": 2.2185136354173896, + "grad_norm": 1.1332497596740723, + "learning_rate": 8.720227616542681e-05, + "loss": 0.4467, + "step": 31971 + }, + { + "epoch": 2.2185830268544864, + "grad_norm": 1.0461560487747192, + "learning_rate": 8.719670377144449e-05, + "loss": 0.4133, + "step": 31972 + }, + { + "epoch": 2.2186524182915828, + "grad_norm": 1.155158519744873, + "learning_rate": 8.719113141788067e-05, + "loss": 0.2547, + "step": 31973 + }, + { + "epoch": 2.2187218097286796, + "grad_norm": 1.287348747253418, + "learning_rate": 8.718555910475299e-05, + "loss": 0.4258, + "step": 31974 + }, + { + "epoch": 2.218791201165776, + "grad_norm": 1.523848056793213, + "learning_rate": 8.7179986832079e-05, + "loss": 0.4475, + "step": 31975 + }, + { + "epoch": 2.218860592602873, + "grad_norm": 1.1872828006744385, + "learning_rate": 8.71744145998763e-05, + "loss": 0.3182, + "step": 31976 + }, + { + "epoch": 2.2189299840399697, + "grad_norm": 1.1025056838989258, + "learning_rate": 8.716884240816249e-05, + "loss": 0.4729, + "step": 31977 + }, + { + "epoch": 2.218999375477066, + "grad_norm": 1.5473010540008545, + "learning_rate": 8.71632702569552e-05, + "loss": 0.3923, + "step": 31978 + }, + { + "epoch": 2.219068766914163, + "grad_norm": 1.064772129058838, + "learning_rate": 8.715769814627193e-05, + "loss": 0.3981, + "step": 31979 + }, + { + "epoch": 2.2191381583512593, + "grad_norm": 1.524225115776062, + "learning_rate": 8.715212607613032e-05, + "loss": 0.4232, + "step": 31980 + }, + { + "epoch": 2.219207549788356, + "grad_norm": 1.8287971019744873, + "learning_rate": 8.714655404654796e-05, + "loss": 0.666, + "step": 31981 + }, + { + "epoch": 2.219276941225453, + "grad_norm": 1.5076771974563599, + "learning_rate": 8.714098205754245e-05, + "loss": 0.508, + "step": 31982 + }, + { + "epoch": 2.2193463326625493, + "grad_norm": 1.0506776571273804, + "learning_rate": 8.713541010913136e-05, + "loss": 0.4296, + "step": 31983 + }, + { + "epoch": 2.219415724099646, + "grad_norm": 1.1529524326324463, + "learning_rate": 8.712983820133228e-05, + "loss": 0.5136, + "step": 31984 + }, + { + "epoch": 2.2194851155367425, + "grad_norm": 1.1505146026611328, + "learning_rate": 8.712426633416284e-05, + "loss": 0.385, + "step": 31985 + }, + { + "epoch": 2.2195545069738394, + "grad_norm": 1.271081805229187, + "learning_rate": 8.711869450764058e-05, + "loss": 0.3693, + "step": 31986 + }, + { + "epoch": 2.219623898410936, + "grad_norm": 1.364820957183838, + "learning_rate": 8.711312272178308e-05, + "loss": 0.5995, + "step": 31987 + }, + { + "epoch": 2.2196932898480326, + "grad_norm": 1.4466485977172852, + "learning_rate": 8.710755097660797e-05, + "loss": 0.4253, + "step": 31988 + }, + { + "epoch": 2.2197626812851294, + "grad_norm": 1.851226568222046, + "learning_rate": 8.710197927213283e-05, + "loss": 0.3354, + "step": 31989 + }, + { + "epoch": 2.2198320727222263, + "grad_norm": 1.1907836198806763, + "learning_rate": 8.709640760837523e-05, + "loss": 0.3464, + "step": 31990 + }, + { + "epoch": 2.2199014641593227, + "grad_norm": 1.4181180000305176, + "learning_rate": 8.70908359853528e-05, + "loss": 0.5099, + "step": 31991 + }, + { + "epoch": 2.2199708555964195, + "grad_norm": 1.3046191930770874, + "learning_rate": 8.708526440308311e-05, + "loss": 0.3623, + "step": 31992 + }, + { + "epoch": 2.220040247033516, + "grad_norm": 1.9411224126815796, + "learning_rate": 8.707969286158372e-05, + "loss": 0.5535, + "step": 31993 + }, + { + "epoch": 2.2201096384706127, + "grad_norm": 1.9560045003890991, + "learning_rate": 8.707412136087224e-05, + "loss": 0.6842, + "step": 31994 + }, + { + "epoch": 2.2201790299077095, + "grad_norm": 1.7246932983398438, + "learning_rate": 8.706854990096626e-05, + "loss": 0.436, + "step": 31995 + }, + { + "epoch": 2.220248421344806, + "grad_norm": 1.6483526229858398, + "learning_rate": 8.706297848188336e-05, + "loss": 0.4933, + "step": 31996 + }, + { + "epoch": 2.2203178127819028, + "grad_norm": 1.2118414640426636, + "learning_rate": 8.705740710364117e-05, + "loss": 0.5097, + "step": 31997 + }, + { + "epoch": 2.2203872042189996, + "grad_norm": 1.4265961647033691, + "learning_rate": 8.705183576625722e-05, + "loss": 0.4901, + "step": 31998 + }, + { + "epoch": 2.220456595656096, + "grad_norm": 1.2652546167373657, + "learning_rate": 8.704626446974913e-05, + "loss": 0.5016, + "step": 31999 + }, + { + "epoch": 2.220525987093193, + "grad_norm": 1.2033520936965942, + "learning_rate": 8.704069321413447e-05, + "loss": 0.3971, + "step": 32000 + }, + { + "epoch": 2.220525987093193, + "eval_loss": 0.5528601408004761, + "eval_runtime": 3823.9012, + "eval_samples_per_second": 6.347, + "eval_steps_per_second": 0.793, + "step": 32000 + }, + { + "epoch": 2.220595378530289, + "grad_norm": 1.404106855392456, + "learning_rate": 8.703512199943086e-05, + "loss": 0.415, + "step": 32001 + }, + { + "epoch": 2.220664769967386, + "grad_norm": 2.5165536403656006, + "learning_rate": 8.702955082565584e-05, + "loss": 0.5545, + "step": 32002 + }, + { + "epoch": 2.220734161404483, + "grad_norm": 1.129412293434143, + "learning_rate": 8.702397969282702e-05, + "loss": 0.3531, + "step": 32003 + }, + { + "epoch": 2.2208035528415793, + "grad_norm": 1.1403751373291016, + "learning_rate": 8.7018408600962e-05, + "loss": 0.3355, + "step": 32004 + }, + { + "epoch": 2.220872944278676, + "grad_norm": 1.2611476182937622, + "learning_rate": 8.701283755007835e-05, + "loss": 0.485, + "step": 32005 + }, + { + "epoch": 2.2209423357157725, + "grad_norm": 2.388216018676758, + "learning_rate": 8.700726654019371e-05, + "loss": 0.587, + "step": 32006 + }, + { + "epoch": 2.2210117271528693, + "grad_norm": 1.204573154449463, + "learning_rate": 8.700169557132558e-05, + "loss": 0.2621, + "step": 32007 + }, + { + "epoch": 2.221081118589966, + "grad_norm": 1.258046269416809, + "learning_rate": 8.699612464349159e-05, + "loss": 0.426, + "step": 32008 + }, + { + "epoch": 2.2211505100270625, + "grad_norm": 1.11225426197052, + "learning_rate": 8.699055375670933e-05, + "loss": 0.4126, + "step": 32009 + }, + { + "epoch": 2.2212199014641594, + "grad_norm": 1.4293767213821411, + "learning_rate": 8.698498291099637e-05, + "loss": 0.4064, + "step": 32010 + }, + { + "epoch": 2.2212892929012558, + "grad_norm": 1.4234049320220947, + "learning_rate": 8.697941210637032e-05, + "loss": 0.3872, + "step": 32011 + }, + { + "epoch": 2.2213586843383526, + "grad_norm": 1.2733210325241089, + "learning_rate": 8.697384134284875e-05, + "loss": 0.4468, + "step": 32012 + }, + { + "epoch": 2.2214280757754494, + "grad_norm": 1.590476393699646, + "learning_rate": 8.696827062044929e-05, + "loss": 0.5866, + "step": 32013 + }, + { + "epoch": 2.221497467212546, + "grad_norm": 1.087201476097107, + "learning_rate": 8.696269993918947e-05, + "loss": 0.2938, + "step": 32014 + }, + { + "epoch": 2.2215668586496427, + "grad_norm": 1.0625920295715332, + "learning_rate": 8.695712929908686e-05, + "loss": 0.4607, + "step": 32015 + }, + { + "epoch": 2.2216362500867395, + "grad_norm": 1.0220221281051636, + "learning_rate": 8.69515587001591e-05, + "loss": 0.3215, + "step": 32016 + }, + { + "epoch": 2.221705641523836, + "grad_norm": 2.0147457122802734, + "learning_rate": 8.694598814242376e-05, + "loss": 0.4116, + "step": 32017 + }, + { + "epoch": 2.2217750329609327, + "grad_norm": 1.3237143754959106, + "learning_rate": 8.694041762589841e-05, + "loss": 0.3659, + "step": 32018 + }, + { + "epoch": 2.221844424398029, + "grad_norm": 1.6343923807144165, + "learning_rate": 8.693484715060065e-05, + "loss": 0.3626, + "step": 32019 + }, + { + "epoch": 2.221913815835126, + "grad_norm": 1.4753637313842773, + "learning_rate": 8.692927671654812e-05, + "loss": 0.6247, + "step": 32020 + }, + { + "epoch": 2.2219832072722228, + "grad_norm": 1.4020895957946777, + "learning_rate": 8.692370632375826e-05, + "loss": 0.4357, + "step": 32021 + }, + { + "epoch": 2.222052598709319, + "grad_norm": 1.6467556953430176, + "learning_rate": 8.691813597224878e-05, + "loss": 0.5376, + "step": 32022 + }, + { + "epoch": 2.222121990146416, + "grad_norm": 1.3225891590118408, + "learning_rate": 8.691256566203721e-05, + "loss": 0.3342, + "step": 32023 + }, + { + "epoch": 2.222191381583513, + "grad_norm": 1.3886899948120117, + "learning_rate": 8.690699539314116e-05, + "loss": 0.5762, + "step": 32024 + }, + { + "epoch": 2.222260773020609, + "grad_norm": 0.9571053385734558, + "learning_rate": 8.69014251655782e-05, + "loss": 0.4009, + "step": 32025 + }, + { + "epoch": 2.222330164457706, + "grad_norm": 1.4020881652832031, + "learning_rate": 8.689585497936597e-05, + "loss": 0.3253, + "step": 32026 + }, + { + "epoch": 2.2223995558948024, + "grad_norm": 1.7551802396774292, + "learning_rate": 8.689028483452195e-05, + "loss": 0.4227, + "step": 32027 + }, + { + "epoch": 2.2224689473318993, + "grad_norm": 1.156236171722412, + "learning_rate": 8.688471473106379e-05, + "loss": 0.3875, + "step": 32028 + }, + { + "epoch": 2.222538338768996, + "grad_norm": 4.337749481201172, + "learning_rate": 8.687914466900908e-05, + "loss": 0.318, + "step": 32029 + }, + { + "epoch": 2.2226077302060925, + "grad_norm": 0.986740231513977, + "learning_rate": 8.687357464837536e-05, + "loss": 0.407, + "step": 32030 + }, + { + "epoch": 2.2226771216431893, + "grad_norm": 1.2761727571487427, + "learning_rate": 8.686800466918025e-05, + "loss": 0.4386, + "step": 32031 + }, + { + "epoch": 2.2227465130802857, + "grad_norm": 1.365011215209961, + "learning_rate": 8.686243473144135e-05, + "loss": 0.4117, + "step": 32032 + }, + { + "epoch": 2.2228159045173825, + "grad_norm": 1.1530771255493164, + "learning_rate": 8.685686483517622e-05, + "loss": 0.4514, + "step": 32033 + }, + { + "epoch": 2.2228852959544794, + "grad_norm": 1.7758501768112183, + "learning_rate": 8.685129498040242e-05, + "loss": 0.4685, + "step": 32034 + }, + { + "epoch": 2.2229546873915758, + "grad_norm": 1.8551816940307617, + "learning_rate": 8.684572516713756e-05, + "loss": 0.4982, + "step": 32035 + }, + { + "epoch": 2.2230240788286726, + "grad_norm": 1.01779305934906, + "learning_rate": 8.684015539539924e-05, + "loss": 0.3901, + "step": 32036 + }, + { + "epoch": 2.223093470265769, + "grad_norm": 0.942548394203186, + "learning_rate": 8.683458566520499e-05, + "loss": 0.2736, + "step": 32037 + }, + { + "epoch": 2.223162861702866, + "grad_norm": 1.0710375308990479, + "learning_rate": 8.682901597657243e-05, + "loss": 0.3472, + "step": 32038 + }, + { + "epoch": 2.2232322531399626, + "grad_norm": 1.3972548246383667, + "learning_rate": 8.682344632951915e-05, + "loss": 0.4166, + "step": 32039 + }, + { + "epoch": 2.223301644577059, + "grad_norm": 1.2794227600097656, + "learning_rate": 8.681787672406277e-05, + "loss": 0.4889, + "step": 32040 + }, + { + "epoch": 2.223371036014156, + "grad_norm": 1.4742107391357422, + "learning_rate": 8.681230716022076e-05, + "loss": 0.4573, + "step": 32041 + }, + { + "epoch": 2.2234404274512527, + "grad_norm": 1.2252342700958252, + "learning_rate": 8.680673763801077e-05, + "loss": 0.4386, + "step": 32042 + }, + { + "epoch": 2.223509818888349, + "grad_norm": 1.3995532989501953, + "learning_rate": 8.680116815745039e-05, + "loss": 0.4338, + "step": 32043 + }, + { + "epoch": 2.223579210325446, + "grad_norm": 1.5892645120620728, + "learning_rate": 8.679559871855718e-05, + "loss": 0.4329, + "step": 32044 + }, + { + "epoch": 2.2236486017625423, + "grad_norm": 1.1792078018188477, + "learning_rate": 8.679002932134874e-05, + "loss": 0.3613, + "step": 32045 + }, + { + "epoch": 2.223717993199639, + "grad_norm": 1.297179102897644, + "learning_rate": 8.678445996584264e-05, + "loss": 0.4987, + "step": 32046 + }, + { + "epoch": 2.223787384636736, + "grad_norm": 1.488998532295227, + "learning_rate": 8.67788906520565e-05, + "loss": 0.5921, + "step": 32047 + }, + { + "epoch": 2.2238567760738324, + "grad_norm": 1.5890381336212158, + "learning_rate": 8.677332138000784e-05, + "loss": 0.3804, + "step": 32048 + }, + { + "epoch": 2.223926167510929, + "grad_norm": 0.9876523613929749, + "learning_rate": 8.676775214971426e-05, + "loss": 0.3348, + "step": 32049 + }, + { + "epoch": 2.223995558948026, + "grad_norm": 1.2183256149291992, + "learning_rate": 8.676218296119336e-05, + "loss": 0.4476, + "step": 32050 + }, + { + "epoch": 2.2240649503851224, + "grad_norm": 1.6403685808181763, + "learning_rate": 8.675661381446273e-05, + "loss": 0.5742, + "step": 32051 + }, + { + "epoch": 2.2241343418222193, + "grad_norm": 1.9721941947937012, + "learning_rate": 8.675104470953991e-05, + "loss": 0.5508, + "step": 32052 + }, + { + "epoch": 2.2242037332593156, + "grad_norm": 1.1846200227737427, + "learning_rate": 8.67454756464425e-05, + "loss": 0.3679, + "step": 32053 + }, + { + "epoch": 2.2242731246964125, + "grad_norm": 0.9229316115379333, + "learning_rate": 8.673990662518813e-05, + "loss": 0.3969, + "step": 32054 + }, + { + "epoch": 2.2243425161335093, + "grad_norm": 1.7879979610443115, + "learning_rate": 8.673433764579431e-05, + "loss": 0.4845, + "step": 32055 + }, + { + "epoch": 2.2244119075706057, + "grad_norm": 1.2808703184127808, + "learning_rate": 8.672876870827865e-05, + "loss": 0.428, + "step": 32056 + }, + { + "epoch": 2.2244812990077025, + "grad_norm": 1.5209388732910156, + "learning_rate": 8.67231998126587e-05, + "loss": 0.3584, + "step": 32057 + }, + { + "epoch": 2.224550690444799, + "grad_norm": 1.2605717182159424, + "learning_rate": 8.67176309589521e-05, + "loss": 0.4659, + "step": 32058 + }, + { + "epoch": 2.2246200818818958, + "grad_norm": 1.3762364387512207, + "learning_rate": 8.671206214717638e-05, + "loss": 0.4109, + "step": 32059 + }, + { + "epoch": 2.2246894733189926, + "grad_norm": 1.3566995859146118, + "learning_rate": 8.670649337734913e-05, + "loss": 0.6293, + "step": 32060 + }, + { + "epoch": 2.224758864756089, + "grad_norm": 1.4413058757781982, + "learning_rate": 8.670092464948799e-05, + "loss": 0.4335, + "step": 32061 + }, + { + "epoch": 2.224828256193186, + "grad_norm": 1.039006233215332, + "learning_rate": 8.669535596361044e-05, + "loss": 0.4014, + "step": 32062 + }, + { + "epoch": 2.224897647630282, + "grad_norm": 1.0579653978347778, + "learning_rate": 8.668978731973412e-05, + "loss": 0.3798, + "step": 32063 + }, + { + "epoch": 2.224967039067379, + "grad_norm": 1.4244606494903564, + "learning_rate": 8.668421871787657e-05, + "loss": 0.7081, + "step": 32064 + }, + { + "epoch": 2.225036430504476, + "grad_norm": 0.8532218933105469, + "learning_rate": 8.667865015805543e-05, + "loss": 0.2365, + "step": 32065 + }, + { + "epoch": 2.2251058219415722, + "grad_norm": 0.8419503569602966, + "learning_rate": 8.667308164028823e-05, + "loss": 0.2361, + "step": 32066 + }, + { + "epoch": 2.225175213378669, + "grad_norm": 1.290067434310913, + "learning_rate": 8.666751316459258e-05, + "loss": 0.4872, + "step": 32067 + }, + { + "epoch": 2.225244604815766, + "grad_norm": 1.2864292860031128, + "learning_rate": 8.666194473098606e-05, + "loss": 0.2979, + "step": 32068 + }, + { + "epoch": 2.2253139962528623, + "grad_norm": 1.1697181463241577, + "learning_rate": 8.66563763394862e-05, + "loss": 0.4973, + "step": 32069 + }, + { + "epoch": 2.225383387689959, + "grad_norm": 1.4450359344482422, + "learning_rate": 8.665080799011061e-05, + "loss": 0.4377, + "step": 32070 + }, + { + "epoch": 2.2254527791270555, + "grad_norm": 1.4812992811203003, + "learning_rate": 8.66452396828769e-05, + "loss": 0.5624, + "step": 32071 + }, + { + "epoch": 2.2255221705641524, + "grad_norm": 2.842381477355957, + "learning_rate": 8.66396714178026e-05, + "loss": 0.5938, + "step": 32072 + }, + { + "epoch": 2.225591562001249, + "grad_norm": 1.079250693321228, + "learning_rate": 8.663410319490531e-05, + "loss": 0.4222, + "step": 32073 + }, + { + "epoch": 2.2256609534383456, + "grad_norm": 1.2962356805801392, + "learning_rate": 8.662853501420259e-05, + "loss": 0.4903, + "step": 32074 + }, + { + "epoch": 2.2257303448754424, + "grad_norm": 1.1439133882522583, + "learning_rate": 8.662296687571207e-05, + "loss": 0.4617, + "step": 32075 + }, + { + "epoch": 2.2257997363125392, + "grad_norm": 1.3090171813964844, + "learning_rate": 8.661739877945127e-05, + "loss": 0.4027, + "step": 32076 + }, + { + "epoch": 2.2258691277496356, + "grad_norm": 1.4862641096115112, + "learning_rate": 8.661183072543778e-05, + "loss": 0.4215, + "step": 32077 + }, + { + "epoch": 2.2259385191867325, + "grad_norm": 1.357262372970581, + "learning_rate": 8.660626271368919e-05, + "loss": 0.3792, + "step": 32078 + }, + { + "epoch": 2.226007910623829, + "grad_norm": 1.2743706703186035, + "learning_rate": 8.660069474422308e-05, + "loss": 0.3444, + "step": 32079 + }, + { + "epoch": 2.2260773020609257, + "grad_norm": 1.27507746219635, + "learning_rate": 8.659512681705701e-05, + "loss": 0.4942, + "step": 32080 + }, + { + "epoch": 2.2261466934980225, + "grad_norm": 1.6458613872528076, + "learning_rate": 8.658955893220857e-05, + "loss": 0.4804, + "step": 32081 + }, + { + "epoch": 2.226216084935119, + "grad_norm": 1.954442024230957, + "learning_rate": 8.658399108969539e-05, + "loss": 0.4807, + "step": 32082 + }, + { + "epoch": 2.2262854763722157, + "grad_norm": 1.2916812896728516, + "learning_rate": 8.657842328953494e-05, + "loss": 0.4176, + "step": 32083 + }, + { + "epoch": 2.226354867809312, + "grad_norm": 1.0443699359893799, + "learning_rate": 8.657285553174485e-05, + "loss": 0.3922, + "step": 32084 + }, + { + "epoch": 2.226424259246409, + "grad_norm": 1.0665363073349, + "learning_rate": 8.656728781634271e-05, + "loss": 0.3955, + "step": 32085 + }, + { + "epoch": 2.226493650683506, + "grad_norm": 1.2829700708389282, + "learning_rate": 8.65617201433461e-05, + "loss": 0.3971, + "step": 32086 + }, + { + "epoch": 2.226563042120602, + "grad_norm": 0.909497082233429, + "learning_rate": 8.655615251277256e-05, + "loss": 0.3386, + "step": 32087 + }, + { + "epoch": 2.226632433557699, + "grad_norm": 1.2684723138809204, + "learning_rate": 8.655058492463968e-05, + "loss": 0.3718, + "step": 32088 + }, + { + "epoch": 2.2267018249947954, + "grad_norm": 1.268350601196289, + "learning_rate": 8.654501737896507e-05, + "loss": 0.2933, + "step": 32089 + }, + { + "epoch": 2.2267712164318922, + "grad_norm": 1.1947886943817139, + "learning_rate": 8.653944987576627e-05, + "loss": 0.4745, + "step": 32090 + }, + { + "epoch": 2.226840607868989, + "grad_norm": 1.175292730331421, + "learning_rate": 8.653388241506087e-05, + "loss": 0.4478, + "step": 32091 + }, + { + "epoch": 2.2269099993060855, + "grad_norm": 1.1692407131195068, + "learning_rate": 8.652831499686641e-05, + "loss": 0.4459, + "step": 32092 + }, + { + "epoch": 2.2269793907431823, + "grad_norm": 1.090314507484436, + "learning_rate": 8.652274762120051e-05, + "loss": 0.3362, + "step": 32093 + }, + { + "epoch": 2.227048782180279, + "grad_norm": 2.081141471862793, + "learning_rate": 8.651718028808073e-05, + "loss": 0.4402, + "step": 32094 + }, + { + "epoch": 2.2271181736173755, + "grad_norm": 1.0347682237625122, + "learning_rate": 8.651161299752463e-05, + "loss": 0.3317, + "step": 32095 + }, + { + "epoch": 2.2271875650544724, + "grad_norm": 1.938501238822937, + "learning_rate": 8.650604574954987e-05, + "loss": 0.5736, + "step": 32096 + }, + { + "epoch": 2.2272569564915687, + "grad_norm": 1.076650619506836, + "learning_rate": 8.650047854417391e-05, + "loss": 0.3326, + "step": 32097 + }, + { + "epoch": 2.2273263479286656, + "grad_norm": 1.1973297595977783, + "learning_rate": 8.649491138141438e-05, + "loss": 0.3743, + "step": 32098 + }, + { + "epoch": 2.2273957393657624, + "grad_norm": 1.1689164638519287, + "learning_rate": 8.648934426128885e-05, + "loss": 0.3549, + "step": 32099 + }, + { + "epoch": 2.227465130802859, + "grad_norm": 1.6048951148986816, + "learning_rate": 8.648377718381487e-05, + "loss": 0.5998, + "step": 32100 + }, + { + "epoch": 2.2275345222399556, + "grad_norm": 1.5339267253875732, + "learning_rate": 8.647821014901006e-05, + "loss": 0.5346, + "step": 32101 + }, + { + "epoch": 2.2276039136770525, + "grad_norm": 1.3741860389709473, + "learning_rate": 8.6472643156892e-05, + "loss": 0.3858, + "step": 32102 + }, + { + "epoch": 2.227673305114149, + "grad_norm": 0.9890936613082886, + "learning_rate": 8.646707620747819e-05, + "loss": 0.2553, + "step": 32103 + }, + { + "epoch": 2.2277426965512457, + "grad_norm": 1.039496898651123, + "learning_rate": 8.646150930078626e-05, + "loss": 0.3422, + "step": 32104 + }, + { + "epoch": 2.227812087988342, + "grad_norm": 1.2952255010604858, + "learning_rate": 8.645594243683377e-05, + "loss": 0.3373, + "step": 32105 + }, + { + "epoch": 2.227881479425439, + "grad_norm": 1.194684386253357, + "learning_rate": 8.645037561563832e-05, + "loss": 0.4021, + "step": 32106 + }, + { + "epoch": 2.2279508708625357, + "grad_norm": 1.1965219974517822, + "learning_rate": 8.644480883721743e-05, + "loss": 0.4821, + "step": 32107 + }, + { + "epoch": 2.228020262299632, + "grad_norm": 1.0920217037200928, + "learning_rate": 8.643924210158874e-05, + "loss": 0.4566, + "step": 32108 + }, + { + "epoch": 2.228089653736729, + "grad_norm": 0.9896678328514099, + "learning_rate": 8.64336754087698e-05, + "loss": 0.4665, + "step": 32109 + }, + { + "epoch": 2.2281590451738253, + "grad_norm": 1.7391780614852905, + "learning_rate": 8.642810875877814e-05, + "loss": 0.5322, + "step": 32110 + }, + { + "epoch": 2.228228436610922, + "grad_norm": 1.59073805809021, + "learning_rate": 8.642254215163138e-05, + "loss": 0.385, + "step": 32111 + }, + { + "epoch": 2.228297828048019, + "grad_norm": 0.9417212605476379, + "learning_rate": 8.641697558734705e-05, + "loss": 0.2842, + "step": 32112 + }, + { + "epoch": 2.2283672194851154, + "grad_norm": 1.130954384803772, + "learning_rate": 8.641140906594278e-05, + "loss": 0.4211, + "step": 32113 + }, + { + "epoch": 2.2284366109222122, + "grad_norm": 1.5726698637008667, + "learning_rate": 8.640584258743609e-05, + "loss": 0.4611, + "step": 32114 + }, + { + "epoch": 2.228506002359309, + "grad_norm": 1.8804419040679932, + "learning_rate": 8.64002761518446e-05, + "loss": 0.3349, + "step": 32115 + }, + { + "epoch": 2.2285753937964055, + "grad_norm": 1.580851674079895, + "learning_rate": 8.639470975918589e-05, + "loss": 0.3492, + "step": 32116 + }, + { + "epoch": 2.2286447852335023, + "grad_norm": 1.2482351064682007, + "learning_rate": 8.638914340947745e-05, + "loss": 0.3797, + "step": 32117 + }, + { + "epoch": 2.2287141766705987, + "grad_norm": 1.2380890846252441, + "learning_rate": 8.638357710273693e-05, + "loss": 0.3626, + "step": 32118 + }, + { + "epoch": 2.2287835681076955, + "grad_norm": 1.6326991319656372, + "learning_rate": 8.637801083898187e-05, + "loss": 0.4276, + "step": 32119 + }, + { + "epoch": 2.2288529595447923, + "grad_norm": 1.4578152894973755, + "learning_rate": 8.637244461822984e-05, + "loss": 0.37, + "step": 32120 + }, + { + "epoch": 2.2289223509818887, + "grad_norm": 1.5101866722106934, + "learning_rate": 8.636687844049845e-05, + "loss": 0.5565, + "step": 32121 + }, + { + "epoch": 2.2289917424189856, + "grad_norm": 1.439644455909729, + "learning_rate": 8.636131230580522e-05, + "loss": 0.4957, + "step": 32122 + }, + { + "epoch": 2.229061133856082, + "grad_norm": 1.009350299835205, + "learning_rate": 8.635574621416776e-05, + "loss": 0.3443, + "step": 32123 + }, + { + "epoch": 2.229130525293179, + "grad_norm": 1.0793737173080444, + "learning_rate": 8.635018016560361e-05, + "loss": 0.4103, + "step": 32124 + }, + { + "epoch": 2.2291999167302756, + "grad_norm": 1.2873834371566772, + "learning_rate": 8.634461416013037e-05, + "loss": 0.4974, + "step": 32125 + }, + { + "epoch": 2.229269308167372, + "grad_norm": 1.1901286840438843, + "learning_rate": 8.633904819776559e-05, + "loss": 0.5838, + "step": 32126 + }, + { + "epoch": 2.229338699604469, + "grad_norm": 1.4729182720184326, + "learning_rate": 8.633348227852686e-05, + "loss": 0.4177, + "step": 32127 + }, + { + "epoch": 2.2294080910415657, + "grad_norm": 1.2608999013900757, + "learning_rate": 8.632791640243172e-05, + "loss": 0.3778, + "step": 32128 + }, + { + "epoch": 2.229477482478662, + "grad_norm": 1.749709129333496, + "learning_rate": 8.632235056949778e-05, + "loss": 0.3641, + "step": 32129 + }, + { + "epoch": 2.229546873915759, + "grad_norm": 1.318896770477295, + "learning_rate": 8.631678477974262e-05, + "loss": 0.4558, + "step": 32130 + }, + { + "epoch": 2.2296162653528553, + "grad_norm": 1.3731427192687988, + "learning_rate": 8.631121903318374e-05, + "loss": 0.5088, + "step": 32131 + }, + { + "epoch": 2.229685656789952, + "grad_norm": 1.1303051710128784, + "learning_rate": 8.630565332983875e-05, + "loss": 0.2495, + "step": 32132 + }, + { + "epoch": 2.229755048227049, + "grad_norm": 1.3927321434020996, + "learning_rate": 8.630008766972524e-05, + "loss": 0.3501, + "step": 32133 + }, + { + "epoch": 2.2298244396641453, + "grad_norm": 2.3968148231506348, + "learning_rate": 8.629452205286075e-05, + "loss": 0.5882, + "step": 32134 + }, + { + "epoch": 2.229893831101242, + "grad_norm": 1.5121731758117676, + "learning_rate": 8.628895647926287e-05, + "loss": 0.3986, + "step": 32135 + }, + { + "epoch": 2.2299632225383386, + "grad_norm": 1.0403019189834595, + "learning_rate": 8.628339094894916e-05, + "loss": 0.3636, + "step": 32136 + }, + { + "epoch": 2.2300326139754354, + "grad_norm": 2.2271900177001953, + "learning_rate": 8.627782546193725e-05, + "loss": 0.6024, + "step": 32137 + }, + { + "epoch": 2.2301020054125322, + "grad_norm": 1.005279302597046, + "learning_rate": 8.627226001824458e-05, + "loss": 0.3653, + "step": 32138 + }, + { + "epoch": 2.2301713968496286, + "grad_norm": 1.0017213821411133, + "learning_rate": 8.626669461788882e-05, + "loss": 0.3204, + "step": 32139 + }, + { + "epoch": 2.2302407882867255, + "grad_norm": 1.0645225048065186, + "learning_rate": 8.626112926088754e-05, + "loss": 0.3347, + "step": 32140 + }, + { + "epoch": 2.2303101797238223, + "grad_norm": 1.1471965312957764, + "learning_rate": 8.625556394725823e-05, + "loss": 0.3276, + "step": 32141 + }, + { + "epoch": 2.2303795711609187, + "grad_norm": 2.52280855178833, + "learning_rate": 8.624999867701853e-05, + "loss": 0.48, + "step": 32142 + }, + { + "epoch": 2.2304489625980155, + "grad_norm": 0.9764460921287537, + "learning_rate": 8.624443345018598e-05, + "loss": 0.354, + "step": 32143 + }, + { + "epoch": 2.230518354035112, + "grad_norm": 1.0405960083007812, + "learning_rate": 8.62388682667782e-05, + "loss": 0.3464, + "step": 32144 + }, + { + "epoch": 2.2305877454722087, + "grad_norm": 1.397258996963501, + "learning_rate": 8.623330312681269e-05, + "loss": 0.5291, + "step": 32145 + }, + { + "epoch": 2.2306571369093056, + "grad_norm": 1.598101258277893, + "learning_rate": 8.622773803030702e-05, + "loss": 0.4273, + "step": 32146 + }, + { + "epoch": 2.230726528346402, + "grad_norm": 1.2680439949035645, + "learning_rate": 8.62221729772788e-05, + "loss": 0.3366, + "step": 32147 + }, + { + "epoch": 2.230795919783499, + "grad_norm": 0.9260702729225159, + "learning_rate": 8.621660796774557e-05, + "loss": 0.3622, + "step": 32148 + }, + { + "epoch": 2.230865311220595, + "grad_norm": 1.1811898946762085, + "learning_rate": 8.621104300172491e-05, + "loss": 0.4977, + "step": 32149 + }, + { + "epoch": 2.230934702657692, + "grad_norm": 1.1171886920928955, + "learning_rate": 8.62054780792344e-05, + "loss": 0.2879, + "step": 32150 + }, + { + "epoch": 2.231004094094789, + "grad_norm": 1.0390499830245972, + "learning_rate": 8.619991320029162e-05, + "loss": 0.5595, + "step": 32151 + }, + { + "epoch": 2.2310734855318852, + "grad_norm": 1.7578306198120117, + "learning_rate": 8.619434836491409e-05, + "loss": 0.4607, + "step": 32152 + }, + { + "epoch": 2.231142876968982, + "grad_norm": 1.5326610803604126, + "learning_rate": 8.618878357311939e-05, + "loss": 0.549, + "step": 32153 + }, + { + "epoch": 2.231212268406079, + "grad_norm": 0.7093926668167114, + "learning_rate": 8.61832188249251e-05, + "loss": 0.2562, + "step": 32154 + }, + { + "epoch": 2.2312816598431753, + "grad_norm": 1.283862829208374, + "learning_rate": 8.617765412034879e-05, + "loss": 0.336, + "step": 32155 + }, + { + "epoch": 2.231351051280272, + "grad_norm": 0.879783570766449, + "learning_rate": 8.617208945940804e-05, + "loss": 0.2693, + "step": 32156 + }, + { + "epoch": 2.2314204427173685, + "grad_norm": 1.1992082595825195, + "learning_rate": 8.616652484212037e-05, + "loss": 0.2492, + "step": 32157 + }, + { + "epoch": 2.2314898341544653, + "grad_norm": 1.4596333503723145, + "learning_rate": 8.61609602685034e-05, + "loss": 0.4387, + "step": 32158 + }, + { + "epoch": 2.231559225591562, + "grad_norm": 1.468598484992981, + "learning_rate": 8.615539573857465e-05, + "loss": 0.6072, + "step": 32159 + }, + { + "epoch": 2.2316286170286586, + "grad_norm": 1.3049778938293457, + "learning_rate": 8.614983125235175e-05, + "loss": 0.3043, + "step": 32160 + }, + { + "epoch": 2.2316980084657554, + "grad_norm": 1.30753493309021, + "learning_rate": 8.614426680985219e-05, + "loss": 0.4203, + "step": 32161 + }, + { + "epoch": 2.231767399902852, + "grad_norm": 0.9777091145515442, + "learning_rate": 8.613870241109356e-05, + "loss": 0.5067, + "step": 32162 + }, + { + "epoch": 2.2318367913399486, + "grad_norm": 1.337465763092041, + "learning_rate": 8.613313805609345e-05, + "loss": 0.4585, + "step": 32163 + }, + { + "epoch": 2.2319061827770454, + "grad_norm": 1.1400116682052612, + "learning_rate": 8.612757374486941e-05, + "loss": 0.5163, + "step": 32164 + }, + { + "epoch": 2.231975574214142, + "grad_norm": 0.9524428844451904, + "learning_rate": 8.612200947743907e-05, + "loss": 0.3154, + "step": 32165 + }, + { + "epoch": 2.2320449656512387, + "grad_norm": 1.520910620689392, + "learning_rate": 8.611644525381987e-05, + "loss": 0.533, + "step": 32166 + }, + { + "epoch": 2.2321143570883355, + "grad_norm": 1.2758548259735107, + "learning_rate": 8.611088107402946e-05, + "loss": 0.442, + "step": 32167 + }, + { + "epoch": 2.232183748525432, + "grad_norm": 1.0959941148757935, + "learning_rate": 8.610531693808538e-05, + "loss": 0.4035, + "step": 32168 + }, + { + "epoch": 2.2322531399625287, + "grad_norm": 1.4904460906982422, + "learning_rate": 8.60997528460052e-05, + "loss": 0.4081, + "step": 32169 + }, + { + "epoch": 2.232322531399625, + "grad_norm": 1.6940040588378906, + "learning_rate": 8.609418879780648e-05, + "loss": 0.3852, + "step": 32170 + }, + { + "epoch": 2.232391922836722, + "grad_norm": 1.0060224533081055, + "learning_rate": 8.608862479350681e-05, + "loss": 0.3585, + "step": 32171 + }, + { + "epoch": 2.2324613142738188, + "grad_norm": 1.312031626701355, + "learning_rate": 8.608306083312377e-05, + "loss": 0.508, + "step": 32172 + }, + { + "epoch": 2.232530705710915, + "grad_norm": 1.2342829704284668, + "learning_rate": 8.607749691667485e-05, + "loss": 0.2779, + "step": 32173 + }, + { + "epoch": 2.232600097148012, + "grad_norm": 1.9498323202133179, + "learning_rate": 8.607193304417767e-05, + "loss": 0.3979, + "step": 32174 + }, + { + "epoch": 2.232669488585109, + "grad_norm": 1.3047142028808594, + "learning_rate": 8.606636921564978e-05, + "loss": 0.3827, + "step": 32175 + }, + { + "epoch": 2.232738880022205, + "grad_norm": 1.639373540878296, + "learning_rate": 8.606080543110874e-05, + "loss": 0.4701, + "step": 32176 + }, + { + "epoch": 2.232808271459302, + "grad_norm": 1.4228242635726929, + "learning_rate": 8.605524169057211e-05, + "loss": 0.5803, + "step": 32177 + }, + { + "epoch": 2.2328776628963984, + "grad_norm": 1.0966895818710327, + "learning_rate": 8.604967799405749e-05, + "loss": 0.3647, + "step": 32178 + }, + { + "epoch": 2.2329470543334953, + "grad_norm": 1.4031591415405273, + "learning_rate": 8.604411434158243e-05, + "loss": 0.4705, + "step": 32179 + }, + { + "epoch": 2.233016445770592, + "grad_norm": 1.3856393098831177, + "learning_rate": 8.603855073316444e-05, + "loss": 0.5112, + "step": 32180 + }, + { + "epoch": 2.2330858372076885, + "grad_norm": 1.220651388168335, + "learning_rate": 8.603298716882114e-05, + "loss": 0.4728, + "step": 32181 + }, + { + "epoch": 2.2331552286447853, + "grad_norm": 1.634456992149353, + "learning_rate": 8.602742364857008e-05, + "loss": 0.5377, + "step": 32182 + }, + { + "epoch": 2.2332246200818817, + "grad_norm": 1.9620898962020874, + "learning_rate": 8.602186017242882e-05, + "loss": 0.4096, + "step": 32183 + }, + { + "epoch": 2.2332940115189786, + "grad_norm": 1.6083804368972778, + "learning_rate": 8.601629674041491e-05, + "loss": 0.5658, + "step": 32184 + }, + { + "epoch": 2.2333634029560754, + "grad_norm": 0.9572051167488098, + "learning_rate": 8.601073335254599e-05, + "loss": 0.3823, + "step": 32185 + }, + { + "epoch": 2.2334327943931718, + "grad_norm": 1.6668343544006348, + "learning_rate": 8.600517000883951e-05, + "loss": 0.4483, + "step": 32186 + }, + { + "epoch": 2.2335021858302686, + "grad_norm": 1.5409859418869019, + "learning_rate": 8.599960670931308e-05, + "loss": 0.3028, + "step": 32187 + }, + { + "epoch": 2.233571577267365, + "grad_norm": 1.4084686040878296, + "learning_rate": 8.599404345398427e-05, + "loss": 0.5539, + "step": 32188 + }, + { + "epoch": 2.233640968704462, + "grad_norm": 0.9640219211578369, + "learning_rate": 8.598848024287065e-05, + "loss": 0.3175, + "step": 32189 + }, + { + "epoch": 2.2337103601415587, + "grad_norm": 1.0800445079803467, + "learning_rate": 8.598291707598978e-05, + "loss": 0.3733, + "step": 32190 + }, + { + "epoch": 2.233779751578655, + "grad_norm": 1.3401254415512085, + "learning_rate": 8.597735395335922e-05, + "loss": 0.4926, + "step": 32191 + }, + { + "epoch": 2.233849143015752, + "grad_norm": 1.1147814989089966, + "learning_rate": 8.597179087499652e-05, + "loss": 0.3951, + "step": 32192 + }, + { + "epoch": 2.2339185344528487, + "grad_norm": 1.1123517751693726, + "learning_rate": 8.596622784091924e-05, + "loss": 0.3195, + "step": 32193 + }, + { + "epoch": 2.233987925889945, + "grad_norm": 1.5543051958084106, + "learning_rate": 8.596066485114497e-05, + "loss": 0.5438, + "step": 32194 + }, + { + "epoch": 2.234057317327042, + "grad_norm": 1.5253695249557495, + "learning_rate": 8.595510190569123e-05, + "loss": 0.3687, + "step": 32195 + }, + { + "epoch": 2.2341267087641383, + "grad_norm": 2.5285227298736572, + "learning_rate": 8.594953900457559e-05, + "loss": 0.3646, + "step": 32196 + }, + { + "epoch": 2.234196100201235, + "grad_norm": 1.9130581617355347, + "learning_rate": 8.594397614781566e-05, + "loss": 0.4837, + "step": 32197 + }, + { + "epoch": 2.234265491638332, + "grad_norm": 1.2483474016189575, + "learning_rate": 8.593841333542893e-05, + "loss": 0.4546, + "step": 32198 + }, + { + "epoch": 2.2343348830754284, + "grad_norm": 2.993309736251831, + "learning_rate": 8.593285056743306e-05, + "loss": 0.42, + "step": 32199 + }, + { + "epoch": 2.234404274512525, + "grad_norm": 1.0721477270126343, + "learning_rate": 8.59272878438455e-05, + "loss": 0.3235, + "step": 32200 + }, + { + "epoch": 2.234473665949622, + "grad_norm": 1.6179393529891968, + "learning_rate": 8.592172516468387e-05, + "loss": 0.2796, + "step": 32201 + }, + { + "epoch": 2.2345430573867184, + "grad_norm": 1.3917129039764404, + "learning_rate": 8.59161625299657e-05, + "loss": 0.3767, + "step": 32202 + }, + { + "epoch": 2.2346124488238153, + "grad_norm": 1.351820468902588, + "learning_rate": 8.591059993970858e-05, + "loss": 0.346, + "step": 32203 + }, + { + "epoch": 2.2346818402609117, + "grad_norm": 1.8897978067398071, + "learning_rate": 8.590503739393006e-05, + "loss": 0.5061, + "step": 32204 + }, + { + "epoch": 2.2347512316980085, + "grad_norm": 1.495498538017273, + "learning_rate": 8.589947489264772e-05, + "loss": 0.444, + "step": 32205 + }, + { + "epoch": 2.2348206231351053, + "grad_norm": 1.7193851470947266, + "learning_rate": 8.589391243587912e-05, + "loss": 0.5407, + "step": 32206 + }, + { + "epoch": 2.2348900145722017, + "grad_norm": 1.051241159439087, + "learning_rate": 8.588835002364177e-05, + "loss": 0.4555, + "step": 32207 + }, + { + "epoch": 2.2349594060092985, + "grad_norm": 1.723585605621338, + "learning_rate": 8.588278765595325e-05, + "loss": 0.3996, + "step": 32208 + }, + { + "epoch": 2.235028797446395, + "grad_norm": 1.5361316204071045, + "learning_rate": 8.587722533283115e-05, + "loss": 0.6363, + "step": 32209 + }, + { + "epoch": 2.2350981888834918, + "grad_norm": 1.7716948986053467, + "learning_rate": 8.587166305429302e-05, + "loss": 0.3783, + "step": 32210 + }, + { + "epoch": 2.2351675803205886, + "grad_norm": 1.734426498413086, + "learning_rate": 8.58661008203564e-05, + "loss": 0.3561, + "step": 32211 + }, + { + "epoch": 2.235236971757685, + "grad_norm": 1.5441018342971802, + "learning_rate": 8.586053863103884e-05, + "loss": 0.5378, + "step": 32212 + }, + { + "epoch": 2.235306363194782, + "grad_norm": 1.093727707862854, + "learning_rate": 8.585497648635797e-05, + "loss": 0.4338, + "step": 32213 + }, + { + "epoch": 2.235375754631878, + "grad_norm": 1.237977385520935, + "learning_rate": 8.584941438633128e-05, + "loss": 0.4341, + "step": 32214 + }, + { + "epoch": 2.235445146068975, + "grad_norm": 1.5584160089492798, + "learning_rate": 8.584385233097632e-05, + "loss": 0.4601, + "step": 32215 + }, + { + "epoch": 2.235514537506072, + "grad_norm": 1.316354513168335, + "learning_rate": 8.58382903203107e-05, + "loss": 0.3669, + "step": 32216 + }, + { + "epoch": 2.2355839289431683, + "grad_norm": 1.478775978088379, + "learning_rate": 8.583272835435192e-05, + "loss": 0.3954, + "step": 32217 + }, + { + "epoch": 2.235653320380265, + "grad_norm": 1.1696922779083252, + "learning_rate": 8.582716643311759e-05, + "loss": 0.5577, + "step": 32218 + }, + { + "epoch": 2.235722711817362, + "grad_norm": 1.6881518363952637, + "learning_rate": 8.582160455662525e-05, + "loss": 0.5914, + "step": 32219 + }, + { + "epoch": 2.2357921032544583, + "grad_norm": 1.2551954984664917, + "learning_rate": 8.58160427248925e-05, + "loss": 0.3966, + "step": 32220 + }, + { + "epoch": 2.235861494691555, + "grad_norm": 1.439070701599121, + "learning_rate": 8.581048093793682e-05, + "loss": 0.4732, + "step": 32221 + }, + { + "epoch": 2.2359308861286515, + "grad_norm": 0.8800984025001526, + "learning_rate": 8.580491919577582e-05, + "loss": 0.2639, + "step": 32222 + }, + { + "epoch": 2.2360002775657484, + "grad_norm": 1.1426877975463867, + "learning_rate": 8.579935749842702e-05, + "loss": 0.4255, + "step": 32223 + }, + { + "epoch": 2.236069669002845, + "grad_norm": 1.4561944007873535, + "learning_rate": 8.579379584590801e-05, + "loss": 0.4881, + "step": 32224 + }, + { + "epoch": 2.2361390604399416, + "grad_norm": 1.7369204759597778, + "learning_rate": 8.578823423823634e-05, + "loss": 0.5693, + "step": 32225 + }, + { + "epoch": 2.2362084518770384, + "grad_norm": 1.2509225606918335, + "learning_rate": 8.578267267542956e-05, + "loss": 0.4118, + "step": 32226 + }, + { + "epoch": 2.2362778433141353, + "grad_norm": 1.105512022972107, + "learning_rate": 8.577711115750527e-05, + "loss": 0.3829, + "step": 32227 + }, + { + "epoch": 2.2363472347512316, + "grad_norm": 1.1401907205581665, + "learning_rate": 8.577154968448096e-05, + "loss": 0.3329, + "step": 32228 + }, + { + "epoch": 2.2364166261883285, + "grad_norm": 1.2490241527557373, + "learning_rate": 8.576598825637423e-05, + "loss": 0.4125, + "step": 32229 + }, + { + "epoch": 2.236486017625425, + "grad_norm": 1.2391356229782104, + "learning_rate": 8.57604268732026e-05, + "loss": 0.292, + "step": 32230 + }, + { + "epoch": 2.2365554090625217, + "grad_norm": 1.4848361015319824, + "learning_rate": 8.575486553498366e-05, + "loss": 0.5392, + "step": 32231 + }, + { + "epoch": 2.2366248004996185, + "grad_norm": 1.4278465509414673, + "learning_rate": 8.574930424173495e-05, + "loss": 0.4283, + "step": 32232 + }, + { + "epoch": 2.236694191936715, + "grad_norm": 1.3547232151031494, + "learning_rate": 8.574374299347403e-05, + "loss": 0.5724, + "step": 32233 + }, + { + "epoch": 2.2367635833738118, + "grad_norm": 1.2673547267913818, + "learning_rate": 8.573818179021851e-05, + "loss": 0.3562, + "step": 32234 + }, + { + "epoch": 2.236832974810908, + "grad_norm": 1.3104172945022583, + "learning_rate": 8.573262063198585e-05, + "loss": 0.4295, + "step": 32235 + }, + { + "epoch": 2.236902366248005, + "grad_norm": 1.064680576324463, + "learning_rate": 8.572705951879365e-05, + "loss": 0.3531, + "step": 32236 + }, + { + "epoch": 2.236971757685102, + "grad_norm": 1.2111775875091553, + "learning_rate": 8.572149845065947e-05, + "loss": 0.349, + "step": 32237 + }, + { + "epoch": 2.237041149122198, + "grad_norm": 1.338868498802185, + "learning_rate": 8.571593742760085e-05, + "loss": 0.4437, + "step": 32238 + }, + { + "epoch": 2.237110540559295, + "grad_norm": 1.669830322265625, + "learning_rate": 8.571037644963537e-05, + "loss": 0.4317, + "step": 32239 + }, + { + "epoch": 2.2371799319963914, + "grad_norm": 1.0321511030197144, + "learning_rate": 8.570481551678058e-05, + "loss": 0.3055, + "step": 32240 + }, + { + "epoch": 2.2372493234334883, + "grad_norm": 0.94310462474823, + "learning_rate": 8.569925462905406e-05, + "loss": 0.3344, + "step": 32241 + }, + { + "epoch": 2.237318714870585, + "grad_norm": 1.6502233743667603, + "learning_rate": 8.56936937864733e-05, + "loss": 0.5411, + "step": 32242 + }, + { + "epoch": 2.2373881063076815, + "grad_norm": 1.3977020978927612, + "learning_rate": 8.56881329890559e-05, + "loss": 0.3964, + "step": 32243 + }, + { + "epoch": 2.2374574977447783, + "grad_norm": 1.7580996751785278, + "learning_rate": 8.568257223681937e-05, + "loss": 0.402, + "step": 32244 + }, + { + "epoch": 2.237526889181875, + "grad_norm": 1.2760505676269531, + "learning_rate": 8.567701152978133e-05, + "loss": 0.3498, + "step": 32245 + }, + { + "epoch": 2.2375962806189715, + "grad_norm": 1.2591943740844727, + "learning_rate": 8.56714508679593e-05, + "loss": 0.3217, + "step": 32246 + }, + { + "epoch": 2.2376656720560684, + "grad_norm": 1.1159294843673706, + "learning_rate": 8.566589025137083e-05, + "loss": 0.5346, + "step": 32247 + }, + { + "epoch": 2.2377350634931648, + "grad_norm": 1.3747148513793945, + "learning_rate": 8.56603296800335e-05, + "loss": 0.583, + "step": 32248 + }, + { + "epoch": 2.2378044549302616, + "grad_norm": 1.362197756767273, + "learning_rate": 8.565476915396485e-05, + "loss": 0.322, + "step": 32249 + }, + { + "epoch": 2.2378738463673584, + "grad_norm": 1.3836709260940552, + "learning_rate": 8.56492086731824e-05, + "loss": 0.4832, + "step": 32250 + }, + { + "epoch": 2.237943237804455, + "grad_norm": 1.2804088592529297, + "learning_rate": 8.564364823770373e-05, + "loss": 0.4006, + "step": 32251 + }, + { + "epoch": 2.2380126292415516, + "grad_norm": 0.7895991206169128, + "learning_rate": 8.56380878475464e-05, + "loss": 0.274, + "step": 32252 + }, + { + "epoch": 2.2380820206786485, + "grad_norm": 0.9207952618598938, + "learning_rate": 8.563252750272796e-05, + "loss": 0.3259, + "step": 32253 + }, + { + "epoch": 2.238151412115745, + "grad_norm": 2.6216044425964355, + "learning_rate": 8.562696720326601e-05, + "loss": 0.7525, + "step": 32254 + }, + { + "epoch": 2.2382208035528417, + "grad_norm": 1.0511566400527954, + "learning_rate": 8.562140694917802e-05, + "loss": 0.2704, + "step": 32255 + }, + { + "epoch": 2.238290194989938, + "grad_norm": 1.4793977737426758, + "learning_rate": 8.561584674048158e-05, + "loss": 0.3913, + "step": 32256 + }, + { + "epoch": 2.238359586427035, + "grad_norm": 1.7661360502243042, + "learning_rate": 8.561028657719423e-05, + "loss": 0.6404, + "step": 32257 + }, + { + "epoch": 2.2384289778641318, + "grad_norm": 1.506975769996643, + "learning_rate": 8.560472645933355e-05, + "loss": 0.4721, + "step": 32258 + }, + { + "epoch": 2.238498369301228, + "grad_norm": 1.1818199157714844, + "learning_rate": 8.559916638691707e-05, + "loss": 0.3414, + "step": 32259 + }, + { + "epoch": 2.238567760738325, + "grad_norm": 1.4895668029785156, + "learning_rate": 8.559360635996234e-05, + "loss": 0.3418, + "step": 32260 + }, + { + "epoch": 2.2386371521754214, + "grad_norm": 1.2291311025619507, + "learning_rate": 8.558804637848698e-05, + "loss": 0.475, + "step": 32261 + }, + { + "epoch": 2.238706543612518, + "grad_norm": 1.3111611604690552, + "learning_rate": 8.558248644250845e-05, + "loss": 0.434, + "step": 32262 + }, + { + "epoch": 2.238775935049615, + "grad_norm": 1.2387208938598633, + "learning_rate": 8.557692655204433e-05, + "loss": 0.4526, + "step": 32263 + }, + { + "epoch": 2.2388453264867114, + "grad_norm": 1.367783546447754, + "learning_rate": 8.55713667071122e-05, + "loss": 0.678, + "step": 32264 + }, + { + "epoch": 2.2389147179238083, + "grad_norm": 1.3794257640838623, + "learning_rate": 8.556580690772956e-05, + "loss": 0.3647, + "step": 32265 + }, + { + "epoch": 2.2389841093609046, + "grad_norm": 1.6733585596084595, + "learning_rate": 8.556024715391401e-05, + "loss": 0.5276, + "step": 32266 + }, + { + "epoch": 2.2390535007980015, + "grad_norm": 1.6463615894317627, + "learning_rate": 8.555468744568306e-05, + "loss": 0.5792, + "step": 32267 + }, + { + "epoch": 2.2391228922350983, + "grad_norm": 1.6541297435760498, + "learning_rate": 8.554912778305436e-05, + "loss": 0.5185, + "step": 32268 + }, + { + "epoch": 2.2391922836721947, + "grad_norm": 1.3530620336532593, + "learning_rate": 8.554356816604533e-05, + "loss": 0.518, + "step": 32269 + }, + { + "epoch": 2.2392616751092915, + "grad_norm": 1.0350843667984009, + "learning_rate": 8.553800859467358e-05, + "loss": 0.4194, + "step": 32270 + }, + { + "epoch": 2.2393310665463884, + "grad_norm": 1.3074196577072144, + "learning_rate": 8.553244906895666e-05, + "loss": 0.3871, + "step": 32271 + }, + { + "epoch": 2.2394004579834847, + "grad_norm": 1.7764018774032593, + "learning_rate": 8.55268895889121e-05, + "loss": 0.5673, + "step": 32272 + }, + { + "epoch": 2.2394698494205816, + "grad_norm": 1.2135778665542603, + "learning_rate": 8.55213301545575e-05, + "loss": 0.3009, + "step": 32273 + }, + { + "epoch": 2.239539240857678, + "grad_norm": 1.158173680305481, + "learning_rate": 8.551577076591036e-05, + "loss": 0.4743, + "step": 32274 + }, + { + "epoch": 2.239608632294775, + "grad_norm": 1.3161728382110596, + "learning_rate": 8.55102114229883e-05, + "loss": 0.4064, + "step": 32275 + }, + { + "epoch": 2.2396780237318716, + "grad_norm": 1.2572007179260254, + "learning_rate": 8.550465212580878e-05, + "loss": 0.4569, + "step": 32276 + }, + { + "epoch": 2.239747415168968, + "grad_norm": 1.1987957954406738, + "learning_rate": 8.54990928743894e-05, + "loss": 0.3289, + "step": 32277 + }, + { + "epoch": 2.239816806606065, + "grad_norm": 1.7123453617095947, + "learning_rate": 8.549353366874768e-05, + "loss": 0.4578, + "step": 32278 + }, + { + "epoch": 2.2398861980431617, + "grad_norm": 2.1419239044189453, + "learning_rate": 8.548797450890122e-05, + "loss": 0.3598, + "step": 32279 + }, + { + "epoch": 2.239955589480258, + "grad_norm": 1.2661123275756836, + "learning_rate": 8.548241539486754e-05, + "loss": 0.4238, + "step": 32280 + }, + { + "epoch": 2.240024980917355, + "grad_norm": 0.9799096584320068, + "learning_rate": 8.547685632666417e-05, + "loss": 0.3382, + "step": 32281 + }, + { + "epoch": 2.2400943723544513, + "grad_norm": 1.1168301105499268, + "learning_rate": 8.54712973043087e-05, + "loss": 0.3128, + "step": 32282 + }, + { + "epoch": 2.240163763791548, + "grad_norm": 1.1678459644317627, + "learning_rate": 8.546573832781868e-05, + "loss": 0.3127, + "step": 32283 + }, + { + "epoch": 2.240233155228645, + "grad_norm": 1.307983160018921, + "learning_rate": 8.546017939721159e-05, + "loss": 0.5295, + "step": 32284 + }, + { + "epoch": 2.2403025466657414, + "grad_norm": 0.9539090394973755, + "learning_rate": 8.545462051250504e-05, + "loss": 0.3076, + "step": 32285 + }, + { + "epoch": 2.240371938102838, + "grad_norm": 1.277500867843628, + "learning_rate": 8.544906167371656e-05, + "loss": 0.5559, + "step": 32286 + }, + { + "epoch": 2.2404413295399346, + "grad_norm": 0.9279569387435913, + "learning_rate": 8.544350288086372e-05, + "loss": 0.4103, + "step": 32287 + }, + { + "epoch": 2.2405107209770314, + "grad_norm": 1.3619142770767212, + "learning_rate": 8.543794413396403e-05, + "loss": 0.5437, + "step": 32288 + }, + { + "epoch": 2.2405801124141282, + "grad_norm": 1.61642324924469, + "learning_rate": 8.543238543303512e-05, + "loss": 0.5198, + "step": 32289 + }, + { + "epoch": 2.2406495038512246, + "grad_norm": 1.4334357976913452, + "learning_rate": 8.542682677809442e-05, + "loss": 0.4188, + "step": 32290 + }, + { + "epoch": 2.2407188952883215, + "grad_norm": 1.615496039390564, + "learning_rate": 8.542126816915955e-05, + "loss": 0.5132, + "step": 32291 + }, + { + "epoch": 2.240788286725418, + "grad_norm": 1.3245412111282349, + "learning_rate": 8.541570960624804e-05, + "loss": 0.4253, + "step": 32292 + }, + { + "epoch": 2.2408576781625147, + "grad_norm": 1.4297844171524048, + "learning_rate": 8.541015108937744e-05, + "loss": 0.4557, + "step": 32293 + }, + { + "epoch": 2.2409270695996115, + "grad_norm": 1.8078218698501587, + "learning_rate": 8.540459261856532e-05, + "loss": 0.639, + "step": 32294 + }, + { + "epoch": 2.240996461036708, + "grad_norm": 0.8042577505111694, + "learning_rate": 8.53990341938292e-05, + "loss": 0.3247, + "step": 32295 + }, + { + "epoch": 2.2410658524738047, + "grad_norm": 1.5783885717391968, + "learning_rate": 8.539347581518665e-05, + "loss": 0.5788, + "step": 32296 + }, + { + "epoch": 2.2411352439109016, + "grad_norm": 0.8026924729347229, + "learning_rate": 8.538791748265519e-05, + "loss": 0.2374, + "step": 32297 + }, + { + "epoch": 2.241204635347998, + "grad_norm": 1.2566053867340088, + "learning_rate": 8.538235919625236e-05, + "loss": 0.2709, + "step": 32298 + }, + { + "epoch": 2.241274026785095, + "grad_norm": 1.6801402568817139, + "learning_rate": 8.537680095599576e-05, + "loss": 0.5055, + "step": 32299 + }, + { + "epoch": 2.241343418222191, + "grad_norm": 1.7346200942993164, + "learning_rate": 8.537124276190289e-05, + "loss": 0.4215, + "step": 32300 + }, + { + "epoch": 2.241412809659288, + "grad_norm": 1.3592511415481567, + "learning_rate": 8.536568461399128e-05, + "loss": 0.4175, + "step": 32301 + }, + { + "epoch": 2.241482201096385, + "grad_norm": 1.0038869380950928, + "learning_rate": 8.536012651227853e-05, + "loss": 0.3301, + "step": 32302 + }, + { + "epoch": 2.2415515925334812, + "grad_norm": 0.9628776907920837, + "learning_rate": 8.53545684567822e-05, + "loss": 0.2282, + "step": 32303 + }, + { + "epoch": 2.241620983970578, + "grad_norm": 1.4024256467819214, + "learning_rate": 8.534901044751975e-05, + "loss": 0.4924, + "step": 32304 + }, + { + "epoch": 2.241690375407675, + "grad_norm": 1.1034810543060303, + "learning_rate": 8.534345248450878e-05, + "loss": 0.4146, + "step": 32305 + }, + { + "epoch": 2.2417597668447713, + "grad_norm": 1.7451149225234985, + "learning_rate": 8.533789456776681e-05, + "loss": 0.4809, + "step": 32306 + }, + { + "epoch": 2.241829158281868, + "grad_norm": 1.2176026105880737, + "learning_rate": 8.533233669731143e-05, + "loss": 0.3837, + "step": 32307 + }, + { + "epoch": 2.2418985497189645, + "grad_norm": 1.0021991729736328, + "learning_rate": 8.532677887316016e-05, + "loss": 0.4597, + "step": 32308 + }, + { + "epoch": 2.2419679411560614, + "grad_norm": 1.2517508268356323, + "learning_rate": 8.532122109533054e-05, + "loss": 0.4215, + "step": 32309 + }, + { + "epoch": 2.242037332593158, + "grad_norm": 1.2119697332382202, + "learning_rate": 8.531566336384017e-05, + "loss": 0.5511, + "step": 32310 + }, + { + "epoch": 2.2421067240302546, + "grad_norm": 1.2262656688690186, + "learning_rate": 8.53101056787065e-05, + "loss": 0.4684, + "step": 32311 + }, + { + "epoch": 2.2421761154673514, + "grad_norm": 1.3164201974868774, + "learning_rate": 8.530454803994712e-05, + "loss": 0.4067, + "step": 32312 + }, + { + "epoch": 2.242245506904448, + "grad_norm": 1.2180718183517456, + "learning_rate": 8.529899044757958e-05, + "loss": 0.3033, + "step": 32313 + }, + { + "epoch": 2.2423148983415446, + "grad_norm": 1.5471343994140625, + "learning_rate": 8.529343290162142e-05, + "loss": 0.4654, + "step": 32314 + }, + { + "epoch": 2.2423842897786415, + "grad_norm": 1.6868646144866943, + "learning_rate": 8.52878754020902e-05, + "loss": 0.4088, + "step": 32315 + }, + { + "epoch": 2.242453681215738, + "grad_norm": 1.1584609746932983, + "learning_rate": 8.528231794900344e-05, + "loss": 0.315, + "step": 32316 + }, + { + "epoch": 2.2425230726528347, + "grad_norm": 1.0368708372116089, + "learning_rate": 8.527676054237872e-05, + "loss": 0.4019, + "step": 32317 + }, + { + "epoch": 2.242592464089931, + "grad_norm": 1.2835047245025635, + "learning_rate": 8.527120318223355e-05, + "loss": 0.3341, + "step": 32318 + }, + { + "epoch": 2.242661855527028, + "grad_norm": 1.6301552057266235, + "learning_rate": 8.526564586858547e-05, + "loss": 0.581, + "step": 32319 + }, + { + "epoch": 2.2427312469641247, + "grad_norm": 1.2826229333877563, + "learning_rate": 8.526008860145203e-05, + "loss": 0.3606, + "step": 32320 + }, + { + "epoch": 2.242800638401221, + "grad_norm": 1.2417173385620117, + "learning_rate": 8.525453138085076e-05, + "loss": 0.4834, + "step": 32321 + }, + { + "epoch": 2.242870029838318, + "grad_norm": 1.1593271493911743, + "learning_rate": 8.524897420679925e-05, + "loss": 0.3163, + "step": 32322 + }, + { + "epoch": 2.242939421275415, + "grad_norm": 1.3473306894302368, + "learning_rate": 8.524341707931501e-05, + "loss": 0.4163, + "step": 32323 + }, + { + "epoch": 2.243008812712511, + "grad_norm": 1.2027281522750854, + "learning_rate": 8.523785999841563e-05, + "loss": 0.6101, + "step": 32324 + }, + { + "epoch": 2.243078204149608, + "grad_norm": 1.2546970844268799, + "learning_rate": 8.523230296411858e-05, + "loss": 0.4178, + "step": 32325 + }, + { + "epoch": 2.2431475955867044, + "grad_norm": 1.4423832893371582, + "learning_rate": 8.522674597644143e-05, + "loss": 0.4487, + "step": 32326 + }, + { + "epoch": 2.2432169870238012, + "grad_norm": 0.9663313627243042, + "learning_rate": 8.522118903540173e-05, + "loss": 0.3199, + "step": 32327 + }, + { + "epoch": 2.243286378460898, + "grad_norm": 1.544971227645874, + "learning_rate": 8.521563214101703e-05, + "loss": 0.6292, + "step": 32328 + }, + { + "epoch": 2.2433557698979945, + "grad_norm": 1.222280740737915, + "learning_rate": 8.521007529330485e-05, + "loss": 0.479, + "step": 32329 + }, + { + "epoch": 2.2434251613350913, + "grad_norm": 1.7552883625030518, + "learning_rate": 8.520451849228279e-05, + "loss": 0.536, + "step": 32330 + }, + { + "epoch": 2.243494552772188, + "grad_norm": 1.5638023614883423, + "learning_rate": 8.519896173796831e-05, + "loss": 0.6649, + "step": 32331 + }, + { + "epoch": 2.2435639442092845, + "grad_norm": 1.592309594154358, + "learning_rate": 8.519340503037898e-05, + "loss": 0.6919, + "step": 32332 + }, + { + "epoch": 2.2436333356463813, + "grad_norm": 1.0545557737350464, + "learning_rate": 8.518784836953238e-05, + "loss": 0.3147, + "step": 32333 + }, + { + "epoch": 2.2437027270834777, + "grad_norm": 1.2519761323928833, + "learning_rate": 8.518229175544601e-05, + "loss": 0.3321, + "step": 32334 + }, + { + "epoch": 2.2437721185205746, + "grad_norm": 1.5680038928985596, + "learning_rate": 8.517673518813742e-05, + "loss": 0.3603, + "step": 32335 + }, + { + "epoch": 2.2438415099576714, + "grad_norm": 1.0428433418273926, + "learning_rate": 8.517117866762417e-05, + "loss": 0.3664, + "step": 32336 + }, + { + "epoch": 2.243910901394768, + "grad_norm": 1.4966245889663696, + "learning_rate": 8.516562219392383e-05, + "loss": 0.544, + "step": 32337 + }, + { + "epoch": 2.2439802928318646, + "grad_norm": 0.9672608971595764, + "learning_rate": 8.516006576705385e-05, + "loss": 0.3704, + "step": 32338 + }, + { + "epoch": 2.244049684268961, + "grad_norm": 1.2660754919052124, + "learning_rate": 8.51545093870318e-05, + "loss": 0.4227, + "step": 32339 + }, + { + "epoch": 2.244119075706058, + "grad_norm": 0.8507369756698608, + "learning_rate": 8.514895305387527e-05, + "loss": 0.3327, + "step": 32340 + }, + { + "epoch": 2.2441884671431547, + "grad_norm": 1.1511764526367188, + "learning_rate": 8.514339676760176e-05, + "loss": 0.4691, + "step": 32341 + }, + { + "epoch": 2.244257858580251, + "grad_norm": 1.0891468524932861, + "learning_rate": 8.513784052822881e-05, + "loss": 0.4443, + "step": 32342 + }, + { + "epoch": 2.244327250017348, + "grad_norm": 1.1342251300811768, + "learning_rate": 8.5132284335774e-05, + "loss": 0.4643, + "step": 32343 + }, + { + "epoch": 2.2443966414544447, + "grad_norm": 1.7552833557128906, + "learning_rate": 8.512672819025486e-05, + "loss": 0.7013, + "step": 32344 + }, + { + "epoch": 2.244466032891541, + "grad_norm": 1.0201472043991089, + "learning_rate": 8.51211720916889e-05, + "loss": 0.3814, + "step": 32345 + }, + { + "epoch": 2.244535424328638, + "grad_norm": 1.4603043794631958, + "learning_rate": 8.511561604009365e-05, + "loss": 0.4012, + "step": 32346 + }, + { + "epoch": 2.2446048157657343, + "grad_norm": 1.2333674430847168, + "learning_rate": 8.511006003548668e-05, + "loss": 0.3258, + "step": 32347 + }, + { + "epoch": 2.244674207202831, + "grad_norm": 1.1186871528625488, + "learning_rate": 8.510450407788553e-05, + "loss": 0.4676, + "step": 32348 + }, + { + "epoch": 2.244743598639928, + "grad_norm": 1.0016955137252808, + "learning_rate": 8.509894816730775e-05, + "loss": 0.4039, + "step": 32349 + }, + { + "epoch": 2.2448129900770244, + "grad_norm": 1.293245553970337, + "learning_rate": 8.509339230377084e-05, + "loss": 0.5582, + "step": 32350 + }, + { + "epoch": 2.2448823815141212, + "grad_norm": 1.6289105415344238, + "learning_rate": 8.508783648729238e-05, + "loss": 0.3066, + "step": 32351 + }, + { + "epoch": 2.2449517729512176, + "grad_norm": 1.1469485759735107, + "learning_rate": 8.508228071788986e-05, + "loss": 0.4054, + "step": 32352 + }, + { + "epoch": 2.2450211643883144, + "grad_norm": 1.5095864534378052, + "learning_rate": 8.507672499558089e-05, + "loss": 0.3277, + "step": 32353 + }, + { + "epoch": 2.2450905558254113, + "grad_norm": 1.2104063034057617, + "learning_rate": 8.507116932038292e-05, + "loss": 0.345, + "step": 32354 + }, + { + "epoch": 2.2451599472625077, + "grad_norm": 1.24384343624115, + "learning_rate": 8.506561369231356e-05, + "loss": 0.4471, + "step": 32355 + }, + { + "epoch": 2.2452293386996045, + "grad_norm": 1.3780516386032104, + "learning_rate": 8.506005811139032e-05, + "loss": 0.4436, + "step": 32356 + }, + { + "epoch": 2.2452987301367013, + "grad_norm": 1.2757669687271118, + "learning_rate": 8.505450257763075e-05, + "loss": 0.2822, + "step": 32357 + }, + { + "epoch": 2.2453681215737977, + "grad_norm": 1.797967553138733, + "learning_rate": 8.504894709105241e-05, + "loss": 0.7098, + "step": 32358 + }, + { + "epoch": 2.2454375130108946, + "grad_norm": 1.1617710590362549, + "learning_rate": 8.504339165167277e-05, + "loss": 0.3902, + "step": 32359 + }, + { + "epoch": 2.245506904447991, + "grad_norm": 1.2597448825836182, + "learning_rate": 8.503783625950941e-05, + "loss": 0.4253, + "step": 32360 + }, + { + "epoch": 2.245576295885088, + "grad_norm": 1.8042876720428467, + "learning_rate": 8.503228091457987e-05, + "loss": 0.5327, + "step": 32361 + }, + { + "epoch": 2.2456456873221846, + "grad_norm": 0.9997574687004089, + "learning_rate": 8.502672561690167e-05, + "loss": 0.3426, + "step": 32362 + }, + { + "epoch": 2.245715078759281, + "grad_norm": 1.4354208707809448, + "learning_rate": 8.502117036649237e-05, + "loss": 0.7135, + "step": 32363 + }, + { + "epoch": 2.245784470196378, + "grad_norm": 1.6352909803390503, + "learning_rate": 8.501561516336949e-05, + "loss": 0.3899, + "step": 32364 + }, + { + "epoch": 2.2458538616334742, + "grad_norm": 0.8190707564353943, + "learning_rate": 8.501006000755062e-05, + "loss": 0.2659, + "step": 32365 + }, + { + "epoch": 2.245923253070571, + "grad_norm": 1.470528483390808, + "learning_rate": 8.50045048990532e-05, + "loss": 0.3805, + "step": 32366 + }, + { + "epoch": 2.245992644507668, + "grad_norm": 1.4735161066055298, + "learning_rate": 8.499894983789483e-05, + "loss": 0.4245, + "step": 32367 + }, + { + "epoch": 2.2460620359447643, + "grad_norm": 1.3351404666900635, + "learning_rate": 8.499339482409303e-05, + "loss": 0.5131, + "step": 32368 + }, + { + "epoch": 2.246131427381861, + "grad_norm": 1.6507842540740967, + "learning_rate": 8.498783985766538e-05, + "loss": 0.4392, + "step": 32369 + }, + { + "epoch": 2.246200818818958, + "grad_norm": 1.039793610572815, + "learning_rate": 8.498228493862933e-05, + "loss": 0.3075, + "step": 32370 + }, + { + "epoch": 2.2462702102560543, + "grad_norm": 1.0008453130722046, + "learning_rate": 8.497673006700248e-05, + "loss": 0.3297, + "step": 32371 + }, + { + "epoch": 2.246339601693151, + "grad_norm": 1.9326430559158325, + "learning_rate": 8.497117524280239e-05, + "loss": 0.3804, + "step": 32372 + }, + { + "epoch": 2.2464089931302476, + "grad_norm": 1.3925178050994873, + "learning_rate": 8.496562046604651e-05, + "loss": 0.2237, + "step": 32373 + }, + { + "epoch": 2.2464783845673444, + "grad_norm": 1.6662592887878418, + "learning_rate": 8.496006573675243e-05, + "loss": 0.6447, + "step": 32374 + }, + { + "epoch": 2.246547776004441, + "grad_norm": 1.5122675895690918, + "learning_rate": 8.495451105493767e-05, + "loss": 0.308, + "step": 32375 + }, + { + "epoch": 2.2466171674415376, + "grad_norm": 1.3546257019042969, + "learning_rate": 8.494895642061977e-05, + "loss": 0.5586, + "step": 32376 + }, + { + "epoch": 2.2466865588786344, + "grad_norm": 1.3482354879379272, + "learning_rate": 8.494340183381627e-05, + "loss": 0.4432, + "step": 32377 + }, + { + "epoch": 2.246755950315731, + "grad_norm": 1.2059192657470703, + "learning_rate": 8.493784729454471e-05, + "loss": 0.3761, + "step": 32378 + }, + { + "epoch": 2.2468253417528277, + "grad_norm": 1.3837047815322876, + "learning_rate": 8.493229280282266e-05, + "loss": 0.5491, + "step": 32379 + }, + { + "epoch": 2.2468947331899245, + "grad_norm": 2.673128604888916, + "learning_rate": 8.492673835866756e-05, + "loss": 0.4609, + "step": 32380 + }, + { + "epoch": 2.246964124627021, + "grad_norm": 1.426941156387329, + "learning_rate": 8.4921183962097e-05, + "loss": 0.4459, + "step": 32381 + }, + { + "epoch": 2.2470335160641177, + "grad_norm": 1.219262719154358, + "learning_rate": 8.491562961312853e-05, + "loss": 0.3848, + "step": 32382 + }, + { + "epoch": 2.2471029075012146, + "grad_norm": 0.890649139881134, + "learning_rate": 8.491007531177965e-05, + "loss": 0.3283, + "step": 32383 + }, + { + "epoch": 2.247172298938311, + "grad_norm": 1.0551443099975586, + "learning_rate": 8.490452105806795e-05, + "loss": 0.2857, + "step": 32384 + }, + { + "epoch": 2.2472416903754078, + "grad_norm": 1.5438156127929688, + "learning_rate": 8.489896685201089e-05, + "loss": 0.2774, + "step": 32385 + }, + { + "epoch": 2.247311081812504, + "grad_norm": 1.4295457601547241, + "learning_rate": 8.489341269362606e-05, + "loss": 0.4295, + "step": 32386 + }, + { + "epoch": 2.247380473249601, + "grad_norm": 1.156928300857544, + "learning_rate": 8.488785858293096e-05, + "loss": 0.4129, + "step": 32387 + }, + { + "epoch": 2.247449864686698, + "grad_norm": 1.9050697088241577, + "learning_rate": 8.488230451994316e-05, + "loss": 0.5777, + "step": 32388 + }, + { + "epoch": 2.247519256123794, + "grad_norm": 1.3357019424438477, + "learning_rate": 8.487675050468014e-05, + "loss": 0.3753, + "step": 32389 + }, + { + "epoch": 2.247588647560891, + "grad_norm": 1.5596708059310913, + "learning_rate": 8.487119653715947e-05, + "loss": 0.5247, + "step": 32390 + }, + { + "epoch": 2.2476580389979874, + "grad_norm": 1.6714279651641846, + "learning_rate": 8.486564261739867e-05, + "loss": 0.539, + "step": 32391 + }, + { + "epoch": 2.2477274304350843, + "grad_norm": 0.885022759437561, + "learning_rate": 8.486008874541528e-05, + "loss": 0.2866, + "step": 32392 + }, + { + "epoch": 2.247796821872181, + "grad_norm": 1.301269769668579, + "learning_rate": 8.485453492122689e-05, + "loss": 0.4966, + "step": 32393 + }, + { + "epoch": 2.2478662133092775, + "grad_norm": 1.2656654119491577, + "learning_rate": 8.484898114485094e-05, + "loss": 0.3612, + "step": 32394 + }, + { + "epoch": 2.2479356047463743, + "grad_norm": 1.3804031610488892, + "learning_rate": 8.484342741630497e-05, + "loss": 0.4916, + "step": 32395 + }, + { + "epoch": 2.248004996183471, + "grad_norm": 1.5944488048553467, + "learning_rate": 8.483787373560657e-05, + "loss": 0.4079, + "step": 32396 + }, + { + "epoch": 2.2480743876205675, + "grad_norm": 0.76606285572052, + "learning_rate": 8.483232010277322e-05, + "loss": 0.2965, + "step": 32397 + }, + { + "epoch": 2.2481437790576644, + "grad_norm": 1.3691366910934448, + "learning_rate": 8.482676651782249e-05, + "loss": 0.3188, + "step": 32398 + }, + { + "epoch": 2.2482131704947608, + "grad_norm": 1.2115365266799927, + "learning_rate": 8.482121298077187e-05, + "loss": 0.4171, + "step": 32399 + }, + { + "epoch": 2.2482825619318576, + "grad_norm": 1.193769097328186, + "learning_rate": 8.4815659491639e-05, + "loss": 0.2723, + "step": 32400 + }, + { + "epoch": 2.2483519533689544, + "grad_norm": 1.0363601446151733, + "learning_rate": 8.481010605044128e-05, + "loss": 0.4131, + "step": 32401 + }, + { + "epoch": 2.248421344806051, + "grad_norm": 1.7918304204940796, + "learning_rate": 8.480455265719628e-05, + "loss": 0.4307, + "step": 32402 + }, + { + "epoch": 2.2484907362431477, + "grad_norm": 1.0474495887756348, + "learning_rate": 8.479899931192154e-05, + "loss": 0.4292, + "step": 32403 + }, + { + "epoch": 2.2485601276802445, + "grad_norm": 1.3402725458145142, + "learning_rate": 8.479344601463463e-05, + "loss": 0.4879, + "step": 32404 + }, + { + "epoch": 2.248629519117341, + "grad_norm": 1.219623327255249, + "learning_rate": 8.478789276535303e-05, + "loss": 0.3438, + "step": 32405 + }, + { + "epoch": 2.2486989105544377, + "grad_norm": 1.0070987939834595, + "learning_rate": 8.478233956409427e-05, + "loss": 0.4552, + "step": 32406 + }, + { + "epoch": 2.248768301991534, + "grad_norm": 1.0667262077331543, + "learning_rate": 8.477678641087597e-05, + "loss": 0.3243, + "step": 32407 + }, + { + "epoch": 2.248837693428631, + "grad_norm": 1.083032488822937, + "learning_rate": 8.477123330571552e-05, + "loss": 0.417, + "step": 32408 + }, + { + "epoch": 2.2489070848657278, + "grad_norm": 1.08697509765625, + "learning_rate": 8.476568024863052e-05, + "loss": 0.4343, + "step": 32409 + }, + { + "epoch": 2.248976476302824, + "grad_norm": 1.9625760316848755, + "learning_rate": 8.476012723963851e-05, + "loss": 0.5255, + "step": 32410 + }, + { + "epoch": 2.249045867739921, + "grad_norm": 1.466522216796875, + "learning_rate": 8.4754574278757e-05, + "loss": 0.397, + "step": 32411 + }, + { + "epoch": 2.2491152591770174, + "grad_norm": 0.8386698365211487, + "learning_rate": 8.474902136600354e-05, + "loss": 0.2317, + "step": 32412 + }, + { + "epoch": 2.249184650614114, + "grad_norm": 1.1057566404342651, + "learning_rate": 8.474346850139569e-05, + "loss": 0.3706, + "step": 32413 + }, + { + "epoch": 2.249254042051211, + "grad_norm": 2.1409201622009277, + "learning_rate": 8.473791568495089e-05, + "loss": 0.4863, + "step": 32414 + }, + { + "epoch": 2.2493234334883074, + "grad_norm": 1.2911615371704102, + "learning_rate": 8.473236291668673e-05, + "loss": 0.4144, + "step": 32415 + }, + { + "epoch": 2.2493928249254043, + "grad_norm": 0.9569373726844788, + "learning_rate": 8.472681019662073e-05, + "loss": 0.3765, + "step": 32416 + }, + { + "epoch": 2.2494622163625007, + "grad_norm": 1.3656085729599, + "learning_rate": 8.472125752477042e-05, + "loss": 0.5062, + "step": 32417 + }, + { + "epoch": 2.2495316077995975, + "grad_norm": 1.3773081302642822, + "learning_rate": 8.471570490115331e-05, + "loss": 0.3173, + "step": 32418 + }, + { + "epoch": 2.2496009992366943, + "grad_norm": 1.676149845123291, + "learning_rate": 8.471015232578699e-05, + "loss": 0.6991, + "step": 32419 + }, + { + "epoch": 2.2496703906737907, + "grad_norm": 1.0341198444366455, + "learning_rate": 8.470459979868893e-05, + "loss": 0.3933, + "step": 32420 + }, + { + "epoch": 2.2497397821108875, + "grad_norm": 1.5382522344589233, + "learning_rate": 8.469904731987665e-05, + "loss": 0.359, + "step": 32421 + }, + { + "epoch": 2.2498091735479844, + "grad_norm": 1.382631540298462, + "learning_rate": 8.469349488936771e-05, + "loss": 0.5006, + "step": 32422 + }, + { + "epoch": 2.2498785649850808, + "grad_norm": 1.409906268119812, + "learning_rate": 8.468794250717965e-05, + "loss": 0.4583, + "step": 32423 + }, + { + "epoch": 2.2499479564221776, + "grad_norm": 1.2047574520111084, + "learning_rate": 8.468239017332996e-05, + "loss": 0.2929, + "step": 32424 + }, + { + "epoch": 2.250017347859274, + "grad_norm": 1.4201774597167969, + "learning_rate": 8.46768378878362e-05, + "loss": 0.4875, + "step": 32425 + }, + { + "epoch": 2.250086739296371, + "grad_norm": 1.0991542339324951, + "learning_rate": 8.467128565071585e-05, + "loss": 0.3988, + "step": 32426 + }, + { + "epoch": 2.2501561307334677, + "grad_norm": 1.585601568222046, + "learning_rate": 8.466573346198654e-05, + "loss": 0.5064, + "step": 32427 + }, + { + "epoch": 2.250225522170564, + "grad_norm": 1.132604718208313, + "learning_rate": 8.466018132166568e-05, + "loss": 0.2607, + "step": 32428 + }, + { + "epoch": 2.250294913607661, + "grad_norm": 1.4798038005828857, + "learning_rate": 8.465462922977086e-05, + "loss": 0.449, + "step": 32429 + }, + { + "epoch": 2.2503643050447577, + "grad_norm": 1.7371776103973389, + "learning_rate": 8.464907718631959e-05, + "loss": 0.4608, + "step": 32430 + }, + { + "epoch": 2.250433696481854, + "grad_norm": 1.2049258947372437, + "learning_rate": 8.46435251913294e-05, + "loss": 0.5834, + "step": 32431 + }, + { + "epoch": 2.250503087918951, + "grad_norm": 1.3695800304412842, + "learning_rate": 8.463797324481781e-05, + "loss": 0.3465, + "step": 32432 + }, + { + "epoch": 2.2505724793560473, + "grad_norm": 1.032819151878357, + "learning_rate": 8.463242134680238e-05, + "loss": 0.3396, + "step": 32433 + }, + { + "epoch": 2.250641870793144, + "grad_norm": 1.0308969020843506, + "learning_rate": 8.462686949730063e-05, + "loss": 0.2963, + "step": 32434 + }, + { + "epoch": 2.250711262230241, + "grad_norm": 1.0225269794464111, + "learning_rate": 8.462131769633003e-05, + "loss": 0.3343, + "step": 32435 + }, + { + "epoch": 2.2507806536673374, + "grad_norm": 1.2685458660125732, + "learning_rate": 8.461576594390815e-05, + "loss": 0.4721, + "step": 32436 + }, + { + "epoch": 2.250850045104434, + "grad_norm": 2.804431915283203, + "learning_rate": 8.461021424005252e-05, + "loss": 0.5491, + "step": 32437 + }, + { + "epoch": 2.2509194365415306, + "grad_norm": 1.3421719074249268, + "learning_rate": 8.460466258478068e-05, + "loss": 0.4236, + "step": 32438 + }, + { + "epoch": 2.2509888279786274, + "grad_norm": 1.3952192068099976, + "learning_rate": 8.459911097811009e-05, + "loss": 0.5555, + "step": 32439 + }, + { + "epoch": 2.2510582194157243, + "grad_norm": 0.7693162560462952, + "learning_rate": 8.459355942005834e-05, + "loss": 0.2998, + "step": 32440 + }, + { + "epoch": 2.2511276108528206, + "grad_norm": 1.3480887413024902, + "learning_rate": 8.458800791064296e-05, + "loss": 0.5749, + "step": 32441 + }, + { + "epoch": 2.2511970022899175, + "grad_norm": 0.7900643944740295, + "learning_rate": 8.458245644988144e-05, + "loss": 0.2746, + "step": 32442 + }, + { + "epoch": 2.251266393727014, + "grad_norm": 1.4755704402923584, + "learning_rate": 8.45769050377913e-05, + "loss": 0.6272, + "step": 32443 + }, + { + "epoch": 2.2513357851641107, + "grad_norm": 1.4992139339447021, + "learning_rate": 8.457135367439008e-05, + "loss": 0.5589, + "step": 32444 + }, + { + "epoch": 2.2514051766012075, + "grad_norm": 1.1599721908569336, + "learning_rate": 8.456580235969531e-05, + "loss": 0.3167, + "step": 32445 + }, + { + "epoch": 2.251474568038304, + "grad_norm": 1.2250607013702393, + "learning_rate": 8.456025109372451e-05, + "loss": 0.4401, + "step": 32446 + }, + { + "epoch": 2.2515439594754008, + "grad_norm": 0.9114051461219788, + "learning_rate": 8.455469987649521e-05, + "loss": 0.4161, + "step": 32447 + }, + { + "epoch": 2.2516133509124976, + "grad_norm": 1.282382607460022, + "learning_rate": 8.454914870802496e-05, + "loss": 0.3618, + "step": 32448 + }, + { + "epoch": 2.251682742349594, + "grad_norm": 1.4560043811798096, + "learning_rate": 8.454359758833122e-05, + "loss": 0.4888, + "step": 32449 + }, + { + "epoch": 2.251752133786691, + "grad_norm": 1.5275299549102783, + "learning_rate": 8.453804651743154e-05, + "loss": 0.4434, + "step": 32450 + }, + { + "epoch": 2.251821525223787, + "grad_norm": 2.2051076889038086, + "learning_rate": 8.453249549534347e-05, + "loss": 0.4139, + "step": 32451 + }, + { + "epoch": 2.251890916660884, + "grad_norm": 1.7128291130065918, + "learning_rate": 8.452694452208451e-05, + "loss": 0.4579, + "step": 32452 + }, + { + "epoch": 2.251960308097981, + "grad_norm": 1.2494676113128662, + "learning_rate": 8.452139359767218e-05, + "loss": 0.4022, + "step": 32453 + }, + { + "epoch": 2.2520296995350773, + "grad_norm": 1.2616304159164429, + "learning_rate": 8.451584272212404e-05, + "loss": 0.4767, + "step": 32454 + }, + { + "epoch": 2.252099090972174, + "grad_norm": 1.4617218971252441, + "learning_rate": 8.451029189545761e-05, + "loss": 0.4734, + "step": 32455 + }, + { + "epoch": 2.252168482409271, + "grad_norm": 1.2948905229568481, + "learning_rate": 8.450474111769035e-05, + "loss": 0.472, + "step": 32456 + }, + { + "epoch": 2.2522378738463673, + "grad_norm": 0.941222071647644, + "learning_rate": 8.449919038883982e-05, + "loss": 0.3807, + "step": 32457 + }, + { + "epoch": 2.252307265283464, + "grad_norm": 1.0240827798843384, + "learning_rate": 8.449363970892359e-05, + "loss": 0.3268, + "step": 32458 + }, + { + "epoch": 2.2523766567205605, + "grad_norm": 1.4177370071411133, + "learning_rate": 8.44880890779591e-05, + "loss": 0.5135, + "step": 32459 + }, + { + "epoch": 2.2524460481576574, + "grad_norm": 1.0698647499084473, + "learning_rate": 8.448253849596392e-05, + "loss": 0.3484, + "step": 32460 + }, + { + "epoch": 2.252515439594754, + "grad_norm": 1.668414831161499, + "learning_rate": 8.447698796295555e-05, + "loss": 0.5273, + "step": 32461 + }, + { + "epoch": 2.2525848310318506, + "grad_norm": 1.4269640445709229, + "learning_rate": 8.44714374789516e-05, + "loss": 0.5902, + "step": 32462 + }, + { + "epoch": 2.2526542224689474, + "grad_norm": 1.7957814931869507, + "learning_rate": 8.446588704396946e-05, + "loss": 0.4272, + "step": 32463 + }, + { + "epoch": 2.252723613906044, + "grad_norm": 1.5244133472442627, + "learning_rate": 8.446033665802672e-05, + "loss": 0.5329, + "step": 32464 + }, + { + "epoch": 2.2527930053431406, + "grad_norm": 1.3370829820632935, + "learning_rate": 8.445478632114088e-05, + "loss": 0.4212, + "step": 32465 + }, + { + "epoch": 2.2528623967802375, + "grad_norm": 1.467078685760498, + "learning_rate": 8.44492360333295e-05, + "loss": 0.3335, + "step": 32466 + }, + { + "epoch": 2.252931788217334, + "grad_norm": 1.4374229907989502, + "learning_rate": 8.444368579461006e-05, + "loss": 0.4827, + "step": 32467 + }, + { + "epoch": 2.2530011796544307, + "grad_norm": 1.560712218284607, + "learning_rate": 8.44381356050001e-05, + "loss": 0.3002, + "step": 32468 + }, + { + "epoch": 2.253070571091527, + "grad_norm": 0.9821715950965881, + "learning_rate": 8.443258546451718e-05, + "loss": 0.3832, + "step": 32469 + }, + { + "epoch": 2.253139962528624, + "grad_norm": 1.8127007484436035, + "learning_rate": 8.442703537317875e-05, + "loss": 0.6263, + "step": 32470 + }, + { + "epoch": 2.2532093539657208, + "grad_norm": 1.2003010511398315, + "learning_rate": 8.442148533100235e-05, + "loss": 0.419, + "step": 32471 + }, + { + "epoch": 2.253278745402817, + "grad_norm": 1.3953921794891357, + "learning_rate": 8.441593533800552e-05, + "loss": 0.4617, + "step": 32472 + }, + { + "epoch": 2.253348136839914, + "grad_norm": 1.6696194410324097, + "learning_rate": 8.441038539420579e-05, + "loss": 0.4871, + "step": 32473 + }, + { + "epoch": 2.253417528277011, + "grad_norm": 1.0470952987670898, + "learning_rate": 8.440483549962066e-05, + "loss": 0.4312, + "step": 32474 + }, + { + "epoch": 2.253486919714107, + "grad_norm": 1.5700178146362305, + "learning_rate": 8.439928565426763e-05, + "loss": 0.5296, + "step": 32475 + }, + { + "epoch": 2.253556311151204, + "grad_norm": 0.9035137295722961, + "learning_rate": 8.43937358581643e-05, + "loss": 0.3799, + "step": 32476 + }, + { + "epoch": 2.2536257025883004, + "grad_norm": 1.0165183544158936, + "learning_rate": 8.438818611132811e-05, + "loss": 0.343, + "step": 32477 + }, + { + "epoch": 2.2536950940253972, + "grad_norm": 1.3890399932861328, + "learning_rate": 8.438263641377658e-05, + "loss": 0.4711, + "step": 32478 + }, + { + "epoch": 2.253764485462494, + "grad_norm": 1.613993763923645, + "learning_rate": 8.437708676552727e-05, + "loss": 0.5128, + "step": 32479 + }, + { + "epoch": 2.2538338768995905, + "grad_norm": 2.678462028503418, + "learning_rate": 8.437153716659767e-05, + "loss": 0.4397, + "step": 32480 + }, + { + "epoch": 2.2539032683366873, + "grad_norm": 1.6406978368759155, + "learning_rate": 8.436598761700533e-05, + "loss": 0.5617, + "step": 32481 + }, + { + "epoch": 2.253972659773784, + "grad_norm": 0.9364163279533386, + "learning_rate": 8.436043811676773e-05, + "loss": 0.3692, + "step": 32482 + }, + { + "epoch": 2.2540420512108805, + "grad_norm": 1.1964213848114014, + "learning_rate": 8.435488866590249e-05, + "loss": 0.5823, + "step": 32483 + }, + { + "epoch": 2.2541114426479774, + "grad_norm": 1.523164987564087, + "learning_rate": 8.434933926442699e-05, + "loss": 0.4795, + "step": 32484 + }, + { + "epoch": 2.2541808340850737, + "grad_norm": 1.072237491607666, + "learning_rate": 8.43437899123588e-05, + "loss": 0.3992, + "step": 32485 + }, + { + "epoch": 2.2542502255221706, + "grad_norm": 1.1838558912277222, + "learning_rate": 8.433824060971545e-05, + "loss": 0.3978, + "step": 32486 + }, + { + "epoch": 2.2543196169592674, + "grad_norm": 1.198533058166504, + "learning_rate": 8.433269135651447e-05, + "loss": 0.5147, + "step": 32487 + }, + { + "epoch": 2.254389008396364, + "grad_norm": 1.3660240173339844, + "learning_rate": 8.432714215277337e-05, + "loss": 0.4286, + "step": 32488 + }, + { + "epoch": 2.2544583998334606, + "grad_norm": 1.2434457540512085, + "learning_rate": 8.432159299850969e-05, + "loss": 0.4114, + "step": 32489 + }, + { + "epoch": 2.254527791270557, + "grad_norm": 1.4870967864990234, + "learning_rate": 8.43160438937409e-05, + "loss": 0.3542, + "step": 32490 + }, + { + "epoch": 2.254597182707654, + "grad_norm": 1.1416707038879395, + "learning_rate": 8.431049483848452e-05, + "loss": 0.3549, + "step": 32491 + }, + { + "epoch": 2.2546665741447507, + "grad_norm": 1.5717400312423706, + "learning_rate": 8.430494583275811e-05, + "loss": 0.5172, + "step": 32492 + }, + { + "epoch": 2.254735965581847, + "grad_norm": 1.5033254623413086, + "learning_rate": 8.429939687657915e-05, + "loss": 0.3511, + "step": 32493 + }, + { + "epoch": 2.254805357018944, + "grad_norm": 1.0744138956069946, + "learning_rate": 8.429384796996517e-05, + "loss": 0.3441, + "step": 32494 + }, + { + "epoch": 2.2548747484560403, + "grad_norm": 1.4693989753723145, + "learning_rate": 8.42882991129337e-05, + "loss": 0.5171, + "step": 32495 + }, + { + "epoch": 2.254944139893137, + "grad_norm": 1.6403456926345825, + "learning_rate": 8.428275030550228e-05, + "loss": 0.397, + "step": 32496 + }, + { + "epoch": 2.255013531330234, + "grad_norm": 1.4847543239593506, + "learning_rate": 8.427720154768835e-05, + "loss": 0.3961, + "step": 32497 + }, + { + "epoch": 2.2550829227673304, + "grad_norm": 1.1486562490463257, + "learning_rate": 8.427165283950947e-05, + "loss": 0.3873, + "step": 32498 + }, + { + "epoch": 2.255152314204427, + "grad_norm": 1.0982040166854858, + "learning_rate": 8.426610418098316e-05, + "loss": 0.5734, + "step": 32499 + }, + { + "epoch": 2.255221705641524, + "grad_norm": 1.6395145654678345, + "learning_rate": 8.426055557212693e-05, + "loss": 0.5909, + "step": 32500 + }, + { + "epoch": 2.2552910970786204, + "grad_norm": 1.419846773147583, + "learning_rate": 8.42550070129583e-05, + "loss": 0.3542, + "step": 32501 + }, + { + "epoch": 2.2553604885157172, + "grad_norm": 1.425093173980713, + "learning_rate": 8.424945850349479e-05, + "loss": 0.4629, + "step": 32502 + }, + { + "epoch": 2.2554298799528136, + "grad_norm": 1.3503905534744263, + "learning_rate": 8.424391004375396e-05, + "loss": 0.3505, + "step": 32503 + }, + { + "epoch": 2.2554992713899105, + "grad_norm": 1.1352744102478027, + "learning_rate": 8.423836163375323e-05, + "loss": 0.3504, + "step": 32504 + }, + { + "epoch": 2.2555686628270073, + "grad_norm": 0.9622699022293091, + "learning_rate": 8.423281327351016e-05, + "loss": 0.3788, + "step": 32505 + }, + { + "epoch": 2.2556380542641037, + "grad_norm": 1.3156906366348267, + "learning_rate": 8.422726496304228e-05, + "loss": 0.4508, + "step": 32506 + }, + { + "epoch": 2.2557074457012005, + "grad_norm": 1.0157548189163208, + "learning_rate": 8.422171670236707e-05, + "loss": 0.4511, + "step": 32507 + }, + { + "epoch": 2.2557768371382974, + "grad_norm": 2.024081230163574, + "learning_rate": 8.421616849150212e-05, + "loss": 0.251, + "step": 32508 + }, + { + "epoch": 2.2558462285753937, + "grad_norm": 1.285852313041687, + "learning_rate": 8.421062033046485e-05, + "loss": 0.3486, + "step": 32509 + }, + { + "epoch": 2.2559156200124906, + "grad_norm": 0.9134400486946106, + "learning_rate": 8.420507221927285e-05, + "loss": 0.303, + "step": 32510 + }, + { + "epoch": 2.255985011449587, + "grad_norm": 1.1001397371292114, + "learning_rate": 8.419952415794358e-05, + "loss": 0.3901, + "step": 32511 + }, + { + "epoch": 2.256054402886684, + "grad_norm": 1.4716839790344238, + "learning_rate": 8.419397614649462e-05, + "loss": 0.435, + "step": 32512 + }, + { + "epoch": 2.2561237943237806, + "grad_norm": 1.5296857357025146, + "learning_rate": 8.418842818494341e-05, + "loss": 0.4047, + "step": 32513 + }, + { + "epoch": 2.256193185760877, + "grad_norm": 1.3122671842575073, + "learning_rate": 8.41828802733075e-05, + "loss": 0.4351, + "step": 32514 + }, + { + "epoch": 2.256262577197974, + "grad_norm": 1.1182184219360352, + "learning_rate": 8.417733241160439e-05, + "loss": 0.4128, + "step": 32515 + }, + { + "epoch": 2.2563319686350702, + "grad_norm": 0.9606627821922302, + "learning_rate": 8.41717845998516e-05, + "loss": 0.2514, + "step": 32516 + }, + { + "epoch": 2.256401360072167, + "grad_norm": 1.2233059406280518, + "learning_rate": 8.416623683806672e-05, + "loss": 0.4076, + "step": 32517 + }, + { + "epoch": 2.256470751509264, + "grad_norm": 1.1517647504806519, + "learning_rate": 8.416068912626714e-05, + "loss": 0.3359, + "step": 32518 + }, + { + "epoch": 2.2565401429463603, + "grad_norm": 1.1555191278457642, + "learning_rate": 8.415514146447041e-05, + "loss": 0.3349, + "step": 32519 + }, + { + "epoch": 2.256609534383457, + "grad_norm": 1.442673921585083, + "learning_rate": 8.414959385269408e-05, + "loss": 0.6521, + "step": 32520 + }, + { + "epoch": 2.2566789258205535, + "grad_norm": 0.8925551772117615, + "learning_rate": 8.414404629095563e-05, + "loss": 0.3225, + "step": 32521 + }, + { + "epoch": 2.2567483172576503, + "grad_norm": 1.7829777002334595, + "learning_rate": 8.41384987792726e-05, + "loss": 0.4227, + "step": 32522 + }, + { + "epoch": 2.256817708694747, + "grad_norm": 1.2642608880996704, + "learning_rate": 8.413295131766248e-05, + "loss": 0.4547, + "step": 32523 + }, + { + "epoch": 2.2568871001318436, + "grad_norm": 1.5697200298309326, + "learning_rate": 8.412740390614283e-05, + "loss": 0.468, + "step": 32524 + }, + { + "epoch": 2.2569564915689404, + "grad_norm": 1.466611623764038, + "learning_rate": 8.412185654473109e-05, + "loss": 0.4939, + "step": 32525 + }, + { + "epoch": 2.2570258830060372, + "grad_norm": 1.5520763397216797, + "learning_rate": 8.41163092334448e-05, + "loss": 0.5202, + "step": 32526 + }, + { + "epoch": 2.2570952744431336, + "grad_norm": 1.472906231880188, + "learning_rate": 8.41107619723015e-05, + "loss": 0.5063, + "step": 32527 + }, + { + "epoch": 2.2571646658802305, + "grad_norm": 1.1910496950149536, + "learning_rate": 8.410521476131867e-05, + "loss": 0.4192, + "step": 32528 + }, + { + "epoch": 2.2572340573173273, + "grad_norm": 1.1689910888671875, + "learning_rate": 8.409966760051383e-05, + "loss": 0.3799, + "step": 32529 + }, + { + "epoch": 2.2573034487544237, + "grad_norm": 1.596096158027649, + "learning_rate": 8.40941204899045e-05, + "loss": 0.4157, + "step": 32530 + }, + { + "epoch": 2.2573728401915205, + "grad_norm": 1.3487110137939453, + "learning_rate": 8.408857342950821e-05, + "loss": 0.3848, + "step": 32531 + }, + { + "epoch": 2.257442231628617, + "grad_norm": 1.2387871742248535, + "learning_rate": 8.408302641934241e-05, + "loss": 0.3921, + "step": 32532 + }, + { + "epoch": 2.2575116230657137, + "grad_norm": 1.1212648153305054, + "learning_rate": 8.407747945942465e-05, + "loss": 0.3149, + "step": 32533 + }, + { + "epoch": 2.2575810145028106, + "grad_norm": 1.15044105052948, + "learning_rate": 8.407193254977246e-05, + "loss": 0.5002, + "step": 32534 + }, + { + "epoch": 2.257650405939907, + "grad_norm": 1.7000268697738647, + "learning_rate": 8.40663856904033e-05, + "loss": 0.6301, + "step": 32535 + }, + { + "epoch": 2.257719797377004, + "grad_norm": 1.1655081510543823, + "learning_rate": 8.406083888133474e-05, + "loss": 0.3279, + "step": 32536 + }, + { + "epoch": 2.2577891888141, + "grad_norm": 1.3089221715927124, + "learning_rate": 8.405529212258425e-05, + "loss": 0.3296, + "step": 32537 + }, + { + "epoch": 2.257858580251197, + "grad_norm": 1.387565016746521, + "learning_rate": 8.404974541416939e-05, + "loss": 0.7139, + "step": 32538 + }, + { + "epoch": 2.257927971688294, + "grad_norm": 1.3829642534255981, + "learning_rate": 8.40441987561076e-05, + "loss": 0.3941, + "step": 32539 + }, + { + "epoch": 2.2579973631253902, + "grad_norm": 1.2226203680038452, + "learning_rate": 8.403865214841642e-05, + "loss": 0.4403, + "step": 32540 + }, + { + "epoch": 2.258066754562487, + "grad_norm": 1.0502116680145264, + "learning_rate": 8.403310559111335e-05, + "loss": 0.2926, + "step": 32541 + }, + { + "epoch": 2.2581361459995835, + "grad_norm": 1.5140656232833862, + "learning_rate": 8.402755908421594e-05, + "loss": 0.5239, + "step": 32542 + }, + { + "epoch": 2.2582055374366803, + "grad_norm": 1.9053562879562378, + "learning_rate": 8.402201262774167e-05, + "loss": 0.5478, + "step": 32543 + }, + { + "epoch": 2.258274928873777, + "grad_norm": 7.638558387756348, + "learning_rate": 8.401646622170804e-05, + "loss": 0.7333, + "step": 32544 + }, + { + "epoch": 2.2583443203108735, + "grad_norm": 1.8433626890182495, + "learning_rate": 8.401091986613258e-05, + "loss": 0.4512, + "step": 32545 + }, + { + "epoch": 2.2584137117479703, + "grad_norm": 1.3950127363204956, + "learning_rate": 8.40053735610328e-05, + "loss": 0.3458, + "step": 32546 + }, + { + "epoch": 2.2584831031850667, + "grad_norm": 2.465726613998413, + "learning_rate": 8.39998273064262e-05, + "loss": 0.4749, + "step": 32547 + }, + { + "epoch": 2.2585524946221636, + "grad_norm": 0.8789460062980652, + "learning_rate": 8.399428110233028e-05, + "loss": 0.3395, + "step": 32548 + }, + { + "epoch": 2.2586218860592604, + "grad_norm": 1.4892655611038208, + "learning_rate": 8.398873494876254e-05, + "loss": 0.5103, + "step": 32549 + }, + { + "epoch": 2.258691277496357, + "grad_norm": 1.4682729244232178, + "learning_rate": 8.398318884574052e-05, + "loss": 0.5487, + "step": 32550 + }, + { + "epoch": 2.2587606689334536, + "grad_norm": 1.2193324565887451, + "learning_rate": 8.397764279328171e-05, + "loss": 0.3778, + "step": 32551 + }, + { + "epoch": 2.2588300603705505, + "grad_norm": 1.8046005964279175, + "learning_rate": 8.397209679140366e-05, + "loss": 0.513, + "step": 32552 + }, + { + "epoch": 2.258899451807647, + "grad_norm": 1.32156240940094, + "learning_rate": 8.39665508401238e-05, + "loss": 0.4499, + "step": 32553 + }, + { + "epoch": 2.2589688432447437, + "grad_norm": 1.3567225933074951, + "learning_rate": 8.396100493945968e-05, + "loss": 0.5233, + "step": 32554 + }, + { + "epoch": 2.2590382346818405, + "grad_norm": 1.3538570404052734, + "learning_rate": 8.395545908942882e-05, + "loss": 0.4674, + "step": 32555 + }, + { + "epoch": 2.259107626118937, + "grad_norm": 1.6585792303085327, + "learning_rate": 8.39499132900487e-05, + "loss": 0.4436, + "step": 32556 + }, + { + "epoch": 2.2591770175560337, + "grad_norm": 1.4216604232788086, + "learning_rate": 8.394436754133684e-05, + "loss": 0.3897, + "step": 32557 + }, + { + "epoch": 2.25924640899313, + "grad_norm": 1.2561379671096802, + "learning_rate": 8.393882184331075e-05, + "loss": 0.3315, + "step": 32558 + }, + { + "epoch": 2.259315800430227, + "grad_norm": 1.3223226070404053, + "learning_rate": 8.393327619598798e-05, + "loss": 0.3835, + "step": 32559 + }, + { + "epoch": 2.259385191867324, + "grad_norm": 1.3541561365127563, + "learning_rate": 8.392773059938596e-05, + "loss": 0.3236, + "step": 32560 + }, + { + "epoch": 2.25945458330442, + "grad_norm": 1.4342234134674072, + "learning_rate": 8.392218505352222e-05, + "loss": 0.3999, + "step": 32561 + }, + { + "epoch": 2.259523974741517, + "grad_norm": 1.0315629243850708, + "learning_rate": 8.39166395584143e-05, + "loss": 0.3647, + "step": 32562 + }, + { + "epoch": 2.2595933661786134, + "grad_norm": 1.5444656610488892, + "learning_rate": 8.391109411407965e-05, + "loss": 0.581, + "step": 32563 + }, + { + "epoch": 2.2596627576157102, + "grad_norm": 1.3807380199432373, + "learning_rate": 8.390554872053584e-05, + "loss": 0.4557, + "step": 32564 + }, + { + "epoch": 2.259732149052807, + "grad_norm": 1.3210937976837158, + "learning_rate": 8.390000337780034e-05, + "loss": 0.3693, + "step": 32565 + }, + { + "epoch": 2.2598015404899034, + "grad_norm": 1.3659464120864868, + "learning_rate": 8.389445808589067e-05, + "loss": 0.3189, + "step": 32566 + }, + { + "epoch": 2.2598709319270003, + "grad_norm": 1.4287469387054443, + "learning_rate": 8.38889128448243e-05, + "loss": 0.4579, + "step": 32567 + }, + { + "epoch": 2.2599403233640967, + "grad_norm": 1.7856733798980713, + "learning_rate": 8.388336765461877e-05, + "loss": 0.5328, + "step": 32568 + }, + { + "epoch": 2.2600097148011935, + "grad_norm": 1.2761842012405396, + "learning_rate": 8.387782251529158e-05, + "loss": 0.4894, + "step": 32569 + }, + { + "epoch": 2.2600791062382903, + "grad_norm": 1.8201040029525757, + "learning_rate": 8.387227742686022e-05, + "loss": 0.4426, + "step": 32570 + }, + { + "epoch": 2.2601484976753867, + "grad_norm": 1.0959302186965942, + "learning_rate": 8.386673238934222e-05, + "loss": 0.4327, + "step": 32571 + }, + { + "epoch": 2.2602178891124836, + "grad_norm": 1.2068332433700562, + "learning_rate": 8.386118740275511e-05, + "loss": 0.4552, + "step": 32572 + }, + { + "epoch": 2.26028728054958, + "grad_norm": 1.9376863241195679, + "learning_rate": 8.385564246711632e-05, + "loss": 0.5403, + "step": 32573 + }, + { + "epoch": 2.260356671986677, + "grad_norm": 1.4620697498321533, + "learning_rate": 8.385009758244339e-05, + "loss": 0.4196, + "step": 32574 + }, + { + "epoch": 2.2604260634237736, + "grad_norm": 0.9943381547927856, + "learning_rate": 8.384455274875385e-05, + "loss": 0.3532, + "step": 32575 + }, + { + "epoch": 2.26049545486087, + "grad_norm": 1.5249720811843872, + "learning_rate": 8.383900796606517e-05, + "loss": 0.4777, + "step": 32576 + }, + { + "epoch": 2.260564846297967, + "grad_norm": 1.0949571132659912, + "learning_rate": 8.383346323439486e-05, + "loss": 0.3318, + "step": 32577 + }, + { + "epoch": 2.2606342377350637, + "grad_norm": 1.3626937866210938, + "learning_rate": 8.382791855376045e-05, + "loss": 0.4157, + "step": 32578 + }, + { + "epoch": 2.26070362917216, + "grad_norm": 1.3464547395706177, + "learning_rate": 8.382237392417943e-05, + "loss": 0.506, + "step": 32579 + }, + { + "epoch": 2.260773020609257, + "grad_norm": 1.2310293912887573, + "learning_rate": 8.381682934566927e-05, + "loss": 0.3469, + "step": 32580 + }, + { + "epoch": 2.2608424120463537, + "grad_norm": 1.1756917238235474, + "learning_rate": 8.381128481824753e-05, + "loss": 0.4071, + "step": 32581 + }, + { + "epoch": 2.26091180348345, + "grad_norm": 2.9591457843780518, + "learning_rate": 8.380574034193167e-05, + "loss": 0.4106, + "step": 32582 + }, + { + "epoch": 2.260981194920547, + "grad_norm": 2.1069841384887695, + "learning_rate": 8.380019591673922e-05, + "loss": 0.5105, + "step": 32583 + }, + { + "epoch": 2.2610505863576433, + "grad_norm": 1.2598036527633667, + "learning_rate": 8.379465154268764e-05, + "loss": 0.4608, + "step": 32584 + }, + { + "epoch": 2.26111997779474, + "grad_norm": 1.3608438968658447, + "learning_rate": 8.378910721979448e-05, + "loss": 0.3227, + "step": 32585 + }, + { + "epoch": 2.261189369231837, + "grad_norm": 3.12187123298645, + "learning_rate": 8.378356294807728e-05, + "loss": 0.4899, + "step": 32586 + }, + { + "epoch": 2.2612587606689334, + "grad_norm": 1.5987850427627563, + "learning_rate": 8.377801872755343e-05, + "loss": 0.3849, + "step": 32587 + }, + { + "epoch": 2.26132815210603, + "grad_norm": 0.9942344427108765, + "learning_rate": 8.37724745582405e-05, + "loss": 0.368, + "step": 32588 + }, + { + "epoch": 2.2613975435431266, + "grad_norm": 1.2362534999847412, + "learning_rate": 8.376693044015599e-05, + "loss": 0.4035, + "step": 32589 + }, + { + "epoch": 2.2614669349802234, + "grad_norm": 1.904274821281433, + "learning_rate": 8.37613863733174e-05, + "loss": 0.4189, + "step": 32590 + }, + { + "epoch": 2.2615363264173203, + "grad_norm": 1.0530476570129395, + "learning_rate": 8.375584235774222e-05, + "loss": 0.3853, + "step": 32591 + }, + { + "epoch": 2.2616057178544167, + "grad_norm": 1.4315444231033325, + "learning_rate": 8.375029839344797e-05, + "loss": 0.4577, + "step": 32592 + }, + { + "epoch": 2.2616751092915135, + "grad_norm": 1.3827972412109375, + "learning_rate": 8.374475448045217e-05, + "loss": 0.4309, + "step": 32593 + }, + { + "epoch": 2.26174450072861, + "grad_norm": 1.7936986684799194, + "learning_rate": 8.373921061877225e-05, + "loss": 0.5102, + "step": 32594 + }, + { + "epoch": 2.2618138921657067, + "grad_norm": 1.5320252180099487, + "learning_rate": 8.373366680842578e-05, + "loss": 0.3201, + "step": 32595 + }, + { + "epoch": 2.2618832836028036, + "grad_norm": 1.586780309677124, + "learning_rate": 8.372812304943022e-05, + "loss": 0.3569, + "step": 32596 + }, + { + "epoch": 2.2619526750399, + "grad_norm": 1.119531273841858, + "learning_rate": 8.372257934180311e-05, + "loss": 0.3354, + "step": 32597 + }, + { + "epoch": 2.2620220664769968, + "grad_norm": 0.9825839996337891, + "learning_rate": 8.371703568556191e-05, + "loss": 0.33, + "step": 32598 + }, + { + "epoch": 2.262091457914093, + "grad_norm": 1.2812608480453491, + "learning_rate": 8.371149208072414e-05, + "loss": 0.3593, + "step": 32599 + }, + { + "epoch": 2.26216084935119, + "grad_norm": 1.7394031286239624, + "learning_rate": 8.370594852730732e-05, + "loss": 0.5089, + "step": 32600 + }, + { + "epoch": 2.262230240788287, + "grad_norm": 1.633782148361206, + "learning_rate": 8.370040502532894e-05, + "loss": 0.4923, + "step": 32601 + }, + { + "epoch": 2.262299632225383, + "grad_norm": 1.2246116399765015, + "learning_rate": 8.369486157480646e-05, + "loss": 0.3444, + "step": 32602 + }, + { + "epoch": 2.26236902366248, + "grad_norm": 1.999168872833252, + "learning_rate": 8.368931817575739e-05, + "loss": 0.4168, + "step": 32603 + }, + { + "epoch": 2.262438415099577, + "grad_norm": 1.3080154657363892, + "learning_rate": 8.368377482819927e-05, + "loss": 0.3665, + "step": 32604 + }, + { + "epoch": 2.2625078065366733, + "grad_norm": 1.0381736755371094, + "learning_rate": 8.367823153214958e-05, + "loss": 0.2683, + "step": 32605 + }, + { + "epoch": 2.26257719797377, + "grad_norm": 1.4260425567626953, + "learning_rate": 8.367268828762582e-05, + "loss": 0.4359, + "step": 32606 + }, + { + "epoch": 2.262646589410867, + "grad_norm": 1.1470015048980713, + "learning_rate": 8.366714509464552e-05, + "loss": 0.4362, + "step": 32607 + }, + { + "epoch": 2.2627159808479633, + "grad_norm": 1.3871842622756958, + "learning_rate": 8.366160195322613e-05, + "loss": 0.3531, + "step": 32608 + }, + { + "epoch": 2.26278537228506, + "grad_norm": 1.5344148874282837, + "learning_rate": 8.365605886338514e-05, + "loss": 0.5847, + "step": 32609 + }, + { + "epoch": 2.2628547637221565, + "grad_norm": 1.5127289295196533, + "learning_rate": 8.36505158251401e-05, + "loss": 0.4214, + "step": 32610 + }, + { + "epoch": 2.2629241551592534, + "grad_norm": 1.6645839214324951, + "learning_rate": 8.364497283850847e-05, + "loss": 0.4173, + "step": 32611 + }, + { + "epoch": 2.26299354659635, + "grad_norm": 1.3980152606964111, + "learning_rate": 8.363942990350776e-05, + "loss": 0.396, + "step": 32612 + }, + { + "epoch": 2.2630629380334466, + "grad_norm": 1.644057273864746, + "learning_rate": 8.36338870201555e-05, + "loss": 0.4586, + "step": 32613 + }, + { + "epoch": 2.2631323294705434, + "grad_norm": 1.2006471157073975, + "learning_rate": 8.362834418846915e-05, + "loss": 0.3956, + "step": 32614 + }, + { + "epoch": 2.26320172090764, + "grad_norm": 1.2075843811035156, + "learning_rate": 8.362280140846621e-05, + "loss": 0.3648, + "step": 32615 + }, + { + "epoch": 2.2632711123447367, + "grad_norm": 1.4156153202056885, + "learning_rate": 8.36172586801642e-05, + "loss": 0.4223, + "step": 32616 + }, + { + "epoch": 2.2633405037818335, + "grad_norm": 1.2260024547576904, + "learning_rate": 8.361171600358059e-05, + "loss": 0.5642, + "step": 32617 + }, + { + "epoch": 2.26340989521893, + "grad_norm": 1.3526731729507446, + "learning_rate": 8.360617337873288e-05, + "loss": 0.4023, + "step": 32618 + }, + { + "epoch": 2.2634792866560267, + "grad_norm": 1.4664238691329956, + "learning_rate": 8.360063080563859e-05, + "loss": 0.4401, + "step": 32619 + }, + { + "epoch": 2.263548678093123, + "grad_norm": 1.077731728553772, + "learning_rate": 8.359508828431519e-05, + "loss": 0.3212, + "step": 32620 + }, + { + "epoch": 2.26361806953022, + "grad_norm": 1.2082738876342773, + "learning_rate": 8.358954581478024e-05, + "loss": 0.4018, + "step": 32621 + }, + { + "epoch": 2.2636874609673168, + "grad_norm": 1.1758860349655151, + "learning_rate": 8.358400339705116e-05, + "loss": 0.3839, + "step": 32622 + }, + { + "epoch": 2.263756852404413, + "grad_norm": 0.8468275666236877, + "learning_rate": 8.357846103114548e-05, + "loss": 0.1965, + "step": 32623 + }, + { + "epoch": 2.26382624384151, + "grad_norm": 1.0517702102661133, + "learning_rate": 8.357291871708067e-05, + "loss": 0.4726, + "step": 32624 + }, + { + "epoch": 2.2638956352786064, + "grad_norm": 1.5530192852020264, + "learning_rate": 8.356737645487427e-05, + "loss": 0.4808, + "step": 32625 + }, + { + "epoch": 2.263965026715703, + "grad_norm": 1.0009758472442627, + "learning_rate": 8.356183424454375e-05, + "loss": 0.2845, + "step": 32626 + }, + { + "epoch": 2.2640344181528, + "grad_norm": 1.2020533084869385, + "learning_rate": 8.35562920861066e-05, + "loss": 0.3133, + "step": 32627 + }, + { + "epoch": 2.2641038095898964, + "grad_norm": 1.1581307649612427, + "learning_rate": 8.355074997958039e-05, + "loss": 0.3039, + "step": 32628 + }, + { + "epoch": 2.2641732010269933, + "grad_norm": 1.4134140014648438, + "learning_rate": 8.354520792498249e-05, + "loss": 0.6416, + "step": 32629 + }, + { + "epoch": 2.26424259246409, + "grad_norm": 1.4174253940582275, + "learning_rate": 8.353966592233048e-05, + "loss": 0.505, + "step": 32630 + }, + { + "epoch": 2.2643119839011865, + "grad_norm": 1.2949131727218628, + "learning_rate": 8.353412397164182e-05, + "loss": 0.4537, + "step": 32631 + }, + { + "epoch": 2.2643813753382833, + "grad_norm": 1.6394381523132324, + "learning_rate": 8.352858207293406e-05, + "loss": 0.4816, + "step": 32632 + }, + { + "epoch": 2.26445076677538, + "grad_norm": 1.2324453592300415, + "learning_rate": 8.352304022622462e-05, + "loss": 0.4148, + "step": 32633 + }, + { + "epoch": 2.2645201582124765, + "grad_norm": 1.5875828266143799, + "learning_rate": 8.351749843153103e-05, + "loss": 0.5773, + "step": 32634 + }, + { + "epoch": 2.2645895496495734, + "grad_norm": 1.4518178701400757, + "learning_rate": 8.351195668887083e-05, + "loss": 0.2831, + "step": 32635 + }, + { + "epoch": 2.2646589410866698, + "grad_norm": 1.4999659061431885, + "learning_rate": 8.350641499826143e-05, + "loss": 0.5033, + "step": 32636 + }, + { + "epoch": 2.2647283325237666, + "grad_norm": 1.2251437902450562, + "learning_rate": 8.350087335972036e-05, + "loss": 0.4464, + "step": 32637 + }, + { + "epoch": 2.2647977239608634, + "grad_norm": 1.329573631286621, + "learning_rate": 8.349533177326512e-05, + "loss": 0.4002, + "step": 32638 + }, + { + "epoch": 2.26486711539796, + "grad_norm": 0.969353437423706, + "learning_rate": 8.34897902389132e-05, + "loss": 0.2891, + "step": 32639 + }, + { + "epoch": 2.2649365068350567, + "grad_norm": 1.3666167259216309, + "learning_rate": 8.34842487566821e-05, + "loss": 0.2301, + "step": 32640 + }, + { + "epoch": 2.265005898272153, + "grad_norm": 1.0829052925109863, + "learning_rate": 8.347870732658935e-05, + "loss": 0.2529, + "step": 32641 + }, + { + "epoch": 2.26507528970925, + "grad_norm": 1.5264191627502441, + "learning_rate": 8.347316594865237e-05, + "loss": 0.648, + "step": 32642 + }, + { + "epoch": 2.2651446811463467, + "grad_norm": 1.5216078758239746, + "learning_rate": 8.346762462288868e-05, + "loss": 0.3786, + "step": 32643 + }, + { + "epoch": 2.265214072583443, + "grad_norm": 1.3530261516571045, + "learning_rate": 8.346208334931578e-05, + "loss": 0.3443, + "step": 32644 + }, + { + "epoch": 2.26528346402054, + "grad_norm": 1.2003709077835083, + "learning_rate": 8.345654212795117e-05, + "loss": 0.312, + "step": 32645 + }, + { + "epoch": 2.2653528554576363, + "grad_norm": 1.3322311639785767, + "learning_rate": 8.345100095881233e-05, + "loss": 0.4113, + "step": 32646 + }, + { + "epoch": 2.265422246894733, + "grad_norm": 1.3561067581176758, + "learning_rate": 8.344545984191678e-05, + "loss": 0.3371, + "step": 32647 + }, + { + "epoch": 2.26549163833183, + "grad_norm": 1.469557762145996, + "learning_rate": 8.343991877728201e-05, + "loss": 0.4298, + "step": 32648 + }, + { + "epoch": 2.2655610297689264, + "grad_norm": 1.2693397998809814, + "learning_rate": 8.343437776492546e-05, + "loss": 0.3894, + "step": 32649 + }, + { + "epoch": 2.265630421206023, + "grad_norm": 1.3839130401611328, + "learning_rate": 8.342883680486466e-05, + "loss": 0.4637, + "step": 32650 + }, + { + "epoch": 2.2656998126431196, + "grad_norm": 1.5776841640472412, + "learning_rate": 8.342329589711712e-05, + "loss": 0.5464, + "step": 32651 + }, + { + "epoch": 2.2657692040802164, + "grad_norm": 1.024732232093811, + "learning_rate": 8.341775504170028e-05, + "loss": 0.3338, + "step": 32652 + }, + { + "epoch": 2.2658385955173133, + "grad_norm": 2.4962635040283203, + "learning_rate": 8.341221423863168e-05, + "loss": 0.442, + "step": 32653 + }, + { + "epoch": 2.2659079869544096, + "grad_norm": 1.0502620935440063, + "learning_rate": 8.340667348792879e-05, + "loss": 0.4197, + "step": 32654 + }, + { + "epoch": 2.2659773783915065, + "grad_norm": 1.2936331033706665, + "learning_rate": 8.340113278960915e-05, + "loss": 0.4354, + "step": 32655 + }, + { + "epoch": 2.2660467698286033, + "grad_norm": 1.18980872631073, + "learning_rate": 8.339559214369016e-05, + "loss": 0.3794, + "step": 32656 + }, + { + "epoch": 2.2661161612656997, + "grad_norm": 1.131859540939331, + "learning_rate": 8.339005155018937e-05, + "loss": 0.3714, + "step": 32657 + }, + { + "epoch": 2.2661855527027965, + "grad_norm": 1.3776030540466309, + "learning_rate": 8.338451100912424e-05, + "loss": 0.4384, + "step": 32658 + }, + { + "epoch": 2.2662549441398934, + "grad_norm": 1.9629695415496826, + "learning_rate": 8.33789705205123e-05, + "loss": 0.5194, + "step": 32659 + }, + { + "epoch": 2.2663243355769898, + "grad_norm": 1.0237209796905518, + "learning_rate": 8.337343008437102e-05, + "loss": 0.2685, + "step": 32660 + }, + { + "epoch": 2.2663937270140866, + "grad_norm": 1.5698246955871582, + "learning_rate": 8.336788970071789e-05, + "loss": 0.6441, + "step": 32661 + }, + { + "epoch": 2.266463118451183, + "grad_norm": 1.462708592414856, + "learning_rate": 8.336234936957043e-05, + "loss": 0.4451, + "step": 32662 + }, + { + "epoch": 2.26653250988828, + "grad_norm": 1.036507487297058, + "learning_rate": 8.335680909094608e-05, + "loss": 0.2995, + "step": 32663 + }, + { + "epoch": 2.2666019013253766, + "grad_norm": 1.1300547122955322, + "learning_rate": 8.335126886486234e-05, + "loss": 0.3185, + "step": 32664 + }, + { + "epoch": 2.266671292762473, + "grad_norm": 1.703682780265808, + "learning_rate": 8.334572869133672e-05, + "loss": 0.5614, + "step": 32665 + }, + { + "epoch": 2.26674068419957, + "grad_norm": 1.2712653875350952, + "learning_rate": 8.334018857038671e-05, + "loss": 0.4504, + "step": 32666 + }, + { + "epoch": 2.2668100756366663, + "grad_norm": 1.7694646120071411, + "learning_rate": 8.33346485020298e-05, + "loss": 0.6062, + "step": 32667 + }, + { + "epoch": 2.266879467073763, + "grad_norm": 1.2619847059249878, + "learning_rate": 8.332910848628345e-05, + "loss": 0.6467, + "step": 32668 + }, + { + "epoch": 2.26694885851086, + "grad_norm": 1.1413567066192627, + "learning_rate": 8.332356852316519e-05, + "loss": 0.5788, + "step": 32669 + }, + { + "epoch": 2.2670182499479563, + "grad_norm": 1.3501795530319214, + "learning_rate": 8.33180286126925e-05, + "loss": 0.4276, + "step": 32670 + }, + { + "epoch": 2.267087641385053, + "grad_norm": 1.06039297580719, + "learning_rate": 8.331248875488283e-05, + "loss": 0.4157, + "step": 32671 + }, + { + "epoch": 2.2671570328221495, + "grad_norm": 1.6960629224777222, + "learning_rate": 8.33069489497537e-05, + "loss": 0.4746, + "step": 32672 + }, + { + "epoch": 2.2672264242592464, + "grad_norm": 1.681749701499939, + "learning_rate": 8.330140919732259e-05, + "loss": 0.4433, + "step": 32673 + }, + { + "epoch": 2.267295815696343, + "grad_norm": 1.341180682182312, + "learning_rate": 8.329586949760699e-05, + "loss": 0.4511, + "step": 32674 + }, + { + "epoch": 2.2673652071334396, + "grad_norm": 1.4259240627288818, + "learning_rate": 8.32903298506244e-05, + "loss": 0.4861, + "step": 32675 + }, + { + "epoch": 2.2674345985705364, + "grad_norm": 1.1975280046463013, + "learning_rate": 8.328479025639236e-05, + "loss": 0.3685, + "step": 32676 + }, + { + "epoch": 2.2675039900076333, + "grad_norm": 1.292210340499878, + "learning_rate": 8.327925071492823e-05, + "loss": 0.3872, + "step": 32677 + }, + { + "epoch": 2.2675733814447296, + "grad_norm": 1.1045098304748535, + "learning_rate": 8.327371122624958e-05, + "loss": 0.3275, + "step": 32678 + }, + { + "epoch": 2.2676427728818265, + "grad_norm": 1.2545026540756226, + "learning_rate": 8.326817179037387e-05, + "loss": 0.387, + "step": 32679 + }, + { + "epoch": 2.267712164318923, + "grad_norm": 1.2306562662124634, + "learning_rate": 8.32626324073186e-05, + "loss": 0.3864, + "step": 32680 + }, + { + "epoch": 2.2677815557560197, + "grad_norm": 1.7072675228118896, + "learning_rate": 8.325709307710126e-05, + "loss": 0.4498, + "step": 32681 + }, + { + "epoch": 2.2678509471931165, + "grad_norm": 1.1217751502990723, + "learning_rate": 8.325155379973936e-05, + "loss": 0.3436, + "step": 32682 + }, + { + "epoch": 2.267920338630213, + "grad_norm": 1.1448512077331543, + "learning_rate": 8.324601457525036e-05, + "loss": 0.3989, + "step": 32683 + }, + { + "epoch": 2.2679897300673098, + "grad_norm": 1.1579471826553345, + "learning_rate": 8.324047540365173e-05, + "loss": 0.4809, + "step": 32684 + }, + { + "epoch": 2.2680591215044066, + "grad_norm": 1.1456471681594849, + "learning_rate": 8.323493628496098e-05, + "loss": 0.3839, + "step": 32685 + }, + { + "epoch": 2.268128512941503, + "grad_norm": 1.291385293006897, + "learning_rate": 8.32293972191956e-05, + "loss": 0.4661, + "step": 32686 + }, + { + "epoch": 2.2681979043786, + "grad_norm": 1.0575369596481323, + "learning_rate": 8.322385820637304e-05, + "loss": 0.3536, + "step": 32687 + }, + { + "epoch": 2.268267295815696, + "grad_norm": 1.4439482688903809, + "learning_rate": 8.321831924651083e-05, + "loss": 0.6753, + "step": 32688 + }, + { + "epoch": 2.268336687252793, + "grad_norm": 1.673011302947998, + "learning_rate": 8.321278033962644e-05, + "loss": 0.3565, + "step": 32689 + }, + { + "epoch": 2.26840607868989, + "grad_norm": 1.3397756814956665, + "learning_rate": 8.320724148573739e-05, + "loss": 0.621, + "step": 32690 + }, + { + "epoch": 2.2684754701269862, + "grad_norm": 1.2398656606674194, + "learning_rate": 8.32017026848611e-05, + "loss": 0.3643, + "step": 32691 + }, + { + "epoch": 2.268544861564083, + "grad_norm": 1.7105926275253296, + "learning_rate": 8.319616393701505e-05, + "loss": 0.5758, + "step": 32692 + }, + { + "epoch": 2.2686142530011795, + "grad_norm": 1.011116623878479, + "learning_rate": 8.319062524221679e-05, + "loss": 0.4049, + "step": 32693 + }, + { + "epoch": 2.2686836444382763, + "grad_norm": 1.1504080295562744, + "learning_rate": 8.318508660048378e-05, + "loss": 0.3922, + "step": 32694 + }, + { + "epoch": 2.268753035875373, + "grad_norm": 1.3504693508148193, + "learning_rate": 8.317954801183349e-05, + "loss": 0.3582, + "step": 32695 + }, + { + "epoch": 2.2688224273124695, + "grad_norm": 1.4674761295318604, + "learning_rate": 8.317400947628342e-05, + "loss": 0.5265, + "step": 32696 + }, + { + "epoch": 2.2688918187495664, + "grad_norm": 0.9971261024475098, + "learning_rate": 8.316847099385109e-05, + "loss": 0.2624, + "step": 32697 + }, + { + "epoch": 2.2689612101866627, + "grad_norm": 0.9826361536979675, + "learning_rate": 8.31629325645539e-05, + "loss": 0.2634, + "step": 32698 + }, + { + "epoch": 2.2690306016237596, + "grad_norm": 1.5010703802108765, + "learning_rate": 8.315739418840938e-05, + "loss": 0.4832, + "step": 32699 + }, + { + "epoch": 2.2690999930608564, + "grad_norm": 1.4582405090332031, + "learning_rate": 8.315185586543502e-05, + "loss": 0.481, + "step": 32700 + }, + { + "epoch": 2.269169384497953, + "grad_norm": 1.2400212287902832, + "learning_rate": 8.314631759564827e-05, + "loss": 0.3598, + "step": 32701 + }, + { + "epoch": 2.2692387759350496, + "grad_norm": 1.2358136177062988, + "learning_rate": 8.314077937906668e-05, + "loss": 0.5215, + "step": 32702 + }, + { + "epoch": 2.2693081673721465, + "grad_norm": 1.296201467514038, + "learning_rate": 8.313524121570768e-05, + "loss": 0.4258, + "step": 32703 + }, + { + "epoch": 2.269377558809243, + "grad_norm": 1.3785046339035034, + "learning_rate": 8.312970310558878e-05, + "loss": 0.4831, + "step": 32704 + }, + { + "epoch": 2.2694469502463397, + "grad_norm": 1.3822252750396729, + "learning_rate": 8.312416504872744e-05, + "loss": 0.4494, + "step": 32705 + }, + { + "epoch": 2.269516341683436, + "grad_norm": 1.3331234455108643, + "learning_rate": 8.311862704514115e-05, + "loss": 0.4244, + "step": 32706 + }, + { + "epoch": 2.269585733120533, + "grad_norm": 1.4807809591293335, + "learning_rate": 8.311308909484739e-05, + "loss": 0.4765, + "step": 32707 + }, + { + "epoch": 2.2696551245576297, + "grad_norm": 1.40172278881073, + "learning_rate": 8.310755119786364e-05, + "loss": 0.4937, + "step": 32708 + }, + { + "epoch": 2.269724515994726, + "grad_norm": 0.9830442070960999, + "learning_rate": 8.31020133542074e-05, + "loss": 0.2877, + "step": 32709 + }, + { + "epoch": 2.269793907431823, + "grad_norm": 1.1653380393981934, + "learning_rate": 8.309647556389614e-05, + "loss": 0.372, + "step": 32710 + }, + { + "epoch": 2.26986329886892, + "grad_norm": 1.3541538715362549, + "learning_rate": 8.309093782694739e-05, + "loss": 0.6699, + "step": 32711 + }, + { + "epoch": 2.269932690306016, + "grad_norm": 1.1905291080474854, + "learning_rate": 8.308540014337854e-05, + "loss": 0.4701, + "step": 32712 + }, + { + "epoch": 2.270002081743113, + "grad_norm": 1.3852518796920776, + "learning_rate": 8.307986251320712e-05, + "loss": 0.4143, + "step": 32713 + }, + { + "epoch": 2.2700714731802094, + "grad_norm": 2.1358563899993896, + "learning_rate": 8.307432493645061e-05, + "loss": 0.5135, + "step": 32714 + }, + { + "epoch": 2.2701408646173062, + "grad_norm": 1.1090890169143677, + "learning_rate": 8.30687874131265e-05, + "loss": 0.2656, + "step": 32715 + }, + { + "epoch": 2.270210256054403, + "grad_norm": 1.4659403562545776, + "learning_rate": 8.306324994325225e-05, + "loss": 0.3049, + "step": 32716 + }, + { + "epoch": 2.2702796474914995, + "grad_norm": 1.1812914609909058, + "learning_rate": 8.305771252684542e-05, + "loss": 0.2291, + "step": 32717 + }, + { + "epoch": 2.2703490389285963, + "grad_norm": 0.9060866236686707, + "learning_rate": 8.305217516392337e-05, + "loss": 0.3138, + "step": 32718 + }, + { + "epoch": 2.2704184303656927, + "grad_norm": 1.0333112478256226, + "learning_rate": 8.304663785450362e-05, + "loss": 0.3414, + "step": 32719 + }, + { + "epoch": 2.2704878218027895, + "grad_norm": 1.6075568199157715, + "learning_rate": 8.30411005986037e-05, + "loss": 0.4429, + "step": 32720 + }, + { + "epoch": 2.2705572132398864, + "grad_norm": 0.9480907320976257, + "learning_rate": 8.303556339624106e-05, + "loss": 0.2533, + "step": 32721 + }, + { + "epoch": 2.2706266046769827, + "grad_norm": 1.0788508653640747, + "learning_rate": 8.303002624743315e-05, + "loss": 0.3845, + "step": 32722 + }, + { + "epoch": 2.2706959961140796, + "grad_norm": 1.7981185913085938, + "learning_rate": 8.302448915219749e-05, + "loss": 0.4699, + "step": 32723 + }, + { + "epoch": 2.270765387551176, + "grad_norm": 1.777122974395752, + "learning_rate": 8.30189521105516e-05, + "loss": 0.5209, + "step": 32724 + }, + { + "epoch": 2.270834778988273, + "grad_norm": 0.986488938331604, + "learning_rate": 8.301341512251285e-05, + "loss": 0.4011, + "step": 32725 + }, + { + "epoch": 2.2709041704253696, + "grad_norm": 1.4860529899597168, + "learning_rate": 8.300787818809879e-05, + "loss": 0.4411, + "step": 32726 + }, + { + "epoch": 2.270973561862466, + "grad_norm": 1.2039380073547363, + "learning_rate": 8.300234130732687e-05, + "loss": 0.2939, + "step": 32727 + }, + { + "epoch": 2.271042953299563, + "grad_norm": 1.1784772872924805, + "learning_rate": 8.29968044802146e-05, + "loss": 0.3745, + "step": 32728 + }, + { + "epoch": 2.2711123447366597, + "grad_norm": 1.3902024030685425, + "learning_rate": 8.299126770677943e-05, + "loss": 0.523, + "step": 32729 + }, + { + "epoch": 2.271181736173756, + "grad_norm": 1.674307107925415, + "learning_rate": 8.298573098703886e-05, + "loss": 0.3032, + "step": 32730 + }, + { + "epoch": 2.271251127610853, + "grad_norm": 1.2153644561767578, + "learning_rate": 8.298019432101044e-05, + "loss": 0.384, + "step": 32731 + }, + { + "epoch": 2.2713205190479493, + "grad_norm": 1.3194464445114136, + "learning_rate": 8.297465770871149e-05, + "loss": 0.4313, + "step": 32732 + }, + { + "epoch": 2.271389910485046, + "grad_norm": 1.2056021690368652, + "learning_rate": 8.296912115015958e-05, + "loss": 0.3676, + "step": 32733 + }, + { + "epoch": 2.271459301922143, + "grad_norm": 1.2789154052734375, + "learning_rate": 8.296358464537217e-05, + "loss": 0.3843, + "step": 32734 + }, + { + "epoch": 2.2715286933592393, + "grad_norm": 1.2583577632904053, + "learning_rate": 8.295804819436677e-05, + "loss": 0.4276, + "step": 32735 + }, + { + "epoch": 2.271598084796336, + "grad_norm": 0.9767405390739441, + "learning_rate": 8.295251179716083e-05, + "loss": 0.3147, + "step": 32736 + }, + { + "epoch": 2.271667476233433, + "grad_norm": 1.082432508468628, + "learning_rate": 8.294697545377182e-05, + "loss": 0.2972, + "step": 32737 + }, + { + "epoch": 2.2717368676705294, + "grad_norm": 1.9646968841552734, + "learning_rate": 8.294143916421726e-05, + "loss": 0.538, + "step": 32738 + }, + { + "epoch": 2.2718062591076262, + "grad_norm": 1.5610156059265137, + "learning_rate": 8.293590292851458e-05, + "loss": 0.3584, + "step": 32739 + }, + { + "epoch": 2.2718756505447226, + "grad_norm": 1.455505609512329, + "learning_rate": 8.293036674668128e-05, + "loss": 0.5984, + "step": 32740 + }, + { + "epoch": 2.2719450419818195, + "grad_norm": 1.338351845741272, + "learning_rate": 8.292483061873481e-05, + "loss": 0.442, + "step": 32741 + }, + { + "epoch": 2.2720144334189163, + "grad_norm": 1.397353172302246, + "learning_rate": 8.291929454469268e-05, + "loss": 0.4413, + "step": 32742 + }, + { + "epoch": 2.2720838248560127, + "grad_norm": 1.3496063947677612, + "learning_rate": 8.291375852457235e-05, + "loss": 0.3843, + "step": 32743 + }, + { + "epoch": 2.2721532162931095, + "grad_norm": 0.8373847007751465, + "learning_rate": 8.29082225583913e-05, + "loss": 0.3043, + "step": 32744 + }, + { + "epoch": 2.272222607730206, + "grad_norm": 1.696488857269287, + "learning_rate": 8.290268664616706e-05, + "loss": 0.4874, + "step": 32745 + }, + { + "epoch": 2.2722919991673027, + "grad_norm": 1.2399247884750366, + "learning_rate": 8.289715078791701e-05, + "loss": 0.2623, + "step": 32746 + }, + { + "epoch": 2.2723613906043996, + "grad_norm": 1.3136160373687744, + "learning_rate": 8.289161498365866e-05, + "loss": 0.5336, + "step": 32747 + }, + { + "epoch": 2.272430782041496, + "grad_norm": 1.4787743091583252, + "learning_rate": 8.288607923340951e-05, + "loss": 0.4008, + "step": 32748 + }, + { + "epoch": 2.272500173478593, + "grad_norm": 1.5491863489151, + "learning_rate": 8.288054353718701e-05, + "loss": 0.3344, + "step": 32749 + }, + { + "epoch": 2.272569564915689, + "grad_norm": 2.0301620960235596, + "learning_rate": 8.287500789500866e-05, + "loss": 0.4591, + "step": 32750 + }, + { + "epoch": 2.272638956352786, + "grad_norm": 1.1689969301223755, + "learning_rate": 8.286947230689192e-05, + "loss": 0.3981, + "step": 32751 + }, + { + "epoch": 2.272708347789883, + "grad_norm": 2.1006505489349365, + "learning_rate": 8.28639367728543e-05, + "loss": 0.5115, + "step": 32752 + }, + { + "epoch": 2.2727777392269792, + "grad_norm": 1.4988908767700195, + "learning_rate": 8.28584012929132e-05, + "loss": 0.5351, + "step": 32753 + }, + { + "epoch": 2.272847130664076, + "grad_norm": 1.8258236646652222, + "learning_rate": 8.285286586708616e-05, + "loss": 0.4905, + "step": 32754 + }, + { + "epoch": 2.272916522101173, + "grad_norm": 1.1387096643447876, + "learning_rate": 8.284733049539062e-05, + "loss": 0.3464, + "step": 32755 + }, + { + "epoch": 2.2729859135382693, + "grad_norm": 1.2824773788452148, + "learning_rate": 8.284179517784408e-05, + "loss": 0.3571, + "step": 32756 + }, + { + "epoch": 2.273055304975366, + "grad_norm": 2.176370620727539, + "learning_rate": 8.283625991446399e-05, + "loss": 0.5344, + "step": 32757 + }, + { + "epoch": 2.2731246964124625, + "grad_norm": 1.0780822038650513, + "learning_rate": 8.283072470526783e-05, + "loss": 0.3698, + "step": 32758 + }, + { + "epoch": 2.2731940878495593, + "grad_norm": 1.1790133714675903, + "learning_rate": 8.282518955027314e-05, + "loss": 0.3763, + "step": 32759 + }, + { + "epoch": 2.273263479286656, + "grad_norm": 1.0537081956863403, + "learning_rate": 8.281965444949727e-05, + "loss": 0.3498, + "step": 32760 + }, + { + "epoch": 2.2733328707237526, + "grad_norm": 1.189281702041626, + "learning_rate": 8.281411940295777e-05, + "loss": 0.4183, + "step": 32761 + }, + { + "epoch": 2.2734022621608494, + "grad_norm": 1.9928654432296753, + "learning_rate": 8.28085844106721e-05, + "loss": 0.7401, + "step": 32762 + }, + { + "epoch": 2.2734716535979462, + "grad_norm": 2.2913572788238525, + "learning_rate": 8.280304947265773e-05, + "loss": 0.554, + "step": 32763 + }, + { + "epoch": 2.2735410450350426, + "grad_norm": 1.340286374092102, + "learning_rate": 8.279751458893215e-05, + "loss": 0.4378, + "step": 32764 + }, + { + "epoch": 2.2736104364721395, + "grad_norm": 1.287144422531128, + "learning_rate": 8.27919797595128e-05, + "loss": 0.5217, + "step": 32765 + }, + { + "epoch": 2.273679827909236, + "grad_norm": 2.179481029510498, + "learning_rate": 8.278644498441724e-05, + "loss": 0.3647, + "step": 32766 + }, + { + "epoch": 2.2737492193463327, + "grad_norm": 1.7803032398223877, + "learning_rate": 8.278091026366282e-05, + "loss": 0.5647, + "step": 32767 + }, + { + "epoch": 2.2738186107834295, + "grad_norm": 1.6825189590454102, + "learning_rate": 8.277537559726705e-05, + "loss": 0.5183, + "step": 32768 + }, + { + "epoch": 2.273888002220526, + "grad_norm": 1.512624979019165, + "learning_rate": 8.276984098524744e-05, + "loss": 0.4978, + "step": 32769 + }, + { + "epoch": 2.2739573936576227, + "grad_norm": 1.1495803594589233, + "learning_rate": 8.276430642762145e-05, + "loss": 0.3711, + "step": 32770 + }, + { + "epoch": 2.274026785094719, + "grad_norm": 1.776382565498352, + "learning_rate": 8.275877192440657e-05, + "loss": 0.4564, + "step": 32771 + }, + { + "epoch": 2.274096176531816, + "grad_norm": 1.2173292636871338, + "learning_rate": 8.275323747562021e-05, + "loss": 0.5081, + "step": 32772 + }, + { + "epoch": 2.274165567968913, + "grad_norm": 1.3833802938461304, + "learning_rate": 8.274770308127991e-05, + "loss": 0.3885, + "step": 32773 + }, + { + "epoch": 2.274234959406009, + "grad_norm": 1.5456771850585938, + "learning_rate": 8.274216874140307e-05, + "loss": 0.481, + "step": 32774 + }, + { + "epoch": 2.274304350843106, + "grad_norm": 1.1614940166473389, + "learning_rate": 8.273663445600724e-05, + "loss": 0.5363, + "step": 32775 + }, + { + "epoch": 2.2743737422802024, + "grad_norm": 1.7341192960739136, + "learning_rate": 8.273110022510982e-05, + "loss": 0.5452, + "step": 32776 + }, + { + "epoch": 2.2744431337172992, + "grad_norm": 0.9084320068359375, + "learning_rate": 8.272556604872832e-05, + "loss": 0.3426, + "step": 32777 + }, + { + "epoch": 2.274512525154396, + "grad_norm": 0.9687105417251587, + "learning_rate": 8.27200319268802e-05, + "loss": 0.3069, + "step": 32778 + }, + { + "epoch": 2.2745819165914924, + "grad_norm": 1.4021390676498413, + "learning_rate": 8.271449785958294e-05, + "loss": 0.5001, + "step": 32779 + }, + { + "epoch": 2.2746513080285893, + "grad_norm": 1.607747197151184, + "learning_rate": 8.270896384685404e-05, + "loss": 0.4321, + "step": 32780 + }, + { + "epoch": 2.274720699465686, + "grad_norm": 1.055236577987671, + "learning_rate": 8.270342988871089e-05, + "loss": 0.4143, + "step": 32781 + }, + { + "epoch": 2.2747900909027825, + "grad_norm": 1.8243556022644043, + "learning_rate": 8.269789598517101e-05, + "loss": 0.5381, + "step": 32782 + }, + { + "epoch": 2.2748594823398793, + "grad_norm": 1.3496917486190796, + "learning_rate": 8.269236213625188e-05, + "loss": 0.3968, + "step": 32783 + }, + { + "epoch": 2.274928873776976, + "grad_norm": 1.3165148496627808, + "learning_rate": 8.268682834197093e-05, + "loss": 0.4692, + "step": 32784 + }, + { + "epoch": 2.2749982652140726, + "grad_norm": 1.2567970752716064, + "learning_rate": 8.268129460234567e-05, + "loss": 0.3457, + "step": 32785 + }, + { + "epoch": 2.2750676566511694, + "grad_norm": 1.6919411420822144, + "learning_rate": 8.267576091739355e-05, + "loss": 0.4488, + "step": 32786 + }, + { + "epoch": 2.275137048088266, + "grad_norm": 1.414210557937622, + "learning_rate": 8.267022728713209e-05, + "loss": 0.4853, + "step": 32787 + }, + { + "epoch": 2.2752064395253626, + "grad_norm": 1.6671578884124756, + "learning_rate": 8.266469371157865e-05, + "loss": 0.439, + "step": 32788 + }, + { + "epoch": 2.2752758309624594, + "grad_norm": 1.5676811933517456, + "learning_rate": 8.265916019075077e-05, + "loss": 0.487, + "step": 32789 + }, + { + "epoch": 2.275345222399556, + "grad_norm": 1.5336023569107056, + "learning_rate": 8.265362672466595e-05, + "loss": 0.5258, + "step": 32790 + }, + { + "epoch": 2.2754146138366527, + "grad_norm": 1.5536025762557983, + "learning_rate": 8.264809331334158e-05, + "loss": 0.5561, + "step": 32791 + }, + { + "epoch": 2.275484005273749, + "grad_norm": 1.0515848398208618, + "learning_rate": 8.264255995679516e-05, + "loss": 0.4286, + "step": 32792 + }, + { + "epoch": 2.275553396710846, + "grad_norm": 1.5656640529632568, + "learning_rate": 8.263702665504421e-05, + "loss": 0.529, + "step": 32793 + }, + { + "epoch": 2.2756227881479427, + "grad_norm": 1.2475754022598267, + "learning_rate": 8.263149340810612e-05, + "loss": 0.4444, + "step": 32794 + }, + { + "epoch": 2.275692179585039, + "grad_norm": 0.9949484467506409, + "learning_rate": 8.262596021599839e-05, + "loss": 0.3223, + "step": 32795 + }, + { + "epoch": 2.275761571022136, + "grad_norm": 0.9110404253005981, + "learning_rate": 8.262042707873848e-05, + "loss": 0.307, + "step": 32796 + }, + { + "epoch": 2.2758309624592323, + "grad_norm": 1.1465047597885132, + "learning_rate": 8.261489399634386e-05, + "loss": 0.4827, + "step": 32797 + }, + { + "epoch": 2.275900353896329, + "grad_norm": 1.4686157703399658, + "learning_rate": 8.260936096883202e-05, + "loss": 0.4566, + "step": 32798 + }, + { + "epoch": 2.275969745333426, + "grad_norm": 1.2164077758789062, + "learning_rate": 8.26038279962204e-05, + "loss": 0.4251, + "step": 32799 + }, + { + "epoch": 2.2760391367705224, + "grad_norm": 1.8772724866867065, + "learning_rate": 8.259829507852653e-05, + "loss": 0.4577, + "step": 32800 + }, + { + "epoch": 2.276108528207619, + "grad_norm": 1.013710618019104, + "learning_rate": 8.259276221576775e-05, + "loss": 0.4197, + "step": 32801 + }, + { + "epoch": 2.2761779196447156, + "grad_norm": 1.5193655490875244, + "learning_rate": 8.258722940796162e-05, + "loss": 0.4184, + "step": 32802 + }, + { + "epoch": 2.2762473110818124, + "grad_norm": 1.2893542051315308, + "learning_rate": 8.25816966551256e-05, + "loss": 0.5196, + "step": 32803 + }, + { + "epoch": 2.2763167025189093, + "grad_norm": 1.8998147249221802, + "learning_rate": 8.257616395727712e-05, + "loss": 0.5064, + "step": 32804 + }, + { + "epoch": 2.2763860939560057, + "grad_norm": 1.4260504245758057, + "learning_rate": 8.257063131443368e-05, + "loss": 0.4266, + "step": 32805 + }, + { + "epoch": 2.2764554853931025, + "grad_norm": 1.0007845163345337, + "learning_rate": 8.256509872661276e-05, + "loss": 0.2317, + "step": 32806 + }, + { + "epoch": 2.2765248768301993, + "grad_norm": 1.2601910829544067, + "learning_rate": 8.255956619383179e-05, + "loss": 0.3397, + "step": 32807 + }, + { + "epoch": 2.2765942682672957, + "grad_norm": 1.1410644054412842, + "learning_rate": 8.255403371610823e-05, + "loss": 0.3406, + "step": 32808 + }, + { + "epoch": 2.2766636597043926, + "grad_norm": 0.8694874048233032, + "learning_rate": 8.254850129345955e-05, + "loss": 0.4124, + "step": 32809 + }, + { + "epoch": 2.2767330511414894, + "grad_norm": 1.3232238292694092, + "learning_rate": 8.254296892590328e-05, + "loss": 0.3817, + "step": 32810 + }, + { + "epoch": 2.2768024425785858, + "grad_norm": 1.3339930772781372, + "learning_rate": 8.253743661345677e-05, + "loss": 0.4281, + "step": 32811 + }, + { + "epoch": 2.2768718340156826, + "grad_norm": 1.4494061470031738, + "learning_rate": 8.253190435613758e-05, + "loss": 0.4549, + "step": 32812 + }, + { + "epoch": 2.276941225452779, + "grad_norm": 1.0486871004104614, + "learning_rate": 8.252637215396311e-05, + "loss": 0.2939, + "step": 32813 + }, + { + "epoch": 2.277010616889876, + "grad_norm": 0.7969025373458862, + "learning_rate": 8.252084000695092e-05, + "loss": 0.244, + "step": 32814 + }, + { + "epoch": 2.2770800083269727, + "grad_norm": 1.1454492807388306, + "learning_rate": 8.251530791511836e-05, + "loss": 0.3997, + "step": 32815 + }, + { + "epoch": 2.277149399764069, + "grad_norm": 1.291000247001648, + "learning_rate": 8.250977587848295e-05, + "loss": 0.4243, + "step": 32816 + }, + { + "epoch": 2.277218791201166, + "grad_norm": 1.231034755706787, + "learning_rate": 8.250424389706216e-05, + "loss": 0.2497, + "step": 32817 + }, + { + "epoch": 2.2772881826382623, + "grad_norm": 1.5538599491119385, + "learning_rate": 8.249871197087342e-05, + "loss": 0.5038, + "step": 32818 + }, + { + "epoch": 2.277357574075359, + "grad_norm": 1.6788586378097534, + "learning_rate": 8.249318009993422e-05, + "loss": 0.4958, + "step": 32819 + }, + { + "epoch": 2.277426965512456, + "grad_norm": 1.1180801391601562, + "learning_rate": 8.248764828426203e-05, + "loss": 0.4975, + "step": 32820 + }, + { + "epoch": 2.2774963569495523, + "grad_norm": 1.0964089632034302, + "learning_rate": 8.248211652387434e-05, + "loss": 0.3293, + "step": 32821 + }, + { + "epoch": 2.277565748386649, + "grad_norm": 1.3956084251403809, + "learning_rate": 8.247658481878853e-05, + "loss": 0.5236, + "step": 32822 + }, + { + "epoch": 2.2776351398237455, + "grad_norm": 3.6148223876953125, + "learning_rate": 8.24710531690221e-05, + "loss": 0.3475, + "step": 32823 + }, + { + "epoch": 2.2777045312608424, + "grad_norm": 1.3626761436462402, + "learning_rate": 8.246552157459253e-05, + "loss": 0.3714, + "step": 32824 + }, + { + "epoch": 2.277773922697939, + "grad_norm": 2.8256704807281494, + "learning_rate": 8.24599900355173e-05, + "loss": 0.4756, + "step": 32825 + }, + { + "epoch": 2.2778433141350356, + "grad_norm": 1.080317735671997, + "learning_rate": 8.245445855181382e-05, + "loss": 0.407, + "step": 32826 + }, + { + "epoch": 2.2779127055721324, + "grad_norm": 1.1458953619003296, + "learning_rate": 8.244892712349959e-05, + "loss": 0.4016, + "step": 32827 + }, + { + "epoch": 2.277982097009229, + "grad_norm": 1.0364936590194702, + "learning_rate": 8.244339575059207e-05, + "loss": 0.4131, + "step": 32828 + }, + { + "epoch": 2.2780514884463257, + "grad_norm": 1.209725260734558, + "learning_rate": 8.243786443310871e-05, + "loss": 0.5369, + "step": 32829 + }, + { + "epoch": 2.2781208798834225, + "grad_norm": 1.2768144607543945, + "learning_rate": 8.243233317106696e-05, + "loss": 0.4096, + "step": 32830 + }, + { + "epoch": 2.278190271320519, + "grad_norm": 1.7654244899749756, + "learning_rate": 8.24268019644843e-05, + "loss": 0.4039, + "step": 32831 + }, + { + "epoch": 2.2782596627576157, + "grad_norm": 1.4917964935302734, + "learning_rate": 8.242127081337818e-05, + "loss": 0.6058, + "step": 32832 + }, + { + "epoch": 2.2783290541947125, + "grad_norm": 1.6026062965393066, + "learning_rate": 8.241573971776606e-05, + "loss": 0.5505, + "step": 32833 + }, + { + "epoch": 2.278398445631809, + "grad_norm": 1.334136724472046, + "learning_rate": 8.241020867766542e-05, + "loss": 0.3685, + "step": 32834 + }, + { + "epoch": 2.2784678370689058, + "grad_norm": 1.269948124885559, + "learning_rate": 8.240467769309376e-05, + "loss": 0.6067, + "step": 32835 + }, + { + "epoch": 2.2785372285060026, + "grad_norm": 1.2277525663375854, + "learning_rate": 8.239914676406844e-05, + "loss": 0.4643, + "step": 32836 + }, + { + "epoch": 2.278606619943099, + "grad_norm": 1.1462122201919556, + "learning_rate": 8.239361589060697e-05, + "loss": 0.4266, + "step": 32837 + }, + { + "epoch": 2.278676011380196, + "grad_norm": 1.3697320222854614, + "learning_rate": 8.238808507272682e-05, + "loss": 0.5144, + "step": 32838 + }, + { + "epoch": 2.278745402817292, + "grad_norm": 1.4218794107437134, + "learning_rate": 8.238255431044544e-05, + "loss": 0.4217, + "step": 32839 + }, + { + "epoch": 2.278814794254389, + "grad_norm": 1.2195138931274414, + "learning_rate": 8.237702360378028e-05, + "loss": 0.5401, + "step": 32840 + }, + { + "epoch": 2.278884185691486, + "grad_norm": 1.0313013792037964, + "learning_rate": 8.237149295274884e-05, + "loss": 0.2435, + "step": 32841 + }, + { + "epoch": 2.2789535771285823, + "grad_norm": 1.3334118127822876, + "learning_rate": 8.236596235736855e-05, + "loss": 0.5156, + "step": 32842 + }, + { + "epoch": 2.279022968565679, + "grad_norm": 1.3004860877990723, + "learning_rate": 8.236043181765686e-05, + "loss": 0.4792, + "step": 32843 + }, + { + "epoch": 2.2790923600027755, + "grad_norm": 1.2650822401046753, + "learning_rate": 8.235490133363126e-05, + "loss": 0.3242, + "step": 32844 + }, + { + "epoch": 2.2791617514398723, + "grad_norm": 1.156160593032837, + "learning_rate": 8.234937090530916e-05, + "loss": 0.3985, + "step": 32845 + }, + { + "epoch": 2.279231142876969, + "grad_norm": 1.2403761148452759, + "learning_rate": 8.234384053270805e-05, + "loss": 0.3692, + "step": 32846 + }, + { + "epoch": 2.2793005343140655, + "grad_norm": 1.1983857154846191, + "learning_rate": 8.233831021584539e-05, + "loss": 0.4894, + "step": 32847 + }, + { + "epoch": 2.2793699257511624, + "grad_norm": 1.2008264064788818, + "learning_rate": 8.233277995473862e-05, + "loss": 0.5253, + "step": 32848 + }, + { + "epoch": 2.2794393171882588, + "grad_norm": 1.410534381866455, + "learning_rate": 8.232724974940528e-05, + "loss": 0.5261, + "step": 32849 + }, + { + "epoch": 2.2795087086253556, + "grad_norm": 1.565457820892334, + "learning_rate": 8.232171959986272e-05, + "loss": 0.3835, + "step": 32850 + }, + { + "epoch": 2.2795781000624524, + "grad_norm": 1.2836151123046875, + "learning_rate": 8.231618950612842e-05, + "loss": 0.3925, + "step": 32851 + }, + { + "epoch": 2.279647491499549, + "grad_norm": 1.6532399654388428, + "learning_rate": 8.231065946821986e-05, + "loss": 0.4942, + "step": 32852 + }, + { + "epoch": 2.2797168829366456, + "grad_norm": 1.2749106884002686, + "learning_rate": 8.230512948615451e-05, + "loss": 0.3639, + "step": 32853 + }, + { + "epoch": 2.279786274373742, + "grad_norm": 1.4160780906677246, + "learning_rate": 8.22995995599498e-05, + "loss": 0.5654, + "step": 32854 + }, + { + "epoch": 2.279855665810839, + "grad_norm": 1.2829476594924927, + "learning_rate": 8.229406968962321e-05, + "loss": 0.3733, + "step": 32855 + }, + { + "epoch": 2.2799250572479357, + "grad_norm": 1.3384231328964233, + "learning_rate": 8.228853987519222e-05, + "loss": 0.4501, + "step": 32856 + }, + { + "epoch": 2.279994448685032, + "grad_norm": 1.5381697416305542, + "learning_rate": 8.228301011667423e-05, + "loss": 0.6277, + "step": 32857 + }, + { + "epoch": 2.280063840122129, + "grad_norm": 1.3436661958694458, + "learning_rate": 8.227748041408671e-05, + "loss": 0.4051, + "step": 32858 + }, + { + "epoch": 2.2801332315592258, + "grad_norm": 1.726091742515564, + "learning_rate": 8.227195076744712e-05, + "loss": 0.4254, + "step": 32859 + }, + { + "epoch": 2.280202622996322, + "grad_norm": 1.1880602836608887, + "learning_rate": 8.226642117677297e-05, + "loss": 0.428, + "step": 32860 + }, + { + "epoch": 2.280272014433419, + "grad_norm": 0.9015063643455505, + "learning_rate": 8.226089164208162e-05, + "loss": 0.3016, + "step": 32861 + }, + { + "epoch": 2.280341405870516, + "grad_norm": 1.1730499267578125, + "learning_rate": 8.225536216339059e-05, + "loss": 0.4865, + "step": 32862 + }, + { + "epoch": 2.280410797307612, + "grad_norm": 4.705106258392334, + "learning_rate": 8.224983274071736e-05, + "loss": 0.4773, + "step": 32863 + }, + { + "epoch": 2.280480188744709, + "grad_norm": 1.466644287109375, + "learning_rate": 8.224430337407932e-05, + "loss": 0.3686, + "step": 32864 + }, + { + "epoch": 2.2805495801818054, + "grad_norm": 1.1885321140289307, + "learning_rate": 8.223877406349395e-05, + "loss": 0.4907, + "step": 32865 + }, + { + "epoch": 2.2806189716189023, + "grad_norm": 1.0164833068847656, + "learning_rate": 8.223324480897871e-05, + "loss": 0.2955, + "step": 32866 + }, + { + "epoch": 2.280688363055999, + "grad_norm": 1.063265085220337, + "learning_rate": 8.222771561055105e-05, + "loss": 0.3958, + "step": 32867 + }, + { + "epoch": 2.2807577544930955, + "grad_norm": 1.2910507917404175, + "learning_rate": 8.222218646822843e-05, + "loss": 0.3965, + "step": 32868 + }, + { + "epoch": 2.2808271459301923, + "grad_norm": 1.9890583753585815, + "learning_rate": 8.221665738202836e-05, + "loss": 0.3635, + "step": 32869 + }, + { + "epoch": 2.2808965373672887, + "grad_norm": 1.138466715812683, + "learning_rate": 8.221112835196818e-05, + "loss": 0.4466, + "step": 32870 + }, + { + "epoch": 2.2809659288043855, + "grad_norm": 1.3128688335418701, + "learning_rate": 8.22055993780654e-05, + "loss": 0.5614, + "step": 32871 + }, + { + "epoch": 2.2810353202414824, + "grad_norm": 1.2198331356048584, + "learning_rate": 8.220007046033748e-05, + "loss": 0.472, + "step": 32872 + }, + { + "epoch": 2.2811047116785788, + "grad_norm": 1.2275787591934204, + "learning_rate": 8.219454159880189e-05, + "loss": 0.5526, + "step": 32873 + }, + { + "epoch": 2.2811741031156756, + "grad_norm": 1.660687804222107, + "learning_rate": 8.218901279347606e-05, + "loss": 0.4264, + "step": 32874 + }, + { + "epoch": 2.281243494552772, + "grad_norm": 1.2646092176437378, + "learning_rate": 8.218348404437745e-05, + "loss": 0.4871, + "step": 32875 + }, + { + "epoch": 2.281312885989869, + "grad_norm": 0.9851946830749512, + "learning_rate": 8.217795535152355e-05, + "loss": 0.3232, + "step": 32876 + }, + { + "epoch": 2.2813822774269656, + "grad_norm": 1.2912719249725342, + "learning_rate": 8.217242671493174e-05, + "loss": 0.5214, + "step": 32877 + }, + { + "epoch": 2.281451668864062, + "grad_norm": 1.3808554410934448, + "learning_rate": 8.21668981346195e-05, + "loss": 0.4555, + "step": 32878 + }, + { + "epoch": 2.281521060301159, + "grad_norm": 1.3977233171463013, + "learning_rate": 8.216136961060432e-05, + "loss": 0.5081, + "step": 32879 + }, + { + "epoch": 2.2815904517382553, + "grad_norm": 1.1936533451080322, + "learning_rate": 8.215584114290361e-05, + "loss": 0.5601, + "step": 32880 + }, + { + "epoch": 2.281659843175352, + "grad_norm": 1.0968399047851562, + "learning_rate": 8.215031273153485e-05, + "loss": 0.356, + "step": 32881 + }, + { + "epoch": 2.281729234612449, + "grad_norm": 1.0369224548339844, + "learning_rate": 8.214478437651546e-05, + "loss": 0.4048, + "step": 32882 + }, + { + "epoch": 2.2817986260495453, + "grad_norm": 1.5358408689498901, + "learning_rate": 8.213925607786297e-05, + "loss": 0.5142, + "step": 32883 + }, + { + "epoch": 2.281868017486642, + "grad_norm": 1.2620503902435303, + "learning_rate": 8.213372783559473e-05, + "loss": 0.4197, + "step": 32884 + }, + { + "epoch": 2.281937408923739, + "grad_norm": 1.466855764389038, + "learning_rate": 8.212819964972824e-05, + "loss": 0.58, + "step": 32885 + }, + { + "epoch": 2.2820068003608354, + "grad_norm": 0.9485246539115906, + "learning_rate": 8.212267152028095e-05, + "loss": 0.4299, + "step": 32886 + }, + { + "epoch": 2.282076191797932, + "grad_norm": 1.5145909786224365, + "learning_rate": 8.211714344727031e-05, + "loss": 0.4864, + "step": 32887 + }, + { + "epoch": 2.282145583235029, + "grad_norm": 1.0910693407058716, + "learning_rate": 8.211161543071376e-05, + "loss": 0.3471, + "step": 32888 + }, + { + "epoch": 2.2822149746721254, + "grad_norm": 1.1533018350601196, + "learning_rate": 8.210608747062878e-05, + "loss": 0.4184, + "step": 32889 + }, + { + "epoch": 2.2822843661092223, + "grad_norm": 1.156075358390808, + "learning_rate": 8.210055956703284e-05, + "loss": 0.3963, + "step": 32890 + }, + { + "epoch": 2.2823537575463186, + "grad_norm": 1.5624139308929443, + "learning_rate": 8.209503171994333e-05, + "loss": 0.5498, + "step": 32891 + }, + { + "epoch": 2.2824231489834155, + "grad_norm": 1.1646571159362793, + "learning_rate": 8.208950392937771e-05, + "loss": 0.5319, + "step": 32892 + }, + { + "epoch": 2.2824925404205123, + "grad_norm": 1.0813124179840088, + "learning_rate": 8.208397619535347e-05, + "loss": 0.4346, + "step": 32893 + }, + { + "epoch": 2.2825619318576087, + "grad_norm": 1.322109341621399, + "learning_rate": 8.207844851788801e-05, + "loss": 0.6062, + "step": 32894 + }, + { + "epoch": 2.2826313232947055, + "grad_norm": 1.4114924669265747, + "learning_rate": 8.207292089699884e-05, + "loss": 0.4207, + "step": 32895 + }, + { + "epoch": 2.282700714731802, + "grad_norm": 1.3575220108032227, + "learning_rate": 8.206739333270337e-05, + "loss": 0.4976, + "step": 32896 + }, + { + "epoch": 2.2827701061688987, + "grad_norm": 1.797842264175415, + "learning_rate": 8.206186582501907e-05, + "loss": 0.3859, + "step": 32897 + }, + { + "epoch": 2.2828394976059956, + "grad_norm": 1.3830249309539795, + "learning_rate": 8.205633837396336e-05, + "loss": 0.5944, + "step": 32898 + }, + { + "epoch": 2.282908889043092, + "grad_norm": 1.5288820266723633, + "learning_rate": 8.205081097955372e-05, + "loss": 0.3694, + "step": 32899 + }, + { + "epoch": 2.282978280480189, + "grad_norm": 1.6873148679733276, + "learning_rate": 8.20452836418076e-05, + "loss": 0.3411, + "step": 32900 + }, + { + "epoch": 2.283047671917285, + "grad_norm": 1.2772479057312012, + "learning_rate": 8.20397563607424e-05, + "loss": 0.3507, + "step": 32901 + }, + { + "epoch": 2.283117063354382, + "grad_norm": 1.6026313304901123, + "learning_rate": 8.203422913637561e-05, + "loss": 0.3402, + "step": 32902 + }, + { + "epoch": 2.283186454791479, + "grad_norm": 1.7553050518035889, + "learning_rate": 8.202870196872468e-05, + "loss": 0.4811, + "step": 32903 + }, + { + "epoch": 2.2832558462285752, + "grad_norm": 1.1511476039886475, + "learning_rate": 8.202317485780709e-05, + "loss": 0.3691, + "step": 32904 + }, + { + "epoch": 2.283325237665672, + "grad_norm": 1.1672427654266357, + "learning_rate": 8.201764780364022e-05, + "loss": 0.3963, + "step": 32905 + }, + { + "epoch": 2.2833946291027685, + "grad_norm": 1.3973617553710938, + "learning_rate": 8.201212080624154e-05, + "loss": 0.4223, + "step": 32906 + }, + { + "epoch": 2.2834640205398653, + "grad_norm": 1.3791471719741821, + "learning_rate": 8.200659386562851e-05, + "loss": 0.3951, + "step": 32907 + }, + { + "epoch": 2.283533411976962, + "grad_norm": 1.3222912549972534, + "learning_rate": 8.200106698181859e-05, + "loss": 0.353, + "step": 32908 + }, + { + "epoch": 2.2836028034140585, + "grad_norm": 1.0665254592895508, + "learning_rate": 8.19955401548292e-05, + "loss": 0.4608, + "step": 32909 + }, + { + "epoch": 2.2836721948511554, + "grad_norm": 1.3723455667495728, + "learning_rate": 8.19900133846778e-05, + "loss": 0.5436, + "step": 32910 + }, + { + "epoch": 2.283741586288252, + "grad_norm": 1.4105441570281982, + "learning_rate": 8.198448667138187e-05, + "loss": 0.3335, + "step": 32911 + }, + { + "epoch": 2.2838109777253486, + "grad_norm": 1.3946739435195923, + "learning_rate": 8.19789600149588e-05, + "loss": 0.4927, + "step": 32912 + }, + { + "epoch": 2.2838803691624454, + "grad_norm": 1.365586519241333, + "learning_rate": 8.197343341542606e-05, + "loss": 0.3286, + "step": 32913 + }, + { + "epoch": 2.2839497605995422, + "grad_norm": 1.8574472665786743, + "learning_rate": 8.196790687280112e-05, + "loss": 0.3694, + "step": 32914 + }, + { + "epoch": 2.2840191520366386, + "grad_norm": 1.0982681512832642, + "learning_rate": 8.196238038710137e-05, + "loss": 0.3514, + "step": 32915 + }, + { + "epoch": 2.2840885434737355, + "grad_norm": 1.9784841537475586, + "learning_rate": 8.195685395834432e-05, + "loss": 0.4012, + "step": 32916 + }, + { + "epoch": 2.284157934910832, + "grad_norm": 1.2850515842437744, + "learning_rate": 8.195132758654737e-05, + "loss": 0.4997, + "step": 32917 + }, + { + "epoch": 2.2842273263479287, + "grad_norm": 1.8174002170562744, + "learning_rate": 8.194580127172803e-05, + "loss": 0.5215, + "step": 32918 + }, + { + "epoch": 2.2842967177850255, + "grad_norm": 1.3981186151504517, + "learning_rate": 8.194027501390365e-05, + "loss": 0.3983, + "step": 32919 + }, + { + "epoch": 2.284366109222122, + "grad_norm": 1.1744340658187866, + "learning_rate": 8.193474881309174e-05, + "loss": 0.3024, + "step": 32920 + }, + { + "epoch": 2.2844355006592187, + "grad_norm": 1.3180643320083618, + "learning_rate": 8.192922266930973e-05, + "loss": 0.4998, + "step": 32921 + }, + { + "epoch": 2.284504892096315, + "grad_norm": 1.9870129823684692, + "learning_rate": 8.192369658257506e-05, + "loss": 0.4851, + "step": 32922 + }, + { + "epoch": 2.284574283533412, + "grad_norm": 1.162945032119751, + "learning_rate": 8.191817055290518e-05, + "loss": 0.4179, + "step": 32923 + }, + { + "epoch": 2.284643674970509, + "grad_norm": 0.9059111475944519, + "learning_rate": 8.191264458031756e-05, + "loss": 0.3616, + "step": 32924 + }, + { + "epoch": 2.284713066407605, + "grad_norm": 1.4080777168273926, + "learning_rate": 8.190711866482964e-05, + "loss": 0.3492, + "step": 32925 + }, + { + "epoch": 2.284782457844702, + "grad_norm": 1.0117024183273315, + "learning_rate": 8.190159280645882e-05, + "loss": 0.3072, + "step": 32926 + }, + { + "epoch": 2.2848518492817984, + "grad_norm": 1.4492310285568237, + "learning_rate": 8.189606700522257e-05, + "loss": 0.4236, + "step": 32927 + }, + { + "epoch": 2.2849212407188952, + "grad_norm": 1.6408216953277588, + "learning_rate": 8.189054126113833e-05, + "loss": 0.4867, + "step": 32928 + }, + { + "epoch": 2.284990632155992, + "grad_norm": 0.8383097052574158, + "learning_rate": 8.188501557422354e-05, + "loss": 0.2219, + "step": 32929 + }, + { + "epoch": 2.2850600235930885, + "grad_norm": 1.5460624694824219, + "learning_rate": 8.18794899444957e-05, + "loss": 0.387, + "step": 32930 + }, + { + "epoch": 2.2851294150301853, + "grad_norm": 1.2786943912506104, + "learning_rate": 8.187396437197216e-05, + "loss": 0.4154, + "step": 32931 + }, + { + "epoch": 2.285198806467282, + "grad_norm": 1.1086325645446777, + "learning_rate": 8.186843885667046e-05, + "loss": 0.3811, + "step": 32932 + }, + { + "epoch": 2.2852681979043785, + "grad_norm": 1.73335862159729, + "learning_rate": 8.1862913398608e-05, + "loss": 0.5239, + "step": 32933 + }, + { + "epoch": 2.2853375893414754, + "grad_norm": 1.0053470134735107, + "learning_rate": 8.185738799780215e-05, + "loss": 0.3885, + "step": 32934 + }, + { + "epoch": 2.2854069807785717, + "grad_norm": 1.7570115327835083, + "learning_rate": 8.185186265427046e-05, + "loss": 0.5604, + "step": 32935 + }, + { + "epoch": 2.2854763722156686, + "grad_norm": 1.5685464143753052, + "learning_rate": 8.184633736803032e-05, + "loss": 0.5061, + "step": 32936 + }, + { + "epoch": 2.2855457636527654, + "grad_norm": 1.0309178829193115, + "learning_rate": 8.18408121390992e-05, + "loss": 0.4757, + "step": 32937 + }, + { + "epoch": 2.285615155089862, + "grad_norm": 1.4221376180648804, + "learning_rate": 8.183528696749452e-05, + "loss": 0.508, + "step": 32938 + }, + { + "epoch": 2.2856845465269586, + "grad_norm": 1.1293363571166992, + "learning_rate": 8.182976185323377e-05, + "loss": 0.428, + "step": 32939 + }, + { + "epoch": 2.2857539379640555, + "grad_norm": 1.0300147533416748, + "learning_rate": 8.182423679633433e-05, + "loss": 0.3802, + "step": 32940 + }, + { + "epoch": 2.285823329401152, + "grad_norm": 1.1767199039459229, + "learning_rate": 8.181871179681364e-05, + "loss": 0.3137, + "step": 32941 + }, + { + "epoch": 2.2858927208382487, + "grad_norm": 2.0808990001678467, + "learning_rate": 8.181318685468918e-05, + "loss": 0.5409, + "step": 32942 + }, + { + "epoch": 2.285962112275345, + "grad_norm": 1.177531123161316, + "learning_rate": 8.180766196997838e-05, + "loss": 0.456, + "step": 32943 + }, + { + "epoch": 2.286031503712442, + "grad_norm": 1.915691614151001, + "learning_rate": 8.180213714269866e-05, + "loss": 0.3848, + "step": 32944 + }, + { + "epoch": 2.2861008951495387, + "grad_norm": 1.206481695175171, + "learning_rate": 8.179661237286756e-05, + "loss": 0.3011, + "step": 32945 + }, + { + "epoch": 2.286170286586635, + "grad_norm": 1.7936104536056519, + "learning_rate": 8.179108766050237e-05, + "loss": 0.4161, + "step": 32946 + }, + { + "epoch": 2.286239678023732, + "grad_norm": 1.115612506866455, + "learning_rate": 8.178556300562063e-05, + "loss": 0.4432, + "step": 32947 + }, + { + "epoch": 2.2863090694608283, + "grad_norm": 0.8815197348594666, + "learning_rate": 8.178003840823973e-05, + "loss": 0.2261, + "step": 32948 + }, + { + "epoch": 2.286378460897925, + "grad_norm": 1.165421485900879, + "learning_rate": 8.177451386837716e-05, + "loss": 0.4131, + "step": 32949 + }, + { + "epoch": 2.286447852335022, + "grad_norm": 0.9014006853103638, + "learning_rate": 8.176898938605034e-05, + "loss": 0.3297, + "step": 32950 + }, + { + "epoch": 2.2865172437721184, + "grad_norm": 1.4893425703048706, + "learning_rate": 8.176346496127667e-05, + "loss": 0.4076, + "step": 32951 + }, + { + "epoch": 2.2865866352092152, + "grad_norm": 1.7640413045883179, + "learning_rate": 8.175794059407367e-05, + "loss": 0.4848, + "step": 32952 + }, + { + "epoch": 2.2866560266463116, + "grad_norm": 1.8788011074066162, + "learning_rate": 8.175241628445873e-05, + "loss": 0.5518, + "step": 32953 + }, + { + "epoch": 2.2867254180834085, + "grad_norm": 1.188719630241394, + "learning_rate": 8.174689203244926e-05, + "loss": 0.374, + "step": 32954 + }, + { + "epoch": 2.2867948095205053, + "grad_norm": 1.9370207786560059, + "learning_rate": 8.174136783806275e-05, + "loss": 0.5329, + "step": 32955 + }, + { + "epoch": 2.2868642009576017, + "grad_norm": 1.3818409442901611, + "learning_rate": 8.173584370131662e-05, + "loss": 0.4918, + "step": 32956 + }, + { + "epoch": 2.2869335923946985, + "grad_norm": 0.9533323049545288, + "learning_rate": 8.173031962222831e-05, + "loss": 0.3425, + "step": 32957 + }, + { + "epoch": 2.2870029838317953, + "grad_norm": 1.5090677738189697, + "learning_rate": 8.172479560081527e-05, + "loss": 0.3413, + "step": 32958 + }, + { + "epoch": 2.2870723752688917, + "grad_norm": 1.6263943910598755, + "learning_rate": 8.171927163709495e-05, + "loss": 0.4614, + "step": 32959 + }, + { + "epoch": 2.2871417667059886, + "grad_norm": 1.4226305484771729, + "learning_rate": 8.171374773108474e-05, + "loss": 0.3592, + "step": 32960 + }, + { + "epoch": 2.287211158143085, + "grad_norm": 1.3462806940078735, + "learning_rate": 8.170822388280212e-05, + "loss": 0.5541, + "step": 32961 + }, + { + "epoch": 2.287280549580182, + "grad_norm": 1.5181046724319458, + "learning_rate": 8.17027000922645e-05, + "loss": 0.4301, + "step": 32962 + }, + { + "epoch": 2.2873499410172786, + "grad_norm": 1.2441110610961914, + "learning_rate": 8.169717635948932e-05, + "loss": 0.4938, + "step": 32963 + }, + { + "epoch": 2.287419332454375, + "grad_norm": 1.3152740001678467, + "learning_rate": 8.169165268449406e-05, + "loss": 0.5933, + "step": 32964 + }, + { + "epoch": 2.287488723891472, + "grad_norm": 2.664618492126465, + "learning_rate": 8.168612906729613e-05, + "loss": 0.553, + "step": 32965 + }, + { + "epoch": 2.2875581153285687, + "grad_norm": 1.150407075881958, + "learning_rate": 8.168060550791297e-05, + "loss": 0.3923, + "step": 32966 + }, + { + "epoch": 2.287627506765665, + "grad_norm": 1.4839420318603516, + "learning_rate": 8.1675082006362e-05, + "loss": 0.4431, + "step": 32967 + }, + { + "epoch": 2.287696898202762, + "grad_norm": 1.569788932800293, + "learning_rate": 8.166955856266068e-05, + "loss": 0.5852, + "step": 32968 + }, + { + "epoch": 2.2877662896398583, + "grad_norm": 1.4191077947616577, + "learning_rate": 8.166403517682641e-05, + "loss": 0.4443, + "step": 32969 + }, + { + "epoch": 2.287835681076955, + "grad_norm": 1.6055538654327393, + "learning_rate": 8.165851184887667e-05, + "loss": 0.4886, + "step": 32970 + }, + { + "epoch": 2.287905072514052, + "grad_norm": 1.767357587814331, + "learning_rate": 8.165298857882887e-05, + "loss": 0.6894, + "step": 32971 + }, + { + "epoch": 2.2879744639511483, + "grad_norm": 2.190980911254883, + "learning_rate": 8.164746536670046e-05, + "loss": 0.6713, + "step": 32972 + }, + { + "epoch": 2.288043855388245, + "grad_norm": 1.5796092748641968, + "learning_rate": 8.164194221250891e-05, + "loss": 0.5774, + "step": 32973 + }, + { + "epoch": 2.2881132468253416, + "grad_norm": 1.089534044265747, + "learning_rate": 8.163641911627158e-05, + "loss": 0.2336, + "step": 32974 + }, + { + "epoch": 2.2881826382624384, + "grad_norm": 1.8445641994476318, + "learning_rate": 8.163089607800596e-05, + "loss": 0.4197, + "step": 32975 + }, + { + "epoch": 2.2882520296995352, + "grad_norm": 1.2554042339324951, + "learning_rate": 8.162537309772945e-05, + "loss": 0.5496, + "step": 32976 + }, + { + "epoch": 2.2883214211366316, + "grad_norm": 1.461846947669983, + "learning_rate": 8.161985017545951e-05, + "loss": 0.4202, + "step": 32977 + }, + { + "epoch": 2.2883908125737285, + "grad_norm": 1.1780710220336914, + "learning_rate": 8.161432731121357e-05, + "loss": 0.337, + "step": 32978 + }, + { + "epoch": 2.288460204010825, + "grad_norm": 1.5512713193893433, + "learning_rate": 8.160880450500906e-05, + "loss": 0.7284, + "step": 32979 + }, + { + "epoch": 2.2885295954479217, + "grad_norm": 1.3659523725509644, + "learning_rate": 8.160328175686347e-05, + "loss": 0.3449, + "step": 32980 + }, + { + "epoch": 2.2885989868850185, + "grad_norm": 1.243860125541687, + "learning_rate": 8.159775906679415e-05, + "loss": 0.525, + "step": 32981 + }, + { + "epoch": 2.288668378322115, + "grad_norm": 0.9399163722991943, + "learning_rate": 8.159223643481856e-05, + "loss": 0.2799, + "step": 32982 + }, + { + "epoch": 2.2887377697592117, + "grad_norm": 1.026793360710144, + "learning_rate": 8.158671386095414e-05, + "loss": 0.5109, + "step": 32983 + }, + { + "epoch": 2.2888071611963086, + "grad_norm": 1.140348196029663, + "learning_rate": 8.158119134521836e-05, + "loss": 0.3528, + "step": 32984 + }, + { + "epoch": 2.288876552633405, + "grad_norm": 1.1905425786972046, + "learning_rate": 8.15756688876286e-05, + "loss": 0.5626, + "step": 32985 + }, + { + "epoch": 2.288945944070502, + "grad_norm": 1.4339040517807007, + "learning_rate": 8.15701464882023e-05, + "loss": 0.3832, + "step": 32986 + }, + { + "epoch": 2.289015335507598, + "grad_norm": 1.3398975133895874, + "learning_rate": 8.156462414695697e-05, + "loss": 0.4316, + "step": 32987 + }, + { + "epoch": 2.289084726944695, + "grad_norm": 1.8168219327926636, + "learning_rate": 8.155910186390994e-05, + "loss": 0.4856, + "step": 32988 + }, + { + "epoch": 2.289154118381792, + "grad_norm": 1.7203917503356934, + "learning_rate": 8.155357963907867e-05, + "loss": 0.4818, + "step": 32989 + }, + { + "epoch": 2.2892235098188882, + "grad_norm": 1.219261884689331, + "learning_rate": 8.154805747248062e-05, + "loss": 0.5261, + "step": 32990 + }, + { + "epoch": 2.289292901255985, + "grad_norm": 1.7646758556365967, + "learning_rate": 8.15425353641332e-05, + "loss": 0.4971, + "step": 32991 + }, + { + "epoch": 2.289362292693082, + "grad_norm": 1.0765495300292969, + "learning_rate": 8.153701331405387e-05, + "loss": 0.3878, + "step": 32992 + }, + { + "epoch": 2.2894316841301783, + "grad_norm": 0.9948058724403381, + "learning_rate": 8.153149132226003e-05, + "loss": 0.3366, + "step": 32993 + }, + { + "epoch": 2.289501075567275, + "grad_norm": 1.3352034091949463, + "learning_rate": 8.15259693887692e-05, + "loss": 0.438, + "step": 32994 + }, + { + "epoch": 2.2895704670043715, + "grad_norm": 1.109614372253418, + "learning_rate": 8.152044751359868e-05, + "loss": 0.3583, + "step": 32995 + }, + { + "epoch": 2.2896398584414683, + "grad_norm": 1.181917428970337, + "learning_rate": 8.151492569676595e-05, + "loss": 0.3888, + "step": 32996 + }, + { + "epoch": 2.289709249878565, + "grad_norm": 1.3898191452026367, + "learning_rate": 8.150940393828848e-05, + "loss": 0.3935, + "step": 32997 + }, + { + "epoch": 2.2897786413156616, + "grad_norm": 1.5437712669372559, + "learning_rate": 8.150388223818365e-05, + "loss": 0.515, + "step": 32998 + }, + { + "epoch": 2.2898480327527584, + "grad_norm": 1.5731381177902222, + "learning_rate": 8.149836059646897e-05, + "loss": 0.5037, + "step": 32999 + }, + { + "epoch": 2.2899174241898548, + "grad_norm": 1.3758586645126343, + "learning_rate": 8.149283901316179e-05, + "loss": 0.4887, + "step": 33000 + }, + { + "epoch": 2.2899868156269516, + "grad_norm": 1.5640666484832764, + "learning_rate": 8.148731748827958e-05, + "loss": 0.4572, + "step": 33001 + }, + { + "epoch": 2.2900562070640484, + "grad_norm": 1.204128623008728, + "learning_rate": 8.148179602183974e-05, + "loss": 0.4758, + "step": 33002 + }, + { + "epoch": 2.290125598501145, + "grad_norm": 1.156134843826294, + "learning_rate": 8.147627461385973e-05, + "loss": 0.3628, + "step": 33003 + }, + { + "epoch": 2.2901949899382417, + "grad_norm": 1.856318712234497, + "learning_rate": 8.147075326435698e-05, + "loss": 0.4501, + "step": 33004 + }, + { + "epoch": 2.290264381375338, + "grad_norm": 1.5481964349746704, + "learning_rate": 8.14652319733489e-05, + "loss": 0.3995, + "step": 33005 + }, + { + "epoch": 2.290333772812435, + "grad_norm": 0.9595068693161011, + "learning_rate": 8.145971074085292e-05, + "loss": 0.3607, + "step": 33006 + }, + { + "epoch": 2.2904031642495317, + "grad_norm": 1.49338698387146, + "learning_rate": 8.14541895668865e-05, + "loss": 0.4298, + "step": 33007 + }, + { + "epoch": 2.290472555686628, + "grad_norm": 1.331892490386963, + "learning_rate": 8.14486684514671e-05, + "loss": 0.5719, + "step": 33008 + }, + { + "epoch": 2.290541947123725, + "grad_norm": 1.5670220851898193, + "learning_rate": 8.144314739461203e-05, + "loss": 0.5714, + "step": 33009 + }, + { + "epoch": 2.2906113385608218, + "grad_norm": 0.9881023168563843, + "learning_rate": 8.143762639633881e-05, + "loss": 0.3658, + "step": 33010 + }, + { + "epoch": 2.290680729997918, + "grad_norm": 1.0070043802261353, + "learning_rate": 8.143210545666485e-05, + "loss": 0.4794, + "step": 33011 + }, + { + "epoch": 2.290750121435015, + "grad_norm": 1.744993805885315, + "learning_rate": 8.142658457560757e-05, + "loss": 0.415, + "step": 33012 + }, + { + "epoch": 2.290819512872112, + "grad_norm": 1.628781795501709, + "learning_rate": 8.142106375318442e-05, + "loss": 0.653, + "step": 33013 + }, + { + "epoch": 2.290888904309208, + "grad_norm": 1.2535580396652222, + "learning_rate": 8.141554298941281e-05, + "loss": 0.4564, + "step": 33014 + }, + { + "epoch": 2.290958295746305, + "grad_norm": 1.0461759567260742, + "learning_rate": 8.141002228431022e-05, + "loss": 0.3681, + "step": 33015 + }, + { + "epoch": 2.2910276871834014, + "grad_norm": 1.154977798461914, + "learning_rate": 8.140450163789398e-05, + "loss": 0.3636, + "step": 33016 + }, + { + "epoch": 2.2910970786204983, + "grad_norm": 1.4044862985610962, + "learning_rate": 8.139898105018157e-05, + "loss": 0.3544, + "step": 33017 + }, + { + "epoch": 2.291166470057595, + "grad_norm": 1.1100517511367798, + "learning_rate": 8.139346052119043e-05, + "loss": 0.3588, + "step": 33018 + }, + { + "epoch": 2.2912358614946915, + "grad_norm": 1.1256786584854126, + "learning_rate": 8.1387940050938e-05, + "loss": 0.4809, + "step": 33019 + }, + { + "epoch": 2.2913052529317883, + "grad_norm": 1.130987524986267, + "learning_rate": 8.138241963944166e-05, + "loss": 0.4037, + "step": 33020 + }, + { + "epoch": 2.2913746443688847, + "grad_norm": 1.701377272605896, + "learning_rate": 8.137689928671888e-05, + "loss": 0.5155, + "step": 33021 + }, + { + "epoch": 2.2914440358059815, + "grad_norm": 1.4189860820770264, + "learning_rate": 8.137137899278707e-05, + "loss": 0.4312, + "step": 33022 + }, + { + "epoch": 2.2915134272430784, + "grad_norm": 1.3045403957366943, + "learning_rate": 8.136585875766363e-05, + "loss": 0.3483, + "step": 33023 + }, + { + "epoch": 2.2915828186801748, + "grad_norm": 1.403531551361084, + "learning_rate": 8.136033858136601e-05, + "loss": 0.317, + "step": 33024 + }, + { + "epoch": 2.2916522101172716, + "grad_norm": 1.4219774007797241, + "learning_rate": 8.135481846391164e-05, + "loss": 0.4134, + "step": 33025 + }, + { + "epoch": 2.291721601554368, + "grad_norm": 1.5413737297058105, + "learning_rate": 8.134929840531794e-05, + "loss": 0.814, + "step": 33026 + }, + { + "epoch": 2.291790992991465, + "grad_norm": 1.3965959548950195, + "learning_rate": 8.134377840560234e-05, + "loss": 0.4019, + "step": 33027 + }, + { + "epoch": 2.2918603844285617, + "grad_norm": 1.6693357229232788, + "learning_rate": 8.133825846478232e-05, + "loss": 0.504, + "step": 33028 + }, + { + "epoch": 2.291929775865658, + "grad_norm": 1.2104988098144531, + "learning_rate": 8.133273858287521e-05, + "loss": 0.3452, + "step": 33029 + }, + { + "epoch": 2.291999167302755, + "grad_norm": 1.879477620124817, + "learning_rate": 8.132721875989848e-05, + "loss": 0.6706, + "step": 33030 + }, + { + "epoch": 2.2920685587398513, + "grad_norm": 1.4229916334152222, + "learning_rate": 8.132169899586954e-05, + "loss": 0.4932, + "step": 33031 + }, + { + "epoch": 2.292137950176948, + "grad_norm": 1.4944841861724854, + "learning_rate": 8.131617929080585e-05, + "loss": 0.5023, + "step": 33032 + }, + { + "epoch": 2.292207341614045, + "grad_norm": 1.8543078899383545, + "learning_rate": 8.13106596447248e-05, + "loss": 0.561, + "step": 33033 + }, + { + "epoch": 2.2922767330511413, + "grad_norm": 1.1303784847259521, + "learning_rate": 8.130514005764386e-05, + "loss": 0.3801, + "step": 33034 + }, + { + "epoch": 2.292346124488238, + "grad_norm": 1.3267457485198975, + "learning_rate": 8.12996205295804e-05, + "loss": 0.5226, + "step": 33035 + }, + { + "epoch": 2.292415515925335, + "grad_norm": 1.2774159908294678, + "learning_rate": 8.129410106055188e-05, + "loss": 0.3024, + "step": 33036 + }, + { + "epoch": 2.2924849073624314, + "grad_norm": 1.5634769201278687, + "learning_rate": 8.128858165057568e-05, + "loss": 0.3388, + "step": 33037 + }, + { + "epoch": 2.292554298799528, + "grad_norm": 1.1437582969665527, + "learning_rate": 8.128306229966932e-05, + "loss": 0.3514, + "step": 33038 + }, + { + "epoch": 2.292623690236625, + "grad_norm": 1.115695595741272, + "learning_rate": 8.127754300785011e-05, + "loss": 0.4344, + "step": 33039 + }, + { + "epoch": 2.2926930816737214, + "grad_norm": 1.13033127784729, + "learning_rate": 8.127202377513552e-05, + "loss": 0.3889, + "step": 33040 + }, + { + "epoch": 2.2927624731108183, + "grad_norm": 1.3106478452682495, + "learning_rate": 8.1266504601543e-05, + "loss": 0.5714, + "step": 33041 + }, + { + "epoch": 2.2928318645479147, + "grad_norm": 1.539149522781372, + "learning_rate": 8.126098548708998e-05, + "loss": 0.513, + "step": 33042 + }, + { + "epoch": 2.2929012559850115, + "grad_norm": 1.4143340587615967, + "learning_rate": 8.12554664317938e-05, + "loss": 0.4151, + "step": 33043 + }, + { + "epoch": 2.2929706474221083, + "grad_norm": 1.0716365575790405, + "learning_rate": 8.124994743567196e-05, + "loss": 0.4525, + "step": 33044 + }, + { + "epoch": 2.2930400388592047, + "grad_norm": 1.8069655895233154, + "learning_rate": 8.124442849874186e-05, + "loss": 0.5994, + "step": 33045 + }, + { + "epoch": 2.2931094302963015, + "grad_norm": 1.3077778816223145, + "learning_rate": 8.123890962102089e-05, + "loss": 0.4532, + "step": 33046 + }, + { + "epoch": 2.293178821733398, + "grad_norm": 0.9844799637794495, + "learning_rate": 8.123339080252654e-05, + "loss": 0.3366, + "step": 33047 + }, + { + "epoch": 2.2932482131704948, + "grad_norm": 1.359929084777832, + "learning_rate": 8.12278720432762e-05, + "loss": 0.4232, + "step": 33048 + }, + { + "epoch": 2.2933176046075916, + "grad_norm": 1.7967411279678345, + "learning_rate": 8.122235334328732e-05, + "loss": 0.5586, + "step": 33049 + }, + { + "epoch": 2.293386996044688, + "grad_norm": 1.3672504425048828, + "learning_rate": 8.121683470257724e-05, + "loss": 0.5202, + "step": 33050 + }, + { + "epoch": 2.293456387481785, + "grad_norm": 1.1162523031234741, + "learning_rate": 8.121131612116345e-05, + "loss": 0.4038, + "step": 33051 + }, + { + "epoch": 2.293525778918881, + "grad_norm": 1.4311797618865967, + "learning_rate": 8.120579759906335e-05, + "loss": 0.6386, + "step": 33052 + }, + { + "epoch": 2.293595170355978, + "grad_norm": 1.3494470119476318, + "learning_rate": 8.120027913629437e-05, + "loss": 0.4451, + "step": 33053 + }, + { + "epoch": 2.293664561793075, + "grad_norm": 1.71462881565094, + "learning_rate": 8.119476073287395e-05, + "loss": 0.471, + "step": 33054 + }, + { + "epoch": 2.2937339532301713, + "grad_norm": 1.2545493841171265, + "learning_rate": 8.118924238881945e-05, + "loss": 0.5662, + "step": 33055 + }, + { + "epoch": 2.293803344667268, + "grad_norm": 1.5485060214996338, + "learning_rate": 8.118372410414836e-05, + "loss": 0.5609, + "step": 33056 + }, + { + "epoch": 2.2938727361043645, + "grad_norm": 1.5479542016983032, + "learning_rate": 8.117820587887807e-05, + "loss": 0.5515, + "step": 33057 + }, + { + "epoch": 2.2939421275414613, + "grad_norm": 1.077803611755371, + "learning_rate": 8.117268771302599e-05, + "loss": 0.3346, + "step": 33058 + }, + { + "epoch": 2.294011518978558, + "grad_norm": 1.3356218338012695, + "learning_rate": 8.116716960660954e-05, + "loss": 0.6007, + "step": 33059 + }, + { + "epoch": 2.2940809104156545, + "grad_norm": 1.0784268379211426, + "learning_rate": 8.116165155964615e-05, + "loss": 0.4618, + "step": 33060 + }, + { + "epoch": 2.2941503018527514, + "grad_norm": 1.2012546062469482, + "learning_rate": 8.115613357215324e-05, + "loss": 0.2752, + "step": 33061 + }, + { + "epoch": 2.294219693289848, + "grad_norm": 1.34018075466156, + "learning_rate": 8.115061564414823e-05, + "loss": 0.3677, + "step": 33062 + }, + { + "epoch": 2.2942890847269446, + "grad_norm": 1.1894075870513916, + "learning_rate": 8.114509777564859e-05, + "loss": 0.33, + "step": 33063 + }, + { + "epoch": 2.2943584761640414, + "grad_norm": 1.068364143371582, + "learning_rate": 8.113957996667164e-05, + "loss": 0.4396, + "step": 33064 + }, + { + "epoch": 2.2944278676011383, + "grad_norm": 2.0090744495391846, + "learning_rate": 8.113406221723485e-05, + "loss": 0.5667, + "step": 33065 + }, + { + "epoch": 2.2944972590382346, + "grad_norm": 0.9182846546173096, + "learning_rate": 8.112854452735563e-05, + "loss": 0.2798, + "step": 33066 + }, + { + "epoch": 2.2945666504753315, + "grad_norm": 0.9227228760719299, + "learning_rate": 8.112302689705142e-05, + "loss": 0.3058, + "step": 33067 + }, + { + "epoch": 2.294636041912428, + "grad_norm": 1.2428373098373413, + "learning_rate": 8.111750932633961e-05, + "loss": 0.4002, + "step": 33068 + }, + { + "epoch": 2.2947054333495247, + "grad_norm": 1.6152044534683228, + "learning_rate": 8.111199181523765e-05, + "loss": 0.6004, + "step": 33069 + }, + { + "epoch": 2.2947748247866215, + "grad_norm": 1.4084770679473877, + "learning_rate": 8.110647436376296e-05, + "loss": 0.4332, + "step": 33070 + }, + { + "epoch": 2.294844216223718, + "grad_norm": 1.0990374088287354, + "learning_rate": 8.110095697193291e-05, + "loss": 0.3965, + "step": 33071 + }, + { + "epoch": 2.2949136076608148, + "grad_norm": 1.6748003959655762, + "learning_rate": 8.109543963976494e-05, + "loss": 0.4629, + "step": 33072 + }, + { + "epoch": 2.294982999097911, + "grad_norm": 1.3107373714447021, + "learning_rate": 8.108992236727649e-05, + "loss": 0.5722, + "step": 33073 + }, + { + "epoch": 2.295052390535008, + "grad_norm": 0.9288836717605591, + "learning_rate": 8.108440515448495e-05, + "loss": 0.3064, + "step": 33074 + }, + { + "epoch": 2.295121781972105, + "grad_norm": 1.71344792842865, + "learning_rate": 8.107888800140774e-05, + "loss": 0.5308, + "step": 33075 + }, + { + "epoch": 2.295191173409201, + "grad_norm": 1.2308299541473389, + "learning_rate": 8.10733709080623e-05, + "loss": 0.4161, + "step": 33076 + }, + { + "epoch": 2.295260564846298, + "grad_norm": 1.3781228065490723, + "learning_rate": 8.106785387446606e-05, + "loss": 0.521, + "step": 33077 + }, + { + "epoch": 2.2953299562833944, + "grad_norm": 1.1462996006011963, + "learning_rate": 8.106233690063637e-05, + "loss": 0.4183, + "step": 33078 + }, + { + "epoch": 2.2953993477204913, + "grad_norm": 1.293209195137024, + "learning_rate": 8.105681998659069e-05, + "loss": 0.361, + "step": 33079 + }, + { + "epoch": 2.295468739157588, + "grad_norm": 1.4001888036727905, + "learning_rate": 8.105130313234643e-05, + "loss": 0.4949, + "step": 33080 + }, + { + "epoch": 2.2955381305946845, + "grad_norm": 1.531266689300537, + "learning_rate": 8.1045786337921e-05, + "loss": 0.4303, + "step": 33081 + }, + { + "epoch": 2.2956075220317813, + "grad_norm": 1.1524173021316528, + "learning_rate": 8.104026960333183e-05, + "loss": 0.363, + "step": 33082 + }, + { + "epoch": 2.2956769134688777, + "grad_norm": 1.5482600927352905, + "learning_rate": 8.103475292859633e-05, + "loss": 0.3026, + "step": 33083 + }, + { + "epoch": 2.2957463049059745, + "grad_norm": 1.4753135442733765, + "learning_rate": 8.102923631373196e-05, + "loss": 0.4529, + "step": 33084 + }, + { + "epoch": 2.2958156963430714, + "grad_norm": 1.355789065361023, + "learning_rate": 8.102371975875604e-05, + "loss": 0.6139, + "step": 33085 + }, + { + "epoch": 2.2958850877801678, + "grad_norm": 1.313232183456421, + "learning_rate": 8.101820326368604e-05, + "loss": 0.7167, + "step": 33086 + }, + { + "epoch": 2.2959544792172646, + "grad_norm": 1.3700333833694458, + "learning_rate": 8.101268682853936e-05, + "loss": 0.4517, + "step": 33087 + }, + { + "epoch": 2.2960238706543614, + "grad_norm": 1.5827761888504028, + "learning_rate": 8.100717045333346e-05, + "loss": 0.4163, + "step": 33088 + }, + { + "epoch": 2.296093262091458, + "grad_norm": 1.100582242012024, + "learning_rate": 8.100165413808569e-05, + "loss": 0.3652, + "step": 33089 + }, + { + "epoch": 2.2961626535285546, + "grad_norm": 1.1427583694458008, + "learning_rate": 8.099613788281349e-05, + "loss": 0.3997, + "step": 33090 + }, + { + "epoch": 2.2962320449656515, + "grad_norm": 1.3946630954742432, + "learning_rate": 8.09906216875343e-05, + "loss": 0.2252, + "step": 33091 + }, + { + "epoch": 2.296301436402748, + "grad_norm": 1.230190634727478, + "learning_rate": 8.098510555226553e-05, + "loss": 0.4514, + "step": 33092 + }, + { + "epoch": 2.2963708278398447, + "grad_norm": 1.3058079481124878, + "learning_rate": 8.097958947702454e-05, + "loss": 0.3673, + "step": 33093 + }, + { + "epoch": 2.296440219276941, + "grad_norm": 1.1998980045318604, + "learning_rate": 8.097407346182878e-05, + "loss": 0.3432, + "step": 33094 + }, + { + "epoch": 2.296509610714038, + "grad_norm": 1.5272783041000366, + "learning_rate": 8.096855750669566e-05, + "loss": 0.4766, + "step": 33095 + }, + { + "epoch": 2.2965790021511348, + "grad_norm": 1.2319204807281494, + "learning_rate": 8.09630416116426e-05, + "loss": 0.4301, + "step": 33096 + }, + { + "epoch": 2.296648393588231, + "grad_norm": 1.396483302116394, + "learning_rate": 8.095752577668704e-05, + "loss": 0.2893, + "step": 33097 + }, + { + "epoch": 2.296717785025328, + "grad_norm": 1.360511302947998, + "learning_rate": 8.095201000184635e-05, + "loss": 0.3566, + "step": 33098 + }, + { + "epoch": 2.2967871764624244, + "grad_norm": 0.8920717835426331, + "learning_rate": 8.094649428713793e-05, + "loss": 0.3878, + "step": 33099 + }, + { + "epoch": 2.296856567899521, + "grad_norm": 1.3962578773498535, + "learning_rate": 8.094097863257921e-05, + "loss": 0.2978, + "step": 33100 + }, + { + "epoch": 2.296925959336618, + "grad_norm": 1.4453755617141724, + "learning_rate": 8.093546303818764e-05, + "loss": 0.4813, + "step": 33101 + }, + { + "epoch": 2.2969953507737144, + "grad_norm": 3.2759969234466553, + "learning_rate": 8.092994750398057e-05, + "loss": 0.407, + "step": 33102 + }, + { + "epoch": 2.2970647422108113, + "grad_norm": 1.2161979675292969, + "learning_rate": 8.092443202997545e-05, + "loss": 0.4527, + "step": 33103 + }, + { + "epoch": 2.2971341336479076, + "grad_norm": 1.7808159589767456, + "learning_rate": 8.091891661618973e-05, + "loss": 0.3087, + "step": 33104 + }, + { + "epoch": 2.2972035250850045, + "grad_norm": 1.717086911201477, + "learning_rate": 8.091340126264074e-05, + "loss": 0.5025, + "step": 33105 + }, + { + "epoch": 2.2972729165221013, + "grad_norm": 1.2132102251052856, + "learning_rate": 8.090788596934593e-05, + "loss": 0.3973, + "step": 33106 + }, + { + "epoch": 2.2973423079591977, + "grad_norm": 1.1482256650924683, + "learning_rate": 8.09023707363227e-05, + "loss": 0.3635, + "step": 33107 + }, + { + "epoch": 2.2974116993962945, + "grad_norm": 0.9415845274925232, + "learning_rate": 8.089685556358849e-05, + "loss": 0.2659, + "step": 33108 + }, + { + "epoch": 2.297481090833391, + "grad_norm": 1.0248774290084839, + "learning_rate": 8.089134045116067e-05, + "loss": 0.3179, + "step": 33109 + }, + { + "epoch": 2.2975504822704877, + "grad_norm": 1.0372340679168701, + "learning_rate": 8.088582539905668e-05, + "loss": 0.307, + "step": 33110 + }, + { + "epoch": 2.2976198737075846, + "grad_norm": 1.182482123374939, + "learning_rate": 8.088031040729396e-05, + "loss": 0.3151, + "step": 33111 + }, + { + "epoch": 2.297689265144681, + "grad_norm": 1.5624445676803589, + "learning_rate": 8.087479547588983e-05, + "loss": 0.4402, + "step": 33112 + }, + { + "epoch": 2.297758656581778, + "grad_norm": 1.3053719997406006, + "learning_rate": 8.086928060486176e-05, + "loss": 0.3562, + "step": 33113 + }, + { + "epoch": 2.2978280480188746, + "grad_norm": 1.1968488693237305, + "learning_rate": 8.086376579422716e-05, + "loss": 0.422, + "step": 33114 + }, + { + "epoch": 2.297897439455971, + "grad_norm": 1.5371688604354858, + "learning_rate": 8.085825104400342e-05, + "loss": 0.5028, + "step": 33115 + }, + { + "epoch": 2.297966830893068, + "grad_norm": 1.5042920112609863, + "learning_rate": 8.085273635420796e-05, + "loss": 0.5247, + "step": 33116 + }, + { + "epoch": 2.2980362223301647, + "grad_norm": 1.0554190874099731, + "learning_rate": 8.08472217248582e-05, + "loss": 0.454, + "step": 33117 + }, + { + "epoch": 2.298105613767261, + "grad_norm": 1.5329338312149048, + "learning_rate": 8.084170715597156e-05, + "loss": 0.3572, + "step": 33118 + }, + { + "epoch": 2.298175005204358, + "grad_norm": 1.420216679573059, + "learning_rate": 8.083619264756542e-05, + "loss": 0.468, + "step": 33119 + }, + { + "epoch": 2.2982443966414543, + "grad_norm": 1.019776701927185, + "learning_rate": 8.083067819965716e-05, + "loss": 0.4274, + "step": 33120 + }, + { + "epoch": 2.298313788078551, + "grad_norm": 1.1747348308563232, + "learning_rate": 8.082516381226424e-05, + "loss": 0.3118, + "step": 33121 + }, + { + "epoch": 2.298383179515648, + "grad_norm": 1.3477686643600464, + "learning_rate": 8.081964948540406e-05, + "loss": 0.5405, + "step": 33122 + }, + { + "epoch": 2.2984525709527444, + "grad_norm": 1.2327214479446411, + "learning_rate": 8.081413521909404e-05, + "loss": 0.3745, + "step": 33123 + }, + { + "epoch": 2.298521962389841, + "grad_norm": 1.253246545791626, + "learning_rate": 8.080862101335153e-05, + "loss": 0.4077, + "step": 33124 + }, + { + "epoch": 2.2985913538269376, + "grad_norm": 1.1711368560791016, + "learning_rate": 8.080310686819401e-05, + "loss": 0.5919, + "step": 33125 + }, + { + "epoch": 2.2986607452640344, + "grad_norm": 1.2900797128677368, + "learning_rate": 8.079759278363885e-05, + "loss": 0.363, + "step": 33126 + }, + { + "epoch": 2.2987301367011312, + "grad_norm": 1.4211349487304688, + "learning_rate": 8.079207875970347e-05, + "loss": 0.5614, + "step": 33127 + }, + { + "epoch": 2.2987995281382276, + "grad_norm": 1.4005906581878662, + "learning_rate": 8.078656479640522e-05, + "loss": 0.3231, + "step": 33128 + }, + { + "epoch": 2.2988689195753245, + "grad_norm": 0.9351723194122314, + "learning_rate": 8.07810508937616e-05, + "loss": 0.2964, + "step": 33129 + }, + { + "epoch": 2.298938311012421, + "grad_norm": 1.7570314407348633, + "learning_rate": 8.077553705178995e-05, + "loss": 0.6257, + "step": 33130 + }, + { + "epoch": 2.2990077024495177, + "grad_norm": 1.534782886505127, + "learning_rate": 8.07700232705077e-05, + "loss": 0.5017, + "step": 33131 + }, + { + "epoch": 2.2990770938866145, + "grad_norm": 1.1784725189208984, + "learning_rate": 8.07645095499323e-05, + "loss": 0.4968, + "step": 33132 + }, + { + "epoch": 2.299146485323711, + "grad_norm": 1.1444710493087769, + "learning_rate": 8.075899589008107e-05, + "loss": 0.401, + "step": 33133 + }, + { + "epoch": 2.2992158767608077, + "grad_norm": 1.0824657678604126, + "learning_rate": 8.075348229097146e-05, + "loss": 0.4644, + "step": 33134 + }, + { + "epoch": 2.299285268197904, + "grad_norm": 1.543921947479248, + "learning_rate": 8.074796875262088e-05, + "loss": 0.5451, + "step": 33135 + }, + { + "epoch": 2.299354659635001, + "grad_norm": 1.3507670164108276, + "learning_rate": 8.074245527504671e-05, + "loss": 0.4715, + "step": 33136 + }, + { + "epoch": 2.299424051072098, + "grad_norm": 1.2028898000717163, + "learning_rate": 8.07369418582664e-05, + "loss": 0.5076, + "step": 33137 + }, + { + "epoch": 2.299493442509194, + "grad_norm": 2.4369051456451416, + "learning_rate": 8.073142850229731e-05, + "loss": 0.5738, + "step": 33138 + }, + { + "epoch": 2.299562833946291, + "grad_norm": 1.5287549495697021, + "learning_rate": 8.072591520715691e-05, + "loss": 0.4775, + "step": 33139 + }, + { + "epoch": 2.299632225383388, + "grad_norm": 1.4315768480300903, + "learning_rate": 8.072040197286252e-05, + "loss": 0.7238, + "step": 33140 + }, + { + "epoch": 2.2997016168204842, + "grad_norm": 1.2398115396499634, + "learning_rate": 8.071488879943158e-05, + "loss": 0.5966, + "step": 33141 + }, + { + "epoch": 2.299771008257581, + "grad_norm": 1.44577157497406, + "learning_rate": 8.070937568688152e-05, + "loss": 0.4453, + "step": 33142 + }, + { + "epoch": 2.299840399694678, + "grad_norm": 0.9254537224769592, + "learning_rate": 8.07038626352297e-05, + "loss": 0.2, + "step": 33143 + }, + { + "epoch": 2.2999097911317743, + "grad_norm": 1.3022712469100952, + "learning_rate": 8.069834964449355e-05, + "loss": 0.4113, + "step": 33144 + }, + { + "epoch": 2.299979182568871, + "grad_norm": 1.2777645587921143, + "learning_rate": 8.069283671469047e-05, + "loss": 0.4369, + "step": 33145 + }, + { + "epoch": 2.3000485740059675, + "grad_norm": 1.396544098854065, + "learning_rate": 8.068732384583792e-05, + "loss": 0.5089, + "step": 33146 + }, + { + "epoch": 2.3001179654430643, + "grad_norm": 0.904679000377655, + "learning_rate": 8.068181103795318e-05, + "loss": 0.3241, + "step": 33147 + }, + { + "epoch": 2.300187356880161, + "grad_norm": 1.005483865737915, + "learning_rate": 8.067629829105375e-05, + "loss": 0.4767, + "step": 33148 + }, + { + "epoch": 2.3002567483172576, + "grad_norm": 1.8237230777740479, + "learning_rate": 8.067078560515698e-05, + "loss": 0.3575, + "step": 33149 + }, + { + "epoch": 2.3003261397543544, + "grad_norm": 1.4494585990905762, + "learning_rate": 8.066527298028031e-05, + "loss": 0.5065, + "step": 33150 + }, + { + "epoch": 2.300395531191451, + "grad_norm": 1.097707986831665, + "learning_rate": 8.065976041644112e-05, + "loss": 0.3786, + "step": 33151 + }, + { + "epoch": 2.3004649226285476, + "grad_norm": 1.6908893585205078, + "learning_rate": 8.065424791365684e-05, + "loss": 0.5699, + "step": 33152 + }, + { + "epoch": 2.3005343140656445, + "grad_norm": 1.1684094667434692, + "learning_rate": 8.064873547194488e-05, + "loss": 0.592, + "step": 33153 + }, + { + "epoch": 2.300603705502741, + "grad_norm": 1.3179090023040771, + "learning_rate": 8.064322309132258e-05, + "loss": 0.5617, + "step": 33154 + }, + { + "epoch": 2.3006730969398377, + "grad_norm": 0.8286964893341064, + "learning_rate": 8.063771077180737e-05, + "loss": 0.2867, + "step": 33155 + }, + { + "epoch": 2.300742488376934, + "grad_norm": 1.435470700263977, + "learning_rate": 8.063219851341668e-05, + "loss": 0.546, + "step": 33156 + }, + { + "epoch": 2.300811879814031, + "grad_norm": 1.1211751699447632, + "learning_rate": 8.062668631616788e-05, + "loss": 0.3387, + "step": 33157 + }, + { + "epoch": 2.3008812712511277, + "grad_norm": 1.0177747011184692, + "learning_rate": 8.06211741800784e-05, + "loss": 0.2265, + "step": 33158 + }, + { + "epoch": 2.300950662688224, + "grad_norm": 0.9122565984725952, + "learning_rate": 8.061566210516563e-05, + "loss": 0.296, + "step": 33159 + }, + { + "epoch": 2.301020054125321, + "grad_norm": 1.325350046157837, + "learning_rate": 8.061015009144695e-05, + "loss": 0.3275, + "step": 33160 + }, + { + "epoch": 2.301089445562418, + "grad_norm": 1.6052788496017456, + "learning_rate": 8.060463813893978e-05, + "loss": 0.4616, + "step": 33161 + }, + { + "epoch": 2.301158836999514, + "grad_norm": 1.2237311601638794, + "learning_rate": 8.059912624766153e-05, + "loss": 0.5531, + "step": 33162 + }, + { + "epoch": 2.301228228436611, + "grad_norm": 1.3028966188430786, + "learning_rate": 8.059361441762957e-05, + "loss": 0.589, + "step": 33163 + }, + { + "epoch": 2.3012976198737074, + "grad_norm": 1.5599822998046875, + "learning_rate": 8.058810264886132e-05, + "loss": 0.3779, + "step": 33164 + }, + { + "epoch": 2.3013670113108042, + "grad_norm": 0.8654764294624329, + "learning_rate": 8.058259094137417e-05, + "loss": 0.2051, + "step": 33165 + }, + { + "epoch": 2.301436402747901, + "grad_norm": 1.415923833847046, + "learning_rate": 8.057707929518553e-05, + "loss": 0.3683, + "step": 33166 + }, + { + "epoch": 2.3015057941849975, + "grad_norm": 0.9816314578056335, + "learning_rate": 8.057156771031284e-05, + "loss": 0.3264, + "step": 33167 + }, + { + "epoch": 2.3015751856220943, + "grad_norm": 1.2971625328063965, + "learning_rate": 8.056605618677342e-05, + "loss": 0.3831, + "step": 33168 + }, + { + "epoch": 2.301644577059191, + "grad_norm": 1.4654589891433716, + "learning_rate": 8.05605447245847e-05, + "loss": 0.44, + "step": 33169 + }, + { + "epoch": 2.3017139684962875, + "grad_norm": 1.1545183658599854, + "learning_rate": 8.055503332376408e-05, + "loss": 0.4093, + "step": 33170 + }, + { + "epoch": 2.3017833599333843, + "grad_norm": 1.3563588857650757, + "learning_rate": 8.054952198432897e-05, + "loss": 0.6219, + "step": 33171 + }, + { + "epoch": 2.3018527513704807, + "grad_norm": 1.7118289470672607, + "learning_rate": 8.054401070629676e-05, + "loss": 0.477, + "step": 33172 + }, + { + "epoch": 2.3019221428075776, + "grad_norm": 1.040434718132019, + "learning_rate": 8.053849948968489e-05, + "loss": 0.4582, + "step": 33173 + }, + { + "epoch": 2.3019915342446744, + "grad_norm": 1.379046082496643, + "learning_rate": 8.053298833451069e-05, + "loss": 0.597, + "step": 33174 + }, + { + "epoch": 2.302060925681771, + "grad_norm": 1.6730109453201294, + "learning_rate": 8.052747724079157e-05, + "loss": 0.4698, + "step": 33175 + }, + { + "epoch": 2.3021303171188676, + "grad_norm": 1.6200411319732666, + "learning_rate": 8.052196620854496e-05, + "loss": 0.5391, + "step": 33176 + }, + { + "epoch": 2.302199708555964, + "grad_norm": 1.281982421875, + "learning_rate": 8.051645523778826e-05, + "loss": 0.4603, + "step": 33177 + }, + { + "epoch": 2.302269099993061, + "grad_norm": 2.8766000270843506, + "learning_rate": 8.051094432853883e-05, + "loss": 0.4686, + "step": 33178 + }, + { + "epoch": 2.3023384914301577, + "grad_norm": 1.0540918111801147, + "learning_rate": 8.050543348081407e-05, + "loss": 0.3102, + "step": 33179 + }, + { + "epoch": 2.302407882867254, + "grad_norm": 1.4979804754257202, + "learning_rate": 8.049992269463145e-05, + "loss": 0.4156, + "step": 33180 + }, + { + "epoch": 2.302477274304351, + "grad_norm": 1.535468578338623, + "learning_rate": 8.049441197000828e-05, + "loss": 0.5015, + "step": 33181 + }, + { + "epoch": 2.3025466657414473, + "grad_norm": 2.0040013790130615, + "learning_rate": 8.048890130696199e-05, + "loss": 0.5158, + "step": 33182 + }, + { + "epoch": 2.302616057178544, + "grad_norm": 1.0759930610656738, + "learning_rate": 8.048339070550996e-05, + "loss": 0.2711, + "step": 33183 + }, + { + "epoch": 2.302685448615641, + "grad_norm": 1.619096279144287, + "learning_rate": 8.047788016566959e-05, + "loss": 0.4659, + "step": 33184 + }, + { + "epoch": 2.3027548400527373, + "grad_norm": 1.18079674243927, + "learning_rate": 8.047236968745831e-05, + "loss": 0.3696, + "step": 33185 + }, + { + "epoch": 2.302824231489834, + "grad_norm": 1.7019063234329224, + "learning_rate": 8.046685927089349e-05, + "loss": 0.5974, + "step": 33186 + }, + { + "epoch": 2.302893622926931, + "grad_norm": 1.2103976011276245, + "learning_rate": 8.046134891599258e-05, + "loss": 0.4732, + "step": 33187 + }, + { + "epoch": 2.3029630143640274, + "grad_norm": 2.1980085372924805, + "learning_rate": 8.045583862277286e-05, + "loss": 0.6597, + "step": 33188 + }, + { + "epoch": 2.3030324058011242, + "grad_norm": 0.7764403820037842, + "learning_rate": 8.045032839125179e-05, + "loss": 0.1882, + "step": 33189 + }, + { + "epoch": 2.3031017972382206, + "grad_norm": 1.4479776620864868, + "learning_rate": 8.044481822144678e-05, + "loss": 0.4925, + "step": 33190 + }, + { + "epoch": 2.3031711886753174, + "grad_norm": 1.2373542785644531, + "learning_rate": 8.043930811337521e-05, + "loss": 0.4173, + "step": 33191 + }, + { + "epoch": 2.3032405801124143, + "grad_norm": 0.934198260307312, + "learning_rate": 8.043379806705447e-05, + "loss": 0.3357, + "step": 33192 + }, + { + "epoch": 2.3033099715495107, + "grad_norm": 1.0555808544158936, + "learning_rate": 8.042828808250198e-05, + "loss": 0.4583, + "step": 33193 + }, + { + "epoch": 2.3033793629866075, + "grad_norm": 1.1929199695587158, + "learning_rate": 8.042277815973512e-05, + "loss": 0.434, + "step": 33194 + }, + { + "epoch": 2.3034487544237043, + "grad_norm": 1.098427414894104, + "learning_rate": 8.041726829877125e-05, + "loss": 0.3778, + "step": 33195 + }, + { + "epoch": 2.3035181458608007, + "grad_norm": 1.7804499864578247, + "learning_rate": 8.041175849962778e-05, + "loss": 0.7175, + "step": 33196 + }, + { + "epoch": 2.3035875372978976, + "grad_norm": 1.209958791732788, + "learning_rate": 8.040624876232216e-05, + "loss": 0.5154, + "step": 33197 + }, + { + "epoch": 2.303656928734994, + "grad_norm": 1.0005271434783936, + "learning_rate": 8.04007390868717e-05, + "loss": 0.3607, + "step": 33198 + }, + { + "epoch": 2.303726320172091, + "grad_norm": 1.132245421409607, + "learning_rate": 8.039522947329385e-05, + "loss": 0.3985, + "step": 33199 + }, + { + "epoch": 2.3037957116091876, + "grad_norm": 1.5929265022277832, + "learning_rate": 8.038971992160596e-05, + "loss": 0.377, + "step": 33200 + }, + { + "epoch": 2.303865103046284, + "grad_norm": 1.1330606937408447, + "learning_rate": 8.038421043182551e-05, + "loss": 0.3238, + "step": 33201 + }, + { + "epoch": 2.303934494483381, + "grad_norm": 1.0677984952926636, + "learning_rate": 8.037870100396979e-05, + "loss": 0.4053, + "step": 33202 + }, + { + "epoch": 2.3040038859204772, + "grad_norm": 1.6922056674957275, + "learning_rate": 8.037319163805624e-05, + "loss": 0.5266, + "step": 33203 + }, + { + "epoch": 2.304073277357574, + "grad_norm": 1.2410061359405518, + "learning_rate": 8.036768233410224e-05, + "loss": 0.3958, + "step": 33204 + }, + { + "epoch": 2.304142668794671, + "grad_norm": 1.4646258354187012, + "learning_rate": 8.036217309212519e-05, + "loss": 0.3968, + "step": 33205 + }, + { + "epoch": 2.3042120602317673, + "grad_norm": 1.0245496034622192, + "learning_rate": 8.035666391214247e-05, + "loss": 0.3432, + "step": 33206 + }, + { + "epoch": 2.304281451668864, + "grad_norm": 1.0747441053390503, + "learning_rate": 8.035115479417149e-05, + "loss": 0.3773, + "step": 33207 + }, + { + "epoch": 2.3043508431059605, + "grad_norm": 0.928234875202179, + "learning_rate": 8.034564573822968e-05, + "loss": 0.3007, + "step": 33208 + }, + { + "epoch": 2.3044202345430573, + "grad_norm": 1.1393324136734009, + "learning_rate": 8.034013674433434e-05, + "loss": 0.251, + "step": 33209 + }, + { + "epoch": 2.304489625980154, + "grad_norm": 1.4749555587768555, + "learning_rate": 8.033462781250292e-05, + "loss": 0.5507, + "step": 33210 + }, + { + "epoch": 2.3045590174172506, + "grad_norm": 1.3797435760498047, + "learning_rate": 8.032911894275278e-05, + "loss": 0.4631, + "step": 33211 + }, + { + "epoch": 2.3046284088543474, + "grad_norm": 0.9377273321151733, + "learning_rate": 8.032361013510136e-05, + "loss": 0.4679, + "step": 33212 + }, + { + "epoch": 2.304697800291444, + "grad_norm": 1.2372069358825684, + "learning_rate": 8.031810138956599e-05, + "loss": 0.4311, + "step": 33213 + }, + { + "epoch": 2.3047671917285406, + "grad_norm": 1.3955873250961304, + "learning_rate": 8.03125927061641e-05, + "loss": 0.3836, + "step": 33214 + }, + { + "epoch": 2.3048365831656374, + "grad_norm": 1.0534164905548096, + "learning_rate": 8.030708408491309e-05, + "loss": 0.3336, + "step": 33215 + }, + { + "epoch": 2.304905974602734, + "grad_norm": 1.3366928100585938, + "learning_rate": 8.030157552583032e-05, + "loss": 0.4927, + "step": 33216 + }, + { + "epoch": 2.3049753660398307, + "grad_norm": 1.4131618738174438, + "learning_rate": 8.02960670289332e-05, + "loss": 0.5602, + "step": 33217 + }, + { + "epoch": 2.3050447574769275, + "grad_norm": 1.0139240026474, + "learning_rate": 8.029055859423908e-05, + "loss": 0.308, + "step": 33218 + }, + { + "epoch": 2.305114148914024, + "grad_norm": 1.1653969287872314, + "learning_rate": 8.028505022176541e-05, + "loss": 0.5136, + "step": 33219 + }, + { + "epoch": 2.3051835403511207, + "grad_norm": 1.390238642692566, + "learning_rate": 8.027954191152951e-05, + "loss": 0.4704, + "step": 33220 + }, + { + "epoch": 2.3052529317882176, + "grad_norm": 1.450015664100647, + "learning_rate": 8.027403366354883e-05, + "loss": 0.3954, + "step": 33221 + }, + { + "epoch": 2.305322323225314, + "grad_norm": 1.8471521139144897, + "learning_rate": 8.026852547784079e-05, + "loss": 0.443, + "step": 33222 + }, + { + "epoch": 2.3053917146624108, + "grad_norm": 1.6120136976242065, + "learning_rate": 8.026301735442265e-05, + "loss": 0.4916, + "step": 33223 + }, + { + "epoch": 2.305461106099507, + "grad_norm": 1.1177892684936523, + "learning_rate": 8.025750929331191e-05, + "loss": 0.2722, + "step": 33224 + }, + { + "epoch": 2.305530497536604, + "grad_norm": 1.2302879095077515, + "learning_rate": 8.02520012945259e-05, + "loss": 0.4684, + "step": 33225 + }, + { + "epoch": 2.305599888973701, + "grad_norm": 1.1426101922988892, + "learning_rate": 8.024649335808204e-05, + "loss": 0.4906, + "step": 33226 + }, + { + "epoch": 2.305669280410797, + "grad_norm": 1.5484291315078735, + "learning_rate": 8.02409854839977e-05, + "loss": 0.5196, + "step": 33227 + }, + { + "epoch": 2.305738671847894, + "grad_norm": 2.613450288772583, + "learning_rate": 8.023547767229032e-05, + "loss": 0.4777, + "step": 33228 + }, + { + "epoch": 2.3058080632849904, + "grad_norm": 1.4469354152679443, + "learning_rate": 8.022996992297723e-05, + "loss": 0.3408, + "step": 33229 + }, + { + "epoch": 2.3058774547220873, + "grad_norm": 1.31248140335083, + "learning_rate": 8.022446223607579e-05, + "loss": 0.3953, + "step": 33230 + }, + { + "epoch": 2.305946846159184, + "grad_norm": 0.845220148563385, + "learning_rate": 8.021895461160347e-05, + "loss": 0.2625, + "step": 33231 + }, + { + "epoch": 2.3060162375962805, + "grad_norm": 1.5293251276016235, + "learning_rate": 8.02134470495776e-05, + "loss": 0.4941, + "step": 33232 + }, + { + "epoch": 2.3060856290333773, + "grad_norm": 1.0517297983169556, + "learning_rate": 8.020793955001556e-05, + "loss": 0.5225, + "step": 33233 + }, + { + "epoch": 2.3061550204704737, + "grad_norm": 1.3241257667541504, + "learning_rate": 8.020243211293478e-05, + "loss": 0.5279, + "step": 33234 + }, + { + "epoch": 2.3062244119075705, + "grad_norm": 1.325851321220398, + "learning_rate": 8.019692473835261e-05, + "loss": 0.5203, + "step": 33235 + }, + { + "epoch": 2.3062938033446674, + "grad_norm": 1.808702826499939, + "learning_rate": 8.019141742628651e-05, + "loss": 0.3987, + "step": 33236 + }, + { + "epoch": 2.3063631947817638, + "grad_norm": 0.992084264755249, + "learning_rate": 8.018591017675375e-05, + "loss": 0.3163, + "step": 33237 + }, + { + "epoch": 2.3064325862188606, + "grad_norm": 1.4481756687164307, + "learning_rate": 8.018040298977178e-05, + "loss": 0.3988, + "step": 33238 + }, + { + "epoch": 2.3065019776559574, + "grad_norm": 1.2371474504470825, + "learning_rate": 8.017489586535798e-05, + "loss": 0.4123, + "step": 33239 + }, + { + "epoch": 2.306571369093054, + "grad_norm": 0.8394970297813416, + "learning_rate": 8.016938880352972e-05, + "loss": 0.3619, + "step": 33240 + }, + { + "epoch": 2.3066407605301507, + "grad_norm": 1.2868455648422241, + "learning_rate": 8.01638818043044e-05, + "loss": 0.492, + "step": 33241 + }, + { + "epoch": 2.3067101519672475, + "grad_norm": 1.525496482849121, + "learning_rate": 8.015837486769942e-05, + "loss": 0.4491, + "step": 33242 + }, + { + "epoch": 2.306779543404344, + "grad_norm": 1.9208805561065674, + "learning_rate": 8.015286799373219e-05, + "loss": 0.4197, + "step": 33243 + }, + { + "epoch": 2.3068489348414407, + "grad_norm": 1.0186289548873901, + "learning_rate": 8.014736118242e-05, + "loss": 0.3772, + "step": 33244 + }, + { + "epoch": 2.306918326278537, + "grad_norm": 1.1677463054656982, + "learning_rate": 8.01418544337803e-05, + "loss": 0.4008, + "step": 33245 + }, + { + "epoch": 2.306987717715634, + "grad_norm": 1.3209288120269775, + "learning_rate": 8.013634774783044e-05, + "loss": 0.6284, + "step": 33246 + }, + { + "epoch": 2.3070571091527308, + "grad_norm": 0.8639791011810303, + "learning_rate": 8.013084112458786e-05, + "loss": 0.2965, + "step": 33247 + }, + { + "epoch": 2.307126500589827, + "grad_norm": 1.3124632835388184, + "learning_rate": 8.01253345640699e-05, + "loss": 0.4979, + "step": 33248 + }, + { + "epoch": 2.307195892026924, + "grad_norm": 1.3222166299819946, + "learning_rate": 8.011982806629396e-05, + "loss": 0.4667, + "step": 33249 + }, + { + "epoch": 2.3072652834640204, + "grad_norm": 1.462952971458435, + "learning_rate": 8.011432163127738e-05, + "loss": 0.5519, + "step": 33250 + }, + { + "epoch": 2.307334674901117, + "grad_norm": 0.7985084652900696, + "learning_rate": 8.010881525903763e-05, + "loss": 0.2899, + "step": 33251 + }, + { + "epoch": 2.307404066338214, + "grad_norm": 1.640784502029419, + "learning_rate": 8.0103308949592e-05, + "loss": 0.5486, + "step": 33252 + }, + { + "epoch": 2.3074734577753104, + "grad_norm": 1.346799612045288, + "learning_rate": 8.009780270295792e-05, + "loss": 0.3903, + "step": 33253 + }, + { + "epoch": 2.3075428492124073, + "grad_norm": 1.1281155347824097, + "learning_rate": 8.009229651915278e-05, + "loss": 0.5832, + "step": 33254 + }, + { + "epoch": 2.3076122406495037, + "grad_norm": 7.49189567565918, + "learning_rate": 8.008679039819395e-05, + "loss": 0.4795, + "step": 33255 + }, + { + "epoch": 2.3076816320866005, + "grad_norm": 1.442549467086792, + "learning_rate": 8.008128434009884e-05, + "loss": 0.4264, + "step": 33256 + }, + { + "epoch": 2.3077510235236973, + "grad_norm": 1.4546480178833008, + "learning_rate": 8.007577834488477e-05, + "loss": 0.4648, + "step": 33257 + }, + { + "epoch": 2.3078204149607937, + "grad_norm": 1.7074679136276245, + "learning_rate": 8.007027241256914e-05, + "loss": 0.3791, + "step": 33258 + }, + { + "epoch": 2.3078898063978905, + "grad_norm": 1.5941684246063232, + "learning_rate": 8.006476654316936e-05, + "loss": 0.5585, + "step": 33259 + }, + { + "epoch": 2.307959197834987, + "grad_norm": 1.3674304485321045, + "learning_rate": 8.005926073670282e-05, + "loss": 0.3681, + "step": 33260 + }, + { + "epoch": 2.3080285892720838, + "grad_norm": 0.9971142411231995, + "learning_rate": 8.005375499318685e-05, + "loss": 0.317, + "step": 33261 + }, + { + "epoch": 2.3080979807091806, + "grad_norm": 1.4770114421844482, + "learning_rate": 8.004824931263888e-05, + "loss": 0.5936, + "step": 33262 + }, + { + "epoch": 2.308167372146277, + "grad_norm": 1.4886733293533325, + "learning_rate": 8.004274369507631e-05, + "loss": 0.4834, + "step": 33263 + }, + { + "epoch": 2.308236763583374, + "grad_norm": 1.7292568683624268, + "learning_rate": 8.003723814051644e-05, + "loss": 0.483, + "step": 33264 + }, + { + "epoch": 2.3083061550204707, + "grad_norm": 0.9837579131126404, + "learning_rate": 8.003173264897669e-05, + "loss": 0.362, + "step": 33265 + }, + { + "epoch": 2.308375546457567, + "grad_norm": 1.0625598430633545, + "learning_rate": 8.002622722047446e-05, + "loss": 0.3565, + "step": 33266 + }, + { + "epoch": 2.308444937894664, + "grad_norm": 1.5810562372207642, + "learning_rate": 8.00207218550271e-05, + "loss": 0.386, + "step": 33267 + }, + { + "epoch": 2.3085143293317607, + "grad_norm": 1.437984824180603, + "learning_rate": 8.001521655265199e-05, + "loss": 0.5467, + "step": 33268 + }, + { + "epoch": 2.308583720768857, + "grad_norm": 1.7949895858764648, + "learning_rate": 8.000971131336654e-05, + "loss": 0.5305, + "step": 33269 + }, + { + "epoch": 2.308653112205954, + "grad_norm": 0.9667779803276062, + "learning_rate": 8.000420613718816e-05, + "loss": 0.4059, + "step": 33270 + }, + { + "epoch": 2.3087225036430503, + "grad_norm": 1.3031001091003418, + "learning_rate": 7.999870102413412e-05, + "loss": 0.3562, + "step": 33271 + }, + { + "epoch": 2.308791895080147, + "grad_norm": 1.4864039421081543, + "learning_rate": 7.999319597422185e-05, + "loss": 0.5335, + "step": 33272 + }, + { + "epoch": 2.308861286517244, + "grad_norm": 1.3891981840133667, + "learning_rate": 7.998769098746877e-05, + "loss": 0.3921, + "step": 33273 + }, + { + "epoch": 2.3089306779543404, + "grad_norm": 1.512028455734253, + "learning_rate": 7.99821860638922e-05, + "loss": 0.4543, + "step": 33274 + }, + { + "epoch": 2.309000069391437, + "grad_norm": 1.0800632238388062, + "learning_rate": 7.997668120350956e-05, + "loss": 0.3247, + "step": 33275 + }, + { + "epoch": 2.3090694608285336, + "grad_norm": 1.8415496349334717, + "learning_rate": 7.997117640633822e-05, + "loss": 0.4101, + "step": 33276 + }, + { + "epoch": 2.3091388522656304, + "grad_norm": 1.6094173192977905, + "learning_rate": 7.99656716723956e-05, + "loss": 0.386, + "step": 33277 + }, + { + "epoch": 2.3092082437027273, + "grad_norm": 1.3239182233810425, + "learning_rate": 7.996016700169897e-05, + "loss": 0.3736, + "step": 33278 + }, + { + "epoch": 2.3092776351398236, + "grad_norm": 1.0369664430618286, + "learning_rate": 7.995466239426577e-05, + "loss": 0.3571, + "step": 33279 + }, + { + "epoch": 2.3093470265769205, + "grad_norm": 1.1647567749023438, + "learning_rate": 7.994915785011338e-05, + "loss": 0.4106, + "step": 33280 + }, + { + "epoch": 2.309416418014017, + "grad_norm": 1.5290560722351074, + "learning_rate": 7.994365336925917e-05, + "loss": 0.6256, + "step": 33281 + }, + { + "epoch": 2.3094858094511137, + "grad_norm": 1.3548916578292847, + "learning_rate": 7.993814895172054e-05, + "loss": 0.4305, + "step": 33282 + }, + { + "epoch": 2.3095552008882105, + "grad_norm": 1.909179925918579, + "learning_rate": 7.993264459751483e-05, + "loss": 0.4725, + "step": 33283 + }, + { + "epoch": 2.309624592325307, + "grad_norm": 1.0548450946807861, + "learning_rate": 7.992714030665945e-05, + "loss": 0.304, + "step": 33284 + }, + { + "epoch": 2.3096939837624038, + "grad_norm": 1.4616470336914062, + "learning_rate": 7.992163607917175e-05, + "loss": 0.489, + "step": 33285 + }, + { + "epoch": 2.3097633751995, + "grad_norm": 1.2789946794509888, + "learning_rate": 7.991613191506912e-05, + "loss": 0.4165, + "step": 33286 + }, + { + "epoch": 2.309832766636597, + "grad_norm": 1.323900580406189, + "learning_rate": 7.99106278143689e-05, + "loss": 0.4786, + "step": 33287 + }, + { + "epoch": 2.309902158073694, + "grad_norm": 1.0652748346328735, + "learning_rate": 7.990512377708852e-05, + "loss": 0.3385, + "step": 33288 + }, + { + "epoch": 2.30997154951079, + "grad_norm": 1.3075647354125977, + "learning_rate": 7.989961980324532e-05, + "loss": 0.5024, + "step": 33289 + }, + { + "epoch": 2.310040940947887, + "grad_norm": 1.6043133735656738, + "learning_rate": 7.98941158928567e-05, + "loss": 0.5303, + "step": 33290 + }, + { + "epoch": 2.310110332384984, + "grad_norm": 1.2046289443969727, + "learning_rate": 7.988861204594007e-05, + "loss": 0.4714, + "step": 33291 + }, + { + "epoch": 2.3101797238220803, + "grad_norm": 1.6824396848678589, + "learning_rate": 7.988310826251269e-05, + "loss": 0.7917, + "step": 33292 + }, + { + "epoch": 2.310249115259177, + "grad_norm": 0.9930504560470581, + "learning_rate": 7.987760454259203e-05, + "loss": 0.4028, + "step": 33293 + }, + { + "epoch": 2.310318506696274, + "grad_norm": 0.9801927804946899, + "learning_rate": 7.987210088619542e-05, + "loss": 0.4461, + "step": 33294 + }, + { + "epoch": 2.3103878981333703, + "grad_norm": 0.7965368032455444, + "learning_rate": 7.986659729334026e-05, + "loss": 0.2979, + "step": 33295 + }, + { + "epoch": 2.310457289570467, + "grad_norm": 1.526619553565979, + "learning_rate": 7.986109376404392e-05, + "loss": 0.407, + "step": 33296 + }, + { + "epoch": 2.3105266810075635, + "grad_norm": 1.301346778869629, + "learning_rate": 7.985559029832379e-05, + "loss": 0.4724, + "step": 33297 + }, + { + "epoch": 2.3105960724446604, + "grad_norm": 1.4929527044296265, + "learning_rate": 7.985008689619722e-05, + "loss": 0.4681, + "step": 33298 + }, + { + "epoch": 2.310665463881757, + "grad_norm": 1.2366608381271362, + "learning_rate": 7.984458355768156e-05, + "loss": 0.3764, + "step": 33299 + }, + { + "epoch": 2.3107348553188536, + "grad_norm": 1.0890618562698364, + "learning_rate": 7.983908028279423e-05, + "loss": 0.3672, + "step": 33300 + }, + { + "epoch": 2.3108042467559504, + "grad_norm": 1.063886046409607, + "learning_rate": 7.983357707155259e-05, + "loss": 0.3599, + "step": 33301 + }, + { + "epoch": 2.310873638193047, + "grad_norm": 1.1606578826904297, + "learning_rate": 7.982807392397399e-05, + "loss": 0.3312, + "step": 33302 + }, + { + "epoch": 2.3109430296301436, + "grad_norm": 1.639412760734558, + "learning_rate": 7.982257084007582e-05, + "loss": 0.4741, + "step": 33303 + }, + { + "epoch": 2.3110124210672405, + "grad_norm": 1.9112564325332642, + "learning_rate": 7.981706781987546e-05, + "loss": 0.5048, + "step": 33304 + }, + { + "epoch": 2.311081812504337, + "grad_norm": 1.3930202722549438, + "learning_rate": 7.981156486339031e-05, + "loss": 0.498, + "step": 33305 + }, + { + "epoch": 2.3111512039414337, + "grad_norm": 1.7031222581863403, + "learning_rate": 7.980606197063767e-05, + "loss": 0.4944, + "step": 33306 + }, + { + "epoch": 2.31122059537853, + "grad_norm": 1.495469570159912, + "learning_rate": 7.980055914163493e-05, + "loss": 0.4298, + "step": 33307 + }, + { + "epoch": 2.311289986815627, + "grad_norm": 1.6233549118041992, + "learning_rate": 7.97950563763995e-05, + "loss": 0.6448, + "step": 33308 + }, + { + "epoch": 2.3113593782527238, + "grad_norm": 1.8853856325149536, + "learning_rate": 7.978955367494873e-05, + "loss": 0.5002, + "step": 33309 + }, + { + "epoch": 2.31142876968982, + "grad_norm": 2.041459798812866, + "learning_rate": 7.97840510373e-05, + "loss": 0.5588, + "step": 33310 + }, + { + "epoch": 2.311498161126917, + "grad_norm": 1.4362826347351074, + "learning_rate": 7.977854846347067e-05, + "loss": 0.3835, + "step": 33311 + }, + { + "epoch": 2.3115675525640134, + "grad_norm": 1.7248462438583374, + "learning_rate": 7.977304595347816e-05, + "loss": 0.4354, + "step": 33312 + }, + { + "epoch": 2.31163694400111, + "grad_norm": 1.068499207496643, + "learning_rate": 7.976754350733974e-05, + "loss": 0.2622, + "step": 33313 + }, + { + "epoch": 2.311706335438207, + "grad_norm": 1.2000926733016968, + "learning_rate": 7.976204112507286e-05, + "loss": 0.3441, + "step": 33314 + }, + { + "epoch": 2.3117757268753034, + "grad_norm": 1.2911829948425293, + "learning_rate": 7.975653880669487e-05, + "loss": 0.3885, + "step": 33315 + }, + { + "epoch": 2.3118451183124002, + "grad_norm": 1.7484713792800903, + "learning_rate": 7.975103655222312e-05, + "loss": 0.314, + "step": 33316 + }, + { + "epoch": 2.311914509749497, + "grad_norm": 1.287743091583252, + "learning_rate": 7.974553436167503e-05, + "loss": 0.4655, + "step": 33317 + }, + { + "epoch": 2.3119839011865935, + "grad_norm": 1.0299079418182373, + "learning_rate": 7.974003223506792e-05, + "loss": 0.4014, + "step": 33318 + }, + { + "epoch": 2.3120532926236903, + "grad_norm": 1.283678650856018, + "learning_rate": 7.973453017241918e-05, + "loss": 0.3996, + "step": 33319 + }, + { + "epoch": 2.312122684060787, + "grad_norm": 1.4041184186935425, + "learning_rate": 7.97290281737462e-05, + "loss": 0.5268, + "step": 33320 + }, + { + "epoch": 2.3121920754978835, + "grad_norm": 0.9953463673591614, + "learning_rate": 7.972352623906629e-05, + "loss": 0.4536, + "step": 33321 + }, + { + "epoch": 2.3122614669349804, + "grad_norm": 1.3418158292770386, + "learning_rate": 7.971802436839687e-05, + "loss": 0.4922, + "step": 33322 + }, + { + "epoch": 2.3123308583720767, + "grad_norm": 1.086669683456421, + "learning_rate": 7.971252256175527e-05, + "loss": 0.4169, + "step": 33323 + }, + { + "epoch": 2.3124002498091736, + "grad_norm": 1.5309571027755737, + "learning_rate": 7.970702081915891e-05, + "loss": 0.4705, + "step": 33324 + }, + { + "epoch": 2.3124696412462704, + "grad_norm": 1.237727403640747, + "learning_rate": 7.970151914062512e-05, + "loss": 0.3606, + "step": 33325 + }, + { + "epoch": 2.312539032683367, + "grad_norm": 0.9525525569915771, + "learning_rate": 7.96960175261713e-05, + "loss": 0.3201, + "step": 33326 + }, + { + "epoch": 2.3126084241204636, + "grad_norm": 1.6148396730422974, + "learning_rate": 7.969051597581479e-05, + "loss": 0.5417, + "step": 33327 + }, + { + "epoch": 2.31267781555756, + "grad_norm": 1.077752709388733, + "learning_rate": 7.968501448957295e-05, + "loss": 0.5127, + "step": 33328 + }, + { + "epoch": 2.312747206994657, + "grad_norm": 1.6267316341400146, + "learning_rate": 7.967951306746316e-05, + "loss": 0.5181, + "step": 33329 + }, + { + "epoch": 2.3128165984317537, + "grad_norm": 1.055528163909912, + "learning_rate": 7.967401170950278e-05, + "loss": 0.3674, + "step": 33330 + }, + { + "epoch": 2.31288598986885, + "grad_norm": 1.2586935758590698, + "learning_rate": 7.96685104157092e-05, + "loss": 0.4082, + "step": 33331 + }, + { + "epoch": 2.312955381305947, + "grad_norm": 1.4786462783813477, + "learning_rate": 7.966300918609981e-05, + "loss": 0.5589, + "step": 33332 + }, + { + "epoch": 2.3130247727430433, + "grad_norm": 1.2289824485778809, + "learning_rate": 7.965750802069189e-05, + "loss": 0.3043, + "step": 33333 + }, + { + "epoch": 2.31309416418014, + "grad_norm": 1.0828217267990112, + "learning_rate": 7.965200691950287e-05, + "loss": 0.3852, + "step": 33334 + }, + { + "epoch": 2.313163555617237, + "grad_norm": 1.2627100944519043, + "learning_rate": 7.96465058825501e-05, + "loss": 0.2594, + "step": 33335 + }, + { + "epoch": 2.3132329470543334, + "grad_norm": 2.0996668338775635, + "learning_rate": 7.964100490985097e-05, + "loss": 0.5904, + "step": 33336 + }, + { + "epoch": 2.31330233849143, + "grad_norm": 0.948682427406311, + "learning_rate": 7.963550400142281e-05, + "loss": 0.3669, + "step": 33337 + }, + { + "epoch": 2.3133717299285266, + "grad_norm": 1.3148696422576904, + "learning_rate": 7.9630003157283e-05, + "loss": 0.4009, + "step": 33338 + }, + { + "epoch": 2.3134411213656234, + "grad_norm": 1.1185322999954224, + "learning_rate": 7.962450237744892e-05, + "loss": 0.3859, + "step": 33339 + }, + { + "epoch": 2.3135105128027202, + "grad_norm": 1.1468361616134644, + "learning_rate": 7.961900166193793e-05, + "loss": 0.3771, + "step": 33340 + }, + { + "epoch": 2.3135799042398166, + "grad_norm": 1.5572973489761353, + "learning_rate": 7.961350101076736e-05, + "loss": 0.4174, + "step": 33341 + }, + { + "epoch": 2.3136492956769135, + "grad_norm": 1.026712417602539, + "learning_rate": 7.960800042395461e-05, + "loss": 0.4753, + "step": 33342 + }, + { + "epoch": 2.3137186871140103, + "grad_norm": 1.3872482776641846, + "learning_rate": 7.960249990151703e-05, + "loss": 0.5365, + "step": 33343 + }, + { + "epoch": 2.3137880785511067, + "grad_norm": 1.465194582939148, + "learning_rate": 7.9596999443472e-05, + "loss": 0.659, + "step": 33344 + }, + { + "epoch": 2.3138574699882035, + "grad_norm": 1.3394678831100464, + "learning_rate": 7.959149904983687e-05, + "loss": 0.5304, + "step": 33345 + }, + { + "epoch": 2.3139268614253004, + "grad_norm": 0.9697035551071167, + "learning_rate": 7.958599872062904e-05, + "loss": 0.4478, + "step": 33346 + }, + { + "epoch": 2.3139962528623967, + "grad_norm": 1.0870542526245117, + "learning_rate": 7.958049845586581e-05, + "loss": 0.3332, + "step": 33347 + }, + { + "epoch": 2.3140656442994936, + "grad_norm": 0.7975345253944397, + "learning_rate": 7.957499825556458e-05, + "loss": 0.3473, + "step": 33348 + }, + { + "epoch": 2.31413503573659, + "grad_norm": 1.1978412866592407, + "learning_rate": 7.956949811974272e-05, + "loss": 0.4191, + "step": 33349 + }, + { + "epoch": 2.314204427173687, + "grad_norm": 1.3721569776535034, + "learning_rate": 7.956399804841756e-05, + "loss": 0.6555, + "step": 33350 + }, + { + "epoch": 2.3142738186107836, + "grad_norm": 1.275923252105713, + "learning_rate": 7.95584980416065e-05, + "loss": 0.4587, + "step": 33351 + }, + { + "epoch": 2.31434321004788, + "grad_norm": 1.2407026290893555, + "learning_rate": 7.955299809932693e-05, + "loss": 0.4511, + "step": 33352 + }, + { + "epoch": 2.314412601484977, + "grad_norm": 1.3124330043792725, + "learning_rate": 7.954749822159616e-05, + "loss": 0.4919, + "step": 33353 + }, + { + "epoch": 2.3144819929220732, + "grad_norm": 1.4941737651824951, + "learning_rate": 7.954199840843153e-05, + "loss": 0.5484, + "step": 33354 + }, + { + "epoch": 2.31455138435917, + "grad_norm": 1.136532187461853, + "learning_rate": 7.953649865985046e-05, + "loss": 0.5647, + "step": 33355 + }, + { + "epoch": 2.314620775796267, + "grad_norm": 2.134922981262207, + "learning_rate": 7.953099897587028e-05, + "loss": 0.5068, + "step": 33356 + }, + { + "epoch": 2.3146901672333633, + "grad_norm": 1.0321972370147705, + "learning_rate": 7.952549935650835e-05, + "loss": 0.3389, + "step": 33357 + }, + { + "epoch": 2.31475955867046, + "grad_norm": 1.2282971143722534, + "learning_rate": 7.951999980178206e-05, + "loss": 0.4151, + "step": 33358 + }, + { + "epoch": 2.3148289501075565, + "grad_norm": 1.215650200843811, + "learning_rate": 7.951450031170875e-05, + "loss": 0.4978, + "step": 33359 + }, + { + "epoch": 2.3148983415446533, + "grad_norm": 1.158334732055664, + "learning_rate": 7.950900088630581e-05, + "loss": 0.4179, + "step": 33360 + }, + { + "epoch": 2.31496773298175, + "grad_norm": 1.5464155673980713, + "learning_rate": 7.950350152559056e-05, + "loss": 0.6147, + "step": 33361 + }, + { + "epoch": 2.3150371244188466, + "grad_norm": 1.0618476867675781, + "learning_rate": 7.949800222958036e-05, + "loss": 0.3847, + "step": 33362 + }, + { + "epoch": 2.3151065158559434, + "grad_norm": 1.2143757343292236, + "learning_rate": 7.94925029982926e-05, + "loss": 0.4067, + "step": 33363 + }, + { + "epoch": 2.31517590729304, + "grad_norm": 1.570541501045227, + "learning_rate": 7.948700383174461e-05, + "loss": 0.4773, + "step": 33364 + }, + { + "epoch": 2.3152452987301366, + "grad_norm": 1.2050278186798096, + "learning_rate": 7.948150472995379e-05, + "loss": 0.2929, + "step": 33365 + }, + { + "epoch": 2.3153146901672335, + "grad_norm": 1.625411033630371, + "learning_rate": 7.947600569293748e-05, + "loss": 0.4493, + "step": 33366 + }, + { + "epoch": 2.31538408160433, + "grad_norm": 1.2022533416748047, + "learning_rate": 7.947050672071306e-05, + "loss": 0.421, + "step": 33367 + }, + { + "epoch": 2.3154534730414267, + "grad_norm": 1.1849061250686646, + "learning_rate": 7.946500781329785e-05, + "loss": 0.4772, + "step": 33368 + }, + { + "epoch": 2.3155228644785235, + "grad_norm": 1.0726267099380493, + "learning_rate": 7.94595089707092e-05, + "loss": 0.5312, + "step": 33369 + }, + { + "epoch": 2.31559225591562, + "grad_norm": 1.198158860206604, + "learning_rate": 7.945401019296452e-05, + "loss": 0.4275, + "step": 33370 + }, + { + "epoch": 2.3156616473527167, + "grad_norm": 1.4678080081939697, + "learning_rate": 7.944851148008115e-05, + "loss": 0.4956, + "step": 33371 + }, + { + "epoch": 2.3157310387898136, + "grad_norm": 0.9719760417938232, + "learning_rate": 7.944301283207643e-05, + "loss": 0.2499, + "step": 33372 + }, + { + "epoch": 2.31580043022691, + "grad_norm": 1.416397213935852, + "learning_rate": 7.943751424896774e-05, + "loss": 0.5263, + "step": 33373 + }, + { + "epoch": 2.315869821664007, + "grad_norm": 1.437817931175232, + "learning_rate": 7.943201573077247e-05, + "loss": 0.3832, + "step": 33374 + }, + { + "epoch": 2.315939213101103, + "grad_norm": 1.3807018995285034, + "learning_rate": 7.94265172775079e-05, + "loss": 0.5428, + "step": 33375 + }, + { + "epoch": 2.3160086045382, + "grad_norm": 1.0751755237579346, + "learning_rate": 7.942101888919144e-05, + "loss": 0.2912, + "step": 33376 + }, + { + "epoch": 2.316077995975297, + "grad_norm": 1.40067458152771, + "learning_rate": 7.94155205658404e-05, + "loss": 0.3188, + "step": 33377 + }, + { + "epoch": 2.3161473874123932, + "grad_norm": 1.0395398139953613, + "learning_rate": 7.941002230747221e-05, + "loss": 0.1992, + "step": 33378 + }, + { + "epoch": 2.31621677884949, + "grad_norm": 0.9898329377174377, + "learning_rate": 7.940452411410417e-05, + "loss": 0.2711, + "step": 33379 + }, + { + "epoch": 2.3162861702865865, + "grad_norm": 1.0982614755630493, + "learning_rate": 7.939902598575366e-05, + "loss": 0.4107, + "step": 33380 + }, + { + "epoch": 2.3163555617236833, + "grad_norm": 1.6342759132385254, + "learning_rate": 7.93935279224381e-05, + "loss": 0.4369, + "step": 33381 + }, + { + "epoch": 2.31642495316078, + "grad_norm": 1.471566081047058, + "learning_rate": 7.938802992417472e-05, + "loss": 0.4135, + "step": 33382 + }, + { + "epoch": 2.3164943445978765, + "grad_norm": 0.990588903427124, + "learning_rate": 7.938253199098094e-05, + "loss": 0.433, + "step": 33383 + }, + { + "epoch": 2.3165637360349733, + "grad_norm": 1.1899622678756714, + "learning_rate": 7.93770341228741e-05, + "loss": 0.4767, + "step": 33384 + }, + { + "epoch": 2.3166331274720697, + "grad_norm": 0.9766620993614197, + "learning_rate": 7.93715363198716e-05, + "loss": 0.4588, + "step": 33385 + }, + { + "epoch": 2.3167025189091666, + "grad_norm": 1.2208235263824463, + "learning_rate": 7.936603858199075e-05, + "loss": 0.4234, + "step": 33386 + }, + { + "epoch": 2.3167719103462634, + "grad_norm": 1.3896031379699707, + "learning_rate": 7.936054090924892e-05, + "loss": 0.5717, + "step": 33387 + }, + { + "epoch": 2.31684130178336, + "grad_norm": 0.9878929853439331, + "learning_rate": 7.93550433016635e-05, + "loss": 0.3355, + "step": 33388 + }, + { + "epoch": 2.3169106932204566, + "grad_norm": 1.0314308404922485, + "learning_rate": 7.934954575925179e-05, + "loss": 0.2747, + "step": 33389 + }, + { + "epoch": 2.3169800846575535, + "grad_norm": 1.1444058418273926, + "learning_rate": 7.934404828203117e-05, + "loss": 0.3564, + "step": 33390 + }, + { + "epoch": 2.31704947609465, + "grad_norm": 1.0583775043487549, + "learning_rate": 7.933855087001898e-05, + "loss": 0.3814, + "step": 33391 + }, + { + "epoch": 2.3171188675317467, + "grad_norm": 1.4537321329116821, + "learning_rate": 7.93330535232326e-05, + "loss": 0.4784, + "step": 33392 + }, + { + "epoch": 2.317188258968843, + "grad_norm": 0.9920659065246582, + "learning_rate": 7.932755624168937e-05, + "loss": 0.3077, + "step": 33393 + }, + { + "epoch": 2.31725765040594, + "grad_norm": 1.2193231582641602, + "learning_rate": 7.932205902540663e-05, + "loss": 0.4768, + "step": 33394 + }, + { + "epoch": 2.3173270418430367, + "grad_norm": 1.1650882959365845, + "learning_rate": 7.931656187440181e-05, + "loss": 0.3551, + "step": 33395 + }, + { + "epoch": 2.317396433280133, + "grad_norm": 1.4600023031234741, + "learning_rate": 7.931106478869215e-05, + "loss": 0.4815, + "step": 33396 + }, + { + "epoch": 2.31746582471723, + "grad_norm": 1.4821068048477173, + "learning_rate": 7.930556776829508e-05, + "loss": 0.5333, + "step": 33397 + }, + { + "epoch": 2.317535216154327, + "grad_norm": 1.273895263671875, + "learning_rate": 7.93000708132279e-05, + "loss": 0.2728, + "step": 33398 + }, + { + "epoch": 2.317604607591423, + "grad_norm": 1.879085898399353, + "learning_rate": 7.9294573923508e-05, + "loss": 0.7131, + "step": 33399 + }, + { + "epoch": 2.31767399902852, + "grad_norm": 1.5130420923233032, + "learning_rate": 7.928907709915275e-05, + "loss": 0.338, + "step": 33400 + }, + { + "epoch": 2.3177433904656164, + "grad_norm": 0.8710052967071533, + "learning_rate": 7.928358034017947e-05, + "loss": 0.2694, + "step": 33401 + }, + { + "epoch": 2.3178127819027132, + "grad_norm": 1.004650592803955, + "learning_rate": 7.927808364660556e-05, + "loss": 0.4657, + "step": 33402 + }, + { + "epoch": 2.31788217333981, + "grad_norm": 1.1357566118240356, + "learning_rate": 7.92725870184483e-05, + "loss": 0.3359, + "step": 33403 + }, + { + "epoch": 2.3179515647769064, + "grad_norm": 0.8914955258369446, + "learning_rate": 7.926709045572508e-05, + "loss": 0.3113, + "step": 33404 + }, + { + "epoch": 2.3180209562140033, + "grad_norm": 1.4929664134979248, + "learning_rate": 7.926159395845325e-05, + "loss": 0.3618, + "step": 33405 + }, + { + "epoch": 2.3180903476510997, + "grad_norm": 1.2734761238098145, + "learning_rate": 7.925609752665017e-05, + "loss": 0.392, + "step": 33406 + }, + { + "epoch": 2.3181597390881965, + "grad_norm": 1.2417335510253906, + "learning_rate": 7.925060116033317e-05, + "loss": 0.474, + "step": 33407 + }, + { + "epoch": 2.3182291305252933, + "grad_norm": 1.4982918500900269, + "learning_rate": 7.924510485951963e-05, + "loss": 0.474, + "step": 33408 + }, + { + "epoch": 2.3182985219623897, + "grad_norm": 1.040587067604065, + "learning_rate": 7.923960862422691e-05, + "loss": 0.4048, + "step": 33409 + }, + { + "epoch": 2.3183679133994866, + "grad_norm": 0.9203047156333923, + "learning_rate": 7.92341124544723e-05, + "loss": 0.3746, + "step": 33410 + }, + { + "epoch": 2.318437304836583, + "grad_norm": 1.508018136024475, + "learning_rate": 7.922861635027318e-05, + "loss": 0.3757, + "step": 33411 + }, + { + "epoch": 2.31850669627368, + "grad_norm": 1.2737709283828735, + "learning_rate": 7.922312031164691e-05, + "loss": 0.5265, + "step": 33412 + }, + { + "epoch": 2.3185760877107766, + "grad_norm": 1.3706949949264526, + "learning_rate": 7.921762433861084e-05, + "loss": 0.4666, + "step": 33413 + }, + { + "epoch": 2.318645479147873, + "grad_norm": 1.00226891040802, + "learning_rate": 7.921212843118233e-05, + "loss": 0.3046, + "step": 33414 + }, + { + "epoch": 2.31871487058497, + "grad_norm": 1.0891276597976685, + "learning_rate": 7.920663258937875e-05, + "loss": 0.4177, + "step": 33415 + }, + { + "epoch": 2.3187842620220667, + "grad_norm": 1.034401774406433, + "learning_rate": 7.920113681321737e-05, + "loss": 0.3182, + "step": 33416 + }, + { + "epoch": 2.318853653459163, + "grad_norm": 1.3884071111679077, + "learning_rate": 7.91956411027156e-05, + "loss": 0.3049, + "step": 33417 + }, + { + "epoch": 2.31892304489626, + "grad_norm": 0.9096802473068237, + "learning_rate": 7.919014545789079e-05, + "loss": 0.2253, + "step": 33418 + }, + { + "epoch": 2.3189924363333563, + "grad_norm": 1.358394980430603, + "learning_rate": 7.918464987876026e-05, + "loss": 0.3137, + "step": 33419 + }, + { + "epoch": 2.319061827770453, + "grad_norm": 1.043635368347168, + "learning_rate": 7.917915436534137e-05, + "loss": 0.312, + "step": 33420 + }, + { + "epoch": 2.31913121920755, + "grad_norm": 1.3085682392120361, + "learning_rate": 7.917365891765149e-05, + "loss": 0.4898, + "step": 33421 + }, + { + "epoch": 2.3192006106446463, + "grad_norm": 1.1975291967391968, + "learning_rate": 7.916816353570798e-05, + "loss": 0.4562, + "step": 33422 + }, + { + "epoch": 2.319270002081743, + "grad_norm": 1.7722523212432861, + "learning_rate": 7.916266821952812e-05, + "loss": 0.5773, + "step": 33423 + }, + { + "epoch": 2.31933939351884, + "grad_norm": 0.9720253348350525, + "learning_rate": 7.91571729691293e-05, + "loss": 0.2946, + "step": 33424 + }, + { + "epoch": 2.3194087849559364, + "grad_norm": 1.3547165393829346, + "learning_rate": 7.915167778452889e-05, + "loss": 0.3895, + "step": 33425 + }, + { + "epoch": 2.319478176393033, + "grad_norm": 1.0980936288833618, + "learning_rate": 7.914618266574419e-05, + "loss": 0.3025, + "step": 33426 + }, + { + "epoch": 2.3195475678301296, + "grad_norm": 0.9142965078353882, + "learning_rate": 7.914068761279257e-05, + "loss": 0.2794, + "step": 33427 + }, + { + "epoch": 2.3196169592672264, + "grad_norm": 1.163393259048462, + "learning_rate": 7.913519262569139e-05, + "loss": 0.3778, + "step": 33428 + }, + { + "epoch": 2.3196863507043233, + "grad_norm": 1.0854965448379517, + "learning_rate": 7.912969770445802e-05, + "loss": 0.4274, + "step": 33429 + }, + { + "epoch": 2.3197557421414197, + "grad_norm": 1.6443259716033936, + "learning_rate": 7.912420284910973e-05, + "loss": 0.5617, + "step": 33430 + }, + { + "epoch": 2.3198251335785165, + "grad_norm": 1.434261441230774, + "learning_rate": 7.911870805966392e-05, + "loss": 0.5765, + "step": 33431 + }, + { + "epoch": 2.319894525015613, + "grad_norm": 1.1721566915512085, + "learning_rate": 7.911321333613791e-05, + "loss": 0.4011, + "step": 33432 + }, + { + "epoch": 2.3199639164527097, + "grad_norm": 1.3106107711791992, + "learning_rate": 7.910771867854907e-05, + "loss": 0.4777, + "step": 33433 + }, + { + "epoch": 2.3200333078898066, + "grad_norm": 1.1360305547714233, + "learning_rate": 7.910222408691474e-05, + "loss": 0.3991, + "step": 33434 + }, + { + "epoch": 2.320102699326903, + "grad_norm": 1.8749922513961792, + "learning_rate": 7.909672956125227e-05, + "loss": 0.4812, + "step": 33435 + }, + { + "epoch": 2.3201720907639998, + "grad_norm": 1.9510611295700073, + "learning_rate": 7.909123510157903e-05, + "loss": 0.3199, + "step": 33436 + }, + { + "epoch": 2.320241482201096, + "grad_norm": 1.4223964214324951, + "learning_rate": 7.90857407079123e-05, + "loss": 0.4772, + "step": 33437 + }, + { + "epoch": 2.320310873638193, + "grad_norm": 1.05132257938385, + "learning_rate": 7.908024638026945e-05, + "loss": 0.3418, + "step": 33438 + }, + { + "epoch": 2.32038026507529, + "grad_norm": 0.9614399671554565, + "learning_rate": 7.907475211866785e-05, + "loss": 0.4215, + "step": 33439 + }, + { + "epoch": 2.320449656512386, + "grad_norm": 1.9184682369232178, + "learning_rate": 7.906925792312485e-05, + "loss": 0.4187, + "step": 33440 + }, + { + "epoch": 2.320519047949483, + "grad_norm": 1.0077905654907227, + "learning_rate": 7.906376379365774e-05, + "loss": 0.4813, + "step": 33441 + }, + { + "epoch": 2.32058843938658, + "grad_norm": 1.6758058071136475, + "learning_rate": 7.905826973028391e-05, + "loss": 0.3751, + "step": 33442 + }, + { + "epoch": 2.3206578308236763, + "grad_norm": 1.4074416160583496, + "learning_rate": 7.90527757330207e-05, + "loss": 0.5053, + "step": 33443 + }, + { + "epoch": 2.320727222260773, + "grad_norm": 1.0218747854232788, + "learning_rate": 7.904728180188545e-05, + "loss": 0.3701, + "step": 33444 + }, + { + "epoch": 2.3207966136978695, + "grad_norm": 1.2287145853042603, + "learning_rate": 7.904178793689549e-05, + "loss": 0.4621, + "step": 33445 + }, + { + "epoch": 2.3208660051349663, + "grad_norm": 1.2179811000823975, + "learning_rate": 7.903629413806817e-05, + "loss": 0.3707, + "step": 33446 + }, + { + "epoch": 2.320935396572063, + "grad_norm": 2.005512237548828, + "learning_rate": 7.903080040542084e-05, + "loss": 0.559, + "step": 33447 + }, + { + "epoch": 2.3210047880091595, + "grad_norm": 1.4091752767562866, + "learning_rate": 7.902530673897084e-05, + "loss": 0.2971, + "step": 33448 + }, + { + "epoch": 2.3210741794462564, + "grad_norm": 1.3610739707946777, + "learning_rate": 7.90198131387355e-05, + "loss": 0.4648, + "step": 33449 + }, + { + "epoch": 2.321143570883353, + "grad_norm": 1.6263070106506348, + "learning_rate": 7.901431960473223e-05, + "loss": 0.4732, + "step": 33450 + }, + { + "epoch": 2.3212129623204496, + "grad_norm": 1.6013622283935547, + "learning_rate": 7.900882613697827e-05, + "loss": 0.5308, + "step": 33451 + }, + { + "epoch": 2.3212823537575464, + "grad_norm": 1.1477636098861694, + "learning_rate": 7.900333273549102e-05, + "loss": 0.3574, + "step": 33452 + }, + { + "epoch": 2.321351745194643, + "grad_norm": 2.0274431705474854, + "learning_rate": 7.89978394002878e-05, + "loss": 0.4207, + "step": 33453 + }, + { + "epoch": 2.3214211366317397, + "grad_norm": 0.9439166188240051, + "learning_rate": 7.899234613138597e-05, + "loss": 0.2877, + "step": 33454 + }, + { + "epoch": 2.3214905280688365, + "grad_norm": 1.513261318206787, + "learning_rate": 7.898685292880288e-05, + "loss": 0.4885, + "step": 33455 + }, + { + "epoch": 2.321559919505933, + "grad_norm": 1.8764564990997314, + "learning_rate": 7.898135979255585e-05, + "loss": 0.5365, + "step": 33456 + }, + { + "epoch": 2.3216293109430297, + "grad_norm": 1.9494203329086304, + "learning_rate": 7.897586672266225e-05, + "loss": 0.439, + "step": 33457 + }, + { + "epoch": 2.321698702380126, + "grad_norm": 1.219207525253296, + "learning_rate": 7.897037371913937e-05, + "loss": 0.4161, + "step": 33458 + }, + { + "epoch": 2.321768093817223, + "grad_norm": 1.588922381401062, + "learning_rate": 7.89648807820046e-05, + "loss": 0.5156, + "step": 33459 + }, + { + "epoch": 2.3218374852543198, + "grad_norm": 1.2622438669204712, + "learning_rate": 7.895938791127525e-05, + "loss": 0.474, + "step": 33460 + }, + { + "epoch": 2.321906876691416, + "grad_norm": 1.0042083263397217, + "learning_rate": 7.895389510696866e-05, + "loss": 0.496, + "step": 33461 + }, + { + "epoch": 2.321976268128513, + "grad_norm": 0.9165029525756836, + "learning_rate": 7.894840236910219e-05, + "loss": 0.3313, + "step": 33462 + }, + { + "epoch": 2.3220456595656094, + "grad_norm": 0.8582057356834412, + "learning_rate": 7.894290969769317e-05, + "loss": 0.3345, + "step": 33463 + }, + { + "epoch": 2.322115051002706, + "grad_norm": 0.862928032875061, + "learning_rate": 7.893741709275898e-05, + "loss": 0.3066, + "step": 33464 + }, + { + "epoch": 2.322184442439803, + "grad_norm": 1.1613327264785767, + "learning_rate": 7.893192455431689e-05, + "loss": 0.3165, + "step": 33465 + }, + { + "epoch": 2.3222538338768994, + "grad_norm": 1.6000632047653198, + "learning_rate": 7.892643208238425e-05, + "loss": 0.638, + "step": 33466 + }, + { + "epoch": 2.3223232253139963, + "grad_norm": 1.1660099029541016, + "learning_rate": 7.892093967697842e-05, + "loss": 0.4314, + "step": 33467 + }, + { + "epoch": 2.322392616751093, + "grad_norm": 1.600500464439392, + "learning_rate": 7.891544733811676e-05, + "loss": 0.3058, + "step": 33468 + }, + { + "epoch": 2.3224620081881895, + "grad_norm": 1.3767772912979126, + "learning_rate": 7.890995506581655e-05, + "loss": 0.5203, + "step": 33469 + }, + { + "epoch": 2.3225313996252863, + "grad_norm": 1.5736161470413208, + "learning_rate": 7.89044628600952e-05, + "loss": 0.4453, + "step": 33470 + }, + { + "epoch": 2.3226007910623827, + "grad_norm": 0.9327185153961182, + "learning_rate": 7.889897072097004e-05, + "loss": 0.3084, + "step": 33471 + }, + { + "epoch": 2.3226701824994795, + "grad_norm": 4.243345260620117, + "learning_rate": 7.889347864845834e-05, + "loss": 0.4692, + "step": 33472 + }, + { + "epoch": 2.3227395739365764, + "grad_norm": 1.2184056043624878, + "learning_rate": 7.888798664257749e-05, + "loss": 0.5079, + "step": 33473 + }, + { + "epoch": 2.3228089653736728, + "grad_norm": 1.0715906620025635, + "learning_rate": 7.888249470334478e-05, + "loss": 0.4301, + "step": 33474 + }, + { + "epoch": 2.3228783568107696, + "grad_norm": 1.7635542154312134, + "learning_rate": 7.887700283077764e-05, + "loss": 0.563, + "step": 33475 + }, + { + "epoch": 2.3229477482478664, + "grad_norm": 1.1665576696395874, + "learning_rate": 7.887151102489332e-05, + "loss": 0.5669, + "step": 33476 + }, + { + "epoch": 2.323017139684963, + "grad_norm": 1.562988042831421, + "learning_rate": 7.88660192857092e-05, + "loss": 0.5317, + "step": 33477 + }, + { + "epoch": 2.3230865311220597, + "grad_norm": 1.2172056436538696, + "learning_rate": 7.886052761324262e-05, + "loss": 0.2778, + "step": 33478 + }, + { + "epoch": 2.323155922559156, + "grad_norm": 1.229384422302246, + "learning_rate": 7.88550360075109e-05, + "loss": 0.5891, + "step": 33479 + }, + { + "epoch": 2.323225313996253, + "grad_norm": 1.2809908390045166, + "learning_rate": 7.884954446853135e-05, + "loss": 0.394, + "step": 33480 + }, + { + "epoch": 2.3232947054333497, + "grad_norm": 1.3014034032821655, + "learning_rate": 7.884405299632134e-05, + "loss": 0.5298, + "step": 33481 + }, + { + "epoch": 2.323364096870446, + "grad_norm": 1.0830583572387695, + "learning_rate": 7.88385615908982e-05, + "loss": 0.2539, + "step": 33482 + }, + { + "epoch": 2.323433488307543, + "grad_norm": 1.2818410396575928, + "learning_rate": 7.883307025227926e-05, + "loss": 0.4483, + "step": 33483 + }, + { + "epoch": 2.3235028797446393, + "grad_norm": 1.6119608879089355, + "learning_rate": 7.882757898048192e-05, + "loss": 0.5758, + "step": 33484 + }, + { + "epoch": 2.323572271181736, + "grad_norm": 1.217330813407898, + "learning_rate": 7.882208777552341e-05, + "loss": 0.5339, + "step": 33485 + }, + { + "epoch": 2.323641662618833, + "grad_norm": 1.7177034616470337, + "learning_rate": 7.88165966374211e-05, + "loss": 0.4739, + "step": 33486 + }, + { + "epoch": 2.3237110540559294, + "grad_norm": 1.175107717514038, + "learning_rate": 7.881110556619233e-05, + "loss": 0.2758, + "step": 33487 + }, + { + "epoch": 2.323780445493026, + "grad_norm": 1.6332190036773682, + "learning_rate": 7.880561456185447e-05, + "loss": 0.4406, + "step": 33488 + }, + { + "epoch": 2.3238498369301226, + "grad_norm": 1.2831586599349976, + "learning_rate": 7.88001236244248e-05, + "loss": 0.5425, + "step": 33489 + }, + { + "epoch": 2.3239192283672194, + "grad_norm": 1.8001257181167603, + "learning_rate": 7.879463275392069e-05, + "loss": 0.6369, + "step": 33490 + }, + { + "epoch": 2.3239886198043163, + "grad_norm": 1.2509328126907349, + "learning_rate": 7.878914195035951e-05, + "loss": 0.3638, + "step": 33491 + }, + { + "epoch": 2.3240580112414126, + "grad_norm": 1.215411901473999, + "learning_rate": 7.87836512137585e-05, + "loss": 0.2974, + "step": 33492 + }, + { + "epoch": 2.3241274026785095, + "grad_norm": 1.349128246307373, + "learning_rate": 7.877816054413504e-05, + "loss": 0.5544, + "step": 33493 + }, + { + "epoch": 2.3241967941156063, + "grad_norm": 1.3128271102905273, + "learning_rate": 7.877266994150647e-05, + "loss": 0.3722, + "step": 33494 + }, + { + "epoch": 2.3242661855527027, + "grad_norm": 1.135528326034546, + "learning_rate": 7.876717940589014e-05, + "loss": 0.4586, + "step": 33495 + }, + { + "epoch": 2.3243355769897995, + "grad_norm": 1.0733754634857178, + "learning_rate": 7.876168893730334e-05, + "loss": 0.2839, + "step": 33496 + }, + { + "epoch": 2.3244049684268964, + "grad_norm": 0.9569342136383057, + "learning_rate": 7.875619853576342e-05, + "loss": 0.4036, + "step": 33497 + }, + { + "epoch": 2.3244743598639928, + "grad_norm": 1.4561421871185303, + "learning_rate": 7.875070820128776e-05, + "loss": 0.3299, + "step": 33498 + }, + { + "epoch": 2.3245437513010896, + "grad_norm": 1.7424999475479126, + "learning_rate": 7.87452179338936e-05, + "loss": 0.3802, + "step": 33499 + }, + { + "epoch": 2.324613142738186, + "grad_norm": 1.2730079889297485, + "learning_rate": 7.873972773359834e-05, + "loss": 0.3772, + "step": 33500 + }, + { + "epoch": 2.324682534175283, + "grad_norm": 1.2991290092468262, + "learning_rate": 7.873423760041927e-05, + "loss": 0.3738, + "step": 33501 + }, + { + "epoch": 2.3247519256123796, + "grad_norm": 1.1122486591339111, + "learning_rate": 7.872874753437375e-05, + "loss": 0.3262, + "step": 33502 + }, + { + "epoch": 2.324821317049476, + "grad_norm": 1.2783446311950684, + "learning_rate": 7.872325753547912e-05, + "loss": 0.4307, + "step": 33503 + }, + { + "epoch": 2.324890708486573, + "grad_norm": 1.7218822240829468, + "learning_rate": 7.87177676037527e-05, + "loss": 0.4464, + "step": 33504 + }, + { + "epoch": 2.3249600999236693, + "grad_norm": 1.6195720434188843, + "learning_rate": 7.871227773921184e-05, + "loss": 0.476, + "step": 33505 + }, + { + "epoch": 2.325029491360766, + "grad_norm": 1.0897213220596313, + "learning_rate": 7.870678794187382e-05, + "loss": 0.3996, + "step": 33506 + }, + { + "epoch": 2.325098882797863, + "grad_norm": 1.3209573030471802, + "learning_rate": 7.870129821175599e-05, + "loss": 0.3887, + "step": 33507 + }, + { + "epoch": 2.3251682742349593, + "grad_norm": 1.212705373764038, + "learning_rate": 7.86958085488757e-05, + "loss": 0.4359, + "step": 33508 + }, + { + "epoch": 2.325237665672056, + "grad_norm": 1.3697826862335205, + "learning_rate": 7.869031895325027e-05, + "loss": 0.3963, + "step": 33509 + }, + { + "epoch": 2.3253070571091525, + "grad_norm": 0.8785887360572815, + "learning_rate": 7.868482942489704e-05, + "loss": 0.2654, + "step": 33510 + }, + { + "epoch": 2.3253764485462494, + "grad_norm": 1.0020426511764526, + "learning_rate": 7.867933996383333e-05, + "loss": 0.3664, + "step": 33511 + }, + { + "epoch": 2.325445839983346, + "grad_norm": 1.5389519929885864, + "learning_rate": 7.867385057007649e-05, + "loss": 0.5276, + "step": 33512 + }, + { + "epoch": 2.3255152314204426, + "grad_norm": 2.19401478767395, + "learning_rate": 7.866836124364379e-05, + "loss": 0.4505, + "step": 33513 + }, + { + "epoch": 2.3255846228575394, + "grad_norm": 1.4940531253814697, + "learning_rate": 7.866287198455262e-05, + "loss": 0.517, + "step": 33514 + }, + { + "epoch": 2.325654014294636, + "grad_norm": 1.3406577110290527, + "learning_rate": 7.865738279282027e-05, + "loss": 0.3525, + "step": 33515 + }, + { + "epoch": 2.3257234057317326, + "grad_norm": 1.6097825765609741, + "learning_rate": 7.865189366846409e-05, + "loss": 0.4283, + "step": 33516 + }, + { + "epoch": 2.3257927971688295, + "grad_norm": 1.7103062868118286, + "learning_rate": 7.864640461150141e-05, + "loss": 0.3782, + "step": 33517 + }, + { + "epoch": 2.325862188605926, + "grad_norm": 1.0719034671783447, + "learning_rate": 7.864091562194955e-05, + "loss": 0.3054, + "step": 33518 + }, + { + "epoch": 2.3259315800430227, + "grad_norm": 1.5275306701660156, + "learning_rate": 7.863542669982589e-05, + "loss": 0.3656, + "step": 33519 + }, + { + "epoch": 2.3260009714801195, + "grad_norm": 1.1185808181762695, + "learning_rate": 7.862993784514765e-05, + "loss": 0.5162, + "step": 33520 + }, + { + "epoch": 2.326070362917216, + "grad_norm": 0.8484577536582947, + "learning_rate": 7.862444905793222e-05, + "loss": 0.367, + "step": 33521 + }, + { + "epoch": 2.3261397543543127, + "grad_norm": 1.3499088287353516, + "learning_rate": 7.861896033819693e-05, + "loss": 0.4596, + "step": 33522 + }, + { + "epoch": 2.3262091457914096, + "grad_norm": 1.814218282699585, + "learning_rate": 7.861347168595912e-05, + "loss": 0.4905, + "step": 33523 + }, + { + "epoch": 2.326278537228506, + "grad_norm": 1.974433422088623, + "learning_rate": 7.860798310123607e-05, + "loss": 0.4495, + "step": 33524 + }, + { + "epoch": 2.326347928665603, + "grad_norm": 1.352882742881775, + "learning_rate": 7.860249458404515e-05, + "loss": 0.5134, + "step": 33525 + }, + { + "epoch": 2.326417320102699, + "grad_norm": 1.6995604038238525, + "learning_rate": 7.85970061344037e-05, + "loss": 0.465, + "step": 33526 + }, + { + "epoch": 2.326486711539796, + "grad_norm": 1.0641257762908936, + "learning_rate": 7.859151775232899e-05, + "loss": 0.5479, + "step": 33527 + }, + { + "epoch": 2.326556102976893, + "grad_norm": 1.4790185689926147, + "learning_rate": 7.858602943783837e-05, + "loss": 0.435, + "step": 33528 + }, + { + "epoch": 2.3266254944139892, + "grad_norm": 1.6832505464553833, + "learning_rate": 7.85805411909492e-05, + "loss": 0.4209, + "step": 33529 + }, + { + "epoch": 2.326694885851086, + "grad_norm": 1.1818464994430542, + "learning_rate": 7.857505301167875e-05, + "loss": 0.3586, + "step": 33530 + }, + { + "epoch": 2.3267642772881825, + "grad_norm": 1.3861167430877686, + "learning_rate": 7.856956490004437e-05, + "loss": 0.5118, + "step": 33531 + }, + { + "epoch": 2.3268336687252793, + "grad_norm": 1.7749333381652832, + "learning_rate": 7.856407685606339e-05, + "loss": 0.7321, + "step": 33532 + }, + { + "epoch": 2.326903060162376, + "grad_norm": 1.8204410076141357, + "learning_rate": 7.855858887975316e-05, + "loss": 0.3705, + "step": 33533 + }, + { + "epoch": 2.3269724515994725, + "grad_norm": 1.2650351524353027, + "learning_rate": 7.855310097113096e-05, + "loss": 0.2826, + "step": 33534 + }, + { + "epoch": 2.3270418430365694, + "grad_norm": 1.6259962320327759, + "learning_rate": 7.854761313021411e-05, + "loss": 0.5008, + "step": 33535 + }, + { + "epoch": 2.3271112344736657, + "grad_norm": 1.1590962409973145, + "learning_rate": 7.854212535701997e-05, + "loss": 0.3095, + "step": 33536 + }, + { + "epoch": 2.3271806259107626, + "grad_norm": 1.6724913120269775, + "learning_rate": 7.853663765156585e-05, + "loss": 0.3972, + "step": 33537 + }, + { + "epoch": 2.3272500173478594, + "grad_norm": 1.1211243867874146, + "learning_rate": 7.853115001386905e-05, + "loss": 0.4264, + "step": 33538 + }, + { + "epoch": 2.327319408784956, + "grad_norm": 0.9545425772666931, + "learning_rate": 7.852566244394694e-05, + "loss": 0.3683, + "step": 33539 + }, + { + "epoch": 2.3273888002220526, + "grad_norm": 1.0146284103393555, + "learning_rate": 7.852017494181687e-05, + "loss": 0.2614, + "step": 33540 + }, + { + "epoch": 2.327458191659149, + "grad_norm": 1.2892729043960571, + "learning_rate": 7.851468750749607e-05, + "loss": 0.4934, + "step": 33541 + }, + { + "epoch": 2.327527583096246, + "grad_norm": 1.2195950746536255, + "learning_rate": 7.85092001410019e-05, + "loss": 0.3086, + "step": 33542 + }, + { + "epoch": 2.3275969745333427, + "grad_norm": 1.3703794479370117, + "learning_rate": 7.85037128423517e-05, + "loss": 0.3145, + "step": 33543 + }, + { + "epoch": 2.327666365970439, + "grad_norm": 1.0921765565872192, + "learning_rate": 7.849822561156279e-05, + "loss": 0.3571, + "step": 33544 + }, + { + "epoch": 2.327735757407536, + "grad_norm": 1.3864374160766602, + "learning_rate": 7.849273844865251e-05, + "loss": 0.3198, + "step": 33545 + }, + { + "epoch": 2.3278051488446327, + "grad_norm": 1.1471925973892212, + "learning_rate": 7.848725135363811e-05, + "loss": 0.261, + "step": 33546 + }, + { + "epoch": 2.327874540281729, + "grad_norm": 1.3737941980361938, + "learning_rate": 7.848176432653701e-05, + "loss": 0.4566, + "step": 33547 + }, + { + "epoch": 2.327943931718826, + "grad_norm": 1.1860508918762207, + "learning_rate": 7.847627736736645e-05, + "loss": 0.3314, + "step": 33548 + }, + { + "epoch": 2.328013323155923, + "grad_norm": 1.0858430862426758, + "learning_rate": 7.847079047614381e-05, + "loss": 0.3705, + "step": 33549 + }, + { + "epoch": 2.328082714593019, + "grad_norm": 1.399088740348816, + "learning_rate": 7.846530365288637e-05, + "loss": 0.4791, + "step": 33550 + }, + { + "epoch": 2.328152106030116, + "grad_norm": 1.3847659826278687, + "learning_rate": 7.845981689761146e-05, + "loss": 0.4983, + "step": 33551 + }, + { + "epoch": 2.3282214974672124, + "grad_norm": 1.0766812562942505, + "learning_rate": 7.845433021033641e-05, + "loss": 0.3386, + "step": 33552 + }, + { + "epoch": 2.3282908889043092, + "grad_norm": 1.3001247644424438, + "learning_rate": 7.844884359107853e-05, + "loss": 0.3816, + "step": 33553 + }, + { + "epoch": 2.328360280341406, + "grad_norm": 1.4941974878311157, + "learning_rate": 7.844335703985521e-05, + "loss": 0.526, + "step": 33554 + }, + { + "epoch": 2.3284296717785025, + "grad_norm": 1.498260498046875, + "learning_rate": 7.843787055668367e-05, + "loss": 0.417, + "step": 33555 + }, + { + "epoch": 2.3284990632155993, + "grad_norm": 1.0300928354263306, + "learning_rate": 7.843238414158127e-05, + "loss": 0.3809, + "step": 33556 + }, + { + "epoch": 2.3285684546526957, + "grad_norm": 1.6813255548477173, + "learning_rate": 7.842689779456532e-05, + "loss": 0.4593, + "step": 33557 + }, + { + "epoch": 2.3286378460897925, + "grad_norm": 1.0932762622833252, + "learning_rate": 7.842141151565315e-05, + "loss": 0.3229, + "step": 33558 + }, + { + "epoch": 2.3287072375268894, + "grad_norm": 1.9075077772140503, + "learning_rate": 7.841592530486208e-05, + "loss": 0.5835, + "step": 33559 + }, + { + "epoch": 2.3287766289639857, + "grad_norm": 1.3482953310012817, + "learning_rate": 7.841043916220949e-05, + "loss": 0.4122, + "step": 33560 + }, + { + "epoch": 2.3288460204010826, + "grad_norm": 0.9419879913330078, + "learning_rate": 7.840495308771258e-05, + "loss": 0.3225, + "step": 33561 + }, + { + "epoch": 2.328915411838179, + "grad_norm": 1.0948927402496338, + "learning_rate": 7.839946708138874e-05, + "loss": 0.5245, + "step": 33562 + }, + { + "epoch": 2.328984803275276, + "grad_norm": 1.3654825687408447, + "learning_rate": 7.839398114325526e-05, + "loss": 0.4557, + "step": 33563 + }, + { + "epoch": 2.3290541947123726, + "grad_norm": 1.3732686042785645, + "learning_rate": 7.83884952733295e-05, + "loss": 0.3618, + "step": 33564 + }, + { + "epoch": 2.329123586149469, + "grad_norm": 1.3851044178009033, + "learning_rate": 7.838300947162875e-05, + "loss": 0.3448, + "step": 33565 + }, + { + "epoch": 2.329192977586566, + "grad_norm": 1.0643097162246704, + "learning_rate": 7.837752373817032e-05, + "loss": 0.2674, + "step": 33566 + }, + { + "epoch": 2.3292623690236622, + "grad_norm": 1.2130564451217651, + "learning_rate": 7.837203807297155e-05, + "loss": 0.4113, + "step": 33567 + }, + { + "epoch": 2.329331760460759, + "grad_norm": 1.6644281148910522, + "learning_rate": 7.836655247604975e-05, + "loss": 0.4896, + "step": 33568 + }, + { + "epoch": 2.329401151897856, + "grad_norm": 1.2860966920852661, + "learning_rate": 7.83610669474222e-05, + "loss": 0.3078, + "step": 33569 + }, + { + "epoch": 2.3294705433349523, + "grad_norm": 1.512713074684143, + "learning_rate": 7.835558148710627e-05, + "loss": 0.6038, + "step": 33570 + }, + { + "epoch": 2.329539934772049, + "grad_norm": 1.5139473676681519, + "learning_rate": 7.835009609511927e-05, + "loss": 0.4893, + "step": 33571 + }, + { + "epoch": 2.329609326209146, + "grad_norm": 1.2876194715499878, + "learning_rate": 7.834461077147849e-05, + "loss": 0.4577, + "step": 33572 + }, + { + "epoch": 2.3296787176462423, + "grad_norm": 0.8398393392562866, + "learning_rate": 7.833912551620126e-05, + "loss": 0.3046, + "step": 33573 + }, + { + "epoch": 2.329748109083339, + "grad_norm": 1.3047268390655518, + "learning_rate": 7.833364032930493e-05, + "loss": 0.3248, + "step": 33574 + }, + { + "epoch": 2.329817500520436, + "grad_norm": 1.3227192163467407, + "learning_rate": 7.832815521080676e-05, + "loss": 0.3447, + "step": 33575 + }, + { + "epoch": 2.3298868919575324, + "grad_norm": 2.1209819316864014, + "learning_rate": 7.832267016072407e-05, + "loss": 0.5382, + "step": 33576 + }, + { + "epoch": 2.3299562833946292, + "grad_norm": 1.4837532043457031, + "learning_rate": 7.831718517907423e-05, + "loss": 0.4638, + "step": 33577 + }, + { + "epoch": 2.3300256748317256, + "grad_norm": 1.1714186668395996, + "learning_rate": 7.831170026587449e-05, + "loss": 0.45, + "step": 33578 + }, + { + "epoch": 2.3300950662688225, + "grad_norm": 1.0248163938522339, + "learning_rate": 7.83062154211422e-05, + "loss": 0.2289, + "step": 33579 + }, + { + "epoch": 2.3301644577059193, + "grad_norm": 1.4074190855026245, + "learning_rate": 7.830073064489469e-05, + "loss": 0.4687, + "step": 33580 + }, + { + "epoch": 2.3302338491430157, + "grad_norm": 1.208817481994629, + "learning_rate": 7.829524593714928e-05, + "loss": 0.4395, + "step": 33581 + }, + { + "epoch": 2.3303032405801125, + "grad_norm": 0.9735512137413025, + "learning_rate": 7.828976129792322e-05, + "loss": 0.2702, + "step": 33582 + }, + { + "epoch": 2.330372632017209, + "grad_norm": 1.5613963603973389, + "learning_rate": 7.828427672723389e-05, + "loss": 0.5897, + "step": 33583 + }, + { + "epoch": 2.3304420234543057, + "grad_norm": 0.9577376246452332, + "learning_rate": 7.827879222509854e-05, + "loss": 0.449, + "step": 33584 + }, + { + "epoch": 2.3305114148914026, + "grad_norm": 1.6401737928390503, + "learning_rate": 7.827330779153454e-05, + "loss": 0.4467, + "step": 33585 + }, + { + "epoch": 2.330580806328499, + "grad_norm": 1.642778754234314, + "learning_rate": 7.826782342655918e-05, + "loss": 0.3665, + "step": 33586 + }, + { + "epoch": 2.330650197765596, + "grad_norm": 1.6427536010742188, + "learning_rate": 7.826233913018978e-05, + "loss": 0.5522, + "step": 33587 + }, + { + "epoch": 2.330719589202692, + "grad_norm": 1.5851231813430786, + "learning_rate": 7.825685490244371e-05, + "loss": 0.5348, + "step": 33588 + }, + { + "epoch": 2.330788980639789, + "grad_norm": 1.1517397165298462, + "learning_rate": 7.825137074333818e-05, + "loss": 0.504, + "step": 33589 + }, + { + "epoch": 2.330858372076886, + "grad_norm": 1.3398295640945435, + "learning_rate": 7.824588665289055e-05, + "loss": 0.5682, + "step": 33590 + }, + { + "epoch": 2.3309277635139822, + "grad_norm": 1.9277617931365967, + "learning_rate": 7.824040263111811e-05, + "loss": 0.3591, + "step": 33591 + }, + { + "epoch": 2.330997154951079, + "grad_norm": 1.1772834062576294, + "learning_rate": 7.823491867803822e-05, + "loss": 0.3974, + "step": 33592 + }, + { + "epoch": 2.3310665463881755, + "grad_norm": 1.259745717048645, + "learning_rate": 7.822943479366815e-05, + "loss": 0.3193, + "step": 33593 + }, + { + "epoch": 2.3311359378252723, + "grad_norm": 1.4883235692977905, + "learning_rate": 7.822395097802522e-05, + "loss": 0.3806, + "step": 33594 + }, + { + "epoch": 2.331205329262369, + "grad_norm": 1.3049927949905396, + "learning_rate": 7.821846723112682e-05, + "loss": 0.5199, + "step": 33595 + }, + { + "epoch": 2.3312747206994655, + "grad_norm": 1.4388982057571411, + "learning_rate": 7.821298355299014e-05, + "loss": 0.3894, + "step": 33596 + }, + { + "epoch": 2.3313441121365623, + "grad_norm": 1.2450491189956665, + "learning_rate": 7.820749994363253e-05, + "loss": 0.4343, + "step": 33597 + }, + { + "epoch": 2.331413503573659, + "grad_norm": 2.233506679534912, + "learning_rate": 7.820201640307133e-05, + "loss": 0.5462, + "step": 33598 + }, + { + "epoch": 2.3314828950107556, + "grad_norm": 1.3958923816680908, + "learning_rate": 7.819653293132385e-05, + "loss": 0.4565, + "step": 33599 + }, + { + "epoch": 2.3315522864478524, + "grad_norm": 1.4079508781433105, + "learning_rate": 7.819104952840737e-05, + "loss": 0.2953, + "step": 33600 + }, + { + "epoch": 2.3316216778849492, + "grad_norm": 1.2256625890731812, + "learning_rate": 7.818556619433921e-05, + "loss": 0.42, + "step": 33601 + }, + { + "epoch": 2.3316910693220456, + "grad_norm": 1.0761351585388184, + "learning_rate": 7.818008292913672e-05, + "loss": 0.4118, + "step": 33602 + }, + { + "epoch": 2.3317604607591425, + "grad_norm": 1.1753509044647217, + "learning_rate": 7.817459973281716e-05, + "loss": 0.4018, + "step": 33603 + }, + { + "epoch": 2.331829852196239, + "grad_norm": 1.830426812171936, + "learning_rate": 7.816911660539783e-05, + "loss": 0.4932, + "step": 33604 + }, + { + "epoch": 2.3318992436333357, + "grad_norm": 1.4869133234024048, + "learning_rate": 7.816363354689609e-05, + "loss": 0.3539, + "step": 33605 + }, + { + "epoch": 2.3319686350704325, + "grad_norm": 1.3009402751922607, + "learning_rate": 7.81581505573292e-05, + "loss": 0.5037, + "step": 33606 + }, + { + "epoch": 2.332038026507529, + "grad_norm": 1.086793303489685, + "learning_rate": 7.815266763671452e-05, + "loss": 0.3309, + "step": 33607 + }, + { + "epoch": 2.3321074179446257, + "grad_norm": 1.0802819728851318, + "learning_rate": 7.814718478506931e-05, + "loss": 0.4279, + "step": 33608 + }, + { + "epoch": 2.332176809381722, + "grad_norm": 1.1492831707000732, + "learning_rate": 7.814170200241095e-05, + "loss": 0.4226, + "step": 33609 + }, + { + "epoch": 2.332246200818819, + "grad_norm": 1.252812385559082, + "learning_rate": 7.813621928875667e-05, + "loss": 0.4132, + "step": 33610 + }, + { + "epoch": 2.332315592255916, + "grad_norm": 1.1531683206558228, + "learning_rate": 7.81307366441238e-05, + "loss": 0.3849, + "step": 33611 + }, + { + "epoch": 2.332384983693012, + "grad_norm": 1.208473801612854, + "learning_rate": 7.812525406852966e-05, + "loss": 0.4557, + "step": 33612 + }, + { + "epoch": 2.332454375130109, + "grad_norm": 0.9175900220870972, + "learning_rate": 7.811977156199155e-05, + "loss": 0.3974, + "step": 33613 + }, + { + "epoch": 2.3325237665672054, + "grad_norm": 1.2655360698699951, + "learning_rate": 7.81142891245268e-05, + "loss": 0.3854, + "step": 33614 + }, + { + "epoch": 2.3325931580043022, + "grad_norm": 2.2784841060638428, + "learning_rate": 7.81088067561527e-05, + "loss": 0.6551, + "step": 33615 + }, + { + "epoch": 2.332662549441399, + "grad_norm": 1.0819344520568848, + "learning_rate": 7.810332445688658e-05, + "loss": 0.3828, + "step": 33616 + }, + { + "epoch": 2.3327319408784954, + "grad_norm": 1.4534622430801392, + "learning_rate": 7.809784222674568e-05, + "loss": 0.6325, + "step": 33617 + }, + { + "epoch": 2.3328013323155923, + "grad_norm": 1.2699170112609863, + "learning_rate": 7.809236006574738e-05, + "loss": 0.4447, + "step": 33618 + }, + { + "epoch": 2.3328707237526887, + "grad_norm": 0.7829234600067139, + "learning_rate": 7.808687797390893e-05, + "loss": 0.3005, + "step": 33619 + }, + { + "epoch": 2.3329401151897855, + "grad_norm": 1.5397123098373413, + "learning_rate": 7.808139595124769e-05, + "loss": 0.5731, + "step": 33620 + }, + { + "epoch": 2.3330095066268823, + "grad_norm": 1.381321907043457, + "learning_rate": 7.80759139977809e-05, + "loss": 0.4562, + "step": 33621 + }, + { + "epoch": 2.3330788980639787, + "grad_norm": 1.6423051357269287, + "learning_rate": 7.807043211352594e-05, + "loss": 0.5464, + "step": 33622 + }, + { + "epoch": 2.3331482895010756, + "grad_norm": 1.283945918083191, + "learning_rate": 7.80649502985001e-05, + "loss": 0.348, + "step": 33623 + }, + { + "epoch": 2.3332176809381724, + "grad_norm": 1.5871633291244507, + "learning_rate": 7.805946855272064e-05, + "loss": 0.4222, + "step": 33624 + }, + { + "epoch": 2.333287072375269, + "grad_norm": 1.683052659034729, + "learning_rate": 7.80539868762049e-05, + "loss": 0.3811, + "step": 33625 + }, + { + "epoch": 2.3333564638123656, + "grad_norm": 1.406043291091919, + "learning_rate": 7.804850526897015e-05, + "loss": 0.4563, + "step": 33626 + }, + { + "epoch": 2.3334258552494624, + "grad_norm": 1.3116960525512695, + "learning_rate": 7.804302373103374e-05, + "loss": 0.6187, + "step": 33627 + }, + { + "epoch": 2.333495246686559, + "grad_norm": 1.7052353620529175, + "learning_rate": 7.803754226241295e-05, + "loss": 0.4614, + "step": 33628 + }, + { + "epoch": 2.3335646381236557, + "grad_norm": 0.7829123139381409, + "learning_rate": 7.803206086312509e-05, + "loss": 0.4138, + "step": 33629 + }, + { + "epoch": 2.333634029560752, + "grad_norm": 1.6579943895339966, + "learning_rate": 7.802657953318751e-05, + "loss": 0.5622, + "step": 33630 + }, + { + "epoch": 2.333703420997849, + "grad_norm": 0.8850293755531311, + "learning_rate": 7.802109827261744e-05, + "loss": 0.3466, + "step": 33631 + }, + { + "epoch": 2.3337728124349457, + "grad_norm": 1.154587984085083, + "learning_rate": 7.80156170814322e-05, + "loss": 0.4104, + "step": 33632 + }, + { + "epoch": 2.333842203872042, + "grad_norm": 1.2633936405181885, + "learning_rate": 7.80101359596491e-05, + "loss": 0.3455, + "step": 33633 + }, + { + "epoch": 2.333911595309139, + "grad_norm": 1.3067001104354858, + "learning_rate": 7.800465490728547e-05, + "loss": 0.3957, + "step": 33634 + }, + { + "epoch": 2.3339809867462353, + "grad_norm": 1.208205223083496, + "learning_rate": 7.79991739243586e-05, + "loss": 0.4463, + "step": 33635 + }, + { + "epoch": 2.334050378183332, + "grad_norm": 1.3057993650436401, + "learning_rate": 7.799369301088579e-05, + "loss": 0.396, + "step": 33636 + }, + { + "epoch": 2.334119769620429, + "grad_norm": 1.483837366104126, + "learning_rate": 7.79882121668843e-05, + "loss": 0.5271, + "step": 33637 + }, + { + "epoch": 2.3341891610575254, + "grad_norm": 0.8358134031295776, + "learning_rate": 7.798273139237151e-05, + "loss": 0.3941, + "step": 33638 + }, + { + "epoch": 2.334258552494622, + "grad_norm": 1.3507272005081177, + "learning_rate": 7.797725068736465e-05, + "loss": 0.3933, + "step": 33639 + }, + { + "epoch": 2.3343279439317186, + "grad_norm": 1.2091645002365112, + "learning_rate": 7.797177005188106e-05, + "loss": 0.4305, + "step": 33640 + }, + { + "epoch": 2.3343973353688154, + "grad_norm": 1.186668038368225, + "learning_rate": 7.796628948593803e-05, + "loss": 0.3466, + "step": 33641 + }, + { + "epoch": 2.3344667268059123, + "grad_norm": 1.5612783432006836, + "learning_rate": 7.796080898955288e-05, + "loss": 0.3282, + "step": 33642 + }, + { + "epoch": 2.3345361182430087, + "grad_norm": 0.832024872303009, + "learning_rate": 7.795532856274294e-05, + "loss": 0.4977, + "step": 33643 + }, + { + "epoch": 2.3346055096801055, + "grad_norm": 1.4176079034805298, + "learning_rate": 7.794984820552542e-05, + "loss": 0.5639, + "step": 33644 + }, + { + "epoch": 2.3346749011172023, + "grad_norm": 1.3708677291870117, + "learning_rate": 7.794436791791767e-05, + "loss": 0.3847, + "step": 33645 + }, + { + "epoch": 2.3347442925542987, + "grad_norm": 0.9441332817077637, + "learning_rate": 7.793888769993701e-05, + "loss": 0.3864, + "step": 33646 + }, + { + "epoch": 2.3348136839913955, + "grad_norm": 1.2818191051483154, + "learning_rate": 7.79334075516007e-05, + "loss": 0.4511, + "step": 33647 + }, + { + "epoch": 2.334883075428492, + "grad_norm": 1.1048861742019653, + "learning_rate": 7.792792747292607e-05, + "loss": 0.3865, + "step": 33648 + }, + { + "epoch": 2.3349524668655888, + "grad_norm": 0.9411976337432861, + "learning_rate": 7.79224474639304e-05, + "loss": 0.3779, + "step": 33649 + }, + { + "epoch": 2.3350218583026856, + "grad_norm": 1.1509956121444702, + "learning_rate": 7.791696752463107e-05, + "loss": 0.3108, + "step": 33650 + }, + { + "epoch": 2.335091249739782, + "grad_norm": 1.090920090675354, + "learning_rate": 7.791148765504524e-05, + "loss": 0.36, + "step": 33651 + }, + { + "epoch": 2.335160641176879, + "grad_norm": 1.4674285650253296, + "learning_rate": 7.790600785519032e-05, + "loss": 0.3707, + "step": 33652 + }, + { + "epoch": 2.3352300326139757, + "grad_norm": 1.4037050008773804, + "learning_rate": 7.790052812508355e-05, + "loss": 0.3716, + "step": 33653 + }, + { + "epoch": 2.335299424051072, + "grad_norm": 1.2089056968688965, + "learning_rate": 7.789504846474226e-05, + "loss": 0.3553, + "step": 33654 + }, + { + "epoch": 2.335368815488169, + "grad_norm": 1.6104580163955688, + "learning_rate": 7.788956887418371e-05, + "loss": 0.4755, + "step": 33655 + }, + { + "epoch": 2.3354382069252653, + "grad_norm": 1.532365083694458, + "learning_rate": 7.788408935342524e-05, + "loss": 0.458, + "step": 33656 + }, + { + "epoch": 2.335507598362362, + "grad_norm": 1.2335171699523926, + "learning_rate": 7.787860990248416e-05, + "loss": 0.4194, + "step": 33657 + }, + { + "epoch": 2.335576989799459, + "grad_norm": 1.2543026208877563, + "learning_rate": 7.787313052137773e-05, + "loss": 0.3773, + "step": 33658 + }, + { + "epoch": 2.3356463812365553, + "grad_norm": 0.7902342081069946, + "learning_rate": 7.786765121012324e-05, + "loss": 0.2539, + "step": 33659 + }, + { + "epoch": 2.335715772673652, + "grad_norm": 1.2930645942687988, + "learning_rate": 7.7862171968738e-05, + "loss": 0.3456, + "step": 33660 + }, + { + "epoch": 2.3357851641107485, + "grad_norm": 1.1628303527832031, + "learning_rate": 7.785669279723933e-05, + "loss": 0.3718, + "step": 33661 + }, + { + "epoch": 2.3358545555478454, + "grad_norm": 1.2660332918167114, + "learning_rate": 7.785121369564451e-05, + "loss": 0.3467, + "step": 33662 + }, + { + "epoch": 2.335923946984942, + "grad_norm": 1.7440837621688843, + "learning_rate": 7.784573466397082e-05, + "loss": 0.4107, + "step": 33663 + }, + { + "epoch": 2.3359933384220386, + "grad_norm": 1.0818884372711182, + "learning_rate": 7.784025570223564e-05, + "loss": 0.4223, + "step": 33664 + }, + { + "epoch": 2.3360627298591354, + "grad_norm": 1.2102701663970947, + "learning_rate": 7.783477681045616e-05, + "loss": 0.4439, + "step": 33665 + }, + { + "epoch": 2.336132121296232, + "grad_norm": 1.2587933540344238, + "learning_rate": 7.782929798864971e-05, + "loss": 0.3661, + "step": 33666 + }, + { + "epoch": 2.3362015127333287, + "grad_norm": 1.2189861536026, + "learning_rate": 7.78238192368336e-05, + "loss": 0.34, + "step": 33667 + }, + { + "epoch": 2.3362709041704255, + "grad_norm": 1.3323607444763184, + "learning_rate": 7.78183405550251e-05, + "loss": 0.3129, + "step": 33668 + }, + { + "epoch": 2.336340295607522, + "grad_norm": 2.1862237453460693, + "learning_rate": 7.781286194324156e-05, + "loss": 0.5016, + "step": 33669 + }, + { + "epoch": 2.3364096870446187, + "grad_norm": 1.4774552583694458, + "learning_rate": 7.780738340150023e-05, + "loss": 0.6736, + "step": 33670 + }, + { + "epoch": 2.3364790784817155, + "grad_norm": 1.7957578897476196, + "learning_rate": 7.780190492981843e-05, + "loss": 0.495, + "step": 33671 + }, + { + "epoch": 2.336548469918812, + "grad_norm": 1.0561288595199585, + "learning_rate": 7.779642652821344e-05, + "loss": 0.2901, + "step": 33672 + }, + { + "epoch": 2.3366178613559088, + "grad_norm": 1.4977726936340332, + "learning_rate": 7.779094819670252e-05, + "loss": 0.4651, + "step": 33673 + }, + { + "epoch": 2.336687252793005, + "grad_norm": 0.9935848116874695, + "learning_rate": 7.778546993530302e-05, + "loss": 0.4314, + "step": 33674 + }, + { + "epoch": 2.336756644230102, + "grad_norm": 1.50191330909729, + "learning_rate": 7.777999174403221e-05, + "loss": 0.4621, + "step": 33675 + }, + { + "epoch": 2.336826035667199, + "grad_norm": 1.340909481048584, + "learning_rate": 7.777451362290738e-05, + "loss": 0.4148, + "step": 33676 + }, + { + "epoch": 2.336895427104295, + "grad_norm": 1.7553781270980835, + "learning_rate": 7.776903557194584e-05, + "loss": 0.4964, + "step": 33677 + }, + { + "epoch": 2.336964818541392, + "grad_norm": 1.4199022054672241, + "learning_rate": 7.77635575911649e-05, + "loss": 0.3704, + "step": 33678 + }, + { + "epoch": 2.337034209978489, + "grad_norm": 0.9024773240089417, + "learning_rate": 7.77580796805818e-05, + "loss": 0.3706, + "step": 33679 + }, + { + "epoch": 2.3371036014155853, + "grad_norm": 1.2678792476654053, + "learning_rate": 7.775260184021387e-05, + "loss": 0.3591, + "step": 33680 + }, + { + "epoch": 2.337172992852682, + "grad_norm": 1.185464859008789, + "learning_rate": 7.774712407007838e-05, + "loss": 0.508, + "step": 33681 + }, + { + "epoch": 2.3372423842897785, + "grad_norm": 1.3005133867263794, + "learning_rate": 7.774164637019265e-05, + "loss": 0.4366, + "step": 33682 + }, + { + "epoch": 2.3373117757268753, + "grad_norm": 1.5190266370773315, + "learning_rate": 7.773616874057395e-05, + "loss": 0.4971, + "step": 33683 + }, + { + "epoch": 2.337381167163972, + "grad_norm": 1.7693978548049927, + "learning_rate": 7.77306911812396e-05, + "loss": 0.4738, + "step": 33684 + }, + { + "epoch": 2.3374505586010685, + "grad_norm": 1.4545427560806274, + "learning_rate": 7.772521369220689e-05, + "loss": 0.4991, + "step": 33685 + }, + { + "epoch": 2.3375199500381654, + "grad_norm": 1.3334020376205444, + "learning_rate": 7.771973627349307e-05, + "loss": 0.3645, + "step": 33686 + }, + { + "epoch": 2.3375893414752618, + "grad_norm": 1.3554022312164307, + "learning_rate": 7.771425892511546e-05, + "loss": 0.4343, + "step": 33687 + }, + { + "epoch": 2.3376587329123586, + "grad_norm": 1.526380181312561, + "learning_rate": 7.770878164709135e-05, + "loss": 0.4448, + "step": 33688 + }, + { + "epoch": 2.3377281243494554, + "grad_norm": 1.3433295488357544, + "learning_rate": 7.770330443943803e-05, + "loss": 0.4998, + "step": 33689 + }, + { + "epoch": 2.337797515786552, + "grad_norm": 1.266379714012146, + "learning_rate": 7.769782730217278e-05, + "loss": 0.435, + "step": 33690 + }, + { + "epoch": 2.3378669072236486, + "grad_norm": 1.3017836809158325, + "learning_rate": 7.769235023531292e-05, + "loss": 0.3585, + "step": 33691 + }, + { + "epoch": 2.337936298660745, + "grad_norm": 1.6109870672225952, + "learning_rate": 7.768687323887574e-05, + "loss": 0.5369, + "step": 33692 + }, + { + "epoch": 2.338005690097842, + "grad_norm": 1.7930247783660889, + "learning_rate": 7.768139631287849e-05, + "loss": 0.488, + "step": 33693 + }, + { + "epoch": 2.3380750815349387, + "grad_norm": 1.3048465251922607, + "learning_rate": 7.767591945733846e-05, + "loss": 0.4912, + "step": 33694 + }, + { + "epoch": 2.338144472972035, + "grad_norm": 1.794646143913269, + "learning_rate": 7.767044267227298e-05, + "loss": 0.6206, + "step": 33695 + }, + { + "epoch": 2.338213864409132, + "grad_norm": 1.394497275352478, + "learning_rate": 7.766496595769932e-05, + "loss": 0.453, + "step": 33696 + }, + { + "epoch": 2.3382832558462288, + "grad_norm": 1.2145593166351318, + "learning_rate": 7.765948931363476e-05, + "loss": 0.3896, + "step": 33697 + }, + { + "epoch": 2.338352647283325, + "grad_norm": 1.4633312225341797, + "learning_rate": 7.765401274009661e-05, + "loss": 0.5785, + "step": 33698 + }, + { + "epoch": 2.338422038720422, + "grad_norm": 1.6281081438064575, + "learning_rate": 7.76485362371022e-05, + "loss": 0.4288, + "step": 33699 + }, + { + "epoch": 2.3384914301575184, + "grad_norm": 1.3677198886871338, + "learning_rate": 7.76430598046687e-05, + "loss": 0.3626, + "step": 33700 + }, + { + "epoch": 2.338560821594615, + "grad_norm": 1.4652451276779175, + "learning_rate": 7.763758344281349e-05, + "loss": 0.4422, + "step": 33701 + }, + { + "epoch": 2.338630213031712, + "grad_norm": 1.2566859722137451, + "learning_rate": 7.763210715155383e-05, + "loss": 0.3218, + "step": 33702 + }, + { + "epoch": 2.3386996044688084, + "grad_norm": 2.014451026916504, + "learning_rate": 7.762663093090701e-05, + "loss": 0.3206, + "step": 33703 + }, + { + "epoch": 2.3387689959059053, + "grad_norm": 1.1800423860549927, + "learning_rate": 7.762115478089034e-05, + "loss": 0.3171, + "step": 33704 + }, + { + "epoch": 2.338838387343002, + "grad_norm": 1.2376030683517456, + "learning_rate": 7.761567870152107e-05, + "loss": 0.3657, + "step": 33705 + }, + { + "epoch": 2.3389077787800985, + "grad_norm": 1.364427089691162, + "learning_rate": 7.761020269281651e-05, + "loss": 0.4245, + "step": 33706 + }, + { + "epoch": 2.3389771702171953, + "grad_norm": 1.423535943031311, + "learning_rate": 7.760472675479396e-05, + "loss": 0.3867, + "step": 33707 + }, + { + "epoch": 2.3390465616542917, + "grad_norm": 1.144631028175354, + "learning_rate": 7.759925088747067e-05, + "loss": 0.3783, + "step": 33708 + }, + { + "epoch": 2.3391159530913885, + "grad_norm": 1.235866904258728, + "learning_rate": 7.759377509086394e-05, + "loss": 0.4878, + "step": 33709 + }, + { + "epoch": 2.3391853445284854, + "grad_norm": 1.4168593883514404, + "learning_rate": 7.758829936499105e-05, + "loss": 0.55, + "step": 33710 + }, + { + "epoch": 2.3392547359655818, + "grad_norm": 1.4237457513809204, + "learning_rate": 7.758282370986932e-05, + "loss": 0.4602, + "step": 33711 + }, + { + "epoch": 2.3393241274026786, + "grad_norm": 1.4568676948547363, + "learning_rate": 7.757734812551605e-05, + "loss": 0.4488, + "step": 33712 + }, + { + "epoch": 2.339393518839775, + "grad_norm": 1.0010405778884888, + "learning_rate": 7.757187261194844e-05, + "loss": 0.2919, + "step": 33713 + }, + { + "epoch": 2.339462910276872, + "grad_norm": 1.2206056118011475, + "learning_rate": 7.756639716918384e-05, + "loss": 0.5845, + "step": 33714 + }, + { + "epoch": 2.3395323017139686, + "grad_norm": 1.352974534034729, + "learning_rate": 7.75609217972395e-05, + "loss": 0.6253, + "step": 33715 + }, + { + "epoch": 2.339601693151065, + "grad_norm": 1.1906893253326416, + "learning_rate": 7.755544649613275e-05, + "loss": 0.4585, + "step": 33716 + }, + { + "epoch": 2.339671084588162, + "grad_norm": 1.0524978637695312, + "learning_rate": 7.754997126588083e-05, + "loss": 0.3369, + "step": 33717 + }, + { + "epoch": 2.3397404760252583, + "grad_norm": 1.1006557941436768, + "learning_rate": 7.754449610650107e-05, + "loss": 0.3632, + "step": 33718 + }, + { + "epoch": 2.339809867462355, + "grad_norm": 1.487568974494934, + "learning_rate": 7.753902101801075e-05, + "loss": 0.4434, + "step": 33719 + }, + { + "epoch": 2.339879258899452, + "grad_norm": 1.6221522092819214, + "learning_rate": 7.75335460004271e-05, + "loss": 0.439, + "step": 33720 + }, + { + "epoch": 2.3399486503365483, + "grad_norm": 1.428243637084961, + "learning_rate": 7.752807105376745e-05, + "loss": 0.4856, + "step": 33721 + }, + { + "epoch": 2.340018041773645, + "grad_norm": 1.3039146661758423, + "learning_rate": 7.752259617804906e-05, + "loss": 0.4504, + "step": 33722 + }, + { + "epoch": 2.340087433210742, + "grad_norm": 1.7363135814666748, + "learning_rate": 7.751712137328926e-05, + "loss": 0.6902, + "step": 33723 + }, + { + "epoch": 2.3401568246478384, + "grad_norm": 1.3679715394973755, + "learning_rate": 7.751164663950526e-05, + "loss": 0.3938, + "step": 33724 + }, + { + "epoch": 2.340226216084935, + "grad_norm": 1.3528732061386108, + "learning_rate": 7.750617197671439e-05, + "loss": 0.509, + "step": 33725 + }, + { + "epoch": 2.340295607522032, + "grad_norm": 1.0444281101226807, + "learning_rate": 7.750069738493397e-05, + "loss": 0.3944, + "step": 33726 + }, + { + "epoch": 2.3403649989591284, + "grad_norm": 1.0073304176330566, + "learning_rate": 7.74952228641812e-05, + "loss": 0.3088, + "step": 33727 + }, + { + "epoch": 2.3404343903962253, + "grad_norm": 2.160367727279663, + "learning_rate": 7.74897484144734e-05, + "loss": 0.524, + "step": 33728 + }, + { + "epoch": 2.3405037818333216, + "grad_norm": 1.2544925212860107, + "learning_rate": 7.748427403582784e-05, + "loss": 0.4728, + "step": 33729 + }, + { + "epoch": 2.3405731732704185, + "grad_norm": 1.1519675254821777, + "learning_rate": 7.747879972826182e-05, + "loss": 0.3221, + "step": 33730 + }, + { + "epoch": 2.3406425647075153, + "grad_norm": 1.3885997533798218, + "learning_rate": 7.747332549179263e-05, + "loss": 0.5035, + "step": 33731 + }, + { + "epoch": 2.3407119561446117, + "grad_norm": 1.1496375799179077, + "learning_rate": 7.746785132643752e-05, + "loss": 0.3607, + "step": 33732 + }, + { + "epoch": 2.3407813475817085, + "grad_norm": 0.8845146894454956, + "learning_rate": 7.746237723221383e-05, + "loss": 0.4238, + "step": 33733 + }, + { + "epoch": 2.340850739018805, + "grad_norm": 1.2537853717803955, + "learning_rate": 7.745690320913877e-05, + "loss": 0.4906, + "step": 33734 + }, + { + "epoch": 2.3409201304559017, + "grad_norm": 1.3769032955169678, + "learning_rate": 7.745142925722966e-05, + "loss": 0.4584, + "step": 33735 + }, + { + "epoch": 2.3409895218929986, + "grad_norm": 1.141282081604004, + "learning_rate": 7.744595537650374e-05, + "loss": 0.388, + "step": 33736 + }, + { + "epoch": 2.341058913330095, + "grad_norm": 1.2029732465744019, + "learning_rate": 7.744048156697836e-05, + "loss": 0.4235, + "step": 33737 + }, + { + "epoch": 2.341128304767192, + "grad_norm": 0.9259267449378967, + "learning_rate": 7.743500782867076e-05, + "loss": 0.2768, + "step": 33738 + }, + { + "epoch": 2.341197696204288, + "grad_norm": 1.747670292854309, + "learning_rate": 7.74295341615982e-05, + "loss": 0.5125, + "step": 33739 + }, + { + "epoch": 2.341267087641385, + "grad_norm": 1.3479562997817993, + "learning_rate": 7.742406056577802e-05, + "loss": 0.4533, + "step": 33740 + }, + { + "epoch": 2.341336479078482, + "grad_norm": 1.3004035949707031, + "learning_rate": 7.741858704122742e-05, + "loss": 0.5223, + "step": 33741 + }, + { + "epoch": 2.3414058705155782, + "grad_norm": 1.3547120094299316, + "learning_rate": 7.741311358796375e-05, + "loss": 0.652, + "step": 33742 + }, + { + "epoch": 2.341475261952675, + "grad_norm": 1.0468593835830688, + "learning_rate": 7.740764020600424e-05, + "loss": 0.5557, + "step": 33743 + }, + { + "epoch": 2.3415446533897715, + "grad_norm": 1.3413333892822266, + "learning_rate": 7.74021668953662e-05, + "loss": 0.5508, + "step": 33744 + }, + { + "epoch": 2.3416140448268683, + "grad_norm": 1.1547868251800537, + "learning_rate": 7.739669365606687e-05, + "loss": 0.5074, + "step": 33745 + }, + { + "epoch": 2.341683436263965, + "grad_norm": 1.2743672132492065, + "learning_rate": 7.739122048812357e-05, + "loss": 0.5343, + "step": 33746 + }, + { + "epoch": 2.3417528277010615, + "grad_norm": 1.822016716003418, + "learning_rate": 7.738574739155361e-05, + "loss": 0.5444, + "step": 33747 + }, + { + "epoch": 2.3418222191381584, + "grad_norm": 1.1733589172363281, + "learning_rate": 7.738027436637418e-05, + "loss": 0.4636, + "step": 33748 + }, + { + "epoch": 2.341891610575255, + "grad_norm": 1.0607863664627075, + "learning_rate": 7.737480141260259e-05, + "loss": 0.3399, + "step": 33749 + }, + { + "epoch": 2.3419610020123516, + "grad_norm": 0.9709525108337402, + "learning_rate": 7.736932853025614e-05, + "loss": 0.4712, + "step": 33750 + }, + { + "epoch": 2.3420303934494484, + "grad_norm": 1.2727558612823486, + "learning_rate": 7.736385571935208e-05, + "loss": 0.3814, + "step": 33751 + }, + { + "epoch": 2.3420997848865452, + "grad_norm": 1.434216022491455, + "learning_rate": 7.73583829799077e-05, + "loss": 0.5245, + "step": 33752 + }, + { + "epoch": 2.3421691763236416, + "grad_norm": 1.2266584634780884, + "learning_rate": 7.73529103119403e-05, + "loss": 0.3379, + "step": 33753 + }, + { + "epoch": 2.3422385677607385, + "grad_norm": 1.239881157875061, + "learning_rate": 7.734743771546715e-05, + "loss": 0.3333, + "step": 33754 + }, + { + "epoch": 2.342307959197835, + "grad_norm": 1.153255820274353, + "learning_rate": 7.73419651905055e-05, + "loss": 0.3117, + "step": 33755 + }, + { + "epoch": 2.3423773506349317, + "grad_norm": 1.1214109659194946, + "learning_rate": 7.733649273707261e-05, + "loss": 0.4295, + "step": 33756 + }, + { + "epoch": 2.3424467420720285, + "grad_norm": 0.9255803227424622, + "learning_rate": 7.73310203551858e-05, + "loss": 0.3588, + "step": 33757 + }, + { + "epoch": 2.342516133509125, + "grad_norm": 1.0300706624984741, + "learning_rate": 7.732554804486234e-05, + "loss": 0.4513, + "step": 33758 + }, + { + "epoch": 2.3425855249462217, + "grad_norm": 1.6418712139129639, + "learning_rate": 7.732007580611946e-05, + "loss": 0.578, + "step": 33759 + }, + { + "epoch": 2.342654916383318, + "grad_norm": 1.1516703367233276, + "learning_rate": 7.73146036389745e-05, + "loss": 0.2842, + "step": 33760 + }, + { + "epoch": 2.342724307820415, + "grad_norm": 1.0291283130645752, + "learning_rate": 7.730913154344472e-05, + "loss": 0.351, + "step": 33761 + }, + { + "epoch": 2.342793699257512, + "grad_norm": 0.8285666108131409, + "learning_rate": 7.730365951954735e-05, + "loss": 0.2862, + "step": 33762 + }, + { + "epoch": 2.342863090694608, + "grad_norm": 1.1045407056808472, + "learning_rate": 7.72981875672997e-05, + "loss": 0.4762, + "step": 33763 + }, + { + "epoch": 2.342932482131705, + "grad_norm": 1.1665029525756836, + "learning_rate": 7.729271568671901e-05, + "loss": 0.4443, + "step": 33764 + }, + { + "epoch": 2.3430018735688014, + "grad_norm": 1.2300920486450195, + "learning_rate": 7.728724387782261e-05, + "loss": 0.5607, + "step": 33765 + }, + { + "epoch": 2.3430712650058982, + "grad_norm": 1.6456348896026611, + "learning_rate": 7.728177214062774e-05, + "loss": 0.4594, + "step": 33766 + }, + { + "epoch": 2.343140656442995, + "grad_norm": 1.8508965969085693, + "learning_rate": 7.727630047515167e-05, + "loss": 0.5626, + "step": 33767 + }, + { + "epoch": 2.3432100478800915, + "grad_norm": 0.9508097767829895, + "learning_rate": 7.727082888141173e-05, + "loss": 0.3881, + "step": 33768 + }, + { + "epoch": 2.3432794393171883, + "grad_norm": 1.3212831020355225, + "learning_rate": 7.72653573594251e-05, + "loss": 0.3738, + "step": 33769 + }, + { + "epoch": 2.3433488307542847, + "grad_norm": 2.064417600631714, + "learning_rate": 7.725988590920912e-05, + "loss": 0.5966, + "step": 33770 + }, + { + "epoch": 2.3434182221913815, + "grad_norm": 1.4960237741470337, + "learning_rate": 7.725441453078103e-05, + "loss": 0.5265, + "step": 33771 + }, + { + "epoch": 2.3434876136284783, + "grad_norm": 1.3737972974777222, + "learning_rate": 7.72489432241581e-05, + "loss": 0.4111, + "step": 33772 + }, + { + "epoch": 2.3435570050655747, + "grad_norm": 1.5820958614349365, + "learning_rate": 7.724347198935765e-05, + "loss": 0.446, + "step": 33773 + }, + { + "epoch": 2.3436263965026716, + "grad_norm": 1.2013832330703735, + "learning_rate": 7.723800082639689e-05, + "loss": 0.4372, + "step": 33774 + }, + { + "epoch": 2.3436957879397684, + "grad_norm": 1.4518197774887085, + "learning_rate": 7.723252973529316e-05, + "loss": 0.49, + "step": 33775 + }, + { + "epoch": 2.343765179376865, + "grad_norm": 0.8394473195075989, + "learning_rate": 7.722705871606367e-05, + "loss": 0.2397, + "step": 33776 + }, + { + "epoch": 2.3438345708139616, + "grad_norm": 1.5686285495758057, + "learning_rate": 7.722158776872571e-05, + "loss": 0.5275, + "step": 33777 + }, + { + "epoch": 2.3439039622510585, + "grad_norm": 1.611968755722046, + "learning_rate": 7.721611689329656e-05, + "loss": 0.4506, + "step": 33778 + }, + { + "epoch": 2.343973353688155, + "grad_norm": 1.5395708084106445, + "learning_rate": 7.721064608979346e-05, + "loss": 0.5039, + "step": 33779 + }, + { + "epoch": 2.3440427451252517, + "grad_norm": 1.3787052631378174, + "learning_rate": 7.720517535823372e-05, + "loss": 0.4733, + "step": 33780 + }, + { + "epoch": 2.344112136562348, + "grad_norm": 1.672821044921875, + "learning_rate": 7.719970469863459e-05, + "loss": 0.4534, + "step": 33781 + }, + { + "epoch": 2.344181527999445, + "grad_norm": 1.2838588953018188, + "learning_rate": 7.719423411101339e-05, + "loss": 0.493, + "step": 33782 + }, + { + "epoch": 2.3442509194365417, + "grad_norm": 1.1374202966690063, + "learning_rate": 7.718876359538731e-05, + "loss": 0.459, + "step": 33783 + }, + { + "epoch": 2.344320310873638, + "grad_norm": 1.292549967765808, + "learning_rate": 7.718329315177366e-05, + "loss": 0.3925, + "step": 33784 + }, + { + "epoch": 2.344389702310735, + "grad_norm": 1.2623355388641357, + "learning_rate": 7.71778227801897e-05, + "loss": 0.2639, + "step": 33785 + }, + { + "epoch": 2.3444590937478313, + "grad_norm": 1.3369933366775513, + "learning_rate": 7.717235248065271e-05, + "loss": 0.4713, + "step": 33786 + }, + { + "epoch": 2.344528485184928, + "grad_norm": 1.5967161655426025, + "learning_rate": 7.716688225317996e-05, + "loss": 0.5563, + "step": 33787 + }, + { + "epoch": 2.344597876622025, + "grad_norm": 1.4002758264541626, + "learning_rate": 7.716141209778874e-05, + "loss": 0.4918, + "step": 33788 + }, + { + "epoch": 2.3446672680591214, + "grad_norm": 1.212049961090088, + "learning_rate": 7.715594201449626e-05, + "loss": 0.3258, + "step": 33789 + }, + { + "epoch": 2.3447366594962182, + "grad_norm": 1.3154743909835815, + "learning_rate": 7.715047200331982e-05, + "loss": 0.4535, + "step": 33790 + }, + { + "epoch": 2.3448060509333146, + "grad_norm": 1.4091932773590088, + "learning_rate": 7.71450020642767e-05, + "loss": 0.3684, + "step": 33791 + }, + { + "epoch": 2.3448754423704115, + "grad_norm": 1.5282564163208008, + "learning_rate": 7.713953219738415e-05, + "loss": 0.4536, + "step": 33792 + }, + { + "epoch": 2.3449448338075083, + "grad_norm": 1.1848781108856201, + "learning_rate": 7.713406240265944e-05, + "loss": 0.3469, + "step": 33793 + }, + { + "epoch": 2.3450142252446047, + "grad_norm": 0.9957644939422607, + "learning_rate": 7.712859268011985e-05, + "loss": 0.2854, + "step": 33794 + }, + { + "epoch": 2.3450836166817015, + "grad_norm": 1.515021562576294, + "learning_rate": 7.712312302978265e-05, + "loss": 0.395, + "step": 33795 + }, + { + "epoch": 2.345153008118798, + "grad_norm": 1.2296180725097656, + "learning_rate": 7.71176534516651e-05, + "loss": 0.2883, + "step": 33796 + }, + { + "epoch": 2.3452223995558947, + "grad_norm": 0.9267192482948303, + "learning_rate": 7.711218394578443e-05, + "loss": 0.3151, + "step": 33797 + }, + { + "epoch": 2.3452917909929916, + "grad_norm": 1.0558686256408691, + "learning_rate": 7.710671451215797e-05, + "loss": 0.2907, + "step": 33798 + }, + { + "epoch": 2.345361182430088, + "grad_norm": 1.177201271057129, + "learning_rate": 7.710124515080293e-05, + "loss": 0.4643, + "step": 33799 + }, + { + "epoch": 2.345430573867185, + "grad_norm": 1.3626677989959717, + "learning_rate": 7.709577586173661e-05, + "loss": 0.2853, + "step": 33800 + }, + { + "epoch": 2.3454999653042816, + "grad_norm": 1.2531824111938477, + "learning_rate": 7.709030664497627e-05, + "loss": 0.4284, + "step": 33801 + }, + { + "epoch": 2.345569356741378, + "grad_norm": 1.1408782005310059, + "learning_rate": 7.708483750053921e-05, + "loss": 0.2073, + "step": 33802 + }, + { + "epoch": 2.345638748178475, + "grad_norm": 1.838193416595459, + "learning_rate": 7.70793684284426e-05, + "loss": 0.6291, + "step": 33803 + }, + { + "epoch": 2.3457081396155717, + "grad_norm": 1.5076292753219604, + "learning_rate": 7.70738994287038e-05, + "loss": 0.3616, + "step": 33804 + }, + { + "epoch": 2.345777531052668, + "grad_norm": 1.4256662130355835, + "learning_rate": 7.706843050134002e-05, + "loss": 0.481, + "step": 33805 + }, + { + "epoch": 2.345846922489765, + "grad_norm": 1.2065327167510986, + "learning_rate": 7.706296164636855e-05, + "loss": 0.4416, + "step": 33806 + }, + { + "epoch": 2.3459163139268613, + "grad_norm": 2.51505970954895, + "learning_rate": 7.705749286380664e-05, + "loss": 0.4125, + "step": 33807 + }, + { + "epoch": 2.345985705363958, + "grad_norm": 2.262916088104248, + "learning_rate": 7.705202415367161e-05, + "loss": 0.4905, + "step": 33808 + }, + { + "epoch": 2.346055096801055, + "grad_norm": 1.0945028066635132, + "learning_rate": 7.704655551598066e-05, + "loss": 0.33, + "step": 33809 + }, + { + "epoch": 2.3461244882381513, + "grad_norm": 1.4155348539352417, + "learning_rate": 7.704108695075104e-05, + "loss": 0.392, + "step": 33810 + }, + { + "epoch": 2.346193879675248, + "grad_norm": 1.4548075199127197, + "learning_rate": 7.703561845800006e-05, + "loss": 0.5879, + "step": 33811 + }, + { + "epoch": 2.3462632711123446, + "grad_norm": 1.5398958921432495, + "learning_rate": 7.703015003774498e-05, + "loss": 0.4469, + "step": 33812 + }, + { + "epoch": 2.3463326625494414, + "grad_norm": 1.1023956537246704, + "learning_rate": 7.702468169000303e-05, + "loss": 0.3843, + "step": 33813 + }, + { + "epoch": 2.3464020539865382, + "grad_norm": 0.8720857501029968, + "learning_rate": 7.70192134147915e-05, + "loss": 0.2704, + "step": 33814 + }, + { + "epoch": 2.3464714454236346, + "grad_norm": 1.3449053764343262, + "learning_rate": 7.701374521212765e-05, + "loss": 0.3806, + "step": 33815 + }, + { + "epoch": 2.3465408368607314, + "grad_norm": 1.2315913438796997, + "learning_rate": 7.700827708202877e-05, + "loss": 0.6005, + "step": 33816 + }, + { + "epoch": 2.346610228297828, + "grad_norm": 1.0729799270629883, + "learning_rate": 7.700280902451206e-05, + "loss": 0.4064, + "step": 33817 + }, + { + "epoch": 2.3466796197349247, + "grad_norm": 1.1808420419692993, + "learning_rate": 7.699734103959481e-05, + "loss": 0.4757, + "step": 33818 + }, + { + "epoch": 2.3467490111720215, + "grad_norm": 1.1023837327957153, + "learning_rate": 7.699187312729428e-05, + "loss": 0.3997, + "step": 33819 + }, + { + "epoch": 2.346818402609118, + "grad_norm": 1.1126587390899658, + "learning_rate": 7.698640528762776e-05, + "loss": 0.4026, + "step": 33820 + }, + { + "epoch": 2.3468877940462147, + "grad_norm": 1.4675319194793701, + "learning_rate": 7.698093752061246e-05, + "loss": 0.5442, + "step": 33821 + }, + { + "epoch": 2.346957185483311, + "grad_norm": 1.3274866342544556, + "learning_rate": 7.697546982626568e-05, + "loss": 0.4418, + "step": 33822 + }, + { + "epoch": 2.347026576920408, + "grad_norm": 1.8009456396102905, + "learning_rate": 7.697000220460471e-05, + "loss": 0.4171, + "step": 33823 + }, + { + "epoch": 2.347095968357505, + "grad_norm": 1.1057970523834229, + "learning_rate": 7.696453465564673e-05, + "loss": 0.4891, + "step": 33824 + }, + { + "epoch": 2.347165359794601, + "grad_norm": 1.0808275938034058, + "learning_rate": 7.695906717940904e-05, + "loss": 0.4715, + "step": 33825 + }, + { + "epoch": 2.347234751231698, + "grad_norm": 1.2816179990768433, + "learning_rate": 7.69535997759089e-05, + "loss": 0.3739, + "step": 33826 + }, + { + "epoch": 2.347304142668795, + "grad_norm": 1.3525094985961914, + "learning_rate": 7.694813244516361e-05, + "loss": 0.3478, + "step": 33827 + }, + { + "epoch": 2.3473735341058912, + "grad_norm": 1.436128854751587, + "learning_rate": 7.694266518719036e-05, + "loss": 0.6278, + "step": 33828 + }, + { + "epoch": 2.347442925542988, + "grad_norm": 1.3520902395248413, + "learning_rate": 7.693719800200643e-05, + "loss": 0.4818, + "step": 33829 + }, + { + "epoch": 2.347512316980085, + "grad_norm": 0.8301291465759277, + "learning_rate": 7.693173088962913e-05, + "loss": 0.2703, + "step": 33830 + }, + { + "epoch": 2.3475817084171813, + "grad_norm": 1.358359932899475, + "learning_rate": 7.692626385007566e-05, + "loss": 0.3826, + "step": 33831 + }, + { + "epoch": 2.347651099854278, + "grad_norm": 1.9969803094863892, + "learning_rate": 7.69207968833633e-05, + "loss": 0.6497, + "step": 33832 + }, + { + "epoch": 2.3477204912913745, + "grad_norm": 1.4180587530136108, + "learning_rate": 7.691532998950927e-05, + "loss": 0.5169, + "step": 33833 + }, + { + "epoch": 2.3477898827284713, + "grad_norm": 1.718238353729248, + "learning_rate": 7.69098631685309e-05, + "loss": 0.4754, + "step": 33834 + }, + { + "epoch": 2.347859274165568, + "grad_norm": 1.2996960878372192, + "learning_rate": 7.69043964204454e-05, + "loss": 0.6054, + "step": 33835 + }, + { + "epoch": 2.3479286656026646, + "grad_norm": 1.5978994369506836, + "learning_rate": 7.689892974527004e-05, + "loss": 0.5017, + "step": 33836 + }, + { + "epoch": 2.3479980570397614, + "grad_norm": 1.363526463508606, + "learning_rate": 7.689346314302213e-05, + "loss": 0.5597, + "step": 33837 + }, + { + "epoch": 2.3480674484768578, + "grad_norm": 1.5303618907928467, + "learning_rate": 7.688799661371882e-05, + "loss": 0.5308, + "step": 33838 + }, + { + "epoch": 2.3481368399139546, + "grad_norm": 1.2151094675064087, + "learning_rate": 7.688253015737743e-05, + "loss": 0.3889, + "step": 33839 + }, + { + "epoch": 2.3482062313510514, + "grad_norm": 1.3171030282974243, + "learning_rate": 7.68770637740152e-05, + "loss": 0.4205, + "step": 33840 + }, + { + "epoch": 2.348275622788148, + "grad_norm": 1.67574143409729, + "learning_rate": 7.687159746364941e-05, + "loss": 0.5, + "step": 33841 + }, + { + "epoch": 2.3483450142252447, + "grad_norm": 1.444043517112732, + "learning_rate": 7.68661312262973e-05, + "loss": 0.4855, + "step": 33842 + }, + { + "epoch": 2.348414405662341, + "grad_norm": 1.3182884454727173, + "learning_rate": 7.686066506197616e-05, + "loss": 0.399, + "step": 33843 + }, + { + "epoch": 2.348483797099438, + "grad_norm": 1.0591243505477905, + "learning_rate": 7.68551989707032e-05, + "loss": 0.2879, + "step": 33844 + }, + { + "epoch": 2.3485531885365347, + "grad_norm": 1.225417971611023, + "learning_rate": 7.684973295249568e-05, + "loss": 0.649, + "step": 33845 + }, + { + "epoch": 2.348622579973631, + "grad_norm": 1.910524606704712, + "learning_rate": 7.684426700737087e-05, + "loss": 0.5315, + "step": 33846 + }, + { + "epoch": 2.348691971410728, + "grad_norm": 1.1448757648468018, + "learning_rate": 7.683880113534602e-05, + "loss": 0.364, + "step": 33847 + }, + { + "epoch": 2.3487613628478243, + "grad_norm": 1.1359654664993286, + "learning_rate": 7.68333353364384e-05, + "loss": 0.4229, + "step": 33848 + }, + { + "epoch": 2.348830754284921, + "grad_norm": 1.3131545782089233, + "learning_rate": 7.682786961066523e-05, + "loss": 0.3427, + "step": 33849 + }, + { + "epoch": 2.348900145722018, + "grad_norm": 1.15583336353302, + "learning_rate": 7.682240395804379e-05, + "loss": 0.5199, + "step": 33850 + }, + { + "epoch": 2.3489695371591144, + "grad_norm": 1.0705138444900513, + "learning_rate": 7.681693837859139e-05, + "loss": 0.3523, + "step": 33851 + }, + { + "epoch": 2.349038928596211, + "grad_norm": 1.2539647817611694, + "learning_rate": 7.681147287232517e-05, + "loss": 0.5789, + "step": 33852 + }, + { + "epoch": 2.349108320033308, + "grad_norm": 1.769384741783142, + "learning_rate": 7.680600743926243e-05, + "loss": 0.4872, + "step": 33853 + }, + { + "epoch": 2.3491777114704044, + "grad_norm": 1.6396820545196533, + "learning_rate": 7.680054207942045e-05, + "loss": 0.4173, + "step": 33854 + }, + { + "epoch": 2.3492471029075013, + "grad_norm": 1.3699657917022705, + "learning_rate": 7.679507679281646e-05, + "loss": 0.2835, + "step": 33855 + }, + { + "epoch": 2.349316494344598, + "grad_norm": 1.3773465156555176, + "learning_rate": 7.678961157946772e-05, + "loss": 0.4286, + "step": 33856 + }, + { + "epoch": 2.3493858857816945, + "grad_norm": 1.5964276790618896, + "learning_rate": 7.678414643939149e-05, + "loss": 0.3172, + "step": 33857 + }, + { + "epoch": 2.3494552772187913, + "grad_norm": 3.3195154666900635, + "learning_rate": 7.677868137260504e-05, + "loss": 0.5066, + "step": 33858 + }, + { + "epoch": 2.3495246686558877, + "grad_norm": 1.3739705085754395, + "learning_rate": 7.677321637912557e-05, + "loss": 0.3968, + "step": 33859 + }, + { + "epoch": 2.3495940600929845, + "grad_norm": 1.2249120473861694, + "learning_rate": 7.676775145897037e-05, + "loss": 0.3464, + "step": 33860 + }, + { + "epoch": 2.3496634515300814, + "grad_norm": 1.29086172580719, + "learning_rate": 7.676228661215668e-05, + "loss": 0.353, + "step": 33861 + }, + { + "epoch": 2.3497328429671778, + "grad_norm": 1.1468794345855713, + "learning_rate": 7.675682183870177e-05, + "loss": 0.3293, + "step": 33862 + }, + { + "epoch": 2.3498022344042746, + "grad_norm": 1.2691407203674316, + "learning_rate": 7.675135713862284e-05, + "loss": 0.3508, + "step": 33863 + }, + { + "epoch": 2.349871625841371, + "grad_norm": 1.2371293306350708, + "learning_rate": 7.674589251193722e-05, + "loss": 0.3265, + "step": 33864 + }, + { + "epoch": 2.349941017278468, + "grad_norm": 1.2016371488571167, + "learning_rate": 7.674042795866209e-05, + "loss": 0.333, + "step": 33865 + }, + { + "epoch": 2.3500104087155647, + "grad_norm": 0.9670240879058838, + "learning_rate": 7.673496347881477e-05, + "loss": 0.2531, + "step": 33866 + }, + { + "epoch": 2.350079800152661, + "grad_norm": 1.3109289407730103, + "learning_rate": 7.672949907241243e-05, + "loss": 0.4152, + "step": 33867 + }, + { + "epoch": 2.350149191589758, + "grad_norm": 1.3848485946655273, + "learning_rate": 7.672403473947236e-05, + "loss": 0.3838, + "step": 33868 + }, + { + "epoch": 2.3502185830268543, + "grad_norm": 1.566096305847168, + "learning_rate": 7.671857048001183e-05, + "loss": 0.5573, + "step": 33869 + }, + { + "epoch": 2.350287974463951, + "grad_norm": 1.1953564882278442, + "learning_rate": 7.671310629404805e-05, + "loss": 0.3836, + "step": 33870 + }, + { + "epoch": 2.350357365901048, + "grad_norm": 1.6238328218460083, + "learning_rate": 7.670764218159835e-05, + "loss": 0.4035, + "step": 33871 + }, + { + "epoch": 2.3504267573381443, + "grad_norm": 0.8282480835914612, + "learning_rate": 7.670217814267987e-05, + "loss": 0.3229, + "step": 33872 + }, + { + "epoch": 2.350496148775241, + "grad_norm": 1.3606116771697998, + "learning_rate": 7.669671417730991e-05, + "loss": 0.5775, + "step": 33873 + }, + { + "epoch": 2.350565540212338, + "grad_norm": 1.5037286281585693, + "learning_rate": 7.669125028550573e-05, + "loss": 0.5731, + "step": 33874 + }, + { + "epoch": 2.3506349316494344, + "grad_norm": 1.1151434183120728, + "learning_rate": 7.668578646728456e-05, + "loss": 0.3133, + "step": 33875 + }, + { + "epoch": 2.350704323086531, + "grad_norm": 1.267901062965393, + "learning_rate": 7.668032272266366e-05, + "loss": 0.4041, + "step": 33876 + }, + { + "epoch": 2.3507737145236276, + "grad_norm": 1.2975735664367676, + "learning_rate": 7.667485905166027e-05, + "loss": 0.3934, + "step": 33877 + }, + { + "epoch": 2.3508431059607244, + "grad_norm": 1.7395278215408325, + "learning_rate": 7.666939545429169e-05, + "loss": 0.4644, + "step": 33878 + }, + { + "epoch": 2.3509124973978213, + "grad_norm": 1.025269865989685, + "learning_rate": 7.666393193057509e-05, + "loss": 0.3608, + "step": 33879 + }, + { + "epoch": 2.3509818888349177, + "grad_norm": 1.0863285064697266, + "learning_rate": 7.665846848052774e-05, + "loss": 0.3456, + "step": 33880 + }, + { + "epoch": 2.3510512802720145, + "grad_norm": 1.5212205648422241, + "learning_rate": 7.665300510416691e-05, + "loss": 0.4626, + "step": 33881 + }, + { + "epoch": 2.3511206717091113, + "grad_norm": 2.2438836097717285, + "learning_rate": 7.664754180150983e-05, + "loss": 0.3452, + "step": 33882 + }, + { + "epoch": 2.3511900631462077, + "grad_norm": 0.931721568107605, + "learning_rate": 7.664207857257373e-05, + "loss": 0.3685, + "step": 33883 + }, + { + "epoch": 2.3512594545833045, + "grad_norm": 1.2129582166671753, + "learning_rate": 7.663661541737589e-05, + "loss": 0.3384, + "step": 33884 + }, + { + "epoch": 2.351328846020401, + "grad_norm": 1.6623613834381104, + "learning_rate": 7.663115233593359e-05, + "loss": 0.3818, + "step": 33885 + }, + { + "epoch": 2.3513982374574978, + "grad_norm": 1.1735334396362305, + "learning_rate": 7.662568932826398e-05, + "loss": 0.4268, + "step": 33886 + }, + { + "epoch": 2.3514676288945946, + "grad_norm": 1.2974700927734375, + "learning_rate": 7.662022639438437e-05, + "loss": 0.4169, + "step": 33887 + }, + { + "epoch": 2.351537020331691, + "grad_norm": 1.4549981355667114, + "learning_rate": 7.661476353431198e-05, + "loss": 0.3372, + "step": 33888 + }, + { + "epoch": 2.351606411768788, + "grad_norm": 1.8457468748092651, + "learning_rate": 7.660930074806408e-05, + "loss": 0.5453, + "step": 33889 + }, + { + "epoch": 2.351675803205884, + "grad_norm": 1.149165153503418, + "learning_rate": 7.660383803565789e-05, + "loss": 0.4212, + "step": 33890 + }, + { + "epoch": 2.351745194642981, + "grad_norm": 3.4020919799804688, + "learning_rate": 7.659837539711068e-05, + "loss": 0.5718, + "step": 33891 + }, + { + "epoch": 2.351814586080078, + "grad_norm": 1.6781229972839355, + "learning_rate": 7.659291283243972e-05, + "loss": 0.4746, + "step": 33892 + }, + { + "epoch": 2.3518839775171743, + "grad_norm": 1.2640684843063354, + "learning_rate": 7.658745034166217e-05, + "loss": 0.5135, + "step": 33893 + }, + { + "epoch": 2.351953368954271, + "grad_norm": 1.1220817565917969, + "learning_rate": 7.658198792479535e-05, + "loss": 0.4132, + "step": 33894 + }, + { + "epoch": 2.3520227603913675, + "grad_norm": 1.3412822484970093, + "learning_rate": 7.657652558185644e-05, + "loss": 0.3902, + "step": 33895 + }, + { + "epoch": 2.3520921518284643, + "grad_norm": 1.0909243822097778, + "learning_rate": 7.657106331286275e-05, + "loss": 0.4976, + "step": 33896 + }, + { + "epoch": 2.352161543265561, + "grad_norm": 1.3803654909133911, + "learning_rate": 7.656560111783149e-05, + "loss": 0.3741, + "step": 33897 + }, + { + "epoch": 2.3522309347026575, + "grad_norm": 2.127451181411743, + "learning_rate": 7.65601389967799e-05, + "loss": 0.5574, + "step": 33898 + }, + { + "epoch": 2.3523003261397544, + "grad_norm": 0.9888333678245544, + "learning_rate": 7.655467694972527e-05, + "loss": 0.2586, + "step": 33899 + }, + { + "epoch": 2.352369717576851, + "grad_norm": 1.3062553405761719, + "learning_rate": 7.654921497668476e-05, + "loss": 0.3878, + "step": 33900 + }, + { + "epoch": 2.3524391090139476, + "grad_norm": 1.345881462097168, + "learning_rate": 7.65437530776757e-05, + "loss": 0.5611, + "step": 33901 + }, + { + "epoch": 2.3525085004510444, + "grad_norm": 1.3552546501159668, + "learning_rate": 7.653829125271524e-05, + "loss": 0.3261, + "step": 33902 + }, + { + "epoch": 2.352577891888141, + "grad_norm": 1.1846246719360352, + "learning_rate": 7.653282950182068e-05, + "loss": 0.533, + "step": 33903 + }, + { + "epoch": 2.3526472833252376, + "grad_norm": 1.5688755512237549, + "learning_rate": 7.652736782500928e-05, + "loss": 0.457, + "step": 33904 + }, + { + "epoch": 2.3527166747623345, + "grad_norm": 1.1580321788787842, + "learning_rate": 7.652190622229823e-05, + "loss": 0.3869, + "step": 33905 + }, + { + "epoch": 2.352786066199431, + "grad_norm": 1.5511912107467651, + "learning_rate": 7.651644469370486e-05, + "loss": 0.4786, + "step": 33906 + }, + { + "epoch": 2.3528554576365277, + "grad_norm": 1.5888688564300537, + "learning_rate": 7.651098323924629e-05, + "loss": 0.4603, + "step": 33907 + }, + { + "epoch": 2.3529248490736245, + "grad_norm": 0.9880688190460205, + "learning_rate": 7.650552185893984e-05, + "loss": 0.3465, + "step": 33908 + }, + { + "epoch": 2.352994240510721, + "grad_norm": 1.6179773807525635, + "learning_rate": 7.650006055280271e-05, + "loss": 0.3259, + "step": 33909 + }, + { + "epoch": 2.3530636319478178, + "grad_norm": 1.3821312189102173, + "learning_rate": 7.649459932085217e-05, + "loss": 0.3016, + "step": 33910 + }, + { + "epoch": 2.353133023384914, + "grad_norm": 1.7478022575378418, + "learning_rate": 7.648913816310546e-05, + "loss": 0.4933, + "step": 33911 + }, + { + "epoch": 2.353202414822011, + "grad_norm": 1.2114205360412598, + "learning_rate": 7.64836770795798e-05, + "loss": 0.4498, + "step": 33912 + }, + { + "epoch": 2.353271806259108, + "grad_norm": 1.0408121347427368, + "learning_rate": 7.64782160702925e-05, + "loss": 0.4905, + "step": 33913 + }, + { + "epoch": 2.353341197696204, + "grad_norm": 2.710927724838257, + "learning_rate": 7.647275513526069e-05, + "loss": 0.4888, + "step": 33914 + }, + { + "epoch": 2.353410589133301, + "grad_norm": 1.245882511138916, + "learning_rate": 7.646729427450167e-05, + "loss": 0.5459, + "step": 33915 + }, + { + "epoch": 2.3534799805703974, + "grad_norm": 1.4595106840133667, + "learning_rate": 7.646183348803269e-05, + "loss": 0.4505, + "step": 33916 + }, + { + "epoch": 2.3535493720074943, + "grad_norm": 1.3779813051223755, + "learning_rate": 7.645637277587095e-05, + "loss": 0.4887, + "step": 33917 + }, + { + "epoch": 2.353618763444591, + "grad_norm": 1.3224971294403076, + "learning_rate": 7.64509121380337e-05, + "loss": 0.3408, + "step": 33918 + }, + { + "epoch": 2.3536881548816875, + "grad_norm": 1.3438886404037476, + "learning_rate": 7.64454515745382e-05, + "loss": 0.5921, + "step": 33919 + }, + { + "epoch": 2.3537575463187843, + "grad_norm": 1.1351873874664307, + "learning_rate": 7.643999108540171e-05, + "loss": 0.4337, + "step": 33920 + }, + { + "epoch": 2.3538269377558807, + "grad_norm": 1.7004027366638184, + "learning_rate": 7.643453067064139e-05, + "loss": 0.6224, + "step": 33921 + }, + { + "epoch": 2.3538963291929775, + "grad_norm": 1.2106629610061646, + "learning_rate": 7.642907033027453e-05, + "loss": 0.4997, + "step": 33922 + }, + { + "epoch": 2.3539657206300744, + "grad_norm": 0.9323424100875854, + "learning_rate": 7.642361006431837e-05, + "loss": 0.3979, + "step": 33923 + }, + { + "epoch": 2.3540351120671708, + "grad_norm": 1.346266508102417, + "learning_rate": 7.641814987279012e-05, + "loss": 0.3255, + "step": 33924 + }, + { + "epoch": 2.3541045035042676, + "grad_norm": 1.0620393753051758, + "learning_rate": 7.641268975570705e-05, + "loss": 0.3182, + "step": 33925 + }, + { + "epoch": 2.3541738949413644, + "grad_norm": 1.4930137395858765, + "learning_rate": 7.640722971308636e-05, + "loss": 0.5758, + "step": 33926 + }, + { + "epoch": 2.354243286378461, + "grad_norm": 1.3671919107437134, + "learning_rate": 7.640176974494537e-05, + "loss": 0.3969, + "step": 33927 + }, + { + "epoch": 2.3543126778155576, + "grad_norm": 1.227736473083496, + "learning_rate": 7.639630985130121e-05, + "loss": 0.4634, + "step": 33928 + }, + { + "epoch": 2.354382069252654, + "grad_norm": 1.4837970733642578, + "learning_rate": 7.639085003217116e-05, + "loss": 0.5257, + "step": 33929 + }, + { + "epoch": 2.354451460689751, + "grad_norm": 1.2761366367340088, + "learning_rate": 7.638539028757245e-05, + "loss": 0.5718, + "step": 33930 + }, + { + "epoch": 2.3545208521268477, + "grad_norm": 0.9084478616714478, + "learning_rate": 7.637993061752233e-05, + "loss": 0.3608, + "step": 33931 + }, + { + "epoch": 2.354590243563944, + "grad_norm": 1.8642767667770386, + "learning_rate": 7.637447102203805e-05, + "loss": 0.5135, + "step": 33932 + }, + { + "epoch": 2.354659635001041, + "grad_norm": 1.1708765029907227, + "learning_rate": 7.636901150113681e-05, + "loss": 0.4512, + "step": 33933 + }, + { + "epoch": 2.3547290264381378, + "grad_norm": 1.3692305088043213, + "learning_rate": 7.636355205483587e-05, + "loss": 0.3973, + "step": 33934 + }, + { + "epoch": 2.354798417875234, + "grad_norm": 0.9614309072494507, + "learning_rate": 7.635809268315246e-05, + "loss": 0.2899, + "step": 33935 + }, + { + "epoch": 2.354867809312331, + "grad_norm": 1.4747493267059326, + "learning_rate": 7.635263338610378e-05, + "loss": 0.572, + "step": 33936 + }, + { + "epoch": 2.3549372007494274, + "grad_norm": 1.1733286380767822, + "learning_rate": 7.634717416370709e-05, + "loss": 0.3089, + "step": 33937 + }, + { + "epoch": 2.355006592186524, + "grad_norm": 1.023634433746338, + "learning_rate": 7.634171501597964e-05, + "loss": 0.3324, + "step": 33938 + }, + { + "epoch": 2.355075983623621, + "grad_norm": 0.850558340549469, + "learning_rate": 7.633625594293865e-05, + "loss": 0.3652, + "step": 33939 + }, + { + "epoch": 2.3551453750607174, + "grad_norm": 2.1508283615112305, + "learning_rate": 7.633079694460139e-05, + "loss": 0.5184, + "step": 33940 + }, + { + "epoch": 2.3552147664978142, + "grad_norm": 1.81411874294281, + "learning_rate": 7.632533802098501e-05, + "loss": 0.5047, + "step": 33941 + }, + { + "epoch": 2.3552841579349106, + "grad_norm": 1.516008734703064, + "learning_rate": 7.63198791721068e-05, + "loss": 0.3242, + "step": 33942 + }, + { + "epoch": 2.3553535493720075, + "grad_norm": 1.2473646402359009, + "learning_rate": 7.631442039798398e-05, + "loss": 0.5218, + "step": 33943 + }, + { + "epoch": 2.3554229408091043, + "grad_norm": 2.068053722381592, + "learning_rate": 7.630896169863379e-05, + "loss": 0.5655, + "step": 33944 + }, + { + "epoch": 2.3554923322462007, + "grad_norm": 1.3241266012191772, + "learning_rate": 7.630350307407345e-05, + "loss": 0.4826, + "step": 33945 + }, + { + "epoch": 2.3555617236832975, + "grad_norm": 1.53287935256958, + "learning_rate": 7.62980445243202e-05, + "loss": 0.419, + "step": 33946 + }, + { + "epoch": 2.355631115120394, + "grad_norm": 1.12868070602417, + "learning_rate": 7.629258604939133e-05, + "loss": 0.4526, + "step": 33947 + }, + { + "epoch": 2.3557005065574907, + "grad_norm": 1.4481290578842163, + "learning_rate": 7.628712764930397e-05, + "loss": 0.3874, + "step": 33948 + }, + { + "epoch": 2.3557698979945876, + "grad_norm": 1.2080453634262085, + "learning_rate": 7.62816693240754e-05, + "loss": 0.5935, + "step": 33949 + }, + { + "epoch": 2.355839289431684, + "grad_norm": 1.3916347026824951, + "learning_rate": 7.627621107372284e-05, + "loss": 0.3714, + "step": 33950 + }, + { + "epoch": 2.355908680868781, + "grad_norm": 1.36257004737854, + "learning_rate": 7.627075289826355e-05, + "loss": 0.4153, + "step": 33951 + }, + { + "epoch": 2.3559780723058776, + "grad_norm": 1.3736659288406372, + "learning_rate": 7.626529479771472e-05, + "loss": 0.4211, + "step": 33952 + }, + { + "epoch": 2.356047463742974, + "grad_norm": 1.5181387662887573, + "learning_rate": 7.625983677209361e-05, + "loss": 0.6058, + "step": 33953 + }, + { + "epoch": 2.356116855180071, + "grad_norm": 1.0380409955978394, + "learning_rate": 7.625437882141746e-05, + "loss": 0.3471, + "step": 33954 + }, + { + "epoch": 2.3561862466171677, + "grad_norm": 1.3676069974899292, + "learning_rate": 7.624892094570348e-05, + "loss": 0.3121, + "step": 33955 + }, + { + "epoch": 2.356255638054264, + "grad_norm": 1.3691574335098267, + "learning_rate": 7.624346314496886e-05, + "loss": 0.4177, + "step": 33956 + }, + { + "epoch": 2.356325029491361, + "grad_norm": 1.17013418674469, + "learning_rate": 7.62380054192309e-05, + "loss": 0.3329, + "step": 33957 + }, + { + "epoch": 2.3563944209284573, + "grad_norm": 1.557350516319275, + "learning_rate": 7.623254776850678e-05, + "loss": 0.5587, + "step": 33958 + }, + { + "epoch": 2.356463812365554, + "grad_norm": 1.030097246170044, + "learning_rate": 7.622709019281375e-05, + "loss": 0.3488, + "step": 33959 + }, + { + "epoch": 2.356533203802651, + "grad_norm": 1.6869257688522339, + "learning_rate": 7.622163269216904e-05, + "loss": 0.4765, + "step": 33960 + }, + { + "epoch": 2.3566025952397474, + "grad_norm": 1.5980029106140137, + "learning_rate": 7.621617526658993e-05, + "loss": 0.5961, + "step": 33961 + }, + { + "epoch": 2.356671986676844, + "grad_norm": 1.7066740989685059, + "learning_rate": 7.621071791609355e-05, + "loss": 0.6419, + "step": 33962 + }, + { + "epoch": 2.3567413781139406, + "grad_norm": 1.4019702672958374, + "learning_rate": 7.620526064069717e-05, + "loss": 0.5048, + "step": 33963 + }, + { + "epoch": 2.3568107695510374, + "grad_norm": 1.5694472789764404, + "learning_rate": 7.619980344041802e-05, + "loss": 0.4419, + "step": 33964 + }, + { + "epoch": 2.3568801609881342, + "grad_norm": 1.1854139566421509, + "learning_rate": 7.619434631527332e-05, + "loss": 0.316, + "step": 33965 + }, + { + "epoch": 2.3569495524252306, + "grad_norm": 0.969809889793396, + "learning_rate": 7.618888926528032e-05, + "loss": 0.2438, + "step": 33966 + }, + { + "epoch": 2.3570189438623275, + "grad_norm": 1.401208758354187, + "learning_rate": 7.618343229045625e-05, + "loss": 0.4332, + "step": 33967 + }, + { + "epoch": 2.357088335299424, + "grad_norm": 1.3213462829589844, + "learning_rate": 7.617797539081832e-05, + "loss": 0.4155, + "step": 33968 + }, + { + "epoch": 2.3571577267365207, + "grad_norm": 1.4208015203475952, + "learning_rate": 7.617251856638374e-05, + "loss": 0.4477, + "step": 33969 + }, + { + "epoch": 2.3572271181736175, + "grad_norm": 1.0793862342834473, + "learning_rate": 7.616706181716977e-05, + "loss": 0.3333, + "step": 33970 + }, + { + "epoch": 2.357296509610714, + "grad_norm": 0.8776722550392151, + "learning_rate": 7.61616051431936e-05, + "loss": 0.4327, + "step": 33971 + }, + { + "epoch": 2.3573659010478107, + "grad_norm": 1.1144993305206299, + "learning_rate": 7.615614854447248e-05, + "loss": 0.3467, + "step": 33972 + }, + { + "epoch": 2.357435292484907, + "grad_norm": 1.5197770595550537, + "learning_rate": 7.615069202102364e-05, + "loss": 0.5226, + "step": 33973 + }, + { + "epoch": 2.357504683922004, + "grad_norm": 1.205243468284607, + "learning_rate": 7.61452355728643e-05, + "loss": 0.3583, + "step": 33974 + }, + { + "epoch": 2.357574075359101, + "grad_norm": 0.9606473445892334, + "learning_rate": 7.613977920001171e-05, + "loss": 0.2286, + "step": 33975 + }, + { + "epoch": 2.357643466796197, + "grad_norm": 1.281595230102539, + "learning_rate": 7.613432290248303e-05, + "loss": 0.3681, + "step": 33976 + }, + { + "epoch": 2.357712858233294, + "grad_norm": 1.3046945333480835, + "learning_rate": 7.612886668029552e-05, + "loss": 0.3229, + "step": 33977 + }, + { + "epoch": 2.357782249670391, + "grad_norm": 1.444921851158142, + "learning_rate": 7.612341053346642e-05, + "loss": 0.441, + "step": 33978 + }, + { + "epoch": 2.3578516411074872, + "grad_norm": 1.3323216438293457, + "learning_rate": 7.611795446201294e-05, + "loss": 0.4965, + "step": 33979 + }, + { + "epoch": 2.357921032544584, + "grad_norm": 0.8621116876602173, + "learning_rate": 7.61124984659523e-05, + "loss": 0.2594, + "step": 33980 + }, + { + "epoch": 2.357990423981681, + "grad_norm": 1.5452065467834473, + "learning_rate": 7.610704254530175e-05, + "loss": 0.3527, + "step": 33981 + }, + { + "epoch": 2.3580598154187773, + "grad_norm": 0.9930751919746399, + "learning_rate": 7.610158670007852e-05, + "loss": 0.3117, + "step": 33982 + }, + { + "epoch": 2.358129206855874, + "grad_norm": 1.374194622039795, + "learning_rate": 7.609613093029978e-05, + "loss": 0.4022, + "step": 33983 + }, + { + "epoch": 2.3581985982929705, + "grad_norm": 0.9468837380409241, + "learning_rate": 7.609067523598275e-05, + "loss": 0.4521, + "step": 33984 + }, + { + "epoch": 2.3582679897300673, + "grad_norm": 1.0756714344024658, + "learning_rate": 7.608521961714472e-05, + "loss": 0.3232, + "step": 33985 + }, + { + "epoch": 2.358337381167164, + "grad_norm": 1.2121405601501465, + "learning_rate": 7.607976407380289e-05, + "loss": 0.4606, + "step": 33986 + }, + { + "epoch": 2.3584067726042606, + "grad_norm": 1.5272607803344727, + "learning_rate": 7.607430860597444e-05, + "loss": 0.3905, + "step": 33987 + }, + { + "epoch": 2.3584761640413574, + "grad_norm": 1.3469923734664917, + "learning_rate": 7.606885321367661e-05, + "loss": 0.5592, + "step": 33988 + }, + { + "epoch": 2.358545555478454, + "grad_norm": 1.8260228633880615, + "learning_rate": 7.606339789692666e-05, + "loss": 0.6192, + "step": 33989 + }, + { + "epoch": 2.3586149469155506, + "grad_norm": 1.2923599481582642, + "learning_rate": 7.60579426557418e-05, + "loss": 0.4932, + "step": 33990 + }, + { + "epoch": 2.3586843383526475, + "grad_norm": 1.332470417022705, + "learning_rate": 7.605248749013921e-05, + "loss": 0.4092, + "step": 33991 + }, + { + "epoch": 2.358753729789744, + "grad_norm": 1.6868067979812622, + "learning_rate": 7.604703240013615e-05, + "loss": 0.4195, + "step": 33992 + }, + { + "epoch": 2.3588231212268407, + "grad_norm": 1.26604163646698, + "learning_rate": 7.60415773857498e-05, + "loss": 0.5349, + "step": 33993 + }, + { + "epoch": 2.358892512663937, + "grad_norm": 1.492171049118042, + "learning_rate": 7.603612244699744e-05, + "loss": 0.4289, + "step": 33994 + }, + { + "epoch": 2.358961904101034, + "grad_norm": 1.0089348554611206, + "learning_rate": 7.603066758389626e-05, + "loss": 0.4661, + "step": 33995 + }, + { + "epoch": 2.3590312955381307, + "grad_norm": 1.240250587463379, + "learning_rate": 7.60252127964635e-05, + "loss": 0.4496, + "step": 33996 + }, + { + "epoch": 2.359100686975227, + "grad_norm": 1.2524569034576416, + "learning_rate": 7.601975808471633e-05, + "loss": 0.4393, + "step": 33997 + }, + { + "epoch": 2.359170078412324, + "grad_norm": 2.1934010982513428, + "learning_rate": 7.601430344867201e-05, + "loss": 0.3993, + "step": 33998 + }, + { + "epoch": 2.3592394698494203, + "grad_norm": 1.6080858707427979, + "learning_rate": 7.600884888834774e-05, + "loss": 0.6296, + "step": 33999 + }, + { + "epoch": 2.359308861286517, + "grad_norm": 1.254793643951416, + "learning_rate": 7.600339440376076e-05, + "loss": 0.3877, + "step": 34000 + }, + { + "epoch": 2.359378252723614, + "grad_norm": 1.1225438117980957, + "learning_rate": 7.599793999492827e-05, + "loss": 0.3124, + "step": 34001 + }, + { + "epoch": 2.3594476441607104, + "grad_norm": 1.2864409685134888, + "learning_rate": 7.599248566186753e-05, + "loss": 0.4996, + "step": 34002 + }, + { + "epoch": 2.3595170355978072, + "grad_norm": 1.3475546836853027, + "learning_rate": 7.598703140459574e-05, + "loss": 0.3437, + "step": 34003 + }, + { + "epoch": 2.359586427034904, + "grad_norm": 1.3609097003936768, + "learning_rate": 7.598157722313006e-05, + "loss": 0.4711, + "step": 34004 + }, + { + "epoch": 2.3596558184720005, + "grad_norm": 1.6006335020065308, + "learning_rate": 7.597612311748777e-05, + "loss": 0.3202, + "step": 34005 + }, + { + "epoch": 2.3597252099090973, + "grad_norm": 1.266370177268982, + "learning_rate": 7.597066908768606e-05, + "loss": 0.7039, + "step": 34006 + }, + { + "epoch": 2.359794601346194, + "grad_norm": 1.4201184511184692, + "learning_rate": 7.596521513374217e-05, + "loss": 0.4357, + "step": 34007 + }, + { + "epoch": 2.3598639927832905, + "grad_norm": 1.3113535642623901, + "learning_rate": 7.59597612556733e-05, + "loss": 0.4133, + "step": 34008 + }, + { + "epoch": 2.3599333842203873, + "grad_norm": 1.3424078226089478, + "learning_rate": 7.595430745349667e-05, + "loss": 0.4791, + "step": 34009 + }, + { + "epoch": 2.3600027756574837, + "grad_norm": 1.5797423124313354, + "learning_rate": 7.594885372722956e-05, + "loss": 0.4443, + "step": 34010 + }, + { + "epoch": 2.3600721670945806, + "grad_norm": 1.4819095134735107, + "learning_rate": 7.594340007688907e-05, + "loss": 0.4539, + "step": 34011 + }, + { + "epoch": 2.3601415585316774, + "grad_norm": 1.2168524265289307, + "learning_rate": 7.593794650249249e-05, + "loss": 0.4281, + "step": 34012 + }, + { + "epoch": 2.360210949968774, + "grad_norm": 1.3514875173568726, + "learning_rate": 7.5932493004057e-05, + "loss": 0.5257, + "step": 34013 + }, + { + "epoch": 2.3602803414058706, + "grad_norm": 1.2801412343978882, + "learning_rate": 7.592703958159986e-05, + "loss": 0.4771, + "step": 34014 + }, + { + "epoch": 2.360349732842967, + "grad_norm": 0.8879753351211548, + "learning_rate": 7.592158623513825e-05, + "loss": 0.3624, + "step": 34015 + }, + { + "epoch": 2.360419124280064, + "grad_norm": 1.5539264678955078, + "learning_rate": 7.591613296468944e-05, + "loss": 0.6601, + "step": 34016 + }, + { + "epoch": 2.3604885157171607, + "grad_norm": 0.8706486821174622, + "learning_rate": 7.591067977027056e-05, + "loss": 0.3432, + "step": 34017 + }, + { + "epoch": 2.360557907154257, + "grad_norm": 1.2219651937484741, + "learning_rate": 7.590522665189888e-05, + "loss": 0.5061, + "step": 34018 + }, + { + "epoch": 2.360627298591354, + "grad_norm": 1.435164213180542, + "learning_rate": 7.58997736095916e-05, + "loss": 0.4737, + "step": 34019 + }, + { + "epoch": 2.3606966900284503, + "grad_norm": 1.3069510459899902, + "learning_rate": 7.589432064336592e-05, + "loss": 0.3819, + "step": 34020 + }, + { + "epoch": 2.360766081465547, + "grad_norm": 1.2845839262008667, + "learning_rate": 7.588886775323913e-05, + "loss": 0.4138, + "step": 34021 + }, + { + "epoch": 2.360835472902644, + "grad_norm": 0.9771966338157654, + "learning_rate": 7.588341493922834e-05, + "loss": 0.4619, + "step": 34022 + }, + { + "epoch": 2.3609048643397403, + "grad_norm": 1.2209287881851196, + "learning_rate": 7.587796220135084e-05, + "loss": 0.4875, + "step": 34023 + }, + { + "epoch": 2.360974255776837, + "grad_norm": 1.0667141675949097, + "learning_rate": 7.58725095396238e-05, + "loss": 0.4193, + "step": 34024 + }, + { + "epoch": 2.3610436472139336, + "grad_norm": 1.3183314800262451, + "learning_rate": 7.586705695406444e-05, + "loss": 0.4033, + "step": 34025 + }, + { + "epoch": 2.3611130386510304, + "grad_norm": 1.2212868928909302, + "learning_rate": 7.586160444468999e-05, + "loss": 0.4326, + "step": 34026 + }, + { + "epoch": 2.3611824300881272, + "grad_norm": 1.2592713832855225, + "learning_rate": 7.585615201151764e-05, + "loss": 0.3971, + "step": 34027 + }, + { + "epoch": 2.3612518215252236, + "grad_norm": 1.2890313863754272, + "learning_rate": 7.585069965456463e-05, + "loss": 0.4705, + "step": 34028 + }, + { + "epoch": 2.3613212129623204, + "grad_norm": 1.2653026580810547, + "learning_rate": 7.584524737384815e-05, + "loss": 0.3646, + "step": 34029 + }, + { + "epoch": 2.3613906043994173, + "grad_norm": 1.578748345375061, + "learning_rate": 7.583979516938546e-05, + "loss": 0.733, + "step": 34030 + }, + { + "epoch": 2.3614599958365137, + "grad_norm": 1.2011194229125977, + "learning_rate": 7.58343430411937e-05, + "loss": 0.3507, + "step": 34031 + }, + { + "epoch": 2.3615293872736105, + "grad_norm": 1.2848917245864868, + "learning_rate": 7.582889098929011e-05, + "loss": 0.4119, + "step": 34032 + }, + { + "epoch": 2.3615987787107073, + "grad_norm": 1.2239832878112793, + "learning_rate": 7.58234390136919e-05, + "loss": 0.4243, + "step": 34033 + }, + { + "epoch": 2.3616681701478037, + "grad_norm": 1.2539358139038086, + "learning_rate": 7.58179871144163e-05, + "loss": 0.4135, + "step": 34034 + }, + { + "epoch": 2.3617375615849006, + "grad_norm": 0.9171602725982666, + "learning_rate": 7.581253529148052e-05, + "loss": 0.4486, + "step": 34035 + }, + { + "epoch": 2.361806953021997, + "grad_norm": 1.3452259302139282, + "learning_rate": 7.580708354490175e-05, + "loss": 0.4778, + "step": 34036 + }, + { + "epoch": 2.361876344459094, + "grad_norm": 1.3746333122253418, + "learning_rate": 7.580163187469723e-05, + "loss": 0.4312, + "step": 34037 + }, + { + "epoch": 2.3619457358961906, + "grad_norm": 1.2544682025909424, + "learning_rate": 7.579618028088414e-05, + "loss": 0.3372, + "step": 34038 + }, + { + "epoch": 2.362015127333287, + "grad_norm": 1.6619099378585815, + "learning_rate": 7.579072876347968e-05, + "loss": 0.3926, + "step": 34039 + }, + { + "epoch": 2.362084518770384, + "grad_norm": 1.6988191604614258, + "learning_rate": 7.578527732250111e-05, + "loss": 0.5169, + "step": 34040 + }, + { + "epoch": 2.3621539102074802, + "grad_norm": 1.9941712617874146, + "learning_rate": 7.577982595796558e-05, + "loss": 0.5697, + "step": 34041 + }, + { + "epoch": 2.362223301644577, + "grad_norm": 1.2187705039978027, + "learning_rate": 7.577437466989035e-05, + "loss": 0.2471, + "step": 34042 + }, + { + "epoch": 2.362292693081674, + "grad_norm": 1.5767323970794678, + "learning_rate": 7.576892345829258e-05, + "loss": 0.493, + "step": 34043 + }, + { + "epoch": 2.3623620845187703, + "grad_norm": 0.7908327579498291, + "learning_rate": 7.576347232318957e-05, + "loss": 0.3202, + "step": 34044 + }, + { + "epoch": 2.362431475955867, + "grad_norm": 0.7129895091056824, + "learning_rate": 7.575802126459843e-05, + "loss": 0.3475, + "step": 34045 + }, + { + "epoch": 2.3625008673929635, + "grad_norm": 1.076778769493103, + "learning_rate": 7.575257028253638e-05, + "loss": 0.5387, + "step": 34046 + }, + { + "epoch": 2.3625702588300603, + "grad_norm": 0.9894335865974426, + "learning_rate": 7.574711937702067e-05, + "loss": 0.3404, + "step": 34047 + }, + { + "epoch": 2.362639650267157, + "grad_norm": 1.514588475227356, + "learning_rate": 7.574166854806847e-05, + "loss": 0.6081, + "step": 34048 + }, + { + "epoch": 2.3627090417042536, + "grad_norm": 1.2242324352264404, + "learning_rate": 7.573621779569704e-05, + "loss": 0.4902, + "step": 34049 + }, + { + "epoch": 2.3627784331413504, + "grad_norm": 1.2830051183700562, + "learning_rate": 7.573076711992354e-05, + "loss": 0.5718, + "step": 34050 + }, + { + "epoch": 2.3628478245784468, + "grad_norm": 1.2384164333343506, + "learning_rate": 7.572531652076523e-05, + "loss": 0.4038, + "step": 34051 + }, + { + "epoch": 2.3629172160155436, + "grad_norm": 1.4295724630355835, + "learning_rate": 7.571986599823924e-05, + "loss": 0.4051, + "step": 34052 + }, + { + "epoch": 2.3629866074526404, + "grad_norm": 1.3801054954528809, + "learning_rate": 7.571441555236281e-05, + "loss": 0.3197, + "step": 34053 + }, + { + "epoch": 2.363055998889737, + "grad_norm": 1.4742107391357422, + "learning_rate": 7.570896518315317e-05, + "loss": 0.4472, + "step": 34054 + }, + { + "epoch": 2.3631253903268337, + "grad_norm": 1.2571163177490234, + "learning_rate": 7.57035148906275e-05, + "loss": 0.4712, + "step": 34055 + }, + { + "epoch": 2.3631947817639305, + "grad_norm": 1.4282468557357788, + "learning_rate": 7.569806467480302e-05, + "loss": 0.5683, + "step": 34056 + }, + { + "epoch": 2.363264173201027, + "grad_norm": 1.1735130548477173, + "learning_rate": 7.569261453569694e-05, + "loss": 0.4354, + "step": 34057 + }, + { + "epoch": 2.3633335646381237, + "grad_norm": 0.9893437623977661, + "learning_rate": 7.568716447332646e-05, + "loss": 0.3007, + "step": 34058 + }, + { + "epoch": 2.3634029560752206, + "grad_norm": 1.3842543363571167, + "learning_rate": 7.568171448770878e-05, + "loss": 0.3478, + "step": 34059 + }, + { + "epoch": 2.363472347512317, + "grad_norm": 1.3646327257156372, + "learning_rate": 7.567626457886108e-05, + "loss": 0.4396, + "step": 34060 + }, + { + "epoch": 2.3635417389494138, + "grad_norm": 1.2571929693222046, + "learning_rate": 7.56708147468006e-05, + "loss": 0.5103, + "step": 34061 + }, + { + "epoch": 2.36361113038651, + "grad_norm": 0.9758043885231018, + "learning_rate": 7.566536499154453e-05, + "loss": 0.383, + "step": 34062 + }, + { + "epoch": 2.363680521823607, + "grad_norm": 1.3273273706436157, + "learning_rate": 7.565991531311007e-05, + "loss": 0.3612, + "step": 34063 + }, + { + "epoch": 2.363749913260704, + "grad_norm": 1.3798480033874512, + "learning_rate": 7.565446571151444e-05, + "loss": 0.4232, + "step": 34064 + }, + { + "epoch": 2.3638193046978, + "grad_norm": 0.8549271821975708, + "learning_rate": 7.56490161867749e-05, + "loss": 0.2378, + "step": 34065 + }, + { + "epoch": 2.363888696134897, + "grad_norm": 1.153352975845337, + "learning_rate": 7.564356673890852e-05, + "loss": 0.2616, + "step": 34066 + }, + { + "epoch": 2.3639580875719934, + "grad_norm": 1.2373785972595215, + "learning_rate": 7.563811736793259e-05, + "loss": 0.3847, + "step": 34067 + }, + { + "epoch": 2.3640274790090903, + "grad_norm": 1.1159878969192505, + "learning_rate": 7.563266807386429e-05, + "loss": 0.3283, + "step": 34068 + }, + { + "epoch": 2.364096870446187, + "grad_norm": 1.629818320274353, + "learning_rate": 7.562721885672085e-05, + "loss": 0.4473, + "step": 34069 + }, + { + "epoch": 2.3641662618832835, + "grad_norm": 1.2608444690704346, + "learning_rate": 7.562176971651942e-05, + "loss": 0.2991, + "step": 34070 + }, + { + "epoch": 2.3642356533203803, + "grad_norm": 1.8789880275726318, + "learning_rate": 7.561632065327728e-05, + "loss": 0.5594, + "step": 34071 + }, + { + "epoch": 2.3643050447574767, + "grad_norm": 2.374243974685669, + "learning_rate": 7.561087166701158e-05, + "loss": 0.4946, + "step": 34072 + }, + { + "epoch": 2.3643744361945735, + "grad_norm": 1.2913938760757446, + "learning_rate": 7.560542275773949e-05, + "loss": 0.3849, + "step": 34073 + }, + { + "epoch": 2.3644438276316704, + "grad_norm": 2.0699801445007324, + "learning_rate": 7.559997392547826e-05, + "loss": 0.4171, + "step": 34074 + }, + { + "epoch": 2.3645132190687668, + "grad_norm": 1.5105669498443604, + "learning_rate": 7.559452517024511e-05, + "loss": 0.5276, + "step": 34075 + }, + { + "epoch": 2.3645826105058636, + "grad_norm": 1.2445346117019653, + "learning_rate": 7.558907649205719e-05, + "loss": 0.3743, + "step": 34076 + }, + { + "epoch": 2.36465200194296, + "grad_norm": 1.6798019409179688, + "learning_rate": 7.558362789093172e-05, + "loss": 0.4911, + "step": 34077 + }, + { + "epoch": 2.364721393380057, + "grad_norm": 1.3814042806625366, + "learning_rate": 7.557817936688591e-05, + "loss": 0.4734, + "step": 34078 + }, + { + "epoch": 2.3647907848171537, + "grad_norm": 1.30115807056427, + "learning_rate": 7.557273091993699e-05, + "loss": 0.3903, + "step": 34079 + }, + { + "epoch": 2.36486017625425, + "grad_norm": 1.667141318321228, + "learning_rate": 7.556728255010208e-05, + "loss": 0.6784, + "step": 34080 + }, + { + "epoch": 2.364929567691347, + "grad_norm": 0.7584269046783447, + "learning_rate": 7.556183425739844e-05, + "loss": 0.2926, + "step": 34081 + }, + { + "epoch": 2.3649989591284437, + "grad_norm": 1.0482127666473389, + "learning_rate": 7.555638604184323e-05, + "loss": 0.423, + "step": 34082 + }, + { + "epoch": 2.36506835056554, + "grad_norm": 1.4497500658035278, + "learning_rate": 7.555093790345369e-05, + "loss": 0.4975, + "step": 34083 + }, + { + "epoch": 2.365137742002637, + "grad_norm": 1.273262858390808, + "learning_rate": 7.5545489842247e-05, + "loss": 0.4722, + "step": 34084 + }, + { + "epoch": 2.3652071334397338, + "grad_norm": 1.591409683227539, + "learning_rate": 7.554004185824036e-05, + "loss": 0.6224, + "step": 34085 + }, + { + "epoch": 2.36527652487683, + "grad_norm": 1.230686902999878, + "learning_rate": 7.553459395145101e-05, + "loss": 0.5034, + "step": 34086 + }, + { + "epoch": 2.365345916313927, + "grad_norm": 1.3437999486923218, + "learning_rate": 7.552914612189606e-05, + "loss": 0.2969, + "step": 34087 + }, + { + "epoch": 2.3654153077510234, + "grad_norm": 1.0647270679473877, + "learning_rate": 7.552369836959278e-05, + "loss": 0.4205, + "step": 34088 + }, + { + "epoch": 2.36548469918812, + "grad_norm": 1.3250914812088013, + "learning_rate": 7.551825069455832e-05, + "loss": 0.5473, + "step": 34089 + }, + { + "epoch": 2.365554090625217, + "grad_norm": 1.7187960147857666, + "learning_rate": 7.551280309680994e-05, + "loss": 0.5278, + "step": 34090 + }, + { + "epoch": 2.3656234820623134, + "grad_norm": 1.2659235000610352, + "learning_rate": 7.550735557636477e-05, + "loss": 0.3621, + "step": 34091 + }, + { + "epoch": 2.3656928734994103, + "grad_norm": 1.5210165977478027, + "learning_rate": 7.550190813324006e-05, + "loss": 0.4423, + "step": 34092 + }, + { + "epoch": 2.3657622649365067, + "grad_norm": 1.2568844556808472, + "learning_rate": 7.549646076745296e-05, + "loss": 0.4036, + "step": 34093 + }, + { + "epoch": 2.3658316563736035, + "grad_norm": 1.0390641689300537, + "learning_rate": 7.549101347902071e-05, + "loss": 0.3871, + "step": 34094 + }, + { + "epoch": 2.3659010478107003, + "grad_norm": 1.4013439416885376, + "learning_rate": 7.548556626796046e-05, + "loss": 0.468, + "step": 34095 + }, + { + "epoch": 2.3659704392477967, + "grad_norm": 1.1257004737854004, + "learning_rate": 7.548011913428944e-05, + "loss": 0.3842, + "step": 34096 + }, + { + "epoch": 2.3660398306848935, + "grad_norm": 1.3907028436660767, + "learning_rate": 7.547467207802484e-05, + "loss": 0.7159, + "step": 34097 + }, + { + "epoch": 2.36610922212199, + "grad_norm": 1.5595438480377197, + "learning_rate": 7.546922509918386e-05, + "loss": 0.4343, + "step": 34098 + }, + { + "epoch": 2.3661786135590868, + "grad_norm": 0.9960337281227112, + "learning_rate": 7.546377819778373e-05, + "loss": 0.2686, + "step": 34099 + }, + { + "epoch": 2.3662480049961836, + "grad_norm": 1.6737675666809082, + "learning_rate": 7.545833137384156e-05, + "loss": 0.4827, + "step": 34100 + }, + { + "epoch": 2.36631739643328, + "grad_norm": 1.0367997884750366, + "learning_rate": 7.545288462737459e-05, + "loss": 0.3036, + "step": 34101 + }, + { + "epoch": 2.366386787870377, + "grad_norm": 1.4195940494537354, + "learning_rate": 7.544743795840001e-05, + "loss": 0.4017, + "step": 34102 + }, + { + "epoch": 2.3664561793074737, + "grad_norm": 1.1864466667175293, + "learning_rate": 7.544199136693503e-05, + "loss": 0.3632, + "step": 34103 + }, + { + "epoch": 2.36652557074457, + "grad_norm": 1.4279605150222778, + "learning_rate": 7.543654485299684e-05, + "loss": 0.4066, + "step": 34104 + }, + { + "epoch": 2.366594962181667, + "grad_norm": 1.090114951133728, + "learning_rate": 7.543109841660262e-05, + "loss": 0.3747, + "step": 34105 + }, + { + "epoch": 2.3666643536187633, + "grad_norm": 1.1363500356674194, + "learning_rate": 7.542565205776961e-05, + "loss": 0.3983, + "step": 34106 + }, + { + "epoch": 2.36673374505586, + "grad_norm": 3.050198554992676, + "learning_rate": 7.542020577651492e-05, + "loss": 0.4777, + "step": 34107 + }, + { + "epoch": 2.366803136492957, + "grad_norm": 1.4139137268066406, + "learning_rate": 7.541475957285579e-05, + "loss": 0.5547, + "step": 34108 + }, + { + "epoch": 2.3668725279300533, + "grad_norm": 1.130759835243225, + "learning_rate": 7.540931344680942e-05, + "loss": 0.4611, + "step": 34109 + }, + { + "epoch": 2.36694191936715, + "grad_norm": 1.3139299154281616, + "learning_rate": 7.540386739839302e-05, + "loss": 0.4327, + "step": 34110 + }, + { + "epoch": 2.367011310804247, + "grad_norm": 1.169636845588684, + "learning_rate": 7.539842142762373e-05, + "loss": 0.5015, + "step": 34111 + }, + { + "epoch": 2.3670807022413434, + "grad_norm": 1.3857799768447876, + "learning_rate": 7.539297553451877e-05, + "loss": 0.3438, + "step": 34112 + }, + { + "epoch": 2.36715009367844, + "grad_norm": 1.2006667852401733, + "learning_rate": 7.538752971909536e-05, + "loss": 0.39, + "step": 34113 + }, + { + "epoch": 2.3672194851155366, + "grad_norm": 1.26025390625, + "learning_rate": 7.538208398137064e-05, + "loss": 0.5308, + "step": 34114 + }, + { + "epoch": 2.3672888765526334, + "grad_norm": 1.1834050416946411, + "learning_rate": 7.537663832136181e-05, + "loss": 0.4149, + "step": 34115 + }, + { + "epoch": 2.3673582679897303, + "grad_norm": 1.4605305194854736, + "learning_rate": 7.537119273908608e-05, + "loss": 0.3286, + "step": 34116 + }, + { + "epoch": 2.3674276594268266, + "grad_norm": 1.5223652124404907, + "learning_rate": 7.536574723456064e-05, + "loss": 0.35, + "step": 34117 + }, + { + "epoch": 2.3674970508639235, + "grad_norm": 1.237284541130066, + "learning_rate": 7.536030180780266e-05, + "loss": 0.3682, + "step": 34118 + }, + { + "epoch": 2.36756644230102, + "grad_norm": 0.9842182993888855, + "learning_rate": 7.535485645882938e-05, + "loss": 0.2696, + "step": 34119 + }, + { + "epoch": 2.3676358337381167, + "grad_norm": 0.927645742893219, + "learning_rate": 7.534941118765798e-05, + "loss": 0.2912, + "step": 34120 + }, + { + "epoch": 2.3677052251752135, + "grad_norm": 3.5557138919830322, + "learning_rate": 7.53439659943056e-05, + "loss": 0.4226, + "step": 34121 + }, + { + "epoch": 2.36777461661231, + "grad_norm": 1.296197772026062, + "learning_rate": 7.533852087878944e-05, + "loss": 0.4583, + "step": 34122 + }, + { + "epoch": 2.3678440080494068, + "grad_norm": 1.1614936590194702, + "learning_rate": 7.533307584112672e-05, + "loss": 0.2723, + "step": 34123 + }, + { + "epoch": 2.367913399486503, + "grad_norm": 1.375475525856018, + "learning_rate": 7.532763088133462e-05, + "loss": 0.3867, + "step": 34124 + }, + { + "epoch": 2.3679827909236, + "grad_norm": 1.2237648963928223, + "learning_rate": 7.532218599943034e-05, + "loss": 0.4255, + "step": 34125 + }, + { + "epoch": 2.368052182360697, + "grad_norm": 1.5889824628829956, + "learning_rate": 7.531674119543105e-05, + "loss": 0.435, + "step": 34126 + }, + { + "epoch": 2.368121573797793, + "grad_norm": 1.4452887773513794, + "learning_rate": 7.531129646935395e-05, + "loss": 0.4935, + "step": 34127 + }, + { + "epoch": 2.36819096523489, + "grad_norm": 0.8810141086578369, + "learning_rate": 7.530585182121621e-05, + "loss": 0.3084, + "step": 34128 + }, + { + "epoch": 2.368260356671987, + "grad_norm": 1.6673067808151245, + "learning_rate": 7.530040725103505e-05, + "loss": 0.5144, + "step": 34129 + }, + { + "epoch": 2.3683297481090833, + "grad_norm": 1.4180041551589966, + "learning_rate": 7.529496275882762e-05, + "loss": 0.5715, + "step": 34130 + }, + { + "epoch": 2.36839913954618, + "grad_norm": 1.77109956741333, + "learning_rate": 7.52895183446111e-05, + "loss": 0.4258, + "step": 34131 + }, + { + "epoch": 2.3684685309832765, + "grad_norm": 1.5795029401779175, + "learning_rate": 7.528407400840274e-05, + "loss": 0.5027, + "step": 34132 + }, + { + "epoch": 2.3685379224203733, + "grad_norm": 1.5877681970596313, + "learning_rate": 7.527862975021967e-05, + "loss": 0.5765, + "step": 34133 + }, + { + "epoch": 2.36860731385747, + "grad_norm": 1.223031759262085, + "learning_rate": 7.527318557007916e-05, + "loss": 0.4613, + "step": 34134 + }, + { + "epoch": 2.3686767052945665, + "grad_norm": 1.4519339799880981, + "learning_rate": 7.526774146799827e-05, + "loss": 0.3846, + "step": 34135 + }, + { + "epoch": 2.3687460967316634, + "grad_norm": 1.5464924573898315, + "learning_rate": 7.526229744399427e-05, + "loss": 0.4572, + "step": 34136 + }, + { + "epoch": 2.36881548816876, + "grad_norm": 1.2647968530654907, + "learning_rate": 7.525685349808431e-05, + "loss": 0.3654, + "step": 34137 + }, + { + "epoch": 2.3688848796058566, + "grad_norm": 1.4084841012954712, + "learning_rate": 7.52514096302856e-05, + "loss": 0.4229, + "step": 34138 + }, + { + "epoch": 2.3689542710429534, + "grad_norm": 1.3904646635055542, + "learning_rate": 7.524596584061532e-05, + "loss": 0.4749, + "step": 34139 + }, + { + "epoch": 2.36902366248005, + "grad_norm": 0.9357373118400574, + "learning_rate": 7.524052212909065e-05, + "loss": 0.3016, + "step": 34140 + }, + { + "epoch": 2.3690930539171466, + "grad_norm": 1.511628270149231, + "learning_rate": 7.523507849572882e-05, + "loss": 0.4508, + "step": 34141 + }, + { + "epoch": 2.3691624453542435, + "grad_norm": 1.308541178703308, + "learning_rate": 7.522963494054694e-05, + "loss": 0.4287, + "step": 34142 + }, + { + "epoch": 2.36923183679134, + "grad_norm": 1.4639472961425781, + "learning_rate": 7.52241914635622e-05, + "loss": 0.3718, + "step": 34143 + }, + { + "epoch": 2.3693012282284367, + "grad_norm": 1.0824686288833618, + "learning_rate": 7.521874806479186e-05, + "loss": 0.3361, + "step": 34144 + }, + { + "epoch": 2.369370619665533, + "grad_norm": 1.3089287281036377, + "learning_rate": 7.521330474425304e-05, + "loss": 0.3504, + "step": 34145 + }, + { + "epoch": 2.36944001110263, + "grad_norm": 1.7360162734985352, + "learning_rate": 7.520786150196293e-05, + "loss": 0.425, + "step": 34146 + }, + { + "epoch": 2.3695094025397268, + "grad_norm": 1.7716925144195557, + "learning_rate": 7.520241833793874e-05, + "loss": 0.4883, + "step": 34147 + }, + { + "epoch": 2.369578793976823, + "grad_norm": 1.6902742385864258, + "learning_rate": 7.519697525219766e-05, + "loss": 0.6048, + "step": 34148 + }, + { + "epoch": 2.36964818541392, + "grad_norm": 1.0598615407943726, + "learning_rate": 7.519153224475682e-05, + "loss": 0.3791, + "step": 34149 + }, + { + "epoch": 2.3697175768510164, + "grad_norm": 1.5572704076766968, + "learning_rate": 7.518608931563344e-05, + "loss": 0.4438, + "step": 34150 + }, + { + "epoch": 2.369786968288113, + "grad_norm": 1.3499807119369507, + "learning_rate": 7.518064646484469e-05, + "loss": 0.3727, + "step": 34151 + }, + { + "epoch": 2.36985635972521, + "grad_norm": 1.2488049268722534, + "learning_rate": 7.517520369240776e-05, + "loss": 0.4463, + "step": 34152 + }, + { + "epoch": 2.3699257511623064, + "grad_norm": 1.044103980064392, + "learning_rate": 7.516976099833984e-05, + "loss": 0.3383, + "step": 34153 + }, + { + "epoch": 2.3699951425994032, + "grad_norm": 0.9712185263633728, + "learning_rate": 7.516431838265809e-05, + "loss": 0.4135, + "step": 34154 + }, + { + "epoch": 2.3700645340365, + "grad_norm": 1.3066586256027222, + "learning_rate": 7.515887584537976e-05, + "loss": 0.5066, + "step": 34155 + }, + { + "epoch": 2.3701339254735965, + "grad_norm": 0.8416784405708313, + "learning_rate": 7.515343338652192e-05, + "loss": 0.4131, + "step": 34156 + }, + { + "epoch": 2.3702033169106933, + "grad_norm": 1.332425832748413, + "learning_rate": 7.514799100610181e-05, + "loss": 0.4091, + "step": 34157 + }, + { + "epoch": 2.3702727083477897, + "grad_norm": 1.1976865530014038, + "learning_rate": 7.514254870413662e-05, + "loss": 0.38, + "step": 34158 + }, + { + "epoch": 2.3703420997848865, + "grad_norm": 1.6818915605545044, + "learning_rate": 7.513710648064351e-05, + "loss": 0.4019, + "step": 34159 + }, + { + "epoch": 2.3704114912219834, + "grad_norm": 0.959242045879364, + "learning_rate": 7.513166433563969e-05, + "loss": 0.3645, + "step": 34160 + }, + { + "epoch": 2.3704808826590797, + "grad_norm": 1.5686054229736328, + "learning_rate": 7.512622226914231e-05, + "loss": 0.4169, + "step": 34161 + }, + { + "epoch": 2.3705502740961766, + "grad_norm": 1.0665628910064697, + "learning_rate": 7.512078028116857e-05, + "loss": 0.4097, + "step": 34162 + }, + { + "epoch": 2.3706196655332734, + "grad_norm": 1.4079183340072632, + "learning_rate": 7.511533837173563e-05, + "loss": 0.4877, + "step": 34163 + }, + { + "epoch": 2.37068905697037, + "grad_norm": 1.1701562404632568, + "learning_rate": 7.510989654086069e-05, + "loss": 0.451, + "step": 34164 + }, + { + "epoch": 2.3707584484074666, + "grad_norm": 1.3011349439620972, + "learning_rate": 7.51044547885609e-05, + "loss": 0.4855, + "step": 34165 + }, + { + "epoch": 2.370827839844563, + "grad_norm": 1.3235615491867065, + "learning_rate": 7.509901311485348e-05, + "loss": 0.3857, + "step": 34166 + }, + { + "epoch": 2.37089723128166, + "grad_norm": 1.0457417964935303, + "learning_rate": 7.509357151975556e-05, + "loss": 0.3045, + "step": 34167 + }, + { + "epoch": 2.3709666227187567, + "grad_norm": 1.3210272789001465, + "learning_rate": 7.508813000328439e-05, + "loss": 0.5171, + "step": 34168 + }, + { + "epoch": 2.371036014155853, + "grad_norm": 1.064516544342041, + "learning_rate": 7.508268856545708e-05, + "loss": 0.3085, + "step": 34169 + }, + { + "epoch": 2.37110540559295, + "grad_norm": 1.6644494533538818, + "learning_rate": 7.507724720629081e-05, + "loss": 0.404, + "step": 34170 + }, + { + "epoch": 2.3711747970300463, + "grad_norm": 1.1509510278701782, + "learning_rate": 7.507180592580278e-05, + "loss": 0.4335, + "step": 34171 + }, + { + "epoch": 2.371244188467143, + "grad_norm": 1.6773896217346191, + "learning_rate": 7.506636472401019e-05, + "loss": 0.4762, + "step": 34172 + }, + { + "epoch": 2.37131357990424, + "grad_norm": 1.2022064924240112, + "learning_rate": 7.506092360093017e-05, + "loss": 0.4647, + "step": 34173 + }, + { + "epoch": 2.3713829713413364, + "grad_norm": 1.3892440795898438, + "learning_rate": 7.505548255657992e-05, + "loss": 0.599, + "step": 34174 + }, + { + "epoch": 2.371452362778433, + "grad_norm": 1.5254384279251099, + "learning_rate": 7.505004159097667e-05, + "loss": 0.4718, + "step": 34175 + }, + { + "epoch": 2.3715217542155296, + "grad_norm": 1.3982019424438477, + "learning_rate": 7.50446007041375e-05, + "loss": 0.4894, + "step": 34176 + }, + { + "epoch": 2.3715911456526264, + "grad_norm": 1.0699357986450195, + "learning_rate": 7.503915989607964e-05, + "loss": 0.5053, + "step": 34177 + }, + { + "epoch": 2.3716605370897232, + "grad_norm": 1.313104271888733, + "learning_rate": 7.503371916682024e-05, + "loss": 0.6596, + "step": 34178 + }, + { + "epoch": 2.3717299285268196, + "grad_norm": 1.0873953104019165, + "learning_rate": 7.502827851637652e-05, + "loss": 0.4449, + "step": 34179 + }, + { + "epoch": 2.3717993199639165, + "grad_norm": 0.9668839573860168, + "learning_rate": 7.50228379447656e-05, + "loss": 0.3048, + "step": 34180 + }, + { + "epoch": 2.3718687114010133, + "grad_norm": 1.2587887048721313, + "learning_rate": 7.501739745200469e-05, + "loss": 0.4221, + "step": 34181 + }, + { + "epoch": 2.3719381028381097, + "grad_norm": 1.2052441835403442, + "learning_rate": 7.501195703811097e-05, + "loss": 0.3637, + "step": 34182 + }, + { + "epoch": 2.3720074942752065, + "grad_norm": 1.2154511213302612, + "learning_rate": 7.500651670310162e-05, + "loss": 0.2904, + "step": 34183 + }, + { + "epoch": 2.3720768857123034, + "grad_norm": 1.0085172653198242, + "learning_rate": 7.500107644699375e-05, + "loss": 0.4326, + "step": 34184 + }, + { + "epoch": 2.3721462771493997, + "grad_norm": 1.324520468711853, + "learning_rate": 7.499563626980461e-05, + "loss": 0.5365, + "step": 34185 + }, + { + "epoch": 2.3722156685864966, + "grad_norm": 1.2499232292175293, + "learning_rate": 7.499019617155132e-05, + "loss": 0.3237, + "step": 34186 + }, + { + "epoch": 2.372285060023593, + "grad_norm": 1.0743567943572998, + "learning_rate": 7.49847561522511e-05, + "loss": 0.2328, + "step": 34187 + }, + { + "epoch": 2.37235445146069, + "grad_norm": 1.0446993112564087, + "learning_rate": 7.49793162119211e-05, + "loss": 0.3919, + "step": 34188 + }, + { + "epoch": 2.3724238428977866, + "grad_norm": 1.6067724227905273, + "learning_rate": 7.497387635057852e-05, + "loss": 0.3999, + "step": 34189 + }, + { + "epoch": 2.372493234334883, + "grad_norm": 1.5871111154556274, + "learning_rate": 7.496843656824048e-05, + "loss": 0.5541, + "step": 34190 + }, + { + "epoch": 2.37256262577198, + "grad_norm": 1.0353842973709106, + "learning_rate": 7.496299686492418e-05, + "loss": 0.4012, + "step": 34191 + }, + { + "epoch": 2.3726320172090762, + "grad_norm": 0.9951072931289673, + "learning_rate": 7.49575572406468e-05, + "loss": 0.4333, + "step": 34192 + }, + { + "epoch": 2.372701408646173, + "grad_norm": 1.6612359285354614, + "learning_rate": 7.495211769542548e-05, + "loss": 0.5283, + "step": 34193 + }, + { + "epoch": 2.37277080008327, + "grad_norm": 1.0198509693145752, + "learning_rate": 7.494667822927745e-05, + "loss": 0.3913, + "step": 34194 + }, + { + "epoch": 2.3728401915203663, + "grad_norm": 1.5227634906768799, + "learning_rate": 7.494123884221985e-05, + "loss": 0.4275, + "step": 34195 + }, + { + "epoch": 2.372909582957463, + "grad_norm": 0.9950312376022339, + "learning_rate": 7.493579953426986e-05, + "loss": 0.472, + "step": 34196 + }, + { + "epoch": 2.3729789743945595, + "grad_norm": 1.5070030689239502, + "learning_rate": 7.493036030544462e-05, + "loss": 0.3795, + "step": 34197 + }, + { + "epoch": 2.3730483658316563, + "grad_norm": 0.8242715001106262, + "learning_rate": 7.492492115576133e-05, + "loss": 0.2748, + "step": 34198 + }, + { + "epoch": 2.373117757268753, + "grad_norm": 1.1573258638381958, + "learning_rate": 7.491948208523719e-05, + "loss": 0.4271, + "step": 34199 + }, + { + "epoch": 2.3731871487058496, + "grad_norm": 1.1965917348861694, + "learning_rate": 7.491404309388928e-05, + "loss": 0.5321, + "step": 34200 + }, + { + "epoch": 2.3732565401429464, + "grad_norm": 1.111266851425171, + "learning_rate": 7.490860418173484e-05, + "loss": 0.3151, + "step": 34201 + }, + { + "epoch": 2.373325931580043, + "grad_norm": 1.040172815322876, + "learning_rate": 7.490316534879103e-05, + "loss": 0.3979, + "step": 34202 + }, + { + "epoch": 2.3733953230171396, + "grad_norm": 1.1558727025985718, + "learning_rate": 7.489772659507505e-05, + "loss": 0.4728, + "step": 34203 + }, + { + "epoch": 2.3734647144542365, + "grad_norm": 0.8804230093955994, + "learning_rate": 7.4892287920604e-05, + "loss": 0.3845, + "step": 34204 + }, + { + "epoch": 2.373534105891333, + "grad_norm": 1.4312947988510132, + "learning_rate": 7.488684932539508e-05, + "loss": 0.4312, + "step": 34205 + }, + { + "epoch": 2.3736034973284297, + "grad_norm": 1.0333166122436523, + "learning_rate": 7.488141080946547e-05, + "loss": 0.3759, + "step": 34206 + }, + { + "epoch": 2.3736728887655265, + "grad_norm": 1.423905372619629, + "learning_rate": 7.487597237283232e-05, + "loss": 0.4482, + "step": 34207 + }, + { + "epoch": 2.373742280202623, + "grad_norm": 1.3881137371063232, + "learning_rate": 7.487053401551282e-05, + "loss": 0.5318, + "step": 34208 + }, + { + "epoch": 2.3738116716397197, + "grad_norm": 1.7063785791397095, + "learning_rate": 7.486509573752412e-05, + "loss": 0.3559, + "step": 34209 + }, + { + "epoch": 2.3738810630768166, + "grad_norm": 1.0182844400405884, + "learning_rate": 7.485965753888345e-05, + "loss": 0.3251, + "step": 34210 + }, + { + "epoch": 2.373950454513913, + "grad_norm": 1.1429213285446167, + "learning_rate": 7.485421941960786e-05, + "loss": 0.4362, + "step": 34211 + }, + { + "epoch": 2.37401984595101, + "grad_norm": 1.3104593753814697, + "learning_rate": 7.48487813797146e-05, + "loss": 0.3912, + "step": 34212 + }, + { + "epoch": 2.374089237388106, + "grad_norm": 1.1281211376190186, + "learning_rate": 7.484334341922082e-05, + "loss": 0.2738, + "step": 34213 + }, + { + "epoch": 2.374158628825203, + "grad_norm": 1.29869544506073, + "learning_rate": 7.48379055381437e-05, + "loss": 0.3384, + "step": 34214 + }, + { + "epoch": 2.3742280202623, + "grad_norm": 1.3398507833480835, + "learning_rate": 7.483246773650039e-05, + "loss": 0.3073, + "step": 34215 + }, + { + "epoch": 2.3742974116993962, + "grad_norm": 1.5857073068618774, + "learning_rate": 7.482703001430804e-05, + "loss": 0.4756, + "step": 34216 + }, + { + "epoch": 2.374366803136493, + "grad_norm": 1.2712687253952026, + "learning_rate": 7.482159237158385e-05, + "loss": 0.4198, + "step": 34217 + }, + { + "epoch": 2.3744361945735895, + "grad_norm": 1.403458833694458, + "learning_rate": 7.481615480834499e-05, + "loss": 0.3956, + "step": 34218 + }, + { + "epoch": 2.3745055860106863, + "grad_norm": 1.3278383016586304, + "learning_rate": 7.481071732460858e-05, + "loss": 0.4229, + "step": 34219 + }, + { + "epoch": 2.374574977447783, + "grad_norm": 1.2782697677612305, + "learning_rate": 7.480527992039181e-05, + "loss": 0.2697, + "step": 34220 + }, + { + "epoch": 2.3746443688848795, + "grad_norm": 1.5506385564804077, + "learning_rate": 7.479984259571186e-05, + "loss": 0.426, + "step": 34221 + }, + { + "epoch": 2.3747137603219763, + "grad_norm": 1.2463847398757935, + "learning_rate": 7.479440535058586e-05, + "loss": 0.3768, + "step": 34222 + }, + { + "epoch": 2.3747831517590727, + "grad_norm": 1.6996058225631714, + "learning_rate": 7.478896818503101e-05, + "loss": 0.5058, + "step": 34223 + }, + { + "epoch": 2.3748525431961696, + "grad_norm": 1.5634137392044067, + "learning_rate": 7.478353109906451e-05, + "loss": 0.4655, + "step": 34224 + }, + { + "epoch": 2.3749219346332664, + "grad_norm": 1.0513073205947876, + "learning_rate": 7.477809409270344e-05, + "loss": 0.3209, + "step": 34225 + }, + { + "epoch": 2.374991326070363, + "grad_norm": 1.2431981563568115, + "learning_rate": 7.477265716596499e-05, + "loss": 0.4218, + "step": 34226 + }, + { + "epoch": 2.3750607175074596, + "grad_norm": 1.5966230630874634, + "learning_rate": 7.476722031886634e-05, + "loss": 0.3916, + "step": 34227 + }, + { + "epoch": 2.375130108944556, + "grad_norm": 1.1022096872329712, + "learning_rate": 7.476178355142464e-05, + "loss": 0.263, + "step": 34228 + }, + { + "epoch": 2.375199500381653, + "grad_norm": 1.363394856452942, + "learning_rate": 7.475634686365707e-05, + "loss": 0.5057, + "step": 34229 + }, + { + "epoch": 2.3752688918187497, + "grad_norm": 1.176951289176941, + "learning_rate": 7.475091025558079e-05, + "loss": 0.3604, + "step": 34230 + }, + { + "epoch": 2.375338283255846, + "grad_norm": 1.310423493385315, + "learning_rate": 7.474547372721296e-05, + "loss": 0.467, + "step": 34231 + }, + { + "epoch": 2.375407674692943, + "grad_norm": 1.4432486295700073, + "learning_rate": 7.474003727857073e-05, + "loss": 0.4371, + "step": 34232 + }, + { + "epoch": 2.3754770661300397, + "grad_norm": 0.921995997428894, + "learning_rate": 7.47346009096713e-05, + "loss": 0.3273, + "step": 34233 + }, + { + "epoch": 2.375546457567136, + "grad_norm": 1.3628987073898315, + "learning_rate": 7.472916462053176e-05, + "loss": 0.5142, + "step": 34234 + }, + { + "epoch": 2.375615849004233, + "grad_norm": 1.2372738122940063, + "learning_rate": 7.472372841116932e-05, + "loss": 0.5516, + "step": 34235 + }, + { + "epoch": 2.37568524044133, + "grad_norm": 1.3382030725479126, + "learning_rate": 7.471829228160114e-05, + "loss": 0.4358, + "step": 34236 + }, + { + "epoch": 2.375754631878426, + "grad_norm": 1.4881991147994995, + "learning_rate": 7.471285623184439e-05, + "loss": 0.3634, + "step": 34237 + }, + { + "epoch": 2.375824023315523, + "grad_norm": 1.3759825229644775, + "learning_rate": 7.470742026191627e-05, + "loss": 0.4847, + "step": 34238 + }, + { + "epoch": 2.3758934147526194, + "grad_norm": 1.0698652267456055, + "learning_rate": 7.470198437183382e-05, + "loss": 0.3866, + "step": 34239 + }, + { + "epoch": 2.3759628061897162, + "grad_norm": 1.20381498336792, + "learning_rate": 7.469654856161428e-05, + "loss": 0.3933, + "step": 34240 + }, + { + "epoch": 2.376032197626813, + "grad_norm": 0.9248119592666626, + "learning_rate": 7.46911128312748e-05, + "loss": 0.3169, + "step": 34241 + }, + { + "epoch": 2.3761015890639094, + "grad_norm": 1.611363172531128, + "learning_rate": 7.468567718083252e-05, + "loss": 0.3666, + "step": 34242 + }, + { + "epoch": 2.3761709805010063, + "grad_norm": 1.7428284883499146, + "learning_rate": 7.468024161030465e-05, + "loss": 0.5401, + "step": 34243 + }, + { + "epoch": 2.3762403719381027, + "grad_norm": 1.1407653093338013, + "learning_rate": 7.467480611970832e-05, + "loss": 0.3579, + "step": 34244 + }, + { + "epoch": 2.3763097633751995, + "grad_norm": 0.8340096473693848, + "learning_rate": 7.466937070906071e-05, + "loss": 0.2992, + "step": 34245 + }, + { + "epoch": 2.3763791548122963, + "grad_norm": 1.5386053323745728, + "learning_rate": 7.466393537837893e-05, + "loss": 0.5893, + "step": 34246 + }, + { + "epoch": 2.3764485462493927, + "grad_norm": 1.515188217163086, + "learning_rate": 7.465850012768016e-05, + "loss": 0.4982, + "step": 34247 + }, + { + "epoch": 2.3765179376864896, + "grad_norm": 1.3436332941055298, + "learning_rate": 7.465306495698157e-05, + "loss": 0.4188, + "step": 34248 + }, + { + "epoch": 2.376587329123586, + "grad_norm": 1.5231635570526123, + "learning_rate": 7.464762986630034e-05, + "loss": 0.4849, + "step": 34249 + }, + { + "epoch": 2.376656720560683, + "grad_norm": 1.6079127788543701, + "learning_rate": 7.464219485565357e-05, + "loss": 0.6054, + "step": 34250 + }, + { + "epoch": 2.3767261119977796, + "grad_norm": 1.5462371110916138, + "learning_rate": 7.463675992505848e-05, + "loss": 0.4737, + "step": 34251 + }, + { + "epoch": 2.376795503434876, + "grad_norm": 1.4517600536346436, + "learning_rate": 7.463132507453216e-05, + "loss": 0.4241, + "step": 34252 + }, + { + "epoch": 2.376864894871973, + "grad_norm": 1.6133482456207275, + "learning_rate": 7.462589030409184e-05, + "loss": 0.6534, + "step": 34253 + }, + { + "epoch": 2.376934286309069, + "grad_norm": 2.3162472248077393, + "learning_rate": 7.46204556137546e-05, + "loss": 0.4181, + "step": 34254 + }, + { + "epoch": 2.377003677746166, + "grad_norm": 1.3741546869277954, + "learning_rate": 7.461502100353766e-05, + "loss": 0.3625, + "step": 34255 + }, + { + "epoch": 2.377073069183263, + "grad_norm": 1.7333881855010986, + "learning_rate": 7.460958647345813e-05, + "loss": 0.4458, + "step": 34256 + }, + { + "epoch": 2.3771424606203593, + "grad_norm": 1.1137909889221191, + "learning_rate": 7.46041520235332e-05, + "loss": 0.3311, + "step": 34257 + }, + { + "epoch": 2.377211852057456, + "grad_norm": 1.1249619722366333, + "learning_rate": 7.459871765378006e-05, + "loss": 0.441, + "step": 34258 + }, + { + "epoch": 2.377281243494553, + "grad_norm": 1.1974143981933594, + "learning_rate": 7.459328336421578e-05, + "loss": 0.4239, + "step": 34259 + }, + { + "epoch": 2.3773506349316493, + "grad_norm": 1.5423908233642578, + "learning_rate": 7.458784915485756e-05, + "loss": 0.6532, + "step": 34260 + }, + { + "epoch": 2.377420026368746, + "grad_norm": 1.6051816940307617, + "learning_rate": 7.458241502572254e-05, + "loss": 0.3518, + "step": 34261 + }, + { + "epoch": 2.377489417805843, + "grad_norm": 1.2114959955215454, + "learning_rate": 7.45769809768279e-05, + "loss": 0.5983, + "step": 34262 + }, + { + "epoch": 2.3775588092429394, + "grad_norm": 1.304640293121338, + "learning_rate": 7.457154700819078e-05, + "loss": 0.289, + "step": 34263 + }, + { + "epoch": 2.377628200680036, + "grad_norm": 1.6282708644866943, + "learning_rate": 7.456611311982833e-05, + "loss": 0.7929, + "step": 34264 + }, + { + "epoch": 2.3776975921171326, + "grad_norm": 1.9056860208511353, + "learning_rate": 7.456067931175774e-05, + "loss": 0.4404, + "step": 34265 + }, + { + "epoch": 2.3777669835542294, + "grad_norm": 1.403106451034546, + "learning_rate": 7.455524558399611e-05, + "loss": 0.4882, + "step": 34266 + }, + { + "epoch": 2.3778363749913263, + "grad_norm": 1.3191603422164917, + "learning_rate": 7.454981193656061e-05, + "loss": 0.3786, + "step": 34267 + }, + { + "epoch": 2.3779057664284227, + "grad_norm": 1.2370320558547974, + "learning_rate": 7.454437836946844e-05, + "loss": 0.488, + "step": 34268 + }, + { + "epoch": 2.3779751578655195, + "grad_norm": 1.1644322872161865, + "learning_rate": 7.453894488273667e-05, + "loss": 0.3747, + "step": 34269 + }, + { + "epoch": 2.378044549302616, + "grad_norm": 1.3128503561019897, + "learning_rate": 7.45335114763825e-05, + "loss": 0.5464, + "step": 34270 + }, + { + "epoch": 2.3781139407397127, + "grad_norm": 1.3891817331314087, + "learning_rate": 7.452807815042309e-05, + "loss": 0.4598, + "step": 34271 + }, + { + "epoch": 2.3781833321768096, + "grad_norm": 1.1865129470825195, + "learning_rate": 7.452264490487561e-05, + "loss": 0.4119, + "step": 34272 + }, + { + "epoch": 2.378252723613906, + "grad_norm": 1.1288585662841797, + "learning_rate": 7.451721173975715e-05, + "loss": 0.4193, + "step": 34273 + }, + { + "epoch": 2.3783221150510028, + "grad_norm": 1.0957697629928589, + "learning_rate": 7.45117786550849e-05, + "loss": 0.427, + "step": 34274 + }, + { + "epoch": 2.378391506488099, + "grad_norm": 1.3296748399734497, + "learning_rate": 7.4506345650876e-05, + "loss": 0.4275, + "step": 34275 + }, + { + "epoch": 2.378460897925196, + "grad_norm": 1.118686556816101, + "learning_rate": 7.450091272714761e-05, + "loss": 0.3591, + "step": 34276 + }, + { + "epoch": 2.378530289362293, + "grad_norm": 0.89284747838974, + "learning_rate": 7.44954798839169e-05, + "loss": 0.2682, + "step": 34277 + }, + { + "epoch": 2.378599680799389, + "grad_norm": 1.1490793228149414, + "learning_rate": 7.449004712120098e-05, + "loss": 0.4124, + "step": 34278 + }, + { + "epoch": 2.378669072236486, + "grad_norm": 1.5035076141357422, + "learning_rate": 7.448461443901707e-05, + "loss": 0.3671, + "step": 34279 + }, + { + "epoch": 2.3787384636735824, + "grad_norm": 1.6983156204223633, + "learning_rate": 7.447918183738222e-05, + "loss": 0.4703, + "step": 34280 + }, + { + "epoch": 2.3788078551106793, + "grad_norm": 1.202152132987976, + "learning_rate": 7.447374931631363e-05, + "loss": 0.3006, + "step": 34281 + }, + { + "epoch": 2.378877246547776, + "grad_norm": 1.482277750968933, + "learning_rate": 7.446831687582847e-05, + "loss": 0.329, + "step": 34282 + }, + { + "epoch": 2.3789466379848725, + "grad_norm": 1.1530331373214722, + "learning_rate": 7.446288451594385e-05, + "loss": 0.346, + "step": 34283 + }, + { + "epoch": 2.3790160294219693, + "grad_norm": 1.5970927476882935, + "learning_rate": 7.445745223667699e-05, + "loss": 0.4013, + "step": 34284 + }, + { + "epoch": 2.379085420859066, + "grad_norm": 1.5359785556793213, + "learning_rate": 7.445202003804495e-05, + "loss": 0.4619, + "step": 34285 + }, + { + "epoch": 2.3791548122961625, + "grad_norm": 1.463069200515747, + "learning_rate": 7.444658792006494e-05, + "loss": 0.2112, + "step": 34286 + }, + { + "epoch": 2.3792242037332594, + "grad_norm": 1.3164905309677124, + "learning_rate": 7.444115588275407e-05, + "loss": 0.3856, + "step": 34287 + }, + { + "epoch": 2.379293595170356, + "grad_norm": 1.260530710220337, + "learning_rate": 7.443572392612952e-05, + "loss": 0.5838, + "step": 34288 + }, + { + "epoch": 2.3793629866074526, + "grad_norm": 1.2022351026535034, + "learning_rate": 7.443029205020842e-05, + "loss": 0.475, + "step": 34289 + }, + { + "epoch": 2.3794323780445494, + "grad_norm": 1.0216619968414307, + "learning_rate": 7.44248602550079e-05, + "loss": 0.4327, + "step": 34290 + }, + { + "epoch": 2.379501769481646, + "grad_norm": 1.4692020416259766, + "learning_rate": 7.441942854054514e-05, + "loss": 0.4177, + "step": 34291 + }, + { + "epoch": 2.3795711609187427, + "grad_norm": 1.3844953775405884, + "learning_rate": 7.441399690683728e-05, + "loss": 0.3131, + "step": 34292 + }, + { + "epoch": 2.3796405523558395, + "grad_norm": 1.178995132446289, + "learning_rate": 7.44085653539015e-05, + "loss": 0.394, + "step": 34293 + }, + { + "epoch": 2.379709943792936, + "grad_norm": 1.084112524986267, + "learning_rate": 7.440313388175486e-05, + "loss": 0.3456, + "step": 34294 + }, + { + "epoch": 2.3797793352300327, + "grad_norm": 1.182133436203003, + "learning_rate": 7.439770249041457e-05, + "loss": 0.2909, + "step": 34295 + }, + { + "epoch": 2.379848726667129, + "grad_norm": 1.5555799007415771, + "learning_rate": 7.439227117989775e-05, + "loss": 0.5784, + "step": 34296 + }, + { + "epoch": 2.379918118104226, + "grad_norm": 1.2400161027908325, + "learning_rate": 7.438683995022157e-05, + "loss": 0.5214, + "step": 34297 + }, + { + "epoch": 2.3799875095413228, + "grad_norm": 1.1629492044448853, + "learning_rate": 7.438140880140316e-05, + "loss": 0.2966, + "step": 34298 + }, + { + "epoch": 2.380056900978419, + "grad_norm": 1.355970025062561, + "learning_rate": 7.437597773345967e-05, + "loss": 0.461, + "step": 34299 + }, + { + "epoch": 2.380126292415516, + "grad_norm": 1.4881534576416016, + "learning_rate": 7.437054674640829e-05, + "loss": 0.5296, + "step": 34300 + }, + { + "epoch": 2.3801956838526124, + "grad_norm": 1.201892375946045, + "learning_rate": 7.436511584026607e-05, + "loss": 0.3695, + "step": 34301 + }, + { + "epoch": 2.380265075289709, + "grad_norm": 1.2628138065338135, + "learning_rate": 7.435968501505021e-05, + "loss": 0.4728, + "step": 34302 + }, + { + "epoch": 2.380334466726806, + "grad_norm": 1.147246241569519, + "learning_rate": 7.435425427077788e-05, + "loss": 0.4052, + "step": 34303 + }, + { + "epoch": 2.3804038581639024, + "grad_norm": 0.8796569108963013, + "learning_rate": 7.434882360746615e-05, + "loss": 0.2727, + "step": 34304 + }, + { + "epoch": 2.3804732496009993, + "grad_norm": 1.2249339818954468, + "learning_rate": 7.434339302513223e-05, + "loss": 0.5034, + "step": 34305 + }, + { + "epoch": 2.3805426410380957, + "grad_norm": 1.1179507970809937, + "learning_rate": 7.433796252379324e-05, + "loss": 0.2529, + "step": 34306 + }, + { + "epoch": 2.3806120324751925, + "grad_norm": 1.413327693939209, + "learning_rate": 7.433253210346636e-05, + "loss": 0.3442, + "step": 34307 + }, + { + "epoch": 2.3806814239122893, + "grad_norm": 1.7328702211380005, + "learning_rate": 7.432710176416866e-05, + "loss": 0.3038, + "step": 34308 + }, + { + "epoch": 2.3807508153493857, + "grad_norm": 1.328623652458191, + "learning_rate": 7.432167150591731e-05, + "loss": 0.4276, + "step": 34309 + }, + { + "epoch": 2.3808202067864825, + "grad_norm": 2.30359148979187, + "learning_rate": 7.431624132872947e-05, + "loss": 0.3702, + "step": 34310 + }, + { + "epoch": 2.3808895982235794, + "grad_norm": 1.0487604141235352, + "learning_rate": 7.43108112326223e-05, + "loss": 0.3551, + "step": 34311 + }, + { + "epoch": 2.3809589896606758, + "grad_norm": 1.5870527029037476, + "learning_rate": 7.430538121761288e-05, + "loss": 0.4292, + "step": 34312 + }, + { + "epoch": 2.3810283810977726, + "grad_norm": 1.1126172542572021, + "learning_rate": 7.429995128371841e-05, + "loss": 0.4538, + "step": 34313 + }, + { + "epoch": 2.3810977725348694, + "grad_norm": 1.2998967170715332, + "learning_rate": 7.429452143095605e-05, + "loss": 0.6688, + "step": 34314 + }, + { + "epoch": 2.381167163971966, + "grad_norm": 1.1114311218261719, + "learning_rate": 7.428909165934286e-05, + "loss": 0.2872, + "step": 34315 + }, + { + "epoch": 2.3812365554090626, + "grad_norm": 1.3161414861679077, + "learning_rate": 7.428366196889602e-05, + "loss": 0.3765, + "step": 34316 + }, + { + "epoch": 2.381305946846159, + "grad_norm": 1.2423360347747803, + "learning_rate": 7.42782323596327e-05, + "loss": 0.4926, + "step": 34317 + }, + { + "epoch": 2.381375338283256, + "grad_norm": 0.9773404002189636, + "learning_rate": 7.427280283156999e-05, + "loss": 0.3136, + "step": 34318 + }, + { + "epoch": 2.3814447297203527, + "grad_norm": 1.5689105987548828, + "learning_rate": 7.426737338472509e-05, + "loss": 0.5725, + "step": 34319 + }, + { + "epoch": 2.381514121157449, + "grad_norm": 1.2970987558364868, + "learning_rate": 7.426194401911508e-05, + "loss": 0.4383, + "step": 34320 + }, + { + "epoch": 2.381583512594546, + "grad_norm": 2.577112913131714, + "learning_rate": 7.425651473475714e-05, + "loss": 0.5742, + "step": 34321 + }, + { + "epoch": 2.3816529040316423, + "grad_norm": 1.3425205945968628, + "learning_rate": 7.425108553166842e-05, + "loss": 0.5489, + "step": 34322 + }, + { + "epoch": 2.381722295468739, + "grad_norm": 1.50177800655365, + "learning_rate": 7.4245656409866e-05, + "loss": 0.4799, + "step": 34323 + }, + { + "epoch": 2.381791686905836, + "grad_norm": 0.7081409692764282, + "learning_rate": 7.424022736936705e-05, + "loss": 0.2168, + "step": 34324 + }, + { + "epoch": 2.3818610783429324, + "grad_norm": 1.4324589967727661, + "learning_rate": 7.423479841018872e-05, + "loss": 0.4125, + "step": 34325 + }, + { + "epoch": 2.381930469780029, + "grad_norm": 0.7929942607879639, + "learning_rate": 7.422936953234813e-05, + "loss": 0.1838, + "step": 34326 + }, + { + "epoch": 2.3819998612171256, + "grad_norm": 1.1228505373001099, + "learning_rate": 7.42239407358625e-05, + "loss": 0.3595, + "step": 34327 + }, + { + "epoch": 2.3820692526542224, + "grad_norm": 1.3495092391967773, + "learning_rate": 7.421851202074884e-05, + "loss": 0.4549, + "step": 34328 + }, + { + "epoch": 2.3821386440913193, + "grad_norm": 1.5092662572860718, + "learning_rate": 7.421308338702434e-05, + "loss": 0.492, + "step": 34329 + }, + { + "epoch": 2.3822080355284156, + "grad_norm": 1.4481576681137085, + "learning_rate": 7.420765483470616e-05, + "loss": 0.546, + "step": 34330 + }, + { + "epoch": 2.3822774269655125, + "grad_norm": 1.3128368854522705, + "learning_rate": 7.42022263638114e-05, + "loss": 0.3071, + "step": 34331 + }, + { + "epoch": 2.3823468184026093, + "grad_norm": 1.3494985103607178, + "learning_rate": 7.419679797435724e-05, + "loss": 0.5024, + "step": 34332 + }, + { + "epoch": 2.3824162098397057, + "grad_norm": 1.9228267669677734, + "learning_rate": 7.419136966636078e-05, + "loss": 0.5077, + "step": 34333 + }, + { + "epoch": 2.3824856012768025, + "grad_norm": 1.2705330848693848, + "learning_rate": 7.418594143983921e-05, + "loss": 0.3997, + "step": 34334 + }, + { + "epoch": 2.382554992713899, + "grad_norm": 0.7966690063476562, + "learning_rate": 7.41805132948096e-05, + "loss": 0.2304, + "step": 34335 + }, + { + "epoch": 2.3826243841509958, + "grad_norm": 1.2113181352615356, + "learning_rate": 7.417508523128911e-05, + "loss": 0.4149, + "step": 34336 + }, + { + "epoch": 2.3826937755880926, + "grad_norm": 1.3021262884140015, + "learning_rate": 7.416965724929488e-05, + "loss": 0.3049, + "step": 34337 + }, + { + "epoch": 2.382763167025189, + "grad_norm": 1.190466284751892, + "learning_rate": 7.416422934884406e-05, + "loss": 0.4164, + "step": 34338 + }, + { + "epoch": 2.382832558462286, + "grad_norm": 0.8911415934562683, + "learning_rate": 7.415880152995374e-05, + "loss": 0.3165, + "step": 34339 + }, + { + "epoch": 2.3829019498993826, + "grad_norm": 1.4639885425567627, + "learning_rate": 7.41533737926411e-05, + "loss": 0.4128, + "step": 34340 + }, + { + "epoch": 2.382971341336479, + "grad_norm": 1.2267245054244995, + "learning_rate": 7.414794613692329e-05, + "loss": 0.3164, + "step": 34341 + }, + { + "epoch": 2.383040732773576, + "grad_norm": 1.1577755212783813, + "learning_rate": 7.41425185628174e-05, + "loss": 0.4625, + "step": 34342 + }, + { + "epoch": 2.3831101242106723, + "grad_norm": 1.2275093793869019, + "learning_rate": 7.413709107034057e-05, + "loss": 0.5131, + "step": 34343 + }, + { + "epoch": 2.383179515647769, + "grad_norm": 1.1457818746566772, + "learning_rate": 7.413166365950993e-05, + "loss": 0.3489, + "step": 34344 + }, + { + "epoch": 2.383248907084866, + "grad_norm": 1.6092923879623413, + "learning_rate": 7.412623633034264e-05, + "loss": 0.3959, + "step": 34345 + }, + { + "epoch": 2.3833182985219623, + "grad_norm": 1.3870381116867065, + "learning_rate": 7.41208090828558e-05, + "loss": 0.3519, + "step": 34346 + }, + { + "epoch": 2.383387689959059, + "grad_norm": 2.158463478088379, + "learning_rate": 7.411538191706657e-05, + "loss": 0.7635, + "step": 34347 + }, + { + "epoch": 2.3834570813961555, + "grad_norm": 1.5191423892974854, + "learning_rate": 7.410995483299214e-05, + "loss": 0.3257, + "step": 34348 + }, + { + "epoch": 2.3835264728332524, + "grad_norm": 1.1125833988189697, + "learning_rate": 7.410452783064952e-05, + "loss": 0.4483, + "step": 34349 + }, + { + "epoch": 2.383595864270349, + "grad_norm": 1.1561894416809082, + "learning_rate": 7.40991009100559e-05, + "loss": 0.381, + "step": 34350 + }, + { + "epoch": 2.3836652557074456, + "grad_norm": 1.538402795791626, + "learning_rate": 7.40936740712284e-05, + "loss": 0.4243, + "step": 34351 + }, + { + "epoch": 2.3837346471445424, + "grad_norm": 1.5025652647018433, + "learning_rate": 7.408824731418419e-05, + "loss": 0.3711, + "step": 34352 + }, + { + "epoch": 2.383804038581639, + "grad_norm": 1.0732496976852417, + "learning_rate": 7.408282063894036e-05, + "loss": 0.3664, + "step": 34353 + }, + { + "epoch": 2.3838734300187356, + "grad_norm": 1.2066686153411865, + "learning_rate": 7.407739404551408e-05, + "loss": 0.4027, + "step": 34354 + }, + { + "epoch": 2.3839428214558325, + "grad_norm": 1.5643569231033325, + "learning_rate": 7.407196753392248e-05, + "loss": 0.7616, + "step": 34355 + }, + { + "epoch": 2.384012212892929, + "grad_norm": 1.0850094556808472, + "learning_rate": 7.406654110418263e-05, + "loss": 0.3738, + "step": 34356 + }, + { + "epoch": 2.3840816043300257, + "grad_norm": 1.953910231590271, + "learning_rate": 7.406111475631172e-05, + "loss": 0.5435, + "step": 34357 + }, + { + "epoch": 2.3841509957671225, + "grad_norm": 1.2497389316558838, + "learning_rate": 7.405568849032685e-05, + "loss": 0.4215, + "step": 34358 + }, + { + "epoch": 2.384220387204219, + "grad_norm": 1.2125085592269897, + "learning_rate": 7.405026230624516e-05, + "loss": 0.4464, + "step": 34359 + }, + { + "epoch": 2.3842897786413157, + "grad_norm": 1.181562066078186, + "learning_rate": 7.404483620408378e-05, + "loss": 0.4088, + "step": 34360 + }, + { + "epoch": 2.384359170078412, + "grad_norm": 1.1232104301452637, + "learning_rate": 7.403941018385985e-05, + "loss": 0.4168, + "step": 34361 + }, + { + "epoch": 2.384428561515509, + "grad_norm": 1.3008514642715454, + "learning_rate": 7.403398424559052e-05, + "loss": 0.4896, + "step": 34362 + }, + { + "epoch": 2.384497952952606, + "grad_norm": 1.5669634342193604, + "learning_rate": 7.402855838929283e-05, + "loss": 0.3555, + "step": 34363 + }, + { + "epoch": 2.384567344389702, + "grad_norm": 1.0375053882598877, + "learning_rate": 7.4023132614984e-05, + "loss": 0.4437, + "step": 34364 + }, + { + "epoch": 2.384636735826799, + "grad_norm": 1.3515090942382812, + "learning_rate": 7.401770692268111e-05, + "loss": 0.4593, + "step": 34365 + }, + { + "epoch": 2.384706127263896, + "grad_norm": 1.1062736511230469, + "learning_rate": 7.40122813124013e-05, + "loss": 0.4029, + "step": 34366 + }, + { + "epoch": 2.3847755187009922, + "grad_norm": 1.0817323923110962, + "learning_rate": 7.400685578416172e-05, + "loss": 0.3638, + "step": 34367 + }, + { + "epoch": 2.384844910138089, + "grad_norm": 1.1549862623214722, + "learning_rate": 7.400143033797947e-05, + "loss": 0.3442, + "step": 34368 + }, + { + "epoch": 2.3849143015751855, + "grad_norm": 1.0295590162277222, + "learning_rate": 7.399600497387174e-05, + "loss": 0.3143, + "step": 34369 + }, + { + "epoch": 2.3849836930122823, + "grad_norm": 1.764162302017212, + "learning_rate": 7.399057969185554e-05, + "loss": 0.4996, + "step": 34370 + }, + { + "epoch": 2.385053084449379, + "grad_norm": 1.8797574043273926, + "learning_rate": 7.398515449194809e-05, + "loss": 0.4491, + "step": 34371 + }, + { + "epoch": 2.3851224758864755, + "grad_norm": 1.8616187572479248, + "learning_rate": 7.397972937416647e-05, + "loss": 0.5403, + "step": 34372 + }, + { + "epoch": 2.3851918673235724, + "grad_norm": 1.4598513841629028, + "learning_rate": 7.397430433852786e-05, + "loss": 0.6092, + "step": 34373 + }, + { + "epoch": 2.3852612587606687, + "grad_norm": 1.7196003198623657, + "learning_rate": 7.396887938504931e-05, + "loss": 0.4013, + "step": 34374 + }, + { + "epoch": 2.3853306501977656, + "grad_norm": 0.8324199318885803, + "learning_rate": 7.3963454513748e-05, + "loss": 0.208, + "step": 34375 + }, + { + "epoch": 2.3854000416348624, + "grad_norm": 1.2433549165725708, + "learning_rate": 7.39580297246411e-05, + "loss": 0.4304, + "step": 34376 + }, + { + "epoch": 2.385469433071959, + "grad_norm": 1.3417003154754639, + "learning_rate": 7.395260501774562e-05, + "loss": 0.4727, + "step": 34377 + }, + { + "epoch": 2.3855388245090556, + "grad_norm": 1.3020583391189575, + "learning_rate": 7.394718039307875e-05, + "loss": 0.4511, + "step": 34378 + }, + { + "epoch": 2.385608215946152, + "grad_norm": 1.246173620223999, + "learning_rate": 7.39417558506576e-05, + "loss": 0.3515, + "step": 34379 + }, + { + "epoch": 2.385677607383249, + "grad_norm": 0.8599367141723633, + "learning_rate": 7.393633139049932e-05, + "loss": 0.3058, + "step": 34380 + }, + { + "epoch": 2.3857469988203457, + "grad_norm": 1.2389519214630127, + "learning_rate": 7.3930907012621e-05, + "loss": 0.3955, + "step": 34381 + }, + { + "epoch": 2.385816390257442, + "grad_norm": 1.4994316101074219, + "learning_rate": 7.392548271703978e-05, + "loss": 0.2836, + "step": 34382 + }, + { + "epoch": 2.385885781694539, + "grad_norm": 1.0422282218933105, + "learning_rate": 7.392005850377284e-05, + "loss": 0.3385, + "step": 34383 + }, + { + "epoch": 2.3859551731316357, + "grad_norm": 0.9999737739562988, + "learning_rate": 7.391463437283722e-05, + "loss": 0.344, + "step": 34384 + }, + { + "epoch": 2.386024564568732, + "grad_norm": 1.1795337200164795, + "learning_rate": 7.390921032425004e-05, + "loss": 0.3671, + "step": 34385 + }, + { + "epoch": 2.386093956005829, + "grad_norm": 1.7881004810333252, + "learning_rate": 7.390378635802848e-05, + "loss": 0.4688, + "step": 34386 + }, + { + "epoch": 2.3861633474429254, + "grad_norm": 1.5969125032424927, + "learning_rate": 7.389836247418964e-05, + "loss": 0.656, + "step": 34387 + }, + { + "epoch": 2.386232738880022, + "grad_norm": 1.2967016696929932, + "learning_rate": 7.389293867275064e-05, + "loss": 0.3948, + "step": 34388 + }, + { + "epoch": 2.386302130317119, + "grad_norm": 1.468746304512024, + "learning_rate": 7.38875149537286e-05, + "loss": 0.4268, + "step": 34389 + }, + { + "epoch": 2.3863715217542154, + "grad_norm": 1.8800108432769775, + "learning_rate": 7.388209131714065e-05, + "loss": 0.3214, + "step": 34390 + }, + { + "epoch": 2.3864409131913122, + "grad_norm": 1.1416617631912231, + "learning_rate": 7.387666776300392e-05, + "loss": 0.4808, + "step": 34391 + }, + { + "epoch": 2.386510304628409, + "grad_norm": 1.0672709941864014, + "learning_rate": 7.387124429133552e-05, + "loss": 0.3385, + "step": 34392 + }, + { + "epoch": 2.3865796960655055, + "grad_norm": 1.4539237022399902, + "learning_rate": 7.386582090215255e-05, + "loss": 0.4709, + "step": 34393 + }, + { + "epoch": 2.3866490875026023, + "grad_norm": 1.3402628898620605, + "learning_rate": 7.386039759547214e-05, + "loss": 0.4646, + "step": 34394 + }, + { + "epoch": 2.3867184789396987, + "grad_norm": 1.4306193590164185, + "learning_rate": 7.385497437131143e-05, + "loss": 0.3711, + "step": 34395 + }, + { + "epoch": 2.3867878703767955, + "grad_norm": 1.2470414638519287, + "learning_rate": 7.384955122968754e-05, + "loss": 0.287, + "step": 34396 + }, + { + "epoch": 2.3868572618138924, + "grad_norm": 0.985271692276001, + "learning_rate": 7.384412817061763e-05, + "loss": 0.3975, + "step": 34397 + }, + { + "epoch": 2.3869266532509887, + "grad_norm": 0.9192506074905396, + "learning_rate": 7.383870519411871e-05, + "loss": 0.3305, + "step": 34398 + }, + { + "epoch": 2.3869960446880856, + "grad_norm": 1.3047816753387451, + "learning_rate": 7.383328230020797e-05, + "loss": 0.4509, + "step": 34399 + }, + { + "epoch": 2.387065436125182, + "grad_norm": 1.1015936136245728, + "learning_rate": 7.382785948890253e-05, + "loss": 0.4241, + "step": 34400 + }, + { + "epoch": 2.387134827562279, + "grad_norm": 1.257408857345581, + "learning_rate": 7.382243676021949e-05, + "loss": 0.5097, + "step": 34401 + }, + { + "epoch": 2.3872042189993756, + "grad_norm": 1.181114673614502, + "learning_rate": 7.381701411417599e-05, + "loss": 0.352, + "step": 34402 + }, + { + "epoch": 2.387273610436472, + "grad_norm": 1.3065935373306274, + "learning_rate": 7.381159155078919e-05, + "loss": 0.4749, + "step": 34403 + }, + { + "epoch": 2.387343001873569, + "grad_norm": 0.9391103386878967, + "learning_rate": 7.380616907007609e-05, + "loss": 0.3169, + "step": 34404 + }, + { + "epoch": 2.3874123933106652, + "grad_norm": 1.1578178405761719, + "learning_rate": 7.380074667205388e-05, + "loss": 0.392, + "step": 34405 + }, + { + "epoch": 2.387481784747762, + "grad_norm": 1.3378976583480835, + "learning_rate": 7.379532435673968e-05, + "loss": 0.4721, + "step": 34406 + }, + { + "epoch": 2.387551176184859, + "grad_norm": 1.0272172689437866, + "learning_rate": 7.37899021241506e-05, + "loss": 0.4149, + "step": 34407 + }, + { + "epoch": 2.3876205676219553, + "grad_norm": 1.5391161441802979, + "learning_rate": 7.378447997430377e-05, + "loss": 0.541, + "step": 34408 + }, + { + "epoch": 2.387689959059052, + "grad_norm": 1.431909203529358, + "learning_rate": 7.37790579072163e-05, + "loss": 0.2914, + "step": 34409 + }, + { + "epoch": 2.387759350496149, + "grad_norm": 0.9443779587745667, + "learning_rate": 7.37736359229053e-05, + "loss": 0.3218, + "step": 34410 + }, + { + "epoch": 2.3878287419332453, + "grad_norm": 1.4582138061523438, + "learning_rate": 7.37682140213879e-05, + "loss": 0.4192, + "step": 34411 + }, + { + "epoch": 2.387898133370342, + "grad_norm": 1.1131089925765991, + "learning_rate": 7.376279220268118e-05, + "loss": 0.3275, + "step": 34412 + }, + { + "epoch": 2.3879675248074386, + "grad_norm": 1.5442383289337158, + "learning_rate": 7.375737046680227e-05, + "loss": 0.58, + "step": 34413 + }, + { + "epoch": 2.3880369162445354, + "grad_norm": 1.335680365562439, + "learning_rate": 7.375194881376832e-05, + "loss": 0.4521, + "step": 34414 + }, + { + "epoch": 2.3881063076816322, + "grad_norm": 0.8049027919769287, + "learning_rate": 7.37465272435964e-05, + "loss": 0.3474, + "step": 34415 + }, + { + "epoch": 2.3881756991187286, + "grad_norm": 1.4624305963516235, + "learning_rate": 7.374110575630367e-05, + "loss": 0.3131, + "step": 34416 + }, + { + "epoch": 2.3882450905558255, + "grad_norm": 0.8834605813026428, + "learning_rate": 7.373568435190724e-05, + "loss": 0.3902, + "step": 34417 + }, + { + "epoch": 2.3883144819929223, + "grad_norm": 1.1503307819366455, + "learning_rate": 7.373026303042418e-05, + "loss": 0.361, + "step": 34418 + }, + { + "epoch": 2.3883838734300187, + "grad_norm": 1.2703150510787964, + "learning_rate": 7.372484179187163e-05, + "loss": 0.4542, + "step": 34419 + }, + { + "epoch": 2.3884532648671155, + "grad_norm": 1.0038959980010986, + "learning_rate": 7.371942063626671e-05, + "loss": 0.2686, + "step": 34420 + }, + { + "epoch": 2.388522656304212, + "grad_norm": 1.4590119123458862, + "learning_rate": 7.371399956362652e-05, + "loss": 0.6076, + "step": 34421 + }, + { + "epoch": 2.3885920477413087, + "grad_norm": 1.6432727575302124, + "learning_rate": 7.370857857396817e-05, + "loss": 0.5804, + "step": 34422 + }, + { + "epoch": 2.3886614391784056, + "grad_norm": 1.2826967239379883, + "learning_rate": 7.370315766730884e-05, + "loss": 0.3016, + "step": 34423 + }, + { + "epoch": 2.388730830615502, + "grad_norm": 1.6447752714157104, + "learning_rate": 7.369773684366557e-05, + "loss": 0.4793, + "step": 34424 + }, + { + "epoch": 2.388800222052599, + "grad_norm": 1.1724740266799927, + "learning_rate": 7.369231610305549e-05, + "loss": 0.5195, + "step": 34425 + }, + { + "epoch": 2.388869613489695, + "grad_norm": 1.3234446048736572, + "learning_rate": 7.368689544549569e-05, + "loss": 0.5199, + "step": 34426 + }, + { + "epoch": 2.388939004926792, + "grad_norm": 1.3680949211120605, + "learning_rate": 7.368147487100335e-05, + "loss": 0.5756, + "step": 34427 + }, + { + "epoch": 2.389008396363889, + "grad_norm": 1.950921893119812, + "learning_rate": 7.36760543795955e-05, + "loss": 0.5865, + "step": 34428 + }, + { + "epoch": 2.3890777878009852, + "grad_norm": 1.260998010635376, + "learning_rate": 7.367063397128932e-05, + "loss": 0.4933, + "step": 34429 + }, + { + "epoch": 2.389147179238082, + "grad_norm": 1.5533463954925537, + "learning_rate": 7.366521364610189e-05, + "loss": 0.5481, + "step": 34430 + }, + { + "epoch": 2.3892165706751785, + "grad_norm": 1.320621132850647, + "learning_rate": 7.365979340405034e-05, + "loss": 0.4553, + "step": 34431 + }, + { + "epoch": 2.3892859621122753, + "grad_norm": 1.1076231002807617, + "learning_rate": 7.365437324515174e-05, + "loss": 0.3806, + "step": 34432 + }, + { + "epoch": 2.389355353549372, + "grad_norm": 1.4312840700149536, + "learning_rate": 7.364895316942323e-05, + "loss": 0.542, + "step": 34433 + }, + { + "epoch": 2.3894247449864685, + "grad_norm": 1.2160900831222534, + "learning_rate": 7.36435331768819e-05, + "loss": 0.5617, + "step": 34434 + }, + { + "epoch": 2.3894941364235653, + "grad_norm": 1.3404169082641602, + "learning_rate": 7.363811326754491e-05, + "loss": 0.4797, + "step": 34435 + }, + { + "epoch": 2.389563527860662, + "grad_norm": 1.2994320392608643, + "learning_rate": 7.363269344142931e-05, + "loss": 0.5847, + "step": 34436 + }, + { + "epoch": 2.3896329192977586, + "grad_norm": 1.4575031995773315, + "learning_rate": 7.362727369855224e-05, + "loss": 0.4419, + "step": 34437 + }, + { + "epoch": 2.3897023107348554, + "grad_norm": 1.02031409740448, + "learning_rate": 7.362185403893088e-05, + "loss": 0.4131, + "step": 34438 + }, + { + "epoch": 2.3897717021719522, + "grad_norm": 1.264967441558838, + "learning_rate": 7.36164344625822e-05, + "loss": 0.6157, + "step": 34439 + }, + { + "epoch": 2.3898410936090486, + "grad_norm": 1.3519190549850464, + "learning_rate": 7.361101496952337e-05, + "loss": 0.3692, + "step": 34440 + }, + { + "epoch": 2.3899104850461454, + "grad_norm": 1.142785906791687, + "learning_rate": 7.360559555977151e-05, + "loss": 0.5288, + "step": 34441 + }, + { + "epoch": 2.389979876483242, + "grad_norm": 1.4911893606185913, + "learning_rate": 7.360017623334371e-05, + "loss": 0.3887, + "step": 34442 + }, + { + "epoch": 2.3900492679203387, + "grad_norm": 1.5868873596191406, + "learning_rate": 7.359475699025713e-05, + "loss": 0.4457, + "step": 34443 + }, + { + "epoch": 2.3901186593574355, + "grad_norm": 1.0255321264266968, + "learning_rate": 7.358933783052881e-05, + "loss": 0.3825, + "step": 34444 + }, + { + "epoch": 2.390188050794532, + "grad_norm": 1.3202170133590698, + "learning_rate": 7.358391875417592e-05, + "loss": 0.3248, + "step": 34445 + }, + { + "epoch": 2.3902574422316287, + "grad_norm": 1.2560882568359375, + "learning_rate": 7.357849976121552e-05, + "loss": 0.3686, + "step": 34446 + }, + { + "epoch": 2.390326833668725, + "grad_norm": 3.010606050491333, + "learning_rate": 7.357308085166473e-05, + "loss": 0.6091, + "step": 34447 + }, + { + "epoch": 2.390396225105822, + "grad_norm": 0.8154062032699585, + "learning_rate": 7.356766202554061e-05, + "loss": 0.2562, + "step": 34448 + }, + { + "epoch": 2.390465616542919, + "grad_norm": 2.4806995391845703, + "learning_rate": 7.356224328286037e-05, + "loss": 0.7757, + "step": 34449 + }, + { + "epoch": 2.390535007980015, + "grad_norm": 1.2830795049667358, + "learning_rate": 7.355682462364102e-05, + "loss": 0.4829, + "step": 34450 + }, + { + "epoch": 2.390604399417112, + "grad_norm": 1.2276780605316162, + "learning_rate": 7.355140604789973e-05, + "loss": 0.4041, + "step": 34451 + }, + { + "epoch": 2.3906737908542084, + "grad_norm": 1.4954177141189575, + "learning_rate": 7.354598755565361e-05, + "loss": 0.4063, + "step": 34452 + }, + { + "epoch": 2.3907431822913052, + "grad_norm": 1.128328800201416, + "learning_rate": 7.354056914691972e-05, + "loss": 0.4058, + "step": 34453 + }, + { + "epoch": 2.390812573728402, + "grad_norm": 1.1174602508544922, + "learning_rate": 7.353515082171516e-05, + "loss": 0.4496, + "step": 34454 + }, + { + "epoch": 2.3908819651654984, + "grad_norm": 1.3137904405593872, + "learning_rate": 7.352973258005706e-05, + "loss": 0.6224, + "step": 34455 + }, + { + "epoch": 2.3909513566025953, + "grad_norm": 1.0604913234710693, + "learning_rate": 7.352431442196253e-05, + "loss": 0.4513, + "step": 34456 + }, + { + "epoch": 2.3910207480396917, + "grad_norm": 1.1174490451812744, + "learning_rate": 7.351889634744867e-05, + "loss": 0.3319, + "step": 34457 + }, + { + "epoch": 2.3910901394767885, + "grad_norm": 1.660778284072876, + "learning_rate": 7.351347835653259e-05, + "loss": 0.4544, + "step": 34458 + }, + { + "epoch": 2.3911595309138853, + "grad_norm": 1.0181803703308105, + "learning_rate": 7.35080604492314e-05, + "loss": 0.4599, + "step": 34459 + }, + { + "epoch": 2.3912289223509817, + "grad_norm": 1.2244611978530884, + "learning_rate": 7.350264262556215e-05, + "loss": 0.243, + "step": 34460 + }, + { + "epoch": 2.3912983137880786, + "grad_norm": 4.171533107757568, + "learning_rate": 7.349722488554201e-05, + "loss": 0.652, + "step": 34461 + }, + { + "epoch": 2.3913677052251754, + "grad_norm": 0.9399585127830505, + "learning_rate": 7.349180722918805e-05, + "loss": 0.401, + "step": 34462 + }, + { + "epoch": 2.3914370966622718, + "grad_norm": 1.0493501424789429, + "learning_rate": 7.348638965651737e-05, + "loss": 0.286, + "step": 34463 + }, + { + "epoch": 2.3915064880993686, + "grad_norm": 1.3659054040908813, + "learning_rate": 7.348097216754709e-05, + "loss": 0.2546, + "step": 34464 + }, + { + "epoch": 2.3915758795364654, + "grad_norm": 0.7505674362182617, + "learning_rate": 7.34755547622943e-05, + "loss": 0.3447, + "step": 34465 + }, + { + "epoch": 2.391645270973562, + "grad_norm": 1.9450552463531494, + "learning_rate": 7.347013744077613e-05, + "loss": 0.5553, + "step": 34466 + }, + { + "epoch": 2.3917146624106587, + "grad_norm": 1.970372200012207, + "learning_rate": 7.346472020300964e-05, + "loss": 0.489, + "step": 34467 + }, + { + "epoch": 2.391784053847755, + "grad_norm": 1.6766172647476196, + "learning_rate": 7.345930304901194e-05, + "loss": 0.3669, + "step": 34468 + }, + { + "epoch": 2.391853445284852, + "grad_norm": 0.8568928241729736, + "learning_rate": 7.345388597880015e-05, + "loss": 0.2912, + "step": 34469 + }, + { + "epoch": 2.3919228367219487, + "grad_norm": 1.6006810665130615, + "learning_rate": 7.344846899239136e-05, + "loss": 0.4244, + "step": 34470 + }, + { + "epoch": 2.391992228159045, + "grad_norm": 1.225516438484192, + "learning_rate": 7.344305208980264e-05, + "loss": 0.4328, + "step": 34471 + }, + { + "epoch": 2.392061619596142, + "grad_norm": 1.7765058279037476, + "learning_rate": 7.343763527105116e-05, + "loss": 0.4977, + "step": 34472 + }, + { + "epoch": 2.3921310110332383, + "grad_norm": 1.666095495223999, + "learning_rate": 7.343221853615401e-05, + "loss": 0.4137, + "step": 34473 + }, + { + "epoch": 2.392200402470335, + "grad_norm": 1.525766134262085, + "learning_rate": 7.342680188512823e-05, + "loss": 0.5689, + "step": 34474 + }, + { + "epoch": 2.392269793907432, + "grad_norm": 1.247577428817749, + "learning_rate": 7.342138531799095e-05, + "loss": 0.5453, + "step": 34475 + }, + { + "epoch": 2.3923391853445284, + "grad_norm": 1.0728780031204224, + "learning_rate": 7.341596883475927e-05, + "loss": 0.3713, + "step": 34476 + }, + { + "epoch": 2.392408576781625, + "grad_norm": 1.2594026327133179, + "learning_rate": 7.341055243545032e-05, + "loss": 0.5209, + "step": 34477 + }, + { + "epoch": 2.3924779682187216, + "grad_norm": 1.6183483600616455, + "learning_rate": 7.340513612008116e-05, + "loss": 0.5167, + "step": 34478 + }, + { + "epoch": 2.3925473596558184, + "grad_norm": 1.3845969438552856, + "learning_rate": 7.339971988866889e-05, + "loss": 0.4376, + "step": 34479 + }, + { + "epoch": 2.3926167510929153, + "grad_norm": 0.9656965732574463, + "learning_rate": 7.339430374123062e-05, + "loss": 0.3088, + "step": 34480 + }, + { + "epoch": 2.3926861425300117, + "grad_norm": 1.798180341720581, + "learning_rate": 7.338888767778347e-05, + "loss": 0.5069, + "step": 34481 + }, + { + "epoch": 2.3927555339671085, + "grad_norm": 1.4988921880722046, + "learning_rate": 7.338347169834447e-05, + "loss": 0.4236, + "step": 34482 + }, + { + "epoch": 2.392824925404205, + "grad_norm": 1.5781488418579102, + "learning_rate": 7.337805580293077e-05, + "loss": 0.4254, + "step": 34483 + }, + { + "epoch": 2.3928943168413017, + "grad_norm": 1.6145673990249634, + "learning_rate": 7.337263999155948e-05, + "loss": 0.6218, + "step": 34484 + }, + { + "epoch": 2.3929637082783985, + "grad_norm": 1.3302956819534302, + "learning_rate": 7.336722426424765e-05, + "loss": 0.3743, + "step": 34485 + }, + { + "epoch": 2.393033099715495, + "grad_norm": 0.9814159870147705, + "learning_rate": 7.336180862101248e-05, + "loss": 0.3909, + "step": 34486 + }, + { + "epoch": 2.3931024911525918, + "grad_norm": 1.4554615020751953, + "learning_rate": 7.335639306187091e-05, + "loss": 0.3824, + "step": 34487 + }, + { + "epoch": 2.3931718825896886, + "grad_norm": 0.7180501818656921, + "learning_rate": 7.335097758684013e-05, + "loss": 0.2434, + "step": 34488 + }, + { + "epoch": 2.393241274026785, + "grad_norm": 1.5206845998764038, + "learning_rate": 7.334556219593722e-05, + "loss": 0.492, + "step": 34489 + }, + { + "epoch": 2.393310665463882, + "grad_norm": 0.9250586628913879, + "learning_rate": 7.334014688917929e-05, + "loss": 0.3452, + "step": 34490 + }, + { + "epoch": 2.3933800569009787, + "grad_norm": 1.0367668867111206, + "learning_rate": 7.33347316665834e-05, + "loss": 0.3484, + "step": 34491 + }, + { + "epoch": 2.393449448338075, + "grad_norm": 1.0557092428207397, + "learning_rate": 7.332931652816668e-05, + "loss": 0.4218, + "step": 34492 + }, + { + "epoch": 2.393518839775172, + "grad_norm": 0.9923600554466248, + "learning_rate": 7.332390147394626e-05, + "loss": 0.3224, + "step": 34493 + }, + { + "epoch": 2.3935882312122683, + "grad_norm": 1.0732717514038086, + "learning_rate": 7.331848650393915e-05, + "loss": 0.2977, + "step": 34494 + }, + { + "epoch": 2.393657622649365, + "grad_norm": 1.5307435989379883, + "learning_rate": 7.331307161816247e-05, + "loss": 0.327, + "step": 34495 + }, + { + "epoch": 2.393727014086462, + "grad_norm": 2.03900408744812, + "learning_rate": 7.330765681663334e-05, + "loss": 0.4312, + "step": 34496 + }, + { + "epoch": 2.3937964055235583, + "grad_norm": 1.0812623500823975, + "learning_rate": 7.330224209936884e-05, + "loss": 0.4308, + "step": 34497 + }, + { + "epoch": 2.393865796960655, + "grad_norm": 1.1153666973114014, + "learning_rate": 7.329682746638606e-05, + "loss": 0.3245, + "step": 34498 + }, + { + "epoch": 2.3939351883977515, + "grad_norm": 1.4366320371627808, + "learning_rate": 7.329141291770209e-05, + "loss": 0.463, + "step": 34499 + }, + { + "epoch": 2.3940045798348484, + "grad_norm": 1.2213410139083862, + "learning_rate": 7.328599845333408e-05, + "loss": 0.5118, + "step": 34500 + }, + { + "epoch": 2.394073971271945, + "grad_norm": 1.4913913011550903, + "learning_rate": 7.328058407329903e-05, + "loss": 0.2688, + "step": 34501 + }, + { + "epoch": 2.3941433627090416, + "grad_norm": 1.3673914670944214, + "learning_rate": 7.327516977761408e-05, + "loss": 0.4333, + "step": 34502 + }, + { + "epoch": 2.3942127541461384, + "grad_norm": 1.779371976852417, + "learning_rate": 7.326975556629632e-05, + "loss": 0.4964, + "step": 34503 + }, + { + "epoch": 2.394282145583235, + "grad_norm": 1.5945132970809937, + "learning_rate": 7.326434143936282e-05, + "loss": 0.5281, + "step": 34504 + }, + { + "epoch": 2.3943515370203317, + "grad_norm": 1.3144686222076416, + "learning_rate": 7.325892739683071e-05, + "loss": 0.3891, + "step": 34505 + }, + { + "epoch": 2.3944209284574285, + "grad_norm": 1.2181422710418701, + "learning_rate": 7.325351343871705e-05, + "loss": 0.4111, + "step": 34506 + }, + { + "epoch": 2.394490319894525, + "grad_norm": 1.2997747659683228, + "learning_rate": 7.3248099565039e-05, + "loss": 0.2966, + "step": 34507 + }, + { + "epoch": 2.3945597113316217, + "grad_norm": 1.532006025314331, + "learning_rate": 7.324268577581355e-05, + "loss": 0.4002, + "step": 34508 + }, + { + "epoch": 2.394629102768718, + "grad_norm": 1.1909290552139282, + "learning_rate": 7.323727207105784e-05, + "loss": 0.4029, + "step": 34509 + }, + { + "epoch": 2.394698494205815, + "grad_norm": 1.2251137495040894, + "learning_rate": 7.323185845078895e-05, + "loss": 0.3743, + "step": 34510 + }, + { + "epoch": 2.3947678856429118, + "grad_norm": 2.0220139026641846, + "learning_rate": 7.322644491502398e-05, + "loss": 0.4657, + "step": 34511 + }, + { + "epoch": 2.394837277080008, + "grad_norm": 1.350300908088684, + "learning_rate": 7.322103146378004e-05, + "loss": 0.4478, + "step": 34512 + }, + { + "epoch": 2.394906668517105, + "grad_norm": 1.0977165699005127, + "learning_rate": 7.321561809707418e-05, + "loss": 0.3607, + "step": 34513 + }, + { + "epoch": 2.394976059954202, + "grad_norm": 1.0276542901992798, + "learning_rate": 7.321020481492351e-05, + "loss": 0.3705, + "step": 34514 + }, + { + "epoch": 2.395045451391298, + "grad_norm": 1.097051978111267, + "learning_rate": 7.32047916173451e-05, + "loss": 0.4622, + "step": 34515 + }, + { + "epoch": 2.395114842828395, + "grad_norm": 1.4291296005249023, + "learning_rate": 7.31993785043561e-05, + "loss": 0.5357, + "step": 34516 + }, + { + "epoch": 2.395184234265492, + "grad_norm": 1.531759262084961, + "learning_rate": 7.31939654759735e-05, + "loss": 0.5914, + "step": 34517 + }, + { + "epoch": 2.3952536257025883, + "grad_norm": 1.4104562997817993, + "learning_rate": 7.318855253221446e-05, + "loss": 0.4824, + "step": 34518 + }, + { + "epoch": 2.395323017139685, + "grad_norm": 1.4681085348129272, + "learning_rate": 7.318313967309604e-05, + "loss": 0.5587, + "step": 34519 + }, + { + "epoch": 2.3953924085767815, + "grad_norm": 1.0230200290679932, + "learning_rate": 7.317772689863534e-05, + "loss": 0.4269, + "step": 34520 + }, + { + "epoch": 2.3954618000138783, + "grad_norm": 1.1172184944152832, + "learning_rate": 7.317231420884947e-05, + "loss": 0.378, + "step": 34521 + }, + { + "epoch": 2.395531191450975, + "grad_norm": 1.2360578775405884, + "learning_rate": 7.316690160375548e-05, + "loss": 0.5089, + "step": 34522 + }, + { + "epoch": 2.3956005828880715, + "grad_norm": 1.1749682426452637, + "learning_rate": 7.316148908337046e-05, + "loss": 0.3126, + "step": 34523 + }, + { + "epoch": 2.3956699743251684, + "grad_norm": 1.07793128490448, + "learning_rate": 7.31560766477115e-05, + "loss": 0.3091, + "step": 34524 + }, + { + "epoch": 2.3957393657622648, + "grad_norm": 1.085969090461731, + "learning_rate": 7.315066429679566e-05, + "loss": 0.3736, + "step": 34525 + }, + { + "epoch": 2.3958087571993616, + "grad_norm": 1.1074296236038208, + "learning_rate": 7.314525203064011e-05, + "loss": 0.3281, + "step": 34526 + }, + { + "epoch": 2.3958781486364584, + "grad_norm": 1.636734127998352, + "learning_rate": 7.313983984926185e-05, + "loss": 0.6202, + "step": 34527 + }, + { + "epoch": 2.395947540073555, + "grad_norm": 1.3905874490737915, + "learning_rate": 7.313442775267805e-05, + "loss": 0.4297, + "step": 34528 + }, + { + "epoch": 2.3960169315106516, + "grad_norm": 1.4950370788574219, + "learning_rate": 7.312901574090571e-05, + "loss": 0.4699, + "step": 34529 + }, + { + "epoch": 2.396086322947748, + "grad_norm": 1.742406964302063, + "learning_rate": 7.312360381396195e-05, + "loss": 0.4878, + "step": 34530 + }, + { + "epoch": 2.396155714384845, + "grad_norm": 1.3783526420593262, + "learning_rate": 7.311819197186387e-05, + "loss": 0.4958, + "step": 34531 + }, + { + "epoch": 2.3962251058219417, + "grad_norm": 1.3099087476730347, + "learning_rate": 7.31127802146285e-05, + "loss": 0.3568, + "step": 34532 + }, + { + "epoch": 2.396294497259038, + "grad_norm": 1.249657154083252, + "learning_rate": 7.3107368542273e-05, + "loss": 0.5085, + "step": 34533 + }, + { + "epoch": 2.396363888696135, + "grad_norm": 1.4901114702224731, + "learning_rate": 7.310195695481439e-05, + "loss": 0.4455, + "step": 34534 + }, + { + "epoch": 2.3964332801332313, + "grad_norm": 1.6970614194869995, + "learning_rate": 7.309654545226982e-05, + "loss": 0.5673, + "step": 34535 + }, + { + "epoch": 2.396502671570328, + "grad_norm": 1.387721300125122, + "learning_rate": 7.309113403465632e-05, + "loss": 0.5432, + "step": 34536 + }, + { + "epoch": 2.396572063007425, + "grad_norm": 1.1219111680984497, + "learning_rate": 7.308572270199097e-05, + "loss": 0.2957, + "step": 34537 + }, + { + "epoch": 2.3966414544445214, + "grad_norm": 0.9099122881889343, + "learning_rate": 7.308031145429088e-05, + "loss": 0.4175, + "step": 34538 + }, + { + "epoch": 2.396710845881618, + "grad_norm": 1.0272345542907715, + "learning_rate": 7.307490029157312e-05, + "loss": 0.4181, + "step": 34539 + }, + { + "epoch": 2.396780237318715, + "grad_norm": 1.1133605241775513, + "learning_rate": 7.306948921385477e-05, + "loss": 0.4864, + "step": 34540 + }, + { + "epoch": 2.3968496287558114, + "grad_norm": 1.4312745332717896, + "learning_rate": 7.30640782211529e-05, + "loss": 0.5222, + "step": 34541 + }, + { + "epoch": 2.3969190201929083, + "grad_norm": 1.9254168272018433, + "learning_rate": 7.305866731348468e-05, + "loss": 0.5159, + "step": 34542 + }, + { + "epoch": 2.396988411630005, + "grad_norm": 1.4608561992645264, + "learning_rate": 7.305325649086707e-05, + "loss": 0.5532, + "step": 34543 + }, + { + "epoch": 2.3970578030671015, + "grad_norm": 1.6890416145324707, + "learning_rate": 7.304784575331721e-05, + "loss": 0.3987, + "step": 34544 + }, + { + "epoch": 2.3971271945041983, + "grad_norm": 1.3047833442687988, + "learning_rate": 7.304243510085217e-05, + "loss": 0.4996, + "step": 34545 + }, + { + "epoch": 2.3971965859412947, + "grad_norm": 2.0222179889678955, + "learning_rate": 7.303702453348904e-05, + "loss": 0.5035, + "step": 34546 + }, + { + "epoch": 2.3972659773783915, + "grad_norm": 1.3616254329681396, + "learning_rate": 7.30316140512449e-05, + "loss": 0.4081, + "step": 34547 + }, + { + "epoch": 2.3973353688154884, + "grad_norm": 1.09894859790802, + "learning_rate": 7.302620365413681e-05, + "loss": 0.3686, + "step": 34548 + }, + { + "epoch": 2.3974047602525848, + "grad_norm": 1.3421813249588013, + "learning_rate": 7.302079334218191e-05, + "loss": 0.5368, + "step": 34549 + }, + { + "epoch": 2.3974741516896816, + "grad_norm": 1.74979567527771, + "learning_rate": 7.301538311539719e-05, + "loss": 0.4916, + "step": 34550 + }, + { + "epoch": 2.397543543126778, + "grad_norm": 1.584739327430725, + "learning_rate": 7.300997297379979e-05, + "loss": 0.3985, + "step": 34551 + }, + { + "epoch": 2.397612934563875, + "grad_norm": 1.0138473510742188, + "learning_rate": 7.300456291740677e-05, + "loss": 0.305, + "step": 34552 + }, + { + "epoch": 2.3976823260009716, + "grad_norm": 1.2533713579177856, + "learning_rate": 7.299915294623521e-05, + "loss": 0.3725, + "step": 34553 + }, + { + "epoch": 2.397751717438068, + "grad_norm": 1.6251296997070312, + "learning_rate": 7.299374306030218e-05, + "loss": 0.4774, + "step": 34554 + }, + { + "epoch": 2.397821108875165, + "grad_norm": 1.4238003492355347, + "learning_rate": 7.298833325962482e-05, + "loss": 0.4104, + "step": 34555 + }, + { + "epoch": 2.3978905003122613, + "grad_norm": 1.7883354425430298, + "learning_rate": 7.298292354422012e-05, + "loss": 0.3756, + "step": 34556 + }, + { + "epoch": 2.397959891749358, + "grad_norm": 1.6991491317749023, + "learning_rate": 7.297751391410519e-05, + "loss": 0.467, + "step": 34557 + }, + { + "epoch": 2.398029283186455, + "grad_norm": 1.6185859441757202, + "learning_rate": 7.29721043692971e-05, + "loss": 0.413, + "step": 34558 + }, + { + "epoch": 2.3980986746235513, + "grad_norm": 1.3440569639205933, + "learning_rate": 7.296669490981297e-05, + "loss": 0.5255, + "step": 34559 + }, + { + "epoch": 2.398168066060648, + "grad_norm": 1.4078692197799683, + "learning_rate": 7.296128553566984e-05, + "loss": 0.5257, + "step": 34560 + }, + { + "epoch": 2.3982374574977445, + "grad_norm": 1.804860234260559, + "learning_rate": 7.295587624688478e-05, + "loss": 0.5733, + "step": 34561 + }, + { + "epoch": 2.3983068489348414, + "grad_norm": 1.2428350448608398, + "learning_rate": 7.295046704347494e-05, + "loss": 0.4456, + "step": 34562 + }, + { + "epoch": 2.398376240371938, + "grad_norm": 1.253785490989685, + "learning_rate": 7.294505792545728e-05, + "loss": 0.4394, + "step": 34563 + }, + { + "epoch": 2.3984456318090346, + "grad_norm": 1.3964464664459229, + "learning_rate": 7.293964889284895e-05, + "loss": 0.7246, + "step": 34564 + }, + { + "epoch": 2.3985150232461314, + "grad_norm": 1.2323956489562988, + "learning_rate": 7.293423994566699e-05, + "loss": 0.4188, + "step": 34565 + }, + { + "epoch": 2.3985844146832282, + "grad_norm": 1.2770229578018188, + "learning_rate": 7.292883108392851e-05, + "loss": 0.5593, + "step": 34566 + }, + { + "epoch": 2.3986538061203246, + "grad_norm": 1.0204087495803833, + "learning_rate": 7.292342230765057e-05, + "loss": 0.3813, + "step": 34567 + }, + { + "epoch": 2.3987231975574215, + "grad_norm": 1.1478053331375122, + "learning_rate": 7.291801361685025e-05, + "loss": 0.382, + "step": 34568 + }, + { + "epoch": 2.3987925889945183, + "grad_norm": 1.565273642539978, + "learning_rate": 7.291260501154462e-05, + "loss": 0.4859, + "step": 34569 + }, + { + "epoch": 2.3988619804316147, + "grad_norm": 1.1864410638809204, + "learning_rate": 7.290719649175075e-05, + "loss": 0.3772, + "step": 34570 + }, + { + "epoch": 2.3989313718687115, + "grad_norm": 1.5307577848434448, + "learning_rate": 7.290178805748571e-05, + "loss": 0.6178, + "step": 34571 + }, + { + "epoch": 2.399000763305808, + "grad_norm": 1.1111253499984741, + "learning_rate": 7.289637970876657e-05, + "loss": 0.3415, + "step": 34572 + }, + { + "epoch": 2.3990701547429047, + "grad_norm": 1.254254698753357, + "learning_rate": 7.289097144561044e-05, + "loss": 0.3083, + "step": 34573 + }, + { + "epoch": 2.3991395461800016, + "grad_norm": 1.266472339630127, + "learning_rate": 7.288556326803432e-05, + "loss": 0.4305, + "step": 34574 + }, + { + "epoch": 2.399208937617098, + "grad_norm": 1.0297060012817383, + "learning_rate": 7.288015517605538e-05, + "loss": 0.3836, + "step": 34575 + }, + { + "epoch": 2.399278329054195, + "grad_norm": 1.0315568447113037, + "learning_rate": 7.287474716969067e-05, + "loss": 0.4414, + "step": 34576 + }, + { + "epoch": 2.399347720491291, + "grad_norm": 1.4240161180496216, + "learning_rate": 7.286933924895718e-05, + "loss": 0.5806, + "step": 34577 + }, + { + "epoch": 2.399417111928388, + "grad_norm": 1.2131545543670654, + "learning_rate": 7.286393141387204e-05, + "loss": 0.3392, + "step": 34578 + }, + { + "epoch": 2.399486503365485, + "grad_norm": 0.9303617477416992, + "learning_rate": 7.285852366445234e-05, + "loss": 0.3962, + "step": 34579 + }, + { + "epoch": 2.3995558948025812, + "grad_norm": 0.9463997483253479, + "learning_rate": 7.285311600071511e-05, + "loss": 0.3542, + "step": 34580 + }, + { + "epoch": 2.399625286239678, + "grad_norm": 1.1036525964736938, + "learning_rate": 7.284770842267747e-05, + "loss": 0.2326, + "step": 34581 + }, + { + "epoch": 2.3996946776767745, + "grad_norm": 1.2282297611236572, + "learning_rate": 7.284230093035647e-05, + "loss": 0.4119, + "step": 34582 + }, + { + "epoch": 2.3997640691138713, + "grad_norm": 1.244158148765564, + "learning_rate": 7.283689352376917e-05, + "loss": 0.2527, + "step": 34583 + }, + { + "epoch": 2.399833460550968, + "grad_norm": 1.2100849151611328, + "learning_rate": 7.283148620293263e-05, + "loss": 0.2924, + "step": 34584 + }, + { + "epoch": 2.3999028519880645, + "grad_norm": 1.3846148252487183, + "learning_rate": 7.282607896786393e-05, + "loss": 0.3829, + "step": 34585 + }, + { + "epoch": 2.3999722434251614, + "grad_norm": 3.3546738624572754, + "learning_rate": 7.282067181858016e-05, + "loss": 0.5411, + "step": 34586 + }, + { + "epoch": 2.400041634862258, + "grad_norm": 1.2146583795547485, + "learning_rate": 7.281526475509837e-05, + "loss": 0.4231, + "step": 34587 + }, + { + "epoch": 2.4001110262993546, + "grad_norm": 2.3111259937286377, + "learning_rate": 7.280985777743563e-05, + "loss": 0.6462, + "step": 34588 + }, + { + "epoch": 2.4001804177364514, + "grad_norm": 1.3807458877563477, + "learning_rate": 7.280445088560902e-05, + "loss": 0.4573, + "step": 34589 + }, + { + "epoch": 2.400249809173548, + "grad_norm": 1.041103482246399, + "learning_rate": 7.279904407963564e-05, + "loss": 0.4344, + "step": 34590 + }, + { + "epoch": 2.4003192006106446, + "grad_norm": 1.2194477319717407, + "learning_rate": 7.279363735953248e-05, + "loss": 0.4571, + "step": 34591 + }, + { + "epoch": 2.4003885920477415, + "grad_norm": 1.3288034200668335, + "learning_rate": 7.278823072531665e-05, + "loss": 0.3909, + "step": 34592 + }, + { + "epoch": 2.400457983484838, + "grad_norm": 1.0157177448272705, + "learning_rate": 7.278282417700522e-05, + "loss": 0.3652, + "step": 34593 + }, + { + "epoch": 2.4005273749219347, + "grad_norm": 0.983523964881897, + "learning_rate": 7.277741771461526e-05, + "loss": 0.4358, + "step": 34594 + }, + { + "epoch": 2.4005967663590315, + "grad_norm": 1.0646677017211914, + "learning_rate": 7.277201133816382e-05, + "loss": 0.4883, + "step": 34595 + }, + { + "epoch": 2.400666157796128, + "grad_norm": 1.093233346939087, + "learning_rate": 7.276660504766799e-05, + "loss": 0.3161, + "step": 34596 + }, + { + "epoch": 2.4007355492332247, + "grad_norm": 1.4835559129714966, + "learning_rate": 7.276119884314487e-05, + "loss": 0.4599, + "step": 34597 + }, + { + "epoch": 2.400804940670321, + "grad_norm": 1.423097014427185, + "learning_rate": 7.275579272461146e-05, + "loss": 0.4233, + "step": 34598 + }, + { + "epoch": 2.400874332107418, + "grad_norm": 1.4773491621017456, + "learning_rate": 7.275038669208483e-05, + "loss": 0.4258, + "step": 34599 + }, + { + "epoch": 2.400943723544515, + "grad_norm": 1.4070227146148682, + "learning_rate": 7.274498074558207e-05, + "loss": 0.6668, + "step": 34600 + }, + { + "epoch": 2.401013114981611, + "grad_norm": 1.3869621753692627, + "learning_rate": 7.273957488512027e-05, + "loss": 0.432, + "step": 34601 + }, + { + "epoch": 2.401082506418708, + "grad_norm": 1.3514366149902344, + "learning_rate": 7.273416911071645e-05, + "loss": 0.4319, + "step": 34602 + }, + { + "epoch": 2.4011518978558044, + "grad_norm": 1.6009838581085205, + "learning_rate": 7.272876342238768e-05, + "loss": 0.5186, + "step": 34603 + }, + { + "epoch": 2.4012212892929012, + "grad_norm": 1.1554088592529297, + "learning_rate": 7.272335782015107e-05, + "loss": 0.426, + "step": 34604 + }, + { + "epoch": 2.401290680729998, + "grad_norm": 2.114912986755371, + "learning_rate": 7.271795230402366e-05, + "loss": 0.3663, + "step": 34605 + }, + { + "epoch": 2.4013600721670945, + "grad_norm": 1.287825107574463, + "learning_rate": 7.271254687402248e-05, + "loss": 0.4664, + "step": 34606 + }, + { + "epoch": 2.4014294636041913, + "grad_norm": 1.440092921257019, + "learning_rate": 7.270714153016463e-05, + "loss": 0.4228, + "step": 34607 + }, + { + "epoch": 2.4014988550412877, + "grad_norm": 1.1229251623153687, + "learning_rate": 7.270173627246717e-05, + "loss": 0.5188, + "step": 34608 + }, + { + "epoch": 2.4015682464783845, + "grad_norm": 1.101286768913269, + "learning_rate": 7.269633110094715e-05, + "loss": 0.4438, + "step": 34609 + }, + { + "epoch": 2.4016376379154813, + "grad_norm": 1.756604552268982, + "learning_rate": 7.269092601562167e-05, + "loss": 0.5298, + "step": 34610 + }, + { + "epoch": 2.4017070293525777, + "grad_norm": 1.3751047849655151, + "learning_rate": 7.268552101650778e-05, + "loss": 0.5578, + "step": 34611 + }, + { + "epoch": 2.4017764207896746, + "grad_norm": 1.4495842456817627, + "learning_rate": 7.268011610362251e-05, + "loss": 0.615, + "step": 34612 + }, + { + "epoch": 2.4018458122267714, + "grad_norm": 1.1842610836029053, + "learning_rate": 7.267471127698293e-05, + "loss": 0.4553, + "step": 34613 + }, + { + "epoch": 2.401915203663868, + "grad_norm": 1.633291482925415, + "learning_rate": 7.266930653660612e-05, + "loss": 0.3697, + "step": 34614 + }, + { + "epoch": 2.4019845951009646, + "grad_norm": 1.396957516670227, + "learning_rate": 7.266390188250914e-05, + "loss": 0.4066, + "step": 34615 + }, + { + "epoch": 2.402053986538061, + "grad_norm": 1.2808659076690674, + "learning_rate": 7.265849731470905e-05, + "loss": 0.4467, + "step": 34616 + }, + { + "epoch": 2.402123377975158, + "grad_norm": 1.3479664325714111, + "learning_rate": 7.265309283322293e-05, + "loss": 0.5166, + "step": 34617 + }, + { + "epoch": 2.4021927694122547, + "grad_norm": 1.2130968570709229, + "learning_rate": 7.264768843806784e-05, + "loss": 0.4325, + "step": 34618 + }, + { + "epoch": 2.402262160849351, + "grad_norm": 1.1902296543121338, + "learning_rate": 7.264228412926077e-05, + "loss": 0.5345, + "step": 34619 + }, + { + "epoch": 2.402331552286448, + "grad_norm": 1.1280097961425781, + "learning_rate": 7.263687990681888e-05, + "loss": 0.4267, + "step": 34620 + }, + { + "epoch": 2.4024009437235447, + "grad_norm": 1.2840781211853027, + "learning_rate": 7.263147577075916e-05, + "loss": 0.4795, + "step": 34621 + }, + { + "epoch": 2.402470335160641, + "grad_norm": 1.173191785812378, + "learning_rate": 7.26260717210987e-05, + "loss": 0.5116, + "step": 34622 + }, + { + "epoch": 2.402539726597738, + "grad_norm": 1.3152472972869873, + "learning_rate": 7.262066775785454e-05, + "loss": 0.2742, + "step": 34623 + }, + { + "epoch": 2.4026091180348343, + "grad_norm": 1.3855583667755127, + "learning_rate": 7.261526388104377e-05, + "loss": 0.4446, + "step": 34624 + }, + { + "epoch": 2.402678509471931, + "grad_norm": 1.0444416999816895, + "learning_rate": 7.260986009068347e-05, + "loss": 0.3549, + "step": 34625 + }, + { + "epoch": 2.402747900909028, + "grad_norm": 1.3277164697647095, + "learning_rate": 7.260445638679063e-05, + "loss": 0.4076, + "step": 34626 + }, + { + "epoch": 2.4028172923461244, + "grad_norm": 1.2115669250488281, + "learning_rate": 7.259905276938233e-05, + "loss": 0.4404, + "step": 34627 + }, + { + "epoch": 2.4028866837832212, + "grad_norm": 0.8938391208648682, + "learning_rate": 7.259364923847564e-05, + "loss": 0.2925, + "step": 34628 + }, + { + "epoch": 2.4029560752203176, + "grad_norm": 1.386393666267395, + "learning_rate": 7.258824579408764e-05, + "loss": 0.4208, + "step": 34629 + }, + { + "epoch": 2.4030254666574145, + "grad_norm": 1.3961273431777954, + "learning_rate": 7.258284243623536e-05, + "loss": 0.4834, + "step": 34630 + }, + { + "epoch": 2.4030948580945113, + "grad_norm": 1.2932422161102295, + "learning_rate": 7.25774391649359e-05, + "loss": 0.4072, + "step": 34631 + }, + { + "epoch": 2.4031642495316077, + "grad_norm": 1.0227994918823242, + "learning_rate": 7.257203598020624e-05, + "loss": 0.4112, + "step": 34632 + }, + { + "epoch": 2.4032336409687045, + "grad_norm": 1.4073008298873901, + "learning_rate": 7.256663288206348e-05, + "loss": 0.3868, + "step": 34633 + }, + { + "epoch": 2.403303032405801, + "grad_norm": 1.1855038404464722, + "learning_rate": 7.256122987052469e-05, + "loss": 0.2776, + "step": 34634 + }, + { + "epoch": 2.4033724238428977, + "grad_norm": 0.7647286653518677, + "learning_rate": 7.255582694560689e-05, + "loss": 0.2628, + "step": 34635 + }, + { + "epoch": 2.4034418152799946, + "grad_norm": 0.8946401476860046, + "learning_rate": 7.255042410732721e-05, + "loss": 0.3724, + "step": 34636 + }, + { + "epoch": 2.403511206717091, + "grad_norm": 1.1720341444015503, + "learning_rate": 7.254502135570262e-05, + "loss": 0.4476, + "step": 34637 + }, + { + "epoch": 2.403580598154188, + "grad_norm": 1.3603299856185913, + "learning_rate": 7.253961869075023e-05, + "loss": 0.4237, + "step": 34638 + }, + { + "epoch": 2.4036499895912846, + "grad_norm": 1.159559726715088, + "learning_rate": 7.253421611248706e-05, + "loss": 0.3541, + "step": 34639 + }, + { + "epoch": 2.403719381028381, + "grad_norm": 1.3447469472885132, + "learning_rate": 7.25288136209302e-05, + "loss": 0.4663, + "step": 34640 + }, + { + "epoch": 2.403788772465478, + "grad_norm": 1.2487715482711792, + "learning_rate": 7.252341121609668e-05, + "loss": 0.3703, + "step": 34641 + }, + { + "epoch": 2.4038581639025742, + "grad_norm": 1.835933804512024, + "learning_rate": 7.251800889800355e-05, + "loss": 0.3882, + "step": 34642 + }, + { + "epoch": 2.403927555339671, + "grad_norm": 1.0208075046539307, + "learning_rate": 7.25126066666679e-05, + "loss": 0.3401, + "step": 34643 + }, + { + "epoch": 2.403996946776768, + "grad_norm": 1.296060562133789, + "learning_rate": 7.250720452210673e-05, + "loss": 0.3622, + "step": 34644 + }, + { + "epoch": 2.4040663382138643, + "grad_norm": 1.0174121856689453, + "learning_rate": 7.250180246433718e-05, + "loss": 0.4069, + "step": 34645 + }, + { + "epoch": 2.404135729650961, + "grad_norm": 0.9058216214179993, + "learning_rate": 7.249640049337622e-05, + "loss": 0.3644, + "step": 34646 + }, + { + "epoch": 2.404205121088058, + "grad_norm": 1.3212378025054932, + "learning_rate": 7.249099860924092e-05, + "loss": 0.5163, + "step": 34647 + }, + { + "epoch": 2.4042745125251543, + "grad_norm": 1.258843183517456, + "learning_rate": 7.248559681194835e-05, + "loss": 0.3837, + "step": 34648 + }, + { + "epoch": 2.404343903962251, + "grad_norm": 1.191658854484558, + "learning_rate": 7.248019510151556e-05, + "loss": 0.5385, + "step": 34649 + }, + { + "epoch": 2.4044132953993476, + "grad_norm": 1.2078670263290405, + "learning_rate": 7.247479347795959e-05, + "loss": 0.4388, + "step": 34650 + }, + { + "epoch": 2.4044826868364444, + "grad_norm": 1.6140589714050293, + "learning_rate": 7.246939194129752e-05, + "loss": 0.6343, + "step": 34651 + }, + { + "epoch": 2.4045520782735412, + "grad_norm": 1.2336621284484863, + "learning_rate": 7.246399049154642e-05, + "loss": 0.3356, + "step": 34652 + }, + { + "epoch": 2.4046214697106376, + "grad_norm": 1.0626648664474487, + "learning_rate": 7.245858912872325e-05, + "loss": 0.4282, + "step": 34653 + }, + { + "epoch": 2.4046908611477344, + "grad_norm": 1.3620331287384033, + "learning_rate": 7.245318785284514e-05, + "loss": 0.396, + "step": 34654 + }, + { + "epoch": 2.404760252584831, + "grad_norm": 3.638544797897339, + "learning_rate": 7.244778666392913e-05, + "loss": 0.4026, + "step": 34655 + }, + { + "epoch": 2.4048296440219277, + "grad_norm": 1.3802179098129272, + "learning_rate": 7.244238556199225e-05, + "loss": 0.3765, + "step": 34656 + }, + { + "epoch": 2.4048990354590245, + "grad_norm": 1.0377552509307861, + "learning_rate": 7.243698454705155e-05, + "loss": 0.4627, + "step": 34657 + }, + { + "epoch": 2.404968426896121, + "grad_norm": 1.2537518739700317, + "learning_rate": 7.24315836191241e-05, + "loss": 0.4285, + "step": 34658 + }, + { + "epoch": 2.4050378183332177, + "grad_norm": 0.9217780232429504, + "learning_rate": 7.242618277822698e-05, + "loss": 0.271, + "step": 34659 + }, + { + "epoch": 2.405107209770314, + "grad_norm": 1.388497233390808, + "learning_rate": 7.242078202437717e-05, + "loss": 0.3847, + "step": 34660 + }, + { + "epoch": 2.405176601207411, + "grad_norm": 1.2284010648727417, + "learning_rate": 7.241538135759174e-05, + "loss": 0.4133, + "step": 34661 + }, + { + "epoch": 2.405245992644508, + "grad_norm": 1.110900640487671, + "learning_rate": 7.240998077788774e-05, + "loss": 0.4361, + "step": 34662 + }, + { + "epoch": 2.405315384081604, + "grad_norm": 1.3140093088150024, + "learning_rate": 7.240458028528225e-05, + "loss": 0.4734, + "step": 34663 + }, + { + "epoch": 2.405384775518701, + "grad_norm": 1.538963794708252, + "learning_rate": 7.239917987979229e-05, + "loss": 0.5349, + "step": 34664 + }, + { + "epoch": 2.405454166955798, + "grad_norm": 1.1868431568145752, + "learning_rate": 7.239377956143492e-05, + "loss": 0.2293, + "step": 34665 + }, + { + "epoch": 2.4055235583928942, + "grad_norm": 1.0520390272140503, + "learning_rate": 7.238837933022722e-05, + "loss": 0.3271, + "step": 34666 + }, + { + "epoch": 2.405592949829991, + "grad_norm": 0.9906743168830872, + "learning_rate": 7.238297918618617e-05, + "loss": 0.2975, + "step": 34667 + }, + { + "epoch": 2.405662341267088, + "grad_norm": 1.4730521440505981, + "learning_rate": 7.237757912932884e-05, + "loss": 0.5305, + "step": 34668 + }, + { + "epoch": 2.4057317327041843, + "grad_norm": 2.41908860206604, + "learning_rate": 7.237217915967229e-05, + "loss": 0.5321, + "step": 34669 + }, + { + "epoch": 2.405801124141281, + "grad_norm": 1.3904805183410645, + "learning_rate": 7.236677927723357e-05, + "loss": 0.3389, + "step": 34670 + }, + { + "epoch": 2.4058705155783775, + "grad_norm": 1.1821104288101196, + "learning_rate": 7.236137948202975e-05, + "loss": 0.4856, + "step": 34671 + }, + { + "epoch": 2.4059399070154743, + "grad_norm": 2.0647175312042236, + "learning_rate": 7.235597977407782e-05, + "loss": 0.4725, + "step": 34672 + }, + { + "epoch": 2.406009298452571, + "grad_norm": 0.8802003860473633, + "learning_rate": 7.235058015339486e-05, + "loss": 0.3409, + "step": 34673 + }, + { + "epoch": 2.4060786898896676, + "grad_norm": 1.433805227279663, + "learning_rate": 7.234518061999793e-05, + "loss": 0.2907, + "step": 34674 + }, + { + "epoch": 2.4061480813267644, + "grad_norm": 0.9557947516441345, + "learning_rate": 7.233978117390403e-05, + "loss": 0.3009, + "step": 34675 + }, + { + "epoch": 2.4062174727638608, + "grad_norm": 1.1525475978851318, + "learning_rate": 7.233438181513024e-05, + "loss": 0.4311, + "step": 34676 + }, + { + "epoch": 2.4062868642009576, + "grad_norm": 1.5432510375976562, + "learning_rate": 7.232898254369359e-05, + "loss": 0.4143, + "step": 34677 + }, + { + "epoch": 2.4063562556380544, + "grad_norm": 0.9624992609024048, + "learning_rate": 7.232358335961112e-05, + "loss": 0.2962, + "step": 34678 + }, + { + "epoch": 2.406425647075151, + "grad_norm": 1.0474369525909424, + "learning_rate": 7.23181842628999e-05, + "loss": 0.3524, + "step": 34679 + }, + { + "epoch": 2.4064950385122477, + "grad_norm": 1.1052418947219849, + "learning_rate": 7.2312785253577e-05, + "loss": 0.4616, + "step": 34680 + }, + { + "epoch": 2.406564429949344, + "grad_norm": 1.1767644882202148, + "learning_rate": 7.230738633165938e-05, + "loss": 0.4391, + "step": 34681 + }, + { + "epoch": 2.406633821386441, + "grad_norm": 1.6263028383255005, + "learning_rate": 7.230198749716413e-05, + "loss": 0.4574, + "step": 34682 + }, + { + "epoch": 2.4067032128235377, + "grad_norm": 1.1706815958023071, + "learning_rate": 7.229658875010829e-05, + "loss": 0.2514, + "step": 34683 + }, + { + "epoch": 2.406772604260634, + "grad_norm": 1.931641936302185, + "learning_rate": 7.229119009050892e-05, + "loss": 0.3013, + "step": 34684 + }, + { + "epoch": 2.406841995697731, + "grad_norm": 1.19324791431427, + "learning_rate": 7.228579151838303e-05, + "loss": 0.3272, + "step": 34685 + }, + { + "epoch": 2.4069113871348273, + "grad_norm": 0.8571451306343079, + "learning_rate": 7.228039303374771e-05, + "loss": 0.3222, + "step": 34686 + }, + { + "epoch": 2.406980778571924, + "grad_norm": 1.3464046716690063, + "learning_rate": 7.227499463661997e-05, + "loss": 0.3777, + "step": 34687 + }, + { + "epoch": 2.407050170009021, + "grad_norm": 1.1846469640731812, + "learning_rate": 7.226959632701683e-05, + "loss": 0.4143, + "step": 34688 + }, + { + "epoch": 2.4071195614461174, + "grad_norm": 1.7414895296096802, + "learning_rate": 7.226419810495536e-05, + "loss": 0.4518, + "step": 34689 + }, + { + "epoch": 2.407188952883214, + "grad_norm": 1.2494149208068848, + "learning_rate": 7.225879997045262e-05, + "loss": 0.2427, + "step": 34690 + }, + { + "epoch": 2.407258344320311, + "grad_norm": 1.4503978490829468, + "learning_rate": 7.22534019235256e-05, + "loss": 0.3563, + "step": 34691 + }, + { + "epoch": 2.4073277357574074, + "grad_norm": 2.6591362953186035, + "learning_rate": 7.224800396419139e-05, + "loss": 0.5206, + "step": 34692 + }, + { + "epoch": 2.4073971271945043, + "grad_norm": 2.3835043907165527, + "learning_rate": 7.2242606092467e-05, + "loss": 0.7483, + "step": 34693 + }, + { + "epoch": 2.407466518631601, + "grad_norm": 1.5824017524719238, + "learning_rate": 7.223720830836953e-05, + "loss": 0.3534, + "step": 34694 + }, + { + "epoch": 2.4075359100686975, + "grad_norm": 0.7813633680343628, + "learning_rate": 7.223181061191592e-05, + "loss": 0.2782, + "step": 34695 + }, + { + "epoch": 2.4076053015057943, + "grad_norm": 1.3220939636230469, + "learning_rate": 7.222641300312327e-05, + "loss": 0.3531, + "step": 34696 + }, + { + "epoch": 2.4076746929428907, + "grad_norm": 1.3072304725646973, + "learning_rate": 7.222101548200862e-05, + "loss": 0.3955, + "step": 34697 + }, + { + "epoch": 2.4077440843799875, + "grad_norm": 1.441821813583374, + "learning_rate": 7.221561804858899e-05, + "loss": 0.5486, + "step": 34698 + }, + { + "epoch": 2.4078134758170844, + "grad_norm": 1.1815272569656372, + "learning_rate": 7.221022070288144e-05, + "loss": 0.4252, + "step": 34699 + }, + { + "epoch": 2.4078828672541808, + "grad_norm": 1.0914969444274902, + "learning_rate": 7.220482344490298e-05, + "loss": 0.3626, + "step": 34700 + }, + { + "epoch": 2.4079522586912776, + "grad_norm": 1.3395429849624634, + "learning_rate": 7.219942627467073e-05, + "loss": 0.4516, + "step": 34701 + }, + { + "epoch": 2.408021650128374, + "grad_norm": 1.3461812734603882, + "learning_rate": 7.219402919220161e-05, + "loss": 0.3901, + "step": 34702 + }, + { + "epoch": 2.408091041565471, + "grad_norm": 0.786614179611206, + "learning_rate": 7.218863219751273e-05, + "loss": 0.3088, + "step": 34703 + }, + { + "epoch": 2.4081604330025677, + "grad_norm": 1.34452223777771, + "learning_rate": 7.21832352906211e-05, + "loss": 0.4174, + "step": 34704 + }, + { + "epoch": 2.408229824439664, + "grad_norm": 1.275364875793457, + "learning_rate": 7.217783847154375e-05, + "loss": 0.2836, + "step": 34705 + }, + { + "epoch": 2.408299215876761, + "grad_norm": 1.2514373064041138, + "learning_rate": 7.217244174029778e-05, + "loss": 0.4073, + "step": 34706 + }, + { + "epoch": 2.4083686073138573, + "grad_norm": 6.8680291175842285, + "learning_rate": 7.216704509690017e-05, + "loss": 0.5431, + "step": 34707 + }, + { + "epoch": 2.408437998750954, + "grad_norm": 1.156468391418457, + "learning_rate": 7.216164854136798e-05, + "loss": 0.312, + "step": 34708 + }, + { + "epoch": 2.408507390188051, + "grad_norm": 1.901496410369873, + "learning_rate": 7.215625207371823e-05, + "loss": 0.3217, + "step": 34709 + }, + { + "epoch": 2.4085767816251473, + "grad_norm": 1.434242844581604, + "learning_rate": 7.215085569396793e-05, + "loss": 0.5097, + "step": 34710 + }, + { + "epoch": 2.408646173062244, + "grad_norm": 1.2160753011703491, + "learning_rate": 7.214545940213418e-05, + "loss": 0.5495, + "step": 34711 + }, + { + "epoch": 2.4087155644993405, + "grad_norm": 1.4285249710083008, + "learning_rate": 7.214006319823397e-05, + "loss": 0.3711, + "step": 34712 + }, + { + "epoch": 2.4087849559364374, + "grad_norm": 1.728671908378601, + "learning_rate": 7.213466708228433e-05, + "loss": 0.4913, + "step": 34713 + }, + { + "epoch": 2.408854347373534, + "grad_norm": 1.2874646186828613, + "learning_rate": 7.212927105430236e-05, + "loss": 0.4945, + "step": 34714 + }, + { + "epoch": 2.4089237388106306, + "grad_norm": 1.2380326986312866, + "learning_rate": 7.212387511430502e-05, + "loss": 0.5043, + "step": 34715 + }, + { + "epoch": 2.4089931302477274, + "grad_norm": 1.1529425382614136, + "learning_rate": 7.211847926230937e-05, + "loss": 0.4336, + "step": 34716 + }, + { + "epoch": 2.4090625216848243, + "grad_norm": 1.2293915748596191, + "learning_rate": 7.211308349833244e-05, + "loss": 0.569, + "step": 34717 + }, + { + "epoch": 2.4091319131219207, + "grad_norm": 1.1433830261230469, + "learning_rate": 7.210768782239128e-05, + "loss": 0.3478, + "step": 34718 + }, + { + "epoch": 2.4092013045590175, + "grad_norm": 1.2263695001602173, + "learning_rate": 7.21022922345029e-05, + "loss": 0.3614, + "step": 34719 + }, + { + "epoch": 2.4092706959961143, + "grad_norm": 1.8003733158111572, + "learning_rate": 7.209689673468435e-05, + "loss": 0.5487, + "step": 34720 + }, + { + "epoch": 2.4093400874332107, + "grad_norm": 1.4328782558441162, + "learning_rate": 7.20915013229527e-05, + "loss": 0.3807, + "step": 34721 + }, + { + "epoch": 2.4094094788703075, + "grad_norm": 1.7901809215545654, + "learning_rate": 7.20861059993249e-05, + "loss": 0.55, + "step": 34722 + }, + { + "epoch": 2.409478870307404, + "grad_norm": 1.3895847797393799, + "learning_rate": 7.208071076381804e-05, + "loss": 0.3251, + "step": 34723 + }, + { + "epoch": 2.4095482617445008, + "grad_norm": 1.1407406330108643, + "learning_rate": 7.207531561644912e-05, + "loss": 0.4447, + "step": 34724 + }, + { + "epoch": 2.4096176531815976, + "grad_norm": 1.1508921384811401, + "learning_rate": 7.206992055723522e-05, + "loss": 0.4559, + "step": 34725 + }, + { + "epoch": 2.409687044618694, + "grad_norm": 1.838356852531433, + "learning_rate": 7.206452558619331e-05, + "loss": 0.4314, + "step": 34726 + }, + { + "epoch": 2.409756436055791, + "grad_norm": 0.9999113082885742, + "learning_rate": 7.205913070334045e-05, + "loss": 0.2564, + "step": 34727 + }, + { + "epoch": 2.409825827492887, + "grad_norm": 1.6683375835418701, + "learning_rate": 7.205373590869373e-05, + "loss": 0.5022, + "step": 34728 + }, + { + "epoch": 2.409895218929984, + "grad_norm": 1.1203570365905762, + "learning_rate": 7.204834120227006e-05, + "loss": 0.3631, + "step": 34729 + }, + { + "epoch": 2.409964610367081, + "grad_norm": 1.012708306312561, + "learning_rate": 7.204294658408656e-05, + "loss": 0.3482, + "step": 34730 + }, + { + "epoch": 2.4100340018041773, + "grad_norm": 1.351492166519165, + "learning_rate": 7.203755205416021e-05, + "loss": 0.4443, + "step": 34731 + }, + { + "epoch": 2.410103393241274, + "grad_norm": 1.3185447454452515, + "learning_rate": 7.203215761250808e-05, + "loss": 0.3055, + "step": 34732 + }, + { + "epoch": 2.4101727846783705, + "grad_norm": 1.0797390937805176, + "learning_rate": 7.202676325914717e-05, + "loss": 0.2572, + "step": 34733 + }, + { + "epoch": 2.4102421761154673, + "grad_norm": 1.4099255800247192, + "learning_rate": 7.202136899409454e-05, + "loss": 0.3838, + "step": 34734 + }, + { + "epoch": 2.410311567552564, + "grad_norm": 1.2130628824234009, + "learning_rate": 7.201597481736724e-05, + "loss": 0.4529, + "step": 34735 + }, + { + "epoch": 2.4103809589896605, + "grad_norm": 1.0227893590927124, + "learning_rate": 7.201058072898222e-05, + "loss": 0.3586, + "step": 34736 + }, + { + "epoch": 2.4104503504267574, + "grad_norm": 1.4150216579437256, + "learning_rate": 7.200518672895653e-05, + "loss": 0.407, + "step": 34737 + }, + { + "epoch": 2.4105197418638538, + "grad_norm": 0.8633798956871033, + "learning_rate": 7.199979281730722e-05, + "loss": 0.2761, + "step": 34738 + }, + { + "epoch": 2.4105891333009506, + "grad_norm": 1.4526636600494385, + "learning_rate": 7.199439899405134e-05, + "loss": 0.3689, + "step": 34739 + }, + { + "epoch": 2.4106585247380474, + "grad_norm": 1.087948203086853, + "learning_rate": 7.198900525920589e-05, + "loss": 0.4486, + "step": 34740 + }, + { + "epoch": 2.410727916175144, + "grad_norm": 1.2252628803253174, + "learning_rate": 7.19836116127879e-05, + "loss": 0.4659, + "step": 34741 + }, + { + "epoch": 2.4107973076122406, + "grad_norm": 1.0815554857254028, + "learning_rate": 7.197821805481439e-05, + "loss": 0.3428, + "step": 34742 + }, + { + "epoch": 2.4108666990493375, + "grad_norm": 1.6667742729187012, + "learning_rate": 7.19728245853024e-05, + "loss": 0.5504, + "step": 34743 + }, + { + "epoch": 2.410936090486434, + "grad_norm": 1.2034426927566528, + "learning_rate": 7.196743120426897e-05, + "loss": 0.4406, + "step": 34744 + }, + { + "epoch": 2.4110054819235307, + "grad_norm": 1.3761872053146362, + "learning_rate": 7.196203791173107e-05, + "loss": 0.5644, + "step": 34745 + }, + { + "epoch": 2.4110748733606275, + "grad_norm": 1.5058417320251465, + "learning_rate": 7.195664470770577e-05, + "loss": 0.4942, + "step": 34746 + }, + { + "epoch": 2.411144264797724, + "grad_norm": 1.2505682706832886, + "learning_rate": 7.19512515922101e-05, + "loss": 0.4095, + "step": 34747 + }, + { + "epoch": 2.4112136562348208, + "grad_norm": 1.788392424583435, + "learning_rate": 7.194585856526106e-05, + "loss": 0.5607, + "step": 34748 + }, + { + "epoch": 2.411283047671917, + "grad_norm": 0.9601188898086548, + "learning_rate": 7.194046562687575e-05, + "loss": 0.2463, + "step": 34749 + }, + { + "epoch": 2.411352439109014, + "grad_norm": 0.9583092331886292, + "learning_rate": 7.193507277707108e-05, + "loss": 0.4437, + "step": 34750 + }, + { + "epoch": 2.411421830546111, + "grad_norm": 1.9989731311798096, + "learning_rate": 7.192968001586413e-05, + "loss": 0.301, + "step": 34751 + }, + { + "epoch": 2.411491221983207, + "grad_norm": 1.370935320854187, + "learning_rate": 7.192428734327192e-05, + "loss": 0.5292, + "step": 34752 + }, + { + "epoch": 2.411560613420304, + "grad_norm": 1.1598061323165894, + "learning_rate": 7.191889475931148e-05, + "loss": 0.3981, + "step": 34753 + }, + { + "epoch": 2.4116300048574004, + "grad_norm": 1.2298372983932495, + "learning_rate": 7.191350226399983e-05, + "loss": 0.4434, + "step": 34754 + }, + { + "epoch": 2.4116993962944973, + "grad_norm": 1.2716143131256104, + "learning_rate": 7.190810985735398e-05, + "loss": 0.3672, + "step": 34755 + }, + { + "epoch": 2.411768787731594, + "grad_norm": 1.1691899299621582, + "learning_rate": 7.190271753939104e-05, + "loss": 0.4704, + "step": 34756 + }, + { + "epoch": 2.4118381791686905, + "grad_norm": 1.604170322418213, + "learning_rate": 7.189732531012791e-05, + "loss": 0.4922, + "step": 34757 + }, + { + "epoch": 2.4119075706057873, + "grad_norm": 1.90709388256073, + "learning_rate": 7.189193316958165e-05, + "loss": 0.6444, + "step": 34758 + }, + { + "epoch": 2.4119769620428837, + "grad_norm": 1.003096580505371, + "learning_rate": 7.188654111776931e-05, + "loss": 0.3804, + "step": 34759 + }, + { + "epoch": 2.4120463534799805, + "grad_norm": 2.288477897644043, + "learning_rate": 7.188114915470791e-05, + "loss": 0.795, + "step": 34760 + }, + { + "epoch": 2.4121157449170774, + "grad_norm": 1.4266294240951538, + "learning_rate": 7.187575728041444e-05, + "loss": 0.4809, + "step": 34761 + }, + { + "epoch": 2.4121851363541738, + "grad_norm": 1.1606600284576416, + "learning_rate": 7.187036549490596e-05, + "loss": 0.5332, + "step": 34762 + }, + { + "epoch": 2.4122545277912706, + "grad_norm": 0.9347432851791382, + "learning_rate": 7.186497379819948e-05, + "loss": 0.3041, + "step": 34763 + }, + { + "epoch": 2.412323919228367, + "grad_norm": 1.2756190299987793, + "learning_rate": 7.1859582190312e-05, + "loss": 0.4918, + "step": 34764 + }, + { + "epoch": 2.412393310665464, + "grad_norm": 1.8806657791137695, + "learning_rate": 7.185419067126054e-05, + "loss": 0.4171, + "step": 34765 + }, + { + "epoch": 2.4124627021025606, + "grad_norm": 1.460635781288147, + "learning_rate": 7.184879924106213e-05, + "loss": 0.4382, + "step": 34766 + }, + { + "epoch": 2.412532093539657, + "grad_norm": 1.3009535074234009, + "learning_rate": 7.18434078997338e-05, + "loss": 0.4138, + "step": 34767 + }, + { + "epoch": 2.412601484976754, + "grad_norm": 1.179288625717163, + "learning_rate": 7.183801664729257e-05, + "loss": 0.4615, + "step": 34768 + }, + { + "epoch": 2.4126708764138507, + "grad_norm": 1.5099396705627441, + "learning_rate": 7.183262548375546e-05, + "loss": 0.6263, + "step": 34769 + }, + { + "epoch": 2.412740267850947, + "grad_norm": 1.3066879510879517, + "learning_rate": 7.182723440913952e-05, + "loss": 0.5177, + "step": 34770 + }, + { + "epoch": 2.412809659288044, + "grad_norm": 1.6367257833480835, + "learning_rate": 7.182184342346168e-05, + "loss": 0.5461, + "step": 34771 + }, + { + "epoch": 2.4128790507251408, + "grad_norm": 1.1098461151123047, + "learning_rate": 7.181645252673901e-05, + "loss": 0.3169, + "step": 34772 + }, + { + "epoch": 2.412948442162237, + "grad_norm": 1.1457449197769165, + "learning_rate": 7.181106171898854e-05, + "loss": 0.4307, + "step": 34773 + }, + { + "epoch": 2.413017833599334, + "grad_norm": 1.4799284934997559, + "learning_rate": 7.180567100022729e-05, + "loss": 0.614, + "step": 34774 + }, + { + "epoch": 2.4130872250364304, + "grad_norm": 1.036378264427185, + "learning_rate": 7.180028037047226e-05, + "loss": 0.3764, + "step": 34775 + }, + { + "epoch": 2.413156616473527, + "grad_norm": 1.3193968534469604, + "learning_rate": 7.179488982974047e-05, + "loss": 0.4206, + "step": 34776 + }, + { + "epoch": 2.413226007910624, + "grad_norm": 1.3517497777938843, + "learning_rate": 7.178949937804895e-05, + "loss": 0.4874, + "step": 34777 + }, + { + "epoch": 2.4132953993477204, + "grad_norm": 1.1217492818832397, + "learning_rate": 7.17841090154147e-05, + "loss": 0.2599, + "step": 34778 + }, + { + "epoch": 2.4133647907848172, + "grad_norm": 1.913487195968628, + "learning_rate": 7.177871874185476e-05, + "loss": 0.567, + "step": 34779 + }, + { + "epoch": 2.4134341822219136, + "grad_norm": 1.1035553216934204, + "learning_rate": 7.17733285573861e-05, + "loss": 0.3746, + "step": 34780 + }, + { + "epoch": 2.4135035736590105, + "grad_norm": 1.6341869831085205, + "learning_rate": 7.176793846202576e-05, + "loss": 0.6538, + "step": 34781 + }, + { + "epoch": 2.4135729650961073, + "grad_norm": 1.7059804201126099, + "learning_rate": 7.176254845579077e-05, + "loss": 0.4688, + "step": 34782 + }, + { + "epoch": 2.4136423565332037, + "grad_norm": 1.057028889656067, + "learning_rate": 7.17571585386982e-05, + "loss": 0.4605, + "step": 34783 + }, + { + "epoch": 2.4137117479703005, + "grad_norm": 1.154286503791809, + "learning_rate": 7.175176871076495e-05, + "loss": 0.3865, + "step": 34784 + }, + { + "epoch": 2.413781139407397, + "grad_norm": 1.4318394660949707, + "learning_rate": 7.174637897200809e-05, + "loss": 0.519, + "step": 34785 + }, + { + "epoch": 2.4138505308444937, + "grad_norm": 1.5863642692565918, + "learning_rate": 7.174098932244463e-05, + "loss": 0.5642, + "step": 34786 + }, + { + "epoch": 2.4139199222815906, + "grad_norm": 1.585988998413086, + "learning_rate": 7.173559976209158e-05, + "loss": 0.3588, + "step": 34787 + }, + { + "epoch": 2.413989313718687, + "grad_norm": 1.4652976989746094, + "learning_rate": 7.173021029096599e-05, + "loss": 0.5977, + "step": 34788 + }, + { + "epoch": 2.414058705155784, + "grad_norm": 1.3498889207839966, + "learning_rate": 7.172482090908482e-05, + "loss": 0.3769, + "step": 34789 + }, + { + "epoch": 2.41412809659288, + "grad_norm": 1.7005953788757324, + "learning_rate": 7.171943161646515e-05, + "loss": 0.4632, + "step": 34790 + }, + { + "epoch": 2.414197488029977, + "grad_norm": 1.0483355522155762, + "learning_rate": 7.171404241312391e-05, + "loss": 0.2374, + "step": 34791 + }, + { + "epoch": 2.414266879467074, + "grad_norm": 1.2325847148895264, + "learning_rate": 7.170865329907818e-05, + "loss": 0.2824, + "step": 34792 + }, + { + "epoch": 2.4143362709041702, + "grad_norm": 1.6767001152038574, + "learning_rate": 7.170326427434492e-05, + "loss": 0.4389, + "step": 34793 + }, + { + "epoch": 2.414405662341267, + "grad_norm": 0.9245907068252563, + "learning_rate": 7.16978753389412e-05, + "loss": 0.2453, + "step": 34794 + }, + { + "epoch": 2.414475053778364, + "grad_norm": 1.2151634693145752, + "learning_rate": 7.169248649288401e-05, + "loss": 0.3385, + "step": 34795 + }, + { + "epoch": 2.4145444452154603, + "grad_norm": 1.0512341260910034, + "learning_rate": 7.168709773619034e-05, + "loss": 0.3764, + "step": 34796 + }, + { + "epoch": 2.414613836652557, + "grad_norm": 1.453835129737854, + "learning_rate": 7.168170906887723e-05, + "loss": 0.5078, + "step": 34797 + }, + { + "epoch": 2.414683228089654, + "grad_norm": 1.8849568367004395, + "learning_rate": 7.167632049096169e-05, + "loss": 0.4378, + "step": 34798 + }, + { + "epoch": 2.4147526195267504, + "grad_norm": 1.3449746370315552, + "learning_rate": 7.167093200246069e-05, + "loss": 0.3633, + "step": 34799 + }, + { + "epoch": 2.414822010963847, + "grad_norm": 1.1487737894058228, + "learning_rate": 7.166554360339128e-05, + "loss": 0.2772, + "step": 34800 + }, + { + "epoch": 2.4148914024009436, + "grad_norm": 1.0014351606369019, + "learning_rate": 7.166015529377045e-05, + "loss": 0.4301, + "step": 34801 + }, + { + "epoch": 2.4149607938380404, + "grad_norm": 1.0888255834579468, + "learning_rate": 7.165476707361523e-05, + "loss": 0.3912, + "step": 34802 + }, + { + "epoch": 2.4150301852751372, + "grad_norm": 1.0326156616210938, + "learning_rate": 7.164937894294261e-05, + "loss": 0.4259, + "step": 34803 + }, + { + "epoch": 2.4150995767122336, + "grad_norm": 1.2490469217300415, + "learning_rate": 7.164399090176967e-05, + "loss": 0.3498, + "step": 34804 + }, + { + "epoch": 2.4151689681493305, + "grad_norm": 1.253253698348999, + "learning_rate": 7.163860295011331e-05, + "loss": 0.3779, + "step": 34805 + }, + { + "epoch": 2.415238359586427, + "grad_norm": 1.019378423690796, + "learning_rate": 7.163321508799059e-05, + "loss": 0.384, + "step": 34806 + }, + { + "epoch": 2.4153077510235237, + "grad_norm": 1.0878615379333496, + "learning_rate": 7.162782731541852e-05, + "loss": 0.4119, + "step": 34807 + }, + { + "epoch": 2.4153771424606205, + "grad_norm": 1.0990492105484009, + "learning_rate": 7.16224396324141e-05, + "loss": 0.4092, + "step": 34808 + }, + { + "epoch": 2.415446533897717, + "grad_norm": 1.1407301425933838, + "learning_rate": 7.161705203899436e-05, + "loss": 0.5269, + "step": 34809 + }, + { + "epoch": 2.4155159253348137, + "grad_norm": 1.4161462783813477, + "learning_rate": 7.161166453517632e-05, + "loss": 0.4702, + "step": 34810 + }, + { + "epoch": 2.41558531677191, + "grad_norm": 1.1305493116378784, + "learning_rate": 7.160627712097694e-05, + "loss": 0.3114, + "step": 34811 + }, + { + "epoch": 2.415654708209007, + "grad_norm": 1.3946093320846558, + "learning_rate": 7.160088979641325e-05, + "loss": 0.3477, + "step": 34812 + }, + { + "epoch": 2.415724099646104, + "grad_norm": 1.553063988685608, + "learning_rate": 7.159550256150223e-05, + "loss": 0.6809, + "step": 34813 + }, + { + "epoch": 2.4157934910832, + "grad_norm": 1.2113336324691772, + "learning_rate": 7.159011541626095e-05, + "loss": 0.2596, + "step": 34814 + }, + { + "epoch": 2.415862882520297, + "grad_norm": 1.112504243850708, + "learning_rate": 7.158472836070635e-05, + "loss": 0.39, + "step": 34815 + }, + { + "epoch": 2.415932273957394, + "grad_norm": 1.397729516029358, + "learning_rate": 7.157934139485547e-05, + "loss": 0.4899, + "step": 34816 + }, + { + "epoch": 2.4160016653944902, + "grad_norm": 1.1097640991210938, + "learning_rate": 7.157395451872532e-05, + "loss": 0.2994, + "step": 34817 + }, + { + "epoch": 2.416071056831587, + "grad_norm": 1.366876244544983, + "learning_rate": 7.156856773233294e-05, + "loss": 0.3392, + "step": 34818 + }, + { + "epoch": 2.4161404482686835, + "grad_norm": 1.3814743757247925, + "learning_rate": 7.156318103569524e-05, + "loss": 0.4019, + "step": 34819 + }, + { + "epoch": 2.4162098397057803, + "grad_norm": 1.5051140785217285, + "learning_rate": 7.155779442882929e-05, + "loss": 0.6489, + "step": 34820 + }, + { + "epoch": 2.416279231142877, + "grad_norm": 1.61922025680542, + "learning_rate": 7.155240791175206e-05, + "loss": 0.5072, + "step": 34821 + }, + { + "epoch": 2.4163486225799735, + "grad_norm": 0.8511173725128174, + "learning_rate": 7.154702148448059e-05, + "loss": 0.2638, + "step": 34822 + }, + { + "epoch": 2.4164180140170703, + "grad_norm": 0.9388003349304199, + "learning_rate": 7.154163514703188e-05, + "loss": 0.4121, + "step": 34823 + }, + { + "epoch": 2.416487405454167, + "grad_norm": 1.149350643157959, + "learning_rate": 7.153624889942292e-05, + "loss": 0.5042, + "step": 34824 + }, + { + "epoch": 2.4165567968912636, + "grad_norm": 1.4809606075286865, + "learning_rate": 7.153086274167076e-05, + "loss": 0.3744, + "step": 34825 + }, + { + "epoch": 2.4166261883283604, + "grad_norm": 0.9135017395019531, + "learning_rate": 7.152547667379232e-05, + "loss": 0.4087, + "step": 34826 + }, + { + "epoch": 2.416695579765457, + "grad_norm": 1.1354038715362549, + "learning_rate": 7.152009069580466e-05, + "loss": 0.4573, + "step": 34827 + }, + { + "epoch": 2.4167649712025536, + "grad_norm": 2.0898168087005615, + "learning_rate": 7.151470480772475e-05, + "loss": 0.7011, + "step": 34828 + }, + { + "epoch": 2.4168343626396505, + "grad_norm": 1.3503113985061646, + "learning_rate": 7.150931900956965e-05, + "loss": 0.4483, + "step": 34829 + }, + { + "epoch": 2.416903754076747, + "grad_norm": 2.0705907344818115, + "learning_rate": 7.150393330135629e-05, + "loss": 0.4843, + "step": 34830 + }, + { + "epoch": 2.4169731455138437, + "grad_norm": 1.2309930324554443, + "learning_rate": 7.149854768310171e-05, + "loss": 0.4331, + "step": 34831 + }, + { + "epoch": 2.41704253695094, + "grad_norm": 1.2646598815917969, + "learning_rate": 7.149316215482294e-05, + "loss": 0.258, + "step": 34832 + }, + { + "epoch": 2.417111928388037, + "grad_norm": 1.2213854789733887, + "learning_rate": 7.148777671653694e-05, + "loss": 0.4354, + "step": 34833 + }, + { + "epoch": 2.4171813198251337, + "grad_norm": 1.3874483108520508, + "learning_rate": 7.14823913682607e-05, + "loss": 0.4617, + "step": 34834 + }, + { + "epoch": 2.41725071126223, + "grad_norm": 1.1909621953964233, + "learning_rate": 7.147700611001124e-05, + "loss": 0.2808, + "step": 34835 + }, + { + "epoch": 2.417320102699327, + "grad_norm": 1.3383066654205322, + "learning_rate": 7.147162094180558e-05, + "loss": 0.4333, + "step": 34836 + }, + { + "epoch": 2.4173894941364233, + "grad_norm": 1.3584791421890259, + "learning_rate": 7.146623586366068e-05, + "loss": 0.3068, + "step": 34837 + }, + { + "epoch": 2.41745888557352, + "grad_norm": 1.1166625022888184, + "learning_rate": 7.146085087559357e-05, + "loss": 0.4216, + "step": 34838 + }, + { + "epoch": 2.417528277010617, + "grad_norm": 1.26176118850708, + "learning_rate": 7.145546597762129e-05, + "loss": 0.3886, + "step": 34839 + }, + { + "epoch": 2.4175976684477134, + "grad_norm": 1.867343544960022, + "learning_rate": 7.145008116976073e-05, + "loss": 0.5709, + "step": 34840 + }, + { + "epoch": 2.4176670598848102, + "grad_norm": 1.2148144245147705, + "learning_rate": 7.144469645202898e-05, + "loss": 0.406, + "step": 34841 + }, + { + "epoch": 2.417736451321907, + "grad_norm": 1.2921195030212402, + "learning_rate": 7.1439311824443e-05, + "loss": 0.3981, + "step": 34842 + }, + { + "epoch": 2.4178058427590035, + "grad_norm": 1.3197057247161865, + "learning_rate": 7.143392728701979e-05, + "loss": 0.4187, + "step": 34843 + }, + { + "epoch": 2.4178752341961003, + "grad_norm": 1.3837913274765015, + "learning_rate": 7.142854283977635e-05, + "loss": 0.5266, + "step": 34844 + }, + { + "epoch": 2.4179446256331967, + "grad_norm": 1.2048444747924805, + "learning_rate": 7.142315848272973e-05, + "loss": 0.3943, + "step": 34845 + }, + { + "epoch": 2.4180140170702935, + "grad_norm": 1.0423307418823242, + "learning_rate": 7.141777421589689e-05, + "loss": 0.4092, + "step": 34846 + }, + { + "epoch": 2.4180834085073903, + "grad_norm": 1.216024398803711, + "learning_rate": 7.141239003929477e-05, + "loss": 0.6035, + "step": 34847 + }, + { + "epoch": 2.4181527999444867, + "grad_norm": 1.5791370868682861, + "learning_rate": 7.140700595294043e-05, + "loss": 0.3757, + "step": 34848 + }, + { + "epoch": 2.4182221913815836, + "grad_norm": 1.4933990240097046, + "learning_rate": 7.140162195685088e-05, + "loss": 0.6262, + "step": 34849 + }, + { + "epoch": 2.4182915828186804, + "grad_norm": 1.5957216024398804, + "learning_rate": 7.139623805104306e-05, + "loss": 0.5958, + "step": 34850 + }, + { + "epoch": 2.418360974255777, + "grad_norm": 1.2965528964996338, + "learning_rate": 7.139085423553401e-05, + "loss": 0.395, + "step": 34851 + }, + { + "epoch": 2.4184303656928736, + "grad_norm": 1.7085680961608887, + "learning_rate": 7.13854705103407e-05, + "loss": 0.4349, + "step": 34852 + }, + { + "epoch": 2.41849975712997, + "grad_norm": 1.0750170946121216, + "learning_rate": 7.13800868754802e-05, + "loss": 0.3976, + "step": 34853 + }, + { + "epoch": 2.418569148567067, + "grad_norm": 1.7802953720092773, + "learning_rate": 7.137470333096938e-05, + "loss": 0.4221, + "step": 34854 + }, + { + "epoch": 2.4186385400041637, + "grad_norm": 1.7396489381790161, + "learning_rate": 7.136931987682532e-05, + "loss": 0.3861, + "step": 34855 + }, + { + "epoch": 2.41870793144126, + "grad_norm": 1.2463780641555786, + "learning_rate": 7.136393651306499e-05, + "loss": 0.465, + "step": 34856 + }, + { + "epoch": 2.418777322878357, + "grad_norm": 1.6865516901016235, + "learning_rate": 7.13585532397054e-05, + "loss": 0.5991, + "step": 34857 + }, + { + "epoch": 2.4188467143154533, + "grad_norm": 0.9877678155899048, + "learning_rate": 7.135317005676352e-05, + "loss": 0.3785, + "step": 34858 + }, + { + "epoch": 2.41891610575255, + "grad_norm": 1.1672354936599731, + "learning_rate": 7.134778696425641e-05, + "loss": 0.3412, + "step": 34859 + }, + { + "epoch": 2.418985497189647, + "grad_norm": 1.3308004140853882, + "learning_rate": 7.134240396220095e-05, + "loss": 0.446, + "step": 34860 + }, + { + "epoch": 2.4190548886267433, + "grad_norm": 1.6351383924484253, + "learning_rate": 7.133702105061421e-05, + "loss": 0.4204, + "step": 34861 + }, + { + "epoch": 2.41912428006384, + "grad_norm": 1.1117327213287354, + "learning_rate": 7.133163822951317e-05, + "loss": 0.4134, + "step": 34862 + }, + { + "epoch": 2.4191936715009366, + "grad_norm": 1.330324411392212, + "learning_rate": 7.132625549891483e-05, + "loss": 0.4447, + "step": 34863 + }, + { + "epoch": 2.4192630629380334, + "grad_norm": 0.7858085632324219, + "learning_rate": 7.132087285883618e-05, + "loss": 0.2216, + "step": 34864 + }, + { + "epoch": 2.4193324543751302, + "grad_norm": 1.463798999786377, + "learning_rate": 7.13154903092942e-05, + "loss": 0.4355, + "step": 34865 + }, + { + "epoch": 2.4194018458122266, + "grad_norm": 1.5022516250610352, + "learning_rate": 7.13101078503059e-05, + "loss": 0.4141, + "step": 34866 + }, + { + "epoch": 2.4194712372493234, + "grad_norm": 3.1688177585601807, + "learning_rate": 7.130472548188825e-05, + "loss": 0.5454, + "step": 34867 + }, + { + "epoch": 2.4195406286864203, + "grad_norm": 1.0763294696807861, + "learning_rate": 7.129934320405825e-05, + "loss": 0.3406, + "step": 34868 + }, + { + "epoch": 2.4196100201235167, + "grad_norm": 1.62697434425354, + "learning_rate": 7.12939610168329e-05, + "loss": 0.6347, + "step": 34869 + }, + { + "epoch": 2.4196794115606135, + "grad_norm": 1.0854920148849487, + "learning_rate": 7.128857892022918e-05, + "loss": 0.2804, + "step": 34870 + }, + { + "epoch": 2.41974880299771, + "grad_norm": 1.3818647861480713, + "learning_rate": 7.128319691426406e-05, + "loss": 0.3521, + "step": 34871 + }, + { + "epoch": 2.4198181944348067, + "grad_norm": 1.8384827375411987, + "learning_rate": 7.127781499895458e-05, + "loss": 0.4631, + "step": 34872 + }, + { + "epoch": 2.4198875858719036, + "grad_norm": 1.3905044794082642, + "learning_rate": 7.127243317431774e-05, + "loss": 0.3848, + "step": 34873 + }, + { + "epoch": 2.419956977309, + "grad_norm": 1.1589633226394653, + "learning_rate": 7.126705144037046e-05, + "loss": 0.3451, + "step": 34874 + }, + { + "epoch": 2.420026368746097, + "grad_norm": 0.8289499878883362, + "learning_rate": 7.126166979712975e-05, + "loss": 0.2651, + "step": 34875 + }, + { + "epoch": 2.4200957601831936, + "grad_norm": 2.091662883758545, + "learning_rate": 7.125628824461264e-05, + "loss": 0.3607, + "step": 34876 + }, + { + "epoch": 2.42016515162029, + "grad_norm": 0.919278621673584, + "learning_rate": 7.125090678283609e-05, + "loss": 0.2947, + "step": 34877 + }, + { + "epoch": 2.420234543057387, + "grad_norm": 1.241646647453308, + "learning_rate": 7.124552541181705e-05, + "loss": 0.4896, + "step": 34878 + }, + { + "epoch": 2.4203039344944832, + "grad_norm": 1.2440476417541504, + "learning_rate": 7.124014413157258e-05, + "loss": 0.3277, + "step": 34879 + }, + { + "epoch": 2.42037332593158, + "grad_norm": 1.2376829385757446, + "learning_rate": 7.12347629421197e-05, + "loss": 0.3627, + "step": 34880 + }, + { + "epoch": 2.420442717368677, + "grad_norm": 1.3193538188934326, + "learning_rate": 7.122938184347526e-05, + "loss": 0.5185, + "step": 34881 + }, + { + "epoch": 2.4205121088057733, + "grad_norm": 1.1001472473144531, + "learning_rate": 7.122400083565634e-05, + "loss": 0.3417, + "step": 34882 + }, + { + "epoch": 2.42058150024287, + "grad_norm": 2.381995677947998, + "learning_rate": 7.121861991867993e-05, + "loss": 0.5289, + "step": 34883 + }, + { + "epoch": 2.4206508916799665, + "grad_norm": 1.323562741279602, + "learning_rate": 7.121323909256298e-05, + "loss": 0.4699, + "step": 34884 + }, + { + "epoch": 2.4207202831170633, + "grad_norm": 1.3948698043823242, + "learning_rate": 7.120785835732249e-05, + "loss": 0.3142, + "step": 34885 + }, + { + "epoch": 2.42078967455416, + "grad_norm": 1.2610241174697876, + "learning_rate": 7.120247771297545e-05, + "loss": 0.3667, + "step": 34886 + }, + { + "epoch": 2.4208590659912566, + "grad_norm": 1.34848952293396, + "learning_rate": 7.119709715953888e-05, + "loss": 0.3646, + "step": 34887 + }, + { + "epoch": 2.4209284574283534, + "grad_norm": 1.2579717636108398, + "learning_rate": 7.11917166970297e-05, + "loss": 0.4163, + "step": 34888 + }, + { + "epoch": 2.4209978488654498, + "grad_norm": 2.046433925628662, + "learning_rate": 7.118633632546493e-05, + "loss": 0.5672, + "step": 34889 + }, + { + "epoch": 2.4210672403025466, + "grad_norm": 1.3298221826553345, + "learning_rate": 7.118095604486154e-05, + "loss": 0.4295, + "step": 34890 + }, + { + "epoch": 2.4211366317396434, + "grad_norm": 1.0304701328277588, + "learning_rate": 7.117557585523655e-05, + "loss": 0.2394, + "step": 34891 + }, + { + "epoch": 2.42120602317674, + "grad_norm": 1.5629572868347168, + "learning_rate": 7.117019575660691e-05, + "loss": 0.4876, + "step": 34892 + }, + { + "epoch": 2.4212754146138367, + "grad_norm": 2.059903144836426, + "learning_rate": 7.116481574898962e-05, + "loss": 0.5349, + "step": 34893 + }, + { + "epoch": 2.4213448060509335, + "grad_norm": 1.3654611110687256, + "learning_rate": 7.11594358324017e-05, + "loss": 0.5424, + "step": 34894 + }, + { + "epoch": 2.42141419748803, + "grad_norm": 14.39626693725586, + "learning_rate": 7.115405600686006e-05, + "loss": 0.4003, + "step": 34895 + }, + { + "epoch": 2.4214835889251267, + "grad_norm": 1.616426944732666, + "learning_rate": 7.114867627238171e-05, + "loss": 0.45, + "step": 34896 + }, + { + "epoch": 2.4215529803622236, + "grad_norm": 0.9092296957969666, + "learning_rate": 7.114329662898364e-05, + "loss": 0.3208, + "step": 34897 + }, + { + "epoch": 2.42162237179932, + "grad_norm": 0.943810224533081, + "learning_rate": 7.113791707668284e-05, + "loss": 0.3516, + "step": 34898 + }, + { + "epoch": 2.4216917632364168, + "grad_norm": 1.410543441772461, + "learning_rate": 7.113253761549632e-05, + "loss": 0.3953, + "step": 34899 + }, + { + "epoch": 2.421761154673513, + "grad_norm": 1.3483940362930298, + "learning_rate": 7.1127158245441e-05, + "loss": 0.4743, + "step": 34900 + }, + { + "epoch": 2.42183054611061, + "grad_norm": 1.3265862464904785, + "learning_rate": 7.112177896653391e-05, + "loss": 0.4554, + "step": 34901 + }, + { + "epoch": 2.421899937547707, + "grad_norm": 1.2035064697265625, + "learning_rate": 7.111639977879198e-05, + "loss": 0.5211, + "step": 34902 + }, + { + "epoch": 2.421969328984803, + "grad_norm": 0.7496364712715149, + "learning_rate": 7.111102068223226e-05, + "loss": 0.2782, + "step": 34903 + }, + { + "epoch": 2.4220387204219, + "grad_norm": 0.9993057250976562, + "learning_rate": 7.110564167687168e-05, + "loss": 0.3576, + "step": 34904 + }, + { + "epoch": 2.4221081118589964, + "grad_norm": 0.920879065990448, + "learning_rate": 7.110026276272722e-05, + "loss": 0.3525, + "step": 34905 + }, + { + "epoch": 2.4221775032960933, + "grad_norm": 1.7244412899017334, + "learning_rate": 7.10948839398159e-05, + "loss": 0.4542, + "step": 34906 + }, + { + "epoch": 2.42224689473319, + "grad_norm": 1.1367127895355225, + "learning_rate": 7.108950520815464e-05, + "loss": 0.5862, + "step": 34907 + }, + { + "epoch": 2.4223162861702865, + "grad_norm": 1.2087198495864868, + "learning_rate": 7.108412656776053e-05, + "loss": 0.4489, + "step": 34908 + }, + { + "epoch": 2.4223856776073833, + "grad_norm": 1.2660260200500488, + "learning_rate": 7.107874801865044e-05, + "loss": 0.4997, + "step": 34909 + }, + { + "epoch": 2.4224550690444797, + "grad_norm": 1.544786810874939, + "learning_rate": 7.107336956084138e-05, + "loss": 0.4955, + "step": 34910 + }, + { + "epoch": 2.4225244604815765, + "grad_norm": 1.8644312620162964, + "learning_rate": 7.106799119435033e-05, + "loss": 0.4756, + "step": 34911 + }, + { + "epoch": 2.4225938519186734, + "grad_norm": 1.2993450164794922, + "learning_rate": 7.106261291919427e-05, + "loss": 0.5605, + "step": 34912 + }, + { + "epoch": 2.4226632433557698, + "grad_norm": 1.5576645135879517, + "learning_rate": 7.105723473539019e-05, + "loss": 0.4941, + "step": 34913 + }, + { + "epoch": 2.4227326347928666, + "grad_norm": 1.8469315767288208, + "learning_rate": 7.105185664295507e-05, + "loss": 0.4769, + "step": 34914 + }, + { + "epoch": 2.422802026229963, + "grad_norm": 1.0431774854660034, + "learning_rate": 7.104647864190592e-05, + "loss": 0.3712, + "step": 34915 + }, + { + "epoch": 2.42287141766706, + "grad_norm": 1.419996976852417, + "learning_rate": 7.104110073225962e-05, + "loss": 0.3587, + "step": 34916 + }, + { + "epoch": 2.4229408091041567, + "grad_norm": 1.5754578113555908, + "learning_rate": 7.103572291403321e-05, + "loss": 0.5216, + "step": 34917 + }, + { + "epoch": 2.423010200541253, + "grad_norm": 1.151258945465088, + "learning_rate": 7.10303451872437e-05, + "loss": 0.3979, + "step": 34918 + }, + { + "epoch": 2.42307959197835, + "grad_norm": 1.7181963920593262, + "learning_rate": 7.1024967551908e-05, + "loss": 0.4172, + "step": 34919 + }, + { + "epoch": 2.4231489834154467, + "grad_norm": 1.5670033693313599, + "learning_rate": 7.101959000804312e-05, + "loss": 0.3813, + "step": 34920 + }, + { + "epoch": 2.423218374852543, + "grad_norm": 1.3802011013031006, + "learning_rate": 7.101421255566602e-05, + "loss": 0.2827, + "step": 34921 + }, + { + "epoch": 2.42328776628964, + "grad_norm": 3.5858519077301025, + "learning_rate": 7.100883519479373e-05, + "loss": 0.5966, + "step": 34922 + }, + { + "epoch": 2.4233571577267368, + "grad_norm": 1.5781902074813843, + "learning_rate": 7.100345792544315e-05, + "loss": 0.4575, + "step": 34923 + }, + { + "epoch": 2.423426549163833, + "grad_norm": 1.353310227394104, + "learning_rate": 7.099808074763128e-05, + "loss": 0.4507, + "step": 34924 + }, + { + "epoch": 2.42349594060093, + "grad_norm": 1.1629754304885864, + "learning_rate": 7.099270366137511e-05, + "loss": 0.4019, + "step": 34925 + }, + { + "epoch": 2.4235653320380264, + "grad_norm": 1.5544074773788452, + "learning_rate": 7.098732666669161e-05, + "loss": 0.6103, + "step": 34926 + }, + { + "epoch": 2.423634723475123, + "grad_norm": 1.494125247001648, + "learning_rate": 7.098194976359776e-05, + "loss": 0.3299, + "step": 34927 + }, + { + "epoch": 2.42370411491222, + "grad_norm": 2.086763858795166, + "learning_rate": 7.097657295211053e-05, + "loss": 0.498, + "step": 34928 + }, + { + "epoch": 2.4237735063493164, + "grad_norm": 1.830804705619812, + "learning_rate": 7.097119623224692e-05, + "loss": 0.5102, + "step": 34929 + }, + { + "epoch": 2.4238428977864133, + "grad_norm": 1.463559627532959, + "learning_rate": 7.096581960402383e-05, + "loss": 0.3326, + "step": 34930 + }, + { + "epoch": 2.4239122892235097, + "grad_norm": 1.7464956045150757, + "learning_rate": 7.096044306745829e-05, + "loss": 0.5236, + "step": 34931 + }, + { + "epoch": 2.4239816806606065, + "grad_norm": 1.2618762254714966, + "learning_rate": 7.095506662256726e-05, + "loss": 0.2776, + "step": 34932 + }, + { + "epoch": 2.4240510720977033, + "grad_norm": 1.3237621784210205, + "learning_rate": 7.094969026936771e-05, + "loss": 0.3973, + "step": 34933 + }, + { + "epoch": 2.4241204635347997, + "grad_norm": 2.2769296169281006, + "learning_rate": 7.094431400787664e-05, + "loss": 0.6316, + "step": 34934 + }, + { + "epoch": 2.4241898549718965, + "grad_norm": 1.2533575296401978, + "learning_rate": 7.093893783811101e-05, + "loss": 0.3379, + "step": 34935 + }, + { + "epoch": 2.424259246408993, + "grad_norm": 1.2510454654693604, + "learning_rate": 7.093356176008775e-05, + "loss": 0.4083, + "step": 34936 + }, + { + "epoch": 2.4243286378460898, + "grad_norm": 1.5088434219360352, + "learning_rate": 7.092818577382388e-05, + "loss": 0.4676, + "step": 34937 + }, + { + "epoch": 2.4243980292831866, + "grad_norm": 1.4656784534454346, + "learning_rate": 7.092280987933635e-05, + "loss": 0.3659, + "step": 34938 + }, + { + "epoch": 2.424467420720283, + "grad_norm": 1.1610686779022217, + "learning_rate": 7.091743407664213e-05, + "loss": 0.4028, + "step": 34939 + }, + { + "epoch": 2.42453681215738, + "grad_norm": 1.3845582008361816, + "learning_rate": 7.091205836575818e-05, + "loss": 0.3565, + "step": 34940 + }, + { + "epoch": 2.424606203594476, + "grad_norm": 1.2889389991760254, + "learning_rate": 7.090668274670152e-05, + "loss": 0.4184, + "step": 34941 + }, + { + "epoch": 2.424675595031573, + "grad_norm": 1.2161635160446167, + "learning_rate": 7.090130721948908e-05, + "loss": 0.3053, + "step": 34942 + }, + { + "epoch": 2.42474498646867, + "grad_norm": 1.0333291292190552, + "learning_rate": 7.089593178413784e-05, + "loss": 0.2498, + "step": 34943 + }, + { + "epoch": 2.4248143779057663, + "grad_norm": 1.270764946937561, + "learning_rate": 7.089055644066475e-05, + "loss": 0.3723, + "step": 34944 + }, + { + "epoch": 2.424883769342863, + "grad_norm": 1.5795230865478516, + "learning_rate": 7.08851811890868e-05, + "loss": 0.5112, + "step": 34945 + }, + { + "epoch": 2.42495316077996, + "grad_norm": 1.1859389543533325, + "learning_rate": 7.087980602942095e-05, + "loss": 0.3178, + "step": 34946 + }, + { + "epoch": 2.4250225522170563, + "grad_norm": 2.248095989227295, + "learning_rate": 7.087443096168417e-05, + "loss": 0.5465, + "step": 34947 + }, + { + "epoch": 2.425091943654153, + "grad_norm": 1.2911020517349243, + "learning_rate": 7.086905598589344e-05, + "loss": 0.4296, + "step": 34948 + }, + { + "epoch": 2.42516133509125, + "grad_norm": 1.0954275131225586, + "learning_rate": 7.086368110206575e-05, + "loss": 0.3664, + "step": 34949 + }, + { + "epoch": 2.4252307265283464, + "grad_norm": 2.2087247371673584, + "learning_rate": 7.085830631021801e-05, + "loss": 0.5798, + "step": 34950 + }, + { + "epoch": 2.425300117965443, + "grad_norm": 1.9936074018478394, + "learning_rate": 7.085293161036721e-05, + "loss": 0.4656, + "step": 34951 + }, + { + "epoch": 2.4253695094025396, + "grad_norm": 1.80522620677948, + "learning_rate": 7.084755700253033e-05, + "loss": 0.4867, + "step": 34952 + }, + { + "epoch": 2.4254389008396364, + "grad_norm": 1.0113966464996338, + "learning_rate": 7.084218248672434e-05, + "loss": 0.37, + "step": 34953 + }, + { + "epoch": 2.4255082922767333, + "grad_norm": 1.246733546257019, + "learning_rate": 7.083680806296619e-05, + "loss": 0.4099, + "step": 34954 + }, + { + "epoch": 2.4255776837138296, + "grad_norm": 1.2168278694152832, + "learning_rate": 7.083143373127285e-05, + "loss": 0.3182, + "step": 34955 + }, + { + "epoch": 2.4256470751509265, + "grad_norm": 1.0477094650268555, + "learning_rate": 7.08260594916613e-05, + "loss": 0.3349, + "step": 34956 + }, + { + "epoch": 2.425716466588023, + "grad_norm": 1.355176329612732, + "learning_rate": 7.08206853441485e-05, + "loss": 0.3415, + "step": 34957 + }, + { + "epoch": 2.4257858580251197, + "grad_norm": 1.1628745794296265, + "learning_rate": 7.08153112887514e-05, + "loss": 0.3923, + "step": 34958 + }, + { + "epoch": 2.4258552494622165, + "grad_norm": 1.424133062362671, + "learning_rate": 7.080993732548697e-05, + "loss": 0.4736, + "step": 34959 + }, + { + "epoch": 2.425924640899313, + "grad_norm": 1.280022382736206, + "learning_rate": 7.080456345437216e-05, + "loss": 0.4005, + "step": 34960 + }, + { + "epoch": 2.4259940323364098, + "grad_norm": 1.1615872383117676, + "learning_rate": 7.079918967542398e-05, + "loss": 0.4147, + "step": 34961 + }, + { + "epoch": 2.426063423773506, + "grad_norm": 1.1748791933059692, + "learning_rate": 7.079381598865937e-05, + "loss": 0.4557, + "step": 34962 + }, + { + "epoch": 2.426132815210603, + "grad_norm": 1.6613878011703491, + "learning_rate": 7.078844239409533e-05, + "loss": 0.4184, + "step": 34963 + }, + { + "epoch": 2.4262022066477, + "grad_norm": 1.3600795269012451, + "learning_rate": 7.078306889174875e-05, + "loss": 0.526, + "step": 34964 + }, + { + "epoch": 2.426271598084796, + "grad_norm": 1.1332582235336304, + "learning_rate": 7.077769548163663e-05, + "loss": 0.2956, + "step": 34965 + }, + { + "epoch": 2.426340989521893, + "grad_norm": 1.121131181716919, + "learning_rate": 7.077232216377594e-05, + "loss": 0.2719, + "step": 34966 + }, + { + "epoch": 2.4264103809589894, + "grad_norm": 0.9187420010566711, + "learning_rate": 7.076694893818363e-05, + "loss": 0.325, + "step": 34967 + }, + { + "epoch": 2.4264797723960863, + "grad_norm": 1.2545204162597656, + "learning_rate": 7.076157580487666e-05, + "loss": 0.3894, + "step": 34968 + }, + { + "epoch": 2.426549163833183, + "grad_norm": 1.2193177938461304, + "learning_rate": 7.075620276387204e-05, + "loss": 0.4848, + "step": 34969 + }, + { + "epoch": 2.4266185552702795, + "grad_norm": 1.092109203338623, + "learning_rate": 7.07508298151867e-05, + "loss": 0.3908, + "step": 34970 + }, + { + "epoch": 2.4266879467073763, + "grad_norm": 2.19419527053833, + "learning_rate": 7.074545695883756e-05, + "loss": 0.6905, + "step": 34971 + }, + { + "epoch": 2.426757338144473, + "grad_norm": 1.3780601024627686, + "learning_rate": 7.074008419484165e-05, + "loss": 0.3722, + "step": 34972 + }, + { + "epoch": 2.4268267295815695, + "grad_norm": 1.1372547149658203, + "learning_rate": 7.073471152321587e-05, + "loss": 0.4055, + "step": 34973 + }, + { + "epoch": 2.4268961210186664, + "grad_norm": 1.5981112718582153, + "learning_rate": 7.072933894397722e-05, + "loss": 0.4167, + "step": 34974 + }, + { + "epoch": 2.426965512455763, + "grad_norm": 0.9939419031143188, + "learning_rate": 7.072396645714265e-05, + "loss": 0.3796, + "step": 34975 + }, + { + "epoch": 2.4270349038928596, + "grad_norm": 1.4727836847305298, + "learning_rate": 7.071859406272911e-05, + "loss": 0.4097, + "step": 34976 + }, + { + "epoch": 2.4271042953299564, + "grad_norm": 1.8183144330978394, + "learning_rate": 7.071322176075364e-05, + "loss": 0.7474, + "step": 34977 + }, + { + "epoch": 2.427173686767053, + "grad_norm": 1.8028011322021484, + "learning_rate": 7.070784955123306e-05, + "loss": 0.4352, + "step": 34978 + }, + { + "epoch": 2.4272430782041496, + "grad_norm": 1.8921480178833008, + "learning_rate": 7.07024774341844e-05, + "loss": 0.4071, + "step": 34979 + }, + { + "epoch": 2.4273124696412465, + "grad_norm": 1.3319159746170044, + "learning_rate": 7.069710540962464e-05, + "loss": 0.3455, + "step": 34980 + }, + { + "epoch": 2.427381861078343, + "grad_norm": 1.336190104484558, + "learning_rate": 7.06917334775707e-05, + "loss": 0.3926, + "step": 34981 + }, + { + "epoch": 2.4274512525154397, + "grad_norm": 1.0888772010803223, + "learning_rate": 7.068636163803956e-05, + "loss": 0.3478, + "step": 34982 + }, + { + "epoch": 2.427520643952536, + "grad_norm": 1.3351950645446777, + "learning_rate": 7.06809898910482e-05, + "loss": 0.3625, + "step": 34983 + }, + { + "epoch": 2.427590035389633, + "grad_norm": 0.9656636118888855, + "learning_rate": 7.067561823661357e-05, + "loss": 0.3612, + "step": 34984 + }, + { + "epoch": 2.4276594268267297, + "grad_norm": 1.2146310806274414, + "learning_rate": 7.067024667475258e-05, + "loss": 0.3975, + "step": 34985 + }, + { + "epoch": 2.427728818263826, + "grad_norm": 1.661264419555664, + "learning_rate": 7.066487520548221e-05, + "loss": 0.4232, + "step": 34986 + }, + { + "epoch": 2.427798209700923, + "grad_norm": 2.037261486053467, + "learning_rate": 7.065950382881943e-05, + "loss": 0.423, + "step": 34987 + }, + { + "epoch": 2.4278676011380194, + "grad_norm": 1.1941255331039429, + "learning_rate": 7.06541325447812e-05, + "loss": 0.3669, + "step": 34988 + }, + { + "epoch": 2.427936992575116, + "grad_norm": 1.3601559400558472, + "learning_rate": 7.064876135338446e-05, + "loss": 0.4303, + "step": 34989 + }, + { + "epoch": 2.428006384012213, + "grad_norm": 1.6148382425308228, + "learning_rate": 7.064339025464618e-05, + "loss": 0.461, + "step": 34990 + }, + { + "epoch": 2.4280757754493094, + "grad_norm": 1.529587745666504, + "learning_rate": 7.063801924858333e-05, + "loss": 0.4752, + "step": 34991 + }, + { + "epoch": 2.4281451668864062, + "grad_norm": 1.3314920663833618, + "learning_rate": 7.063264833521284e-05, + "loss": 0.2987, + "step": 34992 + }, + { + "epoch": 2.4282145583235026, + "grad_norm": 1.1016747951507568, + "learning_rate": 7.062727751455165e-05, + "loss": 0.4582, + "step": 34993 + }, + { + "epoch": 2.4282839497605995, + "grad_norm": 1.396820306777954, + "learning_rate": 7.062190678661676e-05, + "loss": 0.3694, + "step": 34994 + }, + { + "epoch": 2.4283533411976963, + "grad_norm": 4.736161231994629, + "learning_rate": 7.061653615142507e-05, + "loss": 0.5453, + "step": 34995 + }, + { + "epoch": 2.4284227326347927, + "grad_norm": 1.4278103113174438, + "learning_rate": 7.061116560899358e-05, + "loss": 0.3565, + "step": 34996 + }, + { + "epoch": 2.4284921240718895, + "grad_norm": 1.2614045143127441, + "learning_rate": 7.060579515933923e-05, + "loss": 0.4887, + "step": 34997 + }, + { + "epoch": 2.4285615155089864, + "grad_norm": 1.6280900239944458, + "learning_rate": 7.060042480247901e-05, + "loss": 0.5678, + "step": 34998 + }, + { + "epoch": 2.4286309069460827, + "grad_norm": 2.0396993160247803, + "learning_rate": 7.05950545384298e-05, + "loss": 0.7373, + "step": 34999 + }, + { + "epoch": 2.4287002983831796, + "grad_norm": 1.2677923440933228, + "learning_rate": 7.058968436720861e-05, + "loss": 0.5224, + "step": 35000 + }, + { + "epoch": 2.4287696898202764, + "grad_norm": 1.8340860605239868, + "learning_rate": 7.058431428883235e-05, + "loss": 0.3823, + "step": 35001 + }, + { + "epoch": 2.428839081257373, + "grad_norm": 1.3674529790878296, + "learning_rate": 7.0578944303318e-05, + "loss": 0.3688, + "step": 35002 + }, + { + "epoch": 2.4289084726944696, + "grad_norm": 1.2259806394577026, + "learning_rate": 7.05735744106825e-05, + "loss": 0.3407, + "step": 35003 + }, + { + "epoch": 2.428977864131566, + "grad_norm": 1.5380668640136719, + "learning_rate": 7.056820461094287e-05, + "loss": 0.493, + "step": 35004 + }, + { + "epoch": 2.429047255568663, + "grad_norm": 1.4133318662643433, + "learning_rate": 7.056283490411597e-05, + "loss": 0.4444, + "step": 35005 + }, + { + "epoch": 2.4291166470057597, + "grad_norm": 1.1909643411636353, + "learning_rate": 7.055746529021878e-05, + "loss": 0.339, + "step": 35006 + }, + { + "epoch": 2.429186038442856, + "grad_norm": 1.3160183429718018, + "learning_rate": 7.055209576926827e-05, + "loss": 0.5172, + "step": 35007 + }, + { + "epoch": 2.429255429879953, + "grad_norm": 1.4661120176315308, + "learning_rate": 7.054672634128135e-05, + "loss": 0.4471, + "step": 35008 + }, + { + "epoch": 2.4293248213170493, + "grad_norm": 1.4568912982940674, + "learning_rate": 7.0541357006275e-05, + "loss": 0.406, + "step": 35009 + }, + { + "epoch": 2.429394212754146, + "grad_norm": 1.236809492111206, + "learning_rate": 7.053598776426617e-05, + "loss": 0.4547, + "step": 35010 + }, + { + "epoch": 2.429463604191243, + "grad_norm": 1.636451005935669, + "learning_rate": 7.053061861527186e-05, + "loss": 0.3658, + "step": 35011 + }, + { + "epoch": 2.4295329956283394, + "grad_norm": 1.3739053010940552, + "learning_rate": 7.052524955930892e-05, + "loss": 0.5376, + "step": 35012 + }, + { + "epoch": 2.429602387065436, + "grad_norm": 1.3243640661239624, + "learning_rate": 7.051988059639435e-05, + "loss": 0.4091, + "step": 35013 + }, + { + "epoch": 2.4296717785025326, + "grad_norm": 1.0762368440628052, + "learning_rate": 7.051451172654509e-05, + "loss": 0.4246, + "step": 35014 + }, + { + "epoch": 2.4297411699396294, + "grad_norm": 1.7474606037139893, + "learning_rate": 7.050914294977811e-05, + "loss": 0.5337, + "step": 35015 + }, + { + "epoch": 2.4298105613767262, + "grad_norm": 1.2442960739135742, + "learning_rate": 7.050377426611032e-05, + "loss": 0.5511, + "step": 35016 + }, + { + "epoch": 2.4298799528138226, + "grad_norm": 1.4768279790878296, + "learning_rate": 7.04984056755587e-05, + "loss": 0.5527, + "step": 35017 + }, + { + "epoch": 2.4299493442509195, + "grad_norm": 1.1193512678146362, + "learning_rate": 7.049303717814024e-05, + "loss": 0.3278, + "step": 35018 + }, + { + "epoch": 2.430018735688016, + "grad_norm": 1.6211193799972534, + "learning_rate": 7.048766877387179e-05, + "loss": 0.5237, + "step": 35019 + }, + { + "epoch": 2.4300881271251127, + "grad_norm": 0.9574204087257385, + "learning_rate": 7.048230046277035e-05, + "loss": 0.288, + "step": 35020 + }, + { + "epoch": 2.4301575185622095, + "grad_norm": 1.1991664171218872, + "learning_rate": 7.047693224485287e-05, + "loss": 0.4139, + "step": 35021 + }, + { + "epoch": 2.430226909999306, + "grad_norm": 1.9284636974334717, + "learning_rate": 7.047156412013627e-05, + "loss": 0.656, + "step": 35022 + }, + { + "epoch": 2.4302963014364027, + "grad_norm": 1.0020872354507446, + "learning_rate": 7.046619608863755e-05, + "loss": 0.2689, + "step": 35023 + }, + { + "epoch": 2.4303656928734996, + "grad_norm": 1.4029603004455566, + "learning_rate": 7.046082815037361e-05, + "loss": 0.4261, + "step": 35024 + }, + { + "epoch": 2.430435084310596, + "grad_norm": 1.4708904027938843, + "learning_rate": 7.045546030536142e-05, + "loss": 0.6658, + "step": 35025 + }, + { + "epoch": 2.430504475747693, + "grad_norm": 1.7188054323196411, + "learning_rate": 7.045009255361792e-05, + "loss": 0.361, + "step": 35026 + }, + { + "epoch": 2.4305738671847896, + "grad_norm": 1.3184123039245605, + "learning_rate": 7.044472489516e-05, + "loss": 0.5539, + "step": 35027 + }, + { + "epoch": 2.430643258621886, + "grad_norm": 1.25288987159729, + "learning_rate": 7.043935733000469e-05, + "loss": 0.6327, + "step": 35028 + }, + { + "epoch": 2.430712650058983, + "grad_norm": 1.3158175945281982, + "learning_rate": 7.04339898581689e-05, + "loss": 0.6562, + "step": 35029 + }, + { + "epoch": 2.4307820414960792, + "grad_norm": 1.2345222234725952, + "learning_rate": 7.042862247966955e-05, + "loss": 0.4068, + "step": 35030 + }, + { + "epoch": 2.430851432933176, + "grad_norm": 0.808944046497345, + "learning_rate": 7.042325519452362e-05, + "loss": 0.2735, + "step": 35031 + }, + { + "epoch": 2.430920824370273, + "grad_norm": 1.3305506706237793, + "learning_rate": 7.041788800274806e-05, + "loss": 0.5867, + "step": 35032 + }, + { + "epoch": 2.4309902158073693, + "grad_norm": 1.0364480018615723, + "learning_rate": 7.04125209043598e-05, + "loss": 0.3437, + "step": 35033 + }, + { + "epoch": 2.431059607244466, + "grad_norm": 0.8974506258964539, + "learning_rate": 7.040715389937574e-05, + "loss": 0.3238, + "step": 35034 + }, + { + "epoch": 2.4311289986815625, + "grad_norm": 0.9519392848014832, + "learning_rate": 7.040178698781288e-05, + "loss": 0.3015, + "step": 35035 + }, + { + "epoch": 2.4311983901186593, + "grad_norm": 1.1242151260375977, + "learning_rate": 7.039642016968813e-05, + "loss": 0.3294, + "step": 35036 + }, + { + "epoch": 2.431267781555756, + "grad_norm": 1.3803530931472778, + "learning_rate": 7.039105344501847e-05, + "loss": 0.4671, + "step": 35037 + }, + { + "epoch": 2.4313371729928526, + "grad_norm": 1.1241734027862549, + "learning_rate": 7.038568681382082e-05, + "loss": 0.3123, + "step": 35038 + }, + { + "epoch": 2.4314065644299494, + "grad_norm": 1.5313851833343506, + "learning_rate": 7.038032027611212e-05, + "loss": 0.5406, + "step": 35039 + }, + { + "epoch": 2.431475955867046, + "grad_norm": 2.0217156410217285, + "learning_rate": 7.03749538319093e-05, + "loss": 0.5338, + "step": 35040 + }, + { + "epoch": 2.4315453473041426, + "grad_norm": 1.10798978805542, + "learning_rate": 7.03695874812293e-05, + "loss": 0.4648, + "step": 35041 + }, + { + "epoch": 2.4316147387412395, + "grad_norm": 2.0896623134613037, + "learning_rate": 7.036422122408912e-05, + "loss": 0.6806, + "step": 35042 + }, + { + "epoch": 2.431684130178336, + "grad_norm": 1.345888614654541, + "learning_rate": 7.035885506050562e-05, + "loss": 0.4597, + "step": 35043 + }, + { + "epoch": 2.4317535216154327, + "grad_norm": 1.493878960609436, + "learning_rate": 7.035348899049578e-05, + "loss": 0.4904, + "step": 35044 + }, + { + "epoch": 2.4318229130525295, + "grad_norm": 1.231056809425354, + "learning_rate": 7.034812301407654e-05, + "loss": 0.4143, + "step": 35045 + }, + { + "epoch": 2.431892304489626, + "grad_norm": 0.9028039574623108, + "learning_rate": 7.034275713126485e-05, + "loss": 0.3541, + "step": 35046 + }, + { + "epoch": 2.4319616959267227, + "grad_norm": 1.1663622856140137, + "learning_rate": 7.033739134207761e-05, + "loss": 0.368, + "step": 35047 + }, + { + "epoch": 2.432031087363819, + "grad_norm": 1.045354962348938, + "learning_rate": 7.03320256465318e-05, + "loss": 0.3134, + "step": 35048 + }, + { + "epoch": 2.432100478800916, + "grad_norm": 1.1483420133590698, + "learning_rate": 7.032666004464434e-05, + "loss": 0.3582, + "step": 35049 + }, + { + "epoch": 2.432169870238013, + "grad_norm": 1.4759713411331177, + "learning_rate": 7.032129453643214e-05, + "loss": 0.5053, + "step": 35050 + }, + { + "epoch": 2.432239261675109, + "grad_norm": 1.4272804260253906, + "learning_rate": 7.03159291219122e-05, + "loss": 0.3519, + "step": 35051 + }, + { + "epoch": 2.432308653112206, + "grad_norm": 1.5131306648254395, + "learning_rate": 7.031056380110142e-05, + "loss": 0.3259, + "step": 35052 + }, + { + "epoch": 2.432378044549303, + "grad_norm": 1.1049458980560303, + "learning_rate": 7.030519857401678e-05, + "loss": 0.456, + "step": 35053 + }, + { + "epoch": 2.4324474359863992, + "grad_norm": 1.508609652519226, + "learning_rate": 7.029983344067516e-05, + "loss": 0.4157, + "step": 35054 + }, + { + "epoch": 2.432516827423496, + "grad_norm": 1.3937715291976929, + "learning_rate": 7.029446840109351e-05, + "loss": 0.2489, + "step": 35055 + }, + { + "epoch": 2.4325862188605925, + "grad_norm": 1.6019999980926514, + "learning_rate": 7.028910345528876e-05, + "loss": 0.4214, + "step": 35056 + }, + { + "epoch": 2.4326556102976893, + "grad_norm": 1.0112028121948242, + "learning_rate": 7.028373860327788e-05, + "loss": 0.3795, + "step": 35057 + }, + { + "epoch": 2.432725001734786, + "grad_norm": 1.2217142581939697, + "learning_rate": 7.027837384507781e-05, + "loss": 0.486, + "step": 35058 + }, + { + "epoch": 2.4327943931718825, + "grad_norm": 1.9933128356933594, + "learning_rate": 7.027300918070545e-05, + "loss": 0.5667, + "step": 35059 + }, + { + "epoch": 2.4328637846089793, + "grad_norm": 1.153735637664795, + "learning_rate": 7.026764461017776e-05, + "loss": 0.4674, + "step": 35060 + }, + { + "epoch": 2.4329331760460757, + "grad_norm": 1.1571248769760132, + "learning_rate": 7.026228013351167e-05, + "loss": 0.2688, + "step": 35061 + }, + { + "epoch": 2.4330025674831726, + "grad_norm": 1.1141037940979004, + "learning_rate": 7.02569157507241e-05, + "loss": 0.5522, + "step": 35062 + }, + { + "epoch": 2.4330719589202694, + "grad_norm": 1.1818186044692993, + "learning_rate": 7.025155146183197e-05, + "loss": 0.3385, + "step": 35063 + }, + { + "epoch": 2.433141350357366, + "grad_norm": 1.10015869140625, + "learning_rate": 7.024618726685227e-05, + "loss": 0.2061, + "step": 35064 + }, + { + "epoch": 2.4332107417944626, + "grad_norm": 1.6769037246704102, + "learning_rate": 7.02408231658019e-05, + "loss": 0.4245, + "step": 35065 + }, + { + "epoch": 2.433280133231559, + "grad_norm": 1.4389262199401855, + "learning_rate": 7.023545915869778e-05, + "loss": 0.482, + "step": 35066 + }, + { + "epoch": 2.433349524668656, + "grad_norm": 1.3506778478622437, + "learning_rate": 7.023009524555692e-05, + "loss": 0.4546, + "step": 35067 + }, + { + "epoch": 2.4334189161057527, + "grad_norm": 1.0266371965408325, + "learning_rate": 7.022473142639615e-05, + "loss": 0.2665, + "step": 35068 + }, + { + "epoch": 2.433488307542849, + "grad_norm": 1.2905231714248657, + "learning_rate": 7.021936770123246e-05, + "loss": 0.391, + "step": 35069 + }, + { + "epoch": 2.433557698979946, + "grad_norm": 1.642172932624817, + "learning_rate": 7.021400407008273e-05, + "loss": 0.3871, + "step": 35070 + }, + { + "epoch": 2.4336270904170427, + "grad_norm": 1.0394659042358398, + "learning_rate": 7.020864053296398e-05, + "loss": 0.4167, + "step": 35071 + }, + { + "epoch": 2.433696481854139, + "grad_norm": 0.9956875443458557, + "learning_rate": 7.020327708989307e-05, + "loss": 0.2289, + "step": 35072 + }, + { + "epoch": 2.433765873291236, + "grad_norm": 1.5569738149642944, + "learning_rate": 7.019791374088698e-05, + "loss": 0.3577, + "step": 35073 + }, + { + "epoch": 2.4338352647283323, + "grad_norm": 1.0846043825149536, + "learning_rate": 7.019255048596262e-05, + "loss": 0.364, + "step": 35074 + }, + { + "epoch": 2.433904656165429, + "grad_norm": 0.9061812162399292, + "learning_rate": 7.018718732513691e-05, + "loss": 0.2836, + "step": 35075 + }, + { + "epoch": 2.433974047602526, + "grad_norm": 2.222020149230957, + "learning_rate": 7.018182425842678e-05, + "loss": 0.3611, + "step": 35076 + }, + { + "epoch": 2.4340434390396224, + "grad_norm": 1.2660895586013794, + "learning_rate": 7.01764612858492e-05, + "loss": 0.5118, + "step": 35077 + }, + { + "epoch": 2.4341128304767192, + "grad_norm": 1.5377216339111328, + "learning_rate": 7.017109840742105e-05, + "loss": 0.3915, + "step": 35078 + }, + { + "epoch": 2.434182221913816, + "grad_norm": 1.6196645498275757, + "learning_rate": 7.016573562315927e-05, + "loss": 0.5979, + "step": 35079 + }, + { + "epoch": 2.4342516133509124, + "grad_norm": 1.4330227375030518, + "learning_rate": 7.016037293308081e-05, + "loss": 0.5696, + "step": 35080 + }, + { + "epoch": 2.4343210047880093, + "grad_norm": 1.3672300577163696, + "learning_rate": 7.015501033720263e-05, + "loss": 0.506, + "step": 35081 + }, + { + "epoch": 2.4343903962251057, + "grad_norm": 1.2255096435546875, + "learning_rate": 7.014964783554158e-05, + "loss": 0.3494, + "step": 35082 + }, + { + "epoch": 2.4344597876622025, + "grad_norm": 0.9933603405952454, + "learning_rate": 7.014428542811463e-05, + "loss": 0.4023, + "step": 35083 + }, + { + "epoch": 2.4345291790992993, + "grad_norm": 1.0532948970794678, + "learning_rate": 7.013892311493871e-05, + "loss": 0.5095, + "step": 35084 + }, + { + "epoch": 2.4345985705363957, + "grad_norm": 0.8046673536300659, + "learning_rate": 7.013356089603074e-05, + "loss": 0.349, + "step": 35085 + }, + { + "epoch": 2.4346679619734926, + "grad_norm": 1.2320175170898438, + "learning_rate": 7.012819877140764e-05, + "loss": 0.4004, + "step": 35086 + }, + { + "epoch": 2.434737353410589, + "grad_norm": 0.922226071357727, + "learning_rate": 7.012283674108642e-05, + "loss": 0.317, + "step": 35087 + }, + { + "epoch": 2.4348067448476858, + "grad_norm": 1.8533775806427002, + "learning_rate": 7.011747480508387e-05, + "loss": 0.5064, + "step": 35088 + }, + { + "epoch": 2.4348761362847826, + "grad_norm": 1.2072076797485352, + "learning_rate": 7.0112112963417e-05, + "loss": 0.3223, + "step": 35089 + }, + { + "epoch": 2.434945527721879, + "grad_norm": 1.1077314615249634, + "learning_rate": 7.010675121610272e-05, + "loss": 0.3518, + "step": 35090 + }, + { + "epoch": 2.435014919158976, + "grad_norm": 1.2328182458877563, + "learning_rate": 7.010138956315795e-05, + "loss": 0.5054, + "step": 35091 + }, + { + "epoch": 2.435084310596072, + "grad_norm": 1.806854486465454, + "learning_rate": 7.009602800459963e-05, + "loss": 0.5434, + "step": 35092 + }, + { + "epoch": 2.435153702033169, + "grad_norm": 1.4724551439285278, + "learning_rate": 7.00906665404447e-05, + "loss": 0.4991, + "step": 35093 + }, + { + "epoch": 2.435223093470266, + "grad_norm": 1.405463695526123, + "learning_rate": 7.008530517071008e-05, + "loss": 0.2957, + "step": 35094 + }, + { + "epoch": 2.4352924849073623, + "grad_norm": 1.3073647022247314, + "learning_rate": 7.007994389541264e-05, + "loss": 0.4358, + "step": 35095 + }, + { + "epoch": 2.435361876344459, + "grad_norm": 1.2580585479736328, + "learning_rate": 7.007458271456937e-05, + "loss": 0.348, + "step": 35096 + }, + { + "epoch": 2.435431267781556, + "grad_norm": 1.4897737503051758, + "learning_rate": 7.006922162819715e-05, + "loss": 0.4475, + "step": 35097 + }, + { + "epoch": 2.4355006592186523, + "grad_norm": 1.4110395908355713, + "learning_rate": 7.006386063631294e-05, + "loss": 0.4158, + "step": 35098 + }, + { + "epoch": 2.435570050655749, + "grad_norm": 1.3674812316894531, + "learning_rate": 7.005849973893364e-05, + "loss": 0.5176, + "step": 35099 + }, + { + "epoch": 2.4356394420928456, + "grad_norm": 1.2437318563461304, + "learning_rate": 7.005313893607617e-05, + "loss": 0.4701, + "step": 35100 + }, + { + "epoch": 2.4357088335299424, + "grad_norm": 0.8983495831489563, + "learning_rate": 7.004777822775752e-05, + "loss": 0.3165, + "step": 35101 + }, + { + "epoch": 2.435778224967039, + "grad_norm": 1.6470799446105957, + "learning_rate": 7.004241761399451e-05, + "loss": 0.3624, + "step": 35102 + }, + { + "epoch": 2.4358476164041356, + "grad_norm": 1.2516870498657227, + "learning_rate": 7.003705709480413e-05, + "loss": 0.538, + "step": 35103 + }, + { + "epoch": 2.4359170078412324, + "grad_norm": 1.4633369445800781, + "learning_rate": 7.003169667020326e-05, + "loss": 0.4397, + "step": 35104 + }, + { + "epoch": 2.4359863992783293, + "grad_norm": 1.303637146949768, + "learning_rate": 7.002633634020885e-05, + "loss": 0.3828, + "step": 35105 + }, + { + "epoch": 2.4360557907154257, + "grad_norm": 1.297562599182129, + "learning_rate": 7.002097610483784e-05, + "loss": 0.4597, + "step": 35106 + }, + { + "epoch": 2.4361251821525225, + "grad_norm": 0.9177536368370056, + "learning_rate": 7.001561596410711e-05, + "loss": 0.3154, + "step": 35107 + }, + { + "epoch": 2.436194573589619, + "grad_norm": 0.9301772117614746, + "learning_rate": 7.001025591803365e-05, + "loss": 0.3337, + "step": 35108 + }, + { + "epoch": 2.4362639650267157, + "grad_norm": 1.0674729347229004, + "learning_rate": 7.000489596663429e-05, + "loss": 0.3213, + "step": 35109 + }, + { + "epoch": 2.4363333564638125, + "grad_norm": 1.7959870100021362, + "learning_rate": 6.999953610992599e-05, + "loss": 0.4031, + "step": 35110 + }, + { + "epoch": 2.436402747900909, + "grad_norm": 1.8105748891830444, + "learning_rate": 6.999417634792568e-05, + "loss": 0.4908, + "step": 35111 + }, + { + "epoch": 2.4364721393380058, + "grad_norm": 1.6527318954467773, + "learning_rate": 6.998881668065029e-05, + "loss": 0.4706, + "step": 35112 + }, + { + "epoch": 2.436541530775102, + "grad_norm": 1.1607744693756104, + "learning_rate": 6.998345710811671e-05, + "loss": 0.3269, + "step": 35113 + }, + { + "epoch": 2.436610922212199, + "grad_norm": 1.5729283094406128, + "learning_rate": 6.997809763034187e-05, + "loss": 0.4062, + "step": 35114 + }, + { + "epoch": 2.436680313649296, + "grad_norm": 1.198124885559082, + "learning_rate": 6.997273824734273e-05, + "loss": 0.3735, + "step": 35115 + }, + { + "epoch": 2.436749705086392, + "grad_norm": 1.5244828462600708, + "learning_rate": 6.996737895913613e-05, + "loss": 0.4516, + "step": 35116 + }, + { + "epoch": 2.436819096523489, + "grad_norm": 1.464112639427185, + "learning_rate": 6.996201976573905e-05, + "loss": 0.4979, + "step": 35117 + }, + { + "epoch": 2.4368884879605854, + "grad_norm": 1.3279166221618652, + "learning_rate": 6.995666066716834e-05, + "loss": 0.5053, + "step": 35118 + }, + { + "epoch": 2.4369578793976823, + "grad_norm": 1.111183762550354, + "learning_rate": 6.995130166344102e-05, + "loss": 0.3128, + "step": 35119 + }, + { + "epoch": 2.437027270834779, + "grad_norm": 1.4738991260528564, + "learning_rate": 6.994594275457392e-05, + "loss": 0.4924, + "step": 35120 + }, + { + "epoch": 2.4370966622718755, + "grad_norm": 1.0267951488494873, + "learning_rate": 6.994058394058401e-05, + "loss": 0.3217, + "step": 35121 + }, + { + "epoch": 2.4371660537089723, + "grad_norm": 2.086968421936035, + "learning_rate": 6.993522522148822e-05, + "loss": 0.4784, + "step": 35122 + }, + { + "epoch": 2.437235445146069, + "grad_norm": 1.3474571704864502, + "learning_rate": 6.992986659730341e-05, + "loss": 0.4111, + "step": 35123 + }, + { + "epoch": 2.4373048365831655, + "grad_norm": 0.9667059183120728, + "learning_rate": 6.99245080680465e-05, + "loss": 0.4749, + "step": 35124 + }, + { + "epoch": 2.4373742280202624, + "grad_norm": 1.1988052129745483, + "learning_rate": 6.991914963373445e-05, + "loss": 0.4921, + "step": 35125 + }, + { + "epoch": 2.4374436194573588, + "grad_norm": 1.2550606727600098, + "learning_rate": 6.991379129438414e-05, + "loss": 0.5023, + "step": 35126 + }, + { + "epoch": 2.4375130108944556, + "grad_norm": 1.1987090110778809, + "learning_rate": 6.990843305001253e-05, + "loss": 0.2364, + "step": 35127 + }, + { + "epoch": 2.4375824023315524, + "grad_norm": 1.5143262147903442, + "learning_rate": 6.990307490063647e-05, + "loss": 0.4066, + "step": 35128 + }, + { + "epoch": 2.437651793768649, + "grad_norm": 1.0339860916137695, + "learning_rate": 6.989771684627295e-05, + "loss": 0.2617, + "step": 35129 + }, + { + "epoch": 2.4377211852057457, + "grad_norm": 1.1774628162384033, + "learning_rate": 6.989235888693882e-05, + "loss": 0.4088, + "step": 35130 + }, + { + "epoch": 2.4377905766428425, + "grad_norm": 1.1777809858322144, + "learning_rate": 6.988700102265102e-05, + "loss": 0.4061, + "step": 35131 + }, + { + "epoch": 2.437859968079939, + "grad_norm": 1.184924840927124, + "learning_rate": 6.988164325342647e-05, + "loss": 0.5957, + "step": 35132 + }, + { + "epoch": 2.4379293595170357, + "grad_norm": 1.0687165260314941, + "learning_rate": 6.987628557928206e-05, + "loss": 0.4003, + "step": 35133 + }, + { + "epoch": 2.437998750954132, + "grad_norm": 1.0489939451217651, + "learning_rate": 6.987092800023474e-05, + "loss": 0.3479, + "step": 35134 + }, + { + "epoch": 2.438068142391229, + "grad_norm": 1.3836387395858765, + "learning_rate": 6.986557051630138e-05, + "loss": 0.3969, + "step": 35135 + }, + { + "epoch": 2.4381375338283258, + "grad_norm": 1.1773099899291992, + "learning_rate": 6.986021312749898e-05, + "loss": 0.2925, + "step": 35136 + }, + { + "epoch": 2.438206925265422, + "grad_norm": 1.91454017162323, + "learning_rate": 6.985485583384434e-05, + "loss": 0.5399, + "step": 35137 + }, + { + "epoch": 2.438276316702519, + "grad_norm": 1.113368034362793, + "learning_rate": 6.984949863535443e-05, + "loss": 0.3678, + "step": 35138 + }, + { + "epoch": 2.4383457081396154, + "grad_norm": 1.2236143350601196, + "learning_rate": 6.984414153204615e-05, + "loss": 0.3643, + "step": 35139 + }, + { + "epoch": 2.438415099576712, + "grad_norm": 1.9470906257629395, + "learning_rate": 6.983878452393642e-05, + "loss": 0.5896, + "step": 35140 + }, + { + "epoch": 2.438484491013809, + "grad_norm": 1.2072917222976685, + "learning_rate": 6.983342761104214e-05, + "loss": 0.5576, + "step": 35141 + }, + { + "epoch": 2.4385538824509054, + "grad_norm": 1.5523101091384888, + "learning_rate": 6.982807079338023e-05, + "loss": 0.3334, + "step": 35142 + }, + { + "epoch": 2.4386232738880023, + "grad_norm": 1.354587197303772, + "learning_rate": 6.982271407096765e-05, + "loss": 0.5263, + "step": 35143 + }, + { + "epoch": 2.4386926653250987, + "grad_norm": 1.009569525718689, + "learning_rate": 6.98173574438212e-05, + "loss": 0.385, + "step": 35144 + }, + { + "epoch": 2.4387620567621955, + "grad_norm": 0.930461585521698, + "learning_rate": 6.981200091195788e-05, + "loss": 0.3041, + "step": 35145 + }, + { + "epoch": 2.4388314481992923, + "grad_norm": 1.6051058769226074, + "learning_rate": 6.980664447539455e-05, + "loss": 0.366, + "step": 35146 + }, + { + "epoch": 2.4389008396363887, + "grad_norm": 1.4327547550201416, + "learning_rate": 6.980128813414816e-05, + "loss": 0.3986, + "step": 35147 + }, + { + "epoch": 2.4389702310734855, + "grad_norm": 1.1952685117721558, + "learning_rate": 6.979593188823558e-05, + "loss": 0.3683, + "step": 35148 + }, + { + "epoch": 2.4390396225105824, + "grad_norm": 1.015637755393982, + "learning_rate": 6.979057573767376e-05, + "loss": 0.3137, + "step": 35149 + }, + { + "epoch": 2.4391090139476788, + "grad_norm": 1.325155258178711, + "learning_rate": 6.97852196824796e-05, + "loss": 0.3112, + "step": 35150 + }, + { + "epoch": 2.4391784053847756, + "grad_norm": 1.338898777961731, + "learning_rate": 6.977986372266998e-05, + "loss": 0.3703, + "step": 35151 + }, + { + "epoch": 2.4392477968218724, + "grad_norm": 1.6529067754745483, + "learning_rate": 6.97745078582618e-05, + "loss": 0.3769, + "step": 35152 + }, + { + "epoch": 2.439317188258969, + "grad_norm": 1.164028525352478, + "learning_rate": 6.976915208927199e-05, + "loss": 0.4133, + "step": 35153 + }, + { + "epoch": 2.4393865796960656, + "grad_norm": 1.3560909032821655, + "learning_rate": 6.976379641571749e-05, + "loss": 0.4547, + "step": 35154 + }, + { + "epoch": 2.439455971133162, + "grad_norm": 1.805287480354309, + "learning_rate": 6.975844083761516e-05, + "loss": 0.3482, + "step": 35155 + }, + { + "epoch": 2.439525362570259, + "grad_norm": 1.029524326324463, + "learning_rate": 6.975308535498192e-05, + "loss": 0.3139, + "step": 35156 + }, + { + "epoch": 2.4395947540073557, + "grad_norm": 1.2260174751281738, + "learning_rate": 6.974772996783472e-05, + "loss": 0.3926, + "step": 35157 + }, + { + "epoch": 2.439664145444452, + "grad_norm": 1.1513985395431519, + "learning_rate": 6.974237467619038e-05, + "loss": 0.2595, + "step": 35158 + }, + { + "epoch": 2.439733536881549, + "grad_norm": 1.5758625268936157, + "learning_rate": 6.973701948006586e-05, + "loss": 0.4247, + "step": 35159 + }, + { + "epoch": 2.4398029283186453, + "grad_norm": 1.5295871496200562, + "learning_rate": 6.973166437947807e-05, + "loss": 0.3718, + "step": 35160 + }, + { + "epoch": 2.439872319755742, + "grad_norm": 1.4641343355178833, + "learning_rate": 6.972630937444389e-05, + "loss": 0.5122, + "step": 35161 + }, + { + "epoch": 2.439941711192839, + "grad_norm": 1.1407309770584106, + "learning_rate": 6.972095446498025e-05, + "loss": 0.3389, + "step": 35162 + }, + { + "epoch": 2.4400111026299354, + "grad_norm": 1.2905405759811401, + "learning_rate": 6.971559965110406e-05, + "loss": 0.349, + "step": 35163 + }, + { + "epoch": 2.440080494067032, + "grad_norm": 1.3636387586593628, + "learning_rate": 6.97102449328322e-05, + "loss": 0.4915, + "step": 35164 + }, + { + "epoch": 2.4401498855041286, + "grad_norm": 1.4286396503448486, + "learning_rate": 6.970489031018157e-05, + "loss": 0.3997, + "step": 35165 + }, + { + "epoch": 2.4402192769412254, + "grad_norm": 1.232596516609192, + "learning_rate": 6.96995357831691e-05, + "loss": 0.517, + "step": 35166 + }, + { + "epoch": 2.4402886683783223, + "grad_norm": 1.4793609380722046, + "learning_rate": 6.969418135181167e-05, + "loss": 0.429, + "step": 35167 + }, + { + "epoch": 2.4403580598154186, + "grad_norm": 1.5815329551696777, + "learning_rate": 6.968882701612618e-05, + "loss": 0.2975, + "step": 35168 + }, + { + "epoch": 2.4404274512525155, + "grad_norm": 1.4007827043533325, + "learning_rate": 6.968347277612956e-05, + "loss": 0.4982, + "step": 35169 + }, + { + "epoch": 2.440496842689612, + "grad_norm": 1.2115697860717773, + "learning_rate": 6.967811863183873e-05, + "loss": 0.3826, + "step": 35170 + }, + { + "epoch": 2.4405662341267087, + "grad_norm": 0.8357123732566833, + "learning_rate": 6.967276458327052e-05, + "loss": 0.1902, + "step": 35171 + }, + { + "epoch": 2.4406356255638055, + "grad_norm": 1.4866325855255127, + "learning_rate": 6.966741063044189e-05, + "loss": 0.3604, + "step": 35172 + }, + { + "epoch": 2.440705017000902, + "grad_norm": 1.3331842422485352, + "learning_rate": 6.96620567733697e-05, + "loss": 0.515, + "step": 35173 + }, + { + "epoch": 2.4407744084379988, + "grad_norm": 1.5369371175765991, + "learning_rate": 6.96567030120709e-05, + "loss": 0.407, + "step": 35174 + }, + { + "epoch": 2.4408437998750956, + "grad_norm": 0.975073516368866, + "learning_rate": 6.965134934656234e-05, + "loss": 0.3522, + "step": 35175 + }, + { + "epoch": 2.440913191312192, + "grad_norm": 1.4885731935501099, + "learning_rate": 6.964599577686097e-05, + "loss": 0.4461, + "step": 35176 + }, + { + "epoch": 2.440982582749289, + "grad_norm": 1.5261796712875366, + "learning_rate": 6.964064230298371e-05, + "loss": 0.4166, + "step": 35177 + }, + { + "epoch": 2.4410519741863856, + "grad_norm": 1.104190468788147, + "learning_rate": 6.963528892494738e-05, + "loss": 0.3273, + "step": 35178 + }, + { + "epoch": 2.441121365623482, + "grad_norm": 1.0789053440093994, + "learning_rate": 6.962993564276892e-05, + "loss": 0.4448, + "step": 35179 + }, + { + "epoch": 2.441190757060579, + "grad_norm": 1.4179441928863525, + "learning_rate": 6.962458245646523e-05, + "loss": 0.4051, + "step": 35180 + }, + { + "epoch": 2.4412601484976753, + "grad_norm": 1.2685483694076538, + "learning_rate": 6.961922936605324e-05, + "loss": 0.488, + "step": 35181 + }, + { + "epoch": 2.441329539934772, + "grad_norm": 1.1983548402786255, + "learning_rate": 6.961387637154976e-05, + "loss": 0.299, + "step": 35182 + }, + { + "epoch": 2.441398931371869, + "grad_norm": 1.7182121276855469, + "learning_rate": 6.960852347297177e-05, + "loss": 0.4327, + "step": 35183 + }, + { + "epoch": 2.4414683228089653, + "grad_norm": 1.537102460861206, + "learning_rate": 6.960317067033617e-05, + "loss": 0.6212, + "step": 35184 + }, + { + "epoch": 2.441537714246062, + "grad_norm": 1.1833996772766113, + "learning_rate": 6.959781796365982e-05, + "loss": 0.3192, + "step": 35185 + }, + { + "epoch": 2.4416071056831585, + "grad_norm": 1.5325963497161865, + "learning_rate": 6.959246535295962e-05, + "loss": 0.3631, + "step": 35186 + }, + { + "epoch": 2.4416764971202554, + "grad_norm": 1.3408337831497192, + "learning_rate": 6.958711283825249e-05, + "loss": 0.4248, + "step": 35187 + }, + { + "epoch": 2.441745888557352, + "grad_norm": 1.3583749532699585, + "learning_rate": 6.958176041955531e-05, + "loss": 0.4284, + "step": 35188 + }, + { + "epoch": 2.4418152799944486, + "grad_norm": 1.4734904766082764, + "learning_rate": 6.957640809688497e-05, + "loss": 0.4278, + "step": 35189 + }, + { + "epoch": 2.4418846714315454, + "grad_norm": 1.4106346368789673, + "learning_rate": 6.957105587025839e-05, + "loss": 0.6017, + "step": 35190 + }, + { + "epoch": 2.441954062868642, + "grad_norm": 1.443432331085205, + "learning_rate": 6.95657037396925e-05, + "loss": 0.4957, + "step": 35191 + }, + { + "epoch": 2.4420234543057386, + "grad_norm": 1.325253963470459, + "learning_rate": 6.95603517052041e-05, + "loss": 0.3232, + "step": 35192 + }, + { + "epoch": 2.4420928457428355, + "grad_norm": 1.012115478515625, + "learning_rate": 6.955499976681014e-05, + "loss": 0.3411, + "step": 35193 + }, + { + "epoch": 2.442162237179932, + "grad_norm": 2.5074102878570557, + "learning_rate": 6.954964792452753e-05, + "loss": 0.3813, + "step": 35194 + }, + { + "epoch": 2.4422316286170287, + "grad_norm": 1.7978253364562988, + "learning_rate": 6.954429617837314e-05, + "loss": 0.6288, + "step": 35195 + }, + { + "epoch": 2.442301020054125, + "grad_norm": 1.420800805091858, + "learning_rate": 6.953894452836387e-05, + "loss": 0.4398, + "step": 35196 + }, + { + "epoch": 2.442370411491222, + "grad_norm": 1.1523168087005615, + "learning_rate": 6.953359297451663e-05, + "loss": 0.4244, + "step": 35197 + }, + { + "epoch": 2.4424398029283187, + "grad_norm": 1.5798259973526, + "learning_rate": 6.952824151684831e-05, + "loss": 0.5555, + "step": 35198 + }, + { + "epoch": 2.442509194365415, + "grad_norm": 1.1225075721740723, + "learning_rate": 6.952289015537579e-05, + "loss": 0.2371, + "step": 35199 + }, + { + "epoch": 2.442578585802512, + "grad_norm": 1.2706494331359863, + "learning_rate": 6.951753889011594e-05, + "loss": 0.4699, + "step": 35200 + }, + { + "epoch": 2.442647977239609, + "grad_norm": 0.9836789965629578, + "learning_rate": 6.951218772108573e-05, + "loss": 0.3434, + "step": 35201 + }, + { + "epoch": 2.442717368676705, + "grad_norm": 1.0348668098449707, + "learning_rate": 6.950683664830197e-05, + "loss": 0.4007, + "step": 35202 + }, + { + "epoch": 2.442786760113802, + "grad_norm": 1.3173528909683228, + "learning_rate": 6.95014856717816e-05, + "loss": 0.7337, + "step": 35203 + }, + { + "epoch": 2.442856151550899, + "grad_norm": 1.5043089389801025, + "learning_rate": 6.94961347915415e-05, + "loss": 0.4966, + "step": 35204 + }, + { + "epoch": 2.4429255429879952, + "grad_norm": 1.2109359502792358, + "learning_rate": 6.94907840075986e-05, + "loss": 0.4243, + "step": 35205 + }, + { + "epoch": 2.442994934425092, + "grad_norm": 1.156022548675537, + "learning_rate": 6.948543331996971e-05, + "loss": 0.3166, + "step": 35206 + }, + { + "epoch": 2.4430643258621885, + "grad_norm": 0.8739880919456482, + "learning_rate": 6.948008272867176e-05, + "loss": 0.339, + "step": 35207 + }, + { + "epoch": 2.4431337172992853, + "grad_norm": 1.5092053413391113, + "learning_rate": 6.947473223372167e-05, + "loss": 0.3693, + "step": 35208 + }, + { + "epoch": 2.443203108736382, + "grad_norm": 0.966795802116394, + "learning_rate": 6.94693818351363e-05, + "loss": 0.2985, + "step": 35209 + }, + { + "epoch": 2.4432725001734785, + "grad_norm": 1.4084292650222778, + "learning_rate": 6.946403153293255e-05, + "loss": 0.3528, + "step": 35210 + }, + { + "epoch": 2.4433418916105754, + "grad_norm": 1.2822012901306152, + "learning_rate": 6.94586813271273e-05, + "loss": 0.4437, + "step": 35211 + }, + { + "epoch": 2.4434112830476717, + "grad_norm": 1.0839821100234985, + "learning_rate": 6.94533312177375e-05, + "loss": 0.4288, + "step": 35212 + }, + { + "epoch": 2.4434806744847686, + "grad_norm": 1.3439898490905762, + "learning_rate": 6.944798120477994e-05, + "loss": 0.3476, + "step": 35213 + }, + { + "epoch": 2.4435500659218654, + "grad_norm": 1.5291059017181396, + "learning_rate": 6.944263128827158e-05, + "loss": 0.4496, + "step": 35214 + }, + { + "epoch": 2.443619457358962, + "grad_norm": 1.4532279968261719, + "learning_rate": 6.943728146822925e-05, + "loss": 0.5162, + "step": 35215 + }, + { + "epoch": 2.4436888487960586, + "grad_norm": 3.0510036945343018, + "learning_rate": 6.943193174466994e-05, + "loss": 0.5821, + "step": 35216 + }, + { + "epoch": 2.443758240233155, + "grad_norm": 1.1979409456253052, + "learning_rate": 6.942658211761042e-05, + "loss": 0.4236, + "step": 35217 + }, + { + "epoch": 2.443827631670252, + "grad_norm": 1.455483078956604, + "learning_rate": 6.942123258706766e-05, + "loss": 0.4932, + "step": 35218 + }, + { + "epoch": 2.4438970231073487, + "grad_norm": 1.2923314571380615, + "learning_rate": 6.941588315305851e-05, + "loss": 0.2997, + "step": 35219 + }, + { + "epoch": 2.443966414544445, + "grad_norm": 1.6512277126312256, + "learning_rate": 6.941053381559989e-05, + "loss": 0.2819, + "step": 35220 + }, + { + "epoch": 2.444035805981542, + "grad_norm": 1.7092264890670776, + "learning_rate": 6.940518457470864e-05, + "loss": 0.3804, + "step": 35221 + }, + { + "epoch": 2.4441051974186383, + "grad_norm": 1.89292573928833, + "learning_rate": 6.939983543040167e-05, + "loss": 0.4818, + "step": 35222 + }, + { + "epoch": 2.444174588855735, + "grad_norm": 1.037874460220337, + "learning_rate": 6.939448638269587e-05, + "loss": 0.3947, + "step": 35223 + }, + { + "epoch": 2.444243980292832, + "grad_norm": 1.6691035032272339, + "learning_rate": 6.938913743160814e-05, + "loss": 0.3692, + "step": 35224 + }, + { + "epoch": 2.4443133717299284, + "grad_norm": 1.5216501951217651, + "learning_rate": 6.938378857715534e-05, + "loss": 0.3889, + "step": 35225 + }, + { + "epoch": 2.444382763167025, + "grad_norm": 1.3055797815322876, + "learning_rate": 6.937843981935441e-05, + "loss": 0.3344, + "step": 35226 + }, + { + "epoch": 2.444452154604122, + "grad_norm": 0.8857874274253845, + "learning_rate": 6.937309115822215e-05, + "loss": 0.3439, + "step": 35227 + }, + { + "epoch": 2.4445215460412184, + "grad_norm": 0.7713494300842285, + "learning_rate": 6.93677425937755e-05, + "loss": 0.2799, + "step": 35228 + }, + { + "epoch": 2.4445909374783152, + "grad_norm": 1.2070995569229126, + "learning_rate": 6.936239412603132e-05, + "loss": 0.4347, + "step": 35229 + }, + { + "epoch": 2.444660328915412, + "grad_norm": 1.624788522720337, + "learning_rate": 6.93570457550065e-05, + "loss": 0.4679, + "step": 35230 + }, + { + "epoch": 2.4447297203525085, + "grad_norm": 1.2156827449798584, + "learning_rate": 6.935169748071794e-05, + "loss": 0.358, + "step": 35231 + }, + { + "epoch": 2.4447991117896053, + "grad_norm": 1.2188355922698975, + "learning_rate": 6.934634930318254e-05, + "loss": 0.3757, + "step": 35232 + }, + { + "epoch": 2.4448685032267017, + "grad_norm": 1.22077476978302, + "learning_rate": 6.934100122241715e-05, + "loss": 0.3552, + "step": 35233 + }, + { + "epoch": 2.4449378946637985, + "grad_norm": 1.400586485862732, + "learning_rate": 6.933565323843864e-05, + "loss": 0.4904, + "step": 35234 + }, + { + "epoch": 2.4450072861008953, + "grad_norm": 1.5357401371002197, + "learning_rate": 6.933030535126393e-05, + "loss": 0.4072, + "step": 35235 + }, + { + "epoch": 2.4450766775379917, + "grad_norm": 1.1104798316955566, + "learning_rate": 6.932495756090988e-05, + "loss": 0.3857, + "step": 35236 + }, + { + "epoch": 2.4451460689750886, + "grad_norm": 1.328879952430725, + "learning_rate": 6.931960986739338e-05, + "loss": 0.5274, + "step": 35237 + }, + { + "epoch": 2.445215460412185, + "grad_norm": 1.0993263721466064, + "learning_rate": 6.931426227073132e-05, + "loss": 0.3604, + "step": 35238 + }, + { + "epoch": 2.445284851849282, + "grad_norm": 1.095839500427246, + "learning_rate": 6.930891477094056e-05, + "loss": 0.4269, + "step": 35239 + }, + { + "epoch": 2.4453542432863786, + "grad_norm": 2.0191714763641357, + "learning_rate": 6.930356736803805e-05, + "loss": 0.3759, + "step": 35240 + }, + { + "epoch": 2.445423634723475, + "grad_norm": 1.654549241065979, + "learning_rate": 6.929822006204057e-05, + "loss": 0.4666, + "step": 35241 + }, + { + "epoch": 2.445493026160572, + "grad_norm": 1.4304124116897583, + "learning_rate": 6.929287285296504e-05, + "loss": 0.5998, + "step": 35242 + }, + { + "epoch": 2.4455624175976682, + "grad_norm": 1.0884218215942383, + "learning_rate": 6.928752574082836e-05, + "loss": 0.3504, + "step": 35243 + }, + { + "epoch": 2.445631809034765, + "grad_norm": 1.7074079513549805, + "learning_rate": 6.928217872564738e-05, + "loss": 0.4329, + "step": 35244 + }, + { + "epoch": 2.445701200471862, + "grad_norm": 1.2604832649230957, + "learning_rate": 6.9276831807439e-05, + "loss": 0.4706, + "step": 35245 + }, + { + "epoch": 2.4457705919089583, + "grad_norm": 1.1186765432357788, + "learning_rate": 6.927148498622015e-05, + "loss": 0.3506, + "step": 35246 + }, + { + "epoch": 2.445839983346055, + "grad_norm": 1.6864783763885498, + "learning_rate": 6.926613826200761e-05, + "loss": 0.3547, + "step": 35247 + }, + { + "epoch": 2.4459093747831515, + "grad_norm": 1.29188871383667, + "learning_rate": 6.92607916348183e-05, + "loss": 0.508, + "step": 35248 + }, + { + "epoch": 2.4459787662202483, + "grad_norm": 0.8629302978515625, + "learning_rate": 6.925544510466912e-05, + "loss": 0.3593, + "step": 35249 + }, + { + "epoch": 2.446048157657345, + "grad_norm": 1.2491050958633423, + "learning_rate": 6.925009867157692e-05, + "loss": 0.4816, + "step": 35250 + }, + { + "epoch": 2.4461175490944416, + "grad_norm": 1.0767902135849, + "learning_rate": 6.92447523355586e-05, + "loss": 0.4231, + "step": 35251 + }, + { + "epoch": 2.4461869405315384, + "grad_norm": 1.2654284238815308, + "learning_rate": 6.923940609663103e-05, + "loss": 0.472, + "step": 35252 + }, + { + "epoch": 2.4462563319686352, + "grad_norm": 2.122967481613159, + "learning_rate": 6.92340599548111e-05, + "loss": 0.4377, + "step": 35253 + }, + { + "epoch": 2.4463257234057316, + "grad_norm": 1.2034345865249634, + "learning_rate": 6.922871391011564e-05, + "loss": 0.3616, + "step": 35254 + }, + { + "epoch": 2.4463951148428285, + "grad_norm": 0.886681079864502, + "learning_rate": 6.922336796256161e-05, + "loss": 0.3185, + "step": 35255 + }, + { + "epoch": 2.4464645062799253, + "grad_norm": 1.1369471549987793, + "learning_rate": 6.921802211216579e-05, + "loss": 0.3638, + "step": 35256 + }, + { + "epoch": 2.4465338977170217, + "grad_norm": 1.2414294481277466, + "learning_rate": 6.921267635894513e-05, + "loss": 0.419, + "step": 35257 + }, + { + "epoch": 2.4466032891541185, + "grad_norm": 1.2420690059661865, + "learning_rate": 6.920733070291646e-05, + "loss": 0.448, + "step": 35258 + }, + { + "epoch": 2.446672680591215, + "grad_norm": 1.4326943159103394, + "learning_rate": 6.920198514409666e-05, + "loss": 0.4231, + "step": 35259 + }, + { + "epoch": 2.4467420720283117, + "grad_norm": 1.4650630950927734, + "learning_rate": 6.919663968250267e-05, + "loss": 0.6185, + "step": 35260 + }, + { + "epoch": 2.4468114634654086, + "grad_norm": 1.7194992303848267, + "learning_rate": 6.91912943181513e-05, + "loss": 0.4527, + "step": 35261 + }, + { + "epoch": 2.446880854902505, + "grad_norm": 1.2838791608810425, + "learning_rate": 6.91859490510594e-05, + "loss": 0.395, + "step": 35262 + }, + { + "epoch": 2.446950246339602, + "grad_norm": 1.2382551431655884, + "learning_rate": 6.91806038812439e-05, + "loss": 0.3632, + "step": 35263 + }, + { + "epoch": 2.447019637776698, + "grad_norm": 1.404233455657959, + "learning_rate": 6.917525880872166e-05, + "loss": 0.4111, + "step": 35264 + }, + { + "epoch": 2.447089029213795, + "grad_norm": 1.1564888954162598, + "learning_rate": 6.916991383350957e-05, + "loss": 0.4857, + "step": 35265 + }, + { + "epoch": 2.447158420650892, + "grad_norm": 1.6056607961654663, + "learning_rate": 6.916456895562447e-05, + "loss": 0.5281, + "step": 35266 + }, + { + "epoch": 2.4472278120879882, + "grad_norm": 1.308270812034607, + "learning_rate": 6.915922417508327e-05, + "loss": 0.419, + "step": 35267 + }, + { + "epoch": 2.447297203525085, + "grad_norm": 1.309706211090088, + "learning_rate": 6.91538794919028e-05, + "loss": 0.4946, + "step": 35268 + }, + { + "epoch": 2.4473665949621815, + "grad_norm": 1.1951135396957397, + "learning_rate": 6.914853490609995e-05, + "loss": 0.4166, + "step": 35269 + }, + { + "epoch": 2.4474359863992783, + "grad_norm": 1.9340415000915527, + "learning_rate": 6.914319041769162e-05, + "loss": 0.4216, + "step": 35270 + }, + { + "epoch": 2.447505377836375, + "grad_norm": 1.3663851022720337, + "learning_rate": 6.913784602669464e-05, + "loss": 0.3548, + "step": 35271 + }, + { + "epoch": 2.4475747692734715, + "grad_norm": 1.0857537984848022, + "learning_rate": 6.913250173312589e-05, + "loss": 0.414, + "step": 35272 + }, + { + "epoch": 2.4476441607105683, + "grad_norm": 1.2084190845489502, + "learning_rate": 6.912715753700225e-05, + "loss": 0.4156, + "step": 35273 + }, + { + "epoch": 2.4477135521476647, + "grad_norm": 1.8287996053695679, + "learning_rate": 6.912181343834063e-05, + "loss": 0.3997, + "step": 35274 + }, + { + "epoch": 2.4477829435847616, + "grad_norm": 2.409874200820923, + "learning_rate": 6.911646943715784e-05, + "loss": 0.7198, + "step": 35275 + }, + { + "epoch": 2.4478523350218584, + "grad_norm": 0.9051225185394287, + "learning_rate": 6.911112553347074e-05, + "loss": 0.3177, + "step": 35276 + }, + { + "epoch": 2.447921726458955, + "grad_norm": 1.5069360733032227, + "learning_rate": 6.910578172729627e-05, + "loss": 0.5332, + "step": 35277 + }, + { + "epoch": 2.4479911178960516, + "grad_norm": 1.6665499210357666, + "learning_rate": 6.910043801865124e-05, + "loss": 0.4998, + "step": 35278 + }, + { + "epoch": 2.4480605093331484, + "grad_norm": 1.902246356010437, + "learning_rate": 6.909509440755255e-05, + "loss": 0.3327, + "step": 35279 + }, + { + "epoch": 2.448129900770245, + "grad_norm": 1.158388614654541, + "learning_rate": 6.908975089401705e-05, + "loss": 0.4099, + "step": 35280 + }, + { + "epoch": 2.4481992922073417, + "grad_norm": 1.417466163635254, + "learning_rate": 6.908440747806169e-05, + "loss": 0.4394, + "step": 35281 + }, + { + "epoch": 2.4482686836444385, + "grad_norm": 1.2500439882278442, + "learning_rate": 6.90790641597032e-05, + "loss": 0.3916, + "step": 35282 + }, + { + "epoch": 2.448338075081535, + "grad_norm": 1.1011978387832642, + "learning_rate": 6.907372093895852e-05, + "loss": 0.3732, + "step": 35283 + }, + { + "epoch": 2.4484074665186317, + "grad_norm": 1.1199009418487549, + "learning_rate": 6.906837781584452e-05, + "loss": 0.4243, + "step": 35284 + }, + { + "epoch": 2.448476857955728, + "grad_norm": 1.5898140668869019, + "learning_rate": 6.906303479037807e-05, + "loss": 0.4686, + "step": 35285 + }, + { + "epoch": 2.448546249392825, + "grad_norm": 1.1267846822738647, + "learning_rate": 6.905769186257605e-05, + "loss": 0.3359, + "step": 35286 + }, + { + "epoch": 2.448615640829922, + "grad_norm": 1.5562690496444702, + "learning_rate": 6.905234903245527e-05, + "loss": 0.5531, + "step": 35287 + }, + { + "epoch": 2.448685032267018, + "grad_norm": 1.1320114135742188, + "learning_rate": 6.904700630003267e-05, + "loss": 0.3759, + "step": 35288 + }, + { + "epoch": 2.448754423704115, + "grad_norm": 1.2732905149459839, + "learning_rate": 6.904166366532505e-05, + "loss": 0.4441, + "step": 35289 + }, + { + "epoch": 2.4488238151412114, + "grad_norm": 1.5484496355056763, + "learning_rate": 6.903632112834933e-05, + "loss": 0.6489, + "step": 35290 + }, + { + "epoch": 2.4488932065783082, + "grad_norm": 1.9889435768127441, + "learning_rate": 6.903097868912233e-05, + "loss": 0.4223, + "step": 35291 + }, + { + "epoch": 2.448962598015405, + "grad_norm": 1.976928949356079, + "learning_rate": 6.902563634766092e-05, + "loss": 0.7588, + "step": 35292 + }, + { + "epoch": 2.4490319894525014, + "grad_norm": 1.0786404609680176, + "learning_rate": 6.902029410398201e-05, + "loss": 0.3716, + "step": 35293 + }, + { + "epoch": 2.4491013808895983, + "grad_norm": 0.8997547626495361, + "learning_rate": 6.901495195810243e-05, + "loss": 0.3239, + "step": 35294 + }, + { + "epoch": 2.4491707723266947, + "grad_norm": 1.6077237129211426, + "learning_rate": 6.900960991003909e-05, + "loss": 0.4549, + "step": 35295 + }, + { + "epoch": 2.4492401637637915, + "grad_norm": 1.5503896474838257, + "learning_rate": 6.900426795980878e-05, + "loss": 0.3742, + "step": 35296 + }, + { + "epoch": 2.4493095552008883, + "grad_norm": 1.0911571979522705, + "learning_rate": 6.89989261074284e-05, + "loss": 0.3073, + "step": 35297 + }, + { + "epoch": 2.4493789466379847, + "grad_norm": 1.0309593677520752, + "learning_rate": 6.89935843529148e-05, + "loss": 0.3632, + "step": 35298 + }, + { + "epoch": 2.4494483380750816, + "grad_norm": 1.7272182703018188, + "learning_rate": 6.898824269628486e-05, + "loss": 0.5196, + "step": 35299 + }, + { + "epoch": 2.4495177295121784, + "grad_norm": 1.1750805377960205, + "learning_rate": 6.898290113755546e-05, + "loss": 0.4001, + "step": 35300 + }, + { + "epoch": 2.4495871209492748, + "grad_norm": 1.3194562196731567, + "learning_rate": 6.897755967674341e-05, + "loss": 0.4714, + "step": 35301 + }, + { + "epoch": 2.4496565123863716, + "grad_norm": 1.5434174537658691, + "learning_rate": 6.897221831386567e-05, + "loss": 0.4625, + "step": 35302 + }, + { + "epoch": 2.449725903823468, + "grad_norm": 1.6240928173065186, + "learning_rate": 6.8966877048939e-05, + "loss": 0.5339, + "step": 35303 + }, + { + "epoch": 2.449795295260565, + "grad_norm": 1.4340461492538452, + "learning_rate": 6.896153588198028e-05, + "loss": 0.549, + "step": 35304 + }, + { + "epoch": 2.4498646866976617, + "grad_norm": 1.2499985694885254, + "learning_rate": 6.895619481300642e-05, + "loss": 0.3854, + "step": 35305 + }, + { + "epoch": 2.449934078134758, + "grad_norm": 0.8998050093650818, + "learning_rate": 6.895085384203424e-05, + "loss": 0.2126, + "step": 35306 + }, + { + "epoch": 2.450003469571855, + "grad_norm": 1.2295582294464111, + "learning_rate": 6.894551296908058e-05, + "loss": 0.44, + "step": 35307 + }, + { + "epoch": 2.4500728610089517, + "grad_norm": 1.4360711574554443, + "learning_rate": 6.894017219416236e-05, + "loss": 0.4108, + "step": 35308 + }, + { + "epoch": 2.450142252446048, + "grad_norm": 1.3896242380142212, + "learning_rate": 6.893483151729645e-05, + "loss": 0.391, + "step": 35309 + }, + { + "epoch": 2.450211643883145, + "grad_norm": 1.1251298189163208, + "learning_rate": 6.892949093849963e-05, + "loss": 0.4837, + "step": 35310 + }, + { + "epoch": 2.4502810353202413, + "grad_norm": 1.5340197086334229, + "learning_rate": 6.89241504577888e-05, + "loss": 0.4622, + "step": 35311 + }, + { + "epoch": 2.450350426757338, + "grad_norm": 1.4501978158950806, + "learning_rate": 6.891881007518082e-05, + "loss": 0.4113, + "step": 35312 + }, + { + "epoch": 2.450419818194435, + "grad_norm": 1.3158282041549683, + "learning_rate": 6.891346979069255e-05, + "loss": 0.4411, + "step": 35313 + }, + { + "epoch": 2.4504892096315314, + "grad_norm": 1.5386756658554077, + "learning_rate": 6.890812960434085e-05, + "loss": 0.5589, + "step": 35314 + }, + { + "epoch": 2.450558601068628, + "grad_norm": 1.215728521347046, + "learning_rate": 6.890278951614259e-05, + "loss": 0.3504, + "step": 35315 + }, + { + "epoch": 2.4506279925057246, + "grad_norm": 1.3496639728546143, + "learning_rate": 6.889744952611466e-05, + "loss": 0.5582, + "step": 35316 + }, + { + "epoch": 2.4506973839428214, + "grad_norm": 1.4853936433792114, + "learning_rate": 6.88921096342738e-05, + "loss": 0.3169, + "step": 35317 + }, + { + "epoch": 2.4507667753799183, + "grad_norm": 1.1413780450820923, + "learning_rate": 6.888676984063696e-05, + "loss": 0.4086, + "step": 35318 + }, + { + "epoch": 2.4508361668170147, + "grad_norm": 1.1978737115859985, + "learning_rate": 6.888143014522098e-05, + "loss": 0.3639, + "step": 35319 + }, + { + "epoch": 2.4509055582541115, + "grad_norm": 1.4345465898513794, + "learning_rate": 6.887609054804272e-05, + "loss": 0.3978, + "step": 35320 + }, + { + "epoch": 2.450974949691208, + "grad_norm": 1.3532307147979736, + "learning_rate": 6.887075104911904e-05, + "loss": 0.3636, + "step": 35321 + }, + { + "epoch": 2.4510443411283047, + "grad_norm": 1.2431631088256836, + "learning_rate": 6.886541164846679e-05, + "loss": 0.4542, + "step": 35322 + }, + { + "epoch": 2.4511137325654015, + "grad_norm": 1.2029311656951904, + "learning_rate": 6.886007234610281e-05, + "loss": 0.3387, + "step": 35323 + }, + { + "epoch": 2.451183124002498, + "grad_norm": 1.4461555480957031, + "learning_rate": 6.885473314204398e-05, + "loss": 0.5286, + "step": 35324 + }, + { + "epoch": 2.4512525154395948, + "grad_norm": 1.2599860429763794, + "learning_rate": 6.884939403630713e-05, + "loss": 0.3824, + "step": 35325 + }, + { + "epoch": 2.4513219068766916, + "grad_norm": 1.2986631393432617, + "learning_rate": 6.884405502890912e-05, + "loss": 0.4507, + "step": 35326 + }, + { + "epoch": 2.451391298313788, + "grad_norm": 1.0342241525650024, + "learning_rate": 6.883871611986682e-05, + "loss": 0.4119, + "step": 35327 + }, + { + "epoch": 2.451460689750885, + "grad_norm": 1.4548273086547852, + "learning_rate": 6.883337730919708e-05, + "loss": 0.5179, + "step": 35328 + }, + { + "epoch": 2.451530081187981, + "grad_norm": 1.2887102365493774, + "learning_rate": 6.882803859691677e-05, + "loss": 0.3926, + "step": 35329 + }, + { + "epoch": 2.451599472625078, + "grad_norm": 1.504287838935852, + "learning_rate": 6.882269998304272e-05, + "loss": 0.4419, + "step": 35330 + }, + { + "epoch": 2.451668864062175, + "grad_norm": 1.0497664213180542, + "learning_rate": 6.881736146759177e-05, + "loss": 0.3704, + "step": 35331 + }, + { + "epoch": 2.4517382554992713, + "grad_norm": 1.2210825681686401, + "learning_rate": 6.88120230505808e-05, + "loss": 0.3539, + "step": 35332 + }, + { + "epoch": 2.451807646936368, + "grad_norm": 1.2254639863967896, + "learning_rate": 6.880668473202664e-05, + "loss": 0.3848, + "step": 35333 + }, + { + "epoch": 2.451877038373465, + "grad_norm": 0.9591969847679138, + "learning_rate": 6.880134651194618e-05, + "loss": 0.3622, + "step": 35334 + }, + { + "epoch": 2.4519464298105613, + "grad_norm": 1.3596906661987305, + "learning_rate": 6.879600839035622e-05, + "loss": 0.4157, + "step": 35335 + }, + { + "epoch": 2.452015821247658, + "grad_norm": 1.5662405490875244, + "learning_rate": 6.879067036727372e-05, + "loss": 0.5303, + "step": 35336 + }, + { + "epoch": 2.4520852126847545, + "grad_norm": 1.4705235958099365, + "learning_rate": 6.878533244271538e-05, + "loss": 0.4692, + "step": 35337 + }, + { + "epoch": 2.4521546041218514, + "grad_norm": 1.3617316484451294, + "learning_rate": 6.877999461669812e-05, + "loss": 0.477, + "step": 35338 + }, + { + "epoch": 2.452223995558948, + "grad_norm": 1.2783724069595337, + "learning_rate": 6.877465688923882e-05, + "loss": 0.4758, + "step": 35339 + }, + { + "epoch": 2.4522933869960446, + "grad_norm": 1.3889400959014893, + "learning_rate": 6.876931926035431e-05, + "loss": 0.4617, + "step": 35340 + }, + { + "epoch": 2.4523627784331414, + "grad_norm": 0.9305471181869507, + "learning_rate": 6.876398173006143e-05, + "loss": 0.3031, + "step": 35341 + }, + { + "epoch": 2.452432169870238, + "grad_norm": 1.0102609395980835, + "learning_rate": 6.875864429837701e-05, + "loss": 0.3883, + "step": 35342 + }, + { + "epoch": 2.4525015613073347, + "grad_norm": 1.3568968772888184, + "learning_rate": 6.875330696531796e-05, + "loss": 0.3462, + "step": 35343 + }, + { + "epoch": 2.4525709527444315, + "grad_norm": 1.117208480834961, + "learning_rate": 6.874796973090111e-05, + "loss": 0.476, + "step": 35344 + }, + { + "epoch": 2.452640344181528, + "grad_norm": 1.2035194635391235, + "learning_rate": 6.874263259514326e-05, + "loss": 0.4848, + "step": 35345 + }, + { + "epoch": 2.4527097356186247, + "grad_norm": 1.3246123790740967, + "learning_rate": 6.873729555806129e-05, + "loss": 0.4548, + "step": 35346 + }, + { + "epoch": 2.452779127055721, + "grad_norm": 1.5142686367034912, + "learning_rate": 6.873195861967205e-05, + "loss": 0.4724, + "step": 35347 + }, + { + "epoch": 2.452848518492818, + "grad_norm": 1.1183291673660278, + "learning_rate": 6.872662177999238e-05, + "loss": 0.494, + "step": 35348 + }, + { + "epoch": 2.4529179099299148, + "grad_norm": 1.139696717262268, + "learning_rate": 6.872128503903916e-05, + "loss": 0.4646, + "step": 35349 + }, + { + "epoch": 2.452987301367011, + "grad_norm": 1.1549724340438843, + "learning_rate": 6.871594839682925e-05, + "loss": 0.3565, + "step": 35350 + }, + { + "epoch": 2.453056692804108, + "grad_norm": 1.442154049873352, + "learning_rate": 6.871061185337942e-05, + "loss": 0.4611, + "step": 35351 + }, + { + "epoch": 2.453126084241205, + "grad_norm": 1.6526118516921997, + "learning_rate": 6.870527540870654e-05, + "loss": 0.4423, + "step": 35352 + }, + { + "epoch": 2.453195475678301, + "grad_norm": 1.0641348361968994, + "learning_rate": 6.869993906282747e-05, + "loss": 0.3116, + "step": 35353 + }, + { + "epoch": 2.453264867115398, + "grad_norm": 1.0304384231567383, + "learning_rate": 6.869460281575909e-05, + "loss": 0.4053, + "step": 35354 + }, + { + "epoch": 2.4533342585524944, + "grad_norm": 1.665966510772705, + "learning_rate": 6.86892666675182e-05, + "loss": 0.3571, + "step": 35355 + }, + { + "epoch": 2.4534036499895913, + "grad_norm": 1.448595643043518, + "learning_rate": 6.868393061812169e-05, + "loss": 0.6464, + "step": 35356 + }, + { + "epoch": 2.453473041426688, + "grad_norm": 0.9572982788085938, + "learning_rate": 6.867859466758637e-05, + "loss": 0.4297, + "step": 35357 + }, + { + "epoch": 2.4535424328637845, + "grad_norm": 0.8843015432357788, + "learning_rate": 6.867325881592908e-05, + "loss": 0.3564, + "step": 35358 + }, + { + "epoch": 2.4536118243008813, + "grad_norm": 1.0844064950942993, + "learning_rate": 6.86679230631667e-05, + "loss": 0.6573, + "step": 35359 + }, + { + "epoch": 2.453681215737978, + "grad_norm": 1.6728019714355469, + "learning_rate": 6.866258740931601e-05, + "loss": 0.6603, + "step": 35360 + }, + { + "epoch": 2.4537506071750745, + "grad_norm": 1.2743010520935059, + "learning_rate": 6.865725185439392e-05, + "loss": 0.2956, + "step": 35361 + }, + { + "epoch": 2.4538199986121714, + "grad_norm": 1.1151065826416016, + "learning_rate": 6.865191639841724e-05, + "loss": 0.3655, + "step": 35362 + }, + { + "epoch": 2.4538893900492678, + "grad_norm": 1.7464754581451416, + "learning_rate": 6.864658104140282e-05, + "loss": 0.4482, + "step": 35363 + }, + { + "epoch": 2.4539587814863646, + "grad_norm": 1.273598313331604, + "learning_rate": 6.864124578336756e-05, + "loss": 0.402, + "step": 35364 + }, + { + "epoch": 2.4540281729234614, + "grad_norm": 1.482237458229065, + "learning_rate": 6.863591062432819e-05, + "loss": 0.5482, + "step": 35365 + }, + { + "epoch": 2.454097564360558, + "grad_norm": 1.354857325553894, + "learning_rate": 6.863057556430162e-05, + "loss": 0.4442, + "step": 35366 + }, + { + "epoch": 2.4541669557976546, + "grad_norm": 0.8227863907814026, + "learning_rate": 6.862524060330468e-05, + "loss": 0.3326, + "step": 35367 + }, + { + "epoch": 2.454236347234751, + "grad_norm": 1.255320429801941, + "learning_rate": 6.861990574135423e-05, + "loss": 0.4311, + "step": 35368 + }, + { + "epoch": 2.454305738671848, + "grad_norm": 1.1563613414764404, + "learning_rate": 6.861457097846708e-05, + "loss": 0.401, + "step": 35369 + }, + { + "epoch": 2.4543751301089447, + "grad_norm": 1.0703284740447998, + "learning_rate": 6.86092363146601e-05, + "loss": 0.2751, + "step": 35370 + }, + { + "epoch": 2.454444521546041, + "grad_norm": 1.3614140748977661, + "learning_rate": 6.860390174995015e-05, + "loss": 0.3509, + "step": 35371 + }, + { + "epoch": 2.454513912983138, + "grad_norm": 0.9205460548400879, + "learning_rate": 6.8598567284354e-05, + "loss": 0.2242, + "step": 35372 + }, + { + "epoch": 2.4545833044202343, + "grad_norm": 1.380057692527771, + "learning_rate": 6.859323291788853e-05, + "loss": 0.5037, + "step": 35373 + }, + { + "epoch": 2.454652695857331, + "grad_norm": 0.9552251100540161, + "learning_rate": 6.858789865057058e-05, + "loss": 0.2904, + "step": 35374 + }, + { + "epoch": 2.454722087294428, + "grad_norm": 1.6303784847259521, + "learning_rate": 6.858256448241703e-05, + "loss": 0.456, + "step": 35375 + }, + { + "epoch": 2.4547914787315244, + "grad_norm": 1.3569594621658325, + "learning_rate": 6.857723041344462e-05, + "loss": 0.3275, + "step": 35376 + }, + { + "epoch": 2.454860870168621, + "grad_norm": 1.0183424949645996, + "learning_rate": 6.857189644367026e-05, + "loss": 0.4062, + "step": 35377 + }, + { + "epoch": 2.454930261605718, + "grad_norm": 1.1409578323364258, + "learning_rate": 6.856656257311081e-05, + "loss": 0.4313, + "step": 35378 + }, + { + "epoch": 2.4549996530428144, + "grad_norm": 1.183423638343811, + "learning_rate": 6.856122880178309e-05, + "loss": 0.3065, + "step": 35379 + }, + { + "epoch": 2.4550690444799113, + "grad_norm": 1.9506077766418457, + "learning_rate": 6.855589512970386e-05, + "loss": 0.6043, + "step": 35380 + }, + { + "epoch": 2.455138435917008, + "grad_norm": 1.934221863746643, + "learning_rate": 6.855056155689006e-05, + "loss": 0.5367, + "step": 35381 + }, + { + "epoch": 2.4552078273541045, + "grad_norm": 1.3715821504592896, + "learning_rate": 6.854522808335846e-05, + "loss": 0.3447, + "step": 35382 + }, + { + "epoch": 2.4552772187912013, + "grad_norm": 1.0444608926773071, + "learning_rate": 6.853989470912594e-05, + "loss": 0.3966, + "step": 35383 + }, + { + "epoch": 2.4553466102282977, + "grad_norm": 1.198586106300354, + "learning_rate": 6.853456143420931e-05, + "loss": 0.4398, + "step": 35384 + }, + { + "epoch": 2.4554160016653945, + "grad_norm": 1.1482456922531128, + "learning_rate": 6.852922825862548e-05, + "loss": 0.3613, + "step": 35385 + }, + { + "epoch": 2.4554853931024914, + "grad_norm": 1.5692553520202637, + "learning_rate": 6.852389518239117e-05, + "loss": 0.3644, + "step": 35386 + }, + { + "epoch": 2.4555547845395878, + "grad_norm": 1.4514931440353394, + "learning_rate": 6.851856220552328e-05, + "loss": 0.3414, + "step": 35387 + }, + { + "epoch": 2.4556241759766846, + "grad_norm": 1.205909013748169, + "learning_rate": 6.851322932803863e-05, + "loss": 0.371, + "step": 35388 + }, + { + "epoch": 2.455693567413781, + "grad_norm": 1.3101712465286255, + "learning_rate": 6.850789654995407e-05, + "loss": 0.4679, + "step": 35389 + }, + { + "epoch": 2.455762958850878, + "grad_norm": 0.8056755065917969, + "learning_rate": 6.85025638712864e-05, + "loss": 0.2657, + "step": 35390 + }, + { + "epoch": 2.4558323502879746, + "grad_norm": 1.4769985675811768, + "learning_rate": 6.849723129205252e-05, + "loss": 0.5485, + "step": 35391 + }, + { + "epoch": 2.455901741725071, + "grad_norm": 1.9199657440185547, + "learning_rate": 6.849189881226923e-05, + "loss": 0.5414, + "step": 35392 + }, + { + "epoch": 2.455971133162168, + "grad_norm": 1.9453496932983398, + "learning_rate": 6.848656643195335e-05, + "loss": 0.8267, + "step": 35393 + }, + { + "epoch": 2.4560405245992643, + "grad_norm": 1.1163036823272705, + "learning_rate": 6.848123415112173e-05, + "loss": 0.3796, + "step": 35394 + }, + { + "epoch": 2.456109916036361, + "grad_norm": 1.4344069957733154, + "learning_rate": 6.847590196979117e-05, + "loss": 0.5235, + "step": 35395 + }, + { + "epoch": 2.456179307473458, + "grad_norm": 1.3447448015213013, + "learning_rate": 6.847056988797854e-05, + "loss": 0.3938, + "step": 35396 + }, + { + "epoch": 2.4562486989105543, + "grad_norm": 1.3996845483779907, + "learning_rate": 6.846523790570065e-05, + "loss": 0.4016, + "step": 35397 + }, + { + "epoch": 2.456318090347651, + "grad_norm": 0.9554126858711243, + "learning_rate": 6.845990602297441e-05, + "loss": 0.366, + "step": 35398 + }, + { + "epoch": 2.4563874817847475, + "grad_norm": 1.3385509252548218, + "learning_rate": 6.845457423981652e-05, + "loss": 0.4242, + "step": 35399 + }, + { + "epoch": 2.4564568732218444, + "grad_norm": 1.5217808485031128, + "learning_rate": 6.84492425562439e-05, + "loss": 0.4125, + "step": 35400 + }, + { + "epoch": 2.456526264658941, + "grad_norm": 1.8067759275436401, + "learning_rate": 6.844391097227335e-05, + "loss": 0.3649, + "step": 35401 + }, + { + "epoch": 2.4565956560960376, + "grad_norm": 1.0363407135009766, + "learning_rate": 6.843857948792172e-05, + "loss": 0.4657, + "step": 35402 + }, + { + "epoch": 2.4566650475331344, + "grad_norm": 0.9946461915969849, + "learning_rate": 6.843324810320581e-05, + "loss": 0.4055, + "step": 35403 + }, + { + "epoch": 2.4567344389702312, + "grad_norm": 1.49611496925354, + "learning_rate": 6.84279168181425e-05, + "loss": 0.244, + "step": 35404 + }, + { + "epoch": 2.4568038304073276, + "grad_norm": 1.4698264598846436, + "learning_rate": 6.842258563274864e-05, + "loss": 0.3191, + "step": 35405 + }, + { + "epoch": 2.4568732218444245, + "grad_norm": 1.493988275527954, + "learning_rate": 6.841725454704095e-05, + "loss": 0.4312, + "step": 35406 + }, + { + "epoch": 2.4569426132815213, + "grad_norm": 1.5134541988372803, + "learning_rate": 6.841192356103633e-05, + "loss": 0.5248, + "step": 35407 + }, + { + "epoch": 2.4570120047186177, + "grad_norm": 1.319568395614624, + "learning_rate": 6.84065926747516e-05, + "loss": 0.3725, + "step": 35408 + }, + { + "epoch": 2.4570813961557145, + "grad_norm": 1.3153492212295532, + "learning_rate": 6.84012618882036e-05, + "loss": 0.5663, + "step": 35409 + }, + { + "epoch": 2.457150787592811, + "grad_norm": 1.33306086063385, + "learning_rate": 6.839593120140915e-05, + "loss": 0.5385, + "step": 35410 + }, + { + "epoch": 2.4572201790299077, + "grad_norm": 1.1586885452270508, + "learning_rate": 6.839060061438508e-05, + "loss": 0.3727, + "step": 35411 + }, + { + "epoch": 2.4572895704670046, + "grad_norm": 1.3844881057739258, + "learning_rate": 6.838527012714823e-05, + "loss": 0.4084, + "step": 35412 + }, + { + "epoch": 2.457358961904101, + "grad_norm": 1.4610859155654907, + "learning_rate": 6.837993973971541e-05, + "loss": 0.4917, + "step": 35413 + }, + { + "epoch": 2.457428353341198, + "grad_norm": 1.093473196029663, + "learning_rate": 6.837460945210344e-05, + "loss": 0.4695, + "step": 35414 + }, + { + "epoch": 2.457497744778294, + "grad_norm": 1.2040091753005981, + "learning_rate": 6.836927926432916e-05, + "loss": 0.3769, + "step": 35415 + }, + { + "epoch": 2.457567136215391, + "grad_norm": 1.1554741859436035, + "learning_rate": 6.836394917640938e-05, + "loss": 0.3816, + "step": 35416 + }, + { + "epoch": 2.457636527652488, + "grad_norm": 1.3723113536834717, + "learning_rate": 6.835861918836096e-05, + "loss": 0.4514, + "step": 35417 + }, + { + "epoch": 2.4577059190895842, + "grad_norm": 1.4816455841064453, + "learning_rate": 6.83532893002007e-05, + "loss": 0.5497, + "step": 35418 + }, + { + "epoch": 2.457775310526681, + "grad_norm": 0.9832215905189514, + "learning_rate": 6.834795951194547e-05, + "loss": 0.2905, + "step": 35419 + }, + { + "epoch": 2.4578447019637775, + "grad_norm": 1.8887066841125488, + "learning_rate": 6.834262982361202e-05, + "loss": 0.5079, + "step": 35420 + }, + { + "epoch": 2.4579140934008743, + "grad_norm": 0.9686295390129089, + "learning_rate": 6.833730023521723e-05, + "loss": 0.303, + "step": 35421 + }, + { + "epoch": 2.457983484837971, + "grad_norm": 1.2844380140304565, + "learning_rate": 6.833197074677789e-05, + "loss": 0.2816, + "step": 35422 + }, + { + "epoch": 2.4580528762750675, + "grad_norm": 0.9677481055259705, + "learning_rate": 6.832664135831084e-05, + "loss": 0.2787, + "step": 35423 + }, + { + "epoch": 2.4581222677121644, + "grad_norm": 1.412757396697998, + "learning_rate": 6.832131206983293e-05, + "loss": 0.3493, + "step": 35424 + }, + { + "epoch": 2.4581916591492607, + "grad_norm": 1.4485831260681152, + "learning_rate": 6.831598288136098e-05, + "loss": 0.5843, + "step": 35425 + }, + { + "epoch": 2.4582610505863576, + "grad_norm": 1.5086525678634644, + "learning_rate": 6.831065379291178e-05, + "loss": 0.3986, + "step": 35426 + }, + { + "epoch": 2.4583304420234544, + "grad_norm": 0.9856176376342773, + "learning_rate": 6.830532480450216e-05, + "loss": 0.2974, + "step": 35427 + }, + { + "epoch": 2.458399833460551, + "grad_norm": 1.4343444108963013, + "learning_rate": 6.829999591614895e-05, + "loss": 0.4461, + "step": 35428 + }, + { + "epoch": 2.4584692248976476, + "grad_norm": 0.9257775545120239, + "learning_rate": 6.8294667127869e-05, + "loss": 0.3102, + "step": 35429 + }, + { + "epoch": 2.4585386163347445, + "grad_norm": 1.7886544466018677, + "learning_rate": 6.828933843967908e-05, + "loss": 0.6056, + "step": 35430 + }, + { + "epoch": 2.458608007771841, + "grad_norm": 1.4878586530685425, + "learning_rate": 6.828400985159605e-05, + "loss": 0.5094, + "step": 35431 + }, + { + "epoch": 2.4586773992089377, + "grad_norm": 2.4809136390686035, + "learning_rate": 6.827868136363671e-05, + "loss": 0.3029, + "step": 35432 + }, + { + "epoch": 2.4587467906460345, + "grad_norm": 1.2016196250915527, + "learning_rate": 6.827335297581794e-05, + "loss": 0.3398, + "step": 35433 + }, + { + "epoch": 2.458816182083131, + "grad_norm": 1.200027585029602, + "learning_rate": 6.826802468815646e-05, + "loss": 0.4259, + "step": 35434 + }, + { + "epoch": 2.4588855735202277, + "grad_norm": 1.0841937065124512, + "learning_rate": 6.826269650066916e-05, + "loss": 0.3947, + "step": 35435 + }, + { + "epoch": 2.458954964957324, + "grad_norm": 1.4922548532485962, + "learning_rate": 6.825736841337284e-05, + "loss": 0.4976, + "step": 35436 + }, + { + "epoch": 2.459024356394421, + "grad_norm": 1.1375190019607544, + "learning_rate": 6.825204042628432e-05, + "loss": 0.5484, + "step": 35437 + }, + { + "epoch": 2.459093747831518, + "grad_norm": 1.2222886085510254, + "learning_rate": 6.824671253942043e-05, + "loss": 0.3186, + "step": 35438 + }, + { + "epoch": 2.459163139268614, + "grad_norm": 1.7966417074203491, + "learning_rate": 6.824138475279798e-05, + "loss": 0.363, + "step": 35439 + }, + { + "epoch": 2.459232530705711, + "grad_norm": 1.1628056764602661, + "learning_rate": 6.823605706643383e-05, + "loss": 0.474, + "step": 35440 + }, + { + "epoch": 2.4593019221428074, + "grad_norm": 2.045382261276245, + "learning_rate": 6.823072948034473e-05, + "loss": 0.5618, + "step": 35441 + }, + { + "epoch": 2.4593713135799042, + "grad_norm": 1.5167580842971802, + "learning_rate": 6.822540199454755e-05, + "loss": 0.6356, + "step": 35442 + }, + { + "epoch": 2.459440705017001, + "grad_norm": 1.1677666902542114, + "learning_rate": 6.822007460905906e-05, + "loss": 0.3058, + "step": 35443 + }, + { + "epoch": 2.4595100964540975, + "grad_norm": 1.3084028959274292, + "learning_rate": 6.82147473238961e-05, + "loss": 0.4446, + "step": 35444 + }, + { + "epoch": 2.4595794878911943, + "grad_norm": 1.6701524257659912, + "learning_rate": 6.820942013907555e-05, + "loss": 0.5896, + "step": 35445 + }, + { + "epoch": 2.4596488793282907, + "grad_norm": 1.043662428855896, + "learning_rate": 6.820409305461413e-05, + "loss": 0.3655, + "step": 35446 + }, + { + "epoch": 2.4597182707653875, + "grad_norm": 1.3755062818527222, + "learning_rate": 6.819876607052874e-05, + "loss": 0.3339, + "step": 35447 + }, + { + "epoch": 2.4597876622024843, + "grad_norm": 1.194472312927246, + "learning_rate": 6.819343918683613e-05, + "loss": 0.3945, + "step": 35448 + }, + { + "epoch": 2.4598570536395807, + "grad_norm": 1.0831409692764282, + "learning_rate": 6.818811240355313e-05, + "loss": 0.4658, + "step": 35449 + }, + { + "epoch": 2.4599264450766776, + "grad_norm": 1.5670322179794312, + "learning_rate": 6.818278572069656e-05, + "loss": 0.4986, + "step": 35450 + }, + { + "epoch": 2.459995836513774, + "grad_norm": 1.2035681009292603, + "learning_rate": 6.817745913828326e-05, + "loss": 0.3701, + "step": 35451 + }, + { + "epoch": 2.460065227950871, + "grad_norm": 1.3222538232803345, + "learning_rate": 6.817213265633002e-05, + "loss": 0.5212, + "step": 35452 + }, + { + "epoch": 2.4601346193879676, + "grad_norm": 1.12490713596344, + "learning_rate": 6.816680627485368e-05, + "loss": 0.5138, + "step": 35453 + }, + { + "epoch": 2.460204010825064, + "grad_norm": 1.0321807861328125, + "learning_rate": 6.816147999387105e-05, + "loss": 0.2827, + "step": 35454 + }, + { + "epoch": 2.460273402262161, + "grad_norm": 1.0541218519210815, + "learning_rate": 6.815615381339892e-05, + "loss": 0.3928, + "step": 35455 + }, + { + "epoch": 2.4603427936992577, + "grad_norm": 1.102212905883789, + "learning_rate": 6.815082773345412e-05, + "loss": 0.4046, + "step": 35456 + }, + { + "epoch": 2.460412185136354, + "grad_norm": 1.3426917791366577, + "learning_rate": 6.814550175405345e-05, + "loss": 0.3768, + "step": 35457 + }, + { + "epoch": 2.460481576573451, + "grad_norm": 1.9818204641342163, + "learning_rate": 6.814017587521373e-05, + "loss": 0.3991, + "step": 35458 + }, + { + "epoch": 2.4605509680105477, + "grad_norm": 1.1385630369186401, + "learning_rate": 6.813485009695177e-05, + "loss": 0.3831, + "step": 35459 + }, + { + "epoch": 2.460620359447644, + "grad_norm": 0.8929520845413208, + "learning_rate": 6.812952441928444e-05, + "loss": 0.3597, + "step": 35460 + }, + { + "epoch": 2.460689750884741, + "grad_norm": 1.0446364879608154, + "learning_rate": 6.812419884222849e-05, + "loss": 0.4408, + "step": 35461 + }, + { + "epoch": 2.4607591423218373, + "grad_norm": 1.7263246774673462, + "learning_rate": 6.811887336580071e-05, + "loss": 0.7269, + "step": 35462 + }, + { + "epoch": 2.460828533758934, + "grad_norm": 1.7501616477966309, + "learning_rate": 6.811354799001796e-05, + "loss": 0.5712, + "step": 35463 + }, + { + "epoch": 2.460897925196031, + "grad_norm": 1.170009970664978, + "learning_rate": 6.810822271489707e-05, + "loss": 0.4869, + "step": 35464 + }, + { + "epoch": 2.4609673166331274, + "grad_norm": 1.849332571029663, + "learning_rate": 6.81028975404548e-05, + "loss": 0.4682, + "step": 35465 + }, + { + "epoch": 2.4610367080702242, + "grad_norm": 1.261147141456604, + "learning_rate": 6.809757246670797e-05, + "loss": 0.4232, + "step": 35466 + }, + { + "epoch": 2.4611060995073206, + "grad_norm": 1.675736665725708, + "learning_rate": 6.809224749367341e-05, + "loss": 0.4578, + "step": 35467 + }, + { + "epoch": 2.4611754909444175, + "grad_norm": 1.3295741081237793, + "learning_rate": 6.808692262136795e-05, + "loss": 0.3423, + "step": 35468 + }, + { + "epoch": 2.4612448823815143, + "grad_norm": 0.9429380893707275, + "learning_rate": 6.808159784980836e-05, + "loss": 0.3964, + "step": 35469 + }, + { + "epoch": 2.4613142738186107, + "grad_norm": 1.7915607690811157, + "learning_rate": 6.807627317901144e-05, + "loss": 0.713, + "step": 35470 + }, + { + "epoch": 2.4613836652557075, + "grad_norm": 1.068705439567566, + "learning_rate": 6.807094860899403e-05, + "loss": 0.2636, + "step": 35471 + }, + { + "epoch": 2.461453056692804, + "grad_norm": 1.26560640335083, + "learning_rate": 6.806562413977292e-05, + "loss": 0.4125, + "step": 35472 + }, + { + "epoch": 2.4615224481299007, + "grad_norm": 1.1848429441452026, + "learning_rate": 6.806029977136494e-05, + "loss": 0.3268, + "step": 35473 + }, + { + "epoch": 2.4615918395669976, + "grad_norm": 1.0948656797409058, + "learning_rate": 6.805497550378692e-05, + "loss": 0.4246, + "step": 35474 + }, + { + "epoch": 2.461661231004094, + "grad_norm": 1.1460165977478027, + "learning_rate": 6.80496513370556e-05, + "loss": 0.4165, + "step": 35475 + }, + { + "epoch": 2.461730622441191, + "grad_norm": 1.139109492301941, + "learning_rate": 6.804432727118784e-05, + "loss": 0.5975, + "step": 35476 + }, + { + "epoch": 2.461800013878287, + "grad_norm": 1.5033860206604004, + "learning_rate": 6.803900330620041e-05, + "loss": 0.4061, + "step": 35477 + }, + { + "epoch": 2.461869405315384, + "grad_norm": 1.22842276096344, + "learning_rate": 6.803367944211015e-05, + "loss": 0.4272, + "step": 35478 + }, + { + "epoch": 2.461938796752481, + "grad_norm": 1.2501133680343628, + "learning_rate": 6.802835567893386e-05, + "loss": 0.375, + "step": 35479 + }, + { + "epoch": 2.4620081881895772, + "grad_norm": 1.1420997381210327, + "learning_rate": 6.802303201668834e-05, + "loss": 0.3645, + "step": 35480 + }, + { + "epoch": 2.462077579626674, + "grad_norm": 1.2942928075790405, + "learning_rate": 6.801770845539039e-05, + "loss": 0.4208, + "step": 35481 + }, + { + "epoch": 2.462146971063771, + "grad_norm": 1.0101361274719238, + "learning_rate": 6.801238499505683e-05, + "loss": 0.2125, + "step": 35482 + }, + { + "epoch": 2.4622163625008673, + "grad_norm": 1.3203682899475098, + "learning_rate": 6.800706163570448e-05, + "loss": 0.452, + "step": 35483 + }, + { + "epoch": 2.462285753937964, + "grad_norm": 1.4298996925354004, + "learning_rate": 6.800173837735008e-05, + "loss": 0.4344, + "step": 35484 + }, + { + "epoch": 2.462355145375061, + "grad_norm": 1.094289779663086, + "learning_rate": 6.79964152200105e-05, + "loss": 0.3718, + "step": 35485 + }, + { + "epoch": 2.4624245368121573, + "grad_norm": 1.2694177627563477, + "learning_rate": 6.799109216370253e-05, + "loss": 0.4378, + "step": 35486 + }, + { + "epoch": 2.462493928249254, + "grad_norm": 1.1599289178848267, + "learning_rate": 6.798576920844295e-05, + "loss": 0.4439, + "step": 35487 + }, + { + "epoch": 2.4625633196863506, + "grad_norm": 1.5284059047698975, + "learning_rate": 6.798044635424864e-05, + "loss": 0.4668, + "step": 35488 + }, + { + "epoch": 2.4626327111234474, + "grad_norm": 1.4146462678909302, + "learning_rate": 6.79751236011363e-05, + "loss": 0.4213, + "step": 35489 + }, + { + "epoch": 2.4627021025605442, + "grad_norm": 1.257910966873169, + "learning_rate": 6.796980094912278e-05, + "loss": 0.4099, + "step": 35490 + }, + { + "epoch": 2.4627714939976406, + "grad_norm": 1.4371219873428345, + "learning_rate": 6.796447839822486e-05, + "loss": 0.4029, + "step": 35491 + }, + { + "epoch": 2.4628408854347374, + "grad_norm": 1.1570050716400146, + "learning_rate": 6.795915594845939e-05, + "loss": 0.3108, + "step": 35492 + }, + { + "epoch": 2.462910276871834, + "grad_norm": 1.8607772588729858, + "learning_rate": 6.795383359984314e-05, + "loss": 0.4814, + "step": 35493 + }, + { + "epoch": 2.4629796683089307, + "grad_norm": 1.3734018802642822, + "learning_rate": 6.794851135239292e-05, + "loss": 0.3197, + "step": 35494 + }, + { + "epoch": 2.4630490597460275, + "grad_norm": 2.066091775894165, + "learning_rate": 6.794318920612556e-05, + "loss": 0.3628, + "step": 35495 + }, + { + "epoch": 2.463118451183124, + "grad_norm": 1.5444391965866089, + "learning_rate": 6.793786716105781e-05, + "loss": 0.4363, + "step": 35496 + }, + { + "epoch": 2.4631878426202207, + "grad_norm": 1.6268680095672607, + "learning_rate": 6.793254521720649e-05, + "loss": 0.4687, + "step": 35497 + }, + { + "epoch": 2.463257234057317, + "grad_norm": 1.6559512615203857, + "learning_rate": 6.79272233745884e-05, + "loss": 0.4727, + "step": 35498 + }, + { + "epoch": 2.463326625494414, + "grad_norm": 1.0468467473983765, + "learning_rate": 6.792190163322036e-05, + "loss": 0.4108, + "step": 35499 + }, + { + "epoch": 2.463396016931511, + "grad_norm": 0.9096129536628723, + "learning_rate": 6.791657999311916e-05, + "loss": 0.3417, + "step": 35500 + }, + { + "epoch": 2.463465408368607, + "grad_norm": 1.4412522315979004, + "learning_rate": 6.791125845430156e-05, + "loss": 0.3107, + "step": 35501 + }, + { + "epoch": 2.463534799805704, + "grad_norm": 1.262292504310608, + "learning_rate": 6.790593701678444e-05, + "loss": 0.4342, + "step": 35502 + }, + { + "epoch": 2.4636041912428004, + "grad_norm": 1.623360514640808, + "learning_rate": 6.790061568058453e-05, + "loss": 0.355, + "step": 35503 + }, + { + "epoch": 2.4636735826798972, + "grad_norm": 1.1632018089294434, + "learning_rate": 6.789529444571865e-05, + "loss": 0.3654, + "step": 35504 + }, + { + "epoch": 2.463742974116994, + "grad_norm": 1.33834707736969, + "learning_rate": 6.788997331220359e-05, + "loss": 0.406, + "step": 35505 + }, + { + "epoch": 2.4638123655540904, + "grad_norm": 1.5278657674789429, + "learning_rate": 6.788465228005616e-05, + "loss": 0.4059, + "step": 35506 + }, + { + "epoch": 2.4638817569911873, + "grad_norm": 1.2478286027908325, + "learning_rate": 6.787933134929317e-05, + "loss": 0.4546, + "step": 35507 + }, + { + "epoch": 2.463951148428284, + "grad_norm": 2.2638344764709473, + "learning_rate": 6.787401051993138e-05, + "loss": 0.712, + "step": 35508 + }, + { + "epoch": 2.4640205398653805, + "grad_norm": 1.1798216104507446, + "learning_rate": 6.786868979198768e-05, + "loss": 0.4125, + "step": 35509 + }, + { + "epoch": 2.4640899313024773, + "grad_norm": 1.4147744178771973, + "learning_rate": 6.786336916547876e-05, + "loss": 0.4198, + "step": 35510 + }, + { + "epoch": 2.464159322739574, + "grad_norm": 1.0484931468963623, + "learning_rate": 6.785804864042143e-05, + "loss": 0.3913, + "step": 35511 + }, + { + "epoch": 2.4642287141766706, + "grad_norm": 1.2443804740905762, + "learning_rate": 6.785272821683251e-05, + "loss": 0.4492, + "step": 35512 + }, + { + "epoch": 2.4642981056137674, + "grad_norm": 1.05368173122406, + "learning_rate": 6.78474078947288e-05, + "loss": 0.2444, + "step": 35513 + }, + { + "epoch": 2.4643674970508638, + "grad_norm": 1.7423714399337769, + "learning_rate": 6.784208767412713e-05, + "loss": 0.4039, + "step": 35514 + }, + { + "epoch": 2.4644368884879606, + "grad_norm": 1.4503281116485596, + "learning_rate": 6.783676755504423e-05, + "loss": 0.4894, + "step": 35515 + }, + { + "epoch": 2.4645062799250574, + "grad_norm": 0.9809496402740479, + "learning_rate": 6.783144753749695e-05, + "loss": 0.2576, + "step": 35516 + }, + { + "epoch": 2.464575671362154, + "grad_norm": 1.5998152494430542, + "learning_rate": 6.782612762150203e-05, + "loss": 0.5757, + "step": 35517 + }, + { + "epoch": 2.4646450627992507, + "grad_norm": 1.0683485269546509, + "learning_rate": 6.782080780707631e-05, + "loss": 0.3457, + "step": 35518 + }, + { + "epoch": 2.464714454236347, + "grad_norm": 1.0679569244384766, + "learning_rate": 6.781548809423655e-05, + "loss": 0.393, + "step": 35519 + }, + { + "epoch": 2.464783845673444, + "grad_norm": 1.729323148727417, + "learning_rate": 6.781016848299954e-05, + "loss": 0.6469, + "step": 35520 + }, + { + "epoch": 2.4648532371105407, + "grad_norm": 0.9702744483947754, + "learning_rate": 6.780484897338213e-05, + "loss": 0.3263, + "step": 35521 + }, + { + "epoch": 2.464922628547637, + "grad_norm": 1.2111490964889526, + "learning_rate": 6.779952956540106e-05, + "loss": 0.3633, + "step": 35522 + }, + { + "epoch": 2.464992019984734, + "grad_norm": 1.515305519104004, + "learning_rate": 6.779421025907318e-05, + "loss": 0.3231, + "step": 35523 + }, + { + "epoch": 2.4650614114218303, + "grad_norm": 1.3186274766921997, + "learning_rate": 6.77888910544152e-05, + "loss": 0.3336, + "step": 35524 + }, + { + "epoch": 2.465130802858927, + "grad_norm": 1.1636037826538086, + "learning_rate": 6.778357195144394e-05, + "loss": 0.3622, + "step": 35525 + }, + { + "epoch": 2.465200194296024, + "grad_norm": 1.3541181087493896, + "learning_rate": 6.777825295017623e-05, + "loss": 0.4478, + "step": 35526 + }, + { + "epoch": 2.4652695857331204, + "grad_norm": 1.5091415643692017, + "learning_rate": 6.777293405062881e-05, + "loss": 0.4456, + "step": 35527 + }, + { + "epoch": 2.465338977170217, + "grad_norm": 1.4881788492202759, + "learning_rate": 6.776761525281852e-05, + "loss": 0.543, + "step": 35528 + }, + { + "epoch": 2.465408368607314, + "grad_norm": 1.1764243841171265, + "learning_rate": 6.776229655676213e-05, + "loss": 0.3436, + "step": 35529 + }, + { + "epoch": 2.4654777600444104, + "grad_norm": 1.478342890739441, + "learning_rate": 6.775697796247647e-05, + "loss": 0.4041, + "step": 35530 + }, + { + "epoch": 2.4655471514815073, + "grad_norm": 1.77183198928833, + "learning_rate": 6.775165946997823e-05, + "loss": 0.5321, + "step": 35531 + }, + { + "epoch": 2.4656165429186037, + "grad_norm": 0.9907779097557068, + "learning_rate": 6.774634107928427e-05, + "loss": 0.2988, + "step": 35532 + }, + { + "epoch": 2.4656859343557005, + "grad_norm": 1.336350679397583, + "learning_rate": 6.774102279041136e-05, + "loss": 0.4467, + "step": 35533 + }, + { + "epoch": 2.4657553257927973, + "grad_norm": 1.1093261241912842, + "learning_rate": 6.773570460337635e-05, + "loss": 0.3908, + "step": 35534 + }, + { + "epoch": 2.4658247172298937, + "grad_norm": 1.4698327779769897, + "learning_rate": 6.773038651819593e-05, + "loss": 0.5024, + "step": 35535 + }, + { + "epoch": 2.4658941086669905, + "grad_norm": 1.38720703125, + "learning_rate": 6.772506853488693e-05, + "loss": 0.5218, + "step": 35536 + }, + { + "epoch": 2.4659635001040874, + "grad_norm": 0.9441705346107483, + "learning_rate": 6.77197506534662e-05, + "loss": 0.316, + "step": 35537 + }, + { + "epoch": 2.4660328915411838, + "grad_norm": 1.4043679237365723, + "learning_rate": 6.771443287395043e-05, + "loss": 0.5118, + "step": 35538 + }, + { + "epoch": 2.4661022829782806, + "grad_norm": 1.6605583429336548, + "learning_rate": 6.770911519635645e-05, + "loss": 0.7283, + "step": 35539 + }, + { + "epoch": 2.466171674415377, + "grad_norm": 0.9068515300750732, + "learning_rate": 6.770379762070105e-05, + "loss": 0.4203, + "step": 35540 + }, + { + "epoch": 2.466241065852474, + "grad_norm": 1.1745655536651611, + "learning_rate": 6.7698480147001e-05, + "loss": 0.4448, + "step": 35541 + }, + { + "epoch": 2.4663104572895707, + "grad_norm": 1.7191191911697388, + "learning_rate": 6.769316277527311e-05, + "loss": 0.496, + "step": 35542 + }, + { + "epoch": 2.466379848726667, + "grad_norm": 1.1924930810928345, + "learning_rate": 6.768784550553415e-05, + "loss": 0.5842, + "step": 35543 + }, + { + "epoch": 2.466449240163764, + "grad_norm": 1.4348794221878052, + "learning_rate": 6.768252833780098e-05, + "loss": 0.4592, + "step": 35544 + }, + { + "epoch": 2.4665186316008603, + "grad_norm": 0.9953605532646179, + "learning_rate": 6.767721127209025e-05, + "loss": 0.3327, + "step": 35545 + }, + { + "epoch": 2.466588023037957, + "grad_norm": 0.8760862946510315, + "learning_rate": 6.767189430841882e-05, + "loss": 0.2816, + "step": 35546 + }, + { + "epoch": 2.466657414475054, + "grad_norm": 1.2282097339630127, + "learning_rate": 6.766657744680347e-05, + "loss": 0.4558, + "step": 35547 + }, + { + "epoch": 2.4667268059121503, + "grad_norm": 1.1468806266784668, + "learning_rate": 6.766126068726098e-05, + "loss": 0.4409, + "step": 35548 + }, + { + "epoch": 2.466796197349247, + "grad_norm": 1.2322582006454468, + "learning_rate": 6.765594402980816e-05, + "loss": 0.5926, + "step": 35549 + }, + { + "epoch": 2.4668655887863435, + "grad_norm": 1.3140383958816528, + "learning_rate": 6.765062747446178e-05, + "loss": 0.332, + "step": 35550 + }, + { + "epoch": 2.4669349802234404, + "grad_norm": 0.8850905299186707, + "learning_rate": 6.764531102123858e-05, + "loss": 0.3526, + "step": 35551 + }, + { + "epoch": 2.467004371660537, + "grad_norm": 0.9112408757209778, + "learning_rate": 6.76399946701554e-05, + "loss": 0.4171, + "step": 35552 + }, + { + "epoch": 2.4670737630976336, + "grad_norm": 1.4158216714859009, + "learning_rate": 6.763467842122901e-05, + "loss": 0.4795, + "step": 35553 + }, + { + "epoch": 2.4671431545347304, + "grad_norm": 1.3656953573226929, + "learning_rate": 6.762936227447616e-05, + "loss": 0.3446, + "step": 35554 + }, + { + "epoch": 2.4672125459718273, + "grad_norm": 1.1176432371139526, + "learning_rate": 6.762404622991367e-05, + "loss": 0.3787, + "step": 35555 + }, + { + "epoch": 2.4672819374089237, + "grad_norm": 1.4563701152801514, + "learning_rate": 6.761873028755829e-05, + "loss": 0.3361, + "step": 35556 + }, + { + "epoch": 2.4673513288460205, + "grad_norm": 0.8618559837341309, + "learning_rate": 6.761341444742688e-05, + "loss": 0.3824, + "step": 35557 + }, + { + "epoch": 2.467420720283117, + "grad_norm": 1.229587435722351, + "learning_rate": 6.760809870953613e-05, + "loss": 0.3941, + "step": 35558 + }, + { + "epoch": 2.4674901117202137, + "grad_norm": 1.37052583694458, + "learning_rate": 6.760278307390285e-05, + "loss": 0.3248, + "step": 35559 + }, + { + "epoch": 2.4675595031573105, + "grad_norm": 1.1542158126831055, + "learning_rate": 6.759746754054381e-05, + "loss": 0.4107, + "step": 35560 + }, + { + "epoch": 2.467628894594407, + "grad_norm": 1.071714997291565, + "learning_rate": 6.75921521094758e-05, + "loss": 0.3651, + "step": 35561 + }, + { + "epoch": 2.4676982860315038, + "grad_norm": 1.2113322019577026, + "learning_rate": 6.758683678071563e-05, + "loss": 0.4489, + "step": 35562 + }, + { + "epoch": 2.4677676774686006, + "grad_norm": 0.8917766213417053, + "learning_rate": 6.758152155428003e-05, + "loss": 0.2974, + "step": 35563 + }, + { + "epoch": 2.467837068905697, + "grad_norm": 1.159592866897583, + "learning_rate": 6.757620643018587e-05, + "loss": 0.3596, + "step": 35564 + }, + { + "epoch": 2.467906460342794, + "grad_norm": 1.1434545516967773, + "learning_rate": 6.757089140844981e-05, + "loss": 0.4269, + "step": 35565 + }, + { + "epoch": 2.46797585177989, + "grad_norm": 0.9948063492774963, + "learning_rate": 6.75655764890887e-05, + "loss": 0.353, + "step": 35566 + }, + { + "epoch": 2.468045243216987, + "grad_norm": 1.2296526432037354, + "learning_rate": 6.756026167211929e-05, + "loss": 0.4117, + "step": 35567 + }, + { + "epoch": 2.468114634654084, + "grad_norm": 1.3777989149093628, + "learning_rate": 6.755494695755839e-05, + "loss": 0.4138, + "step": 35568 + }, + { + "epoch": 2.4681840260911803, + "grad_norm": 1.317258596420288, + "learning_rate": 6.754963234542273e-05, + "loss": 0.7071, + "step": 35569 + }, + { + "epoch": 2.468253417528277, + "grad_norm": 1.262938141822815, + "learning_rate": 6.754431783572912e-05, + "loss": 0.3829, + "step": 35570 + }, + { + "epoch": 2.4683228089653735, + "grad_norm": 1.0715854167938232, + "learning_rate": 6.753900342849436e-05, + "loss": 0.3467, + "step": 35571 + }, + { + "epoch": 2.4683922004024703, + "grad_norm": 1.1071792840957642, + "learning_rate": 6.75336891237352e-05, + "loss": 0.3463, + "step": 35572 + }, + { + "epoch": 2.468461591839567, + "grad_norm": 1.0921175479888916, + "learning_rate": 6.752837492146839e-05, + "loss": 0.314, + "step": 35573 + }, + { + "epoch": 2.4685309832766635, + "grad_norm": 1.1944060325622559, + "learning_rate": 6.752306082171074e-05, + "loss": 0.3149, + "step": 35574 + }, + { + "epoch": 2.4686003747137604, + "grad_norm": 1.9109402894973755, + "learning_rate": 6.7517746824479e-05, + "loss": 0.4973, + "step": 35575 + }, + { + "epoch": 2.4686697661508568, + "grad_norm": 1.2619123458862305, + "learning_rate": 6.751243292979e-05, + "loss": 0.5698, + "step": 35576 + }, + { + "epoch": 2.4687391575879536, + "grad_norm": 1.2889575958251953, + "learning_rate": 6.750711913766046e-05, + "loss": 0.3502, + "step": 35577 + }, + { + "epoch": 2.4688085490250504, + "grad_norm": 1.243028998374939, + "learning_rate": 6.750180544810721e-05, + "loss": 0.3947, + "step": 35578 + }, + { + "epoch": 2.468877940462147, + "grad_norm": 0.9489595890045166, + "learning_rate": 6.749649186114695e-05, + "loss": 0.3759, + "step": 35579 + }, + { + "epoch": 2.4689473318992436, + "grad_norm": 1.3699331283569336, + "learning_rate": 6.749117837679651e-05, + "loss": 0.5395, + "step": 35580 + }, + { + "epoch": 2.4690167233363405, + "grad_norm": 1.1665031909942627, + "learning_rate": 6.748586499507263e-05, + "loss": 0.3317, + "step": 35581 + }, + { + "epoch": 2.469086114773437, + "grad_norm": 1.0337567329406738, + "learning_rate": 6.74805517159921e-05, + "loss": 0.3594, + "step": 35582 + }, + { + "epoch": 2.4691555062105337, + "grad_norm": 1.3650423288345337, + "learning_rate": 6.747523853957171e-05, + "loss": 0.455, + "step": 35583 + }, + { + "epoch": 2.46922489764763, + "grad_norm": 1.331725001335144, + "learning_rate": 6.746992546582823e-05, + "loss": 0.3618, + "step": 35584 + }, + { + "epoch": 2.469294289084727, + "grad_norm": 1.8274413347244263, + "learning_rate": 6.746461249477842e-05, + "loss": 0.4101, + "step": 35585 + }, + { + "epoch": 2.4693636805218238, + "grad_norm": 1.2191448211669922, + "learning_rate": 6.745929962643905e-05, + "loss": 0.372, + "step": 35586 + }, + { + "epoch": 2.46943307195892, + "grad_norm": 1.294692873954773, + "learning_rate": 6.745398686082687e-05, + "loss": 0.5022, + "step": 35587 + }, + { + "epoch": 2.469502463396017, + "grad_norm": 1.4259216785430908, + "learning_rate": 6.744867419795871e-05, + "loss": 0.5696, + "step": 35588 + }, + { + "epoch": 2.469571854833114, + "grad_norm": 1.0278277397155762, + "learning_rate": 6.74433616378513e-05, + "loss": 0.3788, + "step": 35589 + }, + { + "epoch": 2.46964124627021, + "grad_norm": 1.288512110710144, + "learning_rate": 6.74380491805214e-05, + "loss": 0.4912, + "step": 35590 + }, + { + "epoch": 2.469710637707307, + "grad_norm": 1.307775855064392, + "learning_rate": 6.74327368259858e-05, + "loss": 0.3922, + "step": 35591 + }, + { + "epoch": 2.4697800291444034, + "grad_norm": 1.0132389068603516, + "learning_rate": 6.742742457426133e-05, + "loss": 0.389, + "step": 35592 + }, + { + "epoch": 2.4698494205815003, + "grad_norm": 1.5513648986816406, + "learning_rate": 6.742211242536466e-05, + "loss": 0.4182, + "step": 35593 + }, + { + "epoch": 2.469918812018597, + "grad_norm": 1.2038459777832031, + "learning_rate": 6.74168003793126e-05, + "loss": 0.5119, + "step": 35594 + }, + { + "epoch": 2.4699882034556935, + "grad_norm": 1.2829279899597168, + "learning_rate": 6.74114884361219e-05, + "loss": 0.5051, + "step": 35595 + }, + { + "epoch": 2.4700575948927903, + "grad_norm": 1.282313585281372, + "learning_rate": 6.740617659580937e-05, + "loss": 0.5248, + "step": 35596 + }, + { + "epoch": 2.4701269863298867, + "grad_norm": 1.1134605407714844, + "learning_rate": 6.740086485839176e-05, + "loss": 0.4447, + "step": 35597 + }, + { + "epoch": 2.4701963777669835, + "grad_norm": 1.4267163276672363, + "learning_rate": 6.739555322388582e-05, + "loss": 0.4786, + "step": 35598 + }, + { + "epoch": 2.4702657692040804, + "grad_norm": 1.3184911012649536, + "learning_rate": 6.739024169230838e-05, + "loss": 0.4124, + "step": 35599 + }, + { + "epoch": 2.4703351606411768, + "grad_norm": 1.2321292161941528, + "learning_rate": 6.738493026367614e-05, + "loss": 0.3919, + "step": 35600 + }, + { + "epoch": 2.4704045520782736, + "grad_norm": 1.0190221071243286, + "learning_rate": 6.737961893800589e-05, + "loss": 0.3154, + "step": 35601 + }, + { + "epoch": 2.47047394351537, + "grad_norm": 1.120162844657898, + "learning_rate": 6.737430771531439e-05, + "loss": 0.4049, + "step": 35602 + }, + { + "epoch": 2.470543334952467, + "grad_norm": 1.1367499828338623, + "learning_rate": 6.736899659561843e-05, + "loss": 0.3761, + "step": 35603 + }, + { + "epoch": 2.4706127263895636, + "grad_norm": 2.0963308811187744, + "learning_rate": 6.736368557893475e-05, + "loss": 0.5918, + "step": 35604 + }, + { + "epoch": 2.47068211782666, + "grad_norm": 0.749457061290741, + "learning_rate": 6.735837466528013e-05, + "loss": 0.2715, + "step": 35605 + }, + { + "epoch": 2.470751509263757, + "grad_norm": 2.394287586212158, + "learning_rate": 6.735306385467134e-05, + "loss": 0.5437, + "step": 35606 + }, + { + "epoch": 2.4708209007008537, + "grad_norm": 1.1320726871490479, + "learning_rate": 6.734775314712515e-05, + "loss": 0.3653, + "step": 35607 + }, + { + "epoch": 2.47089029213795, + "grad_norm": 0.8920925259590149, + "learning_rate": 6.734244254265829e-05, + "loss": 0.3813, + "step": 35608 + }, + { + "epoch": 2.470959683575047, + "grad_norm": 1.0974377393722534, + "learning_rate": 6.733713204128755e-05, + "loss": 0.3245, + "step": 35609 + }, + { + "epoch": 2.4710290750121437, + "grad_norm": 1.233246922492981, + "learning_rate": 6.733182164302972e-05, + "loss": 0.3137, + "step": 35610 + }, + { + "epoch": 2.47109846644924, + "grad_norm": 0.9543541073799133, + "learning_rate": 6.732651134790151e-05, + "loss": 0.2626, + "step": 35611 + }, + { + "epoch": 2.471167857886337, + "grad_norm": 1.4444142580032349, + "learning_rate": 6.732120115591971e-05, + "loss": 0.5255, + "step": 35612 + }, + { + "epoch": 2.4712372493234334, + "grad_norm": 1.0031393766403198, + "learning_rate": 6.731589106710114e-05, + "loss": 0.3099, + "step": 35613 + }, + { + "epoch": 2.47130664076053, + "grad_norm": 1.397207498550415, + "learning_rate": 6.731058108146247e-05, + "loss": 0.3801, + "step": 35614 + }, + { + "epoch": 2.471376032197627, + "grad_norm": 0.9781216382980347, + "learning_rate": 6.730527119902051e-05, + "loss": 0.3183, + "step": 35615 + }, + { + "epoch": 2.4714454236347234, + "grad_norm": 0.9308521747589111, + "learning_rate": 6.729996141979198e-05, + "loss": 0.3357, + "step": 35616 + }, + { + "epoch": 2.4715148150718202, + "grad_norm": 1.2308143377304077, + "learning_rate": 6.729465174379371e-05, + "loss": 0.3016, + "step": 35617 + }, + { + "epoch": 2.4715842065089166, + "grad_norm": 1.0260156393051147, + "learning_rate": 6.728934217104242e-05, + "loss": 0.3195, + "step": 35618 + }, + { + "epoch": 2.4716535979460135, + "grad_norm": 1.3119159936904907, + "learning_rate": 6.728403270155491e-05, + "loss": 0.355, + "step": 35619 + }, + { + "epoch": 2.4717229893831103, + "grad_norm": 1.152513027191162, + "learning_rate": 6.72787233353479e-05, + "loss": 0.3707, + "step": 35620 + }, + { + "epoch": 2.4717923808202067, + "grad_norm": 1.5089510679244995, + "learning_rate": 6.727341407243815e-05, + "loss": 0.4593, + "step": 35621 + }, + { + "epoch": 2.4718617722573035, + "grad_norm": 1.340259313583374, + "learning_rate": 6.726810491284247e-05, + "loss": 0.2915, + "step": 35622 + }, + { + "epoch": 2.4719311636944, + "grad_norm": 0.9925808906555176, + "learning_rate": 6.726279585657756e-05, + "loss": 0.3174, + "step": 35623 + }, + { + "epoch": 2.4720005551314967, + "grad_norm": 1.9098405838012695, + "learning_rate": 6.725748690366018e-05, + "loss": 0.2885, + "step": 35624 + }, + { + "epoch": 2.4720699465685936, + "grad_norm": 0.9553334712982178, + "learning_rate": 6.725217805410713e-05, + "loss": 0.2577, + "step": 35625 + }, + { + "epoch": 2.47213933800569, + "grad_norm": 2.110452175140381, + "learning_rate": 6.72468693079352e-05, + "loss": 0.6391, + "step": 35626 + }, + { + "epoch": 2.472208729442787, + "grad_norm": 1.3498343229293823, + "learning_rate": 6.724156066516107e-05, + "loss": 0.4045, + "step": 35627 + }, + { + "epoch": 2.472278120879883, + "grad_norm": 1.2721023559570312, + "learning_rate": 6.723625212580151e-05, + "loss": 0.4691, + "step": 35628 + }, + { + "epoch": 2.47234751231698, + "grad_norm": 1.8094022274017334, + "learning_rate": 6.723094368987332e-05, + "loss": 0.4229, + "step": 35629 + }, + { + "epoch": 2.472416903754077, + "grad_norm": 1.006567358970642, + "learning_rate": 6.722563535739322e-05, + "loss": 0.2753, + "step": 35630 + }, + { + "epoch": 2.4724862951911732, + "grad_norm": 1.938413143157959, + "learning_rate": 6.7220327128378e-05, + "loss": 0.3438, + "step": 35631 + }, + { + "epoch": 2.47255568662827, + "grad_norm": 0.9666098952293396, + "learning_rate": 6.721501900284441e-05, + "loss": 0.2657, + "step": 35632 + }, + { + "epoch": 2.472625078065367, + "grad_norm": 1.1444447040557861, + "learning_rate": 6.720971098080923e-05, + "loss": 0.3431, + "step": 35633 + }, + { + "epoch": 2.4726944695024633, + "grad_norm": 1.222625732421875, + "learning_rate": 6.720440306228914e-05, + "loss": 0.3711, + "step": 35634 + }, + { + "epoch": 2.47276386093956, + "grad_norm": 1.2998948097229004, + "learning_rate": 6.719909524730096e-05, + "loss": 0.4039, + "step": 35635 + }, + { + "epoch": 2.472833252376657, + "grad_norm": 1.1803181171417236, + "learning_rate": 6.719378753586143e-05, + "loss": 0.4674, + "step": 35636 + }, + { + "epoch": 2.4729026438137534, + "grad_norm": 1.7917265892028809, + "learning_rate": 6.718847992798732e-05, + "loss": 0.382, + "step": 35637 + }, + { + "epoch": 2.47297203525085, + "grad_norm": 1.4899420738220215, + "learning_rate": 6.718317242369537e-05, + "loss": 0.5105, + "step": 35638 + }, + { + "epoch": 2.4730414266879466, + "grad_norm": 1.469110369682312, + "learning_rate": 6.717786502300232e-05, + "loss": 0.4983, + "step": 35639 + }, + { + "epoch": 2.4731108181250434, + "grad_norm": 0.7808083295822144, + "learning_rate": 6.717255772592496e-05, + "loss": 0.3107, + "step": 35640 + }, + { + "epoch": 2.4731802095621402, + "grad_norm": 1.2617954015731812, + "learning_rate": 6.716725053248e-05, + "loss": 0.2993, + "step": 35641 + }, + { + "epoch": 2.4732496009992366, + "grad_norm": 1.7675344944000244, + "learning_rate": 6.716194344268426e-05, + "loss": 0.4132, + "step": 35642 + }, + { + "epoch": 2.4733189924363335, + "grad_norm": 1.6974081993103027, + "learning_rate": 6.715663645655443e-05, + "loss": 0.5614, + "step": 35643 + }, + { + "epoch": 2.47338838387343, + "grad_norm": 1.778827428817749, + "learning_rate": 6.715132957410728e-05, + "loss": 0.3989, + "step": 35644 + }, + { + "epoch": 2.4734577753105267, + "grad_norm": 1.554471731185913, + "learning_rate": 6.714602279535958e-05, + "loss": 0.4725, + "step": 35645 + }, + { + "epoch": 2.4735271667476235, + "grad_norm": 1.7770875692367554, + "learning_rate": 6.714071612032806e-05, + "loss": 0.6342, + "step": 35646 + }, + { + "epoch": 2.47359655818472, + "grad_norm": 1.4915072917938232, + "learning_rate": 6.713540954902955e-05, + "loss": 0.4947, + "step": 35647 + }, + { + "epoch": 2.4736659496218167, + "grad_norm": 1.2664164304733276, + "learning_rate": 6.713010308148069e-05, + "loss": 0.5107, + "step": 35648 + }, + { + "epoch": 2.473735341058913, + "grad_norm": 1.2910808324813843, + "learning_rate": 6.712479671769827e-05, + "loss": 0.4518, + "step": 35649 + }, + { + "epoch": 2.47380473249601, + "grad_norm": 1.3838263750076294, + "learning_rate": 6.711949045769905e-05, + "loss": 0.3584, + "step": 35650 + }, + { + "epoch": 2.473874123933107, + "grad_norm": 1.4883601665496826, + "learning_rate": 6.711418430149981e-05, + "loss": 0.5522, + "step": 35651 + }, + { + "epoch": 2.473943515370203, + "grad_norm": 0.9478099346160889, + "learning_rate": 6.710887824911725e-05, + "loss": 0.3406, + "step": 35652 + }, + { + "epoch": 2.4740129068073, + "grad_norm": 1.0446425676345825, + "learning_rate": 6.710357230056815e-05, + "loss": 0.4083, + "step": 35653 + }, + { + "epoch": 2.4740822982443964, + "grad_norm": 1.1618419885635376, + "learning_rate": 6.709826645586929e-05, + "loss": 0.4215, + "step": 35654 + }, + { + "epoch": 2.4741516896814932, + "grad_norm": 1.204331636428833, + "learning_rate": 6.709296071503736e-05, + "loss": 0.3457, + "step": 35655 + }, + { + "epoch": 2.47422108111859, + "grad_norm": 1.2393016815185547, + "learning_rate": 6.708765507808913e-05, + "loss": 0.4806, + "step": 35656 + }, + { + "epoch": 2.4742904725556865, + "grad_norm": 1.428592562675476, + "learning_rate": 6.708234954504138e-05, + "loss": 0.463, + "step": 35657 + }, + { + "epoch": 2.4743598639927833, + "grad_norm": 1.0296030044555664, + "learning_rate": 6.70770441159108e-05, + "loss": 0.2117, + "step": 35658 + }, + { + "epoch": 2.47442925542988, + "grad_norm": 1.2474759817123413, + "learning_rate": 6.707173879071419e-05, + "loss": 0.5, + "step": 35659 + }, + { + "epoch": 2.4744986468669765, + "grad_norm": 1.203109622001648, + "learning_rate": 6.706643356946826e-05, + "loss": 0.4111, + "step": 35660 + }, + { + "epoch": 2.4745680383040733, + "grad_norm": 1.2106884717941284, + "learning_rate": 6.706112845218983e-05, + "loss": 0.3983, + "step": 35661 + }, + { + "epoch": 2.47463742974117, + "grad_norm": 1.3593326807022095, + "learning_rate": 6.705582343889556e-05, + "loss": 0.4989, + "step": 35662 + }, + { + "epoch": 2.4747068211782666, + "grad_norm": 0.9644030928611755, + "learning_rate": 6.705051852960222e-05, + "loss": 0.4283, + "step": 35663 + }, + { + "epoch": 2.4747762126153634, + "grad_norm": 1.3089187145233154, + "learning_rate": 6.704521372432659e-05, + "loss": 0.3299, + "step": 35664 + }, + { + "epoch": 2.47484560405246, + "grad_norm": 1.1255896091461182, + "learning_rate": 6.703990902308537e-05, + "loss": 0.3033, + "step": 35665 + }, + { + "epoch": 2.4749149954895566, + "grad_norm": 1.5781718492507935, + "learning_rate": 6.703460442589536e-05, + "loss": 0.4989, + "step": 35666 + }, + { + "epoch": 2.4749843869266535, + "grad_norm": 0.8853803277015686, + "learning_rate": 6.702929993277326e-05, + "loss": 0.2744, + "step": 35667 + }, + { + "epoch": 2.47505377836375, + "grad_norm": 2.679516553878784, + "learning_rate": 6.702399554373588e-05, + "loss": 0.3996, + "step": 35668 + }, + { + "epoch": 2.4751231698008467, + "grad_norm": 0.8781343698501587, + "learning_rate": 6.701869125879988e-05, + "loss": 0.3317, + "step": 35669 + }, + { + "epoch": 2.475192561237943, + "grad_norm": 1.5258681774139404, + "learning_rate": 6.701338707798204e-05, + "loss": 0.235, + "step": 35670 + }, + { + "epoch": 2.47526195267504, + "grad_norm": 1.2753503322601318, + "learning_rate": 6.700808300129913e-05, + "loss": 0.445, + "step": 35671 + }, + { + "epoch": 2.4753313441121367, + "grad_norm": 1.2292115688323975, + "learning_rate": 6.700277902876787e-05, + "loss": 0.5426, + "step": 35672 + }, + { + "epoch": 2.475400735549233, + "grad_norm": 1.0514780282974243, + "learning_rate": 6.6997475160405e-05, + "loss": 0.3842, + "step": 35673 + }, + { + "epoch": 2.47547012698633, + "grad_norm": 0.919509768486023, + "learning_rate": 6.699217139622726e-05, + "loss": 0.3224, + "step": 35674 + }, + { + "epoch": 2.4755395184234263, + "grad_norm": 1.2279982566833496, + "learning_rate": 6.698686773625145e-05, + "loss": 0.3436, + "step": 35675 + }, + { + "epoch": 2.475608909860523, + "grad_norm": 1.143165111541748, + "learning_rate": 6.698156418049425e-05, + "loss": 0.4595, + "step": 35676 + }, + { + "epoch": 2.47567830129762, + "grad_norm": 1.4163926839828491, + "learning_rate": 6.69762607289724e-05, + "loss": 0.3445, + "step": 35677 + }, + { + "epoch": 2.4757476927347164, + "grad_norm": 1.2161409854888916, + "learning_rate": 6.697095738170267e-05, + "loss": 0.3918, + "step": 35678 + }, + { + "epoch": 2.4758170841718132, + "grad_norm": 1.7262475490570068, + "learning_rate": 6.696565413870179e-05, + "loss": 0.4391, + "step": 35679 + }, + { + "epoch": 2.4758864756089096, + "grad_norm": 0.9865689277648926, + "learning_rate": 6.696035099998652e-05, + "loss": 0.3685, + "step": 35680 + }, + { + "epoch": 2.4759558670460065, + "grad_norm": 0.9820435047149658, + "learning_rate": 6.695504796557357e-05, + "loss": 0.4075, + "step": 35681 + }, + { + "epoch": 2.4760252584831033, + "grad_norm": 1.2593172788619995, + "learning_rate": 6.694974503547975e-05, + "loss": 0.4821, + "step": 35682 + }, + { + "epoch": 2.4760946499201997, + "grad_norm": 1.2690906524658203, + "learning_rate": 6.694444220972171e-05, + "loss": 0.4593, + "step": 35683 + }, + { + "epoch": 2.4761640413572965, + "grad_norm": 1.2630302906036377, + "learning_rate": 6.693913948831623e-05, + "loss": 0.4158, + "step": 35684 + }, + { + "epoch": 2.4762334327943933, + "grad_norm": 1.5855566263198853, + "learning_rate": 6.693383687128006e-05, + "loss": 0.6041, + "step": 35685 + }, + { + "epoch": 2.4763028242314897, + "grad_norm": 1.135930061340332, + "learning_rate": 6.692853435862992e-05, + "loss": 0.4931, + "step": 35686 + }, + { + "epoch": 2.4763722156685866, + "grad_norm": 1.8172144889831543, + "learning_rate": 6.692323195038256e-05, + "loss": 0.5357, + "step": 35687 + }, + { + "epoch": 2.4764416071056834, + "grad_norm": 0.75162273645401, + "learning_rate": 6.691792964655475e-05, + "loss": 0.194, + "step": 35688 + }, + { + "epoch": 2.47651099854278, + "grad_norm": 1.1909834146499634, + "learning_rate": 6.691262744716318e-05, + "loss": 0.4551, + "step": 35689 + }, + { + "epoch": 2.4765803899798766, + "grad_norm": 0.9937722682952881, + "learning_rate": 6.690732535222459e-05, + "loss": 0.5394, + "step": 35690 + }, + { + "epoch": 2.476649781416973, + "grad_norm": 1.0810538530349731, + "learning_rate": 6.690202336175574e-05, + "loss": 0.417, + "step": 35691 + }, + { + "epoch": 2.47671917285407, + "grad_norm": 1.358343482017517, + "learning_rate": 6.689672147577339e-05, + "loss": 0.3905, + "step": 35692 + }, + { + "epoch": 2.4767885642911667, + "grad_norm": 0.9699543118476868, + "learning_rate": 6.689141969429422e-05, + "loss": 0.381, + "step": 35693 + }, + { + "epoch": 2.476857955728263, + "grad_norm": 1.2997660636901855, + "learning_rate": 6.6886118017335e-05, + "loss": 0.509, + "step": 35694 + }, + { + "epoch": 2.47692734716536, + "grad_norm": 0.7343630194664001, + "learning_rate": 6.688081644491246e-05, + "loss": 0.2568, + "step": 35695 + }, + { + "epoch": 2.4769967386024563, + "grad_norm": 0.9571326375007629, + "learning_rate": 6.687551497704339e-05, + "loss": 0.2315, + "step": 35696 + }, + { + "epoch": 2.477066130039553, + "grad_norm": 0.7116069793701172, + "learning_rate": 6.687021361374441e-05, + "loss": 0.3575, + "step": 35697 + }, + { + "epoch": 2.47713552147665, + "grad_norm": 1.4723204374313354, + "learning_rate": 6.686491235503235e-05, + "loss": 0.5647, + "step": 35698 + }, + { + "epoch": 2.4772049129137463, + "grad_norm": 1.1249783039093018, + "learning_rate": 6.68596112009239e-05, + "loss": 0.2533, + "step": 35699 + }, + { + "epoch": 2.477274304350843, + "grad_norm": 1.240566372871399, + "learning_rate": 6.685431015143582e-05, + "loss": 0.3778, + "step": 35700 + }, + { + "epoch": 2.4773436957879396, + "grad_norm": 1.6784985065460205, + "learning_rate": 6.684900920658483e-05, + "loss": 0.4987, + "step": 35701 + }, + { + "epoch": 2.4774130872250364, + "grad_norm": 1.7562391757965088, + "learning_rate": 6.684370836638773e-05, + "loss": 0.6396, + "step": 35702 + }, + { + "epoch": 2.4774824786621332, + "grad_norm": 0.9106032252311707, + "learning_rate": 6.683840763086114e-05, + "loss": 0.3892, + "step": 35703 + }, + { + "epoch": 2.4775518700992296, + "grad_norm": 1.3817863464355469, + "learning_rate": 6.683310700002185e-05, + "loss": 0.3204, + "step": 35704 + }, + { + "epoch": 2.4776212615363264, + "grad_norm": 1.0710387229919434, + "learning_rate": 6.68278064738866e-05, + "loss": 0.347, + "step": 35705 + }, + { + "epoch": 2.477690652973423, + "grad_norm": 1.1240355968475342, + "learning_rate": 6.682250605247211e-05, + "loss": 0.3958, + "step": 35706 + }, + { + "epoch": 2.4777600444105197, + "grad_norm": 1.814179539680481, + "learning_rate": 6.681720573579513e-05, + "loss": 0.4226, + "step": 35707 + }, + { + "epoch": 2.4778294358476165, + "grad_norm": 0.7838389277458191, + "learning_rate": 6.681190552387239e-05, + "loss": 0.2474, + "step": 35708 + }, + { + "epoch": 2.477898827284713, + "grad_norm": 1.34891939163208, + "learning_rate": 6.680660541672062e-05, + "loss": 0.4302, + "step": 35709 + }, + { + "epoch": 2.4779682187218097, + "grad_norm": 1.4021185636520386, + "learning_rate": 6.680130541435652e-05, + "loss": 0.3918, + "step": 35710 + }, + { + "epoch": 2.4780376101589066, + "grad_norm": 1.176459550857544, + "learning_rate": 6.679600551679687e-05, + "loss": 0.368, + "step": 35711 + }, + { + "epoch": 2.478107001596003, + "grad_norm": 1.2364773750305176, + "learning_rate": 6.679070572405836e-05, + "loss": 0.3829, + "step": 35712 + }, + { + "epoch": 2.4781763930331, + "grad_norm": 1.080341100692749, + "learning_rate": 6.678540603615775e-05, + "loss": 0.3122, + "step": 35713 + }, + { + "epoch": 2.4782457844701966, + "grad_norm": 1.278199553489685, + "learning_rate": 6.678010645311174e-05, + "loss": 0.436, + "step": 35714 + }, + { + "epoch": 2.478315175907293, + "grad_norm": 1.3065345287322998, + "learning_rate": 6.677480697493708e-05, + "loss": 0.5251, + "step": 35715 + }, + { + "epoch": 2.47838456734439, + "grad_norm": 1.3019673824310303, + "learning_rate": 6.676950760165056e-05, + "loss": 0.3949, + "step": 35716 + }, + { + "epoch": 2.478453958781486, + "grad_norm": 1.5455081462860107, + "learning_rate": 6.676420833326878e-05, + "loss": 0.3425, + "step": 35717 + }, + { + "epoch": 2.478523350218583, + "grad_norm": 2.274463415145874, + "learning_rate": 6.675890916980856e-05, + "loss": 0.5216, + "step": 35718 + }, + { + "epoch": 2.47859274165568, + "grad_norm": 2.054990291595459, + "learning_rate": 6.675361011128661e-05, + "loss": 0.4452, + "step": 35719 + }, + { + "epoch": 2.4786621330927763, + "grad_norm": 0.9923755526542664, + "learning_rate": 6.674831115771962e-05, + "loss": 0.317, + "step": 35720 + }, + { + "epoch": 2.478731524529873, + "grad_norm": 1.253804326057434, + "learning_rate": 6.674301230912439e-05, + "loss": 0.4303, + "step": 35721 + }, + { + "epoch": 2.4788009159669695, + "grad_norm": 1.1901283264160156, + "learning_rate": 6.67377135655176e-05, + "loss": 0.4202, + "step": 35722 + }, + { + "epoch": 2.4788703074040663, + "grad_norm": 1.492011308670044, + "learning_rate": 6.673241492691602e-05, + "loss": 0.4274, + "step": 35723 + }, + { + "epoch": 2.478939698841163, + "grad_norm": 1.3620144128799438, + "learning_rate": 6.672711639333633e-05, + "loss": 0.4331, + "step": 35724 + }, + { + "epoch": 2.4790090902782596, + "grad_norm": 0.7934439778327942, + "learning_rate": 6.672181796479526e-05, + "loss": 0.2484, + "step": 35725 + }, + { + "epoch": 2.4790784817153564, + "grad_norm": 1.189756155014038, + "learning_rate": 6.671651964130955e-05, + "loss": 0.4998, + "step": 35726 + }, + { + "epoch": 2.4791478731524528, + "grad_norm": 0.8680423498153687, + "learning_rate": 6.671122142289595e-05, + "loss": 0.354, + "step": 35727 + }, + { + "epoch": 2.4792172645895496, + "grad_norm": 1.3273100852966309, + "learning_rate": 6.670592330957113e-05, + "loss": 0.3809, + "step": 35728 + }, + { + "epoch": 2.4792866560266464, + "grad_norm": 1.3530919551849365, + "learning_rate": 6.670062530135186e-05, + "loss": 0.4826, + "step": 35729 + }, + { + "epoch": 2.479356047463743, + "grad_norm": 1.6400480270385742, + "learning_rate": 6.669532739825485e-05, + "loss": 0.2869, + "step": 35730 + }, + { + "epoch": 2.4794254389008397, + "grad_norm": 1.8225730657577515, + "learning_rate": 6.669002960029686e-05, + "loss": 0.5288, + "step": 35731 + }, + { + "epoch": 2.479494830337936, + "grad_norm": 1.1969516277313232, + "learning_rate": 6.668473190749455e-05, + "loss": 0.4248, + "step": 35732 + }, + { + "epoch": 2.479564221775033, + "grad_norm": 1.1698569059371948, + "learning_rate": 6.667943431986468e-05, + "loss": 0.3193, + "step": 35733 + }, + { + "epoch": 2.4796336132121297, + "grad_norm": 1.312243938446045, + "learning_rate": 6.667413683742395e-05, + "loss": 0.3853, + "step": 35734 + }, + { + "epoch": 2.479703004649226, + "grad_norm": 1.2663005590438843, + "learning_rate": 6.666883946018911e-05, + "loss": 0.4009, + "step": 35735 + }, + { + "epoch": 2.479772396086323, + "grad_norm": 1.3704426288604736, + "learning_rate": 6.666354218817689e-05, + "loss": 0.4348, + "step": 35736 + }, + { + "epoch": 2.4798417875234198, + "grad_norm": 1.932930827140808, + "learning_rate": 6.665824502140403e-05, + "loss": 0.454, + "step": 35737 + }, + { + "epoch": 2.479911178960516, + "grad_norm": 1.1321702003479004, + "learning_rate": 6.66529479598872e-05, + "loss": 0.3354, + "step": 35738 + }, + { + "epoch": 2.479980570397613, + "grad_norm": 1.484060525894165, + "learning_rate": 6.664765100364312e-05, + "loss": 0.5104, + "step": 35739 + }, + { + "epoch": 2.48004996183471, + "grad_norm": 1.2127987146377563, + "learning_rate": 6.664235415268853e-05, + "loss": 0.4717, + "step": 35740 + }, + { + "epoch": 2.480119353271806, + "grad_norm": 1.1602429151535034, + "learning_rate": 6.663705740704019e-05, + "loss": 0.3178, + "step": 35741 + }, + { + "epoch": 2.480188744708903, + "grad_norm": 0.9946579933166504, + "learning_rate": 6.663176076671475e-05, + "loss": 0.3052, + "step": 35742 + }, + { + "epoch": 2.4802581361459994, + "grad_norm": 1.5677777528762817, + "learning_rate": 6.662646423172902e-05, + "loss": 0.5246, + "step": 35743 + }, + { + "epoch": 2.4803275275830963, + "grad_norm": 1.0474272966384888, + "learning_rate": 6.662116780209966e-05, + "loss": 0.3006, + "step": 35744 + }, + { + "epoch": 2.480396919020193, + "grad_norm": 1.3786182403564453, + "learning_rate": 6.661587147784338e-05, + "loss": 0.5594, + "step": 35745 + }, + { + "epoch": 2.4804663104572895, + "grad_norm": 0.9467613101005554, + "learning_rate": 6.661057525897694e-05, + "loss": 0.308, + "step": 35746 + }, + { + "epoch": 2.4805357018943863, + "grad_norm": 0.9841901063919067, + "learning_rate": 6.660527914551703e-05, + "loss": 0.3184, + "step": 35747 + }, + { + "epoch": 2.4806050933314827, + "grad_norm": 1.3233948945999146, + "learning_rate": 6.659998313748037e-05, + "loss": 0.5843, + "step": 35748 + }, + { + "epoch": 2.4806744847685795, + "grad_norm": 1.0581687688827515, + "learning_rate": 6.659468723488369e-05, + "loss": 0.5442, + "step": 35749 + }, + { + "epoch": 2.4807438762056764, + "grad_norm": 1.8879883289337158, + "learning_rate": 6.65893914377437e-05, + "loss": 0.8353, + "step": 35750 + }, + { + "epoch": 2.4808132676427728, + "grad_norm": 3.078761577606201, + "learning_rate": 6.658409574607718e-05, + "loss": 0.6004, + "step": 35751 + }, + { + "epoch": 2.4808826590798696, + "grad_norm": 1.455436110496521, + "learning_rate": 6.657880015990074e-05, + "loss": 0.4294, + "step": 35752 + }, + { + "epoch": 2.480952050516966, + "grad_norm": 1.5694293975830078, + "learning_rate": 6.657350467923116e-05, + "loss": 0.5331, + "step": 35753 + }, + { + "epoch": 2.481021441954063, + "grad_norm": 1.1896882057189941, + "learning_rate": 6.656820930408514e-05, + "loss": 0.4248, + "step": 35754 + }, + { + "epoch": 2.4810908333911597, + "grad_norm": 0.8644146919250488, + "learning_rate": 6.65629140344794e-05, + "loss": 0.3153, + "step": 35755 + }, + { + "epoch": 2.481160224828256, + "grad_norm": 1.0889865159988403, + "learning_rate": 6.655761887043068e-05, + "loss": 0.4222, + "step": 35756 + }, + { + "epoch": 2.481229616265353, + "grad_norm": 1.1600031852722168, + "learning_rate": 6.655232381195567e-05, + "loss": 0.4264, + "step": 35757 + }, + { + "epoch": 2.4812990077024497, + "grad_norm": 1.2306852340698242, + "learning_rate": 6.654702885907111e-05, + "loss": 0.4724, + "step": 35758 + }, + { + "epoch": 2.481368399139546, + "grad_norm": 0.8952910304069519, + "learning_rate": 6.654173401179369e-05, + "loss": 0.3076, + "step": 35759 + }, + { + "epoch": 2.481437790576643, + "grad_norm": 1.3111125230789185, + "learning_rate": 6.653643927014012e-05, + "loss": 0.3633, + "step": 35760 + }, + { + "epoch": 2.4815071820137393, + "grad_norm": 1.2113702297210693, + "learning_rate": 6.653114463412713e-05, + "loss": 0.4955, + "step": 35761 + }, + { + "epoch": 2.481576573450836, + "grad_norm": 1.315079927444458, + "learning_rate": 6.652585010377145e-05, + "loss": 0.5411, + "step": 35762 + }, + { + "epoch": 2.481645964887933, + "grad_norm": 0.972725510597229, + "learning_rate": 6.652055567908976e-05, + "loss": 0.4653, + "step": 35763 + }, + { + "epoch": 2.4817153563250294, + "grad_norm": 1.375040054321289, + "learning_rate": 6.651526136009879e-05, + "loss": 0.4325, + "step": 35764 + }, + { + "epoch": 2.481784747762126, + "grad_norm": 2.225680112838745, + "learning_rate": 6.65099671468153e-05, + "loss": 0.4197, + "step": 35765 + }, + { + "epoch": 2.481854139199223, + "grad_norm": 1.3134512901306152, + "learning_rate": 6.650467303925592e-05, + "loss": 0.3688, + "step": 35766 + }, + { + "epoch": 2.4819235306363194, + "grad_norm": 1.1551426649093628, + "learning_rate": 6.64993790374374e-05, + "loss": 0.3741, + "step": 35767 + }, + { + "epoch": 2.4819929220734163, + "grad_norm": 1.294970154762268, + "learning_rate": 6.649408514137644e-05, + "loss": 0.5016, + "step": 35768 + }, + { + "epoch": 2.4820623135105127, + "grad_norm": 1.0542452335357666, + "learning_rate": 6.648879135108976e-05, + "loss": 0.3513, + "step": 35769 + }, + { + "epoch": 2.4821317049476095, + "grad_norm": 1.3679208755493164, + "learning_rate": 6.64834976665941e-05, + "loss": 0.5543, + "step": 35770 + }, + { + "epoch": 2.4822010963847063, + "grad_norm": 1.2633289098739624, + "learning_rate": 6.647820408790614e-05, + "loss": 0.3858, + "step": 35771 + }, + { + "epoch": 2.4822704878218027, + "grad_norm": 1.4608408212661743, + "learning_rate": 6.647291061504265e-05, + "loss": 0.5289, + "step": 35772 + }, + { + "epoch": 2.4823398792588995, + "grad_norm": 1.1675406694412231, + "learning_rate": 6.646761724802025e-05, + "loss": 0.4913, + "step": 35773 + }, + { + "epoch": 2.482409270695996, + "grad_norm": 1.6988633871078491, + "learning_rate": 6.646232398685568e-05, + "loss": 0.4688, + "step": 35774 + }, + { + "epoch": 2.4824786621330928, + "grad_norm": 1.8843275308609009, + "learning_rate": 6.645703083156567e-05, + "loss": 0.4303, + "step": 35775 + }, + { + "epoch": 2.4825480535701896, + "grad_norm": 1.5185332298278809, + "learning_rate": 6.645173778216691e-05, + "loss": 0.4079, + "step": 35776 + }, + { + "epoch": 2.482617445007286, + "grad_norm": 1.5043362379074097, + "learning_rate": 6.644644483867614e-05, + "loss": 0.4969, + "step": 35777 + }, + { + "epoch": 2.482686836444383, + "grad_norm": 0.8597505688667297, + "learning_rate": 6.644115200111007e-05, + "loss": 0.3799, + "step": 35778 + }, + { + "epoch": 2.482756227881479, + "grad_norm": 1.578208088874817, + "learning_rate": 6.643585926948537e-05, + "loss": 0.3464, + "step": 35779 + }, + { + "epoch": 2.482825619318576, + "grad_norm": 1.4939364194869995, + "learning_rate": 6.643056664381875e-05, + "loss": 0.4924, + "step": 35780 + }, + { + "epoch": 2.482895010755673, + "grad_norm": 0.9011442065238953, + "learning_rate": 6.642527412412698e-05, + "loss": 0.3076, + "step": 35781 + }, + { + "epoch": 2.4829644021927693, + "grad_norm": 1.108034372329712, + "learning_rate": 6.641998171042668e-05, + "loss": 0.3196, + "step": 35782 + }, + { + "epoch": 2.483033793629866, + "grad_norm": 1.834145188331604, + "learning_rate": 6.641468940273461e-05, + "loss": 0.5007, + "step": 35783 + }, + { + "epoch": 2.483103185066963, + "grad_norm": 0.9184581637382507, + "learning_rate": 6.640939720106747e-05, + "loss": 0.3303, + "step": 35784 + }, + { + "epoch": 2.4831725765040593, + "grad_norm": 1.4077738523483276, + "learning_rate": 6.640410510544202e-05, + "loss": 0.4768, + "step": 35785 + }, + { + "epoch": 2.483241967941156, + "grad_norm": 1.0051164627075195, + "learning_rate": 6.639881311587485e-05, + "loss": 0.285, + "step": 35786 + }, + { + "epoch": 2.4833113593782525, + "grad_norm": 1.2554020881652832, + "learning_rate": 6.639352123238273e-05, + "loss": 0.5125, + "step": 35787 + }, + { + "epoch": 2.4833807508153494, + "grad_norm": 1.0402640104293823, + "learning_rate": 6.638822945498237e-05, + "loss": 0.567, + "step": 35788 + }, + { + "epoch": 2.483450142252446, + "grad_norm": 1.5137007236480713, + "learning_rate": 6.638293778369048e-05, + "loss": 0.3818, + "step": 35789 + }, + { + "epoch": 2.4835195336895426, + "grad_norm": 1.2580811977386475, + "learning_rate": 6.637764621852375e-05, + "loss": 0.3797, + "step": 35790 + }, + { + "epoch": 2.4835889251266394, + "grad_norm": 1.4652754068374634, + "learning_rate": 6.637235475949888e-05, + "loss": 0.3109, + "step": 35791 + }, + { + "epoch": 2.4836583165637363, + "grad_norm": 1.1322412490844727, + "learning_rate": 6.636706340663263e-05, + "loss": 0.4059, + "step": 35792 + }, + { + "epoch": 2.4837277080008326, + "grad_norm": 1.0925865173339844, + "learning_rate": 6.636177215994163e-05, + "loss": 0.3786, + "step": 35793 + }, + { + "epoch": 2.4837970994379295, + "grad_norm": 1.034688949584961, + "learning_rate": 6.635648101944259e-05, + "loss": 0.3747, + "step": 35794 + }, + { + "epoch": 2.483866490875026, + "grad_norm": 1.2032673358917236, + "learning_rate": 6.635118998515226e-05, + "loss": 0.3702, + "step": 35795 + }, + { + "epoch": 2.4839358823121227, + "grad_norm": 0.8850001096725464, + "learning_rate": 6.63458990570873e-05, + "loss": 0.2683, + "step": 35796 + }, + { + "epoch": 2.4840052737492195, + "grad_norm": 1.3323071002960205, + "learning_rate": 6.634060823526447e-05, + "loss": 0.3838, + "step": 35797 + }, + { + "epoch": 2.484074665186316, + "grad_norm": 1.3129091262817383, + "learning_rate": 6.633531751970039e-05, + "loss": 0.4149, + "step": 35798 + }, + { + "epoch": 2.4841440566234128, + "grad_norm": 1.2498292922973633, + "learning_rate": 6.633002691041184e-05, + "loss": 0.4439, + "step": 35799 + }, + { + "epoch": 2.484213448060509, + "grad_norm": 1.1008374691009521, + "learning_rate": 6.63247364074155e-05, + "loss": 0.2846, + "step": 35800 + }, + { + "epoch": 2.484282839497606, + "grad_norm": 1.2619284391403198, + "learning_rate": 6.631944601072802e-05, + "loss": 0.4209, + "step": 35801 + }, + { + "epoch": 2.484352230934703, + "grad_norm": 1.2766423225402832, + "learning_rate": 6.631415572036616e-05, + "loss": 0.3652, + "step": 35802 + }, + { + "epoch": 2.484421622371799, + "grad_norm": 1.434206247329712, + "learning_rate": 6.630886553634659e-05, + "loss": 0.4864, + "step": 35803 + }, + { + "epoch": 2.484491013808896, + "grad_norm": 1.169816255569458, + "learning_rate": 6.630357545868601e-05, + "loss": 0.4498, + "step": 35804 + }, + { + "epoch": 2.4845604052459924, + "grad_norm": 1.3580259084701538, + "learning_rate": 6.629828548740116e-05, + "loss": 0.5097, + "step": 35805 + }, + { + "epoch": 2.4846297966830893, + "grad_norm": 1.4773237705230713, + "learning_rate": 6.629299562250875e-05, + "loss": 0.3815, + "step": 35806 + }, + { + "epoch": 2.484699188120186, + "grad_norm": 1.4394999742507935, + "learning_rate": 6.628770586402539e-05, + "loss": 0.4772, + "step": 35807 + }, + { + "epoch": 2.4847685795572825, + "grad_norm": 1.5989644527435303, + "learning_rate": 6.628241621196783e-05, + "loss": 0.375, + "step": 35808 + }, + { + "epoch": 2.4848379709943793, + "grad_norm": 0.9092918038368225, + "learning_rate": 6.627712666635277e-05, + "loss": 0.2818, + "step": 35809 + }, + { + "epoch": 2.484907362431476, + "grad_norm": 1.3869932889938354, + "learning_rate": 6.627183722719692e-05, + "loss": 0.4247, + "step": 35810 + }, + { + "epoch": 2.4849767538685725, + "grad_norm": 1.1543529033660889, + "learning_rate": 6.626654789451697e-05, + "loss": 0.4059, + "step": 35811 + }, + { + "epoch": 2.4850461453056694, + "grad_norm": 0.9618620872497559, + "learning_rate": 6.626125866832962e-05, + "loss": 0.4062, + "step": 35812 + }, + { + "epoch": 2.4851155367427658, + "grad_norm": 1.582147240638733, + "learning_rate": 6.625596954865156e-05, + "loss": 0.4445, + "step": 35813 + }, + { + "epoch": 2.4851849281798626, + "grad_norm": 1.599958062171936, + "learning_rate": 6.625068053549948e-05, + "loss": 0.4826, + "step": 35814 + }, + { + "epoch": 2.4852543196169594, + "grad_norm": 0.9644948840141296, + "learning_rate": 6.624539162889007e-05, + "loss": 0.3352, + "step": 35815 + }, + { + "epoch": 2.485323711054056, + "grad_norm": 1.273973822593689, + "learning_rate": 6.624010282884008e-05, + "loss": 0.3903, + "step": 35816 + }, + { + "epoch": 2.4853931024911526, + "grad_norm": 1.0771338939666748, + "learning_rate": 6.623481413536614e-05, + "loss": 0.4767, + "step": 35817 + }, + { + "epoch": 2.4854624939282495, + "grad_norm": 1.2695261240005493, + "learning_rate": 6.622952554848499e-05, + "loss": 0.4398, + "step": 35818 + }, + { + "epoch": 2.485531885365346, + "grad_norm": 1.37769615650177, + "learning_rate": 6.622423706821328e-05, + "loss": 0.6674, + "step": 35819 + }, + { + "epoch": 2.4856012768024427, + "grad_norm": 1.1804221868515015, + "learning_rate": 6.621894869456779e-05, + "loss": 0.4014, + "step": 35820 + }, + { + "epoch": 2.485670668239539, + "grad_norm": 1.0562808513641357, + "learning_rate": 6.621366042756513e-05, + "loss": 0.3166, + "step": 35821 + }, + { + "epoch": 2.485740059676636, + "grad_norm": 1.187464714050293, + "learning_rate": 6.620837226722201e-05, + "loss": 0.2625, + "step": 35822 + }, + { + "epoch": 2.4858094511137327, + "grad_norm": 2.2970852851867676, + "learning_rate": 6.620308421355515e-05, + "loss": 0.5052, + "step": 35823 + }, + { + "epoch": 2.485878842550829, + "grad_norm": 1.379767656326294, + "learning_rate": 6.619779626658123e-05, + "loss": 0.494, + "step": 35824 + }, + { + "epoch": 2.485948233987926, + "grad_norm": 1.4591997861862183, + "learning_rate": 6.619250842631692e-05, + "loss": 0.4074, + "step": 35825 + }, + { + "epoch": 2.4860176254250224, + "grad_norm": 1.493949055671692, + "learning_rate": 6.618722069277896e-05, + "loss": 0.4962, + "step": 35826 + }, + { + "epoch": 2.486087016862119, + "grad_norm": 1.4315123558044434, + "learning_rate": 6.618193306598406e-05, + "loss": 0.4585, + "step": 35827 + }, + { + "epoch": 2.486156408299216, + "grad_norm": 1.5259991884231567, + "learning_rate": 6.617664554594885e-05, + "loss": 0.4795, + "step": 35828 + }, + { + "epoch": 2.4862257997363124, + "grad_norm": 1.5973284244537354, + "learning_rate": 6.617135813269002e-05, + "loss": 0.5413, + "step": 35829 + }, + { + "epoch": 2.4862951911734092, + "grad_norm": 0.9812132120132446, + "learning_rate": 6.61660708262243e-05, + "loss": 0.3551, + "step": 35830 + }, + { + "epoch": 2.4863645826105056, + "grad_norm": 1.1981239318847656, + "learning_rate": 6.61607836265684e-05, + "loss": 0.3879, + "step": 35831 + }, + { + "epoch": 2.4864339740476025, + "grad_norm": 1.0039033889770508, + "learning_rate": 6.615549653373894e-05, + "loss": 0.2898, + "step": 35832 + }, + { + "epoch": 2.4865033654846993, + "grad_norm": 1.250017762184143, + "learning_rate": 6.615020954775264e-05, + "loss": 0.4605, + "step": 35833 + }, + { + "epoch": 2.4865727569217957, + "grad_norm": 1.29623281955719, + "learning_rate": 6.614492266862624e-05, + "loss": 0.3439, + "step": 35834 + }, + { + "epoch": 2.4866421483588925, + "grad_norm": 1.3876219987869263, + "learning_rate": 6.613963589637639e-05, + "loss": 0.4752, + "step": 35835 + }, + { + "epoch": 2.4867115397959894, + "grad_norm": 1.5578712224960327, + "learning_rate": 6.613434923101977e-05, + "loss": 0.463, + "step": 35836 + }, + { + "epoch": 2.4867809312330857, + "grad_norm": 0.9442269206047058, + "learning_rate": 6.612906267257307e-05, + "loss": 0.3513, + "step": 35837 + }, + { + "epoch": 2.4868503226701826, + "grad_norm": 1.6059380769729614, + "learning_rate": 6.612377622105299e-05, + "loss": 0.4861, + "step": 35838 + }, + { + "epoch": 2.4869197141072794, + "grad_norm": 1.532157301902771, + "learning_rate": 6.611848987647622e-05, + "loss": 0.5498, + "step": 35839 + }, + { + "epoch": 2.486989105544376, + "grad_norm": 1.2044254541397095, + "learning_rate": 6.611320363885943e-05, + "loss": 0.442, + "step": 35840 + }, + { + "epoch": 2.4870584969814726, + "grad_norm": 1.2355176210403442, + "learning_rate": 6.610791750821938e-05, + "loss": 0.3398, + "step": 35841 + }, + { + "epoch": 2.487127888418569, + "grad_norm": 1.4482510089874268, + "learning_rate": 6.610263148457268e-05, + "loss": 0.4128, + "step": 35842 + }, + { + "epoch": 2.487197279855666, + "grad_norm": 1.9503408670425415, + "learning_rate": 6.6097345567936e-05, + "loss": 0.4877, + "step": 35843 + }, + { + "epoch": 2.4872666712927627, + "grad_norm": 1.5008883476257324, + "learning_rate": 6.60920597583261e-05, + "loss": 0.4204, + "step": 35844 + }, + { + "epoch": 2.487336062729859, + "grad_norm": 1.1223915815353394, + "learning_rate": 6.608677405575961e-05, + "loss": 0.2607, + "step": 35845 + }, + { + "epoch": 2.487405454166956, + "grad_norm": 1.2154755592346191, + "learning_rate": 6.608148846025326e-05, + "loss": 0.4441, + "step": 35846 + }, + { + "epoch": 2.4874748456040523, + "grad_norm": 2.3304903507232666, + "learning_rate": 6.607620297182372e-05, + "loss": 0.5613, + "step": 35847 + }, + { + "epoch": 2.487544237041149, + "grad_norm": 1.4048632383346558, + "learning_rate": 6.607091759048768e-05, + "loss": 0.3932, + "step": 35848 + }, + { + "epoch": 2.487613628478246, + "grad_norm": 1.0213940143585205, + "learning_rate": 6.60656323162618e-05, + "loss": 0.3041, + "step": 35849 + }, + { + "epoch": 2.4876830199153424, + "grad_norm": 1.037954568862915, + "learning_rate": 6.606034714916276e-05, + "loss": 0.289, + "step": 35850 + }, + { + "epoch": 2.487752411352439, + "grad_norm": 1.5552059412002563, + "learning_rate": 6.605506208920731e-05, + "loss": 0.3856, + "step": 35851 + }, + { + "epoch": 2.4878218027895356, + "grad_norm": 1.3356250524520874, + "learning_rate": 6.604977713641206e-05, + "loss": 0.4685, + "step": 35852 + }, + { + "epoch": 2.4878911942266324, + "grad_norm": 0.9954423904418945, + "learning_rate": 6.604449229079372e-05, + "loss": 0.3348, + "step": 35853 + }, + { + "epoch": 2.4879605856637292, + "grad_norm": 1.9232646226882935, + "learning_rate": 6.603920755236903e-05, + "loss": 0.3279, + "step": 35854 + }, + { + "epoch": 2.4880299771008256, + "grad_norm": 0.9387965202331543, + "learning_rate": 6.603392292115458e-05, + "loss": 0.4456, + "step": 35855 + }, + { + "epoch": 2.4880993685379225, + "grad_norm": 0.882290244102478, + "learning_rate": 6.602863839716709e-05, + "loss": 0.2671, + "step": 35856 + }, + { + "epoch": 2.488168759975019, + "grad_norm": 2.129485607147217, + "learning_rate": 6.602335398042325e-05, + "loss": 0.4861, + "step": 35857 + }, + { + "epoch": 2.4882381514121157, + "grad_norm": 1.526943325996399, + "learning_rate": 6.601806967093974e-05, + "loss": 0.5363, + "step": 35858 + }, + { + "epoch": 2.4883075428492125, + "grad_norm": 1.3903084993362427, + "learning_rate": 6.601278546873325e-05, + "loss": 0.3383, + "step": 35859 + }, + { + "epoch": 2.488376934286309, + "grad_norm": 1.2175960540771484, + "learning_rate": 6.600750137382044e-05, + "loss": 0.4839, + "step": 35860 + }, + { + "epoch": 2.4884463257234057, + "grad_norm": 0.7547219395637512, + "learning_rate": 6.600221738621803e-05, + "loss": 0.2763, + "step": 35861 + }, + { + "epoch": 2.4885157171605026, + "grad_norm": 0.9354838132858276, + "learning_rate": 6.599693350594265e-05, + "loss": 0.4, + "step": 35862 + }, + { + "epoch": 2.488585108597599, + "grad_norm": 1.3986866474151611, + "learning_rate": 6.599164973301102e-05, + "loss": 0.4881, + "step": 35863 + }, + { + "epoch": 2.488654500034696, + "grad_norm": 1.4573763608932495, + "learning_rate": 6.59863660674398e-05, + "loss": 0.3974, + "step": 35864 + }, + { + "epoch": 2.4887238914717926, + "grad_norm": 1.2973994016647339, + "learning_rate": 6.598108250924566e-05, + "loss": 0.4078, + "step": 35865 + }, + { + "epoch": 2.488793282908889, + "grad_norm": 1.3061214685440063, + "learning_rate": 6.597579905844534e-05, + "loss": 0.4404, + "step": 35866 + }, + { + "epoch": 2.488862674345986, + "grad_norm": 1.3241161108016968, + "learning_rate": 6.597051571505544e-05, + "loss": 0.5218, + "step": 35867 + }, + { + "epoch": 2.4889320657830822, + "grad_norm": 1.201788067817688, + "learning_rate": 6.596523247909268e-05, + "loss": 0.4907, + "step": 35868 + }, + { + "epoch": 2.489001457220179, + "grad_norm": 0.8861435651779175, + "learning_rate": 6.595994935057372e-05, + "loss": 0.2891, + "step": 35869 + }, + { + "epoch": 2.489070848657276, + "grad_norm": 1.2892731428146362, + "learning_rate": 6.595466632951529e-05, + "loss": 0.5791, + "step": 35870 + }, + { + "epoch": 2.4891402400943723, + "grad_norm": 1.494612216949463, + "learning_rate": 6.5949383415934e-05, + "loss": 0.4687, + "step": 35871 + }, + { + "epoch": 2.489209631531469, + "grad_norm": 1.2416456937789917, + "learning_rate": 6.594410060984655e-05, + "loss": 0.4462, + "step": 35872 + }, + { + "epoch": 2.4892790229685655, + "grad_norm": 1.4096648693084717, + "learning_rate": 6.593881791126961e-05, + "loss": 0.4734, + "step": 35873 + }, + { + "epoch": 2.4893484144056623, + "grad_norm": 1.6574344635009766, + "learning_rate": 6.593353532021988e-05, + "loss": 0.5177, + "step": 35874 + }, + { + "epoch": 2.489417805842759, + "grad_norm": 1.9607702493667603, + "learning_rate": 6.592825283671409e-05, + "loss": 0.4893, + "step": 35875 + }, + { + "epoch": 2.4894871972798556, + "grad_norm": 1.1363914012908936, + "learning_rate": 6.592297046076879e-05, + "loss": 0.3558, + "step": 35876 + }, + { + "epoch": 2.4895565887169524, + "grad_norm": 1.2523401975631714, + "learning_rate": 6.591768819240072e-05, + "loss": 0.2961, + "step": 35877 + }, + { + "epoch": 2.489625980154049, + "grad_norm": 1.2533109188079834, + "learning_rate": 6.591240603162656e-05, + "loss": 0.3786, + "step": 35878 + }, + { + "epoch": 2.4896953715911456, + "grad_norm": 0.8336349129676819, + "learning_rate": 6.590712397846297e-05, + "loss": 0.2023, + "step": 35879 + }, + { + "epoch": 2.4897647630282425, + "grad_norm": 0.8047417402267456, + "learning_rate": 6.590184203292664e-05, + "loss": 0.3389, + "step": 35880 + }, + { + "epoch": 2.489834154465339, + "grad_norm": 1.2975541353225708, + "learning_rate": 6.589656019503423e-05, + "loss": 0.4584, + "step": 35881 + }, + { + "epoch": 2.4899035459024357, + "grad_norm": 1.2319245338439941, + "learning_rate": 6.589127846480247e-05, + "loss": 0.3837, + "step": 35882 + }, + { + "epoch": 2.489972937339532, + "grad_norm": 1.2750601768493652, + "learning_rate": 6.588599684224797e-05, + "loss": 0.4377, + "step": 35883 + }, + { + "epoch": 2.490042328776629, + "grad_norm": 1.3733694553375244, + "learning_rate": 6.588071532738738e-05, + "loss": 0.4579, + "step": 35884 + }, + { + "epoch": 2.4901117202137257, + "grad_norm": 1.2694857120513916, + "learning_rate": 6.587543392023744e-05, + "loss": 0.4161, + "step": 35885 + }, + { + "epoch": 2.490181111650822, + "grad_norm": 1.051923394203186, + "learning_rate": 6.587015262081481e-05, + "loss": 0.4224, + "step": 35886 + }, + { + "epoch": 2.490250503087919, + "grad_norm": 1.0501798391342163, + "learning_rate": 6.586487142913612e-05, + "loss": 0.383, + "step": 35887 + }, + { + "epoch": 2.490319894525016, + "grad_norm": 1.480956792831421, + "learning_rate": 6.585959034521807e-05, + "loss": 0.5326, + "step": 35888 + }, + { + "epoch": 2.490389285962112, + "grad_norm": 1.2769352197647095, + "learning_rate": 6.585430936907738e-05, + "loss": 0.4293, + "step": 35889 + }, + { + "epoch": 2.490458677399209, + "grad_norm": 1.5754280090332031, + "learning_rate": 6.584902850073064e-05, + "loss": 0.61, + "step": 35890 + }, + { + "epoch": 2.490528068836306, + "grad_norm": 1.261399507522583, + "learning_rate": 6.584374774019455e-05, + "loss": 0.4571, + "step": 35891 + }, + { + "epoch": 2.4905974602734022, + "grad_norm": 0.8752728700637817, + "learning_rate": 6.583846708748578e-05, + "loss": 0.4121, + "step": 35892 + }, + { + "epoch": 2.490666851710499, + "grad_norm": 1.444499135017395, + "learning_rate": 6.583318654262101e-05, + "loss": 0.4209, + "step": 35893 + }, + { + "epoch": 2.4907362431475955, + "grad_norm": 1.552120327949524, + "learning_rate": 6.58279061056169e-05, + "loss": 0.4992, + "step": 35894 + }, + { + "epoch": 2.4908056345846923, + "grad_norm": 1.6937264204025269, + "learning_rate": 6.582262577649012e-05, + "loss": 0.4081, + "step": 35895 + }, + { + "epoch": 2.490875026021789, + "grad_norm": 1.5755562782287598, + "learning_rate": 6.58173455552574e-05, + "loss": 0.4946, + "step": 35896 + }, + { + "epoch": 2.4909444174588855, + "grad_norm": 1.1749684810638428, + "learning_rate": 6.581206544193532e-05, + "loss": 0.4385, + "step": 35897 + }, + { + "epoch": 2.4910138088959823, + "grad_norm": 1.1033216714859009, + "learning_rate": 6.580678543654057e-05, + "loss": 0.4418, + "step": 35898 + }, + { + "epoch": 2.4910832003330787, + "grad_norm": 1.5708461999893188, + "learning_rate": 6.580150553908983e-05, + "loss": 0.4237, + "step": 35899 + }, + { + "epoch": 2.4911525917701756, + "grad_norm": 1.0958706140518188, + "learning_rate": 6.579622574959976e-05, + "loss": 0.3691, + "step": 35900 + }, + { + "epoch": 2.4912219832072724, + "grad_norm": 1.5508164167404175, + "learning_rate": 6.579094606808708e-05, + "loss": 0.4011, + "step": 35901 + }, + { + "epoch": 2.491291374644369, + "grad_norm": 1.1064703464508057, + "learning_rate": 6.578566649456838e-05, + "loss": 0.3857, + "step": 35902 + }, + { + "epoch": 2.4913607660814656, + "grad_norm": 1.1416726112365723, + "learning_rate": 6.578038702906038e-05, + "loss": 0.3642, + "step": 35903 + }, + { + "epoch": 2.491430157518562, + "grad_norm": 1.3325690031051636, + "learning_rate": 6.577510767157971e-05, + "loss": 0.4091, + "step": 35904 + }, + { + "epoch": 2.491499548955659, + "grad_norm": 0.9732681512832642, + "learning_rate": 6.576982842214307e-05, + "loss": 0.3008, + "step": 35905 + }, + { + "epoch": 2.4915689403927557, + "grad_norm": 1.2264713048934937, + "learning_rate": 6.576454928076711e-05, + "loss": 0.4434, + "step": 35906 + }, + { + "epoch": 2.491638331829852, + "grad_norm": 1.0580646991729736, + "learning_rate": 6.575927024746848e-05, + "loss": 0.3963, + "step": 35907 + }, + { + "epoch": 2.491707723266949, + "grad_norm": 1.0061415433883667, + "learning_rate": 6.575399132226388e-05, + "loss": 0.2861, + "step": 35908 + }, + { + "epoch": 2.4917771147040453, + "grad_norm": 1.038720726966858, + "learning_rate": 6.574871250516993e-05, + "loss": 0.324, + "step": 35909 + }, + { + "epoch": 2.491846506141142, + "grad_norm": 0.9450017213821411, + "learning_rate": 6.574343379620337e-05, + "loss": 0.3076, + "step": 35910 + }, + { + "epoch": 2.491915897578239, + "grad_norm": 1.0502943992614746, + "learning_rate": 6.573815519538078e-05, + "loss": 0.4022, + "step": 35911 + }, + { + "epoch": 2.4919852890153353, + "grad_norm": 1.1769330501556396, + "learning_rate": 6.573287670271884e-05, + "loss": 0.4105, + "step": 35912 + }, + { + "epoch": 2.492054680452432, + "grad_norm": 1.0371636152267456, + "learning_rate": 6.572759831823425e-05, + "loss": 0.5314, + "step": 35913 + }, + { + "epoch": 2.492124071889529, + "grad_norm": 1.0613477230072021, + "learning_rate": 6.572232004194366e-05, + "loss": 0.2881, + "step": 35914 + }, + { + "epoch": 2.4921934633266254, + "grad_norm": 1.4762672185897827, + "learning_rate": 6.571704187386372e-05, + "loss": 0.615, + "step": 35915 + }, + { + "epoch": 2.4922628547637222, + "grad_norm": 1.1080352067947388, + "learning_rate": 6.571176381401111e-05, + "loss": 0.4716, + "step": 35916 + }, + { + "epoch": 2.492332246200819, + "grad_norm": 1.3577356338500977, + "learning_rate": 6.570648586240251e-05, + "loss": 0.519, + "step": 35917 + }, + { + "epoch": 2.4924016376379154, + "grad_norm": 1.077265977859497, + "learning_rate": 6.570120801905452e-05, + "loss": 0.3443, + "step": 35918 + }, + { + "epoch": 2.4924710290750123, + "grad_norm": 1.0247604846954346, + "learning_rate": 6.569593028398384e-05, + "loss": 0.3127, + "step": 35919 + }, + { + "epoch": 2.4925404205121087, + "grad_norm": 0.7961330413818359, + "learning_rate": 6.569065265720716e-05, + "loss": 0.3499, + "step": 35920 + }, + { + "epoch": 2.4926098119492055, + "grad_norm": 2.17006516456604, + "learning_rate": 6.568537513874107e-05, + "loss": 0.4795, + "step": 35921 + }, + { + "epoch": 2.4926792033863023, + "grad_norm": 1.7433431148529053, + "learning_rate": 6.568009772860228e-05, + "loss": 0.4699, + "step": 35922 + }, + { + "epoch": 2.4927485948233987, + "grad_norm": 1.1766858100891113, + "learning_rate": 6.567482042680743e-05, + "loss": 0.3638, + "step": 35923 + }, + { + "epoch": 2.4928179862604956, + "grad_norm": 1.1849297285079956, + "learning_rate": 6.566954323337323e-05, + "loss": 0.3537, + "step": 35924 + }, + { + "epoch": 2.492887377697592, + "grad_norm": 1.408441185951233, + "learning_rate": 6.566426614831626e-05, + "loss": 0.3834, + "step": 35925 + }, + { + "epoch": 2.4929567691346888, + "grad_norm": 1.1584155559539795, + "learning_rate": 6.565898917165323e-05, + "loss": 0.311, + "step": 35926 + }, + { + "epoch": 2.4930261605717856, + "grad_norm": 1.2463428974151611, + "learning_rate": 6.565371230340078e-05, + "loss": 0.3345, + "step": 35927 + }, + { + "epoch": 2.493095552008882, + "grad_norm": 1.0250543355941772, + "learning_rate": 6.564843554357558e-05, + "loss": 0.3615, + "step": 35928 + }, + { + "epoch": 2.493164943445979, + "grad_norm": 1.1033867597579956, + "learning_rate": 6.564315889219426e-05, + "loss": 0.3477, + "step": 35929 + }, + { + "epoch": 2.493234334883075, + "grad_norm": 1.3791637420654297, + "learning_rate": 6.563788234927357e-05, + "loss": 0.435, + "step": 35930 + }, + { + "epoch": 2.493303726320172, + "grad_norm": 1.1179578304290771, + "learning_rate": 6.563260591483005e-05, + "loss": 0.4015, + "step": 35931 + }, + { + "epoch": 2.493373117757269, + "grad_norm": 1.4283167123794556, + "learning_rate": 6.56273295888804e-05, + "loss": 0.4951, + "step": 35932 + }, + { + "epoch": 2.4934425091943653, + "grad_norm": 0.9233805537223816, + "learning_rate": 6.562205337144129e-05, + "loss": 0.4601, + "step": 35933 + }, + { + "epoch": 2.493511900631462, + "grad_norm": 1.1303192377090454, + "learning_rate": 6.561677726252935e-05, + "loss": 0.3626, + "step": 35934 + }, + { + "epoch": 2.4935812920685585, + "grad_norm": 1.4367754459381104, + "learning_rate": 6.561150126216127e-05, + "loss": 0.4643, + "step": 35935 + }, + { + "epoch": 2.4936506835056553, + "grad_norm": 1.870662808418274, + "learning_rate": 6.560622537035371e-05, + "loss": 0.6157, + "step": 35936 + }, + { + "epoch": 2.493720074942752, + "grad_norm": 1.5938085317611694, + "learning_rate": 6.560094958712331e-05, + "loss": 0.5829, + "step": 35937 + }, + { + "epoch": 2.4937894663798486, + "grad_norm": 1.2196019887924194, + "learning_rate": 6.55956739124867e-05, + "loss": 0.4067, + "step": 35938 + }, + { + "epoch": 2.4938588578169454, + "grad_norm": 1.1009939908981323, + "learning_rate": 6.559039834646054e-05, + "loss": 0.35, + "step": 35939 + }, + { + "epoch": 2.493928249254042, + "grad_norm": 1.2616699934005737, + "learning_rate": 6.558512288906155e-05, + "loss": 0.2454, + "step": 35940 + }, + { + "epoch": 2.4939976406911386, + "grad_norm": 1.219706416130066, + "learning_rate": 6.557984754030629e-05, + "loss": 0.3302, + "step": 35941 + }, + { + "epoch": 2.4940670321282354, + "grad_norm": 0.9188030362129211, + "learning_rate": 6.557457230021147e-05, + "loss": 0.2626, + "step": 35942 + }, + { + "epoch": 2.4941364235653323, + "grad_norm": 1.2486705780029297, + "learning_rate": 6.556929716879372e-05, + "loss": 0.4136, + "step": 35943 + }, + { + "epoch": 2.4942058150024287, + "grad_norm": 1.0582695007324219, + "learning_rate": 6.556402214606977e-05, + "loss": 0.3268, + "step": 35944 + }, + { + "epoch": 2.4942752064395255, + "grad_norm": 1.5370155572891235, + "learning_rate": 6.555874723205612e-05, + "loss": 0.3391, + "step": 35945 + }, + { + "epoch": 2.494344597876622, + "grad_norm": 0.9849160313606262, + "learning_rate": 6.555347242676955e-05, + "loss": 0.2953, + "step": 35946 + }, + { + "epoch": 2.4944139893137187, + "grad_norm": 1.5255653858184814, + "learning_rate": 6.554819773022666e-05, + "loss": 0.432, + "step": 35947 + }, + { + "epoch": 2.4944833807508155, + "grad_norm": 1.2748501300811768, + "learning_rate": 6.55429231424441e-05, + "loss": 0.4193, + "step": 35948 + }, + { + "epoch": 2.494552772187912, + "grad_norm": 1.4130268096923828, + "learning_rate": 6.553764866343855e-05, + "loss": 0.4627, + "step": 35949 + }, + { + "epoch": 2.4946221636250088, + "grad_norm": 1.656497597694397, + "learning_rate": 6.553237429322664e-05, + "loss": 0.6728, + "step": 35950 + }, + { + "epoch": 2.494691555062105, + "grad_norm": 1.0295532941818237, + "learning_rate": 6.552710003182507e-05, + "loss": 0.2938, + "step": 35951 + }, + { + "epoch": 2.494760946499202, + "grad_norm": 1.5107266902923584, + "learning_rate": 6.55218258792504e-05, + "loss": 0.3963, + "step": 35952 + }, + { + "epoch": 2.494830337936299, + "grad_norm": 0.8643089532852173, + "learning_rate": 6.551655183551932e-05, + "loss": 0.2804, + "step": 35953 + }, + { + "epoch": 2.494899729373395, + "grad_norm": 1.3117872476577759, + "learning_rate": 6.551127790064849e-05, + "loss": 0.4237, + "step": 35954 + }, + { + "epoch": 2.494969120810492, + "grad_norm": 1.578545331954956, + "learning_rate": 6.550600407465457e-05, + "loss": 0.3873, + "step": 35955 + }, + { + "epoch": 2.4950385122475884, + "grad_norm": 1.2923420667648315, + "learning_rate": 6.550073035755417e-05, + "loss": 0.4602, + "step": 35956 + }, + { + "epoch": 2.4951079036846853, + "grad_norm": 1.0984277725219727, + "learning_rate": 6.549545674936398e-05, + "loss": 0.4035, + "step": 35957 + }, + { + "epoch": 2.495177295121782, + "grad_norm": 1.7916473150253296, + "learning_rate": 6.549018325010064e-05, + "loss": 0.4691, + "step": 35958 + }, + { + "epoch": 2.4952466865588785, + "grad_norm": 1.444370150566101, + "learning_rate": 6.548490985978079e-05, + "loss": 0.3019, + "step": 35959 + }, + { + "epoch": 2.4953160779959753, + "grad_norm": 1.272493839263916, + "learning_rate": 6.547963657842104e-05, + "loss": 0.4408, + "step": 35960 + }, + { + "epoch": 2.4953854694330717, + "grad_norm": 1.3867523670196533, + "learning_rate": 6.54743634060381e-05, + "loss": 0.4235, + "step": 35961 + }, + { + "epoch": 2.4954548608701685, + "grad_norm": 2.520551919937134, + "learning_rate": 6.546909034264857e-05, + "loss": 0.6563, + "step": 35962 + }, + { + "epoch": 2.4955242523072654, + "grad_norm": 1.3117541074752808, + "learning_rate": 6.546381738826911e-05, + "loss": 0.2554, + "step": 35963 + }, + { + "epoch": 2.4955936437443618, + "grad_norm": 1.1796149015426636, + "learning_rate": 6.545854454291638e-05, + "loss": 0.3541, + "step": 35964 + }, + { + "epoch": 2.4956630351814586, + "grad_norm": 0.9771077036857605, + "learning_rate": 6.545327180660706e-05, + "loss": 0.301, + "step": 35965 + }, + { + "epoch": 2.4957324266185554, + "grad_norm": 1.0200121402740479, + "learning_rate": 6.544799917935771e-05, + "loss": 0.4543, + "step": 35966 + }, + { + "epoch": 2.495801818055652, + "grad_norm": 1.1940116882324219, + "learning_rate": 6.544272666118502e-05, + "loss": 0.3677, + "step": 35967 + }, + { + "epoch": 2.4958712094927487, + "grad_norm": 1.5472991466522217, + "learning_rate": 6.543745425210563e-05, + "loss": 0.3446, + "step": 35968 + }, + { + "epoch": 2.4959406009298455, + "grad_norm": 1.7388193607330322, + "learning_rate": 6.543218195213618e-05, + "loss": 0.4197, + "step": 35969 + }, + { + "epoch": 2.496009992366942, + "grad_norm": 1.296409249305725, + "learning_rate": 6.542690976129332e-05, + "loss": 0.4654, + "step": 35970 + }, + { + "epoch": 2.4960793838040387, + "grad_norm": 1.2940226793289185, + "learning_rate": 6.542163767959372e-05, + "loss": 0.5871, + "step": 35971 + }, + { + "epoch": 2.496148775241135, + "grad_norm": 1.073451042175293, + "learning_rate": 6.541636570705402e-05, + "loss": 0.3706, + "step": 35972 + }, + { + "epoch": 2.496218166678232, + "grad_norm": 1.0128945112228394, + "learning_rate": 6.54110938436908e-05, + "loss": 0.2875, + "step": 35973 + }, + { + "epoch": 2.4962875581153288, + "grad_norm": 1.0152307748794556, + "learning_rate": 6.540582208952077e-05, + "loss": 0.3973, + "step": 35974 + }, + { + "epoch": 2.496356949552425, + "grad_norm": 1.1776515245437622, + "learning_rate": 6.540055044456051e-05, + "loss": 0.4217, + "step": 35975 + }, + { + "epoch": 2.496426340989522, + "grad_norm": 0.7500163912773132, + "learning_rate": 6.539527890882672e-05, + "loss": 0.2401, + "step": 35976 + }, + { + "epoch": 2.4964957324266184, + "grad_norm": 1.3899534940719604, + "learning_rate": 6.539000748233601e-05, + "loss": 0.6141, + "step": 35977 + }, + { + "epoch": 2.496565123863715, + "grad_norm": 1.4543640613555908, + "learning_rate": 6.538473616510503e-05, + "loss": 0.4116, + "step": 35978 + }, + { + "epoch": 2.496634515300812, + "grad_norm": 1.525234580039978, + "learning_rate": 6.537946495715046e-05, + "loss": 0.344, + "step": 35979 + }, + { + "epoch": 2.4967039067379084, + "grad_norm": 1.1852166652679443, + "learning_rate": 6.537419385848886e-05, + "loss": 0.4598, + "step": 35980 + }, + { + "epoch": 2.4967732981750053, + "grad_norm": 1.4897987842559814, + "learning_rate": 6.536892286913692e-05, + "loss": 0.4434, + "step": 35981 + }, + { + "epoch": 2.4968426896121017, + "grad_norm": 2.726614475250244, + "learning_rate": 6.536365198911128e-05, + "loss": 0.5256, + "step": 35982 + }, + { + "epoch": 2.4969120810491985, + "grad_norm": 1.4705246686935425, + "learning_rate": 6.535838121842855e-05, + "loss": 0.4058, + "step": 35983 + }, + { + "epoch": 2.4969814724862953, + "grad_norm": 1.231168270111084, + "learning_rate": 6.53531105571054e-05, + "loss": 0.3137, + "step": 35984 + }, + { + "epoch": 2.4970508639233917, + "grad_norm": 1.0865243673324585, + "learning_rate": 6.534784000515846e-05, + "loss": 0.3796, + "step": 35985 + }, + { + "epoch": 2.4971202553604885, + "grad_norm": 1.438250184059143, + "learning_rate": 6.534256956260442e-05, + "loss": 0.3667, + "step": 35986 + }, + { + "epoch": 2.4971896467975854, + "grad_norm": 0.8473228216171265, + "learning_rate": 6.533729922945982e-05, + "loss": 0.2826, + "step": 35987 + }, + { + "epoch": 2.4972590382346818, + "grad_norm": 1.5250109434127808, + "learning_rate": 6.533202900574133e-05, + "loss": 0.4327, + "step": 35988 + }, + { + "epoch": 2.4973284296717786, + "grad_norm": 1.3965365886688232, + "learning_rate": 6.532675889146562e-05, + "loss": 0.4609, + "step": 35989 + }, + { + "epoch": 2.497397821108875, + "grad_norm": 1.3049407005310059, + "learning_rate": 6.532148888664932e-05, + "loss": 0.4162, + "step": 35990 + }, + { + "epoch": 2.497467212545972, + "grad_norm": 1.4205753803253174, + "learning_rate": 6.531621899130905e-05, + "loss": 0.4058, + "step": 35991 + }, + { + "epoch": 2.4975366039830686, + "grad_norm": 1.4798948764801025, + "learning_rate": 6.531094920546143e-05, + "loss": 0.3265, + "step": 35992 + }, + { + "epoch": 2.497605995420165, + "grad_norm": 1.3187366724014282, + "learning_rate": 6.530567952912317e-05, + "loss": 0.4265, + "step": 35993 + }, + { + "epoch": 2.497675386857262, + "grad_norm": 1.0661815404891968, + "learning_rate": 6.530040996231083e-05, + "loss": 0.4011, + "step": 35994 + }, + { + "epoch": 2.4977447782943587, + "grad_norm": 1.10293710231781, + "learning_rate": 6.529514050504106e-05, + "loss": 0.3969, + "step": 35995 + }, + { + "epoch": 2.497814169731455, + "grad_norm": 1.1588186025619507, + "learning_rate": 6.52898711573305e-05, + "loss": 0.4412, + "step": 35996 + }, + { + "epoch": 2.497883561168552, + "grad_norm": 1.1069607734680176, + "learning_rate": 6.528460191919578e-05, + "loss": 0.307, + "step": 35997 + }, + { + "epoch": 2.4979529526056483, + "grad_norm": 1.6020046472549438, + "learning_rate": 6.527933279065356e-05, + "loss": 0.5143, + "step": 35998 + }, + { + "epoch": 2.498022344042745, + "grad_norm": 1.082516074180603, + "learning_rate": 6.527406377172046e-05, + "loss": 0.3773, + "step": 35999 + }, + { + "epoch": 2.498091735479842, + "grad_norm": 1.073448657989502, + "learning_rate": 6.526879486241315e-05, + "loss": 0.3588, + "step": 36000 + }, + { + "epoch": 2.498091735479842, + "eval_loss": 0.5200383067131042, + "eval_runtime": 3838.5666, + "eval_samples_per_second": 6.323, + "eval_steps_per_second": 0.79, + "step": 36000 + }, + { + "epoch": 2.4981611269169384, + "grad_norm": 1.120690941810608, + "learning_rate": 6.526352606274817e-05, + "loss": 0.3434, + "step": 36001 + }, + { + "epoch": 2.498230518354035, + "grad_norm": 1.246719479560852, + "learning_rate": 6.525825737274222e-05, + "loss": 0.4667, + "step": 36002 + }, + { + "epoch": 2.4982999097911316, + "grad_norm": 1.1027480363845825, + "learning_rate": 6.525298879241193e-05, + "loss": 0.2834, + "step": 36003 + }, + { + "epoch": 2.4983693012282284, + "grad_norm": 1.1030991077423096, + "learning_rate": 6.524772032177392e-05, + "loss": 0.5311, + "step": 36004 + }, + { + "epoch": 2.4984386926653253, + "grad_norm": 1.8105536699295044, + "learning_rate": 6.52424519608448e-05, + "loss": 0.5662, + "step": 36005 + }, + { + "epoch": 2.4985080841024216, + "grad_norm": 1.3892228603363037, + "learning_rate": 6.523718370964128e-05, + "loss": 0.412, + "step": 36006 + }, + { + "epoch": 2.4985774755395185, + "grad_norm": 1.1884626150131226, + "learning_rate": 6.523191556817992e-05, + "loss": 0.3392, + "step": 36007 + }, + { + "epoch": 2.498646866976615, + "grad_norm": 0.7960643172264099, + "learning_rate": 6.522664753647733e-05, + "loss": 0.2861, + "step": 36008 + }, + { + "epoch": 2.4987162584137117, + "grad_norm": 1.3895540237426758, + "learning_rate": 6.522137961455022e-05, + "loss": 0.4807, + "step": 36009 + }, + { + "epoch": 2.4987856498508085, + "grad_norm": 1.0175271034240723, + "learning_rate": 6.521611180241515e-05, + "loss": 0.4054, + "step": 36010 + }, + { + "epoch": 2.498855041287905, + "grad_norm": 1.5723680257797241, + "learning_rate": 6.52108441000888e-05, + "loss": 0.6189, + "step": 36011 + }, + { + "epoch": 2.4989244327250018, + "grad_norm": 0.9453793168067932, + "learning_rate": 6.520557650758776e-05, + "loss": 0.4306, + "step": 36012 + }, + { + "epoch": 2.4989938241620986, + "grad_norm": 1.7281783819198608, + "learning_rate": 6.520030902492873e-05, + "loss": 0.36, + "step": 36013 + }, + { + "epoch": 2.499063215599195, + "grad_norm": 0.8056609034538269, + "learning_rate": 6.519504165212824e-05, + "loss": 0.2447, + "step": 36014 + }, + { + "epoch": 2.499132607036292, + "grad_norm": 1.4642878770828247, + "learning_rate": 6.518977438920298e-05, + "loss": 0.6448, + "step": 36015 + }, + { + "epoch": 2.499201998473388, + "grad_norm": 1.4501566886901855, + "learning_rate": 6.518450723616953e-05, + "loss": 0.4921, + "step": 36016 + }, + { + "epoch": 2.499271389910485, + "grad_norm": 0.8474621176719666, + "learning_rate": 6.517924019304458e-05, + "loss": 0.2932, + "step": 36017 + }, + { + "epoch": 2.499340781347582, + "grad_norm": 1.2028363943099976, + "learning_rate": 6.517397325984473e-05, + "loss": 0.4331, + "step": 36018 + }, + { + "epoch": 2.4994101727846783, + "grad_norm": 1.2474850416183472, + "learning_rate": 6.51687064365866e-05, + "loss": 0.5312, + "step": 36019 + }, + { + "epoch": 2.499479564221775, + "grad_norm": 1.3004227876663208, + "learning_rate": 6.516343972328685e-05, + "loss": 0.5912, + "step": 36020 + }, + { + "epoch": 2.499548955658872, + "grad_norm": 1.3019849061965942, + "learning_rate": 6.515817311996204e-05, + "loss": 0.5108, + "step": 36021 + }, + { + "epoch": 2.4996183470959683, + "grad_norm": 1.3371165990829468, + "learning_rate": 6.515290662662885e-05, + "loss": 0.5649, + "step": 36022 + }, + { + "epoch": 2.499687738533065, + "grad_norm": 1.8576446771621704, + "learning_rate": 6.514764024330388e-05, + "loss": 0.3526, + "step": 36023 + }, + { + "epoch": 2.4997571299701615, + "grad_norm": 0.9820464253425598, + "learning_rate": 6.514237397000378e-05, + "loss": 0.3069, + "step": 36024 + }, + { + "epoch": 2.4998265214072584, + "grad_norm": 1.3684148788452148, + "learning_rate": 6.513710780674516e-05, + "loss": 0.3551, + "step": 36025 + }, + { + "epoch": 2.499895912844355, + "grad_norm": 0.9290204644203186, + "learning_rate": 6.513184175354464e-05, + "loss": 0.343, + "step": 36026 + }, + { + "epoch": 2.4999653042814516, + "grad_norm": 1.2636830806732178, + "learning_rate": 6.512657581041888e-05, + "loss": 0.4122, + "step": 36027 + }, + { + "epoch": 2.5000346957185484, + "grad_norm": 0.9150175452232361, + "learning_rate": 6.512130997738443e-05, + "loss": 0.2681, + "step": 36028 + }, + { + "epoch": 2.500104087155645, + "grad_norm": 1.2963855266571045, + "learning_rate": 6.511604425445799e-05, + "loss": 0.4811, + "step": 36029 + }, + { + "epoch": 2.5001734785927416, + "grad_norm": 1.053215742111206, + "learning_rate": 6.511077864165613e-05, + "loss": 0.4541, + "step": 36030 + }, + { + "epoch": 2.5002428700298385, + "grad_norm": 1.235621452331543, + "learning_rate": 6.51055131389955e-05, + "loss": 0.3711, + "step": 36031 + }, + { + "epoch": 2.500312261466935, + "grad_norm": 1.0265525579452515, + "learning_rate": 6.510024774649271e-05, + "loss": 0.3631, + "step": 36032 + }, + { + "epoch": 2.5003816529040317, + "grad_norm": 1.3064095973968506, + "learning_rate": 6.50949824641644e-05, + "loss": 0.3766, + "step": 36033 + }, + { + "epoch": 2.500451044341128, + "grad_norm": 1.1616851091384888, + "learning_rate": 6.508971729202719e-05, + "loss": 0.3594, + "step": 36034 + }, + { + "epoch": 2.500520435778225, + "grad_norm": 1.3026093244552612, + "learning_rate": 6.508445223009766e-05, + "loss": 0.2735, + "step": 36035 + }, + { + "epoch": 2.5005898272153217, + "grad_norm": 1.9087321758270264, + "learning_rate": 6.507918727839248e-05, + "loss": 0.4571, + "step": 36036 + }, + { + "epoch": 2.500659218652418, + "grad_norm": 1.5822662115097046, + "learning_rate": 6.507392243692827e-05, + "loss": 0.3594, + "step": 36037 + }, + { + "epoch": 2.500728610089515, + "grad_norm": 1.1173481941223145, + "learning_rate": 6.50686577057216e-05, + "loss": 0.3435, + "step": 36038 + }, + { + "epoch": 2.5007980015266114, + "grad_norm": 1.4873063564300537, + "learning_rate": 6.506339308478915e-05, + "loss": 0.4581, + "step": 36039 + }, + { + "epoch": 2.500867392963708, + "grad_norm": 1.3496791124343872, + "learning_rate": 6.50581285741475e-05, + "loss": 0.2573, + "step": 36040 + }, + { + "epoch": 2.500936784400805, + "grad_norm": 1.8430290222167969, + "learning_rate": 6.505286417381333e-05, + "loss": 0.595, + "step": 36041 + }, + { + "epoch": 2.501006175837902, + "grad_norm": 2.192768096923828, + "learning_rate": 6.504759988380317e-05, + "loss": 0.4887, + "step": 36042 + }, + { + "epoch": 2.5010755672749982, + "grad_norm": 1.3226897716522217, + "learning_rate": 6.504233570413367e-05, + "loss": 0.5294, + "step": 36043 + }, + { + "epoch": 2.501144958712095, + "grad_norm": 1.434762954711914, + "learning_rate": 6.50370716348215e-05, + "loss": 0.3821, + "step": 36044 + }, + { + "epoch": 2.5012143501491915, + "grad_norm": 1.7918325662612915, + "learning_rate": 6.503180767588321e-05, + "loss": 0.4977, + "step": 36045 + }, + { + "epoch": 2.5012837415862883, + "grad_norm": 1.723212718963623, + "learning_rate": 6.502654382733546e-05, + "loss": 0.4765, + "step": 36046 + }, + { + "epoch": 2.501353133023385, + "grad_norm": 1.2137105464935303, + "learning_rate": 6.502128008919484e-05, + "loss": 0.3412, + "step": 36047 + }, + { + "epoch": 2.5014225244604815, + "grad_norm": 1.86715829372406, + "learning_rate": 6.501601646147802e-05, + "loss": 0.5185, + "step": 36048 + }, + { + "epoch": 2.5014919158975784, + "grad_norm": 1.1967167854309082, + "learning_rate": 6.501075294420154e-05, + "loss": 0.4652, + "step": 36049 + }, + { + "epoch": 2.5015613073346747, + "grad_norm": 1.4590644836425781, + "learning_rate": 6.500548953738203e-05, + "loss": 0.3577, + "step": 36050 + }, + { + "epoch": 2.5016306987717716, + "grad_norm": 1.3058702945709229, + "learning_rate": 6.500022624103617e-05, + "loss": 0.3243, + "step": 36051 + }, + { + "epoch": 2.5017000902088684, + "grad_norm": 0.9549804329872131, + "learning_rate": 6.499496305518052e-05, + "loss": 0.3619, + "step": 36052 + }, + { + "epoch": 2.501769481645965, + "grad_norm": 1.145668387413025, + "learning_rate": 6.49896999798317e-05, + "loss": 0.32, + "step": 36053 + }, + { + "epoch": 2.5018388730830616, + "grad_norm": 1.363898515701294, + "learning_rate": 6.498443701500634e-05, + "loss": 0.4054, + "step": 36054 + }, + { + "epoch": 2.501908264520158, + "grad_norm": 1.9815431833267212, + "learning_rate": 6.49791741607211e-05, + "loss": 0.6164, + "step": 36055 + }, + { + "epoch": 2.501977655957255, + "grad_norm": 1.146263599395752, + "learning_rate": 6.49739114169925e-05, + "loss": 0.3858, + "step": 36056 + }, + { + "epoch": 2.5020470473943517, + "grad_norm": 1.215742588043213, + "learning_rate": 6.496864878383718e-05, + "loss": 0.3922, + "step": 36057 + }, + { + "epoch": 2.502116438831448, + "grad_norm": 1.119729995727539, + "learning_rate": 6.496338626127178e-05, + "loss": 0.3631, + "step": 36058 + }, + { + "epoch": 2.502185830268545, + "grad_norm": 1.1503894329071045, + "learning_rate": 6.495812384931292e-05, + "loss": 0.3275, + "step": 36059 + }, + { + "epoch": 2.5022552217056413, + "grad_norm": 1.1329981088638306, + "learning_rate": 6.49528615479772e-05, + "loss": 0.3675, + "step": 36060 + }, + { + "epoch": 2.502324613142738, + "grad_norm": 1.0381734371185303, + "learning_rate": 6.494759935728123e-05, + "loss": 0.4232, + "step": 36061 + }, + { + "epoch": 2.502394004579835, + "grad_norm": 1.3020960092544556, + "learning_rate": 6.494233727724162e-05, + "loss": 0.3673, + "step": 36062 + }, + { + "epoch": 2.5024633960169314, + "grad_norm": 1.5218734741210938, + "learning_rate": 6.4937075307875e-05, + "loss": 0.44, + "step": 36063 + }, + { + "epoch": 2.502532787454028, + "grad_norm": 1.449611783027649, + "learning_rate": 6.493181344919793e-05, + "loss": 0.5209, + "step": 36064 + }, + { + "epoch": 2.5026021788911246, + "grad_norm": 1.4453200101852417, + "learning_rate": 6.492655170122707e-05, + "loss": 0.3776, + "step": 36065 + }, + { + "epoch": 2.5026715703282214, + "grad_norm": 1.4278558492660522, + "learning_rate": 6.4921290063979e-05, + "loss": 0.5836, + "step": 36066 + }, + { + "epoch": 2.5027409617653182, + "grad_norm": 1.2283390760421753, + "learning_rate": 6.491602853747036e-05, + "loss": 0.3789, + "step": 36067 + }, + { + "epoch": 2.502810353202415, + "grad_norm": 0.8992352485656738, + "learning_rate": 6.491076712171775e-05, + "loss": 0.3504, + "step": 36068 + }, + { + "epoch": 2.5028797446395115, + "grad_norm": 0.9847376346588135, + "learning_rate": 6.490550581673782e-05, + "loss": 0.4604, + "step": 36069 + }, + { + "epoch": 2.5029491360766083, + "grad_norm": 1.5093923807144165, + "learning_rate": 6.490024462254709e-05, + "loss": 0.3985, + "step": 36070 + }, + { + "epoch": 2.5030185275137047, + "grad_norm": 1.3633173704147339, + "learning_rate": 6.489498353916222e-05, + "loss": 0.4332, + "step": 36071 + }, + { + "epoch": 2.5030879189508015, + "grad_norm": 1.0281517505645752, + "learning_rate": 6.48897225665998e-05, + "loss": 0.3133, + "step": 36072 + }, + { + "epoch": 2.5031573103878983, + "grad_norm": 1.6060707569122314, + "learning_rate": 6.488446170487645e-05, + "loss": 0.4779, + "step": 36073 + }, + { + "epoch": 2.5032267018249947, + "grad_norm": 1.4948662519454956, + "learning_rate": 6.48792009540088e-05, + "loss": 0.5949, + "step": 36074 + }, + { + "epoch": 2.5032960932620916, + "grad_norm": 1.8767600059509277, + "learning_rate": 6.487394031401346e-05, + "loss": 0.4499, + "step": 36075 + }, + { + "epoch": 2.503365484699188, + "grad_norm": 1.2061315774917603, + "learning_rate": 6.486867978490701e-05, + "loss": 0.3677, + "step": 36076 + }, + { + "epoch": 2.503434876136285, + "grad_norm": 1.0140142440795898, + "learning_rate": 6.486341936670603e-05, + "loss": 0.3557, + "step": 36077 + }, + { + "epoch": 2.5035042675733816, + "grad_norm": 1.113927960395813, + "learning_rate": 6.485815905942716e-05, + "loss": 0.3722, + "step": 36078 + }, + { + "epoch": 2.503573659010478, + "grad_norm": 1.923897385597229, + "learning_rate": 6.485289886308703e-05, + "loss": 0.4377, + "step": 36079 + }, + { + "epoch": 2.503643050447575, + "grad_norm": 1.244155764579773, + "learning_rate": 6.484763877770219e-05, + "loss": 0.3979, + "step": 36080 + }, + { + "epoch": 2.5037124418846712, + "grad_norm": 1.1095609664916992, + "learning_rate": 6.48423788032893e-05, + "loss": 0.2928, + "step": 36081 + }, + { + "epoch": 2.503781833321768, + "grad_norm": 0.9573453068733215, + "learning_rate": 6.483711893986493e-05, + "loss": 0.29, + "step": 36082 + }, + { + "epoch": 2.503851224758865, + "grad_norm": 1.045583963394165, + "learning_rate": 6.483185918744572e-05, + "loss": 0.3851, + "step": 36083 + }, + { + "epoch": 2.5039206161959613, + "grad_norm": 1.520732045173645, + "learning_rate": 6.482659954604823e-05, + "loss": 0.4726, + "step": 36084 + }, + { + "epoch": 2.503990007633058, + "grad_norm": 2.491912364959717, + "learning_rate": 6.482134001568907e-05, + "loss": 0.4665, + "step": 36085 + }, + { + "epoch": 2.5040593990701545, + "grad_norm": 1.0727921724319458, + "learning_rate": 6.481608059638487e-05, + "loss": 0.4825, + "step": 36086 + }, + { + "epoch": 2.5041287905072513, + "grad_norm": 1.1522746086120605, + "learning_rate": 6.481082128815222e-05, + "loss": 0.3829, + "step": 36087 + }, + { + "epoch": 2.504198181944348, + "grad_norm": 1.414971947669983, + "learning_rate": 6.480556209100772e-05, + "loss": 0.357, + "step": 36088 + }, + { + "epoch": 2.5042675733814446, + "grad_norm": 1.4617047309875488, + "learning_rate": 6.480030300496802e-05, + "loss": 0.4456, + "step": 36089 + }, + { + "epoch": 2.5043369648185414, + "grad_norm": 1.2797958850860596, + "learning_rate": 6.479504403004963e-05, + "loss": 0.2626, + "step": 36090 + }, + { + "epoch": 2.504406356255638, + "grad_norm": 1.368610143661499, + "learning_rate": 6.47897851662692e-05, + "loss": 0.5193, + "step": 36091 + }, + { + "epoch": 2.5044757476927346, + "grad_norm": 1.419948935508728, + "learning_rate": 6.478452641364337e-05, + "loss": 0.4446, + "step": 36092 + }, + { + "epoch": 2.5045451391298315, + "grad_norm": 1.542051911354065, + "learning_rate": 6.477926777218867e-05, + "loss": 0.4653, + "step": 36093 + }, + { + "epoch": 2.5046145305669283, + "grad_norm": 1.6589120626449585, + "learning_rate": 6.477400924192174e-05, + "loss": 0.5835, + "step": 36094 + }, + { + "epoch": 2.5046839220040247, + "grad_norm": 1.2773196697235107, + "learning_rate": 6.476875082285921e-05, + "loss": 0.4321, + "step": 36095 + }, + { + "epoch": 2.5047533134411215, + "grad_norm": 0.8499945998191833, + "learning_rate": 6.476349251501762e-05, + "loss": 0.2637, + "step": 36096 + }, + { + "epoch": 2.504822704878218, + "grad_norm": 3.606131076812744, + "learning_rate": 6.47582343184136e-05, + "loss": 0.7147, + "step": 36097 + }, + { + "epoch": 2.5048920963153147, + "grad_norm": 1.2581807374954224, + "learning_rate": 6.475297623306375e-05, + "loss": 0.4762, + "step": 36098 + }, + { + "epoch": 2.5049614877524116, + "grad_norm": 1.446254014968872, + "learning_rate": 6.474771825898465e-05, + "loss": 0.5145, + "step": 36099 + }, + { + "epoch": 2.505030879189508, + "grad_norm": 0.9242468476295471, + "learning_rate": 6.474246039619291e-05, + "loss": 0.377, + "step": 36100 + }, + { + "epoch": 2.505100270626605, + "grad_norm": 1.4328234195709229, + "learning_rate": 6.473720264470513e-05, + "loss": 0.4709, + "step": 36101 + }, + { + "epoch": 2.505169662063701, + "grad_norm": 1.212059497833252, + "learning_rate": 6.473194500453792e-05, + "loss": 0.5416, + "step": 36102 + }, + { + "epoch": 2.505239053500798, + "grad_norm": 1.0253703594207764, + "learning_rate": 6.472668747570792e-05, + "loss": 0.3555, + "step": 36103 + }, + { + "epoch": 2.505308444937895, + "grad_norm": 1.206121802330017, + "learning_rate": 6.472143005823161e-05, + "loss": 0.3171, + "step": 36104 + }, + { + "epoch": 2.5053778363749912, + "grad_norm": 1.3017529249191284, + "learning_rate": 6.471617275212566e-05, + "loss": 0.3907, + "step": 36105 + }, + { + "epoch": 2.505447227812088, + "grad_norm": 1.3216993808746338, + "learning_rate": 6.471091555740666e-05, + "loss": 0.6022, + "step": 36106 + }, + { + "epoch": 2.5055166192491845, + "grad_norm": 1.4732261896133423, + "learning_rate": 6.47056584740912e-05, + "loss": 0.4943, + "step": 36107 + }, + { + "epoch": 2.5055860106862813, + "grad_norm": 0.7479767203330994, + "learning_rate": 6.47004015021959e-05, + "loss": 0.2898, + "step": 36108 + }, + { + "epoch": 2.505655402123378, + "grad_norm": 1.4417197704315186, + "learning_rate": 6.469514464173731e-05, + "loss": 0.322, + "step": 36109 + }, + { + "epoch": 2.5057247935604745, + "grad_norm": 1.0615906715393066, + "learning_rate": 6.468988789273212e-05, + "loss": 0.418, + "step": 36110 + }, + { + "epoch": 2.5057941849975713, + "grad_norm": 1.2516107559204102, + "learning_rate": 6.468463125519679e-05, + "loss": 0.423, + "step": 36111 + }, + { + "epoch": 2.5058635764346677, + "grad_norm": 0.9725518226623535, + "learning_rate": 6.4679374729148e-05, + "loss": 0.3231, + "step": 36112 + }, + { + "epoch": 2.5059329678717646, + "grad_norm": 1.409787893295288, + "learning_rate": 6.467411831460232e-05, + "loss": 0.54, + "step": 36113 + }, + { + "epoch": 2.5060023593088614, + "grad_norm": 0.878995954990387, + "learning_rate": 6.466886201157636e-05, + "loss": 0.3876, + "step": 36114 + }, + { + "epoch": 2.506071750745958, + "grad_norm": 1.3723134994506836, + "learning_rate": 6.46636058200867e-05, + "loss": 0.5741, + "step": 36115 + }, + { + "epoch": 2.5061411421830546, + "grad_norm": 0.9964607357978821, + "learning_rate": 6.465834974014993e-05, + "loss": 0.372, + "step": 36116 + }, + { + "epoch": 2.506210533620151, + "grad_norm": 1.339205265045166, + "learning_rate": 6.465309377178269e-05, + "loss": 0.3523, + "step": 36117 + }, + { + "epoch": 2.506279925057248, + "grad_norm": 1.0508537292480469, + "learning_rate": 6.46478379150015e-05, + "loss": 0.2841, + "step": 36118 + }, + { + "epoch": 2.5063493164943447, + "grad_norm": 1.4292570352554321, + "learning_rate": 6.464258216982298e-05, + "loss": 0.5621, + "step": 36119 + }, + { + "epoch": 2.5064187079314415, + "grad_norm": 1.264432430267334, + "learning_rate": 6.463732653626371e-05, + "loss": 0.4859, + "step": 36120 + }, + { + "epoch": 2.506488099368538, + "grad_norm": 0.9343243837356567, + "learning_rate": 6.463207101434031e-05, + "loss": 0.3044, + "step": 36121 + }, + { + "epoch": 2.5065574908056347, + "grad_norm": 1.356821060180664, + "learning_rate": 6.462681560406936e-05, + "loss": 0.5102, + "step": 36122 + }, + { + "epoch": 2.506626882242731, + "grad_norm": 1.1283612251281738, + "learning_rate": 6.462156030546745e-05, + "loss": 0.3614, + "step": 36123 + }, + { + "epoch": 2.506696273679828, + "grad_norm": 1.3040695190429688, + "learning_rate": 6.46163051185512e-05, + "loss": 0.3507, + "step": 36124 + }, + { + "epoch": 2.506765665116925, + "grad_norm": 1.2827030420303345, + "learning_rate": 6.461105004333715e-05, + "loss": 0.4034, + "step": 36125 + }, + { + "epoch": 2.506835056554021, + "grad_norm": 1.0582901239395142, + "learning_rate": 6.460579507984188e-05, + "loss": 0.3691, + "step": 36126 + }, + { + "epoch": 2.506904447991118, + "grad_norm": 1.121698021888733, + "learning_rate": 6.460054022808202e-05, + "loss": 0.3117, + "step": 36127 + }, + { + "epoch": 2.5069738394282144, + "grad_norm": 1.27662992477417, + "learning_rate": 6.459528548807417e-05, + "loss": 0.3626, + "step": 36128 + }, + { + "epoch": 2.5070432308653112, + "grad_norm": 1.2723485231399536, + "learning_rate": 6.459003085983489e-05, + "loss": 0.495, + "step": 36129 + }, + { + "epoch": 2.507112622302408, + "grad_norm": 1.2377710342407227, + "learning_rate": 6.458477634338075e-05, + "loss": 0.4151, + "step": 36130 + }, + { + "epoch": 2.5071820137395044, + "grad_norm": 1.0533978939056396, + "learning_rate": 6.457952193872841e-05, + "loss": 0.3041, + "step": 36131 + }, + { + "epoch": 2.5072514051766013, + "grad_norm": 1.0620827674865723, + "learning_rate": 6.457426764589436e-05, + "loss": 0.2772, + "step": 36132 + }, + { + "epoch": 2.5073207966136977, + "grad_norm": 0.8126260042190552, + "learning_rate": 6.456901346489528e-05, + "loss": 0.2603, + "step": 36133 + }, + { + "epoch": 2.5073901880507945, + "grad_norm": 1.146655559539795, + "learning_rate": 6.456375939574769e-05, + "loss": 0.4589, + "step": 36134 + }, + { + "epoch": 2.5074595794878913, + "grad_norm": 1.2037862539291382, + "learning_rate": 6.455850543846818e-05, + "loss": 0.4114, + "step": 36135 + }, + { + "epoch": 2.5075289709249877, + "grad_norm": 1.3752121925354004, + "learning_rate": 6.455325159307338e-05, + "loss": 0.4009, + "step": 36136 + }, + { + "epoch": 2.5075983623620846, + "grad_norm": 2.0909533500671387, + "learning_rate": 6.454799785957984e-05, + "loss": 0.652, + "step": 36137 + }, + { + "epoch": 2.507667753799181, + "grad_norm": 1.1459839344024658, + "learning_rate": 6.45427442380042e-05, + "loss": 0.5495, + "step": 36138 + }, + { + "epoch": 2.5077371452362778, + "grad_norm": 1.7510799169540405, + "learning_rate": 6.453749072836294e-05, + "loss": 0.4617, + "step": 36139 + }, + { + "epoch": 2.5078065366733746, + "grad_norm": 1.3542791604995728, + "learning_rate": 6.453223733067274e-05, + "loss": 0.449, + "step": 36140 + }, + { + "epoch": 2.5078759281104714, + "grad_norm": 1.101251482963562, + "learning_rate": 6.452698404495015e-05, + "loss": 0.3636, + "step": 36141 + }, + { + "epoch": 2.507945319547568, + "grad_norm": 0.9383175373077393, + "learning_rate": 6.452173087121174e-05, + "loss": 0.3261, + "step": 36142 + }, + { + "epoch": 2.508014710984664, + "grad_norm": 1.0144214630126953, + "learning_rate": 6.451647780947412e-05, + "loss": 0.2753, + "step": 36143 + }, + { + "epoch": 2.508084102421761, + "grad_norm": 1.3476649522781372, + "learning_rate": 6.451122485975384e-05, + "loss": 0.2821, + "step": 36144 + }, + { + "epoch": 2.508153493858858, + "grad_norm": 1.2708194255828857, + "learning_rate": 6.450597202206758e-05, + "loss": 0.4363, + "step": 36145 + }, + { + "epoch": 2.5082228852959547, + "grad_norm": 1.7641359567642212, + "learning_rate": 6.450071929643179e-05, + "loss": 0.4678, + "step": 36146 + }, + { + "epoch": 2.508292276733051, + "grad_norm": 1.1866847276687622, + "learning_rate": 6.449546668286313e-05, + "loss": 0.4799, + "step": 36147 + }, + { + "epoch": 2.508361668170148, + "grad_norm": 1.2702865600585938, + "learning_rate": 6.449021418137813e-05, + "loss": 0.4284, + "step": 36148 + }, + { + "epoch": 2.5084310596072443, + "grad_norm": 1.2908934354782104, + "learning_rate": 6.448496179199345e-05, + "loss": 0.3686, + "step": 36149 + }, + { + "epoch": 2.508500451044341, + "grad_norm": 1.5914090871810913, + "learning_rate": 6.44797095147256e-05, + "loss": 0.3611, + "step": 36150 + }, + { + "epoch": 2.508569842481438, + "grad_norm": 1.2003263235092163, + "learning_rate": 6.447445734959116e-05, + "loss": 0.403, + "step": 36151 + }, + { + "epoch": 2.5086392339185344, + "grad_norm": 1.0808500051498413, + "learning_rate": 6.44692052966068e-05, + "loss": 0.361, + "step": 36152 + }, + { + "epoch": 2.508708625355631, + "grad_norm": 1.7363479137420654, + "learning_rate": 6.4463953355789e-05, + "loss": 0.3954, + "step": 36153 + }, + { + "epoch": 2.5087780167927276, + "grad_norm": 0.9461402297019958, + "learning_rate": 6.445870152715437e-05, + "loss": 0.2791, + "step": 36154 + }, + { + "epoch": 2.5088474082298244, + "grad_norm": 1.5349175930023193, + "learning_rate": 6.44534498107195e-05, + "loss": 0.4873, + "step": 36155 + }, + { + "epoch": 2.5089167996669213, + "grad_norm": 1.3552491664886475, + "learning_rate": 6.444819820650097e-05, + "loss": 0.5099, + "step": 36156 + }, + { + "epoch": 2.5089861911040177, + "grad_norm": 1.8028768301010132, + "learning_rate": 6.444294671451534e-05, + "loss": 0.4861, + "step": 36157 + }, + { + "epoch": 2.5090555825411145, + "grad_norm": 1.1160094738006592, + "learning_rate": 6.44376953347792e-05, + "loss": 0.3616, + "step": 36158 + }, + { + "epoch": 2.509124973978211, + "grad_norm": 1.1833724975585938, + "learning_rate": 6.443244406730918e-05, + "loss": 0.3341, + "step": 36159 + }, + { + "epoch": 2.5091943654153077, + "grad_norm": 1.164528727531433, + "learning_rate": 6.442719291212175e-05, + "loss": 0.3496, + "step": 36160 + }, + { + "epoch": 2.5092637568524045, + "grad_norm": 1.6929435729980469, + "learning_rate": 6.442194186923357e-05, + "loss": 0.4691, + "step": 36161 + }, + { + "epoch": 2.509333148289501, + "grad_norm": 1.5392630100250244, + "learning_rate": 6.441669093866119e-05, + "loss": 0.4137, + "step": 36162 + }, + { + "epoch": 2.5094025397265978, + "grad_norm": 0.9850048422813416, + "learning_rate": 6.441144012042118e-05, + "loss": 0.3462, + "step": 36163 + }, + { + "epoch": 2.509471931163694, + "grad_norm": 1.8447191715240479, + "learning_rate": 6.440618941453014e-05, + "loss": 0.2274, + "step": 36164 + }, + { + "epoch": 2.509541322600791, + "grad_norm": 1.8736019134521484, + "learning_rate": 6.440093882100461e-05, + "loss": 0.5844, + "step": 36165 + }, + { + "epoch": 2.509610714037888, + "grad_norm": 1.2558752298355103, + "learning_rate": 6.439568833986121e-05, + "loss": 0.5139, + "step": 36166 + }, + { + "epoch": 2.5096801054749847, + "grad_norm": 1.7936279773712158, + "learning_rate": 6.439043797111646e-05, + "loss": 0.4665, + "step": 36167 + }, + { + "epoch": 2.509749496912081, + "grad_norm": 1.4464226961135864, + "learning_rate": 6.4385187714787e-05, + "loss": 0.4147, + "step": 36168 + }, + { + "epoch": 2.5098188883491774, + "grad_norm": 1.2881664037704468, + "learning_rate": 6.437993757088935e-05, + "loss": 0.3073, + "step": 36169 + }, + { + "epoch": 2.5098882797862743, + "grad_norm": 1.188166618347168, + "learning_rate": 6.43746875394401e-05, + "loss": 0.4577, + "step": 36170 + }, + { + "epoch": 2.509957671223371, + "grad_norm": 1.2215067148208618, + "learning_rate": 6.436943762045582e-05, + "loss": 0.3976, + "step": 36171 + }, + { + "epoch": 2.510027062660468, + "grad_norm": 1.1575469970703125, + "learning_rate": 6.436418781395315e-05, + "loss": 0.4353, + "step": 36172 + }, + { + "epoch": 2.5100964540975643, + "grad_norm": 1.4342130422592163, + "learning_rate": 6.435893811994855e-05, + "loss": 0.4146, + "step": 36173 + }, + { + "epoch": 2.510165845534661, + "grad_norm": 1.2586663961410522, + "learning_rate": 6.435368853845864e-05, + "loss": 0.5267, + "step": 36174 + }, + { + "epoch": 2.5102352369717575, + "grad_norm": 1.4628074169158936, + "learning_rate": 6.434843906950004e-05, + "loss": 0.4766, + "step": 36175 + }, + { + "epoch": 2.5103046284088544, + "grad_norm": 1.278897762298584, + "learning_rate": 6.434318971308924e-05, + "loss": 0.3555, + "step": 36176 + }, + { + "epoch": 2.510374019845951, + "grad_norm": 1.145344853401184, + "learning_rate": 6.433794046924287e-05, + "loss": 0.3643, + "step": 36177 + }, + { + "epoch": 2.5104434112830476, + "grad_norm": 1.2509690523147583, + "learning_rate": 6.433269133797749e-05, + "loss": 0.4243, + "step": 36178 + }, + { + "epoch": 2.5105128027201444, + "grad_norm": 1.0472307205200195, + "learning_rate": 6.432744231930969e-05, + "loss": 0.2791, + "step": 36179 + }, + { + "epoch": 2.510582194157241, + "grad_norm": 1.2200020551681519, + "learning_rate": 6.432219341325598e-05, + "loss": 0.4414, + "step": 36180 + }, + { + "epoch": 2.5106515855943377, + "grad_norm": 2.0670995712280273, + "learning_rate": 6.431694461983296e-05, + "loss": 0.474, + "step": 36181 + }, + { + "epoch": 2.5107209770314345, + "grad_norm": 1.531457543373108, + "learning_rate": 6.431169593905721e-05, + "loss": 0.5509, + "step": 36182 + }, + { + "epoch": 2.510790368468531, + "grad_norm": 1.1080244779586792, + "learning_rate": 6.43064473709453e-05, + "loss": 0.5231, + "step": 36183 + }, + { + "epoch": 2.5108597599056277, + "grad_norm": 1.374060869216919, + "learning_rate": 6.430119891551381e-05, + "loss": 0.505, + "step": 36184 + }, + { + "epoch": 2.510929151342724, + "grad_norm": 0.9488646984100342, + "learning_rate": 6.429595057277928e-05, + "loss": 0.3728, + "step": 36185 + }, + { + "epoch": 2.510998542779821, + "grad_norm": 1.4021726846694946, + "learning_rate": 6.429070234275832e-05, + "loss": 0.3755, + "step": 36186 + }, + { + "epoch": 2.5110679342169178, + "grad_norm": 1.347387671470642, + "learning_rate": 6.428545422546746e-05, + "loss": 0.4343, + "step": 36187 + }, + { + "epoch": 2.511137325654014, + "grad_norm": 1.2989288568496704, + "learning_rate": 6.428020622092325e-05, + "loss": 0.5614, + "step": 36188 + }, + { + "epoch": 2.511206717091111, + "grad_norm": 1.1998820304870605, + "learning_rate": 6.427495832914228e-05, + "loss": 0.3266, + "step": 36189 + }, + { + "epoch": 2.5112761085282074, + "grad_norm": 1.1845518350601196, + "learning_rate": 6.426971055014114e-05, + "loss": 0.2913, + "step": 36190 + }, + { + "epoch": 2.511345499965304, + "grad_norm": 1.0910674333572388, + "learning_rate": 6.426446288393637e-05, + "loss": 0.3568, + "step": 36191 + }, + { + "epoch": 2.511414891402401, + "grad_norm": 1.8288649320602417, + "learning_rate": 6.425921533054454e-05, + "loss": 0.6798, + "step": 36192 + }, + { + "epoch": 2.511484282839498, + "grad_norm": 0.8868498802185059, + "learning_rate": 6.425396788998226e-05, + "loss": 0.3614, + "step": 36193 + }, + { + "epoch": 2.5115536742765943, + "grad_norm": 1.582468867301941, + "learning_rate": 6.424872056226604e-05, + "loss": 0.7143, + "step": 36194 + }, + { + "epoch": 2.511623065713691, + "grad_norm": 1.279631495475769, + "learning_rate": 6.424347334741243e-05, + "loss": 0.4405, + "step": 36195 + }, + { + "epoch": 2.5116924571507875, + "grad_norm": 0.9813477396965027, + "learning_rate": 6.423822624543804e-05, + "loss": 0.2658, + "step": 36196 + }, + { + "epoch": 2.5117618485878843, + "grad_norm": 1.33998441696167, + "learning_rate": 6.423297925635941e-05, + "loss": 0.5204, + "step": 36197 + }, + { + "epoch": 2.511831240024981, + "grad_norm": 1.3385555744171143, + "learning_rate": 6.422773238019313e-05, + "loss": 0.4055, + "step": 36198 + }, + { + "epoch": 2.5119006314620775, + "grad_norm": 0.9548995494842529, + "learning_rate": 6.422248561695577e-05, + "loss": 0.2902, + "step": 36199 + }, + { + "epoch": 2.5119700228991744, + "grad_norm": 0.9730538725852966, + "learning_rate": 6.421723896666385e-05, + "loss": 0.21, + "step": 36200 + }, + { + "epoch": 2.5120394143362708, + "grad_norm": 1.180024266242981, + "learning_rate": 6.421199242933396e-05, + "loss": 0.4769, + "step": 36201 + }, + { + "epoch": 2.5121088057733676, + "grad_norm": 1.4230997562408447, + "learning_rate": 6.420674600498264e-05, + "loss": 0.468, + "step": 36202 + }, + { + "epoch": 2.5121781972104644, + "grad_norm": 1.4710639715194702, + "learning_rate": 6.42014996936265e-05, + "loss": 0.4545, + "step": 36203 + }, + { + "epoch": 2.512247588647561, + "grad_norm": 1.1408592462539673, + "learning_rate": 6.419625349528205e-05, + "loss": 0.3718, + "step": 36204 + }, + { + "epoch": 2.5123169800846576, + "grad_norm": 1.0237737894058228, + "learning_rate": 6.419100740996587e-05, + "loss": 0.4854, + "step": 36205 + }, + { + "epoch": 2.512386371521754, + "grad_norm": 1.1659575700759888, + "learning_rate": 6.418576143769452e-05, + "loss": 0.3778, + "step": 36206 + }, + { + "epoch": 2.512455762958851, + "grad_norm": 1.1877919435501099, + "learning_rate": 6.418051557848462e-05, + "loss": 0.4925, + "step": 36207 + }, + { + "epoch": 2.5125251543959477, + "grad_norm": 0.9008255004882812, + "learning_rate": 6.417526983235263e-05, + "loss": 0.285, + "step": 36208 + }, + { + "epoch": 2.512594545833044, + "grad_norm": 1.0362436771392822, + "learning_rate": 6.417002419931516e-05, + "loss": 0.3302, + "step": 36209 + }, + { + "epoch": 2.512663937270141, + "grad_norm": 1.0478301048278809, + "learning_rate": 6.416477867938877e-05, + "loss": 0.2776, + "step": 36210 + }, + { + "epoch": 2.5127333287072373, + "grad_norm": 1.0959908962249756, + "learning_rate": 6.415953327259e-05, + "loss": 0.3641, + "step": 36211 + }, + { + "epoch": 2.512802720144334, + "grad_norm": 1.7308204174041748, + "learning_rate": 6.415428797893544e-05, + "loss": 0.5968, + "step": 36212 + }, + { + "epoch": 2.512872111581431, + "grad_norm": 1.106898307800293, + "learning_rate": 6.414904279844163e-05, + "loss": 0.2773, + "step": 36213 + }, + { + "epoch": 2.5129415030185274, + "grad_norm": 0.9456244707107544, + "learning_rate": 6.414379773112516e-05, + "loss": 0.2771, + "step": 36214 + }, + { + "epoch": 2.513010894455624, + "grad_norm": 1.207517147064209, + "learning_rate": 6.413855277700253e-05, + "loss": 0.3576, + "step": 36215 + }, + { + "epoch": 2.5130802858927206, + "grad_norm": 1.3321470022201538, + "learning_rate": 6.413330793609034e-05, + "loss": 0.3811, + "step": 36216 + }, + { + "epoch": 2.5131496773298174, + "grad_norm": 1.71150803565979, + "learning_rate": 6.412806320840512e-05, + "loss": 0.5177, + "step": 36217 + }, + { + "epoch": 2.5132190687669143, + "grad_norm": 1.165801763534546, + "learning_rate": 6.412281859396347e-05, + "loss": 0.3738, + "step": 36218 + }, + { + "epoch": 2.513288460204011, + "grad_norm": 1.285695195198059, + "learning_rate": 6.41175740927819e-05, + "loss": 0.331, + "step": 36219 + }, + { + "epoch": 2.5133578516411075, + "grad_norm": 1.3985862731933594, + "learning_rate": 6.411232970487698e-05, + "loss": 0.5735, + "step": 36220 + }, + { + "epoch": 2.5134272430782043, + "grad_norm": 1.1534547805786133, + "learning_rate": 6.41070854302653e-05, + "loss": 0.4234, + "step": 36221 + }, + { + "epoch": 2.5134966345153007, + "grad_norm": 1.22653067111969, + "learning_rate": 6.410184126896337e-05, + "loss": 0.4157, + "step": 36222 + }, + { + "epoch": 2.5135660259523975, + "grad_norm": 0.8892115950584412, + "learning_rate": 6.409659722098776e-05, + "loss": 0.2016, + "step": 36223 + }, + { + "epoch": 2.5136354173894944, + "grad_norm": 1.4980487823486328, + "learning_rate": 6.409135328635502e-05, + "loss": 0.5872, + "step": 36224 + }, + { + "epoch": 2.5137048088265908, + "grad_norm": 1.0492238998413086, + "learning_rate": 6.408610946508172e-05, + "loss": 0.426, + "step": 36225 + }, + { + "epoch": 2.5137742002636876, + "grad_norm": 1.1183974742889404, + "learning_rate": 6.408086575718439e-05, + "loss": 0.3103, + "step": 36226 + }, + { + "epoch": 2.513843591700784, + "grad_norm": 0.8026852011680603, + "learning_rate": 6.407562216267961e-05, + "loss": 0.3099, + "step": 36227 + }, + { + "epoch": 2.513912983137881, + "grad_norm": 1.0697588920593262, + "learning_rate": 6.407037868158395e-05, + "loss": 0.324, + "step": 36228 + }, + { + "epoch": 2.5139823745749776, + "grad_norm": 1.8178513050079346, + "learning_rate": 6.406513531391391e-05, + "loss": 0.6318, + "step": 36229 + }, + { + "epoch": 2.514051766012074, + "grad_norm": 0.9358404874801636, + "learning_rate": 6.405989205968605e-05, + "loss": 0.4044, + "step": 36230 + }, + { + "epoch": 2.514121157449171, + "grad_norm": 1.1688416004180908, + "learning_rate": 6.405464891891697e-05, + "loss": 0.3834, + "step": 36231 + }, + { + "epoch": 2.5141905488862673, + "grad_norm": 1.227440595626831, + "learning_rate": 6.404940589162318e-05, + "loss": 0.4355, + "step": 36232 + }, + { + "epoch": 2.514259940323364, + "grad_norm": 1.3972300291061401, + "learning_rate": 6.404416297782123e-05, + "loss": 0.5001, + "step": 36233 + }, + { + "epoch": 2.514329331760461, + "grad_norm": 1.287674069404602, + "learning_rate": 6.403892017752773e-05, + "loss": 0.4122, + "step": 36234 + }, + { + "epoch": 2.5143987231975573, + "grad_norm": 0.9568973183631897, + "learning_rate": 6.403367749075917e-05, + "loss": 0.3525, + "step": 36235 + }, + { + "epoch": 2.514468114634654, + "grad_norm": 0.9377146363258362, + "learning_rate": 6.40284349175321e-05, + "loss": 0.3352, + "step": 36236 + }, + { + "epoch": 2.5145375060717505, + "grad_norm": 1.2510713338851929, + "learning_rate": 6.402319245786309e-05, + "loss": 0.3132, + "step": 36237 + }, + { + "epoch": 2.5146068975088474, + "grad_norm": 0.982144832611084, + "learning_rate": 6.401795011176871e-05, + "loss": 0.4015, + "step": 36238 + }, + { + "epoch": 2.514676288945944, + "grad_norm": 0.972881555557251, + "learning_rate": 6.401270787926547e-05, + "loss": 0.2484, + "step": 36239 + }, + { + "epoch": 2.5147456803830406, + "grad_norm": 1.2757104635238647, + "learning_rate": 6.400746576036993e-05, + "loss": 0.4518, + "step": 36240 + }, + { + "epoch": 2.5148150718201374, + "grad_norm": 1.3629366159439087, + "learning_rate": 6.400222375509868e-05, + "loss": 0.4637, + "step": 36241 + }, + { + "epoch": 2.514884463257234, + "grad_norm": 2.3781661987304688, + "learning_rate": 6.399698186346819e-05, + "loss": 0.3431, + "step": 36242 + }, + { + "epoch": 2.5149538546943306, + "grad_norm": 0.9757696390151978, + "learning_rate": 6.399174008549506e-05, + "loss": 0.3598, + "step": 36243 + }, + { + "epoch": 2.5150232461314275, + "grad_norm": 1.1327972412109375, + "learning_rate": 6.398649842119584e-05, + "loss": 0.4156, + "step": 36244 + }, + { + "epoch": 2.5150926375685243, + "grad_norm": 1.2639962434768677, + "learning_rate": 6.398125687058705e-05, + "loss": 0.439, + "step": 36245 + }, + { + "epoch": 2.5151620290056207, + "grad_norm": 1.8431717157363892, + "learning_rate": 6.397601543368525e-05, + "loss": 0.3497, + "step": 36246 + }, + { + "epoch": 2.5152314204427175, + "grad_norm": 1.3589539527893066, + "learning_rate": 6.3970774110507e-05, + "loss": 0.3553, + "step": 36247 + }, + { + "epoch": 2.515300811879814, + "grad_norm": 1.1237939596176147, + "learning_rate": 6.396553290106887e-05, + "loss": 0.4571, + "step": 36248 + }, + { + "epoch": 2.5153702033169107, + "grad_norm": 0.9667950868606567, + "learning_rate": 6.396029180538734e-05, + "loss": 0.3574, + "step": 36249 + }, + { + "epoch": 2.5154395947540076, + "grad_norm": 0.9992928504943848, + "learning_rate": 6.395505082347896e-05, + "loss": 0.408, + "step": 36250 + }, + { + "epoch": 2.515508986191104, + "grad_norm": 1.541932463645935, + "learning_rate": 6.394980995536034e-05, + "loss": 0.5049, + "step": 36251 + }, + { + "epoch": 2.515578377628201, + "grad_norm": 0.8926253914833069, + "learning_rate": 6.394456920104797e-05, + "loss": 0.2124, + "step": 36252 + }, + { + "epoch": 2.515647769065297, + "grad_norm": 1.0907878875732422, + "learning_rate": 6.393932856055844e-05, + "loss": 0.4374, + "step": 36253 + }, + { + "epoch": 2.515717160502394, + "grad_norm": 1.7892447710037231, + "learning_rate": 6.393408803390824e-05, + "loss": 0.6373, + "step": 36254 + }, + { + "epoch": 2.515786551939491, + "grad_norm": 1.3104320764541626, + "learning_rate": 6.392884762111395e-05, + "loss": 0.2933, + "step": 36255 + }, + { + "epoch": 2.5158559433765872, + "grad_norm": 1.3754838705062866, + "learning_rate": 6.39236073221921e-05, + "loss": 0.3721, + "step": 36256 + }, + { + "epoch": 2.515925334813684, + "grad_norm": 1.5361607074737549, + "learning_rate": 6.391836713715924e-05, + "loss": 0.505, + "step": 36257 + }, + { + "epoch": 2.5159947262507805, + "grad_norm": 1.0400543212890625, + "learning_rate": 6.39131270660319e-05, + "loss": 0.3723, + "step": 36258 + }, + { + "epoch": 2.5160641176878773, + "grad_norm": 1.113193392753601, + "learning_rate": 6.390788710882662e-05, + "loss": 0.4025, + "step": 36259 + }, + { + "epoch": 2.516133509124974, + "grad_norm": 1.2382230758666992, + "learning_rate": 6.390264726555997e-05, + "loss": 0.4823, + "step": 36260 + }, + { + "epoch": 2.5162029005620705, + "grad_norm": 1.082680344581604, + "learning_rate": 6.389740753624847e-05, + "loss": 0.3317, + "step": 36261 + }, + { + "epoch": 2.5162722919991674, + "grad_norm": 1.455227017402649, + "learning_rate": 6.38921679209087e-05, + "loss": 0.6593, + "step": 36262 + }, + { + "epoch": 2.5163416834362637, + "grad_norm": 1.0534203052520752, + "learning_rate": 6.388692841955713e-05, + "loss": 0.4251, + "step": 36263 + }, + { + "epoch": 2.5164110748733606, + "grad_norm": 1.3917447328567505, + "learning_rate": 6.388168903221032e-05, + "loss": 0.4234, + "step": 36264 + }, + { + "epoch": 2.5164804663104574, + "grad_norm": 1.2691586017608643, + "learning_rate": 6.387644975888485e-05, + "loss": 0.3705, + "step": 36265 + }, + { + "epoch": 2.516549857747554, + "grad_norm": 1.1679115295410156, + "learning_rate": 6.387121059959722e-05, + "loss": 0.4113, + "step": 36266 + }, + { + "epoch": 2.5166192491846506, + "grad_norm": 1.4009677171707153, + "learning_rate": 6.3865971554364e-05, + "loss": 0.4006, + "step": 36267 + }, + { + "epoch": 2.516688640621747, + "grad_norm": 1.637352466583252, + "learning_rate": 6.386073262320172e-05, + "loss": 0.6421, + "step": 36268 + }, + { + "epoch": 2.516758032058844, + "grad_norm": 1.1110934019088745, + "learning_rate": 6.385549380612695e-05, + "loss": 0.326, + "step": 36269 + }, + { + "epoch": 2.5168274234959407, + "grad_norm": 1.8027064800262451, + "learning_rate": 6.385025510315615e-05, + "loss": 0.6016, + "step": 36270 + }, + { + "epoch": 2.5168968149330375, + "grad_norm": 1.1701284646987915, + "learning_rate": 6.384501651430588e-05, + "loss": 0.4533, + "step": 36271 + }, + { + "epoch": 2.516966206370134, + "grad_norm": 1.1347601413726807, + "learning_rate": 6.383977803959274e-05, + "loss": 0.5089, + "step": 36272 + }, + { + "epoch": 2.5170355978072307, + "grad_norm": 2.1311228275299072, + "learning_rate": 6.38345396790332e-05, + "loss": 0.2887, + "step": 36273 + }, + { + "epoch": 2.517104989244327, + "grad_norm": 1.5022886991500854, + "learning_rate": 6.382930143264383e-05, + "loss": 0.42, + "step": 36274 + }, + { + "epoch": 2.517174380681424, + "grad_norm": 1.266139030456543, + "learning_rate": 6.382406330044115e-05, + "loss": 0.436, + "step": 36275 + }, + { + "epoch": 2.517243772118521, + "grad_norm": 1.0467206239700317, + "learning_rate": 6.381882528244174e-05, + "loss": 0.3713, + "step": 36276 + }, + { + "epoch": 2.517313163555617, + "grad_norm": 1.074163794517517, + "learning_rate": 6.381358737866207e-05, + "loss": 0.4502, + "step": 36277 + }, + { + "epoch": 2.517382554992714, + "grad_norm": 1.459949016571045, + "learning_rate": 6.38083495891187e-05, + "loss": 0.3727, + "step": 36278 + }, + { + "epoch": 2.5174519464298104, + "grad_norm": 1.8012206554412842, + "learning_rate": 6.380311191382818e-05, + "loss": 0.627, + "step": 36279 + }, + { + "epoch": 2.5175213378669072, + "grad_norm": 1.1312206983566284, + "learning_rate": 6.379787435280705e-05, + "loss": 0.3117, + "step": 36280 + }, + { + "epoch": 2.517590729304004, + "grad_norm": 1.2796871662139893, + "learning_rate": 6.37926369060718e-05, + "loss": 0.3721, + "step": 36281 + }, + { + "epoch": 2.5176601207411005, + "grad_norm": 1.6886203289031982, + "learning_rate": 6.3787399573639e-05, + "loss": 0.4496, + "step": 36282 + }, + { + "epoch": 2.5177295121781973, + "grad_norm": 0.9987310171127319, + "learning_rate": 6.378216235552523e-05, + "loss": 0.3691, + "step": 36283 + }, + { + "epoch": 2.5177989036152937, + "grad_norm": 0.9067211747169495, + "learning_rate": 6.377692525174692e-05, + "loss": 0.3623, + "step": 36284 + }, + { + "epoch": 2.5178682950523905, + "grad_norm": 1.5646028518676758, + "learning_rate": 6.377168826232067e-05, + "loss": 0.6222, + "step": 36285 + }, + { + "epoch": 2.5179376864894873, + "grad_norm": 1.3716342449188232, + "learning_rate": 6.376645138726299e-05, + "loss": 0.4631, + "step": 36286 + }, + { + "epoch": 2.5180070779265837, + "grad_norm": 1.50757896900177, + "learning_rate": 6.376121462659041e-05, + "loss": 0.4451, + "step": 36287 + }, + { + "epoch": 2.5180764693636806, + "grad_norm": 1.0479776859283447, + "learning_rate": 6.375597798031949e-05, + "loss": 0.287, + "step": 36288 + }, + { + "epoch": 2.518145860800777, + "grad_norm": 1.3959121704101562, + "learning_rate": 6.375074144846673e-05, + "loss": 0.4324, + "step": 36289 + }, + { + "epoch": 2.518215252237874, + "grad_norm": 1.4395627975463867, + "learning_rate": 6.37455050310487e-05, + "loss": 0.4154, + "step": 36290 + }, + { + "epoch": 2.5182846436749706, + "grad_norm": 1.0331284999847412, + "learning_rate": 6.374026872808187e-05, + "loss": 0.3881, + "step": 36291 + }, + { + "epoch": 2.518354035112067, + "grad_norm": 1.0850322246551514, + "learning_rate": 6.373503253958284e-05, + "loss": 0.4073, + "step": 36292 + }, + { + "epoch": 2.518423426549164, + "grad_norm": 0.8922473192214966, + "learning_rate": 6.372979646556808e-05, + "loss": 0.2263, + "step": 36293 + }, + { + "epoch": 2.5184928179862602, + "grad_norm": 1.0482409000396729, + "learning_rate": 6.372456050605414e-05, + "loss": 0.3703, + "step": 36294 + }, + { + "epoch": 2.518562209423357, + "grad_norm": 1.3095203638076782, + "learning_rate": 6.371932466105755e-05, + "loss": 0.3984, + "step": 36295 + }, + { + "epoch": 2.518631600860454, + "grad_norm": 1.0725765228271484, + "learning_rate": 6.371408893059486e-05, + "loss": 0.4016, + "step": 36296 + }, + { + "epoch": 2.5187009922975507, + "grad_norm": 1.7856215238571167, + "learning_rate": 6.37088533146826e-05, + "loss": 0.5254, + "step": 36297 + }, + { + "epoch": 2.518770383734647, + "grad_norm": 1.037917137145996, + "learning_rate": 6.370361781333726e-05, + "loss": 0.3305, + "step": 36298 + }, + { + "epoch": 2.518839775171744, + "grad_norm": 1.3441569805145264, + "learning_rate": 6.369838242657538e-05, + "loss": 0.3594, + "step": 36299 + }, + { + "epoch": 2.5189091666088403, + "grad_norm": 1.7254140377044678, + "learning_rate": 6.369314715441348e-05, + "loss": 0.4737, + "step": 36300 + }, + { + "epoch": 2.518978558045937, + "grad_norm": 0.9350656867027283, + "learning_rate": 6.368791199686812e-05, + "loss": 0.3053, + "step": 36301 + }, + { + "epoch": 2.519047949483034, + "grad_norm": 1.4737670421600342, + "learning_rate": 6.36826769539558e-05, + "loss": 0.444, + "step": 36302 + }, + { + "epoch": 2.5191173409201304, + "grad_norm": 1.4964678287506104, + "learning_rate": 6.367744202569306e-05, + "loss": 0.2791, + "step": 36303 + }, + { + "epoch": 2.5191867323572272, + "grad_norm": 1.161903738975525, + "learning_rate": 6.367220721209646e-05, + "loss": 0.3022, + "step": 36304 + }, + { + "epoch": 2.5192561237943236, + "grad_norm": 1.2550560235977173, + "learning_rate": 6.366697251318243e-05, + "loss": 0.4751, + "step": 36305 + }, + { + "epoch": 2.5193255152314205, + "grad_norm": 1.2939907312393188, + "learning_rate": 6.366173792896758e-05, + "loss": 0.4077, + "step": 36306 + }, + { + "epoch": 2.5193949066685173, + "grad_norm": 1.386499047279358, + "learning_rate": 6.365650345946841e-05, + "loss": 0.5274, + "step": 36307 + }, + { + "epoch": 2.5194642981056137, + "grad_norm": 1.5328013896942139, + "learning_rate": 6.365126910470143e-05, + "loss": 0.5601, + "step": 36308 + }, + { + "epoch": 2.5195336895427105, + "grad_norm": 1.0990535020828247, + "learning_rate": 6.364603486468317e-05, + "loss": 0.3298, + "step": 36309 + }, + { + "epoch": 2.519603080979807, + "grad_norm": 1.5853941440582275, + "learning_rate": 6.364080073943017e-05, + "loss": 0.5466, + "step": 36310 + }, + { + "epoch": 2.5196724724169037, + "grad_norm": 1.1169662475585938, + "learning_rate": 6.363556672895898e-05, + "loss": 0.3389, + "step": 36311 + }, + { + "epoch": 2.5197418638540006, + "grad_norm": 1.1093379259109497, + "learning_rate": 6.363033283328605e-05, + "loss": 0.4638, + "step": 36312 + }, + { + "epoch": 2.519811255291097, + "grad_norm": 1.69455087184906, + "learning_rate": 6.362509905242792e-05, + "loss": 0.5882, + "step": 36313 + }, + { + "epoch": 2.519880646728194, + "grad_norm": 1.0512235164642334, + "learning_rate": 6.361986538640114e-05, + "loss": 0.3774, + "step": 36314 + }, + { + "epoch": 2.51995003816529, + "grad_norm": 1.0275492668151855, + "learning_rate": 6.361463183522225e-05, + "loss": 0.4245, + "step": 36315 + }, + { + "epoch": 2.520019429602387, + "grad_norm": 1.8278729915618896, + "learning_rate": 6.360939839890772e-05, + "loss": 0.5774, + "step": 36316 + }, + { + "epoch": 2.520088821039484, + "grad_norm": 1.0841330289840698, + "learning_rate": 6.360416507747415e-05, + "loss": 0.4088, + "step": 36317 + }, + { + "epoch": 2.5201582124765802, + "grad_norm": 1.6452343463897705, + "learning_rate": 6.359893187093796e-05, + "loss": 0.4904, + "step": 36318 + }, + { + "epoch": 2.520227603913677, + "grad_norm": 1.5536330938339233, + "learning_rate": 6.359369877931571e-05, + "loss": 0.5155, + "step": 36319 + }, + { + "epoch": 2.5202969953507735, + "grad_norm": 1.6965302228927612, + "learning_rate": 6.358846580262395e-05, + "loss": 0.5386, + "step": 36320 + }, + { + "epoch": 2.5203663867878703, + "grad_norm": 1.6771621704101562, + "learning_rate": 6.358323294087916e-05, + "loss": 0.6333, + "step": 36321 + }, + { + "epoch": 2.520435778224967, + "grad_norm": 1.0682227611541748, + "learning_rate": 6.357800019409788e-05, + "loss": 0.3179, + "step": 36322 + }, + { + "epoch": 2.520505169662064, + "grad_norm": 1.6569015979766846, + "learning_rate": 6.357276756229666e-05, + "loss": 0.4881, + "step": 36323 + }, + { + "epoch": 2.5205745610991603, + "grad_norm": 1.175998330116272, + "learning_rate": 6.356753504549197e-05, + "loss": 0.3225, + "step": 36324 + }, + { + "epoch": 2.520643952536257, + "grad_norm": 1.03213369846344, + "learning_rate": 6.356230264370034e-05, + "loss": 0.397, + "step": 36325 + }, + { + "epoch": 2.5207133439733536, + "grad_norm": 1.5040838718414307, + "learning_rate": 6.35570703569383e-05, + "loss": 0.3849, + "step": 36326 + }, + { + "epoch": 2.5207827354104504, + "grad_norm": 1.6046364307403564, + "learning_rate": 6.355183818522236e-05, + "loss": 0.4219, + "step": 36327 + }, + { + "epoch": 2.5208521268475472, + "grad_norm": 1.1556618213653564, + "learning_rate": 6.354660612856903e-05, + "loss": 0.4011, + "step": 36328 + }, + { + "epoch": 2.5209215182846436, + "grad_norm": 1.183638572692871, + "learning_rate": 6.354137418699483e-05, + "loss": 0.2383, + "step": 36329 + }, + { + "epoch": 2.5209909097217404, + "grad_norm": 0.8742333054542542, + "learning_rate": 6.353614236051629e-05, + "loss": 0.2796, + "step": 36330 + }, + { + "epoch": 2.521060301158837, + "grad_norm": 1.3263627290725708, + "learning_rate": 6.353091064914996e-05, + "loss": 0.4398, + "step": 36331 + }, + { + "epoch": 2.5211296925959337, + "grad_norm": 1.1220752000808716, + "learning_rate": 6.352567905291225e-05, + "loss": 0.3282, + "step": 36332 + }, + { + "epoch": 2.5211990840330305, + "grad_norm": 1.7535779476165771, + "learning_rate": 6.352044757181976e-05, + "loss": 0.4453, + "step": 36333 + }, + { + "epoch": 2.521268475470127, + "grad_norm": 1.147842288017273, + "learning_rate": 6.351521620588898e-05, + "loss": 0.3225, + "step": 36334 + }, + { + "epoch": 2.5213378669072237, + "grad_norm": 1.262523889541626, + "learning_rate": 6.350998495513643e-05, + "loss": 0.4183, + "step": 36335 + }, + { + "epoch": 2.52140725834432, + "grad_norm": 1.4520319700241089, + "learning_rate": 6.350475381957862e-05, + "loss": 0.4572, + "step": 36336 + }, + { + "epoch": 2.521476649781417, + "grad_norm": 1.477828025817871, + "learning_rate": 6.349952279923207e-05, + "loss": 0.3933, + "step": 36337 + }, + { + "epoch": 2.521546041218514, + "grad_norm": 2.1914501190185547, + "learning_rate": 6.349429189411333e-05, + "loss": 0.689, + "step": 36338 + }, + { + "epoch": 2.52161543265561, + "grad_norm": 1.3657917976379395, + "learning_rate": 6.348906110423884e-05, + "loss": 0.4259, + "step": 36339 + }, + { + "epoch": 2.521684824092707, + "grad_norm": 1.3898080587387085, + "learning_rate": 6.348383042962515e-05, + "loss": 0.3474, + "step": 36340 + }, + { + "epoch": 2.5217542155298034, + "grad_norm": 1.624498963356018, + "learning_rate": 6.347859987028877e-05, + "loss": 0.4995, + "step": 36341 + }, + { + "epoch": 2.5218236069669, + "grad_norm": 1.4919564723968506, + "learning_rate": 6.347336942624622e-05, + "loss": 0.4345, + "step": 36342 + }, + { + "epoch": 2.521892998403997, + "grad_norm": 1.9533429145812988, + "learning_rate": 6.3468139097514e-05, + "loss": 0.6064, + "step": 36343 + }, + { + "epoch": 2.5219623898410934, + "grad_norm": 1.1245795488357544, + "learning_rate": 6.346290888410862e-05, + "loss": 0.3751, + "step": 36344 + }, + { + "epoch": 2.5220317812781903, + "grad_norm": 1.1976224184036255, + "learning_rate": 6.345767878604662e-05, + "loss": 0.3766, + "step": 36345 + }, + { + "epoch": 2.5221011727152867, + "grad_norm": 1.0853049755096436, + "learning_rate": 6.345244880334449e-05, + "loss": 0.3836, + "step": 36346 + }, + { + "epoch": 2.5221705641523835, + "grad_norm": 1.5306651592254639, + "learning_rate": 6.34472189360187e-05, + "loss": 0.3719, + "step": 36347 + }, + { + "epoch": 2.5222399555894803, + "grad_norm": 1.6264334917068481, + "learning_rate": 6.344198918408583e-05, + "loss": 0.4387, + "step": 36348 + }, + { + "epoch": 2.522309347026577, + "grad_norm": 1.3221688270568848, + "learning_rate": 6.343675954756234e-05, + "loss": 0.4962, + "step": 36349 + }, + { + "epoch": 2.5223787384636736, + "grad_norm": 1.2017611265182495, + "learning_rate": 6.343153002646476e-05, + "loss": 0.3842, + "step": 36350 + }, + { + "epoch": 2.5224481299007704, + "grad_norm": 1.4420360326766968, + "learning_rate": 6.34263006208096e-05, + "loss": 0.4858, + "step": 36351 + }, + { + "epoch": 2.5225175213378668, + "grad_norm": 1.7638766765594482, + "learning_rate": 6.34210713306134e-05, + "loss": 0.4528, + "step": 36352 + }, + { + "epoch": 2.5225869127749636, + "grad_norm": 2.008253335952759, + "learning_rate": 6.34158421558926e-05, + "loss": 0.6755, + "step": 36353 + }, + { + "epoch": 2.5226563042120604, + "grad_norm": 0.9299045205116272, + "learning_rate": 6.341061309666375e-05, + "loss": 0.3613, + "step": 36354 + }, + { + "epoch": 2.522725695649157, + "grad_norm": 1.105514407157898, + "learning_rate": 6.340538415294332e-05, + "loss": 0.3487, + "step": 36355 + }, + { + "epoch": 2.5227950870862537, + "grad_norm": 1.2326651811599731, + "learning_rate": 6.340015532474786e-05, + "loss": 0.3617, + "step": 36356 + }, + { + "epoch": 2.52286447852335, + "grad_norm": 1.3883675336837769, + "learning_rate": 6.339492661209387e-05, + "loss": 0.432, + "step": 36357 + }, + { + "epoch": 2.522933869960447, + "grad_norm": 1.1886736154556274, + "learning_rate": 6.338969801499788e-05, + "loss": 0.4055, + "step": 36358 + }, + { + "epoch": 2.5230032613975437, + "grad_norm": 1.6143397092819214, + "learning_rate": 6.338446953347634e-05, + "loss": 0.4804, + "step": 36359 + }, + { + "epoch": 2.52307265283464, + "grad_norm": 1.142167329788208, + "learning_rate": 6.337924116754579e-05, + "loss": 0.3808, + "step": 36360 + }, + { + "epoch": 2.523142044271737, + "grad_norm": 1.1216269731521606, + "learning_rate": 6.337401291722273e-05, + "loss": 0.457, + "step": 36361 + }, + { + "epoch": 2.5232114357088333, + "grad_norm": 1.5277378559112549, + "learning_rate": 6.336878478252364e-05, + "loss": 0.4022, + "step": 36362 + }, + { + "epoch": 2.52328082714593, + "grad_norm": 1.126434326171875, + "learning_rate": 6.336355676346505e-05, + "loss": 0.5049, + "step": 36363 + }, + { + "epoch": 2.523350218583027, + "grad_norm": 1.458249807357788, + "learning_rate": 6.335832886006346e-05, + "loss": 0.3026, + "step": 36364 + }, + { + "epoch": 2.5234196100201234, + "grad_norm": 1.1247659921646118, + "learning_rate": 6.335310107233538e-05, + "loss": 0.3144, + "step": 36365 + }, + { + "epoch": 2.52348900145722, + "grad_norm": 0.8326703906059265, + "learning_rate": 6.334787340029735e-05, + "loss": 0.3793, + "step": 36366 + }, + { + "epoch": 2.5235583928943166, + "grad_norm": 1.4695490598678589, + "learning_rate": 6.334264584396578e-05, + "loss": 0.4205, + "step": 36367 + }, + { + "epoch": 2.5236277843314134, + "grad_norm": 0.9180739521980286, + "learning_rate": 6.333741840335723e-05, + "loss": 0.3056, + "step": 36368 + }, + { + "epoch": 2.5236971757685103, + "grad_norm": 1.1137926578521729, + "learning_rate": 6.333219107848819e-05, + "loss": 0.3106, + "step": 36369 + }, + { + "epoch": 2.523766567205607, + "grad_norm": 1.2611393928527832, + "learning_rate": 6.332696386937518e-05, + "loss": 0.2509, + "step": 36370 + }, + { + "epoch": 2.5238359586427035, + "grad_norm": 1.8159139156341553, + "learning_rate": 6.332173677603469e-05, + "loss": 0.4529, + "step": 36371 + }, + { + "epoch": 2.5239053500798, + "grad_norm": 1.3680952787399292, + "learning_rate": 6.33165097984832e-05, + "loss": 0.2805, + "step": 36372 + }, + { + "epoch": 2.5239747415168967, + "grad_norm": 1.2011620998382568, + "learning_rate": 6.331128293673729e-05, + "loss": 0.4698, + "step": 36373 + }, + { + "epoch": 2.5240441329539935, + "grad_norm": 1.31033456325531, + "learning_rate": 6.330605619081336e-05, + "loss": 0.3602, + "step": 36374 + }, + { + "epoch": 2.5241135243910904, + "grad_norm": 1.1810017824172974, + "learning_rate": 6.330082956072797e-05, + "loss": 0.2653, + "step": 36375 + }, + { + "epoch": 2.5241829158281868, + "grad_norm": 1.433320164680481, + "learning_rate": 6.329560304649758e-05, + "loss": 0.3716, + "step": 36376 + }, + { + "epoch": 2.5242523072652836, + "grad_norm": 1.3377115726470947, + "learning_rate": 6.329037664813874e-05, + "loss": 0.5172, + "step": 36377 + }, + { + "epoch": 2.52432169870238, + "grad_norm": 1.0836291313171387, + "learning_rate": 6.32851503656679e-05, + "loss": 0.4594, + "step": 36378 + }, + { + "epoch": 2.524391090139477, + "grad_norm": 1.1953377723693848, + "learning_rate": 6.327992419910158e-05, + "loss": 0.3396, + "step": 36379 + }, + { + "epoch": 2.5244604815765737, + "grad_norm": 1.6399402618408203, + "learning_rate": 6.32746981484563e-05, + "loss": 0.4868, + "step": 36380 + }, + { + "epoch": 2.52452987301367, + "grad_norm": 1.2913461923599243, + "learning_rate": 6.326947221374855e-05, + "loss": 0.4886, + "step": 36381 + }, + { + "epoch": 2.524599264450767, + "grad_norm": 1.6327887773513794, + "learning_rate": 6.326424639499478e-05, + "loss": 0.4606, + "step": 36382 + }, + { + "epoch": 2.5246686558878633, + "grad_norm": 1.4444854259490967, + "learning_rate": 6.325902069221152e-05, + "loss": 0.4394, + "step": 36383 + }, + { + "epoch": 2.52473804732496, + "grad_norm": 0.8657618761062622, + "learning_rate": 6.325379510541526e-05, + "loss": 0.3137, + "step": 36384 + }, + { + "epoch": 2.524807438762057, + "grad_norm": 2.381157398223877, + "learning_rate": 6.324856963462254e-05, + "loss": 0.4278, + "step": 36385 + }, + { + "epoch": 2.5248768301991533, + "grad_norm": 1.2430427074432373, + "learning_rate": 6.324334427984979e-05, + "loss": 0.4035, + "step": 36386 + }, + { + "epoch": 2.52494622163625, + "grad_norm": 1.065925121307373, + "learning_rate": 6.32381190411136e-05, + "loss": 0.3333, + "step": 36387 + }, + { + "epoch": 2.5250156130733465, + "grad_norm": 1.7204424142837524, + "learning_rate": 6.323289391843035e-05, + "loss": 0.512, + "step": 36388 + }, + { + "epoch": 2.5250850045104434, + "grad_norm": 1.0469706058502197, + "learning_rate": 6.32276689118166e-05, + "loss": 0.3892, + "step": 36389 + }, + { + "epoch": 2.52515439594754, + "grad_norm": 1.382140040397644, + "learning_rate": 6.322244402128881e-05, + "loss": 0.4142, + "step": 36390 + }, + { + "epoch": 2.5252237873846366, + "grad_norm": 0.9118960499763489, + "learning_rate": 6.321721924686352e-05, + "loss": 0.3039, + "step": 36391 + }, + { + "epoch": 2.5252931788217334, + "grad_norm": 1.1972301006317139, + "learning_rate": 6.321199458855717e-05, + "loss": 0.5301, + "step": 36392 + }, + { + "epoch": 2.52536257025883, + "grad_norm": 1.2056785821914673, + "learning_rate": 6.320677004638636e-05, + "loss": 0.5141, + "step": 36393 + }, + { + "epoch": 2.5254319616959267, + "grad_norm": 1.1215578317642212, + "learning_rate": 6.320154562036747e-05, + "loss": 0.4101, + "step": 36394 + }, + { + "epoch": 2.5255013531330235, + "grad_norm": 1.2994836568832397, + "learning_rate": 6.319632131051701e-05, + "loss": 0.3164, + "step": 36395 + }, + { + "epoch": 2.5255707445701203, + "grad_norm": 0.824547529220581, + "learning_rate": 6.319109711685152e-05, + "loss": 0.2218, + "step": 36396 + }, + { + "epoch": 2.5256401360072167, + "grad_norm": 1.437996745109558, + "learning_rate": 6.318587303938743e-05, + "loss": 0.4547, + "step": 36397 + }, + { + "epoch": 2.525709527444313, + "grad_norm": 1.2697776556015015, + "learning_rate": 6.31806490781413e-05, + "loss": 0.4199, + "step": 36398 + }, + { + "epoch": 2.52577891888141, + "grad_norm": 1.2796785831451416, + "learning_rate": 6.317542523312956e-05, + "loss": 0.3333, + "step": 36399 + }, + { + "epoch": 2.5258483103185068, + "grad_norm": 1.5864338874816895, + "learning_rate": 6.317020150436878e-05, + "loss": 0.5088, + "step": 36400 + }, + { + "epoch": 2.5259177017556036, + "grad_norm": 1.4281076192855835, + "learning_rate": 6.316497789187537e-05, + "loss": 0.3889, + "step": 36401 + }, + { + "epoch": 2.5259870931927, + "grad_norm": 1.2062528133392334, + "learning_rate": 6.315975439566583e-05, + "loss": 0.3927, + "step": 36402 + }, + { + "epoch": 2.526056484629797, + "grad_norm": 1.1195287704467773, + "learning_rate": 6.31545310157567e-05, + "loss": 0.3974, + "step": 36403 + }, + { + "epoch": 2.526125876066893, + "grad_norm": 1.1112757921218872, + "learning_rate": 6.314930775216442e-05, + "loss": 0.3361, + "step": 36404 + }, + { + "epoch": 2.52619526750399, + "grad_norm": 1.234535574913025, + "learning_rate": 6.314408460490551e-05, + "loss": 0.3986, + "step": 36405 + }, + { + "epoch": 2.526264658941087, + "grad_norm": 1.126030445098877, + "learning_rate": 6.313886157399645e-05, + "loss": 0.2811, + "step": 36406 + }, + { + "epoch": 2.5263340503781833, + "grad_norm": 1.9563499689102173, + "learning_rate": 6.313363865945375e-05, + "loss": 0.6758, + "step": 36407 + }, + { + "epoch": 2.52640344181528, + "grad_norm": 1.0474281311035156, + "learning_rate": 6.312841586129386e-05, + "loss": 0.3522, + "step": 36408 + }, + { + "epoch": 2.5264728332523765, + "grad_norm": 1.395068883895874, + "learning_rate": 6.312319317953325e-05, + "loss": 0.434, + "step": 36409 + }, + { + "epoch": 2.5265422246894733, + "grad_norm": 1.5241838693618774, + "learning_rate": 6.311797061418847e-05, + "loss": 0.3851, + "step": 36410 + }, + { + "epoch": 2.52661161612657, + "grad_norm": 1.1466243267059326, + "learning_rate": 6.311274816527598e-05, + "loss": 0.3117, + "step": 36411 + }, + { + "epoch": 2.5266810075636665, + "grad_norm": 1.678823709487915, + "learning_rate": 6.310752583281226e-05, + "loss": 0.4668, + "step": 36412 + }, + { + "epoch": 2.5267503990007634, + "grad_norm": 1.0765289068222046, + "learning_rate": 6.31023036168138e-05, + "loss": 0.4085, + "step": 36413 + }, + { + "epoch": 2.5268197904378598, + "grad_norm": 1.9147117137908936, + "learning_rate": 6.30970815172971e-05, + "loss": 0.3298, + "step": 36414 + }, + { + "epoch": 2.5268891818749566, + "grad_norm": 1.5647945404052734, + "learning_rate": 6.309185953427863e-05, + "loss": 0.5533, + "step": 36415 + }, + { + "epoch": 2.5269585733120534, + "grad_norm": 1.3762867450714111, + "learning_rate": 6.308663766777486e-05, + "loss": 0.512, + "step": 36416 + }, + { + "epoch": 2.52702796474915, + "grad_norm": 1.4298878908157349, + "learning_rate": 6.30814159178023e-05, + "loss": 0.3959, + "step": 36417 + }, + { + "epoch": 2.5270973561862466, + "grad_norm": 1.44021475315094, + "learning_rate": 6.307619428437741e-05, + "loss": 0.4284, + "step": 36418 + }, + { + "epoch": 2.527166747623343, + "grad_norm": 1.0607941150665283, + "learning_rate": 6.30709727675167e-05, + "loss": 0.318, + "step": 36419 + }, + { + "epoch": 2.52723613906044, + "grad_norm": 0.8811218738555908, + "learning_rate": 6.306575136723665e-05, + "loss": 0.2868, + "step": 36420 + }, + { + "epoch": 2.5273055304975367, + "grad_norm": 0.9229932427406311, + "learning_rate": 6.306053008355378e-05, + "loss": 0.3326, + "step": 36421 + }, + { + "epoch": 2.5273749219346335, + "grad_norm": 1.4018787145614624, + "learning_rate": 6.305530891648447e-05, + "loss": 0.3983, + "step": 36422 + }, + { + "epoch": 2.52744431337173, + "grad_norm": 1.085906744003296, + "learning_rate": 6.305008786604526e-05, + "loss": 0.3373, + "step": 36423 + }, + { + "epoch": 2.5275137048088268, + "grad_norm": 1.1796363592147827, + "learning_rate": 6.304486693225267e-05, + "loss": 0.5222, + "step": 36424 + }, + { + "epoch": 2.527583096245923, + "grad_norm": 0.8773688077926636, + "learning_rate": 6.303964611512312e-05, + "loss": 0.3105, + "step": 36425 + }, + { + "epoch": 2.52765248768302, + "grad_norm": 1.2233995199203491, + "learning_rate": 6.303442541467313e-05, + "loss": 0.398, + "step": 36426 + }, + { + "epoch": 2.527721879120117, + "grad_norm": 1.0445775985717773, + "learning_rate": 6.302920483091919e-05, + "loss": 0.3419, + "step": 36427 + }, + { + "epoch": 2.527791270557213, + "grad_norm": 1.1203718185424805, + "learning_rate": 6.302398436387776e-05, + "loss": 0.3968, + "step": 36428 + }, + { + "epoch": 2.52786066199431, + "grad_norm": 1.767724633216858, + "learning_rate": 6.30187640135653e-05, + "loss": 0.4395, + "step": 36429 + }, + { + "epoch": 2.5279300534314064, + "grad_norm": 1.2180421352386475, + "learning_rate": 6.301354377999831e-05, + "loss": 0.2521, + "step": 36430 + }, + { + "epoch": 2.5279994448685033, + "grad_norm": 1.1874370574951172, + "learning_rate": 6.30083236631933e-05, + "loss": 0.3364, + "step": 36431 + }, + { + "epoch": 2.5280688363056, + "grad_norm": 1.689942717552185, + "learning_rate": 6.300310366316668e-05, + "loss": 0.444, + "step": 36432 + }, + { + "epoch": 2.5281382277426965, + "grad_norm": 1.4824196100234985, + "learning_rate": 6.299788377993498e-05, + "loss": 0.3892, + "step": 36433 + }, + { + "epoch": 2.5282076191797933, + "grad_norm": 1.1088500022888184, + "learning_rate": 6.299266401351467e-05, + "loss": 0.357, + "step": 36434 + }, + { + "epoch": 2.5282770106168897, + "grad_norm": 1.2421902418136597, + "learning_rate": 6.298744436392227e-05, + "loss": 0.42, + "step": 36435 + }, + { + "epoch": 2.5283464020539865, + "grad_norm": 1.5055336952209473, + "learning_rate": 6.298222483117417e-05, + "loss": 0.4816, + "step": 36436 + }, + { + "epoch": 2.5284157934910834, + "grad_norm": 1.17872154712677, + "learning_rate": 6.297700541528689e-05, + "loss": 0.5327, + "step": 36437 + }, + { + "epoch": 2.5284851849281798, + "grad_norm": 1.4783085584640503, + "learning_rate": 6.297178611627692e-05, + "loss": 0.4591, + "step": 36438 + }, + { + "epoch": 2.5285545763652766, + "grad_norm": 1.4297542572021484, + "learning_rate": 6.29665669341607e-05, + "loss": 0.5888, + "step": 36439 + }, + { + "epoch": 2.528623967802373, + "grad_norm": 1.428486943244934, + "learning_rate": 6.296134786895476e-05, + "loss": 0.3924, + "step": 36440 + }, + { + "epoch": 2.52869335923947, + "grad_norm": 1.1882290840148926, + "learning_rate": 6.295612892067552e-05, + "loss": 0.388, + "step": 36441 + }, + { + "epoch": 2.5287627506765666, + "grad_norm": 1.5334967374801636, + "learning_rate": 6.295091008933955e-05, + "loss": 0.5281, + "step": 36442 + }, + { + "epoch": 2.528832142113663, + "grad_norm": 1.0959954261779785, + "learning_rate": 6.29456913749632e-05, + "loss": 0.4127, + "step": 36443 + }, + { + "epoch": 2.52890153355076, + "grad_norm": 0.9763160347938538, + "learning_rate": 6.2940472777563e-05, + "loss": 0.2324, + "step": 36444 + }, + { + "epoch": 2.5289709249878563, + "grad_norm": 1.0163484811782837, + "learning_rate": 6.293525429715546e-05, + "loss": 0.3253, + "step": 36445 + }, + { + "epoch": 2.529040316424953, + "grad_norm": 1.1734551191329956, + "learning_rate": 6.2930035933757e-05, + "loss": 0.4752, + "step": 36446 + }, + { + "epoch": 2.52910970786205, + "grad_norm": 1.5812745094299316, + "learning_rate": 6.292481768738412e-05, + "loss": 0.4524, + "step": 36447 + }, + { + "epoch": 2.5291790992991467, + "grad_norm": 1.21164870262146, + "learning_rate": 6.291959955805328e-05, + "loss": 0.5403, + "step": 36448 + }, + { + "epoch": 2.529248490736243, + "grad_norm": 1.5771571397781372, + "learning_rate": 6.291438154578099e-05, + "loss": 0.4759, + "step": 36449 + }, + { + "epoch": 2.52931788217334, + "grad_norm": 1.3141999244689941, + "learning_rate": 6.290916365058368e-05, + "loss": 0.4645, + "step": 36450 + }, + { + "epoch": 2.5293872736104364, + "grad_norm": 1.0752683877944946, + "learning_rate": 6.290394587247783e-05, + "loss": 0.3301, + "step": 36451 + }, + { + "epoch": 2.529456665047533, + "grad_norm": 1.8940571546554565, + "learning_rate": 6.289872821147992e-05, + "loss": 0.3448, + "step": 36452 + }, + { + "epoch": 2.52952605648463, + "grad_norm": 1.534909725189209, + "learning_rate": 6.289351066760641e-05, + "loss": 0.4562, + "step": 36453 + }, + { + "epoch": 2.5295954479217264, + "grad_norm": 1.2540438175201416, + "learning_rate": 6.28882932408738e-05, + "loss": 0.3788, + "step": 36454 + }, + { + "epoch": 2.5296648393588232, + "grad_norm": 1.6668040752410889, + "learning_rate": 6.288307593129852e-05, + "loss": 0.7174, + "step": 36455 + }, + { + "epoch": 2.5297342307959196, + "grad_norm": 1.331001877784729, + "learning_rate": 6.28778587388971e-05, + "loss": 0.244, + "step": 36456 + }, + { + "epoch": 2.5298036222330165, + "grad_norm": 1.323566198348999, + "learning_rate": 6.287264166368596e-05, + "loss": 0.579, + "step": 36457 + }, + { + "epoch": 2.5298730136701133, + "grad_norm": 1.1015828847885132, + "learning_rate": 6.286742470568156e-05, + "loss": 0.4515, + "step": 36458 + }, + { + "epoch": 2.5299424051072097, + "grad_norm": 1.283338189125061, + "learning_rate": 6.286220786490042e-05, + "loss": 0.4719, + "step": 36459 + }, + { + "epoch": 2.5300117965443065, + "grad_norm": 1.400364637374878, + "learning_rate": 6.285699114135895e-05, + "loss": 0.3215, + "step": 36460 + }, + { + "epoch": 2.530081187981403, + "grad_norm": 1.6478580236434937, + "learning_rate": 6.285177453507368e-05, + "loss": 0.3753, + "step": 36461 + }, + { + "epoch": 2.5301505794184997, + "grad_norm": 1.059557318687439, + "learning_rate": 6.284655804606104e-05, + "loss": 0.4226, + "step": 36462 + }, + { + "epoch": 2.5302199708555966, + "grad_norm": 0.9341456890106201, + "learning_rate": 6.284134167433751e-05, + "loss": 0.3949, + "step": 36463 + }, + { + "epoch": 2.530289362292693, + "grad_norm": 1.366722822189331, + "learning_rate": 6.283612541991955e-05, + "loss": 0.5449, + "step": 36464 + }, + { + "epoch": 2.53035875372979, + "grad_norm": 1.0047717094421387, + "learning_rate": 6.28309092828236e-05, + "loss": 0.2837, + "step": 36465 + }, + { + "epoch": 2.530428145166886, + "grad_norm": 1.7011661529541016, + "learning_rate": 6.28256932630662e-05, + "loss": 0.5925, + "step": 36466 + }, + { + "epoch": 2.530497536603983, + "grad_norm": 1.2566771507263184, + "learning_rate": 6.282047736066376e-05, + "loss": 0.3311, + "step": 36467 + }, + { + "epoch": 2.53056692804108, + "grad_norm": 1.1357862949371338, + "learning_rate": 6.281526157563275e-05, + "loss": 0.3971, + "step": 36468 + }, + { + "epoch": 2.5306363194781762, + "grad_norm": 1.5574003458023071, + "learning_rate": 6.281004590798968e-05, + "loss": 0.4861, + "step": 36469 + }, + { + "epoch": 2.530705710915273, + "grad_norm": 1.9609041213989258, + "learning_rate": 6.280483035775093e-05, + "loss": 0.4762, + "step": 36470 + }, + { + "epoch": 2.5307751023523695, + "grad_norm": 1.2931631803512573, + "learning_rate": 6.279961492493304e-05, + "loss": 0.3765, + "step": 36471 + }, + { + "epoch": 2.5308444937894663, + "grad_norm": 1.0696395635604858, + "learning_rate": 6.279439960955243e-05, + "loss": 0.3875, + "step": 36472 + }, + { + "epoch": 2.530913885226563, + "grad_norm": 1.0995080471038818, + "learning_rate": 6.27891844116256e-05, + "loss": 0.309, + "step": 36473 + }, + { + "epoch": 2.53098327666366, + "grad_norm": 1.5223113298416138, + "learning_rate": 6.278396933116899e-05, + "loss": 0.4023, + "step": 36474 + }, + { + "epoch": 2.5310526681007564, + "grad_norm": 0.9074804186820984, + "learning_rate": 6.277875436819907e-05, + "loss": 0.3315, + "step": 36475 + }, + { + "epoch": 2.531122059537853, + "grad_norm": 0.9962441325187683, + "learning_rate": 6.277353952273235e-05, + "loss": 0.4447, + "step": 36476 + }, + { + "epoch": 2.5311914509749496, + "grad_norm": 0.9270637631416321, + "learning_rate": 6.276832479478519e-05, + "loss": 0.2896, + "step": 36477 + }, + { + "epoch": 2.5312608424120464, + "grad_norm": 1.1321253776550293, + "learning_rate": 6.276311018437413e-05, + "loss": 0.4022, + "step": 36478 + }, + { + "epoch": 2.5313302338491432, + "grad_norm": 1.2255630493164062, + "learning_rate": 6.27578956915156e-05, + "loss": 0.3551, + "step": 36479 + }, + { + "epoch": 2.5313996252862396, + "grad_norm": 0.856165885925293, + "learning_rate": 6.275268131622606e-05, + "loss": 0.2778, + "step": 36480 + }, + { + "epoch": 2.5314690167233365, + "grad_norm": 1.4157990217208862, + "learning_rate": 6.2747467058522e-05, + "loss": 0.4347, + "step": 36481 + }, + { + "epoch": 2.531538408160433, + "grad_norm": 1.9335322380065918, + "learning_rate": 6.274225291841988e-05, + "loss": 0.4435, + "step": 36482 + }, + { + "epoch": 2.5316077995975297, + "grad_norm": 1.275041937828064, + "learning_rate": 6.273703889593615e-05, + "loss": 0.3931, + "step": 36483 + }, + { + "epoch": 2.5316771910346265, + "grad_norm": 0.9912154078483582, + "learning_rate": 6.273182499108724e-05, + "loss": 0.3786, + "step": 36484 + }, + { + "epoch": 2.531746582471723, + "grad_norm": 0.8121829032897949, + "learning_rate": 6.272661120388965e-05, + "loss": 0.3154, + "step": 36485 + }, + { + "epoch": 2.5318159739088197, + "grad_norm": 1.1980607509613037, + "learning_rate": 6.272139753435981e-05, + "loss": 0.3746, + "step": 36486 + }, + { + "epoch": 2.531885365345916, + "grad_norm": 1.4220073223114014, + "learning_rate": 6.27161839825142e-05, + "loss": 0.3779, + "step": 36487 + }, + { + "epoch": 2.531954756783013, + "grad_norm": 1.6614068746566772, + "learning_rate": 6.271097054836927e-05, + "loss": 0.5471, + "step": 36488 + }, + { + "epoch": 2.53202414822011, + "grad_norm": 1.024654507637024, + "learning_rate": 6.270575723194146e-05, + "loss": 0.2278, + "step": 36489 + }, + { + "epoch": 2.532093539657206, + "grad_norm": 1.925251841545105, + "learning_rate": 6.27005440332473e-05, + "loss": 0.4387, + "step": 36490 + }, + { + "epoch": 2.532162931094303, + "grad_norm": 0.9812014698982239, + "learning_rate": 6.269533095230317e-05, + "loss": 0.3257, + "step": 36491 + }, + { + "epoch": 2.5322323225313994, + "grad_norm": 1.4601277112960815, + "learning_rate": 6.269011798912553e-05, + "loss": 0.3366, + "step": 36492 + }, + { + "epoch": 2.5323017139684962, + "grad_norm": 1.1584662199020386, + "learning_rate": 6.268490514373087e-05, + "loss": 0.3235, + "step": 36493 + }, + { + "epoch": 2.532371105405593, + "grad_norm": 1.2134908437728882, + "learning_rate": 6.267969241613563e-05, + "loss": 0.3584, + "step": 36494 + }, + { + "epoch": 2.5324404968426895, + "grad_norm": 0.8742434978485107, + "learning_rate": 6.267447980635628e-05, + "loss": 0.4108, + "step": 36495 + }, + { + "epoch": 2.5325098882797863, + "grad_norm": 0.9919306039810181, + "learning_rate": 6.266926731440925e-05, + "loss": 0.3419, + "step": 36496 + }, + { + "epoch": 2.5325792797168827, + "grad_norm": 1.0217986106872559, + "learning_rate": 6.266405494031107e-05, + "loss": 0.3345, + "step": 36497 + }, + { + "epoch": 2.5326486711539795, + "grad_norm": 1.173327088356018, + "learning_rate": 6.265884268407808e-05, + "loss": 0.4033, + "step": 36498 + }, + { + "epoch": 2.5327180625910763, + "grad_norm": 1.1547085046768188, + "learning_rate": 6.265363054572681e-05, + "loss": 0.4194, + "step": 36499 + }, + { + "epoch": 2.532787454028173, + "grad_norm": 1.1617836952209473, + "learning_rate": 6.264841852527368e-05, + "loss": 0.3846, + "step": 36500 + }, + { + "epoch": 2.5328568454652696, + "grad_norm": 1.6324248313903809, + "learning_rate": 6.264320662273519e-05, + "loss": 0.3998, + "step": 36501 + }, + { + "epoch": 2.5329262369023664, + "grad_norm": 0.9657561182975769, + "learning_rate": 6.263799483812773e-05, + "loss": 0.3342, + "step": 36502 + }, + { + "epoch": 2.532995628339463, + "grad_norm": 1.3438148498535156, + "learning_rate": 6.263278317146779e-05, + "loss": 0.3633, + "step": 36503 + }, + { + "epoch": 2.5330650197765596, + "grad_norm": 1.8836919069290161, + "learning_rate": 6.262757162277186e-05, + "loss": 0.5591, + "step": 36504 + }, + { + "epoch": 2.5331344112136565, + "grad_norm": 1.335882544517517, + "learning_rate": 6.262236019205631e-05, + "loss": 0.4404, + "step": 36505 + }, + { + "epoch": 2.533203802650753, + "grad_norm": 1.0156878232955933, + "learning_rate": 6.261714887933762e-05, + "loss": 0.359, + "step": 36506 + }, + { + "epoch": 2.5332731940878497, + "grad_norm": 1.2512884140014648, + "learning_rate": 6.261193768463228e-05, + "loss": 0.4115, + "step": 36507 + }, + { + "epoch": 2.533342585524946, + "grad_norm": 1.0558407306671143, + "learning_rate": 6.26067266079567e-05, + "loss": 0.2836, + "step": 36508 + }, + { + "epoch": 2.533411976962043, + "grad_norm": 1.6172065734863281, + "learning_rate": 6.260151564932735e-05, + "loss": 0.4914, + "step": 36509 + }, + { + "epoch": 2.5334813683991397, + "grad_norm": 1.2379388809204102, + "learning_rate": 6.259630480876068e-05, + "loss": 0.3878, + "step": 36510 + }, + { + "epoch": 2.533550759836236, + "grad_norm": 1.031054973602295, + "learning_rate": 6.259109408627317e-05, + "loss": 0.342, + "step": 36511 + }, + { + "epoch": 2.533620151273333, + "grad_norm": 1.811355471611023, + "learning_rate": 6.25858834818812e-05, + "loss": 0.6256, + "step": 36512 + }, + { + "epoch": 2.5336895427104293, + "grad_norm": 1.6723638772964478, + "learning_rate": 6.258067299560124e-05, + "loss": 0.502, + "step": 36513 + }, + { + "epoch": 2.533758934147526, + "grad_norm": 1.0940817594528198, + "learning_rate": 6.257546262744977e-05, + "loss": 0.3288, + "step": 36514 + }, + { + "epoch": 2.533828325584623, + "grad_norm": 2.0761077404022217, + "learning_rate": 6.257025237744321e-05, + "loss": 0.4746, + "step": 36515 + }, + { + "epoch": 2.5338977170217194, + "grad_norm": 0.8824291825294495, + "learning_rate": 6.256504224559805e-05, + "loss": 0.2064, + "step": 36516 + }, + { + "epoch": 2.5339671084588162, + "grad_norm": 1.4304759502410889, + "learning_rate": 6.255983223193068e-05, + "loss": 0.3689, + "step": 36517 + }, + { + "epoch": 2.5340364998959126, + "grad_norm": 1.2839220762252808, + "learning_rate": 6.25546223364576e-05, + "loss": 0.4567, + "step": 36518 + }, + { + "epoch": 2.5341058913330095, + "grad_norm": 1.096651554107666, + "learning_rate": 6.254941255919522e-05, + "loss": 0.2973, + "step": 36519 + }, + { + "epoch": 2.5341752827701063, + "grad_norm": 1.7134627103805542, + "learning_rate": 6.254420290016001e-05, + "loss": 0.4672, + "step": 36520 + }, + { + "epoch": 2.5342446742072027, + "grad_norm": 1.3903342485427856, + "learning_rate": 6.25389933593684e-05, + "loss": 0.4261, + "step": 36521 + }, + { + "epoch": 2.5343140656442995, + "grad_norm": 1.4909192323684692, + "learning_rate": 6.253378393683682e-05, + "loss": 0.5196, + "step": 36522 + }, + { + "epoch": 2.534383457081396, + "grad_norm": 1.3794869184494019, + "learning_rate": 6.252857463258175e-05, + "loss": 0.3621, + "step": 36523 + }, + { + "epoch": 2.5344528485184927, + "grad_norm": 1.2070554494857788, + "learning_rate": 6.252336544661962e-05, + "loss": 0.425, + "step": 36524 + }, + { + "epoch": 2.5345222399555896, + "grad_norm": 1.2630696296691895, + "learning_rate": 6.251815637896692e-05, + "loss": 0.4603, + "step": 36525 + }, + { + "epoch": 2.5345916313926864, + "grad_norm": 1.1915658712387085, + "learning_rate": 6.251294742964e-05, + "loss": 0.3757, + "step": 36526 + }, + { + "epoch": 2.534661022829783, + "grad_norm": 0.8722254633903503, + "learning_rate": 6.250773859865537e-05, + "loss": 0.3179, + "step": 36527 + }, + { + "epoch": 2.5347304142668796, + "grad_norm": 1.2376748323440552, + "learning_rate": 6.250252988602946e-05, + "loss": 0.4521, + "step": 36528 + }, + { + "epoch": 2.534799805703976, + "grad_norm": 0.922240138053894, + "learning_rate": 6.24973212917787e-05, + "loss": 0.3885, + "step": 36529 + }, + { + "epoch": 2.534869197141073, + "grad_norm": 1.425260305404663, + "learning_rate": 6.249211281591955e-05, + "loss": 0.5715, + "step": 36530 + }, + { + "epoch": 2.5349385885781697, + "grad_norm": 1.3196241855621338, + "learning_rate": 6.248690445846846e-05, + "loss": 0.4262, + "step": 36531 + }, + { + "epoch": 2.535007980015266, + "grad_norm": 0.8646925091743469, + "learning_rate": 6.24816962194419e-05, + "loss": 0.2845, + "step": 36532 + }, + { + "epoch": 2.535077371452363, + "grad_norm": 1.3830541372299194, + "learning_rate": 6.247648809885622e-05, + "loss": 0.4337, + "step": 36533 + }, + { + "epoch": 2.5351467628894593, + "grad_norm": 1.4965180158615112, + "learning_rate": 6.247128009672791e-05, + "loss": 0.4508, + "step": 36534 + }, + { + "epoch": 2.535216154326556, + "grad_norm": 1.532796025276184, + "learning_rate": 6.24660722130734e-05, + "loss": 0.4671, + "step": 36535 + }, + { + "epoch": 2.535285545763653, + "grad_norm": 1.4798085689544678, + "learning_rate": 6.24608644479092e-05, + "loss": 0.3084, + "step": 36536 + }, + { + "epoch": 2.5353549372007493, + "grad_norm": 1.5172252655029297, + "learning_rate": 6.245565680125168e-05, + "loss": 0.3922, + "step": 36537 + }, + { + "epoch": 2.535424328637846, + "grad_norm": 1.2209829092025757, + "learning_rate": 6.245044927311727e-05, + "loss": 0.3967, + "step": 36538 + }, + { + "epoch": 2.5354937200749426, + "grad_norm": 1.2966907024383545, + "learning_rate": 6.244524186352247e-05, + "loss": 0.5307, + "step": 36539 + }, + { + "epoch": 2.5355631115120394, + "grad_norm": 1.0018035173416138, + "learning_rate": 6.244003457248366e-05, + "loss": 0.3382, + "step": 36540 + }, + { + "epoch": 2.5356325029491362, + "grad_norm": 1.0495554208755493, + "learning_rate": 6.243482740001732e-05, + "loss": 0.2481, + "step": 36541 + }, + { + "epoch": 2.5357018943862326, + "grad_norm": 0.9986132979393005, + "learning_rate": 6.242962034613984e-05, + "loss": 0.5118, + "step": 36542 + }, + { + "epoch": 2.5357712858233294, + "grad_norm": 1.1471480131149292, + "learning_rate": 6.24244134108677e-05, + "loss": 0.4291, + "step": 36543 + }, + { + "epoch": 2.535840677260426, + "grad_norm": 1.2394487857818604, + "learning_rate": 6.241920659421733e-05, + "loss": 0.3241, + "step": 36544 + }, + { + "epoch": 2.5359100686975227, + "grad_norm": 0.8371142745018005, + "learning_rate": 6.24139998962052e-05, + "loss": 0.2139, + "step": 36545 + }, + { + "epoch": 2.5359794601346195, + "grad_norm": 1.0564699172973633, + "learning_rate": 6.240879331684768e-05, + "loss": 0.3372, + "step": 36546 + }, + { + "epoch": 2.536048851571716, + "grad_norm": 1.2679471969604492, + "learning_rate": 6.240358685616123e-05, + "loss": 0.4505, + "step": 36547 + }, + { + "epoch": 2.5361182430088127, + "grad_norm": 1.9382866621017456, + "learning_rate": 6.239838051416229e-05, + "loss": 0.5507, + "step": 36548 + }, + { + "epoch": 2.536187634445909, + "grad_norm": 1.2320590019226074, + "learning_rate": 6.239317429086731e-05, + "loss": 0.5503, + "step": 36549 + }, + { + "epoch": 2.536257025883006, + "grad_norm": 0.7855899333953857, + "learning_rate": 6.23879681862927e-05, + "loss": 0.2145, + "step": 36550 + }, + { + "epoch": 2.5363264173201028, + "grad_norm": 1.0798859596252441, + "learning_rate": 6.238276220045496e-05, + "loss": 0.3104, + "step": 36551 + }, + { + "epoch": 2.5363958087571996, + "grad_norm": 1.5896475315093994, + "learning_rate": 6.237755633337046e-05, + "loss": 0.3972, + "step": 36552 + }, + { + "epoch": 2.536465200194296, + "grad_norm": 1.0570157766342163, + "learning_rate": 6.237235058505563e-05, + "loss": 0.243, + "step": 36553 + }, + { + "epoch": 2.536534591631393, + "grad_norm": 0.8658997416496277, + "learning_rate": 6.236714495552692e-05, + "loss": 0.1943, + "step": 36554 + }, + { + "epoch": 2.536603983068489, + "grad_norm": 1.135378360748291, + "learning_rate": 6.236193944480078e-05, + "loss": 0.3284, + "step": 36555 + }, + { + "epoch": 2.536673374505586, + "grad_norm": 1.2449383735656738, + "learning_rate": 6.235673405289363e-05, + "loss": 0.3444, + "step": 36556 + }, + { + "epoch": 2.536742765942683, + "grad_norm": 1.4069017171859741, + "learning_rate": 6.235152877982189e-05, + "loss": 0.3054, + "step": 36557 + }, + { + "epoch": 2.5368121573797793, + "grad_norm": 1.0176374912261963, + "learning_rate": 6.234632362560201e-05, + "loss": 0.318, + "step": 36558 + }, + { + "epoch": 2.536881548816876, + "grad_norm": 1.2333202362060547, + "learning_rate": 6.234111859025046e-05, + "loss": 0.3573, + "step": 36559 + }, + { + "epoch": 2.5369509402539725, + "grad_norm": 0.8570724129676819, + "learning_rate": 6.233591367378357e-05, + "loss": 0.4574, + "step": 36560 + }, + { + "epoch": 2.5370203316910693, + "grad_norm": 1.2878224849700928, + "learning_rate": 6.233070887621785e-05, + "loss": 0.343, + "step": 36561 + }, + { + "epoch": 2.537089723128166, + "grad_norm": 0.9022243618965149, + "learning_rate": 6.23255041975697e-05, + "loss": 0.352, + "step": 36562 + }, + { + "epoch": 2.5371591145652626, + "grad_norm": 1.467952013015747, + "learning_rate": 6.232029963785558e-05, + "loss": 0.4545, + "step": 36563 + }, + { + "epoch": 2.5372285060023594, + "grad_norm": 1.0508337020874023, + "learning_rate": 6.231509519709188e-05, + "loss": 0.3595, + "step": 36564 + }, + { + "epoch": 2.5372978974394558, + "grad_norm": 1.4368908405303955, + "learning_rate": 6.230989087529509e-05, + "loss": 0.4121, + "step": 36565 + }, + { + "epoch": 2.5373672888765526, + "grad_norm": 0.953752338886261, + "learning_rate": 6.23046866724816e-05, + "loss": 0.2878, + "step": 36566 + }, + { + "epoch": 2.5374366803136494, + "grad_norm": 1.040631651878357, + "learning_rate": 6.229948258866782e-05, + "loss": 0.3617, + "step": 36567 + }, + { + "epoch": 2.537506071750746, + "grad_norm": 1.6017988920211792, + "learning_rate": 6.22942786238702e-05, + "loss": 0.402, + "step": 36568 + }, + { + "epoch": 2.5375754631878427, + "grad_norm": 1.0987598896026611, + "learning_rate": 6.228907477810516e-05, + "loss": 0.3086, + "step": 36569 + }, + { + "epoch": 2.537644854624939, + "grad_norm": 1.240529179573059, + "learning_rate": 6.228387105138917e-05, + "loss": 0.408, + "step": 36570 + }, + { + "epoch": 2.537714246062036, + "grad_norm": 1.0426489114761353, + "learning_rate": 6.227866744373859e-05, + "loss": 0.423, + "step": 36571 + }, + { + "epoch": 2.5377836374991327, + "grad_norm": 1.1779149770736694, + "learning_rate": 6.227346395516989e-05, + "loss": 0.3945, + "step": 36572 + }, + { + "epoch": 2.537853028936229, + "grad_norm": 1.1002087593078613, + "learning_rate": 6.226826058569948e-05, + "loss": 0.3178, + "step": 36573 + }, + { + "epoch": 2.537922420373326, + "grad_norm": 1.763778567314148, + "learning_rate": 6.226305733534383e-05, + "loss": 0.52, + "step": 36574 + }, + { + "epoch": 2.5379918118104223, + "grad_norm": 1.4121276140213013, + "learning_rate": 6.225785420411929e-05, + "loss": 0.5627, + "step": 36575 + }, + { + "epoch": 2.538061203247519, + "grad_norm": 1.3424681425094604, + "learning_rate": 6.225265119204233e-05, + "loss": 0.4087, + "step": 36576 + }, + { + "epoch": 2.538130594684616, + "grad_norm": 1.1873819828033447, + "learning_rate": 6.224744829912936e-05, + "loss": 0.3973, + "step": 36577 + }, + { + "epoch": 2.538199986121713, + "grad_norm": 1.3343760967254639, + "learning_rate": 6.224224552539684e-05, + "loss": 0.5805, + "step": 36578 + }, + { + "epoch": 2.538269377558809, + "grad_norm": 0.8010313510894775, + "learning_rate": 6.223704287086114e-05, + "loss": 0.2915, + "step": 36579 + }, + { + "epoch": 2.538338768995906, + "grad_norm": 1.5853006839752197, + "learning_rate": 6.223184033553877e-05, + "loss": 0.4062, + "step": 36580 + }, + { + "epoch": 2.5384081604330024, + "grad_norm": 1.2885479927062988, + "learning_rate": 6.222663791944608e-05, + "loss": 0.5871, + "step": 36581 + }, + { + "epoch": 2.5384775518700993, + "grad_norm": 1.2339537143707275, + "learning_rate": 6.222143562259947e-05, + "loss": 0.5529, + "step": 36582 + }, + { + "epoch": 2.538546943307196, + "grad_norm": 1.4928244352340698, + "learning_rate": 6.221623344501542e-05, + "loss": 0.5189, + "step": 36583 + }, + { + "epoch": 2.5386163347442925, + "grad_norm": 0.9333059191703796, + "learning_rate": 6.221103138671035e-05, + "loss": 0.3375, + "step": 36584 + }, + { + "epoch": 2.5386857261813893, + "grad_norm": 1.568095326423645, + "learning_rate": 6.220582944770066e-05, + "loss": 0.3775, + "step": 36585 + }, + { + "epoch": 2.5387551176184857, + "grad_norm": 1.5141923427581787, + "learning_rate": 6.22006276280028e-05, + "loss": 0.3333, + "step": 36586 + }, + { + "epoch": 2.5388245090555825, + "grad_norm": 1.3394056558609009, + "learning_rate": 6.219542592763318e-05, + "loss": 0.3806, + "step": 36587 + }, + { + "epoch": 2.5388939004926794, + "grad_norm": 1.0728182792663574, + "learning_rate": 6.21902243466082e-05, + "loss": 0.3784, + "step": 36588 + }, + { + "epoch": 2.5389632919297758, + "grad_norm": 1.39841890335083, + "learning_rate": 6.218502288494428e-05, + "loss": 0.4255, + "step": 36589 + }, + { + "epoch": 2.5390326833668726, + "grad_norm": 1.1142040491104126, + "learning_rate": 6.217982154265787e-05, + "loss": 0.4245, + "step": 36590 + }, + { + "epoch": 2.539102074803969, + "grad_norm": 1.1144031286239624, + "learning_rate": 6.217462031976538e-05, + "loss": 0.4303, + "step": 36591 + }, + { + "epoch": 2.539171466241066, + "grad_norm": 1.3953657150268555, + "learning_rate": 6.216941921628321e-05, + "loss": 0.3779, + "step": 36592 + }, + { + "epoch": 2.5392408576781627, + "grad_norm": 1.7214657068252563, + "learning_rate": 6.216421823222778e-05, + "loss": 0.3525, + "step": 36593 + }, + { + "epoch": 2.539310249115259, + "grad_norm": 1.1770933866500854, + "learning_rate": 6.215901736761558e-05, + "loss": 0.4096, + "step": 36594 + }, + { + "epoch": 2.539379640552356, + "grad_norm": 1.4307441711425781, + "learning_rate": 6.215381662246293e-05, + "loss": 0.4752, + "step": 36595 + }, + { + "epoch": 2.5394490319894523, + "grad_norm": 0.9805658459663391, + "learning_rate": 6.21486159967863e-05, + "loss": 0.3494, + "step": 36596 + }, + { + "epoch": 2.539518423426549, + "grad_norm": 1.111525297164917, + "learning_rate": 6.214341549060208e-05, + "loss": 0.2368, + "step": 36597 + }, + { + "epoch": 2.539587814863646, + "grad_norm": 1.1396522521972656, + "learning_rate": 6.213821510392672e-05, + "loss": 0.3528, + "step": 36598 + }, + { + "epoch": 2.5396572063007428, + "grad_norm": 1.3594417572021484, + "learning_rate": 6.21330148367766e-05, + "loss": 0.4943, + "step": 36599 + }, + { + "epoch": 2.539726597737839, + "grad_norm": 1.1813380718231201, + "learning_rate": 6.212781468916819e-05, + "loss": 0.4516, + "step": 36600 + }, + { + "epoch": 2.5397959891749355, + "grad_norm": 0.87660813331604, + "learning_rate": 6.21226146611179e-05, + "loss": 0.2487, + "step": 36601 + }, + { + "epoch": 2.5398653806120324, + "grad_norm": 1.262047290802002, + "learning_rate": 6.211741475264208e-05, + "loss": 0.3925, + "step": 36602 + }, + { + "epoch": 2.539934772049129, + "grad_norm": 0.9786025285720825, + "learning_rate": 6.211221496375718e-05, + "loss": 0.3332, + "step": 36603 + }, + { + "epoch": 2.540004163486226, + "grad_norm": 1.177451491355896, + "learning_rate": 6.210701529447962e-05, + "loss": 0.2931, + "step": 36604 + }, + { + "epoch": 2.5400735549233224, + "grad_norm": 1.1235734224319458, + "learning_rate": 6.210181574482585e-05, + "loss": 0.3986, + "step": 36605 + }, + { + "epoch": 2.5401429463604193, + "grad_norm": 1.5118392705917358, + "learning_rate": 6.209661631481222e-05, + "loss": 0.4215, + "step": 36606 + }, + { + "epoch": 2.5402123377975157, + "grad_norm": 0.8616457581520081, + "learning_rate": 6.209141700445519e-05, + "loss": 0.3016, + "step": 36607 + }, + { + "epoch": 2.5402817292346125, + "grad_norm": 1.454620122909546, + "learning_rate": 6.208621781377116e-05, + "loss": 0.3939, + "step": 36608 + }, + { + "epoch": 2.5403511206717093, + "grad_norm": 1.1076046228408813, + "learning_rate": 6.208101874277655e-05, + "loss": 0.2689, + "step": 36609 + }, + { + "epoch": 2.5404205121088057, + "grad_norm": 1.1163432598114014, + "learning_rate": 6.207581979148775e-05, + "loss": 0.3592, + "step": 36610 + }, + { + "epoch": 2.5404899035459025, + "grad_norm": 1.3534332513809204, + "learning_rate": 6.207062095992117e-05, + "loss": 0.4656, + "step": 36611 + }, + { + "epoch": 2.540559294982999, + "grad_norm": 1.0217883586883545, + "learning_rate": 6.206542224809326e-05, + "loss": 0.2313, + "step": 36612 + }, + { + "epoch": 2.5406286864200958, + "grad_norm": 1.047655701637268, + "learning_rate": 6.20602236560204e-05, + "loss": 0.3852, + "step": 36613 + }, + { + "epoch": 2.5406980778571926, + "grad_norm": 0.9486278891563416, + "learning_rate": 6.205502518371902e-05, + "loss": 0.2647, + "step": 36614 + }, + { + "epoch": 2.540767469294289, + "grad_norm": 1.28017258644104, + "learning_rate": 6.204982683120556e-05, + "loss": 0.4509, + "step": 36615 + }, + { + "epoch": 2.540836860731386, + "grad_norm": 0.9863921403884888, + "learning_rate": 6.204462859849634e-05, + "loss": 0.2696, + "step": 36616 + }, + { + "epoch": 2.540906252168482, + "grad_norm": 1.4205740690231323, + "learning_rate": 6.203943048560783e-05, + "loss": 0.4847, + "step": 36617 + }, + { + "epoch": 2.540975643605579, + "grad_norm": 1.5721735954284668, + "learning_rate": 6.203423249255644e-05, + "loss": 0.467, + "step": 36618 + }, + { + "epoch": 2.541045035042676, + "grad_norm": 1.4798932075500488, + "learning_rate": 6.202903461935857e-05, + "loss": 0.4549, + "step": 36619 + }, + { + "epoch": 2.5411144264797723, + "grad_norm": 1.5262948274612427, + "learning_rate": 6.202383686603063e-05, + "loss": 0.4264, + "step": 36620 + }, + { + "epoch": 2.541183817916869, + "grad_norm": 1.4600313901901245, + "learning_rate": 6.201863923258908e-05, + "loss": 0.3492, + "step": 36621 + }, + { + "epoch": 2.5412532093539655, + "grad_norm": 1.7197777032852173, + "learning_rate": 6.201344171905023e-05, + "loss": 0.4822, + "step": 36622 + }, + { + "epoch": 2.5413226007910623, + "grad_norm": 1.1333328485488892, + "learning_rate": 6.200824432543053e-05, + "loss": 0.2768, + "step": 36623 + }, + { + "epoch": 2.541391992228159, + "grad_norm": 0.9290351867675781, + "learning_rate": 6.200304705174643e-05, + "loss": 0.2832, + "step": 36624 + }, + { + "epoch": 2.541461383665256, + "grad_norm": 1.1372976303100586, + "learning_rate": 6.199784989801427e-05, + "loss": 0.4328, + "step": 36625 + }, + { + "epoch": 2.5415307751023524, + "grad_norm": 1.2025216817855835, + "learning_rate": 6.19926528642505e-05, + "loss": 0.3298, + "step": 36626 + }, + { + "epoch": 2.5416001665394488, + "grad_norm": 1.394241452217102, + "learning_rate": 6.198745595047151e-05, + "loss": 0.5555, + "step": 36627 + }, + { + "epoch": 2.5416695579765456, + "grad_norm": 0.8475452661514282, + "learning_rate": 6.198225915669376e-05, + "loss": 0.1977, + "step": 36628 + }, + { + "epoch": 2.5417389494136424, + "grad_norm": 1.1751664876937866, + "learning_rate": 6.197706248293356e-05, + "loss": 0.3225, + "step": 36629 + }, + { + "epoch": 2.5418083408507393, + "grad_norm": 1.204533338546753, + "learning_rate": 6.197186592920736e-05, + "loss": 0.4518, + "step": 36630 + }, + { + "epoch": 2.5418777322878356, + "grad_norm": 0.993646502494812, + "learning_rate": 6.196666949553156e-05, + "loss": 0.3098, + "step": 36631 + }, + { + "epoch": 2.5419471237249325, + "grad_norm": 1.2411366701126099, + "learning_rate": 6.196147318192258e-05, + "loss": 0.3491, + "step": 36632 + }, + { + "epoch": 2.542016515162029, + "grad_norm": 1.344581127166748, + "learning_rate": 6.19562769883968e-05, + "loss": 0.4651, + "step": 36633 + }, + { + "epoch": 2.5420859065991257, + "grad_norm": 1.5237873792648315, + "learning_rate": 6.195108091497068e-05, + "loss": 0.4623, + "step": 36634 + }, + { + "epoch": 2.5421552980362225, + "grad_norm": 1.3112610578536987, + "learning_rate": 6.19458849616606e-05, + "loss": 0.3871, + "step": 36635 + }, + { + "epoch": 2.542224689473319, + "grad_norm": 1.2404818534851074, + "learning_rate": 6.19406891284829e-05, + "loss": 0.4179, + "step": 36636 + }, + { + "epoch": 2.5422940809104158, + "grad_norm": 2.371267795562744, + "learning_rate": 6.193549341545404e-05, + "loss": 0.3763, + "step": 36637 + }, + { + "epoch": 2.542363472347512, + "grad_norm": 1.8227475881576538, + "learning_rate": 6.193029782259039e-05, + "loss": 0.4847, + "step": 36638 + }, + { + "epoch": 2.542432863784609, + "grad_norm": 0.9290395975112915, + "learning_rate": 6.192510234990839e-05, + "loss": 0.2818, + "step": 36639 + }, + { + "epoch": 2.542502255221706, + "grad_norm": 1.9714967012405396, + "learning_rate": 6.191990699742445e-05, + "loss": 0.4234, + "step": 36640 + }, + { + "epoch": 2.542571646658802, + "grad_norm": 1.0628148317337036, + "learning_rate": 6.19147117651549e-05, + "loss": 0.3176, + "step": 36641 + }, + { + "epoch": 2.542641038095899, + "grad_norm": 3.3703761100769043, + "learning_rate": 6.190951665311624e-05, + "loss": 0.3942, + "step": 36642 + }, + { + "epoch": 2.5427104295329954, + "grad_norm": 1.2079987525939941, + "learning_rate": 6.190432166132478e-05, + "loss": 0.3492, + "step": 36643 + }, + { + "epoch": 2.5427798209700923, + "grad_norm": 1.0283339023590088, + "learning_rate": 6.189912678979696e-05, + "loss": 0.3888, + "step": 36644 + }, + { + "epoch": 2.542849212407189, + "grad_norm": 1.5380709171295166, + "learning_rate": 6.189393203854919e-05, + "loss": 0.4095, + "step": 36645 + }, + { + "epoch": 2.5429186038442855, + "grad_norm": 1.29621160030365, + "learning_rate": 6.188873740759783e-05, + "loss": 0.5117, + "step": 36646 + }, + { + "epoch": 2.5429879952813823, + "grad_norm": 1.0604338645935059, + "learning_rate": 6.188354289695933e-05, + "loss": 0.3549, + "step": 36647 + }, + { + "epoch": 2.5430573867184787, + "grad_norm": 1.2433900833129883, + "learning_rate": 6.187834850665003e-05, + "loss": 0.3128, + "step": 36648 + }, + { + "epoch": 2.5431267781555755, + "grad_norm": 1.4011523723602295, + "learning_rate": 6.187315423668642e-05, + "loss": 0.3753, + "step": 36649 + }, + { + "epoch": 2.5431961695926724, + "grad_norm": 0.9828513264656067, + "learning_rate": 6.18679600870848e-05, + "loss": 0.307, + "step": 36650 + }, + { + "epoch": 2.543265561029769, + "grad_norm": 1.2966445684432983, + "learning_rate": 6.18627660578616e-05, + "loss": 0.4192, + "step": 36651 + }, + { + "epoch": 2.5433349524668656, + "grad_norm": 1.304234504699707, + "learning_rate": 6.185757214903322e-05, + "loss": 0.4496, + "step": 36652 + }, + { + "epoch": 2.5434043439039624, + "grad_norm": 1.190361738204956, + "learning_rate": 6.185237836061606e-05, + "loss": 0.3894, + "step": 36653 + }, + { + "epoch": 2.543473735341059, + "grad_norm": 1.106459617614746, + "learning_rate": 6.184718469262651e-05, + "loss": 0.2965, + "step": 36654 + }, + { + "epoch": 2.5435431267781556, + "grad_norm": 1.318056344985962, + "learning_rate": 6.184199114508099e-05, + "loss": 0.2904, + "step": 36655 + }, + { + "epoch": 2.5436125182152525, + "grad_norm": 1.3773497343063354, + "learning_rate": 6.18367977179959e-05, + "loss": 0.3413, + "step": 36656 + }, + { + "epoch": 2.543681909652349, + "grad_norm": 1.7969599962234497, + "learning_rate": 6.183160441138758e-05, + "loss": 0.4093, + "step": 36657 + }, + { + "epoch": 2.5437513010894457, + "grad_norm": 1.08328378200531, + "learning_rate": 6.182641122527246e-05, + "loss": 0.4678, + "step": 36658 + }, + { + "epoch": 2.543820692526542, + "grad_norm": 1.5193915367126465, + "learning_rate": 6.182121815966694e-05, + "loss": 0.3367, + "step": 36659 + }, + { + "epoch": 2.543890083963639, + "grad_norm": 1.5448256731033325, + "learning_rate": 6.18160252145874e-05, + "loss": 0.4174, + "step": 36660 + }, + { + "epoch": 2.5439594754007357, + "grad_norm": 0.9298325777053833, + "learning_rate": 6.181083239005022e-05, + "loss": 0.3218, + "step": 36661 + }, + { + "epoch": 2.544028866837832, + "grad_norm": 1.0555851459503174, + "learning_rate": 6.180563968607182e-05, + "loss": 0.3688, + "step": 36662 + }, + { + "epoch": 2.544098258274929, + "grad_norm": 1.2361830472946167, + "learning_rate": 6.180044710266863e-05, + "loss": 0.36, + "step": 36663 + }, + { + "epoch": 2.5441676497120254, + "grad_norm": 1.0937201976776123, + "learning_rate": 6.179525463985694e-05, + "loss": 0.4449, + "step": 36664 + }, + { + "epoch": 2.544237041149122, + "grad_norm": 1.2148064374923706, + "learning_rate": 6.179006229765322e-05, + "loss": 0.4415, + "step": 36665 + }, + { + "epoch": 2.544306432586219, + "grad_norm": 1.090685248374939, + "learning_rate": 6.178487007607383e-05, + "loss": 0.316, + "step": 36666 + }, + { + "epoch": 2.5443758240233154, + "grad_norm": 1.5139515399932861, + "learning_rate": 6.177967797513517e-05, + "loss": 0.5436, + "step": 36667 + }, + { + "epoch": 2.5444452154604122, + "grad_norm": 1.2755378484725952, + "learning_rate": 6.177448599485362e-05, + "loss": 0.5592, + "step": 36668 + }, + { + "epoch": 2.5445146068975086, + "grad_norm": 1.3570448160171509, + "learning_rate": 6.17692941352456e-05, + "loss": 0.4221, + "step": 36669 + }, + { + "epoch": 2.5445839983346055, + "grad_norm": 1.6069835424423218, + "learning_rate": 6.176410239632753e-05, + "loss": 0.3999, + "step": 36670 + }, + { + "epoch": 2.5446533897717023, + "grad_norm": 1.3897749185562134, + "learning_rate": 6.17589107781157e-05, + "loss": 0.3777, + "step": 36671 + }, + { + "epoch": 2.5447227812087987, + "grad_norm": 1.2082393169403076, + "learning_rate": 6.175371928062655e-05, + "loss": 0.4051, + "step": 36672 + }, + { + "epoch": 2.5447921726458955, + "grad_norm": 1.2324254512786865, + "learning_rate": 6.174852790387648e-05, + "loss": 0.3961, + "step": 36673 + }, + { + "epoch": 2.544861564082992, + "grad_norm": 1.2035843133926392, + "learning_rate": 6.174333664788186e-05, + "loss": 0.3902, + "step": 36674 + }, + { + "epoch": 2.5449309555200887, + "grad_norm": 1.363236665725708, + "learning_rate": 6.17381455126591e-05, + "loss": 0.4759, + "step": 36675 + }, + { + "epoch": 2.5450003469571856, + "grad_norm": 1.4476569890975952, + "learning_rate": 6.173295449822457e-05, + "loss": 0.4382, + "step": 36676 + }, + { + "epoch": 2.5450697383942824, + "grad_norm": 1.0405994653701782, + "learning_rate": 6.172776360459469e-05, + "loss": 0.3634, + "step": 36677 + }, + { + "epoch": 2.545139129831379, + "grad_norm": 1.2475800514221191, + "learning_rate": 6.172257283178578e-05, + "loss": 0.3952, + "step": 36678 + }, + { + "epoch": 2.5452085212684756, + "grad_norm": 1.1692514419555664, + "learning_rate": 6.17173821798143e-05, + "loss": 0.4555, + "step": 36679 + }, + { + "epoch": 2.545277912705572, + "grad_norm": 1.1394016742706299, + "learning_rate": 6.171219164869657e-05, + "loss": 0.3408, + "step": 36680 + }, + { + "epoch": 2.545347304142669, + "grad_norm": 1.2329280376434326, + "learning_rate": 6.170700123844902e-05, + "loss": 0.559, + "step": 36681 + }, + { + "epoch": 2.5454166955797657, + "grad_norm": 1.735201358795166, + "learning_rate": 6.170181094908802e-05, + "loss": 0.5946, + "step": 36682 + }, + { + "epoch": 2.545486087016862, + "grad_norm": 1.9238533973693848, + "learning_rate": 6.169662078062997e-05, + "loss": 0.4216, + "step": 36683 + }, + { + "epoch": 2.545555478453959, + "grad_norm": 1.2518988847732544, + "learning_rate": 6.169143073309127e-05, + "loss": 0.3383, + "step": 36684 + }, + { + "epoch": 2.5456248698910553, + "grad_norm": 1.88031005859375, + "learning_rate": 6.168624080648824e-05, + "loss": 0.4064, + "step": 36685 + }, + { + "epoch": 2.545694261328152, + "grad_norm": 1.3673464059829712, + "learning_rate": 6.168105100083732e-05, + "loss": 0.3798, + "step": 36686 + }, + { + "epoch": 2.545763652765249, + "grad_norm": 1.7044410705566406, + "learning_rate": 6.167586131615485e-05, + "loss": 0.3718, + "step": 36687 + }, + { + "epoch": 2.5458330442023454, + "grad_norm": 0.9187467694282532, + "learning_rate": 6.167067175245726e-05, + "loss": 0.4093, + "step": 36688 + }, + { + "epoch": 2.545902435639442, + "grad_norm": 1.224491834640503, + "learning_rate": 6.16654823097609e-05, + "loss": 0.352, + "step": 36689 + }, + { + "epoch": 2.5459718270765386, + "grad_norm": 0.8815099000930786, + "learning_rate": 6.166029298808216e-05, + "loss": 0.3281, + "step": 36690 + }, + { + "epoch": 2.5460412185136354, + "grad_norm": 1.2835688591003418, + "learning_rate": 6.165510378743748e-05, + "loss": 0.4221, + "step": 36691 + }, + { + "epoch": 2.5461106099507322, + "grad_norm": 1.3254621028900146, + "learning_rate": 6.164991470784315e-05, + "loss": 0.5249, + "step": 36692 + }, + { + "epoch": 2.5461800013878286, + "grad_norm": 1.0404940843582153, + "learning_rate": 6.164472574931559e-05, + "loss": 0.3829, + "step": 36693 + }, + { + "epoch": 2.5462493928249255, + "grad_norm": 0.9481539130210876, + "learning_rate": 6.163953691187119e-05, + "loss": 0.2892, + "step": 36694 + }, + { + "epoch": 2.546318784262022, + "grad_norm": 1.334196925163269, + "learning_rate": 6.16343481955263e-05, + "loss": 0.6149, + "step": 36695 + }, + { + "epoch": 2.5463881756991187, + "grad_norm": 0.9743919372558594, + "learning_rate": 6.162915960029734e-05, + "loss": 0.3217, + "step": 36696 + }, + { + "epoch": 2.5464575671362155, + "grad_norm": 1.1469630002975464, + "learning_rate": 6.162397112620066e-05, + "loss": 0.388, + "step": 36697 + }, + { + "epoch": 2.546526958573312, + "grad_norm": 1.1109278202056885, + "learning_rate": 6.161878277325268e-05, + "loss": 0.3325, + "step": 36698 + }, + { + "epoch": 2.5465963500104087, + "grad_norm": 1.4501920938491821, + "learning_rate": 6.161359454146972e-05, + "loss": 0.5426, + "step": 36699 + }, + { + "epoch": 2.546665741447505, + "grad_norm": 1.7320531606674194, + "learning_rate": 6.160840643086818e-05, + "loss": 0.5829, + "step": 36700 + }, + { + "epoch": 2.546735132884602, + "grad_norm": 1.2437435388565063, + "learning_rate": 6.160321844146446e-05, + "loss": 0.4541, + "step": 36701 + }, + { + "epoch": 2.546804524321699, + "grad_norm": 1.1888022422790527, + "learning_rate": 6.159803057327491e-05, + "loss": 0.4182, + "step": 36702 + }, + { + "epoch": 2.5468739157587956, + "grad_norm": 1.7123055458068848, + "learning_rate": 6.159284282631594e-05, + "loss": 0.4157, + "step": 36703 + }, + { + "epoch": 2.546943307195892, + "grad_norm": 2.2495017051696777, + "learning_rate": 6.158765520060394e-05, + "loss": 0.5898, + "step": 36704 + }, + { + "epoch": 2.547012698632989, + "grad_norm": 1.1723310947418213, + "learning_rate": 6.158246769615521e-05, + "loss": 0.2306, + "step": 36705 + }, + { + "epoch": 2.5470820900700852, + "grad_norm": 1.3747408390045166, + "learning_rate": 6.157728031298618e-05, + "loss": 0.4343, + "step": 36706 + }, + { + "epoch": 2.547151481507182, + "grad_norm": 1.5130923986434937, + "learning_rate": 6.157209305111323e-05, + "loss": 0.4434, + "step": 36707 + }, + { + "epoch": 2.547220872944279, + "grad_norm": 1.6501487493515015, + "learning_rate": 6.15669059105527e-05, + "loss": 0.5132, + "step": 36708 + }, + { + "epoch": 2.5472902643813753, + "grad_norm": 1.2045915126800537, + "learning_rate": 6.156171889132101e-05, + "loss": 0.422, + "step": 36709 + }, + { + "epoch": 2.547359655818472, + "grad_norm": 1.3391796350479126, + "learning_rate": 6.155653199343451e-05, + "loss": 0.4479, + "step": 36710 + }, + { + "epoch": 2.5474290472555685, + "grad_norm": 1.4243500232696533, + "learning_rate": 6.155134521690959e-05, + "loss": 0.4582, + "step": 36711 + }, + { + "epoch": 2.5474984386926653, + "grad_norm": 1.4795873165130615, + "learning_rate": 6.154615856176261e-05, + "loss": 0.5356, + "step": 36712 + }, + { + "epoch": 2.547567830129762, + "grad_norm": 1.3972710371017456, + "learning_rate": 6.154097202800993e-05, + "loss": 0.5042, + "step": 36713 + }, + { + "epoch": 2.5476372215668586, + "grad_norm": 0.884987473487854, + "learning_rate": 6.153578561566795e-05, + "loss": 0.4518, + "step": 36714 + }, + { + "epoch": 2.5477066130039554, + "grad_norm": 1.6477166414260864, + "learning_rate": 6.153059932475302e-05, + "loss": 0.6004, + "step": 36715 + }, + { + "epoch": 2.547776004441052, + "grad_norm": 1.3023285865783691, + "learning_rate": 6.152541315528153e-05, + "loss": 0.3422, + "step": 36716 + }, + { + "epoch": 2.5478453958781486, + "grad_norm": 0.8977706432342529, + "learning_rate": 6.152022710726983e-05, + "loss": 0.2901, + "step": 36717 + }, + { + "epoch": 2.5479147873152455, + "grad_norm": 1.016322374343872, + "learning_rate": 6.151504118073437e-05, + "loss": 0.1917, + "step": 36718 + }, + { + "epoch": 2.547984178752342, + "grad_norm": 1.2823883295059204, + "learning_rate": 6.15098553756914e-05, + "loss": 0.5437, + "step": 36719 + }, + { + "epoch": 2.5480535701894387, + "grad_norm": 0.9210065603256226, + "learning_rate": 6.150466969215737e-05, + "loss": 0.3569, + "step": 36720 + }, + { + "epoch": 2.548122961626535, + "grad_norm": 1.0935288667678833, + "learning_rate": 6.149948413014862e-05, + "loss": 0.2152, + "step": 36721 + }, + { + "epoch": 2.548192353063632, + "grad_norm": 3.6854279041290283, + "learning_rate": 6.149429868968153e-05, + "loss": 0.4695, + "step": 36722 + }, + { + "epoch": 2.5482617445007287, + "grad_norm": 1.254056692123413, + "learning_rate": 6.148911337077248e-05, + "loss": 0.5592, + "step": 36723 + }, + { + "epoch": 2.548331135937825, + "grad_norm": 1.5273240804672241, + "learning_rate": 6.148392817343781e-05, + "loss": 0.4379, + "step": 36724 + }, + { + "epoch": 2.548400527374922, + "grad_norm": 0.9886549115180969, + "learning_rate": 6.1478743097694e-05, + "loss": 0.2871, + "step": 36725 + }, + { + "epoch": 2.5484699188120183, + "grad_norm": 1.324951171875, + "learning_rate": 6.147355814355724e-05, + "loss": 0.4311, + "step": 36726 + }, + { + "epoch": 2.548539310249115, + "grad_norm": 1.2998789548873901, + "learning_rate": 6.1468373311044e-05, + "loss": 0.4482, + "step": 36727 + }, + { + "epoch": 2.548608701686212, + "grad_norm": 1.0697516202926636, + "learning_rate": 6.146318860017064e-05, + "loss": 0.3355, + "step": 36728 + }, + { + "epoch": 2.548678093123309, + "grad_norm": 1.751441240310669, + "learning_rate": 6.145800401095354e-05, + "loss": 0.3814, + "step": 36729 + }, + { + "epoch": 2.5487474845604052, + "grad_norm": 1.1455596685409546, + "learning_rate": 6.145281954340904e-05, + "loss": 0.3651, + "step": 36730 + }, + { + "epoch": 2.548816875997502, + "grad_norm": 3.0354208946228027, + "learning_rate": 6.14476351975535e-05, + "loss": 0.4183, + "step": 36731 + }, + { + "epoch": 2.5488862674345985, + "grad_norm": 1.4956774711608887, + "learning_rate": 6.144245097340333e-05, + "loss": 0.3311, + "step": 36732 + }, + { + "epoch": 2.5489556588716953, + "grad_norm": 1.586740255355835, + "learning_rate": 6.143726687097486e-05, + "loss": 0.4681, + "step": 36733 + }, + { + "epoch": 2.549025050308792, + "grad_norm": 1.1675578355789185, + "learning_rate": 6.143208289028446e-05, + "loss": 0.3201, + "step": 36734 + }, + { + "epoch": 2.5490944417458885, + "grad_norm": 0.6814881563186646, + "learning_rate": 6.142689903134849e-05, + "loss": 0.208, + "step": 36735 + }, + { + "epoch": 2.5491638331829853, + "grad_norm": 0.8168811798095703, + "learning_rate": 6.142171529418332e-05, + "loss": 0.2681, + "step": 36736 + }, + { + "epoch": 2.5492332246200817, + "grad_norm": 1.2797105312347412, + "learning_rate": 6.141653167880533e-05, + "loss": 0.451, + "step": 36737 + }, + { + "epoch": 2.5493026160571786, + "grad_norm": 1.5867936611175537, + "learning_rate": 6.141134818523088e-05, + "loss": 0.5816, + "step": 36738 + }, + { + "epoch": 2.5493720074942754, + "grad_norm": 2.4753096103668213, + "learning_rate": 6.140616481347636e-05, + "loss": 0.4467, + "step": 36739 + }, + { + "epoch": 2.549441398931372, + "grad_norm": 1.3320587873458862, + "learning_rate": 6.140098156355804e-05, + "loss": 0.3867, + "step": 36740 + }, + { + "epoch": 2.5495107903684686, + "grad_norm": 1.4122625589370728, + "learning_rate": 6.139579843549237e-05, + "loss": 0.4009, + "step": 36741 + }, + { + "epoch": 2.549580181805565, + "grad_norm": 1.4039661884307861, + "learning_rate": 6.139061542929567e-05, + "loss": 0.5346, + "step": 36742 + }, + { + "epoch": 2.549649573242662, + "grad_norm": 1.6182525157928467, + "learning_rate": 6.138543254498432e-05, + "loss": 0.4342, + "step": 36743 + }, + { + "epoch": 2.5497189646797587, + "grad_norm": 0.9154033064842224, + "learning_rate": 6.138024978257468e-05, + "loss": 0.291, + "step": 36744 + }, + { + "epoch": 2.549788356116855, + "grad_norm": 1.6169849634170532, + "learning_rate": 6.137506714208313e-05, + "loss": 0.4595, + "step": 36745 + }, + { + "epoch": 2.549857747553952, + "grad_norm": 1.6062426567077637, + "learning_rate": 6.136988462352602e-05, + "loss": 0.5256, + "step": 36746 + }, + { + "epoch": 2.5499271389910483, + "grad_norm": 1.5960944890975952, + "learning_rate": 6.136470222691967e-05, + "loss": 0.4399, + "step": 36747 + }, + { + "epoch": 2.549996530428145, + "grad_norm": 1.5601047277450562, + "learning_rate": 6.13595199522805e-05, + "loss": 0.4458, + "step": 36748 + }, + { + "epoch": 2.550065921865242, + "grad_norm": 1.4436553716659546, + "learning_rate": 6.135433779962482e-05, + "loss": 0.5714, + "step": 36749 + }, + { + "epoch": 2.5501353133023383, + "grad_norm": 1.2483562231063843, + "learning_rate": 6.1349155768969e-05, + "loss": 0.4935, + "step": 36750 + }, + { + "epoch": 2.550204704739435, + "grad_norm": 1.0003328323364258, + "learning_rate": 6.134397386032943e-05, + "loss": 0.4376, + "step": 36751 + }, + { + "epoch": 2.5502740961765316, + "grad_norm": 1.1666741371154785, + "learning_rate": 6.133879207372244e-05, + "loss": 0.3107, + "step": 36752 + }, + { + "epoch": 2.5503434876136284, + "grad_norm": 1.6982768774032593, + "learning_rate": 6.133361040916445e-05, + "loss": 0.4644, + "step": 36753 + }, + { + "epoch": 2.5504128790507252, + "grad_norm": 1.1624457836151123, + "learning_rate": 6.132842886667171e-05, + "loss": 0.3067, + "step": 36754 + }, + { + "epoch": 2.550482270487822, + "grad_norm": 1.1865229606628418, + "learning_rate": 6.132324744626063e-05, + "loss": 0.4383, + "step": 36755 + }, + { + "epoch": 2.5505516619249184, + "grad_norm": 1.1982275247573853, + "learning_rate": 6.131806614794759e-05, + "loss": 0.3511, + "step": 36756 + }, + { + "epoch": 2.5506210533620153, + "grad_norm": 1.225709080696106, + "learning_rate": 6.131288497174892e-05, + "loss": 0.4481, + "step": 36757 + }, + { + "epoch": 2.5506904447991117, + "grad_norm": 1.5896716117858887, + "learning_rate": 6.130770391768097e-05, + "loss": 0.3901, + "step": 36758 + }, + { + "epoch": 2.5507598362362085, + "grad_norm": 1.087049126625061, + "learning_rate": 6.130252298576013e-05, + "loss": 0.3559, + "step": 36759 + }, + { + "epoch": 2.5508292276733053, + "grad_norm": 1.056968331336975, + "learning_rate": 6.129734217600276e-05, + "loss": 0.4627, + "step": 36760 + }, + { + "epoch": 2.5508986191104017, + "grad_norm": 1.262241244316101, + "learning_rate": 6.129216148842515e-05, + "loss": 0.4599, + "step": 36761 + }, + { + "epoch": 2.5509680105474986, + "grad_norm": 1.6301063299179077, + "learning_rate": 6.12869809230437e-05, + "loss": 0.5396, + "step": 36762 + }, + { + "epoch": 2.551037401984595, + "grad_norm": 1.1494783163070679, + "learning_rate": 6.128180047987478e-05, + "loss": 0.3621, + "step": 36763 + }, + { + "epoch": 2.5511067934216918, + "grad_norm": 1.0203200578689575, + "learning_rate": 6.127662015893472e-05, + "loss": 0.3862, + "step": 36764 + }, + { + "epoch": 2.5511761848587886, + "grad_norm": 2.114022731781006, + "learning_rate": 6.127143996023985e-05, + "loss": 0.5081, + "step": 36765 + }, + { + "epoch": 2.551245576295885, + "grad_norm": 1.0915242433547974, + "learning_rate": 6.126625988380657e-05, + "loss": 0.3634, + "step": 36766 + }, + { + "epoch": 2.551314967732982, + "grad_norm": 1.3458958864212036, + "learning_rate": 6.126107992965124e-05, + "loss": 0.2271, + "step": 36767 + }, + { + "epoch": 2.551384359170078, + "grad_norm": 1.241403341293335, + "learning_rate": 6.125590009779016e-05, + "loss": 0.3496, + "step": 36768 + }, + { + "epoch": 2.551453750607175, + "grad_norm": 1.3470929861068726, + "learning_rate": 6.12507203882397e-05, + "loss": 0.3693, + "step": 36769 + }, + { + "epoch": 2.551523142044272, + "grad_norm": 1.1104892492294312, + "learning_rate": 6.124554080101622e-05, + "loss": 0.3211, + "step": 36770 + }, + { + "epoch": 2.5515925334813683, + "grad_norm": 0.8986684083938599, + "learning_rate": 6.124036133613606e-05, + "loss": 0.3079, + "step": 36771 + }, + { + "epoch": 2.551661924918465, + "grad_norm": 1.2338322401046753, + "learning_rate": 6.12351819936156e-05, + "loss": 0.3529, + "step": 36772 + }, + { + "epoch": 2.5517313163555615, + "grad_norm": 1.3390594720840454, + "learning_rate": 6.123000277347121e-05, + "loss": 0.5093, + "step": 36773 + }, + { + "epoch": 2.5518007077926583, + "grad_norm": 1.2055965662002563, + "learning_rate": 6.122482367571915e-05, + "loss": 0.4777, + "step": 36774 + }, + { + "epoch": 2.551870099229755, + "grad_norm": 1.2666867971420288, + "learning_rate": 6.121964470037585e-05, + "loss": 0.3452, + "step": 36775 + }, + { + "epoch": 2.5519394906668516, + "grad_norm": 1.1277337074279785, + "learning_rate": 6.121446584745761e-05, + "loss": 0.3399, + "step": 36776 + }, + { + "epoch": 2.5520088821039484, + "grad_norm": 0.9751396775245667, + "learning_rate": 6.12092871169808e-05, + "loss": 0.3444, + "step": 36777 + }, + { + "epoch": 2.5520782735410448, + "grad_norm": 1.7505316734313965, + "learning_rate": 6.120410850896179e-05, + "loss": 0.6286, + "step": 36778 + }, + { + "epoch": 2.5521476649781416, + "grad_norm": 1.799242615699768, + "learning_rate": 6.119893002341691e-05, + "loss": 0.3077, + "step": 36779 + }, + { + "epoch": 2.5522170564152384, + "grad_norm": 1.2305930852890015, + "learning_rate": 6.119375166036251e-05, + "loss": 0.3939, + "step": 36780 + }, + { + "epoch": 2.5522864478523353, + "grad_norm": 1.048416256904602, + "learning_rate": 6.118857341981492e-05, + "loss": 0.3539, + "step": 36781 + }, + { + "epoch": 2.5523558392894317, + "grad_norm": 1.1711310148239136, + "learning_rate": 6.118339530179049e-05, + "loss": 0.4088, + "step": 36782 + }, + { + "epoch": 2.5524252307265285, + "grad_norm": 0.890920102596283, + "learning_rate": 6.11782173063056e-05, + "loss": 0.311, + "step": 36783 + }, + { + "epoch": 2.552494622163625, + "grad_norm": 1.0134433507919312, + "learning_rate": 6.117303943337657e-05, + "loss": 0.4147, + "step": 36784 + }, + { + "epoch": 2.5525640136007217, + "grad_norm": 1.3776930570602417, + "learning_rate": 6.116786168301973e-05, + "loss": 0.4728, + "step": 36785 + }, + { + "epoch": 2.5526334050378185, + "grad_norm": 1.1197832822799683, + "learning_rate": 6.116268405525146e-05, + "loss": 0.2865, + "step": 36786 + }, + { + "epoch": 2.552702796474915, + "grad_norm": 1.054080843925476, + "learning_rate": 6.115750655008811e-05, + "loss": 0.4273, + "step": 36787 + }, + { + "epoch": 2.5527721879120118, + "grad_norm": 1.2847199440002441, + "learning_rate": 6.115232916754599e-05, + "loss": 0.3703, + "step": 36788 + }, + { + "epoch": 2.552841579349108, + "grad_norm": 1.8685024976730347, + "learning_rate": 6.114715190764144e-05, + "loss": 0.5248, + "step": 36789 + }, + { + "epoch": 2.552910970786205, + "grad_norm": 1.0978306531906128, + "learning_rate": 6.114197477039083e-05, + "loss": 0.3391, + "step": 36790 + }, + { + "epoch": 2.552980362223302, + "grad_norm": 1.0478852987289429, + "learning_rate": 6.11367977558105e-05, + "loss": 0.3424, + "step": 36791 + }, + { + "epoch": 2.553049753660398, + "grad_norm": 1.4939851760864258, + "learning_rate": 6.113162086391679e-05, + "loss": 0.4184, + "step": 36792 + }, + { + "epoch": 2.553119145097495, + "grad_norm": 1.3880068063735962, + "learning_rate": 6.112644409472604e-05, + "loss": 0.3613, + "step": 36793 + }, + { + "epoch": 2.5531885365345914, + "grad_norm": 1.2769805192947388, + "learning_rate": 6.112126744825464e-05, + "loss": 0.3338, + "step": 36794 + }, + { + "epoch": 2.5532579279716883, + "grad_norm": 1.3604801893234253, + "learning_rate": 6.111609092451884e-05, + "loss": 0.4547, + "step": 36795 + }, + { + "epoch": 2.553327319408785, + "grad_norm": 2.150425910949707, + "learning_rate": 6.111091452353503e-05, + "loss": 0.5575, + "step": 36796 + }, + { + "epoch": 2.5533967108458815, + "grad_norm": 1.3539401292800903, + "learning_rate": 6.110573824531956e-05, + "loss": 0.425, + "step": 36797 + }, + { + "epoch": 2.5534661022829783, + "grad_norm": 1.3512015342712402, + "learning_rate": 6.110056208988877e-05, + "loss": 0.3169, + "step": 36798 + }, + { + "epoch": 2.5535354937200747, + "grad_norm": 1.2808334827423096, + "learning_rate": 6.109538605725899e-05, + "loss": 0.4935, + "step": 36799 + }, + { + "epoch": 2.5536048851571715, + "grad_norm": 1.3094121217727661, + "learning_rate": 6.109021014744656e-05, + "loss": 0.296, + "step": 36800 + }, + { + "epoch": 2.5536742765942684, + "grad_norm": 1.2126184701919556, + "learning_rate": 6.108503436046784e-05, + "loss": 0.3627, + "step": 36801 + }, + { + "epoch": 2.5537436680313648, + "grad_norm": 2.5066890716552734, + "learning_rate": 6.107985869633915e-05, + "loss": 0.4327, + "step": 36802 + }, + { + "epoch": 2.5538130594684616, + "grad_norm": 1.2426338195800781, + "learning_rate": 6.107468315507681e-05, + "loss": 0.3962, + "step": 36803 + }, + { + "epoch": 2.553882450905558, + "grad_norm": 1.082162618637085, + "learning_rate": 6.106950773669718e-05, + "loss": 0.3015, + "step": 36804 + }, + { + "epoch": 2.553951842342655, + "grad_norm": 1.4122599363327026, + "learning_rate": 6.106433244121659e-05, + "loss": 0.6502, + "step": 36805 + }, + { + "epoch": 2.5540212337797517, + "grad_norm": 1.1601639986038208, + "learning_rate": 6.105915726865138e-05, + "loss": 0.3578, + "step": 36806 + }, + { + "epoch": 2.5540906252168485, + "grad_norm": 1.0458039045333862, + "learning_rate": 6.10539822190179e-05, + "loss": 0.3461, + "step": 36807 + }, + { + "epoch": 2.554160016653945, + "grad_norm": 1.2931464910507202, + "learning_rate": 6.104880729233253e-05, + "loss": 0.3751, + "step": 36808 + }, + { + "epoch": 2.5542294080910417, + "grad_norm": 0.8428283333778381, + "learning_rate": 6.104363248861148e-05, + "loss": 0.2474, + "step": 36809 + }, + { + "epoch": 2.554298799528138, + "grad_norm": 1.075575351715088, + "learning_rate": 6.103845780787121e-05, + "loss": 0.3148, + "step": 36810 + }, + { + "epoch": 2.554368190965235, + "grad_norm": 1.2170006036758423, + "learning_rate": 6.103328325012798e-05, + "loss": 0.3562, + "step": 36811 + }, + { + "epoch": 2.5544375824023318, + "grad_norm": 1.0977267026901245, + "learning_rate": 6.1028108815398165e-05, + "loss": 0.3254, + "step": 36812 + }, + { + "epoch": 2.554506973839428, + "grad_norm": 1.3833767175674438, + "learning_rate": 6.102293450369809e-05, + "loss": 0.4523, + "step": 36813 + }, + { + "epoch": 2.554576365276525, + "grad_norm": 1.034453272819519, + "learning_rate": 6.10177603150441e-05, + "loss": 0.3355, + "step": 36814 + }, + { + "epoch": 2.5546457567136214, + "grad_norm": 0.9746540784835815, + "learning_rate": 6.101258624945252e-05, + "loss": 0.2396, + "step": 36815 + }, + { + "epoch": 2.554715148150718, + "grad_norm": 1.09913170337677, + "learning_rate": 6.100741230693967e-05, + "loss": 0.2976, + "step": 36816 + }, + { + "epoch": 2.554784539587815, + "grad_norm": 1.2171063423156738, + "learning_rate": 6.100223848752189e-05, + "loss": 0.5029, + "step": 36817 + }, + { + "epoch": 2.5548539310249114, + "grad_norm": 1.865459680557251, + "learning_rate": 6.0997064791215543e-05, + "loss": 0.3984, + "step": 36818 + }, + { + "epoch": 2.5549233224620083, + "grad_norm": 1.5209636688232422, + "learning_rate": 6.09918912180369e-05, + "loss": 0.5498, + "step": 36819 + }, + { + "epoch": 2.5549927138991047, + "grad_norm": 1.2272026538848877, + "learning_rate": 6.098671776800236e-05, + "loss": 0.3297, + "step": 36820 + }, + { + "epoch": 2.5550621053362015, + "grad_norm": 1.0236213207244873, + "learning_rate": 6.0981544441128205e-05, + "loss": 0.3342, + "step": 36821 + }, + { + "epoch": 2.5551314967732983, + "grad_norm": 1.1081798076629639, + "learning_rate": 6.097637123743083e-05, + "loss": 0.398, + "step": 36822 + }, + { + "epoch": 2.5552008882103947, + "grad_norm": 1.0535495281219482, + "learning_rate": 6.097119815692648e-05, + "loss": 0.4227, + "step": 36823 + }, + { + "epoch": 2.5552702796474915, + "grad_norm": 1.0417598485946655, + "learning_rate": 6.0966025199631524e-05, + "loss": 0.457, + "step": 36824 + }, + { + "epoch": 2.555339671084588, + "grad_norm": 1.225372314453125, + "learning_rate": 6.096085236556231e-05, + "loss": 0.2944, + "step": 36825 + }, + { + "epoch": 2.5554090625216848, + "grad_norm": 1.847082257270813, + "learning_rate": 6.0955679654735156e-05, + "loss": 0.4346, + "step": 36826 + }, + { + "epoch": 2.5554784539587816, + "grad_norm": 1.3656967878341675, + "learning_rate": 6.095050706716638e-05, + "loss": 0.4668, + "step": 36827 + }, + { + "epoch": 2.5555478453958784, + "grad_norm": 1.2777140140533447, + "learning_rate": 6.094533460287234e-05, + "loss": 0.4506, + "step": 36828 + }, + { + "epoch": 2.555617236832975, + "grad_norm": 1.4450148344039917, + "learning_rate": 6.094016226186938e-05, + "loss": 0.5173, + "step": 36829 + }, + { + "epoch": 2.555686628270071, + "grad_norm": 1.5756183862686157, + "learning_rate": 6.0934990044173755e-05, + "loss": 0.2976, + "step": 36830 + }, + { + "epoch": 2.555756019707168, + "grad_norm": 1.1326619386672974, + "learning_rate": 6.092981794980183e-05, + "loss": 0.3547, + "step": 36831 + }, + { + "epoch": 2.555825411144265, + "grad_norm": 1.2060320377349854, + "learning_rate": 6.092464597876995e-05, + "loss": 0.3606, + "step": 36832 + }, + { + "epoch": 2.5558948025813617, + "grad_norm": 0.9748684763908386, + "learning_rate": 6.09194741310944e-05, + "loss": 0.4174, + "step": 36833 + }, + { + "epoch": 2.555964194018458, + "grad_norm": 0.7424429059028625, + "learning_rate": 6.091430240679159e-05, + "loss": 0.2556, + "step": 36834 + }, + { + "epoch": 2.556033585455555, + "grad_norm": 1.10316801071167, + "learning_rate": 6.090913080587774e-05, + "loss": 0.3788, + "step": 36835 + }, + { + "epoch": 2.5561029768926513, + "grad_norm": 1.386116862297058, + "learning_rate": 6.0903959328369276e-05, + "loss": 0.4792, + "step": 36836 + }, + { + "epoch": 2.556172368329748, + "grad_norm": 1.5430854558944702, + "learning_rate": 6.089878797428247e-05, + "loss": 0.5306, + "step": 36837 + }, + { + "epoch": 2.556241759766845, + "grad_norm": 1.1889159679412842, + "learning_rate": 6.089361674363363e-05, + "loss": 0.4619, + "step": 36838 + }, + { + "epoch": 2.5563111512039414, + "grad_norm": 1.5426158905029297, + "learning_rate": 6.08884456364391e-05, + "loss": 0.31, + "step": 36839 + }, + { + "epoch": 2.556380542641038, + "grad_norm": 1.7349411249160767, + "learning_rate": 6.0883274652715216e-05, + "loss": 0.3878, + "step": 36840 + }, + { + "epoch": 2.5564499340781346, + "grad_norm": 1.2173779010772705, + "learning_rate": 6.087810379247829e-05, + "loss": 0.6673, + "step": 36841 + }, + { + "epoch": 2.5565193255152314, + "grad_norm": 1.720418930053711, + "learning_rate": 6.087293305574466e-05, + "loss": 0.3839, + "step": 36842 + }, + { + "epoch": 2.5565887169523283, + "grad_norm": 1.0927457809448242, + "learning_rate": 6.086776244253067e-05, + "loss": 0.4511, + "step": 36843 + }, + { + "epoch": 2.5566581083894246, + "grad_norm": 0.9383202195167542, + "learning_rate": 6.086259195285259e-05, + "loss": 0.3585, + "step": 36844 + }, + { + "epoch": 2.5567274998265215, + "grad_norm": 1.1489911079406738, + "learning_rate": 6.085742158672676e-05, + "loss": 0.3612, + "step": 36845 + }, + { + "epoch": 2.556796891263618, + "grad_norm": 0.979637086391449, + "learning_rate": 6.0852251344169496e-05, + "loss": 0.3104, + "step": 36846 + }, + { + "epoch": 2.5568662827007147, + "grad_norm": 1.4210028648376465, + "learning_rate": 6.0847081225197136e-05, + "loss": 0.4406, + "step": 36847 + }, + { + "epoch": 2.5569356741378115, + "grad_norm": 1.4101923704147339, + "learning_rate": 6.0841911229826006e-05, + "loss": 0.3793, + "step": 36848 + }, + { + "epoch": 2.557005065574908, + "grad_norm": 1.1258084774017334, + "learning_rate": 6.083674135807246e-05, + "loss": 0.5255, + "step": 36849 + }, + { + "epoch": 2.5570744570120048, + "grad_norm": 1.323193907737732, + "learning_rate": 6.0831571609952745e-05, + "loss": 0.3841, + "step": 36850 + }, + { + "epoch": 2.557143848449101, + "grad_norm": 1.7263725996017456, + "learning_rate": 6.082640198548319e-05, + "loss": 0.3114, + "step": 36851 + }, + { + "epoch": 2.557213239886198, + "grad_norm": 1.611772060394287, + "learning_rate": 6.0821232484680154e-05, + "loss": 0.4086, + "step": 36852 + }, + { + "epoch": 2.557282631323295, + "grad_norm": 1.2737524509429932, + "learning_rate": 6.0816063107559965e-05, + "loss": 0.2822, + "step": 36853 + }, + { + "epoch": 2.5573520227603916, + "grad_norm": 1.5819551944732666, + "learning_rate": 6.0810893854138895e-05, + "loss": 0.8353, + "step": 36854 + }, + { + "epoch": 2.557421414197488, + "grad_norm": 1.2023004293441772, + "learning_rate": 6.0805724724433275e-05, + "loss": 0.2925, + "step": 36855 + }, + { + "epoch": 2.5574908056345844, + "grad_norm": 1.5721186399459839, + "learning_rate": 6.0800555718459484e-05, + "loss": 0.2616, + "step": 36856 + }, + { + "epoch": 2.5575601970716813, + "grad_norm": 1.326021432876587, + "learning_rate": 6.079538683623376e-05, + "loss": 0.5157, + "step": 36857 + }, + { + "epoch": 2.557629588508778, + "grad_norm": 1.1134192943572998, + "learning_rate": 6.079021807777243e-05, + "loss": 0.2462, + "step": 36858 + }, + { + "epoch": 2.557698979945875, + "grad_norm": 1.1849610805511475, + "learning_rate": 6.078504944309185e-05, + "loss": 0.6262, + "step": 36859 + }, + { + "epoch": 2.5577683713829713, + "grad_norm": 1.0805572271347046, + "learning_rate": 6.077988093220832e-05, + "loss": 0.3868, + "step": 36860 + }, + { + "epoch": 2.557837762820068, + "grad_norm": 1.3003599643707275, + "learning_rate": 6.0774712545138145e-05, + "loss": 0.439, + "step": 36861 + }, + { + "epoch": 2.5579071542571645, + "grad_norm": 1.3144930601119995, + "learning_rate": 6.076954428189765e-05, + "loss": 0.3352, + "step": 36862 + }, + { + "epoch": 2.5579765456942614, + "grad_norm": 1.147834300994873, + "learning_rate": 6.076437614250321e-05, + "loss": 0.4155, + "step": 36863 + }, + { + "epoch": 2.558045937131358, + "grad_norm": 1.1788426637649536, + "learning_rate": 6.0759208126971026e-05, + "loss": 0.3335, + "step": 36864 + }, + { + "epoch": 2.5581153285684546, + "grad_norm": 1.1002719402313232, + "learning_rate": 6.075404023531748e-05, + "loss": 0.2877, + "step": 36865 + }, + { + "epoch": 2.5581847200055514, + "grad_norm": 1.4084103107452393, + "learning_rate": 6.074887246755886e-05, + "loss": 0.2699, + "step": 36866 + }, + { + "epoch": 2.558254111442648, + "grad_norm": 1.3649907112121582, + "learning_rate": 6.07437048237115e-05, + "loss": 0.5047, + "step": 36867 + }, + { + "epoch": 2.5583235028797446, + "grad_norm": 0.9020290374755859, + "learning_rate": 6.073853730379173e-05, + "loss": 0.3502, + "step": 36868 + }, + { + "epoch": 2.5583928943168415, + "grad_norm": 1.4447767734527588, + "learning_rate": 6.073336990781584e-05, + "loss": 0.6384, + "step": 36869 + }, + { + "epoch": 2.558462285753938, + "grad_norm": 1.0391484498977661, + "learning_rate": 6.0728202635800147e-05, + "loss": 0.3191, + "step": 36870 + }, + { + "epoch": 2.5585316771910347, + "grad_norm": 1.1370911598205566, + "learning_rate": 6.072303548776095e-05, + "loss": 0.3404, + "step": 36871 + }, + { + "epoch": 2.558601068628131, + "grad_norm": 1.730584979057312, + "learning_rate": 6.0717868463714586e-05, + "loss": 0.3861, + "step": 36872 + }, + { + "epoch": 2.558670460065228, + "grad_norm": 0.8675180077552795, + "learning_rate": 6.071270156367734e-05, + "loss": 0.3348, + "step": 36873 + }, + { + "epoch": 2.5587398515023247, + "grad_norm": 1.8398621082305908, + "learning_rate": 6.0707534787665534e-05, + "loss": 0.3522, + "step": 36874 + }, + { + "epoch": 2.558809242939421, + "grad_norm": 0.9962478280067444, + "learning_rate": 6.0702368135695474e-05, + "loss": 0.4012, + "step": 36875 + }, + { + "epoch": 2.558878634376518, + "grad_norm": 2.1522092819213867, + "learning_rate": 6.069720160778349e-05, + "loss": 0.547, + "step": 36876 + }, + { + "epoch": 2.5589480258136144, + "grad_norm": 0.9998903870582581, + "learning_rate": 6.069203520394592e-05, + "loss": 0.397, + "step": 36877 + }, + { + "epoch": 2.559017417250711, + "grad_norm": 2.1249918937683105, + "learning_rate": 6.068686892419899e-05, + "loss": 0.6347, + "step": 36878 + }, + { + "epoch": 2.559086808687808, + "grad_norm": 1.3583711385726929, + "learning_rate": 6.0681702768559056e-05, + "loss": 0.5073, + "step": 36879 + }, + { + "epoch": 2.559156200124905, + "grad_norm": 1.3213701248168945, + "learning_rate": 6.0676536737042425e-05, + "loss": 0.4729, + "step": 36880 + }, + { + "epoch": 2.5592255915620012, + "grad_norm": 1.0435056686401367, + "learning_rate": 6.0671370829665405e-05, + "loss": 0.4268, + "step": 36881 + }, + { + "epoch": 2.5592949829990976, + "grad_norm": 0.9168371558189392, + "learning_rate": 6.06662050464443e-05, + "loss": 0.2892, + "step": 36882 + }, + { + "epoch": 2.5593643744361945, + "grad_norm": 1.1222999095916748, + "learning_rate": 6.0661039387395425e-05, + "loss": 0.3023, + "step": 36883 + }, + { + "epoch": 2.5594337658732913, + "grad_norm": 1.2619242668151855, + "learning_rate": 6.065587385253513e-05, + "loss": 0.4658, + "step": 36884 + }, + { + "epoch": 2.559503157310388, + "grad_norm": 1.4111474752426147, + "learning_rate": 6.065070844187963e-05, + "loss": 0.4747, + "step": 36885 + }, + { + "epoch": 2.5595725487474845, + "grad_norm": 1.1623008251190186, + "learning_rate": 6.064554315544528e-05, + "loss": 0.5597, + "step": 36886 + }, + { + "epoch": 2.5596419401845814, + "grad_norm": 1.3882386684417725, + "learning_rate": 6.064037799324838e-05, + "loss": 0.3007, + "step": 36887 + }, + { + "epoch": 2.5597113316216777, + "grad_norm": 1.0876479148864746, + "learning_rate": 6.0635212955305264e-05, + "loss": 0.3342, + "step": 36888 + }, + { + "epoch": 2.5597807230587746, + "grad_norm": 1.2089498043060303, + "learning_rate": 6.063004804163219e-05, + "loss": 0.38, + "step": 36889 + }, + { + "epoch": 2.5598501144958714, + "grad_norm": 1.687401294708252, + "learning_rate": 6.062488325224549e-05, + "loss": 0.4174, + "step": 36890 + }, + { + "epoch": 2.559919505932968, + "grad_norm": 1.2942800521850586, + "learning_rate": 6.061971858716151e-05, + "loss": 0.3347, + "step": 36891 + }, + { + "epoch": 2.5599888973700646, + "grad_norm": 1.3184162378311157, + "learning_rate": 6.061455404639647e-05, + "loss": 0.3465, + "step": 36892 + }, + { + "epoch": 2.560058288807161, + "grad_norm": 1.2985215187072754, + "learning_rate": 6.060938962996671e-05, + "loss": 0.4316, + "step": 36893 + }, + { + "epoch": 2.560127680244258, + "grad_norm": 1.7092207670211792, + "learning_rate": 6.0604225337888524e-05, + "loss": 0.3445, + "step": 36894 + }, + { + "epoch": 2.5601970716813547, + "grad_norm": 1.1092710494995117, + "learning_rate": 6.059906117017824e-05, + "loss": 0.3113, + "step": 36895 + }, + { + "epoch": 2.560266463118451, + "grad_norm": 1.1786562204360962, + "learning_rate": 6.059389712685215e-05, + "loss": 0.2854, + "step": 36896 + }, + { + "epoch": 2.560335854555548, + "grad_norm": 2.3476059436798096, + "learning_rate": 6.058873320792655e-05, + "loss": 0.4037, + "step": 36897 + }, + { + "epoch": 2.5604052459926443, + "grad_norm": 0.9111179709434509, + "learning_rate": 6.058356941341778e-05, + "loss": 0.3425, + "step": 36898 + }, + { + "epoch": 2.560474637429741, + "grad_norm": 1.2643041610717773, + "learning_rate": 6.0578405743342084e-05, + "loss": 0.4362, + "step": 36899 + }, + { + "epoch": 2.560544028866838, + "grad_norm": 1.2667547464370728, + "learning_rate": 6.0573242197715785e-05, + "loss": 0.3865, + "step": 36900 + }, + { + "epoch": 2.5606134203039344, + "grad_norm": 1.399066686630249, + "learning_rate": 6.0568078776555185e-05, + "loss": 0.2987, + "step": 36901 + }, + { + "epoch": 2.560682811741031, + "grad_norm": 1.0564777851104736, + "learning_rate": 6.056291547987658e-05, + "loss": 0.3516, + "step": 36902 + }, + { + "epoch": 2.5607522031781276, + "grad_norm": 1.0110851526260376, + "learning_rate": 6.055775230769629e-05, + "loss": 0.3117, + "step": 36903 + }, + { + "epoch": 2.5608215946152244, + "grad_norm": 1.0886929035186768, + "learning_rate": 6.055258926003059e-05, + "loss": 0.4511, + "step": 36904 + }, + { + "epoch": 2.5608909860523212, + "grad_norm": 1.1895941495895386, + "learning_rate": 6.054742633689581e-05, + "loss": 0.3672, + "step": 36905 + }, + { + "epoch": 2.560960377489418, + "grad_norm": 1.7723942995071411, + "learning_rate": 6.05422635383082e-05, + "loss": 0.4983, + "step": 36906 + }, + { + "epoch": 2.5610297689265145, + "grad_norm": 1.4008415937423706, + "learning_rate": 6.05371008642841e-05, + "loss": 0.3164, + "step": 36907 + }, + { + "epoch": 2.5610991603636113, + "grad_norm": 1.5119751691818237, + "learning_rate": 6.053193831483979e-05, + "loss": 0.4584, + "step": 36908 + }, + { + "epoch": 2.5611685518007077, + "grad_norm": 1.0772885084152222, + "learning_rate": 6.052677588999156e-05, + "loss": 0.2828, + "step": 36909 + }, + { + "epoch": 2.5612379432378045, + "grad_norm": 1.0332528352737427, + "learning_rate": 6.052161358975573e-05, + "loss": 0.2832, + "step": 36910 + }, + { + "epoch": 2.5613073346749013, + "grad_norm": 1.5943405628204346, + "learning_rate": 6.0516451414148576e-05, + "loss": 0.3545, + "step": 36911 + }, + { + "epoch": 2.5613767261119977, + "grad_norm": 0.9619333148002625, + "learning_rate": 6.051128936318645e-05, + "loss": 0.4783, + "step": 36912 + }, + { + "epoch": 2.5614461175490946, + "grad_norm": 0.8884682655334473, + "learning_rate": 6.050612743688556e-05, + "loss": 0.2631, + "step": 36913 + }, + { + "epoch": 2.561515508986191, + "grad_norm": 1.8444414138793945, + "learning_rate": 6.050096563526224e-05, + "loss": 0.5067, + "step": 36914 + }, + { + "epoch": 2.561584900423288, + "grad_norm": 1.4195224046707153, + "learning_rate": 6.04958039583328e-05, + "loss": 0.6476, + "step": 36915 + }, + { + "epoch": 2.5616542918603846, + "grad_norm": 1.1703044176101685, + "learning_rate": 6.0490642406113506e-05, + "loss": 0.3719, + "step": 36916 + }, + { + "epoch": 2.561723683297481, + "grad_norm": 1.573043942451477, + "learning_rate": 6.048548097862068e-05, + "loss": 0.3141, + "step": 36917 + }, + { + "epoch": 2.561793074734578, + "grad_norm": 1.209387183189392, + "learning_rate": 6.04803196758706e-05, + "loss": 0.3161, + "step": 36918 + }, + { + "epoch": 2.5618624661716742, + "grad_norm": 1.0910792350769043, + "learning_rate": 6.0475158497879615e-05, + "loss": 0.2961, + "step": 36919 + }, + { + "epoch": 2.561931857608771, + "grad_norm": 1.6883970499038696, + "learning_rate": 6.0469997444663926e-05, + "loss": 0.6304, + "step": 36920 + }, + { + "epoch": 2.562001249045868, + "grad_norm": 1.4304651021957397, + "learning_rate": 6.0464836516239865e-05, + "loss": 0.3133, + "step": 36921 + }, + { + "epoch": 2.5620706404829643, + "grad_norm": 2.0114169120788574, + "learning_rate": 6.0459675712623765e-05, + "loss": 0.3812, + "step": 36922 + }, + { + "epoch": 2.562140031920061, + "grad_norm": 1.2949855327606201, + "learning_rate": 6.045451503383184e-05, + "loss": 0.3735, + "step": 36923 + }, + { + "epoch": 2.5622094233571575, + "grad_norm": 1.1624623537063599, + "learning_rate": 6.044935447988043e-05, + "loss": 0.3735, + "step": 36924 + }, + { + "epoch": 2.5622788147942543, + "grad_norm": 1.181104302406311, + "learning_rate": 6.0444194050785815e-05, + "loss": 0.4052, + "step": 36925 + }, + { + "epoch": 2.562348206231351, + "grad_norm": 1.2827759981155396, + "learning_rate": 6.043903374656434e-05, + "loss": 0.4448, + "step": 36926 + }, + { + "epoch": 2.5624175976684476, + "grad_norm": 1.3212310075759888, + "learning_rate": 6.0433873567232205e-05, + "loss": 0.4097, + "step": 36927 + }, + { + "epoch": 2.5624869891055444, + "grad_norm": 1.568320631980896, + "learning_rate": 6.042871351280572e-05, + "loss": 0.488, + "step": 36928 + }, + { + "epoch": 2.562556380542641, + "grad_norm": 1.3548957109451294, + "learning_rate": 6.0423553583301216e-05, + "loss": 0.4628, + "step": 36929 + }, + { + "epoch": 2.5626257719797376, + "grad_norm": 1.589311957359314, + "learning_rate": 6.041839377873495e-05, + "loss": 0.6136, + "step": 36930 + }, + { + "epoch": 2.5626951634168345, + "grad_norm": 1.4050203561782837, + "learning_rate": 6.041323409912323e-05, + "loss": 0.3677, + "step": 36931 + }, + { + "epoch": 2.5627645548539313, + "grad_norm": 1.70590341091156, + "learning_rate": 6.0408074544482385e-05, + "loss": 0.4219, + "step": 36932 + }, + { + "epoch": 2.5628339462910277, + "grad_norm": 2.250723361968994, + "learning_rate": 6.040291511482861e-05, + "loss": 0.5965, + "step": 36933 + }, + { + "epoch": 2.5629033377281245, + "grad_norm": 1.2610328197479248, + "learning_rate": 6.039775581017822e-05, + "loss": 0.4173, + "step": 36934 + }, + { + "epoch": 2.562972729165221, + "grad_norm": 0.9953765273094177, + "learning_rate": 6.039259663054754e-05, + "loss": 0.2955, + "step": 36935 + }, + { + "epoch": 2.5630421206023177, + "grad_norm": 3.0585920810699463, + "learning_rate": 6.038743757595283e-05, + "loss": 0.3131, + "step": 36936 + }, + { + "epoch": 2.5631115120394146, + "grad_norm": 0.8813270926475525, + "learning_rate": 6.038227864641038e-05, + "loss": 0.3227, + "step": 36937 + }, + { + "epoch": 2.563180903476511, + "grad_norm": 1.251787781715393, + "learning_rate": 6.03771198419365e-05, + "loss": 0.4265, + "step": 36938 + }, + { + "epoch": 2.563250294913608, + "grad_norm": 1.5019840002059937, + "learning_rate": 6.037196116254746e-05, + "loss": 0.5017, + "step": 36939 + }, + { + "epoch": 2.563319686350704, + "grad_norm": 1.2742682695388794, + "learning_rate": 6.0366802608259535e-05, + "loss": 0.4029, + "step": 36940 + }, + { + "epoch": 2.563389077787801, + "grad_norm": 0.984775185585022, + "learning_rate": 6.0361644179089005e-05, + "loss": 0.2822, + "step": 36941 + }, + { + "epoch": 2.563458469224898, + "grad_norm": 1.3535051345825195, + "learning_rate": 6.0356485875052183e-05, + "loss": 0.4446, + "step": 36942 + }, + { + "epoch": 2.5635278606619942, + "grad_norm": 1.0093847513198853, + "learning_rate": 6.035132769616531e-05, + "loss": 0.3352, + "step": 36943 + }, + { + "epoch": 2.563597252099091, + "grad_norm": 1.058925986289978, + "learning_rate": 6.0346169642444706e-05, + "loss": 0.4058, + "step": 36944 + }, + { + "epoch": 2.5636666435361875, + "grad_norm": 1.8470230102539062, + "learning_rate": 6.034101171390662e-05, + "loss": 0.4069, + "step": 36945 + }, + { + "epoch": 2.5637360349732843, + "grad_norm": 1.8501836061477661, + "learning_rate": 6.033585391056742e-05, + "loss": 0.4989, + "step": 36946 + }, + { + "epoch": 2.563805426410381, + "grad_norm": 1.8346532583236694, + "learning_rate": 6.0330696232443294e-05, + "loss": 0.2961, + "step": 36947 + }, + { + "epoch": 2.5638748178474775, + "grad_norm": 1.2328591346740723, + "learning_rate": 6.032553867955054e-05, + "loss": 0.2954, + "step": 36948 + }, + { + "epoch": 2.5639442092845743, + "grad_norm": 1.502807378768921, + "learning_rate": 6.0320381251905465e-05, + "loss": 0.4271, + "step": 36949 + }, + { + "epoch": 2.5640136007216707, + "grad_norm": 1.123870849609375, + "learning_rate": 6.0315223949524345e-05, + "loss": 0.5511, + "step": 36950 + }, + { + "epoch": 2.5640829921587676, + "grad_norm": 1.7361680269241333, + "learning_rate": 6.031006677242346e-05, + "loss": 0.4076, + "step": 36951 + }, + { + "epoch": 2.5641523835958644, + "grad_norm": 1.0328525304794312, + "learning_rate": 6.0304909720619084e-05, + "loss": 0.2359, + "step": 36952 + }, + { + "epoch": 2.564221775032961, + "grad_norm": 0.9487125277519226, + "learning_rate": 6.029975279412754e-05, + "loss": 0.3875, + "step": 36953 + }, + { + "epoch": 2.5642911664700576, + "grad_norm": 1.231052041053772, + "learning_rate": 6.029459599296503e-05, + "loss": 0.3952, + "step": 36954 + }, + { + "epoch": 2.564360557907154, + "grad_norm": 1.232567310333252, + "learning_rate": 6.028943931714788e-05, + "loss": 0.4378, + "step": 36955 + }, + { + "epoch": 2.564429949344251, + "grad_norm": 1.2370887994766235, + "learning_rate": 6.0284282766692366e-05, + "loss": 0.3575, + "step": 36956 + }, + { + "epoch": 2.5644993407813477, + "grad_norm": 1.3074595928192139, + "learning_rate": 6.027912634161478e-05, + "loss": 0.3881, + "step": 36957 + }, + { + "epoch": 2.5645687322184445, + "grad_norm": 1.178937315940857, + "learning_rate": 6.027397004193136e-05, + "loss": 0.3996, + "step": 36958 + }, + { + "epoch": 2.564638123655541, + "grad_norm": 0.9332690238952637, + "learning_rate": 6.026881386765841e-05, + "loss": 0.4618, + "step": 36959 + }, + { + "epoch": 2.5647075150926377, + "grad_norm": 1.1955033540725708, + "learning_rate": 6.026365781881223e-05, + "loss": 0.3836, + "step": 36960 + }, + { + "epoch": 2.564776906529734, + "grad_norm": 1.6769044399261475, + "learning_rate": 6.025850189540907e-05, + "loss": 0.6561, + "step": 36961 + }, + { + "epoch": 2.564846297966831, + "grad_norm": 1.371705174446106, + "learning_rate": 6.025334609746518e-05, + "loss": 0.2807, + "step": 36962 + }, + { + "epoch": 2.564915689403928, + "grad_norm": 1.3366247415542603, + "learning_rate": 6.0248190424996867e-05, + "loss": 0.4876, + "step": 36963 + }, + { + "epoch": 2.564985080841024, + "grad_norm": 1.3520467281341553, + "learning_rate": 6.024303487802041e-05, + "loss": 0.5083, + "step": 36964 + }, + { + "epoch": 2.565054472278121, + "grad_norm": 1.4443293809890747, + "learning_rate": 6.0237879456552084e-05, + "loss": 0.4797, + "step": 36965 + }, + { + "epoch": 2.5651238637152174, + "grad_norm": 1.2195004224777222, + "learning_rate": 6.023272416060814e-05, + "loss": 0.4359, + "step": 36966 + }, + { + "epoch": 2.5651932551523142, + "grad_norm": 1.435289978981018, + "learning_rate": 6.022756899020493e-05, + "loss": 0.5216, + "step": 36967 + }, + { + "epoch": 2.565262646589411, + "grad_norm": 1.080973505973816, + "learning_rate": 6.022241394535862e-05, + "loss": 0.3851, + "step": 36968 + }, + { + "epoch": 2.5653320380265074, + "grad_norm": 1.1933726072311401, + "learning_rate": 6.021725902608555e-05, + "loss": 0.3596, + "step": 36969 + }, + { + "epoch": 2.5654014294636043, + "grad_norm": 1.2568104267120361, + "learning_rate": 6.021210423240196e-05, + "loss": 0.4431, + "step": 36970 + }, + { + "epoch": 2.5654708209007007, + "grad_norm": 1.2060439586639404, + "learning_rate": 6.020694956432414e-05, + "loss": 0.3422, + "step": 36971 + }, + { + "epoch": 2.5655402123377975, + "grad_norm": 1.2249561548233032, + "learning_rate": 6.020179502186838e-05, + "loss": 0.4459, + "step": 36972 + }, + { + "epoch": 2.5656096037748943, + "grad_norm": 1.2333499193191528, + "learning_rate": 6.0196640605050944e-05, + "loss": 0.3416, + "step": 36973 + }, + { + "epoch": 2.5656789952119907, + "grad_norm": 1.9977023601531982, + "learning_rate": 6.019148631388809e-05, + "loss": 0.4378, + "step": 36974 + }, + { + "epoch": 2.5657483866490876, + "grad_norm": 1.4978516101837158, + "learning_rate": 6.0186332148396084e-05, + "loss": 0.4653, + "step": 36975 + }, + { + "epoch": 2.565817778086184, + "grad_norm": 1.131758689880371, + "learning_rate": 6.018117810859121e-05, + "loss": 0.4143, + "step": 36976 + }, + { + "epoch": 2.5658871695232808, + "grad_norm": 1.248748779296875, + "learning_rate": 6.0176024194489756e-05, + "loss": 0.439, + "step": 36977 + }, + { + "epoch": 2.5659565609603776, + "grad_norm": 1.1821119785308838, + "learning_rate": 6.017087040610795e-05, + "loss": 0.3816, + "step": 36978 + }, + { + "epoch": 2.566025952397474, + "grad_norm": 0.808421790599823, + "learning_rate": 6.0165716743462084e-05, + "loss": 0.3009, + "step": 36979 + }, + { + "epoch": 2.566095343834571, + "grad_norm": 1.2906720638275146, + "learning_rate": 6.016056320656843e-05, + "loss": 0.4952, + "step": 36980 + }, + { + "epoch": 2.566164735271667, + "grad_norm": 0.8227280974388123, + "learning_rate": 6.0155409795443296e-05, + "loss": 0.2651, + "step": 36981 + }, + { + "epoch": 2.566234126708764, + "grad_norm": 0.9527426362037659, + "learning_rate": 6.0150256510102864e-05, + "loss": 0.3096, + "step": 36982 + }, + { + "epoch": 2.566303518145861, + "grad_norm": 1.1987792253494263, + "learning_rate": 6.014510335056347e-05, + "loss": 0.4673, + "step": 36983 + }, + { + "epoch": 2.5663729095829577, + "grad_norm": 0.9602339267730713, + "learning_rate": 6.013995031684134e-05, + "loss": 0.3174, + "step": 36984 + }, + { + "epoch": 2.566442301020054, + "grad_norm": 1.0911976099014282, + "learning_rate": 6.013479740895277e-05, + "loss": 0.4092, + "step": 36985 + }, + { + "epoch": 2.566511692457151, + "grad_norm": 1.4393759965896606, + "learning_rate": 6.0129644626914016e-05, + "loss": 0.4247, + "step": 36986 + }, + { + "epoch": 2.5665810838942473, + "grad_norm": 1.1325265169143677, + "learning_rate": 6.012449197074135e-05, + "loss": 0.4026, + "step": 36987 + }, + { + "epoch": 2.566650475331344, + "grad_norm": 1.442008376121521, + "learning_rate": 6.0119339440451073e-05, + "loss": 0.4899, + "step": 36988 + }, + { + "epoch": 2.566719866768441, + "grad_norm": 1.3230259418487549, + "learning_rate": 6.011418703605938e-05, + "loss": 0.4494, + "step": 36989 + }, + { + "epoch": 2.5667892582055374, + "grad_norm": 1.063446283340454, + "learning_rate": 6.010903475758256e-05, + "loss": 0.3809, + "step": 36990 + }, + { + "epoch": 2.566858649642634, + "grad_norm": 1.1093106269836426, + "learning_rate": 6.01038826050369e-05, + "loss": 0.3845, + "step": 36991 + }, + { + "epoch": 2.5669280410797306, + "grad_norm": 1.8312865495681763, + "learning_rate": 6.009873057843868e-05, + "loss": 0.3324, + "step": 36992 + }, + { + "epoch": 2.5669974325168274, + "grad_norm": 1.1995128393173218, + "learning_rate": 6.009357867780411e-05, + "loss": 0.5364, + "step": 36993 + }, + { + "epoch": 2.5670668239539243, + "grad_norm": 1.621145248413086, + "learning_rate": 6.0088426903149486e-05, + "loss": 0.4127, + "step": 36994 + }, + { + "epoch": 2.5671362153910207, + "grad_norm": 1.0524907112121582, + "learning_rate": 6.0083275254491076e-05, + "loss": 0.2939, + "step": 36995 + }, + { + "epoch": 2.5672056068281175, + "grad_norm": 1.2265781164169312, + "learning_rate": 6.007812373184514e-05, + "loss": 0.4859, + "step": 36996 + }, + { + "epoch": 2.567274998265214, + "grad_norm": 1.1175651550292969, + "learning_rate": 6.007297233522793e-05, + "loss": 0.2892, + "step": 36997 + }, + { + "epoch": 2.5673443897023107, + "grad_norm": 1.222071886062622, + "learning_rate": 6.00678210646557e-05, + "loss": 0.3938, + "step": 36998 + }, + { + "epoch": 2.5674137811394075, + "grad_norm": 1.0255687236785889, + "learning_rate": 6.0062669920144723e-05, + "loss": 0.3203, + "step": 36999 + }, + { + "epoch": 2.567483172576504, + "grad_norm": 0.8886049389839172, + "learning_rate": 6.005751890171128e-05, + "loss": 0.3545, + "step": 37000 + }, + { + "epoch": 2.5675525640136008, + "grad_norm": 1.0870721340179443, + "learning_rate": 6.005236800937159e-05, + "loss": 0.3308, + "step": 37001 + }, + { + "epoch": 2.567621955450697, + "grad_norm": 2.042130947113037, + "learning_rate": 6.0047217243141997e-05, + "loss": 0.5611, + "step": 37002 + }, + { + "epoch": 2.567691346887794, + "grad_norm": 0.915542721748352, + "learning_rate": 6.004206660303866e-05, + "loss": 0.3701, + "step": 37003 + }, + { + "epoch": 2.567760738324891, + "grad_norm": 1.4908695220947266, + "learning_rate": 6.003691608907788e-05, + "loss": 0.4258, + "step": 37004 + }, + { + "epoch": 2.567830129761987, + "grad_norm": 1.2342066764831543, + "learning_rate": 6.003176570127592e-05, + "loss": 0.3361, + "step": 37005 + }, + { + "epoch": 2.567899521199084, + "grad_norm": 1.3056727647781372, + "learning_rate": 6.002661543964905e-05, + "loss": 0.3543, + "step": 37006 + }, + { + "epoch": 2.5679689126361804, + "grad_norm": 1.1504361629486084, + "learning_rate": 6.0021465304213495e-05, + "loss": 0.3964, + "step": 37007 + }, + { + "epoch": 2.5680383040732773, + "grad_norm": 0.9184243083000183, + "learning_rate": 6.001631529498559e-05, + "loss": 0.3972, + "step": 37008 + }, + { + "epoch": 2.568107695510374, + "grad_norm": 1.388527750968933, + "learning_rate": 6.001116541198151e-05, + "loss": 0.3917, + "step": 37009 + }, + { + "epoch": 2.568177086947471, + "grad_norm": 0.9960137605667114, + "learning_rate": 6.000601565521752e-05, + "loss": 0.2798, + "step": 37010 + }, + { + "epoch": 2.5682464783845673, + "grad_norm": 1.3857405185699463, + "learning_rate": 6.000086602470992e-05, + "loss": 0.4021, + "step": 37011 + }, + { + "epoch": 2.568315869821664, + "grad_norm": 1.3505016565322876, + "learning_rate": 5.999571652047493e-05, + "loss": 0.4156, + "step": 37012 + }, + { + "epoch": 2.5683852612587605, + "grad_norm": 1.323769450187683, + "learning_rate": 5.999056714252883e-05, + "loss": 0.5697, + "step": 37013 + }, + { + "epoch": 2.5684546526958574, + "grad_norm": 1.0780011415481567, + "learning_rate": 5.9985417890887864e-05, + "loss": 0.3015, + "step": 37014 + }, + { + "epoch": 2.568524044132954, + "grad_norm": 1.1066839694976807, + "learning_rate": 5.9980268765568325e-05, + "loss": 0.3556, + "step": 37015 + }, + { + "epoch": 2.5685934355700506, + "grad_norm": 0.9557966589927673, + "learning_rate": 5.9975119766586406e-05, + "loss": 0.2913, + "step": 37016 + }, + { + "epoch": 2.5686628270071474, + "grad_norm": 1.2704756259918213, + "learning_rate": 5.996997089395837e-05, + "loss": 0.3604, + "step": 37017 + }, + { + "epoch": 2.568732218444244, + "grad_norm": 1.6071856021881104, + "learning_rate": 5.9964822147700515e-05, + "loss": 0.2982, + "step": 37018 + }, + { + "epoch": 2.5688016098813407, + "grad_norm": 1.4358296394348145, + "learning_rate": 5.995967352782905e-05, + "loss": 0.3384, + "step": 37019 + }, + { + "epoch": 2.5688710013184375, + "grad_norm": 1.279611349105835, + "learning_rate": 5.995452503436026e-05, + "loss": 0.3044, + "step": 37020 + }, + { + "epoch": 2.568940392755534, + "grad_norm": 0.7437955141067505, + "learning_rate": 5.99493766673104e-05, + "loss": 0.25, + "step": 37021 + }, + { + "epoch": 2.5690097841926307, + "grad_norm": 1.3691496849060059, + "learning_rate": 5.994422842669574e-05, + "loss": 0.528, + "step": 37022 + }, + { + "epoch": 2.569079175629727, + "grad_norm": 1.4643540382385254, + "learning_rate": 5.993908031253246e-05, + "loss": 0.4457, + "step": 37023 + }, + { + "epoch": 2.569148567066824, + "grad_norm": 1.3667378425598145, + "learning_rate": 5.9933932324836864e-05, + "loss": 0.432, + "step": 37024 + }, + { + "epoch": 2.5692179585039208, + "grad_norm": 1.305594563484192, + "learning_rate": 5.9928784463625175e-05, + "loss": 0.3484, + "step": 37025 + }, + { + "epoch": 2.569287349941017, + "grad_norm": 1.060725212097168, + "learning_rate": 5.9923636728913676e-05, + "loss": 0.3053, + "step": 37026 + }, + { + "epoch": 2.569356741378114, + "grad_norm": 1.547142505645752, + "learning_rate": 5.9918489120718636e-05, + "loss": 0.4087, + "step": 37027 + }, + { + "epoch": 2.5694261328152104, + "grad_norm": 1.2983510494232178, + "learning_rate": 5.9913341639056245e-05, + "loss": 0.4099, + "step": 37028 + }, + { + "epoch": 2.569495524252307, + "grad_norm": 1.1724697351455688, + "learning_rate": 5.990819428394281e-05, + "loss": 0.4006, + "step": 37029 + }, + { + "epoch": 2.569564915689404, + "grad_norm": 1.0949585437774658, + "learning_rate": 5.990304705539453e-05, + "loss": 0.382, + "step": 37030 + }, + { + "epoch": 2.5696343071265004, + "grad_norm": 0.9349281787872314, + "learning_rate": 5.9897899953427704e-05, + "loss": 0.388, + "step": 37031 + }, + { + "epoch": 2.5697036985635973, + "grad_norm": 1.0789419412612915, + "learning_rate": 5.989275297805853e-05, + "loss": 0.3122, + "step": 37032 + }, + { + "epoch": 2.5697730900006936, + "grad_norm": 1.3055062294006348, + "learning_rate": 5.988760612930329e-05, + "loss": 0.3925, + "step": 37033 + }, + { + "epoch": 2.5698424814377905, + "grad_norm": 1.224684238433838, + "learning_rate": 5.988245940717821e-05, + "loss": 0.3621, + "step": 37034 + }, + { + "epoch": 2.5699118728748873, + "grad_norm": 1.4549574851989746, + "learning_rate": 5.987731281169955e-05, + "loss": 0.4454, + "step": 37035 + }, + { + "epoch": 2.569981264311984, + "grad_norm": 1.3048807382583618, + "learning_rate": 5.987216634288362e-05, + "loss": 0.2762, + "step": 37036 + }, + { + "epoch": 2.5700506557490805, + "grad_norm": 1.073610782623291, + "learning_rate": 5.9867020000746544e-05, + "loss": 0.4136, + "step": 37037 + }, + { + "epoch": 2.5701200471861774, + "grad_norm": 1.050964593887329, + "learning_rate": 5.986187378530463e-05, + "loss": 0.4124, + "step": 37038 + }, + { + "epoch": 2.5701894386232738, + "grad_norm": 1.5940024852752686, + "learning_rate": 5.985672769657413e-05, + "loss": 0.5481, + "step": 37039 + }, + { + "epoch": 2.5702588300603706, + "grad_norm": 1.712015151977539, + "learning_rate": 5.985158173457128e-05, + "loss": 0.5195, + "step": 37040 + }, + { + "epoch": 2.5703282214974674, + "grad_norm": 2.562274694442749, + "learning_rate": 5.9846435899312335e-05, + "loss": 0.4671, + "step": 37041 + }, + { + "epoch": 2.570397612934564, + "grad_norm": 1.2876838445663452, + "learning_rate": 5.984129019081352e-05, + "loss": 0.3987, + "step": 37042 + }, + { + "epoch": 2.5704670043716606, + "grad_norm": 1.173682689666748, + "learning_rate": 5.983614460909115e-05, + "loss": 0.4222, + "step": 37043 + }, + { + "epoch": 2.570536395808757, + "grad_norm": 1.7826898097991943, + "learning_rate": 5.983099915416137e-05, + "loss": 0.4396, + "step": 37044 + }, + { + "epoch": 2.570605787245854, + "grad_norm": 0.9686753153800964, + "learning_rate": 5.982585382604046e-05, + "loss": 0.3758, + "step": 37045 + }, + { + "epoch": 2.5706751786829507, + "grad_norm": 1.5315386056900024, + "learning_rate": 5.982070862474469e-05, + "loss": 0.4347, + "step": 37046 + }, + { + "epoch": 2.570744570120047, + "grad_norm": 1.082506537437439, + "learning_rate": 5.981556355029026e-05, + "loss": 0.3339, + "step": 37047 + }, + { + "epoch": 2.570813961557144, + "grad_norm": 1.0267714262008667, + "learning_rate": 5.981041860269343e-05, + "loss": 0.4591, + "step": 37048 + }, + { + "epoch": 2.5708833529942403, + "grad_norm": 0.9354674220085144, + "learning_rate": 5.980527378197045e-05, + "loss": 0.3718, + "step": 37049 + }, + { + "epoch": 2.570952744431337, + "grad_norm": 0.9983411431312561, + "learning_rate": 5.980012908813759e-05, + "loss": 0.2697, + "step": 37050 + }, + { + "epoch": 2.571022135868434, + "grad_norm": 1.9958617687225342, + "learning_rate": 5.979498452121102e-05, + "loss": 0.4305, + "step": 37051 + }, + { + "epoch": 2.5710915273055304, + "grad_norm": 1.0577856302261353, + "learning_rate": 5.9789840081207025e-05, + "loss": 0.5183, + "step": 37052 + }, + { + "epoch": 2.571160918742627, + "grad_norm": 1.082759141921997, + "learning_rate": 5.978469576814183e-05, + "loss": 0.3335, + "step": 37053 + }, + { + "epoch": 2.5712303101797236, + "grad_norm": 1.6606297492980957, + "learning_rate": 5.97795515820317e-05, + "loss": 0.3499, + "step": 37054 + }, + { + "epoch": 2.5712997016168204, + "grad_norm": 0.9553864002227783, + "learning_rate": 5.977440752289285e-05, + "loss": 0.2954, + "step": 37055 + }, + { + "epoch": 2.5713690930539173, + "grad_norm": 1.2943029403686523, + "learning_rate": 5.976926359074153e-05, + "loss": 0.5018, + "step": 37056 + }, + { + "epoch": 2.5714384844910136, + "grad_norm": 0.9126467108726501, + "learning_rate": 5.976411978559402e-05, + "loss": 0.2548, + "step": 37057 + }, + { + "epoch": 2.5715078759281105, + "grad_norm": 1.0093334913253784, + "learning_rate": 5.975897610746647e-05, + "loss": 0.3843, + "step": 37058 + }, + { + "epoch": 2.571577267365207, + "grad_norm": 1.0116660594940186, + "learning_rate": 5.975383255637517e-05, + "loss": 0.2734, + "step": 37059 + }, + { + "epoch": 2.5716466588023037, + "grad_norm": 1.3285632133483887, + "learning_rate": 5.9748689132336344e-05, + "loss": 0.452, + "step": 37060 + }, + { + "epoch": 2.5717160502394005, + "grad_norm": 1.2895102500915527, + "learning_rate": 5.974354583536623e-05, + "loss": 0.4518, + "step": 37061 + }, + { + "epoch": 2.5717854416764974, + "grad_norm": 0.9255148768424988, + "learning_rate": 5.973840266548111e-05, + "loss": 0.2683, + "step": 37062 + }, + { + "epoch": 2.5718548331135938, + "grad_norm": 0.9532141089439392, + "learning_rate": 5.973325962269716e-05, + "loss": 0.355, + "step": 37063 + }, + { + "epoch": 2.5719242245506906, + "grad_norm": 1.3863273859024048, + "learning_rate": 5.9728116707030644e-05, + "loss": 0.4541, + "step": 37064 + }, + { + "epoch": 2.571993615987787, + "grad_norm": 1.1205202341079712, + "learning_rate": 5.972297391849779e-05, + "loss": 0.4624, + "step": 37065 + }, + { + "epoch": 2.572063007424884, + "grad_norm": 1.6992026567459106, + "learning_rate": 5.971783125711483e-05, + "loss": 0.4974, + "step": 37066 + }, + { + "epoch": 2.5721323988619806, + "grad_norm": 0.7676100134849548, + "learning_rate": 5.9712688722897994e-05, + "loss": 0.2655, + "step": 37067 + }, + { + "epoch": 2.572201790299077, + "grad_norm": 1.3492261171340942, + "learning_rate": 5.970754631586353e-05, + "loss": 0.4524, + "step": 37068 + }, + { + "epoch": 2.572271181736174, + "grad_norm": 0.9784783720970154, + "learning_rate": 5.970240403602765e-05, + "loss": 0.3508, + "step": 37069 + }, + { + "epoch": 2.5723405731732703, + "grad_norm": 1.323024868965149, + "learning_rate": 5.969726188340663e-05, + "loss": 0.4072, + "step": 37070 + }, + { + "epoch": 2.572409964610367, + "grad_norm": 1.1612519025802612, + "learning_rate": 5.9692119858016704e-05, + "loss": 0.2888, + "step": 37071 + }, + { + "epoch": 2.572479356047464, + "grad_norm": 1.3075803518295288, + "learning_rate": 5.9686977959874036e-05, + "loss": 0.3625, + "step": 37072 + }, + { + "epoch": 2.5725487474845603, + "grad_norm": 0.9766340851783752, + "learning_rate": 5.968183618899491e-05, + "loss": 0.3887, + "step": 37073 + }, + { + "epoch": 2.572618138921657, + "grad_norm": 1.2154557704925537, + "learning_rate": 5.967669454539555e-05, + "loss": 0.4216, + "step": 37074 + }, + { + "epoch": 2.5726875303587535, + "grad_norm": 1.3157930374145508, + "learning_rate": 5.967155302909218e-05, + "loss": 0.4858, + "step": 37075 + }, + { + "epoch": 2.5727569217958504, + "grad_norm": 0.8867691159248352, + "learning_rate": 5.966641164010105e-05, + "loss": 0.3711, + "step": 37076 + }, + { + "epoch": 2.572826313232947, + "grad_norm": 0.9690468311309814, + "learning_rate": 5.9661270378438385e-05, + "loss": 0.3692, + "step": 37077 + }, + { + "epoch": 2.5728957046700436, + "grad_norm": 1.1329008340835571, + "learning_rate": 5.9656129244120403e-05, + "loss": 0.2898, + "step": 37078 + }, + { + "epoch": 2.5729650961071404, + "grad_norm": 1.248279094696045, + "learning_rate": 5.965098823716334e-05, + "loss": 0.5414, + "step": 37079 + }, + { + "epoch": 2.573034487544237, + "grad_norm": 0.899504542350769, + "learning_rate": 5.9645847357583415e-05, + "loss": 0.3862, + "step": 37080 + }, + { + "epoch": 2.5731038789813336, + "grad_norm": 1.0940135717391968, + "learning_rate": 5.964070660539689e-05, + "loss": 0.3146, + "step": 37081 + }, + { + "epoch": 2.5731732704184305, + "grad_norm": 1.3490146398544312, + "learning_rate": 5.9635565980619956e-05, + "loss": 0.2985, + "step": 37082 + }, + { + "epoch": 2.5732426618555273, + "grad_norm": 0.6779846549034119, + "learning_rate": 5.963042548326885e-05, + "loss": 0.2329, + "step": 37083 + }, + { + "epoch": 2.5733120532926237, + "grad_norm": 1.0803091526031494, + "learning_rate": 5.962528511335983e-05, + "loss": 0.4496, + "step": 37084 + }, + { + "epoch": 2.57338144472972, + "grad_norm": 2.2958061695098877, + "learning_rate": 5.96201448709091e-05, + "loss": 0.2764, + "step": 37085 + }, + { + "epoch": 2.573450836166817, + "grad_norm": 1.6434457302093506, + "learning_rate": 5.961500475593287e-05, + "loss": 0.5245, + "step": 37086 + }, + { + "epoch": 2.5735202276039137, + "grad_norm": 1.4463388919830322, + "learning_rate": 5.960986476844739e-05, + "loss": 0.4983, + "step": 37087 + }, + { + "epoch": 2.5735896190410106, + "grad_norm": 1.017720103263855, + "learning_rate": 5.9604724908468865e-05, + "loss": 0.3892, + "step": 37088 + }, + { + "epoch": 2.573659010478107, + "grad_norm": 1.1046242713928223, + "learning_rate": 5.959958517601356e-05, + "loss": 0.3943, + "step": 37089 + }, + { + "epoch": 2.573728401915204, + "grad_norm": 1.021822452545166, + "learning_rate": 5.9594445571097656e-05, + "loss": 0.2698, + "step": 37090 + }, + { + "epoch": 2.5737977933523, + "grad_norm": 1.102108359336853, + "learning_rate": 5.958930609373745e-05, + "loss": 0.3775, + "step": 37091 + }, + { + "epoch": 2.573867184789397, + "grad_norm": 1.9301364421844482, + "learning_rate": 5.958416674394906e-05, + "loss": 0.3415, + "step": 37092 + }, + { + "epoch": 2.573936576226494, + "grad_norm": 1.0586775541305542, + "learning_rate": 5.9579027521748796e-05, + "loss": 0.4625, + "step": 37093 + }, + { + "epoch": 2.5740059676635902, + "grad_norm": 1.701870083808899, + "learning_rate": 5.9573888427152836e-05, + "loss": 0.3856, + "step": 37094 + }, + { + "epoch": 2.574075359100687, + "grad_norm": 0.9676422476768494, + "learning_rate": 5.9568749460177406e-05, + "loss": 0.2555, + "step": 37095 + }, + { + "epoch": 2.5741447505377835, + "grad_norm": 1.6407312154769897, + "learning_rate": 5.956361062083877e-05, + "loss": 0.4374, + "step": 37096 + }, + { + "epoch": 2.5742141419748803, + "grad_norm": 1.1030198335647583, + "learning_rate": 5.955847190915313e-05, + "loss": 0.3501, + "step": 37097 + }, + { + "epoch": 2.574283533411977, + "grad_norm": 1.3549835681915283, + "learning_rate": 5.9553333325136695e-05, + "loss": 0.3816, + "step": 37098 + }, + { + "epoch": 2.5743529248490735, + "grad_norm": 0.9987194538116455, + "learning_rate": 5.954819486880568e-05, + "loss": 0.2584, + "step": 37099 + }, + { + "epoch": 2.5744223162861704, + "grad_norm": 1.779966115951538, + "learning_rate": 5.9543056540176345e-05, + "loss": 0.5332, + "step": 37100 + }, + { + "epoch": 2.5744917077232667, + "grad_norm": 1.2759673595428467, + "learning_rate": 5.9537918339264865e-05, + "loss": 0.5155, + "step": 37101 + }, + { + "epoch": 2.5745610991603636, + "grad_norm": 1.095849871635437, + "learning_rate": 5.9532780266087485e-05, + "loss": 0.2904, + "step": 37102 + }, + { + "epoch": 2.5746304905974604, + "grad_norm": 1.5485422611236572, + "learning_rate": 5.952764232066041e-05, + "loss": 0.6275, + "step": 37103 + }, + { + "epoch": 2.574699882034557, + "grad_norm": 1.227479100227356, + "learning_rate": 5.9522504502999886e-05, + "loss": 0.4907, + "step": 37104 + }, + { + "epoch": 2.5747692734716536, + "grad_norm": 1.0995701551437378, + "learning_rate": 5.951736681312215e-05, + "loss": 0.3501, + "step": 37105 + }, + { + "epoch": 2.57483866490875, + "grad_norm": 0.9280716180801392, + "learning_rate": 5.951222925104336e-05, + "loss": 0.2792, + "step": 37106 + }, + { + "epoch": 2.574908056345847, + "grad_norm": 2.4847912788391113, + "learning_rate": 5.9507091816779756e-05, + "loss": 0.5512, + "step": 37107 + }, + { + "epoch": 2.5749774477829437, + "grad_norm": 1.5301363468170166, + "learning_rate": 5.950195451034757e-05, + "loss": 0.4439, + "step": 37108 + }, + { + "epoch": 2.5750468392200405, + "grad_norm": 1.1324400901794434, + "learning_rate": 5.949681733176301e-05, + "loss": 0.4647, + "step": 37109 + }, + { + "epoch": 2.575116230657137, + "grad_norm": 1.5587314367294312, + "learning_rate": 5.94916802810423e-05, + "loss": 0.4953, + "step": 37110 + }, + { + "epoch": 2.5751856220942333, + "grad_norm": 1.0583926439285278, + "learning_rate": 5.9486543358201666e-05, + "loss": 0.404, + "step": 37111 + }, + { + "epoch": 2.57525501353133, + "grad_norm": 1.4934957027435303, + "learning_rate": 5.948140656325734e-05, + "loss": 0.4114, + "step": 37112 + }, + { + "epoch": 2.575324404968427, + "grad_norm": 1.1745766401290894, + "learning_rate": 5.947626989622548e-05, + "loss": 0.4098, + "step": 37113 + }, + { + "epoch": 2.575393796405524, + "grad_norm": 1.7808011770248413, + "learning_rate": 5.9471133357122336e-05, + "loss": 0.5192, + "step": 37114 + }, + { + "epoch": 2.57546318784262, + "grad_norm": 1.246416449546814, + "learning_rate": 5.946599694596411e-05, + "loss": 0.3336, + "step": 37115 + }, + { + "epoch": 2.575532579279717, + "grad_norm": 1.834173321723938, + "learning_rate": 5.946086066276706e-05, + "loss": 0.4551, + "step": 37116 + }, + { + "epoch": 2.5756019707168134, + "grad_norm": 1.0843029022216797, + "learning_rate": 5.945572450754735e-05, + "loss": 0.3294, + "step": 37117 + }, + { + "epoch": 2.5756713621539102, + "grad_norm": 1.3831431865692139, + "learning_rate": 5.945058848032121e-05, + "loss": 0.3973, + "step": 37118 + }, + { + "epoch": 2.575740753591007, + "grad_norm": 1.0783047676086426, + "learning_rate": 5.9445452581104874e-05, + "loss": 0.3613, + "step": 37119 + }, + { + "epoch": 2.5758101450281035, + "grad_norm": 1.1570700407028198, + "learning_rate": 5.9440316809914545e-05, + "loss": 0.3259, + "step": 37120 + }, + { + "epoch": 2.5758795364652003, + "grad_norm": 1.0716620683670044, + "learning_rate": 5.9435181166766406e-05, + "loss": 0.3822, + "step": 37121 + }, + { + "epoch": 2.5759489279022967, + "grad_norm": 1.0802366733551025, + "learning_rate": 5.943004565167669e-05, + "loss": 0.2491, + "step": 37122 + }, + { + "epoch": 2.5760183193393935, + "grad_norm": 1.1381075382232666, + "learning_rate": 5.942491026466161e-05, + "loss": 0.3237, + "step": 37123 + }, + { + "epoch": 2.5760877107764903, + "grad_norm": 1.1543904542922974, + "learning_rate": 5.9419775005737396e-05, + "loss": 0.3795, + "step": 37124 + }, + { + "epoch": 2.5761571022135867, + "grad_norm": 1.0656934976577759, + "learning_rate": 5.941463987492024e-05, + "loss": 0.3277, + "step": 37125 + }, + { + "epoch": 2.5762264936506836, + "grad_norm": 1.2533611059188843, + "learning_rate": 5.9409504872226396e-05, + "loss": 0.4316, + "step": 37126 + }, + { + "epoch": 2.57629588508778, + "grad_norm": 0.9628336429595947, + "learning_rate": 5.9404369997671996e-05, + "loss": 0.3998, + "step": 37127 + }, + { + "epoch": 2.576365276524877, + "grad_norm": 1.0651086568832397, + "learning_rate": 5.939923525127329e-05, + "loss": 0.3643, + "step": 37128 + }, + { + "epoch": 2.5764346679619736, + "grad_norm": 0.9115355014801025, + "learning_rate": 5.939410063304648e-05, + "loss": 0.2274, + "step": 37129 + }, + { + "epoch": 2.57650405939907, + "grad_norm": 1.2631200551986694, + "learning_rate": 5.9388966143007795e-05, + "loss": 0.6679, + "step": 37130 + }, + { + "epoch": 2.576573450836167, + "grad_norm": 1.1306986808776855, + "learning_rate": 5.9383831781173415e-05, + "loss": 0.3323, + "step": 37131 + }, + { + "epoch": 2.5766428422732632, + "grad_norm": 1.488388180732727, + "learning_rate": 5.93786975475596e-05, + "loss": 0.366, + "step": 37132 + }, + { + "epoch": 2.57671223371036, + "grad_norm": 1.5129814147949219, + "learning_rate": 5.937356344218252e-05, + "loss": 0.3582, + "step": 37133 + }, + { + "epoch": 2.576781625147457, + "grad_norm": 0.9505618214607239, + "learning_rate": 5.936842946505836e-05, + "loss": 0.3454, + "step": 37134 + }, + { + "epoch": 2.5768510165845537, + "grad_norm": 1.254008412361145, + "learning_rate": 5.93632956162034e-05, + "loss": 0.2934, + "step": 37135 + }, + { + "epoch": 2.57692040802165, + "grad_norm": 1.0606224536895752, + "learning_rate": 5.935816189563374e-05, + "loss": 0.5168, + "step": 37136 + }, + { + "epoch": 2.576989799458747, + "grad_norm": 1.0863006114959717, + "learning_rate": 5.9353028303365686e-05, + "loss": 0.2964, + "step": 37137 + }, + { + "epoch": 2.5770591908958433, + "grad_norm": 1.3727397918701172, + "learning_rate": 5.934789483941539e-05, + "loss": 0.3782, + "step": 37138 + }, + { + "epoch": 2.57712858233294, + "grad_norm": 1.4317936897277832, + "learning_rate": 5.934276150379907e-05, + "loss": 0.4925, + "step": 37139 + }, + { + "epoch": 2.577197973770037, + "grad_norm": 1.5096732378005981, + "learning_rate": 5.9337628296532976e-05, + "loss": 0.6257, + "step": 37140 + }, + { + "epoch": 2.5772673652071334, + "grad_norm": 1.0186588764190674, + "learning_rate": 5.933249521763323e-05, + "loss": 0.3059, + "step": 37141 + }, + { + "epoch": 2.5773367566442302, + "grad_norm": 1.4220736026763916, + "learning_rate": 5.932736226711608e-05, + "loss": 0.4532, + "step": 37142 + }, + { + "epoch": 2.5774061480813266, + "grad_norm": 1.3456523418426514, + "learning_rate": 5.9322229444997736e-05, + "loss": 0.4568, + "step": 37143 + }, + { + "epoch": 2.5774755395184235, + "grad_norm": 1.2878559827804565, + "learning_rate": 5.931709675129439e-05, + "loss": 0.4711, + "step": 37144 + }, + { + "epoch": 2.5775449309555203, + "grad_norm": 2.5646812915802, + "learning_rate": 5.931196418602225e-05, + "loss": 0.2587, + "step": 37145 + }, + { + "epoch": 2.5776143223926167, + "grad_norm": 1.242414951324463, + "learning_rate": 5.930683174919751e-05, + "loss": 0.3965, + "step": 37146 + }, + { + "epoch": 2.5776837138297135, + "grad_norm": 1.1820342540740967, + "learning_rate": 5.930169944083643e-05, + "loss": 0.3925, + "step": 37147 + }, + { + "epoch": 2.57775310526681, + "grad_norm": 1.2904021739959717, + "learning_rate": 5.929656726095512e-05, + "loss": 0.5088, + "step": 37148 + }, + { + "epoch": 2.5778224967039067, + "grad_norm": 1.4447866678237915, + "learning_rate": 5.9291435209569824e-05, + "loss": 0.5819, + "step": 37149 + }, + { + "epoch": 2.5778918881410036, + "grad_norm": 1.6649198532104492, + "learning_rate": 5.928630328669674e-05, + "loss": 0.3002, + "step": 37150 + }, + { + "epoch": 2.5779612795781, + "grad_norm": 1.4068635702133179, + "learning_rate": 5.928117149235211e-05, + "loss": 0.4998, + "step": 37151 + }, + { + "epoch": 2.578030671015197, + "grad_norm": 1.4874316453933716, + "learning_rate": 5.9276039826552056e-05, + "loss": 0.4706, + "step": 37152 + }, + { + "epoch": 2.578100062452293, + "grad_norm": 1.0034072399139404, + "learning_rate": 5.9270908289312835e-05, + "loss": 0.4256, + "step": 37153 + }, + { + "epoch": 2.57816945388939, + "grad_norm": 1.1108652353286743, + "learning_rate": 5.926577688065066e-05, + "loss": 0.3791, + "step": 37154 + }, + { + "epoch": 2.578238845326487, + "grad_norm": 1.056757926940918, + "learning_rate": 5.926064560058167e-05, + "loss": 0.2903, + "step": 37155 + }, + { + "epoch": 2.5783082367635832, + "grad_norm": 0.9923108816146851, + "learning_rate": 5.9255514449122084e-05, + "loss": 0.4411, + "step": 37156 + }, + { + "epoch": 2.57837762820068, + "grad_norm": 1.44039785861969, + "learning_rate": 5.9250383426288126e-05, + "loss": 0.4994, + "step": 37157 + }, + { + "epoch": 2.5784470196377764, + "grad_norm": 1.135671615600586, + "learning_rate": 5.924525253209596e-05, + "loss": 0.3729, + "step": 37158 + }, + { + "epoch": 2.5785164110748733, + "grad_norm": 1.1492412090301514, + "learning_rate": 5.9240121766561815e-05, + "loss": 0.3152, + "step": 37159 + }, + { + "epoch": 2.57858580251197, + "grad_norm": 1.6593916416168213, + "learning_rate": 5.923499112970192e-05, + "loss": 0.3476, + "step": 37160 + }, + { + "epoch": 2.578655193949067, + "grad_norm": 1.3958320617675781, + "learning_rate": 5.92298606215324e-05, + "loss": 0.4871, + "step": 37161 + }, + { + "epoch": 2.5787245853861633, + "grad_norm": 1.1341661214828491, + "learning_rate": 5.922473024206945e-05, + "loss": 0.2881, + "step": 37162 + }, + { + "epoch": 2.57879397682326, + "grad_norm": 1.2206230163574219, + "learning_rate": 5.921959999132931e-05, + "loss": 0.3068, + "step": 37163 + }, + { + "epoch": 2.5788633682603566, + "grad_norm": 1.3417837619781494, + "learning_rate": 5.921446986932817e-05, + "loss": 0.3594, + "step": 37164 + }, + { + "epoch": 2.5789327596974534, + "grad_norm": 1.2733029127120972, + "learning_rate": 5.9209339876082216e-05, + "loss": 0.4273, + "step": 37165 + }, + { + "epoch": 2.5790021511345502, + "grad_norm": 2.1244523525238037, + "learning_rate": 5.920421001160765e-05, + "loss": 0.361, + "step": 37166 + }, + { + "epoch": 2.5790715425716466, + "grad_norm": 3.320547580718994, + "learning_rate": 5.919908027592067e-05, + "loss": 0.2194, + "step": 37167 + }, + { + "epoch": 2.5791409340087434, + "grad_norm": 1.461254596710205, + "learning_rate": 5.9193950669037436e-05, + "loss": 0.39, + "step": 37168 + }, + { + "epoch": 2.57921032544584, + "grad_norm": 1.0886918306350708, + "learning_rate": 5.918882119097415e-05, + "loss": 0.3902, + "step": 37169 + }, + { + "epoch": 2.5792797168829367, + "grad_norm": 1.0887160301208496, + "learning_rate": 5.918369184174707e-05, + "loss": 0.3267, + "step": 37170 + }, + { + "epoch": 2.5793491083200335, + "grad_norm": 1.0286413431167603, + "learning_rate": 5.91785626213723e-05, + "loss": 0.3267, + "step": 37171 + }, + { + "epoch": 2.57941849975713, + "grad_norm": 0.8157746195793152, + "learning_rate": 5.9173433529866084e-05, + "loss": 0.2417, + "step": 37172 + }, + { + "epoch": 2.5794878911942267, + "grad_norm": 0.8693293929100037, + "learning_rate": 5.916830456724458e-05, + "loss": 0.2692, + "step": 37173 + }, + { + "epoch": 2.579557282631323, + "grad_norm": 1.5912119150161743, + "learning_rate": 5.916317573352407e-05, + "loss": 0.5688, + "step": 37174 + }, + { + "epoch": 2.57962667406842, + "grad_norm": 1.2468770742416382, + "learning_rate": 5.915804702872062e-05, + "loss": 0.3636, + "step": 37175 + }, + { + "epoch": 2.5796960655055168, + "grad_norm": 1.515460729598999, + "learning_rate": 5.915291845285047e-05, + "loss": 0.4552, + "step": 37176 + }, + { + "epoch": 2.579765456942613, + "grad_norm": 1.08902907371521, + "learning_rate": 5.914779000592983e-05, + "loss": 0.2698, + "step": 37177 + }, + { + "epoch": 2.57983484837971, + "grad_norm": 1.5225673913955688, + "learning_rate": 5.9142661687974877e-05, + "loss": 0.3621, + "step": 37178 + }, + { + "epoch": 2.5799042398168064, + "grad_norm": 1.1158983707427979, + "learning_rate": 5.91375334990018e-05, + "loss": 0.4896, + "step": 37179 + }, + { + "epoch": 2.579973631253903, + "grad_norm": 1.0887835025787354, + "learning_rate": 5.913240543902678e-05, + "loss": 0.3287, + "step": 37180 + }, + { + "epoch": 2.580043022691, + "grad_norm": 0.8815441131591797, + "learning_rate": 5.912727750806606e-05, + "loss": 0.3434, + "step": 37181 + }, + { + "epoch": 2.5801124141280964, + "grad_norm": 1.1485977172851562, + "learning_rate": 5.912214970613574e-05, + "loss": 0.4356, + "step": 37182 + }, + { + "epoch": 2.5801818055651933, + "grad_norm": 0.7786910533905029, + "learning_rate": 5.911702203325207e-05, + "loss": 0.4426, + "step": 37183 + }, + { + "epoch": 2.5802511970022897, + "grad_norm": 1.4679173231124878, + "learning_rate": 5.91118944894312e-05, + "loss": 0.4845, + "step": 37184 + }, + { + "epoch": 2.5803205884393865, + "grad_norm": 1.1069860458374023, + "learning_rate": 5.910676707468933e-05, + "loss": 0.3818, + "step": 37185 + }, + { + "epoch": 2.5803899798764833, + "grad_norm": 0.964501142501831, + "learning_rate": 5.910163978904268e-05, + "loss": 0.5158, + "step": 37186 + }, + { + "epoch": 2.58045937131358, + "grad_norm": 1.4817858934402466, + "learning_rate": 5.90965126325074e-05, + "loss": 0.3377, + "step": 37187 + }, + { + "epoch": 2.5805287627506766, + "grad_norm": 0.8133713603019714, + "learning_rate": 5.9091385605099684e-05, + "loss": 0.316, + "step": 37188 + }, + { + "epoch": 2.5805981541877734, + "grad_norm": 1.2788251638412476, + "learning_rate": 5.9086258706835715e-05, + "loss": 0.3856, + "step": 37189 + }, + { + "epoch": 2.5806675456248698, + "grad_norm": 1.518323540687561, + "learning_rate": 5.908113193773168e-05, + "loss": 0.33, + "step": 37190 + }, + { + "epoch": 2.5807369370619666, + "grad_norm": 1.561537742614746, + "learning_rate": 5.907600529780375e-05, + "loss": 0.5641, + "step": 37191 + }, + { + "epoch": 2.5808063284990634, + "grad_norm": 1.237463355064392, + "learning_rate": 5.907087878706812e-05, + "loss": 0.3511, + "step": 37192 + }, + { + "epoch": 2.58087571993616, + "grad_norm": 1.1257785558700562, + "learning_rate": 5.9065752405541e-05, + "loss": 0.5694, + "step": 37193 + }, + { + "epoch": 2.5809451113732567, + "grad_norm": 1.1465147733688354, + "learning_rate": 5.906062615323853e-05, + "loss": 0.3024, + "step": 37194 + }, + { + "epoch": 2.581014502810353, + "grad_norm": 1.4944779872894287, + "learning_rate": 5.905550003017695e-05, + "loss": 0.4052, + "step": 37195 + }, + { + "epoch": 2.58108389424745, + "grad_norm": 1.524232268333435, + "learning_rate": 5.905037403637237e-05, + "loss": 0.5229, + "step": 37196 + }, + { + "epoch": 2.5811532856845467, + "grad_norm": 2.4118878841400146, + "learning_rate": 5.9045248171841016e-05, + "loss": 0.3413, + "step": 37197 + }, + { + "epoch": 2.581222677121643, + "grad_norm": 1.0276449918746948, + "learning_rate": 5.9040122436599064e-05, + "loss": 0.4232, + "step": 37198 + }, + { + "epoch": 2.58129206855874, + "grad_norm": 1.2652959823608398, + "learning_rate": 5.903499683066268e-05, + "loss": 0.3176, + "step": 37199 + }, + { + "epoch": 2.5813614599958363, + "grad_norm": 1.2504632472991943, + "learning_rate": 5.902987135404807e-05, + "loss": 0.3811, + "step": 37200 + }, + { + "epoch": 2.581430851432933, + "grad_norm": 1.2088420391082764, + "learning_rate": 5.9024746006771415e-05, + "loss": 0.3372, + "step": 37201 + }, + { + "epoch": 2.58150024287003, + "grad_norm": 1.1640722751617432, + "learning_rate": 5.9019620788848887e-05, + "loss": 0.3987, + "step": 37202 + }, + { + "epoch": 2.5815696343071264, + "grad_norm": 1.2013988494873047, + "learning_rate": 5.901449570029665e-05, + "loss": 0.2837, + "step": 37203 + }, + { + "epoch": 2.581639025744223, + "grad_norm": 1.1105482578277588, + "learning_rate": 5.900937074113087e-05, + "loss": 0.4067, + "step": 37204 + }, + { + "epoch": 2.5817084171813196, + "grad_norm": 1.2222591638565063, + "learning_rate": 5.900424591136779e-05, + "loss": 0.4145, + "step": 37205 + }, + { + "epoch": 2.5817778086184164, + "grad_norm": 0.877221405506134, + "learning_rate": 5.8999121211023534e-05, + "loss": 0.2723, + "step": 37206 + }, + { + "epoch": 2.5818472000555133, + "grad_norm": 1.1818243265151978, + "learning_rate": 5.899399664011429e-05, + "loss": 0.4032, + "step": 37207 + }, + { + "epoch": 2.5819165914926097, + "grad_norm": 1.3016480207443237, + "learning_rate": 5.898887219865623e-05, + "loss": 0.4925, + "step": 37208 + }, + { + "epoch": 2.5819859829297065, + "grad_norm": 1.6880618333816528, + "learning_rate": 5.898374788666558e-05, + "loss": 0.3826, + "step": 37209 + }, + { + "epoch": 2.582055374366803, + "grad_norm": 0.9463716149330139, + "learning_rate": 5.897862370415845e-05, + "loss": 0.2562, + "step": 37210 + }, + { + "epoch": 2.5821247658038997, + "grad_norm": 2.0315325260162354, + "learning_rate": 5.897349965115105e-05, + "loss": 0.5671, + "step": 37211 + }, + { + "epoch": 2.5821941572409965, + "grad_norm": 1.188133955001831, + "learning_rate": 5.896837572765955e-05, + "loss": 0.3846, + "step": 37212 + }, + { + "epoch": 2.5822635486780934, + "grad_norm": 1.0588213205337524, + "learning_rate": 5.896325193370012e-05, + "loss": 0.2804, + "step": 37213 + }, + { + "epoch": 2.5823329401151898, + "grad_norm": 1.3448760509490967, + "learning_rate": 5.895812826928895e-05, + "loss": 0.4571, + "step": 37214 + }, + { + "epoch": 2.5824023315522866, + "grad_norm": 1.450470209121704, + "learning_rate": 5.8953004734442205e-05, + "loss": 0.3838, + "step": 37215 + }, + { + "epoch": 2.582471722989383, + "grad_norm": 1.3383514881134033, + "learning_rate": 5.8947881329176105e-05, + "loss": 0.4706, + "step": 37216 + }, + { + "epoch": 2.58254111442648, + "grad_norm": 1.2143056392669678, + "learning_rate": 5.894275805350673e-05, + "loss": 0.4341, + "step": 37217 + }, + { + "epoch": 2.5826105058635767, + "grad_norm": 2.6768295764923096, + "learning_rate": 5.893763490745031e-05, + "loss": 0.442, + "step": 37218 + }, + { + "epoch": 2.582679897300673, + "grad_norm": 0.8482765555381775, + "learning_rate": 5.893251189102302e-05, + "loss": 0.3314, + "step": 37219 + }, + { + "epoch": 2.58274928873777, + "grad_norm": 1.2668706178665161, + "learning_rate": 5.892738900424104e-05, + "loss": 0.5055, + "step": 37220 + }, + { + "epoch": 2.5828186801748663, + "grad_norm": 1.059888243675232, + "learning_rate": 5.89222662471205e-05, + "loss": 0.3374, + "step": 37221 + }, + { + "epoch": 2.582888071611963, + "grad_norm": 1.4303052425384521, + "learning_rate": 5.8917143619677605e-05, + "loss": 0.4403, + "step": 37222 + }, + { + "epoch": 2.58295746304906, + "grad_norm": 1.655226469039917, + "learning_rate": 5.891202112192854e-05, + "loss": 0.4294, + "step": 37223 + }, + { + "epoch": 2.5830268544861563, + "grad_norm": 1.0640044212341309, + "learning_rate": 5.890689875388946e-05, + "loss": 0.4034, + "step": 37224 + }, + { + "epoch": 2.583096245923253, + "grad_norm": 1.2253717184066772, + "learning_rate": 5.8901776515576514e-05, + "loss": 0.485, + "step": 37225 + }, + { + "epoch": 2.5831656373603495, + "grad_norm": 1.1353000402450562, + "learning_rate": 5.889665440700589e-05, + "loss": 0.3846, + "step": 37226 + }, + { + "epoch": 2.5832350287974464, + "grad_norm": 1.5971319675445557, + "learning_rate": 5.889153242819376e-05, + "loss": 0.4635, + "step": 37227 + }, + { + "epoch": 2.583304420234543, + "grad_norm": 1.1410362720489502, + "learning_rate": 5.888641057915629e-05, + "loss": 0.4478, + "step": 37228 + }, + { + "epoch": 2.5833738116716396, + "grad_norm": 0.7116817831993103, + "learning_rate": 5.8881288859909646e-05, + "loss": 0.207, + "step": 37229 + }, + { + "epoch": 2.5834432031087364, + "grad_norm": 1.9241827726364136, + "learning_rate": 5.887616727047005e-05, + "loss": 0.6559, + "step": 37230 + }, + { + "epoch": 2.583512594545833, + "grad_norm": 1.4909080266952515, + "learning_rate": 5.8871045810853584e-05, + "loss": 0.3973, + "step": 37231 + }, + { + "epoch": 2.5835819859829297, + "grad_norm": 0.7869929075241089, + "learning_rate": 5.886592448107645e-05, + "loss": 0.3379, + "step": 37232 + }, + { + "epoch": 2.5836513774200265, + "grad_norm": 0.9979549050331116, + "learning_rate": 5.886080328115482e-05, + "loss": 0.3459, + "step": 37233 + }, + { + "epoch": 2.583720768857123, + "grad_norm": 1.1683839559555054, + "learning_rate": 5.8855682211104866e-05, + "loss": 0.4618, + "step": 37234 + }, + { + "epoch": 2.5837901602942197, + "grad_norm": 1.3487699031829834, + "learning_rate": 5.885056127094275e-05, + "loss": 0.3077, + "step": 37235 + }, + { + "epoch": 2.583859551731316, + "grad_norm": 1.1508183479309082, + "learning_rate": 5.884544046068467e-05, + "loss": 0.3706, + "step": 37236 + }, + { + "epoch": 2.583928943168413, + "grad_norm": 1.208940029144287, + "learning_rate": 5.884031978034672e-05, + "loss": 0.4684, + "step": 37237 + }, + { + "epoch": 2.5839983346055098, + "grad_norm": 1.5647801160812378, + "learning_rate": 5.883519922994512e-05, + "loss": 0.4179, + "step": 37238 + }, + { + "epoch": 2.5840677260426066, + "grad_norm": 1.5068960189819336, + "learning_rate": 5.8830078809496e-05, + "loss": 0.4303, + "step": 37239 + }, + { + "epoch": 2.584137117479703, + "grad_norm": 1.1659694910049438, + "learning_rate": 5.8824958519015574e-05, + "loss": 0.349, + "step": 37240 + }, + { + "epoch": 2.5842065089168, + "grad_norm": 1.211834192276001, + "learning_rate": 5.881983835851996e-05, + "loss": 0.3545, + "step": 37241 + }, + { + "epoch": 2.584275900353896, + "grad_norm": 1.0013256072998047, + "learning_rate": 5.881471832802534e-05, + "loss": 0.3152, + "step": 37242 + }, + { + "epoch": 2.584345291790993, + "grad_norm": 0.9357336759567261, + "learning_rate": 5.8809598427547905e-05, + "loss": 0.3456, + "step": 37243 + }, + { + "epoch": 2.58441468322809, + "grad_norm": 1.0161439180374146, + "learning_rate": 5.8804478657103755e-05, + "loss": 0.1785, + "step": 37244 + }, + { + "epoch": 2.5844840746651863, + "grad_norm": 1.1170345544815063, + "learning_rate": 5.8799359016709086e-05, + "loss": 0.4022, + "step": 37245 + }, + { + "epoch": 2.584553466102283, + "grad_norm": 2.0501723289489746, + "learning_rate": 5.8794239506380054e-05, + "loss": 0.4847, + "step": 37246 + }, + { + "epoch": 2.5846228575393795, + "grad_norm": 1.4120864868164062, + "learning_rate": 5.8789120126132846e-05, + "loss": 0.4229, + "step": 37247 + }, + { + "epoch": 2.5846922489764763, + "grad_norm": 1.599907398223877, + "learning_rate": 5.878400087598359e-05, + "loss": 0.5207, + "step": 37248 + }, + { + "epoch": 2.584761640413573, + "grad_norm": 0.9668037295341492, + "learning_rate": 5.8778881755948454e-05, + "loss": 0.3265, + "step": 37249 + }, + { + "epoch": 2.5848310318506695, + "grad_norm": 1.388304591178894, + "learning_rate": 5.8773762766043674e-05, + "loss": 0.474, + "step": 37250 + }, + { + "epoch": 2.5849004232877664, + "grad_norm": 1.3449299335479736, + "learning_rate": 5.876864390628527e-05, + "loss": 0.3511, + "step": 37251 + }, + { + "epoch": 2.5849698147248628, + "grad_norm": 0.9802449345588684, + "learning_rate": 5.876352517668948e-05, + "loss": 0.2441, + "step": 37252 + }, + { + "epoch": 2.5850392061619596, + "grad_norm": 1.599531888961792, + "learning_rate": 5.8758406577272475e-05, + "loss": 0.4022, + "step": 37253 + }, + { + "epoch": 2.5851085975990564, + "grad_norm": 1.3090680837631226, + "learning_rate": 5.875328810805039e-05, + "loss": 0.3189, + "step": 37254 + }, + { + "epoch": 2.585177989036153, + "grad_norm": 1.2625207901000977, + "learning_rate": 5.8748169769039396e-05, + "loss": 0.3743, + "step": 37255 + }, + { + "epoch": 2.5852473804732496, + "grad_norm": 1.058165192604065, + "learning_rate": 5.8743051560255634e-05, + "loss": 0.3615, + "step": 37256 + }, + { + "epoch": 2.585316771910346, + "grad_norm": 1.3734875917434692, + "learning_rate": 5.8737933481715276e-05, + "loss": 0.383, + "step": 37257 + }, + { + "epoch": 2.585386163347443, + "grad_norm": 1.0521761178970337, + "learning_rate": 5.873281553343447e-05, + "loss": 0.327, + "step": 37258 + }, + { + "epoch": 2.5854555547845397, + "grad_norm": 1.054670810699463, + "learning_rate": 5.87276977154294e-05, + "loss": 0.3425, + "step": 37259 + }, + { + "epoch": 2.585524946221636, + "grad_norm": 1.036130666732788, + "learning_rate": 5.8722580027716176e-05, + "loss": 0.3602, + "step": 37260 + }, + { + "epoch": 2.585594337658733, + "grad_norm": 1.3245913982391357, + "learning_rate": 5.871746247031097e-05, + "loss": 0.3775, + "step": 37261 + }, + { + "epoch": 2.5856637290958293, + "grad_norm": 1.4011051654815674, + "learning_rate": 5.871234504322995e-05, + "loss": 0.5486, + "step": 37262 + }, + { + "epoch": 2.585733120532926, + "grad_norm": 1.0765430927276611, + "learning_rate": 5.870722774648927e-05, + "loss": 0.3746, + "step": 37263 + }, + { + "epoch": 2.585802511970023, + "grad_norm": 1.9216448068618774, + "learning_rate": 5.870211058010512e-05, + "loss": 0.5308, + "step": 37264 + }, + { + "epoch": 2.58587190340712, + "grad_norm": 1.1483045816421509, + "learning_rate": 5.869699354409357e-05, + "loss": 0.2882, + "step": 37265 + }, + { + "epoch": 2.585941294844216, + "grad_norm": 1.4029932022094727, + "learning_rate": 5.8691876638470824e-05, + "loss": 0.4817, + "step": 37266 + }, + { + "epoch": 2.586010686281313, + "grad_norm": 1.1542853116989136, + "learning_rate": 5.868675986325303e-05, + "loss": 0.4334, + "step": 37267 + }, + { + "epoch": 2.5860800777184094, + "grad_norm": 1.2162322998046875, + "learning_rate": 5.868164321845635e-05, + "loss": 0.4199, + "step": 37268 + }, + { + "epoch": 2.5861494691555063, + "grad_norm": 1.5115410089492798, + "learning_rate": 5.8676526704096914e-05, + "loss": 0.5299, + "step": 37269 + }, + { + "epoch": 2.586218860592603, + "grad_norm": 1.473922610282898, + "learning_rate": 5.86714103201909e-05, + "loss": 0.5049, + "step": 37270 + }, + { + "epoch": 2.5862882520296995, + "grad_norm": 1.3254752159118652, + "learning_rate": 5.866629406675448e-05, + "loss": 0.4724, + "step": 37271 + }, + { + "epoch": 2.5863576434667963, + "grad_norm": 0.9390634894371033, + "learning_rate": 5.866117794380375e-05, + "loss": 0.287, + "step": 37272 + }, + { + "epoch": 2.5864270349038927, + "grad_norm": 1.567824125289917, + "learning_rate": 5.865606195135487e-05, + "loss": 0.5301, + "step": 37273 + }, + { + "epoch": 2.5864964263409895, + "grad_norm": 1.508719801902771, + "learning_rate": 5.865094608942401e-05, + "loss": 0.5066, + "step": 37274 + }, + { + "epoch": 2.5865658177780864, + "grad_norm": 1.016008734703064, + "learning_rate": 5.864583035802733e-05, + "loss": 0.3548, + "step": 37275 + }, + { + "epoch": 2.5866352092151828, + "grad_norm": 1.0181339979171753, + "learning_rate": 5.8640714757180947e-05, + "loss": 0.3078, + "step": 37276 + }, + { + "epoch": 2.5867046006522796, + "grad_norm": 1.2826497554779053, + "learning_rate": 5.863559928690103e-05, + "loss": 0.473, + "step": 37277 + }, + { + "epoch": 2.586773992089376, + "grad_norm": 1.4487907886505127, + "learning_rate": 5.8630483947203776e-05, + "loss": 0.4533, + "step": 37278 + }, + { + "epoch": 2.586843383526473, + "grad_norm": 1.8734341859817505, + "learning_rate": 5.862536873810524e-05, + "loss": 0.4157, + "step": 37279 + }, + { + "epoch": 2.5869127749635696, + "grad_norm": 1.1126552820205688, + "learning_rate": 5.8620253659621605e-05, + "loss": 0.2995, + "step": 37280 + }, + { + "epoch": 2.586982166400666, + "grad_norm": 0.9120679497718811, + "learning_rate": 5.8615138711769034e-05, + "loss": 0.4129, + "step": 37281 + }, + { + "epoch": 2.587051557837763, + "grad_norm": 1.1641343832015991, + "learning_rate": 5.861002389456367e-05, + "loss": 0.3696, + "step": 37282 + }, + { + "epoch": 2.5871209492748592, + "grad_norm": 1.1707204580307007, + "learning_rate": 5.8604909208021666e-05, + "loss": 0.4515, + "step": 37283 + }, + { + "epoch": 2.587190340711956, + "grad_norm": 1.4407349824905396, + "learning_rate": 5.8599794652159145e-05, + "loss": 0.6839, + "step": 37284 + }, + { + "epoch": 2.587259732149053, + "grad_norm": 0.8982254862785339, + "learning_rate": 5.859468022699233e-05, + "loss": 0.2639, + "step": 37285 + }, + { + "epoch": 2.5873291235861493, + "grad_norm": 1.2420872449874878, + "learning_rate": 5.858956593253724e-05, + "loss": 0.4727, + "step": 37286 + }, + { + "epoch": 2.587398515023246, + "grad_norm": 1.4675627946853638, + "learning_rate": 5.858445176881011e-05, + "loss": 0.4023, + "step": 37287 + }, + { + "epoch": 2.5874679064603425, + "grad_norm": 1.5931618213653564, + "learning_rate": 5.8579337735827047e-05, + "loss": 0.5724, + "step": 37288 + }, + { + "epoch": 2.5875372978974394, + "grad_norm": 0.9217814803123474, + "learning_rate": 5.8574223833604204e-05, + "loss": 0.3047, + "step": 37289 + }, + { + "epoch": 2.587606689334536, + "grad_norm": 1.2981278896331787, + "learning_rate": 5.856911006215776e-05, + "loss": 0.4001, + "step": 37290 + }, + { + "epoch": 2.587676080771633, + "grad_norm": 1.6344541311264038, + "learning_rate": 5.856399642150381e-05, + "loss": 0.5127, + "step": 37291 + }, + { + "epoch": 2.5877454722087294, + "grad_norm": 1.1366115808486938, + "learning_rate": 5.855888291165853e-05, + "loss": 0.3661, + "step": 37292 + }, + { + "epoch": 2.5878148636458262, + "grad_norm": 0.8008020520210266, + "learning_rate": 5.8553769532638024e-05, + "loss": 0.4316, + "step": 37293 + }, + { + "epoch": 2.5878842550829226, + "grad_norm": 1.4801825284957886, + "learning_rate": 5.854865628445849e-05, + "loss": 0.4526, + "step": 37294 + }, + { + "epoch": 2.5879536465200195, + "grad_norm": 1.1961324214935303, + "learning_rate": 5.8543543167136016e-05, + "loss": 0.358, + "step": 37295 + }, + { + "epoch": 2.5880230379571163, + "grad_norm": 1.5044289827346802, + "learning_rate": 5.853843018068677e-05, + "loss": 0.4314, + "step": 37296 + }, + { + "epoch": 2.5880924293942127, + "grad_norm": 1.5102311372756958, + "learning_rate": 5.853331732512688e-05, + "loss": 0.3714, + "step": 37297 + }, + { + "epoch": 2.5881618208313095, + "grad_norm": 1.7558259963989258, + "learning_rate": 5.85282046004725e-05, + "loss": 0.5956, + "step": 37298 + }, + { + "epoch": 2.588231212268406, + "grad_norm": 0.7279237508773804, + "learning_rate": 5.852309200673981e-05, + "loss": 0.2457, + "step": 37299 + }, + { + "epoch": 2.5883006037055027, + "grad_norm": 1.3391672372817993, + "learning_rate": 5.851797954394487e-05, + "loss": 0.355, + "step": 37300 + }, + { + "epoch": 2.5883699951425996, + "grad_norm": 0.9295980334281921, + "learning_rate": 5.8512867212103836e-05, + "loss": 0.3473, + "step": 37301 + }, + { + "epoch": 2.588439386579696, + "grad_norm": 1.5480722188949585, + "learning_rate": 5.850775501123289e-05, + "loss": 0.5182, + "step": 37302 + }, + { + "epoch": 2.588508778016793, + "grad_norm": 1.2754639387130737, + "learning_rate": 5.850264294134813e-05, + "loss": 0.6026, + "step": 37303 + }, + { + "epoch": 2.588578169453889, + "grad_norm": 1.4044140577316284, + "learning_rate": 5.8497531002465725e-05, + "loss": 0.4791, + "step": 37304 + }, + { + "epoch": 2.588647560890986, + "grad_norm": 1.3687255382537842, + "learning_rate": 5.849241919460179e-05, + "loss": 0.3621, + "step": 37305 + }, + { + "epoch": 2.588716952328083, + "grad_norm": 1.5731558799743652, + "learning_rate": 5.8487307517772515e-05, + "loss": 0.3829, + "step": 37306 + }, + { + "epoch": 2.5887863437651792, + "grad_norm": 1.0817744731903076, + "learning_rate": 5.8482195971993956e-05, + "loss": 0.2652, + "step": 37307 + }, + { + "epoch": 2.588855735202276, + "grad_norm": 1.3111342191696167, + "learning_rate": 5.8477084557282294e-05, + "loss": 0.5271, + "step": 37308 + }, + { + "epoch": 2.5889251266393725, + "grad_norm": 1.260852336883545, + "learning_rate": 5.8471973273653656e-05, + "loss": 0.468, + "step": 37309 + }, + { + "epoch": 2.5889945180764693, + "grad_norm": 1.2538315057754517, + "learning_rate": 5.846686212112418e-05, + "loss": 0.3984, + "step": 37310 + }, + { + "epoch": 2.589063909513566, + "grad_norm": 1.0984523296356201, + "learning_rate": 5.846175109971e-05, + "loss": 0.3278, + "step": 37311 + }, + { + "epoch": 2.589133300950663, + "grad_norm": 1.2099336385726929, + "learning_rate": 5.845664020942726e-05, + "loss": 0.3979, + "step": 37312 + }, + { + "epoch": 2.5892026923877594, + "grad_norm": 1.3969011306762695, + "learning_rate": 5.84515294502921e-05, + "loss": 0.3841, + "step": 37313 + }, + { + "epoch": 2.5892720838248557, + "grad_norm": 0.9825198650360107, + "learning_rate": 5.844641882232061e-05, + "loss": 0.3197, + "step": 37314 + }, + { + "epoch": 2.5893414752619526, + "grad_norm": 1.1961956024169922, + "learning_rate": 5.8441308325528966e-05, + "loss": 0.3142, + "step": 37315 + }, + { + "epoch": 2.5894108666990494, + "grad_norm": 1.2467435598373413, + "learning_rate": 5.8436197959933284e-05, + "loss": 0.3327, + "step": 37316 + }, + { + "epoch": 2.5894802581361462, + "grad_norm": 1.3855221271514893, + "learning_rate": 5.843108772554971e-05, + "loss": 0.4747, + "step": 37317 + }, + { + "epoch": 2.5895496495732426, + "grad_norm": 1.4313446283340454, + "learning_rate": 5.842597762239436e-05, + "loss": 0.4888, + "step": 37318 + }, + { + "epoch": 2.5896190410103395, + "grad_norm": 1.4850364923477173, + "learning_rate": 5.8420867650483415e-05, + "loss": 0.4708, + "step": 37319 + }, + { + "epoch": 2.589688432447436, + "grad_norm": 1.1202588081359863, + "learning_rate": 5.841575780983293e-05, + "loss": 0.4561, + "step": 37320 + }, + { + "epoch": 2.5897578238845327, + "grad_norm": 1.4081768989562988, + "learning_rate": 5.841064810045907e-05, + "loss": 0.3201, + "step": 37321 + }, + { + "epoch": 2.5898272153216295, + "grad_norm": 1.0714170932769775, + "learning_rate": 5.840553852237798e-05, + "loss": 0.286, + "step": 37322 + }, + { + "epoch": 2.589896606758726, + "grad_norm": 0.9488697648048401, + "learning_rate": 5.8400429075605765e-05, + "loss": 0.3697, + "step": 37323 + }, + { + "epoch": 2.5899659981958227, + "grad_norm": 1.2312109470367432, + "learning_rate": 5.839531976015857e-05, + "loss": 0.5773, + "step": 37324 + }, + { + "epoch": 2.590035389632919, + "grad_norm": 1.4509836435317993, + "learning_rate": 5.839021057605255e-05, + "loss": 0.4265, + "step": 37325 + }, + { + "epoch": 2.590104781070016, + "grad_norm": 1.6587995290756226, + "learning_rate": 5.8385101523303806e-05, + "loss": 0.3359, + "step": 37326 + }, + { + "epoch": 2.590174172507113, + "grad_norm": 0.9734371900558472, + "learning_rate": 5.8379992601928444e-05, + "loss": 0.3422, + "step": 37327 + }, + { + "epoch": 2.590243563944209, + "grad_norm": 1.243517518043518, + "learning_rate": 5.8374883811942625e-05, + "loss": 0.5842, + "step": 37328 + }, + { + "epoch": 2.590312955381306, + "grad_norm": 1.6247506141662598, + "learning_rate": 5.836977515336246e-05, + "loss": 0.5222, + "step": 37329 + }, + { + "epoch": 2.5903823468184024, + "grad_norm": 1.4967564344406128, + "learning_rate": 5.836466662620409e-05, + "loss": 0.3846, + "step": 37330 + }, + { + "epoch": 2.5904517382554992, + "grad_norm": 1.4256442785263062, + "learning_rate": 5.8359558230483636e-05, + "loss": 0.6313, + "step": 37331 + }, + { + "epoch": 2.590521129692596, + "grad_norm": 0.8821545839309692, + "learning_rate": 5.835444996621722e-05, + "loss": 0.3231, + "step": 37332 + }, + { + "epoch": 2.5905905211296925, + "grad_norm": 2.1660993099212646, + "learning_rate": 5.8349341833421025e-05, + "loss": 0.2736, + "step": 37333 + }, + { + "epoch": 2.5906599125667893, + "grad_norm": 1.2273508310317993, + "learning_rate": 5.834423383211107e-05, + "loss": 0.4898, + "step": 37334 + }, + { + "epoch": 2.5907293040038857, + "grad_norm": 1.0984114408493042, + "learning_rate": 5.8339125962303554e-05, + "loss": 0.3105, + "step": 37335 + }, + { + "epoch": 2.5907986954409825, + "grad_norm": 3.066646099090576, + "learning_rate": 5.833401822401461e-05, + "loss": 0.4989, + "step": 37336 + }, + { + "epoch": 2.5908680868780793, + "grad_norm": 1.1898707151412964, + "learning_rate": 5.83289106172603e-05, + "loss": 0.3576, + "step": 37337 + }, + { + "epoch": 2.590937478315176, + "grad_norm": 1.1126127243041992, + "learning_rate": 5.832380314205679e-05, + "loss": 0.4704, + "step": 37338 + }, + { + "epoch": 2.5910068697522726, + "grad_norm": 0.9806907773017883, + "learning_rate": 5.831869579842019e-05, + "loss": 0.3584, + "step": 37339 + }, + { + "epoch": 2.591076261189369, + "grad_norm": 1.1199402809143066, + "learning_rate": 5.831358858636667e-05, + "loss": 0.5193, + "step": 37340 + }, + { + "epoch": 2.591145652626466, + "grad_norm": 1.0808701515197754, + "learning_rate": 5.830848150591228e-05, + "loss": 0.4692, + "step": 37341 + }, + { + "epoch": 2.5912150440635626, + "grad_norm": 1.4292833805084229, + "learning_rate": 5.830337455707318e-05, + "loss": 0.51, + "step": 37342 + }, + { + "epoch": 2.5912844355006595, + "grad_norm": 1.2229067087173462, + "learning_rate": 5.8298267739865485e-05, + "loss": 0.4887, + "step": 37343 + }, + { + "epoch": 2.591353826937756, + "grad_norm": 1.1236615180969238, + "learning_rate": 5.8293161054305315e-05, + "loss": 0.4334, + "step": 37344 + }, + { + "epoch": 2.5914232183748527, + "grad_norm": 1.4108614921569824, + "learning_rate": 5.8288054500408806e-05, + "loss": 0.4973, + "step": 37345 + }, + { + "epoch": 2.591492609811949, + "grad_norm": 1.103623390197754, + "learning_rate": 5.828294807819206e-05, + "loss": 0.2423, + "step": 37346 + }, + { + "epoch": 2.591562001249046, + "grad_norm": 1.0913453102111816, + "learning_rate": 5.8277841787671264e-05, + "loss": 0.3996, + "step": 37347 + }, + { + "epoch": 2.5916313926861427, + "grad_norm": 1.0879449844360352, + "learning_rate": 5.827273562886242e-05, + "loss": 0.292, + "step": 37348 + }, + { + "epoch": 2.591700784123239, + "grad_norm": 0.9685550332069397, + "learning_rate": 5.826762960178172e-05, + "loss": 0.4067, + "step": 37349 + }, + { + "epoch": 2.591770175560336, + "grad_norm": 1.1568572521209717, + "learning_rate": 5.8262523706445274e-05, + "loss": 0.4031, + "step": 37350 + }, + { + "epoch": 2.5918395669974323, + "grad_norm": 1.3096766471862793, + "learning_rate": 5.8257417942869184e-05, + "loss": 0.3795, + "step": 37351 + }, + { + "epoch": 2.591908958434529, + "grad_norm": 0.9666821360588074, + "learning_rate": 5.82523123110696e-05, + "loss": 0.2669, + "step": 37352 + }, + { + "epoch": 2.591978349871626, + "grad_norm": 1.1552046537399292, + "learning_rate": 5.824720681106261e-05, + "loss": 0.3506, + "step": 37353 + }, + { + "epoch": 2.5920477413087224, + "grad_norm": 0.9605035781860352, + "learning_rate": 5.8242101442864385e-05, + "loss": 0.3095, + "step": 37354 + }, + { + "epoch": 2.5921171327458192, + "grad_norm": 0.9513471126556396, + "learning_rate": 5.8236996206490966e-05, + "loss": 0.2832, + "step": 37355 + }, + { + "epoch": 2.5921865241829156, + "grad_norm": 1.6718361377716064, + "learning_rate": 5.8231891101958504e-05, + "loss": 0.4652, + "step": 37356 + }, + { + "epoch": 2.5922559156200125, + "grad_norm": 1.1499079465866089, + "learning_rate": 5.822678612928312e-05, + "loss": 0.3646, + "step": 37357 + }, + { + "epoch": 2.5923253070571093, + "grad_norm": 1.10734224319458, + "learning_rate": 5.822168128848091e-05, + "loss": 0.4469, + "step": 37358 + }, + { + "epoch": 2.5923946984942057, + "grad_norm": 1.1414026021957397, + "learning_rate": 5.821657657956802e-05, + "loss": 0.3322, + "step": 37359 + }, + { + "epoch": 2.5924640899313025, + "grad_norm": 1.2630529403686523, + "learning_rate": 5.821147200256054e-05, + "loss": 0.3963, + "step": 37360 + }, + { + "epoch": 2.592533481368399, + "grad_norm": 1.2134939432144165, + "learning_rate": 5.820636755747464e-05, + "loss": 0.6175, + "step": 37361 + }, + { + "epoch": 2.5926028728054957, + "grad_norm": 1.7320822477340698, + "learning_rate": 5.8201263244326356e-05, + "loss": 0.4101, + "step": 37362 + }, + { + "epoch": 2.5926722642425926, + "grad_norm": 0.8461487293243408, + "learning_rate": 5.819615906313182e-05, + "loss": 0.3116, + "step": 37363 + }, + { + "epoch": 2.5927416556796894, + "grad_norm": 1.0413569211959839, + "learning_rate": 5.819105501390716e-05, + "loss": 0.311, + "step": 37364 + }, + { + "epoch": 2.592811047116786, + "grad_norm": 0.9009100198745728, + "learning_rate": 5.818595109666849e-05, + "loss": 0.3246, + "step": 37365 + }, + { + "epoch": 2.5928804385538826, + "grad_norm": 1.4901745319366455, + "learning_rate": 5.818084731143193e-05, + "loss": 0.2311, + "step": 37366 + }, + { + "epoch": 2.592949829990979, + "grad_norm": 1.0384008884429932, + "learning_rate": 5.8175743658213586e-05, + "loss": 0.3611, + "step": 37367 + }, + { + "epoch": 2.593019221428076, + "grad_norm": 1.2196944952011108, + "learning_rate": 5.81706401370296e-05, + "loss": 0.3937, + "step": 37368 + }, + { + "epoch": 2.5930886128651727, + "grad_norm": 1.1383529901504517, + "learning_rate": 5.8165536747896e-05, + "loss": 0.3775, + "step": 37369 + }, + { + "epoch": 2.593158004302269, + "grad_norm": 1.2986027002334595, + "learning_rate": 5.816043349082896e-05, + "loss": 0.4855, + "step": 37370 + }, + { + "epoch": 2.593227395739366, + "grad_norm": 1.0883665084838867, + "learning_rate": 5.815533036584462e-05, + "loss": 0.5817, + "step": 37371 + }, + { + "epoch": 2.5932967871764623, + "grad_norm": 0.9612851142883301, + "learning_rate": 5.8150227372958986e-05, + "loss": 0.3651, + "step": 37372 + }, + { + "epoch": 2.593366178613559, + "grad_norm": 1.339662790298462, + "learning_rate": 5.814512451218825e-05, + "loss": 0.5162, + "step": 37373 + }, + { + "epoch": 2.593435570050656, + "grad_norm": 1.145783543586731, + "learning_rate": 5.814002178354849e-05, + "loss": 0.3273, + "step": 37374 + }, + { + "epoch": 2.5935049614877523, + "grad_norm": 1.0255646705627441, + "learning_rate": 5.813491918705587e-05, + "loss": 0.3194, + "step": 37375 + }, + { + "epoch": 2.593574352924849, + "grad_norm": 1.0001883506774902, + "learning_rate": 5.8129816722726414e-05, + "loss": 0.2989, + "step": 37376 + }, + { + "epoch": 2.5936437443619456, + "grad_norm": 1.1700387001037598, + "learning_rate": 5.8124714390576265e-05, + "loss": 0.4688, + "step": 37377 + }, + { + "epoch": 2.5937131357990424, + "grad_norm": 1.349492073059082, + "learning_rate": 5.8119612190621544e-05, + "loss": 0.3866, + "step": 37378 + }, + { + "epoch": 2.5937825272361392, + "grad_norm": 0.7300019264221191, + "learning_rate": 5.811451012287833e-05, + "loss": 0.2111, + "step": 37379 + }, + { + "epoch": 2.5938519186732356, + "grad_norm": 1.2031184434890747, + "learning_rate": 5.810940818736276e-05, + "loss": 0.4382, + "step": 37380 + }, + { + "epoch": 2.5939213101103324, + "grad_norm": 0.8093898296356201, + "learning_rate": 5.810430638409092e-05, + "loss": 0.2822, + "step": 37381 + }, + { + "epoch": 2.593990701547429, + "grad_norm": 2.6595096588134766, + "learning_rate": 5.809920471307897e-05, + "loss": 0.4653, + "step": 37382 + }, + { + "epoch": 2.5940600929845257, + "grad_norm": 1.186538815498352, + "learning_rate": 5.8094103174342926e-05, + "loss": 0.2955, + "step": 37383 + }, + { + "epoch": 2.5941294844216225, + "grad_norm": 1.4847476482391357, + "learning_rate": 5.808900176789893e-05, + "loss": 0.4156, + "step": 37384 + }, + { + "epoch": 2.594198875858719, + "grad_norm": 2.206636905670166, + "learning_rate": 5.808390049376311e-05, + "loss": 0.4315, + "step": 37385 + }, + { + "epoch": 2.5942682672958157, + "grad_norm": 0.9208358526229858, + "learning_rate": 5.807879935195155e-05, + "loss": 0.2449, + "step": 37386 + }, + { + "epoch": 2.594337658732912, + "grad_norm": 1.3378093242645264, + "learning_rate": 5.807369834248035e-05, + "loss": 0.4464, + "step": 37387 + }, + { + "epoch": 2.594407050170009, + "grad_norm": 1.142906904220581, + "learning_rate": 5.806859746536566e-05, + "loss": 0.3407, + "step": 37388 + }, + { + "epoch": 2.5944764416071058, + "grad_norm": 1.342431902885437, + "learning_rate": 5.80634967206235e-05, + "loss": 0.3841, + "step": 37389 + }, + { + "epoch": 2.5945458330442026, + "grad_norm": 0.963773250579834, + "learning_rate": 5.805839610827002e-05, + "loss": 0.2577, + "step": 37390 + }, + { + "epoch": 2.594615224481299, + "grad_norm": 1.0097343921661377, + "learning_rate": 5.805329562832132e-05, + "loss": 0.3698, + "step": 37391 + }, + { + "epoch": 2.594684615918396, + "grad_norm": 2.053615093231201, + "learning_rate": 5.804819528079349e-05, + "loss": 0.3316, + "step": 37392 + }, + { + "epoch": 2.594754007355492, + "grad_norm": 1.2738149166107178, + "learning_rate": 5.8043095065702645e-05, + "loss": 0.5073, + "step": 37393 + }, + { + "epoch": 2.594823398792589, + "grad_norm": 1.2369714975357056, + "learning_rate": 5.803799498306488e-05, + "loss": 0.543, + "step": 37394 + }, + { + "epoch": 2.594892790229686, + "grad_norm": 1.3906750679016113, + "learning_rate": 5.803289503289635e-05, + "loss": 0.6928, + "step": 37395 + }, + { + "epoch": 2.5949621816667823, + "grad_norm": 1.0525490045547485, + "learning_rate": 5.802779521521304e-05, + "loss": 0.3671, + "step": 37396 + }, + { + "epoch": 2.595031573103879, + "grad_norm": 1.3534530401229858, + "learning_rate": 5.8022695530031127e-05, + "loss": 0.3719, + "step": 37397 + }, + { + "epoch": 2.5951009645409755, + "grad_norm": 1.0854039192199707, + "learning_rate": 5.801759597736669e-05, + "loss": 0.4873, + "step": 37398 + }, + { + "epoch": 2.5951703559780723, + "grad_norm": 1.2344789505004883, + "learning_rate": 5.801249655723582e-05, + "loss": 0.3764, + "step": 37399 + }, + { + "epoch": 2.595239747415169, + "grad_norm": 2.2823150157928467, + "learning_rate": 5.800739726965463e-05, + "loss": 0.44, + "step": 37400 + }, + { + "epoch": 2.5953091388522656, + "grad_norm": 1.4570612907409668, + "learning_rate": 5.800229811463922e-05, + "loss": 0.5578, + "step": 37401 + }, + { + "epoch": 2.5953785302893624, + "grad_norm": 0.7493152022361755, + "learning_rate": 5.7997199092205735e-05, + "loss": 0.2392, + "step": 37402 + }, + { + "epoch": 2.5954479217264588, + "grad_norm": 1.3454557657241821, + "learning_rate": 5.799210020237016e-05, + "loss": 0.3797, + "step": 37403 + }, + { + "epoch": 2.5955173131635556, + "grad_norm": 1.3399245738983154, + "learning_rate": 5.798700144514866e-05, + "loss": 0.5058, + "step": 37404 + }, + { + "epoch": 2.5955867046006524, + "grad_norm": 1.0874143838882446, + "learning_rate": 5.798190282055731e-05, + "loss": 0.3766, + "step": 37405 + }, + { + "epoch": 2.595656096037749, + "grad_norm": 1.0681381225585938, + "learning_rate": 5.7976804328612274e-05, + "loss": 0.3075, + "step": 37406 + }, + { + "epoch": 2.5957254874748457, + "grad_norm": 1.001291036605835, + "learning_rate": 5.797170596932955e-05, + "loss": 0.4054, + "step": 37407 + }, + { + "epoch": 2.595794878911942, + "grad_norm": 1.3671609163284302, + "learning_rate": 5.7966607742725265e-05, + "loss": 0.4228, + "step": 37408 + }, + { + "epoch": 2.595864270349039, + "grad_norm": 1.7595652341842651, + "learning_rate": 5.796150964881552e-05, + "loss": 0.3625, + "step": 37409 + }, + { + "epoch": 2.5959336617861357, + "grad_norm": 1.0123741626739502, + "learning_rate": 5.7956411687616464e-05, + "loss": 0.4369, + "step": 37410 + }, + { + "epoch": 2.596003053223232, + "grad_norm": 1.339557409286499, + "learning_rate": 5.79513138591441e-05, + "loss": 0.395, + "step": 37411 + }, + { + "epoch": 2.596072444660329, + "grad_norm": 1.4821945428848267, + "learning_rate": 5.7946216163414544e-05, + "loss": 0.6221, + "step": 37412 + }, + { + "epoch": 2.5961418360974253, + "grad_norm": 1.1561057567596436, + "learning_rate": 5.794111860044391e-05, + "loss": 0.4151, + "step": 37413 + }, + { + "epoch": 2.596211227534522, + "grad_norm": 1.5789730548858643, + "learning_rate": 5.793602117024829e-05, + "loss": 0.4108, + "step": 37414 + }, + { + "epoch": 2.596280618971619, + "grad_norm": 0.8875080943107605, + "learning_rate": 5.793092387284377e-05, + "loss": 0.3346, + "step": 37415 + }, + { + "epoch": 2.596350010408716, + "grad_norm": 1.428969383239746, + "learning_rate": 5.7925826708246475e-05, + "loss": 0.4179, + "step": 37416 + }, + { + "epoch": 2.596419401845812, + "grad_norm": 0.860265851020813, + "learning_rate": 5.7920729676472443e-05, + "loss": 0.2931, + "step": 37417 + }, + { + "epoch": 2.596488793282909, + "grad_norm": 1.308082103729248, + "learning_rate": 5.7915632777537756e-05, + "loss": 0.4335, + "step": 37418 + }, + { + "epoch": 2.5965581847200054, + "grad_norm": 1.0888816118240356, + "learning_rate": 5.791053601145855e-05, + "loss": 0.2392, + "step": 37419 + }, + { + "epoch": 2.5966275761571023, + "grad_norm": 1.2234501838684082, + "learning_rate": 5.79054393782509e-05, + "loss": 0.3179, + "step": 37420 + }, + { + "epoch": 2.596696967594199, + "grad_norm": 1.1070033311843872, + "learning_rate": 5.790034287793089e-05, + "loss": 0.3042, + "step": 37421 + }, + { + "epoch": 2.5967663590312955, + "grad_norm": 0.8161754608154297, + "learning_rate": 5.789524651051461e-05, + "loss": 0.3284, + "step": 37422 + }, + { + "epoch": 2.5968357504683923, + "grad_norm": 1.0698726177215576, + "learning_rate": 5.789015027601821e-05, + "loss": 0.303, + "step": 37423 + }, + { + "epoch": 2.5969051419054887, + "grad_norm": 1.2202950716018677, + "learning_rate": 5.7885054174457664e-05, + "loss": 0.4657, + "step": 37424 + }, + { + "epoch": 2.5969745333425855, + "grad_norm": 1.4397497177124023, + "learning_rate": 5.787995820584912e-05, + "loss": 0.5211, + "step": 37425 + }, + { + "epoch": 2.5970439247796824, + "grad_norm": 2.2150721549987793, + "learning_rate": 5.787486237020865e-05, + "loss": 0.543, + "step": 37426 + }, + { + "epoch": 2.5971133162167788, + "grad_norm": 1.7476388216018677, + "learning_rate": 5.786976666755237e-05, + "loss": 0.6427, + "step": 37427 + }, + { + "epoch": 2.5971827076538756, + "grad_norm": 1.6435433626174927, + "learning_rate": 5.786467109789633e-05, + "loss": 0.5477, + "step": 37428 + }, + { + "epoch": 2.597252099090972, + "grad_norm": 1.023762822151184, + "learning_rate": 5.785957566125665e-05, + "loss": 0.36, + "step": 37429 + }, + { + "epoch": 2.597321490528069, + "grad_norm": 1.2487728595733643, + "learning_rate": 5.7854480357649435e-05, + "loss": 0.3877, + "step": 37430 + }, + { + "epoch": 2.5973908819651657, + "grad_norm": 1.0678716897964478, + "learning_rate": 5.784938518709069e-05, + "loss": 0.4066, + "step": 37431 + }, + { + "epoch": 2.597460273402262, + "grad_norm": 1.4729175567626953, + "learning_rate": 5.784429014959656e-05, + "loss": 0.5849, + "step": 37432 + }, + { + "epoch": 2.597529664839359, + "grad_norm": 1.1063721179962158, + "learning_rate": 5.7839195245183105e-05, + "loss": 0.4354, + "step": 37433 + }, + { + "epoch": 2.5975990562764553, + "grad_norm": 1.0015190839767456, + "learning_rate": 5.783410047386641e-05, + "loss": 0.3453, + "step": 37434 + }, + { + "epoch": 2.597668447713552, + "grad_norm": 0.8322934508323669, + "learning_rate": 5.7829005835662585e-05, + "loss": 0.3435, + "step": 37435 + }, + { + "epoch": 2.597737839150649, + "grad_norm": 1.5743367671966553, + "learning_rate": 5.7823911330587687e-05, + "loss": 0.411, + "step": 37436 + }, + { + "epoch": 2.5978072305877453, + "grad_norm": 1.3175498247146606, + "learning_rate": 5.781881695865785e-05, + "loss": 0.3693, + "step": 37437 + }, + { + "epoch": 2.597876622024842, + "grad_norm": 1.1608973741531372, + "learning_rate": 5.781372271988907e-05, + "loss": 0.4247, + "step": 37438 + }, + { + "epoch": 2.5979460134619385, + "grad_norm": 1.0794371366500854, + "learning_rate": 5.7808628614297476e-05, + "loss": 0.3404, + "step": 37439 + }, + { + "epoch": 2.5980154048990354, + "grad_norm": 1.3525989055633545, + "learning_rate": 5.780353464189915e-05, + "loss": 0.4697, + "step": 37440 + }, + { + "epoch": 2.598084796336132, + "grad_norm": 1.4225735664367676, + "learning_rate": 5.779844080271021e-05, + "loss": 0.3486, + "step": 37441 + }, + { + "epoch": 2.598154187773229, + "grad_norm": 1.2585445642471313, + "learning_rate": 5.779334709674664e-05, + "loss": 0.254, + "step": 37442 + }, + { + "epoch": 2.5982235792103254, + "grad_norm": 1.4077227115631104, + "learning_rate": 5.778825352402459e-05, + "loss": 0.4637, + "step": 37443 + }, + { + "epoch": 2.5982929706474223, + "grad_norm": 1.2139232158660889, + "learning_rate": 5.778316008456013e-05, + "loss": 0.3713, + "step": 37444 + }, + { + "epoch": 2.5983623620845187, + "grad_norm": 1.2229880094528198, + "learning_rate": 5.777806677836937e-05, + "loss": 0.2987, + "step": 37445 + }, + { + "epoch": 2.5984317535216155, + "grad_norm": 1.0240986347198486, + "learning_rate": 5.7772973605468316e-05, + "loss": 0.3438, + "step": 37446 + }, + { + "epoch": 2.5985011449587123, + "grad_norm": 1.2282321453094482, + "learning_rate": 5.776788056587308e-05, + "loss": 0.3903, + "step": 37447 + }, + { + "epoch": 2.5985705363958087, + "grad_norm": 1.3169375658035278, + "learning_rate": 5.776278765959975e-05, + "loss": 0.3994, + "step": 37448 + }, + { + "epoch": 2.5986399278329055, + "grad_norm": 1.0003947019577026, + "learning_rate": 5.775769488666441e-05, + "loss": 0.3726, + "step": 37449 + }, + { + "epoch": 2.598709319270002, + "grad_norm": 1.2141902446746826, + "learning_rate": 5.77526022470831e-05, + "loss": 0.6037, + "step": 37450 + }, + { + "epoch": 2.5987787107070988, + "grad_norm": 1.008312463760376, + "learning_rate": 5.7747509740871984e-05, + "loss": 0.3638, + "step": 37451 + }, + { + "epoch": 2.5988481021441956, + "grad_norm": 1.0906015634536743, + "learning_rate": 5.774241736804703e-05, + "loss": 0.3076, + "step": 37452 + }, + { + "epoch": 2.598917493581292, + "grad_norm": 1.7575690746307373, + "learning_rate": 5.773732512862437e-05, + "loss": 0.4349, + "step": 37453 + }, + { + "epoch": 2.598986885018389, + "grad_norm": 1.4421976804733276, + "learning_rate": 5.7732233022620055e-05, + "loss": 0.5345, + "step": 37454 + }, + { + "epoch": 2.599056276455485, + "grad_norm": 1.1506997346878052, + "learning_rate": 5.772714105005019e-05, + "loss": 0.3648, + "step": 37455 + }, + { + "epoch": 2.599125667892582, + "grad_norm": 0.9060435891151428, + "learning_rate": 5.772204921093083e-05, + "loss": 0.3243, + "step": 37456 + }, + { + "epoch": 2.599195059329679, + "grad_norm": 1.4654322862625122, + "learning_rate": 5.7716957505278056e-05, + "loss": 0.5804, + "step": 37457 + }, + { + "epoch": 2.5992644507667753, + "grad_norm": 1.414117693901062, + "learning_rate": 5.771186593310798e-05, + "loss": 0.4599, + "step": 37458 + }, + { + "epoch": 2.599333842203872, + "grad_norm": 1.497198224067688, + "learning_rate": 5.77067744944366e-05, + "loss": 0.4213, + "step": 37459 + }, + { + "epoch": 2.5994032336409685, + "grad_norm": 1.1616343259811401, + "learning_rate": 5.770168318928002e-05, + "loss": 0.4283, + "step": 37460 + }, + { + "epoch": 2.5994726250780653, + "grad_norm": 1.8901642560958862, + "learning_rate": 5.7696592017654325e-05, + "loss": 0.6732, + "step": 37461 + }, + { + "epoch": 2.599542016515162, + "grad_norm": 1.2116167545318604, + "learning_rate": 5.769150097957559e-05, + "loss": 0.3328, + "step": 37462 + }, + { + "epoch": 2.5996114079522585, + "grad_norm": 1.2184441089630127, + "learning_rate": 5.768641007505986e-05, + "loss": 0.5141, + "step": 37463 + }, + { + "epoch": 2.5996807993893554, + "grad_norm": 1.3624924421310425, + "learning_rate": 5.7681319304123275e-05, + "loss": 0.4248, + "step": 37464 + }, + { + "epoch": 2.5997501908264518, + "grad_norm": 2.4811956882476807, + "learning_rate": 5.767622866678181e-05, + "loss": 0.5066, + "step": 37465 + }, + { + "epoch": 2.5998195822635486, + "grad_norm": 1.5043461322784424, + "learning_rate": 5.767113816305159e-05, + "loss": 0.3886, + "step": 37466 + }, + { + "epoch": 2.5998889737006454, + "grad_norm": 1.1370744705200195, + "learning_rate": 5.766604779294866e-05, + "loss": 0.461, + "step": 37467 + }, + { + "epoch": 2.5999583651377423, + "grad_norm": 1.3906776905059814, + "learning_rate": 5.7660957556489104e-05, + "loss": 0.4795, + "step": 37468 + }, + { + "epoch": 2.6000277565748386, + "grad_norm": 1.6396901607513428, + "learning_rate": 5.765586745368901e-05, + "loss": 0.3926, + "step": 37469 + }, + { + "epoch": 2.6000971480119355, + "grad_norm": 1.1427944898605347, + "learning_rate": 5.765077748456442e-05, + "loss": 0.3891, + "step": 37470 + }, + { + "epoch": 2.600166539449032, + "grad_norm": 1.0591546297073364, + "learning_rate": 5.764568764913145e-05, + "loss": 0.3876, + "step": 37471 + }, + { + "epoch": 2.6002359308861287, + "grad_norm": 1.3037713766098022, + "learning_rate": 5.76405979474061e-05, + "loss": 0.5537, + "step": 37472 + }, + { + "epoch": 2.6003053223232255, + "grad_norm": 1.4766448736190796, + "learning_rate": 5.763550837940446e-05, + "loss": 0.4987, + "step": 37473 + }, + { + "epoch": 2.600374713760322, + "grad_norm": 1.2026647329330444, + "learning_rate": 5.7630418945142604e-05, + "loss": 0.4114, + "step": 37474 + }, + { + "epoch": 2.6004441051974188, + "grad_norm": 1.6712948083877563, + "learning_rate": 5.7625329644636604e-05, + "loss": 0.3658, + "step": 37475 + }, + { + "epoch": 2.600513496634515, + "grad_norm": 1.0679336786270142, + "learning_rate": 5.762024047790255e-05, + "loss": 0.3442, + "step": 37476 + }, + { + "epoch": 2.600582888071612, + "grad_norm": 0.8765162825584412, + "learning_rate": 5.761515144495646e-05, + "loss": 0.316, + "step": 37477 + }, + { + "epoch": 2.600652279508709, + "grad_norm": 2.1194379329681396, + "learning_rate": 5.7610062545814406e-05, + "loss": 0.5535, + "step": 37478 + }, + { + "epoch": 2.600721670945805, + "grad_norm": 1.1459294557571411, + "learning_rate": 5.760497378049251e-05, + "loss": 0.4075, + "step": 37479 + }, + { + "epoch": 2.600791062382902, + "grad_norm": 0.9828515648841858, + "learning_rate": 5.759988514900676e-05, + "loss": 0.4078, + "step": 37480 + }, + { + "epoch": 2.6008604538199984, + "grad_norm": 1.0999151468276978, + "learning_rate": 5.759479665137324e-05, + "loss": 0.5307, + "step": 37481 + }, + { + "epoch": 2.6009298452570953, + "grad_norm": 1.7249094247817993, + "learning_rate": 5.7589708287608035e-05, + "loss": 0.4583, + "step": 37482 + }, + { + "epoch": 2.600999236694192, + "grad_norm": 2.2362966537475586, + "learning_rate": 5.75846200577272e-05, + "loss": 0.5146, + "step": 37483 + }, + { + "epoch": 2.6010686281312885, + "grad_norm": 1.159713864326477, + "learning_rate": 5.7579531961746793e-05, + "loss": 0.4385, + "step": 37484 + }, + { + "epoch": 2.6011380195683853, + "grad_norm": 1.6146882772445679, + "learning_rate": 5.757444399968293e-05, + "loss": 0.3907, + "step": 37485 + }, + { + "epoch": 2.6012074110054817, + "grad_norm": 1.0015445947647095, + "learning_rate": 5.756935617155157e-05, + "loss": 0.3156, + "step": 37486 + }, + { + "epoch": 2.6012768024425785, + "grad_norm": 1.4947776794433594, + "learning_rate": 5.756426847736884e-05, + "loss": 0.6181, + "step": 37487 + }, + { + "epoch": 2.6013461938796754, + "grad_norm": 1.1157857179641724, + "learning_rate": 5.7559180917150804e-05, + "loss": 0.4688, + "step": 37488 + }, + { + "epoch": 2.6014155853167718, + "grad_norm": 0.7660271525382996, + "learning_rate": 5.7554093490913495e-05, + "loss": 0.2269, + "step": 37489 + }, + { + "epoch": 2.6014849767538686, + "grad_norm": 1.3810259103775024, + "learning_rate": 5.7549006198672984e-05, + "loss": 0.3745, + "step": 37490 + }, + { + "epoch": 2.601554368190965, + "grad_norm": 1.1838394403457642, + "learning_rate": 5.754391904044534e-05, + "loss": 0.5154, + "step": 37491 + }, + { + "epoch": 2.601623759628062, + "grad_norm": 1.0918394327163696, + "learning_rate": 5.753883201624666e-05, + "loss": 0.3815, + "step": 37492 + }, + { + "epoch": 2.6016931510651586, + "grad_norm": 0.896087646484375, + "learning_rate": 5.753374512609292e-05, + "loss": 0.3876, + "step": 37493 + }, + { + "epoch": 2.6017625425022555, + "grad_norm": 1.061120867729187, + "learning_rate": 5.752865837000022e-05, + "loss": 0.3301, + "step": 37494 + }, + { + "epoch": 2.601831933939352, + "grad_norm": 1.38890540599823, + "learning_rate": 5.7523571747984614e-05, + "loss": 0.3142, + "step": 37495 + }, + { + "epoch": 2.6019013253764487, + "grad_norm": 0.9146710634231567, + "learning_rate": 5.751848526006217e-05, + "loss": 0.3348, + "step": 37496 + }, + { + "epoch": 2.601970716813545, + "grad_norm": 1.244737982749939, + "learning_rate": 5.751339890624894e-05, + "loss": 0.3972, + "step": 37497 + }, + { + "epoch": 2.602040108250642, + "grad_norm": 0.9439429044723511, + "learning_rate": 5.750831268656098e-05, + "loss": 0.4068, + "step": 37498 + }, + { + "epoch": 2.6021094996877387, + "grad_norm": 0.9997959136962891, + "learning_rate": 5.750322660101438e-05, + "loss": 0.4139, + "step": 37499 + }, + { + "epoch": 2.602178891124835, + "grad_norm": 1.13053297996521, + "learning_rate": 5.749814064962513e-05, + "loss": 0.3523, + "step": 37500 + }, + { + "epoch": 2.602248282561932, + "grad_norm": 0.978289008140564, + "learning_rate": 5.7493054832409324e-05, + "loss": 0.3238, + "step": 37501 + }, + { + "epoch": 2.6023176739990284, + "grad_norm": 1.6692140102386475, + "learning_rate": 5.748796914938299e-05, + "loss": 0.5563, + "step": 37502 + }, + { + "epoch": 2.602387065436125, + "grad_norm": 1.0395519733428955, + "learning_rate": 5.748288360056222e-05, + "loss": 0.2794, + "step": 37503 + }, + { + "epoch": 2.602456456873222, + "grad_norm": 0.9829789996147156, + "learning_rate": 5.747779818596306e-05, + "loss": 0.4068, + "step": 37504 + }, + { + "epoch": 2.6025258483103184, + "grad_norm": 1.0424636602401733, + "learning_rate": 5.7472712905601544e-05, + "loss": 0.3204, + "step": 37505 + }, + { + "epoch": 2.6025952397474152, + "grad_norm": 1.0802534818649292, + "learning_rate": 5.746762775949379e-05, + "loss": 0.381, + "step": 37506 + }, + { + "epoch": 2.6026646311845116, + "grad_norm": 0.9183382391929626, + "learning_rate": 5.7462542747655744e-05, + "loss": 0.2521, + "step": 37507 + }, + { + "epoch": 2.6027340226216085, + "grad_norm": 1.169664740562439, + "learning_rate": 5.7457457870103535e-05, + "loss": 0.3691, + "step": 37508 + }, + { + "epoch": 2.6028034140587053, + "grad_norm": 1.286376953125, + "learning_rate": 5.7452373126853184e-05, + "loss": 0.3554, + "step": 37509 + }, + { + "epoch": 2.6028728054958017, + "grad_norm": 1.2942827939987183, + "learning_rate": 5.7447288517920764e-05, + "loss": 0.3537, + "step": 37510 + }, + { + "epoch": 2.6029421969328985, + "grad_norm": 0.8873911499977112, + "learning_rate": 5.744220404332235e-05, + "loss": 0.2741, + "step": 37511 + }, + { + "epoch": 2.603011588369995, + "grad_norm": 1.7791063785552979, + "learning_rate": 5.743711970307393e-05, + "loss": 0.4579, + "step": 37512 + }, + { + "epoch": 2.6030809798070917, + "grad_norm": 1.3130238056182861, + "learning_rate": 5.743203549719157e-05, + "loss": 0.5035, + "step": 37513 + }, + { + "epoch": 2.6031503712441886, + "grad_norm": 1.104153037071228, + "learning_rate": 5.742695142569139e-05, + "loss": 0.3338, + "step": 37514 + }, + { + "epoch": 2.603219762681285, + "grad_norm": 1.2671542167663574, + "learning_rate": 5.742186748858933e-05, + "loss": 0.4035, + "step": 37515 + }, + { + "epoch": 2.603289154118382, + "grad_norm": 1.09714937210083, + "learning_rate": 5.7416783685901506e-05, + "loss": 0.3955, + "step": 37516 + }, + { + "epoch": 2.603358545555478, + "grad_norm": 1.0707708597183228, + "learning_rate": 5.7411700017643945e-05, + "loss": 0.3342, + "step": 37517 + }, + { + "epoch": 2.603427936992575, + "grad_norm": 0.7371609807014465, + "learning_rate": 5.740661648383272e-05, + "loss": 0.239, + "step": 37518 + }, + { + "epoch": 2.603497328429672, + "grad_norm": 1.565545916557312, + "learning_rate": 5.740153308448385e-05, + "loss": 0.4072, + "step": 37519 + }, + { + "epoch": 2.6035667198667687, + "grad_norm": 1.7204148769378662, + "learning_rate": 5.739644981961345e-05, + "loss": 0.4804, + "step": 37520 + }, + { + "epoch": 2.603636111303865, + "grad_norm": 1.2452807426452637, + "learning_rate": 5.739136668923746e-05, + "loss": 0.3661, + "step": 37521 + }, + { + "epoch": 2.603705502740962, + "grad_norm": 1.2910354137420654, + "learning_rate": 5.7386283693372e-05, + "loss": 0.3888, + "step": 37522 + }, + { + "epoch": 2.6037748941780583, + "grad_norm": 1.3274556398391724, + "learning_rate": 5.738120083203309e-05, + "loss": 0.3052, + "step": 37523 + }, + { + "epoch": 2.603844285615155, + "grad_norm": 1.3927522897720337, + "learning_rate": 5.737611810523677e-05, + "loss": 0.3804, + "step": 37524 + }, + { + "epoch": 2.603913677052252, + "grad_norm": 1.1796364784240723, + "learning_rate": 5.7371035512999114e-05, + "loss": 0.2905, + "step": 37525 + }, + { + "epoch": 2.6039830684893484, + "grad_norm": 1.7508174180984497, + "learning_rate": 5.7365953055336165e-05, + "loss": 0.3893, + "step": 37526 + }, + { + "epoch": 2.604052459926445, + "grad_norm": 1.3924068212509155, + "learning_rate": 5.7360870732263974e-05, + "loss": 0.3201, + "step": 37527 + }, + { + "epoch": 2.6041218513635416, + "grad_norm": 1.0243735313415527, + "learning_rate": 5.735578854379854e-05, + "loss": 0.3105, + "step": 37528 + }, + { + "epoch": 2.6041912428006384, + "grad_norm": 1.308716893196106, + "learning_rate": 5.735070648995593e-05, + "loss": 0.4158, + "step": 37529 + }, + { + "epoch": 2.6042606342377352, + "grad_norm": 1.232298731803894, + "learning_rate": 5.734562457075219e-05, + "loss": 0.4112, + "step": 37530 + }, + { + "epoch": 2.6043300256748316, + "grad_norm": 1.1066938638687134, + "learning_rate": 5.7340542786203363e-05, + "loss": 0.4201, + "step": 37531 + }, + { + "epoch": 2.6043994171119285, + "grad_norm": 1.2410507202148438, + "learning_rate": 5.733546113632551e-05, + "loss": 0.3959, + "step": 37532 + }, + { + "epoch": 2.604468808549025, + "grad_norm": 1.5102099180221558, + "learning_rate": 5.733037962113463e-05, + "loss": 0.4426, + "step": 37533 + }, + { + "epoch": 2.6045381999861217, + "grad_norm": 1.4518934488296509, + "learning_rate": 5.732529824064685e-05, + "loss": 0.3327, + "step": 37534 + }, + { + "epoch": 2.6046075914232185, + "grad_norm": 1.003022313117981, + "learning_rate": 5.732021699487811e-05, + "loss": 0.3278, + "step": 37535 + }, + { + "epoch": 2.604676982860315, + "grad_norm": 1.105146884918213, + "learning_rate": 5.7315135883844476e-05, + "loss": 0.3807, + "step": 37536 + }, + { + "epoch": 2.6047463742974117, + "grad_norm": 1.0348618030548096, + "learning_rate": 5.731005490756203e-05, + "loss": 0.241, + "step": 37537 + }, + { + "epoch": 2.604815765734508, + "grad_norm": 1.1945741176605225, + "learning_rate": 5.730497406604677e-05, + "loss": 0.3717, + "step": 37538 + }, + { + "epoch": 2.604885157171605, + "grad_norm": 0.9905992150306702, + "learning_rate": 5.729989335931476e-05, + "loss": 0.3742, + "step": 37539 + }, + { + "epoch": 2.604954548608702, + "grad_norm": 1.2405470609664917, + "learning_rate": 5.729481278738207e-05, + "loss": 0.4232, + "step": 37540 + }, + { + "epoch": 2.6050239400457986, + "grad_norm": 1.3329931497573853, + "learning_rate": 5.728973235026467e-05, + "loss": 0.3271, + "step": 37541 + }, + { + "epoch": 2.605093331482895, + "grad_norm": 1.8652527332305908, + "learning_rate": 5.7284652047978615e-05, + "loss": 0.4289, + "step": 37542 + }, + { + "epoch": 2.6051627229199914, + "grad_norm": 0.9677611589431763, + "learning_rate": 5.7279571880539965e-05, + "loss": 0.312, + "step": 37543 + }, + { + "epoch": 2.6052321143570882, + "grad_norm": 1.1967360973358154, + "learning_rate": 5.727449184796475e-05, + "loss": 0.409, + "step": 37544 + }, + { + "epoch": 2.605301505794185, + "grad_norm": 1.113317608833313, + "learning_rate": 5.7269411950269046e-05, + "loss": 0.3729, + "step": 37545 + }, + { + "epoch": 2.605370897231282, + "grad_norm": 0.9339604377746582, + "learning_rate": 5.7264332187468816e-05, + "loss": 0.2736, + "step": 37546 + }, + { + "epoch": 2.6054402886683783, + "grad_norm": 2.137319803237915, + "learning_rate": 5.725925255958014e-05, + "loss": 0.4289, + "step": 37547 + }, + { + "epoch": 2.605509680105475, + "grad_norm": 1.6885006427764893, + "learning_rate": 5.7254173066619024e-05, + "loss": 0.5611, + "step": 37548 + }, + { + "epoch": 2.6055790715425715, + "grad_norm": 1.021209955215454, + "learning_rate": 5.724909370860156e-05, + "loss": 0.3936, + "step": 37549 + }, + { + "epoch": 2.6056484629796683, + "grad_norm": 1.1478086709976196, + "learning_rate": 5.724401448554373e-05, + "loss": 0.4433, + "step": 37550 + }, + { + "epoch": 2.605717854416765, + "grad_norm": 1.5205695629119873, + "learning_rate": 5.723893539746157e-05, + "loss": 0.5875, + "step": 37551 + }, + { + "epoch": 2.6057872458538616, + "grad_norm": 1.195351481437683, + "learning_rate": 5.723385644437114e-05, + "loss": 0.2767, + "step": 37552 + }, + { + "epoch": 2.6058566372909584, + "grad_norm": 1.4414321184158325, + "learning_rate": 5.7228777626288464e-05, + "loss": 0.4354, + "step": 37553 + }, + { + "epoch": 2.605926028728055, + "grad_norm": 1.206697940826416, + "learning_rate": 5.722369894322961e-05, + "loss": 0.3159, + "step": 37554 + }, + { + "epoch": 2.6059954201651516, + "grad_norm": 1.5399749279022217, + "learning_rate": 5.721862039521053e-05, + "loss": 0.5145, + "step": 37555 + }, + { + "epoch": 2.6060648116022485, + "grad_norm": 1.3068947792053223, + "learning_rate": 5.721354198224731e-05, + "loss": 0.4021, + "step": 37556 + }, + { + "epoch": 2.606134203039345, + "grad_norm": 1.1932917833328247, + "learning_rate": 5.7208463704355965e-05, + "loss": 0.2883, + "step": 37557 + }, + { + "epoch": 2.6062035944764417, + "grad_norm": 0.8361765742301941, + "learning_rate": 5.7203385561552535e-05, + "loss": 0.3051, + "step": 37558 + }, + { + "epoch": 2.606272985913538, + "grad_norm": 1.0691102743148804, + "learning_rate": 5.719830755385307e-05, + "loss": 0.3396, + "step": 37559 + }, + { + "epoch": 2.606342377350635, + "grad_norm": 1.6588962078094482, + "learning_rate": 5.7193229681273566e-05, + "loss": 0.5737, + "step": 37560 + }, + { + "epoch": 2.6064117687877317, + "grad_norm": 1.0876622200012207, + "learning_rate": 5.718815194383011e-05, + "loss": 0.3877, + "step": 37561 + }, + { + "epoch": 2.606481160224828, + "grad_norm": 1.580540418624878, + "learning_rate": 5.718307434153865e-05, + "loss": 0.3287, + "step": 37562 + }, + { + "epoch": 2.606550551661925, + "grad_norm": 1.243748664855957, + "learning_rate": 5.717799687441526e-05, + "loss": 0.4788, + "step": 37563 + }, + { + "epoch": 2.6066199430990213, + "grad_norm": 1.030666708946228, + "learning_rate": 5.7172919542475964e-05, + "loss": 0.3568, + "step": 37564 + }, + { + "epoch": 2.606689334536118, + "grad_norm": 1.1410927772521973, + "learning_rate": 5.71678423457368e-05, + "loss": 0.5095, + "step": 37565 + }, + { + "epoch": 2.606758725973215, + "grad_norm": 1.6317031383514404, + "learning_rate": 5.716276528421378e-05, + "loss": 0.4939, + "step": 37566 + }, + { + "epoch": 2.606828117410312, + "grad_norm": 1.1197751760482788, + "learning_rate": 5.7157688357922944e-05, + "loss": 0.4917, + "step": 37567 + }, + { + "epoch": 2.6068975088474082, + "grad_norm": 1.7407606840133667, + "learning_rate": 5.715261156688035e-05, + "loss": 0.4751, + "step": 37568 + }, + { + "epoch": 2.6069669002845046, + "grad_norm": 1.2692577838897705, + "learning_rate": 5.7147534911101965e-05, + "loss": 0.4085, + "step": 37569 + }, + { + "epoch": 2.6070362917216015, + "grad_norm": 0.949849545955658, + "learning_rate": 5.714245839060381e-05, + "loss": 0.2831, + "step": 37570 + }, + { + "epoch": 2.6071056831586983, + "grad_norm": 0.9894232749938965, + "learning_rate": 5.713738200540197e-05, + "loss": 0.3148, + "step": 37571 + }, + { + "epoch": 2.607175074595795, + "grad_norm": 1.8000034093856812, + "learning_rate": 5.713230575551243e-05, + "loss": 0.5898, + "step": 37572 + }, + { + "epoch": 2.6072444660328915, + "grad_norm": 1.493811845779419, + "learning_rate": 5.712722964095123e-05, + "loss": 0.5643, + "step": 37573 + }, + { + "epoch": 2.6073138574699883, + "grad_norm": 1.270134687423706, + "learning_rate": 5.712215366173439e-05, + "loss": 0.4174, + "step": 37574 + }, + { + "epoch": 2.6073832489070847, + "grad_norm": 1.266244888305664, + "learning_rate": 5.7117077817877985e-05, + "loss": 0.4151, + "step": 37575 + }, + { + "epoch": 2.6074526403441816, + "grad_norm": 0.9545316100120544, + "learning_rate": 5.711200210939794e-05, + "loss": 0.451, + "step": 37576 + }, + { + "epoch": 2.6075220317812784, + "grad_norm": 0.7564358115196228, + "learning_rate": 5.7106926536310333e-05, + "loss": 0.4501, + "step": 37577 + }, + { + "epoch": 2.607591423218375, + "grad_norm": 1.2978878021240234, + "learning_rate": 5.7101851098631175e-05, + "loss": 0.52, + "step": 37578 + }, + { + "epoch": 2.6076608146554716, + "grad_norm": 0.9590240716934204, + "learning_rate": 5.709677579637651e-05, + "loss": 0.3025, + "step": 37579 + }, + { + "epoch": 2.607730206092568, + "grad_norm": 1.0638768672943115, + "learning_rate": 5.7091700629562375e-05, + "loss": 0.2429, + "step": 37580 + }, + { + "epoch": 2.607799597529665, + "grad_norm": 1.0221925973892212, + "learning_rate": 5.7086625598204725e-05, + "loss": 0.3933, + "step": 37581 + }, + { + "epoch": 2.6078689889667617, + "grad_norm": 0.9200994968414307, + "learning_rate": 5.708155070231962e-05, + "loss": 0.3193, + "step": 37582 + }, + { + "epoch": 2.607938380403858, + "grad_norm": 1.3127484321594238, + "learning_rate": 5.707647594192308e-05, + "loss": 0.3747, + "step": 37583 + }, + { + "epoch": 2.608007771840955, + "grad_norm": 1.4432283639907837, + "learning_rate": 5.707140131703117e-05, + "loss": 0.3723, + "step": 37584 + }, + { + "epoch": 2.6080771632780513, + "grad_norm": 1.0732076168060303, + "learning_rate": 5.7066326827659824e-05, + "loss": 0.4416, + "step": 37585 + }, + { + "epoch": 2.608146554715148, + "grad_norm": 1.0164223909378052, + "learning_rate": 5.706125247382509e-05, + "loss": 0.4179, + "step": 37586 + }, + { + "epoch": 2.608215946152245, + "grad_norm": 1.0989152193069458, + "learning_rate": 5.7056178255543014e-05, + "loss": 0.3954, + "step": 37587 + }, + { + "epoch": 2.6082853375893413, + "grad_norm": 2.225696325302124, + "learning_rate": 5.7051104172829596e-05, + "loss": 0.4395, + "step": 37588 + }, + { + "epoch": 2.608354729026438, + "grad_norm": 1.4965403079986572, + "learning_rate": 5.70460302257009e-05, + "loss": 0.4768, + "step": 37589 + }, + { + "epoch": 2.6084241204635346, + "grad_norm": 1.2153793573379517, + "learning_rate": 5.7040956414172865e-05, + "loss": 0.3915, + "step": 37590 + }, + { + "epoch": 2.6084935119006314, + "grad_norm": 1.0306506156921387, + "learning_rate": 5.7035882738261535e-05, + "loss": 0.3627, + "step": 37591 + }, + { + "epoch": 2.6085629033377282, + "grad_norm": 1.1772302389144897, + "learning_rate": 5.703080919798295e-05, + "loss": 0.4349, + "step": 37592 + }, + { + "epoch": 2.608632294774825, + "grad_norm": 0.8653579950332642, + "learning_rate": 5.702573579335311e-05, + "loss": 0.3091, + "step": 37593 + }, + { + "epoch": 2.6087016862119214, + "grad_norm": 1.3222440481185913, + "learning_rate": 5.7020662524388046e-05, + "loss": 0.4076, + "step": 37594 + }, + { + "epoch": 2.6087710776490183, + "grad_norm": 1.2622047662734985, + "learning_rate": 5.7015589391103755e-05, + "loss": 0.4385, + "step": 37595 + }, + { + "epoch": 2.6088404690861147, + "grad_norm": 1.4106003046035767, + "learning_rate": 5.70105163935163e-05, + "loss": 0.4645, + "step": 37596 + }, + { + "epoch": 2.6089098605232115, + "grad_norm": 1.4731683731079102, + "learning_rate": 5.700544353164161e-05, + "loss": 0.4072, + "step": 37597 + }, + { + "epoch": 2.6089792519603083, + "grad_norm": 2.0630404949188232, + "learning_rate": 5.700037080549576e-05, + "loss": 0.4091, + "step": 37598 + }, + { + "epoch": 2.6090486433974047, + "grad_norm": 1.2076663970947266, + "learning_rate": 5.699529821509475e-05, + "loss": 0.3341, + "step": 37599 + }, + { + "epoch": 2.6091180348345016, + "grad_norm": 0.9925633072853088, + "learning_rate": 5.699022576045458e-05, + "loss": 0.2043, + "step": 37600 + }, + { + "epoch": 2.609187426271598, + "grad_norm": 1.0203328132629395, + "learning_rate": 5.69851534415913e-05, + "loss": 0.4085, + "step": 37601 + }, + { + "epoch": 2.6092568177086948, + "grad_norm": 1.1725634336471558, + "learning_rate": 5.698008125852089e-05, + "loss": 0.4407, + "step": 37602 + }, + { + "epoch": 2.6093262091457916, + "grad_norm": 1.3904123306274414, + "learning_rate": 5.69750092112594e-05, + "loss": 0.3761, + "step": 37603 + }, + { + "epoch": 2.609395600582888, + "grad_norm": 1.068469762802124, + "learning_rate": 5.6969937299822795e-05, + "loss": 0.3238, + "step": 37604 + }, + { + "epoch": 2.609464992019985, + "grad_norm": 1.2683626413345337, + "learning_rate": 5.696486552422708e-05, + "loss": 0.4197, + "step": 37605 + }, + { + "epoch": 2.609534383457081, + "grad_norm": 1.165368914604187, + "learning_rate": 5.695979388448831e-05, + "loss": 0.3426, + "step": 37606 + }, + { + "epoch": 2.609603774894178, + "grad_norm": 1.1111360788345337, + "learning_rate": 5.695472238062248e-05, + "loss": 0.2562, + "step": 37607 + }, + { + "epoch": 2.609673166331275, + "grad_norm": 1.8257914781570435, + "learning_rate": 5.69496510126456e-05, + "loss": 0.4579, + "step": 37608 + }, + { + "epoch": 2.6097425577683713, + "grad_norm": 1.2214393615722656, + "learning_rate": 5.694457978057366e-05, + "loss": 0.334, + "step": 37609 + }, + { + "epoch": 2.609811949205468, + "grad_norm": 1.0915422439575195, + "learning_rate": 5.693950868442273e-05, + "loss": 0.4291, + "step": 37610 + }, + { + "epoch": 2.6098813406425645, + "grad_norm": 1.2419519424438477, + "learning_rate": 5.693443772420875e-05, + "loss": 0.3756, + "step": 37611 + }, + { + "epoch": 2.6099507320796613, + "grad_norm": 2.197554349899292, + "learning_rate": 5.692936689994774e-05, + "loss": 0.373, + "step": 37612 + }, + { + "epoch": 2.610020123516758, + "grad_norm": 1.3086163997650146, + "learning_rate": 5.6924296211655714e-05, + "loss": 0.3258, + "step": 37613 + }, + { + "epoch": 2.6100895149538546, + "grad_norm": 0.9609914422035217, + "learning_rate": 5.6919225659348705e-05, + "loss": 0.3129, + "step": 37614 + }, + { + "epoch": 2.6101589063909514, + "grad_norm": 0.9677923917770386, + "learning_rate": 5.691415524304273e-05, + "loss": 0.4242, + "step": 37615 + }, + { + "epoch": 2.6102282978280478, + "grad_norm": 1.7430287599563599, + "learning_rate": 5.690908496275373e-05, + "loss": 0.4676, + "step": 37616 + }, + { + "epoch": 2.6102976892651446, + "grad_norm": 1.0040017366409302, + "learning_rate": 5.6904014818497763e-05, + "loss": 0.299, + "step": 37617 + }, + { + "epoch": 2.6103670807022414, + "grad_norm": 0.8647371530532837, + "learning_rate": 5.689894481029081e-05, + "loss": 0.3079, + "step": 37618 + }, + { + "epoch": 2.6104364721393383, + "grad_norm": 1.430944800376892, + "learning_rate": 5.689387493814893e-05, + "loss": 0.4684, + "step": 37619 + }, + { + "epoch": 2.6105058635764347, + "grad_norm": 1.2656841278076172, + "learning_rate": 5.688880520208804e-05, + "loss": 0.3917, + "step": 37620 + }, + { + "epoch": 2.6105752550135315, + "grad_norm": 1.0602641105651855, + "learning_rate": 5.688373560212419e-05, + "loss": 0.4237, + "step": 37621 + }, + { + "epoch": 2.610644646450628, + "grad_norm": 1.183182716369629, + "learning_rate": 5.6878666138273394e-05, + "loss": 0.2683, + "step": 37622 + }, + { + "epoch": 2.6107140378877247, + "grad_norm": 1.2961816787719727, + "learning_rate": 5.6873596810551665e-05, + "loss": 0.6099, + "step": 37623 + }, + { + "epoch": 2.6107834293248215, + "grad_norm": 1.002020001411438, + "learning_rate": 5.6868527618974964e-05, + "loss": 0.1879, + "step": 37624 + }, + { + "epoch": 2.610852820761918, + "grad_norm": 2.473475694656372, + "learning_rate": 5.6863458563559325e-05, + "loss": 0.4118, + "step": 37625 + }, + { + "epoch": 2.6109222121990148, + "grad_norm": 0.9962228536605835, + "learning_rate": 5.685838964432073e-05, + "loss": 0.3655, + "step": 37626 + }, + { + "epoch": 2.610991603636111, + "grad_norm": 1.9900044202804565, + "learning_rate": 5.68533208612752e-05, + "loss": 0.4356, + "step": 37627 + }, + { + "epoch": 2.611060995073208, + "grad_norm": 1.4759660959243774, + "learning_rate": 5.684825221443872e-05, + "loss": 0.367, + "step": 37628 + }, + { + "epoch": 2.611130386510305, + "grad_norm": 1.7117902040481567, + "learning_rate": 5.684318370382731e-05, + "loss": 0.4667, + "step": 37629 + }, + { + "epoch": 2.611199777947401, + "grad_norm": 1.2295440435409546, + "learning_rate": 5.6838115329457e-05, + "loss": 0.4669, + "step": 37630 + }, + { + "epoch": 2.611269169384498, + "grad_norm": 1.5143035650253296, + "learning_rate": 5.683304709134372e-05, + "loss": 0.4189, + "step": 37631 + }, + { + "epoch": 2.6113385608215944, + "grad_norm": 1.0255591869354248, + "learning_rate": 5.682797898950348e-05, + "loss": 0.2986, + "step": 37632 + }, + { + "epoch": 2.6114079522586913, + "grad_norm": 1.2753068208694458, + "learning_rate": 5.682291102395232e-05, + "loss": 0.4471, + "step": 37633 + }, + { + "epoch": 2.611477343695788, + "grad_norm": 1.9654453992843628, + "learning_rate": 5.6817843194706225e-05, + "loss": 0.5003, + "step": 37634 + }, + { + "epoch": 2.6115467351328845, + "grad_norm": 1.8083456754684448, + "learning_rate": 5.681277550178117e-05, + "loss": 0.4588, + "step": 37635 + }, + { + "epoch": 2.6116161265699813, + "grad_norm": 1.0325188636779785, + "learning_rate": 5.680770794519319e-05, + "loss": 0.2956, + "step": 37636 + }, + { + "epoch": 2.6116855180070777, + "grad_norm": 1.8511059284210205, + "learning_rate": 5.68026405249583e-05, + "loss": 0.315, + "step": 37637 + }, + { + "epoch": 2.6117549094441745, + "grad_norm": 0.8517943024635315, + "learning_rate": 5.67975732410924e-05, + "loss": 0.3245, + "step": 37638 + }, + { + "epoch": 2.6118243008812714, + "grad_norm": 1.3200030326843262, + "learning_rate": 5.679250609361157e-05, + "loss": 0.3061, + "step": 37639 + }, + { + "epoch": 2.6118936923183678, + "grad_norm": 1.634903073310852, + "learning_rate": 5.678743908253179e-05, + "loss": 0.4829, + "step": 37640 + }, + { + "epoch": 2.6119630837554646, + "grad_norm": 0.7955511212348938, + "learning_rate": 5.678237220786904e-05, + "loss": 0.2711, + "step": 37641 + }, + { + "epoch": 2.612032475192561, + "grad_norm": 1.2545156478881836, + "learning_rate": 5.677730546963933e-05, + "loss": 0.3913, + "step": 37642 + }, + { + "epoch": 2.612101866629658, + "grad_norm": 1.2222942113876343, + "learning_rate": 5.677223886785865e-05, + "loss": 0.4191, + "step": 37643 + }, + { + "epoch": 2.6121712580667547, + "grad_norm": 0.8644701242446899, + "learning_rate": 5.676717240254305e-05, + "loss": 0.2948, + "step": 37644 + }, + { + "epoch": 2.6122406495038515, + "grad_norm": 1.1747702360153198, + "learning_rate": 5.676210607370842e-05, + "loss": 0.3734, + "step": 37645 + }, + { + "epoch": 2.612310040940948, + "grad_norm": 3.179425001144409, + "learning_rate": 5.67570398813708e-05, + "loss": 0.6011, + "step": 37646 + }, + { + "epoch": 2.6123794323780447, + "grad_norm": 1.2869985103607178, + "learning_rate": 5.675197382554619e-05, + "loss": 0.4073, + "step": 37647 + }, + { + "epoch": 2.612448823815141, + "grad_norm": 1.0616908073425293, + "learning_rate": 5.674690790625059e-05, + "loss": 0.3688, + "step": 37648 + }, + { + "epoch": 2.612518215252238, + "grad_norm": 1.3952528238296509, + "learning_rate": 5.674184212349998e-05, + "loss": 0.4865, + "step": 37649 + }, + { + "epoch": 2.6125876066893348, + "grad_norm": 2.6962673664093018, + "learning_rate": 5.673677647731038e-05, + "loss": 0.4375, + "step": 37650 + }, + { + "epoch": 2.612656998126431, + "grad_norm": 1.3694335222244263, + "learning_rate": 5.6731710967697736e-05, + "loss": 0.445, + "step": 37651 + }, + { + "epoch": 2.612726389563528, + "grad_norm": 1.5183299779891968, + "learning_rate": 5.672664559467805e-05, + "loss": 0.4717, + "step": 37652 + }, + { + "epoch": 2.6127957810006244, + "grad_norm": 1.3199453353881836, + "learning_rate": 5.672158035826733e-05, + "loss": 0.4039, + "step": 37653 + }, + { + "epoch": 2.612865172437721, + "grad_norm": 0.9520692229270935, + "learning_rate": 5.6716515258481597e-05, + "loss": 0.4362, + "step": 37654 + }, + { + "epoch": 2.612934563874818, + "grad_norm": 1.158089280128479, + "learning_rate": 5.671145029533676e-05, + "loss": 0.4549, + "step": 37655 + }, + { + "epoch": 2.6130039553119144, + "grad_norm": 1.31282639503479, + "learning_rate": 5.670638546884885e-05, + "loss": 0.4724, + "step": 37656 + }, + { + "epoch": 2.6130733467490113, + "grad_norm": 0.9102343916893005, + "learning_rate": 5.6701320779033874e-05, + "loss": 0.4256, + "step": 37657 + }, + { + "epoch": 2.6131427381861076, + "grad_norm": 0.8184326887130737, + "learning_rate": 5.669625622590783e-05, + "loss": 0.3112, + "step": 37658 + }, + { + "epoch": 2.6132121296232045, + "grad_norm": 1.0087796449661255, + "learning_rate": 5.6691191809486646e-05, + "loss": 0.4853, + "step": 37659 + }, + { + "epoch": 2.6132815210603013, + "grad_norm": 1.345096230506897, + "learning_rate": 5.6686127529786345e-05, + "loss": 0.3281, + "step": 37660 + }, + { + "epoch": 2.6133509124973977, + "grad_norm": 1.0442140102386475, + "learning_rate": 5.6681063386822905e-05, + "loss": 0.5094, + "step": 37661 + }, + { + "epoch": 2.6134203039344945, + "grad_norm": 1.1854099035263062, + "learning_rate": 5.667599938061233e-05, + "loss": 0.4359, + "step": 37662 + }, + { + "epoch": 2.613489695371591, + "grad_norm": 1.2827479839324951, + "learning_rate": 5.6670935511170596e-05, + "loss": 0.4047, + "step": 37663 + }, + { + "epoch": 2.6135590868086878, + "grad_norm": 1.31931471824646, + "learning_rate": 5.6665871778513704e-05, + "loss": 0.4968, + "step": 37664 + }, + { + "epoch": 2.6136284782457846, + "grad_norm": 1.4855244159698486, + "learning_rate": 5.6660808182657654e-05, + "loss": 0.574, + "step": 37665 + }, + { + "epoch": 2.613697869682881, + "grad_norm": 1.5858901739120483, + "learning_rate": 5.6655744723618364e-05, + "loss": 0.3474, + "step": 37666 + }, + { + "epoch": 2.613767261119978, + "grad_norm": 2.048996686935425, + "learning_rate": 5.6650681401411866e-05, + "loss": 0.3537, + "step": 37667 + }, + { + "epoch": 2.613836652557074, + "grad_norm": 2.100525140762329, + "learning_rate": 5.664561821605413e-05, + "loss": 0.5077, + "step": 37668 + }, + { + "epoch": 2.613906043994171, + "grad_norm": 1.973764181137085, + "learning_rate": 5.6640555167561147e-05, + "loss": 0.4675, + "step": 37669 + }, + { + "epoch": 2.613975435431268, + "grad_norm": 0.9122058749198914, + "learning_rate": 5.663549225594891e-05, + "loss": 0.4514, + "step": 37670 + }, + { + "epoch": 2.6140448268683647, + "grad_norm": 1.3839980363845825, + "learning_rate": 5.6630429481233385e-05, + "loss": 0.4617, + "step": 37671 + }, + { + "epoch": 2.614114218305461, + "grad_norm": 1.0202970504760742, + "learning_rate": 5.6625366843430606e-05, + "loss": 0.313, + "step": 37672 + }, + { + "epoch": 2.614183609742558, + "grad_norm": 1.4790786504745483, + "learning_rate": 5.6620304342556477e-05, + "loss": 0.4501, + "step": 37673 + }, + { + "epoch": 2.6142530011796543, + "grad_norm": 1.0847595930099487, + "learning_rate": 5.6615241978626995e-05, + "loss": 0.4762, + "step": 37674 + }, + { + "epoch": 2.614322392616751, + "grad_norm": 1.2891184091567993, + "learning_rate": 5.661017975165819e-05, + "loss": 0.4945, + "step": 37675 + }, + { + "epoch": 2.614391784053848, + "grad_norm": 1.286024808883667, + "learning_rate": 5.6605117661665985e-05, + "loss": 0.4615, + "step": 37676 + }, + { + "epoch": 2.6144611754909444, + "grad_norm": 1.0452038049697876, + "learning_rate": 5.66000557086664e-05, + "loss": 0.333, + "step": 37677 + }, + { + "epoch": 2.614530566928041, + "grad_norm": 1.2590687274932861, + "learning_rate": 5.6594993892675396e-05, + "loss": 0.3313, + "step": 37678 + }, + { + "epoch": 2.6145999583651376, + "grad_norm": 1.2368078231811523, + "learning_rate": 5.658993221370902e-05, + "loss": 0.3695, + "step": 37679 + }, + { + "epoch": 2.6146693498022344, + "grad_norm": 1.3308653831481934, + "learning_rate": 5.658487067178313e-05, + "loss": 0.4409, + "step": 37680 + }, + { + "epoch": 2.6147387412393313, + "grad_norm": 1.8843166828155518, + "learning_rate": 5.657980926691378e-05, + "loss": 0.3789, + "step": 37681 + }, + { + "epoch": 2.6148081326764276, + "grad_norm": 1.3405243158340454, + "learning_rate": 5.657474799911693e-05, + "loss": 0.4628, + "step": 37682 + }, + { + "epoch": 2.6148775241135245, + "grad_norm": 0.9723250865936279, + "learning_rate": 5.6569686868408566e-05, + "loss": 0.4058, + "step": 37683 + }, + { + "epoch": 2.614946915550621, + "grad_norm": 1.0883910655975342, + "learning_rate": 5.6564625874804656e-05, + "loss": 0.2982, + "step": 37684 + }, + { + "epoch": 2.6150163069877177, + "grad_norm": 1.635390281677246, + "learning_rate": 5.6559565018321225e-05, + "loss": 0.3938, + "step": 37685 + }, + { + "epoch": 2.6150856984248145, + "grad_norm": 1.5443556308746338, + "learning_rate": 5.6554504298974176e-05, + "loss": 0.4843, + "step": 37686 + }, + { + "epoch": 2.615155089861911, + "grad_norm": 0.9131660461425781, + "learning_rate": 5.65494437167795e-05, + "loss": 0.3875, + "step": 37687 + }, + { + "epoch": 2.6152244812990078, + "grad_norm": 1.5568419694900513, + "learning_rate": 5.654438327175324e-05, + "loss": 0.4253, + "step": 37688 + }, + { + "epoch": 2.615293872736104, + "grad_norm": 1.256661295890808, + "learning_rate": 5.6539322963911266e-05, + "loss": 0.4603, + "step": 37689 + }, + { + "epoch": 2.615363264173201, + "grad_norm": 1.1016252040863037, + "learning_rate": 5.653426279326963e-05, + "loss": 0.4344, + "step": 37690 + }, + { + "epoch": 2.615432655610298, + "grad_norm": 1.1628543138504028, + "learning_rate": 5.652920275984427e-05, + "loss": 0.607, + "step": 37691 + }, + { + "epoch": 2.615502047047394, + "grad_norm": 1.4116390943527222, + "learning_rate": 5.652414286365121e-05, + "loss": 0.4642, + "step": 37692 + }, + { + "epoch": 2.615571438484491, + "grad_norm": 1.523003339767456, + "learning_rate": 5.651908310470635e-05, + "loss": 0.5732, + "step": 37693 + }, + { + "epoch": 2.6156408299215874, + "grad_norm": 1.027073860168457, + "learning_rate": 5.65140234830257e-05, + "loss": 0.3349, + "step": 37694 + }, + { + "epoch": 2.6157102213586843, + "grad_norm": 2.8912529945373535, + "learning_rate": 5.6508963998625244e-05, + "loss": 0.3727, + "step": 37695 + }, + { + "epoch": 2.615779612795781, + "grad_norm": 0.9506128430366516, + "learning_rate": 5.6503904651520935e-05, + "loss": 0.4229, + "step": 37696 + }, + { + "epoch": 2.615849004232878, + "grad_norm": 1.105878472328186, + "learning_rate": 5.6498845441728744e-05, + "loss": 0.2677, + "step": 37697 + }, + { + "epoch": 2.6159183956699743, + "grad_norm": 0.7897796630859375, + "learning_rate": 5.6493786369264657e-05, + "loss": 0.3054, + "step": 37698 + }, + { + "epoch": 2.615987787107071, + "grad_norm": 1.2581197023391724, + "learning_rate": 5.648872743414467e-05, + "loss": 0.2232, + "step": 37699 + }, + { + "epoch": 2.6160571785441675, + "grad_norm": 1.0675612688064575, + "learning_rate": 5.64836686363847e-05, + "loss": 0.3464, + "step": 37700 + }, + { + "epoch": 2.6161265699812644, + "grad_norm": 1.199649691581726, + "learning_rate": 5.647860997600073e-05, + "loss": 0.5705, + "step": 37701 + }, + { + "epoch": 2.616195961418361, + "grad_norm": 1.395005702972412, + "learning_rate": 5.6473551453008745e-05, + "loss": 0.4021, + "step": 37702 + }, + { + "epoch": 2.6162653528554576, + "grad_norm": 1.512900710105896, + "learning_rate": 5.64684930674247e-05, + "loss": 0.3723, + "step": 37703 + }, + { + "epoch": 2.6163347442925544, + "grad_norm": 0.986237645149231, + "learning_rate": 5.646343481926457e-05, + "loss": 0.3785, + "step": 37704 + }, + { + "epoch": 2.616404135729651, + "grad_norm": 1.0606226921081543, + "learning_rate": 5.645837670854432e-05, + "loss": 0.4353, + "step": 37705 + }, + { + "epoch": 2.6164735271667476, + "grad_norm": 1.3755549192428589, + "learning_rate": 5.645331873527998e-05, + "loss": 0.4792, + "step": 37706 + }, + { + "epoch": 2.6165429186038445, + "grad_norm": 1.2936762571334839, + "learning_rate": 5.6448260899487404e-05, + "loss": 0.5391, + "step": 37707 + }, + { + "epoch": 2.616612310040941, + "grad_norm": 1.4521235227584839, + "learning_rate": 5.6443203201182634e-05, + "loss": 0.597, + "step": 37708 + }, + { + "epoch": 2.6166817014780377, + "grad_norm": 0.8612931966781616, + "learning_rate": 5.6438145640381615e-05, + "loss": 0.3627, + "step": 37709 + }, + { + "epoch": 2.616751092915134, + "grad_norm": 1.6049426794052124, + "learning_rate": 5.643308821710029e-05, + "loss": 0.3836, + "step": 37710 + }, + { + "epoch": 2.616820484352231, + "grad_norm": 1.6287325620651245, + "learning_rate": 5.642803093135468e-05, + "loss": 0.5592, + "step": 37711 + }, + { + "epoch": 2.6168898757893277, + "grad_norm": 1.1014491319656372, + "learning_rate": 5.6422973783160704e-05, + "loss": 0.43, + "step": 37712 + }, + { + "epoch": 2.616959267226424, + "grad_norm": 1.42830491065979, + "learning_rate": 5.64179167725344e-05, + "loss": 0.5123, + "step": 37713 + }, + { + "epoch": 2.617028658663521, + "grad_norm": 1.860878586769104, + "learning_rate": 5.641285989949162e-05, + "loss": 0.3014, + "step": 37714 + }, + { + "epoch": 2.6170980501006174, + "grad_norm": 1.0979293584823608, + "learning_rate": 5.640780316404839e-05, + "loss": 0.5886, + "step": 37715 + }, + { + "epoch": 2.617167441537714, + "grad_norm": 1.2319347858428955, + "learning_rate": 5.640274656622067e-05, + "loss": 0.3777, + "step": 37716 + }, + { + "epoch": 2.617236832974811, + "grad_norm": 2.19502329826355, + "learning_rate": 5.6397690106024425e-05, + "loss": 0.3367, + "step": 37717 + }, + { + "epoch": 2.6173062244119074, + "grad_norm": 1.457053303718567, + "learning_rate": 5.639263378347561e-05, + "loss": 0.5652, + "step": 37718 + }, + { + "epoch": 2.6173756158490042, + "grad_norm": 1.395447850227356, + "learning_rate": 5.6387577598590194e-05, + "loss": 0.5053, + "step": 37719 + }, + { + "epoch": 2.6174450072861006, + "grad_norm": 0.988741934299469, + "learning_rate": 5.638252155138417e-05, + "loss": 0.2967, + "step": 37720 + }, + { + "epoch": 2.6175143987231975, + "grad_norm": 0.9352303743362427, + "learning_rate": 5.637746564187342e-05, + "loss": 0.2568, + "step": 37721 + }, + { + "epoch": 2.6175837901602943, + "grad_norm": 1.0759488344192505, + "learning_rate": 5.637240987007395e-05, + "loss": 0.3411, + "step": 37722 + }, + { + "epoch": 2.617653181597391, + "grad_norm": 0.7353484630584717, + "learning_rate": 5.636735423600177e-05, + "loss": 0.1784, + "step": 37723 + }, + { + "epoch": 2.6177225730344875, + "grad_norm": 1.418583631515503, + "learning_rate": 5.636229873967275e-05, + "loss": 0.4292, + "step": 37724 + }, + { + "epoch": 2.6177919644715844, + "grad_norm": 1.1384804248809814, + "learning_rate": 5.6357243381102885e-05, + "loss": 0.4497, + "step": 37725 + }, + { + "epoch": 2.6178613559086807, + "grad_norm": 1.914247989654541, + "learning_rate": 5.635218816030814e-05, + "loss": 0.7687, + "step": 37726 + }, + { + "epoch": 2.6179307473457776, + "grad_norm": 1.1037379503250122, + "learning_rate": 5.63471330773045e-05, + "loss": 0.3345, + "step": 37727 + }, + { + "epoch": 2.6180001387828744, + "grad_norm": 1.0993846654891968, + "learning_rate": 5.634207813210788e-05, + "loss": 0.3126, + "step": 37728 + }, + { + "epoch": 2.618069530219971, + "grad_norm": 1.2973780632019043, + "learning_rate": 5.6337023324734254e-05, + "loss": 0.3811, + "step": 37729 + }, + { + "epoch": 2.6181389216570676, + "grad_norm": 0.8899520039558411, + "learning_rate": 5.633196865519956e-05, + "loss": 0.3404, + "step": 37730 + }, + { + "epoch": 2.618208313094164, + "grad_norm": 1.1359295845031738, + "learning_rate": 5.632691412351978e-05, + "loss": 0.3135, + "step": 37731 + }, + { + "epoch": 2.618277704531261, + "grad_norm": 1.4066362380981445, + "learning_rate": 5.632185972971087e-05, + "loss": 0.4452, + "step": 37732 + }, + { + "epoch": 2.6183470959683577, + "grad_norm": 1.4827488660812378, + "learning_rate": 5.631680547378878e-05, + "loss": 0.5275, + "step": 37733 + }, + { + "epoch": 2.618416487405454, + "grad_norm": 0.9499058127403259, + "learning_rate": 5.6311751355769516e-05, + "loss": 0.345, + "step": 37734 + }, + { + "epoch": 2.618485878842551, + "grad_norm": 1.0947973728179932, + "learning_rate": 5.630669737566892e-05, + "loss": 0.3308, + "step": 37735 + }, + { + "epoch": 2.6185552702796473, + "grad_norm": 1.3213398456573486, + "learning_rate": 5.630164353350303e-05, + "loss": 0.3823, + "step": 37736 + }, + { + "epoch": 2.618624661716744, + "grad_norm": 0.8512281179428101, + "learning_rate": 5.6296589829287785e-05, + "loss": 0.3064, + "step": 37737 + }, + { + "epoch": 2.618694053153841, + "grad_norm": 1.145331859588623, + "learning_rate": 5.629153626303912e-05, + "loss": 0.3068, + "step": 37738 + }, + { + "epoch": 2.6187634445909374, + "grad_norm": 1.2434768676757812, + "learning_rate": 5.6286482834773014e-05, + "loss": 0.473, + "step": 37739 + }, + { + "epoch": 2.618832836028034, + "grad_norm": 1.187172293663025, + "learning_rate": 5.628142954450541e-05, + "loss": 0.486, + "step": 37740 + }, + { + "epoch": 2.6189022274651306, + "grad_norm": 1.3383337259292603, + "learning_rate": 5.6276376392252294e-05, + "loss": 0.285, + "step": 37741 + }, + { + "epoch": 2.6189716189022274, + "grad_norm": 1.6928547620773315, + "learning_rate": 5.627132337802955e-05, + "loss": 0.394, + "step": 37742 + }, + { + "epoch": 2.6190410103393242, + "grad_norm": 1.5245708227157593, + "learning_rate": 5.6266270501853155e-05, + "loss": 0.3406, + "step": 37743 + }, + { + "epoch": 2.6191104017764206, + "grad_norm": 1.2157269716262817, + "learning_rate": 5.626121776373909e-05, + "loss": 0.3818, + "step": 37744 + }, + { + "epoch": 2.6191797932135175, + "grad_norm": 1.0480575561523438, + "learning_rate": 5.625616516370327e-05, + "loss": 0.3012, + "step": 37745 + }, + { + "epoch": 2.619249184650614, + "grad_norm": 1.471053123474121, + "learning_rate": 5.625111270176167e-05, + "loss": 0.4516, + "step": 37746 + }, + { + "epoch": 2.6193185760877107, + "grad_norm": 1.9738004207611084, + "learning_rate": 5.624606037793023e-05, + "loss": 0.3933, + "step": 37747 + }, + { + "epoch": 2.6193879675248075, + "grad_norm": 1.246358036994934, + "learning_rate": 5.624100819222493e-05, + "loss": 0.3381, + "step": 37748 + }, + { + "epoch": 2.6194573589619043, + "grad_norm": 0.8598490357398987, + "learning_rate": 5.623595614466166e-05, + "loss": 0.4219, + "step": 37749 + }, + { + "epoch": 2.6195267503990007, + "grad_norm": 2.8435511589050293, + "learning_rate": 5.6230904235256396e-05, + "loss": 0.4104, + "step": 37750 + }, + { + "epoch": 2.6195961418360976, + "grad_norm": 1.2939249277114868, + "learning_rate": 5.6225852464025094e-05, + "loss": 0.4395, + "step": 37751 + }, + { + "epoch": 2.619665533273194, + "grad_norm": 1.8511278629302979, + "learning_rate": 5.622080083098369e-05, + "loss": 0.5185, + "step": 37752 + }, + { + "epoch": 2.619734924710291, + "grad_norm": 1.3304815292358398, + "learning_rate": 5.621574933614815e-05, + "loss": 0.4366, + "step": 37753 + }, + { + "epoch": 2.6198043161473876, + "grad_norm": 0.9041262865066528, + "learning_rate": 5.62106979795344e-05, + "loss": 0.2884, + "step": 37754 + }, + { + "epoch": 2.619873707584484, + "grad_norm": 1.3402903079986572, + "learning_rate": 5.620564676115844e-05, + "loss": 0.3784, + "step": 37755 + }, + { + "epoch": 2.619943099021581, + "grad_norm": 1.4584546089172363, + "learning_rate": 5.620059568103613e-05, + "loss": 0.2406, + "step": 37756 + }, + { + "epoch": 2.6200124904586772, + "grad_norm": 1.3073904514312744, + "learning_rate": 5.619554473918345e-05, + "loss": 0.3636, + "step": 37757 + }, + { + "epoch": 2.620081881895774, + "grad_norm": 1.0790060758590698, + "learning_rate": 5.6190493935616396e-05, + "loss": 0.3726, + "step": 37758 + }, + { + "epoch": 2.620151273332871, + "grad_norm": 0.9150481820106506, + "learning_rate": 5.618544327035084e-05, + "loss": 0.2747, + "step": 37759 + }, + { + "epoch": 2.6202206647699673, + "grad_norm": 1.4157172441482544, + "learning_rate": 5.618039274340274e-05, + "loss": 0.5372, + "step": 37760 + }, + { + "epoch": 2.620290056207064, + "grad_norm": 1.0323762893676758, + "learning_rate": 5.617534235478808e-05, + "loss": 0.4165, + "step": 37761 + }, + { + "epoch": 2.6203594476441605, + "grad_norm": 1.313989520072937, + "learning_rate": 5.6170292104522807e-05, + "loss": 0.4149, + "step": 37762 + }, + { + "epoch": 2.6204288390812573, + "grad_norm": 1.3772683143615723, + "learning_rate": 5.616524199262279e-05, + "loss": 0.4498, + "step": 37763 + }, + { + "epoch": 2.620498230518354, + "grad_norm": 1.4069737195968628, + "learning_rate": 5.616019201910403e-05, + "loss": 0.5496, + "step": 37764 + }, + { + "epoch": 2.6205676219554506, + "grad_norm": 1.124793291091919, + "learning_rate": 5.615514218398244e-05, + "loss": 0.2392, + "step": 37765 + }, + { + "epoch": 2.6206370133925474, + "grad_norm": 1.2900587320327759, + "learning_rate": 5.6150092487274e-05, + "loss": 0.3387, + "step": 37766 + }, + { + "epoch": 2.620706404829644, + "grad_norm": 1.2454932928085327, + "learning_rate": 5.614504292899462e-05, + "loss": 0.388, + "step": 37767 + }, + { + "epoch": 2.6207757962667406, + "grad_norm": 1.1231454610824585, + "learning_rate": 5.6139993509160295e-05, + "loss": 0.449, + "step": 37768 + }, + { + "epoch": 2.6208451877038375, + "grad_norm": 1.3242565393447876, + "learning_rate": 5.613494422778687e-05, + "loss": 0.4786, + "step": 37769 + }, + { + "epoch": 2.6209145791409343, + "grad_norm": 0.9434362649917603, + "learning_rate": 5.612989508489035e-05, + "loss": 0.2322, + "step": 37770 + }, + { + "epoch": 2.6209839705780307, + "grad_norm": 1.0716254711151123, + "learning_rate": 5.612484608048666e-05, + "loss": 0.2968, + "step": 37771 + }, + { + "epoch": 2.621053362015127, + "grad_norm": 1.5247060060501099, + "learning_rate": 5.611979721459173e-05, + "loss": 0.5623, + "step": 37772 + }, + { + "epoch": 2.621122753452224, + "grad_norm": 1.4302421808242798, + "learning_rate": 5.611474848722152e-05, + "loss": 0.3658, + "step": 37773 + }, + { + "epoch": 2.6211921448893207, + "grad_norm": 0.946600079536438, + "learning_rate": 5.610969989839196e-05, + "loss": 0.2667, + "step": 37774 + }, + { + "epoch": 2.6212615363264176, + "grad_norm": 1.1385624408721924, + "learning_rate": 5.610465144811902e-05, + "loss": 0.3745, + "step": 37775 + }, + { + "epoch": 2.621330927763514, + "grad_norm": 1.4210830926895142, + "learning_rate": 5.6099603136418555e-05, + "loss": 0.3863, + "step": 37776 + }, + { + "epoch": 2.621400319200611, + "grad_norm": 1.3274198770523071, + "learning_rate": 5.609455496330655e-05, + "loss": 0.458, + "step": 37777 + }, + { + "epoch": 2.621469710637707, + "grad_norm": 1.0429123640060425, + "learning_rate": 5.608950692879895e-05, + "loss": 0.3411, + "step": 37778 + }, + { + "epoch": 2.621539102074804, + "grad_norm": 2.9318814277648926, + "learning_rate": 5.6084459032911685e-05, + "loss": 0.5708, + "step": 37779 + }, + { + "epoch": 2.621608493511901, + "grad_norm": 1.0237056016921997, + "learning_rate": 5.6079411275660687e-05, + "loss": 0.3961, + "step": 37780 + }, + { + "epoch": 2.6216778849489972, + "grad_norm": 1.2223466634750366, + "learning_rate": 5.607436365706189e-05, + "loss": 0.3853, + "step": 37781 + }, + { + "epoch": 2.621747276386094, + "grad_norm": 1.437605857849121, + "learning_rate": 5.606931617713127e-05, + "loss": 0.4487, + "step": 37782 + }, + { + "epoch": 2.6218166678231904, + "grad_norm": 1.7670025825500488, + "learning_rate": 5.606426883588467e-05, + "loss": 0.2801, + "step": 37783 + }, + { + "epoch": 2.6218860592602873, + "grad_norm": 1.3581207990646362, + "learning_rate": 5.6059221633338086e-05, + "loss": 0.3677, + "step": 37784 + }, + { + "epoch": 2.621955450697384, + "grad_norm": 1.1288050413131714, + "learning_rate": 5.6054174569507445e-05, + "loss": 0.4106, + "step": 37785 + }, + { + "epoch": 2.6220248421344805, + "grad_norm": 1.122693419456482, + "learning_rate": 5.604912764440866e-05, + "loss": 0.3565, + "step": 37786 + }, + { + "epoch": 2.6220942335715773, + "grad_norm": 1.433011770248413, + "learning_rate": 5.6044080858057704e-05, + "loss": 0.5226, + "step": 37787 + }, + { + "epoch": 2.6221636250086737, + "grad_norm": 1.4300552606582642, + "learning_rate": 5.603903421047048e-05, + "loss": 0.5079, + "step": 37788 + }, + { + "epoch": 2.6222330164457706, + "grad_norm": 1.3205701112747192, + "learning_rate": 5.603398770166295e-05, + "loss": 0.4479, + "step": 37789 + }, + { + "epoch": 2.6223024078828674, + "grad_norm": 0.9302279353141785, + "learning_rate": 5.602894133165099e-05, + "loss": 0.2741, + "step": 37790 + }, + { + "epoch": 2.622371799319964, + "grad_norm": 1.3562828302383423, + "learning_rate": 5.602389510045055e-05, + "loss": 0.5049, + "step": 37791 + }, + { + "epoch": 2.6224411907570606, + "grad_norm": 0.952068030834198, + "learning_rate": 5.601884900807758e-05, + "loss": 0.2999, + "step": 37792 + }, + { + "epoch": 2.622510582194157, + "grad_norm": 1.6430411338806152, + "learning_rate": 5.6013803054548045e-05, + "loss": 0.5069, + "step": 37793 + }, + { + "epoch": 2.622579973631254, + "grad_norm": 1.9709293842315674, + "learning_rate": 5.600875723987777e-05, + "loss": 0.4153, + "step": 37794 + }, + { + "epoch": 2.6226493650683507, + "grad_norm": 1.8361178636550903, + "learning_rate": 5.6003711564082774e-05, + "loss": 0.3556, + "step": 37795 + }, + { + "epoch": 2.6227187565054475, + "grad_norm": 1.4281954765319824, + "learning_rate": 5.599866602717894e-05, + "loss": 0.3719, + "step": 37796 + }, + { + "epoch": 2.622788147942544, + "grad_norm": 0.9364468455314636, + "learning_rate": 5.599362062918225e-05, + "loss": 0.2575, + "step": 37797 + }, + { + "epoch": 2.6228575393796403, + "grad_norm": 1.0286556482315063, + "learning_rate": 5.598857537010855e-05, + "loss": 0.3076, + "step": 37798 + }, + { + "epoch": 2.622926930816737, + "grad_norm": 1.1036441326141357, + "learning_rate": 5.598353024997383e-05, + "loss": 0.4206, + "step": 37799 + }, + { + "epoch": 2.622996322253834, + "grad_norm": 1.1229808330535889, + "learning_rate": 5.597848526879399e-05, + "loss": 0.291, + "step": 37800 + }, + { + "epoch": 2.623065713690931, + "grad_norm": 1.4078497886657715, + "learning_rate": 5.597344042658495e-05, + "loss": 0.3899, + "step": 37801 + }, + { + "epoch": 2.623135105128027, + "grad_norm": 1.1036771535873413, + "learning_rate": 5.596839572336268e-05, + "loss": 0.3959, + "step": 37802 + }, + { + "epoch": 2.623204496565124, + "grad_norm": 1.0319573879241943, + "learning_rate": 5.5963351159143084e-05, + "loss": 0.2813, + "step": 37803 + }, + { + "epoch": 2.6232738880022204, + "grad_norm": 1.330644130706787, + "learning_rate": 5.5958306733942054e-05, + "loss": 0.3611, + "step": 37804 + }, + { + "epoch": 2.623343279439317, + "grad_norm": 1.231223225593567, + "learning_rate": 5.5953262447775546e-05, + "loss": 0.3695, + "step": 37805 + }, + { + "epoch": 2.623412670876414, + "grad_norm": 1.5354278087615967, + "learning_rate": 5.594821830065946e-05, + "loss": 0.4216, + "step": 37806 + }, + { + "epoch": 2.6234820623135104, + "grad_norm": 1.9005192518234253, + "learning_rate": 5.594317429260976e-05, + "loss": 0.4857, + "step": 37807 + }, + { + "epoch": 2.6235514537506073, + "grad_norm": 0.9357596635818481, + "learning_rate": 5.593813042364233e-05, + "loss": 0.3114, + "step": 37808 + }, + { + "epoch": 2.6236208451877037, + "grad_norm": 1.4598780870437622, + "learning_rate": 5.593308669377312e-05, + "loss": 0.4493, + "step": 37809 + }, + { + "epoch": 2.6236902366248005, + "grad_norm": 1.453399658203125, + "learning_rate": 5.592804310301808e-05, + "loss": 0.294, + "step": 37810 + }, + { + "epoch": 2.6237596280618973, + "grad_norm": 1.2929586172103882, + "learning_rate": 5.5922999651393046e-05, + "loss": 0.4317, + "step": 37811 + }, + { + "epoch": 2.6238290194989937, + "grad_norm": 1.4798249006271362, + "learning_rate": 5.5917956338914005e-05, + "loss": 0.4487, + "step": 37812 + }, + { + "epoch": 2.6238984109360906, + "grad_norm": 1.2487473487854004, + "learning_rate": 5.591291316559686e-05, + "loss": 0.3271, + "step": 37813 + }, + { + "epoch": 2.623967802373187, + "grad_norm": 1.4088771343231201, + "learning_rate": 5.590787013145752e-05, + "loss": 0.3529, + "step": 37814 + }, + { + "epoch": 2.6240371938102838, + "grad_norm": 1.3239279985427856, + "learning_rate": 5.5902827236511924e-05, + "loss": 0.4259, + "step": 37815 + }, + { + "epoch": 2.6241065852473806, + "grad_norm": 1.3251845836639404, + "learning_rate": 5.5897784480775984e-05, + "loss": 0.3753, + "step": 37816 + }, + { + "epoch": 2.624175976684477, + "grad_norm": 0.9135540127754211, + "learning_rate": 5.589274186426566e-05, + "loss": 0.2023, + "step": 37817 + }, + { + "epoch": 2.624245368121574, + "grad_norm": 1.9517301321029663, + "learning_rate": 5.5887699386996805e-05, + "loss": 0.6437, + "step": 37818 + }, + { + "epoch": 2.62431475955867, + "grad_norm": 1.5186196565628052, + "learning_rate": 5.5882657048985354e-05, + "loss": 0.4298, + "step": 37819 + }, + { + "epoch": 2.624384150995767, + "grad_norm": 1.8608170747756958, + "learning_rate": 5.587761485024723e-05, + "loss": 0.3963, + "step": 37820 + }, + { + "epoch": 2.624453542432864, + "grad_norm": 1.5217063426971436, + "learning_rate": 5.587257279079837e-05, + "loss": 0.4655, + "step": 37821 + }, + { + "epoch": 2.6245229338699607, + "grad_norm": 1.0173712968826294, + "learning_rate": 5.5867530870654686e-05, + "loss": 0.412, + "step": 37822 + }, + { + "epoch": 2.624592325307057, + "grad_norm": 1.66558039188385, + "learning_rate": 5.586248908983207e-05, + "loss": 0.3475, + "step": 37823 + }, + { + "epoch": 2.6246617167441535, + "grad_norm": 1.1936781406402588, + "learning_rate": 5.585744744834649e-05, + "loss": 0.357, + "step": 37824 + }, + { + "epoch": 2.6247311081812503, + "grad_norm": 1.0757861137390137, + "learning_rate": 5.585240594621379e-05, + "loss": 0.4696, + "step": 37825 + }, + { + "epoch": 2.624800499618347, + "grad_norm": 0.9624986052513123, + "learning_rate": 5.584736458344993e-05, + "loss": 0.3381, + "step": 37826 + }, + { + "epoch": 2.624869891055444, + "grad_norm": 1.1101233959197998, + "learning_rate": 5.584232336007083e-05, + "loss": 0.4205, + "step": 37827 + }, + { + "epoch": 2.6249392824925404, + "grad_norm": 1.1989251375198364, + "learning_rate": 5.5837282276092414e-05, + "loss": 0.3149, + "step": 37828 + }, + { + "epoch": 2.625008673929637, + "grad_norm": 1.6641346216201782, + "learning_rate": 5.5832241331530535e-05, + "loss": 0.5103, + "step": 37829 + }, + { + "epoch": 2.6250780653667336, + "grad_norm": 1.3876625299453735, + "learning_rate": 5.582720052640115e-05, + "loss": 0.384, + "step": 37830 + }, + { + "epoch": 2.6251474568038304, + "grad_norm": 1.150412917137146, + "learning_rate": 5.5822159860720204e-05, + "loss": 0.3087, + "step": 37831 + }, + { + "epoch": 2.6252168482409273, + "grad_norm": 1.1911052465438843, + "learning_rate": 5.581711933450353e-05, + "loss": 0.5025, + "step": 37832 + }, + { + "epoch": 2.6252862396780237, + "grad_norm": 0.8076039552688599, + "learning_rate": 5.58120789477671e-05, + "loss": 0.2265, + "step": 37833 + }, + { + "epoch": 2.6253556311151205, + "grad_norm": 0.8851074576377869, + "learning_rate": 5.5807038700526795e-05, + "loss": 0.251, + "step": 37834 + }, + { + "epoch": 2.625425022552217, + "grad_norm": 1.4499493837356567, + "learning_rate": 5.5801998592798556e-05, + "loss": 0.4802, + "step": 37835 + }, + { + "epoch": 2.6254944139893137, + "grad_norm": 1.1779931783676147, + "learning_rate": 5.579695862459826e-05, + "loss": 0.3021, + "step": 37836 + }, + { + "epoch": 2.6255638054264105, + "grad_norm": 1.2274060249328613, + "learning_rate": 5.579191879594186e-05, + "loss": 0.3626, + "step": 37837 + }, + { + "epoch": 2.625633196863507, + "grad_norm": 2.1355602741241455, + "learning_rate": 5.578687910684526e-05, + "loss": 0.4736, + "step": 37838 + }, + { + "epoch": 2.6257025883006038, + "grad_norm": 1.6176694631576538, + "learning_rate": 5.578183955732432e-05, + "loss": 0.4386, + "step": 37839 + }, + { + "epoch": 2.6257719797377, + "grad_norm": 1.038016676902771, + "learning_rate": 5.577680014739499e-05, + "loss": 0.3324, + "step": 37840 + }, + { + "epoch": 2.625841371174797, + "grad_norm": 1.2196378707885742, + "learning_rate": 5.5771760877073165e-05, + "loss": 0.3539, + "step": 37841 + }, + { + "epoch": 2.625910762611894, + "grad_norm": 1.3153939247131348, + "learning_rate": 5.5766721746374764e-05, + "loss": 0.3308, + "step": 37842 + }, + { + "epoch": 2.62598015404899, + "grad_norm": 1.0122249126434326, + "learning_rate": 5.576168275531568e-05, + "loss": 0.4491, + "step": 37843 + }, + { + "epoch": 2.626049545486087, + "grad_norm": 1.380637288093567, + "learning_rate": 5.5756643903911845e-05, + "loss": 0.5161, + "step": 37844 + }, + { + "epoch": 2.6261189369231834, + "grad_norm": 0.8924541473388672, + "learning_rate": 5.575160519217919e-05, + "loss": 0.4031, + "step": 37845 + }, + { + "epoch": 2.6261883283602803, + "grad_norm": 1.001431941986084, + "learning_rate": 5.5746566620133535e-05, + "loss": 0.4102, + "step": 37846 + }, + { + "epoch": 2.626257719797377, + "grad_norm": 2.928248405456543, + "learning_rate": 5.5741528187790835e-05, + "loss": 0.5974, + "step": 37847 + }, + { + "epoch": 2.626327111234474, + "grad_norm": 1.2177897691726685, + "learning_rate": 5.5736489895167e-05, + "loss": 0.5716, + "step": 37848 + }, + { + "epoch": 2.6263965026715703, + "grad_norm": 1.4560080766677856, + "learning_rate": 5.573145174227794e-05, + "loss": 0.4872, + "step": 37849 + }, + { + "epoch": 2.626465894108667, + "grad_norm": 1.2408639192581177, + "learning_rate": 5.572641372913954e-05, + "loss": 0.4192, + "step": 37850 + }, + { + "epoch": 2.6265352855457635, + "grad_norm": 1.6540405750274658, + "learning_rate": 5.572137585576776e-05, + "loss": 0.49, + "step": 37851 + }, + { + "epoch": 2.6266046769828604, + "grad_norm": 1.0861115455627441, + "learning_rate": 5.57163381221784e-05, + "loss": 0.3589, + "step": 37852 + }, + { + "epoch": 2.626674068419957, + "grad_norm": 1.010433316230774, + "learning_rate": 5.571130052838744e-05, + "loss": 0.3154, + "step": 37853 + }, + { + "epoch": 2.6267434598570536, + "grad_norm": 1.1002211570739746, + "learning_rate": 5.570626307441077e-05, + "loss": 0.4434, + "step": 37854 + }, + { + "epoch": 2.6268128512941504, + "grad_norm": 1.1210590600967407, + "learning_rate": 5.5701225760264284e-05, + "loss": 0.3948, + "step": 37855 + }, + { + "epoch": 2.626882242731247, + "grad_norm": 1.0256261825561523, + "learning_rate": 5.5696188585963885e-05, + "loss": 0.3258, + "step": 37856 + }, + { + "epoch": 2.6269516341683437, + "grad_norm": 1.3386118412017822, + "learning_rate": 5.569115155152549e-05, + "loss": 0.2713, + "step": 37857 + }, + { + "epoch": 2.6270210256054405, + "grad_norm": 1.6720560789108276, + "learning_rate": 5.5686114656965015e-05, + "loss": 0.4226, + "step": 37858 + }, + { + "epoch": 2.627090417042537, + "grad_norm": 1.454484462738037, + "learning_rate": 5.56810779022983e-05, + "loss": 0.472, + "step": 37859 + }, + { + "epoch": 2.6271598084796337, + "grad_norm": 1.424948811531067, + "learning_rate": 5.567604128754127e-05, + "loss": 0.456, + "step": 37860 + }, + { + "epoch": 2.62722919991673, + "grad_norm": 1.1594505310058594, + "learning_rate": 5.567100481270986e-05, + "loss": 0.3833, + "step": 37861 + }, + { + "epoch": 2.627298591353827, + "grad_norm": 1.0964146852493286, + "learning_rate": 5.5665968477819916e-05, + "loss": 0.3452, + "step": 37862 + }, + { + "epoch": 2.6273679827909238, + "grad_norm": 1.3755712509155273, + "learning_rate": 5.5660932282887424e-05, + "loss": 0.3558, + "step": 37863 + }, + { + "epoch": 2.62743737422802, + "grad_norm": 1.678002119064331, + "learning_rate": 5.565589622792817e-05, + "loss": 0.3898, + "step": 37864 + }, + { + "epoch": 2.627506765665117, + "grad_norm": 1.758205533027649, + "learning_rate": 5.565086031295812e-05, + "loss": 0.5637, + "step": 37865 + }, + { + "epoch": 2.6275761571022134, + "grad_norm": 1.4535902738571167, + "learning_rate": 5.564582453799321e-05, + "loss": 0.5664, + "step": 37866 + }, + { + "epoch": 2.62764554853931, + "grad_norm": 1.2610951662063599, + "learning_rate": 5.5640788903049226e-05, + "loss": 0.4734, + "step": 37867 + }, + { + "epoch": 2.627714939976407, + "grad_norm": 1.0667580366134644, + "learning_rate": 5.563575340814213e-05, + "loss": 0.4572, + "step": 37868 + }, + { + "epoch": 2.6277843314135034, + "grad_norm": 1.211196780204773, + "learning_rate": 5.563071805328781e-05, + "loss": 0.3383, + "step": 37869 + }, + { + "epoch": 2.6278537228506003, + "grad_norm": 1.0894280672073364, + "learning_rate": 5.5625682838502156e-05, + "loss": 0.3083, + "step": 37870 + }, + { + "epoch": 2.6279231142876966, + "grad_norm": 1.4405643939971924, + "learning_rate": 5.5620647763801094e-05, + "loss": 0.266, + "step": 37871 + }, + { + "epoch": 2.6279925057247935, + "grad_norm": 0.9738864302635193, + "learning_rate": 5.5615612829200535e-05, + "loss": 0.3013, + "step": 37872 + }, + { + "epoch": 2.6280618971618903, + "grad_norm": 1.3821207284927368, + "learning_rate": 5.561057803471629e-05, + "loss": 0.3416, + "step": 37873 + }, + { + "epoch": 2.628131288598987, + "grad_norm": 1.4166899919509888, + "learning_rate": 5.56055433803643e-05, + "loss": 0.5105, + "step": 37874 + }, + { + "epoch": 2.6282006800360835, + "grad_norm": 0.9530677795410156, + "learning_rate": 5.560050886616046e-05, + "loss": 0.2512, + "step": 37875 + }, + { + "epoch": 2.6282700714731804, + "grad_norm": 0.9274014234542847, + "learning_rate": 5.5595474492120656e-05, + "loss": 0.3848, + "step": 37876 + }, + { + "epoch": 2.6283394629102768, + "grad_norm": 1.541263461112976, + "learning_rate": 5.559044025826079e-05, + "loss": 0.3825, + "step": 37877 + }, + { + "epoch": 2.6284088543473736, + "grad_norm": 1.129640817642212, + "learning_rate": 5.558540616459675e-05, + "loss": 0.3416, + "step": 37878 + }, + { + "epoch": 2.6284782457844704, + "grad_norm": 1.8488754034042358, + "learning_rate": 5.558037221114448e-05, + "loss": 0.3919, + "step": 37879 + }, + { + "epoch": 2.628547637221567, + "grad_norm": 1.9121575355529785, + "learning_rate": 5.5575338397919775e-05, + "loss": 0.518, + "step": 37880 + }, + { + "epoch": 2.6286170286586636, + "grad_norm": 0.8775162696838379, + "learning_rate": 5.557030472493858e-05, + "loss": 0.3399, + "step": 37881 + }, + { + "epoch": 2.62868642009576, + "grad_norm": 1.2340646982192993, + "learning_rate": 5.5565271192216775e-05, + "loss": 0.374, + "step": 37882 + }, + { + "epoch": 2.628755811532857, + "grad_norm": 1.0521585941314697, + "learning_rate": 5.556023779977024e-05, + "loss": 0.333, + "step": 37883 + }, + { + "epoch": 2.6288252029699537, + "grad_norm": 0.8391807079315186, + "learning_rate": 5.5555204547614895e-05, + "loss": 0.3247, + "step": 37884 + }, + { + "epoch": 2.62889459440705, + "grad_norm": 1.0626047849655151, + "learning_rate": 5.55501714357666e-05, + "loss": 0.3343, + "step": 37885 + }, + { + "epoch": 2.628963985844147, + "grad_norm": 1.167834997177124, + "learning_rate": 5.554513846424131e-05, + "loss": 0.3819, + "step": 37886 + }, + { + "epoch": 2.6290333772812433, + "grad_norm": 1.4419505596160889, + "learning_rate": 5.554010563305481e-05, + "loss": 0.3531, + "step": 37887 + }, + { + "epoch": 2.62910276871834, + "grad_norm": 1.383039116859436, + "learning_rate": 5.5535072942223045e-05, + "loss": 0.3895, + "step": 37888 + }, + { + "epoch": 2.629172160155437, + "grad_norm": 1.5003643035888672, + "learning_rate": 5.553004039176188e-05, + "loss": 0.5771, + "step": 37889 + }, + { + "epoch": 2.6292415515925334, + "grad_norm": 1.1575629711151123, + "learning_rate": 5.552500798168724e-05, + "loss": 0.2455, + "step": 37890 + }, + { + "epoch": 2.62931094302963, + "grad_norm": 1.0230789184570312, + "learning_rate": 5.551997571201497e-05, + "loss": 0.4294, + "step": 37891 + }, + { + "epoch": 2.6293803344667266, + "grad_norm": 1.0173311233520508, + "learning_rate": 5.551494358276099e-05, + "loss": 0.3856, + "step": 37892 + }, + { + "epoch": 2.6294497259038234, + "grad_norm": 1.8817058801651, + "learning_rate": 5.550991159394121e-05, + "loss": 0.4119, + "step": 37893 + }, + { + "epoch": 2.6295191173409203, + "grad_norm": 1.7177029848098755, + "learning_rate": 5.550487974557144e-05, + "loss": 0.406, + "step": 37894 + }, + { + "epoch": 2.6295885087780166, + "grad_norm": 1.0077804327011108, + "learning_rate": 5.549984803766759e-05, + "loss": 0.3818, + "step": 37895 + }, + { + "epoch": 2.6296579002151135, + "grad_norm": 1.3872920274734497, + "learning_rate": 5.549481647024557e-05, + "loss": 0.4857, + "step": 37896 + }, + { + "epoch": 2.62972729165221, + "grad_norm": 1.494964599609375, + "learning_rate": 5.5489785043321274e-05, + "loss": 0.4847, + "step": 37897 + }, + { + "epoch": 2.6297966830893067, + "grad_norm": 2.5253374576568604, + "learning_rate": 5.548475375691053e-05, + "loss": 0.6411, + "step": 37898 + }, + { + "epoch": 2.6298660745264035, + "grad_norm": 0.9553446173667908, + "learning_rate": 5.547972261102926e-05, + "loss": 0.2639, + "step": 37899 + }, + { + "epoch": 2.6299354659635004, + "grad_norm": 1.8040313720703125, + "learning_rate": 5.547469160569334e-05, + "loss": 0.3821, + "step": 37900 + }, + { + "epoch": 2.6300048574005968, + "grad_norm": 1.436246395111084, + "learning_rate": 5.5469660740918685e-05, + "loss": 0.4133, + "step": 37901 + }, + { + "epoch": 2.6300742488376936, + "grad_norm": 1.321454644203186, + "learning_rate": 5.546463001672111e-05, + "loss": 0.3943, + "step": 37902 + }, + { + "epoch": 2.63014364027479, + "grad_norm": 1.0424753427505493, + "learning_rate": 5.545959943311654e-05, + "loss": 0.2387, + "step": 37903 + }, + { + "epoch": 2.630213031711887, + "grad_norm": 1.3751040697097778, + "learning_rate": 5.545456899012082e-05, + "loss": 0.33, + "step": 37904 + }, + { + "epoch": 2.6302824231489836, + "grad_norm": 1.0184041261672974, + "learning_rate": 5.5449538687749895e-05, + "loss": 0.311, + "step": 37905 + }, + { + "epoch": 2.63035181458608, + "grad_norm": 1.1815435886383057, + "learning_rate": 5.544450852601958e-05, + "loss": 0.4123, + "step": 37906 + }, + { + "epoch": 2.630421206023177, + "grad_norm": 1.1468791961669922, + "learning_rate": 5.543947850494583e-05, + "loss": 0.326, + "step": 37907 + }, + { + "epoch": 2.6304905974602732, + "grad_norm": 1.1876307725906372, + "learning_rate": 5.543444862454443e-05, + "loss": 0.3885, + "step": 37908 + }, + { + "epoch": 2.63055998889737, + "grad_norm": 1.3810186386108398, + "learning_rate": 5.542941888483131e-05, + "loss": 0.4576, + "step": 37909 + }, + { + "epoch": 2.630629380334467, + "grad_norm": 1.4423433542251587, + "learning_rate": 5.5424389285822344e-05, + "loss": 0.415, + "step": 37910 + }, + { + "epoch": 2.6306987717715633, + "grad_norm": 0.9673619270324707, + "learning_rate": 5.541935982753341e-05, + "loss": 0.344, + "step": 37911 + }, + { + "epoch": 2.63076816320866, + "grad_norm": 1.079947829246521, + "learning_rate": 5.5414330509980395e-05, + "loss": 0.3525, + "step": 37912 + }, + { + "epoch": 2.6308375546457565, + "grad_norm": 1.2852131128311157, + "learning_rate": 5.5409301333179144e-05, + "loss": 0.427, + "step": 37913 + }, + { + "epoch": 2.6309069460828534, + "grad_norm": 1.333640694618225, + "learning_rate": 5.540427229714562e-05, + "loss": 0.4067, + "step": 37914 + }, + { + "epoch": 2.63097633751995, + "grad_norm": 1.075456976890564, + "learning_rate": 5.539924340189557e-05, + "loss": 0.2808, + "step": 37915 + }, + { + "epoch": 2.6310457289570466, + "grad_norm": 1.6059757471084595, + "learning_rate": 5.5394214647444945e-05, + "loss": 0.6323, + "step": 37916 + }, + { + "epoch": 2.6311151203941434, + "grad_norm": 1.0355679988861084, + "learning_rate": 5.53891860338096e-05, + "loss": 0.2869, + "step": 37917 + }, + { + "epoch": 2.63118451183124, + "grad_norm": 1.634059190750122, + "learning_rate": 5.538415756100543e-05, + "loss": 0.5151, + "step": 37918 + }, + { + "epoch": 2.6312539032683366, + "grad_norm": 1.1363472938537598, + "learning_rate": 5.537912922904831e-05, + "loss": 0.3728, + "step": 37919 + }, + { + "epoch": 2.6313232947054335, + "grad_norm": 1.248334527015686, + "learning_rate": 5.537410103795409e-05, + "loss": 0.4481, + "step": 37920 + }, + { + "epoch": 2.63139268614253, + "grad_norm": 1.8169177770614624, + "learning_rate": 5.536907298773868e-05, + "loss": 0.5497, + "step": 37921 + }, + { + "epoch": 2.6314620775796267, + "grad_norm": 1.0716081857681274, + "learning_rate": 5.53640450784179e-05, + "loss": 0.3299, + "step": 37922 + }, + { + "epoch": 2.631531469016723, + "grad_norm": 1.4452823400497437, + "learning_rate": 5.535901731000766e-05, + "loss": 0.3493, + "step": 37923 + }, + { + "epoch": 2.63160086045382, + "grad_norm": 0.8641960024833679, + "learning_rate": 5.535398968252381e-05, + "loss": 0.2811, + "step": 37924 + }, + { + "epoch": 2.6316702518909167, + "grad_norm": 2.080167770385742, + "learning_rate": 5.5348962195982246e-05, + "loss": 0.618, + "step": 37925 + }, + { + "epoch": 2.6317396433280136, + "grad_norm": 1.1837068796157837, + "learning_rate": 5.5343934850398816e-05, + "loss": 0.3571, + "step": 37926 + }, + { + "epoch": 2.63180903476511, + "grad_norm": 1.3623270988464355, + "learning_rate": 5.5338907645789465e-05, + "loss": 0.3194, + "step": 37927 + }, + { + "epoch": 2.631878426202207, + "grad_norm": 1.1964932680130005, + "learning_rate": 5.5333880582169926e-05, + "loss": 0.4257, + "step": 37928 + }, + { + "epoch": 2.631947817639303, + "grad_norm": 1.3503206968307495, + "learning_rate": 5.532885365955618e-05, + "loss": 0.4439, + "step": 37929 + }, + { + "epoch": 2.6320172090764, + "grad_norm": 1.2016609907150269, + "learning_rate": 5.5323826877964034e-05, + "loss": 0.4604, + "step": 37930 + }, + { + "epoch": 2.632086600513497, + "grad_norm": 1.1646296977996826, + "learning_rate": 5.5318800237409386e-05, + "loss": 0.3425, + "step": 37931 + }, + { + "epoch": 2.6321559919505932, + "grad_norm": 1.0349009037017822, + "learning_rate": 5.531377373790816e-05, + "loss": 0.3947, + "step": 37932 + }, + { + "epoch": 2.63222538338769, + "grad_norm": 1.0576893091201782, + "learning_rate": 5.530874737947611e-05, + "loss": 0.4288, + "step": 37933 + }, + { + "epoch": 2.6322947748247865, + "grad_norm": 0.9899251461029053, + "learning_rate": 5.530372116212916e-05, + "loss": 0.2493, + "step": 37934 + }, + { + "epoch": 2.6323641662618833, + "grad_norm": 0.9180014133453369, + "learning_rate": 5.529869508588318e-05, + "loss": 0.3071, + "step": 37935 + }, + { + "epoch": 2.63243355769898, + "grad_norm": 1.1103397607803345, + "learning_rate": 5.529366915075407e-05, + "loss": 0.3085, + "step": 37936 + }, + { + "epoch": 2.6325029491360765, + "grad_norm": 0.7056336402893066, + "learning_rate": 5.528864335675763e-05, + "loss": 0.2144, + "step": 37937 + }, + { + "epoch": 2.6325723405731734, + "grad_norm": 1.2872453927993774, + "learning_rate": 5.528361770390974e-05, + "loss": 0.3673, + "step": 37938 + }, + { + "epoch": 2.6326417320102697, + "grad_norm": 1.1565766334533691, + "learning_rate": 5.527859219222629e-05, + "loss": 0.3625, + "step": 37939 + }, + { + "epoch": 2.6327111234473666, + "grad_norm": 1.4645026922225952, + "learning_rate": 5.5273566821723134e-05, + "loss": 0.4193, + "step": 37940 + }, + { + "epoch": 2.6327805148844634, + "grad_norm": 1.1142687797546387, + "learning_rate": 5.526854159241618e-05, + "loss": 0.3315, + "step": 37941 + }, + { + "epoch": 2.63284990632156, + "grad_norm": 0.9453916549682617, + "learning_rate": 5.52635165043212e-05, + "loss": 0.3065, + "step": 37942 + }, + { + "epoch": 2.6329192977586566, + "grad_norm": 1.3104323148727417, + "learning_rate": 5.525849155745413e-05, + "loss": 0.5033, + "step": 37943 + }, + { + "epoch": 2.632988689195753, + "grad_norm": 1.137142539024353, + "learning_rate": 5.525346675183078e-05, + "loss": 0.4044, + "step": 37944 + }, + { + "epoch": 2.63305808063285, + "grad_norm": 3.175759792327881, + "learning_rate": 5.524844208746707e-05, + "loss": 0.4678, + "step": 37945 + }, + { + "epoch": 2.6331274720699467, + "grad_norm": 1.1427805423736572, + "learning_rate": 5.524341756437883e-05, + "loss": 0.4694, + "step": 37946 + }, + { + "epoch": 2.633196863507043, + "grad_norm": 1.3329617977142334, + "learning_rate": 5.5238393182581926e-05, + "loss": 0.3834, + "step": 37947 + }, + { + "epoch": 2.63326625494414, + "grad_norm": 0.918407142162323, + "learning_rate": 5.5233368942092254e-05, + "loss": 0.2986, + "step": 37948 + }, + { + "epoch": 2.6333356463812363, + "grad_norm": 1.127662181854248, + "learning_rate": 5.5228344842925625e-05, + "loss": 0.4134, + "step": 37949 + }, + { + "epoch": 2.633405037818333, + "grad_norm": 1.1312968730926514, + "learning_rate": 5.5223320885097895e-05, + "loss": 0.2887, + "step": 37950 + }, + { + "epoch": 2.63347442925543, + "grad_norm": 1.1622697114944458, + "learning_rate": 5.521829706862496e-05, + "loss": 0.3248, + "step": 37951 + }, + { + "epoch": 2.633543820692527, + "grad_norm": 1.7330831289291382, + "learning_rate": 5.5213273393522646e-05, + "loss": 0.4526, + "step": 37952 + }, + { + "epoch": 2.633613212129623, + "grad_norm": 1.7758573293685913, + "learning_rate": 5.520824985980686e-05, + "loss": 0.6434, + "step": 37953 + }, + { + "epoch": 2.63368260356672, + "grad_norm": 1.3306043148040771, + "learning_rate": 5.5203226467493406e-05, + "loss": 0.3783, + "step": 37954 + }, + { + "epoch": 2.6337519950038164, + "grad_norm": 1.166420817375183, + "learning_rate": 5.519820321659822e-05, + "loss": 0.3832, + "step": 37955 + }, + { + "epoch": 2.6338213864409132, + "grad_norm": 0.9740299582481384, + "learning_rate": 5.519318010713707e-05, + "loss": 0.3243, + "step": 37956 + }, + { + "epoch": 2.63389077787801, + "grad_norm": 1.3675206899642944, + "learning_rate": 5.518815713912584e-05, + "loss": 0.4089, + "step": 37957 + }, + { + "epoch": 2.6339601693151065, + "grad_norm": 0.8738727569580078, + "learning_rate": 5.518313431258042e-05, + "loss": 0.2898, + "step": 37958 + }, + { + "epoch": 2.6340295607522033, + "grad_norm": 1.0938293933868408, + "learning_rate": 5.5178111627516623e-05, + "loss": 0.4303, + "step": 37959 + }, + { + "epoch": 2.6340989521892997, + "grad_norm": 1.693780779838562, + "learning_rate": 5.517308908395035e-05, + "loss": 0.4129, + "step": 37960 + }, + { + "epoch": 2.6341683436263965, + "grad_norm": 1.12035071849823, + "learning_rate": 5.516806668189742e-05, + "loss": 0.3052, + "step": 37961 + }, + { + "epoch": 2.6342377350634933, + "grad_norm": 1.1926385164260864, + "learning_rate": 5.5163044421373745e-05, + "loss": 0.3066, + "step": 37962 + }, + { + "epoch": 2.6343071265005897, + "grad_norm": 1.3615193367004395, + "learning_rate": 5.515802230239508e-05, + "loss": 0.509, + "step": 37963 + }, + { + "epoch": 2.6343765179376866, + "grad_norm": 0.727400004863739, + "learning_rate": 5.515300032497737e-05, + "loss": 0.2771, + "step": 37964 + }, + { + "epoch": 2.634445909374783, + "grad_norm": 0.9578968286514282, + "learning_rate": 5.5147978489136395e-05, + "loss": 0.3534, + "step": 37965 + }, + { + "epoch": 2.63451530081188, + "grad_norm": 1.4399811029434204, + "learning_rate": 5.514295679488808e-05, + "loss": 0.384, + "step": 37966 + }, + { + "epoch": 2.6345846922489766, + "grad_norm": 1.1450601816177368, + "learning_rate": 5.5137935242248274e-05, + "loss": 0.4656, + "step": 37967 + }, + { + "epoch": 2.634654083686073, + "grad_norm": 0.7644761204719543, + "learning_rate": 5.513291383123276e-05, + "loss": 0.3035, + "step": 37968 + }, + { + "epoch": 2.63472347512317, + "grad_norm": 1.0967825651168823, + "learning_rate": 5.512789256185744e-05, + "loss": 0.2943, + "step": 37969 + }, + { + "epoch": 2.6347928665602662, + "grad_norm": 1.2696185111999512, + "learning_rate": 5.5122871434138146e-05, + "loss": 0.4523, + "step": 37970 + }, + { + "epoch": 2.634862257997363, + "grad_norm": 1.3708980083465576, + "learning_rate": 5.511785044809079e-05, + "loss": 0.3375, + "step": 37971 + }, + { + "epoch": 2.63493164943446, + "grad_norm": 1.5801197290420532, + "learning_rate": 5.511282960373112e-05, + "loss": 0.5275, + "step": 37972 + }, + { + "epoch": 2.6350010408715563, + "grad_norm": 1.1780190467834473, + "learning_rate": 5.5107808901075055e-05, + "loss": 0.5356, + "step": 37973 + }, + { + "epoch": 2.635070432308653, + "grad_norm": 1.7667001485824585, + "learning_rate": 5.5102788340138424e-05, + "loss": 0.6101, + "step": 37974 + }, + { + "epoch": 2.6351398237457495, + "grad_norm": 1.1785415410995483, + "learning_rate": 5.509776792093708e-05, + "loss": 0.4656, + "step": 37975 + }, + { + "epoch": 2.6352092151828463, + "grad_norm": 1.5279418230056763, + "learning_rate": 5.509274764348691e-05, + "loss": 0.4052, + "step": 37976 + }, + { + "epoch": 2.635278606619943, + "grad_norm": 1.0943949222564697, + "learning_rate": 5.508772750780369e-05, + "loss": 0.4128, + "step": 37977 + }, + { + "epoch": 2.63534799805704, + "grad_norm": 1.1651841402053833, + "learning_rate": 5.50827075139033e-05, + "loss": 0.4653, + "step": 37978 + }, + { + "epoch": 2.6354173894941364, + "grad_norm": 1.246625304222107, + "learning_rate": 5.5077687661801594e-05, + "loss": 0.3592, + "step": 37979 + }, + { + "epoch": 2.6354867809312332, + "grad_norm": 1.3433932065963745, + "learning_rate": 5.50726679515144e-05, + "loss": 0.3691, + "step": 37980 + }, + { + "epoch": 2.6355561723683296, + "grad_norm": 1.0453673601150513, + "learning_rate": 5.5067648383057605e-05, + "loss": 0.3843, + "step": 37981 + }, + { + "epoch": 2.6356255638054265, + "grad_norm": 0.9697123169898987, + "learning_rate": 5.5062628956447005e-05, + "loss": 0.3555, + "step": 37982 + }, + { + "epoch": 2.6356949552425233, + "grad_norm": 1.1081116199493408, + "learning_rate": 5.505760967169853e-05, + "loss": 0.5029, + "step": 37983 + }, + { + "epoch": 2.6357643466796197, + "grad_norm": 1.049638032913208, + "learning_rate": 5.505259052882793e-05, + "loss": 0.3257, + "step": 37984 + }, + { + "epoch": 2.6358337381167165, + "grad_norm": 1.7281570434570312, + "learning_rate": 5.504757152785107e-05, + "loss": 0.4871, + "step": 37985 + }, + { + "epoch": 2.635903129553813, + "grad_norm": 2.2118473052978516, + "learning_rate": 5.504255266878383e-05, + "loss": 0.4429, + "step": 37986 + }, + { + "epoch": 2.6359725209909097, + "grad_norm": 1.1554629802703857, + "learning_rate": 5.503753395164203e-05, + "loss": 0.3408, + "step": 37987 + }, + { + "epoch": 2.6360419124280066, + "grad_norm": 1.2265616655349731, + "learning_rate": 5.503251537644152e-05, + "loss": 0.4982, + "step": 37988 + }, + { + "epoch": 2.636111303865103, + "grad_norm": 1.1663506031036377, + "learning_rate": 5.502749694319813e-05, + "loss": 0.3991, + "step": 37989 + }, + { + "epoch": 2.6361806953022, + "grad_norm": 1.2795350551605225, + "learning_rate": 5.5022478651927755e-05, + "loss": 0.3516, + "step": 37990 + }, + { + "epoch": 2.636250086739296, + "grad_norm": 0.9977702498435974, + "learning_rate": 5.5017460502646166e-05, + "loss": 0.3501, + "step": 37991 + }, + { + "epoch": 2.636319478176393, + "grad_norm": 1.1966360807418823, + "learning_rate": 5.501244249536923e-05, + "loss": 0.3746, + "step": 37992 + }, + { + "epoch": 2.63638886961349, + "grad_norm": 1.3001914024353027, + "learning_rate": 5.50074246301128e-05, + "loss": 0.311, + "step": 37993 + }, + { + "epoch": 2.6364582610505862, + "grad_norm": 1.3987879753112793, + "learning_rate": 5.5002406906892715e-05, + "loss": 0.5599, + "step": 37994 + }, + { + "epoch": 2.636527652487683, + "grad_norm": 1.2632684707641602, + "learning_rate": 5.4997389325724805e-05, + "loss": 0.4386, + "step": 37995 + }, + { + "epoch": 2.6365970439247794, + "grad_norm": 1.547214388847351, + "learning_rate": 5.4992371886624907e-05, + "loss": 0.3989, + "step": 37996 + }, + { + "epoch": 2.6366664353618763, + "grad_norm": 1.6056311130523682, + "learning_rate": 5.498735458960891e-05, + "loss": 0.3689, + "step": 37997 + }, + { + "epoch": 2.636735826798973, + "grad_norm": 1.149048089981079, + "learning_rate": 5.498233743469258e-05, + "loss": 0.5119, + "step": 37998 + }, + { + "epoch": 2.6368052182360695, + "grad_norm": 1.0740673542022705, + "learning_rate": 5.4977320421891784e-05, + "loss": 0.4007, + "step": 37999 + }, + { + "epoch": 2.6368746096731663, + "grad_norm": 1.8945097923278809, + "learning_rate": 5.497230355122236e-05, + "loss": 0.5408, + "step": 38000 + }, + { + "epoch": 2.6369440011102627, + "grad_norm": 1.233897089958191, + "learning_rate": 5.496728682270016e-05, + "loss": 0.4162, + "step": 38001 + }, + { + "epoch": 2.6370133925473596, + "grad_norm": 1.0817768573760986, + "learning_rate": 5.496227023634104e-05, + "loss": 0.4427, + "step": 38002 + }, + { + "epoch": 2.6370827839844564, + "grad_norm": 1.2502130270004272, + "learning_rate": 5.4957253792160765e-05, + "loss": 0.2879, + "step": 38003 + }, + { + "epoch": 2.6371521754215532, + "grad_norm": 1.0848140716552734, + "learning_rate": 5.495223749017523e-05, + "loss": 0.482, + "step": 38004 + }, + { + "epoch": 2.6372215668586496, + "grad_norm": 1.1187734603881836, + "learning_rate": 5.494722133040024e-05, + "loss": 0.4389, + "step": 38005 + }, + { + "epoch": 2.6372909582957464, + "grad_norm": 0.786873459815979, + "learning_rate": 5.4942205312851705e-05, + "loss": 0.2969, + "step": 38006 + }, + { + "epoch": 2.637360349732843, + "grad_norm": 0.8823023438453674, + "learning_rate": 5.4937189437545335e-05, + "loss": 0.4672, + "step": 38007 + }, + { + "epoch": 2.6374297411699397, + "grad_norm": 0.9136298298835754, + "learning_rate": 5.493217370449705e-05, + "loss": 0.2566, + "step": 38008 + }, + { + "epoch": 2.6374991326070365, + "grad_norm": 1.1685746908187866, + "learning_rate": 5.492715811372264e-05, + "loss": 0.3683, + "step": 38009 + }, + { + "epoch": 2.637568524044133, + "grad_norm": 1.265696406364441, + "learning_rate": 5.492214266523803e-05, + "loss": 0.5415, + "step": 38010 + }, + { + "epoch": 2.6376379154812297, + "grad_norm": 1.0871726274490356, + "learning_rate": 5.491712735905893e-05, + "loss": 0.3974, + "step": 38011 + }, + { + "epoch": 2.637707306918326, + "grad_norm": 1.0706037282943726, + "learning_rate": 5.491211219520123e-05, + "loss": 0.395, + "step": 38012 + }, + { + "epoch": 2.637776698355423, + "grad_norm": 1.1721835136413574, + "learning_rate": 5.4907097173680765e-05, + "loss": 0.432, + "step": 38013 + }, + { + "epoch": 2.6378460897925198, + "grad_norm": 0.9732682108879089, + "learning_rate": 5.490208229451336e-05, + "loss": 0.3651, + "step": 38014 + }, + { + "epoch": 2.637915481229616, + "grad_norm": 1.16836416721344, + "learning_rate": 5.489706755771484e-05, + "loss": 0.3104, + "step": 38015 + }, + { + "epoch": 2.637984872666713, + "grad_norm": 0.9509138464927673, + "learning_rate": 5.4892052963301065e-05, + "loss": 0.3711, + "step": 38016 + }, + { + "epoch": 2.6380542641038094, + "grad_norm": 1.242910623550415, + "learning_rate": 5.4887038511287866e-05, + "loss": 0.3635, + "step": 38017 + }, + { + "epoch": 2.638123655540906, + "grad_norm": 1.4624947309494019, + "learning_rate": 5.488202420169102e-05, + "loss": 0.3678, + "step": 38018 + }, + { + "epoch": 2.638193046978003, + "grad_norm": 1.1886602640151978, + "learning_rate": 5.487701003452639e-05, + "loss": 0.3568, + "step": 38019 + }, + { + "epoch": 2.6382624384150994, + "grad_norm": 0.9881294369697571, + "learning_rate": 5.48719960098098e-05, + "loss": 0.4134, + "step": 38020 + }, + { + "epoch": 2.6383318298521963, + "grad_norm": 1.0960066318511963, + "learning_rate": 5.486698212755709e-05, + "loss": 0.3981, + "step": 38021 + }, + { + "epoch": 2.6384012212892927, + "grad_norm": 1.2217823266983032, + "learning_rate": 5.486196838778408e-05, + "loss": 0.3748, + "step": 38022 + }, + { + "epoch": 2.6384706127263895, + "grad_norm": 0.9798653721809387, + "learning_rate": 5.4856954790506584e-05, + "loss": 0.3984, + "step": 38023 + }, + { + "epoch": 2.6385400041634863, + "grad_norm": 1.2408493757247925, + "learning_rate": 5.48519413357405e-05, + "loss": 0.4479, + "step": 38024 + }, + { + "epoch": 2.638609395600583, + "grad_norm": 1.1972277164459229, + "learning_rate": 5.4846928023501554e-05, + "loss": 0.2726, + "step": 38025 + }, + { + "epoch": 2.6386787870376796, + "grad_norm": 1.6105480194091797, + "learning_rate": 5.484191485380561e-05, + "loss": 0.3397, + "step": 38026 + }, + { + "epoch": 2.638748178474776, + "grad_norm": 0.9885740280151367, + "learning_rate": 5.483690182666852e-05, + "loss": 0.3206, + "step": 38027 + }, + { + "epoch": 2.6388175699118728, + "grad_norm": 1.300260066986084, + "learning_rate": 5.4831888942106066e-05, + "loss": 0.3728, + "step": 38028 + }, + { + "epoch": 2.6388869613489696, + "grad_norm": 4.505725383758545, + "learning_rate": 5.482687620013412e-05, + "loss": 0.4375, + "step": 38029 + }, + { + "epoch": 2.6389563527860664, + "grad_norm": 1.2333606481552124, + "learning_rate": 5.482186360076849e-05, + "loss": 0.3565, + "step": 38030 + }, + { + "epoch": 2.639025744223163, + "grad_norm": 1.0165563821792603, + "learning_rate": 5.4816851144025016e-05, + "loss": 0.4533, + "step": 38031 + }, + { + "epoch": 2.6390951356602597, + "grad_norm": 1.240401029586792, + "learning_rate": 5.481183882991947e-05, + "loss": 0.3553, + "step": 38032 + }, + { + "epoch": 2.639164527097356, + "grad_norm": 1.458408236503601, + "learning_rate": 5.48068266584677e-05, + "loss": 0.5772, + "step": 38033 + }, + { + "epoch": 2.639233918534453, + "grad_norm": 1.291211724281311, + "learning_rate": 5.480181462968553e-05, + "loss": 0.4268, + "step": 38034 + }, + { + "epoch": 2.6393033099715497, + "grad_norm": 1.34537672996521, + "learning_rate": 5.479680274358881e-05, + "loss": 0.3565, + "step": 38035 + }, + { + "epoch": 2.639372701408646, + "grad_norm": 1.0780482292175293, + "learning_rate": 5.4791791000193315e-05, + "loss": 0.3779, + "step": 38036 + }, + { + "epoch": 2.639442092845743, + "grad_norm": 1.53898286819458, + "learning_rate": 5.478677939951493e-05, + "loss": 0.5072, + "step": 38037 + }, + { + "epoch": 2.6395114842828393, + "grad_norm": 1.0207440853118896, + "learning_rate": 5.4781767941569415e-05, + "loss": 0.4397, + "step": 38038 + }, + { + "epoch": 2.639580875719936, + "grad_norm": 1.3439607620239258, + "learning_rate": 5.4776756626372604e-05, + "loss": 0.36, + "step": 38039 + }, + { + "epoch": 2.639650267157033, + "grad_norm": 1.1868244409561157, + "learning_rate": 5.477174545394037e-05, + "loss": 0.4727, + "step": 38040 + }, + { + "epoch": 2.6397196585941294, + "grad_norm": 1.1145833730697632, + "learning_rate": 5.476673442428844e-05, + "loss": 0.3975, + "step": 38041 + }, + { + "epoch": 2.639789050031226, + "grad_norm": 1.1567474603652954, + "learning_rate": 5.4761723537432685e-05, + "loss": 0.3179, + "step": 38042 + }, + { + "epoch": 2.6398584414683226, + "grad_norm": 1.3772789239883423, + "learning_rate": 5.475671279338892e-05, + "loss": 0.3608, + "step": 38043 + }, + { + "epoch": 2.6399278329054194, + "grad_norm": 1.6236822605133057, + "learning_rate": 5.475170219217296e-05, + "loss": 0.5267, + "step": 38044 + }, + { + "epoch": 2.6399972243425163, + "grad_norm": 1.1012364625930786, + "learning_rate": 5.474669173380067e-05, + "loss": 0.3671, + "step": 38045 + }, + { + "epoch": 2.6400666157796127, + "grad_norm": 1.413362979888916, + "learning_rate": 5.474168141828779e-05, + "loss": 0.3929, + "step": 38046 + }, + { + "epoch": 2.6401360072167095, + "grad_norm": 1.3218231201171875, + "learning_rate": 5.473667124565016e-05, + "loss": 0.4361, + "step": 38047 + }, + { + "epoch": 2.640205398653806, + "grad_norm": 0.9861844778060913, + "learning_rate": 5.473166121590362e-05, + "loss": 0.2958, + "step": 38048 + }, + { + "epoch": 2.6402747900909027, + "grad_norm": 1.012450933456421, + "learning_rate": 5.4726651329063963e-05, + "loss": 0.3474, + "step": 38049 + }, + { + "epoch": 2.6403441815279995, + "grad_norm": 1.281059980392456, + "learning_rate": 5.472164158514701e-05, + "loss": 0.3647, + "step": 38050 + }, + { + "epoch": 2.6404135729650964, + "grad_norm": 1.2433221340179443, + "learning_rate": 5.47166319841686e-05, + "loss": 0.3744, + "step": 38051 + }, + { + "epoch": 2.6404829644021928, + "grad_norm": 1.1712909936904907, + "learning_rate": 5.4711622526144544e-05, + "loss": 0.5454, + "step": 38052 + }, + { + "epoch": 2.640552355839289, + "grad_norm": 1.4700595140457153, + "learning_rate": 5.470661321109063e-05, + "loss": 0.3452, + "step": 38053 + }, + { + "epoch": 2.640621747276386, + "grad_norm": 1.136850357055664, + "learning_rate": 5.470160403902267e-05, + "loss": 0.4216, + "step": 38054 + }, + { + "epoch": 2.640691138713483, + "grad_norm": 1.5419683456420898, + "learning_rate": 5.469659500995647e-05, + "loss": 0.368, + "step": 38055 + }, + { + "epoch": 2.6407605301505797, + "grad_norm": 1.1532608270645142, + "learning_rate": 5.46915861239079e-05, + "loss": 0.5344, + "step": 38056 + }, + { + "epoch": 2.640829921587676, + "grad_norm": 1.5597248077392578, + "learning_rate": 5.4686577380892715e-05, + "loss": 0.358, + "step": 38057 + }, + { + "epoch": 2.640899313024773, + "grad_norm": 1.1486239433288574, + "learning_rate": 5.468156878092675e-05, + "loss": 0.4045, + "step": 38058 + }, + { + "epoch": 2.6409687044618693, + "grad_norm": 1.1056592464447021, + "learning_rate": 5.467656032402586e-05, + "loss": 0.317, + "step": 38059 + }, + { + "epoch": 2.641038095898966, + "grad_norm": 0.9907265305519104, + "learning_rate": 5.467155201020576e-05, + "loss": 0.2745, + "step": 38060 + }, + { + "epoch": 2.641107487336063, + "grad_norm": 1.5827144384384155, + "learning_rate": 5.4666543839482306e-05, + "loss": 0.6554, + "step": 38061 + }, + { + "epoch": 2.6411768787731593, + "grad_norm": 1.3811204433441162, + "learning_rate": 5.4661535811871336e-05, + "loss": 0.4381, + "step": 38062 + }, + { + "epoch": 2.641246270210256, + "grad_norm": 1.2565370798110962, + "learning_rate": 5.4656527927388627e-05, + "loss": 0.4723, + "step": 38063 + }, + { + "epoch": 2.6413156616473525, + "grad_norm": 1.0226306915283203, + "learning_rate": 5.465152018605e-05, + "loss": 0.3045, + "step": 38064 + }, + { + "epoch": 2.6413850530844494, + "grad_norm": 1.1053681373596191, + "learning_rate": 5.4646512587871256e-05, + "loss": 0.3186, + "step": 38065 + }, + { + "epoch": 2.641454444521546, + "grad_norm": 1.0580382347106934, + "learning_rate": 5.4641505132868253e-05, + "loss": 0.5521, + "step": 38066 + }, + { + "epoch": 2.6415238359586426, + "grad_norm": 0.9550095796585083, + "learning_rate": 5.4636497821056706e-05, + "loss": 0.4042, + "step": 38067 + }, + { + "epoch": 2.6415932273957394, + "grad_norm": 1.7717857360839844, + "learning_rate": 5.46314906524525e-05, + "loss": 0.5965, + "step": 38068 + }, + { + "epoch": 2.641662618832836, + "grad_norm": 1.0055369138717651, + "learning_rate": 5.462648362707139e-05, + "loss": 0.3775, + "step": 38069 + }, + { + "epoch": 2.6417320102699327, + "grad_norm": 1.195346713066101, + "learning_rate": 5.462147674492921e-05, + "loss": 0.4537, + "step": 38070 + }, + { + "epoch": 2.6418014017070295, + "grad_norm": 0.84554123878479, + "learning_rate": 5.461647000604178e-05, + "loss": 0.3338, + "step": 38071 + }, + { + "epoch": 2.641870793144126, + "grad_norm": 1.1914265155792236, + "learning_rate": 5.461146341042491e-05, + "loss": 0.3802, + "step": 38072 + }, + { + "epoch": 2.6419401845812227, + "grad_norm": 1.4152112007141113, + "learning_rate": 5.460645695809434e-05, + "loss": 0.321, + "step": 38073 + }, + { + "epoch": 2.642009576018319, + "grad_norm": 1.289204716682434, + "learning_rate": 5.460145064906593e-05, + "loss": 0.4987, + "step": 38074 + }, + { + "epoch": 2.642078967455416, + "grad_norm": 1.0352720022201538, + "learning_rate": 5.4596444483355514e-05, + "loss": 0.2924, + "step": 38075 + }, + { + "epoch": 2.6421483588925128, + "grad_norm": 1.5229183435440063, + "learning_rate": 5.459143846097881e-05, + "loss": 0.4384, + "step": 38076 + }, + { + "epoch": 2.6422177503296096, + "grad_norm": 0.8626216650009155, + "learning_rate": 5.4586432581951666e-05, + "loss": 0.2611, + "step": 38077 + }, + { + "epoch": 2.642287141766706, + "grad_norm": 1.5181366205215454, + "learning_rate": 5.458142684628989e-05, + "loss": 0.5325, + "step": 38078 + }, + { + "epoch": 2.642356533203803, + "grad_norm": 1.361752986907959, + "learning_rate": 5.457642125400932e-05, + "loss": 0.3956, + "step": 38079 + }, + { + "epoch": 2.642425924640899, + "grad_norm": 1.931859016418457, + "learning_rate": 5.457141580512568e-05, + "loss": 0.4532, + "step": 38080 + }, + { + "epoch": 2.642495316077996, + "grad_norm": 1.2519527673721313, + "learning_rate": 5.456641049965481e-05, + "loss": 0.3912, + "step": 38081 + }, + { + "epoch": 2.642564707515093, + "grad_norm": 1.8124233484268188, + "learning_rate": 5.4561405337612496e-05, + "loss": 0.4337, + "step": 38082 + }, + { + "epoch": 2.6426340989521893, + "grad_norm": 1.3956798315048218, + "learning_rate": 5.455640031901457e-05, + "loss": 0.3816, + "step": 38083 + }, + { + "epoch": 2.642703490389286, + "grad_norm": 1.3908194303512573, + "learning_rate": 5.455139544387681e-05, + "loss": 0.27, + "step": 38084 + }, + { + "epoch": 2.6427728818263825, + "grad_norm": 1.2583848237991333, + "learning_rate": 5.454639071221501e-05, + "loss": 0.4458, + "step": 38085 + }, + { + "epoch": 2.6428422732634793, + "grad_norm": 1.0747097730636597, + "learning_rate": 5.454138612404505e-05, + "loss": 0.3793, + "step": 38086 + }, + { + "epoch": 2.642911664700576, + "grad_norm": 1.3385554552078247, + "learning_rate": 5.45363816793826e-05, + "loss": 0.3858, + "step": 38087 + }, + { + "epoch": 2.6429810561376725, + "grad_norm": 1.276566505432129, + "learning_rate": 5.4531377378243506e-05, + "loss": 0.3853, + "step": 38088 + }, + { + "epoch": 2.6430504475747694, + "grad_norm": 1.473797082901001, + "learning_rate": 5.45263732206436e-05, + "loss": 0.5344, + "step": 38089 + }, + { + "epoch": 2.6431198390118658, + "grad_norm": 1.5315213203430176, + "learning_rate": 5.452136920659865e-05, + "loss": 0.3645, + "step": 38090 + }, + { + "epoch": 2.6431892304489626, + "grad_norm": 1.3083903789520264, + "learning_rate": 5.4516365336124455e-05, + "loss": 0.3053, + "step": 38091 + }, + { + "epoch": 2.6432586218860594, + "grad_norm": 1.1337405443191528, + "learning_rate": 5.4511361609236824e-05, + "loss": 0.4502, + "step": 38092 + }, + { + "epoch": 2.643328013323156, + "grad_norm": 1.4692349433898926, + "learning_rate": 5.450635802595159e-05, + "loss": 0.4544, + "step": 38093 + }, + { + "epoch": 2.6433974047602526, + "grad_norm": 1.1923103332519531, + "learning_rate": 5.4501354586284467e-05, + "loss": 0.3873, + "step": 38094 + }, + { + "epoch": 2.643466796197349, + "grad_norm": 0.8603568077087402, + "learning_rate": 5.449635129025128e-05, + "loss": 0.3501, + "step": 38095 + }, + { + "epoch": 2.643536187634446, + "grad_norm": 1.1765180826187134, + "learning_rate": 5.4491348137867835e-05, + "loss": 0.3373, + "step": 38096 + }, + { + "epoch": 2.6436055790715427, + "grad_norm": 0.8588748574256897, + "learning_rate": 5.4486345129149916e-05, + "loss": 0.2159, + "step": 38097 + }, + { + "epoch": 2.643674970508639, + "grad_norm": 1.1343978643417358, + "learning_rate": 5.448134226411334e-05, + "loss": 0.3749, + "step": 38098 + }, + { + "epoch": 2.643744361945736, + "grad_norm": 1.7388660907745361, + "learning_rate": 5.4476339542773894e-05, + "loss": 0.64, + "step": 38099 + }, + { + "epoch": 2.6438137533828323, + "grad_norm": 1.4076248407363892, + "learning_rate": 5.447133696514738e-05, + "loss": 0.4406, + "step": 38100 + }, + { + "epoch": 2.643883144819929, + "grad_norm": 1.0060853958129883, + "learning_rate": 5.446633453124954e-05, + "loss": 0.4436, + "step": 38101 + }, + { + "epoch": 2.643952536257026, + "grad_norm": 1.3747354745864868, + "learning_rate": 5.44613322410962e-05, + "loss": 0.4367, + "step": 38102 + }, + { + "epoch": 2.644021927694123, + "grad_norm": 1.1097038984298706, + "learning_rate": 5.4456330094703147e-05, + "loss": 0.3888, + "step": 38103 + }, + { + "epoch": 2.644091319131219, + "grad_norm": 1.3840610980987549, + "learning_rate": 5.445132809208617e-05, + "loss": 0.4081, + "step": 38104 + }, + { + "epoch": 2.644160710568316, + "grad_norm": 1.233741283416748, + "learning_rate": 5.444632623326109e-05, + "loss": 0.2405, + "step": 38105 + }, + { + "epoch": 2.6442301020054124, + "grad_norm": 1.4825160503387451, + "learning_rate": 5.4441324518243653e-05, + "loss": 0.4776, + "step": 38106 + }, + { + "epoch": 2.6442994934425093, + "grad_norm": 1.097743272781372, + "learning_rate": 5.443632294704971e-05, + "loss": 0.3439, + "step": 38107 + }, + { + "epoch": 2.644368884879606, + "grad_norm": 1.254254937171936, + "learning_rate": 5.443132151969498e-05, + "loss": 0.4981, + "step": 38108 + }, + { + "epoch": 2.6444382763167025, + "grad_norm": 0.7519970536231995, + "learning_rate": 5.442632023619527e-05, + "loss": 0.3203, + "step": 38109 + }, + { + "epoch": 2.6445076677537993, + "grad_norm": 1.1200095415115356, + "learning_rate": 5.4421319096566424e-05, + "loss": 0.3599, + "step": 38110 + }, + { + "epoch": 2.6445770591908957, + "grad_norm": 1.6861878633499146, + "learning_rate": 5.4416318100824146e-05, + "loss": 0.6755, + "step": 38111 + }, + { + "epoch": 2.6446464506279925, + "grad_norm": 1.25596284866333, + "learning_rate": 5.441131724898425e-05, + "loss": 0.3296, + "step": 38112 + }, + { + "epoch": 2.6447158420650894, + "grad_norm": 1.7237579822540283, + "learning_rate": 5.440631654106256e-05, + "loss": 0.5116, + "step": 38113 + }, + { + "epoch": 2.6447852335021858, + "grad_norm": 1.4495904445648193, + "learning_rate": 5.4401315977074864e-05, + "loss": 0.3989, + "step": 38114 + }, + { + "epoch": 2.6448546249392826, + "grad_norm": 1.4282267093658447, + "learning_rate": 5.439631555703688e-05, + "loss": 0.5678, + "step": 38115 + }, + { + "epoch": 2.644924016376379, + "grad_norm": 1.1939871311187744, + "learning_rate": 5.439131528096444e-05, + "loss": 0.4436, + "step": 38116 + }, + { + "epoch": 2.644993407813476, + "grad_norm": 1.1749716997146606, + "learning_rate": 5.438631514887334e-05, + "loss": 0.3029, + "step": 38117 + }, + { + "epoch": 2.6450627992505726, + "grad_norm": 0.9941607117652893, + "learning_rate": 5.438131516077934e-05, + "loss": 0.352, + "step": 38118 + }, + { + "epoch": 2.645132190687669, + "grad_norm": 1.2802335023880005, + "learning_rate": 5.4376315316698224e-05, + "loss": 0.3403, + "step": 38119 + }, + { + "epoch": 2.645201582124766, + "grad_norm": 2.092630624771118, + "learning_rate": 5.43713156166458e-05, + "loss": 0.4701, + "step": 38120 + }, + { + "epoch": 2.6452709735618622, + "grad_norm": 1.537523865699768, + "learning_rate": 5.436631606063788e-05, + "loss": 0.3669, + "step": 38121 + }, + { + "epoch": 2.645340364998959, + "grad_norm": 1.427470088005066, + "learning_rate": 5.436131664869016e-05, + "loss": 0.536, + "step": 38122 + }, + { + "epoch": 2.645409756436056, + "grad_norm": 2.0067708492279053, + "learning_rate": 5.435631738081848e-05, + "loss": 0.3258, + "step": 38123 + }, + { + "epoch": 2.6454791478731523, + "grad_norm": 1.015535593032837, + "learning_rate": 5.4351318257038584e-05, + "loss": 0.3963, + "step": 38124 + }, + { + "epoch": 2.645548539310249, + "grad_norm": 1.8088421821594238, + "learning_rate": 5.434631927736631e-05, + "loss": 0.4488, + "step": 38125 + }, + { + "epoch": 2.6456179307473455, + "grad_norm": 1.4168874025344849, + "learning_rate": 5.434132044181739e-05, + "loss": 0.5852, + "step": 38126 + }, + { + "epoch": 2.6456873221844424, + "grad_norm": 1.297390341758728, + "learning_rate": 5.433632175040764e-05, + "loss": 0.3846, + "step": 38127 + }, + { + "epoch": 2.645756713621539, + "grad_norm": 1.1255888938903809, + "learning_rate": 5.433132320315284e-05, + "loss": 0.3137, + "step": 38128 + }, + { + "epoch": 2.645826105058636, + "grad_norm": 1.1632932424545288, + "learning_rate": 5.432632480006874e-05, + "loss": 0.4163, + "step": 38129 + }, + { + "epoch": 2.6458954964957324, + "grad_norm": 1.3397495746612549, + "learning_rate": 5.432132654117113e-05, + "loss": 0.3295, + "step": 38130 + }, + { + "epoch": 2.6459648879328292, + "grad_norm": 1.110739827156067, + "learning_rate": 5.431632842647577e-05, + "loss": 0.4747, + "step": 38131 + }, + { + "epoch": 2.6460342793699256, + "grad_norm": 1.1860332489013672, + "learning_rate": 5.431133045599848e-05, + "loss": 0.2819, + "step": 38132 + }, + { + "epoch": 2.6461036708070225, + "grad_norm": 1.1493921279907227, + "learning_rate": 5.4306332629755016e-05, + "loss": 0.3694, + "step": 38133 + }, + { + "epoch": 2.6461730622441193, + "grad_norm": 0.9296635389328003, + "learning_rate": 5.430133494776115e-05, + "loss": 0.4096, + "step": 38134 + }, + { + "epoch": 2.6462424536812157, + "grad_norm": 1.3918322324752808, + "learning_rate": 5.429633741003272e-05, + "loss": 0.4739, + "step": 38135 + }, + { + "epoch": 2.6463118451183125, + "grad_norm": 1.1690624952316284, + "learning_rate": 5.42913400165854e-05, + "loss": 0.4735, + "step": 38136 + }, + { + "epoch": 2.646381236555409, + "grad_norm": 1.114176630973816, + "learning_rate": 5.428634276743503e-05, + "loss": 0.4113, + "step": 38137 + }, + { + "epoch": 2.6464506279925057, + "grad_norm": 0.9517456293106079, + "learning_rate": 5.428134566259735e-05, + "loss": 0.3102, + "step": 38138 + }, + { + "epoch": 2.6465200194296026, + "grad_norm": 1.3100618124008179, + "learning_rate": 5.427634870208817e-05, + "loss": 0.4008, + "step": 38139 + }, + { + "epoch": 2.646589410866699, + "grad_norm": 1.1657428741455078, + "learning_rate": 5.427135188592325e-05, + "loss": 0.3908, + "step": 38140 + }, + { + "epoch": 2.646658802303796, + "grad_norm": 1.5046501159667969, + "learning_rate": 5.426635521411841e-05, + "loss": 0.4457, + "step": 38141 + }, + { + "epoch": 2.646728193740892, + "grad_norm": 0.9711755514144897, + "learning_rate": 5.4261358686689346e-05, + "loss": 0.2304, + "step": 38142 + }, + { + "epoch": 2.646797585177989, + "grad_norm": 1.2504253387451172, + "learning_rate": 5.425636230365185e-05, + "loss": 0.3719, + "step": 38143 + }, + { + "epoch": 2.646866976615086, + "grad_norm": 0.8879294395446777, + "learning_rate": 5.4251366065021726e-05, + "loss": 0.2422, + "step": 38144 + }, + { + "epoch": 2.6469363680521822, + "grad_norm": 1.1408236026763916, + "learning_rate": 5.424636997081477e-05, + "loss": 0.3497, + "step": 38145 + }, + { + "epoch": 2.647005759489279, + "grad_norm": 1.604877233505249, + "learning_rate": 5.4241374021046676e-05, + "loss": 0.4516, + "step": 38146 + }, + { + "epoch": 2.6470751509263755, + "grad_norm": 1.0129499435424805, + "learning_rate": 5.423637821573324e-05, + "loss": 0.3623, + "step": 38147 + }, + { + "epoch": 2.6471445423634723, + "grad_norm": 1.04151451587677, + "learning_rate": 5.423138255489026e-05, + "loss": 0.4224, + "step": 38148 + }, + { + "epoch": 2.647213933800569, + "grad_norm": 1.176016926765442, + "learning_rate": 5.422638703853353e-05, + "loss": 0.3728, + "step": 38149 + }, + { + "epoch": 2.6472833252376655, + "grad_norm": 1.1102803945541382, + "learning_rate": 5.422139166667875e-05, + "loss": 0.4242, + "step": 38150 + }, + { + "epoch": 2.6473527166747624, + "grad_norm": 1.1192433834075928, + "learning_rate": 5.421639643934173e-05, + "loss": 0.3677, + "step": 38151 + }, + { + "epoch": 2.6474221081118587, + "grad_norm": 1.5060889720916748, + "learning_rate": 5.421140135653824e-05, + "loss": 0.4303, + "step": 38152 + }, + { + "epoch": 2.6474914995489556, + "grad_norm": 1.0965381860733032, + "learning_rate": 5.4206406418284026e-05, + "loss": 0.4917, + "step": 38153 + }, + { + "epoch": 2.6475608909860524, + "grad_norm": 2.0665338039398193, + "learning_rate": 5.420141162459489e-05, + "loss": 0.3984, + "step": 38154 + }, + { + "epoch": 2.6476302824231492, + "grad_norm": 0.9400511980056763, + "learning_rate": 5.419641697548661e-05, + "loss": 0.3813, + "step": 38155 + }, + { + "epoch": 2.6476996738602456, + "grad_norm": 1.7833402156829834, + "learning_rate": 5.419142247097489e-05, + "loss": 0.5881, + "step": 38156 + }, + { + "epoch": 2.6477690652973425, + "grad_norm": 1.0079729557037354, + "learning_rate": 5.418642811107553e-05, + "loss": 0.3048, + "step": 38157 + }, + { + "epoch": 2.647838456734439, + "grad_norm": 1.6129422187805176, + "learning_rate": 5.418143389580431e-05, + "loss": 0.3502, + "step": 38158 + }, + { + "epoch": 2.6479078481715357, + "grad_norm": 1.097630500793457, + "learning_rate": 5.4176439825176995e-05, + "loss": 0.4598, + "step": 38159 + }, + { + "epoch": 2.6479772396086325, + "grad_norm": 1.4833813905715942, + "learning_rate": 5.417144589920933e-05, + "loss": 0.3605, + "step": 38160 + }, + { + "epoch": 2.648046631045729, + "grad_norm": 1.1666003465652466, + "learning_rate": 5.416645211791709e-05, + "loss": 0.3263, + "step": 38161 + }, + { + "epoch": 2.6481160224828257, + "grad_norm": 2.8520116806030273, + "learning_rate": 5.416145848131608e-05, + "loss": 0.4878, + "step": 38162 + }, + { + "epoch": 2.648185413919922, + "grad_norm": 1.1185581684112549, + "learning_rate": 5.415646498942201e-05, + "loss": 0.2712, + "step": 38163 + }, + { + "epoch": 2.648254805357019, + "grad_norm": 0.882237434387207, + "learning_rate": 5.415147164225065e-05, + "loss": 0.3543, + "step": 38164 + }, + { + "epoch": 2.648324196794116, + "grad_norm": 1.0615662336349487, + "learning_rate": 5.414647843981777e-05, + "loss": 0.2835, + "step": 38165 + }, + { + "epoch": 2.648393588231212, + "grad_norm": 1.0261473655700684, + "learning_rate": 5.414148538213913e-05, + "loss": 0.3206, + "step": 38166 + }, + { + "epoch": 2.648462979668309, + "grad_norm": 2.1724886894226074, + "learning_rate": 5.4136492469230514e-05, + "loss": 0.4509, + "step": 38167 + }, + { + "epoch": 2.6485323711054054, + "grad_norm": 1.0048229694366455, + "learning_rate": 5.413149970110767e-05, + "loss": 0.2848, + "step": 38168 + }, + { + "epoch": 2.6486017625425022, + "grad_norm": 1.257463812828064, + "learning_rate": 5.412650707778639e-05, + "loss": 0.3577, + "step": 38169 + }, + { + "epoch": 2.648671153979599, + "grad_norm": 1.0715619325637817, + "learning_rate": 5.4121514599282375e-05, + "loss": 0.3133, + "step": 38170 + }, + { + "epoch": 2.6487405454166955, + "grad_norm": 1.0828691720962524, + "learning_rate": 5.4116522265611415e-05, + "loss": 0.3391, + "step": 38171 + }, + { + "epoch": 2.6488099368537923, + "grad_norm": 0.8346876502037048, + "learning_rate": 5.411153007678926e-05, + "loss": 0.3103, + "step": 38172 + }, + { + "epoch": 2.6488793282908887, + "grad_norm": 1.5923417806625366, + "learning_rate": 5.4106538032831687e-05, + "loss": 0.3917, + "step": 38173 + }, + { + "epoch": 2.6489487197279855, + "grad_norm": 1.0080420970916748, + "learning_rate": 5.4101546133754454e-05, + "loss": 0.4419, + "step": 38174 + }, + { + "epoch": 2.6490181111650823, + "grad_norm": 1.8861666917800903, + "learning_rate": 5.4096554379573306e-05, + "loss": 0.4323, + "step": 38175 + }, + { + "epoch": 2.6490875026021787, + "grad_norm": 0.8781725168228149, + "learning_rate": 5.409156277030405e-05, + "loss": 0.329, + "step": 38176 + }, + { + "epoch": 2.6491568940392756, + "grad_norm": 0.6977089047431946, + "learning_rate": 5.4086571305962374e-05, + "loss": 0.2212, + "step": 38177 + }, + { + "epoch": 2.649226285476372, + "grad_norm": 1.2209746837615967, + "learning_rate": 5.408157998656406e-05, + "loss": 0.4611, + "step": 38178 + }, + { + "epoch": 2.649295676913469, + "grad_norm": 1.3575146198272705, + "learning_rate": 5.4076588812124873e-05, + "loss": 0.4513, + "step": 38179 + }, + { + "epoch": 2.6493650683505656, + "grad_norm": 1.4568952322006226, + "learning_rate": 5.4071597782660596e-05, + "loss": 0.4915, + "step": 38180 + }, + { + "epoch": 2.6494344597876625, + "grad_norm": 1.1709098815917969, + "learning_rate": 5.4066606898186925e-05, + "loss": 0.4358, + "step": 38181 + }, + { + "epoch": 2.649503851224759, + "grad_norm": 0.9645494818687439, + "learning_rate": 5.406161615871965e-05, + "loss": 0.349, + "step": 38182 + }, + { + "epoch": 2.6495732426618557, + "grad_norm": 0.9569976329803467, + "learning_rate": 5.4056625564274554e-05, + "loss": 0.3094, + "step": 38183 + }, + { + "epoch": 2.649642634098952, + "grad_norm": 0.8935368061065674, + "learning_rate": 5.405163511486733e-05, + "loss": 0.2517, + "step": 38184 + }, + { + "epoch": 2.649712025536049, + "grad_norm": 1.1377298831939697, + "learning_rate": 5.404664481051376e-05, + "loss": 0.4463, + "step": 38185 + }, + { + "epoch": 2.6497814169731457, + "grad_norm": 1.4203261137008667, + "learning_rate": 5.404165465122959e-05, + "loss": 0.3878, + "step": 38186 + }, + { + "epoch": 2.649850808410242, + "grad_norm": 0.8765668869018555, + "learning_rate": 5.40366646370306e-05, + "loss": 0.3184, + "step": 38187 + }, + { + "epoch": 2.649920199847339, + "grad_norm": 1.0122716426849365, + "learning_rate": 5.403167476793251e-05, + "loss": 0.2722, + "step": 38188 + }, + { + "epoch": 2.6499895912844353, + "grad_norm": 0.8743422627449036, + "learning_rate": 5.402668504395111e-05, + "loss": 0.3577, + "step": 38189 + }, + { + "epoch": 2.650058982721532, + "grad_norm": 1.222787618637085, + "learning_rate": 5.4021695465102164e-05, + "loss": 0.443, + "step": 38190 + }, + { + "epoch": 2.650128374158629, + "grad_norm": 0.9209710955619812, + "learning_rate": 5.4016706031401345e-05, + "loss": 0.2946, + "step": 38191 + }, + { + "epoch": 2.6501977655957254, + "grad_norm": 0.9953347444534302, + "learning_rate": 5.401171674286445e-05, + "loss": 0.3632, + "step": 38192 + }, + { + "epoch": 2.6502671570328222, + "grad_norm": 0.9391133189201355, + "learning_rate": 5.400672759950723e-05, + "loss": 0.322, + "step": 38193 + }, + { + "epoch": 2.6503365484699186, + "grad_norm": 1.1279430389404297, + "learning_rate": 5.400173860134544e-05, + "loss": 0.3096, + "step": 38194 + }, + { + "epoch": 2.6504059399070155, + "grad_norm": 1.158394455909729, + "learning_rate": 5.3996749748394814e-05, + "loss": 0.4303, + "step": 38195 + }, + { + "epoch": 2.6504753313441123, + "grad_norm": 1.068767786026001, + "learning_rate": 5.399176104067113e-05, + "loss": 0.3736, + "step": 38196 + }, + { + "epoch": 2.6505447227812087, + "grad_norm": 0.9095953106880188, + "learning_rate": 5.398677247819015e-05, + "loss": 0.2954, + "step": 38197 + }, + { + "epoch": 2.6506141142183055, + "grad_norm": 1.3417799472808838, + "learning_rate": 5.398178406096754e-05, + "loss": 0.4554, + "step": 38198 + }, + { + "epoch": 2.650683505655402, + "grad_norm": 1.046717643737793, + "learning_rate": 5.3976795789019106e-05, + "loss": 0.3771, + "step": 38199 + }, + { + "epoch": 2.6507528970924987, + "grad_norm": 0.9634729027748108, + "learning_rate": 5.39718076623606e-05, + "loss": 0.3665, + "step": 38200 + }, + { + "epoch": 2.6508222885295956, + "grad_norm": 1.2334227561950684, + "learning_rate": 5.396681968100774e-05, + "loss": 0.4418, + "step": 38201 + }, + { + "epoch": 2.650891679966692, + "grad_norm": 1.1807122230529785, + "learning_rate": 5.39618318449763e-05, + "loss": 0.373, + "step": 38202 + }, + { + "epoch": 2.650961071403789, + "grad_norm": 1.3277475833892822, + "learning_rate": 5.3956844154282e-05, + "loss": 0.2684, + "step": 38203 + }, + { + "epoch": 2.651030462840885, + "grad_norm": 1.3585399389266968, + "learning_rate": 5.395185660894067e-05, + "loss": 0.7061, + "step": 38204 + }, + { + "epoch": 2.651099854277982, + "grad_norm": 1.161083459854126, + "learning_rate": 5.394686920896792e-05, + "loss": 0.4584, + "step": 38205 + }, + { + "epoch": 2.651169245715079, + "grad_norm": 0.8408169746398926, + "learning_rate": 5.394188195437957e-05, + "loss": 0.3064, + "step": 38206 + }, + { + "epoch": 2.6512386371521757, + "grad_norm": 1.0491019487380981, + "learning_rate": 5.3936894845191354e-05, + "loss": 0.3886, + "step": 38207 + }, + { + "epoch": 2.651308028589272, + "grad_norm": 1.2909375429153442, + "learning_rate": 5.393190788141902e-05, + "loss": 0.3937, + "step": 38208 + }, + { + "epoch": 2.651377420026369, + "grad_norm": 1.4592301845550537, + "learning_rate": 5.39269210630783e-05, + "loss": 0.478, + "step": 38209 + }, + { + "epoch": 2.6514468114634653, + "grad_norm": 1.251715898513794, + "learning_rate": 5.392193439018495e-05, + "loss": 0.4197, + "step": 38210 + }, + { + "epoch": 2.651516202900562, + "grad_norm": 1.2404639720916748, + "learning_rate": 5.391694786275474e-05, + "loss": 0.3194, + "step": 38211 + }, + { + "epoch": 2.651585594337659, + "grad_norm": 0.6859397292137146, + "learning_rate": 5.391196148080335e-05, + "loss": 0.2152, + "step": 38212 + }, + { + "epoch": 2.6516549857747553, + "grad_norm": 1.2044397592544556, + "learning_rate": 5.390697524434654e-05, + "loss": 0.3821, + "step": 38213 + }, + { + "epoch": 2.651724377211852, + "grad_norm": 1.3326051235198975, + "learning_rate": 5.3901989153400076e-05, + "loss": 0.4939, + "step": 38214 + }, + { + "epoch": 2.6517937686489486, + "grad_norm": 0.9550879597663879, + "learning_rate": 5.38970032079797e-05, + "loss": 0.2841, + "step": 38215 + }, + { + "epoch": 2.6518631600860454, + "grad_norm": 1.0200896263122559, + "learning_rate": 5.389201740810112e-05, + "loss": 0.4304, + "step": 38216 + }, + { + "epoch": 2.6519325515231422, + "grad_norm": 1.3714303970336914, + "learning_rate": 5.388703175378008e-05, + "loss": 0.2854, + "step": 38217 + }, + { + "epoch": 2.6520019429602386, + "grad_norm": 0.9954567551612854, + "learning_rate": 5.388204624503237e-05, + "loss": 0.251, + "step": 38218 + }, + { + "epoch": 2.6520713343973354, + "grad_norm": 3.1942782402038574, + "learning_rate": 5.387706088187365e-05, + "loss": 0.3998, + "step": 38219 + }, + { + "epoch": 2.652140725834432, + "grad_norm": 0.8939774632453918, + "learning_rate": 5.387207566431971e-05, + "loss": 0.2086, + "step": 38220 + }, + { + "epoch": 2.6522101172715287, + "grad_norm": 1.2782591581344604, + "learning_rate": 5.386709059238626e-05, + "loss": 0.5394, + "step": 38221 + }, + { + "epoch": 2.6522795087086255, + "grad_norm": 1.8601073026657104, + "learning_rate": 5.386210566608907e-05, + "loss": 0.56, + "step": 38222 + }, + { + "epoch": 2.652348900145722, + "grad_norm": 1.064748764038086, + "learning_rate": 5.3857120885443836e-05, + "loss": 0.232, + "step": 38223 + }, + { + "epoch": 2.6524182915828187, + "grad_norm": 1.1953577995300293, + "learning_rate": 5.385213625046633e-05, + "loss": 0.2367, + "step": 38224 + }, + { + "epoch": 2.652487683019915, + "grad_norm": 1.5053939819335938, + "learning_rate": 5.384715176117232e-05, + "loss": 0.4159, + "step": 38225 + }, + { + "epoch": 2.652557074457012, + "grad_norm": 1.2782447338104248, + "learning_rate": 5.384216741757745e-05, + "loss": 0.5168, + "step": 38226 + }, + { + "epoch": 2.6526264658941088, + "grad_norm": 1.4331799745559692, + "learning_rate": 5.3837183219697504e-05, + "loss": 0.5227, + "step": 38227 + }, + { + "epoch": 2.652695857331205, + "grad_norm": 1.371110200881958, + "learning_rate": 5.383219916754821e-05, + "loss": 0.4282, + "step": 38228 + }, + { + "epoch": 2.652765248768302, + "grad_norm": 1.8882758617401123, + "learning_rate": 5.382721526114531e-05, + "loss": 0.2674, + "step": 38229 + }, + { + "epoch": 2.6528346402053984, + "grad_norm": 0.7794345617294312, + "learning_rate": 5.382223150050454e-05, + "loss": 0.4164, + "step": 38230 + }, + { + "epoch": 2.652904031642495, + "grad_norm": 1.351029634475708, + "learning_rate": 5.381724788564166e-05, + "loss": 0.3989, + "step": 38231 + }, + { + "epoch": 2.652973423079592, + "grad_norm": 1.1639589071273804, + "learning_rate": 5.381226441657234e-05, + "loss": 0.4221, + "step": 38232 + }, + { + "epoch": 2.653042814516689, + "grad_norm": 1.7876238822937012, + "learning_rate": 5.380728109331233e-05, + "loss": 0.5258, + "step": 38233 + }, + { + "epoch": 2.6531122059537853, + "grad_norm": 1.2705317735671997, + "learning_rate": 5.380229791587739e-05, + "loss": 0.4549, + "step": 38234 + }, + { + "epoch": 2.653181597390882, + "grad_norm": 1.1383119821548462, + "learning_rate": 5.3797314884283226e-05, + "loss": 0.52, + "step": 38235 + }, + { + "epoch": 2.6532509888279785, + "grad_norm": 1.100251317024231, + "learning_rate": 5.3792331998545574e-05, + "loss": 0.5383, + "step": 38236 + }, + { + "epoch": 2.6533203802650753, + "grad_norm": 1.5005838871002197, + "learning_rate": 5.378734925868016e-05, + "loss": 0.3537, + "step": 38237 + }, + { + "epoch": 2.653389771702172, + "grad_norm": 1.5672465562820435, + "learning_rate": 5.3782366664702785e-05, + "loss": 0.4173, + "step": 38238 + }, + { + "epoch": 2.6534591631392686, + "grad_norm": 2.1838412284851074, + "learning_rate": 5.377738421662907e-05, + "loss": 0.6661, + "step": 38239 + }, + { + "epoch": 2.6535285545763654, + "grad_norm": 1.3678486347198486, + "learning_rate": 5.377240191447479e-05, + "loss": 0.4117, + "step": 38240 + }, + { + "epoch": 2.6535979460134618, + "grad_norm": 1.4076720476150513, + "learning_rate": 5.376741975825566e-05, + "loss": 0.5089, + "step": 38241 + }, + { + "epoch": 2.6536673374505586, + "grad_norm": 1.0831844806671143, + "learning_rate": 5.376243774798744e-05, + "loss": 0.3279, + "step": 38242 + }, + { + "epoch": 2.6537367288876554, + "grad_norm": 1.4249149560928345, + "learning_rate": 5.375745588368584e-05, + "loss": 0.5738, + "step": 38243 + }, + { + "epoch": 2.653806120324752, + "grad_norm": 1.364692211151123, + "learning_rate": 5.3752474165366574e-05, + "loss": 0.5833, + "step": 38244 + }, + { + "epoch": 2.6538755117618487, + "grad_norm": 1.3084917068481445, + "learning_rate": 5.374749259304542e-05, + "loss": 0.3779, + "step": 38245 + }, + { + "epoch": 2.653944903198945, + "grad_norm": 1.5544843673706055, + "learning_rate": 5.374251116673803e-05, + "loss": 0.3669, + "step": 38246 + }, + { + "epoch": 2.654014294636042, + "grad_norm": 1.2562605142593384, + "learning_rate": 5.373752988646017e-05, + "loss": 0.4466, + "step": 38247 + }, + { + "epoch": 2.6540836860731387, + "grad_norm": 1.2421345710754395, + "learning_rate": 5.3732548752227564e-05, + "loss": 0.3601, + "step": 38248 + }, + { + "epoch": 2.654153077510235, + "grad_norm": 0.9382283091545105, + "learning_rate": 5.372756776405592e-05, + "loss": 0.2657, + "step": 38249 + }, + { + "epoch": 2.654222468947332, + "grad_norm": 1.4481632709503174, + "learning_rate": 5.3722586921961036e-05, + "loss": 0.5163, + "step": 38250 + }, + { + "epoch": 2.6542918603844283, + "grad_norm": 0.7348549365997314, + "learning_rate": 5.371760622595852e-05, + "loss": 0.3191, + "step": 38251 + }, + { + "epoch": 2.654361251821525, + "grad_norm": 1.7150154113769531, + "learning_rate": 5.371262567606416e-05, + "loss": 0.3798, + "step": 38252 + }, + { + "epoch": 2.654430643258622, + "grad_norm": 1.2489817142486572, + "learning_rate": 5.3707645272293707e-05, + "loss": 0.3795, + "step": 38253 + }, + { + "epoch": 2.654500034695719, + "grad_norm": 1.180949091911316, + "learning_rate": 5.370266501466279e-05, + "loss": 0.5674, + "step": 38254 + }, + { + "epoch": 2.654569426132815, + "grad_norm": 1.4486831426620483, + "learning_rate": 5.369768490318721e-05, + "loss": 0.382, + "step": 38255 + }, + { + "epoch": 2.6546388175699116, + "grad_norm": 1.6043455600738525, + "learning_rate": 5.369270493788265e-05, + "loss": 0.4452, + "step": 38256 + }, + { + "epoch": 2.6547082090070084, + "grad_norm": 0.9647479057312012, + "learning_rate": 5.3687725118764874e-05, + "loss": 0.266, + "step": 38257 + }, + { + "epoch": 2.6547776004441053, + "grad_norm": 1.114732027053833, + "learning_rate": 5.3682745445849544e-05, + "loss": 0.3707, + "step": 38258 + }, + { + "epoch": 2.654846991881202, + "grad_norm": 1.3412444591522217, + "learning_rate": 5.367776591915248e-05, + "loss": 0.4163, + "step": 38259 + }, + { + "epoch": 2.6549163833182985, + "grad_norm": 1.6407454013824463, + "learning_rate": 5.367278653868928e-05, + "loss": 0.4784, + "step": 38260 + }, + { + "epoch": 2.6549857747553953, + "grad_norm": 1.529333472251892, + "learning_rate": 5.3667807304475724e-05, + "loss": 0.4512, + "step": 38261 + }, + { + "epoch": 2.6550551661924917, + "grad_norm": 1.0378910303115845, + "learning_rate": 5.366282821652753e-05, + "loss": 0.2797, + "step": 38262 + }, + { + "epoch": 2.6551245576295885, + "grad_norm": 1.087209939956665, + "learning_rate": 5.3657849274860396e-05, + "loss": 0.5205, + "step": 38263 + }, + { + "epoch": 2.6551939490666854, + "grad_norm": 1.213140845298767, + "learning_rate": 5.365287047949006e-05, + "loss": 0.3518, + "step": 38264 + }, + { + "epoch": 2.6552633405037818, + "grad_norm": 1.5107008218765259, + "learning_rate": 5.364789183043224e-05, + "loss": 0.438, + "step": 38265 + }, + { + "epoch": 2.6553327319408786, + "grad_norm": 1.5961788892745972, + "learning_rate": 5.3642913327702685e-05, + "loss": 0.4904, + "step": 38266 + }, + { + "epoch": 2.655402123377975, + "grad_norm": 1.4076567888259888, + "learning_rate": 5.363793497131704e-05, + "loss": 0.3711, + "step": 38267 + }, + { + "epoch": 2.655471514815072, + "grad_norm": 1.8823699951171875, + "learning_rate": 5.3632956761291054e-05, + "loss": 0.4773, + "step": 38268 + }, + { + "epoch": 2.6555409062521687, + "grad_norm": 0.9464420676231384, + "learning_rate": 5.3627978697640434e-05, + "loss": 0.4612, + "step": 38269 + }, + { + "epoch": 2.655610297689265, + "grad_norm": 1.0081005096435547, + "learning_rate": 5.362300078038092e-05, + "loss": 0.4016, + "step": 38270 + }, + { + "epoch": 2.655679689126362, + "grad_norm": 1.1725516319274902, + "learning_rate": 5.3618023009528196e-05, + "loss": 0.4144, + "step": 38271 + }, + { + "epoch": 2.6557490805634583, + "grad_norm": 1.1335052251815796, + "learning_rate": 5.361304538509799e-05, + "loss": 0.3181, + "step": 38272 + }, + { + "epoch": 2.655818472000555, + "grad_norm": 1.241365909576416, + "learning_rate": 5.360806790710607e-05, + "loss": 0.4169, + "step": 38273 + }, + { + "epoch": 2.655887863437652, + "grad_norm": 1.1401646137237549, + "learning_rate": 5.360309057556806e-05, + "loss": 0.3461, + "step": 38274 + }, + { + "epoch": 2.6559572548747483, + "grad_norm": 1.3839855194091797, + "learning_rate": 5.359811339049969e-05, + "loss": 0.5271, + "step": 38275 + }, + { + "epoch": 2.656026646311845, + "grad_norm": 1.6672645807266235, + "learning_rate": 5.3593136351916716e-05, + "loss": 0.4224, + "step": 38276 + }, + { + "epoch": 2.6560960377489415, + "grad_norm": 1.3592910766601562, + "learning_rate": 5.358815945983482e-05, + "loss": 0.4184, + "step": 38277 + }, + { + "epoch": 2.6561654291860384, + "grad_norm": 0.8988872170448303, + "learning_rate": 5.358318271426972e-05, + "loss": 0.4053, + "step": 38278 + }, + { + "epoch": 2.656234820623135, + "grad_norm": 1.209446907043457, + "learning_rate": 5.3578206115237116e-05, + "loss": 0.3687, + "step": 38279 + }, + { + "epoch": 2.656304212060232, + "grad_norm": 1.429610252380371, + "learning_rate": 5.3573229662752774e-05, + "loss": 0.3467, + "step": 38280 + }, + { + "epoch": 2.6563736034973284, + "grad_norm": 1.6314526796340942, + "learning_rate": 5.356825335683233e-05, + "loss": 0.4961, + "step": 38281 + }, + { + "epoch": 2.656442994934425, + "grad_norm": 1.3440182209014893, + "learning_rate": 5.356327719749151e-05, + "loss": 0.4578, + "step": 38282 + }, + { + "epoch": 2.6565123863715217, + "grad_norm": 1.2451839447021484, + "learning_rate": 5.355830118474605e-05, + "loss": 0.4018, + "step": 38283 + }, + { + "epoch": 2.6565817778086185, + "grad_norm": 1.0685787200927734, + "learning_rate": 5.355332531861167e-05, + "loss": 0.3147, + "step": 38284 + }, + { + "epoch": 2.6566511692457153, + "grad_norm": 1.1909124851226807, + "learning_rate": 5.354834959910402e-05, + "loss": 0.4277, + "step": 38285 + }, + { + "epoch": 2.6567205606828117, + "grad_norm": 1.4519236087799072, + "learning_rate": 5.354337402623883e-05, + "loss": 0.4059, + "step": 38286 + }, + { + "epoch": 2.6567899521199085, + "grad_norm": 1.094448208808899, + "learning_rate": 5.3538398600031835e-05, + "loss": 0.3116, + "step": 38287 + }, + { + "epoch": 2.656859343557005, + "grad_norm": 1.6840564012527466, + "learning_rate": 5.3533423320498756e-05, + "loss": 0.4413, + "step": 38288 + }, + { + "epoch": 2.6569287349941018, + "grad_norm": 1.3719652891159058, + "learning_rate": 5.3528448187655235e-05, + "loss": 0.2006, + "step": 38289 + }, + { + "epoch": 2.6569981264311986, + "grad_norm": 1.066267490386963, + "learning_rate": 5.3523473201517e-05, + "loss": 0.376, + "step": 38290 + }, + { + "epoch": 2.657067517868295, + "grad_norm": 1.350780725479126, + "learning_rate": 5.351849836209979e-05, + "loss": 0.518, + "step": 38291 + }, + { + "epoch": 2.657136909305392, + "grad_norm": 1.0145591497421265, + "learning_rate": 5.3513523669419264e-05, + "loss": 0.4178, + "step": 38292 + }, + { + "epoch": 2.657206300742488, + "grad_norm": 1.1191308498382568, + "learning_rate": 5.350854912349117e-05, + "loss": 0.4667, + "step": 38293 + }, + { + "epoch": 2.657275692179585, + "grad_norm": 1.4055335521697998, + "learning_rate": 5.350357472433121e-05, + "loss": 0.4202, + "step": 38294 + }, + { + "epoch": 2.657345083616682, + "grad_norm": 1.0032386779785156, + "learning_rate": 5.349860047195505e-05, + "loss": 0.3586, + "step": 38295 + }, + { + "epoch": 2.6574144750537783, + "grad_norm": 1.0151104927062988, + "learning_rate": 5.34936263663784e-05, + "loss": 0.1873, + "step": 38296 + }, + { + "epoch": 2.657483866490875, + "grad_norm": 1.125146746635437, + "learning_rate": 5.348865240761698e-05, + "loss": 0.3402, + "step": 38297 + }, + { + "epoch": 2.6575532579279715, + "grad_norm": 1.7577694654464722, + "learning_rate": 5.3483678595686495e-05, + "loss": 0.5941, + "step": 38298 + }, + { + "epoch": 2.6576226493650683, + "grad_norm": 1.1176735162734985, + "learning_rate": 5.347870493060263e-05, + "loss": 0.3294, + "step": 38299 + }, + { + "epoch": 2.657692040802165, + "grad_norm": 1.1524308919906616, + "learning_rate": 5.34737314123811e-05, + "loss": 0.372, + "step": 38300 + }, + { + "epoch": 2.6577614322392615, + "grad_norm": 1.1855303049087524, + "learning_rate": 5.346875804103764e-05, + "loss": 0.2919, + "step": 38301 + }, + { + "epoch": 2.6578308236763584, + "grad_norm": 0.8024687767028809, + "learning_rate": 5.346378481658787e-05, + "loss": 0.1952, + "step": 38302 + }, + { + "epoch": 2.6579002151134548, + "grad_norm": 1.206291913986206, + "learning_rate": 5.345881173904754e-05, + "loss": 0.4539, + "step": 38303 + }, + { + "epoch": 2.6579696065505516, + "grad_norm": 1.4664441347122192, + "learning_rate": 5.345383880843233e-05, + "loss": 0.4181, + "step": 38304 + }, + { + "epoch": 2.6580389979876484, + "grad_norm": 1.6753969192504883, + "learning_rate": 5.344886602475796e-05, + "loss": 0.5154, + "step": 38305 + }, + { + "epoch": 2.6581083894247453, + "grad_norm": 1.4610360860824585, + "learning_rate": 5.3443893388040124e-05, + "loss": 0.3433, + "step": 38306 + }, + { + "epoch": 2.6581777808618416, + "grad_norm": 1.4675709009170532, + "learning_rate": 5.343892089829454e-05, + "loss": 0.514, + "step": 38307 + }, + { + "epoch": 2.6582471722989385, + "grad_norm": 1.2666786909103394, + "learning_rate": 5.343394855553685e-05, + "loss": 0.4387, + "step": 38308 + }, + { + "epoch": 2.658316563736035, + "grad_norm": 1.0523600578308105, + "learning_rate": 5.342897635978279e-05, + "loss": 0.3503, + "step": 38309 + }, + { + "epoch": 2.6583859551731317, + "grad_norm": 1.756534457206726, + "learning_rate": 5.3424004311048035e-05, + "loss": 0.2979, + "step": 38310 + }, + { + "epoch": 2.6584553466102285, + "grad_norm": 1.8166894912719727, + "learning_rate": 5.34190324093483e-05, + "loss": 0.5512, + "step": 38311 + }, + { + "epoch": 2.658524738047325, + "grad_norm": 1.5490922927856445, + "learning_rate": 5.3414060654699275e-05, + "loss": 0.5966, + "step": 38312 + }, + { + "epoch": 2.6585941294844218, + "grad_norm": 0.9661470055580139, + "learning_rate": 5.3409089047116655e-05, + "loss": 0.3647, + "step": 38313 + }, + { + "epoch": 2.658663520921518, + "grad_norm": 1.2215101718902588, + "learning_rate": 5.340411758661618e-05, + "loss": 0.5133, + "step": 38314 + }, + { + "epoch": 2.658732912358615, + "grad_norm": 1.381055474281311, + "learning_rate": 5.339914627321345e-05, + "loss": 0.5632, + "step": 38315 + }, + { + "epoch": 2.658802303795712, + "grad_norm": 1.0047162771224976, + "learning_rate": 5.339417510692422e-05, + "loss": 0.2778, + "step": 38316 + }, + { + "epoch": 2.658871695232808, + "grad_norm": 1.0941262245178223, + "learning_rate": 5.338920408776416e-05, + "loss": 0.3261, + "step": 38317 + }, + { + "epoch": 2.658941086669905, + "grad_norm": 1.3149769306182861, + "learning_rate": 5.338423321574897e-05, + "loss": 0.4207, + "step": 38318 + }, + { + "epoch": 2.6590104781070014, + "grad_norm": 1.4621593952178955, + "learning_rate": 5.33792624908944e-05, + "loss": 0.3941, + "step": 38319 + }, + { + "epoch": 2.6590798695440983, + "grad_norm": 1.1311819553375244, + "learning_rate": 5.337429191321606e-05, + "loss": 0.3269, + "step": 38320 + }, + { + "epoch": 2.659149260981195, + "grad_norm": 1.488552212715149, + "learning_rate": 5.336932148272964e-05, + "loss": 0.4175, + "step": 38321 + }, + { + "epoch": 2.6592186524182915, + "grad_norm": 1.446488618850708, + "learning_rate": 5.336435119945089e-05, + "loss": 0.5372, + "step": 38322 + }, + { + "epoch": 2.6592880438553883, + "grad_norm": 1.010456919670105, + "learning_rate": 5.3359381063395494e-05, + "loss": 0.3182, + "step": 38323 + }, + { + "epoch": 2.6593574352924847, + "grad_norm": 1.02768075466156, + "learning_rate": 5.335441107457909e-05, + "loss": 0.3691, + "step": 38324 + }, + { + "epoch": 2.6594268267295815, + "grad_norm": 1.0447546243667603, + "learning_rate": 5.3349441233017395e-05, + "loss": 0.334, + "step": 38325 + }, + { + "epoch": 2.6594962181666784, + "grad_norm": 1.3257392644882202, + "learning_rate": 5.3344471538726105e-05, + "loss": 0.4147, + "step": 38326 + }, + { + "epoch": 2.6595656096037747, + "grad_norm": 0.9087225198745728, + "learning_rate": 5.333950199172091e-05, + "loss": 0.3757, + "step": 38327 + }, + { + "epoch": 2.6596350010408716, + "grad_norm": 1.5865808725357056, + "learning_rate": 5.333453259201753e-05, + "loss": 0.524, + "step": 38328 + }, + { + "epoch": 2.659704392477968, + "grad_norm": 2.413426399230957, + "learning_rate": 5.332956333963157e-05, + "loss": 0.4318, + "step": 38329 + }, + { + "epoch": 2.659773783915065, + "grad_norm": 1.1242051124572754, + "learning_rate": 5.332459423457876e-05, + "loss": 0.4298, + "step": 38330 + }, + { + "epoch": 2.6598431753521616, + "grad_norm": 1.233498454093933, + "learning_rate": 5.3319625276874794e-05, + "loss": 0.3573, + "step": 38331 + }, + { + "epoch": 2.6599125667892585, + "grad_norm": 1.8578318357467651, + "learning_rate": 5.3314656466535365e-05, + "loss": 0.471, + "step": 38332 + }, + { + "epoch": 2.659981958226355, + "grad_norm": 1.554996371269226, + "learning_rate": 5.3309687803576145e-05, + "loss": 0.4092, + "step": 38333 + }, + { + "epoch": 2.6600513496634517, + "grad_norm": 0.9794264435768127, + "learning_rate": 5.3304719288012815e-05, + "loss": 0.3868, + "step": 38334 + }, + { + "epoch": 2.660120741100548, + "grad_norm": 1.2479099035263062, + "learning_rate": 5.329975091986112e-05, + "loss": 0.5567, + "step": 38335 + }, + { + "epoch": 2.660190132537645, + "grad_norm": 1.3251206874847412, + "learning_rate": 5.3294782699136645e-05, + "loss": 0.4044, + "step": 38336 + }, + { + "epoch": 2.6602595239747417, + "grad_norm": 1.1955915689468384, + "learning_rate": 5.3289814625855126e-05, + "loss": 0.4926, + "step": 38337 + }, + { + "epoch": 2.660328915411838, + "grad_norm": 1.2353501319885254, + "learning_rate": 5.3284846700032244e-05, + "loss": 0.5564, + "step": 38338 + }, + { + "epoch": 2.660398306848935, + "grad_norm": 1.0612618923187256, + "learning_rate": 5.327987892168368e-05, + "loss": 0.3268, + "step": 38339 + }, + { + "epoch": 2.6604676982860314, + "grad_norm": 1.558967113494873, + "learning_rate": 5.3274911290825114e-05, + "loss": 0.4543, + "step": 38340 + }, + { + "epoch": 2.660537089723128, + "grad_norm": 1.273998498916626, + "learning_rate": 5.3269943807472233e-05, + "loss": 0.2911, + "step": 38341 + }, + { + "epoch": 2.660606481160225, + "grad_norm": 1.0635238885879517, + "learning_rate": 5.326497647164077e-05, + "loss": 0.3647, + "step": 38342 + }, + { + "epoch": 2.6606758725973214, + "grad_norm": 1.252472162246704, + "learning_rate": 5.326000928334629e-05, + "loss": 0.3874, + "step": 38343 + }, + { + "epoch": 2.6607452640344182, + "grad_norm": 1.6575655937194824, + "learning_rate": 5.325504224260456e-05, + "loss": 0.3444, + "step": 38344 + }, + { + "epoch": 2.6608146554715146, + "grad_norm": 1.1253366470336914, + "learning_rate": 5.325007534943123e-05, + "loss": 0.3369, + "step": 38345 + }, + { + "epoch": 2.6608840469086115, + "grad_norm": 0.8232235908508301, + "learning_rate": 5.324510860384199e-05, + "loss": 0.389, + "step": 38346 + }, + { + "epoch": 2.6609534383457083, + "grad_norm": 1.1889026165008545, + "learning_rate": 5.324014200585251e-05, + "loss": 0.3893, + "step": 38347 + }, + { + "epoch": 2.6610228297828047, + "grad_norm": 1.5144481658935547, + "learning_rate": 5.323517555547848e-05, + "loss": 0.5127, + "step": 38348 + }, + { + "epoch": 2.6610922212199015, + "grad_norm": 1.3840700387954712, + "learning_rate": 5.323020925273562e-05, + "loss": 0.8079, + "step": 38349 + }, + { + "epoch": 2.661161612656998, + "grad_norm": 0.9000358581542969, + "learning_rate": 5.322524309763952e-05, + "loss": 0.3632, + "step": 38350 + }, + { + "epoch": 2.6612310040940947, + "grad_norm": 0.9375554919242859, + "learning_rate": 5.322027709020589e-05, + "loss": 0.2648, + "step": 38351 + }, + { + "epoch": 2.6613003955311916, + "grad_norm": 1.480207085609436, + "learning_rate": 5.321531123045044e-05, + "loss": 0.2886, + "step": 38352 + }, + { + "epoch": 2.661369786968288, + "grad_norm": 1.2189818620681763, + "learning_rate": 5.3210345518388817e-05, + "loss": 0.6156, + "step": 38353 + }, + { + "epoch": 2.661439178405385, + "grad_norm": 1.3316774368286133, + "learning_rate": 5.3205379954036735e-05, + "loss": 0.4642, + "step": 38354 + }, + { + "epoch": 2.661508569842481, + "grad_norm": 1.6426407098770142, + "learning_rate": 5.32004145374098e-05, + "loss": 0.482, + "step": 38355 + }, + { + "epoch": 2.661577961279578, + "grad_norm": 0.7282587289810181, + "learning_rate": 5.319544926852373e-05, + "loss": 0.3167, + "step": 38356 + }, + { + "epoch": 2.661647352716675, + "grad_norm": 1.0544610023498535, + "learning_rate": 5.3190484147394203e-05, + "loss": 0.3905, + "step": 38357 + }, + { + "epoch": 2.6617167441537717, + "grad_norm": 1.000556230545044, + "learning_rate": 5.3185519174036925e-05, + "loss": 0.2385, + "step": 38358 + }, + { + "epoch": 2.661786135590868, + "grad_norm": 1.4564226865768433, + "learning_rate": 5.3180554348467496e-05, + "loss": 0.3937, + "step": 38359 + }, + { + "epoch": 2.661855527027965, + "grad_norm": 1.146040916442871, + "learning_rate": 5.317558967070161e-05, + "loss": 0.426, + "step": 38360 + }, + { + "epoch": 2.6619249184650613, + "grad_norm": 1.024158000946045, + "learning_rate": 5.317062514075496e-05, + "loss": 0.3888, + "step": 38361 + }, + { + "epoch": 2.661994309902158, + "grad_norm": 1.5464855432510376, + "learning_rate": 5.316566075864322e-05, + "loss": 0.5043, + "step": 38362 + }, + { + "epoch": 2.662063701339255, + "grad_norm": 1.0759109258651733, + "learning_rate": 5.316069652438208e-05, + "loss": 0.2886, + "step": 38363 + }, + { + "epoch": 2.6621330927763514, + "grad_norm": 1.1625715494155884, + "learning_rate": 5.315573243798716e-05, + "loss": 0.4256, + "step": 38364 + }, + { + "epoch": 2.662202484213448, + "grad_norm": 0.879810631275177, + "learning_rate": 5.315076849947416e-05, + "loss": 0.2366, + "step": 38365 + }, + { + "epoch": 2.6622718756505446, + "grad_norm": 1.8117406368255615, + "learning_rate": 5.314580470885874e-05, + "loss": 0.4495, + "step": 38366 + }, + { + "epoch": 2.6623412670876414, + "grad_norm": 1.074141025543213, + "learning_rate": 5.3140841066156577e-05, + "loss": 0.2793, + "step": 38367 + }, + { + "epoch": 2.6624106585247382, + "grad_norm": 1.14736807346344, + "learning_rate": 5.313587757138334e-05, + "loss": 0.3088, + "step": 38368 + }, + { + "epoch": 2.6624800499618346, + "grad_norm": 1.1890407800674438, + "learning_rate": 5.313091422455472e-05, + "loss": 0.351, + "step": 38369 + }, + { + "epoch": 2.6625494413989315, + "grad_norm": 1.5539934635162354, + "learning_rate": 5.312595102568638e-05, + "loss": 0.3921, + "step": 38370 + }, + { + "epoch": 2.662618832836028, + "grad_norm": 1.1840620040893555, + "learning_rate": 5.312098797479395e-05, + "loss": 0.4775, + "step": 38371 + }, + { + "epoch": 2.6626882242731247, + "grad_norm": 1.4218393564224243, + "learning_rate": 5.311602507189313e-05, + "loss": 0.4369, + "step": 38372 + }, + { + "epoch": 2.6627576157102215, + "grad_norm": 0.9049053192138672, + "learning_rate": 5.311106231699956e-05, + "loss": 0.2618, + "step": 38373 + }, + { + "epoch": 2.662827007147318, + "grad_norm": 1.2663213014602661, + "learning_rate": 5.310609971012893e-05, + "loss": 0.3873, + "step": 38374 + }, + { + "epoch": 2.6628963985844147, + "grad_norm": 1.170756459236145, + "learning_rate": 5.310113725129691e-05, + "loss": 0.4615, + "step": 38375 + }, + { + "epoch": 2.662965790021511, + "grad_norm": 1.0982118844985962, + "learning_rate": 5.309617494051915e-05, + "loss": 0.4298, + "step": 38376 + }, + { + "epoch": 2.663035181458608, + "grad_norm": 1.2626991271972656, + "learning_rate": 5.309121277781137e-05, + "loss": 0.4099, + "step": 38377 + }, + { + "epoch": 2.663104572895705, + "grad_norm": 1.2648125886917114, + "learning_rate": 5.308625076318915e-05, + "loss": 0.31, + "step": 38378 + }, + { + "epoch": 2.663173964332801, + "grad_norm": 1.5080652236938477, + "learning_rate": 5.3081288896668194e-05, + "loss": 0.3571, + "step": 38379 + }, + { + "epoch": 2.663243355769898, + "grad_norm": 1.0936470031738281, + "learning_rate": 5.307632717826416e-05, + "loss": 0.3555, + "step": 38380 + }, + { + "epoch": 2.6633127472069944, + "grad_norm": 0.9611606001853943, + "learning_rate": 5.307136560799272e-05, + "loss": 0.3581, + "step": 38381 + }, + { + "epoch": 2.6633821386440912, + "grad_norm": 1.624646544456482, + "learning_rate": 5.306640418586954e-05, + "loss": 0.5702, + "step": 38382 + }, + { + "epoch": 2.663451530081188, + "grad_norm": 1.3135850429534912, + "learning_rate": 5.3061442911910306e-05, + "loss": 0.3984, + "step": 38383 + }, + { + "epoch": 2.663520921518285, + "grad_norm": 1.2437855005264282, + "learning_rate": 5.3056481786130605e-05, + "loss": 0.3366, + "step": 38384 + }, + { + "epoch": 2.6635903129553813, + "grad_norm": 2.126821756362915, + "learning_rate": 5.305152080854616e-05, + "loss": 0.6484, + "step": 38385 + }, + { + "epoch": 2.663659704392478, + "grad_norm": 1.3732936382293701, + "learning_rate": 5.3046559979172604e-05, + "loss": 0.4236, + "step": 38386 + }, + { + "epoch": 2.6637290958295745, + "grad_norm": 1.4743013381958008, + "learning_rate": 5.304159929802562e-05, + "loss": 0.5928, + "step": 38387 + }, + { + "epoch": 2.6637984872666713, + "grad_norm": 1.2392784357070923, + "learning_rate": 5.303663876512086e-05, + "loss": 0.3579, + "step": 38388 + }, + { + "epoch": 2.663867878703768, + "grad_norm": 1.6975177526474, + "learning_rate": 5.303167838047403e-05, + "loss": 0.6716, + "step": 38389 + }, + { + "epoch": 2.6639372701408646, + "grad_norm": 1.4030754566192627, + "learning_rate": 5.302671814410068e-05, + "loss": 0.4934, + "step": 38390 + }, + { + "epoch": 2.6640066615779614, + "grad_norm": 1.1379753351211548, + "learning_rate": 5.3021758056016545e-05, + "loss": 0.3792, + "step": 38391 + }, + { + "epoch": 2.664076053015058, + "grad_norm": 1.0384474992752075, + "learning_rate": 5.3016798116237266e-05, + "loss": 0.2926, + "step": 38392 + }, + { + "epoch": 2.6641454444521546, + "grad_norm": 1.398051142692566, + "learning_rate": 5.301183832477854e-05, + "loss": 0.4388, + "step": 38393 + }, + { + "epoch": 2.6642148358892515, + "grad_norm": 1.8164639472961426, + "learning_rate": 5.300687868165596e-05, + "loss": 0.3853, + "step": 38394 + }, + { + "epoch": 2.664284227326348, + "grad_norm": 1.5204392671585083, + "learning_rate": 5.3001919186885206e-05, + "loss": 0.4566, + "step": 38395 + }, + { + "epoch": 2.6643536187634447, + "grad_norm": 1.027681827545166, + "learning_rate": 5.299695984048194e-05, + "loss": 0.449, + "step": 38396 + }, + { + "epoch": 2.664423010200541, + "grad_norm": 1.2567037343978882, + "learning_rate": 5.2992000642461856e-05, + "loss": 0.4346, + "step": 38397 + }, + { + "epoch": 2.664492401637638, + "grad_norm": 1.2612011432647705, + "learning_rate": 5.298704159284053e-05, + "loss": 0.2296, + "step": 38398 + }, + { + "epoch": 2.6645617930747347, + "grad_norm": 0.9725635051727295, + "learning_rate": 5.2982082691633673e-05, + "loss": 0.3593, + "step": 38399 + }, + { + "epoch": 2.664631184511831, + "grad_norm": 1.2646615505218506, + "learning_rate": 5.297712393885691e-05, + "loss": 0.3538, + "step": 38400 + }, + { + "epoch": 2.664700575948928, + "grad_norm": 1.407055139541626, + "learning_rate": 5.2972165334525917e-05, + "loss": 0.3638, + "step": 38401 + }, + { + "epoch": 2.6647699673860243, + "grad_norm": 1.2330632209777832, + "learning_rate": 5.296720687865635e-05, + "loss": 0.4491, + "step": 38402 + }, + { + "epoch": 2.664839358823121, + "grad_norm": 1.2936733961105347, + "learning_rate": 5.296224857126384e-05, + "loss": 0.3783, + "step": 38403 + }, + { + "epoch": 2.664908750260218, + "grad_norm": 0.9966455698013306, + "learning_rate": 5.29572904123641e-05, + "loss": 0.5096, + "step": 38404 + }, + { + "epoch": 2.6649781416973144, + "grad_norm": 0.9897196292877197, + "learning_rate": 5.295233240197269e-05, + "loss": 0.4141, + "step": 38405 + }, + { + "epoch": 2.6650475331344112, + "grad_norm": 1.6294986009597778, + "learning_rate": 5.2947374540105324e-05, + "loss": 0.3928, + "step": 38406 + }, + { + "epoch": 2.6651169245715076, + "grad_norm": 1.2335104942321777, + "learning_rate": 5.294241682677763e-05, + "loss": 0.3987, + "step": 38407 + }, + { + "epoch": 2.6651863160086045, + "grad_norm": 1.615396499633789, + "learning_rate": 5.293745926200526e-05, + "loss": 0.4349, + "step": 38408 + }, + { + "epoch": 2.6652557074457013, + "grad_norm": 1.1195673942565918, + "learning_rate": 5.293250184580387e-05, + "loss": 0.391, + "step": 38409 + }, + { + "epoch": 2.665325098882798, + "grad_norm": 1.166973352432251, + "learning_rate": 5.292754457818911e-05, + "loss": 0.3973, + "step": 38410 + }, + { + "epoch": 2.6653944903198945, + "grad_norm": 1.0104931592941284, + "learning_rate": 5.292258745917669e-05, + "loss": 0.3087, + "step": 38411 + }, + { + "epoch": 2.6654638817569913, + "grad_norm": 0.8143792152404785, + "learning_rate": 5.2917630488782136e-05, + "loss": 0.2973, + "step": 38412 + }, + { + "epoch": 2.6655332731940877, + "grad_norm": 1.1621536016464233, + "learning_rate": 5.291267366702117e-05, + "loss": 0.3875, + "step": 38413 + }, + { + "epoch": 2.6656026646311846, + "grad_norm": 0.9746607542037964, + "learning_rate": 5.290771699390944e-05, + "loss": 0.3051, + "step": 38414 + }, + { + "epoch": 2.6656720560682814, + "grad_norm": 1.0638508796691895, + "learning_rate": 5.2902760469462565e-05, + "loss": 0.5088, + "step": 38415 + }, + { + "epoch": 2.665741447505378, + "grad_norm": 1.5128451585769653, + "learning_rate": 5.289780409369622e-05, + "loss": 0.4194, + "step": 38416 + }, + { + "epoch": 2.6658108389424746, + "grad_norm": 1.0807716846466064, + "learning_rate": 5.289284786662605e-05, + "loss": 0.3886, + "step": 38417 + }, + { + "epoch": 2.665880230379571, + "grad_norm": 1.232024073600769, + "learning_rate": 5.288789178826772e-05, + "loss": 0.4738, + "step": 38418 + }, + { + "epoch": 2.665949621816668, + "grad_norm": 2.3263585567474365, + "learning_rate": 5.2882935858636815e-05, + "loss": 0.4834, + "step": 38419 + }, + { + "epoch": 2.6660190132537647, + "grad_norm": 1.2703505754470825, + "learning_rate": 5.287798007774902e-05, + "loss": 0.3696, + "step": 38420 + }, + { + "epoch": 2.666088404690861, + "grad_norm": 1.0144450664520264, + "learning_rate": 5.287302444561997e-05, + "loss": 0.322, + "step": 38421 + }, + { + "epoch": 2.666157796127958, + "grad_norm": 0.9031859636306763, + "learning_rate": 5.2868068962265304e-05, + "loss": 0.3479, + "step": 38422 + }, + { + "epoch": 2.6662271875650543, + "grad_norm": 1.2146793603897095, + "learning_rate": 5.286311362770069e-05, + "loss": 0.479, + "step": 38423 + }, + { + "epoch": 2.666296579002151, + "grad_norm": 0.8639112710952759, + "learning_rate": 5.28581584419418e-05, + "loss": 0.3293, + "step": 38424 + }, + { + "epoch": 2.666365970439248, + "grad_norm": 0.923464834690094, + "learning_rate": 5.285320340500418e-05, + "loss": 0.3663, + "step": 38425 + }, + { + "epoch": 2.6664353618763443, + "grad_norm": 1.5711641311645508, + "learning_rate": 5.2848248516903534e-05, + "loss": 0.5448, + "step": 38426 + }, + { + "epoch": 2.666504753313441, + "grad_norm": 1.626429796218872, + "learning_rate": 5.284329377765553e-05, + "loss": 0.3435, + "step": 38427 + }, + { + "epoch": 2.6665741447505376, + "grad_norm": 1.5069900751113892, + "learning_rate": 5.283833918727575e-05, + "loss": 0.405, + "step": 38428 + }, + { + "epoch": 2.6666435361876344, + "grad_norm": 1.4706460237503052, + "learning_rate": 5.283338474577986e-05, + "loss": 0.6769, + "step": 38429 + }, + { + "epoch": 2.666712927624731, + "grad_norm": 1.0474199056625366, + "learning_rate": 5.28284304531835e-05, + "loss": 0.292, + "step": 38430 + }, + { + "epoch": 2.6667823190618276, + "grad_norm": 0.953061580657959, + "learning_rate": 5.282347630950231e-05, + "loss": 0.465, + "step": 38431 + }, + { + "epoch": 2.6668517104989244, + "grad_norm": 1.025244116783142, + "learning_rate": 5.281852231475197e-05, + "loss": 0.3171, + "step": 38432 + }, + { + "epoch": 2.666921101936021, + "grad_norm": 1.4016351699829102, + "learning_rate": 5.281356846894804e-05, + "loss": 0.6475, + "step": 38433 + }, + { + "epoch": 2.6669904933731177, + "grad_norm": 1.1826430559158325, + "learning_rate": 5.28086147721062e-05, + "loss": 0.3869, + "step": 38434 + }, + { + "epoch": 2.6670598848102145, + "grad_norm": 1.3242638111114502, + "learning_rate": 5.28036612242421e-05, + "loss": 0.4493, + "step": 38435 + }, + { + "epoch": 2.6671292762473113, + "grad_norm": 1.3487250804901123, + "learning_rate": 5.279870782537135e-05, + "loss": 0.4037, + "step": 38436 + }, + { + "epoch": 2.6671986676844077, + "grad_norm": 1.2923389673233032, + "learning_rate": 5.27937545755096e-05, + "loss": 0.3451, + "step": 38437 + }, + { + "epoch": 2.6672680591215046, + "grad_norm": 0.9595087170600891, + "learning_rate": 5.27888014746725e-05, + "loss": 0.3659, + "step": 38438 + }, + { + "epoch": 2.667337450558601, + "grad_norm": 0.9368408918380737, + "learning_rate": 5.278384852287572e-05, + "loss": 0.3617, + "step": 38439 + }, + { + "epoch": 2.6674068419956978, + "grad_norm": 0.9279530048370361, + "learning_rate": 5.2778895720134805e-05, + "loss": 0.3444, + "step": 38440 + }, + { + "epoch": 2.6674762334327946, + "grad_norm": 2.000558614730835, + "learning_rate": 5.277394306646544e-05, + "loss": 0.4985, + "step": 38441 + }, + { + "epoch": 2.667545624869891, + "grad_norm": 0.8059353232383728, + "learning_rate": 5.2768990561883244e-05, + "loss": 0.272, + "step": 38442 + }, + { + "epoch": 2.667615016306988, + "grad_norm": 1.3515428304672241, + "learning_rate": 5.2764038206403886e-05, + "loss": 0.4961, + "step": 38443 + }, + { + "epoch": 2.667684407744084, + "grad_norm": 1.3224745988845825, + "learning_rate": 5.275908600004298e-05, + "loss": 0.4525, + "step": 38444 + }, + { + "epoch": 2.667753799181181, + "grad_norm": 1.9752075672149658, + "learning_rate": 5.2754133942816145e-05, + "loss": 0.4603, + "step": 38445 + }, + { + "epoch": 2.667823190618278, + "grad_norm": 1.4339535236358643, + "learning_rate": 5.274918203473906e-05, + "loss": 0.3232, + "step": 38446 + }, + { + "epoch": 2.6678925820553743, + "grad_norm": 1.111229658126831, + "learning_rate": 5.274423027582731e-05, + "loss": 0.4624, + "step": 38447 + }, + { + "epoch": 2.667961973492471, + "grad_norm": 1.1951680183410645, + "learning_rate": 5.273927866609652e-05, + "loss": 0.4686, + "step": 38448 + }, + { + "epoch": 2.6680313649295675, + "grad_norm": 1.1935149431228638, + "learning_rate": 5.273432720556236e-05, + "loss": 0.5116, + "step": 38449 + }, + { + "epoch": 2.6681007563666643, + "grad_norm": 1.0863654613494873, + "learning_rate": 5.272937589424044e-05, + "loss": 0.2126, + "step": 38450 + }, + { + "epoch": 2.668170147803761, + "grad_norm": 1.0144788026809692, + "learning_rate": 5.272442473214639e-05, + "loss": 0.3045, + "step": 38451 + }, + { + "epoch": 2.6682395392408575, + "grad_norm": 1.6492489576339722, + "learning_rate": 5.2719473719295854e-05, + "loss": 0.4746, + "step": 38452 + }, + { + "epoch": 2.6683089306779544, + "grad_norm": 1.7069083452224731, + "learning_rate": 5.271452285570448e-05, + "loss": 0.4686, + "step": 38453 + }, + { + "epoch": 2.6683783221150508, + "grad_norm": 1.0787748098373413, + "learning_rate": 5.2709572141387855e-05, + "loss": 0.3513, + "step": 38454 + }, + { + "epoch": 2.6684477135521476, + "grad_norm": 1.5363450050354004, + "learning_rate": 5.27046215763616e-05, + "loss": 0.418, + "step": 38455 + }, + { + "epoch": 2.6685171049892444, + "grad_norm": 1.3545870780944824, + "learning_rate": 5.2699671160641384e-05, + "loss": 0.5738, + "step": 38456 + }, + { + "epoch": 2.668586496426341, + "grad_norm": 1.3164544105529785, + "learning_rate": 5.269472089424281e-05, + "loss": 0.304, + "step": 38457 + }, + { + "epoch": 2.6686558878634377, + "grad_norm": 1.224318265914917, + "learning_rate": 5.268977077718151e-05, + "loss": 0.3214, + "step": 38458 + }, + { + "epoch": 2.668725279300534, + "grad_norm": 1.0684552192687988, + "learning_rate": 5.268482080947317e-05, + "loss": 0.2617, + "step": 38459 + }, + { + "epoch": 2.668794670737631, + "grad_norm": 0.9737393260002136, + "learning_rate": 5.267987099113332e-05, + "loss": 0.3565, + "step": 38460 + }, + { + "epoch": 2.6688640621747277, + "grad_norm": 1.7929517030715942, + "learning_rate": 5.267492132217762e-05, + "loss": 0.4904, + "step": 38461 + }, + { + "epoch": 2.6689334536118245, + "grad_norm": 1.0324410200119019, + "learning_rate": 5.266997180262173e-05, + "loss": 0.3598, + "step": 38462 + }, + { + "epoch": 2.669002845048921, + "grad_norm": 1.3617243766784668, + "learning_rate": 5.2665022432481225e-05, + "loss": 0.4632, + "step": 38463 + }, + { + "epoch": 2.6690722364860178, + "grad_norm": 1.1325249671936035, + "learning_rate": 5.266007321177174e-05, + "loss": 0.4563, + "step": 38464 + }, + { + "epoch": 2.669141627923114, + "grad_norm": 1.2120407819747925, + "learning_rate": 5.2655124140508924e-05, + "loss": 0.3108, + "step": 38465 + }, + { + "epoch": 2.669211019360211, + "grad_norm": 1.114152431488037, + "learning_rate": 5.265017521870842e-05, + "loss": 0.4422, + "step": 38466 + }, + { + "epoch": 2.669280410797308, + "grad_norm": 1.3850529193878174, + "learning_rate": 5.264522644638579e-05, + "loss": 0.3507, + "step": 38467 + }, + { + "epoch": 2.669349802234404, + "grad_norm": 1.3176931142807007, + "learning_rate": 5.264027782355667e-05, + "loss": 0.3515, + "step": 38468 + }, + { + "epoch": 2.669419193671501, + "grad_norm": 1.079308032989502, + "learning_rate": 5.2635329350236706e-05, + "loss": 0.2917, + "step": 38469 + }, + { + "epoch": 2.6694885851085974, + "grad_norm": 1.5506521463394165, + "learning_rate": 5.2630381026441514e-05, + "loss": 0.3899, + "step": 38470 + }, + { + "epoch": 2.6695579765456943, + "grad_norm": 1.6175369024276733, + "learning_rate": 5.262543285218671e-05, + "loss": 0.5574, + "step": 38471 + }, + { + "epoch": 2.669627367982791, + "grad_norm": 0.9677188992500305, + "learning_rate": 5.262048482748792e-05, + "loss": 0.282, + "step": 38472 + }, + { + "epoch": 2.6696967594198875, + "grad_norm": 0.8421090841293335, + "learning_rate": 5.26155369523608e-05, + "loss": 0.1842, + "step": 38473 + }, + { + "epoch": 2.6697661508569843, + "grad_norm": 0.9433256387710571, + "learning_rate": 5.261058922682089e-05, + "loss": 0.3605, + "step": 38474 + }, + { + "epoch": 2.6698355422940807, + "grad_norm": 0.9047799110412598, + "learning_rate": 5.260564165088386e-05, + "loss": 0.3661, + "step": 38475 + }, + { + "epoch": 2.6699049337311775, + "grad_norm": 0.8179202675819397, + "learning_rate": 5.2600694224565303e-05, + "loss": 0.296, + "step": 38476 + }, + { + "epoch": 2.6699743251682744, + "grad_norm": 1.5214890241622925, + "learning_rate": 5.2595746947880874e-05, + "loss": 0.4888, + "step": 38477 + }, + { + "epoch": 2.6700437166053708, + "grad_norm": 1.1583439111709595, + "learning_rate": 5.2590799820846174e-05, + "loss": 0.4833, + "step": 38478 + }, + { + "epoch": 2.6701131080424676, + "grad_norm": 0.9849352240562439, + "learning_rate": 5.2585852843476814e-05, + "loss": 0.4229, + "step": 38479 + }, + { + "epoch": 2.670182499479564, + "grad_norm": 1.1818478107452393, + "learning_rate": 5.258090601578844e-05, + "loss": 0.3806, + "step": 38480 + }, + { + "epoch": 2.670251890916661, + "grad_norm": 1.2734382152557373, + "learning_rate": 5.257595933779662e-05, + "loss": 0.3897, + "step": 38481 + }, + { + "epoch": 2.6703212823537577, + "grad_norm": 1.208622932434082, + "learning_rate": 5.257101280951699e-05, + "loss": 0.4138, + "step": 38482 + }, + { + "epoch": 2.6703906737908545, + "grad_norm": 1.0448012351989746, + "learning_rate": 5.2566066430965165e-05, + "loss": 0.311, + "step": 38483 + }, + { + "epoch": 2.670460065227951, + "grad_norm": 1.2693246603012085, + "learning_rate": 5.2561120202156775e-05, + "loss": 0.4499, + "step": 38484 + }, + { + "epoch": 2.6705294566650473, + "grad_norm": 1.0808873176574707, + "learning_rate": 5.2556174123107415e-05, + "loss": 0.4002, + "step": 38485 + }, + { + "epoch": 2.670598848102144, + "grad_norm": 0.9573160409927368, + "learning_rate": 5.255122819383271e-05, + "loss": 0.35, + "step": 38486 + }, + { + "epoch": 2.670668239539241, + "grad_norm": 1.3044788837432861, + "learning_rate": 5.254628241434831e-05, + "loss": 0.4564, + "step": 38487 + }, + { + "epoch": 2.6707376309763378, + "grad_norm": 1.1858532428741455, + "learning_rate": 5.254133678466977e-05, + "loss": 0.2625, + "step": 38488 + }, + { + "epoch": 2.670807022413434, + "grad_norm": 1.5987128019332886, + "learning_rate": 5.25363913048127e-05, + "loss": 0.5791, + "step": 38489 + }, + { + "epoch": 2.670876413850531, + "grad_norm": 0.9575749635696411, + "learning_rate": 5.2531445974792756e-05, + "loss": 0.2205, + "step": 38490 + }, + { + "epoch": 2.6709458052876274, + "grad_norm": 1.4582455158233643, + "learning_rate": 5.2526500794625514e-05, + "loss": 0.4353, + "step": 38491 + }, + { + "epoch": 2.671015196724724, + "grad_norm": 1.0711731910705566, + "learning_rate": 5.25215557643266e-05, + "loss": 0.379, + "step": 38492 + }, + { + "epoch": 2.671084588161821, + "grad_norm": 0.8830723166465759, + "learning_rate": 5.2516610883911676e-05, + "loss": 0.3794, + "step": 38493 + }, + { + "epoch": 2.6711539795989174, + "grad_norm": 1.5459235906600952, + "learning_rate": 5.2511666153396265e-05, + "loss": 0.4546, + "step": 38494 + }, + { + "epoch": 2.6712233710360143, + "grad_norm": 1.4462049007415771, + "learning_rate": 5.2506721572796e-05, + "loss": 0.3922, + "step": 38495 + }, + { + "epoch": 2.6712927624731106, + "grad_norm": 1.429117202758789, + "learning_rate": 5.250177714212652e-05, + "loss": 0.479, + "step": 38496 + }, + { + "epoch": 2.6713621539102075, + "grad_norm": 1.4205893278121948, + "learning_rate": 5.249683286140345e-05, + "loss": 0.4061, + "step": 38497 + }, + { + "epoch": 2.6714315453473043, + "grad_norm": 1.6991785764694214, + "learning_rate": 5.2491888730642325e-05, + "loss": 0.3678, + "step": 38498 + }, + { + "epoch": 2.6715009367844007, + "grad_norm": 1.3786238431930542, + "learning_rate": 5.24869447498588e-05, + "loss": 0.4683, + "step": 38499 + }, + { + "epoch": 2.6715703282214975, + "grad_norm": 0.9004248976707458, + "learning_rate": 5.2482000919068475e-05, + "loss": 0.3135, + "step": 38500 + }, + { + "epoch": 2.671639719658594, + "grad_norm": 1.1632131338119507, + "learning_rate": 5.2477057238287e-05, + "loss": 0.4105, + "step": 38501 + }, + { + "epoch": 2.6717091110956908, + "grad_norm": 1.3783807754516602, + "learning_rate": 5.247211370752991e-05, + "loss": 0.4593, + "step": 38502 + }, + { + "epoch": 2.6717785025327876, + "grad_norm": 1.100756049156189, + "learning_rate": 5.246717032681283e-05, + "loss": 0.5513, + "step": 38503 + }, + { + "epoch": 2.671847893969884, + "grad_norm": 1.2025740146636963, + "learning_rate": 5.2462227096151387e-05, + "loss": 0.5057, + "step": 38504 + }, + { + "epoch": 2.671917285406981, + "grad_norm": 2.221337080001831, + "learning_rate": 5.245728401556118e-05, + "loss": 0.5758, + "step": 38505 + }, + { + "epoch": 2.671986676844077, + "grad_norm": 2.979201316833496, + "learning_rate": 5.245234108505781e-05, + "loss": 0.3711, + "step": 38506 + }, + { + "epoch": 2.672056068281174, + "grad_norm": 0.9151641130447388, + "learning_rate": 5.244739830465687e-05, + "loss": 0.3442, + "step": 38507 + }, + { + "epoch": 2.672125459718271, + "grad_norm": 0.9654542803764343, + "learning_rate": 5.2442455674374024e-05, + "loss": 0.3386, + "step": 38508 + }, + { + "epoch": 2.6721948511553677, + "grad_norm": 1.0912156105041504, + "learning_rate": 5.2437513194224786e-05, + "loss": 0.387, + "step": 38509 + }, + { + "epoch": 2.672264242592464, + "grad_norm": 1.6101047992706299, + "learning_rate": 5.243257086422482e-05, + "loss": 0.2789, + "step": 38510 + }, + { + "epoch": 2.6723336340295605, + "grad_norm": 0.9772582054138184, + "learning_rate": 5.242762868438968e-05, + "loss": 0.4057, + "step": 38511 + }, + { + "epoch": 2.6724030254666573, + "grad_norm": 1.1753499507904053, + "learning_rate": 5.242268665473502e-05, + "loss": 0.3329, + "step": 38512 + }, + { + "epoch": 2.672472416903754, + "grad_norm": 1.9622701406478882, + "learning_rate": 5.24177447752764e-05, + "loss": 0.5964, + "step": 38513 + }, + { + "epoch": 2.672541808340851, + "grad_norm": 1.1404545307159424, + "learning_rate": 5.2412803046029446e-05, + "loss": 0.5745, + "step": 38514 + }, + { + "epoch": 2.6726111997779474, + "grad_norm": 1.3804343938827515, + "learning_rate": 5.24078614670098e-05, + "loss": 0.4898, + "step": 38515 + }, + { + "epoch": 2.672680591215044, + "grad_norm": 1.524827003479004, + "learning_rate": 5.240292003823296e-05, + "loss": 0.3358, + "step": 38516 + }, + { + "epoch": 2.6727499826521406, + "grad_norm": 1.1101086139678955, + "learning_rate": 5.239797875971459e-05, + "loss": 0.4094, + "step": 38517 + }, + { + "epoch": 2.6728193740892374, + "grad_norm": 0.9017484784126282, + "learning_rate": 5.239303763147028e-05, + "loss": 0.2282, + "step": 38518 + }, + { + "epoch": 2.6728887655263343, + "grad_norm": 0.9432478547096252, + "learning_rate": 5.238809665351561e-05, + "loss": 0.2723, + "step": 38519 + }, + { + "epoch": 2.6729581569634306, + "grad_norm": 2.490875482559204, + "learning_rate": 5.238315582586622e-05, + "loss": 0.5343, + "step": 38520 + }, + { + "epoch": 2.6730275484005275, + "grad_norm": 2.1190757751464844, + "learning_rate": 5.2378215148537656e-05, + "loss": 0.4394, + "step": 38521 + }, + { + "epoch": 2.673096939837624, + "grad_norm": 1.6099234819412231, + "learning_rate": 5.23732746215456e-05, + "loss": 0.4861, + "step": 38522 + }, + { + "epoch": 2.6731663312747207, + "grad_norm": 0.9903795123100281, + "learning_rate": 5.2368334244905534e-05, + "loss": 0.3318, + "step": 38523 + }, + { + "epoch": 2.6732357227118175, + "grad_norm": 1.1586377620697021, + "learning_rate": 5.236339401863313e-05, + "loss": 0.3733, + "step": 38524 + }, + { + "epoch": 2.673305114148914, + "grad_norm": 1.488677740097046, + "learning_rate": 5.2358453942743945e-05, + "loss": 0.2763, + "step": 38525 + }, + { + "epoch": 2.6733745055860108, + "grad_norm": 1.324015498161316, + "learning_rate": 5.235351401725361e-05, + "loss": 0.5453, + "step": 38526 + }, + { + "epoch": 2.673443897023107, + "grad_norm": 1.2032394409179688, + "learning_rate": 5.2348574242177695e-05, + "loss": 0.4604, + "step": 38527 + }, + { + "epoch": 2.673513288460204, + "grad_norm": 1.1021459102630615, + "learning_rate": 5.234363461753185e-05, + "loss": 0.305, + "step": 38528 + }, + { + "epoch": 2.673582679897301, + "grad_norm": 1.9018330574035645, + "learning_rate": 5.233869514333156e-05, + "loss": 0.4522, + "step": 38529 + }, + { + "epoch": 2.673652071334397, + "grad_norm": 1.2000428438186646, + "learning_rate": 5.23337558195925e-05, + "loss": 0.3758, + "step": 38530 + }, + { + "epoch": 2.673721462771494, + "grad_norm": 2.080944299697876, + "learning_rate": 5.232881664633025e-05, + "loss": 0.5725, + "step": 38531 + }, + { + "epoch": 2.6737908542085904, + "grad_norm": 1.0675878524780273, + "learning_rate": 5.2323877623560414e-05, + "loss": 0.3628, + "step": 38532 + }, + { + "epoch": 2.6738602456456873, + "grad_norm": 0.8891000747680664, + "learning_rate": 5.231893875129853e-05, + "loss": 0.4129, + "step": 38533 + }, + { + "epoch": 2.673929637082784, + "grad_norm": 1.071301817893982, + "learning_rate": 5.231400002956023e-05, + "loss": 0.52, + "step": 38534 + }, + { + "epoch": 2.673999028519881, + "grad_norm": 1.0504944324493408, + "learning_rate": 5.2309061458361095e-05, + "loss": 0.5086, + "step": 38535 + }, + { + "epoch": 2.6740684199569773, + "grad_norm": 1.0125254392623901, + "learning_rate": 5.230412303771677e-05, + "loss": 0.2691, + "step": 38536 + }, + { + "epoch": 2.6741378113940737, + "grad_norm": 1.898679494857788, + "learning_rate": 5.2299184767642736e-05, + "loss": 0.3507, + "step": 38537 + }, + { + "epoch": 2.6742072028311705, + "grad_norm": 0.9794492125511169, + "learning_rate": 5.229424664815466e-05, + "loss": 0.2906, + "step": 38538 + }, + { + "epoch": 2.6742765942682674, + "grad_norm": 1.2333588600158691, + "learning_rate": 5.22893086792681e-05, + "loss": 0.3909, + "step": 38539 + }, + { + "epoch": 2.674345985705364, + "grad_norm": 1.420390248298645, + "learning_rate": 5.2284370860998666e-05, + "loss": 0.4374, + "step": 38540 + }, + { + "epoch": 2.6744153771424606, + "grad_norm": 1.8079578876495361, + "learning_rate": 5.2279433193361924e-05, + "loss": 0.2661, + "step": 38541 + }, + { + "epoch": 2.6744847685795574, + "grad_norm": 1.2035104036331177, + "learning_rate": 5.2274495676373524e-05, + "loss": 0.4076, + "step": 38542 + }, + { + "epoch": 2.674554160016654, + "grad_norm": 1.0295460224151611, + "learning_rate": 5.226955831004896e-05, + "loss": 0.3164, + "step": 38543 + }, + { + "epoch": 2.6746235514537506, + "grad_norm": 1.4092040061950684, + "learning_rate": 5.226462109440387e-05, + "loss": 0.4478, + "step": 38544 + }, + { + "epoch": 2.6746929428908475, + "grad_norm": 1.5240529775619507, + "learning_rate": 5.225968402945383e-05, + "loss": 0.4714, + "step": 38545 + }, + { + "epoch": 2.674762334327944, + "grad_norm": 1.123051404953003, + "learning_rate": 5.225474711521443e-05, + "loss": 0.3542, + "step": 38546 + }, + { + "epoch": 2.6748317257650407, + "grad_norm": 1.692186951637268, + "learning_rate": 5.224981035170124e-05, + "loss": 0.3803, + "step": 38547 + }, + { + "epoch": 2.674901117202137, + "grad_norm": 1.4708260297775269, + "learning_rate": 5.224487373892988e-05, + "loss": 0.3658, + "step": 38548 + }, + { + "epoch": 2.674970508639234, + "grad_norm": 1.0594269037246704, + "learning_rate": 5.223993727691594e-05, + "loss": 0.3215, + "step": 38549 + }, + { + "epoch": 2.6750399000763307, + "grad_norm": 1.47074556350708, + "learning_rate": 5.223500096567495e-05, + "loss": 0.547, + "step": 38550 + }, + { + "epoch": 2.675109291513427, + "grad_norm": 1.2282153367996216, + "learning_rate": 5.22300648052225e-05, + "loss": 0.3416, + "step": 38551 + }, + { + "epoch": 2.675178682950524, + "grad_norm": 1.3136025667190552, + "learning_rate": 5.222512879557421e-05, + "loss": 0.5314, + "step": 38552 + }, + { + "epoch": 2.6752480743876204, + "grad_norm": 1.376363754272461, + "learning_rate": 5.222019293674564e-05, + "loss": 0.4329, + "step": 38553 + }, + { + "epoch": 2.675317465824717, + "grad_norm": 1.780063271522522, + "learning_rate": 5.221525722875238e-05, + "loss": 0.4648, + "step": 38554 + }, + { + "epoch": 2.675386857261814, + "grad_norm": 1.120192527770996, + "learning_rate": 5.221032167161e-05, + "loss": 0.5584, + "step": 38555 + }, + { + "epoch": 2.6754562486989104, + "grad_norm": 1.2259819507598877, + "learning_rate": 5.2205386265334136e-05, + "loss": 0.2998, + "step": 38556 + }, + { + "epoch": 2.6755256401360072, + "grad_norm": 1.4707971811294556, + "learning_rate": 5.220045100994028e-05, + "loss": 0.371, + "step": 38557 + }, + { + "epoch": 2.6755950315731036, + "grad_norm": 1.1406587362289429, + "learning_rate": 5.219551590544405e-05, + "loss": 0.4662, + "step": 38558 + }, + { + "epoch": 2.6756644230102005, + "grad_norm": 1.1880515813827515, + "learning_rate": 5.219058095186105e-05, + "loss": 0.5291, + "step": 38559 + }, + { + "epoch": 2.6757338144472973, + "grad_norm": 1.0211421251296997, + "learning_rate": 5.2185646149206815e-05, + "loss": 0.3305, + "step": 38560 + }, + { + "epoch": 2.675803205884394, + "grad_norm": 1.2622939348220825, + "learning_rate": 5.218071149749695e-05, + "loss": 0.5161, + "step": 38561 + }, + { + "epoch": 2.6758725973214905, + "grad_norm": 0.8907747864723206, + "learning_rate": 5.217577699674705e-05, + "loss": 0.3948, + "step": 38562 + }, + { + "epoch": 2.6759419887585874, + "grad_norm": 0.8765878677368164, + "learning_rate": 5.2170842646972696e-05, + "loss": 0.3132, + "step": 38563 + }, + { + "epoch": 2.6760113801956837, + "grad_norm": 1.297363042831421, + "learning_rate": 5.21659084481894e-05, + "loss": 0.4661, + "step": 38564 + }, + { + "epoch": 2.6760807716327806, + "grad_norm": 0.9116132259368896, + "learning_rate": 5.216097440041279e-05, + "loss": 0.3092, + "step": 38565 + }, + { + "epoch": 2.6761501630698774, + "grad_norm": 1.1790156364440918, + "learning_rate": 5.215604050365843e-05, + "loss": 0.382, + "step": 38566 + }, + { + "epoch": 2.676219554506974, + "grad_norm": 0.9538807272911072, + "learning_rate": 5.2151106757941947e-05, + "loss": 0.3829, + "step": 38567 + }, + { + "epoch": 2.6762889459440706, + "grad_norm": 1.067317008972168, + "learning_rate": 5.214617316327882e-05, + "loss": 0.2806, + "step": 38568 + }, + { + "epoch": 2.676358337381167, + "grad_norm": 1.4287513494491577, + "learning_rate": 5.2141239719684676e-05, + "loss": 0.5595, + "step": 38569 + }, + { + "epoch": 2.676427728818264, + "grad_norm": 0.8896589875221252, + "learning_rate": 5.213630642717513e-05, + "loss": 0.2714, + "step": 38570 + }, + { + "epoch": 2.6764971202553607, + "grad_norm": 0.7798924446105957, + "learning_rate": 5.2131373285765674e-05, + "loss": 0.2361, + "step": 38571 + }, + { + "epoch": 2.676566511692457, + "grad_norm": 1.1628015041351318, + "learning_rate": 5.21264402954719e-05, + "loss": 0.4033, + "step": 38572 + }, + { + "epoch": 2.676635903129554, + "grad_norm": 0.8551054000854492, + "learning_rate": 5.212150745630941e-05, + "loss": 0.2889, + "step": 38573 + }, + { + "epoch": 2.6767052945666503, + "grad_norm": 0.9524736404418945, + "learning_rate": 5.211657476829377e-05, + "loss": 0.334, + "step": 38574 + }, + { + "epoch": 2.676774686003747, + "grad_norm": 1.37167489528656, + "learning_rate": 5.211164223144054e-05, + "loss": 0.3137, + "step": 38575 + }, + { + "epoch": 2.676844077440844, + "grad_norm": 1.1867351531982422, + "learning_rate": 5.210670984576531e-05, + "loss": 0.3417, + "step": 38576 + }, + { + "epoch": 2.6769134688779403, + "grad_norm": 1.1277587413787842, + "learning_rate": 5.2101777611283675e-05, + "loss": 0.3527, + "step": 38577 + }, + { + "epoch": 2.676982860315037, + "grad_norm": 1.3578227758407593, + "learning_rate": 5.209684552801113e-05, + "loss": 0.4099, + "step": 38578 + }, + { + "epoch": 2.6770522517521336, + "grad_norm": 1.0709844827651978, + "learning_rate": 5.2091913595963285e-05, + "loss": 0.4641, + "step": 38579 + }, + { + "epoch": 2.6771216431892304, + "grad_norm": 1.7442412376403809, + "learning_rate": 5.208698181515571e-05, + "loss": 0.4023, + "step": 38580 + }, + { + "epoch": 2.6771910346263272, + "grad_norm": 1.7251012325286865, + "learning_rate": 5.208205018560397e-05, + "loss": 0.4118, + "step": 38581 + }, + { + "epoch": 2.6772604260634236, + "grad_norm": 1.3620671033859253, + "learning_rate": 5.207711870732366e-05, + "loss": 0.4422, + "step": 38582 + }, + { + "epoch": 2.6773298175005205, + "grad_norm": 2.4304511547088623, + "learning_rate": 5.2072187380330306e-05, + "loss": 0.5227, + "step": 38583 + }, + { + "epoch": 2.677399208937617, + "grad_norm": 1.2964507341384888, + "learning_rate": 5.2067256204639536e-05, + "loss": 0.3984, + "step": 38584 + }, + { + "epoch": 2.6774686003747137, + "grad_norm": 1.042051076889038, + "learning_rate": 5.206232518026683e-05, + "loss": 0.368, + "step": 38585 + }, + { + "epoch": 2.6775379918118105, + "grad_norm": 1.189184546470642, + "learning_rate": 5.2057394307227826e-05, + "loss": 0.3948, + "step": 38586 + }, + { + "epoch": 2.6776073832489073, + "grad_norm": 1.4218180179595947, + "learning_rate": 5.205246358553805e-05, + "loss": 0.483, + "step": 38587 + }, + { + "epoch": 2.6776767746860037, + "grad_norm": 0.8297024965286255, + "learning_rate": 5.204753301521308e-05, + "loss": 0.336, + "step": 38588 + }, + { + "epoch": 2.6777461661231006, + "grad_norm": 1.31866455078125, + "learning_rate": 5.20426025962685e-05, + "loss": 0.3464, + "step": 38589 + }, + { + "epoch": 2.677815557560197, + "grad_norm": 1.6460413932800293, + "learning_rate": 5.203767232871984e-05, + "loss": 0.4912, + "step": 38590 + }, + { + "epoch": 2.677884948997294, + "grad_norm": 1.2339143753051758, + "learning_rate": 5.203274221258274e-05, + "loss": 0.5278, + "step": 38591 + }, + { + "epoch": 2.6779543404343906, + "grad_norm": 0.8833408951759338, + "learning_rate": 5.2027812247872655e-05, + "loss": 0.4216, + "step": 38592 + }, + { + "epoch": 2.678023731871487, + "grad_norm": 2.0180139541625977, + "learning_rate": 5.2022882434605216e-05, + "loss": 0.4271, + "step": 38593 + }, + { + "epoch": 2.678093123308584, + "grad_norm": 1.0296069383621216, + "learning_rate": 5.2017952772795955e-05, + "loss": 0.3095, + "step": 38594 + }, + { + "epoch": 2.6781625147456802, + "grad_norm": 1.0998499393463135, + "learning_rate": 5.201302326246047e-05, + "loss": 0.3963, + "step": 38595 + }, + { + "epoch": 2.678231906182777, + "grad_norm": 1.5737831592559814, + "learning_rate": 5.200809390361428e-05, + "loss": 0.6666, + "step": 38596 + }, + { + "epoch": 2.678301297619874, + "grad_norm": 0.936774730682373, + "learning_rate": 5.2003164696272974e-05, + "loss": 0.2828, + "step": 38597 + }, + { + "epoch": 2.6783706890569703, + "grad_norm": 1.4864097833633423, + "learning_rate": 5.1998235640452164e-05, + "loss": 0.3719, + "step": 38598 + }, + { + "epoch": 2.678440080494067, + "grad_norm": 1.858940839767456, + "learning_rate": 5.199330673616731e-05, + "loss": 0.2778, + "step": 38599 + }, + { + "epoch": 2.6785094719311635, + "grad_norm": 0.9634416699409485, + "learning_rate": 5.198837798343401e-05, + "loss": 0.2041, + "step": 38600 + }, + { + "epoch": 2.6785788633682603, + "grad_norm": 1.1583331823349, + "learning_rate": 5.198344938226782e-05, + "loss": 0.4027, + "step": 38601 + }, + { + "epoch": 2.678648254805357, + "grad_norm": 1.2750425338745117, + "learning_rate": 5.197852093268437e-05, + "loss": 0.2, + "step": 38602 + }, + { + "epoch": 2.6787176462424536, + "grad_norm": 1.3136520385742188, + "learning_rate": 5.1973592634699105e-05, + "loss": 0.4037, + "step": 38603 + }, + { + "epoch": 2.6787870376795504, + "grad_norm": 1.2589504718780518, + "learning_rate": 5.1968664488327644e-05, + "loss": 0.3408, + "step": 38604 + }, + { + "epoch": 2.678856429116647, + "grad_norm": 1.1320865154266357, + "learning_rate": 5.196373649358557e-05, + "loss": 0.3187, + "step": 38605 + }, + { + "epoch": 2.6789258205537436, + "grad_norm": 1.0202690362930298, + "learning_rate": 5.195880865048837e-05, + "loss": 0.2623, + "step": 38606 + }, + { + "epoch": 2.6789952119908405, + "grad_norm": 0.9775280356407166, + "learning_rate": 5.195388095905165e-05, + "loss": 0.4101, + "step": 38607 + }, + { + "epoch": 2.679064603427937, + "grad_norm": 1.1022697687149048, + "learning_rate": 5.1948953419290936e-05, + "loss": 0.3636, + "step": 38608 + }, + { + "epoch": 2.6791339948650337, + "grad_norm": 1.8179444074630737, + "learning_rate": 5.1944026031221804e-05, + "loss": 0.4755, + "step": 38609 + }, + { + "epoch": 2.67920338630213, + "grad_norm": 1.0869474411010742, + "learning_rate": 5.193909879485981e-05, + "loss": 0.364, + "step": 38610 + }, + { + "epoch": 2.679272777739227, + "grad_norm": 1.46116304397583, + "learning_rate": 5.1934171710220533e-05, + "loss": 0.4251, + "step": 38611 + }, + { + "epoch": 2.6793421691763237, + "grad_norm": 0.8401824831962585, + "learning_rate": 5.192924477731948e-05, + "loss": 0.2286, + "step": 38612 + }, + { + "epoch": 2.6794115606134206, + "grad_norm": 1.3624658584594727, + "learning_rate": 5.1924317996172214e-05, + "loss": 0.3568, + "step": 38613 + }, + { + "epoch": 2.679480952050517, + "grad_norm": 0.7062464356422424, + "learning_rate": 5.191939136679429e-05, + "loss": 0.2361, + "step": 38614 + }, + { + "epoch": 2.679550343487614, + "grad_norm": 1.237763524055481, + "learning_rate": 5.191446488920128e-05, + "loss": 0.415, + "step": 38615 + }, + { + "epoch": 2.67961973492471, + "grad_norm": 1.2594046592712402, + "learning_rate": 5.190953856340872e-05, + "loss": 0.3017, + "step": 38616 + }, + { + "epoch": 2.679689126361807, + "grad_norm": 1.2769180536270142, + "learning_rate": 5.190461238943216e-05, + "loss": 0.4549, + "step": 38617 + }, + { + "epoch": 2.679758517798904, + "grad_norm": 1.0631860494613647, + "learning_rate": 5.189968636728722e-05, + "loss": 0.3474, + "step": 38618 + }, + { + "epoch": 2.6798279092360002, + "grad_norm": 1.1364415884017944, + "learning_rate": 5.189476049698934e-05, + "loss": 0.4556, + "step": 38619 + }, + { + "epoch": 2.679897300673097, + "grad_norm": 1.5667182207107544, + "learning_rate": 5.1889834778554113e-05, + "loss": 0.5845, + "step": 38620 + }, + { + "epoch": 2.6799666921101934, + "grad_norm": 1.1864346265792847, + "learning_rate": 5.188490921199709e-05, + "loss": 0.32, + "step": 38621 + }, + { + "epoch": 2.6800360835472903, + "grad_norm": 1.0717856884002686, + "learning_rate": 5.187998379733384e-05, + "loss": 0.2958, + "step": 38622 + }, + { + "epoch": 2.680105474984387, + "grad_norm": 1.3181266784667969, + "learning_rate": 5.18750585345799e-05, + "loss": 0.4064, + "step": 38623 + }, + { + "epoch": 2.6801748664214835, + "grad_norm": 1.3705497980117798, + "learning_rate": 5.187013342375081e-05, + "loss": 0.3317, + "step": 38624 + }, + { + "epoch": 2.6802442578585803, + "grad_norm": 1.4217662811279297, + "learning_rate": 5.186520846486216e-05, + "loss": 0.4878, + "step": 38625 + }, + { + "epoch": 2.6803136492956767, + "grad_norm": 1.1728971004486084, + "learning_rate": 5.186028365792943e-05, + "loss": 0.3242, + "step": 38626 + }, + { + "epoch": 2.6803830407327736, + "grad_norm": 1.3155461549758911, + "learning_rate": 5.185535900296819e-05, + "loss": 0.4542, + "step": 38627 + }, + { + "epoch": 2.6804524321698704, + "grad_norm": 1.0515069961547852, + "learning_rate": 5.185043449999401e-05, + "loss": 0.3917, + "step": 38628 + }, + { + "epoch": 2.680521823606967, + "grad_norm": 1.1636754274368286, + "learning_rate": 5.1845510149022414e-05, + "loss": 0.4403, + "step": 38629 + }, + { + "epoch": 2.6805912150440636, + "grad_norm": 1.128145694732666, + "learning_rate": 5.1840585950068954e-05, + "loss": 0.3675, + "step": 38630 + }, + { + "epoch": 2.68066060648116, + "grad_norm": 1.0916831493377686, + "learning_rate": 5.1835661903149194e-05, + "loss": 0.3513, + "step": 38631 + }, + { + "epoch": 2.680729997918257, + "grad_norm": 1.4223525524139404, + "learning_rate": 5.1830738008278665e-05, + "loss": 0.4133, + "step": 38632 + }, + { + "epoch": 2.6807993893553537, + "grad_norm": 1.576707124710083, + "learning_rate": 5.1825814265472905e-05, + "loss": 0.5317, + "step": 38633 + }, + { + "epoch": 2.68086878079245, + "grad_norm": 1.1542158126831055, + "learning_rate": 5.182089067474744e-05, + "loss": 0.3763, + "step": 38634 + }, + { + "epoch": 2.680938172229547, + "grad_norm": 1.8746966123580933, + "learning_rate": 5.181596723611784e-05, + "loss": 0.4816, + "step": 38635 + }, + { + "epoch": 2.6810075636666433, + "grad_norm": 1.427519679069519, + "learning_rate": 5.181104394959968e-05, + "loss": 0.3273, + "step": 38636 + }, + { + "epoch": 2.68107695510374, + "grad_norm": 1.3527836799621582, + "learning_rate": 5.1806120815208415e-05, + "loss": 0.5215, + "step": 38637 + }, + { + "epoch": 2.681146346540837, + "grad_norm": 0.9880020022392273, + "learning_rate": 5.1801197832959644e-05, + "loss": 0.3425, + "step": 38638 + }, + { + "epoch": 2.6812157379779338, + "grad_norm": 1.8549598455429077, + "learning_rate": 5.1796275002868896e-05, + "loss": 0.6053, + "step": 38639 + }, + { + "epoch": 2.68128512941503, + "grad_norm": 1.4232255220413208, + "learning_rate": 5.179135232495176e-05, + "loss": 0.4044, + "step": 38640 + }, + { + "epoch": 2.681354520852127, + "grad_norm": 0.9515239596366882, + "learning_rate": 5.178642979922368e-05, + "loss": 0.3371, + "step": 38641 + }, + { + "epoch": 2.6814239122892234, + "grad_norm": 1.2914822101593018, + "learning_rate": 5.178150742570025e-05, + "loss": 0.359, + "step": 38642 + }, + { + "epoch": 2.68149330372632, + "grad_norm": 0.9768327474594116, + "learning_rate": 5.177658520439701e-05, + "loss": 0.2266, + "step": 38643 + }, + { + "epoch": 2.681562695163417, + "grad_norm": 0.9271623492240906, + "learning_rate": 5.177166313532951e-05, + "loss": 0.3833, + "step": 38644 + }, + { + "epoch": 2.6816320866005134, + "grad_norm": 1.549643635749817, + "learning_rate": 5.176674121851326e-05, + "loss": 0.5722, + "step": 38645 + }, + { + "epoch": 2.6817014780376103, + "grad_norm": 1.461166262626648, + "learning_rate": 5.176181945396385e-05, + "loss": 0.2661, + "step": 38646 + }, + { + "epoch": 2.6817708694747067, + "grad_norm": 1.7368324995040894, + "learning_rate": 5.175689784169674e-05, + "loss": 0.6853, + "step": 38647 + }, + { + "epoch": 2.6818402609118035, + "grad_norm": 1.1097662448883057, + "learning_rate": 5.175197638172751e-05, + "loss": 0.3445, + "step": 38648 + }, + { + "epoch": 2.6819096523489003, + "grad_norm": 1.2437914609909058, + "learning_rate": 5.174705507407169e-05, + "loss": 0.3434, + "step": 38649 + }, + { + "epoch": 2.6819790437859967, + "grad_norm": 1.008392095565796, + "learning_rate": 5.1742133918744814e-05, + "loss": 0.301, + "step": 38650 + }, + { + "epoch": 2.6820484352230936, + "grad_norm": 1.3872438669204712, + "learning_rate": 5.173721291576244e-05, + "loss": 0.3304, + "step": 38651 + }, + { + "epoch": 2.68211782666019, + "grad_norm": 1.5162832736968994, + "learning_rate": 5.173229206514006e-05, + "loss": 0.407, + "step": 38652 + }, + { + "epoch": 2.6821872180972868, + "grad_norm": 1.1312822103500366, + "learning_rate": 5.172737136689329e-05, + "loss": 0.325, + "step": 38653 + }, + { + "epoch": 2.6822566095343836, + "grad_norm": 1.3658792972564697, + "learning_rate": 5.172245082103757e-05, + "loss": 0.4391, + "step": 38654 + }, + { + "epoch": 2.68232600097148, + "grad_norm": 0.8108816146850586, + "learning_rate": 5.171753042758847e-05, + "loss": 0.2303, + "step": 38655 + }, + { + "epoch": 2.682395392408577, + "grad_norm": 1.0616849660873413, + "learning_rate": 5.171261018656152e-05, + "loss": 0.4006, + "step": 38656 + }, + { + "epoch": 2.682464783845673, + "grad_norm": 1.288050651550293, + "learning_rate": 5.170769009797225e-05, + "loss": 0.4156, + "step": 38657 + }, + { + "epoch": 2.68253417528277, + "grad_norm": 1.325414776802063, + "learning_rate": 5.170277016183622e-05, + "loss": 0.5081, + "step": 38658 + }, + { + "epoch": 2.682603566719867, + "grad_norm": 0.8125797510147095, + "learning_rate": 5.169785037816892e-05, + "loss": 0.2943, + "step": 38659 + }, + { + "epoch": 2.6826729581569633, + "grad_norm": 1.1873605251312256, + "learning_rate": 5.169293074698596e-05, + "loss": 0.3845, + "step": 38660 + }, + { + "epoch": 2.68274234959406, + "grad_norm": 1.3442820310592651, + "learning_rate": 5.168801126830276e-05, + "loss": 0.3072, + "step": 38661 + }, + { + "epoch": 2.6828117410311565, + "grad_norm": 1.6644819974899292, + "learning_rate": 5.168309194213491e-05, + "loss": 0.4919, + "step": 38662 + }, + { + "epoch": 2.6828811324682533, + "grad_norm": 1.1118435859680176, + "learning_rate": 5.1678172768497936e-05, + "loss": 0.3202, + "step": 38663 + }, + { + "epoch": 2.68295052390535, + "grad_norm": 0.803609311580658, + "learning_rate": 5.167325374740735e-05, + "loss": 0.325, + "step": 38664 + }, + { + "epoch": 2.683019915342447, + "grad_norm": 1.2969502210617065, + "learning_rate": 5.16683348788787e-05, + "loss": 0.4671, + "step": 38665 + }, + { + "epoch": 2.6830893067795434, + "grad_norm": 1.5663069486618042, + "learning_rate": 5.1663416162927514e-05, + "loss": 0.6971, + "step": 38666 + }, + { + "epoch": 2.68315869821664, + "grad_norm": 1.2378647327423096, + "learning_rate": 5.1658497599569357e-05, + "loss": 0.3976, + "step": 38667 + }, + { + "epoch": 2.6832280896537366, + "grad_norm": 1.0732126235961914, + "learning_rate": 5.1653579188819675e-05, + "loss": 0.3972, + "step": 38668 + }, + { + "epoch": 2.6832974810908334, + "grad_norm": 1.2677018642425537, + "learning_rate": 5.164866093069403e-05, + "loss": 0.377, + "step": 38669 + }, + { + "epoch": 2.6833668725279303, + "grad_norm": 1.1219849586486816, + "learning_rate": 5.1643742825207954e-05, + "loss": 0.451, + "step": 38670 + }, + { + "epoch": 2.6834362639650267, + "grad_norm": 1.1392484903335571, + "learning_rate": 5.163882487237701e-05, + "loss": 0.4964, + "step": 38671 + }, + { + "epoch": 2.6835056554021235, + "grad_norm": 1.2360270023345947, + "learning_rate": 5.1633907072216646e-05, + "loss": 0.4283, + "step": 38672 + }, + { + "epoch": 2.68357504683922, + "grad_norm": 1.3438498973846436, + "learning_rate": 5.162898942474242e-05, + "loss": 0.3082, + "step": 38673 + }, + { + "epoch": 2.6836444382763167, + "grad_norm": 1.0140914916992188, + "learning_rate": 5.162407192996987e-05, + "loss": 0.2601, + "step": 38674 + }, + { + "epoch": 2.6837138297134135, + "grad_norm": 1.041087031364441, + "learning_rate": 5.1619154587914553e-05, + "loss": 0.3766, + "step": 38675 + }, + { + "epoch": 2.68378322115051, + "grad_norm": 0.9570097327232361, + "learning_rate": 5.1614237398591905e-05, + "loss": 0.3572, + "step": 38676 + }, + { + "epoch": 2.6838526125876068, + "grad_norm": 1.2115005254745483, + "learning_rate": 5.160932036201751e-05, + "loss": 0.3478, + "step": 38677 + }, + { + "epoch": 2.683922004024703, + "grad_norm": 0.8900604248046875, + "learning_rate": 5.1604403478206855e-05, + "loss": 0.3575, + "step": 38678 + }, + { + "epoch": 2.6839913954618, + "grad_norm": 1.0014642477035522, + "learning_rate": 5.1599486747175496e-05, + "loss": 0.3609, + "step": 38679 + }, + { + "epoch": 2.684060786898897, + "grad_norm": 1.2217278480529785, + "learning_rate": 5.159457016893894e-05, + "loss": 0.5291, + "step": 38680 + }, + { + "epoch": 2.684130178335993, + "grad_norm": 1.4901264905929565, + "learning_rate": 5.158965374351276e-05, + "loss": 0.4538, + "step": 38681 + }, + { + "epoch": 2.68419956977309, + "grad_norm": 1.3057023286819458, + "learning_rate": 5.158473747091237e-05, + "loss": 0.3687, + "step": 38682 + }, + { + "epoch": 2.6842689612101864, + "grad_norm": 1.1760472059249878, + "learning_rate": 5.157982135115336e-05, + "loss": 0.324, + "step": 38683 + }, + { + "epoch": 2.6843383526472833, + "grad_norm": 1.167775273323059, + "learning_rate": 5.1574905384251225e-05, + "loss": 0.2623, + "step": 38684 + }, + { + "epoch": 2.68440774408438, + "grad_norm": 1.2645076513290405, + "learning_rate": 5.15699895702215e-05, + "loss": 0.5243, + "step": 38685 + }, + { + "epoch": 2.6844771355214765, + "grad_norm": 1.1666158437728882, + "learning_rate": 5.1565073909079694e-05, + "loss": 0.3905, + "step": 38686 + }, + { + "epoch": 2.6845465269585733, + "grad_norm": 0.8750214576721191, + "learning_rate": 5.156015840084139e-05, + "loss": 0.2551, + "step": 38687 + }, + { + "epoch": 2.6846159183956697, + "grad_norm": 0.8988392949104309, + "learning_rate": 5.155524304552197e-05, + "loss": 0.2406, + "step": 38688 + }, + { + "epoch": 2.6846853098327665, + "grad_norm": 1.454715371131897, + "learning_rate": 5.155032784313706e-05, + "loss": 0.5209, + "step": 38689 + }, + { + "epoch": 2.6847547012698634, + "grad_norm": 1.0652490854263306, + "learning_rate": 5.154541279370213e-05, + "loss": 0.3946, + "step": 38690 + }, + { + "epoch": 2.68482409270696, + "grad_norm": 1.0108574628829956, + "learning_rate": 5.154049789723271e-05, + "loss": 0.3533, + "step": 38691 + }, + { + "epoch": 2.6848934841440566, + "grad_norm": 1.287899374961853, + "learning_rate": 5.1535583153744315e-05, + "loss": 0.5315, + "step": 38692 + }, + { + "epoch": 2.6849628755811534, + "grad_norm": 0.8313099145889282, + "learning_rate": 5.153066856325246e-05, + "loss": 0.2329, + "step": 38693 + }, + { + "epoch": 2.68503226701825, + "grad_norm": 0.9390955567359924, + "learning_rate": 5.152575412577272e-05, + "loss": 0.2737, + "step": 38694 + }, + { + "epoch": 2.6851016584553467, + "grad_norm": 0.9611245393753052, + "learning_rate": 5.152083984132048e-05, + "loss": 0.3026, + "step": 38695 + }, + { + "epoch": 2.6851710498924435, + "grad_norm": 1.2875657081604004, + "learning_rate": 5.151592570991135e-05, + "loss": 0.4661, + "step": 38696 + }, + { + "epoch": 2.68524044132954, + "grad_norm": 1.132872223854065, + "learning_rate": 5.15110117315608e-05, + "loss": 0.4353, + "step": 38697 + }, + { + "epoch": 2.6853098327666367, + "grad_norm": 1.655972957611084, + "learning_rate": 5.150609790628437e-05, + "loss": 0.4493, + "step": 38698 + }, + { + "epoch": 2.685379224203733, + "grad_norm": 1.2310086488723755, + "learning_rate": 5.1501184234097556e-05, + "loss": 0.2631, + "step": 38699 + }, + { + "epoch": 2.68544861564083, + "grad_norm": 1.1517010927200317, + "learning_rate": 5.1496270715015884e-05, + "loss": 0.3518, + "step": 38700 + }, + { + "epoch": 2.6855180070779268, + "grad_norm": 1.3185638189315796, + "learning_rate": 5.149135734905488e-05, + "loss": 0.3058, + "step": 38701 + }, + { + "epoch": 2.685587398515023, + "grad_norm": 1.1729375123977661, + "learning_rate": 5.148644413623002e-05, + "loss": 0.3458, + "step": 38702 + }, + { + "epoch": 2.68565678995212, + "grad_norm": 1.0591082572937012, + "learning_rate": 5.148153107655681e-05, + "loss": 0.395, + "step": 38703 + }, + { + "epoch": 2.6857261813892164, + "grad_norm": 0.9779178500175476, + "learning_rate": 5.1476618170050785e-05, + "loss": 0.3739, + "step": 38704 + }, + { + "epoch": 2.685795572826313, + "grad_norm": 1.447955846786499, + "learning_rate": 5.1471705416727434e-05, + "loss": 0.4036, + "step": 38705 + }, + { + "epoch": 2.68586496426341, + "grad_norm": 1.4288614988327026, + "learning_rate": 5.1466792816602324e-05, + "loss": 0.4389, + "step": 38706 + }, + { + "epoch": 2.6859343557005064, + "grad_norm": 0.9324381351470947, + "learning_rate": 5.1461880369690885e-05, + "loss": 0.2216, + "step": 38707 + }, + { + "epoch": 2.6860037471376033, + "grad_norm": 1.2344636917114258, + "learning_rate": 5.145696807600866e-05, + "loss": 0.478, + "step": 38708 + }, + { + "epoch": 2.6860731385746996, + "grad_norm": 1.0779142379760742, + "learning_rate": 5.1452055935571145e-05, + "loss": 0.3082, + "step": 38709 + }, + { + "epoch": 2.6861425300117965, + "grad_norm": 1.6251623630523682, + "learning_rate": 5.144714394839389e-05, + "loss": 0.5234, + "step": 38710 + }, + { + "epoch": 2.6862119214488933, + "grad_norm": 1.4180164337158203, + "learning_rate": 5.1442232114492354e-05, + "loss": 0.4523, + "step": 38711 + }, + { + "epoch": 2.6862813128859897, + "grad_norm": 1.0792453289031982, + "learning_rate": 5.1437320433882034e-05, + "loss": 0.3815, + "step": 38712 + }, + { + "epoch": 2.6863507043230865, + "grad_norm": 1.0771615505218506, + "learning_rate": 5.143240890657848e-05, + "loss": 0.4284, + "step": 38713 + }, + { + "epoch": 2.686420095760183, + "grad_norm": 1.2547944784164429, + "learning_rate": 5.1427497532597146e-05, + "loss": 0.4682, + "step": 38714 + }, + { + "epoch": 2.6864894871972798, + "grad_norm": 1.2146320343017578, + "learning_rate": 5.142258631195362e-05, + "loss": 0.3983, + "step": 38715 + }, + { + "epoch": 2.6865588786343766, + "grad_norm": 1.0901098251342773, + "learning_rate": 5.141767524466333e-05, + "loss": 0.3721, + "step": 38716 + }, + { + "epoch": 2.6866282700714734, + "grad_norm": 1.4093095064163208, + "learning_rate": 5.1412764330741793e-05, + "loss": 0.446, + "step": 38717 + }, + { + "epoch": 2.68669766150857, + "grad_norm": 1.1809473037719727, + "learning_rate": 5.1407853570204526e-05, + "loss": 0.3671, + "step": 38718 + }, + { + "epoch": 2.6867670529456666, + "grad_norm": 1.3458316326141357, + "learning_rate": 5.140294296306701e-05, + "loss": 0.5315, + "step": 38719 + }, + { + "epoch": 2.686836444382763, + "grad_norm": 0.904160737991333, + "learning_rate": 5.139803250934478e-05, + "loss": 0.3949, + "step": 38720 + }, + { + "epoch": 2.68690583581986, + "grad_norm": 1.142762303352356, + "learning_rate": 5.139312220905332e-05, + "loss": 0.5007, + "step": 38721 + }, + { + "epoch": 2.6869752272569567, + "grad_norm": 1.0600463151931763, + "learning_rate": 5.138821206220816e-05, + "loss": 0.3284, + "step": 38722 + }, + { + "epoch": 2.687044618694053, + "grad_norm": 1.1466374397277832, + "learning_rate": 5.138330206882476e-05, + "loss": 0.4444, + "step": 38723 + }, + { + "epoch": 2.68711401013115, + "grad_norm": 1.4568164348602295, + "learning_rate": 5.137839222891862e-05, + "loss": 0.4578, + "step": 38724 + }, + { + "epoch": 2.6871834015682463, + "grad_norm": 0.9276643991470337, + "learning_rate": 5.1373482542505246e-05, + "loss": 0.3261, + "step": 38725 + }, + { + "epoch": 2.687252793005343, + "grad_norm": 1.03219473361969, + "learning_rate": 5.1368573009600164e-05, + "loss": 0.2793, + "step": 38726 + }, + { + "epoch": 2.68732218444244, + "grad_norm": 1.148784875869751, + "learning_rate": 5.136366363021885e-05, + "loss": 0.3566, + "step": 38727 + }, + { + "epoch": 2.6873915758795364, + "grad_norm": 1.5227031707763672, + "learning_rate": 5.1358754404376796e-05, + "loss": 0.3936, + "step": 38728 + }, + { + "epoch": 2.687460967316633, + "grad_norm": 1.0119777917861938, + "learning_rate": 5.135384533208956e-05, + "loss": 0.3437, + "step": 38729 + }, + { + "epoch": 2.6875303587537296, + "grad_norm": 1.119304895401001, + "learning_rate": 5.1348936413372553e-05, + "loss": 0.461, + "step": 38730 + }, + { + "epoch": 2.6875997501908264, + "grad_norm": 1.3103291988372803, + "learning_rate": 5.1344027648241296e-05, + "loss": 0.4726, + "step": 38731 + }, + { + "epoch": 2.6876691416279233, + "grad_norm": 1.2059907913208008, + "learning_rate": 5.133911903671131e-05, + "loss": 0.2851, + "step": 38732 + }, + { + "epoch": 2.6877385330650196, + "grad_norm": 1.7406107187271118, + "learning_rate": 5.133421057879808e-05, + "loss": 0.5145, + "step": 38733 + }, + { + "epoch": 2.6878079245021165, + "grad_norm": 1.3951950073242188, + "learning_rate": 5.1329302274517085e-05, + "loss": 0.4693, + "step": 38734 + }, + { + "epoch": 2.687877315939213, + "grad_norm": 0.8993955850601196, + "learning_rate": 5.132439412388386e-05, + "loss": 0.3662, + "step": 38735 + }, + { + "epoch": 2.6879467073763097, + "grad_norm": 1.1945863962173462, + "learning_rate": 5.1319486126913905e-05, + "loss": 0.3892, + "step": 38736 + }, + { + "epoch": 2.6880160988134065, + "grad_norm": 1.4866151809692383, + "learning_rate": 5.131457828362264e-05, + "loss": 0.6161, + "step": 38737 + }, + { + "epoch": 2.6880854902505034, + "grad_norm": 1.0904607772827148, + "learning_rate": 5.130967059402561e-05, + "loss": 0.4324, + "step": 38738 + }, + { + "epoch": 2.6881548816875998, + "grad_norm": 1.1261250972747803, + "learning_rate": 5.13047630581383e-05, + "loss": 0.3833, + "step": 38739 + }, + { + "epoch": 2.688224273124696, + "grad_norm": 1.1579855680465698, + "learning_rate": 5.129985567597619e-05, + "loss": 0.3507, + "step": 38740 + }, + { + "epoch": 2.688293664561793, + "grad_norm": 0.8652751445770264, + "learning_rate": 5.1294948447554835e-05, + "loss": 0.2571, + "step": 38741 + }, + { + "epoch": 2.68836305599889, + "grad_norm": 0.9803771376609802, + "learning_rate": 5.1290041372889644e-05, + "loss": 0.3242, + "step": 38742 + }, + { + "epoch": 2.6884324474359866, + "grad_norm": 1.0867934226989746, + "learning_rate": 5.128513445199613e-05, + "loss": 0.3815, + "step": 38743 + }, + { + "epoch": 2.688501838873083, + "grad_norm": 1.0013962984085083, + "learning_rate": 5.1280227684889806e-05, + "loss": 0.4179, + "step": 38744 + }, + { + "epoch": 2.68857123031018, + "grad_norm": 1.2864519357681274, + "learning_rate": 5.127532107158617e-05, + "loss": 0.4549, + "step": 38745 + }, + { + "epoch": 2.6886406217472762, + "grad_norm": 0.9392678141593933, + "learning_rate": 5.127041461210067e-05, + "loss": 0.4272, + "step": 38746 + }, + { + "epoch": 2.688710013184373, + "grad_norm": 1.7609143257141113, + "learning_rate": 5.126550830644882e-05, + "loss": 0.3298, + "step": 38747 + }, + { + "epoch": 2.68877940462147, + "grad_norm": 1.094145655632019, + "learning_rate": 5.12606021546461e-05, + "loss": 0.4628, + "step": 38748 + }, + { + "epoch": 2.6888487960585663, + "grad_norm": 1.4581005573272705, + "learning_rate": 5.1255696156708e-05, + "loss": 0.4461, + "step": 38749 + }, + { + "epoch": 2.688918187495663, + "grad_norm": 1.3471895456314087, + "learning_rate": 5.125079031265005e-05, + "loss": 0.4553, + "step": 38750 + }, + { + "epoch": 2.6889875789327595, + "grad_norm": 1.441309928894043, + "learning_rate": 5.124588462248767e-05, + "loss": 0.3516, + "step": 38751 + }, + { + "epoch": 2.6890569703698564, + "grad_norm": 2.8441061973571777, + "learning_rate": 5.124097908623637e-05, + "loss": 0.4178, + "step": 38752 + }, + { + "epoch": 2.689126361806953, + "grad_norm": 0.9418030381202698, + "learning_rate": 5.1236073703911635e-05, + "loss": 0.3023, + "step": 38753 + }, + { + "epoch": 2.6891957532440496, + "grad_norm": 1.0781303644180298, + "learning_rate": 5.1231168475528976e-05, + "loss": 0.315, + "step": 38754 + }, + { + "epoch": 2.6892651446811464, + "grad_norm": 1.0710960626602173, + "learning_rate": 5.1226263401103846e-05, + "loss": 0.3097, + "step": 38755 + }, + { + "epoch": 2.689334536118243, + "grad_norm": 1.0249736309051514, + "learning_rate": 5.122135848065175e-05, + "loss": 0.2805, + "step": 38756 + }, + { + "epoch": 2.6894039275553396, + "grad_norm": 1.011083960533142, + "learning_rate": 5.1216453714188194e-05, + "loss": 0.343, + "step": 38757 + }, + { + "epoch": 2.6894733189924365, + "grad_norm": 0.9542443752288818, + "learning_rate": 5.1211549101728606e-05, + "loss": 0.313, + "step": 38758 + }, + { + "epoch": 2.689542710429533, + "grad_norm": 1.14085054397583, + "learning_rate": 5.1206644643288495e-05, + "loss": 0.3136, + "step": 38759 + }, + { + "epoch": 2.6896121018666297, + "grad_norm": 1.6538904905319214, + "learning_rate": 5.1201740338883345e-05, + "loss": 0.4562, + "step": 38760 + }, + { + "epoch": 2.689681493303726, + "grad_norm": 0.9020410180091858, + "learning_rate": 5.119683618852863e-05, + "loss": 0.2764, + "step": 38761 + }, + { + "epoch": 2.689750884740823, + "grad_norm": 1.1394749879837036, + "learning_rate": 5.1191932192239854e-05, + "loss": 0.3897, + "step": 38762 + }, + { + "epoch": 2.6898202761779197, + "grad_norm": 1.2168543338775635, + "learning_rate": 5.118702835003248e-05, + "loss": 0.3671, + "step": 38763 + }, + { + "epoch": 2.6898896676150166, + "grad_norm": 1.0053164958953857, + "learning_rate": 5.1182124661922026e-05, + "loss": 0.3925, + "step": 38764 + }, + { + "epoch": 2.689959059052113, + "grad_norm": 1.3974287509918213, + "learning_rate": 5.1177221127923916e-05, + "loss": 0.4948, + "step": 38765 + }, + { + "epoch": 2.6900284504892094, + "grad_norm": 1.142174482345581, + "learning_rate": 5.117231774805364e-05, + "loss": 0.5644, + "step": 38766 + }, + { + "epoch": 2.690097841926306, + "grad_norm": 1.0332703590393066, + "learning_rate": 5.1167414522326695e-05, + "loss": 0.3691, + "step": 38767 + }, + { + "epoch": 2.690167233363403, + "grad_norm": 1.7599400281906128, + "learning_rate": 5.1162511450758565e-05, + "loss": 0.5691, + "step": 38768 + }, + { + "epoch": 2.6902366248005, + "grad_norm": 1.4931501150131226, + "learning_rate": 5.115760853336472e-05, + "loss": 0.5283, + "step": 38769 + }, + { + "epoch": 2.6903060162375962, + "grad_norm": 1.4450558423995972, + "learning_rate": 5.115270577016067e-05, + "loss": 0.5427, + "step": 38770 + }, + { + "epoch": 2.690375407674693, + "grad_norm": 0.8992951512336731, + "learning_rate": 5.1147803161161835e-05, + "loss": 0.4008, + "step": 38771 + }, + { + "epoch": 2.6904447991117895, + "grad_norm": 1.3158071041107178, + "learning_rate": 5.1142900706383704e-05, + "loss": 0.3236, + "step": 38772 + }, + { + "epoch": 2.6905141905488863, + "grad_norm": 0.8784991502761841, + "learning_rate": 5.113799840584177e-05, + "loss": 0.3056, + "step": 38773 + }, + { + "epoch": 2.690583581985983, + "grad_norm": 1.5148335695266724, + "learning_rate": 5.113309625955149e-05, + "loss": 0.5656, + "step": 38774 + }, + { + "epoch": 2.6906529734230795, + "grad_norm": 1.1676586866378784, + "learning_rate": 5.112819426752838e-05, + "loss": 0.4064, + "step": 38775 + }, + { + "epoch": 2.6907223648601764, + "grad_norm": 1.4870315790176392, + "learning_rate": 5.112329242978793e-05, + "loss": 0.5227, + "step": 38776 + }, + { + "epoch": 2.6907917562972727, + "grad_norm": 1.1141870021820068, + "learning_rate": 5.1118390746345525e-05, + "loss": 0.2668, + "step": 38777 + }, + { + "epoch": 2.6908611477343696, + "grad_norm": 1.3835192918777466, + "learning_rate": 5.111348921721668e-05, + "loss": 0.7029, + "step": 38778 + }, + { + "epoch": 2.6909305391714664, + "grad_norm": 0.9114719033241272, + "learning_rate": 5.1108587842416934e-05, + "loss": 0.3282, + "step": 38779 + }, + { + "epoch": 2.690999930608563, + "grad_norm": 1.0616276264190674, + "learning_rate": 5.1103686621961655e-05, + "loss": 0.3171, + "step": 38780 + }, + { + "epoch": 2.6910693220456596, + "grad_norm": 0.8025444149971008, + "learning_rate": 5.109878555586638e-05, + "loss": 0.2651, + "step": 38781 + }, + { + "epoch": 2.691138713482756, + "grad_norm": 1.6908557415008545, + "learning_rate": 5.109388464414655e-05, + "loss": 0.4602, + "step": 38782 + }, + { + "epoch": 2.691208104919853, + "grad_norm": 1.055148959159851, + "learning_rate": 5.108898388681764e-05, + "loss": 0.4386, + "step": 38783 + }, + { + "epoch": 2.6912774963569497, + "grad_norm": 0.9858805537223816, + "learning_rate": 5.10840832838952e-05, + "loss": 0.3247, + "step": 38784 + }, + { + "epoch": 2.691346887794046, + "grad_norm": 1.8041553497314453, + "learning_rate": 5.107918283539458e-05, + "loss": 0.449, + "step": 38785 + }, + { + "epoch": 2.691416279231143, + "grad_norm": 1.8362740278244019, + "learning_rate": 5.107428254133131e-05, + "loss": 0.7295, + "step": 38786 + }, + { + "epoch": 2.6914856706682393, + "grad_norm": 1.3992283344268799, + "learning_rate": 5.1069382401720846e-05, + "loss": 0.4558, + "step": 38787 + }, + { + "epoch": 2.691555062105336, + "grad_norm": 1.3706032037734985, + "learning_rate": 5.106448241657866e-05, + "loss": 0.3264, + "step": 38788 + }, + { + "epoch": 2.691624453542433, + "grad_norm": 1.039409875869751, + "learning_rate": 5.105958258592024e-05, + "loss": 0.2976, + "step": 38789 + }, + { + "epoch": 2.69169384497953, + "grad_norm": 0.9510029554367065, + "learning_rate": 5.105468290976103e-05, + "loss": 0.3529, + "step": 38790 + }, + { + "epoch": 2.691763236416626, + "grad_norm": 0.8121045827865601, + "learning_rate": 5.1049783388116555e-05, + "loss": 0.4602, + "step": 38791 + }, + { + "epoch": 2.691832627853723, + "grad_norm": 1.2409768104553223, + "learning_rate": 5.1044884021002195e-05, + "loss": 0.3907, + "step": 38792 + }, + { + "epoch": 2.6919020192908194, + "grad_norm": 1.0041015148162842, + "learning_rate": 5.103998480843345e-05, + "loss": 0.3133, + "step": 38793 + }, + { + "epoch": 2.6919714107279162, + "grad_norm": 1.3483577966690063, + "learning_rate": 5.103508575042579e-05, + "loss": 0.3344, + "step": 38794 + }, + { + "epoch": 2.692040802165013, + "grad_norm": 0.9968024492263794, + "learning_rate": 5.10301868469947e-05, + "loss": 0.4027, + "step": 38795 + }, + { + "epoch": 2.6921101936021095, + "grad_norm": 1.377393364906311, + "learning_rate": 5.102528809815561e-05, + "loss": 0.43, + "step": 38796 + }, + { + "epoch": 2.6921795850392063, + "grad_norm": 1.7276926040649414, + "learning_rate": 5.1020389503924004e-05, + "loss": 0.4181, + "step": 38797 + }, + { + "epoch": 2.6922489764763027, + "grad_norm": 1.4403382539749146, + "learning_rate": 5.101549106431539e-05, + "loss": 0.6053, + "step": 38798 + }, + { + "epoch": 2.6923183679133995, + "grad_norm": 1.178672194480896, + "learning_rate": 5.101059277934514e-05, + "loss": 0.3729, + "step": 38799 + }, + { + "epoch": 2.6923877593504963, + "grad_norm": 1.1648242473602295, + "learning_rate": 5.100569464902878e-05, + "loss": 0.4584, + "step": 38800 + }, + { + "epoch": 2.6924571507875927, + "grad_norm": 1.1487350463867188, + "learning_rate": 5.100079667338176e-05, + "loss": 0.2031, + "step": 38801 + }, + { + "epoch": 2.6925265422246896, + "grad_norm": 1.1908820867538452, + "learning_rate": 5.099589885241953e-05, + "loss": 0.5088, + "step": 38802 + }, + { + "epoch": 2.692595933661786, + "grad_norm": 1.1623345613479614, + "learning_rate": 5.0991001186157563e-05, + "loss": 0.288, + "step": 38803 + }, + { + "epoch": 2.692665325098883, + "grad_norm": 0.9347201585769653, + "learning_rate": 5.098610367461132e-05, + "loss": 0.35, + "step": 38804 + }, + { + "epoch": 2.6927347165359796, + "grad_norm": 0.9867463111877441, + "learning_rate": 5.0981206317796295e-05, + "loss": 0.2884, + "step": 38805 + }, + { + "epoch": 2.692804107973076, + "grad_norm": 0.8979681134223938, + "learning_rate": 5.097630911572788e-05, + "loss": 0.3887, + "step": 38806 + }, + { + "epoch": 2.692873499410173, + "grad_norm": 1.3987491130828857, + "learning_rate": 5.0971412068421564e-05, + "loss": 0.4482, + "step": 38807 + }, + { + "epoch": 2.6929428908472692, + "grad_norm": 0.8761946558952332, + "learning_rate": 5.096651517589282e-05, + "loss": 0.3091, + "step": 38808 + }, + { + "epoch": 2.693012282284366, + "grad_norm": 1.1658859252929688, + "learning_rate": 5.09616184381571e-05, + "loss": 0.5246, + "step": 38809 + }, + { + "epoch": 2.693081673721463, + "grad_norm": 2.452770233154297, + "learning_rate": 5.095672185522985e-05, + "loss": 0.3784, + "step": 38810 + }, + { + "epoch": 2.6931510651585593, + "grad_norm": 1.4503988027572632, + "learning_rate": 5.0951825427126586e-05, + "loss": 0.372, + "step": 38811 + }, + { + "epoch": 2.693220456595656, + "grad_norm": 1.3816627264022827, + "learning_rate": 5.094692915386268e-05, + "loss": 0.32, + "step": 38812 + }, + { + "epoch": 2.6932898480327525, + "grad_norm": 0.8918179273605347, + "learning_rate": 5.094203303545362e-05, + "loss": 0.3172, + "step": 38813 + }, + { + "epoch": 2.6933592394698493, + "grad_norm": 1.133205771446228, + "learning_rate": 5.093713707191492e-05, + "loss": 0.3242, + "step": 38814 + }, + { + "epoch": 2.693428630906946, + "grad_norm": 1.4141541719436646, + "learning_rate": 5.093224126326194e-05, + "loss": 0.4019, + "step": 38815 + }, + { + "epoch": 2.693498022344043, + "grad_norm": 1.2308475971221924, + "learning_rate": 5.0927345609510176e-05, + "loss": 0.3429, + "step": 38816 + }, + { + "epoch": 2.6935674137811394, + "grad_norm": 1.1801234483718872, + "learning_rate": 5.092245011067509e-05, + "loss": 0.3586, + "step": 38817 + }, + { + "epoch": 2.6936368052182362, + "grad_norm": 1.6195367574691772, + "learning_rate": 5.0917554766772147e-05, + "loss": 0.4858, + "step": 38818 + }, + { + "epoch": 2.6937061966553326, + "grad_norm": 1.5509417057037354, + "learning_rate": 5.0912659577816826e-05, + "loss": 0.3609, + "step": 38819 + }, + { + "epoch": 2.6937755880924295, + "grad_norm": 1.6038881540298462, + "learning_rate": 5.090776454382449e-05, + "loss": 0.6736, + "step": 38820 + }, + { + "epoch": 2.6938449795295263, + "grad_norm": 1.0034253597259521, + "learning_rate": 5.090286966481066e-05, + "loss": 0.243, + "step": 38821 + }, + { + "epoch": 2.6939143709666227, + "grad_norm": 0.993597149848938, + "learning_rate": 5.089797494079077e-05, + "loss": 0.3585, + "step": 38822 + }, + { + "epoch": 2.6939837624037195, + "grad_norm": 0.7686764597892761, + "learning_rate": 5.0893080371780265e-05, + "loss": 0.2528, + "step": 38823 + }, + { + "epoch": 2.694053153840816, + "grad_norm": 1.894943356513977, + "learning_rate": 5.088818595779463e-05, + "loss": 0.4569, + "step": 38824 + }, + { + "epoch": 2.6941225452779127, + "grad_norm": 1.0810099840164185, + "learning_rate": 5.088329169884928e-05, + "loss": 0.3873, + "step": 38825 + }, + { + "epoch": 2.6941919367150096, + "grad_norm": 0.9606112837791443, + "learning_rate": 5.087839759495971e-05, + "loss": 0.4808, + "step": 38826 + }, + { + "epoch": 2.694261328152106, + "grad_norm": 1.242720603942871, + "learning_rate": 5.087350364614132e-05, + "loss": 0.5045, + "step": 38827 + }, + { + "epoch": 2.694330719589203, + "grad_norm": 1.5643136501312256, + "learning_rate": 5.0868609852409574e-05, + "loss": 0.4644, + "step": 38828 + }, + { + "epoch": 2.694400111026299, + "grad_norm": 0.8878412246704102, + "learning_rate": 5.0863716213779924e-05, + "loss": 0.3469, + "step": 38829 + }, + { + "epoch": 2.694469502463396, + "grad_norm": 1.2225661277770996, + "learning_rate": 5.085882273026781e-05, + "loss": 0.425, + "step": 38830 + }, + { + "epoch": 2.694538893900493, + "grad_norm": 1.4275184869766235, + "learning_rate": 5.085392940188871e-05, + "loss": 0.3642, + "step": 38831 + }, + { + "epoch": 2.6946082853375892, + "grad_norm": 0.9292939305305481, + "learning_rate": 5.084903622865804e-05, + "loss": 0.458, + "step": 38832 + }, + { + "epoch": 2.694677676774686, + "grad_norm": 1.1752203702926636, + "learning_rate": 5.08441432105913e-05, + "loss": 0.4602, + "step": 38833 + }, + { + "epoch": 2.6947470682117824, + "grad_norm": 1.2071959972381592, + "learning_rate": 5.083925034770387e-05, + "loss": 0.3398, + "step": 38834 + }, + { + "epoch": 2.6948164596488793, + "grad_norm": 0.8906675577163696, + "learning_rate": 5.083435764001121e-05, + "loss": 0.3092, + "step": 38835 + }, + { + "epoch": 2.694885851085976, + "grad_norm": 1.91422438621521, + "learning_rate": 5.082946508752878e-05, + "loss": 0.4557, + "step": 38836 + }, + { + "epoch": 2.6949552425230725, + "grad_norm": 0.89192795753479, + "learning_rate": 5.082457269027203e-05, + "loss": 0.4062, + "step": 38837 + }, + { + "epoch": 2.6950246339601693, + "grad_norm": 0.9454939961433411, + "learning_rate": 5.081968044825639e-05, + "loss": 0.3461, + "step": 38838 + }, + { + "epoch": 2.6950940253972657, + "grad_norm": 1.8099138736724854, + "learning_rate": 5.081478836149731e-05, + "loss": 0.6744, + "step": 38839 + }, + { + "epoch": 2.6951634168343626, + "grad_norm": 1.8948948383331299, + "learning_rate": 5.080989643001029e-05, + "loss": 0.3289, + "step": 38840 + }, + { + "epoch": 2.6952328082714594, + "grad_norm": 1.4614747762680054, + "learning_rate": 5.080500465381068e-05, + "loss": 0.4805, + "step": 38841 + }, + { + "epoch": 2.6953021997085562, + "grad_norm": 1.1726961135864258, + "learning_rate": 5.0800113032913956e-05, + "loss": 0.3815, + "step": 38842 + }, + { + "epoch": 2.6953715911456526, + "grad_norm": 0.9049230813980103, + "learning_rate": 5.079522156733556e-05, + "loss": 0.2947, + "step": 38843 + }, + { + "epoch": 2.6954409825827494, + "grad_norm": 1.4382725954055786, + "learning_rate": 5.079033025709095e-05, + "loss": 0.3614, + "step": 38844 + }, + { + "epoch": 2.695510374019846, + "grad_norm": 2.2226977348327637, + "learning_rate": 5.07854391021956e-05, + "loss": 0.4024, + "step": 38845 + }, + { + "epoch": 2.6955797654569427, + "grad_norm": 1.163869023323059, + "learning_rate": 5.078054810266486e-05, + "loss": 0.3344, + "step": 38846 + }, + { + "epoch": 2.6956491568940395, + "grad_norm": 1.0827596187591553, + "learning_rate": 5.077565725851423e-05, + "loss": 0.357, + "step": 38847 + }, + { + "epoch": 2.695718548331136, + "grad_norm": 1.2067477703094482, + "learning_rate": 5.077076656975913e-05, + "loss": 0.4296, + "step": 38848 + }, + { + "epoch": 2.6957879397682327, + "grad_norm": 1.5850484371185303, + "learning_rate": 5.0765876036415046e-05, + "loss": 0.4291, + "step": 38849 + }, + { + "epoch": 2.695857331205329, + "grad_norm": 1.305349349975586, + "learning_rate": 5.076098565849735e-05, + "loss": 0.4549, + "step": 38850 + }, + { + "epoch": 2.695926722642426, + "grad_norm": 1.13322114944458, + "learning_rate": 5.075609543602151e-05, + "loss": 0.4012, + "step": 38851 + }, + { + "epoch": 2.6959961140795228, + "grad_norm": 1.1809728145599365, + "learning_rate": 5.0751205369002954e-05, + "loss": 0.4396, + "step": 38852 + }, + { + "epoch": 2.696065505516619, + "grad_norm": 1.3040519952774048, + "learning_rate": 5.074631545745716e-05, + "loss": 0.3421, + "step": 38853 + }, + { + "epoch": 2.696134896953716, + "grad_norm": 1.1499241590499878, + "learning_rate": 5.074142570139951e-05, + "loss": 0.4195, + "step": 38854 + }, + { + "epoch": 2.6962042883908124, + "grad_norm": 1.492773175239563, + "learning_rate": 5.073653610084547e-05, + "loss": 0.5188, + "step": 38855 + }, + { + "epoch": 2.696273679827909, + "grad_norm": 1.022739291191101, + "learning_rate": 5.073164665581045e-05, + "loss": 0.3692, + "step": 38856 + }, + { + "epoch": 2.696343071265006, + "grad_norm": 1.4875930547714233, + "learning_rate": 5.072675736630992e-05, + "loss": 0.4612, + "step": 38857 + }, + { + "epoch": 2.6964124627021024, + "grad_norm": 1.3661376237869263, + "learning_rate": 5.072186823235929e-05, + "loss": 0.392, + "step": 38858 + }, + { + "epoch": 2.6964818541391993, + "grad_norm": 1.0485891103744507, + "learning_rate": 5.071697925397402e-05, + "loss": 0.4077, + "step": 38859 + }, + { + "epoch": 2.6965512455762957, + "grad_norm": 1.4063047170639038, + "learning_rate": 5.0712090431169556e-05, + "loss": 0.4615, + "step": 38860 + }, + { + "epoch": 2.6966206370133925, + "grad_norm": 1.3115479946136475, + "learning_rate": 5.0707201763961266e-05, + "loss": 0.5153, + "step": 38861 + }, + { + "epoch": 2.6966900284504893, + "grad_norm": 0.9911465644836426, + "learning_rate": 5.0702313252364616e-05, + "loss": 0.2903, + "step": 38862 + }, + { + "epoch": 2.6967594198875857, + "grad_norm": 1.6586196422576904, + "learning_rate": 5.069742489639505e-05, + "loss": 0.3673, + "step": 38863 + }, + { + "epoch": 2.6968288113246826, + "grad_norm": 1.0120095014572144, + "learning_rate": 5.0692536696067994e-05, + "loss": 0.4043, + "step": 38864 + }, + { + "epoch": 2.696898202761779, + "grad_norm": 1.0243196487426758, + "learning_rate": 5.068764865139888e-05, + "loss": 0.439, + "step": 38865 + }, + { + "epoch": 2.6969675941988758, + "grad_norm": 1.2799159288406372, + "learning_rate": 5.0682760762403134e-05, + "loss": 0.4242, + "step": 38866 + }, + { + "epoch": 2.6970369856359726, + "grad_norm": 1.4192529916763306, + "learning_rate": 5.0677873029096237e-05, + "loss": 0.5203, + "step": 38867 + }, + { + "epoch": 2.6971063770730694, + "grad_norm": 1.2074006795883179, + "learning_rate": 5.0672985451493524e-05, + "loss": 0.3095, + "step": 38868 + }, + { + "epoch": 2.697175768510166, + "grad_norm": 1.4675979614257812, + "learning_rate": 5.066809802961048e-05, + "loss": 0.4236, + "step": 38869 + }, + { + "epoch": 2.6972451599472627, + "grad_norm": 1.5227922201156616, + "learning_rate": 5.066321076346252e-05, + "loss": 0.5093, + "step": 38870 + }, + { + "epoch": 2.697314551384359, + "grad_norm": 1.9465346336364746, + "learning_rate": 5.0658323653065084e-05, + "loss": 0.3691, + "step": 38871 + }, + { + "epoch": 2.697383942821456, + "grad_norm": 1.2640328407287598, + "learning_rate": 5.065343669843359e-05, + "loss": 0.3751, + "step": 38872 + }, + { + "epoch": 2.6974533342585527, + "grad_norm": 2.387390613555908, + "learning_rate": 5.0648549899583476e-05, + "loss": 0.5689, + "step": 38873 + }, + { + "epoch": 2.697522725695649, + "grad_norm": 1.1569322347640991, + "learning_rate": 5.0643663256530206e-05, + "loss": 0.4712, + "step": 38874 + }, + { + "epoch": 2.697592117132746, + "grad_norm": 1.3953979015350342, + "learning_rate": 5.063877676928912e-05, + "loss": 0.4356, + "step": 38875 + }, + { + "epoch": 2.6976615085698423, + "grad_norm": 1.1226383447647095, + "learning_rate": 5.0633890437875684e-05, + "loss": 0.4472, + "step": 38876 + }, + { + "epoch": 2.697730900006939, + "grad_norm": 1.63132643699646, + "learning_rate": 5.0629004262305345e-05, + "loss": 0.3966, + "step": 38877 + }, + { + "epoch": 2.697800291444036, + "grad_norm": 0.9541690349578857, + "learning_rate": 5.06241182425935e-05, + "loss": 0.3601, + "step": 38878 + }, + { + "epoch": 2.6978696828811324, + "grad_norm": 1.7096632719039917, + "learning_rate": 5.0619232378755586e-05, + "loss": 0.5881, + "step": 38879 + }, + { + "epoch": 2.697939074318229, + "grad_norm": 1.0091732740402222, + "learning_rate": 5.0614346670807065e-05, + "loss": 0.353, + "step": 38880 + }, + { + "epoch": 2.6980084657553256, + "grad_norm": 1.5574802160263062, + "learning_rate": 5.060946111876328e-05, + "loss": 0.4509, + "step": 38881 + }, + { + "epoch": 2.6980778571924224, + "grad_norm": 1.1214008331298828, + "learning_rate": 5.06045757226397e-05, + "loss": 0.3607, + "step": 38882 + }, + { + "epoch": 2.6981472486295193, + "grad_norm": 1.7695221900939941, + "learning_rate": 5.059969048245175e-05, + "loss": 0.5901, + "step": 38883 + }, + { + "epoch": 2.6982166400666157, + "grad_norm": 1.408158302307129, + "learning_rate": 5.0594805398214874e-05, + "loss": 0.4122, + "step": 38884 + }, + { + "epoch": 2.6982860315037125, + "grad_norm": 1.2792145013809204, + "learning_rate": 5.058992046994443e-05, + "loss": 0.443, + "step": 38885 + }, + { + "epoch": 2.698355422940809, + "grad_norm": 1.3622629642486572, + "learning_rate": 5.058503569765587e-05, + "loss": 0.4314, + "step": 38886 + }, + { + "epoch": 2.6984248143779057, + "grad_norm": 0.9186203479766846, + "learning_rate": 5.058015108136462e-05, + "loss": 0.3351, + "step": 38887 + }, + { + "epoch": 2.6984942058150025, + "grad_norm": 1.8291435241699219, + "learning_rate": 5.057526662108615e-05, + "loss": 0.452, + "step": 38888 + }, + { + "epoch": 2.698563597252099, + "grad_norm": 1.4266548156738281, + "learning_rate": 5.0570382316835774e-05, + "loss": 0.472, + "step": 38889 + }, + { + "epoch": 2.6986329886891958, + "grad_norm": 1.5538486242294312, + "learning_rate": 5.0565498168628976e-05, + "loss": 0.4549, + "step": 38890 + }, + { + "epoch": 2.698702380126292, + "grad_norm": 0.9873765110969543, + "learning_rate": 5.056061417648117e-05, + "loss": 0.3782, + "step": 38891 + }, + { + "epoch": 2.698771771563389, + "grad_norm": 1.3859938383102417, + "learning_rate": 5.055573034040776e-05, + "loss": 0.4527, + "step": 38892 + }, + { + "epoch": 2.698841163000486, + "grad_norm": 1.4041262865066528, + "learning_rate": 5.055084666042418e-05, + "loss": 0.4314, + "step": 38893 + }, + { + "epoch": 2.6989105544375827, + "grad_norm": 1.2384897470474243, + "learning_rate": 5.054596313654582e-05, + "loss": 0.3348, + "step": 38894 + }, + { + "epoch": 2.698979945874679, + "grad_norm": 1.4707465171813965, + "learning_rate": 5.054107976878818e-05, + "loss": 0.6387, + "step": 38895 + }, + { + "epoch": 2.699049337311776, + "grad_norm": 1.2991043329238892, + "learning_rate": 5.053619655716656e-05, + "loss": 0.3984, + "step": 38896 + }, + { + "epoch": 2.6991187287488723, + "grad_norm": 1.3259923458099365, + "learning_rate": 5.053131350169643e-05, + "loss": 0.527, + "step": 38897 + }, + { + "epoch": 2.699188120185969, + "grad_norm": 1.334313988685608, + "learning_rate": 5.052643060239322e-05, + "loss": 0.4856, + "step": 38898 + }, + { + "epoch": 2.699257511623066, + "grad_norm": 1.242546558380127, + "learning_rate": 5.05215478592723e-05, + "loss": 0.3588, + "step": 38899 + }, + { + "epoch": 2.6993269030601623, + "grad_norm": 0.9832791686058044, + "learning_rate": 5.051666527234913e-05, + "loss": 0.2918, + "step": 38900 + }, + { + "epoch": 2.699396294497259, + "grad_norm": 0.9530963897705078, + "learning_rate": 5.051178284163911e-05, + "loss": 0.3169, + "step": 38901 + }, + { + "epoch": 2.6994656859343555, + "grad_norm": 1.3315532207489014, + "learning_rate": 5.050690056715768e-05, + "loss": 0.3341, + "step": 38902 + }, + { + "epoch": 2.6995350773714524, + "grad_norm": 1.5303767919540405, + "learning_rate": 5.0502018448920186e-05, + "loss": 0.2764, + "step": 38903 + }, + { + "epoch": 2.699604468808549, + "grad_norm": 1.4543240070343018, + "learning_rate": 5.049713648694208e-05, + "loss": 0.5119, + "step": 38904 + }, + { + "epoch": 2.6996738602456456, + "grad_norm": 1.0330877304077148, + "learning_rate": 5.049225468123877e-05, + "loss": 0.369, + "step": 38905 + }, + { + "epoch": 2.6997432516827424, + "grad_norm": 1.4921945333480835, + "learning_rate": 5.048737303182567e-05, + "loss": 0.465, + "step": 38906 + }, + { + "epoch": 2.699812643119839, + "grad_norm": 1.328399658203125, + "learning_rate": 5.048249153871818e-05, + "loss": 0.3952, + "step": 38907 + }, + { + "epoch": 2.6998820345569357, + "grad_norm": 1.2328262329101562, + "learning_rate": 5.0477610201931726e-05, + "loss": 0.4332, + "step": 38908 + }, + { + "epoch": 2.6999514259940325, + "grad_norm": 2.9801621437072754, + "learning_rate": 5.0472729021481746e-05, + "loss": 0.2647, + "step": 38909 + }, + { + "epoch": 2.700020817431129, + "grad_norm": 1.0164299011230469, + "learning_rate": 5.046784799738358e-05, + "loss": 0.2871, + "step": 38910 + }, + { + "epoch": 2.7000902088682257, + "grad_norm": 1.3126260042190552, + "learning_rate": 5.046296712965266e-05, + "loss": 0.3381, + "step": 38911 + }, + { + "epoch": 2.700159600305322, + "grad_norm": 1.407486081123352, + "learning_rate": 5.045808641830443e-05, + "loss": 0.4192, + "step": 38912 + }, + { + "epoch": 2.700228991742419, + "grad_norm": 1.032322645187378, + "learning_rate": 5.045320586335425e-05, + "loss": 0.3793, + "step": 38913 + }, + { + "epoch": 2.7002983831795158, + "grad_norm": 1.206350564956665, + "learning_rate": 5.044832546481756e-05, + "loss": 0.4651, + "step": 38914 + }, + { + "epoch": 2.700367774616612, + "grad_norm": 0.8873113393783569, + "learning_rate": 5.0443445222709785e-05, + "loss": 0.3475, + "step": 38915 + }, + { + "epoch": 2.700437166053709, + "grad_norm": 1.0026605129241943, + "learning_rate": 5.043856513704628e-05, + "loss": 0.3706, + "step": 38916 + }, + { + "epoch": 2.7005065574908054, + "grad_norm": 1.2188489437103271, + "learning_rate": 5.043368520784246e-05, + "loss": 0.4127, + "step": 38917 + }, + { + "epoch": 2.700575948927902, + "grad_norm": 1.3814460039138794, + "learning_rate": 5.0428805435113765e-05, + "loss": 0.4095, + "step": 38918 + }, + { + "epoch": 2.700645340364999, + "grad_norm": 1.0597079992294312, + "learning_rate": 5.0423925818875605e-05, + "loss": 0.3134, + "step": 38919 + }, + { + "epoch": 2.700714731802096, + "grad_norm": 1.1627287864685059, + "learning_rate": 5.041904635914332e-05, + "loss": 0.5261, + "step": 38920 + }, + { + "epoch": 2.7007841232391923, + "grad_norm": 2.372485876083374, + "learning_rate": 5.041416705593236e-05, + "loss": 0.4117, + "step": 38921 + }, + { + "epoch": 2.700853514676289, + "grad_norm": 1.1452053785324097, + "learning_rate": 5.0409287909258166e-05, + "loss": 0.3691, + "step": 38922 + }, + { + "epoch": 2.7009229061133855, + "grad_norm": 1.1756287813186646, + "learning_rate": 5.040440891913605e-05, + "loss": 0.3717, + "step": 38923 + }, + { + "epoch": 2.7009922975504823, + "grad_norm": 1.6004852056503296, + "learning_rate": 5.039953008558145e-05, + "loss": 0.4411, + "step": 38924 + }, + { + "epoch": 2.701061688987579, + "grad_norm": 1.1146515607833862, + "learning_rate": 5.039465140860979e-05, + "loss": 0.3446, + "step": 38925 + }, + { + "epoch": 2.7011310804246755, + "grad_norm": 1.1429603099822998, + "learning_rate": 5.038977288823645e-05, + "loss": 0.4523, + "step": 38926 + }, + { + "epoch": 2.7012004718617724, + "grad_norm": 1.9065139293670654, + "learning_rate": 5.0384894524476854e-05, + "loss": 0.3383, + "step": 38927 + }, + { + "epoch": 2.7012698632988688, + "grad_norm": 1.2698184251785278, + "learning_rate": 5.038001631734639e-05, + "loss": 0.5563, + "step": 38928 + }, + { + "epoch": 2.7013392547359656, + "grad_norm": 1.3714419603347778, + "learning_rate": 5.037513826686049e-05, + "loss": 0.2917, + "step": 38929 + }, + { + "epoch": 2.7014086461730624, + "grad_norm": 1.2660821676254272, + "learning_rate": 5.037026037303448e-05, + "loss": 0.2975, + "step": 38930 + }, + { + "epoch": 2.701478037610159, + "grad_norm": 0.9625850319862366, + "learning_rate": 5.036538263588381e-05, + "loss": 0.3761, + "step": 38931 + }, + { + "epoch": 2.7015474290472556, + "grad_norm": 1.4702345132827759, + "learning_rate": 5.036050505542386e-05, + "loss": 0.358, + "step": 38932 + }, + { + "epoch": 2.701616820484352, + "grad_norm": 0.9286733865737915, + "learning_rate": 5.035562763167003e-05, + "loss": 0.3921, + "step": 38933 + }, + { + "epoch": 2.701686211921449, + "grad_norm": 0.9797583818435669, + "learning_rate": 5.0350750364637745e-05, + "loss": 0.2843, + "step": 38934 + }, + { + "epoch": 2.7017556033585457, + "grad_norm": 1.3009791374206543, + "learning_rate": 5.0345873254342366e-05, + "loss": 0.3937, + "step": 38935 + }, + { + "epoch": 2.701824994795642, + "grad_norm": 1.5056568384170532, + "learning_rate": 5.034099630079934e-05, + "loss": 0.458, + "step": 38936 + }, + { + "epoch": 2.701894386232739, + "grad_norm": 1.3456346988677979, + "learning_rate": 5.0336119504023993e-05, + "loss": 0.315, + "step": 38937 + }, + { + "epoch": 2.7019637776698353, + "grad_norm": 1.2360363006591797, + "learning_rate": 5.0331242864031745e-05, + "loss": 0.3956, + "step": 38938 + }, + { + "epoch": 2.702033169106932, + "grad_norm": 1.3463177680969238, + "learning_rate": 5.032636638083802e-05, + "loss": 0.29, + "step": 38939 + }, + { + "epoch": 2.702102560544029, + "grad_norm": 1.2084882259368896, + "learning_rate": 5.032149005445819e-05, + "loss": 0.5369, + "step": 38940 + }, + { + "epoch": 2.7021719519811254, + "grad_norm": 0.9616431593894958, + "learning_rate": 5.031661388490764e-05, + "loss": 0.2829, + "step": 38941 + }, + { + "epoch": 2.702241343418222, + "grad_norm": 0.9788913726806641, + "learning_rate": 5.031173787220178e-05, + "loss": 0.2808, + "step": 38942 + }, + { + "epoch": 2.7023107348553186, + "grad_norm": 1.0814324617385864, + "learning_rate": 5.030686201635604e-05, + "loss": 0.3373, + "step": 38943 + }, + { + "epoch": 2.7023801262924154, + "grad_norm": 1.1409002542495728, + "learning_rate": 5.030198631738573e-05, + "loss": 0.3063, + "step": 38944 + }, + { + "epoch": 2.7024495177295123, + "grad_norm": 1.1297158002853394, + "learning_rate": 5.0297110775306276e-05, + "loss": 0.3712, + "step": 38945 + }, + { + "epoch": 2.702518909166609, + "grad_norm": 1.165006399154663, + "learning_rate": 5.029223539013309e-05, + "loss": 0.3191, + "step": 38946 + }, + { + "epoch": 2.7025883006037055, + "grad_norm": 0.9433461427688599, + "learning_rate": 5.028736016188155e-05, + "loss": 0.3168, + "step": 38947 + }, + { + "epoch": 2.7026576920408023, + "grad_norm": 0.9189340472221375, + "learning_rate": 5.028248509056703e-05, + "loss": 0.3076, + "step": 38948 + }, + { + "epoch": 2.7027270834778987, + "grad_norm": 1.5046272277832031, + "learning_rate": 5.027761017620496e-05, + "loss": 0.4609, + "step": 38949 + }, + { + "epoch": 2.7027964749149955, + "grad_norm": 0.8011420369148254, + "learning_rate": 5.0272735418810724e-05, + "loss": 0.3058, + "step": 38950 + }, + { + "epoch": 2.7028658663520924, + "grad_norm": 1.0806174278259277, + "learning_rate": 5.0267860818399674e-05, + "loss": 0.4188, + "step": 38951 + }, + { + "epoch": 2.7029352577891887, + "grad_norm": 1.0184420347213745, + "learning_rate": 5.026298637498719e-05, + "loss": 0.3487, + "step": 38952 + }, + { + "epoch": 2.7030046492262856, + "grad_norm": 1.4809651374816895, + "learning_rate": 5.025811208858871e-05, + "loss": 0.3606, + "step": 38953 + }, + { + "epoch": 2.703074040663382, + "grad_norm": 1.3173588514328003, + "learning_rate": 5.0253237959219625e-05, + "loss": 0.256, + "step": 38954 + }, + { + "epoch": 2.703143432100479, + "grad_norm": 2.8528172969818115, + "learning_rate": 5.024836398689526e-05, + "loss": 0.4443, + "step": 38955 + }, + { + "epoch": 2.7032128235375756, + "grad_norm": 1.6172598600387573, + "learning_rate": 5.0243490171631034e-05, + "loss": 0.4808, + "step": 38956 + }, + { + "epoch": 2.703282214974672, + "grad_norm": 1.1088662147521973, + "learning_rate": 5.023861651344239e-05, + "loss": 0.3522, + "step": 38957 + }, + { + "epoch": 2.703351606411769, + "grad_norm": 1.4267606735229492, + "learning_rate": 5.0233743012344615e-05, + "loss": 0.4006, + "step": 38958 + }, + { + "epoch": 2.7034209978488652, + "grad_norm": 1.183128833770752, + "learning_rate": 5.022886966835314e-05, + "loss": 0.4955, + "step": 38959 + }, + { + "epoch": 2.703490389285962, + "grad_norm": 0.8774023056030273, + "learning_rate": 5.022399648148335e-05, + "loss": 0.2243, + "step": 38960 + }, + { + "epoch": 2.703559780723059, + "grad_norm": 1.510482668876648, + "learning_rate": 5.0219123451750614e-05, + "loss": 0.4972, + "step": 38961 + }, + { + "epoch": 2.7036291721601553, + "grad_norm": 1.1508820056915283, + "learning_rate": 5.021425057917035e-05, + "loss": 0.3037, + "step": 38962 + }, + { + "epoch": 2.703698563597252, + "grad_norm": 1.4627326726913452, + "learning_rate": 5.02093778637579e-05, + "loss": 0.3371, + "step": 38963 + }, + { + "epoch": 2.7037679550343485, + "grad_norm": 1.6542752981185913, + "learning_rate": 5.020450530552872e-05, + "loss": 0.6527, + "step": 38964 + }, + { + "epoch": 2.7038373464714454, + "grad_norm": 0.9809470772743225, + "learning_rate": 5.0199632904498096e-05, + "loss": 0.3003, + "step": 38965 + }, + { + "epoch": 2.703906737908542, + "grad_norm": 1.3207805156707764, + "learning_rate": 5.019476066068145e-05, + "loss": 0.3708, + "step": 38966 + }, + { + "epoch": 2.703976129345639, + "grad_norm": 1.8869651556015015, + "learning_rate": 5.018988857409418e-05, + "loss": 0.4387, + "step": 38967 + }, + { + "epoch": 2.7040455207827354, + "grad_norm": 0.9461386799812317, + "learning_rate": 5.018501664475164e-05, + "loss": 0.3426, + "step": 38968 + }, + { + "epoch": 2.704114912219832, + "grad_norm": 1.4187037944793701, + "learning_rate": 5.0180144872669224e-05, + "loss": 0.3821, + "step": 38969 + }, + { + "epoch": 2.7041843036569286, + "grad_norm": 1.5047547817230225, + "learning_rate": 5.017527325786231e-05, + "loss": 0.3865, + "step": 38970 + }, + { + "epoch": 2.7042536950940255, + "grad_norm": 1.3715457916259766, + "learning_rate": 5.017040180034632e-05, + "loss": 0.3832, + "step": 38971 + }, + { + "epoch": 2.7043230865311223, + "grad_norm": 1.1366503238677979, + "learning_rate": 5.016553050013655e-05, + "loss": 0.4378, + "step": 38972 + }, + { + "epoch": 2.7043924779682187, + "grad_norm": 1.2949789762496948, + "learning_rate": 5.016065935724843e-05, + "loss": 0.415, + "step": 38973 + }, + { + "epoch": 2.7044618694053155, + "grad_norm": 1.2081705331802368, + "learning_rate": 5.015578837169731e-05, + "loss": 0.3365, + "step": 38974 + }, + { + "epoch": 2.704531260842412, + "grad_norm": 1.5631769895553589, + "learning_rate": 5.015091754349859e-05, + "loss": 0.3155, + "step": 38975 + }, + { + "epoch": 2.7046006522795087, + "grad_norm": 1.6641165018081665, + "learning_rate": 5.014604687266764e-05, + "loss": 0.415, + "step": 38976 + }, + { + "epoch": 2.7046700437166056, + "grad_norm": 1.2103025913238525, + "learning_rate": 5.0141176359219823e-05, + "loss": 0.2568, + "step": 38977 + }, + { + "epoch": 2.704739435153702, + "grad_norm": 1.862207055091858, + "learning_rate": 5.0136306003170586e-05, + "loss": 0.4096, + "step": 38978 + }, + { + "epoch": 2.704808826590799, + "grad_norm": 1.0183701515197754, + "learning_rate": 5.0131435804535185e-05, + "loss": 0.2959, + "step": 38979 + }, + { + "epoch": 2.704878218027895, + "grad_norm": 1.1212824583053589, + "learning_rate": 5.012656576332908e-05, + "loss": 0.3578, + "step": 38980 + }, + { + "epoch": 2.704947609464992, + "grad_norm": 1.1765012741088867, + "learning_rate": 5.012169587956761e-05, + "loss": 0.4475, + "step": 38981 + }, + { + "epoch": 2.705017000902089, + "grad_norm": 1.0314382314682007, + "learning_rate": 5.011682615326616e-05, + "loss": 0.3182, + "step": 38982 + }, + { + "epoch": 2.7050863923391852, + "grad_norm": 1.5016757249832153, + "learning_rate": 5.01119565844401e-05, + "loss": 0.4597, + "step": 38983 + }, + { + "epoch": 2.705155783776282, + "grad_norm": 1.0694090127944946, + "learning_rate": 5.0107087173104794e-05, + "loss": 0.4418, + "step": 38984 + }, + { + "epoch": 2.7052251752133785, + "grad_norm": 1.421423316001892, + "learning_rate": 5.010221791927567e-05, + "loss": 0.4017, + "step": 38985 + }, + { + "epoch": 2.7052945666504753, + "grad_norm": 1.5377357006072998, + "learning_rate": 5.009734882296804e-05, + "loss": 0.3689, + "step": 38986 + }, + { + "epoch": 2.705363958087572, + "grad_norm": 1.5797652006149292, + "learning_rate": 5.009247988419726e-05, + "loss": 0.5513, + "step": 38987 + }, + { + "epoch": 2.7054333495246685, + "grad_norm": 1.3979008197784424, + "learning_rate": 5.0087611102978784e-05, + "loss": 0.4109, + "step": 38988 + }, + { + "epoch": 2.7055027409617654, + "grad_norm": 1.285222053527832, + "learning_rate": 5.0082742479327883e-05, + "loss": 0.6306, + "step": 38989 + }, + { + "epoch": 2.7055721323988617, + "grad_norm": 1.4171098470687866, + "learning_rate": 5.007787401325998e-05, + "loss": 0.3716, + "step": 38990 + }, + { + "epoch": 2.7056415238359586, + "grad_norm": 1.5863772630691528, + "learning_rate": 5.0073005704790434e-05, + "loss": 0.3933, + "step": 38991 + }, + { + "epoch": 2.7057109152730554, + "grad_norm": 1.1928263902664185, + "learning_rate": 5.006813755393465e-05, + "loss": 0.5299, + "step": 38992 + }, + { + "epoch": 2.7057803067101522, + "grad_norm": 2.236523151397705, + "learning_rate": 5.006326956070793e-05, + "loss": 0.5805, + "step": 38993 + }, + { + "epoch": 2.7058496981472486, + "grad_norm": 0.8534892797470093, + "learning_rate": 5.005840172512567e-05, + "loss": 0.3095, + "step": 38994 + }, + { + "epoch": 2.705919089584345, + "grad_norm": 1.3980424404144287, + "learning_rate": 5.0053534047203246e-05, + "loss": 0.4263, + "step": 38995 + }, + { + "epoch": 2.705988481021442, + "grad_norm": 1.0164812803268433, + "learning_rate": 5.004866652695601e-05, + "loss": 0.3459, + "step": 38996 + }, + { + "epoch": 2.7060578724585387, + "grad_norm": 1.3424763679504395, + "learning_rate": 5.004379916439935e-05, + "loss": 0.41, + "step": 38997 + }, + { + "epoch": 2.7061272638956355, + "grad_norm": 1.258744716644287, + "learning_rate": 5.0038931959548655e-05, + "loss": 0.324, + "step": 38998 + }, + { + "epoch": 2.706196655332732, + "grad_norm": 1.2811715602874756, + "learning_rate": 5.003406491241922e-05, + "loss": 0.4833, + "step": 38999 + }, + { + "epoch": 2.7062660467698287, + "grad_norm": 1.1358729600906372, + "learning_rate": 5.0029198023026435e-05, + "loss": 0.4124, + "step": 39000 + }, + { + "epoch": 2.706335438206925, + "grad_norm": 1.6412307024002075, + "learning_rate": 5.002433129138568e-05, + "loss": 0.5337, + "step": 39001 + }, + { + "epoch": 2.706404829644022, + "grad_norm": 1.1744718551635742, + "learning_rate": 5.001946471751231e-05, + "loss": 0.4872, + "step": 39002 + }, + { + "epoch": 2.706474221081119, + "grad_norm": 1.118297815322876, + "learning_rate": 5.00145983014217e-05, + "loss": 0.4165, + "step": 39003 + }, + { + "epoch": 2.706543612518215, + "grad_norm": 1.7261748313903809, + "learning_rate": 5.00097320431292e-05, + "loss": 0.4135, + "step": 39004 + }, + { + "epoch": 2.706613003955312, + "grad_norm": 1.4153143167495728, + "learning_rate": 5.000486594265021e-05, + "loss": 0.4953, + "step": 39005 + }, + { + "epoch": 2.7066823953924084, + "grad_norm": 2.3428096771240234, + "learning_rate": 5.000000000000002e-05, + "loss": 0.5765, + "step": 39006 + }, + { + "epoch": 2.7067517868295052, + "grad_norm": 1.2722047567367554, + "learning_rate": 4.999513421519403e-05, + "loss": 0.4007, + "step": 39007 + }, + { + "epoch": 2.706821178266602, + "grad_norm": 1.193562388420105, + "learning_rate": 4.9990268588247604e-05, + "loss": 0.466, + "step": 39008 + }, + { + "epoch": 2.7068905697036985, + "grad_norm": 1.056601881980896, + "learning_rate": 4.998540311917609e-05, + "loss": 0.3721, + "step": 39009 + }, + { + "epoch": 2.7069599611407953, + "grad_norm": 1.09878408908844, + "learning_rate": 4.998053780799487e-05, + "loss": 0.3483, + "step": 39010 + }, + { + "epoch": 2.7070293525778917, + "grad_norm": 1.8970000743865967, + "learning_rate": 4.997567265471927e-05, + "loss": 0.4957, + "step": 39011 + }, + { + "epoch": 2.7070987440149885, + "grad_norm": 1.3633819818496704, + "learning_rate": 4.997080765936472e-05, + "loss": 0.344, + "step": 39012 + }, + { + "epoch": 2.7071681354520853, + "grad_norm": 1.4267598390579224, + "learning_rate": 4.996594282194648e-05, + "loss": 0.497, + "step": 39013 + }, + { + "epoch": 2.7072375268891817, + "grad_norm": 1.172202467918396, + "learning_rate": 4.996107814247995e-05, + "loss": 0.5101, + "step": 39014 + }, + { + "epoch": 2.7073069183262786, + "grad_norm": 1.1207780838012695, + "learning_rate": 4.99562136209805e-05, + "loss": 0.3787, + "step": 39015 + }, + { + "epoch": 2.707376309763375, + "grad_norm": 0.7500752806663513, + "learning_rate": 4.9951349257463475e-05, + "loss": 0.1961, + "step": 39016 + }, + { + "epoch": 2.707445701200472, + "grad_norm": 1.2057740688323975, + "learning_rate": 4.994648505194423e-05, + "loss": 0.4151, + "step": 39017 + }, + { + "epoch": 2.7075150926375686, + "grad_norm": 0.865790069103241, + "learning_rate": 4.994162100443812e-05, + "loss": 0.2537, + "step": 39018 + }, + { + "epoch": 2.7075844840746655, + "grad_norm": 1.2350537776947021, + "learning_rate": 4.9936757114960553e-05, + "loss": 0.3388, + "step": 39019 + }, + { + "epoch": 2.707653875511762, + "grad_norm": 0.8129552006721497, + "learning_rate": 4.99318933835268e-05, + "loss": 0.2241, + "step": 39020 + }, + { + "epoch": 2.7077232669488587, + "grad_norm": 1.0041979551315308, + "learning_rate": 4.9927029810152236e-05, + "loss": 0.2669, + "step": 39021 + }, + { + "epoch": 2.707792658385955, + "grad_norm": 1.4617198705673218, + "learning_rate": 4.992216639485224e-05, + "loss": 0.4778, + "step": 39022 + }, + { + "epoch": 2.707862049823052, + "grad_norm": 0.8779043555259705, + "learning_rate": 4.991730313764219e-05, + "loss": 0.2904, + "step": 39023 + }, + { + "epoch": 2.7079314412601487, + "grad_norm": 1.4264335632324219, + "learning_rate": 4.9912440038537366e-05, + "loss": 0.3707, + "step": 39024 + }, + { + "epoch": 2.708000832697245, + "grad_norm": 0.783057689666748, + "learning_rate": 4.990757709755316e-05, + "loss": 0.2824, + "step": 39025 + }, + { + "epoch": 2.708070224134342, + "grad_norm": 0.9355727434158325, + "learning_rate": 4.9902714314704925e-05, + "loss": 0.2837, + "step": 39026 + }, + { + "epoch": 2.7081396155714383, + "grad_norm": 1.3238576650619507, + "learning_rate": 4.9897851690008036e-05, + "loss": 0.5451, + "step": 39027 + }, + { + "epoch": 2.708209007008535, + "grad_norm": 1.0144224166870117, + "learning_rate": 4.989298922347778e-05, + "loss": 0.4098, + "step": 39028 + }, + { + "epoch": 2.708278398445632, + "grad_norm": 0.9670843482017517, + "learning_rate": 4.988812691512955e-05, + "loss": 0.2799, + "step": 39029 + }, + { + "epoch": 2.7083477898827284, + "grad_norm": 1.0055731534957886, + "learning_rate": 4.9883264764978676e-05, + "loss": 0.368, + "step": 39030 + }, + { + "epoch": 2.7084171813198252, + "grad_norm": 1.2687747478485107, + "learning_rate": 4.987840277304053e-05, + "loss": 0.3347, + "step": 39031 + }, + { + "epoch": 2.7084865727569216, + "grad_norm": 1.4713878631591797, + "learning_rate": 4.9873540939330454e-05, + "loss": 0.425, + "step": 39032 + }, + { + "epoch": 2.7085559641940185, + "grad_norm": 1.185899257659912, + "learning_rate": 4.986867926386383e-05, + "loss": 0.4165, + "step": 39033 + }, + { + "epoch": 2.7086253556311153, + "grad_norm": 1.3338567018508911, + "learning_rate": 4.9863817746655926e-05, + "loss": 0.6244, + "step": 39034 + }, + { + "epoch": 2.7086947470682117, + "grad_norm": 1.1716878414154053, + "learning_rate": 4.9858956387722147e-05, + "loss": 0.3611, + "step": 39035 + }, + { + "epoch": 2.7087641385053085, + "grad_norm": 1.1250298023223877, + "learning_rate": 4.985409518707781e-05, + "loss": 0.3571, + "step": 39036 + }, + { + "epoch": 2.708833529942405, + "grad_norm": 1.0992717742919922, + "learning_rate": 4.9849234144738276e-05, + "loss": 0.3828, + "step": 39037 + }, + { + "epoch": 2.7089029213795017, + "grad_norm": 0.8214186429977417, + "learning_rate": 4.98443732607189e-05, + "loss": 0.2224, + "step": 39038 + }, + { + "epoch": 2.7089723128165986, + "grad_norm": 0.966719388961792, + "learning_rate": 4.983951253503502e-05, + "loss": 0.2871, + "step": 39039 + }, + { + "epoch": 2.709041704253695, + "grad_norm": 0.8514065146446228, + "learning_rate": 4.9834651967702005e-05, + "loss": 0.4144, + "step": 39040 + }, + { + "epoch": 2.709111095690792, + "grad_norm": 1.1436525583267212, + "learning_rate": 4.982979155873514e-05, + "loss": 0.3551, + "step": 39041 + }, + { + "epoch": 2.709180487127888, + "grad_norm": 1.557712197303772, + "learning_rate": 4.98249313081498e-05, + "loss": 0.4401, + "step": 39042 + }, + { + "epoch": 2.709249878564985, + "grad_norm": 1.1986773014068604, + "learning_rate": 4.982007121596134e-05, + "loss": 0.4044, + "step": 39043 + }, + { + "epoch": 2.709319270002082, + "grad_norm": 1.2612829208374023, + "learning_rate": 4.981521128218509e-05, + "loss": 0.3695, + "step": 39044 + }, + { + "epoch": 2.7093886614391787, + "grad_norm": 1.4835280179977417, + "learning_rate": 4.9810351506836393e-05, + "loss": 0.3162, + "step": 39045 + }, + { + "epoch": 2.709458052876275, + "grad_norm": 1.0199658870697021, + "learning_rate": 4.980549188993059e-05, + "loss": 0.3265, + "step": 39046 + }, + { + "epoch": 2.709527444313372, + "grad_norm": 1.1907166242599487, + "learning_rate": 4.9800632431483076e-05, + "loss": 0.4416, + "step": 39047 + }, + { + "epoch": 2.7095968357504683, + "grad_norm": 1.184195637702942, + "learning_rate": 4.97957731315091e-05, + "loss": 0.4987, + "step": 39048 + }, + { + "epoch": 2.709666227187565, + "grad_norm": 1.352706789970398, + "learning_rate": 4.979091399002402e-05, + "loss": 0.3779, + "step": 39049 + }, + { + "epoch": 2.709735618624662, + "grad_norm": 1.0673089027404785, + "learning_rate": 4.978605500704323e-05, + "loss": 0.3206, + "step": 39050 + }, + { + "epoch": 2.7098050100617583, + "grad_norm": 1.3548074960708618, + "learning_rate": 4.9781196182582024e-05, + "loss": 0.327, + "step": 39051 + }, + { + "epoch": 2.709874401498855, + "grad_norm": 1.2679015398025513, + "learning_rate": 4.977633751665576e-05, + "loss": 0.437, + "step": 39052 + }, + { + "epoch": 2.7099437929359516, + "grad_norm": 1.7581747770309448, + "learning_rate": 4.9771479009279766e-05, + "loss": 0.5385, + "step": 39053 + }, + { + "epoch": 2.7100131843730484, + "grad_norm": 1.2823878526687622, + "learning_rate": 4.976662066046942e-05, + "loss": 0.5085, + "step": 39054 + }, + { + "epoch": 2.7100825758101452, + "grad_norm": 1.510107398033142, + "learning_rate": 4.9761762470239995e-05, + "loss": 0.5629, + "step": 39055 + }, + { + "epoch": 2.7101519672472416, + "grad_norm": 1.2886223793029785, + "learning_rate": 4.9756904438606856e-05, + "loss": 0.3449, + "step": 39056 + }, + { + "epoch": 2.7102213586843384, + "grad_norm": 1.066226601600647, + "learning_rate": 4.975204656558533e-05, + "loss": 0.3267, + "step": 39057 + }, + { + "epoch": 2.710290750121435, + "grad_norm": 1.3220587968826294, + "learning_rate": 4.974718885119082e-05, + "loss": 0.3858, + "step": 39058 + }, + { + "epoch": 2.7103601415585317, + "grad_norm": 1.5649317502975464, + "learning_rate": 4.974233129543855e-05, + "loss": 0.4405, + "step": 39059 + }, + { + "epoch": 2.7104295329956285, + "grad_norm": 1.3833266496658325, + "learning_rate": 4.973747389834391e-05, + "loss": 0.3878, + "step": 39060 + }, + { + "epoch": 2.710498924432725, + "grad_norm": 0.9940348267555237, + "learning_rate": 4.973261665992225e-05, + "loss": 0.2721, + "step": 39061 + }, + { + "epoch": 2.7105683158698217, + "grad_norm": 1.461290717124939, + "learning_rate": 4.97277595801889e-05, + "loss": 0.3398, + "step": 39062 + }, + { + "epoch": 2.710637707306918, + "grad_norm": 1.2567179203033447, + "learning_rate": 4.9722902659159166e-05, + "loss": 0.3361, + "step": 39063 + }, + { + "epoch": 2.710707098744015, + "grad_norm": 1.2375632524490356, + "learning_rate": 4.971804589684839e-05, + "loss": 0.3445, + "step": 39064 + }, + { + "epoch": 2.7107764901811118, + "grad_norm": 0.9094449281692505, + "learning_rate": 4.97131892932719e-05, + "loss": 0.4452, + "step": 39065 + }, + { + "epoch": 2.710845881618208, + "grad_norm": 1.0903007984161377, + "learning_rate": 4.970833284844504e-05, + "loss": 0.3316, + "step": 39066 + }, + { + "epoch": 2.710915273055305, + "grad_norm": 0.949618399143219, + "learning_rate": 4.9703476562383145e-05, + "loss": 0.3485, + "step": 39067 + }, + { + "epoch": 2.7109846644924014, + "grad_norm": 1.028067946434021, + "learning_rate": 4.969862043510156e-05, + "loss": 0.3685, + "step": 39068 + }, + { + "epoch": 2.711054055929498, + "grad_norm": 1.1619443893432617, + "learning_rate": 4.969376446661558e-05, + "loss": 0.4443, + "step": 39069 + }, + { + "epoch": 2.711123447366595, + "grad_norm": 0.8666784763336182, + "learning_rate": 4.968890865694054e-05, + "loss": 0.3041, + "step": 39070 + }, + { + "epoch": 2.711192838803692, + "grad_norm": 1.1791132688522339, + "learning_rate": 4.968405300609177e-05, + "loss": 0.4401, + "step": 39071 + }, + { + "epoch": 2.7112622302407883, + "grad_norm": 1.4797269105911255, + "learning_rate": 4.9679197514084606e-05, + "loss": 0.4371, + "step": 39072 + }, + { + "epoch": 2.711331621677885, + "grad_norm": 1.9363855123519897, + "learning_rate": 4.9674342180934384e-05, + "loss": 0.4345, + "step": 39073 + }, + { + "epoch": 2.7114010131149815, + "grad_norm": 1.0754351615905762, + "learning_rate": 4.9669487006656467e-05, + "loss": 0.3269, + "step": 39074 + }, + { + "epoch": 2.7114704045520783, + "grad_norm": 1.67363440990448, + "learning_rate": 4.96646319912661e-05, + "loss": 0.411, + "step": 39075 + }, + { + "epoch": 2.711539795989175, + "grad_norm": 1.1234443187713623, + "learning_rate": 4.965977713477864e-05, + "loss": 0.3329, + "step": 39076 + }, + { + "epoch": 2.7116091874262715, + "grad_norm": 1.4894393682479858, + "learning_rate": 4.965492243720943e-05, + "loss": 0.4577, + "step": 39077 + }, + { + "epoch": 2.7116785788633684, + "grad_norm": 1.76906418800354, + "learning_rate": 4.9650067898573795e-05, + "loss": 0.5128, + "step": 39078 + }, + { + "epoch": 2.7117479703004648, + "grad_norm": 1.1796307563781738, + "learning_rate": 4.964521351888704e-05, + "loss": 0.6013, + "step": 39079 + }, + { + "epoch": 2.7118173617375616, + "grad_norm": 1.227571725845337, + "learning_rate": 4.9640359298164516e-05, + "loss": 0.29, + "step": 39080 + }, + { + "epoch": 2.7118867531746584, + "grad_norm": 1.2663389444351196, + "learning_rate": 4.963550523642156e-05, + "loss": 0.3793, + "step": 39081 + }, + { + "epoch": 2.711956144611755, + "grad_norm": 0.7578246593475342, + "learning_rate": 4.963065133367345e-05, + "loss": 0.2915, + "step": 39082 + }, + { + "epoch": 2.7120255360488517, + "grad_norm": 1.1230279207229614, + "learning_rate": 4.962579758993552e-05, + "loss": 0.249, + "step": 39083 + }, + { + "epoch": 2.712094927485948, + "grad_norm": 1.152642011642456, + "learning_rate": 4.96209440052231e-05, + "loss": 0.3528, + "step": 39084 + }, + { + "epoch": 2.712164318923045, + "grad_norm": 1.4630237817764282, + "learning_rate": 4.961609057955152e-05, + "loss": 0.427, + "step": 39085 + }, + { + "epoch": 2.7122337103601417, + "grad_norm": 0.9155113697052002, + "learning_rate": 4.96112373129361e-05, + "loss": 0.3377, + "step": 39086 + }, + { + "epoch": 2.712303101797238, + "grad_norm": 1.1666195392608643, + "learning_rate": 4.960638420539214e-05, + "loss": 0.6706, + "step": 39087 + }, + { + "epoch": 2.712372493234335, + "grad_norm": 1.1633269786834717, + "learning_rate": 4.960153125693504e-05, + "loss": 0.3853, + "step": 39088 + }, + { + "epoch": 2.7124418846714313, + "grad_norm": 1.1489770412445068, + "learning_rate": 4.959667846758001e-05, + "loss": 0.5275, + "step": 39089 + }, + { + "epoch": 2.712511276108528, + "grad_norm": 1.2938157320022583, + "learning_rate": 4.959182583734241e-05, + "loss": 0.3661, + "step": 39090 + }, + { + "epoch": 2.712580667545625, + "grad_norm": 1.2165168523788452, + "learning_rate": 4.9586973366237576e-05, + "loss": 0.5612, + "step": 39091 + }, + { + "epoch": 2.7126500589827214, + "grad_norm": 1.432824730873108, + "learning_rate": 4.958212105428082e-05, + "loss": 0.3817, + "step": 39092 + }, + { + "epoch": 2.712719450419818, + "grad_norm": 1.9033113718032837, + "learning_rate": 4.957726890148749e-05, + "loss": 0.4142, + "step": 39093 + }, + { + "epoch": 2.7127888418569146, + "grad_norm": 1.1198793649673462, + "learning_rate": 4.9572416907872834e-05, + "loss": 0.4387, + "step": 39094 + }, + { + "epoch": 2.7128582332940114, + "grad_norm": 0.7993512153625488, + "learning_rate": 4.956756507345222e-05, + "loss": 0.2589, + "step": 39095 + }, + { + "epoch": 2.7129276247311083, + "grad_norm": 1.2156659364700317, + "learning_rate": 4.956271339824094e-05, + "loss": 0.4281, + "step": 39096 + }, + { + "epoch": 2.712997016168205, + "grad_norm": 1.229124903678894, + "learning_rate": 4.9557861882254355e-05, + "loss": 0.485, + "step": 39097 + }, + { + "epoch": 2.7130664076053015, + "grad_norm": 1.461025357246399, + "learning_rate": 4.9553010525507716e-05, + "loss": 0.3522, + "step": 39098 + }, + { + "epoch": 2.7131357990423983, + "grad_norm": 0.8636562824249268, + "learning_rate": 4.954815932801638e-05, + "loss": 0.3252, + "step": 39099 + }, + { + "epoch": 2.7132051904794947, + "grad_norm": 1.454287052154541, + "learning_rate": 4.954330828979563e-05, + "loss": 0.4566, + "step": 39100 + }, + { + "epoch": 2.7132745819165915, + "grad_norm": 1.003708004951477, + "learning_rate": 4.9538457410860815e-05, + "loss": 0.4295, + "step": 39101 + }, + { + "epoch": 2.7133439733536884, + "grad_norm": 1.1937246322631836, + "learning_rate": 4.9533606691227274e-05, + "loss": 0.514, + "step": 39102 + }, + { + "epoch": 2.7134133647907848, + "grad_norm": 1.2353339195251465, + "learning_rate": 4.952875613091025e-05, + "loss": 0.3948, + "step": 39103 + }, + { + "epoch": 2.7134827562278816, + "grad_norm": 1.0193318128585815, + "learning_rate": 4.952390572992507e-05, + "loss": 0.4343, + "step": 39104 + }, + { + "epoch": 2.713552147664978, + "grad_norm": 1.2459670305252075, + "learning_rate": 4.951905548828707e-05, + "loss": 0.3992, + "step": 39105 + }, + { + "epoch": 2.713621539102075, + "grad_norm": 1.3722822666168213, + "learning_rate": 4.9514205406011563e-05, + "loss": 0.4985, + "step": 39106 + }, + { + "epoch": 2.7136909305391717, + "grad_norm": 1.4497261047363281, + "learning_rate": 4.950935548311384e-05, + "loss": 0.5049, + "step": 39107 + }, + { + "epoch": 2.713760321976268, + "grad_norm": 1.376393437385559, + "learning_rate": 4.950450571960923e-05, + "loss": 0.4927, + "step": 39108 + }, + { + "epoch": 2.713829713413365, + "grad_norm": 1.433790922164917, + "learning_rate": 4.949965611551307e-05, + "loss": 0.4762, + "step": 39109 + }, + { + "epoch": 2.7138991048504613, + "grad_norm": 1.064419150352478, + "learning_rate": 4.949480667084061e-05, + "loss": 0.3131, + "step": 39110 + }, + { + "epoch": 2.713968496287558, + "grad_norm": 0.7583814859390259, + "learning_rate": 4.948995738560718e-05, + "loss": 0.285, + "step": 39111 + }, + { + "epoch": 2.714037887724655, + "grad_norm": 1.3788875341415405, + "learning_rate": 4.9485108259828095e-05, + "loss": 0.3862, + "step": 39112 + }, + { + "epoch": 2.7141072791617513, + "grad_norm": 1.00083589553833, + "learning_rate": 4.9480259293518664e-05, + "loss": 0.3869, + "step": 39113 + }, + { + "epoch": 2.714176670598848, + "grad_norm": 1.0033880472183228, + "learning_rate": 4.9475410486694186e-05, + "loss": 0.3456, + "step": 39114 + }, + { + "epoch": 2.7142460620359445, + "grad_norm": 1.227293610572815, + "learning_rate": 4.9470561839369976e-05, + "loss": 0.3476, + "step": 39115 + }, + { + "epoch": 2.7143154534730414, + "grad_norm": 1.4938688278198242, + "learning_rate": 4.946571335156138e-05, + "loss": 0.4604, + "step": 39116 + }, + { + "epoch": 2.714384844910138, + "grad_norm": 0.6411321759223938, + "learning_rate": 4.946086502328362e-05, + "loss": 0.2569, + "step": 39117 + }, + { + "epoch": 2.7144542363472346, + "grad_norm": 1.0768314599990845, + "learning_rate": 4.945601685455206e-05, + "loss": 0.4341, + "step": 39118 + }, + { + "epoch": 2.7145236277843314, + "grad_norm": 1.9713853597640991, + "learning_rate": 4.9451168845381976e-05, + "loss": 0.6905, + "step": 39119 + }, + { + "epoch": 2.714593019221428, + "grad_norm": 1.2824503183364868, + "learning_rate": 4.94463209957887e-05, + "loss": 0.4549, + "step": 39120 + }, + { + "epoch": 2.7146624106585246, + "grad_norm": 1.2386913299560547, + "learning_rate": 4.944147330578751e-05, + "loss": 0.3281, + "step": 39121 + }, + { + "epoch": 2.7147318020956215, + "grad_norm": 0.9450876712799072, + "learning_rate": 4.943662577539373e-05, + "loss": 0.2923, + "step": 39122 + }, + { + "epoch": 2.7148011935327183, + "grad_norm": 1.116678237915039, + "learning_rate": 4.9431778404622695e-05, + "loss": 0.474, + "step": 39123 + }, + { + "epoch": 2.7148705849698147, + "grad_norm": 1.8425525426864624, + "learning_rate": 4.942693119348963e-05, + "loss": 0.6532, + "step": 39124 + }, + { + "epoch": 2.7149399764069115, + "grad_norm": 1.4906054735183716, + "learning_rate": 4.942208414200987e-05, + "loss": 0.3315, + "step": 39125 + }, + { + "epoch": 2.715009367844008, + "grad_norm": 1.7065668106079102, + "learning_rate": 4.941723725019873e-05, + "loss": 0.3981, + "step": 39126 + }, + { + "epoch": 2.7150787592811048, + "grad_norm": 1.315259575843811, + "learning_rate": 4.941239051807151e-05, + "loss": 0.4513, + "step": 39127 + }, + { + "epoch": 2.7151481507182016, + "grad_norm": 1.2071889638900757, + "learning_rate": 4.940754394564352e-05, + "loss": 0.4438, + "step": 39128 + }, + { + "epoch": 2.715217542155298, + "grad_norm": 1.4549477100372314, + "learning_rate": 4.940269753293002e-05, + "loss": 0.4374, + "step": 39129 + }, + { + "epoch": 2.715286933592395, + "grad_norm": 1.61500084400177, + "learning_rate": 4.939785127994634e-05, + "loss": 0.4613, + "step": 39130 + }, + { + "epoch": 2.715356325029491, + "grad_norm": 1.0457890033721924, + "learning_rate": 4.93930051867078e-05, + "loss": 0.3222, + "step": 39131 + }, + { + "epoch": 2.715425716466588, + "grad_norm": 1.3365601301193237, + "learning_rate": 4.938815925322964e-05, + "loss": 0.4248, + "step": 39132 + }, + { + "epoch": 2.715495107903685, + "grad_norm": 1.445160150527954, + "learning_rate": 4.9383313479527185e-05, + "loss": 0.4984, + "step": 39133 + }, + { + "epoch": 2.7155644993407813, + "grad_norm": 1.4009034633636475, + "learning_rate": 4.937846786561573e-05, + "loss": 0.5334, + "step": 39134 + }, + { + "epoch": 2.715633890777878, + "grad_norm": 0.920983612537384, + "learning_rate": 4.937362241151059e-05, + "loss": 0.3364, + "step": 39135 + }, + { + "epoch": 2.7157032822149745, + "grad_norm": 1.425200343132019, + "learning_rate": 4.936877711722704e-05, + "loss": 0.5845, + "step": 39136 + }, + { + "epoch": 2.7157726736520713, + "grad_norm": 1.1092250347137451, + "learning_rate": 4.9363931982780434e-05, + "loss": 0.3679, + "step": 39137 + }, + { + "epoch": 2.715842065089168, + "grad_norm": 1.0032389163970947, + "learning_rate": 4.9359087008185975e-05, + "loss": 0.3265, + "step": 39138 + }, + { + "epoch": 2.7159114565262645, + "grad_norm": 1.1494224071502686, + "learning_rate": 4.9354242193459e-05, + "loss": 0.3867, + "step": 39139 + }, + { + "epoch": 2.7159808479633614, + "grad_norm": 1.0105448961257935, + "learning_rate": 4.934939753861482e-05, + "loss": 0.303, + "step": 39140 + }, + { + "epoch": 2.7160502394004578, + "grad_norm": 0.8760766386985779, + "learning_rate": 4.934455304366871e-05, + "loss": 0.3389, + "step": 39141 + }, + { + "epoch": 2.7161196308375546, + "grad_norm": 1.439893364906311, + "learning_rate": 4.9339708708635956e-05, + "loss": 0.3411, + "step": 39142 + }, + { + "epoch": 2.7161890222746514, + "grad_norm": 0.9157266020774841, + "learning_rate": 4.933486453353188e-05, + "loss": 0.3608, + "step": 39143 + }, + { + "epoch": 2.716258413711748, + "grad_norm": 1.3462810516357422, + "learning_rate": 4.933002051837179e-05, + "loss": 0.4792, + "step": 39144 + }, + { + "epoch": 2.7163278051488446, + "grad_norm": 0.9213237762451172, + "learning_rate": 4.93251766631709e-05, + "loss": 0.4767, + "step": 39145 + }, + { + "epoch": 2.716397196585941, + "grad_norm": 1.6005375385284424, + "learning_rate": 4.932033296794456e-05, + "loss": 0.2998, + "step": 39146 + }, + { + "epoch": 2.716466588023038, + "grad_norm": 1.1521915197372437, + "learning_rate": 4.931548943270805e-05, + "loss": 0.2859, + "step": 39147 + }, + { + "epoch": 2.7165359794601347, + "grad_norm": 1.3814408779144287, + "learning_rate": 4.931064605747665e-05, + "loss": 0.4165, + "step": 39148 + }, + { + "epoch": 2.7166053708972315, + "grad_norm": 1.986621618270874, + "learning_rate": 4.9305802842265667e-05, + "loss": 0.5254, + "step": 39149 + }, + { + "epoch": 2.716674762334328, + "grad_norm": 1.1999887228012085, + "learning_rate": 4.930095978709042e-05, + "loss": 0.4058, + "step": 39150 + }, + { + "epoch": 2.7167441537714248, + "grad_norm": 1.331780195236206, + "learning_rate": 4.929611689196613e-05, + "loss": 0.469, + "step": 39151 + }, + { + "epoch": 2.716813545208521, + "grad_norm": 1.2757620811462402, + "learning_rate": 4.92912741569081e-05, + "loss": 0.4242, + "step": 39152 + }, + { + "epoch": 2.716882936645618, + "grad_norm": 1.0752023458480835, + "learning_rate": 4.9286431581931646e-05, + "loss": 0.3958, + "step": 39153 + }, + { + "epoch": 2.716952328082715, + "grad_norm": 1.576589584350586, + "learning_rate": 4.928158916705204e-05, + "loss": 0.4477, + "step": 39154 + }, + { + "epoch": 2.717021719519811, + "grad_norm": 1.341351866722107, + "learning_rate": 4.9276746912284565e-05, + "loss": 0.4462, + "step": 39155 + }, + { + "epoch": 2.717091110956908, + "grad_norm": 1.1199369430541992, + "learning_rate": 4.927190481764452e-05, + "loss": 0.2666, + "step": 39156 + }, + { + "epoch": 2.7171605023940044, + "grad_norm": 1.607921838760376, + "learning_rate": 4.926706288314723e-05, + "loss": 0.2429, + "step": 39157 + }, + { + "epoch": 2.7172298938311013, + "grad_norm": 0.9562339782714844, + "learning_rate": 4.9262221108807896e-05, + "loss": 0.3106, + "step": 39158 + }, + { + "epoch": 2.717299285268198, + "grad_norm": 1.239168643951416, + "learning_rate": 4.925737949464184e-05, + "loss": 0.4366, + "step": 39159 + }, + { + "epoch": 2.7173686767052945, + "grad_norm": 0.6931412220001221, + "learning_rate": 4.925253804066435e-05, + "loss": 0.2474, + "step": 39160 + }, + { + "epoch": 2.7174380681423913, + "grad_norm": 1.0558778047561646, + "learning_rate": 4.924769674689072e-05, + "loss": 0.3295, + "step": 39161 + }, + { + "epoch": 2.7175074595794877, + "grad_norm": 1.1299479007720947, + "learning_rate": 4.9242855613336204e-05, + "loss": 0.3758, + "step": 39162 + }, + { + "epoch": 2.7175768510165845, + "grad_norm": 1.1475962400436401, + "learning_rate": 4.923801464001615e-05, + "loss": 0.362, + "step": 39163 + }, + { + "epoch": 2.7176462424536814, + "grad_norm": 1.2103183269500732, + "learning_rate": 4.923317382694576e-05, + "loss": 0.3458, + "step": 39164 + }, + { + "epoch": 2.7177156338907777, + "grad_norm": 1.0971442461013794, + "learning_rate": 4.922833317414035e-05, + "loss": 0.3855, + "step": 39165 + }, + { + "epoch": 2.7177850253278746, + "grad_norm": 1.0772494077682495, + "learning_rate": 4.922349268161524e-05, + "loss": 0.2838, + "step": 39166 + }, + { + "epoch": 2.717854416764971, + "grad_norm": 1.1187340021133423, + "learning_rate": 4.921865234938563e-05, + "loss": 0.3132, + "step": 39167 + }, + { + "epoch": 2.717923808202068, + "grad_norm": 0.9369348287582397, + "learning_rate": 4.9213812177466846e-05, + "loss": 0.3059, + "step": 39168 + }, + { + "epoch": 2.7179931996391646, + "grad_norm": 1.1376662254333496, + "learning_rate": 4.9208972165874156e-05, + "loss": 0.3543, + "step": 39169 + }, + { + "epoch": 2.718062591076261, + "grad_norm": 1.4453022480010986, + "learning_rate": 4.9204132314622866e-05, + "loss": 0.5735, + "step": 39170 + }, + { + "epoch": 2.718131982513358, + "grad_norm": 0.8176921010017395, + "learning_rate": 4.9199292623728255e-05, + "loss": 0.3345, + "step": 39171 + }, + { + "epoch": 2.7182013739504542, + "grad_norm": 1.1838241815567017, + "learning_rate": 4.9194453093205564e-05, + "loss": 0.3906, + "step": 39172 + }, + { + "epoch": 2.718270765387551, + "grad_norm": 1.7108646631240845, + "learning_rate": 4.918961372307007e-05, + "loss": 0.3745, + "step": 39173 + }, + { + "epoch": 2.718340156824648, + "grad_norm": 1.4697856903076172, + "learning_rate": 4.918477451333709e-05, + "loss": 0.4533, + "step": 39174 + }, + { + "epoch": 2.7184095482617447, + "grad_norm": 1.0210260152816772, + "learning_rate": 4.917993546402186e-05, + "loss": 0.3931, + "step": 39175 + }, + { + "epoch": 2.718478939698841, + "grad_norm": 1.3950570821762085, + "learning_rate": 4.91750965751397e-05, + "loss": 0.607, + "step": 39176 + }, + { + "epoch": 2.718548331135938, + "grad_norm": 0.7455658912658691, + "learning_rate": 4.9170257846705845e-05, + "loss": 0.2575, + "step": 39177 + }, + { + "epoch": 2.7186177225730344, + "grad_norm": 0.9047746658325195, + "learning_rate": 4.916541927873563e-05, + "loss": 0.3242, + "step": 39178 + }, + { + "epoch": 2.718687114010131, + "grad_norm": 1.3628264665603638, + "learning_rate": 4.916058087124426e-05, + "loss": 0.5109, + "step": 39179 + }, + { + "epoch": 2.718756505447228, + "grad_norm": 1.0713136196136475, + "learning_rate": 4.9155742624247035e-05, + "loss": 0.4314, + "step": 39180 + }, + { + "epoch": 2.7188258968843244, + "grad_norm": 1.0868520736694336, + "learning_rate": 4.915090453775924e-05, + "loss": 0.4029, + "step": 39181 + }, + { + "epoch": 2.7188952883214212, + "grad_norm": 0.9123558402061462, + "learning_rate": 4.914606661179612e-05, + "loss": 0.4315, + "step": 39182 + }, + { + "epoch": 2.7189646797585176, + "grad_norm": 1.4236260652542114, + "learning_rate": 4.914122884637298e-05, + "loss": 0.4665, + "step": 39183 + }, + { + "epoch": 2.7190340711956145, + "grad_norm": 1.1920257806777954, + "learning_rate": 4.913639124150507e-05, + "loss": 0.4467, + "step": 39184 + }, + { + "epoch": 2.7191034626327113, + "grad_norm": 1.2344205379486084, + "learning_rate": 4.913155379720772e-05, + "loss": 0.4088, + "step": 39185 + }, + { + "epoch": 2.7191728540698077, + "grad_norm": 1.02189302444458, + "learning_rate": 4.9126716513496115e-05, + "loss": 0.5511, + "step": 39186 + }, + { + "epoch": 2.7192422455069045, + "grad_norm": 1.3928145170211792, + "learning_rate": 4.912187939038556e-05, + "loss": 0.3647, + "step": 39187 + }, + { + "epoch": 2.719311636944001, + "grad_norm": 1.2117881774902344, + "learning_rate": 4.911704242789132e-05, + "loss": 0.4371, + "step": 39188 + }, + { + "epoch": 2.7193810283810977, + "grad_norm": 0.9259148836135864, + "learning_rate": 4.911220562602868e-05, + "loss": 0.2975, + "step": 39189 + }, + { + "epoch": 2.7194504198181946, + "grad_norm": 0.8565690517425537, + "learning_rate": 4.910736898481292e-05, + "loss": 0.3531, + "step": 39190 + }, + { + "epoch": 2.719519811255291, + "grad_norm": 1.194813847541809, + "learning_rate": 4.9102532504259266e-05, + "loss": 0.4549, + "step": 39191 + }, + { + "epoch": 2.719589202692388, + "grad_norm": 1.3071585893630981, + "learning_rate": 4.9097696184383065e-05, + "loss": 0.3762, + "step": 39192 + }, + { + "epoch": 2.719658594129484, + "grad_norm": 0.9992847442626953, + "learning_rate": 4.9092860025199484e-05, + "loss": 0.4323, + "step": 39193 + }, + { + "epoch": 2.719727985566581, + "grad_norm": 1.1467759609222412, + "learning_rate": 4.908802402672384e-05, + "loss": 0.5095, + "step": 39194 + }, + { + "epoch": 2.719797377003678, + "grad_norm": 2.2728495597839355, + "learning_rate": 4.908318818897141e-05, + "loss": 0.4861, + "step": 39195 + }, + { + "epoch": 2.7198667684407747, + "grad_norm": 0.9829422831535339, + "learning_rate": 4.907835251195742e-05, + "loss": 0.3379, + "step": 39196 + }, + { + "epoch": 2.719936159877871, + "grad_norm": 1.2131054401397705, + "learning_rate": 4.907351699569719e-05, + "loss": 0.5048, + "step": 39197 + }, + { + "epoch": 2.7200055513149675, + "grad_norm": 1.3392328023910522, + "learning_rate": 4.906868164020598e-05, + "loss": 0.4154, + "step": 39198 + }, + { + "epoch": 2.7200749427520643, + "grad_norm": 1.4507297277450562, + "learning_rate": 4.906384644549901e-05, + "loss": 0.3186, + "step": 39199 + }, + { + "epoch": 2.720144334189161, + "grad_norm": 1.5754021406173706, + "learning_rate": 4.905901141159156e-05, + "loss": 0.383, + "step": 39200 + }, + { + "epoch": 2.720213725626258, + "grad_norm": 0.7551518082618713, + "learning_rate": 4.905417653849893e-05, + "loss": 0.2966, + "step": 39201 + }, + { + "epoch": 2.7202831170633544, + "grad_norm": 0.6606706976890564, + "learning_rate": 4.904934182623633e-05, + "loss": 0.3017, + "step": 39202 + }, + { + "epoch": 2.720352508500451, + "grad_norm": 1.366647720336914, + "learning_rate": 4.904450727481903e-05, + "loss": 0.3418, + "step": 39203 + }, + { + "epoch": 2.7204218999375476, + "grad_norm": 0.8921118974685669, + "learning_rate": 4.903967288426232e-05, + "loss": 0.2535, + "step": 39204 + }, + { + "epoch": 2.7204912913746444, + "grad_norm": 1.1052838563919067, + "learning_rate": 4.903483865458145e-05, + "loss": 0.5315, + "step": 39205 + }, + { + "epoch": 2.7205606828117412, + "grad_norm": 0.697727382183075, + "learning_rate": 4.9030004585791724e-05, + "loss": 0.2408, + "step": 39206 + }, + { + "epoch": 2.7206300742488376, + "grad_norm": 1.0749108791351318, + "learning_rate": 4.90251706779083e-05, + "loss": 0.4263, + "step": 39207 + }, + { + "epoch": 2.7206994656859345, + "grad_norm": 0.616226851940155, + "learning_rate": 4.902033693094652e-05, + "loss": 0.2226, + "step": 39208 + }, + { + "epoch": 2.720768857123031, + "grad_norm": 0.7201557159423828, + "learning_rate": 4.901550334492162e-05, + "loss": 0.1799, + "step": 39209 + }, + { + "epoch": 2.7208382485601277, + "grad_norm": 1.3511146306991577, + "learning_rate": 4.901066991984885e-05, + "loss": 0.5093, + "step": 39210 + }, + { + "epoch": 2.7209076399972245, + "grad_norm": 1.3211833238601685, + "learning_rate": 4.900583665574349e-05, + "loss": 0.3602, + "step": 39211 + }, + { + "epoch": 2.720977031434321, + "grad_norm": 0.9004718065261841, + "learning_rate": 4.900100355262077e-05, + "loss": 0.3774, + "step": 39212 + }, + { + "epoch": 2.7210464228714177, + "grad_norm": 1.7292217016220093, + "learning_rate": 4.899617061049602e-05, + "loss": 0.5882, + "step": 39213 + }, + { + "epoch": 2.721115814308514, + "grad_norm": 0.9934225082397461, + "learning_rate": 4.89913378293844e-05, + "loss": 0.2838, + "step": 39214 + }, + { + "epoch": 2.721185205745611, + "grad_norm": 1.4761784076690674, + "learning_rate": 4.89865052093012e-05, + "loss": 0.6264, + "step": 39215 + }, + { + "epoch": 2.721254597182708, + "grad_norm": 0.8815251588821411, + "learning_rate": 4.89816727502617e-05, + "loss": 0.3455, + "step": 39216 + }, + { + "epoch": 2.721323988619804, + "grad_norm": 1.2259305715560913, + "learning_rate": 4.897684045228113e-05, + "loss": 0.5439, + "step": 39217 + }, + { + "epoch": 2.721393380056901, + "grad_norm": 1.3180993795394897, + "learning_rate": 4.8972008315374754e-05, + "loss": 0.3017, + "step": 39218 + }, + { + "epoch": 2.7214627714939974, + "grad_norm": 1.0070480108261108, + "learning_rate": 4.896717633955783e-05, + "loss": 0.3445, + "step": 39219 + }, + { + "epoch": 2.7215321629310942, + "grad_norm": 0.9385769367218018, + "learning_rate": 4.896234452484565e-05, + "loss": 0.4359, + "step": 39220 + }, + { + "epoch": 2.721601554368191, + "grad_norm": 1.0740020275115967, + "learning_rate": 4.8957512871253386e-05, + "loss": 0.4274, + "step": 39221 + }, + { + "epoch": 2.721670945805288, + "grad_norm": 0.9085135459899902, + "learning_rate": 4.895268137879633e-05, + "loss": 0.4493, + "step": 39222 + }, + { + "epoch": 2.7217403372423843, + "grad_norm": 1.1107276678085327, + "learning_rate": 4.894785004748974e-05, + "loss": 0.2604, + "step": 39223 + }, + { + "epoch": 2.7218097286794807, + "grad_norm": 3.8057398796081543, + "learning_rate": 4.894301887734886e-05, + "loss": 0.537, + "step": 39224 + }, + { + "epoch": 2.7218791201165775, + "grad_norm": 1.0128346681594849, + "learning_rate": 4.8938187868388955e-05, + "loss": 0.4078, + "step": 39225 + }, + { + "epoch": 2.7219485115536743, + "grad_norm": 1.2057700157165527, + "learning_rate": 4.893335702062529e-05, + "loss": 0.3969, + "step": 39226 + }, + { + "epoch": 2.722017902990771, + "grad_norm": 1.4043980836868286, + "learning_rate": 4.892852633407306e-05, + "loss": 0.2507, + "step": 39227 + }, + { + "epoch": 2.7220872944278676, + "grad_norm": 1.4998360872268677, + "learning_rate": 4.8923695808747546e-05, + "loss": 0.4655, + "step": 39228 + }, + { + "epoch": 2.7221566858649644, + "grad_norm": 1.0878735780715942, + "learning_rate": 4.891886544466401e-05, + "loss": 0.2639, + "step": 39229 + }, + { + "epoch": 2.722226077302061, + "grad_norm": 0.96293705701828, + "learning_rate": 4.891403524183767e-05, + "loss": 0.2076, + "step": 39230 + }, + { + "epoch": 2.7222954687391576, + "grad_norm": 0.9725903868675232, + "learning_rate": 4.89092052002838e-05, + "loss": 0.4714, + "step": 39231 + }, + { + "epoch": 2.7223648601762545, + "grad_norm": 1.2684228420257568, + "learning_rate": 4.890437532001769e-05, + "loss": 0.3281, + "step": 39232 + }, + { + "epoch": 2.722434251613351, + "grad_norm": 1.1538033485412598, + "learning_rate": 4.8899545601054496e-05, + "loss": 0.4401, + "step": 39233 + }, + { + "epoch": 2.7225036430504477, + "grad_norm": 1.4178602695465088, + "learning_rate": 4.8894716043409505e-05, + "loss": 0.4127, + "step": 39234 + }, + { + "epoch": 2.722573034487544, + "grad_norm": 1.3791507482528687, + "learning_rate": 4.888988664709797e-05, + "loss": 0.3525, + "step": 39235 + }, + { + "epoch": 2.722642425924641, + "grad_norm": 1.9362870454788208, + "learning_rate": 4.888505741213517e-05, + "loss": 0.3855, + "step": 39236 + }, + { + "epoch": 2.7227118173617377, + "grad_norm": 1.0735745429992676, + "learning_rate": 4.888022833853627e-05, + "loss": 0.3735, + "step": 39237 + }, + { + "epoch": 2.722781208798834, + "grad_norm": 1.5960336923599243, + "learning_rate": 4.887539942631656e-05, + "loss": 0.5559, + "step": 39238 + }, + { + "epoch": 2.722850600235931, + "grad_norm": 1.0948818922042847, + "learning_rate": 4.887057067549129e-05, + "loss": 0.3996, + "step": 39239 + }, + { + "epoch": 2.7229199916730273, + "grad_norm": 1.7839019298553467, + "learning_rate": 4.886574208607573e-05, + "loss": 0.4431, + "step": 39240 + }, + { + "epoch": 2.722989383110124, + "grad_norm": 1.5218894481658936, + "learning_rate": 4.886091365808504e-05, + "loss": 0.4769, + "step": 39241 + }, + { + "epoch": 2.723058774547221, + "grad_norm": 0.9275079965591431, + "learning_rate": 4.885608539153452e-05, + "loss": 0.357, + "step": 39242 + }, + { + "epoch": 2.7231281659843174, + "grad_norm": 1.0728912353515625, + "learning_rate": 4.885125728643941e-05, + "loss": 0.4021, + "step": 39243 + }, + { + "epoch": 2.7231975574214142, + "grad_norm": 1.5719205141067505, + "learning_rate": 4.884642934281495e-05, + "loss": 0.494, + "step": 39244 + }, + { + "epoch": 2.7232669488585106, + "grad_norm": 1.2572194337844849, + "learning_rate": 4.884160156067636e-05, + "loss": 0.413, + "step": 39245 + }, + { + "epoch": 2.7233363402956074, + "grad_norm": 1.0963282585144043, + "learning_rate": 4.883677394003892e-05, + "loss": 0.4007, + "step": 39246 + }, + { + "epoch": 2.7234057317327043, + "grad_norm": 1.1657063961029053, + "learning_rate": 4.883194648091788e-05, + "loss": 0.3833, + "step": 39247 + }, + { + "epoch": 2.723475123169801, + "grad_norm": 1.0437114238739014, + "learning_rate": 4.882711918332839e-05, + "loss": 0.303, + "step": 39248 + }, + { + "epoch": 2.7235445146068975, + "grad_norm": 1.0099234580993652, + "learning_rate": 4.882229204728578e-05, + "loss": 0.3089, + "step": 39249 + }, + { + "epoch": 2.7236139060439943, + "grad_norm": 0.8943188786506653, + "learning_rate": 4.8817465072805235e-05, + "loss": 0.318, + "step": 39250 + }, + { + "epoch": 2.7236832974810907, + "grad_norm": 1.0910203456878662, + "learning_rate": 4.881263825990202e-05, + "loss": 0.3527, + "step": 39251 + }, + { + "epoch": 2.7237526889181876, + "grad_norm": 1.1098181009292603, + "learning_rate": 4.880781160859137e-05, + "loss": 0.3636, + "step": 39252 + }, + { + "epoch": 2.7238220803552844, + "grad_norm": 1.2588437795639038, + "learning_rate": 4.8802985118888514e-05, + "loss": 0.3547, + "step": 39253 + }, + { + "epoch": 2.723891471792381, + "grad_norm": 1.3120256662368774, + "learning_rate": 4.879815879080873e-05, + "loss": 0.5024, + "step": 39254 + }, + { + "epoch": 2.7239608632294776, + "grad_norm": 0.9129072427749634, + "learning_rate": 4.879333262436718e-05, + "loss": 0.3134, + "step": 39255 + }, + { + "epoch": 2.724030254666574, + "grad_norm": 1.466850996017456, + "learning_rate": 4.878850661957914e-05, + "loss": 0.361, + "step": 39256 + }, + { + "epoch": 2.724099646103671, + "grad_norm": 1.492959976196289, + "learning_rate": 4.878368077645985e-05, + "loss": 0.3969, + "step": 39257 + }, + { + "epoch": 2.7241690375407677, + "grad_norm": 2.1715128421783447, + "learning_rate": 4.877885509502453e-05, + "loss": 0.4014, + "step": 39258 + }, + { + "epoch": 2.724238428977864, + "grad_norm": 1.238026738166809, + "learning_rate": 4.8774029575288424e-05, + "loss": 0.3924, + "step": 39259 + }, + { + "epoch": 2.724307820414961, + "grad_norm": 1.0582568645477295, + "learning_rate": 4.8769204217266775e-05, + "loss": 0.3578, + "step": 39260 + }, + { + "epoch": 2.7243772118520573, + "grad_norm": 1.4354584217071533, + "learning_rate": 4.8764379020974826e-05, + "loss": 0.4988, + "step": 39261 + }, + { + "epoch": 2.724446603289154, + "grad_norm": 1.3233284950256348, + "learning_rate": 4.8759553986427755e-05, + "loss": 0.4005, + "step": 39262 + }, + { + "epoch": 2.724515994726251, + "grad_norm": 1.6695513725280762, + "learning_rate": 4.875472911364083e-05, + "loss": 0.5899, + "step": 39263 + }, + { + "epoch": 2.7245853861633473, + "grad_norm": 1.0470045804977417, + "learning_rate": 4.8749904402629275e-05, + "loss": 0.3721, + "step": 39264 + }, + { + "epoch": 2.724654777600444, + "grad_norm": 1.5122408866882324, + "learning_rate": 4.8745079853408325e-05, + "loss": 0.3744, + "step": 39265 + }, + { + "epoch": 2.7247241690375406, + "grad_norm": 1.3956239223480225, + "learning_rate": 4.8740255465993226e-05, + "loss": 0.3622, + "step": 39266 + }, + { + "epoch": 2.7247935604746374, + "grad_norm": 2.563361883163452, + "learning_rate": 4.873543124039921e-05, + "loss": 0.5179, + "step": 39267 + }, + { + "epoch": 2.724862951911734, + "grad_norm": 1.0729718208312988, + "learning_rate": 4.873060717664146e-05, + "loss": 0.3729, + "step": 39268 + }, + { + "epoch": 2.7249323433488306, + "grad_norm": 1.4043467044830322, + "learning_rate": 4.872578327473524e-05, + "loss": 0.3812, + "step": 39269 + }, + { + "epoch": 2.7250017347859274, + "grad_norm": 1.456457495689392, + "learning_rate": 4.8720959534695766e-05, + "loss": 0.4309, + "step": 39270 + }, + { + "epoch": 2.725071126223024, + "grad_norm": 1.3958067893981934, + "learning_rate": 4.871613595653831e-05, + "loss": 0.5845, + "step": 39271 + }, + { + "epoch": 2.7251405176601207, + "grad_norm": 0.9522707462310791, + "learning_rate": 4.871131254027803e-05, + "loss": 0.2444, + "step": 39272 + }, + { + "epoch": 2.7252099090972175, + "grad_norm": 1.2391585111618042, + "learning_rate": 4.870648928593019e-05, + "loss": 0.5712, + "step": 39273 + }, + { + "epoch": 2.7252793005343143, + "grad_norm": 1.842248558998108, + "learning_rate": 4.870166619351001e-05, + "loss": 0.451, + "step": 39274 + }, + { + "epoch": 2.7253486919714107, + "grad_norm": 1.2455298900604248, + "learning_rate": 4.8696843263032755e-05, + "loss": 0.3789, + "step": 39275 + }, + { + "epoch": 2.7254180834085076, + "grad_norm": 1.115381121635437, + "learning_rate": 4.869202049451357e-05, + "loss": 0.3622, + "step": 39276 + }, + { + "epoch": 2.725487474845604, + "grad_norm": 1.2493278980255127, + "learning_rate": 4.868719788796773e-05, + "loss": 0.3889, + "step": 39277 + }, + { + "epoch": 2.7255568662827008, + "grad_norm": 1.7806984186172485, + "learning_rate": 4.868237544341044e-05, + "loss": 0.4327, + "step": 39278 + }, + { + "epoch": 2.7256262577197976, + "grad_norm": 1.2098380327224731, + "learning_rate": 4.8677553160856945e-05, + "loss": 0.3237, + "step": 39279 + }, + { + "epoch": 2.725695649156894, + "grad_norm": 1.6368343830108643, + "learning_rate": 4.867273104032247e-05, + "loss": 0.4787, + "step": 39280 + }, + { + "epoch": 2.725765040593991, + "grad_norm": 1.1260236501693726, + "learning_rate": 4.866790908182222e-05, + "loss": 0.3161, + "step": 39281 + }, + { + "epoch": 2.725834432031087, + "grad_norm": 1.732722520828247, + "learning_rate": 4.866308728537147e-05, + "loss": 0.3257, + "step": 39282 + }, + { + "epoch": 2.725903823468184, + "grad_norm": 1.0840810537338257, + "learning_rate": 4.865826565098535e-05, + "loss": 0.4136, + "step": 39283 + }, + { + "epoch": 2.725973214905281, + "grad_norm": 0.7125611305236816, + "learning_rate": 4.865344417867913e-05, + "loss": 0.2644, + "step": 39284 + }, + { + "epoch": 2.7260426063423773, + "grad_norm": 1.6928728818893433, + "learning_rate": 4.864862286846803e-05, + "loss": 0.5122, + "step": 39285 + }, + { + "epoch": 2.726111997779474, + "grad_norm": 1.260379433631897, + "learning_rate": 4.864380172036727e-05, + "loss": 0.453, + "step": 39286 + }, + { + "epoch": 2.7261813892165705, + "grad_norm": 1.3950796127319336, + "learning_rate": 4.863898073439207e-05, + "loss": 0.5158, + "step": 39287 + }, + { + "epoch": 2.7262507806536673, + "grad_norm": 0.8778393268585205, + "learning_rate": 4.8634159910557644e-05, + "loss": 0.2618, + "step": 39288 + }, + { + "epoch": 2.726320172090764, + "grad_norm": 5.23573637008667, + "learning_rate": 4.862933924887926e-05, + "loss": 0.4136, + "step": 39289 + }, + { + "epoch": 2.7263895635278605, + "grad_norm": 1.1867352724075317, + "learning_rate": 4.862451874937205e-05, + "loss": 0.5276, + "step": 39290 + }, + { + "epoch": 2.7264589549649574, + "grad_norm": 2.112647533416748, + "learning_rate": 4.8619698412051286e-05, + "loss": 0.5224, + "step": 39291 + }, + { + "epoch": 2.7265283464020538, + "grad_norm": 1.1287840604782104, + "learning_rate": 4.861487823693216e-05, + "loss": 0.3626, + "step": 39292 + }, + { + "epoch": 2.7265977378391506, + "grad_norm": 1.314681887626648, + "learning_rate": 4.86100582240299e-05, + "loss": 0.4636, + "step": 39293 + }, + { + "epoch": 2.7266671292762474, + "grad_norm": 1.4819148778915405, + "learning_rate": 4.860523837335973e-05, + "loss": 0.4282, + "step": 39294 + }, + { + "epoch": 2.726736520713344, + "grad_norm": 1.5356934070587158, + "learning_rate": 4.8600418684936854e-05, + "loss": 0.5512, + "step": 39295 + }, + { + "epoch": 2.7268059121504407, + "grad_norm": 1.7949250936508179, + "learning_rate": 4.859559915877654e-05, + "loss": 0.3147, + "step": 39296 + }, + { + "epoch": 2.726875303587537, + "grad_norm": 1.252518892288208, + "learning_rate": 4.8590779794893914e-05, + "loss": 0.384, + "step": 39297 + }, + { + "epoch": 2.726944695024634, + "grad_norm": 1.2684345245361328, + "learning_rate": 4.8585960593304224e-05, + "loss": 0.3826, + "step": 39298 + }, + { + "epoch": 2.7270140864617307, + "grad_norm": 0.9293981194496155, + "learning_rate": 4.85811415540227e-05, + "loss": 0.3033, + "step": 39299 + }, + { + "epoch": 2.7270834778988275, + "grad_norm": 0.9746467471122742, + "learning_rate": 4.857632267706454e-05, + "loss": 0.3773, + "step": 39300 + }, + { + "epoch": 2.727152869335924, + "grad_norm": 0.8409731388092041, + "learning_rate": 4.857150396244496e-05, + "loss": 0.3589, + "step": 39301 + }, + { + "epoch": 2.7272222607730208, + "grad_norm": 1.5023797750473022, + "learning_rate": 4.8566685410179224e-05, + "loss": 0.3857, + "step": 39302 + }, + { + "epoch": 2.727291652210117, + "grad_norm": 1.6173704862594604, + "learning_rate": 4.856186702028246e-05, + "loss": 0.6121, + "step": 39303 + }, + { + "epoch": 2.727361043647214, + "grad_norm": 1.3765099048614502, + "learning_rate": 4.8557048792769896e-05, + "loss": 0.3461, + "step": 39304 + }, + { + "epoch": 2.727430435084311, + "grad_norm": 1.088538408279419, + "learning_rate": 4.855223072765678e-05, + "loss": 0.4182, + "step": 39305 + }, + { + "epoch": 2.727499826521407, + "grad_norm": 1.2989896535873413, + "learning_rate": 4.854741282495833e-05, + "loss": 0.3993, + "step": 39306 + }, + { + "epoch": 2.727569217958504, + "grad_norm": 1.4187263250350952, + "learning_rate": 4.854259508468969e-05, + "loss": 0.4421, + "step": 39307 + }, + { + "epoch": 2.7276386093956004, + "grad_norm": 0.9322051405906677, + "learning_rate": 4.853777750686611e-05, + "loss": 0.3604, + "step": 39308 + }, + { + "epoch": 2.7277080008326973, + "grad_norm": 1.3112990856170654, + "learning_rate": 4.853296009150283e-05, + "loss": 0.4704, + "step": 39309 + }, + { + "epoch": 2.727777392269794, + "grad_norm": 0.7961640357971191, + "learning_rate": 4.852814283861499e-05, + "loss": 0.2972, + "step": 39310 + }, + { + "epoch": 2.7278467837068905, + "grad_norm": 0.9748188853263855, + "learning_rate": 4.852332574821783e-05, + "loss": 0.3734, + "step": 39311 + }, + { + "epoch": 2.7279161751439873, + "grad_norm": 1.383086085319519, + "learning_rate": 4.8518508820326556e-05, + "loss": 0.5064, + "step": 39312 + }, + { + "epoch": 2.7279855665810837, + "grad_norm": 0.8926776647567749, + "learning_rate": 4.8513692054956386e-05, + "loss": 0.2673, + "step": 39313 + }, + { + "epoch": 2.7280549580181805, + "grad_norm": 1.6444389820098877, + "learning_rate": 4.85088754521225e-05, + "loss": 0.5904, + "step": 39314 + }, + { + "epoch": 2.7281243494552774, + "grad_norm": 1.291946291923523, + "learning_rate": 4.850405901184013e-05, + "loss": 0.4, + "step": 39315 + }, + { + "epoch": 2.7281937408923738, + "grad_norm": 1.4920967817306519, + "learning_rate": 4.8499242734124505e-05, + "loss": 0.3566, + "step": 39316 + }, + { + "epoch": 2.7282631323294706, + "grad_norm": 1.4003267288208008, + "learning_rate": 4.8494426618990765e-05, + "loss": 0.5321, + "step": 39317 + }, + { + "epoch": 2.728332523766567, + "grad_norm": 0.9470771551132202, + "learning_rate": 4.848961066645415e-05, + "loss": 0.3572, + "step": 39318 + }, + { + "epoch": 2.728401915203664, + "grad_norm": 1.24342679977417, + "learning_rate": 4.848479487652984e-05, + "loss": 0.5012, + "step": 39319 + }, + { + "epoch": 2.7284713066407607, + "grad_norm": 1.0570539236068726, + "learning_rate": 4.847997924923307e-05, + "loss": 0.395, + "step": 39320 + }, + { + "epoch": 2.728540698077857, + "grad_norm": 1.6840187311172485, + "learning_rate": 4.847516378457901e-05, + "loss": 0.4501, + "step": 39321 + }, + { + "epoch": 2.728610089514954, + "grad_norm": 0.97407466173172, + "learning_rate": 4.8470348482582894e-05, + "loss": 0.4488, + "step": 39322 + }, + { + "epoch": 2.7286794809520503, + "grad_norm": 1.1714770793914795, + "learning_rate": 4.846553334325994e-05, + "loss": 0.2826, + "step": 39323 + }, + { + "epoch": 2.728748872389147, + "grad_norm": 1.3263731002807617, + "learning_rate": 4.846071836662529e-05, + "loss": 0.421, + "step": 39324 + }, + { + "epoch": 2.728818263826244, + "grad_norm": 1.2242363691329956, + "learning_rate": 4.845590355269417e-05, + "loss": 0.3122, + "step": 39325 + }, + { + "epoch": 2.7288876552633408, + "grad_norm": 1.2585152387619019, + "learning_rate": 4.845108890148178e-05, + "loss": 0.4374, + "step": 39326 + }, + { + "epoch": 2.728957046700437, + "grad_norm": 0.9681022763252258, + "learning_rate": 4.844627441300331e-05, + "loss": 0.359, + "step": 39327 + }, + { + "epoch": 2.729026438137534, + "grad_norm": 1.1496881246566772, + "learning_rate": 4.8441460087273997e-05, + "loss": 0.3361, + "step": 39328 + }, + { + "epoch": 2.7290958295746304, + "grad_norm": 1.422633409500122, + "learning_rate": 4.843664592430899e-05, + "loss": 0.4135, + "step": 39329 + }, + { + "epoch": 2.729165221011727, + "grad_norm": 0.7555365562438965, + "learning_rate": 4.843183192412355e-05, + "loss": 0.254, + "step": 39330 + }, + { + "epoch": 2.729234612448824, + "grad_norm": 1.2209712266921997, + "learning_rate": 4.84270180867328e-05, + "loss": 0.4996, + "step": 39331 + }, + { + "epoch": 2.7293040038859204, + "grad_norm": 0.8661304116249084, + "learning_rate": 4.842220441215197e-05, + "loss": 0.2599, + "step": 39332 + }, + { + "epoch": 2.7293733953230173, + "grad_norm": 1.0209641456604004, + "learning_rate": 4.841739090039627e-05, + "loss": 0.3063, + "step": 39333 + }, + { + "epoch": 2.7294427867601136, + "grad_norm": 1.0127469301223755, + "learning_rate": 4.8412577551480867e-05, + "loss": 0.3609, + "step": 39334 + }, + { + "epoch": 2.7295121781972105, + "grad_norm": 1.6829699277877808, + "learning_rate": 4.840776436542097e-05, + "loss": 0.6562, + "step": 39335 + }, + { + "epoch": 2.7295815696343073, + "grad_norm": 1.4382673501968384, + "learning_rate": 4.840295134223177e-05, + "loss": 0.3954, + "step": 39336 + }, + { + "epoch": 2.7296509610714037, + "grad_norm": 0.6868923902511597, + "learning_rate": 4.839813848192852e-05, + "loss": 0.1529, + "step": 39337 + }, + { + "epoch": 2.7297203525085005, + "grad_norm": 1.436874270439148, + "learning_rate": 4.839332578452631e-05, + "loss": 0.5233, + "step": 39338 + }, + { + "epoch": 2.729789743945597, + "grad_norm": 1.3088215589523315, + "learning_rate": 4.838851325004039e-05, + "loss": 0.46, + "step": 39339 + }, + { + "epoch": 2.7298591353826938, + "grad_norm": 1.2247679233551025, + "learning_rate": 4.8383700878485946e-05, + "loss": 0.433, + "step": 39340 + }, + { + "epoch": 2.7299285268197906, + "grad_norm": 1.3655532598495483, + "learning_rate": 4.837888866987821e-05, + "loss": 0.4228, + "step": 39341 + }, + { + "epoch": 2.729997918256887, + "grad_norm": 1.6949560642242432, + "learning_rate": 4.8374076624232277e-05, + "loss": 0.3247, + "step": 39342 + }, + { + "epoch": 2.730067309693984, + "grad_norm": 1.1663670539855957, + "learning_rate": 4.836926474156341e-05, + "loss": 0.3156, + "step": 39343 + }, + { + "epoch": 2.73013670113108, + "grad_norm": 0.9995930790901184, + "learning_rate": 4.836445302188682e-05, + "loss": 0.418, + "step": 39344 + }, + { + "epoch": 2.730206092568177, + "grad_norm": 1.104317307472229, + "learning_rate": 4.835964146521763e-05, + "loss": 0.4427, + "step": 39345 + }, + { + "epoch": 2.730275484005274, + "grad_norm": 1.3447654247283936, + "learning_rate": 4.8354830071571054e-05, + "loss": 0.3637, + "step": 39346 + }, + { + "epoch": 2.7303448754423703, + "grad_norm": 0.95606529712677, + "learning_rate": 4.835001884096229e-05, + "loss": 0.3052, + "step": 39347 + }, + { + "epoch": 2.730414266879467, + "grad_norm": 1.4115402698516846, + "learning_rate": 4.834520777340651e-05, + "loss": 0.3933, + "step": 39348 + }, + { + "epoch": 2.7304836583165635, + "grad_norm": 1.4327856302261353, + "learning_rate": 4.8340396868918935e-05, + "loss": 0.3042, + "step": 39349 + }, + { + "epoch": 2.7305530497536603, + "grad_norm": 1.894888997077942, + "learning_rate": 4.833558612751472e-05, + "loss": 0.6023, + "step": 39350 + }, + { + "epoch": 2.730622441190757, + "grad_norm": 2.2927238941192627, + "learning_rate": 4.833077554920911e-05, + "loss": 0.4463, + "step": 39351 + }, + { + "epoch": 2.730691832627854, + "grad_norm": 1.6204029321670532, + "learning_rate": 4.83259651340172e-05, + "loss": 0.4863, + "step": 39352 + }, + { + "epoch": 2.7307612240649504, + "grad_norm": 0.8469702005386353, + "learning_rate": 4.8321154881954233e-05, + "loss": 0.2925, + "step": 39353 + }, + { + "epoch": 2.730830615502047, + "grad_norm": 1.2396798133850098, + "learning_rate": 4.831634479303539e-05, + "loss": 0.3936, + "step": 39354 + }, + { + "epoch": 2.7309000069391436, + "grad_norm": 1.1862465143203735, + "learning_rate": 4.831153486727584e-05, + "loss": 0.4638, + "step": 39355 + }, + { + "epoch": 2.7309693983762404, + "grad_norm": 0.9569802284240723, + "learning_rate": 4.8306725104690776e-05, + "loss": 0.3802, + "step": 39356 + }, + { + "epoch": 2.7310387898133373, + "grad_norm": 1.2186338901519775, + "learning_rate": 4.830191550529539e-05, + "loss": 0.5299, + "step": 39357 + }, + { + "epoch": 2.7311081812504336, + "grad_norm": 0.851592481136322, + "learning_rate": 4.829710606910489e-05, + "loss": 0.2712, + "step": 39358 + }, + { + "epoch": 2.7311775726875305, + "grad_norm": 1.1465243101119995, + "learning_rate": 4.829229679613439e-05, + "loss": 0.411, + "step": 39359 + }, + { + "epoch": 2.731246964124627, + "grad_norm": 1.538689136505127, + "learning_rate": 4.828748768639911e-05, + "loss": 0.6189, + "step": 39360 + }, + { + "epoch": 2.7313163555617237, + "grad_norm": 1.5959150791168213, + "learning_rate": 4.8282678739914244e-05, + "loss": 0.4587, + "step": 39361 + }, + { + "epoch": 2.7313857469988205, + "grad_norm": 1.0771573781967163, + "learning_rate": 4.8277869956694945e-05, + "loss": 0.3651, + "step": 39362 + }, + { + "epoch": 2.731455138435917, + "grad_norm": 1.3387395143508911, + "learning_rate": 4.827306133675642e-05, + "loss": 0.3913, + "step": 39363 + }, + { + "epoch": 2.7315245298730138, + "grad_norm": 1.1196576356887817, + "learning_rate": 4.826825288011383e-05, + "loss": 0.3974, + "step": 39364 + }, + { + "epoch": 2.73159392131011, + "grad_norm": 0.9881464838981628, + "learning_rate": 4.82634445867824e-05, + "loss": 0.5799, + "step": 39365 + }, + { + "epoch": 2.731663312747207, + "grad_norm": 1.3461798429489136, + "learning_rate": 4.825863645677725e-05, + "loss": 0.3462, + "step": 39366 + }, + { + "epoch": 2.731732704184304, + "grad_norm": 1.0432350635528564, + "learning_rate": 4.825382849011357e-05, + "loss": 0.2764, + "step": 39367 + }, + { + "epoch": 2.7318020956214, + "grad_norm": 1.2071378231048584, + "learning_rate": 4.8249020686806555e-05, + "loss": 0.2864, + "step": 39368 + }, + { + "epoch": 2.731871487058497, + "grad_norm": 1.0679564476013184, + "learning_rate": 4.824421304687138e-05, + "loss": 0.388, + "step": 39369 + }, + { + "epoch": 2.7319408784955934, + "grad_norm": 0.9096198678016663, + "learning_rate": 4.8239405570323206e-05, + "loss": 0.3338, + "step": 39370 + }, + { + "epoch": 2.7320102699326902, + "grad_norm": 0.7886230945587158, + "learning_rate": 4.823459825717723e-05, + "loss": 0.2558, + "step": 39371 + }, + { + "epoch": 2.732079661369787, + "grad_norm": 1.1361310482025146, + "learning_rate": 4.822979110744866e-05, + "loss": 0.4497, + "step": 39372 + }, + { + "epoch": 2.7321490528068835, + "grad_norm": 1.4453637599945068, + "learning_rate": 4.82249841211526e-05, + "loss": 0.3393, + "step": 39373 + }, + { + "epoch": 2.7322184442439803, + "grad_norm": 0.9501438140869141, + "learning_rate": 4.822017729830426e-05, + "loss": 0.3244, + "step": 39374 + }, + { + "epoch": 2.7322878356810767, + "grad_norm": 1.3829951286315918, + "learning_rate": 4.8215370638918847e-05, + "loss": 0.4458, + "step": 39375 + }, + { + "epoch": 2.7323572271181735, + "grad_norm": 0.9579158425331116, + "learning_rate": 4.821056414301145e-05, + "loss": 0.2672, + "step": 39376 + }, + { + "epoch": 2.7324266185552704, + "grad_norm": 1.1024610996246338, + "learning_rate": 4.8205757810597304e-05, + "loss": 0.3704, + "step": 39377 + }, + { + "epoch": 2.732496009992367, + "grad_norm": 0.9157585501670837, + "learning_rate": 4.820095164169157e-05, + "loss": 0.2521, + "step": 39378 + }, + { + "epoch": 2.7325654014294636, + "grad_norm": 1.116348385810852, + "learning_rate": 4.819614563630947e-05, + "loss": 0.3472, + "step": 39379 + }, + { + "epoch": 2.7326347928665604, + "grad_norm": 1.0516501665115356, + "learning_rate": 4.819133979446607e-05, + "loss": 0.3932, + "step": 39380 + }, + { + "epoch": 2.732704184303657, + "grad_norm": 1.1032493114471436, + "learning_rate": 4.8186534116176616e-05, + "loss": 0.3992, + "step": 39381 + }, + { + "epoch": 2.7327735757407536, + "grad_norm": 1.856557011604309, + "learning_rate": 4.818172860145625e-05, + "loss": 0.3697, + "step": 39382 + }, + { + "epoch": 2.7328429671778505, + "grad_norm": 1.1204783916473389, + "learning_rate": 4.817692325032017e-05, + "loss": 0.2639, + "step": 39383 + }, + { + "epoch": 2.732912358614947, + "grad_norm": 1.4092886447906494, + "learning_rate": 4.817211806278352e-05, + "loss": 0.48, + "step": 39384 + }, + { + "epoch": 2.7329817500520437, + "grad_norm": 1.0265109539031982, + "learning_rate": 4.816731303886153e-05, + "loss": 0.351, + "step": 39385 + }, + { + "epoch": 2.73305114148914, + "grad_norm": 1.9533206224441528, + "learning_rate": 4.8162508178569264e-05, + "loss": 0.5275, + "step": 39386 + }, + { + "epoch": 2.733120532926237, + "grad_norm": 1.037127137184143, + "learning_rate": 4.815770348192195e-05, + "loss": 0.399, + "step": 39387 + }, + { + "epoch": 2.7331899243633337, + "grad_norm": 1.3495454788208008, + "learning_rate": 4.815289894893476e-05, + "loss": 0.3692, + "step": 39388 + }, + { + "epoch": 2.73325931580043, + "grad_norm": 0.9710584282875061, + "learning_rate": 4.814809457962285e-05, + "loss": 0.462, + "step": 39389 + }, + { + "epoch": 2.733328707237527, + "grad_norm": 1.1984822750091553, + "learning_rate": 4.8143290374001384e-05, + "loss": 0.4319, + "step": 39390 + }, + { + "epoch": 2.7333980986746234, + "grad_norm": 1.2276594638824463, + "learning_rate": 4.813848633208553e-05, + "loss": 0.5174, + "step": 39391 + }, + { + "epoch": 2.73346749011172, + "grad_norm": 1.7207703590393066, + "learning_rate": 4.8133682453890506e-05, + "loss": 0.5184, + "step": 39392 + }, + { + "epoch": 2.733536881548817, + "grad_norm": 0.9704475402832031, + "learning_rate": 4.812887873943138e-05, + "loss": 0.2523, + "step": 39393 + }, + { + "epoch": 2.7336062729859134, + "grad_norm": 0.9733821153640747, + "learning_rate": 4.812407518872337e-05, + "loss": 0.4003, + "step": 39394 + }, + { + "epoch": 2.7336756644230102, + "grad_norm": 0.8692581057548523, + "learning_rate": 4.811927180178164e-05, + "loss": 0.4437, + "step": 39395 + }, + { + "epoch": 2.7337450558601066, + "grad_norm": 1.381314992904663, + "learning_rate": 4.811446857862135e-05, + "loss": 0.4718, + "step": 39396 + }, + { + "epoch": 2.7338144472972035, + "grad_norm": 1.2674378156661987, + "learning_rate": 4.810966551925766e-05, + "loss": 0.4241, + "step": 39397 + }, + { + "epoch": 2.7338838387343003, + "grad_norm": 1.290381908416748, + "learning_rate": 4.8104862623705735e-05, + "loss": 0.4925, + "step": 39398 + }, + { + "epoch": 2.7339532301713967, + "grad_norm": 1.4410818815231323, + "learning_rate": 4.810005989198079e-05, + "loss": 0.6188, + "step": 39399 + }, + { + "epoch": 2.7340226216084935, + "grad_norm": 1.1046231985092163, + "learning_rate": 4.8095257324097875e-05, + "loss": 0.2918, + "step": 39400 + }, + { + "epoch": 2.73409201304559, + "grad_norm": 0.9587092399597168, + "learning_rate": 4.8090454920072216e-05, + "loss": 0.4039, + "step": 39401 + }, + { + "epoch": 2.7341614044826867, + "grad_norm": 0.9422168731689453, + "learning_rate": 4.8085652679918966e-05, + "loss": 0.2505, + "step": 39402 + }, + { + "epoch": 2.7342307959197836, + "grad_norm": 1.3060952425003052, + "learning_rate": 4.80808506036533e-05, + "loss": 0.3884, + "step": 39403 + }, + { + "epoch": 2.7343001873568804, + "grad_norm": 1.4622225761413574, + "learning_rate": 4.807604869129035e-05, + "loss": 0.2801, + "step": 39404 + }, + { + "epoch": 2.734369578793977, + "grad_norm": 1.3118364810943604, + "learning_rate": 4.80712469428453e-05, + "loss": 0.4159, + "step": 39405 + }, + { + "epoch": 2.7344389702310736, + "grad_norm": 1.04787278175354, + "learning_rate": 4.806644535833333e-05, + "loss": 0.4029, + "step": 39406 + }, + { + "epoch": 2.73450836166817, + "grad_norm": 1.03554105758667, + "learning_rate": 4.806164393776954e-05, + "loss": 0.4506, + "step": 39407 + }, + { + "epoch": 2.734577753105267, + "grad_norm": 1.0701674222946167, + "learning_rate": 4.805684268116911e-05, + "loss": 0.3203, + "step": 39408 + }, + { + "epoch": 2.7346471445423637, + "grad_norm": 1.425141453742981, + "learning_rate": 4.8052041588547195e-05, + "loss": 0.3631, + "step": 39409 + }, + { + "epoch": 2.73471653597946, + "grad_norm": 1.0513551235198975, + "learning_rate": 4.8047240659919004e-05, + "loss": 0.2748, + "step": 39410 + }, + { + "epoch": 2.734785927416557, + "grad_norm": 1.8703992366790771, + "learning_rate": 4.8042439895299594e-05, + "loss": 0.4206, + "step": 39411 + }, + { + "epoch": 2.7348553188536533, + "grad_norm": 1.2409353256225586, + "learning_rate": 4.8037639294704184e-05, + "loss": 0.3826, + "step": 39412 + }, + { + "epoch": 2.73492471029075, + "grad_norm": 1.169243574142456, + "learning_rate": 4.803283885814792e-05, + "loss": 0.4608, + "step": 39413 + }, + { + "epoch": 2.734994101727847, + "grad_norm": 1.0993741750717163, + "learning_rate": 4.8028038585645985e-05, + "loss": 0.3338, + "step": 39414 + }, + { + "epoch": 2.7350634931649433, + "grad_norm": 1.359204888343811, + "learning_rate": 4.802323847721347e-05, + "loss": 0.6472, + "step": 39415 + }, + { + "epoch": 2.73513288460204, + "grad_norm": 1.0149354934692383, + "learning_rate": 4.801843853286556e-05, + "loss": 0.3547, + "step": 39416 + }, + { + "epoch": 2.7352022760391366, + "grad_norm": 0.917103111743927, + "learning_rate": 4.8013638752617405e-05, + "loss": 0.3108, + "step": 39417 + }, + { + "epoch": 2.7352716674762334, + "grad_norm": 1.0606294870376587, + "learning_rate": 4.800883913648416e-05, + "loss": 0.4237, + "step": 39418 + }, + { + "epoch": 2.7353410589133302, + "grad_norm": 1.226482629776001, + "learning_rate": 4.8004039684480985e-05, + "loss": 0.2685, + "step": 39419 + }, + { + "epoch": 2.7354104503504266, + "grad_norm": 0.9985478520393372, + "learning_rate": 4.799924039662306e-05, + "loss": 0.2598, + "step": 39420 + }, + { + "epoch": 2.7354798417875235, + "grad_norm": 0.8684371113777161, + "learning_rate": 4.799444127292545e-05, + "loss": 0.2432, + "step": 39421 + }, + { + "epoch": 2.73554923322462, + "grad_norm": 1.2400485277175903, + "learning_rate": 4.798964231340337e-05, + "loss": 0.4256, + "step": 39422 + }, + { + "epoch": 2.7356186246617167, + "grad_norm": 1.2338050603866577, + "learning_rate": 4.7984843518071954e-05, + "loss": 0.4712, + "step": 39423 + }, + { + "epoch": 2.7356880160988135, + "grad_norm": 0.8400164246559143, + "learning_rate": 4.798004488694634e-05, + "loss": 0.3483, + "step": 39424 + }, + { + "epoch": 2.7357574075359103, + "grad_norm": 1.255450963973999, + "learning_rate": 4.7975246420041694e-05, + "loss": 0.4897, + "step": 39425 + }, + { + "epoch": 2.7358267989730067, + "grad_norm": 1.248938798904419, + "learning_rate": 4.797044811737317e-05, + "loss": 0.3344, + "step": 39426 + }, + { + "epoch": 2.735896190410103, + "grad_norm": 2.105524778366089, + "learning_rate": 4.796564997895593e-05, + "loss": 0.5349, + "step": 39427 + }, + { + "epoch": 2.7359655818472, + "grad_norm": 1.6393351554870605, + "learning_rate": 4.796085200480507e-05, + "loss": 0.6133, + "step": 39428 + }, + { + "epoch": 2.736034973284297, + "grad_norm": 1.3863179683685303, + "learning_rate": 4.795605419493575e-05, + "loss": 0.3309, + "step": 39429 + }, + { + "epoch": 2.7361043647213936, + "grad_norm": 0.8673497438430786, + "learning_rate": 4.7951256549363136e-05, + "loss": 0.354, + "step": 39430 + }, + { + "epoch": 2.73617375615849, + "grad_norm": 1.374274730682373, + "learning_rate": 4.794645906810236e-05, + "loss": 0.4142, + "step": 39431 + }, + { + "epoch": 2.736243147595587, + "grad_norm": 0.8808214664459229, + "learning_rate": 4.794166175116858e-05, + "loss": 0.3043, + "step": 39432 + }, + { + "epoch": 2.7363125390326832, + "grad_norm": 0.934015691280365, + "learning_rate": 4.793686459857693e-05, + "loss": 0.38, + "step": 39433 + }, + { + "epoch": 2.73638193046978, + "grad_norm": 1.3658065795898438, + "learning_rate": 4.79320676103426e-05, + "loss": 0.4583, + "step": 39434 + }, + { + "epoch": 2.736451321906877, + "grad_norm": 1.5866256952285767, + "learning_rate": 4.792727078648065e-05, + "loss": 0.4435, + "step": 39435 + }, + { + "epoch": 2.7365207133439733, + "grad_norm": 1.31089448928833, + "learning_rate": 4.7922474127006264e-05, + "loss": 0.5147, + "step": 39436 + }, + { + "epoch": 2.73659010478107, + "grad_norm": 1.0315173864364624, + "learning_rate": 4.791767763193459e-05, + "loss": 0.3659, + "step": 39437 + }, + { + "epoch": 2.7366594962181665, + "grad_norm": 1.163891315460205, + "learning_rate": 4.791288130128076e-05, + "loss": 0.4406, + "step": 39438 + }, + { + "epoch": 2.7367288876552633, + "grad_norm": 0.8864922523498535, + "learning_rate": 4.790808513505992e-05, + "loss": 0.3104, + "step": 39439 + }, + { + "epoch": 2.73679827909236, + "grad_norm": 1.0358506441116333, + "learning_rate": 4.790328913328721e-05, + "loss": 0.2856, + "step": 39440 + }, + { + "epoch": 2.7368676705294566, + "grad_norm": 1.234442114830017, + "learning_rate": 4.789849329597782e-05, + "loss": 0.2748, + "step": 39441 + }, + { + "epoch": 2.7369370619665534, + "grad_norm": 0.9117891788482666, + "learning_rate": 4.7893697623146795e-05, + "loss": 0.2631, + "step": 39442 + }, + { + "epoch": 2.73700645340365, + "grad_norm": 1.6788727045059204, + "learning_rate": 4.788890211480932e-05, + "loss": 0.6079, + "step": 39443 + }, + { + "epoch": 2.7370758448407466, + "grad_norm": 0.903116762638092, + "learning_rate": 4.788410677098054e-05, + "loss": 0.3771, + "step": 39444 + }, + { + "epoch": 2.7371452362778435, + "grad_norm": 1.003158450126648, + "learning_rate": 4.7879311591675625e-05, + "loss": 0.3608, + "step": 39445 + }, + { + "epoch": 2.73721462771494, + "grad_norm": 1.0602436065673828, + "learning_rate": 4.7874516576909635e-05, + "loss": 0.3008, + "step": 39446 + }, + { + "epoch": 2.7372840191520367, + "grad_norm": 0.9755099415779114, + "learning_rate": 4.786972172669776e-05, + "loss": 0.3375, + "step": 39447 + }, + { + "epoch": 2.737353410589133, + "grad_norm": 2.330249786376953, + "learning_rate": 4.786492704105511e-05, + "loss": 0.4144, + "step": 39448 + }, + { + "epoch": 2.73742280202623, + "grad_norm": 0.855941891670227, + "learning_rate": 4.786013251999688e-05, + "loss": 0.3365, + "step": 39449 + }, + { + "epoch": 2.7374921934633267, + "grad_norm": 1.0237737894058228, + "learning_rate": 4.7855338163538135e-05, + "loss": 0.3627, + "step": 39450 + }, + { + "epoch": 2.7375615849004236, + "grad_norm": 1.2125762701034546, + "learning_rate": 4.7850543971694015e-05, + "loss": 0.3542, + "step": 39451 + }, + { + "epoch": 2.73763097633752, + "grad_norm": 1.5465770959854126, + "learning_rate": 4.78457499444797e-05, + "loss": 0.366, + "step": 39452 + }, + { + "epoch": 2.7377003677746163, + "grad_norm": 1.1652270555496216, + "learning_rate": 4.784095608191028e-05, + "loss": 0.3519, + "step": 39453 + }, + { + "epoch": 2.737769759211713, + "grad_norm": 1.4674773216247559, + "learning_rate": 4.783616238400097e-05, + "loss": 0.4962, + "step": 39454 + }, + { + "epoch": 2.73783915064881, + "grad_norm": 1.7416478395462036, + "learning_rate": 4.783136885076679e-05, + "loss": 0.3326, + "step": 39455 + }, + { + "epoch": 2.737908542085907, + "grad_norm": 1.1216119527816772, + "learning_rate": 4.782657548222294e-05, + "loss": 0.3479, + "step": 39456 + }, + { + "epoch": 2.7379779335230032, + "grad_norm": 1.2514605522155762, + "learning_rate": 4.782178227838452e-05, + "loss": 0.7662, + "step": 39457 + }, + { + "epoch": 2.7380473249601, + "grad_norm": 1.0305242538452148, + "learning_rate": 4.7816989239266685e-05, + "loss": 0.3383, + "step": 39458 + }, + { + "epoch": 2.7381167163971964, + "grad_norm": 0.9098873734474182, + "learning_rate": 4.781219636488457e-05, + "loss": 0.3339, + "step": 39459 + }, + { + "epoch": 2.7381861078342933, + "grad_norm": 1.1264230012893677, + "learning_rate": 4.780740365525329e-05, + "loss": 0.3191, + "step": 39460 + }, + { + "epoch": 2.73825549927139, + "grad_norm": 1.0060266256332397, + "learning_rate": 4.7802611110388016e-05, + "loss": 0.4189, + "step": 39461 + }, + { + "epoch": 2.7383248907084865, + "grad_norm": 0.888786256313324, + "learning_rate": 4.779781873030381e-05, + "loss": 0.2978, + "step": 39462 + }, + { + "epoch": 2.7383942821455833, + "grad_norm": 1.3744239807128906, + "learning_rate": 4.779302651501584e-05, + "loss": 0.4742, + "step": 39463 + }, + { + "epoch": 2.7384636735826797, + "grad_norm": 0.7666060328483582, + "learning_rate": 4.778823446453922e-05, + "loss": 0.2462, + "step": 39464 + }, + { + "epoch": 2.7385330650197766, + "grad_norm": 1.488538146018982, + "learning_rate": 4.7783442578889085e-05, + "loss": 0.593, + "step": 39465 + }, + { + "epoch": 2.7386024564568734, + "grad_norm": 1.0756661891937256, + "learning_rate": 4.7778650858080554e-05, + "loss": 0.3769, + "step": 39466 + }, + { + "epoch": 2.73867184789397, + "grad_norm": 1.4960856437683105, + "learning_rate": 4.777385930212877e-05, + "loss": 0.5925, + "step": 39467 + }, + { + "epoch": 2.7387412393310666, + "grad_norm": 0.9766416549682617, + "learning_rate": 4.7769067911048905e-05, + "loss": 0.3401, + "step": 39468 + }, + { + "epoch": 2.738810630768163, + "grad_norm": 1.2651742696762085, + "learning_rate": 4.776427668485598e-05, + "loss": 0.2968, + "step": 39469 + }, + { + "epoch": 2.73888002220526, + "grad_norm": 1.1565152406692505, + "learning_rate": 4.775948562356518e-05, + "loss": 0.4884, + "step": 39470 + }, + { + "epoch": 2.7389494136423567, + "grad_norm": 0.8054254651069641, + "learning_rate": 4.775469472719162e-05, + "loss": 0.207, + "step": 39471 + }, + { + "epoch": 2.739018805079453, + "grad_norm": 1.4384229183197021, + "learning_rate": 4.774990399575042e-05, + "loss": 0.5297, + "step": 39472 + }, + { + "epoch": 2.73908819651655, + "grad_norm": 0.9556218385696411, + "learning_rate": 4.774511342925673e-05, + "loss": 0.2866, + "step": 39473 + }, + { + "epoch": 2.7391575879536463, + "grad_norm": 1.3490660190582275, + "learning_rate": 4.774032302772564e-05, + "loss": 0.5976, + "step": 39474 + }, + { + "epoch": 2.739226979390743, + "grad_norm": 1.3410155773162842, + "learning_rate": 4.773553279117232e-05, + "loss": 0.4364, + "step": 39475 + }, + { + "epoch": 2.73929637082784, + "grad_norm": 0.9863776564598083, + "learning_rate": 4.773074271961183e-05, + "loss": 0.3651, + "step": 39476 + }, + { + "epoch": 2.7393657622649368, + "grad_norm": 2.764658212661743, + "learning_rate": 4.772595281305933e-05, + "loss": 0.3417, + "step": 39477 + }, + { + "epoch": 2.739435153702033, + "grad_norm": 1.3715494871139526, + "learning_rate": 4.772116307152992e-05, + "loss": 0.4876, + "step": 39478 + }, + { + "epoch": 2.7395045451391296, + "grad_norm": 1.0554436445236206, + "learning_rate": 4.771637349503874e-05, + "loss": 0.4826, + "step": 39479 + }, + { + "epoch": 2.7395739365762264, + "grad_norm": 0.9742176532745361, + "learning_rate": 4.771158408360093e-05, + "loss": 0.2829, + "step": 39480 + }, + { + "epoch": 2.739643328013323, + "grad_norm": 1.148314118385315, + "learning_rate": 4.770679483723155e-05, + "loss": 0.4625, + "step": 39481 + }, + { + "epoch": 2.73971271945042, + "grad_norm": 1.2657068967819214, + "learning_rate": 4.770200575594575e-05, + "loss": 0.4334, + "step": 39482 + }, + { + "epoch": 2.7397821108875164, + "grad_norm": 1.5138769149780273, + "learning_rate": 4.7697216839758654e-05, + "loss": 0.4041, + "step": 39483 + }, + { + "epoch": 2.7398515023246133, + "grad_norm": 1.0012017488479614, + "learning_rate": 4.76924280886854e-05, + "loss": 0.438, + "step": 39484 + }, + { + "epoch": 2.7399208937617097, + "grad_norm": 1.4133132696151733, + "learning_rate": 4.7687639502741065e-05, + "loss": 0.4431, + "step": 39485 + }, + { + "epoch": 2.7399902851988065, + "grad_norm": 1.1857696771621704, + "learning_rate": 4.768285108194077e-05, + "loss": 0.3931, + "step": 39486 + }, + { + "epoch": 2.7400596766359033, + "grad_norm": 1.0968607664108276, + "learning_rate": 4.767806282629965e-05, + "loss": 0.3112, + "step": 39487 + }, + { + "epoch": 2.7401290680729997, + "grad_norm": 1.6148805618286133, + "learning_rate": 4.767327473583281e-05, + "loss": 0.5377, + "step": 39488 + }, + { + "epoch": 2.7401984595100966, + "grad_norm": 1.1541649103164673, + "learning_rate": 4.7668486810555414e-05, + "loss": 0.5141, + "step": 39489 + }, + { + "epoch": 2.740267850947193, + "grad_norm": 1.1336623430252075, + "learning_rate": 4.766369905048249e-05, + "loss": 0.3524, + "step": 39490 + }, + { + "epoch": 2.7403372423842898, + "grad_norm": 1.0890629291534424, + "learning_rate": 4.7658911455629196e-05, + "loss": 0.3964, + "step": 39491 + }, + { + "epoch": 2.7404066338213866, + "grad_norm": 1.3036574125289917, + "learning_rate": 4.765412402601065e-05, + "loss": 0.3841, + "step": 39492 + }, + { + "epoch": 2.740476025258483, + "grad_norm": 1.3659422397613525, + "learning_rate": 4.764933676164196e-05, + "loss": 0.5368, + "step": 39493 + }, + { + "epoch": 2.74054541669558, + "grad_norm": 1.0511876344680786, + "learning_rate": 4.764454966253824e-05, + "loss": 0.2532, + "step": 39494 + }, + { + "epoch": 2.740614808132676, + "grad_norm": 1.2232316732406616, + "learning_rate": 4.76397627287146e-05, + "loss": 0.5021, + "step": 39495 + }, + { + "epoch": 2.740684199569773, + "grad_norm": 1.1792359352111816, + "learning_rate": 4.76349759601862e-05, + "loss": 0.2571, + "step": 39496 + }, + { + "epoch": 2.74075359100687, + "grad_norm": 1.7052162885665894, + "learning_rate": 4.7630189356968056e-05, + "loss": 0.5276, + "step": 39497 + }, + { + "epoch": 2.7408229824439663, + "grad_norm": 0.8010218739509583, + "learning_rate": 4.7625402919075334e-05, + "loss": 0.2959, + "step": 39498 + }, + { + "epoch": 2.740892373881063, + "grad_norm": 1.4128532409667969, + "learning_rate": 4.7620616646523145e-05, + "loss": 0.4294, + "step": 39499 + }, + { + "epoch": 2.7409617653181595, + "grad_norm": 1.068383812904358, + "learning_rate": 4.761583053932658e-05, + "loss": 0.413, + "step": 39500 + }, + { + "epoch": 2.7410311567552563, + "grad_norm": 1.0159149169921875, + "learning_rate": 4.7611044597500764e-05, + "loss": 0.3009, + "step": 39501 + }, + { + "epoch": 2.741100548192353, + "grad_norm": 0.9565480351448059, + "learning_rate": 4.7606258821060813e-05, + "loss": 0.3624, + "step": 39502 + }, + { + "epoch": 2.74116993962945, + "grad_norm": 1.1981980800628662, + "learning_rate": 4.7601473210021854e-05, + "loss": 0.5239, + "step": 39503 + }, + { + "epoch": 2.7412393310665464, + "grad_norm": 1.191756248474121, + "learning_rate": 4.7596687764398916e-05, + "loss": 0.3464, + "step": 39504 + }, + { + "epoch": 2.741308722503643, + "grad_norm": 0.9991710782051086, + "learning_rate": 4.759190248420716e-05, + "loss": 0.2758, + "step": 39505 + }, + { + "epoch": 2.7413781139407396, + "grad_norm": 1.3362793922424316, + "learning_rate": 4.75871173694617e-05, + "loss": 0.4396, + "step": 39506 + }, + { + "epoch": 2.7414475053778364, + "grad_norm": 0.8902217149734497, + "learning_rate": 4.758233242017762e-05, + "loss": 0.239, + "step": 39507 + }, + { + "epoch": 2.7415168968149333, + "grad_norm": 2.0281240940093994, + "learning_rate": 4.757754763637004e-05, + "loss": 0.548, + "step": 39508 + }, + { + "epoch": 2.7415862882520297, + "grad_norm": 0.8068147301673889, + "learning_rate": 4.757276301805406e-05, + "loss": 0.2715, + "step": 39509 + }, + { + "epoch": 2.7416556796891265, + "grad_norm": 1.5572552680969238, + "learning_rate": 4.756797856524481e-05, + "loss": 0.4525, + "step": 39510 + }, + { + "epoch": 2.741725071126223, + "grad_norm": 0.8781014084815979, + "learning_rate": 4.756319427795735e-05, + "loss": 0.3933, + "step": 39511 + }, + { + "epoch": 2.7417944625633197, + "grad_norm": 1.064756155014038, + "learning_rate": 4.7558410156206776e-05, + "loss": 0.4889, + "step": 39512 + }, + { + "epoch": 2.7418638540004165, + "grad_norm": 0.9664748311042786, + "learning_rate": 4.755362620000824e-05, + "loss": 0.3074, + "step": 39513 + }, + { + "epoch": 2.741933245437513, + "grad_norm": 1.9246528148651123, + "learning_rate": 4.754884240937681e-05, + "loss": 0.6186, + "step": 39514 + }, + { + "epoch": 2.7420026368746098, + "grad_norm": 1.1937180757522583, + "learning_rate": 4.7544058784327636e-05, + "loss": 0.3377, + "step": 39515 + }, + { + "epoch": 2.742072028311706, + "grad_norm": 0.8101937770843506, + "learning_rate": 4.753927532487575e-05, + "loss": 0.246, + "step": 39516 + }, + { + "epoch": 2.742141419748803, + "grad_norm": 1.2078083753585815, + "learning_rate": 4.7534492031036285e-05, + "loss": 0.5596, + "step": 39517 + }, + { + "epoch": 2.7422108111859, + "grad_norm": 1.206881046295166, + "learning_rate": 4.752970890282438e-05, + "loss": 0.2797, + "step": 39518 + }, + { + "epoch": 2.742280202622996, + "grad_norm": 1.2618976831436157, + "learning_rate": 4.752492594025506e-05, + "loss": 0.3671, + "step": 39519 + }, + { + "epoch": 2.742349594060093, + "grad_norm": 1.0262097120285034, + "learning_rate": 4.752014314334346e-05, + "loss": 0.4529, + "step": 39520 + }, + { + "epoch": 2.7424189854971894, + "grad_norm": 0.8553709983825684, + "learning_rate": 4.751536051210467e-05, + "loss": 0.212, + "step": 39521 + }, + { + "epoch": 2.7424883769342863, + "grad_norm": 1.519588828086853, + "learning_rate": 4.75105780465538e-05, + "loss": 0.2978, + "step": 39522 + }, + { + "epoch": 2.742557768371383, + "grad_norm": 1.0768014192581177, + "learning_rate": 4.750579574670595e-05, + "loss": 0.3231, + "step": 39523 + }, + { + "epoch": 2.7426271598084795, + "grad_norm": 1.3035355806350708, + "learning_rate": 4.7501013612576254e-05, + "loss": 0.3447, + "step": 39524 + }, + { + "epoch": 2.7426965512455763, + "grad_norm": 1.3574351072311401, + "learning_rate": 4.749623164417972e-05, + "loss": 0.2671, + "step": 39525 + }, + { + "epoch": 2.7427659426826727, + "grad_norm": 1.5961352586746216, + "learning_rate": 4.74914498415315e-05, + "loss": 0.4562, + "step": 39526 + }, + { + "epoch": 2.7428353341197695, + "grad_norm": 0.8704122304916382, + "learning_rate": 4.7486668204646677e-05, + "loss": 0.2585, + "step": 39527 + }, + { + "epoch": 2.7429047255568664, + "grad_norm": 1.1768755912780762, + "learning_rate": 4.748188673354035e-05, + "loss": 0.5243, + "step": 39528 + }, + { + "epoch": 2.742974116993963, + "grad_norm": 1.245286226272583, + "learning_rate": 4.747710542822761e-05, + "loss": 0.3805, + "step": 39529 + }, + { + "epoch": 2.7430435084310596, + "grad_norm": 0.6962853670120239, + "learning_rate": 4.747232428872359e-05, + "loss": 0.3927, + "step": 39530 + }, + { + "epoch": 2.7431128998681564, + "grad_norm": 1.0837125778198242, + "learning_rate": 4.746754331504332e-05, + "loss": 0.4861, + "step": 39531 + }, + { + "epoch": 2.743182291305253, + "grad_norm": 1.1414616107940674, + "learning_rate": 4.746276250720192e-05, + "loss": 0.5133, + "step": 39532 + }, + { + "epoch": 2.7432516827423497, + "grad_norm": 1.125145673751831, + "learning_rate": 4.7457981865214476e-05, + "loss": 0.4424, + "step": 39533 + }, + { + "epoch": 2.7433210741794465, + "grad_norm": 0.7351963520050049, + "learning_rate": 4.745320138909609e-05, + "loss": 0.253, + "step": 39534 + }, + { + "epoch": 2.743390465616543, + "grad_norm": 1.0426177978515625, + "learning_rate": 4.7448421078861846e-05, + "loss": 0.4344, + "step": 39535 + }, + { + "epoch": 2.7434598570536397, + "grad_norm": 1.1458392143249512, + "learning_rate": 4.7443640934526847e-05, + "loss": 0.3049, + "step": 39536 + }, + { + "epoch": 2.743529248490736, + "grad_norm": 0.9593973755836487, + "learning_rate": 4.7438860956106215e-05, + "loss": 0.339, + "step": 39537 + }, + { + "epoch": 2.743598639927833, + "grad_norm": 1.093740701675415, + "learning_rate": 4.743408114361495e-05, + "loss": 0.3633, + "step": 39538 + }, + { + "epoch": 2.7436680313649298, + "grad_norm": 1.086174488067627, + "learning_rate": 4.74293014970682e-05, + "loss": 0.3395, + "step": 39539 + }, + { + "epoch": 2.743737422802026, + "grad_norm": 1.0912082195281982, + "learning_rate": 4.742452201648105e-05, + "loss": 0.4868, + "step": 39540 + }, + { + "epoch": 2.743806814239123, + "grad_norm": 1.2520562410354614, + "learning_rate": 4.741974270186859e-05, + "loss": 0.4398, + "step": 39541 + }, + { + "epoch": 2.7438762056762194, + "grad_norm": 1.1219850778579712, + "learning_rate": 4.741496355324588e-05, + "loss": 0.3819, + "step": 39542 + }, + { + "epoch": 2.743945597113316, + "grad_norm": 1.0499211549758911, + "learning_rate": 4.741018457062805e-05, + "loss": 0.3868, + "step": 39543 + }, + { + "epoch": 2.744014988550413, + "grad_norm": 1.0871750116348267, + "learning_rate": 4.74054057540302e-05, + "loss": 0.4341, + "step": 39544 + }, + { + "epoch": 2.7440843799875094, + "grad_norm": 1.1098390817642212, + "learning_rate": 4.740062710346733e-05, + "loss": 0.4725, + "step": 39545 + }, + { + "epoch": 2.7441537714246063, + "grad_norm": 0.9663684964179993, + "learning_rate": 4.7395848618954584e-05, + "loss": 0.3455, + "step": 39546 + }, + { + "epoch": 2.7442231628617026, + "grad_norm": 0.9182147979736328, + "learning_rate": 4.739107030050705e-05, + "loss": 0.3495, + "step": 39547 + }, + { + "epoch": 2.7442925542987995, + "grad_norm": 0.9471951723098755, + "learning_rate": 4.7386292148139797e-05, + "loss": 0.3738, + "step": 39548 + }, + { + "epoch": 2.7443619457358963, + "grad_norm": 0.9737802743911743, + "learning_rate": 4.7381514161867913e-05, + "loss": 0.2427, + "step": 39549 + }, + { + "epoch": 2.7444313371729927, + "grad_norm": 1.261540174484253, + "learning_rate": 4.737673634170653e-05, + "loss": 0.366, + "step": 39550 + }, + { + "epoch": 2.7445007286100895, + "grad_norm": 1.032220482826233, + "learning_rate": 4.7371958687670635e-05, + "loss": 0.2434, + "step": 39551 + }, + { + "epoch": 2.744570120047186, + "grad_norm": 1.014064908027649, + "learning_rate": 4.736718119977537e-05, + "loss": 0.4163, + "step": 39552 + }, + { + "epoch": 2.7446395114842828, + "grad_norm": 0.7842938303947449, + "learning_rate": 4.7362403878035845e-05, + "loss": 0.2508, + "step": 39553 + }, + { + "epoch": 2.7447089029213796, + "grad_norm": 1.016076922416687, + "learning_rate": 4.735762672246706e-05, + "loss": 0.3621, + "step": 39554 + }, + { + "epoch": 2.7447782943584764, + "grad_norm": 0.7069135904312134, + "learning_rate": 4.7352849733084145e-05, + "loss": 0.2911, + "step": 39555 + }, + { + "epoch": 2.744847685795573, + "grad_norm": 1.2245941162109375, + "learning_rate": 4.734807290990218e-05, + "loss": 0.5103, + "step": 39556 + }, + { + "epoch": 2.7449170772326696, + "grad_norm": 1.0290920734405518, + "learning_rate": 4.734329625293622e-05, + "loss": 0.4249, + "step": 39557 + }, + { + "epoch": 2.744986468669766, + "grad_norm": 1.0315442085266113, + "learning_rate": 4.733851976220142e-05, + "loss": 0.5361, + "step": 39558 + }, + { + "epoch": 2.745055860106863, + "grad_norm": 1.4381192922592163, + "learning_rate": 4.733374343771276e-05, + "loss": 0.348, + "step": 39559 + }, + { + "epoch": 2.7451252515439597, + "grad_norm": 1.9807084798812866, + "learning_rate": 4.732896727948536e-05, + "loss": 0.8006, + "step": 39560 + }, + { + "epoch": 2.745194642981056, + "grad_norm": 1.2891688346862793, + "learning_rate": 4.7324191287534304e-05, + "loss": 0.3963, + "step": 39561 + }, + { + "epoch": 2.745264034418153, + "grad_norm": 1.6616592407226562, + "learning_rate": 4.731941546187466e-05, + "loss": 0.3503, + "step": 39562 + }, + { + "epoch": 2.7453334258552493, + "grad_norm": 1.4957051277160645, + "learning_rate": 4.731463980252152e-05, + "loss": 0.4391, + "step": 39563 + }, + { + "epoch": 2.745402817292346, + "grad_norm": 1.32536780834198, + "learning_rate": 4.7309864309489936e-05, + "loss": 0.3977, + "step": 39564 + }, + { + "epoch": 2.745472208729443, + "grad_norm": 0.7570125460624695, + "learning_rate": 4.7305088982795045e-05, + "loss": 0.267, + "step": 39565 + }, + { + "epoch": 2.7455416001665394, + "grad_norm": 1.8206853866577148, + "learning_rate": 4.730031382245183e-05, + "loss": 0.4213, + "step": 39566 + }, + { + "epoch": 2.745610991603636, + "grad_norm": 1.7592228651046753, + "learning_rate": 4.72955388284754e-05, + "loss": 0.417, + "step": 39567 + }, + { + "epoch": 2.7456803830407326, + "grad_norm": 1.3537964820861816, + "learning_rate": 4.729076400088086e-05, + "loss": 0.4895, + "step": 39568 + }, + { + "epoch": 2.7457497744778294, + "grad_norm": 1.396307349205017, + "learning_rate": 4.7285989339683255e-05, + "loss": 0.4245, + "step": 39569 + }, + { + "epoch": 2.7458191659149263, + "grad_norm": 1.3370879888534546, + "learning_rate": 4.728121484489766e-05, + "loss": 0.5689, + "step": 39570 + }, + { + "epoch": 2.7458885573520226, + "grad_norm": 1.8087191581726074, + "learning_rate": 4.727644051653915e-05, + "loss": 0.6716, + "step": 39571 + }, + { + "epoch": 2.7459579487891195, + "grad_norm": 1.0936802625656128, + "learning_rate": 4.7271666354622856e-05, + "loss": 0.3575, + "step": 39572 + }, + { + "epoch": 2.746027340226216, + "grad_norm": 1.1495548486709595, + "learning_rate": 4.726689235916375e-05, + "loss": 0.271, + "step": 39573 + }, + { + "epoch": 2.7460967316633127, + "grad_norm": 1.2434396743774414, + "learning_rate": 4.7262118530176934e-05, + "loss": 0.4025, + "step": 39574 + }, + { + "epoch": 2.7461661231004095, + "grad_norm": 1.1791653633117676, + "learning_rate": 4.72573448676775e-05, + "loss": 0.3465, + "step": 39575 + }, + { + "epoch": 2.746235514537506, + "grad_norm": 1.302778959274292, + "learning_rate": 4.7252571371680506e-05, + "loss": 0.3497, + "step": 39576 + }, + { + "epoch": 2.7463049059746028, + "grad_norm": 1.0570091009140015, + "learning_rate": 4.724779804220103e-05, + "loss": 0.3792, + "step": 39577 + }, + { + "epoch": 2.746374297411699, + "grad_norm": 1.264384388923645, + "learning_rate": 4.724302487925413e-05, + "loss": 0.3241, + "step": 39578 + }, + { + "epoch": 2.746443688848796, + "grad_norm": 1.1208187341690063, + "learning_rate": 4.723825188285492e-05, + "loss": 0.5103, + "step": 39579 + }, + { + "epoch": 2.746513080285893, + "grad_norm": 1.1024774312973022, + "learning_rate": 4.723347905301839e-05, + "loss": 0.3286, + "step": 39580 + }, + { + "epoch": 2.7465824717229896, + "grad_norm": 0.8513790369033813, + "learning_rate": 4.7228706389759636e-05, + "loss": 0.3163, + "step": 39581 + }, + { + "epoch": 2.746651863160086, + "grad_norm": 1.839951992034912, + "learning_rate": 4.722393389309374e-05, + "loss": 0.4373, + "step": 39582 + }, + { + "epoch": 2.746721254597183, + "grad_norm": 1.4044954776763916, + "learning_rate": 4.721916156303576e-05, + "loss": 0.311, + "step": 39583 + }, + { + "epoch": 2.7467906460342792, + "grad_norm": 0.9929968118667603, + "learning_rate": 4.721438939960081e-05, + "loss": 0.4752, + "step": 39584 + }, + { + "epoch": 2.746860037471376, + "grad_norm": 1.1332980394363403, + "learning_rate": 4.7209617402803864e-05, + "loss": 0.3988, + "step": 39585 + }, + { + "epoch": 2.746929428908473, + "grad_norm": 1.0245578289031982, + "learning_rate": 4.720484557266003e-05, + "loss": 0.4087, + "step": 39586 + }, + { + "epoch": 2.7469988203455693, + "grad_norm": 1.3854769468307495, + "learning_rate": 4.720007390918437e-05, + "loss": 0.3679, + "step": 39587 + }, + { + "epoch": 2.747068211782666, + "grad_norm": 1.8133025169372559, + "learning_rate": 4.7195302412391994e-05, + "loss": 0.5618, + "step": 39588 + }, + { + "epoch": 2.7471376032197625, + "grad_norm": 0.9656505584716797, + "learning_rate": 4.7190531082297883e-05, + "loss": 0.4695, + "step": 39589 + }, + { + "epoch": 2.7472069946568594, + "grad_norm": 1.0580037832260132, + "learning_rate": 4.7185759918917136e-05, + "loss": 0.4343, + "step": 39590 + }, + { + "epoch": 2.747276386093956, + "grad_norm": 1.1249752044677734, + "learning_rate": 4.718098892226482e-05, + "loss": 0.4204, + "step": 39591 + }, + { + "epoch": 2.7473457775310526, + "grad_norm": 1.5790119171142578, + "learning_rate": 4.7176218092355993e-05, + "loss": 0.5321, + "step": 39592 + }, + { + "epoch": 2.7474151689681494, + "grad_norm": 1.424373984336853, + "learning_rate": 4.717144742920575e-05, + "loss": 0.3963, + "step": 39593 + }, + { + "epoch": 2.747484560405246, + "grad_norm": 1.2102071046829224, + "learning_rate": 4.716667693282909e-05, + "loss": 0.3353, + "step": 39594 + }, + { + "epoch": 2.7475539518423426, + "grad_norm": 0.9305770397186279, + "learning_rate": 4.716190660324109e-05, + "loss": 0.4185, + "step": 39595 + }, + { + "epoch": 2.7476233432794395, + "grad_norm": 0.8644077181816101, + "learning_rate": 4.715713644045683e-05, + "loss": 0.3285, + "step": 39596 + }, + { + "epoch": 2.747692734716536, + "grad_norm": 1.1575084924697876, + "learning_rate": 4.715236644449136e-05, + "loss": 0.3366, + "step": 39597 + }, + { + "epoch": 2.7477621261536327, + "grad_norm": 1.281341314315796, + "learning_rate": 4.714759661535974e-05, + "loss": 0.5216, + "step": 39598 + }, + { + "epoch": 2.747831517590729, + "grad_norm": 1.3780547380447388, + "learning_rate": 4.714282695307701e-05, + "loss": 0.3073, + "step": 39599 + }, + { + "epoch": 2.747900909027826, + "grad_norm": 1.1947550773620605, + "learning_rate": 4.713805745765829e-05, + "loss": 0.2962, + "step": 39600 + }, + { + "epoch": 2.7479703004649227, + "grad_norm": 1.296309471130371, + "learning_rate": 4.713328812911855e-05, + "loss": 0.4057, + "step": 39601 + }, + { + "epoch": 2.748039691902019, + "grad_norm": 1.1534768342971802, + "learning_rate": 4.712851896747289e-05, + "loss": 0.3606, + "step": 39602 + }, + { + "epoch": 2.748109083339116, + "grad_norm": 1.9553531408309937, + "learning_rate": 4.7123749972736355e-05, + "loss": 0.2878, + "step": 39603 + }, + { + "epoch": 2.7481784747762124, + "grad_norm": 1.2941408157348633, + "learning_rate": 4.711898114492401e-05, + "loss": 0.422, + "step": 39604 + }, + { + "epoch": 2.748247866213309, + "grad_norm": 1.2734417915344238, + "learning_rate": 4.71142124840509e-05, + "loss": 0.5404, + "step": 39605 + }, + { + "epoch": 2.748317257650406, + "grad_norm": 0.7864357829093933, + "learning_rate": 4.710944399013213e-05, + "loss": 0.2001, + "step": 39606 + }, + { + "epoch": 2.748386649087503, + "grad_norm": 1.156002402305603, + "learning_rate": 4.710467566318267e-05, + "loss": 0.3881, + "step": 39607 + }, + { + "epoch": 2.7484560405245992, + "grad_norm": 1.1279969215393066, + "learning_rate": 4.709990750321761e-05, + "loss": 0.2488, + "step": 39608 + }, + { + "epoch": 2.748525431961696, + "grad_norm": 1.261887550354004, + "learning_rate": 4.7095139510252006e-05, + "loss": 0.4355, + "step": 39609 + }, + { + "epoch": 2.7485948233987925, + "grad_norm": 1.3968887329101562, + "learning_rate": 4.7090371684300916e-05, + "loss": 0.4409, + "step": 39610 + }, + { + "epoch": 2.7486642148358893, + "grad_norm": 1.054031252861023, + "learning_rate": 4.708560402537937e-05, + "loss": 0.3568, + "step": 39611 + }, + { + "epoch": 2.748733606272986, + "grad_norm": 1.0823255777359009, + "learning_rate": 4.7080836533502445e-05, + "loss": 0.3377, + "step": 39612 + }, + { + "epoch": 2.7488029977100825, + "grad_norm": 1.3064115047454834, + "learning_rate": 4.707606920868521e-05, + "loss": 0.4326, + "step": 39613 + }, + { + "epoch": 2.7488723891471794, + "grad_norm": 1.4110087156295776, + "learning_rate": 4.7071302050942645e-05, + "loss": 0.4212, + "step": 39614 + }, + { + "epoch": 2.7489417805842757, + "grad_norm": 1.8120640516281128, + "learning_rate": 4.706653506028985e-05, + "loss": 0.5206, + "step": 39615 + }, + { + "epoch": 2.7490111720213726, + "grad_norm": 1.0111348628997803, + "learning_rate": 4.7061768236741856e-05, + "loss": 0.2317, + "step": 39616 + }, + { + "epoch": 2.7490805634584694, + "grad_norm": 1.1457728147506714, + "learning_rate": 4.705700158031371e-05, + "loss": 0.2849, + "step": 39617 + }, + { + "epoch": 2.749149954895566, + "grad_norm": 1.3186662197113037, + "learning_rate": 4.705223509102047e-05, + "loss": 0.4457, + "step": 39618 + }, + { + "epoch": 2.7492193463326626, + "grad_norm": 1.4106731414794922, + "learning_rate": 4.704746876887722e-05, + "loss": 0.4474, + "step": 39619 + }, + { + "epoch": 2.749288737769759, + "grad_norm": 1.1059170961380005, + "learning_rate": 4.7042702613898934e-05, + "loss": 0.3999, + "step": 39620 + }, + { + "epoch": 2.749358129206856, + "grad_norm": 1.1900031566619873, + "learning_rate": 4.70379366261007e-05, + "loss": 0.317, + "step": 39621 + }, + { + "epoch": 2.7494275206439527, + "grad_norm": 1.1233587265014648, + "learning_rate": 4.703317080549754e-05, + "loss": 0.3808, + "step": 39622 + }, + { + "epoch": 2.749496912081049, + "grad_norm": 0.7636473178863525, + "learning_rate": 4.702840515210455e-05, + "loss": 0.334, + "step": 39623 + }, + { + "epoch": 2.749566303518146, + "grad_norm": 1.1587785482406616, + "learning_rate": 4.7023639665936715e-05, + "loss": 0.3273, + "step": 39624 + }, + { + "epoch": 2.7496356949552423, + "grad_norm": 0.9038227796554565, + "learning_rate": 4.701887434700909e-05, + "loss": 0.333, + "step": 39625 + }, + { + "epoch": 2.749705086392339, + "grad_norm": 1.1850489377975464, + "learning_rate": 4.701410919533674e-05, + "loss": 0.3813, + "step": 39626 + }, + { + "epoch": 2.749774477829436, + "grad_norm": 1.2517577409744263, + "learning_rate": 4.7009344210934735e-05, + "loss": 0.3616, + "step": 39627 + }, + { + "epoch": 2.7498438692665323, + "grad_norm": 1.010740041732788, + "learning_rate": 4.700457939381805e-05, + "loss": 0.3523, + "step": 39628 + }, + { + "epoch": 2.749913260703629, + "grad_norm": 1.377482533454895, + "learning_rate": 4.6999814744001745e-05, + "loss": 0.5039, + "step": 39629 + }, + { + "epoch": 2.7499826521407256, + "grad_norm": 1.3982969522476196, + "learning_rate": 4.6995050261500894e-05, + "loss": 0.382, + "step": 39630 + }, + { + "epoch": 2.7500520435778224, + "grad_norm": 0.8650837540626526, + "learning_rate": 4.6990285946330506e-05, + "loss": 0.4108, + "step": 39631 + }, + { + "epoch": 2.7501214350149192, + "grad_norm": 0.9776937961578369, + "learning_rate": 4.698552179850564e-05, + "loss": 0.3823, + "step": 39632 + }, + { + "epoch": 2.750190826452016, + "grad_norm": 1.080227255821228, + "learning_rate": 4.698075781804133e-05, + "loss": 0.4451, + "step": 39633 + }, + { + "epoch": 2.7502602178891125, + "grad_norm": 1.0405114889144897, + "learning_rate": 4.697599400495266e-05, + "loss": 0.3926, + "step": 39634 + }, + { + "epoch": 2.7503296093262093, + "grad_norm": 1.8730595111846924, + "learning_rate": 4.697123035925458e-05, + "loss": 0.3902, + "step": 39635 + }, + { + "epoch": 2.7503990007633057, + "grad_norm": 1.401352882385254, + "learning_rate": 4.696646688096218e-05, + "loss": 0.4185, + "step": 39636 + }, + { + "epoch": 2.7504683922004025, + "grad_norm": 1.2030584812164307, + "learning_rate": 4.696170357009047e-05, + "loss": 0.2429, + "step": 39637 + }, + { + "epoch": 2.7505377836374993, + "grad_norm": 1.0011475086212158, + "learning_rate": 4.695694042665454e-05, + "loss": 0.3613, + "step": 39638 + }, + { + "epoch": 2.7506071750745957, + "grad_norm": 1.246755838394165, + "learning_rate": 4.695217745066938e-05, + "loss": 0.4215, + "step": 39639 + }, + { + "epoch": 2.7506765665116926, + "grad_norm": 1.3253518342971802, + "learning_rate": 4.694741464215003e-05, + "loss": 0.4687, + "step": 39640 + }, + { + "epoch": 2.750745957948789, + "grad_norm": 1.1723496913909912, + "learning_rate": 4.694265200111159e-05, + "loss": 0.4295, + "step": 39641 + }, + { + "epoch": 2.750815349385886, + "grad_norm": 1.5064181089401245, + "learning_rate": 4.6937889527569e-05, + "loss": 0.4356, + "step": 39642 + }, + { + "epoch": 2.7508847408229826, + "grad_norm": 1.495388150215149, + "learning_rate": 4.6933127221537334e-05, + "loss": 0.5311, + "step": 39643 + }, + { + "epoch": 2.750954132260079, + "grad_norm": 2.7969560623168945, + "learning_rate": 4.692836508303163e-05, + "loss": 0.4983, + "step": 39644 + }, + { + "epoch": 2.751023523697176, + "grad_norm": 1.1222003698349, + "learning_rate": 4.692360311206692e-05, + "loss": 0.356, + "step": 39645 + }, + { + "epoch": 2.7510929151342722, + "grad_norm": 1.2612874507904053, + "learning_rate": 4.691884130865824e-05, + "loss": 0.4452, + "step": 39646 + }, + { + "epoch": 2.751162306571369, + "grad_norm": 1.1165522336959839, + "learning_rate": 4.691407967282061e-05, + "loss": 0.3125, + "step": 39647 + }, + { + "epoch": 2.751231698008466, + "grad_norm": 1.096492052078247, + "learning_rate": 4.690931820456912e-05, + "loss": 0.3884, + "step": 39648 + }, + { + "epoch": 2.7513010894455623, + "grad_norm": 0.9024727940559387, + "learning_rate": 4.690455690391871e-05, + "loss": 0.3013, + "step": 39649 + }, + { + "epoch": 2.751370480882659, + "grad_norm": 1.8005181550979614, + "learning_rate": 4.689979577088445e-05, + "loss": 0.4571, + "step": 39650 + }, + { + "epoch": 2.7514398723197555, + "grad_norm": 1.2470201253890991, + "learning_rate": 4.689503480548138e-05, + "loss": 0.5005, + "step": 39651 + }, + { + "epoch": 2.7515092637568523, + "grad_norm": 1.2091225385665894, + "learning_rate": 4.689027400772452e-05, + "loss": 0.4574, + "step": 39652 + }, + { + "epoch": 2.751578655193949, + "grad_norm": 0.9335790276527405, + "learning_rate": 4.688551337762891e-05, + "loss": 0.4123, + "step": 39653 + }, + { + "epoch": 2.7516480466310456, + "grad_norm": 0.8006832003593445, + "learning_rate": 4.68807529152096e-05, + "loss": 0.2056, + "step": 39654 + }, + { + "epoch": 2.7517174380681424, + "grad_norm": 1.615311622619629, + "learning_rate": 4.687599262048156e-05, + "loss": 0.4065, + "step": 39655 + }, + { + "epoch": 2.751786829505239, + "grad_norm": 1.0970572233200073, + "learning_rate": 4.687123249345984e-05, + "loss": 0.3002, + "step": 39656 + }, + { + "epoch": 2.7518562209423356, + "grad_norm": 1.032545804977417, + "learning_rate": 4.686647253415948e-05, + "loss": 0.3074, + "step": 39657 + }, + { + "epoch": 2.7519256123794325, + "grad_norm": 1.23636794090271, + "learning_rate": 4.6861712742595545e-05, + "loss": 0.3721, + "step": 39658 + }, + { + "epoch": 2.7519950038165293, + "grad_norm": 1.9973727464675903, + "learning_rate": 4.685695311878297e-05, + "loss": 0.4283, + "step": 39659 + }, + { + "epoch": 2.7520643952536257, + "grad_norm": 1.5515977144241333, + "learning_rate": 4.685219366273683e-05, + "loss": 0.4607, + "step": 39660 + }, + { + "epoch": 2.7521337866907225, + "grad_norm": 1.6406117677688599, + "learning_rate": 4.6847434374472156e-05, + "loss": 0.4611, + "step": 39661 + }, + { + "epoch": 2.752203178127819, + "grad_norm": 1.4782136678695679, + "learning_rate": 4.684267525400399e-05, + "loss": 0.4867, + "step": 39662 + }, + { + "epoch": 2.7522725695649157, + "grad_norm": 1.6819840669631958, + "learning_rate": 4.68379163013473e-05, + "loss": 0.539, + "step": 39663 + }, + { + "epoch": 2.7523419610020126, + "grad_norm": 1.3613845109939575, + "learning_rate": 4.683315751651715e-05, + "loss": 0.3479, + "step": 39664 + }, + { + "epoch": 2.752411352439109, + "grad_norm": 1.061450719833374, + "learning_rate": 4.6828398899528534e-05, + "loss": 0.3922, + "step": 39665 + }, + { + "epoch": 2.752480743876206, + "grad_norm": 1.1620995998382568, + "learning_rate": 4.68236404503965e-05, + "loss": 0.3728, + "step": 39666 + }, + { + "epoch": 2.752550135313302, + "grad_norm": 1.0502885580062866, + "learning_rate": 4.681888216913607e-05, + "loss": 0.4163, + "step": 39667 + }, + { + "epoch": 2.752619526750399, + "grad_norm": 1.049812912940979, + "learning_rate": 4.6814124055762254e-05, + "loss": 0.5388, + "step": 39668 + }, + { + "epoch": 2.752688918187496, + "grad_norm": 0.9048234820365906, + "learning_rate": 4.680936611029012e-05, + "loss": 0.2097, + "step": 39669 + }, + { + "epoch": 2.7527583096245922, + "grad_norm": 1.233319878578186, + "learning_rate": 4.6804608332734604e-05, + "loss": 0.3896, + "step": 39670 + }, + { + "epoch": 2.752827701061689, + "grad_norm": 1.1054160594940186, + "learning_rate": 4.679985072311077e-05, + "loss": 0.3694, + "step": 39671 + }, + { + "epoch": 2.7528970924987854, + "grad_norm": 1.288210391998291, + "learning_rate": 4.679509328143363e-05, + "loss": 0.3761, + "step": 39672 + }, + { + "epoch": 2.7529664839358823, + "grad_norm": 1.3497371673583984, + "learning_rate": 4.6790336007718215e-05, + "loss": 0.5677, + "step": 39673 + }, + { + "epoch": 2.753035875372979, + "grad_norm": 1.0994092226028442, + "learning_rate": 4.6785578901979535e-05, + "loss": 0.4748, + "step": 39674 + }, + { + "epoch": 2.7531052668100755, + "grad_norm": 1.084261417388916, + "learning_rate": 4.67808219642326e-05, + "loss": 0.4314, + "step": 39675 + }, + { + "epoch": 2.7531746582471723, + "grad_norm": 1.137755274772644, + "learning_rate": 4.677606519449248e-05, + "loss": 0.3575, + "step": 39676 + }, + { + "epoch": 2.7532440496842687, + "grad_norm": 1.0513055324554443, + "learning_rate": 4.677130859277411e-05, + "loss": 0.4274, + "step": 39677 + }, + { + "epoch": 2.7533134411213656, + "grad_norm": 0.9665644764900208, + "learning_rate": 4.6766552159092536e-05, + "loss": 0.4637, + "step": 39678 + }, + { + "epoch": 2.7533828325584624, + "grad_norm": 0.7884703278541565, + "learning_rate": 4.676179589346279e-05, + "loss": 0.2685, + "step": 39679 + }, + { + "epoch": 2.7534522239955592, + "grad_norm": 1.6883913278579712, + "learning_rate": 4.675703979589987e-05, + "loss": 0.5906, + "step": 39680 + }, + { + "epoch": 2.7535216154326556, + "grad_norm": 1.2169389724731445, + "learning_rate": 4.67522838664188e-05, + "loss": 0.3424, + "step": 39681 + }, + { + "epoch": 2.753591006869752, + "grad_norm": 1.098596215248108, + "learning_rate": 4.67475281050346e-05, + "loss": 0.1807, + "step": 39682 + }, + { + "epoch": 2.753660398306849, + "grad_norm": 1.1517858505249023, + "learning_rate": 4.67427725117623e-05, + "loss": 0.3955, + "step": 39683 + }, + { + "epoch": 2.7537297897439457, + "grad_norm": 1.1851520538330078, + "learning_rate": 4.6738017086616856e-05, + "loss": 0.378, + "step": 39684 + }, + { + "epoch": 2.7537991811810425, + "grad_norm": 1.3373116254806519, + "learning_rate": 4.673326182961332e-05, + "loss": 0.3439, + "step": 39685 + }, + { + "epoch": 2.753868572618139, + "grad_norm": 1.3513058423995972, + "learning_rate": 4.672850674076669e-05, + "loss": 0.4933, + "step": 39686 + }, + { + "epoch": 2.7539379640552357, + "grad_norm": 1.5767433643341064, + "learning_rate": 4.6723751820091974e-05, + "loss": 0.5776, + "step": 39687 + }, + { + "epoch": 2.754007355492332, + "grad_norm": 0.9243062138557434, + "learning_rate": 4.671899706760421e-05, + "loss": 0.3488, + "step": 39688 + }, + { + "epoch": 2.754076746929429, + "grad_norm": 1.509400725364685, + "learning_rate": 4.671424248331842e-05, + "loss": 0.5449, + "step": 39689 + }, + { + "epoch": 2.7541461383665258, + "grad_norm": 1.0336246490478516, + "learning_rate": 4.670948806724954e-05, + "loss": 0.3893, + "step": 39690 + }, + { + "epoch": 2.754215529803622, + "grad_norm": 1.0787595510482788, + "learning_rate": 4.670473381941263e-05, + "loss": 0.3674, + "step": 39691 + }, + { + "epoch": 2.754284921240719, + "grad_norm": 1.1418262720108032, + "learning_rate": 4.66999797398227e-05, + "loss": 0.4785, + "step": 39692 + }, + { + "epoch": 2.7543543126778154, + "grad_norm": 1.1505488157272339, + "learning_rate": 4.669522582849477e-05, + "loss": 0.3365, + "step": 39693 + }, + { + "epoch": 2.754423704114912, + "grad_norm": 1.0654324293136597, + "learning_rate": 4.6690472085443804e-05, + "loss": 0.3543, + "step": 39694 + }, + { + "epoch": 2.754493095552009, + "grad_norm": 1.0505613088607788, + "learning_rate": 4.6685718510684826e-05, + "loss": 0.3175, + "step": 39695 + }, + { + "epoch": 2.7545624869891054, + "grad_norm": 1.053991436958313, + "learning_rate": 4.6680965104232886e-05, + "loss": 0.4089, + "step": 39696 + }, + { + "epoch": 2.7546318784262023, + "grad_norm": 1.4507262706756592, + "learning_rate": 4.6676211866102914e-05, + "loss": 0.296, + "step": 39697 + }, + { + "epoch": 2.7547012698632987, + "grad_norm": 1.1876561641693115, + "learning_rate": 4.6671458796309965e-05, + "loss": 0.3886, + "step": 39698 + }, + { + "epoch": 2.7547706613003955, + "grad_norm": 1.237327218055725, + "learning_rate": 4.666670589486902e-05, + "loss": 0.3539, + "step": 39699 + }, + { + "epoch": 2.7548400527374923, + "grad_norm": 1.0950348377227783, + "learning_rate": 4.666195316179511e-05, + "loss": 0.4568, + "step": 39700 + }, + { + "epoch": 2.7549094441745887, + "grad_norm": 2.1190972328186035, + "learning_rate": 4.665720059710321e-05, + "loss": 0.4465, + "step": 39701 + }, + { + "epoch": 2.7549788356116856, + "grad_norm": 1.2287030220031738, + "learning_rate": 4.665244820080835e-05, + "loss": 0.592, + "step": 39702 + }, + { + "epoch": 2.755048227048782, + "grad_norm": 1.2344032526016235, + "learning_rate": 4.664769597292557e-05, + "loss": 0.395, + "step": 39703 + }, + { + "epoch": 2.7551176184858788, + "grad_norm": 0.8021227717399597, + "learning_rate": 4.6642943913469764e-05, + "loss": 0.3118, + "step": 39704 + }, + { + "epoch": 2.7551870099229756, + "grad_norm": 0.7686235308647156, + "learning_rate": 4.6638192022456004e-05, + "loss": 0.3558, + "step": 39705 + }, + { + "epoch": 2.7552564013600724, + "grad_norm": 1.623076319694519, + "learning_rate": 4.6633440299899275e-05, + "loss": 0.2727, + "step": 39706 + }, + { + "epoch": 2.755325792797169, + "grad_norm": 1.050960659980774, + "learning_rate": 4.662868874581459e-05, + "loss": 0.4592, + "step": 39707 + }, + { + "epoch": 2.755395184234265, + "grad_norm": 1.364732027053833, + "learning_rate": 4.662393736021694e-05, + "loss": 0.369, + "step": 39708 + }, + { + "epoch": 2.755464575671362, + "grad_norm": 1.3428014516830444, + "learning_rate": 4.661918614312133e-05, + "loss": 0.4284, + "step": 39709 + }, + { + "epoch": 2.755533967108459, + "grad_norm": 1.0760061740875244, + "learning_rate": 4.661443509454278e-05, + "loss": 0.3488, + "step": 39710 + }, + { + "epoch": 2.7556033585455557, + "grad_norm": 1.0610827207565308, + "learning_rate": 4.660968421449624e-05, + "loss": 0.2748, + "step": 39711 + }, + { + "epoch": 2.755672749982652, + "grad_norm": 1.1825940608978271, + "learning_rate": 4.660493350299674e-05, + "loss": 0.3131, + "step": 39712 + }, + { + "epoch": 2.755742141419749, + "grad_norm": 1.0094561576843262, + "learning_rate": 4.660018296005926e-05, + "loss": 0.297, + "step": 39713 + }, + { + "epoch": 2.7558115328568453, + "grad_norm": 0.7442787289619446, + "learning_rate": 4.6595432585698806e-05, + "loss": 0.2935, + "step": 39714 + }, + { + "epoch": 2.755880924293942, + "grad_norm": 1.377293586730957, + "learning_rate": 4.659068237993037e-05, + "loss": 0.4134, + "step": 39715 + }, + { + "epoch": 2.755950315731039, + "grad_norm": 0.7669578790664673, + "learning_rate": 4.6585932342768956e-05, + "loss": 0.3058, + "step": 39716 + }, + { + "epoch": 2.7560197071681354, + "grad_norm": 0.8070973753929138, + "learning_rate": 4.65811824742296e-05, + "loss": 0.3222, + "step": 39717 + }, + { + "epoch": 2.756089098605232, + "grad_norm": 0.9643106460571289, + "learning_rate": 4.6576432774327215e-05, + "loss": 0.3669, + "step": 39718 + }, + { + "epoch": 2.7561584900423286, + "grad_norm": 1.0988785028457642, + "learning_rate": 4.6571683243076826e-05, + "loss": 0.3478, + "step": 39719 + }, + { + "epoch": 2.7562278814794254, + "grad_norm": 1.33283531665802, + "learning_rate": 4.6566933880493446e-05, + "loss": 0.5884, + "step": 39720 + }, + { + "epoch": 2.7562972729165223, + "grad_norm": 1.2607200145721436, + "learning_rate": 4.656218468659205e-05, + "loss": 0.3937, + "step": 39721 + }, + { + "epoch": 2.7563666643536187, + "grad_norm": 1.367462396621704, + "learning_rate": 4.655743566138764e-05, + "loss": 0.4276, + "step": 39722 + }, + { + "epoch": 2.7564360557907155, + "grad_norm": 0.9179089069366455, + "learning_rate": 4.655268680489519e-05, + "loss": 0.3457, + "step": 39723 + }, + { + "epoch": 2.756505447227812, + "grad_norm": 1.3001176118850708, + "learning_rate": 4.654793811712976e-05, + "loss": 0.3546, + "step": 39724 + }, + { + "epoch": 2.7565748386649087, + "grad_norm": 1.4871876239776611, + "learning_rate": 4.654318959810625e-05, + "loss": 0.3543, + "step": 39725 + }, + { + "epoch": 2.7566442301020055, + "grad_norm": 1.3409535884857178, + "learning_rate": 4.653844124783968e-05, + "loss": 0.4428, + "step": 39726 + }, + { + "epoch": 2.756713621539102, + "grad_norm": 1.0504268407821655, + "learning_rate": 4.6533693066345095e-05, + "loss": 0.3406, + "step": 39727 + }, + { + "epoch": 2.7567830129761988, + "grad_norm": 1.508689522743225, + "learning_rate": 4.6528945053637397e-05, + "loss": 0.6532, + "step": 39728 + }, + { + "epoch": 2.756852404413295, + "grad_norm": 1.1927779912948608, + "learning_rate": 4.6524197209731615e-05, + "loss": 0.3735, + "step": 39729 + }, + { + "epoch": 2.756921795850392, + "grad_norm": 1.3612347841262817, + "learning_rate": 4.651944953464273e-05, + "loss": 0.3792, + "step": 39730 + }, + { + "epoch": 2.756991187287489, + "grad_norm": 1.6881963014602661, + "learning_rate": 4.6514702028385784e-05, + "loss": 0.412, + "step": 39731 + }, + { + "epoch": 2.7570605787245857, + "grad_norm": 1.095645785331726, + "learning_rate": 4.650995469097568e-05, + "loss": 0.3482, + "step": 39732 + }, + { + "epoch": 2.757129970161682, + "grad_norm": 1.454954743385315, + "learning_rate": 4.650520752242744e-05, + "loss": 0.6531, + "step": 39733 + }, + { + "epoch": 2.757199361598779, + "grad_norm": 1.1904436349868774, + "learning_rate": 4.6500460522756064e-05, + "loss": 0.2274, + "step": 39734 + }, + { + "epoch": 2.7572687530358753, + "grad_norm": 1.1112622022628784, + "learning_rate": 4.649571369197652e-05, + "loss": 0.5111, + "step": 39735 + }, + { + "epoch": 2.757338144472972, + "grad_norm": 1.1865988969802856, + "learning_rate": 4.649096703010379e-05, + "loss": 0.3836, + "step": 39736 + }, + { + "epoch": 2.757407535910069, + "grad_norm": 1.294664740562439, + "learning_rate": 4.648622053715288e-05, + "loss": 0.5581, + "step": 39737 + }, + { + "epoch": 2.7574769273471653, + "grad_norm": 1.009562611579895, + "learning_rate": 4.64814742131388e-05, + "loss": 0.2782, + "step": 39738 + }, + { + "epoch": 2.757546318784262, + "grad_norm": 0.9697675704956055, + "learning_rate": 4.647672805807646e-05, + "loss": 0.2987, + "step": 39739 + }, + { + "epoch": 2.7576157102213585, + "grad_norm": 0.9234620332717896, + "learning_rate": 4.647198207198087e-05, + "loss": 0.3277, + "step": 39740 + }, + { + "epoch": 2.7576851016584554, + "grad_norm": 1.3008811473846436, + "learning_rate": 4.646723625486703e-05, + "loss": 0.3367, + "step": 39741 + }, + { + "epoch": 2.757754493095552, + "grad_norm": 0.9467839598655701, + "learning_rate": 4.646249060674992e-05, + "loss": 0.2596, + "step": 39742 + }, + { + "epoch": 2.7578238845326486, + "grad_norm": 0.9409823417663574, + "learning_rate": 4.64577451276445e-05, + "loss": 0.4438, + "step": 39743 + }, + { + "epoch": 2.7578932759697454, + "grad_norm": 0.9943633675575256, + "learning_rate": 4.645299981756578e-05, + "loss": 0.2956, + "step": 39744 + }, + { + "epoch": 2.757962667406842, + "grad_norm": 1.1585553884506226, + "learning_rate": 4.644825467652876e-05, + "loss": 0.3966, + "step": 39745 + }, + { + "epoch": 2.7580320588439386, + "grad_norm": 0.7967986464500427, + "learning_rate": 4.6443509704548346e-05, + "loss": 0.3686, + "step": 39746 + }, + { + "epoch": 2.7581014502810355, + "grad_norm": 1.9166796207427979, + "learning_rate": 4.6438764901639564e-05, + "loss": 0.4502, + "step": 39747 + }, + { + "epoch": 2.758170841718132, + "grad_norm": 1.9398458003997803, + "learning_rate": 4.643402026781739e-05, + "loss": 0.5244, + "step": 39748 + }, + { + "epoch": 2.7582402331552287, + "grad_norm": 0.975098729133606, + "learning_rate": 4.64292758030968e-05, + "loss": 0.4475, + "step": 39749 + }, + { + "epoch": 2.758309624592325, + "grad_norm": 0.74789959192276, + "learning_rate": 4.6424531507492764e-05, + "loss": 0.3211, + "step": 39750 + }, + { + "epoch": 2.758379016029422, + "grad_norm": 0.9077175855636597, + "learning_rate": 4.6419787381020265e-05, + "loss": 0.2899, + "step": 39751 + }, + { + "epoch": 2.7584484074665188, + "grad_norm": 0.7738667726516724, + "learning_rate": 4.6415043423694316e-05, + "loss": 0.2833, + "step": 39752 + }, + { + "epoch": 2.758517798903615, + "grad_norm": 1.2675105333328247, + "learning_rate": 4.641029963552983e-05, + "loss": 0.4769, + "step": 39753 + }, + { + "epoch": 2.758587190340712, + "grad_norm": 1.2400314807891846, + "learning_rate": 4.6405556016541796e-05, + "loss": 0.4524, + "step": 39754 + }, + { + "epoch": 2.7586565817778084, + "grad_norm": 1.1602319478988647, + "learning_rate": 4.640081256674522e-05, + "loss": 0.4416, + "step": 39755 + }, + { + "epoch": 2.758725973214905, + "grad_norm": 1.444042682647705, + "learning_rate": 4.6396069286155055e-05, + "loss": 0.4474, + "step": 39756 + }, + { + "epoch": 2.758795364652002, + "grad_norm": 1.4100006818771362, + "learning_rate": 4.6391326174786274e-05, + "loss": 0.3839, + "step": 39757 + }, + { + "epoch": 2.758864756089099, + "grad_norm": 1.6859506368637085, + "learning_rate": 4.6386583232653856e-05, + "loss": 0.6069, + "step": 39758 + }, + { + "epoch": 2.7589341475261953, + "grad_norm": 1.7073003053665161, + "learning_rate": 4.638184045977282e-05, + "loss": 0.4042, + "step": 39759 + }, + { + "epoch": 2.759003538963292, + "grad_norm": 1.061216950416565, + "learning_rate": 4.6377097856158045e-05, + "loss": 0.2204, + "step": 39760 + }, + { + "epoch": 2.7590729304003885, + "grad_norm": 2.134833574295044, + "learning_rate": 4.637235542182456e-05, + "loss": 0.4067, + "step": 39761 + }, + { + "epoch": 2.7591423218374853, + "grad_norm": 1.3250486850738525, + "learning_rate": 4.636761315678737e-05, + "loss": 0.3448, + "step": 39762 + }, + { + "epoch": 2.759211713274582, + "grad_norm": 1.9512914419174194, + "learning_rate": 4.6362871061061355e-05, + "loss": 0.5853, + "step": 39763 + }, + { + "epoch": 2.7592811047116785, + "grad_norm": 2.680962085723877, + "learning_rate": 4.6358129134661534e-05, + "loss": 0.3892, + "step": 39764 + }, + { + "epoch": 2.7593504961487754, + "grad_norm": 1.0489667654037476, + "learning_rate": 4.635338737760288e-05, + "loss": 0.3405, + "step": 39765 + }, + { + "epoch": 2.7594198875858718, + "grad_norm": 1.2333595752716064, + "learning_rate": 4.6348645789900394e-05, + "loss": 0.385, + "step": 39766 + }, + { + "epoch": 2.7594892790229686, + "grad_norm": 1.321344017982483, + "learning_rate": 4.634390437156897e-05, + "loss": 0.4605, + "step": 39767 + }, + { + "epoch": 2.7595586704600654, + "grad_norm": 0.9208200573921204, + "learning_rate": 4.633916312262361e-05, + "loss": 0.2924, + "step": 39768 + }, + { + "epoch": 2.759628061897162, + "grad_norm": 1.3432585000991821, + "learning_rate": 4.6334422043079284e-05, + "loss": 0.4986, + "step": 39769 + }, + { + "epoch": 2.7596974533342586, + "grad_norm": 1.144394040107727, + "learning_rate": 4.6329681132950975e-05, + "loss": 0.3357, + "step": 39770 + }, + { + "epoch": 2.759766844771355, + "grad_norm": 0.9776185750961304, + "learning_rate": 4.632494039225362e-05, + "loss": 0.3392, + "step": 39771 + }, + { + "epoch": 2.759836236208452, + "grad_norm": 1.1999951601028442, + "learning_rate": 4.632019982100225e-05, + "loss": 0.375, + "step": 39772 + }, + { + "epoch": 2.7599056276455487, + "grad_norm": 1.428541660308838, + "learning_rate": 4.631545941921174e-05, + "loss": 0.3686, + "step": 39773 + }, + { + "epoch": 2.759975019082645, + "grad_norm": 1.2874616384506226, + "learning_rate": 4.63107191868971e-05, + "loss": 0.3818, + "step": 39774 + }, + { + "epoch": 2.760044410519742, + "grad_norm": 1.5140498876571655, + "learning_rate": 4.630597912407329e-05, + "loss": 0.4498, + "step": 39775 + }, + { + "epoch": 2.7601138019568383, + "grad_norm": 0.8793973922729492, + "learning_rate": 4.6301239230755265e-05, + "loss": 0.3345, + "step": 39776 + }, + { + "epoch": 2.760183193393935, + "grad_norm": 1.2416633367538452, + "learning_rate": 4.6296499506958005e-05, + "loss": 0.5229, + "step": 39777 + }, + { + "epoch": 2.760252584831032, + "grad_norm": 1.1745957136154175, + "learning_rate": 4.6291759952696456e-05, + "loss": 0.3931, + "step": 39778 + }, + { + "epoch": 2.7603219762681284, + "grad_norm": 1.5748929977416992, + "learning_rate": 4.628702056798564e-05, + "loss": 0.3858, + "step": 39779 + }, + { + "epoch": 2.760391367705225, + "grad_norm": 1.3207404613494873, + "learning_rate": 4.628228135284043e-05, + "loss": 0.3402, + "step": 39780 + }, + { + "epoch": 2.7604607591423216, + "grad_norm": 0.8874020576477051, + "learning_rate": 4.627754230727581e-05, + "loss": 0.2314, + "step": 39781 + }, + { + "epoch": 2.7605301505794184, + "grad_norm": 1.1269539594650269, + "learning_rate": 4.6272803431306775e-05, + "loss": 0.3837, + "step": 39782 + }, + { + "epoch": 2.7605995420165153, + "grad_norm": 2.1599299907684326, + "learning_rate": 4.626806472494826e-05, + "loss": 0.437, + "step": 39783 + }, + { + "epoch": 2.760668933453612, + "grad_norm": 1.3598049879074097, + "learning_rate": 4.6263326188215226e-05, + "loss": 0.3844, + "step": 39784 + }, + { + "epoch": 2.7607383248907085, + "grad_norm": 1.3327299356460571, + "learning_rate": 4.625858782112263e-05, + "loss": 0.3492, + "step": 39785 + }, + { + "epoch": 2.7608077163278053, + "grad_norm": 1.1888459920883179, + "learning_rate": 4.6253849623685485e-05, + "loss": 0.2932, + "step": 39786 + }, + { + "epoch": 2.7608771077649017, + "grad_norm": 0.7594896554946899, + "learning_rate": 4.6249111595918674e-05, + "loss": 0.2767, + "step": 39787 + }, + { + "epoch": 2.7609464992019985, + "grad_norm": 1.1809992790222168, + "learning_rate": 4.6244373737837166e-05, + "loss": 0.4604, + "step": 39788 + }, + { + "epoch": 2.7610158906390954, + "grad_norm": 1.296575903892517, + "learning_rate": 4.623963604945594e-05, + "loss": 0.3879, + "step": 39789 + }, + { + "epoch": 2.7610852820761917, + "grad_norm": 1.4500337839126587, + "learning_rate": 4.623489853078994e-05, + "loss": 0.3884, + "step": 39790 + }, + { + "epoch": 2.7611546735132886, + "grad_norm": 1.0740467309951782, + "learning_rate": 4.623016118185413e-05, + "loss": 0.3391, + "step": 39791 + }, + { + "epoch": 2.761224064950385, + "grad_norm": 1.6476205587387085, + "learning_rate": 4.6225424002663456e-05, + "loss": 0.3682, + "step": 39792 + }, + { + "epoch": 2.761293456387482, + "grad_norm": 0.8872443437576294, + "learning_rate": 4.622068699323292e-05, + "loss": 0.2931, + "step": 39793 + }, + { + "epoch": 2.7613628478245786, + "grad_norm": 1.1796742677688599, + "learning_rate": 4.62159501535774e-05, + "loss": 0.4275, + "step": 39794 + }, + { + "epoch": 2.761432239261675, + "grad_norm": 1.1052972078323364, + "learning_rate": 4.621121348371189e-05, + "loss": 0.2874, + "step": 39795 + }, + { + "epoch": 2.761501630698772, + "grad_norm": 0.8283626437187195, + "learning_rate": 4.620647698365134e-05, + "loss": 0.2189, + "step": 39796 + }, + { + "epoch": 2.7615710221358682, + "grad_norm": 0.6212873458862305, + "learning_rate": 4.620174065341073e-05, + "loss": 0.2009, + "step": 39797 + }, + { + "epoch": 2.761640413572965, + "grad_norm": 0.9460880756378174, + "learning_rate": 4.6197004493004945e-05, + "loss": 0.3114, + "step": 39798 + }, + { + "epoch": 2.761709805010062, + "grad_norm": 1.2783397436141968, + "learning_rate": 4.619226850244899e-05, + "loss": 0.3424, + "step": 39799 + }, + { + "epoch": 2.7617791964471583, + "grad_norm": 0.9267919063568115, + "learning_rate": 4.6187532681757783e-05, + "loss": 0.3697, + "step": 39800 + }, + { + "epoch": 2.761848587884255, + "grad_norm": 0.7623544931411743, + "learning_rate": 4.6182797030946335e-05, + "loss": 0.2943, + "step": 39801 + }, + { + "epoch": 2.7619179793213515, + "grad_norm": 1.0347846746444702, + "learning_rate": 4.6178061550029515e-05, + "loss": 0.3368, + "step": 39802 + }, + { + "epoch": 2.7619873707584484, + "grad_norm": 1.293819785118103, + "learning_rate": 4.617332623902232e-05, + "loss": 0.3678, + "step": 39803 + }, + { + "epoch": 2.762056762195545, + "grad_norm": 1.0111790895462036, + "learning_rate": 4.6168591097939674e-05, + "loss": 0.3362, + "step": 39804 + }, + { + "epoch": 2.7621261536326416, + "grad_norm": 1.2328836917877197, + "learning_rate": 4.616385612679654e-05, + "loss": 0.4014, + "step": 39805 + }, + { + "epoch": 2.7621955450697384, + "grad_norm": 1.1606297492980957, + "learning_rate": 4.615912132560788e-05, + "loss": 0.363, + "step": 39806 + }, + { + "epoch": 2.762264936506835, + "grad_norm": 1.4028444290161133, + "learning_rate": 4.6154386694388653e-05, + "loss": 0.322, + "step": 39807 + }, + { + "epoch": 2.7623343279439316, + "grad_norm": 1.6034584045410156, + "learning_rate": 4.614965223315375e-05, + "loss": 0.376, + "step": 39808 + }, + { + "epoch": 2.7624037193810285, + "grad_norm": 1.620659351348877, + "learning_rate": 4.614491794191814e-05, + "loss": 0.4873, + "step": 39809 + }, + { + "epoch": 2.7624731108181253, + "grad_norm": 0.9751870036125183, + "learning_rate": 4.614018382069678e-05, + "loss": 0.2914, + "step": 39810 + }, + { + "epoch": 2.7625425022552217, + "grad_norm": 1.5211830139160156, + "learning_rate": 4.613544986950461e-05, + "loss": 0.5317, + "step": 39811 + }, + { + "epoch": 2.7626118936923185, + "grad_norm": 1.9698710441589355, + "learning_rate": 4.6130716088356574e-05, + "loss": 0.3233, + "step": 39812 + }, + { + "epoch": 2.762681285129415, + "grad_norm": 1.1694740056991577, + "learning_rate": 4.612598247726763e-05, + "loss": 0.3824, + "step": 39813 + }, + { + "epoch": 2.7627506765665117, + "grad_norm": 1.1038317680358887, + "learning_rate": 4.6121249036252735e-05, + "loss": 0.3809, + "step": 39814 + }, + { + "epoch": 2.7628200680036086, + "grad_norm": 1.17425537109375, + "learning_rate": 4.611651576532676e-05, + "loss": 0.3187, + "step": 39815 + }, + { + "epoch": 2.762889459440705, + "grad_norm": 1.6260685920715332, + "learning_rate": 4.61117826645047e-05, + "loss": 0.2973, + "step": 39816 + }, + { + "epoch": 2.762958850877802, + "grad_norm": 1.263256311416626, + "learning_rate": 4.61070497338015e-05, + "loss": 0.4698, + "step": 39817 + }, + { + "epoch": 2.763028242314898, + "grad_norm": 1.0982129573822021, + "learning_rate": 4.610231697323207e-05, + "loss": 0.3903, + "step": 39818 + }, + { + "epoch": 2.763097633751995, + "grad_norm": 1.0051316022872925, + "learning_rate": 4.609758438281139e-05, + "loss": 0.2571, + "step": 39819 + }, + { + "epoch": 2.763167025189092, + "grad_norm": 1.2493466138839722, + "learning_rate": 4.6092851962554384e-05, + "loss": 0.5223, + "step": 39820 + }, + { + "epoch": 2.7632364166261882, + "grad_norm": 1.1848154067993164, + "learning_rate": 4.608811971247602e-05, + "loss": 0.3346, + "step": 39821 + }, + { + "epoch": 2.763305808063285, + "grad_norm": 1.8681131601333618, + "learning_rate": 4.6083387632591165e-05, + "loss": 0.5836, + "step": 39822 + }, + { + "epoch": 2.7633751995003815, + "grad_norm": 1.4177592992782593, + "learning_rate": 4.607865572291481e-05, + "loss": 0.3317, + "step": 39823 + }, + { + "epoch": 2.7634445909374783, + "grad_norm": 1.1811392307281494, + "learning_rate": 4.607392398346188e-05, + "loss": 0.2393, + "step": 39824 + }, + { + "epoch": 2.763513982374575, + "grad_norm": 0.7737924456596375, + "learning_rate": 4.606919241424731e-05, + "loss": 0.2475, + "step": 39825 + }, + { + "epoch": 2.7635833738116715, + "grad_norm": 1.332481861114502, + "learning_rate": 4.6064461015286044e-05, + "loss": 0.466, + "step": 39826 + }, + { + "epoch": 2.7636527652487684, + "grad_norm": 1.4583793878555298, + "learning_rate": 4.605972978659302e-05, + "loss": 0.4847, + "step": 39827 + }, + { + "epoch": 2.7637221566858647, + "grad_norm": 1.4136689901351929, + "learning_rate": 4.6054998728183206e-05, + "loss": 0.4087, + "step": 39828 + }, + { + "epoch": 2.7637915481229616, + "grad_norm": 1.3618648052215576, + "learning_rate": 4.605026784007146e-05, + "loss": 0.4317, + "step": 39829 + }, + { + "epoch": 2.7638609395600584, + "grad_norm": 1.3773788213729858, + "learning_rate": 4.604553712227276e-05, + "loss": 0.4712, + "step": 39830 + }, + { + "epoch": 2.763930330997155, + "grad_norm": 0.7822592258453369, + "learning_rate": 4.604080657480205e-05, + "loss": 0.2221, + "step": 39831 + }, + { + "epoch": 2.7639997224342516, + "grad_norm": 1.3509812355041504, + "learning_rate": 4.603607619767428e-05, + "loss": 0.4258, + "step": 39832 + }, + { + "epoch": 2.764069113871348, + "grad_norm": 1.9667787551879883, + "learning_rate": 4.603134599090433e-05, + "loss": 0.435, + "step": 39833 + }, + { + "epoch": 2.764138505308445, + "grad_norm": 1.6827359199523926, + "learning_rate": 4.6026615954507156e-05, + "loss": 0.4017, + "step": 39834 + }, + { + "epoch": 2.7642078967455417, + "grad_norm": 1.222359299659729, + "learning_rate": 4.602188608849769e-05, + "loss": 0.4875, + "step": 39835 + }, + { + "epoch": 2.7642772881826385, + "grad_norm": 0.913834810256958, + "learning_rate": 4.60171563928909e-05, + "loss": 0.2359, + "step": 39836 + }, + { + "epoch": 2.764346679619735, + "grad_norm": 1.3051435947418213, + "learning_rate": 4.6012426867701654e-05, + "loss": 0.3186, + "step": 39837 + }, + { + "epoch": 2.7644160710568317, + "grad_norm": 1.0590344667434692, + "learning_rate": 4.600769751294491e-05, + "loss": 0.3247, + "step": 39838 + }, + { + "epoch": 2.764485462493928, + "grad_norm": 1.0323526859283447, + "learning_rate": 4.6002968328635596e-05, + "loss": 0.3382, + "step": 39839 + }, + { + "epoch": 2.764554853931025, + "grad_norm": 1.1534706354141235, + "learning_rate": 4.599823931478865e-05, + "loss": 0.4178, + "step": 39840 + }, + { + "epoch": 2.764624245368122, + "grad_norm": 1.0379576683044434, + "learning_rate": 4.5993510471419046e-05, + "loss": 0.3175, + "step": 39841 + }, + { + "epoch": 2.764693636805218, + "grad_norm": 1.3744534254074097, + "learning_rate": 4.5988781798541616e-05, + "loss": 0.4352, + "step": 39842 + }, + { + "epoch": 2.764763028242315, + "grad_norm": 1.1813255548477173, + "learning_rate": 4.598405329617133e-05, + "loss": 0.4436, + "step": 39843 + }, + { + "epoch": 2.7648324196794114, + "grad_norm": 1.084287166595459, + "learning_rate": 4.5979324964323125e-05, + "loss": 0.3542, + "step": 39844 + }, + { + "epoch": 2.7649018111165082, + "grad_norm": 1.0135146379470825, + "learning_rate": 4.597459680301193e-05, + "loss": 0.2952, + "step": 39845 + }, + { + "epoch": 2.764971202553605, + "grad_norm": 1.1431218385696411, + "learning_rate": 4.596986881225265e-05, + "loss": 0.3746, + "step": 39846 + }, + { + "epoch": 2.7650405939907015, + "grad_norm": 1.05313241481781, + "learning_rate": 4.596514099206024e-05, + "loss": 0.435, + "step": 39847 + }, + { + "epoch": 2.7651099854277983, + "grad_norm": 1.9668965339660645, + "learning_rate": 4.596041334244964e-05, + "loss": 0.6234, + "step": 39848 + }, + { + "epoch": 2.7651793768648947, + "grad_norm": 1.6598013639450073, + "learning_rate": 4.59556858634357e-05, + "loss": 0.6386, + "step": 39849 + }, + { + "epoch": 2.7652487683019915, + "grad_norm": 0.8724278211593628, + "learning_rate": 4.595095855503341e-05, + "loss": 0.3991, + "step": 39850 + }, + { + "epoch": 2.7653181597390883, + "grad_norm": 0.7811035513877869, + "learning_rate": 4.5946231417257654e-05, + "loss": 0.2434, + "step": 39851 + }, + { + "epoch": 2.7653875511761847, + "grad_norm": 1.2642637491226196, + "learning_rate": 4.5941504450123396e-05, + "loss": 0.6398, + "step": 39852 + }, + { + "epoch": 2.7654569426132816, + "grad_norm": 1.0633442401885986, + "learning_rate": 4.5936777653645514e-05, + "loss": 0.3197, + "step": 39853 + }, + { + "epoch": 2.765526334050378, + "grad_norm": 0.7499555945396423, + "learning_rate": 4.5932051027838964e-05, + "loss": 0.3198, + "step": 39854 + }, + { + "epoch": 2.765595725487475, + "grad_norm": 2.1680235862731934, + "learning_rate": 4.592732457271869e-05, + "loss": 0.3729, + "step": 39855 + }, + { + "epoch": 2.7656651169245716, + "grad_norm": 0.9154797792434692, + "learning_rate": 4.592259828829954e-05, + "loss": 0.3432, + "step": 39856 + }, + { + "epoch": 2.765734508361668, + "grad_norm": 1.2196050882339478, + "learning_rate": 4.5917872174596475e-05, + "loss": 0.2538, + "step": 39857 + }, + { + "epoch": 2.765803899798765, + "grad_norm": 0.9871355295181274, + "learning_rate": 4.591314623162442e-05, + "loss": 0.3297, + "step": 39858 + }, + { + "epoch": 2.7658732912358612, + "grad_norm": 1.2882122993469238, + "learning_rate": 4.5908420459398284e-05, + "loss": 0.4651, + "step": 39859 + }, + { + "epoch": 2.765942682672958, + "grad_norm": 1.411746859550476, + "learning_rate": 4.5903694857932976e-05, + "loss": 0.4906, + "step": 39860 + }, + { + "epoch": 2.766012074110055, + "grad_norm": 1.3965425491333008, + "learning_rate": 4.589896942724344e-05, + "loss": 0.5427, + "step": 39861 + }, + { + "epoch": 2.7660814655471517, + "grad_norm": 0.8651453256607056, + "learning_rate": 4.589424416734461e-05, + "loss": 0.2818, + "step": 39862 + }, + { + "epoch": 2.766150856984248, + "grad_norm": 1.324053168296814, + "learning_rate": 4.5889519078251343e-05, + "loss": 0.4524, + "step": 39863 + }, + { + "epoch": 2.766220248421345, + "grad_norm": 1.4898873567581177, + "learning_rate": 4.5884794159978586e-05, + "loss": 0.3378, + "step": 39864 + }, + { + "epoch": 2.7662896398584413, + "grad_norm": 1.1860007047653198, + "learning_rate": 4.588006941254126e-05, + "loss": 0.3904, + "step": 39865 + }, + { + "epoch": 2.766359031295538, + "grad_norm": 1.2000340223312378, + "learning_rate": 4.587534483595426e-05, + "loss": 0.2271, + "step": 39866 + }, + { + "epoch": 2.766428422732635, + "grad_norm": 1.0726803541183472, + "learning_rate": 4.5870620430232566e-05, + "loss": 0.4479, + "step": 39867 + }, + { + "epoch": 2.7664978141697314, + "grad_norm": 1.0337536334991455, + "learning_rate": 4.5865896195391e-05, + "loss": 0.3666, + "step": 39868 + }, + { + "epoch": 2.7665672056068282, + "grad_norm": 1.3214837312698364, + "learning_rate": 4.586117213144453e-05, + "loss": 0.3818, + "step": 39869 + }, + { + "epoch": 2.7666365970439246, + "grad_norm": 1.3586865663528442, + "learning_rate": 4.585644823840808e-05, + "loss": 0.3459, + "step": 39870 + }, + { + "epoch": 2.7667059884810214, + "grad_norm": 1.1219764947891235, + "learning_rate": 4.585172451629651e-05, + "loss": 0.3678, + "step": 39871 + }, + { + "epoch": 2.7667753799181183, + "grad_norm": 1.1095749139785767, + "learning_rate": 4.5847000965124766e-05, + "loss": 0.51, + "step": 39872 + }, + { + "epoch": 2.7668447713552147, + "grad_norm": 1.371648907661438, + "learning_rate": 4.5842277584907757e-05, + "loss": 0.5489, + "step": 39873 + }, + { + "epoch": 2.7669141627923115, + "grad_norm": 0.9386845231056213, + "learning_rate": 4.58375543756604e-05, + "loss": 0.2759, + "step": 39874 + }, + { + "epoch": 2.766983554229408, + "grad_norm": 1.1806663274765015, + "learning_rate": 4.583283133739759e-05, + "loss": 0.5338, + "step": 39875 + }, + { + "epoch": 2.7670529456665047, + "grad_norm": 1.3494064807891846, + "learning_rate": 4.582810847013429e-05, + "loss": 0.5492, + "step": 39876 + }, + { + "epoch": 2.7671223371036016, + "grad_norm": 1.3816438913345337, + "learning_rate": 4.582338577388532e-05, + "loss": 0.589, + "step": 39877 + }, + { + "epoch": 2.767191728540698, + "grad_norm": 1.551637053489685, + "learning_rate": 4.581866324866564e-05, + "loss": 0.4774, + "step": 39878 + }, + { + "epoch": 2.767261119977795, + "grad_norm": 1.5808684825897217, + "learning_rate": 4.581394089449016e-05, + "loss": 0.465, + "step": 39879 + }, + { + "epoch": 2.767330511414891, + "grad_norm": 1.5062695741653442, + "learning_rate": 4.580921871137378e-05, + "loss": 0.4592, + "step": 39880 + }, + { + "epoch": 2.767399902851988, + "grad_norm": 1.9309508800506592, + "learning_rate": 4.58044966993314e-05, + "loss": 0.5845, + "step": 39881 + }, + { + "epoch": 2.767469294289085, + "grad_norm": 1.4245362281799316, + "learning_rate": 4.579977485837794e-05, + "loss": 0.3771, + "step": 39882 + }, + { + "epoch": 2.7675386857261812, + "grad_norm": 1.6784251928329468, + "learning_rate": 4.579505318852835e-05, + "loss": 0.337, + "step": 39883 + }, + { + "epoch": 2.767608077163278, + "grad_norm": 1.3572641611099243, + "learning_rate": 4.579033168979744e-05, + "loss": 0.3639, + "step": 39884 + }, + { + "epoch": 2.7676774686003744, + "grad_norm": 1.20088529586792, + "learning_rate": 4.578561036220017e-05, + "loss": 0.4002, + "step": 39885 + }, + { + "epoch": 2.7677468600374713, + "grad_norm": 1.4895044565200806, + "learning_rate": 4.578088920575143e-05, + "loss": 0.3891, + "step": 39886 + }, + { + "epoch": 2.767816251474568, + "grad_norm": 1.6174026727676392, + "learning_rate": 4.5776168220466145e-05, + "loss": 0.4826, + "step": 39887 + }, + { + "epoch": 2.767885642911665, + "grad_norm": 1.289798617362976, + "learning_rate": 4.577144740635919e-05, + "loss": 0.4679, + "step": 39888 + }, + { + "epoch": 2.7679550343487613, + "grad_norm": 0.8426569104194641, + "learning_rate": 4.576672676344549e-05, + "loss": 0.3564, + "step": 39889 + }, + { + "epoch": 2.768024425785858, + "grad_norm": 1.0191890001296997, + "learning_rate": 4.5762006291739976e-05, + "loss": 0.2385, + "step": 39890 + }, + { + "epoch": 2.7680938172229546, + "grad_norm": 1.1286498308181763, + "learning_rate": 4.575728599125749e-05, + "loss": 0.4316, + "step": 39891 + }, + { + "epoch": 2.7681632086600514, + "grad_norm": 1.1670119762420654, + "learning_rate": 4.575256586201295e-05, + "loss": 0.3515, + "step": 39892 + }, + { + "epoch": 2.768232600097148, + "grad_norm": 1.237552285194397, + "learning_rate": 4.574784590402127e-05, + "loss": 0.3611, + "step": 39893 + }, + { + "epoch": 2.7683019915342446, + "grad_norm": 1.1437187194824219, + "learning_rate": 4.574312611729734e-05, + "loss": 0.3544, + "step": 39894 + }, + { + "epoch": 2.7683713829713414, + "grad_norm": 1.141711711883545, + "learning_rate": 4.573840650185607e-05, + "loss": 0.3791, + "step": 39895 + }, + { + "epoch": 2.768440774408438, + "grad_norm": 1.4699026346206665, + "learning_rate": 4.573368705771235e-05, + "loss": 0.4354, + "step": 39896 + }, + { + "epoch": 2.7685101658455347, + "grad_norm": 1.021207332611084, + "learning_rate": 4.5728967784881124e-05, + "loss": 0.3631, + "step": 39897 + }, + { + "epoch": 2.7685795572826315, + "grad_norm": 1.2228111028671265, + "learning_rate": 4.572424868337721e-05, + "loss": 0.4294, + "step": 39898 + }, + { + "epoch": 2.768648948719728, + "grad_norm": 1.5042537450790405, + "learning_rate": 4.571952975321556e-05, + "loss": 0.4125, + "step": 39899 + }, + { + "epoch": 2.7687183401568247, + "grad_norm": 1.3206639289855957, + "learning_rate": 4.571481099441104e-05, + "loss": 0.3239, + "step": 39900 + }, + { + "epoch": 2.768787731593921, + "grad_norm": 0.9084690809249878, + "learning_rate": 4.5710092406978575e-05, + "loss": 0.3284, + "step": 39901 + }, + { + "epoch": 2.768857123031018, + "grad_norm": 1.2337952852249146, + "learning_rate": 4.5705373990933075e-05, + "loss": 0.3644, + "step": 39902 + }, + { + "epoch": 2.7689265144681148, + "grad_norm": 1.175386667251587, + "learning_rate": 4.570065574628939e-05, + "loss": 0.3789, + "step": 39903 + }, + { + "epoch": 2.768995905905211, + "grad_norm": 1.2042089700698853, + "learning_rate": 4.5695937673062414e-05, + "loss": 0.4446, + "step": 39904 + }, + { + "epoch": 2.769065297342308, + "grad_norm": 1.2939406633377075, + "learning_rate": 4.569121977126711e-05, + "loss": 0.4546, + "step": 39905 + }, + { + "epoch": 2.7691346887794044, + "grad_norm": 1.2097817659378052, + "learning_rate": 4.568650204091829e-05, + "loss": 0.4649, + "step": 39906 + }, + { + "epoch": 2.769204080216501, + "grad_norm": 0.9432533979415894, + "learning_rate": 4.5681784482030875e-05, + "loss": 0.3972, + "step": 39907 + }, + { + "epoch": 2.769273471653598, + "grad_norm": 1.1029247045516968, + "learning_rate": 4.5677067094619774e-05, + "loss": 0.4546, + "step": 39908 + }, + { + "epoch": 2.769342863090695, + "grad_norm": 0.8523438572883606, + "learning_rate": 4.567234987869986e-05, + "loss": 0.2893, + "step": 39909 + }, + { + "epoch": 2.7694122545277913, + "grad_norm": 1.6846566200256348, + "learning_rate": 4.5667632834286034e-05, + "loss": 0.4698, + "step": 39910 + }, + { + "epoch": 2.7694816459648877, + "grad_norm": 1.1066834926605225, + "learning_rate": 4.566291596139324e-05, + "loss": 0.3813, + "step": 39911 + }, + { + "epoch": 2.7695510374019845, + "grad_norm": 1.2335234880447388, + "learning_rate": 4.565819926003626e-05, + "loss": 0.3484, + "step": 39912 + }, + { + "epoch": 2.7696204288390813, + "grad_norm": 1.3371728658676147, + "learning_rate": 4.565348273023005e-05, + "loss": 0.3467, + "step": 39913 + }, + { + "epoch": 2.769689820276178, + "grad_norm": 1.0004314184188843, + "learning_rate": 4.5648766371989484e-05, + "loss": 0.3767, + "step": 39914 + }, + { + "epoch": 2.7697592117132745, + "grad_norm": 1.3714419603347778, + "learning_rate": 4.5644050185329456e-05, + "loss": 0.3811, + "step": 39915 + }, + { + "epoch": 2.7698286031503714, + "grad_norm": 1.0453169345855713, + "learning_rate": 4.563933417026486e-05, + "loss": 0.4228, + "step": 39916 + }, + { + "epoch": 2.7698979945874678, + "grad_norm": 1.005858063697815, + "learning_rate": 4.563461832681062e-05, + "loss": 0.2951, + "step": 39917 + }, + { + "epoch": 2.7699673860245646, + "grad_norm": 1.3247870206832886, + "learning_rate": 4.5629902654981535e-05, + "loss": 0.3884, + "step": 39918 + }, + { + "epoch": 2.7700367774616614, + "grad_norm": 1.1040431261062622, + "learning_rate": 4.562518715479255e-05, + "loss": 0.3015, + "step": 39919 + }, + { + "epoch": 2.770106168898758, + "grad_norm": 1.1005178689956665, + "learning_rate": 4.5620471826258524e-05, + "loss": 0.3124, + "step": 39920 + }, + { + "epoch": 2.7701755603358547, + "grad_norm": 1.2319532632827759, + "learning_rate": 4.5615756669394374e-05, + "loss": 0.3192, + "step": 39921 + }, + { + "epoch": 2.770244951772951, + "grad_norm": 1.3036689758300781, + "learning_rate": 4.561104168421497e-05, + "loss": 0.3214, + "step": 39922 + }, + { + "epoch": 2.770314343210048, + "grad_norm": 3.2360692024230957, + "learning_rate": 4.5606326870735194e-05, + "loss": 0.3455, + "step": 39923 + }, + { + "epoch": 2.7703837346471447, + "grad_norm": 1.5433099269866943, + "learning_rate": 4.560161222896997e-05, + "loss": 0.4026, + "step": 39924 + }, + { + "epoch": 2.770453126084241, + "grad_norm": 1.0022332668304443, + "learning_rate": 4.559689775893411e-05, + "loss": 0.4158, + "step": 39925 + }, + { + "epoch": 2.770522517521338, + "grad_norm": 1.1487512588500977, + "learning_rate": 4.5592183460642524e-05, + "loss": 0.4796, + "step": 39926 + }, + { + "epoch": 2.7705919089584343, + "grad_norm": 1.431044101715088, + "learning_rate": 4.558746933411011e-05, + "loss": 0.3625, + "step": 39927 + }, + { + "epoch": 2.770661300395531, + "grad_norm": 1.6429139375686646, + "learning_rate": 4.558275537935174e-05, + "loss": 0.3313, + "step": 39928 + }, + { + "epoch": 2.770730691832628, + "grad_norm": 0.6362351775169373, + "learning_rate": 4.557804159638229e-05, + "loss": 0.2489, + "step": 39929 + }, + { + "epoch": 2.7708000832697244, + "grad_norm": 1.4776804447174072, + "learning_rate": 4.557332798521666e-05, + "loss": 0.3275, + "step": 39930 + }, + { + "epoch": 2.770869474706821, + "grad_norm": 1.8979486227035522, + "learning_rate": 4.556861454586976e-05, + "loss": 0.4783, + "step": 39931 + }, + { + "epoch": 2.7709388661439176, + "grad_norm": 0.9377026557922363, + "learning_rate": 4.556390127835638e-05, + "loss": 0.5114, + "step": 39932 + }, + { + "epoch": 2.7710082575810144, + "grad_norm": 1.3345322608947754, + "learning_rate": 4.555918818269145e-05, + "loss": 0.4269, + "step": 39933 + }, + { + "epoch": 2.7710776490181113, + "grad_norm": 1.5495131015777588, + "learning_rate": 4.555447525888985e-05, + "loss": 0.4311, + "step": 39934 + }, + { + "epoch": 2.771147040455208, + "grad_norm": 1.1196544170379639, + "learning_rate": 4.5549762506966456e-05, + "loss": 0.4873, + "step": 39935 + }, + { + "epoch": 2.7712164318923045, + "grad_norm": 1.3311971426010132, + "learning_rate": 4.554504992693618e-05, + "loss": 0.4315, + "step": 39936 + }, + { + "epoch": 2.771285823329401, + "grad_norm": 1.034231424331665, + "learning_rate": 4.5540337518813816e-05, + "loss": 0.3585, + "step": 39937 + }, + { + "epoch": 2.7713552147664977, + "grad_norm": 0.9979843497276306, + "learning_rate": 4.5535625282614305e-05, + "loss": 0.3439, + "step": 39938 + }, + { + "epoch": 2.7714246062035945, + "grad_norm": 1.627886176109314, + "learning_rate": 4.553091321835249e-05, + "loss": 0.5782, + "step": 39939 + }, + { + "epoch": 2.7714939976406914, + "grad_norm": 0.788648247718811, + "learning_rate": 4.5526201326043304e-05, + "loss": 0.2464, + "step": 39940 + }, + { + "epoch": 2.7715633890777878, + "grad_norm": 1.1233865022659302, + "learning_rate": 4.552148960570155e-05, + "loss": 0.4304, + "step": 39941 + }, + { + "epoch": 2.7716327805148846, + "grad_norm": 1.1671329736709595, + "learning_rate": 4.5516778057342136e-05, + "loss": 0.5328, + "step": 39942 + }, + { + "epoch": 2.771702171951981, + "grad_norm": 1.5586767196655273, + "learning_rate": 4.5512066680979925e-05, + "loss": 0.3451, + "step": 39943 + }, + { + "epoch": 2.771771563389078, + "grad_norm": 0.9343401789665222, + "learning_rate": 4.55073554766298e-05, + "loss": 0.2858, + "step": 39944 + }, + { + "epoch": 2.7718409548261747, + "grad_norm": 0.8482872247695923, + "learning_rate": 4.550264444430666e-05, + "loss": 0.3076, + "step": 39945 + }, + { + "epoch": 2.771910346263271, + "grad_norm": 1.2240071296691895, + "learning_rate": 4.549793358402532e-05, + "loss": 0.3584, + "step": 39946 + }, + { + "epoch": 2.771979737700368, + "grad_norm": 0.9802775382995605, + "learning_rate": 4.549322289580068e-05, + "loss": 0.377, + "step": 39947 + }, + { + "epoch": 2.7720491291374643, + "grad_norm": 1.3824539184570312, + "learning_rate": 4.548851237964761e-05, + "loss": 0.4508, + "step": 39948 + }, + { + "epoch": 2.772118520574561, + "grad_norm": 1.128422737121582, + "learning_rate": 4.548380203558098e-05, + "loss": 0.3444, + "step": 39949 + }, + { + "epoch": 2.772187912011658, + "grad_norm": 1.4366270303726196, + "learning_rate": 4.547909186361566e-05, + "loss": 0.4242, + "step": 39950 + }, + { + "epoch": 2.7722573034487543, + "grad_norm": 1.200152039527893, + "learning_rate": 4.5474381863766524e-05, + "loss": 0.2547, + "step": 39951 + }, + { + "epoch": 2.772326694885851, + "grad_norm": 1.126425862312317, + "learning_rate": 4.546967203604848e-05, + "loss": 0.3615, + "step": 39952 + }, + { + "epoch": 2.7723960863229475, + "grad_norm": 1.3221131563186646, + "learning_rate": 4.546496238047632e-05, + "loss": 0.38, + "step": 39953 + }, + { + "epoch": 2.7724654777600444, + "grad_norm": 1.209001064300537, + "learning_rate": 4.546025289706494e-05, + "loss": 0.3921, + "step": 39954 + }, + { + "epoch": 2.772534869197141, + "grad_norm": 1.3021061420440674, + "learning_rate": 4.5455543585829206e-05, + "loss": 0.5197, + "step": 39955 + }, + { + "epoch": 2.7726042606342376, + "grad_norm": 1.5633025169372559, + "learning_rate": 4.545083444678401e-05, + "loss": 0.3313, + "step": 39956 + }, + { + "epoch": 2.7726736520713344, + "grad_norm": 1.0828979015350342, + "learning_rate": 4.544612547994419e-05, + "loss": 0.3712, + "step": 39957 + }, + { + "epoch": 2.772743043508431, + "grad_norm": 1.4927477836608887, + "learning_rate": 4.5441416685324626e-05, + "loss": 0.4687, + "step": 39958 + }, + { + "epoch": 2.7728124349455276, + "grad_norm": 0.9972364902496338, + "learning_rate": 4.543670806294021e-05, + "loss": 0.2996, + "step": 39959 + }, + { + "epoch": 2.7728818263826245, + "grad_norm": 3.6664624214172363, + "learning_rate": 4.5431999612805744e-05, + "loss": 0.5889, + "step": 39960 + }, + { + "epoch": 2.7729512178197213, + "grad_norm": 1.2647510766983032, + "learning_rate": 4.5427291334936126e-05, + "loss": 0.4582, + "step": 39961 + }, + { + "epoch": 2.7730206092568177, + "grad_norm": 0.7624346613883972, + "learning_rate": 4.542258322934621e-05, + "loss": 0.3135, + "step": 39962 + }, + { + "epoch": 2.7730900006939145, + "grad_norm": 1.1411126852035522, + "learning_rate": 4.5417875296050886e-05, + "loss": 0.4392, + "step": 39963 + }, + { + "epoch": 2.773159392131011, + "grad_norm": 1.6614373922348022, + "learning_rate": 4.541316753506497e-05, + "loss": 0.4921, + "step": 39964 + }, + { + "epoch": 2.7732287835681078, + "grad_norm": 1.2437065839767456, + "learning_rate": 4.540845994640337e-05, + "loss": 0.3931, + "step": 39965 + }, + { + "epoch": 2.7732981750052046, + "grad_norm": 1.4298123121261597, + "learning_rate": 4.540375253008096e-05, + "loss": 0.3994, + "step": 39966 + }, + { + "epoch": 2.773367566442301, + "grad_norm": 1.1779134273529053, + "learning_rate": 4.539904528611253e-05, + "loss": 0.2342, + "step": 39967 + }, + { + "epoch": 2.773436957879398, + "grad_norm": 1.2883012294769287, + "learning_rate": 4.5394338214512975e-05, + "loss": 0.4438, + "step": 39968 + }, + { + "epoch": 2.773506349316494, + "grad_norm": 1.324256181716919, + "learning_rate": 4.538963131529717e-05, + "loss": 0.4643, + "step": 39969 + }, + { + "epoch": 2.773575740753591, + "grad_norm": 1.0088109970092773, + "learning_rate": 4.538492458847996e-05, + "loss": 0.3978, + "step": 39970 + }, + { + "epoch": 2.773645132190688, + "grad_norm": 1.6902343034744263, + "learning_rate": 4.5380218034076236e-05, + "loss": 0.4614, + "step": 39971 + }, + { + "epoch": 2.7737145236277843, + "grad_norm": 1.392328143119812, + "learning_rate": 4.53755116521008e-05, + "loss": 0.3365, + "step": 39972 + }, + { + "epoch": 2.773783915064881, + "grad_norm": 0.852790117263794, + "learning_rate": 4.537080544256853e-05, + "loss": 0.3721, + "step": 39973 + }, + { + "epoch": 2.7738533065019775, + "grad_norm": 0.9987885355949402, + "learning_rate": 4.5366099405494286e-05, + "loss": 0.3325, + "step": 39974 + }, + { + "epoch": 2.7739226979390743, + "grad_norm": 1.097573161125183, + "learning_rate": 4.536139354089297e-05, + "loss": 0.3745, + "step": 39975 + }, + { + "epoch": 2.773992089376171, + "grad_norm": 1.3922010660171509, + "learning_rate": 4.5356687848779356e-05, + "loss": 0.3885, + "step": 39976 + }, + { + "epoch": 2.7740614808132675, + "grad_norm": 1.1707119941711426, + "learning_rate": 4.535198232916834e-05, + "loss": 0.3523, + "step": 39977 + }, + { + "epoch": 2.7741308722503644, + "grad_norm": 1.3970013856887817, + "learning_rate": 4.534727698207477e-05, + "loss": 0.5542, + "step": 39978 + }, + { + "epoch": 2.7742002636874608, + "grad_norm": 1.144559383392334, + "learning_rate": 4.534257180751351e-05, + "loss": 0.3589, + "step": 39979 + }, + { + "epoch": 2.7742696551245576, + "grad_norm": 8.709380149841309, + "learning_rate": 4.533786680549945e-05, + "loss": 0.5516, + "step": 39980 + }, + { + "epoch": 2.7743390465616544, + "grad_norm": 0.9972277879714966, + "learning_rate": 4.533316197604737e-05, + "loss": 0.2553, + "step": 39981 + }, + { + "epoch": 2.774408437998751, + "grad_norm": 1.4401100873947144, + "learning_rate": 4.532845731917213e-05, + "loss": 0.5466, + "step": 39982 + }, + { + "epoch": 2.7744778294358476, + "grad_norm": 1.4370137453079224, + "learning_rate": 4.532375283488863e-05, + "loss": 0.4138, + "step": 39983 + }, + { + "epoch": 2.774547220872944, + "grad_norm": 1.500791072845459, + "learning_rate": 4.531904852321169e-05, + "loss": 0.3077, + "step": 39984 + }, + { + "epoch": 2.774616612310041, + "grad_norm": 1.3334853649139404, + "learning_rate": 4.531434438415618e-05, + "loss": 0.3673, + "step": 39985 + }, + { + "epoch": 2.7746860037471377, + "grad_norm": 0.7239691019058228, + "learning_rate": 4.530964041773693e-05, + "loss": 0.2002, + "step": 39986 + }, + { + "epoch": 2.7747553951842345, + "grad_norm": 1.90242338180542, + "learning_rate": 4.530493662396884e-05, + "loss": 0.6079, + "step": 39987 + }, + { + "epoch": 2.774824786621331, + "grad_norm": 1.2972283363342285, + "learning_rate": 4.5300233002866685e-05, + "loss": 0.4279, + "step": 39988 + }, + { + "epoch": 2.7748941780584278, + "grad_norm": 1.624908447265625, + "learning_rate": 4.529552955444535e-05, + "loss": 0.5671, + "step": 39989 + }, + { + "epoch": 2.774963569495524, + "grad_norm": 1.4010989665985107, + "learning_rate": 4.529082627871967e-05, + "loss": 0.389, + "step": 39990 + }, + { + "epoch": 2.775032960932621, + "grad_norm": 1.3600865602493286, + "learning_rate": 4.5286123175704506e-05, + "loss": 0.5448, + "step": 39991 + }, + { + "epoch": 2.775102352369718, + "grad_norm": 1.2969011068344116, + "learning_rate": 4.528142024541471e-05, + "loss": 0.3152, + "step": 39992 + }, + { + "epoch": 2.775171743806814, + "grad_norm": 1.1963472366333008, + "learning_rate": 4.527671748786516e-05, + "loss": 0.3133, + "step": 39993 + }, + { + "epoch": 2.775241135243911, + "grad_norm": 1.34316885471344, + "learning_rate": 4.527201490307062e-05, + "loss": 0.526, + "step": 39994 + }, + { + "epoch": 2.7753105266810074, + "grad_norm": 1.002394199371338, + "learning_rate": 4.5267312491045985e-05, + "loss": 0.268, + "step": 39995 + }, + { + "epoch": 2.7753799181181042, + "grad_norm": 1.0765622854232788, + "learning_rate": 4.5262610251806093e-05, + "loss": 0.32, + "step": 39996 + }, + { + "epoch": 2.775449309555201, + "grad_norm": 1.3915736675262451, + "learning_rate": 4.525790818536578e-05, + "loss": 0.2934, + "step": 39997 + }, + { + "epoch": 2.7755187009922975, + "grad_norm": 2.04174542427063, + "learning_rate": 4.525320629173991e-05, + "loss": 0.454, + "step": 39998 + }, + { + "epoch": 2.7755880924293943, + "grad_norm": 1.4218859672546387, + "learning_rate": 4.524850457094332e-05, + "loss": 0.569, + "step": 39999 + }, + { + "epoch": 2.7756574838664907, + "grad_norm": 1.187986135482788, + "learning_rate": 4.5243803022990883e-05, + "loss": 0.3686, + "step": 40000 + }, + { + "epoch": 2.7756574838664907, + "eval_loss": 0.49565672874450684, + "eval_runtime": 3864.3444, + "eval_samples_per_second": 6.281, + "eval_steps_per_second": 0.785, + "step": 40000 + }, + { + "epoch": 2.7757268753035875, + "grad_norm": 1.0715175867080688, + "learning_rate": 4.5239101647897366e-05, + "loss": 0.1774, + "step": 40001 + }, + { + "epoch": 2.7757962667406844, + "grad_norm": 1.2843455076217651, + "learning_rate": 4.523440044567765e-05, + "loss": 0.4263, + "step": 40002 + }, + { + "epoch": 2.7758656581777807, + "grad_norm": 1.4004173278808594, + "learning_rate": 4.5229699416346586e-05, + "loss": 0.4408, + "step": 40003 + }, + { + "epoch": 2.7759350496148776, + "grad_norm": 1.0630900859832764, + "learning_rate": 4.5224998559918996e-05, + "loss": 0.4203, + "step": 40004 + }, + { + "epoch": 2.776004441051974, + "grad_norm": 1.6728357076644897, + "learning_rate": 4.5220297876409734e-05, + "loss": 0.3377, + "step": 40005 + }, + { + "epoch": 2.776073832489071, + "grad_norm": 1.2203609943389893, + "learning_rate": 4.521559736583367e-05, + "loss": 0.3602, + "step": 40006 + }, + { + "epoch": 2.7761432239261676, + "grad_norm": 1.4426593780517578, + "learning_rate": 4.521089702820558e-05, + "loss": 0.4834, + "step": 40007 + }, + { + "epoch": 2.776212615363264, + "grad_norm": 1.2811133861541748, + "learning_rate": 4.520619686354033e-05, + "loss": 0.3309, + "step": 40008 + }, + { + "epoch": 2.776282006800361, + "grad_norm": 1.1540807485580444, + "learning_rate": 4.5201496871852754e-05, + "loss": 0.3237, + "step": 40009 + }, + { + "epoch": 2.7763513982374572, + "grad_norm": 1.0387853384017944, + "learning_rate": 4.5196797053157734e-05, + "loss": 0.4489, + "step": 40010 + }, + { + "epoch": 2.776420789674554, + "grad_norm": 1.2416547536849976, + "learning_rate": 4.5192097407470026e-05, + "loss": 0.4045, + "step": 40011 + }, + { + "epoch": 2.776490181111651, + "grad_norm": 1.3425172567367554, + "learning_rate": 4.518739793480451e-05, + "loss": 0.3808, + "step": 40012 + }, + { + "epoch": 2.7765595725487477, + "grad_norm": 0.862299382686615, + "learning_rate": 4.5182698635176026e-05, + "loss": 0.3052, + "step": 40013 + }, + { + "epoch": 2.776628963985844, + "grad_norm": 1.262917399406433, + "learning_rate": 4.517799950859943e-05, + "loss": 0.2863, + "step": 40014 + }, + { + "epoch": 2.776698355422941, + "grad_norm": 1.1085636615753174, + "learning_rate": 4.5173300555089506e-05, + "loss": 0.4076, + "step": 40015 + }, + { + "epoch": 2.7767677468600374, + "grad_norm": 0.910393476486206, + "learning_rate": 4.51686017746611e-05, + "loss": 0.342, + "step": 40016 + }, + { + "epoch": 2.776837138297134, + "grad_norm": 0.8309308290481567, + "learning_rate": 4.516390316732906e-05, + "loss": 0.339, + "step": 40017 + }, + { + "epoch": 2.776906529734231, + "grad_norm": 1.519287347793579, + "learning_rate": 4.5159204733108206e-05, + "loss": 0.4743, + "step": 40018 + }, + { + "epoch": 2.7769759211713274, + "grad_norm": 2.7906835079193115, + "learning_rate": 4.515450647201339e-05, + "loss": 0.3521, + "step": 40019 + }, + { + "epoch": 2.7770453126084242, + "grad_norm": 1.1586766242980957, + "learning_rate": 4.514980838405943e-05, + "loss": 0.397, + "step": 40020 + }, + { + "epoch": 2.7771147040455206, + "grad_norm": 1.446871280670166, + "learning_rate": 4.5145110469261196e-05, + "loss": 0.514, + "step": 40021 + }, + { + "epoch": 2.7771840954826175, + "grad_norm": 0.9619102478027344, + "learning_rate": 4.5140412727633454e-05, + "loss": 0.3927, + "step": 40022 + }, + { + "epoch": 2.7772534869197143, + "grad_norm": 1.8870515823364258, + "learning_rate": 4.5135715159191064e-05, + "loss": 0.4796, + "step": 40023 + }, + { + "epoch": 2.7773228783568107, + "grad_norm": 1.2580565214157104, + "learning_rate": 4.513101776394886e-05, + "loss": 0.4215, + "step": 40024 + }, + { + "epoch": 2.7773922697939075, + "grad_norm": 0.9513170719146729, + "learning_rate": 4.5126320541921654e-05, + "loss": 0.3955, + "step": 40025 + }, + { + "epoch": 2.777461661231004, + "grad_norm": 0.9778249263763428, + "learning_rate": 4.512162349312429e-05, + "loss": 0.3394, + "step": 40026 + }, + { + "epoch": 2.7775310526681007, + "grad_norm": 1.5045804977416992, + "learning_rate": 4.51169266175716e-05, + "loss": 0.5703, + "step": 40027 + }, + { + "epoch": 2.7776004441051976, + "grad_norm": 1.1067405939102173, + "learning_rate": 4.511222991527844e-05, + "loss": 0.2982, + "step": 40028 + }, + { + "epoch": 2.777669835542294, + "grad_norm": 0.9618380069732666, + "learning_rate": 4.510753338625956e-05, + "loss": 0.2908, + "step": 40029 + }, + { + "epoch": 2.777739226979391, + "grad_norm": 1.1804372072219849, + "learning_rate": 4.510283703052983e-05, + "loss": 0.2404, + "step": 40030 + }, + { + "epoch": 2.777808618416487, + "grad_norm": 1.2339850664138794, + "learning_rate": 4.5098140848104075e-05, + "loss": 0.4271, + "step": 40031 + }, + { + "epoch": 2.777878009853584, + "grad_norm": 0.7494481205940247, + "learning_rate": 4.509344483899712e-05, + "loss": 0.1891, + "step": 40032 + }, + { + "epoch": 2.777947401290681, + "grad_norm": 1.398508071899414, + "learning_rate": 4.5088749003223795e-05, + "loss": 0.3288, + "step": 40033 + }, + { + "epoch": 2.7780167927277772, + "grad_norm": 1.333851933479309, + "learning_rate": 4.508405334079892e-05, + "loss": 0.4413, + "step": 40034 + }, + { + "epoch": 2.778086184164874, + "grad_norm": 1.1875818967819214, + "learning_rate": 4.5079357851737344e-05, + "loss": 0.4137, + "step": 40035 + }, + { + "epoch": 2.7781555756019705, + "grad_norm": 1.2428390979766846, + "learning_rate": 4.507466253605383e-05, + "loss": 0.3995, + "step": 40036 + }, + { + "epoch": 2.7782249670390673, + "grad_norm": 1.0676119327545166, + "learning_rate": 4.506996739376323e-05, + "loss": 0.3683, + "step": 40037 + }, + { + "epoch": 2.778294358476164, + "grad_norm": 1.2175239324569702, + "learning_rate": 4.506527242488038e-05, + "loss": 0.4193, + "step": 40038 + }, + { + "epoch": 2.778363749913261, + "grad_norm": 0.8542088270187378, + "learning_rate": 4.5060577629420084e-05, + "loss": 0.2676, + "step": 40039 + }, + { + "epoch": 2.7784331413503573, + "grad_norm": 1.4550007581710815, + "learning_rate": 4.505588300739717e-05, + "loss": 0.6359, + "step": 40040 + }, + { + "epoch": 2.778502532787454, + "grad_norm": 1.4772262573242188, + "learning_rate": 4.50511885588265e-05, + "loss": 0.3496, + "step": 40041 + }, + { + "epoch": 2.7785719242245506, + "grad_norm": 1.610063076019287, + "learning_rate": 4.5046494283722807e-05, + "loss": 0.5357, + "step": 40042 + }, + { + "epoch": 2.7786413156616474, + "grad_norm": 1.0143002271652222, + "learning_rate": 4.504180018210097e-05, + "loss": 0.3379, + "step": 40043 + }, + { + "epoch": 2.7787107070987442, + "grad_norm": 1.2167431116104126, + "learning_rate": 4.503710625397578e-05, + "loss": 0.3805, + "step": 40044 + }, + { + "epoch": 2.7787800985358406, + "grad_norm": 0.9736671447753906, + "learning_rate": 4.5032412499362106e-05, + "loss": 0.4295, + "step": 40045 + }, + { + "epoch": 2.7788494899729375, + "grad_norm": 1.3306891918182373, + "learning_rate": 4.50277189182747e-05, + "loss": 0.5138, + "step": 40046 + }, + { + "epoch": 2.778918881410034, + "grad_norm": 1.3608065843582153, + "learning_rate": 4.50230255107284e-05, + "loss": 0.4469, + "step": 40047 + }, + { + "epoch": 2.7789882728471307, + "grad_norm": 1.3854641914367676, + "learning_rate": 4.501833227673804e-05, + "loss": 0.3415, + "step": 40048 + }, + { + "epoch": 2.7790576642842275, + "grad_norm": 1.038961410522461, + "learning_rate": 4.501363921631845e-05, + "loss": 0.4986, + "step": 40049 + }, + { + "epoch": 2.779127055721324, + "grad_norm": 1.577564001083374, + "learning_rate": 4.500894632948439e-05, + "loss": 0.5513, + "step": 40050 + }, + { + "epoch": 2.7791964471584207, + "grad_norm": 1.2417734861373901, + "learning_rate": 4.500425361625071e-05, + "loss": 0.3926, + "step": 40051 + }, + { + "epoch": 2.779265838595517, + "grad_norm": 0.9623902440071106, + "learning_rate": 4.4999561076632225e-05, + "loss": 0.3325, + "step": 40052 + }, + { + "epoch": 2.779335230032614, + "grad_norm": 1.1084039211273193, + "learning_rate": 4.4994868710643734e-05, + "loss": 0.3129, + "step": 40053 + }, + { + "epoch": 2.779404621469711, + "grad_norm": 1.2130250930786133, + "learning_rate": 4.4990176518300066e-05, + "loss": 0.3509, + "step": 40054 + }, + { + "epoch": 2.779474012906807, + "grad_norm": 1.1328802108764648, + "learning_rate": 4.4985484499616015e-05, + "loss": 0.3311, + "step": 40055 + }, + { + "epoch": 2.779543404343904, + "grad_norm": 0.780912458896637, + "learning_rate": 4.498079265460646e-05, + "loss": 0.3251, + "step": 40056 + }, + { + "epoch": 2.7796127957810004, + "grad_norm": 1.286026120185852, + "learning_rate": 4.4976100983286116e-05, + "loss": 0.4295, + "step": 40057 + }, + { + "epoch": 2.7796821872180972, + "grad_norm": 0.9214844703674316, + "learning_rate": 4.497140948566983e-05, + "loss": 0.2305, + "step": 40058 + }, + { + "epoch": 2.779751578655194, + "grad_norm": 0.8822916746139526, + "learning_rate": 4.496671816177244e-05, + "loss": 0.2921, + "step": 40059 + }, + { + "epoch": 2.7798209700922905, + "grad_norm": 1.333644151687622, + "learning_rate": 4.4962027011608723e-05, + "loss": 0.4789, + "step": 40060 + }, + { + "epoch": 2.7798903615293873, + "grad_norm": 0.6903641223907471, + "learning_rate": 4.4957336035193496e-05, + "loss": 0.2538, + "step": 40061 + }, + { + "epoch": 2.7799597529664837, + "grad_norm": 1.4749201536178589, + "learning_rate": 4.4952645232541576e-05, + "loss": 0.39, + "step": 40062 + }, + { + "epoch": 2.7800291444035805, + "grad_norm": 1.4416981935501099, + "learning_rate": 4.4947954603667816e-05, + "loss": 0.4712, + "step": 40063 + }, + { + "epoch": 2.7800985358406773, + "grad_norm": 0.7508521676063538, + "learning_rate": 4.494326414858693e-05, + "loss": 0.3142, + "step": 40064 + }, + { + "epoch": 2.780167927277774, + "grad_norm": 1.0360006093978882, + "learning_rate": 4.493857386731377e-05, + "loss": 0.4143, + "step": 40065 + }, + { + "epoch": 2.7802373187148706, + "grad_norm": 1.0645452737808228, + "learning_rate": 4.493388375986315e-05, + "loss": 0.3965, + "step": 40066 + }, + { + "epoch": 2.7803067101519674, + "grad_norm": 1.5776396989822388, + "learning_rate": 4.492919382624987e-05, + "loss": 0.3365, + "step": 40067 + }, + { + "epoch": 2.780376101589064, + "grad_norm": 1.5351247787475586, + "learning_rate": 4.492450406648873e-05, + "loss": 0.3804, + "step": 40068 + }, + { + "epoch": 2.7804454930261606, + "grad_norm": 1.2722735404968262, + "learning_rate": 4.491981448059458e-05, + "loss": 0.3123, + "step": 40069 + }, + { + "epoch": 2.7805148844632575, + "grad_norm": 1.1433361768722534, + "learning_rate": 4.4915125068582145e-05, + "loss": 0.2858, + "step": 40070 + }, + { + "epoch": 2.780584275900354, + "grad_norm": 0.9944037199020386, + "learning_rate": 4.491043583046627e-05, + "loss": 0.2691, + "step": 40071 + }, + { + "epoch": 2.7806536673374507, + "grad_norm": 0.9913355112075806, + "learning_rate": 4.490574676626177e-05, + "loss": 0.4122, + "step": 40072 + }, + { + "epoch": 2.780723058774547, + "grad_norm": 1.335675597190857, + "learning_rate": 4.490105787598342e-05, + "loss": 0.3719, + "step": 40073 + }, + { + "epoch": 2.780792450211644, + "grad_norm": 0.929878294467926, + "learning_rate": 4.489636915964605e-05, + "loss": 0.2664, + "step": 40074 + }, + { + "epoch": 2.7808618416487407, + "grad_norm": 1.1686437129974365, + "learning_rate": 4.489168061726443e-05, + "loss": 0.3803, + "step": 40075 + }, + { + "epoch": 2.780931233085837, + "grad_norm": 1.3185298442840576, + "learning_rate": 4.488699224885343e-05, + "loss": 0.4337, + "step": 40076 + }, + { + "epoch": 2.781000624522934, + "grad_norm": 0.9118351936340332, + "learning_rate": 4.488230405442777e-05, + "loss": 0.3804, + "step": 40077 + }, + { + "epoch": 2.7810700159600303, + "grad_norm": 0.7687956690788269, + "learning_rate": 4.487761603400227e-05, + "loss": 0.2151, + "step": 40078 + }, + { + "epoch": 2.781139407397127, + "grad_norm": 1.2338858842849731, + "learning_rate": 4.487292818759179e-05, + "loss": 0.2526, + "step": 40079 + }, + { + "epoch": 2.781208798834224, + "grad_norm": 1.1129525899887085, + "learning_rate": 4.486824051521104e-05, + "loss": 0.4312, + "step": 40080 + }, + { + "epoch": 2.7812781902713204, + "grad_norm": 1.1591920852661133, + "learning_rate": 4.486355301687485e-05, + "loss": 0.244, + "step": 40081 + }, + { + "epoch": 2.7813475817084172, + "grad_norm": 1.4958490133285522, + "learning_rate": 4.485886569259803e-05, + "loss": 0.3488, + "step": 40082 + }, + { + "epoch": 2.7814169731455136, + "grad_norm": 1.554514765739441, + "learning_rate": 4.485417854239541e-05, + "loss": 0.2929, + "step": 40083 + }, + { + "epoch": 2.7814863645826104, + "grad_norm": 1.167047142982483, + "learning_rate": 4.484949156628171e-05, + "loss": 0.4441, + "step": 40084 + }, + { + "epoch": 2.7815557560197073, + "grad_norm": 1.0246731042861938, + "learning_rate": 4.4844804764271776e-05, + "loss": 0.3262, + "step": 40085 + }, + { + "epoch": 2.7816251474568037, + "grad_norm": 1.0330626964569092, + "learning_rate": 4.484011813638038e-05, + "loss": 0.4007, + "step": 40086 + }, + { + "epoch": 2.7816945388939005, + "grad_norm": 1.2768856287002563, + "learning_rate": 4.483543168262234e-05, + "loss": 0.3889, + "step": 40087 + }, + { + "epoch": 2.781763930330997, + "grad_norm": 1.3117908239364624, + "learning_rate": 4.483074540301244e-05, + "loss": 0.3364, + "step": 40088 + }, + { + "epoch": 2.7818333217680937, + "grad_norm": 1.035269021987915, + "learning_rate": 4.482605929756548e-05, + "loss": 0.3658, + "step": 40089 + }, + { + "epoch": 2.7819027132051906, + "grad_norm": 0.9185494780540466, + "learning_rate": 4.4821373366296285e-05, + "loss": 0.32, + "step": 40090 + }, + { + "epoch": 2.7819721046422874, + "grad_norm": 1.0807496309280396, + "learning_rate": 4.4816687609219576e-05, + "loss": 0.3294, + "step": 40091 + }, + { + "epoch": 2.782041496079384, + "grad_norm": 0.9612891674041748, + "learning_rate": 4.481200202635019e-05, + "loss": 0.3815, + "step": 40092 + }, + { + "epoch": 2.7821108875164806, + "grad_norm": 1.2508989572525024, + "learning_rate": 4.480731661770289e-05, + "loss": 0.4746, + "step": 40093 + }, + { + "epoch": 2.782180278953577, + "grad_norm": 1.1379081010818481, + "learning_rate": 4.480263138329251e-05, + "loss": 0.4309, + "step": 40094 + }, + { + "epoch": 2.782249670390674, + "grad_norm": 1.3067790269851685, + "learning_rate": 4.47979463231338e-05, + "loss": 0.5403, + "step": 40095 + }, + { + "epoch": 2.7823190618277707, + "grad_norm": 1.1921327114105225, + "learning_rate": 4.4793261437241575e-05, + "loss": 0.348, + "step": 40096 + }, + { + "epoch": 2.782388453264867, + "grad_norm": 2.0457704067230225, + "learning_rate": 4.478857672563066e-05, + "loss": 0.4467, + "step": 40097 + }, + { + "epoch": 2.782457844701964, + "grad_norm": 1.3523560762405396, + "learning_rate": 4.4783892188315754e-05, + "loss": 0.4591, + "step": 40098 + }, + { + "epoch": 2.7825272361390603, + "grad_norm": 0.766789972782135, + "learning_rate": 4.477920782531171e-05, + "loss": 0.1968, + "step": 40099 + }, + { + "epoch": 2.782596627576157, + "grad_norm": 1.626356601715088, + "learning_rate": 4.477452363663329e-05, + "loss": 0.5571, + "step": 40100 + }, + { + "epoch": 2.782666019013254, + "grad_norm": 3.2354719638824463, + "learning_rate": 4.476983962229529e-05, + "loss": 0.4027, + "step": 40101 + }, + { + "epoch": 2.7827354104503503, + "grad_norm": 1.0077272653579712, + "learning_rate": 4.476515578231251e-05, + "loss": 0.2793, + "step": 40102 + }, + { + "epoch": 2.782804801887447, + "grad_norm": 1.3975625038146973, + "learning_rate": 4.4760472116699717e-05, + "loss": 0.3251, + "step": 40103 + }, + { + "epoch": 2.7828741933245436, + "grad_norm": 1.6310011148452759, + "learning_rate": 4.475578862547174e-05, + "loss": 0.6565, + "step": 40104 + }, + { + "epoch": 2.7829435847616404, + "grad_norm": 1.0291600227355957, + "learning_rate": 4.4751105308643295e-05, + "loss": 0.5184, + "step": 40105 + }, + { + "epoch": 2.783012976198737, + "grad_norm": 1.166709303855896, + "learning_rate": 4.4746422166229195e-05, + "loss": 0.4096, + "step": 40106 + }, + { + "epoch": 2.7830823676358336, + "grad_norm": 2.004326343536377, + "learning_rate": 4.474173919824424e-05, + "loss": 0.7599, + "step": 40107 + }, + { + "epoch": 2.7831517590729304, + "grad_norm": 1.991735816001892, + "learning_rate": 4.473705640470319e-05, + "loss": 0.3985, + "step": 40108 + }, + { + "epoch": 2.783221150510027, + "grad_norm": 1.4848564863204956, + "learning_rate": 4.4732373785620855e-05, + "loss": 0.6798, + "step": 40109 + }, + { + "epoch": 2.7832905419471237, + "grad_norm": 1.6627272367477417, + "learning_rate": 4.472769134101199e-05, + "loss": 0.4177, + "step": 40110 + }, + { + "epoch": 2.7833599333842205, + "grad_norm": 1.0601950883865356, + "learning_rate": 4.472300907089143e-05, + "loss": 0.3609, + "step": 40111 + }, + { + "epoch": 2.783429324821317, + "grad_norm": 1.1617863178253174, + "learning_rate": 4.47183269752739e-05, + "loss": 0.3884, + "step": 40112 + }, + { + "epoch": 2.7834987162584137, + "grad_norm": 1.3448927402496338, + "learning_rate": 4.471364505417417e-05, + "loss": 0.3436, + "step": 40113 + }, + { + "epoch": 2.78356810769551, + "grad_norm": 1.1493096351623535, + "learning_rate": 4.4708963307607096e-05, + "loss": 0.6336, + "step": 40114 + }, + { + "epoch": 2.783637499132607, + "grad_norm": 0.9176126718521118, + "learning_rate": 4.4704281735587385e-05, + "loss": 0.3513, + "step": 40115 + }, + { + "epoch": 2.7837068905697038, + "grad_norm": 0.8616272211074829, + "learning_rate": 4.469960033812982e-05, + "loss": 0.3884, + "step": 40116 + }, + { + "epoch": 2.7837762820068006, + "grad_norm": 1.042289137840271, + "learning_rate": 4.469491911524921e-05, + "loss": 0.4017, + "step": 40117 + }, + { + "epoch": 2.783845673443897, + "grad_norm": 1.7598657608032227, + "learning_rate": 4.4690238066960365e-05, + "loss": 0.4888, + "step": 40118 + }, + { + "epoch": 2.783915064880994, + "grad_norm": 1.1520919799804688, + "learning_rate": 4.4685557193277984e-05, + "loss": 0.419, + "step": 40119 + }, + { + "epoch": 2.78398445631809, + "grad_norm": 1.5232638120651245, + "learning_rate": 4.468087649421687e-05, + "loss": 0.5219, + "step": 40120 + }, + { + "epoch": 2.784053847755187, + "grad_norm": 1.1036059856414795, + "learning_rate": 4.4676195969791824e-05, + "loss": 0.4063, + "step": 40121 + }, + { + "epoch": 2.784123239192284, + "grad_norm": 0.8286070823669434, + "learning_rate": 4.467151562001761e-05, + "loss": 0.3112, + "step": 40122 + }, + { + "epoch": 2.7841926306293803, + "grad_norm": 1.4367754459381104, + "learning_rate": 4.4666835444908995e-05, + "loss": 0.4573, + "step": 40123 + }, + { + "epoch": 2.784262022066477, + "grad_norm": 1.0850454568862915, + "learning_rate": 4.466215544448076e-05, + "loss": 0.3398, + "step": 40124 + }, + { + "epoch": 2.7843314135035735, + "grad_norm": 0.8676068782806396, + "learning_rate": 4.465747561874771e-05, + "loss": 0.3142, + "step": 40125 + }, + { + "epoch": 2.7844008049406703, + "grad_norm": 0.8916944265365601, + "learning_rate": 4.465279596772457e-05, + "loss": 0.2457, + "step": 40126 + }, + { + "epoch": 2.784470196377767, + "grad_norm": 1.4173848628997803, + "learning_rate": 4.4648116491426114e-05, + "loss": 0.4137, + "step": 40127 + }, + { + "epoch": 2.7845395878148635, + "grad_norm": 0.8277060985565186, + "learning_rate": 4.4643437189867146e-05, + "loss": 0.3037, + "step": 40128 + }, + { + "epoch": 2.7846089792519604, + "grad_norm": 1.2654070854187012, + "learning_rate": 4.463875806306241e-05, + "loss": 0.3943, + "step": 40129 + }, + { + "epoch": 2.7846783706890568, + "grad_norm": 1.3090481758117676, + "learning_rate": 4.46340791110267e-05, + "loss": 0.4662, + "step": 40130 + }, + { + "epoch": 2.7847477621261536, + "grad_norm": 0.9660142064094543, + "learning_rate": 4.4629400333774786e-05, + "loss": 0.3223, + "step": 40131 + }, + { + "epoch": 2.7848171535632504, + "grad_norm": 0.9054101705551147, + "learning_rate": 4.462472173132146e-05, + "loss": 0.3845, + "step": 40132 + }, + { + "epoch": 2.784886545000347, + "grad_norm": 1.1071704626083374, + "learning_rate": 4.4620043303681424e-05, + "loss": 0.4404, + "step": 40133 + }, + { + "epoch": 2.7849559364374437, + "grad_norm": 0.9541314840316772, + "learning_rate": 4.461536505086948e-05, + "loss": 0.3311, + "step": 40134 + }, + { + "epoch": 2.78502532787454, + "grad_norm": 1.4335066080093384, + "learning_rate": 4.4610686972900415e-05, + "loss": 0.4242, + "step": 40135 + }, + { + "epoch": 2.785094719311637, + "grad_norm": 1.5273795127868652, + "learning_rate": 4.460600906978899e-05, + "loss": 0.3912, + "step": 40136 + }, + { + "epoch": 2.7851641107487337, + "grad_norm": 1.0650594234466553, + "learning_rate": 4.460133134154996e-05, + "loss": 0.2964, + "step": 40137 + }, + { + "epoch": 2.7852335021858305, + "grad_norm": 1.6543997526168823, + "learning_rate": 4.45966537881981e-05, + "loss": 0.3954, + "step": 40138 + }, + { + "epoch": 2.785302893622927, + "grad_norm": 1.6969040632247925, + "learning_rate": 4.45919764097482e-05, + "loss": 0.3691, + "step": 40139 + }, + { + "epoch": 2.7853722850600233, + "grad_norm": 1.1230857372283936, + "learning_rate": 4.4587299206214985e-05, + "loss": 0.3254, + "step": 40140 + }, + { + "epoch": 2.78544167649712, + "grad_norm": 1.1520484685897827, + "learning_rate": 4.458262217761322e-05, + "loss": 0.3203, + "step": 40141 + }, + { + "epoch": 2.785511067934217, + "grad_norm": 1.1496795415878296, + "learning_rate": 4.4577945323957693e-05, + "loss": 0.3144, + "step": 40142 + }, + { + "epoch": 2.785580459371314, + "grad_norm": 1.2394903898239136, + "learning_rate": 4.457326864526317e-05, + "loss": 0.2754, + "step": 40143 + }, + { + "epoch": 2.78564985080841, + "grad_norm": 1.2700464725494385, + "learning_rate": 4.456859214154439e-05, + "loss": 0.3435, + "step": 40144 + }, + { + "epoch": 2.785719242245507, + "grad_norm": 0.9717888236045837, + "learning_rate": 4.4563915812816194e-05, + "loss": 0.5553, + "step": 40145 + }, + { + "epoch": 2.7857886336826034, + "grad_norm": 1.5540978908538818, + "learning_rate": 4.4559239659093235e-05, + "loss": 0.4267, + "step": 40146 + }, + { + "epoch": 2.7858580251197003, + "grad_norm": 1.2690118551254272, + "learning_rate": 4.455456368039033e-05, + "loss": 0.5402, + "step": 40147 + }, + { + "epoch": 2.785927416556797, + "grad_norm": 1.2990798950195312, + "learning_rate": 4.454988787672222e-05, + "loss": 0.3831, + "step": 40148 + }, + { + "epoch": 2.7859968079938935, + "grad_norm": 1.1088402271270752, + "learning_rate": 4.454521224810373e-05, + "loss": 0.4347, + "step": 40149 + }, + { + "epoch": 2.7860661994309903, + "grad_norm": 0.6981542110443115, + "learning_rate": 4.454053679454953e-05, + "loss": 0.2874, + "step": 40150 + }, + { + "epoch": 2.7861355908680867, + "grad_norm": 1.0519959926605225, + "learning_rate": 4.4535861516074415e-05, + "loss": 0.2912, + "step": 40151 + }, + { + "epoch": 2.7862049823051835, + "grad_norm": 1.0080899000167847, + "learning_rate": 4.453118641269316e-05, + "loss": 0.3759, + "step": 40152 + }, + { + "epoch": 2.7862743737422804, + "grad_norm": 1.1925089359283447, + "learning_rate": 4.452651148442055e-05, + "loss": 0.4095, + "step": 40153 + }, + { + "epoch": 2.7863437651793768, + "grad_norm": 0.8057091236114502, + "learning_rate": 4.452183673127126e-05, + "loss": 0.3164, + "step": 40154 + }, + { + "epoch": 2.7864131566164736, + "grad_norm": 1.4215749502182007, + "learning_rate": 4.4517162153260106e-05, + "loss": 0.4897, + "step": 40155 + }, + { + "epoch": 2.78648254805357, + "grad_norm": 1.4065284729003906, + "learning_rate": 4.451248775040183e-05, + "loss": 0.4892, + "step": 40156 + }, + { + "epoch": 2.786551939490667, + "grad_norm": 0.8933526277542114, + "learning_rate": 4.45078135227112e-05, + "loss": 0.2591, + "step": 40157 + }, + { + "epoch": 2.7866213309277637, + "grad_norm": 1.620565414428711, + "learning_rate": 4.450313947020295e-05, + "loss": 0.4615, + "step": 40158 + }, + { + "epoch": 2.78669072236486, + "grad_norm": 1.3967251777648926, + "learning_rate": 4.449846559289189e-05, + "loss": 0.3142, + "step": 40159 + }, + { + "epoch": 2.786760113801957, + "grad_norm": 1.2051063776016235, + "learning_rate": 4.44937918907927e-05, + "loss": 0.2922, + "step": 40160 + }, + { + "epoch": 2.7868295052390533, + "grad_norm": 0.88448166847229, + "learning_rate": 4.448911836392018e-05, + "loss": 0.3814, + "step": 40161 + }, + { + "epoch": 2.78689889667615, + "grad_norm": 1.090152621269226, + "learning_rate": 4.4484445012289066e-05, + "loss": 0.3158, + "step": 40162 + }, + { + "epoch": 2.786968288113247, + "grad_norm": 1.0637046098709106, + "learning_rate": 4.447977183591411e-05, + "loss": 0.4841, + "step": 40163 + }, + { + "epoch": 2.7870376795503438, + "grad_norm": 1.270584225654602, + "learning_rate": 4.4475098834810084e-05, + "loss": 0.4227, + "step": 40164 + }, + { + "epoch": 2.78710707098744, + "grad_norm": 1.0603457689285278, + "learning_rate": 4.4470426008991714e-05, + "loss": 0.3351, + "step": 40165 + }, + { + "epoch": 2.7871764624245365, + "grad_norm": 1.2334450483322144, + "learning_rate": 4.446575335847382e-05, + "loss": 0.4067, + "step": 40166 + }, + { + "epoch": 2.7872458538616334, + "grad_norm": 1.3549144268035889, + "learning_rate": 4.4461080883271055e-05, + "loss": 0.4255, + "step": 40167 + }, + { + "epoch": 2.78731524529873, + "grad_norm": 0.888595700263977, + "learning_rate": 4.445640858339821e-05, + "loss": 0.3727, + "step": 40168 + }, + { + "epoch": 2.787384636735827, + "grad_norm": 1.5555223226547241, + "learning_rate": 4.445173645887004e-05, + "loss": 0.4912, + "step": 40169 + }, + { + "epoch": 2.7874540281729234, + "grad_norm": 1.0823477506637573, + "learning_rate": 4.4447064509701295e-05, + "loss": 0.4524, + "step": 40170 + }, + { + "epoch": 2.7875234196100203, + "grad_norm": 0.9130859375, + "learning_rate": 4.4442392735906736e-05, + "loss": 0.2266, + "step": 40171 + }, + { + "epoch": 2.7875928110471166, + "grad_norm": 1.116112232208252, + "learning_rate": 4.4437721137501075e-05, + "loss": 0.3591, + "step": 40172 + }, + { + "epoch": 2.7876622024842135, + "grad_norm": 0.9409789443016052, + "learning_rate": 4.443304971449913e-05, + "loss": 0.2807, + "step": 40173 + }, + { + "epoch": 2.7877315939213103, + "grad_norm": 1.080539345741272, + "learning_rate": 4.4428378466915565e-05, + "loss": 0.3419, + "step": 40174 + }, + { + "epoch": 2.7878009853584067, + "grad_norm": 1.2953699827194214, + "learning_rate": 4.442370739476516e-05, + "loss": 0.3118, + "step": 40175 + }, + { + "epoch": 2.7878703767955035, + "grad_norm": 0.998127281665802, + "learning_rate": 4.441903649806266e-05, + "loss": 0.3184, + "step": 40176 + }, + { + "epoch": 2.7879397682326, + "grad_norm": 2.0217337608337402, + "learning_rate": 4.441436577682281e-05, + "loss": 0.5847, + "step": 40177 + }, + { + "epoch": 2.7880091596696968, + "grad_norm": 1.6802866458892822, + "learning_rate": 4.4409695231060365e-05, + "loss": 0.3438, + "step": 40178 + }, + { + "epoch": 2.7880785511067936, + "grad_norm": 1.0966095924377441, + "learning_rate": 4.4405024860790055e-05, + "loss": 0.3877, + "step": 40179 + }, + { + "epoch": 2.78814794254389, + "grad_norm": 1.3299025297164917, + "learning_rate": 4.440035466602668e-05, + "loss": 0.3294, + "step": 40180 + }, + { + "epoch": 2.788217333980987, + "grad_norm": 1.6784026622772217, + "learning_rate": 4.439568464678489e-05, + "loss": 0.3717, + "step": 40181 + }, + { + "epoch": 2.788286725418083, + "grad_norm": 1.3436931371688843, + "learning_rate": 4.4391014803079466e-05, + "loss": 0.4379, + "step": 40182 + }, + { + "epoch": 2.78835611685518, + "grad_norm": 1.237011432647705, + "learning_rate": 4.438634513492517e-05, + "loss": 0.4944, + "step": 40183 + }, + { + "epoch": 2.788425508292277, + "grad_norm": 2.0015878677368164, + "learning_rate": 4.4381675642336764e-05, + "loss": 0.4509, + "step": 40184 + }, + { + "epoch": 2.7884948997293733, + "grad_norm": 1.2057770490646362, + "learning_rate": 4.4377006325328904e-05, + "loss": 0.3386, + "step": 40185 + }, + { + "epoch": 2.78856429116647, + "grad_norm": 1.7570037841796875, + "learning_rate": 4.43723371839164e-05, + "loss": 0.3582, + "step": 40186 + }, + { + "epoch": 2.7886336826035665, + "grad_norm": 1.4890248775482178, + "learning_rate": 4.436766821811397e-05, + "loss": 0.5453, + "step": 40187 + }, + { + "epoch": 2.7887030740406633, + "grad_norm": 1.2734448909759521, + "learning_rate": 4.436299942793638e-05, + "loss": 0.412, + "step": 40188 + }, + { + "epoch": 2.78877246547776, + "grad_norm": 1.2907027006149292, + "learning_rate": 4.435833081339832e-05, + "loss": 0.3669, + "step": 40189 + }, + { + "epoch": 2.788841856914857, + "grad_norm": 1.9356153011322021, + "learning_rate": 4.435366237451455e-05, + "loss": 0.4494, + "step": 40190 + }, + { + "epoch": 2.7889112483519534, + "grad_norm": 1.2389339208602905, + "learning_rate": 4.434899411129981e-05, + "loss": 0.388, + "step": 40191 + }, + { + "epoch": 2.78898063978905, + "grad_norm": 1.177941918373108, + "learning_rate": 4.434432602376885e-05, + "loss": 0.4356, + "step": 40192 + }, + { + "epoch": 2.7890500312261466, + "grad_norm": 1.3112294673919678, + "learning_rate": 4.433965811193639e-05, + "loss": 0.3668, + "step": 40193 + }, + { + "epoch": 2.7891194226632434, + "grad_norm": 1.3236913681030273, + "learning_rate": 4.433499037581721e-05, + "loss": 0.4081, + "step": 40194 + }, + { + "epoch": 2.7891888141003403, + "grad_norm": 1.2251389026641846, + "learning_rate": 4.4330322815425965e-05, + "loss": 0.3208, + "step": 40195 + }, + { + "epoch": 2.7892582055374366, + "grad_norm": 1.2403732538223267, + "learning_rate": 4.4325655430777435e-05, + "loss": 0.334, + "step": 40196 + }, + { + "epoch": 2.7893275969745335, + "grad_norm": 1.2568827867507935, + "learning_rate": 4.432098822188635e-05, + "loss": 0.3271, + "step": 40197 + }, + { + "epoch": 2.78939698841163, + "grad_norm": 1.09078049659729, + "learning_rate": 4.431632118876744e-05, + "loss": 0.3194, + "step": 40198 + }, + { + "epoch": 2.7894663798487267, + "grad_norm": 1.860666275024414, + "learning_rate": 4.4311654331435446e-05, + "loss": 0.4824, + "step": 40199 + }, + { + "epoch": 2.7895357712858235, + "grad_norm": 1.1982767581939697, + "learning_rate": 4.430698764990511e-05, + "loss": 0.4014, + "step": 40200 + }, + { + "epoch": 2.78960516272292, + "grad_norm": 1.2693076133728027, + "learning_rate": 4.430232114419117e-05, + "loss": 0.4635, + "step": 40201 + }, + { + "epoch": 2.7896745541600168, + "grad_norm": 3.2485811710357666, + "learning_rate": 4.429765481430832e-05, + "loss": 0.4081, + "step": 40202 + }, + { + "epoch": 2.789743945597113, + "grad_norm": 2.244136333465576, + "learning_rate": 4.429298866027131e-05, + "loss": 0.5986, + "step": 40203 + }, + { + "epoch": 2.78981333703421, + "grad_norm": 1.8048356771469116, + "learning_rate": 4.4288322682094864e-05, + "loss": 0.6297, + "step": 40204 + }, + { + "epoch": 2.789882728471307, + "grad_norm": 0.7604512572288513, + "learning_rate": 4.428365687979372e-05, + "loss": 0.3382, + "step": 40205 + }, + { + "epoch": 2.789952119908403, + "grad_norm": 1.2662551403045654, + "learning_rate": 4.4278991253382604e-05, + "loss": 0.4627, + "step": 40206 + }, + { + "epoch": 2.7900215113455, + "grad_norm": 1.2455471754074097, + "learning_rate": 4.427432580287626e-05, + "loss": 0.3691, + "step": 40207 + }, + { + "epoch": 2.7900909027825964, + "grad_norm": 0.7954803705215454, + "learning_rate": 4.4269660528289426e-05, + "loss": 0.2073, + "step": 40208 + }, + { + "epoch": 2.7901602942196932, + "grad_norm": 1.4378085136413574, + "learning_rate": 4.426499542963678e-05, + "loss": 0.4084, + "step": 40209 + }, + { + "epoch": 2.79022968565679, + "grad_norm": 1.0334241390228271, + "learning_rate": 4.426033050693308e-05, + "loss": 0.2316, + "step": 40210 + }, + { + "epoch": 2.7902990770938865, + "grad_norm": 0.8971740007400513, + "learning_rate": 4.4255665760193044e-05, + "loss": 0.2694, + "step": 40211 + }, + { + "epoch": 2.7903684685309833, + "grad_norm": 1.1414088010787964, + "learning_rate": 4.425100118943141e-05, + "loss": 0.5367, + "step": 40212 + }, + { + "epoch": 2.7904378599680797, + "grad_norm": 1.1482855081558228, + "learning_rate": 4.424633679466289e-05, + "loss": 0.4083, + "step": 40213 + }, + { + "epoch": 2.7905072514051765, + "grad_norm": 1.7918235063552856, + "learning_rate": 4.424167257590222e-05, + "loss": 0.5856, + "step": 40214 + }, + { + "epoch": 2.7905766428422734, + "grad_norm": 1.1768027544021606, + "learning_rate": 4.423700853316415e-05, + "loss": 0.4907, + "step": 40215 + }, + { + "epoch": 2.79064603427937, + "grad_norm": 1.0270200967788696, + "learning_rate": 4.4232344666463345e-05, + "loss": 0.2652, + "step": 40216 + }, + { + "epoch": 2.7907154257164666, + "grad_norm": 1.2340683937072754, + "learning_rate": 4.422768097581456e-05, + "loss": 0.4459, + "step": 40217 + }, + { + "epoch": 2.7907848171535634, + "grad_norm": 1.4299266338348389, + "learning_rate": 4.4223017461232506e-05, + "loss": 0.3671, + "step": 40218 + }, + { + "epoch": 2.79085420859066, + "grad_norm": 1.2496922016143799, + "learning_rate": 4.4218354122731954e-05, + "loss": 0.4241, + "step": 40219 + }, + { + "epoch": 2.7909236000277566, + "grad_norm": 1.4527966976165771, + "learning_rate": 4.421369096032755e-05, + "loss": 0.5764, + "step": 40220 + }, + { + "epoch": 2.7909929914648535, + "grad_norm": 0.9353010058403015, + "learning_rate": 4.4209027974034056e-05, + "loss": 0.3378, + "step": 40221 + }, + { + "epoch": 2.79106238290195, + "grad_norm": 1.0709737539291382, + "learning_rate": 4.4204365163866226e-05, + "loss": 0.4265, + "step": 40222 + }, + { + "epoch": 2.7911317743390467, + "grad_norm": 1.071535348892212, + "learning_rate": 4.4199702529838694e-05, + "loss": 0.3308, + "step": 40223 + }, + { + "epoch": 2.791201165776143, + "grad_norm": 1.0911047458648682, + "learning_rate": 4.419504007196623e-05, + "loss": 0.259, + "step": 40224 + }, + { + "epoch": 2.79127055721324, + "grad_norm": 1.1470155715942383, + "learning_rate": 4.419037779026356e-05, + "loss": 0.4067, + "step": 40225 + }, + { + "epoch": 2.7913399486503367, + "grad_norm": 0.9976587295532227, + "learning_rate": 4.418571568474538e-05, + "loss": 0.4019, + "step": 40226 + }, + { + "epoch": 2.791409340087433, + "grad_norm": 1.0205341577529907, + "learning_rate": 4.418105375542643e-05, + "loss": 0.1964, + "step": 40227 + }, + { + "epoch": 2.79147873152453, + "grad_norm": 1.1865097284317017, + "learning_rate": 4.4176392002321444e-05, + "loss": 0.4636, + "step": 40228 + }, + { + "epoch": 2.7915481229616264, + "grad_norm": 0.9776434302330017, + "learning_rate": 4.417173042544508e-05, + "loss": 0.3836, + "step": 40229 + }, + { + "epoch": 2.791617514398723, + "grad_norm": 1.5561482906341553, + "learning_rate": 4.416706902481209e-05, + "loss": 0.3722, + "step": 40230 + }, + { + "epoch": 2.79168690583582, + "grad_norm": 1.113502025604248, + "learning_rate": 4.416240780043718e-05, + "loss": 0.3629, + "step": 40231 + }, + { + "epoch": 2.7917562972729164, + "grad_norm": 1.1878527402877808, + "learning_rate": 4.415774675233507e-05, + "loss": 0.2531, + "step": 40232 + }, + { + "epoch": 2.7918256887100132, + "grad_norm": 1.281140923500061, + "learning_rate": 4.415308588052047e-05, + "loss": 0.3438, + "step": 40233 + }, + { + "epoch": 2.7918950801471096, + "grad_norm": 1.3518739938735962, + "learning_rate": 4.414842518500811e-05, + "loss": 0.4605, + "step": 40234 + }, + { + "epoch": 2.7919644715842065, + "grad_norm": 1.6138097047805786, + "learning_rate": 4.4143764665812726e-05, + "loss": 0.4362, + "step": 40235 + }, + { + "epoch": 2.7920338630213033, + "grad_norm": 2.3799755573272705, + "learning_rate": 4.413910432294895e-05, + "loss": 0.5968, + "step": 40236 + }, + { + "epoch": 2.7921032544583997, + "grad_norm": 1.3045202493667603, + "learning_rate": 4.413444415643155e-05, + "loss": 0.3808, + "step": 40237 + }, + { + "epoch": 2.7921726458954965, + "grad_norm": 0.9392461180686951, + "learning_rate": 4.4129784166275226e-05, + "loss": 0.3887, + "step": 40238 + }, + { + "epoch": 2.792242037332593, + "grad_norm": 0.8728974461555481, + "learning_rate": 4.412512435249469e-05, + "loss": 0.3171, + "step": 40239 + }, + { + "epoch": 2.7923114287696897, + "grad_norm": 1.2033578157424927, + "learning_rate": 4.412046471510466e-05, + "loss": 0.2378, + "step": 40240 + }, + { + "epoch": 2.7923808202067866, + "grad_norm": 1.6222018003463745, + "learning_rate": 4.411580525411983e-05, + "loss": 0.5319, + "step": 40241 + }, + { + "epoch": 2.7924502116438834, + "grad_norm": 1.1085965633392334, + "learning_rate": 4.411114596955496e-05, + "loss": 0.3903, + "step": 40242 + }, + { + "epoch": 2.79251960308098, + "grad_norm": 0.9477143287658691, + "learning_rate": 4.410648686142468e-05, + "loss": 0.3046, + "step": 40243 + }, + { + "epoch": 2.7925889945180766, + "grad_norm": 1.3199315071105957, + "learning_rate": 4.4101827929743735e-05, + "loss": 0.3821, + "step": 40244 + }, + { + "epoch": 2.792658385955173, + "grad_norm": 1.5744181871414185, + "learning_rate": 4.409716917452684e-05, + "loss": 0.5583, + "step": 40245 + }, + { + "epoch": 2.79272777739227, + "grad_norm": 1.3383454084396362, + "learning_rate": 4.4092510595788697e-05, + "loss": 0.3423, + "step": 40246 + }, + { + "epoch": 2.7927971688293667, + "grad_norm": 1.237434983253479, + "learning_rate": 4.4087852193543997e-05, + "loss": 0.7619, + "step": 40247 + }, + { + "epoch": 2.792866560266463, + "grad_norm": 1.4818929433822632, + "learning_rate": 4.408319396780747e-05, + "loss": 0.4631, + "step": 40248 + }, + { + "epoch": 2.79293595170356, + "grad_norm": 1.1264276504516602, + "learning_rate": 4.407853591859385e-05, + "loss": 0.4186, + "step": 40249 + }, + { + "epoch": 2.7930053431406563, + "grad_norm": 1.4893755912780762, + "learning_rate": 4.407387804591777e-05, + "loss": 0.2543, + "step": 40250 + }, + { + "epoch": 2.793074734577753, + "grad_norm": 1.9812209606170654, + "learning_rate": 4.406922034979396e-05, + "loss": 0.5502, + "step": 40251 + }, + { + "epoch": 2.79314412601485, + "grad_norm": 1.4749153852462769, + "learning_rate": 4.4064562830237145e-05, + "loss": 0.4838, + "step": 40252 + }, + { + "epoch": 2.7932135174519463, + "grad_norm": 1.556775450706482, + "learning_rate": 4.4059905487262e-05, + "loss": 0.4228, + "step": 40253 + }, + { + "epoch": 2.793282908889043, + "grad_norm": 1.0895944833755493, + "learning_rate": 4.405524832088329e-05, + "loss": 0.4389, + "step": 40254 + }, + { + "epoch": 2.7933523003261396, + "grad_norm": 1.0001802444458008, + "learning_rate": 4.405059133111563e-05, + "loss": 0.377, + "step": 40255 + }, + { + "epoch": 2.7934216917632364, + "grad_norm": 1.0954781770706177, + "learning_rate": 4.4045934517973764e-05, + "loss": 0.3124, + "step": 40256 + }, + { + "epoch": 2.7934910832003332, + "grad_norm": 1.5162708759307861, + "learning_rate": 4.404127788147242e-05, + "loss": 0.4338, + "step": 40257 + }, + { + "epoch": 2.7935604746374296, + "grad_norm": 1.143547534942627, + "learning_rate": 4.4036621421626244e-05, + "loss": 0.2937, + "step": 40258 + }, + { + "epoch": 2.7936298660745265, + "grad_norm": 0.9388720989227295, + "learning_rate": 4.4031965138449946e-05, + "loss": 0.435, + "step": 40259 + }, + { + "epoch": 2.793699257511623, + "grad_norm": 0.9773024320602417, + "learning_rate": 4.402730903195825e-05, + "loss": 0.241, + "step": 40260 + }, + { + "epoch": 2.7937686489487197, + "grad_norm": 1.1546239852905273, + "learning_rate": 4.4022653102165854e-05, + "loss": 0.3284, + "step": 40261 + }, + { + "epoch": 2.7938380403858165, + "grad_norm": 1.5099897384643555, + "learning_rate": 4.4017997349087434e-05, + "loss": 0.4409, + "step": 40262 + }, + { + "epoch": 2.793907431822913, + "grad_norm": 1.3959437608718872, + "learning_rate": 4.401334177273775e-05, + "loss": 0.4285, + "step": 40263 + }, + { + "epoch": 2.7939768232600097, + "grad_norm": 1.3281699419021606, + "learning_rate": 4.40086863731314e-05, + "loss": 0.5252, + "step": 40264 + }, + { + "epoch": 2.794046214697106, + "grad_norm": 1.2733715772628784, + "learning_rate": 4.4004031150283146e-05, + "loss": 0.4124, + "step": 40265 + }, + { + "epoch": 2.794115606134203, + "grad_norm": 1.3735291957855225, + "learning_rate": 4.399937610420766e-05, + "loss": 0.4189, + "step": 40266 + }, + { + "epoch": 2.7941849975713, + "grad_norm": 1.0981348752975464, + "learning_rate": 4.3994721234919646e-05, + "loss": 0.4197, + "step": 40267 + }, + { + "epoch": 2.7942543890083966, + "grad_norm": 1.1885898113250732, + "learning_rate": 4.399006654243379e-05, + "loss": 0.2772, + "step": 40268 + }, + { + "epoch": 2.794323780445493, + "grad_norm": 1.5296238660812378, + "learning_rate": 4.398541202676482e-05, + "loss": 0.4452, + "step": 40269 + }, + { + "epoch": 2.79439317188259, + "grad_norm": 1.2050021886825562, + "learning_rate": 4.3980757687927423e-05, + "loss": 0.3338, + "step": 40270 + }, + { + "epoch": 2.7944625633196862, + "grad_norm": 1.5459411144256592, + "learning_rate": 4.3976103525936244e-05, + "loss": 0.5143, + "step": 40271 + }, + { + "epoch": 2.794531954756783, + "grad_norm": 1.0045140981674194, + "learning_rate": 4.397144954080601e-05, + "loss": 0.2497, + "step": 40272 + }, + { + "epoch": 2.79460134619388, + "grad_norm": 0.85909503698349, + "learning_rate": 4.39667957325514e-05, + "loss": 0.3206, + "step": 40273 + }, + { + "epoch": 2.7946707376309763, + "grad_norm": 1.4139671325683594, + "learning_rate": 4.3962142101187124e-05, + "loss": 0.4153, + "step": 40274 + }, + { + "epoch": 2.794740129068073, + "grad_norm": 1.286495327949524, + "learning_rate": 4.395748864672785e-05, + "loss": 0.4458, + "step": 40275 + }, + { + "epoch": 2.7948095205051695, + "grad_norm": 0.8029871582984924, + "learning_rate": 4.3952835369188295e-05, + "loss": 0.2601, + "step": 40276 + }, + { + "epoch": 2.7948789119422663, + "grad_norm": 1.1291316747665405, + "learning_rate": 4.394818226858317e-05, + "loss": 0.2694, + "step": 40277 + }, + { + "epoch": 2.794948303379363, + "grad_norm": 1.322532296180725, + "learning_rate": 4.394352934492709e-05, + "loss": 0.3526, + "step": 40278 + }, + { + "epoch": 2.7950176948164596, + "grad_norm": 1.1533071994781494, + "learning_rate": 4.393887659823479e-05, + "loss": 0.4187, + "step": 40279 + }, + { + "epoch": 2.7950870862535564, + "grad_norm": 1.2977702617645264, + "learning_rate": 4.3934224028520944e-05, + "loss": 0.3742, + "step": 40280 + }, + { + "epoch": 2.795156477690653, + "grad_norm": 1.4119833707809448, + "learning_rate": 4.392957163580025e-05, + "loss": 0.4005, + "step": 40281 + }, + { + "epoch": 2.7952258691277496, + "grad_norm": 1.2818078994750977, + "learning_rate": 4.3924919420087395e-05, + "loss": 0.3701, + "step": 40282 + }, + { + "epoch": 2.7952952605648465, + "grad_norm": 1.2531381845474243, + "learning_rate": 4.392026738139707e-05, + "loss": 0.3967, + "step": 40283 + }, + { + "epoch": 2.795364652001943, + "grad_norm": 0.7455248832702637, + "learning_rate": 4.391561551974398e-05, + "loss": 0.254, + "step": 40284 + }, + { + "epoch": 2.7954340434390397, + "grad_norm": 1.5412728786468506, + "learning_rate": 4.391096383514276e-05, + "loss": 0.5064, + "step": 40285 + }, + { + "epoch": 2.795503434876136, + "grad_norm": 1.1858909130096436, + "learning_rate": 4.3906312327608115e-05, + "loss": 0.4622, + "step": 40286 + }, + { + "epoch": 2.795572826313233, + "grad_norm": 1.9613333940505981, + "learning_rate": 4.390166099715473e-05, + "loss": 0.4908, + "step": 40287 + }, + { + "epoch": 2.7956422177503297, + "grad_norm": 1.2519229650497437, + "learning_rate": 4.389700984379729e-05, + "loss": 0.4032, + "step": 40288 + }, + { + "epoch": 2.795711609187426, + "grad_norm": 1.3946253061294556, + "learning_rate": 4.389235886755052e-05, + "loss": 0.4463, + "step": 40289 + }, + { + "epoch": 2.795781000624523, + "grad_norm": 1.5758647918701172, + "learning_rate": 4.388770806842903e-05, + "loss": 0.3478, + "step": 40290 + }, + { + "epoch": 2.7958503920616193, + "grad_norm": 1.224314570426941, + "learning_rate": 4.388305744644754e-05, + "loss": 0.4605, + "step": 40291 + }, + { + "epoch": 2.795919783498716, + "grad_norm": 1.2419995069503784, + "learning_rate": 4.3878407001620756e-05, + "loss": 0.4107, + "step": 40292 + }, + { + "epoch": 2.795989174935813, + "grad_norm": 1.010604977607727, + "learning_rate": 4.387375673396329e-05, + "loss": 0.298, + "step": 40293 + }, + { + "epoch": 2.79605856637291, + "grad_norm": 1.4832661151885986, + "learning_rate": 4.386910664348988e-05, + "loss": 0.5066, + "step": 40294 + }, + { + "epoch": 2.7961279578100062, + "grad_norm": 1.4295654296875, + "learning_rate": 4.3864456730215164e-05, + "loss": 0.2877, + "step": 40295 + }, + { + "epoch": 2.796197349247103, + "grad_norm": 0.9314865469932556, + "learning_rate": 4.385980699415386e-05, + "loss": 0.3483, + "step": 40296 + }, + { + "epoch": 2.7962667406841994, + "grad_norm": 1.3410171270370483, + "learning_rate": 4.385515743532066e-05, + "loss": 0.3591, + "step": 40297 + }, + { + "epoch": 2.7963361321212963, + "grad_norm": 1.6778093576431274, + "learning_rate": 4.385050805373019e-05, + "loss": 0.5025, + "step": 40298 + }, + { + "epoch": 2.796405523558393, + "grad_norm": 1.218058705329895, + "learning_rate": 4.3845858849397136e-05, + "loss": 0.2592, + "step": 40299 + }, + { + "epoch": 2.7964749149954895, + "grad_norm": 1.8632944822311401, + "learning_rate": 4.3841209822336205e-05, + "loss": 0.319, + "step": 40300 + }, + { + "epoch": 2.7965443064325863, + "grad_norm": 1.3868221044540405, + "learning_rate": 4.383656097256205e-05, + "loss": 0.3713, + "step": 40301 + }, + { + "epoch": 2.7966136978696827, + "grad_norm": 1.1716878414154053, + "learning_rate": 4.383191230008935e-05, + "loss": 0.5027, + "step": 40302 + }, + { + "epoch": 2.7966830893067796, + "grad_norm": 1.2659746408462524, + "learning_rate": 4.382726380493279e-05, + "loss": 0.4838, + "step": 40303 + }, + { + "epoch": 2.7967524807438764, + "grad_norm": 0.9249099493026733, + "learning_rate": 4.382261548710709e-05, + "loss": 0.3615, + "step": 40304 + }, + { + "epoch": 2.796821872180973, + "grad_norm": 0.81572026014328, + "learning_rate": 4.381796734662683e-05, + "loss": 0.2519, + "step": 40305 + }, + { + "epoch": 2.7968912636180696, + "grad_norm": 0.9740525484085083, + "learning_rate": 4.381331938350673e-05, + "loss": 0.2966, + "step": 40306 + }, + { + "epoch": 2.796960655055166, + "grad_norm": 1.1193368434906006, + "learning_rate": 4.3808671597761465e-05, + "loss": 0.4469, + "step": 40307 + }, + { + "epoch": 2.797030046492263, + "grad_norm": 1.4508073329925537, + "learning_rate": 4.380402398940569e-05, + "loss": 0.3578, + "step": 40308 + }, + { + "epoch": 2.7970994379293597, + "grad_norm": 1.193308711051941, + "learning_rate": 4.379937655845411e-05, + "loss": 0.5586, + "step": 40309 + }, + { + "epoch": 2.797168829366456, + "grad_norm": 1.0681551694869995, + "learning_rate": 4.3794729304921367e-05, + "loss": 0.336, + "step": 40310 + }, + { + "epoch": 2.797238220803553, + "grad_norm": 1.074489951133728, + "learning_rate": 4.379008222882218e-05, + "loss": 0.2813, + "step": 40311 + }, + { + "epoch": 2.7973076122406493, + "grad_norm": 1.312034249305725, + "learning_rate": 4.3785435330171144e-05, + "loss": 0.4608, + "step": 40312 + }, + { + "epoch": 2.797377003677746, + "grad_norm": 1.1707342863082886, + "learning_rate": 4.378078860898297e-05, + "loss": 0.3341, + "step": 40313 + }, + { + "epoch": 2.797446395114843, + "grad_norm": 1.1341333389282227, + "learning_rate": 4.377614206527233e-05, + "loss": 0.4054, + "step": 40314 + }, + { + "epoch": 2.7975157865519393, + "grad_norm": 1.1722397804260254, + "learning_rate": 4.377149569905388e-05, + "loss": 0.3519, + "step": 40315 + }, + { + "epoch": 2.797585177989036, + "grad_norm": 1.4774320125579834, + "learning_rate": 4.376684951034229e-05, + "loss": 0.5086, + "step": 40316 + }, + { + "epoch": 2.7976545694261326, + "grad_norm": 1.0413986444473267, + "learning_rate": 4.3762203499152234e-05, + "loss": 0.3274, + "step": 40317 + }, + { + "epoch": 2.7977239608632294, + "grad_norm": 1.6588687896728516, + "learning_rate": 4.3757557665498416e-05, + "loss": 0.3756, + "step": 40318 + }, + { + "epoch": 2.797793352300326, + "grad_norm": 0.8621796369552612, + "learning_rate": 4.375291200939542e-05, + "loss": 0.2848, + "step": 40319 + }, + { + "epoch": 2.797862743737423, + "grad_norm": 2.263369083404541, + "learning_rate": 4.374826653085796e-05, + "loss": 0.6004, + "step": 40320 + }, + { + "epoch": 2.7979321351745194, + "grad_norm": 0.8137166500091553, + "learning_rate": 4.3743621229900696e-05, + "loss": 0.3783, + "step": 40321 + }, + { + "epoch": 2.7980015266116163, + "grad_norm": 1.6989428997039795, + "learning_rate": 4.3738976106538297e-05, + "loss": 0.4845, + "step": 40322 + }, + { + "epoch": 2.7980709180487127, + "grad_norm": 0.907428503036499, + "learning_rate": 4.373433116078546e-05, + "loss": 0.4004, + "step": 40323 + }, + { + "epoch": 2.7981403094858095, + "grad_norm": 1.1662075519561768, + "learning_rate": 4.3729686392656764e-05, + "loss": 0.1985, + "step": 40324 + }, + { + "epoch": 2.7982097009229063, + "grad_norm": 1.2335187196731567, + "learning_rate": 4.3725041802166924e-05, + "loss": 0.487, + "step": 40325 + }, + { + "epoch": 2.7982790923600027, + "grad_norm": 1.0610401630401611, + "learning_rate": 4.3720397389330604e-05, + "loss": 0.3036, + "step": 40326 + }, + { + "epoch": 2.7983484837970996, + "grad_norm": 1.379831075668335, + "learning_rate": 4.371575315416249e-05, + "loss": 0.5287, + "step": 40327 + }, + { + "epoch": 2.798417875234196, + "grad_norm": 0.8378221392631531, + "learning_rate": 4.371110909667718e-05, + "loss": 0.2226, + "step": 40328 + }, + { + "epoch": 2.7984872666712928, + "grad_norm": 1.016599178314209, + "learning_rate": 4.370646521688937e-05, + "loss": 0.361, + "step": 40329 + }, + { + "epoch": 2.7985566581083896, + "grad_norm": 1.2477993965148926, + "learning_rate": 4.370182151481371e-05, + "loss": 0.5228, + "step": 40330 + }, + { + "epoch": 2.798626049545486, + "grad_norm": 1.048357605934143, + "learning_rate": 4.3697177990464874e-05, + "loss": 0.5032, + "step": 40331 + }, + { + "epoch": 2.798695440982583, + "grad_norm": 1.9050573110580444, + "learning_rate": 4.3692534643857555e-05, + "loss": 0.319, + "step": 40332 + }, + { + "epoch": 2.798764832419679, + "grad_norm": 1.6916435956954956, + "learning_rate": 4.368789147500633e-05, + "loss": 0.5794, + "step": 40333 + }, + { + "epoch": 2.798834223856776, + "grad_norm": 1.119785189628601, + "learning_rate": 4.368324848392589e-05, + "loss": 0.3683, + "step": 40334 + }, + { + "epoch": 2.798903615293873, + "grad_norm": 0.9697985053062439, + "learning_rate": 4.3678605670630915e-05, + "loss": 0.4399, + "step": 40335 + }, + { + "epoch": 2.7989730067309693, + "grad_norm": 0.9842323064804077, + "learning_rate": 4.367396303513603e-05, + "loss": 0.3138, + "step": 40336 + }, + { + "epoch": 2.799042398168066, + "grad_norm": 1.3875685930252075, + "learning_rate": 4.366932057745593e-05, + "loss": 0.4241, + "step": 40337 + }, + { + "epoch": 2.7991117896051625, + "grad_norm": 1.4539988040924072, + "learning_rate": 4.366467829760523e-05, + "loss": 0.4654, + "step": 40338 + }, + { + "epoch": 2.7991811810422593, + "grad_norm": 1.0661804676055908, + "learning_rate": 4.3660036195598655e-05, + "loss": 0.3433, + "step": 40339 + }, + { + "epoch": 2.799250572479356, + "grad_norm": 1.6832329034805298, + "learning_rate": 4.365539427145077e-05, + "loss": 0.5008, + "step": 40340 + }, + { + "epoch": 2.7993199639164525, + "grad_norm": 0.9778301119804382, + "learning_rate": 4.3650752525176264e-05, + "loss": 0.3489, + "step": 40341 + }, + { + "epoch": 2.7993893553535494, + "grad_norm": 0.9824504852294922, + "learning_rate": 4.36461109567898e-05, + "loss": 0.3368, + "step": 40342 + }, + { + "epoch": 2.7994587467906458, + "grad_norm": 1.0867420434951782, + "learning_rate": 4.3641469566306016e-05, + "loss": 0.3094, + "step": 40343 + }, + { + "epoch": 2.7995281382277426, + "grad_norm": 2.1154184341430664, + "learning_rate": 4.363682835373958e-05, + "loss": 0.4693, + "step": 40344 + }, + { + "epoch": 2.7995975296648394, + "grad_norm": 1.3556445837020874, + "learning_rate": 4.363218731910513e-05, + "loss": 0.4385, + "step": 40345 + }, + { + "epoch": 2.7996669211019363, + "grad_norm": 2.260568380355835, + "learning_rate": 4.362754646241737e-05, + "loss": 0.3955, + "step": 40346 + }, + { + "epoch": 2.7997363125390327, + "grad_norm": 1.3332418203353882, + "learning_rate": 4.3622905783690874e-05, + "loss": 0.3662, + "step": 40347 + }, + { + "epoch": 2.7998057039761295, + "grad_norm": 1.6467571258544922, + "learning_rate": 4.361826528294031e-05, + "loss": 0.5454, + "step": 40348 + }, + { + "epoch": 2.799875095413226, + "grad_norm": 1.0213454961776733, + "learning_rate": 4.361362496018036e-05, + "loss": 0.3929, + "step": 40349 + }, + { + "epoch": 2.7999444868503227, + "grad_norm": 1.188068151473999, + "learning_rate": 4.360898481542563e-05, + "loss": 0.3682, + "step": 40350 + }, + { + "epoch": 2.8000138782874195, + "grad_norm": 0.9234535098075867, + "learning_rate": 4.3604344848690807e-05, + "loss": 0.4194, + "step": 40351 + }, + { + "epoch": 2.800083269724516, + "grad_norm": 0.7720056772232056, + "learning_rate": 4.3599705059990515e-05, + "loss": 0.3703, + "step": 40352 + }, + { + "epoch": 2.8001526611616128, + "grad_norm": 1.2356011867523193, + "learning_rate": 4.3595065449339455e-05, + "loss": 0.4111, + "step": 40353 + }, + { + "epoch": 2.800222052598709, + "grad_norm": 1.2690935134887695, + "learning_rate": 4.35904260167522e-05, + "loss": 0.3143, + "step": 40354 + }, + { + "epoch": 2.800291444035806, + "grad_norm": 1.0223931074142456, + "learning_rate": 4.3585786762243405e-05, + "loss": 0.3522, + "step": 40355 + }, + { + "epoch": 2.800360835472903, + "grad_norm": 0.7536935210227966, + "learning_rate": 4.358114768582775e-05, + "loss": 0.2647, + "step": 40356 + }, + { + "epoch": 2.800430226909999, + "grad_norm": 1.242042064666748, + "learning_rate": 4.3576508787519864e-05, + "loss": 0.4441, + "step": 40357 + }, + { + "epoch": 2.800499618347096, + "grad_norm": 1.3123869895935059, + "learning_rate": 4.357187006733443e-05, + "loss": 0.4675, + "step": 40358 + }, + { + "epoch": 2.8005690097841924, + "grad_norm": 1.3771055936813354, + "learning_rate": 4.3567231525286015e-05, + "loss": 0.3165, + "step": 40359 + }, + { + "epoch": 2.8006384012212893, + "grad_norm": 1.527026891708374, + "learning_rate": 4.356259316138931e-05, + "loss": 0.4799, + "step": 40360 + }, + { + "epoch": 2.800707792658386, + "grad_norm": 1.3262971639633179, + "learning_rate": 4.3557954975658945e-05, + "loss": 0.619, + "step": 40361 + }, + { + "epoch": 2.8007771840954825, + "grad_norm": 1.6258386373519897, + "learning_rate": 4.3553316968109604e-05, + "loss": 0.3983, + "step": 40362 + }, + { + "epoch": 2.8008465755325793, + "grad_norm": 1.2622278928756714, + "learning_rate": 4.3548679138755856e-05, + "loss": 0.3153, + "step": 40363 + }, + { + "epoch": 2.8009159669696757, + "grad_norm": 1.275949478149414, + "learning_rate": 4.354404148761237e-05, + "loss": 0.345, + "step": 40364 + }, + { + "epoch": 2.8009853584067725, + "grad_norm": 1.2908278703689575, + "learning_rate": 4.353940401469381e-05, + "loss": 0.4428, + "step": 40365 + }, + { + "epoch": 2.8010547498438694, + "grad_norm": 1.189902424812317, + "learning_rate": 4.3534766720014796e-05, + "loss": 0.4518, + "step": 40366 + }, + { + "epoch": 2.8011241412809658, + "grad_norm": 0.8708195686340332, + "learning_rate": 4.353012960359001e-05, + "loss": 0.2513, + "step": 40367 + }, + { + "epoch": 2.8011935327180626, + "grad_norm": 1.8198236227035522, + "learning_rate": 4.352549266543401e-05, + "loss": 0.4145, + "step": 40368 + }, + { + "epoch": 2.801262924155159, + "grad_norm": 2.347085952758789, + "learning_rate": 4.352085590556148e-05, + "loss": 0.3709, + "step": 40369 + }, + { + "epoch": 2.801332315592256, + "grad_norm": 1.528455138206482, + "learning_rate": 4.351621932398706e-05, + "loss": 0.577, + "step": 40370 + }, + { + "epoch": 2.8014017070293527, + "grad_norm": 1.0855015516281128, + "learning_rate": 4.351158292072538e-05, + "loss": 0.3459, + "step": 40371 + }, + { + "epoch": 2.8014710984664495, + "grad_norm": 1.1593868732452393, + "learning_rate": 4.3506946695791075e-05, + "loss": 0.283, + "step": 40372 + }, + { + "epoch": 2.801540489903546, + "grad_norm": 1.375498652458191, + "learning_rate": 4.350231064919883e-05, + "loss": 0.4365, + "step": 40373 + }, + { + "epoch": 2.8016098813406427, + "grad_norm": 1.572731375694275, + "learning_rate": 4.3497674780963196e-05, + "loss": 0.6121, + "step": 40374 + }, + { + "epoch": 2.801679272777739, + "grad_norm": 1.3829319477081299, + "learning_rate": 4.349303909109884e-05, + "loss": 0.4881, + "step": 40375 + }, + { + "epoch": 2.801748664214836, + "grad_norm": 1.2900067567825317, + "learning_rate": 4.348840357962041e-05, + "loss": 0.328, + "step": 40376 + }, + { + "epoch": 2.8018180556519328, + "grad_norm": 1.3300738334655762, + "learning_rate": 4.3483768246542523e-05, + "loss": 0.4675, + "step": 40377 + }, + { + "epoch": 2.801887447089029, + "grad_norm": 1.0976275205612183, + "learning_rate": 4.347913309187983e-05, + "loss": 0.4913, + "step": 40378 + }, + { + "epoch": 2.801956838526126, + "grad_norm": 0.9651568531990051, + "learning_rate": 4.347449811564695e-05, + "loss": 0.2864, + "step": 40379 + }, + { + "epoch": 2.8020262299632224, + "grad_norm": 0.9746438264846802, + "learning_rate": 4.3469863317858574e-05, + "loss": 0.2892, + "step": 40380 + }, + { + "epoch": 2.802095621400319, + "grad_norm": 1.6286214590072632, + "learning_rate": 4.3465228698529235e-05, + "loss": 0.4137, + "step": 40381 + }, + { + "epoch": 2.802165012837416, + "grad_norm": 0.9013366103172302, + "learning_rate": 4.3460594257673615e-05, + "loss": 0.2999, + "step": 40382 + }, + { + "epoch": 2.8022344042745124, + "grad_norm": 2.1416685581207275, + "learning_rate": 4.345595999530633e-05, + "loss": 0.3549, + "step": 40383 + }, + { + "epoch": 2.8023037957116093, + "grad_norm": 1.2481213808059692, + "learning_rate": 4.345132591144202e-05, + "loss": 0.4122, + "step": 40384 + }, + { + "epoch": 2.8023731871487056, + "grad_norm": 1.1036863327026367, + "learning_rate": 4.344669200609533e-05, + "loss": 0.3754, + "step": 40385 + }, + { + "epoch": 2.8024425785858025, + "grad_norm": 1.637969970703125, + "learning_rate": 4.344205827928085e-05, + "loss": 0.4506, + "step": 40386 + }, + { + "epoch": 2.8025119700228993, + "grad_norm": 1.422669768333435, + "learning_rate": 4.343742473101328e-05, + "loss": 0.451, + "step": 40387 + }, + { + "epoch": 2.8025813614599957, + "grad_norm": 1.9259274005889893, + "learning_rate": 4.3432791361307146e-05, + "loss": 0.4574, + "step": 40388 + }, + { + "epoch": 2.8026507528970925, + "grad_norm": 0.997072160243988, + "learning_rate": 4.3428158170177136e-05, + "loss": 0.2758, + "step": 40389 + }, + { + "epoch": 2.802720144334189, + "grad_norm": 0.9968132376670837, + "learning_rate": 4.342352515763787e-05, + "loss": 0.2816, + "step": 40390 + }, + { + "epoch": 2.8027895357712858, + "grad_norm": 1.1025665998458862, + "learning_rate": 4.3418892323703965e-05, + "loss": 0.3457, + "step": 40391 + }, + { + "epoch": 2.8028589272083826, + "grad_norm": 2.406628370285034, + "learning_rate": 4.3414259668390056e-05, + "loss": 0.5239, + "step": 40392 + }, + { + "epoch": 2.8029283186454794, + "grad_norm": 1.3212791681289673, + "learning_rate": 4.34096271917108e-05, + "loss": 0.6026, + "step": 40393 + }, + { + "epoch": 2.802997710082576, + "grad_norm": 1.0328922271728516, + "learning_rate": 4.340499489368075e-05, + "loss": 0.2755, + "step": 40394 + }, + { + "epoch": 2.803067101519672, + "grad_norm": 2.1882150173187256, + "learning_rate": 4.340036277431456e-05, + "loss": 0.5081, + "step": 40395 + }, + { + "epoch": 2.803136492956769, + "grad_norm": 1.0814473628997803, + "learning_rate": 4.3395730833626866e-05, + "loss": 0.3737, + "step": 40396 + }, + { + "epoch": 2.803205884393866, + "grad_norm": 0.9788290858268738, + "learning_rate": 4.339109907163231e-05, + "loss": 0.328, + "step": 40397 + }, + { + "epoch": 2.8032752758309627, + "grad_norm": 1.1836848258972168, + "learning_rate": 4.338646748834546e-05, + "loss": 0.3649, + "step": 40398 + }, + { + "epoch": 2.803344667268059, + "grad_norm": 1.2381694316864014, + "learning_rate": 4.338183608378096e-05, + "loss": 0.508, + "step": 40399 + }, + { + "epoch": 2.803414058705156, + "grad_norm": 1.2496923208236694, + "learning_rate": 4.337720485795344e-05, + "loss": 0.4303, + "step": 40400 + }, + { + "epoch": 2.8034834501422523, + "grad_norm": 1.7960177659988403, + "learning_rate": 4.3372573810877534e-05, + "loss": 0.5244, + "step": 40401 + }, + { + "epoch": 2.803552841579349, + "grad_norm": 1.2141627073287964, + "learning_rate": 4.336794294256782e-05, + "loss": 0.3616, + "step": 40402 + }, + { + "epoch": 2.803622233016446, + "grad_norm": 0.8729479312896729, + "learning_rate": 4.3363312253038935e-05, + "loss": 0.2722, + "step": 40403 + }, + { + "epoch": 2.8036916244535424, + "grad_norm": 1.2254539728164673, + "learning_rate": 4.3358681742305506e-05, + "loss": 0.6058, + "step": 40404 + }, + { + "epoch": 2.803761015890639, + "grad_norm": 1.229224443435669, + "learning_rate": 4.335405141038215e-05, + "loss": 0.5022, + "step": 40405 + }, + { + "epoch": 2.8038304073277356, + "grad_norm": 0.9447742104530334, + "learning_rate": 4.334942125728347e-05, + "loss": 0.2467, + "step": 40406 + }, + { + "epoch": 2.8038997987648324, + "grad_norm": 1.4575523138046265, + "learning_rate": 4.33447912830241e-05, + "loss": 0.3561, + "step": 40407 + }, + { + "epoch": 2.8039691902019293, + "grad_norm": 1.969527006149292, + "learning_rate": 4.334016148761869e-05, + "loss": 0.4848, + "step": 40408 + }, + { + "epoch": 2.8040385816390256, + "grad_norm": 1.4927639961242676, + "learning_rate": 4.333553187108178e-05, + "loss": 0.5079, + "step": 40409 + }, + { + "epoch": 2.8041079730761225, + "grad_norm": 1.7971903085708618, + "learning_rate": 4.3330902433428014e-05, + "loss": 0.4197, + "step": 40410 + }, + { + "epoch": 2.804177364513219, + "grad_norm": 0.8423730134963989, + "learning_rate": 4.3326273174672016e-05, + "loss": 0.3391, + "step": 40411 + }, + { + "epoch": 2.8042467559503157, + "grad_norm": 1.254734992980957, + "learning_rate": 4.3321644094828404e-05, + "loss": 0.5081, + "step": 40412 + }, + { + "epoch": 2.8043161473874125, + "grad_norm": 0.8104230165481567, + "learning_rate": 4.331701519391178e-05, + "loss": 0.2923, + "step": 40413 + }, + { + "epoch": 2.804385538824509, + "grad_norm": 2.018634557723999, + "learning_rate": 4.331238647193677e-05, + "loss": 0.3824, + "step": 40414 + }, + { + "epoch": 2.8044549302616057, + "grad_norm": 1.2856594324111938, + "learning_rate": 4.3307757928918016e-05, + "loss": 0.4034, + "step": 40415 + }, + { + "epoch": 2.804524321698702, + "grad_norm": 0.9551953077316284, + "learning_rate": 4.330312956487005e-05, + "loss": 0.3871, + "step": 40416 + }, + { + "epoch": 2.804593713135799, + "grad_norm": 1.1349122524261475, + "learning_rate": 4.329850137980752e-05, + "loss": 0.3665, + "step": 40417 + }, + { + "epoch": 2.804663104572896, + "grad_norm": 1.0301086902618408, + "learning_rate": 4.329387337374505e-05, + "loss": 0.324, + "step": 40418 + }, + { + "epoch": 2.8047324960099926, + "grad_norm": 1.1674238443374634, + "learning_rate": 4.3289245546697245e-05, + "loss": 0.3762, + "step": 40419 + }, + { + "epoch": 2.804801887447089, + "grad_norm": 0.9471697211265564, + "learning_rate": 4.3284617898678717e-05, + "loss": 0.362, + "step": 40420 + }, + { + "epoch": 2.8048712788841854, + "grad_norm": 1.1469981670379639, + "learning_rate": 4.3279990429704054e-05, + "loss": 0.3211, + "step": 40421 + }, + { + "epoch": 2.8049406703212822, + "grad_norm": 0.9246501326560974, + "learning_rate": 4.327536313978793e-05, + "loss": 0.3888, + "step": 40422 + }, + { + "epoch": 2.805010061758379, + "grad_norm": 1.3589974641799927, + "learning_rate": 4.327073602894486e-05, + "loss": 0.4454, + "step": 40423 + }, + { + "epoch": 2.805079453195476, + "grad_norm": 1.035083293914795, + "learning_rate": 4.32661090971895e-05, + "loss": 0.2679, + "step": 40424 + }, + { + "epoch": 2.8051488446325723, + "grad_norm": 1.0880306959152222, + "learning_rate": 4.326148234453644e-05, + "loss": 0.4003, + "step": 40425 + }, + { + "epoch": 2.805218236069669, + "grad_norm": 1.0978296995162964, + "learning_rate": 4.3256855771000315e-05, + "loss": 0.4086, + "step": 40426 + }, + { + "epoch": 2.8052876275067655, + "grad_norm": 1.7799828052520752, + "learning_rate": 4.32522293765957e-05, + "loss": 0.4422, + "step": 40427 + }, + { + "epoch": 2.8053570189438624, + "grad_norm": 1.094177484512329, + "learning_rate": 4.324760316133725e-05, + "loss": 0.4097, + "step": 40428 + }, + { + "epoch": 2.805426410380959, + "grad_norm": 1.5373508930206299, + "learning_rate": 4.32429771252395e-05, + "loss": 0.4279, + "step": 40429 + }, + { + "epoch": 2.8054958018180556, + "grad_norm": 1.5031476020812988, + "learning_rate": 4.323835126831709e-05, + "loss": 0.5105, + "step": 40430 + }, + { + "epoch": 2.8055651932551524, + "grad_norm": 0.9405916929244995, + "learning_rate": 4.323372559058461e-05, + "loss": 0.4174, + "step": 40431 + }, + { + "epoch": 2.805634584692249, + "grad_norm": 1.2498726844787598, + "learning_rate": 4.3229100092056716e-05, + "loss": 0.6526, + "step": 40432 + }, + { + "epoch": 2.8057039761293456, + "grad_norm": 0.9002657532691956, + "learning_rate": 4.322447477274792e-05, + "loss": 0.3738, + "step": 40433 + }, + { + "epoch": 2.8057733675664425, + "grad_norm": 1.6865798234939575, + "learning_rate": 4.321984963267288e-05, + "loss": 0.4479, + "step": 40434 + }, + { + "epoch": 2.805842759003539, + "grad_norm": 0.8898730278015137, + "learning_rate": 4.321522467184619e-05, + "loss": 0.3853, + "step": 40435 + }, + { + "epoch": 2.8059121504406357, + "grad_norm": 0.9671227931976318, + "learning_rate": 4.3210599890282477e-05, + "loss": 0.2482, + "step": 40436 + }, + { + "epoch": 2.805981541877732, + "grad_norm": 1.1801564693450928, + "learning_rate": 4.320597528799627e-05, + "loss": 0.4469, + "step": 40437 + }, + { + "epoch": 2.806050933314829, + "grad_norm": 1.485316276550293, + "learning_rate": 4.320135086500222e-05, + "loss": 0.3469, + "step": 40438 + }, + { + "epoch": 2.8061203247519257, + "grad_norm": 1.027872085571289, + "learning_rate": 4.3196726621314907e-05, + "loss": 0.3136, + "step": 40439 + }, + { + "epoch": 2.806189716189022, + "grad_norm": 0.9984025955200195, + "learning_rate": 4.3192102556948945e-05, + "loss": 0.3773, + "step": 40440 + }, + { + "epoch": 2.806259107626119, + "grad_norm": 1.0969005823135376, + "learning_rate": 4.3187478671918914e-05, + "loss": 0.2765, + "step": 40441 + }, + { + "epoch": 2.8063284990632154, + "grad_norm": 1.4000808000564575, + "learning_rate": 4.318285496623943e-05, + "loss": 0.5197, + "step": 40442 + }, + { + "epoch": 2.806397890500312, + "grad_norm": 1.335140585899353, + "learning_rate": 4.317823143992512e-05, + "loss": 0.5012, + "step": 40443 + }, + { + "epoch": 2.806467281937409, + "grad_norm": 1.4206827878952026, + "learning_rate": 4.3173608092990494e-05, + "loss": 0.4612, + "step": 40444 + }, + { + "epoch": 2.806536673374506, + "grad_norm": 0.9807228446006775, + "learning_rate": 4.316898492545021e-05, + "loss": 0.4008, + "step": 40445 + }, + { + "epoch": 2.8066060648116022, + "grad_norm": 1.0885331630706787, + "learning_rate": 4.316436193731883e-05, + "loss": 0.3534, + "step": 40446 + }, + { + "epoch": 2.806675456248699, + "grad_norm": 0.7864665389060974, + "learning_rate": 4.3159739128610975e-05, + "loss": 0.3214, + "step": 40447 + }, + { + "epoch": 2.8067448476857955, + "grad_norm": 0.965855598449707, + "learning_rate": 4.3155116499341216e-05, + "loss": 0.4223, + "step": 40448 + }, + { + "epoch": 2.8068142391228923, + "grad_norm": 1.0696461200714111, + "learning_rate": 4.315049404952422e-05, + "loss": 0.3261, + "step": 40449 + }, + { + "epoch": 2.806883630559989, + "grad_norm": 1.0219049453735352, + "learning_rate": 4.3145871779174465e-05, + "loss": 0.3734, + "step": 40450 + }, + { + "epoch": 2.8069530219970855, + "grad_norm": 0.7370311617851257, + "learning_rate": 4.3141249688306597e-05, + "loss": 0.2277, + "step": 40451 + }, + { + "epoch": 2.8070224134341824, + "grad_norm": 1.01727294921875, + "learning_rate": 4.31366277769352e-05, + "loss": 0.3621, + "step": 40452 + }, + { + "epoch": 2.8070918048712787, + "grad_norm": 0.9779333472251892, + "learning_rate": 4.313200604507489e-05, + "loss": 0.3395, + "step": 40453 + }, + { + "epoch": 2.8071611963083756, + "grad_norm": 1.353319525718689, + "learning_rate": 4.312738449274023e-05, + "loss": 0.4294, + "step": 40454 + }, + { + "epoch": 2.8072305877454724, + "grad_norm": 1.1011494398117065, + "learning_rate": 4.312276311994582e-05, + "loss": 0.3154, + "step": 40455 + }, + { + "epoch": 2.807299979182569, + "grad_norm": 0.8810011148452759, + "learning_rate": 4.311814192670628e-05, + "loss": 0.3503, + "step": 40456 + }, + { + "epoch": 2.8073693706196656, + "grad_norm": 1.1263792514801025, + "learning_rate": 4.3113520913036134e-05, + "loss": 0.4819, + "step": 40457 + }, + { + "epoch": 2.807438762056762, + "grad_norm": 1.4397872686386108, + "learning_rate": 4.310890007895e-05, + "loss": 0.392, + "step": 40458 + }, + { + "epoch": 2.807508153493859, + "grad_norm": 1.3278415203094482, + "learning_rate": 4.310427942446247e-05, + "loss": 0.3335, + "step": 40459 + }, + { + "epoch": 2.8075775449309557, + "grad_norm": 1.0136144161224365, + "learning_rate": 4.309965894958813e-05, + "loss": 0.3458, + "step": 40460 + }, + { + "epoch": 2.807646936368052, + "grad_norm": 1.2111852169036865, + "learning_rate": 4.3095038654341555e-05, + "loss": 0.5046, + "step": 40461 + }, + { + "epoch": 2.807716327805149, + "grad_norm": 1.5564700365066528, + "learning_rate": 4.3090418538737355e-05, + "loss": 0.6729, + "step": 40462 + }, + { + "epoch": 2.8077857192422453, + "grad_norm": 2.058340072631836, + "learning_rate": 4.308579860279013e-05, + "loss": 0.4136, + "step": 40463 + }, + { + "epoch": 2.807855110679342, + "grad_norm": 1.257023572921753, + "learning_rate": 4.308117884651439e-05, + "loss": 0.3887, + "step": 40464 + }, + { + "epoch": 2.807924502116439, + "grad_norm": 1.6548724174499512, + "learning_rate": 4.307655926992478e-05, + "loss": 0.4358, + "step": 40465 + }, + { + "epoch": 2.8079938935535353, + "grad_norm": 1.3636285066604614, + "learning_rate": 4.307193987303589e-05, + "loss": 0.4535, + "step": 40466 + }, + { + "epoch": 2.808063284990632, + "grad_norm": 1.1503854990005493, + "learning_rate": 4.306732065586226e-05, + "loss": 0.3175, + "step": 40467 + }, + { + "epoch": 2.8081326764277286, + "grad_norm": 1.1282449960708618, + "learning_rate": 4.306270161841849e-05, + "loss": 0.365, + "step": 40468 + }, + { + "epoch": 2.8082020678648254, + "grad_norm": 1.905279278755188, + "learning_rate": 4.305808276071916e-05, + "loss": 0.4516, + "step": 40469 + }, + { + "epoch": 2.8082714593019222, + "grad_norm": 0.8094700574874878, + "learning_rate": 4.3053464082778895e-05, + "loss": 0.3539, + "step": 40470 + }, + { + "epoch": 2.808340850739019, + "grad_norm": 1.6605513095855713, + "learning_rate": 4.3048845584612196e-05, + "loss": 0.4394, + "step": 40471 + }, + { + "epoch": 2.8084102421761155, + "grad_norm": 1.458448052406311, + "learning_rate": 4.3044227266233695e-05, + "loss": 0.366, + "step": 40472 + }, + { + "epoch": 2.8084796336132123, + "grad_norm": 0.8848315477371216, + "learning_rate": 4.303960912765795e-05, + "loss": 0.38, + "step": 40473 + }, + { + "epoch": 2.8085490250503087, + "grad_norm": 1.1855947971343994, + "learning_rate": 4.3034991168899555e-05, + "loss": 0.3534, + "step": 40474 + }, + { + "epoch": 2.8086184164874055, + "grad_norm": 0.9441655874252319, + "learning_rate": 4.3030373389973074e-05, + "loss": 0.2799, + "step": 40475 + }, + { + "epoch": 2.8086878079245023, + "grad_norm": 1.1125181913375854, + "learning_rate": 4.302575579089311e-05, + "loss": 0.3966, + "step": 40476 + }, + { + "epoch": 2.8087571993615987, + "grad_norm": 1.898333191871643, + "learning_rate": 4.3021138371674253e-05, + "loss": 0.5612, + "step": 40477 + }, + { + "epoch": 2.8088265907986956, + "grad_norm": 1.3493707180023193, + "learning_rate": 4.3016521132331024e-05, + "loss": 0.4335, + "step": 40478 + }, + { + "epoch": 2.808895982235792, + "grad_norm": 0.8862612247467041, + "learning_rate": 4.3011904072878015e-05, + "loss": 0.3433, + "step": 40479 + }, + { + "epoch": 2.808965373672889, + "grad_norm": 0.9448130130767822, + "learning_rate": 4.300728719332982e-05, + "loss": 0.3388, + "step": 40480 + }, + { + "epoch": 2.8090347651099856, + "grad_norm": 0.8114845752716064, + "learning_rate": 4.300267049370099e-05, + "loss": 0.209, + "step": 40481 + }, + { + "epoch": 2.809104156547082, + "grad_norm": 1.6355847120285034, + "learning_rate": 4.2998053974006136e-05, + "loss": 0.4396, + "step": 40482 + }, + { + "epoch": 2.809173547984179, + "grad_norm": 1.123950719833374, + "learning_rate": 4.2993437634259804e-05, + "loss": 0.3418, + "step": 40483 + }, + { + "epoch": 2.8092429394212752, + "grad_norm": 1.234453558921814, + "learning_rate": 4.298882147447661e-05, + "loss": 0.5394, + "step": 40484 + }, + { + "epoch": 2.809312330858372, + "grad_norm": 1.504184603691101, + "learning_rate": 4.298420549467107e-05, + "loss": 0.4436, + "step": 40485 + }, + { + "epoch": 2.809381722295469, + "grad_norm": 1.190224289894104, + "learning_rate": 4.297958969485776e-05, + "loss": 0.4847, + "step": 40486 + }, + { + "epoch": 2.8094511137325653, + "grad_norm": 1.27708101272583, + "learning_rate": 4.2974974075051276e-05, + "loss": 0.3034, + "step": 40487 + }, + { + "epoch": 2.809520505169662, + "grad_norm": 1.5436079502105713, + "learning_rate": 4.297035863526618e-05, + "loss": 0.3804, + "step": 40488 + }, + { + "epoch": 2.8095898966067585, + "grad_norm": 1.069415807723999, + "learning_rate": 4.296574337551705e-05, + "loss": 0.3959, + "step": 40489 + }, + { + "epoch": 2.8096592880438553, + "grad_norm": 1.1496319770812988, + "learning_rate": 4.296112829581845e-05, + "loss": 0.3158, + "step": 40490 + }, + { + "epoch": 2.809728679480952, + "grad_norm": 1.2222623825073242, + "learning_rate": 4.2956513396184994e-05, + "loss": 0.5705, + "step": 40491 + }, + { + "epoch": 2.8097980709180486, + "grad_norm": 1.1033140420913696, + "learning_rate": 4.295189867663116e-05, + "loss": 0.3689, + "step": 40492 + }, + { + "epoch": 2.8098674623551454, + "grad_norm": 1.3186938762664795, + "learning_rate": 4.2947284137171576e-05, + "loss": 0.2665, + "step": 40493 + }, + { + "epoch": 2.809936853792242, + "grad_norm": 1.0119191408157349, + "learning_rate": 4.2942669777820774e-05, + "loss": 0.3583, + "step": 40494 + }, + { + "epoch": 2.8100062452293386, + "grad_norm": 1.1543338298797607, + "learning_rate": 4.2938055598593364e-05, + "loss": 0.308, + "step": 40495 + }, + { + "epoch": 2.8100756366664355, + "grad_norm": 0.9272451996803284, + "learning_rate": 4.293344159950389e-05, + "loss": 0.3319, + "step": 40496 + }, + { + "epoch": 2.8101450281035323, + "grad_norm": 1.0770689249038696, + "learning_rate": 4.2928827780566904e-05, + "loss": 0.3817, + "step": 40497 + }, + { + "epoch": 2.8102144195406287, + "grad_norm": 1.092808723449707, + "learning_rate": 4.292421414179704e-05, + "loss": 0.3123, + "step": 40498 + }, + { + "epoch": 2.8102838109777255, + "grad_norm": 0.6679819822311401, + "learning_rate": 4.2919600683208774e-05, + "loss": 0.2236, + "step": 40499 + }, + { + "epoch": 2.810353202414822, + "grad_norm": 1.3873955011367798, + "learning_rate": 4.29149874048167e-05, + "loss": 0.5091, + "step": 40500 + }, + { + "epoch": 2.8104225938519187, + "grad_norm": 1.5208029747009277, + "learning_rate": 4.291037430663542e-05, + "loss": 0.2373, + "step": 40501 + }, + { + "epoch": 2.8104919852890156, + "grad_norm": 1.0742607116699219, + "learning_rate": 4.290576138867943e-05, + "loss": 0.3298, + "step": 40502 + }, + { + "epoch": 2.810561376726112, + "grad_norm": 0.7058163285255432, + "learning_rate": 4.290114865096333e-05, + "loss": 0.2497, + "step": 40503 + }, + { + "epoch": 2.810630768163209, + "grad_norm": 1.2259249687194824, + "learning_rate": 4.289653609350168e-05, + "loss": 0.4219, + "step": 40504 + }, + { + "epoch": 2.810700159600305, + "grad_norm": 1.2047233581542969, + "learning_rate": 4.289192371630907e-05, + "loss": 0.3408, + "step": 40505 + }, + { + "epoch": 2.810769551037402, + "grad_norm": 2.242619752883911, + "learning_rate": 4.288731151940001e-05, + "loss": 0.488, + "step": 40506 + }, + { + "epoch": 2.810838942474499, + "grad_norm": 1.1612205505371094, + "learning_rate": 4.288269950278906e-05, + "loss": 0.3995, + "step": 40507 + }, + { + "epoch": 2.8109083339115952, + "grad_norm": 1.3020715713500977, + "learning_rate": 4.287808766649082e-05, + "loss": 0.4879, + "step": 40508 + }, + { + "epoch": 2.810977725348692, + "grad_norm": 1.3525162935256958, + "learning_rate": 4.287347601051982e-05, + "loss": 0.3085, + "step": 40509 + }, + { + "epoch": 2.8110471167857884, + "grad_norm": 1.2840219736099243, + "learning_rate": 4.286886453489063e-05, + "loss": 0.3193, + "step": 40510 + }, + { + "epoch": 2.8111165082228853, + "grad_norm": 1.3637499809265137, + "learning_rate": 4.28642532396178e-05, + "loss": 0.4396, + "step": 40511 + }, + { + "epoch": 2.811185899659982, + "grad_norm": 1.5918946266174316, + "learning_rate": 4.285964212471594e-05, + "loss": 0.6415, + "step": 40512 + }, + { + "epoch": 2.8112552910970785, + "grad_norm": 1.1039084196090698, + "learning_rate": 4.285503119019952e-05, + "loss": 0.3834, + "step": 40513 + }, + { + "epoch": 2.8113246825341753, + "grad_norm": 1.2992465496063232, + "learning_rate": 4.2850420436083136e-05, + "loss": 0.3933, + "step": 40514 + }, + { + "epoch": 2.8113940739712717, + "grad_norm": 0.8905726671218872, + "learning_rate": 4.284580986238135e-05, + "loss": 0.4074, + "step": 40515 + }, + { + "epoch": 2.8114634654083686, + "grad_norm": 1.1262996196746826, + "learning_rate": 4.28411994691087e-05, + "loss": 0.3914, + "step": 40516 + }, + { + "epoch": 2.8115328568454654, + "grad_norm": 1.457189917564392, + "learning_rate": 4.283658925627976e-05, + "loss": 0.4493, + "step": 40517 + }, + { + "epoch": 2.8116022482825618, + "grad_norm": 1.746280550956726, + "learning_rate": 4.283197922390907e-05, + "loss": 0.5032, + "step": 40518 + }, + { + "epoch": 2.8116716397196586, + "grad_norm": 1.2205020189285278, + "learning_rate": 4.282736937201123e-05, + "loss": 0.3555, + "step": 40519 + }, + { + "epoch": 2.811741031156755, + "grad_norm": 1.0188921689987183, + "learning_rate": 4.282275970060071e-05, + "loss": 0.3152, + "step": 40520 + }, + { + "epoch": 2.811810422593852, + "grad_norm": 0.79329514503479, + "learning_rate": 4.281815020969211e-05, + "loss": 0.2732, + "step": 40521 + }, + { + "epoch": 2.8118798140309487, + "grad_norm": 1.0309562683105469, + "learning_rate": 4.281354089929998e-05, + "loss": 0.2535, + "step": 40522 + }, + { + "epoch": 2.8119492054680455, + "grad_norm": 1.1941735744476318, + "learning_rate": 4.2808931769438855e-05, + "loss": 0.436, + "step": 40523 + }, + { + "epoch": 2.812018596905142, + "grad_norm": 1.2707691192626953, + "learning_rate": 4.28043228201233e-05, + "loss": 0.384, + "step": 40524 + }, + { + "epoch": 2.8120879883422387, + "grad_norm": 1.2768317461013794, + "learning_rate": 4.279971405136787e-05, + "loss": 0.4603, + "step": 40525 + }, + { + "epoch": 2.812157379779335, + "grad_norm": 1.335993766784668, + "learning_rate": 4.279510546318713e-05, + "loss": 0.3232, + "step": 40526 + }, + { + "epoch": 2.812226771216432, + "grad_norm": 0.9523324966430664, + "learning_rate": 4.279049705559558e-05, + "loss": 0.4092, + "step": 40527 + }, + { + "epoch": 2.8122961626535288, + "grad_norm": 0.9072941541671753, + "learning_rate": 4.278588882860778e-05, + "loss": 0.4074, + "step": 40528 + }, + { + "epoch": 2.812365554090625, + "grad_norm": 1.3024795055389404, + "learning_rate": 4.2781280782238296e-05, + "loss": 0.3374, + "step": 40529 + }, + { + "epoch": 2.812434945527722, + "grad_norm": 0.7974620461463928, + "learning_rate": 4.277667291650168e-05, + "loss": 0.2615, + "step": 40530 + }, + { + "epoch": 2.8125043369648184, + "grad_norm": 1.128762125968933, + "learning_rate": 4.277206523141246e-05, + "loss": 0.3315, + "step": 40531 + }, + { + "epoch": 2.812573728401915, + "grad_norm": 1.5646919012069702, + "learning_rate": 4.276745772698523e-05, + "loss": 0.5022, + "step": 40532 + }, + { + "epoch": 2.812643119839012, + "grad_norm": 1.2504119873046875, + "learning_rate": 4.276285040323445e-05, + "loss": 0.449, + "step": 40533 + }, + { + "epoch": 2.8127125112761084, + "grad_norm": 0.7758260369300842, + "learning_rate": 4.2758243260174726e-05, + "loss": 0.2884, + "step": 40534 + }, + { + "epoch": 2.8127819027132053, + "grad_norm": 1.238999843597412, + "learning_rate": 4.2753636297820574e-05, + "loss": 0.497, + "step": 40535 + }, + { + "epoch": 2.8128512941503017, + "grad_norm": 1.2203201055526733, + "learning_rate": 4.274902951618659e-05, + "loss": 0.4058, + "step": 40536 + }, + { + "epoch": 2.8129206855873985, + "grad_norm": 1.8391704559326172, + "learning_rate": 4.274442291528724e-05, + "loss": 0.4267, + "step": 40537 + }, + { + "epoch": 2.8129900770244953, + "grad_norm": 1.0622608661651611, + "learning_rate": 4.273981649513711e-05, + "loss": 0.3384, + "step": 40538 + }, + { + "epoch": 2.8130594684615917, + "grad_norm": 1.2767926454544067, + "learning_rate": 4.2735210255750734e-05, + "loss": 0.3078, + "step": 40539 + }, + { + "epoch": 2.8131288598986885, + "grad_norm": 1.1216481924057007, + "learning_rate": 4.273060419714269e-05, + "loss": 0.3215, + "step": 40540 + }, + { + "epoch": 2.813198251335785, + "grad_norm": 1.0752475261688232, + "learning_rate": 4.2725998319327455e-05, + "loss": 0.3415, + "step": 40541 + }, + { + "epoch": 2.8132676427728818, + "grad_norm": 1.3137935400009155, + "learning_rate": 4.272139262231959e-05, + "loss": 0.5036, + "step": 40542 + }, + { + "epoch": 2.8133370342099786, + "grad_norm": 1.0495301485061646, + "learning_rate": 4.271678710613364e-05, + "loss": 0.2868, + "step": 40543 + }, + { + "epoch": 2.813406425647075, + "grad_norm": 1.0027817487716675, + "learning_rate": 4.2712181770784155e-05, + "loss": 0.2942, + "step": 40544 + }, + { + "epoch": 2.813475817084172, + "grad_norm": 1.7598387002944946, + "learning_rate": 4.270757661628566e-05, + "loss": 0.387, + "step": 40545 + }, + { + "epoch": 2.813545208521268, + "grad_norm": 1.2300364971160889, + "learning_rate": 4.2702971642652735e-05, + "loss": 0.3627, + "step": 40546 + }, + { + "epoch": 2.813614599958365, + "grad_norm": 1.1136384010314941, + "learning_rate": 4.269836684989984e-05, + "loss": 0.335, + "step": 40547 + }, + { + "epoch": 2.813683991395462, + "grad_norm": 0.9352974891662598, + "learning_rate": 4.269376223804157e-05, + "loss": 0.5045, + "step": 40548 + }, + { + "epoch": 2.8137533828325587, + "grad_norm": 0.8644570112228394, + "learning_rate": 4.268915780709243e-05, + "loss": 0.2016, + "step": 40549 + }, + { + "epoch": 2.813822774269655, + "grad_norm": 0.9551741480827332, + "learning_rate": 4.2684553557066965e-05, + "loss": 0.3265, + "step": 40550 + }, + { + "epoch": 2.813892165706752, + "grad_norm": 1.3226828575134277, + "learning_rate": 4.267994948797972e-05, + "loss": 0.4357, + "step": 40551 + }, + { + "epoch": 2.8139615571438483, + "grad_norm": 1.365488052368164, + "learning_rate": 4.267534559984523e-05, + "loss": 0.5138, + "step": 40552 + }, + { + "epoch": 2.814030948580945, + "grad_norm": 1.0726966857910156, + "learning_rate": 4.267074189267806e-05, + "loss": 0.3365, + "step": 40553 + }, + { + "epoch": 2.814100340018042, + "grad_norm": 1.3604909181594849, + "learning_rate": 4.2666138366492656e-05, + "loss": 0.3948, + "step": 40554 + }, + { + "epoch": 2.8141697314551384, + "grad_norm": 1.4685859680175781, + "learning_rate": 4.266153502130361e-05, + "loss": 0.4229, + "step": 40555 + }, + { + "epoch": 2.814239122892235, + "grad_norm": 1.4395182132720947, + "learning_rate": 4.2656931857125446e-05, + "loss": 0.6514, + "step": 40556 + }, + { + "epoch": 2.8143085143293316, + "grad_norm": 1.5940215587615967, + "learning_rate": 4.26523288739727e-05, + "loss": 0.4004, + "step": 40557 + }, + { + "epoch": 2.8143779057664284, + "grad_norm": 1.0942045450210571, + "learning_rate": 4.2647726071859896e-05, + "loss": 0.4721, + "step": 40558 + }, + { + "epoch": 2.8144472972035253, + "grad_norm": 0.9713225960731506, + "learning_rate": 4.264312345080156e-05, + "loss": 0.3745, + "step": 40559 + }, + { + "epoch": 2.8145166886406217, + "grad_norm": 1.0949779748916626, + "learning_rate": 4.263852101081227e-05, + "loss": 0.4001, + "step": 40560 + }, + { + "epoch": 2.8145860800777185, + "grad_norm": 0.9194564819335938, + "learning_rate": 4.263391875190648e-05, + "loss": 0.2741, + "step": 40561 + }, + { + "epoch": 2.814655471514815, + "grad_norm": 1.3965955972671509, + "learning_rate": 4.262931667409874e-05, + "loss": 0.4437, + "step": 40562 + }, + { + "epoch": 2.8147248629519117, + "grad_norm": 1.289391279220581, + "learning_rate": 4.262471477740361e-05, + "loss": 0.5498, + "step": 40563 + }, + { + "epoch": 2.8147942543890085, + "grad_norm": 0.9559727907180786, + "learning_rate": 4.2620113061835585e-05, + "loss": 0.2815, + "step": 40564 + }, + { + "epoch": 2.814863645826105, + "grad_norm": 1.1393247842788696, + "learning_rate": 4.2615511527409214e-05, + "loss": 0.4477, + "step": 40565 + }, + { + "epoch": 2.8149330372632018, + "grad_norm": 0.9621694684028625, + "learning_rate": 4.2610910174139006e-05, + "loss": 0.4035, + "step": 40566 + }, + { + "epoch": 2.815002428700298, + "grad_norm": 1.3574351072311401, + "learning_rate": 4.260630900203954e-05, + "loss": 0.4012, + "step": 40567 + }, + { + "epoch": 2.815071820137395, + "grad_norm": 1.1519966125488281, + "learning_rate": 4.2601708011125254e-05, + "loss": 0.4448, + "step": 40568 + }, + { + "epoch": 2.815141211574492, + "grad_norm": 1.0848079919815063, + "learning_rate": 4.2597107201410725e-05, + "loss": 0.4615, + "step": 40569 + }, + { + "epoch": 2.815210603011588, + "grad_norm": 0.9629559516906738, + "learning_rate": 4.2592506572910465e-05, + "loss": 0.2885, + "step": 40570 + }, + { + "epoch": 2.815279994448685, + "grad_norm": 1.0055038928985596, + "learning_rate": 4.258790612563903e-05, + "loss": 0.3993, + "step": 40571 + }, + { + "epoch": 2.8153493858857814, + "grad_norm": 1.464532732963562, + "learning_rate": 4.258330585961089e-05, + "loss": 0.4863, + "step": 40572 + }, + { + "epoch": 2.8154187773228783, + "grad_norm": 1.3349088430404663, + "learning_rate": 4.2578705774840586e-05, + "loss": 0.382, + "step": 40573 + }, + { + "epoch": 2.815488168759975, + "grad_norm": 1.1030787229537964, + "learning_rate": 4.257410587134269e-05, + "loss": 0.3528, + "step": 40574 + }, + { + "epoch": 2.815557560197072, + "grad_norm": 1.2139090299606323, + "learning_rate": 4.256950614913163e-05, + "loss": 0.3715, + "step": 40575 + }, + { + "epoch": 2.8156269516341683, + "grad_norm": 1.516327977180481, + "learning_rate": 4.256490660822199e-05, + "loss": 0.6588, + "step": 40576 + }, + { + "epoch": 2.815696343071265, + "grad_norm": 1.1971017122268677, + "learning_rate": 4.256030724862826e-05, + "loss": 0.4307, + "step": 40577 + }, + { + "epoch": 2.8157657345083615, + "grad_norm": 0.9307875633239746, + "learning_rate": 4.2555708070364984e-05, + "loss": 0.358, + "step": 40578 + }, + { + "epoch": 2.8158351259454584, + "grad_norm": 1.1740412712097168, + "learning_rate": 4.255110907344667e-05, + "loss": 0.4351, + "step": 40579 + }, + { + "epoch": 2.815904517382555, + "grad_norm": 1.1693792343139648, + "learning_rate": 4.254651025788785e-05, + "loss": 0.3921, + "step": 40580 + }, + { + "epoch": 2.8159739088196516, + "grad_norm": 1.1723034381866455, + "learning_rate": 4.254191162370306e-05, + "loss": 0.3393, + "step": 40581 + }, + { + "epoch": 2.8160433002567484, + "grad_norm": 0.8008821606636047, + "learning_rate": 4.2537313170906745e-05, + "loss": 0.2793, + "step": 40582 + }, + { + "epoch": 2.816112691693845, + "grad_norm": 1.606734275817871, + "learning_rate": 4.253271489951347e-05, + "loss": 0.4766, + "step": 40583 + }, + { + "epoch": 2.8161820831309416, + "grad_norm": 1.3636674880981445, + "learning_rate": 4.2528116809537754e-05, + "loss": 0.3605, + "step": 40584 + }, + { + "epoch": 2.8162514745680385, + "grad_norm": 0.8996577262878418, + "learning_rate": 4.25235189009941e-05, + "loss": 0.3398, + "step": 40585 + }, + { + "epoch": 2.816320866005135, + "grad_norm": 1.879988431930542, + "learning_rate": 4.251892117389703e-05, + "loss": 0.718, + "step": 40586 + }, + { + "epoch": 2.8163902574422317, + "grad_norm": 1.181509017944336, + "learning_rate": 4.251432362826106e-05, + "loss": 0.3315, + "step": 40587 + }, + { + "epoch": 2.816459648879328, + "grad_norm": 1.130831003189087, + "learning_rate": 4.250972626410073e-05, + "loss": 0.4819, + "step": 40588 + }, + { + "epoch": 2.816529040316425, + "grad_norm": 1.435033917427063, + "learning_rate": 4.250512908143048e-05, + "loss": 0.4532, + "step": 40589 + }, + { + "epoch": 2.8165984317535218, + "grad_norm": 1.3058085441589355, + "learning_rate": 4.250053208026489e-05, + "loss": 0.479, + "step": 40590 + }, + { + "epoch": 2.816667823190618, + "grad_norm": 1.4719889163970947, + "learning_rate": 4.2495935260618434e-05, + "loss": 0.4969, + "step": 40591 + }, + { + "epoch": 2.816737214627715, + "grad_norm": 1.027759075164795, + "learning_rate": 4.2491338622505636e-05, + "loss": 0.3551, + "step": 40592 + }, + { + "epoch": 2.8168066060648114, + "grad_norm": 1.295147180557251, + "learning_rate": 4.248674216594102e-05, + "loss": 0.3918, + "step": 40593 + }, + { + "epoch": 2.816875997501908, + "grad_norm": 1.4930707216262817, + "learning_rate": 4.248214589093907e-05, + "loss": 0.5261, + "step": 40594 + }, + { + "epoch": 2.816945388939005, + "grad_norm": 1.0705692768096924, + "learning_rate": 4.247754979751437e-05, + "loss": 0.4141, + "step": 40595 + }, + { + "epoch": 2.8170147803761014, + "grad_norm": 0.8503950834274292, + "learning_rate": 4.247295388568132e-05, + "loss": 0.2929, + "step": 40596 + }, + { + "epoch": 2.8170841718131983, + "grad_norm": 1.604600191116333, + "learning_rate": 4.246835815545448e-05, + "loss": 0.4416, + "step": 40597 + }, + { + "epoch": 2.8171535632502946, + "grad_norm": 1.0729941129684448, + "learning_rate": 4.246376260684837e-05, + "loss": 0.2821, + "step": 40598 + }, + { + "epoch": 2.8172229546873915, + "grad_norm": 0.9165025353431702, + "learning_rate": 4.245916723987747e-05, + "loss": 0.3566, + "step": 40599 + }, + { + "epoch": 2.8172923461244883, + "grad_norm": 1.002756118774414, + "learning_rate": 4.2454572054556317e-05, + "loss": 0.4612, + "step": 40600 + }, + { + "epoch": 2.817361737561585, + "grad_norm": 1.2560957670211792, + "learning_rate": 4.24499770508994e-05, + "loss": 0.4093, + "step": 40601 + }, + { + "epoch": 2.8174311289986815, + "grad_norm": 0.8814780712127686, + "learning_rate": 4.244538222892126e-05, + "loss": 0.2053, + "step": 40602 + }, + { + "epoch": 2.8175005204357784, + "grad_norm": 1.277687430381775, + "learning_rate": 4.2440787588636344e-05, + "loss": 0.4053, + "step": 40603 + }, + { + "epoch": 2.8175699118728748, + "grad_norm": 1.0808513164520264, + "learning_rate": 4.243619313005918e-05, + "loss": 0.4398, + "step": 40604 + }, + { + "epoch": 2.8176393033099716, + "grad_norm": 1.3577687740325928, + "learning_rate": 4.243159885320427e-05, + "loss": 0.328, + "step": 40605 + }, + { + "epoch": 2.8177086947470684, + "grad_norm": 1.1217947006225586, + "learning_rate": 4.242700475808616e-05, + "loss": 0.3747, + "step": 40606 + }, + { + "epoch": 2.817778086184165, + "grad_norm": 1.1669107675552368, + "learning_rate": 4.242241084471928e-05, + "loss": 0.3479, + "step": 40607 + }, + { + "epoch": 2.8178474776212616, + "grad_norm": 0.7867707014083862, + "learning_rate": 4.241781711311817e-05, + "loss": 0.3643, + "step": 40608 + }, + { + "epoch": 2.817916869058358, + "grad_norm": 0.7833222150802612, + "learning_rate": 4.2413223563297366e-05, + "loss": 0.2682, + "step": 40609 + }, + { + "epoch": 2.817986260495455, + "grad_norm": 1.2678951025009155, + "learning_rate": 4.2408630195271304e-05, + "loss": 0.4567, + "step": 40610 + }, + { + "epoch": 2.8180556519325517, + "grad_norm": 1.626897931098938, + "learning_rate": 4.240403700905451e-05, + "loss": 0.4292, + "step": 40611 + }, + { + "epoch": 2.818125043369648, + "grad_norm": 0.8967178463935852, + "learning_rate": 4.239944400466149e-05, + "loss": 0.3431, + "step": 40612 + }, + { + "epoch": 2.818194434806745, + "grad_norm": 1.0976269245147705, + "learning_rate": 4.2394851182106746e-05, + "loss": 0.3342, + "step": 40613 + }, + { + "epoch": 2.8182638262438413, + "grad_norm": 1.358251929283142, + "learning_rate": 4.239025854140476e-05, + "loss": 0.4427, + "step": 40614 + }, + { + "epoch": 2.818333217680938, + "grad_norm": 1.0615648031234741, + "learning_rate": 4.2385666082570097e-05, + "loss": 0.3942, + "step": 40615 + }, + { + "epoch": 2.818402609118035, + "grad_norm": 1.2843594551086426, + "learning_rate": 4.2381073805617155e-05, + "loss": 0.418, + "step": 40616 + }, + { + "epoch": 2.8184720005551314, + "grad_norm": 0.8204070329666138, + "learning_rate": 4.237648171056048e-05, + "loss": 0.3067, + "step": 40617 + }, + { + "epoch": 2.818541391992228, + "grad_norm": 1.7804127931594849, + "learning_rate": 4.237188979741457e-05, + "loss": 0.523, + "step": 40618 + }, + { + "epoch": 2.8186107834293246, + "grad_norm": 1.0276509523391724, + "learning_rate": 4.236729806619391e-05, + "loss": 0.3202, + "step": 40619 + }, + { + "epoch": 2.8186801748664214, + "grad_norm": 1.2310141324996948, + "learning_rate": 4.2362706516913006e-05, + "loss": 0.3374, + "step": 40620 + }, + { + "epoch": 2.8187495663035183, + "grad_norm": 1.0563675165176392, + "learning_rate": 4.2358115149586355e-05, + "loss": 0.3715, + "step": 40621 + }, + { + "epoch": 2.818818957740615, + "grad_norm": 0.866023063659668, + "learning_rate": 4.235352396422848e-05, + "loss": 0.2744, + "step": 40622 + }, + { + "epoch": 2.8188883491777115, + "grad_norm": 0.9788052439689636, + "learning_rate": 4.2348932960853804e-05, + "loss": 0.3234, + "step": 40623 + }, + { + "epoch": 2.818957740614808, + "grad_norm": 0.9693113565444946, + "learning_rate": 4.234434213947686e-05, + "loss": 0.3699, + "step": 40624 + }, + { + "epoch": 2.8190271320519047, + "grad_norm": 1.3239314556121826, + "learning_rate": 4.2339751500112135e-05, + "loss": 0.3707, + "step": 40625 + }, + { + "epoch": 2.8190965234890015, + "grad_norm": 0.9686722159385681, + "learning_rate": 4.233516104277413e-05, + "loss": 0.4159, + "step": 40626 + }, + { + "epoch": 2.8191659149260984, + "grad_norm": 1.2129194736480713, + "learning_rate": 4.233057076747733e-05, + "loss": 0.3003, + "step": 40627 + }, + { + "epoch": 2.8192353063631947, + "grad_norm": 1.5911064147949219, + "learning_rate": 4.232598067423623e-05, + "loss": 0.593, + "step": 40628 + }, + { + "epoch": 2.8193046978002916, + "grad_norm": 1.3064069747924805, + "learning_rate": 4.232139076306535e-05, + "loss": 0.4521, + "step": 40629 + }, + { + "epoch": 2.819374089237388, + "grad_norm": 1.1423407793045044, + "learning_rate": 4.231680103397912e-05, + "loss": 0.3442, + "step": 40630 + }, + { + "epoch": 2.819443480674485, + "grad_norm": 1.417296290397644, + "learning_rate": 4.231221148699204e-05, + "loss": 0.416, + "step": 40631 + }, + { + "epoch": 2.8195128721115816, + "grad_norm": 1.3477782011032104, + "learning_rate": 4.230762212211862e-05, + "loss": 0.4575, + "step": 40632 + }, + { + "epoch": 2.819582263548678, + "grad_norm": 1.1568481922149658, + "learning_rate": 4.230303293937334e-05, + "loss": 0.3743, + "step": 40633 + }, + { + "epoch": 2.819651654985775, + "grad_norm": 1.3674501180648804, + "learning_rate": 4.229844393877071e-05, + "loss": 0.4228, + "step": 40634 + }, + { + "epoch": 2.8197210464228712, + "grad_norm": 0.9955394268035889, + "learning_rate": 4.229385512032518e-05, + "loss": 0.3882, + "step": 40635 + }, + { + "epoch": 2.819790437859968, + "grad_norm": 1.3061883449554443, + "learning_rate": 4.228926648405129e-05, + "loss": 0.4464, + "step": 40636 + }, + { + "epoch": 2.819859829297065, + "grad_norm": 1.6720423698425293, + "learning_rate": 4.2284678029963464e-05, + "loss": 0.4336, + "step": 40637 + }, + { + "epoch": 2.8199292207341613, + "grad_norm": 1.8251550197601318, + "learning_rate": 4.228008975807621e-05, + "loss": 0.2983, + "step": 40638 + }, + { + "epoch": 2.819998612171258, + "grad_norm": 1.4058607816696167, + "learning_rate": 4.2275501668404014e-05, + "loss": 0.417, + "step": 40639 + }, + { + "epoch": 2.8200680036083545, + "grad_norm": 1.1224762201309204, + "learning_rate": 4.227091376096136e-05, + "loss": 0.4331, + "step": 40640 + }, + { + "epoch": 2.8201373950454514, + "grad_norm": 1.070090651512146, + "learning_rate": 4.2266326035762785e-05, + "loss": 0.3137, + "step": 40641 + }, + { + "epoch": 2.820206786482548, + "grad_norm": 1.1907544136047363, + "learning_rate": 4.226173849282267e-05, + "loss": 0.339, + "step": 40642 + }, + { + "epoch": 2.8202761779196446, + "grad_norm": 1.1948696374893188, + "learning_rate": 4.225715113215555e-05, + "loss": 0.3888, + "step": 40643 + }, + { + "epoch": 2.8203455693567414, + "grad_norm": 1.076969027519226, + "learning_rate": 4.225256395377594e-05, + "loss": 0.5025, + "step": 40644 + }, + { + "epoch": 2.820414960793838, + "grad_norm": 1.3555614948272705, + "learning_rate": 4.224797695769824e-05, + "loss": 0.4644, + "step": 40645 + }, + { + "epoch": 2.8204843522309346, + "grad_norm": 1.3615690469741821, + "learning_rate": 4.224339014393699e-05, + "loss": 0.4155, + "step": 40646 + }, + { + "epoch": 2.8205537436680315, + "grad_norm": 1.095492959022522, + "learning_rate": 4.223880351250665e-05, + "loss": 0.3663, + "step": 40647 + }, + { + "epoch": 2.8206231351051283, + "grad_norm": 0.9983401298522949, + "learning_rate": 4.223421706342171e-05, + "loss": 0.3313, + "step": 40648 + }, + { + "epoch": 2.8206925265422247, + "grad_norm": 0.9645739793777466, + "learning_rate": 4.222963079669664e-05, + "loss": 0.3629, + "step": 40649 + }, + { + "epoch": 2.820761917979321, + "grad_norm": 1.3153131008148193, + "learning_rate": 4.2225044712345964e-05, + "loss": 0.3852, + "step": 40650 + }, + { + "epoch": 2.820831309416418, + "grad_norm": 1.648358702659607, + "learning_rate": 4.222045881038408e-05, + "loss": 0.4949, + "step": 40651 + }, + { + "epoch": 2.8209007008535147, + "grad_norm": 1.267121434211731, + "learning_rate": 4.221587309082551e-05, + "loss": 0.2995, + "step": 40652 + }, + { + "epoch": 2.8209700922906116, + "grad_norm": 1.1515334844589233, + "learning_rate": 4.221128755368471e-05, + "loss": 0.4174, + "step": 40653 + }, + { + "epoch": 2.821039483727708, + "grad_norm": 0.7873069047927856, + "learning_rate": 4.220670219897618e-05, + "loss": 0.3029, + "step": 40654 + }, + { + "epoch": 2.821108875164805, + "grad_norm": 1.3959084749221802, + "learning_rate": 4.220211702671438e-05, + "loss": 0.4101, + "step": 40655 + }, + { + "epoch": 2.821178266601901, + "grad_norm": 1.2487590312957764, + "learning_rate": 4.21975320369138e-05, + "loss": 0.3079, + "step": 40656 + }, + { + "epoch": 2.821247658038998, + "grad_norm": 1.068647861480713, + "learning_rate": 4.219294722958894e-05, + "loss": 0.4128, + "step": 40657 + }, + { + "epoch": 2.821317049476095, + "grad_norm": 1.0513546466827393, + "learning_rate": 4.2188362604754196e-05, + "loss": 0.3987, + "step": 40658 + }, + { + "epoch": 2.8213864409131912, + "grad_norm": 0.8727914094924927, + "learning_rate": 4.218377816242409e-05, + "loss": 0.3119, + "step": 40659 + }, + { + "epoch": 2.821455832350288, + "grad_norm": 1.5041768550872803, + "learning_rate": 4.217919390261309e-05, + "loss": 0.6019, + "step": 40660 + }, + { + "epoch": 2.8215252237873845, + "grad_norm": 1.9287854433059692, + "learning_rate": 4.2174609825335655e-05, + "loss": 0.5471, + "step": 40661 + }, + { + "epoch": 2.8215946152244813, + "grad_norm": 1.871135950088501, + "learning_rate": 4.217002593060627e-05, + "loss": 0.3925, + "step": 40662 + }, + { + "epoch": 2.821664006661578, + "grad_norm": 1.0678807497024536, + "learning_rate": 4.216544221843941e-05, + "loss": 0.4888, + "step": 40663 + }, + { + "epoch": 2.8217333980986745, + "grad_norm": 1.8635848760604858, + "learning_rate": 4.216085868884958e-05, + "loss": 0.3782, + "step": 40664 + }, + { + "epoch": 2.8218027895357713, + "grad_norm": 1.177426815032959, + "learning_rate": 4.215627534185117e-05, + "loss": 0.5042, + "step": 40665 + }, + { + "epoch": 2.8218721809728677, + "grad_norm": 1.025967001914978, + "learning_rate": 4.215169217745868e-05, + "loss": 0.2936, + "step": 40666 + }, + { + "epoch": 2.8219415724099646, + "grad_norm": 0.9035949110984802, + "learning_rate": 4.2147109195686586e-05, + "loss": 0.2565, + "step": 40667 + }, + { + "epoch": 2.8220109638470614, + "grad_norm": 1.0353423357009888, + "learning_rate": 4.214252639654937e-05, + "loss": 0.3829, + "step": 40668 + }, + { + "epoch": 2.822080355284158, + "grad_norm": 1.5346276760101318, + "learning_rate": 4.213794378006146e-05, + "loss": 0.3526, + "step": 40669 + }, + { + "epoch": 2.8221497467212546, + "grad_norm": 0.941085696220398, + "learning_rate": 4.213336134623737e-05, + "loss": 0.3945, + "step": 40670 + }, + { + "epoch": 2.822219138158351, + "grad_norm": 1.1162278652191162, + "learning_rate": 4.2128779095091576e-05, + "loss": 0.2931, + "step": 40671 + }, + { + "epoch": 2.822288529595448, + "grad_norm": 0.884527862071991, + "learning_rate": 4.2124197026638465e-05, + "loss": 0.3043, + "step": 40672 + }, + { + "epoch": 2.8223579210325447, + "grad_norm": 0.8792065978050232, + "learning_rate": 4.2119615140892564e-05, + "loss": 0.2795, + "step": 40673 + }, + { + "epoch": 2.8224273124696415, + "grad_norm": 1.14231538772583, + "learning_rate": 4.2115033437868325e-05, + "loss": 0.3768, + "step": 40674 + }, + { + "epoch": 2.822496703906738, + "grad_norm": 1.3964205980300903, + "learning_rate": 4.211045191758023e-05, + "loss": 0.5472, + "step": 40675 + }, + { + "epoch": 2.8225660953438347, + "grad_norm": 1.348993182182312, + "learning_rate": 4.210587058004269e-05, + "loss": 0.5578, + "step": 40676 + }, + { + "epoch": 2.822635486780931, + "grad_norm": 1.1312577724456787, + "learning_rate": 4.2101289425270204e-05, + "loss": 0.3325, + "step": 40677 + }, + { + "epoch": 2.822704878218028, + "grad_norm": 1.376461386680603, + "learning_rate": 4.2096708453277235e-05, + "loss": 0.4521, + "step": 40678 + }, + { + "epoch": 2.822774269655125, + "grad_norm": 1.8415775299072266, + "learning_rate": 4.209212766407826e-05, + "loss": 0.4136, + "step": 40679 + }, + { + "epoch": 2.822843661092221, + "grad_norm": 1.4683104753494263, + "learning_rate": 4.20875470576877e-05, + "loss": 0.5216, + "step": 40680 + }, + { + "epoch": 2.822913052529318, + "grad_norm": 0.841262936592102, + "learning_rate": 4.2082966634120025e-05, + "loss": 0.2933, + "step": 40681 + }, + { + "epoch": 2.8229824439664144, + "grad_norm": 1.211642861366272, + "learning_rate": 4.2078386393389716e-05, + "loss": 0.4425, + "step": 40682 + }, + { + "epoch": 2.8230518354035112, + "grad_norm": 1.603870153427124, + "learning_rate": 4.20738063355112e-05, + "loss": 0.3908, + "step": 40683 + }, + { + "epoch": 2.823121226840608, + "grad_norm": 1.386137843132019, + "learning_rate": 4.2069226460499014e-05, + "loss": 0.461, + "step": 40684 + }, + { + "epoch": 2.8231906182777045, + "grad_norm": 1.0998096466064453, + "learning_rate": 4.206464676836751e-05, + "loss": 0.3424, + "step": 40685 + }, + { + "epoch": 2.8232600097148013, + "grad_norm": 0.8361411094665527, + "learning_rate": 4.2060067259131194e-05, + "loss": 0.301, + "step": 40686 + }, + { + "epoch": 2.8233294011518977, + "grad_norm": 1.25791597366333, + "learning_rate": 4.205548793280453e-05, + "loss": 0.2755, + "step": 40687 + }, + { + "epoch": 2.8233987925889945, + "grad_norm": 1.5746963024139404, + "learning_rate": 4.205090878940197e-05, + "loss": 0.6159, + "step": 40688 + }, + { + "epoch": 2.8234681840260913, + "grad_norm": 1.4124298095703125, + "learning_rate": 4.2046329828937955e-05, + "loss": 0.3674, + "step": 40689 + }, + { + "epoch": 2.8235375754631877, + "grad_norm": 2.0604469776153564, + "learning_rate": 4.2041751051426955e-05, + "loss": 0.5551, + "step": 40690 + }, + { + "epoch": 2.8236069669002846, + "grad_norm": 1.5584282875061035, + "learning_rate": 4.203717245688347e-05, + "loss": 0.431, + "step": 40691 + }, + { + "epoch": 2.823676358337381, + "grad_norm": 0.912356972694397, + "learning_rate": 4.203259404532187e-05, + "loss": 0.2468, + "step": 40692 + }, + { + "epoch": 2.823745749774478, + "grad_norm": 0.9617254734039307, + "learning_rate": 4.202801581675664e-05, + "loss": 0.3219, + "step": 40693 + }, + { + "epoch": 2.8238151412115746, + "grad_norm": 1.4486050605773926, + "learning_rate": 4.202343777120224e-05, + "loss": 0.4212, + "step": 40694 + }, + { + "epoch": 2.823884532648671, + "grad_norm": 0.7036596536636353, + "learning_rate": 4.201885990867311e-05, + "loss": 0.2984, + "step": 40695 + }, + { + "epoch": 2.823953924085768, + "grad_norm": 1.2917520999908447, + "learning_rate": 4.201428222918372e-05, + "loss": 0.4321, + "step": 40696 + }, + { + "epoch": 2.8240233155228642, + "grad_norm": 1.3423084020614624, + "learning_rate": 4.200970473274852e-05, + "loss": 0.54, + "step": 40697 + }, + { + "epoch": 2.824092706959961, + "grad_norm": 0.9511104822158813, + "learning_rate": 4.2005127419381986e-05, + "loss": 0.3293, + "step": 40698 + }, + { + "epoch": 2.824162098397058, + "grad_norm": 1.2687031030654907, + "learning_rate": 4.200055028909849e-05, + "loss": 0.4033, + "step": 40699 + }, + { + "epoch": 2.8242314898341547, + "grad_norm": 0.9954302310943604, + "learning_rate": 4.199597334191254e-05, + "loss": 0.2849, + "step": 40700 + }, + { + "epoch": 2.824300881271251, + "grad_norm": 1.154538869857788, + "learning_rate": 4.199139657783857e-05, + "loss": 0.372, + "step": 40701 + }, + { + "epoch": 2.824370272708348, + "grad_norm": 1.3420048952102661, + "learning_rate": 4.198681999689103e-05, + "loss": 0.3731, + "step": 40702 + }, + { + "epoch": 2.8244396641454443, + "grad_norm": 1.7181273698806763, + "learning_rate": 4.198224359908436e-05, + "loss": 0.6249, + "step": 40703 + }, + { + "epoch": 2.824509055582541, + "grad_norm": 1.2380880117416382, + "learning_rate": 4.197766738443303e-05, + "loss": 0.3285, + "step": 40704 + }, + { + "epoch": 2.824578447019638, + "grad_norm": 1.3948523998260498, + "learning_rate": 4.19730913529515e-05, + "loss": 0.4194, + "step": 40705 + }, + { + "epoch": 2.8246478384567344, + "grad_norm": 0.7838547825813293, + "learning_rate": 4.196851550465415e-05, + "loss": 0.2729, + "step": 40706 + }, + { + "epoch": 2.8247172298938312, + "grad_norm": 1.2660187482833862, + "learning_rate": 4.196393983955547e-05, + "loss": 0.3076, + "step": 40707 + }, + { + "epoch": 2.8247866213309276, + "grad_norm": 0.8723859190940857, + "learning_rate": 4.195936435766989e-05, + "loss": 0.3259, + "step": 40708 + }, + { + "epoch": 2.8248560127680244, + "grad_norm": 1.3087555170059204, + "learning_rate": 4.195478905901188e-05, + "loss": 0.4559, + "step": 40709 + }, + { + "epoch": 2.8249254042051213, + "grad_norm": 1.188372254371643, + "learning_rate": 4.195021394359588e-05, + "loss": 0.389, + "step": 40710 + }, + { + "epoch": 2.8249947956422177, + "grad_norm": 1.1786277294158936, + "learning_rate": 4.19456390114363e-05, + "loss": 0.3533, + "step": 40711 + }, + { + "epoch": 2.8250641870793145, + "grad_norm": 0.90833580493927, + "learning_rate": 4.194106426254759e-05, + "loss": 0.2776, + "step": 40712 + }, + { + "epoch": 2.825133578516411, + "grad_norm": 1.29411780834198, + "learning_rate": 4.1936489696944215e-05, + "loss": 0.427, + "step": 40713 + }, + { + "epoch": 2.8252029699535077, + "grad_norm": 1.0411512851715088, + "learning_rate": 4.193191531464063e-05, + "loss": 0.2886, + "step": 40714 + }, + { + "epoch": 2.8252723613906046, + "grad_norm": 1.2081876993179321, + "learning_rate": 4.192734111565123e-05, + "loss": 0.3505, + "step": 40715 + }, + { + "epoch": 2.825341752827701, + "grad_norm": 1.1298444271087646, + "learning_rate": 4.1922767099990465e-05, + "loss": 0.3565, + "step": 40716 + }, + { + "epoch": 2.825411144264798, + "grad_norm": 0.8671739101409912, + "learning_rate": 4.191819326767279e-05, + "loss": 0.3316, + "step": 40717 + }, + { + "epoch": 2.825480535701894, + "grad_norm": 1.0871243476867676, + "learning_rate": 4.191361961871264e-05, + "loss": 0.3106, + "step": 40718 + }, + { + "epoch": 2.825549927138991, + "grad_norm": 1.2568042278289795, + "learning_rate": 4.1909046153124496e-05, + "loss": 0.2876, + "step": 40719 + }, + { + "epoch": 2.825619318576088, + "grad_norm": 1.5616520643234253, + "learning_rate": 4.190447287092271e-05, + "loss": 0.486, + "step": 40720 + }, + { + "epoch": 2.8256887100131842, + "grad_norm": 1.1978487968444824, + "learning_rate": 4.1899899772121766e-05, + "loss": 0.3606, + "step": 40721 + }, + { + "epoch": 2.825758101450281, + "grad_norm": 0.9792284369468689, + "learning_rate": 4.189532685673609e-05, + "loss": 0.3497, + "step": 40722 + }, + { + "epoch": 2.8258274928873774, + "grad_norm": 1.4922490119934082, + "learning_rate": 4.1890754124780126e-05, + "loss": 0.5537, + "step": 40723 + }, + { + "epoch": 2.8258968843244743, + "grad_norm": 2.3995940685272217, + "learning_rate": 4.188618157626832e-05, + "loss": 0.3788, + "step": 40724 + }, + { + "epoch": 2.825966275761571, + "grad_norm": 1.5476261377334595, + "learning_rate": 4.188160921121509e-05, + "loss": 0.5345, + "step": 40725 + }, + { + "epoch": 2.826035667198668, + "grad_norm": 0.900413990020752, + "learning_rate": 4.187703702963491e-05, + "loss": 0.282, + "step": 40726 + }, + { + "epoch": 2.8261050586357643, + "grad_norm": 1.0312319993972778, + "learning_rate": 4.187246503154215e-05, + "loss": 0.2631, + "step": 40727 + }, + { + "epoch": 2.826174450072861, + "grad_norm": 1.007325291633606, + "learning_rate": 4.186789321695127e-05, + "loss": 0.3752, + "step": 40728 + }, + { + "epoch": 2.8262438415099576, + "grad_norm": 1.5199042558670044, + "learning_rate": 4.186332158587671e-05, + "loss": 0.3225, + "step": 40729 + }, + { + "epoch": 2.8263132329470544, + "grad_norm": 1.2280395030975342, + "learning_rate": 4.1858750138332904e-05, + "loss": 0.5123, + "step": 40730 + }, + { + "epoch": 2.826382624384151, + "grad_norm": 1.1134793758392334, + "learning_rate": 4.185417887433427e-05, + "loss": 0.3105, + "step": 40731 + }, + { + "epoch": 2.8264520158212476, + "grad_norm": 1.3794366121292114, + "learning_rate": 4.184960779389524e-05, + "loss": 0.3658, + "step": 40732 + }, + { + "epoch": 2.8265214072583444, + "grad_norm": 0.9871639013290405, + "learning_rate": 4.1845036897030295e-05, + "loss": 0.2682, + "step": 40733 + }, + { + "epoch": 2.826590798695441, + "grad_norm": 0.8506274819374084, + "learning_rate": 4.184046618375379e-05, + "loss": 0.316, + "step": 40734 + }, + { + "epoch": 2.8266601901325377, + "grad_norm": 1.2100406885147095, + "learning_rate": 4.183589565408018e-05, + "loss": 0.4066, + "step": 40735 + }, + { + "epoch": 2.8267295815696345, + "grad_norm": 1.411473035812378, + "learning_rate": 4.1831325308023906e-05, + "loss": 0.5387, + "step": 40736 + }, + { + "epoch": 2.826798973006731, + "grad_norm": 1.2594175338745117, + "learning_rate": 4.182675514559938e-05, + "loss": 0.3667, + "step": 40737 + }, + { + "epoch": 2.8268683644438277, + "grad_norm": 1.0284987688064575, + "learning_rate": 4.182218516682105e-05, + "loss": 0.3189, + "step": 40738 + }, + { + "epoch": 2.826937755880924, + "grad_norm": 1.263206958770752, + "learning_rate": 4.1817615371703325e-05, + "loss": 0.2906, + "step": 40739 + }, + { + "epoch": 2.827007147318021, + "grad_norm": 1.2510159015655518, + "learning_rate": 4.1813045760260674e-05, + "loss": 0.3546, + "step": 40740 + }, + { + "epoch": 2.8270765387551178, + "grad_norm": 1.7249655723571777, + "learning_rate": 4.180847633250744e-05, + "loss": 0.4626, + "step": 40741 + }, + { + "epoch": 2.827145930192214, + "grad_norm": 0.9206743240356445, + "learning_rate": 4.1803907088458104e-05, + "loss": 0.3574, + "step": 40742 + }, + { + "epoch": 2.827215321629311, + "grad_norm": 1.620676875114441, + "learning_rate": 4.1799338028127086e-05, + "loss": 0.4528, + "step": 40743 + }, + { + "epoch": 2.8272847130664074, + "grad_norm": 1.148112177848816, + "learning_rate": 4.1794769151528804e-05, + "loss": 0.3418, + "step": 40744 + }, + { + "epoch": 2.827354104503504, + "grad_norm": 1.2357498407363892, + "learning_rate": 4.179020045867771e-05, + "loss": 0.5218, + "step": 40745 + }, + { + "epoch": 2.827423495940601, + "grad_norm": 1.2076917886734009, + "learning_rate": 4.1785631949588166e-05, + "loss": 0.4705, + "step": 40746 + }, + { + "epoch": 2.8274928873776974, + "grad_norm": 1.0506373643875122, + "learning_rate": 4.178106362427463e-05, + "loss": 0.4027, + "step": 40747 + }, + { + "epoch": 2.8275622788147943, + "grad_norm": 1.7171645164489746, + "learning_rate": 4.177649548275153e-05, + "loss": 0.5999, + "step": 40748 + }, + { + "epoch": 2.8276316702518907, + "grad_norm": 1.2205616235733032, + "learning_rate": 4.1771927525033304e-05, + "loss": 0.3968, + "step": 40749 + }, + { + "epoch": 2.8277010616889875, + "grad_norm": 1.4513925313949585, + "learning_rate": 4.1767359751134313e-05, + "loss": 0.4877, + "step": 40750 + }, + { + "epoch": 2.8277704531260843, + "grad_norm": 1.0043132305145264, + "learning_rate": 4.176279216106902e-05, + "loss": 0.2764, + "step": 40751 + }, + { + "epoch": 2.827839844563181, + "grad_norm": 1.191680669784546, + "learning_rate": 4.1758224754851824e-05, + "loss": 0.3934, + "step": 40752 + }, + { + "epoch": 2.8279092360002775, + "grad_norm": 1.2696468830108643, + "learning_rate": 4.175365753249716e-05, + "loss": 0.3327, + "step": 40753 + }, + { + "epoch": 2.8279786274373744, + "grad_norm": 1.1868504285812378, + "learning_rate": 4.1749090494019474e-05, + "loss": 0.3432, + "step": 40754 + }, + { + "epoch": 2.8280480188744708, + "grad_norm": 1.2071589231491089, + "learning_rate": 4.1744523639433116e-05, + "loss": 0.4215, + "step": 40755 + }, + { + "epoch": 2.8281174103115676, + "grad_norm": 0.8963668346405029, + "learning_rate": 4.1739956968752534e-05, + "loss": 0.2155, + "step": 40756 + }, + { + "epoch": 2.8281868017486644, + "grad_norm": 1.2191400527954102, + "learning_rate": 4.173539048199215e-05, + "loss": 0.4178, + "step": 40757 + }, + { + "epoch": 2.828256193185761, + "grad_norm": 1.3029115200042725, + "learning_rate": 4.17308241791664e-05, + "loss": 0.4096, + "step": 40758 + }, + { + "epoch": 2.8283255846228577, + "grad_norm": 1.242087960243225, + "learning_rate": 4.172625806028965e-05, + "loss": 0.3312, + "step": 40759 + }, + { + "epoch": 2.828394976059954, + "grad_norm": 1.3351014852523804, + "learning_rate": 4.172169212537639e-05, + "loss": 0.5402, + "step": 40760 + }, + { + "epoch": 2.828464367497051, + "grad_norm": 1.2714438438415527, + "learning_rate": 4.171712637444095e-05, + "loss": 0.4721, + "step": 40761 + }, + { + "epoch": 2.8285337589341477, + "grad_norm": 2.1231350898742676, + "learning_rate": 4.171256080749777e-05, + "loss": 0.3323, + "step": 40762 + }, + { + "epoch": 2.828603150371244, + "grad_norm": 1.8488651514053345, + "learning_rate": 4.1707995424561286e-05, + "loss": 0.3372, + "step": 40763 + }, + { + "epoch": 2.828672541808341, + "grad_norm": 2.0679397583007812, + "learning_rate": 4.170343022564589e-05, + "loss": 0.3619, + "step": 40764 + }, + { + "epoch": 2.8287419332454373, + "grad_norm": 1.3461856842041016, + "learning_rate": 4.1698865210766e-05, + "loss": 0.4074, + "step": 40765 + }, + { + "epoch": 2.828811324682534, + "grad_norm": 0.9826362729072571, + "learning_rate": 4.169430037993604e-05, + "loss": 0.2974, + "step": 40766 + }, + { + "epoch": 2.828880716119631, + "grad_norm": 1.1898256540298462, + "learning_rate": 4.168973573317043e-05, + "loss": 0.2941, + "step": 40767 + }, + { + "epoch": 2.8289501075567274, + "grad_norm": 1.345017671585083, + "learning_rate": 4.168517127048354e-05, + "loss": 0.3119, + "step": 40768 + }, + { + "epoch": 2.829019498993824, + "grad_norm": 1.705016851425171, + "learning_rate": 4.168060699188978e-05, + "loss": 0.4917, + "step": 40769 + }, + { + "epoch": 2.8290888904309206, + "grad_norm": 1.0080255270004272, + "learning_rate": 4.167604289740359e-05, + "loss": 0.2859, + "step": 40770 + }, + { + "epoch": 2.8291582818680174, + "grad_norm": 1.2656522989273071, + "learning_rate": 4.167147898703936e-05, + "loss": 0.7949, + "step": 40771 + }, + { + "epoch": 2.8292276733051143, + "grad_norm": 0.9909164905548096, + "learning_rate": 4.16669152608115e-05, + "loss": 0.3483, + "step": 40772 + }, + { + "epoch": 2.8292970647422107, + "grad_norm": 1.1388516426086426, + "learning_rate": 4.166235171873443e-05, + "loss": 0.4564, + "step": 40773 + }, + { + "epoch": 2.8293664561793075, + "grad_norm": 1.2693028450012207, + "learning_rate": 4.165778836082258e-05, + "loss": 0.4148, + "step": 40774 + }, + { + "epoch": 2.829435847616404, + "grad_norm": 1.1779098510742188, + "learning_rate": 4.1653225187090284e-05, + "loss": 0.3504, + "step": 40775 + }, + { + "epoch": 2.8295052390535007, + "grad_norm": 1.0239031314849854, + "learning_rate": 4.164866219755199e-05, + "loss": 0.4178, + "step": 40776 + }, + { + "epoch": 2.8295746304905975, + "grad_norm": 0.973050057888031, + "learning_rate": 4.16440993922221e-05, + "loss": 0.579, + "step": 40777 + }, + { + "epoch": 2.8296440219276944, + "grad_norm": 1.0739636421203613, + "learning_rate": 4.1639536771115016e-05, + "loss": 0.4129, + "step": 40778 + }, + { + "epoch": 2.8297134133647908, + "grad_norm": 1.2263718843460083, + "learning_rate": 4.163497433424515e-05, + "loss": 0.4, + "step": 40779 + }, + { + "epoch": 2.8297828048018876, + "grad_norm": 1.7019329071044922, + "learning_rate": 4.1630412081626925e-05, + "loss": 0.2617, + "step": 40780 + }, + { + "epoch": 2.829852196238984, + "grad_norm": 1.2847391366958618, + "learning_rate": 4.162585001327468e-05, + "loss": 0.4615, + "step": 40781 + }, + { + "epoch": 2.829921587676081, + "grad_norm": 1.4785242080688477, + "learning_rate": 4.162128812920285e-05, + "loss": 0.4378, + "step": 40782 + }, + { + "epoch": 2.8299909791131777, + "grad_norm": 1.4921233654022217, + "learning_rate": 4.161672642942586e-05, + "loss": 0.3352, + "step": 40783 + }, + { + "epoch": 2.830060370550274, + "grad_norm": 0.8581548929214478, + "learning_rate": 4.161216491395811e-05, + "loss": 0.3326, + "step": 40784 + }, + { + "epoch": 2.830129761987371, + "grad_norm": 1.125669240951538, + "learning_rate": 4.160760358281395e-05, + "loss": 0.3827, + "step": 40785 + }, + { + "epoch": 2.8301991534244673, + "grad_norm": 0.9976955652236938, + "learning_rate": 4.1603042436007814e-05, + "loss": 0.3461, + "step": 40786 + }, + { + "epoch": 2.830268544861564, + "grad_norm": 1.8418278694152832, + "learning_rate": 4.1598481473554084e-05, + "loss": 0.3975, + "step": 40787 + }, + { + "epoch": 2.830337936298661, + "grad_norm": 1.2263675928115845, + "learning_rate": 4.1593920695467226e-05, + "loss": 0.3314, + "step": 40788 + }, + { + "epoch": 2.8304073277357573, + "grad_norm": 1.293515682220459, + "learning_rate": 4.158936010176153e-05, + "loss": 0.41, + "step": 40789 + }, + { + "epoch": 2.830476719172854, + "grad_norm": 1.2815194129943848, + "learning_rate": 4.158479969245146e-05, + "loss": 0.4776, + "step": 40790 + }, + { + "epoch": 2.8305461106099505, + "grad_norm": 1.06027352809906, + "learning_rate": 4.15802394675514e-05, + "loss": 0.5125, + "step": 40791 + }, + { + "epoch": 2.8306155020470474, + "grad_norm": 1.06461501121521, + "learning_rate": 4.1575679427075745e-05, + "loss": 0.345, + "step": 40792 + }, + { + "epoch": 2.830684893484144, + "grad_norm": 1.3348404169082642, + "learning_rate": 4.157111957103889e-05, + "loss": 0.372, + "step": 40793 + }, + { + "epoch": 2.8307542849212406, + "grad_norm": 0.9247518181800842, + "learning_rate": 4.1566559899455224e-05, + "loss": 0.2338, + "step": 40794 + }, + { + "epoch": 2.8308236763583374, + "grad_norm": 0.8908724784851074, + "learning_rate": 4.1562000412339196e-05, + "loss": 0.2986, + "step": 40795 + }, + { + "epoch": 2.830893067795434, + "grad_norm": 0.835712730884552, + "learning_rate": 4.155744110970512e-05, + "loss": 0.2823, + "step": 40796 + }, + { + "epoch": 2.8309624592325306, + "grad_norm": 1.0335737466812134, + "learning_rate": 4.155288199156742e-05, + "loss": 0.26, + "step": 40797 + }, + { + "epoch": 2.8310318506696275, + "grad_norm": 1.010640025138855, + "learning_rate": 4.1548323057940475e-05, + "loss": 0.3173, + "step": 40798 + }, + { + "epoch": 2.831101242106724, + "grad_norm": 0.8271385431289673, + "learning_rate": 4.1543764308838704e-05, + "loss": 0.4453, + "step": 40799 + }, + { + "epoch": 2.8311706335438207, + "grad_norm": 1.393509864807129, + "learning_rate": 4.153920574427649e-05, + "loss": 0.4926, + "step": 40800 + }, + { + "epoch": 2.831240024980917, + "grad_norm": 0.9605583548545837, + "learning_rate": 4.153464736426821e-05, + "loss": 0.3718, + "step": 40801 + }, + { + "epoch": 2.831309416418014, + "grad_norm": 0.8488134145736694, + "learning_rate": 4.153008916882831e-05, + "loss": 0.2904, + "step": 40802 + }, + { + "epoch": 2.8313788078551108, + "grad_norm": 1.0222320556640625, + "learning_rate": 4.15255311579711e-05, + "loss": 0.3689, + "step": 40803 + }, + { + "epoch": 2.8314481992922076, + "grad_norm": 1.0120744705200195, + "learning_rate": 4.1520973331711e-05, + "loss": 0.4102, + "step": 40804 + }, + { + "epoch": 2.831517590729304, + "grad_norm": 1.3651400804519653, + "learning_rate": 4.15164156900624e-05, + "loss": 0.4843, + "step": 40805 + }, + { + "epoch": 2.831586982166401, + "grad_norm": 1.4947423934936523, + "learning_rate": 4.151185823303969e-05, + "loss": 0.4451, + "step": 40806 + }, + { + "epoch": 2.831656373603497, + "grad_norm": 1.0763442516326904, + "learning_rate": 4.1507300960657255e-05, + "loss": 0.3797, + "step": 40807 + }, + { + "epoch": 2.831725765040594, + "grad_norm": 1.065279483795166, + "learning_rate": 4.15027438729295e-05, + "loss": 0.3219, + "step": 40808 + }, + { + "epoch": 2.831795156477691, + "grad_norm": 0.6866188645362854, + "learning_rate": 4.149818696987082e-05, + "loss": 0.2393, + "step": 40809 + }, + { + "epoch": 2.8318645479147873, + "grad_norm": 1.067297339439392, + "learning_rate": 4.149363025149554e-05, + "loss": 0.2659, + "step": 40810 + }, + { + "epoch": 2.831933939351884, + "grad_norm": 1.5461342334747314, + "learning_rate": 4.1489073717818086e-05, + "loss": 0.3784, + "step": 40811 + }, + { + "epoch": 2.8320033307889805, + "grad_norm": 0.9041587710380554, + "learning_rate": 4.1484517368852835e-05, + "loss": 0.3797, + "step": 40812 + }, + { + "epoch": 2.8320727222260773, + "grad_norm": 1.2319906949996948, + "learning_rate": 4.1479961204614174e-05, + "loss": 0.2783, + "step": 40813 + }, + { + "epoch": 2.832142113663174, + "grad_norm": 2.4863393306732178, + "learning_rate": 4.147540522511648e-05, + "loss": 0.4634, + "step": 40814 + }, + { + "epoch": 2.8322115051002705, + "grad_norm": 0.9795360565185547, + "learning_rate": 4.147084943037418e-05, + "loss": 0.283, + "step": 40815 + }, + { + "epoch": 2.8322808965373674, + "grad_norm": 1.2374120950698853, + "learning_rate": 4.146629382040159e-05, + "loss": 0.4104, + "step": 40816 + }, + { + "epoch": 2.8323502879744638, + "grad_norm": 1.1515480279922485, + "learning_rate": 4.146173839521311e-05, + "loss": 0.3757, + "step": 40817 + }, + { + "epoch": 2.8324196794115606, + "grad_norm": 1.1002812385559082, + "learning_rate": 4.1457183154823176e-05, + "loss": 0.3758, + "step": 40818 + }, + { + "epoch": 2.8324890708486574, + "grad_norm": 1.787975788116455, + "learning_rate": 4.1452628099246085e-05, + "loss": 0.3719, + "step": 40819 + }, + { + "epoch": 2.832558462285754, + "grad_norm": 0.8328431248664856, + "learning_rate": 4.1448073228496245e-05, + "loss": 0.3136, + "step": 40820 + }, + { + "epoch": 2.8326278537228506, + "grad_norm": 0.9061217308044434, + "learning_rate": 4.144351854258804e-05, + "loss": 0.3477, + "step": 40821 + }, + { + "epoch": 2.832697245159947, + "grad_norm": 0.915575385093689, + "learning_rate": 4.1438964041535876e-05, + "loss": 0.4437, + "step": 40822 + }, + { + "epoch": 2.832766636597044, + "grad_norm": 0.6884983777999878, + "learning_rate": 4.1434409725354126e-05, + "loss": 0.2198, + "step": 40823 + }, + { + "epoch": 2.8328360280341407, + "grad_norm": 1.5521926879882812, + "learning_rate": 4.142985559405712e-05, + "loss": 0.4423, + "step": 40824 + }, + { + "epoch": 2.832905419471237, + "grad_norm": 0.9281569123268127, + "learning_rate": 4.142530164765927e-05, + "loss": 0.354, + "step": 40825 + }, + { + "epoch": 2.832974810908334, + "grad_norm": 1.3070646524429321, + "learning_rate": 4.142074788617494e-05, + "loss": 0.5699, + "step": 40826 + }, + { + "epoch": 2.8330442023454303, + "grad_norm": 1.1556651592254639, + "learning_rate": 4.14161943096185e-05, + "loss": 0.3019, + "step": 40827 + }, + { + "epoch": 2.833113593782527, + "grad_norm": 1.4305005073547363, + "learning_rate": 4.141164091800436e-05, + "loss": 0.4968, + "step": 40828 + }, + { + "epoch": 2.833182985219624, + "grad_norm": 1.181506872177124, + "learning_rate": 4.140708771134686e-05, + "loss": 0.3612, + "step": 40829 + }, + { + "epoch": 2.833252376656721, + "grad_norm": 1.3292992115020752, + "learning_rate": 4.140253468966042e-05, + "loss": 0.327, + "step": 40830 + }, + { + "epoch": 2.833321768093817, + "grad_norm": 1.000226378440857, + "learning_rate": 4.1397981852959346e-05, + "loss": 0.2429, + "step": 40831 + }, + { + "epoch": 2.833391159530914, + "grad_norm": 1.4710758924484253, + "learning_rate": 4.139342920125805e-05, + "loss": 0.3834, + "step": 40832 + }, + { + "epoch": 2.8334605509680104, + "grad_norm": 2.1206724643707275, + "learning_rate": 4.138887673457089e-05, + "loss": 0.3302, + "step": 40833 + }, + { + "epoch": 2.8335299424051072, + "grad_norm": 1.1869993209838867, + "learning_rate": 4.1384324452912246e-05, + "loss": 0.37, + "step": 40834 + }, + { + "epoch": 2.833599333842204, + "grad_norm": 1.1635708808898926, + "learning_rate": 4.13797723562965e-05, + "loss": 0.2989, + "step": 40835 + }, + { + "epoch": 2.8336687252793005, + "grad_norm": 1.1151156425476074, + "learning_rate": 4.137522044473804e-05, + "loss": 0.372, + "step": 40836 + }, + { + "epoch": 2.8337381167163973, + "grad_norm": 1.0218827724456787, + "learning_rate": 4.137066871825117e-05, + "loss": 0.4304, + "step": 40837 + }, + { + "epoch": 2.8338075081534937, + "grad_norm": 1.2126951217651367, + "learning_rate": 4.136611717685029e-05, + "loss": 0.3736, + "step": 40838 + }, + { + "epoch": 2.8338768995905905, + "grad_norm": 0.9243044257164001, + "learning_rate": 4.1361565820549796e-05, + "loss": 0.3551, + "step": 40839 + }, + { + "epoch": 2.8339462910276874, + "grad_norm": 0.8395277261734009, + "learning_rate": 4.135701464936402e-05, + "loss": 0.2345, + "step": 40840 + }, + { + "epoch": 2.8340156824647837, + "grad_norm": 1.007630467414856, + "learning_rate": 4.1352463663307354e-05, + "loss": 0.3365, + "step": 40841 + }, + { + "epoch": 2.8340850739018806, + "grad_norm": 1.223460078239441, + "learning_rate": 4.1347912862394144e-05, + "loss": 0.3519, + "step": 40842 + }, + { + "epoch": 2.834154465338977, + "grad_norm": 1.541452169418335, + "learning_rate": 4.134336224663881e-05, + "loss": 0.4421, + "step": 40843 + }, + { + "epoch": 2.834223856776074, + "grad_norm": 0.9854142665863037, + "learning_rate": 4.1338811816055646e-05, + "loss": 0.3067, + "step": 40844 + }, + { + "epoch": 2.8342932482131706, + "grad_norm": 1.196418046951294, + "learning_rate": 4.133426157065904e-05, + "loss": 0.4603, + "step": 40845 + }, + { + "epoch": 2.834362639650267, + "grad_norm": 1.2733782529830933, + "learning_rate": 4.1329711510463364e-05, + "loss": 0.3502, + "step": 40846 + }, + { + "epoch": 2.834432031087364, + "grad_norm": 1.1379510164260864, + "learning_rate": 4.132516163548298e-05, + "loss": 0.3806, + "step": 40847 + }, + { + "epoch": 2.8345014225244602, + "grad_norm": 1.8566564321517944, + "learning_rate": 4.132061194573226e-05, + "loss": 0.3356, + "step": 40848 + }, + { + "epoch": 2.834570813961557, + "grad_norm": 1.1915881633758545, + "learning_rate": 4.131606244122556e-05, + "loss": 0.3683, + "step": 40849 + }, + { + "epoch": 2.834640205398654, + "grad_norm": 0.9663358330726624, + "learning_rate": 4.1311513121977266e-05, + "loss": 0.3663, + "step": 40850 + }, + { + "epoch": 2.8347095968357507, + "grad_norm": 1.4330635070800781, + "learning_rate": 4.130696398800169e-05, + "loss": 0.3579, + "step": 40851 + }, + { + "epoch": 2.834778988272847, + "grad_norm": 1.4199626445770264, + "learning_rate": 4.130241503931321e-05, + "loss": 0.2932, + "step": 40852 + }, + { + "epoch": 2.8348483797099435, + "grad_norm": 1.1262893676757812, + "learning_rate": 4.1297866275926244e-05, + "loss": 0.41, + "step": 40853 + }, + { + "epoch": 2.8349177711470404, + "grad_norm": 1.6331557035446167, + "learning_rate": 4.129331769785506e-05, + "loss": 0.5467, + "step": 40854 + }, + { + "epoch": 2.834987162584137, + "grad_norm": 1.2168306112289429, + "learning_rate": 4.128876930511406e-05, + "loss": 0.6072, + "step": 40855 + }, + { + "epoch": 2.835056554021234, + "grad_norm": 1.6566275358200073, + "learning_rate": 4.1284221097717615e-05, + "loss": 0.5931, + "step": 40856 + }, + { + "epoch": 2.8351259454583304, + "grad_norm": 0.8437013030052185, + "learning_rate": 4.127967307568009e-05, + "loss": 0.2837, + "step": 40857 + }, + { + "epoch": 2.8351953368954272, + "grad_norm": 1.3703947067260742, + "learning_rate": 4.12751252390158e-05, + "loss": 0.2987, + "step": 40858 + }, + { + "epoch": 2.8352647283325236, + "grad_norm": 1.0453240871429443, + "learning_rate": 4.127057758773912e-05, + "loss": 0.4091, + "step": 40859 + }, + { + "epoch": 2.8353341197696205, + "grad_norm": 1.0805901288986206, + "learning_rate": 4.1266030121864415e-05, + "loss": 0.3555, + "step": 40860 + }, + { + "epoch": 2.8354035112067173, + "grad_norm": 0.852079451084137, + "learning_rate": 4.126148284140605e-05, + "loss": 0.3125, + "step": 40861 + }, + { + "epoch": 2.8354729026438137, + "grad_norm": 1.6697546243667603, + "learning_rate": 4.125693574637836e-05, + "loss": 0.4271, + "step": 40862 + }, + { + "epoch": 2.8355422940809105, + "grad_norm": 1.3473225831985474, + "learning_rate": 4.12523888367957e-05, + "loss": 0.5363, + "step": 40863 + }, + { + "epoch": 2.835611685518007, + "grad_norm": 0.9590115547180176, + "learning_rate": 4.1247842112672475e-05, + "loss": 0.3363, + "step": 40864 + }, + { + "epoch": 2.8356810769551037, + "grad_norm": 1.3183009624481201, + "learning_rate": 4.1243295574022964e-05, + "loss": 0.4071, + "step": 40865 + }, + { + "epoch": 2.8357504683922006, + "grad_norm": 2.2422733306884766, + "learning_rate": 4.123874922086155e-05, + "loss": 0.3673, + "step": 40866 + }, + { + "epoch": 2.835819859829297, + "grad_norm": 0.9704864621162415, + "learning_rate": 4.123420305320258e-05, + "loss": 0.2803, + "step": 40867 + }, + { + "epoch": 2.835889251266394, + "grad_norm": 0.8778725862503052, + "learning_rate": 4.1229657071060425e-05, + "loss": 0.2753, + "step": 40868 + }, + { + "epoch": 2.83595864270349, + "grad_norm": 0.8334062695503235, + "learning_rate": 4.122511127444941e-05, + "loss": 0.2911, + "step": 40869 + }, + { + "epoch": 2.836028034140587, + "grad_norm": 1.0093743801116943, + "learning_rate": 4.122056566338391e-05, + "loss": 0.3795, + "step": 40870 + }, + { + "epoch": 2.836097425577684, + "grad_norm": 1.013069748878479, + "learning_rate": 4.1216020237878295e-05, + "loss": 0.4625, + "step": 40871 + }, + { + "epoch": 2.8361668170147802, + "grad_norm": 1.3243869543075562, + "learning_rate": 4.1211474997946855e-05, + "loss": 0.3477, + "step": 40872 + }, + { + "epoch": 2.836236208451877, + "grad_norm": 1.5910372734069824, + "learning_rate": 4.120692994360396e-05, + "loss": 0.4321, + "step": 40873 + }, + { + "epoch": 2.8363055998889735, + "grad_norm": 1.072927713394165, + "learning_rate": 4.1202385074863966e-05, + "loss": 0.3608, + "step": 40874 + }, + { + "epoch": 2.8363749913260703, + "grad_norm": 1.399455189704895, + "learning_rate": 4.119784039174123e-05, + "loss": 0.4169, + "step": 40875 + }, + { + "epoch": 2.836444382763167, + "grad_norm": 1.563733696937561, + "learning_rate": 4.1193295894250084e-05, + "loss": 0.4376, + "step": 40876 + }, + { + "epoch": 2.836513774200264, + "grad_norm": 1.1555944681167603, + "learning_rate": 4.118875158240487e-05, + "loss": 0.4676, + "step": 40877 + }, + { + "epoch": 2.8365831656373603, + "grad_norm": 1.0804498195648193, + "learning_rate": 4.118420745621999e-05, + "loss": 0.325, + "step": 40878 + }, + { + "epoch": 2.8366525570744567, + "grad_norm": 1.8392045497894287, + "learning_rate": 4.117966351570971e-05, + "loss": 0.3818, + "step": 40879 + }, + { + "epoch": 2.8367219485115536, + "grad_norm": 1.1515612602233887, + "learning_rate": 4.1175119760888406e-05, + "loss": 0.2989, + "step": 40880 + }, + { + "epoch": 2.8367913399486504, + "grad_norm": 1.0650094747543335, + "learning_rate": 4.117057619177042e-05, + "loss": 0.3433, + "step": 40881 + }, + { + "epoch": 2.8368607313857472, + "grad_norm": 1.06454336643219, + "learning_rate": 4.11660328083701e-05, + "loss": 0.3454, + "step": 40882 + }, + { + "epoch": 2.8369301228228436, + "grad_norm": 1.6052136421203613, + "learning_rate": 4.116148961070179e-05, + "loss": 0.5559, + "step": 40883 + }, + { + "epoch": 2.8369995142599405, + "grad_norm": 1.483129858970642, + "learning_rate": 4.115694659877987e-05, + "loss": 0.3537, + "step": 40884 + }, + { + "epoch": 2.837068905697037, + "grad_norm": 1.7305275201797485, + "learning_rate": 4.11524037726186e-05, + "loss": 0.3816, + "step": 40885 + }, + { + "epoch": 2.8371382971341337, + "grad_norm": 1.6433582305908203, + "learning_rate": 4.1147861132232366e-05, + "loss": 0.5196, + "step": 40886 + }, + { + "epoch": 2.8372076885712305, + "grad_norm": 1.010682463645935, + "learning_rate": 4.114331867763551e-05, + "loss": 0.3776, + "step": 40887 + }, + { + "epoch": 2.837277080008327, + "grad_norm": 1.0941420793533325, + "learning_rate": 4.1138776408842394e-05, + "loss": 0.4115, + "step": 40888 + }, + { + "epoch": 2.8373464714454237, + "grad_norm": 0.9588284492492676, + "learning_rate": 4.1134234325867307e-05, + "loss": 0.301, + "step": 40889 + }, + { + "epoch": 2.83741586288252, + "grad_norm": 1.8380656242370605, + "learning_rate": 4.1129692428724607e-05, + "loss": 0.5043, + "step": 40890 + }, + { + "epoch": 2.837485254319617, + "grad_norm": 1.010929822921753, + "learning_rate": 4.112515071742864e-05, + "loss": 0.322, + "step": 40891 + }, + { + "epoch": 2.837554645756714, + "grad_norm": 1.0037118196487427, + "learning_rate": 4.112060919199378e-05, + "loss": 0.3465, + "step": 40892 + }, + { + "epoch": 2.83762403719381, + "grad_norm": 1.2978146076202393, + "learning_rate": 4.111606785243428e-05, + "loss": 0.2123, + "step": 40893 + }, + { + "epoch": 2.837693428630907, + "grad_norm": 0.7990165948867798, + "learning_rate": 4.111152669876454e-05, + "loss": 0.353, + "step": 40894 + }, + { + "epoch": 2.8377628200680034, + "grad_norm": 1.4794481992721558, + "learning_rate": 4.1106985730998856e-05, + "loss": 0.5016, + "step": 40895 + }, + { + "epoch": 2.8378322115051002, + "grad_norm": 0.9092438817024231, + "learning_rate": 4.11024449491516e-05, + "loss": 0.2879, + "step": 40896 + }, + { + "epoch": 2.837901602942197, + "grad_norm": 1.5666594505310059, + "learning_rate": 4.109790435323708e-05, + "loss": 0.4203, + "step": 40897 + }, + { + "epoch": 2.8379709943792935, + "grad_norm": 0.9393077492713928, + "learning_rate": 4.109336394326966e-05, + "loss": 0.2925, + "step": 40898 + }, + { + "epoch": 2.8380403858163903, + "grad_norm": 1.225937008857727, + "learning_rate": 4.108882371926368e-05, + "loss": 0.4791, + "step": 40899 + }, + { + "epoch": 2.8381097772534867, + "grad_norm": 0.8268942832946777, + "learning_rate": 4.108428368123341e-05, + "loss": 0.2345, + "step": 40900 + }, + { + "epoch": 2.8381791686905835, + "grad_norm": 0.9601516127586365, + "learning_rate": 4.107974382919323e-05, + "loss": 0.3747, + "step": 40901 + }, + { + "epoch": 2.8382485601276803, + "grad_norm": 1.4079514741897583, + "learning_rate": 4.107520416315745e-05, + "loss": 0.3932, + "step": 40902 + }, + { + "epoch": 2.838317951564777, + "grad_norm": 1.3362889289855957, + "learning_rate": 4.107066468314043e-05, + "loss": 0.401, + "step": 40903 + }, + { + "epoch": 2.8383873430018736, + "grad_norm": 1.0027514696121216, + "learning_rate": 4.106612538915647e-05, + "loss": 0.2974, + "step": 40904 + }, + { + "epoch": 2.8384567344389704, + "grad_norm": 1.1977944374084473, + "learning_rate": 4.106158628121992e-05, + "loss": 0.3506, + "step": 40905 + }, + { + "epoch": 2.838526125876067, + "grad_norm": 1.2515684366226196, + "learning_rate": 4.1057047359345134e-05, + "loss": 0.2447, + "step": 40906 + }, + { + "epoch": 2.8385955173131636, + "grad_norm": 1.1757677793502808, + "learning_rate": 4.105250862354638e-05, + "loss": 0.5049, + "step": 40907 + }, + { + "epoch": 2.8386649087502605, + "grad_norm": 2.7020444869995117, + "learning_rate": 4.1047970073838026e-05, + "loss": 0.4485, + "step": 40908 + }, + { + "epoch": 2.838734300187357, + "grad_norm": 1.877471923828125, + "learning_rate": 4.104343171023438e-05, + "loss": 0.6373, + "step": 40909 + }, + { + "epoch": 2.8388036916244537, + "grad_norm": 1.6239399909973145, + "learning_rate": 4.103889353274979e-05, + "loss": 0.3324, + "step": 40910 + }, + { + "epoch": 2.83887308306155, + "grad_norm": 1.3596665859222412, + "learning_rate": 4.103435554139856e-05, + "loss": 0.3821, + "step": 40911 + }, + { + "epoch": 2.838942474498647, + "grad_norm": 0.9093688130378723, + "learning_rate": 4.1029817736195065e-05, + "loss": 0.2325, + "step": 40912 + }, + { + "epoch": 2.8390118659357437, + "grad_norm": 1.881116271018982, + "learning_rate": 4.102528011715356e-05, + "loss": 0.3304, + "step": 40913 + }, + { + "epoch": 2.83908125737284, + "grad_norm": 1.2063665390014648, + "learning_rate": 4.1020742684288414e-05, + "loss": 0.4863, + "step": 40914 + }, + { + "epoch": 2.839150648809937, + "grad_norm": 1.3218454122543335, + "learning_rate": 4.101620543761393e-05, + "loss": 0.3021, + "step": 40915 + }, + { + "epoch": 2.8392200402470333, + "grad_norm": 0.9042892456054688, + "learning_rate": 4.101166837714444e-05, + "loss": 0.3579, + "step": 40916 + }, + { + "epoch": 2.83928943168413, + "grad_norm": 0.9795360565185547, + "learning_rate": 4.100713150289428e-05, + "loss": 0.2943, + "step": 40917 + }, + { + "epoch": 2.839358823121227, + "grad_norm": 1.2437477111816406, + "learning_rate": 4.100259481487775e-05, + "loss": 0.409, + "step": 40918 + }, + { + "epoch": 2.8394282145583234, + "grad_norm": 0.8795381188392639, + "learning_rate": 4.099805831310922e-05, + "loss": 0.2126, + "step": 40919 + }, + { + "epoch": 2.8394976059954202, + "grad_norm": 1.4031050205230713, + "learning_rate": 4.0993521997602944e-05, + "loss": 0.6263, + "step": 40920 + }, + { + "epoch": 2.8395669974325166, + "grad_norm": 1.4088071584701538, + "learning_rate": 4.0988985868373274e-05, + "loss": 0.3785, + "step": 40921 + }, + { + "epoch": 2.8396363888696134, + "grad_norm": 0.7209720015525818, + "learning_rate": 4.098444992543452e-05, + "loss": 0.3137, + "step": 40922 + }, + { + "epoch": 2.8397057803067103, + "grad_norm": 1.1712409257888794, + "learning_rate": 4.097991416880106e-05, + "loss": 0.3437, + "step": 40923 + }, + { + "epoch": 2.8397751717438067, + "grad_norm": 0.9008560180664062, + "learning_rate": 4.0975378598487115e-05, + "loss": 0.2931, + "step": 40924 + }, + { + "epoch": 2.8398445631809035, + "grad_norm": 1.4088304042816162, + "learning_rate": 4.097084321450706e-05, + "loss": 0.3291, + "step": 40925 + }, + { + "epoch": 2.839913954618, + "grad_norm": 1.0438592433929443, + "learning_rate": 4.0966308016875196e-05, + "loss": 0.2274, + "step": 40926 + }, + { + "epoch": 2.8399833460550967, + "grad_norm": 1.6634505987167358, + "learning_rate": 4.09617730056059e-05, + "loss": 0.381, + "step": 40927 + }, + { + "epoch": 2.8400527374921936, + "grad_norm": 1.4280074834823608, + "learning_rate": 4.095723818071339e-05, + "loss": 0.5076, + "step": 40928 + }, + { + "epoch": 2.8401221289292904, + "grad_norm": 1.0278282165527344, + "learning_rate": 4.095270354221202e-05, + "loss": 0.3532, + "step": 40929 + }, + { + "epoch": 2.840191520366387, + "grad_norm": 1.2751888036727905, + "learning_rate": 4.094816909011613e-05, + "loss": 0.4318, + "step": 40930 + }, + { + "epoch": 2.8402609118034836, + "grad_norm": 1.1905876398086548, + "learning_rate": 4.094363482444e-05, + "loss": 0.4118, + "step": 40931 + }, + { + "epoch": 2.84033030324058, + "grad_norm": 1.0952125787734985, + "learning_rate": 4.093910074519798e-05, + "loss": 0.4155, + "step": 40932 + }, + { + "epoch": 2.840399694677677, + "grad_norm": 1.2707568407058716, + "learning_rate": 4.0934566852404396e-05, + "loss": 0.3057, + "step": 40933 + }, + { + "epoch": 2.8404690861147737, + "grad_norm": 1.4859368801116943, + "learning_rate": 4.09300331460735e-05, + "loss": 0.4055, + "step": 40934 + }, + { + "epoch": 2.84053847755187, + "grad_norm": 0.8594744801521301, + "learning_rate": 4.0925499626219634e-05, + "loss": 0.351, + "step": 40935 + }, + { + "epoch": 2.840607868988967, + "grad_norm": 1.3540595769882202, + "learning_rate": 4.092096629285711e-05, + "loss": 0.4016, + "step": 40936 + }, + { + "epoch": 2.8406772604260633, + "grad_norm": 1.4440560340881348, + "learning_rate": 4.091643314600024e-05, + "loss": 0.4255, + "step": 40937 + }, + { + "epoch": 2.84074665186316, + "grad_norm": 1.1486189365386963, + "learning_rate": 4.0911900185663344e-05, + "loss": 0.2865, + "step": 40938 + }, + { + "epoch": 2.840816043300257, + "grad_norm": 2.00341534614563, + "learning_rate": 4.090736741186071e-05, + "loss": 0.3251, + "step": 40939 + }, + { + "epoch": 2.8408854347373533, + "grad_norm": 1.004533052444458, + "learning_rate": 4.0902834824606704e-05, + "loss": 0.3586, + "step": 40940 + }, + { + "epoch": 2.84095482617445, + "grad_norm": 1.3093457221984863, + "learning_rate": 4.0898302423915555e-05, + "loss": 0.434, + "step": 40941 + }, + { + "epoch": 2.8410242176115466, + "grad_norm": 1.3181861639022827, + "learning_rate": 4.089377020980162e-05, + "loss": 0.3937, + "step": 40942 + }, + { + "epoch": 2.8410936090486434, + "grad_norm": 1.814435362815857, + "learning_rate": 4.088923818227918e-05, + "loss": 0.5246, + "step": 40943 + }, + { + "epoch": 2.84116300048574, + "grad_norm": 1.4517498016357422, + "learning_rate": 4.088470634136255e-05, + "loss": 0.5064, + "step": 40944 + }, + { + "epoch": 2.8412323919228366, + "grad_norm": 1.1939865350723267, + "learning_rate": 4.088017468706606e-05, + "loss": 0.3868, + "step": 40945 + }, + { + "epoch": 2.8413017833599334, + "grad_norm": 1.2236413955688477, + "learning_rate": 4.0875643219404e-05, + "loss": 0.3158, + "step": 40946 + }, + { + "epoch": 2.84137117479703, + "grad_norm": 1.1297235488891602, + "learning_rate": 4.087111193839071e-05, + "loss": 0.3196, + "step": 40947 + }, + { + "epoch": 2.8414405662341267, + "grad_norm": 1.6801856756210327, + "learning_rate": 4.086658084404041e-05, + "loss": 0.5265, + "step": 40948 + }, + { + "epoch": 2.8415099576712235, + "grad_norm": 1.9972556829452515, + "learning_rate": 4.086204993636746e-05, + "loss": 0.4452, + "step": 40949 + }, + { + "epoch": 2.84157934910832, + "grad_norm": 1.002291202545166, + "learning_rate": 4.0857519215386165e-05, + "loss": 0.3774, + "step": 40950 + }, + { + "epoch": 2.8416487405454167, + "grad_norm": 1.147166132926941, + "learning_rate": 4.085298868111082e-05, + "loss": 0.5169, + "step": 40951 + }, + { + "epoch": 2.841718131982513, + "grad_norm": 1.3277499675750732, + "learning_rate": 4.0848458333555705e-05, + "loss": 0.4124, + "step": 40952 + }, + { + "epoch": 2.84178752341961, + "grad_norm": 1.2963106632232666, + "learning_rate": 4.084392817273517e-05, + "loss": 0.2938, + "step": 40953 + }, + { + "epoch": 2.8418569148567068, + "grad_norm": 1.3192546367645264, + "learning_rate": 4.0839398198663516e-05, + "loss": 0.2307, + "step": 40954 + }, + { + "epoch": 2.8419263062938036, + "grad_norm": 0.804314374923706, + "learning_rate": 4.0834868411354975e-05, + "loss": 0.364, + "step": 40955 + }, + { + "epoch": 2.8419956977309, + "grad_norm": 1.00830078125, + "learning_rate": 4.08303388108239e-05, + "loss": 0.5014, + "step": 40956 + }, + { + "epoch": 2.842065089167997, + "grad_norm": 1.0410226583480835, + "learning_rate": 4.082580939708458e-05, + "loss": 0.4242, + "step": 40957 + }, + { + "epoch": 2.842134480605093, + "grad_norm": 1.0042521953582764, + "learning_rate": 4.082128017015134e-05, + "loss": 0.3085, + "step": 40958 + }, + { + "epoch": 2.84220387204219, + "grad_norm": 0.9375962018966675, + "learning_rate": 4.0816751130038434e-05, + "loss": 0.3225, + "step": 40959 + }, + { + "epoch": 2.842273263479287, + "grad_norm": 0.9979764819145203, + "learning_rate": 4.081222227676017e-05, + "loss": 0.354, + "step": 40960 + }, + { + "epoch": 2.8423426549163833, + "grad_norm": 1.341341257095337, + "learning_rate": 4.080769361033089e-05, + "loss": 0.3032, + "step": 40961 + }, + { + "epoch": 2.84241204635348, + "grad_norm": 1.6783063411712646, + "learning_rate": 4.080316513076481e-05, + "loss": 0.5068, + "step": 40962 + }, + { + "epoch": 2.8424814377905765, + "grad_norm": 1.4064034223556519, + "learning_rate": 4.0798636838076276e-05, + "loss": 0.276, + "step": 40963 + }, + { + "epoch": 2.8425508292276733, + "grad_norm": 1.51470148563385, + "learning_rate": 4.079410873227958e-05, + "loss": 0.3366, + "step": 40964 + }, + { + "epoch": 2.84262022066477, + "grad_norm": 1.1217526197433472, + "learning_rate": 4.078958081338901e-05, + "loss": 0.4639, + "step": 40965 + }, + { + "epoch": 2.8426896121018665, + "grad_norm": 3.3986222743988037, + "learning_rate": 4.078505308141887e-05, + "loss": 0.4444, + "step": 40966 + }, + { + "epoch": 2.8427590035389634, + "grad_norm": 0.8245589137077332, + "learning_rate": 4.0780525536383426e-05, + "loss": 0.3446, + "step": 40967 + }, + { + "epoch": 2.8428283949760598, + "grad_norm": 1.2959784269332886, + "learning_rate": 4.0775998178297046e-05, + "loss": 0.3186, + "step": 40968 + }, + { + "epoch": 2.8428977864131566, + "grad_norm": 1.5859694480895996, + "learning_rate": 4.077147100717394e-05, + "loss": 0.3703, + "step": 40969 + }, + { + "epoch": 2.8429671778502534, + "grad_norm": 0.721923828125, + "learning_rate": 4.076694402302841e-05, + "loss": 0.278, + "step": 40970 + }, + { + "epoch": 2.84303656928735, + "grad_norm": 1.2010128498077393, + "learning_rate": 4.0762417225874774e-05, + "loss": 0.3464, + "step": 40971 + }, + { + "epoch": 2.8431059607244467, + "grad_norm": 1.3869258165359497, + "learning_rate": 4.0757890615727315e-05, + "loss": 0.4456, + "step": 40972 + }, + { + "epoch": 2.843175352161543, + "grad_norm": 1.2154645919799805, + "learning_rate": 4.075336419260032e-05, + "loss": 0.4152, + "step": 40973 + }, + { + "epoch": 2.84324474359864, + "grad_norm": 1.2927024364471436, + "learning_rate": 4.074883795650808e-05, + "loss": 0.4966, + "step": 40974 + }, + { + "epoch": 2.8433141350357367, + "grad_norm": 1.06342613697052, + "learning_rate": 4.074431190746492e-05, + "loss": 0.2479, + "step": 40975 + }, + { + "epoch": 2.843383526472833, + "grad_norm": 0.9125651121139526, + "learning_rate": 4.073978604548506e-05, + "loss": 0.3901, + "step": 40976 + }, + { + "epoch": 2.84345291790993, + "grad_norm": 0.8747063279151917, + "learning_rate": 4.0735260370582817e-05, + "loss": 0.2883, + "step": 40977 + }, + { + "epoch": 2.8435223093470263, + "grad_norm": 1.0826820135116577, + "learning_rate": 4.073073488277248e-05, + "loss": 0.5124, + "step": 40978 + }, + { + "epoch": 2.843591700784123, + "grad_norm": 1.3664005994796753, + "learning_rate": 4.072620958206833e-05, + "loss": 0.3482, + "step": 40979 + }, + { + "epoch": 2.84366109222122, + "grad_norm": 1.04667067527771, + "learning_rate": 4.0721684468484677e-05, + "loss": 0.3542, + "step": 40980 + }, + { + "epoch": 2.843730483658317, + "grad_norm": 1.135886788368225, + "learning_rate": 4.0717159542035775e-05, + "loss": 0.3781, + "step": 40981 + }, + { + "epoch": 2.843799875095413, + "grad_norm": 1.1824935674667358, + "learning_rate": 4.0712634802735964e-05, + "loss": 0.3419, + "step": 40982 + }, + { + "epoch": 2.84386926653251, + "grad_norm": 1.538381814956665, + "learning_rate": 4.070811025059944e-05, + "loss": 0.4757, + "step": 40983 + }, + { + "epoch": 2.8439386579696064, + "grad_norm": 1.2782068252563477, + "learning_rate": 4.0703585885640546e-05, + "loss": 0.4108, + "step": 40984 + }, + { + "epoch": 2.8440080494067033, + "grad_norm": 1.5434914827346802, + "learning_rate": 4.069906170787354e-05, + "loss": 0.4964, + "step": 40985 + }, + { + "epoch": 2.8440774408438, + "grad_norm": 0.962635338306427, + "learning_rate": 4.0694537717312734e-05, + "loss": 0.2735, + "step": 40986 + }, + { + "epoch": 2.8441468322808965, + "grad_norm": 1.2274694442749023, + "learning_rate": 4.069001391397237e-05, + "loss": 0.5245, + "step": 40987 + }, + { + "epoch": 2.8442162237179933, + "grad_norm": 1.0712499618530273, + "learning_rate": 4.06854902978668e-05, + "loss": 0.5086, + "step": 40988 + }, + { + "epoch": 2.8442856151550897, + "grad_norm": 1.0542700290679932, + "learning_rate": 4.068096686901022e-05, + "loss": 0.3926, + "step": 40989 + }, + { + "epoch": 2.8443550065921865, + "grad_norm": 1.1548726558685303, + "learning_rate": 4.0676443627416935e-05, + "loss": 0.3996, + "step": 40990 + }, + { + "epoch": 2.8444243980292834, + "grad_norm": 0.9819570183753967, + "learning_rate": 4.067192057310124e-05, + "loss": 0.4076, + "step": 40991 + }, + { + "epoch": 2.8444937894663798, + "grad_norm": 1.3793442249298096, + "learning_rate": 4.066739770607742e-05, + "loss": 0.422, + "step": 40992 + }, + { + "epoch": 2.8445631809034766, + "grad_norm": 0.7361540794372559, + "learning_rate": 4.066287502635976e-05, + "loss": 0.3216, + "step": 40993 + }, + { + "epoch": 2.844632572340573, + "grad_norm": 1.607470989227295, + "learning_rate": 4.0658352533962494e-05, + "loss": 0.4932, + "step": 40994 + }, + { + "epoch": 2.84470196377767, + "grad_norm": 1.5086348056793213, + "learning_rate": 4.065383022889992e-05, + "loss": 0.4602, + "step": 40995 + }, + { + "epoch": 2.8447713552147667, + "grad_norm": 1.1254440546035767, + "learning_rate": 4.0649308111186355e-05, + "loss": 0.3205, + "step": 40996 + }, + { + "epoch": 2.844840746651863, + "grad_norm": 1.0637049674987793, + "learning_rate": 4.0644786180836e-05, + "loss": 0.3739, + "step": 40997 + }, + { + "epoch": 2.84491013808896, + "grad_norm": 0.8251523971557617, + "learning_rate": 4.064026443786317e-05, + "loss": 0.2695, + "step": 40998 + }, + { + "epoch": 2.8449795295260563, + "grad_norm": 1.3005797863006592, + "learning_rate": 4.0635742882282145e-05, + "loss": 0.4661, + "step": 40999 + }, + { + "epoch": 2.845048920963153, + "grad_norm": 1.1149184703826904, + "learning_rate": 4.0631221514107185e-05, + "loss": 0.2473, + "step": 41000 + }, + { + "epoch": 2.84511831240025, + "grad_norm": 1.3256748914718628, + "learning_rate": 4.0626700333352576e-05, + "loss": 0.428, + "step": 41001 + }, + { + "epoch": 2.8451877038373463, + "grad_norm": 2.4460437297821045, + "learning_rate": 4.0622179340032616e-05, + "loss": 0.5063, + "step": 41002 + }, + { + "epoch": 2.845257095274443, + "grad_norm": 1.1584885120391846, + "learning_rate": 4.061765853416151e-05, + "loss": 0.3405, + "step": 41003 + }, + { + "epoch": 2.8453264867115395, + "grad_norm": 0.961596667766571, + "learning_rate": 4.061313791575357e-05, + "loss": 0.4724, + "step": 41004 + }, + { + "epoch": 2.8453958781486364, + "grad_norm": 0.889873206615448, + "learning_rate": 4.060861748482306e-05, + "loss": 0.3003, + "step": 41005 + }, + { + "epoch": 2.845465269585733, + "grad_norm": 1.2005482912063599, + "learning_rate": 4.0604097241384256e-05, + "loss": 0.4275, + "step": 41006 + }, + { + "epoch": 2.84553466102283, + "grad_norm": 1.2383780479431152, + "learning_rate": 4.0599577185451434e-05, + "loss": 0.3815, + "step": 41007 + }, + { + "epoch": 2.8456040524599264, + "grad_norm": 1.066947102546692, + "learning_rate": 4.0595057317038844e-05, + "loss": 0.3176, + "step": 41008 + }, + { + "epoch": 2.8456734438970233, + "grad_norm": 1.564592957496643, + "learning_rate": 4.05905376361608e-05, + "loss": 0.3688, + "step": 41009 + }, + { + "epoch": 2.8457428353341196, + "grad_norm": 1.0665161609649658, + "learning_rate": 4.05860181428315e-05, + "loss": 0.3304, + "step": 41010 + }, + { + "epoch": 2.8458122267712165, + "grad_norm": 1.2258288860321045, + "learning_rate": 4.058149883706526e-05, + "loss": 0.3645, + "step": 41011 + }, + { + "epoch": 2.8458816182083133, + "grad_norm": 0.751020073890686, + "learning_rate": 4.057697971887633e-05, + "loss": 0.2281, + "step": 41012 + }, + { + "epoch": 2.8459510096454097, + "grad_norm": 1.7382590770721436, + "learning_rate": 4.057246078827898e-05, + "loss": 0.5804, + "step": 41013 + }, + { + "epoch": 2.8460204010825065, + "grad_norm": 1.2262924909591675, + "learning_rate": 4.056794204528747e-05, + "loss": 0.4356, + "step": 41014 + }, + { + "epoch": 2.846089792519603, + "grad_norm": 1.0435621738433838, + "learning_rate": 4.056342348991607e-05, + "loss": 0.2512, + "step": 41015 + }, + { + "epoch": 2.8461591839566998, + "grad_norm": 1.209458589553833, + "learning_rate": 4.055890512217909e-05, + "loss": 0.4732, + "step": 41016 + }, + { + "epoch": 2.8462285753937966, + "grad_norm": 1.1048752069473267, + "learning_rate": 4.0554386942090706e-05, + "loss": 0.2802, + "step": 41017 + }, + { + "epoch": 2.846297966830893, + "grad_norm": 0.9250602126121521, + "learning_rate": 4.0549868949665235e-05, + "loss": 0.3244, + "step": 41018 + }, + { + "epoch": 2.84636735826799, + "grad_norm": 0.9930047988891602, + "learning_rate": 4.054535114491692e-05, + "loss": 0.3501, + "step": 41019 + }, + { + "epoch": 2.846436749705086, + "grad_norm": 1.0230456590652466, + "learning_rate": 4.0540833527860043e-05, + "loss": 0.3167, + "step": 41020 + }, + { + "epoch": 2.846506141142183, + "grad_norm": 1.2821965217590332, + "learning_rate": 4.053631609850885e-05, + "loss": 0.406, + "step": 41021 + }, + { + "epoch": 2.84657553257928, + "grad_norm": 1.477034330368042, + "learning_rate": 4.053179885687761e-05, + "loss": 0.5206, + "step": 41022 + }, + { + "epoch": 2.8466449240163763, + "grad_norm": 1.2550195455551147, + "learning_rate": 4.052728180298062e-05, + "loss": 0.4547, + "step": 41023 + }, + { + "epoch": 2.846714315453473, + "grad_norm": 1.2846213579177856, + "learning_rate": 4.052276493683207e-05, + "loss": 0.4324, + "step": 41024 + }, + { + "epoch": 2.8467837068905695, + "grad_norm": 1.0151015520095825, + "learning_rate": 4.051824825844623e-05, + "loss": 0.3434, + "step": 41025 + }, + { + "epoch": 2.8468530983276663, + "grad_norm": 1.000616431236267, + "learning_rate": 4.05137317678374e-05, + "loss": 0.3958, + "step": 41026 + }, + { + "epoch": 2.846922489764763, + "grad_norm": 0.9212082028388977, + "learning_rate": 4.0509215465019844e-05, + "loss": 0.329, + "step": 41027 + }, + { + "epoch": 2.8469918812018595, + "grad_norm": 1.061315655708313, + "learning_rate": 4.050469935000776e-05, + "loss": 0.3414, + "step": 41028 + }, + { + "epoch": 2.8470612726389564, + "grad_norm": 1.3123599290847778, + "learning_rate": 4.050018342281544e-05, + "loss": 0.4298, + "step": 41029 + }, + { + "epoch": 2.8471306640760528, + "grad_norm": 1.0684949159622192, + "learning_rate": 4.0495667683457125e-05, + "loss": 0.2061, + "step": 41030 + }, + { + "epoch": 2.8472000555131496, + "grad_norm": 1.063755750656128, + "learning_rate": 4.049115213194713e-05, + "loss": 0.3613, + "step": 41031 + }, + { + "epoch": 2.8472694469502464, + "grad_norm": 0.9883999824523926, + "learning_rate": 4.048663676829963e-05, + "loss": 0.2986, + "step": 41032 + }, + { + "epoch": 2.8473388383873433, + "grad_norm": 0.96517413854599, + "learning_rate": 4.048212159252891e-05, + "loss": 0.3707, + "step": 41033 + }, + { + "epoch": 2.8474082298244396, + "grad_norm": 1.3788113594055176, + "learning_rate": 4.047760660464922e-05, + "loss": 0.4435, + "step": 41034 + }, + { + "epoch": 2.8474776212615365, + "grad_norm": 1.4195629358291626, + "learning_rate": 4.047309180467482e-05, + "loss": 0.3504, + "step": 41035 + }, + { + "epoch": 2.847547012698633, + "grad_norm": 0.8794300556182861, + "learning_rate": 4.046857719261997e-05, + "loss": 0.3254, + "step": 41036 + }, + { + "epoch": 2.8476164041357297, + "grad_norm": 1.6949872970581055, + "learning_rate": 4.046406276849895e-05, + "loss": 0.4511, + "step": 41037 + }, + { + "epoch": 2.8476857955728265, + "grad_norm": 1.2580515146255493, + "learning_rate": 4.045954853232594e-05, + "loss": 0.3533, + "step": 41038 + }, + { + "epoch": 2.847755187009923, + "grad_norm": 1.5186501741409302, + "learning_rate": 4.045503448411523e-05, + "loss": 0.3833, + "step": 41039 + }, + { + "epoch": 2.8478245784470197, + "grad_norm": 0.7442593574523926, + "learning_rate": 4.045052062388106e-05, + "loss": 0.234, + "step": 41040 + }, + { + "epoch": 2.847893969884116, + "grad_norm": 1.3867090940475464, + "learning_rate": 4.0446006951637696e-05, + "loss": 0.4795, + "step": 41041 + }, + { + "epoch": 2.847963361321213, + "grad_norm": 0.7003527283668518, + "learning_rate": 4.044149346739937e-05, + "loss": 0.3121, + "step": 41042 + }, + { + "epoch": 2.84803275275831, + "grad_norm": 1.5929569005966187, + "learning_rate": 4.0436980171180335e-05, + "loss": 0.5175, + "step": 41043 + }, + { + "epoch": 2.848102144195406, + "grad_norm": 0.8596017956733704, + "learning_rate": 4.04324670629949e-05, + "loss": 0.331, + "step": 41044 + }, + { + "epoch": 2.848171535632503, + "grad_norm": 1.0106275081634521, + "learning_rate": 4.04279541428572e-05, + "loss": 0.2666, + "step": 41045 + }, + { + "epoch": 2.8482409270695994, + "grad_norm": 1.7804386615753174, + "learning_rate": 4.0423441410781545e-05, + "loss": 0.3744, + "step": 41046 + }, + { + "epoch": 2.8483103185066962, + "grad_norm": 1.2136110067367554, + "learning_rate": 4.041892886678217e-05, + "loss": 0.4225, + "step": 41047 + }, + { + "epoch": 2.848379709943793, + "grad_norm": 1.6123067140579224, + "learning_rate": 4.0414416510873334e-05, + "loss": 0.5378, + "step": 41048 + }, + { + "epoch": 2.8484491013808895, + "grad_norm": 1.0544437170028687, + "learning_rate": 4.040990434306926e-05, + "loss": 0.3516, + "step": 41049 + }, + { + "epoch": 2.8485184928179863, + "grad_norm": 1.0433324575424194, + "learning_rate": 4.040539236338421e-05, + "loss": 0.3388, + "step": 41050 + }, + { + "epoch": 2.8485878842550827, + "grad_norm": 1.106245517730713, + "learning_rate": 4.040088057183246e-05, + "loss": 0.2991, + "step": 41051 + }, + { + "epoch": 2.8486572756921795, + "grad_norm": 1.063374400138855, + "learning_rate": 4.039636896842817e-05, + "loss": 0.4472, + "step": 41052 + }, + { + "epoch": 2.8487266671292764, + "grad_norm": 1.3628844022750854, + "learning_rate": 4.039185755318563e-05, + "loss": 0.3393, + "step": 41053 + }, + { + "epoch": 2.8487960585663727, + "grad_norm": 1.9809876680374146, + "learning_rate": 4.038734632611908e-05, + "loss": 0.446, + "step": 41054 + }, + { + "epoch": 2.8488654500034696, + "grad_norm": 1.3158448934555054, + "learning_rate": 4.038283528724277e-05, + "loss": 0.4966, + "step": 41055 + }, + { + "epoch": 2.848934841440566, + "grad_norm": 1.2305183410644531, + "learning_rate": 4.037832443657092e-05, + "loss": 0.3947, + "step": 41056 + }, + { + "epoch": 2.849004232877663, + "grad_norm": 0.9989683628082275, + "learning_rate": 4.037381377411779e-05, + "loss": 0.3384, + "step": 41057 + }, + { + "epoch": 2.8490736243147596, + "grad_norm": 1.1154276132583618, + "learning_rate": 4.036930329989764e-05, + "loss": 0.4642, + "step": 41058 + }, + { + "epoch": 2.8491430157518565, + "grad_norm": 1.2377352714538574, + "learning_rate": 4.0364793013924664e-05, + "loss": 0.4168, + "step": 41059 + }, + { + "epoch": 2.849212407188953, + "grad_norm": 1.217472791671753, + "learning_rate": 4.03602829162131e-05, + "loss": 0.3754, + "step": 41060 + }, + { + "epoch": 2.8492817986260497, + "grad_norm": 1.462589144706726, + "learning_rate": 4.035577300677721e-05, + "loss": 0.5233, + "step": 41061 + }, + { + "epoch": 2.849351190063146, + "grad_norm": 1.1350494623184204, + "learning_rate": 4.035126328563126e-05, + "loss": 0.4277, + "step": 41062 + }, + { + "epoch": 2.849420581500243, + "grad_norm": 0.8505650758743286, + "learning_rate": 4.0346753752789415e-05, + "loss": 0.2501, + "step": 41063 + }, + { + "epoch": 2.8494899729373397, + "grad_norm": 1.3380452394485474, + "learning_rate": 4.034224440826594e-05, + "loss": 0.3865, + "step": 41064 + }, + { + "epoch": 2.849559364374436, + "grad_norm": 1.8817836046218872, + "learning_rate": 4.0337735252075095e-05, + "loss": 0.4313, + "step": 41065 + }, + { + "epoch": 2.849628755811533, + "grad_norm": 1.2045496702194214, + "learning_rate": 4.033322628423112e-05, + "loss": 0.4078, + "step": 41066 + }, + { + "epoch": 2.8496981472486294, + "grad_norm": 4.369763374328613, + "learning_rate": 4.032871750474819e-05, + "loss": 0.4424, + "step": 41067 + }, + { + "epoch": 2.849767538685726, + "grad_norm": 0.9478738307952881, + "learning_rate": 4.032420891364058e-05, + "loss": 0.4251, + "step": 41068 + }, + { + "epoch": 2.849836930122823, + "grad_norm": 1.0714848041534424, + "learning_rate": 4.031970051092252e-05, + "loss": 0.3868, + "step": 41069 + }, + { + "epoch": 2.8499063215599194, + "grad_norm": 1.0498179197311401, + "learning_rate": 4.031519229660824e-05, + "loss": 0.4003, + "step": 41070 + }, + { + "epoch": 2.8499757129970162, + "grad_norm": 1.2408679723739624, + "learning_rate": 4.0310684270712004e-05, + "loss": 0.4173, + "step": 41071 + }, + { + "epoch": 2.8500451044341126, + "grad_norm": 1.111157774925232, + "learning_rate": 4.0306176433247986e-05, + "loss": 0.3925, + "step": 41072 + }, + { + "epoch": 2.8501144958712095, + "grad_norm": 1.3629978895187378, + "learning_rate": 4.030166878423044e-05, + "loss": 0.4072, + "step": 41073 + }, + { + "epoch": 2.8501838873083063, + "grad_norm": 1.6325777769088745, + "learning_rate": 4.0297161323673594e-05, + "loss": 0.5022, + "step": 41074 + }, + { + "epoch": 2.8502532787454027, + "grad_norm": 1.1715848445892334, + "learning_rate": 4.0292654051591695e-05, + "loss": 0.3279, + "step": 41075 + }, + { + "epoch": 2.8503226701824995, + "grad_norm": 0.9472123384475708, + "learning_rate": 4.028814696799895e-05, + "loss": 0.2374, + "step": 41076 + }, + { + "epoch": 2.850392061619596, + "grad_norm": 1.0747127532958984, + "learning_rate": 4.02836400729096e-05, + "loss": 0.3133, + "step": 41077 + }, + { + "epoch": 2.8504614530566927, + "grad_norm": 1.1883575916290283, + "learning_rate": 4.0279133366337896e-05, + "loss": 0.4744, + "step": 41078 + }, + { + "epoch": 2.8505308444937896, + "grad_norm": 1.183260202407837, + "learning_rate": 4.0274626848298015e-05, + "loss": 0.4119, + "step": 41079 + }, + { + "epoch": 2.8506002359308864, + "grad_norm": 1.2278544902801514, + "learning_rate": 4.027012051880421e-05, + "loss": 0.3872, + "step": 41080 + }, + { + "epoch": 2.850669627367983, + "grad_norm": 1.19392728805542, + "learning_rate": 4.0265614377870696e-05, + "loss": 0.3423, + "step": 41081 + }, + { + "epoch": 2.850739018805079, + "grad_norm": 0.7670843005180359, + "learning_rate": 4.0261108425511706e-05, + "loss": 0.2941, + "step": 41082 + }, + { + "epoch": 2.850808410242176, + "grad_norm": 1.2894340753555298, + "learning_rate": 4.025660266174147e-05, + "loss": 0.4365, + "step": 41083 + }, + { + "epoch": 2.850877801679273, + "grad_norm": 1.518393635749817, + "learning_rate": 4.025209708657421e-05, + "loss": 0.4047, + "step": 41084 + }, + { + "epoch": 2.8509471931163697, + "grad_norm": 0.89057856798172, + "learning_rate": 4.024759170002418e-05, + "loss": 0.3603, + "step": 41085 + }, + { + "epoch": 2.851016584553466, + "grad_norm": 0.8851302266120911, + "learning_rate": 4.024308650210554e-05, + "loss": 0.3213, + "step": 41086 + }, + { + "epoch": 2.851085975990563, + "grad_norm": 1.8606868982315063, + "learning_rate": 4.023858149283254e-05, + "loss": 0.53, + "step": 41087 + }, + { + "epoch": 2.8511553674276593, + "grad_norm": 0.9773709177970886, + "learning_rate": 4.0234076672219404e-05, + "loss": 0.3353, + "step": 41088 + }, + { + "epoch": 2.851224758864756, + "grad_norm": 1.2914223670959473, + "learning_rate": 4.022957204028035e-05, + "loss": 0.3992, + "step": 41089 + }, + { + "epoch": 2.851294150301853, + "grad_norm": 1.1659494638442993, + "learning_rate": 4.0225067597029606e-05, + "loss": 0.4498, + "step": 41090 + }, + { + "epoch": 2.8513635417389493, + "grad_norm": 1.1992727518081665, + "learning_rate": 4.0220563342481385e-05, + "loss": 0.3725, + "step": 41091 + }, + { + "epoch": 2.851432933176046, + "grad_norm": 1.6482586860656738, + "learning_rate": 4.021605927664996e-05, + "loss": 0.5338, + "step": 41092 + }, + { + "epoch": 2.8515023246131426, + "grad_norm": 1.4426358938217163, + "learning_rate": 4.0211555399549436e-05, + "loss": 0.3169, + "step": 41093 + }, + { + "epoch": 2.8515717160502394, + "grad_norm": 1.3961925506591797, + "learning_rate": 4.020705171119412e-05, + "loss": 0.4744, + "step": 41094 + }, + { + "epoch": 2.8516411074873362, + "grad_norm": 1.2915524244308472, + "learning_rate": 4.020254821159819e-05, + "loss": 0.3018, + "step": 41095 + }, + { + "epoch": 2.8517104989244326, + "grad_norm": 0.8947917819023132, + "learning_rate": 4.019804490077589e-05, + "loss": 0.3405, + "step": 41096 + }, + { + "epoch": 2.8517798903615295, + "grad_norm": 1.1247878074645996, + "learning_rate": 4.019354177874145e-05, + "loss": 0.2954, + "step": 41097 + }, + { + "epoch": 2.851849281798626, + "grad_norm": 1.606826901435852, + "learning_rate": 4.018903884550903e-05, + "loss": 0.3414, + "step": 41098 + }, + { + "epoch": 2.8519186732357227, + "grad_norm": 0.8152243494987488, + "learning_rate": 4.0184536101092874e-05, + "loss": 0.2361, + "step": 41099 + }, + { + "epoch": 2.8519880646728195, + "grad_norm": 0.8896542191505432, + "learning_rate": 4.018003354550719e-05, + "loss": 0.3176, + "step": 41100 + }, + { + "epoch": 2.852057456109916, + "grad_norm": 1.4139436483383179, + "learning_rate": 4.017553117876624e-05, + "loss": 0.4313, + "step": 41101 + }, + { + "epoch": 2.8521268475470127, + "grad_norm": 1.0825035572052002, + "learning_rate": 4.0171029000884165e-05, + "loss": 0.3615, + "step": 41102 + }, + { + "epoch": 2.852196238984109, + "grad_norm": 1.4013506174087524, + "learning_rate": 4.016652701187521e-05, + "loss": 0.4765, + "step": 41103 + }, + { + "epoch": 2.852265630421206, + "grad_norm": 1.2929213047027588, + "learning_rate": 4.016202521175358e-05, + "loss": 0.3349, + "step": 41104 + }, + { + "epoch": 2.852335021858303, + "grad_norm": 0.9904000163078308, + "learning_rate": 4.01575236005335e-05, + "loss": 0.3821, + "step": 41105 + }, + { + "epoch": 2.8524044132953996, + "grad_norm": 1.2284624576568604, + "learning_rate": 4.0153022178229203e-05, + "loss": 0.4391, + "step": 41106 + }, + { + "epoch": 2.852473804732496, + "grad_norm": 1.2727659940719604, + "learning_rate": 4.014852094485485e-05, + "loss": 0.4123, + "step": 41107 + }, + { + "epoch": 2.8525431961695924, + "grad_norm": 1.2737150192260742, + "learning_rate": 4.0144019900424665e-05, + "loss": 0.4397, + "step": 41108 + }, + { + "epoch": 2.8526125876066892, + "grad_norm": 0.9605419039726257, + "learning_rate": 4.0139519044952855e-05, + "loss": 0.2751, + "step": 41109 + }, + { + "epoch": 2.852681979043786, + "grad_norm": 1.4816243648529053, + "learning_rate": 4.013501837845365e-05, + "loss": 0.3021, + "step": 41110 + }, + { + "epoch": 2.852751370480883, + "grad_norm": 1.178921103477478, + "learning_rate": 4.0130517900941246e-05, + "loss": 0.4556, + "step": 41111 + }, + { + "epoch": 2.8528207619179793, + "grad_norm": 1.0571752786636353, + "learning_rate": 4.012601761242985e-05, + "loss": 0.4205, + "step": 41112 + }, + { + "epoch": 2.852890153355076, + "grad_norm": 0.7955881357192993, + "learning_rate": 4.01215175129337e-05, + "loss": 0.2119, + "step": 41113 + }, + { + "epoch": 2.8529595447921725, + "grad_norm": 0.9820027947425842, + "learning_rate": 4.0117017602466945e-05, + "loss": 0.3511, + "step": 41114 + }, + { + "epoch": 2.8530289362292693, + "grad_norm": 1.2122005224227905, + "learning_rate": 4.011251788104381e-05, + "loss": 0.3196, + "step": 41115 + }, + { + "epoch": 2.853098327666366, + "grad_norm": 2.346144914627075, + "learning_rate": 4.010801834867851e-05, + "loss": 0.5337, + "step": 41116 + }, + { + "epoch": 2.8531677191034626, + "grad_norm": 0.9787795543670654, + "learning_rate": 4.010351900538525e-05, + "loss": 0.3696, + "step": 41117 + }, + { + "epoch": 2.8532371105405594, + "grad_norm": 1.5105016231536865, + "learning_rate": 4.0099019851178234e-05, + "loss": 0.3963, + "step": 41118 + }, + { + "epoch": 2.853306501977656, + "grad_norm": 1.5447757244110107, + "learning_rate": 4.0094520886071654e-05, + "loss": 0.4427, + "step": 41119 + }, + { + "epoch": 2.8533758934147526, + "grad_norm": 1.025879979133606, + "learning_rate": 4.009002211007976e-05, + "loss": 0.3023, + "step": 41120 + }, + { + "epoch": 2.8534452848518495, + "grad_norm": 1.1532726287841797, + "learning_rate": 4.008552352321668e-05, + "loss": 0.3415, + "step": 41121 + }, + { + "epoch": 2.853514676288946, + "grad_norm": 1.0366623401641846, + "learning_rate": 4.008102512549665e-05, + "loss": 0.2872, + "step": 41122 + }, + { + "epoch": 2.8535840677260427, + "grad_norm": 0.7425228953361511, + "learning_rate": 4.007652691693387e-05, + "loss": 0.2642, + "step": 41123 + }, + { + "epoch": 2.853653459163139, + "grad_norm": 0.7150042653083801, + "learning_rate": 4.007202889754255e-05, + "loss": 0.2179, + "step": 41124 + }, + { + "epoch": 2.853722850600236, + "grad_norm": 1.0341945886611938, + "learning_rate": 4.006753106733687e-05, + "loss": 0.3374, + "step": 41125 + }, + { + "epoch": 2.8537922420373327, + "grad_norm": 1.6343462467193604, + "learning_rate": 4.006303342633105e-05, + "loss": 0.4621, + "step": 41126 + }, + { + "epoch": 2.853861633474429, + "grad_norm": 1.1476832628250122, + "learning_rate": 4.0058535974539316e-05, + "loss": 0.5123, + "step": 41127 + }, + { + "epoch": 2.853931024911526, + "grad_norm": 1.1039937734603882, + "learning_rate": 4.00540387119758e-05, + "loss": 0.3982, + "step": 41128 + }, + { + "epoch": 2.8540004163486223, + "grad_norm": 0.9866127371788025, + "learning_rate": 4.0049541638654706e-05, + "loss": 0.3749, + "step": 41129 + }, + { + "epoch": 2.854069807785719, + "grad_norm": 1.434625267982483, + "learning_rate": 4.004504475459028e-05, + "loss": 0.4481, + "step": 41130 + }, + { + "epoch": 2.854139199222816, + "grad_norm": 1.0390784740447998, + "learning_rate": 4.0040548059796666e-05, + "loss": 0.2955, + "step": 41131 + }, + { + "epoch": 2.854208590659913, + "grad_norm": 1.1900907754898071, + "learning_rate": 4.0036051554288136e-05, + "loss": 0.3426, + "step": 41132 + }, + { + "epoch": 2.8542779820970092, + "grad_norm": 0.9843214750289917, + "learning_rate": 4.00315552380788e-05, + "loss": 0.2545, + "step": 41133 + }, + { + "epoch": 2.8543473735341056, + "grad_norm": 0.8612868785858154, + "learning_rate": 4.002705911118288e-05, + "loss": 0.2165, + "step": 41134 + }, + { + "epoch": 2.8544167649712024, + "grad_norm": 1.1680858135223389, + "learning_rate": 4.002256317361458e-05, + "loss": 0.3664, + "step": 41135 + }, + { + "epoch": 2.8544861564082993, + "grad_norm": 0.9335424900054932, + "learning_rate": 4.001806742538813e-05, + "loss": 0.2758, + "step": 41136 + }, + { + "epoch": 2.854555547845396, + "grad_norm": 1.1464205980300903, + "learning_rate": 4.001357186651763e-05, + "loss": 0.3554, + "step": 41137 + }, + { + "epoch": 2.8546249392824925, + "grad_norm": 1.095712423324585, + "learning_rate": 4.000907649701734e-05, + "loss": 0.3755, + "step": 41138 + }, + { + "epoch": 2.8546943307195893, + "grad_norm": 1.0283269882202148, + "learning_rate": 4.000458131690143e-05, + "loss": 0.4763, + "step": 41139 + }, + { + "epoch": 2.8547637221566857, + "grad_norm": 1.0793821811676025, + "learning_rate": 4.0000086326184126e-05, + "loss": 0.3174, + "step": 41140 + }, + { + "epoch": 2.8548331135937826, + "grad_norm": 0.7149763107299805, + "learning_rate": 3.999559152487956e-05, + "loss": 0.2258, + "step": 41141 + }, + { + "epoch": 2.8549025050308794, + "grad_norm": 1.3233015537261963, + "learning_rate": 3.9991096913001946e-05, + "loss": 0.5547, + "step": 41142 + }, + { + "epoch": 2.854971896467976, + "grad_norm": 1.10956871509552, + "learning_rate": 3.998660249056548e-05, + "loss": 0.429, + "step": 41143 + }, + { + "epoch": 2.8550412879050726, + "grad_norm": 0.9866469502449036, + "learning_rate": 3.998210825758435e-05, + "loss": 0.324, + "step": 41144 + }, + { + "epoch": 2.855110679342169, + "grad_norm": 1.2958557605743408, + "learning_rate": 3.997761421407274e-05, + "loss": 0.4962, + "step": 41145 + }, + { + "epoch": 2.855180070779266, + "grad_norm": 1.145233154296875, + "learning_rate": 3.9973120360044825e-05, + "loss": 0.4923, + "step": 41146 + }, + { + "epoch": 2.8552494622163627, + "grad_norm": 1.5116852521896362, + "learning_rate": 3.996862669551485e-05, + "loss": 0.2979, + "step": 41147 + }, + { + "epoch": 2.855318853653459, + "grad_norm": 0.9020809531211853, + "learning_rate": 3.996413322049692e-05, + "loss": 0.3787, + "step": 41148 + }, + { + "epoch": 2.855388245090556, + "grad_norm": 1.188250184059143, + "learning_rate": 3.995963993500526e-05, + "loss": 0.472, + "step": 41149 + }, + { + "epoch": 2.8554576365276523, + "grad_norm": 1.5188000202178955, + "learning_rate": 3.995514683905404e-05, + "loss": 0.3847, + "step": 41150 + }, + { + "epoch": 2.855527027964749, + "grad_norm": 1.2146958112716675, + "learning_rate": 3.9950653932657456e-05, + "loss": 0.587, + "step": 41151 + }, + { + "epoch": 2.855596419401846, + "grad_norm": 1.9499802589416504, + "learning_rate": 3.99461612158297e-05, + "loss": 0.4352, + "step": 41152 + }, + { + "epoch": 2.8556658108389423, + "grad_norm": 1.1131622791290283, + "learning_rate": 3.994166868858494e-05, + "loss": 0.5492, + "step": 41153 + }, + { + "epoch": 2.855735202276039, + "grad_norm": 1.2308193445205688, + "learning_rate": 3.99371763509374e-05, + "loss": 0.3899, + "step": 41154 + }, + { + "epoch": 2.8558045937131356, + "grad_norm": 1.606249451637268, + "learning_rate": 3.9932684202901194e-05, + "loss": 0.5014, + "step": 41155 + }, + { + "epoch": 2.8558739851502324, + "grad_norm": 1.1874662637710571, + "learning_rate": 3.9928192244490524e-05, + "loss": 0.5233, + "step": 41156 + }, + { + "epoch": 2.855943376587329, + "grad_norm": 1.2171378135681152, + "learning_rate": 3.9923700475719594e-05, + "loss": 0.344, + "step": 41157 + }, + { + "epoch": 2.856012768024426, + "grad_norm": 1.2506511211395264, + "learning_rate": 3.9919208896602566e-05, + "loss": 0.3829, + "step": 41158 + }, + { + "epoch": 2.8560821594615224, + "grad_norm": 1.2251440286636353, + "learning_rate": 3.991471750715363e-05, + "loss": 0.4908, + "step": 41159 + }, + { + "epoch": 2.8561515508986193, + "grad_norm": 0.8851630091667175, + "learning_rate": 3.9910226307386956e-05, + "loss": 0.3064, + "step": 41160 + }, + { + "epoch": 2.8562209423357157, + "grad_norm": 0.8952000141143799, + "learning_rate": 3.990573529731676e-05, + "loss": 0.2605, + "step": 41161 + }, + { + "epoch": 2.8562903337728125, + "grad_norm": 1.0754578113555908, + "learning_rate": 3.990124447695714e-05, + "loss": 0.244, + "step": 41162 + }, + { + "epoch": 2.8563597252099093, + "grad_norm": 1.2423080205917358, + "learning_rate": 3.989675384632233e-05, + "loss": 0.5057, + "step": 41163 + }, + { + "epoch": 2.8564291166470057, + "grad_norm": 1.0260088443756104, + "learning_rate": 3.98922634054265e-05, + "loss": 0.3863, + "step": 41164 + }, + { + "epoch": 2.8564985080841025, + "grad_norm": 1.6681991815567017, + "learning_rate": 3.988777315428381e-05, + "loss": 0.3579, + "step": 41165 + }, + { + "epoch": 2.856567899521199, + "grad_norm": 0.9134418368339539, + "learning_rate": 3.9883283092908454e-05, + "loss": 0.3549, + "step": 41166 + }, + { + "epoch": 2.8566372909582958, + "grad_norm": 0.9758260250091553, + "learning_rate": 3.987879322131463e-05, + "loss": 0.4133, + "step": 41167 + }, + { + "epoch": 2.8567066823953926, + "grad_norm": 0.8415502905845642, + "learning_rate": 3.9874303539516443e-05, + "loss": 0.2738, + "step": 41168 + }, + { + "epoch": 2.856776073832489, + "grad_norm": 0.6690188646316528, + "learning_rate": 3.986981404752811e-05, + "loss": 0.2586, + "step": 41169 + }, + { + "epoch": 2.856845465269586, + "grad_norm": 1.3614232540130615, + "learning_rate": 3.986532474536382e-05, + "loss": 0.4053, + "step": 41170 + }, + { + "epoch": 2.856914856706682, + "grad_norm": 1.1347695589065552, + "learning_rate": 3.9860835633037695e-05, + "loss": 0.3592, + "step": 41171 + }, + { + "epoch": 2.856984248143779, + "grad_norm": 1.2625694274902344, + "learning_rate": 3.985634671056394e-05, + "loss": 0.3876, + "step": 41172 + }, + { + "epoch": 2.857053639580876, + "grad_norm": 1.0385937690734863, + "learning_rate": 3.985185797795672e-05, + "loss": 0.3668, + "step": 41173 + }, + { + "epoch": 2.8571230310179723, + "grad_norm": 0.7458078265190125, + "learning_rate": 3.9847369435230196e-05, + "loss": 0.1443, + "step": 41174 + }, + { + "epoch": 2.857192422455069, + "grad_norm": 0.9619347453117371, + "learning_rate": 3.9842881082398585e-05, + "loss": 0.192, + "step": 41175 + }, + { + "epoch": 2.8572618138921655, + "grad_norm": 0.8002005815505981, + "learning_rate": 3.9838392919475995e-05, + "loss": 0.2004, + "step": 41176 + }, + { + "epoch": 2.8573312053292623, + "grad_norm": 1.165939450263977, + "learning_rate": 3.983390494647661e-05, + "loss": 0.4492, + "step": 41177 + }, + { + "epoch": 2.857400596766359, + "grad_norm": 1.0923155546188354, + "learning_rate": 3.9829417163414606e-05, + "loss": 0.3574, + "step": 41178 + }, + { + "epoch": 2.8574699882034555, + "grad_norm": 1.1316499710083008, + "learning_rate": 3.982492957030416e-05, + "loss": 0.2972, + "step": 41179 + }, + { + "epoch": 2.8575393796405524, + "grad_norm": 1.2269879579544067, + "learning_rate": 3.982044216715942e-05, + "loss": 0.5347, + "step": 41180 + }, + { + "epoch": 2.8576087710776488, + "grad_norm": 0.9310880899429321, + "learning_rate": 3.981595495399456e-05, + "loss": 0.3606, + "step": 41181 + }, + { + "epoch": 2.8576781625147456, + "grad_norm": 1.552562952041626, + "learning_rate": 3.9811467930823784e-05, + "loss": 0.4385, + "step": 41182 + }, + { + "epoch": 2.8577475539518424, + "grad_norm": 1.2484221458435059, + "learning_rate": 3.9806981097661186e-05, + "loss": 0.4611, + "step": 41183 + }, + { + "epoch": 2.8578169453889393, + "grad_norm": 1.4267710447311401, + "learning_rate": 3.980249445452098e-05, + "loss": 0.4372, + "step": 41184 + }, + { + "epoch": 2.8578863368260357, + "grad_norm": 0.9629202485084534, + "learning_rate": 3.97980080014173e-05, + "loss": 0.415, + "step": 41185 + }, + { + "epoch": 2.8579557282631325, + "grad_norm": 1.007912039756775, + "learning_rate": 3.979352173836433e-05, + "loss": 0.3312, + "step": 41186 + }, + { + "epoch": 2.858025119700229, + "grad_norm": 1.1666409969329834, + "learning_rate": 3.978903566537622e-05, + "loss": 0.4347, + "step": 41187 + }, + { + "epoch": 2.8580945111373257, + "grad_norm": 1.0543550252914429, + "learning_rate": 3.978454978246714e-05, + "loss": 0.4236, + "step": 41188 + }, + { + "epoch": 2.8581639025744225, + "grad_norm": 1.0892616510391235, + "learning_rate": 3.978006408965128e-05, + "loss": 0.3474, + "step": 41189 + }, + { + "epoch": 2.858233294011519, + "grad_norm": 1.2714499235153198, + "learning_rate": 3.977557858694275e-05, + "loss": 0.4329, + "step": 41190 + }, + { + "epoch": 2.8583026854486158, + "grad_norm": 1.3933756351470947, + "learning_rate": 3.977109327435572e-05, + "loss": 0.4528, + "step": 41191 + }, + { + "epoch": 2.858372076885712, + "grad_norm": 0.9734860062599182, + "learning_rate": 3.976660815190436e-05, + "loss": 0.2199, + "step": 41192 + }, + { + "epoch": 2.858441468322809, + "grad_norm": 1.328102707862854, + "learning_rate": 3.9762123219602846e-05, + "loss": 0.5991, + "step": 41193 + }, + { + "epoch": 2.858510859759906, + "grad_norm": 1.1174005270004272, + "learning_rate": 3.97576384774653e-05, + "loss": 0.3874, + "step": 41194 + }, + { + "epoch": 2.858580251197002, + "grad_norm": 1.6456493139266968, + "learning_rate": 3.9753153925505904e-05, + "loss": 0.4494, + "step": 41195 + }, + { + "epoch": 2.858649642634099, + "grad_norm": 1.0403215885162354, + "learning_rate": 3.974866956373885e-05, + "loss": 0.3437, + "step": 41196 + }, + { + "epoch": 2.8587190340711954, + "grad_norm": 1.0022364854812622, + "learning_rate": 3.9744185392178223e-05, + "loss": 0.4082, + "step": 41197 + }, + { + "epoch": 2.8587884255082923, + "grad_norm": 1.3078539371490479, + "learning_rate": 3.973970141083822e-05, + "loss": 0.5275, + "step": 41198 + }, + { + "epoch": 2.858857816945389, + "grad_norm": 1.2492365837097168, + "learning_rate": 3.973521761973297e-05, + "loss": 0.5282, + "step": 41199 + }, + { + "epoch": 2.8589272083824855, + "grad_norm": 1.2677353620529175, + "learning_rate": 3.973073401887666e-05, + "loss": 0.499, + "step": 41200 + }, + { + "epoch": 2.8589965998195823, + "grad_norm": 0.9812806248664856, + "learning_rate": 3.972625060828342e-05, + "loss": 0.2906, + "step": 41201 + }, + { + "epoch": 2.8590659912566787, + "grad_norm": 1.5741482973098755, + "learning_rate": 3.972176738796745e-05, + "loss": 0.2991, + "step": 41202 + }, + { + "epoch": 2.8591353826937755, + "grad_norm": 0.6366034150123596, + "learning_rate": 3.971728435794283e-05, + "loss": 0.1632, + "step": 41203 + }, + { + "epoch": 2.8592047741308724, + "grad_norm": 1.0599117279052734, + "learning_rate": 3.971280151822376e-05, + "loss": 0.416, + "step": 41204 + }, + { + "epoch": 2.8592741655679688, + "grad_norm": 1.586780071258545, + "learning_rate": 3.970831886882441e-05, + "loss": 0.3821, + "step": 41205 + }, + { + "epoch": 2.8593435570050656, + "grad_norm": 1.069637656211853, + "learning_rate": 3.9703836409758866e-05, + "loss": 0.3273, + "step": 41206 + }, + { + "epoch": 2.859412948442162, + "grad_norm": 1.15119206905365, + "learning_rate": 3.9699354141041315e-05, + "loss": 0.523, + "step": 41207 + }, + { + "epoch": 2.859482339879259, + "grad_norm": 1.600436806678772, + "learning_rate": 3.969487206268591e-05, + "loss": 0.5327, + "step": 41208 + }, + { + "epoch": 2.8595517313163556, + "grad_norm": 1.1610771417617798, + "learning_rate": 3.9690390174706794e-05, + "loss": 0.2393, + "step": 41209 + }, + { + "epoch": 2.8596211227534525, + "grad_norm": 1.2239981889724731, + "learning_rate": 3.968590847711815e-05, + "loss": 0.425, + "step": 41210 + }, + { + "epoch": 2.859690514190549, + "grad_norm": 1.3044246435165405, + "learning_rate": 3.968142696993407e-05, + "loss": 0.5052, + "step": 41211 + }, + { + "epoch": 2.8597599056276457, + "grad_norm": 1.6584922075271606, + "learning_rate": 3.967694565316872e-05, + "loss": 0.4443, + "step": 41212 + }, + { + "epoch": 2.859829297064742, + "grad_norm": 0.8807470202445984, + "learning_rate": 3.967246452683625e-05, + "loss": 0.3106, + "step": 41213 + }, + { + "epoch": 2.859898688501839, + "grad_norm": 1.682665228843689, + "learning_rate": 3.966798359095082e-05, + "loss": 0.4334, + "step": 41214 + }, + { + "epoch": 2.8599680799389358, + "grad_norm": 1.1715084314346313, + "learning_rate": 3.966350284552656e-05, + "loss": 0.3283, + "step": 41215 + }, + { + "epoch": 2.860037471376032, + "grad_norm": 1.6498292684555054, + "learning_rate": 3.965902229057766e-05, + "loss": 0.4488, + "step": 41216 + }, + { + "epoch": 2.860106862813129, + "grad_norm": 1.6102690696716309, + "learning_rate": 3.9654541926118194e-05, + "loss": 0.5141, + "step": 41217 + }, + { + "epoch": 2.8601762542502254, + "grad_norm": 1.669503927230835, + "learning_rate": 3.9650061752162325e-05, + "loss": 0.4008, + "step": 41218 + }, + { + "epoch": 2.860245645687322, + "grad_norm": 1.2071467638015747, + "learning_rate": 3.9645581768724216e-05, + "loss": 0.338, + "step": 41219 + }, + { + "epoch": 2.860315037124419, + "grad_norm": 1.1458890438079834, + "learning_rate": 3.964110197581801e-05, + "loss": 0.3813, + "step": 41220 + }, + { + "epoch": 2.8603844285615154, + "grad_norm": 1.1343127489089966, + "learning_rate": 3.9636622373457836e-05, + "loss": 0.3444, + "step": 41221 + }, + { + "epoch": 2.8604538199986123, + "grad_norm": 1.0138472318649292, + "learning_rate": 3.9632142961657834e-05, + "loss": 0.2634, + "step": 41222 + }, + { + "epoch": 2.8605232114357086, + "grad_norm": 1.0708084106445312, + "learning_rate": 3.9627663740432196e-05, + "loss": 0.4714, + "step": 41223 + }, + { + "epoch": 2.8605926028728055, + "grad_norm": 0.9729138016700745, + "learning_rate": 3.962318470979498e-05, + "loss": 0.3459, + "step": 41224 + }, + { + "epoch": 2.8606619943099023, + "grad_norm": 0.8977912664413452, + "learning_rate": 3.9618705869760374e-05, + "loss": 0.3125, + "step": 41225 + }, + { + "epoch": 2.8607313857469987, + "grad_norm": 1.1785789728164673, + "learning_rate": 3.96142272203425e-05, + "loss": 0.3, + "step": 41226 + }, + { + "epoch": 2.8608007771840955, + "grad_norm": 1.5051032304763794, + "learning_rate": 3.960974876155551e-05, + "loss": 0.4152, + "step": 41227 + }, + { + "epoch": 2.860870168621192, + "grad_norm": 1.3761895895004272, + "learning_rate": 3.960527049341353e-05, + "loss": 0.4703, + "step": 41228 + }, + { + "epoch": 2.8609395600582888, + "grad_norm": 0.9600945115089417, + "learning_rate": 3.960079241593071e-05, + "loss": 0.2782, + "step": 41229 + }, + { + "epoch": 2.8610089514953856, + "grad_norm": 1.2493360042572021, + "learning_rate": 3.959631452912121e-05, + "loss": 0.351, + "step": 41230 + }, + { + "epoch": 2.861078342932482, + "grad_norm": 0.9874134063720703, + "learning_rate": 3.959183683299911e-05, + "loss": 0.2407, + "step": 41231 + }, + { + "epoch": 2.861147734369579, + "grad_norm": 1.384412169456482, + "learning_rate": 3.958735932757857e-05, + "loss": 0.3668, + "step": 41232 + }, + { + "epoch": 2.861217125806675, + "grad_norm": 1.3529928922653198, + "learning_rate": 3.958288201287372e-05, + "loss": 0.5921, + "step": 41233 + }, + { + "epoch": 2.861286517243772, + "grad_norm": 1.0526564121246338, + "learning_rate": 3.95784048888987e-05, + "loss": 0.2897, + "step": 41234 + }, + { + "epoch": 2.861355908680869, + "grad_norm": 0.8866806626319885, + "learning_rate": 3.957392795566766e-05, + "loss": 0.3708, + "step": 41235 + }, + { + "epoch": 2.8614253001179657, + "grad_norm": 1.0899310111999512, + "learning_rate": 3.956945121319471e-05, + "loss": 0.3447, + "step": 41236 + }, + { + "epoch": 2.861494691555062, + "grad_norm": 1.058812141418457, + "learning_rate": 3.9564974661494026e-05, + "loss": 0.3091, + "step": 41237 + }, + { + "epoch": 2.861564082992159, + "grad_norm": 2.4847469329833984, + "learning_rate": 3.956049830057967e-05, + "loss": 0.6727, + "step": 41238 + }, + { + "epoch": 2.8616334744292553, + "grad_norm": 1.431623101234436, + "learning_rate": 3.955602213046581e-05, + "loss": 0.5616, + "step": 41239 + }, + { + "epoch": 2.861702865866352, + "grad_norm": 1.499023675918579, + "learning_rate": 3.955154615116661e-05, + "loss": 0.4446, + "step": 41240 + }, + { + "epoch": 2.861772257303449, + "grad_norm": 1.1651650667190552, + "learning_rate": 3.954707036269614e-05, + "loss": 0.4035, + "step": 41241 + }, + { + "epoch": 2.8618416487405454, + "grad_norm": 1.4931765794754028, + "learning_rate": 3.9542594765068544e-05, + "loss": 0.5441, + "step": 41242 + }, + { + "epoch": 2.861911040177642, + "grad_norm": 1.6060152053833008, + "learning_rate": 3.953811935829795e-05, + "loss": 0.3906, + "step": 41243 + }, + { + "epoch": 2.8619804316147386, + "grad_norm": 1.1935232877731323, + "learning_rate": 3.953364414239855e-05, + "loss": 0.3555, + "step": 41244 + }, + { + "epoch": 2.8620498230518354, + "grad_norm": 0.9356208443641663, + "learning_rate": 3.952916911738439e-05, + "loss": 0.3409, + "step": 41245 + }, + { + "epoch": 2.8621192144889323, + "grad_norm": 1.0302650928497314, + "learning_rate": 3.952469428326961e-05, + "loss": 0.286, + "step": 41246 + }, + { + "epoch": 2.8621886059260286, + "grad_norm": 1.725224494934082, + "learning_rate": 3.952021964006837e-05, + "loss": 0.3778, + "step": 41247 + }, + { + "epoch": 2.8622579973631255, + "grad_norm": 0.8105183243751526, + "learning_rate": 3.9515745187794775e-05, + "loss": 0.304, + "step": 41248 + }, + { + "epoch": 2.862327388800222, + "grad_norm": 1.2656774520874023, + "learning_rate": 3.9511270926462954e-05, + "loss": 0.4601, + "step": 41249 + }, + { + "epoch": 2.8623967802373187, + "grad_norm": 1.210240364074707, + "learning_rate": 3.9506796856087026e-05, + "loss": 0.3103, + "step": 41250 + }, + { + "epoch": 2.8624661716744155, + "grad_norm": 1.8053803443908691, + "learning_rate": 3.950232297668116e-05, + "loss": 0.481, + "step": 41251 + }, + { + "epoch": 2.862535563111512, + "grad_norm": 1.236262559890747, + "learning_rate": 3.9497849288259414e-05, + "loss": 0.4519, + "step": 41252 + }, + { + "epoch": 2.8626049545486087, + "grad_norm": 0.8795818090438843, + "learning_rate": 3.949337579083593e-05, + "loss": 0.4906, + "step": 41253 + }, + { + "epoch": 2.862674345985705, + "grad_norm": 1.1951544284820557, + "learning_rate": 3.948890248442484e-05, + "loss": 0.4032, + "step": 41254 + }, + { + "epoch": 2.862743737422802, + "grad_norm": 1.179526686668396, + "learning_rate": 3.9484429369040265e-05, + "loss": 0.4486, + "step": 41255 + }, + { + "epoch": 2.862813128859899, + "grad_norm": 1.4206360578536987, + "learning_rate": 3.947995644469633e-05, + "loss": 0.3646, + "step": 41256 + }, + { + "epoch": 2.862882520296995, + "grad_norm": 1.2644423246383667, + "learning_rate": 3.947548371140715e-05, + "loss": 0.421, + "step": 41257 + }, + { + "epoch": 2.862951911734092, + "grad_norm": 1.1698338985443115, + "learning_rate": 3.9471011169186875e-05, + "loss": 0.3684, + "step": 41258 + }, + { + "epoch": 2.8630213031711884, + "grad_norm": 0.9022922515869141, + "learning_rate": 3.9466538818049556e-05, + "loss": 0.3655, + "step": 41259 + }, + { + "epoch": 2.8630906946082852, + "grad_norm": 1.1047375202178955, + "learning_rate": 3.946206665800937e-05, + "loss": 0.4377, + "step": 41260 + }, + { + "epoch": 2.863160086045382, + "grad_norm": 1.335349678993225, + "learning_rate": 3.94575946890804e-05, + "loss": 0.4119, + "step": 41261 + }, + { + "epoch": 2.863229477482479, + "grad_norm": 1.1011743545532227, + "learning_rate": 3.945312291127679e-05, + "loss": 0.2796, + "step": 41262 + }, + { + "epoch": 2.8632988689195753, + "grad_norm": 1.0362505912780762, + "learning_rate": 3.944865132461263e-05, + "loss": 0.309, + "step": 41263 + }, + { + "epoch": 2.863368260356672, + "grad_norm": 1.1734329462051392, + "learning_rate": 3.9444179929102064e-05, + "loss": 0.3731, + "step": 41264 + }, + { + "epoch": 2.8634376517937685, + "grad_norm": 1.37583327293396, + "learning_rate": 3.9439708724759225e-05, + "loss": 0.4096, + "step": 41265 + }, + { + "epoch": 2.8635070432308654, + "grad_norm": 0.9248060584068298, + "learning_rate": 3.943523771159817e-05, + "loss": 0.2279, + "step": 41266 + }, + { + "epoch": 2.863576434667962, + "grad_norm": 1.5373553037643433, + "learning_rate": 3.943076688963304e-05, + "loss": 0.3809, + "step": 41267 + }, + { + "epoch": 2.8636458261050586, + "grad_norm": 1.4289844036102295, + "learning_rate": 3.942629625887795e-05, + "loss": 0.4713, + "step": 41268 + }, + { + "epoch": 2.8637152175421554, + "grad_norm": 1.061647653579712, + "learning_rate": 3.9421825819347016e-05, + "loss": 0.2203, + "step": 41269 + }, + { + "epoch": 2.863784608979252, + "grad_norm": 1.3378328084945679, + "learning_rate": 3.941735557105435e-05, + "loss": 0.319, + "step": 41270 + }, + { + "epoch": 2.8638540004163486, + "grad_norm": 1.131081223487854, + "learning_rate": 3.94128855140141e-05, + "loss": 0.3386, + "step": 41271 + }, + { + "epoch": 2.8639233918534455, + "grad_norm": 1.6857643127441406, + "learning_rate": 3.9408415648240306e-05, + "loss": 0.3401, + "step": 41272 + }, + { + "epoch": 2.863992783290542, + "grad_norm": 1.5525120496749878, + "learning_rate": 3.940394597374711e-05, + "loss": 0.4659, + "step": 41273 + }, + { + "epoch": 2.8640621747276387, + "grad_norm": 1.8542286157608032, + "learning_rate": 3.939947649054864e-05, + "loss": 0.5365, + "step": 41274 + }, + { + "epoch": 2.864131566164735, + "grad_norm": 2.033975124359131, + "learning_rate": 3.9395007198659005e-05, + "loss": 0.3856, + "step": 41275 + }, + { + "epoch": 2.864200957601832, + "grad_norm": 1.152925968170166, + "learning_rate": 3.9390538098092276e-05, + "loss": 0.3837, + "step": 41276 + }, + { + "epoch": 2.8642703490389287, + "grad_norm": 1.0132167339324951, + "learning_rate": 3.93860691888626e-05, + "loss": 0.2932, + "step": 41277 + }, + { + "epoch": 2.864339740476025, + "grad_norm": 1.8974334001541138, + "learning_rate": 3.938160047098405e-05, + "loss": 0.5629, + "step": 41278 + }, + { + "epoch": 2.864409131913122, + "grad_norm": 1.094854474067688, + "learning_rate": 3.9377131944470804e-05, + "loss": 0.4358, + "step": 41279 + }, + { + "epoch": 2.8644785233502184, + "grad_norm": 1.179438591003418, + "learning_rate": 3.937266360933687e-05, + "loss": 0.4579, + "step": 41280 + }, + { + "epoch": 2.864547914787315, + "grad_norm": 1.1005792617797852, + "learning_rate": 3.936819546559641e-05, + "loss": 0.3415, + "step": 41281 + }, + { + "epoch": 2.864617306224412, + "grad_norm": 1.0840035676956177, + "learning_rate": 3.936372751326353e-05, + "loss": 0.3363, + "step": 41282 + }, + { + "epoch": 2.8646866976615084, + "grad_norm": 1.1762919425964355, + "learning_rate": 3.935925975235232e-05, + "loss": 0.4658, + "step": 41283 + }, + { + "epoch": 2.8647560890986052, + "grad_norm": 0.9859733581542969, + "learning_rate": 3.935479218287689e-05, + "loss": 0.4015, + "step": 41284 + }, + { + "epoch": 2.8648254805357016, + "grad_norm": 1.4945675134658813, + "learning_rate": 3.935032480485134e-05, + "loss": 0.2828, + "step": 41285 + }, + { + "epoch": 2.8648948719727985, + "grad_norm": 1.1107795238494873, + "learning_rate": 3.934585761828982e-05, + "loss": 0.393, + "step": 41286 + }, + { + "epoch": 2.8649642634098953, + "grad_norm": 0.9819322824478149, + "learning_rate": 3.9341390623206356e-05, + "loss": 0.2317, + "step": 41287 + }, + { + "epoch": 2.865033654846992, + "grad_norm": 0.993096649646759, + "learning_rate": 3.9336923819615076e-05, + "loss": 0.2964, + "step": 41288 + }, + { + "epoch": 2.8651030462840885, + "grad_norm": 1.503212332725525, + "learning_rate": 3.933245720753009e-05, + "loss": 0.4433, + "step": 41289 + }, + { + "epoch": 2.8651724377211854, + "grad_norm": 1.3069961071014404, + "learning_rate": 3.932799078696551e-05, + "loss": 0.3904, + "step": 41290 + }, + { + "epoch": 2.8652418291582817, + "grad_norm": 0.9521887302398682, + "learning_rate": 3.932352455793542e-05, + "loss": 0.2525, + "step": 41291 + }, + { + "epoch": 2.8653112205953786, + "grad_norm": 1.2166035175323486, + "learning_rate": 3.931905852045395e-05, + "loss": 0.4216, + "step": 41292 + }, + { + "epoch": 2.8653806120324754, + "grad_norm": 1.2892045974731445, + "learning_rate": 3.9314592674535144e-05, + "loss": 0.4502, + "step": 41293 + }, + { + "epoch": 2.865450003469572, + "grad_norm": 1.071054458618164, + "learning_rate": 3.931012702019313e-05, + "loss": 0.381, + "step": 41294 + }, + { + "epoch": 2.8655193949066686, + "grad_norm": 1.1909481287002563, + "learning_rate": 3.9305661557442e-05, + "loss": 0.4335, + "step": 41295 + }, + { + "epoch": 2.865588786343765, + "grad_norm": 1.094687819480896, + "learning_rate": 3.930119628629585e-05, + "loss": 0.4623, + "step": 41296 + }, + { + "epoch": 2.865658177780862, + "grad_norm": 1.345294713973999, + "learning_rate": 3.929673120676879e-05, + "loss": 0.2733, + "step": 41297 + }, + { + "epoch": 2.8657275692179587, + "grad_norm": 1.2208086252212524, + "learning_rate": 3.92922663188749e-05, + "loss": 0.6093, + "step": 41298 + }, + { + "epoch": 2.865796960655055, + "grad_norm": 1.232427716255188, + "learning_rate": 3.9287801622628315e-05, + "loss": 0.4128, + "step": 41299 + }, + { + "epoch": 2.865866352092152, + "grad_norm": 1.4666482210159302, + "learning_rate": 3.928333711804307e-05, + "loss": 0.4311, + "step": 41300 + }, + { + "epoch": 2.8659357435292483, + "grad_norm": 0.9400672316551208, + "learning_rate": 3.927887280513328e-05, + "loss": 0.263, + "step": 41301 + }, + { + "epoch": 2.866005134966345, + "grad_norm": 0.9888506531715393, + "learning_rate": 3.927440868391304e-05, + "loss": 0.3625, + "step": 41302 + }, + { + "epoch": 2.866074526403442, + "grad_norm": 1.2368698120117188, + "learning_rate": 3.926994475439646e-05, + "loss": 0.3898, + "step": 41303 + }, + { + "epoch": 2.8661439178405383, + "grad_norm": 1.5631420612335205, + "learning_rate": 3.9265481016597604e-05, + "loss": 0.3778, + "step": 41304 + }, + { + "epoch": 2.866213309277635, + "grad_norm": 1.2322072982788086, + "learning_rate": 3.926101747053058e-05, + "loss": 0.3848, + "step": 41305 + }, + { + "epoch": 2.8662827007147316, + "grad_norm": 1.1170148849487305, + "learning_rate": 3.925655411620952e-05, + "loss": 0.3769, + "step": 41306 + }, + { + "epoch": 2.8663520921518284, + "grad_norm": 1.2765129804611206, + "learning_rate": 3.925209095364844e-05, + "loss": 0.3409, + "step": 41307 + }, + { + "epoch": 2.8664214835889252, + "grad_norm": 1.1669625043869019, + "learning_rate": 3.9247627982861444e-05, + "loss": 0.3458, + "step": 41308 + }, + { + "epoch": 2.8664908750260216, + "grad_norm": 0.8589138984680176, + "learning_rate": 3.9243165203862644e-05, + "loss": 0.4196, + "step": 41309 + }, + { + "epoch": 2.8665602664631185, + "grad_norm": 1.3196650743484497, + "learning_rate": 3.9238702616666155e-05, + "loss": 0.4704, + "step": 41310 + }, + { + "epoch": 2.866629657900215, + "grad_norm": 0.9540936350822449, + "learning_rate": 3.9234240221286e-05, + "loss": 0.4018, + "step": 41311 + }, + { + "epoch": 2.8666990493373117, + "grad_norm": 1.1452587842941284, + "learning_rate": 3.922977801773631e-05, + "loss": 0.3526, + "step": 41312 + }, + { + "epoch": 2.8667684407744085, + "grad_norm": 1.4436638355255127, + "learning_rate": 3.9225316006031175e-05, + "loss": 0.3662, + "step": 41313 + }, + { + "epoch": 2.8668378322115053, + "grad_norm": 0.7739512920379639, + "learning_rate": 3.9220854186184655e-05, + "loss": 0.23, + "step": 41314 + }, + { + "epoch": 2.8669072236486017, + "grad_norm": 1.485654592514038, + "learning_rate": 3.9216392558210824e-05, + "loss": 0.4172, + "step": 41315 + }, + { + "epoch": 2.8669766150856986, + "grad_norm": 1.3720179796218872, + "learning_rate": 3.92119311221238e-05, + "loss": 0.3793, + "step": 41316 + }, + { + "epoch": 2.867046006522795, + "grad_norm": 1.5896254777908325, + "learning_rate": 3.920746987793765e-05, + "loss": 0.4664, + "step": 41317 + }, + { + "epoch": 2.867115397959892, + "grad_norm": 1.295837163925171, + "learning_rate": 3.920300882566648e-05, + "loss": 0.3187, + "step": 41318 + }, + { + "epoch": 2.8671847893969886, + "grad_norm": 1.0542182922363281, + "learning_rate": 3.919854796532434e-05, + "loss": 0.3297, + "step": 41319 + }, + { + "epoch": 2.867254180834085, + "grad_norm": 1.1821767091751099, + "learning_rate": 3.9194087296925376e-05, + "loss": 0.5165, + "step": 41320 + }, + { + "epoch": 2.867323572271182, + "grad_norm": 0.9665867686271667, + "learning_rate": 3.918962682048358e-05, + "loss": 0.4967, + "step": 41321 + }, + { + "epoch": 2.8673929637082782, + "grad_norm": 1.1251047849655151, + "learning_rate": 3.918516653601309e-05, + "loss": 0.4627, + "step": 41322 + }, + { + "epoch": 2.867462355145375, + "grad_norm": 1.2397657632827759, + "learning_rate": 3.918070644352796e-05, + "loss": 0.4085, + "step": 41323 + }, + { + "epoch": 2.867531746582472, + "grad_norm": 1.5617473125457764, + "learning_rate": 3.917624654304228e-05, + "loss": 0.5335, + "step": 41324 + }, + { + "epoch": 2.8676011380195683, + "grad_norm": 1.1081860065460205, + "learning_rate": 3.917178683457014e-05, + "loss": 0.3283, + "step": 41325 + }, + { + "epoch": 2.867670529456665, + "grad_norm": 1.5601904392242432, + "learning_rate": 3.916732731812561e-05, + "loss": 0.5731, + "step": 41326 + }, + { + "epoch": 2.8677399208937615, + "grad_norm": 1.3767789602279663, + "learning_rate": 3.9162867993722805e-05, + "loss": 0.3685, + "step": 41327 + }, + { + "epoch": 2.8678093123308583, + "grad_norm": 1.5039290189743042, + "learning_rate": 3.915840886137574e-05, + "loss": 0.294, + "step": 41328 + }, + { + "epoch": 2.867878703767955, + "grad_norm": 0.8506431579589844, + "learning_rate": 3.915394992109852e-05, + "loss": 0.2759, + "step": 41329 + }, + { + "epoch": 2.8679480952050516, + "grad_norm": 2.12203311920166, + "learning_rate": 3.9149491172905205e-05, + "loss": 0.4729, + "step": 41330 + }, + { + "epoch": 2.8680174866421484, + "grad_norm": 1.1239112615585327, + "learning_rate": 3.914503261680989e-05, + "loss": 0.3541, + "step": 41331 + }, + { + "epoch": 2.868086878079245, + "grad_norm": 1.3801308870315552, + "learning_rate": 3.914057425282667e-05, + "loss": 0.4763, + "step": 41332 + }, + { + "epoch": 2.8681562695163416, + "grad_norm": 1.6048094034194946, + "learning_rate": 3.913611608096956e-05, + "loss": 0.3996, + "step": 41333 + }, + { + "epoch": 2.8682256609534384, + "grad_norm": 1.166809320449829, + "learning_rate": 3.9131658101252735e-05, + "loss": 0.4484, + "step": 41334 + }, + { + "epoch": 2.8682950523905353, + "grad_norm": 1.0689971446990967, + "learning_rate": 3.912720031369016e-05, + "loss": 0.3385, + "step": 41335 + }, + { + "epoch": 2.8683644438276317, + "grad_norm": 1.23665189743042, + "learning_rate": 3.9122742718295946e-05, + "loss": 0.683, + "step": 41336 + }, + { + "epoch": 2.868433835264728, + "grad_norm": 1.2097917795181274, + "learning_rate": 3.911828531508417e-05, + "loss": 0.5495, + "step": 41337 + }, + { + "epoch": 2.868503226701825, + "grad_norm": 1.7775321006774902, + "learning_rate": 3.911382810406891e-05, + "loss": 0.5784, + "step": 41338 + }, + { + "epoch": 2.8685726181389217, + "grad_norm": 0.8884047269821167, + "learning_rate": 3.910937108526422e-05, + "loss": 0.3734, + "step": 41339 + }, + { + "epoch": 2.8686420095760186, + "grad_norm": 0.815230131149292, + "learning_rate": 3.910491425868419e-05, + "loss": 0.2921, + "step": 41340 + }, + { + "epoch": 2.868711401013115, + "grad_norm": 1.2117432355880737, + "learning_rate": 3.9100457624342915e-05, + "loss": 0.4078, + "step": 41341 + }, + { + "epoch": 2.868780792450212, + "grad_norm": 0.8471556901931763, + "learning_rate": 3.909600118225438e-05, + "loss": 0.3131, + "step": 41342 + }, + { + "epoch": 2.868850183887308, + "grad_norm": 1.1281242370605469, + "learning_rate": 3.9091544932432724e-05, + "loss": 0.3758, + "step": 41343 + }, + { + "epoch": 2.868919575324405, + "grad_norm": 1.3648966550827026, + "learning_rate": 3.9087088874891984e-05, + "loss": 0.3763, + "step": 41344 + }, + { + "epoch": 2.868988966761502, + "grad_norm": 1.1525403261184692, + "learning_rate": 3.9082633009646277e-05, + "loss": 0.3796, + "step": 41345 + }, + { + "epoch": 2.8690583581985982, + "grad_norm": 1.0881271362304688, + "learning_rate": 3.907817733670959e-05, + "loss": 0.3803, + "step": 41346 + }, + { + "epoch": 2.869127749635695, + "grad_norm": 1.3557755947113037, + "learning_rate": 3.9073721856096026e-05, + "loss": 0.258, + "step": 41347 + }, + { + "epoch": 2.8691971410727914, + "grad_norm": 1.0753647089004517, + "learning_rate": 3.906926656781969e-05, + "loss": 0.1692, + "step": 41348 + }, + { + "epoch": 2.8692665325098883, + "grad_norm": 1.0194615125656128, + "learning_rate": 3.906481147189458e-05, + "loss": 0.3231, + "step": 41349 + }, + { + "epoch": 2.869335923946985, + "grad_norm": 1.03274667263031, + "learning_rate": 3.906035656833479e-05, + "loss": 0.4485, + "step": 41350 + }, + { + "epoch": 2.8694053153840815, + "grad_norm": 1.3255349397659302, + "learning_rate": 3.905590185715438e-05, + "loss": 0.5827, + "step": 41351 + }, + { + "epoch": 2.8694747068211783, + "grad_norm": 1.5058170557022095, + "learning_rate": 3.905144733836742e-05, + "loss": 0.434, + "step": 41352 + }, + { + "epoch": 2.8695440982582747, + "grad_norm": 1.641282558441162, + "learning_rate": 3.9046993011987964e-05, + "loss": 0.5079, + "step": 41353 + }, + { + "epoch": 2.8696134896953716, + "grad_norm": 0.9740924835205078, + "learning_rate": 3.9042538878030086e-05, + "loss": 0.3774, + "step": 41354 + }, + { + "epoch": 2.8696828811324684, + "grad_norm": 1.428983449935913, + "learning_rate": 3.903808493650788e-05, + "loss": 0.443, + "step": 41355 + }, + { + "epoch": 2.8697522725695648, + "grad_norm": 1.694732427597046, + "learning_rate": 3.9033631187435316e-05, + "loss": 0.4483, + "step": 41356 + }, + { + "epoch": 2.8698216640066616, + "grad_norm": 1.396458387374878, + "learning_rate": 3.902917763082652e-05, + "loss": 0.3387, + "step": 41357 + }, + { + "epoch": 2.869891055443758, + "grad_norm": 0.9272170066833496, + "learning_rate": 3.902472426669553e-05, + "loss": 0.2132, + "step": 41358 + }, + { + "epoch": 2.869960446880855, + "grad_norm": 1.1706751585006714, + "learning_rate": 3.9020271095056406e-05, + "loss": 0.381, + "step": 41359 + }, + { + "epoch": 2.8700298383179517, + "grad_norm": 1.1389918327331543, + "learning_rate": 3.901581811592322e-05, + "loss": 0.3963, + "step": 41360 + }, + { + "epoch": 2.8700992297550485, + "grad_norm": 1.1121208667755127, + "learning_rate": 3.901136532931001e-05, + "loss": 0.3616, + "step": 41361 + }, + { + "epoch": 2.870168621192145, + "grad_norm": 1.1545460224151611, + "learning_rate": 3.9006912735230894e-05, + "loss": 0.3863, + "step": 41362 + }, + { + "epoch": 2.8702380126292413, + "grad_norm": 1.3556100130081177, + "learning_rate": 3.900246033369983e-05, + "loss": 0.4025, + "step": 41363 + }, + { + "epoch": 2.870307404066338, + "grad_norm": 0.8549799919128418, + "learning_rate": 3.899800812473093e-05, + "loss": 0.283, + "step": 41364 + }, + { + "epoch": 2.870376795503435, + "grad_norm": 1.0019766092300415, + "learning_rate": 3.8993556108338236e-05, + "loss": 0.367, + "step": 41365 + }, + { + "epoch": 2.8704461869405318, + "grad_norm": 0.9813392162322998, + "learning_rate": 3.8989104284535816e-05, + "loss": 0.3034, + "step": 41366 + }, + { + "epoch": 2.870515578377628, + "grad_norm": 1.0447516441345215, + "learning_rate": 3.89846526533377e-05, + "loss": 0.3896, + "step": 41367 + }, + { + "epoch": 2.870584969814725, + "grad_norm": 1.1657180786132812, + "learning_rate": 3.898020121475801e-05, + "loss": 0.3819, + "step": 41368 + }, + { + "epoch": 2.8706543612518214, + "grad_norm": 2.162968397140503, + "learning_rate": 3.897574996881069e-05, + "loss": 0.4313, + "step": 41369 + }, + { + "epoch": 2.870723752688918, + "grad_norm": 1.1836128234863281, + "learning_rate": 3.8971298915509854e-05, + "loss": 0.3916, + "step": 41370 + }, + { + "epoch": 2.870793144126015, + "grad_norm": 0.862200915813446, + "learning_rate": 3.896684805486955e-05, + "loss": 0.3322, + "step": 41371 + }, + { + "epoch": 2.8708625355631114, + "grad_norm": 0.8775528073310852, + "learning_rate": 3.8962397386903823e-05, + "loss": 0.379, + "step": 41372 + }, + { + "epoch": 2.8709319270002083, + "grad_norm": 1.0715038776397705, + "learning_rate": 3.895794691162673e-05, + "loss": 0.2429, + "step": 41373 + }, + { + "epoch": 2.8710013184373047, + "grad_norm": 1.2015810012817383, + "learning_rate": 3.895349662905232e-05, + "loss": 0.3554, + "step": 41374 + }, + { + "epoch": 2.8710707098744015, + "grad_norm": 1.3720780611038208, + "learning_rate": 3.894904653919467e-05, + "loss": 0.4194, + "step": 41375 + }, + { + "epoch": 2.8711401013114983, + "grad_norm": 1.0614701509475708, + "learning_rate": 3.8944596642067754e-05, + "loss": 0.3, + "step": 41376 + }, + { + "epoch": 2.8712094927485947, + "grad_norm": 1.2218438386917114, + "learning_rate": 3.8940146937685675e-05, + "loss": 0.3421, + "step": 41377 + }, + { + "epoch": 2.8712788841856915, + "grad_norm": 1.3062834739685059, + "learning_rate": 3.893569742606246e-05, + "loss": 0.5469, + "step": 41378 + }, + { + "epoch": 2.871348275622788, + "grad_norm": 1.3110804557800293, + "learning_rate": 3.893124810721216e-05, + "loss": 0.5469, + "step": 41379 + }, + { + "epoch": 2.8714176670598848, + "grad_norm": 1.1315308809280396, + "learning_rate": 3.892679898114886e-05, + "loss": 0.238, + "step": 41380 + }, + { + "epoch": 2.8714870584969816, + "grad_norm": 1.5116610527038574, + "learning_rate": 3.892235004788655e-05, + "loss": 0.5455, + "step": 41381 + }, + { + "epoch": 2.871556449934078, + "grad_norm": 1.4409780502319336, + "learning_rate": 3.891790130743927e-05, + "loss": 0.3886, + "step": 41382 + }, + { + "epoch": 2.871625841371175, + "grad_norm": 1.9346766471862793, + "learning_rate": 3.891345275982114e-05, + "loss": 0.508, + "step": 41383 + }, + { + "epoch": 2.871695232808271, + "grad_norm": 0.9679474830627441, + "learning_rate": 3.8909004405046114e-05, + "loss": 0.2763, + "step": 41384 + }, + { + "epoch": 2.871764624245368, + "grad_norm": 1.1927975416183472, + "learning_rate": 3.890455624312828e-05, + "loss": 0.3564, + "step": 41385 + }, + { + "epoch": 2.871834015682465, + "grad_norm": 1.390116810798645, + "learning_rate": 3.890010827408166e-05, + "loss": 0.4963, + "step": 41386 + }, + { + "epoch": 2.8719034071195617, + "grad_norm": 1.1627618074417114, + "learning_rate": 3.8895660497920313e-05, + "loss": 0.3884, + "step": 41387 + }, + { + "epoch": 2.871972798556658, + "grad_norm": 1.0233510732650757, + "learning_rate": 3.889121291465828e-05, + "loss": 0.3553, + "step": 41388 + }, + { + "epoch": 2.872042189993755, + "grad_norm": 1.5587389469146729, + "learning_rate": 3.888676552430964e-05, + "loss": 0.4804, + "step": 41389 + }, + { + "epoch": 2.8721115814308513, + "grad_norm": 1.3423328399658203, + "learning_rate": 3.8882318326888345e-05, + "loss": 0.4035, + "step": 41390 + }, + { + "epoch": 2.872180972867948, + "grad_norm": 1.209022045135498, + "learning_rate": 3.8877871322408485e-05, + "loss": 0.4251, + "step": 41391 + }, + { + "epoch": 2.872250364305045, + "grad_norm": 0.8396954536437988, + "learning_rate": 3.887342451088409e-05, + "loss": 0.225, + "step": 41392 + }, + { + "epoch": 2.8723197557421414, + "grad_norm": 0.8788377642631531, + "learning_rate": 3.886897789232921e-05, + "loss": 0.3263, + "step": 41393 + }, + { + "epoch": 2.872389147179238, + "grad_norm": 1.30445396900177, + "learning_rate": 3.886453146675787e-05, + "loss": 0.4, + "step": 41394 + }, + { + "epoch": 2.8724585386163346, + "grad_norm": 0.9988659620285034, + "learning_rate": 3.886008523418411e-05, + "loss": 0.3713, + "step": 41395 + }, + { + "epoch": 2.8725279300534314, + "grad_norm": 0.8926131129264832, + "learning_rate": 3.8855639194621996e-05, + "loss": 0.3179, + "step": 41396 + }, + { + "epoch": 2.8725973214905283, + "grad_norm": 1.0744551420211792, + "learning_rate": 3.885119334808551e-05, + "loss": 0.3983, + "step": 41397 + }, + { + "epoch": 2.8726667129276247, + "grad_norm": 1.340575933456421, + "learning_rate": 3.8846747694588705e-05, + "loss": 0.4089, + "step": 41398 + }, + { + "epoch": 2.8727361043647215, + "grad_norm": 1.2078230381011963, + "learning_rate": 3.884230223414562e-05, + "loss": 0.4782, + "step": 41399 + }, + { + "epoch": 2.872805495801818, + "grad_norm": 1.0704574584960938, + "learning_rate": 3.88378569667703e-05, + "loss": 0.2852, + "step": 41400 + }, + { + "epoch": 2.8728748872389147, + "grad_norm": 0.8547680974006653, + "learning_rate": 3.883341189247677e-05, + "loss": 0.2756, + "step": 41401 + }, + { + "epoch": 2.8729442786760115, + "grad_norm": 1.4211702346801758, + "learning_rate": 3.8828967011279064e-05, + "loss": 0.311, + "step": 41402 + }, + { + "epoch": 2.873013670113108, + "grad_norm": 0.9471937417984009, + "learning_rate": 3.8824522323191236e-05, + "loss": 0.3716, + "step": 41403 + }, + { + "epoch": 2.8730830615502048, + "grad_norm": 1.5557292699813843, + "learning_rate": 3.8820077828227266e-05, + "loss": 0.4211, + "step": 41404 + }, + { + "epoch": 2.873152452987301, + "grad_norm": 1.5073421001434326, + "learning_rate": 3.881563352640121e-05, + "loss": 0.4649, + "step": 41405 + }, + { + "epoch": 2.873221844424398, + "grad_norm": 1.5163073539733887, + "learning_rate": 3.881118941772709e-05, + "loss": 0.2852, + "step": 41406 + }, + { + "epoch": 2.873291235861495, + "grad_norm": 0.8907381892204285, + "learning_rate": 3.880674550221895e-05, + "loss": 0.3172, + "step": 41407 + }, + { + "epoch": 2.873360627298591, + "grad_norm": 0.8956363201141357, + "learning_rate": 3.880230177989082e-05, + "loss": 0.2933, + "step": 41408 + }, + { + "epoch": 2.873430018735688, + "grad_norm": 0.9753234386444092, + "learning_rate": 3.879785825075673e-05, + "loss": 0.2364, + "step": 41409 + }, + { + "epoch": 2.8734994101727844, + "grad_norm": 1.0247727632522583, + "learning_rate": 3.879341491483073e-05, + "loss": 0.3383, + "step": 41410 + }, + { + "epoch": 2.8735688016098813, + "grad_norm": 1.560625433921814, + "learning_rate": 3.878897177212678e-05, + "loss": 0.351, + "step": 41411 + }, + { + "epoch": 2.873638193046978, + "grad_norm": 1.3137794733047485, + "learning_rate": 3.878452882265895e-05, + "loss": 0.4207, + "step": 41412 + }, + { + "epoch": 2.873707584484075, + "grad_norm": 0.974716067314148, + "learning_rate": 3.878008606644126e-05, + "loss": 0.4696, + "step": 41413 + }, + { + "epoch": 2.8737769759211713, + "grad_norm": 1.1675763130187988, + "learning_rate": 3.877564350348777e-05, + "loss": 0.2927, + "step": 41414 + }, + { + "epoch": 2.873846367358268, + "grad_norm": 1.3555629253387451, + "learning_rate": 3.877120113381243e-05, + "loss": 0.6117, + "step": 41415 + }, + { + "epoch": 2.8739157587953645, + "grad_norm": 1.0613043308258057, + "learning_rate": 3.876675895742931e-05, + "loss": 0.3088, + "step": 41416 + }, + { + "epoch": 2.8739851502324614, + "grad_norm": 1.2653241157531738, + "learning_rate": 3.876231697435242e-05, + "loss": 0.5229, + "step": 41417 + }, + { + "epoch": 2.874054541669558, + "grad_norm": 1.0041897296905518, + "learning_rate": 3.8757875184595823e-05, + "loss": 0.2703, + "step": 41418 + }, + { + "epoch": 2.8741239331066546, + "grad_norm": 0.9156665802001953, + "learning_rate": 3.875343358817348e-05, + "loss": 0.2636, + "step": 41419 + }, + { + "epoch": 2.8741933245437514, + "grad_norm": 1.0059514045715332, + "learning_rate": 3.874899218509944e-05, + "loss": 0.5085, + "step": 41420 + }, + { + "epoch": 2.874262715980848, + "grad_norm": 0.9054659008979797, + "learning_rate": 3.874455097538773e-05, + "loss": 0.2535, + "step": 41421 + }, + { + "epoch": 2.8743321074179446, + "grad_norm": 1.3612254858016968, + "learning_rate": 3.874010995905236e-05, + "loss": 0.3058, + "step": 41422 + }, + { + "epoch": 2.8744014988550415, + "grad_norm": 1.326767921447754, + "learning_rate": 3.873566913610736e-05, + "loss": 0.3481, + "step": 41423 + }, + { + "epoch": 2.874470890292138, + "grad_norm": 1.327012062072754, + "learning_rate": 3.8731228506566776e-05, + "loss": 0.3085, + "step": 41424 + }, + { + "epoch": 2.8745402817292347, + "grad_norm": 0.8444441556930542, + "learning_rate": 3.872678807044456e-05, + "loss": 0.2623, + "step": 41425 + }, + { + "epoch": 2.874609673166331, + "grad_norm": 1.3048505783081055, + "learning_rate": 3.872234782775477e-05, + "loss": 0.37, + "step": 41426 + }, + { + "epoch": 2.874679064603428, + "grad_norm": 1.0690162181854248, + "learning_rate": 3.871790777851141e-05, + "loss": 0.3961, + "step": 41427 + }, + { + "epoch": 2.8747484560405248, + "grad_norm": 2.044405698776245, + "learning_rate": 3.8713467922728505e-05, + "loss": 0.4527, + "step": 41428 + }, + { + "epoch": 2.874817847477621, + "grad_norm": 0.9285107254981995, + "learning_rate": 3.870902826042008e-05, + "loss": 0.346, + "step": 41429 + }, + { + "epoch": 2.874887238914718, + "grad_norm": 1.1876434087753296, + "learning_rate": 3.870458879160014e-05, + "loss": 0.308, + "step": 41430 + }, + { + "epoch": 2.8749566303518144, + "grad_norm": 1.363242745399475, + "learning_rate": 3.8700149516282736e-05, + "loss": 0.4469, + "step": 41431 + }, + { + "epoch": 2.875026021788911, + "grad_norm": 1.5473099946975708, + "learning_rate": 3.8695710434481813e-05, + "loss": 0.3769, + "step": 41432 + }, + { + "epoch": 2.875095413226008, + "grad_norm": 1.5304911136627197, + "learning_rate": 3.869127154621142e-05, + "loss": 0.2951, + "step": 41433 + }, + { + "epoch": 2.8751648046631044, + "grad_norm": 0.8884008526802063, + "learning_rate": 3.868683285148555e-05, + "loss": 0.3304, + "step": 41434 + }, + { + "epoch": 2.8752341961002013, + "grad_norm": 1.6707723140716553, + "learning_rate": 3.868239435031826e-05, + "loss": 0.3471, + "step": 41435 + }, + { + "epoch": 2.8753035875372976, + "grad_norm": 0.9491007924079895, + "learning_rate": 3.8677956042723526e-05, + "loss": 0.293, + "step": 41436 + }, + { + "epoch": 2.8753729789743945, + "grad_norm": 0.9292174577713013, + "learning_rate": 3.867351792871537e-05, + "loss": 0.2744, + "step": 41437 + }, + { + "epoch": 2.8754423704114913, + "grad_norm": 1.3262965679168701, + "learning_rate": 3.8669080008307846e-05, + "loss": 0.2667, + "step": 41438 + }, + { + "epoch": 2.875511761848588, + "grad_norm": 1.4005321264266968, + "learning_rate": 3.866464228151489e-05, + "loss": 0.3977, + "step": 41439 + }, + { + "epoch": 2.8755811532856845, + "grad_norm": 1.672297716140747, + "learning_rate": 3.866020474835052e-05, + "loss": 0.4398, + "step": 41440 + }, + { + "epoch": 2.8756505447227814, + "grad_norm": 1.1806223392486572, + "learning_rate": 3.865576740882878e-05, + "loss": 0.4551, + "step": 41441 + }, + { + "epoch": 2.8757199361598778, + "grad_norm": 1.2485934495925903, + "learning_rate": 3.865133026296367e-05, + "loss": 0.4266, + "step": 41442 + }, + { + "epoch": 2.8757893275969746, + "grad_norm": 0.8178853988647461, + "learning_rate": 3.8646893310769184e-05, + "loss": 0.3589, + "step": 41443 + }, + { + "epoch": 2.8758587190340714, + "grad_norm": 1.498160481452942, + "learning_rate": 3.8642456552259343e-05, + "loss": 0.4679, + "step": 41444 + }, + { + "epoch": 2.875928110471168, + "grad_norm": 0.9018086194992065, + "learning_rate": 3.8638019987448184e-05, + "loss": 0.2868, + "step": 41445 + }, + { + "epoch": 2.8759975019082646, + "grad_norm": 1.1349953413009644, + "learning_rate": 3.863358361634963e-05, + "loss": 0.4645, + "step": 41446 + }, + { + "epoch": 2.876066893345361, + "grad_norm": 0.942380964756012, + "learning_rate": 3.862914743897774e-05, + "loss": 0.2539, + "step": 41447 + }, + { + "epoch": 2.876136284782458, + "grad_norm": 0.989401638507843, + "learning_rate": 3.862471145534652e-05, + "loss": 0.3411, + "step": 41448 + }, + { + "epoch": 2.8762056762195547, + "grad_norm": 0.9035803079605103, + "learning_rate": 3.862027566547e-05, + "loss": 0.3404, + "step": 41449 + }, + { + "epoch": 2.876275067656651, + "grad_norm": 1.7859492301940918, + "learning_rate": 3.86158400693621e-05, + "loss": 0.6036, + "step": 41450 + }, + { + "epoch": 2.876344459093748, + "grad_norm": 1.1389024257659912, + "learning_rate": 3.861140466703688e-05, + "loss": 0.4565, + "step": 41451 + }, + { + "epoch": 2.8764138505308443, + "grad_norm": 1.1617425680160522, + "learning_rate": 3.8606969458508326e-05, + "loss": 0.4819, + "step": 41452 + }, + { + "epoch": 2.876483241967941, + "grad_norm": 1.1292517185211182, + "learning_rate": 3.8602534443790487e-05, + "loss": 0.4592, + "step": 41453 + }, + { + "epoch": 2.876552633405038, + "grad_norm": 1.550342082977295, + "learning_rate": 3.859809962289728e-05, + "loss": 0.4397, + "step": 41454 + }, + { + "epoch": 2.8766220248421344, + "grad_norm": 1.2665648460388184, + "learning_rate": 3.859366499584276e-05, + "loss": 0.3198, + "step": 41455 + }, + { + "epoch": 2.876691416279231, + "grad_norm": 1.2891680002212524, + "learning_rate": 3.858923056264091e-05, + "loss": 0.3401, + "step": 41456 + }, + { + "epoch": 2.8767608077163276, + "grad_norm": 1.3590468168258667, + "learning_rate": 3.8584796323305726e-05, + "loss": 0.4829, + "step": 41457 + }, + { + "epoch": 2.8768301991534244, + "grad_norm": 1.251814842224121, + "learning_rate": 3.8580362277851266e-05, + "loss": 0.6016, + "step": 41458 + }, + { + "epoch": 2.8768995905905212, + "grad_norm": 0.9832074642181396, + "learning_rate": 3.857592842629143e-05, + "loss": 0.3266, + "step": 41459 + }, + { + "epoch": 2.8769689820276176, + "grad_norm": 0.9528534412384033, + "learning_rate": 3.8571494768640257e-05, + "loss": 0.4008, + "step": 41460 + }, + { + "epoch": 2.8770383734647145, + "grad_norm": 1.13552725315094, + "learning_rate": 3.856706130491177e-05, + "loss": 0.3697, + "step": 41461 + }, + { + "epoch": 2.877107764901811, + "grad_norm": 1.357168436050415, + "learning_rate": 3.8562628035119916e-05, + "loss": 0.3773, + "step": 41462 + }, + { + "epoch": 2.8771771563389077, + "grad_norm": 1.9634822607040405, + "learning_rate": 3.8558194959278734e-05, + "loss": 0.5199, + "step": 41463 + }, + { + "epoch": 2.8772465477760045, + "grad_norm": 1.1300700902938843, + "learning_rate": 3.85537620774022e-05, + "loss": 0.4177, + "step": 41464 + }, + { + "epoch": 2.8773159392131014, + "grad_norm": 1.0339014530181885, + "learning_rate": 3.854932938950434e-05, + "loss": 0.4136, + "step": 41465 + }, + { + "epoch": 2.8773853306501977, + "grad_norm": 1.2730764150619507, + "learning_rate": 3.854489689559908e-05, + "loss": 0.3821, + "step": 41466 + }, + { + "epoch": 2.8774547220872946, + "grad_norm": 2.1427674293518066, + "learning_rate": 3.854046459570047e-05, + "loss": 0.5605, + "step": 41467 + }, + { + "epoch": 2.877524113524391, + "grad_norm": 1.9040669202804565, + "learning_rate": 3.853603248982246e-05, + "loss": 0.4186, + "step": 41468 + }, + { + "epoch": 2.877593504961488, + "grad_norm": 0.8584247827529907, + "learning_rate": 3.853160057797908e-05, + "loss": 0.234, + "step": 41469 + }, + { + "epoch": 2.8776628963985846, + "grad_norm": 1.376307487487793, + "learning_rate": 3.852716886018429e-05, + "loss": 0.327, + "step": 41470 + }, + { + "epoch": 2.877732287835681, + "grad_norm": 1.0001310110092163, + "learning_rate": 3.852273733645211e-05, + "loss": 0.2905, + "step": 41471 + }, + { + "epoch": 2.877801679272778, + "grad_norm": 1.0044643878936768, + "learning_rate": 3.8518306006796535e-05, + "loss": 0.3467, + "step": 41472 + }, + { + "epoch": 2.8778710707098742, + "grad_norm": 1.39860999584198, + "learning_rate": 3.851387487123151e-05, + "loss": 0.4984, + "step": 41473 + }, + { + "epoch": 2.877940462146971, + "grad_norm": 0.9457232356071472, + "learning_rate": 3.850944392977105e-05, + "loss": 0.3651, + "step": 41474 + }, + { + "epoch": 2.878009853584068, + "grad_norm": 1.068923830986023, + "learning_rate": 3.850501318242914e-05, + "loss": 0.3759, + "step": 41475 + }, + { + "epoch": 2.8780792450211643, + "grad_norm": 0.8864417672157288, + "learning_rate": 3.850058262921976e-05, + "loss": 0.3689, + "step": 41476 + }, + { + "epoch": 2.878148636458261, + "grad_norm": 1.3013595342636108, + "learning_rate": 3.849615227015692e-05, + "loss": 0.4332, + "step": 41477 + }, + { + "epoch": 2.8782180278953575, + "grad_norm": 1.4345701932907104, + "learning_rate": 3.849172210525458e-05, + "loss": 0.4482, + "step": 41478 + }, + { + "epoch": 2.8782874193324544, + "grad_norm": 0.8309757113456726, + "learning_rate": 3.8487292134526774e-05, + "loss": 0.4177, + "step": 41479 + }, + { + "epoch": 2.878356810769551, + "grad_norm": 1.1233971118927002, + "learning_rate": 3.848286235798742e-05, + "loss": 0.3204, + "step": 41480 + }, + { + "epoch": 2.8784262022066476, + "grad_norm": 1.502603530883789, + "learning_rate": 3.847843277565053e-05, + "loss": 0.3045, + "step": 41481 + }, + { + "epoch": 2.8784955936437444, + "grad_norm": 1.3376126289367676, + "learning_rate": 3.847400338753009e-05, + "loss": 0.4135, + "step": 41482 + }, + { + "epoch": 2.878564985080841, + "grad_norm": 1.0628093481063843, + "learning_rate": 3.846957419364008e-05, + "loss": 0.307, + "step": 41483 + }, + { + "epoch": 2.8786343765179376, + "grad_norm": 0.9392099380493164, + "learning_rate": 3.846514519399452e-05, + "loss": 0.321, + "step": 41484 + }, + { + "epoch": 2.8787037679550345, + "grad_norm": 1.3890968561172485, + "learning_rate": 3.846071638860733e-05, + "loss": 0.3933, + "step": 41485 + }, + { + "epoch": 2.878773159392131, + "grad_norm": 1.0485985279083252, + "learning_rate": 3.8456287777492516e-05, + "loss": 0.3959, + "step": 41486 + }, + { + "epoch": 2.8788425508292277, + "grad_norm": 1.907641053199768, + "learning_rate": 3.845185936066405e-05, + "loss": 0.4364, + "step": 41487 + }, + { + "epoch": 2.878911942266324, + "grad_norm": 1.2730177640914917, + "learning_rate": 3.844743113813597e-05, + "loss": 0.2605, + "step": 41488 + }, + { + "epoch": 2.878981333703421, + "grad_norm": 1.3036035299301147, + "learning_rate": 3.844300310992217e-05, + "loss": 0.413, + "step": 41489 + }, + { + "epoch": 2.8790507251405177, + "grad_norm": 1.0493731498718262, + "learning_rate": 3.8438575276036674e-05, + "loss": 0.3328, + "step": 41490 + }, + { + "epoch": 2.8791201165776146, + "grad_norm": 1.6363285779953003, + "learning_rate": 3.843414763649344e-05, + "loss": 0.5681, + "step": 41491 + }, + { + "epoch": 2.879189508014711, + "grad_norm": 1.5273853540420532, + "learning_rate": 3.842972019130646e-05, + "loss": 0.5908, + "step": 41492 + }, + { + "epoch": 2.879258899451808, + "grad_norm": 0.8666058778762817, + "learning_rate": 3.8425292940489756e-05, + "loss": 0.3339, + "step": 41493 + }, + { + "epoch": 2.879328290888904, + "grad_norm": 1.1277753114700317, + "learning_rate": 3.842086588405722e-05, + "loss": 0.435, + "step": 41494 + }, + { + "epoch": 2.879397682326001, + "grad_norm": 1.304896593093872, + "learning_rate": 3.8416439022022864e-05, + "loss": 0.4196, + "step": 41495 + }, + { + "epoch": 2.879467073763098, + "grad_norm": 0.9989566206932068, + "learning_rate": 3.841201235440066e-05, + "loss": 0.3724, + "step": 41496 + }, + { + "epoch": 2.8795364652001942, + "grad_norm": 1.1619857549667358, + "learning_rate": 3.840758588120459e-05, + "loss": 0.24, + "step": 41497 + }, + { + "epoch": 2.879605856637291, + "grad_norm": 1.215820074081421, + "learning_rate": 3.840315960244863e-05, + "loss": 0.4084, + "step": 41498 + }, + { + "epoch": 2.8796752480743875, + "grad_norm": 1.5545157194137573, + "learning_rate": 3.839873351814674e-05, + "loss": 0.4654, + "step": 41499 + }, + { + "epoch": 2.8797446395114843, + "grad_norm": 1.025069236755371, + "learning_rate": 3.839430762831295e-05, + "loss": 0.3925, + "step": 41500 + }, + { + "epoch": 2.879814030948581, + "grad_norm": 1.4963281154632568, + "learning_rate": 3.838988193296114e-05, + "loss": 0.3411, + "step": 41501 + }, + { + "epoch": 2.8798834223856775, + "grad_norm": 1.4310520887374878, + "learning_rate": 3.838545643210533e-05, + "loss": 0.3668, + "step": 41502 + }, + { + "epoch": 2.8799528138227743, + "grad_norm": 2.4989311695098877, + "learning_rate": 3.838103112575947e-05, + "loss": 0.5427, + "step": 41503 + }, + { + "epoch": 2.8800222052598707, + "grad_norm": 1.116854190826416, + "learning_rate": 3.8376606013937566e-05, + "loss": 0.3926, + "step": 41504 + }, + { + "epoch": 2.8800915966969676, + "grad_norm": 1.3355447053909302, + "learning_rate": 3.837218109665355e-05, + "loss": 0.3935, + "step": 41505 + }, + { + "epoch": 2.8801609881340644, + "grad_norm": 1.4220738410949707, + "learning_rate": 3.8367756373921416e-05, + "loss": 0.3786, + "step": 41506 + }, + { + "epoch": 2.880230379571161, + "grad_norm": 1.041373372077942, + "learning_rate": 3.836333184575516e-05, + "loss": 0.2638, + "step": 41507 + }, + { + "epoch": 2.8802997710082576, + "grad_norm": 1.3799645900726318, + "learning_rate": 3.835890751216867e-05, + "loss": 0.3751, + "step": 41508 + }, + { + "epoch": 2.880369162445354, + "grad_norm": 1.112324833869934, + "learning_rate": 3.835448337317598e-05, + "loss": 0.3359, + "step": 41509 + }, + { + "epoch": 2.880438553882451, + "grad_norm": 0.9620168209075928, + "learning_rate": 3.835005942879101e-05, + "loss": 0.2954, + "step": 41510 + }, + { + "epoch": 2.8805079453195477, + "grad_norm": 1.1672563552856445, + "learning_rate": 3.834563567902776e-05, + "loss": 0.6328, + "step": 41511 + }, + { + "epoch": 2.880577336756644, + "grad_norm": 0.791133463382721, + "learning_rate": 3.834121212390018e-05, + "loss": 0.2744, + "step": 41512 + }, + { + "epoch": 2.880646728193741, + "grad_norm": 1.1140838861465454, + "learning_rate": 3.8336788763422246e-05, + "loss": 0.2891, + "step": 41513 + }, + { + "epoch": 2.8807161196308373, + "grad_norm": 0.8492300510406494, + "learning_rate": 3.8332365597607945e-05, + "loss": 0.3719, + "step": 41514 + }, + { + "epoch": 2.880785511067934, + "grad_norm": 1.1839711666107178, + "learning_rate": 3.832794262647118e-05, + "loss": 0.4307, + "step": 41515 + }, + { + "epoch": 2.880854902505031, + "grad_norm": 1.0061177015304565, + "learning_rate": 3.832351985002595e-05, + "loss": 0.3401, + "step": 41516 + }, + { + "epoch": 2.880924293942128, + "grad_norm": 1.0395615100860596, + "learning_rate": 3.831909726828621e-05, + "loss": 0.3861, + "step": 41517 + }, + { + "epoch": 2.880993685379224, + "grad_norm": 1.3090333938598633, + "learning_rate": 3.831467488126591e-05, + "loss": 0.4692, + "step": 41518 + }, + { + "epoch": 2.881063076816321, + "grad_norm": 1.4383546113967896, + "learning_rate": 3.831025268897907e-05, + "loss": 0.4504, + "step": 41519 + }, + { + "epoch": 2.8811324682534174, + "grad_norm": 1.1589807271957397, + "learning_rate": 3.830583069143957e-05, + "loss": 0.2788, + "step": 41520 + }, + { + "epoch": 2.8812018596905142, + "grad_norm": 1.5751838684082031, + "learning_rate": 3.8301408888661405e-05, + "loss": 0.5419, + "step": 41521 + }, + { + "epoch": 2.881271251127611, + "grad_norm": 1.585972785949707, + "learning_rate": 3.829698728065857e-05, + "loss": 0.6039, + "step": 41522 + }, + { + "epoch": 2.8813406425647075, + "grad_norm": 1.4725466966629028, + "learning_rate": 3.829256586744494e-05, + "loss": 0.5361, + "step": 41523 + }, + { + "epoch": 2.8814100340018043, + "grad_norm": 1.2904525995254517, + "learning_rate": 3.828814464903454e-05, + "loss": 0.2773, + "step": 41524 + }, + { + "epoch": 2.8814794254389007, + "grad_norm": 1.083870768547058, + "learning_rate": 3.828372362544129e-05, + "loss": 0.2968, + "step": 41525 + }, + { + "epoch": 2.8815488168759975, + "grad_norm": 1.0764634609222412, + "learning_rate": 3.827930279667917e-05, + "loss": 0.3004, + "step": 41526 + }, + { + "epoch": 2.8816182083130943, + "grad_norm": 1.3151077032089233, + "learning_rate": 3.827488216276217e-05, + "loss": 0.4329, + "step": 41527 + }, + { + "epoch": 2.8816875997501907, + "grad_norm": 1.2018861770629883, + "learning_rate": 3.827046172370416e-05, + "loss": 0.3473, + "step": 41528 + }, + { + "epoch": 2.8817569911872876, + "grad_norm": 1.2595796585083008, + "learning_rate": 3.8266041479519156e-05, + "loss": 0.3395, + "step": 41529 + }, + { + "epoch": 2.881826382624384, + "grad_norm": 1.235376238822937, + "learning_rate": 3.826162143022108e-05, + "loss": 0.4574, + "step": 41530 + }, + { + "epoch": 2.881895774061481, + "grad_norm": 1.085328459739685, + "learning_rate": 3.825720157582392e-05, + "loss": 0.2722, + "step": 41531 + }, + { + "epoch": 2.8819651654985776, + "grad_norm": 1.016046166419983, + "learning_rate": 3.8252781916341596e-05, + "loss": 0.3453, + "step": 41532 + }, + { + "epoch": 2.882034556935674, + "grad_norm": 1.132536768913269, + "learning_rate": 3.824836245178809e-05, + "loss": 0.3426, + "step": 41533 + }, + { + "epoch": 2.882103948372771, + "grad_norm": 1.222449541091919, + "learning_rate": 3.824394318217736e-05, + "loss": 0.3609, + "step": 41534 + }, + { + "epoch": 2.8821733398098672, + "grad_norm": 1.0735588073730469, + "learning_rate": 3.823952410752331e-05, + "loss": 0.2948, + "step": 41535 + }, + { + "epoch": 2.882242731246964, + "grad_norm": 1.8543065786361694, + "learning_rate": 3.8235105227839906e-05, + "loss": 0.6621, + "step": 41536 + }, + { + "epoch": 2.882312122684061, + "grad_norm": 1.6911001205444336, + "learning_rate": 3.8230686543141115e-05, + "loss": 0.4258, + "step": 41537 + }, + { + "epoch": 2.8823815141211573, + "grad_norm": 1.8353080749511719, + "learning_rate": 3.8226268053440883e-05, + "loss": 0.4735, + "step": 41538 + }, + { + "epoch": 2.882450905558254, + "grad_norm": 1.47085702419281, + "learning_rate": 3.8221849758753155e-05, + "loss": 0.2402, + "step": 41539 + }, + { + "epoch": 2.8825202969953505, + "grad_norm": 1.1495709419250488, + "learning_rate": 3.821743165909188e-05, + "loss": 0.3023, + "step": 41540 + }, + { + "epoch": 2.8825896884324473, + "grad_norm": 0.8830721378326416, + "learning_rate": 3.821301375447104e-05, + "loss": 0.3124, + "step": 41541 + }, + { + "epoch": 2.882659079869544, + "grad_norm": 1.3437118530273438, + "learning_rate": 3.820859604490451e-05, + "loss": 0.5293, + "step": 41542 + }, + { + "epoch": 2.882728471306641, + "grad_norm": 1.5444281101226807, + "learning_rate": 3.8204178530406274e-05, + "loss": 0.5508, + "step": 41543 + }, + { + "epoch": 2.8827978627437374, + "grad_norm": 1.2332851886749268, + "learning_rate": 3.819976121099027e-05, + "loss": 0.4443, + "step": 41544 + }, + { + "epoch": 2.8828672541808342, + "grad_norm": 0.7701936960220337, + "learning_rate": 3.819534408667046e-05, + "loss": 0.1883, + "step": 41545 + }, + { + "epoch": 2.8829366456179306, + "grad_norm": 1.1703143119812012, + "learning_rate": 3.8190927157460776e-05, + "loss": 0.399, + "step": 41546 + }, + { + "epoch": 2.8830060370550274, + "grad_norm": 1.2468101978302002, + "learning_rate": 3.818651042337515e-05, + "loss": 0.3335, + "step": 41547 + }, + { + "epoch": 2.8830754284921243, + "grad_norm": 1.446810007095337, + "learning_rate": 3.81820938844276e-05, + "loss": 0.4114, + "step": 41548 + }, + { + "epoch": 2.8831448199292207, + "grad_norm": 1.3984782695770264, + "learning_rate": 3.8177677540631954e-05, + "loss": 0.4027, + "step": 41549 + }, + { + "epoch": 2.8832142113663175, + "grad_norm": 1.2153241634368896, + "learning_rate": 3.8173261392002216e-05, + "loss": 0.3768, + "step": 41550 + }, + { + "epoch": 2.883283602803414, + "grad_norm": 0.9634305834770203, + "learning_rate": 3.8168845438552314e-05, + "loss": 0.3923, + "step": 41551 + }, + { + "epoch": 2.8833529942405107, + "grad_norm": 1.4190088510513306, + "learning_rate": 3.816442968029619e-05, + "loss": 0.4305, + "step": 41552 + }, + { + "epoch": 2.8834223856776076, + "grad_norm": 1.152977466583252, + "learning_rate": 3.816001411724779e-05, + "loss": 0.3086, + "step": 41553 + }, + { + "epoch": 2.883491777114704, + "grad_norm": 1.9037374258041382, + "learning_rate": 3.8155598749421086e-05, + "loss": 0.4887, + "step": 41554 + }, + { + "epoch": 2.883561168551801, + "grad_norm": 1.1576783657073975, + "learning_rate": 3.815118357682994e-05, + "loss": 0.4247, + "step": 41555 + }, + { + "epoch": 2.883630559988897, + "grad_norm": 1.1163311004638672, + "learning_rate": 3.8146768599488346e-05, + "loss": 0.3509, + "step": 41556 + }, + { + "epoch": 2.883699951425994, + "grad_norm": 1.1925699710845947, + "learning_rate": 3.814235381741025e-05, + "loss": 0.5122, + "step": 41557 + }, + { + "epoch": 2.883769342863091, + "grad_norm": 1.080609917640686, + "learning_rate": 3.8137939230609544e-05, + "loss": 0.5121, + "step": 41558 + }, + { + "epoch": 2.8838387343001872, + "grad_norm": 0.943261444568634, + "learning_rate": 3.8133524839100174e-05, + "loss": 0.1835, + "step": 41559 + }, + { + "epoch": 2.883908125737284, + "grad_norm": 0.9744963645935059, + "learning_rate": 3.812911064289609e-05, + "loss": 0.3354, + "step": 41560 + }, + { + "epoch": 2.8839775171743804, + "grad_norm": 1.7069001197814941, + "learning_rate": 3.8124696642011225e-05, + "loss": 0.4462, + "step": 41561 + }, + { + "epoch": 2.8840469086114773, + "grad_norm": 0.894138514995575, + "learning_rate": 3.8120282836459556e-05, + "loss": 0.3673, + "step": 41562 + }, + { + "epoch": 2.884116300048574, + "grad_norm": 1.194506287574768, + "learning_rate": 3.811586922625493e-05, + "loss": 0.4527, + "step": 41563 + }, + { + "epoch": 2.884185691485671, + "grad_norm": 1.1936030387878418, + "learning_rate": 3.8111455811411335e-05, + "loss": 0.2459, + "step": 41564 + }, + { + "epoch": 2.8842550829227673, + "grad_norm": 1.232234239578247, + "learning_rate": 3.810704259194269e-05, + "loss": 0.3688, + "step": 41565 + }, + { + "epoch": 2.8843244743598637, + "grad_norm": 1.0877586603164673, + "learning_rate": 3.8102629567862925e-05, + "loss": 0.3559, + "step": 41566 + }, + { + "epoch": 2.8843938657969606, + "grad_norm": 0.9806371331214905, + "learning_rate": 3.8098216739185986e-05, + "loss": 0.2157, + "step": 41567 + }, + { + "epoch": 2.8844632572340574, + "grad_norm": 0.6894857883453369, + "learning_rate": 3.809380410592579e-05, + "loss": 0.2925, + "step": 41568 + }, + { + "epoch": 2.884532648671154, + "grad_norm": 1.1683398485183716, + "learning_rate": 3.8089391668096306e-05, + "loss": 0.386, + "step": 41569 + }, + { + "epoch": 2.8846020401082506, + "grad_norm": 1.1589796543121338, + "learning_rate": 3.8084979425711395e-05, + "loss": 0.3566, + "step": 41570 + }, + { + "epoch": 2.8846714315453474, + "grad_norm": 1.6320449113845825, + "learning_rate": 3.808056737878503e-05, + "loss": 0.5234, + "step": 41571 + }, + { + "epoch": 2.884740822982444, + "grad_norm": 0.9682841897010803, + "learning_rate": 3.807615552733111e-05, + "loss": 0.2938, + "step": 41572 + }, + { + "epoch": 2.8848102144195407, + "grad_norm": 1.0562855005264282, + "learning_rate": 3.80717438713636e-05, + "loss": 0.3617, + "step": 41573 + }, + { + "epoch": 2.8848796058566375, + "grad_norm": 1.2864689826965332, + "learning_rate": 3.80673324108964e-05, + "loss": 0.3841, + "step": 41574 + }, + { + "epoch": 2.884948997293734, + "grad_norm": 2.391216993331909, + "learning_rate": 3.8062921145943453e-05, + "loss": 0.4462, + "step": 41575 + }, + { + "epoch": 2.8850183887308307, + "grad_norm": 1.3284196853637695, + "learning_rate": 3.80585100765187e-05, + "loss": 0.3583, + "step": 41576 + }, + { + "epoch": 2.885087780167927, + "grad_norm": 0.865479588508606, + "learning_rate": 3.805409920263603e-05, + "loss": 0.2317, + "step": 41577 + }, + { + "epoch": 2.885157171605024, + "grad_norm": 1.507010817527771, + "learning_rate": 3.8049688524309366e-05, + "loss": 0.2385, + "step": 41578 + }, + { + "epoch": 2.8852265630421208, + "grad_norm": 1.4327706098556519, + "learning_rate": 3.804527804155266e-05, + "loss": 0.4573, + "step": 41579 + }, + { + "epoch": 2.885295954479217, + "grad_norm": 1.1157169342041016, + "learning_rate": 3.804086775437981e-05, + "loss": 0.3904, + "step": 41580 + }, + { + "epoch": 2.885365345916314, + "grad_norm": 1.414603590965271, + "learning_rate": 3.803645766280476e-05, + "loss": 0.5005, + "step": 41581 + }, + { + "epoch": 2.8854347373534104, + "grad_norm": 1.7660655975341797, + "learning_rate": 3.803204776684142e-05, + "loss": 0.409, + "step": 41582 + }, + { + "epoch": 2.885504128790507, + "grad_norm": 1.3010629415512085, + "learning_rate": 3.802763806650375e-05, + "loss": 0.4306, + "step": 41583 + }, + { + "epoch": 2.885573520227604, + "grad_norm": 1.181929588317871, + "learning_rate": 3.80232285618056e-05, + "loss": 0.351, + "step": 41584 + }, + { + "epoch": 2.8856429116647004, + "grad_norm": 1.7006422281265259, + "learning_rate": 3.801881925276093e-05, + "loss": 0.5808, + "step": 41585 + }, + { + "epoch": 2.8857123031017973, + "grad_norm": 1.101462483406067, + "learning_rate": 3.801441013938366e-05, + "loss": 0.2607, + "step": 41586 + }, + { + "epoch": 2.8857816945388937, + "grad_norm": 1.2252196073532104, + "learning_rate": 3.80100012216877e-05, + "loss": 0.4618, + "step": 41587 + }, + { + "epoch": 2.8858510859759905, + "grad_norm": 1.3168095350265503, + "learning_rate": 3.800559249968697e-05, + "loss": 0.3648, + "step": 41588 + }, + { + "epoch": 2.8859204774130873, + "grad_norm": 1.5347094535827637, + "learning_rate": 3.8001183973395435e-05, + "loss": 0.3616, + "step": 41589 + }, + { + "epoch": 2.885989868850184, + "grad_norm": 2.3147125244140625, + "learning_rate": 3.799677564282693e-05, + "loss": 0.4289, + "step": 41590 + }, + { + "epoch": 2.8860592602872805, + "grad_norm": 1.1055622100830078, + "learning_rate": 3.799236750799542e-05, + "loss": 0.3568, + "step": 41591 + }, + { + "epoch": 2.886128651724377, + "grad_norm": 1.316808819770813, + "learning_rate": 3.798795956891483e-05, + "loss": 0.3378, + "step": 41592 + }, + { + "epoch": 2.8861980431614738, + "grad_norm": 0.9565632939338684, + "learning_rate": 3.798355182559904e-05, + "loss": 0.2517, + "step": 41593 + }, + { + "epoch": 2.8862674345985706, + "grad_norm": 1.2404268980026245, + "learning_rate": 3.7979144278061974e-05, + "loss": 0.5676, + "step": 41594 + }, + { + "epoch": 2.8863368260356674, + "grad_norm": 1.33457612991333, + "learning_rate": 3.797473692631755e-05, + "loss": 0.4571, + "step": 41595 + }, + { + "epoch": 2.886406217472764, + "grad_norm": 1.0716301202774048, + "learning_rate": 3.797032977037969e-05, + "loss": 0.2881, + "step": 41596 + }, + { + "epoch": 2.8864756089098607, + "grad_norm": 0.9870481491088867, + "learning_rate": 3.7965922810262334e-05, + "loss": 0.4045, + "step": 41597 + }, + { + "epoch": 2.886545000346957, + "grad_norm": 1.019803524017334, + "learning_rate": 3.7961516045979316e-05, + "loss": 0.2676, + "step": 41598 + }, + { + "epoch": 2.886614391784054, + "grad_norm": 1.3000849485397339, + "learning_rate": 3.795710947754462e-05, + "loss": 0.3851, + "step": 41599 + }, + { + "epoch": 2.8866837832211507, + "grad_norm": 1.0536167621612549, + "learning_rate": 3.795270310497211e-05, + "loss": 0.4061, + "step": 41600 + }, + { + "epoch": 2.886753174658247, + "grad_norm": 1.026812195777893, + "learning_rate": 3.7948296928275726e-05, + "loss": 0.2176, + "step": 41601 + }, + { + "epoch": 2.886822566095344, + "grad_norm": 1.8300470113754272, + "learning_rate": 3.7943890947469364e-05, + "loss": 0.445, + "step": 41602 + }, + { + "epoch": 2.8868919575324403, + "grad_norm": 1.046987533569336, + "learning_rate": 3.793948516256698e-05, + "loss": 0.4194, + "step": 41603 + }, + { + "epoch": 2.886961348969537, + "grad_norm": 1.3024650812149048, + "learning_rate": 3.79350795735824e-05, + "loss": 0.4007, + "step": 41604 + }, + { + "epoch": 2.887030740406634, + "grad_norm": 0.9144369959831238, + "learning_rate": 3.793067418052958e-05, + "loss": 0.3491, + "step": 41605 + }, + { + "epoch": 2.8871001318437304, + "grad_norm": 0.897286057472229, + "learning_rate": 3.792626898342242e-05, + "loss": 0.2811, + "step": 41606 + }, + { + "epoch": 2.887169523280827, + "grad_norm": 1.3138328790664673, + "learning_rate": 3.792186398227482e-05, + "loss": 0.5232, + "step": 41607 + }, + { + "epoch": 2.8872389147179236, + "grad_norm": 1.4337129592895508, + "learning_rate": 3.79174591771007e-05, + "loss": 0.457, + "step": 41608 + }, + { + "epoch": 2.8873083061550204, + "grad_norm": 1.2590157985687256, + "learning_rate": 3.791305456791395e-05, + "loss": 0.3154, + "step": 41609 + }, + { + "epoch": 2.8873776975921173, + "grad_norm": 0.9286985397338867, + "learning_rate": 3.790865015472852e-05, + "loss": 0.2898, + "step": 41610 + }, + { + "epoch": 2.8874470890292137, + "grad_norm": 0.982742190361023, + "learning_rate": 3.790424593755825e-05, + "loss": 0.3065, + "step": 41611 + }, + { + "epoch": 2.8875164804663105, + "grad_norm": 1.0670675039291382, + "learning_rate": 3.7899841916417066e-05, + "loss": 0.3068, + "step": 41612 + }, + { + "epoch": 2.887585871903407, + "grad_norm": 0.8737162947654724, + "learning_rate": 3.789543809131888e-05, + "loss": 0.3546, + "step": 41613 + }, + { + "epoch": 2.8876552633405037, + "grad_norm": 0.9087973237037659, + "learning_rate": 3.789103446227759e-05, + "loss": 0.3223, + "step": 41614 + }, + { + "epoch": 2.8877246547776005, + "grad_norm": 0.9171520471572876, + "learning_rate": 3.7886631029307085e-05, + "loss": 0.2908, + "step": 41615 + }, + { + "epoch": 2.8877940462146974, + "grad_norm": 1.2787350416183472, + "learning_rate": 3.78822277924213e-05, + "loss": 0.3515, + "step": 41616 + }, + { + "epoch": 2.8878634376517938, + "grad_norm": 1.2439889907836914, + "learning_rate": 3.787782475163414e-05, + "loss": 0.4595, + "step": 41617 + }, + { + "epoch": 2.8879328290888906, + "grad_norm": 1.0442633628845215, + "learning_rate": 3.787342190695944e-05, + "loss": 0.3247, + "step": 41618 + }, + { + "epoch": 2.888002220525987, + "grad_norm": 1.1474637985229492, + "learning_rate": 3.786901925841114e-05, + "loss": 0.3469, + "step": 41619 + }, + { + "epoch": 2.888071611963084, + "grad_norm": 1.208778738975525, + "learning_rate": 3.786461680600315e-05, + "loss": 0.3359, + "step": 41620 + }, + { + "epoch": 2.8881410034001807, + "grad_norm": 0.8631961345672607, + "learning_rate": 3.786021454974934e-05, + "loss": 0.3172, + "step": 41621 + }, + { + "epoch": 2.888210394837277, + "grad_norm": 1.1164703369140625, + "learning_rate": 3.785581248966363e-05, + "loss": 0.2876, + "step": 41622 + }, + { + "epoch": 2.888279786274374, + "grad_norm": 1.388283133506775, + "learning_rate": 3.7851410625759944e-05, + "loss": 0.5125, + "step": 41623 + }, + { + "epoch": 2.8883491777114703, + "grad_norm": 0.8694639205932617, + "learning_rate": 3.784700895805211e-05, + "loss": 0.3371, + "step": 41624 + }, + { + "epoch": 2.888418569148567, + "grad_norm": 1.3318336009979248, + "learning_rate": 3.784260748655406e-05, + "loss": 0.5149, + "step": 41625 + }, + { + "epoch": 2.888487960585664, + "grad_norm": 1.5466502904891968, + "learning_rate": 3.783820621127968e-05, + "loss": 0.43, + "step": 41626 + }, + { + "epoch": 2.8885573520227603, + "grad_norm": 0.8153098821640015, + "learning_rate": 3.7833805132242914e-05, + "loss": 0.2591, + "step": 41627 + }, + { + "epoch": 2.888626743459857, + "grad_norm": 1.1291087865829468, + "learning_rate": 3.782940424945758e-05, + "loss": 0.3214, + "step": 41628 + }, + { + "epoch": 2.8886961348969535, + "grad_norm": 1.818401575088501, + "learning_rate": 3.782500356293759e-05, + "loss": 0.476, + "step": 41629 + }, + { + "epoch": 2.8887655263340504, + "grad_norm": 1.0273964405059814, + "learning_rate": 3.782060307269686e-05, + "loss": 0.3543, + "step": 41630 + }, + { + "epoch": 2.888834917771147, + "grad_norm": 1.0904297828674316, + "learning_rate": 3.781620277874931e-05, + "loss": 0.2726, + "step": 41631 + }, + { + "epoch": 2.8889043092082436, + "grad_norm": 1.2083585262298584, + "learning_rate": 3.781180268110874e-05, + "loss": 0.324, + "step": 41632 + }, + { + "epoch": 2.8889737006453404, + "grad_norm": 1.3473143577575684, + "learning_rate": 3.7807402779789114e-05, + "loss": 0.3236, + "step": 41633 + }, + { + "epoch": 2.889043092082437, + "grad_norm": 1.3917672634124756, + "learning_rate": 3.78030030748043e-05, + "loss": 0.4103, + "step": 41634 + }, + { + "epoch": 2.8891124835195336, + "grad_norm": 0.728371262550354, + "learning_rate": 3.779860356616818e-05, + "loss": 0.2464, + "step": 41635 + }, + { + "epoch": 2.8891818749566305, + "grad_norm": 0.884477436542511, + "learning_rate": 3.779420425389465e-05, + "loss": 0.2555, + "step": 41636 + }, + { + "epoch": 2.889251266393727, + "grad_norm": 1.3513305187225342, + "learning_rate": 3.7789805137997614e-05, + "loss": 0.2769, + "step": 41637 + }, + { + "epoch": 2.8893206578308237, + "grad_norm": 1.3905375003814697, + "learning_rate": 3.778540621849097e-05, + "loss": 0.4599, + "step": 41638 + }, + { + "epoch": 2.88939004926792, + "grad_norm": 1.4878840446472168, + "learning_rate": 3.778100749538854e-05, + "loss": 0.4798, + "step": 41639 + }, + { + "epoch": 2.889459440705017, + "grad_norm": 0.9777414202690125, + "learning_rate": 3.7776608968704265e-05, + "loss": 0.2388, + "step": 41640 + }, + { + "epoch": 2.8895288321421138, + "grad_norm": 1.4090564250946045, + "learning_rate": 3.777221063845201e-05, + "loss": 0.4369, + "step": 41641 + }, + { + "epoch": 2.8895982235792106, + "grad_norm": 1.1340765953063965, + "learning_rate": 3.776781250464566e-05, + "loss": 0.2685, + "step": 41642 + }, + { + "epoch": 2.889667615016307, + "grad_norm": 1.0563539266586304, + "learning_rate": 3.7763414567299114e-05, + "loss": 0.2711, + "step": 41643 + }, + { + "epoch": 2.889737006453404, + "grad_norm": 0.9194267392158508, + "learning_rate": 3.775901682642624e-05, + "loss": 0.3199, + "step": 41644 + }, + { + "epoch": 2.8898063978905, + "grad_norm": 1.2502694129943848, + "learning_rate": 3.775461928204097e-05, + "loss": 0.4101, + "step": 41645 + }, + { + "epoch": 2.889875789327597, + "grad_norm": 1.6936637163162231, + "learning_rate": 3.7750221934157103e-05, + "loss": 0.3839, + "step": 41646 + }, + { + "epoch": 2.889945180764694, + "grad_norm": 1.0806899070739746, + "learning_rate": 3.7745824782788574e-05, + "loss": 0.3577, + "step": 41647 + }, + { + "epoch": 2.8900145722017903, + "grad_norm": 1.2500723600387573, + "learning_rate": 3.774142782794925e-05, + "loss": 0.3029, + "step": 41648 + }, + { + "epoch": 2.890083963638887, + "grad_norm": 0.9820776581764221, + "learning_rate": 3.7737031069653e-05, + "loss": 0.2845, + "step": 41649 + }, + { + "epoch": 2.8901533550759835, + "grad_norm": 1.2872421741485596, + "learning_rate": 3.773263450791373e-05, + "loss": 0.5179, + "step": 41650 + }, + { + "epoch": 2.8902227465130803, + "grad_norm": 1.1693135499954224, + "learning_rate": 3.772823814274532e-05, + "loss": 0.4486, + "step": 41651 + }, + { + "epoch": 2.890292137950177, + "grad_norm": 1.5727852582931519, + "learning_rate": 3.7723841974161654e-05, + "loss": 0.4871, + "step": 41652 + }, + { + "epoch": 2.8903615293872735, + "grad_norm": 1.3272918462753296, + "learning_rate": 3.771944600217657e-05, + "loss": 0.2713, + "step": 41653 + }, + { + "epoch": 2.8904309208243704, + "grad_norm": 1.0236992835998535, + "learning_rate": 3.771505022680397e-05, + "loss": 0.3108, + "step": 41654 + }, + { + "epoch": 2.8905003122614668, + "grad_norm": 1.0310930013656616, + "learning_rate": 3.771065464805772e-05, + "loss": 0.3241, + "step": 41655 + }, + { + "epoch": 2.8905697036985636, + "grad_norm": 1.1359519958496094, + "learning_rate": 3.770625926595171e-05, + "loss": 0.4119, + "step": 41656 + }, + { + "epoch": 2.8906390951356604, + "grad_norm": 1.0157356262207031, + "learning_rate": 3.7701864080499816e-05, + "loss": 0.4233, + "step": 41657 + }, + { + "epoch": 2.890708486572757, + "grad_norm": 1.5544861555099487, + "learning_rate": 3.769746909171594e-05, + "loss": 0.3861, + "step": 41658 + }, + { + "epoch": 2.8907778780098536, + "grad_norm": 1.4209671020507812, + "learning_rate": 3.769307429961389e-05, + "loss": 0.2744, + "step": 41659 + }, + { + "epoch": 2.89084726944695, + "grad_norm": 1.348474144935608, + "learning_rate": 3.7688679704207586e-05, + "loss": 0.4182, + "step": 41660 + }, + { + "epoch": 2.890916660884047, + "grad_norm": 1.261252522468567, + "learning_rate": 3.768428530551089e-05, + "loss": 0.4291, + "step": 41661 + }, + { + "epoch": 2.8909860523211437, + "grad_norm": 1.3581405878067017, + "learning_rate": 3.76798911035377e-05, + "loss": 0.3903, + "step": 41662 + }, + { + "epoch": 2.89105544375824, + "grad_norm": 1.129470944404602, + "learning_rate": 3.767549709830184e-05, + "loss": 0.4131, + "step": 41663 + }, + { + "epoch": 2.891124835195337, + "grad_norm": 1.3036086559295654, + "learning_rate": 3.767110328981719e-05, + "loss": 0.5558, + "step": 41664 + }, + { + "epoch": 2.8911942266324333, + "grad_norm": 0.9853824377059937, + "learning_rate": 3.766670967809765e-05, + "loss": 0.3253, + "step": 41665 + }, + { + "epoch": 2.89126361806953, + "grad_norm": 1.2723275423049927, + "learning_rate": 3.7662316263157095e-05, + "loss": 0.4519, + "step": 41666 + }, + { + "epoch": 2.891333009506627, + "grad_norm": 1.4215456247329712, + "learning_rate": 3.7657923045009346e-05, + "loss": 0.3408, + "step": 41667 + }, + { + "epoch": 2.891402400943724, + "grad_norm": 1.3471477031707764, + "learning_rate": 3.7653530023668306e-05, + "loss": 0.3855, + "step": 41668 + }, + { + "epoch": 2.89147179238082, + "grad_norm": 1.188416600227356, + "learning_rate": 3.7649137199147834e-05, + "loss": 0.4081, + "step": 41669 + }, + { + "epoch": 2.891541183817917, + "grad_norm": 1.2627118825912476, + "learning_rate": 3.76447445714618e-05, + "loss": 0.4075, + "step": 41670 + }, + { + "epoch": 2.8916105752550134, + "grad_norm": 0.9767494797706604, + "learning_rate": 3.764035214062408e-05, + "loss": 0.413, + "step": 41671 + }, + { + "epoch": 2.8916799666921102, + "grad_norm": 1.314438819885254, + "learning_rate": 3.763595990664852e-05, + "loss": 0.3117, + "step": 41672 + }, + { + "epoch": 2.891749358129207, + "grad_norm": 1.2014552354812622, + "learning_rate": 3.7631567869549045e-05, + "loss": 0.3512, + "step": 41673 + }, + { + "epoch": 2.8918187495663035, + "grad_norm": 1.0549993515014648, + "learning_rate": 3.762717602933944e-05, + "loss": 0.4201, + "step": 41674 + }, + { + "epoch": 2.8918881410034003, + "grad_norm": 0.8199288249015808, + "learning_rate": 3.76227843860336e-05, + "loss": 0.2382, + "step": 41675 + }, + { + "epoch": 2.8919575324404967, + "grad_norm": 1.143122911453247, + "learning_rate": 3.76183929396454e-05, + "loss": 0.5003, + "step": 41676 + }, + { + "epoch": 2.8920269238775935, + "grad_norm": 0.8605979084968567, + "learning_rate": 3.761400169018868e-05, + "loss": 0.3034, + "step": 41677 + }, + { + "epoch": 2.8920963153146904, + "grad_norm": 1.8920012712478638, + "learning_rate": 3.760961063767733e-05, + "loss": 0.5797, + "step": 41678 + }, + { + "epoch": 2.8921657067517867, + "grad_norm": 1.2650561332702637, + "learning_rate": 3.7605219782125235e-05, + "loss": 0.3617, + "step": 41679 + }, + { + "epoch": 2.8922350981888836, + "grad_norm": 0.9796171188354492, + "learning_rate": 3.760082912354618e-05, + "loss": 0.309, + "step": 41680 + }, + { + "epoch": 2.89230448962598, + "grad_norm": 1.8613898754119873, + "learning_rate": 3.759643866195407e-05, + "loss": 0.4073, + "step": 41681 + }, + { + "epoch": 2.892373881063077, + "grad_norm": 1.086069107055664, + "learning_rate": 3.7592048397362766e-05, + "loss": 0.3979, + "step": 41682 + }, + { + "epoch": 2.8924432725001736, + "grad_norm": 1.0575276613235474, + "learning_rate": 3.7587658329786136e-05, + "loss": 0.368, + "step": 41683 + }, + { + "epoch": 2.89251266393727, + "grad_norm": 1.2547613382339478, + "learning_rate": 3.758326845923801e-05, + "loss": 0.3453, + "step": 41684 + }, + { + "epoch": 2.892582055374367, + "grad_norm": 1.0638046264648438, + "learning_rate": 3.757887878573227e-05, + "loss": 0.285, + "step": 41685 + }, + { + "epoch": 2.8926514468114632, + "grad_norm": 1.6283138990402222, + "learning_rate": 3.75744893092828e-05, + "loss": 0.5581, + "step": 41686 + }, + { + "epoch": 2.89272083824856, + "grad_norm": 1.1065101623535156, + "learning_rate": 3.757010002990339e-05, + "loss": 0.3672, + "step": 41687 + }, + { + "epoch": 2.892790229685657, + "grad_norm": 1.1801607608795166, + "learning_rate": 3.756571094760794e-05, + "loss": 0.3827, + "step": 41688 + }, + { + "epoch": 2.8928596211227533, + "grad_norm": 1.5657025575637817, + "learning_rate": 3.7561322062410285e-05, + "loss": 0.4193, + "step": 41689 + }, + { + "epoch": 2.89292901255985, + "grad_norm": 0.7290158271789551, + "learning_rate": 3.75569333743243e-05, + "loss": 0.2396, + "step": 41690 + }, + { + "epoch": 2.8929984039969465, + "grad_norm": 1.0797821283340454, + "learning_rate": 3.755254488336383e-05, + "loss": 0.3895, + "step": 41691 + }, + { + "epoch": 2.8930677954340434, + "grad_norm": 1.0482503175735474, + "learning_rate": 3.7548156589542724e-05, + "loss": 0.2634, + "step": 41692 + }, + { + "epoch": 2.89313718687114, + "grad_norm": 1.2590453624725342, + "learning_rate": 3.754376849287489e-05, + "loss": 0.4141, + "step": 41693 + }, + { + "epoch": 2.893206578308237, + "grad_norm": 1.6706031560897827, + "learning_rate": 3.7539380593374085e-05, + "loss": 0.3479, + "step": 41694 + }, + { + "epoch": 2.8932759697453334, + "grad_norm": 0.8678137063980103, + "learning_rate": 3.753499289105421e-05, + "loss": 0.2482, + "step": 41695 + }, + { + "epoch": 2.8933453611824302, + "grad_norm": 1.116187572479248, + "learning_rate": 3.753060538592912e-05, + "loss": 0.3286, + "step": 41696 + }, + { + "epoch": 2.8934147526195266, + "grad_norm": 1.7285884618759155, + "learning_rate": 3.7526218078012686e-05, + "loss": 0.4506, + "step": 41697 + }, + { + "epoch": 2.8934841440566235, + "grad_norm": 1.7546864748001099, + "learning_rate": 3.752183096731871e-05, + "loss": 0.3256, + "step": 41698 + }, + { + "epoch": 2.8935535354937203, + "grad_norm": 1.5281338691711426, + "learning_rate": 3.7517444053861065e-05, + "loss": 0.3873, + "step": 41699 + }, + { + "epoch": 2.8936229269308167, + "grad_norm": 1.001823902130127, + "learning_rate": 3.751305733765363e-05, + "loss": 0.3384, + "step": 41700 + }, + { + "epoch": 2.8936923183679135, + "grad_norm": 1.9290261268615723, + "learning_rate": 3.750867081871019e-05, + "loss": 0.3018, + "step": 41701 + }, + { + "epoch": 2.89376170980501, + "grad_norm": 1.414125919342041, + "learning_rate": 3.750428449704463e-05, + "loss": 0.4091, + "step": 41702 + }, + { + "epoch": 2.8938311012421067, + "grad_norm": 1.656597375869751, + "learning_rate": 3.7499898372670784e-05, + "loss": 0.3788, + "step": 41703 + }, + { + "epoch": 2.8939004926792036, + "grad_norm": 1.037987232208252, + "learning_rate": 3.749551244560251e-05, + "loss": 0.3449, + "step": 41704 + }, + { + "epoch": 2.8939698841163, + "grad_norm": 1.5970923900604248, + "learning_rate": 3.7491126715853655e-05, + "loss": 0.3752, + "step": 41705 + }, + { + "epoch": 2.894039275553397, + "grad_norm": 1.969664454460144, + "learning_rate": 3.7486741183438056e-05, + "loss": 0.4447, + "step": 41706 + }, + { + "epoch": 2.894108666990493, + "grad_norm": 1.2797837257385254, + "learning_rate": 3.74823558483696e-05, + "loss": 0.4194, + "step": 41707 + }, + { + "epoch": 2.89417805842759, + "grad_norm": 1.1248416900634766, + "learning_rate": 3.747797071066206e-05, + "loss": 0.6162, + "step": 41708 + }, + { + "epoch": 2.894247449864687, + "grad_norm": 1.283166527748108, + "learning_rate": 3.747358577032931e-05, + "loss": 0.4314, + "step": 41709 + }, + { + "epoch": 2.8943168413017832, + "grad_norm": 1.1967103481292725, + "learning_rate": 3.746920102738519e-05, + "loss": 0.4666, + "step": 41710 + }, + { + "epoch": 2.89438623273888, + "grad_norm": 1.5061309337615967, + "learning_rate": 3.746481648184355e-05, + "loss": 0.4382, + "step": 41711 + }, + { + "epoch": 2.8944556241759765, + "grad_norm": 1.1774295568466187, + "learning_rate": 3.746043213371824e-05, + "loss": 0.3958, + "step": 41712 + }, + { + "epoch": 2.8945250156130733, + "grad_norm": 1.5024956464767456, + "learning_rate": 3.745604798302308e-05, + "loss": 0.5179, + "step": 41713 + }, + { + "epoch": 2.89459440705017, + "grad_norm": 1.3742597103118896, + "learning_rate": 3.745166402977196e-05, + "loss": 0.3447, + "step": 41714 + }, + { + "epoch": 2.8946637984872665, + "grad_norm": 1.950812816619873, + "learning_rate": 3.7447280273978636e-05, + "loss": 0.5954, + "step": 41715 + }, + { + "epoch": 2.8947331899243633, + "grad_norm": 1.7784852981567383, + "learning_rate": 3.744289671565701e-05, + "loss": 0.4802, + "step": 41716 + }, + { + "epoch": 2.8948025813614597, + "grad_norm": 0.8858321309089661, + "learning_rate": 3.743851335482088e-05, + "loss": 0.3953, + "step": 41717 + }, + { + "epoch": 2.8948719727985566, + "grad_norm": 1.3310890197753906, + "learning_rate": 3.743413019148411e-05, + "loss": 0.386, + "step": 41718 + }, + { + "epoch": 2.8949413642356534, + "grad_norm": 0.6292072534561157, + "learning_rate": 3.742974722566054e-05, + "loss": 0.2571, + "step": 41719 + }, + { + "epoch": 2.8950107556727502, + "grad_norm": 0.9068267941474915, + "learning_rate": 3.742536445736399e-05, + "loss": 0.2454, + "step": 41720 + }, + { + "epoch": 2.8950801471098466, + "grad_norm": 1.0903661251068115, + "learning_rate": 3.742098188660834e-05, + "loss": 0.408, + "step": 41721 + }, + { + "epoch": 2.8951495385469435, + "grad_norm": 0.9793239831924438, + "learning_rate": 3.741659951340736e-05, + "loss": 0.3787, + "step": 41722 + }, + { + "epoch": 2.89521892998404, + "grad_norm": 0.8686397671699524, + "learning_rate": 3.7412217337774914e-05, + "loss": 0.3153, + "step": 41723 + }, + { + "epoch": 2.8952883214211367, + "grad_norm": 1.0165956020355225, + "learning_rate": 3.740783535972483e-05, + "loss": 0.3807, + "step": 41724 + }, + { + "epoch": 2.8953577128582335, + "grad_norm": 1.1469141244888306, + "learning_rate": 3.740345357927095e-05, + "loss": 0.4301, + "step": 41725 + }, + { + "epoch": 2.89542710429533, + "grad_norm": 1.5191779136657715, + "learning_rate": 3.73990719964271e-05, + "loss": 0.5529, + "step": 41726 + }, + { + "epoch": 2.8954964957324267, + "grad_norm": 1.2884635925292969, + "learning_rate": 3.7394690611207125e-05, + "loss": 0.2815, + "step": 41727 + }, + { + "epoch": 2.895565887169523, + "grad_norm": 1.2336355447769165, + "learning_rate": 3.739030942362488e-05, + "loss": 0.414, + "step": 41728 + }, + { + "epoch": 2.89563527860662, + "grad_norm": 1.4546998739242554, + "learning_rate": 3.738592843369413e-05, + "loss": 0.4171, + "step": 41729 + }, + { + "epoch": 2.895704670043717, + "grad_norm": 0.7774395942687988, + "learning_rate": 3.738154764142874e-05, + "loss": 0.2905, + "step": 41730 + }, + { + "epoch": 2.895774061480813, + "grad_norm": 0.9767879247665405, + "learning_rate": 3.737716704684254e-05, + "loss": 0.299, + "step": 41731 + }, + { + "epoch": 2.89584345291791, + "grad_norm": 1.6026296615600586, + "learning_rate": 3.7372786649949384e-05, + "loss": 0.4002, + "step": 41732 + }, + { + "epoch": 2.8959128443550064, + "grad_norm": 1.3425480127334595, + "learning_rate": 3.736840645076306e-05, + "loss": 0.2864, + "step": 41733 + }, + { + "epoch": 2.8959822357921032, + "grad_norm": 1.1680136919021606, + "learning_rate": 3.736402644929739e-05, + "loss": 0.2338, + "step": 41734 + }, + { + "epoch": 2.8960516272292, + "grad_norm": 1.1691066026687622, + "learning_rate": 3.735964664556627e-05, + "loss": 0.3302, + "step": 41735 + }, + { + "epoch": 2.8961210186662965, + "grad_norm": 0.8186773657798767, + "learning_rate": 3.735526703958343e-05, + "loss": 0.2492, + "step": 41736 + }, + { + "epoch": 2.8961904101033933, + "grad_norm": 1.2662218809127808, + "learning_rate": 3.735088763136275e-05, + "loss": 0.3417, + "step": 41737 + }, + { + "epoch": 2.8962598015404897, + "grad_norm": 1.530508041381836, + "learning_rate": 3.734650842091805e-05, + "loss": 0.5536, + "step": 41738 + }, + { + "epoch": 2.8963291929775865, + "grad_norm": 1.3541977405548096, + "learning_rate": 3.734212940826315e-05, + "loss": 0.4033, + "step": 41739 + }, + { + "epoch": 2.8963985844146833, + "grad_norm": 1.0896973609924316, + "learning_rate": 3.733775059341188e-05, + "loss": 0.3961, + "step": 41740 + }, + { + "epoch": 2.8964679758517797, + "grad_norm": 0.9221115112304688, + "learning_rate": 3.733337197637806e-05, + "loss": 0.3717, + "step": 41741 + }, + { + "epoch": 2.8965373672888766, + "grad_norm": 1.0240086317062378, + "learning_rate": 3.7328993557175564e-05, + "loss": 0.2682, + "step": 41742 + }, + { + "epoch": 2.896606758725973, + "grad_norm": 1.5086389780044556, + "learning_rate": 3.732461533581812e-05, + "loss": 0.5221, + "step": 41743 + }, + { + "epoch": 2.89667615016307, + "grad_norm": 1.2561155557632446, + "learning_rate": 3.7320237312319584e-05, + "loss": 0.4353, + "step": 41744 + }, + { + "epoch": 2.8967455416001666, + "grad_norm": 1.4676241874694824, + "learning_rate": 3.731585948669378e-05, + "loss": 0.4415, + "step": 41745 + }, + { + "epoch": 2.8968149330372635, + "grad_norm": 1.6072386503219604, + "learning_rate": 3.7311481858954555e-05, + "loss": 0.4661, + "step": 41746 + }, + { + "epoch": 2.89688432447436, + "grad_norm": 0.8195646405220032, + "learning_rate": 3.7307104429115705e-05, + "loss": 0.2739, + "step": 41747 + }, + { + "epoch": 2.8969537159114567, + "grad_norm": 0.9480087757110596, + "learning_rate": 3.7302727197191044e-05, + "loss": 0.3402, + "step": 41748 + }, + { + "epoch": 2.897023107348553, + "grad_norm": 1.34059739112854, + "learning_rate": 3.729835016319444e-05, + "loss": 0.3375, + "step": 41749 + }, + { + "epoch": 2.89709249878565, + "grad_norm": 1.3286716938018799, + "learning_rate": 3.7293973327139634e-05, + "loss": 0.3356, + "step": 41750 + }, + { + "epoch": 2.8971618902227467, + "grad_norm": 1.6379666328430176, + "learning_rate": 3.728959668904047e-05, + "loss": 0.3967, + "step": 41751 + }, + { + "epoch": 2.897231281659843, + "grad_norm": 2.652172088623047, + "learning_rate": 3.728522024891079e-05, + "loss": 0.3314, + "step": 41752 + }, + { + "epoch": 2.89730067309694, + "grad_norm": 1.5366040468215942, + "learning_rate": 3.728084400676438e-05, + "loss": 0.3736, + "step": 41753 + }, + { + "epoch": 2.8973700645340363, + "grad_norm": 0.8572622537612915, + "learning_rate": 3.727646796261508e-05, + "loss": 0.2544, + "step": 41754 + }, + { + "epoch": 2.897439455971133, + "grad_norm": 1.0415611267089844, + "learning_rate": 3.727209211647672e-05, + "loss": 0.3358, + "step": 41755 + }, + { + "epoch": 2.89750884740823, + "grad_norm": 0.8396332859992981, + "learning_rate": 3.726771646836306e-05, + "loss": 0.3191, + "step": 41756 + }, + { + "epoch": 2.8975782388453264, + "grad_norm": 1.1320571899414062, + "learning_rate": 3.7263341018287934e-05, + "loss": 0.4549, + "step": 41757 + }, + { + "epoch": 2.8976476302824232, + "grad_norm": 1.2044686079025269, + "learning_rate": 3.7258965766265165e-05, + "loss": 0.3838, + "step": 41758 + }, + { + "epoch": 2.8977170217195196, + "grad_norm": 1.0067391395568848, + "learning_rate": 3.7254590712308556e-05, + "loss": 0.2466, + "step": 41759 + }, + { + "epoch": 2.8977864131566164, + "grad_norm": 1.251160740852356, + "learning_rate": 3.725021585643192e-05, + "loss": 0.4781, + "step": 41760 + }, + { + "epoch": 2.8978558045937133, + "grad_norm": 1.5803884267807007, + "learning_rate": 3.724584119864909e-05, + "loss": 0.3338, + "step": 41761 + }, + { + "epoch": 2.8979251960308097, + "grad_norm": 1.0219056606292725, + "learning_rate": 3.724146673897387e-05, + "loss": 0.2598, + "step": 41762 + }, + { + "epoch": 2.8979945874679065, + "grad_norm": 5.934032440185547, + "learning_rate": 3.7237092477420046e-05, + "loss": 0.4825, + "step": 41763 + }, + { + "epoch": 2.898063978905003, + "grad_norm": 1.2402458190917969, + "learning_rate": 3.723271841400143e-05, + "loss": 0.4519, + "step": 41764 + }, + { + "epoch": 2.8981333703420997, + "grad_norm": 1.0031516551971436, + "learning_rate": 3.722834454873184e-05, + "loss": 0.2977, + "step": 41765 + }, + { + "epoch": 2.8982027617791966, + "grad_norm": 0.9003226161003113, + "learning_rate": 3.7223970881625115e-05, + "loss": 0.3086, + "step": 41766 + }, + { + "epoch": 2.898272153216293, + "grad_norm": 0.8535423278808594, + "learning_rate": 3.721959741269501e-05, + "loss": 0.2343, + "step": 41767 + }, + { + "epoch": 2.89834154465339, + "grad_norm": 1.0418064594268799, + "learning_rate": 3.721522414195533e-05, + "loss": 0.335, + "step": 41768 + }, + { + "epoch": 2.898410936090486, + "grad_norm": 1.5167678594589233, + "learning_rate": 3.721085106941992e-05, + "loss": 0.5842, + "step": 41769 + }, + { + "epoch": 2.898480327527583, + "grad_norm": 1.5689605474472046, + "learning_rate": 3.720647819510259e-05, + "loss": 0.3198, + "step": 41770 + }, + { + "epoch": 2.89854971896468, + "grad_norm": 1.0678770542144775, + "learning_rate": 3.7202105519017095e-05, + "loss": 0.3189, + "step": 41771 + }, + { + "epoch": 2.8986191104017767, + "grad_norm": 0.7976406216621399, + "learning_rate": 3.7197733041177275e-05, + "loss": 0.2903, + "step": 41772 + }, + { + "epoch": 2.898688501838873, + "grad_norm": 0.9751486778259277, + "learning_rate": 3.719336076159692e-05, + "loss": 0.364, + "step": 41773 + }, + { + "epoch": 2.89875789327597, + "grad_norm": 1.5062034130096436, + "learning_rate": 3.718898868028983e-05, + "loss": 0.3395, + "step": 41774 + }, + { + "epoch": 2.8988272847130663, + "grad_norm": 0.967028796672821, + "learning_rate": 3.718461679726982e-05, + "loss": 0.3102, + "step": 41775 + }, + { + "epoch": 2.898896676150163, + "grad_norm": 1.0374373197555542, + "learning_rate": 3.7180245112550726e-05, + "loss": 0.4067, + "step": 41776 + }, + { + "epoch": 2.89896606758726, + "grad_norm": 0.95042884349823, + "learning_rate": 3.717587362614626e-05, + "loss": 0.3521, + "step": 41777 + }, + { + "epoch": 2.8990354590243563, + "grad_norm": 0.9112287759780884, + "learning_rate": 3.71715023380703e-05, + "loss": 0.3002, + "step": 41778 + }, + { + "epoch": 2.899104850461453, + "grad_norm": 0.8189573287963867, + "learning_rate": 3.716713124833659e-05, + "loss": 0.2272, + "step": 41779 + }, + { + "epoch": 2.8991742418985496, + "grad_norm": 1.3287394046783447, + "learning_rate": 3.716276035695897e-05, + "loss": 0.3377, + "step": 41780 + }, + { + "epoch": 2.8992436333356464, + "grad_norm": 1.4052163362503052, + "learning_rate": 3.7158389663951224e-05, + "loss": 0.4744, + "step": 41781 + }, + { + "epoch": 2.899313024772743, + "grad_norm": 1.3473255634307861, + "learning_rate": 3.715401916932715e-05, + "loss": 0.5139, + "step": 41782 + }, + { + "epoch": 2.8993824162098396, + "grad_norm": 1.1422966718673706, + "learning_rate": 3.714964887310058e-05, + "loss": 0.4859, + "step": 41783 + }, + { + "epoch": 2.8994518076469364, + "grad_norm": 1.0542980432510376, + "learning_rate": 3.714527877528524e-05, + "loss": 0.3106, + "step": 41784 + }, + { + "epoch": 2.899521199084033, + "grad_norm": 1.1525037288665771, + "learning_rate": 3.714090887589496e-05, + "loss": 0.3418, + "step": 41785 + }, + { + "epoch": 2.8995905905211297, + "grad_norm": 1.1723394393920898, + "learning_rate": 3.7136539174943555e-05, + "loss": 0.3051, + "step": 41786 + }, + { + "epoch": 2.8996599819582265, + "grad_norm": 1.1086128950119019, + "learning_rate": 3.7132169672444785e-05, + "loss": 0.3586, + "step": 41787 + }, + { + "epoch": 2.899729373395323, + "grad_norm": 1.2446178197860718, + "learning_rate": 3.712780036841247e-05, + "loss": 0.3332, + "step": 41788 + }, + { + "epoch": 2.8997987648324197, + "grad_norm": 1.5129762887954712, + "learning_rate": 3.712343126286039e-05, + "loss": 0.453, + "step": 41789 + }, + { + "epoch": 2.899868156269516, + "grad_norm": 1.3072054386138916, + "learning_rate": 3.711906235580237e-05, + "loss": 0.3324, + "step": 41790 + }, + { + "epoch": 2.899937547706613, + "grad_norm": 1.4676710367202759, + "learning_rate": 3.711469364725215e-05, + "loss": 0.4301, + "step": 41791 + }, + { + "epoch": 2.9000069391437098, + "grad_norm": 1.107263207435608, + "learning_rate": 3.711032513722354e-05, + "loss": 0.372, + "step": 41792 + }, + { + "epoch": 2.9000763305808066, + "grad_norm": 0.9318835139274597, + "learning_rate": 3.7105956825730346e-05, + "loss": 0.342, + "step": 41793 + }, + { + "epoch": 2.900145722017903, + "grad_norm": 1.3194228410720825, + "learning_rate": 3.710158871278634e-05, + "loss": 0.3889, + "step": 41794 + }, + { + "epoch": 2.9002151134549994, + "grad_norm": 1.2027100324630737, + "learning_rate": 3.709722079840533e-05, + "loss": 0.4161, + "step": 41795 + }, + { + "epoch": 2.900284504892096, + "grad_norm": 1.3870054483413696, + "learning_rate": 3.7092853082601084e-05, + "loss": 0.4004, + "step": 41796 + }, + { + "epoch": 2.900353896329193, + "grad_norm": 1.3020069599151611, + "learning_rate": 3.708848556538745e-05, + "loss": 0.3261, + "step": 41797 + }, + { + "epoch": 2.90042328776629, + "grad_norm": 1.5336434841156006, + "learning_rate": 3.708411824677813e-05, + "loss": 0.344, + "step": 41798 + }, + { + "epoch": 2.9004926792033863, + "grad_norm": 0.9138832092285156, + "learning_rate": 3.7079751126786944e-05, + "loss": 0.2953, + "step": 41799 + }, + { + "epoch": 2.900562070640483, + "grad_norm": 1.03754723072052, + "learning_rate": 3.7075384205427686e-05, + "loss": 0.3222, + "step": 41800 + }, + { + "epoch": 2.9006314620775795, + "grad_norm": 0.983113706111908, + "learning_rate": 3.7071017482714174e-05, + "loss": 0.3348, + "step": 41801 + }, + { + "epoch": 2.9007008535146763, + "grad_norm": 1.0805318355560303, + "learning_rate": 3.706665095866012e-05, + "loss": 0.3216, + "step": 41802 + }, + { + "epoch": 2.900770244951773, + "grad_norm": 1.4136791229248047, + "learning_rate": 3.7062284633279354e-05, + "loss": 0.4381, + "step": 41803 + }, + { + "epoch": 2.9008396363888695, + "grad_norm": 1.1554328203201294, + "learning_rate": 3.705791850658565e-05, + "loss": 0.4006, + "step": 41804 + }, + { + "epoch": 2.9009090278259664, + "grad_norm": 1.0874085426330566, + "learning_rate": 3.7053552578592833e-05, + "loss": 0.3071, + "step": 41805 + }, + { + "epoch": 2.9009784192630628, + "grad_norm": 1.06411612033844, + "learning_rate": 3.704918684931461e-05, + "loss": 0.3517, + "step": 41806 + }, + { + "epoch": 2.9010478107001596, + "grad_norm": 1.271235704421997, + "learning_rate": 3.70448213187648e-05, + "loss": 0.3665, + "step": 41807 + }, + { + "epoch": 2.9011172021372564, + "grad_norm": 1.4061747789382935, + "learning_rate": 3.704045598695718e-05, + "loss": 0.3722, + "step": 41808 + }, + { + "epoch": 2.901186593574353, + "grad_norm": 1.3485170602798462, + "learning_rate": 3.7036090853905544e-05, + "loss": 0.356, + "step": 41809 + }, + { + "epoch": 2.9012559850114497, + "grad_norm": 1.80241858959198, + "learning_rate": 3.703172591962366e-05, + "loss": 0.4392, + "step": 41810 + }, + { + "epoch": 2.901325376448546, + "grad_norm": 0.9720348119735718, + "learning_rate": 3.7027361184125356e-05, + "loss": 0.4034, + "step": 41811 + }, + { + "epoch": 2.901394767885643, + "grad_norm": 1.0922378301620483, + "learning_rate": 3.702299664742431e-05, + "loss": 0.4458, + "step": 41812 + }, + { + "epoch": 2.9014641593227397, + "grad_norm": 1.1283031702041626, + "learning_rate": 3.701863230953437e-05, + "loss": 0.3498, + "step": 41813 + }, + { + "epoch": 2.901533550759836, + "grad_norm": 1.444799542427063, + "learning_rate": 3.70142681704693e-05, + "loss": 0.4268, + "step": 41814 + }, + { + "epoch": 2.901602942196933, + "grad_norm": 1.114794373512268, + "learning_rate": 3.7009904230242885e-05, + "loss": 0.3061, + "step": 41815 + }, + { + "epoch": 2.9016723336340293, + "grad_norm": 0.9253867268562317, + "learning_rate": 3.700554048886889e-05, + "loss": 0.301, + "step": 41816 + }, + { + "epoch": 2.901741725071126, + "grad_norm": 1.1521921157836914, + "learning_rate": 3.7001176946361094e-05, + "loss": 0.3364, + "step": 41817 + }, + { + "epoch": 2.901811116508223, + "grad_norm": 1.2275210618972778, + "learning_rate": 3.6996813602733304e-05, + "loss": 0.3258, + "step": 41818 + }, + { + "epoch": 2.90188050794532, + "grad_norm": 1.044878602027893, + "learning_rate": 3.699245045799923e-05, + "loss": 0.3295, + "step": 41819 + }, + { + "epoch": 2.901949899382416, + "grad_norm": 1.1826329231262207, + "learning_rate": 3.698808751217268e-05, + "loss": 0.3526, + "step": 41820 + }, + { + "epoch": 2.9020192908195126, + "grad_norm": 1.4466568231582642, + "learning_rate": 3.698372476526742e-05, + "loss": 0.4837, + "step": 41821 + }, + { + "epoch": 2.9020886822566094, + "grad_norm": 1.175660490989685, + "learning_rate": 3.697936221729724e-05, + "loss": 0.4364, + "step": 41822 + }, + { + "epoch": 2.9021580736937063, + "grad_norm": 1.3060635328292847, + "learning_rate": 3.6974999868275905e-05, + "loss": 0.3109, + "step": 41823 + }, + { + "epoch": 2.902227465130803, + "grad_norm": 1.0486435890197754, + "learning_rate": 3.697063771821717e-05, + "loss": 0.2849, + "step": 41824 + }, + { + "epoch": 2.9022968565678995, + "grad_norm": 1.3250977993011475, + "learning_rate": 3.696627576713486e-05, + "loss": 0.4027, + "step": 41825 + }, + { + "epoch": 2.9023662480049963, + "grad_norm": 1.0071909427642822, + "learning_rate": 3.6961914015042676e-05, + "loss": 0.2806, + "step": 41826 + }, + { + "epoch": 2.9024356394420927, + "grad_norm": 0.9581249356269836, + "learning_rate": 3.695755246195442e-05, + "loss": 0.4065, + "step": 41827 + }, + { + "epoch": 2.9025050308791895, + "grad_norm": 1.2424172163009644, + "learning_rate": 3.6953191107883835e-05, + "loss": 0.3987, + "step": 41828 + }, + { + "epoch": 2.9025744223162864, + "grad_norm": 1.0633430480957031, + "learning_rate": 3.6948829952844724e-05, + "loss": 0.3284, + "step": 41829 + }, + { + "epoch": 2.9026438137533828, + "grad_norm": 1.229828953742981, + "learning_rate": 3.694446899685085e-05, + "loss": 0.3682, + "step": 41830 + }, + { + "epoch": 2.9027132051904796, + "grad_norm": 0.958867073059082, + "learning_rate": 3.694010823991598e-05, + "loss": 0.2529, + "step": 41831 + }, + { + "epoch": 2.902782596627576, + "grad_norm": 2.9291059970855713, + "learning_rate": 3.693574768205387e-05, + "loss": 0.3628, + "step": 41832 + }, + { + "epoch": 2.902851988064673, + "grad_norm": 0.9786421656608582, + "learning_rate": 3.6931387323278264e-05, + "loss": 0.4444, + "step": 41833 + }, + { + "epoch": 2.9029213795017696, + "grad_norm": 1.1548233032226562, + "learning_rate": 3.692702716360296e-05, + "loss": 0.3584, + "step": 41834 + }, + { + "epoch": 2.902990770938866, + "grad_norm": 1.3733887672424316, + "learning_rate": 3.6922667203041706e-05, + "loss": 0.3806, + "step": 41835 + }, + { + "epoch": 2.903060162375963, + "grad_norm": 1.0340334177017212, + "learning_rate": 3.691830744160831e-05, + "loss": 0.3594, + "step": 41836 + }, + { + "epoch": 2.9031295538130593, + "grad_norm": 1.5743483304977417, + "learning_rate": 3.691394787931648e-05, + "loss": 0.3003, + "step": 41837 + }, + { + "epoch": 2.903198945250156, + "grad_norm": 1.0779904127120972, + "learning_rate": 3.6909588516179984e-05, + "loss": 0.3499, + "step": 41838 + }, + { + "epoch": 2.903268336687253, + "grad_norm": 1.2694412469863892, + "learning_rate": 3.690522935221259e-05, + "loss": 0.4782, + "step": 41839 + }, + { + "epoch": 2.9033377281243493, + "grad_norm": 1.5580511093139648, + "learning_rate": 3.690087038742811e-05, + "loss": 0.484, + "step": 41840 + }, + { + "epoch": 2.903407119561446, + "grad_norm": 0.8124770522117615, + "learning_rate": 3.689651162184024e-05, + "loss": 0.2821, + "step": 41841 + }, + { + "epoch": 2.9034765109985425, + "grad_norm": 1.3982993364334106, + "learning_rate": 3.689215305546274e-05, + "loss": 0.5575, + "step": 41842 + }, + { + "epoch": 2.9035459024356394, + "grad_norm": 1.2282088994979858, + "learning_rate": 3.6887794688309395e-05, + "loss": 0.3974, + "step": 41843 + }, + { + "epoch": 2.903615293872736, + "grad_norm": 1.9925349950790405, + "learning_rate": 3.688343652039397e-05, + "loss": 0.4384, + "step": 41844 + }, + { + "epoch": 2.903684685309833, + "grad_norm": 0.9654726982116699, + "learning_rate": 3.687907855173024e-05, + "loss": 0.356, + "step": 41845 + }, + { + "epoch": 2.9037540767469294, + "grad_norm": 0.7100779414176941, + "learning_rate": 3.68747207823319e-05, + "loss": 0.2405, + "step": 41846 + }, + { + "epoch": 2.9038234681840263, + "grad_norm": 1.429304599761963, + "learning_rate": 3.6870363212212745e-05, + "loss": 0.4112, + "step": 41847 + }, + { + "epoch": 2.9038928596211226, + "grad_norm": 1.408433198928833, + "learning_rate": 3.6866005841386534e-05, + "loss": 0.5226, + "step": 41848 + }, + { + "epoch": 2.9039622510582195, + "grad_norm": 1.8465399742126465, + "learning_rate": 3.6861648669867e-05, + "loss": 0.4437, + "step": 41849 + }, + { + "epoch": 2.9040316424953163, + "grad_norm": 1.407037615776062, + "learning_rate": 3.685729169766793e-05, + "loss": 0.3345, + "step": 41850 + }, + { + "epoch": 2.9041010339324127, + "grad_norm": 0.8776957392692566, + "learning_rate": 3.685293492480306e-05, + "loss": 0.4051, + "step": 41851 + }, + { + "epoch": 2.9041704253695095, + "grad_norm": 1.4271748065948486, + "learning_rate": 3.684857835128619e-05, + "loss": 0.5636, + "step": 41852 + }, + { + "epoch": 2.904239816806606, + "grad_norm": 1.1014362573623657, + "learning_rate": 3.684422197713099e-05, + "loss": 0.439, + "step": 41853 + }, + { + "epoch": 2.9043092082437028, + "grad_norm": 1.0901844501495361, + "learning_rate": 3.683986580235126e-05, + "loss": 0.3568, + "step": 41854 + }, + { + "epoch": 2.9043785996807996, + "grad_norm": 0.99900221824646, + "learning_rate": 3.6835509826960734e-05, + "loss": 0.3557, + "step": 41855 + }, + { + "epoch": 2.904447991117896, + "grad_norm": 1.2706166505813599, + "learning_rate": 3.6831154050973184e-05, + "loss": 0.5414, + "step": 41856 + }, + { + "epoch": 2.904517382554993, + "grad_norm": 1.2102289199829102, + "learning_rate": 3.682679847440235e-05, + "loss": 0.3231, + "step": 41857 + }, + { + "epoch": 2.904586773992089, + "grad_norm": 1.1435542106628418, + "learning_rate": 3.682244309726198e-05, + "loss": 0.3356, + "step": 41858 + }, + { + "epoch": 2.904656165429186, + "grad_norm": 1.1037728786468506, + "learning_rate": 3.6818087919565866e-05, + "loss": 0.405, + "step": 41859 + }, + { + "epoch": 2.904725556866283, + "grad_norm": 1.0864976644515991, + "learning_rate": 3.681373294132768e-05, + "loss": 0.3919, + "step": 41860 + }, + { + "epoch": 2.9047949483033793, + "grad_norm": 0.8765677213668823, + "learning_rate": 3.6809378162561215e-05, + "loss": 0.4056, + "step": 41861 + }, + { + "epoch": 2.904864339740476, + "grad_norm": 1.070288062095642, + "learning_rate": 3.680502358328021e-05, + "loss": 0.429, + "step": 41862 + }, + { + "epoch": 2.9049337311775725, + "grad_norm": 0.9371625185012817, + "learning_rate": 3.680066920349841e-05, + "loss": 0.2899, + "step": 41863 + }, + { + "epoch": 2.9050031226146693, + "grad_norm": 1.3736547231674194, + "learning_rate": 3.6796315023229567e-05, + "loss": 0.4878, + "step": 41864 + }, + { + "epoch": 2.905072514051766, + "grad_norm": 0.9937587976455688, + "learning_rate": 3.679196104248742e-05, + "loss": 0.322, + "step": 41865 + }, + { + "epoch": 2.9051419054888625, + "grad_norm": 0.7717457413673401, + "learning_rate": 3.6787607261285763e-05, + "loss": 0.1963, + "step": 41866 + }, + { + "epoch": 2.9052112969259594, + "grad_norm": 0.780862033367157, + "learning_rate": 3.678325367963825e-05, + "loss": 0.3051, + "step": 41867 + }, + { + "epoch": 2.9052806883630558, + "grad_norm": 0.8904052376747131, + "learning_rate": 3.677890029755867e-05, + "loss": 0.2565, + "step": 41868 + }, + { + "epoch": 2.9053500798001526, + "grad_norm": 1.4978270530700684, + "learning_rate": 3.677454711506077e-05, + "loss": 0.4297, + "step": 41869 + }, + { + "epoch": 2.9054194712372494, + "grad_norm": 1.1255568265914917, + "learning_rate": 3.6770194132158295e-05, + "loss": 0.3769, + "step": 41870 + }, + { + "epoch": 2.9054888626743463, + "grad_norm": 0.7897615432739258, + "learning_rate": 3.676584134886502e-05, + "loss": 0.2756, + "step": 41871 + }, + { + "epoch": 2.9055582541114426, + "grad_norm": 1.5519698858261108, + "learning_rate": 3.676148876519459e-05, + "loss": 0.47, + "step": 41872 + }, + { + "epoch": 2.9056276455485395, + "grad_norm": 1.318682074546814, + "learning_rate": 3.675713638116083e-05, + "loss": 0.407, + "step": 41873 + }, + { + "epoch": 2.905697036985636, + "grad_norm": 1.3758282661437988, + "learning_rate": 3.675278419677744e-05, + "loss": 0.5252, + "step": 41874 + }, + { + "epoch": 2.9057664284227327, + "grad_norm": 1.0709521770477295, + "learning_rate": 3.6748432212058205e-05, + "loss": 0.2956, + "step": 41875 + }, + { + "epoch": 2.9058358198598295, + "grad_norm": 1.921492576599121, + "learning_rate": 3.6744080427016805e-05, + "loss": 0.5309, + "step": 41876 + }, + { + "epoch": 2.905905211296926, + "grad_norm": 1.016510009765625, + "learning_rate": 3.673972884166699e-05, + "loss": 0.3437, + "step": 41877 + }, + { + "epoch": 2.9059746027340227, + "grad_norm": 0.8460122346878052, + "learning_rate": 3.6735377456022526e-05, + "loss": 0.3024, + "step": 41878 + }, + { + "epoch": 2.906043994171119, + "grad_norm": 0.9268073439598083, + "learning_rate": 3.6731026270097125e-05, + "loss": 0.2994, + "step": 41879 + }, + { + "epoch": 2.906113385608216, + "grad_norm": 1.539153814315796, + "learning_rate": 3.672667528390458e-05, + "loss": 0.5422, + "step": 41880 + }, + { + "epoch": 2.906182777045313, + "grad_norm": 1.195716142654419, + "learning_rate": 3.6722324497458524e-05, + "loss": 0.2675, + "step": 41881 + }, + { + "epoch": 2.906252168482409, + "grad_norm": 1.1319317817687988, + "learning_rate": 3.671797391077276e-05, + "loss": 0.355, + "step": 41882 + }, + { + "epoch": 2.906321559919506, + "grad_norm": 0.9820546507835388, + "learning_rate": 3.671362352386102e-05, + "loss": 0.3689, + "step": 41883 + }, + { + "epoch": 2.9063909513566024, + "grad_norm": 1.0294976234436035, + "learning_rate": 3.670927333673702e-05, + "loss": 0.3973, + "step": 41884 + }, + { + "epoch": 2.9064603427936992, + "grad_norm": 1.0440891981124878, + "learning_rate": 3.67049233494145e-05, + "loss": 0.5726, + "step": 41885 + }, + { + "epoch": 2.906529734230796, + "grad_norm": 0.8508288264274597, + "learning_rate": 3.6700573561907195e-05, + "loss": 0.2663, + "step": 41886 + }, + { + "epoch": 2.9065991256678925, + "grad_norm": 1.4644834995269775, + "learning_rate": 3.669622397422886e-05, + "loss": 0.4643, + "step": 41887 + }, + { + "epoch": 2.9066685171049893, + "grad_norm": 1.417457103729248, + "learning_rate": 3.669187458639318e-05, + "loss": 0.4078, + "step": 41888 + }, + { + "epoch": 2.9067379085420857, + "grad_norm": 1.1683354377746582, + "learning_rate": 3.66875253984139e-05, + "loss": 0.4209, + "step": 41889 + }, + { + "epoch": 2.9068072999791825, + "grad_norm": 0.8664402961730957, + "learning_rate": 3.668317641030477e-05, + "loss": 0.3242, + "step": 41890 + }, + { + "epoch": 2.9068766914162794, + "grad_norm": 0.8725019693374634, + "learning_rate": 3.6678827622079495e-05, + "loss": 0.2835, + "step": 41891 + }, + { + "epoch": 2.9069460828533757, + "grad_norm": 1.1285102367401123, + "learning_rate": 3.667447903375181e-05, + "loss": 0.3116, + "step": 41892 + }, + { + "epoch": 2.9070154742904726, + "grad_norm": 0.8179855942726135, + "learning_rate": 3.6670130645335444e-05, + "loss": 0.1884, + "step": 41893 + }, + { + "epoch": 2.907084865727569, + "grad_norm": 0.922105073928833, + "learning_rate": 3.666578245684418e-05, + "loss": 0.3114, + "step": 41894 + }, + { + "epoch": 2.907154257164666, + "grad_norm": 0.9030975103378296, + "learning_rate": 3.6661434468291646e-05, + "loss": 0.2409, + "step": 41895 + }, + { + "epoch": 2.9072236486017626, + "grad_norm": 1.1205610036849976, + "learning_rate": 3.665708667969161e-05, + "loss": 0.3396, + "step": 41896 + }, + { + "epoch": 2.9072930400388595, + "grad_norm": 1.0202640295028687, + "learning_rate": 3.665273909105782e-05, + "loss": 0.3203, + "step": 41897 + }, + { + "epoch": 2.907362431475956, + "grad_norm": 1.29179847240448, + "learning_rate": 3.664839170240397e-05, + "loss": 0.5572, + "step": 41898 + }, + { + "epoch": 2.9074318229130527, + "grad_norm": 0.8774092197418213, + "learning_rate": 3.66440445137438e-05, + "loss": 0.3812, + "step": 41899 + }, + { + "epoch": 2.907501214350149, + "grad_norm": 1.3336931467056274, + "learning_rate": 3.663969752509103e-05, + "loss": 0.4373, + "step": 41900 + }, + { + "epoch": 2.907570605787246, + "grad_norm": 1.1086866855621338, + "learning_rate": 3.663535073645942e-05, + "loss": 0.3428, + "step": 41901 + }, + { + "epoch": 2.9076399972243427, + "grad_norm": 1.1961747407913208, + "learning_rate": 3.663100414786261e-05, + "loss": 0.5147, + "step": 41902 + }, + { + "epoch": 2.907709388661439, + "grad_norm": 1.1929572820663452, + "learning_rate": 3.6626657759314387e-05, + "loss": 0.3437, + "step": 41903 + }, + { + "epoch": 2.907778780098536, + "grad_norm": 1.0448442697525024, + "learning_rate": 3.662231157082844e-05, + "loss": 0.2999, + "step": 41904 + }, + { + "epoch": 2.9078481715356324, + "grad_norm": 1.5108500719070435, + "learning_rate": 3.661796558241851e-05, + "loss": 0.3882, + "step": 41905 + }, + { + "epoch": 2.907917562972729, + "grad_norm": 0.9915630221366882, + "learning_rate": 3.6613619794098345e-05, + "loss": 0.4281, + "step": 41906 + }, + { + "epoch": 2.907986954409826, + "grad_norm": 2.3785903453826904, + "learning_rate": 3.660927420588158e-05, + "loss": 0.5422, + "step": 41907 + }, + { + "epoch": 2.9080563458469224, + "grad_norm": 1.0200355052947998, + "learning_rate": 3.660492881778199e-05, + "loss": 0.4694, + "step": 41908 + }, + { + "epoch": 2.9081257372840192, + "grad_norm": 1.5654325485229492, + "learning_rate": 3.660058362981332e-05, + "loss": 0.3737, + "step": 41909 + }, + { + "epoch": 2.9081951287211156, + "grad_norm": 1.3569859266281128, + "learning_rate": 3.659623864198922e-05, + "loss": 0.3792, + "step": 41910 + }, + { + "epoch": 2.9082645201582125, + "grad_norm": 1.0405573844909668, + "learning_rate": 3.659189385432343e-05, + "loss": 0.3928, + "step": 41911 + }, + { + "epoch": 2.9083339115953093, + "grad_norm": 0.9422658085823059, + "learning_rate": 3.658754926682968e-05, + "loss": 0.3388, + "step": 41912 + }, + { + "epoch": 2.9084033030324057, + "grad_norm": 1.279545783996582, + "learning_rate": 3.6583204879521685e-05, + "loss": 0.4731, + "step": 41913 + }, + { + "epoch": 2.9084726944695025, + "grad_norm": 0.9946120381355286, + "learning_rate": 3.6578860692413175e-05, + "loss": 0.4055, + "step": 41914 + }, + { + "epoch": 2.908542085906599, + "grad_norm": 1.3462175130844116, + "learning_rate": 3.657451670551782e-05, + "loss": 0.3606, + "step": 41915 + }, + { + "epoch": 2.9086114773436957, + "grad_norm": 1.7931419610977173, + "learning_rate": 3.6570172918849355e-05, + "loss": 0.4574, + "step": 41916 + }, + { + "epoch": 2.9086808687807926, + "grad_norm": 1.1936416625976562, + "learning_rate": 3.6565829332421496e-05, + "loss": 0.2939, + "step": 41917 + }, + { + "epoch": 2.908750260217889, + "grad_norm": 1.3923150300979614, + "learning_rate": 3.6561485946247954e-05, + "loss": 0.5022, + "step": 41918 + }, + { + "epoch": 2.908819651654986, + "grad_norm": 1.1342852115631104, + "learning_rate": 3.6557142760342436e-05, + "loss": 0.3999, + "step": 41919 + }, + { + "epoch": 2.908889043092082, + "grad_norm": 1.3010843992233276, + "learning_rate": 3.655279977471866e-05, + "loss": 0.3515, + "step": 41920 + }, + { + "epoch": 2.908958434529179, + "grad_norm": 1.3024238348007202, + "learning_rate": 3.654845698939037e-05, + "loss": 0.3856, + "step": 41921 + }, + { + "epoch": 2.909027825966276, + "grad_norm": 1.2430716753005981, + "learning_rate": 3.65441144043712e-05, + "loss": 0.3726, + "step": 41922 + }, + { + "epoch": 2.9090972174033727, + "grad_norm": 1.2190277576446533, + "learning_rate": 3.6539772019674904e-05, + "loss": 0.4931, + "step": 41923 + }, + { + "epoch": 2.909166608840469, + "grad_norm": 1.5207973718643188, + "learning_rate": 3.6535429835315183e-05, + "loss": 0.4266, + "step": 41924 + }, + { + "epoch": 2.909236000277566, + "grad_norm": 1.0540040731430054, + "learning_rate": 3.6531087851305745e-05, + "loss": 0.39, + "step": 41925 + }, + { + "epoch": 2.9093053917146623, + "grad_norm": 1.2018946409225464, + "learning_rate": 3.65267460676603e-05, + "loss": 0.4652, + "step": 41926 + }, + { + "epoch": 2.909374783151759, + "grad_norm": 1.2576260566711426, + "learning_rate": 3.652240448439256e-05, + "loss": 0.2764, + "step": 41927 + }, + { + "epoch": 2.909444174588856, + "grad_norm": 1.362175464630127, + "learning_rate": 3.651806310151625e-05, + "loss": 0.4185, + "step": 41928 + }, + { + "epoch": 2.9095135660259523, + "grad_norm": 1.373797059059143, + "learning_rate": 3.651372191904502e-05, + "loss": 0.3207, + "step": 41929 + }, + { + "epoch": 2.909582957463049, + "grad_norm": 1.1590895652770996, + "learning_rate": 3.65093809369926e-05, + "loss": 0.4831, + "step": 41930 + }, + { + "epoch": 2.9096523489001456, + "grad_norm": 1.2651374340057373, + "learning_rate": 3.65050401553727e-05, + "loss": 0.4389, + "step": 41931 + }, + { + "epoch": 2.9097217403372424, + "grad_norm": 1.444878339767456, + "learning_rate": 3.6500699574199025e-05, + "loss": 0.4439, + "step": 41932 + }, + { + "epoch": 2.9097911317743392, + "grad_norm": 1.063271403312683, + "learning_rate": 3.649635919348526e-05, + "loss": 0.3207, + "step": 41933 + }, + { + "epoch": 2.9098605232114356, + "grad_norm": 0.98734050989151, + "learning_rate": 3.6492019013245136e-05, + "loss": 0.4342, + "step": 41934 + }, + { + "epoch": 2.9099299146485325, + "grad_norm": 1.3351601362228394, + "learning_rate": 3.648767903349237e-05, + "loss": 0.5346, + "step": 41935 + }, + { + "epoch": 2.909999306085629, + "grad_norm": 0.893265962600708, + "learning_rate": 3.64833392542406e-05, + "loss": 0.296, + "step": 41936 + }, + { + "epoch": 2.9100686975227257, + "grad_norm": 1.0471436977386475, + "learning_rate": 3.6478999675503556e-05, + "loss": 0.2704, + "step": 41937 + }, + { + "epoch": 2.9101380889598225, + "grad_norm": 0.8536232113838196, + "learning_rate": 3.6474660297294946e-05, + "loss": 0.299, + "step": 41938 + }, + { + "epoch": 2.910207480396919, + "grad_norm": 1.8871341943740845, + "learning_rate": 3.647032111962846e-05, + "loss": 0.4163, + "step": 41939 + }, + { + "epoch": 2.9102768718340157, + "grad_norm": 1.2242718935012817, + "learning_rate": 3.6465982142517796e-05, + "loss": 0.4252, + "step": 41940 + }, + { + "epoch": 2.910346263271112, + "grad_norm": 1.1163203716278076, + "learning_rate": 3.646164336597669e-05, + "loss": 0.3128, + "step": 41941 + }, + { + "epoch": 2.910415654708209, + "grad_norm": 1.0103216171264648, + "learning_rate": 3.645730479001878e-05, + "loss": 0.3853, + "step": 41942 + }, + { + "epoch": 2.910485046145306, + "grad_norm": 1.0944302082061768, + "learning_rate": 3.645296641465776e-05, + "loss": 0.3542, + "step": 41943 + }, + { + "epoch": 2.910554437582402, + "grad_norm": 1.3344026803970337, + "learning_rate": 3.644862823990741e-05, + "loss": 0.4087, + "step": 41944 + }, + { + "epoch": 2.910623829019499, + "grad_norm": 0.9276654124259949, + "learning_rate": 3.644429026578133e-05, + "loss": 0.3007, + "step": 41945 + }, + { + "epoch": 2.9106932204565954, + "grad_norm": 1.1251400709152222, + "learning_rate": 3.643995249229325e-05, + "loss": 0.3344, + "step": 41946 + }, + { + "epoch": 2.9107626118936922, + "grad_norm": 1.1130937337875366, + "learning_rate": 3.6435614919456865e-05, + "loss": 0.317, + "step": 41947 + }, + { + "epoch": 2.910832003330789, + "grad_norm": 0.9056435823440552, + "learning_rate": 3.643127754728587e-05, + "loss": 0.4035, + "step": 41948 + }, + { + "epoch": 2.910901394767886, + "grad_norm": 0.8731787204742432, + "learning_rate": 3.6426940375793985e-05, + "loss": 0.3972, + "step": 41949 + }, + { + "epoch": 2.9109707862049823, + "grad_norm": 1.072829008102417, + "learning_rate": 3.642260340499484e-05, + "loss": 0.3258, + "step": 41950 + }, + { + "epoch": 2.911040177642079, + "grad_norm": 0.7954230904579163, + "learning_rate": 3.641826663490216e-05, + "loss": 0.2618, + "step": 41951 + }, + { + "epoch": 2.9111095690791755, + "grad_norm": 1.4370137453079224, + "learning_rate": 3.641393006552964e-05, + "loss": 0.4507, + "step": 41952 + }, + { + "epoch": 2.9111789605162723, + "grad_norm": 1.7053574323654175, + "learning_rate": 3.640959369689096e-05, + "loss": 0.4393, + "step": 41953 + }, + { + "epoch": 2.911248351953369, + "grad_norm": 1.2028472423553467, + "learning_rate": 3.640525752899983e-05, + "loss": 0.4087, + "step": 41954 + }, + { + "epoch": 2.9113177433904656, + "grad_norm": 1.3940619230270386, + "learning_rate": 3.64009215618699e-05, + "loss": 0.5062, + "step": 41955 + }, + { + "epoch": 2.9113871348275624, + "grad_norm": 1.146039366722107, + "learning_rate": 3.6396585795514935e-05, + "loss": 0.389, + "step": 41956 + }, + { + "epoch": 2.911456526264659, + "grad_norm": 0.9259560704231262, + "learning_rate": 3.639225022994853e-05, + "loss": 0.3289, + "step": 41957 + }, + { + "epoch": 2.9115259177017556, + "grad_norm": 1.2979793548583984, + "learning_rate": 3.6387914865184405e-05, + "loss": 0.3164, + "step": 41958 + }, + { + "epoch": 2.9115953091388524, + "grad_norm": 1.5066286325454712, + "learning_rate": 3.638357970123626e-05, + "loss": 0.339, + "step": 41959 + }, + { + "epoch": 2.911664700575949, + "grad_norm": 1.0286694765090942, + "learning_rate": 3.637924473811777e-05, + "loss": 0.2978, + "step": 41960 + }, + { + "epoch": 2.9117340920130457, + "grad_norm": 1.6067581176757812, + "learning_rate": 3.637490997584262e-05, + "loss": 0.4259, + "step": 41961 + }, + { + "epoch": 2.911803483450142, + "grad_norm": 0.9622563719749451, + "learning_rate": 3.63705754144245e-05, + "loss": 0.2872, + "step": 41962 + }, + { + "epoch": 2.911872874887239, + "grad_norm": 1.036198616027832, + "learning_rate": 3.6366241053877126e-05, + "loss": 0.2898, + "step": 41963 + }, + { + "epoch": 2.9119422663243357, + "grad_norm": 1.0777095556259155, + "learning_rate": 3.636190689421412e-05, + "loss": 0.4197, + "step": 41964 + }, + { + "epoch": 2.912011657761432, + "grad_norm": 1.022697925567627, + "learning_rate": 3.635757293544917e-05, + "loss": 0.3269, + "step": 41965 + }, + { + "epoch": 2.912081049198529, + "grad_norm": 1.373852014541626, + "learning_rate": 3.6353239177596e-05, + "loss": 0.5088, + "step": 41966 + }, + { + "epoch": 2.9121504406356253, + "grad_norm": 1.4876190423965454, + "learning_rate": 3.634890562066825e-05, + "loss": 0.5376, + "step": 41967 + }, + { + "epoch": 2.912219832072722, + "grad_norm": 1.0415844917297363, + "learning_rate": 3.634457226467963e-05, + "loss": 0.3204, + "step": 41968 + }, + { + "epoch": 2.912289223509819, + "grad_norm": 1.0188649892807007, + "learning_rate": 3.6340239109643815e-05, + "loss": 0.4856, + "step": 41969 + }, + { + "epoch": 2.9123586149469154, + "grad_norm": 1.0861120223999023, + "learning_rate": 3.633590615557451e-05, + "loss": 0.3458, + "step": 41970 + }, + { + "epoch": 2.9124280063840122, + "grad_norm": 0.9482064843177795, + "learning_rate": 3.633157340248532e-05, + "loss": 0.2535, + "step": 41971 + }, + { + "epoch": 2.9124973978211086, + "grad_norm": 0.973273754119873, + "learning_rate": 3.632724085038998e-05, + "loss": 0.3519, + "step": 41972 + }, + { + "epoch": 2.9125667892582054, + "grad_norm": 0.8247746229171753, + "learning_rate": 3.6322908499302144e-05, + "loss": 0.3577, + "step": 41973 + }, + { + "epoch": 2.9126361806953023, + "grad_norm": 1.0377591848373413, + "learning_rate": 3.63185763492355e-05, + "loss": 0.4698, + "step": 41974 + }, + { + "epoch": 2.912705572132399, + "grad_norm": 1.1468579769134521, + "learning_rate": 3.6314244400203754e-05, + "loss": 0.4776, + "step": 41975 + }, + { + "epoch": 2.9127749635694955, + "grad_norm": 1.6031630039215088, + "learning_rate": 3.6309912652220515e-05, + "loss": 0.3468, + "step": 41976 + }, + { + "epoch": 2.9128443550065923, + "grad_norm": 1.2109174728393555, + "learning_rate": 3.63055811052995e-05, + "loss": 0.5828, + "step": 41977 + }, + { + "epoch": 2.9129137464436887, + "grad_norm": 1.6936520338058472, + "learning_rate": 3.630124975945437e-05, + "loss": 0.4201, + "step": 41978 + }, + { + "epoch": 2.9129831378807856, + "grad_norm": 1.08649742603302, + "learning_rate": 3.629691861469884e-05, + "loss": 0.4094, + "step": 41979 + }, + { + "epoch": 2.9130525293178824, + "grad_norm": 1.2701151371002197, + "learning_rate": 3.629258767104651e-05, + "loss": 0.3797, + "step": 41980 + }, + { + "epoch": 2.9131219207549788, + "grad_norm": 0.872582733631134, + "learning_rate": 3.628825692851109e-05, + "loss": 0.3737, + "step": 41981 + }, + { + "epoch": 2.9131913121920756, + "grad_norm": 1.26236093044281, + "learning_rate": 3.6283926387106245e-05, + "loss": 0.4553, + "step": 41982 + }, + { + "epoch": 2.913260703629172, + "grad_norm": 0.8509743213653564, + "learning_rate": 3.62795960468457e-05, + "loss": 0.3335, + "step": 41983 + }, + { + "epoch": 2.913330095066269, + "grad_norm": 0.9293652176856995, + "learning_rate": 3.627526590774303e-05, + "loss": 0.3286, + "step": 41984 + }, + { + "epoch": 2.9133994865033657, + "grad_norm": 0.8237786293029785, + "learning_rate": 3.627093596981196e-05, + "loss": 0.2422, + "step": 41985 + }, + { + "epoch": 2.913468877940462, + "grad_norm": 0.8062194585800171, + "learning_rate": 3.626660623306615e-05, + "loss": 0.3618, + "step": 41986 + }, + { + "epoch": 2.913538269377559, + "grad_norm": 0.8614299297332764, + "learning_rate": 3.626227669751927e-05, + "loss": 0.3374, + "step": 41987 + }, + { + "epoch": 2.9136076608146553, + "grad_norm": 0.8408482670783997, + "learning_rate": 3.6257947363184986e-05, + "loss": 0.2595, + "step": 41988 + }, + { + "epoch": 2.913677052251752, + "grad_norm": 0.9765698909759521, + "learning_rate": 3.625361823007697e-05, + "loss": 0.2956, + "step": 41989 + }, + { + "epoch": 2.913746443688849, + "grad_norm": 1.0328500270843506, + "learning_rate": 3.624928929820891e-05, + "loss": 0.3107, + "step": 41990 + }, + { + "epoch": 2.9138158351259453, + "grad_norm": 1.0951673984527588, + "learning_rate": 3.624496056759442e-05, + "loss": 0.4025, + "step": 41991 + }, + { + "epoch": 2.913885226563042, + "grad_norm": 1.1811094284057617, + "learning_rate": 3.624063203824719e-05, + "loss": 0.3448, + "step": 41992 + }, + { + "epoch": 2.9139546180001386, + "grad_norm": 1.7308869361877441, + "learning_rate": 3.6236303710180896e-05, + "loss": 0.3612, + "step": 41993 + }, + { + "epoch": 2.9140240094372354, + "grad_norm": 1.406875491142273, + "learning_rate": 3.623197558340918e-05, + "loss": 0.3063, + "step": 41994 + }, + { + "epoch": 2.914093400874332, + "grad_norm": 0.9675381183624268, + "learning_rate": 3.622764765794573e-05, + "loss": 0.3566, + "step": 41995 + }, + { + "epoch": 2.9141627923114286, + "grad_norm": 1.5359145402908325, + "learning_rate": 3.622331993380419e-05, + "loss": 0.4728, + "step": 41996 + }, + { + "epoch": 2.9142321837485254, + "grad_norm": 1.109103798866272, + "learning_rate": 3.621899241099827e-05, + "loss": 0.3371, + "step": 41997 + }, + { + "epoch": 2.914301575185622, + "grad_norm": 1.360114574432373, + "learning_rate": 3.6214665089541555e-05, + "loss": 0.598, + "step": 41998 + }, + { + "epoch": 2.9143709666227187, + "grad_norm": 1.4402722120285034, + "learning_rate": 3.621033796944774e-05, + "loss": 0.5261, + "step": 41999 + }, + { + "epoch": 2.9144403580598155, + "grad_norm": 1.080800175666809, + "learning_rate": 3.620601105073048e-05, + "loss": 0.3782, + "step": 42000 + }, + { + "epoch": 2.9145097494969123, + "grad_norm": 2.0498836040496826, + "learning_rate": 3.620168433340345e-05, + "loss": 0.6446, + "step": 42001 + }, + { + "epoch": 2.9145791409340087, + "grad_norm": 1.5092006921768188, + "learning_rate": 3.61973578174803e-05, + "loss": 0.509, + "step": 42002 + }, + { + "epoch": 2.9146485323711055, + "grad_norm": 1.0129798650741577, + "learning_rate": 3.619303150297469e-05, + "loss": 0.3547, + "step": 42003 + }, + { + "epoch": 2.914717923808202, + "grad_norm": 1.080397129058838, + "learning_rate": 3.618870538990031e-05, + "loss": 0.3724, + "step": 42004 + }, + { + "epoch": 2.9147873152452988, + "grad_norm": 1.0822144746780396, + "learning_rate": 3.6184379478270745e-05, + "loss": 0.3339, + "step": 42005 + }, + { + "epoch": 2.9148567066823956, + "grad_norm": 1.5090758800506592, + "learning_rate": 3.6180053768099696e-05, + "loss": 0.445, + "step": 42006 + }, + { + "epoch": 2.914926098119492, + "grad_norm": 1.1819584369659424, + "learning_rate": 3.617572825940081e-05, + "loss": 0.4078, + "step": 42007 + }, + { + "epoch": 2.914995489556589, + "grad_norm": 1.685465693473816, + "learning_rate": 3.617140295218775e-05, + "loss": 0.4972, + "step": 42008 + }, + { + "epoch": 2.915064880993685, + "grad_norm": 1.0779500007629395, + "learning_rate": 3.616707784647415e-05, + "loss": 0.385, + "step": 42009 + }, + { + "epoch": 2.915134272430782, + "grad_norm": 1.2346714735031128, + "learning_rate": 3.6162752942273726e-05, + "loss": 0.3315, + "step": 42010 + }, + { + "epoch": 2.915203663867879, + "grad_norm": 1.0735692977905273, + "learning_rate": 3.615842823960005e-05, + "loss": 0.4192, + "step": 42011 + }, + { + "epoch": 2.9152730553049753, + "grad_norm": 1.4683183431625366, + "learning_rate": 3.6154103738466805e-05, + "loss": 0.363, + "step": 42012 + }, + { + "epoch": 2.915342446742072, + "grad_norm": 1.1127853393554688, + "learning_rate": 3.6149779438887655e-05, + "loss": 0.4564, + "step": 42013 + }, + { + "epoch": 2.9154118381791685, + "grad_norm": 1.5211808681488037, + "learning_rate": 3.6145455340876265e-05, + "loss": 0.276, + "step": 42014 + }, + { + "epoch": 2.9154812296162653, + "grad_norm": 1.040279507637024, + "learning_rate": 3.614113144444624e-05, + "loss": 0.409, + "step": 42015 + }, + { + "epoch": 2.915550621053362, + "grad_norm": 1.3934288024902344, + "learning_rate": 3.6136807749611245e-05, + "loss": 0.4363, + "step": 42016 + }, + { + "epoch": 2.9156200124904585, + "grad_norm": 1.2201300859451294, + "learning_rate": 3.613248425638494e-05, + "loss": 0.6042, + "step": 42017 + }, + { + "epoch": 2.9156894039275554, + "grad_norm": 0.9364071488380432, + "learning_rate": 3.6128160964781025e-05, + "loss": 0.2131, + "step": 42018 + }, + { + "epoch": 2.9157587953646518, + "grad_norm": 0.92835932970047, + "learning_rate": 3.6123837874813035e-05, + "loss": 0.2805, + "step": 42019 + }, + { + "epoch": 2.9158281868017486, + "grad_norm": 1.522050380706787, + "learning_rate": 3.6119514986494696e-05, + "loss": 0.4228, + "step": 42020 + }, + { + "epoch": 2.9158975782388454, + "grad_norm": 1.2939332723617554, + "learning_rate": 3.611519229983963e-05, + "loss": 0.4042, + "step": 42021 + }, + { + "epoch": 2.915966969675942, + "grad_norm": 1.4640693664550781, + "learning_rate": 3.611086981486149e-05, + "loss": 0.3776, + "step": 42022 + }, + { + "epoch": 2.9160363611130387, + "grad_norm": 1.4467979669570923, + "learning_rate": 3.610654753157392e-05, + "loss": 0.3583, + "step": 42023 + }, + { + "epoch": 2.916105752550135, + "grad_norm": 1.238768458366394, + "learning_rate": 3.6102225449990566e-05, + "loss": 0.4538, + "step": 42024 + }, + { + "epoch": 2.916175143987232, + "grad_norm": 1.3024497032165527, + "learning_rate": 3.609790357012511e-05, + "loss": 0.4755, + "step": 42025 + }, + { + "epoch": 2.9162445354243287, + "grad_norm": 1.5853327512741089, + "learning_rate": 3.609358189199112e-05, + "loss": 0.4603, + "step": 42026 + }, + { + "epoch": 2.9163139268614255, + "grad_norm": 1.2286827564239502, + "learning_rate": 3.608926041560229e-05, + "loss": 0.2634, + "step": 42027 + }, + { + "epoch": 2.916383318298522, + "grad_norm": 1.2354214191436768, + "learning_rate": 3.6084939140972226e-05, + "loss": 0.5259, + "step": 42028 + }, + { + "epoch": 2.9164527097356188, + "grad_norm": 1.6145095825195312, + "learning_rate": 3.608061806811461e-05, + "loss": 0.464, + "step": 42029 + }, + { + "epoch": 2.916522101172715, + "grad_norm": 1.184546947479248, + "learning_rate": 3.6076297197043076e-05, + "loss": 0.4495, + "step": 42030 + }, + { + "epoch": 2.916591492609812, + "grad_norm": 1.4139901399612427, + "learning_rate": 3.607197652777123e-05, + "loss": 0.4626, + "step": 42031 + }, + { + "epoch": 2.916660884046909, + "grad_norm": 0.9658908247947693, + "learning_rate": 3.606765606031279e-05, + "loss": 0.2919, + "step": 42032 + }, + { + "epoch": 2.916730275484005, + "grad_norm": 0.8799425363540649, + "learning_rate": 3.606333579468131e-05, + "loss": 0.3317, + "step": 42033 + }, + { + "epoch": 2.916799666921102, + "grad_norm": 1.2934197187423706, + "learning_rate": 3.6059015730890455e-05, + "loss": 0.3437, + "step": 42034 + }, + { + "epoch": 2.9168690583581984, + "grad_norm": 1.088354229927063, + "learning_rate": 3.605469586895387e-05, + "loss": 0.3075, + "step": 42035 + }, + { + "epoch": 2.9169384497952953, + "grad_norm": 1.033320426940918, + "learning_rate": 3.6050376208885186e-05, + "loss": 0.3885, + "step": 42036 + }, + { + "epoch": 2.917007841232392, + "grad_norm": 1.1472870111465454, + "learning_rate": 3.604605675069804e-05, + "loss": 0.4941, + "step": 42037 + }, + { + "epoch": 2.9170772326694885, + "grad_norm": 1.6026928424835205, + "learning_rate": 3.604173749440608e-05, + "loss": 0.3498, + "step": 42038 + }, + { + "epoch": 2.9171466241065853, + "grad_norm": 1.039086937904358, + "learning_rate": 3.6037418440022975e-05, + "loss": 0.3034, + "step": 42039 + }, + { + "epoch": 2.9172160155436817, + "grad_norm": 0.8841633796691895, + "learning_rate": 3.603309958756228e-05, + "loss": 0.3154, + "step": 42040 + }, + { + "epoch": 2.9172854069807785, + "grad_norm": 1.1228067874908447, + "learning_rate": 3.602878093703767e-05, + "loss": 0.3689, + "step": 42041 + }, + { + "epoch": 2.9173547984178754, + "grad_norm": 0.9646426439285278, + "learning_rate": 3.602446248846276e-05, + "loss": 0.3668, + "step": 42042 + }, + { + "epoch": 2.9174241898549718, + "grad_norm": 1.0151866674423218, + "learning_rate": 3.6020144241851215e-05, + "loss": 0.3071, + "step": 42043 + }, + { + "epoch": 2.9174935812920686, + "grad_norm": 1.1049214601516724, + "learning_rate": 3.601582619721665e-05, + "loss": 0.3647, + "step": 42044 + }, + { + "epoch": 2.917562972729165, + "grad_norm": 1.3957289457321167, + "learning_rate": 3.601150835457273e-05, + "loss": 0.3577, + "step": 42045 + }, + { + "epoch": 2.917632364166262, + "grad_norm": 1.1380436420440674, + "learning_rate": 3.6007190713933025e-05, + "loss": 0.3853, + "step": 42046 + }, + { + "epoch": 2.9177017556033586, + "grad_norm": 1.0948035717010498, + "learning_rate": 3.6002873275311187e-05, + "loss": 0.4593, + "step": 42047 + }, + { + "epoch": 2.9177711470404555, + "grad_norm": 1.401071548461914, + "learning_rate": 3.599855603872087e-05, + "loss": 0.3915, + "step": 42048 + }, + { + "epoch": 2.917840538477552, + "grad_norm": 1.164353847503662, + "learning_rate": 3.59942390041757e-05, + "loss": 0.3355, + "step": 42049 + }, + { + "epoch": 2.9179099299146483, + "grad_norm": 1.6103315353393555, + "learning_rate": 3.598992217168926e-05, + "loss": 0.5401, + "step": 42050 + }, + { + "epoch": 2.917979321351745, + "grad_norm": 1.654023289680481, + "learning_rate": 3.598560554127521e-05, + "loss": 0.43, + "step": 42051 + }, + { + "epoch": 2.918048712788842, + "grad_norm": 1.0222139358520508, + "learning_rate": 3.598128911294719e-05, + "loss": 0.4873, + "step": 42052 + }, + { + "epoch": 2.9181181042259388, + "grad_norm": 1.2145309448242188, + "learning_rate": 3.597697288671884e-05, + "loss": 0.365, + "step": 42053 + }, + { + "epoch": 2.918187495663035, + "grad_norm": 1.177183985710144, + "learning_rate": 3.597265686260372e-05, + "loss": 0.3087, + "step": 42054 + }, + { + "epoch": 2.918256887100132, + "grad_norm": 1.2196937799453735, + "learning_rate": 3.596834104061549e-05, + "loss": 0.4614, + "step": 42055 + }, + { + "epoch": 2.9183262785372284, + "grad_norm": 1.333876371383667, + "learning_rate": 3.596402542076779e-05, + "loss": 0.3299, + "step": 42056 + }, + { + "epoch": 2.918395669974325, + "grad_norm": 0.9053855538368225, + "learning_rate": 3.595971000307422e-05, + "loss": 0.2422, + "step": 42057 + }, + { + "epoch": 2.918465061411422, + "grad_norm": 1.457223892211914, + "learning_rate": 3.595539478754843e-05, + "loss": 0.6141, + "step": 42058 + }, + { + "epoch": 2.9185344528485184, + "grad_norm": 1.1114879846572876, + "learning_rate": 3.595107977420404e-05, + "loss": 0.3717, + "step": 42059 + }, + { + "epoch": 2.9186038442856153, + "grad_norm": 1.0402677059173584, + "learning_rate": 3.594676496305465e-05, + "loss": 0.3299, + "step": 42060 + }, + { + "epoch": 2.9186732357227116, + "grad_norm": 1.4521195888519287, + "learning_rate": 3.5942450354113874e-05, + "loss": 0.3947, + "step": 42061 + }, + { + "epoch": 2.9187426271598085, + "grad_norm": 1.2090356349945068, + "learning_rate": 3.593813594739536e-05, + "loss": 0.3864, + "step": 42062 + }, + { + "epoch": 2.9188120185969053, + "grad_norm": 1.0592361688613892, + "learning_rate": 3.593382174291271e-05, + "loss": 0.2741, + "step": 42063 + }, + { + "epoch": 2.9188814100340017, + "grad_norm": 0.9938839077949524, + "learning_rate": 3.592950774067956e-05, + "loss": 0.185, + "step": 42064 + }, + { + "epoch": 2.9189508014710985, + "grad_norm": 0.9092065691947937, + "learning_rate": 3.592519394070951e-05, + "loss": 0.3061, + "step": 42065 + }, + { + "epoch": 2.919020192908195, + "grad_norm": 1.0446434020996094, + "learning_rate": 3.592088034301624e-05, + "loss": 0.2955, + "step": 42066 + }, + { + "epoch": 2.9190895843452918, + "grad_norm": 1.0016660690307617, + "learning_rate": 3.591656694761326e-05, + "loss": 0.2256, + "step": 42067 + }, + { + "epoch": 2.9191589757823886, + "grad_norm": 1.09017813205719, + "learning_rate": 3.591225375451426e-05, + "loss": 0.5025, + "step": 42068 + }, + { + "epoch": 2.919228367219485, + "grad_norm": 1.1971310377120972, + "learning_rate": 3.590794076373284e-05, + "loss": 0.3011, + "step": 42069 + }, + { + "epoch": 2.919297758656582, + "grad_norm": 1.4097919464111328, + "learning_rate": 3.5903627975282606e-05, + "loss": 0.393, + "step": 42070 + }, + { + "epoch": 2.919367150093678, + "grad_norm": 1.1196907758712769, + "learning_rate": 3.589931538917718e-05, + "loss": 0.3878, + "step": 42071 + }, + { + "epoch": 2.919436541530775, + "grad_norm": 1.3169454336166382, + "learning_rate": 3.589500300543018e-05, + "loss": 0.2665, + "step": 42072 + }, + { + "epoch": 2.919505932967872, + "grad_norm": 1.4116846323013306, + "learning_rate": 3.589069082405526e-05, + "loss": 0.6756, + "step": 42073 + }, + { + "epoch": 2.9195753244049687, + "grad_norm": 1.1123254299163818, + "learning_rate": 3.588637884506596e-05, + "loss": 0.2698, + "step": 42074 + }, + { + "epoch": 2.919644715842065, + "grad_norm": 1.3220537900924683, + "learning_rate": 3.588206706847591e-05, + "loss": 0.5436, + "step": 42075 + }, + { + "epoch": 2.9197141072791615, + "grad_norm": 1.1000909805297852, + "learning_rate": 3.587775549429875e-05, + "loss": 0.3992, + "step": 42076 + }, + { + "epoch": 2.9197834987162583, + "grad_norm": 1.1870776414871216, + "learning_rate": 3.5873444122548064e-05, + "loss": 0.4289, + "step": 42077 + }, + { + "epoch": 2.919852890153355, + "grad_norm": 0.9643579721450806, + "learning_rate": 3.5869132953237475e-05, + "loss": 0.3056, + "step": 42078 + }, + { + "epoch": 2.919922281590452, + "grad_norm": 3.013302803039551, + "learning_rate": 3.5864821986380594e-05, + "loss": 0.3461, + "step": 42079 + }, + { + "epoch": 2.9199916730275484, + "grad_norm": 1.831779956817627, + "learning_rate": 3.586051122199107e-05, + "loss": 0.7037, + "step": 42080 + }, + { + "epoch": 2.920061064464645, + "grad_norm": 1.2422966957092285, + "learning_rate": 3.5856200660082426e-05, + "loss": 0.4341, + "step": 42081 + }, + { + "epoch": 2.9201304559017416, + "grad_norm": 1.273627519607544, + "learning_rate": 3.5851890300668325e-05, + "loss": 0.3285, + "step": 42082 + }, + { + "epoch": 2.9201998473388384, + "grad_norm": 1.263713002204895, + "learning_rate": 3.584758014376236e-05, + "loss": 0.3385, + "step": 42083 + }, + { + "epoch": 2.9202692387759352, + "grad_norm": 0.9047614932060242, + "learning_rate": 3.584327018937818e-05, + "loss": 0.2548, + "step": 42084 + }, + { + "epoch": 2.9203386302130316, + "grad_norm": 1.3328994512557983, + "learning_rate": 3.58389604375293e-05, + "loss": 0.3293, + "step": 42085 + }, + { + "epoch": 2.9204080216501285, + "grad_norm": 1.0669535398483276, + "learning_rate": 3.583465088822939e-05, + "loss": 0.3894, + "step": 42086 + }, + { + "epoch": 2.920477413087225, + "grad_norm": 1.2024065256118774, + "learning_rate": 3.5830341541492083e-05, + "loss": 0.297, + "step": 42087 + }, + { + "epoch": 2.9205468045243217, + "grad_norm": 1.310713291168213, + "learning_rate": 3.58260323973309e-05, + "loss": 0.542, + "step": 42088 + }, + { + "epoch": 2.9206161959614185, + "grad_norm": 1.0122884511947632, + "learning_rate": 3.5821723455759493e-05, + "loss": 0.3395, + "step": 42089 + }, + { + "epoch": 2.920685587398515, + "grad_norm": 1.331133484840393, + "learning_rate": 3.581741471679145e-05, + "loss": 0.4837, + "step": 42090 + }, + { + "epoch": 2.9207549788356117, + "grad_norm": 1.0971037149429321, + "learning_rate": 3.5813106180440394e-05, + "loss": 0.3208, + "step": 42091 + }, + { + "epoch": 2.920824370272708, + "grad_norm": 1.672120213508606, + "learning_rate": 3.580879784671991e-05, + "loss": 0.4613, + "step": 42092 + }, + { + "epoch": 2.920893761709805, + "grad_norm": 1.3773880004882812, + "learning_rate": 3.58044897156436e-05, + "loss": 0.4499, + "step": 42093 + }, + { + "epoch": 2.920963153146902, + "grad_norm": 1.5566489696502686, + "learning_rate": 3.580018178722511e-05, + "loss": 0.517, + "step": 42094 + }, + { + "epoch": 2.921032544583998, + "grad_norm": 1.131285548210144, + "learning_rate": 3.579587406147795e-05, + "loss": 0.3214, + "step": 42095 + }, + { + "epoch": 2.921101936021095, + "grad_norm": 1.31355881690979, + "learning_rate": 3.579156653841578e-05, + "loss": 0.6325, + "step": 42096 + }, + { + "epoch": 2.9211713274581914, + "grad_norm": 1.0962612628936768, + "learning_rate": 3.578725921805217e-05, + "loss": 0.3939, + "step": 42097 + }, + { + "epoch": 2.9212407188952882, + "grad_norm": 1.385300636291504, + "learning_rate": 3.578295210040075e-05, + "loss": 0.3025, + "step": 42098 + }, + { + "epoch": 2.921310110332385, + "grad_norm": 0.9776261448860168, + "learning_rate": 3.577864518547509e-05, + "loss": 0.3195, + "step": 42099 + }, + { + "epoch": 2.921379501769482, + "grad_norm": 1.0293573141098022, + "learning_rate": 3.577433847328879e-05, + "loss": 0.3055, + "step": 42100 + }, + { + "epoch": 2.9214488932065783, + "grad_norm": 0.9773455858230591, + "learning_rate": 3.57700319638555e-05, + "loss": 0.3027, + "step": 42101 + }, + { + "epoch": 2.921518284643675, + "grad_norm": 1.6252520084381104, + "learning_rate": 3.576572565718873e-05, + "loss": 0.6153, + "step": 42102 + }, + { + "epoch": 2.9215876760807715, + "grad_norm": 1.3749758005142212, + "learning_rate": 3.576141955330211e-05, + "loss": 0.2963, + "step": 42103 + }, + { + "epoch": 2.9216570675178684, + "grad_norm": 0.942101001739502, + "learning_rate": 3.575711365220923e-05, + "loss": 0.1763, + "step": 42104 + }, + { + "epoch": 2.921726458954965, + "grad_norm": 1.2808339595794678, + "learning_rate": 3.5752807953923694e-05, + "loss": 0.3047, + "step": 42105 + }, + { + "epoch": 2.9217958503920616, + "grad_norm": 1.1559311151504517, + "learning_rate": 3.574850245845909e-05, + "loss": 0.283, + "step": 42106 + }, + { + "epoch": 2.9218652418291584, + "grad_norm": 0.7149419188499451, + "learning_rate": 3.5744197165828994e-05, + "loss": 0.2841, + "step": 42107 + }, + { + "epoch": 2.921934633266255, + "grad_norm": 1.3365235328674316, + "learning_rate": 3.573989207604706e-05, + "loss": 0.438, + "step": 42108 + }, + { + "epoch": 2.9220040247033516, + "grad_norm": 1.0809452533721924, + "learning_rate": 3.5735587189126806e-05, + "loss": 0.4713, + "step": 42109 + }, + { + "epoch": 2.9220734161404485, + "grad_norm": 1.0998430252075195, + "learning_rate": 3.573128250508183e-05, + "loss": 0.4279, + "step": 42110 + }, + { + "epoch": 2.922142807577545, + "grad_norm": 1.3531394004821777, + "learning_rate": 3.572697802392575e-05, + "loss": 0.4604, + "step": 42111 + }, + { + "epoch": 2.9222121990146417, + "grad_norm": 1.6631802320480347, + "learning_rate": 3.572267374567214e-05, + "loss": 0.443, + "step": 42112 + }, + { + "epoch": 2.922281590451738, + "grad_norm": 1.1459523439407349, + "learning_rate": 3.571836967033459e-05, + "loss": 0.3257, + "step": 42113 + }, + { + "epoch": 2.922350981888835, + "grad_norm": 1.0135198831558228, + "learning_rate": 3.571406579792669e-05, + "loss": 0.2848, + "step": 42114 + }, + { + "epoch": 2.9224203733259317, + "grad_norm": 1.3152425289154053, + "learning_rate": 3.5709762128462065e-05, + "loss": 0.4424, + "step": 42115 + }, + { + "epoch": 2.922489764763028, + "grad_norm": 1.1552784442901611, + "learning_rate": 3.570545866195422e-05, + "loss": 0.395, + "step": 42116 + }, + { + "epoch": 2.922559156200125, + "grad_norm": 0.8640164136886597, + "learning_rate": 3.570115539841679e-05, + "loss": 0.3073, + "step": 42117 + }, + { + "epoch": 2.9226285476372214, + "grad_norm": 0.6853345036506653, + "learning_rate": 3.569685233786337e-05, + "loss": 0.1855, + "step": 42118 + }, + { + "epoch": 2.922697939074318, + "grad_norm": 1.0116181373596191, + "learning_rate": 3.5692549480307503e-05, + "loss": 0.321, + "step": 42119 + }, + { + "epoch": 2.922767330511415, + "grad_norm": 1.2824454307556152, + "learning_rate": 3.5688246825762806e-05, + "loss": 0.626, + "step": 42120 + }, + { + "epoch": 2.9228367219485114, + "grad_norm": 1.390916109085083, + "learning_rate": 3.568394437424284e-05, + "loss": 0.4881, + "step": 42121 + }, + { + "epoch": 2.9229061133856082, + "grad_norm": 1.9835295677185059, + "learning_rate": 3.5679642125761236e-05, + "loss": 0.4379, + "step": 42122 + }, + { + "epoch": 2.9229755048227046, + "grad_norm": 0.8961754441261292, + "learning_rate": 3.567534008033151e-05, + "loss": 0.3875, + "step": 42123 + }, + { + "epoch": 2.9230448962598015, + "grad_norm": 1.6199575662612915, + "learning_rate": 3.567103823796727e-05, + "loss": 0.5027, + "step": 42124 + }, + { + "epoch": 2.9231142876968983, + "grad_norm": 1.1613047122955322, + "learning_rate": 3.5666736598682095e-05, + "loss": 0.5018, + "step": 42125 + }, + { + "epoch": 2.923183679133995, + "grad_norm": 1.2971373796463013, + "learning_rate": 3.566243516248957e-05, + "loss": 0.3969, + "step": 42126 + }, + { + "epoch": 2.9232530705710915, + "grad_norm": 0.7510924935340881, + "learning_rate": 3.565813392940327e-05, + "loss": 0.2071, + "step": 42127 + }, + { + "epoch": 2.9233224620081883, + "grad_norm": 1.0211260318756104, + "learning_rate": 3.565383289943678e-05, + "loss": 0.3438, + "step": 42128 + }, + { + "epoch": 2.9233918534452847, + "grad_norm": 1.0111799240112305, + "learning_rate": 3.5649532072603706e-05, + "loss": 0.3692, + "step": 42129 + }, + { + "epoch": 2.9234612448823816, + "grad_norm": 1.2842752933502197, + "learning_rate": 3.564523144891756e-05, + "loss": 0.4068, + "step": 42130 + }, + { + "epoch": 2.9235306363194784, + "grad_norm": 1.1503819227218628, + "learning_rate": 3.564093102839195e-05, + "loss": 0.3513, + "step": 42131 + }, + { + "epoch": 2.923600027756575, + "grad_norm": 0.8792408108711243, + "learning_rate": 3.563663081104046e-05, + "loss": 0.2433, + "step": 42132 + }, + { + "epoch": 2.9236694191936716, + "grad_norm": 1.3637844324111938, + "learning_rate": 3.563233079687665e-05, + "loss": 0.3535, + "step": 42133 + }, + { + "epoch": 2.923738810630768, + "grad_norm": 1.3906974792480469, + "learning_rate": 3.5628030985914116e-05, + "loss": 0.5222, + "step": 42134 + }, + { + "epoch": 2.923808202067865, + "grad_norm": 1.3866252899169922, + "learning_rate": 3.5623731378166445e-05, + "loss": 0.4448, + "step": 42135 + }, + { + "epoch": 2.9238775935049617, + "grad_norm": 0.8310904502868652, + "learning_rate": 3.561943197364715e-05, + "loss": 0.3938, + "step": 42136 + }, + { + "epoch": 2.923946984942058, + "grad_norm": 1.228103756904602, + "learning_rate": 3.5615132772369844e-05, + "loss": 0.4683, + "step": 42137 + }, + { + "epoch": 2.924016376379155, + "grad_norm": 1.146423101425171, + "learning_rate": 3.561083377434809e-05, + "loss": 0.394, + "step": 42138 + }, + { + "epoch": 2.9240857678162513, + "grad_norm": 1.3008508682250977, + "learning_rate": 3.5606534979595465e-05, + "loss": 0.3346, + "step": 42139 + }, + { + "epoch": 2.924155159253348, + "grad_norm": 1.3854120969772339, + "learning_rate": 3.560223638812554e-05, + "loss": 0.4553, + "step": 42140 + }, + { + "epoch": 2.924224550690445, + "grad_norm": 1.6177078485488892, + "learning_rate": 3.559793799995188e-05, + "loss": 0.4912, + "step": 42141 + }, + { + "epoch": 2.9242939421275413, + "grad_norm": 0.9376275539398193, + "learning_rate": 3.5593639815088086e-05, + "loss": 0.2259, + "step": 42142 + }, + { + "epoch": 2.924363333564638, + "grad_norm": 1.0782259702682495, + "learning_rate": 3.5589341833547675e-05, + "loss": 0.3088, + "step": 42143 + }, + { + "epoch": 2.9244327250017346, + "grad_norm": 0.7000982165336609, + "learning_rate": 3.558504405534423e-05, + "loss": 0.2715, + "step": 42144 + }, + { + "epoch": 2.9245021164388314, + "grad_norm": 1.06236732006073, + "learning_rate": 3.5580746480491334e-05, + "loss": 0.4838, + "step": 42145 + }, + { + "epoch": 2.9245715078759282, + "grad_norm": 1.200992465019226, + "learning_rate": 3.557644910900255e-05, + "loss": 0.3886, + "step": 42146 + }, + { + "epoch": 2.9246408993130246, + "grad_norm": 1.4305680990219116, + "learning_rate": 3.557215194089144e-05, + "loss": 0.4207, + "step": 42147 + }, + { + "epoch": 2.9247102907501215, + "grad_norm": 0.8789859414100647, + "learning_rate": 3.5567854976171566e-05, + "loss": 0.3125, + "step": 42148 + }, + { + "epoch": 2.924779682187218, + "grad_norm": 1.1147750616073608, + "learning_rate": 3.556355821485653e-05, + "loss": 0.3793, + "step": 42149 + }, + { + "epoch": 2.9248490736243147, + "grad_norm": 1.0338423252105713, + "learning_rate": 3.555926165695984e-05, + "loss": 0.3886, + "step": 42150 + }, + { + "epoch": 2.9249184650614115, + "grad_norm": 0.7231321334838867, + "learning_rate": 3.555496530249508e-05, + "loss": 0.3392, + "step": 42151 + }, + { + "epoch": 2.9249878564985083, + "grad_norm": 4.1779351234436035, + "learning_rate": 3.555066915147581e-05, + "loss": 0.5488, + "step": 42152 + }, + { + "epoch": 2.9250572479356047, + "grad_norm": 1.5285069942474365, + "learning_rate": 3.554637320391565e-05, + "loss": 0.3564, + "step": 42153 + }, + { + "epoch": 2.9251266393727016, + "grad_norm": 1.1963673830032349, + "learning_rate": 3.554207745982806e-05, + "loss": 0.3467, + "step": 42154 + }, + { + "epoch": 2.925196030809798, + "grad_norm": 0.7774038910865784, + "learning_rate": 3.553778191922666e-05, + "loss": 0.2236, + "step": 42155 + }, + { + "epoch": 2.925265422246895, + "grad_norm": 0.8667685985565186, + "learning_rate": 3.553348658212501e-05, + "loss": 0.2855, + "step": 42156 + }, + { + "epoch": 2.9253348136839916, + "grad_norm": 1.0584582090377808, + "learning_rate": 3.552919144853669e-05, + "loss": 0.4635, + "step": 42157 + }, + { + "epoch": 2.925404205121088, + "grad_norm": 0.9230299592018127, + "learning_rate": 3.55248965184752e-05, + "loss": 0.4114, + "step": 42158 + }, + { + "epoch": 2.925473596558185, + "grad_norm": 1.2612777948379517, + "learning_rate": 3.5520601791954145e-05, + "loss": 0.3646, + "step": 42159 + }, + { + "epoch": 2.9255429879952812, + "grad_norm": 1.269404411315918, + "learning_rate": 3.5516307268987045e-05, + "loss": 0.358, + "step": 42160 + }, + { + "epoch": 2.925612379432378, + "grad_norm": 1.0949163436889648, + "learning_rate": 3.55120129495875e-05, + "loss": 0.4938, + "step": 42161 + }, + { + "epoch": 2.925681770869475, + "grad_norm": 1.2436176538467407, + "learning_rate": 3.550771883376904e-05, + "loss": 0.4536, + "step": 42162 + }, + { + "epoch": 2.9257511623065713, + "grad_norm": 1.047554612159729, + "learning_rate": 3.550342492154527e-05, + "loss": 0.2103, + "step": 42163 + }, + { + "epoch": 2.925820553743668, + "grad_norm": 1.2947297096252441, + "learning_rate": 3.549913121292966e-05, + "loss": 0.5797, + "step": 42164 + }, + { + "epoch": 2.9258899451807645, + "grad_norm": 1.3243160247802734, + "learning_rate": 3.549483770793581e-05, + "loss": 0.4812, + "step": 42165 + }, + { + "epoch": 2.9259593366178613, + "grad_norm": 1.1651421785354614, + "learning_rate": 3.549054440657729e-05, + "loss": 0.4035, + "step": 42166 + }, + { + "epoch": 2.926028728054958, + "grad_norm": 0.7499014139175415, + "learning_rate": 3.548625130886762e-05, + "loss": 0.3872, + "step": 42167 + }, + { + "epoch": 2.9260981194920546, + "grad_norm": 0.8403280377388, + "learning_rate": 3.548195841482037e-05, + "loss": 0.2684, + "step": 42168 + }, + { + "epoch": 2.9261675109291514, + "grad_norm": 1.244762659072876, + "learning_rate": 3.547766572444909e-05, + "loss": 0.4064, + "step": 42169 + }, + { + "epoch": 2.926236902366248, + "grad_norm": 0.9117634296417236, + "learning_rate": 3.5473373237767384e-05, + "loss": 0.2475, + "step": 42170 + }, + { + "epoch": 2.9263062938033446, + "grad_norm": 1.2512747049331665, + "learning_rate": 3.5469080954788705e-05, + "loss": 0.399, + "step": 42171 + }, + { + "epoch": 2.9263756852404414, + "grad_norm": 1.2799618244171143, + "learning_rate": 3.5464788875526655e-05, + "loss": 0.3402, + "step": 42172 + }, + { + "epoch": 2.926445076677538, + "grad_norm": 0.8132128119468689, + "learning_rate": 3.546049699999478e-05, + "loss": 0.2869, + "step": 42173 + }, + { + "epoch": 2.9265144681146347, + "grad_norm": 1.4896050691604614, + "learning_rate": 3.545620532820663e-05, + "loss": 0.4744, + "step": 42174 + }, + { + "epoch": 2.926583859551731, + "grad_norm": 0.9053963422775269, + "learning_rate": 3.5451913860175735e-05, + "loss": 0.2938, + "step": 42175 + }, + { + "epoch": 2.926653250988828, + "grad_norm": 1.206382155418396, + "learning_rate": 3.5447622595915674e-05, + "loss": 0.3435, + "step": 42176 + }, + { + "epoch": 2.9267226424259247, + "grad_norm": 0.8718389272689819, + "learning_rate": 3.5443331535440004e-05, + "loss": 0.3315, + "step": 42177 + }, + { + "epoch": 2.9267920338630216, + "grad_norm": 1.4354523420333862, + "learning_rate": 3.543904067876222e-05, + "loss": 0.5367, + "step": 42178 + }, + { + "epoch": 2.926861425300118, + "grad_norm": 1.0891697406768799, + "learning_rate": 3.543475002589589e-05, + "loss": 0.2616, + "step": 42179 + }, + { + "epoch": 2.926930816737215, + "grad_norm": 1.0066282749176025, + "learning_rate": 3.543045957685457e-05, + "loss": 0.3692, + "step": 42180 + }, + { + "epoch": 2.927000208174311, + "grad_norm": 1.0687202215194702, + "learning_rate": 3.542616933165179e-05, + "loss": 0.2933, + "step": 42181 + }, + { + "epoch": 2.927069599611408, + "grad_norm": 1.2386736869812012, + "learning_rate": 3.5421879290301105e-05, + "loss": 0.379, + "step": 42182 + }, + { + "epoch": 2.927138991048505, + "grad_norm": 1.4824588298797607, + "learning_rate": 3.5417589452816056e-05, + "loss": 0.4039, + "step": 42183 + }, + { + "epoch": 2.9272083824856012, + "grad_norm": 1.0211015939712524, + "learning_rate": 3.541329981921021e-05, + "loss": 0.3599, + "step": 42184 + }, + { + "epoch": 2.927277773922698, + "grad_norm": 0.8636837005615234, + "learning_rate": 3.5409010389497055e-05, + "loss": 0.2738, + "step": 42185 + }, + { + "epoch": 2.9273471653597944, + "grad_norm": 1.679226040840149, + "learning_rate": 3.540472116369017e-05, + "loss": 0.4683, + "step": 42186 + }, + { + "epoch": 2.9274165567968913, + "grad_norm": 1.2371273040771484, + "learning_rate": 3.5400432141803075e-05, + "loss": 0.4028, + "step": 42187 + }, + { + "epoch": 2.927485948233988, + "grad_norm": 0.950347900390625, + "learning_rate": 3.539614332384935e-05, + "loss": 0.3378, + "step": 42188 + }, + { + "epoch": 2.9275553396710845, + "grad_norm": 1.4311999082565308, + "learning_rate": 3.539185470984247e-05, + "loss": 0.4523, + "step": 42189 + }, + { + "epoch": 2.9276247311081813, + "grad_norm": 1.1406995058059692, + "learning_rate": 3.538756629979603e-05, + "loss": 0.2637, + "step": 42190 + }, + { + "epoch": 2.9276941225452777, + "grad_norm": 0.8183029890060425, + "learning_rate": 3.538327809372353e-05, + "loss": 0.3566, + "step": 42191 + }, + { + "epoch": 2.9277635139823746, + "grad_norm": 1.1399837732315063, + "learning_rate": 3.5378990091638564e-05, + "loss": 0.3183, + "step": 42192 + }, + { + "epoch": 2.9278329054194714, + "grad_norm": 1.507830023765564, + "learning_rate": 3.5374702293554586e-05, + "loss": 0.5292, + "step": 42193 + }, + { + "epoch": 2.9279022968565678, + "grad_norm": 1.4150886535644531, + "learning_rate": 3.5370414699485176e-05, + "loss": 0.3414, + "step": 42194 + }, + { + "epoch": 2.9279716882936646, + "grad_norm": 0.8941318392753601, + "learning_rate": 3.536612730944388e-05, + "loss": 0.3781, + "step": 42195 + }, + { + "epoch": 2.928041079730761, + "grad_norm": 1.2162880897521973, + "learning_rate": 3.536184012344421e-05, + "loss": 0.3416, + "step": 42196 + }, + { + "epoch": 2.928110471167858, + "grad_norm": 1.0141664743423462, + "learning_rate": 3.535755314149972e-05, + "loss": 0.5121, + "step": 42197 + }, + { + "epoch": 2.9281798626049547, + "grad_norm": 1.789412498474121, + "learning_rate": 3.535326636362396e-05, + "loss": 0.2714, + "step": 42198 + }, + { + "epoch": 2.928249254042051, + "grad_norm": 1.0392820835113525, + "learning_rate": 3.5348979789830406e-05, + "loss": 0.3235, + "step": 42199 + }, + { + "epoch": 2.928318645479148, + "grad_norm": 1.1545560359954834, + "learning_rate": 3.5344693420132626e-05, + "loss": 0.3542, + "step": 42200 + }, + { + "epoch": 2.9283880369162443, + "grad_norm": 1.2411272525787354, + "learning_rate": 3.534040725454415e-05, + "loss": 0.2548, + "step": 42201 + }, + { + "epoch": 2.928457428353341, + "grad_norm": 1.4502702951431274, + "learning_rate": 3.5336121293078495e-05, + "loss": 0.4658, + "step": 42202 + }, + { + "epoch": 2.928526819790438, + "grad_norm": 0.999503493309021, + "learning_rate": 3.5331835535749215e-05, + "loss": 0.3151, + "step": 42203 + }, + { + "epoch": 2.9285962112275348, + "grad_norm": 1.1009950637817383, + "learning_rate": 3.532754998256982e-05, + "loss": 0.3159, + "step": 42204 + }, + { + "epoch": 2.928665602664631, + "grad_norm": 1.0080569982528687, + "learning_rate": 3.5323264633553874e-05, + "loss": 0.2212, + "step": 42205 + }, + { + "epoch": 2.928734994101728, + "grad_norm": 1.222038984298706, + "learning_rate": 3.531897948871486e-05, + "loss": 0.4137, + "step": 42206 + }, + { + "epoch": 2.9288043855388244, + "grad_norm": 2.4513912200927734, + "learning_rate": 3.531469454806632e-05, + "loss": 0.3375, + "step": 42207 + }, + { + "epoch": 2.928873776975921, + "grad_norm": 1.2850512266159058, + "learning_rate": 3.531040981162177e-05, + "loss": 0.4417, + "step": 42208 + }, + { + "epoch": 2.928943168413018, + "grad_norm": 1.1019363403320312, + "learning_rate": 3.5306125279394765e-05, + "loss": 0.4537, + "step": 42209 + }, + { + "epoch": 2.9290125598501144, + "grad_norm": 0.7258356213569641, + "learning_rate": 3.5301840951398815e-05, + "loss": 0.2436, + "step": 42210 + }, + { + "epoch": 2.9290819512872113, + "grad_norm": 1.3807090520858765, + "learning_rate": 3.529755682764747e-05, + "loss": 0.6455, + "step": 42211 + }, + { + "epoch": 2.9291513427243077, + "grad_norm": 1.3069584369659424, + "learning_rate": 3.529327290815421e-05, + "loss": 0.316, + "step": 42212 + }, + { + "epoch": 2.9292207341614045, + "grad_norm": 5.058200359344482, + "learning_rate": 3.5288989192932574e-05, + "loss": 0.4042, + "step": 42213 + }, + { + "epoch": 2.9292901255985013, + "grad_norm": 2.301593780517578, + "learning_rate": 3.528470568199609e-05, + "loss": 0.4106, + "step": 42214 + }, + { + "epoch": 2.9293595170355977, + "grad_norm": 1.1505694389343262, + "learning_rate": 3.528042237535828e-05, + "loss": 0.3312, + "step": 42215 + }, + { + "epoch": 2.9294289084726945, + "grad_norm": 1.3056467771530151, + "learning_rate": 3.5276139273032674e-05, + "loss": 0.4128, + "step": 42216 + }, + { + "epoch": 2.929498299909791, + "grad_norm": 1.3221466541290283, + "learning_rate": 3.5271856375032785e-05, + "loss": 0.4677, + "step": 42217 + }, + { + "epoch": 2.9295676913468878, + "grad_norm": 0.9628995060920715, + "learning_rate": 3.526757368137217e-05, + "loss": 0.432, + "step": 42218 + }, + { + "epoch": 2.9296370827839846, + "grad_norm": 1.2190028429031372, + "learning_rate": 3.5263291192064274e-05, + "loss": 0.404, + "step": 42219 + }, + { + "epoch": 2.929706474221081, + "grad_norm": 1.2843050956726074, + "learning_rate": 3.525900890712267e-05, + "loss": 0.3749, + "step": 42220 + }, + { + "epoch": 2.929775865658178, + "grad_norm": 1.0753988027572632, + "learning_rate": 3.5254726826560856e-05, + "loss": 0.407, + "step": 42221 + }, + { + "epoch": 2.929845257095274, + "grad_norm": 1.4475926160812378, + "learning_rate": 3.525044495039236e-05, + "loss": 0.4461, + "step": 42222 + }, + { + "epoch": 2.929914648532371, + "grad_norm": 1.1905834674835205, + "learning_rate": 3.524616327863073e-05, + "loss": 0.3386, + "step": 42223 + }, + { + "epoch": 2.929984039969468, + "grad_norm": 1.5612155199050903, + "learning_rate": 3.524188181128941e-05, + "loss": 0.3077, + "step": 42224 + }, + { + "epoch": 2.9300534314065643, + "grad_norm": 1.1336960792541504, + "learning_rate": 3.5237600548381954e-05, + "loss": 0.3592, + "step": 42225 + }, + { + "epoch": 2.930122822843661, + "grad_norm": 0.8265459537506104, + "learning_rate": 3.523331948992189e-05, + "loss": 0.3405, + "step": 42226 + }, + { + "epoch": 2.9301922142807575, + "grad_norm": 1.0459097623825073, + "learning_rate": 3.522903863592275e-05, + "loss": 0.4387, + "step": 42227 + }, + { + "epoch": 2.9302616057178543, + "grad_norm": 1.1384145021438599, + "learning_rate": 3.522475798639799e-05, + "loss": 0.2778, + "step": 42228 + }, + { + "epoch": 2.930330997154951, + "grad_norm": 1.3635517358779907, + "learning_rate": 3.5220477541361155e-05, + "loss": 0.314, + "step": 42229 + }, + { + "epoch": 2.930400388592048, + "grad_norm": 1.1541513204574585, + "learning_rate": 3.5216197300825745e-05, + "loss": 0.454, + "step": 42230 + }, + { + "epoch": 2.9304697800291444, + "grad_norm": 1.084946632385254, + "learning_rate": 3.5211917264805295e-05, + "loss": 0.382, + "step": 42231 + }, + { + "epoch": 2.930539171466241, + "grad_norm": 1.058318018913269, + "learning_rate": 3.520763743331334e-05, + "loss": 0.3603, + "step": 42232 + }, + { + "epoch": 2.9306085629033376, + "grad_norm": 1.1808029413223267, + "learning_rate": 3.5203357806363316e-05, + "loss": 0.3348, + "step": 42233 + }, + { + "epoch": 2.9306779543404344, + "grad_norm": 1.2190021276474, + "learning_rate": 3.5199078383968776e-05, + "loss": 0.3529, + "step": 42234 + }, + { + "epoch": 2.9307473457775313, + "grad_norm": 1.1306103467941284, + "learning_rate": 3.5194799166143223e-05, + "loss": 0.4606, + "step": 42235 + }, + { + "epoch": 2.9308167372146277, + "grad_norm": 1.0794552564620972, + "learning_rate": 3.5190520152900176e-05, + "loss": 0.3799, + "step": 42236 + }, + { + "epoch": 2.9308861286517245, + "grad_norm": 1.0193568468093872, + "learning_rate": 3.5186241344253134e-05, + "loss": 0.4233, + "step": 42237 + }, + { + "epoch": 2.930955520088821, + "grad_norm": 0.9729064702987671, + "learning_rate": 3.5181962740215613e-05, + "loss": 0.372, + "step": 42238 + }, + { + "epoch": 2.9310249115259177, + "grad_norm": 1.016730785369873, + "learning_rate": 3.517768434080114e-05, + "loss": 0.3391, + "step": 42239 + }, + { + "epoch": 2.9310943029630145, + "grad_norm": 1.3604686260223389, + "learning_rate": 3.517340614602318e-05, + "loss": 0.3407, + "step": 42240 + }, + { + "epoch": 2.931163694400111, + "grad_norm": 1.032348871231079, + "learning_rate": 3.516912815589524e-05, + "loss": 0.3447, + "step": 42241 + }, + { + "epoch": 2.9312330858372078, + "grad_norm": 1.1967414617538452, + "learning_rate": 3.516485037043085e-05, + "loss": 0.3192, + "step": 42242 + }, + { + "epoch": 2.931302477274304, + "grad_norm": 1.0403298139572144, + "learning_rate": 3.51605727896435e-05, + "loss": 0.3329, + "step": 42243 + }, + { + "epoch": 2.931371868711401, + "grad_norm": 0.9156932234764099, + "learning_rate": 3.515629541354669e-05, + "loss": 0.3182, + "step": 42244 + }, + { + "epoch": 2.931441260148498, + "grad_norm": 1.0177569389343262, + "learning_rate": 3.515201824215395e-05, + "loss": 0.2468, + "step": 42245 + }, + { + "epoch": 2.931510651585594, + "grad_norm": 1.3506919145584106, + "learning_rate": 3.5147741275478785e-05, + "loss": 0.3796, + "step": 42246 + }, + { + "epoch": 2.931580043022691, + "grad_norm": 1.4009255170822144, + "learning_rate": 3.514346451353465e-05, + "loss": 0.4696, + "step": 42247 + }, + { + "epoch": 2.9316494344597874, + "grad_norm": 1.009835958480835, + "learning_rate": 3.513918795633508e-05, + "loss": 0.395, + "step": 42248 + }, + { + "epoch": 2.9317188258968843, + "grad_norm": 1.3341219425201416, + "learning_rate": 3.513491160389355e-05, + "loss": 0.3715, + "step": 42249 + }, + { + "epoch": 2.931788217333981, + "grad_norm": 1.0180169343948364, + "learning_rate": 3.5130635456223595e-05, + "loss": 0.3938, + "step": 42250 + }, + { + "epoch": 2.9318576087710775, + "grad_norm": 0.8701237440109253, + "learning_rate": 3.512635951333868e-05, + "loss": 0.2295, + "step": 42251 + }, + { + "epoch": 2.9319270002081743, + "grad_norm": 0.9302391409873962, + "learning_rate": 3.512208377525233e-05, + "loss": 0.3973, + "step": 42252 + }, + { + "epoch": 2.9319963916452707, + "grad_norm": 1.1459470987319946, + "learning_rate": 3.511780824197808e-05, + "loss": 0.4229, + "step": 42253 + }, + { + "epoch": 2.9320657830823675, + "grad_norm": 0.8940250873565674, + "learning_rate": 3.511353291352933e-05, + "loss": 0.2834, + "step": 42254 + }, + { + "epoch": 2.9321351745194644, + "grad_norm": 1.3315902948379517, + "learning_rate": 3.510925778991964e-05, + "loss": 0.2833, + "step": 42255 + }, + { + "epoch": 2.932204565956561, + "grad_norm": 0.9670912623405457, + "learning_rate": 3.5104982871162484e-05, + "loss": 0.3057, + "step": 42256 + }, + { + "epoch": 2.9322739573936576, + "grad_norm": 0.936872661113739, + "learning_rate": 3.5100708157271375e-05, + "loss": 0.3767, + "step": 42257 + }, + { + "epoch": 2.9323433488307544, + "grad_norm": 1.162376880645752, + "learning_rate": 3.509643364825984e-05, + "loss": 0.2856, + "step": 42258 + }, + { + "epoch": 2.932412740267851, + "grad_norm": 1.1049779653549194, + "learning_rate": 3.5092159344141284e-05, + "loss": 0.3942, + "step": 42259 + }, + { + "epoch": 2.9324821317049476, + "grad_norm": 1.492403507232666, + "learning_rate": 3.5087885244929266e-05, + "loss": 0.4924, + "step": 42260 + }, + { + "epoch": 2.9325515231420445, + "grad_norm": 1.3932946920394897, + "learning_rate": 3.508361135063729e-05, + "loss": 0.4255, + "step": 42261 + }, + { + "epoch": 2.932620914579141, + "grad_norm": 1.0572643280029297, + "learning_rate": 3.507933766127879e-05, + "loss": 0.4273, + "step": 42262 + }, + { + "epoch": 2.9326903060162377, + "grad_norm": 0.9769446849822998, + "learning_rate": 3.507506417686729e-05, + "loss": 0.2956, + "step": 42263 + }, + { + "epoch": 2.932759697453334, + "grad_norm": 1.2293448448181152, + "learning_rate": 3.507079089741627e-05, + "loss": 0.4436, + "step": 42264 + }, + { + "epoch": 2.932829088890431, + "grad_norm": 1.2452994585037231, + "learning_rate": 3.506651782293924e-05, + "loss": 0.4182, + "step": 42265 + }, + { + "epoch": 2.9328984803275278, + "grad_norm": 1.2574093341827393, + "learning_rate": 3.506224495344967e-05, + "loss": 0.3704, + "step": 42266 + }, + { + "epoch": 2.932967871764624, + "grad_norm": 1.2300524711608887, + "learning_rate": 3.505797228896111e-05, + "loss": 0.5145, + "step": 42267 + }, + { + "epoch": 2.933037263201721, + "grad_norm": 1.2026557922363281, + "learning_rate": 3.505369982948694e-05, + "loss": 0.3624, + "step": 42268 + }, + { + "epoch": 2.9331066546388174, + "grad_norm": 1.28340744972229, + "learning_rate": 3.504942757504072e-05, + "loss": 0.4476, + "step": 42269 + }, + { + "epoch": 2.933176046075914, + "grad_norm": 1.419566035270691, + "learning_rate": 3.504515552563592e-05, + "loss": 0.382, + "step": 42270 + }, + { + "epoch": 2.933245437513011, + "grad_norm": 0.9746535420417786, + "learning_rate": 3.504088368128602e-05, + "loss": 0.4349, + "step": 42271 + }, + { + "epoch": 2.9333148289501074, + "grad_norm": 1.2938446998596191, + "learning_rate": 3.503661204200452e-05, + "loss": 0.502, + "step": 42272 + }, + { + "epoch": 2.9333842203872043, + "grad_norm": 1.3593299388885498, + "learning_rate": 3.503234060780489e-05, + "loss": 0.4834, + "step": 42273 + }, + { + "epoch": 2.9334536118243006, + "grad_norm": 0.8468018770217896, + "learning_rate": 3.5028069378700665e-05, + "loss": 0.2929, + "step": 42274 + }, + { + "epoch": 2.9335230032613975, + "grad_norm": 1.0295263528823853, + "learning_rate": 3.502379835470526e-05, + "loss": 0.401, + "step": 42275 + }, + { + "epoch": 2.9335923946984943, + "grad_norm": 1.3915725946426392, + "learning_rate": 3.5019527535832175e-05, + "loss": 0.3283, + "step": 42276 + }, + { + "epoch": 2.933661786135591, + "grad_norm": 1.03326416015625, + "learning_rate": 3.501525692209491e-05, + "loss": 0.4683, + "step": 42277 + }, + { + "epoch": 2.9337311775726875, + "grad_norm": 1.1565948724746704, + "learning_rate": 3.501098651350693e-05, + "loss": 0.4493, + "step": 42278 + }, + { + "epoch": 2.933800569009784, + "grad_norm": 1.4748824834823608, + "learning_rate": 3.5006716310081735e-05, + "loss": 0.4093, + "step": 42279 + }, + { + "epoch": 2.9338699604468808, + "grad_norm": 0.9615399241447449, + "learning_rate": 3.500244631183278e-05, + "loss": 0.4264, + "step": 42280 + }, + { + "epoch": 2.9339393518839776, + "grad_norm": 1.3279041051864624, + "learning_rate": 3.499817651877361e-05, + "loss": 0.3312, + "step": 42281 + }, + { + "epoch": 2.9340087433210744, + "grad_norm": 0.8053078055381775, + "learning_rate": 3.499390693091762e-05, + "loss": 0.2139, + "step": 42282 + }, + { + "epoch": 2.934078134758171, + "grad_norm": 0.9040886759757996, + "learning_rate": 3.498963754827832e-05, + "loss": 0.2501, + "step": 42283 + }, + { + "epoch": 2.9341475261952676, + "grad_norm": 0.8881788849830627, + "learning_rate": 3.4985368370869186e-05, + "loss": 0.351, + "step": 42284 + }, + { + "epoch": 2.934216917632364, + "grad_norm": 1.3177740573883057, + "learning_rate": 3.498109939870371e-05, + "loss": 0.3959, + "step": 42285 + }, + { + "epoch": 2.934286309069461, + "grad_norm": 1.1168757677078247, + "learning_rate": 3.497683063179535e-05, + "loss": 0.3922, + "step": 42286 + }, + { + "epoch": 2.9343557005065577, + "grad_norm": 0.8907211422920227, + "learning_rate": 3.497256207015763e-05, + "loss": 0.2602, + "step": 42287 + }, + { + "epoch": 2.934425091943654, + "grad_norm": 1.237313151359558, + "learning_rate": 3.496829371380396e-05, + "loss": 0.4241, + "step": 42288 + }, + { + "epoch": 2.934494483380751, + "grad_norm": 1.0575523376464844, + "learning_rate": 3.496402556274784e-05, + "loss": 0.2108, + "step": 42289 + }, + { + "epoch": 2.9345638748178473, + "grad_norm": 0.980888307094574, + "learning_rate": 3.495975761700274e-05, + "loss": 0.5264, + "step": 42290 + }, + { + "epoch": 2.934633266254944, + "grad_norm": 0.9455319046974182, + "learning_rate": 3.4955489876582136e-05, + "loss": 0.3368, + "step": 42291 + }, + { + "epoch": 2.934702657692041, + "grad_norm": 1.3900609016418457, + "learning_rate": 3.495122234149951e-05, + "loss": 0.3211, + "step": 42292 + }, + { + "epoch": 2.9347720491291374, + "grad_norm": 1.1209691762924194, + "learning_rate": 3.494695501176837e-05, + "loss": 0.392, + "step": 42293 + }, + { + "epoch": 2.934841440566234, + "grad_norm": 1.2128212451934814, + "learning_rate": 3.4942687887402095e-05, + "loss": 0.2864, + "step": 42294 + }, + { + "epoch": 2.9349108320033306, + "grad_norm": 2.2161195278167725, + "learning_rate": 3.493842096841422e-05, + "loss": 0.5043, + "step": 42295 + }, + { + "epoch": 2.9349802234404274, + "grad_norm": 1.47017502784729, + "learning_rate": 3.493415425481823e-05, + "loss": 0.4405, + "step": 42296 + }, + { + "epoch": 2.9350496148775242, + "grad_norm": 1.3033872842788696, + "learning_rate": 3.492988774662753e-05, + "loss": 0.5356, + "step": 42297 + }, + { + "epoch": 2.9351190063146206, + "grad_norm": 1.1198270320892334, + "learning_rate": 3.4925621443855626e-05, + "loss": 0.3711, + "step": 42298 + }, + { + "epoch": 2.9351883977517175, + "grad_norm": 1.9056044816970825, + "learning_rate": 3.4921355346515986e-05, + "loss": 0.5631, + "step": 42299 + }, + { + "epoch": 2.935257789188814, + "grad_norm": 0.7981072664260864, + "learning_rate": 3.491708945462208e-05, + "loss": 0.3832, + "step": 42300 + }, + { + "epoch": 2.9353271806259107, + "grad_norm": 0.9609571695327759, + "learning_rate": 3.49128237681874e-05, + "loss": 0.3691, + "step": 42301 + }, + { + "epoch": 2.9353965720630075, + "grad_norm": 1.2217614650726318, + "learning_rate": 3.490855828722535e-05, + "loss": 0.5241, + "step": 42302 + }, + { + "epoch": 2.9354659635001044, + "grad_norm": 1.099647045135498, + "learning_rate": 3.490429301174942e-05, + "loss": 0.323, + "step": 42303 + }, + { + "epoch": 2.9355353549372007, + "grad_norm": 1.0784181356430054, + "learning_rate": 3.49000279417731e-05, + "loss": 0.4006, + "step": 42304 + }, + { + "epoch": 2.935604746374297, + "grad_norm": 0.8898319005966187, + "learning_rate": 3.489576307730984e-05, + "loss": 0.296, + "step": 42305 + }, + { + "epoch": 2.935674137811394, + "grad_norm": 1.131524920463562, + "learning_rate": 3.489149841837309e-05, + "loss": 0.3657, + "step": 42306 + }, + { + "epoch": 2.935743529248491, + "grad_norm": 1.429607629776001, + "learning_rate": 3.488723396497634e-05, + "loss": 0.4983, + "step": 42307 + }, + { + "epoch": 2.9358129206855876, + "grad_norm": 1.2916103601455688, + "learning_rate": 3.4882969717133065e-05, + "loss": 0.4829, + "step": 42308 + }, + { + "epoch": 2.935882312122684, + "grad_norm": 0.9903582334518433, + "learning_rate": 3.487870567485666e-05, + "loss": 0.3936, + "step": 42309 + }, + { + "epoch": 2.935951703559781, + "grad_norm": 1.018634557723999, + "learning_rate": 3.4874441838160634e-05, + "loss": 0.3062, + "step": 42310 + }, + { + "epoch": 2.9360210949968772, + "grad_norm": 1.1356605291366577, + "learning_rate": 3.4870178207058434e-05, + "loss": 0.4327, + "step": 42311 + }, + { + "epoch": 2.936090486433974, + "grad_norm": 1.2001216411590576, + "learning_rate": 3.486591478156353e-05, + "loss": 0.2476, + "step": 42312 + }, + { + "epoch": 2.936159877871071, + "grad_norm": 1.1707711219787598, + "learning_rate": 3.4861651561689365e-05, + "loss": 0.3794, + "step": 42313 + }, + { + "epoch": 2.9362292693081673, + "grad_norm": 1.2419543266296387, + "learning_rate": 3.485738854744942e-05, + "loss": 0.2961, + "step": 42314 + }, + { + "epoch": 2.936298660745264, + "grad_norm": 0.8993836641311646, + "learning_rate": 3.485312573885717e-05, + "loss": 0.3379, + "step": 42315 + }, + { + "epoch": 2.9363680521823605, + "grad_norm": 1.9663619995117188, + "learning_rate": 3.4848863135926016e-05, + "loss": 0.4117, + "step": 42316 + }, + { + "epoch": 2.9364374436194574, + "grad_norm": 1.4142907857894897, + "learning_rate": 3.484460073866943e-05, + "loss": 0.3815, + "step": 42317 + }, + { + "epoch": 2.936506835056554, + "grad_norm": 0.8859854340553284, + "learning_rate": 3.4840338547100894e-05, + "loss": 0.2906, + "step": 42318 + }, + { + "epoch": 2.9365762264936506, + "grad_norm": 0.983943521976471, + "learning_rate": 3.4836076561233844e-05, + "loss": 0.4291, + "step": 42319 + }, + { + "epoch": 2.9366456179307474, + "grad_norm": 1.4927525520324707, + "learning_rate": 3.483181478108174e-05, + "loss": 0.5962, + "step": 42320 + }, + { + "epoch": 2.936715009367844, + "grad_norm": 1.1886252164840698, + "learning_rate": 3.4827553206658024e-05, + "loss": 0.361, + "step": 42321 + }, + { + "epoch": 2.9367844008049406, + "grad_norm": 0.8963851928710938, + "learning_rate": 3.482329183797621e-05, + "loss": 0.2302, + "step": 42322 + }, + { + "epoch": 2.9368537922420375, + "grad_norm": 1.0115420818328857, + "learning_rate": 3.481903067504967e-05, + "loss": 0.5287, + "step": 42323 + }, + { + "epoch": 2.936923183679134, + "grad_norm": 1.306207299232483, + "learning_rate": 3.481476971789187e-05, + "loss": 0.3724, + "step": 42324 + }, + { + "epoch": 2.9369925751162307, + "grad_norm": 1.1544231176376343, + "learning_rate": 3.48105089665163e-05, + "loss": 0.5398, + "step": 42325 + }, + { + "epoch": 2.937061966553327, + "grad_norm": 1.1404175758361816, + "learning_rate": 3.4806248420936374e-05, + "loss": 0.3558, + "step": 42326 + }, + { + "epoch": 2.937131357990424, + "grad_norm": 1.067721962928772, + "learning_rate": 3.480198808116556e-05, + "loss": 0.3273, + "step": 42327 + }, + { + "epoch": 2.9372007494275207, + "grad_norm": 1.0051332712173462, + "learning_rate": 3.479772794721735e-05, + "loss": 0.4353, + "step": 42328 + }, + { + "epoch": 2.9372701408646176, + "grad_norm": 0.6665438413619995, + "learning_rate": 3.47934680191051e-05, + "loss": 0.2265, + "step": 42329 + }, + { + "epoch": 2.937339532301714, + "grad_norm": 0.8592153787612915, + "learning_rate": 3.4789208296842314e-05, + "loss": 0.3379, + "step": 42330 + }, + { + "epoch": 2.937408923738811, + "grad_norm": 1.2957217693328857, + "learning_rate": 3.4784948780442474e-05, + "loss": 0.5088, + "step": 42331 + }, + { + "epoch": 2.937478315175907, + "grad_norm": 0.7181663513183594, + "learning_rate": 3.478068946991894e-05, + "loss": 0.2536, + "step": 42332 + }, + { + "epoch": 2.937547706613004, + "grad_norm": 1.0369541645050049, + "learning_rate": 3.47764303652852e-05, + "loss": 0.3893, + "step": 42333 + }, + { + "epoch": 2.937617098050101, + "grad_norm": 0.7833779454231262, + "learning_rate": 3.4772171466554705e-05, + "loss": 0.2909, + "step": 42334 + }, + { + "epoch": 2.9376864894871972, + "grad_norm": 1.144585371017456, + "learning_rate": 3.4767912773740896e-05, + "loss": 0.4493, + "step": 42335 + }, + { + "epoch": 2.937755880924294, + "grad_norm": 0.9967533349990845, + "learning_rate": 3.476365428685725e-05, + "loss": 0.3239, + "step": 42336 + }, + { + "epoch": 2.9378252723613905, + "grad_norm": 0.8065508604049683, + "learning_rate": 3.4759396005917156e-05, + "loss": 0.3054, + "step": 42337 + }, + { + "epoch": 2.9378946637984873, + "grad_norm": 1.0502020120620728, + "learning_rate": 3.4755137930934064e-05, + "loss": 0.3991, + "step": 42338 + }, + { + "epoch": 2.937964055235584, + "grad_norm": 1.2597668170928955, + "learning_rate": 3.475088006192144e-05, + "loss": 0.4601, + "step": 42339 + }, + { + "epoch": 2.9380334466726805, + "grad_norm": 0.8169155120849609, + "learning_rate": 3.474662239889273e-05, + "loss": 0.2958, + "step": 42340 + }, + { + "epoch": 2.9381028381097773, + "grad_norm": 1.4600191116333008, + "learning_rate": 3.474236494186134e-05, + "loss": 0.4297, + "step": 42341 + }, + { + "epoch": 2.9381722295468737, + "grad_norm": 1.019065499305725, + "learning_rate": 3.473810769084074e-05, + "loss": 0.4648, + "step": 42342 + }, + { + "epoch": 2.9382416209839706, + "grad_norm": 1.6367605924606323, + "learning_rate": 3.47338506458444e-05, + "loss": 0.376, + "step": 42343 + }, + { + "epoch": 2.9383110124210674, + "grad_norm": 1.1862841844558716, + "learning_rate": 3.472959380688569e-05, + "loss": 0.45, + "step": 42344 + }, + { + "epoch": 2.938380403858164, + "grad_norm": 1.0644078254699707, + "learning_rate": 3.4725337173978066e-05, + "loss": 0.2903, + "step": 42345 + }, + { + "epoch": 2.9384497952952606, + "grad_norm": 1.1752499341964722, + "learning_rate": 3.4721080747134993e-05, + "loss": 0.5318, + "step": 42346 + }, + { + "epoch": 2.938519186732357, + "grad_norm": 1.573637843132019, + "learning_rate": 3.471682452636989e-05, + "loss": 0.5696, + "step": 42347 + }, + { + "epoch": 2.938588578169454, + "grad_norm": 0.8074957728385925, + "learning_rate": 3.47125685116962e-05, + "loss": 0.2585, + "step": 42348 + }, + { + "epoch": 2.9386579696065507, + "grad_norm": 1.695307731628418, + "learning_rate": 3.470831270312734e-05, + "loss": 0.4332, + "step": 42349 + }, + { + "epoch": 2.938727361043647, + "grad_norm": 1.4621144533157349, + "learning_rate": 3.470405710067681e-05, + "loss": 0.3835, + "step": 42350 + }, + { + "epoch": 2.938796752480744, + "grad_norm": 1.5236350297927856, + "learning_rate": 3.469980170435797e-05, + "loss": 0.4408, + "step": 42351 + }, + { + "epoch": 2.9388661439178403, + "grad_norm": 1.733991265296936, + "learning_rate": 3.4695546514184275e-05, + "loss": 0.3902, + "step": 42352 + }, + { + "epoch": 2.938935535354937, + "grad_norm": 1.282625436782837, + "learning_rate": 3.4691291530169156e-05, + "loss": 0.3764, + "step": 42353 + }, + { + "epoch": 2.939004926792034, + "grad_norm": 1.3312432765960693, + "learning_rate": 3.468703675232605e-05, + "loss": 0.4039, + "step": 42354 + }, + { + "epoch": 2.939074318229131, + "grad_norm": 1.3742194175720215, + "learning_rate": 3.468278218066839e-05, + "loss": 0.5332, + "step": 42355 + }, + { + "epoch": 2.939143709666227, + "grad_norm": 1.458479881286621, + "learning_rate": 3.4678527815209625e-05, + "loss": 0.4721, + "step": 42356 + }, + { + "epoch": 2.939213101103324, + "grad_norm": 0.9269689321517944, + "learning_rate": 3.46742736559632e-05, + "loss": 0.3185, + "step": 42357 + }, + { + "epoch": 2.9392824925404204, + "grad_norm": 2.5347774028778076, + "learning_rate": 3.467001970294247e-05, + "loss": 0.4793, + "step": 42358 + }, + { + "epoch": 2.9393518839775172, + "grad_norm": 1.4333900213241577, + "learning_rate": 3.46657659561609e-05, + "loss": 0.3607, + "step": 42359 + }, + { + "epoch": 2.939421275414614, + "grad_norm": 1.1772210597991943, + "learning_rate": 3.4661512415631945e-05, + "loss": 0.4027, + "step": 42360 + }, + { + "epoch": 2.9394906668517105, + "grad_norm": 1.161909818649292, + "learning_rate": 3.4657259081369006e-05, + "loss": 0.2807, + "step": 42361 + }, + { + "epoch": 2.9395600582888073, + "grad_norm": 1.2315216064453125, + "learning_rate": 3.465300595338555e-05, + "loss": 0.4941, + "step": 42362 + }, + { + "epoch": 2.9396294497259037, + "grad_norm": 1.0043712854385376, + "learning_rate": 3.464875303169495e-05, + "loss": 0.4405, + "step": 42363 + }, + { + "epoch": 2.9396988411630005, + "grad_norm": 1.4641879796981812, + "learning_rate": 3.464450031631065e-05, + "loss": 0.4, + "step": 42364 + }, + { + "epoch": 2.9397682326000973, + "grad_norm": 1.1695585250854492, + "learning_rate": 3.464024780724607e-05, + "loss": 0.4554, + "step": 42365 + }, + { + "epoch": 2.9398376240371937, + "grad_norm": 1.1261684894561768, + "learning_rate": 3.463599550451468e-05, + "loss": 0.3481, + "step": 42366 + }, + { + "epoch": 2.9399070154742906, + "grad_norm": 1.0248600244522095, + "learning_rate": 3.463174340812985e-05, + "loss": 0.2964, + "step": 42367 + }, + { + "epoch": 2.939976406911387, + "grad_norm": 1.4124608039855957, + "learning_rate": 3.4627491518105004e-05, + "loss": 0.3909, + "step": 42368 + }, + { + "epoch": 2.940045798348484, + "grad_norm": 0.9044243097305298, + "learning_rate": 3.462323983445359e-05, + "loss": 0.2699, + "step": 42369 + }, + { + "epoch": 2.9401151897855806, + "grad_norm": 1.1340482234954834, + "learning_rate": 3.4618988357189066e-05, + "loss": 0.4142, + "step": 42370 + }, + { + "epoch": 2.940184581222677, + "grad_norm": 1.041602373123169, + "learning_rate": 3.461473708632477e-05, + "loss": 0.354, + "step": 42371 + }, + { + "epoch": 2.940253972659774, + "grad_norm": 1.1262587308883667, + "learning_rate": 3.461048602187416e-05, + "loss": 0.3808, + "step": 42372 + }, + { + "epoch": 2.9403233640968702, + "grad_norm": 1.322075605392456, + "learning_rate": 3.460623516385065e-05, + "loss": 0.3368, + "step": 42373 + }, + { + "epoch": 2.940392755533967, + "grad_norm": 1.252229928970337, + "learning_rate": 3.460198451226768e-05, + "loss": 0.4127, + "step": 42374 + }, + { + "epoch": 2.940462146971064, + "grad_norm": 1.2227100133895874, + "learning_rate": 3.459773406713865e-05, + "loss": 0.4254, + "step": 42375 + }, + { + "epoch": 2.9405315384081603, + "grad_norm": 1.3083711862564087, + "learning_rate": 3.4593483828477003e-05, + "loss": 0.3302, + "step": 42376 + }, + { + "epoch": 2.940600929845257, + "grad_norm": 1.3853360414505005, + "learning_rate": 3.458923379629616e-05, + "loss": 0.4881, + "step": 42377 + }, + { + "epoch": 2.9406703212823535, + "grad_norm": 0.9571930170059204, + "learning_rate": 3.4584983970609486e-05, + "loss": 0.2345, + "step": 42378 + }, + { + "epoch": 2.9407397127194503, + "grad_norm": 0.9526572823524475, + "learning_rate": 3.4580734351430425e-05, + "loss": 0.3685, + "step": 42379 + }, + { + "epoch": 2.940809104156547, + "grad_norm": 1.1770633459091187, + "learning_rate": 3.457648493877239e-05, + "loss": 0.3682, + "step": 42380 + }, + { + "epoch": 2.940878495593644, + "grad_norm": 0.9878354072570801, + "learning_rate": 3.457223573264881e-05, + "loss": 0.3034, + "step": 42381 + }, + { + "epoch": 2.9409478870307404, + "grad_norm": 1.360113501548767, + "learning_rate": 3.456798673307309e-05, + "loss": 0.357, + "step": 42382 + }, + { + "epoch": 2.9410172784678372, + "grad_norm": 1.8813598155975342, + "learning_rate": 3.4563737940058636e-05, + "loss": 0.6641, + "step": 42383 + }, + { + "epoch": 2.9410866699049336, + "grad_norm": 0.9137597680091858, + "learning_rate": 3.455948935361891e-05, + "loss": 0.2996, + "step": 42384 + }, + { + "epoch": 2.9411560613420304, + "grad_norm": 1.1076643466949463, + "learning_rate": 3.455524097376726e-05, + "loss": 0.3609, + "step": 42385 + }, + { + "epoch": 2.9412254527791273, + "grad_norm": 0.8836668729782104, + "learning_rate": 3.455099280051711e-05, + "loss": 0.2346, + "step": 42386 + }, + { + "epoch": 2.9412948442162237, + "grad_norm": 0.8787795305252075, + "learning_rate": 3.4546744833881886e-05, + "loss": 0.3295, + "step": 42387 + }, + { + "epoch": 2.9413642356533205, + "grad_norm": 1.3372719287872314, + "learning_rate": 3.454249707387499e-05, + "loss": 0.3005, + "step": 42388 + }, + { + "epoch": 2.941433627090417, + "grad_norm": 0.9063104391098022, + "learning_rate": 3.453824952050984e-05, + "loss": 0.3388, + "step": 42389 + }, + { + "epoch": 2.9415030185275137, + "grad_norm": 1.3965195417404175, + "learning_rate": 3.453400217379984e-05, + "loss": 0.4081, + "step": 42390 + }, + { + "epoch": 2.9415724099646106, + "grad_norm": 0.8100735545158386, + "learning_rate": 3.452975503375842e-05, + "loss": 0.2534, + "step": 42391 + }, + { + "epoch": 2.941641801401707, + "grad_norm": 0.7659441828727722, + "learning_rate": 3.452550810039895e-05, + "loss": 0.3065, + "step": 42392 + }, + { + "epoch": 2.941711192838804, + "grad_norm": 1.0534641742706299, + "learning_rate": 3.452126137373486e-05, + "loss": 0.2284, + "step": 42393 + }, + { + "epoch": 2.9417805842759, + "grad_norm": 1.445413589477539, + "learning_rate": 3.451701485377954e-05, + "loss": 0.3239, + "step": 42394 + }, + { + "epoch": 2.941849975712997, + "grad_norm": 1.1817221641540527, + "learning_rate": 3.45127685405464e-05, + "loss": 0.3023, + "step": 42395 + }, + { + "epoch": 2.941919367150094, + "grad_norm": 0.8420083522796631, + "learning_rate": 3.450852243404887e-05, + "loss": 0.1782, + "step": 42396 + }, + { + "epoch": 2.9419887585871902, + "grad_norm": 1.363192081451416, + "learning_rate": 3.450427653430036e-05, + "loss": 0.3871, + "step": 42397 + }, + { + "epoch": 2.942058150024287, + "grad_norm": 1.0944099426269531, + "learning_rate": 3.4500030841314215e-05, + "loss": 0.3282, + "step": 42398 + }, + { + "epoch": 2.9421275414613834, + "grad_norm": 1.438831090927124, + "learning_rate": 3.449578535510388e-05, + "loss": 0.2838, + "step": 42399 + }, + { + "epoch": 2.9421969328984803, + "grad_norm": 1.1098456382751465, + "learning_rate": 3.4491540075682735e-05, + "loss": 0.3132, + "step": 42400 + }, + { + "epoch": 2.942266324335577, + "grad_norm": 1.476550579071045, + "learning_rate": 3.4487295003064244e-05, + "loss": 0.373, + "step": 42401 + }, + { + "epoch": 2.9423357157726735, + "grad_norm": 1.1474223136901855, + "learning_rate": 3.448305013726173e-05, + "loss": 0.1986, + "step": 42402 + }, + { + "epoch": 2.9424051072097703, + "grad_norm": 1.2272082567214966, + "learning_rate": 3.4478805478288626e-05, + "loss": 0.4283, + "step": 42403 + }, + { + "epoch": 2.9424744986468667, + "grad_norm": 1.590462565422058, + "learning_rate": 3.4474561026158326e-05, + "loss": 0.3927, + "step": 42404 + }, + { + "epoch": 2.9425438900839636, + "grad_norm": 1.256857991218567, + "learning_rate": 3.447031678088427e-05, + "loss": 0.3023, + "step": 42405 + }, + { + "epoch": 2.9426132815210604, + "grad_norm": 1.0679374933242798, + "learning_rate": 3.446607274247978e-05, + "loss": 0.315, + "step": 42406 + }, + { + "epoch": 2.942682672958157, + "grad_norm": 1.000506043434143, + "learning_rate": 3.4461828910958306e-05, + "loss": 0.2624, + "step": 42407 + }, + { + "epoch": 2.9427520643952536, + "grad_norm": 1.371160864830017, + "learning_rate": 3.4457585286333226e-05, + "loss": 0.3499, + "step": 42408 + }, + { + "epoch": 2.9428214558323504, + "grad_norm": 0.8232013583183289, + "learning_rate": 3.445334186861795e-05, + "loss": 0.28, + "step": 42409 + }, + { + "epoch": 2.942890847269447, + "grad_norm": 1.0324217081069946, + "learning_rate": 3.4449098657825875e-05, + "loss": 0.3868, + "step": 42410 + }, + { + "epoch": 2.9429602387065437, + "grad_norm": 1.0970598459243774, + "learning_rate": 3.444485565397038e-05, + "loss": 0.3609, + "step": 42411 + }, + { + "epoch": 2.9430296301436405, + "grad_norm": 1.0887603759765625, + "learning_rate": 3.4440612857064904e-05, + "loss": 0.4269, + "step": 42412 + }, + { + "epoch": 2.943099021580737, + "grad_norm": 1.3306989669799805, + "learning_rate": 3.4436370267122776e-05, + "loss": 0.463, + "step": 42413 + }, + { + "epoch": 2.9431684130178337, + "grad_norm": 0.926368772983551, + "learning_rate": 3.443212788415742e-05, + "loss": 0.3725, + "step": 42414 + }, + { + "epoch": 2.94323780445493, + "grad_norm": 1.0943478345870972, + "learning_rate": 3.4427885708182226e-05, + "loss": 0.2654, + "step": 42415 + }, + { + "epoch": 2.943307195892027, + "grad_norm": 1.2231019735336304, + "learning_rate": 3.44236437392106e-05, + "loss": 0.2894, + "step": 42416 + }, + { + "epoch": 2.9433765873291238, + "grad_norm": 1.5182029008865356, + "learning_rate": 3.441940197725591e-05, + "loss": 0.4595, + "step": 42417 + }, + { + "epoch": 2.94344597876622, + "grad_norm": 0.7108277678489685, + "learning_rate": 3.441516042233156e-05, + "loss": 0.3135, + "step": 42418 + }, + { + "epoch": 2.943515370203317, + "grad_norm": 1.0698330402374268, + "learning_rate": 3.4410919074450975e-05, + "loss": 0.3844, + "step": 42419 + }, + { + "epoch": 2.9435847616404134, + "grad_norm": 1.254582166671753, + "learning_rate": 3.4406677933627484e-05, + "loss": 0.3834, + "step": 42420 + }, + { + "epoch": 2.94365415307751, + "grad_norm": 1.3734116554260254, + "learning_rate": 3.4402436999874486e-05, + "loss": 0.4675, + "step": 42421 + }, + { + "epoch": 2.943723544514607, + "grad_norm": 1.2652024030685425, + "learning_rate": 3.4398196273205396e-05, + "loss": 0.5318, + "step": 42422 + }, + { + "epoch": 2.9437929359517034, + "grad_norm": 1.674056053161621, + "learning_rate": 3.439395575363359e-05, + "loss": 0.3914, + "step": 42423 + }, + { + "epoch": 2.9438623273888003, + "grad_norm": 1.0040147304534912, + "learning_rate": 3.438971544117244e-05, + "loss": 0.4169, + "step": 42424 + }, + { + "epoch": 2.9439317188258967, + "grad_norm": 1.317413568496704, + "learning_rate": 3.438547533583536e-05, + "loss": 0.3988, + "step": 42425 + }, + { + "epoch": 2.9440011102629935, + "grad_norm": 0.739342451095581, + "learning_rate": 3.438123543763574e-05, + "loss": 0.2851, + "step": 42426 + }, + { + "epoch": 2.9440705017000903, + "grad_norm": 1.1932978630065918, + "learning_rate": 3.437699574658693e-05, + "loss": 0.3107, + "step": 42427 + }, + { + "epoch": 2.9441398931371867, + "grad_norm": 1.4458729028701782, + "learning_rate": 3.437275626270232e-05, + "loss": 0.4554, + "step": 42428 + }, + { + "epoch": 2.9442092845742835, + "grad_norm": 1.1588971614837646, + "learning_rate": 3.43685169859953e-05, + "loss": 0.3645, + "step": 42429 + }, + { + "epoch": 2.94427867601138, + "grad_norm": 0.9136581420898438, + "learning_rate": 3.436427791647927e-05, + "loss": 0.3393, + "step": 42430 + }, + { + "epoch": 2.9443480674484768, + "grad_norm": 1.1819796562194824, + "learning_rate": 3.436003905416759e-05, + "loss": 0.3011, + "step": 42431 + }, + { + "epoch": 2.9444174588855736, + "grad_norm": 1.1521503925323486, + "learning_rate": 3.435580039907369e-05, + "loss": 0.4408, + "step": 42432 + }, + { + "epoch": 2.9444868503226704, + "grad_norm": 1.2905848026275635, + "learning_rate": 3.435156195121089e-05, + "loss": 0.279, + "step": 42433 + }, + { + "epoch": 2.944556241759767, + "grad_norm": 1.2749412059783936, + "learning_rate": 3.4347323710592574e-05, + "loss": 0.3611, + "step": 42434 + }, + { + "epoch": 2.9446256331968637, + "grad_norm": 0.9004039764404297, + "learning_rate": 3.434308567723215e-05, + "loss": 0.3397, + "step": 42435 + }, + { + "epoch": 2.94469502463396, + "grad_norm": 1.2227284908294678, + "learning_rate": 3.433884785114301e-05, + "loss": 0.5433, + "step": 42436 + }, + { + "epoch": 2.944764416071057, + "grad_norm": 1.228324055671692, + "learning_rate": 3.4334610232338483e-05, + "loss": 0.4928, + "step": 42437 + }, + { + "epoch": 2.9448338075081537, + "grad_norm": 1.3796645402908325, + "learning_rate": 3.433037282083197e-05, + "loss": 0.6024, + "step": 42438 + }, + { + "epoch": 2.94490319894525, + "grad_norm": 0.8887342214584351, + "learning_rate": 3.4326135616636854e-05, + "loss": 0.2258, + "step": 42439 + }, + { + "epoch": 2.944972590382347, + "grad_norm": 1.2908388376235962, + "learning_rate": 3.4321898619766535e-05, + "loss": 0.4625, + "step": 42440 + }, + { + "epoch": 2.9450419818194433, + "grad_norm": 1.2462656497955322, + "learning_rate": 3.431766183023434e-05, + "loss": 0.4361, + "step": 42441 + }, + { + "epoch": 2.94511137325654, + "grad_norm": 0.9759596586227417, + "learning_rate": 3.431342524805366e-05, + "loss": 0.3246, + "step": 42442 + }, + { + "epoch": 2.945180764693637, + "grad_norm": 0.9750998020172119, + "learning_rate": 3.430918887323788e-05, + "loss": 0.4078, + "step": 42443 + }, + { + "epoch": 2.9452501561307334, + "grad_norm": 1.2996792793273926, + "learning_rate": 3.4304952705800354e-05, + "loss": 0.4845, + "step": 42444 + }, + { + "epoch": 2.94531954756783, + "grad_norm": 1.0216493606567383, + "learning_rate": 3.430071674575448e-05, + "loss": 0.342, + "step": 42445 + }, + { + "epoch": 2.9453889390049266, + "grad_norm": 0.9670411944389343, + "learning_rate": 3.429648099311366e-05, + "loss": 0.2803, + "step": 42446 + }, + { + "epoch": 2.9454583304420234, + "grad_norm": 0.9745684266090393, + "learning_rate": 3.429224544789118e-05, + "loss": 0.2903, + "step": 42447 + }, + { + "epoch": 2.9455277218791203, + "grad_norm": 1.2627332210540771, + "learning_rate": 3.428801011010047e-05, + "loss": 0.3461, + "step": 42448 + }, + { + "epoch": 2.9455971133162167, + "grad_norm": 1.706781029701233, + "learning_rate": 3.428377497975488e-05, + "loss": 0.352, + "step": 42449 + }, + { + "epoch": 2.9456665047533135, + "grad_norm": 1.5856380462646484, + "learning_rate": 3.42795400568678e-05, + "loss": 0.3609, + "step": 42450 + }, + { + "epoch": 2.94573589619041, + "grad_norm": 1.592732548713684, + "learning_rate": 3.427530534145257e-05, + "loss": 0.6578, + "step": 42451 + }, + { + "epoch": 2.9458052876275067, + "grad_norm": 1.0099530220031738, + "learning_rate": 3.427107083352259e-05, + "loss": 0.3615, + "step": 42452 + }, + { + "epoch": 2.9458746790646035, + "grad_norm": 1.0000801086425781, + "learning_rate": 3.426683653309124e-05, + "loss": 0.3085, + "step": 42453 + }, + { + "epoch": 2.9459440705017, + "grad_norm": 1.2290658950805664, + "learning_rate": 3.426260244017182e-05, + "loss": 0.3088, + "step": 42454 + }, + { + "epoch": 2.9460134619387968, + "grad_norm": 1.534979224205017, + "learning_rate": 3.425836855477775e-05, + "loss": 0.5598, + "step": 42455 + }, + { + "epoch": 2.946082853375893, + "grad_norm": 0.8715418577194214, + "learning_rate": 3.425413487692238e-05, + "loss": 0.2586, + "step": 42456 + }, + { + "epoch": 2.94615224481299, + "grad_norm": 1.0026148557662964, + "learning_rate": 3.4249901406619076e-05, + "loss": 0.289, + "step": 42457 + }, + { + "epoch": 2.946221636250087, + "grad_norm": 0.8582665920257568, + "learning_rate": 3.4245668143881206e-05, + "loss": 0.3601, + "step": 42458 + }, + { + "epoch": 2.9462910276871837, + "grad_norm": 1.8188949823379517, + "learning_rate": 3.424143508872213e-05, + "loss": 0.5628, + "step": 42459 + }, + { + "epoch": 2.94636041912428, + "grad_norm": 1.3687165975570679, + "learning_rate": 3.423720224115525e-05, + "loss": 0.4136, + "step": 42460 + }, + { + "epoch": 2.946429810561377, + "grad_norm": 0.9276875257492065, + "learning_rate": 3.4232969601193856e-05, + "loss": 0.3618, + "step": 42461 + }, + { + "epoch": 2.9464992019984733, + "grad_norm": 0.7836611866950989, + "learning_rate": 3.422873716885136e-05, + "loss": 0.2581, + "step": 42462 + }, + { + "epoch": 2.94656859343557, + "grad_norm": 1.2399466037750244, + "learning_rate": 3.422450494414109e-05, + "loss": 0.2573, + "step": 42463 + }, + { + "epoch": 2.946637984872667, + "grad_norm": 2.6339914798736572, + "learning_rate": 3.4220272927076434e-05, + "loss": 0.5735, + "step": 42464 + }, + { + "epoch": 2.9467073763097633, + "grad_norm": 0.8036618232727051, + "learning_rate": 3.421604111767075e-05, + "loss": 0.3231, + "step": 42465 + }, + { + "epoch": 2.94677676774686, + "grad_norm": 1.5821247100830078, + "learning_rate": 3.4211809515937385e-05, + "loss": 0.2805, + "step": 42466 + }, + { + "epoch": 2.9468461591839565, + "grad_norm": 1.0480372905731201, + "learning_rate": 3.4207578121889736e-05, + "loss": 0.3674, + "step": 42467 + }, + { + "epoch": 2.9469155506210534, + "grad_norm": 1.2674803733825684, + "learning_rate": 3.4203346935541105e-05, + "loss": 0.3579, + "step": 42468 + }, + { + "epoch": 2.94698494205815, + "grad_norm": 1.1445116996765137, + "learning_rate": 3.419911595690487e-05, + "loss": 0.4759, + "step": 42469 + }, + { + "epoch": 2.9470543334952466, + "grad_norm": 1.2575381994247437, + "learning_rate": 3.419488518599443e-05, + "loss": 0.5115, + "step": 42470 + }, + { + "epoch": 2.9471237249323434, + "grad_norm": 0.9117036461830139, + "learning_rate": 3.419065462282306e-05, + "loss": 0.2843, + "step": 42471 + }, + { + "epoch": 2.94719311636944, + "grad_norm": 1.0780967473983765, + "learning_rate": 3.418642426740416e-05, + "loss": 0.581, + "step": 42472 + }, + { + "epoch": 2.9472625078065366, + "grad_norm": 0.8735588788986206, + "learning_rate": 3.418219411975109e-05, + "loss": 0.3463, + "step": 42473 + }, + { + "epoch": 2.9473318992436335, + "grad_norm": 1.2639214992523193, + "learning_rate": 3.417796417987722e-05, + "loss": 0.3147, + "step": 42474 + }, + { + "epoch": 2.94740129068073, + "grad_norm": 1.257796049118042, + "learning_rate": 3.417373444779586e-05, + "loss": 0.5324, + "step": 42475 + }, + { + "epoch": 2.9474706821178267, + "grad_norm": 0.9529324769973755, + "learning_rate": 3.416950492352037e-05, + "loss": 0.2466, + "step": 42476 + }, + { + "epoch": 2.947540073554923, + "grad_norm": 1.2425367832183838, + "learning_rate": 3.416527560706412e-05, + "loss": 0.3554, + "step": 42477 + }, + { + "epoch": 2.94760946499202, + "grad_norm": 1.5571658611297607, + "learning_rate": 3.416104649844046e-05, + "loss": 0.2248, + "step": 42478 + }, + { + "epoch": 2.9476788564291168, + "grad_norm": 1.1255664825439453, + "learning_rate": 3.415681759766274e-05, + "loss": 0.4214, + "step": 42479 + }, + { + "epoch": 2.947748247866213, + "grad_norm": 1.126346468925476, + "learning_rate": 3.41525889047443e-05, + "loss": 0.3204, + "step": 42480 + }, + { + "epoch": 2.94781763930331, + "grad_norm": 0.8592971563339233, + "learning_rate": 3.4148360419698534e-05, + "loss": 0.287, + "step": 42481 + }, + { + "epoch": 2.9478870307404064, + "grad_norm": 1.7567299604415894, + "learning_rate": 3.414413214253871e-05, + "loss": 0.3922, + "step": 42482 + }, + { + "epoch": 2.947956422177503, + "grad_norm": 0.9235590100288391, + "learning_rate": 3.413990407327823e-05, + "loss": 0.3289, + "step": 42483 + }, + { + "epoch": 2.9480258136146, + "grad_norm": 1.0420266389846802, + "learning_rate": 3.4135676211930435e-05, + "loss": 0.3606, + "step": 42484 + }, + { + "epoch": 2.948095205051697, + "grad_norm": 0.8718050122261047, + "learning_rate": 3.413144855850866e-05, + "loss": 0.3021, + "step": 42485 + }, + { + "epoch": 2.9481645964887933, + "grad_norm": 1.2780365943908691, + "learning_rate": 3.4127221113026265e-05, + "loss": 0.3112, + "step": 42486 + }, + { + "epoch": 2.94823398792589, + "grad_norm": 1.0790501832962036, + "learning_rate": 3.412299387549659e-05, + "loss": 0.5038, + "step": 42487 + }, + { + "epoch": 2.9483033793629865, + "grad_norm": 1.0552220344543457, + "learning_rate": 3.4118766845933014e-05, + "loss": 0.327, + "step": 42488 + }, + { + "epoch": 2.9483727708000833, + "grad_norm": 1.0292929410934448, + "learning_rate": 3.411454002434882e-05, + "loss": 0.237, + "step": 42489 + }, + { + "epoch": 2.94844216223718, + "grad_norm": 1.3673089742660522, + "learning_rate": 3.4110313410757366e-05, + "loss": 0.4563, + "step": 42490 + }, + { + "epoch": 2.9485115536742765, + "grad_norm": 1.254469394683838, + "learning_rate": 3.410608700517202e-05, + "loss": 0.3715, + "step": 42491 + }, + { + "epoch": 2.9485809451113734, + "grad_norm": 0.8234046101570129, + "learning_rate": 3.41018608076061e-05, + "loss": 0.2466, + "step": 42492 + }, + { + "epoch": 2.9486503365484698, + "grad_norm": 0.7545185685157776, + "learning_rate": 3.409763481807297e-05, + "loss": 0.3203, + "step": 42493 + }, + { + "epoch": 2.9487197279855666, + "grad_norm": 1.2238839864730835, + "learning_rate": 3.4093409036585954e-05, + "loss": 0.4488, + "step": 42494 + }, + { + "epoch": 2.9487891194226634, + "grad_norm": 0.8600780963897705, + "learning_rate": 3.4089183463158434e-05, + "loss": 0.3134, + "step": 42495 + }, + { + "epoch": 2.94885851085976, + "grad_norm": 1.1233127117156982, + "learning_rate": 3.408495809780369e-05, + "loss": 0.2774, + "step": 42496 + }, + { + "epoch": 2.9489279022968566, + "grad_norm": 0.9703748226165771, + "learning_rate": 3.408073294053508e-05, + "loss": 0.3152, + "step": 42497 + }, + { + "epoch": 2.948997293733953, + "grad_norm": 1.2130229473114014, + "learning_rate": 3.407650799136595e-05, + "loss": 0.5198, + "step": 42498 + }, + { + "epoch": 2.94906668517105, + "grad_norm": 1.1666496992111206, + "learning_rate": 3.407228325030962e-05, + "loss": 0.4651, + "step": 42499 + }, + { + "epoch": 2.9491360766081467, + "grad_norm": 1.4232584238052368, + "learning_rate": 3.4068058717379456e-05, + "loss": 0.515, + "step": 42500 + }, + { + "epoch": 2.949205468045243, + "grad_norm": 1.2447456121444702, + "learning_rate": 3.406383439258879e-05, + "loss": 0.3658, + "step": 42501 + }, + { + "epoch": 2.94927485948234, + "grad_norm": 1.1389588117599487, + "learning_rate": 3.405961027595097e-05, + "loss": 0.3063, + "step": 42502 + }, + { + "epoch": 2.9493442509194363, + "grad_norm": 1.1185604333877563, + "learning_rate": 3.405538636747928e-05, + "loss": 0.3113, + "step": 42503 + }, + { + "epoch": 2.949413642356533, + "grad_norm": 1.4790955781936646, + "learning_rate": 3.4051162667187076e-05, + "loss": 0.4268, + "step": 42504 + }, + { + "epoch": 2.94948303379363, + "grad_norm": 0.6663240194320679, + "learning_rate": 3.404693917508774e-05, + "loss": 0.1836, + "step": 42505 + }, + { + "epoch": 2.949552425230727, + "grad_norm": 0.8718065619468689, + "learning_rate": 3.4042715891194534e-05, + "loss": 0.2762, + "step": 42506 + }, + { + "epoch": 2.949621816667823, + "grad_norm": 1.136337161064148, + "learning_rate": 3.4038492815520815e-05, + "loss": 0.3, + "step": 42507 + }, + { + "epoch": 2.9496912081049196, + "grad_norm": 0.9623289108276367, + "learning_rate": 3.4034269948079935e-05, + "loss": 0.2925, + "step": 42508 + }, + { + "epoch": 2.9497605995420164, + "grad_norm": 1.2180732488632202, + "learning_rate": 3.4030047288885235e-05, + "loss": 0.345, + "step": 42509 + }, + { + "epoch": 2.9498299909791132, + "grad_norm": 1.8196775913238525, + "learning_rate": 3.4025824837949984e-05, + "loss": 0.3787, + "step": 42510 + }, + { + "epoch": 2.94989938241621, + "grad_norm": 0.9831900000572205, + "learning_rate": 3.402160259528756e-05, + "loss": 0.3734, + "step": 42511 + }, + { + "epoch": 2.9499687738533065, + "grad_norm": 0.8968014121055603, + "learning_rate": 3.401738056091127e-05, + "loss": 0.2738, + "step": 42512 + }, + { + "epoch": 2.9500381652904033, + "grad_norm": 1.5488293170928955, + "learning_rate": 3.401315873483446e-05, + "loss": 0.3121, + "step": 42513 + }, + { + "epoch": 2.9501075567274997, + "grad_norm": 2.6825497150421143, + "learning_rate": 3.4008937117070463e-05, + "loss": 0.3631, + "step": 42514 + }, + { + "epoch": 2.9501769481645965, + "grad_norm": 1.1778068542480469, + "learning_rate": 3.4004715707632586e-05, + "loss": 0.3676, + "step": 42515 + }, + { + "epoch": 2.9502463396016934, + "grad_norm": 1.4065626859664917, + "learning_rate": 3.4000494506534184e-05, + "loss": 0.526, + "step": 42516 + }, + { + "epoch": 2.9503157310387897, + "grad_norm": 1.089431643486023, + "learning_rate": 3.399627351378855e-05, + "loss": 0.3258, + "step": 42517 + }, + { + "epoch": 2.9503851224758866, + "grad_norm": 1.272573471069336, + "learning_rate": 3.399205272940902e-05, + "loss": 0.3018, + "step": 42518 + }, + { + "epoch": 2.950454513912983, + "grad_norm": 1.2588082551956177, + "learning_rate": 3.3987832153408916e-05, + "loss": 0.2993, + "step": 42519 + }, + { + "epoch": 2.95052390535008, + "grad_norm": 1.2796993255615234, + "learning_rate": 3.398361178580156e-05, + "loss": 0.4458, + "step": 42520 + }, + { + "epoch": 2.9505932967871766, + "grad_norm": 1.4376832246780396, + "learning_rate": 3.397939162660029e-05, + "loss": 0.3479, + "step": 42521 + }, + { + "epoch": 2.950662688224273, + "grad_norm": 0.9254965782165527, + "learning_rate": 3.397517167581845e-05, + "loss": 0.3189, + "step": 42522 + }, + { + "epoch": 2.95073207966137, + "grad_norm": 1.203447699546814, + "learning_rate": 3.397095193346931e-05, + "loss": 0.3774, + "step": 42523 + }, + { + "epoch": 2.9508014710984662, + "grad_norm": 0.981539249420166, + "learning_rate": 3.3966732399566206e-05, + "loss": 0.292, + "step": 42524 + }, + { + "epoch": 2.950870862535563, + "grad_norm": 1.3009239435195923, + "learning_rate": 3.396251307412247e-05, + "loss": 0.4359, + "step": 42525 + }, + { + "epoch": 2.95094025397266, + "grad_norm": 1.5596057176589966, + "learning_rate": 3.395829395715142e-05, + "loss": 0.4319, + "step": 42526 + }, + { + "epoch": 2.9510096454097563, + "grad_norm": 1.6556023359298706, + "learning_rate": 3.3954075048666365e-05, + "loss": 0.3821, + "step": 42527 + }, + { + "epoch": 2.951079036846853, + "grad_norm": 1.5082197189331055, + "learning_rate": 3.394985634868064e-05, + "loss": 0.5646, + "step": 42528 + }, + { + "epoch": 2.9511484282839495, + "grad_norm": 1.642962098121643, + "learning_rate": 3.394563785720758e-05, + "loss": 0.4211, + "step": 42529 + }, + { + "epoch": 2.9512178197210464, + "grad_norm": 1.7775934934616089, + "learning_rate": 3.394141957426046e-05, + "loss": 0.3985, + "step": 42530 + }, + { + "epoch": 2.951287211158143, + "grad_norm": 1.4574949741363525, + "learning_rate": 3.39372014998526e-05, + "loss": 0.3807, + "step": 42531 + }, + { + "epoch": 2.95135660259524, + "grad_norm": 1.2766121625900269, + "learning_rate": 3.393298363399734e-05, + "loss": 0.3643, + "step": 42532 + }, + { + "epoch": 2.9514259940323364, + "grad_norm": 0.9012949466705322, + "learning_rate": 3.3928765976707976e-05, + "loss": 0.4827, + "step": 42533 + }, + { + "epoch": 2.951495385469433, + "grad_norm": 1.4826310873031616, + "learning_rate": 3.392454852799783e-05, + "loss": 0.5168, + "step": 42534 + }, + { + "epoch": 2.9515647769065296, + "grad_norm": 1.0591561794281006, + "learning_rate": 3.3920331287880226e-05, + "loss": 0.3803, + "step": 42535 + }, + { + "epoch": 2.9516341683436265, + "grad_norm": 1.5167666673660278, + "learning_rate": 3.3916114256368505e-05, + "loss": 0.507, + "step": 42536 + }, + { + "epoch": 2.9517035597807233, + "grad_norm": 1.2584999799728394, + "learning_rate": 3.39118974334759e-05, + "loss": 0.4145, + "step": 42537 + }, + { + "epoch": 2.9517729512178197, + "grad_norm": 1.7721052169799805, + "learning_rate": 3.390768081921578e-05, + "loss": 0.4047, + "step": 42538 + }, + { + "epoch": 2.9518423426549165, + "grad_norm": 1.8110336065292358, + "learning_rate": 3.390346441360144e-05, + "loss": 0.3183, + "step": 42539 + }, + { + "epoch": 2.951911734092013, + "grad_norm": 1.2307966947555542, + "learning_rate": 3.389924821664623e-05, + "loss": 0.5427, + "step": 42540 + }, + { + "epoch": 2.9519811255291097, + "grad_norm": 1.2774219512939453, + "learning_rate": 3.389503222836338e-05, + "loss": 0.4701, + "step": 42541 + }, + { + "epoch": 2.9520505169662066, + "grad_norm": 1.0607142448425293, + "learning_rate": 3.389081644876625e-05, + "loss": 0.3853, + "step": 42542 + }, + { + "epoch": 2.952119908403303, + "grad_norm": 1.3565716743469238, + "learning_rate": 3.388660087786815e-05, + "loss": 0.4308, + "step": 42543 + }, + { + "epoch": 2.9521892998404, + "grad_norm": 0.87016761302948, + "learning_rate": 3.3882385515682405e-05, + "loss": 0.2845, + "step": 42544 + }, + { + "epoch": 2.952258691277496, + "grad_norm": 1.9792946577072144, + "learning_rate": 3.3878170362222264e-05, + "loss": 0.4391, + "step": 42545 + }, + { + "epoch": 2.952328082714593, + "grad_norm": 1.1248005628585815, + "learning_rate": 3.387395541750107e-05, + "loss": 0.2756, + "step": 42546 + }, + { + "epoch": 2.95239747415169, + "grad_norm": 1.4085692167282104, + "learning_rate": 3.386974068153212e-05, + "loss": 0.4768, + "step": 42547 + }, + { + "epoch": 2.9524668655887862, + "grad_norm": 1.1587518453598022, + "learning_rate": 3.386552615432873e-05, + "loss": 0.2448, + "step": 42548 + }, + { + "epoch": 2.952536257025883, + "grad_norm": 1.0429542064666748, + "learning_rate": 3.3861311835904206e-05, + "loss": 0.3326, + "step": 42549 + }, + { + "epoch": 2.9526056484629795, + "grad_norm": 1.166443109512329, + "learning_rate": 3.3857097726271883e-05, + "loss": 0.4215, + "step": 42550 + }, + { + "epoch": 2.9526750399000763, + "grad_norm": 0.8588929176330566, + "learning_rate": 3.385288382544498e-05, + "loss": 0.345, + "step": 42551 + }, + { + "epoch": 2.952744431337173, + "grad_norm": 1.0133672952651978, + "learning_rate": 3.384867013343685e-05, + "loss": 0.3608, + "step": 42552 + }, + { + "epoch": 2.9528138227742695, + "grad_norm": 0.8273351788520813, + "learning_rate": 3.3844456650260814e-05, + "loss": 0.2751, + "step": 42553 + }, + { + "epoch": 2.9528832142113663, + "grad_norm": 0.7581399083137512, + "learning_rate": 3.3840243375930135e-05, + "loss": 0.3383, + "step": 42554 + }, + { + "epoch": 2.9529526056484627, + "grad_norm": 1.0197299718856812, + "learning_rate": 3.383603031045814e-05, + "loss": 0.3891, + "step": 42555 + }, + { + "epoch": 2.9530219970855596, + "grad_norm": 0.8795846104621887, + "learning_rate": 3.383181745385812e-05, + "loss": 0.2632, + "step": 42556 + }, + { + "epoch": 2.9530913885226564, + "grad_norm": 1.3300141096115112, + "learning_rate": 3.38276048061434e-05, + "loss": 0.4096, + "step": 42557 + }, + { + "epoch": 2.9531607799597532, + "grad_norm": 1.129702091217041, + "learning_rate": 3.382339236732723e-05, + "loss": 0.4205, + "step": 42558 + }, + { + "epoch": 2.9532301713968496, + "grad_norm": 1.3047139644622803, + "learning_rate": 3.381918013742293e-05, + "loss": 0.3933, + "step": 42559 + }, + { + "epoch": 2.9532995628339465, + "grad_norm": 0.9334124326705933, + "learning_rate": 3.38149681164438e-05, + "loss": 0.3104, + "step": 42560 + }, + { + "epoch": 2.953368954271043, + "grad_norm": 1.4735862016677856, + "learning_rate": 3.3810756304403134e-05, + "loss": 0.4519, + "step": 42561 + }, + { + "epoch": 2.9534383457081397, + "grad_norm": 1.0939677953720093, + "learning_rate": 3.380654470131423e-05, + "loss": 0.2969, + "step": 42562 + }, + { + "epoch": 2.9535077371452365, + "grad_norm": 1.950319766998291, + "learning_rate": 3.380233330719039e-05, + "loss": 0.4564, + "step": 42563 + }, + { + "epoch": 2.953577128582333, + "grad_norm": 1.9733260869979858, + "learning_rate": 3.3798122122044936e-05, + "loss": 0.5438, + "step": 42564 + }, + { + "epoch": 2.9536465200194297, + "grad_norm": 1.5125014781951904, + "learning_rate": 3.37939111458911e-05, + "loss": 0.6156, + "step": 42565 + }, + { + "epoch": 2.953715911456526, + "grad_norm": 1.5383954048156738, + "learning_rate": 3.3789700378742196e-05, + "loss": 0.461, + "step": 42566 + }, + { + "epoch": 2.953785302893623, + "grad_norm": 1.0605672597885132, + "learning_rate": 3.378548982061152e-05, + "loss": 0.3764, + "step": 42567 + }, + { + "epoch": 2.95385469433072, + "grad_norm": 1.2203032970428467, + "learning_rate": 3.378127947151239e-05, + "loss": 0.4356, + "step": 42568 + }, + { + "epoch": 2.953924085767816, + "grad_norm": 1.306803584098816, + "learning_rate": 3.377706933145807e-05, + "loss": 0.3121, + "step": 42569 + }, + { + "epoch": 2.953993477204913, + "grad_norm": 1.6169155836105347, + "learning_rate": 3.3772859400461854e-05, + "loss": 0.4902, + "step": 42570 + }, + { + "epoch": 2.9540628686420094, + "grad_norm": 1.1766902208328247, + "learning_rate": 3.3768649678537075e-05, + "loss": 0.3494, + "step": 42571 + }, + { + "epoch": 2.9541322600791062, + "grad_norm": 0.6835522055625916, + "learning_rate": 3.376444016569695e-05, + "loss": 0.2176, + "step": 42572 + }, + { + "epoch": 2.954201651516203, + "grad_norm": 1.468122124671936, + "learning_rate": 3.37602308619548e-05, + "loss": 0.4396, + "step": 42573 + }, + { + "epoch": 2.9542710429532995, + "grad_norm": 1.2861262559890747, + "learning_rate": 3.375602176732392e-05, + "loss": 0.3207, + "step": 42574 + }, + { + "epoch": 2.9543404343903963, + "grad_norm": 1.3351314067840576, + "learning_rate": 3.3751812881817626e-05, + "loss": 0.3276, + "step": 42575 + }, + { + "epoch": 2.9544098258274927, + "grad_norm": 1.2115484476089478, + "learning_rate": 3.3747604205449135e-05, + "loss": 0.3635, + "step": 42576 + }, + { + "epoch": 2.9544792172645895, + "grad_norm": 0.8890538811683655, + "learning_rate": 3.3743395738231774e-05, + "loss": 0.2653, + "step": 42577 + }, + { + "epoch": 2.9545486087016863, + "grad_norm": 1.4468035697937012, + "learning_rate": 3.373918748017883e-05, + "loss": 0.3958, + "step": 42578 + }, + { + "epoch": 2.9546180001387827, + "grad_norm": 0.7476900815963745, + "learning_rate": 3.373497943130362e-05, + "loss": 0.2571, + "step": 42579 + }, + { + "epoch": 2.9546873915758796, + "grad_norm": 1.2363828420639038, + "learning_rate": 3.373077159161935e-05, + "loss": 0.4118, + "step": 42580 + }, + { + "epoch": 2.954756783012976, + "grad_norm": 1.5864710807800293, + "learning_rate": 3.3726563961139346e-05, + "loss": 0.5176, + "step": 42581 + }, + { + "epoch": 2.954826174450073, + "grad_norm": 3.187532424926758, + "learning_rate": 3.372235653987689e-05, + "loss": 0.491, + "step": 42582 + }, + { + "epoch": 2.9548955658871696, + "grad_norm": 1.0308568477630615, + "learning_rate": 3.371814932784526e-05, + "loss": 0.395, + "step": 42583 + }, + { + "epoch": 2.9549649573242665, + "grad_norm": 1.6752227544784546, + "learning_rate": 3.371394232505775e-05, + "loss": 0.461, + "step": 42584 + }, + { + "epoch": 2.955034348761363, + "grad_norm": 1.2224253416061401, + "learning_rate": 3.3709735531527665e-05, + "loss": 0.3817, + "step": 42585 + }, + { + "epoch": 2.9551037401984597, + "grad_norm": 0.5716179013252258, + "learning_rate": 3.370552894726822e-05, + "loss": 0.2612, + "step": 42586 + }, + { + "epoch": 2.955173131635556, + "grad_norm": 0.9972409605979919, + "learning_rate": 3.3701322572292724e-05, + "loss": 0.3652, + "step": 42587 + }, + { + "epoch": 2.955242523072653, + "grad_norm": 0.9834387302398682, + "learning_rate": 3.369711640661447e-05, + "loss": 0.2998, + "step": 42588 + }, + { + "epoch": 2.9553119145097497, + "grad_norm": 1.5344706773757935, + "learning_rate": 3.369291045024673e-05, + "loss": 0.4299, + "step": 42589 + }, + { + "epoch": 2.955381305946846, + "grad_norm": 1.5180163383483887, + "learning_rate": 3.368870470320277e-05, + "loss": 0.5084, + "step": 42590 + }, + { + "epoch": 2.955450697383943, + "grad_norm": 1.4933432340621948, + "learning_rate": 3.3684499165495875e-05, + "loss": 0.5394, + "step": 42591 + }, + { + "epoch": 2.9555200888210393, + "grad_norm": 1.3539469242095947, + "learning_rate": 3.3680293837139356e-05, + "loss": 0.5559, + "step": 42592 + }, + { + "epoch": 2.955589480258136, + "grad_norm": 4.382140159606934, + "learning_rate": 3.367608871814642e-05, + "loss": 0.4635, + "step": 42593 + }, + { + "epoch": 2.955658871695233, + "grad_norm": 1.4631515741348267, + "learning_rate": 3.367188380853038e-05, + "loss": 0.5403, + "step": 42594 + }, + { + "epoch": 2.9557282631323294, + "grad_norm": 1.2330060005187988, + "learning_rate": 3.3667679108304505e-05, + "loss": 0.4677, + "step": 42595 + }, + { + "epoch": 2.9557976545694262, + "grad_norm": 1.4015493392944336, + "learning_rate": 3.3663474617482075e-05, + "loss": 0.3782, + "step": 42596 + }, + { + "epoch": 2.9558670460065226, + "grad_norm": 1.2165662050247192, + "learning_rate": 3.365927033607636e-05, + "loss": 0.4443, + "step": 42597 + }, + { + "epoch": 2.9559364374436194, + "grad_norm": 0.9765894412994385, + "learning_rate": 3.365506626410067e-05, + "loss": 0.3667, + "step": 42598 + }, + { + "epoch": 2.9560058288807163, + "grad_norm": 1.5403103828430176, + "learning_rate": 3.3650862401568186e-05, + "loss": 0.4687, + "step": 42599 + }, + { + "epoch": 2.9560752203178127, + "grad_norm": 0.7913837432861328, + "learning_rate": 3.364665874849224e-05, + "loss": 0.2403, + "step": 42600 + }, + { + "epoch": 2.9561446117549095, + "grad_norm": 1.436391830444336, + "learning_rate": 3.36424553048861e-05, + "loss": 0.4233, + "step": 42601 + }, + { + "epoch": 2.956214003192006, + "grad_norm": 0.9497317671775818, + "learning_rate": 3.363825207076303e-05, + "loss": 0.3064, + "step": 42602 + }, + { + "epoch": 2.9562833946291027, + "grad_norm": 1.4306076765060425, + "learning_rate": 3.3634049046136297e-05, + "loss": 0.3137, + "step": 42603 + }, + { + "epoch": 2.9563527860661996, + "grad_norm": 1.640874981880188, + "learning_rate": 3.362984623101917e-05, + "loss": 0.386, + "step": 42604 + }, + { + "epoch": 2.956422177503296, + "grad_norm": 1.0677872896194458, + "learning_rate": 3.3625643625424954e-05, + "loss": 0.3285, + "step": 42605 + }, + { + "epoch": 2.9564915689403928, + "grad_norm": 0.9436873197555542, + "learning_rate": 3.3621441229366834e-05, + "loss": 0.2515, + "step": 42606 + }, + { + "epoch": 2.956560960377489, + "grad_norm": 1.1929388046264648, + "learning_rate": 3.361723904285814e-05, + "loss": 0.5211, + "step": 42607 + }, + { + "epoch": 2.956630351814586, + "grad_norm": 1.0276358127593994, + "learning_rate": 3.361303706591211e-05, + "loss": 0.2341, + "step": 42608 + }, + { + "epoch": 2.956699743251683, + "grad_norm": 1.1242879629135132, + "learning_rate": 3.360883529854202e-05, + "loss": 0.325, + "step": 42609 + }, + { + "epoch": 2.9567691346887797, + "grad_norm": 0.9644256830215454, + "learning_rate": 3.360463374076117e-05, + "loss": 0.3985, + "step": 42610 + }, + { + "epoch": 2.956838526125876, + "grad_norm": 1.0205941200256348, + "learning_rate": 3.360043239258276e-05, + "loss": 0.4102, + "step": 42611 + }, + { + "epoch": 2.956907917562973, + "grad_norm": 1.370921015739441, + "learning_rate": 3.359623125402007e-05, + "loss": 0.3013, + "step": 42612 + }, + { + "epoch": 2.9569773090000693, + "grad_norm": 1.333446741104126, + "learning_rate": 3.359203032508642e-05, + "loss": 0.5482, + "step": 42613 + }, + { + "epoch": 2.957046700437166, + "grad_norm": 1.0927199125289917, + "learning_rate": 3.3587829605794986e-05, + "loss": 0.3325, + "step": 42614 + }, + { + "epoch": 2.957116091874263, + "grad_norm": 1.113774299621582, + "learning_rate": 3.3583629096159064e-05, + "loss": 0.2741, + "step": 42615 + }, + { + "epoch": 2.9571854833113593, + "grad_norm": 0.9975767135620117, + "learning_rate": 3.3579428796191925e-05, + "loss": 0.4228, + "step": 42616 + }, + { + "epoch": 2.957254874748456, + "grad_norm": 0.9625735282897949, + "learning_rate": 3.3575228705906816e-05, + "loss": 0.3005, + "step": 42617 + }, + { + "epoch": 2.9573242661855526, + "grad_norm": 1.8321295976638794, + "learning_rate": 3.3571028825317005e-05, + "loss": 0.2808, + "step": 42618 + }, + { + "epoch": 2.9573936576226494, + "grad_norm": 1.236584186553955, + "learning_rate": 3.3566829154435795e-05, + "loss": 0.2641, + "step": 42619 + }, + { + "epoch": 2.957463049059746, + "grad_norm": 1.1411770582199097, + "learning_rate": 3.356262969327635e-05, + "loss": 0.3181, + "step": 42620 + }, + { + "epoch": 2.9575324404968426, + "grad_norm": 1.1023691892623901, + "learning_rate": 3.355843044185198e-05, + "loss": 0.3587, + "step": 42621 + }, + { + "epoch": 2.9576018319339394, + "grad_norm": 2.0478930473327637, + "learning_rate": 3.355423140017594e-05, + "loss": 0.4236, + "step": 42622 + }, + { + "epoch": 2.957671223371036, + "grad_norm": 1.2997002601623535, + "learning_rate": 3.355003256826148e-05, + "loss": 0.2668, + "step": 42623 + }, + { + "epoch": 2.9577406148081327, + "grad_norm": 0.9732263088226318, + "learning_rate": 3.354583394612185e-05, + "loss": 0.3874, + "step": 42624 + }, + { + "epoch": 2.9578100062452295, + "grad_norm": 1.1080843210220337, + "learning_rate": 3.354163553377032e-05, + "loss": 0.3995, + "step": 42625 + }, + { + "epoch": 2.957879397682326, + "grad_norm": 1.2355427742004395, + "learning_rate": 3.353743733122017e-05, + "loss": 0.4086, + "step": 42626 + }, + { + "epoch": 2.9579487891194227, + "grad_norm": 1.0411701202392578, + "learning_rate": 3.353323933848458e-05, + "loss": 0.4453, + "step": 42627 + }, + { + "epoch": 2.958018180556519, + "grad_norm": 1.1675012111663818, + "learning_rate": 3.352904155557685e-05, + "loss": 0.2873, + "step": 42628 + }, + { + "epoch": 2.958087571993616, + "grad_norm": 1.4985519647598267, + "learning_rate": 3.3524843982510215e-05, + "loss": 0.3171, + "step": 42629 + }, + { + "epoch": 2.9581569634307128, + "grad_norm": 1.3668652772903442, + "learning_rate": 3.3520646619297946e-05, + "loss": 0.3698, + "step": 42630 + }, + { + "epoch": 2.958226354867809, + "grad_norm": 1.2984228134155273, + "learning_rate": 3.351644946595327e-05, + "loss": 0.3951, + "step": 42631 + }, + { + "epoch": 2.958295746304906, + "grad_norm": 1.0583086013793945, + "learning_rate": 3.3512252522489464e-05, + "loss": 0.4138, + "step": 42632 + }, + { + "epoch": 2.9583651377420024, + "grad_norm": 1.807863712310791, + "learning_rate": 3.350805578891978e-05, + "loss": 0.546, + "step": 42633 + }, + { + "epoch": 2.958434529179099, + "grad_norm": 0.9641680717468262, + "learning_rate": 3.350385926525742e-05, + "loss": 0.3881, + "step": 42634 + }, + { + "epoch": 2.958503920616196, + "grad_norm": 1.5920008420944214, + "learning_rate": 3.3499662951515664e-05, + "loss": 0.4087, + "step": 42635 + }, + { + "epoch": 2.958573312053293, + "grad_norm": 1.329684853553772, + "learning_rate": 3.349546684770776e-05, + "loss": 0.5131, + "step": 42636 + }, + { + "epoch": 2.9586427034903893, + "grad_norm": 1.4533233642578125, + "learning_rate": 3.349127095384695e-05, + "loss": 0.3775, + "step": 42637 + }, + { + "epoch": 2.958712094927486, + "grad_norm": 0.9723921418190002, + "learning_rate": 3.348707526994648e-05, + "loss": 0.2856, + "step": 42638 + }, + { + "epoch": 2.9587814863645825, + "grad_norm": 1.0287362337112427, + "learning_rate": 3.3482879796019595e-05, + "loss": 0.2711, + "step": 42639 + }, + { + "epoch": 2.9588508778016793, + "grad_norm": 1.1459529399871826, + "learning_rate": 3.347868453207957e-05, + "loss": 0.3157, + "step": 42640 + }, + { + "epoch": 2.958920269238776, + "grad_norm": 1.4849544763565063, + "learning_rate": 3.3474489478139605e-05, + "loss": 0.5219, + "step": 42641 + }, + { + "epoch": 2.9589896606758725, + "grad_norm": 1.1118179559707642, + "learning_rate": 3.347029463421294e-05, + "loss": 0.2854, + "step": 42642 + }, + { + "epoch": 2.9590590521129694, + "grad_norm": 1.2648019790649414, + "learning_rate": 3.346610000031284e-05, + "loss": 0.5805, + "step": 42643 + }, + { + "epoch": 2.9591284435500658, + "grad_norm": 1.3555631637573242, + "learning_rate": 3.346190557645255e-05, + "loss": 0.4509, + "step": 42644 + }, + { + "epoch": 2.9591978349871626, + "grad_norm": 1.1702898740768433, + "learning_rate": 3.3457711362645325e-05, + "loss": 0.3184, + "step": 42645 + }, + { + "epoch": 2.9592672264242594, + "grad_norm": 1.1492091417312622, + "learning_rate": 3.345351735890436e-05, + "loss": 0.2551, + "step": 42646 + }, + { + "epoch": 2.959336617861356, + "grad_norm": 0.9276050329208374, + "learning_rate": 3.3449323565242915e-05, + "loss": 0.3424, + "step": 42647 + }, + { + "epoch": 2.9594060092984527, + "grad_norm": 1.192422866821289, + "learning_rate": 3.344512998167427e-05, + "loss": 0.3661, + "step": 42648 + }, + { + "epoch": 2.959475400735549, + "grad_norm": 0.942384660243988, + "learning_rate": 3.344093660821161e-05, + "loss": 0.2524, + "step": 42649 + }, + { + "epoch": 2.959544792172646, + "grad_norm": 1.2057939767837524, + "learning_rate": 3.343674344486817e-05, + "loss": 0.4754, + "step": 42650 + }, + { + "epoch": 2.9596141836097427, + "grad_norm": 1.0235576629638672, + "learning_rate": 3.343255049165721e-05, + "loss": 0.2683, + "step": 42651 + }, + { + "epoch": 2.959683575046839, + "grad_norm": 0.9160522818565369, + "learning_rate": 3.342835774859197e-05, + "loss": 0.2227, + "step": 42652 + }, + { + "epoch": 2.959752966483936, + "grad_norm": 0.8068366646766663, + "learning_rate": 3.3424165215685687e-05, + "loss": 0.2752, + "step": 42653 + }, + { + "epoch": 2.9598223579210323, + "grad_norm": 1.176743984222412, + "learning_rate": 3.341997289295161e-05, + "loss": 0.3618, + "step": 42654 + }, + { + "epoch": 2.959891749358129, + "grad_norm": 1.266984462738037, + "learning_rate": 3.341578078040293e-05, + "loss": 0.373, + "step": 42655 + }, + { + "epoch": 2.959961140795226, + "grad_norm": 0.9317529201507568, + "learning_rate": 3.3411588878052904e-05, + "loss": 0.2417, + "step": 42656 + }, + { + "epoch": 2.9600305322323224, + "grad_norm": 1.1527527570724487, + "learning_rate": 3.340739718591477e-05, + "loss": 0.3936, + "step": 42657 + }, + { + "epoch": 2.960099923669419, + "grad_norm": 1.6211284399032593, + "learning_rate": 3.340320570400175e-05, + "loss": 0.5635, + "step": 42658 + }, + { + "epoch": 2.9601693151065156, + "grad_norm": 1.2820277214050293, + "learning_rate": 3.339901443232708e-05, + "loss": 0.4139, + "step": 42659 + }, + { + "epoch": 2.9602387065436124, + "grad_norm": 1.3857160806655884, + "learning_rate": 3.3394823370903995e-05, + "loss": 0.3124, + "step": 42660 + }, + { + "epoch": 2.9603080979807093, + "grad_norm": 0.9120272994041443, + "learning_rate": 3.3390632519745765e-05, + "loss": 0.209, + "step": 42661 + }, + { + "epoch": 2.960377489417806, + "grad_norm": 1.4658088684082031, + "learning_rate": 3.338644187886555e-05, + "loss": 0.4466, + "step": 42662 + }, + { + "epoch": 2.9604468808549025, + "grad_norm": 1.0706393718719482, + "learning_rate": 3.33822514482766e-05, + "loss": 0.2765, + "step": 42663 + }, + { + "epoch": 2.9605162722919993, + "grad_norm": 0.8127623200416565, + "learning_rate": 3.337806122799215e-05, + "loss": 0.326, + "step": 42664 + }, + { + "epoch": 2.9605856637290957, + "grad_norm": 1.3127504587173462, + "learning_rate": 3.3373871218025444e-05, + "loss": 0.579, + "step": 42665 + }, + { + "epoch": 2.9606550551661925, + "grad_norm": 1.5033843517303467, + "learning_rate": 3.336968141838969e-05, + "loss": 0.3975, + "step": 42666 + }, + { + "epoch": 2.9607244466032894, + "grad_norm": 1.3221079111099243, + "learning_rate": 3.3365491829098115e-05, + "loss": 0.4413, + "step": 42667 + }, + { + "epoch": 2.9607938380403858, + "grad_norm": 1.3057652711868286, + "learning_rate": 3.336130245016399e-05, + "loss": 0.3118, + "step": 42668 + }, + { + "epoch": 2.9608632294774826, + "grad_norm": 1.360795259475708, + "learning_rate": 3.335711328160047e-05, + "loss": 0.3043, + "step": 42669 + }, + { + "epoch": 2.960932620914579, + "grad_norm": 0.8181046843528748, + "learning_rate": 3.335292432342081e-05, + "loss": 0.282, + "step": 42670 + }, + { + "epoch": 2.961002012351676, + "grad_norm": 0.8796766400337219, + "learning_rate": 3.3348735575638235e-05, + "loss": 0.2781, + "step": 42671 + }, + { + "epoch": 2.9610714037887726, + "grad_norm": 0.8086782693862915, + "learning_rate": 3.3344547038265975e-05, + "loss": 0.2544, + "step": 42672 + }, + { + "epoch": 2.961140795225869, + "grad_norm": 1.2644188404083252, + "learning_rate": 3.3340358711317244e-05, + "loss": 0.4549, + "step": 42673 + }, + { + "epoch": 2.961210186662966, + "grad_norm": 1.272581696510315, + "learning_rate": 3.333617059480529e-05, + "loss": 0.3556, + "step": 42674 + }, + { + "epoch": 2.9612795781000623, + "grad_norm": 1.0102663040161133, + "learning_rate": 3.333198268874329e-05, + "loss": 0.35, + "step": 42675 + }, + { + "epoch": 2.961348969537159, + "grad_norm": 1.4305065870285034, + "learning_rate": 3.332779499314448e-05, + "loss": 0.4737, + "step": 42676 + }, + { + "epoch": 2.961418360974256, + "grad_norm": 1.21087646484375, + "learning_rate": 3.3323607508022084e-05, + "loss": 0.3518, + "step": 42677 + }, + { + "epoch": 2.9614877524113523, + "grad_norm": 1.0704593658447266, + "learning_rate": 3.331942023338932e-05, + "loss": 0.3583, + "step": 42678 + }, + { + "epoch": 2.961557143848449, + "grad_norm": 1.33338463306427, + "learning_rate": 3.331523316925942e-05, + "loss": 0.303, + "step": 42679 + }, + { + "epoch": 2.9616265352855455, + "grad_norm": 1.4096601009368896, + "learning_rate": 3.331104631564562e-05, + "loss": 0.5397, + "step": 42680 + }, + { + "epoch": 2.9616959267226424, + "grad_norm": 1.0504590272903442, + "learning_rate": 3.330685967256107e-05, + "loss": 0.3816, + "step": 42681 + }, + { + "epoch": 2.961765318159739, + "grad_norm": 1.1088035106658936, + "learning_rate": 3.3302673240019046e-05, + "loss": 0.3182, + "step": 42682 + }, + { + "epoch": 2.9618347095968356, + "grad_norm": 1.2312875986099243, + "learning_rate": 3.329848701803275e-05, + "loss": 0.2532, + "step": 42683 + }, + { + "epoch": 2.9619041010339324, + "grad_norm": 0.9940983653068542, + "learning_rate": 3.329430100661538e-05, + "loss": 0.3339, + "step": 42684 + }, + { + "epoch": 2.961973492471029, + "grad_norm": 1.2569679021835327, + "learning_rate": 3.329011520578016e-05, + "loss": 0.3907, + "step": 42685 + }, + { + "epoch": 2.9620428839081256, + "grad_norm": 0.9788886308670044, + "learning_rate": 3.32859296155403e-05, + "loss": 0.4946, + "step": 42686 + }, + { + "epoch": 2.9621122753452225, + "grad_norm": 0.9932334423065186, + "learning_rate": 3.328174423590902e-05, + "loss": 0.2731, + "step": 42687 + }, + { + "epoch": 2.9621816667823193, + "grad_norm": 2.250070571899414, + "learning_rate": 3.3277559066899566e-05, + "loss": 0.3582, + "step": 42688 + }, + { + "epoch": 2.9622510582194157, + "grad_norm": 1.3721857070922852, + "learning_rate": 3.327337410852509e-05, + "loss": 0.3923, + "step": 42689 + }, + { + "epoch": 2.9623204496565125, + "grad_norm": 1.0058039426803589, + "learning_rate": 3.326918936079883e-05, + "loss": 0.2165, + "step": 42690 + }, + { + "epoch": 2.962389841093609, + "grad_norm": 1.137210488319397, + "learning_rate": 3.326500482373398e-05, + "loss": 0.3358, + "step": 42691 + }, + { + "epoch": 2.9624592325307058, + "grad_norm": 0.9771386981010437, + "learning_rate": 3.3260820497343784e-05, + "loss": 0.36, + "step": 42692 + }, + { + "epoch": 2.9625286239678026, + "grad_norm": 1.3646970987319946, + "learning_rate": 3.325663638164143e-05, + "loss": 0.4968, + "step": 42693 + }, + { + "epoch": 2.962598015404899, + "grad_norm": 0.6845124959945679, + "learning_rate": 3.3252452476640126e-05, + "loss": 0.2489, + "step": 42694 + }, + { + "epoch": 2.962667406841996, + "grad_norm": 1.121559739112854, + "learning_rate": 3.324826878235313e-05, + "loss": 0.3537, + "step": 42695 + }, + { + "epoch": 2.962736798279092, + "grad_norm": 1.2812869548797607, + "learning_rate": 3.324408529879357e-05, + "loss": 0.5645, + "step": 42696 + }, + { + "epoch": 2.962806189716189, + "grad_norm": 1.1357187032699585, + "learning_rate": 3.323990202597468e-05, + "loss": 0.3991, + "step": 42697 + }, + { + "epoch": 2.962875581153286, + "grad_norm": 1.1804907321929932, + "learning_rate": 3.323571896390968e-05, + "loss": 0.3642, + "step": 42698 + }, + { + "epoch": 2.9629449725903823, + "grad_norm": 1.2313730716705322, + "learning_rate": 3.3231536112611764e-05, + "loss": 0.3468, + "step": 42699 + }, + { + "epoch": 2.963014364027479, + "grad_norm": 1.2092792987823486, + "learning_rate": 3.3227353472094136e-05, + "loss": 0.5502, + "step": 42700 + }, + { + "epoch": 2.9630837554645755, + "grad_norm": 1.133103847503662, + "learning_rate": 3.3223171042370016e-05, + "loss": 0.253, + "step": 42701 + }, + { + "epoch": 2.9631531469016723, + "grad_norm": 1.5307753086090088, + "learning_rate": 3.3218988823452636e-05, + "loss": 0.451, + "step": 42702 + }, + { + "epoch": 2.963222538338769, + "grad_norm": 1.0969551801681519, + "learning_rate": 3.321480681535512e-05, + "loss": 0.4418, + "step": 42703 + }, + { + "epoch": 2.9632919297758655, + "grad_norm": 1.447335124015808, + "learning_rate": 3.3210625018090705e-05, + "loss": 0.4735, + "step": 42704 + }, + { + "epoch": 2.9633613212129624, + "grad_norm": 1.2979981899261475, + "learning_rate": 3.320644343167261e-05, + "loss": 0.6385, + "step": 42705 + }, + { + "epoch": 2.9634307126500588, + "grad_norm": 0.839678168296814, + "learning_rate": 3.320226205611401e-05, + "loss": 0.3458, + "step": 42706 + }, + { + "epoch": 2.9635001040871556, + "grad_norm": 1.1948491334915161, + "learning_rate": 3.319808089142813e-05, + "loss": 0.3949, + "step": 42707 + }, + { + "epoch": 2.9635694955242524, + "grad_norm": 1.4314908981323242, + "learning_rate": 3.319389993762815e-05, + "loss": 0.3497, + "step": 42708 + }, + { + "epoch": 2.963638886961349, + "grad_norm": 1.417956829071045, + "learning_rate": 3.318971919472731e-05, + "loss": 0.3614, + "step": 42709 + }, + { + "epoch": 2.9637082783984456, + "grad_norm": 1.0380927324295044, + "learning_rate": 3.3185538662738744e-05, + "loss": 0.3615, + "step": 42710 + }, + { + "epoch": 2.963777669835542, + "grad_norm": 1.005523443222046, + "learning_rate": 3.318135834167568e-05, + "loss": 0.4156, + "step": 42711 + }, + { + "epoch": 2.963847061272639, + "grad_norm": 1.111707091331482, + "learning_rate": 3.317717823155132e-05, + "loss": 0.331, + "step": 42712 + }, + { + "epoch": 2.9639164527097357, + "grad_norm": 2.151418685913086, + "learning_rate": 3.3172998332378856e-05, + "loss": 0.3846, + "step": 42713 + }, + { + "epoch": 2.9639858441468325, + "grad_norm": 1.1444743871688843, + "learning_rate": 3.316881864417151e-05, + "loss": 0.2668, + "step": 42714 + }, + { + "epoch": 2.964055235583929, + "grad_norm": 1.1440426111221313, + "learning_rate": 3.3164639166942426e-05, + "loss": 0.3339, + "step": 42715 + }, + { + "epoch": 2.9641246270210257, + "grad_norm": 1.0377038717269897, + "learning_rate": 3.316045990070481e-05, + "loss": 0.4035, + "step": 42716 + }, + { + "epoch": 2.964194018458122, + "grad_norm": 1.2274211645126343, + "learning_rate": 3.315628084547188e-05, + "loss": 0.4685, + "step": 42717 + }, + { + "epoch": 2.964263409895219, + "grad_norm": 1.1369671821594238, + "learning_rate": 3.315210200125684e-05, + "loss": 0.3421, + "step": 42718 + }, + { + "epoch": 2.964332801332316, + "grad_norm": 0.9463236331939697, + "learning_rate": 3.314792336807282e-05, + "loss": 0.2699, + "step": 42719 + }, + { + "epoch": 2.964402192769412, + "grad_norm": 1.481870412826538, + "learning_rate": 3.3143744945933056e-05, + "loss": 0.4554, + "step": 42720 + }, + { + "epoch": 2.964471584206509, + "grad_norm": 0.9096069931983948, + "learning_rate": 3.3139566734850735e-05, + "loss": 0.3105, + "step": 42721 + }, + { + "epoch": 2.9645409756436054, + "grad_norm": 0.9875932335853577, + "learning_rate": 3.313538873483904e-05, + "loss": 0.3252, + "step": 42722 + }, + { + "epoch": 2.9646103670807022, + "grad_norm": 1.012380838394165, + "learning_rate": 3.3131210945911196e-05, + "loss": 0.3668, + "step": 42723 + }, + { + "epoch": 2.964679758517799, + "grad_norm": 1.297707200050354, + "learning_rate": 3.312703336808034e-05, + "loss": 0.362, + "step": 42724 + }, + { + "epoch": 2.9647491499548955, + "grad_norm": 1.3113852739334106, + "learning_rate": 3.312285600135967e-05, + "loss": 0.5129, + "step": 42725 + }, + { + "epoch": 2.9648185413919923, + "grad_norm": 0.9752359390258789, + "learning_rate": 3.311867884576238e-05, + "loss": 0.2301, + "step": 42726 + }, + { + "epoch": 2.9648879328290887, + "grad_norm": 1.052729845046997, + "learning_rate": 3.3114501901301676e-05, + "loss": 0.2203, + "step": 42727 + }, + { + "epoch": 2.9649573242661855, + "grad_norm": 1.1423981189727783, + "learning_rate": 3.311032516799072e-05, + "loss": 0.6353, + "step": 42728 + }, + { + "epoch": 2.9650267157032824, + "grad_norm": 1.5638681650161743, + "learning_rate": 3.31061486458427e-05, + "loss": 0.4946, + "step": 42729 + }, + { + "epoch": 2.9650961071403787, + "grad_norm": 1.100867748260498, + "learning_rate": 3.3101972334870856e-05, + "loss": 0.3154, + "step": 42730 + }, + { + "epoch": 2.9651654985774756, + "grad_norm": 0.9516605138778687, + "learning_rate": 3.3097796235088285e-05, + "loss": 0.2567, + "step": 42731 + }, + { + "epoch": 2.965234890014572, + "grad_norm": 1.5330126285552979, + "learning_rate": 3.3093620346508215e-05, + "loss": 0.4606, + "step": 42732 + }, + { + "epoch": 2.965304281451669, + "grad_norm": 1.170284390449524, + "learning_rate": 3.308944466914381e-05, + "loss": 0.521, + "step": 42733 + }, + { + "epoch": 2.9653736728887656, + "grad_norm": 0.89886873960495, + "learning_rate": 3.308526920300827e-05, + "loss": 0.2516, + "step": 42734 + }, + { + "epoch": 2.9654430643258625, + "grad_norm": 0.8832342624664307, + "learning_rate": 3.3081093948114774e-05, + "loss": 0.2493, + "step": 42735 + }, + { + "epoch": 2.965512455762959, + "grad_norm": 1.6069049835205078, + "learning_rate": 3.3076918904476503e-05, + "loss": 0.3845, + "step": 42736 + }, + { + "epoch": 2.9655818472000552, + "grad_norm": 1.3651833534240723, + "learning_rate": 3.307274407210667e-05, + "loss": 0.3987, + "step": 42737 + }, + { + "epoch": 2.965651238637152, + "grad_norm": 1.1571896076202393, + "learning_rate": 3.306856945101837e-05, + "loss": 0.3323, + "step": 42738 + }, + { + "epoch": 2.965720630074249, + "grad_norm": 1.284228801727295, + "learning_rate": 3.306439504122485e-05, + "loss": 0.5084, + "step": 42739 + }, + { + "epoch": 2.9657900215113457, + "grad_norm": 1.551285982131958, + "learning_rate": 3.3060220842739256e-05, + "loss": 0.5347, + "step": 42740 + }, + { + "epoch": 2.965859412948442, + "grad_norm": 1.5788449048995972, + "learning_rate": 3.305604685557478e-05, + "loss": 0.4832, + "step": 42741 + }, + { + "epoch": 2.965928804385539, + "grad_norm": 1.2313286066055298, + "learning_rate": 3.30518730797446e-05, + "loss": 0.3695, + "step": 42742 + }, + { + "epoch": 2.9659981958226354, + "grad_norm": 1.1729035377502441, + "learning_rate": 3.304769951526189e-05, + "loss": 0.3055, + "step": 42743 + }, + { + "epoch": 2.966067587259732, + "grad_norm": 1.0627777576446533, + "learning_rate": 3.304352616213985e-05, + "loss": 0.292, + "step": 42744 + }, + { + "epoch": 2.966136978696829, + "grad_norm": 1.0149714946746826, + "learning_rate": 3.3039353020391606e-05, + "loss": 0.3257, + "step": 42745 + }, + { + "epoch": 2.9662063701339254, + "grad_norm": 1.0827337503433228, + "learning_rate": 3.303518009003035e-05, + "loss": 0.2737, + "step": 42746 + }, + { + "epoch": 2.9662757615710222, + "grad_norm": 1.451094150543213, + "learning_rate": 3.3031007371069266e-05, + "loss": 0.4471, + "step": 42747 + }, + { + "epoch": 2.9663451530081186, + "grad_norm": 1.1953119039535522, + "learning_rate": 3.3026834863521514e-05, + "loss": 0.3367, + "step": 42748 + }, + { + "epoch": 2.9664145444452155, + "grad_norm": 1.0047491788864136, + "learning_rate": 3.302266256740032e-05, + "loss": 0.3934, + "step": 42749 + }, + { + "epoch": 2.9664839358823123, + "grad_norm": 1.3673012256622314, + "learning_rate": 3.3018490482718766e-05, + "loss": 0.4501, + "step": 42750 + }, + { + "epoch": 2.9665533273194087, + "grad_norm": 1.044620394706726, + "learning_rate": 3.301431860949007e-05, + "loss": 0.3712, + "step": 42751 + }, + { + "epoch": 2.9666227187565055, + "grad_norm": 1.090854525566101, + "learning_rate": 3.3010146947727394e-05, + "loss": 0.3955, + "step": 42752 + }, + { + "epoch": 2.966692110193602, + "grad_norm": 1.1741551160812378, + "learning_rate": 3.300597549744395e-05, + "loss": 0.4461, + "step": 42753 + }, + { + "epoch": 2.9667615016306987, + "grad_norm": 1.1940416097640991, + "learning_rate": 3.300180425865284e-05, + "loss": 0.3858, + "step": 42754 + }, + { + "epoch": 2.9668308930677956, + "grad_norm": 1.2957167625427246, + "learning_rate": 3.299763323136724e-05, + "loss": 0.4196, + "step": 42755 + }, + { + "epoch": 2.966900284504892, + "grad_norm": 1.4677205085754395, + "learning_rate": 3.2993462415600366e-05, + "loss": 0.5274, + "step": 42756 + }, + { + "epoch": 2.966969675941989, + "grad_norm": 1.4033011198043823, + "learning_rate": 3.2989291811365374e-05, + "loss": 0.532, + "step": 42757 + }, + { + "epoch": 2.967039067379085, + "grad_norm": 1.3158625364303589, + "learning_rate": 3.2985121418675377e-05, + "loss": 0.419, + "step": 42758 + }, + { + "epoch": 2.967108458816182, + "grad_norm": 1.569950819015503, + "learning_rate": 3.298095123754359e-05, + "loss": 0.4685, + "step": 42759 + }, + { + "epoch": 2.967177850253279, + "grad_norm": 1.258837342262268, + "learning_rate": 3.2976781267983146e-05, + "loss": 0.3584, + "step": 42760 + }, + { + "epoch": 2.9672472416903757, + "grad_norm": 1.119466781616211, + "learning_rate": 3.2972611510007245e-05, + "loss": 0.3759, + "step": 42761 + }, + { + "epoch": 2.967316633127472, + "grad_norm": 1.3786237239837646, + "learning_rate": 3.2968441963629024e-05, + "loss": 0.359, + "step": 42762 + }, + { + "epoch": 2.9673860245645685, + "grad_norm": 7.76230525970459, + "learning_rate": 3.296427262886166e-05, + "loss": 0.3364, + "step": 42763 + }, + { + "epoch": 2.9674554160016653, + "grad_norm": 1.4880889654159546, + "learning_rate": 3.296010350571833e-05, + "loss": 0.3917, + "step": 42764 + }, + { + "epoch": 2.967524807438762, + "grad_norm": 1.9991016387939453, + "learning_rate": 3.295593459421216e-05, + "loss": 0.398, + "step": 42765 + }, + { + "epoch": 2.967594198875859, + "grad_norm": 1.1739501953125, + "learning_rate": 3.2951765894356315e-05, + "loss": 0.3571, + "step": 42766 + }, + { + "epoch": 2.9676635903129553, + "grad_norm": 1.4784040451049805, + "learning_rate": 3.294759740616397e-05, + "loss": 0.4546, + "step": 42767 + }, + { + "epoch": 2.967732981750052, + "grad_norm": 1.6348592042922974, + "learning_rate": 3.294342912964827e-05, + "loss": 0.43, + "step": 42768 + }, + { + "epoch": 2.9678023731871486, + "grad_norm": 1.5378824472427368, + "learning_rate": 3.29392610648224e-05, + "loss": 0.4894, + "step": 42769 + }, + { + "epoch": 2.9678717646242454, + "grad_norm": 1.9718397855758667, + "learning_rate": 3.29350932116995e-05, + "loss": 0.6239, + "step": 42770 + }, + { + "epoch": 2.9679411560613422, + "grad_norm": 0.9786093235015869, + "learning_rate": 3.293092557029275e-05, + "loss": 0.3851, + "step": 42771 + }, + { + "epoch": 2.9680105474984386, + "grad_norm": 1.3604542016983032, + "learning_rate": 3.2926758140615265e-05, + "loss": 0.4672, + "step": 42772 + }, + { + "epoch": 2.9680799389355355, + "grad_norm": 0.7933774590492249, + "learning_rate": 3.292259092268022e-05, + "loss": 0.2736, + "step": 42773 + }, + { + "epoch": 2.968149330372632, + "grad_norm": 1.259392261505127, + "learning_rate": 3.291842391650077e-05, + "loss": 0.3829, + "step": 42774 + }, + { + "epoch": 2.9682187218097287, + "grad_norm": 1.2201849222183228, + "learning_rate": 3.2914257122090075e-05, + "loss": 0.2336, + "step": 42775 + }, + { + "epoch": 2.9682881132468255, + "grad_norm": 1.2873516082763672, + "learning_rate": 3.291009053946128e-05, + "loss": 0.4004, + "step": 42776 + }, + { + "epoch": 2.968357504683922, + "grad_norm": 1.010748267173767, + "learning_rate": 3.2905924168627555e-05, + "loss": 0.2877, + "step": 42777 + }, + { + "epoch": 2.9684268961210187, + "grad_norm": 1.3498872518539429, + "learning_rate": 3.2901758009602076e-05, + "loss": 0.4332, + "step": 42778 + }, + { + "epoch": 2.968496287558115, + "grad_norm": 0.9567951560020447, + "learning_rate": 3.289759206239793e-05, + "loss": 0.4432, + "step": 42779 + }, + { + "epoch": 2.968565678995212, + "grad_norm": 1.493552327156067, + "learning_rate": 3.28934263270283e-05, + "loss": 0.385, + "step": 42780 + }, + { + "epoch": 2.968635070432309, + "grad_norm": 0.9853642582893372, + "learning_rate": 3.288926080350633e-05, + "loss": 0.2795, + "step": 42781 + }, + { + "epoch": 2.968704461869405, + "grad_norm": 1.0720833539962769, + "learning_rate": 3.288509549184517e-05, + "loss": 0.3161, + "step": 42782 + }, + { + "epoch": 2.968773853306502, + "grad_norm": 0.9499545097351074, + "learning_rate": 3.2880930392057996e-05, + "loss": 0.2683, + "step": 42783 + }, + { + "epoch": 2.9688432447435984, + "grad_norm": 1.0630345344543457, + "learning_rate": 3.2876765504157955e-05, + "loss": 0.324, + "step": 42784 + }, + { + "epoch": 2.9689126361806952, + "grad_norm": 1.0426617860794067, + "learning_rate": 3.287260082815816e-05, + "loss": 0.4301, + "step": 42785 + }, + { + "epoch": 2.968982027617792, + "grad_norm": 1.0104113817214966, + "learning_rate": 3.2868436364071766e-05, + "loss": 0.3437, + "step": 42786 + }, + { + "epoch": 2.969051419054889, + "grad_norm": 1.2906275987625122, + "learning_rate": 3.286427211191193e-05, + "loss": 0.3159, + "step": 42787 + }, + { + "epoch": 2.9691208104919853, + "grad_norm": 1.6693848371505737, + "learning_rate": 3.286010807169184e-05, + "loss": 0.4713, + "step": 42788 + }, + { + "epoch": 2.9691902019290817, + "grad_norm": 2.0453109741210938, + "learning_rate": 3.2855944243424555e-05, + "loss": 0.4791, + "step": 42789 + }, + { + "epoch": 2.9692595933661785, + "grad_norm": 1.3304506540298462, + "learning_rate": 3.285178062712327e-05, + "loss": 0.2477, + "step": 42790 + }, + { + "epoch": 2.9693289848032753, + "grad_norm": 1.2798621654510498, + "learning_rate": 3.284761722280112e-05, + "loss": 0.4006, + "step": 42791 + }, + { + "epoch": 2.969398376240372, + "grad_norm": 1.898273229598999, + "learning_rate": 3.284345403047129e-05, + "loss": 0.4129, + "step": 42792 + }, + { + "epoch": 2.9694677676774686, + "grad_norm": 1.1717441082000732, + "learning_rate": 3.2839291050146835e-05, + "loss": 0.5037, + "step": 42793 + }, + { + "epoch": 2.9695371591145654, + "grad_norm": 1.097213625907898, + "learning_rate": 3.283512828184097e-05, + "loss": 0.424, + "step": 42794 + }, + { + "epoch": 2.969606550551662, + "grad_norm": 0.9865909814834595, + "learning_rate": 3.28309657255668e-05, + "loss": 0.3077, + "step": 42795 + }, + { + "epoch": 2.9696759419887586, + "grad_norm": 1.1913295984268188, + "learning_rate": 3.282680338133747e-05, + "loss": 0.4144, + "step": 42796 + }, + { + "epoch": 2.9697453334258554, + "grad_norm": 1.3427348136901855, + "learning_rate": 3.2822641249166144e-05, + "loss": 0.4336, + "step": 42797 + }, + { + "epoch": 2.969814724862952, + "grad_norm": 1.6454460620880127, + "learning_rate": 3.281847932906593e-05, + "loss": 0.3913, + "step": 42798 + }, + { + "epoch": 2.9698841163000487, + "grad_norm": 1.1594189405441284, + "learning_rate": 3.2814317621050014e-05, + "loss": 0.33, + "step": 42799 + }, + { + "epoch": 2.969953507737145, + "grad_norm": 1.1710134744644165, + "learning_rate": 3.2810156125131485e-05, + "loss": 0.3847, + "step": 42800 + }, + { + "epoch": 2.970022899174242, + "grad_norm": 1.0785839557647705, + "learning_rate": 3.280599484132349e-05, + "loss": 0.4547, + "step": 42801 + }, + { + "epoch": 2.9700922906113387, + "grad_norm": 1.2922765016555786, + "learning_rate": 3.280183376963918e-05, + "loss": 0.4143, + "step": 42802 + }, + { + "epoch": 2.970161682048435, + "grad_norm": 1.1667109727859497, + "learning_rate": 3.279767291009167e-05, + "loss": 0.4061, + "step": 42803 + }, + { + "epoch": 2.970231073485532, + "grad_norm": 1.2152670621871948, + "learning_rate": 3.2793512262694126e-05, + "loss": 0.3304, + "step": 42804 + }, + { + "epoch": 2.9703004649226283, + "grad_norm": 0.8861262202262878, + "learning_rate": 3.2789351827459657e-05, + "loss": 0.383, + "step": 42805 + }, + { + "epoch": 2.970369856359725, + "grad_norm": 1.106304407119751, + "learning_rate": 3.2785191604401445e-05, + "loss": 0.3296, + "step": 42806 + }, + { + "epoch": 2.970439247796822, + "grad_norm": 1.8107755184173584, + "learning_rate": 3.278103159353254e-05, + "loss": 0.5394, + "step": 42807 + }, + { + "epoch": 2.9705086392339184, + "grad_norm": 1.6144001483917236, + "learning_rate": 3.277687179486613e-05, + "loss": 0.4416, + "step": 42808 + }, + { + "epoch": 2.9705780306710152, + "grad_norm": 1.5506097078323364, + "learning_rate": 3.2772712208415335e-05, + "loss": 0.344, + "step": 42809 + }, + { + "epoch": 2.9706474221081116, + "grad_norm": 0.864913284778595, + "learning_rate": 3.2768552834193275e-05, + "loss": 0.2683, + "step": 42810 + }, + { + "epoch": 2.9707168135452084, + "grad_norm": 1.3946471214294434, + "learning_rate": 3.2764393672213114e-05, + "loss": 0.1676, + "step": 42811 + }, + { + "epoch": 2.9707862049823053, + "grad_norm": 1.1107993125915527, + "learning_rate": 3.2760234722487945e-05, + "loss": 0.3321, + "step": 42812 + }, + { + "epoch": 2.970855596419402, + "grad_norm": 0.9736884236335754, + "learning_rate": 3.275607598503095e-05, + "loss": 0.4103, + "step": 42813 + }, + { + "epoch": 2.9709249878564985, + "grad_norm": 0.8845785856246948, + "learning_rate": 3.27519174598552e-05, + "loss": 0.2749, + "step": 42814 + }, + { + "epoch": 2.9709943792935953, + "grad_norm": 1.3929955959320068, + "learning_rate": 3.2747759146973836e-05, + "loss": 0.4916, + "step": 42815 + }, + { + "epoch": 2.9710637707306917, + "grad_norm": 1.333426594734192, + "learning_rate": 3.2743601046399996e-05, + "loss": 0.4301, + "step": 42816 + }, + { + "epoch": 2.9711331621677886, + "grad_norm": 1.0728987455368042, + "learning_rate": 3.2739443158146795e-05, + "loss": 0.3652, + "step": 42817 + }, + { + "epoch": 2.9712025536048854, + "grad_norm": 1.05364191532135, + "learning_rate": 3.273528548222739e-05, + "loss": 0.3773, + "step": 42818 + }, + { + "epoch": 2.9712719450419818, + "grad_norm": 1.2675960063934326, + "learning_rate": 3.27311280186549e-05, + "loss": 0.4491, + "step": 42819 + }, + { + "epoch": 2.9713413364790786, + "grad_norm": 1.469659447669983, + "learning_rate": 3.27269707674424e-05, + "loss": 0.3851, + "step": 42820 + }, + { + "epoch": 2.971410727916175, + "grad_norm": 1.8832087516784668, + "learning_rate": 3.272281372860305e-05, + "loss": 0.4359, + "step": 42821 + }, + { + "epoch": 2.971480119353272, + "grad_norm": 0.8587567806243896, + "learning_rate": 3.271865690214998e-05, + "loss": 0.2249, + "step": 42822 + }, + { + "epoch": 2.9715495107903687, + "grad_norm": 1.4021130800247192, + "learning_rate": 3.271450028809633e-05, + "loss": 0.5163, + "step": 42823 + }, + { + "epoch": 2.971618902227465, + "grad_norm": 1.1732430458068848, + "learning_rate": 3.2710343886455164e-05, + "loss": 0.3874, + "step": 42824 + }, + { + "epoch": 2.971688293664562, + "grad_norm": 1.1460634469985962, + "learning_rate": 3.270618769723963e-05, + "loss": 0.2988, + "step": 42825 + }, + { + "epoch": 2.9717576851016583, + "grad_norm": 1.2424174547195435, + "learning_rate": 3.27020317204629e-05, + "loss": 0.3465, + "step": 42826 + }, + { + "epoch": 2.971827076538755, + "grad_norm": 0.8990272283554077, + "learning_rate": 3.2697875956138e-05, + "loss": 0.213, + "step": 42827 + }, + { + "epoch": 2.971896467975852, + "grad_norm": 1.1555334329605103, + "learning_rate": 3.26937204042781e-05, + "loss": 0.2899, + "step": 42828 + }, + { + "epoch": 2.9719658594129483, + "grad_norm": 1.0479907989501953, + "learning_rate": 3.2689565064896334e-05, + "loss": 0.3798, + "step": 42829 + }, + { + "epoch": 2.972035250850045, + "grad_norm": 1.1180250644683838, + "learning_rate": 3.2685409938005785e-05, + "loss": 0.3516, + "step": 42830 + }, + { + "epoch": 2.9721046422871416, + "grad_norm": 1.4070484638214111, + "learning_rate": 3.268125502361959e-05, + "loss": 0.241, + "step": 42831 + }, + { + "epoch": 2.9721740337242384, + "grad_norm": 0.9549462795257568, + "learning_rate": 3.267710032175085e-05, + "loss": 0.2282, + "step": 42832 + }, + { + "epoch": 2.972243425161335, + "grad_norm": 0.9264118075370789, + "learning_rate": 3.267294583241275e-05, + "loss": 0.3284, + "step": 42833 + }, + { + "epoch": 2.9723128165984316, + "grad_norm": 0.8084405064582825, + "learning_rate": 3.2668791555618306e-05, + "loss": 0.378, + "step": 42834 + }, + { + "epoch": 2.9723822080355284, + "grad_norm": 0.9317237734794617, + "learning_rate": 3.2664637491380665e-05, + "loss": 0.2415, + "step": 42835 + }, + { + "epoch": 2.972451599472625, + "grad_norm": 1.3172399997711182, + "learning_rate": 3.266048363971296e-05, + "loss": 0.5649, + "step": 42836 + }, + { + "epoch": 2.9725209909097217, + "grad_norm": 1.2807245254516602, + "learning_rate": 3.26563300006283e-05, + "loss": 0.4411, + "step": 42837 + }, + { + "epoch": 2.9725903823468185, + "grad_norm": 0.9921834468841553, + "learning_rate": 3.265217657413978e-05, + "loss": 0.4118, + "step": 42838 + }, + { + "epoch": 2.9726597737839153, + "grad_norm": 0.8277469873428345, + "learning_rate": 3.264802336026054e-05, + "loss": 0.2231, + "step": 42839 + }, + { + "epoch": 2.9727291652210117, + "grad_norm": 1.5613558292388916, + "learning_rate": 3.2643870359003706e-05, + "loss": 0.3069, + "step": 42840 + }, + { + "epoch": 2.9727985566581085, + "grad_norm": 0.9607189893722534, + "learning_rate": 3.263971757038231e-05, + "loss": 0.3446, + "step": 42841 + }, + { + "epoch": 2.972867948095205, + "grad_norm": 1.8680840730667114, + "learning_rate": 3.263556499440953e-05, + "loss": 0.3582, + "step": 42842 + }, + { + "epoch": 2.9729373395323018, + "grad_norm": 1.198831558227539, + "learning_rate": 3.2631412631098445e-05, + "loss": 0.5002, + "step": 42843 + }, + { + "epoch": 2.9730067309693986, + "grad_norm": 1.5122753381729126, + "learning_rate": 3.2627260480462176e-05, + "loss": 0.627, + "step": 42844 + }, + { + "epoch": 2.973076122406495, + "grad_norm": 1.228419303894043, + "learning_rate": 3.2623108542513826e-05, + "loss": 0.2904, + "step": 42845 + }, + { + "epoch": 2.973145513843592, + "grad_norm": 1.278681755065918, + "learning_rate": 3.261895681726651e-05, + "loss": 0.3999, + "step": 42846 + }, + { + "epoch": 2.973214905280688, + "grad_norm": 0.7718316912651062, + "learning_rate": 3.2614805304733356e-05, + "loss": 0.2413, + "step": 42847 + }, + { + "epoch": 2.973284296717785, + "grad_norm": 1.1944183111190796, + "learning_rate": 3.261065400492742e-05, + "loss": 0.2915, + "step": 42848 + }, + { + "epoch": 2.973353688154882, + "grad_norm": 1.853063702583313, + "learning_rate": 3.260650291786183e-05, + "loss": 0.3357, + "step": 42849 + }, + { + "epoch": 2.9734230795919783, + "grad_norm": 0.8914254903793335, + "learning_rate": 3.260235204354968e-05, + "loss": 0.2747, + "step": 42850 + }, + { + "epoch": 2.973492471029075, + "grad_norm": 0.7479250431060791, + "learning_rate": 3.259820138200409e-05, + "loss": 0.4036, + "step": 42851 + }, + { + "epoch": 2.9735618624661715, + "grad_norm": 1.0777666568756104, + "learning_rate": 3.2594050933238164e-05, + "loss": 0.5217, + "step": 42852 + }, + { + "epoch": 2.9736312539032683, + "grad_norm": 1.0962918996810913, + "learning_rate": 3.2589900697265e-05, + "loss": 0.3273, + "step": 42853 + }, + { + "epoch": 2.973700645340365, + "grad_norm": 1.1820683479309082, + "learning_rate": 3.2585750674097726e-05, + "loss": 0.4571, + "step": 42854 + }, + { + "epoch": 2.9737700367774615, + "grad_norm": 0.7885865569114685, + "learning_rate": 3.258160086374937e-05, + "loss": 0.199, + "step": 42855 + }, + { + "epoch": 2.9738394282145584, + "grad_norm": 0.938089907169342, + "learning_rate": 3.257745126623309e-05, + "loss": 0.3477, + "step": 42856 + }, + { + "epoch": 2.9739088196516548, + "grad_norm": 1.8516128063201904, + "learning_rate": 3.257330188156201e-05, + "loss": 0.4243, + "step": 42857 + }, + { + "epoch": 2.9739782110887516, + "grad_norm": 0.8199607133865356, + "learning_rate": 3.256915270974916e-05, + "loss": 0.311, + "step": 42858 + }, + { + "epoch": 2.9740476025258484, + "grad_norm": 1.1527929306030273, + "learning_rate": 3.256500375080768e-05, + "loss": 0.4219, + "step": 42859 + }, + { + "epoch": 2.974116993962945, + "grad_norm": 0.8213740587234497, + "learning_rate": 3.256085500475065e-05, + "loss": 0.3141, + "step": 42860 + }, + { + "epoch": 2.9741863854000417, + "grad_norm": 1.2563512325286865, + "learning_rate": 3.25567064715912e-05, + "loss": 0.4641, + "step": 42861 + }, + { + "epoch": 2.974255776837138, + "grad_norm": 1.5511420965194702, + "learning_rate": 3.255255815134238e-05, + "loss": 0.3406, + "step": 42862 + }, + { + "epoch": 2.974325168274235, + "grad_norm": 1.054431438446045, + "learning_rate": 3.254841004401731e-05, + "loss": 0.307, + "step": 42863 + }, + { + "epoch": 2.9743945597113317, + "grad_norm": 0.8572244048118591, + "learning_rate": 3.254426214962908e-05, + "loss": 0.3166, + "step": 42864 + }, + { + "epoch": 2.9744639511484285, + "grad_norm": 1.934453010559082, + "learning_rate": 3.254011446819078e-05, + "loss": 0.4068, + "step": 42865 + }, + { + "epoch": 2.974533342585525, + "grad_norm": 1.0543164014816284, + "learning_rate": 3.253596699971553e-05, + "loss": 0.3764, + "step": 42866 + }, + { + "epoch": 2.9746027340226218, + "grad_norm": 1.2033191919326782, + "learning_rate": 3.25318197442164e-05, + "loss": 0.4137, + "step": 42867 + }, + { + "epoch": 2.974672125459718, + "grad_norm": 0.987166166305542, + "learning_rate": 3.25276727017065e-05, + "loss": 0.3655, + "step": 42868 + }, + { + "epoch": 2.974741516896815, + "grad_norm": 1.9622278213500977, + "learning_rate": 3.25235258721989e-05, + "loss": 0.5837, + "step": 42869 + }, + { + "epoch": 2.974810908333912, + "grad_norm": 0.8700876832008362, + "learning_rate": 3.251937925570669e-05, + "loss": 0.3167, + "step": 42870 + }, + { + "epoch": 2.974880299771008, + "grad_norm": 0.8745121955871582, + "learning_rate": 3.251523285224297e-05, + "loss": 0.3067, + "step": 42871 + }, + { + "epoch": 2.974949691208105, + "grad_norm": 0.8045698404312134, + "learning_rate": 3.2511086661820834e-05, + "loss": 0.2809, + "step": 42872 + }, + { + "epoch": 2.9750190826452014, + "grad_norm": 1.1904067993164062, + "learning_rate": 3.250694068445337e-05, + "loss": 0.3251, + "step": 42873 + }, + { + "epoch": 2.9750884740822983, + "grad_norm": 1.64844810962677, + "learning_rate": 3.2502794920153656e-05, + "loss": 0.333, + "step": 42874 + }, + { + "epoch": 2.975157865519395, + "grad_norm": 1.3523778915405273, + "learning_rate": 3.249864936893482e-05, + "loss": 0.3589, + "step": 42875 + }, + { + "epoch": 2.9752272569564915, + "grad_norm": 0.8136110305786133, + "learning_rate": 3.249450403080989e-05, + "loss": 0.3379, + "step": 42876 + }, + { + "epoch": 2.9752966483935883, + "grad_norm": 1.4718748331069946, + "learning_rate": 3.249035890579197e-05, + "loss": 0.5468, + "step": 42877 + }, + { + "epoch": 2.9753660398306847, + "grad_norm": 1.703614592552185, + "learning_rate": 3.2486213993894165e-05, + "loss": 0.4166, + "step": 42878 + }, + { + "epoch": 2.9754354312677815, + "grad_norm": 1.4700571298599243, + "learning_rate": 3.248206929512955e-05, + "loss": 0.4978, + "step": 42879 + }, + { + "epoch": 2.9755048227048784, + "grad_norm": 1.5456910133361816, + "learning_rate": 3.247792480951119e-05, + "loss": 0.3769, + "step": 42880 + }, + { + "epoch": 2.9755742141419748, + "grad_norm": 0.7977350354194641, + "learning_rate": 3.24737805370522e-05, + "loss": 0.241, + "step": 42881 + }, + { + "epoch": 2.9756436055790716, + "grad_norm": 0.6552748084068298, + "learning_rate": 3.2469636477765685e-05, + "loss": 0.1888, + "step": 42882 + }, + { + "epoch": 2.975712997016168, + "grad_norm": 1.9445390701293945, + "learning_rate": 3.2465492631664664e-05, + "loss": 0.3288, + "step": 42883 + }, + { + "epoch": 2.975782388453265, + "grad_norm": 0.8820446133613586, + "learning_rate": 3.246134899876224e-05, + "loss": 0.2962, + "step": 42884 + }, + { + "epoch": 2.9758517798903616, + "grad_norm": 1.0658807754516602, + "learning_rate": 3.2457205579071505e-05, + "loss": 0.2871, + "step": 42885 + }, + { + "epoch": 2.975921171327458, + "grad_norm": 1.1601321697235107, + "learning_rate": 3.245306237260554e-05, + "loss": 0.403, + "step": 42886 + }, + { + "epoch": 2.975990562764555, + "grad_norm": 1.4768245220184326, + "learning_rate": 3.244891937937741e-05, + "loss": 0.3393, + "step": 42887 + }, + { + "epoch": 2.9760599542016513, + "grad_norm": 1.1295799016952515, + "learning_rate": 3.244477659940021e-05, + "loss": 0.2868, + "step": 42888 + }, + { + "epoch": 2.976129345638748, + "grad_norm": 0.7621057033538818, + "learning_rate": 3.244063403268703e-05, + "loss": 0.3167, + "step": 42889 + }, + { + "epoch": 2.976198737075845, + "grad_norm": 1.5636439323425293, + "learning_rate": 3.2436491679250915e-05, + "loss": 0.5629, + "step": 42890 + }, + { + "epoch": 2.9762681285129418, + "grad_norm": 1.2313066720962524, + "learning_rate": 3.243234953910495e-05, + "loss": 0.3733, + "step": 42891 + }, + { + "epoch": 2.976337519950038, + "grad_norm": 1.6175886392593384, + "learning_rate": 3.242820761226225e-05, + "loss": 0.563, + "step": 42892 + }, + { + "epoch": 2.976406911387135, + "grad_norm": 1.4084450006484985, + "learning_rate": 3.2424065898735836e-05, + "loss": 0.3971, + "step": 42893 + }, + { + "epoch": 2.9764763028242314, + "grad_norm": 0.9207173585891724, + "learning_rate": 3.2419924398538804e-05, + "loss": 0.2869, + "step": 42894 + }, + { + "epoch": 2.976545694261328, + "grad_norm": 1.3357161283493042, + "learning_rate": 3.241578311168422e-05, + "loss": 0.4639, + "step": 42895 + }, + { + "epoch": 2.976615085698425, + "grad_norm": 0.8150368332862854, + "learning_rate": 3.241164203818521e-05, + "loss": 0.2654, + "step": 42896 + }, + { + "epoch": 2.9766844771355214, + "grad_norm": 1.0187050104141235, + "learning_rate": 3.2407501178054756e-05, + "loss": 0.2602, + "step": 42897 + }, + { + "epoch": 2.9767538685726183, + "grad_norm": 0.9760719537734985, + "learning_rate": 3.2403360531305996e-05, + "loss": 0.3126, + "step": 42898 + }, + { + "epoch": 2.9768232600097146, + "grad_norm": 1.1955697536468506, + "learning_rate": 3.2399220097951975e-05, + "loss": 0.2889, + "step": 42899 + }, + { + "epoch": 2.9768926514468115, + "grad_norm": 1.8028632402420044, + "learning_rate": 3.239507987800577e-05, + "loss": 0.5228, + "step": 42900 + }, + { + "epoch": 2.9769620428839083, + "grad_norm": 0.9609285593032837, + "learning_rate": 3.239093987148046e-05, + "loss": 0.3573, + "step": 42901 + }, + { + "epoch": 2.9770314343210047, + "grad_norm": 1.3493504524230957, + "learning_rate": 3.2386800078389145e-05, + "loss": 0.3343, + "step": 42902 + }, + { + "epoch": 2.9771008257581015, + "grad_norm": 0.8893547058105469, + "learning_rate": 3.238266049874482e-05, + "loss": 0.3055, + "step": 42903 + }, + { + "epoch": 2.977170217195198, + "grad_norm": 1.1375746726989746, + "learning_rate": 3.237852113256059e-05, + "loss": 0.3512, + "step": 42904 + }, + { + "epoch": 2.9772396086322948, + "grad_norm": 1.3438628911972046, + "learning_rate": 3.237438197984953e-05, + "loss": 0.4244, + "step": 42905 + }, + { + "epoch": 2.9773090000693916, + "grad_norm": 1.0399668216705322, + "learning_rate": 3.237024304062469e-05, + "loss": 0.3819, + "step": 42906 + }, + { + "epoch": 2.977378391506488, + "grad_norm": 1.1522239446640015, + "learning_rate": 3.236610431489916e-05, + "loss": 0.4153, + "step": 42907 + }, + { + "epoch": 2.977447782943585, + "grad_norm": 1.5730637311935425, + "learning_rate": 3.236196580268598e-05, + "loss": 0.4552, + "step": 42908 + }, + { + "epoch": 2.977517174380681, + "grad_norm": 1.0261669158935547, + "learning_rate": 3.235782750399826e-05, + "loss": 0.3708, + "step": 42909 + }, + { + "epoch": 2.977586565817778, + "grad_norm": 1.38441002368927, + "learning_rate": 3.2353689418848995e-05, + "loss": 0.3936, + "step": 42910 + }, + { + "epoch": 2.977655957254875, + "grad_norm": 1.6787201166152954, + "learning_rate": 3.234955154725129e-05, + "loss": 0.4819, + "step": 42911 + }, + { + "epoch": 2.9777253486919713, + "grad_norm": 1.2263625860214233, + "learning_rate": 3.23454138892182e-05, + "loss": 0.3478, + "step": 42912 + }, + { + "epoch": 2.977794740129068, + "grad_norm": 1.4582158327102661, + "learning_rate": 3.234127644476279e-05, + "loss": 0.4132, + "step": 42913 + }, + { + "epoch": 2.9778641315661645, + "grad_norm": 1.312116265296936, + "learning_rate": 3.2337139213898126e-05, + "loss": 0.4471, + "step": 42914 + }, + { + "epoch": 2.9779335230032613, + "grad_norm": 0.8775119781494141, + "learning_rate": 3.233300219663725e-05, + "loss": 0.3374, + "step": 42915 + }, + { + "epoch": 2.978002914440358, + "grad_norm": 1.0523581504821777, + "learning_rate": 3.232886539299328e-05, + "loss": 0.323, + "step": 42916 + }, + { + "epoch": 2.978072305877455, + "grad_norm": 1.3349372148513794, + "learning_rate": 3.23247288029792e-05, + "loss": 0.322, + "step": 42917 + }, + { + "epoch": 2.9781416973145514, + "grad_norm": 1.1691462993621826, + "learning_rate": 3.2320592426608085e-05, + "loss": 0.403, + "step": 42918 + }, + { + "epoch": 2.978211088751648, + "grad_norm": 1.151538372039795, + "learning_rate": 3.2316456263893014e-05, + "loss": 0.428, + "step": 42919 + }, + { + "epoch": 2.9782804801887446, + "grad_norm": 0.8786848187446594, + "learning_rate": 3.2312320314847046e-05, + "loss": 0.3279, + "step": 42920 + }, + { + "epoch": 2.9783498716258414, + "grad_norm": 1.1574283838272095, + "learning_rate": 3.2308184579483225e-05, + "loss": 0.3858, + "step": 42921 + }, + { + "epoch": 2.9784192630629382, + "grad_norm": 0.9055167436599731, + "learning_rate": 3.23040490578146e-05, + "loss": 0.3247, + "step": 42922 + }, + { + "epoch": 2.9784886545000346, + "grad_norm": 1.6460564136505127, + "learning_rate": 3.229991374985428e-05, + "loss": 0.4847, + "step": 42923 + }, + { + "epoch": 2.9785580459371315, + "grad_norm": 1.6249120235443115, + "learning_rate": 3.229577865561525e-05, + "loss": 0.4027, + "step": 42924 + }, + { + "epoch": 2.978627437374228, + "grad_norm": 1.3534647226333618, + "learning_rate": 3.229164377511058e-05, + "loss": 0.5057, + "step": 42925 + }, + { + "epoch": 2.9786968288113247, + "grad_norm": 1.0893573760986328, + "learning_rate": 3.228750910835334e-05, + "loss": 0.3369, + "step": 42926 + }, + { + "epoch": 2.9787662202484215, + "grad_norm": 2.284787893295288, + "learning_rate": 3.2283374655356614e-05, + "loss": 0.4345, + "step": 42927 + }, + { + "epoch": 2.978835611685518, + "grad_norm": 0.805646538734436, + "learning_rate": 3.2279240416133386e-05, + "loss": 0.2376, + "step": 42928 + }, + { + "epoch": 2.9789050031226147, + "grad_norm": 1.1568975448608398, + "learning_rate": 3.227510639069674e-05, + "loss": 0.3926, + "step": 42929 + }, + { + "epoch": 2.978974394559711, + "grad_norm": 1.0986117124557495, + "learning_rate": 3.227097257905972e-05, + "loss": 0.4309, + "step": 42930 + }, + { + "epoch": 2.979043785996808, + "grad_norm": 1.231033205986023, + "learning_rate": 3.2266838981235416e-05, + "loss": 0.362, + "step": 42931 + }, + { + "epoch": 2.979113177433905, + "grad_norm": 1.6836289167404175, + "learning_rate": 3.226270559723681e-05, + "loss": 0.4657, + "step": 42932 + }, + { + "epoch": 2.979182568871001, + "grad_norm": 1.1523361206054688, + "learning_rate": 3.225857242707698e-05, + "loss": 0.38, + "step": 42933 + }, + { + "epoch": 2.979251960308098, + "grad_norm": 1.7665555477142334, + "learning_rate": 3.225443947076898e-05, + "loss": 0.4471, + "step": 42934 + }, + { + "epoch": 2.9793213517451944, + "grad_norm": 1.5230648517608643, + "learning_rate": 3.2250306728325855e-05, + "loss": 0.5855, + "step": 42935 + }, + { + "epoch": 2.9793907431822912, + "grad_norm": 1.2701598405838013, + "learning_rate": 3.224617419976065e-05, + "loss": 0.3264, + "step": 42936 + }, + { + "epoch": 2.979460134619388, + "grad_norm": 0.9433780908584595, + "learning_rate": 3.224204188508645e-05, + "loss": 0.2793, + "step": 42937 + }, + { + "epoch": 2.9795295260564845, + "grad_norm": 1.0974723100662231, + "learning_rate": 3.2237909784316225e-05, + "loss": 0.3567, + "step": 42938 + }, + { + "epoch": 2.9795989174935813, + "grad_norm": 1.15776526927948, + "learning_rate": 3.223377789746306e-05, + "loss": 0.4022, + "step": 42939 + }, + { + "epoch": 2.9796683089306777, + "grad_norm": 1.1063557863235474, + "learning_rate": 3.222964622453999e-05, + "loss": 0.4249, + "step": 42940 + }, + { + "epoch": 2.9797377003677745, + "grad_norm": 1.2676275968551636, + "learning_rate": 3.222551476556007e-05, + "loss": 0.6231, + "step": 42941 + }, + { + "epoch": 2.9798070918048714, + "grad_norm": 1.4247913360595703, + "learning_rate": 3.222138352053633e-05, + "loss": 0.4508, + "step": 42942 + }, + { + "epoch": 2.979876483241968, + "grad_norm": 1.1480977535247803, + "learning_rate": 3.221725248948182e-05, + "loss": 0.3231, + "step": 42943 + }, + { + "epoch": 2.9799458746790646, + "grad_norm": 1.4314998388290405, + "learning_rate": 3.221312167240962e-05, + "loss": 0.5471, + "step": 42944 + }, + { + "epoch": 2.9800152661161614, + "grad_norm": 1.5125157833099365, + "learning_rate": 3.2208991069332694e-05, + "loss": 0.5578, + "step": 42945 + }, + { + "epoch": 2.980084657553258, + "grad_norm": 0.7805265784263611, + "learning_rate": 3.220486068026412e-05, + "loss": 0.2349, + "step": 42946 + }, + { + "epoch": 2.9801540489903546, + "grad_norm": 1.2100350856781006, + "learning_rate": 3.220073050521693e-05, + "loss": 0.5053, + "step": 42947 + }, + { + "epoch": 2.9802234404274515, + "grad_norm": 1.184118390083313, + "learning_rate": 3.219660054420417e-05, + "loss": 0.416, + "step": 42948 + }, + { + "epoch": 2.980292831864548, + "grad_norm": 1.1453160047531128, + "learning_rate": 3.2192470797238874e-05, + "loss": 0.2865, + "step": 42949 + }, + { + "epoch": 2.9803622233016447, + "grad_norm": 1.3090050220489502, + "learning_rate": 3.218834126433408e-05, + "loss": 0.4324, + "step": 42950 + }, + { + "epoch": 2.980431614738741, + "grad_norm": 0.6765922904014587, + "learning_rate": 3.218421194550285e-05, + "loss": 0.2273, + "step": 42951 + }, + { + "epoch": 2.980501006175838, + "grad_norm": 1.172221064567566, + "learning_rate": 3.218008284075817e-05, + "loss": 0.4147, + "step": 42952 + }, + { + "epoch": 2.9805703976129347, + "grad_norm": 1.18870210647583, + "learning_rate": 3.21759539501131e-05, + "loss": 0.439, + "step": 42953 + }, + { + "epoch": 2.980639789050031, + "grad_norm": 0.7061259150505066, + "learning_rate": 3.217182527358068e-05, + "loss": 0.2223, + "step": 42954 + }, + { + "epoch": 2.980709180487128, + "grad_norm": 1.1849933862686157, + "learning_rate": 3.216769681117393e-05, + "loss": 0.3473, + "step": 42955 + }, + { + "epoch": 2.9807785719242244, + "grad_norm": 1.753690481185913, + "learning_rate": 3.21635685629059e-05, + "loss": 0.24, + "step": 42956 + }, + { + "epoch": 2.980847963361321, + "grad_norm": 1.1097252368927002, + "learning_rate": 3.215944052878961e-05, + "loss": 0.3509, + "step": 42957 + }, + { + "epoch": 2.980917354798418, + "grad_norm": 1.0591437816619873, + "learning_rate": 3.2155312708838126e-05, + "loss": 0.3299, + "step": 42958 + }, + { + "epoch": 2.9809867462355144, + "grad_norm": 1.3223495483398438, + "learning_rate": 3.215118510306442e-05, + "loss": 0.4555, + "step": 42959 + }, + { + "epoch": 2.9810561376726112, + "grad_norm": 1.1186211109161377, + "learning_rate": 3.2147057711481545e-05, + "loss": 0.3644, + "step": 42960 + }, + { + "epoch": 2.9811255291097076, + "grad_norm": 1.664115071296692, + "learning_rate": 3.2142930534102536e-05, + "loss": 0.444, + "step": 42961 + }, + { + "epoch": 2.9811949205468045, + "grad_norm": 1.0178675651550293, + "learning_rate": 3.213880357094046e-05, + "loss": 0.336, + "step": 42962 + }, + { + "epoch": 2.9812643119839013, + "grad_norm": 1.2662807703018188, + "learning_rate": 3.213467682200827e-05, + "loss": 0.383, + "step": 42963 + }, + { + "epoch": 2.9813337034209977, + "grad_norm": 1.2120407819747925, + "learning_rate": 3.213055028731903e-05, + "loss": 0.3819, + "step": 42964 + }, + { + "epoch": 2.9814030948580945, + "grad_norm": 1.0637308359146118, + "learning_rate": 3.2126423966885766e-05, + "loss": 0.3575, + "step": 42965 + }, + { + "epoch": 2.981472486295191, + "grad_norm": 0.9752535820007324, + "learning_rate": 3.212229786072155e-05, + "loss": 0.2509, + "step": 42966 + }, + { + "epoch": 2.9815418777322877, + "grad_norm": 1.1753463745117188, + "learning_rate": 3.211817196883933e-05, + "loss": 0.3463, + "step": 42967 + }, + { + "epoch": 2.9816112691693846, + "grad_norm": 0.9626726508140564, + "learning_rate": 3.211404629125217e-05, + "loss": 0.2827, + "step": 42968 + }, + { + "epoch": 2.9816806606064814, + "grad_norm": 1.1306077241897583, + "learning_rate": 3.210992082797307e-05, + "loss": 0.3335, + "step": 42969 + }, + { + "epoch": 2.981750052043578, + "grad_norm": 0.9422601461410522, + "learning_rate": 3.210579557901509e-05, + "loss": 0.3415, + "step": 42970 + }, + { + "epoch": 2.9818194434806746, + "grad_norm": 0.6691814064979553, + "learning_rate": 3.210167054439123e-05, + "loss": 0.2226, + "step": 42971 + }, + { + "epoch": 2.981888834917771, + "grad_norm": 1.1638143062591553, + "learning_rate": 3.2097545724114554e-05, + "loss": 0.2355, + "step": 42972 + }, + { + "epoch": 2.981958226354868, + "grad_norm": 1.3645128011703491, + "learning_rate": 3.209342111819801e-05, + "loss": 0.5305, + "step": 42973 + }, + { + "epoch": 2.9820276177919647, + "grad_norm": 1.1327764987945557, + "learning_rate": 3.208929672665466e-05, + "loss": 0.5017, + "step": 42974 + }, + { + "epoch": 2.982097009229061, + "grad_norm": 1.3779622316360474, + "learning_rate": 3.208517254949752e-05, + "loss": 0.6675, + "step": 42975 + }, + { + "epoch": 2.982166400666158, + "grad_norm": 1.2988076210021973, + "learning_rate": 3.208104858673961e-05, + "loss": 0.561, + "step": 42976 + }, + { + "epoch": 2.9822357921032543, + "grad_norm": 1.1600229740142822, + "learning_rate": 3.207692483839396e-05, + "loss": 0.3744, + "step": 42977 + }, + { + "epoch": 2.982305183540351, + "grad_norm": 1.170028805732727, + "learning_rate": 3.20728013044736e-05, + "loss": 0.3315, + "step": 42978 + }, + { + "epoch": 2.982374574977448, + "grad_norm": 1.2508933544158936, + "learning_rate": 3.206867798499149e-05, + "loss": 0.3249, + "step": 42979 + }, + { + "epoch": 2.9824439664145443, + "grad_norm": 1.2191088199615479, + "learning_rate": 3.2064554879960695e-05, + "loss": 0.2584, + "step": 42980 + }, + { + "epoch": 2.982513357851641, + "grad_norm": 1.3690375089645386, + "learning_rate": 3.2060431989394216e-05, + "loss": 0.444, + "step": 42981 + }, + { + "epoch": 2.9825827492887376, + "grad_norm": 0.9569831490516663, + "learning_rate": 3.205630931330506e-05, + "loss": 0.1444, + "step": 42982 + }, + { + "epoch": 2.9826521407258344, + "grad_norm": 0.8519121408462524, + "learning_rate": 3.205218685170626e-05, + "loss": 0.3306, + "step": 42983 + }, + { + "epoch": 2.9827215321629312, + "grad_norm": 1.152133822441101, + "learning_rate": 3.2048064604610814e-05, + "loss": 0.435, + "step": 42984 + }, + { + "epoch": 2.9827909236000276, + "grad_norm": 1.0989032983779907, + "learning_rate": 3.204394257203178e-05, + "loss": 0.2699, + "step": 42985 + }, + { + "epoch": 2.9828603150371245, + "grad_norm": 2.422283887863159, + "learning_rate": 3.203982075398211e-05, + "loss": 0.3713, + "step": 42986 + }, + { + "epoch": 2.982929706474221, + "grad_norm": 2.0102121829986572, + "learning_rate": 3.203569915047484e-05, + "loss": 0.2688, + "step": 42987 + }, + { + "epoch": 2.9829990979113177, + "grad_norm": 1.0309982299804688, + "learning_rate": 3.203157776152298e-05, + "loss": 0.4263, + "step": 42988 + }, + { + "epoch": 2.9830684893484145, + "grad_norm": 1.0903807878494263, + "learning_rate": 3.202745658713954e-05, + "loss": 0.3937, + "step": 42989 + }, + { + "epoch": 2.9831378807855113, + "grad_norm": 0.7360156178474426, + "learning_rate": 3.2023335627337535e-05, + "loss": 0.3068, + "step": 42990 + }, + { + "epoch": 2.9832072722226077, + "grad_norm": 0.7692247629165649, + "learning_rate": 3.201921488212998e-05, + "loss": 0.2762, + "step": 42991 + }, + { + "epoch": 2.983276663659704, + "grad_norm": 1.3626505136489868, + "learning_rate": 3.2015094351529896e-05, + "loss": 0.631, + "step": 42992 + }, + { + "epoch": 2.983346055096801, + "grad_norm": 1.011939287185669, + "learning_rate": 3.201097403555025e-05, + "loss": 0.3337, + "step": 42993 + }, + { + "epoch": 2.983415446533898, + "grad_norm": 1.413417935371399, + "learning_rate": 3.200685393420406e-05, + "loss": 0.5385, + "step": 42994 + }, + { + "epoch": 2.9834848379709946, + "grad_norm": 2.186911106109619, + "learning_rate": 3.200273404750434e-05, + "loss": 0.3043, + "step": 42995 + }, + { + "epoch": 2.983554229408091, + "grad_norm": 1.0235034227371216, + "learning_rate": 3.199861437546412e-05, + "loss": 0.398, + "step": 42996 + }, + { + "epoch": 2.983623620845188, + "grad_norm": 1.2499157190322876, + "learning_rate": 3.1994494918096406e-05, + "loss": 0.4067, + "step": 42997 + }, + { + "epoch": 2.9836930122822842, + "grad_norm": 0.9764722585678101, + "learning_rate": 3.1990375675414155e-05, + "loss": 0.3255, + "step": 42998 + }, + { + "epoch": 2.983762403719381, + "grad_norm": 1.228952407836914, + "learning_rate": 3.198625664743038e-05, + "loss": 0.3802, + "step": 42999 + }, + { + "epoch": 2.983831795156478, + "grad_norm": 1.0891631841659546, + "learning_rate": 3.1982137834158156e-05, + "loss": 0.3817, + "step": 43000 + }, + { + "epoch": 2.9839011865935743, + "grad_norm": 0.9195609092712402, + "learning_rate": 3.1978019235610376e-05, + "loss": 0.3718, + "step": 43001 + }, + { + "epoch": 2.983970578030671, + "grad_norm": 1.0488547086715698, + "learning_rate": 3.197390085180012e-05, + "loss": 0.4065, + "step": 43002 + }, + { + "epoch": 2.9840399694677675, + "grad_norm": 0.9687744975090027, + "learning_rate": 3.1969782682740354e-05, + "loss": 0.351, + "step": 43003 + }, + { + "epoch": 2.9841093609048643, + "grad_norm": 0.858084499835968, + "learning_rate": 3.196566472844409e-05, + "loss": 0.3421, + "step": 43004 + }, + { + "epoch": 2.984178752341961, + "grad_norm": 1.1316461563110352, + "learning_rate": 3.196154698892434e-05, + "loss": 0.291, + "step": 43005 + }, + { + "epoch": 2.9842481437790576, + "grad_norm": 1.3088078498840332, + "learning_rate": 3.195742946419411e-05, + "loss": 0.3682, + "step": 43006 + }, + { + "epoch": 2.9843175352161544, + "grad_norm": 1.3581923246383667, + "learning_rate": 3.1953312154266354e-05, + "loss": 0.4512, + "step": 43007 + }, + { + "epoch": 2.984386926653251, + "grad_norm": 1.0309369564056396, + "learning_rate": 3.19491950591541e-05, + "loss": 0.3579, + "step": 43008 + }, + { + "epoch": 2.9844563180903476, + "grad_norm": 1.2082957029342651, + "learning_rate": 3.1945078178870346e-05, + "loss": 0.4152, + "step": 43009 + }, + { + "epoch": 2.9845257095274444, + "grad_norm": 1.4286898374557495, + "learning_rate": 3.194096151342807e-05, + "loss": 0.4273, + "step": 43010 + }, + { + "epoch": 2.984595100964541, + "grad_norm": 1.4271600246429443, + "learning_rate": 3.1936845062840296e-05, + "loss": 0.5081, + "step": 43011 + }, + { + "epoch": 2.9846644924016377, + "grad_norm": 1.0359888076782227, + "learning_rate": 3.193272882711999e-05, + "loss": 0.3684, + "step": 43012 + }, + { + "epoch": 2.984733883838734, + "grad_norm": 0.9919776916503906, + "learning_rate": 3.192861280628021e-05, + "loss": 0.5211, + "step": 43013 + }, + { + "epoch": 2.984803275275831, + "grad_norm": 1.1293699741363525, + "learning_rate": 3.192449700033386e-05, + "loss": 0.2868, + "step": 43014 + }, + { + "epoch": 2.9848726667129277, + "grad_norm": 0.9196469783782959, + "learning_rate": 3.1920381409293986e-05, + "loss": 0.302, + "step": 43015 + }, + { + "epoch": 2.9849420581500246, + "grad_norm": 1.0190023183822632, + "learning_rate": 3.1916266033173556e-05, + "loss": 0.3893, + "step": 43016 + }, + { + "epoch": 2.985011449587121, + "grad_norm": 1.3519701957702637, + "learning_rate": 3.191215087198558e-05, + "loss": 0.5142, + "step": 43017 + }, + { + "epoch": 2.9850808410242173, + "grad_norm": 1.3216158151626587, + "learning_rate": 3.190803592574305e-05, + "loss": 0.3644, + "step": 43018 + }, + { + "epoch": 2.985150232461314, + "grad_norm": 1.0675852298736572, + "learning_rate": 3.190392119445894e-05, + "loss": 0.428, + "step": 43019 + }, + { + "epoch": 2.985219623898411, + "grad_norm": 1.177841067314148, + "learning_rate": 3.189980667814629e-05, + "loss": 0.3409, + "step": 43020 + }, + { + "epoch": 2.985289015335508, + "grad_norm": 0.8718773126602173, + "learning_rate": 3.189569237681801e-05, + "loss": 0.2844, + "step": 43021 + }, + { + "epoch": 2.9853584067726042, + "grad_norm": 1.5256966352462769, + "learning_rate": 3.1891578290487135e-05, + "loss": 0.5054, + "step": 43022 + }, + { + "epoch": 2.985427798209701, + "grad_norm": 1.5887919664382935, + "learning_rate": 3.1887464419166645e-05, + "loss": 0.5464, + "step": 43023 + }, + { + "epoch": 2.9854971896467974, + "grad_norm": 1.3177591562271118, + "learning_rate": 3.188335076286952e-05, + "loss": 0.544, + "step": 43024 + }, + { + "epoch": 2.9855665810838943, + "grad_norm": 1.0269640684127808, + "learning_rate": 3.187923732160876e-05, + "loss": 0.3884, + "step": 43025 + }, + { + "epoch": 2.985635972520991, + "grad_norm": 0.7009844183921814, + "learning_rate": 3.1875124095397345e-05, + "loss": 0.2772, + "step": 43026 + }, + { + "epoch": 2.9857053639580875, + "grad_norm": 0.9128760099411011, + "learning_rate": 3.187101108424828e-05, + "loss": 0.3454, + "step": 43027 + }, + { + "epoch": 2.9857747553951843, + "grad_norm": 1.0439114570617676, + "learning_rate": 3.1866898288174506e-05, + "loss": 0.3612, + "step": 43028 + }, + { + "epoch": 2.9858441468322807, + "grad_norm": 1.7236595153808594, + "learning_rate": 3.1862785707189036e-05, + "loss": 0.5607, + "step": 43029 + }, + { + "epoch": 2.9859135382693776, + "grad_norm": 1.281989574432373, + "learning_rate": 3.185867334130483e-05, + "loss": 0.6681, + "step": 43030 + }, + { + "epoch": 2.9859829297064744, + "grad_norm": 1.3312199115753174, + "learning_rate": 3.18545611905349e-05, + "loss": 0.4823, + "step": 43031 + }, + { + "epoch": 2.9860523211435708, + "grad_norm": 1.399495244026184, + "learning_rate": 3.185044925489225e-05, + "loss": 0.5413, + "step": 43032 + }, + { + "epoch": 2.9861217125806676, + "grad_norm": 1.7767263650894165, + "learning_rate": 3.184633753438978e-05, + "loss": 0.4363, + "step": 43033 + }, + { + "epoch": 2.986191104017764, + "grad_norm": 1.3355194330215454, + "learning_rate": 3.1842226029040514e-05, + "loss": 0.5421, + "step": 43034 + }, + { + "epoch": 2.986260495454861, + "grad_norm": 1.180247187614441, + "learning_rate": 3.183811473885747e-05, + "loss": 0.3584, + "step": 43035 + }, + { + "epoch": 2.9863298868919577, + "grad_norm": 1.9011797904968262, + "learning_rate": 3.1834003663853563e-05, + "loss": 0.5245, + "step": 43036 + }, + { + "epoch": 2.986399278329054, + "grad_norm": 0.8260186314582825, + "learning_rate": 3.1829892804041786e-05, + "loss": 0.2286, + "step": 43037 + }, + { + "epoch": 2.986468669766151, + "grad_norm": 1.0755648612976074, + "learning_rate": 3.1825782159435135e-05, + "loss": 0.2931, + "step": 43038 + }, + { + "epoch": 2.9865380612032473, + "grad_norm": 1.0317835807800293, + "learning_rate": 3.182167173004658e-05, + "loss": 0.2578, + "step": 43039 + }, + { + "epoch": 2.986607452640344, + "grad_norm": 1.0201984643936157, + "learning_rate": 3.1817561515889094e-05, + "loss": 0.3209, + "step": 43040 + }, + { + "epoch": 2.986676844077441, + "grad_norm": 1.1554287672042847, + "learning_rate": 3.181345151697569e-05, + "loss": 0.2953, + "step": 43041 + }, + { + "epoch": 2.9867462355145378, + "grad_norm": 1.0786964893341064, + "learning_rate": 3.1809341733319275e-05, + "loss": 0.4148, + "step": 43042 + }, + { + "epoch": 2.986815626951634, + "grad_norm": 0.9261377453804016, + "learning_rate": 3.1805232164932866e-05, + "loss": 0.3686, + "step": 43043 + }, + { + "epoch": 2.986885018388731, + "grad_norm": 0.8093616366386414, + "learning_rate": 3.180112281182941e-05, + "loss": 0.2382, + "step": 43044 + }, + { + "epoch": 2.9869544098258274, + "grad_norm": 1.2944773435592651, + "learning_rate": 3.179701367402192e-05, + "loss": 0.4085, + "step": 43045 + }, + { + "epoch": 2.987023801262924, + "grad_norm": 1.3238179683685303, + "learning_rate": 3.1792904751523324e-05, + "loss": 0.2863, + "step": 43046 + }, + { + "epoch": 2.987093192700021, + "grad_norm": 1.456695556640625, + "learning_rate": 3.178879604434661e-05, + "loss": 0.5029, + "step": 43047 + }, + { + "epoch": 2.9871625841371174, + "grad_norm": 1.303607702255249, + "learning_rate": 3.17846875525048e-05, + "loss": 0.3484, + "step": 43048 + }, + { + "epoch": 2.9872319755742143, + "grad_norm": 1.2634950876235962, + "learning_rate": 3.178057927601078e-05, + "loss": 0.4435, + "step": 43049 + }, + { + "epoch": 2.9873013670113107, + "grad_norm": 0.9822535514831543, + "learning_rate": 3.177647121487756e-05, + "loss": 0.2708, + "step": 43050 + }, + { + "epoch": 2.9873707584484075, + "grad_norm": 1.0920014381408691, + "learning_rate": 3.1772363369118107e-05, + "loss": 0.3076, + "step": 43051 + }, + { + "epoch": 2.9874401498855043, + "grad_norm": 1.3098423480987549, + "learning_rate": 3.176825573874538e-05, + "loss": 0.4304, + "step": 43052 + }, + { + "epoch": 2.9875095413226007, + "grad_norm": 1.114179253578186, + "learning_rate": 3.1764148323772355e-05, + "loss": 0.454, + "step": 43053 + }, + { + "epoch": 2.9875789327596975, + "grad_norm": 1.300667643547058, + "learning_rate": 3.176004112421203e-05, + "loss": 0.4917, + "step": 43054 + }, + { + "epoch": 2.987648324196794, + "grad_norm": 0.8862536549568176, + "learning_rate": 3.1755934140077306e-05, + "loss": 0.3193, + "step": 43055 + }, + { + "epoch": 2.9877177156338908, + "grad_norm": 1.2716784477233887, + "learning_rate": 3.175182737138117e-05, + "loss": 0.3293, + "step": 43056 + }, + { + "epoch": 2.9877871070709876, + "grad_norm": 1.4586420059204102, + "learning_rate": 3.174772081813662e-05, + "loss": 0.2563, + "step": 43057 + }, + { + "epoch": 2.987856498508084, + "grad_norm": 1.4600720405578613, + "learning_rate": 3.1743614480356576e-05, + "loss": 0.5096, + "step": 43058 + }, + { + "epoch": 2.987925889945181, + "grad_norm": 1.3431826829910278, + "learning_rate": 3.1739508358054025e-05, + "loss": 0.3062, + "step": 43059 + }, + { + "epoch": 2.987995281382277, + "grad_norm": 1.0686843395233154, + "learning_rate": 3.1735402451241935e-05, + "loss": 0.3688, + "step": 43060 + }, + { + "epoch": 2.988064672819374, + "grad_norm": 1.0715832710266113, + "learning_rate": 3.173129675993328e-05, + "loss": 0.2791, + "step": 43061 + }, + { + "epoch": 2.988134064256471, + "grad_norm": 0.9994996190071106, + "learning_rate": 3.1727191284140975e-05, + "loss": 0.3632, + "step": 43062 + }, + { + "epoch": 2.9882034556935673, + "grad_norm": 1.2911373376846313, + "learning_rate": 3.172308602387801e-05, + "loss": 0.4436, + "step": 43063 + }, + { + "epoch": 2.988272847130664, + "grad_norm": 1.7688572406768799, + "learning_rate": 3.171898097915733e-05, + "loss": 0.338, + "step": 43064 + }, + { + "epoch": 2.9883422385677605, + "grad_norm": 1.2564653158187866, + "learning_rate": 3.1714876149991915e-05, + "loss": 0.3622, + "step": 43065 + }, + { + "epoch": 2.9884116300048573, + "grad_norm": 1.1900278329849243, + "learning_rate": 3.171077153639474e-05, + "loss": 0.6154, + "step": 43066 + }, + { + "epoch": 2.988481021441954, + "grad_norm": 1.1403521299362183, + "learning_rate": 3.170666713837871e-05, + "loss": 0.3277, + "step": 43067 + }, + { + "epoch": 2.988550412879051, + "grad_norm": 1.159382700920105, + "learning_rate": 3.17025629559568e-05, + "loss": 0.5685, + "step": 43068 + }, + { + "epoch": 2.9886198043161474, + "grad_norm": 1.040545105934143, + "learning_rate": 3.169845898914199e-05, + "loss": 0.3499, + "step": 43069 + }, + { + "epoch": 2.988689195753244, + "grad_norm": 1.2164705991744995, + "learning_rate": 3.169435523794724e-05, + "loss": 0.4451, + "step": 43070 + }, + { + "epoch": 2.9887585871903406, + "grad_norm": 1.0240209102630615, + "learning_rate": 3.169025170238545e-05, + "loss": 0.3028, + "step": 43071 + }, + { + "epoch": 2.9888279786274374, + "grad_norm": 0.9585163593292236, + "learning_rate": 3.1686148382469616e-05, + "loss": 0.2832, + "step": 43072 + }, + { + "epoch": 2.9888973700645343, + "grad_norm": 0.8422459363937378, + "learning_rate": 3.168204527821268e-05, + "loss": 0.3828, + "step": 43073 + }, + { + "epoch": 2.9889667615016307, + "grad_norm": 1.5762522220611572, + "learning_rate": 3.167794238962761e-05, + "loss": 0.415, + "step": 43074 + }, + { + "epoch": 2.9890361529387275, + "grad_norm": 1.5749984979629517, + "learning_rate": 3.1673839716727386e-05, + "loss": 0.6474, + "step": 43075 + }, + { + "epoch": 2.989105544375824, + "grad_norm": 1.196217656135559, + "learning_rate": 3.166973725952488e-05, + "loss": 0.3394, + "step": 43076 + }, + { + "epoch": 2.9891749358129207, + "grad_norm": 0.9922430515289307, + "learning_rate": 3.166563501803308e-05, + "loss": 0.4712, + "step": 43077 + }, + { + "epoch": 2.9892443272500175, + "grad_norm": 1.242690920829773, + "learning_rate": 3.1661532992264955e-05, + "loss": 0.3889, + "step": 43078 + }, + { + "epoch": 2.989313718687114, + "grad_norm": 1.2548810243606567, + "learning_rate": 3.165743118223343e-05, + "loss": 0.3996, + "step": 43079 + }, + { + "epoch": 2.9893831101242108, + "grad_norm": 1.3557699918746948, + "learning_rate": 3.165332958795148e-05, + "loss": 0.4209, + "step": 43080 + }, + { + "epoch": 2.989452501561307, + "grad_norm": 1.0416122674942017, + "learning_rate": 3.1649228209432025e-05, + "loss": 0.3163, + "step": 43081 + }, + { + "epoch": 2.989521892998404, + "grad_norm": 0.8882326483726501, + "learning_rate": 3.164512704668806e-05, + "loss": 0.2713, + "step": 43082 + }, + { + "epoch": 2.989591284435501, + "grad_norm": 1.0278493165969849, + "learning_rate": 3.164102609973247e-05, + "loss": 0.3035, + "step": 43083 + }, + { + "epoch": 2.989660675872597, + "grad_norm": 1.1316033601760864, + "learning_rate": 3.163692536857823e-05, + "loss": 0.4685, + "step": 43084 + }, + { + "epoch": 2.989730067309694, + "grad_norm": 1.0750505924224854, + "learning_rate": 3.163282485323828e-05, + "loss": 0.3661, + "step": 43085 + }, + { + "epoch": 2.9897994587467904, + "grad_norm": 0.8869460821151733, + "learning_rate": 3.162872455372557e-05, + "loss": 0.2945, + "step": 43086 + }, + { + "epoch": 2.9898688501838873, + "grad_norm": 1.2139827013015747, + "learning_rate": 3.1624624470053054e-05, + "loss": 0.2588, + "step": 43087 + }, + { + "epoch": 2.989938241620984, + "grad_norm": 1.245469331741333, + "learning_rate": 3.162052460223366e-05, + "loss": 0.5043, + "step": 43088 + }, + { + "epoch": 2.9900076330580805, + "grad_norm": 1.2926733493804932, + "learning_rate": 3.161642495028038e-05, + "loss": 0.3226, + "step": 43089 + }, + { + "epoch": 2.9900770244951773, + "grad_norm": 1.226776123046875, + "learning_rate": 3.1612325514206066e-05, + "loss": 0.4907, + "step": 43090 + }, + { + "epoch": 2.9901464159322737, + "grad_norm": 1.0836042165756226, + "learning_rate": 3.16082262940237e-05, + "loss": 0.4677, + "step": 43091 + }, + { + "epoch": 2.9902158073693705, + "grad_norm": 0.9674732685089111, + "learning_rate": 3.1604127289746246e-05, + "loss": 0.3472, + "step": 43092 + }, + { + "epoch": 2.9902851988064674, + "grad_norm": 0.9128106236457825, + "learning_rate": 3.160002850138662e-05, + "loss": 0.3659, + "step": 43093 + }, + { + "epoch": 2.990354590243564, + "grad_norm": 1.3427969217300415, + "learning_rate": 3.1595929928957757e-05, + "loss": 0.3465, + "step": 43094 + }, + { + "epoch": 2.9904239816806606, + "grad_norm": 1.1355541944503784, + "learning_rate": 3.159183157247262e-05, + "loss": 0.5281, + "step": 43095 + }, + { + "epoch": 2.9904933731177574, + "grad_norm": 1.5239558219909668, + "learning_rate": 3.158773343194415e-05, + "loss": 0.3688, + "step": 43096 + }, + { + "epoch": 2.990562764554854, + "grad_norm": 1.0907018184661865, + "learning_rate": 3.1583635507385254e-05, + "loss": 0.5086, + "step": 43097 + }, + { + "epoch": 2.9906321559919506, + "grad_norm": 1.432781457901001, + "learning_rate": 3.157953779880888e-05, + "loss": 0.4454, + "step": 43098 + }, + { + "epoch": 2.9907015474290475, + "grad_norm": 1.0590664148330688, + "learning_rate": 3.157544030622797e-05, + "loss": 0.3397, + "step": 43099 + }, + { + "epoch": 2.990770938866144, + "grad_norm": 1.5334516763687134, + "learning_rate": 3.157134302965544e-05, + "loss": 0.3085, + "step": 43100 + }, + { + "epoch": 2.9908403303032407, + "grad_norm": 0.978735089302063, + "learning_rate": 3.156724596910429e-05, + "loss": 0.3717, + "step": 43101 + }, + { + "epoch": 2.990909721740337, + "grad_norm": 1.2849388122558594, + "learning_rate": 3.156314912458736e-05, + "loss": 0.3449, + "step": 43102 + }, + { + "epoch": 2.990979113177434, + "grad_norm": 1.264113187789917, + "learning_rate": 3.1559052496117636e-05, + "loss": 0.5088, + "step": 43103 + }, + { + "epoch": 2.9910485046145308, + "grad_norm": 0.9709743857383728, + "learning_rate": 3.155495608370804e-05, + "loss": 0.2958, + "step": 43104 + }, + { + "epoch": 2.991117896051627, + "grad_norm": 1.960991621017456, + "learning_rate": 3.155085988737154e-05, + "loss": 0.4191, + "step": 43105 + }, + { + "epoch": 2.991187287488724, + "grad_norm": 1.308298110961914, + "learning_rate": 3.1546763907121003e-05, + "loss": 0.4198, + "step": 43106 + }, + { + "epoch": 2.9912566789258204, + "grad_norm": 1.3674776554107666, + "learning_rate": 3.154266814296939e-05, + "loss": 0.405, + "step": 43107 + }, + { + "epoch": 2.991326070362917, + "grad_norm": 1.2613030672073364, + "learning_rate": 3.153857259492963e-05, + "loss": 0.4095, + "step": 43108 + }, + { + "epoch": 2.991395461800014, + "grad_norm": 1.0714430809020996, + "learning_rate": 3.153447726301465e-05, + "loss": 0.3231, + "step": 43109 + }, + { + "epoch": 2.9914648532371104, + "grad_norm": 1.0859700441360474, + "learning_rate": 3.1530382147237416e-05, + "loss": 0.3735, + "step": 43110 + }, + { + "epoch": 2.9915342446742073, + "grad_norm": 1.0763461589813232, + "learning_rate": 3.15262872476108e-05, + "loss": 0.3591, + "step": 43111 + }, + { + "epoch": 2.9916036361113036, + "grad_norm": 1.0499111413955688, + "learning_rate": 3.1522192564147734e-05, + "loss": 0.3291, + "step": 43112 + }, + { + "epoch": 2.9916730275484005, + "grad_norm": 1.5809532403945923, + "learning_rate": 3.151809809686117e-05, + "loss": 0.3361, + "step": 43113 + }, + { + "epoch": 2.9917424189854973, + "grad_norm": 1.390875220298767, + "learning_rate": 3.1514003845764016e-05, + "loss": 0.4573, + "step": 43114 + }, + { + "epoch": 2.9918118104225937, + "grad_norm": 1.346702218055725, + "learning_rate": 3.150990981086922e-05, + "loss": 0.3748, + "step": 43115 + }, + { + "epoch": 2.9918812018596905, + "grad_norm": 1.0850090980529785, + "learning_rate": 3.150581599218968e-05, + "loss": 0.4673, + "step": 43116 + }, + { + "epoch": 2.991950593296787, + "grad_norm": 1.2115354537963867, + "learning_rate": 3.1501722389738376e-05, + "loss": 0.3131, + "step": 43117 + }, + { + "epoch": 2.9920199847338838, + "grad_norm": 1.060304880142212, + "learning_rate": 3.149762900352815e-05, + "loss": 0.4316, + "step": 43118 + }, + { + "epoch": 2.9920893761709806, + "grad_norm": 0.6577731966972351, + "learning_rate": 3.149353583357195e-05, + "loss": 0.2674, + "step": 43119 + }, + { + "epoch": 2.9921587676080774, + "grad_norm": 1.167441725730896, + "learning_rate": 3.148944287988273e-05, + "loss": 0.4001, + "step": 43120 + }, + { + "epoch": 2.992228159045174, + "grad_norm": 1.1592769622802734, + "learning_rate": 3.148535014247338e-05, + "loss": 0.3204, + "step": 43121 + }, + { + "epoch": 2.9922975504822706, + "grad_norm": 1.0505338907241821, + "learning_rate": 3.148125762135683e-05, + "loss": 0.3572, + "step": 43122 + }, + { + "epoch": 2.992366941919367, + "grad_norm": 1.018386960029602, + "learning_rate": 3.1477165316546e-05, + "loss": 0.5442, + "step": 43123 + }, + { + "epoch": 2.992436333356464, + "grad_norm": 1.3108059167861938, + "learning_rate": 3.147307322805384e-05, + "loss": 0.305, + "step": 43124 + }, + { + "epoch": 2.9925057247935607, + "grad_norm": 1.4157661199569702, + "learning_rate": 3.14689813558932e-05, + "loss": 0.4812, + "step": 43125 + }, + { + "epoch": 2.992575116230657, + "grad_norm": 1.832830786705017, + "learning_rate": 3.146488970007704e-05, + "loss": 0.3641, + "step": 43126 + }, + { + "epoch": 2.992644507667754, + "grad_norm": 1.0806751251220703, + "learning_rate": 3.146079826061827e-05, + "loss": 0.2881, + "step": 43127 + }, + { + "epoch": 2.9927138991048503, + "grad_norm": 1.5888400077819824, + "learning_rate": 3.145670703752982e-05, + "loss": 0.4823, + "step": 43128 + }, + { + "epoch": 2.992783290541947, + "grad_norm": 0.9287980198860168, + "learning_rate": 3.145261603082458e-05, + "loss": 0.3364, + "step": 43129 + }, + { + "epoch": 2.992852681979044, + "grad_norm": 0.9617122411727905, + "learning_rate": 3.144852524051551e-05, + "loss": 0.2685, + "step": 43130 + }, + { + "epoch": 2.9929220734161404, + "grad_norm": 1.0689283609390259, + "learning_rate": 3.144443466661547e-05, + "loss": 0.5078, + "step": 43131 + }, + { + "epoch": 2.992991464853237, + "grad_norm": 1.2080023288726807, + "learning_rate": 3.144034430913739e-05, + "loss": 0.4881, + "step": 43132 + }, + { + "epoch": 2.9930608562903336, + "grad_norm": 1.3562047481536865, + "learning_rate": 3.143625416809419e-05, + "loss": 0.2804, + "step": 43133 + }, + { + "epoch": 2.9931302477274304, + "grad_norm": 1.2950754165649414, + "learning_rate": 3.143216424349877e-05, + "loss": 0.3601, + "step": 43134 + }, + { + "epoch": 2.9931996391645272, + "grad_norm": 0.9925583600997925, + "learning_rate": 3.142807453536407e-05, + "loss": 0.3073, + "step": 43135 + }, + { + "epoch": 2.9932690306016236, + "grad_norm": 1.4272342920303345, + "learning_rate": 3.1423985043703005e-05, + "loss": 0.4015, + "step": 43136 + }, + { + "epoch": 2.9933384220387205, + "grad_norm": 1.1892021894454956, + "learning_rate": 3.141989576852843e-05, + "loss": 0.4761, + "step": 43137 + }, + { + "epoch": 2.993407813475817, + "grad_norm": 1.0529462099075317, + "learning_rate": 3.141580670985329e-05, + "loss": 0.323, + "step": 43138 + }, + { + "epoch": 2.9934772049129137, + "grad_norm": 1.6877501010894775, + "learning_rate": 3.141171786769049e-05, + "loss": 0.4678, + "step": 43139 + }, + { + "epoch": 2.9935465963500105, + "grad_norm": 1.470483660697937, + "learning_rate": 3.140762924205298e-05, + "loss": 0.2264, + "step": 43140 + }, + { + "epoch": 2.993615987787107, + "grad_norm": 1.1949416399002075, + "learning_rate": 3.140354083295358e-05, + "loss": 0.4137, + "step": 43141 + }, + { + "epoch": 2.9936853792242037, + "grad_norm": 1.1927450895309448, + "learning_rate": 3.139945264040525e-05, + "loss": 0.5118, + "step": 43142 + }, + { + "epoch": 2.9937547706613, + "grad_norm": 1.569319725036621, + "learning_rate": 3.139536466442089e-05, + "loss": 0.3734, + "step": 43143 + }, + { + "epoch": 2.993824162098397, + "grad_norm": 0.9540514349937439, + "learning_rate": 3.1391276905013437e-05, + "loss": 0.3657, + "step": 43144 + }, + { + "epoch": 2.993893553535494, + "grad_norm": 1.2078428268432617, + "learning_rate": 3.1387189362195727e-05, + "loss": 0.5065, + "step": 43145 + }, + { + "epoch": 2.9939629449725906, + "grad_norm": 1.0657848119735718, + "learning_rate": 3.138310203598069e-05, + "loss": 0.2547, + "step": 43146 + }, + { + "epoch": 2.994032336409687, + "grad_norm": 1.1512293815612793, + "learning_rate": 3.1379014926381256e-05, + "loss": 0.4061, + "step": 43147 + }, + { + "epoch": 2.994101727846784, + "grad_norm": 1.7074968814849854, + "learning_rate": 3.137492803341031e-05, + "loss": 0.493, + "step": 43148 + }, + { + "epoch": 2.9941711192838802, + "grad_norm": 0.9411042332649231, + "learning_rate": 3.137084135708075e-05, + "loss": 0.3723, + "step": 43149 + }, + { + "epoch": 2.994240510720977, + "grad_norm": 0.9838456511497498, + "learning_rate": 3.136675489740547e-05, + "loss": 0.2525, + "step": 43150 + }, + { + "epoch": 2.994309902158074, + "grad_norm": 1.160910964012146, + "learning_rate": 3.136266865439743e-05, + "loss": 0.2833, + "step": 43151 + }, + { + "epoch": 2.9943792935951703, + "grad_norm": 1.1048202514648438, + "learning_rate": 3.135858262806944e-05, + "loss": 0.2945, + "step": 43152 + }, + { + "epoch": 2.994448685032267, + "grad_norm": 1.0273162126541138, + "learning_rate": 3.135449681843444e-05, + "loss": 0.2608, + "step": 43153 + }, + { + "epoch": 2.9945180764693635, + "grad_norm": 1.4984865188598633, + "learning_rate": 3.135041122550534e-05, + "loss": 0.3827, + "step": 43154 + }, + { + "epoch": 2.9945874679064604, + "grad_norm": 0.8537891507148743, + "learning_rate": 3.134632584929501e-05, + "loss": 0.2878, + "step": 43155 + }, + { + "epoch": 2.994656859343557, + "grad_norm": 1.2218570709228516, + "learning_rate": 3.1342240689816384e-05, + "loss": 0.4521, + "step": 43156 + }, + { + "epoch": 2.9947262507806536, + "grad_norm": 1.467162013053894, + "learning_rate": 3.133815574708232e-05, + "loss": 0.4996, + "step": 43157 + }, + { + "epoch": 2.9947956422177504, + "grad_norm": 0.9974749684333801, + "learning_rate": 3.1334071021105773e-05, + "loss": 0.3313, + "step": 43158 + }, + { + "epoch": 2.994865033654847, + "grad_norm": 1.3175324201583862, + "learning_rate": 3.1329986511899566e-05, + "loss": 0.546, + "step": 43159 + }, + { + "epoch": 2.9949344250919436, + "grad_norm": 0.9548898339271545, + "learning_rate": 3.132590221947662e-05, + "loss": 0.2996, + "step": 43160 + }, + { + "epoch": 2.9950038165290405, + "grad_norm": 0.9094663858413696, + "learning_rate": 3.1321818143849835e-05, + "loss": 0.29, + "step": 43161 + }, + { + "epoch": 2.995073207966137, + "grad_norm": 1.024185061454773, + "learning_rate": 3.131773428503211e-05, + "loss": 0.2862, + "step": 43162 + }, + { + "epoch": 2.9951425994032337, + "grad_norm": 1.2265828847885132, + "learning_rate": 3.131365064303632e-05, + "loss": 0.4364, + "step": 43163 + }, + { + "epoch": 2.99521199084033, + "grad_norm": 1.7158081531524658, + "learning_rate": 3.1309567217875365e-05, + "loss": 0.4107, + "step": 43164 + }, + { + "epoch": 2.995281382277427, + "grad_norm": 1.0668193101882935, + "learning_rate": 3.130548400956217e-05, + "loss": 0.3035, + "step": 43165 + }, + { + "epoch": 2.9953507737145237, + "grad_norm": 0.9975417256355286, + "learning_rate": 3.1301401018109565e-05, + "loss": 0.4388, + "step": 43166 + }, + { + "epoch": 2.99542016515162, + "grad_norm": 1.1849887371063232, + "learning_rate": 3.129731824353046e-05, + "loss": 0.3104, + "step": 43167 + }, + { + "epoch": 2.995489556588717, + "grad_norm": 1.1467533111572266, + "learning_rate": 3.129323568583775e-05, + "loss": 0.2778, + "step": 43168 + }, + { + "epoch": 2.9955589480258134, + "grad_norm": 1.2937811613082886, + "learning_rate": 3.128915334504433e-05, + "loss": 0.4822, + "step": 43169 + }, + { + "epoch": 2.99562833946291, + "grad_norm": 0.7319396734237671, + "learning_rate": 3.128507122116308e-05, + "loss": 0.3392, + "step": 43170 + }, + { + "epoch": 2.995697730900007, + "grad_norm": 0.881148636341095, + "learning_rate": 3.128098931420692e-05, + "loss": 0.294, + "step": 43171 + }, + { + "epoch": 2.995767122337104, + "grad_norm": 1.4884133338928223, + "learning_rate": 3.1276907624188666e-05, + "loss": 0.4743, + "step": 43172 + }, + { + "epoch": 2.9958365137742002, + "grad_norm": 1.053670048713684, + "learning_rate": 3.127282615112125e-05, + "loss": 0.4808, + "step": 43173 + }, + { + "epoch": 2.995905905211297, + "grad_norm": 1.2873011827468872, + "learning_rate": 3.1268744895017534e-05, + "loss": 0.4122, + "step": 43174 + }, + { + "epoch": 2.9959752966483935, + "grad_norm": 1.6957430839538574, + "learning_rate": 3.1264663855890463e-05, + "loss": 0.3746, + "step": 43175 + }, + { + "epoch": 2.9960446880854903, + "grad_norm": 1.0404993295669556, + "learning_rate": 3.126058303375283e-05, + "loss": 0.2742, + "step": 43176 + }, + { + "epoch": 2.996114079522587, + "grad_norm": 0.7167146801948547, + "learning_rate": 3.1256502428617564e-05, + "loss": 0.1816, + "step": 43177 + }, + { + "epoch": 2.9961834709596835, + "grad_norm": 1.2205772399902344, + "learning_rate": 3.125242204049755e-05, + "loss": 0.4731, + "step": 43178 + }, + { + "epoch": 2.9962528623967803, + "grad_norm": 1.2948038578033447, + "learning_rate": 3.124834186940569e-05, + "loss": 0.3393, + "step": 43179 + }, + { + "epoch": 2.9963222538338767, + "grad_norm": 1.0780303478240967, + "learning_rate": 3.124426191535481e-05, + "loss": 0.3255, + "step": 43180 + }, + { + "epoch": 2.9963916452709736, + "grad_norm": 0.9519073367118835, + "learning_rate": 3.1240182178357814e-05, + "loss": 0.4178, + "step": 43181 + }, + { + "epoch": 2.9964610367080704, + "grad_norm": 1.0954225063323975, + "learning_rate": 3.123610265842759e-05, + "loss": 0.3184, + "step": 43182 + }, + { + "epoch": 2.996530428145167, + "grad_norm": 1.0174243450164795, + "learning_rate": 3.1232023355577e-05, + "loss": 0.3219, + "step": 43183 + }, + { + "epoch": 2.9965998195822636, + "grad_norm": 1.3256982564926147, + "learning_rate": 3.1227944269818944e-05, + "loss": 0.2537, + "step": 43184 + }, + { + "epoch": 2.99666921101936, + "grad_norm": 1.0232125520706177, + "learning_rate": 3.1223865401166285e-05, + "loss": 0.3365, + "step": 43185 + }, + { + "epoch": 2.996738602456457, + "grad_norm": 1.0865790843963623, + "learning_rate": 3.121978674963194e-05, + "loss": 0.3465, + "step": 43186 + }, + { + "epoch": 2.9968079938935537, + "grad_norm": 0.9218743443489075, + "learning_rate": 3.121570831522872e-05, + "loss": 0.2806, + "step": 43187 + }, + { + "epoch": 2.99687738533065, + "grad_norm": 1.3363004922866821, + "learning_rate": 3.1211630097969525e-05, + "loss": 0.3908, + "step": 43188 + }, + { + "epoch": 2.996946776767747, + "grad_norm": 0.7454349994659424, + "learning_rate": 3.120755209786722e-05, + "loss": 0.2653, + "step": 43189 + }, + { + "epoch": 2.9970161682048433, + "grad_norm": 0.923323392868042, + "learning_rate": 3.120347431493471e-05, + "loss": 0.3414, + "step": 43190 + }, + { + "epoch": 2.99708555964194, + "grad_norm": 1.0084748268127441, + "learning_rate": 3.119939674918484e-05, + "loss": 0.2593, + "step": 43191 + }, + { + "epoch": 2.997154951079037, + "grad_norm": 1.503231406211853, + "learning_rate": 3.11953194006305e-05, + "loss": 0.481, + "step": 43192 + }, + { + "epoch": 2.9972243425161333, + "grad_norm": 1.2119663953781128, + "learning_rate": 3.119124226928457e-05, + "loss": 0.4749, + "step": 43193 + }, + { + "epoch": 2.99729373395323, + "grad_norm": 1.2975983619689941, + "learning_rate": 3.118716535515989e-05, + "loss": 0.3933, + "step": 43194 + }, + { + "epoch": 2.9973631253903266, + "grad_norm": 1.7366502285003662, + "learning_rate": 3.1183088658269334e-05, + "loss": 0.6019, + "step": 43195 + }, + { + "epoch": 2.9974325168274234, + "grad_norm": 1.6211538314819336, + "learning_rate": 3.117901217862579e-05, + "loss": 0.509, + "step": 43196 + }, + { + "epoch": 2.9975019082645202, + "grad_norm": 1.4422624111175537, + "learning_rate": 3.1174935916242106e-05, + "loss": 0.4719, + "step": 43197 + }, + { + "epoch": 2.997571299701617, + "grad_norm": 1.3898450136184692, + "learning_rate": 3.1170859871131175e-05, + "loss": 0.3946, + "step": 43198 + }, + { + "epoch": 2.9976406911387135, + "grad_norm": 0.9251410961151123, + "learning_rate": 3.116678404330585e-05, + "loss": 0.3558, + "step": 43199 + }, + { + "epoch": 2.9977100825758103, + "grad_norm": 1.157271146774292, + "learning_rate": 3.1162708432779044e-05, + "loss": 0.4598, + "step": 43200 + }, + { + "epoch": 2.9977794740129067, + "grad_norm": 1.2043644189834595, + "learning_rate": 3.115863303956355e-05, + "loss": 0.4444, + "step": 43201 + }, + { + "epoch": 2.9978488654500035, + "grad_norm": 0.8422122597694397, + "learning_rate": 3.115455786367225e-05, + "loss": 0.2724, + "step": 43202 + }, + { + "epoch": 2.9979182568871003, + "grad_norm": 1.2543429136276245, + "learning_rate": 3.1150482905118025e-05, + "loss": 0.4448, + "step": 43203 + }, + { + "epoch": 2.9979876483241967, + "grad_norm": 1.7097314596176147, + "learning_rate": 3.114640816391374e-05, + "loss": 0.6532, + "step": 43204 + }, + { + "epoch": 2.9980570397612936, + "grad_norm": 1.277869701385498, + "learning_rate": 3.114233364007226e-05, + "loss": 0.385, + "step": 43205 + }, + { + "epoch": 2.99812643119839, + "grad_norm": 0.7256812453269958, + "learning_rate": 3.113825933360648e-05, + "loss": 0.2725, + "step": 43206 + }, + { + "epoch": 2.998195822635487, + "grad_norm": 1.0217788219451904, + "learning_rate": 3.113418524452918e-05, + "loss": 0.3666, + "step": 43207 + }, + { + "epoch": 2.9982652140725836, + "grad_norm": 1.1975812911987305, + "learning_rate": 3.113011137285328e-05, + "loss": 0.5114, + "step": 43208 + }, + { + "epoch": 2.99833460550968, + "grad_norm": 1.3772293329238892, + "learning_rate": 3.112603771859165e-05, + "loss": 0.4851, + "step": 43209 + }, + { + "epoch": 2.998403996946777, + "grad_norm": 1.708666205406189, + "learning_rate": 3.11219642817571e-05, + "loss": 0.5455, + "step": 43210 + }, + { + "epoch": 2.9984733883838732, + "grad_norm": 0.9944754242897034, + "learning_rate": 3.111789106236253e-05, + "loss": 0.2371, + "step": 43211 + }, + { + "epoch": 2.99854277982097, + "grad_norm": 0.8812329769134521, + "learning_rate": 3.111381806042076e-05, + "loss": 0.3211, + "step": 43212 + }, + { + "epoch": 2.998612171258067, + "grad_norm": 1.1272501945495605, + "learning_rate": 3.110974527594472e-05, + "loss": 0.2887, + "step": 43213 + }, + { + "epoch": 2.9986815626951633, + "grad_norm": 0.9910596013069153, + "learning_rate": 3.1105672708947196e-05, + "loss": 0.3369, + "step": 43214 + }, + { + "epoch": 2.99875095413226, + "grad_norm": 1.634814977645874, + "learning_rate": 3.110160035944106e-05, + "loss": 0.557, + "step": 43215 + }, + { + "epoch": 2.9988203455693565, + "grad_norm": 1.3409531116485596, + "learning_rate": 3.109752822743919e-05, + "loss": 0.5031, + "step": 43216 + }, + { + "epoch": 2.9988897370064533, + "grad_norm": 0.9505719542503357, + "learning_rate": 3.109345631295442e-05, + "loss": 0.2736, + "step": 43217 + }, + { + "epoch": 2.99895912844355, + "grad_norm": 1.2155908346176147, + "learning_rate": 3.1089384615999605e-05, + "loss": 0.3544, + "step": 43218 + }, + { + "epoch": 2.999028519880647, + "grad_norm": 1.243876338005066, + "learning_rate": 3.108531313658762e-05, + "loss": 0.4325, + "step": 43219 + }, + { + "epoch": 2.9990979113177434, + "grad_norm": 1.3478766679763794, + "learning_rate": 3.108124187473134e-05, + "loss": 0.3361, + "step": 43220 + }, + { + "epoch": 2.99916730275484, + "grad_norm": 1.5630639791488647, + "learning_rate": 3.107717083044355e-05, + "loss": 0.4191, + "step": 43221 + }, + { + "epoch": 2.9992366941919366, + "grad_norm": 1.2772876024246216, + "learning_rate": 3.107310000373712e-05, + "loss": 0.1734, + "step": 43222 + }, + { + "epoch": 2.9993060856290334, + "grad_norm": 1.1935298442840576, + "learning_rate": 3.1069029394624935e-05, + "loss": 0.3049, + "step": 43223 + }, + { + "epoch": 2.9993754770661303, + "grad_norm": 1.3107578754425049, + "learning_rate": 3.106495900311982e-05, + "loss": 0.2697, + "step": 43224 + }, + { + "epoch": 2.9994448685032267, + "grad_norm": 0.7963207960128784, + "learning_rate": 3.106088882923464e-05, + "loss": 0.2647, + "step": 43225 + }, + { + "epoch": 2.9995142599403235, + "grad_norm": 1.195956826210022, + "learning_rate": 3.105681887298222e-05, + "loss": 0.3305, + "step": 43226 + }, + { + "epoch": 2.99958365137742, + "grad_norm": 1.0812238454818726, + "learning_rate": 3.105274913437546e-05, + "loss": 0.3052, + "step": 43227 + }, + { + "epoch": 2.9996530428145167, + "grad_norm": 0.832921028137207, + "learning_rate": 3.1048679613427145e-05, + "loss": 0.2648, + "step": 43228 + }, + { + "epoch": 2.9997224342516136, + "grad_norm": 1.2563962936401367, + "learning_rate": 3.1044610310150145e-05, + "loss": 0.4018, + "step": 43229 + }, + { + "epoch": 2.99979182568871, + "grad_norm": 1.506673812866211, + "learning_rate": 3.1040541224557305e-05, + "loss": 0.4654, + "step": 43230 + }, + { + "epoch": 2.999861217125807, + "grad_norm": 1.1192914247512817, + "learning_rate": 3.103647235666148e-05, + "loss": 0.3729, + "step": 43231 + }, + { + "epoch": 2.999930608562903, + "grad_norm": 2.168976068496704, + "learning_rate": 3.103240370647552e-05, + "loss": 0.4328, + "step": 43232 + }, + { + "epoch": 3.0, + "grad_norm": 1.1944360733032227, + "learning_rate": 3.102833527401224e-05, + "loss": 0.2733, + "step": 43233 + }, + { + "epoch": 3.000069391437097, + "grad_norm": 1.4353939294815063, + "learning_rate": 3.102426705928455e-05, + "loss": 0.287, + "step": 43234 + }, + { + "epoch": 3.000138782874193, + "grad_norm": 1.0526676177978516, + "learning_rate": 3.1020199062305225e-05, + "loss": 0.5215, + "step": 43235 + }, + { + "epoch": 3.00020817431129, + "grad_norm": 0.8532492518424988, + "learning_rate": 3.10161312830871e-05, + "loss": 0.2616, + "step": 43236 + }, + { + "epoch": 3.0002775657483864, + "grad_norm": 1.0963513851165771, + "learning_rate": 3.101206372164307e-05, + "loss": 0.2479, + "step": 43237 + }, + { + "epoch": 3.0003469571854833, + "grad_norm": 1.0077060461044312, + "learning_rate": 3.100799637798594e-05, + "loss": 0.1491, + "step": 43238 + }, + { + "epoch": 3.00041634862258, + "grad_norm": 0.4528180658817291, + "learning_rate": 3.100392925212856e-05, + "loss": 0.2117, + "step": 43239 + }, + { + "epoch": 3.0004857400596765, + "grad_norm": 0.7243607640266418, + "learning_rate": 3.0999862344083774e-05, + "loss": 0.2867, + "step": 43240 + }, + { + "epoch": 3.0005551314967733, + "grad_norm": 0.7281962633132935, + "learning_rate": 3.0995795653864446e-05, + "loss": 0.3275, + "step": 43241 + }, + { + "epoch": 3.00062452293387, + "grad_norm": 0.8579512238502502, + "learning_rate": 3.099172918148335e-05, + "loss": 0.3227, + "step": 43242 + }, + { + "epoch": 3.0006939143709666, + "grad_norm": 0.7916656732559204, + "learning_rate": 3.098766292695336e-05, + "loss": 0.2027, + "step": 43243 + }, + { + "epoch": 3.0007633058080634, + "grad_norm": 0.7462568879127502, + "learning_rate": 3.098359689028735e-05, + "loss": 0.2355, + "step": 43244 + }, + { + "epoch": 3.0008326972451598, + "grad_norm": 0.7946965098381042, + "learning_rate": 3.097953107149808e-05, + "loss": 0.2354, + "step": 43245 + }, + { + "epoch": 3.0009020886822566, + "grad_norm": 0.9606568217277527, + "learning_rate": 3.097546547059842e-05, + "loss": 0.2285, + "step": 43246 + }, + { + "epoch": 3.0009714801193534, + "grad_norm": 0.7340520024299622, + "learning_rate": 3.097140008760121e-05, + "loss": 0.2438, + "step": 43247 + }, + { + "epoch": 3.00104087155645, + "grad_norm": 0.4902731776237488, + "learning_rate": 3.096733492251932e-05, + "loss": 0.2472, + "step": 43248 + }, + { + "epoch": 3.0011102629935467, + "grad_norm": 0.7403565645217896, + "learning_rate": 3.0963269975365504e-05, + "loss": 0.2768, + "step": 43249 + }, + { + "epoch": 3.001179654430643, + "grad_norm": 0.8445318341255188, + "learning_rate": 3.095920524615263e-05, + "loss": 0.2702, + "step": 43250 + }, + { + "epoch": 3.00124904586774, + "grad_norm": 0.9155913591384888, + "learning_rate": 3.095514073489355e-05, + "loss": 0.2752, + "step": 43251 + }, + { + "epoch": 3.0013184373048367, + "grad_norm": 0.8022767305374146, + "learning_rate": 3.095107644160107e-05, + "loss": 0.2844, + "step": 43252 + }, + { + "epoch": 3.001387828741933, + "grad_norm": 0.650432825088501, + "learning_rate": 3.0947012366288034e-05, + "loss": 0.21, + "step": 43253 + }, + { + "epoch": 3.00145722017903, + "grad_norm": 0.9425303936004639, + "learning_rate": 3.094294850896726e-05, + "loss": 0.3302, + "step": 43254 + }, + { + "epoch": 3.0015266116161268, + "grad_norm": 0.7803428769111633, + "learning_rate": 3.093888486965163e-05, + "loss": 0.2633, + "step": 43255 + }, + { + "epoch": 3.001596003053223, + "grad_norm": 0.9497696161270142, + "learning_rate": 3.093482144835389e-05, + "loss": 0.3192, + "step": 43256 + }, + { + "epoch": 3.00166539449032, + "grad_norm": 0.6985189318656921, + "learning_rate": 3.0930758245086914e-05, + "loss": 0.3219, + "step": 43257 + }, + { + "epoch": 3.0017347859274164, + "grad_norm": 0.6469232439994812, + "learning_rate": 3.0926695259863504e-05, + "loss": 0.2919, + "step": 43258 + }, + { + "epoch": 3.001804177364513, + "grad_norm": 0.9318621754646301, + "learning_rate": 3.0922632492696516e-05, + "loss": 0.2582, + "step": 43259 + }, + { + "epoch": 3.00187356880161, + "grad_norm": 0.670939564704895, + "learning_rate": 3.091856994359875e-05, + "loss": 0.1918, + "step": 43260 + }, + { + "epoch": 3.0019429602387064, + "grad_norm": 0.811033308506012, + "learning_rate": 3.091450761258305e-05, + "loss": 0.2733, + "step": 43261 + }, + { + "epoch": 3.0020123516758033, + "grad_norm": 0.7438433766365051, + "learning_rate": 3.091044549966228e-05, + "loss": 0.3654, + "step": 43262 + }, + { + "epoch": 3.0020817431128997, + "grad_norm": 0.8169713616371155, + "learning_rate": 3.0906383604849165e-05, + "loss": 0.2481, + "step": 43263 + }, + { + "epoch": 3.0021511345499965, + "grad_norm": 0.8211733102798462, + "learning_rate": 3.0902321928156594e-05, + "loss": 0.2998, + "step": 43264 + }, + { + "epoch": 3.0022205259870933, + "grad_norm": 0.8240317106246948, + "learning_rate": 3.0898260469597376e-05, + "loss": 0.3228, + "step": 43265 + }, + { + "epoch": 3.0022899174241897, + "grad_norm": 0.7203791737556458, + "learning_rate": 3.089419922918432e-05, + "loss": 0.1984, + "step": 43266 + }, + { + "epoch": 3.0023593088612865, + "grad_norm": 0.8361931443214417, + "learning_rate": 3.0890138206930266e-05, + "loss": 0.2112, + "step": 43267 + }, + { + "epoch": 3.0024287002983834, + "grad_norm": 1.1013613939285278, + "learning_rate": 3.0886077402848036e-05, + "loss": 0.3029, + "step": 43268 + }, + { + "epoch": 3.0024980917354798, + "grad_norm": 1.0993376970291138, + "learning_rate": 3.088201681695047e-05, + "loss": 0.2967, + "step": 43269 + }, + { + "epoch": 3.0025674831725766, + "grad_norm": 0.8291440010070801, + "learning_rate": 3.0877956449250315e-05, + "loss": 0.3294, + "step": 43270 + }, + { + "epoch": 3.002636874609673, + "grad_norm": 0.922644853591919, + "learning_rate": 3.087389629976043e-05, + "loss": 0.2568, + "step": 43271 + }, + { + "epoch": 3.00270626604677, + "grad_norm": 1.1973451375961304, + "learning_rate": 3.0869836368493646e-05, + "loss": 0.5581, + "step": 43272 + }, + { + "epoch": 3.0027756574838667, + "grad_norm": 1.0018901824951172, + "learning_rate": 3.0865776655462764e-05, + "loss": 0.3628, + "step": 43273 + }, + { + "epoch": 3.002845048920963, + "grad_norm": 1.066150426864624, + "learning_rate": 3.08617171606806e-05, + "loss": 0.1988, + "step": 43274 + }, + { + "epoch": 3.00291444035806, + "grad_norm": 0.8285674452781677, + "learning_rate": 3.0857657884159984e-05, + "loss": 0.2941, + "step": 43275 + }, + { + "epoch": 3.0029838317951563, + "grad_norm": 0.6225870847702026, + "learning_rate": 3.0853598825913746e-05, + "loss": 0.1638, + "step": 43276 + }, + { + "epoch": 3.003053223232253, + "grad_norm": 0.7888987064361572, + "learning_rate": 3.084953998595465e-05, + "loss": 0.2429, + "step": 43277 + }, + { + "epoch": 3.00312261466935, + "grad_norm": 0.7872958779335022, + "learning_rate": 3.084548136429552e-05, + "loss": 0.3141, + "step": 43278 + }, + { + "epoch": 3.0031920061064463, + "grad_norm": 0.7406520247459412, + "learning_rate": 3.0841422960949215e-05, + "loss": 0.3372, + "step": 43279 + }, + { + "epoch": 3.003261397543543, + "grad_norm": 1.0814337730407715, + "learning_rate": 3.083736477592849e-05, + "loss": 0.242, + "step": 43280 + }, + { + "epoch": 3.00333078898064, + "grad_norm": 0.7004013657569885, + "learning_rate": 3.0833306809246175e-05, + "loss": 0.23, + "step": 43281 + }, + { + "epoch": 3.0034001804177364, + "grad_norm": 0.8582659959793091, + "learning_rate": 3.082924906091508e-05, + "loss": 0.309, + "step": 43282 + }, + { + "epoch": 3.003469571854833, + "grad_norm": 0.9737263917922974, + "learning_rate": 3.082519153094806e-05, + "loss": 0.2624, + "step": 43283 + }, + { + "epoch": 3.0035389632919296, + "grad_norm": 0.5351364016532898, + "learning_rate": 3.082113421935785e-05, + "loss": 0.1693, + "step": 43284 + }, + { + "epoch": 3.0036083547290264, + "grad_norm": 1.0152418613433838, + "learning_rate": 3.08170771261573e-05, + "loss": 0.2217, + "step": 43285 + }, + { + "epoch": 3.0036777461661233, + "grad_norm": 1.0741666555404663, + "learning_rate": 3.08130202513592e-05, + "loss": 0.2007, + "step": 43286 + }, + { + "epoch": 3.0037471376032197, + "grad_norm": 0.7676712274551392, + "learning_rate": 3.080896359497638e-05, + "loss": 0.2183, + "step": 43287 + }, + { + "epoch": 3.0038165290403165, + "grad_norm": 1.022558569908142, + "learning_rate": 3.0804907157021626e-05, + "loss": 0.3336, + "step": 43288 + }, + { + "epoch": 3.003885920477413, + "grad_norm": 0.709130585193634, + "learning_rate": 3.0800850937507784e-05, + "loss": 0.2045, + "step": 43289 + }, + { + "epoch": 3.0039553119145097, + "grad_norm": 0.773216187953949, + "learning_rate": 3.0796794936447595e-05, + "loss": 0.178, + "step": 43290 + }, + { + "epoch": 3.0040247033516065, + "grad_norm": 1.0337779521942139, + "learning_rate": 3.07927391538539e-05, + "loss": 0.1994, + "step": 43291 + }, + { + "epoch": 3.004094094788703, + "grad_norm": 0.8964592814445496, + "learning_rate": 3.0788683589739485e-05, + "loss": 0.2129, + "step": 43292 + }, + { + "epoch": 3.0041634862257998, + "grad_norm": 0.9121667742729187, + "learning_rate": 3.0784628244117186e-05, + "loss": 0.284, + "step": 43293 + }, + { + "epoch": 3.0042328776628966, + "grad_norm": 0.7266495227813721, + "learning_rate": 3.078057311699977e-05, + "loss": 0.2751, + "step": 43294 + }, + { + "epoch": 3.004302269099993, + "grad_norm": 0.7889487147331238, + "learning_rate": 3.0776518208400065e-05, + "loss": 0.218, + "step": 43295 + }, + { + "epoch": 3.00437166053709, + "grad_norm": 0.9200797080993652, + "learning_rate": 3.0772463518330885e-05, + "loss": 0.2864, + "step": 43296 + }, + { + "epoch": 3.004441051974186, + "grad_norm": 0.939743161201477, + "learning_rate": 3.076840904680497e-05, + "loss": 0.2448, + "step": 43297 + }, + { + "epoch": 3.004510443411283, + "grad_norm": 0.7487042546272278, + "learning_rate": 3.076435479383516e-05, + "loss": 0.2843, + "step": 43298 + }, + { + "epoch": 3.00457983484838, + "grad_norm": 0.8742517232894897, + "learning_rate": 3.076030075943425e-05, + "loss": 0.2383, + "step": 43299 + }, + { + "epoch": 3.0046492262854763, + "grad_norm": 0.7652464509010315, + "learning_rate": 3.075624694361504e-05, + "loss": 0.4295, + "step": 43300 + }, + { + "epoch": 3.004718617722573, + "grad_norm": 0.6626985669136047, + "learning_rate": 3.075219334639032e-05, + "loss": 0.2641, + "step": 43301 + }, + { + "epoch": 3.0047880091596695, + "grad_norm": 0.6450586915016174, + "learning_rate": 3.074813996777289e-05, + "loss": 0.1997, + "step": 43302 + }, + { + "epoch": 3.0048574005967663, + "grad_norm": 0.8717252612113953, + "learning_rate": 3.0744086807775583e-05, + "loss": 0.2552, + "step": 43303 + }, + { + "epoch": 3.004926792033863, + "grad_norm": 0.7251635193824768, + "learning_rate": 3.074003386641112e-05, + "loss": 0.173, + "step": 43304 + }, + { + "epoch": 3.0049961834709595, + "grad_norm": 0.9042537212371826, + "learning_rate": 3.073598114369235e-05, + "loss": 0.3339, + "step": 43305 + }, + { + "epoch": 3.0050655749080564, + "grad_norm": 0.8435424566268921, + "learning_rate": 3.073192863963204e-05, + "loss": 0.2553, + "step": 43306 + }, + { + "epoch": 3.005134966345153, + "grad_norm": 1.3449733257293701, + "learning_rate": 3.072787635424299e-05, + "loss": 0.3923, + "step": 43307 + }, + { + "epoch": 3.0052043577822496, + "grad_norm": 1.0843440294265747, + "learning_rate": 3.0723824287538e-05, + "loss": 0.2766, + "step": 43308 + }, + { + "epoch": 3.0052737492193464, + "grad_norm": 1.0750483274459839, + "learning_rate": 3.071977243952987e-05, + "loss": 0.2597, + "step": 43309 + }, + { + "epoch": 3.005343140656443, + "grad_norm": 1.058854341506958, + "learning_rate": 3.071572081023141e-05, + "loss": 0.4133, + "step": 43310 + }, + { + "epoch": 3.0054125320935396, + "grad_norm": 0.819951593875885, + "learning_rate": 3.071166939965534e-05, + "loss": 0.2962, + "step": 43311 + }, + { + "epoch": 3.0054819235306365, + "grad_norm": 1.01606023311615, + "learning_rate": 3.070761820781449e-05, + "loss": 0.3647, + "step": 43312 + }, + { + "epoch": 3.005551314967733, + "grad_norm": 0.7984592318534851, + "learning_rate": 3.0703567234721664e-05, + "loss": 0.2134, + "step": 43313 + }, + { + "epoch": 3.0056207064048297, + "grad_norm": 0.7556284070014954, + "learning_rate": 3.069951648038966e-05, + "loss": 0.2489, + "step": 43314 + }, + { + "epoch": 3.005690097841926, + "grad_norm": 0.7594335675239563, + "learning_rate": 3.0695465944831204e-05, + "loss": 0.2934, + "step": 43315 + }, + { + "epoch": 3.005759489279023, + "grad_norm": 1.0147639513015747, + "learning_rate": 3.0691415628059125e-05, + "loss": 0.3458, + "step": 43316 + }, + { + "epoch": 3.0058288807161198, + "grad_norm": 1.1438757181167603, + "learning_rate": 3.0687365530086206e-05, + "loss": 0.2911, + "step": 43317 + }, + { + "epoch": 3.005898272153216, + "grad_norm": 0.9733648300170898, + "learning_rate": 3.068331565092527e-05, + "loss": 0.3006, + "step": 43318 + }, + { + "epoch": 3.005967663590313, + "grad_norm": 1.6578162908554077, + "learning_rate": 3.067926599058903e-05, + "loss": 0.32, + "step": 43319 + }, + { + "epoch": 3.00603705502741, + "grad_norm": 0.7719467282295227, + "learning_rate": 3.0675216549090305e-05, + "loss": 0.3143, + "step": 43320 + }, + { + "epoch": 3.006106446464506, + "grad_norm": 0.8768706321716309, + "learning_rate": 3.0671167326441865e-05, + "loss": 0.2523, + "step": 43321 + }, + { + "epoch": 3.006175837901603, + "grad_norm": 1.1536331176757812, + "learning_rate": 3.066711832265652e-05, + "loss": 0.2792, + "step": 43322 + }, + { + "epoch": 3.0062452293386994, + "grad_norm": 0.8894128203392029, + "learning_rate": 3.0663069537747035e-05, + "loss": 0.2978, + "step": 43323 + }, + { + "epoch": 3.0063146207757963, + "grad_norm": 0.9177780747413635, + "learning_rate": 3.0659020971726226e-05, + "loss": 0.1706, + "step": 43324 + }, + { + "epoch": 3.006384012212893, + "grad_norm": 0.9263755679130554, + "learning_rate": 3.065497262460681e-05, + "loss": 0.3632, + "step": 43325 + }, + { + "epoch": 3.0064534036499895, + "grad_norm": 1.0214308500289917, + "learning_rate": 3.06509244964016e-05, + "loss": 0.2691, + "step": 43326 + }, + { + "epoch": 3.0065227950870863, + "grad_norm": 1.4666742086410522, + "learning_rate": 3.064687658712339e-05, + "loss": 0.2792, + "step": 43327 + }, + { + "epoch": 3.0065921865241827, + "grad_norm": 0.7940655946731567, + "learning_rate": 3.0642828896784926e-05, + "loss": 0.229, + "step": 43328 + }, + { + "epoch": 3.0066615779612795, + "grad_norm": 1.0276962518692017, + "learning_rate": 3.063878142539901e-05, + "loss": 0.2598, + "step": 43329 + }, + { + "epoch": 3.0067309693983764, + "grad_norm": 0.8297512531280518, + "learning_rate": 3.0634734172978405e-05, + "loss": 0.1957, + "step": 43330 + }, + { + "epoch": 3.0068003608354728, + "grad_norm": 0.8973343968391418, + "learning_rate": 3.063068713953594e-05, + "loss": 0.2535, + "step": 43331 + }, + { + "epoch": 3.0068697522725696, + "grad_norm": 1.9166975021362305, + "learning_rate": 3.062664032508431e-05, + "loss": 0.3402, + "step": 43332 + }, + { + "epoch": 3.0069391437096664, + "grad_norm": 0.7350906729698181, + "learning_rate": 3.062259372963633e-05, + "loss": 0.2257, + "step": 43333 + }, + { + "epoch": 3.007008535146763, + "grad_norm": 0.8310381770133972, + "learning_rate": 3.061854735320478e-05, + "loss": 0.213, + "step": 43334 + }, + { + "epoch": 3.0070779265838596, + "grad_norm": 1.2522813081741333, + "learning_rate": 3.061450119580241e-05, + "loss": 0.2223, + "step": 43335 + }, + { + "epoch": 3.007147318020956, + "grad_norm": 0.6377027034759521, + "learning_rate": 3.0610455257442015e-05, + "loss": 0.144, + "step": 43336 + }, + { + "epoch": 3.007216709458053, + "grad_norm": 0.8956596255302429, + "learning_rate": 3.060640953813637e-05, + "loss": 0.2784, + "step": 43337 + }, + { + "epoch": 3.0072861008951497, + "grad_norm": 1.1558154821395874, + "learning_rate": 3.060236403789826e-05, + "loss": 0.3821, + "step": 43338 + }, + { + "epoch": 3.007355492332246, + "grad_norm": 0.9126007556915283, + "learning_rate": 3.0598318756740405e-05, + "loss": 0.2657, + "step": 43339 + }, + { + "epoch": 3.007424883769343, + "grad_norm": 0.8123077154159546, + "learning_rate": 3.059427369467561e-05, + "loss": 0.225, + "step": 43340 + }, + { + "epoch": 3.0074942752064393, + "grad_norm": 0.8950236439704895, + "learning_rate": 3.059022885171664e-05, + "loss": 0.3307, + "step": 43341 + }, + { + "epoch": 3.007563666643536, + "grad_norm": 0.6867150664329529, + "learning_rate": 3.058618422787627e-05, + "loss": 0.2285, + "step": 43342 + }, + { + "epoch": 3.007633058080633, + "grad_norm": 0.7535484433174133, + "learning_rate": 3.058213982316726e-05, + "loss": 0.2504, + "step": 43343 + }, + { + "epoch": 3.0077024495177294, + "grad_norm": 0.7662637233734131, + "learning_rate": 3.0578095637602375e-05, + "loss": 0.2239, + "step": 43344 + }, + { + "epoch": 3.007771840954826, + "grad_norm": 1.217054009437561, + "learning_rate": 3.057405167119441e-05, + "loss": 0.2286, + "step": 43345 + }, + { + "epoch": 3.007841232391923, + "grad_norm": 0.9461026787757874, + "learning_rate": 3.05700079239561e-05, + "loss": 0.2521, + "step": 43346 + }, + { + "epoch": 3.0079106238290194, + "grad_norm": 1.23697829246521, + "learning_rate": 3.0565964395900204e-05, + "loss": 0.2747, + "step": 43347 + }, + { + "epoch": 3.0079800152661162, + "grad_norm": 1.0100905895233154, + "learning_rate": 3.0561921087039514e-05, + "loss": 0.3867, + "step": 43348 + }, + { + "epoch": 3.0080494067032126, + "grad_norm": 0.7740107774734497, + "learning_rate": 3.055787799738681e-05, + "loss": 0.1857, + "step": 43349 + }, + { + "epoch": 3.0081187981403095, + "grad_norm": 0.8222496509552002, + "learning_rate": 3.055383512695479e-05, + "loss": 0.2617, + "step": 43350 + }, + { + "epoch": 3.0081881895774063, + "grad_norm": 0.8845312595367432, + "learning_rate": 3.0549792475756256e-05, + "loss": 0.2265, + "step": 43351 + }, + { + "epoch": 3.0082575810145027, + "grad_norm": 0.9277716875076294, + "learning_rate": 3.054575004380401e-05, + "loss": 0.2985, + "step": 43352 + }, + { + "epoch": 3.0083269724515995, + "grad_norm": 1.0715441703796387, + "learning_rate": 3.054170783111073e-05, + "loss": 0.3892, + "step": 43353 + }, + { + "epoch": 3.0083963638886964, + "grad_norm": 0.963041365146637, + "learning_rate": 3.0537665837689234e-05, + "loss": 0.2303, + "step": 43354 + }, + { + "epoch": 3.0084657553257927, + "grad_norm": 1.0531219244003296, + "learning_rate": 3.053362406355226e-05, + "loss": 0.2726, + "step": 43355 + }, + { + "epoch": 3.0085351467628896, + "grad_norm": 0.8937464952468872, + "learning_rate": 3.0529582508712576e-05, + "loss": 0.268, + "step": 43356 + }, + { + "epoch": 3.008604538199986, + "grad_norm": 1.1873207092285156, + "learning_rate": 3.052554117318294e-05, + "loss": 0.2597, + "step": 43357 + }, + { + "epoch": 3.008673929637083, + "grad_norm": 0.5125587582588196, + "learning_rate": 3.052150005697611e-05, + "loss": 0.1346, + "step": 43358 + }, + { + "epoch": 3.0087433210741796, + "grad_norm": 0.7979469299316406, + "learning_rate": 3.0517459160104865e-05, + "loss": 0.1852, + "step": 43359 + }, + { + "epoch": 3.008812712511276, + "grad_norm": 1.002828598022461, + "learning_rate": 3.0513418482581925e-05, + "loss": 0.3197, + "step": 43360 + }, + { + "epoch": 3.008882103948373, + "grad_norm": 0.8257035613059998, + "learning_rate": 3.0509378024420044e-05, + "loss": 0.252, + "step": 43361 + }, + { + "epoch": 3.0089514953854692, + "grad_norm": 1.2256723642349243, + "learning_rate": 3.0505337785632005e-05, + "loss": 0.5072, + "step": 43362 + }, + { + "epoch": 3.009020886822566, + "grad_norm": 2.216642379760742, + "learning_rate": 3.050129776623055e-05, + "loss": 0.2322, + "step": 43363 + }, + { + "epoch": 3.009090278259663, + "grad_norm": 0.7988092303276062, + "learning_rate": 3.0497257966228432e-05, + "loss": 0.3222, + "step": 43364 + }, + { + "epoch": 3.0091596696967593, + "grad_norm": 1.109102487564087, + "learning_rate": 3.0493218385638445e-05, + "loss": 0.2551, + "step": 43365 + }, + { + "epoch": 3.009229061133856, + "grad_norm": 0.8308202028274536, + "learning_rate": 3.0489179024473267e-05, + "loss": 0.4191, + "step": 43366 + }, + { + "epoch": 3.009298452570953, + "grad_norm": 0.9314638376235962, + "learning_rate": 3.0485139882745693e-05, + "loss": 0.1883, + "step": 43367 + }, + { + "epoch": 3.0093678440080494, + "grad_norm": 0.9411931037902832, + "learning_rate": 3.0481100960468455e-05, + "loss": 0.3749, + "step": 43368 + }, + { + "epoch": 3.009437235445146, + "grad_norm": 1.2270292043685913, + "learning_rate": 3.047706225765432e-05, + "loss": 0.2007, + "step": 43369 + }, + { + "epoch": 3.0095066268822426, + "grad_norm": 1.1730303764343262, + "learning_rate": 3.0473023774316033e-05, + "loss": 0.26, + "step": 43370 + }, + { + "epoch": 3.0095760183193394, + "grad_norm": 0.6413175463676453, + "learning_rate": 3.0468985510466352e-05, + "loss": 0.215, + "step": 43371 + }, + { + "epoch": 3.0096454097564362, + "grad_norm": 0.6397503614425659, + "learning_rate": 3.0464947466118042e-05, + "loss": 0.2143, + "step": 43372 + }, + { + "epoch": 3.0097148011935326, + "grad_norm": 0.9853636622428894, + "learning_rate": 3.0460909641283796e-05, + "loss": 0.2549, + "step": 43373 + }, + { + "epoch": 3.0097841926306295, + "grad_norm": 1.0804803371429443, + "learning_rate": 3.045687203597638e-05, + "loss": 0.2429, + "step": 43374 + }, + { + "epoch": 3.009853584067726, + "grad_norm": 0.7586706876754761, + "learning_rate": 3.0452834650208563e-05, + "loss": 0.1412, + "step": 43375 + }, + { + "epoch": 3.0099229755048227, + "grad_norm": 1.389725923538208, + "learning_rate": 3.0448797483993075e-05, + "loss": 0.3947, + "step": 43376 + }, + { + "epoch": 3.0099923669419195, + "grad_norm": 0.624494731426239, + "learning_rate": 3.0444760537342653e-05, + "loss": 0.1698, + "step": 43377 + }, + { + "epoch": 3.010061758379016, + "grad_norm": 0.8648837804794312, + "learning_rate": 3.0440723810270065e-05, + "loss": 0.3184, + "step": 43378 + }, + { + "epoch": 3.0101311498161127, + "grad_norm": 1.4145803451538086, + "learning_rate": 3.043668730278807e-05, + "loss": 0.4151, + "step": 43379 + }, + { + "epoch": 3.0102005412532096, + "grad_norm": 0.6528947353363037, + "learning_rate": 3.0432651014909352e-05, + "loss": 0.1921, + "step": 43380 + }, + { + "epoch": 3.010269932690306, + "grad_norm": 0.9652631878852844, + "learning_rate": 3.0428614946646682e-05, + "loss": 0.2628, + "step": 43381 + }, + { + "epoch": 3.010339324127403, + "grad_norm": 0.9526745676994324, + "learning_rate": 3.0424579098012794e-05, + "loss": 0.2493, + "step": 43382 + }, + { + "epoch": 3.010408715564499, + "grad_norm": 1.3882339000701904, + "learning_rate": 3.0420543469020444e-05, + "loss": 0.3992, + "step": 43383 + }, + { + "epoch": 3.010478107001596, + "grad_norm": 0.4303670823574066, + "learning_rate": 3.0416508059682403e-05, + "loss": 0.1211, + "step": 43384 + }, + { + "epoch": 3.010547498438693, + "grad_norm": 0.8501124382019043, + "learning_rate": 3.0412472870011342e-05, + "loss": 0.3712, + "step": 43385 + }, + { + "epoch": 3.0106168898757892, + "grad_norm": 0.8757393956184387, + "learning_rate": 3.0408437900020025e-05, + "loss": 0.2908, + "step": 43386 + }, + { + "epoch": 3.010686281312886, + "grad_norm": 1.1992418766021729, + "learning_rate": 3.0404403149721227e-05, + "loss": 0.2748, + "step": 43387 + }, + { + "epoch": 3.0107556727499825, + "grad_norm": 1.0945669412612915, + "learning_rate": 3.0400368619127617e-05, + "loss": 0.1999, + "step": 43388 + }, + { + "epoch": 3.0108250641870793, + "grad_norm": 1.04666268825531, + "learning_rate": 3.0396334308251985e-05, + "loss": 0.1954, + "step": 43389 + }, + { + "epoch": 3.010894455624176, + "grad_norm": 0.7858566045761108, + "learning_rate": 3.0392300217107027e-05, + "loss": 0.2267, + "step": 43390 + }, + { + "epoch": 3.0109638470612725, + "grad_norm": 1.0425487756729126, + "learning_rate": 3.038826634570552e-05, + "loss": 0.3445, + "step": 43391 + }, + { + "epoch": 3.0110332384983693, + "grad_norm": 1.5397682189941406, + "learning_rate": 3.0384232694060166e-05, + "loss": 0.2482, + "step": 43392 + }, + { + "epoch": 3.011102629935466, + "grad_norm": 1.0415706634521484, + "learning_rate": 3.0380199262183763e-05, + "loss": 0.2759, + "step": 43393 + }, + { + "epoch": 3.0111720213725626, + "grad_norm": 1.2227907180786133, + "learning_rate": 3.037616605008894e-05, + "loss": 0.2569, + "step": 43394 + }, + { + "epoch": 3.0112414128096594, + "grad_norm": 0.6279845237731934, + "learning_rate": 3.0372133057788498e-05, + "loss": 0.1437, + "step": 43395 + }, + { + "epoch": 3.011310804246756, + "grad_norm": 1.06001615524292, + "learning_rate": 3.0368100285295142e-05, + "loss": 0.236, + "step": 43396 + }, + { + "epoch": 3.0113801956838526, + "grad_norm": 0.6198078989982605, + "learning_rate": 3.0364067732621625e-05, + "loss": 0.2037, + "step": 43397 + }, + { + "epoch": 3.0114495871209495, + "grad_norm": 0.7549017667770386, + "learning_rate": 3.0360035399780662e-05, + "loss": 0.1697, + "step": 43398 + }, + { + "epoch": 3.011518978558046, + "grad_norm": 0.7973992824554443, + "learning_rate": 3.0356003286784997e-05, + "loss": 0.219, + "step": 43399 + }, + { + "epoch": 3.0115883699951427, + "grad_norm": 1.3300018310546875, + "learning_rate": 3.0351971393647362e-05, + "loss": 0.4539, + "step": 43400 + }, + { + "epoch": 3.011657761432239, + "grad_norm": 0.7372947931289673, + "learning_rate": 3.034793972038046e-05, + "loss": 0.3653, + "step": 43401 + }, + { + "epoch": 3.011727152869336, + "grad_norm": 1.072175145149231, + "learning_rate": 3.0343908266997024e-05, + "loss": 0.4537, + "step": 43402 + }, + { + "epoch": 3.0117965443064327, + "grad_norm": 0.8880119323730469, + "learning_rate": 3.033987703350978e-05, + "loss": 0.2977, + "step": 43403 + }, + { + "epoch": 3.011865935743529, + "grad_norm": 1.0322673320770264, + "learning_rate": 3.033584601993147e-05, + "loss": 0.2283, + "step": 43404 + }, + { + "epoch": 3.011935327180626, + "grad_norm": 0.7934232950210571, + "learning_rate": 3.0331815226274807e-05, + "loss": 0.3383, + "step": 43405 + }, + { + "epoch": 3.012004718617723, + "grad_norm": 1.1466128826141357, + "learning_rate": 3.0327784652552527e-05, + "loss": 0.2991, + "step": 43406 + }, + { + "epoch": 3.012074110054819, + "grad_norm": 1.4484074115753174, + "learning_rate": 3.0323754298777374e-05, + "loss": 0.3422, + "step": 43407 + }, + { + "epoch": 3.012143501491916, + "grad_norm": 1.4448401927947998, + "learning_rate": 3.0319724164962017e-05, + "loss": 0.2513, + "step": 43408 + }, + { + "epoch": 3.0122128929290124, + "grad_norm": 1.0617467164993286, + "learning_rate": 3.03156942511192e-05, + "loss": 0.2994, + "step": 43409 + }, + { + "epoch": 3.0122822843661092, + "grad_norm": 1.345445990562439, + "learning_rate": 3.0311664557261653e-05, + "loss": 0.4143, + "step": 43410 + }, + { + "epoch": 3.012351675803206, + "grad_norm": 0.861354410648346, + "learning_rate": 3.030763508340211e-05, + "loss": 0.365, + "step": 43411 + }, + { + "epoch": 3.0124210672403025, + "grad_norm": 0.690760612487793, + "learning_rate": 3.030360582955326e-05, + "loss": 0.2245, + "step": 43412 + }, + { + "epoch": 3.0124904586773993, + "grad_norm": 0.8653578758239746, + "learning_rate": 3.0299576795727848e-05, + "loss": 0.2514, + "step": 43413 + }, + { + "epoch": 3.0125598501144957, + "grad_norm": 0.7729160189628601, + "learning_rate": 3.029554798193861e-05, + "loss": 0.3653, + "step": 43414 + }, + { + "epoch": 3.0126292415515925, + "grad_norm": 0.8878541588783264, + "learning_rate": 3.029151938819822e-05, + "loss": 0.2218, + "step": 43415 + }, + { + "epoch": 3.0126986329886893, + "grad_norm": 0.6457002758979797, + "learning_rate": 3.0287491014519398e-05, + "loss": 0.1747, + "step": 43416 + }, + { + "epoch": 3.0127680244257857, + "grad_norm": 1.2381457090377808, + "learning_rate": 3.028346286091489e-05, + "loss": 0.2923, + "step": 43417 + }, + { + "epoch": 3.0128374158628826, + "grad_norm": 1.0149000883102417, + "learning_rate": 3.027943492739743e-05, + "loss": 0.391, + "step": 43418 + }, + { + "epoch": 3.0129068072999794, + "grad_norm": 1.0735517740249634, + "learning_rate": 3.027540721397968e-05, + "loss": 0.4035, + "step": 43419 + }, + { + "epoch": 3.012976198737076, + "grad_norm": 1.3438376188278198, + "learning_rate": 3.0271379720674377e-05, + "loss": 0.4228, + "step": 43420 + }, + { + "epoch": 3.0130455901741726, + "grad_norm": 0.9893856048583984, + "learning_rate": 3.0267352447494234e-05, + "loss": 0.456, + "step": 43421 + }, + { + "epoch": 3.013114981611269, + "grad_norm": 1.1308839321136475, + "learning_rate": 3.0263325394452002e-05, + "loss": 0.3442, + "step": 43422 + }, + { + "epoch": 3.013184373048366, + "grad_norm": 0.9315109848976135, + "learning_rate": 3.0259298561560335e-05, + "loss": 0.2915, + "step": 43423 + }, + { + "epoch": 3.0132537644854627, + "grad_norm": 0.7371487617492676, + "learning_rate": 3.0255271948831964e-05, + "loss": 0.2408, + "step": 43424 + }, + { + "epoch": 3.013323155922559, + "grad_norm": 0.8577212691307068, + "learning_rate": 3.0251245556279607e-05, + "loss": 0.2292, + "step": 43425 + }, + { + "epoch": 3.013392547359656, + "grad_norm": 0.8847052454948425, + "learning_rate": 3.0247219383915982e-05, + "loss": 0.2172, + "step": 43426 + }, + { + "epoch": 3.0134619387967523, + "grad_norm": 0.7818620800971985, + "learning_rate": 3.024319343175378e-05, + "loss": 0.1594, + "step": 43427 + }, + { + "epoch": 3.013531330233849, + "grad_norm": 1.2917078733444214, + "learning_rate": 3.023916769980577e-05, + "loss": 0.3034, + "step": 43428 + }, + { + "epoch": 3.013600721670946, + "grad_norm": 1.057097315788269, + "learning_rate": 3.0235142188084577e-05, + "loss": 0.3729, + "step": 43429 + }, + { + "epoch": 3.0136701131080423, + "grad_norm": 1.4108190536499023, + "learning_rate": 3.023111689660294e-05, + "loss": 0.411, + "step": 43430 + }, + { + "epoch": 3.013739504545139, + "grad_norm": 1.5229767560958862, + "learning_rate": 3.0227091825373576e-05, + "loss": 0.4079, + "step": 43431 + }, + { + "epoch": 3.013808895982236, + "grad_norm": 0.8115563988685608, + "learning_rate": 3.0223066974409185e-05, + "loss": 0.1624, + "step": 43432 + }, + { + "epoch": 3.0138782874193324, + "grad_norm": 1.2080039978027344, + "learning_rate": 3.021904234372248e-05, + "loss": 0.3314, + "step": 43433 + }, + { + "epoch": 3.0139476788564292, + "grad_norm": 1.0150936841964722, + "learning_rate": 3.021501793332616e-05, + "loss": 0.425, + "step": 43434 + }, + { + "epoch": 3.0140170702935256, + "grad_norm": 0.9267988204956055, + "learning_rate": 3.0210993743232973e-05, + "loss": 0.269, + "step": 43435 + }, + { + "epoch": 3.0140864617306224, + "grad_norm": 1.021264910697937, + "learning_rate": 3.020696977345554e-05, + "loss": 0.2652, + "step": 43436 + }, + { + "epoch": 3.0141558531677193, + "grad_norm": 1.3771218061447144, + "learning_rate": 3.0202946024006605e-05, + "loss": 0.4277, + "step": 43437 + }, + { + "epoch": 3.0142252446048157, + "grad_norm": 0.5363849997520447, + "learning_rate": 3.0198922494898874e-05, + "loss": 0.1269, + "step": 43438 + }, + { + "epoch": 3.0142946360419125, + "grad_norm": 0.6022589802742004, + "learning_rate": 3.0194899186145053e-05, + "loss": 0.1351, + "step": 43439 + }, + { + "epoch": 3.014364027479009, + "grad_norm": 1.4206570386886597, + "learning_rate": 3.0190876097757824e-05, + "loss": 0.2351, + "step": 43440 + }, + { + "epoch": 3.0144334189161057, + "grad_norm": 1.072728157043457, + "learning_rate": 3.0186853229749934e-05, + "loss": 0.3532, + "step": 43441 + }, + { + "epoch": 3.0145028103532026, + "grad_norm": 0.8174221515655518, + "learning_rate": 3.0182830582134025e-05, + "loss": 0.2303, + "step": 43442 + }, + { + "epoch": 3.014572201790299, + "grad_norm": 0.7505819201469421, + "learning_rate": 3.0178808154922812e-05, + "loss": 0.183, + "step": 43443 + }, + { + "epoch": 3.0146415932273958, + "grad_norm": 0.7915443778038025, + "learning_rate": 3.0174785948129003e-05, + "loss": 0.2295, + "step": 43444 + }, + { + "epoch": 3.0147109846644926, + "grad_norm": 0.7916473746299744, + "learning_rate": 3.0170763961765303e-05, + "loss": 0.2269, + "step": 43445 + }, + { + "epoch": 3.014780376101589, + "grad_norm": 1.0822257995605469, + "learning_rate": 3.0166742195844387e-05, + "loss": 0.3363, + "step": 43446 + }, + { + "epoch": 3.014849767538686, + "grad_norm": 1.012395977973938, + "learning_rate": 3.016272065037896e-05, + "loss": 0.3654, + "step": 43447 + }, + { + "epoch": 3.014919158975782, + "grad_norm": 1.2554303407669067, + "learning_rate": 3.0158699325381768e-05, + "loss": 0.2588, + "step": 43448 + }, + { + "epoch": 3.014988550412879, + "grad_norm": 1.5141586065292358, + "learning_rate": 3.0154678220865417e-05, + "loss": 0.3083, + "step": 43449 + }, + { + "epoch": 3.015057941849976, + "grad_norm": 0.8597115278244019, + "learning_rate": 3.0150657336842646e-05, + "loss": 0.3078, + "step": 43450 + }, + { + "epoch": 3.0151273332870723, + "grad_norm": 1.1972733736038208, + "learning_rate": 3.0146636673326147e-05, + "loss": 0.3336, + "step": 43451 + }, + { + "epoch": 3.015196724724169, + "grad_norm": 0.8598061203956604, + "learning_rate": 3.0142616230328613e-05, + "loss": 0.2695, + "step": 43452 + }, + { + "epoch": 3.0152661161612655, + "grad_norm": 0.7672167420387268, + "learning_rate": 3.013859600786276e-05, + "loss": 0.2015, + "step": 43453 + }, + { + "epoch": 3.0153355075983623, + "grad_norm": 0.8557226657867432, + "learning_rate": 3.013457600594123e-05, + "loss": 0.2229, + "step": 43454 + }, + { + "epoch": 3.015404899035459, + "grad_norm": 0.9102931618690491, + "learning_rate": 3.0130556224576734e-05, + "loss": 0.3228, + "step": 43455 + }, + { + "epoch": 3.0154742904725556, + "grad_norm": 0.7824970483779907, + "learning_rate": 3.012653666378197e-05, + "loss": 0.1952, + "step": 43456 + }, + { + "epoch": 3.0155436819096524, + "grad_norm": 1.574571967124939, + "learning_rate": 3.0122517323569643e-05, + "loss": 0.3047, + "step": 43457 + }, + { + "epoch": 3.015613073346749, + "grad_norm": 0.9003354907035828, + "learning_rate": 3.0118498203952395e-05, + "loss": 0.3063, + "step": 43458 + }, + { + "epoch": 3.0156824647838456, + "grad_norm": 0.6917663812637329, + "learning_rate": 3.0114479304942944e-05, + "loss": 0.1366, + "step": 43459 + }, + { + "epoch": 3.0157518562209424, + "grad_norm": 0.7368433475494385, + "learning_rate": 3.011046062655397e-05, + "loss": 0.3081, + "step": 43460 + }, + { + "epoch": 3.015821247658039, + "grad_norm": 0.7320535778999329, + "learning_rate": 3.0106442168798164e-05, + "loss": 0.1953, + "step": 43461 + }, + { + "epoch": 3.0158906390951357, + "grad_norm": 1.2836968898773193, + "learning_rate": 3.010242393168825e-05, + "loss": 0.3653, + "step": 43462 + }, + { + "epoch": 3.0159600305322325, + "grad_norm": 0.8640172481536865, + "learning_rate": 3.0098405915236836e-05, + "loss": 0.2728, + "step": 43463 + }, + { + "epoch": 3.016029421969329, + "grad_norm": 0.8062240481376648, + "learning_rate": 3.009438811945664e-05, + "loss": 0.1979, + "step": 43464 + }, + { + "epoch": 3.0160988134064257, + "grad_norm": 0.5580402612686157, + "learning_rate": 3.0090370544360357e-05, + "loss": 0.1714, + "step": 43465 + }, + { + "epoch": 3.016168204843522, + "grad_norm": 0.6998552083969116, + "learning_rate": 3.008635318996066e-05, + "loss": 0.2086, + "step": 43466 + }, + { + "epoch": 3.016237596280619, + "grad_norm": 2.3616011142730713, + "learning_rate": 3.0082336056270232e-05, + "loss": 0.2397, + "step": 43467 + }, + { + "epoch": 3.0163069877177158, + "grad_norm": 0.8279358148574829, + "learning_rate": 3.007831914330177e-05, + "loss": 0.3031, + "step": 43468 + }, + { + "epoch": 3.016376379154812, + "grad_norm": 1.0380305051803589, + "learning_rate": 3.0074302451067958e-05, + "loss": 0.2056, + "step": 43469 + }, + { + "epoch": 3.016445770591909, + "grad_norm": 1.0331369638442993, + "learning_rate": 3.0070285979581435e-05, + "loss": 0.3812, + "step": 43470 + }, + { + "epoch": 3.016515162029006, + "grad_norm": 0.7500349879264832, + "learning_rate": 3.0066269728854913e-05, + "loss": 0.1948, + "step": 43471 + }, + { + "epoch": 3.016584553466102, + "grad_norm": 1.2802097797393799, + "learning_rate": 3.0062253698901054e-05, + "loss": 0.169, + "step": 43472 + }, + { + "epoch": 3.016653944903199, + "grad_norm": 1.3101094961166382, + "learning_rate": 3.0058237889732543e-05, + "loss": 0.2234, + "step": 43473 + }, + { + "epoch": 3.0167233363402954, + "grad_norm": 1.0715175867080688, + "learning_rate": 3.0054222301362066e-05, + "loss": 0.2599, + "step": 43474 + }, + { + "epoch": 3.0167927277773923, + "grad_norm": 0.9985202550888062, + "learning_rate": 3.0050206933802293e-05, + "loss": 0.3257, + "step": 43475 + }, + { + "epoch": 3.016862119214489, + "grad_norm": 0.8914768695831299, + "learning_rate": 3.0046191787065934e-05, + "loss": 0.2524, + "step": 43476 + }, + { + "epoch": 3.0169315106515855, + "grad_norm": 0.813180148601532, + "learning_rate": 3.0042176861165593e-05, + "loss": 0.2697, + "step": 43477 + }, + { + "epoch": 3.0170009020886823, + "grad_norm": 1.283286690711975, + "learning_rate": 3.0038162156113992e-05, + "loss": 0.2686, + "step": 43478 + }, + { + "epoch": 3.0170702935257787, + "grad_norm": 1.096562147140503, + "learning_rate": 3.0034147671923796e-05, + "loss": 0.3906, + "step": 43479 + }, + { + "epoch": 3.0171396849628755, + "grad_norm": 2.043654203414917, + "learning_rate": 3.0030133408607673e-05, + "loss": 0.3655, + "step": 43480 + }, + { + "epoch": 3.0172090763999724, + "grad_norm": 0.8375125527381897, + "learning_rate": 3.0026119366178305e-05, + "loss": 0.2327, + "step": 43481 + }, + { + "epoch": 3.0172784678370688, + "grad_norm": 0.7133716940879822, + "learning_rate": 3.002210554464836e-05, + "loss": 0.2019, + "step": 43482 + }, + { + "epoch": 3.0173478592741656, + "grad_norm": 0.9106301069259644, + "learning_rate": 3.0018091944030534e-05, + "loss": 0.3885, + "step": 43483 + }, + { + "epoch": 3.0174172507112624, + "grad_norm": 0.7383921146392822, + "learning_rate": 3.0014078564337444e-05, + "loss": 0.1807, + "step": 43484 + }, + { + "epoch": 3.017486642148359, + "grad_norm": 1.0780845880508423, + "learning_rate": 3.0010065405581788e-05, + "loss": 0.2324, + "step": 43485 + }, + { + "epoch": 3.0175560335854557, + "grad_norm": 0.7965662479400635, + "learning_rate": 3.0006052467776235e-05, + "loss": 0.2392, + "step": 43486 + }, + { + "epoch": 3.017625425022552, + "grad_norm": 0.6308576464653015, + "learning_rate": 3.0002039750933453e-05, + "loss": 0.1422, + "step": 43487 + }, + { + "epoch": 3.017694816459649, + "grad_norm": 0.7958690524101257, + "learning_rate": 2.999802725506615e-05, + "loss": 0.2384, + "step": 43488 + }, + { + "epoch": 3.0177642078967457, + "grad_norm": 0.7134976387023926, + "learning_rate": 2.9994014980186912e-05, + "loss": 0.2115, + "step": 43489 + }, + { + "epoch": 3.017833599333842, + "grad_norm": 0.8301382064819336, + "learning_rate": 2.9990002926308446e-05, + "loss": 0.2188, + "step": 43490 + }, + { + "epoch": 3.017902990770939, + "grad_norm": 0.901807963848114, + "learning_rate": 2.998599109344342e-05, + "loss": 0.2726, + "step": 43491 + }, + { + "epoch": 3.0179723822080353, + "grad_norm": 0.8887767791748047, + "learning_rate": 2.9981979481604537e-05, + "loss": 0.1966, + "step": 43492 + }, + { + "epoch": 3.018041773645132, + "grad_norm": 1.0410327911376953, + "learning_rate": 2.9977968090804377e-05, + "loss": 0.3024, + "step": 43493 + }, + { + "epoch": 3.018111165082229, + "grad_norm": 0.9799679517745972, + "learning_rate": 2.9973956921055656e-05, + "loss": 0.3129, + "step": 43494 + }, + { + "epoch": 3.0181805565193254, + "grad_norm": 0.7398421168327332, + "learning_rate": 2.9969945972371027e-05, + "loss": 0.1938, + "step": 43495 + }, + { + "epoch": 3.018249947956422, + "grad_norm": 0.9622771143913269, + "learning_rate": 2.9965935244763156e-05, + "loss": 0.29, + "step": 43496 + }, + { + "epoch": 3.018319339393519, + "grad_norm": 0.8799148201942444, + "learning_rate": 2.996192473824473e-05, + "loss": 0.2975, + "step": 43497 + }, + { + "epoch": 3.0183887308306154, + "grad_norm": 0.49105772376060486, + "learning_rate": 2.9957914452828352e-05, + "loss": 0.0967, + "step": 43498 + }, + { + "epoch": 3.0184581222677123, + "grad_norm": 0.6976600885391235, + "learning_rate": 2.995390438852671e-05, + "loss": 0.1552, + "step": 43499 + }, + { + "epoch": 3.0185275137048087, + "grad_norm": 1.014824628829956, + "learning_rate": 2.9949894545352462e-05, + "loss": 0.306, + "step": 43500 + }, + { + "epoch": 3.0185969051419055, + "grad_norm": 0.8713850975036621, + "learning_rate": 2.994588492331827e-05, + "loss": 0.2649, + "step": 43501 + }, + { + "epoch": 3.0186662965790023, + "grad_norm": 0.8428568243980408, + "learning_rate": 2.994187552243679e-05, + "loss": 0.228, + "step": 43502 + }, + { + "epoch": 3.0187356880160987, + "grad_norm": 0.9355049729347229, + "learning_rate": 2.9937866342720688e-05, + "loss": 0.3037, + "step": 43503 + }, + { + "epoch": 3.0188050794531955, + "grad_norm": 0.7590944766998291, + "learning_rate": 2.993385738418264e-05, + "loss": 0.2164, + "step": 43504 + }, + { + "epoch": 3.018874470890292, + "grad_norm": 1.7463312149047852, + "learning_rate": 2.9929848646835247e-05, + "loss": 0.388, + "step": 43505 + }, + { + "epoch": 3.0189438623273888, + "grad_norm": 1.1867934465408325, + "learning_rate": 2.9925840130691183e-05, + "loss": 0.2954, + "step": 43506 + }, + { + "epoch": 3.0190132537644856, + "grad_norm": 1.468070387840271, + "learning_rate": 2.9921831835763127e-05, + "loss": 0.2313, + "step": 43507 + }, + { + "epoch": 3.019082645201582, + "grad_norm": 0.9857138991355896, + "learning_rate": 2.991782376206371e-05, + "loss": 0.2141, + "step": 43508 + }, + { + "epoch": 3.019152036638679, + "grad_norm": 0.5765519142150879, + "learning_rate": 2.991381590960559e-05, + "loss": 0.197, + "step": 43509 + }, + { + "epoch": 3.0192214280757756, + "grad_norm": 1.0345972776412964, + "learning_rate": 2.990980827840142e-05, + "loss": 0.3196, + "step": 43510 + }, + { + "epoch": 3.019290819512872, + "grad_norm": 0.6463043093681335, + "learning_rate": 2.9905800868463895e-05, + "loss": 0.2122, + "step": 43511 + }, + { + "epoch": 3.019360210949969, + "grad_norm": 0.9028619527816772, + "learning_rate": 2.990179367980558e-05, + "loss": 0.231, + "step": 43512 + }, + { + "epoch": 3.0194296023870653, + "grad_norm": 0.9061259627342224, + "learning_rate": 2.989778671243918e-05, + "loss": 0.2667, + "step": 43513 + }, + { + "epoch": 3.019498993824162, + "grad_norm": 0.7644907832145691, + "learning_rate": 2.9893779966377322e-05, + "loss": 0.1895, + "step": 43514 + }, + { + "epoch": 3.019568385261259, + "grad_norm": 1.2452847957611084, + "learning_rate": 2.9889773441632675e-05, + "loss": 0.1982, + "step": 43515 + }, + { + "epoch": 3.0196377766983553, + "grad_norm": 0.9104862213134766, + "learning_rate": 2.9885767138217878e-05, + "loss": 0.2336, + "step": 43516 + }, + { + "epoch": 3.019707168135452, + "grad_norm": 0.8404643535614014, + "learning_rate": 2.988176105614562e-05, + "loss": 0.243, + "step": 43517 + }, + { + "epoch": 3.0197765595725485, + "grad_norm": 1.0471047163009644, + "learning_rate": 2.987775519542846e-05, + "loss": 0.3314, + "step": 43518 + }, + { + "epoch": 3.0198459510096454, + "grad_norm": 1.0770514011383057, + "learning_rate": 2.98737495560791e-05, + "loss": 0.4098, + "step": 43519 + }, + { + "epoch": 3.019915342446742, + "grad_norm": 1.2436083555221558, + "learning_rate": 2.986974413811017e-05, + "loss": 0.2416, + "step": 43520 + }, + { + "epoch": 3.0199847338838386, + "grad_norm": 0.7534613013267517, + "learning_rate": 2.9865738941534328e-05, + "loss": 0.3491, + "step": 43521 + }, + { + "epoch": 3.0200541253209354, + "grad_norm": 0.8024226427078247, + "learning_rate": 2.9861733966364202e-05, + "loss": 0.2033, + "step": 43522 + }, + { + "epoch": 3.0201235167580323, + "grad_norm": 0.7427532076835632, + "learning_rate": 2.9857729212612483e-05, + "loss": 0.1592, + "step": 43523 + }, + { + "epoch": 3.0201929081951286, + "grad_norm": 0.7354908585548401, + "learning_rate": 2.985372468029175e-05, + "loss": 0.2092, + "step": 43524 + }, + { + "epoch": 3.0202622996322255, + "grad_norm": 1.0356636047363281, + "learning_rate": 2.9849720369414658e-05, + "loss": 0.2597, + "step": 43525 + }, + { + "epoch": 3.020331691069322, + "grad_norm": 1.4456344842910767, + "learning_rate": 2.9845716279993864e-05, + "loss": 0.2687, + "step": 43526 + }, + { + "epoch": 3.0204010825064187, + "grad_norm": 1.0124200582504272, + "learning_rate": 2.9841712412042032e-05, + "loss": 0.2203, + "step": 43527 + }, + { + "epoch": 3.0204704739435155, + "grad_norm": 0.8710760474205017, + "learning_rate": 2.9837708765571748e-05, + "loss": 0.2425, + "step": 43528 + }, + { + "epoch": 3.020539865380612, + "grad_norm": 0.9724698066711426, + "learning_rate": 2.9833705340595674e-05, + "loss": 0.2846, + "step": 43529 + }, + { + "epoch": 3.0206092568177088, + "grad_norm": 0.708138108253479, + "learning_rate": 2.9829702137126457e-05, + "loss": 0.2646, + "step": 43530 + }, + { + "epoch": 3.020678648254805, + "grad_norm": 0.777810275554657, + "learning_rate": 2.9825699155176756e-05, + "loss": 0.164, + "step": 43531 + }, + { + "epoch": 3.020748039691902, + "grad_norm": 1.1286531686782837, + "learning_rate": 2.9821696394759148e-05, + "loss": 0.3305, + "step": 43532 + }, + { + "epoch": 3.020817431128999, + "grad_norm": 0.6826087832450867, + "learning_rate": 2.9817693855886307e-05, + "loss": 0.2559, + "step": 43533 + }, + { + "epoch": 3.020886822566095, + "grad_norm": 0.7191289067268372, + "learning_rate": 2.9813691538570864e-05, + "loss": 0.2426, + "step": 43534 + }, + { + "epoch": 3.020956214003192, + "grad_norm": 1.026167392730713, + "learning_rate": 2.980968944282545e-05, + "loss": 0.2628, + "step": 43535 + }, + { + "epoch": 3.021025605440289, + "grad_norm": 0.9044832587242126, + "learning_rate": 2.98056875686627e-05, + "loss": 0.1879, + "step": 43536 + }, + { + "epoch": 3.0210949968773853, + "grad_norm": 0.9388047456741333, + "learning_rate": 2.9801685916095255e-05, + "loss": 0.2632, + "step": 43537 + }, + { + "epoch": 3.021164388314482, + "grad_norm": 0.7050222754478455, + "learning_rate": 2.979768448513577e-05, + "loss": 0.1635, + "step": 43538 + }, + { + "epoch": 3.0212337797515785, + "grad_norm": 1.0989691019058228, + "learning_rate": 2.9793683275796824e-05, + "loss": 0.3009, + "step": 43539 + }, + { + "epoch": 3.0213031711886753, + "grad_norm": 1.6225754022598267, + "learning_rate": 2.978968228809107e-05, + "loss": 0.4888, + "step": 43540 + }, + { + "epoch": 3.021372562625772, + "grad_norm": 1.1527830362319946, + "learning_rate": 2.978568152203115e-05, + "loss": 0.2289, + "step": 43541 + }, + { + "epoch": 3.0214419540628685, + "grad_norm": 0.6993261575698853, + "learning_rate": 2.9781680977629678e-05, + "loss": 0.2363, + "step": 43542 + }, + { + "epoch": 3.0215113454999654, + "grad_norm": 1.053493857383728, + "learning_rate": 2.9777680654899298e-05, + "loss": 0.2584, + "step": 43543 + }, + { + "epoch": 3.0215807369370618, + "grad_norm": 0.8754347562789917, + "learning_rate": 2.9773680553852633e-05, + "loss": 0.2084, + "step": 43544 + }, + { + "epoch": 3.0216501283741586, + "grad_norm": 0.8458728790283203, + "learning_rate": 2.9769680674502344e-05, + "loss": 0.2437, + "step": 43545 + }, + { + "epoch": 3.0217195198112554, + "grad_norm": 1.448791742324829, + "learning_rate": 2.976568101686099e-05, + "loss": 0.2723, + "step": 43546 + }, + { + "epoch": 3.021788911248352, + "grad_norm": 0.7377706170082092, + "learning_rate": 2.9761681580941238e-05, + "loss": 0.2123, + "step": 43547 + }, + { + "epoch": 3.0218583026854486, + "grad_norm": 0.930445671081543, + "learning_rate": 2.97576823667557e-05, + "loss": 0.2182, + "step": 43548 + }, + { + "epoch": 3.0219276941225455, + "grad_norm": 0.6951121687889099, + "learning_rate": 2.9753683374317008e-05, + "loss": 0.1534, + "step": 43549 + }, + { + "epoch": 3.021997085559642, + "grad_norm": 1.9289411306381226, + "learning_rate": 2.9749684603637796e-05, + "loss": 0.4485, + "step": 43550 + }, + { + "epoch": 3.0220664769967387, + "grad_norm": 1.0838221311569214, + "learning_rate": 2.9745686054730672e-05, + "loss": 0.3226, + "step": 43551 + }, + { + "epoch": 3.022135868433835, + "grad_norm": 0.6653031706809998, + "learning_rate": 2.9741687727608303e-05, + "loss": 0.2112, + "step": 43552 + }, + { + "epoch": 3.022205259870932, + "grad_norm": 0.7333603501319885, + "learning_rate": 2.973768962228324e-05, + "loss": 0.3031, + "step": 43553 + }, + { + "epoch": 3.0222746513080287, + "grad_norm": 0.9597737193107605, + "learning_rate": 2.9733691738768132e-05, + "loss": 0.3379, + "step": 43554 + }, + { + "epoch": 3.022344042745125, + "grad_norm": 0.8933359384536743, + "learning_rate": 2.9729694077075622e-05, + "loss": 0.3786, + "step": 43555 + }, + { + "epoch": 3.022413434182222, + "grad_norm": 1.1024848222732544, + "learning_rate": 2.9725696637218304e-05, + "loss": 0.238, + "step": 43556 + }, + { + "epoch": 3.0224828256193184, + "grad_norm": 0.6129850149154663, + "learning_rate": 2.9721699419208814e-05, + "loss": 0.219, + "step": 43557 + }, + { + "epoch": 3.022552217056415, + "grad_norm": 0.7541505098342896, + "learning_rate": 2.971770242305979e-05, + "loss": 0.2268, + "step": 43558 + }, + { + "epoch": 3.022621608493512, + "grad_norm": 0.903399646282196, + "learning_rate": 2.9713705648783795e-05, + "loss": 0.323, + "step": 43559 + }, + { + "epoch": 3.0226909999306084, + "grad_norm": 0.8988575339317322, + "learning_rate": 2.970970909639348e-05, + "loss": 0.2768, + "step": 43560 + }, + { + "epoch": 3.0227603913677052, + "grad_norm": 0.976336658000946, + "learning_rate": 2.9705712765901494e-05, + "loss": 0.2284, + "step": 43561 + }, + { + "epoch": 3.022829782804802, + "grad_norm": 0.9343780279159546, + "learning_rate": 2.9701716657320377e-05, + "loss": 0.317, + "step": 43562 + }, + { + "epoch": 3.0228991742418985, + "grad_norm": 0.6612064242362976, + "learning_rate": 2.9697720770662797e-05, + "loss": 0.2166, + "step": 43563 + }, + { + "epoch": 3.0229685656789953, + "grad_norm": 0.8920551538467407, + "learning_rate": 2.969372510594134e-05, + "loss": 0.2093, + "step": 43564 + }, + { + "epoch": 3.0230379571160917, + "grad_norm": 1.1597785949707031, + "learning_rate": 2.968972966316864e-05, + "loss": 0.2676, + "step": 43565 + }, + { + "epoch": 3.0231073485531885, + "grad_norm": 0.8574179410934448, + "learning_rate": 2.968573444235735e-05, + "loss": 0.2305, + "step": 43566 + }, + { + "epoch": 3.0231767399902854, + "grad_norm": 1.1003952026367188, + "learning_rate": 2.9681739443519994e-05, + "loss": 0.3085, + "step": 43567 + }, + { + "epoch": 3.0232461314273817, + "grad_norm": 1.6284376382827759, + "learning_rate": 2.9677744666669227e-05, + "loss": 0.3335, + "step": 43568 + }, + { + "epoch": 3.0233155228644786, + "grad_norm": 0.9351140260696411, + "learning_rate": 2.9673750111817667e-05, + "loss": 0.2112, + "step": 43569 + }, + { + "epoch": 3.023384914301575, + "grad_norm": 0.9178498387336731, + "learning_rate": 2.9669755778977914e-05, + "loss": 0.3204, + "step": 43570 + }, + { + "epoch": 3.023454305738672, + "grad_norm": 1.143038272857666, + "learning_rate": 2.9665761668162584e-05, + "loss": 0.3138, + "step": 43571 + }, + { + "epoch": 3.0235236971757686, + "grad_norm": 0.9144985675811768, + "learning_rate": 2.9661767779384286e-05, + "loss": 0.2495, + "step": 43572 + }, + { + "epoch": 3.023593088612865, + "grad_norm": 1.338477373123169, + "learning_rate": 2.9657774112655658e-05, + "loss": 0.3674, + "step": 43573 + }, + { + "epoch": 3.023662480049962, + "grad_norm": 0.8347742557525635, + "learning_rate": 2.965378066798925e-05, + "loss": 0.2082, + "step": 43574 + }, + { + "epoch": 3.0237318714870587, + "grad_norm": 1.0697237253189087, + "learning_rate": 2.9649787445397682e-05, + "loss": 0.2913, + "step": 43575 + }, + { + "epoch": 3.023801262924155, + "grad_norm": 0.9658710956573486, + "learning_rate": 2.9645794444893583e-05, + "loss": 0.2931, + "step": 43576 + }, + { + "epoch": 3.023870654361252, + "grad_norm": 1.024294376373291, + "learning_rate": 2.9641801666489542e-05, + "loss": 0.3434, + "step": 43577 + }, + { + "epoch": 3.0239400457983483, + "grad_norm": 0.5879400968551636, + "learning_rate": 2.9637809110198177e-05, + "loss": 0.1741, + "step": 43578 + }, + { + "epoch": 3.024009437235445, + "grad_norm": 0.9430599212646484, + "learning_rate": 2.9633816776032075e-05, + "loss": 0.2604, + "step": 43579 + }, + { + "epoch": 3.024078828672542, + "grad_norm": 1.0671350955963135, + "learning_rate": 2.962982466400389e-05, + "loss": 0.27, + "step": 43580 + }, + { + "epoch": 3.0241482201096384, + "grad_norm": 0.7066904306411743, + "learning_rate": 2.962583277412615e-05, + "loss": 0.2382, + "step": 43581 + }, + { + "epoch": 3.024217611546735, + "grad_norm": 0.7213060259819031, + "learning_rate": 2.9621841106411486e-05, + "loss": 0.2947, + "step": 43582 + }, + { + "epoch": 3.024287002983832, + "grad_norm": 0.9292177557945251, + "learning_rate": 2.961784966087251e-05, + "loss": 0.2898, + "step": 43583 + }, + { + "epoch": 3.0243563944209284, + "grad_norm": 0.7168262600898743, + "learning_rate": 2.9613858437521814e-05, + "loss": 0.1863, + "step": 43584 + }, + { + "epoch": 3.0244257858580252, + "grad_norm": 0.625335693359375, + "learning_rate": 2.9609867436372007e-05, + "loss": 0.1921, + "step": 43585 + }, + { + "epoch": 3.0244951772951216, + "grad_norm": 0.673444926738739, + "learning_rate": 2.9605876657435672e-05, + "loss": 0.2366, + "step": 43586 + }, + { + "epoch": 3.0245645687322185, + "grad_norm": 1.062463641166687, + "learning_rate": 2.9601886100725452e-05, + "loss": 0.1649, + "step": 43587 + }, + { + "epoch": 3.0246339601693153, + "grad_norm": 0.7996595501899719, + "learning_rate": 2.959789576625388e-05, + "loss": 0.1956, + "step": 43588 + }, + { + "epoch": 3.0247033516064117, + "grad_norm": 0.9143112897872925, + "learning_rate": 2.959390565403358e-05, + "loss": 0.2629, + "step": 43589 + }, + { + "epoch": 3.0247727430435085, + "grad_norm": 0.7413088083267212, + "learning_rate": 2.958991576407716e-05, + "loss": 0.2514, + "step": 43590 + }, + { + "epoch": 3.024842134480605, + "grad_norm": 1.072493314743042, + "learning_rate": 2.95859260963972e-05, + "loss": 0.317, + "step": 43591 + }, + { + "epoch": 3.0249115259177017, + "grad_norm": 0.6727190017700195, + "learning_rate": 2.9581936651006305e-05, + "loss": 0.1935, + "step": 43592 + }, + { + "epoch": 3.0249809173547986, + "grad_norm": 1.1184157133102417, + "learning_rate": 2.9577947427917097e-05, + "loss": 0.2864, + "step": 43593 + }, + { + "epoch": 3.025050308791895, + "grad_norm": 0.8775438666343689, + "learning_rate": 2.95739584271421e-05, + "loss": 0.3515, + "step": 43594 + }, + { + "epoch": 3.025119700228992, + "grad_norm": 0.7596313953399658, + "learning_rate": 2.956996964869395e-05, + "loss": 0.2347, + "step": 43595 + }, + { + "epoch": 3.025189091666088, + "grad_norm": 1.1384080648422241, + "learning_rate": 2.956598109258527e-05, + "loss": 0.3792, + "step": 43596 + }, + { + "epoch": 3.025258483103185, + "grad_norm": 1.2190347909927368, + "learning_rate": 2.956199275882858e-05, + "loss": 0.1825, + "step": 43597 + }, + { + "epoch": 3.025327874540282, + "grad_norm": 0.7370041608810425, + "learning_rate": 2.955800464743651e-05, + "loss": 0.1787, + "step": 43598 + }, + { + "epoch": 3.0253972659773782, + "grad_norm": 0.8474717736244202, + "learning_rate": 2.9554016758421645e-05, + "loss": 0.2762, + "step": 43599 + }, + { + "epoch": 3.025466657414475, + "grad_norm": 0.9169741272926331, + "learning_rate": 2.9550029091796605e-05, + "loss": 0.371, + "step": 43600 + }, + { + "epoch": 3.025536048851572, + "grad_norm": 0.9114072322845459, + "learning_rate": 2.954604164757392e-05, + "loss": 0.3659, + "step": 43601 + }, + { + "epoch": 3.0256054402886683, + "grad_norm": 1.0077403783798218, + "learning_rate": 2.9542054425766197e-05, + "loss": 0.306, + "step": 43602 + }, + { + "epoch": 3.025674831725765, + "grad_norm": 1.1216832399368286, + "learning_rate": 2.9538067426386053e-05, + "loss": 0.4547, + "step": 43603 + }, + { + "epoch": 3.0257442231628615, + "grad_norm": 0.7451304197311401, + "learning_rate": 2.953408064944604e-05, + "loss": 0.173, + "step": 43604 + }, + { + "epoch": 3.0258136145999583, + "grad_norm": 1.3346318006515503, + "learning_rate": 2.953009409495876e-05, + "loss": 0.2858, + "step": 43605 + }, + { + "epoch": 3.025883006037055, + "grad_norm": 0.9825106263160706, + "learning_rate": 2.9526107762936807e-05, + "loss": 0.2715, + "step": 43606 + }, + { + "epoch": 3.0259523974741516, + "grad_norm": 1.1163389682769775, + "learning_rate": 2.9522121653392776e-05, + "loss": 0.2417, + "step": 43607 + }, + { + "epoch": 3.0260217889112484, + "grad_norm": 1.0061744451522827, + "learning_rate": 2.9518135766339205e-05, + "loss": 0.3338, + "step": 43608 + }, + { + "epoch": 3.0260911803483452, + "grad_norm": 0.9843569397926331, + "learning_rate": 2.9514150101788697e-05, + "loss": 0.2256, + "step": 43609 + }, + { + "epoch": 3.0261605717854416, + "grad_norm": 0.8782671093940735, + "learning_rate": 2.9510164659753846e-05, + "loss": 0.2418, + "step": 43610 + }, + { + "epoch": 3.0262299632225385, + "grad_norm": 2.518200159072876, + "learning_rate": 2.9506179440247216e-05, + "loss": 0.4128, + "step": 43611 + }, + { + "epoch": 3.026299354659635, + "grad_norm": 0.634813666343689, + "learning_rate": 2.9502194443281405e-05, + "loss": 0.2209, + "step": 43612 + }, + { + "epoch": 3.0263687460967317, + "grad_norm": 0.7333199381828308, + "learning_rate": 2.9498209668868992e-05, + "loss": 0.1416, + "step": 43613 + }, + { + "epoch": 3.0264381375338285, + "grad_norm": 0.9427652955055237, + "learning_rate": 2.9494225117022577e-05, + "loss": 0.3485, + "step": 43614 + }, + { + "epoch": 3.026507528970925, + "grad_norm": 0.6550818085670471, + "learning_rate": 2.9490240787754685e-05, + "loss": 0.2302, + "step": 43615 + }, + { + "epoch": 3.0265769204080217, + "grad_norm": 0.7909093499183655, + "learning_rate": 2.9486256681077918e-05, + "loss": 0.2765, + "step": 43616 + }, + { + "epoch": 3.026646311845118, + "grad_norm": 0.5612912178039551, + "learning_rate": 2.948227279700486e-05, + "loss": 0.1479, + "step": 43617 + }, + { + "epoch": 3.026715703282215, + "grad_norm": 0.7490330338478088, + "learning_rate": 2.9478289135548086e-05, + "loss": 0.2579, + "step": 43618 + }, + { + "epoch": 3.026785094719312, + "grad_norm": 1.482056736946106, + "learning_rate": 2.947430569672017e-05, + "loss": 0.4819, + "step": 43619 + }, + { + "epoch": 3.026854486156408, + "grad_norm": 0.8320590257644653, + "learning_rate": 2.9470322480533685e-05, + "loss": 0.2605, + "step": 43620 + }, + { + "epoch": 3.026923877593505, + "grad_norm": 0.7236607074737549, + "learning_rate": 2.946633948700125e-05, + "loss": 0.2696, + "step": 43621 + }, + { + "epoch": 3.026993269030602, + "grad_norm": 0.9552645683288574, + "learning_rate": 2.9462356716135363e-05, + "loss": 0.4081, + "step": 43622 + }, + { + "epoch": 3.0270626604676982, + "grad_norm": 1.3055347204208374, + "learning_rate": 2.945837416794862e-05, + "loss": 0.3887, + "step": 43623 + }, + { + "epoch": 3.027132051904795, + "grad_norm": 1.5027660131454468, + "learning_rate": 2.9454391842453622e-05, + "loss": 0.3323, + "step": 43624 + }, + { + "epoch": 3.0272014433418915, + "grad_norm": 0.8271377086639404, + "learning_rate": 2.9450409739662908e-05, + "loss": 0.2211, + "step": 43625 + }, + { + "epoch": 3.0272708347789883, + "grad_norm": 1.0842344760894775, + "learning_rate": 2.9446427859589076e-05, + "loss": 0.3993, + "step": 43626 + }, + { + "epoch": 3.027340226216085, + "grad_norm": 1.0949519872665405, + "learning_rate": 2.944244620224469e-05, + "loss": 0.3879, + "step": 43627 + }, + { + "epoch": 3.0274096176531815, + "grad_norm": 0.8851403594017029, + "learning_rate": 2.943846476764234e-05, + "loss": 0.41, + "step": 43628 + }, + { + "epoch": 3.0274790090902783, + "grad_norm": 0.9392628073692322, + "learning_rate": 2.9434483555794534e-05, + "loss": 0.2679, + "step": 43629 + }, + { + "epoch": 3.0275484005273747, + "grad_norm": 0.7914354801177979, + "learning_rate": 2.943050256671388e-05, + "loss": 0.2096, + "step": 43630 + }, + { + "epoch": 3.0276177919644716, + "grad_norm": 1.3320351839065552, + "learning_rate": 2.9426521800412977e-05, + "loss": 0.2326, + "step": 43631 + }, + { + "epoch": 3.0276871834015684, + "grad_norm": 0.7905106544494629, + "learning_rate": 2.9422541256904324e-05, + "loss": 0.1966, + "step": 43632 + }, + { + "epoch": 3.027756574838665, + "grad_norm": 0.7277395129203796, + "learning_rate": 2.9418560936200523e-05, + "loss": 0.2331, + "step": 43633 + }, + { + "epoch": 3.0278259662757616, + "grad_norm": 0.8885990977287292, + "learning_rate": 2.9414580838314132e-05, + "loss": 0.2308, + "step": 43634 + }, + { + "epoch": 3.0278953577128584, + "grad_norm": 1.055223822593689, + "learning_rate": 2.941060096325775e-05, + "loss": 0.3222, + "step": 43635 + }, + { + "epoch": 3.027964749149955, + "grad_norm": 1.1326426267623901, + "learning_rate": 2.940662131104388e-05, + "loss": 0.1678, + "step": 43636 + }, + { + "epoch": 3.0280341405870517, + "grad_norm": 0.7551819682121277, + "learning_rate": 2.940264188168511e-05, + "loss": 0.2851, + "step": 43637 + }, + { + "epoch": 3.028103532024148, + "grad_norm": 0.5517140030860901, + "learning_rate": 2.939866267519402e-05, + "loss": 0.1888, + "step": 43638 + }, + { + "epoch": 3.028172923461245, + "grad_norm": 1.1523840427398682, + "learning_rate": 2.939468369158316e-05, + "loss": 0.2722, + "step": 43639 + }, + { + "epoch": 3.0282423148983417, + "grad_norm": 1.0988037586212158, + "learning_rate": 2.9390704930865086e-05, + "loss": 0.3693, + "step": 43640 + }, + { + "epoch": 3.028311706335438, + "grad_norm": 0.7660189270973206, + "learning_rate": 2.9386726393052367e-05, + "loss": 0.236, + "step": 43641 + }, + { + "epoch": 3.028381097772535, + "grad_norm": 1.1058814525604248, + "learning_rate": 2.9382748078157584e-05, + "loss": 0.2268, + "step": 43642 + }, + { + "epoch": 3.0284504892096313, + "grad_norm": 0.7919471263885498, + "learning_rate": 2.9378769986193243e-05, + "loss": 0.2121, + "step": 43643 + }, + { + "epoch": 3.028519880646728, + "grad_norm": 0.8162013292312622, + "learning_rate": 2.9374792117171945e-05, + "loss": 0.2231, + "step": 43644 + }, + { + "epoch": 3.028589272083825, + "grad_norm": 0.8421609997749329, + "learning_rate": 2.9370814471106222e-05, + "loss": 0.2399, + "step": 43645 + }, + { + "epoch": 3.0286586635209214, + "grad_norm": 0.7920985817909241, + "learning_rate": 2.9366837048008644e-05, + "loss": 0.4117, + "step": 43646 + }, + { + "epoch": 3.0287280549580182, + "grad_norm": 0.9284127950668335, + "learning_rate": 2.936285984789178e-05, + "loss": 0.4164, + "step": 43647 + }, + { + "epoch": 3.028797446395115, + "grad_norm": 1.0859894752502441, + "learning_rate": 2.9358882870768157e-05, + "loss": 0.359, + "step": 43648 + }, + { + "epoch": 3.0288668378322114, + "grad_norm": 0.8724030256271362, + "learning_rate": 2.935490611665037e-05, + "loss": 0.2208, + "step": 43649 + }, + { + "epoch": 3.0289362292693083, + "grad_norm": 1.0719873905181885, + "learning_rate": 2.9350929585550935e-05, + "loss": 0.2291, + "step": 43650 + }, + { + "epoch": 3.0290056207064047, + "grad_norm": 0.7714253664016724, + "learning_rate": 2.9346953277482403e-05, + "loss": 0.265, + "step": 43651 + }, + { + "epoch": 3.0290750121435015, + "grad_norm": 0.7566471099853516, + "learning_rate": 2.9342977192457355e-05, + "loss": 0.3102, + "step": 43652 + }, + { + "epoch": 3.0291444035805983, + "grad_norm": 0.9405591487884521, + "learning_rate": 2.9339001330488327e-05, + "loss": 0.2896, + "step": 43653 + }, + { + "epoch": 3.0292137950176947, + "grad_norm": 0.9985042214393616, + "learning_rate": 2.933502569158787e-05, + "loss": 0.293, + "step": 43654 + }, + { + "epoch": 3.0292831864547916, + "grad_norm": 1.1718767881393433, + "learning_rate": 2.9331050275768535e-05, + "loss": 0.279, + "step": 43655 + }, + { + "epoch": 3.029352577891888, + "grad_norm": 1.0776439905166626, + "learning_rate": 2.932707508304291e-05, + "loss": 0.3569, + "step": 43656 + }, + { + "epoch": 3.0294219693289848, + "grad_norm": 0.7575071454048157, + "learning_rate": 2.9323100113423486e-05, + "loss": 0.1755, + "step": 43657 + }, + { + "epoch": 3.0294913607660816, + "grad_norm": 0.874347984790802, + "learning_rate": 2.9319125366922828e-05, + "loss": 0.2439, + "step": 43658 + }, + { + "epoch": 3.029560752203178, + "grad_norm": 1.0300376415252686, + "learning_rate": 2.9315150843553485e-05, + "loss": 0.5236, + "step": 43659 + }, + { + "epoch": 3.029630143640275, + "grad_norm": 1.1083624362945557, + "learning_rate": 2.9311176543328023e-05, + "loss": 0.248, + "step": 43660 + }, + { + "epoch": 3.0296995350773717, + "grad_norm": 0.8813040256500244, + "learning_rate": 2.9307202466258964e-05, + "loss": 0.1675, + "step": 43661 + }, + { + "epoch": 3.029768926514468, + "grad_norm": 0.8874408602714539, + "learning_rate": 2.9303228612358902e-05, + "loss": 0.1836, + "step": 43662 + }, + { + "epoch": 3.029838317951565, + "grad_norm": 0.7616990804672241, + "learning_rate": 2.9299254981640313e-05, + "loss": 0.2331, + "step": 43663 + }, + { + "epoch": 3.0299077093886613, + "grad_norm": 0.7816190719604492, + "learning_rate": 2.9295281574115772e-05, + "loss": 0.2147, + "step": 43664 + }, + { + "epoch": 3.029977100825758, + "grad_norm": 1.0195616483688354, + "learning_rate": 2.9291308389797824e-05, + "loss": 0.3358, + "step": 43665 + }, + { + "epoch": 3.030046492262855, + "grad_norm": 0.9110891222953796, + "learning_rate": 2.9287335428699037e-05, + "loss": 0.2069, + "step": 43666 + }, + { + "epoch": 3.0301158836999513, + "grad_norm": 0.7222393155097961, + "learning_rate": 2.9283362690831907e-05, + "loss": 0.2633, + "step": 43667 + }, + { + "epoch": 3.030185275137048, + "grad_norm": 0.8641601800918579, + "learning_rate": 2.9279390176208988e-05, + "loss": 0.2945, + "step": 43668 + }, + { + "epoch": 3.0302546665741446, + "grad_norm": 1.0914443731307983, + "learning_rate": 2.927541788484284e-05, + "loss": 0.3486, + "step": 43669 + }, + { + "epoch": 3.0303240580112414, + "grad_norm": 1.0006028413772583, + "learning_rate": 2.9271445816746012e-05, + "loss": 0.1975, + "step": 43670 + }, + { + "epoch": 3.030393449448338, + "grad_norm": 0.710132360458374, + "learning_rate": 2.926747397193099e-05, + "loss": 0.3004, + "step": 43671 + }, + { + "epoch": 3.0304628408854346, + "grad_norm": 0.6140178442001343, + "learning_rate": 2.9263502350410355e-05, + "loss": 0.146, + "step": 43672 + }, + { + "epoch": 3.0305322323225314, + "grad_norm": 0.6750196218490601, + "learning_rate": 2.9259530952196633e-05, + "loss": 0.2333, + "step": 43673 + }, + { + "epoch": 3.0306016237596283, + "grad_norm": 0.8605504631996155, + "learning_rate": 2.9255559777302365e-05, + "loss": 0.2516, + "step": 43674 + }, + { + "epoch": 3.0306710151967247, + "grad_norm": 0.6716688871383667, + "learning_rate": 2.9251588825740094e-05, + "loss": 0.1698, + "step": 43675 + }, + { + "epoch": 3.0307404066338215, + "grad_norm": 0.726882815361023, + "learning_rate": 2.9247618097522366e-05, + "loss": 0.1938, + "step": 43676 + }, + { + "epoch": 3.030809798070918, + "grad_norm": 0.8454005718231201, + "learning_rate": 2.9243647592661683e-05, + "loss": 0.1535, + "step": 43677 + }, + { + "epoch": 3.0308791895080147, + "grad_norm": 0.7997973561286926, + "learning_rate": 2.923967731117059e-05, + "loss": 0.2183, + "step": 43678 + }, + { + "epoch": 3.0309485809451115, + "grad_norm": 1.6871240139007568, + "learning_rate": 2.9235707253061618e-05, + "loss": 0.2292, + "step": 43679 + }, + { + "epoch": 3.031017972382208, + "grad_norm": 0.7702519297599792, + "learning_rate": 2.9231737418347317e-05, + "loss": 0.1934, + "step": 43680 + }, + { + "epoch": 3.0310873638193048, + "grad_norm": 1.357893705368042, + "learning_rate": 2.9227767807040197e-05, + "loss": 0.2702, + "step": 43681 + }, + { + "epoch": 3.031156755256401, + "grad_norm": 1.259365200996399, + "learning_rate": 2.9223798419152815e-05, + "loss": 0.3667, + "step": 43682 + }, + { + "epoch": 3.031226146693498, + "grad_norm": 0.6702893376350403, + "learning_rate": 2.921982925469772e-05, + "loss": 0.2099, + "step": 43683 + }, + { + "epoch": 3.031295538130595, + "grad_norm": 1.003859043121338, + "learning_rate": 2.9215860313687383e-05, + "loss": 0.305, + "step": 43684 + }, + { + "epoch": 3.031364929567691, + "grad_norm": 1.0880221128463745, + "learning_rate": 2.9211891596134354e-05, + "loss": 0.3587, + "step": 43685 + }, + { + "epoch": 3.031434321004788, + "grad_norm": 0.9164647459983826, + "learning_rate": 2.920792310205117e-05, + "loss": 0.2084, + "step": 43686 + }, + { + "epoch": 3.031503712441885, + "grad_norm": 0.9192325472831726, + "learning_rate": 2.9203954831450375e-05, + "loss": 0.3198, + "step": 43687 + }, + { + "epoch": 3.0315731038789813, + "grad_norm": 0.7491157054901123, + "learning_rate": 2.919998678434447e-05, + "loss": 0.3038, + "step": 43688 + }, + { + "epoch": 3.031642495316078, + "grad_norm": 0.991387665271759, + "learning_rate": 2.9196018960745984e-05, + "loss": 0.4642, + "step": 43689 + }, + { + "epoch": 3.0317118867531745, + "grad_norm": 1.1738039255142212, + "learning_rate": 2.9192051360667493e-05, + "loss": 0.3024, + "step": 43690 + }, + { + "epoch": 3.0317812781902713, + "grad_norm": 1.1604982614517212, + "learning_rate": 2.918808398412145e-05, + "loss": 0.2335, + "step": 43691 + }, + { + "epoch": 3.031850669627368, + "grad_norm": 1.0933650732040405, + "learning_rate": 2.918411683112041e-05, + "loss": 0.3299, + "step": 43692 + }, + { + "epoch": 3.0319200610644645, + "grad_norm": 1.4179770946502686, + "learning_rate": 2.918014990167689e-05, + "loss": 0.3143, + "step": 43693 + }, + { + "epoch": 3.0319894525015614, + "grad_norm": 0.7767290472984314, + "learning_rate": 2.917618319580343e-05, + "loss": 0.2316, + "step": 43694 + }, + { + "epoch": 3.0320588439386578, + "grad_norm": 3.9424889087677, + "learning_rate": 2.9172216713512534e-05, + "loss": 0.3003, + "step": 43695 + }, + { + "epoch": 3.0321282353757546, + "grad_norm": 0.5249865055084229, + "learning_rate": 2.9168250454816736e-05, + "loss": 0.1831, + "step": 43696 + }, + { + "epoch": 3.0321976268128514, + "grad_norm": 1.032891035079956, + "learning_rate": 2.9164284419728584e-05, + "loss": 0.2046, + "step": 43697 + }, + { + "epoch": 3.032267018249948, + "grad_norm": 1.0239192247390747, + "learning_rate": 2.9160318608260538e-05, + "loss": 0.3215, + "step": 43698 + }, + { + "epoch": 3.0323364096870447, + "grad_norm": 0.7928954362869263, + "learning_rate": 2.915635302042514e-05, + "loss": 0.1839, + "step": 43699 + }, + { + "epoch": 3.0324058011241415, + "grad_norm": 0.8541154861450195, + "learning_rate": 2.9152387656234924e-05, + "loss": 0.2733, + "step": 43700 + }, + { + "epoch": 3.032475192561238, + "grad_norm": 1.2605295181274414, + "learning_rate": 2.9148422515702434e-05, + "loss": 0.4291, + "step": 43701 + }, + { + "epoch": 3.0325445839983347, + "grad_norm": 0.807489275932312, + "learning_rate": 2.9144457598840114e-05, + "loss": 0.22, + "step": 43702 + }, + { + "epoch": 3.032613975435431, + "grad_norm": 1.1449862718582153, + "learning_rate": 2.914049290566052e-05, + "loss": 0.2777, + "step": 43703 + }, + { + "epoch": 3.032683366872528, + "grad_norm": 1.384921908378601, + "learning_rate": 2.9136528436176212e-05, + "loss": 0.3353, + "step": 43704 + }, + { + "epoch": 3.0327527583096248, + "grad_norm": 1.5549613237380981, + "learning_rate": 2.9132564190399625e-05, + "loss": 0.258, + "step": 43705 + }, + { + "epoch": 3.032822149746721, + "grad_norm": 0.7652251124382019, + "learning_rate": 2.91286001683433e-05, + "loss": 0.2148, + "step": 43706 + }, + { + "epoch": 3.032891541183818, + "grad_norm": 0.8990711569786072, + "learning_rate": 2.9124636370019774e-05, + "loss": 0.3082, + "step": 43707 + }, + { + "epoch": 3.0329609326209144, + "grad_norm": 0.7655835151672363, + "learning_rate": 2.912067279544155e-05, + "loss": 0.2096, + "step": 43708 + }, + { + "epoch": 3.033030324058011, + "grad_norm": 1.2705260515213013, + "learning_rate": 2.9116709444621126e-05, + "loss": 0.3954, + "step": 43709 + }, + { + "epoch": 3.033099715495108, + "grad_norm": 0.5357019901275635, + "learning_rate": 2.911274631757104e-05, + "loss": 0.2554, + "step": 43710 + }, + { + "epoch": 3.0331691069322044, + "grad_norm": 0.7573447823524475, + "learning_rate": 2.910878341430381e-05, + "loss": 0.2017, + "step": 43711 + }, + { + "epoch": 3.0332384983693013, + "grad_norm": 0.9602231383323669, + "learning_rate": 2.9104820734831905e-05, + "loss": 0.2781, + "step": 43712 + }, + { + "epoch": 3.033307889806398, + "grad_norm": 0.9856070280075073, + "learning_rate": 2.9100858279167842e-05, + "loss": 0.2937, + "step": 43713 + }, + { + "epoch": 3.0333772812434945, + "grad_norm": 0.701262354850769, + "learning_rate": 2.9096896047324153e-05, + "loss": 0.2532, + "step": 43714 + }, + { + "epoch": 3.0334466726805913, + "grad_norm": 0.8987666368484497, + "learning_rate": 2.9092934039313337e-05, + "loss": 0.2696, + "step": 43715 + }, + { + "epoch": 3.0335160641176877, + "grad_norm": 0.8226123452186584, + "learning_rate": 2.9088972255147895e-05, + "loss": 0.1608, + "step": 43716 + }, + { + "epoch": 3.0335854555547845, + "grad_norm": 0.8798120021820068, + "learning_rate": 2.908501069484034e-05, + "loss": 0.3223, + "step": 43717 + }, + { + "epoch": 3.0336548469918814, + "grad_norm": 1.3160711526870728, + "learning_rate": 2.9081049358403224e-05, + "loss": 0.4267, + "step": 43718 + }, + { + "epoch": 3.0337242384289778, + "grad_norm": 1.2492992877960205, + "learning_rate": 2.907708824584896e-05, + "loss": 0.2492, + "step": 43719 + }, + { + "epoch": 3.0337936298660746, + "grad_norm": 0.762198805809021, + "learning_rate": 2.907312735719011e-05, + "loss": 0.2602, + "step": 43720 + }, + { + "epoch": 3.033863021303171, + "grad_norm": 0.6704473495483398, + "learning_rate": 2.9069166692439165e-05, + "loss": 0.1799, + "step": 43721 + }, + { + "epoch": 3.033932412740268, + "grad_norm": 0.8773960471153259, + "learning_rate": 2.9065206251608634e-05, + "loss": 0.2933, + "step": 43722 + }, + { + "epoch": 3.0340018041773646, + "grad_norm": 3.453033208847046, + "learning_rate": 2.906124603471101e-05, + "loss": 0.4034, + "step": 43723 + }, + { + "epoch": 3.034071195614461, + "grad_norm": 0.5884382724761963, + "learning_rate": 2.90572860417588e-05, + "loss": 0.1521, + "step": 43724 + }, + { + "epoch": 3.034140587051558, + "grad_norm": 1.3405137062072754, + "learning_rate": 2.9053326272764548e-05, + "loss": 0.1894, + "step": 43725 + }, + { + "epoch": 3.0342099784886547, + "grad_norm": 0.8312490582466125, + "learning_rate": 2.904936672774068e-05, + "loss": 0.2128, + "step": 43726 + }, + { + "epoch": 3.034279369925751, + "grad_norm": 1.090958833694458, + "learning_rate": 2.904540740669972e-05, + "loss": 0.3516, + "step": 43727 + }, + { + "epoch": 3.034348761362848, + "grad_norm": 0.7920105457305908, + "learning_rate": 2.9041448309654186e-05, + "loss": 0.2071, + "step": 43728 + }, + { + "epoch": 3.0344181527999443, + "grad_norm": 0.8743970990180969, + "learning_rate": 2.903748943661656e-05, + "loss": 0.3272, + "step": 43729 + }, + { + "epoch": 3.034487544237041, + "grad_norm": 0.677833616733551, + "learning_rate": 2.903353078759936e-05, + "loss": 0.2177, + "step": 43730 + }, + { + "epoch": 3.034556935674138, + "grad_norm": 0.8694079518318176, + "learning_rate": 2.9029572362615053e-05, + "loss": 0.2152, + "step": 43731 + }, + { + "epoch": 3.0346263271112344, + "grad_norm": 0.6977030634880066, + "learning_rate": 2.9025614161676194e-05, + "loss": 0.2118, + "step": 43732 + }, + { + "epoch": 3.034695718548331, + "grad_norm": 0.8430972099304199, + "learning_rate": 2.902165618479521e-05, + "loss": 0.225, + "step": 43733 + }, + { + "epoch": 3.0347651099854276, + "grad_norm": 0.8038887977600098, + "learning_rate": 2.9017698431984607e-05, + "loss": 0.3226, + "step": 43734 + }, + { + "epoch": 3.0348345014225244, + "grad_norm": 1.6625746488571167, + "learning_rate": 2.9013740903256903e-05, + "loss": 0.4698, + "step": 43735 + }, + { + "epoch": 3.0349038928596213, + "grad_norm": 0.7572070360183716, + "learning_rate": 2.900978359862462e-05, + "loss": 0.2195, + "step": 43736 + }, + { + "epoch": 3.0349732842967176, + "grad_norm": 1.4430335760116577, + "learning_rate": 2.9005826518100178e-05, + "loss": 0.2692, + "step": 43737 + }, + { + "epoch": 3.0350426757338145, + "grad_norm": 1.302847146987915, + "learning_rate": 2.90018696616961e-05, + "loss": 0.3078, + "step": 43738 + }, + { + "epoch": 3.0351120671709113, + "grad_norm": 0.7249595522880554, + "learning_rate": 2.8997913029424916e-05, + "loss": 0.205, + "step": 43739 + }, + { + "epoch": 3.0351814586080077, + "grad_norm": 1.3455253839492798, + "learning_rate": 2.8993956621299046e-05, + "loss": 0.2597, + "step": 43740 + }, + { + "epoch": 3.0352508500451045, + "grad_norm": 1.0504281520843506, + "learning_rate": 2.899000043733102e-05, + "loss": 0.3041, + "step": 43741 + }, + { + "epoch": 3.035320241482201, + "grad_norm": 0.8500596284866333, + "learning_rate": 2.8986044477533337e-05, + "loss": 0.2116, + "step": 43742 + }, + { + "epoch": 3.0353896329192978, + "grad_norm": 0.9442214369773865, + "learning_rate": 2.8982088741918458e-05, + "loss": 0.2524, + "step": 43743 + }, + { + "epoch": 3.0354590243563946, + "grad_norm": 0.885615885257721, + "learning_rate": 2.8978133230498895e-05, + "loss": 0.2687, + "step": 43744 + }, + { + "epoch": 3.035528415793491, + "grad_norm": 0.697527289390564, + "learning_rate": 2.897417794328715e-05, + "loss": 0.2759, + "step": 43745 + }, + { + "epoch": 3.035597807230588, + "grad_norm": 0.7541470527648926, + "learning_rate": 2.897022288029565e-05, + "loss": 0.2268, + "step": 43746 + }, + { + "epoch": 3.035667198667684, + "grad_norm": 0.8453264832496643, + "learning_rate": 2.8966268041536927e-05, + "loss": 0.2232, + "step": 43747 + }, + { + "epoch": 3.035736590104781, + "grad_norm": 0.8638284206390381, + "learning_rate": 2.8962313427023445e-05, + "loss": 0.2288, + "step": 43748 + }, + { + "epoch": 3.035805981541878, + "grad_norm": 1.1050320863723755, + "learning_rate": 2.895835903676769e-05, + "loss": 0.3662, + "step": 43749 + }, + { + "epoch": 3.0358753729789743, + "grad_norm": 1.0600972175598145, + "learning_rate": 2.895440487078217e-05, + "loss": 0.2984, + "step": 43750 + }, + { + "epoch": 3.035944764416071, + "grad_norm": 0.9726309180259705, + "learning_rate": 2.8950450929079344e-05, + "loss": 0.2976, + "step": 43751 + }, + { + "epoch": 3.036014155853168, + "grad_norm": 0.7165212035179138, + "learning_rate": 2.894649721167173e-05, + "loss": 0.2464, + "step": 43752 + }, + { + "epoch": 3.0360835472902643, + "grad_norm": 1.0395108461380005, + "learning_rate": 2.8942543718571758e-05, + "loss": 0.436, + "step": 43753 + }, + { + "epoch": 3.036152938727361, + "grad_norm": 1.302416205406189, + "learning_rate": 2.8938590449791924e-05, + "loss": 0.4287, + "step": 43754 + }, + { + "epoch": 3.0362223301644575, + "grad_norm": 0.579114556312561, + "learning_rate": 2.893463740534471e-05, + "loss": 0.1987, + "step": 43755 + }, + { + "epoch": 3.0362917216015544, + "grad_norm": 1.0173510313034058, + "learning_rate": 2.8930684585242606e-05, + "loss": 0.1814, + "step": 43756 + }, + { + "epoch": 3.036361113038651, + "grad_norm": 0.8962090015411377, + "learning_rate": 2.8926731989498092e-05, + "loss": 0.1923, + "step": 43757 + }, + { + "epoch": 3.0364305044757476, + "grad_norm": 1.2095352411270142, + "learning_rate": 2.892277961812363e-05, + "loss": 0.2868, + "step": 43758 + }, + { + "epoch": 3.0364998959128444, + "grad_norm": 1.2929933071136475, + "learning_rate": 2.891882747113174e-05, + "loss": 0.3084, + "step": 43759 + }, + { + "epoch": 3.036569287349941, + "grad_norm": 1.038803219795227, + "learning_rate": 2.8914875548534835e-05, + "loss": 0.2884, + "step": 43760 + }, + { + "epoch": 3.0366386787870376, + "grad_norm": 0.7989490032196045, + "learning_rate": 2.8910923850345416e-05, + "loss": 0.2469, + "step": 43761 + }, + { + "epoch": 3.0367080702241345, + "grad_norm": 0.7808680534362793, + "learning_rate": 2.890697237657597e-05, + "loss": 0.2113, + "step": 43762 + }, + { + "epoch": 3.036777461661231, + "grad_norm": 0.9046652913093567, + "learning_rate": 2.890302112723896e-05, + "loss": 0.3441, + "step": 43763 + }, + { + "epoch": 3.0368468530983277, + "grad_norm": 0.6685486435890198, + "learning_rate": 2.889907010234686e-05, + "loss": 0.2681, + "step": 43764 + }, + { + "epoch": 3.0369162445354245, + "grad_norm": 1.1225658655166626, + "learning_rate": 2.8895119301912154e-05, + "loss": 0.2114, + "step": 43765 + }, + { + "epoch": 3.036985635972521, + "grad_norm": 1.156410813331604, + "learning_rate": 2.889116872594734e-05, + "loss": 0.51, + "step": 43766 + }, + { + "epoch": 3.0370550274096177, + "grad_norm": 1.1856462955474854, + "learning_rate": 2.888721837446481e-05, + "loss": 0.4213, + "step": 43767 + }, + { + "epoch": 3.037124418846714, + "grad_norm": 0.9973747134208679, + "learning_rate": 2.8883268247477092e-05, + "loss": 0.2286, + "step": 43768 + }, + { + "epoch": 3.037193810283811, + "grad_norm": 1.108824610710144, + "learning_rate": 2.887931834499664e-05, + "loss": 0.339, + "step": 43769 + }, + { + "epoch": 3.037263201720908, + "grad_norm": 0.8801562190055847, + "learning_rate": 2.887536866703594e-05, + "loss": 0.2485, + "step": 43770 + }, + { + "epoch": 3.037332593158004, + "grad_norm": 0.9382248520851135, + "learning_rate": 2.8871419213607464e-05, + "loss": 0.3607, + "step": 43771 + }, + { + "epoch": 3.037401984595101, + "grad_norm": 0.8650305271148682, + "learning_rate": 2.886746998472365e-05, + "loss": 0.2287, + "step": 43772 + }, + { + "epoch": 3.0374713760321974, + "grad_norm": 0.9759156703948975, + "learning_rate": 2.8863520980396964e-05, + "loss": 0.222, + "step": 43773 + }, + { + "epoch": 3.0375407674692942, + "grad_norm": 1.6722140312194824, + "learning_rate": 2.8859572200639927e-05, + "loss": 0.2893, + "step": 43774 + }, + { + "epoch": 3.037610158906391, + "grad_norm": 1.2130967378616333, + "learning_rate": 2.8855623645464945e-05, + "loss": 0.3905, + "step": 43775 + }, + { + "epoch": 3.0376795503434875, + "grad_norm": 0.9550670981407166, + "learning_rate": 2.8851675314884497e-05, + "loss": 0.3007, + "step": 43776 + }, + { + "epoch": 3.0377489417805843, + "grad_norm": 1.0223630666732788, + "learning_rate": 2.8847727208911046e-05, + "loss": 0.1945, + "step": 43777 + }, + { + "epoch": 3.037818333217681, + "grad_norm": 1.2008781433105469, + "learning_rate": 2.8843779327557076e-05, + "loss": 0.2581, + "step": 43778 + }, + { + "epoch": 3.0378877246547775, + "grad_norm": 1.0045112371444702, + "learning_rate": 2.8839831670835037e-05, + "loss": 0.248, + "step": 43779 + }, + { + "epoch": 3.0379571160918744, + "grad_norm": 1.0993818044662476, + "learning_rate": 2.8835884238757428e-05, + "loss": 0.2361, + "step": 43780 + }, + { + "epoch": 3.0380265075289707, + "grad_norm": 0.9503302574157715, + "learning_rate": 2.8831937031336643e-05, + "loss": 0.2289, + "step": 43781 + }, + { + "epoch": 3.0380958989660676, + "grad_norm": 0.8858156800270081, + "learning_rate": 2.8827990048585175e-05, + "loss": 0.2309, + "step": 43782 + }, + { + "epoch": 3.0381652904031644, + "grad_norm": 0.9058998227119446, + "learning_rate": 2.882404329051548e-05, + "loss": 0.2869, + "step": 43783 + }, + { + "epoch": 3.038234681840261, + "grad_norm": 2.0474789142608643, + "learning_rate": 2.8820096757140025e-05, + "loss": 0.3509, + "step": 43784 + }, + { + "epoch": 3.0383040732773576, + "grad_norm": 0.6897709965705872, + "learning_rate": 2.8816150448471267e-05, + "loss": 0.2071, + "step": 43785 + }, + { + "epoch": 3.038373464714454, + "grad_norm": 1.3774477243423462, + "learning_rate": 2.8812204364521665e-05, + "loss": 0.2956, + "step": 43786 + }, + { + "epoch": 3.038442856151551, + "grad_norm": 0.8537589311599731, + "learning_rate": 2.880825850530371e-05, + "loss": 0.2274, + "step": 43787 + }, + { + "epoch": 3.0385122475886477, + "grad_norm": 1.353851318359375, + "learning_rate": 2.8804312870829786e-05, + "loss": 0.238, + "step": 43788 + }, + { + "epoch": 3.038581639025744, + "grad_norm": 0.924576997756958, + "learning_rate": 2.8800367461112376e-05, + "loss": 0.2118, + "step": 43789 + }, + { + "epoch": 3.038651030462841, + "grad_norm": 1.0140329599380493, + "learning_rate": 2.879642227616396e-05, + "loss": 0.4102, + "step": 43790 + }, + { + "epoch": 3.0387204218999377, + "grad_norm": 0.6798906326293945, + "learning_rate": 2.8792477315996967e-05, + "loss": 0.2217, + "step": 43791 + }, + { + "epoch": 3.038789813337034, + "grad_norm": 1.0200409889221191, + "learning_rate": 2.8788532580623872e-05, + "loss": 0.2341, + "step": 43792 + }, + { + "epoch": 3.038859204774131, + "grad_norm": 0.9913458228111267, + "learning_rate": 2.8784588070057115e-05, + "loss": 0.2808, + "step": 43793 + }, + { + "epoch": 3.0389285962112274, + "grad_norm": 0.9800493717193604, + "learning_rate": 2.878064378430918e-05, + "loss": 0.2387, + "step": 43794 + }, + { + "epoch": 3.038997987648324, + "grad_norm": 0.8991262316703796, + "learning_rate": 2.8776699723392465e-05, + "loss": 0.2108, + "step": 43795 + }, + { + "epoch": 3.039067379085421, + "grad_norm": 0.9704573750495911, + "learning_rate": 2.8772755887319437e-05, + "loss": 0.439, + "step": 43796 + }, + { + "epoch": 3.0391367705225174, + "grad_norm": 0.8163243532180786, + "learning_rate": 2.8768812276102553e-05, + "loss": 0.27, + "step": 43797 + }, + { + "epoch": 3.0392061619596142, + "grad_norm": 1.0529613494873047, + "learning_rate": 2.876486888975427e-05, + "loss": 0.3034, + "step": 43798 + }, + { + "epoch": 3.0392755533967106, + "grad_norm": 0.7793115973472595, + "learning_rate": 2.8760925728287025e-05, + "loss": 0.3079, + "step": 43799 + }, + { + "epoch": 3.0393449448338075, + "grad_norm": 1.0686630010604858, + "learning_rate": 2.8756982791713283e-05, + "loss": 0.265, + "step": 43800 + }, + { + "epoch": 3.0394143362709043, + "grad_norm": 1.1701117753982544, + "learning_rate": 2.87530400800455e-05, + "loss": 0.3563, + "step": 43801 + }, + { + "epoch": 3.0394837277080007, + "grad_norm": 0.9736099243164062, + "learning_rate": 2.874909759329608e-05, + "loss": 0.2348, + "step": 43802 + }, + { + "epoch": 3.0395531191450975, + "grad_norm": 0.875536322593689, + "learning_rate": 2.874515533147749e-05, + "loss": 0.255, + "step": 43803 + }, + { + "epoch": 3.0396225105821943, + "grad_norm": 0.6848970651626587, + "learning_rate": 2.8741213294602176e-05, + "loss": 0.2232, + "step": 43804 + }, + { + "epoch": 3.0396919020192907, + "grad_norm": 0.7246642708778381, + "learning_rate": 2.8737271482682613e-05, + "loss": 0.2237, + "step": 43805 + }, + { + "epoch": 3.0397612934563876, + "grad_norm": 0.7816585898399353, + "learning_rate": 2.8733329895731196e-05, + "loss": 0.2701, + "step": 43806 + }, + { + "epoch": 3.039830684893484, + "grad_norm": 0.8081578016281128, + "learning_rate": 2.872938853376038e-05, + "loss": 0.2783, + "step": 43807 + }, + { + "epoch": 3.039900076330581, + "grad_norm": 1.0954854488372803, + "learning_rate": 2.8725447396782613e-05, + "loss": 0.1822, + "step": 43808 + }, + { + "epoch": 3.0399694677676776, + "grad_norm": 0.8570713996887207, + "learning_rate": 2.872150648481038e-05, + "loss": 0.2359, + "step": 43809 + }, + { + "epoch": 3.040038859204774, + "grad_norm": 1.0961666107177734, + "learning_rate": 2.871756579785604e-05, + "loss": 0.2367, + "step": 43810 + }, + { + "epoch": 3.040108250641871, + "grad_norm": 1.330314040184021, + "learning_rate": 2.8713625335932082e-05, + "loss": 0.2596, + "step": 43811 + }, + { + "epoch": 3.0401776420789677, + "grad_norm": 0.7538453340530396, + "learning_rate": 2.870968509905093e-05, + "loss": 0.2194, + "step": 43812 + }, + { + "epoch": 3.040247033516064, + "grad_norm": 0.7090739607810974, + "learning_rate": 2.8705745087225032e-05, + "loss": 0.2376, + "step": 43813 + }, + { + "epoch": 3.040316424953161, + "grad_norm": 1.2263070344924927, + "learning_rate": 2.8701805300466832e-05, + "loss": 0.257, + "step": 43814 + }, + { + "epoch": 3.0403858163902573, + "grad_norm": 0.8991519808769226, + "learning_rate": 2.8697865738788787e-05, + "loss": 0.2358, + "step": 43815 + }, + { + "epoch": 3.040455207827354, + "grad_norm": 0.7317352890968323, + "learning_rate": 2.8693926402203276e-05, + "loss": 0.2899, + "step": 43816 + }, + { + "epoch": 3.040524599264451, + "grad_norm": 0.8166071176528931, + "learning_rate": 2.8689987290722764e-05, + "loss": 0.231, + "step": 43817 + }, + { + "epoch": 3.0405939907015473, + "grad_norm": 1.0264540910720825, + "learning_rate": 2.86860484043597e-05, + "loss": 0.1605, + "step": 43818 + }, + { + "epoch": 3.040663382138644, + "grad_norm": 0.9084092974662781, + "learning_rate": 2.8682109743126496e-05, + "loss": 0.2753, + "step": 43819 + }, + { + "epoch": 3.0407327735757406, + "grad_norm": 1.4128382205963135, + "learning_rate": 2.8678171307035596e-05, + "loss": 0.3167, + "step": 43820 + }, + { + "epoch": 3.0408021650128374, + "grad_norm": 0.9624930620193481, + "learning_rate": 2.8674233096099467e-05, + "loss": 0.3714, + "step": 43821 + }, + { + "epoch": 3.0408715564499342, + "grad_norm": 1.22248113155365, + "learning_rate": 2.8670295110330482e-05, + "loss": 0.4209, + "step": 43822 + }, + { + "epoch": 3.0409409478870306, + "grad_norm": 1.2885727882385254, + "learning_rate": 2.8666357349741103e-05, + "loss": 0.3536, + "step": 43823 + }, + { + "epoch": 3.0410103393241275, + "grad_norm": 2.953380823135376, + "learning_rate": 2.8662419814343754e-05, + "loss": 0.2572, + "step": 43824 + }, + { + "epoch": 3.041079730761224, + "grad_norm": 1.0158559083938599, + "learning_rate": 2.8658482504150864e-05, + "loss": 0.2504, + "step": 43825 + }, + { + "epoch": 3.0411491221983207, + "grad_norm": 0.8566833734512329, + "learning_rate": 2.8654545419174872e-05, + "loss": 0.2203, + "step": 43826 + }, + { + "epoch": 3.0412185136354175, + "grad_norm": 0.803348183631897, + "learning_rate": 2.8650608559428204e-05, + "loss": 0.3119, + "step": 43827 + }, + { + "epoch": 3.041287905072514, + "grad_norm": 1.1693569421768188, + "learning_rate": 2.8646671924923307e-05, + "loss": 0.2619, + "step": 43828 + }, + { + "epoch": 3.0413572965096107, + "grad_norm": 1.0782369375228882, + "learning_rate": 2.864273551567257e-05, + "loss": 0.3471, + "step": 43829 + }, + { + "epoch": 3.0414266879467076, + "grad_norm": 0.819211483001709, + "learning_rate": 2.8638799331688437e-05, + "loss": 0.1484, + "step": 43830 + }, + { + "epoch": 3.041496079383804, + "grad_norm": 0.8761317729949951, + "learning_rate": 2.8634863372983323e-05, + "loss": 0.2301, + "step": 43831 + }, + { + "epoch": 3.041565470820901, + "grad_norm": 0.8514412045478821, + "learning_rate": 2.8630927639569684e-05, + "loss": 0.1854, + "step": 43832 + }, + { + "epoch": 3.041634862257997, + "grad_norm": 1.019474744796753, + "learning_rate": 2.8626992131459907e-05, + "loss": 0.2314, + "step": 43833 + }, + { + "epoch": 3.041704253695094, + "grad_norm": 0.9797388315200806, + "learning_rate": 2.862305684866645e-05, + "loss": 0.1898, + "step": 43834 + }, + { + "epoch": 3.041773645132191, + "grad_norm": 1.1765230894088745, + "learning_rate": 2.8619121791201742e-05, + "loss": 0.2776, + "step": 43835 + }, + { + "epoch": 3.0418430365692872, + "grad_norm": 0.6859154105186462, + "learning_rate": 2.8615186959078156e-05, + "loss": 0.2293, + "step": 43836 + }, + { + "epoch": 3.041912428006384, + "grad_norm": 0.8243829011917114, + "learning_rate": 2.861125235230815e-05, + "loss": 0.2283, + "step": 43837 + }, + { + "epoch": 3.041981819443481, + "grad_norm": 1.1393582820892334, + "learning_rate": 2.8607317970904136e-05, + "loss": 0.3699, + "step": 43838 + }, + { + "epoch": 3.0420512108805773, + "grad_norm": 1.0787591934204102, + "learning_rate": 2.8603383814878528e-05, + "loss": 0.58, + "step": 43839 + }, + { + "epoch": 3.042120602317674, + "grad_norm": 1.0012940168380737, + "learning_rate": 2.859944988424379e-05, + "loss": 0.217, + "step": 43840 + }, + { + "epoch": 3.0421899937547705, + "grad_norm": 0.9762204885482788, + "learning_rate": 2.8595516179012273e-05, + "loss": 0.1848, + "step": 43841 + }, + { + "epoch": 3.0422593851918673, + "grad_norm": 1.3147263526916504, + "learning_rate": 2.8591582699196428e-05, + "loss": 0.3212, + "step": 43842 + }, + { + "epoch": 3.042328776628964, + "grad_norm": 1.167302131652832, + "learning_rate": 2.858764944480866e-05, + "loss": 0.2781, + "step": 43843 + }, + { + "epoch": 3.0423981680660606, + "grad_norm": 0.6625606417655945, + "learning_rate": 2.858371641586145e-05, + "loss": 0.153, + "step": 43844 + }, + { + "epoch": 3.0424675595031574, + "grad_norm": 1.157934308052063, + "learning_rate": 2.8579783612367118e-05, + "loss": 0.3618, + "step": 43845 + }, + { + "epoch": 3.042536950940254, + "grad_norm": 1.2237873077392578, + "learning_rate": 2.857585103433812e-05, + "loss": 0.2851, + "step": 43846 + }, + { + "epoch": 3.0426063423773506, + "grad_norm": 1.1093246936798096, + "learning_rate": 2.857191868178688e-05, + "loss": 0.2877, + "step": 43847 + }, + { + "epoch": 3.0426757338144474, + "grad_norm": 0.8060447573661804, + "learning_rate": 2.85679865547258e-05, + "loss": 0.2818, + "step": 43848 + }, + { + "epoch": 3.042745125251544, + "grad_norm": 0.7756507396697998, + "learning_rate": 2.8564054653167337e-05, + "loss": 0.2416, + "step": 43849 + }, + { + "epoch": 3.0428145166886407, + "grad_norm": 0.7909820675849915, + "learning_rate": 2.8560122977123826e-05, + "loss": 0.2246, + "step": 43850 + }, + { + "epoch": 3.0428839081257375, + "grad_norm": 0.6001265645027161, + "learning_rate": 2.8556191526607724e-05, + "loss": 0.2434, + "step": 43851 + }, + { + "epoch": 3.042953299562834, + "grad_norm": 1.3830875158309937, + "learning_rate": 2.855226030163144e-05, + "loss": 0.3305, + "step": 43852 + }, + { + "epoch": 3.0430226909999307, + "grad_norm": 0.7685546278953552, + "learning_rate": 2.8548329302207368e-05, + "loss": 0.213, + "step": 43853 + }, + { + "epoch": 3.043092082437027, + "grad_norm": 1.1764253377914429, + "learning_rate": 2.8544398528347936e-05, + "loss": 0.273, + "step": 43854 + }, + { + "epoch": 3.043161473874124, + "grad_norm": 1.1121459007263184, + "learning_rate": 2.8540467980065555e-05, + "loss": 0.3028, + "step": 43855 + }, + { + "epoch": 3.043230865311221, + "grad_norm": 0.8732079267501831, + "learning_rate": 2.8536537657372653e-05, + "loss": 0.2074, + "step": 43856 + }, + { + "epoch": 3.043300256748317, + "grad_norm": 0.7799206972122192, + "learning_rate": 2.8532607560281575e-05, + "loss": 0.2529, + "step": 43857 + }, + { + "epoch": 3.043369648185414, + "grad_norm": 0.8269470930099487, + "learning_rate": 2.852867768880477e-05, + "loss": 0.2943, + "step": 43858 + }, + { + "epoch": 3.0434390396225104, + "grad_norm": 0.6752244234085083, + "learning_rate": 2.852474804295463e-05, + "loss": 0.3326, + "step": 43859 + }, + { + "epoch": 3.0435084310596072, + "grad_norm": 0.8538930416107178, + "learning_rate": 2.8520818622743572e-05, + "loss": 0.2797, + "step": 43860 + }, + { + "epoch": 3.043577822496704, + "grad_norm": 1.0934619903564453, + "learning_rate": 2.851688942818399e-05, + "loss": 0.4114, + "step": 43861 + }, + { + "epoch": 3.0436472139338004, + "grad_norm": 0.5785646438598633, + "learning_rate": 2.8512960459288308e-05, + "loss": 0.1726, + "step": 43862 + }, + { + "epoch": 3.0437166053708973, + "grad_norm": 1.134507656097412, + "learning_rate": 2.8509031716068934e-05, + "loss": 0.3577, + "step": 43863 + }, + { + "epoch": 3.043785996807994, + "grad_norm": 0.7417473196983337, + "learning_rate": 2.850510319853823e-05, + "loss": 0.2467, + "step": 43864 + }, + { + "epoch": 3.0438553882450905, + "grad_norm": 0.8010101914405823, + "learning_rate": 2.8501174906708617e-05, + "loss": 0.1961, + "step": 43865 + }, + { + "epoch": 3.0439247796821873, + "grad_norm": 0.90811687707901, + "learning_rate": 2.8497246840592496e-05, + "loss": 0.2836, + "step": 43866 + }, + { + "epoch": 3.0439941711192837, + "grad_norm": 0.9439486861228943, + "learning_rate": 2.8493319000202267e-05, + "loss": 0.4219, + "step": 43867 + }, + { + "epoch": 3.0440635625563806, + "grad_norm": 1.0335150957107544, + "learning_rate": 2.8489391385550347e-05, + "loss": 0.2836, + "step": 43868 + }, + { + "epoch": 3.0441329539934774, + "grad_norm": 0.9083585739135742, + "learning_rate": 2.848546399664911e-05, + "loss": 0.3008, + "step": 43869 + }, + { + "epoch": 3.0442023454305738, + "grad_norm": 0.8400281667709351, + "learning_rate": 2.8481536833511e-05, + "loss": 0.2393, + "step": 43870 + }, + { + "epoch": 3.0442717368676706, + "grad_norm": 0.6693915128707886, + "learning_rate": 2.847760989614836e-05, + "loss": 0.2623, + "step": 43871 + }, + { + "epoch": 3.044341128304767, + "grad_norm": 0.6923505663871765, + "learning_rate": 2.8473683184573597e-05, + "loss": 0.2284, + "step": 43872 + }, + { + "epoch": 3.044410519741864, + "grad_norm": 0.685096263885498, + "learning_rate": 2.846975669879912e-05, + "loss": 0.1753, + "step": 43873 + }, + { + "epoch": 3.0444799111789607, + "grad_norm": 1.0687450170516968, + "learning_rate": 2.8465830438837314e-05, + "loss": 0.2408, + "step": 43874 + }, + { + "epoch": 3.044549302616057, + "grad_norm": 1.0140653848648071, + "learning_rate": 2.8461904404700623e-05, + "loss": 0.231, + "step": 43875 + }, + { + "epoch": 3.044618694053154, + "grad_norm": 0.910933792591095, + "learning_rate": 2.8457978596401367e-05, + "loss": 0.2989, + "step": 43876 + }, + { + "epoch": 3.0446880854902507, + "grad_norm": 1.2332074642181396, + "learning_rate": 2.8454053013951975e-05, + "loss": 0.2805, + "step": 43877 + }, + { + "epoch": 3.044757476927347, + "grad_norm": 1.2371833324432373, + "learning_rate": 2.8450127657364833e-05, + "loss": 0.3002, + "step": 43878 + }, + { + "epoch": 3.044826868364444, + "grad_norm": 0.67134690284729, + "learning_rate": 2.8446202526652365e-05, + "loss": 0.2101, + "step": 43879 + }, + { + "epoch": 3.0448962598015403, + "grad_norm": 1.2977808713912964, + "learning_rate": 2.8442277621826908e-05, + "loss": 0.2859, + "step": 43880 + }, + { + "epoch": 3.044965651238637, + "grad_norm": 1.1739424467086792, + "learning_rate": 2.843835294290087e-05, + "loss": 0.2047, + "step": 43881 + }, + { + "epoch": 3.045035042675734, + "grad_norm": 0.5646804571151733, + "learning_rate": 2.8434428489886655e-05, + "loss": 0.172, + "step": 43882 + }, + { + "epoch": 3.0451044341128304, + "grad_norm": 1.582371711730957, + "learning_rate": 2.8430504262796653e-05, + "loss": 0.3071, + "step": 43883 + }, + { + "epoch": 3.045173825549927, + "grad_norm": 0.8057247400283813, + "learning_rate": 2.8426580261643265e-05, + "loss": 0.2098, + "step": 43884 + }, + { + "epoch": 3.0452432169870236, + "grad_norm": 0.9832282662391663, + "learning_rate": 2.8422656486438836e-05, + "loss": 0.2134, + "step": 43885 + }, + { + "epoch": 3.0453126084241204, + "grad_norm": 1.4928200244903564, + "learning_rate": 2.841873293719577e-05, + "loss": 0.3274, + "step": 43886 + }, + { + "epoch": 3.0453819998612173, + "grad_norm": 0.9391759634017944, + "learning_rate": 2.8414809613926464e-05, + "loss": 0.3364, + "step": 43887 + }, + { + "epoch": 3.0454513912983137, + "grad_norm": 0.7617693543434143, + "learning_rate": 2.84108865166433e-05, + "loss": 0.2788, + "step": 43888 + }, + { + "epoch": 3.0455207827354105, + "grad_norm": 0.9380422830581665, + "learning_rate": 2.8406963645358654e-05, + "loss": 0.2469, + "step": 43889 + }, + { + "epoch": 3.0455901741725073, + "grad_norm": 0.8731606602668762, + "learning_rate": 2.8403041000084918e-05, + "loss": 0.3071, + "step": 43890 + }, + { + "epoch": 3.0456595656096037, + "grad_norm": 0.9395721554756165, + "learning_rate": 2.8399118580834503e-05, + "loss": 0.3983, + "step": 43891 + }, + { + "epoch": 3.0457289570467005, + "grad_norm": 1.1217708587646484, + "learning_rate": 2.839519638761975e-05, + "loss": 0.2165, + "step": 43892 + }, + { + "epoch": 3.045798348483797, + "grad_norm": 0.5609440803527832, + "learning_rate": 2.839127442045304e-05, + "loss": 0.1302, + "step": 43893 + }, + { + "epoch": 3.0458677399208938, + "grad_norm": 1.2882521152496338, + "learning_rate": 2.8387352679346767e-05, + "loss": 0.451, + "step": 43894 + }, + { + "epoch": 3.0459371313579906, + "grad_norm": 0.6317625045776367, + "learning_rate": 2.8383431164313324e-05, + "loss": 0.248, + "step": 43895 + }, + { + "epoch": 3.046006522795087, + "grad_norm": 0.9627262353897095, + "learning_rate": 2.8379509875365064e-05, + "loss": 0.2861, + "step": 43896 + }, + { + "epoch": 3.046075914232184, + "grad_norm": 0.6904631853103638, + "learning_rate": 2.837558881251442e-05, + "loss": 0.2236, + "step": 43897 + }, + { + "epoch": 3.04614530566928, + "grad_norm": 0.8688303232192993, + "learning_rate": 2.837166797577371e-05, + "loss": 0.3156, + "step": 43898 + }, + { + "epoch": 3.046214697106377, + "grad_norm": 0.9288737773895264, + "learning_rate": 2.836774736515533e-05, + "loss": 0.1797, + "step": 43899 + }, + { + "epoch": 3.046284088543474, + "grad_norm": 1.1085439920425415, + "learning_rate": 2.8363826980671648e-05, + "loss": 0.3759, + "step": 43900 + }, + { + "epoch": 3.0463534799805703, + "grad_norm": 1.1258193254470825, + "learning_rate": 2.8359906822335068e-05, + "loss": 0.2716, + "step": 43901 + }, + { + "epoch": 3.046422871417667, + "grad_norm": 1.1545021533966064, + "learning_rate": 2.835598689015794e-05, + "loss": 0.3295, + "step": 43902 + }, + { + "epoch": 3.046492262854764, + "grad_norm": 0.9685917496681213, + "learning_rate": 2.8352067184152653e-05, + "loss": 0.1972, + "step": 43903 + }, + { + "epoch": 3.0465616542918603, + "grad_norm": 1.1665784120559692, + "learning_rate": 2.8348147704331608e-05, + "loss": 0.2692, + "step": 43904 + }, + { + "epoch": 3.046631045728957, + "grad_norm": 0.86556476354599, + "learning_rate": 2.834422845070712e-05, + "loss": 0.2234, + "step": 43905 + }, + { + "epoch": 3.0467004371660535, + "grad_norm": 0.9055364727973938, + "learning_rate": 2.8340309423291588e-05, + "loss": 0.2374, + "step": 43906 + }, + { + "epoch": 3.0467698286031504, + "grad_norm": 1.1722677946090698, + "learning_rate": 2.8336390622097386e-05, + "loss": 0.2762, + "step": 43907 + }, + { + "epoch": 3.046839220040247, + "grad_norm": 0.7711220383644104, + "learning_rate": 2.8332472047136882e-05, + "loss": 0.3682, + "step": 43908 + }, + { + "epoch": 3.0469086114773436, + "grad_norm": 0.8938001990318298, + "learning_rate": 2.8328553698422443e-05, + "loss": 0.3021, + "step": 43909 + }, + { + "epoch": 3.0469780029144404, + "grad_norm": 0.9710903763771057, + "learning_rate": 2.8324635575966486e-05, + "loss": 0.2176, + "step": 43910 + }, + { + "epoch": 3.047047394351537, + "grad_norm": 0.9438758492469788, + "learning_rate": 2.8320717679781306e-05, + "loss": 0.3033, + "step": 43911 + }, + { + "epoch": 3.0471167857886337, + "grad_norm": 0.921809196472168, + "learning_rate": 2.8316800009879297e-05, + "loss": 0.3096, + "step": 43912 + }, + { + "epoch": 3.0471861772257305, + "grad_norm": 0.743623673915863, + "learning_rate": 2.8312882566272837e-05, + "loss": 0.2371, + "step": 43913 + }, + { + "epoch": 3.047255568662827, + "grad_norm": 0.9652242660522461, + "learning_rate": 2.8308965348974315e-05, + "loss": 0.2338, + "step": 43914 + }, + { + "epoch": 3.0473249600999237, + "grad_norm": 0.8866890072822571, + "learning_rate": 2.8305048357996055e-05, + "loss": 0.2883, + "step": 43915 + }, + { + "epoch": 3.0473943515370205, + "grad_norm": 0.969035804271698, + "learning_rate": 2.8301131593350426e-05, + "loss": 0.4129, + "step": 43916 + }, + { + "epoch": 3.047463742974117, + "grad_norm": 0.7521989941596985, + "learning_rate": 2.8297215055049808e-05, + "loss": 0.1866, + "step": 43917 + }, + { + "epoch": 3.0475331344112138, + "grad_norm": 0.594866156578064, + "learning_rate": 2.82932987431066e-05, + "loss": 0.1636, + "step": 43918 + }, + { + "epoch": 3.04760252584831, + "grad_norm": 0.7855798602104187, + "learning_rate": 2.828938265753309e-05, + "loss": 0.2822, + "step": 43919 + }, + { + "epoch": 3.047671917285407, + "grad_norm": 0.728733241558075, + "learning_rate": 2.8285466798341688e-05, + "loss": 0.2289, + "step": 43920 + }, + { + "epoch": 3.047741308722504, + "grad_norm": 0.913021981716156, + "learning_rate": 2.828155116554474e-05, + "loss": 0.2728, + "step": 43921 + }, + { + "epoch": 3.0478107001596, + "grad_norm": 0.7059386372566223, + "learning_rate": 2.8277635759154618e-05, + "loss": 0.2107, + "step": 43922 + }, + { + "epoch": 3.047880091596697, + "grad_norm": 1.3504689931869507, + "learning_rate": 2.8273720579183672e-05, + "loss": 0.3726, + "step": 43923 + }, + { + "epoch": 3.0479494830337934, + "grad_norm": 0.785926878452301, + "learning_rate": 2.8269805625644273e-05, + "loss": 0.2818, + "step": 43924 + }, + { + "epoch": 3.0480188744708903, + "grad_norm": 0.824287474155426, + "learning_rate": 2.826589089854881e-05, + "loss": 0.2784, + "step": 43925 + }, + { + "epoch": 3.048088265907987, + "grad_norm": 0.7159253358840942, + "learning_rate": 2.8261976397909572e-05, + "loss": 0.2173, + "step": 43926 + }, + { + "epoch": 3.0481576573450835, + "grad_norm": 1.1225824356079102, + "learning_rate": 2.825806212373895e-05, + "loss": 0.3675, + "step": 43927 + }, + { + "epoch": 3.0482270487821803, + "grad_norm": 0.7883551120758057, + "learning_rate": 2.8254148076049304e-05, + "loss": 0.1701, + "step": 43928 + }, + { + "epoch": 3.048296440219277, + "grad_norm": 1.2403247356414795, + "learning_rate": 2.8250234254852992e-05, + "loss": 0.3139, + "step": 43929 + }, + { + "epoch": 3.0483658316563735, + "grad_norm": 0.8568710684776306, + "learning_rate": 2.8246320660162364e-05, + "loss": 0.268, + "step": 43930 + }, + { + "epoch": 3.0484352230934704, + "grad_norm": 0.801889955997467, + "learning_rate": 2.824240729198977e-05, + "loss": 0.1505, + "step": 43931 + }, + { + "epoch": 3.0485046145305668, + "grad_norm": 1.091323733329773, + "learning_rate": 2.8238494150347604e-05, + "loss": 0.1984, + "step": 43932 + }, + { + "epoch": 3.0485740059676636, + "grad_norm": 0.8664121031761169, + "learning_rate": 2.8234581235248158e-05, + "loss": 0.2496, + "step": 43933 + }, + { + "epoch": 3.0486433974047604, + "grad_norm": 0.8577644228935242, + "learning_rate": 2.823066854670382e-05, + "loss": 0.2657, + "step": 43934 + }, + { + "epoch": 3.048712788841857, + "grad_norm": 0.7317853569984436, + "learning_rate": 2.8226756084726925e-05, + "loss": 0.2493, + "step": 43935 + }, + { + "epoch": 3.0487821802789536, + "grad_norm": 0.5481078028678894, + "learning_rate": 2.8222843849329827e-05, + "loss": 0.1768, + "step": 43936 + }, + { + "epoch": 3.04885157171605, + "grad_norm": 1.1072138547897339, + "learning_rate": 2.8218931840524898e-05, + "loss": 0.3402, + "step": 43937 + }, + { + "epoch": 3.048920963153147, + "grad_norm": 0.981380045413971, + "learning_rate": 2.821502005832447e-05, + "loss": 0.363, + "step": 43938 + }, + { + "epoch": 3.0489903545902437, + "grad_norm": 0.7777704000473022, + "learning_rate": 2.8211108502740922e-05, + "loss": 0.2303, + "step": 43939 + }, + { + "epoch": 3.04905974602734, + "grad_norm": 1.440295696258545, + "learning_rate": 2.8207197173786548e-05, + "loss": 0.3957, + "step": 43940 + }, + { + "epoch": 3.049129137464437, + "grad_norm": 0.696552574634552, + "learning_rate": 2.820328607147371e-05, + "loss": 0.1809, + "step": 43941 + }, + { + "epoch": 3.0491985289015338, + "grad_norm": 0.7856448888778687, + "learning_rate": 2.819937519581478e-05, + "loss": 0.3449, + "step": 43942 + }, + { + "epoch": 3.04926792033863, + "grad_norm": 1.0580717325210571, + "learning_rate": 2.819546454682209e-05, + "loss": 0.2295, + "step": 43943 + }, + { + "epoch": 3.049337311775727, + "grad_norm": 1.2473065853118896, + "learning_rate": 2.819155412450799e-05, + "loss": 0.3331, + "step": 43944 + }, + { + "epoch": 3.0494067032128234, + "grad_norm": 0.9654088616371155, + "learning_rate": 2.8187643928884853e-05, + "loss": 0.2137, + "step": 43945 + }, + { + "epoch": 3.04947609464992, + "grad_norm": 1.2980273962020874, + "learning_rate": 2.8183733959964963e-05, + "loss": 0.2785, + "step": 43946 + }, + { + "epoch": 3.049545486087017, + "grad_norm": 0.9862164258956909, + "learning_rate": 2.8179824217760685e-05, + "loss": 0.2917, + "step": 43947 + }, + { + "epoch": 3.0496148775241134, + "grad_norm": 0.7955218553543091, + "learning_rate": 2.8175914702284413e-05, + "loss": 0.2491, + "step": 43948 + }, + { + "epoch": 3.0496842689612103, + "grad_norm": 0.680835485458374, + "learning_rate": 2.8172005413548398e-05, + "loss": 0.2201, + "step": 43949 + }, + { + "epoch": 3.0497536603983066, + "grad_norm": 0.8821448087692261, + "learning_rate": 2.8168096351565044e-05, + "loss": 0.2957, + "step": 43950 + }, + { + "epoch": 3.0498230518354035, + "grad_norm": 0.7817618250846863, + "learning_rate": 2.816418751634666e-05, + "loss": 0.2562, + "step": 43951 + }, + { + "epoch": 3.0498924432725003, + "grad_norm": 0.7001160383224487, + "learning_rate": 2.8160278907905612e-05, + "loss": 0.1614, + "step": 43952 + }, + { + "epoch": 3.0499618347095967, + "grad_norm": 0.8412532210350037, + "learning_rate": 2.815637052625426e-05, + "loss": 0.241, + "step": 43953 + }, + { + "epoch": 3.0500312261466935, + "grad_norm": 1.087530255317688, + "learning_rate": 2.8152462371404877e-05, + "loss": 0.2407, + "step": 43954 + }, + { + "epoch": 3.0501006175837904, + "grad_norm": 0.6310089230537415, + "learning_rate": 2.8148554443369836e-05, + "loss": 0.2024, + "step": 43955 + }, + { + "epoch": 3.0501700090208868, + "grad_norm": 0.8699630498886108, + "learning_rate": 2.8144646742161463e-05, + "loss": 0.321, + "step": 43956 + }, + { + "epoch": 3.0502394004579836, + "grad_norm": 0.7976424098014832, + "learning_rate": 2.8140739267792116e-05, + "loss": 0.2761, + "step": 43957 + }, + { + "epoch": 3.05030879189508, + "grad_norm": 1.1661373376846313, + "learning_rate": 2.8136832020274105e-05, + "loss": 0.3566, + "step": 43958 + }, + { + "epoch": 3.050378183332177, + "grad_norm": 0.9370436072349548, + "learning_rate": 2.8132924999619792e-05, + "loss": 0.3114, + "step": 43959 + }, + { + "epoch": 3.0504475747692736, + "grad_norm": 0.8037644624710083, + "learning_rate": 2.812901820584152e-05, + "loss": 0.2912, + "step": 43960 + }, + { + "epoch": 3.05051696620637, + "grad_norm": 0.9363127946853638, + "learning_rate": 2.8125111638951575e-05, + "loss": 0.2601, + "step": 43961 + }, + { + "epoch": 3.050586357643467, + "grad_norm": 0.6556947827339172, + "learning_rate": 2.8121205298962306e-05, + "loss": 0.1591, + "step": 43962 + }, + { + "epoch": 3.0506557490805633, + "grad_norm": 0.5173394680023193, + "learning_rate": 2.811729918588605e-05, + "loss": 0.1566, + "step": 43963 + }, + { + "epoch": 3.05072514051766, + "grad_norm": 1.1005483865737915, + "learning_rate": 2.8113393299735146e-05, + "loss": 0.2864, + "step": 43964 + }, + { + "epoch": 3.050794531954757, + "grad_norm": 0.6205269694328308, + "learning_rate": 2.8109487640521914e-05, + "loss": 0.2061, + "step": 43965 + }, + { + "epoch": 3.0508639233918533, + "grad_norm": 0.9701460003852844, + "learning_rate": 2.8105582208258695e-05, + "loss": 0.3217, + "step": 43966 + }, + { + "epoch": 3.05093331482895, + "grad_norm": 0.8514886498451233, + "learning_rate": 2.810167700295784e-05, + "loss": 0.2254, + "step": 43967 + }, + { + "epoch": 3.051002706266047, + "grad_norm": 0.91538405418396, + "learning_rate": 2.809777202463162e-05, + "loss": 0.2733, + "step": 43968 + }, + { + "epoch": 3.0510720977031434, + "grad_norm": 1.2963876724243164, + "learning_rate": 2.8093867273292384e-05, + "loss": 0.2024, + "step": 43969 + }, + { + "epoch": 3.05114148914024, + "grad_norm": 0.8010004758834839, + "learning_rate": 2.808996274895248e-05, + "loss": 0.261, + "step": 43970 + }, + { + "epoch": 3.0512108805773366, + "grad_norm": 1.1012799739837646, + "learning_rate": 2.808605845162421e-05, + "loss": 0.2753, + "step": 43971 + }, + { + "epoch": 3.0512802720144334, + "grad_norm": 1.1789206266403198, + "learning_rate": 2.8082154381319912e-05, + "loss": 0.2333, + "step": 43972 + }, + { + "epoch": 3.0513496634515302, + "grad_norm": 1.2001742124557495, + "learning_rate": 2.8078250538051942e-05, + "loss": 0.3781, + "step": 43973 + }, + { + "epoch": 3.0514190548886266, + "grad_norm": 0.806291937828064, + "learning_rate": 2.8074346921832552e-05, + "loss": 0.2165, + "step": 43974 + }, + { + "epoch": 3.0514884463257235, + "grad_norm": 1.0471079349517822, + "learning_rate": 2.8070443532674107e-05, + "loss": 0.3234, + "step": 43975 + }, + { + "epoch": 3.05155783776282, + "grad_norm": 0.5746634602546692, + "learning_rate": 2.8066540370588935e-05, + "loss": 0.1597, + "step": 43976 + }, + { + "epoch": 3.0516272291999167, + "grad_norm": 0.9049685597419739, + "learning_rate": 2.8062637435589345e-05, + "loss": 0.2629, + "step": 43977 + }, + { + "epoch": 3.0516966206370135, + "grad_norm": 0.7036727070808411, + "learning_rate": 2.805873472768765e-05, + "loss": 0.2397, + "step": 43978 + }, + { + "epoch": 3.05176601207411, + "grad_norm": 1.1181941032409668, + "learning_rate": 2.80548322468962e-05, + "loss": 0.5156, + "step": 43979 + }, + { + "epoch": 3.0518354035112067, + "grad_norm": 0.7258144617080688, + "learning_rate": 2.8050929993227326e-05, + "loss": 0.1746, + "step": 43980 + }, + { + "epoch": 3.0519047949483036, + "grad_norm": 1.0258890390396118, + "learning_rate": 2.8047027966693273e-05, + "loss": 0.3305, + "step": 43981 + }, + { + "epoch": 3.0519741863854, + "grad_norm": 1.0836201906204224, + "learning_rate": 2.8043126167306422e-05, + "loss": 0.3394, + "step": 43982 + }, + { + "epoch": 3.052043577822497, + "grad_norm": 0.929608941078186, + "learning_rate": 2.8039224595079094e-05, + "loss": 0.3384, + "step": 43983 + }, + { + "epoch": 3.052112969259593, + "grad_norm": 0.7784901857376099, + "learning_rate": 2.803532325002356e-05, + "loss": 0.2064, + "step": 43984 + }, + { + "epoch": 3.05218236069669, + "grad_norm": 1.030395269393921, + "learning_rate": 2.8031422132152153e-05, + "loss": 0.2441, + "step": 43985 + }, + { + "epoch": 3.052251752133787, + "grad_norm": 0.7355748414993286, + "learning_rate": 2.8027521241477207e-05, + "loss": 0.1566, + "step": 43986 + }, + { + "epoch": 3.0523211435708832, + "grad_norm": 0.953644871711731, + "learning_rate": 2.8023620578011047e-05, + "loss": 0.2068, + "step": 43987 + }, + { + "epoch": 3.05239053500798, + "grad_norm": 1.0835524797439575, + "learning_rate": 2.8019720141765936e-05, + "loss": 0.2576, + "step": 43988 + }, + { + "epoch": 3.0524599264450765, + "grad_norm": 0.7819769978523254, + "learning_rate": 2.801581993275422e-05, + "loss": 0.2043, + "step": 43989 + }, + { + "epoch": 3.0525293178821733, + "grad_norm": 0.8437821865081787, + "learning_rate": 2.8011919950988207e-05, + "loss": 0.2157, + "step": 43990 + }, + { + "epoch": 3.05259870931927, + "grad_norm": 0.726898729801178, + "learning_rate": 2.800802019648021e-05, + "loss": 0.2623, + "step": 43991 + }, + { + "epoch": 3.0526681007563665, + "grad_norm": 0.9414178133010864, + "learning_rate": 2.800412066924255e-05, + "loss": 0.2513, + "step": 43992 + }, + { + "epoch": 3.0527374921934634, + "grad_norm": 0.8767415881156921, + "learning_rate": 2.8000221369287514e-05, + "loss": 0.2444, + "step": 43993 + }, + { + "epoch": 3.05280688363056, + "grad_norm": 1.1095807552337646, + "learning_rate": 2.7996322296627474e-05, + "loss": 0.2611, + "step": 43994 + }, + { + "epoch": 3.0528762750676566, + "grad_norm": 1.3077646493911743, + "learning_rate": 2.799242345127464e-05, + "loss": 0.3984, + "step": 43995 + }, + { + "epoch": 3.0529456665047534, + "grad_norm": 0.6805675625801086, + "learning_rate": 2.798852483324138e-05, + "loss": 0.1901, + "step": 43996 + }, + { + "epoch": 3.05301505794185, + "grad_norm": 0.8875991106033325, + "learning_rate": 2.7984626442539998e-05, + "loss": 0.3194, + "step": 43997 + }, + { + "epoch": 3.0530844493789466, + "grad_norm": 0.7730635404586792, + "learning_rate": 2.798072827918279e-05, + "loss": 0.3135, + "step": 43998 + }, + { + "epoch": 3.0531538408160435, + "grad_norm": 1.0650768280029297, + "learning_rate": 2.797683034318207e-05, + "loss": 0.1838, + "step": 43999 + }, + { + "epoch": 3.05322323225314, + "grad_norm": 1.179710030555725, + "learning_rate": 2.797293263455013e-05, + "loss": 0.2903, + "step": 44000 + }, + { + "epoch": 3.05322323225314, + "eval_loss": 0.4787708818912506, + "eval_runtime": 3840.0512, + "eval_samples_per_second": 6.321, + "eval_steps_per_second": 0.79, + "step": 44000 + }, + { + "epoch": 3.0532926236902367, + "grad_norm": 0.7561846971511841, + "learning_rate": 2.7969035153299328e-05, + "loss": 0.2512, + "step": 44001 + }, + { + "epoch": 3.053362015127333, + "grad_norm": 1.1596802473068237, + "learning_rate": 2.7965137899441883e-05, + "loss": 0.3827, + "step": 44002 + }, + { + "epoch": 3.05343140656443, + "grad_norm": 0.9157540798187256, + "learning_rate": 2.7961240872990145e-05, + "loss": 0.3068, + "step": 44003 + }, + { + "epoch": 3.0535007980015267, + "grad_norm": 0.8035908937454224, + "learning_rate": 2.7957344073956416e-05, + "loss": 0.115, + "step": 44004 + }, + { + "epoch": 3.053570189438623, + "grad_norm": 1.00114905834198, + "learning_rate": 2.7953447502352992e-05, + "loss": 0.3255, + "step": 44005 + }, + { + "epoch": 3.05363958087572, + "grad_norm": 0.6736446619033813, + "learning_rate": 2.7949551158192168e-05, + "loss": 0.1943, + "step": 44006 + }, + { + "epoch": 3.053708972312817, + "grad_norm": 1.104393720626831, + "learning_rate": 2.7945655041486264e-05, + "loss": 0.2115, + "step": 44007 + }, + { + "epoch": 3.053778363749913, + "grad_norm": 0.9168259501457214, + "learning_rate": 2.794175915224758e-05, + "loss": 0.2812, + "step": 44008 + }, + { + "epoch": 3.05384775518701, + "grad_norm": 1.7509084939956665, + "learning_rate": 2.793786349048839e-05, + "loss": 0.2905, + "step": 44009 + }, + { + "epoch": 3.0539171466241064, + "grad_norm": 0.9694458246231079, + "learning_rate": 2.7933968056220995e-05, + "loss": 0.3014, + "step": 44010 + }, + { + "epoch": 3.0539865380612032, + "grad_norm": 0.8756529688835144, + "learning_rate": 2.79300728494577e-05, + "loss": 0.3502, + "step": 44011 + }, + { + "epoch": 3.0540559294983, + "grad_norm": 0.780051589012146, + "learning_rate": 2.7926177870210802e-05, + "loss": 0.1996, + "step": 44012 + }, + { + "epoch": 3.0541253209353965, + "grad_norm": 0.7601562738418579, + "learning_rate": 2.792228311849261e-05, + "loss": 0.2394, + "step": 44013 + }, + { + "epoch": 3.0541947123724933, + "grad_norm": 1.6433398723602295, + "learning_rate": 2.7918388594315427e-05, + "loss": 0.2808, + "step": 44014 + }, + { + "epoch": 3.0542641038095897, + "grad_norm": 0.9025479555130005, + "learning_rate": 2.7914494297691497e-05, + "loss": 0.2828, + "step": 44015 + }, + { + "epoch": 3.0543334952466865, + "grad_norm": 0.7690916657447815, + "learning_rate": 2.7910600228633156e-05, + "loss": 0.2416, + "step": 44016 + }, + { + "epoch": 3.0544028866837833, + "grad_norm": 1.2678858041763306, + "learning_rate": 2.790670638715268e-05, + "loss": 0.2953, + "step": 44017 + }, + { + "epoch": 3.0544722781208797, + "grad_norm": 0.8503758907318115, + "learning_rate": 2.7902812773262398e-05, + "loss": 0.2965, + "step": 44018 + }, + { + "epoch": 3.0545416695579766, + "grad_norm": 1.3131828308105469, + "learning_rate": 2.789891938697453e-05, + "loss": 0.2823, + "step": 44019 + }, + { + "epoch": 3.0546110609950734, + "grad_norm": 0.9649742841720581, + "learning_rate": 2.7895026228301425e-05, + "loss": 0.291, + "step": 44020 + }, + { + "epoch": 3.05468045243217, + "grad_norm": 2.125471830368042, + "learning_rate": 2.7891133297255345e-05, + "loss": 0.2964, + "step": 44021 + }, + { + "epoch": 3.0547498438692666, + "grad_norm": 1.1290210485458374, + "learning_rate": 2.788724059384863e-05, + "loss": 0.2399, + "step": 44022 + }, + { + "epoch": 3.054819235306363, + "grad_norm": 2.220444917678833, + "learning_rate": 2.78833481180935e-05, + "loss": 0.4301, + "step": 44023 + }, + { + "epoch": 3.05488862674346, + "grad_norm": 3.2651288509368896, + "learning_rate": 2.7879455870002268e-05, + "loss": 0.2218, + "step": 44024 + }, + { + "epoch": 3.0549580181805567, + "grad_norm": 1.0551010370254517, + "learning_rate": 2.7875563849587226e-05, + "loss": 0.3721, + "step": 44025 + }, + { + "epoch": 3.055027409617653, + "grad_norm": 0.6958938241004944, + "learning_rate": 2.7871672056860654e-05, + "loss": 0.2468, + "step": 44026 + }, + { + "epoch": 3.05509680105475, + "grad_norm": 1.3500220775604248, + "learning_rate": 2.786778049183486e-05, + "loss": 0.2975, + "step": 44027 + }, + { + "epoch": 3.0551661924918463, + "grad_norm": 0.9045283198356628, + "learning_rate": 2.7863889154522106e-05, + "loss": 0.2134, + "step": 44028 + }, + { + "epoch": 3.055235583928943, + "grad_norm": 1.232002854347229, + "learning_rate": 2.785999804493471e-05, + "loss": 0.3039, + "step": 44029 + }, + { + "epoch": 3.05530497536604, + "grad_norm": 0.9665306210517883, + "learning_rate": 2.785610716308491e-05, + "loss": 0.2701, + "step": 44030 + }, + { + "epoch": 3.0553743668031363, + "grad_norm": 1.0597375631332397, + "learning_rate": 2.7852216508985006e-05, + "loss": 0.2997, + "step": 44031 + }, + { + "epoch": 3.055443758240233, + "grad_norm": 0.7942545413970947, + "learning_rate": 2.784832608264728e-05, + "loss": 0.301, + "step": 44032 + }, + { + "epoch": 3.05551314967733, + "grad_norm": 0.6368644833564758, + "learning_rate": 2.7844435884084018e-05, + "loss": 0.1945, + "step": 44033 + }, + { + "epoch": 3.0555825411144264, + "grad_norm": 0.7862067222595215, + "learning_rate": 2.7840545913307504e-05, + "loss": 0.2899, + "step": 44034 + }, + { + "epoch": 3.0556519325515232, + "grad_norm": 1.089289903640747, + "learning_rate": 2.783665617033001e-05, + "loss": 0.3555, + "step": 44035 + }, + { + "epoch": 3.0557213239886196, + "grad_norm": 0.6461989879608154, + "learning_rate": 2.7832766655163856e-05, + "loss": 0.2406, + "step": 44036 + }, + { + "epoch": 3.0557907154257165, + "grad_norm": 0.8651943206787109, + "learning_rate": 2.782887736782126e-05, + "loss": 0.4368, + "step": 44037 + }, + { + "epoch": 3.0558601068628133, + "grad_norm": 1.1168663501739502, + "learning_rate": 2.7824988308314513e-05, + "loss": 0.3067, + "step": 44038 + }, + { + "epoch": 3.0559294982999097, + "grad_norm": 1.0423548221588135, + "learning_rate": 2.782109947665591e-05, + "loss": 0.2952, + "step": 44039 + }, + { + "epoch": 3.0559988897370065, + "grad_norm": 0.839867889881134, + "learning_rate": 2.7817210872857714e-05, + "loss": 0.2993, + "step": 44040 + }, + { + "epoch": 3.0560682811741033, + "grad_norm": 1.0502945184707642, + "learning_rate": 2.7813322496932214e-05, + "loss": 0.3265, + "step": 44041 + }, + { + "epoch": 3.0561376726111997, + "grad_norm": 0.9233506321907043, + "learning_rate": 2.7809434348891683e-05, + "loss": 0.2503, + "step": 44042 + }, + { + "epoch": 3.0562070640482966, + "grad_norm": 0.743245005607605, + "learning_rate": 2.780554642874842e-05, + "loss": 0.2728, + "step": 44043 + }, + { + "epoch": 3.056276455485393, + "grad_norm": 0.9752629399299622, + "learning_rate": 2.7801658736514636e-05, + "loss": 0.3455, + "step": 44044 + }, + { + "epoch": 3.05634584692249, + "grad_norm": 1.0519179105758667, + "learning_rate": 2.7797771272202644e-05, + "loss": 0.2879, + "step": 44045 + }, + { + "epoch": 3.0564152383595866, + "grad_norm": 0.48109856247901917, + "learning_rate": 2.7793884035824714e-05, + "loss": 0.1012, + "step": 44046 + }, + { + "epoch": 3.056484629796683, + "grad_norm": 1.1665494441986084, + "learning_rate": 2.77899970273931e-05, + "loss": 0.2207, + "step": 44047 + }, + { + "epoch": 3.05655402123378, + "grad_norm": 2.215090036392212, + "learning_rate": 2.7786110246920105e-05, + "loss": 0.2165, + "step": 44048 + }, + { + "epoch": 3.0566234126708762, + "grad_norm": 1.1312637329101562, + "learning_rate": 2.7782223694417996e-05, + "loss": 0.3743, + "step": 44049 + }, + { + "epoch": 3.056692804107973, + "grad_norm": 0.808489203453064, + "learning_rate": 2.777833736989901e-05, + "loss": 0.221, + "step": 44050 + }, + { + "epoch": 3.05676219554507, + "grad_norm": 0.5997743010520935, + "learning_rate": 2.7774451273375434e-05, + "loss": 0.2093, + "step": 44051 + }, + { + "epoch": 3.0568315869821663, + "grad_norm": 0.709953784942627, + "learning_rate": 2.777056540485953e-05, + "loss": 0.2813, + "step": 44052 + }, + { + "epoch": 3.056900978419263, + "grad_norm": 1.1429027318954468, + "learning_rate": 2.77666797643636e-05, + "loss": 0.3238, + "step": 44053 + }, + { + "epoch": 3.0569703698563595, + "grad_norm": 1.2843581438064575, + "learning_rate": 2.7762794351899856e-05, + "loss": 0.2289, + "step": 44054 + }, + { + "epoch": 3.0570397612934563, + "grad_norm": 0.7542033195495605, + "learning_rate": 2.7758909167480583e-05, + "loss": 0.1671, + "step": 44055 + }, + { + "epoch": 3.057109152730553, + "grad_norm": 1.1332497596740723, + "learning_rate": 2.7755024211118087e-05, + "loss": 0.2121, + "step": 44056 + }, + { + "epoch": 3.0571785441676496, + "grad_norm": 0.7978813052177429, + "learning_rate": 2.775113948282457e-05, + "loss": 0.3125, + "step": 44057 + }, + { + "epoch": 3.0572479356047464, + "grad_norm": 1.217110514640808, + "learning_rate": 2.7747254982612316e-05, + "loss": 0.3703, + "step": 44058 + }, + { + "epoch": 3.0573173270418432, + "grad_norm": 0.6453571319580078, + "learning_rate": 2.77433707104936e-05, + "loss": 0.2523, + "step": 44059 + }, + { + "epoch": 3.0573867184789396, + "grad_norm": 1.3928613662719727, + "learning_rate": 2.7739486666480674e-05, + "loss": 0.3046, + "step": 44060 + }, + { + "epoch": 3.0574561099160364, + "grad_norm": 0.9565592408180237, + "learning_rate": 2.7735602850585818e-05, + "loss": 0.3401, + "step": 44061 + }, + { + "epoch": 3.057525501353133, + "grad_norm": 0.7592796087265015, + "learning_rate": 2.7731719262821265e-05, + "loss": 0.2252, + "step": 44062 + }, + { + "epoch": 3.0575948927902297, + "grad_norm": 1.20164954662323, + "learning_rate": 2.7727835903199327e-05, + "loss": 0.2007, + "step": 44063 + }, + { + "epoch": 3.0576642842273265, + "grad_norm": 0.8378297090530396, + "learning_rate": 2.77239527717322e-05, + "loss": 0.2184, + "step": 44064 + }, + { + "epoch": 3.057733675664423, + "grad_norm": 0.9169852137565613, + "learning_rate": 2.772006986843215e-05, + "loss": 0.2104, + "step": 44065 + }, + { + "epoch": 3.0578030671015197, + "grad_norm": 0.9880591034889221, + "learning_rate": 2.771618719331147e-05, + "loss": 0.3054, + "step": 44066 + }, + { + "epoch": 3.0578724585386166, + "grad_norm": 0.9165895581245422, + "learning_rate": 2.77123047463824e-05, + "loss": 0.2597, + "step": 44067 + }, + { + "epoch": 3.057941849975713, + "grad_norm": 0.9311310648918152, + "learning_rate": 2.7708422527657197e-05, + "loss": 0.3226, + "step": 44068 + }, + { + "epoch": 3.0580112414128098, + "grad_norm": 0.49086663126945496, + "learning_rate": 2.7704540537148103e-05, + "loss": 0.1383, + "step": 44069 + }, + { + "epoch": 3.058080632849906, + "grad_norm": 0.887367844581604, + "learning_rate": 2.770065877486744e-05, + "loss": 0.2389, + "step": 44070 + }, + { + "epoch": 3.058150024287003, + "grad_norm": 0.7557563781738281, + "learning_rate": 2.7696777240827367e-05, + "loss": 0.3014, + "step": 44071 + }, + { + "epoch": 3.0582194157241, + "grad_norm": 0.8910141587257385, + "learning_rate": 2.7692895935040175e-05, + "loss": 0.2482, + "step": 44072 + }, + { + "epoch": 3.058288807161196, + "grad_norm": 0.9022027254104614, + "learning_rate": 2.7689014857518124e-05, + "loss": 0.2044, + "step": 44073 + }, + { + "epoch": 3.058358198598293, + "grad_norm": 0.958626925945282, + "learning_rate": 2.7685134008273462e-05, + "loss": 0.3059, + "step": 44074 + }, + { + "epoch": 3.0584275900353894, + "grad_norm": 0.9332433938980103, + "learning_rate": 2.768125338731844e-05, + "loss": 0.2059, + "step": 44075 + }, + { + "epoch": 3.0584969814724863, + "grad_norm": 0.9784977436065674, + "learning_rate": 2.767737299466532e-05, + "loss": 0.3958, + "step": 44076 + }, + { + "epoch": 3.058566372909583, + "grad_norm": 0.7190515995025635, + "learning_rate": 2.7673492830326364e-05, + "loss": 0.2221, + "step": 44077 + }, + { + "epoch": 3.0586357643466795, + "grad_norm": 1.2771259546279907, + "learning_rate": 2.766961289431378e-05, + "loss": 0.2921, + "step": 44078 + }, + { + "epoch": 3.0587051557837763, + "grad_norm": 0.8779252767562866, + "learning_rate": 2.766573318663983e-05, + "loss": 0.2303, + "step": 44079 + }, + { + "epoch": 3.0587745472208727, + "grad_norm": 0.6351211667060852, + "learning_rate": 2.7661853707316765e-05, + "loss": 0.2286, + "step": 44080 + }, + { + "epoch": 3.0588439386579696, + "grad_norm": 0.9395821690559387, + "learning_rate": 2.765797445635684e-05, + "loss": 0.2297, + "step": 44081 + }, + { + "epoch": 3.0589133300950664, + "grad_norm": 1.1184816360473633, + "learning_rate": 2.7654095433772287e-05, + "loss": 0.315, + "step": 44082 + }, + { + "epoch": 3.0589827215321628, + "grad_norm": 1.0513522624969482, + "learning_rate": 2.7650216639575366e-05, + "loss": 0.303, + "step": 44083 + }, + { + "epoch": 3.0590521129692596, + "grad_norm": 0.8989686369895935, + "learning_rate": 2.7646338073778356e-05, + "loss": 0.3023, + "step": 44084 + }, + { + "epoch": 3.0591215044063564, + "grad_norm": 0.6934911608695984, + "learning_rate": 2.7642459736393422e-05, + "loss": 0.1709, + "step": 44085 + }, + { + "epoch": 3.059190895843453, + "grad_norm": 0.8422802090644836, + "learning_rate": 2.7638581627432847e-05, + "loss": 0.2156, + "step": 44086 + }, + { + "epoch": 3.0592602872805497, + "grad_norm": 0.7393664717674255, + "learning_rate": 2.7634703746908875e-05, + "loss": 0.2169, + "step": 44087 + }, + { + "epoch": 3.059329678717646, + "grad_norm": 1.0460842847824097, + "learning_rate": 2.7630826094833783e-05, + "loss": 0.2658, + "step": 44088 + }, + { + "epoch": 3.059399070154743, + "grad_norm": 0.8963404893875122, + "learning_rate": 2.7626948671219742e-05, + "loss": 0.3673, + "step": 44089 + }, + { + "epoch": 3.0594684615918397, + "grad_norm": 0.6815760731697083, + "learning_rate": 2.7623071476079033e-05, + "loss": 0.2962, + "step": 44090 + }, + { + "epoch": 3.059537853028936, + "grad_norm": 1.1123039722442627, + "learning_rate": 2.761919450942392e-05, + "loss": 0.2325, + "step": 44091 + }, + { + "epoch": 3.059607244466033, + "grad_norm": 0.7130491137504578, + "learning_rate": 2.761531777126658e-05, + "loss": 0.2073, + "step": 44092 + }, + { + "epoch": 3.0596766359031298, + "grad_norm": 0.6834646463394165, + "learning_rate": 2.7611441261619275e-05, + "loss": 0.2126, + "step": 44093 + }, + { + "epoch": 3.059746027340226, + "grad_norm": 1.080055594444275, + "learning_rate": 2.7607564980494262e-05, + "loss": 0.3054, + "step": 44094 + }, + { + "epoch": 3.059815418777323, + "grad_norm": 0.721955418586731, + "learning_rate": 2.760368892790376e-05, + "loss": 0.2465, + "step": 44095 + }, + { + "epoch": 3.0598848102144194, + "grad_norm": 0.8466538190841675, + "learning_rate": 2.7599813103860017e-05, + "loss": 0.2178, + "step": 44096 + }, + { + "epoch": 3.059954201651516, + "grad_norm": 0.7804415225982666, + "learning_rate": 2.759593750837526e-05, + "loss": 0.2297, + "step": 44097 + }, + { + "epoch": 3.060023593088613, + "grad_norm": 0.8631153106689453, + "learning_rate": 2.7592062141461762e-05, + "loss": 0.1886, + "step": 44098 + }, + { + "epoch": 3.0600929845257094, + "grad_norm": 0.9023857712745667, + "learning_rate": 2.7588187003131693e-05, + "loss": 0.2069, + "step": 44099 + }, + { + "epoch": 3.0601623759628063, + "grad_norm": 0.8188347220420837, + "learning_rate": 2.7584312093397314e-05, + "loss": 0.2337, + "step": 44100 + }, + { + "epoch": 3.0602317673999027, + "grad_norm": 0.8038988709449768, + "learning_rate": 2.7580437412270867e-05, + "loss": 0.1435, + "step": 44101 + }, + { + "epoch": 3.0603011588369995, + "grad_norm": 2.1940066814422607, + "learning_rate": 2.757656295976456e-05, + "loss": 0.3312, + "step": 44102 + }, + { + "epoch": 3.0603705502740963, + "grad_norm": 1.2966008186340332, + "learning_rate": 2.7572688735890663e-05, + "loss": 0.34, + "step": 44103 + }, + { + "epoch": 3.0604399417111927, + "grad_norm": 0.9415680170059204, + "learning_rate": 2.7568814740661365e-05, + "loss": 0.2992, + "step": 44104 + }, + { + "epoch": 3.0605093331482895, + "grad_norm": 0.8666366934776306, + "learning_rate": 2.756494097408896e-05, + "loss": 0.3593, + "step": 44105 + }, + { + "epoch": 3.0605787245853864, + "grad_norm": 1.104841709136963, + "learning_rate": 2.7561067436185595e-05, + "loss": 0.3157, + "step": 44106 + }, + { + "epoch": 3.0606481160224828, + "grad_norm": 0.6735548973083496, + "learning_rate": 2.755719412696354e-05, + "loss": 0.2062, + "step": 44107 + }, + { + "epoch": 3.0607175074595796, + "grad_norm": 0.4971572458744049, + "learning_rate": 2.755332104643501e-05, + "loss": 0.0946, + "step": 44108 + }, + { + "epoch": 3.060786898896676, + "grad_norm": 0.9559951424598694, + "learning_rate": 2.754944819461225e-05, + "loss": 0.2196, + "step": 44109 + }, + { + "epoch": 3.060856290333773, + "grad_norm": 1.4689695835113525, + "learning_rate": 2.7545575571507466e-05, + "loss": 0.2703, + "step": 44110 + }, + { + "epoch": 3.0609256817708697, + "grad_norm": 0.5649235844612122, + "learning_rate": 2.7541703177132905e-05, + "loss": 0.1807, + "step": 44111 + }, + { + "epoch": 3.060995073207966, + "grad_norm": 1.1588610410690308, + "learning_rate": 2.7537831011500804e-05, + "loss": 0.281, + "step": 44112 + }, + { + "epoch": 3.061064464645063, + "grad_norm": 1.9111944437026978, + "learning_rate": 2.7533959074623327e-05, + "loss": 0.2933, + "step": 44113 + }, + { + "epoch": 3.0611338560821593, + "grad_norm": 0.8318114876747131, + "learning_rate": 2.7530087366512746e-05, + "loss": 0.2173, + "step": 44114 + }, + { + "epoch": 3.061203247519256, + "grad_norm": 1.326949954032898, + "learning_rate": 2.752621588718126e-05, + "loss": 0.2839, + "step": 44115 + }, + { + "epoch": 3.061272638956353, + "grad_norm": 1.047745943069458, + "learning_rate": 2.7522344636641106e-05, + "loss": 0.1885, + "step": 44116 + }, + { + "epoch": 3.0613420303934493, + "grad_norm": 0.5947290062904358, + "learning_rate": 2.7518473614904507e-05, + "loss": 0.236, + "step": 44117 + }, + { + "epoch": 3.061411421830546, + "grad_norm": 0.8024067878723145, + "learning_rate": 2.751460282198367e-05, + "loss": 0.2288, + "step": 44118 + }, + { + "epoch": 3.061480813267643, + "grad_norm": 0.859688937664032, + "learning_rate": 2.7510732257890858e-05, + "loss": 0.3081, + "step": 44119 + }, + { + "epoch": 3.0615502047047394, + "grad_norm": 0.9602181911468506, + "learning_rate": 2.7506861922638216e-05, + "loss": 0.2349, + "step": 44120 + }, + { + "epoch": 3.061619596141836, + "grad_norm": 1.165769338607788, + "learning_rate": 2.750299181623801e-05, + "loss": 0.3261, + "step": 44121 + }, + { + "epoch": 3.0616889875789326, + "grad_norm": 0.8461020588874817, + "learning_rate": 2.7499121938702444e-05, + "loss": 0.1437, + "step": 44122 + }, + { + "epoch": 3.0617583790160294, + "grad_norm": 0.8723413348197937, + "learning_rate": 2.749525229004376e-05, + "loss": 0.2522, + "step": 44123 + }, + { + "epoch": 3.0618277704531263, + "grad_norm": 0.7389600872993469, + "learning_rate": 2.749138287027413e-05, + "loss": 0.2324, + "step": 44124 + }, + { + "epoch": 3.0618971618902227, + "grad_norm": 0.9887120127677917, + "learning_rate": 2.7487513679405797e-05, + "loss": 0.2881, + "step": 44125 + }, + { + "epoch": 3.0619665533273195, + "grad_norm": 1.2878421545028687, + "learning_rate": 2.7483644717450983e-05, + "loss": 0.3559, + "step": 44126 + }, + { + "epoch": 3.062035944764416, + "grad_norm": 1.326614260673523, + "learning_rate": 2.7479775984421873e-05, + "loss": 0.3067, + "step": 44127 + }, + { + "epoch": 3.0621053362015127, + "grad_norm": 1.1295698881149292, + "learning_rate": 2.74759074803307e-05, + "loss": 0.353, + "step": 44128 + }, + { + "epoch": 3.0621747276386095, + "grad_norm": 0.7303272485733032, + "learning_rate": 2.747203920518966e-05, + "loss": 0.1837, + "step": 44129 + }, + { + "epoch": 3.062244119075706, + "grad_norm": 0.9043522477149963, + "learning_rate": 2.746817115901099e-05, + "loss": 0.257, + "step": 44130 + }, + { + "epoch": 3.0623135105128028, + "grad_norm": 1.070300817489624, + "learning_rate": 2.7464303341806884e-05, + "loss": 0.3083, + "step": 44131 + }, + { + "epoch": 3.0623829019498996, + "grad_norm": 1.486438512802124, + "learning_rate": 2.7460435753589585e-05, + "loss": 0.2889, + "step": 44132 + }, + { + "epoch": 3.062452293386996, + "grad_norm": 0.7671436667442322, + "learning_rate": 2.7456568394371242e-05, + "loss": 0.2981, + "step": 44133 + }, + { + "epoch": 3.062521684824093, + "grad_norm": 0.9240952134132385, + "learning_rate": 2.7452701264164092e-05, + "loss": 0.2207, + "step": 44134 + }, + { + "epoch": 3.062591076261189, + "grad_norm": 0.8613308072090149, + "learning_rate": 2.7448834362980357e-05, + "loss": 0.2588, + "step": 44135 + }, + { + "epoch": 3.062660467698286, + "grad_norm": 0.846682071685791, + "learning_rate": 2.7444967690832234e-05, + "loss": 0.2482, + "step": 44136 + }, + { + "epoch": 3.062729859135383, + "grad_norm": 0.8317235708236694, + "learning_rate": 2.7441101247731926e-05, + "loss": 0.2334, + "step": 44137 + }, + { + "epoch": 3.0627992505724793, + "grad_norm": 1.249240756034851, + "learning_rate": 2.7437235033691645e-05, + "loss": 0.332, + "step": 44138 + }, + { + "epoch": 3.062868642009576, + "grad_norm": 0.7487064003944397, + "learning_rate": 2.743336904872362e-05, + "loss": 0.2242, + "step": 44139 + }, + { + "epoch": 3.0629380334466725, + "grad_norm": 1.1582658290863037, + "learning_rate": 2.742950329284e-05, + "loss": 0.3611, + "step": 44140 + }, + { + "epoch": 3.0630074248837693, + "grad_norm": 1.023162841796875, + "learning_rate": 2.7425637766053025e-05, + "loss": 0.2381, + "step": 44141 + }, + { + "epoch": 3.063076816320866, + "grad_norm": 1.1028053760528564, + "learning_rate": 2.7421772468374883e-05, + "loss": 0.2927, + "step": 44142 + }, + { + "epoch": 3.0631462077579625, + "grad_norm": 1.16311776638031, + "learning_rate": 2.741790739981779e-05, + "loss": 0.2578, + "step": 44143 + }, + { + "epoch": 3.0632155991950594, + "grad_norm": 0.791324257850647, + "learning_rate": 2.741404256039394e-05, + "loss": 0.1368, + "step": 44144 + }, + { + "epoch": 3.063284990632156, + "grad_norm": 0.7537769079208374, + "learning_rate": 2.7410177950115533e-05, + "loss": 0.2097, + "step": 44145 + }, + { + "epoch": 3.0633543820692526, + "grad_norm": 0.6698954701423645, + "learning_rate": 2.7406313568994814e-05, + "loss": 0.3111, + "step": 44146 + }, + { + "epoch": 3.0634237735063494, + "grad_norm": 1.1541935205459595, + "learning_rate": 2.7402449417043906e-05, + "loss": 0.3401, + "step": 44147 + }, + { + "epoch": 3.063493164943446, + "grad_norm": 0.7444598078727722, + "learning_rate": 2.7398585494275032e-05, + "loss": 0.231, + "step": 44148 + }, + { + "epoch": 3.0635625563805426, + "grad_norm": 1.5288236141204834, + "learning_rate": 2.7394721800700406e-05, + "loss": 0.2733, + "step": 44149 + }, + { + "epoch": 3.0636319478176395, + "grad_norm": 0.5971788167953491, + "learning_rate": 2.7390858336332227e-05, + "loss": 0.1665, + "step": 44150 + }, + { + "epoch": 3.063701339254736, + "grad_norm": 0.8732200264930725, + "learning_rate": 2.7386995101182677e-05, + "loss": 0.3539, + "step": 44151 + }, + { + "epoch": 3.0637707306918327, + "grad_norm": 0.9233349561691284, + "learning_rate": 2.7383132095263962e-05, + "loss": 0.2596, + "step": 44152 + }, + { + "epoch": 3.063840122128929, + "grad_norm": 0.8857583999633789, + "learning_rate": 2.7379269318588298e-05, + "loss": 0.2137, + "step": 44153 + }, + { + "epoch": 3.063909513566026, + "grad_norm": 0.5875202417373657, + "learning_rate": 2.7375406771167833e-05, + "loss": 0.1534, + "step": 44154 + }, + { + "epoch": 3.0639789050031228, + "grad_norm": 0.8731650710105896, + "learning_rate": 2.7371544453014776e-05, + "loss": 0.2974, + "step": 44155 + }, + { + "epoch": 3.064048296440219, + "grad_norm": 0.9549754858016968, + "learning_rate": 2.7367682364141334e-05, + "loss": 0.331, + "step": 44156 + }, + { + "epoch": 3.064117687877316, + "grad_norm": 1.2126405239105225, + "learning_rate": 2.736382050455971e-05, + "loss": 0.4032, + "step": 44157 + }, + { + "epoch": 3.064187079314413, + "grad_norm": 0.8102703094482422, + "learning_rate": 2.7359958874282056e-05, + "loss": 0.2257, + "step": 44158 + }, + { + "epoch": 3.064256470751509, + "grad_norm": 0.7700521349906921, + "learning_rate": 2.7356097473320587e-05, + "loss": 0.1731, + "step": 44159 + }, + { + "epoch": 3.064325862188606, + "grad_norm": 1.0679285526275635, + "learning_rate": 2.7352236301687485e-05, + "loss": 0.3311, + "step": 44160 + }, + { + "epoch": 3.0643952536257024, + "grad_norm": 0.7125512957572937, + "learning_rate": 2.7348375359394974e-05, + "loss": 0.263, + "step": 44161 + }, + { + "epoch": 3.0644646450627993, + "grad_norm": 0.9142247438430786, + "learning_rate": 2.7344514646455187e-05, + "loss": 0.3252, + "step": 44162 + }, + { + "epoch": 3.064534036499896, + "grad_norm": 0.9526596665382385, + "learning_rate": 2.734065416288033e-05, + "loss": 0.2727, + "step": 44163 + }, + { + "epoch": 3.0646034279369925, + "grad_norm": 1.1557210683822632, + "learning_rate": 2.73367939086826e-05, + "loss": 0.2189, + "step": 44164 + }, + { + "epoch": 3.0646728193740893, + "grad_norm": 0.788215696811676, + "learning_rate": 2.733293388387419e-05, + "loss": 0.1825, + "step": 44165 + }, + { + "epoch": 3.0647422108111857, + "grad_norm": 0.7319158315658569, + "learning_rate": 2.7329074088467266e-05, + "loss": 0.1838, + "step": 44166 + }, + { + "epoch": 3.0648116022482825, + "grad_norm": 0.9451262354850769, + "learning_rate": 2.7325214522474065e-05, + "loss": 0.2071, + "step": 44167 + }, + { + "epoch": 3.0648809936853794, + "grad_norm": 0.8278728127479553, + "learning_rate": 2.7321355185906696e-05, + "loss": 0.2254, + "step": 44168 + }, + { + "epoch": 3.0649503851224758, + "grad_norm": 0.6197590231895447, + "learning_rate": 2.731749607877736e-05, + "loss": 0.1677, + "step": 44169 + }, + { + "epoch": 3.0650197765595726, + "grad_norm": 0.8281680941581726, + "learning_rate": 2.7313637201098273e-05, + "loss": 0.1729, + "step": 44170 + }, + { + "epoch": 3.0650891679966694, + "grad_norm": 0.8191649913787842, + "learning_rate": 2.7309778552881604e-05, + "loss": 0.2651, + "step": 44171 + }, + { + "epoch": 3.065158559433766, + "grad_norm": 0.8657371997833252, + "learning_rate": 2.730592013413952e-05, + "loss": 0.2472, + "step": 44172 + }, + { + "epoch": 3.0652279508708626, + "grad_norm": 0.8872888088226318, + "learning_rate": 2.7302061944884217e-05, + "loss": 0.297, + "step": 44173 + }, + { + "epoch": 3.065297342307959, + "grad_norm": 0.812847375869751, + "learning_rate": 2.729820398512789e-05, + "loss": 0.1886, + "step": 44174 + }, + { + "epoch": 3.065366733745056, + "grad_norm": 1.017757534980774, + "learning_rate": 2.7294346254882686e-05, + "loss": 0.2802, + "step": 44175 + }, + { + "epoch": 3.0654361251821527, + "grad_norm": 0.7992355823516846, + "learning_rate": 2.729048875416078e-05, + "loss": 0.23, + "step": 44176 + }, + { + "epoch": 3.065505516619249, + "grad_norm": 0.6911103129386902, + "learning_rate": 2.7286631482974367e-05, + "loss": 0.1749, + "step": 44177 + }, + { + "epoch": 3.065574908056346, + "grad_norm": 1.004167079925537, + "learning_rate": 2.7282774441335636e-05, + "loss": 0.3198, + "step": 44178 + }, + { + "epoch": 3.0656442994934423, + "grad_norm": 0.9990859627723694, + "learning_rate": 2.7278917629256728e-05, + "loss": 0.2087, + "step": 44179 + }, + { + "epoch": 3.065713690930539, + "grad_norm": 0.8721263408660889, + "learning_rate": 2.727506104674985e-05, + "loss": 0.3227, + "step": 44180 + }, + { + "epoch": 3.065783082367636, + "grad_norm": 1.0498685836791992, + "learning_rate": 2.7271204693827203e-05, + "loss": 0.2799, + "step": 44181 + }, + { + "epoch": 3.0658524738047324, + "grad_norm": 1.1494957208633423, + "learning_rate": 2.7267348570500895e-05, + "loss": 0.3559, + "step": 44182 + }, + { + "epoch": 3.065921865241829, + "grad_norm": 1.1036107540130615, + "learning_rate": 2.726349267678312e-05, + "loss": 0.2843, + "step": 44183 + }, + { + "epoch": 3.065991256678926, + "grad_norm": 1.0993787050247192, + "learning_rate": 2.7259637012686057e-05, + "loss": 0.2831, + "step": 44184 + }, + { + "epoch": 3.0660606481160224, + "grad_norm": 0.7176108956336975, + "learning_rate": 2.7255781578221894e-05, + "loss": 0.2573, + "step": 44185 + }, + { + "epoch": 3.0661300395531192, + "grad_norm": 1.4067455530166626, + "learning_rate": 2.725192637340278e-05, + "loss": 0.3619, + "step": 44186 + }, + { + "epoch": 3.0661994309902156, + "grad_norm": 1.0250577926635742, + "learning_rate": 2.7248071398240893e-05, + "loss": 0.307, + "step": 44187 + }, + { + "epoch": 3.0662688224273125, + "grad_norm": 0.7688761353492737, + "learning_rate": 2.7244216652748434e-05, + "loss": 0.2064, + "step": 44188 + }, + { + "epoch": 3.0663382138644093, + "grad_norm": 0.8583618998527527, + "learning_rate": 2.7240362136937513e-05, + "loss": 0.2565, + "step": 44189 + }, + { + "epoch": 3.0664076053015057, + "grad_norm": 1.0307294130325317, + "learning_rate": 2.7236507850820325e-05, + "loss": 0.3004, + "step": 44190 + }, + { + "epoch": 3.0664769967386025, + "grad_norm": 0.8009227514266968, + "learning_rate": 2.7232653794409046e-05, + "loss": 0.2917, + "step": 44191 + }, + { + "epoch": 3.066546388175699, + "grad_norm": 0.8642564415931702, + "learning_rate": 2.7228799967715867e-05, + "loss": 0.2137, + "step": 44192 + }, + { + "epoch": 3.0666157796127957, + "grad_norm": 1.5030893087387085, + "learning_rate": 2.722494637075289e-05, + "loss": 0.3061, + "step": 44193 + }, + { + "epoch": 3.0666851710498926, + "grad_norm": 1.3383362293243408, + "learning_rate": 2.7221093003532304e-05, + "loss": 0.3784, + "step": 44194 + }, + { + "epoch": 3.066754562486989, + "grad_norm": 0.7909834384918213, + "learning_rate": 2.72172398660663e-05, + "loss": 0.131, + "step": 44195 + }, + { + "epoch": 3.066823953924086, + "grad_norm": 1.2268593311309814, + "learning_rate": 2.7213386958367047e-05, + "loss": 0.1985, + "step": 44196 + }, + { + "epoch": 3.0668933453611826, + "grad_norm": 0.7961472868919373, + "learning_rate": 2.7209534280446657e-05, + "loss": 0.1725, + "step": 44197 + }, + { + "epoch": 3.066962736798279, + "grad_norm": 1.0465586185455322, + "learning_rate": 2.720568183231732e-05, + "loss": 0.3494, + "step": 44198 + }, + { + "epoch": 3.067032128235376, + "grad_norm": 0.9561812877655029, + "learning_rate": 2.7201829613991202e-05, + "loss": 0.271, + "step": 44199 + }, + { + "epoch": 3.0671015196724722, + "grad_norm": 1.0399665832519531, + "learning_rate": 2.7197977625480465e-05, + "loss": 0.228, + "step": 44200 + }, + { + "epoch": 3.067170911109569, + "grad_norm": 1.1020680665969849, + "learning_rate": 2.719412586679726e-05, + "loss": 0.248, + "step": 44201 + }, + { + "epoch": 3.067240302546666, + "grad_norm": 0.8375552892684937, + "learning_rate": 2.719027433795379e-05, + "loss": 0.353, + "step": 44202 + }, + { + "epoch": 3.0673096939837623, + "grad_norm": 0.828569233417511, + "learning_rate": 2.7186423038962137e-05, + "loss": 0.2588, + "step": 44203 + }, + { + "epoch": 3.067379085420859, + "grad_norm": 0.830193817615509, + "learning_rate": 2.7182571969834514e-05, + "loss": 0.1805, + "step": 44204 + }, + { + "epoch": 3.0674484768579555, + "grad_norm": 0.8074812293052673, + "learning_rate": 2.717872113058305e-05, + "loss": 0.2152, + "step": 44205 + }, + { + "epoch": 3.0675178682950524, + "grad_norm": 0.8799959421157837, + "learning_rate": 2.717487052121992e-05, + "loss": 0.2662, + "step": 44206 + }, + { + "epoch": 3.067587259732149, + "grad_norm": 1.1850731372833252, + "learning_rate": 2.7171020141757265e-05, + "loss": 0.2133, + "step": 44207 + }, + { + "epoch": 3.0676566511692456, + "grad_norm": 1.0188790559768677, + "learning_rate": 2.7167169992207288e-05, + "loss": 0.3925, + "step": 44208 + }, + { + "epoch": 3.0677260426063424, + "grad_norm": 0.9372043609619141, + "learning_rate": 2.7163320072582076e-05, + "loss": 0.1868, + "step": 44209 + }, + { + "epoch": 3.0677954340434392, + "grad_norm": 0.9245284199714661, + "learning_rate": 2.7159470382893805e-05, + "loss": 0.3244, + "step": 44210 + }, + { + "epoch": 3.0678648254805356, + "grad_norm": 0.6474133133888245, + "learning_rate": 2.7155620923154646e-05, + "loss": 0.2216, + "step": 44211 + }, + { + "epoch": 3.0679342169176325, + "grad_norm": 0.9402358531951904, + "learning_rate": 2.7151771693376726e-05, + "loss": 0.2942, + "step": 44212 + }, + { + "epoch": 3.068003608354729, + "grad_norm": 0.9355787038803101, + "learning_rate": 2.7147922693572226e-05, + "loss": 0.2382, + "step": 44213 + }, + { + "epoch": 3.0680729997918257, + "grad_norm": 0.595196008682251, + "learning_rate": 2.7144073923753267e-05, + "loss": 0.1796, + "step": 44214 + }, + { + "epoch": 3.0681423912289225, + "grad_norm": 0.9118601083755493, + "learning_rate": 2.714022538393205e-05, + "loss": 0.1735, + "step": 44215 + }, + { + "epoch": 3.068211782666019, + "grad_norm": 0.8642613887786865, + "learning_rate": 2.7136377074120665e-05, + "loss": 0.3, + "step": 44216 + }, + { + "epoch": 3.0682811741031157, + "grad_norm": 0.889793336391449, + "learning_rate": 2.7132528994331274e-05, + "loss": 0.2853, + "step": 44217 + }, + { + "epoch": 3.068350565540212, + "grad_norm": 0.9454255104064941, + "learning_rate": 2.712868114457604e-05, + "loss": 0.2414, + "step": 44218 + }, + { + "epoch": 3.068419956977309, + "grad_norm": 0.8321868777275085, + "learning_rate": 2.7124833524867098e-05, + "loss": 0.1935, + "step": 44219 + }, + { + "epoch": 3.068489348414406, + "grad_norm": 0.836368978023529, + "learning_rate": 2.7120986135216608e-05, + "loss": 0.3386, + "step": 44220 + }, + { + "epoch": 3.068558739851502, + "grad_norm": 0.8564911484718323, + "learning_rate": 2.711713897563671e-05, + "loss": 0.2308, + "step": 44221 + }, + { + "epoch": 3.068628131288599, + "grad_norm": 1.3152968883514404, + "learning_rate": 2.7113292046139584e-05, + "loss": 0.2846, + "step": 44222 + }, + { + "epoch": 3.068697522725696, + "grad_norm": 1.2877380847930908, + "learning_rate": 2.7109445346737296e-05, + "loss": 0.254, + "step": 44223 + }, + { + "epoch": 3.0687669141627922, + "grad_norm": 1.0733122825622559, + "learning_rate": 2.710559887744204e-05, + "loss": 0.2862, + "step": 44224 + }, + { + "epoch": 3.068836305599889, + "grad_norm": 0.7095326781272888, + "learning_rate": 2.7101752638265944e-05, + "loss": 0.2572, + "step": 44225 + }, + { + "epoch": 3.0689056970369855, + "grad_norm": 0.8551134467124939, + "learning_rate": 2.709790662922116e-05, + "loss": 0.2988, + "step": 44226 + }, + { + "epoch": 3.0689750884740823, + "grad_norm": 0.9180625081062317, + "learning_rate": 2.7094060850319858e-05, + "loss": 0.4799, + "step": 44227 + }, + { + "epoch": 3.069044479911179, + "grad_norm": 1.5369319915771484, + "learning_rate": 2.709021530157412e-05, + "loss": 0.2907, + "step": 44228 + }, + { + "epoch": 3.0691138713482755, + "grad_norm": 0.9619260430335999, + "learning_rate": 2.708636998299611e-05, + "loss": 0.1989, + "step": 44229 + }, + { + "epoch": 3.0691832627853723, + "grad_norm": 1.0497785806655884, + "learning_rate": 2.708252489459797e-05, + "loss": 0.1947, + "step": 44230 + }, + { + "epoch": 3.0692526542224687, + "grad_norm": 0.7011074423789978, + "learning_rate": 2.7078680036391867e-05, + "loss": 0.2493, + "step": 44231 + }, + { + "epoch": 3.0693220456595656, + "grad_norm": 0.6903491020202637, + "learning_rate": 2.7074835408389888e-05, + "loss": 0.2209, + "step": 44232 + }, + { + "epoch": 3.0693914370966624, + "grad_norm": 0.9508424997329712, + "learning_rate": 2.7070991010604186e-05, + "loss": 0.3024, + "step": 44233 + }, + { + "epoch": 3.069460828533759, + "grad_norm": 0.9287758469581604, + "learning_rate": 2.7067146843046908e-05, + "loss": 0.2417, + "step": 44234 + }, + { + "epoch": 3.0695302199708556, + "grad_norm": 0.787177562713623, + "learning_rate": 2.706330290573018e-05, + "loss": 0.2546, + "step": 44235 + }, + { + "epoch": 3.0695996114079525, + "grad_norm": 0.874794065952301, + "learning_rate": 2.7059459198666182e-05, + "loss": 0.2051, + "step": 44236 + }, + { + "epoch": 3.069669002845049, + "grad_norm": 1.0156455039978027, + "learning_rate": 2.7055615721866967e-05, + "loss": 0.1812, + "step": 44237 + }, + { + "epoch": 3.0697383942821457, + "grad_norm": 0.7839825749397278, + "learning_rate": 2.7051772475344716e-05, + "loss": 0.225, + "step": 44238 + }, + { + "epoch": 3.069807785719242, + "grad_norm": 0.819709837436676, + "learning_rate": 2.7047929459111564e-05, + "loss": 0.2082, + "step": 44239 + }, + { + "epoch": 3.069877177156339, + "grad_norm": 0.8662432432174683, + "learning_rate": 2.7044086673179624e-05, + "loss": 0.2052, + "step": 44240 + }, + { + "epoch": 3.0699465685934357, + "grad_norm": 0.7545931935310364, + "learning_rate": 2.704024411756104e-05, + "loss": 0.1726, + "step": 44241 + }, + { + "epoch": 3.070015960030532, + "grad_norm": 0.8243928551673889, + "learning_rate": 2.703640179226794e-05, + "loss": 0.2039, + "step": 44242 + }, + { + "epoch": 3.070085351467629, + "grad_norm": 0.7958853840827942, + "learning_rate": 2.7032559697312488e-05, + "loss": 0.2083, + "step": 44243 + }, + { + "epoch": 3.0701547429047253, + "grad_norm": 1.4993606805801392, + "learning_rate": 2.702871783270674e-05, + "loss": 0.4165, + "step": 44244 + }, + { + "epoch": 3.070224134341822, + "grad_norm": 0.7991405725479126, + "learning_rate": 2.702487619846288e-05, + "loss": 0.1871, + "step": 44245 + }, + { + "epoch": 3.070293525778919, + "grad_norm": 0.739732563495636, + "learning_rate": 2.7021034794593003e-05, + "loss": 0.2599, + "step": 44246 + }, + { + "epoch": 3.0703629172160154, + "grad_norm": 0.7967778444290161, + "learning_rate": 2.701719362110926e-05, + "loss": 0.1858, + "step": 44247 + }, + { + "epoch": 3.0704323086531122, + "grad_norm": 0.8026668429374695, + "learning_rate": 2.7013352678023762e-05, + "loss": 0.1703, + "step": 44248 + }, + { + "epoch": 3.070501700090209, + "grad_norm": 1.141664743423462, + "learning_rate": 2.7009511965348654e-05, + "loss": 0.4616, + "step": 44249 + }, + { + "epoch": 3.0705710915273055, + "grad_norm": 1.0048538446426392, + "learning_rate": 2.7005671483096072e-05, + "loss": 0.2714, + "step": 44250 + }, + { + "epoch": 3.0706404829644023, + "grad_norm": 0.9737595915794373, + "learning_rate": 2.7001831231278085e-05, + "loss": 0.28, + "step": 44251 + }, + { + "epoch": 3.0707098744014987, + "grad_norm": 0.8811574578285217, + "learning_rate": 2.6997991209906846e-05, + "loss": 0.2722, + "step": 44252 + }, + { + "epoch": 3.0707792658385955, + "grad_norm": 0.7804234623908997, + "learning_rate": 2.6994151418994486e-05, + "loss": 0.2069, + "step": 44253 + }, + { + "epoch": 3.0708486572756923, + "grad_norm": 1.0581692457199097, + "learning_rate": 2.699031185855311e-05, + "loss": 0.2459, + "step": 44254 + }, + { + "epoch": 3.0709180487127887, + "grad_norm": 0.7260430455207825, + "learning_rate": 2.698647252859485e-05, + "loss": 0.2143, + "step": 44255 + }, + { + "epoch": 3.0709874401498856, + "grad_norm": 0.947384774684906, + "learning_rate": 2.6982633429131833e-05, + "loss": 0.2825, + "step": 44256 + }, + { + "epoch": 3.071056831586982, + "grad_norm": 1.0526561737060547, + "learning_rate": 2.697879456017619e-05, + "loss": 0.3616, + "step": 44257 + }, + { + "epoch": 3.071126223024079, + "grad_norm": 0.7073780298233032, + "learning_rate": 2.6974955921739998e-05, + "loss": 0.2648, + "step": 44258 + }, + { + "epoch": 3.0711956144611756, + "grad_norm": 0.8097119927406311, + "learning_rate": 2.6971117513835397e-05, + "loss": 0.1972, + "step": 44259 + }, + { + "epoch": 3.071265005898272, + "grad_norm": 0.7232989072799683, + "learning_rate": 2.6967279336474505e-05, + "loss": 0.2456, + "step": 44260 + }, + { + "epoch": 3.071334397335369, + "grad_norm": 0.8446601033210754, + "learning_rate": 2.6963441389669442e-05, + "loss": 0.2459, + "step": 44261 + }, + { + "epoch": 3.0714037887724657, + "grad_norm": 0.880179226398468, + "learning_rate": 2.6959603673432342e-05, + "loss": 0.3123, + "step": 44262 + }, + { + "epoch": 3.071473180209562, + "grad_norm": 0.6576330661773682, + "learning_rate": 2.695576618777528e-05, + "loss": 0.2171, + "step": 44263 + }, + { + "epoch": 3.071542571646659, + "grad_norm": 0.9724107980728149, + "learning_rate": 2.695192893271038e-05, + "loss": 0.1587, + "step": 44264 + }, + { + "epoch": 3.0716119630837553, + "grad_norm": 0.8859158754348755, + "learning_rate": 2.6948091908249762e-05, + "loss": 0.2263, + "step": 44265 + }, + { + "epoch": 3.071681354520852, + "grad_norm": 1.0125230550765991, + "learning_rate": 2.694425511440558e-05, + "loss": 0.1747, + "step": 44266 + }, + { + "epoch": 3.071750745957949, + "grad_norm": 1.5666157007217407, + "learning_rate": 2.694041855118988e-05, + "loss": 0.3894, + "step": 44267 + }, + { + "epoch": 3.0718201373950453, + "grad_norm": 0.7681808471679688, + "learning_rate": 2.6936582218614802e-05, + "loss": 0.2328, + "step": 44268 + }, + { + "epoch": 3.071889528832142, + "grad_norm": 1.0470527410507202, + "learning_rate": 2.6932746116692452e-05, + "loss": 0.2671, + "step": 44269 + }, + { + "epoch": 3.071958920269239, + "grad_norm": 1.1024298667907715, + "learning_rate": 2.6928910245434945e-05, + "loss": 0.3049, + "step": 44270 + }, + { + "epoch": 3.0720283117063354, + "grad_norm": 1.398269534111023, + "learning_rate": 2.6925074604854426e-05, + "loss": 0.3025, + "step": 44271 + }, + { + "epoch": 3.0720977031434322, + "grad_norm": 1.249385118484497, + "learning_rate": 2.692123919496293e-05, + "loss": 0.1781, + "step": 44272 + }, + { + "epoch": 3.0721670945805286, + "grad_norm": 0.8102847933769226, + "learning_rate": 2.691740401577262e-05, + "loss": 0.255, + "step": 44273 + }, + { + "epoch": 3.0722364860176254, + "grad_norm": 0.5742053985595703, + "learning_rate": 2.6913569067295563e-05, + "loss": 0.1824, + "step": 44274 + }, + { + "epoch": 3.0723058774547223, + "grad_norm": 0.7852613925933838, + "learning_rate": 2.69097343495439e-05, + "loss": 0.2141, + "step": 44275 + }, + { + "epoch": 3.0723752688918187, + "grad_norm": 0.8515074253082275, + "learning_rate": 2.6905899862529727e-05, + "loss": 0.2702, + "step": 44276 + }, + { + "epoch": 3.0724446603289155, + "grad_norm": 1.4702228307724, + "learning_rate": 2.690206560626515e-05, + "loss": 0.3111, + "step": 44277 + }, + { + "epoch": 3.072514051766012, + "grad_norm": 1.0885987281799316, + "learning_rate": 2.68982315807623e-05, + "loss": 0.2587, + "step": 44278 + }, + { + "epoch": 3.0725834432031087, + "grad_norm": 1.0355448722839355, + "learning_rate": 2.6894397786033222e-05, + "loss": 0.2321, + "step": 44279 + }, + { + "epoch": 3.0726528346402056, + "grad_norm": 0.9893555045127869, + "learning_rate": 2.689056422209004e-05, + "loss": 0.2689, + "step": 44280 + }, + { + "epoch": 3.072722226077302, + "grad_norm": 0.775684654712677, + "learning_rate": 2.6886730888944876e-05, + "loss": 0.2256, + "step": 44281 + }, + { + "epoch": 3.0727916175143988, + "grad_norm": 0.9327194690704346, + "learning_rate": 2.6882897786609806e-05, + "loss": 0.2134, + "step": 44282 + }, + { + "epoch": 3.072861008951495, + "grad_norm": 0.9326897859573364, + "learning_rate": 2.6879064915096954e-05, + "loss": 0.2422, + "step": 44283 + }, + { + "epoch": 3.072930400388592, + "grad_norm": 0.6618403792381287, + "learning_rate": 2.6875232274418438e-05, + "loss": 0.1956, + "step": 44284 + }, + { + "epoch": 3.072999791825689, + "grad_norm": 0.9349377751350403, + "learning_rate": 2.687139986458631e-05, + "loss": 0.2371, + "step": 44285 + }, + { + "epoch": 3.073069183262785, + "grad_norm": 0.7655128240585327, + "learning_rate": 2.6867567685612672e-05, + "loss": 0.1641, + "step": 44286 + }, + { + "epoch": 3.073138574699882, + "grad_norm": 0.80789715051651, + "learning_rate": 2.6863735737509643e-05, + "loss": 0.2701, + "step": 44287 + }, + { + "epoch": 3.073207966136979, + "grad_norm": 0.9188749194145203, + "learning_rate": 2.6859904020289317e-05, + "loss": 0.193, + "step": 44288 + }, + { + "epoch": 3.0732773575740753, + "grad_norm": 0.7276052832603455, + "learning_rate": 2.6856072533963795e-05, + "loss": 0.2168, + "step": 44289 + }, + { + "epoch": 3.073346749011172, + "grad_norm": 1.2555185556411743, + "learning_rate": 2.6852241278545154e-05, + "loss": 0.2169, + "step": 44290 + }, + { + "epoch": 3.0734161404482685, + "grad_norm": 0.8527147173881531, + "learning_rate": 2.6848410254045543e-05, + "loss": 0.1864, + "step": 44291 + }, + { + "epoch": 3.0734855318853653, + "grad_norm": 1.00533127784729, + "learning_rate": 2.684457946047697e-05, + "loss": 0.1871, + "step": 44292 + }, + { + "epoch": 3.073554923322462, + "grad_norm": 1.5276037454605103, + "learning_rate": 2.6840748897851577e-05, + "loss": 0.3758, + "step": 44293 + }, + { + "epoch": 3.0736243147595586, + "grad_norm": 0.8517042994499207, + "learning_rate": 2.6836918566181457e-05, + "loss": 0.2438, + "step": 44294 + }, + { + "epoch": 3.0736937061966554, + "grad_norm": 1.3917269706726074, + "learning_rate": 2.6833088465478685e-05, + "loss": 0.223, + "step": 44295 + }, + { + "epoch": 3.073763097633752, + "grad_norm": 0.8523831963539124, + "learning_rate": 2.682925859575538e-05, + "loss": 0.2346, + "step": 44296 + }, + { + "epoch": 3.0738324890708486, + "grad_norm": 1.127328634262085, + "learning_rate": 2.6825428957023634e-05, + "loss": 0.266, + "step": 44297 + }, + { + "epoch": 3.0739018805079454, + "grad_norm": 0.7187556624412537, + "learning_rate": 2.6821599549295494e-05, + "loss": 0.1498, + "step": 44298 + }, + { + "epoch": 3.073971271945042, + "grad_norm": 0.9604288935661316, + "learning_rate": 2.6817770372583072e-05, + "loss": 0.1837, + "step": 44299 + }, + { + "epoch": 3.0740406633821387, + "grad_norm": 1.3603134155273438, + "learning_rate": 2.6813941426898483e-05, + "loss": 0.2189, + "step": 44300 + }, + { + "epoch": 3.0741100548192355, + "grad_norm": 1.1733890771865845, + "learning_rate": 2.6810112712253765e-05, + "loss": 0.3417, + "step": 44301 + }, + { + "epoch": 3.074179446256332, + "grad_norm": 0.5603910088539124, + "learning_rate": 2.6806284228661027e-05, + "loss": 0.2371, + "step": 44302 + }, + { + "epoch": 3.0742488376934287, + "grad_norm": 1.308597207069397, + "learning_rate": 2.680245597613237e-05, + "loss": 0.3698, + "step": 44303 + }, + { + "epoch": 3.074318229130525, + "grad_norm": 0.5761896967887878, + "learning_rate": 2.6798627954679856e-05, + "loss": 0.1397, + "step": 44304 + }, + { + "epoch": 3.074387620567622, + "grad_norm": 1.0789716243743896, + "learning_rate": 2.6794800164315613e-05, + "loss": 0.357, + "step": 44305 + }, + { + "epoch": 3.0744570120047188, + "grad_norm": 0.8436456322669983, + "learning_rate": 2.6790972605051655e-05, + "loss": 0.2777, + "step": 44306 + }, + { + "epoch": 3.074526403441815, + "grad_norm": 1.1856305599212646, + "learning_rate": 2.6787145276900115e-05, + "loss": 0.3559, + "step": 44307 + }, + { + "epoch": 3.074595794878912, + "grad_norm": 0.8255226016044617, + "learning_rate": 2.678331817987305e-05, + "loss": 0.2826, + "step": 44308 + }, + { + "epoch": 3.0746651863160084, + "grad_norm": 0.879214346408844, + "learning_rate": 2.6779491313982563e-05, + "loss": 0.271, + "step": 44309 + }, + { + "epoch": 3.074734577753105, + "grad_norm": 0.8939313292503357, + "learning_rate": 2.677566467924073e-05, + "loss": 0.2827, + "step": 44310 + }, + { + "epoch": 3.074803969190202, + "grad_norm": 0.8423704504966736, + "learning_rate": 2.6771838275659623e-05, + "loss": 0.2102, + "step": 44311 + }, + { + "epoch": 3.0748733606272984, + "grad_norm": 0.667966365814209, + "learning_rate": 2.676801210325136e-05, + "loss": 0.3743, + "step": 44312 + }, + { + "epoch": 3.0749427520643953, + "grad_norm": 0.772891104221344, + "learning_rate": 2.676418616202795e-05, + "loss": 0.2137, + "step": 44313 + }, + { + "epoch": 3.075012143501492, + "grad_norm": 0.9560642838478088, + "learning_rate": 2.676036045200151e-05, + "loss": 0.2592, + "step": 44314 + }, + { + "epoch": 3.0750815349385885, + "grad_norm": 0.8799802660942078, + "learning_rate": 2.6756534973184122e-05, + "loss": 0.1693, + "step": 44315 + }, + { + "epoch": 3.0751509263756853, + "grad_norm": 1.0010830163955688, + "learning_rate": 2.6752709725587844e-05, + "loss": 0.2923, + "step": 44316 + }, + { + "epoch": 3.0752203178127817, + "grad_norm": 0.6669821739196777, + "learning_rate": 2.6748884709224763e-05, + "loss": 0.1377, + "step": 44317 + }, + { + "epoch": 3.0752897092498785, + "grad_norm": 1.2270638942718506, + "learning_rate": 2.6745059924106962e-05, + "loss": 0.2252, + "step": 44318 + }, + { + "epoch": 3.0753591006869754, + "grad_norm": 0.9922764897346497, + "learning_rate": 2.674123537024654e-05, + "loss": 0.317, + "step": 44319 + }, + { + "epoch": 3.0754284921240718, + "grad_norm": 0.8867892026901245, + "learning_rate": 2.6737411047655503e-05, + "loss": 0.1693, + "step": 44320 + }, + { + "epoch": 3.0754978835611686, + "grad_norm": 0.8157872557640076, + "learning_rate": 2.6733586956345957e-05, + "loss": 0.2541, + "step": 44321 + }, + { + "epoch": 3.0755672749982654, + "grad_norm": 1.3828011751174927, + "learning_rate": 2.6729763096329986e-05, + "loss": 0.3708, + "step": 44322 + }, + { + "epoch": 3.075636666435362, + "grad_norm": 0.6446294784545898, + "learning_rate": 2.6725939467619644e-05, + "loss": 0.1427, + "step": 44323 + }, + { + "epoch": 3.0757060578724587, + "grad_norm": 1.2491559982299805, + "learning_rate": 2.672211607022701e-05, + "loss": 0.132, + "step": 44324 + }, + { + "epoch": 3.075775449309555, + "grad_norm": 0.8592446446418762, + "learning_rate": 2.6718292904164153e-05, + "loss": 0.1808, + "step": 44325 + }, + { + "epoch": 3.075844840746652, + "grad_norm": 0.7882893085479736, + "learning_rate": 2.671446996944317e-05, + "loss": 0.2111, + "step": 44326 + }, + { + "epoch": 3.0759142321837487, + "grad_norm": 0.8945042490959167, + "learning_rate": 2.6710647266076085e-05, + "loss": 0.194, + "step": 44327 + }, + { + "epoch": 3.075983623620845, + "grad_norm": 0.862342357635498, + "learning_rate": 2.670682479407497e-05, + "loss": 0.2544, + "step": 44328 + }, + { + "epoch": 3.076053015057942, + "grad_norm": 0.8059945106506348, + "learning_rate": 2.6703002553451918e-05, + "loss": 0.2169, + "step": 44329 + }, + { + "epoch": 3.0761224064950383, + "grad_norm": 0.9564675092697144, + "learning_rate": 2.6699180544218972e-05, + "loss": 0.3007, + "step": 44330 + }, + { + "epoch": 3.076191797932135, + "grad_norm": 0.8405334949493408, + "learning_rate": 2.669535876638821e-05, + "loss": 0.211, + "step": 44331 + }, + { + "epoch": 3.076261189369232, + "grad_norm": 1.0292681455612183, + "learning_rate": 2.669153721997173e-05, + "loss": 0.1767, + "step": 44332 + }, + { + "epoch": 3.0763305808063284, + "grad_norm": 0.7743035554885864, + "learning_rate": 2.6687715904981524e-05, + "loss": 0.2829, + "step": 44333 + }, + { + "epoch": 3.076399972243425, + "grad_norm": 0.7745179533958435, + "learning_rate": 2.66838948214297e-05, + "loss": 0.1576, + "step": 44334 + }, + { + "epoch": 3.0764693636805216, + "grad_norm": 1.1119252443313599, + "learning_rate": 2.6680073969328345e-05, + "loss": 0.357, + "step": 44335 + }, + { + "epoch": 3.0765387551176184, + "grad_norm": 0.7900203466415405, + "learning_rate": 2.6676253348689463e-05, + "loss": 0.231, + "step": 44336 + }, + { + "epoch": 3.0766081465547153, + "grad_norm": 0.8343676328659058, + "learning_rate": 2.667243295952513e-05, + "loss": 0.2176, + "step": 44337 + }, + { + "epoch": 3.0766775379918117, + "grad_norm": 0.9246475696563721, + "learning_rate": 2.6668612801847426e-05, + "loss": 0.2874, + "step": 44338 + }, + { + "epoch": 3.0767469294289085, + "grad_norm": 0.8964622616767883, + "learning_rate": 2.666479287566841e-05, + "loss": 0.2906, + "step": 44339 + }, + { + "epoch": 3.0768163208660053, + "grad_norm": 0.9939713478088379, + "learning_rate": 2.6660973181000147e-05, + "loss": 0.2823, + "step": 44340 + }, + { + "epoch": 3.0768857123031017, + "grad_norm": 1.0184508562088013, + "learning_rate": 2.6657153717854667e-05, + "loss": 0.376, + "step": 44341 + }, + { + "epoch": 3.0769551037401985, + "grad_norm": 0.6808511018753052, + "learning_rate": 2.665333448624404e-05, + "loss": 0.2065, + "step": 44342 + }, + { + "epoch": 3.077024495177295, + "grad_norm": 1.298599362373352, + "learning_rate": 2.664951548618032e-05, + "loss": 0.3571, + "step": 44343 + }, + { + "epoch": 3.0770938866143918, + "grad_norm": 1.0025646686553955, + "learning_rate": 2.6645696717675584e-05, + "loss": 0.2153, + "step": 44344 + }, + { + "epoch": 3.0771632780514886, + "grad_norm": 0.7376207709312439, + "learning_rate": 2.6641878180741865e-05, + "loss": 0.1751, + "step": 44345 + }, + { + "epoch": 3.077232669488585, + "grad_norm": 1.286210536956787, + "learning_rate": 2.663805987539122e-05, + "loss": 0.2258, + "step": 44346 + }, + { + "epoch": 3.077302060925682, + "grad_norm": 1.1746909618377686, + "learning_rate": 2.663424180163574e-05, + "loss": 0.2432, + "step": 44347 + }, + { + "epoch": 3.0773714523627786, + "grad_norm": 0.8088943362236023, + "learning_rate": 2.6630423959487418e-05, + "loss": 0.2914, + "step": 44348 + }, + { + "epoch": 3.077440843799875, + "grad_norm": 0.7506882548332214, + "learning_rate": 2.662660634895834e-05, + "loss": 0.2804, + "step": 44349 + }, + { + "epoch": 3.077510235236972, + "grad_norm": 1.0626699924468994, + "learning_rate": 2.6622788970060552e-05, + "loss": 0.2124, + "step": 44350 + }, + { + "epoch": 3.0775796266740683, + "grad_norm": 0.9241155385971069, + "learning_rate": 2.6618971822806095e-05, + "loss": 0.276, + "step": 44351 + }, + { + "epoch": 3.077649018111165, + "grad_norm": 1.2035558223724365, + "learning_rate": 2.661515490720704e-05, + "loss": 0.2701, + "step": 44352 + }, + { + "epoch": 3.077718409548262, + "grad_norm": 1.134960412979126, + "learning_rate": 2.661133822327543e-05, + "loss": 0.2044, + "step": 44353 + }, + { + "epoch": 3.0777878009853583, + "grad_norm": 1.4907346963882446, + "learning_rate": 2.6607521771023338e-05, + "loss": 0.4579, + "step": 44354 + }, + { + "epoch": 3.077857192422455, + "grad_norm": 0.8560102581977844, + "learning_rate": 2.6603705550462743e-05, + "loss": 0.2533, + "step": 44355 + }, + { + "epoch": 3.0779265838595515, + "grad_norm": 1.37688148021698, + "learning_rate": 2.659988956160575e-05, + "loss": 0.2322, + "step": 44356 + }, + { + "epoch": 3.0779959752966484, + "grad_norm": 0.9627751111984253, + "learning_rate": 2.6596073804464382e-05, + "loss": 0.2247, + "step": 44357 + }, + { + "epoch": 3.078065366733745, + "grad_norm": 0.9874812364578247, + "learning_rate": 2.6592258279050686e-05, + "loss": 0.3401, + "step": 44358 + }, + { + "epoch": 3.0781347581708416, + "grad_norm": 0.8276384472846985, + "learning_rate": 2.6588442985376727e-05, + "loss": 0.278, + "step": 44359 + }, + { + "epoch": 3.0782041496079384, + "grad_norm": 0.6611929535865784, + "learning_rate": 2.6584627923454564e-05, + "loss": 0.2527, + "step": 44360 + }, + { + "epoch": 3.0782735410450353, + "grad_norm": 1.153616189956665, + "learning_rate": 2.6580813093296177e-05, + "loss": 0.4491, + "step": 44361 + }, + { + "epoch": 3.0783429324821316, + "grad_norm": 1.678326964378357, + "learning_rate": 2.657699849491364e-05, + "loss": 0.2526, + "step": 44362 + }, + { + "epoch": 3.0784123239192285, + "grad_norm": 0.8247469067573547, + "learning_rate": 2.6573184128319008e-05, + "loss": 0.205, + "step": 44363 + }, + { + "epoch": 3.078481715356325, + "grad_norm": 1.6360183954238892, + "learning_rate": 2.656936999352432e-05, + "loss": 0.2832, + "step": 44364 + }, + { + "epoch": 3.0785511067934217, + "grad_norm": 0.9025084376335144, + "learning_rate": 2.65655560905416e-05, + "loss": 0.2277, + "step": 44365 + }, + { + "epoch": 3.0786204982305185, + "grad_norm": 0.7016757726669312, + "learning_rate": 2.656174241938294e-05, + "loss": 0.1836, + "step": 44366 + }, + { + "epoch": 3.078689889667615, + "grad_norm": 0.8843620419502258, + "learning_rate": 2.6557928980060297e-05, + "loss": 0.3271, + "step": 44367 + }, + { + "epoch": 3.0787592811047118, + "grad_norm": 0.7041894197463989, + "learning_rate": 2.6554115772585763e-05, + "loss": 0.1782, + "step": 44368 + }, + { + "epoch": 3.078828672541808, + "grad_norm": 1.3082035779953003, + "learning_rate": 2.655030279697136e-05, + "loss": 0.2942, + "step": 44369 + }, + { + "epoch": 3.078898063978905, + "grad_norm": 0.9055866599082947, + "learning_rate": 2.6546490053229156e-05, + "loss": 0.313, + "step": 44370 + }, + { + "epoch": 3.078967455416002, + "grad_norm": 0.7525302171707153, + "learning_rate": 2.654267754137113e-05, + "loss": 0.1686, + "step": 44371 + }, + { + "epoch": 3.079036846853098, + "grad_norm": 1.294542670249939, + "learning_rate": 2.6538865261409352e-05, + "loss": 0.3434, + "step": 44372 + }, + { + "epoch": 3.079106238290195, + "grad_norm": 0.8864291310310364, + "learning_rate": 2.6535053213355854e-05, + "loss": 0.3139, + "step": 44373 + }, + { + "epoch": 3.079175629727292, + "grad_norm": 1.0425060987472534, + "learning_rate": 2.6531241397222694e-05, + "loss": 0.4261, + "step": 44374 + }, + { + "epoch": 3.0792450211643883, + "grad_norm": 1.0352565050125122, + "learning_rate": 2.6527429813021866e-05, + "loss": 0.2604, + "step": 44375 + }, + { + "epoch": 3.079314412601485, + "grad_norm": 1.065358281135559, + "learning_rate": 2.65236184607654e-05, + "loss": 0.2664, + "step": 44376 + }, + { + "epoch": 3.0793838040385815, + "grad_norm": 1.203244686126709, + "learning_rate": 2.6519807340465353e-05, + "loss": 0.3327, + "step": 44377 + }, + { + "epoch": 3.0794531954756783, + "grad_norm": 1.0706062316894531, + "learning_rate": 2.651599645213375e-05, + "loss": 0.3265, + "step": 44378 + }, + { + "epoch": 3.079522586912775, + "grad_norm": 0.749633252620697, + "learning_rate": 2.6512185795782618e-05, + "loss": 0.1979, + "step": 44379 + }, + { + "epoch": 3.0795919783498715, + "grad_norm": 0.7553061246871948, + "learning_rate": 2.6508375371423988e-05, + "loss": 0.2274, + "step": 44380 + }, + { + "epoch": 3.0796613697869684, + "grad_norm": 1.0541292428970337, + "learning_rate": 2.650456517906993e-05, + "loss": 0.2524, + "step": 44381 + }, + { + "epoch": 3.0797307612240648, + "grad_norm": 1.1370737552642822, + "learning_rate": 2.6500755218732387e-05, + "loss": 0.2533, + "step": 44382 + }, + { + "epoch": 3.0798001526611616, + "grad_norm": 0.5781093835830688, + "learning_rate": 2.649694549042344e-05, + "loss": 0.2281, + "step": 44383 + }, + { + "epoch": 3.0798695440982584, + "grad_norm": 1.0596035718917847, + "learning_rate": 2.6493135994155118e-05, + "loss": 0.2895, + "step": 44384 + }, + { + "epoch": 3.079938935535355, + "grad_norm": 0.6521660685539246, + "learning_rate": 2.6489326729939423e-05, + "loss": 0.2376, + "step": 44385 + }, + { + "epoch": 3.0800083269724516, + "grad_norm": 0.9535241723060608, + "learning_rate": 2.6485517697788397e-05, + "loss": 0.281, + "step": 44386 + }, + { + "epoch": 3.0800777184095485, + "grad_norm": 1.136468529701233, + "learning_rate": 2.6481708897714065e-05, + "loss": 0.2996, + "step": 44387 + }, + { + "epoch": 3.080147109846645, + "grad_norm": 1.3330987691879272, + "learning_rate": 2.647790032972848e-05, + "loss": 0.3178, + "step": 44388 + }, + { + "epoch": 3.0802165012837417, + "grad_norm": 0.8539707660675049, + "learning_rate": 2.6474091993843598e-05, + "loss": 0.2832, + "step": 44389 + }, + { + "epoch": 3.080285892720838, + "grad_norm": 0.8559499382972717, + "learning_rate": 2.647028389007148e-05, + "loss": 0.1431, + "step": 44390 + }, + { + "epoch": 3.080355284157935, + "grad_norm": 1.0793242454528809, + "learning_rate": 2.6466476018424148e-05, + "loss": 0.2867, + "step": 44391 + }, + { + "epoch": 3.0804246755950317, + "grad_norm": 1.0538711547851562, + "learning_rate": 2.646266837891361e-05, + "loss": 0.2511, + "step": 44392 + }, + { + "epoch": 3.080494067032128, + "grad_norm": 0.9224835634231567, + "learning_rate": 2.6458860971551912e-05, + "loss": 0.2497, + "step": 44393 + }, + { + "epoch": 3.080563458469225, + "grad_norm": 1.205180048942566, + "learning_rate": 2.6455053796351048e-05, + "loss": 0.3604, + "step": 44394 + }, + { + "epoch": 3.0806328499063214, + "grad_norm": 1.018383502960205, + "learning_rate": 2.6451246853323076e-05, + "loss": 0.3177, + "step": 44395 + }, + { + "epoch": 3.080702241343418, + "grad_norm": 0.998365044593811, + "learning_rate": 2.644744014247996e-05, + "loss": 0.2869, + "step": 44396 + }, + { + "epoch": 3.080771632780515, + "grad_norm": 0.9075584411621094, + "learning_rate": 2.6443633663833733e-05, + "loss": 0.2792, + "step": 44397 + }, + { + "epoch": 3.0808410242176114, + "grad_norm": 0.7193336486816406, + "learning_rate": 2.6439827417396424e-05, + "loss": 0.1684, + "step": 44398 + }, + { + "epoch": 3.0809104156547082, + "grad_norm": 0.7728385329246521, + "learning_rate": 2.6436021403180046e-05, + "loss": 0.2222, + "step": 44399 + }, + { + "epoch": 3.080979807091805, + "grad_norm": 1.0624315738677979, + "learning_rate": 2.643221562119662e-05, + "loss": 0.3072, + "step": 44400 + }, + { + "epoch": 3.0810491985289015, + "grad_norm": 0.9070367813110352, + "learning_rate": 2.6428410071458176e-05, + "loss": 0.298, + "step": 44401 + }, + { + "epoch": 3.0811185899659983, + "grad_norm": 0.8629488945007324, + "learning_rate": 2.642460475397668e-05, + "loss": 0.3393, + "step": 44402 + }, + { + "epoch": 3.0811879814030947, + "grad_norm": 0.8738399147987366, + "learning_rate": 2.6420799668764164e-05, + "loss": 0.3385, + "step": 44403 + }, + { + "epoch": 3.0812573728401915, + "grad_norm": 0.809628427028656, + "learning_rate": 2.6416994815832653e-05, + "loss": 0.2433, + "step": 44404 + }, + { + "epoch": 3.0813267642772884, + "grad_norm": 1.8306161165237427, + "learning_rate": 2.641319019519417e-05, + "loss": 0.3269, + "step": 44405 + }, + { + "epoch": 3.0813961557143847, + "grad_norm": 1.0603139400482178, + "learning_rate": 2.6409385806860686e-05, + "loss": 0.3708, + "step": 44406 + }, + { + "epoch": 3.0814655471514816, + "grad_norm": 1.1313680410385132, + "learning_rate": 2.6405581650844235e-05, + "loss": 0.2391, + "step": 44407 + }, + { + "epoch": 3.081534938588578, + "grad_norm": 0.7362122535705566, + "learning_rate": 2.640177772715683e-05, + "loss": 0.2958, + "step": 44408 + }, + { + "epoch": 3.081604330025675, + "grad_norm": 0.7936232686042786, + "learning_rate": 2.6397974035810492e-05, + "loss": 0.3, + "step": 44409 + }, + { + "epoch": 3.0816737214627716, + "grad_norm": 1.1932481527328491, + "learning_rate": 2.639417057681719e-05, + "loss": 0.2555, + "step": 44410 + }, + { + "epoch": 3.081743112899868, + "grad_norm": 0.7110249400138855, + "learning_rate": 2.639036735018895e-05, + "loss": 0.2316, + "step": 44411 + }, + { + "epoch": 3.081812504336965, + "grad_norm": 1.2485452890396118, + "learning_rate": 2.638656435593777e-05, + "loss": 0.3192, + "step": 44412 + }, + { + "epoch": 3.0818818957740617, + "grad_norm": 0.8386895060539246, + "learning_rate": 2.638276159407568e-05, + "loss": 0.2431, + "step": 44413 + }, + { + "epoch": 3.081951287211158, + "grad_norm": 1.0067763328552246, + "learning_rate": 2.6378959064614662e-05, + "loss": 0.2172, + "step": 44414 + }, + { + "epoch": 3.082020678648255, + "grad_norm": 0.7159746885299683, + "learning_rate": 2.6375156767566723e-05, + "loss": 0.3054, + "step": 44415 + }, + { + "epoch": 3.0820900700853513, + "grad_norm": 1.0185343027114868, + "learning_rate": 2.637135470294392e-05, + "loss": 0.315, + "step": 44416 + }, + { + "epoch": 3.082159461522448, + "grad_norm": 0.9803375005722046, + "learning_rate": 2.636755287075816e-05, + "loss": 0.3002, + "step": 44417 + }, + { + "epoch": 3.082228852959545, + "grad_norm": 1.5580511093139648, + "learning_rate": 2.6363751271021496e-05, + "loss": 0.3496, + "step": 44418 + }, + { + "epoch": 3.0822982443966414, + "grad_norm": 1.0517243146896362, + "learning_rate": 2.635994990374593e-05, + "loss": 0.3065, + "step": 44419 + }, + { + "epoch": 3.082367635833738, + "grad_norm": 1.083743929862976, + "learning_rate": 2.6356148768943456e-05, + "loss": 0.3408, + "step": 44420 + }, + { + "epoch": 3.0824370272708346, + "grad_norm": 1.2565964460372925, + "learning_rate": 2.635234786662608e-05, + "loss": 0.2988, + "step": 44421 + }, + { + "epoch": 3.0825064187079314, + "grad_norm": 0.8090313673019409, + "learning_rate": 2.6348547196805797e-05, + "loss": 0.2327, + "step": 44422 + }, + { + "epoch": 3.0825758101450282, + "grad_norm": 0.7548696994781494, + "learning_rate": 2.634474675949463e-05, + "loss": 0.1449, + "step": 44423 + }, + { + "epoch": 3.0826452015821246, + "grad_norm": 0.8306558728218079, + "learning_rate": 2.6340946554704526e-05, + "loss": 0.2373, + "step": 44424 + }, + { + "epoch": 3.0827145930192215, + "grad_norm": 0.8035799860954285, + "learning_rate": 2.6337146582447503e-05, + "loss": 0.2253, + "step": 44425 + }, + { + "epoch": 3.0827839844563183, + "grad_norm": 0.9863942861557007, + "learning_rate": 2.6333346842735573e-05, + "loss": 0.2575, + "step": 44426 + }, + { + "epoch": 3.0828533758934147, + "grad_norm": 0.6849063634872437, + "learning_rate": 2.6329547335580706e-05, + "loss": 0.3304, + "step": 44427 + }, + { + "epoch": 3.0829227673305115, + "grad_norm": 0.7705814242362976, + "learning_rate": 2.6325748060994916e-05, + "loss": 0.2614, + "step": 44428 + }, + { + "epoch": 3.082992158767608, + "grad_norm": 1.2393417358398438, + "learning_rate": 2.63219490189902e-05, + "loss": 0.4771, + "step": 44429 + }, + { + "epoch": 3.0830615502047047, + "grad_norm": 0.9623107314109802, + "learning_rate": 2.631815020957856e-05, + "loss": 0.3073, + "step": 44430 + }, + { + "epoch": 3.0831309416418016, + "grad_norm": 0.9171028733253479, + "learning_rate": 2.631435163277195e-05, + "loss": 0.381, + "step": 44431 + }, + { + "epoch": 3.083200333078898, + "grad_norm": 1.1603567600250244, + "learning_rate": 2.6310553288582373e-05, + "loss": 0.364, + "step": 44432 + }, + { + "epoch": 3.083269724515995, + "grad_norm": 0.6712163090705872, + "learning_rate": 2.6306755177021836e-05, + "loss": 0.1586, + "step": 44433 + }, + { + "epoch": 3.083339115953091, + "grad_norm": 0.9909200668334961, + "learning_rate": 2.6302957298102327e-05, + "loss": 0.219, + "step": 44434 + }, + { + "epoch": 3.083408507390188, + "grad_norm": 1.0982513427734375, + "learning_rate": 2.6299159651835826e-05, + "loss": 0.2865, + "step": 44435 + }, + { + "epoch": 3.083477898827285, + "grad_norm": 1.0202651023864746, + "learning_rate": 2.6295362238234354e-05, + "loss": 0.3201, + "step": 44436 + }, + { + "epoch": 3.0835472902643812, + "grad_norm": 0.8386155962944031, + "learning_rate": 2.629156505730985e-05, + "loss": 0.224, + "step": 44437 + }, + { + "epoch": 3.083616681701478, + "grad_norm": 1.010201096534729, + "learning_rate": 2.6287768109074318e-05, + "loss": 0.3463, + "step": 44438 + }, + { + "epoch": 3.083686073138575, + "grad_norm": 0.7098134756088257, + "learning_rate": 2.6283971393539752e-05, + "loss": 0.1855, + "step": 44439 + }, + { + "epoch": 3.0837554645756713, + "grad_norm": 0.6388649344444275, + "learning_rate": 2.6280174910718168e-05, + "loss": 0.1213, + "step": 44440 + }, + { + "epoch": 3.083824856012768, + "grad_norm": 0.8116192817687988, + "learning_rate": 2.627637866062148e-05, + "loss": 0.2715, + "step": 44441 + }, + { + "epoch": 3.0838942474498645, + "grad_norm": 0.6054348945617676, + "learning_rate": 2.6272582643261724e-05, + "loss": 0.2101, + "step": 44442 + }, + { + "epoch": 3.0839636388869613, + "grad_norm": 0.6245182752609253, + "learning_rate": 2.6268786858650895e-05, + "loss": 0.1332, + "step": 44443 + }, + { + "epoch": 3.084033030324058, + "grad_norm": 1.1587941646575928, + "learning_rate": 2.6264991306800913e-05, + "loss": 0.5604, + "step": 44444 + }, + { + "epoch": 3.0841024217611546, + "grad_norm": 1.0246180295944214, + "learning_rate": 2.6261195987723807e-05, + "loss": 0.2917, + "step": 44445 + }, + { + "epoch": 3.0841718131982514, + "grad_norm": 1.1348627805709839, + "learning_rate": 2.625740090143155e-05, + "loss": 0.2867, + "step": 44446 + }, + { + "epoch": 3.084241204635348, + "grad_norm": 0.9819961190223694, + "learning_rate": 2.625360604793612e-05, + "loss": 0.236, + "step": 44447 + }, + { + "epoch": 3.0843105960724446, + "grad_norm": 1.3945691585540771, + "learning_rate": 2.6249811427249504e-05, + "loss": 0.2178, + "step": 44448 + }, + { + "epoch": 3.0843799875095415, + "grad_norm": 1.1035799980163574, + "learning_rate": 2.6246017039383673e-05, + "loss": 0.2765, + "step": 44449 + }, + { + "epoch": 3.084449378946638, + "grad_norm": 0.852289617061615, + "learning_rate": 2.624222288435064e-05, + "loss": 0.2797, + "step": 44450 + }, + { + "epoch": 3.0845187703837347, + "grad_norm": 0.9213836193084717, + "learning_rate": 2.623842896216232e-05, + "loss": 0.4108, + "step": 44451 + }, + { + "epoch": 3.0845881618208315, + "grad_norm": 0.6768964529037476, + "learning_rate": 2.6234635272830722e-05, + "loss": 0.1929, + "step": 44452 + }, + { + "epoch": 3.084657553257928, + "grad_norm": 0.6478129625320435, + "learning_rate": 2.6230841816367825e-05, + "loss": 0.1462, + "step": 44453 + }, + { + "epoch": 3.0847269446950247, + "grad_norm": 1.3241021633148193, + "learning_rate": 2.622704859278561e-05, + "loss": 0.2523, + "step": 44454 + }, + { + "epoch": 3.084796336132121, + "grad_norm": 0.8613808751106262, + "learning_rate": 2.622325560209603e-05, + "loss": 0.2311, + "step": 44455 + }, + { + "epoch": 3.084865727569218, + "grad_norm": 0.9649606943130493, + "learning_rate": 2.6219462844311073e-05, + "loss": 0.3324, + "step": 44456 + }, + { + "epoch": 3.084935119006315, + "grad_norm": 1.0765398740768433, + "learning_rate": 2.6215670319442743e-05, + "loss": 0.2987, + "step": 44457 + }, + { + "epoch": 3.085004510443411, + "grad_norm": 0.860139012336731, + "learning_rate": 2.6211878027502955e-05, + "loss": 0.3358, + "step": 44458 + }, + { + "epoch": 3.085073901880508, + "grad_norm": 0.7899995446205139, + "learning_rate": 2.6208085968503702e-05, + "loss": 0.2102, + "step": 44459 + }, + { + "epoch": 3.0851432933176044, + "grad_norm": 0.9560293555259705, + "learning_rate": 2.6204294142456965e-05, + "loss": 0.2708, + "step": 44460 + }, + { + "epoch": 3.0852126847547012, + "grad_norm": 0.8833569288253784, + "learning_rate": 2.6200502549374707e-05, + "loss": 0.2211, + "step": 44461 + }, + { + "epoch": 3.085282076191798, + "grad_norm": 1.9672696590423584, + "learning_rate": 2.6196711189268895e-05, + "loss": 0.4179, + "step": 44462 + }, + { + "epoch": 3.0853514676288945, + "grad_norm": 0.7861941456794739, + "learning_rate": 2.6192920062151516e-05, + "loss": 0.1921, + "step": 44463 + }, + { + "epoch": 3.0854208590659913, + "grad_norm": 0.8120971322059631, + "learning_rate": 2.6189129168034544e-05, + "loss": 0.2064, + "step": 44464 + }, + { + "epoch": 3.085490250503088, + "grad_norm": 0.6598656177520752, + "learning_rate": 2.6185338506929902e-05, + "loss": 0.2369, + "step": 44465 + }, + { + "epoch": 3.0855596419401845, + "grad_norm": 0.9199396967887878, + "learning_rate": 2.6181548078849583e-05, + "loss": 0.1531, + "step": 44466 + }, + { + "epoch": 3.0856290333772813, + "grad_norm": 0.6784421801567078, + "learning_rate": 2.6177757883805552e-05, + "loss": 0.3156, + "step": 44467 + }, + { + "epoch": 3.0856984248143777, + "grad_norm": 0.9703254699707031, + "learning_rate": 2.617396792180977e-05, + "loss": 0.3379, + "step": 44468 + }, + { + "epoch": 3.0857678162514746, + "grad_norm": 0.9625577330589294, + "learning_rate": 2.617017819287422e-05, + "loss": 0.1906, + "step": 44469 + }, + { + "epoch": 3.0858372076885714, + "grad_norm": 1.012467861175537, + "learning_rate": 2.6166388697010837e-05, + "loss": 0.1924, + "step": 44470 + }, + { + "epoch": 3.085906599125668, + "grad_norm": 0.8685722947120667, + "learning_rate": 2.6162599434231627e-05, + "loss": 0.2127, + "step": 44471 + }, + { + "epoch": 3.0859759905627646, + "grad_norm": 0.8838121891021729, + "learning_rate": 2.6158810404548494e-05, + "loss": 0.2819, + "step": 44472 + }, + { + "epoch": 3.086045381999861, + "grad_norm": 0.7942826747894287, + "learning_rate": 2.615502160797344e-05, + "loss": 0.2664, + "step": 44473 + }, + { + "epoch": 3.086114773436958, + "grad_norm": 0.7494845986366272, + "learning_rate": 2.615123304451841e-05, + "loss": 0.2361, + "step": 44474 + }, + { + "epoch": 3.0861841648740547, + "grad_norm": 1.1817331314086914, + "learning_rate": 2.6147444714195403e-05, + "loss": 0.3424, + "step": 44475 + }, + { + "epoch": 3.086253556311151, + "grad_norm": 0.8181073069572449, + "learning_rate": 2.614365661701631e-05, + "loss": 0.2242, + "step": 44476 + }, + { + "epoch": 3.086322947748248, + "grad_norm": 0.8289077877998352, + "learning_rate": 2.6139868752993123e-05, + "loss": 0.2518, + "step": 44477 + }, + { + "epoch": 3.0863923391853447, + "grad_norm": 1.0174692869186401, + "learning_rate": 2.6136081122137835e-05, + "loss": 0.2695, + "step": 44478 + }, + { + "epoch": 3.086461730622441, + "grad_norm": 1.270139455795288, + "learning_rate": 2.613229372446233e-05, + "loss": 0.2426, + "step": 44479 + }, + { + "epoch": 3.086531122059538, + "grad_norm": 0.7209979891777039, + "learning_rate": 2.6128506559978617e-05, + "loss": 0.2069, + "step": 44480 + }, + { + "epoch": 3.0866005134966343, + "grad_norm": 0.8798627853393555, + "learning_rate": 2.6124719628698624e-05, + "loss": 0.2583, + "step": 44481 + }, + { + "epoch": 3.086669904933731, + "grad_norm": 0.9580260515213013, + "learning_rate": 2.6120932930634334e-05, + "loss": 0.2015, + "step": 44482 + }, + { + "epoch": 3.086739296370828, + "grad_norm": 0.7570315599441528, + "learning_rate": 2.6117146465797682e-05, + "loss": 0.2357, + "step": 44483 + }, + { + "epoch": 3.0868086878079244, + "grad_norm": 0.8457432389259338, + "learning_rate": 2.6113360234200622e-05, + "loss": 0.1389, + "step": 44484 + }, + { + "epoch": 3.0868780792450212, + "grad_norm": 1.1984585523605347, + "learning_rate": 2.610957423585514e-05, + "loss": 0.3189, + "step": 44485 + }, + { + "epoch": 3.0869474706821176, + "grad_norm": 1.2576090097427368, + "learning_rate": 2.610578847077314e-05, + "loss": 0.2201, + "step": 44486 + }, + { + "epoch": 3.0870168621192144, + "grad_norm": 0.6353623270988464, + "learning_rate": 2.6102002938966587e-05, + "loss": 0.2374, + "step": 44487 + }, + { + "epoch": 3.0870862535563113, + "grad_norm": 1.27743399143219, + "learning_rate": 2.6098217640447432e-05, + "loss": 0.2534, + "step": 44488 + }, + { + "epoch": 3.0871556449934077, + "grad_norm": 0.9934810996055603, + "learning_rate": 2.6094432575227633e-05, + "loss": 0.35, + "step": 44489 + }, + { + "epoch": 3.0872250364305045, + "grad_norm": 1.1861540079116821, + "learning_rate": 2.6090647743319142e-05, + "loss": 0.3177, + "step": 44490 + }, + { + "epoch": 3.0872944278676013, + "grad_norm": 0.9349198341369629, + "learning_rate": 2.6086863144733886e-05, + "loss": 0.3461, + "step": 44491 + }, + { + "epoch": 3.0873638193046977, + "grad_norm": 1.0542268753051758, + "learning_rate": 2.6083078779483873e-05, + "loss": 0.264, + "step": 44492 + }, + { + "epoch": 3.0874332107417946, + "grad_norm": 1.018376350402832, + "learning_rate": 2.6079294647580964e-05, + "loss": 0.2363, + "step": 44493 + }, + { + "epoch": 3.087502602178891, + "grad_norm": 0.7073501944541931, + "learning_rate": 2.6075510749037156e-05, + "loss": 0.1363, + "step": 44494 + }, + { + "epoch": 3.0875719936159878, + "grad_norm": 0.8024802207946777, + "learning_rate": 2.6071727083864373e-05, + "loss": 0.292, + "step": 44495 + }, + { + "epoch": 3.0876413850530846, + "grad_norm": 0.6982554793357849, + "learning_rate": 2.6067943652074578e-05, + "loss": 0.2018, + "step": 44496 + }, + { + "epoch": 3.087710776490181, + "grad_norm": 1.264467477798462, + "learning_rate": 2.6064160453679697e-05, + "loss": 0.2611, + "step": 44497 + }, + { + "epoch": 3.087780167927278, + "grad_norm": 1.1699938774108887, + "learning_rate": 2.60603774886917e-05, + "loss": 0.3545, + "step": 44498 + }, + { + "epoch": 3.0878495593643747, + "grad_norm": 0.9631401896476746, + "learning_rate": 2.6056594757122532e-05, + "loss": 0.2898, + "step": 44499 + }, + { + "epoch": 3.087918950801471, + "grad_norm": 0.7684724926948547, + "learning_rate": 2.6052812258984083e-05, + "loss": 0.2436, + "step": 44500 + }, + { + "epoch": 3.087988342238568, + "grad_norm": 1.0390214920043945, + "learning_rate": 2.604902999428833e-05, + "loss": 0.2189, + "step": 44501 + }, + { + "epoch": 3.0880577336756643, + "grad_norm": 1.1696949005126953, + "learning_rate": 2.6045247963047216e-05, + "loss": 0.284, + "step": 44502 + }, + { + "epoch": 3.088127125112761, + "grad_norm": 0.8497242331504822, + "learning_rate": 2.6041466165272667e-05, + "loss": 0.175, + "step": 44503 + }, + { + "epoch": 3.088196516549858, + "grad_norm": 1.2050143480300903, + "learning_rate": 2.603768460097662e-05, + "loss": 0.3212, + "step": 44504 + }, + { + "epoch": 3.0882659079869543, + "grad_norm": 0.8567396402359009, + "learning_rate": 2.6033903270171035e-05, + "loss": 0.125, + "step": 44505 + }, + { + "epoch": 3.088335299424051, + "grad_norm": 1.2653555870056152, + "learning_rate": 2.6030122172867854e-05, + "loss": 0.2305, + "step": 44506 + }, + { + "epoch": 3.0884046908611476, + "grad_norm": 1.1075810194015503, + "learning_rate": 2.6026341309078972e-05, + "loss": 0.242, + "step": 44507 + }, + { + "epoch": 3.0884740822982444, + "grad_norm": 0.8167598843574524, + "learning_rate": 2.6022560678816356e-05, + "loss": 0.1415, + "step": 44508 + }, + { + "epoch": 3.088543473735341, + "grad_norm": 0.9585596919059753, + "learning_rate": 2.6018780282091947e-05, + "loss": 0.3664, + "step": 44509 + }, + { + "epoch": 3.0886128651724376, + "grad_norm": 0.7932007312774658, + "learning_rate": 2.6015000118917643e-05, + "loss": 0.2353, + "step": 44510 + }, + { + "epoch": 3.0886822566095344, + "grad_norm": 0.9244131445884705, + "learning_rate": 2.6011220189305407e-05, + "loss": 0.2251, + "step": 44511 + }, + { + "epoch": 3.088751648046631, + "grad_norm": 0.8570660948753357, + "learning_rate": 2.6007440493267154e-05, + "loss": 0.3014, + "step": 44512 + }, + { + "epoch": 3.0888210394837277, + "grad_norm": 0.7420376539230347, + "learning_rate": 2.6003661030814862e-05, + "loss": 0.2307, + "step": 44513 + }, + { + "epoch": 3.0888904309208245, + "grad_norm": 0.8262513875961304, + "learning_rate": 2.599988180196039e-05, + "loss": 0.2789, + "step": 44514 + }, + { + "epoch": 3.088959822357921, + "grad_norm": 1.507470965385437, + "learning_rate": 2.5996102806715716e-05, + "loss": 0.2751, + "step": 44515 + }, + { + "epoch": 3.0890292137950177, + "grad_norm": 0.8131290078163147, + "learning_rate": 2.5992324045092752e-05, + "loss": 0.2742, + "step": 44516 + }, + { + "epoch": 3.0890986052321145, + "grad_norm": 0.9938179850578308, + "learning_rate": 2.5988545517103436e-05, + "loss": 0.3572, + "step": 44517 + }, + { + "epoch": 3.089167996669211, + "grad_norm": 1.1863361597061157, + "learning_rate": 2.59847672227597e-05, + "loss": 0.2525, + "step": 44518 + }, + { + "epoch": 3.0892373881063078, + "grad_norm": 1.258249044418335, + "learning_rate": 2.5980989162073487e-05, + "loss": 0.2087, + "step": 44519 + }, + { + "epoch": 3.089306779543404, + "grad_norm": 0.8115279078483582, + "learning_rate": 2.5977211335056672e-05, + "loss": 0.1927, + "step": 44520 + }, + { + "epoch": 3.089376170980501, + "grad_norm": 0.9399133920669556, + "learning_rate": 2.597343374172122e-05, + "loss": 0.2313, + "step": 44521 + }, + { + "epoch": 3.089445562417598, + "grad_norm": 1.1720058917999268, + "learning_rate": 2.5969656382079055e-05, + "loss": 0.361, + "step": 44522 + }, + { + "epoch": 3.089514953854694, + "grad_norm": 1.0443365573883057, + "learning_rate": 2.596587925614209e-05, + "loss": 0.3083, + "step": 44523 + }, + { + "epoch": 3.089584345291791, + "grad_norm": 1.6290582418441772, + "learning_rate": 2.5962102363922246e-05, + "loss": 0.4502, + "step": 44524 + }, + { + "epoch": 3.089653736728888, + "grad_norm": 0.7622470259666443, + "learning_rate": 2.5958325705431465e-05, + "loss": 0.2444, + "step": 44525 + }, + { + "epoch": 3.0897231281659843, + "grad_norm": 1.115726113319397, + "learning_rate": 2.5954549280681685e-05, + "loss": 0.275, + "step": 44526 + }, + { + "epoch": 3.089792519603081, + "grad_norm": 0.8456157445907593, + "learning_rate": 2.5950773089684775e-05, + "loss": 0.2259, + "step": 44527 + }, + { + "epoch": 3.0898619110401775, + "grad_norm": 0.9467167854309082, + "learning_rate": 2.5946997132452676e-05, + "loss": 0.1935, + "step": 44528 + }, + { + "epoch": 3.0899313024772743, + "grad_norm": 1.1857351064682007, + "learning_rate": 2.5943221408997333e-05, + "loss": 0.2654, + "step": 44529 + }, + { + "epoch": 3.090000693914371, + "grad_norm": 0.8542999029159546, + "learning_rate": 2.593944591933063e-05, + "loss": 0.2072, + "step": 44530 + }, + { + "epoch": 3.0900700853514675, + "grad_norm": 0.547632098197937, + "learning_rate": 2.5935670663464507e-05, + "loss": 0.1221, + "step": 44531 + }, + { + "epoch": 3.0901394767885644, + "grad_norm": 1.1982347965240479, + "learning_rate": 2.5931895641410886e-05, + "loss": 0.4628, + "step": 44532 + }, + { + "epoch": 3.0902088682256608, + "grad_norm": 1.36290442943573, + "learning_rate": 2.5928120853181703e-05, + "loss": 0.2641, + "step": 44533 + }, + { + "epoch": 3.0902782596627576, + "grad_norm": 0.9362960457801819, + "learning_rate": 2.5924346298788817e-05, + "loss": 0.2293, + "step": 44534 + }, + { + "epoch": 3.0903476510998544, + "grad_norm": 0.7202136516571045, + "learning_rate": 2.592057197824417e-05, + "loss": 0.1944, + "step": 44535 + }, + { + "epoch": 3.090417042536951, + "grad_norm": 0.822981059551239, + "learning_rate": 2.5916797891559684e-05, + "loss": 0.2184, + "step": 44536 + }, + { + "epoch": 3.0904864339740477, + "grad_norm": 1.1314455270767212, + "learning_rate": 2.591302403874728e-05, + "loss": 0.2083, + "step": 44537 + }, + { + "epoch": 3.090555825411144, + "grad_norm": 0.7789680361747742, + "learning_rate": 2.5909250419818866e-05, + "loss": 0.2659, + "step": 44538 + }, + { + "epoch": 3.090625216848241, + "grad_norm": 0.8505771160125732, + "learning_rate": 2.590547703478634e-05, + "loss": 0.3399, + "step": 44539 + }, + { + "epoch": 3.0906946082853377, + "grad_norm": 0.8611390590667725, + "learning_rate": 2.5901703883661665e-05, + "loss": 0.2232, + "step": 44540 + }, + { + "epoch": 3.090763999722434, + "grad_norm": 0.9077962636947632, + "learning_rate": 2.5897930966456685e-05, + "loss": 0.2824, + "step": 44541 + }, + { + "epoch": 3.090833391159531, + "grad_norm": 0.7939990162849426, + "learning_rate": 2.5894158283183333e-05, + "loss": 0.2434, + "step": 44542 + }, + { + "epoch": 3.0909027825966278, + "grad_norm": 1.0004013776779175, + "learning_rate": 2.5890385833853526e-05, + "loss": 0.2108, + "step": 44543 + }, + { + "epoch": 3.090972174033724, + "grad_norm": 1.0106475353240967, + "learning_rate": 2.5886613618479205e-05, + "loss": 0.2417, + "step": 44544 + }, + { + "epoch": 3.091041565470821, + "grad_norm": 1.4324009418487549, + "learning_rate": 2.588284163707222e-05, + "loss": 0.4168, + "step": 44545 + }, + { + "epoch": 3.0911109569079174, + "grad_norm": 0.7940011024475098, + "learning_rate": 2.587906988964449e-05, + "loss": 0.2736, + "step": 44546 + }, + { + "epoch": 3.091180348345014, + "grad_norm": 1.002012848854065, + "learning_rate": 2.5875298376207956e-05, + "loss": 0.1915, + "step": 44547 + }, + { + "epoch": 3.091249739782111, + "grad_norm": 0.8814887404441833, + "learning_rate": 2.5871527096774515e-05, + "loss": 0.192, + "step": 44548 + }, + { + "epoch": 3.0913191312192074, + "grad_norm": 0.9788652062416077, + "learning_rate": 2.586775605135605e-05, + "loss": 0.2636, + "step": 44549 + }, + { + "epoch": 3.0913885226563043, + "grad_norm": 1.0275788307189941, + "learning_rate": 2.5863985239964472e-05, + "loss": 0.2892, + "step": 44550 + }, + { + "epoch": 3.091457914093401, + "grad_norm": 0.6672163605690002, + "learning_rate": 2.586021466261169e-05, + "loss": 0.2099, + "step": 44551 + }, + { + "epoch": 3.0915273055304975, + "grad_norm": 1.4540807008743286, + "learning_rate": 2.5856444319309614e-05, + "loss": 0.3741, + "step": 44552 + }, + { + "epoch": 3.0915966969675943, + "grad_norm": 0.9236897826194763, + "learning_rate": 2.5852674210070137e-05, + "loss": 0.2941, + "step": 44553 + }, + { + "epoch": 3.0916660884046907, + "grad_norm": 1.1512757539749146, + "learning_rate": 2.584890433490519e-05, + "loss": 0.2536, + "step": 44554 + }, + { + "epoch": 3.0917354798417875, + "grad_norm": 0.9994334578514099, + "learning_rate": 2.5845134693826622e-05, + "loss": 0.1813, + "step": 44555 + }, + { + "epoch": 3.0918048712788844, + "grad_norm": 0.9463122487068176, + "learning_rate": 2.5841365286846363e-05, + "loss": 0.2831, + "step": 44556 + }, + { + "epoch": 3.0918742627159808, + "grad_norm": 0.7335960865020752, + "learning_rate": 2.5837596113976313e-05, + "loss": 0.2428, + "step": 44557 + }, + { + "epoch": 3.0919436541530776, + "grad_norm": 1.349655032157898, + "learning_rate": 2.5833827175228366e-05, + "loss": 0.4369, + "step": 44558 + }, + { + "epoch": 3.092013045590174, + "grad_norm": 0.977756917476654, + "learning_rate": 2.5830058470614416e-05, + "loss": 0.3884, + "step": 44559 + }, + { + "epoch": 3.092082437027271, + "grad_norm": 0.903167188167572, + "learning_rate": 2.582629000014637e-05, + "loss": 0.2192, + "step": 44560 + }, + { + "epoch": 3.0921518284643676, + "grad_norm": 0.9858518838882446, + "learning_rate": 2.582252176383615e-05, + "loss": 0.2265, + "step": 44561 + }, + { + "epoch": 3.092221219901464, + "grad_norm": 0.8762786388397217, + "learning_rate": 2.581875376169559e-05, + "loss": 0.2428, + "step": 44562 + }, + { + "epoch": 3.092290611338561, + "grad_norm": 0.9334184527397156, + "learning_rate": 2.581498599373662e-05, + "loss": 0.2629, + "step": 44563 + }, + { + "epoch": 3.0923600027756573, + "grad_norm": 0.7633947134017944, + "learning_rate": 2.5811218459971132e-05, + "loss": 0.1921, + "step": 44564 + }, + { + "epoch": 3.092429394212754, + "grad_norm": 0.6678022146224976, + "learning_rate": 2.5807451160411024e-05, + "loss": 0.1883, + "step": 44565 + }, + { + "epoch": 3.092498785649851, + "grad_norm": 1.160605549812317, + "learning_rate": 2.580368409506817e-05, + "loss": 0.3907, + "step": 44566 + }, + { + "epoch": 3.0925681770869473, + "grad_norm": 0.9424981474876404, + "learning_rate": 2.5799917263954487e-05, + "loss": 0.2428, + "step": 44567 + }, + { + "epoch": 3.092637568524044, + "grad_norm": 1.1193865537643433, + "learning_rate": 2.5796150667081887e-05, + "loss": 0.3591, + "step": 44568 + }, + { + "epoch": 3.092706959961141, + "grad_norm": 0.8780977129936218, + "learning_rate": 2.5792384304462202e-05, + "loss": 0.2297, + "step": 44569 + }, + { + "epoch": 3.0927763513982374, + "grad_norm": 0.961403489112854, + "learning_rate": 2.578861817610735e-05, + "loss": 0.2572, + "step": 44570 + }, + { + "epoch": 3.092845742835334, + "grad_norm": 1.1429027318954468, + "learning_rate": 2.5784852282029215e-05, + "loss": 0.3239, + "step": 44571 + }, + { + "epoch": 3.0929151342724306, + "grad_norm": 0.7699694633483887, + "learning_rate": 2.57810866222397e-05, + "loss": 0.1446, + "step": 44572 + }, + { + "epoch": 3.0929845257095274, + "grad_norm": 0.6992621421813965, + "learning_rate": 2.5777321196750682e-05, + "loss": 0.2562, + "step": 44573 + }, + { + "epoch": 3.0930539171466243, + "grad_norm": 1.4864720106124878, + "learning_rate": 2.577355600557404e-05, + "loss": 0.2995, + "step": 44574 + }, + { + "epoch": 3.0931233085837206, + "grad_norm": 1.1202367544174194, + "learning_rate": 2.57697910487217e-05, + "loss": 0.3184, + "step": 44575 + }, + { + "epoch": 3.0931927000208175, + "grad_norm": 0.7149022221565247, + "learning_rate": 2.5766026326205495e-05, + "loss": 0.2657, + "step": 44576 + }, + { + "epoch": 3.0932620914579143, + "grad_norm": 1.0917398929595947, + "learning_rate": 2.5762261838037338e-05, + "loss": 0.4039, + "step": 44577 + }, + { + "epoch": 3.0933314828950107, + "grad_norm": 0.918931245803833, + "learning_rate": 2.5758497584229092e-05, + "loss": 0.2115, + "step": 44578 + }, + { + "epoch": 3.0934008743321075, + "grad_norm": 0.9250279068946838, + "learning_rate": 2.57547335647927e-05, + "loss": 0.2772, + "step": 44579 + }, + { + "epoch": 3.093470265769204, + "grad_norm": 1.070205807685852, + "learning_rate": 2.5750969779739954e-05, + "loss": 0.2029, + "step": 44580 + }, + { + "epoch": 3.0935396572063008, + "grad_norm": 0.7450166940689087, + "learning_rate": 2.5747206229082798e-05, + "loss": 0.2012, + "step": 44581 + }, + { + "epoch": 3.0936090486433976, + "grad_norm": 1.0386042594909668, + "learning_rate": 2.5743442912833082e-05, + "loss": 0.2504, + "step": 44582 + }, + { + "epoch": 3.093678440080494, + "grad_norm": 0.7100386023521423, + "learning_rate": 2.573967983100274e-05, + "loss": 0.1949, + "step": 44583 + }, + { + "epoch": 3.093747831517591, + "grad_norm": 0.8111453652381897, + "learning_rate": 2.5735916983603582e-05, + "loss": 0.2076, + "step": 44584 + }, + { + "epoch": 3.093817222954687, + "grad_norm": 2.08457350730896, + "learning_rate": 2.573215437064752e-05, + "loss": 0.2595, + "step": 44585 + }, + { + "epoch": 3.093886614391784, + "grad_norm": 0.7141116857528687, + "learning_rate": 2.5728391992146426e-05, + "loss": 0.1912, + "step": 44586 + }, + { + "epoch": 3.093956005828881, + "grad_norm": 0.9146908521652222, + "learning_rate": 2.572462984811218e-05, + "loss": 0.2654, + "step": 44587 + }, + { + "epoch": 3.0940253972659773, + "grad_norm": 0.5648170113563538, + "learning_rate": 2.5720867938556693e-05, + "loss": 0.1756, + "step": 44588 + }, + { + "epoch": 3.094094788703074, + "grad_norm": 1.058436632156372, + "learning_rate": 2.5717106263491785e-05, + "loss": 0.3083, + "step": 44589 + }, + { + "epoch": 3.094164180140171, + "grad_norm": 0.8434172868728638, + "learning_rate": 2.571334482292934e-05, + "loss": 0.2766, + "step": 44590 + }, + { + "epoch": 3.0942335715772673, + "grad_norm": 1.1199320554733276, + "learning_rate": 2.570958361688126e-05, + "loss": 0.1363, + "step": 44591 + }, + { + "epoch": 3.094302963014364, + "grad_norm": 1.0947825908660889, + "learning_rate": 2.57058226453594e-05, + "loss": 0.264, + "step": 44592 + }, + { + "epoch": 3.0943723544514605, + "grad_norm": 0.678540825843811, + "learning_rate": 2.570206190837563e-05, + "loss": 0.2315, + "step": 44593 + }, + { + "epoch": 3.0944417458885574, + "grad_norm": 0.9453098177909851, + "learning_rate": 2.5698301405941838e-05, + "loss": 0.2153, + "step": 44594 + }, + { + "epoch": 3.094511137325654, + "grad_norm": 0.8857366442680359, + "learning_rate": 2.5694541138069916e-05, + "loss": 0.1889, + "step": 44595 + }, + { + "epoch": 3.0945805287627506, + "grad_norm": 1.0823791027069092, + "learning_rate": 2.569078110477169e-05, + "loss": 0.3572, + "step": 44596 + }, + { + "epoch": 3.0946499201998474, + "grad_norm": 1.1994469165802002, + "learning_rate": 2.5687021306059032e-05, + "loss": 0.2383, + "step": 44597 + }, + { + "epoch": 3.094719311636944, + "grad_norm": 1.3491195440292358, + "learning_rate": 2.568326174194382e-05, + "loss": 0.2937, + "step": 44598 + }, + { + "epoch": 3.0947887030740406, + "grad_norm": 0.8775956034660339, + "learning_rate": 2.567950241243794e-05, + "loss": 0.2193, + "step": 44599 + }, + { + "epoch": 3.0948580945111375, + "grad_norm": 0.9376315474510193, + "learning_rate": 2.5675743317553248e-05, + "loss": 0.2721, + "step": 44600 + }, + { + "epoch": 3.094927485948234, + "grad_norm": 0.8638089895248413, + "learning_rate": 2.5671984457301613e-05, + "loss": 0.3362, + "step": 44601 + }, + { + "epoch": 3.0949968773853307, + "grad_norm": 0.8634494543075562, + "learning_rate": 2.5668225831694913e-05, + "loss": 0.3086, + "step": 44602 + }, + { + "epoch": 3.0950662688224275, + "grad_norm": 0.987469494342804, + "learning_rate": 2.566446744074499e-05, + "loss": 0.2846, + "step": 44603 + }, + { + "epoch": 3.095135660259524, + "grad_norm": 0.9027338027954102, + "learning_rate": 2.56607092844637e-05, + "loss": 0.2316, + "step": 44604 + }, + { + "epoch": 3.0952050516966207, + "grad_norm": 1.2432246208190918, + "learning_rate": 2.5656951362862946e-05, + "loss": 0.4587, + "step": 44605 + }, + { + "epoch": 3.095274443133717, + "grad_norm": 0.8006539344787598, + "learning_rate": 2.5653193675954556e-05, + "loss": 0.2217, + "step": 44606 + }, + { + "epoch": 3.095343834570814, + "grad_norm": 0.7770971059799194, + "learning_rate": 2.564943622375041e-05, + "loss": 0.2036, + "step": 44607 + }, + { + "epoch": 3.095413226007911, + "grad_norm": 1.018734097480774, + "learning_rate": 2.564567900626238e-05, + "loss": 0.302, + "step": 44608 + }, + { + "epoch": 3.095482617445007, + "grad_norm": 1.0465834140777588, + "learning_rate": 2.5641922023502328e-05, + "loss": 0.2739, + "step": 44609 + }, + { + "epoch": 3.095552008882104, + "grad_norm": 1.794076919555664, + "learning_rate": 2.5638165275482083e-05, + "loss": 0.3026, + "step": 44610 + }, + { + "epoch": 3.0956214003192004, + "grad_norm": 1.2405190467834473, + "learning_rate": 2.5634408762213512e-05, + "loss": 0.2976, + "step": 44611 + }, + { + "epoch": 3.0956907917562972, + "grad_norm": 0.6832403540611267, + "learning_rate": 2.5630652483708495e-05, + "loss": 0.2121, + "step": 44612 + }, + { + "epoch": 3.095760183193394, + "grad_norm": 1.0359903573989868, + "learning_rate": 2.5626896439978877e-05, + "loss": 0.3532, + "step": 44613 + }, + { + "epoch": 3.0958295746304905, + "grad_norm": 1.0355054140090942, + "learning_rate": 2.5623140631036545e-05, + "loss": 0.2651, + "step": 44614 + }, + { + "epoch": 3.0958989660675873, + "grad_norm": 1.1236306428909302, + "learning_rate": 2.56193850568933e-05, + "loss": 0.2554, + "step": 44615 + }, + { + "epoch": 3.095968357504684, + "grad_norm": 1.130265712738037, + "learning_rate": 2.561562971756103e-05, + "loss": 0.3801, + "step": 44616 + }, + { + "epoch": 3.0960377489417805, + "grad_norm": 1.2842369079589844, + "learning_rate": 2.5611874613051578e-05, + "loss": 0.4234, + "step": 44617 + }, + { + "epoch": 3.0961071403788774, + "grad_norm": 1.061399221420288, + "learning_rate": 2.5608119743376845e-05, + "loss": 0.2817, + "step": 44618 + }, + { + "epoch": 3.0961765318159737, + "grad_norm": 1.4575759172439575, + "learning_rate": 2.560436510854861e-05, + "loss": 0.5093, + "step": 44619 + }, + { + "epoch": 3.0962459232530706, + "grad_norm": 0.7512929439544678, + "learning_rate": 2.5600610708578765e-05, + "loss": 0.1331, + "step": 44620 + }, + { + "epoch": 3.0963153146901674, + "grad_norm": 0.8891457915306091, + "learning_rate": 2.5596856543479163e-05, + "loss": 0.1956, + "step": 44621 + }, + { + "epoch": 3.096384706127264, + "grad_norm": 0.803192138671875, + "learning_rate": 2.5593102613261655e-05, + "loss": 0.1886, + "step": 44622 + }, + { + "epoch": 3.0964540975643606, + "grad_norm": 1.051030158996582, + "learning_rate": 2.5589348917938116e-05, + "loss": 0.2503, + "step": 44623 + }, + { + "epoch": 3.096523489001457, + "grad_norm": 1.2315049171447754, + "learning_rate": 2.558559545752034e-05, + "loss": 0.2323, + "step": 44624 + }, + { + "epoch": 3.096592880438554, + "grad_norm": 1.0129938125610352, + "learning_rate": 2.5581842232020205e-05, + "loss": 0.3924, + "step": 44625 + }, + { + "epoch": 3.0966622718756507, + "grad_norm": 1.073890209197998, + "learning_rate": 2.5578089241449555e-05, + "loss": 0.2583, + "step": 44626 + }, + { + "epoch": 3.096731663312747, + "grad_norm": 1.0236200094223022, + "learning_rate": 2.5574336485820248e-05, + "loss": 0.3072, + "step": 44627 + }, + { + "epoch": 3.096801054749844, + "grad_norm": 1.3115265369415283, + "learning_rate": 2.5570583965144135e-05, + "loss": 0.3272, + "step": 44628 + }, + { + "epoch": 3.0968704461869407, + "grad_norm": 1.1840213537216187, + "learning_rate": 2.5566831679433034e-05, + "loss": 0.2052, + "step": 44629 + }, + { + "epoch": 3.096939837624037, + "grad_norm": 0.7292435765266418, + "learning_rate": 2.556307962869885e-05, + "loss": 0.2252, + "step": 44630 + }, + { + "epoch": 3.097009229061134, + "grad_norm": 0.8771979808807373, + "learning_rate": 2.555932781295336e-05, + "loss": 0.2602, + "step": 44631 + }, + { + "epoch": 3.0970786204982304, + "grad_norm": 0.7453383207321167, + "learning_rate": 2.5555576232208435e-05, + "loss": 0.3179, + "step": 44632 + }, + { + "epoch": 3.097148011935327, + "grad_norm": 1.1840580701828003, + "learning_rate": 2.5551824886475904e-05, + "loss": 0.2112, + "step": 44633 + }, + { + "epoch": 3.097217403372424, + "grad_norm": 0.872481644153595, + "learning_rate": 2.5548073775767644e-05, + "loss": 0.229, + "step": 44634 + }, + { + "epoch": 3.0972867948095204, + "grad_norm": 1.047585368156433, + "learning_rate": 2.5544322900095464e-05, + "loss": 0.2588, + "step": 44635 + }, + { + "epoch": 3.0973561862466172, + "grad_norm": 0.8475011587142944, + "learning_rate": 2.554057225947123e-05, + "loss": 0.1719, + "step": 44636 + }, + { + "epoch": 3.0974255776837136, + "grad_norm": 0.7894859313964844, + "learning_rate": 2.553682185390679e-05, + "loss": 0.1701, + "step": 44637 + }, + { + "epoch": 3.0974949691208105, + "grad_norm": 0.717204213142395, + "learning_rate": 2.5533071683413933e-05, + "loss": 0.2141, + "step": 44638 + }, + { + "epoch": 3.0975643605579073, + "grad_norm": 0.914520800113678, + "learning_rate": 2.552932174800453e-05, + "loss": 0.213, + "step": 44639 + }, + { + "epoch": 3.0976337519950037, + "grad_norm": 0.8480232357978821, + "learning_rate": 2.5525572047690427e-05, + "loss": 0.286, + "step": 44640 + }, + { + "epoch": 3.0977031434321005, + "grad_norm": 0.7283748984336853, + "learning_rate": 2.5521822582483447e-05, + "loss": 0.2214, + "step": 44641 + }, + { + "epoch": 3.0977725348691973, + "grad_norm": 0.9853072762489319, + "learning_rate": 2.5518073352395432e-05, + "loss": 0.2243, + "step": 44642 + }, + { + "epoch": 3.0978419263062937, + "grad_norm": 0.984700620174408, + "learning_rate": 2.5514324357438213e-05, + "loss": 0.3223, + "step": 44643 + }, + { + "epoch": 3.0979113177433906, + "grad_norm": 0.8799157738685608, + "learning_rate": 2.5510575597623663e-05, + "loss": 0.2622, + "step": 44644 + }, + { + "epoch": 3.097980709180487, + "grad_norm": 1.3396482467651367, + "learning_rate": 2.550682707296356e-05, + "loss": 0.3378, + "step": 44645 + }, + { + "epoch": 3.098050100617584, + "grad_norm": 0.987825334072113, + "learning_rate": 2.5503078783469758e-05, + "loss": 0.4115, + "step": 44646 + }, + { + "epoch": 3.0981194920546806, + "grad_norm": 1.0828673839569092, + "learning_rate": 2.5499330729154093e-05, + "loss": 0.2317, + "step": 44647 + }, + { + "epoch": 3.098188883491777, + "grad_norm": 1.0729241371154785, + "learning_rate": 2.549558291002839e-05, + "loss": 0.3861, + "step": 44648 + }, + { + "epoch": 3.098258274928874, + "grad_norm": 0.7548777461051941, + "learning_rate": 2.5491835326104528e-05, + "loss": 0.1767, + "step": 44649 + }, + { + "epoch": 3.0983276663659702, + "grad_norm": 0.6795552968978882, + "learning_rate": 2.548808797739426e-05, + "loss": 0.1503, + "step": 44650 + }, + { + "epoch": 3.098397057803067, + "grad_norm": 0.6807668209075928, + "learning_rate": 2.548434086390945e-05, + "loss": 0.19, + "step": 44651 + }, + { + "epoch": 3.098466449240164, + "grad_norm": 1.1824089288711548, + "learning_rate": 2.548059398566197e-05, + "loss": 0.3941, + "step": 44652 + }, + { + "epoch": 3.0985358406772603, + "grad_norm": 1.2493993043899536, + "learning_rate": 2.547684734266358e-05, + "loss": 0.4287, + "step": 44653 + }, + { + "epoch": 3.098605232114357, + "grad_norm": 0.7653309106826782, + "learning_rate": 2.547310093492613e-05, + "loss": 0.2962, + "step": 44654 + }, + { + "epoch": 3.098674623551454, + "grad_norm": 1.3885464668273926, + "learning_rate": 2.546935476246145e-05, + "loss": 0.3774, + "step": 44655 + }, + { + "epoch": 3.0987440149885503, + "grad_norm": 0.7440468668937683, + "learning_rate": 2.5465608825281383e-05, + "loss": 0.1865, + "step": 44656 + }, + { + "epoch": 3.098813406425647, + "grad_norm": 0.7642331719398499, + "learning_rate": 2.5461863123397735e-05, + "loss": 0.2278, + "step": 44657 + }, + { + "epoch": 3.0988827978627436, + "grad_norm": 1.204654574394226, + "learning_rate": 2.5458117656822357e-05, + "loss": 0.3476, + "step": 44658 + }, + { + "epoch": 3.0989521892998404, + "grad_norm": 0.8825735449790955, + "learning_rate": 2.545437242556704e-05, + "loss": 0.2663, + "step": 44659 + }, + { + "epoch": 3.0990215807369372, + "grad_norm": 0.8208959698677063, + "learning_rate": 2.5450627429643602e-05, + "loss": 0.2622, + "step": 44660 + }, + { + "epoch": 3.0990909721740336, + "grad_norm": 1.1331056356430054, + "learning_rate": 2.5446882669063898e-05, + "loss": 0.2463, + "step": 44661 + }, + { + "epoch": 3.0991603636111305, + "grad_norm": 1.1501224040985107, + "learning_rate": 2.5443138143839728e-05, + "loss": 0.337, + "step": 44662 + }, + { + "epoch": 3.099229755048227, + "grad_norm": 0.8194695711135864, + "learning_rate": 2.543939385398293e-05, + "loss": 0.2454, + "step": 44663 + }, + { + "epoch": 3.0992991464853237, + "grad_norm": 1.1510010957717896, + "learning_rate": 2.5435649799505334e-05, + "loss": 0.3546, + "step": 44664 + }, + { + "epoch": 3.0993685379224205, + "grad_norm": 0.6946366429328918, + "learning_rate": 2.5431905980418725e-05, + "loss": 0.189, + "step": 44665 + }, + { + "epoch": 3.099437929359517, + "grad_norm": 1.027697205543518, + "learning_rate": 2.5428162396734923e-05, + "loss": 0.2986, + "step": 44666 + }, + { + "epoch": 3.0995073207966137, + "grad_norm": 0.9602401256561279, + "learning_rate": 2.5424419048465764e-05, + "loss": 0.2681, + "step": 44667 + }, + { + "epoch": 3.0995767122337106, + "grad_norm": 1.0715477466583252, + "learning_rate": 2.5420675935623063e-05, + "loss": 0.2954, + "step": 44668 + }, + { + "epoch": 3.099646103670807, + "grad_norm": 0.8730539679527283, + "learning_rate": 2.541693305821864e-05, + "loss": 0.1795, + "step": 44669 + }, + { + "epoch": 3.099715495107904, + "grad_norm": 0.8739388585090637, + "learning_rate": 2.5413190416264298e-05, + "loss": 0.2702, + "step": 44670 + }, + { + "epoch": 3.099784886545, + "grad_norm": 0.82578045129776, + "learning_rate": 2.5409448009771896e-05, + "loss": 0.2066, + "step": 44671 + }, + { + "epoch": 3.099854277982097, + "grad_norm": 1.04578697681427, + "learning_rate": 2.540570583875318e-05, + "loss": 0.2809, + "step": 44672 + }, + { + "epoch": 3.099923669419194, + "grad_norm": 1.237209677696228, + "learning_rate": 2.5401963903219994e-05, + "loss": 0.5155, + "step": 44673 + }, + { + "epoch": 3.0999930608562902, + "grad_norm": 0.8951730132102966, + "learning_rate": 2.5398222203184163e-05, + "loss": 0.223, + "step": 44674 + }, + { + "epoch": 3.100062452293387, + "grad_norm": 1.022648572921753, + "learning_rate": 2.5394480738657478e-05, + "loss": 0.3247, + "step": 44675 + }, + { + "epoch": 3.1001318437304834, + "grad_norm": 0.7025517821311951, + "learning_rate": 2.5390739509651763e-05, + "loss": 0.2376, + "step": 44676 + }, + { + "epoch": 3.1002012351675803, + "grad_norm": 1.0880697965621948, + "learning_rate": 2.538699851617883e-05, + "loss": 0.3166, + "step": 44677 + }, + { + "epoch": 3.100270626604677, + "grad_norm": 0.6907352209091187, + "learning_rate": 2.538325775825051e-05, + "loss": 0.1431, + "step": 44678 + }, + { + "epoch": 3.1003400180417735, + "grad_norm": 0.7476900219917297, + "learning_rate": 2.537951723587857e-05, + "loss": 0.2429, + "step": 44679 + }, + { + "epoch": 3.1004094094788703, + "grad_norm": 0.8262800574302673, + "learning_rate": 2.5375776949074826e-05, + "loss": 0.3075, + "step": 44680 + }, + { + "epoch": 3.100478800915967, + "grad_norm": 0.7713396549224854, + "learning_rate": 2.5372036897851104e-05, + "loss": 0.2003, + "step": 44681 + }, + { + "epoch": 3.1005481923530636, + "grad_norm": 0.7753056883811951, + "learning_rate": 2.5368297082219195e-05, + "loss": 0.1196, + "step": 44682 + }, + { + "epoch": 3.1006175837901604, + "grad_norm": 0.966526210308075, + "learning_rate": 2.5364557502190923e-05, + "loss": 0.3063, + "step": 44683 + }, + { + "epoch": 3.100686975227257, + "grad_norm": 1.1456760168075562, + "learning_rate": 2.536081815777811e-05, + "loss": 0.4264, + "step": 44684 + }, + { + "epoch": 3.1007563666643536, + "grad_norm": 0.9515067934989929, + "learning_rate": 2.5357079048992495e-05, + "loss": 0.4373, + "step": 44685 + }, + { + "epoch": 3.1008257581014504, + "grad_norm": 0.8760546445846558, + "learning_rate": 2.535334017584593e-05, + "loss": 0.3114, + "step": 44686 + }, + { + "epoch": 3.100895149538547, + "grad_norm": 0.5380197167396545, + "learning_rate": 2.5349601538350243e-05, + "loss": 0.158, + "step": 44687 + }, + { + "epoch": 3.1009645409756437, + "grad_norm": 1.080424427986145, + "learning_rate": 2.5345863136517167e-05, + "loss": 0.2431, + "step": 44688 + }, + { + "epoch": 3.10103393241274, + "grad_norm": 0.728079080581665, + "learning_rate": 2.5342124970358537e-05, + "loss": 0.2383, + "step": 44689 + }, + { + "epoch": 3.101103323849837, + "grad_norm": 1.2333989143371582, + "learning_rate": 2.5338387039886157e-05, + "loss": 0.2847, + "step": 44690 + }, + { + "epoch": 3.1011727152869337, + "grad_norm": 1.6157342195510864, + "learning_rate": 2.5334649345111828e-05, + "loss": 0.4403, + "step": 44691 + }, + { + "epoch": 3.10124210672403, + "grad_norm": 5.199967861175537, + "learning_rate": 2.533091188604738e-05, + "loss": 0.31, + "step": 44692 + }, + { + "epoch": 3.101311498161127, + "grad_norm": 0.8328163623809814, + "learning_rate": 2.5327174662704557e-05, + "loss": 0.2753, + "step": 44693 + }, + { + "epoch": 3.1013808895982238, + "grad_norm": 0.7091754674911499, + "learning_rate": 2.5323437675095173e-05, + "loss": 0.2429, + "step": 44694 + }, + { + "epoch": 3.10145028103532, + "grad_norm": 1.0911741256713867, + "learning_rate": 2.5319700923231026e-05, + "loss": 0.377, + "step": 44695 + }, + { + "epoch": 3.101519672472417, + "grad_norm": 0.8540477752685547, + "learning_rate": 2.5315964407123927e-05, + "loss": 0.2462, + "step": 44696 + }, + { + "epoch": 3.1015890639095134, + "grad_norm": 0.6832461357116699, + "learning_rate": 2.5312228126785665e-05, + "loss": 0.2416, + "step": 44697 + }, + { + "epoch": 3.10165845534661, + "grad_norm": 1.0012420415878296, + "learning_rate": 2.530849208222803e-05, + "loss": 0.2128, + "step": 44698 + }, + { + "epoch": 3.101727846783707, + "grad_norm": 0.8980221152305603, + "learning_rate": 2.5304756273462847e-05, + "loss": 0.3126, + "step": 44699 + }, + { + "epoch": 3.1017972382208034, + "grad_norm": 0.7318108081817627, + "learning_rate": 2.5301020700501853e-05, + "loss": 0.2139, + "step": 44700 + }, + { + "epoch": 3.1018666296579003, + "grad_norm": 0.8235574960708618, + "learning_rate": 2.529728536335687e-05, + "loss": 0.2049, + "step": 44701 + }, + { + "epoch": 3.1019360210949967, + "grad_norm": 0.9171142578125, + "learning_rate": 2.529355026203969e-05, + "loss": 0.2251, + "step": 44702 + }, + { + "epoch": 3.1020054125320935, + "grad_norm": 0.9307078719139099, + "learning_rate": 2.528981539656211e-05, + "loss": 0.2257, + "step": 44703 + }, + { + "epoch": 3.1020748039691903, + "grad_norm": 0.9898384809494019, + "learning_rate": 2.5286080766935906e-05, + "loss": 0.3247, + "step": 44704 + }, + { + "epoch": 3.1021441954062867, + "grad_norm": 0.9064143300056458, + "learning_rate": 2.5282346373172883e-05, + "loss": 0.2646, + "step": 44705 + }, + { + "epoch": 3.1022135868433836, + "grad_norm": 0.8046214580535889, + "learning_rate": 2.5278612215284847e-05, + "loss": 0.2709, + "step": 44706 + }, + { + "epoch": 3.1022829782804804, + "grad_norm": 1.1179465055465698, + "learning_rate": 2.5274878293283543e-05, + "loss": 0.3627, + "step": 44707 + }, + { + "epoch": 3.1023523697175768, + "grad_norm": 1.0027618408203125, + "learning_rate": 2.5271144607180775e-05, + "loss": 0.2196, + "step": 44708 + }, + { + "epoch": 3.1024217611546736, + "grad_norm": 0.7009956240653992, + "learning_rate": 2.526741115698833e-05, + "loss": 0.1465, + "step": 44709 + }, + { + "epoch": 3.10249115259177, + "grad_norm": 0.7986167669296265, + "learning_rate": 2.5263677942718e-05, + "loss": 0.2149, + "step": 44710 + }, + { + "epoch": 3.102560544028867, + "grad_norm": 0.6638970971107483, + "learning_rate": 2.5259944964381565e-05, + "loss": 0.1738, + "step": 44711 + }, + { + "epoch": 3.1026299354659637, + "grad_norm": 0.9415666460990906, + "learning_rate": 2.5256212221990815e-05, + "loss": 0.2705, + "step": 44712 + }, + { + "epoch": 3.10269932690306, + "grad_norm": 0.8104736804962158, + "learning_rate": 2.5252479715557552e-05, + "loss": 0.2399, + "step": 44713 + }, + { + "epoch": 3.102768718340157, + "grad_norm": 0.9837777018547058, + "learning_rate": 2.5248747445093523e-05, + "loss": 0.1954, + "step": 44714 + }, + { + "epoch": 3.1028381097772533, + "grad_norm": 0.9211798310279846, + "learning_rate": 2.5245015410610518e-05, + "loss": 0.2795, + "step": 44715 + }, + { + "epoch": 3.10290750121435, + "grad_norm": 0.6716299057006836, + "learning_rate": 2.5241283612120325e-05, + "loss": 0.2186, + "step": 44716 + }, + { + "epoch": 3.102976892651447, + "grad_norm": 0.9298258423805237, + "learning_rate": 2.5237552049634727e-05, + "loss": 0.1682, + "step": 44717 + }, + { + "epoch": 3.1030462840885433, + "grad_norm": 1.0068577527999878, + "learning_rate": 2.5233820723165502e-05, + "loss": 0.1334, + "step": 44718 + }, + { + "epoch": 3.10311567552564, + "grad_norm": 1.4318500757217407, + "learning_rate": 2.5230089632724464e-05, + "loss": 0.3092, + "step": 44719 + }, + { + "epoch": 3.103185066962737, + "grad_norm": 0.5953439474105835, + "learning_rate": 2.5226358778323333e-05, + "loss": 0.2339, + "step": 44720 + }, + { + "epoch": 3.1032544583998334, + "grad_norm": 0.8711426854133606, + "learning_rate": 2.5222628159973903e-05, + "loss": 0.3525, + "step": 44721 + }, + { + "epoch": 3.10332384983693, + "grad_norm": 0.9714680910110474, + "learning_rate": 2.5218897777687988e-05, + "loss": 0.2974, + "step": 44722 + }, + { + "epoch": 3.1033932412740266, + "grad_norm": 0.9388777017593384, + "learning_rate": 2.521516763147731e-05, + "loss": 0.307, + "step": 44723 + }, + { + "epoch": 3.1034626327111234, + "grad_norm": 1.5337902307510376, + "learning_rate": 2.521143772135368e-05, + "loss": 0.2538, + "step": 44724 + }, + { + "epoch": 3.1035320241482203, + "grad_norm": 0.9294021129608154, + "learning_rate": 2.5207708047328858e-05, + "loss": 0.2305, + "step": 44725 + }, + { + "epoch": 3.1036014155853167, + "grad_norm": 0.8084582686424255, + "learning_rate": 2.5203978609414626e-05, + "loss": 0.2869, + "step": 44726 + }, + { + "epoch": 3.1036708070224135, + "grad_norm": 0.9339015483856201, + "learning_rate": 2.520024940762278e-05, + "loss": 0.2642, + "step": 44727 + }, + { + "epoch": 3.1037401984595103, + "grad_norm": 0.8968790173530579, + "learning_rate": 2.5196520441965034e-05, + "loss": 0.2854, + "step": 44728 + }, + { + "epoch": 3.1038095898966067, + "grad_norm": 0.6845057010650635, + "learning_rate": 2.51927917124532e-05, + "loss": 0.1956, + "step": 44729 + }, + { + "epoch": 3.1038789813337035, + "grad_norm": 0.8707348704338074, + "learning_rate": 2.5189063219099052e-05, + "loss": 0.208, + "step": 44730 + }, + { + "epoch": 3.1039483727708, + "grad_norm": 0.7271638512611389, + "learning_rate": 2.5185334961914343e-05, + "loss": 0.212, + "step": 44731 + }, + { + "epoch": 3.1040177642078968, + "grad_norm": 0.8526099324226379, + "learning_rate": 2.518160694091085e-05, + "loss": 0.3181, + "step": 44732 + }, + { + "epoch": 3.1040871556449936, + "grad_norm": 0.6912098526954651, + "learning_rate": 2.5177879156100346e-05, + "loss": 0.2142, + "step": 44733 + }, + { + "epoch": 3.10415654708209, + "grad_norm": 1.2850096225738525, + "learning_rate": 2.517415160749462e-05, + "loss": 0.3004, + "step": 44734 + }, + { + "epoch": 3.104225938519187, + "grad_norm": 1.096372127532959, + "learning_rate": 2.51704242951054e-05, + "loss": 0.3684, + "step": 44735 + }, + { + "epoch": 3.104295329956283, + "grad_norm": 0.9309507608413696, + "learning_rate": 2.5166697218944457e-05, + "loss": 0.26, + "step": 44736 + }, + { + "epoch": 3.10436472139338, + "grad_norm": 1.4973959922790527, + "learning_rate": 2.5162970379023565e-05, + "loss": 0.4255, + "step": 44737 + }, + { + "epoch": 3.104434112830477, + "grad_norm": 0.6705449223518372, + "learning_rate": 2.51592437753545e-05, + "loss": 0.1986, + "step": 44738 + }, + { + "epoch": 3.1045035042675733, + "grad_norm": 1.0260708332061768, + "learning_rate": 2.5155517407949024e-05, + "loss": 0.3217, + "step": 44739 + }, + { + "epoch": 3.10457289570467, + "grad_norm": 0.8390262722969055, + "learning_rate": 2.5151791276818926e-05, + "loss": 0.2516, + "step": 44740 + }, + { + "epoch": 3.1046422871417665, + "grad_norm": 0.9266583919525146, + "learning_rate": 2.5148065381975906e-05, + "loss": 0.2528, + "step": 44741 + }, + { + "epoch": 3.1047116785788633, + "grad_norm": 0.6901194453239441, + "learning_rate": 2.5144339723431754e-05, + "loss": 0.2421, + "step": 44742 + }, + { + "epoch": 3.10478107001596, + "grad_norm": 0.7239407300949097, + "learning_rate": 2.514061430119824e-05, + "loss": 0.2737, + "step": 44743 + }, + { + "epoch": 3.1048504614530565, + "grad_norm": 0.794436514377594, + "learning_rate": 2.513688911528712e-05, + "loss": 0.2808, + "step": 44744 + }, + { + "epoch": 3.1049198528901534, + "grad_norm": 0.9784688353538513, + "learning_rate": 2.5133164165710155e-05, + "loss": 0.2331, + "step": 44745 + }, + { + "epoch": 3.10498924432725, + "grad_norm": 0.7807585597038269, + "learning_rate": 2.512943945247911e-05, + "loss": 0.2121, + "step": 44746 + }, + { + "epoch": 3.1050586357643466, + "grad_norm": 1.0148358345031738, + "learning_rate": 2.5125714975605763e-05, + "loss": 0.3263, + "step": 44747 + }, + { + "epoch": 3.1051280272014434, + "grad_norm": 0.8478595614433289, + "learning_rate": 2.512199073510182e-05, + "loss": 0.2148, + "step": 44748 + }, + { + "epoch": 3.10519741863854, + "grad_norm": 1.2404595613479614, + "learning_rate": 2.5118266730979056e-05, + "loss": 0.3279, + "step": 44749 + }, + { + "epoch": 3.1052668100756367, + "grad_norm": 0.8722612857818604, + "learning_rate": 2.5114542963249242e-05, + "loss": 0.2455, + "step": 44750 + }, + { + "epoch": 3.1053362015127335, + "grad_norm": 0.7539102435112, + "learning_rate": 2.511081943192414e-05, + "loss": 0.2277, + "step": 44751 + }, + { + "epoch": 3.10540559294983, + "grad_norm": 0.6083895564079285, + "learning_rate": 2.5107096137015475e-05, + "loss": 0.112, + "step": 44752 + }, + { + "epoch": 3.1054749843869267, + "grad_norm": 0.6753592491149902, + "learning_rate": 2.5103373078535052e-05, + "loss": 0.2085, + "step": 44753 + }, + { + "epoch": 3.1055443758240235, + "grad_norm": 1.0747829675674438, + "learning_rate": 2.5099650256494566e-05, + "loss": 0.3038, + "step": 44754 + }, + { + "epoch": 3.10561376726112, + "grad_norm": 0.989790678024292, + "learning_rate": 2.5095927670905784e-05, + "loss": 0.3054, + "step": 44755 + }, + { + "epoch": 3.1056831586982168, + "grad_norm": 0.8612237572669983, + "learning_rate": 2.5092205321780482e-05, + "loss": 0.2541, + "step": 44756 + }, + { + "epoch": 3.105752550135313, + "grad_norm": 1.102055549621582, + "learning_rate": 2.508848320913042e-05, + "loss": 0.2037, + "step": 44757 + }, + { + "epoch": 3.10582194157241, + "grad_norm": 1.3084142208099365, + "learning_rate": 2.5084761332967287e-05, + "loss": 0.438, + "step": 44758 + }, + { + "epoch": 3.105891333009507, + "grad_norm": 1.2423737049102783, + "learning_rate": 2.5081039693302877e-05, + "loss": 0.2342, + "step": 44759 + }, + { + "epoch": 3.105960724446603, + "grad_norm": 1.236672282218933, + "learning_rate": 2.5077318290148944e-05, + "loss": 0.3645, + "step": 44760 + }, + { + "epoch": 3.1060301158837, + "grad_norm": 0.8456174731254578, + "learning_rate": 2.5073597123517233e-05, + "loss": 0.2773, + "step": 44761 + }, + { + "epoch": 3.1060995073207964, + "grad_norm": 1.0497468709945679, + "learning_rate": 2.5069876193419472e-05, + "loss": 0.2782, + "step": 44762 + }, + { + "epoch": 3.1061688987578933, + "grad_norm": 1.098435640335083, + "learning_rate": 2.5066155499867404e-05, + "loss": 0.3194, + "step": 44763 + }, + { + "epoch": 3.10623829019499, + "grad_norm": 0.9533299207687378, + "learning_rate": 2.50624350428728e-05, + "loss": 0.3386, + "step": 44764 + }, + { + "epoch": 3.1063076816320865, + "grad_norm": 0.9904155731201172, + "learning_rate": 2.5058714822447392e-05, + "loss": 0.2988, + "step": 44765 + }, + { + "epoch": 3.1063770730691833, + "grad_norm": 1.2861378192901611, + "learning_rate": 2.505499483860292e-05, + "loss": 0.3262, + "step": 44766 + }, + { + "epoch": 3.1064464645062797, + "grad_norm": 0.6879734396934509, + "learning_rate": 2.5051275091351145e-05, + "loss": 0.185, + "step": 44767 + }, + { + "epoch": 3.1065158559433765, + "grad_norm": 0.6951280832290649, + "learning_rate": 2.5047555580703817e-05, + "loss": 0.1695, + "step": 44768 + }, + { + "epoch": 3.1065852473804734, + "grad_norm": 0.7947139739990234, + "learning_rate": 2.504383630667264e-05, + "loss": 0.2333, + "step": 44769 + }, + { + "epoch": 3.1066546388175698, + "grad_norm": 0.9408588409423828, + "learning_rate": 2.504011726926937e-05, + "loss": 0.2321, + "step": 44770 + }, + { + "epoch": 3.1067240302546666, + "grad_norm": 0.8110026717185974, + "learning_rate": 2.5036398468505752e-05, + "loss": 0.2101, + "step": 44771 + }, + { + "epoch": 3.1067934216917634, + "grad_norm": 0.7445458173751831, + "learning_rate": 2.503267990439353e-05, + "loss": 0.2683, + "step": 44772 + }, + { + "epoch": 3.10686281312886, + "grad_norm": 0.8431615829467773, + "learning_rate": 2.502896157694443e-05, + "loss": 0.2196, + "step": 44773 + }, + { + "epoch": 3.1069322045659566, + "grad_norm": 0.7964254021644592, + "learning_rate": 2.5025243486170213e-05, + "loss": 0.1825, + "step": 44774 + }, + { + "epoch": 3.107001596003053, + "grad_norm": 0.9071550369262695, + "learning_rate": 2.5021525632082632e-05, + "loss": 0.2818, + "step": 44775 + }, + { + "epoch": 3.10707098744015, + "grad_norm": 0.8156747817993164, + "learning_rate": 2.5017808014693357e-05, + "loss": 0.1715, + "step": 44776 + }, + { + "epoch": 3.1071403788772467, + "grad_norm": 0.7215175628662109, + "learning_rate": 2.5014090634014164e-05, + "loss": 0.1538, + "step": 44777 + }, + { + "epoch": 3.107209770314343, + "grad_norm": 0.7246344685554504, + "learning_rate": 2.5010373490056793e-05, + "loss": 0.2368, + "step": 44778 + }, + { + "epoch": 3.10727916175144, + "grad_norm": 0.7242457270622253, + "learning_rate": 2.5006656582832978e-05, + "loss": 0.2752, + "step": 44779 + }, + { + "epoch": 3.1073485531885368, + "grad_norm": 1.4075862169265747, + "learning_rate": 2.5002939912354428e-05, + "loss": 0.3044, + "step": 44780 + }, + { + "epoch": 3.107417944625633, + "grad_norm": 0.6909558773040771, + "learning_rate": 2.4999223478632917e-05, + "loss": 0.2284, + "step": 44781 + }, + { + "epoch": 3.10748733606273, + "grad_norm": 0.7475515007972717, + "learning_rate": 2.4995507281680163e-05, + "loss": 0.2065, + "step": 44782 + }, + { + "epoch": 3.1075567274998264, + "grad_norm": 1.0996170043945312, + "learning_rate": 2.4991791321507875e-05, + "loss": 0.4696, + "step": 44783 + }, + { + "epoch": 3.107626118936923, + "grad_norm": 1.0109845399856567, + "learning_rate": 2.4988075598127803e-05, + "loss": 0.2999, + "step": 44784 + }, + { + "epoch": 3.10769551037402, + "grad_norm": 0.8053009510040283, + "learning_rate": 2.4984360111551662e-05, + "loss": 0.3708, + "step": 44785 + }, + { + "epoch": 3.1077649018111164, + "grad_norm": 1.0396424531936646, + "learning_rate": 2.4980644861791202e-05, + "loss": 0.3821, + "step": 44786 + }, + { + "epoch": 3.1078342932482133, + "grad_norm": 1.0700201988220215, + "learning_rate": 2.497692984885813e-05, + "loss": 0.2806, + "step": 44787 + }, + { + "epoch": 3.1079036846853096, + "grad_norm": 1.0676836967468262, + "learning_rate": 2.4973215072764234e-05, + "loss": 0.2788, + "step": 44788 + }, + { + "epoch": 3.1079730761224065, + "grad_norm": 0.7158902287483215, + "learning_rate": 2.496950053352115e-05, + "loss": 0.2022, + "step": 44789 + }, + { + "epoch": 3.1080424675595033, + "grad_norm": 0.9309860467910767, + "learning_rate": 2.4965786231140654e-05, + "loss": 0.2836, + "step": 44790 + }, + { + "epoch": 3.1081118589965997, + "grad_norm": 0.757843554019928, + "learning_rate": 2.4962072165634463e-05, + "loss": 0.1903, + "step": 44791 + }, + { + "epoch": 3.1081812504336965, + "grad_norm": 1.0140273571014404, + "learning_rate": 2.495835833701433e-05, + "loss": 0.3263, + "step": 44792 + }, + { + "epoch": 3.108250641870793, + "grad_norm": 0.9638352990150452, + "learning_rate": 2.4954644745291934e-05, + "loss": 0.306, + "step": 44793 + }, + { + "epoch": 3.1083200333078898, + "grad_norm": 0.8926337957382202, + "learning_rate": 2.4950931390479004e-05, + "loss": 0.3067, + "step": 44794 + }, + { + "epoch": 3.1083894247449866, + "grad_norm": 0.6613554954528809, + "learning_rate": 2.494721827258729e-05, + "loss": 0.2093, + "step": 44795 + }, + { + "epoch": 3.108458816182083, + "grad_norm": 0.8436132669448853, + "learning_rate": 2.494350539162852e-05, + "loss": 0.2101, + "step": 44796 + }, + { + "epoch": 3.10852820761918, + "grad_norm": 0.7180424928665161, + "learning_rate": 2.4939792747614365e-05, + "loss": 0.2117, + "step": 44797 + }, + { + "epoch": 3.1085975990562766, + "grad_norm": 1.0595396757125854, + "learning_rate": 2.4936080340556578e-05, + "loss": 0.2783, + "step": 44798 + }, + { + "epoch": 3.108666990493373, + "grad_norm": 0.6961773633956909, + "learning_rate": 2.493236817046688e-05, + "loss": 0.1783, + "step": 44799 + }, + { + "epoch": 3.10873638193047, + "grad_norm": 1.962947130203247, + "learning_rate": 2.4928656237356986e-05, + "loss": 0.3681, + "step": 44800 + }, + { + "epoch": 3.1088057733675662, + "grad_norm": 1.0453797578811646, + "learning_rate": 2.4924944541238616e-05, + "loss": 0.2541, + "step": 44801 + }, + { + "epoch": 3.108875164804663, + "grad_norm": 0.8685462474822998, + "learning_rate": 2.4921233082123474e-05, + "loss": 0.1388, + "step": 44802 + }, + { + "epoch": 3.10894455624176, + "grad_norm": 0.6467155814170837, + "learning_rate": 2.491752186002333e-05, + "loss": 0.1681, + "step": 44803 + }, + { + "epoch": 3.1090139476788563, + "grad_norm": 0.7936059832572937, + "learning_rate": 2.4913810874949828e-05, + "loss": 0.2258, + "step": 44804 + }, + { + "epoch": 3.109083339115953, + "grad_norm": 1.0019294023513794, + "learning_rate": 2.4910100126914705e-05, + "loss": 0.4957, + "step": 44805 + }, + { + "epoch": 3.10915273055305, + "grad_norm": 1.4591299295425415, + "learning_rate": 2.4906389615929693e-05, + "loss": 0.436, + "step": 44806 + }, + { + "epoch": 3.1092221219901464, + "grad_norm": 0.822303056716919, + "learning_rate": 2.4902679342006497e-05, + "loss": 0.3341, + "step": 44807 + }, + { + "epoch": 3.109291513427243, + "grad_norm": 0.7923442125320435, + "learning_rate": 2.4898969305156818e-05, + "loss": 0.3544, + "step": 44808 + }, + { + "epoch": 3.1093609048643396, + "grad_norm": 0.6814828515052795, + "learning_rate": 2.489525950539239e-05, + "loss": 0.1425, + "step": 44809 + }, + { + "epoch": 3.1094302963014364, + "grad_norm": 0.7652433514595032, + "learning_rate": 2.4891549942724934e-05, + "loss": 0.1616, + "step": 44810 + }, + { + "epoch": 3.1094996877385332, + "grad_norm": 1.2265033721923828, + "learning_rate": 2.4887840617166124e-05, + "loss": 0.4459, + "step": 44811 + }, + { + "epoch": 3.1095690791756296, + "grad_norm": 0.7977038025856018, + "learning_rate": 2.4884131528727682e-05, + "loss": 0.331, + "step": 44812 + }, + { + "epoch": 3.1096384706127265, + "grad_norm": 0.9479016065597534, + "learning_rate": 2.488042267742131e-05, + "loss": 0.2555, + "step": 44813 + }, + { + "epoch": 3.109707862049823, + "grad_norm": 1.0278925895690918, + "learning_rate": 2.4876714063258745e-05, + "loss": 0.2704, + "step": 44814 + }, + { + "epoch": 3.1097772534869197, + "grad_norm": 1.2887264490127563, + "learning_rate": 2.487300568625167e-05, + "loss": 0.349, + "step": 44815 + }, + { + "epoch": 3.1098466449240165, + "grad_norm": 1.1336053609848022, + "learning_rate": 2.486929754641183e-05, + "loss": 0.3432, + "step": 44816 + }, + { + "epoch": 3.109916036361113, + "grad_norm": 0.5393611192703247, + "learning_rate": 2.4865589643750865e-05, + "loss": 0.1557, + "step": 44817 + }, + { + "epoch": 3.1099854277982097, + "grad_norm": 0.6991551518440247, + "learning_rate": 2.486188197828052e-05, + "loss": 0.2155, + "step": 44818 + }, + { + "epoch": 3.1100548192353066, + "grad_norm": 1.0047107934951782, + "learning_rate": 2.4858174550012504e-05, + "loss": 0.225, + "step": 44819 + }, + { + "epoch": 3.110124210672403, + "grad_norm": 0.9151492118835449, + "learning_rate": 2.485446735895851e-05, + "loss": 0.3051, + "step": 44820 + }, + { + "epoch": 3.1101936021095, + "grad_norm": 1.0227940082550049, + "learning_rate": 2.4850760405130234e-05, + "loss": 0.2081, + "step": 44821 + }, + { + "epoch": 3.110262993546596, + "grad_norm": 0.8620362281799316, + "learning_rate": 2.48470536885394e-05, + "loss": 0.1495, + "step": 44822 + }, + { + "epoch": 3.110332384983693, + "grad_norm": 0.6617456674575806, + "learning_rate": 2.484334720919772e-05, + "loss": 0.1414, + "step": 44823 + }, + { + "epoch": 3.11040177642079, + "grad_norm": 0.7629582285881042, + "learning_rate": 2.483964096711685e-05, + "loss": 0.2328, + "step": 44824 + }, + { + "epoch": 3.1104711678578862, + "grad_norm": 0.8081070780754089, + "learning_rate": 2.4835934962308517e-05, + "loss": 0.194, + "step": 44825 + }, + { + "epoch": 3.110540559294983, + "grad_norm": 0.6594908833503723, + "learning_rate": 2.4832229194784407e-05, + "loss": 0.1212, + "step": 44826 + }, + { + "epoch": 3.1106099507320795, + "grad_norm": 0.9475570917129517, + "learning_rate": 2.4828523664556257e-05, + "loss": 0.2785, + "step": 44827 + }, + { + "epoch": 3.1106793421691763, + "grad_norm": 0.9414734840393066, + "learning_rate": 2.4824818371635715e-05, + "loss": 0.21, + "step": 44828 + }, + { + "epoch": 3.110748733606273, + "grad_norm": 0.8644090890884399, + "learning_rate": 2.4821113316034495e-05, + "loss": 0.1931, + "step": 44829 + }, + { + "epoch": 3.1108181250433695, + "grad_norm": 0.8523988127708435, + "learning_rate": 2.4817408497764328e-05, + "loss": 0.2378, + "step": 44830 + }, + { + "epoch": 3.1108875164804664, + "grad_norm": 0.6948854923248291, + "learning_rate": 2.481370391683685e-05, + "loss": 0.1898, + "step": 44831 + }, + { + "epoch": 3.110956907917563, + "grad_norm": 0.737903892993927, + "learning_rate": 2.4809999573263776e-05, + "loss": 0.2439, + "step": 44832 + }, + { + "epoch": 3.1110262993546596, + "grad_norm": 1.2115743160247803, + "learning_rate": 2.4806295467056816e-05, + "loss": 0.2685, + "step": 44833 + }, + { + "epoch": 3.1110956907917564, + "grad_norm": 1.1752036809921265, + "learning_rate": 2.4802591598227665e-05, + "loss": 0.3007, + "step": 44834 + }, + { + "epoch": 3.111165082228853, + "grad_norm": 0.6693901419639587, + "learning_rate": 2.479888796678801e-05, + "loss": 0.1858, + "step": 44835 + }, + { + "epoch": 3.1112344736659496, + "grad_norm": 0.9011726379394531, + "learning_rate": 2.479518457274953e-05, + "loss": 0.2701, + "step": 44836 + }, + { + "epoch": 3.1113038651030465, + "grad_norm": 0.6923406720161438, + "learning_rate": 2.4791481416123952e-05, + "loss": 0.1542, + "step": 44837 + }, + { + "epoch": 3.111373256540143, + "grad_norm": 0.7770766019821167, + "learning_rate": 2.478777849692292e-05, + "loss": 0.3757, + "step": 44838 + }, + { + "epoch": 3.1114426479772397, + "grad_norm": 0.7407844662666321, + "learning_rate": 2.478407581515815e-05, + "loss": 0.2535, + "step": 44839 + }, + { + "epoch": 3.111512039414336, + "grad_norm": 0.6872583031654358, + "learning_rate": 2.4780373370841314e-05, + "loss": 0.1719, + "step": 44840 + }, + { + "epoch": 3.111581430851433, + "grad_norm": 0.7855591177940369, + "learning_rate": 2.4776671163984122e-05, + "loss": 0.2484, + "step": 44841 + }, + { + "epoch": 3.1116508222885297, + "grad_norm": 0.9416021704673767, + "learning_rate": 2.477296919459825e-05, + "loss": 0.4475, + "step": 44842 + }, + { + "epoch": 3.111720213725626, + "grad_norm": 0.9774675965309143, + "learning_rate": 2.476926746269539e-05, + "loss": 0.3007, + "step": 44843 + }, + { + "epoch": 3.111789605162723, + "grad_norm": 0.6667593717575073, + "learning_rate": 2.476556596828724e-05, + "loss": 0.2053, + "step": 44844 + }, + { + "epoch": 3.11185899659982, + "grad_norm": 1.3116620779037476, + "learning_rate": 2.476186471138545e-05, + "loss": 0.2109, + "step": 44845 + }, + { + "epoch": 3.111928388036916, + "grad_norm": 1.4407696723937988, + "learning_rate": 2.4758163692001725e-05, + "loss": 0.4787, + "step": 44846 + }, + { + "epoch": 3.111997779474013, + "grad_norm": 0.9754564762115479, + "learning_rate": 2.4754462910147748e-05, + "loss": 0.3193, + "step": 44847 + }, + { + "epoch": 3.1120671709111094, + "grad_norm": 0.8560084104537964, + "learning_rate": 2.47507623658352e-05, + "loss": 0.2221, + "step": 44848 + }, + { + "epoch": 3.1121365623482062, + "grad_norm": 0.814560055732727, + "learning_rate": 2.4747062059075756e-05, + "loss": 0.1967, + "step": 44849 + }, + { + "epoch": 3.112205953785303, + "grad_norm": 1.4993375539779663, + "learning_rate": 2.474336198988112e-05, + "loss": 0.2402, + "step": 44850 + }, + { + "epoch": 3.1122753452223995, + "grad_norm": 1.0405522584915161, + "learning_rate": 2.473966215826298e-05, + "loss": 0.3167, + "step": 44851 + }, + { + "epoch": 3.1123447366594963, + "grad_norm": 0.7273461222648621, + "learning_rate": 2.473596256423296e-05, + "loss": 0.2445, + "step": 44852 + }, + { + "epoch": 3.1124141280965927, + "grad_norm": 0.7316051125526428, + "learning_rate": 2.4732263207802785e-05, + "loss": 0.1991, + "step": 44853 + }, + { + "epoch": 3.1124835195336895, + "grad_norm": 0.8223941326141357, + "learning_rate": 2.472856408898412e-05, + "loss": 0.1363, + "step": 44854 + }, + { + "epoch": 3.1125529109707863, + "grad_norm": 0.782708466053009, + "learning_rate": 2.472486520778864e-05, + "loss": 0.2348, + "step": 44855 + }, + { + "epoch": 3.1126223024078827, + "grad_norm": 0.8903416991233826, + "learning_rate": 2.4721166564228026e-05, + "loss": 0.1837, + "step": 44856 + }, + { + "epoch": 3.1126916938449796, + "grad_norm": 0.7066115736961365, + "learning_rate": 2.4717468158313962e-05, + "loss": 0.2012, + "step": 44857 + }, + { + "epoch": 3.1127610852820764, + "grad_norm": 0.7029722929000854, + "learning_rate": 2.4713769990058146e-05, + "loss": 0.2883, + "step": 44858 + }, + { + "epoch": 3.112830476719173, + "grad_norm": 1.1530576944351196, + "learning_rate": 2.4710072059472188e-05, + "loss": 0.349, + "step": 44859 + }, + { + "epoch": 3.1128998681562696, + "grad_norm": 0.8317326903343201, + "learning_rate": 2.470637436656781e-05, + "loss": 0.1937, + "step": 44860 + }, + { + "epoch": 3.112969259593366, + "grad_norm": 0.764397919178009, + "learning_rate": 2.470267691135666e-05, + "loss": 0.1596, + "step": 44861 + }, + { + "epoch": 3.113038651030463, + "grad_norm": 1.3609702587127686, + "learning_rate": 2.4698979693850455e-05, + "loss": 0.456, + "step": 44862 + }, + { + "epoch": 3.1131080424675597, + "grad_norm": 0.7497732043266296, + "learning_rate": 2.4695282714060797e-05, + "loss": 0.2098, + "step": 44863 + }, + { + "epoch": 3.113177433904656, + "grad_norm": 0.7694002985954285, + "learning_rate": 2.469158597199941e-05, + "loss": 0.2197, + "step": 44864 + }, + { + "epoch": 3.113246825341753, + "grad_norm": 0.8711977005004883, + "learning_rate": 2.468788946767797e-05, + "loss": 0.415, + "step": 44865 + }, + { + "epoch": 3.1133162167788493, + "grad_norm": 0.9445041418075562, + "learning_rate": 2.4684193201108096e-05, + "loss": 0.3262, + "step": 44866 + }, + { + "epoch": 3.113385608215946, + "grad_norm": 1.071089744567871, + "learning_rate": 2.4680497172301488e-05, + "loss": 0.3725, + "step": 44867 + }, + { + "epoch": 3.113454999653043, + "grad_norm": 0.8937728404998779, + "learning_rate": 2.467680138126981e-05, + "loss": 0.1778, + "step": 44868 + }, + { + "epoch": 3.1135243910901393, + "grad_norm": 1.0990853309631348, + "learning_rate": 2.4673105828024735e-05, + "loss": 0.261, + "step": 44869 + }, + { + "epoch": 3.113593782527236, + "grad_norm": 1.082881212234497, + "learning_rate": 2.4669410512577918e-05, + "loss": 0.2404, + "step": 44870 + }, + { + "epoch": 3.113663173964333, + "grad_norm": 0.8911322951316833, + "learning_rate": 2.4665715434941038e-05, + "loss": 0.2519, + "step": 44871 + }, + { + "epoch": 3.1137325654014294, + "grad_norm": 1.2801084518432617, + "learning_rate": 2.466202059512578e-05, + "loss": 0.3792, + "step": 44872 + }, + { + "epoch": 3.1138019568385262, + "grad_norm": 0.9648451805114746, + "learning_rate": 2.4658325993143748e-05, + "loss": 0.2188, + "step": 44873 + }, + { + "epoch": 3.1138713482756226, + "grad_norm": 0.7642826437950134, + "learning_rate": 2.4654631629006654e-05, + "loss": 0.2519, + "step": 44874 + }, + { + "epoch": 3.1139407397127195, + "grad_norm": 0.9370253682136536, + "learning_rate": 2.4650937502726136e-05, + "loss": 0.2411, + "step": 44875 + }, + { + "epoch": 3.1140101311498163, + "grad_norm": 1.8561822175979614, + "learning_rate": 2.4647243614313865e-05, + "loss": 0.4737, + "step": 44876 + }, + { + "epoch": 3.1140795225869127, + "grad_norm": 0.595680832862854, + "learning_rate": 2.464354996378151e-05, + "loss": 0.1742, + "step": 44877 + }, + { + "epoch": 3.1141489140240095, + "grad_norm": 0.8839362859725952, + "learning_rate": 2.4639856551140717e-05, + "loss": 0.2199, + "step": 44878 + }, + { + "epoch": 3.114218305461106, + "grad_norm": 1.1959071159362793, + "learning_rate": 2.463616337640319e-05, + "loss": 0.3372, + "step": 44879 + }, + { + "epoch": 3.1142876968982027, + "grad_norm": 1.2231096029281616, + "learning_rate": 2.4632470439580514e-05, + "loss": 0.374, + "step": 44880 + }, + { + "epoch": 3.1143570883352996, + "grad_norm": 0.7656000256538391, + "learning_rate": 2.4628777740684394e-05, + "loss": 0.2389, + "step": 44881 + }, + { + "epoch": 3.114426479772396, + "grad_norm": 1.179623007774353, + "learning_rate": 2.4625085279726466e-05, + "loss": 0.3538, + "step": 44882 + }, + { + "epoch": 3.114495871209493, + "grad_norm": 1.0712802410125732, + "learning_rate": 2.4621393056718412e-05, + "loss": 0.2473, + "step": 44883 + }, + { + "epoch": 3.1145652626465896, + "grad_norm": 1.217754602432251, + "learning_rate": 2.461770107167186e-05, + "loss": 0.3131, + "step": 44884 + }, + { + "epoch": 3.114634654083686, + "grad_norm": 1.0915744304656982, + "learning_rate": 2.46140093245985e-05, + "loss": 0.3045, + "step": 44885 + }, + { + "epoch": 3.114704045520783, + "grad_norm": 1.3811084032058716, + "learning_rate": 2.4610317815509976e-05, + "loss": 0.461, + "step": 44886 + }, + { + "epoch": 3.1147734369578792, + "grad_norm": 0.9591004252433777, + "learning_rate": 2.460662654441791e-05, + "loss": 0.3777, + "step": 44887 + }, + { + "epoch": 3.114842828394976, + "grad_norm": 0.77860027551651, + "learning_rate": 2.460293551133398e-05, + "loss": 0.264, + "step": 44888 + }, + { + "epoch": 3.114912219832073, + "grad_norm": 0.9850372076034546, + "learning_rate": 2.459924471626983e-05, + "loss": 0.2742, + "step": 44889 + }, + { + "epoch": 3.1149816112691693, + "grad_norm": 0.9483935236930847, + "learning_rate": 2.459555415923712e-05, + "loss": 0.2684, + "step": 44890 + }, + { + "epoch": 3.115051002706266, + "grad_norm": 0.8139939904212952, + "learning_rate": 2.4591863840247498e-05, + "loss": 0.2745, + "step": 44891 + }, + { + "epoch": 3.1151203941433625, + "grad_norm": 1.2746987342834473, + "learning_rate": 2.458817375931264e-05, + "loss": 0.4178, + "step": 44892 + }, + { + "epoch": 3.1151897855804593, + "grad_norm": 1.1138206720352173, + "learning_rate": 2.4584483916444134e-05, + "loss": 0.2686, + "step": 44893 + }, + { + "epoch": 3.115259177017556, + "grad_norm": 1.1721556186676025, + "learning_rate": 2.4580794311653664e-05, + "loss": 0.5023, + "step": 44894 + }, + { + "epoch": 3.1153285684546526, + "grad_norm": 0.7576115727424622, + "learning_rate": 2.457710494495289e-05, + "loss": 0.2053, + "step": 44895 + }, + { + "epoch": 3.1153979598917494, + "grad_norm": 0.9176598191261292, + "learning_rate": 2.4573415816353463e-05, + "loss": 0.3152, + "step": 44896 + }, + { + "epoch": 3.1154673513288462, + "grad_norm": 1.4692108631134033, + "learning_rate": 2.4569726925866975e-05, + "loss": 0.3625, + "step": 44897 + }, + { + "epoch": 3.1155367427659426, + "grad_norm": 0.7714634537696838, + "learning_rate": 2.456603827350512e-05, + "loss": 0.1525, + "step": 44898 + }, + { + "epoch": 3.1156061342030394, + "grad_norm": 1.3555684089660645, + "learning_rate": 2.456234985927952e-05, + "loss": 0.255, + "step": 44899 + }, + { + "epoch": 3.115675525640136, + "grad_norm": 0.7865859866142273, + "learning_rate": 2.4558661683201867e-05, + "loss": 0.1977, + "step": 44900 + }, + { + "epoch": 3.1157449170772327, + "grad_norm": 0.8625088334083557, + "learning_rate": 2.455497374528373e-05, + "loss": 0.2303, + "step": 44901 + }, + { + "epoch": 3.1158143085143295, + "grad_norm": 0.933490514755249, + "learning_rate": 2.455128604553679e-05, + "loss": 0.2921, + "step": 44902 + }, + { + "epoch": 3.115883699951426, + "grad_norm": 0.9576590061187744, + "learning_rate": 2.4547598583972686e-05, + "loss": 0.3356, + "step": 44903 + }, + { + "epoch": 3.1159530913885227, + "grad_norm": 0.7188206315040588, + "learning_rate": 2.4543911360603068e-05, + "loss": 0.1884, + "step": 44904 + }, + { + "epoch": 3.116022482825619, + "grad_norm": 0.708088755607605, + "learning_rate": 2.454022437543956e-05, + "loss": 0.2455, + "step": 44905 + }, + { + "epoch": 3.116091874262716, + "grad_norm": 1.1097766160964966, + "learning_rate": 2.4536537628493828e-05, + "loss": 0.471, + "step": 44906 + }, + { + "epoch": 3.1161612656998128, + "grad_norm": 0.5698590278625488, + "learning_rate": 2.453285111977748e-05, + "loss": 0.2038, + "step": 44907 + }, + { + "epoch": 3.116230657136909, + "grad_norm": 1.0050806999206543, + "learning_rate": 2.4529164849302156e-05, + "loss": 0.2747, + "step": 44908 + }, + { + "epoch": 3.116300048574006, + "grad_norm": 0.6557195782661438, + "learning_rate": 2.4525478817079495e-05, + "loss": 0.2386, + "step": 44909 + }, + { + "epoch": 3.116369440011103, + "grad_norm": 1.4906749725341797, + "learning_rate": 2.4521793023121154e-05, + "loss": 0.2299, + "step": 44910 + }, + { + "epoch": 3.116438831448199, + "grad_norm": 0.7363883852958679, + "learning_rate": 2.4518107467438744e-05, + "loss": 0.2305, + "step": 44911 + }, + { + "epoch": 3.116508222885296, + "grad_norm": 0.8903813362121582, + "learning_rate": 2.4514422150043925e-05, + "loss": 0.2602, + "step": 44912 + }, + { + "epoch": 3.1165776143223924, + "grad_norm": 0.7415410280227661, + "learning_rate": 2.451073707094834e-05, + "loss": 0.309, + "step": 44913 + }, + { + "epoch": 3.1166470057594893, + "grad_norm": 1.0516128540039062, + "learning_rate": 2.450705223016356e-05, + "loss": 0.3948, + "step": 44914 + }, + { + "epoch": 3.116716397196586, + "grad_norm": 0.8425201773643494, + "learning_rate": 2.450336762770128e-05, + "loss": 0.2019, + "step": 44915 + }, + { + "epoch": 3.1167857886336825, + "grad_norm": 0.6041681170463562, + "learning_rate": 2.449968326357309e-05, + "loss": 0.1877, + "step": 44916 + }, + { + "epoch": 3.1168551800707793, + "grad_norm": 0.6441674828529358, + "learning_rate": 2.449599913779065e-05, + "loss": 0.226, + "step": 44917 + }, + { + "epoch": 3.1169245715078757, + "grad_norm": 0.9697873592376709, + "learning_rate": 2.4492315250365572e-05, + "loss": 0.1925, + "step": 44918 + }, + { + "epoch": 3.1169939629449726, + "grad_norm": 0.7037637829780579, + "learning_rate": 2.44886316013095e-05, + "loss": 0.1807, + "step": 44919 + }, + { + "epoch": 3.1170633543820694, + "grad_norm": 1.116782784461975, + "learning_rate": 2.4484948190634083e-05, + "loss": 0.3145, + "step": 44920 + }, + { + "epoch": 3.1171327458191658, + "grad_norm": 1.398970603942871, + "learning_rate": 2.4481265018350897e-05, + "loss": 0.3129, + "step": 44921 + }, + { + "epoch": 3.1172021372562626, + "grad_norm": 0.8177086114883423, + "learning_rate": 2.4477582084471596e-05, + "loss": 0.2452, + "step": 44922 + }, + { + "epoch": 3.1172715286933594, + "grad_norm": 1.1555602550506592, + "learning_rate": 2.4473899389007803e-05, + "loss": 0.2271, + "step": 44923 + }, + { + "epoch": 3.117340920130456, + "grad_norm": 1.009432077407837, + "learning_rate": 2.4470216931971157e-05, + "loss": 0.1873, + "step": 44924 + }, + { + "epoch": 3.1174103115675527, + "grad_norm": 0.8737172484397888, + "learning_rate": 2.4466534713373256e-05, + "loss": 0.2682, + "step": 44925 + }, + { + "epoch": 3.117479703004649, + "grad_norm": 0.8244720101356506, + "learning_rate": 2.446285273322575e-05, + "loss": 0.2253, + "step": 44926 + }, + { + "epoch": 3.117549094441746, + "grad_norm": 1.3401354551315308, + "learning_rate": 2.445917099154029e-05, + "loss": 0.2211, + "step": 44927 + }, + { + "epoch": 3.1176184858788427, + "grad_norm": 0.9820684790611267, + "learning_rate": 2.445548948832842e-05, + "loss": 0.3379, + "step": 44928 + }, + { + "epoch": 3.117687877315939, + "grad_norm": 0.6155511140823364, + "learning_rate": 2.4451808223601814e-05, + "loss": 0.1486, + "step": 44929 + }, + { + "epoch": 3.117757268753036, + "grad_norm": 0.848659336566925, + "learning_rate": 2.4448127197372085e-05, + "loss": 0.2353, + "step": 44930 + }, + { + "epoch": 3.1178266601901323, + "grad_norm": 0.6959164142608643, + "learning_rate": 2.4444446409650878e-05, + "loss": 0.1598, + "step": 44931 + }, + { + "epoch": 3.117896051627229, + "grad_norm": 1.1916813850402832, + "learning_rate": 2.4440765860449765e-05, + "loss": 0.2592, + "step": 44932 + }, + { + "epoch": 3.117965443064326, + "grad_norm": 0.6877869367599487, + "learning_rate": 2.443708554978038e-05, + "loss": 0.1773, + "step": 44933 + }, + { + "epoch": 3.1180348345014224, + "grad_norm": 0.7310125827789307, + "learning_rate": 2.4433405477654346e-05, + "loss": 0.2008, + "step": 44934 + }, + { + "epoch": 3.118104225938519, + "grad_norm": 0.9521505832672119, + "learning_rate": 2.4429725644083324e-05, + "loss": 0.2811, + "step": 44935 + }, + { + "epoch": 3.118173617375616, + "grad_norm": 0.9545972347259521, + "learning_rate": 2.4426046049078853e-05, + "loss": 0.3035, + "step": 44936 + }, + { + "epoch": 3.1182430088127124, + "grad_norm": 0.8287777900695801, + "learning_rate": 2.442236669265259e-05, + "loss": 0.2273, + "step": 44937 + }, + { + "epoch": 3.1183124002498093, + "grad_norm": 0.7218388319015503, + "learning_rate": 2.4418687574816146e-05, + "loss": 0.1901, + "step": 44938 + }, + { + "epoch": 3.1183817916869057, + "grad_norm": 1.3548407554626465, + "learning_rate": 2.4415008695581133e-05, + "loss": 0.3515, + "step": 44939 + }, + { + "epoch": 3.1184511831240025, + "grad_norm": 1.1554720401763916, + "learning_rate": 2.441133005495918e-05, + "loss": 0.2245, + "step": 44940 + }, + { + "epoch": 3.1185205745610993, + "grad_norm": 1.0673645734786987, + "learning_rate": 2.4407651652961906e-05, + "loss": 0.3453, + "step": 44941 + }, + { + "epoch": 3.1185899659981957, + "grad_norm": 1.0667932033538818, + "learning_rate": 2.4403973489600874e-05, + "loss": 0.2707, + "step": 44942 + }, + { + "epoch": 3.1186593574352925, + "grad_norm": 0.9301247000694275, + "learning_rate": 2.4400295564887733e-05, + "loss": 0.2075, + "step": 44943 + }, + { + "epoch": 3.118728748872389, + "grad_norm": 0.6507503390312195, + "learning_rate": 2.4396617878834084e-05, + "loss": 0.2537, + "step": 44944 + }, + { + "epoch": 3.1187981403094858, + "grad_norm": 0.9739923477172852, + "learning_rate": 2.439294043145155e-05, + "loss": 0.3051, + "step": 44945 + }, + { + "epoch": 3.1188675317465826, + "grad_norm": 0.9276453256607056, + "learning_rate": 2.4389263222751724e-05, + "loss": 0.2222, + "step": 44946 + }, + { + "epoch": 3.118936923183679, + "grad_norm": 0.7615301609039307, + "learning_rate": 2.4385586252746216e-05, + "loss": 0.2191, + "step": 44947 + }, + { + "epoch": 3.119006314620776, + "grad_norm": 1.074172854423523, + "learning_rate": 2.4381909521446678e-05, + "loss": 0.2503, + "step": 44948 + }, + { + "epoch": 3.1190757060578727, + "grad_norm": 0.7345530390739441, + "learning_rate": 2.437823302886464e-05, + "loss": 0.19, + "step": 44949 + }, + { + "epoch": 3.119145097494969, + "grad_norm": 0.8946821689605713, + "learning_rate": 2.437455677501176e-05, + "loss": 0.2219, + "step": 44950 + }, + { + "epoch": 3.119214488932066, + "grad_norm": 0.9256985187530518, + "learning_rate": 2.4370880759899616e-05, + "loss": 0.2987, + "step": 44951 + }, + { + "epoch": 3.1192838803691623, + "grad_norm": 0.9370371699333191, + "learning_rate": 2.436720498353984e-05, + "loss": 0.2413, + "step": 44952 + }, + { + "epoch": 3.119353271806259, + "grad_norm": 1.0422804355621338, + "learning_rate": 2.436352944594401e-05, + "loss": 0.2518, + "step": 44953 + }, + { + "epoch": 3.119422663243356, + "grad_norm": 0.7509401440620422, + "learning_rate": 2.435985414712374e-05, + "loss": 0.2344, + "step": 44954 + }, + { + "epoch": 3.1194920546804523, + "grad_norm": 0.7362405061721802, + "learning_rate": 2.435617908709068e-05, + "loss": 0.3026, + "step": 44955 + }, + { + "epoch": 3.119561446117549, + "grad_norm": 0.7843990325927734, + "learning_rate": 2.4352504265856346e-05, + "loss": 0.2405, + "step": 44956 + }, + { + "epoch": 3.119630837554646, + "grad_norm": 0.6149212121963501, + "learning_rate": 2.4348829683432385e-05, + "loss": 0.1585, + "step": 44957 + }, + { + "epoch": 3.1197002289917424, + "grad_norm": 0.7372169494628906, + "learning_rate": 2.4345155339830383e-05, + "loss": 0.1826, + "step": 44958 + }, + { + "epoch": 3.119769620428839, + "grad_norm": 0.6522491574287415, + "learning_rate": 2.434148123506196e-05, + "loss": 0.1844, + "step": 44959 + }, + { + "epoch": 3.1198390118659356, + "grad_norm": 0.780413806438446, + "learning_rate": 2.4337807369138688e-05, + "loss": 0.2029, + "step": 44960 + }, + { + "epoch": 3.1199084033030324, + "grad_norm": 1.6821556091308594, + "learning_rate": 2.4334133742072184e-05, + "loss": 0.2555, + "step": 44961 + }, + { + "epoch": 3.1199777947401293, + "grad_norm": 0.7736489176750183, + "learning_rate": 2.4330460353874073e-05, + "loss": 0.2721, + "step": 44962 + }, + { + "epoch": 3.1200471861772257, + "grad_norm": 0.7937049269676208, + "learning_rate": 2.4326787204555902e-05, + "loss": 0.2405, + "step": 44963 + }, + { + "epoch": 3.1201165776143225, + "grad_norm": 0.6858432292938232, + "learning_rate": 2.4323114294129266e-05, + "loss": 0.1957, + "step": 44964 + }, + { + "epoch": 3.120185969051419, + "grad_norm": 0.8599296808242798, + "learning_rate": 2.4319441622605786e-05, + "loss": 0.1594, + "step": 44965 + }, + { + "epoch": 3.1202553604885157, + "grad_norm": 0.7799213528633118, + "learning_rate": 2.4315769189997073e-05, + "loss": 0.2182, + "step": 44966 + }, + { + "epoch": 3.1203247519256125, + "grad_norm": 0.8695921897888184, + "learning_rate": 2.431209699631467e-05, + "loss": 0.2711, + "step": 44967 + }, + { + "epoch": 3.120394143362709, + "grad_norm": 0.6746662259101868, + "learning_rate": 2.4308425041570203e-05, + "loss": 0.2221, + "step": 44968 + }, + { + "epoch": 3.1204635347998058, + "grad_norm": 1.6813961267471313, + "learning_rate": 2.430475332577524e-05, + "loss": 0.3032, + "step": 44969 + }, + { + "epoch": 3.120532926236902, + "grad_norm": 0.9337809085845947, + "learning_rate": 2.4301081848941433e-05, + "loss": 0.3373, + "step": 44970 + }, + { + "epoch": 3.120602317673999, + "grad_norm": 0.8905223608016968, + "learning_rate": 2.4297410611080284e-05, + "loss": 0.2212, + "step": 44971 + }, + { + "epoch": 3.120671709111096, + "grad_norm": 0.8344375491142273, + "learning_rate": 2.4293739612203438e-05, + "loss": 0.2123, + "step": 44972 + }, + { + "epoch": 3.120741100548192, + "grad_norm": 1.1812855005264282, + "learning_rate": 2.4290068852322467e-05, + "loss": 0.3459, + "step": 44973 + }, + { + "epoch": 3.120810491985289, + "grad_norm": 0.7788622379302979, + "learning_rate": 2.428639833144897e-05, + "loss": 0.323, + "step": 44974 + }, + { + "epoch": 3.120879883422386, + "grad_norm": 0.9576297998428345, + "learning_rate": 2.4282728049594562e-05, + "loss": 0.2427, + "step": 44975 + }, + { + "epoch": 3.1209492748594823, + "grad_norm": 0.8696954250335693, + "learning_rate": 2.427905800677075e-05, + "loss": 0.2435, + "step": 44976 + }, + { + "epoch": 3.121018666296579, + "grad_norm": 0.8629331588745117, + "learning_rate": 2.427538820298918e-05, + "loss": 0.3213, + "step": 44977 + }, + { + "epoch": 3.1210880577336755, + "grad_norm": 0.8237106204032898, + "learning_rate": 2.427171863826142e-05, + "loss": 0.182, + "step": 44978 + }, + { + "epoch": 3.1211574491707723, + "grad_norm": 0.6588080525398254, + "learning_rate": 2.426804931259906e-05, + "loss": 0.1341, + "step": 44979 + }, + { + "epoch": 3.121226840607869, + "grad_norm": 0.8844638466835022, + "learning_rate": 2.4264380226013682e-05, + "loss": 0.2336, + "step": 44980 + }, + { + "epoch": 3.1212962320449655, + "grad_norm": 1.0840916633605957, + "learning_rate": 2.4260711378516865e-05, + "loss": 0.3939, + "step": 44981 + }, + { + "epoch": 3.1213656234820624, + "grad_norm": 1.1151262521743774, + "learning_rate": 2.425704277012023e-05, + "loss": 0.3323, + "step": 44982 + }, + { + "epoch": 3.121435014919159, + "grad_norm": 0.7440863847732544, + "learning_rate": 2.4253374400835292e-05, + "loss": 0.2041, + "step": 44983 + }, + { + "epoch": 3.1215044063562556, + "grad_norm": 0.9149256348609924, + "learning_rate": 2.4249706270673656e-05, + "loss": 0.2174, + "step": 44984 + }, + { + "epoch": 3.1215737977933524, + "grad_norm": 0.8580596446990967, + "learning_rate": 2.4246038379646918e-05, + "loss": 0.2559, + "step": 44985 + }, + { + "epoch": 3.121643189230449, + "grad_norm": 0.8517102003097534, + "learning_rate": 2.424237072776664e-05, + "loss": 0.3085, + "step": 44986 + }, + { + "epoch": 3.1217125806675456, + "grad_norm": 1.6067371368408203, + "learning_rate": 2.4238703315044408e-05, + "loss": 0.2377, + "step": 44987 + }, + { + "epoch": 3.1217819721046425, + "grad_norm": 0.8048750758171082, + "learning_rate": 2.4235036141491808e-05, + "loss": 0.3042, + "step": 44988 + }, + { + "epoch": 3.121851363541739, + "grad_norm": 0.5736500024795532, + "learning_rate": 2.4231369207120435e-05, + "loss": 0.2429, + "step": 44989 + }, + { + "epoch": 3.1219207549788357, + "grad_norm": 0.8498724102973938, + "learning_rate": 2.4227702511941808e-05, + "loss": 0.1894, + "step": 44990 + }, + { + "epoch": 3.121990146415932, + "grad_norm": 0.8858970999717712, + "learning_rate": 2.4224036055967535e-05, + "loss": 0.3557, + "step": 44991 + }, + { + "epoch": 3.122059537853029, + "grad_norm": 0.9337400794029236, + "learning_rate": 2.422036983920918e-05, + "loss": 0.2495, + "step": 44992 + }, + { + "epoch": 3.1221289292901258, + "grad_norm": 1.6047401428222656, + "learning_rate": 2.4216703861678336e-05, + "loss": 0.4525, + "step": 44993 + }, + { + "epoch": 3.122198320727222, + "grad_norm": 0.9986638426780701, + "learning_rate": 2.4213038123386555e-05, + "loss": 0.2189, + "step": 44994 + }, + { + "epoch": 3.122267712164319, + "grad_norm": 0.6392764449119568, + "learning_rate": 2.4209372624345426e-05, + "loss": 0.2059, + "step": 44995 + }, + { + "epoch": 3.1223371036014154, + "grad_norm": 1.186018466949463, + "learning_rate": 2.4205707364566545e-05, + "loss": 0.3328, + "step": 44996 + }, + { + "epoch": 3.122406495038512, + "grad_norm": 0.8588526844978333, + "learning_rate": 2.4202042344061427e-05, + "loss": 0.2863, + "step": 44997 + }, + { + "epoch": 3.122475886475609, + "grad_norm": 0.7587540149688721, + "learning_rate": 2.4198377562841656e-05, + "loss": 0.2938, + "step": 44998 + }, + { + "epoch": 3.1225452779127054, + "grad_norm": 0.9151185154914856, + "learning_rate": 2.4194713020918813e-05, + "loss": 0.2603, + "step": 44999 + }, + { + "epoch": 3.1226146693498023, + "grad_norm": 1.10377037525177, + "learning_rate": 2.4191048718304475e-05, + "loss": 0.2829, + "step": 45000 + }, + { + "epoch": 3.122684060786899, + "grad_norm": 0.8294756412506104, + "learning_rate": 2.4187384655010227e-05, + "loss": 0.2146, + "step": 45001 + }, + { + "epoch": 3.1227534522239955, + "grad_norm": 0.5551549196243286, + "learning_rate": 2.4183720831047574e-05, + "loss": 0.2396, + "step": 45002 + }, + { + "epoch": 3.1228228436610923, + "grad_norm": 0.949446439743042, + "learning_rate": 2.4180057246428124e-05, + "loss": 0.3377, + "step": 45003 + }, + { + "epoch": 3.1228922350981887, + "grad_norm": 1.3131836652755737, + "learning_rate": 2.4176393901163464e-05, + "loss": 0.3762, + "step": 45004 + }, + { + "epoch": 3.1229616265352855, + "grad_norm": 0.9405927062034607, + "learning_rate": 2.4172730795265108e-05, + "loss": 0.2141, + "step": 45005 + }, + { + "epoch": 3.1230310179723824, + "grad_norm": 0.9663246273994446, + "learning_rate": 2.416906792874464e-05, + "loss": 0.2105, + "step": 45006 + }, + { + "epoch": 3.1231004094094788, + "grad_norm": 1.0888375043869019, + "learning_rate": 2.416540530161362e-05, + "loss": 0.2646, + "step": 45007 + }, + { + "epoch": 3.1231698008465756, + "grad_norm": 0.5511639714241028, + "learning_rate": 2.4161742913883624e-05, + "loss": 0.1504, + "step": 45008 + }, + { + "epoch": 3.1232391922836724, + "grad_norm": 1.187135934829712, + "learning_rate": 2.4158080765566203e-05, + "loss": 0.2541, + "step": 45009 + }, + { + "epoch": 3.123308583720769, + "grad_norm": 1.032984972000122, + "learning_rate": 2.4154418856672954e-05, + "loss": 0.2099, + "step": 45010 + }, + { + "epoch": 3.1233779751578656, + "grad_norm": 0.9575166702270508, + "learning_rate": 2.4150757187215367e-05, + "loss": 0.3272, + "step": 45011 + }, + { + "epoch": 3.123447366594962, + "grad_norm": 0.8055926561355591, + "learning_rate": 2.414709575720505e-05, + "loss": 0.2275, + "step": 45012 + }, + { + "epoch": 3.123516758032059, + "grad_norm": 0.7481327652931213, + "learning_rate": 2.4143434566653545e-05, + "loss": 0.1178, + "step": 45013 + }, + { + "epoch": 3.1235861494691557, + "grad_norm": 1.007185697555542, + "learning_rate": 2.4139773615572424e-05, + "loss": 0.1725, + "step": 45014 + }, + { + "epoch": 3.123655540906252, + "grad_norm": 1.230092167854309, + "learning_rate": 2.413611290397323e-05, + "loss": 0.264, + "step": 45015 + }, + { + "epoch": 3.123724932343349, + "grad_norm": 0.726970911026001, + "learning_rate": 2.4132452431867526e-05, + "loss": 0.3302, + "step": 45016 + }, + { + "epoch": 3.1237943237804453, + "grad_norm": 0.600215494632721, + "learning_rate": 2.4128792199266893e-05, + "loss": 0.1629, + "step": 45017 + }, + { + "epoch": 3.123863715217542, + "grad_norm": 0.8584138751029968, + "learning_rate": 2.4125132206182844e-05, + "loss": 0.3522, + "step": 45018 + }, + { + "epoch": 3.123933106654639, + "grad_norm": 0.8704333901405334, + "learning_rate": 2.4121472452626935e-05, + "loss": 0.2199, + "step": 45019 + }, + { + "epoch": 3.1240024980917354, + "grad_norm": 0.8544483184814453, + "learning_rate": 2.411781293861074e-05, + "loss": 0.2811, + "step": 45020 + }, + { + "epoch": 3.124071889528832, + "grad_norm": 0.8918529748916626, + "learning_rate": 2.411415366414581e-05, + "loss": 0.2327, + "step": 45021 + }, + { + "epoch": 3.1241412809659286, + "grad_norm": 0.7587891221046448, + "learning_rate": 2.4110494629243686e-05, + "loss": 0.2908, + "step": 45022 + }, + { + "epoch": 3.1242106724030254, + "grad_norm": 0.7525522112846375, + "learning_rate": 2.410683583391593e-05, + "loss": 0.2121, + "step": 45023 + }, + { + "epoch": 3.1242800638401222, + "grad_norm": 1.2181748151779175, + "learning_rate": 2.4103177278174118e-05, + "loss": 0.3017, + "step": 45024 + }, + { + "epoch": 3.1243494552772186, + "grad_norm": 0.6294850707054138, + "learning_rate": 2.4099518962029733e-05, + "loss": 0.1957, + "step": 45025 + }, + { + "epoch": 3.1244188467143155, + "grad_norm": 0.779677152633667, + "learning_rate": 2.409586088549436e-05, + "loss": 0.1542, + "step": 45026 + }, + { + "epoch": 3.1244882381514123, + "grad_norm": 0.8918408751487732, + "learning_rate": 2.4092203048579553e-05, + "loss": 0.254, + "step": 45027 + }, + { + "epoch": 3.1245576295885087, + "grad_norm": 1.1669739484786987, + "learning_rate": 2.408854545129685e-05, + "loss": 0.3951, + "step": 45028 + }, + { + "epoch": 3.1246270210256055, + "grad_norm": 0.8355004191398621, + "learning_rate": 2.408488809365781e-05, + "loss": 0.2935, + "step": 45029 + }, + { + "epoch": 3.124696412462702, + "grad_norm": 1.146567702293396, + "learning_rate": 2.4081230975673963e-05, + "loss": 0.2502, + "step": 45030 + }, + { + "epoch": 3.1247658038997987, + "grad_norm": 1.1428143978118896, + "learning_rate": 2.4077574097356893e-05, + "loss": 0.3343, + "step": 45031 + }, + { + "epoch": 3.1248351953368956, + "grad_norm": 1.0526714324951172, + "learning_rate": 2.407391745871809e-05, + "loss": 0.3222, + "step": 45032 + }, + { + "epoch": 3.124904586773992, + "grad_norm": 0.7242319583892822, + "learning_rate": 2.4070261059769117e-05, + "loss": 0.263, + "step": 45033 + }, + { + "epoch": 3.124973978211089, + "grad_norm": 1.005088448524475, + "learning_rate": 2.4066604900521518e-05, + "loss": 0.4486, + "step": 45034 + }, + { + "epoch": 3.1250433696481856, + "grad_norm": 1.0443878173828125, + "learning_rate": 2.406294898098683e-05, + "loss": 0.3177, + "step": 45035 + }, + { + "epoch": 3.125112761085282, + "grad_norm": 0.7385225296020508, + "learning_rate": 2.4059293301176645e-05, + "loss": 0.2047, + "step": 45036 + }, + { + "epoch": 3.125182152522379, + "grad_norm": 1.5238006114959717, + "learning_rate": 2.405563786110242e-05, + "loss": 0.2929, + "step": 45037 + }, + { + "epoch": 3.1252515439594752, + "grad_norm": 0.6548656225204468, + "learning_rate": 2.4051982660775742e-05, + "loss": 0.2279, + "step": 45038 + }, + { + "epoch": 3.125320935396572, + "grad_norm": 0.60373854637146, + "learning_rate": 2.4048327700208173e-05, + "loss": 0.1637, + "step": 45039 + }, + { + "epoch": 3.125390326833669, + "grad_norm": 1.1007020473480225, + "learning_rate": 2.4044672979411187e-05, + "loss": 0.3084, + "step": 45040 + }, + { + "epoch": 3.1254597182707653, + "grad_norm": 0.862251877784729, + "learning_rate": 2.404101849839635e-05, + "loss": 0.2363, + "step": 45041 + }, + { + "epoch": 3.125529109707862, + "grad_norm": 1.0092087984085083, + "learning_rate": 2.403736425717522e-05, + "loss": 0.2748, + "step": 45042 + }, + { + "epoch": 3.1255985011449585, + "grad_norm": 0.7745072245597839, + "learning_rate": 2.4033710255759312e-05, + "loss": 0.1879, + "step": 45043 + }, + { + "epoch": 3.1256678925820554, + "grad_norm": 0.96878582239151, + "learning_rate": 2.403005649416016e-05, + "loss": 0.2477, + "step": 45044 + }, + { + "epoch": 3.125737284019152, + "grad_norm": 0.9234841465950012, + "learning_rate": 2.402640297238934e-05, + "loss": 0.271, + "step": 45045 + }, + { + "epoch": 3.1258066754562486, + "grad_norm": 0.7660887241363525, + "learning_rate": 2.4022749690458324e-05, + "loss": 0.1821, + "step": 45046 + }, + { + "epoch": 3.1258760668933454, + "grad_norm": 0.7694817781448364, + "learning_rate": 2.401909664837867e-05, + "loss": 0.2207, + "step": 45047 + }, + { + "epoch": 3.125945458330442, + "grad_norm": 0.7605868577957153, + "learning_rate": 2.401544384616191e-05, + "loss": 0.2473, + "step": 45048 + }, + { + "epoch": 3.1260148497675386, + "grad_norm": 1.0504578351974487, + "learning_rate": 2.4011791283819583e-05, + "loss": 0.1106, + "step": 45049 + }, + { + "epoch": 3.1260842412046355, + "grad_norm": 0.7420949935913086, + "learning_rate": 2.4008138961363214e-05, + "loss": 0.128, + "step": 45050 + }, + { + "epoch": 3.126153632641732, + "grad_norm": 1.261633038520813, + "learning_rate": 2.4004486878804366e-05, + "loss": 0.2832, + "step": 45051 + }, + { + "epoch": 3.1262230240788287, + "grad_norm": 0.9945547580718994, + "learning_rate": 2.40008350361545e-05, + "loss": 0.332, + "step": 45052 + }, + { + "epoch": 3.1262924155159255, + "grad_norm": 1.0929133892059326, + "learning_rate": 2.3997183433425185e-05, + "loss": 0.3333, + "step": 45053 + }, + { + "epoch": 3.126361806953022, + "grad_norm": 0.8648033142089844, + "learning_rate": 2.3993532070627932e-05, + "loss": 0.278, + "step": 45054 + }, + { + "epoch": 3.1264311983901187, + "grad_norm": 0.9483857750892639, + "learning_rate": 2.39898809477743e-05, + "loss": 0.2623, + "step": 45055 + }, + { + "epoch": 3.126500589827215, + "grad_norm": 0.6743650436401367, + "learning_rate": 2.398623006487578e-05, + "loss": 0.3692, + "step": 45056 + }, + { + "epoch": 3.126569981264312, + "grad_norm": 1.0608419179916382, + "learning_rate": 2.398257942194392e-05, + "loss": 0.2829, + "step": 45057 + }, + { + "epoch": 3.126639372701409, + "grad_norm": 0.6763004660606384, + "learning_rate": 2.397892901899026e-05, + "loss": 0.2151, + "step": 45058 + }, + { + "epoch": 3.126708764138505, + "grad_norm": 0.925472617149353, + "learning_rate": 2.3975278856026273e-05, + "loss": 0.2298, + "step": 45059 + }, + { + "epoch": 3.126778155575602, + "grad_norm": 0.7074524760246277, + "learning_rate": 2.3971628933063518e-05, + "loss": 0.1751, + "step": 45060 + }, + { + "epoch": 3.126847547012699, + "grad_norm": 0.9533671140670776, + "learning_rate": 2.3967979250113502e-05, + "loss": 0.2677, + "step": 45061 + }, + { + "epoch": 3.1269169384497952, + "grad_norm": 0.8861160278320312, + "learning_rate": 2.3964329807187758e-05, + "loss": 0.2541, + "step": 45062 + }, + { + "epoch": 3.126986329886892, + "grad_norm": 0.9859704375267029, + "learning_rate": 2.3960680604297792e-05, + "loss": 0.2494, + "step": 45063 + }, + { + "epoch": 3.1270557213239885, + "grad_norm": 1.126042366027832, + "learning_rate": 2.3957031641455153e-05, + "loss": 0.2828, + "step": 45064 + }, + { + "epoch": 3.1271251127610853, + "grad_norm": 0.9591740965843201, + "learning_rate": 2.3953382918671353e-05, + "loss": 0.2554, + "step": 45065 + }, + { + "epoch": 3.127194504198182, + "grad_norm": 0.8072221875190735, + "learning_rate": 2.3949734435957882e-05, + "loss": 0.2177, + "step": 45066 + }, + { + "epoch": 3.1272638956352785, + "grad_norm": 0.9635196924209595, + "learning_rate": 2.3946086193326277e-05, + "loss": 0.2477, + "step": 45067 + }, + { + "epoch": 3.1273332870723753, + "grad_norm": 0.7648927569389343, + "learning_rate": 2.3942438190788053e-05, + "loss": 0.1819, + "step": 45068 + }, + { + "epoch": 3.1274026785094717, + "grad_norm": 0.9855961799621582, + "learning_rate": 2.3938790428354717e-05, + "loss": 0.2329, + "step": 45069 + }, + { + "epoch": 3.1274720699465686, + "grad_norm": 0.8383955955505371, + "learning_rate": 2.393514290603781e-05, + "loss": 0.2133, + "step": 45070 + }, + { + "epoch": 3.1275414613836654, + "grad_norm": 0.7651471495628357, + "learning_rate": 2.3931495623848855e-05, + "loss": 0.2317, + "step": 45071 + }, + { + "epoch": 3.127610852820762, + "grad_norm": 0.9527589082717896, + "learning_rate": 2.3927848581799318e-05, + "loss": 0.284, + "step": 45072 + }, + { + "epoch": 3.1276802442578586, + "grad_norm": 0.8763996362686157, + "learning_rate": 2.3924201779900724e-05, + "loss": 0.3021, + "step": 45073 + }, + { + "epoch": 3.127749635694955, + "grad_norm": 1.0646402835845947, + "learning_rate": 2.392055521816464e-05, + "loss": 0.3194, + "step": 45074 + }, + { + "epoch": 3.127819027132052, + "grad_norm": 1.5334392786026, + "learning_rate": 2.3916908896602498e-05, + "loss": 0.2869, + "step": 45075 + }, + { + "epoch": 3.1278884185691487, + "grad_norm": 0.7591291069984436, + "learning_rate": 2.3913262815225857e-05, + "loss": 0.2119, + "step": 45076 + }, + { + "epoch": 3.127957810006245, + "grad_norm": 0.8757861852645874, + "learning_rate": 2.390961697404621e-05, + "loss": 0.2337, + "step": 45077 + }, + { + "epoch": 3.128027201443342, + "grad_norm": 0.8445764183998108, + "learning_rate": 2.3905971373075074e-05, + "loss": 0.2873, + "step": 45078 + }, + { + "epoch": 3.1280965928804387, + "grad_norm": 0.6317397952079773, + "learning_rate": 2.3902326012323995e-05, + "loss": 0.1485, + "step": 45079 + }, + { + "epoch": 3.128165984317535, + "grad_norm": 0.7231806516647339, + "learning_rate": 2.3898680891804402e-05, + "loss": 0.1593, + "step": 45080 + }, + { + "epoch": 3.128235375754632, + "grad_norm": 1.1718701124191284, + "learning_rate": 2.389503601152786e-05, + "loss": 0.4067, + "step": 45081 + }, + { + "epoch": 3.1283047671917283, + "grad_norm": 0.8895435929298401, + "learning_rate": 2.3891391371505844e-05, + "loss": 0.2541, + "step": 45082 + }, + { + "epoch": 3.128374158628825, + "grad_norm": 0.6653412580490112, + "learning_rate": 2.3887746971749893e-05, + "loss": 0.1655, + "step": 45083 + }, + { + "epoch": 3.128443550065922, + "grad_norm": 0.7503894567489624, + "learning_rate": 2.388410281227148e-05, + "loss": 0.2375, + "step": 45084 + }, + { + "epoch": 3.1285129415030184, + "grad_norm": 0.8040645718574524, + "learning_rate": 2.3880458893082125e-05, + "loss": 0.185, + "step": 45085 + }, + { + "epoch": 3.1285823329401152, + "grad_norm": 1.1892277002334595, + "learning_rate": 2.387681521419336e-05, + "loss": 0.2806, + "step": 45086 + }, + { + "epoch": 3.128651724377212, + "grad_norm": 1.0234026908874512, + "learning_rate": 2.3873171775616633e-05, + "loss": 0.3709, + "step": 45087 + }, + { + "epoch": 3.1287211158143085, + "grad_norm": 0.7172013521194458, + "learning_rate": 2.3869528577363463e-05, + "loss": 0.1648, + "step": 45088 + }, + { + "epoch": 3.1287905072514053, + "grad_norm": 1.0639957189559937, + "learning_rate": 2.386588561944536e-05, + "loss": 0.2769, + "step": 45089 + }, + { + "epoch": 3.1288598986885017, + "grad_norm": 0.9451000094413757, + "learning_rate": 2.3862242901873823e-05, + "loss": 0.2485, + "step": 45090 + }, + { + "epoch": 3.1289292901255985, + "grad_norm": 1.6988273859024048, + "learning_rate": 2.385860042466036e-05, + "loss": 0.3054, + "step": 45091 + }, + { + "epoch": 3.1289986815626953, + "grad_norm": 0.8498106598854065, + "learning_rate": 2.3854958187816457e-05, + "loss": 0.2534, + "step": 45092 + }, + { + "epoch": 3.1290680729997917, + "grad_norm": 0.7856335639953613, + "learning_rate": 2.3851316191353633e-05, + "loss": 0.2037, + "step": 45093 + }, + { + "epoch": 3.1291374644368886, + "grad_norm": 0.8272308111190796, + "learning_rate": 2.384767443528335e-05, + "loss": 0.2191, + "step": 45094 + }, + { + "epoch": 3.129206855873985, + "grad_norm": 1.0314903259277344, + "learning_rate": 2.3844032919617133e-05, + "loss": 0.3607, + "step": 45095 + }, + { + "epoch": 3.129276247311082, + "grad_norm": 0.87114018201828, + "learning_rate": 2.3840391644366455e-05, + "loss": 0.2184, + "step": 45096 + }, + { + "epoch": 3.1293456387481786, + "grad_norm": 1.148698091506958, + "learning_rate": 2.3836750609542824e-05, + "loss": 0.2791, + "step": 45097 + }, + { + "epoch": 3.129415030185275, + "grad_norm": 0.921466052532196, + "learning_rate": 2.383310981515774e-05, + "loss": 0.2349, + "step": 45098 + }, + { + "epoch": 3.129484421622372, + "grad_norm": 1.1801321506500244, + "learning_rate": 2.3829469261222692e-05, + "loss": 0.2162, + "step": 45099 + }, + { + "epoch": 3.1295538130594682, + "grad_norm": 0.5978344678878784, + "learning_rate": 2.3825828947749208e-05, + "loss": 0.1478, + "step": 45100 + }, + { + "epoch": 3.129623204496565, + "grad_norm": 1.3129104375839233, + "learning_rate": 2.3822188874748696e-05, + "loss": 0.3442, + "step": 45101 + }, + { + "epoch": 3.129692595933662, + "grad_norm": 0.651354968547821, + "learning_rate": 2.3818549042232707e-05, + "loss": 0.1669, + "step": 45102 + }, + { + "epoch": 3.1297619873707583, + "grad_norm": 0.8913907408714294, + "learning_rate": 2.381490945021272e-05, + "loss": 0.3032, + "step": 45103 + }, + { + "epoch": 3.129831378807855, + "grad_norm": 1.767611026763916, + "learning_rate": 2.381127009870022e-05, + "loss": 0.347, + "step": 45104 + }, + { + "epoch": 3.129900770244952, + "grad_norm": 0.8202996253967285, + "learning_rate": 2.380763098770674e-05, + "loss": 0.2345, + "step": 45105 + }, + { + "epoch": 3.1299701616820483, + "grad_norm": 1.052362322807312, + "learning_rate": 2.380399211724369e-05, + "loss": 0.2989, + "step": 45106 + }, + { + "epoch": 3.130039553119145, + "grad_norm": 0.663139283657074, + "learning_rate": 2.3800353487322603e-05, + "loss": 0.1906, + "step": 45107 + }, + { + "epoch": 3.1301089445562416, + "grad_norm": 1.2187217473983765, + "learning_rate": 2.379671509795496e-05, + "loss": 0.279, + "step": 45108 + }, + { + "epoch": 3.1301783359933384, + "grad_norm": 0.7479267716407776, + "learning_rate": 2.379307694915227e-05, + "loss": 0.2161, + "step": 45109 + }, + { + "epoch": 3.1302477274304352, + "grad_norm": 2.6213881969451904, + "learning_rate": 2.3789439040925965e-05, + "loss": 0.3286, + "step": 45110 + }, + { + "epoch": 3.1303171188675316, + "grad_norm": 1.410007119178772, + "learning_rate": 2.3785801373287563e-05, + "loss": 0.296, + "step": 45111 + }, + { + "epoch": 3.1303865103046284, + "grad_norm": 1.1381239891052246, + "learning_rate": 2.378216394624855e-05, + "loss": 0.2774, + "step": 45112 + }, + { + "epoch": 3.1304559017417253, + "grad_norm": 0.7379522919654846, + "learning_rate": 2.3778526759820386e-05, + "loss": 0.3166, + "step": 45113 + }, + { + "epoch": 3.1305252931788217, + "grad_norm": 1.241109848022461, + "learning_rate": 2.3774889814014612e-05, + "loss": 0.2319, + "step": 45114 + }, + { + "epoch": 3.1305946846159185, + "grad_norm": 0.8672340512275696, + "learning_rate": 2.3771253108842628e-05, + "loss": 0.2928, + "step": 45115 + }, + { + "epoch": 3.130664076053015, + "grad_norm": 0.9974434971809387, + "learning_rate": 2.376761664431596e-05, + "loss": 0.2491, + "step": 45116 + }, + { + "epoch": 3.1307334674901117, + "grad_norm": 0.7474615573883057, + "learning_rate": 2.3763980420446084e-05, + "loss": 0.1955, + "step": 45117 + }, + { + "epoch": 3.1308028589272086, + "grad_norm": 1.4680057764053345, + "learning_rate": 2.3760344437244476e-05, + "loss": 0.3782, + "step": 45118 + }, + { + "epoch": 3.130872250364305, + "grad_norm": 0.6575485467910767, + "learning_rate": 2.3756708694722608e-05, + "loss": 0.2144, + "step": 45119 + }, + { + "epoch": 3.1309416418014018, + "grad_norm": 0.8216129541397095, + "learning_rate": 2.375307319289197e-05, + "loss": 0.2051, + "step": 45120 + }, + { + "epoch": 3.131011033238498, + "grad_norm": 0.8438489437103271, + "learning_rate": 2.3749437931764063e-05, + "loss": 0.2405, + "step": 45121 + }, + { + "epoch": 3.131080424675595, + "grad_norm": 0.5894724726676941, + "learning_rate": 2.37458029113503e-05, + "loss": 0.2084, + "step": 45122 + }, + { + "epoch": 3.131149816112692, + "grad_norm": 1.1136212348937988, + "learning_rate": 2.3742168131662192e-05, + "loss": 0.2947, + "step": 45123 + }, + { + "epoch": 3.131219207549788, + "grad_norm": 0.8123213648796082, + "learning_rate": 2.373853359271121e-05, + "loss": 0.2725, + "step": 45124 + }, + { + "epoch": 3.131288598986885, + "grad_norm": 0.8776993751525879, + "learning_rate": 2.373489929450884e-05, + "loss": 0.2485, + "step": 45125 + }, + { + "epoch": 3.131357990423982, + "grad_norm": 0.9181239604949951, + "learning_rate": 2.3731265237066536e-05, + "loss": 0.2348, + "step": 45126 + }, + { + "epoch": 3.1314273818610783, + "grad_norm": 1.164918303489685, + "learning_rate": 2.3727631420395812e-05, + "loss": 0.2265, + "step": 45127 + }, + { + "epoch": 3.131496773298175, + "grad_norm": 0.7329376935958862, + "learning_rate": 2.372399784450807e-05, + "loss": 0.2524, + "step": 45128 + }, + { + "epoch": 3.1315661647352715, + "grad_norm": 1.3664058446884155, + "learning_rate": 2.372036450941483e-05, + "loss": 0.2811, + "step": 45129 + }, + { + "epoch": 3.1316355561723683, + "grad_norm": 0.8237148523330688, + "learning_rate": 2.3716731415127535e-05, + "loss": 0.2187, + "step": 45130 + }, + { + "epoch": 3.131704947609465, + "grad_norm": 0.8250108361244202, + "learning_rate": 2.3713098561657677e-05, + "loss": 0.1968, + "step": 45131 + }, + { + "epoch": 3.1317743390465616, + "grad_norm": 0.8080863356590271, + "learning_rate": 2.370946594901671e-05, + "loss": 0.2254, + "step": 45132 + }, + { + "epoch": 3.1318437304836584, + "grad_norm": 0.717357873916626, + "learning_rate": 2.3705833577216107e-05, + "loss": 0.2392, + "step": 45133 + }, + { + "epoch": 3.1319131219207548, + "grad_norm": 1.3618143796920776, + "learning_rate": 2.370220144626737e-05, + "loss": 0.3305, + "step": 45134 + }, + { + "epoch": 3.1319825133578516, + "grad_norm": 0.7536410689353943, + "learning_rate": 2.3698569556181903e-05, + "loss": 0.1845, + "step": 45135 + }, + { + "epoch": 3.1320519047949484, + "grad_norm": 1.4941325187683105, + "learning_rate": 2.3694937906971194e-05, + "loss": 0.2931, + "step": 45136 + }, + { + "epoch": 3.132121296232045, + "grad_norm": 0.8719547390937805, + "learning_rate": 2.3691306498646715e-05, + "loss": 0.2295, + "step": 45137 + }, + { + "epoch": 3.1321906876691417, + "grad_norm": 1.7600486278533936, + "learning_rate": 2.3687675331219927e-05, + "loss": 0.4571, + "step": 45138 + }, + { + "epoch": 3.1322600791062385, + "grad_norm": 1.1037683486938477, + "learning_rate": 2.36840444047023e-05, + "loss": 0.2858, + "step": 45139 + }, + { + "epoch": 3.132329470543335, + "grad_norm": 1.1702263355255127, + "learning_rate": 2.368041371910532e-05, + "loss": 0.2522, + "step": 45140 + }, + { + "epoch": 3.1323988619804317, + "grad_norm": 1.2987457513809204, + "learning_rate": 2.3676783274440383e-05, + "loss": 0.3954, + "step": 45141 + }, + { + "epoch": 3.132468253417528, + "grad_norm": 1.7656278610229492, + "learning_rate": 2.367315307071899e-05, + "loss": 0.3759, + "step": 45142 + }, + { + "epoch": 3.132537644854625, + "grad_norm": 0.8498369455337524, + "learning_rate": 2.3669523107952595e-05, + "loss": 0.2028, + "step": 45143 + }, + { + "epoch": 3.1326070362917218, + "grad_norm": 1.4315564632415771, + "learning_rate": 2.3665893386152682e-05, + "loss": 0.3201, + "step": 45144 + }, + { + "epoch": 3.132676427728818, + "grad_norm": 0.877278745174408, + "learning_rate": 2.3662263905330672e-05, + "loss": 0.2227, + "step": 45145 + }, + { + "epoch": 3.132745819165915, + "grad_norm": 0.941604733467102, + "learning_rate": 2.3658634665498026e-05, + "loss": 0.3592, + "step": 45146 + }, + { + "epoch": 3.1328152106030114, + "grad_norm": 0.8809286952018738, + "learning_rate": 2.365500566666622e-05, + "loss": 0.3356, + "step": 45147 + }, + { + "epoch": 3.132884602040108, + "grad_norm": 0.857666015625, + "learning_rate": 2.3651376908846735e-05, + "loss": 0.2638, + "step": 45148 + }, + { + "epoch": 3.132953993477205, + "grad_norm": 1.015115737915039, + "learning_rate": 2.364774839205096e-05, + "loss": 0.3738, + "step": 45149 + }, + { + "epoch": 3.1330233849143014, + "grad_norm": 0.7998127341270447, + "learning_rate": 2.364412011629039e-05, + "loss": 0.2452, + "step": 45150 + }, + { + "epoch": 3.1330927763513983, + "grad_norm": 0.8118219375610352, + "learning_rate": 2.3640492081576472e-05, + "loss": 0.2874, + "step": 45151 + }, + { + "epoch": 3.133162167788495, + "grad_norm": 0.996843159198761, + "learning_rate": 2.3636864287920657e-05, + "loss": 0.2806, + "step": 45152 + }, + { + "epoch": 3.1332315592255915, + "grad_norm": 1.4313775300979614, + "learning_rate": 2.363323673533441e-05, + "loss": 0.3435, + "step": 45153 + }, + { + "epoch": 3.1333009506626883, + "grad_norm": 1.3250969648361206, + "learning_rate": 2.3629609423829167e-05, + "loss": 0.3824, + "step": 45154 + }, + { + "epoch": 3.1333703420997847, + "grad_norm": 0.5975522994995117, + "learning_rate": 2.3625982353416408e-05, + "loss": 0.2446, + "step": 45155 + }, + { + "epoch": 3.1334397335368815, + "grad_norm": 0.6234197616577148, + "learning_rate": 2.3622355524107555e-05, + "loss": 0.1557, + "step": 45156 + }, + { + "epoch": 3.1335091249739784, + "grad_norm": 0.9354624152183533, + "learning_rate": 2.3618728935914047e-05, + "loss": 0.3574, + "step": 45157 + }, + { + "epoch": 3.1335785164110748, + "grad_norm": 0.7260840535163879, + "learning_rate": 2.361510258884735e-05, + "loss": 0.2175, + "step": 45158 + }, + { + "epoch": 3.1336479078481716, + "grad_norm": 1.3279953002929688, + "learning_rate": 2.3611476482918915e-05, + "loss": 0.2847, + "step": 45159 + }, + { + "epoch": 3.1337172992852684, + "grad_norm": 0.8835667967796326, + "learning_rate": 2.3607850618140194e-05, + "loss": 0.3557, + "step": 45160 + }, + { + "epoch": 3.133786690722365, + "grad_norm": 0.8385189771652222, + "learning_rate": 2.3604224994522616e-05, + "loss": 0.3413, + "step": 45161 + }, + { + "epoch": 3.1338560821594617, + "grad_norm": 0.9819769263267517, + "learning_rate": 2.3600599612077666e-05, + "loss": 0.2989, + "step": 45162 + }, + { + "epoch": 3.133925473596558, + "grad_norm": 0.9556541442871094, + "learning_rate": 2.3596974470816735e-05, + "loss": 0.2768, + "step": 45163 + }, + { + "epoch": 3.133994865033655, + "grad_norm": 1.1703424453735352, + "learning_rate": 2.359334957075129e-05, + "loss": 0.2567, + "step": 45164 + }, + { + "epoch": 3.1340642564707517, + "grad_norm": 1.0834465026855469, + "learning_rate": 2.3589724911892775e-05, + "loss": 0.1935, + "step": 45165 + }, + { + "epoch": 3.134133647907848, + "grad_norm": 1.0279794931411743, + "learning_rate": 2.3586100494252626e-05, + "loss": 0.3022, + "step": 45166 + }, + { + "epoch": 3.134203039344945, + "grad_norm": 0.8813967704772949, + "learning_rate": 2.3582476317842294e-05, + "loss": 0.2228, + "step": 45167 + }, + { + "epoch": 3.1342724307820413, + "grad_norm": 1.0249686241149902, + "learning_rate": 2.3578852382673223e-05, + "loss": 0.2656, + "step": 45168 + }, + { + "epoch": 3.134341822219138, + "grad_norm": 0.9020690321922302, + "learning_rate": 2.357522868875688e-05, + "loss": 0.2091, + "step": 45169 + }, + { + "epoch": 3.134411213656235, + "grad_norm": 0.6185152530670166, + "learning_rate": 2.3571605236104644e-05, + "loss": 0.2204, + "step": 45170 + }, + { + "epoch": 3.1344806050933314, + "grad_norm": 0.8323894143104553, + "learning_rate": 2.3567982024727974e-05, + "loss": 0.2933, + "step": 45171 + }, + { + "epoch": 3.134549996530428, + "grad_norm": 1.0420042276382446, + "learning_rate": 2.356435905463833e-05, + "loss": 0.1946, + "step": 45172 + }, + { + "epoch": 3.1346193879675246, + "grad_norm": 1.0279817581176758, + "learning_rate": 2.3560736325847122e-05, + "loss": 0.3041, + "step": 45173 + }, + { + "epoch": 3.1346887794046214, + "grad_norm": 0.7605175971984863, + "learning_rate": 2.355711383836581e-05, + "loss": 0.3834, + "step": 45174 + }, + { + "epoch": 3.1347581708417183, + "grad_norm": 0.8839420080184937, + "learning_rate": 2.3553491592205846e-05, + "loss": 0.3013, + "step": 45175 + }, + { + "epoch": 3.1348275622788146, + "grad_norm": 0.6571767926216125, + "learning_rate": 2.3549869587378614e-05, + "loss": 0.2293, + "step": 45176 + }, + { + "epoch": 3.1348969537159115, + "grad_norm": 1.2675232887268066, + "learning_rate": 2.354624782389557e-05, + "loss": 0.2188, + "step": 45177 + }, + { + "epoch": 3.1349663451530083, + "grad_norm": 0.9496719241142273, + "learning_rate": 2.3542626301768157e-05, + "loss": 0.2211, + "step": 45178 + }, + { + "epoch": 3.1350357365901047, + "grad_norm": 1.0119414329528809, + "learning_rate": 2.353900502100782e-05, + "loss": 0.3311, + "step": 45179 + }, + { + "epoch": 3.1351051280272015, + "grad_norm": 0.9482215642929077, + "learning_rate": 2.353538398162596e-05, + "loss": 0.3408, + "step": 45180 + }, + { + "epoch": 3.135174519464298, + "grad_norm": 0.9994434714317322, + "learning_rate": 2.3531763183634005e-05, + "loss": 0.4092, + "step": 45181 + }, + { + "epoch": 3.1352439109013948, + "grad_norm": 0.80745929479599, + "learning_rate": 2.3528142627043414e-05, + "loss": 0.2884, + "step": 45182 + }, + { + "epoch": 3.1353133023384916, + "grad_norm": 1.2141249179840088, + "learning_rate": 2.3524522311865627e-05, + "loss": 0.2926, + "step": 45183 + }, + { + "epoch": 3.135382693775588, + "grad_norm": 1.2581565380096436, + "learning_rate": 2.352090223811202e-05, + "loss": 0.23, + "step": 45184 + }, + { + "epoch": 3.135452085212685, + "grad_norm": 0.6133684515953064, + "learning_rate": 2.351728240579406e-05, + "loss": 0.1492, + "step": 45185 + }, + { + "epoch": 3.1355214766497816, + "grad_norm": 0.8654347062110901, + "learning_rate": 2.3513662814923156e-05, + "loss": 0.2125, + "step": 45186 + }, + { + "epoch": 3.135590868086878, + "grad_norm": 0.6661157608032227, + "learning_rate": 2.3510043465510745e-05, + "loss": 0.1903, + "step": 45187 + }, + { + "epoch": 3.135660259523975, + "grad_norm": 0.9431607127189636, + "learning_rate": 2.350642435756826e-05, + "loss": 0.2759, + "step": 45188 + }, + { + "epoch": 3.1357296509610713, + "grad_norm": 0.8567787408828735, + "learning_rate": 2.3502805491107116e-05, + "loss": 0.16, + "step": 45189 + }, + { + "epoch": 3.135799042398168, + "grad_norm": 1.1011943817138672, + "learning_rate": 2.3499186866138767e-05, + "loss": 0.149, + "step": 45190 + }, + { + "epoch": 3.135868433835265, + "grad_norm": 1.059145212173462, + "learning_rate": 2.3495568482674578e-05, + "loss": 0.3037, + "step": 45191 + }, + { + "epoch": 3.1359378252723613, + "grad_norm": 0.7567523717880249, + "learning_rate": 2.349195034072601e-05, + "loss": 0.1692, + "step": 45192 + }, + { + "epoch": 3.136007216709458, + "grad_norm": 0.9484298229217529, + "learning_rate": 2.348833244030447e-05, + "loss": 0.1697, + "step": 45193 + }, + { + "epoch": 3.1360766081465545, + "grad_norm": 1.0938501358032227, + "learning_rate": 2.348471478142139e-05, + "loss": 0.1822, + "step": 45194 + }, + { + "epoch": 3.1361459995836514, + "grad_norm": 1.0986543893814087, + "learning_rate": 2.3481097364088188e-05, + "loss": 0.3335, + "step": 45195 + }, + { + "epoch": 3.136215391020748, + "grad_norm": 0.96346515417099, + "learning_rate": 2.347748018831629e-05, + "loss": 0.1841, + "step": 45196 + }, + { + "epoch": 3.1362847824578446, + "grad_norm": 0.9180934429168701, + "learning_rate": 2.3473863254117136e-05, + "loss": 0.2135, + "step": 45197 + }, + { + "epoch": 3.1363541738949414, + "grad_norm": 1.3277227878570557, + "learning_rate": 2.3470246561502075e-05, + "loss": 0.3162, + "step": 45198 + }, + { + "epoch": 3.136423565332038, + "grad_norm": 0.9581111669540405, + "learning_rate": 2.346663011048258e-05, + "loss": 0.3202, + "step": 45199 + }, + { + "epoch": 3.1364929567691346, + "grad_norm": 0.9360209107398987, + "learning_rate": 2.3463013901070052e-05, + "loss": 0.2565, + "step": 45200 + }, + { + "epoch": 3.1365623482062315, + "grad_norm": 1.0794490575790405, + "learning_rate": 2.3459397933275907e-05, + "loss": 0.2524, + "step": 45201 + }, + { + "epoch": 3.136631739643328, + "grad_norm": 0.8405042886734009, + "learning_rate": 2.3455782207111564e-05, + "loss": 0.231, + "step": 45202 + }, + { + "epoch": 3.1367011310804247, + "grad_norm": 1.4405015707015991, + "learning_rate": 2.3452166722588465e-05, + "loss": 0.3964, + "step": 45203 + }, + { + "epoch": 3.1367705225175215, + "grad_norm": 0.7998762130737305, + "learning_rate": 2.344855147971796e-05, + "loss": 0.223, + "step": 45204 + }, + { + "epoch": 3.136839913954618, + "grad_norm": 0.932511568069458, + "learning_rate": 2.3444936478511502e-05, + "loss": 0.2695, + "step": 45205 + }, + { + "epoch": 3.1369093053917148, + "grad_norm": 0.6837620139122009, + "learning_rate": 2.344132171898049e-05, + "loss": 0.1786, + "step": 45206 + }, + { + "epoch": 3.136978696828811, + "grad_norm": 0.7737235426902771, + "learning_rate": 2.343770720113635e-05, + "loss": 0.2844, + "step": 45207 + }, + { + "epoch": 3.137048088265908, + "grad_norm": 1.0217254161834717, + "learning_rate": 2.3434092924990493e-05, + "loss": 0.3032, + "step": 45208 + }, + { + "epoch": 3.137117479703005, + "grad_norm": 1.0841524600982666, + "learning_rate": 2.343047889055431e-05, + "loss": 0.3581, + "step": 45209 + }, + { + "epoch": 3.137186871140101, + "grad_norm": 0.936311662197113, + "learning_rate": 2.3426865097839256e-05, + "loss": 0.3765, + "step": 45210 + }, + { + "epoch": 3.137256262577198, + "grad_norm": 1.2700799703598022, + "learning_rate": 2.3423251546856674e-05, + "loss": 0.4729, + "step": 45211 + }, + { + "epoch": 3.137325654014295, + "grad_norm": 1.406689167022705, + "learning_rate": 2.3419638237618003e-05, + "loss": 0.3081, + "step": 45212 + }, + { + "epoch": 3.1373950454513913, + "grad_norm": 0.874478816986084, + "learning_rate": 2.3416025170134637e-05, + "loss": 0.198, + "step": 45213 + }, + { + "epoch": 3.137464436888488, + "grad_norm": 0.9733192920684814, + "learning_rate": 2.3412412344418032e-05, + "loss": 0.2745, + "step": 45214 + }, + { + "epoch": 3.1375338283255845, + "grad_norm": 0.7872303128242493, + "learning_rate": 2.340879976047954e-05, + "loss": 0.1959, + "step": 45215 + }, + { + "epoch": 3.1376032197626813, + "grad_norm": 0.8298403024673462, + "learning_rate": 2.3405187418330564e-05, + "loss": 0.1782, + "step": 45216 + }, + { + "epoch": 3.137672611199778, + "grad_norm": 0.8988802433013916, + "learning_rate": 2.340157531798256e-05, + "loss": 0.2985, + "step": 45217 + }, + { + "epoch": 3.1377420026368745, + "grad_norm": 0.8877331018447876, + "learning_rate": 2.339796345944686e-05, + "loss": 0.2542, + "step": 45218 + }, + { + "epoch": 3.1378113940739714, + "grad_norm": 0.7368435263633728, + "learning_rate": 2.3394351842734906e-05, + "loss": 0.2818, + "step": 45219 + }, + { + "epoch": 3.1378807855110677, + "grad_norm": 0.8825321793556213, + "learning_rate": 2.3390740467858095e-05, + "loss": 0.3328, + "step": 45220 + }, + { + "epoch": 3.1379501769481646, + "grad_norm": 1.0449867248535156, + "learning_rate": 2.338712933482784e-05, + "loss": 0.4683, + "step": 45221 + }, + { + "epoch": 3.1380195683852614, + "grad_norm": 0.9872543811798096, + "learning_rate": 2.3383518443655515e-05, + "loss": 0.3198, + "step": 45222 + }, + { + "epoch": 3.138088959822358, + "grad_norm": 1.0124661922454834, + "learning_rate": 2.337990779435254e-05, + "loss": 0.2556, + "step": 45223 + }, + { + "epoch": 3.1381583512594546, + "grad_norm": 1.0690498352050781, + "learning_rate": 2.3376297386930322e-05, + "loss": 0.3368, + "step": 45224 + }, + { + "epoch": 3.138227742696551, + "grad_norm": 1.0597165822982788, + "learning_rate": 2.3372687221400235e-05, + "loss": 0.2905, + "step": 45225 + }, + { + "epoch": 3.138297134133648, + "grad_norm": 1.3499563932418823, + "learning_rate": 2.3369077297773667e-05, + "loss": 0.4247, + "step": 45226 + }, + { + "epoch": 3.1383665255707447, + "grad_norm": 0.6991201639175415, + "learning_rate": 2.3365467616062043e-05, + "loss": 0.1675, + "step": 45227 + }, + { + "epoch": 3.138435917007841, + "grad_norm": 0.7413440346717834, + "learning_rate": 2.336185817627674e-05, + "loss": 0.1432, + "step": 45228 + }, + { + "epoch": 3.138505308444938, + "grad_norm": 1.2020149230957031, + "learning_rate": 2.3358248978429154e-05, + "loss": 0.3184, + "step": 45229 + }, + { + "epoch": 3.1385746998820347, + "grad_norm": 1.0882145166397095, + "learning_rate": 2.3354640022530695e-05, + "loss": 0.4081, + "step": 45230 + }, + { + "epoch": 3.138644091319131, + "grad_norm": 1.011560082435608, + "learning_rate": 2.3351031308592763e-05, + "loss": 0.1917, + "step": 45231 + }, + { + "epoch": 3.138713482756228, + "grad_norm": 1.1477010250091553, + "learning_rate": 2.334742283662671e-05, + "loss": 0.3209, + "step": 45232 + }, + { + "epoch": 3.1387828741933244, + "grad_norm": 1.2195686101913452, + "learning_rate": 2.3343814606643953e-05, + "loss": 0.2378, + "step": 45233 + }, + { + "epoch": 3.138852265630421, + "grad_norm": 1.113675594329834, + "learning_rate": 2.334020661865588e-05, + "loss": 0.39, + "step": 45234 + }, + { + "epoch": 3.138921657067518, + "grad_norm": 1.5708519220352173, + "learning_rate": 2.3336598872673875e-05, + "loss": 0.2988, + "step": 45235 + }, + { + "epoch": 3.1389910485046144, + "grad_norm": 0.9412149786949158, + "learning_rate": 2.3332991368709333e-05, + "loss": 0.2545, + "step": 45236 + }, + { + "epoch": 3.1390604399417112, + "grad_norm": 0.9419254064559937, + "learning_rate": 2.3329384106773643e-05, + "loss": 0.2817, + "step": 45237 + }, + { + "epoch": 3.139129831378808, + "grad_norm": 0.9035120010375977, + "learning_rate": 2.332577708687822e-05, + "loss": 0.2934, + "step": 45238 + }, + { + "epoch": 3.1391992228159045, + "grad_norm": 0.8829771876335144, + "learning_rate": 2.3322170309034397e-05, + "loss": 0.232, + "step": 45239 + }, + { + "epoch": 3.1392686142530013, + "grad_norm": 0.6613840460777283, + "learning_rate": 2.3318563773253575e-05, + "loss": 0.1307, + "step": 45240 + }, + { + "epoch": 3.1393380056900977, + "grad_norm": 0.9423418045043945, + "learning_rate": 2.3314957479547163e-05, + "loss": 0.2796, + "step": 45241 + }, + { + "epoch": 3.1394073971271945, + "grad_norm": 0.6775540709495544, + "learning_rate": 2.331135142792653e-05, + "loss": 0.214, + "step": 45242 + }, + { + "epoch": 3.1394767885642914, + "grad_norm": 1.1722266674041748, + "learning_rate": 2.3307745618403065e-05, + "loss": 0.2397, + "step": 45243 + }, + { + "epoch": 3.1395461800013877, + "grad_norm": 1.0244659185409546, + "learning_rate": 2.3304140050988143e-05, + "loss": 0.1994, + "step": 45244 + }, + { + "epoch": 3.1396155714384846, + "grad_norm": 0.8314455151557922, + "learning_rate": 2.3300534725693192e-05, + "loss": 0.335, + "step": 45245 + }, + { + "epoch": 3.139684962875581, + "grad_norm": 1.215757966041565, + "learning_rate": 2.3296929642529507e-05, + "loss": 0.3743, + "step": 45246 + }, + { + "epoch": 3.139754354312678, + "grad_norm": 0.7055997252464294, + "learning_rate": 2.3293324801508533e-05, + "loss": 0.1698, + "step": 45247 + }, + { + "epoch": 3.1398237457497746, + "grad_norm": 1.1398367881774902, + "learning_rate": 2.328972020264165e-05, + "loss": 0.2173, + "step": 45248 + }, + { + "epoch": 3.139893137186871, + "grad_norm": 0.7750896215438843, + "learning_rate": 2.3286115845940205e-05, + "loss": 0.3002, + "step": 45249 + }, + { + "epoch": 3.139962528623968, + "grad_norm": 1.8643953800201416, + "learning_rate": 2.328251173141558e-05, + "loss": 0.3562, + "step": 45250 + }, + { + "epoch": 3.1400319200610642, + "grad_norm": 0.9228726029396057, + "learning_rate": 2.327890785907917e-05, + "loss": 0.2101, + "step": 45251 + }, + { + "epoch": 3.140101311498161, + "grad_norm": 1.0438380241394043, + "learning_rate": 2.3275304228942375e-05, + "loss": 0.2675, + "step": 45252 + }, + { + "epoch": 3.140170702935258, + "grad_norm": 0.8590624928474426, + "learning_rate": 2.327170084101651e-05, + "loss": 0.1163, + "step": 45253 + }, + { + "epoch": 3.1402400943723543, + "grad_norm": 0.6571321487426758, + "learning_rate": 2.3268097695312985e-05, + "loss": 0.151, + "step": 45254 + }, + { + "epoch": 3.140309485809451, + "grad_norm": 1.5689260959625244, + "learning_rate": 2.326449479184317e-05, + "loss": 0.4103, + "step": 45255 + }, + { + "epoch": 3.140378877246548, + "grad_norm": 0.9676522612571716, + "learning_rate": 2.326089213061845e-05, + "loss": 0.1901, + "step": 45256 + }, + { + "epoch": 3.1404482686836444, + "grad_norm": 1.0354316234588623, + "learning_rate": 2.325728971165019e-05, + "loss": 0.3183, + "step": 45257 + }, + { + "epoch": 3.140517660120741, + "grad_norm": 0.7075691223144531, + "learning_rate": 2.3253687534949765e-05, + "loss": 0.2191, + "step": 45258 + }, + { + "epoch": 3.1405870515578376, + "grad_norm": 0.8660216331481934, + "learning_rate": 2.325008560052856e-05, + "loss": 0.23, + "step": 45259 + }, + { + "epoch": 3.1406564429949344, + "grad_norm": 1.466950535774231, + "learning_rate": 2.3246483908397908e-05, + "loss": 0.1577, + "step": 45260 + }, + { + "epoch": 3.1407258344320312, + "grad_norm": 1.4747605323791504, + "learning_rate": 2.3242882458569206e-05, + "loss": 0.3749, + "step": 45261 + }, + { + "epoch": 3.1407952258691276, + "grad_norm": 0.6937659978866577, + "learning_rate": 2.3239281251053814e-05, + "loss": 0.2001, + "step": 45262 + }, + { + "epoch": 3.1408646173062245, + "grad_norm": 1.0715962648391724, + "learning_rate": 2.3235680285863103e-05, + "loss": 0.3517, + "step": 45263 + }, + { + "epoch": 3.1409340087433213, + "grad_norm": 0.9724109172821045, + "learning_rate": 2.3232079563008448e-05, + "loss": 0.2621, + "step": 45264 + }, + { + "epoch": 3.1410034001804177, + "grad_norm": 0.7298134565353394, + "learning_rate": 2.3228479082501208e-05, + "loss": 0.2141, + "step": 45265 + }, + { + "epoch": 3.1410727916175145, + "grad_norm": 0.7216483950614929, + "learning_rate": 2.322487884435278e-05, + "loss": 0.1643, + "step": 45266 + }, + { + "epoch": 3.141142183054611, + "grad_norm": 0.9037635326385498, + "learning_rate": 2.322127884857448e-05, + "loss": 0.2461, + "step": 45267 + }, + { + "epoch": 3.1412115744917077, + "grad_norm": 1.0503095388412476, + "learning_rate": 2.3217679095177693e-05, + "loss": 0.2779, + "step": 45268 + }, + { + "epoch": 3.1412809659288046, + "grad_norm": 1.1557495594024658, + "learning_rate": 2.3214079584173776e-05, + "loss": 0.3346, + "step": 45269 + }, + { + "epoch": 3.141350357365901, + "grad_norm": 0.9245873093605042, + "learning_rate": 2.32104803155741e-05, + "loss": 0.2509, + "step": 45270 + }, + { + "epoch": 3.141419748802998, + "grad_norm": 1.0775827169418335, + "learning_rate": 2.3206881289390037e-05, + "loss": 0.1704, + "step": 45271 + }, + { + "epoch": 3.141489140240094, + "grad_norm": 0.9093495011329651, + "learning_rate": 2.320328250563294e-05, + "loss": 0.1811, + "step": 45272 + }, + { + "epoch": 3.141558531677191, + "grad_norm": 0.9102795124053955, + "learning_rate": 2.31996839643142e-05, + "loss": 0.2596, + "step": 45273 + }, + { + "epoch": 3.141627923114288, + "grad_norm": 0.7273774743080139, + "learning_rate": 2.319608566544511e-05, + "loss": 0.2559, + "step": 45274 + }, + { + "epoch": 3.1416973145513842, + "grad_norm": 0.9729095697402954, + "learning_rate": 2.3192487609037072e-05, + "loss": 0.2677, + "step": 45275 + }, + { + "epoch": 3.141766705988481, + "grad_norm": 0.7284733653068542, + "learning_rate": 2.3188889795101443e-05, + "loss": 0.2918, + "step": 45276 + }, + { + "epoch": 3.1418360974255775, + "grad_norm": 0.6573286652565002, + "learning_rate": 2.3185292223649568e-05, + "loss": 0.1988, + "step": 45277 + }, + { + "epoch": 3.1419054888626743, + "grad_norm": 1.2510545253753662, + "learning_rate": 2.3181694894692817e-05, + "loss": 0.257, + "step": 45278 + }, + { + "epoch": 3.141974880299771, + "grad_norm": 0.9138762950897217, + "learning_rate": 2.317809780824258e-05, + "loss": 0.297, + "step": 45279 + }, + { + "epoch": 3.1420442717368675, + "grad_norm": 0.8302096128463745, + "learning_rate": 2.317450096431014e-05, + "loss": 0.236, + "step": 45280 + }, + { + "epoch": 3.1421136631739643, + "grad_norm": 1.164415717124939, + "learning_rate": 2.3170904362906885e-05, + "loss": 0.3284, + "step": 45281 + }, + { + "epoch": 3.142183054611061, + "grad_norm": 0.7467427849769592, + "learning_rate": 2.3167308004044175e-05, + "loss": 0.1696, + "step": 45282 + }, + { + "epoch": 3.1422524460481576, + "grad_norm": 1.1758877038955688, + "learning_rate": 2.3163711887733385e-05, + "loss": 0.3154, + "step": 45283 + }, + { + "epoch": 3.1423218374852544, + "grad_norm": 0.926292359828949, + "learning_rate": 2.316011601398581e-05, + "loss": 0.1995, + "step": 45284 + }, + { + "epoch": 3.142391228922351, + "grad_norm": 0.9359236359596252, + "learning_rate": 2.315652038281284e-05, + "loss": 0.228, + "step": 45285 + }, + { + "epoch": 3.1424606203594476, + "grad_norm": 0.6464989185333252, + "learning_rate": 2.3152924994225822e-05, + "loss": 0.2701, + "step": 45286 + }, + { + "epoch": 3.1425300117965445, + "grad_norm": 0.776332437992096, + "learning_rate": 2.3149329848236122e-05, + "loss": 0.2474, + "step": 45287 + }, + { + "epoch": 3.142599403233641, + "grad_norm": 0.7497934699058533, + "learning_rate": 2.3145734944855045e-05, + "loss": 0.2188, + "step": 45288 + }, + { + "epoch": 3.1426687946707377, + "grad_norm": 1.0688834190368652, + "learning_rate": 2.3142140284093972e-05, + "loss": 0.2722, + "step": 45289 + }, + { + "epoch": 3.1427381861078345, + "grad_norm": 1.3963584899902344, + "learning_rate": 2.3138545865964244e-05, + "loss": 0.3005, + "step": 45290 + }, + { + "epoch": 3.142807577544931, + "grad_norm": 0.6324653625488281, + "learning_rate": 2.3134951690477214e-05, + "loss": 0.2072, + "step": 45291 + }, + { + "epoch": 3.1428769689820277, + "grad_norm": 0.6728149056434631, + "learning_rate": 2.313135775764421e-05, + "loss": 0.2178, + "step": 45292 + }, + { + "epoch": 3.142946360419124, + "grad_norm": 0.8297364115715027, + "learning_rate": 2.312776406747663e-05, + "loss": 0.1729, + "step": 45293 + }, + { + "epoch": 3.143015751856221, + "grad_norm": 1.1092009544372559, + "learning_rate": 2.312417061998575e-05, + "loss": 0.1478, + "step": 45294 + }, + { + "epoch": 3.143085143293318, + "grad_norm": 1.199319839477539, + "learning_rate": 2.3120577415182944e-05, + "loss": 0.3091, + "step": 45295 + }, + { + "epoch": 3.143154534730414, + "grad_norm": 0.9890388250350952, + "learning_rate": 2.311698445307955e-05, + "loss": 0.1822, + "step": 45296 + }, + { + "epoch": 3.143223926167511, + "grad_norm": 0.8539787530899048, + "learning_rate": 2.311339173368692e-05, + "loss": 0.2414, + "step": 45297 + }, + { + "epoch": 3.1432933176046074, + "grad_norm": 0.8201122879981995, + "learning_rate": 2.3109799257016397e-05, + "loss": 0.2536, + "step": 45298 + }, + { + "epoch": 3.1433627090417042, + "grad_norm": 0.8340057134628296, + "learning_rate": 2.3106207023079308e-05, + "loss": 0.2089, + "step": 45299 + }, + { + "epoch": 3.143432100478801, + "grad_norm": 0.8364703059196472, + "learning_rate": 2.3102615031887032e-05, + "loss": 0.153, + "step": 45300 + }, + { + "epoch": 3.1435014919158975, + "grad_norm": 0.8088594675064087, + "learning_rate": 2.3099023283450848e-05, + "loss": 0.3146, + "step": 45301 + }, + { + "epoch": 3.1435708833529943, + "grad_norm": 0.5982499718666077, + "learning_rate": 2.309543177778213e-05, + "loss": 0.2121, + "step": 45302 + }, + { + "epoch": 3.1436402747900907, + "grad_norm": 1.0219082832336426, + "learning_rate": 2.3091840514892217e-05, + "loss": 0.2275, + "step": 45303 + }, + { + "epoch": 3.1437096662271875, + "grad_norm": 0.9068940281867981, + "learning_rate": 2.3088249494792437e-05, + "loss": 0.3423, + "step": 45304 + }, + { + "epoch": 3.1437790576642843, + "grad_norm": 1.111116647720337, + "learning_rate": 2.3084658717494123e-05, + "loss": 0.2958, + "step": 45305 + }, + { + "epoch": 3.1438484491013807, + "grad_norm": 1.003699779510498, + "learning_rate": 2.3081068183008625e-05, + "loss": 0.3329, + "step": 45306 + }, + { + "epoch": 3.1439178405384776, + "grad_norm": 1.6083648204803467, + "learning_rate": 2.3077477891347286e-05, + "loss": 0.3416, + "step": 45307 + }, + { + "epoch": 3.1439872319755744, + "grad_norm": 1.4577215909957886, + "learning_rate": 2.3073887842521413e-05, + "loss": 0.305, + "step": 45308 + }, + { + "epoch": 3.144056623412671, + "grad_norm": 0.961919903755188, + "learning_rate": 2.3070298036542336e-05, + "loss": 0.3532, + "step": 45309 + }, + { + "epoch": 3.1441260148497676, + "grad_norm": 0.9576616287231445, + "learning_rate": 2.3066708473421418e-05, + "loss": 0.3852, + "step": 45310 + }, + { + "epoch": 3.144195406286864, + "grad_norm": 0.9331623315811157, + "learning_rate": 2.3063119153169965e-05, + "loss": 0.2565, + "step": 45311 + }, + { + "epoch": 3.144264797723961, + "grad_norm": 0.6046170592308044, + "learning_rate": 2.305953007579932e-05, + "loss": 0.1936, + "step": 45312 + }, + { + "epoch": 3.1443341891610577, + "grad_norm": 1.0157077312469482, + "learning_rate": 2.305594124132082e-05, + "loss": 0.3512, + "step": 45313 + }, + { + "epoch": 3.144403580598154, + "grad_norm": 0.8194742202758789, + "learning_rate": 2.3052352649745803e-05, + "loss": 0.138, + "step": 45314 + }, + { + "epoch": 3.144472972035251, + "grad_norm": 1.397279977798462, + "learning_rate": 2.3048764301085556e-05, + "loss": 0.3495, + "step": 45315 + }, + { + "epoch": 3.1445423634723477, + "grad_norm": 1.4329781532287598, + "learning_rate": 2.3045176195351438e-05, + "loss": 0.3627, + "step": 45316 + }, + { + "epoch": 3.144611754909444, + "grad_norm": 0.7596926689147949, + "learning_rate": 2.3041588332554764e-05, + "loss": 0.1613, + "step": 45317 + }, + { + "epoch": 3.144681146346541, + "grad_norm": 1.2062604427337646, + "learning_rate": 2.3038000712706897e-05, + "loss": 0.2584, + "step": 45318 + }, + { + "epoch": 3.1447505377836373, + "grad_norm": 0.9869847297668457, + "learning_rate": 2.3034413335819104e-05, + "loss": 0.3953, + "step": 45319 + }, + { + "epoch": 3.144819929220734, + "grad_norm": 0.781278669834137, + "learning_rate": 2.3030826201902745e-05, + "loss": 0.22, + "step": 45320 + }, + { + "epoch": 3.144889320657831, + "grad_norm": 0.7954033613204956, + "learning_rate": 2.3027239310969128e-05, + "loss": 0.2133, + "step": 45321 + }, + { + "epoch": 3.1449587120949274, + "grad_norm": 0.931708037853241, + "learning_rate": 2.3023652663029615e-05, + "loss": 0.2496, + "step": 45322 + }, + { + "epoch": 3.145028103532024, + "grad_norm": 0.738491952419281, + "learning_rate": 2.3020066258095473e-05, + "loss": 0.1898, + "step": 45323 + }, + { + "epoch": 3.1450974949691206, + "grad_norm": 0.9789555668830872, + "learning_rate": 2.301648009617806e-05, + "loss": 0.2371, + "step": 45324 + }, + { + "epoch": 3.1451668864062174, + "grad_norm": 0.867700457572937, + "learning_rate": 2.301289417728868e-05, + "loss": 0.3081, + "step": 45325 + }, + { + "epoch": 3.1452362778433143, + "grad_norm": 1.0047175884246826, + "learning_rate": 2.3009308501438655e-05, + "loss": 0.3368, + "step": 45326 + }, + { + "epoch": 3.1453056692804107, + "grad_norm": 1.1697056293487549, + "learning_rate": 2.3005723068639317e-05, + "loss": 0.32, + "step": 45327 + }, + { + "epoch": 3.1453750607175075, + "grad_norm": 0.7243513464927673, + "learning_rate": 2.300213787890201e-05, + "loss": 0.2875, + "step": 45328 + }, + { + "epoch": 3.145444452154604, + "grad_norm": 1.5858988761901855, + "learning_rate": 2.2998552932237983e-05, + "loss": 0.3237, + "step": 45329 + }, + { + "epoch": 3.1455138435917007, + "grad_norm": 0.7275933623313904, + "learning_rate": 2.29949682286586e-05, + "loss": 0.1959, + "step": 45330 + }, + { + "epoch": 3.1455832350287976, + "grad_norm": 1.0570045709609985, + "learning_rate": 2.2991383768175157e-05, + "loss": 0.3756, + "step": 45331 + }, + { + "epoch": 3.145652626465894, + "grad_norm": 0.8007593154907227, + "learning_rate": 2.2987799550798984e-05, + "loss": 0.2369, + "step": 45332 + }, + { + "epoch": 3.1457220179029908, + "grad_norm": 0.8437095284461975, + "learning_rate": 2.2984215576541402e-05, + "loss": 0.2515, + "step": 45333 + }, + { + "epoch": 3.1457914093400876, + "grad_norm": 1.2756050825119019, + "learning_rate": 2.29806318454137e-05, + "loss": 0.3403, + "step": 45334 + }, + { + "epoch": 3.145860800777184, + "grad_norm": 1.164544939994812, + "learning_rate": 2.2977048357427255e-05, + "loss": 0.3259, + "step": 45335 + }, + { + "epoch": 3.145930192214281, + "grad_norm": 1.052635908126831, + "learning_rate": 2.2973465112593285e-05, + "loss": 0.1913, + "step": 45336 + }, + { + "epoch": 3.145999583651377, + "grad_norm": 0.8165395855903625, + "learning_rate": 2.2969882110923158e-05, + "loss": 0.2386, + "step": 45337 + }, + { + "epoch": 3.146068975088474, + "grad_norm": 0.8931487798690796, + "learning_rate": 2.2966299352428178e-05, + "loss": 0.1915, + "step": 45338 + }, + { + "epoch": 3.146138366525571, + "grad_norm": 1.310701847076416, + "learning_rate": 2.2962716837119656e-05, + "loss": 0.265, + "step": 45339 + }, + { + "epoch": 3.1462077579626673, + "grad_norm": 0.7144867181777954, + "learning_rate": 2.2959134565008887e-05, + "loss": 0.114, + "step": 45340 + }, + { + "epoch": 3.146277149399764, + "grad_norm": 0.7740366458892822, + "learning_rate": 2.2955552536107206e-05, + "loss": 0.1797, + "step": 45341 + }, + { + "epoch": 3.146346540836861, + "grad_norm": 0.8451919555664062, + "learning_rate": 2.295197075042592e-05, + "loss": 0.2455, + "step": 45342 + }, + { + "epoch": 3.1464159322739573, + "grad_norm": 0.9435306787490845, + "learning_rate": 2.2948389207976308e-05, + "loss": 0.2143, + "step": 45343 + }, + { + "epoch": 3.146485323711054, + "grad_norm": 0.8300119638442993, + "learning_rate": 2.2944807908769682e-05, + "loss": 0.1612, + "step": 45344 + }, + { + "epoch": 3.1465547151481505, + "grad_norm": 1.0155060291290283, + "learning_rate": 2.2941226852817367e-05, + "loss": 0.3727, + "step": 45345 + }, + { + "epoch": 3.1466241065852474, + "grad_norm": 0.9031139612197876, + "learning_rate": 2.2937646040130657e-05, + "loss": 0.28, + "step": 45346 + }, + { + "epoch": 3.146693498022344, + "grad_norm": 0.823401689529419, + "learning_rate": 2.2934065470720846e-05, + "loss": 0.2555, + "step": 45347 + }, + { + "epoch": 3.1467628894594406, + "grad_norm": 1.011946439743042, + "learning_rate": 2.2930485144599257e-05, + "loss": 0.2468, + "step": 45348 + }, + { + "epoch": 3.1468322808965374, + "grad_norm": 0.9464555978775024, + "learning_rate": 2.2926905061777216e-05, + "loss": 0.3302, + "step": 45349 + }, + { + "epoch": 3.146901672333634, + "grad_norm": 1.7310073375701904, + "learning_rate": 2.2923325222265968e-05, + "loss": 0.4581, + "step": 45350 + }, + { + "epoch": 3.1469710637707307, + "grad_norm": 1.960919976234436, + "learning_rate": 2.2919745626076827e-05, + "loss": 0.2891, + "step": 45351 + }, + { + "epoch": 3.1470404552078275, + "grad_norm": 1.0785878896713257, + "learning_rate": 2.291616627322112e-05, + "loss": 0.2467, + "step": 45352 + }, + { + "epoch": 3.147109846644924, + "grad_norm": 0.770562469959259, + "learning_rate": 2.2912587163710153e-05, + "loss": 0.1559, + "step": 45353 + }, + { + "epoch": 3.1471792380820207, + "grad_norm": 1.2886419296264648, + "learning_rate": 2.290900829755517e-05, + "loss": 0.5095, + "step": 45354 + }, + { + "epoch": 3.1472486295191175, + "grad_norm": 0.8279454708099365, + "learning_rate": 2.290542967476751e-05, + "loss": 0.2202, + "step": 45355 + }, + { + "epoch": 3.147318020956214, + "grad_norm": 0.8258597254753113, + "learning_rate": 2.290185129535847e-05, + "loss": 0.2012, + "step": 45356 + }, + { + "epoch": 3.1473874123933108, + "grad_norm": 1.1927026510238647, + "learning_rate": 2.2898273159339356e-05, + "loss": 0.33, + "step": 45357 + }, + { + "epoch": 3.147456803830407, + "grad_norm": 0.9020527005195618, + "learning_rate": 2.2894695266721435e-05, + "loss": 0.3242, + "step": 45358 + }, + { + "epoch": 3.147526195267504, + "grad_norm": 1.0309908390045166, + "learning_rate": 2.2891117617516e-05, + "loss": 0.1646, + "step": 45359 + }, + { + "epoch": 3.147595586704601, + "grad_norm": 0.9446073174476624, + "learning_rate": 2.288754021173437e-05, + "loss": 0.1835, + "step": 45360 + }, + { + "epoch": 3.147664978141697, + "grad_norm": 1.0422394275665283, + "learning_rate": 2.2883963049387823e-05, + "loss": 0.212, + "step": 45361 + }, + { + "epoch": 3.147734369578794, + "grad_norm": 1.102561354637146, + "learning_rate": 2.2880386130487696e-05, + "loss": 0.2624, + "step": 45362 + }, + { + "epoch": 3.1478037610158904, + "grad_norm": 1.051164984703064, + "learning_rate": 2.287680945504521e-05, + "loss": 0.2425, + "step": 45363 + }, + { + "epoch": 3.1478731524529873, + "grad_norm": 0.7574042081832886, + "learning_rate": 2.2873233023071683e-05, + "loss": 0.2266, + "step": 45364 + }, + { + "epoch": 3.147942543890084, + "grad_norm": 0.7398970723152161, + "learning_rate": 2.2869656834578414e-05, + "loss": 0.1993, + "step": 45365 + }, + { + "epoch": 3.1480119353271805, + "grad_norm": 0.7765833139419556, + "learning_rate": 2.286608088957669e-05, + "loss": 0.2462, + "step": 45366 + }, + { + "epoch": 3.1480813267642773, + "grad_norm": 0.8780566453933716, + "learning_rate": 2.2862505188077798e-05, + "loss": 0.1491, + "step": 45367 + }, + { + "epoch": 3.148150718201374, + "grad_norm": 0.8427233695983887, + "learning_rate": 2.2858929730093027e-05, + "loss": 0.1872, + "step": 45368 + }, + { + "epoch": 3.1482201096384705, + "grad_norm": 1.0841645002365112, + "learning_rate": 2.2855354515633698e-05, + "loss": 0.2744, + "step": 45369 + }, + { + "epoch": 3.1482895010755674, + "grad_norm": 0.9153153896331787, + "learning_rate": 2.2851779544711028e-05, + "loss": 0.2334, + "step": 45370 + }, + { + "epoch": 3.1483588925126638, + "grad_norm": 0.7297101616859436, + "learning_rate": 2.284820481733635e-05, + "loss": 0.185, + "step": 45371 + }, + { + "epoch": 3.1484282839497606, + "grad_norm": 0.6857261061668396, + "learning_rate": 2.2844630333520934e-05, + "loss": 0.2414, + "step": 45372 + }, + { + "epoch": 3.1484976753868574, + "grad_norm": 1.0133473873138428, + "learning_rate": 2.2841056093276058e-05, + "loss": 0.1812, + "step": 45373 + }, + { + "epoch": 3.148567066823954, + "grad_norm": 0.800318717956543, + "learning_rate": 2.283748209661303e-05, + "loss": 0.1986, + "step": 45374 + }, + { + "epoch": 3.1486364582610507, + "grad_norm": 0.8571597337722778, + "learning_rate": 2.283390834354311e-05, + "loss": 0.2243, + "step": 45375 + }, + { + "epoch": 3.148705849698147, + "grad_norm": 1.1101669073104858, + "learning_rate": 2.2830334834077626e-05, + "loss": 0.3319, + "step": 45376 + }, + { + "epoch": 3.148775241135244, + "grad_norm": 0.9056833386421204, + "learning_rate": 2.2826761568227783e-05, + "loss": 0.3003, + "step": 45377 + }, + { + "epoch": 3.1488446325723407, + "grad_norm": 0.8952959179878235, + "learning_rate": 2.2823188546004904e-05, + "loss": 0.2163, + "step": 45378 + }, + { + "epoch": 3.148914024009437, + "grad_norm": 1.2184830904006958, + "learning_rate": 2.2819615767420265e-05, + "loss": 0.2245, + "step": 45379 + }, + { + "epoch": 3.148983415446534, + "grad_norm": 1.2661263942718506, + "learning_rate": 2.281604323248514e-05, + "loss": 0.4178, + "step": 45380 + }, + { + "epoch": 3.1490528068836308, + "grad_norm": 1.7071685791015625, + "learning_rate": 2.2812470941210817e-05, + "loss": 0.1834, + "step": 45381 + }, + { + "epoch": 3.149122198320727, + "grad_norm": 1.2225277423858643, + "learning_rate": 2.2808898893608567e-05, + "loss": 0.2858, + "step": 45382 + }, + { + "epoch": 3.149191589757824, + "grad_norm": 1.0979012250900269, + "learning_rate": 2.28053270896897e-05, + "loss": 0.2563, + "step": 45383 + }, + { + "epoch": 3.1492609811949204, + "grad_norm": 0.9039633870124817, + "learning_rate": 2.2801755529465417e-05, + "loss": 0.2888, + "step": 45384 + }, + { + "epoch": 3.149330372632017, + "grad_norm": 1.2409772872924805, + "learning_rate": 2.279818421294705e-05, + "loss": 0.2166, + "step": 45385 + }, + { + "epoch": 3.149399764069114, + "grad_norm": 0.8356349468231201, + "learning_rate": 2.279461314014585e-05, + "loss": 0.2426, + "step": 45386 + }, + { + "epoch": 3.1494691555062104, + "grad_norm": 0.864582359790802, + "learning_rate": 2.2791042311073098e-05, + "loss": 0.324, + "step": 45387 + }, + { + "epoch": 3.1495385469433073, + "grad_norm": 0.671729326248169, + "learning_rate": 2.2787471725740105e-05, + "loss": 0.176, + "step": 45388 + }, + { + "epoch": 3.149607938380404, + "grad_norm": 1.031667709350586, + "learning_rate": 2.2783901384158068e-05, + "loss": 0.3934, + "step": 45389 + }, + { + "epoch": 3.1496773298175005, + "grad_norm": 1.5104016065597534, + "learning_rate": 2.2780331286338296e-05, + "loss": 0.3793, + "step": 45390 + }, + { + "epoch": 3.1497467212545973, + "grad_norm": 0.92161625623703, + "learning_rate": 2.2776761432292093e-05, + "loss": 0.2379, + "step": 45391 + }, + { + "epoch": 3.1498161126916937, + "grad_norm": 0.9447241425514221, + "learning_rate": 2.277319182203066e-05, + "loss": 0.3059, + "step": 45392 + }, + { + "epoch": 3.1498855041287905, + "grad_norm": 1.0351393222808838, + "learning_rate": 2.276962245556531e-05, + "loss": 0.223, + "step": 45393 + }, + { + "epoch": 3.1499548955658874, + "grad_norm": 0.9463663101196289, + "learning_rate": 2.276605333290729e-05, + "loss": 0.2007, + "step": 45394 + }, + { + "epoch": 3.1500242870029838, + "grad_norm": 0.6688248515129089, + "learning_rate": 2.276248445406789e-05, + "loss": 0.1759, + "step": 45395 + }, + { + "epoch": 3.1500936784400806, + "grad_norm": 1.0075126886367798, + "learning_rate": 2.2758915819058357e-05, + "loss": 0.2838, + "step": 45396 + }, + { + "epoch": 3.150163069877177, + "grad_norm": 0.7652144432067871, + "learning_rate": 2.2755347427889993e-05, + "loss": 0.2219, + "step": 45397 + }, + { + "epoch": 3.150232461314274, + "grad_norm": 1.2722543478012085, + "learning_rate": 2.2751779280574016e-05, + "loss": 0.4374, + "step": 45398 + }, + { + "epoch": 3.1503018527513706, + "grad_norm": 0.6974453330039978, + "learning_rate": 2.2748211377121698e-05, + "loss": 0.2412, + "step": 45399 + }, + { + "epoch": 3.150371244188467, + "grad_norm": 0.754098117351532, + "learning_rate": 2.2744643717544324e-05, + "loss": 0.2244, + "step": 45400 + }, + { + "epoch": 3.150440635625564, + "grad_norm": 0.9955126643180847, + "learning_rate": 2.2741076301853136e-05, + "loss": 0.2714, + "step": 45401 + }, + { + "epoch": 3.1505100270626603, + "grad_norm": 0.7954394221305847, + "learning_rate": 2.273750913005941e-05, + "loss": 0.2214, + "step": 45402 + }, + { + "epoch": 3.150579418499757, + "grad_norm": 0.7911931872367859, + "learning_rate": 2.273394220217441e-05, + "loss": 0.2768, + "step": 45403 + }, + { + "epoch": 3.150648809936854, + "grad_norm": 1.0314821004867554, + "learning_rate": 2.2730375518209414e-05, + "loss": 0.3055, + "step": 45404 + }, + { + "epoch": 3.1507182013739503, + "grad_norm": 0.7930801510810852, + "learning_rate": 2.2726809078175627e-05, + "loss": 0.2323, + "step": 45405 + }, + { + "epoch": 3.150787592811047, + "grad_norm": 0.6644982099533081, + "learning_rate": 2.272324288208434e-05, + "loss": 0.1374, + "step": 45406 + }, + { + "epoch": 3.150856984248144, + "grad_norm": 1.1652576923370361, + "learning_rate": 2.271967692994681e-05, + "loss": 0.4433, + "step": 45407 + }, + { + "epoch": 3.1509263756852404, + "grad_norm": 1.3835400342941284, + "learning_rate": 2.2716111221774294e-05, + "loss": 0.3176, + "step": 45408 + }, + { + "epoch": 3.150995767122337, + "grad_norm": 1.0576202869415283, + "learning_rate": 2.271254575757804e-05, + "loss": 0.2951, + "step": 45409 + }, + { + "epoch": 3.1510651585594336, + "grad_norm": 1.2678133249282837, + "learning_rate": 2.2708980537369327e-05, + "loss": 0.3419, + "step": 45410 + }, + { + "epoch": 3.1511345499965304, + "grad_norm": 1.1750335693359375, + "learning_rate": 2.2705415561159414e-05, + "loss": 0.3578, + "step": 45411 + }, + { + "epoch": 3.1512039414336273, + "grad_norm": 0.987168550491333, + "learning_rate": 2.2701850828959514e-05, + "loss": 0.3247, + "step": 45412 + }, + { + "epoch": 3.1512733328707236, + "grad_norm": 0.9042500257492065, + "learning_rate": 2.2698286340780895e-05, + "loss": 0.3759, + "step": 45413 + }, + { + "epoch": 3.1513427243078205, + "grad_norm": 0.8628413081169128, + "learning_rate": 2.2694722096634823e-05, + "loss": 0.3003, + "step": 45414 + }, + { + "epoch": 3.1514121157449173, + "grad_norm": 0.7904931902885437, + "learning_rate": 2.269115809653254e-05, + "loss": 0.2232, + "step": 45415 + }, + { + "epoch": 3.1514815071820137, + "grad_norm": 0.8217305541038513, + "learning_rate": 2.2687594340485306e-05, + "loss": 0.3007, + "step": 45416 + }, + { + "epoch": 3.1515508986191105, + "grad_norm": 0.8321089744567871, + "learning_rate": 2.2684030828504366e-05, + "loss": 0.2316, + "step": 45417 + }, + { + "epoch": 3.151620290056207, + "grad_norm": 0.774235725402832, + "learning_rate": 2.2680467560600994e-05, + "loss": 0.1751, + "step": 45418 + }, + { + "epoch": 3.1516896814933038, + "grad_norm": 0.9480097889900208, + "learning_rate": 2.2676904536786392e-05, + "loss": 0.3058, + "step": 45419 + }, + { + "epoch": 3.1517590729304006, + "grad_norm": 0.9270824790000916, + "learning_rate": 2.2673341757071832e-05, + "loss": 0.2321, + "step": 45420 + }, + { + "epoch": 3.151828464367497, + "grad_norm": 0.9031057953834534, + "learning_rate": 2.2669779221468555e-05, + "loss": 0.2651, + "step": 45421 + }, + { + "epoch": 3.151897855804594, + "grad_norm": 0.5116531848907471, + "learning_rate": 2.2666216929987817e-05, + "loss": 0.1354, + "step": 45422 + }, + { + "epoch": 3.15196724724169, + "grad_norm": 0.8550524115562439, + "learning_rate": 2.2662654882640887e-05, + "loss": 0.3235, + "step": 45423 + }, + { + "epoch": 3.152036638678787, + "grad_norm": 0.7699903249740601, + "learning_rate": 2.2659093079438953e-05, + "loss": 0.2002, + "step": 45424 + }, + { + "epoch": 3.152106030115884, + "grad_norm": 1.0170938968658447, + "learning_rate": 2.265553152039329e-05, + "loss": 0.3088, + "step": 45425 + }, + { + "epoch": 3.1521754215529803, + "grad_norm": 0.7134917974472046, + "learning_rate": 2.265197020551517e-05, + "loss": 0.164, + "step": 45426 + }, + { + "epoch": 3.152244812990077, + "grad_norm": 0.9867449402809143, + "learning_rate": 2.2648409134815773e-05, + "loss": 0.319, + "step": 45427 + }, + { + "epoch": 3.1523142044271735, + "grad_norm": 0.6967678070068359, + "learning_rate": 2.2644848308306378e-05, + "loss": 0.1735, + "step": 45428 + }, + { + "epoch": 3.1523835958642703, + "grad_norm": 0.8893631100654602, + "learning_rate": 2.2641287725998228e-05, + "loss": 0.2963, + "step": 45429 + }, + { + "epoch": 3.152452987301367, + "grad_norm": 0.9428272247314453, + "learning_rate": 2.2637727387902552e-05, + "loss": 0.2754, + "step": 45430 + }, + { + "epoch": 3.1525223787384635, + "grad_norm": 1.1332348585128784, + "learning_rate": 2.263416729403062e-05, + "loss": 0.2876, + "step": 45431 + }, + { + "epoch": 3.1525917701755604, + "grad_norm": 0.831504225730896, + "learning_rate": 2.2630607444393615e-05, + "loss": 0.2602, + "step": 45432 + }, + { + "epoch": 3.152661161612657, + "grad_norm": 1.160240650177002, + "learning_rate": 2.262704783900281e-05, + "loss": 0.3659, + "step": 45433 + }, + { + "epoch": 3.1527305530497536, + "grad_norm": 0.6446803212165833, + "learning_rate": 2.2623488477869438e-05, + "loss": 0.2083, + "step": 45434 + }, + { + "epoch": 3.1527999444868504, + "grad_norm": 1.0314784049987793, + "learning_rate": 2.2619929361004744e-05, + "loss": 0.2644, + "step": 45435 + }, + { + "epoch": 3.152869335923947, + "grad_norm": 1.128151774406433, + "learning_rate": 2.2616370488419947e-05, + "loss": 0.3449, + "step": 45436 + }, + { + "epoch": 3.1529387273610436, + "grad_norm": 0.8984603881835938, + "learning_rate": 2.2612811860126282e-05, + "loss": 0.165, + "step": 45437 + }, + { + "epoch": 3.1530081187981405, + "grad_norm": 1.0163809061050415, + "learning_rate": 2.260925347613503e-05, + "loss": 0.3926, + "step": 45438 + }, + { + "epoch": 3.153077510235237, + "grad_norm": 0.8120734691619873, + "learning_rate": 2.2605695336457356e-05, + "loss": 0.3641, + "step": 45439 + }, + { + "epoch": 3.1531469016723337, + "grad_norm": 0.8166589736938477, + "learning_rate": 2.260213744110452e-05, + "loss": 0.2502, + "step": 45440 + }, + { + "epoch": 3.1532162931094305, + "grad_norm": 0.6487416625022888, + "learning_rate": 2.259857979008776e-05, + "loss": 0.1659, + "step": 45441 + }, + { + "epoch": 3.153285684546527, + "grad_norm": 0.543363630771637, + "learning_rate": 2.2595022383418306e-05, + "loss": 0.1397, + "step": 45442 + }, + { + "epoch": 3.1533550759836237, + "grad_norm": 0.9439370632171631, + "learning_rate": 2.2591465221107376e-05, + "loss": 0.1685, + "step": 45443 + }, + { + "epoch": 3.15342446742072, + "grad_norm": 1.0838062763214111, + "learning_rate": 2.2587908303166217e-05, + "loss": 0.211, + "step": 45444 + }, + { + "epoch": 3.153493858857817, + "grad_norm": 1.1341009140014648, + "learning_rate": 2.258435162960608e-05, + "loss": 0.327, + "step": 45445 + }, + { + "epoch": 3.153563250294914, + "grad_norm": 0.8706459403038025, + "learning_rate": 2.258079520043813e-05, + "loss": 0.2608, + "step": 45446 + }, + { + "epoch": 3.15363264173201, + "grad_norm": 0.7719771862030029, + "learning_rate": 2.2577239015673624e-05, + "loss": 0.2025, + "step": 45447 + }, + { + "epoch": 3.153702033169107, + "grad_norm": 0.7700733542442322, + "learning_rate": 2.2573683075323805e-05, + "loss": 0.1965, + "step": 45448 + }, + { + "epoch": 3.1537714246062034, + "grad_norm": 0.9593619704246521, + "learning_rate": 2.2570127379399874e-05, + "loss": 0.2343, + "step": 45449 + }, + { + "epoch": 3.1538408160433002, + "grad_norm": 0.6724267601966858, + "learning_rate": 2.2566571927913073e-05, + "loss": 0.2288, + "step": 45450 + }, + { + "epoch": 3.153910207480397, + "grad_norm": 0.9483231902122498, + "learning_rate": 2.2563016720874608e-05, + "loss": 0.1992, + "step": 45451 + }, + { + "epoch": 3.1539795989174935, + "grad_norm": 0.6427388787269592, + "learning_rate": 2.2559461758295753e-05, + "loss": 0.1817, + "step": 45452 + }, + { + "epoch": 3.1540489903545903, + "grad_norm": 0.5479657649993896, + "learning_rate": 2.2555907040187662e-05, + "loss": 0.212, + "step": 45453 + }, + { + "epoch": 3.1541183817916867, + "grad_norm": 0.8156490921974182, + "learning_rate": 2.2552352566561595e-05, + "loss": 0.2705, + "step": 45454 + }, + { + "epoch": 3.1541877732287835, + "grad_norm": 0.8236250877380371, + "learning_rate": 2.2548798337428756e-05, + "loss": 0.3092, + "step": 45455 + }, + { + "epoch": 3.1542571646658804, + "grad_norm": 0.8414537906646729, + "learning_rate": 2.2545244352800376e-05, + "loss": 0.1795, + "step": 45456 + }, + { + "epoch": 3.1543265561029767, + "grad_norm": 0.9466597437858582, + "learning_rate": 2.2541690612687703e-05, + "loss": 0.2794, + "step": 45457 + }, + { + "epoch": 3.1543959475400736, + "grad_norm": 0.8659215569496155, + "learning_rate": 2.25381371171019e-05, + "loss": 0.3559, + "step": 45458 + }, + { + "epoch": 3.1544653389771704, + "grad_norm": 0.6293696165084839, + "learning_rate": 2.2534583866054203e-05, + "loss": 0.2347, + "step": 45459 + }, + { + "epoch": 3.154534730414267, + "grad_norm": 1.295999526977539, + "learning_rate": 2.2531030859555845e-05, + "loss": 0.3478, + "step": 45460 + }, + { + "epoch": 3.1546041218513636, + "grad_norm": 0.9530354142189026, + "learning_rate": 2.252747809761806e-05, + "loss": 0.3454, + "step": 45461 + }, + { + "epoch": 3.15467351328846, + "grad_norm": 0.753944456577301, + "learning_rate": 2.2523925580252014e-05, + "loss": 0.2292, + "step": 45462 + }, + { + "epoch": 3.154742904725557, + "grad_norm": 1.236782431602478, + "learning_rate": 2.2520373307468946e-05, + "loss": 0.2204, + "step": 45463 + }, + { + "epoch": 3.1548122961626537, + "grad_norm": 1.5895366668701172, + "learning_rate": 2.2516821279280066e-05, + "loss": 0.3155, + "step": 45464 + }, + { + "epoch": 3.15488168759975, + "grad_norm": 0.7258375883102417, + "learning_rate": 2.2513269495696588e-05, + "loss": 0.1531, + "step": 45465 + }, + { + "epoch": 3.154951079036847, + "grad_norm": 1.6342970132827759, + "learning_rate": 2.250971795672976e-05, + "loss": 0.4193, + "step": 45466 + }, + { + "epoch": 3.1550204704739437, + "grad_norm": 0.9894511699676514, + "learning_rate": 2.2506166662390737e-05, + "loss": 0.2597, + "step": 45467 + }, + { + "epoch": 3.15508986191104, + "grad_norm": 0.8764106631278992, + "learning_rate": 2.2502615612690747e-05, + "loss": 0.2163, + "step": 45468 + }, + { + "epoch": 3.155159253348137, + "grad_norm": 0.6684442162513733, + "learning_rate": 2.2499064807641014e-05, + "loss": 0.1414, + "step": 45469 + }, + { + "epoch": 3.1552286447852333, + "grad_norm": 1.060712218284607, + "learning_rate": 2.2495514247252746e-05, + "loss": 0.3643, + "step": 45470 + }, + { + "epoch": 3.15529803622233, + "grad_norm": 0.6883909106254578, + "learning_rate": 2.249196393153714e-05, + "loss": 0.2048, + "step": 45471 + }, + { + "epoch": 3.155367427659427, + "grad_norm": 0.7905903458595276, + "learning_rate": 2.2488413860505408e-05, + "loss": 0.2477, + "step": 45472 + }, + { + "epoch": 3.1554368190965234, + "grad_norm": 1.0432173013687134, + "learning_rate": 2.2484864034168794e-05, + "loss": 0.2238, + "step": 45473 + }, + { + "epoch": 3.1555062105336202, + "grad_norm": 1.1487880945205688, + "learning_rate": 2.2481314452538448e-05, + "loss": 0.3891, + "step": 45474 + }, + { + "epoch": 3.1555756019707166, + "grad_norm": 0.5955283045768738, + "learning_rate": 2.2477765115625592e-05, + "loss": 0.2178, + "step": 45475 + }, + { + "epoch": 3.1556449934078135, + "grad_norm": 1.570049524307251, + "learning_rate": 2.247421602344144e-05, + "loss": 0.1843, + "step": 45476 + }, + { + "epoch": 3.1557143848449103, + "grad_norm": 0.9297417402267456, + "learning_rate": 2.2470667175997195e-05, + "loss": 0.1513, + "step": 45477 + }, + { + "epoch": 3.1557837762820067, + "grad_norm": 0.8424502611160278, + "learning_rate": 2.246711857330405e-05, + "loss": 0.2331, + "step": 45478 + }, + { + "epoch": 3.1558531677191035, + "grad_norm": 0.9960528612136841, + "learning_rate": 2.246357021537322e-05, + "loss": 0.2873, + "step": 45479 + }, + { + "epoch": 3.1559225591562, + "grad_norm": 0.825839102268219, + "learning_rate": 2.246002210221594e-05, + "loss": 0.1896, + "step": 45480 + }, + { + "epoch": 3.1559919505932967, + "grad_norm": 1.1040228605270386, + "learning_rate": 2.245647423384334e-05, + "loss": 0.4534, + "step": 45481 + }, + { + "epoch": 3.1560613420303936, + "grad_norm": 0.8422759771347046, + "learning_rate": 2.2452926610266646e-05, + "loss": 0.1832, + "step": 45482 + }, + { + "epoch": 3.15613073346749, + "grad_norm": 0.6640983819961548, + "learning_rate": 2.244937923149707e-05, + "loss": 0.2061, + "step": 45483 + }, + { + "epoch": 3.156200124904587, + "grad_norm": 0.8112927675247192, + "learning_rate": 2.2445832097545817e-05, + "loss": 0.2306, + "step": 45484 + }, + { + "epoch": 3.1562695163416836, + "grad_norm": 0.7938560247421265, + "learning_rate": 2.2442285208424064e-05, + "loss": 0.19, + "step": 45485 + }, + { + "epoch": 3.15633890777878, + "grad_norm": 1.4311950206756592, + "learning_rate": 2.2438738564143014e-05, + "loss": 0.2843, + "step": 45486 + }, + { + "epoch": 3.156408299215877, + "grad_norm": 0.7827353477478027, + "learning_rate": 2.24351921647139e-05, + "loss": 0.2069, + "step": 45487 + }, + { + "epoch": 3.1564776906529732, + "grad_norm": 0.9119069576263428, + "learning_rate": 2.2431646010147868e-05, + "loss": 0.2266, + "step": 45488 + }, + { + "epoch": 3.15654708209007, + "grad_norm": 0.7587918639183044, + "learning_rate": 2.242810010045612e-05, + "loss": 0.2202, + "step": 45489 + }, + { + "epoch": 3.156616473527167, + "grad_norm": 1.0652730464935303, + "learning_rate": 2.2424554435649858e-05, + "loss": 0.2028, + "step": 45490 + }, + { + "epoch": 3.1566858649642633, + "grad_norm": 0.6134592294692993, + "learning_rate": 2.242100901574028e-05, + "loss": 0.1412, + "step": 45491 + }, + { + "epoch": 3.15675525640136, + "grad_norm": 0.94712895154953, + "learning_rate": 2.2417463840738593e-05, + "loss": 0.3812, + "step": 45492 + }, + { + "epoch": 3.156824647838457, + "grad_norm": 0.8382983207702637, + "learning_rate": 2.2413918910655962e-05, + "loss": 0.1916, + "step": 45493 + }, + { + "epoch": 3.1568940392755533, + "grad_norm": 1.3749451637268066, + "learning_rate": 2.2410374225503573e-05, + "loss": 0.3072, + "step": 45494 + }, + { + "epoch": 3.15696343071265, + "grad_norm": 2.3451061248779297, + "learning_rate": 2.2406829785292637e-05, + "loss": 0.3641, + "step": 45495 + }, + { + "epoch": 3.1570328221497466, + "grad_norm": 0.9336406588554382, + "learning_rate": 2.2403285590034374e-05, + "loss": 0.2785, + "step": 45496 + }, + { + "epoch": 3.1571022135868434, + "grad_norm": 0.6111760139465332, + "learning_rate": 2.239974163973989e-05, + "loss": 0.2238, + "step": 45497 + }, + { + "epoch": 3.1571716050239402, + "grad_norm": 0.8080621957778931, + "learning_rate": 2.2396197934420428e-05, + "loss": 0.3315, + "step": 45498 + }, + { + "epoch": 3.1572409964610366, + "grad_norm": 0.7489392757415771, + "learning_rate": 2.239265447408716e-05, + "loss": 0.2184, + "step": 45499 + }, + { + "epoch": 3.1573103878981335, + "grad_norm": 0.6978267431259155, + "learning_rate": 2.238911125875128e-05, + "loss": 0.1801, + "step": 45500 + }, + { + "epoch": 3.15737977933523, + "grad_norm": 1.1097278594970703, + "learning_rate": 2.238556828842401e-05, + "loss": 0.2709, + "step": 45501 + }, + { + "epoch": 3.1574491707723267, + "grad_norm": 1.1067214012145996, + "learning_rate": 2.238202556311646e-05, + "loss": 0.3709, + "step": 45502 + }, + { + "epoch": 3.1575185622094235, + "grad_norm": 0.7583543062210083, + "learning_rate": 2.2378483082839842e-05, + "loss": 0.2705, + "step": 45503 + }, + { + "epoch": 3.15758795364652, + "grad_norm": 1.092673420906067, + "learning_rate": 2.2374940847605353e-05, + "loss": 0.3445, + "step": 45504 + }, + { + "epoch": 3.1576573450836167, + "grad_norm": 0.7074288129806519, + "learning_rate": 2.2371398857424177e-05, + "loss": 0.1909, + "step": 45505 + }, + { + "epoch": 3.157726736520713, + "grad_norm": 1.031154990196228, + "learning_rate": 2.2367857112307477e-05, + "loss": 0.324, + "step": 45506 + }, + { + "epoch": 3.15779612795781, + "grad_norm": 0.8643720149993896, + "learning_rate": 2.2364315612266483e-05, + "loss": 0.2531, + "step": 45507 + }, + { + "epoch": 3.157865519394907, + "grad_norm": 1.1890181303024292, + "learning_rate": 2.2360774357312297e-05, + "loss": 0.2753, + "step": 45508 + }, + { + "epoch": 3.157934910832003, + "grad_norm": 1.1228106021881104, + "learning_rate": 2.235723334745615e-05, + "loss": 0.3752, + "step": 45509 + }, + { + "epoch": 3.1580043022691, + "grad_norm": 0.7640299797058105, + "learning_rate": 2.235369258270921e-05, + "loss": 0.1961, + "step": 45510 + }, + { + "epoch": 3.158073693706197, + "grad_norm": 0.7925587296485901, + "learning_rate": 2.235015206308264e-05, + "loss": 0.3375, + "step": 45511 + }, + { + "epoch": 3.1581430851432932, + "grad_norm": 0.8596439361572266, + "learning_rate": 2.2346611788587647e-05, + "loss": 0.2731, + "step": 45512 + }, + { + "epoch": 3.15821247658039, + "grad_norm": 1.0413686037063599, + "learning_rate": 2.234307175923538e-05, + "loss": 0.3428, + "step": 45513 + }, + { + "epoch": 3.1582818680174864, + "grad_norm": 0.9173212051391602, + "learning_rate": 2.2339531975037054e-05, + "loss": 0.2332, + "step": 45514 + }, + { + "epoch": 3.1583512594545833, + "grad_norm": 0.810732364654541, + "learning_rate": 2.2335992436003795e-05, + "loss": 0.2328, + "step": 45515 + }, + { + "epoch": 3.15842065089168, + "grad_norm": 0.8330888152122498, + "learning_rate": 2.233245314214679e-05, + "loss": 0.2376, + "step": 45516 + }, + { + "epoch": 3.1584900423287765, + "grad_norm": 0.8500552773475647, + "learning_rate": 2.2328914093477215e-05, + "loss": 0.343, + "step": 45517 + }, + { + "epoch": 3.1585594337658733, + "grad_norm": 0.877623975276947, + "learning_rate": 2.2325375290006256e-05, + "loss": 0.2272, + "step": 45518 + }, + { + "epoch": 3.15862882520297, + "grad_norm": 0.7800691723823547, + "learning_rate": 2.232183673174507e-05, + "loss": 0.2231, + "step": 45519 + }, + { + "epoch": 3.1586982166400666, + "grad_norm": 0.8291434645652771, + "learning_rate": 2.2318298418704843e-05, + "loss": 0.2988, + "step": 45520 + }, + { + "epoch": 3.1587676080771634, + "grad_norm": 0.7674264311790466, + "learning_rate": 2.2314760350896757e-05, + "loss": 0.2853, + "step": 45521 + }, + { + "epoch": 3.15883699951426, + "grad_norm": 4.803855895996094, + "learning_rate": 2.2311222528331933e-05, + "loss": 0.2599, + "step": 45522 + }, + { + "epoch": 3.1589063909513566, + "grad_norm": 0.9756852984428406, + "learning_rate": 2.2307684951021555e-05, + "loss": 0.3087, + "step": 45523 + }, + { + "epoch": 3.1589757823884534, + "grad_norm": 0.6736242771148682, + "learning_rate": 2.230414761897681e-05, + "loss": 0.225, + "step": 45524 + }, + { + "epoch": 3.15904517382555, + "grad_norm": 0.9393788576126099, + "learning_rate": 2.2300610532208855e-05, + "loss": 0.1902, + "step": 45525 + }, + { + "epoch": 3.1591145652626467, + "grad_norm": 0.9004343748092651, + "learning_rate": 2.2297073690728863e-05, + "loss": 0.2721, + "step": 45526 + }, + { + "epoch": 3.159183956699743, + "grad_norm": 0.9727028012275696, + "learning_rate": 2.2293537094548024e-05, + "loss": 0.3565, + "step": 45527 + }, + { + "epoch": 3.15925334813684, + "grad_norm": 0.9676699638366699, + "learning_rate": 2.229000074367743e-05, + "loss": 0.2078, + "step": 45528 + }, + { + "epoch": 3.1593227395739367, + "grad_norm": 0.8863275051116943, + "learning_rate": 2.2286464638128302e-05, + "loss": 0.2213, + "step": 45529 + }, + { + "epoch": 3.159392131011033, + "grad_norm": 0.9265686273574829, + "learning_rate": 2.2282928777911782e-05, + "loss": 0.289, + "step": 45530 + }, + { + "epoch": 3.15946152244813, + "grad_norm": 0.944495439529419, + "learning_rate": 2.2279393163039063e-05, + "loss": 0.2996, + "step": 45531 + }, + { + "epoch": 3.1595309138852263, + "grad_norm": 0.928332507610321, + "learning_rate": 2.2275857793521272e-05, + "loss": 0.3732, + "step": 45532 + }, + { + "epoch": 3.159600305322323, + "grad_norm": 1.1976975202560425, + "learning_rate": 2.2272322669369562e-05, + "loss": 0.3016, + "step": 45533 + }, + { + "epoch": 3.15966969675942, + "grad_norm": 1.252602458000183, + "learning_rate": 2.2268787790595115e-05, + "loss": 0.3896, + "step": 45534 + }, + { + "epoch": 3.1597390881965164, + "grad_norm": 0.7448014616966248, + "learning_rate": 2.2265253157209122e-05, + "loss": 0.2485, + "step": 45535 + }, + { + "epoch": 3.159808479633613, + "grad_norm": 1.055384874343872, + "learning_rate": 2.226171876922267e-05, + "loss": 0.2669, + "step": 45536 + }, + { + "epoch": 3.15987787107071, + "grad_norm": 0.7980487942695618, + "learning_rate": 2.225818462664696e-05, + "loss": 0.3199, + "step": 45537 + }, + { + "epoch": 3.1599472625078064, + "grad_norm": 0.698401153087616, + "learning_rate": 2.225465072949313e-05, + "loss": 0.1762, + "step": 45538 + }, + { + "epoch": 3.1600166539449033, + "grad_norm": 1.0220791101455688, + "learning_rate": 2.2251117077772354e-05, + "loss": 0.3758, + "step": 45539 + }, + { + "epoch": 3.1600860453819997, + "grad_norm": 1.3027619123458862, + "learning_rate": 2.2247583671495788e-05, + "loss": 0.284, + "step": 45540 + }, + { + "epoch": 3.1601554368190965, + "grad_norm": 1.1170486211776733, + "learning_rate": 2.2244050510674574e-05, + "loss": 0.293, + "step": 45541 + }, + { + "epoch": 3.1602248282561933, + "grad_norm": 0.8233125805854797, + "learning_rate": 2.22405175953199e-05, + "loss": 0.2199, + "step": 45542 + }, + { + "epoch": 3.1602942196932897, + "grad_norm": 0.9824064373970032, + "learning_rate": 2.223698492544285e-05, + "loss": 0.3282, + "step": 45543 + }, + { + "epoch": 3.1603636111303866, + "grad_norm": 0.7053279280662537, + "learning_rate": 2.223345250105463e-05, + "loss": 0.1691, + "step": 45544 + }, + { + "epoch": 3.1604330025674834, + "grad_norm": 1.7684330940246582, + "learning_rate": 2.2229920322166376e-05, + "loss": 0.3883, + "step": 45545 + }, + { + "epoch": 3.1605023940045798, + "grad_norm": 0.8506196141242981, + "learning_rate": 2.2226388388789232e-05, + "loss": 0.2582, + "step": 45546 + }, + { + "epoch": 3.1605717854416766, + "grad_norm": 1.3031504154205322, + "learning_rate": 2.222285670093437e-05, + "loss": 0.2729, + "step": 45547 + }, + { + "epoch": 3.160641176878773, + "grad_norm": 0.5067316889762878, + "learning_rate": 2.2219325258612922e-05, + "loss": 0.19, + "step": 45548 + }, + { + "epoch": 3.16071056831587, + "grad_norm": 0.9309865832328796, + "learning_rate": 2.221579406183606e-05, + "loss": 0.2439, + "step": 45549 + }, + { + "epoch": 3.1607799597529667, + "grad_norm": 0.8946124315261841, + "learning_rate": 2.2212263110614883e-05, + "loss": 0.1967, + "step": 45550 + }, + { + "epoch": 3.160849351190063, + "grad_norm": 0.7956726551055908, + "learning_rate": 2.2208732404960574e-05, + "loss": 0.531, + "step": 45551 + }, + { + "epoch": 3.16091874262716, + "grad_norm": 0.6492950320243835, + "learning_rate": 2.220520194488427e-05, + "loss": 0.1609, + "step": 45552 + }, + { + "epoch": 3.1609881340642563, + "grad_norm": 1.3303353786468506, + "learning_rate": 2.2201671730397122e-05, + "loss": 0.2208, + "step": 45553 + }, + { + "epoch": 3.161057525501353, + "grad_norm": 0.7652201056480408, + "learning_rate": 2.219814176151026e-05, + "loss": 0.2076, + "step": 45554 + }, + { + "epoch": 3.16112691693845, + "grad_norm": 0.942408561706543, + "learning_rate": 2.2194612038234842e-05, + "loss": 0.4626, + "step": 45555 + }, + { + "epoch": 3.1611963083755463, + "grad_norm": 0.8748419880867004, + "learning_rate": 2.2191082560582033e-05, + "loss": 0.2822, + "step": 45556 + }, + { + "epoch": 3.161265699812643, + "grad_norm": 1.4631843566894531, + "learning_rate": 2.2187553328562927e-05, + "loss": 0.3516, + "step": 45557 + }, + { + "epoch": 3.1613350912497395, + "grad_norm": 1.0589988231658936, + "learning_rate": 2.2184024342188682e-05, + "loss": 0.3987, + "step": 45558 + }, + { + "epoch": 3.1614044826868364, + "grad_norm": 0.9724079370498657, + "learning_rate": 2.218049560147044e-05, + "loss": 0.2653, + "step": 45559 + }, + { + "epoch": 3.161473874123933, + "grad_norm": 0.9502545595169067, + "learning_rate": 2.2176967106419354e-05, + "loss": 0.2282, + "step": 45560 + }, + { + "epoch": 3.1615432655610296, + "grad_norm": 1.1597208976745605, + "learning_rate": 2.2173438857046548e-05, + "loss": 0.4104, + "step": 45561 + }, + { + "epoch": 3.1616126569981264, + "grad_norm": 0.6858108043670654, + "learning_rate": 2.2169910853363197e-05, + "loss": 0.1898, + "step": 45562 + }, + { + "epoch": 3.1616820484352233, + "grad_norm": 0.9932959675788879, + "learning_rate": 2.2166383095380384e-05, + "loss": 0.4197, + "step": 45563 + }, + { + "epoch": 3.1617514398723197, + "grad_norm": 0.8352394104003906, + "learning_rate": 2.2162855583109256e-05, + "loss": 0.2556, + "step": 45564 + }, + { + "epoch": 3.1618208313094165, + "grad_norm": 1.1960479021072388, + "learning_rate": 2.215932831656097e-05, + "loss": 0.4014, + "step": 45565 + }, + { + "epoch": 3.161890222746513, + "grad_norm": 0.9300647974014282, + "learning_rate": 2.215580129574668e-05, + "loss": 0.1986, + "step": 45566 + }, + { + "epoch": 3.1619596141836097, + "grad_norm": 1.0817948579788208, + "learning_rate": 2.215227452067746e-05, + "loss": 0.223, + "step": 45567 + }, + { + "epoch": 3.1620290056207065, + "grad_norm": 1.1320029497146606, + "learning_rate": 2.2148747991364494e-05, + "loss": 0.4291, + "step": 45568 + }, + { + "epoch": 3.162098397057803, + "grad_norm": 0.7331374287605286, + "learning_rate": 2.2145221707818885e-05, + "loss": 0.184, + "step": 45569 + }, + { + "epoch": 3.1621677884948998, + "grad_norm": 0.5942695736885071, + "learning_rate": 2.214169567005181e-05, + "loss": 0.1518, + "step": 45570 + }, + { + "epoch": 3.1622371799319966, + "grad_norm": 1.146657109260559, + "learning_rate": 2.2138169878074333e-05, + "loss": 0.3655, + "step": 45571 + }, + { + "epoch": 3.162306571369093, + "grad_norm": 0.9267463684082031, + "learning_rate": 2.2134644331897636e-05, + "loss": 0.2752, + "step": 45572 + }, + { + "epoch": 3.16237596280619, + "grad_norm": 1.3361231088638306, + "learning_rate": 2.2131119031532822e-05, + "loss": 0.315, + "step": 45573 + }, + { + "epoch": 3.162445354243286, + "grad_norm": 1.146653652191162, + "learning_rate": 2.2127593976991022e-05, + "loss": 0.2627, + "step": 45574 + }, + { + "epoch": 3.162514745680383, + "grad_norm": 0.8369279503822327, + "learning_rate": 2.212406916828339e-05, + "loss": 0.2278, + "step": 45575 + }, + { + "epoch": 3.16258413711748, + "grad_norm": 1.3313030004501343, + "learning_rate": 2.2120544605421034e-05, + "loss": 0.3358, + "step": 45576 + }, + { + "epoch": 3.1626535285545763, + "grad_norm": 0.6825438141822815, + "learning_rate": 2.21170202884151e-05, + "loss": 0.2519, + "step": 45577 + }, + { + "epoch": 3.162722919991673, + "grad_norm": 0.9215809106826782, + "learning_rate": 2.2113496217276673e-05, + "loss": 0.3629, + "step": 45578 + }, + { + "epoch": 3.1627923114287695, + "grad_norm": 1.1170189380645752, + "learning_rate": 2.2109972392016907e-05, + "loss": 0.421, + "step": 45579 + }, + { + "epoch": 3.1628617028658663, + "grad_norm": 0.7845498919487, + "learning_rate": 2.2106448812646917e-05, + "loss": 0.3144, + "step": 45580 + }, + { + "epoch": 3.162931094302963, + "grad_norm": 0.8055393099784851, + "learning_rate": 2.2102925479177837e-05, + "loss": 0.2091, + "step": 45581 + }, + { + "epoch": 3.1630004857400595, + "grad_norm": 0.6375275254249573, + "learning_rate": 2.2099402391620773e-05, + "loss": 0.2794, + "step": 45582 + }, + { + "epoch": 3.1630698771771564, + "grad_norm": 1.045498013496399, + "learning_rate": 2.2095879549986897e-05, + "loss": 0.2987, + "step": 45583 + }, + { + "epoch": 3.163139268614253, + "grad_norm": 0.8489391803741455, + "learning_rate": 2.2092356954287253e-05, + "loss": 0.1871, + "step": 45584 + }, + { + "epoch": 3.1632086600513496, + "grad_norm": 0.9529160857200623, + "learning_rate": 2.2088834604533006e-05, + "loss": 0.2143, + "step": 45585 + }, + { + "epoch": 3.1632780514884464, + "grad_norm": 1.0644034147262573, + "learning_rate": 2.2085312500735266e-05, + "loss": 0.2815, + "step": 45586 + }, + { + "epoch": 3.163347442925543, + "grad_norm": 0.7305189371109009, + "learning_rate": 2.2081790642905154e-05, + "loss": 0.1537, + "step": 45587 + }, + { + "epoch": 3.1634168343626397, + "grad_norm": 1.2531490325927734, + "learning_rate": 2.207826903105379e-05, + "loss": 0.2707, + "step": 45588 + }, + { + "epoch": 3.1634862257997365, + "grad_norm": 0.9459829330444336, + "learning_rate": 2.2074747665192286e-05, + "loss": 0.2981, + "step": 45589 + }, + { + "epoch": 3.163555617236833, + "grad_norm": 1.0760290622711182, + "learning_rate": 2.2071226545331793e-05, + "loss": 0.2876, + "step": 45590 + }, + { + "epoch": 3.1636250086739297, + "grad_norm": 0.6736446022987366, + "learning_rate": 2.2067705671483372e-05, + "loss": 0.2073, + "step": 45591 + }, + { + "epoch": 3.163694400111026, + "grad_norm": 0.7874877452850342, + "learning_rate": 2.206418504365816e-05, + "loss": 0.1992, + "step": 45592 + }, + { + "epoch": 3.163763791548123, + "grad_norm": 1.012933611869812, + "learning_rate": 2.206066466186727e-05, + "loss": 0.3383, + "step": 45593 + }, + { + "epoch": 3.1638331829852198, + "grad_norm": 0.973759114742279, + "learning_rate": 2.205714452612182e-05, + "loss": 0.3005, + "step": 45594 + }, + { + "epoch": 3.163902574422316, + "grad_norm": 0.7706507444381714, + "learning_rate": 2.2053624636432924e-05, + "loss": 0.1286, + "step": 45595 + }, + { + "epoch": 3.163971965859413, + "grad_norm": 0.9008523225784302, + "learning_rate": 2.2050104992811693e-05, + "loss": 0.193, + "step": 45596 + }, + { + "epoch": 3.16404135729651, + "grad_norm": 0.8061282634735107, + "learning_rate": 2.2046585595269265e-05, + "loss": 0.2152, + "step": 45597 + }, + { + "epoch": 3.164110748733606, + "grad_norm": 0.6902856826782227, + "learning_rate": 2.2043066443816695e-05, + "loss": 0.1632, + "step": 45598 + }, + { + "epoch": 3.164180140170703, + "grad_norm": 0.857245683670044, + "learning_rate": 2.203954753846511e-05, + "loss": 0.2611, + "step": 45599 + }, + { + "epoch": 3.1642495316077994, + "grad_norm": 0.6431915163993835, + "learning_rate": 2.2036028879225668e-05, + "loss": 0.167, + "step": 45600 + }, + { + "epoch": 3.1643189230448963, + "grad_norm": 0.9041566252708435, + "learning_rate": 2.20325104661094e-05, + "loss": 0.2419, + "step": 45601 + }, + { + "epoch": 3.164388314481993, + "grad_norm": 1.0400818586349487, + "learning_rate": 2.202899229912746e-05, + "loss": 0.2639, + "step": 45602 + }, + { + "epoch": 3.1644577059190895, + "grad_norm": 0.5752182006835938, + "learning_rate": 2.2025474378290945e-05, + "loss": 0.1159, + "step": 45603 + }, + { + "epoch": 3.1645270973561863, + "grad_norm": 0.7913774847984314, + "learning_rate": 2.202195670361098e-05, + "loss": 0.2131, + "step": 45604 + }, + { + "epoch": 3.1645964887932827, + "grad_norm": 0.6323282718658447, + "learning_rate": 2.2018439275098625e-05, + "loss": 0.1785, + "step": 45605 + }, + { + "epoch": 3.1646658802303795, + "grad_norm": 0.8564460277557373, + "learning_rate": 2.201492209276501e-05, + "loss": 0.2549, + "step": 45606 + }, + { + "epoch": 3.1647352716674764, + "grad_norm": 0.8688789010047913, + "learning_rate": 2.2011405156621235e-05, + "loss": 0.2042, + "step": 45607 + }, + { + "epoch": 3.1648046631045728, + "grad_norm": 0.9591459631919861, + "learning_rate": 2.2007888466678417e-05, + "loss": 0.1884, + "step": 45608 + }, + { + "epoch": 3.1648740545416696, + "grad_norm": 1.0275514125823975, + "learning_rate": 2.200437202294764e-05, + "loss": 0.2116, + "step": 45609 + }, + { + "epoch": 3.1649434459787664, + "grad_norm": 1.3442023992538452, + "learning_rate": 2.2000855825440002e-05, + "loss": 0.3044, + "step": 45610 + }, + { + "epoch": 3.165012837415863, + "grad_norm": 0.9696335792541504, + "learning_rate": 2.1997339874166655e-05, + "loss": 0.2758, + "step": 45611 + }, + { + "epoch": 3.1650822288529596, + "grad_norm": 0.82699054479599, + "learning_rate": 2.199382416913862e-05, + "loss": 0.3023, + "step": 45612 + }, + { + "epoch": 3.165151620290056, + "grad_norm": 1.2962449789047241, + "learning_rate": 2.1990308710367034e-05, + "loss": 0.3392, + "step": 45613 + }, + { + "epoch": 3.165221011727153, + "grad_norm": 0.8598020076751709, + "learning_rate": 2.198679349786299e-05, + "loss": 0.3159, + "step": 45614 + }, + { + "epoch": 3.1652904031642497, + "grad_norm": 0.7369604706764221, + "learning_rate": 2.1983278531637586e-05, + "loss": 0.2722, + "step": 45615 + }, + { + "epoch": 3.165359794601346, + "grad_norm": 1.482168197631836, + "learning_rate": 2.1979763811701927e-05, + "loss": 0.5048, + "step": 45616 + }, + { + "epoch": 3.165429186038443, + "grad_norm": 0.8960852026939392, + "learning_rate": 2.19762493380671e-05, + "loss": 0.2494, + "step": 45617 + }, + { + "epoch": 3.1654985774755398, + "grad_norm": 1.0460405349731445, + "learning_rate": 2.1972735110744225e-05, + "loss": 0.3014, + "step": 45618 + }, + { + "epoch": 3.165567968912636, + "grad_norm": 0.7955565452575684, + "learning_rate": 2.1969221129744353e-05, + "loss": 0.2424, + "step": 45619 + }, + { + "epoch": 3.165637360349733, + "grad_norm": 1.3758444786071777, + "learning_rate": 2.1965707395078584e-05, + "loss": 0.2949, + "step": 45620 + }, + { + "epoch": 3.1657067517868294, + "grad_norm": 0.9147423505783081, + "learning_rate": 2.1962193906758032e-05, + "loss": 0.2752, + "step": 45621 + }, + { + "epoch": 3.165776143223926, + "grad_norm": 1.0774593353271484, + "learning_rate": 2.1958680664793784e-05, + "loss": 0.3304, + "step": 45622 + }, + { + "epoch": 3.165845534661023, + "grad_norm": 1.1413419246673584, + "learning_rate": 2.195516766919692e-05, + "loss": 0.3401, + "step": 45623 + }, + { + "epoch": 3.1659149260981194, + "grad_norm": 0.8337132930755615, + "learning_rate": 2.1951654919978536e-05, + "loss": 0.1924, + "step": 45624 + }, + { + "epoch": 3.1659843175352163, + "grad_norm": 0.8379994630813599, + "learning_rate": 2.1948142417149753e-05, + "loss": 0.216, + "step": 45625 + }, + { + "epoch": 3.1660537089723126, + "grad_norm": 1.1276148557662964, + "learning_rate": 2.194463016072159e-05, + "loss": 0.3049, + "step": 45626 + }, + { + "epoch": 3.1661231004094095, + "grad_norm": 0.5142741203308105, + "learning_rate": 2.1941118150705188e-05, + "loss": 0.1439, + "step": 45627 + }, + { + "epoch": 3.1661924918465063, + "grad_norm": 1.2149399518966675, + "learning_rate": 2.193760638711161e-05, + "loss": 0.3258, + "step": 45628 + }, + { + "epoch": 3.1662618832836027, + "grad_norm": 0.6980187892913818, + "learning_rate": 2.193409486995196e-05, + "loss": 0.1631, + "step": 45629 + }, + { + "epoch": 3.1663312747206995, + "grad_norm": 0.9830855131149292, + "learning_rate": 2.1930583599237307e-05, + "loss": 0.295, + "step": 45630 + }, + { + "epoch": 3.166400666157796, + "grad_norm": 0.8539049625396729, + "learning_rate": 2.1927072574978757e-05, + "loss": 0.2203, + "step": 45631 + }, + { + "epoch": 3.1664700575948928, + "grad_norm": 1.1065454483032227, + "learning_rate": 2.1923561797187397e-05, + "loss": 0.2142, + "step": 45632 + }, + { + "epoch": 3.1665394490319896, + "grad_norm": 0.6659985184669495, + "learning_rate": 2.1920051265874263e-05, + "loss": 0.1847, + "step": 45633 + }, + { + "epoch": 3.166608840469086, + "grad_norm": 0.9714363813400269, + "learning_rate": 2.1916540981050482e-05, + "loss": 0.2789, + "step": 45634 + }, + { + "epoch": 3.166678231906183, + "grad_norm": 0.6230797171592712, + "learning_rate": 2.1913030942727132e-05, + "loss": 0.229, + "step": 45635 + }, + { + "epoch": 3.1667476233432796, + "grad_norm": 0.8093785643577576, + "learning_rate": 2.190952115091527e-05, + "loss": 0.2103, + "step": 45636 + }, + { + "epoch": 3.166817014780376, + "grad_norm": 2.2103192806243896, + "learning_rate": 2.1906011605625986e-05, + "loss": 0.3416, + "step": 45637 + }, + { + "epoch": 3.166886406217473, + "grad_norm": 1.0136041641235352, + "learning_rate": 2.1902502306870355e-05, + "loss": 0.2507, + "step": 45638 + }, + { + "epoch": 3.1669557976545692, + "grad_norm": 1.2089179754257202, + "learning_rate": 2.1898993254659506e-05, + "loss": 0.3883, + "step": 45639 + }, + { + "epoch": 3.167025189091666, + "grad_norm": 1.0415890216827393, + "learning_rate": 2.1895484449004434e-05, + "loss": 0.1628, + "step": 45640 + }, + { + "epoch": 3.167094580528763, + "grad_norm": 0.8298018574714661, + "learning_rate": 2.189197588991626e-05, + "loss": 0.1983, + "step": 45641 + }, + { + "epoch": 3.1671639719658593, + "grad_norm": 1.0932680368423462, + "learning_rate": 2.188846757740606e-05, + "loss": 0.3653, + "step": 45642 + }, + { + "epoch": 3.167233363402956, + "grad_norm": 0.9764496684074402, + "learning_rate": 2.1884959511484893e-05, + "loss": 0.2064, + "step": 45643 + }, + { + "epoch": 3.167302754840053, + "grad_norm": 0.741253137588501, + "learning_rate": 2.1881451692163855e-05, + "loss": 0.2211, + "step": 45644 + }, + { + "epoch": 3.1673721462771494, + "grad_norm": 0.8957230448722839, + "learning_rate": 2.1877944119454007e-05, + "loss": 0.265, + "step": 45645 + }, + { + "epoch": 3.167441537714246, + "grad_norm": 1.3706425428390503, + "learning_rate": 2.1874436793366458e-05, + "loss": 0.3173, + "step": 45646 + }, + { + "epoch": 3.1675109291513426, + "grad_norm": 1.6698583364486694, + "learning_rate": 2.1870929713912213e-05, + "loss": 0.3328, + "step": 45647 + }, + { + "epoch": 3.1675803205884394, + "grad_norm": 0.9264711141586304, + "learning_rate": 2.1867422881102385e-05, + "loss": 0.2955, + "step": 45648 + }, + { + "epoch": 3.1676497120255362, + "grad_norm": 0.9962772130966187, + "learning_rate": 2.186391629494804e-05, + "loss": 0.2682, + "step": 45649 + }, + { + "epoch": 3.1677191034626326, + "grad_norm": 1.1318742036819458, + "learning_rate": 2.186040995546024e-05, + "loss": 0.2666, + "step": 45650 + }, + { + "epoch": 3.1677884948997295, + "grad_norm": 0.836158037185669, + "learning_rate": 2.185690386265006e-05, + "loss": 0.2079, + "step": 45651 + }, + { + "epoch": 3.167857886336826, + "grad_norm": 0.76095050573349, + "learning_rate": 2.1853398016528572e-05, + "loss": 0.2121, + "step": 45652 + }, + { + "epoch": 3.1679272777739227, + "grad_norm": 1.0444056987762451, + "learning_rate": 2.1849892417106864e-05, + "loss": 0.3063, + "step": 45653 + }, + { + "epoch": 3.1679966692110195, + "grad_norm": 0.8513844609260559, + "learning_rate": 2.1846387064395947e-05, + "loss": 0.2884, + "step": 45654 + }, + { + "epoch": 3.168066060648116, + "grad_norm": 0.7671600580215454, + "learning_rate": 2.1842881958406923e-05, + "loss": 0.1862, + "step": 45655 + }, + { + "epoch": 3.1681354520852127, + "grad_norm": 1.2413779497146606, + "learning_rate": 2.1839377099150848e-05, + "loss": 0.3219, + "step": 45656 + }, + { + "epoch": 3.168204843522309, + "grad_norm": 0.6547529697418213, + "learning_rate": 2.1835872486638797e-05, + "loss": 0.1197, + "step": 45657 + }, + { + "epoch": 3.168274234959406, + "grad_norm": 0.8187633752822876, + "learning_rate": 2.183236812088183e-05, + "loss": 0.2443, + "step": 45658 + }, + { + "epoch": 3.168343626396503, + "grad_norm": 0.7594094276428223, + "learning_rate": 2.1828864001891037e-05, + "loss": 0.1982, + "step": 45659 + }, + { + "epoch": 3.168413017833599, + "grad_norm": 0.7317366003990173, + "learning_rate": 2.182536012967742e-05, + "loss": 0.2494, + "step": 45660 + }, + { + "epoch": 3.168482409270696, + "grad_norm": 1.1075716018676758, + "learning_rate": 2.182185650425207e-05, + "loss": 0.2504, + "step": 45661 + }, + { + "epoch": 3.168551800707793, + "grad_norm": 0.5846282243728638, + "learning_rate": 2.1818353125626044e-05, + "loss": 0.128, + "step": 45662 + }, + { + "epoch": 3.1686211921448892, + "grad_norm": 0.8369811177253723, + "learning_rate": 2.1814849993810405e-05, + "loss": 0.2497, + "step": 45663 + }, + { + "epoch": 3.168690583581986, + "grad_norm": 0.7627089619636536, + "learning_rate": 2.1811347108816226e-05, + "loss": 0.2219, + "step": 45664 + }, + { + "epoch": 3.1687599750190825, + "grad_norm": 1.2661199569702148, + "learning_rate": 2.1807844470654547e-05, + "loss": 0.3883, + "step": 45665 + }, + { + "epoch": 3.1688293664561793, + "grad_norm": 1.012750506401062, + "learning_rate": 2.180434207933646e-05, + "loss": 0.2935, + "step": 45666 + }, + { + "epoch": 3.168898757893276, + "grad_norm": 1.0596224069595337, + "learning_rate": 2.180083993487295e-05, + "loss": 0.3, + "step": 45667 + }, + { + "epoch": 3.1689681493303725, + "grad_norm": 0.8844931721687317, + "learning_rate": 2.1797338037275138e-05, + "loss": 0.3201, + "step": 45668 + }, + { + "epoch": 3.1690375407674694, + "grad_norm": 0.7245144248008728, + "learning_rate": 2.1793836386554046e-05, + "loss": 0.2018, + "step": 45669 + }, + { + "epoch": 3.169106932204566, + "grad_norm": 1.061943769454956, + "learning_rate": 2.179033498272076e-05, + "loss": 0.427, + "step": 45670 + }, + { + "epoch": 3.1691763236416626, + "grad_norm": 1.1789401769638062, + "learning_rate": 2.17868338257863e-05, + "loss": 0.4643, + "step": 45671 + }, + { + "epoch": 3.1692457150787594, + "grad_norm": 0.9362871646881104, + "learning_rate": 2.1783332915761722e-05, + "loss": 0.3182, + "step": 45672 + }, + { + "epoch": 3.169315106515856, + "grad_norm": 1.7049671411514282, + "learning_rate": 2.1779832252658082e-05, + "loss": 0.1978, + "step": 45673 + }, + { + "epoch": 3.1693844979529526, + "grad_norm": 1.0656658411026, + "learning_rate": 2.1776331836486474e-05, + "loss": 0.3876, + "step": 45674 + }, + { + "epoch": 3.1694538893900495, + "grad_norm": 0.7017503380775452, + "learning_rate": 2.1772831667257876e-05, + "loss": 0.2143, + "step": 45675 + }, + { + "epoch": 3.169523280827146, + "grad_norm": 0.7301368713378906, + "learning_rate": 2.176933174498338e-05, + "loss": 0.2409, + "step": 45676 + }, + { + "epoch": 3.1695926722642427, + "grad_norm": 0.8633803129196167, + "learning_rate": 2.1765832069674018e-05, + "loss": 0.2483, + "step": 45677 + }, + { + "epoch": 3.169662063701339, + "grad_norm": 1.0764702558517456, + "learning_rate": 2.1762332641340853e-05, + "loss": 0.249, + "step": 45678 + }, + { + "epoch": 3.169731455138436, + "grad_norm": 0.8921145796775818, + "learning_rate": 2.1758833459994933e-05, + "loss": 0.3665, + "step": 45679 + }, + { + "epoch": 3.1698008465755327, + "grad_norm": 0.852810263633728, + "learning_rate": 2.175533452564732e-05, + "loss": 0.3359, + "step": 45680 + }, + { + "epoch": 3.169870238012629, + "grad_norm": 0.8021485805511475, + "learning_rate": 2.175183583830901e-05, + "loss": 0.2175, + "step": 45681 + }, + { + "epoch": 3.169939629449726, + "grad_norm": 0.7347501516342163, + "learning_rate": 2.174833739799107e-05, + "loss": 0.3133, + "step": 45682 + }, + { + "epoch": 3.1700090208868223, + "grad_norm": 0.7480567097663879, + "learning_rate": 2.174483920470456e-05, + "loss": 0.2607, + "step": 45683 + }, + { + "epoch": 3.170078412323919, + "grad_norm": 0.9380248188972473, + "learning_rate": 2.1741341258460503e-05, + "loss": 0.3244, + "step": 45684 + }, + { + "epoch": 3.170147803761016, + "grad_norm": 0.9376634359359741, + "learning_rate": 2.1737843559269954e-05, + "loss": 0.2381, + "step": 45685 + }, + { + "epoch": 3.1702171951981124, + "grad_norm": 0.7171290516853333, + "learning_rate": 2.1734346107143954e-05, + "loss": 0.199, + "step": 45686 + }, + { + "epoch": 3.1702865866352092, + "grad_norm": 0.7594048380851746, + "learning_rate": 2.1730848902093582e-05, + "loss": 0.2653, + "step": 45687 + }, + { + "epoch": 3.170355978072306, + "grad_norm": 1.1134583950042725, + "learning_rate": 2.1727351944129805e-05, + "loss": 0.3273, + "step": 45688 + }, + { + "epoch": 3.1704253695094025, + "grad_norm": 0.6540691256523132, + "learning_rate": 2.172385523326369e-05, + "loss": 0.1568, + "step": 45689 + }, + { + "epoch": 3.1704947609464993, + "grad_norm": 0.8319800496101379, + "learning_rate": 2.1720358769506276e-05, + "loss": 0.253, + "step": 45690 + }, + { + "epoch": 3.1705641523835957, + "grad_norm": 1.0325897932052612, + "learning_rate": 2.171686255286862e-05, + "loss": 0.2645, + "step": 45691 + }, + { + "epoch": 3.1706335438206925, + "grad_norm": 0.5615170001983643, + "learning_rate": 2.171336658336175e-05, + "loss": 0.2328, + "step": 45692 + }, + { + "epoch": 3.1707029352577893, + "grad_norm": 1.2026910781860352, + "learning_rate": 2.1709870860996683e-05, + "loss": 0.298, + "step": 45693 + }, + { + "epoch": 3.1707723266948857, + "grad_norm": 1.1176139116287231, + "learning_rate": 2.1706375385784518e-05, + "loss": 0.2727, + "step": 45694 + }, + { + "epoch": 3.1708417181319826, + "grad_norm": 0.8540140390396118, + "learning_rate": 2.1702880157736194e-05, + "loss": 0.3078, + "step": 45695 + }, + { + "epoch": 3.1709111095690794, + "grad_norm": 0.8444560766220093, + "learning_rate": 2.1699385176862807e-05, + "loss": 0.2209, + "step": 45696 + }, + { + "epoch": 3.170980501006176, + "grad_norm": 1.362261414527893, + "learning_rate": 2.169589044317537e-05, + "loss": 0.4343, + "step": 45697 + }, + { + "epoch": 3.1710498924432726, + "grad_norm": 0.7579066157341003, + "learning_rate": 2.169239595668493e-05, + "loss": 0.2104, + "step": 45698 + }, + { + "epoch": 3.171119283880369, + "grad_norm": 0.967420220375061, + "learning_rate": 2.1688901717402498e-05, + "loss": 0.3861, + "step": 45699 + }, + { + "epoch": 3.171188675317466, + "grad_norm": 1.119874358177185, + "learning_rate": 2.1685407725339124e-05, + "loss": 0.3079, + "step": 45700 + }, + { + "epoch": 3.1712580667545627, + "grad_norm": 0.848927915096283, + "learning_rate": 2.1681913980505864e-05, + "loss": 0.1789, + "step": 45701 + }, + { + "epoch": 3.171327458191659, + "grad_norm": 1.354066014289856, + "learning_rate": 2.1678420482913687e-05, + "loss": 0.2361, + "step": 45702 + }, + { + "epoch": 3.171396849628756, + "grad_norm": 0.8031628727912903, + "learning_rate": 2.167492723257364e-05, + "loss": 0.2481, + "step": 45703 + }, + { + "epoch": 3.1714662410658523, + "grad_norm": 0.9612593650817871, + "learning_rate": 2.1671434229496767e-05, + "loss": 0.2151, + "step": 45704 + }, + { + "epoch": 3.171535632502949, + "grad_norm": 0.8139821887016296, + "learning_rate": 2.166794147369412e-05, + "loss": 0.2804, + "step": 45705 + }, + { + "epoch": 3.171605023940046, + "grad_norm": 0.7032424807548523, + "learning_rate": 2.1664448965176652e-05, + "loss": 0.171, + "step": 45706 + }, + { + "epoch": 3.1716744153771423, + "grad_norm": 0.7996153235435486, + "learning_rate": 2.166095670395545e-05, + "loss": 0.2147, + "step": 45707 + }, + { + "epoch": 3.171743806814239, + "grad_norm": 1.0219554901123047, + "learning_rate": 2.1657464690041495e-05, + "loss": 0.2381, + "step": 45708 + }, + { + "epoch": 3.1718131982513356, + "grad_norm": 0.6493773460388184, + "learning_rate": 2.1653972923445876e-05, + "loss": 0.1801, + "step": 45709 + }, + { + "epoch": 3.1718825896884324, + "grad_norm": 1.116058111190796, + "learning_rate": 2.1650481404179545e-05, + "loss": 0.386, + "step": 45710 + }, + { + "epoch": 3.1719519811255292, + "grad_norm": 0.7319939732551575, + "learning_rate": 2.1646990132253563e-05, + "loss": 0.1923, + "step": 45711 + }, + { + "epoch": 3.1720213725626256, + "grad_norm": 1.232071042060852, + "learning_rate": 2.1643499107678932e-05, + "loss": 0.2148, + "step": 45712 + }, + { + "epoch": 3.1720907639997225, + "grad_norm": 1.065208911895752, + "learning_rate": 2.1640008330466687e-05, + "loss": 0.2775, + "step": 45713 + }, + { + "epoch": 3.1721601554368193, + "grad_norm": 1.1132208108901978, + "learning_rate": 2.1636517800627842e-05, + "loss": 0.4553, + "step": 45714 + }, + { + "epoch": 3.1722295468739157, + "grad_norm": 0.8770056962966919, + "learning_rate": 2.1633027518173444e-05, + "loss": 0.174, + "step": 45715 + }, + { + "epoch": 3.1722989383110125, + "grad_norm": 0.7890492677688599, + "learning_rate": 2.162953748311446e-05, + "loss": 0.2229, + "step": 45716 + }, + { + "epoch": 3.172368329748109, + "grad_norm": 0.9878706336021423, + "learning_rate": 2.162604769546194e-05, + "loss": 0.2492, + "step": 45717 + }, + { + "epoch": 3.1724377211852057, + "grad_norm": 0.9792899489402771, + "learning_rate": 2.1622558155226892e-05, + "loss": 0.3168, + "step": 45718 + }, + { + "epoch": 3.1725071126223026, + "grad_norm": 0.9916718602180481, + "learning_rate": 2.161906886242032e-05, + "loss": 0.3213, + "step": 45719 + }, + { + "epoch": 3.172576504059399, + "grad_norm": 0.8896130919456482, + "learning_rate": 2.1615579817053266e-05, + "loss": 0.1888, + "step": 45720 + }, + { + "epoch": 3.172645895496496, + "grad_norm": 0.6787224411964417, + "learning_rate": 2.161209101913674e-05, + "loss": 0.163, + "step": 45721 + }, + { + "epoch": 3.1727152869335926, + "grad_norm": 0.971139132976532, + "learning_rate": 2.1608602468681762e-05, + "loss": 0.3532, + "step": 45722 + }, + { + "epoch": 3.172784678370689, + "grad_norm": 1.0915873050689697, + "learning_rate": 2.1605114165699315e-05, + "loss": 0.3076, + "step": 45723 + }, + { + "epoch": 3.172854069807786, + "grad_norm": 0.7838045954704285, + "learning_rate": 2.1601626110200414e-05, + "loss": 0.388, + "step": 45724 + }, + { + "epoch": 3.1729234612448822, + "grad_norm": 0.9018279910087585, + "learning_rate": 2.159813830219609e-05, + "loss": 0.2271, + "step": 45725 + }, + { + "epoch": 3.172992852681979, + "grad_norm": 2.84511399269104, + "learning_rate": 2.159465074169734e-05, + "loss": 0.3012, + "step": 45726 + }, + { + "epoch": 3.173062244119076, + "grad_norm": 0.8012571334838867, + "learning_rate": 2.15911634287152e-05, + "loss": 0.2613, + "step": 45727 + }, + { + "epoch": 3.1731316355561723, + "grad_norm": 1.158997654914856, + "learning_rate": 2.1587676363260645e-05, + "loss": 0.2744, + "step": 45728 + }, + { + "epoch": 3.173201026993269, + "grad_norm": 0.718251645565033, + "learning_rate": 2.1584189545344723e-05, + "loss": 0.2326, + "step": 45729 + }, + { + "epoch": 3.1732704184303655, + "grad_norm": 0.859818160533905, + "learning_rate": 2.1580702974978396e-05, + "loss": 0.1969, + "step": 45730 + }, + { + "epoch": 3.1733398098674623, + "grad_norm": 0.855695366859436, + "learning_rate": 2.1577216652172694e-05, + "loss": 0.2303, + "step": 45731 + }, + { + "epoch": 3.173409201304559, + "grad_norm": 1.0773112773895264, + "learning_rate": 2.157373057693861e-05, + "loss": 0.3544, + "step": 45732 + }, + { + "epoch": 3.1734785927416556, + "grad_norm": 0.9961163401603699, + "learning_rate": 2.157024474928717e-05, + "loss": 0.2426, + "step": 45733 + }, + { + "epoch": 3.1735479841787524, + "grad_norm": 0.5015149116516113, + "learning_rate": 2.1566759169229366e-05, + "loss": 0.1206, + "step": 45734 + }, + { + "epoch": 3.173617375615849, + "grad_norm": 0.6885863542556763, + "learning_rate": 2.1563273836776222e-05, + "loss": 0.3129, + "step": 45735 + }, + { + "epoch": 3.1736867670529456, + "grad_norm": 0.9253979325294495, + "learning_rate": 2.1559788751938703e-05, + "loss": 0.2911, + "step": 45736 + }, + { + "epoch": 3.1737561584900424, + "grad_norm": 0.8947243094444275, + "learning_rate": 2.1556303914727828e-05, + "loss": 0.2164, + "step": 45737 + }, + { + "epoch": 3.173825549927139, + "grad_norm": 0.8057863712310791, + "learning_rate": 2.1552819325154607e-05, + "loss": 0.2836, + "step": 45738 + }, + { + "epoch": 3.1738949413642357, + "grad_norm": 1.279064416885376, + "learning_rate": 2.1549334983230018e-05, + "loss": 0.2063, + "step": 45739 + }, + { + "epoch": 3.1739643328013325, + "grad_norm": 1.04537034034729, + "learning_rate": 2.1545850888965113e-05, + "loss": 0.2889, + "step": 45740 + }, + { + "epoch": 3.174033724238429, + "grad_norm": 0.7899054288864136, + "learning_rate": 2.1542367042370836e-05, + "loss": 0.2642, + "step": 45741 + }, + { + "epoch": 3.1741031156755257, + "grad_norm": 0.9482054710388184, + "learning_rate": 2.1538883443458192e-05, + "loss": 0.2932, + "step": 45742 + }, + { + "epoch": 3.174172507112622, + "grad_norm": 0.7761634588241577, + "learning_rate": 2.1535400092238223e-05, + "loss": 0.2119, + "step": 45743 + }, + { + "epoch": 3.174241898549719, + "grad_norm": 0.8049038052558899, + "learning_rate": 2.1531916988721857e-05, + "loss": 0.2928, + "step": 45744 + }, + { + "epoch": 3.1743112899868158, + "grad_norm": 1.2568720579147339, + "learning_rate": 2.152843413292013e-05, + "loss": 0.4577, + "step": 45745 + }, + { + "epoch": 3.174380681423912, + "grad_norm": 0.7621713876724243, + "learning_rate": 2.152495152484404e-05, + "loss": 0.2396, + "step": 45746 + }, + { + "epoch": 3.174450072861009, + "grad_norm": 0.7477173805236816, + "learning_rate": 2.1521469164504558e-05, + "loss": 0.1668, + "step": 45747 + }, + { + "epoch": 3.174519464298106, + "grad_norm": 0.7743068933486938, + "learning_rate": 2.1517987051912702e-05, + "loss": 0.2357, + "step": 45748 + }, + { + "epoch": 3.174588855735202, + "grad_norm": 1.0453754663467407, + "learning_rate": 2.151450518707948e-05, + "loss": 0.3215, + "step": 45749 + }, + { + "epoch": 3.174658247172299, + "grad_norm": 1.130374550819397, + "learning_rate": 2.151102357001583e-05, + "loss": 0.2431, + "step": 45750 + }, + { + "epoch": 3.1747276386093954, + "grad_norm": 0.925315797328949, + "learning_rate": 2.1507542200732777e-05, + "loss": 0.1895, + "step": 45751 + }, + { + "epoch": 3.1747970300464923, + "grad_norm": 0.847381591796875, + "learning_rate": 2.15040610792413e-05, + "loss": 0.2561, + "step": 45752 + }, + { + "epoch": 3.174866421483589, + "grad_norm": 0.6770356893539429, + "learning_rate": 2.15005802055524e-05, + "loss": 0.1811, + "step": 45753 + }, + { + "epoch": 3.1749358129206855, + "grad_norm": 0.7213724851608276, + "learning_rate": 2.149709957967706e-05, + "loss": 0.125, + "step": 45754 + }, + { + "epoch": 3.1750052043577823, + "grad_norm": 0.783941924571991, + "learning_rate": 2.1493619201626268e-05, + "loss": 0.215, + "step": 45755 + }, + { + "epoch": 3.1750745957948787, + "grad_norm": 0.9628376960754395, + "learning_rate": 2.1490139071411043e-05, + "loss": 0.2021, + "step": 45756 + }, + { + "epoch": 3.1751439872319756, + "grad_norm": 1.223008632659912, + "learning_rate": 2.1486659189042314e-05, + "loss": 0.2969, + "step": 45757 + }, + { + "epoch": 3.1752133786690724, + "grad_norm": 2.782029390335083, + "learning_rate": 2.1483179554531085e-05, + "loss": 0.2108, + "step": 45758 + }, + { + "epoch": 3.1752827701061688, + "grad_norm": 0.5787655115127563, + "learning_rate": 2.147970016788835e-05, + "loss": 0.1503, + "step": 45759 + }, + { + "epoch": 3.1753521615432656, + "grad_norm": 0.8345435261726379, + "learning_rate": 2.147622102912509e-05, + "loss": 0.2949, + "step": 45760 + }, + { + "epoch": 3.175421552980362, + "grad_norm": 1.4131429195404053, + "learning_rate": 2.14727421382523e-05, + "loss": 0.3652, + "step": 45761 + }, + { + "epoch": 3.175490944417459, + "grad_norm": 0.9480676651000977, + "learning_rate": 2.1469263495280944e-05, + "loss": 0.3036, + "step": 45762 + }, + { + "epoch": 3.1755603358545557, + "grad_norm": 0.8491470813751221, + "learning_rate": 2.1465785100222035e-05, + "loss": 0.2822, + "step": 45763 + }, + { + "epoch": 3.175629727291652, + "grad_norm": 0.8784075379371643, + "learning_rate": 2.146230695308651e-05, + "loss": 0.1404, + "step": 45764 + }, + { + "epoch": 3.175699118728749, + "grad_norm": 0.8866063356399536, + "learning_rate": 2.1458829053885364e-05, + "loss": 0.1664, + "step": 45765 + }, + { + "epoch": 3.1757685101658457, + "grad_norm": 0.8201375007629395, + "learning_rate": 2.1455351402629587e-05, + "loss": 0.3013, + "step": 45766 + }, + { + "epoch": 3.175837901602942, + "grad_norm": 1.0583893060684204, + "learning_rate": 2.1451873999330152e-05, + "loss": 0.3621, + "step": 45767 + }, + { + "epoch": 3.175907293040039, + "grad_norm": 1.0098185539245605, + "learning_rate": 2.1448396843998044e-05, + "loss": 0.2442, + "step": 45768 + }, + { + "epoch": 3.1759766844771353, + "grad_norm": 0.5243464708328247, + "learning_rate": 2.1444919936644226e-05, + "loss": 0.1874, + "step": 45769 + }, + { + "epoch": 3.176046075914232, + "grad_norm": 0.5033993124961853, + "learning_rate": 2.1441443277279703e-05, + "loss": 0.1101, + "step": 45770 + }, + { + "epoch": 3.176115467351329, + "grad_norm": 0.6800079345703125, + "learning_rate": 2.1437966865915416e-05, + "loss": 0.1708, + "step": 45771 + }, + { + "epoch": 3.1761848587884254, + "grad_norm": 0.9887560606002808, + "learning_rate": 2.143449070256235e-05, + "loss": 0.2357, + "step": 45772 + }, + { + "epoch": 3.176254250225522, + "grad_norm": 1.8356236219406128, + "learning_rate": 2.1431014787231473e-05, + "loss": 0.4428, + "step": 45773 + }, + { + "epoch": 3.176323641662619, + "grad_norm": 0.8389527797698975, + "learning_rate": 2.142753911993377e-05, + "loss": 0.3728, + "step": 45774 + }, + { + "epoch": 3.1763930330997154, + "grad_norm": 1.029467225074768, + "learning_rate": 2.1424063700680242e-05, + "loss": 0.2534, + "step": 45775 + }, + { + "epoch": 3.1764624245368123, + "grad_norm": 1.086830496788025, + "learning_rate": 2.14205885294818e-05, + "loss": 0.2027, + "step": 45776 + }, + { + "epoch": 3.1765318159739087, + "grad_norm": 0.6939398646354675, + "learning_rate": 2.1417113606349437e-05, + "loss": 0.1913, + "step": 45777 + }, + { + "epoch": 3.1766012074110055, + "grad_norm": 1.0263950824737549, + "learning_rate": 2.1413638931294157e-05, + "loss": 0.2258, + "step": 45778 + }, + { + "epoch": 3.1766705988481023, + "grad_norm": 0.7963762879371643, + "learning_rate": 2.1410164504326878e-05, + "loss": 0.2554, + "step": 45779 + }, + { + "epoch": 3.1767399902851987, + "grad_norm": 0.9610347151756287, + "learning_rate": 2.140669032545859e-05, + "loss": 0.431, + "step": 45780 + }, + { + "epoch": 3.1768093817222955, + "grad_norm": 0.8956016898155212, + "learning_rate": 2.1403216394700266e-05, + "loss": 0.2508, + "step": 45781 + }, + { + "epoch": 3.176878773159392, + "grad_norm": 1.319213628768921, + "learning_rate": 2.1399742712062864e-05, + "loss": 0.2998, + "step": 45782 + }, + { + "epoch": 3.1769481645964888, + "grad_norm": 0.8230816125869751, + "learning_rate": 2.1396269277557358e-05, + "loss": 0.1817, + "step": 45783 + }, + { + "epoch": 3.1770175560335856, + "grad_norm": 0.9755006432533264, + "learning_rate": 2.1392796091194734e-05, + "loss": 0.2583, + "step": 45784 + }, + { + "epoch": 3.177086947470682, + "grad_norm": 0.9192429780960083, + "learning_rate": 2.138932315298591e-05, + "loss": 0.1907, + "step": 45785 + }, + { + "epoch": 3.177156338907779, + "grad_norm": 1.2457618713378906, + "learning_rate": 2.138585046294187e-05, + "loss": 0.2947, + "step": 45786 + }, + { + "epoch": 3.177225730344875, + "grad_norm": 0.9161921143531799, + "learning_rate": 2.138237802107358e-05, + "loss": 0.1839, + "step": 45787 + }, + { + "epoch": 3.177295121781972, + "grad_norm": 1.1446475982666016, + "learning_rate": 2.137890582739199e-05, + "loss": 0.2737, + "step": 45788 + }, + { + "epoch": 3.177364513219069, + "grad_norm": 0.7346704602241516, + "learning_rate": 2.137543388190809e-05, + "loss": 0.2172, + "step": 45789 + }, + { + "epoch": 3.1774339046561653, + "grad_norm": 0.8417909145355225, + "learning_rate": 2.137196218463281e-05, + "loss": 0.2583, + "step": 45790 + }, + { + "epoch": 3.177503296093262, + "grad_norm": 1.669861078262329, + "learning_rate": 2.1368490735577164e-05, + "loss": 0.5054, + "step": 45791 + }, + { + "epoch": 3.177572687530359, + "grad_norm": 0.6230789422988892, + "learning_rate": 2.136501953475203e-05, + "loss": 0.2, + "step": 45792 + }, + { + "epoch": 3.1776420789674553, + "grad_norm": 0.6889912486076355, + "learning_rate": 2.136154858216841e-05, + "loss": 0.1826, + "step": 45793 + }, + { + "epoch": 3.177711470404552, + "grad_norm": 0.608065128326416, + "learning_rate": 2.1358077877837257e-05, + "loss": 0.1688, + "step": 45794 + }, + { + "epoch": 3.1777808618416485, + "grad_norm": 0.7079269886016846, + "learning_rate": 2.1354607421769522e-05, + "loss": 0.1912, + "step": 45795 + }, + { + "epoch": 3.1778502532787454, + "grad_norm": 0.7486978769302368, + "learning_rate": 2.135113721397618e-05, + "loss": 0.2335, + "step": 45796 + }, + { + "epoch": 3.177919644715842, + "grad_norm": 0.6369544267654419, + "learning_rate": 2.134766725446816e-05, + "loss": 0.1951, + "step": 45797 + }, + { + "epoch": 3.1779890361529386, + "grad_norm": 0.9961163997650146, + "learning_rate": 2.1344197543256473e-05, + "loss": 0.2962, + "step": 45798 + }, + { + "epoch": 3.1780584275900354, + "grad_norm": 0.9488645792007446, + "learning_rate": 2.134072808035198e-05, + "loss": 0.2105, + "step": 45799 + }, + { + "epoch": 3.1781278190271323, + "grad_norm": 0.9347118139266968, + "learning_rate": 2.1337258865765696e-05, + "loss": 0.2989, + "step": 45800 + }, + { + "epoch": 3.1781972104642287, + "grad_norm": 0.9122614860534668, + "learning_rate": 2.133378989950856e-05, + "loss": 0.1691, + "step": 45801 + }, + { + "epoch": 3.1782666019013255, + "grad_norm": 0.8286989331245422, + "learning_rate": 2.1330321181591516e-05, + "loss": 0.2138, + "step": 45802 + }, + { + "epoch": 3.178335993338422, + "grad_norm": 1.035677433013916, + "learning_rate": 2.1326852712025524e-05, + "loss": 0.4296, + "step": 45803 + }, + { + "epoch": 3.1784053847755187, + "grad_norm": 0.8198595643043518, + "learning_rate": 2.1323384490821528e-05, + "loss": 0.2609, + "step": 45804 + }, + { + "epoch": 3.1784747762126155, + "grad_norm": 0.8134605288505554, + "learning_rate": 2.1319916517990513e-05, + "loss": 0.2226, + "step": 45805 + }, + { + "epoch": 3.178544167649712, + "grad_norm": 0.8412140011787415, + "learning_rate": 2.1316448793543355e-05, + "loss": 0.1621, + "step": 45806 + }, + { + "epoch": 3.1786135590868088, + "grad_norm": 1.0029739141464233, + "learning_rate": 2.1312981317491044e-05, + "loss": 0.2859, + "step": 45807 + }, + { + "epoch": 3.178682950523905, + "grad_norm": 0.7274284958839417, + "learning_rate": 2.130951408984453e-05, + "loss": 0.2461, + "step": 45808 + }, + { + "epoch": 3.178752341961002, + "grad_norm": 1.1196153163909912, + "learning_rate": 2.1306047110614735e-05, + "loss": 0.3064, + "step": 45809 + }, + { + "epoch": 3.178821733398099, + "grad_norm": 1.052839994430542, + "learning_rate": 2.1302580379812654e-05, + "loss": 0.175, + "step": 45810 + }, + { + "epoch": 3.178891124835195, + "grad_norm": 0.9114190340042114, + "learning_rate": 2.1299113897449175e-05, + "loss": 0.2382, + "step": 45811 + }, + { + "epoch": 3.178960516272292, + "grad_norm": 0.6624868512153625, + "learning_rate": 2.1295647663535255e-05, + "loss": 0.1668, + "step": 45812 + }, + { + "epoch": 3.179029907709389, + "grad_norm": 1.0422803163528442, + "learning_rate": 2.129218167808187e-05, + "loss": 0.3664, + "step": 45813 + }, + { + "epoch": 3.1790992991464853, + "grad_norm": 0.914269208908081, + "learning_rate": 2.1288715941099913e-05, + "loss": 0.2407, + "step": 45814 + }, + { + "epoch": 3.179168690583582, + "grad_norm": 0.7956032156944275, + "learning_rate": 2.1285250452600347e-05, + "loss": 0.3042, + "step": 45815 + }, + { + "epoch": 3.1792380820206785, + "grad_norm": 0.8877039551734924, + "learning_rate": 2.128178521259411e-05, + "loss": 0.2244, + "step": 45816 + }, + { + "epoch": 3.1793074734577753, + "grad_norm": 0.9674456715583801, + "learning_rate": 2.127832022109214e-05, + "loss": 0.3093, + "step": 45817 + }, + { + "epoch": 3.179376864894872, + "grad_norm": 0.7175275683403015, + "learning_rate": 2.1274855478105415e-05, + "loss": 0.1808, + "step": 45818 + }, + { + "epoch": 3.1794462563319685, + "grad_norm": 0.9185308218002319, + "learning_rate": 2.12713909836448e-05, + "loss": 0.2927, + "step": 45819 + }, + { + "epoch": 3.1795156477690654, + "grad_norm": 0.8502925038337708, + "learning_rate": 2.126792673772129e-05, + "loss": 0.1964, + "step": 45820 + }, + { + "epoch": 3.1795850392061618, + "grad_norm": 1.2204076051712036, + "learning_rate": 2.1264462740345782e-05, + "loss": 0.276, + "step": 45821 + }, + { + "epoch": 3.1796544306432586, + "grad_norm": 1.8518190383911133, + "learning_rate": 2.1260998991529236e-05, + "loss": 0.4109, + "step": 45822 + }, + { + "epoch": 3.1797238220803554, + "grad_norm": 1.4755982160568237, + "learning_rate": 2.1257535491282576e-05, + "loss": 0.2591, + "step": 45823 + }, + { + "epoch": 3.179793213517452, + "grad_norm": 0.8296052813529968, + "learning_rate": 2.125407223961675e-05, + "loss": 0.2847, + "step": 45824 + }, + { + "epoch": 3.1798626049545486, + "grad_norm": 0.995800256729126, + "learning_rate": 2.1250609236542696e-05, + "loss": 0.2202, + "step": 45825 + }, + { + "epoch": 3.1799319963916455, + "grad_norm": 0.8551025390625, + "learning_rate": 2.1247146482071324e-05, + "loss": 0.2736, + "step": 45826 + }, + { + "epoch": 3.180001387828742, + "grad_norm": 1.1613444089889526, + "learning_rate": 2.1243683976213557e-05, + "loss": 0.2918, + "step": 45827 + }, + { + "epoch": 3.1800707792658387, + "grad_norm": 0.8041744828224182, + "learning_rate": 2.1240221718980347e-05, + "loss": 0.243, + "step": 45828 + }, + { + "epoch": 3.180140170702935, + "grad_norm": 1.3136733770370483, + "learning_rate": 2.1236759710382614e-05, + "loss": 0.2473, + "step": 45829 + }, + { + "epoch": 3.180209562140032, + "grad_norm": 0.7414345145225525, + "learning_rate": 2.1233297950431307e-05, + "loss": 0.2491, + "step": 45830 + }, + { + "epoch": 3.1802789535771288, + "grad_norm": 1.214050531387329, + "learning_rate": 2.122983643913733e-05, + "loss": 0.3433, + "step": 45831 + }, + { + "epoch": 3.180348345014225, + "grad_norm": 0.889729917049408, + "learning_rate": 2.122637517651165e-05, + "loss": 0.3216, + "step": 45832 + }, + { + "epoch": 3.180417736451322, + "grad_norm": 0.8342314958572388, + "learning_rate": 2.1222914162565133e-05, + "loss": 0.2403, + "step": 45833 + }, + { + "epoch": 3.1804871278884184, + "grad_norm": 0.7756126523017883, + "learning_rate": 2.121945339730873e-05, + "loss": 0.2481, + "step": 45834 + }, + { + "epoch": 3.180556519325515, + "grad_norm": 0.8119613528251648, + "learning_rate": 2.121599288075339e-05, + "loss": 0.3547, + "step": 45835 + }, + { + "epoch": 3.180625910762612, + "grad_norm": 0.9768105745315552, + "learning_rate": 2.1212532612910006e-05, + "loss": 0.2834, + "step": 45836 + }, + { + "epoch": 3.1806953021997084, + "grad_norm": 1.0746705532073975, + "learning_rate": 2.1209072593789524e-05, + "loss": 0.2379, + "step": 45837 + }, + { + "epoch": 3.1807646936368053, + "grad_norm": 0.8494361639022827, + "learning_rate": 2.1205612823402854e-05, + "loss": 0.3096, + "step": 45838 + }, + { + "epoch": 3.180834085073902, + "grad_norm": 0.7778304219245911, + "learning_rate": 2.1202153301760952e-05, + "loss": 0.2127, + "step": 45839 + }, + { + "epoch": 3.1809034765109985, + "grad_norm": 0.9765451550483704, + "learning_rate": 2.1198694028874677e-05, + "loss": 0.1498, + "step": 45840 + }, + { + "epoch": 3.1809728679480953, + "grad_norm": 1.3261140584945679, + "learning_rate": 2.1195235004754988e-05, + "loss": 0.3203, + "step": 45841 + }, + { + "epoch": 3.1810422593851917, + "grad_norm": 0.6458507776260376, + "learning_rate": 2.1191776229412796e-05, + "loss": 0.2245, + "step": 45842 + }, + { + "epoch": 3.1811116508222885, + "grad_norm": 1.3688808679580688, + "learning_rate": 2.1188317702859017e-05, + "loss": 0.3695, + "step": 45843 + }, + { + "epoch": 3.1811810422593854, + "grad_norm": 1.0852680206298828, + "learning_rate": 2.118485942510462e-05, + "loss": 0.3783, + "step": 45844 + }, + { + "epoch": 3.1812504336964817, + "grad_norm": 0.9819271564483643, + "learning_rate": 2.118140139616044e-05, + "loss": 0.3684, + "step": 45845 + }, + { + "epoch": 3.1813198251335786, + "grad_norm": 1.5986287593841553, + "learning_rate": 2.117794361603742e-05, + "loss": 0.5009, + "step": 45846 + }, + { + "epoch": 3.181389216570675, + "grad_norm": 0.8759949803352356, + "learning_rate": 2.1174486084746504e-05, + "loss": 0.318, + "step": 45847 + }, + { + "epoch": 3.181458608007772, + "grad_norm": 0.9433926939964294, + "learning_rate": 2.117102880229861e-05, + "loss": 0.2805, + "step": 45848 + }, + { + "epoch": 3.1815279994448686, + "grad_norm": 1.2331061363220215, + "learning_rate": 2.1167571768704607e-05, + "loss": 0.2939, + "step": 45849 + }, + { + "epoch": 3.181597390881965, + "grad_norm": 1.026025652885437, + "learning_rate": 2.1164114983975435e-05, + "loss": 0.1992, + "step": 45850 + }, + { + "epoch": 3.181666782319062, + "grad_norm": 0.8118530511856079, + "learning_rate": 2.1160658448122006e-05, + "loss": 0.2706, + "step": 45851 + }, + { + "epoch": 3.1817361737561587, + "grad_norm": 1.2051609754562378, + "learning_rate": 2.115720216115522e-05, + "loss": 0.452, + "step": 45852 + }, + { + "epoch": 3.181805565193255, + "grad_norm": 0.9148880839347839, + "learning_rate": 2.1153746123086037e-05, + "loss": 0.2208, + "step": 45853 + }, + { + "epoch": 3.181874956630352, + "grad_norm": 1.4813424348831177, + "learning_rate": 2.1150290333925303e-05, + "loss": 0.427, + "step": 45854 + }, + { + "epoch": 3.1819443480674483, + "grad_norm": 0.7250556349754333, + "learning_rate": 2.1146834793683944e-05, + "loss": 0.2959, + "step": 45855 + }, + { + "epoch": 3.182013739504545, + "grad_norm": 0.8705204129219055, + "learning_rate": 2.1143379502372894e-05, + "loss": 0.1414, + "step": 45856 + }, + { + "epoch": 3.182083130941642, + "grad_norm": 0.838524341583252, + "learning_rate": 2.1139924460003034e-05, + "loss": 0.1659, + "step": 45857 + }, + { + "epoch": 3.1821525223787384, + "grad_norm": 0.8289299607276917, + "learning_rate": 2.113646966658529e-05, + "loss": 0.2936, + "step": 45858 + }, + { + "epoch": 3.182221913815835, + "grad_norm": 1.2654180526733398, + "learning_rate": 2.113301512213056e-05, + "loss": 0.2661, + "step": 45859 + }, + { + "epoch": 3.1822913052529316, + "grad_norm": 0.7396421432495117, + "learning_rate": 2.1129560826649776e-05, + "loss": 0.176, + "step": 45860 + }, + { + "epoch": 3.1823606966900284, + "grad_norm": 0.9962454438209534, + "learning_rate": 2.1126106780153788e-05, + "loss": 0.3188, + "step": 45861 + }, + { + "epoch": 3.1824300881271252, + "grad_norm": 1.0710471868515015, + "learning_rate": 2.1122652982653536e-05, + "loss": 0.1895, + "step": 45862 + }, + { + "epoch": 3.1824994795642216, + "grad_norm": 0.6891605854034424, + "learning_rate": 2.111919943415991e-05, + "loss": 0.239, + "step": 45863 + }, + { + "epoch": 3.1825688710013185, + "grad_norm": 1.7743922472000122, + "learning_rate": 2.1115746134683823e-05, + "loss": 0.3069, + "step": 45864 + }, + { + "epoch": 3.1826382624384153, + "grad_norm": 1.0544003248214722, + "learning_rate": 2.111229308423618e-05, + "loss": 0.4293, + "step": 45865 + }, + { + "epoch": 3.1827076538755117, + "grad_norm": 1.669707179069519, + "learning_rate": 2.110884028282787e-05, + "loss": 0.4178, + "step": 45866 + }, + { + "epoch": 3.1827770453126085, + "grad_norm": 0.8486326932907104, + "learning_rate": 2.1105387730469827e-05, + "loss": 0.1922, + "step": 45867 + }, + { + "epoch": 3.182846436749705, + "grad_norm": 0.9451970458030701, + "learning_rate": 2.1101935427172892e-05, + "loss": 0.1844, + "step": 45868 + }, + { + "epoch": 3.1829158281868017, + "grad_norm": 0.6701275110244751, + "learning_rate": 2.1098483372947985e-05, + "loss": 0.2293, + "step": 45869 + }, + { + "epoch": 3.1829852196238986, + "grad_norm": 1.5353367328643799, + "learning_rate": 2.1095031567806024e-05, + "loss": 0.349, + "step": 45870 + }, + { + "epoch": 3.183054611060995, + "grad_norm": 0.7862856388092041, + "learning_rate": 2.1091580011757883e-05, + "loss": 0.2711, + "step": 45871 + }, + { + "epoch": 3.183124002498092, + "grad_norm": 1.0404696464538574, + "learning_rate": 2.108812870481448e-05, + "loss": 0.2119, + "step": 45872 + }, + { + "epoch": 3.1831933939351886, + "grad_norm": 0.6655060648918152, + "learning_rate": 2.1084677646986695e-05, + "loss": 0.2425, + "step": 45873 + }, + { + "epoch": 3.183262785372285, + "grad_norm": 1.1424490213394165, + "learning_rate": 2.108122683828545e-05, + "loss": 0.3198, + "step": 45874 + }, + { + "epoch": 3.183332176809382, + "grad_norm": 0.9166175127029419, + "learning_rate": 2.1077776278721596e-05, + "loss": 0.2672, + "step": 45875 + }, + { + "epoch": 3.1834015682464782, + "grad_norm": 0.9914047122001648, + "learning_rate": 2.107432596830604e-05, + "loss": 0.3409, + "step": 45876 + }, + { + "epoch": 3.183470959683575, + "grad_norm": 0.9040518999099731, + "learning_rate": 2.107087590704968e-05, + "loss": 0.2543, + "step": 45877 + }, + { + "epoch": 3.183540351120672, + "grad_norm": 1.0064733028411865, + "learning_rate": 2.1067426094963404e-05, + "loss": 0.196, + "step": 45878 + }, + { + "epoch": 3.1836097425577683, + "grad_norm": 0.9899826645851135, + "learning_rate": 2.106397653205814e-05, + "loss": 0.2595, + "step": 45879 + }, + { + "epoch": 3.183679133994865, + "grad_norm": 0.9871460795402527, + "learning_rate": 2.1060527218344717e-05, + "loss": 0.2859, + "step": 45880 + }, + { + "epoch": 3.1837485254319615, + "grad_norm": 1.1552696228027344, + "learning_rate": 2.1057078153834043e-05, + "loss": 0.2679, + "step": 45881 + }, + { + "epoch": 3.1838179168690584, + "grad_norm": 0.8345674872398376, + "learning_rate": 2.1053629338537016e-05, + "loss": 0.2384, + "step": 45882 + }, + { + "epoch": 3.183887308306155, + "grad_norm": 0.9308041334152222, + "learning_rate": 2.1050180772464557e-05, + "loss": 0.2808, + "step": 45883 + }, + { + "epoch": 3.1839566997432516, + "grad_norm": 1.2993863821029663, + "learning_rate": 2.1046732455627483e-05, + "loss": 0.3139, + "step": 45884 + }, + { + "epoch": 3.1840260911803484, + "grad_norm": 0.7531890869140625, + "learning_rate": 2.1043284388036722e-05, + "loss": 0.2301, + "step": 45885 + }, + { + "epoch": 3.184095482617445, + "grad_norm": 0.9727701544761658, + "learning_rate": 2.1039836569703142e-05, + "loss": 0.2823, + "step": 45886 + }, + { + "epoch": 3.1841648740545416, + "grad_norm": 0.7476885914802551, + "learning_rate": 2.1036389000637667e-05, + "loss": 0.2612, + "step": 45887 + }, + { + "epoch": 3.1842342654916385, + "grad_norm": 0.983555018901825, + "learning_rate": 2.1032941680851126e-05, + "loss": 0.3001, + "step": 45888 + }, + { + "epoch": 3.184303656928735, + "grad_norm": 0.9847845435142517, + "learning_rate": 2.1029494610354428e-05, + "loss": 0.1891, + "step": 45889 + }, + { + "epoch": 3.1843730483658317, + "grad_norm": 0.9774227142333984, + "learning_rate": 2.1026047789158442e-05, + "loss": 0.2874, + "step": 45890 + }, + { + "epoch": 3.1844424398029285, + "grad_norm": 0.8905367851257324, + "learning_rate": 2.1022601217274075e-05, + "loss": 0.2652, + "step": 45891 + }, + { + "epoch": 3.184511831240025, + "grad_norm": 0.9471863508224487, + "learning_rate": 2.101915489471219e-05, + "loss": 0.2369, + "step": 45892 + }, + { + "epoch": 3.1845812226771217, + "grad_norm": 1.1677461862564087, + "learning_rate": 2.101570882148366e-05, + "loss": 0.3247, + "step": 45893 + }, + { + "epoch": 3.184650614114218, + "grad_norm": 0.8717394471168518, + "learning_rate": 2.1012262997599408e-05, + "loss": 0.2203, + "step": 45894 + }, + { + "epoch": 3.184720005551315, + "grad_norm": 1.1658837795257568, + "learning_rate": 2.1008817423070248e-05, + "loss": 0.2633, + "step": 45895 + }, + { + "epoch": 3.184789396988412, + "grad_norm": 0.7507005333900452, + "learning_rate": 2.1005372097907082e-05, + "loss": 0.2169, + "step": 45896 + }, + { + "epoch": 3.184858788425508, + "grad_norm": 0.6756687164306641, + "learning_rate": 2.1001927022120804e-05, + "loss": 0.1758, + "step": 45897 + }, + { + "epoch": 3.184928179862605, + "grad_norm": 0.6877245903015137, + "learning_rate": 2.099848219572227e-05, + "loss": 0.2169, + "step": 45898 + }, + { + "epoch": 3.184997571299702, + "grad_norm": 0.7855391502380371, + "learning_rate": 2.0995037618722358e-05, + "loss": 0.2432, + "step": 45899 + }, + { + "epoch": 3.1850669627367982, + "grad_norm": 0.8700243234634399, + "learning_rate": 2.099159329113195e-05, + "loss": 0.1976, + "step": 45900 + }, + { + "epoch": 3.185136354173895, + "grad_norm": 1.1181973218917847, + "learning_rate": 2.098814921296194e-05, + "loss": 0.305, + "step": 45901 + }, + { + "epoch": 3.1852057456109915, + "grad_norm": 1.297367811203003, + "learning_rate": 2.0984705384223157e-05, + "loss": 0.3587, + "step": 45902 + }, + { + "epoch": 3.1852751370480883, + "grad_norm": 0.7169619798660278, + "learning_rate": 2.0981261804926477e-05, + "loss": 0.1151, + "step": 45903 + }, + { + "epoch": 3.185344528485185, + "grad_norm": 0.8344098925590515, + "learning_rate": 2.09778184750828e-05, + "loss": 0.2255, + "step": 45904 + }, + { + "epoch": 3.1854139199222815, + "grad_norm": 0.9210313558578491, + "learning_rate": 2.0974375394702973e-05, + "loss": 0.2593, + "step": 45905 + }, + { + "epoch": 3.1854833113593783, + "grad_norm": 0.7188798785209656, + "learning_rate": 2.0970932563797873e-05, + "loss": 0.166, + "step": 45906 + }, + { + "epoch": 3.1855527027964747, + "grad_norm": 0.5371579527854919, + "learning_rate": 2.0967489982378375e-05, + "loss": 0.1522, + "step": 45907 + }, + { + "epoch": 3.1856220942335716, + "grad_norm": 0.5594844222068787, + "learning_rate": 2.0964047650455363e-05, + "loss": 0.1206, + "step": 45908 + }, + { + "epoch": 3.1856914856706684, + "grad_norm": 0.7413569688796997, + "learning_rate": 2.0960605568039672e-05, + "loss": 0.2309, + "step": 45909 + }, + { + "epoch": 3.185760877107765, + "grad_norm": 0.8603000640869141, + "learning_rate": 2.095716373514216e-05, + "loss": 0.275, + "step": 45910 + }, + { + "epoch": 3.1858302685448616, + "grad_norm": 0.710612952709198, + "learning_rate": 2.0953722151773724e-05, + "loss": 0.2766, + "step": 45911 + }, + { + "epoch": 3.185899659981958, + "grad_norm": 0.8133552670478821, + "learning_rate": 2.09502808179452e-05, + "loss": 0.3075, + "step": 45912 + }, + { + "epoch": 3.185969051419055, + "grad_norm": 0.9716134667396545, + "learning_rate": 2.0946839733667488e-05, + "loss": 0.2421, + "step": 45913 + }, + { + "epoch": 3.1860384428561517, + "grad_norm": 0.840939998626709, + "learning_rate": 2.094339889895145e-05, + "loss": 0.2163, + "step": 45914 + }, + { + "epoch": 3.186107834293248, + "grad_norm": 1.0183637142181396, + "learning_rate": 2.0939958313807893e-05, + "loss": 0.3251, + "step": 45915 + }, + { + "epoch": 3.186177225730345, + "grad_norm": 1.0315864086151123, + "learning_rate": 2.0936517978247727e-05, + "loss": 0.4274, + "step": 45916 + }, + { + "epoch": 3.1862466171674417, + "grad_norm": 0.8678873181343079, + "learning_rate": 2.0933077892281795e-05, + "loss": 0.274, + "step": 45917 + }, + { + "epoch": 3.186316008604538, + "grad_norm": 0.9028254747390747, + "learning_rate": 2.0929638055920987e-05, + "loss": 0.215, + "step": 45918 + }, + { + "epoch": 3.186385400041635, + "grad_norm": 1.373095154762268, + "learning_rate": 2.0926198469176118e-05, + "loss": 0.2747, + "step": 45919 + }, + { + "epoch": 3.1864547914787313, + "grad_norm": 0.7016729116439819, + "learning_rate": 2.0922759132058068e-05, + "loss": 0.2094, + "step": 45920 + }, + { + "epoch": 3.186524182915828, + "grad_norm": 0.933979868888855, + "learning_rate": 2.0919320044577683e-05, + "loss": 0.3021, + "step": 45921 + }, + { + "epoch": 3.186593574352925, + "grad_norm": 0.9195874929428101, + "learning_rate": 2.0915881206745857e-05, + "loss": 0.3557, + "step": 45922 + }, + { + "epoch": 3.1866629657900214, + "grad_norm": 1.1863540410995483, + "learning_rate": 2.09124426185734e-05, + "loss": 0.3392, + "step": 45923 + }, + { + "epoch": 3.1867323572271182, + "grad_norm": 0.7745488882064819, + "learning_rate": 2.0909004280071187e-05, + "loss": 0.2466, + "step": 45924 + }, + { + "epoch": 3.186801748664215, + "grad_norm": 0.8346506953239441, + "learning_rate": 2.090556619125006e-05, + "loss": 0.2338, + "step": 45925 + }, + { + "epoch": 3.1868711401013115, + "grad_norm": 0.7934303879737854, + "learning_rate": 2.0902128352120898e-05, + "loss": 0.2756, + "step": 45926 + }, + { + "epoch": 3.1869405315384083, + "grad_norm": 0.7373437285423279, + "learning_rate": 2.0898690762694527e-05, + "loss": 0.2129, + "step": 45927 + }, + { + "epoch": 3.1870099229755047, + "grad_norm": 1.0788484811782837, + "learning_rate": 2.089525342298182e-05, + "loss": 0.247, + "step": 45928 + }, + { + "epoch": 3.1870793144126015, + "grad_norm": 0.6406776309013367, + "learning_rate": 2.089181633299364e-05, + "loss": 0.1856, + "step": 45929 + }, + { + "epoch": 3.1871487058496983, + "grad_norm": 0.9929965138435364, + "learning_rate": 2.0888379492740794e-05, + "loss": 0.3114, + "step": 45930 + }, + { + "epoch": 3.1872180972867947, + "grad_norm": 1.078207015991211, + "learning_rate": 2.088494290223415e-05, + "loss": 0.2532, + "step": 45931 + }, + { + "epoch": 3.1872874887238916, + "grad_norm": 0.9888738393783569, + "learning_rate": 2.0881506561484578e-05, + "loss": 0.3654, + "step": 45932 + }, + { + "epoch": 3.187356880160988, + "grad_norm": 0.7674580216407776, + "learning_rate": 2.0878070470502896e-05, + "loss": 0.2426, + "step": 45933 + }, + { + "epoch": 3.187426271598085, + "grad_norm": 0.8542605638504028, + "learning_rate": 2.0874634629299972e-05, + "loss": 0.1911, + "step": 45934 + }, + { + "epoch": 3.1874956630351816, + "grad_norm": 1.2642688751220703, + "learning_rate": 2.0871199037886637e-05, + "loss": 0.2863, + "step": 45935 + }, + { + "epoch": 3.187565054472278, + "grad_norm": 0.8703367710113525, + "learning_rate": 2.086776369627378e-05, + "loss": 0.1829, + "step": 45936 + }, + { + "epoch": 3.187634445909375, + "grad_norm": 0.7626585960388184, + "learning_rate": 2.0864328604472184e-05, + "loss": 0.2554, + "step": 45937 + }, + { + "epoch": 3.1877038373464712, + "grad_norm": 1.1922372579574585, + "learning_rate": 2.086089376249273e-05, + "loss": 0.3759, + "step": 45938 + }, + { + "epoch": 3.187773228783568, + "grad_norm": 0.7640543580055237, + "learning_rate": 2.085745917034624e-05, + "loss": 0.156, + "step": 45939 + }, + { + "epoch": 3.187842620220665, + "grad_norm": 1.1541039943695068, + "learning_rate": 2.0854024828043573e-05, + "loss": 0.1743, + "step": 45940 + }, + { + "epoch": 3.1879120116577613, + "grad_norm": 0.9528263807296753, + "learning_rate": 2.0850590735595564e-05, + "loss": 0.249, + "step": 45941 + }, + { + "epoch": 3.187981403094858, + "grad_norm": 0.8191843032836914, + "learning_rate": 2.0847156893013066e-05, + "loss": 0.1347, + "step": 45942 + }, + { + "epoch": 3.188050794531955, + "grad_norm": 0.8442957401275635, + "learning_rate": 2.0843723300306927e-05, + "loss": 0.2972, + "step": 45943 + }, + { + "epoch": 3.1881201859690513, + "grad_norm": 0.7534186244010925, + "learning_rate": 2.084028995748796e-05, + "loss": 0.2486, + "step": 45944 + }, + { + "epoch": 3.188189577406148, + "grad_norm": 1.0072990655899048, + "learning_rate": 2.083685686456699e-05, + "loss": 0.2848, + "step": 45945 + }, + { + "epoch": 3.1882589688432446, + "grad_norm": 1.115007758140564, + "learning_rate": 2.0833424021554893e-05, + "loss": 0.3361, + "step": 45946 + }, + { + "epoch": 3.1883283602803414, + "grad_norm": 1.06412672996521, + "learning_rate": 2.0829991428462493e-05, + "loss": 0.3344, + "step": 45947 + }, + { + "epoch": 3.1883977517174382, + "grad_norm": 1.0888630151748657, + "learning_rate": 2.082655908530061e-05, + "loss": 0.2393, + "step": 45948 + }, + { + "epoch": 3.1884671431545346, + "grad_norm": 1.306057095527649, + "learning_rate": 2.0823126992080135e-05, + "loss": 0.312, + "step": 45949 + }, + { + "epoch": 3.1885365345916314, + "grad_norm": 0.8529720306396484, + "learning_rate": 2.0819695148811837e-05, + "loss": 0.3328, + "step": 45950 + }, + { + "epoch": 3.1886059260287283, + "grad_norm": 0.9562587738037109, + "learning_rate": 2.0816263555506566e-05, + "loss": 0.2364, + "step": 45951 + }, + { + "epoch": 3.1886753174658247, + "grad_norm": 1.6732090711593628, + "learning_rate": 2.0812832212175193e-05, + "loss": 0.297, + "step": 45952 + }, + { + "epoch": 3.1887447089029215, + "grad_norm": 1.0792417526245117, + "learning_rate": 2.0809401118828487e-05, + "loss": 0.2374, + "step": 45953 + }, + { + "epoch": 3.188814100340018, + "grad_norm": 0.8454976081848145, + "learning_rate": 2.0805970275477325e-05, + "loss": 0.2629, + "step": 45954 + }, + { + "epoch": 3.1888834917771147, + "grad_norm": 1.7518103122711182, + "learning_rate": 2.080253968213253e-05, + "loss": 0.3453, + "step": 45955 + }, + { + "epoch": 3.1889528832142116, + "grad_norm": 1.2863104343414307, + "learning_rate": 2.0799109338804912e-05, + "loss": 0.359, + "step": 45956 + }, + { + "epoch": 3.189022274651308, + "grad_norm": 0.7740558385848999, + "learning_rate": 2.079567924550535e-05, + "loss": 0.2888, + "step": 45957 + }, + { + "epoch": 3.1890916660884048, + "grad_norm": 0.8012741804122925, + "learning_rate": 2.0792249402244625e-05, + "loss": 0.2997, + "step": 45958 + }, + { + "epoch": 3.189161057525501, + "grad_norm": 0.9073081016540527, + "learning_rate": 2.0788819809033567e-05, + "loss": 0.4376, + "step": 45959 + }, + { + "epoch": 3.189230448962598, + "grad_norm": 0.760221540927887, + "learning_rate": 2.0785390465883014e-05, + "loss": 0.2628, + "step": 45960 + }, + { + "epoch": 3.189299840399695, + "grad_norm": 0.9622036814689636, + "learning_rate": 2.0781961372803793e-05, + "loss": 0.2817, + "step": 45961 + }, + { + "epoch": 3.189369231836791, + "grad_norm": 0.8817417621612549, + "learning_rate": 2.0778532529806737e-05, + "loss": 0.2394, + "step": 45962 + }, + { + "epoch": 3.189438623273888, + "grad_norm": 0.858613908290863, + "learning_rate": 2.077510393690265e-05, + "loss": 0.2831, + "step": 45963 + }, + { + "epoch": 3.1895080147109844, + "grad_norm": 1.2279382944107056, + "learning_rate": 2.0771675594102412e-05, + "loss": 0.3392, + "step": 45964 + }, + { + "epoch": 3.1895774061480813, + "grad_norm": 0.8514189124107361, + "learning_rate": 2.076824750141676e-05, + "loss": 0.3705, + "step": 45965 + }, + { + "epoch": 3.189646797585178, + "grad_norm": 0.9373727440834045, + "learning_rate": 2.0764819658856562e-05, + "loss": 0.2187, + "step": 45966 + }, + { + "epoch": 3.1897161890222745, + "grad_norm": 0.8035327196121216, + "learning_rate": 2.076139206643264e-05, + "loss": 0.1467, + "step": 45967 + }, + { + "epoch": 3.1897855804593713, + "grad_norm": 0.6072701215744019, + "learning_rate": 2.075796472415581e-05, + "loss": 0.1297, + "step": 45968 + }, + { + "epoch": 3.189854971896468, + "grad_norm": 0.9422358870506287, + "learning_rate": 2.0754537632036887e-05, + "loss": 0.2853, + "step": 45969 + }, + { + "epoch": 3.1899243633335645, + "grad_norm": 0.8640759587287903, + "learning_rate": 2.0751110790086715e-05, + "loss": 0.2611, + "step": 45970 + }, + { + "epoch": 3.1899937547706614, + "grad_norm": 0.9262075424194336, + "learning_rate": 2.0747684198316075e-05, + "loss": 0.1936, + "step": 45971 + }, + { + "epoch": 3.1900631462077578, + "grad_norm": 0.6599398255348206, + "learning_rate": 2.0744257856735804e-05, + "loss": 0.1237, + "step": 45972 + }, + { + "epoch": 3.1901325376448546, + "grad_norm": 0.7800513505935669, + "learning_rate": 2.0740831765356706e-05, + "loss": 0.2187, + "step": 45973 + }, + { + "epoch": 3.1902019290819514, + "grad_norm": 0.9400511384010315, + "learning_rate": 2.073740592418961e-05, + "loss": 0.2322, + "step": 45974 + }, + { + "epoch": 3.190271320519048, + "grad_norm": 1.4545296430587769, + "learning_rate": 2.073398033324533e-05, + "loss": 0.2868, + "step": 45975 + }, + { + "epoch": 3.1903407119561447, + "grad_norm": 1.2417629957199097, + "learning_rate": 2.0730554992534666e-05, + "loss": 0.1939, + "step": 45976 + }, + { + "epoch": 3.1904101033932415, + "grad_norm": 0.7016326189041138, + "learning_rate": 2.072712990206849e-05, + "loss": 0.2377, + "step": 45977 + }, + { + "epoch": 3.190479494830338, + "grad_norm": 0.8211874961853027, + "learning_rate": 2.0723705061857525e-05, + "loss": 0.2559, + "step": 45978 + }, + { + "epoch": 3.1905488862674347, + "grad_norm": 0.924282431602478, + "learning_rate": 2.0720280471912634e-05, + "loss": 0.2862, + "step": 45979 + }, + { + "epoch": 3.190618277704531, + "grad_norm": 1.1292179822921753, + "learning_rate": 2.0716856132244612e-05, + "loss": 0.286, + "step": 45980 + }, + { + "epoch": 3.190687669141628, + "grad_norm": 0.8300514817237854, + "learning_rate": 2.0713432042864277e-05, + "loss": 0.172, + "step": 45981 + }, + { + "epoch": 3.1907570605787248, + "grad_norm": 0.8781158328056335, + "learning_rate": 2.0710008203782438e-05, + "loss": 0.2298, + "step": 45982 + }, + { + "epoch": 3.190826452015821, + "grad_norm": 0.9579910039901733, + "learning_rate": 2.0706584615009917e-05, + "loss": 0.2316, + "step": 45983 + }, + { + "epoch": 3.190895843452918, + "grad_norm": 0.6382492780685425, + "learning_rate": 2.0703161276557513e-05, + "loss": 0.1806, + "step": 45984 + }, + { + "epoch": 3.1909652348900144, + "grad_norm": 1.0661187171936035, + "learning_rate": 2.0699738188436023e-05, + "loss": 0.2656, + "step": 45985 + }, + { + "epoch": 3.191034626327111, + "grad_norm": 0.8477088212966919, + "learning_rate": 2.069631535065625e-05, + "loss": 0.2579, + "step": 45986 + }, + { + "epoch": 3.191104017764208, + "grad_norm": 0.8600766062736511, + "learning_rate": 2.0692892763229033e-05, + "loss": 0.2305, + "step": 45987 + }, + { + "epoch": 3.1911734092013044, + "grad_norm": 0.8427049517631531, + "learning_rate": 2.0689470426165126e-05, + "loss": 0.2495, + "step": 45988 + }, + { + "epoch": 3.1912428006384013, + "grad_norm": 0.8464354276657104, + "learning_rate": 2.068604833947536e-05, + "loss": 0.2807, + "step": 45989 + }, + { + "epoch": 3.1913121920754977, + "grad_norm": 0.666280210018158, + "learning_rate": 2.0682626503170543e-05, + "loss": 0.15, + "step": 45990 + }, + { + "epoch": 3.1913815835125945, + "grad_norm": 1.4375624656677246, + "learning_rate": 2.0679204917261497e-05, + "loss": 0.2325, + "step": 45991 + }, + { + "epoch": 3.1914509749496913, + "grad_norm": 0.6225181221961975, + "learning_rate": 2.067578358175897e-05, + "loss": 0.1437, + "step": 45992 + }, + { + "epoch": 3.1915203663867877, + "grad_norm": 1.002570390701294, + "learning_rate": 2.067236249667378e-05, + "loss": 0.3223, + "step": 45993 + }, + { + "epoch": 3.1915897578238845, + "grad_norm": 1.258582353591919, + "learning_rate": 2.066894166201675e-05, + "loss": 0.2857, + "step": 45994 + }, + { + "epoch": 3.1916591492609814, + "grad_norm": 0.6406968235969543, + "learning_rate": 2.0665521077798666e-05, + "loss": 0.1828, + "step": 45995 + }, + { + "epoch": 3.1917285406980778, + "grad_norm": 0.8960352540016174, + "learning_rate": 2.0662100744030322e-05, + "loss": 0.1623, + "step": 45996 + }, + { + "epoch": 3.1917979321351746, + "grad_norm": 0.8975418210029602, + "learning_rate": 2.065868066072253e-05, + "loss": 0.2344, + "step": 45997 + }, + { + "epoch": 3.191867323572271, + "grad_norm": 0.910476803779602, + "learning_rate": 2.0655260827886092e-05, + "loss": 0.2979, + "step": 45998 + }, + { + "epoch": 3.191936715009368, + "grad_norm": 0.7007210850715637, + "learning_rate": 2.0651841245531778e-05, + "loss": 0.1726, + "step": 45999 + }, + { + "epoch": 3.1920061064464647, + "grad_norm": 0.8898947834968567, + "learning_rate": 2.064842191367038e-05, + "loss": 0.3168, + "step": 46000 + }, + { + "epoch": 3.192075497883561, + "grad_norm": 0.8475421071052551, + "learning_rate": 2.064500283231272e-05, + "loss": 0.3013, + "step": 46001 + }, + { + "epoch": 3.192144889320658, + "grad_norm": 1.1446515321731567, + "learning_rate": 2.064158400146957e-05, + "loss": 0.2752, + "step": 46002 + }, + { + "epoch": 3.1922142807577547, + "grad_norm": 1.178894281387329, + "learning_rate": 2.0638165421151733e-05, + "loss": 0.4444, + "step": 46003 + }, + { + "epoch": 3.192283672194851, + "grad_norm": 0.7738179564476013, + "learning_rate": 2.063474709137001e-05, + "loss": 0.2062, + "step": 46004 + }, + { + "epoch": 3.192353063631948, + "grad_norm": 0.887025773525238, + "learning_rate": 2.0631329012135203e-05, + "loss": 0.206, + "step": 46005 + }, + { + "epoch": 3.1924224550690443, + "grad_norm": 1.014075756072998, + "learning_rate": 2.062791118345805e-05, + "loss": 0.228, + "step": 46006 + }, + { + "epoch": 3.192491846506141, + "grad_norm": 0.7499421834945679, + "learning_rate": 2.062449360534938e-05, + "loss": 0.3005, + "step": 46007 + }, + { + "epoch": 3.192561237943238, + "grad_norm": 0.894006609916687, + "learning_rate": 2.0621076277819972e-05, + "loss": 0.2861, + "step": 46008 + }, + { + "epoch": 3.1926306293803344, + "grad_norm": 0.996532678604126, + "learning_rate": 2.061765920088061e-05, + "loss": 0.2845, + "step": 46009 + }, + { + "epoch": 3.192700020817431, + "grad_norm": 1.261800765991211, + "learning_rate": 2.061424237454209e-05, + "loss": 0.4185, + "step": 46010 + }, + { + "epoch": 3.1927694122545276, + "grad_norm": 1.2023004293441772, + "learning_rate": 2.0610825798815205e-05, + "loss": 0.2647, + "step": 46011 + }, + { + "epoch": 3.1928388036916244, + "grad_norm": 1.341143250465393, + "learning_rate": 2.0607409473710758e-05, + "loss": 0.2519, + "step": 46012 + }, + { + "epoch": 3.1929081951287213, + "grad_norm": 0.8513938784599304, + "learning_rate": 2.0603993399239475e-05, + "loss": 0.2845, + "step": 46013 + }, + { + "epoch": 3.1929775865658176, + "grad_norm": 1.248476505279541, + "learning_rate": 2.0600577575412172e-05, + "loss": 0.3024, + "step": 46014 + }, + { + "epoch": 3.1930469780029145, + "grad_norm": 0.9732199907302856, + "learning_rate": 2.0597162002239632e-05, + "loss": 0.2879, + "step": 46015 + }, + { + "epoch": 3.193116369440011, + "grad_norm": 0.8862155079841614, + "learning_rate": 2.059374667973265e-05, + "loss": 0.2077, + "step": 46016 + }, + { + "epoch": 3.1931857608771077, + "grad_norm": 0.7447839975357056, + "learning_rate": 2.059033160790199e-05, + "loss": 0.1601, + "step": 46017 + }, + { + "epoch": 3.1932551523142045, + "grad_norm": 0.889872670173645, + "learning_rate": 2.0586916786758447e-05, + "loss": 0.2785, + "step": 46018 + }, + { + "epoch": 3.193324543751301, + "grad_norm": 0.7471600770950317, + "learning_rate": 2.058350221631281e-05, + "loss": 0.2734, + "step": 46019 + }, + { + "epoch": 3.1933939351883978, + "grad_norm": 0.8958613872528076, + "learning_rate": 2.0580087896575817e-05, + "loss": 0.2437, + "step": 46020 + }, + { + "epoch": 3.1934633266254946, + "grad_norm": 0.9783086776733398, + "learning_rate": 2.057667382755828e-05, + "loss": 0.2998, + "step": 46021 + }, + { + "epoch": 3.193532718062591, + "grad_norm": 0.7531139850616455, + "learning_rate": 2.0573260009270988e-05, + "loss": 0.1707, + "step": 46022 + }, + { + "epoch": 3.193602109499688, + "grad_norm": 0.751028299331665, + "learning_rate": 2.0569846441724683e-05, + "loss": 0.2696, + "step": 46023 + }, + { + "epoch": 3.193671500936784, + "grad_norm": 0.844909131526947, + "learning_rate": 2.056643312493014e-05, + "loss": 0.3779, + "step": 46024 + }, + { + "epoch": 3.193740892373881, + "grad_norm": 0.8427475094795227, + "learning_rate": 2.0563020058898162e-05, + "loss": 0.2459, + "step": 46025 + }, + { + "epoch": 3.193810283810978, + "grad_norm": 0.9752976894378662, + "learning_rate": 2.0559607243639545e-05, + "loss": 0.1654, + "step": 46026 + }, + { + "epoch": 3.1938796752480743, + "grad_norm": 0.838807225227356, + "learning_rate": 2.0556194679164998e-05, + "loss": 0.3598, + "step": 46027 + }, + { + "epoch": 3.193949066685171, + "grad_norm": 0.9018200039863586, + "learning_rate": 2.0552782365485323e-05, + "loss": 0.1856, + "step": 46028 + }, + { + "epoch": 3.194018458122268, + "grad_norm": 0.914247453212738, + "learning_rate": 2.0549370302611304e-05, + "loss": 0.3368, + "step": 46029 + }, + { + "epoch": 3.1940878495593643, + "grad_norm": 0.7489761114120483, + "learning_rate": 2.0545958490553708e-05, + "loss": 0.181, + "step": 46030 + }, + { + "epoch": 3.194157240996461, + "grad_norm": 1.945569634437561, + "learning_rate": 2.0542546929323294e-05, + "loss": 0.205, + "step": 46031 + }, + { + "epoch": 3.1942266324335575, + "grad_norm": 0.8061253428459167, + "learning_rate": 2.0539135618930837e-05, + "loss": 0.2531, + "step": 46032 + }, + { + "epoch": 3.1942960238706544, + "grad_norm": 1.0272510051727295, + "learning_rate": 2.0535724559387148e-05, + "loss": 0.2051, + "step": 46033 + }, + { + "epoch": 3.194365415307751, + "grad_norm": 1.0175809860229492, + "learning_rate": 2.0532313750702925e-05, + "loss": 0.3358, + "step": 46034 + }, + { + "epoch": 3.1944348067448476, + "grad_norm": 1.0934011936187744, + "learning_rate": 2.052890319288897e-05, + "loss": 0.3387, + "step": 46035 + }, + { + "epoch": 3.1945041981819444, + "grad_norm": 1.040634274482727, + "learning_rate": 2.0525492885956042e-05, + "loss": 0.3333, + "step": 46036 + }, + { + "epoch": 3.194573589619041, + "grad_norm": 0.9789066910743713, + "learning_rate": 2.052208282991491e-05, + "loss": 0.3629, + "step": 46037 + }, + { + "epoch": 3.1946429810561376, + "grad_norm": 1.0879685878753662, + "learning_rate": 2.0518673024776347e-05, + "loss": 0.2465, + "step": 46038 + }, + { + "epoch": 3.1947123724932345, + "grad_norm": 0.9737522602081299, + "learning_rate": 2.0515263470551117e-05, + "loss": 0.3645, + "step": 46039 + }, + { + "epoch": 3.194781763930331, + "grad_norm": 1.417963981628418, + "learning_rate": 2.051185416725e-05, + "loss": 0.3346, + "step": 46040 + }, + { + "epoch": 3.1948511553674277, + "grad_norm": 1.222184658050537, + "learning_rate": 2.0508445114883714e-05, + "loss": 0.1883, + "step": 46041 + }, + { + "epoch": 3.194920546804524, + "grad_norm": 1.2061305046081543, + "learning_rate": 2.0505036313463034e-05, + "loss": 0.3417, + "step": 46042 + }, + { + "epoch": 3.194989938241621, + "grad_norm": 0.7007421255111694, + "learning_rate": 2.050162776299873e-05, + "loss": 0.1487, + "step": 46043 + }, + { + "epoch": 3.1950593296787178, + "grad_norm": 0.8563266396522522, + "learning_rate": 2.0498219463501577e-05, + "loss": 0.2158, + "step": 46044 + }, + { + "epoch": 3.195128721115814, + "grad_norm": 0.9578967690467834, + "learning_rate": 2.0494811414982318e-05, + "loss": 0.2166, + "step": 46045 + }, + { + "epoch": 3.195198112552911, + "grad_norm": 0.9298904538154602, + "learning_rate": 2.049140361745173e-05, + "loss": 0.2479, + "step": 46046 + }, + { + "epoch": 3.195267503990008, + "grad_norm": 0.9267597198486328, + "learning_rate": 2.0487996070920544e-05, + "loss": 0.2667, + "step": 46047 + }, + { + "epoch": 3.195336895427104, + "grad_norm": 1.1001611948013306, + "learning_rate": 2.0484588775399528e-05, + "loss": 0.3038, + "step": 46048 + }, + { + "epoch": 3.195406286864201, + "grad_norm": 1.0429893732070923, + "learning_rate": 2.0481181730899422e-05, + "loss": 0.2448, + "step": 46049 + }, + { + "epoch": 3.1954756783012974, + "grad_norm": 0.7765690088272095, + "learning_rate": 2.0477774937431015e-05, + "loss": 0.3003, + "step": 46050 + }, + { + "epoch": 3.1955450697383943, + "grad_norm": 1.019179105758667, + "learning_rate": 2.047436839500505e-05, + "loss": 0.1989, + "step": 46051 + }, + { + "epoch": 3.195614461175491, + "grad_norm": 0.7327064871788025, + "learning_rate": 2.0470962103632262e-05, + "loss": 0.1528, + "step": 46052 + }, + { + "epoch": 3.1956838526125875, + "grad_norm": 2.1331753730773926, + "learning_rate": 2.0467556063323458e-05, + "loss": 0.2439, + "step": 46053 + }, + { + "epoch": 3.1957532440496843, + "grad_norm": 0.9625204205513, + "learning_rate": 2.046415027408932e-05, + "loss": 0.1864, + "step": 46054 + }, + { + "epoch": 3.195822635486781, + "grad_norm": 0.6156449317932129, + "learning_rate": 2.0460744735940642e-05, + "loss": 0.1886, + "step": 46055 + }, + { + "epoch": 3.1958920269238775, + "grad_norm": 0.8359969854354858, + "learning_rate": 2.0457339448888158e-05, + "loss": 0.2606, + "step": 46056 + }, + { + "epoch": 3.1959614183609744, + "grad_norm": 1.00796377658844, + "learning_rate": 2.0453934412942643e-05, + "loss": 0.2386, + "step": 46057 + }, + { + "epoch": 3.1960308097980707, + "grad_norm": 0.8419939279556274, + "learning_rate": 2.045052962811481e-05, + "loss": 0.2311, + "step": 46058 + }, + { + "epoch": 3.1961002012351676, + "grad_norm": 0.8469204902648926, + "learning_rate": 2.044712509441543e-05, + "loss": 0.2384, + "step": 46059 + }, + { + "epoch": 3.1961695926722644, + "grad_norm": 1.2485606670379639, + "learning_rate": 2.044372081185524e-05, + "loss": 0.3414, + "step": 46060 + }, + { + "epoch": 3.196238984109361, + "grad_norm": 0.9504008293151855, + "learning_rate": 2.0440316780445013e-05, + "loss": 0.2257, + "step": 46061 + }, + { + "epoch": 3.1963083755464576, + "grad_norm": 1.0132917165756226, + "learning_rate": 2.0436913000195458e-05, + "loss": 0.1827, + "step": 46062 + }, + { + "epoch": 3.196377766983554, + "grad_norm": 1.200770616531372, + "learning_rate": 2.043350947111734e-05, + "loss": 0.2613, + "step": 46063 + }, + { + "epoch": 3.196447158420651, + "grad_norm": 0.9091033935546875, + "learning_rate": 2.0430106193221398e-05, + "loss": 0.1987, + "step": 46064 + }, + { + "epoch": 3.1965165498577477, + "grad_norm": 0.7955341935157776, + "learning_rate": 2.0426703166518367e-05, + "loss": 0.2166, + "step": 46065 + }, + { + "epoch": 3.196585941294844, + "grad_norm": 1.1583462953567505, + "learning_rate": 2.042330039101902e-05, + "loss": 0.2578, + "step": 46066 + }, + { + "epoch": 3.196655332731941, + "grad_norm": 1.1123851537704468, + "learning_rate": 2.0419897866734096e-05, + "loss": 0.3322, + "step": 46067 + }, + { + "epoch": 3.1967247241690377, + "grad_norm": 1.0987837314605713, + "learning_rate": 2.04164955936743e-05, + "loss": 0.3097, + "step": 46068 + }, + { + "epoch": 3.196794115606134, + "grad_norm": 0.8334192633628845, + "learning_rate": 2.0413093571850395e-05, + "loss": 0.2527, + "step": 46069 + }, + { + "epoch": 3.196863507043231, + "grad_norm": 1.1964493989944458, + "learning_rate": 2.040969180127311e-05, + "loss": 0.3239, + "step": 46070 + }, + { + "epoch": 3.1969328984803274, + "grad_norm": 1.4402925968170166, + "learning_rate": 2.040629028195321e-05, + "loss": 0.2578, + "step": 46071 + }, + { + "epoch": 3.197002289917424, + "grad_norm": 0.9081660509109497, + "learning_rate": 2.0402889013901404e-05, + "loss": 0.1976, + "step": 46072 + }, + { + "epoch": 3.197071681354521, + "grad_norm": 0.9295480847358704, + "learning_rate": 2.0399487997128454e-05, + "loss": 0.2504, + "step": 46073 + }, + { + "epoch": 3.1971410727916174, + "grad_norm": 0.8531882166862488, + "learning_rate": 2.03960872316451e-05, + "loss": 0.2145, + "step": 46074 + }, + { + "epoch": 3.1972104642287142, + "grad_norm": 0.7754867672920227, + "learning_rate": 2.039268671746204e-05, + "loss": 0.1974, + "step": 46075 + }, + { + "epoch": 3.1972798556658106, + "grad_norm": 0.7778425812721252, + "learning_rate": 2.038928645459003e-05, + "loss": 0.1398, + "step": 46076 + }, + { + "epoch": 3.1973492471029075, + "grad_norm": 0.6463174223899841, + "learning_rate": 2.0385886443039813e-05, + "loss": 0.1597, + "step": 46077 + }, + { + "epoch": 3.1974186385400043, + "grad_norm": 0.7943038940429688, + "learning_rate": 2.038248668282211e-05, + "loss": 0.3416, + "step": 46078 + }, + { + "epoch": 3.1974880299771007, + "grad_norm": 0.6865927577018738, + "learning_rate": 2.0379087173947666e-05, + "loss": 0.2247, + "step": 46079 + }, + { + "epoch": 3.1975574214141975, + "grad_norm": 0.818531334400177, + "learning_rate": 2.03756879164272e-05, + "loss": 0.2034, + "step": 46080 + }, + { + "epoch": 3.1976268128512944, + "grad_norm": 0.9483010172843933, + "learning_rate": 2.0372288910271477e-05, + "loss": 0.3005, + "step": 46081 + }, + { + "epoch": 3.1976962042883907, + "grad_norm": 1.1724270582199097, + "learning_rate": 2.0368890155491173e-05, + "loss": 0.2169, + "step": 46082 + }, + { + "epoch": 3.1977655957254876, + "grad_norm": 1.0863879919052124, + "learning_rate": 2.0365491652097034e-05, + "loss": 0.3078, + "step": 46083 + }, + { + "epoch": 3.197834987162584, + "grad_norm": 0.7515571117401123, + "learning_rate": 2.0362093400099814e-05, + "loss": 0.25, + "step": 46084 + }, + { + "epoch": 3.197904378599681, + "grad_norm": 1.071831464767456, + "learning_rate": 2.035869539951022e-05, + "loss": 0.2781, + "step": 46085 + }, + { + "epoch": 3.1979737700367776, + "grad_norm": 0.8721191883087158, + "learning_rate": 2.0355297650338978e-05, + "loss": 0.2674, + "step": 46086 + }, + { + "epoch": 3.198043161473874, + "grad_norm": 1.1127396821975708, + "learning_rate": 2.0351900152596825e-05, + "loss": 0.3768, + "step": 46087 + }, + { + "epoch": 3.198112552910971, + "grad_norm": 0.6669632792472839, + "learning_rate": 2.034850290629451e-05, + "loss": 0.1664, + "step": 46088 + }, + { + "epoch": 3.1981819443480672, + "grad_norm": 1.0767930746078491, + "learning_rate": 2.0345105911442708e-05, + "loss": 0.3297, + "step": 46089 + }, + { + "epoch": 3.198251335785164, + "grad_norm": 0.8717469573020935, + "learning_rate": 2.034170916805216e-05, + "loss": 0.2459, + "step": 46090 + }, + { + "epoch": 3.198320727222261, + "grad_norm": 0.7437397241592407, + "learning_rate": 2.03383126761336e-05, + "loss": 0.2241, + "step": 46091 + }, + { + "epoch": 3.1983901186593573, + "grad_norm": 1.1869899034500122, + "learning_rate": 2.033491643569776e-05, + "loss": 0.2972, + "step": 46092 + }, + { + "epoch": 3.198459510096454, + "grad_norm": 1.1203466653823853, + "learning_rate": 2.0331520446755325e-05, + "loss": 0.1724, + "step": 46093 + }, + { + "epoch": 3.198528901533551, + "grad_norm": 0.8633742928504944, + "learning_rate": 2.032812470931703e-05, + "loss": 0.2779, + "step": 46094 + }, + { + "epoch": 3.1985982929706473, + "grad_norm": 1.0139237642288208, + "learning_rate": 2.0324729223393635e-05, + "loss": 0.2565, + "step": 46095 + }, + { + "epoch": 3.198667684407744, + "grad_norm": 0.9806751608848572, + "learning_rate": 2.0321333988995806e-05, + "loss": 0.299, + "step": 46096 + }, + { + "epoch": 3.1987370758448406, + "grad_norm": 0.7318734526634216, + "learning_rate": 2.0317939006134268e-05, + "loss": 0.225, + "step": 46097 + }, + { + "epoch": 3.1988064672819374, + "grad_norm": 0.830084502696991, + "learning_rate": 2.031454427481976e-05, + "loss": 0.1867, + "step": 46098 + }, + { + "epoch": 3.1988758587190342, + "grad_norm": 1.0499869585037231, + "learning_rate": 2.031114979506299e-05, + "loss": 0.2066, + "step": 46099 + }, + { + "epoch": 3.1989452501561306, + "grad_norm": 0.7412392497062683, + "learning_rate": 2.0307755566874675e-05, + "loss": 0.2624, + "step": 46100 + }, + { + "epoch": 3.1990146415932275, + "grad_norm": 1.2816052436828613, + "learning_rate": 2.030436159026553e-05, + "loss": 0.3809, + "step": 46101 + }, + { + "epoch": 3.1990840330303243, + "grad_norm": 0.9034192562103271, + "learning_rate": 2.030096786524629e-05, + "loss": 0.3342, + "step": 46102 + }, + { + "epoch": 3.1991534244674207, + "grad_norm": 1.0884528160095215, + "learning_rate": 2.0297574391827622e-05, + "loss": 0.2858, + "step": 46103 + }, + { + "epoch": 3.1992228159045175, + "grad_norm": 1.3117256164550781, + "learning_rate": 2.0294181170020266e-05, + "loss": 0.2998, + "step": 46104 + }, + { + "epoch": 3.199292207341614, + "grad_norm": 0.8039535880088806, + "learning_rate": 2.0290788199834933e-05, + "loss": 0.2279, + "step": 46105 + }, + { + "epoch": 3.1993615987787107, + "grad_norm": 0.7516803741455078, + "learning_rate": 2.0287395481282333e-05, + "loss": 0.258, + "step": 46106 + }, + { + "epoch": 3.1994309902158076, + "grad_norm": 0.6917749643325806, + "learning_rate": 2.0284003014373186e-05, + "loss": 0.1517, + "step": 46107 + }, + { + "epoch": 3.199500381652904, + "grad_norm": 0.717083215713501, + "learning_rate": 2.028061079911818e-05, + "loss": 0.2005, + "step": 46108 + }, + { + "epoch": 3.199569773090001, + "grad_norm": 1.2334390878677368, + "learning_rate": 2.0277218835528067e-05, + "loss": 0.298, + "step": 46109 + }, + { + "epoch": 3.199639164527097, + "grad_norm": 1.5735950469970703, + "learning_rate": 2.0273827123613497e-05, + "loss": 0.2851, + "step": 46110 + }, + { + "epoch": 3.199708555964194, + "grad_norm": 0.8224570155143738, + "learning_rate": 2.027043566338521e-05, + "loss": 0.2219, + "step": 46111 + }, + { + "epoch": 3.199777947401291, + "grad_norm": 1.7474653720855713, + "learning_rate": 2.0267044454853902e-05, + "loss": 0.2252, + "step": 46112 + }, + { + "epoch": 3.1998473388383872, + "grad_norm": 0.6524653434753418, + "learning_rate": 2.026365349803029e-05, + "loss": 0.1666, + "step": 46113 + }, + { + "epoch": 3.199916730275484, + "grad_norm": 0.8090639710426331, + "learning_rate": 2.0260262792925068e-05, + "loss": 0.2458, + "step": 46114 + }, + { + "epoch": 3.1999861217125805, + "grad_norm": 1.0728061199188232, + "learning_rate": 2.0256872339548948e-05, + "loss": 0.265, + "step": 46115 + }, + { + "epoch": 3.2000555131496773, + "grad_norm": 1.1745449304580688, + "learning_rate": 2.025348213791265e-05, + "loss": 0.324, + "step": 46116 + }, + { + "epoch": 3.200124904586774, + "grad_norm": 0.6073265671730042, + "learning_rate": 2.0250092188026836e-05, + "loss": 0.212, + "step": 46117 + }, + { + "epoch": 3.2001942960238705, + "grad_norm": 0.766865074634552, + "learning_rate": 2.0246702489902226e-05, + "loss": 0.1747, + "step": 46118 + }, + { + "epoch": 3.2002636874609673, + "grad_norm": 0.8289129137992859, + "learning_rate": 2.0243313043549527e-05, + "loss": 0.2403, + "step": 46119 + }, + { + "epoch": 3.200333078898064, + "grad_norm": 0.8173882961273193, + "learning_rate": 2.0239923848979424e-05, + "loss": 0.3715, + "step": 46120 + }, + { + "epoch": 3.2004024703351606, + "grad_norm": 0.9215728640556335, + "learning_rate": 2.0236534906202642e-05, + "loss": 0.2276, + "step": 46121 + }, + { + "epoch": 3.2004718617722574, + "grad_norm": 1.4386800527572632, + "learning_rate": 2.023314621522988e-05, + "loss": 0.3429, + "step": 46122 + }, + { + "epoch": 3.200541253209354, + "grad_norm": 1.1206068992614746, + "learning_rate": 2.02297577760718e-05, + "loss": 0.4099, + "step": 46123 + }, + { + "epoch": 3.2006106446464506, + "grad_norm": 1.031022310256958, + "learning_rate": 2.0226369588739125e-05, + "loss": 0.3071, + "step": 46124 + }, + { + "epoch": 3.2006800360835475, + "grad_norm": 1.3673549890518188, + "learning_rate": 2.0222981653242533e-05, + "loss": 0.2602, + "step": 46125 + }, + { + "epoch": 3.200749427520644, + "grad_norm": 0.7857504487037659, + "learning_rate": 2.0219593969592743e-05, + "loss": 0.2367, + "step": 46126 + }, + { + "epoch": 3.2008188189577407, + "grad_norm": 0.8811720609664917, + "learning_rate": 2.0216206537800464e-05, + "loss": 0.2845, + "step": 46127 + }, + { + "epoch": 3.2008882103948375, + "grad_norm": 0.9631436467170715, + "learning_rate": 2.0212819357876332e-05, + "loss": 0.2724, + "step": 46128 + }, + { + "epoch": 3.200957601831934, + "grad_norm": 0.9158031940460205, + "learning_rate": 2.0209432429831078e-05, + "loss": 0.2279, + "step": 46129 + }, + { + "epoch": 3.2010269932690307, + "grad_norm": 1.5663994550704956, + "learning_rate": 2.0206045753675408e-05, + "loss": 0.3367, + "step": 46130 + }, + { + "epoch": 3.201096384706127, + "grad_norm": 0.9175224304199219, + "learning_rate": 2.020265932941997e-05, + "loss": 0.3214, + "step": 46131 + }, + { + "epoch": 3.201165776143224, + "grad_norm": 0.8639481067657471, + "learning_rate": 2.019927315707548e-05, + "loss": 0.2464, + "step": 46132 + }, + { + "epoch": 3.201235167580321, + "grad_norm": 1.2749733924865723, + "learning_rate": 2.0195887236652622e-05, + "loss": 0.4161, + "step": 46133 + }, + { + "epoch": 3.201304559017417, + "grad_norm": 0.9627832174301147, + "learning_rate": 2.019250156816209e-05, + "loss": 0.3807, + "step": 46134 + }, + { + "epoch": 3.201373950454514, + "grad_norm": 1.2268829345703125, + "learning_rate": 2.0189116151614574e-05, + "loss": 0.3123, + "step": 46135 + }, + { + "epoch": 3.2014433418916104, + "grad_norm": 1.0597808361053467, + "learning_rate": 2.018573098702079e-05, + "loss": 0.2697, + "step": 46136 + }, + { + "epoch": 3.2015127333287072, + "grad_norm": 0.9087938666343689, + "learning_rate": 2.0182346074391357e-05, + "loss": 0.3069, + "step": 46137 + }, + { + "epoch": 3.201582124765804, + "grad_norm": 0.5878456234931946, + "learning_rate": 2.017896141373701e-05, + "loss": 0.18, + "step": 46138 + }, + { + "epoch": 3.2016515162029004, + "grad_norm": 1.0727907419204712, + "learning_rate": 2.0175577005068402e-05, + "loss": 0.3198, + "step": 46139 + }, + { + "epoch": 3.2017209076399973, + "grad_norm": 0.6168510913848877, + "learning_rate": 2.0172192848396245e-05, + "loss": 0.1193, + "step": 46140 + }, + { + "epoch": 3.2017902990770937, + "grad_norm": 0.6308680772781372, + "learning_rate": 2.0168808943731222e-05, + "loss": 0.1586, + "step": 46141 + }, + { + "epoch": 3.2018596905141905, + "grad_norm": 1.0893065929412842, + "learning_rate": 2.0165425291083994e-05, + "loss": 0.3612, + "step": 46142 + }, + { + "epoch": 3.2019290819512873, + "grad_norm": 1.0837491750717163, + "learning_rate": 2.016204189046529e-05, + "loss": 0.2486, + "step": 46143 + }, + { + "epoch": 3.2019984733883837, + "grad_norm": 0.9993512630462646, + "learning_rate": 2.015865874188573e-05, + "loss": 0.2474, + "step": 46144 + }, + { + "epoch": 3.2020678648254806, + "grad_norm": 0.5751276612281799, + "learning_rate": 2.0155275845356013e-05, + "loss": 0.1625, + "step": 46145 + }, + { + "epoch": 3.2021372562625774, + "grad_norm": 1.2289451360702515, + "learning_rate": 2.015189320088684e-05, + "loss": 0.1997, + "step": 46146 + }, + { + "epoch": 3.202206647699674, + "grad_norm": 0.8069050908088684, + "learning_rate": 2.014851080848886e-05, + "loss": 0.363, + "step": 46147 + }, + { + "epoch": 3.2022760391367706, + "grad_norm": 0.6247028112411499, + "learning_rate": 2.0145128668172773e-05, + "loss": 0.229, + "step": 46148 + }, + { + "epoch": 3.202345430573867, + "grad_norm": 0.9932435154914856, + "learning_rate": 2.0141746779949256e-05, + "loss": 0.2782, + "step": 46149 + }, + { + "epoch": 3.202414822010964, + "grad_norm": 0.7760623097419739, + "learning_rate": 2.0138365143829008e-05, + "loss": 0.2541, + "step": 46150 + }, + { + "epoch": 3.2024842134480607, + "grad_norm": 0.6931678056716919, + "learning_rate": 2.0134983759822635e-05, + "loss": 0.195, + "step": 46151 + }, + { + "epoch": 3.202553604885157, + "grad_norm": 0.8651004433631897, + "learning_rate": 2.013160262794086e-05, + "loss": 0.2791, + "step": 46152 + }, + { + "epoch": 3.202622996322254, + "grad_norm": 0.8667238354682922, + "learning_rate": 2.0128221748194352e-05, + "loss": 0.2576, + "step": 46153 + }, + { + "epoch": 3.2026923877593507, + "grad_norm": 1.0064468383789062, + "learning_rate": 2.0124841120593785e-05, + "loss": 0.3034, + "step": 46154 + }, + { + "epoch": 3.202761779196447, + "grad_norm": 2.3301806449890137, + "learning_rate": 2.0121460745149813e-05, + "loss": 0.4079, + "step": 46155 + }, + { + "epoch": 3.202831170633544, + "grad_norm": 0.6736888289451599, + "learning_rate": 2.0118080621873136e-05, + "loss": 0.1672, + "step": 46156 + }, + { + "epoch": 3.2029005620706403, + "grad_norm": 0.7953614592552185, + "learning_rate": 2.0114700750774428e-05, + "loss": 0.2133, + "step": 46157 + }, + { + "epoch": 3.202969953507737, + "grad_norm": 0.664890706539154, + "learning_rate": 2.0111321131864313e-05, + "loss": 0.1856, + "step": 46158 + }, + { + "epoch": 3.203039344944834, + "grad_norm": 0.9326421022415161, + "learning_rate": 2.0107941765153494e-05, + "loss": 0.2885, + "step": 46159 + }, + { + "epoch": 3.2031087363819304, + "grad_norm": 0.8745913505554199, + "learning_rate": 2.0104562650652638e-05, + "loss": 0.2472, + "step": 46160 + }, + { + "epoch": 3.203178127819027, + "grad_norm": 0.8748732805252075, + "learning_rate": 2.0101183788372402e-05, + "loss": 0.1984, + "step": 46161 + }, + { + "epoch": 3.2032475192561236, + "grad_norm": 0.9439599514007568, + "learning_rate": 2.009780517832348e-05, + "loss": 0.3855, + "step": 46162 + }, + { + "epoch": 3.2033169106932204, + "grad_norm": 0.8506670594215393, + "learning_rate": 2.00944268205165e-05, + "loss": 0.207, + "step": 46163 + }, + { + "epoch": 3.2033863021303173, + "grad_norm": 1.4111610651016235, + "learning_rate": 2.0091048714962136e-05, + "loss": 0.3871, + "step": 46164 + }, + { + "epoch": 3.2034556935674137, + "grad_norm": 0.8633244633674622, + "learning_rate": 2.008767086167109e-05, + "loss": 0.23, + "step": 46165 + }, + { + "epoch": 3.2035250850045105, + "grad_norm": 0.5619423985481262, + "learning_rate": 2.0084293260653974e-05, + "loss": 0.1379, + "step": 46166 + }, + { + "epoch": 3.203594476441607, + "grad_norm": 0.8823177814483643, + "learning_rate": 2.0080915911921472e-05, + "loss": 0.2617, + "step": 46167 + }, + { + "epoch": 3.2036638678787037, + "grad_norm": 0.7999131679534912, + "learning_rate": 2.0077538815484242e-05, + "loss": 0.1747, + "step": 46168 + }, + { + "epoch": 3.2037332593158006, + "grad_norm": 0.9546110033988953, + "learning_rate": 2.0074161971352944e-05, + "loss": 0.1739, + "step": 46169 + }, + { + "epoch": 3.203802650752897, + "grad_norm": 1.301192283630371, + "learning_rate": 2.007078537953825e-05, + "loss": 0.2337, + "step": 46170 + }, + { + "epoch": 3.2038720421899938, + "grad_norm": 1.2758349180221558, + "learning_rate": 2.0067409040050833e-05, + "loss": 0.3849, + "step": 46171 + }, + { + "epoch": 3.2039414336270906, + "grad_norm": 0.9159408211708069, + "learning_rate": 2.0064032952901312e-05, + "loss": 0.3201, + "step": 46172 + }, + { + "epoch": 3.204010825064187, + "grad_norm": 0.9420018792152405, + "learning_rate": 2.006065711810037e-05, + "loss": 0.3575, + "step": 46173 + }, + { + "epoch": 3.204080216501284, + "grad_norm": 1.2770684957504272, + "learning_rate": 2.0057281535658645e-05, + "loss": 0.2954, + "step": 46174 + }, + { + "epoch": 3.20414960793838, + "grad_norm": 0.9331624507904053, + "learning_rate": 2.0053906205586813e-05, + "loss": 0.3268, + "step": 46175 + }, + { + "epoch": 3.204218999375477, + "grad_norm": 0.6989194750785828, + "learning_rate": 2.0050531127895533e-05, + "loss": 0.1967, + "step": 46176 + }, + { + "epoch": 3.204288390812574, + "grad_norm": 0.8997901082038879, + "learning_rate": 2.004715630259544e-05, + "loss": 0.348, + "step": 46177 + }, + { + "epoch": 3.2043577822496703, + "grad_norm": 0.5898043513298035, + "learning_rate": 2.004378172969722e-05, + "loss": 0.1617, + "step": 46178 + }, + { + "epoch": 3.204427173686767, + "grad_norm": 0.9710648059844971, + "learning_rate": 2.004040740921149e-05, + "loss": 0.2526, + "step": 46179 + }, + { + "epoch": 3.204496565123864, + "grad_norm": 0.8110800385475159, + "learning_rate": 2.0037033341148915e-05, + "loss": 0.1847, + "step": 46180 + }, + { + "epoch": 3.2045659565609603, + "grad_norm": 1.0276708602905273, + "learning_rate": 2.0033659525520144e-05, + "loss": 0.2022, + "step": 46181 + }, + { + "epoch": 3.204635347998057, + "grad_norm": 0.7590569257736206, + "learning_rate": 2.0030285962335825e-05, + "loss": 0.3012, + "step": 46182 + }, + { + "epoch": 3.2047047394351535, + "grad_norm": 0.9623140096664429, + "learning_rate": 2.0026912651606623e-05, + "loss": 0.1821, + "step": 46183 + }, + { + "epoch": 3.2047741308722504, + "grad_norm": 0.7717501521110535, + "learning_rate": 2.0023539593343187e-05, + "loss": 0.2428, + "step": 46184 + }, + { + "epoch": 3.204843522309347, + "grad_norm": 6.6361894607543945, + "learning_rate": 2.0020166787556173e-05, + "loss": 0.4178, + "step": 46185 + }, + { + "epoch": 3.2049129137464436, + "grad_norm": 0.9832509756088257, + "learning_rate": 2.0016794234256186e-05, + "loss": 0.2009, + "step": 46186 + }, + { + "epoch": 3.2049823051835404, + "grad_norm": 0.7877191305160522, + "learning_rate": 2.00134219334539e-05, + "loss": 0.2465, + "step": 46187 + }, + { + "epoch": 3.205051696620637, + "grad_norm": 0.8623496294021606, + "learning_rate": 2.0010049885159978e-05, + "loss": 0.2118, + "step": 46188 + }, + { + "epoch": 3.2051210880577337, + "grad_norm": 0.8108837008476257, + "learning_rate": 2.000667808938502e-05, + "loss": 0.2601, + "step": 46189 + }, + { + "epoch": 3.2051904794948305, + "grad_norm": 0.8121871948242188, + "learning_rate": 2.000330654613972e-05, + "loss": 0.2203, + "step": 46190 + }, + { + "epoch": 3.205259870931927, + "grad_norm": 0.8488773107528687, + "learning_rate": 1.9999935255434688e-05, + "loss": 0.2268, + "step": 46191 + }, + { + "epoch": 3.2053292623690237, + "grad_norm": 0.6496952772140503, + "learning_rate": 1.999656421728061e-05, + "loss": 0.1994, + "step": 46192 + }, + { + "epoch": 3.20539865380612, + "grad_norm": 1.4715756177902222, + "learning_rate": 1.9993193431688073e-05, + "loss": 0.3317, + "step": 46193 + }, + { + "epoch": 3.205468045243217, + "grad_norm": 1.1130075454711914, + "learning_rate": 1.9989822898667743e-05, + "loss": 0.2578, + "step": 46194 + }, + { + "epoch": 3.2055374366803138, + "grad_norm": 1.0522027015686035, + "learning_rate": 1.9986452618230257e-05, + "loss": 0.3152, + "step": 46195 + }, + { + "epoch": 3.20560682811741, + "grad_norm": 0.8660951256752014, + "learning_rate": 1.998308259038627e-05, + "loss": 0.1719, + "step": 46196 + }, + { + "epoch": 3.205676219554507, + "grad_norm": 1.1583404541015625, + "learning_rate": 1.99797128151464e-05, + "loss": 0.2867, + "step": 46197 + }, + { + "epoch": 3.205745610991604, + "grad_norm": 0.9328457713127136, + "learning_rate": 1.997634329252128e-05, + "loss": 0.239, + "step": 46198 + }, + { + "epoch": 3.2058150024287, + "grad_norm": 1.037489652633667, + "learning_rate": 1.997297402252156e-05, + "loss": 0.2332, + "step": 46199 + }, + { + "epoch": 3.205884393865797, + "grad_norm": 1.3435428142547607, + "learning_rate": 1.996960500515791e-05, + "loss": 0.2037, + "step": 46200 + }, + { + "epoch": 3.2059537853028934, + "grad_norm": 1.1270920038223267, + "learning_rate": 1.9966236240440904e-05, + "loss": 0.3209, + "step": 46201 + }, + { + "epoch": 3.2060231767399903, + "grad_norm": 0.7049445509910583, + "learning_rate": 1.9962867728381207e-05, + "loss": 0.1816, + "step": 46202 + }, + { + "epoch": 3.206092568177087, + "grad_norm": 1.208896517753601, + "learning_rate": 1.995949946898944e-05, + "loss": 0.3673, + "step": 46203 + }, + { + "epoch": 3.2061619596141835, + "grad_norm": 0.7275880575180054, + "learning_rate": 1.9956131462276263e-05, + "loss": 0.2832, + "step": 46204 + }, + { + "epoch": 3.2062313510512803, + "grad_norm": 0.9154782295227051, + "learning_rate": 1.995276370825231e-05, + "loss": 0.2656, + "step": 46205 + }, + { + "epoch": 3.206300742488377, + "grad_norm": 1.075048565864563, + "learning_rate": 1.9949396206928162e-05, + "loss": 0.2447, + "step": 46206 + }, + { + "epoch": 3.2063701339254735, + "grad_norm": 0.7128656506538391, + "learning_rate": 1.9946028958314488e-05, + "loss": 0.2354, + "step": 46207 + }, + { + "epoch": 3.2064395253625704, + "grad_norm": 0.9234702587127686, + "learning_rate": 1.9942661962421916e-05, + "loss": 0.2828, + "step": 46208 + }, + { + "epoch": 3.2065089167996668, + "grad_norm": 0.6382076740264893, + "learning_rate": 1.993929521926107e-05, + "loss": 0.197, + "step": 46209 + }, + { + "epoch": 3.2065783082367636, + "grad_norm": 0.7709026336669922, + "learning_rate": 1.9935928728842567e-05, + "loss": 0.2014, + "step": 46210 + }, + { + "epoch": 3.2066476996738604, + "grad_norm": 1.0397388935089111, + "learning_rate": 1.9932562491177065e-05, + "loss": 0.2757, + "step": 46211 + }, + { + "epoch": 3.206717091110957, + "grad_norm": 0.88603276014328, + "learning_rate": 1.9929196506275195e-05, + "loss": 0.2724, + "step": 46212 + }, + { + "epoch": 3.2067864825480537, + "grad_norm": 0.8556057810783386, + "learning_rate": 1.9925830774147534e-05, + "loss": 0.4027, + "step": 46213 + }, + { + "epoch": 3.20685587398515, + "grad_norm": 0.6884132623672485, + "learning_rate": 1.9922465294804737e-05, + "loss": 0.1567, + "step": 46214 + }, + { + "epoch": 3.206925265422247, + "grad_norm": 1.1438994407653809, + "learning_rate": 1.9919100068257422e-05, + "loss": 0.323, + "step": 46215 + }, + { + "epoch": 3.2069946568593437, + "grad_norm": 0.8549600839614868, + "learning_rate": 1.9915735094516218e-05, + "loss": 0.1948, + "step": 46216 + }, + { + "epoch": 3.20706404829644, + "grad_norm": 1.5955238342285156, + "learning_rate": 1.991237037359175e-05, + "loss": 0.3442, + "step": 46217 + }, + { + "epoch": 3.207133439733537, + "grad_norm": 0.9787972569465637, + "learning_rate": 1.990900590549464e-05, + "loss": 0.228, + "step": 46218 + }, + { + "epoch": 3.2072028311706333, + "grad_norm": 1.1506078243255615, + "learning_rate": 1.990564169023553e-05, + "loss": 0.3676, + "step": 46219 + }, + { + "epoch": 3.20727222260773, + "grad_norm": 0.7387645840644836, + "learning_rate": 1.990227772782498e-05, + "loss": 0.1536, + "step": 46220 + }, + { + "epoch": 3.207341614044827, + "grad_norm": 1.2726775407791138, + "learning_rate": 1.9898914018273664e-05, + "loss": 0.2497, + "step": 46221 + }, + { + "epoch": 3.2074110054819234, + "grad_norm": 0.7555350661277771, + "learning_rate": 1.9895550561592168e-05, + "loss": 0.2301, + "step": 46222 + }, + { + "epoch": 3.20748039691902, + "grad_norm": 0.9668674468994141, + "learning_rate": 1.9892187357791135e-05, + "loss": 0.3048, + "step": 46223 + }, + { + "epoch": 3.207549788356117, + "grad_norm": 0.6831097602844238, + "learning_rate": 1.9888824406881167e-05, + "loss": 0.1696, + "step": 46224 + }, + { + "epoch": 3.2076191797932134, + "grad_norm": 0.7014611959457397, + "learning_rate": 1.988546170887289e-05, + "loss": 0.1394, + "step": 46225 + }, + { + "epoch": 3.2076885712303103, + "grad_norm": 0.9614906907081604, + "learning_rate": 1.9882099263776934e-05, + "loss": 0.2518, + "step": 46226 + }, + { + "epoch": 3.2077579626674066, + "grad_norm": 0.8629493117332458, + "learning_rate": 1.987873707160387e-05, + "loss": 0.2478, + "step": 46227 + }, + { + "epoch": 3.2078273541045035, + "grad_norm": 1.0347357988357544, + "learning_rate": 1.987537513236434e-05, + "loss": 0.3167, + "step": 46228 + }, + { + "epoch": 3.2078967455416003, + "grad_norm": 0.8203165531158447, + "learning_rate": 1.987201344606896e-05, + "loss": 0.21, + "step": 46229 + }, + { + "epoch": 3.2079661369786967, + "grad_norm": 0.9133678078651428, + "learning_rate": 1.9868652012728327e-05, + "loss": 0.235, + "step": 46230 + }, + { + "epoch": 3.2080355284157935, + "grad_norm": 1.6547795534133911, + "learning_rate": 1.98652908323531e-05, + "loss": 0.2633, + "step": 46231 + }, + { + "epoch": 3.2081049198528904, + "grad_norm": 0.9195645451545715, + "learning_rate": 1.9861929904953812e-05, + "loss": 0.2576, + "step": 46232 + }, + { + "epoch": 3.2081743112899868, + "grad_norm": 0.6101518273353577, + "learning_rate": 1.9858569230541123e-05, + "loss": 0.1687, + "step": 46233 + }, + { + "epoch": 3.2082437027270836, + "grad_norm": 0.7790742516517639, + "learning_rate": 1.9855208809125635e-05, + "loss": 0.228, + "step": 46234 + }, + { + "epoch": 3.20831309416418, + "grad_norm": 1.0272562503814697, + "learning_rate": 1.9851848640717974e-05, + "loss": 0.2842, + "step": 46235 + }, + { + "epoch": 3.208382485601277, + "grad_norm": 1.0196278095245361, + "learning_rate": 1.9848488725328708e-05, + "loss": 0.2598, + "step": 46236 + }, + { + "epoch": 3.2084518770383736, + "grad_norm": 1.3973588943481445, + "learning_rate": 1.984512906296846e-05, + "loss": 0.3856, + "step": 46237 + }, + { + "epoch": 3.20852126847547, + "grad_norm": 1.0649394989013672, + "learning_rate": 1.9841769653647836e-05, + "loss": 0.3528, + "step": 46238 + }, + { + "epoch": 3.208590659912567, + "grad_norm": 0.7699746489524841, + "learning_rate": 1.983841049737746e-05, + "loss": 0.2056, + "step": 46239 + }, + { + "epoch": 3.2086600513496633, + "grad_norm": 1.1763637065887451, + "learning_rate": 1.983505159416793e-05, + "loss": 0.3192, + "step": 46240 + }, + { + "epoch": 3.20872944278676, + "grad_norm": 0.9302101731300354, + "learning_rate": 1.9831692944029824e-05, + "loss": 0.2287, + "step": 46241 + }, + { + "epoch": 3.208798834223857, + "grad_norm": 1.2363497018814087, + "learning_rate": 1.982833454697376e-05, + "loss": 0.2339, + "step": 46242 + }, + { + "epoch": 3.2088682256609533, + "grad_norm": 0.6677768230438232, + "learning_rate": 1.9824976403010343e-05, + "loss": 0.2141, + "step": 46243 + }, + { + "epoch": 3.20893761709805, + "grad_norm": 0.6809736490249634, + "learning_rate": 1.9821618512150177e-05, + "loss": 0.2577, + "step": 46244 + }, + { + "epoch": 3.2090070085351465, + "grad_norm": 0.9784897565841675, + "learning_rate": 1.9818260874403848e-05, + "loss": 0.3071, + "step": 46245 + }, + { + "epoch": 3.2090763999722434, + "grad_norm": 0.9653673768043518, + "learning_rate": 1.9814903489781976e-05, + "loss": 0.2166, + "step": 46246 + }, + { + "epoch": 3.20914579140934, + "grad_norm": 0.9855946898460388, + "learning_rate": 1.9811546358295175e-05, + "loss": 0.282, + "step": 46247 + }, + { + "epoch": 3.2092151828464366, + "grad_norm": 0.9083052277565002, + "learning_rate": 1.9808189479953985e-05, + "loss": 0.2655, + "step": 46248 + }, + { + "epoch": 3.2092845742835334, + "grad_norm": 0.801935076713562, + "learning_rate": 1.980483285476905e-05, + "loss": 0.1747, + "step": 46249 + }, + { + "epoch": 3.2093539657206303, + "grad_norm": 1.1381932497024536, + "learning_rate": 1.9801476482750946e-05, + "loss": 0.2601, + "step": 46250 + }, + { + "epoch": 3.2094233571577266, + "grad_norm": 1.033769130706787, + "learning_rate": 1.9798120363910278e-05, + "loss": 0.3003, + "step": 46251 + }, + { + "epoch": 3.2094927485948235, + "grad_norm": 1.1555755138397217, + "learning_rate": 1.9794764498257635e-05, + "loss": 0.2744, + "step": 46252 + }, + { + "epoch": 3.20956214003192, + "grad_norm": 1.1440876722335815, + "learning_rate": 1.979140888580362e-05, + "loss": 0.345, + "step": 46253 + }, + { + "epoch": 3.2096315314690167, + "grad_norm": 1.0759546756744385, + "learning_rate": 1.9788053526558847e-05, + "loss": 0.2263, + "step": 46254 + }, + { + "epoch": 3.2097009229061135, + "grad_norm": 0.6129821538925171, + "learning_rate": 1.9784698420533852e-05, + "loss": 0.1656, + "step": 46255 + }, + { + "epoch": 3.20977031434321, + "grad_norm": 0.8749545812606812, + "learning_rate": 1.9781343567739263e-05, + "loss": 0.2663, + "step": 46256 + }, + { + "epoch": 3.2098397057803068, + "grad_norm": 0.7544443607330322, + "learning_rate": 1.9777988968185655e-05, + "loss": 0.1807, + "step": 46257 + }, + { + "epoch": 3.2099090972174036, + "grad_norm": 1.1523151397705078, + "learning_rate": 1.9774634621883636e-05, + "loss": 0.2587, + "step": 46258 + }, + { + "epoch": 3.2099784886545, + "grad_norm": 1.544006586074829, + "learning_rate": 1.9771280528843794e-05, + "loss": 0.2784, + "step": 46259 + }, + { + "epoch": 3.210047880091597, + "grad_norm": 0.7829627990722656, + "learning_rate": 1.97679266890767e-05, + "loss": 0.1612, + "step": 46260 + }, + { + "epoch": 3.210117271528693, + "grad_norm": 1.2149584293365479, + "learning_rate": 1.9764573102592986e-05, + "loss": 0.3119, + "step": 46261 + }, + { + "epoch": 3.21018666296579, + "grad_norm": 0.8844226002693176, + "learning_rate": 1.9761219769403172e-05, + "loss": 0.2126, + "step": 46262 + }, + { + "epoch": 3.210256054402887, + "grad_norm": 0.7797023057937622, + "learning_rate": 1.975786668951789e-05, + "loss": 0.1582, + "step": 46263 + }, + { + "epoch": 3.2103254458399832, + "grad_norm": 0.7280135154724121, + "learning_rate": 1.9754513862947698e-05, + "loss": 0.242, + "step": 46264 + }, + { + "epoch": 3.21039483727708, + "grad_norm": 0.850835382938385, + "learning_rate": 1.9751161289703203e-05, + "loss": 0.3227, + "step": 46265 + }, + { + "epoch": 3.2104642287141765, + "grad_norm": 0.9799710512161255, + "learning_rate": 1.974780896979501e-05, + "loss": 0.2626, + "step": 46266 + }, + { + "epoch": 3.2105336201512733, + "grad_norm": 0.6930533051490784, + "learning_rate": 1.9744456903233644e-05, + "loss": 0.2779, + "step": 46267 + }, + { + "epoch": 3.21060301158837, + "grad_norm": 1.127822756767273, + "learning_rate": 1.9741105090029723e-05, + "loss": 0.2192, + "step": 46268 + }, + { + "epoch": 3.2106724030254665, + "grad_norm": 1.0839046239852905, + "learning_rate": 1.9737753530193813e-05, + "loss": 0.2285, + "step": 46269 + }, + { + "epoch": 3.2107417944625634, + "grad_norm": 1.09272301197052, + "learning_rate": 1.9734402223736527e-05, + "loss": 0.3036, + "step": 46270 + }, + { + "epoch": 3.2108111858996597, + "grad_norm": 0.6782175898551941, + "learning_rate": 1.9731051170668402e-05, + "loss": 0.1666, + "step": 46271 + }, + { + "epoch": 3.2108805773367566, + "grad_norm": 0.8734118342399597, + "learning_rate": 1.9727700371000035e-05, + "loss": 0.2995, + "step": 46272 + }, + { + "epoch": 3.2109499687738534, + "grad_norm": 0.6346064209938049, + "learning_rate": 1.9724349824741995e-05, + "loss": 0.2624, + "step": 46273 + }, + { + "epoch": 3.21101936021095, + "grad_norm": 0.7948492169380188, + "learning_rate": 1.9720999531904904e-05, + "loss": 0.1547, + "step": 46274 + }, + { + "epoch": 3.2110887516480466, + "grad_norm": 1.153544545173645, + "learning_rate": 1.971764949249928e-05, + "loss": 0.3297, + "step": 46275 + }, + { + "epoch": 3.2111581430851435, + "grad_norm": 0.6545425057411194, + "learning_rate": 1.971429970653571e-05, + "loss": 0.1406, + "step": 46276 + }, + { + "epoch": 3.21122753452224, + "grad_norm": 1.3669921159744263, + "learning_rate": 1.971095017402479e-05, + "loss": 0.3274, + "step": 46277 + }, + { + "epoch": 3.2112969259593367, + "grad_norm": 1.0099194049835205, + "learning_rate": 1.9707600894977086e-05, + "loss": 0.3751, + "step": 46278 + }, + { + "epoch": 3.211366317396433, + "grad_norm": 0.9944263100624084, + "learning_rate": 1.970425186940317e-05, + "loss": 0.4543, + "step": 46279 + }, + { + "epoch": 3.21143570883353, + "grad_norm": 0.8778966665267944, + "learning_rate": 1.9700903097313604e-05, + "loss": 0.2649, + "step": 46280 + }, + { + "epoch": 3.2115051002706267, + "grad_norm": 0.8251224756240845, + "learning_rate": 1.9697554578719012e-05, + "loss": 0.2208, + "step": 46281 + }, + { + "epoch": 3.211574491707723, + "grad_norm": 0.8415720462799072, + "learning_rate": 1.969420631362989e-05, + "loss": 0.2031, + "step": 46282 + }, + { + "epoch": 3.21164388314482, + "grad_norm": 1.122143030166626, + "learning_rate": 1.9690858302056835e-05, + "loss": 0.2233, + "step": 46283 + }, + { + "epoch": 3.211713274581917, + "grad_norm": 1.2066130638122559, + "learning_rate": 1.9687510544010436e-05, + "loss": 0.2567, + "step": 46284 + }, + { + "epoch": 3.211782666019013, + "grad_norm": 1.005326509475708, + "learning_rate": 1.9684163039501235e-05, + "loss": 0.3126, + "step": 46285 + }, + { + "epoch": 3.21185205745611, + "grad_norm": 0.9722545146942139, + "learning_rate": 1.968081578853983e-05, + "loss": 0.4254, + "step": 46286 + }, + { + "epoch": 3.2119214488932064, + "grad_norm": 0.5410408973693848, + "learning_rate": 1.9677468791136754e-05, + "loss": 0.1301, + "step": 46287 + }, + { + "epoch": 3.2119908403303032, + "grad_norm": 1.0076799392700195, + "learning_rate": 1.9674122047302625e-05, + "loss": 0.264, + "step": 46288 + }, + { + "epoch": 3.2120602317674, + "grad_norm": 1.0283360481262207, + "learning_rate": 1.9670775557047948e-05, + "loss": 0.2476, + "step": 46289 + }, + { + "epoch": 3.2121296232044965, + "grad_norm": 1.0682326555252075, + "learning_rate": 1.96674293203833e-05, + "loss": 0.3478, + "step": 46290 + }, + { + "epoch": 3.2121990146415933, + "grad_norm": 0.9131230711936951, + "learning_rate": 1.966408333731927e-05, + "loss": 0.3698, + "step": 46291 + }, + { + "epoch": 3.2122684060786897, + "grad_norm": 0.8832476735115051, + "learning_rate": 1.966073760786641e-05, + "loss": 0.2831, + "step": 46292 + }, + { + "epoch": 3.2123377975157865, + "grad_norm": 0.6810268759727478, + "learning_rate": 1.965739213203527e-05, + "loss": 0.2119, + "step": 46293 + }, + { + "epoch": 3.2124071889528834, + "grad_norm": 1.1155098676681519, + "learning_rate": 1.9654046909836433e-05, + "loss": 0.3092, + "step": 46294 + }, + { + "epoch": 3.2124765803899797, + "grad_norm": 1.1795414686203003, + "learning_rate": 1.9650701941280458e-05, + "loss": 0.314, + "step": 46295 + }, + { + "epoch": 3.2125459718270766, + "grad_norm": 0.8392233848571777, + "learning_rate": 1.9647357226377883e-05, + "loss": 0.2019, + "step": 46296 + }, + { + "epoch": 3.2126153632641734, + "grad_norm": 0.5879892706871033, + "learning_rate": 1.9644012765139274e-05, + "loss": 0.164, + "step": 46297 + }, + { + "epoch": 3.21268475470127, + "grad_norm": 0.9652966260910034, + "learning_rate": 1.964066855757519e-05, + "loss": 0.3306, + "step": 46298 + }, + { + "epoch": 3.2127541461383666, + "grad_norm": 0.9803025126457214, + "learning_rate": 1.9637324603696206e-05, + "loss": 0.183, + "step": 46299 + }, + { + "epoch": 3.212823537575463, + "grad_norm": 0.7740188241004944, + "learning_rate": 1.963398090351285e-05, + "loss": 0.2632, + "step": 46300 + }, + { + "epoch": 3.21289292901256, + "grad_norm": 1.409867525100708, + "learning_rate": 1.9630637457035718e-05, + "loss": 0.2046, + "step": 46301 + }, + { + "epoch": 3.2129623204496567, + "grad_norm": 1.064802885055542, + "learning_rate": 1.962729426427532e-05, + "loss": 0.2278, + "step": 46302 + }, + { + "epoch": 3.213031711886753, + "grad_norm": 1.272623896598816, + "learning_rate": 1.962395132524223e-05, + "loss": 0.3386, + "step": 46303 + }, + { + "epoch": 3.21310110332385, + "grad_norm": 1.2288308143615723, + "learning_rate": 1.962060863994699e-05, + "loss": 0.2662, + "step": 46304 + }, + { + "epoch": 3.2131704947609463, + "grad_norm": 1.0481154918670654, + "learning_rate": 1.9617266208400198e-05, + "loss": 0.2493, + "step": 46305 + }, + { + "epoch": 3.213239886198043, + "grad_norm": 0.9374614953994751, + "learning_rate": 1.961392403061234e-05, + "loss": 0.2972, + "step": 46306 + }, + { + "epoch": 3.21330927763514, + "grad_norm": 1.1947134733200073, + "learning_rate": 1.9610582106594e-05, + "loss": 0.2655, + "step": 46307 + }, + { + "epoch": 3.2133786690722363, + "grad_norm": 0.94426429271698, + "learning_rate": 1.960724043635572e-05, + "loss": 0.2005, + "step": 46308 + }, + { + "epoch": 3.213448060509333, + "grad_norm": 0.6916705965995789, + "learning_rate": 1.960389901990809e-05, + "loss": 0.1541, + "step": 46309 + }, + { + "epoch": 3.21351745194643, + "grad_norm": 0.7764579057693481, + "learning_rate": 1.960055785726159e-05, + "loss": 0.1783, + "step": 46310 + }, + { + "epoch": 3.2135868433835264, + "grad_norm": 0.8266112804412842, + "learning_rate": 1.9597216948426813e-05, + "loss": 0.2935, + "step": 46311 + }, + { + "epoch": 3.2136562348206232, + "grad_norm": 1.090882420539856, + "learning_rate": 1.9593876293414283e-05, + "loss": 0.3022, + "step": 46312 + }, + { + "epoch": 3.2137256262577196, + "grad_norm": 0.6174705624580383, + "learning_rate": 1.9590535892234553e-05, + "loss": 0.1737, + "step": 46313 + }, + { + "epoch": 3.2137950176948165, + "grad_norm": 1.6246435642242432, + "learning_rate": 1.958719574489818e-05, + "loss": 0.3305, + "step": 46314 + }, + { + "epoch": 3.2138644091319133, + "grad_norm": 1.0481817722320557, + "learning_rate": 1.958385585141569e-05, + "loss": 0.3563, + "step": 46315 + }, + { + "epoch": 3.2139338005690097, + "grad_norm": 1.572105050086975, + "learning_rate": 1.9580516211797672e-05, + "loss": 0.3987, + "step": 46316 + }, + { + "epoch": 3.2140031920061065, + "grad_norm": 1.1140618324279785, + "learning_rate": 1.957717682605461e-05, + "loss": 0.3515, + "step": 46317 + }, + { + "epoch": 3.214072583443203, + "grad_norm": 0.9438315629959106, + "learning_rate": 1.957383769419706e-05, + "loss": 0.2975, + "step": 46318 + }, + { + "epoch": 3.2141419748802997, + "grad_norm": 1.0816102027893066, + "learning_rate": 1.9570498816235584e-05, + "loss": 0.2901, + "step": 46319 + }, + { + "epoch": 3.2142113663173966, + "grad_norm": 0.7901511192321777, + "learning_rate": 1.9567160192180702e-05, + "loss": 0.2776, + "step": 46320 + }, + { + "epoch": 3.214280757754493, + "grad_norm": 0.9042083024978638, + "learning_rate": 1.9563821822042973e-05, + "loss": 0.2104, + "step": 46321 + }, + { + "epoch": 3.21435014919159, + "grad_norm": 1.0200105905532837, + "learning_rate": 1.956048370583292e-05, + "loss": 0.1916, + "step": 46322 + }, + { + "epoch": 3.2144195406286866, + "grad_norm": 0.7565939426422119, + "learning_rate": 1.9557145843561108e-05, + "loss": 0.1728, + "step": 46323 + }, + { + "epoch": 3.214488932065783, + "grad_norm": 0.7934476137161255, + "learning_rate": 1.9553808235238035e-05, + "loss": 0.3218, + "step": 46324 + }, + { + "epoch": 3.21455832350288, + "grad_norm": 0.6190059185028076, + "learning_rate": 1.955047088087426e-05, + "loss": 0.1717, + "step": 46325 + }, + { + "epoch": 3.2146277149399762, + "grad_norm": 0.8791203498840332, + "learning_rate": 1.954713378048031e-05, + "loss": 0.2323, + "step": 46326 + }, + { + "epoch": 3.214697106377073, + "grad_norm": 0.9867808818817139, + "learning_rate": 1.9543796934066717e-05, + "loss": 0.2371, + "step": 46327 + }, + { + "epoch": 3.21476649781417, + "grad_norm": 0.9090983271598816, + "learning_rate": 1.954046034164403e-05, + "loss": 0.3434, + "step": 46328 + }, + { + "epoch": 3.2148358892512663, + "grad_norm": 1.1099443435668945, + "learning_rate": 1.9537124003222773e-05, + "loss": 0.283, + "step": 46329 + }, + { + "epoch": 3.214905280688363, + "grad_norm": 0.9554112553596497, + "learning_rate": 1.9533787918813518e-05, + "loss": 0.2614, + "step": 46330 + }, + { + "epoch": 3.21497467212546, + "grad_norm": 1.2828706502914429, + "learning_rate": 1.953045208842672e-05, + "loss": 0.3512, + "step": 46331 + }, + { + "epoch": 3.2150440635625563, + "grad_norm": 0.8027399182319641, + "learning_rate": 1.952711651207295e-05, + "loss": 0.1282, + "step": 46332 + }, + { + "epoch": 3.215113454999653, + "grad_norm": 0.80841463804245, + "learning_rate": 1.9523781189762735e-05, + "loss": 0.19, + "step": 46333 + }, + { + "epoch": 3.2151828464367496, + "grad_norm": 0.828005313873291, + "learning_rate": 1.9520446121506608e-05, + "loss": 0.2933, + "step": 46334 + }, + { + "epoch": 3.2152522378738464, + "grad_norm": 0.6454311609268188, + "learning_rate": 1.9517111307315093e-05, + "loss": 0.1593, + "step": 46335 + }, + { + "epoch": 3.2153216293109432, + "grad_norm": 0.9246098399162292, + "learning_rate": 1.9513776747198754e-05, + "loss": 0.3354, + "step": 46336 + }, + { + "epoch": 3.2153910207480396, + "grad_norm": 0.657105028629303, + "learning_rate": 1.951044244116804e-05, + "loss": 0.2206, + "step": 46337 + }, + { + "epoch": 3.2154604121851365, + "grad_norm": 0.7560743093490601, + "learning_rate": 1.950710838923354e-05, + "loss": 0.1632, + "step": 46338 + }, + { + "epoch": 3.215529803622233, + "grad_norm": 0.9011959433555603, + "learning_rate": 1.9503774591405776e-05, + "loss": 0.1584, + "step": 46339 + }, + { + "epoch": 3.2155991950593297, + "grad_norm": 1.0598211288452148, + "learning_rate": 1.950044104769523e-05, + "loss": 0.4004, + "step": 46340 + }, + { + "epoch": 3.2156685864964265, + "grad_norm": 0.7776734232902527, + "learning_rate": 1.9497107758112454e-05, + "loss": 0.1458, + "step": 46341 + }, + { + "epoch": 3.215737977933523, + "grad_norm": 0.8058629631996155, + "learning_rate": 1.9493774722667967e-05, + "loss": 0.16, + "step": 46342 + }, + { + "epoch": 3.2158073693706197, + "grad_norm": 0.6298394799232483, + "learning_rate": 1.9490441941372297e-05, + "loss": 0.1613, + "step": 46343 + }, + { + "epoch": 3.215876760807716, + "grad_norm": 0.7991411089897156, + "learning_rate": 1.948710941423598e-05, + "loss": 0.2045, + "step": 46344 + }, + { + "epoch": 3.215946152244813, + "grad_norm": 0.6026244759559631, + "learning_rate": 1.94837771412695e-05, + "loss": 0.1601, + "step": 46345 + }, + { + "epoch": 3.21601554368191, + "grad_norm": 1.060797095298767, + "learning_rate": 1.9480445122483394e-05, + "loss": 0.3709, + "step": 46346 + }, + { + "epoch": 3.216084935119006, + "grad_norm": 0.8827109932899475, + "learning_rate": 1.9477113357888177e-05, + "loss": 0.3864, + "step": 46347 + }, + { + "epoch": 3.216154326556103, + "grad_norm": 0.7577254176139832, + "learning_rate": 1.9473781847494365e-05, + "loss": 0.2034, + "step": 46348 + }, + { + "epoch": 3.2162237179932, + "grad_norm": 0.7715203762054443, + "learning_rate": 1.9470450591312494e-05, + "loss": 0.2262, + "step": 46349 + }, + { + "epoch": 3.2162931094302962, + "grad_norm": 1.2969272136688232, + "learning_rate": 1.9467119589353088e-05, + "loss": 0.2814, + "step": 46350 + }, + { + "epoch": 3.216362500867393, + "grad_norm": 0.7874451875686646, + "learning_rate": 1.9463788841626606e-05, + "loss": 0.2789, + "step": 46351 + }, + { + "epoch": 3.2164318923044894, + "grad_norm": 1.005743145942688, + "learning_rate": 1.9460458348143606e-05, + "loss": 0.3325, + "step": 46352 + }, + { + "epoch": 3.2165012837415863, + "grad_norm": 1.7521756887435913, + "learning_rate": 1.94571281089146e-05, + "loss": 0.3034, + "step": 46353 + }, + { + "epoch": 3.216570675178683, + "grad_norm": 0.7903364300727844, + "learning_rate": 1.9453798123950097e-05, + "loss": 0.2275, + "step": 46354 + }, + { + "epoch": 3.2166400666157795, + "grad_norm": 1.6751062870025635, + "learning_rate": 1.945046839326059e-05, + "loss": 0.3223, + "step": 46355 + }, + { + "epoch": 3.2167094580528763, + "grad_norm": 1.0952197313308716, + "learning_rate": 1.944713891685662e-05, + "loss": 0.218, + "step": 46356 + }, + { + "epoch": 3.216778849489973, + "grad_norm": 1.13743257522583, + "learning_rate": 1.9443809694748715e-05, + "loss": 0.2015, + "step": 46357 + }, + { + "epoch": 3.2168482409270696, + "grad_norm": 0.9234113693237305, + "learning_rate": 1.944048072694733e-05, + "loss": 0.2574, + "step": 46358 + }, + { + "epoch": 3.2169176323641664, + "grad_norm": 1.0886889696121216, + "learning_rate": 1.9437152013463e-05, + "loss": 0.2103, + "step": 46359 + }, + { + "epoch": 3.216987023801263, + "grad_norm": 0.8227772116661072, + "learning_rate": 1.943382355430624e-05, + "loss": 0.2808, + "step": 46360 + }, + { + "epoch": 3.2170564152383596, + "grad_norm": 0.7974900603294373, + "learning_rate": 1.943049534948754e-05, + "loss": 0.2108, + "step": 46361 + }, + { + "epoch": 3.2171258066754564, + "grad_norm": 0.9127123355865479, + "learning_rate": 1.942716739901743e-05, + "loss": 0.2551, + "step": 46362 + }, + { + "epoch": 3.217195198112553, + "grad_norm": 0.8125719428062439, + "learning_rate": 1.94238397029064e-05, + "loss": 0.2408, + "step": 46363 + }, + { + "epoch": 3.2172645895496497, + "grad_norm": 1.219803810119629, + "learning_rate": 1.9420512261164982e-05, + "loss": 0.347, + "step": 46364 + }, + { + "epoch": 3.217333980986746, + "grad_norm": 0.6979904770851135, + "learning_rate": 1.9417185073803634e-05, + "loss": 0.2469, + "step": 46365 + }, + { + "epoch": 3.217403372423843, + "grad_norm": 0.9766406416893005, + "learning_rate": 1.9413858140832884e-05, + "loss": 0.2004, + "step": 46366 + }, + { + "epoch": 3.2174727638609397, + "grad_norm": 1.224173903465271, + "learning_rate": 1.9410531462263225e-05, + "loss": 0.2598, + "step": 46367 + }, + { + "epoch": 3.217542155298036, + "grad_norm": 1.4765143394470215, + "learning_rate": 1.940720503810518e-05, + "loss": 0.5959, + "step": 46368 + }, + { + "epoch": 3.217611546735133, + "grad_norm": 0.7422542572021484, + "learning_rate": 1.9403878868369242e-05, + "loss": 0.3231, + "step": 46369 + }, + { + "epoch": 3.2176809381722293, + "grad_norm": 1.44054114818573, + "learning_rate": 1.9400552953065898e-05, + "loss": 0.208, + "step": 46370 + }, + { + "epoch": 3.217750329609326, + "grad_norm": 0.7048388123512268, + "learning_rate": 1.9397227292205677e-05, + "loss": 0.2484, + "step": 46371 + }, + { + "epoch": 3.217819721046423, + "grad_norm": 0.9747743606567383, + "learning_rate": 1.939390188579905e-05, + "loss": 0.2492, + "step": 46372 + }, + { + "epoch": 3.2178891124835194, + "grad_norm": 1.0405592918395996, + "learning_rate": 1.9390576733856502e-05, + "loss": 0.2388, + "step": 46373 + }, + { + "epoch": 3.217958503920616, + "grad_norm": 0.9973421096801758, + "learning_rate": 1.93872518363886e-05, + "loss": 0.2013, + "step": 46374 + }, + { + "epoch": 3.218027895357713, + "grad_norm": 1.3682079315185547, + "learning_rate": 1.938392719340575e-05, + "loss": 0.3213, + "step": 46375 + }, + { + "epoch": 3.2180972867948094, + "grad_norm": 1.1312639713287354, + "learning_rate": 1.938060280491849e-05, + "loss": 0.2473, + "step": 46376 + }, + { + "epoch": 3.2181666782319063, + "grad_norm": 1.212146282196045, + "learning_rate": 1.9377278670937316e-05, + "loss": 0.2573, + "step": 46377 + }, + { + "epoch": 3.2182360696690027, + "grad_norm": 1.2512127161026, + "learning_rate": 1.9373954791472738e-05, + "loss": 0.2783, + "step": 46378 + }, + { + "epoch": 3.2183054611060995, + "grad_norm": 0.7222139835357666, + "learning_rate": 1.9370631166535214e-05, + "loss": 0.1757, + "step": 46379 + }, + { + "epoch": 3.2183748525431963, + "grad_norm": 1.0003879070281982, + "learning_rate": 1.9367307796135248e-05, + "loss": 0.4411, + "step": 46380 + }, + { + "epoch": 3.2184442439802927, + "grad_norm": 0.6939907073974609, + "learning_rate": 1.936398468028333e-05, + "loss": 0.2208, + "step": 46381 + }, + { + "epoch": 3.2185136354173896, + "grad_norm": 0.9700140357017517, + "learning_rate": 1.9360661818989966e-05, + "loss": 0.3062, + "step": 46382 + }, + { + "epoch": 3.2185830268544864, + "grad_norm": 1.4063186645507812, + "learning_rate": 1.9357339212265635e-05, + "loss": 0.2435, + "step": 46383 + }, + { + "epoch": 3.2186524182915828, + "grad_norm": 1.051939845085144, + "learning_rate": 1.9354016860120828e-05, + "loss": 0.2745, + "step": 46384 + }, + { + "epoch": 3.2187218097286796, + "grad_norm": 0.8563088178634644, + "learning_rate": 1.9350694762566057e-05, + "loss": 0.2468, + "step": 46385 + }, + { + "epoch": 3.218791201165776, + "grad_norm": 0.6883593797683716, + "learning_rate": 1.9347372919611762e-05, + "loss": 0.2409, + "step": 46386 + }, + { + "epoch": 3.218860592602873, + "grad_norm": 0.6948387622833252, + "learning_rate": 1.9344051331268452e-05, + "loss": 0.2028, + "step": 46387 + }, + { + "epoch": 3.2189299840399697, + "grad_norm": 1.0693820714950562, + "learning_rate": 1.9340729997546615e-05, + "loss": 0.2195, + "step": 46388 + }, + { + "epoch": 3.218999375477066, + "grad_norm": 0.8234279155731201, + "learning_rate": 1.9337408918456735e-05, + "loss": 0.3155, + "step": 46389 + }, + { + "epoch": 3.219068766914163, + "grad_norm": 0.9352246522903442, + "learning_rate": 1.9334088094009296e-05, + "loss": 0.3282, + "step": 46390 + }, + { + "epoch": 3.2191381583512593, + "grad_norm": 0.9261341094970703, + "learning_rate": 1.9330767524214787e-05, + "loss": 0.3488, + "step": 46391 + }, + { + "epoch": 3.219207549788356, + "grad_norm": 0.7437145113945007, + "learning_rate": 1.9327447209083706e-05, + "loss": 0.2125, + "step": 46392 + }, + { + "epoch": 3.219276941225453, + "grad_norm": 1.3093702793121338, + "learning_rate": 1.9324127148626494e-05, + "loss": 0.2205, + "step": 46393 + }, + { + "epoch": 3.2193463326625493, + "grad_norm": 1.0962125062942505, + "learning_rate": 1.932080734285364e-05, + "loss": 0.3411, + "step": 46394 + }, + { + "epoch": 3.219415724099646, + "grad_norm": 1.0730106830596924, + "learning_rate": 1.9317487791775645e-05, + "loss": 0.4675, + "step": 46395 + }, + { + "epoch": 3.2194851155367425, + "grad_norm": 0.7389287948608398, + "learning_rate": 1.9314168495402986e-05, + "loss": 0.2628, + "step": 46396 + }, + { + "epoch": 3.2195545069738394, + "grad_norm": 1.081668734550476, + "learning_rate": 1.9310849453746126e-05, + "loss": 0.248, + "step": 46397 + }, + { + "epoch": 3.219623898410936, + "grad_norm": 0.548737108707428, + "learning_rate": 1.9307530666815554e-05, + "loss": 0.1782, + "step": 46398 + }, + { + "epoch": 3.2196932898480326, + "grad_norm": 0.8915917277336121, + "learning_rate": 1.930421213462177e-05, + "loss": 0.2379, + "step": 46399 + }, + { + "epoch": 3.2197626812851294, + "grad_norm": 1.6301473379135132, + "learning_rate": 1.9300893857175206e-05, + "loss": 0.3102, + "step": 46400 + }, + { + "epoch": 3.2198320727222263, + "grad_norm": 1.210556149482727, + "learning_rate": 1.9297575834486348e-05, + "loss": 0.4066, + "step": 46401 + }, + { + "epoch": 3.2199014641593227, + "grad_norm": 0.7036354541778564, + "learning_rate": 1.9294258066565695e-05, + "loss": 0.2294, + "step": 46402 + }, + { + "epoch": 3.2199708555964195, + "grad_norm": 1.062865138053894, + "learning_rate": 1.92909405534237e-05, + "loss": 0.3387, + "step": 46403 + }, + { + "epoch": 3.220040247033516, + "grad_norm": 1.1021679639816284, + "learning_rate": 1.9287623295070833e-05, + "loss": 0.3322, + "step": 46404 + }, + { + "epoch": 3.2201096384706127, + "grad_norm": 1.0725821256637573, + "learning_rate": 1.9284306291517608e-05, + "loss": 0.4217, + "step": 46405 + }, + { + "epoch": 3.2201790299077095, + "grad_norm": 0.9249046444892883, + "learning_rate": 1.928098954277443e-05, + "loss": 0.4327, + "step": 46406 + }, + { + "epoch": 3.220248421344806, + "grad_norm": 0.9900884032249451, + "learning_rate": 1.9277673048851808e-05, + "loss": 0.2474, + "step": 46407 + }, + { + "epoch": 3.2203178127819028, + "grad_norm": 1.1307529211044312, + "learning_rate": 1.9274356809760207e-05, + "loss": 0.1523, + "step": 46408 + }, + { + "epoch": 3.2203872042189996, + "grad_norm": 0.7912355661392212, + "learning_rate": 1.9271040825510123e-05, + "loss": 0.22, + "step": 46409 + }, + { + "epoch": 3.220456595656096, + "grad_norm": 1.2539421319961548, + "learning_rate": 1.926772509611198e-05, + "loss": 0.3364, + "step": 46410 + }, + { + "epoch": 3.220525987093193, + "grad_norm": 1.0583763122558594, + "learning_rate": 1.926440962157624e-05, + "loss": 0.2424, + "step": 46411 + }, + { + "epoch": 3.220595378530289, + "grad_norm": 1.288870096206665, + "learning_rate": 1.9261094401913404e-05, + "loss": 0.4302, + "step": 46412 + }, + { + "epoch": 3.220664769967386, + "grad_norm": 0.7385408282279968, + "learning_rate": 1.925777943713396e-05, + "loss": 0.2313, + "step": 46413 + }, + { + "epoch": 3.220734161404483, + "grad_norm": 1.0468101501464844, + "learning_rate": 1.92544647272483e-05, + "loss": 0.2817, + "step": 46414 + }, + { + "epoch": 3.2208035528415793, + "grad_norm": 0.8171976208686829, + "learning_rate": 1.9251150272266926e-05, + "loss": 0.279, + "step": 46415 + }, + { + "epoch": 3.220872944278676, + "grad_norm": 0.8311832547187805, + "learning_rate": 1.924783607220031e-05, + "loss": 0.1931, + "step": 46416 + }, + { + "epoch": 3.2209423357157725, + "grad_norm": 0.7605814933776855, + "learning_rate": 1.9244522127058905e-05, + "loss": 0.2116, + "step": 46417 + }, + { + "epoch": 3.2210117271528693, + "grad_norm": 0.8372212052345276, + "learning_rate": 1.9241208436853177e-05, + "loss": 0.2334, + "step": 46418 + }, + { + "epoch": 3.221081118589966, + "grad_norm": 0.9444245100021362, + "learning_rate": 1.9237895001593574e-05, + "loss": 0.3475, + "step": 46419 + }, + { + "epoch": 3.2211505100270625, + "grad_norm": 2.0285205841064453, + "learning_rate": 1.9234581821290598e-05, + "loss": 0.4021, + "step": 46420 + }, + { + "epoch": 3.2212199014641594, + "grad_norm": 1.0009006261825562, + "learning_rate": 1.9231268895954645e-05, + "loss": 0.2765, + "step": 46421 + }, + { + "epoch": 3.2212892929012558, + "grad_norm": 1.3051812648773193, + "learning_rate": 1.9227956225596223e-05, + "loss": 0.2584, + "step": 46422 + }, + { + "epoch": 3.2213586843383526, + "grad_norm": 1.2920215129852295, + "learning_rate": 1.922464381022575e-05, + "loss": 0.2918, + "step": 46423 + }, + { + "epoch": 3.2214280757754494, + "grad_norm": 0.8444361686706543, + "learning_rate": 1.9221331649853713e-05, + "loss": 0.2616, + "step": 46424 + }, + { + "epoch": 3.221497467212546, + "grad_norm": 0.7382649183273315, + "learning_rate": 1.9218019744490566e-05, + "loss": 0.1837, + "step": 46425 + }, + { + "epoch": 3.2215668586496427, + "grad_norm": 0.6054681539535522, + "learning_rate": 1.9214708094146783e-05, + "loss": 0.2063, + "step": 46426 + }, + { + "epoch": 3.2216362500867395, + "grad_norm": 1.0973918437957764, + "learning_rate": 1.9211396698832762e-05, + "loss": 0.2545, + "step": 46427 + }, + { + "epoch": 3.221705641523836, + "grad_norm": 1.9486680030822754, + "learning_rate": 1.9208085558558986e-05, + "loss": 0.2988, + "step": 46428 + }, + { + "epoch": 3.2217750329609327, + "grad_norm": 1.0031760931015015, + "learning_rate": 1.9204774673335924e-05, + "loss": 0.2782, + "step": 46429 + }, + { + "epoch": 3.221844424398029, + "grad_norm": 1.2108148336410522, + "learning_rate": 1.9201464043174002e-05, + "loss": 0.3094, + "step": 46430 + }, + { + "epoch": 3.221913815835126, + "grad_norm": 0.8891498446464539, + "learning_rate": 1.9198153668083685e-05, + "loss": 0.2511, + "step": 46431 + }, + { + "epoch": 3.2219832072722228, + "grad_norm": 0.777271568775177, + "learning_rate": 1.9194843548075425e-05, + "loss": 0.2145, + "step": 46432 + }, + { + "epoch": 3.222052598709319, + "grad_norm": 0.7363371849060059, + "learning_rate": 1.919153368315969e-05, + "loss": 0.2404, + "step": 46433 + }, + { + "epoch": 3.222121990146416, + "grad_norm": 1.2745959758758545, + "learning_rate": 1.9188224073346883e-05, + "loss": 0.3419, + "step": 46434 + }, + { + "epoch": 3.222191381583513, + "grad_norm": 1.10756254196167, + "learning_rate": 1.9184914718647484e-05, + "loss": 0.3712, + "step": 46435 + }, + { + "epoch": 3.222260773020609, + "grad_norm": 0.737549364566803, + "learning_rate": 1.918160561907193e-05, + "loss": 0.1999, + "step": 46436 + }, + { + "epoch": 3.222330164457706, + "grad_norm": 1.4245243072509766, + "learning_rate": 1.9178296774630666e-05, + "loss": 0.2765, + "step": 46437 + }, + { + "epoch": 3.2223995558948024, + "grad_norm": 1.2976031303405762, + "learning_rate": 1.9174988185334142e-05, + "loss": 0.2545, + "step": 46438 + }, + { + "epoch": 3.2224689473318993, + "grad_norm": 0.7541211247444153, + "learning_rate": 1.9171679851192815e-05, + "loss": 0.1926, + "step": 46439 + }, + { + "epoch": 3.222538338768996, + "grad_norm": 0.8304406404495239, + "learning_rate": 1.916837177221713e-05, + "loss": 0.1985, + "step": 46440 + }, + { + "epoch": 3.2226077302060925, + "grad_norm": 0.7986340522766113, + "learning_rate": 1.9165063948417495e-05, + "loss": 0.1627, + "step": 46441 + }, + { + "epoch": 3.2226771216431893, + "grad_norm": 0.6723962426185608, + "learning_rate": 1.9161756379804376e-05, + "loss": 0.2263, + "step": 46442 + }, + { + "epoch": 3.2227465130802857, + "grad_norm": 1.1297396421432495, + "learning_rate": 1.9158449066388206e-05, + "loss": 0.2586, + "step": 46443 + }, + { + "epoch": 3.2228159045173825, + "grad_norm": 1.2054188251495361, + "learning_rate": 1.9155142008179462e-05, + "loss": 0.1944, + "step": 46444 + }, + { + "epoch": 3.2228852959544794, + "grad_norm": 0.7948295474052429, + "learning_rate": 1.9151835205188528e-05, + "loss": 0.22, + "step": 46445 + }, + { + "epoch": 3.2229546873915758, + "grad_norm": 0.8530613780021667, + "learning_rate": 1.914852865742588e-05, + "loss": 0.3987, + "step": 46446 + }, + { + "epoch": 3.2230240788286726, + "grad_norm": 0.7932759523391724, + "learning_rate": 1.9145222364901937e-05, + "loss": 0.265, + "step": 46447 + }, + { + "epoch": 3.223093470265769, + "grad_norm": 0.8312280178070068, + "learning_rate": 1.9141916327627175e-05, + "loss": 0.3047, + "step": 46448 + }, + { + "epoch": 3.223162861702866, + "grad_norm": 0.8657221794128418, + "learning_rate": 1.9138610545611978e-05, + "loss": 0.2352, + "step": 46449 + }, + { + "epoch": 3.2232322531399626, + "grad_norm": 0.6412439942359924, + "learning_rate": 1.91353050188668e-05, + "loss": 0.139, + "step": 46450 + }, + { + "epoch": 3.223301644577059, + "grad_norm": 1.0368082523345947, + "learning_rate": 1.9131999747402086e-05, + "loss": 0.3353, + "step": 46451 + }, + { + "epoch": 3.223371036014156, + "grad_norm": 1.0978926420211792, + "learning_rate": 1.912869473122827e-05, + "loss": 0.3652, + "step": 46452 + }, + { + "epoch": 3.2234404274512527, + "grad_norm": 0.8796415328979492, + "learning_rate": 1.912538997035578e-05, + "loss": 0.2557, + "step": 46453 + }, + { + "epoch": 3.223509818888349, + "grad_norm": 0.9251955151557922, + "learning_rate": 1.912208546479507e-05, + "loss": 0.2759, + "step": 46454 + }, + { + "epoch": 3.223579210325446, + "grad_norm": 0.8770474195480347, + "learning_rate": 1.9118781214556538e-05, + "loss": 0.2648, + "step": 46455 + }, + { + "epoch": 3.2236486017625423, + "grad_norm": 0.724732518196106, + "learning_rate": 1.9115477219650613e-05, + "loss": 0.1612, + "step": 46456 + }, + { + "epoch": 3.223717993199639, + "grad_norm": 1.1687883138656616, + "learning_rate": 1.9112173480087757e-05, + "loss": 0.3441, + "step": 46457 + }, + { + "epoch": 3.223787384636736, + "grad_norm": 1.2085710763931274, + "learning_rate": 1.9108869995878376e-05, + "loss": 0.397, + "step": 46458 + }, + { + "epoch": 3.2238567760738324, + "grad_norm": 1.2347286939620972, + "learning_rate": 1.910556676703291e-05, + "loss": 0.4272, + "step": 46459 + }, + { + "epoch": 3.223926167510929, + "grad_norm": 0.7977100610733032, + "learning_rate": 1.910226379356179e-05, + "loss": 0.3023, + "step": 46460 + }, + { + "epoch": 3.223995558948026, + "grad_norm": 0.5657722353935242, + "learning_rate": 1.909896107547545e-05, + "loss": 0.1167, + "step": 46461 + }, + { + "epoch": 3.2240649503851224, + "grad_norm": 0.9601913690567017, + "learning_rate": 1.9095658612784294e-05, + "loss": 0.3553, + "step": 46462 + }, + { + "epoch": 3.2241343418222193, + "grad_norm": 1.1633687019348145, + "learning_rate": 1.9092356405498745e-05, + "loss": 0.1975, + "step": 46463 + }, + { + "epoch": 3.2242037332593156, + "grad_norm": 0.7451165318489075, + "learning_rate": 1.908905445362925e-05, + "loss": 0.2109, + "step": 46464 + }, + { + "epoch": 3.2242731246964125, + "grad_norm": 1.0281931161880493, + "learning_rate": 1.9085752757186216e-05, + "loss": 0.3073, + "step": 46465 + }, + { + "epoch": 3.2243425161335093, + "grad_norm": 0.7937695980072021, + "learning_rate": 1.908245131618007e-05, + "loss": 0.2235, + "step": 46466 + }, + { + "epoch": 3.2244119075706057, + "grad_norm": 1.1516131162643433, + "learning_rate": 1.9079150130621237e-05, + "loss": 0.2809, + "step": 46467 + }, + { + "epoch": 3.2244812990077025, + "grad_norm": 1.0816909074783325, + "learning_rate": 1.9075849200520168e-05, + "loss": 0.2126, + "step": 46468 + }, + { + "epoch": 3.224550690444799, + "grad_norm": 0.6657052636146545, + "learning_rate": 1.9072548525887234e-05, + "loss": 0.2223, + "step": 46469 + }, + { + "epoch": 3.2246200818818958, + "grad_norm": 0.8379685878753662, + "learning_rate": 1.906924810673286e-05, + "loss": 0.2059, + "step": 46470 + }, + { + "epoch": 3.2246894733189926, + "grad_norm": 0.8135282397270203, + "learning_rate": 1.9065947943067498e-05, + "loss": 0.2094, + "step": 46471 + }, + { + "epoch": 3.224758864756089, + "grad_norm": 0.6485392451286316, + "learning_rate": 1.9062648034901543e-05, + "loss": 0.2087, + "step": 46472 + }, + { + "epoch": 3.224828256193186, + "grad_norm": 0.9802784323692322, + "learning_rate": 1.9059348382245414e-05, + "loss": 0.3018, + "step": 46473 + }, + { + "epoch": 3.224897647630282, + "grad_norm": 0.7001740336418152, + "learning_rate": 1.905604898510952e-05, + "loss": 0.1364, + "step": 46474 + }, + { + "epoch": 3.224967039067379, + "grad_norm": 0.7843836545944214, + "learning_rate": 1.9052749843504335e-05, + "loss": 0.2304, + "step": 46475 + }, + { + "epoch": 3.225036430504476, + "grad_norm": 0.698114275932312, + "learning_rate": 1.9049450957440186e-05, + "loss": 0.2084, + "step": 46476 + }, + { + "epoch": 3.2251058219415722, + "grad_norm": 0.9168362617492676, + "learning_rate": 1.9046152326927534e-05, + "loss": 0.2997, + "step": 46477 + }, + { + "epoch": 3.225175213378669, + "grad_norm": 1.204344630241394, + "learning_rate": 1.904285395197679e-05, + "loss": 0.2321, + "step": 46478 + }, + { + "epoch": 3.225244604815766, + "grad_norm": 1.2294936180114746, + "learning_rate": 1.9039555832598388e-05, + "loss": 0.3343, + "step": 46479 + }, + { + "epoch": 3.2253139962528623, + "grad_norm": 0.9281690716743469, + "learning_rate": 1.9036257968802683e-05, + "loss": 0.2152, + "step": 46480 + }, + { + "epoch": 3.225383387689959, + "grad_norm": 0.9869213700294495, + "learning_rate": 1.9032960360600126e-05, + "loss": 0.3029, + "step": 46481 + }, + { + "epoch": 3.2254527791270555, + "grad_norm": 0.8384886384010315, + "learning_rate": 1.902966300800114e-05, + "loss": 0.2646, + "step": 46482 + }, + { + "epoch": 3.2255221705641524, + "grad_norm": 0.9947566986083984, + "learning_rate": 1.9026365911016087e-05, + "loss": 0.2429, + "step": 46483 + }, + { + "epoch": 3.225591562001249, + "grad_norm": 0.797476589679718, + "learning_rate": 1.9023069069655408e-05, + "loss": 0.2012, + "step": 46484 + }, + { + "epoch": 3.2256609534383456, + "grad_norm": 0.7837579250335693, + "learning_rate": 1.90197724839295e-05, + "loss": 0.3425, + "step": 46485 + }, + { + "epoch": 3.2257303448754424, + "grad_norm": 0.8895343542098999, + "learning_rate": 1.9016476153848774e-05, + "loss": 0.2345, + "step": 46486 + }, + { + "epoch": 3.2257997363125392, + "grad_norm": 1.1483744382858276, + "learning_rate": 1.901318007942363e-05, + "loss": 0.3257, + "step": 46487 + }, + { + "epoch": 3.2258691277496356, + "grad_norm": 1.0453118085861206, + "learning_rate": 1.900988426066449e-05, + "loss": 0.2439, + "step": 46488 + }, + { + "epoch": 3.2259385191867325, + "grad_norm": 1.158215880393982, + "learning_rate": 1.9006588697581786e-05, + "loss": 0.2239, + "step": 46489 + }, + { + "epoch": 3.226007910623829, + "grad_norm": 1.369652509689331, + "learning_rate": 1.9003293390185838e-05, + "loss": 0.3644, + "step": 46490 + }, + { + "epoch": 3.2260773020609257, + "grad_norm": 0.8694060444831848, + "learning_rate": 1.89999983384871e-05, + "loss": 0.1482, + "step": 46491 + }, + { + "epoch": 3.2261466934980225, + "grad_norm": 0.8338845372200012, + "learning_rate": 1.8996703542495975e-05, + "loss": 0.2006, + "step": 46492 + }, + { + "epoch": 3.226216084935119, + "grad_norm": 1.0650674104690552, + "learning_rate": 1.8993409002222863e-05, + "loss": 0.214, + "step": 46493 + }, + { + "epoch": 3.2262854763722157, + "grad_norm": 0.7741092443466187, + "learning_rate": 1.8990114717678153e-05, + "loss": 0.3365, + "step": 46494 + }, + { + "epoch": 3.226354867809312, + "grad_norm": 0.6547027826309204, + "learning_rate": 1.8986820688872252e-05, + "loss": 0.1244, + "step": 46495 + }, + { + "epoch": 3.226424259246409, + "grad_norm": 0.6921118497848511, + "learning_rate": 1.8983526915815587e-05, + "loss": 0.162, + "step": 46496 + }, + { + "epoch": 3.226493650683506, + "grad_norm": 0.8348175287246704, + "learning_rate": 1.89802333985185e-05, + "loss": 0.2402, + "step": 46497 + }, + { + "epoch": 3.226563042120602, + "grad_norm": 1.1565322875976562, + "learning_rate": 1.897694013699142e-05, + "loss": 0.3009, + "step": 46498 + }, + { + "epoch": 3.226632433557699, + "grad_norm": 1.1201900243759155, + "learning_rate": 1.8973647131244732e-05, + "loss": 0.2922, + "step": 46499 + }, + { + "epoch": 3.2267018249947954, + "grad_norm": 0.9415774345397949, + "learning_rate": 1.8970354381288847e-05, + "loss": 0.2362, + "step": 46500 + }, + { + "epoch": 3.2267712164318922, + "grad_norm": 1.0192711353302002, + "learning_rate": 1.896706188713414e-05, + "loss": 0.2497, + "step": 46501 + }, + { + "epoch": 3.226840607868989, + "grad_norm": 0.8575276136398315, + "learning_rate": 1.8963769648791064e-05, + "loss": 0.2399, + "step": 46502 + }, + { + "epoch": 3.2269099993060855, + "grad_norm": 0.7997991442680359, + "learning_rate": 1.8960477666269916e-05, + "loss": 0.2418, + "step": 46503 + }, + { + "epoch": 3.2269793907431823, + "grad_norm": 0.720028817653656, + "learning_rate": 1.8957185939581157e-05, + "loss": 0.1998, + "step": 46504 + }, + { + "epoch": 3.227048782180279, + "grad_norm": 0.9102209210395813, + "learning_rate": 1.8953894468735146e-05, + "loss": 0.1817, + "step": 46505 + }, + { + "epoch": 3.2271181736173755, + "grad_norm": 1.036014199256897, + "learning_rate": 1.8950603253742295e-05, + "loss": 0.3071, + "step": 46506 + }, + { + "epoch": 3.2271875650544724, + "grad_norm": 0.7727447748184204, + "learning_rate": 1.8947312294612984e-05, + "loss": 0.1625, + "step": 46507 + }, + { + "epoch": 3.2272569564915687, + "grad_norm": 0.7744892835617065, + "learning_rate": 1.8944021591357598e-05, + "loss": 0.2336, + "step": 46508 + }, + { + "epoch": 3.2273263479286656, + "grad_norm": 0.582267165184021, + "learning_rate": 1.8940731143986566e-05, + "loss": 0.1337, + "step": 46509 + }, + { + "epoch": 3.2273957393657624, + "grad_norm": 1.1416507959365845, + "learning_rate": 1.893744095251021e-05, + "loss": 0.2427, + "step": 46510 + }, + { + "epoch": 3.227465130802859, + "grad_norm": 1.084450364112854, + "learning_rate": 1.893415101693895e-05, + "loss": 0.2709, + "step": 46511 + }, + { + "epoch": 3.2275345222399556, + "grad_norm": 0.8138262629508972, + "learning_rate": 1.893086133728318e-05, + "loss": 0.2623, + "step": 46512 + }, + { + "epoch": 3.2276039136770525, + "grad_norm": 0.8065065145492554, + "learning_rate": 1.8927571913553256e-05, + "loss": 0.2329, + "step": 46513 + }, + { + "epoch": 3.227673305114149, + "grad_norm": 0.9058849215507507, + "learning_rate": 1.8924282745759613e-05, + "loss": 0.2191, + "step": 46514 + }, + { + "epoch": 3.2277426965512457, + "grad_norm": 0.697155773639679, + "learning_rate": 1.892099383391258e-05, + "loss": 0.2127, + "step": 46515 + }, + { + "epoch": 3.227812087988342, + "grad_norm": 0.843283474445343, + "learning_rate": 1.8917705178022572e-05, + "loss": 0.2616, + "step": 46516 + }, + { + "epoch": 3.227881479425439, + "grad_norm": 0.9434918761253357, + "learning_rate": 1.8914416778099976e-05, + "loss": 0.2657, + "step": 46517 + }, + { + "epoch": 3.2279508708625357, + "grad_norm": 0.7942313551902771, + "learning_rate": 1.891112863415513e-05, + "loss": 0.3126, + "step": 46518 + }, + { + "epoch": 3.228020262299632, + "grad_norm": 0.9485875368118286, + "learning_rate": 1.8907840746198447e-05, + "loss": 0.2295, + "step": 46519 + }, + { + "epoch": 3.228089653736729, + "grad_norm": 0.7804301977157593, + "learning_rate": 1.890455311424031e-05, + "loss": 0.0917, + "step": 46520 + }, + { + "epoch": 3.2281590451738253, + "grad_norm": 0.8886798620223999, + "learning_rate": 1.8901265738291084e-05, + "loss": 0.2722, + "step": 46521 + }, + { + "epoch": 3.228228436610922, + "grad_norm": 0.8795570731163025, + "learning_rate": 1.8897978618361144e-05, + "loss": 0.2658, + "step": 46522 + }, + { + "epoch": 3.228297828048019, + "grad_norm": 1.382565975189209, + "learning_rate": 1.8894691754460912e-05, + "loss": 0.2616, + "step": 46523 + }, + { + "epoch": 3.2283672194851154, + "grad_norm": 2.1148252487182617, + "learning_rate": 1.8891405146600706e-05, + "loss": 0.4351, + "step": 46524 + }, + { + "epoch": 3.2284366109222122, + "grad_norm": 1.0142130851745605, + "learning_rate": 1.888811879479091e-05, + "loss": 0.4149, + "step": 46525 + }, + { + "epoch": 3.228506002359309, + "grad_norm": 1.1009209156036377, + "learning_rate": 1.8884832699041922e-05, + "loss": 0.2249, + "step": 46526 + }, + { + "epoch": 3.2285753937964055, + "grad_norm": 0.6518292427062988, + "learning_rate": 1.888154685936411e-05, + "loss": 0.1948, + "step": 46527 + }, + { + "epoch": 3.2286447852335023, + "grad_norm": 0.7575855255126953, + "learning_rate": 1.8878261275767827e-05, + "loss": 0.2529, + "step": 46528 + }, + { + "epoch": 3.2287141766705987, + "grad_norm": 1.210094690322876, + "learning_rate": 1.8874975948263474e-05, + "loss": 0.1939, + "step": 46529 + }, + { + "epoch": 3.2287835681076955, + "grad_norm": 0.8994550704956055, + "learning_rate": 1.8871690876861436e-05, + "loss": 0.2559, + "step": 46530 + }, + { + "epoch": 3.2288529595447923, + "grad_norm": 0.822188675403595, + "learning_rate": 1.886840606157203e-05, + "loss": 0.2342, + "step": 46531 + }, + { + "epoch": 3.2289223509818887, + "grad_norm": 0.7849745750427246, + "learning_rate": 1.8865121502405647e-05, + "loss": 0.1976, + "step": 46532 + }, + { + "epoch": 3.2289917424189856, + "grad_norm": 1.1139602661132812, + "learning_rate": 1.8861837199372668e-05, + "loss": 0.3999, + "step": 46533 + }, + { + "epoch": 3.229061133856082, + "grad_norm": 0.7249305844306946, + "learning_rate": 1.8858553152483448e-05, + "loss": 0.1429, + "step": 46534 + }, + { + "epoch": 3.229130525293179, + "grad_norm": 0.8312787413597107, + "learning_rate": 1.8855269361748372e-05, + "loss": 0.2558, + "step": 46535 + }, + { + "epoch": 3.2291999167302756, + "grad_norm": 0.6932803988456726, + "learning_rate": 1.8851985827177788e-05, + "loss": 0.1655, + "step": 46536 + }, + { + "epoch": 3.229269308167372, + "grad_norm": 1.0421611070632935, + "learning_rate": 1.8848702548782106e-05, + "loss": 0.3146, + "step": 46537 + }, + { + "epoch": 3.229338699604469, + "grad_norm": 0.9090625047683716, + "learning_rate": 1.8845419526571618e-05, + "loss": 0.2458, + "step": 46538 + }, + { + "epoch": 3.2294080910415657, + "grad_norm": 0.7529069781303406, + "learning_rate": 1.8842136760556727e-05, + "loss": 0.2126, + "step": 46539 + }, + { + "epoch": 3.229477482478662, + "grad_norm": 0.9099327325820923, + "learning_rate": 1.8838854250747795e-05, + "loss": 0.2561, + "step": 46540 + }, + { + "epoch": 3.229546873915759, + "grad_norm": 1.1255046129226685, + "learning_rate": 1.883557199715519e-05, + "loss": 0.3548, + "step": 46541 + }, + { + "epoch": 3.2296162653528553, + "grad_norm": 0.6679121851921082, + "learning_rate": 1.8832289999789253e-05, + "loss": 0.1854, + "step": 46542 + }, + { + "epoch": 3.229685656789952, + "grad_norm": 0.7089213132858276, + "learning_rate": 1.8829008258660373e-05, + "loss": 0.235, + "step": 46543 + }, + { + "epoch": 3.229755048227049, + "grad_norm": 0.9667813777923584, + "learning_rate": 1.882572677377892e-05, + "loss": 0.2366, + "step": 46544 + }, + { + "epoch": 3.2298244396641453, + "grad_norm": 1.0432028770446777, + "learning_rate": 1.88224455451552e-05, + "loss": 0.1937, + "step": 46545 + }, + { + "epoch": 3.229893831101242, + "grad_norm": 0.9862748980522156, + "learning_rate": 1.8819164572799596e-05, + "loss": 0.2642, + "step": 46546 + }, + { + "epoch": 3.2299632225383386, + "grad_norm": 0.802182137966156, + "learning_rate": 1.8815883856722483e-05, + "loss": 0.1836, + "step": 46547 + }, + { + "epoch": 3.2300326139754354, + "grad_norm": 0.8401647806167603, + "learning_rate": 1.8812603396934227e-05, + "loss": 0.2095, + "step": 46548 + }, + { + "epoch": 3.2301020054125322, + "grad_norm": 0.9183486104011536, + "learning_rate": 1.880932319344514e-05, + "loss": 0.3312, + "step": 46549 + }, + { + "epoch": 3.2301713968496286, + "grad_norm": 2.256967544555664, + "learning_rate": 1.8806043246265594e-05, + "loss": 0.236, + "step": 46550 + }, + { + "epoch": 3.2302407882867255, + "grad_norm": 0.9261696338653564, + "learning_rate": 1.8802763555405945e-05, + "loss": 0.2224, + "step": 46551 + }, + { + "epoch": 3.2303101797238223, + "grad_norm": 0.9693259000778198, + "learning_rate": 1.8799484120876577e-05, + "loss": 0.2638, + "step": 46552 + }, + { + "epoch": 3.2303795711609187, + "grad_norm": 1.1101484298706055, + "learning_rate": 1.87962049426878e-05, + "loss": 0.2515, + "step": 46553 + }, + { + "epoch": 3.2304489625980155, + "grad_norm": 0.5861524939537048, + "learning_rate": 1.8792926020849976e-05, + "loss": 0.1248, + "step": 46554 + }, + { + "epoch": 3.230518354035112, + "grad_norm": 0.8208444714546204, + "learning_rate": 1.878964735537346e-05, + "loss": 0.2163, + "step": 46555 + }, + { + "epoch": 3.2305877454722087, + "grad_norm": 0.8856073021888733, + "learning_rate": 1.8786368946268616e-05, + "loss": 0.2707, + "step": 46556 + }, + { + "epoch": 3.2306571369093056, + "grad_norm": 1.006787896156311, + "learning_rate": 1.8783090793545777e-05, + "loss": 0.2952, + "step": 46557 + }, + { + "epoch": 3.230726528346402, + "grad_norm": 0.7181389331817627, + "learning_rate": 1.877981289721532e-05, + "loss": 0.2846, + "step": 46558 + }, + { + "epoch": 3.230795919783499, + "grad_norm": 1.2344691753387451, + "learning_rate": 1.8776535257287553e-05, + "loss": 0.4665, + "step": 46559 + }, + { + "epoch": 3.230865311220595, + "grad_norm": 1.3609691858291626, + "learning_rate": 1.8773257873772833e-05, + "loss": 0.2264, + "step": 46560 + }, + { + "epoch": 3.230934702657692, + "grad_norm": 1.0225993394851685, + "learning_rate": 1.8769980746681515e-05, + "loss": 0.3837, + "step": 46561 + }, + { + "epoch": 3.231004094094789, + "grad_norm": 1.052363634109497, + "learning_rate": 1.876670387602395e-05, + "loss": 0.2347, + "step": 46562 + }, + { + "epoch": 3.2310734855318852, + "grad_norm": 0.9390946626663208, + "learning_rate": 1.8763427261810462e-05, + "loss": 0.1755, + "step": 46563 + }, + { + "epoch": 3.231142876968982, + "grad_norm": 1.0551038980484009, + "learning_rate": 1.876015090405142e-05, + "loss": 0.2972, + "step": 46564 + }, + { + "epoch": 3.231212268406079, + "grad_norm": 0.730892539024353, + "learning_rate": 1.8756874802757185e-05, + "loss": 0.2299, + "step": 46565 + }, + { + "epoch": 3.2312816598431753, + "grad_norm": 0.8496353030204773, + "learning_rate": 1.8753598957938035e-05, + "loss": 0.2261, + "step": 46566 + }, + { + "epoch": 3.231351051280272, + "grad_norm": 0.8889583945274353, + "learning_rate": 1.875032336960435e-05, + "loss": 0.2566, + "step": 46567 + }, + { + "epoch": 3.2314204427173685, + "grad_norm": 0.9454810619354248, + "learning_rate": 1.8747048037766468e-05, + "loss": 0.1816, + "step": 46568 + }, + { + "epoch": 3.2314898341544653, + "grad_norm": 0.7644630074501038, + "learning_rate": 1.8743772962434724e-05, + "loss": 0.1929, + "step": 46569 + }, + { + "epoch": 3.231559225591562, + "grad_norm": 0.9218060374259949, + "learning_rate": 1.8740498143619467e-05, + "loss": 0.2796, + "step": 46570 + }, + { + "epoch": 3.2316286170286586, + "grad_norm": 0.8728572130203247, + "learning_rate": 1.873722358133103e-05, + "loss": 0.2094, + "step": 46571 + }, + { + "epoch": 3.2316980084657554, + "grad_norm": 0.7301061749458313, + "learning_rate": 1.8733949275579775e-05, + "loss": 0.1324, + "step": 46572 + }, + { + "epoch": 3.231767399902852, + "grad_norm": 0.9806867241859436, + "learning_rate": 1.873067522637598e-05, + "loss": 0.3085, + "step": 46573 + }, + { + "epoch": 3.2318367913399486, + "grad_norm": 1.1247074604034424, + "learning_rate": 1.872740143373003e-05, + "loss": 0.2917, + "step": 46574 + }, + { + "epoch": 3.2319061827770454, + "grad_norm": 0.8877482414245605, + "learning_rate": 1.8724127897652232e-05, + "loss": 0.2738, + "step": 46575 + }, + { + "epoch": 3.231975574214142, + "grad_norm": 1.126927375793457, + "learning_rate": 1.872085461815294e-05, + "loss": 0.1875, + "step": 46576 + }, + { + "epoch": 3.2320449656512387, + "grad_norm": 0.8368698358535767, + "learning_rate": 1.8717581595242473e-05, + "loss": 0.213, + "step": 46577 + }, + { + "epoch": 3.2321143570883355, + "grad_norm": 0.7139174342155457, + "learning_rate": 1.8714308828931192e-05, + "loss": 0.2089, + "step": 46578 + }, + { + "epoch": 3.232183748525432, + "grad_norm": 1.2445098161697388, + "learning_rate": 1.871103631922939e-05, + "loss": 0.2133, + "step": 46579 + }, + { + "epoch": 3.2322531399625287, + "grad_norm": 0.6619267463684082, + "learning_rate": 1.8707764066147414e-05, + "loss": 0.2253, + "step": 46580 + }, + { + "epoch": 3.232322531399625, + "grad_norm": 0.9886103868484497, + "learning_rate": 1.87044920696956e-05, + "loss": 0.1842, + "step": 46581 + }, + { + "epoch": 3.232391922836722, + "grad_norm": 0.9538880586624146, + "learning_rate": 1.870122032988426e-05, + "loss": 0.2355, + "step": 46582 + }, + { + "epoch": 3.2324613142738188, + "grad_norm": 0.8676025867462158, + "learning_rate": 1.8697948846723767e-05, + "loss": 0.1926, + "step": 46583 + }, + { + "epoch": 3.232530705710915, + "grad_norm": 0.8555080890655518, + "learning_rate": 1.8694677620224398e-05, + "loss": 0.2628, + "step": 46584 + }, + { + "epoch": 3.232600097148012, + "grad_norm": 0.9956382513046265, + "learning_rate": 1.8691406650396493e-05, + "loss": 0.2807, + "step": 46585 + }, + { + "epoch": 3.232669488585109, + "grad_norm": 0.7062440514564514, + "learning_rate": 1.8688135937250383e-05, + "loss": 0.1541, + "step": 46586 + }, + { + "epoch": 3.232738880022205, + "grad_norm": 1.017397403717041, + "learning_rate": 1.868486548079642e-05, + "loss": 0.2309, + "step": 46587 + }, + { + "epoch": 3.232808271459302, + "grad_norm": 0.9545493125915527, + "learning_rate": 1.8681595281044872e-05, + "loss": 0.2553, + "step": 46588 + }, + { + "epoch": 3.2328776628963984, + "grad_norm": 0.860287606716156, + "learning_rate": 1.8678325338006098e-05, + "loss": 0.2109, + "step": 46589 + }, + { + "epoch": 3.2329470543334953, + "grad_norm": 0.7371987104415894, + "learning_rate": 1.867505565169041e-05, + "loss": 0.2474, + "step": 46590 + }, + { + "epoch": 3.233016445770592, + "grad_norm": 0.7427359819412231, + "learning_rate": 1.8671786222108144e-05, + "loss": 0.1283, + "step": 46591 + }, + { + "epoch": 3.2330858372076885, + "grad_norm": 0.8388674855232239, + "learning_rate": 1.866851704926964e-05, + "loss": 0.2029, + "step": 46592 + }, + { + "epoch": 3.2331552286447853, + "grad_norm": 0.6889266967773438, + "learning_rate": 1.8665248133185153e-05, + "loss": 0.2473, + "step": 46593 + }, + { + "epoch": 3.2332246200818817, + "grad_norm": 0.6337286829948425, + "learning_rate": 1.8661979473865055e-05, + "loss": 0.1909, + "step": 46594 + }, + { + "epoch": 3.2332940115189786, + "grad_norm": 1.1139633655548096, + "learning_rate": 1.8658711071319647e-05, + "loss": 0.3279, + "step": 46595 + }, + { + "epoch": 3.2333634029560754, + "grad_norm": 1.0637651681900024, + "learning_rate": 1.865544292555924e-05, + "loss": 0.2921, + "step": 46596 + }, + { + "epoch": 3.2334327943931718, + "grad_norm": 1.0628607273101807, + "learning_rate": 1.8652175036594177e-05, + "loss": 0.1971, + "step": 46597 + }, + { + "epoch": 3.2335021858302686, + "grad_norm": 1.4531937837600708, + "learning_rate": 1.8648907404434755e-05, + "loss": 0.484, + "step": 46598 + }, + { + "epoch": 3.233571577267365, + "grad_norm": 0.6857951879501343, + "learning_rate": 1.8645640029091315e-05, + "loss": 0.1753, + "step": 46599 + }, + { + "epoch": 3.233640968704462, + "grad_norm": 0.781612753868103, + "learning_rate": 1.8642372910574123e-05, + "loss": 0.2445, + "step": 46600 + }, + { + "epoch": 3.2337103601415587, + "grad_norm": 0.8932279348373413, + "learning_rate": 1.863910604889353e-05, + "loss": 0.3455, + "step": 46601 + }, + { + "epoch": 3.233779751578655, + "grad_norm": 1.410231351852417, + "learning_rate": 1.8635839444059833e-05, + "loss": 0.3182, + "step": 46602 + }, + { + "epoch": 3.233849143015752, + "grad_norm": 0.88127201795578, + "learning_rate": 1.863257309608336e-05, + "loss": 0.2281, + "step": 46603 + }, + { + "epoch": 3.2339185344528487, + "grad_norm": 1.0227985382080078, + "learning_rate": 1.8629307004974405e-05, + "loss": 0.2621, + "step": 46604 + }, + { + "epoch": 3.233987925889945, + "grad_norm": 1.1470800638198853, + "learning_rate": 1.862604117074328e-05, + "loss": 0.3555, + "step": 46605 + }, + { + "epoch": 3.234057317327042, + "grad_norm": 0.900191605091095, + "learning_rate": 1.8622775593400345e-05, + "loss": 0.2596, + "step": 46606 + }, + { + "epoch": 3.2341267087641383, + "grad_norm": 1.3284119367599487, + "learning_rate": 1.8619510272955833e-05, + "loss": 0.4749, + "step": 46607 + }, + { + "epoch": 3.234196100201235, + "grad_norm": 1.153022050857544, + "learning_rate": 1.861624520942008e-05, + "loss": 0.2834, + "step": 46608 + }, + { + "epoch": 3.234265491638332, + "grad_norm": 0.9170002341270447, + "learning_rate": 1.8612980402803402e-05, + "loss": 0.1225, + "step": 46609 + }, + { + "epoch": 3.2343348830754284, + "grad_norm": 0.785864531993866, + "learning_rate": 1.8609715853116094e-05, + "loss": 0.2355, + "step": 46610 + }, + { + "epoch": 3.234404274512525, + "grad_norm": 0.9645792245864868, + "learning_rate": 1.8606451560368488e-05, + "loss": 0.2958, + "step": 46611 + }, + { + "epoch": 3.234473665949622, + "grad_norm": 1.0093101263046265, + "learning_rate": 1.8603187524570855e-05, + "loss": 0.3088, + "step": 46612 + }, + { + "epoch": 3.2345430573867184, + "grad_norm": 0.9939772486686707, + "learning_rate": 1.859992374573355e-05, + "loss": 0.3622, + "step": 46613 + }, + { + "epoch": 3.2346124488238153, + "grad_norm": 0.753909707069397, + "learning_rate": 1.859666022386681e-05, + "loss": 0.2933, + "step": 46614 + }, + { + "epoch": 3.2346818402609117, + "grad_norm": 0.8864548802375793, + "learning_rate": 1.8593396958980968e-05, + "loss": 0.2821, + "step": 46615 + }, + { + "epoch": 3.2347512316980085, + "grad_norm": 0.6449956297874451, + "learning_rate": 1.859013395108633e-05, + "loss": 0.1726, + "step": 46616 + }, + { + "epoch": 3.2348206231351053, + "grad_norm": 1.0527122020721436, + "learning_rate": 1.8586871200193202e-05, + "loss": 0.3862, + "step": 46617 + }, + { + "epoch": 3.2348900145722017, + "grad_norm": 1.264673113822937, + "learning_rate": 1.8583608706311896e-05, + "loss": 0.2847, + "step": 46618 + }, + { + "epoch": 3.2349594060092985, + "grad_norm": 1.0694706439971924, + "learning_rate": 1.8580346469452657e-05, + "loss": 0.2596, + "step": 46619 + }, + { + "epoch": 3.235028797446395, + "grad_norm": 0.8772523999214172, + "learning_rate": 1.8577084489625828e-05, + "loss": 0.3195, + "step": 46620 + }, + { + "epoch": 3.2350981888834918, + "grad_norm": 1.199515461921692, + "learning_rate": 1.8573822766841697e-05, + "loss": 0.3458, + "step": 46621 + }, + { + "epoch": 3.2351675803205886, + "grad_norm": 0.8275002241134644, + "learning_rate": 1.8570561301110578e-05, + "loss": 0.2746, + "step": 46622 + }, + { + "epoch": 3.235236971757685, + "grad_norm": 0.8152729868888855, + "learning_rate": 1.856730009244273e-05, + "loss": 0.2084, + "step": 46623 + }, + { + "epoch": 3.235306363194782, + "grad_norm": 1.0002598762512207, + "learning_rate": 1.8564039140848468e-05, + "loss": 0.2641, + "step": 46624 + }, + { + "epoch": 3.235375754631878, + "grad_norm": 0.8657529950141907, + "learning_rate": 1.856077844633808e-05, + "loss": 0.2145, + "step": 46625 + }, + { + "epoch": 3.235445146068975, + "grad_norm": 1.2016681432724, + "learning_rate": 1.855751800892187e-05, + "loss": 0.2791, + "step": 46626 + }, + { + "epoch": 3.235514537506072, + "grad_norm": 0.9361673593521118, + "learning_rate": 1.8554257828610155e-05, + "loss": 0.3337, + "step": 46627 + }, + { + "epoch": 3.2355839289431683, + "grad_norm": 1.0500396490097046, + "learning_rate": 1.855099790541317e-05, + "loss": 0.3709, + "step": 46628 + }, + { + "epoch": 3.235653320380265, + "grad_norm": 0.8687628507614136, + "learning_rate": 1.8547738239341227e-05, + "loss": 0.1822, + "step": 46629 + }, + { + "epoch": 3.235722711817362, + "grad_norm": 0.994985044002533, + "learning_rate": 1.854447883040463e-05, + "loss": 0.3756, + "step": 46630 + }, + { + "epoch": 3.2357921032544583, + "grad_norm": 0.9357550144195557, + "learning_rate": 1.8541219678613663e-05, + "loss": 0.3363, + "step": 46631 + }, + { + "epoch": 3.235861494691555, + "grad_norm": 1.1496661901474, + "learning_rate": 1.8537960783978624e-05, + "loss": 0.2016, + "step": 46632 + }, + { + "epoch": 3.2359308861286515, + "grad_norm": 0.793530285358429, + "learning_rate": 1.8534702146509774e-05, + "loss": 0.2711, + "step": 46633 + }, + { + "epoch": 3.2360002775657484, + "grad_norm": 1.0162286758422852, + "learning_rate": 1.853144376621746e-05, + "loss": 0.2141, + "step": 46634 + }, + { + "epoch": 3.236069669002845, + "grad_norm": 1.1844096183776855, + "learning_rate": 1.8528185643111895e-05, + "loss": 0.3728, + "step": 46635 + }, + { + "epoch": 3.2361390604399416, + "grad_norm": 1.011134147644043, + "learning_rate": 1.8524927777203395e-05, + "loss": 0.2472, + "step": 46636 + }, + { + "epoch": 3.2362084518770384, + "grad_norm": 1.44309401512146, + "learning_rate": 1.8521670168502237e-05, + "loss": 0.2166, + "step": 46637 + }, + { + "epoch": 3.2362778433141353, + "grad_norm": 0.7590434551239014, + "learning_rate": 1.8518412817018716e-05, + "loss": 0.2982, + "step": 46638 + }, + { + "epoch": 3.2363472347512316, + "grad_norm": 0.9249574542045593, + "learning_rate": 1.8515155722763123e-05, + "loss": 0.218, + "step": 46639 + }, + { + "epoch": 3.2364166261883285, + "grad_norm": 0.8950073719024658, + "learning_rate": 1.8511898885745717e-05, + "loss": 0.2305, + "step": 46640 + }, + { + "epoch": 3.236486017625425, + "grad_norm": 0.7736849784851074, + "learning_rate": 1.850864230597682e-05, + "loss": 0.199, + "step": 46641 + }, + { + "epoch": 3.2365554090625217, + "grad_norm": 0.8927138447761536, + "learning_rate": 1.850538598346666e-05, + "loss": 0.2745, + "step": 46642 + }, + { + "epoch": 3.2366248004996185, + "grad_norm": 0.7011379599571228, + "learning_rate": 1.850212991822555e-05, + "loss": 0.1916, + "step": 46643 + }, + { + "epoch": 3.236694191936715, + "grad_norm": 0.8113133311271667, + "learning_rate": 1.8498874110263753e-05, + "loss": 0.1919, + "step": 46644 + }, + { + "epoch": 3.2367635833738118, + "grad_norm": 0.6858553886413574, + "learning_rate": 1.8495618559591553e-05, + "loss": 0.1975, + "step": 46645 + }, + { + "epoch": 3.236832974810908, + "grad_norm": 1.0986236333847046, + "learning_rate": 1.8492363266219225e-05, + "loss": 0.2744, + "step": 46646 + }, + { + "epoch": 3.236902366248005, + "grad_norm": 1.1065902709960938, + "learning_rate": 1.848910823015706e-05, + "loss": 0.2714, + "step": 46647 + }, + { + "epoch": 3.236971757685102, + "grad_norm": 1.0273529291152954, + "learning_rate": 1.8485853451415347e-05, + "loss": 0.2424, + "step": 46648 + }, + { + "epoch": 3.237041149122198, + "grad_norm": 0.6454037427902222, + "learning_rate": 1.8482598930004303e-05, + "loss": 0.2509, + "step": 46649 + }, + { + "epoch": 3.237110540559295, + "grad_norm": 0.9055516719818115, + "learning_rate": 1.847934466593424e-05, + "loss": 0.2898, + "step": 46650 + }, + { + "epoch": 3.2371799319963914, + "grad_norm": 0.7494869828224182, + "learning_rate": 1.847609065921543e-05, + "loss": 0.2208, + "step": 46651 + }, + { + "epoch": 3.2372493234334883, + "grad_norm": 0.5642123222351074, + "learning_rate": 1.847283690985815e-05, + "loss": 0.1453, + "step": 46652 + }, + { + "epoch": 3.237318714870585, + "grad_norm": 0.896737813949585, + "learning_rate": 1.8469583417872683e-05, + "loss": 0.3205, + "step": 46653 + }, + { + "epoch": 3.2373881063076815, + "grad_norm": 0.9273011684417725, + "learning_rate": 1.8466330183269254e-05, + "loss": 0.293, + "step": 46654 + }, + { + "epoch": 3.2374574977447783, + "grad_norm": 0.6056384444236755, + "learning_rate": 1.8463077206058166e-05, + "loss": 0.1731, + "step": 46655 + }, + { + "epoch": 3.237526889181875, + "grad_norm": 1.0375875234603882, + "learning_rate": 1.8459824486249676e-05, + "loss": 0.2822, + "step": 46656 + }, + { + "epoch": 3.2375962806189715, + "grad_norm": 1.0132931470870972, + "learning_rate": 1.845657202385409e-05, + "loss": 0.3041, + "step": 46657 + }, + { + "epoch": 3.2376656720560684, + "grad_norm": 0.763523519039154, + "learning_rate": 1.845331981888162e-05, + "loss": 0.1998, + "step": 46658 + }, + { + "epoch": 3.2377350634931648, + "grad_norm": 0.6973297595977783, + "learning_rate": 1.8450067871342557e-05, + "loss": 0.2099, + "step": 46659 + }, + { + "epoch": 3.2378044549302616, + "grad_norm": 0.7924265265464783, + "learning_rate": 1.8446816181247173e-05, + "loss": 0.1696, + "step": 46660 + }, + { + "epoch": 3.2378738463673584, + "grad_norm": 0.9107685685157776, + "learning_rate": 1.8443564748605746e-05, + "loss": 0.3141, + "step": 46661 + }, + { + "epoch": 3.237943237804455, + "grad_norm": 0.6851638555526733, + "learning_rate": 1.8440313573428504e-05, + "loss": 0.2266, + "step": 46662 + }, + { + "epoch": 3.2380126292415516, + "grad_norm": 1.2985970973968506, + "learning_rate": 1.8437062655725724e-05, + "loss": 0.2717, + "step": 46663 + }, + { + "epoch": 3.2380820206786485, + "grad_norm": 0.9978873133659363, + "learning_rate": 1.843381199550768e-05, + "loss": 0.2208, + "step": 46664 + }, + { + "epoch": 3.238151412115745, + "grad_norm": 0.6530131101608276, + "learning_rate": 1.8430561592784623e-05, + "loss": 0.2018, + "step": 46665 + }, + { + "epoch": 3.2382208035528417, + "grad_norm": 0.8328512907028198, + "learning_rate": 1.842731144756682e-05, + "loss": 0.2255, + "step": 46666 + }, + { + "epoch": 3.238290194989938, + "grad_norm": 1.215813398361206, + "learning_rate": 1.8424061559864535e-05, + "loss": 0.4022, + "step": 46667 + }, + { + "epoch": 3.238359586427035, + "grad_norm": 0.9415367245674133, + "learning_rate": 1.8420811929688054e-05, + "loss": 0.331, + "step": 46668 + }, + { + "epoch": 3.2384289778641318, + "grad_norm": 0.6659110188484192, + "learning_rate": 1.8417562557047563e-05, + "loss": 0.1876, + "step": 46669 + }, + { + "epoch": 3.238498369301228, + "grad_norm": 1.064091444015503, + "learning_rate": 1.841431344195337e-05, + "loss": 0.2673, + "step": 46670 + }, + { + "epoch": 3.238567760738325, + "grad_norm": 0.9572625160217285, + "learning_rate": 1.8411064584415727e-05, + "loss": 0.2918, + "step": 46671 + }, + { + "epoch": 3.2386371521754214, + "grad_norm": 1.8239519596099854, + "learning_rate": 1.8407815984444878e-05, + "loss": 0.462, + "step": 46672 + }, + { + "epoch": 3.238706543612518, + "grad_norm": 0.8202567100524902, + "learning_rate": 1.8404567642051096e-05, + "loss": 0.2917, + "step": 46673 + }, + { + "epoch": 3.238775935049615, + "grad_norm": 0.7048453092575073, + "learning_rate": 1.8401319557244624e-05, + "loss": 0.1236, + "step": 46674 + }, + { + "epoch": 3.2388453264867114, + "grad_norm": 1.345017433166504, + "learning_rate": 1.839807173003575e-05, + "loss": 0.28, + "step": 46675 + }, + { + "epoch": 3.2389147179238083, + "grad_norm": 0.6783936023712158, + "learning_rate": 1.8394824160434655e-05, + "loss": 0.1552, + "step": 46676 + }, + { + "epoch": 3.2389841093609046, + "grad_norm": 0.6083570122718811, + "learning_rate": 1.8391576848451642e-05, + "loss": 0.1155, + "step": 46677 + }, + { + "epoch": 3.2390535007980015, + "grad_norm": 0.9032129049301147, + "learning_rate": 1.838832979409696e-05, + "loss": 0.2345, + "step": 46678 + }, + { + "epoch": 3.2391228922350983, + "grad_norm": 1.3031271696090698, + "learning_rate": 1.8385082997380843e-05, + "loss": 0.2857, + "step": 46679 + }, + { + "epoch": 3.2391922836721947, + "grad_norm": 1.0695641040802002, + "learning_rate": 1.8381836458313552e-05, + "loss": 0.3212, + "step": 46680 + }, + { + "epoch": 3.2392616751092915, + "grad_norm": 0.8947038650512695, + "learning_rate": 1.837859017690533e-05, + "loss": 0.3782, + "step": 46681 + }, + { + "epoch": 3.2393310665463884, + "grad_norm": 0.9162452220916748, + "learning_rate": 1.837534415316646e-05, + "loss": 0.2342, + "step": 46682 + }, + { + "epoch": 3.2394004579834847, + "grad_norm": 1.2473254203796387, + "learning_rate": 1.837209838710713e-05, + "loss": 0.3581, + "step": 46683 + }, + { + "epoch": 3.2394698494205816, + "grad_norm": 0.8317973017692566, + "learning_rate": 1.836885287873762e-05, + "loss": 0.2746, + "step": 46684 + }, + { + "epoch": 3.239539240857678, + "grad_norm": 0.8765377998352051, + "learning_rate": 1.836560762806816e-05, + "loss": 0.2667, + "step": 46685 + }, + { + "epoch": 3.239608632294775, + "grad_norm": 0.9889958500862122, + "learning_rate": 1.836236263510901e-05, + "loss": 0.3502, + "step": 46686 + }, + { + "epoch": 3.2396780237318716, + "grad_norm": 0.7803793549537659, + "learning_rate": 1.8359117899870414e-05, + "loss": 0.1885, + "step": 46687 + }, + { + "epoch": 3.239747415168968, + "grad_norm": 0.7670183181762695, + "learning_rate": 1.8355873422362636e-05, + "loss": 0.2284, + "step": 46688 + }, + { + "epoch": 3.239816806606065, + "grad_norm": 0.8158411979675293, + "learning_rate": 1.835262920259586e-05, + "loss": 0.2525, + "step": 46689 + }, + { + "epoch": 3.2398861980431617, + "grad_norm": 0.9115098714828491, + "learning_rate": 1.8349385240580362e-05, + "loss": 0.2255, + "step": 46690 + }, + { + "epoch": 3.239955589480258, + "grad_norm": 0.8818845748901367, + "learning_rate": 1.834614153632642e-05, + "loss": 0.2873, + "step": 46691 + }, + { + "epoch": 3.240024980917355, + "grad_norm": 0.7613835334777832, + "learning_rate": 1.8342898089844197e-05, + "loss": 0.1683, + "step": 46692 + }, + { + "epoch": 3.2400943723544513, + "grad_norm": 1.3141520023345947, + "learning_rate": 1.8339654901143978e-05, + "loss": 0.2423, + "step": 46693 + }, + { + "epoch": 3.240163763791548, + "grad_norm": 1.363938570022583, + "learning_rate": 1.833641197023599e-05, + "loss": 0.3354, + "step": 46694 + }, + { + "epoch": 3.240233155228645, + "grad_norm": 1.1727381944656372, + "learning_rate": 1.8333169297130472e-05, + "loss": 0.3203, + "step": 46695 + }, + { + "epoch": 3.2403025466657414, + "grad_norm": 0.8306686878204346, + "learning_rate": 1.832992688183769e-05, + "loss": 0.1826, + "step": 46696 + }, + { + "epoch": 3.240371938102838, + "grad_norm": 1.1300703287124634, + "learning_rate": 1.8326684724367836e-05, + "loss": 0.3365, + "step": 46697 + }, + { + "epoch": 3.2404413295399346, + "grad_norm": 1.0143978595733643, + "learning_rate": 1.8323442824731152e-05, + "loss": 0.2294, + "step": 46698 + }, + { + "epoch": 3.2405107209770314, + "grad_norm": 0.7975401282310486, + "learning_rate": 1.832020118293789e-05, + "loss": 0.2053, + "step": 46699 + }, + { + "epoch": 3.2405801124141282, + "grad_norm": 0.8928385972976685, + "learning_rate": 1.8316959798998267e-05, + "loss": 0.2098, + "step": 46700 + }, + { + "epoch": 3.2406495038512246, + "grad_norm": 2.6043899059295654, + "learning_rate": 1.831371867292253e-05, + "loss": 0.2356, + "step": 46701 + }, + { + "epoch": 3.2407188952883215, + "grad_norm": 1.1088197231292725, + "learning_rate": 1.8310477804720915e-05, + "loss": 0.1729, + "step": 46702 + }, + { + "epoch": 3.240788286725418, + "grad_norm": 0.836941659450531, + "learning_rate": 1.8307237194403658e-05, + "loss": 0.3836, + "step": 46703 + }, + { + "epoch": 3.2408576781625147, + "grad_norm": 0.9585374593734741, + "learning_rate": 1.8303996841980953e-05, + "loss": 0.3602, + "step": 46704 + }, + { + "epoch": 3.2409270695996115, + "grad_norm": 0.7580797672271729, + "learning_rate": 1.8300756747463054e-05, + "loss": 0.2077, + "step": 46705 + }, + { + "epoch": 3.240996461036708, + "grad_norm": 0.9208083152770996, + "learning_rate": 1.829751691086018e-05, + "loss": 0.27, + "step": 46706 + }, + { + "epoch": 3.2410658524738047, + "grad_norm": 1.1033234596252441, + "learning_rate": 1.8294277332182574e-05, + "loss": 0.3323, + "step": 46707 + }, + { + "epoch": 3.2411352439109016, + "grad_norm": 1.0037251710891724, + "learning_rate": 1.8291038011440453e-05, + "loss": 0.2872, + "step": 46708 + }, + { + "epoch": 3.241204635347998, + "grad_norm": 1.050063967704773, + "learning_rate": 1.8287798948644042e-05, + "loss": 0.3573, + "step": 46709 + }, + { + "epoch": 3.241274026785095, + "grad_norm": 0.7528858780860901, + "learning_rate": 1.8284560143803597e-05, + "loss": 0.2381, + "step": 46710 + }, + { + "epoch": 3.241343418222191, + "grad_norm": 1.0743714570999146, + "learning_rate": 1.828132159692929e-05, + "loss": 0.2862, + "step": 46711 + }, + { + "epoch": 3.241412809659288, + "grad_norm": 1.0612291097640991, + "learning_rate": 1.827808330803137e-05, + "loss": 0.1793, + "step": 46712 + }, + { + "epoch": 3.241482201096385, + "grad_norm": 0.9876545667648315, + "learning_rate": 1.8274845277120056e-05, + "loss": 0.2903, + "step": 46713 + }, + { + "epoch": 3.2415515925334812, + "grad_norm": 0.9618071913719177, + "learning_rate": 1.8271607504205566e-05, + "loss": 0.3141, + "step": 46714 + }, + { + "epoch": 3.241620983970578, + "grad_norm": 0.8830422759056091, + "learning_rate": 1.826836998929814e-05, + "loss": 0.1988, + "step": 46715 + }, + { + "epoch": 3.241690375407675, + "grad_norm": 0.8783818483352661, + "learning_rate": 1.8265132732407987e-05, + "loss": 0.1668, + "step": 46716 + }, + { + "epoch": 3.2417597668447713, + "grad_norm": 0.7847543954849243, + "learning_rate": 1.8261895733545354e-05, + "loss": 0.1936, + "step": 46717 + }, + { + "epoch": 3.241829158281868, + "grad_norm": 1.1028443574905396, + "learning_rate": 1.8258658992720402e-05, + "loss": 0.2573, + "step": 46718 + }, + { + "epoch": 3.2418985497189645, + "grad_norm": 0.8462181687355042, + "learning_rate": 1.8255422509943387e-05, + "loss": 0.2307, + "step": 46719 + }, + { + "epoch": 3.2419679411560614, + "grad_norm": 0.784273087978363, + "learning_rate": 1.8252186285224514e-05, + "loss": 0.2485, + "step": 46720 + }, + { + "epoch": 3.242037332593158, + "grad_norm": 1.1290502548217773, + "learning_rate": 1.8248950318574e-05, + "loss": 0.2627, + "step": 46721 + }, + { + "epoch": 3.2421067240302546, + "grad_norm": 1.0537689924240112, + "learning_rate": 1.8245714610002075e-05, + "loss": 0.3138, + "step": 46722 + }, + { + "epoch": 3.2421761154673514, + "grad_norm": 1.237688422203064, + "learning_rate": 1.8242479159518956e-05, + "loss": 0.34, + "step": 46723 + }, + { + "epoch": 3.242245506904448, + "grad_norm": 0.9974484443664551, + "learning_rate": 1.8239243967134834e-05, + "loss": 0.1603, + "step": 46724 + }, + { + "epoch": 3.2423148983415446, + "grad_norm": 1.1395620107650757, + "learning_rate": 1.8236009032859926e-05, + "loss": 0.3239, + "step": 46725 + }, + { + "epoch": 3.2423842897786415, + "grad_norm": 1.2136527299880981, + "learning_rate": 1.8232774356704475e-05, + "loss": 0.3865, + "step": 46726 + }, + { + "epoch": 3.242453681215738, + "grad_norm": 0.8712109327316284, + "learning_rate": 1.822953993867864e-05, + "loss": 0.2926, + "step": 46727 + }, + { + "epoch": 3.2425230726528347, + "grad_norm": 1.1511330604553223, + "learning_rate": 1.8226305778792662e-05, + "loss": 0.277, + "step": 46728 + }, + { + "epoch": 3.242592464089931, + "grad_norm": 1.2274847030639648, + "learning_rate": 1.822307187705675e-05, + "loss": 0.2706, + "step": 46729 + }, + { + "epoch": 3.242661855527028, + "grad_norm": 0.9045506715774536, + "learning_rate": 1.8219838233481144e-05, + "loss": 0.2346, + "step": 46730 + }, + { + "epoch": 3.2427312469641247, + "grad_norm": 1.0061225891113281, + "learning_rate": 1.8216604848075992e-05, + "loss": 0.3273, + "step": 46731 + }, + { + "epoch": 3.242800638401221, + "grad_norm": 0.8795945644378662, + "learning_rate": 1.8213371720851525e-05, + "loss": 0.2153, + "step": 46732 + }, + { + "epoch": 3.242870029838318, + "grad_norm": 0.9826155304908752, + "learning_rate": 1.821013885181796e-05, + "loss": 0.2816, + "step": 46733 + }, + { + "epoch": 3.242939421275415, + "grad_norm": 1.2791950702667236, + "learning_rate": 1.8206906240985498e-05, + "loss": 0.2861, + "step": 46734 + }, + { + "epoch": 3.243008812712511, + "grad_norm": 0.5249474048614502, + "learning_rate": 1.8203673888364336e-05, + "loss": 0.156, + "step": 46735 + }, + { + "epoch": 3.243078204149608, + "grad_norm": 0.6756646037101746, + "learning_rate": 1.8200441793964695e-05, + "loss": 0.2467, + "step": 46736 + }, + { + "epoch": 3.2431475955867044, + "grad_norm": 0.7624618411064148, + "learning_rate": 1.819720995779679e-05, + "loss": 0.2033, + "step": 46737 + }, + { + "epoch": 3.2432169870238012, + "grad_norm": 1.8004411458969116, + "learning_rate": 1.819397837987077e-05, + "loss": 0.2886, + "step": 46738 + }, + { + "epoch": 3.243286378460898, + "grad_norm": 0.7441430687904358, + "learning_rate": 1.8190747060196877e-05, + "loss": 0.1807, + "step": 46739 + }, + { + "epoch": 3.2433557698979945, + "grad_norm": 0.8068796396255493, + "learning_rate": 1.81875159987853e-05, + "loss": 0.2879, + "step": 46740 + }, + { + "epoch": 3.2434251613350913, + "grad_norm": 0.8276040554046631, + "learning_rate": 1.8184285195646244e-05, + "loss": 0.1319, + "step": 46741 + }, + { + "epoch": 3.243494552772188, + "grad_norm": 0.6836187243461609, + "learning_rate": 1.818105465078991e-05, + "loss": 0.1651, + "step": 46742 + }, + { + "epoch": 3.2435639442092845, + "grad_norm": 0.9530971646308899, + "learning_rate": 1.817782436422649e-05, + "loss": 0.4333, + "step": 46743 + }, + { + "epoch": 3.2436333356463813, + "grad_norm": 0.9132247567176819, + "learning_rate": 1.817459433596621e-05, + "loss": 0.2816, + "step": 46744 + }, + { + "epoch": 3.2437027270834777, + "grad_norm": 1.003934383392334, + "learning_rate": 1.817136456601922e-05, + "loss": 0.261, + "step": 46745 + }, + { + "epoch": 3.2437721185205746, + "grad_norm": 0.8955672383308411, + "learning_rate": 1.816813505439574e-05, + "loss": 0.2505, + "step": 46746 + }, + { + "epoch": 3.2438415099576714, + "grad_norm": 0.7229447364807129, + "learning_rate": 1.8164905801105957e-05, + "loss": 0.1678, + "step": 46747 + }, + { + "epoch": 3.243910901394768, + "grad_norm": 0.9873794913291931, + "learning_rate": 1.8161676806160077e-05, + "loss": 0.2867, + "step": 46748 + }, + { + "epoch": 3.2439802928318646, + "grad_norm": 0.8360777497291565, + "learning_rate": 1.815844806956829e-05, + "loss": 0.1873, + "step": 46749 + }, + { + "epoch": 3.244049684268961, + "grad_norm": 0.7931199073791504, + "learning_rate": 1.815521959134078e-05, + "loss": 0.1528, + "step": 46750 + }, + { + "epoch": 3.244119075706058, + "grad_norm": 0.8018729090690613, + "learning_rate": 1.8151991371487777e-05, + "loss": 0.1538, + "step": 46751 + }, + { + "epoch": 3.2441884671431547, + "grad_norm": 1.275181770324707, + "learning_rate": 1.814876341001941e-05, + "loss": 0.2568, + "step": 46752 + }, + { + "epoch": 3.244257858580251, + "grad_norm": 0.9287477731704712, + "learning_rate": 1.8145535706945905e-05, + "loss": 0.4513, + "step": 46753 + }, + { + "epoch": 3.244327250017348, + "grad_norm": 0.8527137041091919, + "learning_rate": 1.8142308262277453e-05, + "loss": 0.2744, + "step": 46754 + }, + { + "epoch": 3.2443966414544447, + "grad_norm": 1.4251785278320312, + "learning_rate": 1.813908107602422e-05, + "loss": 0.3271, + "step": 46755 + }, + { + "epoch": 3.244466032891541, + "grad_norm": 1.0411618947982788, + "learning_rate": 1.8135854148196428e-05, + "loss": 0.2603, + "step": 46756 + }, + { + "epoch": 3.244535424328638, + "grad_norm": 0.9175259470939636, + "learning_rate": 1.8132627478804232e-05, + "loss": 0.245, + "step": 46757 + }, + { + "epoch": 3.2446048157657343, + "grad_norm": 0.6227042078971863, + "learning_rate": 1.8129401067857865e-05, + "loss": 0.1003, + "step": 46758 + }, + { + "epoch": 3.244674207202831, + "grad_norm": 1.374266505241394, + "learning_rate": 1.8126174915367454e-05, + "loss": 0.3698, + "step": 46759 + }, + { + "epoch": 3.244743598639928, + "grad_norm": 2.182981252670288, + "learning_rate": 1.8122949021343217e-05, + "loss": 0.2144, + "step": 46760 + }, + { + "epoch": 3.2448129900770244, + "grad_norm": 0.7895352840423584, + "learning_rate": 1.811972338579535e-05, + "loss": 0.3024, + "step": 46761 + }, + { + "epoch": 3.2448823815141212, + "grad_norm": 1.4820773601531982, + "learning_rate": 1.8116498008733983e-05, + "loss": 0.3617, + "step": 46762 + }, + { + "epoch": 3.2449517729512176, + "grad_norm": 1.5630639791488647, + "learning_rate": 1.8113272890169342e-05, + "loss": 0.2986, + "step": 46763 + }, + { + "epoch": 3.2450211643883144, + "grad_norm": 1.5168286561965942, + "learning_rate": 1.8110048030111593e-05, + "loss": 0.2944, + "step": 46764 + }, + { + "epoch": 3.2450905558254113, + "grad_norm": 1.0643900632858276, + "learning_rate": 1.810682342857095e-05, + "loss": 0.2657, + "step": 46765 + }, + { + "epoch": 3.2451599472625077, + "grad_norm": 0.5530665516853333, + "learning_rate": 1.8103599085557532e-05, + "loss": 0.1722, + "step": 46766 + }, + { + "epoch": 3.2452293386996045, + "grad_norm": 1.2992172241210938, + "learning_rate": 1.810037500108156e-05, + "loss": 0.2105, + "step": 46767 + }, + { + "epoch": 3.2452987301367013, + "grad_norm": 0.7578660845756531, + "learning_rate": 1.8097151175153192e-05, + "loss": 0.2309, + "step": 46768 + }, + { + "epoch": 3.2453681215737977, + "grad_norm": 0.7375741004943848, + "learning_rate": 1.8093927607782623e-05, + "loss": 0.2385, + "step": 46769 + }, + { + "epoch": 3.2454375130108946, + "grad_norm": 0.9350765943527222, + "learning_rate": 1.8090704298980022e-05, + "loss": 0.3912, + "step": 46770 + }, + { + "epoch": 3.245506904447991, + "grad_norm": 0.9535643458366394, + "learning_rate": 1.8087481248755556e-05, + "loss": 0.2471, + "step": 46771 + }, + { + "epoch": 3.245576295885088, + "grad_norm": 1.975947618484497, + "learning_rate": 1.808425845711944e-05, + "loss": 0.2911, + "step": 46772 + }, + { + "epoch": 3.2456456873221846, + "grad_norm": 1.0569459199905396, + "learning_rate": 1.808103592408179e-05, + "loss": 0.3099, + "step": 46773 + }, + { + "epoch": 3.245715078759281, + "grad_norm": 0.7608860731124878, + "learning_rate": 1.8077813649652807e-05, + "loss": 0.1995, + "step": 46774 + }, + { + "epoch": 3.245784470196378, + "grad_norm": 1.1448289155960083, + "learning_rate": 1.8074591633842652e-05, + "loss": 0.2979, + "step": 46775 + }, + { + "epoch": 3.2458538616334742, + "grad_norm": 1.0390517711639404, + "learning_rate": 1.8071369876661516e-05, + "loss": 0.2859, + "step": 46776 + }, + { + "epoch": 3.245923253070571, + "grad_norm": 0.8891675472259521, + "learning_rate": 1.8068148378119555e-05, + "loss": 0.2445, + "step": 46777 + }, + { + "epoch": 3.245992644507668, + "grad_norm": 0.9239131212234497, + "learning_rate": 1.8064927138226952e-05, + "loss": 0.2159, + "step": 46778 + }, + { + "epoch": 3.2460620359447643, + "grad_norm": 1.1715843677520752, + "learning_rate": 1.8061706156993884e-05, + "loss": 0.3317, + "step": 46779 + }, + { + "epoch": 3.246131427381861, + "grad_norm": 0.8430559635162354, + "learning_rate": 1.8058485434430482e-05, + "loss": 0.2243, + "step": 46780 + }, + { + "epoch": 3.246200818818958, + "grad_norm": 0.6130515933036804, + "learning_rate": 1.8055264970546927e-05, + "loss": 0.2772, + "step": 46781 + }, + { + "epoch": 3.2462702102560543, + "grad_norm": 1.3595463037490845, + "learning_rate": 1.8052044765353404e-05, + "loss": 0.1835, + "step": 46782 + }, + { + "epoch": 3.246339601693151, + "grad_norm": 0.9120560884475708, + "learning_rate": 1.8048824818860053e-05, + "loss": 0.2858, + "step": 46783 + }, + { + "epoch": 3.2464089931302476, + "grad_norm": 0.8541526794433594, + "learning_rate": 1.804560513107707e-05, + "loss": 0.1995, + "step": 46784 + }, + { + "epoch": 3.2464783845673444, + "grad_norm": 0.8584619164466858, + "learning_rate": 1.8042385702014595e-05, + "loss": 0.2487, + "step": 46785 + }, + { + "epoch": 3.246547776004441, + "grad_norm": 0.7538508772850037, + "learning_rate": 1.8039166531682827e-05, + "loss": 0.1868, + "step": 46786 + }, + { + "epoch": 3.2466171674415376, + "grad_norm": 1.1200765371322632, + "learning_rate": 1.8035947620091875e-05, + "loss": 0.3307, + "step": 46787 + }, + { + "epoch": 3.2466865588786344, + "grad_norm": 0.7804496884346008, + "learning_rate": 1.803272896725193e-05, + "loss": 0.2253, + "step": 46788 + }, + { + "epoch": 3.246755950315731, + "grad_norm": 0.831535279750824, + "learning_rate": 1.8029510573173146e-05, + "loss": 0.1772, + "step": 46789 + }, + { + "epoch": 3.2468253417528277, + "grad_norm": 0.8405562043190002, + "learning_rate": 1.8026292437865688e-05, + "loss": 0.2502, + "step": 46790 + }, + { + "epoch": 3.2468947331899245, + "grad_norm": 0.8296696543693542, + "learning_rate": 1.8023074561339714e-05, + "loss": 0.1842, + "step": 46791 + }, + { + "epoch": 3.246964124627021, + "grad_norm": 0.9611924290657043, + "learning_rate": 1.8019856943605406e-05, + "loss": 0.2844, + "step": 46792 + }, + { + "epoch": 3.2470335160641177, + "grad_norm": 0.8358528017997742, + "learning_rate": 1.8016639584672877e-05, + "loss": 0.2805, + "step": 46793 + }, + { + "epoch": 3.2471029075012146, + "grad_norm": 0.7416982650756836, + "learning_rate": 1.8013422484552302e-05, + "loss": 0.2108, + "step": 46794 + }, + { + "epoch": 3.247172298938311, + "grad_norm": 1.188877820968628, + "learning_rate": 1.801020564325384e-05, + "loss": 0.2519, + "step": 46795 + }, + { + "epoch": 3.2472416903754078, + "grad_norm": 0.9418323040008545, + "learning_rate": 1.800698906078767e-05, + "loss": 0.2766, + "step": 46796 + }, + { + "epoch": 3.247311081812504, + "grad_norm": 0.8937512040138245, + "learning_rate": 1.8003772737163903e-05, + "loss": 0.2477, + "step": 46797 + }, + { + "epoch": 3.247380473249601, + "grad_norm": 0.8734515905380249, + "learning_rate": 1.800055667239271e-05, + "loss": 0.192, + "step": 46798 + }, + { + "epoch": 3.247449864686698, + "grad_norm": 0.5940255522727966, + "learning_rate": 1.7997340866484246e-05, + "loss": 0.1633, + "step": 46799 + }, + { + "epoch": 3.247519256123794, + "grad_norm": 1.0005874633789062, + "learning_rate": 1.7994125319448684e-05, + "loss": 0.2059, + "step": 46800 + }, + { + "epoch": 3.247588647560891, + "grad_norm": 0.9531801342964172, + "learning_rate": 1.7990910031296127e-05, + "loss": 0.2732, + "step": 46801 + }, + { + "epoch": 3.2476580389979874, + "grad_norm": 1.1341975927352905, + "learning_rate": 1.7987695002036766e-05, + "loss": 0.2915, + "step": 46802 + }, + { + "epoch": 3.2477274304350843, + "grad_norm": 0.6995089650154114, + "learning_rate": 1.7984480231680722e-05, + "loss": 0.3077, + "step": 46803 + }, + { + "epoch": 3.247796821872181, + "grad_norm": 1.2703547477722168, + "learning_rate": 1.798126572023816e-05, + "loss": 0.4251, + "step": 46804 + }, + { + "epoch": 3.2478662133092775, + "grad_norm": 0.8435859680175781, + "learning_rate": 1.7978051467719236e-05, + "loss": 0.2989, + "step": 46805 + }, + { + "epoch": 3.2479356047463743, + "grad_norm": 0.7234935760498047, + "learning_rate": 1.79748374741341e-05, + "loss": 0.1779, + "step": 46806 + }, + { + "epoch": 3.248004996183471, + "grad_norm": 0.9949169158935547, + "learning_rate": 1.797162373949287e-05, + "loss": 0.2522, + "step": 46807 + }, + { + "epoch": 3.2480743876205675, + "grad_norm": 0.7922956943511963, + "learning_rate": 1.79684102638057e-05, + "loss": 0.2946, + "step": 46808 + }, + { + "epoch": 3.2481437790576644, + "grad_norm": 1.1264328956604004, + "learning_rate": 1.796519704708275e-05, + "loss": 0.4496, + "step": 46809 + }, + { + "epoch": 3.2482131704947608, + "grad_norm": 0.9750795960426331, + "learning_rate": 1.796198408933415e-05, + "loss": 0.3653, + "step": 46810 + }, + { + "epoch": 3.2482825619318576, + "grad_norm": 0.9273973703384399, + "learning_rate": 1.795877139057004e-05, + "loss": 0.2465, + "step": 46811 + }, + { + "epoch": 3.2483519533689544, + "grad_norm": 0.9594783186912537, + "learning_rate": 1.795555895080058e-05, + "loss": 0.2179, + "step": 46812 + }, + { + "epoch": 3.248421344806051, + "grad_norm": 1.2259994745254517, + "learning_rate": 1.795234677003592e-05, + "loss": 0.2922, + "step": 46813 + }, + { + "epoch": 3.2484907362431477, + "grad_norm": 0.7538979649543762, + "learning_rate": 1.7949134848286163e-05, + "loss": 0.223, + "step": 46814 + }, + { + "epoch": 3.2485601276802445, + "grad_norm": 0.7871238589286804, + "learning_rate": 1.7945923185561476e-05, + "loss": 0.2588, + "step": 46815 + }, + { + "epoch": 3.248629519117341, + "grad_norm": 0.7269983291625977, + "learning_rate": 1.794271178187198e-05, + "loss": 0.275, + "step": 46816 + }, + { + "epoch": 3.2486989105544377, + "grad_norm": 0.7162107229232788, + "learning_rate": 1.7939500637227812e-05, + "loss": 0.241, + "step": 46817 + }, + { + "epoch": 3.248768301991534, + "grad_norm": 0.8040220737457275, + "learning_rate": 1.7936289751639134e-05, + "loss": 0.3363, + "step": 46818 + }, + { + "epoch": 3.248837693428631, + "grad_norm": 0.9560747146606445, + "learning_rate": 1.793307912511606e-05, + "loss": 0.1553, + "step": 46819 + }, + { + "epoch": 3.2489070848657278, + "grad_norm": 0.973702073097229, + "learning_rate": 1.7929868757668767e-05, + "loss": 0.1935, + "step": 46820 + }, + { + "epoch": 3.248976476302824, + "grad_norm": 1.745633840560913, + "learning_rate": 1.7926658649307326e-05, + "loss": 0.4177, + "step": 46821 + }, + { + "epoch": 3.249045867739921, + "grad_norm": 0.8389727473258972, + "learning_rate": 1.7923448800041897e-05, + "loss": 0.2281, + "step": 46822 + }, + { + "epoch": 3.2491152591770174, + "grad_norm": 0.8739764094352722, + "learning_rate": 1.7920239209882626e-05, + "loss": 0.3154, + "step": 46823 + }, + { + "epoch": 3.249184650614114, + "grad_norm": 0.641623318195343, + "learning_rate": 1.791702987883963e-05, + "loss": 0.1984, + "step": 46824 + }, + { + "epoch": 3.249254042051211, + "grad_norm": 1.098259687423706, + "learning_rate": 1.7913820806923053e-05, + "loss": 0.1702, + "step": 46825 + }, + { + "epoch": 3.2493234334883074, + "grad_norm": 0.9306393265724182, + "learning_rate": 1.7910611994143022e-05, + "loss": 0.2113, + "step": 46826 + }, + { + "epoch": 3.2493928249254043, + "grad_norm": 0.8386228084564209, + "learning_rate": 1.7907403440509685e-05, + "loss": 0.3164, + "step": 46827 + }, + { + "epoch": 3.2494622163625007, + "grad_norm": 0.9329797029495239, + "learning_rate": 1.7904195146033122e-05, + "loss": 0.2394, + "step": 46828 + }, + { + "epoch": 3.2495316077995975, + "grad_norm": 0.634053647518158, + "learning_rate": 1.79009871107235e-05, + "loss": 0.1767, + "step": 46829 + }, + { + "epoch": 3.2496009992366943, + "grad_norm": 1.1057521104812622, + "learning_rate": 1.7897779334590926e-05, + "loss": 0.233, + "step": 46830 + }, + { + "epoch": 3.2496703906737907, + "grad_norm": 0.7035872936248779, + "learning_rate": 1.7894571817645566e-05, + "loss": 0.2129, + "step": 46831 + }, + { + "epoch": 3.2497397821108875, + "grad_norm": 1.1571379899978638, + "learning_rate": 1.7891364559897493e-05, + "loss": 0.2066, + "step": 46832 + }, + { + "epoch": 3.2498091735479844, + "grad_norm": 1.0423388481140137, + "learning_rate": 1.7888157561356854e-05, + "loss": 0.2913, + "step": 46833 + }, + { + "epoch": 3.2498785649850808, + "grad_norm": 1.029770016670227, + "learning_rate": 1.788495082203381e-05, + "loss": 0.2519, + "step": 46834 + }, + { + "epoch": 3.2499479564221776, + "grad_norm": 0.9336835741996765, + "learning_rate": 1.788174434193841e-05, + "loss": 0.2544, + "step": 46835 + }, + { + "epoch": 3.250017347859274, + "grad_norm": 1.2263853549957275, + "learning_rate": 1.7878538121080824e-05, + "loss": 0.3297, + "step": 46836 + }, + { + "epoch": 3.250086739296371, + "grad_norm": 1.0754144191741943, + "learning_rate": 1.787533215947116e-05, + "loss": 0.3731, + "step": 46837 + }, + { + "epoch": 3.2501561307334677, + "grad_norm": 1.034642219543457, + "learning_rate": 1.7872126457119544e-05, + "loss": 0.3632, + "step": 46838 + }, + { + "epoch": 3.250225522170564, + "grad_norm": 0.9905682802200317, + "learning_rate": 1.78689210140361e-05, + "loss": 0.2832, + "step": 46839 + }, + { + "epoch": 3.250294913607661, + "grad_norm": 0.859036386013031, + "learning_rate": 1.7865715830230946e-05, + "loss": 0.3182, + "step": 46840 + }, + { + "epoch": 3.2503643050447577, + "grad_norm": 0.8496035933494568, + "learning_rate": 1.7862510905714215e-05, + "loss": 0.2368, + "step": 46841 + }, + { + "epoch": 3.250433696481854, + "grad_norm": 0.9862287640571594, + "learning_rate": 1.7859306240495988e-05, + "loss": 0.3388, + "step": 46842 + }, + { + "epoch": 3.250503087918951, + "grad_norm": 1.0981457233428955, + "learning_rate": 1.7856101834586404e-05, + "loss": 0.3849, + "step": 46843 + }, + { + "epoch": 3.2505724793560473, + "grad_norm": 1.4730079174041748, + "learning_rate": 1.7852897687995564e-05, + "loss": 0.2376, + "step": 46844 + }, + { + "epoch": 3.250641870793144, + "grad_norm": 1.3032444715499878, + "learning_rate": 1.78496938007336e-05, + "loss": 0.3153, + "step": 46845 + }, + { + "epoch": 3.250711262230241, + "grad_norm": 0.7434435486793518, + "learning_rate": 1.7846490172810627e-05, + "loss": 0.1939, + "step": 46846 + }, + { + "epoch": 3.2507806536673374, + "grad_norm": 0.7329076528549194, + "learning_rate": 1.7843286804236747e-05, + "loss": 0.2343, + "step": 46847 + }, + { + "epoch": 3.250850045104434, + "grad_norm": 0.8038451671600342, + "learning_rate": 1.7840083695022113e-05, + "loss": 0.2736, + "step": 46848 + }, + { + "epoch": 3.2509194365415306, + "grad_norm": 1.0246611833572388, + "learning_rate": 1.7836880845176773e-05, + "loss": 0.3978, + "step": 46849 + }, + { + "epoch": 3.2509888279786274, + "grad_norm": 0.8476282954216003, + "learning_rate": 1.7833678254710862e-05, + "loss": 0.2431, + "step": 46850 + }, + { + "epoch": 3.2510582194157243, + "grad_norm": 1.221427083015442, + "learning_rate": 1.7830475923634494e-05, + "loss": 0.2968, + "step": 46851 + }, + { + "epoch": 3.2511276108528206, + "grad_norm": 0.9642326235771179, + "learning_rate": 1.7827273851957794e-05, + "loss": 0.2183, + "step": 46852 + }, + { + "epoch": 3.2511970022899175, + "grad_norm": 1.2351306676864624, + "learning_rate": 1.7824072039690844e-05, + "loss": 0.2031, + "step": 46853 + }, + { + "epoch": 3.251266393727014, + "grad_norm": 0.8950639367103577, + "learning_rate": 1.7820870486843766e-05, + "loss": 0.2781, + "step": 46854 + }, + { + "epoch": 3.2513357851641107, + "grad_norm": 1.226081371307373, + "learning_rate": 1.7817669193426688e-05, + "loss": 0.3287, + "step": 46855 + }, + { + "epoch": 3.2514051766012075, + "grad_norm": 0.8071466088294983, + "learning_rate": 1.7814468159449683e-05, + "loss": 0.2764, + "step": 46856 + }, + { + "epoch": 3.251474568038304, + "grad_norm": 0.8109843134880066, + "learning_rate": 1.7811267384922847e-05, + "loss": 0.1466, + "step": 46857 + }, + { + "epoch": 3.2515439594754008, + "grad_norm": 0.7023177146911621, + "learning_rate": 1.780806686985632e-05, + "loss": 0.1242, + "step": 46858 + }, + { + "epoch": 3.2516133509124976, + "grad_norm": 0.7336330413818359, + "learning_rate": 1.7804866614260185e-05, + "loss": 0.1339, + "step": 46859 + }, + { + "epoch": 3.251682742349594, + "grad_norm": 0.9254457950592041, + "learning_rate": 1.7801666618144543e-05, + "loss": 0.2411, + "step": 46860 + }, + { + "epoch": 3.251752133786691, + "grad_norm": 1.1988489627838135, + "learning_rate": 1.7798466881519517e-05, + "loss": 0.2355, + "step": 46861 + }, + { + "epoch": 3.251821525223787, + "grad_norm": 0.9656444191932678, + "learning_rate": 1.7795267404395198e-05, + "loss": 0.2695, + "step": 46862 + }, + { + "epoch": 3.251890916660884, + "grad_norm": 1.0579540729522705, + "learning_rate": 1.7792068186781676e-05, + "loss": 0.3508, + "step": 46863 + }, + { + "epoch": 3.251960308097981, + "grad_norm": 1.8955931663513184, + "learning_rate": 1.7788869228689053e-05, + "loss": 0.1536, + "step": 46864 + }, + { + "epoch": 3.2520296995350773, + "grad_norm": 0.7389753460884094, + "learning_rate": 1.7785670530127428e-05, + "loss": 0.1763, + "step": 46865 + }, + { + "epoch": 3.252099090972174, + "grad_norm": 1.0022295713424683, + "learning_rate": 1.7782472091106926e-05, + "loss": 0.2081, + "step": 46866 + }, + { + "epoch": 3.252168482409271, + "grad_norm": 0.85509192943573, + "learning_rate": 1.7779273911637594e-05, + "loss": 0.153, + "step": 46867 + }, + { + "epoch": 3.2522378738463673, + "grad_norm": 0.7985396981239319, + "learning_rate": 1.777607599172957e-05, + "loss": 0.2082, + "step": 46868 + }, + { + "epoch": 3.252307265283464, + "grad_norm": 0.9674611687660217, + "learning_rate": 1.777287833139294e-05, + "loss": 0.2607, + "step": 46869 + }, + { + "epoch": 3.2523766567205605, + "grad_norm": 0.6432492733001709, + "learning_rate": 1.7769680930637778e-05, + "loss": 0.1891, + "step": 46870 + }, + { + "epoch": 3.2524460481576574, + "grad_norm": 0.9259761571884155, + "learning_rate": 1.7766483789474186e-05, + "loss": 0.3069, + "step": 46871 + }, + { + "epoch": 3.252515439594754, + "grad_norm": 0.6383349895477295, + "learning_rate": 1.776328690791227e-05, + "loss": 0.1744, + "step": 46872 + }, + { + "epoch": 3.2525848310318506, + "grad_norm": 1.2701078653335571, + "learning_rate": 1.7760090285962117e-05, + "loss": 0.1624, + "step": 46873 + }, + { + "epoch": 3.2526542224689474, + "grad_norm": 0.7223964929580688, + "learning_rate": 1.7756893923633823e-05, + "loss": 0.1677, + "step": 46874 + }, + { + "epoch": 3.252723613906044, + "grad_norm": 0.934008777141571, + "learning_rate": 1.7753697820937455e-05, + "loss": 0.2805, + "step": 46875 + }, + { + "epoch": 3.2527930053431406, + "grad_norm": 0.763984739780426, + "learning_rate": 1.775050197788316e-05, + "loss": 0.3027, + "step": 46876 + }, + { + "epoch": 3.2528623967802375, + "grad_norm": 0.8381831645965576, + "learning_rate": 1.7747306394480955e-05, + "loss": 0.266, + "step": 46877 + }, + { + "epoch": 3.252931788217334, + "grad_norm": 1.082934856414795, + "learning_rate": 1.7744111070740965e-05, + "loss": 0.2183, + "step": 46878 + }, + { + "epoch": 3.2530011796544307, + "grad_norm": 1.0131661891937256, + "learning_rate": 1.774091600667327e-05, + "loss": 0.2034, + "step": 46879 + }, + { + "epoch": 3.253070571091527, + "grad_norm": 1.0080128908157349, + "learning_rate": 1.773772120228796e-05, + "loss": 0.2446, + "step": 46880 + }, + { + "epoch": 3.253139962528624, + "grad_norm": 0.6784303188323975, + "learning_rate": 1.773452665759513e-05, + "loss": 0.2007, + "step": 46881 + }, + { + "epoch": 3.2532093539657208, + "grad_norm": 0.7487040162086487, + "learning_rate": 1.773133237260485e-05, + "loss": 0.1843, + "step": 46882 + }, + { + "epoch": 3.253278745402817, + "grad_norm": 0.7646598815917969, + "learning_rate": 1.7728138347327227e-05, + "loss": 0.2728, + "step": 46883 + }, + { + "epoch": 3.253348136839914, + "grad_norm": 0.8548696637153625, + "learning_rate": 1.7724944581772295e-05, + "loss": 0.2021, + "step": 46884 + }, + { + "epoch": 3.253417528277011, + "grad_norm": 1.0858373641967773, + "learning_rate": 1.7721751075950178e-05, + "loss": 0.2525, + "step": 46885 + }, + { + "epoch": 3.253486919714107, + "grad_norm": 1.4268449544906616, + "learning_rate": 1.771855782987095e-05, + "loss": 0.2671, + "step": 46886 + }, + { + "epoch": 3.253556311151204, + "grad_norm": 0.7065498232841492, + "learning_rate": 1.771536484354468e-05, + "loss": 0.2124, + "step": 46887 + }, + { + "epoch": 3.2536257025883004, + "grad_norm": 0.9460871815681458, + "learning_rate": 1.7712172116981464e-05, + "loss": 0.2223, + "step": 46888 + }, + { + "epoch": 3.2536950940253972, + "grad_norm": 1.2056390047073364, + "learning_rate": 1.7708979650191393e-05, + "loss": 0.3314, + "step": 46889 + }, + { + "epoch": 3.253764485462494, + "grad_norm": 0.6655226945877075, + "learning_rate": 1.7705787443184498e-05, + "loss": 0.1576, + "step": 46890 + }, + { + "epoch": 3.2538338768995905, + "grad_norm": 1.0626704692840576, + "learning_rate": 1.7702595495970888e-05, + "loss": 0.2708, + "step": 46891 + }, + { + "epoch": 3.2539032683366873, + "grad_norm": 0.7929635047912598, + "learning_rate": 1.7699403808560633e-05, + "loss": 0.3011, + "step": 46892 + }, + { + "epoch": 3.253972659773784, + "grad_norm": 1.011573076248169, + "learning_rate": 1.7696212380963816e-05, + "loss": 0.1997, + "step": 46893 + }, + { + "epoch": 3.2540420512108805, + "grad_norm": 0.8415670394897461, + "learning_rate": 1.7693021213190507e-05, + "loss": 0.2265, + "step": 46894 + }, + { + "epoch": 3.2541114426479774, + "grad_norm": 0.879314124584198, + "learning_rate": 1.7689830305250776e-05, + "loss": 0.2809, + "step": 46895 + }, + { + "epoch": 3.2541808340850737, + "grad_norm": 0.736526608467102, + "learning_rate": 1.7686639657154735e-05, + "loss": 0.2341, + "step": 46896 + }, + { + "epoch": 3.2542502255221706, + "grad_norm": 1.1464214324951172, + "learning_rate": 1.768344926891239e-05, + "loss": 0.3449, + "step": 46897 + }, + { + "epoch": 3.2543196169592674, + "grad_norm": 1.235001564025879, + "learning_rate": 1.768025914053385e-05, + "loss": 0.5613, + "step": 46898 + }, + { + "epoch": 3.254389008396364, + "grad_norm": 1.0419747829437256, + "learning_rate": 1.7677069272029177e-05, + "loss": 0.2718, + "step": 46899 + }, + { + "epoch": 3.2544583998334606, + "grad_norm": 0.9507400989532471, + "learning_rate": 1.7673879663408465e-05, + "loss": 0.2465, + "step": 46900 + }, + { + "epoch": 3.254527791270557, + "grad_norm": 0.6267780065536499, + "learning_rate": 1.767069031468175e-05, + "loss": 0.1941, + "step": 46901 + }, + { + "epoch": 3.254597182707654, + "grad_norm": 0.8550606966018677, + "learning_rate": 1.7667501225859105e-05, + "loss": 0.1939, + "step": 46902 + }, + { + "epoch": 3.2546665741447507, + "grad_norm": 0.9112508296966553, + "learning_rate": 1.7664312396950612e-05, + "loss": 0.2012, + "step": 46903 + }, + { + "epoch": 3.254735965581847, + "grad_norm": 0.9577624797821045, + "learning_rate": 1.7661123827966353e-05, + "loss": 0.2974, + "step": 46904 + }, + { + "epoch": 3.254805357018944, + "grad_norm": 1.265398383140564, + "learning_rate": 1.7657935518916356e-05, + "loss": 0.2333, + "step": 46905 + }, + { + "epoch": 3.2548747484560403, + "grad_norm": 1.2229187488555908, + "learning_rate": 1.7654747469810694e-05, + "loss": 0.3239, + "step": 46906 + }, + { + "epoch": 3.254944139893137, + "grad_norm": 0.5438734889030457, + "learning_rate": 1.7651559680659447e-05, + "loss": 0.1343, + "step": 46907 + }, + { + "epoch": 3.255013531330234, + "grad_norm": 0.7168398499488831, + "learning_rate": 1.7648372151472658e-05, + "loss": 0.1714, + "step": 46908 + }, + { + "epoch": 3.2550829227673304, + "grad_norm": 1.2711188793182373, + "learning_rate": 1.7645184882260425e-05, + "loss": 0.18, + "step": 46909 + }, + { + "epoch": 3.255152314204427, + "grad_norm": 0.9099565744400024, + "learning_rate": 1.76419978730328e-05, + "loss": 0.2742, + "step": 46910 + }, + { + "epoch": 3.255221705641524, + "grad_norm": 0.8668842315673828, + "learning_rate": 1.7638811123799803e-05, + "loss": 0.2112, + "step": 46911 + }, + { + "epoch": 3.2552910970786204, + "grad_norm": 0.8893439173698425, + "learning_rate": 1.7635624634571534e-05, + "loss": 0.2304, + "step": 46912 + }, + { + "epoch": 3.2553604885157172, + "grad_norm": 0.9816778302192688, + "learning_rate": 1.763243840535803e-05, + "loss": 0.2923, + "step": 46913 + }, + { + "epoch": 3.2554298799528136, + "grad_norm": 0.8040448427200317, + "learning_rate": 1.7629252436169373e-05, + "loss": 0.2181, + "step": 46914 + }, + { + "epoch": 3.2554992713899105, + "grad_norm": 1.0487858057022095, + "learning_rate": 1.762606672701561e-05, + "loss": 0.3611, + "step": 46915 + }, + { + "epoch": 3.2555686628270073, + "grad_norm": 0.7687195539474487, + "learning_rate": 1.762288127790679e-05, + "loss": 0.2202, + "step": 46916 + }, + { + "epoch": 3.2556380542641037, + "grad_norm": 1.5844981670379639, + "learning_rate": 1.761969608885301e-05, + "loss": 0.1598, + "step": 46917 + }, + { + "epoch": 3.2557074457012005, + "grad_norm": 0.7201143503189087, + "learning_rate": 1.761651115986426e-05, + "loss": 0.0978, + "step": 46918 + }, + { + "epoch": 3.2557768371382974, + "grad_norm": 1.1850175857543945, + "learning_rate": 1.761332649095062e-05, + "loss": 0.2863, + "step": 46919 + }, + { + "epoch": 3.2558462285753937, + "grad_norm": 0.9082477688789368, + "learning_rate": 1.761014208212216e-05, + "loss": 0.2539, + "step": 46920 + }, + { + "epoch": 3.2559156200124906, + "grad_norm": 0.6266504526138306, + "learning_rate": 1.7606957933388924e-05, + "loss": 0.1686, + "step": 46921 + }, + { + "epoch": 3.255985011449587, + "grad_norm": 1.2601984739303589, + "learning_rate": 1.7603774044760956e-05, + "loss": 0.2349, + "step": 46922 + }, + { + "epoch": 3.256054402886684, + "grad_norm": 0.7430381774902344, + "learning_rate": 1.760059041624831e-05, + "loss": 0.2404, + "step": 46923 + }, + { + "epoch": 3.2561237943237806, + "grad_norm": 0.9292841553688049, + "learning_rate": 1.759740704786107e-05, + "loss": 0.247, + "step": 46924 + }, + { + "epoch": 3.256193185760877, + "grad_norm": 1.2605109214782715, + "learning_rate": 1.7594223939609235e-05, + "loss": 0.2571, + "step": 46925 + }, + { + "epoch": 3.256262577197974, + "grad_norm": 1.2129409313201904, + "learning_rate": 1.759104109150287e-05, + "loss": 0.4014, + "step": 46926 + }, + { + "epoch": 3.2563319686350702, + "grad_norm": 0.9522193074226379, + "learning_rate": 1.7587858503552024e-05, + "loss": 0.2268, + "step": 46927 + }, + { + "epoch": 3.256401360072167, + "grad_norm": 0.7542626261711121, + "learning_rate": 1.7584676175766745e-05, + "loss": 0.2367, + "step": 46928 + }, + { + "epoch": 3.256470751509264, + "grad_norm": 0.9033963680267334, + "learning_rate": 1.7581494108157093e-05, + "loss": 0.4798, + "step": 46929 + }, + { + "epoch": 3.2565401429463603, + "grad_norm": 0.9084759950637817, + "learning_rate": 1.7578312300733092e-05, + "loss": 0.2326, + "step": 46930 + }, + { + "epoch": 3.256609534383457, + "grad_norm": 1.4231390953063965, + "learning_rate": 1.7575130753504832e-05, + "loss": 0.388, + "step": 46931 + }, + { + "epoch": 3.2566789258205535, + "grad_norm": 1.4162073135375977, + "learning_rate": 1.7571949466482285e-05, + "loss": 0.356, + "step": 46932 + }, + { + "epoch": 3.2567483172576503, + "grad_norm": 0.892022967338562, + "learning_rate": 1.756876843967553e-05, + "loss": 0.2987, + "step": 46933 + }, + { + "epoch": 3.256817708694747, + "grad_norm": 0.914274275302887, + "learning_rate": 1.7565587673094617e-05, + "loss": 0.328, + "step": 46934 + }, + { + "epoch": 3.2568871001318436, + "grad_norm": 1.0464956760406494, + "learning_rate": 1.7562407166749595e-05, + "loss": 0.2708, + "step": 46935 + }, + { + "epoch": 3.2569564915689404, + "grad_norm": 1.0938730239868164, + "learning_rate": 1.7559226920650472e-05, + "loss": 0.2087, + "step": 46936 + }, + { + "epoch": 3.2570258830060372, + "grad_norm": 0.783483624458313, + "learning_rate": 1.75560469348073e-05, + "loss": 0.2184, + "step": 46937 + }, + { + "epoch": 3.2570952744431336, + "grad_norm": 0.808458149433136, + "learning_rate": 1.7552867209230127e-05, + "loss": 0.2553, + "step": 46938 + }, + { + "epoch": 3.2571646658802305, + "grad_norm": 1.0100587606430054, + "learning_rate": 1.754968774392901e-05, + "loss": 0.2191, + "step": 46939 + }, + { + "epoch": 3.2572340573173273, + "grad_norm": 0.8254392743110657, + "learning_rate": 1.7546508538913943e-05, + "loss": 0.2494, + "step": 46940 + }, + { + "epoch": 3.2573034487544237, + "grad_norm": 0.9159327149391174, + "learning_rate": 1.754332959419497e-05, + "loss": 0.3121, + "step": 46941 + }, + { + "epoch": 3.2573728401915205, + "grad_norm": 0.7959943413734436, + "learning_rate": 1.754015090978215e-05, + "loss": 0.2519, + "step": 46942 + }, + { + "epoch": 3.257442231628617, + "grad_norm": 0.770836353302002, + "learning_rate": 1.75369724856855e-05, + "loss": 0.1323, + "step": 46943 + }, + { + "epoch": 3.2575116230657137, + "grad_norm": 1.1125245094299316, + "learning_rate": 1.7533794321915063e-05, + "loss": 0.3206, + "step": 46944 + }, + { + "epoch": 3.2575810145028106, + "grad_norm": 0.7692753076553345, + "learning_rate": 1.7530616418480895e-05, + "loss": 0.1983, + "step": 46945 + }, + { + "epoch": 3.257650405939907, + "grad_norm": 0.44686436653137207, + "learning_rate": 1.752743877539297e-05, + "loss": 0.138, + "step": 46946 + }, + { + "epoch": 3.257719797377004, + "grad_norm": 0.7283634543418884, + "learning_rate": 1.7524261392661358e-05, + "loss": 0.2508, + "step": 46947 + }, + { + "epoch": 3.2577891888141, + "grad_norm": 0.8919868469238281, + "learning_rate": 1.7521084270296084e-05, + "loss": 0.1713, + "step": 46948 + }, + { + "epoch": 3.257858580251197, + "grad_norm": 0.8633528351783752, + "learning_rate": 1.7517907408307177e-05, + "loss": 0.2611, + "step": 46949 + }, + { + "epoch": 3.257927971688294, + "grad_norm": 0.7448242902755737, + "learning_rate": 1.751473080670467e-05, + "loss": 0.1853, + "step": 46950 + }, + { + "epoch": 3.2579973631253902, + "grad_norm": 0.6699389219284058, + "learning_rate": 1.7511554465498582e-05, + "loss": 0.2071, + "step": 46951 + }, + { + "epoch": 3.258066754562487, + "grad_norm": 0.5836447477340698, + "learning_rate": 1.7508378384698976e-05, + "loss": 0.1196, + "step": 46952 + }, + { + "epoch": 3.2581361459995835, + "grad_norm": 1.0837914943695068, + "learning_rate": 1.7505202564315825e-05, + "loss": 0.3087, + "step": 46953 + }, + { + "epoch": 3.2582055374366803, + "grad_norm": 0.9434207081794739, + "learning_rate": 1.7502027004359177e-05, + "loss": 0.2668, + "step": 46954 + }, + { + "epoch": 3.258274928873777, + "grad_norm": 1.271134376525879, + "learning_rate": 1.749885170483905e-05, + "loss": 0.2136, + "step": 46955 + }, + { + "epoch": 3.2583443203108735, + "grad_norm": 0.7131756544113159, + "learning_rate": 1.749567666576548e-05, + "loss": 0.1547, + "step": 46956 + }, + { + "epoch": 3.2584137117479703, + "grad_norm": 0.9003409147262573, + "learning_rate": 1.7492501887148494e-05, + "loss": 0.269, + "step": 46957 + }, + { + "epoch": 3.2584831031850667, + "grad_norm": 1.3199230432510376, + "learning_rate": 1.74893273689981e-05, + "loss": 0.4064, + "step": 46958 + }, + { + "epoch": 3.2585524946221636, + "grad_norm": 0.8300343751907349, + "learning_rate": 1.7486153111324355e-05, + "loss": 0.2919, + "step": 46959 + }, + { + "epoch": 3.2586218860592604, + "grad_norm": 0.7898284196853638, + "learning_rate": 1.748297911413722e-05, + "loss": 0.1792, + "step": 46960 + }, + { + "epoch": 3.258691277496357, + "grad_norm": 0.7579343318939209, + "learning_rate": 1.7479805377446757e-05, + "loss": 0.2155, + "step": 46961 + }, + { + "epoch": 3.2587606689334536, + "grad_norm": 0.672256588935852, + "learning_rate": 1.7476631901262974e-05, + "loss": 0.3272, + "step": 46962 + }, + { + "epoch": 3.2588300603705505, + "grad_norm": 0.8512763381004333, + "learning_rate": 1.7473458685595877e-05, + "loss": 0.2379, + "step": 46963 + }, + { + "epoch": 3.258899451807647, + "grad_norm": 0.7790940403938293, + "learning_rate": 1.7470285730455505e-05, + "loss": 0.349, + "step": 46964 + }, + { + "epoch": 3.2589688432447437, + "grad_norm": 1.440972089767456, + "learning_rate": 1.746711303585189e-05, + "loss": 0.196, + "step": 46965 + }, + { + "epoch": 3.2590382346818405, + "grad_norm": 0.7049575448036194, + "learning_rate": 1.746394060179499e-05, + "loss": 0.1517, + "step": 46966 + }, + { + "epoch": 3.259107626118937, + "grad_norm": 0.9133327603340149, + "learning_rate": 1.746076842829487e-05, + "loss": 0.3268, + "step": 46967 + }, + { + "epoch": 3.2591770175560337, + "grad_norm": 0.8966413140296936, + "learning_rate": 1.745759651536152e-05, + "loss": 0.2805, + "step": 46968 + }, + { + "epoch": 3.25924640899313, + "grad_norm": 0.9883648157119751, + "learning_rate": 1.745442486300496e-05, + "loss": 0.2059, + "step": 46969 + }, + { + "epoch": 3.259315800430227, + "grad_norm": 1.260014533996582, + "learning_rate": 1.7451253471235228e-05, + "loss": 0.2817, + "step": 46970 + }, + { + "epoch": 3.259385191867324, + "grad_norm": 1.4291038513183594, + "learning_rate": 1.744808234006229e-05, + "loss": 0.3502, + "step": 46971 + }, + { + "epoch": 3.25945458330442, + "grad_norm": 1.184117317199707, + "learning_rate": 1.7444911469496173e-05, + "loss": 0.3608, + "step": 46972 + }, + { + "epoch": 3.259523974741517, + "grad_norm": 0.857018768787384, + "learning_rate": 1.7441740859546885e-05, + "loss": 0.2242, + "step": 46973 + }, + { + "epoch": 3.2595933661786134, + "grad_norm": 0.8638913631439209, + "learning_rate": 1.7438570510224483e-05, + "loss": 0.2614, + "step": 46974 + }, + { + "epoch": 3.2596627576157102, + "grad_norm": 0.8502447605133057, + "learning_rate": 1.7435400421538905e-05, + "loss": 0.2942, + "step": 46975 + }, + { + "epoch": 3.259732149052807, + "grad_norm": 0.9207910895347595, + "learning_rate": 1.7432230593500188e-05, + "loss": 0.2511, + "step": 46976 + }, + { + "epoch": 3.2598015404899034, + "grad_norm": 1.2201381921768188, + "learning_rate": 1.742906102611833e-05, + "loss": 0.416, + "step": 46977 + }, + { + "epoch": 3.2598709319270003, + "grad_norm": 1.0217480659484863, + "learning_rate": 1.7425891719403365e-05, + "loss": 0.2273, + "step": 46978 + }, + { + "epoch": 3.2599403233640967, + "grad_norm": 1.0059970617294312, + "learning_rate": 1.742272267336529e-05, + "loss": 0.2498, + "step": 46979 + }, + { + "epoch": 3.2600097148011935, + "grad_norm": 0.9477478861808777, + "learning_rate": 1.7419553888014085e-05, + "loss": 0.1979, + "step": 46980 + }, + { + "epoch": 3.2600791062382903, + "grad_norm": 0.8540311455726624, + "learning_rate": 1.7416385363359754e-05, + "loss": 0.2672, + "step": 46981 + }, + { + "epoch": 3.2601484976753867, + "grad_norm": 0.6851257085800171, + "learning_rate": 1.7413217099412328e-05, + "loss": 0.1914, + "step": 46982 + }, + { + "epoch": 3.2602178891124836, + "grad_norm": 0.899337112903595, + "learning_rate": 1.741004909618178e-05, + "loss": 0.4017, + "step": 46983 + }, + { + "epoch": 3.26028728054958, + "grad_norm": 0.7822021245956421, + "learning_rate": 1.740688135367813e-05, + "loss": 0.2624, + "step": 46984 + }, + { + "epoch": 3.260356671986677, + "grad_norm": 0.8710291981697083, + "learning_rate": 1.7403713871911375e-05, + "loss": 0.2053, + "step": 46985 + }, + { + "epoch": 3.2604260634237736, + "grad_norm": 0.8895184397697449, + "learning_rate": 1.7400546650891534e-05, + "loss": 0.2792, + "step": 46986 + }, + { + "epoch": 3.26049545486087, + "grad_norm": 0.9798687100410461, + "learning_rate": 1.7397379690628557e-05, + "loss": 0.2214, + "step": 46987 + }, + { + "epoch": 3.260564846297967, + "grad_norm": 0.8863018751144409, + "learning_rate": 1.7394212991132476e-05, + "loss": 0.3098, + "step": 46988 + }, + { + "epoch": 3.2606342377350637, + "grad_norm": 1.2774564027786255, + "learning_rate": 1.739104655241327e-05, + "loss": 0.3621, + "step": 46989 + }, + { + "epoch": 3.26070362917216, + "grad_norm": 0.9478096961975098, + "learning_rate": 1.738788037448096e-05, + "loss": 0.2354, + "step": 46990 + }, + { + "epoch": 3.260773020609257, + "grad_norm": 1.0106457471847534, + "learning_rate": 1.7384714457345517e-05, + "loss": 0.3398, + "step": 46991 + }, + { + "epoch": 3.2608424120463537, + "grad_norm": 0.8613165616989136, + "learning_rate": 1.7381548801016945e-05, + "loss": 0.2047, + "step": 46992 + }, + { + "epoch": 3.26091180348345, + "grad_norm": 0.6772844195365906, + "learning_rate": 1.7378383405505273e-05, + "loss": 0.1882, + "step": 46993 + }, + { + "epoch": 3.260981194920547, + "grad_norm": 0.8119269609451294, + "learning_rate": 1.7375218270820426e-05, + "loss": 0.3352, + "step": 46994 + }, + { + "epoch": 3.2610505863576433, + "grad_norm": 0.9788399934768677, + "learning_rate": 1.737205339697242e-05, + "loss": 0.1784, + "step": 46995 + }, + { + "epoch": 3.26111997779474, + "grad_norm": 0.831630289554596, + "learning_rate": 1.7368888783971273e-05, + "loss": 0.2716, + "step": 46996 + }, + { + "epoch": 3.261189369231837, + "grad_norm": 0.9833624958992004, + "learning_rate": 1.7365724431826945e-05, + "loss": 0.2743, + "step": 46997 + }, + { + "epoch": 3.2612587606689334, + "grad_norm": 1.6295613050460815, + "learning_rate": 1.736256034054944e-05, + "loss": 0.5619, + "step": 46998 + }, + { + "epoch": 3.26132815210603, + "grad_norm": 0.981928825378418, + "learning_rate": 1.735939651014874e-05, + "loss": 0.2864, + "step": 46999 + }, + { + "epoch": 3.2613975435431266, + "grad_norm": 1.043258547782898, + "learning_rate": 1.7356232940634864e-05, + "loss": 0.2158, + "step": 47000 + }, + { + "epoch": 3.2614669349802234, + "grad_norm": 0.7723898887634277, + "learning_rate": 1.735306963201776e-05, + "loss": 0.2616, + "step": 47001 + }, + { + "epoch": 3.2615363264173203, + "grad_norm": 0.9998471140861511, + "learning_rate": 1.7349906584307408e-05, + "loss": 0.226, + "step": 47002 + }, + { + "epoch": 3.2616057178544167, + "grad_norm": 0.9370837807655334, + "learning_rate": 1.734674379751382e-05, + "loss": 0.2672, + "step": 47003 + }, + { + "epoch": 3.2616751092915135, + "grad_norm": 0.9053914546966553, + "learning_rate": 1.734358127164698e-05, + "loss": 0.154, + "step": 47004 + }, + { + "epoch": 3.26174450072861, + "grad_norm": 2.28692626953125, + "learning_rate": 1.734041900671688e-05, + "loss": 0.5512, + "step": 47005 + }, + { + "epoch": 3.2618138921657067, + "grad_norm": 1.191633939743042, + "learning_rate": 1.7337257002733463e-05, + "loss": 0.3187, + "step": 47006 + }, + { + "epoch": 3.2618832836028036, + "grad_norm": 1.3655295372009277, + "learning_rate": 1.733409525970673e-05, + "loss": 0.2684, + "step": 47007 + }, + { + "epoch": 3.2619526750399, + "grad_norm": 1.06563401222229, + "learning_rate": 1.7330933777646675e-05, + "loss": 0.2971, + "step": 47008 + }, + { + "epoch": 3.2620220664769968, + "grad_norm": 0.7139651775360107, + "learning_rate": 1.7327772556563293e-05, + "loss": 0.2376, + "step": 47009 + }, + { + "epoch": 3.262091457914093, + "grad_norm": 0.9304356575012207, + "learning_rate": 1.7324611596466524e-05, + "loss": 0.244, + "step": 47010 + }, + { + "epoch": 3.26216084935119, + "grad_norm": 1.3128540515899658, + "learning_rate": 1.7321450897366352e-05, + "loss": 0.264, + "step": 47011 + }, + { + "epoch": 3.262230240788287, + "grad_norm": 0.8035224676132202, + "learning_rate": 1.731829045927277e-05, + "loss": 0.2456, + "step": 47012 + }, + { + "epoch": 3.262299632225383, + "grad_norm": 1.033960223197937, + "learning_rate": 1.7315130282195757e-05, + "loss": 0.3472, + "step": 47013 + }, + { + "epoch": 3.26236902366248, + "grad_norm": 0.7393543720245361, + "learning_rate": 1.7311970366145303e-05, + "loss": 0.2722, + "step": 47014 + }, + { + "epoch": 3.262438415099577, + "grad_norm": 0.7924416661262512, + "learning_rate": 1.730881071113134e-05, + "loss": 0.2848, + "step": 47015 + }, + { + "epoch": 3.2625078065366733, + "grad_norm": 0.9966463446617126, + "learning_rate": 1.7305651317163873e-05, + "loss": 0.4183, + "step": 47016 + }, + { + "epoch": 3.26257719797377, + "grad_norm": 1.0964449644088745, + "learning_rate": 1.7302492184252872e-05, + "loss": 0.3071, + "step": 47017 + }, + { + "epoch": 3.262646589410867, + "grad_norm": 0.7639622688293457, + "learning_rate": 1.7299333312408305e-05, + "loss": 0.2884, + "step": 47018 + }, + { + "epoch": 3.2627159808479633, + "grad_norm": 0.6964291334152222, + "learning_rate": 1.7296174701640143e-05, + "loss": 0.2065, + "step": 47019 + }, + { + "epoch": 3.26278537228506, + "grad_norm": 1.0721949338912964, + "learning_rate": 1.729301635195837e-05, + "loss": 0.2877, + "step": 47020 + }, + { + "epoch": 3.2628547637221565, + "grad_norm": 1.0002070665359497, + "learning_rate": 1.7289858263372972e-05, + "loss": 0.2685, + "step": 47021 + }, + { + "epoch": 3.2629241551592534, + "grad_norm": 1.4829946756362915, + "learning_rate": 1.7286700435893866e-05, + "loss": 0.2822, + "step": 47022 + }, + { + "epoch": 3.26299354659635, + "grad_norm": 0.9089826345443726, + "learning_rate": 1.7283542869531057e-05, + "loss": 0.2945, + "step": 47023 + }, + { + "epoch": 3.2630629380334466, + "grad_norm": 1.028603434562683, + "learning_rate": 1.72803855642945e-05, + "loss": 0.264, + "step": 47024 + }, + { + "epoch": 3.2631323294705434, + "grad_norm": 0.6683157682418823, + "learning_rate": 1.7277228520194166e-05, + "loss": 0.1439, + "step": 47025 + }, + { + "epoch": 3.26320172090764, + "grad_norm": 1.567299485206604, + "learning_rate": 1.7274071737240027e-05, + "loss": 0.317, + "step": 47026 + }, + { + "epoch": 3.2632711123447367, + "grad_norm": 1.146759033203125, + "learning_rate": 1.7270915215442053e-05, + "loss": 0.3369, + "step": 47027 + }, + { + "epoch": 3.2633405037818335, + "grad_norm": 1.3064935207366943, + "learning_rate": 1.7267758954810222e-05, + "loss": 0.2385, + "step": 47028 + }, + { + "epoch": 3.26340989521893, + "grad_norm": 1.2123374938964844, + "learning_rate": 1.7264602955354447e-05, + "loss": 0.5934, + "step": 47029 + }, + { + "epoch": 3.2634792866560267, + "grad_norm": 0.7523715496063232, + "learning_rate": 1.7261447217084725e-05, + "loss": 0.1516, + "step": 47030 + }, + { + "epoch": 3.263548678093123, + "grad_norm": 0.839025616645813, + "learning_rate": 1.725829174001101e-05, + "loss": 0.309, + "step": 47031 + }, + { + "epoch": 3.26361806953022, + "grad_norm": 0.6544777154922485, + "learning_rate": 1.7255136524143278e-05, + "loss": 0.1592, + "step": 47032 + }, + { + "epoch": 3.2636874609673168, + "grad_norm": 0.891697108745575, + "learning_rate": 1.725198156949147e-05, + "loss": 0.3706, + "step": 47033 + }, + { + "epoch": 3.263756852404413, + "grad_norm": 1.6961305141448975, + "learning_rate": 1.724882687606556e-05, + "loss": 0.4461, + "step": 47034 + }, + { + "epoch": 3.26382624384151, + "grad_norm": 0.7741097211837769, + "learning_rate": 1.7245672443875527e-05, + "loss": 0.2499, + "step": 47035 + }, + { + "epoch": 3.2638956352786064, + "grad_norm": 0.7461000084877014, + "learning_rate": 1.7242518272931284e-05, + "loss": 0.1994, + "step": 47036 + }, + { + "epoch": 3.263965026715703, + "grad_norm": 1.0446114540100098, + "learning_rate": 1.7239364363242806e-05, + "loss": 0.3682, + "step": 47037 + }, + { + "epoch": 3.2640344181528, + "grad_norm": 0.8947322368621826, + "learning_rate": 1.7236210714820056e-05, + "loss": 0.2315, + "step": 47038 + }, + { + "epoch": 3.2641038095898964, + "grad_norm": 1.4466990232467651, + "learning_rate": 1.7233057327672973e-05, + "loss": 0.3494, + "step": 47039 + }, + { + "epoch": 3.2641732010269933, + "grad_norm": 1.0307090282440186, + "learning_rate": 1.7229904201811563e-05, + "loss": 0.1783, + "step": 47040 + }, + { + "epoch": 3.26424259246409, + "grad_norm": 0.6787943840026855, + "learning_rate": 1.7226751337245716e-05, + "loss": 0.2356, + "step": 47041 + }, + { + "epoch": 3.2643119839011865, + "grad_norm": 1.228618860244751, + "learning_rate": 1.7223598733985413e-05, + "loss": 0.37, + "step": 47042 + }, + { + "epoch": 3.2643813753382833, + "grad_norm": 0.8269269466400146, + "learning_rate": 1.7220446392040622e-05, + "loss": 0.2556, + "step": 47043 + }, + { + "epoch": 3.26445076677538, + "grad_norm": 1.047523856163025, + "learning_rate": 1.721729431142126e-05, + "loss": 0.3133, + "step": 47044 + }, + { + "epoch": 3.2645201582124765, + "grad_norm": 1.4768368005752563, + "learning_rate": 1.7214142492137296e-05, + "loss": 0.3359, + "step": 47045 + }, + { + "epoch": 3.2645895496495734, + "grad_norm": 0.8434934020042419, + "learning_rate": 1.7210990934198666e-05, + "loss": 0.2561, + "step": 47046 + }, + { + "epoch": 3.2646589410866698, + "grad_norm": 1.1376246213912964, + "learning_rate": 1.7207839637615352e-05, + "loss": 0.3084, + "step": 47047 + }, + { + "epoch": 3.2647283325237666, + "grad_norm": 1.0670500993728638, + "learning_rate": 1.720468860239729e-05, + "loss": 0.33, + "step": 47048 + }, + { + "epoch": 3.2647977239608634, + "grad_norm": 0.8222702741622925, + "learning_rate": 1.7201537828554404e-05, + "loss": 0.2089, + "step": 47049 + }, + { + "epoch": 3.26486711539796, + "grad_norm": 0.6501958966255188, + "learning_rate": 1.7198387316096666e-05, + "loss": 0.1658, + "step": 47050 + }, + { + "epoch": 3.2649365068350567, + "grad_norm": 0.9620513916015625, + "learning_rate": 1.7195237065034e-05, + "loss": 0.2637, + "step": 47051 + }, + { + "epoch": 3.265005898272153, + "grad_norm": 0.8184677958488464, + "learning_rate": 1.7192087075376363e-05, + "loss": 0.2316, + "step": 47052 + }, + { + "epoch": 3.26507528970925, + "grad_norm": 0.7183536291122437, + "learning_rate": 1.7188937347133704e-05, + "loss": 0.2458, + "step": 47053 + }, + { + "epoch": 3.2651446811463467, + "grad_norm": 0.9016485214233398, + "learning_rate": 1.7185787880315964e-05, + "loss": 0.213, + "step": 47054 + }, + { + "epoch": 3.265214072583443, + "grad_norm": 0.741981565952301, + "learning_rate": 1.7182638674933092e-05, + "loss": 0.2126, + "step": 47055 + }, + { + "epoch": 3.26528346402054, + "grad_norm": 0.9338934421539307, + "learning_rate": 1.7179489730995014e-05, + "loss": 0.3121, + "step": 47056 + }, + { + "epoch": 3.2653528554576363, + "grad_norm": 0.8621644377708435, + "learning_rate": 1.717634104851167e-05, + "loss": 0.2801, + "step": 47057 + }, + { + "epoch": 3.265422246894733, + "grad_norm": 1.1246548891067505, + "learning_rate": 1.7173192627493017e-05, + "loss": 0.2399, + "step": 47058 + }, + { + "epoch": 3.26549163833183, + "grad_norm": 1.0710581541061401, + "learning_rate": 1.7170044467948976e-05, + "loss": 0.4061, + "step": 47059 + }, + { + "epoch": 3.2655610297689264, + "grad_norm": 0.5867817401885986, + "learning_rate": 1.7166896569889502e-05, + "loss": 0.1587, + "step": 47060 + }, + { + "epoch": 3.265630421206023, + "grad_norm": 0.5295984745025635, + "learning_rate": 1.7163748933324518e-05, + "loss": 0.1325, + "step": 47061 + }, + { + "epoch": 3.2656998126431196, + "grad_norm": 1.058465838432312, + "learning_rate": 1.716060155826399e-05, + "loss": 0.2455, + "step": 47062 + }, + { + "epoch": 3.2657692040802164, + "grad_norm": 1.4218502044677734, + "learning_rate": 1.7157454444717814e-05, + "loss": 0.3848, + "step": 47063 + }, + { + "epoch": 3.2658385955173133, + "grad_norm": 1.0382287502288818, + "learning_rate": 1.7154307592695938e-05, + "loss": 0.2333, + "step": 47064 + }, + { + "epoch": 3.2659079869544096, + "grad_norm": 1.2541834115982056, + "learning_rate": 1.7151161002208306e-05, + "loss": 0.2965, + "step": 47065 + }, + { + "epoch": 3.2659773783915065, + "grad_norm": 1.1692347526550293, + "learning_rate": 1.7148014673264844e-05, + "loss": 0.2825, + "step": 47066 + }, + { + "epoch": 3.2660467698286033, + "grad_norm": 0.6565874218940735, + "learning_rate": 1.7144868605875486e-05, + "loss": 0.2167, + "step": 47067 + }, + { + "epoch": 3.2661161612656997, + "grad_norm": 0.8542739152908325, + "learning_rate": 1.7141722800050173e-05, + "loss": 0.2622, + "step": 47068 + }, + { + "epoch": 3.2661855527027965, + "grad_norm": 1.617228627204895, + "learning_rate": 1.7138577255798837e-05, + "loss": 0.2874, + "step": 47069 + }, + { + "epoch": 3.2662549441398934, + "grad_norm": 0.7602376937866211, + "learning_rate": 1.713543197313139e-05, + "loss": 0.2468, + "step": 47070 + }, + { + "epoch": 3.2663243355769898, + "grad_norm": 1.2405405044555664, + "learning_rate": 1.7132286952057762e-05, + "loss": 0.3704, + "step": 47071 + }, + { + "epoch": 3.2663937270140866, + "grad_norm": 0.8968556523323059, + "learning_rate": 1.712914219258789e-05, + "loss": 0.2808, + "step": 47072 + }, + { + "epoch": 3.266463118451183, + "grad_norm": 0.8566410541534424, + "learning_rate": 1.71259976947317e-05, + "loss": 0.2707, + "step": 47073 + }, + { + "epoch": 3.26653250988828, + "grad_norm": 0.648983895778656, + "learning_rate": 1.712285345849912e-05, + "loss": 0.2172, + "step": 47074 + }, + { + "epoch": 3.2666019013253766, + "grad_norm": 0.8785334229469299, + "learning_rate": 1.7119709483900103e-05, + "loss": 0.5064, + "step": 47075 + }, + { + "epoch": 3.266671292762473, + "grad_norm": 0.856178879737854, + "learning_rate": 1.7116565770944525e-05, + "loss": 0.3469, + "step": 47076 + }, + { + "epoch": 3.26674068419957, + "grad_norm": 1.0403105020523071, + "learning_rate": 1.7113422319642334e-05, + "loss": 0.3231, + "step": 47077 + }, + { + "epoch": 3.2668100756366663, + "grad_norm": 1.368582844734192, + "learning_rate": 1.711027913000347e-05, + "loss": 0.4712, + "step": 47078 + }, + { + "epoch": 3.266879467073763, + "grad_norm": 1.0897639989852905, + "learning_rate": 1.710713620203782e-05, + "loss": 0.2908, + "step": 47079 + }, + { + "epoch": 3.26694885851086, + "grad_norm": 1.067295789718628, + "learning_rate": 1.7103993535755314e-05, + "loss": 0.3563, + "step": 47080 + }, + { + "epoch": 3.2670182499479563, + "grad_norm": 1.0104353427886963, + "learning_rate": 1.7100851131165895e-05, + "loss": 0.3082, + "step": 47081 + }, + { + "epoch": 3.267087641385053, + "grad_norm": 0.8811454772949219, + "learning_rate": 1.7097708988279458e-05, + "loss": 0.2484, + "step": 47082 + }, + { + "epoch": 3.2671570328221495, + "grad_norm": 0.7629802227020264, + "learning_rate": 1.7094567107105975e-05, + "loss": 0.2614, + "step": 47083 + }, + { + "epoch": 3.2672264242592464, + "grad_norm": 1.3567912578582764, + "learning_rate": 1.7091425487655288e-05, + "loss": 0.2952, + "step": 47084 + }, + { + "epoch": 3.267295815696343, + "grad_norm": 0.9032691121101379, + "learning_rate": 1.7088284129937348e-05, + "loss": 0.224, + "step": 47085 + }, + { + "epoch": 3.2673652071334396, + "grad_norm": 0.9720858931541443, + "learning_rate": 1.7085143033962092e-05, + "loss": 0.2441, + "step": 47086 + }, + { + "epoch": 3.2674345985705364, + "grad_norm": 0.7000833749771118, + "learning_rate": 1.708200219973941e-05, + "loss": 0.1798, + "step": 47087 + }, + { + "epoch": 3.2675039900076333, + "grad_norm": 1.061623215675354, + "learning_rate": 1.7078861627279218e-05, + "loss": 0.2821, + "step": 47088 + }, + { + "epoch": 3.2675733814447296, + "grad_norm": 2.085771322250366, + "learning_rate": 1.7075721316591453e-05, + "loss": 0.2339, + "step": 47089 + }, + { + "epoch": 3.2676427728818265, + "grad_norm": 0.7415015697479248, + "learning_rate": 1.7072581267686038e-05, + "loss": 0.1732, + "step": 47090 + }, + { + "epoch": 3.267712164318923, + "grad_norm": 1.0194385051727295, + "learning_rate": 1.7069441480572835e-05, + "loss": 0.2478, + "step": 47091 + }, + { + "epoch": 3.2677815557560197, + "grad_norm": 0.7721012234687805, + "learning_rate": 1.7066301955261778e-05, + "loss": 0.1857, + "step": 47092 + }, + { + "epoch": 3.2678509471931165, + "grad_norm": 1.1280757188796997, + "learning_rate": 1.706316269176279e-05, + "loss": 0.3572, + "step": 47093 + }, + { + "epoch": 3.267920338630213, + "grad_norm": 0.8768872022628784, + "learning_rate": 1.7060023690085767e-05, + "loss": 0.23, + "step": 47094 + }, + { + "epoch": 3.2679897300673098, + "grad_norm": 1.1505545377731323, + "learning_rate": 1.7056884950240638e-05, + "loss": 0.2358, + "step": 47095 + }, + { + "epoch": 3.2680591215044066, + "grad_norm": 0.8887135982513428, + "learning_rate": 1.70537464722373e-05, + "loss": 0.3204, + "step": 47096 + }, + { + "epoch": 3.268128512941503, + "grad_norm": 1.3507375717163086, + "learning_rate": 1.7050608256085678e-05, + "loss": 0.2239, + "step": 47097 + }, + { + "epoch": 3.2681979043786, + "grad_norm": 0.8827003240585327, + "learning_rate": 1.7047470301795644e-05, + "loss": 0.2366, + "step": 47098 + }, + { + "epoch": 3.268267295815696, + "grad_norm": 0.7698726058006287, + "learning_rate": 1.704433260937711e-05, + "loss": 0.2231, + "step": 47099 + }, + { + "epoch": 3.268336687252793, + "grad_norm": 0.7466473579406738, + "learning_rate": 1.7041195178840007e-05, + "loss": 0.1815, + "step": 47100 + }, + { + "epoch": 3.26840607868989, + "grad_norm": 0.7317022681236267, + "learning_rate": 1.7038058010194215e-05, + "loss": 0.2788, + "step": 47101 + }, + { + "epoch": 3.2684754701269862, + "grad_norm": 1.1722590923309326, + "learning_rate": 1.703492110344965e-05, + "loss": 0.3201, + "step": 47102 + }, + { + "epoch": 3.268544861564083, + "grad_norm": 0.6993798613548279, + "learning_rate": 1.703178445861622e-05, + "loss": 0.1572, + "step": 47103 + }, + { + "epoch": 3.2686142530011795, + "grad_norm": 0.8208770155906677, + "learning_rate": 1.7028648075703847e-05, + "loss": 0.2269, + "step": 47104 + }, + { + "epoch": 3.2686836444382763, + "grad_norm": 1.1870969533920288, + "learning_rate": 1.7025511954722373e-05, + "loss": 0.2407, + "step": 47105 + }, + { + "epoch": 3.268753035875373, + "grad_norm": 1.6745330095291138, + "learning_rate": 1.7022376095681735e-05, + "loss": 0.3116, + "step": 47106 + }, + { + "epoch": 3.2688224273124695, + "grad_norm": 1.007737159729004, + "learning_rate": 1.7019240498591827e-05, + "loss": 0.2516, + "step": 47107 + }, + { + "epoch": 3.2688918187495664, + "grad_norm": 0.8470851182937622, + "learning_rate": 1.7016105163462548e-05, + "loss": 0.2378, + "step": 47108 + }, + { + "epoch": 3.2689612101866627, + "grad_norm": 0.7669568657875061, + "learning_rate": 1.701297009030379e-05, + "loss": 0.1511, + "step": 47109 + }, + { + "epoch": 3.2690306016237596, + "grad_norm": 0.7459577322006226, + "learning_rate": 1.7009835279125498e-05, + "loss": 0.2234, + "step": 47110 + }, + { + "epoch": 3.2690999930608564, + "grad_norm": 1.1148691177368164, + "learning_rate": 1.7006700729937497e-05, + "loss": 0.2605, + "step": 47111 + }, + { + "epoch": 3.269169384497953, + "grad_norm": 0.8069899678230286, + "learning_rate": 1.7003566442749706e-05, + "loss": 0.226, + "step": 47112 + }, + { + "epoch": 3.2692387759350496, + "grad_norm": 0.8836886882781982, + "learning_rate": 1.7000432417572056e-05, + "loss": 0.3189, + "step": 47113 + }, + { + "epoch": 3.2693081673721465, + "grad_norm": 0.7776733636856079, + "learning_rate": 1.6997298654414384e-05, + "loss": 0.1879, + "step": 47114 + }, + { + "epoch": 3.269377558809243, + "grad_norm": 0.9205312728881836, + "learning_rate": 1.6994165153286612e-05, + "loss": 0.2457, + "step": 47115 + }, + { + "epoch": 3.2694469502463397, + "grad_norm": 1.228692889213562, + "learning_rate": 1.699103191419863e-05, + "loss": 0.2749, + "step": 47116 + }, + { + "epoch": 3.269516341683436, + "grad_norm": 1.408706784248352, + "learning_rate": 1.698789893716035e-05, + "loss": 0.2681, + "step": 47117 + }, + { + "epoch": 3.269585733120533, + "grad_norm": 1.327579140663147, + "learning_rate": 1.6984766222181625e-05, + "loss": 0.3672, + "step": 47118 + }, + { + "epoch": 3.2696551245576297, + "grad_norm": 0.5742344856262207, + "learning_rate": 1.6981633769272365e-05, + "loss": 0.1598, + "step": 47119 + }, + { + "epoch": 3.269724515994726, + "grad_norm": 0.7746668457984924, + "learning_rate": 1.697850157844244e-05, + "loss": 0.2124, + "step": 47120 + }, + { + "epoch": 3.269793907431823, + "grad_norm": 0.9943293333053589, + "learning_rate": 1.6975369649701767e-05, + "loss": 0.301, + "step": 47121 + }, + { + "epoch": 3.26986329886892, + "grad_norm": 1.0587308406829834, + "learning_rate": 1.6972237983060212e-05, + "loss": 0.2063, + "step": 47122 + }, + { + "epoch": 3.269932690306016, + "grad_norm": 0.945587158203125, + "learning_rate": 1.696910657852767e-05, + "loss": 0.1402, + "step": 47123 + }, + { + "epoch": 3.270002081743113, + "grad_norm": 0.9558417201042175, + "learning_rate": 1.6965975436114057e-05, + "loss": 0.3207, + "step": 47124 + }, + { + "epoch": 3.2700714731802094, + "grad_norm": 1.1759614944458008, + "learning_rate": 1.6962844555829193e-05, + "loss": 0.3343, + "step": 47125 + }, + { + "epoch": 3.2701408646173062, + "grad_norm": 0.9332822561264038, + "learning_rate": 1.6959713937682998e-05, + "loss": 0.2323, + "step": 47126 + }, + { + "epoch": 3.270210256054403, + "grad_norm": 1.2473883628845215, + "learning_rate": 1.6956583581685348e-05, + "loss": 0.2691, + "step": 47127 + }, + { + "epoch": 3.2702796474914995, + "grad_norm": 1.1067407131195068, + "learning_rate": 1.6953453487846127e-05, + "loss": 0.2534, + "step": 47128 + }, + { + "epoch": 3.2703490389285963, + "grad_norm": 1.0129046440124512, + "learning_rate": 1.6950323656175227e-05, + "loss": 0.2101, + "step": 47129 + }, + { + "epoch": 3.2704184303656927, + "grad_norm": 0.7892872095108032, + "learning_rate": 1.6947194086682505e-05, + "loss": 0.1819, + "step": 47130 + }, + { + "epoch": 3.2704878218027895, + "grad_norm": 0.6402595043182373, + "learning_rate": 1.69440647793779e-05, + "loss": 0.124, + "step": 47131 + }, + { + "epoch": 3.2705572132398864, + "grad_norm": 0.9787907600402832, + "learning_rate": 1.6940935734271203e-05, + "loss": 0.2394, + "step": 47132 + }, + { + "epoch": 3.2706266046769827, + "grad_norm": 0.9473981857299805, + "learning_rate": 1.6937806951372348e-05, + "loss": 0.3497, + "step": 47133 + }, + { + "epoch": 3.2706959961140796, + "grad_norm": 0.7409324645996094, + "learning_rate": 1.6934678430691197e-05, + "loss": 0.2028, + "step": 47134 + }, + { + "epoch": 3.270765387551176, + "grad_norm": 0.7904589176177979, + "learning_rate": 1.693155017223762e-05, + "loss": 0.171, + "step": 47135 + }, + { + "epoch": 3.270834778988273, + "grad_norm": 1.009685754776001, + "learning_rate": 1.692842217602151e-05, + "loss": 0.2566, + "step": 47136 + }, + { + "epoch": 3.2709041704253696, + "grad_norm": 0.6667447090148926, + "learning_rate": 1.692529444205273e-05, + "loss": 0.199, + "step": 47137 + }, + { + "epoch": 3.270973561862466, + "grad_norm": 1.945852518081665, + "learning_rate": 1.692216697034118e-05, + "loss": 0.303, + "step": 47138 + }, + { + "epoch": 3.271042953299563, + "grad_norm": 0.9472081661224365, + "learning_rate": 1.6919039760896695e-05, + "loss": 0.122, + "step": 47139 + }, + { + "epoch": 3.2711123447366597, + "grad_norm": 0.8479140400886536, + "learning_rate": 1.691591281372915e-05, + "loss": 0.2091, + "step": 47140 + }, + { + "epoch": 3.271181736173756, + "grad_norm": 1.023292064666748, + "learning_rate": 1.6912786128848436e-05, + "loss": 0.2552, + "step": 47141 + }, + { + "epoch": 3.271251127610853, + "grad_norm": 0.9202150702476501, + "learning_rate": 1.6909659706264413e-05, + "loss": 0.3457, + "step": 47142 + }, + { + "epoch": 3.2713205190479493, + "grad_norm": 1.2912062406539917, + "learning_rate": 1.6906533545986957e-05, + "loss": 0.2683, + "step": 47143 + }, + { + "epoch": 3.271389910485046, + "grad_norm": 1.2509267330169678, + "learning_rate": 1.690340764802596e-05, + "loss": 0.2631, + "step": 47144 + }, + { + "epoch": 3.271459301922143, + "grad_norm": 1.0915645360946655, + "learning_rate": 1.6900282012391234e-05, + "loss": 0.2967, + "step": 47145 + }, + { + "epoch": 3.2715286933592393, + "grad_norm": 0.8929974436759949, + "learning_rate": 1.6897156639092682e-05, + "loss": 0.1572, + "step": 47146 + }, + { + "epoch": 3.271598084796336, + "grad_norm": 0.677456796169281, + "learning_rate": 1.6894031528140163e-05, + "loss": 0.2537, + "step": 47147 + }, + { + "epoch": 3.271667476233433, + "grad_norm": 1.1488605737686157, + "learning_rate": 1.689090667954356e-05, + "loss": 0.3294, + "step": 47148 + }, + { + "epoch": 3.2717368676705294, + "grad_norm": 0.8091408014297485, + "learning_rate": 1.688778209331271e-05, + "loss": 0.2451, + "step": 47149 + }, + { + "epoch": 3.2718062591076262, + "grad_norm": 1.034538745880127, + "learning_rate": 1.688465776945748e-05, + "loss": 0.4689, + "step": 47150 + }, + { + "epoch": 3.2718756505447226, + "grad_norm": 1.4030404090881348, + "learning_rate": 1.688153370798775e-05, + "loss": 0.2551, + "step": 47151 + }, + { + "epoch": 3.2719450419818195, + "grad_norm": 1.0439037084579468, + "learning_rate": 1.6878409908913395e-05, + "loss": 0.3657, + "step": 47152 + }, + { + "epoch": 3.2720144334189163, + "grad_norm": 0.9839107990264893, + "learning_rate": 1.6875286372244237e-05, + "loss": 0.2313, + "step": 47153 + }, + { + "epoch": 3.2720838248560127, + "grad_norm": 1.1850985288619995, + "learning_rate": 1.6872163097990145e-05, + "loss": 0.3739, + "step": 47154 + }, + { + "epoch": 3.2721532162931095, + "grad_norm": 0.9852069616317749, + "learning_rate": 1.6869040086161005e-05, + "loss": 0.3057, + "step": 47155 + }, + { + "epoch": 3.272222607730206, + "grad_norm": 0.6338034868240356, + "learning_rate": 1.6865917336766646e-05, + "loss": 0.1288, + "step": 47156 + }, + { + "epoch": 3.2722919991673027, + "grad_norm": 0.9680497050285339, + "learning_rate": 1.6862794849816942e-05, + "loss": 0.2329, + "step": 47157 + }, + { + "epoch": 3.2723613906043996, + "grad_norm": 1.0038928985595703, + "learning_rate": 1.6859672625321755e-05, + "loss": 0.3969, + "step": 47158 + }, + { + "epoch": 3.272430782041496, + "grad_norm": 0.8891764283180237, + "learning_rate": 1.685655066329096e-05, + "loss": 0.1492, + "step": 47159 + }, + { + "epoch": 3.272500173478593, + "grad_norm": 0.8091293573379517, + "learning_rate": 1.6853428963734352e-05, + "loss": 0.2139, + "step": 47160 + }, + { + "epoch": 3.272569564915689, + "grad_norm": 1.0471093654632568, + "learning_rate": 1.6850307526661833e-05, + "loss": 0.2977, + "step": 47161 + }, + { + "epoch": 3.272638956352786, + "grad_norm": 0.90537428855896, + "learning_rate": 1.6847186352083243e-05, + "loss": 0.322, + "step": 47162 + }, + { + "epoch": 3.272708347789883, + "grad_norm": 1.235721230506897, + "learning_rate": 1.684406544000844e-05, + "loss": 0.3722, + "step": 47163 + }, + { + "epoch": 3.2727777392269792, + "grad_norm": 1.0006366968154907, + "learning_rate": 1.6840944790447277e-05, + "loss": 0.3425, + "step": 47164 + }, + { + "epoch": 3.272847130664076, + "grad_norm": 1.0335465669631958, + "learning_rate": 1.6837824403409598e-05, + "loss": 0.2968, + "step": 47165 + }, + { + "epoch": 3.272916522101173, + "grad_norm": 0.8880339860916138, + "learning_rate": 1.6834704278905287e-05, + "loss": 0.2397, + "step": 47166 + }, + { + "epoch": 3.2729859135382693, + "grad_norm": 1.0542800426483154, + "learning_rate": 1.6831584416944134e-05, + "loss": 0.2693, + "step": 47167 + }, + { + "epoch": 3.273055304975366, + "grad_norm": 0.6313795447349548, + "learning_rate": 1.6828464817536026e-05, + "loss": 0.2754, + "step": 47168 + }, + { + "epoch": 3.2731246964124625, + "grad_norm": 1.444972038269043, + "learning_rate": 1.6825345480690814e-05, + "loss": 0.3039, + "step": 47169 + }, + { + "epoch": 3.2731940878495593, + "grad_norm": 0.9834609627723694, + "learning_rate": 1.6822226406418318e-05, + "loss": 0.2148, + "step": 47170 + }, + { + "epoch": 3.273263479286656, + "grad_norm": 0.8795164823532104, + "learning_rate": 1.681910759472841e-05, + "loss": 0.1786, + "step": 47171 + }, + { + "epoch": 3.2733328707237526, + "grad_norm": 0.9891856908798218, + "learning_rate": 1.681598904563093e-05, + "loss": 0.1337, + "step": 47172 + }, + { + "epoch": 3.2734022621608494, + "grad_norm": 1.1380386352539062, + "learning_rate": 1.6812870759135746e-05, + "loss": 0.2356, + "step": 47173 + }, + { + "epoch": 3.2734716535979462, + "grad_norm": 0.9886236190795898, + "learning_rate": 1.680975273525265e-05, + "loss": 0.1707, + "step": 47174 + }, + { + "epoch": 3.2735410450350426, + "grad_norm": 1.3941229581832886, + "learning_rate": 1.680663497399152e-05, + "loss": 0.4733, + "step": 47175 + }, + { + "epoch": 3.2736104364721395, + "grad_norm": 0.8305443525314331, + "learning_rate": 1.6803517475362184e-05, + "loss": 0.189, + "step": 47176 + }, + { + "epoch": 3.273679827909236, + "grad_norm": 0.8881478905677795, + "learning_rate": 1.6800400239374492e-05, + "loss": 0.307, + "step": 47177 + }, + { + "epoch": 3.2737492193463327, + "grad_norm": 0.9077866077423096, + "learning_rate": 1.679728326603829e-05, + "loss": 0.3735, + "step": 47178 + }, + { + "epoch": 3.2738186107834295, + "grad_norm": 0.8163923621177673, + "learning_rate": 1.6794166555363423e-05, + "loss": 0.1895, + "step": 47179 + }, + { + "epoch": 3.273888002220526, + "grad_norm": 0.7524274587631226, + "learning_rate": 1.6791050107359707e-05, + "loss": 0.2412, + "step": 47180 + }, + { + "epoch": 3.2739573936576227, + "grad_norm": 0.9429642558097839, + "learning_rate": 1.678793392203698e-05, + "loss": 0.3497, + "step": 47181 + }, + { + "epoch": 3.274026785094719, + "grad_norm": 0.619493305683136, + "learning_rate": 1.67848179994051e-05, + "loss": 0.1129, + "step": 47182 + }, + { + "epoch": 3.274096176531816, + "grad_norm": 1.0474967956542969, + "learning_rate": 1.6781702339473913e-05, + "loss": 0.2557, + "step": 47183 + }, + { + "epoch": 3.274165567968913, + "grad_norm": 0.7050389647483826, + "learning_rate": 1.6778586942253227e-05, + "loss": 0.207, + "step": 47184 + }, + { + "epoch": 3.274234959406009, + "grad_norm": 1.1451619863510132, + "learning_rate": 1.6775471807752875e-05, + "loss": 0.2682, + "step": 47185 + }, + { + "epoch": 3.274304350843106, + "grad_norm": 1.4888919591903687, + "learning_rate": 1.67723569359827e-05, + "loss": 0.2462, + "step": 47186 + }, + { + "epoch": 3.2743737422802024, + "grad_norm": 0.9211450219154358, + "learning_rate": 1.676924232695256e-05, + "loss": 0.248, + "step": 47187 + }, + { + "epoch": 3.2744431337172992, + "grad_norm": 1.4278464317321777, + "learning_rate": 1.676612798067225e-05, + "loss": 0.3045, + "step": 47188 + }, + { + "epoch": 3.274512525154396, + "grad_norm": 1.2135424613952637, + "learning_rate": 1.6763013897151613e-05, + "loss": 0.3345, + "step": 47189 + }, + { + "epoch": 3.2745819165914924, + "grad_norm": 0.7430368661880493, + "learning_rate": 1.675990007640048e-05, + "loss": 0.1714, + "step": 47190 + }, + { + "epoch": 3.2746513080285893, + "grad_norm": 0.8583263754844666, + "learning_rate": 1.675678651842869e-05, + "loss": 0.2407, + "step": 47191 + }, + { + "epoch": 3.274720699465686, + "grad_norm": 0.8895275592803955, + "learning_rate": 1.675367322324607e-05, + "loss": 0.3119, + "step": 47192 + }, + { + "epoch": 3.2747900909027825, + "grad_norm": 0.6532415151596069, + "learning_rate": 1.675056019086246e-05, + "loss": 0.1953, + "step": 47193 + }, + { + "epoch": 3.2748594823398793, + "grad_norm": 1.6005991697311401, + "learning_rate": 1.6747447421287666e-05, + "loss": 0.3192, + "step": 47194 + }, + { + "epoch": 3.274928873776976, + "grad_norm": 1.4530385732650757, + "learning_rate": 1.67443349145315e-05, + "loss": 0.27, + "step": 47195 + }, + { + "epoch": 3.2749982652140726, + "grad_norm": 0.8079380393028259, + "learning_rate": 1.674122267060383e-05, + "loss": 0.1468, + "step": 47196 + }, + { + "epoch": 3.2750676566511694, + "grad_norm": 0.8094096779823303, + "learning_rate": 1.673811068951444e-05, + "loss": 0.2373, + "step": 47197 + }, + { + "epoch": 3.275137048088266, + "grad_norm": 1.3298759460449219, + "learning_rate": 1.6734998971273185e-05, + "loss": 0.3022, + "step": 47198 + }, + { + "epoch": 3.2752064395253626, + "grad_norm": 0.9202873706817627, + "learning_rate": 1.6731887515889878e-05, + "loss": 0.2078, + "step": 47199 + }, + { + "epoch": 3.2752758309624594, + "grad_norm": 0.6637300848960876, + "learning_rate": 1.6728776323374363e-05, + "loss": 0.2136, + "step": 47200 + }, + { + "epoch": 3.275345222399556, + "grad_norm": 0.9817861318588257, + "learning_rate": 1.672566539373641e-05, + "loss": 0.2156, + "step": 47201 + }, + { + "epoch": 3.2754146138366527, + "grad_norm": 0.9524036049842834, + "learning_rate": 1.6722554726985884e-05, + "loss": 0.2031, + "step": 47202 + }, + { + "epoch": 3.275484005273749, + "grad_norm": 1.1684777736663818, + "learning_rate": 1.671944432313258e-05, + "loss": 0.3163, + "step": 47203 + }, + { + "epoch": 3.275553396710846, + "grad_norm": 0.7979189157485962, + "learning_rate": 1.6716334182186332e-05, + "loss": 0.2697, + "step": 47204 + }, + { + "epoch": 3.2756227881479427, + "grad_norm": 0.9825541377067566, + "learning_rate": 1.6713224304156948e-05, + "loss": 0.1927, + "step": 47205 + }, + { + "epoch": 3.275692179585039, + "grad_norm": 1.0554291009902954, + "learning_rate": 1.6710114689054258e-05, + "loss": 0.2565, + "step": 47206 + }, + { + "epoch": 3.275761571022136, + "grad_norm": 0.5740257501602173, + "learning_rate": 1.6707005336888105e-05, + "loss": 0.1122, + "step": 47207 + }, + { + "epoch": 3.2758309624592323, + "grad_norm": 1.468955397605896, + "learning_rate": 1.670389624766824e-05, + "loss": 0.3347, + "step": 47208 + }, + { + "epoch": 3.275900353896329, + "grad_norm": 0.822356641292572, + "learning_rate": 1.6700787421404508e-05, + "loss": 0.2654, + "step": 47209 + }, + { + "epoch": 3.275969745333426, + "grad_norm": 1.298136591911316, + "learning_rate": 1.6697678858106725e-05, + "loss": 0.2402, + "step": 47210 + }, + { + "epoch": 3.2760391367705224, + "grad_norm": 0.9125332832336426, + "learning_rate": 1.6694570557784707e-05, + "loss": 0.3125, + "step": 47211 + }, + { + "epoch": 3.276108528207619, + "grad_norm": 1.2081835269927979, + "learning_rate": 1.6691462520448265e-05, + "loss": 0.2753, + "step": 47212 + }, + { + "epoch": 3.2761779196447156, + "grad_norm": 0.9393717646598816, + "learning_rate": 1.6688354746107205e-05, + "loss": 0.1516, + "step": 47213 + }, + { + "epoch": 3.2762473110818124, + "grad_norm": 0.9364620447158813, + "learning_rate": 1.6685247234771373e-05, + "loss": 0.219, + "step": 47214 + }, + { + "epoch": 3.2763167025189093, + "grad_norm": 1.1154500246047974, + "learning_rate": 1.668213998645052e-05, + "loss": 0.277, + "step": 47215 + }, + { + "epoch": 3.2763860939560057, + "grad_norm": 0.8661090731620789, + "learning_rate": 1.6679033001154487e-05, + "loss": 0.3073, + "step": 47216 + }, + { + "epoch": 3.2764554853931025, + "grad_norm": 0.9653818011283875, + "learning_rate": 1.6675926278893084e-05, + "loss": 0.2837, + "step": 47217 + }, + { + "epoch": 3.2765248768301993, + "grad_norm": 1.072095274925232, + "learning_rate": 1.6672819819676134e-05, + "loss": 0.362, + "step": 47218 + }, + { + "epoch": 3.2765942682672957, + "grad_norm": 0.6283448338508606, + "learning_rate": 1.6669713623513407e-05, + "loss": 0.1903, + "step": 47219 + }, + { + "epoch": 3.2766636597043926, + "grad_norm": 1.55320405960083, + "learning_rate": 1.6666607690414715e-05, + "loss": 0.3481, + "step": 47220 + }, + { + "epoch": 3.2767330511414894, + "grad_norm": 1.08939790725708, + "learning_rate": 1.666350202038991e-05, + "loss": 0.2167, + "step": 47221 + }, + { + "epoch": 3.2768024425785858, + "grad_norm": 1.6041982173919678, + "learning_rate": 1.6660396613448724e-05, + "loss": 0.2426, + "step": 47222 + }, + { + "epoch": 3.2768718340156826, + "grad_norm": 0.9380235075950623, + "learning_rate": 1.665729146960101e-05, + "loss": 0.2007, + "step": 47223 + }, + { + "epoch": 3.276941225452779, + "grad_norm": 0.656853973865509, + "learning_rate": 1.6654186588856545e-05, + "loss": 0.1852, + "step": 47224 + }, + { + "epoch": 3.277010616889876, + "grad_norm": 0.7123876214027405, + "learning_rate": 1.6651081971225147e-05, + "loss": 0.1964, + "step": 47225 + }, + { + "epoch": 3.2770800083269727, + "grad_norm": 0.9085597991943359, + "learning_rate": 1.6647977616716616e-05, + "loss": 0.2536, + "step": 47226 + }, + { + "epoch": 3.277149399764069, + "grad_norm": 1.1621792316436768, + "learning_rate": 1.6644873525340744e-05, + "loss": 0.1804, + "step": 47227 + }, + { + "epoch": 3.277218791201166, + "grad_norm": 1.732594609260559, + "learning_rate": 1.664176969710737e-05, + "loss": 0.3689, + "step": 47228 + }, + { + "epoch": 3.2772881826382623, + "grad_norm": 1.0181480646133423, + "learning_rate": 1.6638666132026227e-05, + "loss": 0.2965, + "step": 47229 + }, + { + "epoch": 3.277357574075359, + "grad_norm": 0.7616724371910095, + "learning_rate": 1.663556283010713e-05, + "loss": 0.1858, + "step": 47230 + }, + { + "epoch": 3.277426965512456, + "grad_norm": 0.9300963878631592, + "learning_rate": 1.6632459791359913e-05, + "loss": 0.3513, + "step": 47231 + }, + { + "epoch": 3.2774963569495523, + "grad_norm": 1.1210845708847046, + "learning_rate": 1.6629357015794335e-05, + "loss": 0.2152, + "step": 47232 + }, + { + "epoch": 3.277565748386649, + "grad_norm": 0.6997535228729248, + "learning_rate": 1.6626254503420203e-05, + "loss": 0.1501, + "step": 47233 + }, + { + "epoch": 3.2776351398237455, + "grad_norm": 0.9320030808448792, + "learning_rate": 1.662315225424731e-05, + "loss": 0.3022, + "step": 47234 + }, + { + "epoch": 3.2777045312608424, + "grad_norm": 0.7531080842018127, + "learning_rate": 1.6620050268285482e-05, + "loss": 0.1813, + "step": 47235 + }, + { + "epoch": 3.277773922697939, + "grad_norm": 1.3778822422027588, + "learning_rate": 1.6616948545544453e-05, + "loss": 0.1986, + "step": 47236 + }, + { + "epoch": 3.2778433141350356, + "grad_norm": 0.9852475523948669, + "learning_rate": 1.661384708603405e-05, + "loss": 0.2473, + "step": 47237 + }, + { + "epoch": 3.2779127055721324, + "grad_norm": 0.8056983351707458, + "learning_rate": 1.6610745889764055e-05, + "loss": 0.2578, + "step": 47238 + }, + { + "epoch": 3.277982097009229, + "grad_norm": 1.1084682941436768, + "learning_rate": 1.660764495674426e-05, + "loss": 0.2377, + "step": 47239 + }, + { + "epoch": 3.2780514884463257, + "grad_norm": 1.0538980960845947, + "learning_rate": 1.6604544286984446e-05, + "loss": 0.2188, + "step": 47240 + }, + { + "epoch": 3.2781208798834225, + "grad_norm": 1.216825008392334, + "learning_rate": 1.6601443880494426e-05, + "loss": 0.2633, + "step": 47241 + }, + { + "epoch": 3.278190271320519, + "grad_norm": 0.83326655626297, + "learning_rate": 1.6598343737283984e-05, + "loss": 0.3536, + "step": 47242 + }, + { + "epoch": 3.2782596627576157, + "grad_norm": 0.9596609473228455, + "learning_rate": 1.6595243857362874e-05, + "loss": 0.2773, + "step": 47243 + }, + { + "epoch": 3.2783290541947125, + "grad_norm": 0.7703666687011719, + "learning_rate": 1.6592144240740903e-05, + "loss": 0.2208, + "step": 47244 + }, + { + "epoch": 3.278398445631809, + "grad_norm": 0.9090961217880249, + "learning_rate": 1.658904488742785e-05, + "loss": 0.2101, + "step": 47245 + }, + { + "epoch": 3.2784678370689058, + "grad_norm": 0.7849027514457703, + "learning_rate": 1.658594579743351e-05, + "loss": 0.2027, + "step": 47246 + }, + { + "epoch": 3.2785372285060026, + "grad_norm": 1.0107617378234863, + "learning_rate": 1.658284697076766e-05, + "loss": 0.2415, + "step": 47247 + }, + { + "epoch": 3.278606619943099, + "grad_norm": 0.8970651030540466, + "learning_rate": 1.6579748407440088e-05, + "loss": 0.2384, + "step": 47248 + }, + { + "epoch": 3.278676011380196, + "grad_norm": 1.7333793640136719, + "learning_rate": 1.657665010746059e-05, + "loss": 0.3745, + "step": 47249 + }, + { + "epoch": 3.278745402817292, + "grad_norm": 1.3764309883117676, + "learning_rate": 1.6573552070838906e-05, + "loss": 0.3723, + "step": 47250 + }, + { + "epoch": 3.278814794254389, + "grad_norm": 1.140914797782898, + "learning_rate": 1.6570454297584837e-05, + "loss": 0.3335, + "step": 47251 + }, + { + "epoch": 3.278884185691486, + "grad_norm": 0.7431280612945557, + "learning_rate": 1.6567356787708175e-05, + "loss": 0.1599, + "step": 47252 + }, + { + "epoch": 3.2789535771285823, + "grad_norm": 1.1048672199249268, + "learning_rate": 1.6564259541218696e-05, + "loss": 0.2835, + "step": 47253 + }, + { + "epoch": 3.279022968565679, + "grad_norm": 0.9785783290863037, + "learning_rate": 1.6561162558126153e-05, + "loss": 0.3681, + "step": 47254 + }, + { + "epoch": 3.2790923600027755, + "grad_norm": 1.5229978561401367, + "learning_rate": 1.655806583844034e-05, + "loss": 0.3536, + "step": 47255 + }, + { + "epoch": 3.2791617514398723, + "grad_norm": 1.1840927600860596, + "learning_rate": 1.655496938217106e-05, + "loss": 0.3887, + "step": 47256 + }, + { + "epoch": 3.279231142876969, + "grad_norm": 1.432266354560852, + "learning_rate": 1.6551873189328028e-05, + "loss": 0.2217, + "step": 47257 + }, + { + "epoch": 3.2793005343140655, + "grad_norm": 1.0260339975357056, + "learning_rate": 1.6548777259921056e-05, + "loss": 0.3431, + "step": 47258 + }, + { + "epoch": 3.2793699257511624, + "grad_norm": 0.8086276650428772, + "learning_rate": 1.6545681593959904e-05, + "loss": 0.3164, + "step": 47259 + }, + { + "epoch": 3.2794393171882588, + "grad_norm": 0.945686399936676, + "learning_rate": 1.6542586191454367e-05, + "loss": 0.3202, + "step": 47260 + }, + { + "epoch": 3.2795087086253556, + "grad_norm": 0.8336130380630493, + "learning_rate": 1.653949105241419e-05, + "loss": 0.2726, + "step": 47261 + }, + { + "epoch": 3.2795781000624524, + "grad_norm": 0.8818267583847046, + "learning_rate": 1.653639617684917e-05, + "loss": 0.3003, + "step": 47262 + }, + { + "epoch": 3.279647491499549, + "grad_norm": 0.9462692141532898, + "learning_rate": 1.6533301564769066e-05, + "loss": 0.2322, + "step": 47263 + }, + { + "epoch": 3.2797168829366456, + "grad_norm": 1.0031942129135132, + "learning_rate": 1.6530207216183634e-05, + "loss": 0.2284, + "step": 47264 + }, + { + "epoch": 3.279786274373742, + "grad_norm": 0.9311997890472412, + "learning_rate": 1.6527113131102657e-05, + "loss": 0.2369, + "step": 47265 + }, + { + "epoch": 3.279855665810839, + "grad_norm": 1.4258078336715698, + "learning_rate": 1.65240193095359e-05, + "loss": 0.3534, + "step": 47266 + }, + { + "epoch": 3.2799250572479357, + "grad_norm": 1.3040053844451904, + "learning_rate": 1.652092575149312e-05, + "loss": 0.3612, + "step": 47267 + }, + { + "epoch": 3.279994448685032, + "grad_norm": 0.5929514765739441, + "learning_rate": 1.6517832456984094e-05, + "loss": 0.2482, + "step": 47268 + }, + { + "epoch": 3.280063840122129, + "grad_norm": 1.322134256362915, + "learning_rate": 1.6514739426018612e-05, + "loss": 0.32, + "step": 47269 + }, + { + "epoch": 3.2801332315592258, + "grad_norm": 1.1555917263031006, + "learning_rate": 1.6511646658606384e-05, + "loss": 0.3728, + "step": 47270 + }, + { + "epoch": 3.280202622996322, + "grad_norm": 1.519537329673767, + "learning_rate": 1.6508554154757206e-05, + "loss": 0.3953, + "step": 47271 + }, + { + "epoch": 3.280272014433419, + "grad_norm": 0.8926810026168823, + "learning_rate": 1.6505461914480836e-05, + "loss": 0.2657, + "step": 47272 + }, + { + "epoch": 3.280341405870516, + "grad_norm": 1.279953956604004, + "learning_rate": 1.6502369937787034e-05, + "loss": 0.4062, + "step": 47273 + }, + { + "epoch": 3.280410797307612, + "grad_norm": 0.7647971510887146, + "learning_rate": 1.6499278224685565e-05, + "loss": 0.2418, + "step": 47274 + }, + { + "epoch": 3.280480188744709, + "grad_norm": 1.1003940105438232, + "learning_rate": 1.6496186775186184e-05, + "loss": 0.4215, + "step": 47275 + }, + { + "epoch": 3.2805495801818054, + "grad_norm": 1.3521521091461182, + "learning_rate": 1.649309558929867e-05, + "loss": 0.2977, + "step": 47276 + }, + { + "epoch": 3.2806189716189023, + "grad_norm": 1.049119234085083, + "learning_rate": 1.6490004667032755e-05, + "loss": 0.4028, + "step": 47277 + }, + { + "epoch": 3.280688363055999, + "grad_norm": 0.9230301976203918, + "learning_rate": 1.648691400839819e-05, + "loss": 0.2308, + "step": 47278 + }, + { + "epoch": 3.2807577544930955, + "grad_norm": 0.8096073269844055, + "learning_rate": 1.648382361340476e-05, + "loss": 0.2189, + "step": 47279 + }, + { + "epoch": 3.2808271459301923, + "grad_norm": 0.9123204350471497, + "learning_rate": 1.6480733482062205e-05, + "loss": 0.3116, + "step": 47280 + }, + { + "epoch": 3.2808965373672887, + "grad_norm": 0.8815705180168152, + "learning_rate": 1.647764361438029e-05, + "loss": 0.2848, + "step": 47281 + }, + { + "epoch": 3.2809659288043855, + "grad_norm": 0.7342422604560852, + "learning_rate": 1.6474554010368758e-05, + "loss": 0.2034, + "step": 47282 + }, + { + "epoch": 3.2810353202414824, + "grad_norm": 0.7256109714508057, + "learning_rate": 1.6471464670037395e-05, + "loss": 0.2636, + "step": 47283 + }, + { + "epoch": 3.2811047116785788, + "grad_norm": 0.8446228504180908, + "learning_rate": 1.6468375593395902e-05, + "loss": 0.2327, + "step": 47284 + }, + { + "epoch": 3.2811741031156756, + "grad_norm": 1.1906100511550903, + "learning_rate": 1.6465286780454048e-05, + "loss": 0.3642, + "step": 47285 + }, + { + "epoch": 3.281243494552772, + "grad_norm": 0.8497340083122253, + "learning_rate": 1.6462198231221605e-05, + "loss": 0.1883, + "step": 47286 + }, + { + "epoch": 3.281312885989869, + "grad_norm": 1.1692700386047363, + "learning_rate": 1.645910994570833e-05, + "loss": 0.2406, + "step": 47287 + }, + { + "epoch": 3.2813822774269656, + "grad_norm": 0.7724258303642273, + "learning_rate": 1.6456021923923925e-05, + "loss": 0.182, + "step": 47288 + }, + { + "epoch": 3.281451668864062, + "grad_norm": 1.062386393547058, + "learning_rate": 1.6452934165878165e-05, + "loss": 0.3612, + "step": 47289 + }, + { + "epoch": 3.281521060301159, + "grad_norm": 0.7448616623878479, + "learning_rate": 1.6449846671580805e-05, + "loss": 0.1521, + "step": 47290 + }, + { + "epoch": 3.2815904517382553, + "grad_norm": 0.8654547929763794, + "learning_rate": 1.6446759441041603e-05, + "loss": 0.2403, + "step": 47291 + }, + { + "epoch": 3.281659843175352, + "grad_norm": 0.6676221489906311, + "learning_rate": 1.6443672474270268e-05, + "loss": 0.2571, + "step": 47292 + }, + { + "epoch": 3.281729234612449, + "grad_norm": 0.8533626794815063, + "learning_rate": 1.6440585771276562e-05, + "loss": 0.1666, + "step": 47293 + }, + { + "epoch": 3.2817986260495453, + "grad_norm": 0.8135328888893127, + "learning_rate": 1.6437499332070228e-05, + "loss": 0.2084, + "step": 47294 + }, + { + "epoch": 3.281868017486642, + "grad_norm": 1.1160329580307007, + "learning_rate": 1.6434413156661022e-05, + "loss": 0.1816, + "step": 47295 + }, + { + "epoch": 3.281937408923739, + "grad_norm": 0.8777616620063782, + "learning_rate": 1.643132724505867e-05, + "loss": 0.4652, + "step": 47296 + }, + { + "epoch": 3.2820068003608354, + "grad_norm": 0.6229572892189026, + "learning_rate": 1.6428241597272952e-05, + "loss": 0.1779, + "step": 47297 + }, + { + "epoch": 3.282076191797932, + "grad_norm": 0.9483271241188049, + "learning_rate": 1.642515621331355e-05, + "loss": 0.311, + "step": 47298 + }, + { + "epoch": 3.282145583235029, + "grad_norm": 0.845139741897583, + "learning_rate": 1.642207109319024e-05, + "loss": 0.1739, + "step": 47299 + }, + { + "epoch": 3.2822149746721254, + "grad_norm": 0.8271388411521912, + "learning_rate": 1.641898623691275e-05, + "loss": 0.2244, + "step": 47300 + }, + { + "epoch": 3.2822843661092223, + "grad_norm": 1.1798694133758545, + "learning_rate": 1.6415901644490828e-05, + "loss": 0.2371, + "step": 47301 + }, + { + "epoch": 3.2823537575463186, + "grad_norm": 1.0532009601593018, + "learning_rate": 1.6412817315934202e-05, + "loss": 0.1988, + "step": 47302 + }, + { + "epoch": 3.2824231489834155, + "grad_norm": 1.1189756393432617, + "learning_rate": 1.6409733251252612e-05, + "loss": 0.3108, + "step": 47303 + }, + { + "epoch": 3.2824925404205123, + "grad_norm": 1.0019738674163818, + "learning_rate": 1.6406649450455825e-05, + "loss": 0.2321, + "step": 47304 + }, + { + "epoch": 3.2825619318576087, + "grad_norm": 1.133440613746643, + "learning_rate": 1.6403565913553522e-05, + "loss": 0.5624, + "step": 47305 + }, + { + "epoch": 3.2826313232947055, + "grad_norm": 0.9184807538986206, + "learning_rate": 1.640048264055546e-05, + "loss": 0.3847, + "step": 47306 + }, + { + "epoch": 3.282700714731802, + "grad_norm": 0.791851818561554, + "learning_rate": 1.6397399631471377e-05, + "loss": 0.1737, + "step": 47307 + }, + { + "epoch": 3.2827701061688987, + "grad_norm": 0.8481200337409973, + "learning_rate": 1.639431688631101e-05, + "loss": 0.1752, + "step": 47308 + }, + { + "epoch": 3.2828394976059956, + "grad_norm": 0.9439972043037415, + "learning_rate": 1.6391234405084077e-05, + "loss": 0.2227, + "step": 47309 + }, + { + "epoch": 3.282908889043092, + "grad_norm": 0.7790788412094116, + "learning_rate": 1.6388152187800322e-05, + "loss": 0.1588, + "step": 47310 + }, + { + "epoch": 3.282978280480189, + "grad_norm": 0.6133601665496826, + "learning_rate": 1.638507023446948e-05, + "loss": 0.1155, + "step": 47311 + }, + { + "epoch": 3.283047671917285, + "grad_norm": 1.245423436164856, + "learning_rate": 1.6381988545101257e-05, + "loss": 0.3366, + "step": 47312 + }, + { + "epoch": 3.283117063354382, + "grad_norm": 1.026817798614502, + "learning_rate": 1.63789071197054e-05, + "loss": 0.2521, + "step": 47313 + }, + { + "epoch": 3.283186454791479, + "grad_norm": 0.675100564956665, + "learning_rate": 1.6375825958291624e-05, + "loss": 0.1621, + "step": 47314 + }, + { + "epoch": 3.2832558462285752, + "grad_norm": 0.8535048961639404, + "learning_rate": 1.6372745060869664e-05, + "loss": 0.336, + "step": 47315 + }, + { + "epoch": 3.283325237665672, + "grad_norm": 0.7549222111701965, + "learning_rate": 1.636966442744924e-05, + "loss": 0.2705, + "step": 47316 + }, + { + "epoch": 3.2833946291027685, + "grad_norm": 0.5981837511062622, + "learning_rate": 1.636658405804009e-05, + "loss": 0.2064, + "step": 47317 + }, + { + "epoch": 3.2834640205398653, + "grad_norm": 1.3005461692810059, + "learning_rate": 1.6363503952651947e-05, + "loss": 0.3887, + "step": 47318 + }, + { + "epoch": 3.283533411976962, + "grad_norm": 1.3535600900650024, + "learning_rate": 1.6360424111294504e-05, + "loss": 0.1853, + "step": 47319 + }, + { + "epoch": 3.2836028034140585, + "grad_norm": 0.9312548041343689, + "learning_rate": 1.6357344533977492e-05, + "loss": 0.3236, + "step": 47320 + }, + { + "epoch": 3.2836721948511554, + "grad_norm": 0.8944690227508545, + "learning_rate": 1.635426522071064e-05, + "loss": 0.2826, + "step": 47321 + }, + { + "epoch": 3.283741586288252, + "grad_norm": 0.8913145661354065, + "learning_rate": 1.6351186171503696e-05, + "loss": 0.3026, + "step": 47322 + }, + { + "epoch": 3.2838109777253486, + "grad_norm": 1.6337305307388306, + "learning_rate": 1.634810738636633e-05, + "loss": 0.3148, + "step": 47323 + }, + { + "epoch": 3.2838803691624454, + "grad_norm": 0.9296249151229858, + "learning_rate": 1.6345028865308276e-05, + "loss": 0.2471, + "step": 47324 + }, + { + "epoch": 3.2839497605995422, + "grad_norm": 0.9859194755554199, + "learning_rate": 1.6341950608339263e-05, + "loss": 0.2431, + "step": 47325 + }, + { + "epoch": 3.2840191520366386, + "grad_norm": 1.213384747505188, + "learning_rate": 1.6338872615469024e-05, + "loss": 0.2344, + "step": 47326 + }, + { + "epoch": 3.2840885434737355, + "grad_norm": 0.847902238368988, + "learning_rate": 1.6335794886707246e-05, + "loss": 0.3016, + "step": 47327 + }, + { + "epoch": 3.284157934910832, + "grad_norm": 1.147966742515564, + "learning_rate": 1.633271742206365e-05, + "loss": 0.3053, + "step": 47328 + }, + { + "epoch": 3.2842273263479287, + "grad_norm": 0.7075480222702026, + "learning_rate": 1.632964022154796e-05, + "loss": 0.1537, + "step": 47329 + }, + { + "epoch": 3.2842967177850255, + "grad_norm": 0.9555512070655823, + "learning_rate": 1.6326563285169884e-05, + "loss": 0.3056, + "step": 47330 + }, + { + "epoch": 3.284366109222122, + "grad_norm": 0.8534666299819946, + "learning_rate": 1.632348661293914e-05, + "loss": 0.2061, + "step": 47331 + }, + { + "epoch": 3.2844355006592187, + "grad_norm": 0.5787025094032288, + "learning_rate": 1.6320410204865467e-05, + "loss": 0.2083, + "step": 47332 + }, + { + "epoch": 3.284504892096315, + "grad_norm": 0.8748605847358704, + "learning_rate": 1.6317334060958522e-05, + "loss": 0.2573, + "step": 47333 + }, + { + "epoch": 3.284574283533412, + "grad_norm": 0.8344401121139526, + "learning_rate": 1.6314258181228058e-05, + "loss": 0.1947, + "step": 47334 + }, + { + "epoch": 3.284643674970509, + "grad_norm": 0.86399906873703, + "learning_rate": 1.631118256568376e-05, + "loss": 0.3083, + "step": 47335 + }, + { + "epoch": 3.284713066407605, + "grad_norm": 0.9726697206497192, + "learning_rate": 1.6308107214335344e-05, + "loss": 0.2624, + "step": 47336 + }, + { + "epoch": 3.284782457844702, + "grad_norm": 0.842542290687561, + "learning_rate": 1.6305032127192532e-05, + "loss": 0.2687, + "step": 47337 + }, + { + "epoch": 3.2848518492817984, + "grad_norm": 1.0874857902526855, + "learning_rate": 1.6301957304265026e-05, + "loss": 0.2934, + "step": 47338 + }, + { + "epoch": 3.2849212407188952, + "grad_norm": 0.7157903909683228, + "learning_rate": 1.629888274556255e-05, + "loss": 0.2482, + "step": 47339 + }, + { + "epoch": 3.284990632155992, + "grad_norm": 1.1702743768692017, + "learning_rate": 1.6295808451094765e-05, + "loss": 0.2174, + "step": 47340 + }, + { + "epoch": 3.2850600235930885, + "grad_norm": 0.9960790276527405, + "learning_rate": 1.6292734420871404e-05, + "loss": 0.2843, + "step": 47341 + }, + { + "epoch": 3.2851294150301853, + "grad_norm": 1.0896885395050049, + "learning_rate": 1.6289660654902163e-05, + "loss": 0.2672, + "step": 47342 + }, + { + "epoch": 3.285198806467282, + "grad_norm": 0.8192414045333862, + "learning_rate": 1.6286587153196753e-05, + "loss": 0.3111, + "step": 47343 + }, + { + "epoch": 3.2852681979043785, + "grad_norm": 0.8054303526878357, + "learning_rate": 1.628351391576488e-05, + "loss": 0.2607, + "step": 47344 + }, + { + "epoch": 3.2853375893414754, + "grad_norm": 0.5708000063896179, + "learning_rate": 1.628044094261626e-05, + "loss": 0.0962, + "step": 47345 + }, + { + "epoch": 3.2854069807785717, + "grad_norm": 0.6710488200187683, + "learning_rate": 1.6277368233760548e-05, + "loss": 0.2456, + "step": 47346 + }, + { + "epoch": 3.2854763722156686, + "grad_norm": 0.8695286512374878, + "learning_rate": 1.6274295789207472e-05, + "loss": 0.2588, + "step": 47347 + }, + { + "epoch": 3.2855457636527654, + "grad_norm": 1.272840976715088, + "learning_rate": 1.6271223608966735e-05, + "loss": 0.3304, + "step": 47348 + }, + { + "epoch": 3.285615155089862, + "grad_norm": 0.7087864279747009, + "learning_rate": 1.626815169304803e-05, + "loss": 0.2523, + "step": 47349 + }, + { + "epoch": 3.2856845465269586, + "grad_norm": 1.6741044521331787, + "learning_rate": 1.6265080041461056e-05, + "loss": 0.3596, + "step": 47350 + }, + { + "epoch": 3.2857539379640555, + "grad_norm": 0.9651508331298828, + "learning_rate": 1.62620086542155e-05, + "loss": 0.2629, + "step": 47351 + }, + { + "epoch": 3.285823329401152, + "grad_norm": 1.0395406484603882, + "learning_rate": 1.6258937531321107e-05, + "loss": 0.3005, + "step": 47352 + }, + { + "epoch": 3.2858927208382487, + "grad_norm": 0.7423252463340759, + "learning_rate": 1.62558666727875e-05, + "loss": 0.2625, + "step": 47353 + }, + { + "epoch": 3.285962112275345, + "grad_norm": 0.6770381331443787, + "learning_rate": 1.62527960786244e-05, + "loss": 0.231, + "step": 47354 + }, + { + "epoch": 3.286031503712442, + "grad_norm": 0.9087145328521729, + "learning_rate": 1.6249725748841516e-05, + "loss": 0.2381, + "step": 47355 + }, + { + "epoch": 3.2861008951495387, + "grad_norm": 0.6760402917861938, + "learning_rate": 1.624665568344852e-05, + "loss": 0.1764, + "step": 47356 + }, + { + "epoch": 3.286170286586635, + "grad_norm": 0.9826154708862305, + "learning_rate": 1.6243585882455147e-05, + "loss": 0.2224, + "step": 47357 + }, + { + "epoch": 3.286239678023732, + "grad_norm": 0.7635176777839661, + "learning_rate": 1.624051634587104e-05, + "loss": 0.1984, + "step": 47358 + }, + { + "epoch": 3.2863090694608283, + "grad_norm": 0.9529001712799072, + "learning_rate": 1.623744707370589e-05, + "loss": 0.2346, + "step": 47359 + }, + { + "epoch": 3.286378460897925, + "grad_norm": 1.0590513944625854, + "learning_rate": 1.62343780659694e-05, + "loss": 0.2969, + "step": 47360 + }, + { + "epoch": 3.286447852335022, + "grad_norm": 0.8526033759117126, + "learning_rate": 1.6231309322671295e-05, + "loss": 0.2484, + "step": 47361 + }, + { + "epoch": 3.2865172437721184, + "grad_norm": 0.6806823015213013, + "learning_rate": 1.6228240843821196e-05, + "loss": 0.2089, + "step": 47362 + }, + { + "epoch": 3.2865866352092152, + "grad_norm": 0.9597861170768738, + "learning_rate": 1.622517262942883e-05, + "loss": 0.2812, + "step": 47363 + }, + { + "epoch": 3.2866560266463116, + "grad_norm": 0.8808082938194275, + "learning_rate": 1.622210467950387e-05, + "loss": 0.2239, + "step": 47364 + }, + { + "epoch": 3.2867254180834085, + "grad_norm": 0.7348169088363647, + "learning_rate": 1.6219036994056003e-05, + "loss": 0.1865, + "step": 47365 + }, + { + "epoch": 3.2867948095205053, + "grad_norm": 1.0999855995178223, + "learning_rate": 1.6215969573094937e-05, + "loss": 0.2133, + "step": 47366 + }, + { + "epoch": 3.2868642009576017, + "grad_norm": 0.827863335609436, + "learning_rate": 1.6212902416630317e-05, + "loss": 0.1282, + "step": 47367 + }, + { + "epoch": 3.2869335923946985, + "grad_norm": 0.9541298747062683, + "learning_rate": 1.6209835524671846e-05, + "loss": 0.2349, + "step": 47368 + }, + { + "epoch": 3.2870029838317953, + "grad_norm": 1.3060654401779175, + "learning_rate": 1.6206768897229195e-05, + "loss": 0.3247, + "step": 47369 + }, + { + "epoch": 3.2870723752688917, + "grad_norm": 2.135867118835449, + "learning_rate": 1.620370253431205e-05, + "loss": 0.1891, + "step": 47370 + }, + { + "epoch": 3.2871417667059886, + "grad_norm": 0.718883752822876, + "learning_rate": 1.6200636435930105e-05, + "loss": 0.2653, + "step": 47371 + }, + { + "epoch": 3.287211158143085, + "grad_norm": 1.1229760646820068, + "learning_rate": 1.6197570602093015e-05, + "loss": 0.3518, + "step": 47372 + }, + { + "epoch": 3.287280549580182, + "grad_norm": 0.7998185753822327, + "learning_rate": 1.61945050328105e-05, + "loss": 0.2135, + "step": 47373 + }, + { + "epoch": 3.2873499410172786, + "grad_norm": 1.0498123168945312, + "learning_rate": 1.6191439728092193e-05, + "loss": 0.3273, + "step": 47374 + }, + { + "epoch": 3.287419332454375, + "grad_norm": 1.1425882577896118, + "learning_rate": 1.6188374687947784e-05, + "loss": 0.2242, + "step": 47375 + }, + { + "epoch": 3.287488723891472, + "grad_norm": 1.2025288343429565, + "learning_rate": 1.6185309912386947e-05, + "loss": 0.2789, + "step": 47376 + }, + { + "epoch": 3.2875581153285687, + "grad_norm": 0.8220654129981995, + "learning_rate": 1.618224540141937e-05, + "loss": 0.2119, + "step": 47377 + }, + { + "epoch": 3.287627506765665, + "grad_norm": 0.7964504957199097, + "learning_rate": 1.617918115505471e-05, + "loss": 0.2149, + "step": 47378 + }, + { + "epoch": 3.287696898202762, + "grad_norm": 1.09116530418396, + "learning_rate": 1.6176117173302662e-05, + "loss": 0.3113, + "step": 47379 + }, + { + "epoch": 3.2877662896398583, + "grad_norm": 0.912175714969635, + "learning_rate": 1.6173053456172905e-05, + "loss": 0.309, + "step": 47380 + }, + { + "epoch": 3.287835681076955, + "grad_norm": 1.415733814239502, + "learning_rate": 1.616999000367506e-05, + "loss": 0.293, + "step": 47381 + }, + { + "epoch": 3.287905072514052, + "grad_norm": 1.2028743028640747, + "learning_rate": 1.6166926815818852e-05, + "loss": 0.273, + "step": 47382 + }, + { + "epoch": 3.2879744639511483, + "grad_norm": 0.6841464042663574, + "learning_rate": 1.6163863892613906e-05, + "loss": 0.1931, + "step": 47383 + }, + { + "epoch": 3.288043855388245, + "grad_norm": 0.705630362033844, + "learning_rate": 1.6160801234069934e-05, + "loss": 0.2862, + "step": 47384 + }, + { + "epoch": 3.2881132468253416, + "grad_norm": 0.7622392773628235, + "learning_rate": 1.6157738840196568e-05, + "loss": 0.2918, + "step": 47385 + }, + { + "epoch": 3.2881826382624384, + "grad_norm": 0.7321042418479919, + "learning_rate": 1.6154676711003502e-05, + "loss": 0.236, + "step": 47386 + }, + { + "epoch": 3.2882520296995352, + "grad_norm": 0.7163752317428589, + "learning_rate": 1.615161484650042e-05, + "loss": 0.1936, + "step": 47387 + }, + { + "epoch": 3.2883214211366316, + "grad_norm": 0.8041055798530579, + "learning_rate": 1.6148553246696928e-05, + "loss": 0.1399, + "step": 47388 + }, + { + "epoch": 3.2883908125737285, + "grad_norm": 0.7743467688560486, + "learning_rate": 1.6145491911602738e-05, + "loss": 0.2243, + "step": 47389 + }, + { + "epoch": 3.288460204010825, + "grad_norm": 0.6293184161186218, + "learning_rate": 1.6142430841227497e-05, + "loss": 0.2326, + "step": 47390 + }, + { + "epoch": 3.2885295954479217, + "grad_norm": 0.6844509243965149, + "learning_rate": 1.613937003558087e-05, + "loss": 0.1961, + "step": 47391 + }, + { + "epoch": 3.2885989868850185, + "grad_norm": 0.7348629236221313, + "learning_rate": 1.613630949467254e-05, + "loss": 0.1997, + "step": 47392 + }, + { + "epoch": 3.288668378322115, + "grad_norm": 0.7839466333389282, + "learning_rate": 1.6133249218512136e-05, + "loss": 0.2061, + "step": 47393 + }, + { + "epoch": 3.2887377697592117, + "grad_norm": 1.0153647661209106, + "learning_rate": 1.6130189207109337e-05, + "loss": 0.3448, + "step": 47394 + }, + { + "epoch": 3.2888071611963086, + "grad_norm": 0.7205612063407898, + "learning_rate": 1.6127129460473788e-05, + "loss": 0.2932, + "step": 47395 + }, + { + "epoch": 3.288876552633405, + "grad_norm": 0.7231001257896423, + "learning_rate": 1.6124069978615184e-05, + "loss": 0.2141, + "step": 47396 + }, + { + "epoch": 3.288945944070502, + "grad_norm": 1.1375404596328735, + "learning_rate": 1.6121010761543153e-05, + "loss": 0.2763, + "step": 47397 + }, + { + "epoch": 3.289015335507598, + "grad_norm": 0.6843751668930054, + "learning_rate": 1.611795180926734e-05, + "loss": 0.2038, + "step": 47398 + }, + { + "epoch": 3.289084726944695, + "grad_norm": 1.1141453981399536, + "learning_rate": 1.6114893121797437e-05, + "loss": 0.5944, + "step": 47399 + }, + { + "epoch": 3.289154118381792, + "grad_norm": 0.8952156901359558, + "learning_rate": 1.611183469914307e-05, + "loss": 0.2363, + "step": 47400 + }, + { + "epoch": 3.2892235098188882, + "grad_norm": 1.2621088027954102, + "learning_rate": 1.610877654131394e-05, + "loss": 0.2928, + "step": 47401 + }, + { + "epoch": 3.289292901255985, + "grad_norm": 1.0409035682678223, + "learning_rate": 1.6105718648319645e-05, + "loss": 0.3442, + "step": 47402 + }, + { + "epoch": 3.289362292693082, + "grad_norm": 1.5731892585754395, + "learning_rate": 1.610266102016986e-05, + "loss": 0.4007, + "step": 47403 + }, + { + "epoch": 3.2894316841301783, + "grad_norm": 1.049004316329956, + "learning_rate": 1.609960365687424e-05, + "loss": 0.2234, + "step": 47404 + }, + { + "epoch": 3.289501075567275, + "grad_norm": 1.938602328300476, + "learning_rate": 1.609654655844244e-05, + "loss": 0.3784, + "step": 47405 + }, + { + "epoch": 3.2895704670043715, + "grad_norm": 0.7529967427253723, + "learning_rate": 1.6093489724884102e-05, + "loss": 0.2204, + "step": 47406 + }, + { + "epoch": 3.2896398584414683, + "grad_norm": 1.301597237586975, + "learning_rate": 1.6090433156208896e-05, + "loss": 0.2522, + "step": 47407 + }, + { + "epoch": 3.289709249878565, + "grad_norm": 0.5350426435470581, + "learning_rate": 1.6087376852426472e-05, + "loss": 0.1363, + "step": 47408 + }, + { + "epoch": 3.2897786413156616, + "grad_norm": 0.8177952766418457, + "learning_rate": 1.6084320813546437e-05, + "loss": 0.2266, + "step": 47409 + }, + { + "epoch": 3.2898480327527584, + "grad_norm": 0.7810848355293274, + "learning_rate": 1.6081265039578464e-05, + "loss": 0.1912, + "step": 47410 + }, + { + "epoch": 3.2899174241898548, + "grad_norm": 0.8666752576828003, + "learning_rate": 1.6078209530532208e-05, + "loss": 0.3337, + "step": 47411 + }, + { + "epoch": 3.2899868156269516, + "grad_norm": 1.797858715057373, + "learning_rate": 1.6075154286417305e-05, + "loss": 0.3844, + "step": 47412 + }, + { + "epoch": 3.2900562070640484, + "grad_norm": 0.9586302042007446, + "learning_rate": 1.6072099307243392e-05, + "loss": 0.1513, + "step": 47413 + }, + { + "epoch": 3.290125598501145, + "grad_norm": 0.694650411605835, + "learning_rate": 1.6069044593020145e-05, + "loss": 0.2922, + "step": 47414 + }, + { + "epoch": 3.2901949899382417, + "grad_norm": 0.8749974370002747, + "learning_rate": 1.6065990143757193e-05, + "loss": 0.2164, + "step": 47415 + }, + { + "epoch": 3.290264381375338, + "grad_norm": 1.115276575088501, + "learning_rate": 1.6062935959464154e-05, + "loss": 0.3654, + "step": 47416 + }, + { + "epoch": 3.290333772812435, + "grad_norm": 1.6488604545593262, + "learning_rate": 1.6059882040150686e-05, + "loss": 0.4226, + "step": 47417 + }, + { + "epoch": 3.2904031642495317, + "grad_norm": 0.6646860837936401, + "learning_rate": 1.605682838582644e-05, + "loss": 0.1731, + "step": 47418 + }, + { + "epoch": 3.290472555686628, + "grad_norm": 0.9336001873016357, + "learning_rate": 1.6053774996501036e-05, + "loss": 0.3062, + "step": 47419 + }, + { + "epoch": 3.290541947123725, + "grad_norm": 0.7897732853889465, + "learning_rate": 1.6050721872184126e-05, + "loss": 0.2024, + "step": 47420 + }, + { + "epoch": 3.2906113385608218, + "grad_norm": 0.5897248983383179, + "learning_rate": 1.604766901288538e-05, + "loss": 0.1808, + "step": 47421 + }, + { + "epoch": 3.290680729997918, + "grad_norm": 0.942409336566925, + "learning_rate": 1.6044616418614378e-05, + "loss": 0.2473, + "step": 47422 + }, + { + "epoch": 3.290750121435015, + "grad_norm": 1.0114632844924927, + "learning_rate": 1.6041564089380777e-05, + "loss": 0.1887, + "step": 47423 + }, + { + "epoch": 3.290819512872112, + "grad_norm": 0.867686927318573, + "learning_rate": 1.6038512025194218e-05, + "loss": 0.2784, + "step": 47424 + }, + { + "epoch": 3.290888904309208, + "grad_norm": 0.6532585620880127, + "learning_rate": 1.6035460226064335e-05, + "loss": 0.1512, + "step": 47425 + }, + { + "epoch": 3.290958295746305, + "grad_norm": 0.853541910648346, + "learning_rate": 1.6032408692000755e-05, + "loss": 0.2259, + "step": 47426 + }, + { + "epoch": 3.2910276871834014, + "grad_norm": 0.6224337816238403, + "learning_rate": 1.6029357423013148e-05, + "loss": 0.2428, + "step": 47427 + }, + { + "epoch": 3.2910970786204983, + "grad_norm": 0.9598963260650635, + "learning_rate": 1.6026306419111102e-05, + "loss": 0.2145, + "step": 47428 + }, + { + "epoch": 3.291166470057595, + "grad_norm": 0.7032110095024109, + "learning_rate": 1.602325568030425e-05, + "loss": 0.2285, + "step": 47429 + }, + { + "epoch": 3.2912358614946915, + "grad_norm": 1.0298188924789429, + "learning_rate": 1.602020520660227e-05, + "loss": 0.2596, + "step": 47430 + }, + { + "epoch": 3.2913052529317883, + "grad_norm": 0.908759593963623, + "learning_rate": 1.6017154998014728e-05, + "loss": 0.3046, + "step": 47431 + }, + { + "epoch": 3.2913746443688847, + "grad_norm": 0.9002179503440857, + "learning_rate": 1.6014105054551287e-05, + "loss": 0.1692, + "step": 47432 + }, + { + "epoch": 3.2914440358059815, + "grad_norm": 0.8585768938064575, + "learning_rate": 1.6011055376221573e-05, + "loss": 0.3208, + "step": 47433 + }, + { + "epoch": 3.2915134272430784, + "grad_norm": 0.9535524249076843, + "learning_rate": 1.600800596303522e-05, + "loss": 0.211, + "step": 47434 + }, + { + "epoch": 3.2915828186801748, + "grad_norm": 1.1937716007232666, + "learning_rate": 1.600495681500186e-05, + "loss": 0.4159, + "step": 47435 + }, + { + "epoch": 3.2916522101172716, + "grad_norm": 1.1698538064956665, + "learning_rate": 1.6001907932131087e-05, + "loss": 0.2545, + "step": 47436 + }, + { + "epoch": 3.291721601554368, + "grad_norm": 0.9202385544776917, + "learning_rate": 1.5998859314432547e-05, + "loss": 0.1715, + "step": 47437 + }, + { + "epoch": 3.291790992991465, + "grad_norm": 0.95975261926651, + "learning_rate": 1.5995810961915857e-05, + "loss": 0.2982, + "step": 47438 + }, + { + "epoch": 3.2918603844285617, + "grad_norm": 0.6334643363952637, + "learning_rate": 1.599276287459065e-05, + "loss": 0.2218, + "step": 47439 + }, + { + "epoch": 3.291929775865658, + "grad_norm": 1.094986915588379, + "learning_rate": 1.5989715052466537e-05, + "loss": 0.2286, + "step": 47440 + }, + { + "epoch": 3.291999167302755, + "grad_norm": 0.8724699020385742, + "learning_rate": 1.5986667495553153e-05, + "loss": 0.2864, + "step": 47441 + }, + { + "epoch": 3.2920685587398513, + "grad_norm": 0.8839346766471863, + "learning_rate": 1.5983620203860138e-05, + "loss": 0.3265, + "step": 47442 + }, + { + "epoch": 3.292137950176948, + "grad_norm": 0.9202618598937988, + "learning_rate": 1.598057317739706e-05, + "loss": 0.1968, + "step": 47443 + }, + { + "epoch": 3.292207341614045, + "grad_norm": 0.856072723865509, + "learning_rate": 1.597752641617356e-05, + "loss": 0.2348, + "step": 47444 + }, + { + "epoch": 3.2922767330511413, + "grad_norm": 0.9940648674964905, + "learning_rate": 1.597447992019927e-05, + "loss": 0.2139, + "step": 47445 + }, + { + "epoch": 3.292346124488238, + "grad_norm": 0.6426418423652649, + "learning_rate": 1.5971433689483784e-05, + "loss": 0.1449, + "step": 47446 + }, + { + "epoch": 3.292415515925335, + "grad_norm": 0.9702524542808533, + "learning_rate": 1.596838772403675e-05, + "loss": 0.2098, + "step": 47447 + }, + { + "epoch": 3.2924849073624314, + "grad_norm": 1.0301316976547241, + "learning_rate": 1.5965342023867758e-05, + "loss": 0.1831, + "step": 47448 + }, + { + "epoch": 3.292554298799528, + "grad_norm": 1.0437144041061401, + "learning_rate": 1.5962296588986447e-05, + "loss": 0.2487, + "step": 47449 + }, + { + "epoch": 3.292623690236625, + "grad_norm": 1.0045149326324463, + "learning_rate": 1.5959251419402398e-05, + "loss": 0.2296, + "step": 47450 + }, + { + "epoch": 3.2926930816737214, + "grad_norm": 1.1585469245910645, + "learning_rate": 1.595620651512525e-05, + "loss": 0.3897, + "step": 47451 + }, + { + "epoch": 3.2927624731108183, + "grad_norm": 0.7454318404197693, + "learning_rate": 1.59531618761646e-05, + "loss": 0.3044, + "step": 47452 + }, + { + "epoch": 3.2928318645479147, + "grad_norm": 0.8016576766967773, + "learning_rate": 1.595011750253007e-05, + "loss": 0.1934, + "step": 47453 + }, + { + "epoch": 3.2929012559850115, + "grad_norm": 0.8911025524139404, + "learning_rate": 1.594707339423126e-05, + "loss": 0.2306, + "step": 47454 + }, + { + "epoch": 3.2929706474221083, + "grad_norm": 1.3808448314666748, + "learning_rate": 1.5944029551277795e-05, + "loss": 0.2085, + "step": 47455 + }, + { + "epoch": 3.2930400388592047, + "grad_norm": 0.8978317379951477, + "learning_rate": 1.594098597367929e-05, + "loss": 0.2821, + "step": 47456 + }, + { + "epoch": 3.2931094302963015, + "grad_norm": 1.0836803913116455, + "learning_rate": 1.5937942661445325e-05, + "loss": 0.2966, + "step": 47457 + }, + { + "epoch": 3.293178821733398, + "grad_norm": 0.5972420573234558, + "learning_rate": 1.593489961458552e-05, + "loss": 0.2073, + "step": 47458 + }, + { + "epoch": 3.2932482131704948, + "grad_norm": 1.691495656967163, + "learning_rate": 1.5931856833109483e-05, + "loss": 0.3381, + "step": 47459 + }, + { + "epoch": 3.2933176046075916, + "grad_norm": 1.0191582441329956, + "learning_rate": 1.5928814317026818e-05, + "loss": 0.3683, + "step": 47460 + }, + { + "epoch": 3.293386996044688, + "grad_norm": 1.3076972961425781, + "learning_rate": 1.5925772066347145e-05, + "loss": 0.3168, + "step": 47461 + }, + { + "epoch": 3.293456387481785, + "grad_norm": 0.8168329000473022, + "learning_rate": 1.5922730081080062e-05, + "loss": 0.1695, + "step": 47462 + }, + { + "epoch": 3.293525778918881, + "grad_norm": 1.2436879873275757, + "learning_rate": 1.591968836123515e-05, + "loss": 0.4367, + "step": 47463 + }, + { + "epoch": 3.293595170355978, + "grad_norm": 0.6438592672348022, + "learning_rate": 1.591664690682203e-05, + "loss": 0.2316, + "step": 47464 + }, + { + "epoch": 3.293664561793075, + "grad_norm": 1.02953040599823, + "learning_rate": 1.591360571785031e-05, + "loss": 0.2102, + "step": 47465 + }, + { + "epoch": 3.2937339532301713, + "grad_norm": 0.8812642097473145, + "learning_rate": 1.5910564794329576e-05, + "loss": 0.284, + "step": 47466 + }, + { + "epoch": 3.293803344667268, + "grad_norm": 1.4737861156463623, + "learning_rate": 1.5907524136269412e-05, + "loss": 0.2608, + "step": 47467 + }, + { + "epoch": 3.2938727361043645, + "grad_norm": 0.9569985866546631, + "learning_rate": 1.5904483743679454e-05, + "loss": 0.2104, + "step": 47468 + }, + { + "epoch": 3.2939421275414613, + "grad_norm": 0.83930504322052, + "learning_rate": 1.590144361656928e-05, + "loss": 0.2294, + "step": 47469 + }, + { + "epoch": 3.294011518978558, + "grad_norm": 1.1857385635375977, + "learning_rate": 1.589840375494851e-05, + "loss": 0.34, + "step": 47470 + }, + { + "epoch": 3.2940809104156545, + "grad_norm": 0.9240110516548157, + "learning_rate": 1.589536415882671e-05, + "loss": 0.3645, + "step": 47471 + }, + { + "epoch": 3.2941503018527514, + "grad_norm": 0.5343651175498962, + "learning_rate": 1.5892324828213477e-05, + "loss": 0.1393, + "step": 47472 + }, + { + "epoch": 3.294219693289848, + "grad_norm": 1.5046583414077759, + "learning_rate": 1.5889285763118423e-05, + "loss": 0.2666, + "step": 47473 + }, + { + "epoch": 3.2942890847269446, + "grad_norm": 0.8532124757766724, + "learning_rate": 1.588624696355112e-05, + "loss": 0.3546, + "step": 47474 + }, + { + "epoch": 3.2943584761640414, + "grad_norm": 1.0952061414718628, + "learning_rate": 1.588320842952119e-05, + "loss": 0.3169, + "step": 47475 + }, + { + "epoch": 3.2944278676011383, + "grad_norm": 0.9352558255195618, + "learning_rate": 1.588017016103821e-05, + "loss": 0.2248, + "step": 47476 + }, + { + "epoch": 3.2944972590382346, + "grad_norm": 1.0925534963607788, + "learning_rate": 1.5877132158111796e-05, + "loss": 0.3135, + "step": 47477 + }, + { + "epoch": 3.2945666504753315, + "grad_norm": 0.6368869543075562, + "learning_rate": 1.5874094420751495e-05, + "loss": 0.2481, + "step": 47478 + }, + { + "epoch": 3.294636041912428, + "grad_norm": 1.109520435333252, + "learning_rate": 1.5871056948966912e-05, + "loss": 0.3449, + "step": 47479 + }, + { + "epoch": 3.2947054333495247, + "grad_norm": 0.8585535883903503, + "learning_rate": 1.5868019742767637e-05, + "loss": 0.223, + "step": 47480 + }, + { + "epoch": 3.2947748247866215, + "grad_norm": 0.9604312777519226, + "learning_rate": 1.5864982802163276e-05, + "loss": 0.3351, + "step": 47481 + }, + { + "epoch": 3.294844216223718, + "grad_norm": 0.7920132875442505, + "learning_rate": 1.5861946127163397e-05, + "loss": 0.2381, + "step": 47482 + }, + { + "epoch": 3.2949136076608148, + "grad_norm": 0.821674108505249, + "learning_rate": 1.5858909717777593e-05, + "loss": 0.3762, + "step": 47483 + }, + { + "epoch": 3.294982999097911, + "grad_norm": 0.9017291069030762, + "learning_rate": 1.5855873574015466e-05, + "loss": 0.2604, + "step": 47484 + }, + { + "epoch": 3.295052390535008, + "grad_norm": 0.6704252362251282, + "learning_rate": 1.585283769588657e-05, + "loss": 0.1994, + "step": 47485 + }, + { + "epoch": 3.295121781972105, + "grad_norm": 1.2741528749465942, + "learning_rate": 1.5849802083400488e-05, + "loss": 0.2367, + "step": 47486 + }, + { + "epoch": 3.295191173409201, + "grad_norm": 0.8780615925788879, + "learning_rate": 1.5846766736566833e-05, + "loss": 0.1588, + "step": 47487 + }, + { + "epoch": 3.295260564846298, + "grad_norm": 0.8084625005722046, + "learning_rate": 1.5843731655395165e-05, + "loss": 0.2454, + "step": 47488 + }, + { + "epoch": 3.2953299562833944, + "grad_norm": 1.2261883020401, + "learning_rate": 1.5840696839895063e-05, + "loss": 0.2567, + "step": 47489 + }, + { + "epoch": 3.2953993477204913, + "grad_norm": 1.248337984085083, + "learning_rate": 1.5837662290076128e-05, + "loss": 0.2418, + "step": 47490 + }, + { + "epoch": 3.295468739157588, + "grad_norm": 1.114079236984253, + "learning_rate": 1.583462800594796e-05, + "loss": 0.2573, + "step": 47491 + }, + { + "epoch": 3.2955381305946845, + "grad_norm": 1.2180825471878052, + "learning_rate": 1.5831593987520067e-05, + "loss": 0.301, + "step": 47492 + }, + { + "epoch": 3.2956075220317813, + "grad_norm": 0.7419800162315369, + "learning_rate": 1.582856023480207e-05, + "loss": 0.207, + "step": 47493 + }, + { + "epoch": 3.2956769134688777, + "grad_norm": 0.8308727741241455, + "learning_rate": 1.5825526747803553e-05, + "loss": 0.2868, + "step": 47494 + }, + { + "epoch": 3.2957463049059745, + "grad_norm": 0.7064241766929626, + "learning_rate": 1.5822493526534078e-05, + "loss": 0.2417, + "step": 47495 + }, + { + "epoch": 3.2958156963430714, + "grad_norm": 0.9395121335983276, + "learning_rate": 1.5819460571003243e-05, + "loss": 0.2713, + "step": 47496 + }, + { + "epoch": 3.2958850877801678, + "grad_norm": 0.8133070468902588, + "learning_rate": 1.5816427881220573e-05, + "loss": 0.2338, + "step": 47497 + }, + { + "epoch": 3.2959544792172646, + "grad_norm": 1.0600684881210327, + "learning_rate": 1.581339545719569e-05, + "loss": 0.2244, + "step": 47498 + }, + { + "epoch": 3.2960238706543614, + "grad_norm": 0.9498022794723511, + "learning_rate": 1.5810363298938136e-05, + "loss": 0.2765, + "step": 47499 + }, + { + "epoch": 3.296093262091458, + "grad_norm": 0.7006793022155762, + "learning_rate": 1.5807331406457527e-05, + "loss": 0.2528, + "step": 47500 + }, + { + "epoch": 3.2961626535285546, + "grad_norm": 1.2241930961608887, + "learning_rate": 1.5804299779763387e-05, + "loss": 0.4262, + "step": 47501 + }, + { + "epoch": 3.2962320449656515, + "grad_norm": 1.4608416557312012, + "learning_rate": 1.58012684188653e-05, + "loss": 0.3765, + "step": 47502 + }, + { + "epoch": 3.296301436402748, + "grad_norm": 1.0812681913375854, + "learning_rate": 1.5798237323772837e-05, + "loss": 0.2191, + "step": 47503 + }, + { + "epoch": 3.2963708278398447, + "grad_norm": 1.3732445240020752, + "learning_rate": 1.5795206494495596e-05, + "loss": 0.3943, + "step": 47504 + }, + { + "epoch": 3.296440219276941, + "grad_norm": 1.3112512826919556, + "learning_rate": 1.5792175931043084e-05, + "loss": 0.4356, + "step": 47505 + }, + { + "epoch": 3.296509610714038, + "grad_norm": 0.9044344425201416, + "learning_rate": 1.5789145633424917e-05, + "loss": 0.3171, + "step": 47506 + }, + { + "epoch": 3.2965790021511348, + "grad_norm": 0.7275375127792358, + "learning_rate": 1.5786115601650642e-05, + "loss": 0.2094, + "step": 47507 + }, + { + "epoch": 3.296648393588231, + "grad_norm": 0.9509289264678955, + "learning_rate": 1.5783085835729828e-05, + "loss": 0.1937, + "step": 47508 + }, + { + "epoch": 3.296717785025328, + "grad_norm": 1.114924669265747, + "learning_rate": 1.5780056335672043e-05, + "loss": 0.2704, + "step": 47509 + }, + { + "epoch": 3.2967871764624244, + "grad_norm": 1.4415630102157593, + "learning_rate": 1.577702710148685e-05, + "loss": 0.351, + "step": 47510 + }, + { + "epoch": 3.296856567899521, + "grad_norm": 1.0735989809036255, + "learning_rate": 1.577399813318383e-05, + "loss": 0.2908, + "step": 47511 + }, + { + "epoch": 3.296925959336618, + "grad_norm": 0.7203556895256042, + "learning_rate": 1.5770969430772507e-05, + "loss": 0.2226, + "step": 47512 + }, + { + "epoch": 3.2969953507737144, + "grad_norm": 0.8872259259223938, + "learning_rate": 1.5767940994262454e-05, + "loss": 0.2203, + "step": 47513 + }, + { + "epoch": 3.2970647422108113, + "grad_norm": 0.8482672572135925, + "learning_rate": 1.5764912823663246e-05, + "loss": 0.3964, + "step": 47514 + }, + { + "epoch": 3.2971341336479076, + "grad_norm": 0.9643978476524353, + "learning_rate": 1.5761884918984437e-05, + "loss": 0.1957, + "step": 47515 + }, + { + "epoch": 3.2972035250850045, + "grad_norm": 0.5380303859710693, + "learning_rate": 1.5758857280235573e-05, + "loss": 0.166, + "step": 47516 + }, + { + "epoch": 3.2972729165221013, + "grad_norm": 1.255982756614685, + "learning_rate": 1.5755829907426232e-05, + "loss": 0.2038, + "step": 47517 + }, + { + "epoch": 3.2973423079591977, + "grad_norm": 0.8400503993034363, + "learning_rate": 1.575280280056598e-05, + "loss": 0.156, + "step": 47518 + }, + { + "epoch": 3.2974116993962945, + "grad_norm": 1.2287063598632812, + "learning_rate": 1.5749775959664336e-05, + "loss": 0.3019, + "step": 47519 + }, + { + "epoch": 3.297481090833391, + "grad_norm": 0.7986453771591187, + "learning_rate": 1.5746749384730876e-05, + "loss": 0.115, + "step": 47520 + }, + { + "epoch": 3.2975504822704877, + "grad_norm": 1.4267525672912598, + "learning_rate": 1.5743723075775153e-05, + "loss": 0.2866, + "step": 47521 + }, + { + "epoch": 3.2976198737075846, + "grad_norm": 0.7993435263633728, + "learning_rate": 1.5740697032806716e-05, + "loss": 0.2033, + "step": 47522 + }, + { + "epoch": 3.297689265144681, + "grad_norm": 0.6567320227622986, + "learning_rate": 1.5737671255835118e-05, + "loss": 0.1918, + "step": 47523 + }, + { + "epoch": 3.297758656581778, + "grad_norm": 1.237249732017517, + "learning_rate": 1.5734645744869925e-05, + "loss": 0.2615, + "step": 47524 + }, + { + "epoch": 3.2978280480188746, + "grad_norm": 1.5322437286376953, + "learning_rate": 1.5731620499920697e-05, + "loss": 0.3075, + "step": 47525 + }, + { + "epoch": 3.297897439455971, + "grad_norm": 0.7296623587608337, + "learning_rate": 1.572859552099695e-05, + "loss": 0.2479, + "step": 47526 + }, + { + "epoch": 3.297966830893068, + "grad_norm": 1.2358161211013794, + "learning_rate": 1.5725570808108238e-05, + "loss": 0.4452, + "step": 47527 + }, + { + "epoch": 3.2980362223301647, + "grad_norm": 1.0313043594360352, + "learning_rate": 1.5722546361264122e-05, + "loss": 0.2329, + "step": 47528 + }, + { + "epoch": 3.298105613767261, + "grad_norm": 1.0848034620285034, + "learning_rate": 1.5719522180474157e-05, + "loss": 0.2728, + "step": 47529 + }, + { + "epoch": 3.298175005204358, + "grad_norm": 1.121858835220337, + "learning_rate": 1.5716498265747882e-05, + "loss": 0.3502, + "step": 47530 + }, + { + "epoch": 3.2982443966414543, + "grad_norm": 1.418232798576355, + "learning_rate": 1.571347461709487e-05, + "loss": 0.4232, + "step": 47531 + }, + { + "epoch": 3.298313788078551, + "grad_norm": 0.9501292705535889, + "learning_rate": 1.5710451234524603e-05, + "loss": 0.3483, + "step": 47532 + }, + { + "epoch": 3.298383179515648, + "grad_norm": 1.0793557167053223, + "learning_rate": 1.5707428118046674e-05, + "loss": 0.2256, + "step": 47533 + }, + { + "epoch": 3.2984525709527444, + "grad_norm": 1.1251620054244995, + "learning_rate": 1.570440526767061e-05, + "loss": 0.2829, + "step": 47534 + }, + { + "epoch": 3.298521962389841, + "grad_norm": 0.9652982950210571, + "learning_rate": 1.5701382683405985e-05, + "loss": 0.2252, + "step": 47535 + }, + { + "epoch": 3.2985913538269376, + "grad_norm": 0.8568540215492249, + "learning_rate": 1.5698360365262287e-05, + "loss": 0.2399, + "step": 47536 + }, + { + "epoch": 3.2986607452640344, + "grad_norm": 1.1631413698196411, + "learning_rate": 1.5695338313249087e-05, + "loss": 0.3853, + "step": 47537 + }, + { + "epoch": 3.2987301367011312, + "grad_norm": 0.9255599975585938, + "learning_rate": 1.5692316527375916e-05, + "loss": 0.2285, + "step": 47538 + }, + { + "epoch": 3.2987995281382276, + "grad_norm": 0.6847506761550903, + "learning_rate": 1.568929500765235e-05, + "loss": 0.2908, + "step": 47539 + }, + { + "epoch": 3.2988689195753245, + "grad_norm": 0.856203019618988, + "learning_rate": 1.568627375408788e-05, + "loss": 0.2776, + "step": 47540 + }, + { + "epoch": 3.298938311012421, + "grad_norm": 1.4156272411346436, + "learning_rate": 1.5683252766692046e-05, + "loss": 0.3184, + "step": 47541 + }, + { + "epoch": 3.2990077024495177, + "grad_norm": 0.9014653563499451, + "learning_rate": 1.5680232045474418e-05, + "loss": 0.2464, + "step": 47542 + }, + { + "epoch": 3.2990770938866145, + "grad_norm": 0.6414216756820679, + "learning_rate": 1.567721159044451e-05, + "loss": 0.1641, + "step": 47543 + }, + { + "epoch": 3.299146485323711, + "grad_norm": 1.103865146636963, + "learning_rate": 1.5674191401611848e-05, + "loss": 0.2789, + "step": 47544 + }, + { + "epoch": 3.2992158767608077, + "grad_norm": 0.8618003129959106, + "learning_rate": 1.5671171478985992e-05, + "loss": 0.239, + "step": 47545 + }, + { + "epoch": 3.299285268197904, + "grad_norm": 1.2437032461166382, + "learning_rate": 1.5668151822576483e-05, + "loss": 0.3319, + "step": 47546 + }, + { + "epoch": 3.299354659635001, + "grad_norm": 1.2549331188201904, + "learning_rate": 1.5665132432392814e-05, + "loss": 0.2845, + "step": 47547 + }, + { + "epoch": 3.299424051072098, + "grad_norm": 0.8367325663566589, + "learning_rate": 1.5662113308444536e-05, + "loss": 0.2606, + "step": 47548 + }, + { + "epoch": 3.299493442509194, + "grad_norm": 1.4483922719955444, + "learning_rate": 1.5659094450741184e-05, + "loss": 0.2409, + "step": 47549 + }, + { + "epoch": 3.299562833946291, + "grad_norm": 0.9121901392936707, + "learning_rate": 1.5656075859292284e-05, + "loss": 0.2793, + "step": 47550 + }, + { + "epoch": 3.299632225383388, + "grad_norm": 0.9054064750671387, + "learning_rate": 1.5653057534107362e-05, + "loss": 0.2745, + "step": 47551 + }, + { + "epoch": 3.2997016168204842, + "grad_norm": 0.7407326698303223, + "learning_rate": 1.5650039475195954e-05, + "loss": 0.3732, + "step": 47552 + }, + { + "epoch": 3.299771008257581, + "grad_norm": 0.6874055862426758, + "learning_rate": 1.5647021682567608e-05, + "loss": 0.1025, + "step": 47553 + }, + { + "epoch": 3.299840399694678, + "grad_norm": 1.119324803352356, + "learning_rate": 1.5644004156231807e-05, + "loss": 0.233, + "step": 47554 + }, + { + "epoch": 3.2999097911317743, + "grad_norm": 0.994012713432312, + "learning_rate": 1.5640986896198097e-05, + "loss": 0.2135, + "step": 47555 + }, + { + "epoch": 3.299979182568871, + "grad_norm": 0.8314948081970215, + "learning_rate": 1.5637969902476014e-05, + "loss": 0.1925, + "step": 47556 + }, + { + "epoch": 3.3000485740059675, + "grad_norm": 0.786013662815094, + "learning_rate": 1.5634953175075053e-05, + "loss": 0.218, + "step": 47557 + }, + { + "epoch": 3.3001179654430643, + "grad_norm": 1.052644968032837, + "learning_rate": 1.563193671400477e-05, + "loss": 0.2172, + "step": 47558 + }, + { + "epoch": 3.300187356880161, + "grad_norm": 0.8652258515357971, + "learning_rate": 1.562892051927467e-05, + "loss": 0.2383, + "step": 47559 + }, + { + "epoch": 3.3002567483172576, + "grad_norm": 0.8215078711509705, + "learning_rate": 1.5625904590894304e-05, + "loss": 0.1958, + "step": 47560 + }, + { + "epoch": 3.3003261397543544, + "grad_norm": 0.8460332751274109, + "learning_rate": 1.5622888928873135e-05, + "loss": 0.2366, + "step": 47561 + }, + { + "epoch": 3.300395531191451, + "grad_norm": 1.077723741531372, + "learning_rate": 1.561987353322073e-05, + "loss": 0.2209, + "step": 47562 + }, + { + "epoch": 3.3004649226285476, + "grad_norm": 0.7175086140632629, + "learning_rate": 1.561685840394659e-05, + "loss": 0.2035, + "step": 47563 + }, + { + "epoch": 3.3005343140656445, + "grad_norm": 0.8831761479377747, + "learning_rate": 1.5613843541060236e-05, + "loss": 0.2492, + "step": 47564 + }, + { + "epoch": 3.300603705502741, + "grad_norm": 0.9177663326263428, + "learning_rate": 1.5610828944571186e-05, + "loss": 0.2302, + "step": 47565 + }, + { + "epoch": 3.3006730969398377, + "grad_norm": 1.0253779888153076, + "learning_rate": 1.560781461448898e-05, + "loss": 0.3079, + "step": 47566 + }, + { + "epoch": 3.300742488376934, + "grad_norm": 0.6811551451683044, + "learning_rate": 1.5604800550823096e-05, + "loss": 0.2449, + "step": 47567 + }, + { + "epoch": 3.300811879814031, + "grad_norm": 0.8329110145568848, + "learning_rate": 1.560178675358306e-05, + "loss": 0.3402, + "step": 47568 + }, + { + "epoch": 3.3008812712511277, + "grad_norm": 0.6608747839927673, + "learning_rate": 1.5598773222778386e-05, + "loss": 0.2021, + "step": 47569 + }, + { + "epoch": 3.300950662688224, + "grad_norm": 1.4184293746948242, + "learning_rate": 1.5595759958418622e-05, + "loss": 0.4571, + "step": 47570 + }, + { + "epoch": 3.301020054125321, + "grad_norm": 0.8235169053077698, + "learning_rate": 1.5592746960513227e-05, + "loss": 0.1628, + "step": 47571 + }, + { + "epoch": 3.301089445562418, + "grad_norm": 1.7433264255523682, + "learning_rate": 1.558973422907174e-05, + "loss": 0.3251, + "step": 47572 + }, + { + "epoch": 3.301158836999514, + "grad_norm": 1.0429478883743286, + "learning_rate": 1.5586721764103684e-05, + "loss": 0.4264, + "step": 47573 + }, + { + "epoch": 3.301228228436611, + "grad_norm": 1.252447247505188, + "learning_rate": 1.5583709565618542e-05, + "loss": 0.4096, + "step": 47574 + }, + { + "epoch": 3.3012976198737074, + "grad_norm": 0.8494365811347961, + "learning_rate": 1.558069763362583e-05, + "loss": 0.2178, + "step": 47575 + }, + { + "epoch": 3.3013670113108042, + "grad_norm": 0.7628585696220398, + "learning_rate": 1.5577685968135058e-05, + "loss": 0.2334, + "step": 47576 + }, + { + "epoch": 3.301436402747901, + "grad_norm": 0.7888601422309875, + "learning_rate": 1.5574674569155733e-05, + "loss": 0.3146, + "step": 47577 + }, + { + "epoch": 3.3015057941849975, + "grad_norm": 0.8085368275642395, + "learning_rate": 1.557166343669737e-05, + "loss": 0.271, + "step": 47578 + }, + { + "epoch": 3.3015751856220943, + "grad_norm": 0.978888988494873, + "learning_rate": 1.556865257076947e-05, + "loss": 0.3221, + "step": 47579 + }, + { + "epoch": 3.301644577059191, + "grad_norm": 0.7309520244598389, + "learning_rate": 1.5565641971381562e-05, + "loss": 0.1572, + "step": 47580 + }, + { + "epoch": 3.3017139684962875, + "grad_norm": 0.9691342711448669, + "learning_rate": 1.5562631638543103e-05, + "loss": 0.2402, + "step": 47581 + }, + { + "epoch": 3.3017833599333843, + "grad_norm": 1.0808343887329102, + "learning_rate": 1.5559621572263618e-05, + "loss": 0.3424, + "step": 47582 + }, + { + "epoch": 3.3018527513704807, + "grad_norm": 0.8697856068611145, + "learning_rate": 1.5556611772552608e-05, + "loss": 0.2037, + "step": 47583 + }, + { + "epoch": 3.3019221428075776, + "grad_norm": 1.1777836084365845, + "learning_rate": 1.5553602239419584e-05, + "loss": 0.4012, + "step": 47584 + }, + { + "epoch": 3.3019915342446744, + "grad_norm": 0.5975197553634644, + "learning_rate": 1.5550592972874045e-05, + "loss": 0.1867, + "step": 47585 + }, + { + "epoch": 3.302060925681771, + "grad_norm": 0.8689664602279663, + "learning_rate": 1.5547583972925472e-05, + "loss": 0.2106, + "step": 47586 + }, + { + "epoch": 3.3021303171188676, + "grad_norm": 0.8680180311203003, + "learning_rate": 1.5544575239583403e-05, + "loss": 0.3214, + "step": 47587 + }, + { + "epoch": 3.302199708555964, + "grad_norm": 1.0899547338485718, + "learning_rate": 1.5541566772857297e-05, + "loss": 0.2344, + "step": 47588 + }, + { + "epoch": 3.302269099993061, + "grad_norm": 0.7457356452941895, + "learning_rate": 1.5538558572756668e-05, + "loss": 0.1566, + "step": 47589 + }, + { + "epoch": 3.3023384914301577, + "grad_norm": 1.12045156955719, + "learning_rate": 1.5535550639290996e-05, + "loss": 0.3945, + "step": 47590 + }, + { + "epoch": 3.302407882867254, + "grad_norm": 0.8594370484352112, + "learning_rate": 1.5532542972469798e-05, + "loss": 0.1796, + "step": 47591 + }, + { + "epoch": 3.302477274304351, + "grad_norm": 1.0308384895324707, + "learning_rate": 1.5529535572302568e-05, + "loss": 0.2122, + "step": 47592 + }, + { + "epoch": 3.3025466657414473, + "grad_norm": 1.2409650087356567, + "learning_rate": 1.5526528438798782e-05, + "loss": 0.4064, + "step": 47593 + }, + { + "epoch": 3.302616057178544, + "grad_norm": 1.1961021423339844, + "learning_rate": 1.5523521571967982e-05, + "loss": 0.3009, + "step": 47594 + }, + { + "epoch": 3.302685448615641, + "grad_norm": 0.7998080849647522, + "learning_rate": 1.552051497181959e-05, + "loss": 0.3076, + "step": 47595 + }, + { + "epoch": 3.3027548400527373, + "grad_norm": 0.997846245765686, + "learning_rate": 1.5517508638363132e-05, + "loss": 0.2489, + "step": 47596 + }, + { + "epoch": 3.302824231489834, + "grad_norm": 0.9401440024375916, + "learning_rate": 1.5514502571608093e-05, + "loss": 0.2898, + "step": 47597 + }, + { + "epoch": 3.302893622926931, + "grad_norm": 0.8940221667289734, + "learning_rate": 1.5511496771563973e-05, + "loss": 0.2212, + "step": 47598 + }, + { + "epoch": 3.3029630143640274, + "grad_norm": 0.7593586444854736, + "learning_rate": 1.5508491238240252e-05, + "loss": 0.1993, + "step": 47599 + }, + { + "epoch": 3.3030324058011242, + "grad_norm": 1.2255899906158447, + "learning_rate": 1.5505485971646427e-05, + "loss": 0.4593, + "step": 47600 + }, + { + "epoch": 3.3031017972382206, + "grad_norm": 0.8921608924865723, + "learning_rate": 1.5502480971791998e-05, + "loss": 0.2764, + "step": 47601 + }, + { + "epoch": 3.3031711886753174, + "grad_norm": 1.01497483253479, + "learning_rate": 1.5499476238686418e-05, + "loss": 0.3034, + "step": 47602 + }, + { + "epoch": 3.3032405801124143, + "grad_norm": 1.1139638423919678, + "learning_rate": 1.5496471772339184e-05, + "loss": 0.2428, + "step": 47603 + }, + { + "epoch": 3.3033099715495107, + "grad_norm": 0.9397599697113037, + "learning_rate": 1.549346757275978e-05, + "loss": 0.2791, + "step": 47604 + }, + { + "epoch": 3.3033793629866075, + "grad_norm": 1.0129947662353516, + "learning_rate": 1.549046363995772e-05, + "loss": 0.4174, + "step": 47605 + }, + { + "epoch": 3.3034487544237043, + "grad_norm": 1.2216867208480835, + "learning_rate": 1.548745997394243e-05, + "loss": 0.2843, + "step": 47606 + }, + { + "epoch": 3.3035181458608007, + "grad_norm": 1.0898481607437134, + "learning_rate": 1.548445657472344e-05, + "loss": 0.2454, + "step": 47607 + }, + { + "epoch": 3.3035875372978976, + "grad_norm": 1.1699562072753906, + "learning_rate": 1.5481453442310223e-05, + "loss": 0.3062, + "step": 47608 + }, + { + "epoch": 3.303656928734994, + "grad_norm": 0.9574131369590759, + "learning_rate": 1.5478450576712232e-05, + "loss": 0.2091, + "step": 47609 + }, + { + "epoch": 3.303726320172091, + "grad_norm": 1.1144816875457764, + "learning_rate": 1.547544797793897e-05, + "loss": 0.3374, + "step": 47610 + }, + { + "epoch": 3.3037957116091876, + "grad_norm": 0.9456910490989685, + "learning_rate": 1.5472445645999924e-05, + "loss": 0.3476, + "step": 47611 + }, + { + "epoch": 3.303865103046284, + "grad_norm": 0.9881850481033325, + "learning_rate": 1.5469443580904542e-05, + "loss": 0.2648, + "step": 47612 + }, + { + "epoch": 3.303934494483381, + "grad_norm": 1.0447887182235718, + "learning_rate": 1.5466441782662323e-05, + "loss": 0.3126, + "step": 47613 + }, + { + "epoch": 3.3040038859204772, + "grad_norm": 0.5780842900276184, + "learning_rate": 1.5463440251282733e-05, + "loss": 0.1659, + "step": 47614 + }, + { + "epoch": 3.304073277357574, + "grad_norm": 1.080887794494629, + "learning_rate": 1.5460438986775293e-05, + "loss": 0.3106, + "step": 47615 + }, + { + "epoch": 3.304142668794671, + "grad_norm": 1.8491404056549072, + "learning_rate": 1.5457437989149402e-05, + "loss": 0.2993, + "step": 47616 + }, + { + "epoch": 3.3042120602317673, + "grad_norm": 0.9581944942474365, + "learning_rate": 1.5454437258414577e-05, + "loss": 0.3771, + "step": 47617 + }, + { + "epoch": 3.304281451668864, + "grad_norm": 1.275995135307312, + "learning_rate": 1.5451436794580287e-05, + "loss": 0.2531, + "step": 47618 + }, + { + "epoch": 3.3043508431059605, + "grad_norm": 0.6334418058395386, + "learning_rate": 1.5448436597656003e-05, + "loss": 0.1785, + "step": 47619 + }, + { + "epoch": 3.3044202345430573, + "grad_norm": 1.0649956464767456, + "learning_rate": 1.544543666765119e-05, + "loss": 0.2571, + "step": 47620 + }, + { + "epoch": 3.304489625980154, + "grad_norm": 0.7632439732551575, + "learning_rate": 1.5442437004575317e-05, + "loss": 0.2558, + "step": 47621 + }, + { + "epoch": 3.3045590174172506, + "grad_norm": 0.9991478323936462, + "learning_rate": 1.543943760843789e-05, + "loss": 0.2242, + "step": 47622 + }, + { + "epoch": 3.3046284088543474, + "grad_norm": 1.4998663663864136, + "learning_rate": 1.543643847924833e-05, + "loss": 0.2299, + "step": 47623 + }, + { + "epoch": 3.304697800291444, + "grad_norm": 0.9535290598869324, + "learning_rate": 1.543343961701611e-05, + "loss": 0.203, + "step": 47624 + }, + { + "epoch": 3.3047671917285406, + "grad_norm": 0.9159064292907715, + "learning_rate": 1.5430441021750718e-05, + "loss": 0.3632, + "step": 47625 + }, + { + "epoch": 3.3048365831656374, + "grad_norm": 0.8451260328292847, + "learning_rate": 1.5427442693461614e-05, + "loss": 0.2083, + "step": 47626 + }, + { + "epoch": 3.304905974602734, + "grad_norm": 1.3769350051879883, + "learning_rate": 1.5424444632158254e-05, + "loss": 0.257, + "step": 47627 + }, + { + "epoch": 3.3049753660398307, + "grad_norm": 0.9505053162574768, + "learning_rate": 1.5421446837850118e-05, + "loss": 0.2851, + "step": 47628 + }, + { + "epoch": 3.3050447574769275, + "grad_norm": 0.5930526852607727, + "learning_rate": 1.541844931054668e-05, + "loss": 0.1938, + "step": 47629 + }, + { + "epoch": 3.305114148914024, + "grad_norm": 1.0482392311096191, + "learning_rate": 1.5415452050257352e-05, + "loss": 0.2809, + "step": 47630 + }, + { + "epoch": 3.3051835403511207, + "grad_norm": 0.6482624411582947, + "learning_rate": 1.5412455056991638e-05, + "loss": 0.1893, + "step": 47631 + }, + { + "epoch": 3.3052529317882176, + "grad_norm": 0.7376284599304199, + "learning_rate": 1.5409458330758987e-05, + "loss": 0.2899, + "step": 47632 + }, + { + "epoch": 3.305322323225314, + "grad_norm": 0.5084466934204102, + "learning_rate": 1.5406461871568868e-05, + "loss": 0.1211, + "step": 47633 + }, + { + "epoch": 3.3053917146624108, + "grad_norm": 0.9109324812889099, + "learning_rate": 1.5403465679430728e-05, + "loss": 0.3116, + "step": 47634 + }, + { + "epoch": 3.305461106099507, + "grad_norm": 0.7975170016288757, + "learning_rate": 1.540046975435403e-05, + "loss": 0.1555, + "step": 47635 + }, + { + "epoch": 3.305530497536604, + "grad_norm": 1.1703256368637085, + "learning_rate": 1.5397474096348265e-05, + "loss": 0.3422, + "step": 47636 + }, + { + "epoch": 3.305599888973701, + "grad_norm": 0.9283720850944519, + "learning_rate": 1.539447870542283e-05, + "loss": 0.2792, + "step": 47637 + }, + { + "epoch": 3.305669280410797, + "grad_norm": 0.9311518669128418, + "learning_rate": 1.5391483581587208e-05, + "loss": 0.2372, + "step": 47638 + }, + { + "epoch": 3.305738671847894, + "grad_norm": 0.6718111038208008, + "learning_rate": 1.538848872485088e-05, + "loss": 0.1747, + "step": 47639 + }, + { + "epoch": 3.3058080632849904, + "grad_norm": 1.010690689086914, + "learning_rate": 1.5385494135223254e-05, + "loss": 0.4469, + "step": 47640 + }, + { + "epoch": 3.3058774547220873, + "grad_norm": 1.1808209419250488, + "learning_rate": 1.5382499812713803e-05, + "loss": 0.3663, + "step": 47641 + }, + { + "epoch": 3.305946846159184, + "grad_norm": 1.0209589004516602, + "learning_rate": 1.537950575733199e-05, + "loss": 0.2493, + "step": 47642 + }, + { + "epoch": 3.3060162375962805, + "grad_norm": 1.166403889656067, + "learning_rate": 1.5376511969087282e-05, + "loss": 0.339, + "step": 47643 + }, + { + "epoch": 3.3060856290333773, + "grad_norm": 0.9771887063980103, + "learning_rate": 1.537351844798908e-05, + "loss": 0.285, + "step": 47644 + }, + { + "epoch": 3.3061550204704737, + "grad_norm": 0.7459678053855896, + "learning_rate": 1.537052519404686e-05, + "loss": 0.2553, + "step": 47645 + }, + { + "epoch": 3.3062244119075705, + "grad_norm": 0.9172765016555786, + "learning_rate": 1.5367532207270076e-05, + "loss": 0.2905, + "step": 47646 + }, + { + "epoch": 3.3062938033446674, + "grad_norm": 0.9020079374313354, + "learning_rate": 1.5364539487668172e-05, + "loss": 0.3587, + "step": 47647 + }, + { + "epoch": 3.3063631947817638, + "grad_norm": 0.7640618681907654, + "learning_rate": 1.53615470352506e-05, + "loss": 0.2218, + "step": 47648 + }, + { + "epoch": 3.3064325862188606, + "grad_norm": 0.9177742004394531, + "learning_rate": 1.535855485002682e-05, + "loss": 0.3669, + "step": 47649 + }, + { + "epoch": 3.3065019776559574, + "grad_norm": 0.8452896475791931, + "learning_rate": 1.535556293200625e-05, + "loss": 0.2659, + "step": 47650 + }, + { + "epoch": 3.306571369093054, + "grad_norm": 0.9035723209381104, + "learning_rate": 1.5352571281198337e-05, + "loss": 0.2467, + "step": 47651 + }, + { + "epoch": 3.3066407605301507, + "grad_norm": 0.8682355880737305, + "learning_rate": 1.5349579897612543e-05, + "loss": 0.2491, + "step": 47652 + }, + { + "epoch": 3.3067101519672475, + "grad_norm": 0.7185527086257935, + "learning_rate": 1.5346588781258298e-05, + "loss": 0.2163, + "step": 47653 + }, + { + "epoch": 3.306779543404344, + "grad_norm": 0.8783391118049622, + "learning_rate": 1.534359793214506e-05, + "loss": 0.2408, + "step": 47654 + }, + { + "epoch": 3.3068489348414407, + "grad_norm": 0.9429614543914795, + "learning_rate": 1.5340607350282255e-05, + "loss": 0.2508, + "step": 47655 + }, + { + "epoch": 3.306918326278537, + "grad_norm": 0.9245887994766235, + "learning_rate": 1.5337617035679352e-05, + "loss": 0.19, + "step": 47656 + }, + { + "epoch": 3.306987717715634, + "grad_norm": 0.8808264136314392, + "learning_rate": 1.5334626988345746e-05, + "loss": 0.3915, + "step": 47657 + }, + { + "epoch": 3.3070571091527308, + "grad_norm": 1.1006529331207275, + "learning_rate": 1.5331637208290906e-05, + "loss": 0.3989, + "step": 47658 + }, + { + "epoch": 3.307126500589827, + "grad_norm": 0.9869014620780945, + "learning_rate": 1.5328647695524255e-05, + "loss": 0.2397, + "step": 47659 + }, + { + "epoch": 3.307195892026924, + "grad_norm": 0.9696701765060425, + "learning_rate": 1.5325658450055246e-05, + "loss": 0.2797, + "step": 47660 + }, + { + "epoch": 3.3072652834640204, + "grad_norm": 0.9512741565704346, + "learning_rate": 1.5322669471893314e-05, + "loss": 0.3614, + "step": 47661 + }, + { + "epoch": 3.307334674901117, + "grad_norm": 1.11836838722229, + "learning_rate": 1.5319680761047882e-05, + "loss": 0.2249, + "step": 47662 + }, + { + "epoch": 3.307404066338214, + "grad_norm": 0.9109033942222595, + "learning_rate": 1.5316692317528413e-05, + "loss": 0.2563, + "step": 47663 + }, + { + "epoch": 3.3074734577753104, + "grad_norm": 0.7934789061546326, + "learning_rate": 1.531370414134431e-05, + "loss": 0.2911, + "step": 47664 + }, + { + "epoch": 3.3075428492124073, + "grad_norm": 0.8465432524681091, + "learning_rate": 1.5310716232505007e-05, + "loss": 0.2518, + "step": 47665 + }, + { + "epoch": 3.3076122406495037, + "grad_norm": 1.0235847234725952, + "learning_rate": 1.5307728591019955e-05, + "loss": 0.3955, + "step": 47666 + }, + { + "epoch": 3.3076816320866005, + "grad_norm": 0.8671051263809204, + "learning_rate": 1.5304741216898566e-05, + "loss": 0.2425, + "step": 47667 + }, + { + "epoch": 3.3077510235236973, + "grad_norm": 0.9237731695175171, + "learning_rate": 1.530175411015029e-05, + "loss": 0.321, + "step": 47668 + }, + { + "epoch": 3.3078204149607937, + "grad_norm": 1.3566863536834717, + "learning_rate": 1.5298767270784554e-05, + "loss": 0.2351, + "step": 47669 + }, + { + "epoch": 3.3078898063978905, + "grad_norm": 1.1123746633529663, + "learning_rate": 1.5295780698810803e-05, + "loss": 0.2854, + "step": 47670 + }, + { + "epoch": 3.307959197834987, + "grad_norm": 0.9010574221611023, + "learning_rate": 1.5292794394238418e-05, + "loss": 0.225, + "step": 47671 + }, + { + "epoch": 3.3080285892720838, + "grad_norm": 0.9841448664665222, + "learning_rate": 1.528980835707685e-05, + "loss": 0.1898, + "step": 47672 + }, + { + "epoch": 3.3080979807091806, + "grad_norm": 0.8974496126174927, + "learning_rate": 1.5286822587335537e-05, + "loss": 0.2056, + "step": 47673 + }, + { + "epoch": 3.308167372146277, + "grad_norm": 0.7072232365608215, + "learning_rate": 1.5283837085023912e-05, + "loss": 0.1541, + "step": 47674 + }, + { + "epoch": 3.308236763583374, + "grad_norm": 0.9310945272445679, + "learning_rate": 1.5280851850151367e-05, + "loss": 0.2334, + "step": 47675 + }, + { + "epoch": 3.3083061550204707, + "grad_norm": 0.8985764980316162, + "learning_rate": 1.5277866882727344e-05, + "loss": 0.3105, + "step": 47676 + }, + { + "epoch": 3.308375546457567, + "grad_norm": 1.0128155946731567, + "learning_rate": 1.5274882182761262e-05, + "loss": 0.2846, + "step": 47677 + }, + { + "epoch": 3.308444937894664, + "grad_norm": 0.8752449750900269, + "learning_rate": 1.5271897750262578e-05, + "loss": 0.283, + "step": 47678 + }, + { + "epoch": 3.3085143293317607, + "grad_norm": 0.8764675855636597, + "learning_rate": 1.526891358524065e-05, + "loss": 0.3316, + "step": 47679 + }, + { + "epoch": 3.308583720768857, + "grad_norm": 1.1664767265319824, + "learning_rate": 1.5265929687704937e-05, + "loss": 0.3069, + "step": 47680 + }, + { + "epoch": 3.308653112205954, + "grad_norm": 0.9735689759254456, + "learning_rate": 1.526294605766485e-05, + "loss": 0.2777, + "step": 47681 + }, + { + "epoch": 3.3087225036430503, + "grad_norm": 0.9129835963249207, + "learning_rate": 1.5259962695129815e-05, + "loss": 0.2617, + "step": 47682 + }, + { + "epoch": 3.308791895080147, + "grad_norm": 0.7258065342903137, + "learning_rate": 1.5256979600109234e-05, + "loss": 0.169, + "step": 47683 + }, + { + "epoch": 3.308861286517244, + "grad_norm": 1.0298454761505127, + "learning_rate": 1.5253996772612577e-05, + "loss": 0.2804, + "step": 47684 + }, + { + "epoch": 3.3089306779543404, + "grad_norm": 0.9798367023468018, + "learning_rate": 1.5251014212649183e-05, + "loss": 0.332, + "step": 47685 + }, + { + "epoch": 3.309000069391437, + "grad_norm": 1.1070947647094727, + "learning_rate": 1.5248031920228512e-05, + "loss": 0.212, + "step": 47686 + }, + { + "epoch": 3.3090694608285336, + "grad_norm": 0.9614884257316589, + "learning_rate": 1.5245049895359964e-05, + "loss": 0.1875, + "step": 47687 + }, + { + "epoch": 3.3091388522656304, + "grad_norm": 0.7583927512168884, + "learning_rate": 1.5242068138052967e-05, + "loss": 0.2809, + "step": 47688 + }, + { + "epoch": 3.3092082437027273, + "grad_norm": 0.7675384283065796, + "learning_rate": 1.5239086648316925e-05, + "loss": 0.2089, + "step": 47689 + }, + { + "epoch": 3.3092776351398236, + "grad_norm": 0.9481327533721924, + "learning_rate": 1.5236105426161251e-05, + "loss": 0.2289, + "step": 47690 + }, + { + "epoch": 3.3093470265769205, + "grad_norm": 1.3140416145324707, + "learning_rate": 1.5233124471595384e-05, + "loss": 0.3176, + "step": 47691 + }, + { + "epoch": 3.309416418014017, + "grad_norm": 0.9694843292236328, + "learning_rate": 1.523014378462868e-05, + "loss": 0.2171, + "step": 47692 + }, + { + "epoch": 3.3094858094511137, + "grad_norm": 0.9648555517196655, + "learning_rate": 1.5227163365270581e-05, + "loss": 0.3105, + "step": 47693 + }, + { + "epoch": 3.3095552008882105, + "grad_norm": 0.8248463869094849, + "learning_rate": 1.5224183213530497e-05, + "loss": 0.2317, + "step": 47694 + }, + { + "epoch": 3.309624592325307, + "grad_norm": 0.7759970426559448, + "learning_rate": 1.522120332941782e-05, + "loss": 0.1856, + "step": 47695 + }, + { + "epoch": 3.3096939837624038, + "grad_norm": 0.9369803071022034, + "learning_rate": 1.521822371294197e-05, + "loss": 0.2228, + "step": 47696 + }, + { + "epoch": 3.3097633751995, + "grad_norm": 0.658896267414093, + "learning_rate": 1.5215244364112357e-05, + "loss": 0.1719, + "step": 47697 + }, + { + "epoch": 3.309832766636597, + "grad_norm": 0.7881823778152466, + "learning_rate": 1.5212265282938399e-05, + "loss": 0.1395, + "step": 47698 + }, + { + "epoch": 3.309902158073694, + "grad_norm": 0.8316985964775085, + "learning_rate": 1.5209286469429463e-05, + "loss": 0.1554, + "step": 47699 + }, + { + "epoch": 3.30997154951079, + "grad_norm": 0.8018701076507568, + "learning_rate": 1.5206307923594964e-05, + "loss": 0.1592, + "step": 47700 + }, + { + "epoch": 3.310040940947887, + "grad_norm": 1.6736825704574585, + "learning_rate": 1.520332964544432e-05, + "loss": 0.3486, + "step": 47701 + }, + { + "epoch": 3.310110332384984, + "grad_norm": 1.1642333269119263, + "learning_rate": 1.5200351634986931e-05, + "loss": 0.187, + "step": 47702 + }, + { + "epoch": 3.3101797238220803, + "grad_norm": 0.8171831369400024, + "learning_rate": 1.5197373892232181e-05, + "loss": 0.2116, + "step": 47703 + }, + { + "epoch": 3.310249115259177, + "grad_norm": 0.797413170337677, + "learning_rate": 1.5194396417189494e-05, + "loss": 0.2092, + "step": 47704 + }, + { + "epoch": 3.310318506696274, + "grad_norm": 0.8714427351951599, + "learning_rate": 1.5191419209868274e-05, + "loss": 0.3238, + "step": 47705 + }, + { + "epoch": 3.3103878981333703, + "grad_norm": 0.8422492146492004, + "learning_rate": 1.5188442270277881e-05, + "loss": 0.219, + "step": 47706 + }, + { + "epoch": 3.310457289570467, + "grad_norm": 0.7257623076438904, + "learning_rate": 1.5185465598427739e-05, + "loss": 0.1902, + "step": 47707 + }, + { + "epoch": 3.3105266810075635, + "grad_norm": 0.8262270092964172, + "learning_rate": 1.5182489194327243e-05, + "loss": 0.2053, + "step": 47708 + }, + { + "epoch": 3.3105960724446604, + "grad_norm": 0.8424351811408997, + "learning_rate": 1.5179513057985805e-05, + "loss": 0.2568, + "step": 47709 + }, + { + "epoch": 3.310665463881757, + "grad_norm": 0.9448636174201965, + "learning_rate": 1.5176537189412787e-05, + "loss": 0.3132, + "step": 47710 + }, + { + "epoch": 3.3107348553188536, + "grad_norm": 1.2437736988067627, + "learning_rate": 1.5173561588617601e-05, + "loss": 0.229, + "step": 47711 + }, + { + "epoch": 3.3108042467559504, + "grad_norm": 0.8517574667930603, + "learning_rate": 1.517058625560963e-05, + "loss": 0.2774, + "step": 47712 + }, + { + "epoch": 3.310873638193047, + "grad_norm": 1.6001468896865845, + "learning_rate": 1.5167611190398313e-05, + "loss": 0.223, + "step": 47713 + }, + { + "epoch": 3.3109430296301436, + "grad_norm": 1.9764114618301392, + "learning_rate": 1.516463639299297e-05, + "loss": 0.1339, + "step": 47714 + }, + { + "epoch": 3.3110124210672405, + "grad_norm": 1.0558327436447144, + "learning_rate": 1.5161661863403032e-05, + "loss": 0.4891, + "step": 47715 + }, + { + "epoch": 3.311081812504337, + "grad_norm": 1.376908540725708, + "learning_rate": 1.5158687601637877e-05, + "loss": 0.298, + "step": 47716 + }, + { + "epoch": 3.3111512039414337, + "grad_norm": 1.4358434677124023, + "learning_rate": 1.5155713607706912e-05, + "loss": 0.4304, + "step": 47717 + }, + { + "epoch": 3.31122059537853, + "grad_norm": 0.9635117053985596, + "learning_rate": 1.5152739881619515e-05, + "loss": 0.1966, + "step": 47718 + }, + { + "epoch": 3.311289986815627, + "grad_norm": 1.1248372793197632, + "learning_rate": 1.5149766423385093e-05, + "loss": 0.3053, + "step": 47719 + }, + { + "epoch": 3.3113593782527238, + "grad_norm": 1.1221989393234253, + "learning_rate": 1.514679323301299e-05, + "loss": 0.2949, + "step": 47720 + }, + { + "epoch": 3.31142876968982, + "grad_norm": 0.7156598567962646, + "learning_rate": 1.5143820310512612e-05, + "loss": 0.2544, + "step": 47721 + }, + { + "epoch": 3.311498161126917, + "grad_norm": 0.8737147450447083, + "learning_rate": 1.5140847655893353e-05, + "loss": 0.2709, + "step": 47722 + }, + { + "epoch": 3.3115675525640134, + "grad_norm": 0.5512111186981201, + "learning_rate": 1.5137875269164592e-05, + "loss": 0.1463, + "step": 47723 + }, + { + "epoch": 3.31163694400111, + "grad_norm": 1.053837537765503, + "learning_rate": 1.5134903150335722e-05, + "loss": 0.322, + "step": 47724 + }, + { + "epoch": 3.311706335438207, + "grad_norm": 0.9264988303184509, + "learning_rate": 1.5131931299416102e-05, + "loss": 0.1933, + "step": 47725 + }, + { + "epoch": 3.3117757268753034, + "grad_norm": 0.7658081650733948, + "learning_rate": 1.5128959716415159e-05, + "loss": 0.2177, + "step": 47726 + }, + { + "epoch": 3.3118451183124002, + "grad_norm": 0.7459264993667603, + "learning_rate": 1.5125988401342218e-05, + "loss": 0.1768, + "step": 47727 + }, + { + "epoch": 3.311914509749497, + "grad_norm": 0.8883646130561829, + "learning_rate": 1.5123017354206692e-05, + "loss": 0.2622, + "step": 47728 + }, + { + "epoch": 3.3119839011865935, + "grad_norm": 0.8995992541313171, + "learning_rate": 1.5120046575017944e-05, + "loss": 0.2847, + "step": 47729 + }, + { + "epoch": 3.3120532926236903, + "grad_norm": 1.0406354665756226, + "learning_rate": 1.5117076063785363e-05, + "loss": 0.241, + "step": 47730 + }, + { + "epoch": 3.312122684060787, + "grad_norm": 1.2012218236923218, + "learning_rate": 1.511410582051832e-05, + "loss": 0.1891, + "step": 47731 + }, + { + "epoch": 3.3121920754978835, + "grad_norm": 0.7874333262443542, + "learning_rate": 1.5111135845226232e-05, + "loss": 0.1464, + "step": 47732 + }, + { + "epoch": 3.3122614669349804, + "grad_norm": 1.252097249031067, + "learning_rate": 1.51081661379184e-05, + "loss": 0.3458, + "step": 47733 + }, + { + "epoch": 3.3123308583720767, + "grad_norm": 0.8382012844085693, + "learning_rate": 1.5105196698604252e-05, + "loss": 0.2137, + "step": 47734 + }, + { + "epoch": 3.3124002498091736, + "grad_norm": 0.8319385647773743, + "learning_rate": 1.5102227527293144e-05, + "loss": 0.2354, + "step": 47735 + }, + { + "epoch": 3.3124696412462704, + "grad_norm": 0.628624677658081, + "learning_rate": 1.509925862399445e-05, + "loss": 0.1932, + "step": 47736 + }, + { + "epoch": 3.312539032683367, + "grad_norm": 0.6409154534339905, + "learning_rate": 1.5096289988717549e-05, + "loss": 0.2597, + "step": 47737 + }, + { + "epoch": 3.3126084241204636, + "grad_norm": 0.8290517926216125, + "learning_rate": 1.5093321621471811e-05, + "loss": 0.4046, + "step": 47738 + }, + { + "epoch": 3.31267781555756, + "grad_norm": 0.7991857528686523, + "learning_rate": 1.509035352226662e-05, + "loss": 0.2401, + "step": 47739 + }, + { + "epoch": 3.312747206994657, + "grad_norm": 1.1825038194656372, + "learning_rate": 1.5087385691111322e-05, + "loss": 0.4284, + "step": 47740 + }, + { + "epoch": 3.3128165984317537, + "grad_norm": 1.2651244401931763, + "learning_rate": 1.5084418128015288e-05, + "loss": 0.2478, + "step": 47741 + }, + { + "epoch": 3.31288598986885, + "grad_norm": 0.9135139584541321, + "learning_rate": 1.5081450832987899e-05, + "loss": 0.2253, + "step": 47742 + }, + { + "epoch": 3.312955381305947, + "grad_norm": 0.7490145564079285, + "learning_rate": 1.5078483806038513e-05, + "loss": 0.2938, + "step": 47743 + }, + { + "epoch": 3.3130247727430433, + "grad_norm": 0.801927924156189, + "learning_rate": 1.5075517047176525e-05, + "loss": 0.3388, + "step": 47744 + }, + { + "epoch": 3.31309416418014, + "grad_norm": 0.8775134086608887, + "learning_rate": 1.5072550556411247e-05, + "loss": 0.2773, + "step": 47745 + }, + { + "epoch": 3.313163555617237, + "grad_norm": 0.9374135732650757, + "learning_rate": 1.5069584333752084e-05, + "loss": 0.2908, + "step": 47746 + }, + { + "epoch": 3.3132329470543334, + "grad_norm": 0.796051561832428, + "learning_rate": 1.5066618379208385e-05, + "loss": 0.182, + "step": 47747 + }, + { + "epoch": 3.31330233849143, + "grad_norm": 0.8906428217887878, + "learning_rate": 1.5063652692789543e-05, + "loss": 0.2838, + "step": 47748 + }, + { + "epoch": 3.3133717299285266, + "grad_norm": 0.8187409043312073, + "learning_rate": 1.506068727450487e-05, + "loss": 0.2505, + "step": 47749 + }, + { + "epoch": 3.3134411213656234, + "grad_norm": 1.1224569082260132, + "learning_rate": 1.5057722124363749e-05, + "loss": 0.1433, + "step": 47750 + }, + { + "epoch": 3.3135105128027202, + "grad_norm": 2.2404708862304688, + "learning_rate": 1.505475724237555e-05, + "loss": 0.3161, + "step": 47751 + }, + { + "epoch": 3.3135799042398166, + "grad_norm": 1.1013703346252441, + "learning_rate": 1.5051792628549621e-05, + "loss": 0.3527, + "step": 47752 + }, + { + "epoch": 3.3136492956769135, + "grad_norm": 0.8357861638069153, + "learning_rate": 1.5048828282895356e-05, + "loss": 0.2824, + "step": 47753 + }, + { + "epoch": 3.3137186871140103, + "grad_norm": 0.8568114638328552, + "learning_rate": 1.5045864205422067e-05, + "loss": 0.3364, + "step": 47754 + }, + { + "epoch": 3.3137880785511067, + "grad_norm": 1.1080081462860107, + "learning_rate": 1.5042900396139115e-05, + "loss": 0.36, + "step": 47755 + }, + { + "epoch": 3.3138574699882035, + "grad_norm": 1.0371806621551514, + "learning_rate": 1.5039936855055869e-05, + "loss": 0.3166, + "step": 47756 + }, + { + "epoch": 3.3139268614253004, + "grad_norm": 0.923646867275238, + "learning_rate": 1.503697358218169e-05, + "loss": 0.3213, + "step": 47757 + }, + { + "epoch": 3.3139962528623967, + "grad_norm": 0.8313905000686646, + "learning_rate": 1.5034010577525936e-05, + "loss": 0.1747, + "step": 47758 + }, + { + "epoch": 3.3140656442994936, + "grad_norm": 0.8522977828979492, + "learning_rate": 1.5031047841097946e-05, + "loss": 0.2228, + "step": 47759 + }, + { + "epoch": 3.31413503573659, + "grad_norm": 1.2177534103393555, + "learning_rate": 1.5028085372907097e-05, + "loss": 0.3628, + "step": 47760 + }, + { + "epoch": 3.314204427173687, + "grad_norm": 0.7976195812225342, + "learning_rate": 1.5025123172962707e-05, + "loss": 0.1894, + "step": 47761 + }, + { + "epoch": 3.3142738186107836, + "grad_norm": 0.7426366806030273, + "learning_rate": 1.5022161241274147e-05, + "loss": 0.2374, + "step": 47762 + }, + { + "epoch": 3.31434321004788, + "grad_norm": 1.1135830879211426, + "learning_rate": 1.5019199577850761e-05, + "loss": 0.3606, + "step": 47763 + }, + { + "epoch": 3.314412601484977, + "grad_norm": 0.7346505522727966, + "learning_rate": 1.5016238182701902e-05, + "loss": 0.2144, + "step": 47764 + }, + { + "epoch": 3.3144819929220732, + "grad_norm": 0.9445798397064209, + "learning_rate": 1.5013277055836916e-05, + "loss": 0.3252, + "step": 47765 + }, + { + "epoch": 3.31455138435917, + "grad_norm": 0.8709703087806702, + "learning_rate": 1.501031619726515e-05, + "loss": 0.2282, + "step": 47766 + }, + { + "epoch": 3.314620775796267, + "grad_norm": 0.9542971253395081, + "learning_rate": 1.5007355606995987e-05, + "loss": 0.2709, + "step": 47767 + }, + { + "epoch": 3.3146901672333633, + "grad_norm": 0.9586372971534729, + "learning_rate": 1.500439528503872e-05, + "loss": 0.2253, + "step": 47768 + }, + { + "epoch": 3.31475955867046, + "grad_norm": 0.7812029123306274, + "learning_rate": 1.5001435231402716e-05, + "loss": 0.2149, + "step": 47769 + }, + { + "epoch": 3.3148289501075565, + "grad_norm": 0.8481321930885315, + "learning_rate": 1.4998475446097315e-05, + "loss": 0.2412, + "step": 47770 + }, + { + "epoch": 3.3148983415446533, + "grad_norm": 1.0562348365783691, + "learning_rate": 1.4995515929131865e-05, + "loss": 0.2095, + "step": 47771 + }, + { + "epoch": 3.31496773298175, + "grad_norm": 1.0799636840820312, + "learning_rate": 1.4992556680515713e-05, + "loss": 0.2399, + "step": 47772 + }, + { + "epoch": 3.3150371244188466, + "grad_norm": 0.8075141310691833, + "learning_rate": 1.4989597700258185e-05, + "loss": 0.1843, + "step": 47773 + }, + { + "epoch": 3.3151065158559434, + "grad_norm": 1.075520634651184, + "learning_rate": 1.4986638988368674e-05, + "loss": 0.3762, + "step": 47774 + }, + { + "epoch": 3.31517590729304, + "grad_norm": 0.8879668116569519, + "learning_rate": 1.498368054485645e-05, + "loss": 0.3213, + "step": 47775 + }, + { + "epoch": 3.3152452987301366, + "grad_norm": 1.0452983379364014, + "learning_rate": 1.4980722369730882e-05, + "loss": 0.2755, + "step": 47776 + }, + { + "epoch": 3.3153146901672335, + "grad_norm": 1.0063886642456055, + "learning_rate": 1.497776446300131e-05, + "loss": 0.3654, + "step": 47777 + }, + { + "epoch": 3.31538408160433, + "grad_norm": 0.8702477812767029, + "learning_rate": 1.4974806824677069e-05, + "loss": 0.2287, + "step": 47778 + }, + { + "epoch": 3.3154534730414267, + "grad_norm": 1.1871916055679321, + "learning_rate": 1.4971849454767517e-05, + "loss": 0.3315, + "step": 47779 + }, + { + "epoch": 3.3155228644785235, + "grad_norm": 0.7904735207557678, + "learning_rate": 1.496889235328195e-05, + "loss": 0.2052, + "step": 47780 + }, + { + "epoch": 3.31559225591562, + "grad_norm": 1.0360934734344482, + "learning_rate": 1.4965935520229734e-05, + "loss": 0.3542, + "step": 47781 + }, + { + "epoch": 3.3156616473527167, + "grad_norm": 0.7757530808448792, + "learning_rate": 1.496297895562021e-05, + "loss": 0.301, + "step": 47782 + }, + { + "epoch": 3.3157310387898136, + "grad_norm": 0.8213497400283813, + "learning_rate": 1.4960022659462669e-05, + "loss": 0.2632, + "step": 47783 + }, + { + "epoch": 3.31580043022691, + "grad_norm": 1.0261739492416382, + "learning_rate": 1.4957066631766481e-05, + "loss": 0.2824, + "step": 47784 + }, + { + "epoch": 3.315869821664007, + "grad_norm": 1.4102210998535156, + "learning_rate": 1.495411087254096e-05, + "loss": 0.3095, + "step": 47785 + }, + { + "epoch": 3.315939213101103, + "grad_norm": 0.7323259711265564, + "learning_rate": 1.4951155381795446e-05, + "loss": 0.1743, + "step": 47786 + }, + { + "epoch": 3.3160086045382, + "grad_norm": 1.011082410812378, + "learning_rate": 1.494820015953926e-05, + "loss": 0.2839, + "step": 47787 + }, + { + "epoch": 3.316077995975297, + "grad_norm": 0.8373074531555176, + "learning_rate": 1.4945245205781766e-05, + "loss": 0.2608, + "step": 47788 + }, + { + "epoch": 3.3161473874123932, + "grad_norm": 0.976617157459259, + "learning_rate": 1.4942290520532243e-05, + "loss": 0.3068, + "step": 47789 + }, + { + "epoch": 3.31621677884949, + "grad_norm": 1.227466344833374, + "learning_rate": 1.4939336103800038e-05, + "loss": 0.4594, + "step": 47790 + }, + { + "epoch": 3.3162861702865865, + "grad_norm": 0.8932726979255676, + "learning_rate": 1.4936381955594481e-05, + "loss": 0.1961, + "step": 47791 + }, + { + "epoch": 3.3163555617236833, + "grad_norm": 1.4441664218902588, + "learning_rate": 1.4933428075924894e-05, + "loss": 0.3038, + "step": 47792 + }, + { + "epoch": 3.31642495316078, + "grad_norm": 1.1209789514541626, + "learning_rate": 1.4930474464800615e-05, + "loss": 0.2802, + "step": 47793 + }, + { + "epoch": 3.3164943445978765, + "grad_norm": 0.9870165586471558, + "learning_rate": 1.4927521122230947e-05, + "loss": 0.2388, + "step": 47794 + }, + { + "epoch": 3.3165637360349733, + "grad_norm": 1.196663737297058, + "learning_rate": 1.4924568048225252e-05, + "loss": 0.2679, + "step": 47795 + }, + { + "epoch": 3.3166331274720697, + "grad_norm": 0.9141815900802612, + "learning_rate": 1.4921615242792796e-05, + "loss": 0.4084, + "step": 47796 + }, + { + "epoch": 3.3167025189091666, + "grad_norm": 0.8281623721122742, + "learning_rate": 1.4918662705942943e-05, + "loss": 0.1801, + "step": 47797 + }, + { + "epoch": 3.3167719103462634, + "grad_norm": 1.4806638956069946, + "learning_rate": 1.4915710437684993e-05, + "loss": 0.2576, + "step": 47798 + }, + { + "epoch": 3.31684130178336, + "grad_norm": 0.9457499384880066, + "learning_rate": 1.4912758438028273e-05, + "loss": 0.3212, + "step": 47799 + }, + { + "epoch": 3.3169106932204566, + "grad_norm": 0.8393099904060364, + "learning_rate": 1.49098067069821e-05, + "loss": 0.1995, + "step": 47800 + }, + { + "epoch": 3.3169800846575535, + "grad_norm": 0.9307003617286682, + "learning_rate": 1.4906855244555794e-05, + "loss": 0.358, + "step": 47801 + }, + { + "epoch": 3.31704947609465, + "grad_norm": 0.7481175065040588, + "learning_rate": 1.4903904050758711e-05, + "loss": 0.1728, + "step": 47802 + }, + { + "epoch": 3.3171188675317467, + "grad_norm": 0.8680680990219116, + "learning_rate": 1.4900953125600092e-05, + "loss": 0.2277, + "step": 47803 + }, + { + "epoch": 3.317188258968843, + "grad_norm": 1.1638262271881104, + "learning_rate": 1.48980024690893e-05, + "loss": 0.2178, + "step": 47804 + }, + { + "epoch": 3.31725765040594, + "grad_norm": 0.7728496789932251, + "learning_rate": 1.4895052081235628e-05, + "loss": 0.2396, + "step": 47805 + }, + { + "epoch": 3.3173270418430367, + "grad_norm": 0.7073949575424194, + "learning_rate": 1.489210196204841e-05, + "loss": 0.2801, + "step": 47806 + }, + { + "epoch": 3.317396433280133, + "grad_norm": 0.9416795969009399, + "learning_rate": 1.4889152111536953e-05, + "loss": 0.2544, + "step": 47807 + }, + { + "epoch": 3.31746582471723, + "grad_norm": 0.9267206192016602, + "learning_rate": 1.488620252971059e-05, + "loss": 0.1818, + "step": 47808 + }, + { + "epoch": 3.317535216154327, + "grad_norm": 1.108321189880371, + "learning_rate": 1.4883253216578596e-05, + "loss": 0.2638, + "step": 47809 + }, + { + "epoch": 3.317604607591423, + "grad_norm": 1.0440744161605835, + "learning_rate": 1.4880304172150283e-05, + "loss": 0.295, + "step": 47810 + }, + { + "epoch": 3.31767399902852, + "grad_norm": 1.213809847831726, + "learning_rate": 1.4877355396434978e-05, + "loss": 0.2921, + "step": 47811 + }, + { + "epoch": 3.3177433904656164, + "grad_norm": 0.8506865501403809, + "learning_rate": 1.4874406889441994e-05, + "loss": 0.2441, + "step": 47812 + }, + { + "epoch": 3.3178127819027132, + "grad_norm": 0.9901101589202881, + "learning_rate": 1.4871458651180625e-05, + "loss": 0.2724, + "step": 47813 + }, + { + "epoch": 3.31788217333981, + "grad_norm": 0.816753089427948, + "learning_rate": 1.4868510681660197e-05, + "loss": 0.2286, + "step": 47814 + }, + { + "epoch": 3.3179515647769064, + "grad_norm": 0.827161967754364, + "learning_rate": 1.486556298088999e-05, + "loss": 0.2349, + "step": 47815 + }, + { + "epoch": 3.3180209562140033, + "grad_norm": 0.8502107262611389, + "learning_rate": 1.4862615548879334e-05, + "loss": 0.1447, + "step": 47816 + }, + { + "epoch": 3.3180903476510997, + "grad_norm": 0.8401237726211548, + "learning_rate": 1.4859668385637526e-05, + "loss": 0.2284, + "step": 47817 + }, + { + "epoch": 3.3181597390881965, + "grad_norm": 1.190549612045288, + "learning_rate": 1.4856721491173852e-05, + "loss": 0.2426, + "step": 47818 + }, + { + "epoch": 3.3182291305252933, + "grad_norm": 1.1349400281906128, + "learning_rate": 1.4853774865497638e-05, + "loss": 0.1916, + "step": 47819 + }, + { + "epoch": 3.3182985219623897, + "grad_norm": 1.213667869567871, + "learning_rate": 1.4850828508618164e-05, + "loss": 0.2843, + "step": 47820 + }, + { + "epoch": 3.3183679133994866, + "grad_norm": 1.041995882987976, + "learning_rate": 1.4847882420544756e-05, + "loss": 0.2777, + "step": 47821 + }, + { + "epoch": 3.318437304836583, + "grad_norm": 0.9630112051963806, + "learning_rate": 1.484493660128673e-05, + "loss": 0.1935, + "step": 47822 + }, + { + "epoch": 3.31850669627368, + "grad_norm": 1.1442903280258179, + "learning_rate": 1.484199105085332e-05, + "loss": 0.2291, + "step": 47823 + }, + { + "epoch": 3.3185760877107766, + "grad_norm": 1.1437253952026367, + "learning_rate": 1.4839045769253879e-05, + "loss": 0.3409, + "step": 47824 + }, + { + "epoch": 3.318645479147873, + "grad_norm": 0.8585578799247742, + "learning_rate": 1.4836100756497683e-05, + "loss": 0.2379, + "step": 47825 + }, + { + "epoch": 3.31871487058497, + "grad_norm": 1.0498727560043335, + "learning_rate": 1.483315601259404e-05, + "loss": 0.2878, + "step": 47826 + }, + { + "epoch": 3.3187842620220667, + "grad_norm": 0.8716328740119934, + "learning_rate": 1.483021153755224e-05, + "loss": 0.332, + "step": 47827 + }, + { + "epoch": 3.318853653459163, + "grad_norm": 1.032972812652588, + "learning_rate": 1.4827267331381579e-05, + "loss": 0.3253, + "step": 47828 + }, + { + "epoch": 3.31892304489626, + "grad_norm": 1.0563037395477295, + "learning_rate": 1.4824323394091377e-05, + "loss": 0.3979, + "step": 47829 + }, + { + "epoch": 3.3189924363333563, + "grad_norm": 1.0091520547866821, + "learning_rate": 1.4821379725690876e-05, + "loss": 0.3112, + "step": 47830 + }, + { + "epoch": 3.319061827770453, + "grad_norm": 1.5029494762420654, + "learning_rate": 1.481843632618941e-05, + "loss": 0.3307, + "step": 47831 + }, + { + "epoch": 3.31913121920755, + "grad_norm": 1.1358510255813599, + "learning_rate": 1.4815493195596242e-05, + "loss": 0.333, + "step": 47832 + }, + { + "epoch": 3.3192006106446463, + "grad_norm": 1.125294804573059, + "learning_rate": 1.4812550333920694e-05, + "loss": 0.2397, + "step": 47833 + }, + { + "epoch": 3.319270002081743, + "grad_norm": 0.8449006676673889, + "learning_rate": 1.4809607741172027e-05, + "loss": 0.2082, + "step": 47834 + }, + { + "epoch": 3.31933939351884, + "grad_norm": 0.8369007110595703, + "learning_rate": 1.4806665417359555e-05, + "loss": 0.2536, + "step": 47835 + }, + { + "epoch": 3.3194087849559364, + "grad_norm": 0.801020622253418, + "learning_rate": 1.4803723362492573e-05, + "loss": 0.2294, + "step": 47836 + }, + { + "epoch": 3.319478176393033, + "grad_norm": 0.615757167339325, + "learning_rate": 1.4800781576580335e-05, + "loss": 0.184, + "step": 47837 + }, + { + "epoch": 3.3195475678301296, + "grad_norm": 0.7417094111442566, + "learning_rate": 1.479784005963214e-05, + "loss": 0.207, + "step": 47838 + }, + { + "epoch": 3.3196169592672264, + "grad_norm": 0.8246883749961853, + "learning_rate": 1.4794898811657276e-05, + "loss": 0.2448, + "step": 47839 + }, + { + "epoch": 3.3196863507043233, + "grad_norm": 0.9294000864028931, + "learning_rate": 1.4791957832665038e-05, + "loss": 0.324, + "step": 47840 + }, + { + "epoch": 3.3197557421414197, + "grad_norm": 0.6934851408004761, + "learning_rate": 1.4789017122664694e-05, + "loss": 0.2267, + "step": 47841 + }, + { + "epoch": 3.3198251335785165, + "grad_norm": 0.6606870293617249, + "learning_rate": 1.478607668166554e-05, + "loss": 0.1861, + "step": 47842 + }, + { + "epoch": 3.319894525015613, + "grad_norm": 0.40457046031951904, + "learning_rate": 1.4783136509676886e-05, + "loss": 0.094, + "step": 47843 + }, + { + "epoch": 3.3199639164527097, + "grad_norm": 1.223668098449707, + "learning_rate": 1.4780196606707952e-05, + "loss": 0.4472, + "step": 47844 + }, + { + "epoch": 3.3200333078898066, + "grad_norm": 0.6762481927871704, + "learning_rate": 1.477725697276805e-05, + "loss": 0.2315, + "step": 47845 + }, + { + "epoch": 3.320102699326903, + "grad_norm": 0.7848085761070251, + "learning_rate": 1.477431760786646e-05, + "loss": 0.1766, + "step": 47846 + }, + { + "epoch": 3.3201720907639998, + "grad_norm": 1.3362962007522583, + "learning_rate": 1.4771378512012469e-05, + "loss": 0.3174, + "step": 47847 + }, + { + "epoch": 3.320241482201096, + "grad_norm": 0.6276819705963135, + "learning_rate": 1.4768439685215362e-05, + "loss": 0.2246, + "step": 47848 + }, + { + "epoch": 3.320310873638193, + "grad_norm": 1.550557255744934, + "learning_rate": 1.4765501127484382e-05, + "loss": 0.1986, + "step": 47849 + }, + { + "epoch": 3.32038026507529, + "grad_norm": 0.7766384482383728, + "learning_rate": 1.4762562838828831e-05, + "loss": 0.2333, + "step": 47850 + }, + { + "epoch": 3.320449656512386, + "grad_norm": 0.7240954041481018, + "learning_rate": 1.4759624819257967e-05, + "loss": 0.2691, + "step": 47851 + }, + { + "epoch": 3.320519047949483, + "grad_norm": 1.1562882661819458, + "learning_rate": 1.4756687068781105e-05, + "loss": 0.2041, + "step": 47852 + }, + { + "epoch": 3.32058843938658, + "grad_norm": 0.6951514482498169, + "learning_rate": 1.4753749587407472e-05, + "loss": 0.152, + "step": 47853 + }, + { + "epoch": 3.3206578308236763, + "grad_norm": 0.6172339916229248, + "learning_rate": 1.4750812375146372e-05, + "loss": 0.1336, + "step": 47854 + }, + { + "epoch": 3.320727222260773, + "grad_norm": 0.675794243812561, + "learning_rate": 1.4747875432007052e-05, + "loss": 0.1799, + "step": 47855 + }, + { + "epoch": 3.3207966136978695, + "grad_norm": 1.242720603942871, + "learning_rate": 1.4744938757998806e-05, + "loss": 0.3345, + "step": 47856 + }, + { + "epoch": 3.3208660051349663, + "grad_norm": 0.6763639450073242, + "learning_rate": 1.4742002353130913e-05, + "loss": 0.2604, + "step": 47857 + }, + { + "epoch": 3.320935396572063, + "grad_norm": 1.468195915222168, + "learning_rate": 1.4739066217412612e-05, + "loss": 0.2746, + "step": 47858 + }, + { + "epoch": 3.3210047880091595, + "grad_norm": 1.251207947731018, + "learning_rate": 1.4736130350853173e-05, + "loss": 0.3093, + "step": 47859 + }, + { + "epoch": 3.3210741794462564, + "grad_norm": 0.9374096393585205, + "learning_rate": 1.4733194753461887e-05, + "loss": 0.262, + "step": 47860 + }, + { + "epoch": 3.321143570883353, + "grad_norm": 1.4295395612716675, + "learning_rate": 1.4730259425248017e-05, + "loss": 0.3564, + "step": 47861 + }, + { + "epoch": 3.3212129623204496, + "grad_norm": 0.8722303509712219, + "learning_rate": 1.4727324366220818e-05, + "loss": 0.2951, + "step": 47862 + }, + { + "epoch": 3.3212823537575464, + "grad_norm": 0.9174544811248779, + "learning_rate": 1.4724389576389563e-05, + "loss": 0.1086, + "step": 47863 + }, + { + "epoch": 3.321351745194643, + "grad_norm": 0.921746015548706, + "learning_rate": 1.4721455055763545e-05, + "loss": 0.2448, + "step": 47864 + }, + { + "epoch": 3.3214211366317397, + "grad_norm": 1.054265022277832, + "learning_rate": 1.4718520804351976e-05, + "loss": 0.2243, + "step": 47865 + }, + { + "epoch": 3.3214905280688365, + "grad_norm": 0.8914081454277039, + "learning_rate": 1.471558682216413e-05, + "loss": 0.2529, + "step": 47866 + }, + { + "epoch": 3.321559919505933, + "grad_norm": 0.6949490904808044, + "learning_rate": 1.4712653109209295e-05, + "loss": 0.1359, + "step": 47867 + }, + { + "epoch": 3.3216293109430297, + "grad_norm": 0.7641871571540833, + "learning_rate": 1.4709719665496713e-05, + "loss": 0.1889, + "step": 47868 + }, + { + "epoch": 3.321698702380126, + "grad_norm": 0.6525749564170837, + "learning_rate": 1.4706786491035651e-05, + "loss": 0.1454, + "step": 47869 + }, + { + "epoch": 3.321768093817223, + "grad_norm": 1.1297451257705688, + "learning_rate": 1.470385358583537e-05, + "loss": 0.2844, + "step": 47870 + }, + { + "epoch": 3.3218374852543198, + "grad_norm": 0.9463005661964417, + "learning_rate": 1.4700920949905151e-05, + "loss": 0.2425, + "step": 47871 + }, + { + "epoch": 3.321906876691416, + "grad_norm": 0.9153293371200562, + "learning_rate": 1.4697988583254197e-05, + "loss": 0.2407, + "step": 47872 + }, + { + "epoch": 3.321976268128513, + "grad_norm": 1.0871713161468506, + "learning_rate": 1.4695056485891811e-05, + "loss": 0.3158, + "step": 47873 + }, + { + "epoch": 3.3220456595656094, + "grad_norm": 0.9593184590339661, + "learning_rate": 1.4692124657827222e-05, + "loss": 0.1888, + "step": 47874 + }, + { + "epoch": 3.322115051002706, + "grad_norm": 0.7868916988372803, + "learning_rate": 1.4689193099069698e-05, + "loss": 0.1863, + "step": 47875 + }, + { + "epoch": 3.322184442439803, + "grad_norm": 0.8332489132881165, + "learning_rate": 1.46862618096285e-05, + "loss": 0.2462, + "step": 47876 + }, + { + "epoch": 3.3222538338768994, + "grad_norm": 0.9358068704605103, + "learning_rate": 1.4683330789512872e-05, + "loss": 0.252, + "step": 47877 + }, + { + "epoch": 3.3223232253139963, + "grad_norm": 1.0774985551834106, + "learning_rate": 1.468040003873209e-05, + "loss": 0.4262, + "step": 47878 + }, + { + "epoch": 3.322392616751093, + "grad_norm": 1.4987974166870117, + "learning_rate": 1.4677469557295365e-05, + "loss": 0.4975, + "step": 47879 + }, + { + "epoch": 3.3224620081881895, + "grad_norm": 0.9122974872589111, + "learning_rate": 1.4674539345211979e-05, + "loss": 0.3749, + "step": 47880 + }, + { + "epoch": 3.3225313996252863, + "grad_norm": 0.9270438551902771, + "learning_rate": 1.4671609402491159e-05, + "loss": 0.3072, + "step": 47881 + }, + { + "epoch": 3.3226007910623827, + "grad_norm": 1.1050394773483276, + "learning_rate": 1.4668679729142176e-05, + "loss": 0.259, + "step": 47882 + }, + { + "epoch": 3.3226701824994795, + "grad_norm": 0.6790294051170349, + "learning_rate": 1.4665750325174288e-05, + "loss": 0.2011, + "step": 47883 + }, + { + "epoch": 3.3227395739365764, + "grad_norm": 1.5007034540176392, + "learning_rate": 1.466282119059671e-05, + "loss": 0.2164, + "step": 47884 + }, + { + "epoch": 3.3228089653736728, + "grad_norm": 1.0532809495925903, + "learning_rate": 1.4659892325418701e-05, + "loss": 0.3833, + "step": 47885 + }, + { + "epoch": 3.3228783568107696, + "grad_norm": 0.8329757452011108, + "learning_rate": 1.465696372964951e-05, + "loss": 0.1741, + "step": 47886 + }, + { + "epoch": 3.3229477482478664, + "grad_norm": 0.9579599499702454, + "learning_rate": 1.4654035403298405e-05, + "loss": 0.2667, + "step": 47887 + }, + { + "epoch": 3.323017139684963, + "grad_norm": 0.7037506103515625, + "learning_rate": 1.4651107346374593e-05, + "loss": 0.2486, + "step": 47888 + }, + { + "epoch": 3.3230865311220597, + "grad_norm": 0.7622580528259277, + "learning_rate": 1.464817955888732e-05, + "loss": 0.4049, + "step": 47889 + }, + { + "epoch": 3.323155922559156, + "grad_norm": 0.9963359236717224, + "learning_rate": 1.4645252040845847e-05, + "loss": 0.2227, + "step": 47890 + }, + { + "epoch": 3.323225313996253, + "grad_norm": 0.9808590412139893, + "learning_rate": 1.4642324792259443e-05, + "loss": 0.2611, + "step": 47891 + }, + { + "epoch": 3.3232947054333497, + "grad_norm": 0.7438235878944397, + "learning_rate": 1.4639397813137278e-05, + "loss": 0.228, + "step": 47892 + }, + { + "epoch": 3.323364096870446, + "grad_norm": 0.94547438621521, + "learning_rate": 1.4636471103488636e-05, + "loss": 0.2878, + "step": 47893 + }, + { + "epoch": 3.323433488307543, + "grad_norm": 0.7196076512336731, + "learning_rate": 1.4633544663322751e-05, + "loss": 0.3191, + "step": 47894 + }, + { + "epoch": 3.3235028797446393, + "grad_norm": 0.5880117416381836, + "learning_rate": 1.4630618492648863e-05, + "loss": 0.1364, + "step": 47895 + }, + { + "epoch": 3.323572271181736, + "grad_norm": 0.8272735476493835, + "learning_rate": 1.4627692591476206e-05, + "loss": 0.2463, + "step": 47896 + }, + { + "epoch": 3.323641662618833, + "grad_norm": 0.8394966125488281, + "learning_rate": 1.4624766959814018e-05, + "loss": 0.3506, + "step": 47897 + }, + { + "epoch": 3.3237110540559294, + "grad_norm": 0.8305506110191345, + "learning_rate": 1.462184159767156e-05, + "loss": 0.2234, + "step": 47898 + }, + { + "epoch": 3.323780445493026, + "grad_norm": 0.934556782245636, + "learning_rate": 1.4618916505058022e-05, + "loss": 0.3559, + "step": 47899 + }, + { + "epoch": 3.3238498369301226, + "grad_norm": 0.9108031392097473, + "learning_rate": 1.4615991681982654e-05, + "loss": 0.2014, + "step": 47900 + }, + { + "epoch": 3.3239192283672194, + "grad_norm": 0.8494241833686829, + "learning_rate": 1.4613067128454693e-05, + "loss": 0.1679, + "step": 47901 + }, + { + "epoch": 3.3239886198043163, + "grad_norm": 0.8736143112182617, + "learning_rate": 1.4610142844483377e-05, + "loss": 0.2358, + "step": 47902 + }, + { + "epoch": 3.3240580112414126, + "grad_norm": 0.8783608078956604, + "learning_rate": 1.4607218830077928e-05, + "loss": 0.1768, + "step": 47903 + }, + { + "epoch": 3.3241274026785095, + "grad_norm": 0.9708769917488098, + "learning_rate": 1.4604295085247588e-05, + "loss": 0.4097, + "step": 47904 + }, + { + "epoch": 3.3241967941156063, + "grad_norm": 1.0147197246551514, + "learning_rate": 1.4601371610001602e-05, + "loss": 0.2958, + "step": 47905 + }, + { + "epoch": 3.3242661855527027, + "grad_norm": 0.9618217349052429, + "learning_rate": 1.4598448404349152e-05, + "loss": 0.2578, + "step": 47906 + }, + { + "epoch": 3.3243355769897995, + "grad_norm": 0.6124770045280457, + "learning_rate": 1.4595525468299486e-05, + "loss": 0.177, + "step": 47907 + }, + { + "epoch": 3.3244049684268964, + "grad_norm": 0.7521764039993286, + "learning_rate": 1.4592602801861854e-05, + "loss": 0.2426, + "step": 47908 + }, + { + "epoch": 3.3244743598639928, + "grad_norm": 0.9505897164344788, + "learning_rate": 1.4589680405045447e-05, + "loss": 0.2564, + "step": 47909 + }, + { + "epoch": 3.3245437513010896, + "grad_norm": 0.8181738257408142, + "learning_rate": 1.4586758277859525e-05, + "loss": 0.2031, + "step": 47910 + }, + { + "epoch": 3.324613142738186, + "grad_norm": 0.9884834885597229, + "learning_rate": 1.458383642031329e-05, + "loss": 0.2955, + "step": 47911 + }, + { + "epoch": 3.324682534175283, + "grad_norm": 1.1801868677139282, + "learning_rate": 1.4580914832415992e-05, + "loss": 0.2608, + "step": 47912 + }, + { + "epoch": 3.3247519256123796, + "grad_norm": 0.7443333864212036, + "learning_rate": 1.4577993514176825e-05, + "loss": 0.1757, + "step": 47913 + }, + { + "epoch": 3.324821317049476, + "grad_norm": 1.0119692087173462, + "learning_rate": 1.4575072465605011e-05, + "loss": 0.5525, + "step": 47914 + }, + { + "epoch": 3.324890708486573, + "grad_norm": 1.3978126049041748, + "learning_rate": 1.457215168670979e-05, + "loss": 0.4069, + "step": 47915 + }, + { + "epoch": 3.3249600999236693, + "grad_norm": 0.6877488493919373, + "learning_rate": 1.4569231177500364e-05, + "loss": 0.2144, + "step": 47916 + }, + { + "epoch": 3.325029491360766, + "grad_norm": 1.047836184501648, + "learning_rate": 1.4566310937985972e-05, + "loss": 0.2115, + "step": 47917 + }, + { + "epoch": 3.325098882797863, + "grad_norm": 0.8868567943572998, + "learning_rate": 1.4563390968175838e-05, + "loss": 0.2717, + "step": 47918 + }, + { + "epoch": 3.3251682742349593, + "grad_norm": 0.72853684425354, + "learning_rate": 1.4560471268079156e-05, + "loss": 0.3243, + "step": 47919 + }, + { + "epoch": 3.325237665672056, + "grad_norm": 0.7379990816116333, + "learning_rate": 1.4557551837705153e-05, + "loss": 0.2518, + "step": 47920 + }, + { + "epoch": 3.3253070571091525, + "grad_norm": 0.8117194771766663, + "learning_rate": 1.4554632677063029e-05, + "loss": 0.2293, + "step": 47921 + }, + { + "epoch": 3.3253764485462494, + "grad_norm": 1.245957851409912, + "learning_rate": 1.4551713786162058e-05, + "loss": 0.3104, + "step": 47922 + }, + { + "epoch": 3.325445839983346, + "grad_norm": 0.7747524380683899, + "learning_rate": 1.4548795165011374e-05, + "loss": 0.2761, + "step": 47923 + }, + { + "epoch": 3.3255152314204426, + "grad_norm": 1.8313450813293457, + "learning_rate": 1.454587681362024e-05, + "loss": 0.4601, + "step": 47924 + }, + { + "epoch": 3.3255846228575394, + "grad_norm": 0.8766728639602661, + "learning_rate": 1.4542958731997858e-05, + "loss": 0.2288, + "step": 47925 + }, + { + "epoch": 3.325654014294636, + "grad_norm": 0.7859913110733032, + "learning_rate": 1.4540040920153464e-05, + "loss": 0.3188, + "step": 47926 + }, + { + "epoch": 3.3257234057317326, + "grad_norm": 0.719104528427124, + "learning_rate": 1.4537123378096219e-05, + "loss": 0.2566, + "step": 47927 + }, + { + "epoch": 3.3257927971688295, + "grad_norm": 1.0226367712020874, + "learning_rate": 1.4534206105835358e-05, + "loss": 0.2612, + "step": 47928 + }, + { + "epoch": 3.325862188605926, + "grad_norm": 0.7977219820022583, + "learning_rate": 1.4531289103380098e-05, + "loss": 0.2156, + "step": 47929 + }, + { + "epoch": 3.3259315800430227, + "grad_norm": 0.8449790477752686, + "learning_rate": 1.452837237073964e-05, + "loss": 0.2352, + "step": 47930 + }, + { + "epoch": 3.3260009714801195, + "grad_norm": 0.9304174780845642, + "learning_rate": 1.4525455907923192e-05, + "loss": 0.2197, + "step": 47931 + }, + { + "epoch": 3.326070362917216, + "grad_norm": 1.2259526252746582, + "learning_rate": 1.4522539714939964e-05, + "loss": 0.2558, + "step": 47932 + }, + { + "epoch": 3.3261397543543127, + "grad_norm": 1.494246482849121, + "learning_rate": 1.4519623791799186e-05, + "loss": 0.3242, + "step": 47933 + }, + { + "epoch": 3.3262091457914096, + "grad_norm": 1.1523196697235107, + "learning_rate": 1.4516708138510005e-05, + "loss": 0.3748, + "step": 47934 + }, + { + "epoch": 3.326278537228506, + "grad_norm": 0.7170604467391968, + "learning_rate": 1.4513792755081667e-05, + "loss": 0.2112, + "step": 47935 + }, + { + "epoch": 3.326347928665603, + "grad_norm": 0.7288220524787903, + "learning_rate": 1.4510877641523368e-05, + "loss": 0.2043, + "step": 47936 + }, + { + "epoch": 3.326417320102699, + "grad_norm": 1.3948274850845337, + "learning_rate": 1.450796279784431e-05, + "loss": 0.2729, + "step": 47937 + }, + { + "epoch": 3.326486711539796, + "grad_norm": 0.8047090768814087, + "learning_rate": 1.4505048224053696e-05, + "loss": 0.2402, + "step": 47938 + }, + { + "epoch": 3.326556102976893, + "grad_norm": 0.8886420726776123, + "learning_rate": 1.450213392016072e-05, + "loss": 0.189, + "step": 47939 + }, + { + "epoch": 3.3266254944139892, + "grad_norm": 0.9362377524375916, + "learning_rate": 1.449921988617461e-05, + "loss": 0.304, + "step": 47940 + }, + { + "epoch": 3.326694885851086, + "grad_norm": 0.885423481464386, + "learning_rate": 1.4496306122104531e-05, + "loss": 0.3794, + "step": 47941 + }, + { + "epoch": 3.3267642772881825, + "grad_norm": 0.9400507807731628, + "learning_rate": 1.4493392627959679e-05, + "loss": 0.2053, + "step": 47942 + }, + { + "epoch": 3.3268336687252793, + "grad_norm": 0.8923110365867615, + "learning_rate": 1.4490479403749268e-05, + "loss": 0.262, + "step": 47943 + }, + { + "epoch": 3.326903060162376, + "grad_norm": 1.0721417665481567, + "learning_rate": 1.4487566449482504e-05, + "loss": 0.2876, + "step": 47944 + }, + { + "epoch": 3.3269724515994725, + "grad_norm": 0.793342113494873, + "learning_rate": 1.4484653765168565e-05, + "loss": 0.2745, + "step": 47945 + }, + { + "epoch": 3.3270418430365694, + "grad_norm": 0.8653436899185181, + "learning_rate": 1.4481741350816647e-05, + "loss": 0.2568, + "step": 47946 + }, + { + "epoch": 3.3271112344736657, + "grad_norm": 1.0262093544006348, + "learning_rate": 1.4478829206435973e-05, + "loss": 0.4549, + "step": 47947 + }, + { + "epoch": 3.3271806259107626, + "grad_norm": 1.1080857515335083, + "learning_rate": 1.4475917332035693e-05, + "loss": 0.2158, + "step": 47948 + }, + { + "epoch": 3.3272500173478594, + "grad_norm": 0.822868287563324, + "learning_rate": 1.4473005727625023e-05, + "loss": 0.2575, + "step": 47949 + }, + { + "epoch": 3.327319408784956, + "grad_norm": 1.1529713869094849, + "learning_rate": 1.447009439321315e-05, + "loss": 0.3107, + "step": 47950 + }, + { + "epoch": 3.3273888002220526, + "grad_norm": 0.8232079744338989, + "learning_rate": 1.4467183328809264e-05, + "loss": 0.1856, + "step": 47951 + }, + { + "epoch": 3.327458191659149, + "grad_norm": 0.8784090280532837, + "learning_rate": 1.4464272534422563e-05, + "loss": 0.1779, + "step": 47952 + }, + { + "epoch": 3.327527583096246, + "grad_norm": 1.0276859998703003, + "learning_rate": 1.4461362010062241e-05, + "loss": 0.1649, + "step": 47953 + }, + { + "epoch": 3.3275969745333427, + "grad_norm": 1.0324180126190186, + "learning_rate": 1.4458451755737456e-05, + "loss": 0.2762, + "step": 47954 + }, + { + "epoch": 3.327666365970439, + "grad_norm": 0.739321231842041, + "learning_rate": 1.4455541771457415e-05, + "loss": 0.2801, + "step": 47955 + }, + { + "epoch": 3.327735757407536, + "grad_norm": 0.9798319935798645, + "learning_rate": 1.4452632057231308e-05, + "loss": 0.4236, + "step": 47956 + }, + { + "epoch": 3.3278051488446327, + "grad_norm": 0.932458758354187, + "learning_rate": 1.4449722613068329e-05, + "loss": 0.2501, + "step": 47957 + }, + { + "epoch": 3.327874540281729, + "grad_norm": 1.0701251029968262, + "learning_rate": 1.4446813438977636e-05, + "loss": 0.2631, + "step": 47958 + }, + { + "epoch": 3.327943931718826, + "grad_norm": 0.7240661382675171, + "learning_rate": 1.4443904534968422e-05, + "loss": 0.1934, + "step": 47959 + }, + { + "epoch": 3.328013323155923, + "grad_norm": 1.1770224571228027, + "learning_rate": 1.4440995901049903e-05, + "loss": 0.3856, + "step": 47960 + }, + { + "epoch": 3.328082714593019, + "grad_norm": 2.4605796337127686, + "learning_rate": 1.4438087537231203e-05, + "loss": 0.2101, + "step": 47961 + }, + { + "epoch": 3.328152106030116, + "grad_norm": 0.6897443532943726, + "learning_rate": 1.443517944352154e-05, + "loss": 0.1855, + "step": 47962 + }, + { + "epoch": 3.3282214974672124, + "grad_norm": 0.7665672898292542, + "learning_rate": 1.4432271619930082e-05, + "loss": 0.1956, + "step": 47963 + }, + { + "epoch": 3.3282908889043092, + "grad_norm": 1.6133873462677002, + "learning_rate": 1.4429364066466022e-05, + "loss": 0.2769, + "step": 47964 + }, + { + "epoch": 3.328360280341406, + "grad_norm": 1.0564403533935547, + "learning_rate": 1.4426456783138531e-05, + "loss": 0.3127, + "step": 47965 + }, + { + "epoch": 3.3284296717785025, + "grad_norm": 0.7700970768928528, + "learning_rate": 1.442354976995679e-05, + "loss": 0.2161, + "step": 47966 + }, + { + "epoch": 3.3284990632155993, + "grad_norm": 1.19585120677948, + "learning_rate": 1.4420643026929993e-05, + "loss": 0.3242, + "step": 47967 + }, + { + "epoch": 3.3285684546526957, + "grad_norm": 0.8255268931388855, + "learning_rate": 1.4417736554067262e-05, + "loss": 0.1441, + "step": 47968 + }, + { + "epoch": 3.3286378460897925, + "grad_norm": 1.0051538944244385, + "learning_rate": 1.4414830351377828e-05, + "loss": 0.2449, + "step": 47969 + }, + { + "epoch": 3.3287072375268894, + "grad_norm": 0.9795612096786499, + "learning_rate": 1.4411924418870826e-05, + "loss": 0.2479, + "step": 47970 + }, + { + "epoch": 3.3287766289639857, + "grad_norm": 1.2872889041900635, + "learning_rate": 1.4409018756555459e-05, + "loss": 0.3312, + "step": 47971 + }, + { + "epoch": 3.3288460204010826, + "grad_norm": 0.7026882171630859, + "learning_rate": 1.4406113364440887e-05, + "loss": 0.2129, + "step": 47972 + }, + { + "epoch": 3.328915411838179, + "grad_norm": 1.0523463487625122, + "learning_rate": 1.4403208242536292e-05, + "loss": 0.4049, + "step": 47973 + }, + { + "epoch": 3.328984803275276, + "grad_norm": 1.0243127346038818, + "learning_rate": 1.4400303390850844e-05, + "loss": 0.3408, + "step": 47974 + }, + { + "epoch": 3.3290541947123726, + "grad_norm": 1.0864310264587402, + "learning_rate": 1.4397398809393703e-05, + "loss": 0.3214, + "step": 47975 + }, + { + "epoch": 3.329123586149469, + "grad_norm": 0.7159032821655273, + "learning_rate": 1.4394494498174026e-05, + "loss": 0.2793, + "step": 47976 + }, + { + "epoch": 3.329192977586566, + "grad_norm": 0.7133042216300964, + "learning_rate": 1.439159045720101e-05, + "loss": 0.1883, + "step": 47977 + }, + { + "epoch": 3.3292623690236622, + "grad_norm": 0.7041470408439636, + "learning_rate": 1.4388686686483798e-05, + "loss": 0.1721, + "step": 47978 + }, + { + "epoch": 3.329331760460759, + "grad_norm": 0.8447162508964539, + "learning_rate": 1.4385783186031577e-05, + "loss": 0.2278, + "step": 47979 + }, + { + "epoch": 3.329401151897856, + "grad_norm": 0.815658688545227, + "learning_rate": 1.4382879955853513e-05, + "loss": 0.2281, + "step": 47980 + }, + { + "epoch": 3.3294705433349523, + "grad_norm": 0.9123049378395081, + "learning_rate": 1.4379976995958777e-05, + "loss": 0.3424, + "step": 47981 + }, + { + "epoch": 3.329539934772049, + "grad_norm": 0.8043678998947144, + "learning_rate": 1.4377074306356496e-05, + "loss": 0.1304, + "step": 47982 + }, + { + "epoch": 3.329609326209146, + "grad_norm": 0.8645951151847839, + "learning_rate": 1.4374171887055855e-05, + "loss": 0.1669, + "step": 47983 + }, + { + "epoch": 3.3296787176462423, + "grad_norm": 1.0865107774734497, + "learning_rate": 1.4371269738066029e-05, + "loss": 0.1717, + "step": 47984 + }, + { + "epoch": 3.329748109083339, + "grad_norm": 0.6179049611091614, + "learning_rate": 1.436836785939616e-05, + "loss": 0.1563, + "step": 47985 + }, + { + "epoch": 3.329817500520436, + "grad_norm": 1.3369419574737549, + "learning_rate": 1.4365466251055426e-05, + "loss": 0.412, + "step": 47986 + }, + { + "epoch": 3.3298868919575324, + "grad_norm": 0.8116111159324646, + "learning_rate": 1.4362564913052979e-05, + "loss": 0.3694, + "step": 47987 + }, + { + "epoch": 3.3299562833946292, + "grad_norm": 0.8788337111473083, + "learning_rate": 1.4359663845397997e-05, + "loss": 0.2428, + "step": 47988 + }, + { + "epoch": 3.3300256748317256, + "grad_norm": 0.9863951206207275, + "learning_rate": 1.4356763048099598e-05, + "loss": 0.2913, + "step": 47989 + }, + { + "epoch": 3.3300950662688225, + "grad_norm": 0.7061769366264343, + "learning_rate": 1.4353862521166961e-05, + "loss": 0.204, + "step": 47990 + }, + { + "epoch": 3.3301644577059193, + "grad_norm": 0.7708362340927124, + "learning_rate": 1.4350962264609269e-05, + "loss": 0.256, + "step": 47991 + }, + { + "epoch": 3.3302338491430157, + "grad_norm": 0.6669869422912598, + "learning_rate": 1.4348062278435625e-05, + "loss": 0.2556, + "step": 47992 + }, + { + "epoch": 3.3303032405801125, + "grad_norm": 1.0141196250915527, + "learning_rate": 1.4345162562655212e-05, + "loss": 0.2975, + "step": 47993 + }, + { + "epoch": 3.330372632017209, + "grad_norm": 0.7749674916267395, + "learning_rate": 1.4342263117277189e-05, + "loss": 0.1816, + "step": 47994 + }, + { + "epoch": 3.3304420234543057, + "grad_norm": 0.8265372514724731, + "learning_rate": 1.4339363942310724e-05, + "loss": 0.3158, + "step": 47995 + }, + { + "epoch": 3.3305114148914026, + "grad_norm": 0.7420514225959778, + "learning_rate": 1.4336465037764923e-05, + "loss": 0.2529, + "step": 47996 + }, + { + "epoch": 3.330580806328499, + "grad_norm": 0.7922024726867676, + "learning_rate": 1.4333566403648968e-05, + "loss": 0.1551, + "step": 47997 + }, + { + "epoch": 3.330650197765596, + "grad_norm": 1.1001720428466797, + "learning_rate": 1.4330668039971995e-05, + "loss": 0.194, + "step": 47998 + }, + { + "epoch": 3.330719589202692, + "grad_norm": 0.9541339874267578, + "learning_rate": 1.4327769946743174e-05, + "loss": 0.2563, + "step": 47999 + }, + { + "epoch": 3.330788980639789, + "grad_norm": 0.8635066747665405, + "learning_rate": 1.4324872123971633e-05, + "loss": 0.2512, + "step": 48000 + }, + { + "epoch": 3.330788980639789, + "eval_loss": 0.4727431535720825, + "eval_runtime": 3848.7578, + "eval_samples_per_second": 6.306, + "eval_steps_per_second": 0.788, + "step": 48000 + }, + { + "epoch": 3.330858372076886, + "grad_norm": 1.0059058666229248, + "learning_rate": 1.432197457166654e-05, + "loss": 0.2825, + "step": 48001 + }, + { + "epoch": 3.3309277635139822, + "grad_norm": 1.1831334829330444, + "learning_rate": 1.4319077289837046e-05, + "loss": 0.3513, + "step": 48002 + }, + { + "epoch": 3.330997154951079, + "grad_norm": 1.1457533836364746, + "learning_rate": 1.4316180278492275e-05, + "loss": 0.2595, + "step": 48003 + }, + { + "epoch": 3.3310665463881755, + "grad_norm": 0.8474962115287781, + "learning_rate": 1.4313283537641364e-05, + "loss": 0.3043, + "step": 48004 + }, + { + "epoch": 3.3311359378252723, + "grad_norm": 1.7413439750671387, + "learning_rate": 1.4310387067293495e-05, + "loss": 0.416, + "step": 48005 + }, + { + "epoch": 3.331205329262369, + "grad_norm": 0.8064813017845154, + "learning_rate": 1.4307490867457774e-05, + "loss": 0.2007, + "step": 48006 + }, + { + "epoch": 3.3312747206994655, + "grad_norm": 0.7331943511962891, + "learning_rate": 1.430459493814338e-05, + "loss": 0.2728, + "step": 48007 + }, + { + "epoch": 3.3313441121365623, + "grad_norm": 0.5377798676490784, + "learning_rate": 1.4301699279359426e-05, + "loss": 0.1637, + "step": 48008 + }, + { + "epoch": 3.331413503573659, + "grad_norm": 0.7597735524177551, + "learning_rate": 1.4298803891115087e-05, + "loss": 0.2503, + "step": 48009 + }, + { + "epoch": 3.3314828950107556, + "grad_norm": 0.8889057636260986, + "learning_rate": 1.4295908773419453e-05, + "loss": 0.2596, + "step": 48010 + }, + { + "epoch": 3.3315522864478524, + "grad_norm": 0.6412687301635742, + "learning_rate": 1.4293013926281706e-05, + "loss": 0.1931, + "step": 48011 + }, + { + "epoch": 3.3316216778849492, + "grad_norm": 1.2708698511123657, + "learning_rate": 1.4290119349710962e-05, + "loss": 0.2487, + "step": 48012 + }, + { + "epoch": 3.3316910693220456, + "grad_norm": 0.7900117039680481, + "learning_rate": 1.4287225043716368e-05, + "loss": 0.2412, + "step": 48013 + }, + { + "epoch": 3.3317604607591425, + "grad_norm": 0.6214935183525085, + "learning_rate": 1.4284331008307062e-05, + "loss": 0.15, + "step": 48014 + }, + { + "epoch": 3.331829852196239, + "grad_norm": 0.9973556399345398, + "learning_rate": 1.4281437243492168e-05, + "loss": 0.286, + "step": 48015 + }, + { + "epoch": 3.3318992436333357, + "grad_norm": 0.9787096977233887, + "learning_rate": 1.4278543749280871e-05, + "loss": 0.1616, + "step": 48016 + }, + { + "epoch": 3.3319686350704325, + "grad_norm": 0.7483053803443909, + "learning_rate": 1.4275650525682228e-05, + "loss": 0.1739, + "step": 48017 + }, + { + "epoch": 3.332038026507529, + "grad_norm": 0.8874703645706177, + "learning_rate": 1.4272757572705419e-05, + "loss": 0.2921, + "step": 48018 + }, + { + "epoch": 3.3321074179446257, + "grad_norm": 0.6612038016319275, + "learning_rate": 1.4269864890359564e-05, + "loss": 0.1752, + "step": 48019 + }, + { + "epoch": 3.332176809381722, + "grad_norm": 0.7656506299972534, + "learning_rate": 1.4266972478653806e-05, + "loss": 0.2464, + "step": 48020 + }, + { + "epoch": 3.332246200818819, + "grad_norm": 0.7028480768203735, + "learning_rate": 1.4264080337597263e-05, + "loss": 0.1657, + "step": 48021 + }, + { + "epoch": 3.332315592255916, + "grad_norm": 0.6488137245178223, + "learning_rate": 1.4261188467199071e-05, + "loss": 0.1972, + "step": 48022 + }, + { + "epoch": 3.332384983693012, + "grad_norm": 1.028062343597412, + "learning_rate": 1.4258296867468379e-05, + "loss": 0.4909, + "step": 48023 + }, + { + "epoch": 3.332454375130109, + "grad_norm": 0.9638089537620544, + "learning_rate": 1.425540553841428e-05, + "loss": 0.2346, + "step": 48024 + }, + { + "epoch": 3.3325237665672054, + "grad_norm": 0.7078832387924194, + "learning_rate": 1.425251448004592e-05, + "loss": 0.2452, + "step": 48025 + }, + { + "epoch": 3.3325931580043022, + "grad_norm": 0.5633235573768616, + "learning_rate": 1.4249623692372439e-05, + "loss": 0.1514, + "step": 48026 + }, + { + "epoch": 3.332662549441399, + "grad_norm": 1.2009782791137695, + "learning_rate": 1.4246733175402926e-05, + "loss": 0.2973, + "step": 48027 + }, + { + "epoch": 3.3327319408784954, + "grad_norm": 1.1590882539749146, + "learning_rate": 1.4243842929146522e-05, + "loss": 0.3021, + "step": 48028 + }, + { + "epoch": 3.3328013323155923, + "grad_norm": 1.0541515350341797, + "learning_rate": 1.4240952953612364e-05, + "loss": 0.3607, + "step": 48029 + }, + { + "epoch": 3.3328707237526887, + "grad_norm": 0.8434950113296509, + "learning_rate": 1.4238063248809586e-05, + "loss": 0.201, + "step": 48030 + }, + { + "epoch": 3.3329401151897855, + "grad_norm": 0.8022478222846985, + "learning_rate": 1.4235173814747272e-05, + "loss": 0.1848, + "step": 48031 + }, + { + "epoch": 3.3330095066268823, + "grad_norm": 0.8661524057388306, + "learning_rate": 1.4232284651434568e-05, + "loss": 0.2679, + "step": 48032 + }, + { + "epoch": 3.3330788980639787, + "grad_norm": 0.8704030513763428, + "learning_rate": 1.4229395758880581e-05, + "loss": 0.2441, + "step": 48033 + }, + { + "epoch": 3.3331482895010756, + "grad_norm": 0.7305774688720703, + "learning_rate": 1.4226507137094447e-05, + "loss": 0.1887, + "step": 48034 + }, + { + "epoch": 3.3332176809381724, + "grad_norm": 0.75074702501297, + "learning_rate": 1.4223618786085279e-05, + "loss": 0.1562, + "step": 48035 + }, + { + "epoch": 3.333287072375269, + "grad_norm": 0.9407943487167358, + "learning_rate": 1.4220730705862218e-05, + "loss": 0.2171, + "step": 48036 + }, + { + "epoch": 3.3333564638123656, + "grad_norm": 1.083979606628418, + "learning_rate": 1.421784289643433e-05, + "loss": 0.3117, + "step": 48037 + }, + { + "epoch": 3.3334258552494624, + "grad_norm": 1.3074164390563965, + "learning_rate": 1.4214955357810755e-05, + "loss": 0.2973, + "step": 48038 + }, + { + "epoch": 3.333495246686559, + "grad_norm": 0.6721934080123901, + "learning_rate": 1.4212068090000618e-05, + "loss": 0.1746, + "step": 48039 + }, + { + "epoch": 3.3335646381236557, + "grad_norm": 0.6497955918312073, + "learning_rate": 1.4209181093013036e-05, + "loss": 0.1885, + "step": 48040 + }, + { + "epoch": 3.333634029560752, + "grad_norm": 0.5303329229354858, + "learning_rate": 1.4206294366857108e-05, + "loss": 0.1944, + "step": 48041 + }, + { + "epoch": 3.333703420997849, + "grad_norm": 0.6953983306884766, + "learning_rate": 1.4203407911541954e-05, + "loss": 0.1671, + "step": 48042 + }, + { + "epoch": 3.3337728124349457, + "grad_norm": 0.7264481782913208, + "learning_rate": 1.420052172707671e-05, + "loss": 0.2525, + "step": 48043 + }, + { + "epoch": 3.333842203872042, + "grad_norm": 0.7504560947418213, + "learning_rate": 1.4197635813470444e-05, + "loss": 0.2552, + "step": 48044 + }, + { + "epoch": 3.333911595309139, + "grad_norm": 0.8860438466072083, + "learning_rate": 1.4194750170732296e-05, + "loss": 0.2763, + "step": 48045 + }, + { + "epoch": 3.3339809867462353, + "grad_norm": 1.0044243335723877, + "learning_rate": 1.4191864798871357e-05, + "loss": 0.2878, + "step": 48046 + }, + { + "epoch": 3.334050378183332, + "grad_norm": 0.8784679174423218, + "learning_rate": 1.4188979697896742e-05, + "loss": 0.2981, + "step": 48047 + }, + { + "epoch": 3.334119769620429, + "grad_norm": 0.932053804397583, + "learning_rate": 1.4186094867817567e-05, + "loss": 0.3324, + "step": 48048 + }, + { + "epoch": 3.3341891610575254, + "grad_norm": 0.6473264694213867, + "learning_rate": 1.4183210308642935e-05, + "loss": 0.2155, + "step": 48049 + }, + { + "epoch": 3.334258552494622, + "grad_norm": 0.6076818108558655, + "learning_rate": 1.4180326020381972e-05, + "loss": 0.229, + "step": 48050 + }, + { + "epoch": 3.3343279439317186, + "grad_norm": 0.8388148546218872, + "learning_rate": 1.4177442003043739e-05, + "loss": 0.2727, + "step": 48051 + }, + { + "epoch": 3.3343973353688154, + "grad_norm": 1.861372470855713, + "learning_rate": 1.4174558256637361e-05, + "loss": 0.2328, + "step": 48052 + }, + { + "epoch": 3.3344667268059123, + "grad_norm": 0.732322633266449, + "learning_rate": 1.417167478117195e-05, + "loss": 0.1853, + "step": 48053 + }, + { + "epoch": 3.3345361182430087, + "grad_norm": 0.6464574933052063, + "learning_rate": 1.4168791576656604e-05, + "loss": 0.139, + "step": 48054 + }, + { + "epoch": 3.3346055096801055, + "grad_norm": 0.8372011780738831, + "learning_rate": 1.4165908643100424e-05, + "loss": 0.1602, + "step": 48055 + }, + { + "epoch": 3.3346749011172023, + "grad_norm": 1.4802478551864624, + "learning_rate": 1.4163025980512512e-05, + "loss": 0.3023, + "step": 48056 + }, + { + "epoch": 3.3347442925542987, + "grad_norm": 1.2127820253372192, + "learning_rate": 1.4160143588901987e-05, + "loss": 0.2975, + "step": 48057 + }, + { + "epoch": 3.3348136839913955, + "grad_norm": 1.0000141859054565, + "learning_rate": 1.4157261468277915e-05, + "loss": 0.2451, + "step": 48058 + }, + { + "epoch": 3.334883075428492, + "grad_norm": 0.8594576120376587, + "learning_rate": 1.4154379618649406e-05, + "loss": 0.4435, + "step": 48059 + }, + { + "epoch": 3.3349524668655888, + "grad_norm": 0.7601140141487122, + "learning_rate": 1.4151498040025557e-05, + "loss": 0.2925, + "step": 48060 + }, + { + "epoch": 3.3350218583026856, + "grad_norm": 0.973446786403656, + "learning_rate": 1.4148616732415499e-05, + "loss": 0.2417, + "step": 48061 + }, + { + "epoch": 3.335091249739782, + "grad_norm": 1.0319921970367432, + "learning_rate": 1.4145735695828266e-05, + "loss": 0.2323, + "step": 48062 + }, + { + "epoch": 3.335160641176879, + "grad_norm": 0.9921954274177551, + "learning_rate": 1.4142854930272986e-05, + "loss": 0.3393, + "step": 48063 + }, + { + "epoch": 3.3352300326139757, + "grad_norm": 0.9856935143470764, + "learning_rate": 1.4139974435758752e-05, + "loss": 0.24, + "step": 48064 + }, + { + "epoch": 3.335299424051072, + "grad_norm": 0.9000915288925171, + "learning_rate": 1.4137094212294677e-05, + "loss": 0.2594, + "step": 48065 + }, + { + "epoch": 3.335368815488169, + "grad_norm": 1.224116325378418, + "learning_rate": 1.4134214259889821e-05, + "loss": 0.2538, + "step": 48066 + }, + { + "epoch": 3.3354382069252653, + "grad_norm": 1.2766313552856445, + "learning_rate": 1.4131334578553279e-05, + "loss": 0.2708, + "step": 48067 + }, + { + "epoch": 3.335507598362362, + "grad_norm": 1.2345224618911743, + "learning_rate": 1.4128455168294153e-05, + "loss": 0.3755, + "step": 48068 + }, + { + "epoch": 3.335576989799459, + "grad_norm": 1.1684095859527588, + "learning_rate": 1.4125576029121535e-05, + "loss": 0.2257, + "step": 48069 + }, + { + "epoch": 3.3356463812365553, + "grad_norm": 0.7175374627113342, + "learning_rate": 1.4122697161044507e-05, + "loss": 0.2439, + "step": 48070 + }, + { + "epoch": 3.335715772673652, + "grad_norm": 0.9069969058036804, + "learning_rate": 1.4119818564072185e-05, + "loss": 0.4808, + "step": 48071 + }, + { + "epoch": 3.3357851641107485, + "grad_norm": 1.4072014093399048, + "learning_rate": 1.4116940238213605e-05, + "loss": 0.2571, + "step": 48072 + }, + { + "epoch": 3.3358545555478454, + "grad_norm": 0.9816429018974304, + "learning_rate": 1.4114062183477882e-05, + "loss": 0.3805, + "step": 48073 + }, + { + "epoch": 3.335923946984942, + "grad_norm": 0.6383629441261292, + "learning_rate": 1.41111843998741e-05, + "loss": 0.1436, + "step": 48074 + }, + { + "epoch": 3.3359933384220386, + "grad_norm": 1.1054222583770752, + "learning_rate": 1.4108306887411348e-05, + "loss": 0.2731, + "step": 48075 + }, + { + "epoch": 3.3360627298591354, + "grad_norm": 1.3062160015106201, + "learning_rate": 1.41054296460987e-05, + "loss": 0.2259, + "step": 48076 + }, + { + "epoch": 3.336132121296232, + "grad_norm": 0.8697853088378906, + "learning_rate": 1.4102552675945246e-05, + "loss": 0.2431, + "step": 48077 + }, + { + "epoch": 3.3362015127333287, + "grad_norm": 0.8776026368141174, + "learning_rate": 1.4099675976960091e-05, + "loss": 0.2852, + "step": 48078 + }, + { + "epoch": 3.3362709041704255, + "grad_norm": 0.8121945261955261, + "learning_rate": 1.4096799549152274e-05, + "loss": 0.2628, + "step": 48079 + }, + { + "epoch": 3.336340295607522, + "grad_norm": 1.2668004035949707, + "learning_rate": 1.4093923392530894e-05, + "loss": 0.343, + "step": 48080 + }, + { + "epoch": 3.3364096870446187, + "grad_norm": 1.2165029048919678, + "learning_rate": 1.4091047507105027e-05, + "loss": 0.2791, + "step": 48081 + }, + { + "epoch": 3.3364790784817155, + "grad_norm": 1.8047020435333252, + "learning_rate": 1.408817189288375e-05, + "loss": 0.2914, + "step": 48082 + }, + { + "epoch": 3.336548469918812, + "grad_norm": 0.5957052707672119, + "learning_rate": 1.4085296549876159e-05, + "loss": 0.2143, + "step": 48083 + }, + { + "epoch": 3.3366178613559088, + "grad_norm": 0.9132304787635803, + "learning_rate": 1.4082421478091313e-05, + "loss": 0.2843, + "step": 48084 + }, + { + "epoch": 3.336687252793005, + "grad_norm": 0.907092273235321, + "learning_rate": 1.4079546677538313e-05, + "loss": 0.2357, + "step": 48085 + }, + { + "epoch": 3.336756644230102, + "grad_norm": 0.8503269553184509, + "learning_rate": 1.40766721482262e-05, + "loss": 0.2251, + "step": 48086 + }, + { + "epoch": 3.336826035667199, + "grad_norm": 0.8678933382034302, + "learning_rate": 1.4073797890164064e-05, + "loss": 0.2552, + "step": 48087 + }, + { + "epoch": 3.336895427104295, + "grad_norm": 1.0595208406448364, + "learning_rate": 1.4070923903360966e-05, + "loss": 0.3053, + "step": 48088 + }, + { + "epoch": 3.336964818541392, + "grad_norm": 0.675259530544281, + "learning_rate": 1.406805018782601e-05, + "loss": 0.1609, + "step": 48089 + }, + { + "epoch": 3.337034209978489, + "grad_norm": 0.8192583918571472, + "learning_rate": 1.4065176743568242e-05, + "loss": 0.2916, + "step": 48090 + }, + { + "epoch": 3.3371036014155853, + "grad_norm": 0.9140140414237976, + "learning_rate": 1.4062303570596736e-05, + "loss": 0.2902, + "step": 48091 + }, + { + "epoch": 3.337172992852682, + "grad_norm": 1.019768238067627, + "learning_rate": 1.4059430668920592e-05, + "loss": 0.3258, + "step": 48092 + }, + { + "epoch": 3.3372423842897785, + "grad_norm": 0.7486204504966736, + "learning_rate": 1.4056558038548839e-05, + "loss": 0.2461, + "step": 48093 + }, + { + "epoch": 3.3373117757268753, + "grad_norm": 0.9273965358734131, + "learning_rate": 1.4053685679490546e-05, + "loss": 0.2771, + "step": 48094 + }, + { + "epoch": 3.337381167163972, + "grad_norm": 0.7080788016319275, + "learning_rate": 1.4050813591754808e-05, + "loss": 0.1901, + "step": 48095 + }, + { + "epoch": 3.3374505586010685, + "grad_norm": 0.9800874590873718, + "learning_rate": 1.4047941775350704e-05, + "loss": 0.1611, + "step": 48096 + }, + { + "epoch": 3.3375199500381654, + "grad_norm": 0.8653706312179565, + "learning_rate": 1.4045070230287238e-05, + "loss": 0.2194, + "step": 48097 + }, + { + "epoch": 3.3375893414752618, + "grad_norm": 0.7580340504646301, + "learning_rate": 1.4042198956573527e-05, + "loss": 0.179, + "step": 48098 + }, + { + "epoch": 3.3376587329123586, + "grad_norm": 1.2884085178375244, + "learning_rate": 1.4039327954218617e-05, + "loss": 0.2794, + "step": 48099 + }, + { + "epoch": 3.3377281243494554, + "grad_norm": 1.1594926118850708, + "learning_rate": 1.4036457223231602e-05, + "loss": 0.1513, + "step": 48100 + }, + { + "epoch": 3.337797515786552, + "grad_norm": 1.1546648740768433, + "learning_rate": 1.4033586763621487e-05, + "loss": 0.3097, + "step": 48101 + }, + { + "epoch": 3.3378669072236486, + "grad_norm": 0.9060418605804443, + "learning_rate": 1.4030716575397362e-05, + "loss": 0.3279, + "step": 48102 + }, + { + "epoch": 3.337936298660745, + "grad_norm": 0.665973961353302, + "learning_rate": 1.40278466585683e-05, + "loss": 0.2423, + "step": 48103 + }, + { + "epoch": 3.338005690097842, + "grad_norm": 1.1095224618911743, + "learning_rate": 1.4024977013143348e-05, + "loss": 0.2601, + "step": 48104 + }, + { + "epoch": 3.3380750815349387, + "grad_norm": 1.1360353231430054, + "learning_rate": 1.4022107639131577e-05, + "loss": 0.2579, + "step": 48105 + }, + { + "epoch": 3.338144472972035, + "grad_norm": 0.7686911821365356, + "learning_rate": 1.4019238536542045e-05, + "loss": 0.2077, + "step": 48106 + }, + { + "epoch": 3.338213864409132, + "grad_norm": 0.8744713664054871, + "learning_rate": 1.4016369705383792e-05, + "loss": 0.2111, + "step": 48107 + }, + { + "epoch": 3.3382832558462288, + "grad_norm": 0.7921186685562134, + "learning_rate": 1.4013501145665875e-05, + "loss": 0.196, + "step": 48108 + }, + { + "epoch": 3.338352647283325, + "grad_norm": 0.9544202089309692, + "learning_rate": 1.4010632857397354e-05, + "loss": 0.3342, + "step": 48109 + }, + { + "epoch": 3.338422038720422, + "grad_norm": 0.9795116782188416, + "learning_rate": 1.40077648405873e-05, + "loss": 0.2845, + "step": 48110 + }, + { + "epoch": 3.3384914301575184, + "grad_norm": 0.9677219986915588, + "learning_rate": 1.400489709524475e-05, + "loss": 0.2876, + "step": 48111 + }, + { + "epoch": 3.338560821594615, + "grad_norm": 1.1793416738510132, + "learning_rate": 1.4002029621378788e-05, + "loss": 0.3675, + "step": 48112 + }, + { + "epoch": 3.338630213031712, + "grad_norm": 0.4861081540584564, + "learning_rate": 1.3999162418998413e-05, + "loss": 0.1312, + "step": 48113 + }, + { + "epoch": 3.3386996044688084, + "grad_norm": 0.9537944793701172, + "learning_rate": 1.399629548811271e-05, + "loss": 0.3203, + "step": 48114 + }, + { + "epoch": 3.3387689959059053, + "grad_norm": 1.1374279260635376, + "learning_rate": 1.3993428828730716e-05, + "loss": 0.3191, + "step": 48115 + }, + { + "epoch": 3.338838387343002, + "grad_norm": 0.8479323983192444, + "learning_rate": 1.3990562440861498e-05, + "loss": 0.2522, + "step": 48116 + }, + { + "epoch": 3.3389077787800985, + "grad_norm": 0.7651646137237549, + "learning_rate": 1.3987696324514088e-05, + "loss": 0.2755, + "step": 48117 + }, + { + "epoch": 3.3389771702171953, + "grad_norm": 0.7731497883796692, + "learning_rate": 1.3984830479697542e-05, + "loss": 0.1752, + "step": 48118 + }, + { + "epoch": 3.3390465616542917, + "grad_norm": 1.1230052709579468, + "learning_rate": 1.398196490642093e-05, + "loss": 0.1875, + "step": 48119 + }, + { + "epoch": 3.3391159530913885, + "grad_norm": 0.7740380764007568, + "learning_rate": 1.3979099604693257e-05, + "loss": 0.1581, + "step": 48120 + }, + { + "epoch": 3.3391853445284854, + "grad_norm": 0.7037861943244934, + "learning_rate": 1.3976234574523583e-05, + "loss": 0.1973, + "step": 48121 + }, + { + "epoch": 3.3392547359655818, + "grad_norm": 1.1648927927017212, + "learning_rate": 1.3973369815920955e-05, + "loss": 0.3439, + "step": 48122 + }, + { + "epoch": 3.3393241274026786, + "grad_norm": 1.1533949375152588, + "learning_rate": 1.3970505328894423e-05, + "loss": 0.2417, + "step": 48123 + }, + { + "epoch": 3.339393518839775, + "grad_norm": 0.6768150329589844, + "learning_rate": 1.3967641113453023e-05, + "loss": 0.2066, + "step": 48124 + }, + { + "epoch": 3.339462910276872, + "grad_norm": 0.924760639667511, + "learning_rate": 1.3964777169605792e-05, + "loss": 0.1885, + "step": 48125 + }, + { + "epoch": 3.3395323017139686, + "grad_norm": 0.8419272899627686, + "learning_rate": 1.39619134973618e-05, + "loss": 0.214, + "step": 48126 + }, + { + "epoch": 3.339601693151065, + "grad_norm": 1.1740549802780151, + "learning_rate": 1.3959050096730053e-05, + "loss": 0.2903, + "step": 48127 + }, + { + "epoch": 3.339671084588162, + "grad_norm": 0.8019773364067078, + "learning_rate": 1.3956186967719598e-05, + "loss": 0.1589, + "step": 48128 + }, + { + "epoch": 3.3397404760252583, + "grad_norm": 1.0474703311920166, + "learning_rate": 1.3953324110339472e-05, + "loss": 0.2589, + "step": 48129 + }, + { + "epoch": 3.339809867462355, + "grad_norm": 0.7977354526519775, + "learning_rate": 1.3950461524598724e-05, + "loss": 0.2351, + "step": 48130 + }, + { + "epoch": 3.339879258899452, + "grad_norm": 0.6077736616134644, + "learning_rate": 1.39475992105064e-05, + "loss": 0.2327, + "step": 48131 + }, + { + "epoch": 3.3399486503365483, + "grad_norm": 1.4187103509902954, + "learning_rate": 1.394473716807151e-05, + "loss": 0.3037, + "step": 48132 + }, + { + "epoch": 3.340018041773645, + "grad_norm": 1.4532805681228638, + "learning_rate": 1.3941875397303095e-05, + "loss": 0.2195, + "step": 48133 + }, + { + "epoch": 3.340087433210742, + "grad_norm": 1.3388781547546387, + "learning_rate": 1.3939013898210207e-05, + "loss": 0.2362, + "step": 48134 + }, + { + "epoch": 3.3401568246478384, + "grad_norm": 0.7661651968955994, + "learning_rate": 1.3936152670801862e-05, + "loss": 0.2172, + "step": 48135 + }, + { + "epoch": 3.340226216084935, + "grad_norm": 1.3047587871551514, + "learning_rate": 1.3933291715087082e-05, + "loss": 0.4796, + "step": 48136 + }, + { + "epoch": 3.340295607522032, + "grad_norm": 0.8157334923744202, + "learning_rate": 1.3930431031074931e-05, + "loss": 0.2158, + "step": 48137 + }, + { + "epoch": 3.3403649989591284, + "grad_norm": 0.9266569018363953, + "learning_rate": 1.392757061877441e-05, + "loss": 0.3136, + "step": 48138 + }, + { + "epoch": 3.3404343903962253, + "grad_norm": 0.7490397691726685, + "learning_rate": 1.392471047819457e-05, + "loss": 0.2454, + "step": 48139 + }, + { + "epoch": 3.3405037818333216, + "grad_norm": 0.8678736090660095, + "learning_rate": 1.3921850609344455e-05, + "loss": 0.3436, + "step": 48140 + }, + { + "epoch": 3.3405731732704185, + "grad_norm": 0.7375888824462891, + "learning_rate": 1.391899101223304e-05, + "loss": 0.2254, + "step": 48141 + }, + { + "epoch": 3.3406425647075153, + "grad_norm": 1.394112229347229, + "learning_rate": 1.3916131686869393e-05, + "loss": 0.3454, + "step": 48142 + }, + { + "epoch": 3.3407119561446117, + "grad_norm": 1.599429965019226, + "learning_rate": 1.391327263326253e-05, + "loss": 0.2746, + "step": 48143 + }, + { + "epoch": 3.3407813475817085, + "grad_norm": 0.8376618027687073, + "learning_rate": 1.3910413851421466e-05, + "loss": 0.2913, + "step": 48144 + }, + { + "epoch": 3.340850739018805, + "grad_norm": 1.500846028327942, + "learning_rate": 1.390755534135525e-05, + "loss": 0.2963, + "step": 48145 + }, + { + "epoch": 3.3409201304559017, + "grad_norm": 0.994551420211792, + "learning_rate": 1.3904697103072884e-05, + "loss": 0.2153, + "step": 48146 + }, + { + "epoch": 3.3409895218929986, + "grad_norm": 0.7245627045631409, + "learning_rate": 1.3901839136583427e-05, + "loss": 0.2178, + "step": 48147 + }, + { + "epoch": 3.341058913330095, + "grad_norm": 0.791992723941803, + "learning_rate": 1.3898981441895854e-05, + "loss": 0.1799, + "step": 48148 + }, + { + "epoch": 3.341128304767192, + "grad_norm": 0.6809943318367004, + "learning_rate": 1.3896124019019208e-05, + "loss": 0.1218, + "step": 48149 + }, + { + "epoch": 3.341197696204288, + "grad_norm": 0.9392586946487427, + "learning_rate": 1.3893266867962506e-05, + "loss": 0.3029, + "step": 48150 + }, + { + "epoch": 3.341267087641385, + "grad_norm": 0.9163650274276733, + "learning_rate": 1.3890409988734766e-05, + "loss": 0.3225, + "step": 48151 + }, + { + "epoch": 3.341336479078482, + "grad_norm": 1.162047028541565, + "learning_rate": 1.388755338134502e-05, + "loss": 0.3571, + "step": 48152 + }, + { + "epoch": 3.3414058705155782, + "grad_norm": 0.8849802613258362, + "learning_rate": 1.3884697045802274e-05, + "loss": 0.3069, + "step": 48153 + }, + { + "epoch": 3.341475261952675, + "grad_norm": 0.9507158994674683, + "learning_rate": 1.3881840982115569e-05, + "loss": 0.2126, + "step": 48154 + }, + { + "epoch": 3.3415446533897715, + "grad_norm": 0.6484119892120361, + "learning_rate": 1.3878985190293881e-05, + "loss": 0.2691, + "step": 48155 + }, + { + "epoch": 3.3416140448268683, + "grad_norm": 0.809075653553009, + "learning_rate": 1.3876129670346239e-05, + "loss": 0.2132, + "step": 48156 + }, + { + "epoch": 3.341683436263965, + "grad_norm": 1.024187445640564, + "learning_rate": 1.3873274422281668e-05, + "loss": 0.3338, + "step": 48157 + }, + { + "epoch": 3.3417528277010615, + "grad_norm": 0.9394762516021729, + "learning_rate": 1.3870419446109172e-05, + "loss": 0.4211, + "step": 48158 + }, + { + "epoch": 3.3418222191381584, + "grad_norm": 0.7011258006095886, + "learning_rate": 1.3867564741837768e-05, + "loss": 0.2138, + "step": 48159 + }, + { + "epoch": 3.341891610575255, + "grad_norm": 0.9774205088615417, + "learning_rate": 1.3864710309476481e-05, + "loss": 0.1773, + "step": 48160 + }, + { + "epoch": 3.3419610020123516, + "grad_norm": 0.8029956817626953, + "learning_rate": 1.3861856149034325e-05, + "loss": 0.1546, + "step": 48161 + }, + { + "epoch": 3.3420303934494484, + "grad_norm": 1.2211369276046753, + "learning_rate": 1.3859002260520271e-05, + "loss": 0.227, + "step": 48162 + }, + { + "epoch": 3.3420997848865452, + "grad_norm": 1.0312687158584595, + "learning_rate": 1.3856148643943356e-05, + "loss": 0.2833, + "step": 48163 + }, + { + "epoch": 3.3421691763236416, + "grad_norm": 3.113826036453247, + "learning_rate": 1.3853295299312575e-05, + "loss": 0.1594, + "step": 48164 + }, + { + "epoch": 3.3422385677607385, + "grad_norm": 1.186810851097107, + "learning_rate": 1.3850442226636962e-05, + "loss": 0.3616, + "step": 48165 + }, + { + "epoch": 3.342307959197835, + "grad_norm": 0.7036823630332947, + "learning_rate": 1.3847589425925523e-05, + "loss": 0.2764, + "step": 48166 + }, + { + "epoch": 3.3423773506349317, + "grad_norm": 0.87105393409729, + "learning_rate": 1.3844736897187216e-05, + "loss": 0.3197, + "step": 48167 + }, + { + "epoch": 3.3424467420720285, + "grad_norm": 0.7841121554374695, + "learning_rate": 1.3841884640431091e-05, + "loss": 0.1844, + "step": 48168 + }, + { + "epoch": 3.342516133509125, + "grad_norm": 1.0584611892700195, + "learning_rate": 1.3839032655666161e-05, + "loss": 0.3599, + "step": 48169 + }, + { + "epoch": 3.3425855249462217, + "grad_norm": 0.9464345574378967, + "learning_rate": 1.3836180942901388e-05, + "loss": 0.2527, + "step": 48170 + }, + { + "epoch": 3.342654916383318, + "grad_norm": 0.8377383351325989, + "learning_rate": 1.3833329502145786e-05, + "loss": 0.2515, + "step": 48171 + }, + { + "epoch": 3.342724307820415, + "grad_norm": 1.1006048917770386, + "learning_rate": 1.3830478333408369e-05, + "loss": 0.1606, + "step": 48172 + }, + { + "epoch": 3.342793699257512, + "grad_norm": 1.1350138187408447, + "learning_rate": 1.3827627436698143e-05, + "loss": 0.2397, + "step": 48173 + }, + { + "epoch": 3.342863090694608, + "grad_norm": 0.7776743173599243, + "learning_rate": 1.3824776812024087e-05, + "loss": 0.1615, + "step": 48174 + }, + { + "epoch": 3.342932482131705, + "grad_norm": 0.8376541137695312, + "learning_rate": 1.3821926459395251e-05, + "loss": 0.284, + "step": 48175 + }, + { + "epoch": 3.3430018735688014, + "grad_norm": 1.1459290981292725, + "learning_rate": 1.3819076378820562e-05, + "loss": 0.2392, + "step": 48176 + }, + { + "epoch": 3.3430712650058982, + "grad_norm": 0.9531811475753784, + "learning_rate": 1.3816226570309043e-05, + "loss": 0.2118, + "step": 48177 + }, + { + "epoch": 3.343140656442995, + "grad_norm": 0.7670777440071106, + "learning_rate": 1.3813377033869713e-05, + "loss": 0.2446, + "step": 48178 + }, + { + "epoch": 3.3432100478800915, + "grad_norm": 0.8827512860298157, + "learning_rate": 1.3810527769511538e-05, + "loss": 0.2439, + "step": 48179 + }, + { + "epoch": 3.3432794393171883, + "grad_norm": 2.090684413909912, + "learning_rate": 1.3807678777243538e-05, + "loss": 0.3546, + "step": 48180 + }, + { + "epoch": 3.3433488307542847, + "grad_norm": 0.9353664517402649, + "learning_rate": 1.380483005707468e-05, + "loss": 0.1876, + "step": 48181 + }, + { + "epoch": 3.3434182221913815, + "grad_norm": 0.952815055847168, + "learning_rate": 1.3801981609014015e-05, + "loss": 0.3597, + "step": 48182 + }, + { + "epoch": 3.3434876136284783, + "grad_norm": 1.1917250156402588, + "learning_rate": 1.3799133433070455e-05, + "loss": 0.2153, + "step": 48183 + }, + { + "epoch": 3.3435570050655747, + "grad_norm": 1.0318665504455566, + "learning_rate": 1.3796285529253039e-05, + "loss": 0.2621, + "step": 48184 + }, + { + "epoch": 3.3436263965026716, + "grad_norm": 0.8236427903175354, + "learning_rate": 1.3793437897570748e-05, + "loss": 0.369, + "step": 48185 + }, + { + "epoch": 3.3436957879397684, + "grad_norm": 0.833477258682251, + "learning_rate": 1.3790590538032566e-05, + "loss": 0.2484, + "step": 48186 + }, + { + "epoch": 3.343765179376865, + "grad_norm": 0.9530150890350342, + "learning_rate": 1.3787743450647494e-05, + "loss": 0.271, + "step": 48187 + }, + { + "epoch": 3.3438345708139616, + "grad_norm": 1.314374566078186, + "learning_rate": 1.3784896635424539e-05, + "loss": 0.1725, + "step": 48188 + }, + { + "epoch": 3.3439039622510585, + "grad_norm": 0.90867680311203, + "learning_rate": 1.3782050092372622e-05, + "loss": 0.2817, + "step": 48189 + }, + { + "epoch": 3.343973353688155, + "grad_norm": 0.9219579100608826, + "learning_rate": 1.3779203821500786e-05, + "loss": 0.2383, + "step": 48190 + }, + { + "epoch": 3.3440427451252517, + "grad_norm": 1.0506608486175537, + "learning_rate": 1.3776357822817998e-05, + "loss": 0.2446, + "step": 48191 + }, + { + "epoch": 3.344112136562348, + "grad_norm": 0.8343930840492249, + "learning_rate": 1.377351209633324e-05, + "loss": 0.2273, + "step": 48192 + }, + { + "epoch": 3.344181527999445, + "grad_norm": 1.07343590259552, + "learning_rate": 1.3770666642055508e-05, + "loss": 0.2691, + "step": 48193 + }, + { + "epoch": 3.3442509194365417, + "grad_norm": 0.815571129322052, + "learning_rate": 1.3767821459993768e-05, + "loss": 0.2175, + "step": 48194 + }, + { + "epoch": 3.344320310873638, + "grad_norm": 0.8496905565261841, + "learning_rate": 1.3764976550157038e-05, + "loss": 0.2904, + "step": 48195 + }, + { + "epoch": 3.344389702310735, + "grad_norm": 0.966275155544281, + "learning_rate": 1.3762131912554244e-05, + "loss": 0.2817, + "step": 48196 + }, + { + "epoch": 3.3444590937478313, + "grad_norm": 0.7381757497787476, + "learning_rate": 1.3759287547194399e-05, + "loss": 0.2667, + "step": 48197 + }, + { + "epoch": 3.344528485184928, + "grad_norm": 1.0711685419082642, + "learning_rate": 1.3756443454086476e-05, + "loss": 0.3038, + "step": 48198 + }, + { + "epoch": 3.344597876622025, + "grad_norm": 0.9193187355995178, + "learning_rate": 1.3753599633239455e-05, + "loss": 0.161, + "step": 48199 + }, + { + "epoch": 3.3446672680591214, + "grad_norm": 0.9878288507461548, + "learning_rate": 1.3750756084662319e-05, + "loss": 0.2439, + "step": 48200 + }, + { + "epoch": 3.3447366594962182, + "grad_norm": 0.807556688785553, + "learning_rate": 1.374791280836406e-05, + "loss": 0.2548, + "step": 48201 + }, + { + "epoch": 3.3448060509333146, + "grad_norm": 1.1043126583099365, + "learning_rate": 1.3745069804353605e-05, + "loss": 0.3004, + "step": 48202 + }, + { + "epoch": 3.3448754423704115, + "grad_norm": 0.574134111404419, + "learning_rate": 1.3742227072639957e-05, + "loss": 0.1623, + "step": 48203 + }, + { + "epoch": 3.3449448338075083, + "grad_norm": 1.040521264076233, + "learning_rate": 1.373938461323212e-05, + "loss": 0.3001, + "step": 48204 + }, + { + "epoch": 3.3450142252446047, + "grad_norm": 1.1059504747390747, + "learning_rate": 1.373654242613901e-05, + "loss": 0.2625, + "step": 48205 + }, + { + "epoch": 3.3450836166817015, + "grad_norm": 0.8259087204933167, + "learning_rate": 1.373370051136963e-05, + "loss": 0.1923, + "step": 48206 + }, + { + "epoch": 3.345153008118798, + "grad_norm": 1.36689293384552, + "learning_rate": 1.373085886893294e-05, + "loss": 0.478, + "step": 48207 + }, + { + "epoch": 3.3452223995558947, + "grad_norm": 0.9356188774108887, + "learning_rate": 1.3728017498837931e-05, + "loss": 0.3265, + "step": 48208 + }, + { + "epoch": 3.3452917909929916, + "grad_norm": 1.1563657522201538, + "learning_rate": 1.3725176401093577e-05, + "loss": 0.3132, + "step": 48209 + }, + { + "epoch": 3.345361182430088, + "grad_norm": 1.408848762512207, + "learning_rate": 1.372233557570881e-05, + "loss": 0.3246, + "step": 48210 + }, + { + "epoch": 3.345430573867185, + "grad_norm": 0.8457412123680115, + "learning_rate": 1.3719495022692618e-05, + "loss": 0.1585, + "step": 48211 + }, + { + "epoch": 3.3454999653042816, + "grad_norm": 1.188883900642395, + "learning_rate": 1.371665474205398e-05, + "loss": 0.2784, + "step": 48212 + }, + { + "epoch": 3.345569356741378, + "grad_norm": 0.9341204166412354, + "learning_rate": 1.3713814733801844e-05, + "loss": 0.2338, + "step": 48213 + }, + { + "epoch": 3.345638748178475, + "grad_norm": 0.8836957812309265, + "learning_rate": 1.371097499794518e-05, + "loss": 0.2368, + "step": 48214 + }, + { + "epoch": 3.3457081396155717, + "grad_norm": 0.9075614213943481, + "learning_rate": 1.3708135534492961e-05, + "loss": 0.2262, + "step": 48215 + }, + { + "epoch": 3.345777531052668, + "grad_norm": 1.024244785308838, + "learning_rate": 1.3705296343454165e-05, + "loss": 0.3062, + "step": 48216 + }, + { + "epoch": 3.345846922489765, + "grad_norm": 0.7542946338653564, + "learning_rate": 1.3702457424837722e-05, + "loss": 0.2285, + "step": 48217 + }, + { + "epoch": 3.3459163139268613, + "grad_norm": 0.9249489307403564, + "learning_rate": 1.36996187786526e-05, + "loss": 0.2763, + "step": 48218 + }, + { + "epoch": 3.345985705363958, + "grad_norm": 1.2252520322799683, + "learning_rate": 1.3696780404907783e-05, + "loss": 0.2882, + "step": 48219 + }, + { + "epoch": 3.346055096801055, + "grad_norm": 1.1326841115951538, + "learning_rate": 1.3693942303612207e-05, + "loss": 0.3269, + "step": 48220 + }, + { + "epoch": 3.3461244882381513, + "grad_norm": 1.6199288368225098, + "learning_rate": 1.3691104474774841e-05, + "loss": 0.4188, + "step": 48221 + }, + { + "epoch": 3.346193879675248, + "grad_norm": 1.0716780424118042, + "learning_rate": 1.3688266918404646e-05, + "loss": 0.2291, + "step": 48222 + }, + { + "epoch": 3.3462632711123446, + "grad_norm": 1.0604692697525024, + "learning_rate": 1.3685429634510594e-05, + "loss": 0.2406, + "step": 48223 + }, + { + "epoch": 3.3463326625494414, + "grad_norm": 1.3642841577529907, + "learning_rate": 1.368259262310162e-05, + "loss": 0.2222, + "step": 48224 + }, + { + "epoch": 3.3464020539865382, + "grad_norm": 1.1677180528640747, + "learning_rate": 1.3679755884186673e-05, + "loss": 0.251, + "step": 48225 + }, + { + "epoch": 3.3464714454236346, + "grad_norm": 1.643231749534607, + "learning_rate": 1.3676919417774725e-05, + "loss": 0.5223, + "step": 48226 + }, + { + "epoch": 3.3465408368607314, + "grad_norm": 0.7168435454368591, + "learning_rate": 1.3674083223874723e-05, + "loss": 0.1719, + "step": 48227 + }, + { + "epoch": 3.346610228297828, + "grad_norm": 0.7011266946792603, + "learning_rate": 1.3671247302495637e-05, + "loss": 0.2229, + "step": 48228 + }, + { + "epoch": 3.3466796197349247, + "grad_norm": 1.378385305404663, + "learning_rate": 1.3668411653646396e-05, + "loss": 0.3387, + "step": 48229 + }, + { + "epoch": 3.3467490111720215, + "grad_norm": 0.7709914445877075, + "learning_rate": 1.366557627733599e-05, + "loss": 0.2445, + "step": 48230 + }, + { + "epoch": 3.346818402609118, + "grad_norm": 1.397882103919983, + "learning_rate": 1.3662741173573313e-05, + "loss": 0.3128, + "step": 48231 + }, + { + "epoch": 3.3468877940462147, + "grad_norm": 1.0661569833755493, + "learning_rate": 1.3659906342367357e-05, + "loss": 0.3751, + "step": 48232 + }, + { + "epoch": 3.346957185483311, + "grad_norm": 1.0887515544891357, + "learning_rate": 1.3657071783727049e-05, + "loss": 0.3504, + "step": 48233 + }, + { + "epoch": 3.347026576920408, + "grad_norm": 0.7822506427764893, + "learning_rate": 1.365423749766136e-05, + "loss": 0.2166, + "step": 48234 + }, + { + "epoch": 3.347095968357505, + "grad_norm": 0.9197332262992859, + "learning_rate": 1.3651403484179238e-05, + "loss": 0.2606, + "step": 48235 + }, + { + "epoch": 3.347165359794601, + "grad_norm": 1.1304941177368164, + "learning_rate": 1.3648569743289597e-05, + "loss": 0.2577, + "step": 48236 + }, + { + "epoch": 3.347234751231698, + "grad_norm": 0.9473012089729309, + "learning_rate": 1.3645736275001397e-05, + "loss": 0.3073, + "step": 48237 + }, + { + "epoch": 3.347304142668795, + "grad_norm": 0.8303945064544678, + "learning_rate": 1.36429030793236e-05, + "loss": 0.2764, + "step": 48238 + }, + { + "epoch": 3.3473735341058912, + "grad_norm": 0.8676915168762207, + "learning_rate": 1.364007015626515e-05, + "loss": 0.2387, + "step": 48239 + }, + { + "epoch": 3.347442925542988, + "grad_norm": 1.2256581783294678, + "learning_rate": 1.3637237505834966e-05, + "loss": 0.4823, + "step": 48240 + }, + { + "epoch": 3.347512316980085, + "grad_norm": 0.6757882237434387, + "learning_rate": 1.3634405128042005e-05, + "loss": 0.2101, + "step": 48241 + }, + { + "epoch": 3.3475817084171813, + "grad_norm": 0.8474608659744263, + "learning_rate": 1.3631573022895195e-05, + "loss": 0.2182, + "step": 48242 + }, + { + "epoch": 3.347651099854278, + "grad_norm": 1.2676693201065063, + "learning_rate": 1.3628741190403494e-05, + "loss": 0.317, + "step": 48243 + }, + { + "epoch": 3.3477204912913745, + "grad_norm": 0.7850707173347473, + "learning_rate": 1.3625909630575862e-05, + "loss": 0.2528, + "step": 48244 + }, + { + "epoch": 3.3477898827284713, + "grad_norm": 0.675450325012207, + "learning_rate": 1.362307834342118e-05, + "loss": 0.2094, + "step": 48245 + }, + { + "epoch": 3.347859274165568, + "grad_norm": 0.8414197564125061, + "learning_rate": 1.3620247328948432e-05, + "loss": 0.2906, + "step": 48246 + }, + { + "epoch": 3.3479286656026646, + "grad_norm": 0.8229267001152039, + "learning_rate": 1.3617416587166532e-05, + "loss": 0.2773, + "step": 48247 + }, + { + "epoch": 3.3479980570397614, + "grad_norm": 0.8658197522163391, + "learning_rate": 1.3614586118084426e-05, + "loss": 0.2604, + "step": 48248 + }, + { + "epoch": 3.3480674484768578, + "grad_norm": 1.377783179283142, + "learning_rate": 1.3611755921711056e-05, + "loss": 0.3176, + "step": 48249 + }, + { + "epoch": 3.3481368399139546, + "grad_norm": 0.8606918454170227, + "learning_rate": 1.3608925998055343e-05, + "loss": 0.2368, + "step": 48250 + }, + { + "epoch": 3.3482062313510514, + "grad_norm": 0.7328651547431946, + "learning_rate": 1.360609634712625e-05, + "loss": 0.1682, + "step": 48251 + }, + { + "epoch": 3.348275622788148, + "grad_norm": 0.7227756977081299, + "learning_rate": 1.360326696893266e-05, + "loss": 0.2237, + "step": 48252 + }, + { + "epoch": 3.3483450142252447, + "grad_norm": 0.785446286201477, + "learning_rate": 1.3600437863483539e-05, + "loss": 0.2942, + "step": 48253 + }, + { + "epoch": 3.348414405662341, + "grad_norm": 0.6195264458656311, + "learning_rate": 1.3597609030787816e-05, + "loss": 0.2729, + "step": 48254 + }, + { + "epoch": 3.348483797099438, + "grad_norm": 0.7372583150863647, + "learning_rate": 1.3594780470854407e-05, + "loss": 0.195, + "step": 48255 + }, + { + "epoch": 3.3485531885365347, + "grad_norm": 0.7226945161819458, + "learning_rate": 1.3591952183692258e-05, + "loss": 0.2041, + "step": 48256 + }, + { + "epoch": 3.348622579973631, + "grad_norm": 0.7933021783828735, + "learning_rate": 1.3589124169310286e-05, + "loss": 0.254, + "step": 48257 + }, + { + "epoch": 3.348691971410728, + "grad_norm": 0.880479633808136, + "learning_rate": 1.3586296427717448e-05, + "loss": 0.1855, + "step": 48258 + }, + { + "epoch": 3.3487613628478243, + "grad_norm": 1.0013949871063232, + "learning_rate": 1.3583468958922629e-05, + "loss": 0.2519, + "step": 48259 + }, + { + "epoch": 3.348830754284921, + "grad_norm": 0.9713246822357178, + "learning_rate": 1.3580641762934765e-05, + "loss": 0.2564, + "step": 48260 + }, + { + "epoch": 3.348900145722018, + "grad_norm": 0.7511747479438782, + "learning_rate": 1.3577814839762793e-05, + "loss": 0.2011, + "step": 48261 + }, + { + "epoch": 3.3489695371591144, + "grad_norm": 1.4701743125915527, + "learning_rate": 1.3574988189415627e-05, + "loss": 0.3529, + "step": 48262 + }, + { + "epoch": 3.349038928596211, + "grad_norm": 0.7954042553901672, + "learning_rate": 1.3572161811902195e-05, + "loss": 0.2292, + "step": 48263 + }, + { + "epoch": 3.349108320033308, + "grad_norm": 1.3613704442977905, + "learning_rate": 1.3569335707231457e-05, + "loss": 0.2937, + "step": 48264 + }, + { + "epoch": 3.3491777114704044, + "grad_norm": 0.7858920693397522, + "learning_rate": 1.3566509875412259e-05, + "loss": 0.2406, + "step": 48265 + }, + { + "epoch": 3.3492471029075013, + "grad_norm": 1.0404225587844849, + "learning_rate": 1.3563684316453574e-05, + "loss": 0.2278, + "step": 48266 + }, + { + "epoch": 3.349316494344598, + "grad_norm": 1.3664615154266357, + "learning_rate": 1.3560859030364304e-05, + "loss": 0.4066, + "step": 48267 + }, + { + "epoch": 3.3493858857816945, + "grad_norm": 0.8965705633163452, + "learning_rate": 1.3558034017153375e-05, + "loss": 0.2603, + "step": 48268 + }, + { + "epoch": 3.3494552772187913, + "grad_norm": 0.9807709455490112, + "learning_rate": 1.3555209276829695e-05, + "loss": 0.3026, + "step": 48269 + }, + { + "epoch": 3.3495246686558877, + "grad_norm": 1.0301368236541748, + "learning_rate": 1.3552384809402218e-05, + "loss": 0.1443, + "step": 48270 + }, + { + "epoch": 3.3495940600929845, + "grad_norm": 0.9391477704048157, + "learning_rate": 1.354956061487982e-05, + "loss": 0.2054, + "step": 48271 + }, + { + "epoch": 3.3496634515300814, + "grad_norm": 1.143181324005127, + "learning_rate": 1.3546736693271423e-05, + "loss": 0.2453, + "step": 48272 + }, + { + "epoch": 3.3497328429671778, + "grad_norm": 0.965050995349884, + "learning_rate": 1.3543913044585942e-05, + "loss": 0.2325, + "step": 48273 + }, + { + "epoch": 3.3498022344042746, + "grad_norm": 1.6020410060882568, + "learning_rate": 1.354108966883232e-05, + "loss": 0.2321, + "step": 48274 + }, + { + "epoch": 3.349871625841371, + "grad_norm": 0.6810035705566406, + "learning_rate": 1.353826656601943e-05, + "loss": 0.1805, + "step": 48275 + }, + { + "epoch": 3.349941017278468, + "grad_norm": 0.6647348403930664, + "learning_rate": 1.3535443736156194e-05, + "loss": 0.2775, + "step": 48276 + }, + { + "epoch": 3.3500104087155647, + "grad_norm": 1.5690861940383911, + "learning_rate": 1.3532621179251537e-05, + "loss": 0.3422, + "step": 48277 + }, + { + "epoch": 3.350079800152661, + "grad_norm": 0.9080774784088135, + "learning_rate": 1.3529798895314383e-05, + "loss": 0.2042, + "step": 48278 + }, + { + "epoch": 3.350149191589758, + "grad_norm": 1.2036793231964111, + "learning_rate": 1.3526976884353592e-05, + "loss": 0.1816, + "step": 48279 + }, + { + "epoch": 3.3502185830268543, + "grad_norm": 0.8571022748947144, + "learning_rate": 1.3524155146378115e-05, + "loss": 0.2396, + "step": 48280 + }, + { + "epoch": 3.350287974463951, + "grad_norm": 0.555471658706665, + "learning_rate": 1.352133368139683e-05, + "loss": 0.1289, + "step": 48281 + }, + { + "epoch": 3.350357365901048, + "grad_norm": 0.7506660223007202, + "learning_rate": 1.3518512489418678e-05, + "loss": 0.1777, + "step": 48282 + }, + { + "epoch": 3.3504267573381443, + "grad_norm": 0.7762079238891602, + "learning_rate": 1.3515691570452538e-05, + "loss": 0.1751, + "step": 48283 + }, + { + "epoch": 3.350496148775241, + "grad_norm": 0.7894554138183594, + "learning_rate": 1.3512870924507338e-05, + "loss": 0.2269, + "step": 48284 + }, + { + "epoch": 3.350565540212338, + "grad_norm": 1.0447607040405273, + "learning_rate": 1.351005055159198e-05, + "loss": 0.4139, + "step": 48285 + }, + { + "epoch": 3.3506349316494344, + "grad_norm": 1.0533487796783447, + "learning_rate": 1.3507230451715336e-05, + "loss": 0.3395, + "step": 48286 + }, + { + "epoch": 3.350704323086531, + "grad_norm": 1.1515382528305054, + "learning_rate": 1.3504410624886344e-05, + "loss": 0.3446, + "step": 48287 + }, + { + "epoch": 3.3507737145236276, + "grad_norm": 0.7817995548248291, + "learning_rate": 1.3501591071113873e-05, + "loss": 0.2707, + "step": 48288 + }, + { + "epoch": 3.3508431059607244, + "grad_norm": 1.0324639081954956, + "learning_rate": 1.3498771790406861e-05, + "loss": 0.2592, + "step": 48289 + }, + { + "epoch": 3.3509124973978213, + "grad_norm": 1.0841915607452393, + "learning_rate": 1.349595278277418e-05, + "loss": 0.3195, + "step": 48290 + }, + { + "epoch": 3.3509818888349177, + "grad_norm": 0.6525591015815735, + "learning_rate": 1.3493134048224743e-05, + "loss": 0.2201, + "step": 48291 + }, + { + "epoch": 3.3510512802720145, + "grad_norm": 0.9595968723297119, + "learning_rate": 1.3490315586767466e-05, + "loss": 0.3352, + "step": 48292 + }, + { + "epoch": 3.3511206717091113, + "grad_norm": 0.8535409569740295, + "learning_rate": 1.3487497398411197e-05, + "loss": 0.2245, + "step": 48293 + }, + { + "epoch": 3.3511900631462077, + "grad_norm": 0.7468151450157166, + "learning_rate": 1.3484679483164875e-05, + "loss": 0.294, + "step": 48294 + }, + { + "epoch": 3.3512594545833045, + "grad_norm": 0.8972082734107971, + "learning_rate": 1.3481861841037369e-05, + "loss": 0.2227, + "step": 48295 + }, + { + "epoch": 3.351328846020401, + "grad_norm": 0.8341959118843079, + "learning_rate": 1.3479044472037593e-05, + "loss": 0.1662, + "step": 48296 + }, + { + "epoch": 3.3513982374574978, + "grad_norm": 0.905272901058197, + "learning_rate": 1.3476227376174444e-05, + "loss": 0.2194, + "step": 48297 + }, + { + "epoch": 3.3514676288945946, + "grad_norm": 0.9598821401596069, + "learning_rate": 1.34734105534568e-05, + "loss": 0.2894, + "step": 48298 + }, + { + "epoch": 3.351537020331691, + "grad_norm": 0.6928517818450928, + "learning_rate": 1.3470594003893578e-05, + "loss": 0.1739, + "step": 48299 + }, + { + "epoch": 3.351606411768788, + "grad_norm": 0.9085513353347778, + "learning_rate": 1.3467777727493636e-05, + "loss": 0.3127, + "step": 48300 + }, + { + "epoch": 3.351675803205884, + "grad_norm": 0.7640461325645447, + "learning_rate": 1.3464961724265868e-05, + "loss": 0.2554, + "step": 48301 + }, + { + "epoch": 3.351745194642981, + "grad_norm": 0.7558988928794861, + "learning_rate": 1.3462145994219188e-05, + "loss": 0.1903, + "step": 48302 + }, + { + "epoch": 3.351814586080078, + "grad_norm": 0.8365745544433594, + "learning_rate": 1.345933053736247e-05, + "loss": 0.2401, + "step": 48303 + }, + { + "epoch": 3.3518839775171743, + "grad_norm": 1.210605263710022, + "learning_rate": 1.3456515353704603e-05, + "loss": 0.432, + "step": 48304 + }, + { + "epoch": 3.351953368954271, + "grad_norm": 0.7293115854263306, + "learning_rate": 1.3453700443254503e-05, + "loss": 0.1586, + "step": 48305 + }, + { + "epoch": 3.3520227603913675, + "grad_norm": 1.0352637767791748, + "learning_rate": 1.3450885806020997e-05, + "loss": 0.2107, + "step": 48306 + }, + { + "epoch": 3.3520921518284643, + "grad_norm": 0.9769685864448547, + "learning_rate": 1.344807144201301e-05, + "loss": 0.2239, + "step": 48307 + }, + { + "epoch": 3.352161543265561, + "grad_norm": 0.7073196172714233, + "learning_rate": 1.3445257351239415e-05, + "loss": 0.1803, + "step": 48308 + }, + { + "epoch": 3.3522309347026575, + "grad_norm": 1.8366796970367432, + "learning_rate": 1.3442443533709137e-05, + "loss": 0.1856, + "step": 48309 + }, + { + "epoch": 3.3523003261397544, + "grad_norm": 0.8339884877204895, + "learning_rate": 1.343962998943098e-05, + "loss": 0.1634, + "step": 48310 + }, + { + "epoch": 3.352369717576851, + "grad_norm": 0.8493346571922302, + "learning_rate": 1.3436816718413881e-05, + "loss": 0.2016, + "step": 48311 + }, + { + "epoch": 3.3524391090139476, + "grad_norm": 1.0533610582351685, + "learning_rate": 1.343400372066671e-05, + "loss": 0.2809, + "step": 48312 + }, + { + "epoch": 3.3525085004510444, + "grad_norm": 1.0046755075454712, + "learning_rate": 1.3431190996198362e-05, + "loss": 0.2312, + "step": 48313 + }, + { + "epoch": 3.352577891888141, + "grad_norm": 1.1504530906677246, + "learning_rate": 1.3428378545017673e-05, + "loss": 0.2928, + "step": 48314 + }, + { + "epoch": 3.3526472833252376, + "grad_norm": 1.7133910655975342, + "learning_rate": 1.3425566367133557e-05, + "loss": 0.3602, + "step": 48315 + }, + { + "epoch": 3.3527166747623345, + "grad_norm": 0.7641985416412354, + "learning_rate": 1.3422754462554888e-05, + "loss": 0.2126, + "step": 48316 + }, + { + "epoch": 3.352786066199431, + "grad_norm": 1.224518895149231, + "learning_rate": 1.3419942831290521e-05, + "loss": 0.2187, + "step": 48317 + }, + { + "epoch": 3.3528554576365277, + "grad_norm": 2.536303997039795, + "learning_rate": 1.3417131473349365e-05, + "loss": 0.3643, + "step": 48318 + }, + { + "epoch": 3.3529248490736245, + "grad_norm": 1.0742837190628052, + "learning_rate": 1.3414320388740275e-05, + "loss": 0.2217, + "step": 48319 + }, + { + "epoch": 3.352994240510721, + "grad_norm": 1.036574125289917, + "learning_rate": 1.3411509577472147e-05, + "loss": 0.2246, + "step": 48320 + }, + { + "epoch": 3.3530636319478178, + "grad_norm": 0.7763793468475342, + "learning_rate": 1.340869903955383e-05, + "loss": 0.1978, + "step": 48321 + }, + { + "epoch": 3.353133023384914, + "grad_norm": 0.8564006686210632, + "learning_rate": 1.3405888774994191e-05, + "loss": 0.256, + "step": 48322 + }, + { + "epoch": 3.353202414822011, + "grad_norm": 0.9738046526908875, + "learning_rate": 1.3403078783802114e-05, + "loss": 0.2444, + "step": 48323 + }, + { + "epoch": 3.353271806259108, + "grad_norm": 0.6856417655944824, + "learning_rate": 1.340026906598647e-05, + "loss": 0.2214, + "step": 48324 + }, + { + "epoch": 3.353341197696204, + "grad_norm": 0.8403124809265137, + "learning_rate": 1.339745962155613e-05, + "loss": 0.2403, + "step": 48325 + }, + { + "epoch": 3.353410589133301, + "grad_norm": 1.492539882659912, + "learning_rate": 1.3394650450519962e-05, + "loss": 0.2825, + "step": 48326 + }, + { + "epoch": 3.3534799805703974, + "grad_norm": 0.8542996644973755, + "learning_rate": 1.3391841552886863e-05, + "loss": 0.2289, + "step": 48327 + }, + { + "epoch": 3.3535493720074943, + "grad_norm": 1.048352599143982, + "learning_rate": 1.3389032928665645e-05, + "loss": 0.1995, + "step": 48328 + }, + { + "epoch": 3.353618763444591, + "grad_norm": 1.0946626663208008, + "learning_rate": 1.3386224577865202e-05, + "loss": 0.2721, + "step": 48329 + }, + { + "epoch": 3.3536881548816875, + "grad_norm": 0.9086869955062866, + "learning_rate": 1.3383416500494394e-05, + "loss": 0.1379, + "step": 48330 + }, + { + "epoch": 3.3537575463187843, + "grad_norm": 0.797617495059967, + "learning_rate": 1.3380608696562102e-05, + "loss": 0.2161, + "step": 48331 + }, + { + "epoch": 3.3538269377558807, + "grad_norm": 0.6143446564674377, + "learning_rate": 1.3377801166077176e-05, + "loss": 0.1883, + "step": 48332 + }, + { + "epoch": 3.3538963291929775, + "grad_norm": 0.8281943202018738, + "learning_rate": 1.3374993909048472e-05, + "loss": 0.2554, + "step": 48333 + }, + { + "epoch": 3.3539657206300744, + "grad_norm": 1.2255052328109741, + "learning_rate": 1.3372186925484887e-05, + "loss": 0.4042, + "step": 48334 + }, + { + "epoch": 3.3540351120671708, + "grad_norm": 0.8879346251487732, + "learning_rate": 1.3369380215395244e-05, + "loss": 0.3025, + "step": 48335 + }, + { + "epoch": 3.3541045035042676, + "grad_norm": 0.9158664345741272, + "learning_rate": 1.3366573778788405e-05, + "loss": 0.2599, + "step": 48336 + }, + { + "epoch": 3.3541738949413644, + "grad_norm": 0.7272934317588806, + "learning_rate": 1.336376761567325e-05, + "loss": 0.2056, + "step": 48337 + }, + { + "epoch": 3.354243286378461, + "grad_norm": 1.4183440208435059, + "learning_rate": 1.3360961726058629e-05, + "loss": 0.2644, + "step": 48338 + }, + { + "epoch": 3.3543126778155576, + "grad_norm": 0.7788223028182983, + "learning_rate": 1.335815610995339e-05, + "loss": 0.1654, + "step": 48339 + }, + { + "epoch": 3.354382069252654, + "grad_norm": 1.1918995380401611, + "learning_rate": 1.3355350767366426e-05, + "loss": 0.3057, + "step": 48340 + }, + { + "epoch": 3.354451460689751, + "grad_norm": 0.9577654600143433, + "learning_rate": 1.3352545698306551e-05, + "loss": 0.1573, + "step": 48341 + }, + { + "epoch": 3.3545208521268477, + "grad_norm": 1.1361438035964966, + "learning_rate": 1.3349740902782636e-05, + "loss": 0.4334, + "step": 48342 + }, + { + "epoch": 3.354590243563944, + "grad_norm": 1.0733481645584106, + "learning_rate": 1.3346936380803532e-05, + "loss": 0.285, + "step": 48343 + }, + { + "epoch": 3.354659635001041, + "grad_norm": 0.6430225968360901, + "learning_rate": 1.3344132132378118e-05, + "loss": 0.126, + "step": 48344 + }, + { + "epoch": 3.3547290264381378, + "grad_norm": 0.8364899754524231, + "learning_rate": 1.3341328157515198e-05, + "loss": 0.1168, + "step": 48345 + }, + { + "epoch": 3.354798417875234, + "grad_norm": 0.9942260980606079, + "learning_rate": 1.3338524456223655e-05, + "loss": 0.2947, + "step": 48346 + }, + { + "epoch": 3.354867809312331, + "grad_norm": 0.6962217092514038, + "learning_rate": 1.333572102851236e-05, + "loss": 0.1936, + "step": 48347 + }, + { + "epoch": 3.3549372007494274, + "grad_norm": 1.3466976881027222, + "learning_rate": 1.3332917874390116e-05, + "loss": 0.4371, + "step": 48348 + }, + { + "epoch": 3.355006592186524, + "grad_norm": 0.9097145795822144, + "learning_rate": 1.3330114993865795e-05, + "loss": 0.2294, + "step": 48349 + }, + { + "epoch": 3.355075983623621, + "grad_norm": 1.7559914588928223, + "learning_rate": 1.3327312386948243e-05, + "loss": 0.4224, + "step": 48350 + }, + { + "epoch": 3.3551453750607174, + "grad_norm": 1.0863075256347656, + "learning_rate": 1.3324510053646312e-05, + "loss": 0.2234, + "step": 48351 + }, + { + "epoch": 3.3552147664978142, + "grad_norm": 0.6966812014579773, + "learning_rate": 1.3321707993968846e-05, + "loss": 0.1632, + "step": 48352 + }, + { + "epoch": 3.3552841579349106, + "grad_norm": 1.0683823823928833, + "learning_rate": 1.3318906207924698e-05, + "loss": 0.323, + "step": 48353 + }, + { + "epoch": 3.3553535493720075, + "grad_norm": 1.0157973766326904, + "learning_rate": 1.3316104695522724e-05, + "loss": 0.2711, + "step": 48354 + }, + { + "epoch": 3.3554229408091043, + "grad_norm": 1.2424732446670532, + "learning_rate": 1.3313303456771719e-05, + "loss": 0.5006, + "step": 48355 + }, + { + "epoch": 3.3554923322462007, + "grad_norm": 1.070422649383545, + "learning_rate": 1.3310502491680565e-05, + "loss": 0.3352, + "step": 48356 + }, + { + "epoch": 3.3555617236832975, + "grad_norm": 0.7206574082374573, + "learning_rate": 1.33077018002581e-05, + "loss": 0.1855, + "step": 48357 + }, + { + "epoch": 3.355631115120394, + "grad_norm": 1.0211834907531738, + "learning_rate": 1.3304901382513157e-05, + "loss": 0.2086, + "step": 48358 + }, + { + "epoch": 3.3557005065574907, + "grad_norm": 0.7406349182128906, + "learning_rate": 1.3302101238454578e-05, + "loss": 0.2568, + "step": 48359 + }, + { + "epoch": 3.3557698979945876, + "grad_norm": 3.015200138092041, + "learning_rate": 1.3299301368091221e-05, + "loss": 0.3563, + "step": 48360 + }, + { + "epoch": 3.355839289431684, + "grad_norm": 1.8896081447601318, + "learning_rate": 1.3296501771431924e-05, + "loss": 0.3454, + "step": 48361 + }, + { + "epoch": 3.355908680868781, + "grad_norm": 1.2175565958023071, + "learning_rate": 1.329370244848549e-05, + "loss": 0.2672, + "step": 48362 + }, + { + "epoch": 3.3559780723058776, + "grad_norm": 1.4162800312042236, + "learning_rate": 1.3290903399260769e-05, + "loss": 0.2324, + "step": 48363 + }, + { + "epoch": 3.356047463742974, + "grad_norm": 1.1023179292678833, + "learning_rate": 1.3288104623766618e-05, + "loss": 0.2431, + "step": 48364 + }, + { + "epoch": 3.356116855180071, + "grad_norm": 0.9142754077911377, + "learning_rate": 1.3285306122011853e-05, + "loss": 0.2658, + "step": 48365 + }, + { + "epoch": 3.3561862466171677, + "grad_norm": 0.8143202662467957, + "learning_rate": 1.3282507894005325e-05, + "loss": 0.2518, + "step": 48366 + }, + { + "epoch": 3.356255638054264, + "grad_norm": 1.1342591047286987, + "learning_rate": 1.3279709939755847e-05, + "loss": 0.2765, + "step": 48367 + }, + { + "epoch": 3.356325029491361, + "grad_norm": 1.0456957817077637, + "learning_rate": 1.327691225927229e-05, + "loss": 0.2515, + "step": 48368 + }, + { + "epoch": 3.3563944209284573, + "grad_norm": 1.0562597513198853, + "learning_rate": 1.3274114852563436e-05, + "loss": 0.3319, + "step": 48369 + }, + { + "epoch": 3.356463812365554, + "grad_norm": 2.1070070266723633, + "learning_rate": 1.3271317719638143e-05, + "loss": 0.1872, + "step": 48370 + }, + { + "epoch": 3.356533203802651, + "grad_norm": 0.6804697513580322, + "learning_rate": 1.326852086050524e-05, + "loss": 0.2313, + "step": 48371 + }, + { + "epoch": 3.3566025952397474, + "grad_norm": 0.8330220580101013, + "learning_rate": 1.3265724275173542e-05, + "loss": 0.1481, + "step": 48372 + }, + { + "epoch": 3.356671986676844, + "grad_norm": 0.8606569170951843, + "learning_rate": 1.3262927963651895e-05, + "loss": 0.2428, + "step": 48373 + }, + { + "epoch": 3.3567413781139406, + "grad_norm": 0.9722488522529602, + "learning_rate": 1.3260131925949127e-05, + "loss": 0.3738, + "step": 48374 + }, + { + "epoch": 3.3568107695510374, + "grad_norm": 1.4786696434020996, + "learning_rate": 1.3257336162074074e-05, + "loss": 0.5537, + "step": 48375 + }, + { + "epoch": 3.3568801609881342, + "grad_norm": 1.7356542348861694, + "learning_rate": 1.3254540672035531e-05, + "loss": 0.1799, + "step": 48376 + }, + { + "epoch": 3.3569495524252306, + "grad_norm": 0.9447786211967468, + "learning_rate": 1.3251745455842334e-05, + "loss": 0.3347, + "step": 48377 + }, + { + "epoch": 3.3570189438623275, + "grad_norm": 1.1985797882080078, + "learning_rate": 1.3248950513503332e-05, + "loss": 0.157, + "step": 48378 + }, + { + "epoch": 3.357088335299424, + "grad_norm": 2.015897274017334, + "learning_rate": 1.324615584502732e-05, + "loss": 0.4181, + "step": 48379 + }, + { + "epoch": 3.3571577267365207, + "grad_norm": 0.8407579064369202, + "learning_rate": 1.3243361450423108e-05, + "loss": 0.2322, + "step": 48380 + }, + { + "epoch": 3.3572271181736175, + "grad_norm": 0.8565247654914856, + "learning_rate": 1.324056732969955e-05, + "loss": 0.1674, + "step": 48381 + }, + { + "epoch": 3.357296509610714, + "grad_norm": 1.053499460220337, + "learning_rate": 1.3237773482865479e-05, + "loss": 0.153, + "step": 48382 + }, + { + "epoch": 3.3573659010478107, + "grad_norm": 0.6011362075805664, + "learning_rate": 1.3234979909929667e-05, + "loss": 0.12, + "step": 48383 + }, + { + "epoch": 3.357435292484907, + "grad_norm": 1.2411518096923828, + "learning_rate": 1.3232186610900955e-05, + "loss": 0.3155, + "step": 48384 + }, + { + "epoch": 3.357504683922004, + "grad_norm": 0.7208390235900879, + "learning_rate": 1.3229393585788164e-05, + "loss": 0.212, + "step": 48385 + }, + { + "epoch": 3.357574075359101, + "grad_norm": 1.2007988691329956, + "learning_rate": 1.32266008346001e-05, + "loss": 0.3177, + "step": 48386 + }, + { + "epoch": 3.357643466796197, + "grad_norm": 0.7218056321144104, + "learning_rate": 1.3223808357345601e-05, + "loss": 0.2436, + "step": 48387 + }, + { + "epoch": 3.357712858233294, + "grad_norm": 0.9319543242454529, + "learning_rate": 1.3221016154033471e-05, + "loss": 0.2841, + "step": 48388 + }, + { + "epoch": 3.357782249670391, + "grad_norm": 0.9123266339302063, + "learning_rate": 1.3218224224672549e-05, + "loss": 0.2207, + "step": 48389 + }, + { + "epoch": 3.3578516411074872, + "grad_norm": 0.636526346206665, + "learning_rate": 1.321543256927159e-05, + "loss": 0.2752, + "step": 48390 + }, + { + "epoch": 3.357921032544584, + "grad_norm": 0.810132622718811, + "learning_rate": 1.3212641187839458e-05, + "loss": 0.1723, + "step": 48391 + }, + { + "epoch": 3.357990423981681, + "grad_norm": 0.7331801652908325, + "learning_rate": 1.3209850080384933e-05, + "loss": 0.1893, + "step": 48392 + }, + { + "epoch": 3.3580598154187773, + "grad_norm": 1.1124308109283447, + "learning_rate": 1.3207059246916854e-05, + "loss": 0.2305, + "step": 48393 + }, + { + "epoch": 3.358129206855874, + "grad_norm": 0.8445762991905212, + "learning_rate": 1.320426868744401e-05, + "loss": 0.2784, + "step": 48394 + }, + { + "epoch": 3.3581985982929705, + "grad_norm": 0.7956162691116333, + "learning_rate": 1.3201478401975232e-05, + "loss": 0.2167, + "step": 48395 + }, + { + "epoch": 3.3582679897300673, + "grad_norm": 1.015192985534668, + "learning_rate": 1.3198688390519321e-05, + "loss": 0.2473, + "step": 48396 + }, + { + "epoch": 3.358337381167164, + "grad_norm": 1.2878859043121338, + "learning_rate": 1.319589865308507e-05, + "loss": 0.2056, + "step": 48397 + }, + { + "epoch": 3.3584067726042606, + "grad_norm": 0.9835391044616699, + "learning_rate": 1.3193109189681296e-05, + "loss": 0.2867, + "step": 48398 + }, + { + "epoch": 3.3584761640413574, + "grad_norm": 0.9252178072929382, + "learning_rate": 1.3190320000316803e-05, + "loss": 0.2731, + "step": 48399 + }, + { + "epoch": 3.358545555478454, + "grad_norm": 0.8411034345626831, + "learning_rate": 1.3187531085000393e-05, + "loss": 0.286, + "step": 48400 + }, + { + "epoch": 3.3586149469155506, + "grad_norm": 0.8771146535873413, + "learning_rate": 1.3184742443740872e-05, + "loss": 0.2998, + "step": 48401 + }, + { + "epoch": 3.3586843383526475, + "grad_norm": 0.7070238590240479, + "learning_rate": 1.3181954076547055e-05, + "loss": 0.2367, + "step": 48402 + }, + { + "epoch": 3.358753729789744, + "grad_norm": 0.8295432329177856, + "learning_rate": 1.3179165983427754e-05, + "loss": 0.2708, + "step": 48403 + }, + { + "epoch": 3.3588231212268407, + "grad_norm": 1.4398795366287231, + "learning_rate": 1.3176378164391746e-05, + "loss": 0.2106, + "step": 48404 + }, + { + "epoch": 3.358892512663937, + "grad_norm": 0.9741045832633972, + "learning_rate": 1.3173590619447818e-05, + "loss": 0.239, + "step": 48405 + }, + { + "epoch": 3.358961904101034, + "grad_norm": 1.1546432971954346, + "learning_rate": 1.31708033486048e-05, + "loss": 0.1801, + "step": 48406 + }, + { + "epoch": 3.3590312955381307, + "grad_norm": 0.8075150847434998, + "learning_rate": 1.3168016351871481e-05, + "loss": 0.2554, + "step": 48407 + }, + { + "epoch": 3.359100686975227, + "grad_norm": 0.8674546480178833, + "learning_rate": 1.316522962925666e-05, + "loss": 0.2736, + "step": 48408 + }, + { + "epoch": 3.359170078412324, + "grad_norm": 0.7440155148506165, + "learning_rate": 1.3162443180769135e-05, + "loss": 0.2016, + "step": 48409 + }, + { + "epoch": 3.3592394698494203, + "grad_norm": 0.8945303559303284, + "learning_rate": 1.3159657006417724e-05, + "loss": 0.261, + "step": 48410 + }, + { + "epoch": 3.359308861286517, + "grad_norm": 0.9242152571678162, + "learning_rate": 1.3156871106211177e-05, + "loss": 0.2483, + "step": 48411 + }, + { + "epoch": 3.359378252723614, + "grad_norm": 0.9173086881637573, + "learning_rate": 1.3154085480158306e-05, + "loss": 0.2583, + "step": 48412 + }, + { + "epoch": 3.3594476441607104, + "grad_norm": 0.8853099346160889, + "learning_rate": 1.315130012826794e-05, + "loss": 0.2669, + "step": 48413 + }, + { + "epoch": 3.3595170355978072, + "grad_norm": 0.9427335858345032, + "learning_rate": 1.3148515050548826e-05, + "loss": 0.2857, + "step": 48414 + }, + { + "epoch": 3.359586427034904, + "grad_norm": 0.8323582410812378, + "learning_rate": 1.3145730247009768e-05, + "loss": 0.2628, + "step": 48415 + }, + { + "epoch": 3.3596558184720005, + "grad_norm": 0.9290018677711487, + "learning_rate": 1.314294571765956e-05, + "loss": 0.376, + "step": 48416 + }, + { + "epoch": 3.3597252099090973, + "grad_norm": 0.781200647354126, + "learning_rate": 1.3140161462507028e-05, + "loss": 0.1494, + "step": 48417 + }, + { + "epoch": 3.359794601346194, + "grad_norm": 1.2293976545333862, + "learning_rate": 1.3137377481560898e-05, + "loss": 0.3161, + "step": 48418 + }, + { + "epoch": 3.3598639927832905, + "grad_norm": 1.1347272396087646, + "learning_rate": 1.3134593774829984e-05, + "loss": 0.4095, + "step": 48419 + }, + { + "epoch": 3.3599333842203873, + "grad_norm": 0.7799042463302612, + "learning_rate": 1.3131810342323091e-05, + "loss": 0.2847, + "step": 48420 + }, + { + "epoch": 3.3600027756574837, + "grad_norm": 1.3226995468139648, + "learning_rate": 1.3129027184048982e-05, + "loss": 0.4318, + "step": 48421 + }, + { + "epoch": 3.3600721670945806, + "grad_norm": 0.8550728559494019, + "learning_rate": 1.3126244300016466e-05, + "loss": 0.1884, + "step": 48422 + }, + { + "epoch": 3.3601415585316774, + "grad_norm": 0.6838469505310059, + "learning_rate": 1.312346169023434e-05, + "loss": 0.2897, + "step": 48423 + }, + { + "epoch": 3.360210949968774, + "grad_norm": 1.3644682168960571, + "learning_rate": 1.3120679354711329e-05, + "loss": 0.3356, + "step": 48424 + }, + { + "epoch": 3.3602803414058706, + "grad_norm": 1.0576283931732178, + "learning_rate": 1.3117897293456271e-05, + "loss": 0.2836, + "step": 48425 + }, + { + "epoch": 3.360349732842967, + "grad_norm": 0.9515995979309082, + "learning_rate": 1.3115115506477927e-05, + "loss": 0.2134, + "step": 48426 + }, + { + "epoch": 3.360419124280064, + "grad_norm": 1.0229114294052124, + "learning_rate": 1.3112333993785075e-05, + "loss": 0.3456, + "step": 48427 + }, + { + "epoch": 3.3604885157171607, + "grad_norm": 0.8798730969429016, + "learning_rate": 1.3109552755386511e-05, + "loss": 0.32, + "step": 48428 + }, + { + "epoch": 3.360557907154257, + "grad_norm": 0.8093754649162292, + "learning_rate": 1.3106771791291005e-05, + "loss": 0.2439, + "step": 48429 + }, + { + "epoch": 3.360627298591354, + "grad_norm": 0.758271336555481, + "learning_rate": 1.3103991101507352e-05, + "loss": 0.2377, + "step": 48430 + }, + { + "epoch": 3.3606966900284503, + "grad_norm": 0.6551881432533264, + "learning_rate": 1.3101210686044307e-05, + "loss": 0.2056, + "step": 48431 + }, + { + "epoch": 3.360766081465547, + "grad_norm": 1.2517009973526, + "learning_rate": 1.3098430544910656e-05, + "loss": 0.293, + "step": 48432 + }, + { + "epoch": 3.360835472902644, + "grad_norm": 0.6838763952255249, + "learning_rate": 1.309565067811518e-05, + "loss": 0.184, + "step": 48433 + }, + { + "epoch": 3.3609048643397403, + "grad_norm": 1.2642842531204224, + "learning_rate": 1.3092871085666648e-05, + "loss": 0.2421, + "step": 48434 + }, + { + "epoch": 3.360974255776837, + "grad_norm": 1.2800915241241455, + "learning_rate": 1.3090091767573843e-05, + "loss": 0.4518, + "step": 48435 + }, + { + "epoch": 3.3610436472139336, + "grad_norm": 1.2096502780914307, + "learning_rate": 1.3087312723845524e-05, + "loss": 0.3153, + "step": 48436 + }, + { + "epoch": 3.3611130386510304, + "grad_norm": 1.0234854221343994, + "learning_rate": 1.3084533954490508e-05, + "loss": 0.2759, + "step": 48437 + }, + { + "epoch": 3.3611824300881272, + "grad_norm": 0.7715692520141602, + "learning_rate": 1.3081755459517508e-05, + "loss": 0.185, + "step": 48438 + }, + { + "epoch": 3.3612518215252236, + "grad_norm": 0.8120748996734619, + "learning_rate": 1.307897723893532e-05, + "loss": 0.2428, + "step": 48439 + }, + { + "epoch": 3.3613212129623204, + "grad_norm": 0.7991036176681519, + "learning_rate": 1.3076199292752722e-05, + "loss": 0.1997, + "step": 48440 + }, + { + "epoch": 3.3613906043994173, + "grad_norm": 0.9423400163650513, + "learning_rate": 1.3073421620978465e-05, + "loss": 0.357, + "step": 48441 + }, + { + "epoch": 3.3614599958365137, + "grad_norm": 1.039372444152832, + "learning_rate": 1.3070644223621342e-05, + "loss": 0.273, + "step": 48442 + }, + { + "epoch": 3.3615293872736105, + "grad_norm": 0.9966158866882324, + "learning_rate": 1.3067867100690112e-05, + "loss": 0.2521, + "step": 48443 + }, + { + "epoch": 3.3615987787107073, + "grad_norm": 1.082396149635315, + "learning_rate": 1.3065090252193557e-05, + "loss": 0.2801, + "step": 48444 + }, + { + "epoch": 3.3616681701478037, + "grad_norm": 1.0535039901733398, + "learning_rate": 1.3062313678140403e-05, + "loss": 0.3084, + "step": 48445 + }, + { + "epoch": 3.3617375615849006, + "grad_norm": 1.033685564994812, + "learning_rate": 1.3059537378539433e-05, + "loss": 0.192, + "step": 48446 + }, + { + "epoch": 3.361806953021997, + "grad_norm": 0.8156396746635437, + "learning_rate": 1.3056761353399427e-05, + "loss": 0.2059, + "step": 48447 + }, + { + "epoch": 3.361876344459094, + "grad_norm": 0.9546821713447571, + "learning_rate": 1.3053985602729158e-05, + "loss": 0.2284, + "step": 48448 + }, + { + "epoch": 3.3619457358961906, + "grad_norm": 1.034473180770874, + "learning_rate": 1.305121012653734e-05, + "loss": 0.2699, + "step": 48449 + }, + { + "epoch": 3.362015127333287, + "grad_norm": 1.238432765007019, + "learning_rate": 1.3048434924832764e-05, + "loss": 0.2387, + "step": 48450 + }, + { + "epoch": 3.362084518770384, + "grad_norm": 1.127817153930664, + "learning_rate": 1.3045659997624193e-05, + "loss": 0.366, + "step": 48451 + }, + { + "epoch": 3.3621539102074802, + "grad_norm": 0.6631914973258972, + "learning_rate": 1.3042885344920407e-05, + "loss": 0.2199, + "step": 48452 + }, + { + "epoch": 3.362223301644577, + "grad_norm": 1.025309681892395, + "learning_rate": 1.3040110966730124e-05, + "loss": 0.238, + "step": 48453 + }, + { + "epoch": 3.362292693081674, + "grad_norm": 0.7910332679748535, + "learning_rate": 1.3037336863062122e-05, + "loss": 0.2635, + "step": 48454 + }, + { + "epoch": 3.3623620845187703, + "grad_norm": 0.8914155960083008, + "learning_rate": 1.3034563033925163e-05, + "loss": 0.2753, + "step": 48455 + }, + { + "epoch": 3.362431475955867, + "grad_norm": 0.9142614603042603, + "learning_rate": 1.3031789479327982e-05, + "loss": 0.4011, + "step": 48456 + }, + { + "epoch": 3.3625008673929635, + "grad_norm": 0.8611311316490173, + "learning_rate": 1.3029016199279365e-05, + "loss": 0.2305, + "step": 48457 + }, + { + "epoch": 3.3625702588300603, + "grad_norm": 1.3758940696716309, + "learning_rate": 1.3026243193788069e-05, + "loss": 0.3981, + "step": 48458 + }, + { + "epoch": 3.362639650267157, + "grad_norm": 1.0670198202133179, + "learning_rate": 1.3023470462862819e-05, + "loss": 0.3984, + "step": 48459 + }, + { + "epoch": 3.3627090417042536, + "grad_norm": 0.6268641352653503, + "learning_rate": 1.3020698006512365e-05, + "loss": 0.1642, + "step": 48460 + }, + { + "epoch": 3.3627784331413504, + "grad_norm": 0.9556045532226562, + "learning_rate": 1.3017925824745492e-05, + "loss": 0.254, + "step": 48461 + }, + { + "epoch": 3.3628478245784468, + "grad_norm": 0.844856321811676, + "learning_rate": 1.3015153917570932e-05, + "loss": 0.3379, + "step": 48462 + }, + { + "epoch": 3.3629172160155436, + "grad_norm": 0.6892920732498169, + "learning_rate": 1.3012382284997437e-05, + "loss": 0.2127, + "step": 48463 + }, + { + "epoch": 3.3629866074526404, + "grad_norm": 1.1408368349075317, + "learning_rate": 1.3009610927033755e-05, + "loss": 0.2156, + "step": 48464 + }, + { + "epoch": 3.363055998889737, + "grad_norm": 0.840001106262207, + "learning_rate": 1.3006839843688667e-05, + "loss": 0.2954, + "step": 48465 + }, + { + "epoch": 3.3631253903268337, + "grad_norm": 1.1074939966201782, + "learning_rate": 1.3004069034970866e-05, + "loss": 0.2805, + "step": 48466 + }, + { + "epoch": 3.3631947817639305, + "grad_norm": 1.2187845706939697, + "learning_rate": 1.3001298500889125e-05, + "loss": 0.2156, + "step": 48467 + }, + { + "epoch": 3.363264173201027, + "grad_norm": 0.8863686323165894, + "learning_rate": 1.299852824145219e-05, + "loss": 0.2761, + "step": 48468 + }, + { + "epoch": 3.3633335646381237, + "grad_norm": 0.9391911625862122, + "learning_rate": 1.299575825666881e-05, + "loss": 0.2573, + "step": 48469 + }, + { + "epoch": 3.3634029560752206, + "grad_norm": 0.8573692440986633, + "learning_rate": 1.2992988546547734e-05, + "loss": 0.2279, + "step": 48470 + }, + { + "epoch": 3.363472347512317, + "grad_norm": 0.6914741396903992, + "learning_rate": 1.299021911109769e-05, + "loss": 0.2069, + "step": 48471 + }, + { + "epoch": 3.3635417389494138, + "grad_norm": 0.7293953895568848, + "learning_rate": 1.2987449950327446e-05, + "loss": 0.1809, + "step": 48472 + }, + { + "epoch": 3.36361113038651, + "grad_norm": 2.3644018173217773, + "learning_rate": 1.2984681064245719e-05, + "loss": 0.2578, + "step": 48473 + }, + { + "epoch": 3.363680521823607, + "grad_norm": 0.9607266783714294, + "learning_rate": 1.2981912452861245e-05, + "loss": 0.2315, + "step": 48474 + }, + { + "epoch": 3.363749913260704, + "grad_norm": 0.925621509552002, + "learning_rate": 1.2979144116182784e-05, + "loss": 0.346, + "step": 48475 + }, + { + "epoch": 3.3638193046978, + "grad_norm": 1.121075987815857, + "learning_rate": 1.2976376054219075e-05, + "loss": 0.2482, + "step": 48476 + }, + { + "epoch": 3.363888696134897, + "grad_norm": 0.9732176661491394, + "learning_rate": 1.2973608266978843e-05, + "loss": 0.2515, + "step": 48477 + }, + { + "epoch": 3.3639580875719934, + "grad_norm": 0.8685973882675171, + "learning_rate": 1.2970840754470826e-05, + "loss": 0.3293, + "step": 48478 + }, + { + "epoch": 3.3640274790090903, + "grad_norm": 0.9547193050384521, + "learning_rate": 1.2968073516703804e-05, + "loss": 0.1619, + "step": 48479 + }, + { + "epoch": 3.364096870446187, + "grad_norm": 0.7716862559318542, + "learning_rate": 1.2965306553686441e-05, + "loss": 0.1958, + "step": 48480 + }, + { + "epoch": 3.3641662618832835, + "grad_norm": 0.9229130744934082, + "learning_rate": 1.2962539865427515e-05, + "loss": 0.2838, + "step": 48481 + }, + { + "epoch": 3.3642356533203803, + "grad_norm": 0.8837274312973022, + "learning_rate": 1.2959773451935752e-05, + "loss": 0.2813, + "step": 48482 + }, + { + "epoch": 3.3643050447574767, + "grad_norm": 0.9428698420524597, + "learning_rate": 1.2957007313219905e-05, + "loss": 0.2071, + "step": 48483 + }, + { + "epoch": 3.3643744361945735, + "grad_norm": 0.7594399452209473, + "learning_rate": 1.2954241449288662e-05, + "loss": 0.1878, + "step": 48484 + }, + { + "epoch": 3.3644438276316704, + "grad_norm": 0.8723773956298828, + "learning_rate": 1.2951475860150785e-05, + "loss": 0.3135, + "step": 48485 + }, + { + "epoch": 3.3645132190687668, + "grad_norm": 1.1748902797698975, + "learning_rate": 1.2948710545815001e-05, + "loss": 0.3496, + "step": 48486 + }, + { + "epoch": 3.3645826105058636, + "grad_norm": 1.035382866859436, + "learning_rate": 1.2945945506290059e-05, + "loss": 0.3324, + "step": 48487 + }, + { + "epoch": 3.36465200194296, + "grad_norm": 0.9177533984184265, + "learning_rate": 1.2943180741584649e-05, + "loss": 0.2608, + "step": 48488 + }, + { + "epoch": 3.364721393380057, + "grad_norm": 1.050924301147461, + "learning_rate": 1.294041625170751e-05, + "loss": 0.2862, + "step": 48489 + }, + { + "epoch": 3.3647907848171537, + "grad_norm": 0.9942970871925354, + "learning_rate": 1.2937652036667369e-05, + "loss": 0.2868, + "step": 48490 + }, + { + "epoch": 3.36486017625425, + "grad_norm": 0.8161142468452454, + "learning_rate": 1.2934888096472974e-05, + "loss": 0.1911, + "step": 48491 + }, + { + "epoch": 3.364929567691347, + "grad_norm": 1.2291594743728638, + "learning_rate": 1.2932124431133053e-05, + "loss": 0.2916, + "step": 48492 + }, + { + "epoch": 3.3649989591284437, + "grad_norm": 0.711368203163147, + "learning_rate": 1.2929361040656284e-05, + "loss": 0.285, + "step": 48493 + }, + { + "epoch": 3.36506835056554, + "grad_norm": 0.918310284614563, + "learning_rate": 1.2926597925051432e-05, + "loss": 0.2059, + "step": 48494 + }, + { + "epoch": 3.365137742002637, + "grad_norm": 1.1298472881317139, + "learning_rate": 1.2923835084327196e-05, + "loss": 0.3328, + "step": 48495 + }, + { + "epoch": 3.3652071334397338, + "grad_norm": 0.8071291446685791, + "learning_rate": 1.2921072518492317e-05, + "loss": 0.2457, + "step": 48496 + }, + { + "epoch": 3.36527652487683, + "grad_norm": 0.8713196516036987, + "learning_rate": 1.2918310227555508e-05, + "loss": 0.2305, + "step": 48497 + }, + { + "epoch": 3.365345916313927, + "grad_norm": 0.9012911915779114, + "learning_rate": 1.2915548211525497e-05, + "loss": 0.279, + "step": 48498 + }, + { + "epoch": 3.3654153077510234, + "grad_norm": 0.7497433423995972, + "learning_rate": 1.2912786470411009e-05, + "loss": 0.194, + "step": 48499 + }, + { + "epoch": 3.36548469918812, + "grad_norm": 0.9150092601776123, + "learning_rate": 1.2910025004220726e-05, + "loss": 0.1909, + "step": 48500 + }, + { + "epoch": 3.365554090625217, + "grad_norm": 1.068626880645752, + "learning_rate": 1.29072638129634e-05, + "loss": 0.2202, + "step": 48501 + }, + { + "epoch": 3.3656234820623134, + "grad_norm": 0.912871241569519, + "learning_rate": 1.290450289664773e-05, + "loss": 0.3399, + "step": 48502 + }, + { + "epoch": 3.3656928734994103, + "grad_norm": 0.8026244640350342, + "learning_rate": 1.2901742255282446e-05, + "loss": 0.1693, + "step": 48503 + }, + { + "epoch": 3.3657622649365067, + "grad_norm": 1.1524957418441772, + "learning_rate": 1.2898981888876249e-05, + "loss": 0.2846, + "step": 48504 + }, + { + "epoch": 3.3658316563736035, + "grad_norm": 1.2706276178359985, + "learning_rate": 1.2896221797437858e-05, + "loss": 0.2917, + "step": 48505 + }, + { + "epoch": 3.3659010478107003, + "grad_norm": 0.7597141265869141, + "learning_rate": 1.2893461980976018e-05, + "loss": 0.1842, + "step": 48506 + }, + { + "epoch": 3.3659704392477967, + "grad_norm": 1.0789505243301392, + "learning_rate": 1.2890702439499391e-05, + "loss": 0.2213, + "step": 48507 + }, + { + "epoch": 3.3660398306848935, + "grad_norm": 0.8308202028274536, + "learning_rate": 1.2887943173016703e-05, + "loss": 0.2219, + "step": 48508 + }, + { + "epoch": 3.36610922212199, + "grad_norm": 1.232614517211914, + "learning_rate": 1.288518418153668e-05, + "loss": 0.3609, + "step": 48509 + }, + { + "epoch": 3.3661786135590868, + "grad_norm": 0.7550560832023621, + "learning_rate": 1.2882425465068026e-05, + "loss": 0.1986, + "step": 48510 + }, + { + "epoch": 3.3662480049961836, + "grad_norm": 0.5946915745735168, + "learning_rate": 1.2879667023619435e-05, + "loss": 0.1649, + "step": 48511 + }, + { + "epoch": 3.36631739643328, + "grad_norm": 0.6664948463439941, + "learning_rate": 1.2876908857199632e-05, + "loss": 0.2161, + "step": 48512 + }, + { + "epoch": 3.366386787870377, + "grad_norm": 0.9604995250701904, + "learning_rate": 1.2874150965817344e-05, + "loss": 0.2953, + "step": 48513 + }, + { + "epoch": 3.3664561793074737, + "grad_norm": 1.0809227228164673, + "learning_rate": 1.2871393349481231e-05, + "loss": 0.1889, + "step": 48514 + }, + { + "epoch": 3.36652557074457, + "grad_norm": 1.0150110721588135, + "learning_rate": 1.2868636008200018e-05, + "loss": 0.4224, + "step": 48515 + }, + { + "epoch": 3.366594962181667, + "grad_norm": 1.1766303777694702, + "learning_rate": 1.2865878941982411e-05, + "loss": 0.3016, + "step": 48516 + }, + { + "epoch": 3.3666643536187633, + "grad_norm": 1.3334230184555054, + "learning_rate": 1.2863122150837126e-05, + "loss": 0.2885, + "step": 48517 + }, + { + "epoch": 3.36673374505586, + "grad_norm": 1.0177180767059326, + "learning_rate": 1.2860365634772864e-05, + "loss": 0.3447, + "step": 48518 + }, + { + "epoch": 3.366803136492957, + "grad_norm": 1.5188796520233154, + "learning_rate": 1.2857609393798297e-05, + "loss": 0.2439, + "step": 48519 + }, + { + "epoch": 3.3668725279300533, + "grad_norm": 0.6217052340507507, + "learning_rate": 1.2854853427922154e-05, + "loss": 0.2226, + "step": 48520 + }, + { + "epoch": 3.36694191936715, + "grad_norm": 1.2479826211929321, + "learning_rate": 1.2852097737153145e-05, + "loss": 0.3125, + "step": 48521 + }, + { + "epoch": 3.367011310804247, + "grad_norm": 0.9012230038642883, + "learning_rate": 1.2849342321499924e-05, + "loss": 0.2021, + "step": 48522 + }, + { + "epoch": 3.3670807022413434, + "grad_norm": 1.1200352907180786, + "learning_rate": 1.2846587180971226e-05, + "loss": 0.353, + "step": 48523 + }, + { + "epoch": 3.36715009367844, + "grad_norm": 0.5444571375846863, + "learning_rate": 1.2843832315575743e-05, + "loss": 0.1606, + "step": 48524 + }, + { + "epoch": 3.3672194851155366, + "grad_norm": 0.7952179908752441, + "learning_rate": 1.284107772532216e-05, + "loss": 0.3493, + "step": 48525 + }, + { + "epoch": 3.3672888765526334, + "grad_norm": 0.8618508577346802, + "learning_rate": 1.2838323410219188e-05, + "loss": 0.2195, + "step": 48526 + }, + { + "epoch": 3.3673582679897303, + "grad_norm": 0.8980869054794312, + "learning_rate": 1.2835569370275535e-05, + "loss": 0.1968, + "step": 48527 + }, + { + "epoch": 3.3674276594268266, + "grad_norm": 0.9247922897338867, + "learning_rate": 1.2832815605499859e-05, + "loss": 0.2768, + "step": 48528 + }, + { + "epoch": 3.3674970508639235, + "grad_norm": 1.1076359748840332, + "learning_rate": 1.2830062115900876e-05, + "loss": 0.3342, + "step": 48529 + }, + { + "epoch": 3.36756644230102, + "grad_norm": 1.1905909776687622, + "learning_rate": 1.2827308901487268e-05, + "loss": 0.3076, + "step": 48530 + }, + { + "epoch": 3.3676358337381167, + "grad_norm": 0.8730660676956177, + "learning_rate": 1.2824555962267725e-05, + "loss": 0.2306, + "step": 48531 + }, + { + "epoch": 3.3677052251752135, + "grad_norm": 0.8894117474555969, + "learning_rate": 1.2821803298250956e-05, + "loss": 0.2018, + "step": 48532 + }, + { + "epoch": 3.36777461661231, + "grad_norm": 0.7776122093200684, + "learning_rate": 1.2819050909445629e-05, + "loss": 0.2147, + "step": 48533 + }, + { + "epoch": 3.3678440080494068, + "grad_norm": 0.8709455728530884, + "learning_rate": 1.281629879586047e-05, + "loss": 0.2403, + "step": 48534 + }, + { + "epoch": 3.367913399486503, + "grad_norm": 0.7980528473854065, + "learning_rate": 1.2813546957504119e-05, + "loss": 0.1865, + "step": 48535 + }, + { + "epoch": 3.3679827909236, + "grad_norm": 0.9248086214065552, + "learning_rate": 1.2810795394385288e-05, + "loss": 0.2161, + "step": 48536 + }, + { + "epoch": 3.368052182360697, + "grad_norm": 1.0666165351867676, + "learning_rate": 1.280804410651265e-05, + "loss": 0.3621, + "step": 48537 + }, + { + "epoch": 3.368121573797793, + "grad_norm": 0.810619592666626, + "learning_rate": 1.280529309389491e-05, + "loss": 0.2841, + "step": 48538 + }, + { + "epoch": 3.36819096523489, + "grad_norm": 0.6979926824569702, + "learning_rate": 1.2802542356540736e-05, + "loss": 0.1785, + "step": 48539 + }, + { + "epoch": 3.368260356671987, + "grad_norm": 1.6723499298095703, + "learning_rate": 1.2799791894458823e-05, + "loss": 0.4456, + "step": 48540 + }, + { + "epoch": 3.3683297481090833, + "grad_norm": 0.9664281606674194, + "learning_rate": 1.2797041707657875e-05, + "loss": 0.2167, + "step": 48541 + }, + { + "epoch": 3.36839913954618, + "grad_norm": 1.1157690286636353, + "learning_rate": 1.2794291796146529e-05, + "loss": 0.2624, + "step": 48542 + }, + { + "epoch": 3.3684685309832765, + "grad_norm": 0.6751578450202942, + "learning_rate": 1.2791542159933479e-05, + "loss": 0.198, + "step": 48543 + }, + { + "epoch": 3.3685379224203733, + "grad_norm": 0.7308859825134277, + "learning_rate": 1.2788792799027416e-05, + "loss": 0.2721, + "step": 48544 + }, + { + "epoch": 3.36860731385747, + "grad_norm": 0.5485804080963135, + "learning_rate": 1.2786043713437012e-05, + "loss": 0.2033, + "step": 48545 + }, + { + "epoch": 3.3686767052945665, + "grad_norm": 0.9667307734489441, + "learning_rate": 1.278329490317095e-05, + "loss": 0.241, + "step": 48546 + }, + { + "epoch": 3.3687460967316634, + "grad_norm": 0.9746043682098389, + "learning_rate": 1.278054636823791e-05, + "loss": 0.276, + "step": 48547 + }, + { + "epoch": 3.36881548816876, + "grad_norm": 1.0402095317840576, + "learning_rate": 1.2777798108646588e-05, + "loss": 0.2392, + "step": 48548 + }, + { + "epoch": 3.3688848796058566, + "grad_norm": 1.1199506521224976, + "learning_rate": 1.2775050124405618e-05, + "loss": 0.2073, + "step": 48549 + }, + { + "epoch": 3.3689542710429534, + "grad_norm": 0.9743820428848267, + "learning_rate": 1.2772302415523685e-05, + "loss": 0.2286, + "step": 48550 + }, + { + "epoch": 3.36902366248005, + "grad_norm": 1.0253264904022217, + "learning_rate": 1.276955498200948e-05, + "loss": 0.2545, + "step": 48551 + }, + { + "epoch": 3.3690930539171466, + "grad_norm": 1.0598044395446777, + "learning_rate": 1.2766807823871674e-05, + "loss": 0.2206, + "step": 48552 + }, + { + "epoch": 3.3691624453542435, + "grad_norm": 0.693315863609314, + "learning_rate": 1.2764060941118949e-05, + "loss": 0.1832, + "step": 48553 + }, + { + "epoch": 3.36923183679134, + "grad_norm": 0.79093998670578, + "learning_rate": 1.2761314333759944e-05, + "loss": 0.2899, + "step": 48554 + }, + { + "epoch": 3.3693012282284367, + "grad_norm": 1.2551437616348267, + "learning_rate": 1.275856800180334e-05, + "loss": 0.3061, + "step": 48555 + }, + { + "epoch": 3.369370619665533, + "grad_norm": 0.7776361703872681, + "learning_rate": 1.275582194525785e-05, + "loss": 0.2049, + "step": 48556 + }, + { + "epoch": 3.36944001110263, + "grad_norm": 0.9496458768844604, + "learning_rate": 1.2753076164132072e-05, + "loss": 0.2716, + "step": 48557 + }, + { + "epoch": 3.3695094025397268, + "grad_norm": 1.0161465406417847, + "learning_rate": 1.2750330658434717e-05, + "loss": 0.2686, + "step": 48558 + }, + { + "epoch": 3.369578793976823, + "grad_norm": 2.041738748550415, + "learning_rate": 1.2747585428174447e-05, + "loss": 0.4537, + "step": 48559 + }, + { + "epoch": 3.36964818541392, + "grad_norm": 1.2849212884902954, + "learning_rate": 1.274484047335993e-05, + "loss": 0.1943, + "step": 48560 + }, + { + "epoch": 3.3697175768510164, + "grad_norm": 1.2117362022399902, + "learning_rate": 1.274209579399982e-05, + "loss": 0.2954, + "step": 48561 + }, + { + "epoch": 3.369786968288113, + "grad_norm": 1.3721834421157837, + "learning_rate": 1.2739351390102816e-05, + "loss": 0.4049, + "step": 48562 + }, + { + "epoch": 3.36985635972521, + "grad_norm": 0.9771113991737366, + "learning_rate": 1.2736607261677536e-05, + "loss": 0.2013, + "step": 48563 + }, + { + "epoch": 3.3699257511623064, + "grad_norm": 0.9843951463699341, + "learning_rate": 1.2733863408732661e-05, + "loss": 0.319, + "step": 48564 + }, + { + "epoch": 3.3699951425994032, + "grad_norm": 0.8185514807701111, + "learning_rate": 1.273111983127685e-05, + "loss": 0.1866, + "step": 48565 + }, + { + "epoch": 3.3700645340365, + "grad_norm": 0.8714323043823242, + "learning_rate": 1.2728376529318776e-05, + "loss": 0.3771, + "step": 48566 + }, + { + "epoch": 3.3701339254735965, + "grad_norm": 0.9464332461357117, + "learning_rate": 1.2725633502867085e-05, + "loss": 0.2487, + "step": 48567 + }, + { + "epoch": 3.3702033169106933, + "grad_norm": 0.9107726216316223, + "learning_rate": 1.272289075193046e-05, + "loss": 0.2331, + "step": 48568 + }, + { + "epoch": 3.3702727083477897, + "grad_norm": 0.7210456132888794, + "learning_rate": 1.272014827651753e-05, + "loss": 0.166, + "step": 48569 + }, + { + "epoch": 3.3703420997848865, + "grad_norm": 0.9148133993148804, + "learning_rate": 1.2717406076636962e-05, + "loss": 0.2339, + "step": 48570 + }, + { + "epoch": 3.3704114912219834, + "grad_norm": 2.324634552001953, + "learning_rate": 1.2714664152297417e-05, + "loss": 0.2567, + "step": 48571 + }, + { + "epoch": 3.3704808826590797, + "grad_norm": 1.2246043682098389, + "learning_rate": 1.2711922503507556e-05, + "loss": 0.3065, + "step": 48572 + }, + { + "epoch": 3.3705502740961766, + "grad_norm": 1.0160549879074097, + "learning_rate": 1.2709181130276027e-05, + "loss": 0.3588, + "step": 48573 + }, + { + "epoch": 3.3706196655332734, + "grad_norm": 0.8997920751571655, + "learning_rate": 1.2706440032611477e-05, + "loss": 0.3084, + "step": 48574 + }, + { + "epoch": 3.37068905697037, + "grad_norm": 0.9063040614128113, + "learning_rate": 1.2703699210522601e-05, + "loss": 0.1934, + "step": 48575 + }, + { + "epoch": 3.3707584484074666, + "grad_norm": 1.546560287475586, + "learning_rate": 1.2700958664017992e-05, + "loss": 0.2317, + "step": 48576 + }, + { + "epoch": 3.370827839844563, + "grad_norm": 1.2010314464569092, + "learning_rate": 1.269821839310632e-05, + "loss": 0.2733, + "step": 48577 + }, + { + "epoch": 3.37089723128166, + "grad_norm": 0.9667601585388184, + "learning_rate": 1.2695478397796256e-05, + "loss": 0.247, + "step": 48578 + }, + { + "epoch": 3.3709666227187567, + "grad_norm": 0.9424307346343994, + "learning_rate": 1.2692738678096427e-05, + "loss": 0.2268, + "step": 48579 + }, + { + "epoch": 3.371036014155853, + "grad_norm": 0.7084434032440186, + "learning_rate": 1.2689999234015504e-05, + "loss": 0.198, + "step": 48580 + }, + { + "epoch": 3.37110540559295, + "grad_norm": 0.9711548686027527, + "learning_rate": 1.2687260065562112e-05, + "loss": 0.4702, + "step": 48581 + }, + { + "epoch": 3.3711747970300463, + "grad_norm": 0.6520823240280151, + "learning_rate": 1.2684521172744934e-05, + "loss": 0.2423, + "step": 48582 + }, + { + "epoch": 3.371244188467143, + "grad_norm": 1.3088175058364868, + "learning_rate": 1.2681782555572575e-05, + "loss": 0.2651, + "step": 48583 + }, + { + "epoch": 3.37131357990424, + "grad_norm": 0.7175734639167786, + "learning_rate": 1.2679044214053703e-05, + "loss": 0.1371, + "step": 48584 + }, + { + "epoch": 3.3713829713413364, + "grad_norm": 0.9271020889282227, + "learning_rate": 1.2676306148196948e-05, + "loss": 0.1832, + "step": 48585 + }, + { + "epoch": 3.371452362778433, + "grad_norm": 1.244049310684204, + "learning_rate": 1.2673568358010968e-05, + "loss": 0.356, + "step": 48586 + }, + { + "epoch": 3.3715217542155296, + "grad_norm": 0.8629252314567566, + "learning_rate": 1.267083084350441e-05, + "loss": 0.331, + "step": 48587 + }, + { + "epoch": 3.3715911456526264, + "grad_norm": 1.0697712898254395, + "learning_rate": 1.2668093604685905e-05, + "loss": 0.3408, + "step": 48588 + }, + { + "epoch": 3.3716605370897232, + "grad_norm": 1.010584831237793, + "learning_rate": 1.2665356641564086e-05, + "loss": 0.2411, + "step": 48589 + }, + { + "epoch": 3.3717299285268196, + "grad_norm": 0.7558618783950806, + "learning_rate": 1.2662619954147604e-05, + "loss": 0.2407, + "step": 48590 + }, + { + "epoch": 3.3717993199639165, + "grad_norm": 0.8855978846549988, + "learning_rate": 1.2659883542445116e-05, + "loss": 0.441, + "step": 48591 + }, + { + "epoch": 3.3718687114010133, + "grad_norm": 0.9348558187484741, + "learning_rate": 1.2657147406465219e-05, + "loss": 0.2198, + "step": 48592 + }, + { + "epoch": 3.3719381028381097, + "grad_norm": 1.3621164560317993, + "learning_rate": 1.2654411546216583e-05, + "loss": 0.2739, + "step": 48593 + }, + { + "epoch": 3.3720074942752065, + "grad_norm": 1.364732265472412, + "learning_rate": 1.2651675961707831e-05, + "loss": 0.2315, + "step": 48594 + }, + { + "epoch": 3.3720768857123034, + "grad_norm": 1.1910103559494019, + "learning_rate": 1.2648940652947605e-05, + "loss": 0.3309, + "step": 48595 + }, + { + "epoch": 3.3721462771493997, + "grad_norm": 0.878713846206665, + "learning_rate": 1.2646205619944562e-05, + "loss": 0.203, + "step": 48596 + }, + { + "epoch": 3.3722156685864966, + "grad_norm": 0.8310960531234741, + "learning_rate": 1.2643470862707285e-05, + "loss": 0.2055, + "step": 48597 + }, + { + "epoch": 3.372285060023593, + "grad_norm": 0.7990710139274597, + "learning_rate": 1.2640736381244433e-05, + "loss": 0.3642, + "step": 48598 + }, + { + "epoch": 3.37235445146069, + "grad_norm": 1.1498734951019287, + "learning_rate": 1.2638002175564657e-05, + "loss": 0.3114, + "step": 48599 + }, + { + "epoch": 3.3724238428977866, + "grad_norm": 1.1645102500915527, + "learning_rate": 1.2635268245676556e-05, + "loss": 0.1982, + "step": 48600 + }, + { + "epoch": 3.372493234334883, + "grad_norm": 1.2019641399383545, + "learning_rate": 1.2632534591588786e-05, + "loss": 0.2411, + "step": 48601 + }, + { + "epoch": 3.37256262577198, + "grad_norm": 1.0237478017807007, + "learning_rate": 1.2629801213309966e-05, + "loss": 0.2728, + "step": 48602 + }, + { + "epoch": 3.3726320172090762, + "grad_norm": 2.244922161102295, + "learning_rate": 1.262706811084875e-05, + "loss": 0.2331, + "step": 48603 + }, + { + "epoch": 3.372701408646173, + "grad_norm": 0.9969447255134583, + "learning_rate": 1.2624335284213717e-05, + "loss": 0.2697, + "step": 48604 + }, + { + "epoch": 3.37277080008327, + "grad_norm": 1.004339337348938, + "learning_rate": 1.2621602733413518e-05, + "loss": 0.3656, + "step": 48605 + }, + { + "epoch": 3.3728401915203663, + "grad_norm": 1.2789437770843506, + "learning_rate": 1.2618870458456788e-05, + "loss": 0.3132, + "step": 48606 + }, + { + "epoch": 3.372909582957463, + "grad_norm": 0.7245708107948303, + "learning_rate": 1.2616138459352134e-05, + "loss": 0.2179, + "step": 48607 + }, + { + "epoch": 3.3729789743945595, + "grad_norm": 0.8558701872825623, + "learning_rate": 1.2613406736108203e-05, + "loss": 0.1565, + "step": 48608 + }, + { + "epoch": 3.3730483658316563, + "grad_norm": 1.3794386386871338, + "learning_rate": 1.2610675288733609e-05, + "loss": 0.1968, + "step": 48609 + }, + { + "epoch": 3.373117757268753, + "grad_norm": 0.9563413262367249, + "learning_rate": 1.2607944117236992e-05, + "loss": 0.2626, + "step": 48610 + }, + { + "epoch": 3.3731871487058496, + "grad_norm": 0.999355137348175, + "learning_rate": 1.2605213221626933e-05, + "loss": 0.2791, + "step": 48611 + }, + { + "epoch": 3.3732565401429464, + "grad_norm": 0.5277669429779053, + "learning_rate": 1.260248260191208e-05, + "loss": 0.1687, + "step": 48612 + }, + { + "epoch": 3.373325931580043, + "grad_norm": 1.4658563137054443, + "learning_rate": 1.259975225810105e-05, + "loss": 0.2224, + "step": 48613 + }, + { + "epoch": 3.3733953230171396, + "grad_norm": 1.3525733947753906, + "learning_rate": 1.2597022190202456e-05, + "loss": 0.3126, + "step": 48614 + }, + { + "epoch": 3.3734647144542365, + "grad_norm": 1.0971298217773438, + "learning_rate": 1.2594292398224927e-05, + "loss": 0.3426, + "step": 48615 + }, + { + "epoch": 3.373534105891333, + "grad_norm": 1.5952421426773071, + "learning_rate": 1.2591562882177077e-05, + "loss": 0.2951, + "step": 48616 + }, + { + "epoch": 3.3736034973284297, + "grad_norm": 0.9639642238616943, + "learning_rate": 1.2588833642067543e-05, + "loss": 0.2629, + "step": 48617 + }, + { + "epoch": 3.3736728887655265, + "grad_norm": 1.0490427017211914, + "learning_rate": 1.2586104677904897e-05, + "loss": 0.305, + "step": 48618 + }, + { + "epoch": 3.373742280202623, + "grad_norm": 0.8803061842918396, + "learning_rate": 1.2583375989697777e-05, + "loss": 0.2418, + "step": 48619 + }, + { + "epoch": 3.3738116716397197, + "grad_norm": 1.1109261512756348, + "learning_rate": 1.2580647577454796e-05, + "loss": 0.2126, + "step": 48620 + }, + { + "epoch": 3.3738810630768166, + "grad_norm": 0.8557742238044739, + "learning_rate": 1.2577919441184572e-05, + "loss": 0.2581, + "step": 48621 + }, + { + "epoch": 3.373950454513913, + "grad_norm": 0.9641779661178589, + "learning_rate": 1.2575191580895729e-05, + "loss": 0.2017, + "step": 48622 + }, + { + "epoch": 3.37401984595101, + "grad_norm": 0.9662386775016785, + "learning_rate": 1.2572463996596839e-05, + "loss": 0.3878, + "step": 48623 + }, + { + "epoch": 3.374089237388106, + "grad_norm": 0.8429635167121887, + "learning_rate": 1.2569736688296541e-05, + "loss": 0.1839, + "step": 48624 + }, + { + "epoch": 3.374158628825203, + "grad_norm": 0.8268209099769592, + "learning_rate": 1.2567009656003448e-05, + "loss": 0.2344, + "step": 48625 + }, + { + "epoch": 3.3742280202623, + "grad_norm": 0.9710233211517334, + "learning_rate": 1.2564282899726176e-05, + "loss": 0.1417, + "step": 48626 + }, + { + "epoch": 3.3742974116993962, + "grad_norm": 1.6941893100738525, + "learning_rate": 1.2561556419473297e-05, + "loss": 0.3812, + "step": 48627 + }, + { + "epoch": 3.374366803136493, + "grad_norm": 0.8229745626449585, + "learning_rate": 1.2558830215253436e-05, + "loss": 0.1793, + "step": 48628 + }, + { + "epoch": 3.3744361945735895, + "grad_norm": 0.7916615605354309, + "learning_rate": 1.255610428707521e-05, + "loss": 0.1291, + "step": 48629 + }, + { + "epoch": 3.3745055860106863, + "grad_norm": 0.688351035118103, + "learning_rate": 1.2553378634947211e-05, + "loss": 0.1388, + "step": 48630 + }, + { + "epoch": 3.374574977447783, + "grad_norm": 0.9563993811607361, + "learning_rate": 1.2550653258878076e-05, + "loss": 0.296, + "step": 48631 + }, + { + "epoch": 3.3746443688848795, + "grad_norm": 1.054551601409912, + "learning_rate": 1.2547928158876354e-05, + "loss": 0.3613, + "step": 48632 + }, + { + "epoch": 3.3747137603219763, + "grad_norm": 0.7954046726226807, + "learning_rate": 1.2545203334950683e-05, + "loss": 0.181, + "step": 48633 + }, + { + "epoch": 3.3747831517590727, + "grad_norm": 1.2059292793273926, + "learning_rate": 1.2542478787109658e-05, + "loss": 0.3074, + "step": 48634 + }, + { + "epoch": 3.3748525431961696, + "grad_norm": 0.6732224822044373, + "learning_rate": 1.253975451536188e-05, + "loss": 0.2062, + "step": 48635 + }, + { + "epoch": 3.3749219346332664, + "grad_norm": 0.6923085451126099, + "learning_rate": 1.2537030519715953e-05, + "loss": 0.1758, + "step": 48636 + }, + { + "epoch": 3.374991326070363, + "grad_norm": 0.8138566017150879, + "learning_rate": 1.2534306800180473e-05, + "loss": 0.2272, + "step": 48637 + }, + { + "epoch": 3.3750607175074596, + "grad_norm": 0.7815215587615967, + "learning_rate": 1.2531583356764054e-05, + "loss": 0.1569, + "step": 48638 + }, + { + "epoch": 3.375130108944556, + "grad_norm": 0.9113404154777527, + "learning_rate": 1.2528860189475255e-05, + "loss": 0.2521, + "step": 48639 + }, + { + "epoch": 3.375199500381653, + "grad_norm": 1.4097825288772583, + "learning_rate": 1.2526137298322705e-05, + "loss": 0.2894, + "step": 48640 + }, + { + "epoch": 3.3752688918187497, + "grad_norm": 0.6375142335891724, + "learning_rate": 1.2523414683314983e-05, + "loss": 0.2053, + "step": 48641 + }, + { + "epoch": 3.375338283255846, + "grad_norm": 0.8558904528617859, + "learning_rate": 1.2520692344460693e-05, + "loss": 0.234, + "step": 48642 + }, + { + "epoch": 3.375407674692943, + "grad_norm": 0.7352285981178284, + "learning_rate": 1.2517970281768431e-05, + "loss": 0.2343, + "step": 48643 + }, + { + "epoch": 3.3754770661300397, + "grad_norm": 0.9320256114006042, + "learning_rate": 1.251524849524679e-05, + "loss": 0.2876, + "step": 48644 + }, + { + "epoch": 3.375546457567136, + "grad_norm": 0.8044186234474182, + "learning_rate": 1.2512526984904372e-05, + "loss": 0.2063, + "step": 48645 + }, + { + "epoch": 3.375615849004233, + "grad_norm": 0.8034155964851379, + "learning_rate": 1.2509805750749737e-05, + "loss": 0.2132, + "step": 48646 + }, + { + "epoch": 3.37568524044133, + "grad_norm": 0.6791016459465027, + "learning_rate": 1.25070847927915e-05, + "loss": 0.1482, + "step": 48647 + }, + { + "epoch": 3.375754631878426, + "grad_norm": 0.9152119159698486, + "learning_rate": 1.2504364111038235e-05, + "loss": 0.2865, + "step": 48648 + }, + { + "epoch": 3.375824023315523, + "grad_norm": 1.0302897691726685, + "learning_rate": 1.2501643705498556e-05, + "loss": 0.3766, + "step": 48649 + }, + { + "epoch": 3.3758934147526194, + "grad_norm": 1.1954385042190552, + "learning_rate": 1.2498923576181021e-05, + "loss": 0.359, + "step": 48650 + }, + { + "epoch": 3.3759628061897162, + "grad_norm": 1.0930472612380981, + "learning_rate": 1.2496203723094257e-05, + "loss": 0.3269, + "step": 48651 + }, + { + "epoch": 3.376032197626813, + "grad_norm": 0.8694193959236145, + "learning_rate": 1.2493484146246814e-05, + "loss": 0.3042, + "step": 48652 + }, + { + "epoch": 3.3761015890639094, + "grad_norm": 1.0928289890289307, + "learning_rate": 1.2490764845647285e-05, + "loss": 0.3027, + "step": 48653 + }, + { + "epoch": 3.3761709805010063, + "grad_norm": 0.7442107200622559, + "learning_rate": 1.248804582130425e-05, + "loss": 0.1718, + "step": 48654 + }, + { + "epoch": 3.3762403719381027, + "grad_norm": 1.490393042564392, + "learning_rate": 1.2485327073226304e-05, + "loss": 0.338, + "step": 48655 + }, + { + "epoch": 3.3763097633751995, + "grad_norm": 0.7191770076751709, + "learning_rate": 1.248260860142203e-05, + "loss": 0.1886, + "step": 48656 + }, + { + "epoch": 3.3763791548122963, + "grad_norm": 0.9997097253799438, + "learning_rate": 1.247989040590003e-05, + "loss": 0.218, + "step": 48657 + }, + { + "epoch": 3.3764485462493927, + "grad_norm": 1.1377625465393066, + "learning_rate": 1.2477172486668843e-05, + "loss": 0.1816, + "step": 48658 + }, + { + "epoch": 3.3765179376864896, + "grad_norm": 1.6496673822402954, + "learning_rate": 1.247445484373706e-05, + "loss": 0.2245, + "step": 48659 + }, + { + "epoch": 3.376587329123586, + "grad_norm": 0.8930208086967468, + "learning_rate": 1.2471737477113277e-05, + "loss": 0.2216, + "step": 48660 + }, + { + "epoch": 3.376656720560683, + "grad_norm": 1.3250188827514648, + "learning_rate": 1.2469020386806074e-05, + "loss": 0.1289, + "step": 48661 + }, + { + "epoch": 3.3767261119977796, + "grad_norm": 0.8886973857879639, + "learning_rate": 1.2466303572824e-05, + "loss": 0.2423, + "step": 48662 + }, + { + "epoch": 3.376795503434876, + "grad_norm": 0.8501278758049011, + "learning_rate": 1.246358703517565e-05, + "loss": 0.2104, + "step": 48663 + }, + { + "epoch": 3.376864894871973, + "grad_norm": 0.9079132676124573, + "learning_rate": 1.2460870773869603e-05, + "loss": 0.2201, + "step": 48664 + }, + { + "epoch": 3.376934286309069, + "grad_norm": 0.6761801838874817, + "learning_rate": 1.2458154788914455e-05, + "loss": 0.2469, + "step": 48665 + }, + { + "epoch": 3.377003677746166, + "grad_norm": 0.9143792390823364, + "learning_rate": 1.2455439080318731e-05, + "loss": 0.318, + "step": 48666 + }, + { + "epoch": 3.377073069183263, + "grad_norm": 0.8705251812934875, + "learning_rate": 1.2452723648091025e-05, + "loss": 0.2806, + "step": 48667 + }, + { + "epoch": 3.3771424606203593, + "grad_norm": 0.642676055431366, + "learning_rate": 1.2450008492239917e-05, + "loss": 0.2099, + "step": 48668 + }, + { + "epoch": 3.377211852057456, + "grad_norm": 2.416384220123291, + "learning_rate": 1.2447293612773969e-05, + "loss": 0.3511, + "step": 48669 + }, + { + "epoch": 3.377281243494553, + "grad_norm": 0.6324672102928162, + "learning_rate": 1.2444579009701762e-05, + "loss": 0.1515, + "step": 48670 + }, + { + "epoch": 3.3773506349316493, + "grad_norm": 0.8678277134895325, + "learning_rate": 1.2441864683031856e-05, + "loss": 0.2733, + "step": 48671 + }, + { + "epoch": 3.377420026368746, + "grad_norm": 0.8927626013755798, + "learning_rate": 1.2439150632772856e-05, + "loss": 0.2412, + "step": 48672 + }, + { + "epoch": 3.377489417805843, + "grad_norm": 0.6628906726837158, + "learning_rate": 1.2436436858933276e-05, + "loss": 0.2234, + "step": 48673 + }, + { + "epoch": 3.3775588092429394, + "grad_norm": 0.7521681189537048, + "learning_rate": 1.2433723361521699e-05, + "loss": 0.2256, + "step": 48674 + }, + { + "epoch": 3.377628200680036, + "grad_norm": 1.179674744606018, + "learning_rate": 1.2431010140546695e-05, + "loss": 0.28, + "step": 48675 + }, + { + "epoch": 3.3776975921171326, + "grad_norm": 0.7886717319488525, + "learning_rate": 1.2428297196016835e-05, + "loss": 0.1626, + "step": 48676 + }, + { + "epoch": 3.3777669835542294, + "grad_norm": 1.5188847780227661, + "learning_rate": 1.242558452794068e-05, + "loss": 0.2834, + "step": 48677 + }, + { + "epoch": 3.3778363749913263, + "grad_norm": 1.0531022548675537, + "learning_rate": 1.2422872136326801e-05, + "loss": 0.2403, + "step": 48678 + }, + { + "epoch": 3.3779057664284227, + "grad_norm": 0.9499773383140564, + "learning_rate": 1.242016002118377e-05, + "loss": 0.2165, + "step": 48679 + }, + { + "epoch": 3.3779751578655195, + "grad_norm": 1.6893575191497803, + "learning_rate": 1.2417448182520108e-05, + "loss": 0.4798, + "step": 48680 + }, + { + "epoch": 3.378044549302616, + "grad_norm": 0.7808372378349304, + "learning_rate": 1.2414736620344413e-05, + "loss": 0.1931, + "step": 48681 + }, + { + "epoch": 3.3781139407397127, + "grad_norm": 1.0013433694839478, + "learning_rate": 1.2412025334665222e-05, + "loss": 0.3075, + "step": 48682 + }, + { + "epoch": 3.3781833321768096, + "grad_norm": 0.8679642677307129, + "learning_rate": 1.2409314325491106e-05, + "loss": 0.2866, + "step": 48683 + }, + { + "epoch": 3.378252723613906, + "grad_norm": 0.8134424686431885, + "learning_rate": 1.2406603592830624e-05, + "loss": 0.2229, + "step": 48684 + }, + { + "epoch": 3.3783221150510028, + "grad_norm": 1.2174863815307617, + "learning_rate": 1.2403893136692325e-05, + "loss": 0.3175, + "step": 48685 + }, + { + "epoch": 3.378391506488099, + "grad_norm": 1.1430193185806274, + "learning_rate": 1.24011829570848e-05, + "loss": 0.3032, + "step": 48686 + }, + { + "epoch": 3.378460897925196, + "grad_norm": 0.9565688371658325, + "learning_rate": 1.2398473054016558e-05, + "loss": 0.2883, + "step": 48687 + }, + { + "epoch": 3.378530289362293, + "grad_norm": 0.9064702391624451, + "learning_rate": 1.2395763427496176e-05, + "loss": 0.266, + "step": 48688 + }, + { + "epoch": 3.378599680799389, + "grad_norm": 1.0736169815063477, + "learning_rate": 1.2393054077532196e-05, + "loss": 0.2497, + "step": 48689 + }, + { + "epoch": 3.378669072236486, + "grad_norm": 1.0497182607650757, + "learning_rate": 1.2390345004133186e-05, + "loss": 0.2388, + "step": 48690 + }, + { + "epoch": 3.3787384636735824, + "grad_norm": 0.6422812342643738, + "learning_rate": 1.2387636207307685e-05, + "loss": 0.1806, + "step": 48691 + }, + { + "epoch": 3.3788078551106793, + "grad_norm": 0.8959338665008545, + "learning_rate": 1.2384927687064275e-05, + "loss": 0.1939, + "step": 48692 + }, + { + "epoch": 3.378877246547776, + "grad_norm": 1.1438807249069214, + "learning_rate": 1.2382219443411468e-05, + "loss": 0.211, + "step": 48693 + }, + { + "epoch": 3.3789466379848725, + "grad_norm": 1.0577815771102905, + "learning_rate": 1.2379511476357818e-05, + "loss": 0.3631, + "step": 48694 + }, + { + "epoch": 3.3790160294219693, + "grad_norm": 1.0610594749450684, + "learning_rate": 1.2376803785911895e-05, + "loss": 0.2437, + "step": 48695 + }, + { + "epoch": 3.379085420859066, + "grad_norm": 0.7953948378562927, + "learning_rate": 1.2374096372082256e-05, + "loss": 0.2351, + "step": 48696 + }, + { + "epoch": 3.3791548122961625, + "grad_norm": 0.7775314450263977, + "learning_rate": 1.2371389234877406e-05, + "loss": 0.1999, + "step": 48697 + }, + { + "epoch": 3.3792242037332594, + "grad_norm": 1.3953869342803955, + "learning_rate": 1.2368682374305907e-05, + "loss": 0.233, + "step": 48698 + }, + { + "epoch": 3.379293595170356, + "grad_norm": 0.5131032466888428, + "learning_rate": 1.2365975790376316e-05, + "loss": 0.1102, + "step": 48699 + }, + { + "epoch": 3.3793629866074526, + "grad_norm": 0.6946276426315308, + "learning_rate": 1.2363269483097196e-05, + "loss": 0.2034, + "step": 48700 + }, + { + "epoch": 3.3794323780445494, + "grad_norm": 1.0497589111328125, + "learning_rate": 1.2360563452477047e-05, + "loss": 0.2502, + "step": 48701 + }, + { + "epoch": 3.379501769481646, + "grad_norm": 1.2442851066589355, + "learning_rate": 1.2357857698524422e-05, + "loss": 0.5096, + "step": 48702 + }, + { + "epoch": 3.3795711609187427, + "grad_norm": 1.0073740482330322, + "learning_rate": 1.235515222124788e-05, + "loss": 0.2265, + "step": 48703 + }, + { + "epoch": 3.3796405523558395, + "grad_norm": 1.0420887470245361, + "learning_rate": 1.2352447020655944e-05, + "loss": 0.2499, + "step": 48704 + }, + { + "epoch": 3.379709943792936, + "grad_norm": 1.1045325994491577, + "learning_rate": 1.2349742096757166e-05, + "loss": 0.2379, + "step": 48705 + }, + { + "epoch": 3.3797793352300327, + "grad_norm": 0.8103423118591309, + "learning_rate": 1.2347037449560094e-05, + "loss": 0.1936, + "step": 48706 + }, + { + "epoch": 3.379848726667129, + "grad_norm": 1.1640254259109497, + "learning_rate": 1.2344333079073267e-05, + "loss": 0.2709, + "step": 48707 + }, + { + "epoch": 3.379918118104226, + "grad_norm": 1.2734962701797485, + "learning_rate": 1.2341628985305188e-05, + "loss": 0.2683, + "step": 48708 + }, + { + "epoch": 3.3799875095413228, + "grad_norm": 0.7389503121376038, + "learning_rate": 1.2338925168264426e-05, + "loss": 0.1799, + "step": 48709 + }, + { + "epoch": 3.380056900978419, + "grad_norm": 0.8132947683334351, + "learning_rate": 1.2336221627959498e-05, + "loss": 0.2003, + "step": 48710 + }, + { + "epoch": 3.380126292415516, + "grad_norm": 1.1187883615493774, + "learning_rate": 1.2333518364398943e-05, + "loss": 0.2611, + "step": 48711 + }, + { + "epoch": 3.3801956838526124, + "grad_norm": 0.7619400024414062, + "learning_rate": 1.233081537759131e-05, + "loss": 0.2887, + "step": 48712 + }, + { + "epoch": 3.380265075289709, + "grad_norm": 1.3658554553985596, + "learning_rate": 1.2328112667545122e-05, + "loss": 0.3719, + "step": 48713 + }, + { + "epoch": 3.380334466726806, + "grad_norm": 0.6764602065086365, + "learning_rate": 1.232541023426892e-05, + "loss": 0.2229, + "step": 48714 + }, + { + "epoch": 3.3804038581639024, + "grad_norm": 0.7810477018356323, + "learning_rate": 1.2322708077771216e-05, + "loss": 0.2089, + "step": 48715 + }, + { + "epoch": 3.3804732496009993, + "grad_norm": 1.3554376363754272, + "learning_rate": 1.232000619806054e-05, + "loss": 0.3005, + "step": 48716 + }, + { + "epoch": 3.3805426410380957, + "grad_norm": 1.3825135231018066, + "learning_rate": 1.231730459514544e-05, + "loss": 0.3468, + "step": 48717 + }, + { + "epoch": 3.3806120324751925, + "grad_norm": 1.6135916709899902, + "learning_rate": 1.231460326903443e-05, + "loss": 0.316, + "step": 48718 + }, + { + "epoch": 3.3806814239122893, + "grad_norm": 0.670780599117279, + "learning_rate": 1.2311902219736049e-05, + "loss": 0.2303, + "step": 48719 + }, + { + "epoch": 3.3807508153493857, + "grad_norm": 0.907171905040741, + "learning_rate": 1.2309201447258822e-05, + "loss": 0.2342, + "step": 48720 + }, + { + "epoch": 3.3808202067864825, + "grad_norm": 0.754660964012146, + "learning_rate": 1.2306500951611289e-05, + "loss": 0.204, + "step": 48721 + }, + { + "epoch": 3.3808895982235794, + "grad_norm": 1.1846588850021362, + "learning_rate": 1.230380073280194e-05, + "loss": 0.5283, + "step": 48722 + }, + { + "epoch": 3.3809589896606758, + "grad_norm": 1.4529846906661987, + "learning_rate": 1.2301100790839314e-05, + "loss": 0.4682, + "step": 48723 + }, + { + "epoch": 3.3810283810977726, + "grad_norm": 0.636829674243927, + "learning_rate": 1.229840112573193e-05, + "loss": 0.1987, + "step": 48724 + }, + { + "epoch": 3.3810977725348694, + "grad_norm": 1.1757495403289795, + "learning_rate": 1.229570173748833e-05, + "loss": 0.2439, + "step": 48725 + }, + { + "epoch": 3.381167163971966, + "grad_norm": 0.6951853632926941, + "learning_rate": 1.2293002626117023e-05, + "loss": 0.1868, + "step": 48726 + }, + { + "epoch": 3.3812365554090626, + "grad_norm": 0.7332814931869507, + "learning_rate": 1.2290303791626545e-05, + "loss": 0.1977, + "step": 48727 + }, + { + "epoch": 3.381305946846159, + "grad_norm": 0.8168799877166748, + "learning_rate": 1.228760523402539e-05, + "loss": 0.2268, + "step": 48728 + }, + { + "epoch": 3.381375338283256, + "grad_norm": 0.8860392570495605, + "learning_rate": 1.2284906953322083e-05, + "loss": 0.3002, + "step": 48729 + }, + { + "epoch": 3.3814447297203527, + "grad_norm": 0.5652479529380798, + "learning_rate": 1.2282208949525165e-05, + "loss": 0.2166, + "step": 48730 + }, + { + "epoch": 3.381514121157449, + "grad_norm": 0.9352321624755859, + "learning_rate": 1.2279511222643125e-05, + "loss": 0.1959, + "step": 48731 + }, + { + "epoch": 3.381583512594546, + "grad_norm": 1.0062568187713623, + "learning_rate": 1.2276813772684481e-05, + "loss": 0.2803, + "step": 48732 + }, + { + "epoch": 3.3816529040316423, + "grad_norm": 0.6418304443359375, + "learning_rate": 1.227411659965777e-05, + "loss": 0.1677, + "step": 48733 + }, + { + "epoch": 3.381722295468739, + "grad_norm": 0.8169299364089966, + "learning_rate": 1.2271419703571507e-05, + "loss": 0.212, + "step": 48734 + }, + { + "epoch": 3.381791686905836, + "grad_norm": 0.7787527441978455, + "learning_rate": 1.2268723084434175e-05, + "loss": 0.1457, + "step": 48735 + }, + { + "epoch": 3.3818610783429324, + "grad_norm": 0.9531768560409546, + "learning_rate": 1.22660267422543e-05, + "loss": 0.3293, + "step": 48736 + }, + { + "epoch": 3.381930469780029, + "grad_norm": 0.7708117365837097, + "learning_rate": 1.2263330677040408e-05, + "loss": 0.1812, + "step": 48737 + }, + { + "epoch": 3.3819998612171256, + "grad_norm": 0.8245618343353271, + "learning_rate": 1.2260634888800993e-05, + "loss": 0.3003, + "step": 48738 + }, + { + "epoch": 3.3820692526542224, + "grad_norm": 0.986952543258667, + "learning_rate": 1.225793937754458e-05, + "loss": 0.2808, + "step": 48739 + }, + { + "epoch": 3.3821386440913193, + "grad_norm": 0.8895391821861267, + "learning_rate": 1.2255244143279676e-05, + "loss": 0.202, + "step": 48740 + }, + { + "epoch": 3.3822080355284156, + "grad_norm": 0.7111117243766785, + "learning_rate": 1.2252549186014794e-05, + "loss": 0.2231, + "step": 48741 + }, + { + "epoch": 3.3822774269655125, + "grad_norm": 1.0199507474899292, + "learning_rate": 1.224985450575843e-05, + "loss": 0.3104, + "step": 48742 + }, + { + "epoch": 3.3823468184026093, + "grad_norm": 1.2580803632736206, + "learning_rate": 1.2247160102519084e-05, + "loss": 0.3966, + "step": 48743 + }, + { + "epoch": 3.3824162098397057, + "grad_norm": 0.8032681345939636, + "learning_rate": 1.2244465976305275e-05, + "loss": 0.2281, + "step": 48744 + }, + { + "epoch": 3.3824856012768025, + "grad_norm": 0.7462274432182312, + "learning_rate": 1.2241772127125495e-05, + "loss": 0.2079, + "step": 48745 + }, + { + "epoch": 3.382554992713899, + "grad_norm": 0.7363151907920837, + "learning_rate": 1.223907855498827e-05, + "loss": 0.223, + "step": 48746 + }, + { + "epoch": 3.3826243841509958, + "grad_norm": 0.8208132982254028, + "learning_rate": 1.2236385259902095e-05, + "loss": 0.1718, + "step": 48747 + }, + { + "epoch": 3.3826937755880926, + "grad_norm": 2.0283560752868652, + "learning_rate": 1.2233692241875483e-05, + "loss": 0.2675, + "step": 48748 + }, + { + "epoch": 3.382763167025189, + "grad_norm": 0.6452485918998718, + "learning_rate": 1.2230999500916895e-05, + "loss": 0.1764, + "step": 48749 + }, + { + "epoch": 3.382832558462286, + "grad_norm": 1.0634649991989136, + "learning_rate": 1.222830703703487e-05, + "loss": 0.1992, + "step": 48750 + }, + { + "epoch": 3.3829019498993826, + "grad_norm": 0.9346541166305542, + "learning_rate": 1.2225614850237898e-05, + "loss": 0.2114, + "step": 48751 + }, + { + "epoch": 3.382971341336479, + "grad_norm": 0.7896780967712402, + "learning_rate": 1.2222922940534464e-05, + "loss": 0.2845, + "step": 48752 + }, + { + "epoch": 3.383040732773576, + "grad_norm": 0.9425451755523682, + "learning_rate": 1.2220231307933084e-05, + "loss": 0.2706, + "step": 48753 + }, + { + "epoch": 3.3831101242106723, + "grad_norm": 0.9942271113395691, + "learning_rate": 1.2217539952442258e-05, + "loss": 0.2762, + "step": 48754 + }, + { + "epoch": 3.383179515647769, + "grad_norm": 1.395437240600586, + "learning_rate": 1.2214848874070483e-05, + "loss": 0.2592, + "step": 48755 + }, + { + "epoch": 3.383248907084866, + "grad_norm": 0.912467896938324, + "learning_rate": 1.2212158072826229e-05, + "loss": 0.25, + "step": 48756 + }, + { + "epoch": 3.3833182985219623, + "grad_norm": 0.9912781715393066, + "learning_rate": 1.2209467548718001e-05, + "loss": 0.2574, + "step": 48757 + }, + { + "epoch": 3.383387689959059, + "grad_norm": 0.8562334775924683, + "learning_rate": 1.2206777301754302e-05, + "loss": 0.215, + "step": 48758 + }, + { + "epoch": 3.3834570813961555, + "grad_norm": 1.011940598487854, + "learning_rate": 1.2204087331943625e-05, + "loss": 0.255, + "step": 48759 + }, + { + "epoch": 3.3835264728332524, + "grad_norm": 0.76106196641922, + "learning_rate": 1.2201397639294454e-05, + "loss": 0.1265, + "step": 48760 + }, + { + "epoch": 3.383595864270349, + "grad_norm": 0.7375674843788147, + "learning_rate": 1.2198708223815291e-05, + "loss": 0.1865, + "step": 48761 + }, + { + "epoch": 3.3836652557074456, + "grad_norm": 0.8046929836273193, + "learning_rate": 1.2196019085514632e-05, + "loss": 0.22, + "step": 48762 + }, + { + "epoch": 3.3837346471445424, + "grad_norm": 0.8075157999992371, + "learning_rate": 1.2193330224400935e-05, + "loss": 0.2147, + "step": 48763 + }, + { + "epoch": 3.383804038581639, + "grad_norm": 0.8806634545326233, + "learning_rate": 1.2190641640482703e-05, + "loss": 0.2281, + "step": 48764 + }, + { + "epoch": 3.3838734300187356, + "grad_norm": 0.7522954344749451, + "learning_rate": 1.2187953333768453e-05, + "loss": 0.2365, + "step": 48765 + }, + { + "epoch": 3.3839428214558325, + "grad_norm": 1.0862705707550049, + "learning_rate": 1.2185265304266636e-05, + "loss": 0.3039, + "step": 48766 + }, + { + "epoch": 3.384012212892929, + "grad_norm": 1.3319584131240845, + "learning_rate": 1.218257755198573e-05, + "loss": 0.3559, + "step": 48767 + }, + { + "epoch": 3.3840816043300257, + "grad_norm": 0.8167205452919006, + "learning_rate": 1.2179890076934242e-05, + "loss": 0.2806, + "step": 48768 + }, + { + "epoch": 3.3841509957671225, + "grad_norm": 0.8199507594108582, + "learning_rate": 1.2177202879120686e-05, + "loss": 0.2487, + "step": 48769 + }, + { + "epoch": 3.384220387204219, + "grad_norm": 1.004460096359253, + "learning_rate": 1.2174515958553479e-05, + "loss": 0.1314, + "step": 48770 + }, + { + "epoch": 3.3842897786413157, + "grad_norm": 1.4721170663833618, + "learning_rate": 1.2171829315241135e-05, + "loss": 0.22, + "step": 48771 + }, + { + "epoch": 3.384359170078412, + "grad_norm": 0.7820828557014465, + "learning_rate": 1.2169142949192136e-05, + "loss": 0.2827, + "step": 48772 + }, + { + "epoch": 3.384428561515509, + "grad_norm": 1.6347016096115112, + "learning_rate": 1.2166456860414966e-05, + "loss": 0.6007, + "step": 48773 + }, + { + "epoch": 3.384497952952606, + "grad_norm": 1.0905462503433228, + "learning_rate": 1.2163771048918093e-05, + "loss": 0.277, + "step": 48774 + }, + { + "epoch": 3.384567344389702, + "grad_norm": 1.148818016052246, + "learning_rate": 1.2161085514710013e-05, + "loss": 0.32, + "step": 48775 + }, + { + "epoch": 3.384636735826799, + "grad_norm": 0.8152585029602051, + "learning_rate": 1.2158400257799218e-05, + "loss": 0.3081, + "step": 48776 + }, + { + "epoch": 3.384706127263896, + "grad_norm": 1.0158705711364746, + "learning_rate": 1.2155715278194135e-05, + "loss": 0.2466, + "step": 48777 + }, + { + "epoch": 3.3847755187009922, + "grad_norm": 0.6714869737625122, + "learning_rate": 1.2153030575903268e-05, + "loss": 0.2004, + "step": 48778 + }, + { + "epoch": 3.384844910138089, + "grad_norm": 1.3350715637207031, + "learning_rate": 1.2150346150935088e-05, + "loss": 0.2099, + "step": 48779 + }, + { + "epoch": 3.3849143015751855, + "grad_norm": 1.1391665935516357, + "learning_rate": 1.2147662003298077e-05, + "loss": 0.2872, + "step": 48780 + }, + { + "epoch": 3.3849836930122823, + "grad_norm": 1.4134879112243652, + "learning_rate": 1.2144978133000706e-05, + "loss": 0.2464, + "step": 48781 + }, + { + "epoch": 3.385053084449379, + "grad_norm": 1.6433417797088623, + "learning_rate": 1.2142294540051435e-05, + "loss": 0.2973, + "step": 48782 + }, + { + "epoch": 3.3851224758864755, + "grad_norm": 0.8110584020614624, + "learning_rate": 1.213961122445878e-05, + "loss": 0.2835, + "step": 48783 + }, + { + "epoch": 3.3851918673235724, + "grad_norm": 1.0252389907836914, + "learning_rate": 1.2136928186231156e-05, + "loss": 0.195, + "step": 48784 + }, + { + "epoch": 3.3852612587606687, + "grad_norm": 0.7058817744255066, + "learning_rate": 1.2134245425377056e-05, + "loss": 0.1607, + "step": 48785 + }, + { + "epoch": 3.3853306501977656, + "grad_norm": 0.7042855620384216, + "learning_rate": 1.2131562941904939e-05, + "loss": 0.1601, + "step": 48786 + }, + { + "epoch": 3.3854000416348624, + "grad_norm": 0.8699474930763245, + "learning_rate": 1.21288807358233e-05, + "loss": 0.2238, + "step": 48787 + }, + { + "epoch": 3.385469433071959, + "grad_norm": 1.2828516960144043, + "learning_rate": 1.2126198807140576e-05, + "loss": 0.2267, + "step": 48788 + }, + { + "epoch": 3.3855388245090556, + "grad_norm": 0.7535219788551331, + "learning_rate": 1.212351715586525e-05, + "loss": 0.2109, + "step": 48789 + }, + { + "epoch": 3.385608215946152, + "grad_norm": 0.9926596283912659, + "learning_rate": 1.2120835782005813e-05, + "loss": 0.3319, + "step": 48790 + }, + { + "epoch": 3.385677607383249, + "grad_norm": 0.7699077129364014, + "learning_rate": 1.211815468557067e-05, + "loss": 0.2148, + "step": 48791 + }, + { + "epoch": 3.3857469988203457, + "grad_norm": 1.2313165664672852, + "learning_rate": 1.2115473866568316e-05, + "loss": 0.5122, + "step": 48792 + }, + { + "epoch": 3.385816390257442, + "grad_norm": 0.7862423658370972, + "learning_rate": 1.211279332500722e-05, + "loss": 0.2011, + "step": 48793 + }, + { + "epoch": 3.385885781694539, + "grad_norm": 0.888131320476532, + "learning_rate": 1.2110113060895845e-05, + "loss": 0.306, + "step": 48794 + }, + { + "epoch": 3.3859551731316357, + "grad_norm": 0.8675158023834229, + "learning_rate": 1.2107433074242635e-05, + "loss": 0.1868, + "step": 48795 + }, + { + "epoch": 3.386024564568732, + "grad_norm": 0.6548888087272644, + "learning_rate": 1.2104753365056087e-05, + "loss": 0.0769, + "step": 48796 + }, + { + "epoch": 3.386093956005829, + "grad_norm": 0.7660921216011047, + "learning_rate": 1.2102073933344615e-05, + "loss": 0.1694, + "step": 48797 + }, + { + "epoch": 3.3861633474429254, + "grad_norm": 1.0168817043304443, + "learning_rate": 1.209939477911669e-05, + "loss": 0.214, + "step": 48798 + }, + { + "epoch": 3.386232738880022, + "grad_norm": 0.9774031043052673, + "learning_rate": 1.2096715902380784e-05, + "loss": 0.2262, + "step": 48799 + }, + { + "epoch": 3.386302130317119, + "grad_norm": 1.5527055263519287, + "learning_rate": 1.2094037303145366e-05, + "loss": 0.4623, + "step": 48800 + }, + { + "epoch": 3.3863715217542154, + "grad_norm": 0.8706364035606384, + "learning_rate": 1.2091358981418866e-05, + "loss": 0.306, + "step": 48801 + }, + { + "epoch": 3.3864409131913122, + "grad_norm": 0.8086511492729187, + "learning_rate": 1.2088680937209729e-05, + "loss": 0.2331, + "step": 48802 + }, + { + "epoch": 3.386510304628409, + "grad_norm": 0.9247791767120361, + "learning_rate": 1.2086003170526427e-05, + "loss": 0.2758, + "step": 48803 + }, + { + "epoch": 3.3865796960655055, + "grad_norm": 0.8995643258094788, + "learning_rate": 1.2083325681377445e-05, + "loss": 0.212, + "step": 48804 + }, + { + "epoch": 3.3866490875026023, + "grad_norm": 1.741429090499878, + "learning_rate": 1.2080648469771171e-05, + "loss": 0.3395, + "step": 48805 + }, + { + "epoch": 3.3867184789396987, + "grad_norm": 0.9858139753341675, + "learning_rate": 1.2077971535716104e-05, + "loss": 0.2619, + "step": 48806 + }, + { + "epoch": 3.3867878703767955, + "grad_norm": 0.5658584833145142, + "learning_rate": 1.2075294879220666e-05, + "loss": 0.1979, + "step": 48807 + }, + { + "epoch": 3.3868572618138924, + "grad_norm": 1.1355377435684204, + "learning_rate": 1.207261850029332e-05, + "loss": 0.2428, + "step": 48808 + }, + { + "epoch": 3.3869266532509887, + "grad_norm": 0.8725718259811401, + "learning_rate": 1.2069942398942524e-05, + "loss": 0.2216, + "step": 48809 + }, + { + "epoch": 3.3869960446880856, + "grad_norm": 1.1076910495758057, + "learning_rate": 1.206726657517674e-05, + "loss": 0.2803, + "step": 48810 + }, + { + "epoch": 3.387065436125182, + "grad_norm": 1.0178568363189697, + "learning_rate": 1.2064591029004368e-05, + "loss": 0.2696, + "step": 48811 + }, + { + "epoch": 3.387134827562279, + "grad_norm": 1.156840443611145, + "learning_rate": 1.2061915760433884e-05, + "loss": 0.4173, + "step": 48812 + }, + { + "epoch": 3.3872042189993756, + "grad_norm": 0.6846333146095276, + "learning_rate": 1.2059240769473722e-05, + "loss": 0.1958, + "step": 48813 + }, + { + "epoch": 3.387273610436472, + "grad_norm": 1.5754936933517456, + "learning_rate": 1.2056566056132334e-05, + "loss": 0.4407, + "step": 48814 + }, + { + "epoch": 3.387343001873569, + "grad_norm": 0.8707669377326965, + "learning_rate": 1.2053891620418168e-05, + "loss": 0.2547, + "step": 48815 + }, + { + "epoch": 3.3874123933106652, + "grad_norm": 1.0546801090240479, + "learning_rate": 1.205121746233967e-05, + "loss": 0.3302, + "step": 48816 + }, + { + "epoch": 3.387481784747762, + "grad_norm": 0.9729058742523193, + "learning_rate": 1.2048543581905281e-05, + "loss": 0.2309, + "step": 48817 + }, + { + "epoch": 3.387551176184859, + "grad_norm": 0.9977705478668213, + "learning_rate": 1.2045869979123425e-05, + "loss": 0.2501, + "step": 48818 + }, + { + "epoch": 3.3876205676219553, + "grad_norm": 0.7859445810317993, + "learning_rate": 1.2043196654002554e-05, + "loss": 0.2758, + "step": 48819 + }, + { + "epoch": 3.387689959059052, + "grad_norm": 0.7138611078262329, + "learning_rate": 1.2040523606551102e-05, + "loss": 0.1381, + "step": 48820 + }, + { + "epoch": 3.387759350496149, + "grad_norm": 1.1179853677749634, + "learning_rate": 1.203785083677751e-05, + "loss": 0.3043, + "step": 48821 + }, + { + "epoch": 3.3878287419332453, + "grad_norm": 0.8221741318702698, + "learning_rate": 1.2035178344690213e-05, + "loss": 0.2021, + "step": 48822 + }, + { + "epoch": 3.387898133370342, + "grad_norm": 0.7392385601997375, + "learning_rate": 1.203250613029766e-05, + "loss": 0.2413, + "step": 48823 + }, + { + "epoch": 3.3879675248074386, + "grad_norm": 0.7349449992179871, + "learning_rate": 1.2029834193608292e-05, + "loss": 0.177, + "step": 48824 + }, + { + "epoch": 3.3880369162445354, + "grad_norm": 0.688428521156311, + "learning_rate": 1.202716253463052e-05, + "loss": 0.208, + "step": 48825 + }, + { + "epoch": 3.3881063076816322, + "grad_norm": 0.9728122353553772, + "learning_rate": 1.2024491153372786e-05, + "loss": 0.3058, + "step": 48826 + }, + { + "epoch": 3.3881756991187286, + "grad_norm": 0.9389646649360657, + "learning_rate": 1.2021820049843524e-05, + "loss": 0.2741, + "step": 48827 + }, + { + "epoch": 3.3882450905558255, + "grad_norm": 0.6819685697555542, + "learning_rate": 1.2019149224051174e-05, + "loss": 0.2441, + "step": 48828 + }, + { + "epoch": 3.3883144819929223, + "grad_norm": 1.1645689010620117, + "learning_rate": 1.201647867600415e-05, + "loss": 0.2479, + "step": 48829 + }, + { + "epoch": 3.3883838734300187, + "grad_norm": 0.9282070398330688, + "learning_rate": 1.2013808405710913e-05, + "loss": 0.2731, + "step": 48830 + }, + { + "epoch": 3.3884532648671155, + "grad_norm": 0.8235100507736206, + "learning_rate": 1.2011138413179878e-05, + "loss": 0.2693, + "step": 48831 + }, + { + "epoch": 3.388522656304212, + "grad_norm": 1.0615442991256714, + "learning_rate": 1.2008468698419461e-05, + "loss": 0.2662, + "step": 48832 + }, + { + "epoch": 3.3885920477413087, + "grad_norm": 0.8112691044807434, + "learning_rate": 1.20057992614381e-05, + "loss": 0.1448, + "step": 48833 + }, + { + "epoch": 3.3886614391784056, + "grad_norm": 1.4554545879364014, + "learning_rate": 1.200313010224421e-05, + "loss": 0.3661, + "step": 48834 + }, + { + "epoch": 3.388730830615502, + "grad_norm": 0.7309998273849487, + "learning_rate": 1.200046122084626e-05, + "loss": 0.1368, + "step": 48835 + }, + { + "epoch": 3.388800222052599, + "grad_norm": 0.8177822828292847, + "learning_rate": 1.1997792617252623e-05, + "loss": 0.2368, + "step": 48836 + }, + { + "epoch": 3.388869613489695, + "grad_norm": 0.8443405032157898, + "learning_rate": 1.1995124291471737e-05, + "loss": 0.1953, + "step": 48837 + }, + { + "epoch": 3.388939004926792, + "grad_norm": 1.5862921476364136, + "learning_rate": 1.199245624351204e-05, + "loss": 0.2296, + "step": 48838 + }, + { + "epoch": 3.389008396363889, + "grad_norm": 0.8354364633560181, + "learning_rate": 1.198978847338197e-05, + "loss": 0.2313, + "step": 48839 + }, + { + "epoch": 3.3890777878009852, + "grad_norm": 0.7023648023605347, + "learning_rate": 1.1987120981089894e-05, + "loss": 0.2271, + "step": 48840 + }, + { + "epoch": 3.389147179238082, + "grad_norm": 1.0154776573181152, + "learning_rate": 1.1984453766644277e-05, + "loss": 0.3219, + "step": 48841 + }, + { + "epoch": 3.3892165706751785, + "grad_norm": 0.9909616112709045, + "learning_rate": 1.198178683005352e-05, + "loss": 0.343, + "step": 48842 + }, + { + "epoch": 3.3892859621122753, + "grad_norm": 1.3005930185317993, + "learning_rate": 1.1979120171326064e-05, + "loss": 0.1847, + "step": 48843 + }, + { + "epoch": 3.389355353549372, + "grad_norm": 1.5989797115325928, + "learning_rate": 1.1976453790470299e-05, + "loss": 0.3062, + "step": 48844 + }, + { + "epoch": 3.3894247449864685, + "grad_norm": 0.7546305656433105, + "learning_rate": 1.1973787687494686e-05, + "loss": 0.237, + "step": 48845 + }, + { + "epoch": 3.3894941364235653, + "grad_norm": 0.744929313659668, + "learning_rate": 1.1971121862407587e-05, + "loss": 0.201, + "step": 48846 + }, + { + "epoch": 3.389563527860662, + "grad_norm": 1.0133562088012695, + "learning_rate": 1.1968456315217447e-05, + "loss": 0.3642, + "step": 48847 + }, + { + "epoch": 3.3896329192977586, + "grad_norm": 0.8632287383079529, + "learning_rate": 1.1965791045932673e-05, + "loss": 0.2795, + "step": 48848 + }, + { + "epoch": 3.3897023107348554, + "grad_norm": 0.8175297379493713, + "learning_rate": 1.196312605456168e-05, + "loss": 0.2101, + "step": 48849 + }, + { + "epoch": 3.3897717021719522, + "grad_norm": 0.7598969340324402, + "learning_rate": 1.1960461341112884e-05, + "loss": 0.1969, + "step": 48850 + }, + { + "epoch": 3.3898410936090486, + "grad_norm": 1.6692982912063599, + "learning_rate": 1.195779690559471e-05, + "loss": 0.2912, + "step": 48851 + }, + { + "epoch": 3.3899104850461454, + "grad_norm": 0.8827612996101379, + "learning_rate": 1.1955132748015562e-05, + "loss": 0.2684, + "step": 48852 + }, + { + "epoch": 3.389979876483242, + "grad_norm": 0.6241944432258606, + "learning_rate": 1.1952468868383826e-05, + "loss": 0.1769, + "step": 48853 + }, + { + "epoch": 3.3900492679203387, + "grad_norm": 1.1041460037231445, + "learning_rate": 1.1949805266707936e-05, + "loss": 0.3376, + "step": 48854 + }, + { + "epoch": 3.3901186593574355, + "grad_norm": 0.6162348985671997, + "learning_rate": 1.1947141942996298e-05, + "loss": 0.1802, + "step": 48855 + }, + { + "epoch": 3.390188050794532, + "grad_norm": 0.8470408320426941, + "learning_rate": 1.1944478897257305e-05, + "loss": 0.2628, + "step": 48856 + }, + { + "epoch": 3.3902574422316287, + "grad_norm": 0.920171856880188, + "learning_rate": 1.1941816129499383e-05, + "loss": 0.3094, + "step": 48857 + }, + { + "epoch": 3.390326833668725, + "grad_norm": 0.9501824378967285, + "learning_rate": 1.1939153639730938e-05, + "loss": 0.225, + "step": 48858 + }, + { + "epoch": 3.390396225105822, + "grad_norm": 1.0619412660598755, + "learning_rate": 1.1936491427960383e-05, + "loss": 0.3257, + "step": 48859 + }, + { + "epoch": 3.390465616542919, + "grad_norm": 1.003876805305481, + "learning_rate": 1.1933829494196081e-05, + "loss": 0.2717, + "step": 48860 + }, + { + "epoch": 3.390535007980015, + "grad_norm": 1.1027761697769165, + "learning_rate": 1.1931167838446466e-05, + "loss": 0.2008, + "step": 48861 + }, + { + "epoch": 3.390604399417112, + "grad_norm": 1.9079363346099854, + "learning_rate": 1.1928506460719935e-05, + "loss": 0.1646, + "step": 48862 + }, + { + "epoch": 3.3906737908542084, + "grad_norm": 1.02183198928833, + "learning_rate": 1.192584536102489e-05, + "loss": 0.1853, + "step": 48863 + }, + { + "epoch": 3.3907431822913052, + "grad_norm": 0.7354658246040344, + "learning_rate": 1.1923184539369736e-05, + "loss": 0.2872, + "step": 48864 + }, + { + "epoch": 3.390812573728402, + "grad_norm": 1.0039753913879395, + "learning_rate": 1.1920523995762866e-05, + "loss": 0.2455, + "step": 48865 + }, + { + "epoch": 3.3908819651654984, + "grad_norm": 0.7230811715126038, + "learning_rate": 1.1917863730212708e-05, + "loss": 0.2069, + "step": 48866 + }, + { + "epoch": 3.3909513566025953, + "grad_norm": 1.154908537864685, + "learning_rate": 1.191520374272762e-05, + "loss": 0.2817, + "step": 48867 + }, + { + "epoch": 3.3910207480396917, + "grad_norm": 0.8795780539512634, + "learning_rate": 1.1912544033315998e-05, + "loss": 0.2224, + "step": 48868 + }, + { + "epoch": 3.3910901394767885, + "grad_norm": 1.1211960315704346, + "learning_rate": 1.1909884601986265e-05, + "loss": 0.1941, + "step": 48869 + }, + { + "epoch": 3.3911595309138853, + "grad_norm": 0.9593278169631958, + "learning_rate": 1.1907225448746829e-05, + "loss": 0.1953, + "step": 48870 + }, + { + "epoch": 3.3912289223509817, + "grad_norm": 0.8995845317840576, + "learning_rate": 1.1904566573606034e-05, + "loss": 0.1435, + "step": 48871 + }, + { + "epoch": 3.3912983137880786, + "grad_norm": 1.0098552703857422, + "learning_rate": 1.1901907976572303e-05, + "loss": 0.32, + "step": 48872 + }, + { + "epoch": 3.3913677052251754, + "grad_norm": 1.4505277872085571, + "learning_rate": 1.1899249657654043e-05, + "loss": 0.2516, + "step": 48873 + }, + { + "epoch": 3.3914370966622718, + "grad_norm": 1.098157525062561, + "learning_rate": 1.189659161685962e-05, + "loss": 0.3362, + "step": 48874 + }, + { + "epoch": 3.3915064880993686, + "grad_norm": 1.1259803771972656, + "learning_rate": 1.1893933854197425e-05, + "loss": 0.2612, + "step": 48875 + }, + { + "epoch": 3.3915758795364654, + "grad_norm": 1.133916974067688, + "learning_rate": 1.1891276369675863e-05, + "loss": 0.2353, + "step": 48876 + }, + { + "epoch": 3.391645270973562, + "grad_norm": 0.5481031537055969, + "learning_rate": 1.1888619163303328e-05, + "loss": 0.1534, + "step": 48877 + }, + { + "epoch": 3.3917146624106587, + "grad_norm": 0.6123114824295044, + "learning_rate": 1.1885962235088188e-05, + "loss": 0.1563, + "step": 48878 + }, + { + "epoch": 3.391784053847755, + "grad_norm": 0.9349907040596008, + "learning_rate": 1.1883305585038862e-05, + "loss": 0.1961, + "step": 48879 + }, + { + "epoch": 3.391853445284852, + "grad_norm": 1.3727656602859497, + "learning_rate": 1.1880649213163709e-05, + "loss": 0.2136, + "step": 48880 + }, + { + "epoch": 3.3919228367219487, + "grad_norm": 1.0504533052444458, + "learning_rate": 1.1877993119471108e-05, + "loss": 0.1763, + "step": 48881 + }, + { + "epoch": 3.391992228159045, + "grad_norm": 2.381209135055542, + "learning_rate": 1.187533730396947e-05, + "loss": 0.3344, + "step": 48882 + }, + { + "epoch": 3.392061619596142, + "grad_norm": 0.8622912764549255, + "learning_rate": 1.187268176666716e-05, + "loss": 0.2061, + "step": 48883 + }, + { + "epoch": 3.3921310110332383, + "grad_norm": 0.8542079925537109, + "learning_rate": 1.1870026507572573e-05, + "loss": 0.2437, + "step": 48884 + }, + { + "epoch": 3.392200402470335, + "grad_norm": 0.7952082753181458, + "learning_rate": 1.1867371526694082e-05, + "loss": 0.2327, + "step": 48885 + }, + { + "epoch": 3.392269793907432, + "grad_norm": 0.8173017501831055, + "learning_rate": 1.1864716824040101e-05, + "loss": 0.2377, + "step": 48886 + }, + { + "epoch": 3.3923391853445284, + "grad_norm": 0.8279679417610168, + "learning_rate": 1.1862062399618956e-05, + "loss": 0.3226, + "step": 48887 + }, + { + "epoch": 3.392408576781625, + "grad_norm": 0.9478362798690796, + "learning_rate": 1.1859408253439063e-05, + "loss": 0.2954, + "step": 48888 + }, + { + "epoch": 3.3924779682187216, + "grad_norm": 1.0031318664550781, + "learning_rate": 1.1856754385508784e-05, + "loss": 0.2954, + "step": 48889 + }, + { + "epoch": 3.3925473596558184, + "grad_norm": 1.0375614166259766, + "learning_rate": 1.185410079583651e-05, + "loss": 0.2353, + "step": 48890 + }, + { + "epoch": 3.3926167510929153, + "grad_norm": 0.9977974891662598, + "learning_rate": 1.185144748443061e-05, + "loss": 0.1796, + "step": 48891 + }, + { + "epoch": 3.3926861425300117, + "grad_norm": 1.056217908859253, + "learning_rate": 1.184879445129946e-05, + "loss": 0.2227, + "step": 48892 + }, + { + "epoch": 3.3927555339671085, + "grad_norm": 1.3239845037460327, + "learning_rate": 1.1846141696451462e-05, + "loss": 0.3358, + "step": 48893 + }, + { + "epoch": 3.392824925404205, + "grad_norm": 1.2827316522598267, + "learning_rate": 1.1843489219894955e-05, + "loss": 0.3749, + "step": 48894 + }, + { + "epoch": 3.3928943168413017, + "grad_norm": 1.1799068450927734, + "learning_rate": 1.1840837021638319e-05, + "loss": 0.234, + "step": 48895 + }, + { + "epoch": 3.3929637082783985, + "grad_norm": 0.7123247981071472, + "learning_rate": 1.1838185101689924e-05, + "loss": 0.2476, + "step": 48896 + }, + { + "epoch": 3.393033099715495, + "grad_norm": 0.7179297208786011, + "learning_rate": 1.1835533460058157e-05, + "loss": 0.2147, + "step": 48897 + }, + { + "epoch": 3.3931024911525918, + "grad_norm": 0.7481122612953186, + "learning_rate": 1.1832882096751385e-05, + "loss": 0.1939, + "step": 48898 + }, + { + "epoch": 3.3931718825896886, + "grad_norm": 5.909392356872559, + "learning_rate": 1.183023101177797e-05, + "loss": 0.2107, + "step": 48899 + }, + { + "epoch": 3.393241274026785, + "grad_norm": 1.2074377536773682, + "learning_rate": 1.1827580205146305e-05, + "loss": 0.3595, + "step": 48900 + }, + { + "epoch": 3.393310665463882, + "grad_norm": 0.7202327847480774, + "learning_rate": 1.1824929676864716e-05, + "loss": 0.2257, + "step": 48901 + }, + { + "epoch": 3.3933800569009787, + "grad_norm": 0.9442741870880127, + "learning_rate": 1.1822279426941607e-05, + "loss": 0.2285, + "step": 48902 + }, + { + "epoch": 3.393449448338075, + "grad_norm": 0.9367694854736328, + "learning_rate": 1.1819629455385316e-05, + "loss": 0.1753, + "step": 48903 + }, + { + "epoch": 3.393518839775172, + "grad_norm": 0.9171451926231384, + "learning_rate": 1.1816979762204239e-05, + "loss": 0.2724, + "step": 48904 + }, + { + "epoch": 3.3935882312122683, + "grad_norm": 0.9784044623374939, + "learning_rate": 1.1814330347406733e-05, + "loss": 0.2757, + "step": 48905 + }, + { + "epoch": 3.393657622649365, + "grad_norm": 1.4923752546310425, + "learning_rate": 1.1811681211001147e-05, + "loss": 0.3404, + "step": 48906 + }, + { + "epoch": 3.393727014086462, + "grad_norm": 0.9730826020240784, + "learning_rate": 1.1809032352995842e-05, + "loss": 0.2353, + "step": 48907 + }, + { + "epoch": 3.3937964055235583, + "grad_norm": 0.9966667294502258, + "learning_rate": 1.180638377339921e-05, + "loss": 0.3441, + "step": 48908 + }, + { + "epoch": 3.393865796960655, + "grad_norm": 0.9544702768325806, + "learning_rate": 1.180373547221958e-05, + "loss": 0.252, + "step": 48909 + }, + { + "epoch": 3.3939351883977515, + "grad_norm": 0.728150486946106, + "learning_rate": 1.180108744946532e-05, + "loss": 0.2481, + "step": 48910 + }, + { + "epoch": 3.3940045798348484, + "grad_norm": 1.0483936071395874, + "learning_rate": 1.1798439705144793e-05, + "loss": 0.1835, + "step": 48911 + }, + { + "epoch": 3.394073971271945, + "grad_norm": 0.7229155898094177, + "learning_rate": 1.1795792239266356e-05, + "loss": 0.1399, + "step": 48912 + }, + { + "epoch": 3.3941433627090416, + "grad_norm": 0.7592240571975708, + "learning_rate": 1.1793145051838373e-05, + "loss": 0.2374, + "step": 48913 + }, + { + "epoch": 3.3942127541461384, + "grad_norm": 1.1883121728897095, + "learning_rate": 1.1790498142869211e-05, + "loss": 0.2833, + "step": 48914 + }, + { + "epoch": 3.394282145583235, + "grad_norm": 0.7789404392242432, + "learning_rate": 1.1787851512367198e-05, + "loss": 0.1818, + "step": 48915 + }, + { + "epoch": 3.3943515370203317, + "grad_norm": 0.7591878175735474, + "learning_rate": 1.1785205160340707e-05, + "loss": 0.1459, + "step": 48916 + }, + { + "epoch": 3.3944209284574285, + "grad_norm": 0.9679160118103027, + "learning_rate": 1.1782559086798084e-05, + "loss": 0.34, + "step": 48917 + }, + { + "epoch": 3.394490319894525, + "grad_norm": 0.8472112417221069, + "learning_rate": 1.1779913291747679e-05, + "loss": 0.3198, + "step": 48918 + }, + { + "epoch": 3.3945597113316217, + "grad_norm": 1.3816782236099243, + "learning_rate": 1.1777267775197865e-05, + "loss": 0.2925, + "step": 48919 + }, + { + "epoch": 3.394629102768718, + "grad_norm": 0.42153117060661316, + "learning_rate": 1.1774622537156976e-05, + "loss": 0.0857, + "step": 48920 + }, + { + "epoch": 3.394698494205815, + "grad_norm": 0.9943657517433167, + "learning_rate": 1.1771977577633387e-05, + "loss": 0.3518, + "step": 48921 + }, + { + "epoch": 3.3947678856429118, + "grad_norm": 1.3753976821899414, + "learning_rate": 1.176933289663541e-05, + "loss": 0.306, + "step": 48922 + }, + { + "epoch": 3.394837277080008, + "grad_norm": 0.8009050488471985, + "learning_rate": 1.1766688494171408e-05, + "loss": 0.2291, + "step": 48923 + }, + { + "epoch": 3.394906668517105, + "grad_norm": 1.061633825302124, + "learning_rate": 1.1764044370249738e-05, + "loss": 0.3239, + "step": 48924 + }, + { + "epoch": 3.394976059954202, + "grad_norm": 0.8507673144340515, + "learning_rate": 1.176140052487874e-05, + "loss": 0.2624, + "step": 48925 + }, + { + "epoch": 3.395045451391298, + "grad_norm": 0.7464807033538818, + "learning_rate": 1.1758756958066774e-05, + "loss": 0.1845, + "step": 48926 + }, + { + "epoch": 3.395114842828395, + "grad_norm": 0.8093045353889465, + "learning_rate": 1.1756113669822155e-05, + "loss": 0.3483, + "step": 48927 + }, + { + "epoch": 3.395184234265492, + "grad_norm": 1.028654932975769, + "learning_rate": 1.1753470660153288e-05, + "loss": 0.327, + "step": 48928 + }, + { + "epoch": 3.3952536257025883, + "grad_norm": 1.2319940328598022, + "learning_rate": 1.1750827929068442e-05, + "loss": 0.2282, + "step": 48929 + }, + { + "epoch": 3.395323017139685, + "grad_norm": 0.6568204760551453, + "learning_rate": 1.174818547657599e-05, + "loss": 0.1909, + "step": 48930 + }, + { + "epoch": 3.3953924085767815, + "grad_norm": 1.1563893556594849, + "learning_rate": 1.1745543302684292e-05, + "loss": 0.3339, + "step": 48931 + }, + { + "epoch": 3.3954618000138783, + "grad_norm": 1.0738930702209473, + "learning_rate": 1.1742901407401663e-05, + "loss": 0.2186, + "step": 48932 + }, + { + "epoch": 3.395531191450975, + "grad_norm": 0.9363900423049927, + "learning_rate": 1.1740259790736452e-05, + "loss": 0.2612, + "step": 48933 + }, + { + "epoch": 3.3956005828880715, + "grad_norm": 0.736076831817627, + "learning_rate": 1.1737618452697008e-05, + "loss": 0.3013, + "step": 48934 + }, + { + "epoch": 3.3956699743251684, + "grad_norm": 0.6716084480285645, + "learning_rate": 1.1734977393291679e-05, + "loss": 0.1247, + "step": 48935 + }, + { + "epoch": 3.3957393657622648, + "grad_norm": 0.5766177177429199, + "learning_rate": 1.1732336612528772e-05, + "loss": 0.1854, + "step": 48936 + }, + { + "epoch": 3.3958087571993616, + "grad_norm": 0.6927422881126404, + "learning_rate": 1.1729696110416633e-05, + "loss": 0.2279, + "step": 48937 + }, + { + "epoch": 3.3958781486364584, + "grad_norm": 1.2301828861236572, + "learning_rate": 1.17270558869636e-05, + "loss": 0.2893, + "step": 48938 + }, + { + "epoch": 3.395947540073555, + "grad_norm": 0.8360544443130493, + "learning_rate": 1.1724415942178036e-05, + "loss": 0.2483, + "step": 48939 + }, + { + "epoch": 3.3960169315106516, + "grad_norm": 0.8200786709785461, + "learning_rate": 1.172177627606822e-05, + "loss": 0.1632, + "step": 48940 + }, + { + "epoch": 3.396086322947748, + "grad_norm": 1.2143832445144653, + "learning_rate": 1.1719136888642524e-05, + "loss": 0.3729, + "step": 48941 + }, + { + "epoch": 3.396155714384845, + "grad_norm": 1.30375337600708, + "learning_rate": 1.1716497779909263e-05, + "loss": 0.2326, + "step": 48942 + }, + { + "epoch": 3.3962251058219417, + "grad_norm": 1.0910545587539673, + "learning_rate": 1.17138589498768e-05, + "loss": 0.282, + "step": 48943 + }, + { + "epoch": 3.396294497259038, + "grad_norm": 1.09286367893219, + "learning_rate": 1.1711220398553414e-05, + "loss": 0.2653, + "step": 48944 + }, + { + "epoch": 3.396363888696135, + "grad_norm": 0.6681278347969055, + "learning_rate": 1.1708582125947464e-05, + "loss": 0.1543, + "step": 48945 + }, + { + "epoch": 3.3964332801332313, + "grad_norm": 0.9171357750892639, + "learning_rate": 1.1705944132067281e-05, + "loss": 0.2065, + "step": 48946 + }, + { + "epoch": 3.396502671570328, + "grad_norm": 0.7284382581710815, + "learning_rate": 1.1703306416921189e-05, + "loss": 0.2159, + "step": 48947 + }, + { + "epoch": 3.396572063007425, + "grad_norm": 1.4186034202575684, + "learning_rate": 1.1700668980517515e-05, + "loss": 0.2717, + "step": 48948 + }, + { + "epoch": 3.3966414544445214, + "grad_norm": 1.7357786893844604, + "learning_rate": 1.1698031822864597e-05, + "loss": 0.3751, + "step": 48949 + }, + { + "epoch": 3.396710845881618, + "grad_norm": 0.9733392596244812, + "learning_rate": 1.1695394943970728e-05, + "loss": 0.2235, + "step": 48950 + }, + { + "epoch": 3.396780237318715, + "grad_norm": 0.9654077291488647, + "learning_rate": 1.1692758343844258e-05, + "loss": 0.249, + "step": 48951 + }, + { + "epoch": 3.3968496287558114, + "grad_norm": 0.8442959785461426, + "learning_rate": 1.169012202249351e-05, + "loss": 0.1598, + "step": 48952 + }, + { + "epoch": 3.3969190201929083, + "grad_norm": 0.722722053527832, + "learning_rate": 1.1687485979926793e-05, + "loss": 0.1926, + "step": 48953 + }, + { + "epoch": 3.396988411630005, + "grad_norm": 0.7834563851356506, + "learning_rate": 1.1684850216152432e-05, + "loss": 0.2456, + "step": 48954 + }, + { + "epoch": 3.3970578030671015, + "grad_norm": 0.8191862106323242, + "learning_rate": 1.1682214731178776e-05, + "loss": 0.3368, + "step": 48955 + }, + { + "epoch": 3.3971271945041983, + "grad_norm": 0.9473147988319397, + "learning_rate": 1.1679579525014106e-05, + "loss": 0.2657, + "step": 48956 + }, + { + "epoch": 3.3971965859412947, + "grad_norm": 0.5986428260803223, + "learning_rate": 1.167694459766675e-05, + "loss": 0.1651, + "step": 48957 + }, + { + "epoch": 3.3972659773783915, + "grad_norm": 0.8515954613685608, + "learning_rate": 1.1674309949145035e-05, + "loss": 0.3135, + "step": 48958 + }, + { + "epoch": 3.3973353688154884, + "grad_norm": 1.0039416551589966, + "learning_rate": 1.1671675579457276e-05, + "loss": 0.3438, + "step": 48959 + }, + { + "epoch": 3.3974047602525848, + "grad_norm": 0.9362875819206238, + "learning_rate": 1.1669041488611787e-05, + "loss": 0.1741, + "step": 48960 + }, + { + "epoch": 3.3974741516896816, + "grad_norm": 1.0079702138900757, + "learning_rate": 1.1666407676616898e-05, + "loss": 0.2976, + "step": 48961 + }, + { + "epoch": 3.397543543126778, + "grad_norm": 0.9639332890510559, + "learning_rate": 1.166377414348092e-05, + "loss": 0.2896, + "step": 48962 + }, + { + "epoch": 3.397612934563875, + "grad_norm": 0.9396595358848572, + "learning_rate": 1.166114088921214e-05, + "loss": 0.2601, + "step": 48963 + }, + { + "epoch": 3.3976823260009716, + "grad_norm": 2.3782870769500732, + "learning_rate": 1.1658507913818884e-05, + "loss": 0.2591, + "step": 48964 + }, + { + "epoch": 3.397751717438068, + "grad_norm": 1.029140591621399, + "learning_rate": 1.1655875217309475e-05, + "loss": 0.2693, + "step": 48965 + }, + { + "epoch": 3.397821108875165, + "grad_norm": 0.655452311038971, + "learning_rate": 1.165324279969222e-05, + "loss": 0.1482, + "step": 48966 + }, + { + "epoch": 3.3978905003122613, + "grad_norm": 0.7770769596099854, + "learning_rate": 1.1650610660975425e-05, + "loss": 0.2039, + "step": 48967 + }, + { + "epoch": 3.397959891749358, + "grad_norm": 0.8239768743515015, + "learning_rate": 1.1647978801167403e-05, + "loss": 0.2762, + "step": 48968 + }, + { + "epoch": 3.398029283186455, + "grad_norm": 1.8374073505401611, + "learning_rate": 1.164534722027647e-05, + "loss": 0.3559, + "step": 48969 + }, + { + "epoch": 3.3980986746235513, + "grad_norm": 1.2117321491241455, + "learning_rate": 1.164271591831092e-05, + "loss": 0.2782, + "step": 48970 + }, + { + "epoch": 3.398168066060648, + "grad_norm": 1.1367393732070923, + "learning_rate": 1.1640084895279058e-05, + "loss": 0.3962, + "step": 48971 + }, + { + "epoch": 3.3982374574977445, + "grad_norm": 0.5408863425254822, + "learning_rate": 1.1637454151189197e-05, + "loss": 0.1886, + "step": 48972 + }, + { + "epoch": 3.3983068489348414, + "grad_norm": 0.9020888805389404, + "learning_rate": 1.1634823686049645e-05, + "loss": 0.2401, + "step": 48973 + }, + { + "epoch": 3.398376240371938, + "grad_norm": 1.0139440298080444, + "learning_rate": 1.1632193499868716e-05, + "loss": 0.2819, + "step": 48974 + }, + { + "epoch": 3.3984456318090346, + "grad_norm": 0.9499230980873108, + "learning_rate": 1.1629563592654679e-05, + "loss": 0.2742, + "step": 48975 + }, + { + "epoch": 3.3985150232461314, + "grad_norm": 0.764516294002533, + "learning_rate": 1.1626933964415865e-05, + "loss": 0.2174, + "step": 48976 + }, + { + "epoch": 3.3985844146832282, + "grad_norm": 1.1825692653656006, + "learning_rate": 1.1624304615160564e-05, + "loss": 0.2876, + "step": 48977 + }, + { + "epoch": 3.3986538061203246, + "grad_norm": 0.99368816614151, + "learning_rate": 1.1621675544897092e-05, + "loss": 0.2289, + "step": 48978 + }, + { + "epoch": 3.3987231975574215, + "grad_norm": 0.8352723121643066, + "learning_rate": 1.1619046753633722e-05, + "loss": 0.2995, + "step": 48979 + }, + { + "epoch": 3.3987925889945183, + "grad_norm": 0.8167909979820251, + "learning_rate": 1.1616418241378768e-05, + "loss": 0.2235, + "step": 48980 + }, + { + "epoch": 3.3988619804316147, + "grad_norm": 0.8167203068733215, + "learning_rate": 1.1613790008140535e-05, + "loss": 0.2226, + "step": 48981 + }, + { + "epoch": 3.3989313718687115, + "grad_norm": 0.6841103434562683, + "learning_rate": 1.1611162053927304e-05, + "loss": 0.1778, + "step": 48982 + }, + { + "epoch": 3.399000763305808, + "grad_norm": 1.120656132698059, + "learning_rate": 1.1608534378747393e-05, + "loss": 0.4813, + "step": 48983 + }, + { + "epoch": 3.3990701547429047, + "grad_norm": 1.0026575326919556, + "learning_rate": 1.1605906982609071e-05, + "loss": 0.2183, + "step": 48984 + }, + { + "epoch": 3.3991395461800016, + "grad_norm": 0.9681530594825745, + "learning_rate": 1.1603279865520645e-05, + "loss": 0.3065, + "step": 48985 + }, + { + "epoch": 3.399208937617098, + "grad_norm": 0.8822807669639587, + "learning_rate": 1.1600653027490404e-05, + "loss": 0.282, + "step": 48986 + }, + { + "epoch": 3.399278329054195, + "grad_norm": 1.2129625082015991, + "learning_rate": 1.1598026468526657e-05, + "loss": 0.2839, + "step": 48987 + }, + { + "epoch": 3.399347720491291, + "grad_norm": 0.8136278986930847, + "learning_rate": 1.1595400188637672e-05, + "loss": 0.236, + "step": 48988 + }, + { + "epoch": 3.399417111928388, + "grad_norm": 0.8577807545661926, + "learning_rate": 1.1592774187831756e-05, + "loss": 0.2924, + "step": 48989 + }, + { + "epoch": 3.399486503365485, + "grad_norm": 0.788817286491394, + "learning_rate": 1.1590148466117212e-05, + "loss": 0.2263, + "step": 48990 + }, + { + "epoch": 3.3995558948025812, + "grad_norm": 0.8570560812950134, + "learning_rate": 1.15875230235023e-05, + "loss": 0.262, + "step": 48991 + }, + { + "epoch": 3.399625286239678, + "grad_norm": 0.7284532189369202, + "learning_rate": 1.1584897859995314e-05, + "loss": 0.1611, + "step": 48992 + }, + { + "epoch": 3.3996946776767745, + "grad_norm": 1.026416540145874, + "learning_rate": 1.1582272975604547e-05, + "loss": 0.302, + "step": 48993 + }, + { + "epoch": 3.3997640691138713, + "grad_norm": 0.9001761078834534, + "learning_rate": 1.1579648370338291e-05, + "loss": 0.1638, + "step": 48994 + }, + { + "epoch": 3.399833460550968, + "grad_norm": 0.586582362651825, + "learning_rate": 1.157702404420482e-05, + "loss": 0.2169, + "step": 48995 + }, + { + "epoch": 3.3999028519880645, + "grad_norm": 0.8493924736976624, + "learning_rate": 1.1574399997212427e-05, + "loss": 0.2129, + "step": 48996 + }, + { + "epoch": 3.3999722434251614, + "grad_norm": 0.9108816981315613, + "learning_rate": 1.1571776229369413e-05, + "loss": 0.2505, + "step": 48997 + }, + { + "epoch": 3.400041634862258, + "grad_norm": 0.9904941320419312, + "learning_rate": 1.156915274068402e-05, + "loss": 0.2772, + "step": 48998 + }, + { + "epoch": 3.4001110262993546, + "grad_norm": 0.7343103289604187, + "learning_rate": 1.156652953116456e-05, + "loss": 0.1975, + "step": 48999 + }, + { + "epoch": 3.4001804177364514, + "grad_norm": 0.771575391292572, + "learning_rate": 1.1563906600819297e-05, + "loss": 0.1875, + "step": 49000 + }, + { + "epoch": 3.400249809173548, + "grad_norm": 1.0407495498657227, + "learning_rate": 1.1561283949656533e-05, + "loss": 0.3146, + "step": 49001 + }, + { + "epoch": 3.4003192006106446, + "grad_norm": 1.0693608522415161, + "learning_rate": 1.1558661577684527e-05, + "loss": 0.2561, + "step": 49002 + }, + { + "epoch": 3.4003885920477415, + "grad_norm": 1.1099141836166382, + "learning_rate": 1.1556039484911562e-05, + "loss": 0.2081, + "step": 49003 + }, + { + "epoch": 3.400457983484838, + "grad_norm": 0.9514187574386597, + "learning_rate": 1.1553417671345946e-05, + "loss": 0.2165, + "step": 49004 + }, + { + "epoch": 3.4005273749219347, + "grad_norm": 1.2017712593078613, + "learning_rate": 1.15507961369959e-05, + "loss": 0.2538, + "step": 49005 + }, + { + "epoch": 3.4005967663590315, + "grad_norm": 0.8264671564102173, + "learning_rate": 1.1548174881869733e-05, + "loss": 0.2329, + "step": 49006 + }, + { + "epoch": 3.400666157796128, + "grad_norm": 1.0866224765777588, + "learning_rate": 1.1545553905975715e-05, + "loss": 0.3102, + "step": 49007 + }, + { + "epoch": 3.4007355492332247, + "grad_norm": 0.7780188322067261, + "learning_rate": 1.1542933209322115e-05, + "loss": 0.33, + "step": 49008 + }, + { + "epoch": 3.400804940670321, + "grad_norm": 0.997076153755188, + "learning_rate": 1.154031279191724e-05, + "loss": 0.2817, + "step": 49009 + }, + { + "epoch": 3.400874332107418, + "grad_norm": 0.6146934032440186, + "learning_rate": 1.1537692653769317e-05, + "loss": 0.2303, + "step": 49010 + }, + { + "epoch": 3.400943723544515, + "grad_norm": 0.9898288249969482, + "learning_rate": 1.1535072794886625e-05, + "loss": 0.2727, + "step": 49011 + }, + { + "epoch": 3.401013114981611, + "grad_norm": 1.1468368768692017, + "learning_rate": 1.153245321527745e-05, + "loss": 0.2558, + "step": 49012 + }, + { + "epoch": 3.401082506418708, + "grad_norm": 1.37260103225708, + "learning_rate": 1.1529833914950083e-05, + "loss": 0.3276, + "step": 49013 + }, + { + "epoch": 3.4011518978558044, + "grad_norm": 1.1659806966781616, + "learning_rate": 1.152721489391273e-05, + "loss": 0.2268, + "step": 49014 + }, + { + "epoch": 3.4012212892929012, + "grad_norm": 1.1853326559066772, + "learning_rate": 1.1524596152173706e-05, + "loss": 0.3074, + "step": 49015 + }, + { + "epoch": 3.401290680729998, + "grad_norm": 1.2977631092071533, + "learning_rate": 1.152197768974127e-05, + "loss": 0.3469, + "step": 49016 + }, + { + "epoch": 3.4013600721670945, + "grad_norm": 1.074392557144165, + "learning_rate": 1.1519359506623673e-05, + "loss": 0.2407, + "step": 49017 + }, + { + "epoch": 3.4014294636041913, + "grad_norm": 0.9035768508911133, + "learning_rate": 1.1516741602829218e-05, + "loss": 0.3631, + "step": 49018 + }, + { + "epoch": 3.4014988550412877, + "grad_norm": 1.0726754665374756, + "learning_rate": 1.151412397836613e-05, + "loss": 0.3429, + "step": 49019 + }, + { + "epoch": 3.4015682464783845, + "grad_norm": 0.8133885860443115, + "learning_rate": 1.1511506633242675e-05, + "loss": 0.1462, + "step": 49020 + }, + { + "epoch": 3.4016376379154813, + "grad_norm": 1.393661379814148, + "learning_rate": 1.150888956746714e-05, + "loss": 0.1719, + "step": 49021 + }, + { + "epoch": 3.4017070293525777, + "grad_norm": 1.7481768131256104, + "learning_rate": 1.1506272781047766e-05, + "loss": 0.222, + "step": 49022 + }, + { + "epoch": 3.4017764207896746, + "grad_norm": 1.2658215761184692, + "learning_rate": 1.1503656273992825e-05, + "loss": 0.3023, + "step": 49023 + }, + { + "epoch": 3.4018458122267714, + "grad_norm": 0.7307139039039612, + "learning_rate": 1.1501040046310562e-05, + "loss": 0.2418, + "step": 49024 + }, + { + "epoch": 3.401915203663868, + "grad_norm": 0.6678038239479065, + "learning_rate": 1.1498424098009287e-05, + "loss": 0.2049, + "step": 49025 + }, + { + "epoch": 3.4019845951009646, + "grad_norm": 1.217915415763855, + "learning_rate": 1.1495808429097189e-05, + "loss": 0.2148, + "step": 49026 + }, + { + "epoch": 3.402053986538061, + "grad_norm": 1.1622873544692993, + "learning_rate": 1.149319303958255e-05, + "loss": 0.2732, + "step": 49027 + }, + { + "epoch": 3.402123377975158, + "grad_norm": 0.532268762588501, + "learning_rate": 1.1490577929473634e-05, + "loss": 0.0984, + "step": 49028 + }, + { + "epoch": 3.4021927694122547, + "grad_norm": 0.8732941746711731, + "learning_rate": 1.1487963098778698e-05, + "loss": 0.1562, + "step": 49029 + }, + { + "epoch": 3.402262160849351, + "grad_norm": 0.9373955130577087, + "learning_rate": 1.1485348547505991e-05, + "loss": 0.225, + "step": 49030 + }, + { + "epoch": 3.402331552286448, + "grad_norm": 0.7073463201522827, + "learning_rate": 1.1482734275663799e-05, + "loss": 0.1868, + "step": 49031 + }, + { + "epoch": 3.4024009437235447, + "grad_norm": 1.3949581384658813, + "learning_rate": 1.148012028326031e-05, + "loss": 0.2337, + "step": 49032 + }, + { + "epoch": 3.402470335160641, + "grad_norm": 0.871695876121521, + "learning_rate": 1.147750657030382e-05, + "loss": 0.3574, + "step": 49033 + }, + { + "epoch": 3.402539726597738, + "grad_norm": 1.103082299232483, + "learning_rate": 1.1474893136802567e-05, + "loss": 0.2519, + "step": 49034 + }, + { + "epoch": 3.4026091180348343, + "grad_norm": 0.8745966553688049, + "learning_rate": 1.1472279982764811e-05, + "loss": 0.1608, + "step": 49035 + }, + { + "epoch": 3.402678509471931, + "grad_norm": 0.9989069104194641, + "learning_rate": 1.1469667108198789e-05, + "loss": 0.3857, + "step": 49036 + }, + { + "epoch": 3.402747900909028, + "grad_norm": 0.7583453059196472, + "learning_rate": 1.1467054513112762e-05, + "loss": 0.2643, + "step": 49037 + }, + { + "epoch": 3.4028172923461244, + "grad_norm": 0.7215188145637512, + "learning_rate": 1.1464442197514991e-05, + "loss": 0.2448, + "step": 49038 + }, + { + "epoch": 3.4028866837832212, + "grad_norm": 0.950966477394104, + "learning_rate": 1.146183016141369e-05, + "loss": 0.2694, + "step": 49039 + }, + { + "epoch": 3.4029560752203176, + "grad_norm": 0.8697654604911804, + "learning_rate": 1.1459218404817106e-05, + "loss": 0.2735, + "step": 49040 + }, + { + "epoch": 3.4030254666574145, + "grad_norm": 0.9936951994895935, + "learning_rate": 1.1456606927733504e-05, + "loss": 0.3142, + "step": 49041 + }, + { + "epoch": 3.4030948580945113, + "grad_norm": 0.5911709666252136, + "learning_rate": 1.1453995730171119e-05, + "loss": 0.2043, + "step": 49042 + }, + { + "epoch": 3.4031642495316077, + "grad_norm": 0.8471410870552063, + "learning_rate": 1.14513848121382e-05, + "loss": 0.2386, + "step": 49043 + }, + { + "epoch": 3.4032336409687045, + "grad_norm": 0.7525888085365295, + "learning_rate": 1.1448774173643006e-05, + "loss": 0.2603, + "step": 49044 + }, + { + "epoch": 3.403303032405801, + "grad_norm": 0.6765055060386658, + "learning_rate": 1.1446163814693734e-05, + "loss": 0.1445, + "step": 49045 + }, + { + "epoch": 3.4033724238428977, + "grad_norm": 0.9386348724365234, + "learning_rate": 1.144355373529865e-05, + "loss": 0.2776, + "step": 49046 + }, + { + "epoch": 3.4034418152799946, + "grad_norm": 0.8033105731010437, + "learning_rate": 1.1440943935465986e-05, + "loss": 0.2535, + "step": 49047 + }, + { + "epoch": 3.403511206717091, + "grad_norm": 1.013991117477417, + "learning_rate": 1.143833441520401e-05, + "loss": 0.3484, + "step": 49048 + }, + { + "epoch": 3.403580598154188, + "grad_norm": 0.8324164748191833, + "learning_rate": 1.1435725174520928e-05, + "loss": 0.1839, + "step": 49049 + }, + { + "epoch": 3.4036499895912846, + "grad_norm": 1.1034715175628662, + "learning_rate": 1.1433116213424977e-05, + "loss": 0.2543, + "step": 49050 + }, + { + "epoch": 3.403719381028381, + "grad_norm": 2.419994592666626, + "learning_rate": 1.1430507531924406e-05, + "loss": 0.2432, + "step": 49051 + }, + { + "epoch": 3.403788772465478, + "grad_norm": 1.2337998151779175, + "learning_rate": 1.1427899130027464e-05, + "loss": 0.4061, + "step": 49052 + }, + { + "epoch": 3.4038581639025742, + "grad_norm": 0.6985242366790771, + "learning_rate": 1.1425291007742345e-05, + "loss": 0.2128, + "step": 49053 + }, + { + "epoch": 3.403927555339671, + "grad_norm": 1.6519776582717896, + "learning_rate": 1.142268316507732e-05, + "loss": 0.2265, + "step": 49054 + }, + { + "epoch": 3.403996946776768, + "grad_norm": 0.6799246668815613, + "learning_rate": 1.1420075602040593e-05, + "loss": 0.2778, + "step": 49055 + }, + { + "epoch": 3.4040663382138643, + "grad_norm": 0.6517114639282227, + "learning_rate": 1.1417468318640411e-05, + "loss": 0.1906, + "step": 49056 + }, + { + "epoch": 3.404135729650961, + "grad_norm": 0.6897598505020142, + "learning_rate": 1.1414861314885006e-05, + "loss": 0.2437, + "step": 49057 + }, + { + "epoch": 3.404205121088058, + "grad_norm": 1.2915891408920288, + "learning_rate": 1.1412254590782612e-05, + "loss": 0.2981, + "step": 49058 + }, + { + "epoch": 3.4042745125251543, + "grad_norm": 0.9342392086982727, + "learning_rate": 1.1409648146341467e-05, + "loss": 0.2105, + "step": 49059 + }, + { + "epoch": 3.404343903962251, + "grad_norm": 1.0670101642608643, + "learning_rate": 1.1407041981569767e-05, + "loss": 0.3113, + "step": 49060 + }, + { + "epoch": 3.4044132953993476, + "grad_norm": 0.7654717564582825, + "learning_rate": 1.1404436096475757e-05, + "loss": 0.276, + "step": 49061 + }, + { + "epoch": 3.4044826868364444, + "grad_norm": 0.6939628720283508, + "learning_rate": 1.1401830491067655e-05, + "loss": 0.1639, + "step": 49062 + }, + { + "epoch": 3.4045520782735412, + "grad_norm": 0.9067319631576538, + "learning_rate": 1.13992251653537e-05, + "loss": 0.2791, + "step": 49063 + }, + { + "epoch": 3.4046214697106376, + "grad_norm": 0.8810742497444153, + "learning_rate": 1.1396620119342116e-05, + "loss": 0.2317, + "step": 49064 + }, + { + "epoch": 3.4046908611477344, + "grad_norm": 0.5889590382575989, + "learning_rate": 1.139401535304112e-05, + "loss": 0.1685, + "step": 49065 + }, + { + "epoch": 3.404760252584831, + "grad_norm": 0.7705336213111877, + "learning_rate": 1.1391410866458951e-05, + "loss": 0.2316, + "step": 49066 + }, + { + "epoch": 3.4048296440219277, + "grad_norm": 1.2617219686508179, + "learning_rate": 1.13888066596038e-05, + "loss": 0.4532, + "step": 49067 + }, + { + "epoch": 3.4048990354590245, + "grad_norm": 0.8740434646606445, + "learning_rate": 1.138620273248392e-05, + "loss": 0.2029, + "step": 49068 + }, + { + "epoch": 3.404968426896121, + "grad_norm": 1.0966393947601318, + "learning_rate": 1.1383599085107498e-05, + "loss": 0.3145, + "step": 49069 + }, + { + "epoch": 3.4050378183332177, + "grad_norm": 1.0613985061645508, + "learning_rate": 1.1380995717482778e-05, + "loss": 0.2533, + "step": 49070 + }, + { + "epoch": 3.405107209770314, + "grad_norm": 0.9168829917907715, + "learning_rate": 1.1378392629617973e-05, + "loss": 0.3062, + "step": 49071 + }, + { + "epoch": 3.405176601207411, + "grad_norm": 1.1911505460739136, + "learning_rate": 1.137578982152131e-05, + "loss": 0.3231, + "step": 49072 + }, + { + "epoch": 3.405245992644508, + "grad_norm": 0.9041529297828674, + "learning_rate": 1.1373187293201004e-05, + "loss": 0.1816, + "step": 49073 + }, + { + "epoch": 3.405315384081604, + "grad_norm": 0.8883560299873352, + "learning_rate": 1.137058504466525e-05, + "loss": 0.28, + "step": 49074 + }, + { + "epoch": 3.405384775518701, + "grad_norm": 0.7225112915039062, + "learning_rate": 1.1367983075922272e-05, + "loss": 0.2711, + "step": 49075 + }, + { + "epoch": 3.405454166955798, + "grad_norm": 0.6120979189872742, + "learning_rate": 1.136538138698029e-05, + "loss": 0.1476, + "step": 49076 + }, + { + "epoch": 3.4055235583928942, + "grad_norm": 0.8377140760421753, + "learning_rate": 1.1362779977847514e-05, + "loss": 0.359, + "step": 49077 + }, + { + "epoch": 3.405592949829991, + "grad_norm": 1.0368379354476929, + "learning_rate": 1.1360178848532155e-05, + "loss": 0.2689, + "step": 49078 + }, + { + "epoch": 3.405662341267088, + "grad_norm": 0.7347983121871948, + "learning_rate": 1.1357577999042457e-05, + "loss": 0.2364, + "step": 49079 + }, + { + "epoch": 3.4057317327041843, + "grad_norm": 1.0101031064987183, + "learning_rate": 1.135497742938657e-05, + "loss": 0.2256, + "step": 49080 + }, + { + "epoch": 3.405801124141281, + "grad_norm": 1.2432292699813843, + "learning_rate": 1.1352377139572734e-05, + "loss": 0.2664, + "step": 49081 + }, + { + "epoch": 3.4058705155783775, + "grad_norm": 0.8267167210578918, + "learning_rate": 1.1349777129609185e-05, + "loss": 0.2879, + "step": 49082 + }, + { + "epoch": 3.4059399070154743, + "grad_norm": 0.6429236531257629, + "learning_rate": 1.1347177399504083e-05, + "loss": 0.2073, + "step": 49083 + }, + { + "epoch": 3.406009298452571, + "grad_norm": 0.9678862690925598, + "learning_rate": 1.1344577949265655e-05, + "loss": 0.2543, + "step": 49084 + }, + { + "epoch": 3.4060786898896676, + "grad_norm": 0.9596431255340576, + "learning_rate": 1.1341978778902119e-05, + "loss": 0.2549, + "step": 49085 + }, + { + "epoch": 3.4061480813267644, + "grad_norm": 0.7052220106124878, + "learning_rate": 1.1339379888421653e-05, + "loss": 0.1873, + "step": 49086 + }, + { + "epoch": 3.4062174727638608, + "grad_norm": 0.962193489074707, + "learning_rate": 1.1336781277832508e-05, + "loss": 0.2594, + "step": 49087 + }, + { + "epoch": 3.4062868642009576, + "grad_norm": 1.0569649934768677, + "learning_rate": 1.1334182947142836e-05, + "loss": 0.2644, + "step": 49088 + }, + { + "epoch": 3.4063562556380544, + "grad_norm": 1.1030510663986206, + "learning_rate": 1.133158489636086e-05, + "loss": 0.3386, + "step": 49089 + }, + { + "epoch": 3.406425647075151, + "grad_norm": 1.1121267080307007, + "learning_rate": 1.1328987125494783e-05, + "loss": 0.3438, + "step": 49090 + }, + { + "epoch": 3.4064950385122477, + "grad_norm": 0.9241346120834351, + "learning_rate": 1.1326389634552803e-05, + "loss": 0.1647, + "step": 49091 + }, + { + "epoch": 3.406564429949344, + "grad_norm": 0.742153525352478, + "learning_rate": 1.1323792423543133e-05, + "loss": 0.2454, + "step": 49092 + }, + { + "epoch": 3.406633821386441, + "grad_norm": 1.5968300104141235, + "learning_rate": 1.1321195492473957e-05, + "loss": 0.2043, + "step": 49093 + }, + { + "epoch": 3.4067032128235377, + "grad_norm": 0.9458156824111938, + "learning_rate": 1.131859884135349e-05, + "loss": 0.2362, + "step": 49094 + }, + { + "epoch": 3.406772604260634, + "grad_norm": 0.7894832491874695, + "learning_rate": 1.1316002470189902e-05, + "loss": 0.2574, + "step": 49095 + }, + { + "epoch": 3.406841995697731, + "grad_norm": 1.0075018405914307, + "learning_rate": 1.131340637899141e-05, + "loss": 0.256, + "step": 49096 + }, + { + "epoch": 3.4069113871348273, + "grad_norm": 0.7544105648994446, + "learning_rate": 1.1310810567766206e-05, + "loss": 0.1809, + "step": 49097 + }, + { + "epoch": 3.406980778571924, + "grad_norm": 1.1444333791732788, + "learning_rate": 1.1308215036522474e-05, + "loss": 0.287, + "step": 49098 + }, + { + "epoch": 3.407050170009021, + "grad_norm": 0.7289932370185852, + "learning_rate": 1.1305619785268417e-05, + "loss": 0.1436, + "step": 49099 + }, + { + "epoch": 3.4071195614461174, + "grad_norm": 0.8126696944236755, + "learning_rate": 1.1303024814012243e-05, + "loss": 0.222, + "step": 49100 + }, + { + "epoch": 3.407188952883214, + "grad_norm": 0.7533219456672668, + "learning_rate": 1.130043012276213e-05, + "loss": 0.2651, + "step": 49101 + }, + { + "epoch": 3.407258344320311, + "grad_norm": 1.0720288753509521, + "learning_rate": 1.129783571152626e-05, + "loss": 0.3395, + "step": 49102 + }, + { + "epoch": 3.4073277357574074, + "grad_norm": 1.7822142839431763, + "learning_rate": 1.1295241580312832e-05, + "loss": 0.1931, + "step": 49103 + }, + { + "epoch": 3.4073971271945043, + "grad_norm": 0.7917740345001221, + "learning_rate": 1.1292647729130023e-05, + "loss": 0.1714, + "step": 49104 + }, + { + "epoch": 3.407466518631601, + "grad_norm": 1.7448906898498535, + "learning_rate": 1.129005415798604e-05, + "loss": 0.3764, + "step": 49105 + }, + { + "epoch": 3.4075359100686975, + "grad_norm": 1.1750593185424805, + "learning_rate": 1.1287460866889065e-05, + "loss": 0.283, + "step": 49106 + }, + { + "epoch": 3.4076053015057943, + "grad_norm": 1.1310577392578125, + "learning_rate": 1.1284867855847303e-05, + "loss": 0.2783, + "step": 49107 + }, + { + "epoch": 3.4076746929428907, + "grad_norm": 0.6802605986595154, + "learning_rate": 1.1282275124868901e-05, + "loss": 0.2019, + "step": 49108 + }, + { + "epoch": 3.4077440843799875, + "grad_norm": 0.6831530928611755, + "learning_rate": 1.1279682673962067e-05, + "loss": 0.2157, + "step": 49109 + }, + { + "epoch": 3.4078134758170844, + "grad_norm": 1.0257742404937744, + "learning_rate": 1.1277090503134979e-05, + "loss": 0.2661, + "step": 49110 + }, + { + "epoch": 3.4078828672541808, + "grad_norm": 0.8242462873458862, + "learning_rate": 1.1274498612395812e-05, + "loss": 0.1925, + "step": 49111 + }, + { + "epoch": 3.4079522586912776, + "grad_norm": 0.7490960359573364, + "learning_rate": 1.1271907001752769e-05, + "loss": 0.2606, + "step": 49112 + }, + { + "epoch": 3.408021650128374, + "grad_norm": 1.2811349630355835, + "learning_rate": 1.126931567121402e-05, + "loss": 0.289, + "step": 49113 + }, + { + "epoch": 3.408091041565471, + "grad_norm": 0.7627873420715332, + "learning_rate": 1.1266724620787771e-05, + "loss": 0.1815, + "step": 49114 + }, + { + "epoch": 3.4081604330025677, + "grad_norm": 0.9989739656448364, + "learning_rate": 1.126413385048215e-05, + "loss": 0.2776, + "step": 49115 + }, + { + "epoch": 3.408229824439664, + "grad_norm": 0.9870420098304749, + "learning_rate": 1.126154336030536e-05, + "loss": 0.2607, + "step": 49116 + }, + { + "epoch": 3.408299215876761, + "grad_norm": 0.7578638792037964, + "learning_rate": 1.1258953150265617e-05, + "loss": 0.236, + "step": 49117 + }, + { + "epoch": 3.4083686073138573, + "grad_norm": 0.7882484793663025, + "learning_rate": 1.1256363220371025e-05, + "loss": 0.1944, + "step": 49118 + }, + { + "epoch": 3.408437998750954, + "grad_norm": 0.583916425704956, + "learning_rate": 1.1253773570629812e-05, + "loss": 0.2121, + "step": 49119 + }, + { + "epoch": 3.408507390188051, + "grad_norm": 1.0503658056259155, + "learning_rate": 1.1251184201050136e-05, + "loss": 0.2653, + "step": 49120 + }, + { + "epoch": 3.4085767816251473, + "grad_norm": 1.1304121017456055, + "learning_rate": 1.1248595111640192e-05, + "loss": 0.2558, + "step": 49121 + }, + { + "epoch": 3.408646173062244, + "grad_norm": 0.8131890296936035, + "learning_rate": 1.1246006302408118e-05, + "loss": 0.1855, + "step": 49122 + }, + { + "epoch": 3.4087155644993405, + "grad_norm": 1.6091467142105103, + "learning_rate": 1.1243417773362097e-05, + "loss": 0.328, + "step": 49123 + }, + { + "epoch": 3.4087849559364374, + "grad_norm": 0.9554010033607483, + "learning_rate": 1.1240829524510321e-05, + "loss": 0.3046, + "step": 49124 + }, + { + "epoch": 3.408854347373534, + "grad_norm": 0.6204820275306702, + "learning_rate": 1.123824155586094e-05, + "loss": 0.1509, + "step": 49125 + }, + { + "epoch": 3.4089237388106306, + "grad_norm": 0.9555267095565796, + "learning_rate": 1.1235653867422136e-05, + "loss": 0.2888, + "step": 49126 + }, + { + "epoch": 3.4089931302477274, + "grad_norm": 0.8338959813117981, + "learning_rate": 1.1233066459202068e-05, + "loss": 0.2813, + "step": 49127 + }, + { + "epoch": 3.4090625216848243, + "grad_norm": 0.6679349541664124, + "learning_rate": 1.1230479331208944e-05, + "loss": 0.2538, + "step": 49128 + }, + { + "epoch": 3.4091319131219207, + "grad_norm": 0.6631157994270325, + "learning_rate": 1.1227892483450864e-05, + "loss": 0.1882, + "step": 49129 + }, + { + "epoch": 3.4092013045590175, + "grad_norm": 0.9361758828163147, + "learning_rate": 1.1225305915936034e-05, + "loss": 0.2094, + "step": 49130 + }, + { + "epoch": 3.4092706959961143, + "grad_norm": 1.0253034830093384, + "learning_rate": 1.1222719628672607e-05, + "loss": 0.2725, + "step": 49131 + }, + { + "epoch": 3.4093400874332107, + "grad_norm": 0.7623194456100464, + "learning_rate": 1.122013362166876e-05, + "loss": 0.2635, + "step": 49132 + }, + { + "epoch": 3.4094094788703075, + "grad_norm": 0.938207745552063, + "learning_rate": 1.1217547894932656e-05, + "loss": 0.2399, + "step": 49133 + }, + { + "epoch": 3.409478870307404, + "grad_norm": 0.8679356575012207, + "learning_rate": 1.1214962448472444e-05, + "loss": 0.2588, + "step": 49134 + }, + { + "epoch": 3.4095482617445008, + "grad_norm": 0.8160179257392883, + "learning_rate": 1.1212377282296315e-05, + "loss": 0.1924, + "step": 49135 + }, + { + "epoch": 3.4096176531815976, + "grad_norm": 0.7845442891120911, + "learning_rate": 1.1209792396412399e-05, + "loss": 0.2891, + "step": 49136 + }, + { + "epoch": 3.409687044618694, + "grad_norm": 0.8239366412162781, + "learning_rate": 1.1207207790828854e-05, + "loss": 0.3845, + "step": 49137 + }, + { + "epoch": 3.409756436055791, + "grad_norm": 1.111698865890503, + "learning_rate": 1.1204623465553854e-05, + "loss": 0.2945, + "step": 49138 + }, + { + "epoch": 3.409825827492887, + "grad_norm": 0.9271157383918762, + "learning_rate": 1.1202039420595567e-05, + "loss": 0.1434, + "step": 49139 + }, + { + "epoch": 3.409895218929984, + "grad_norm": 0.7924012541770935, + "learning_rate": 1.1199455655962132e-05, + "loss": 0.2037, + "step": 49140 + }, + { + "epoch": 3.409964610367081, + "grad_norm": 1.3304598331451416, + "learning_rate": 1.119687217166171e-05, + "loss": 0.2534, + "step": 49141 + }, + { + "epoch": 3.4100340018041773, + "grad_norm": 0.9813786149024963, + "learning_rate": 1.1194288967702482e-05, + "loss": 0.277, + "step": 49142 + }, + { + "epoch": 3.410103393241274, + "grad_norm": 1.211971640586853, + "learning_rate": 1.1191706044092554e-05, + "loss": 0.1927, + "step": 49143 + }, + { + "epoch": 3.4101727846783705, + "grad_norm": 0.6722222566604614, + "learning_rate": 1.1189123400840118e-05, + "loss": 0.2099, + "step": 49144 + }, + { + "epoch": 3.4102421761154673, + "grad_norm": 0.8346523642539978, + "learning_rate": 1.1186541037953313e-05, + "loss": 0.1702, + "step": 49145 + }, + { + "epoch": 3.410311567552564, + "grad_norm": 0.9086517691612244, + "learning_rate": 1.1183958955440288e-05, + "loss": 0.1813, + "step": 49146 + }, + { + "epoch": 3.4103809589896605, + "grad_norm": 0.8750752210617065, + "learning_rate": 1.1181377153309202e-05, + "loss": 0.2003, + "step": 49147 + }, + { + "epoch": 3.4104503504267574, + "grad_norm": 0.611168384552002, + "learning_rate": 1.1178795631568206e-05, + "loss": 0.1365, + "step": 49148 + }, + { + "epoch": 3.4105197418638538, + "grad_norm": 0.8302459120750427, + "learning_rate": 1.117621439022546e-05, + "loss": 0.2104, + "step": 49149 + }, + { + "epoch": 3.4105891333009506, + "grad_norm": 0.7389975786209106, + "learning_rate": 1.1173633429289087e-05, + "loss": 0.1877, + "step": 49150 + }, + { + "epoch": 3.4106585247380474, + "grad_norm": 1.0178359746932983, + "learning_rate": 1.1171052748767241e-05, + "loss": 0.3166, + "step": 49151 + }, + { + "epoch": 3.410727916175144, + "grad_norm": 0.8749765753746033, + "learning_rate": 1.1168472348668102e-05, + "loss": 0.2199, + "step": 49152 + }, + { + "epoch": 3.4107973076122406, + "grad_norm": 1.3935179710388184, + "learning_rate": 1.1165892228999764e-05, + "loss": 0.2383, + "step": 49153 + }, + { + "epoch": 3.4108666990493375, + "grad_norm": 0.7960079908370972, + "learning_rate": 1.1163312389770408e-05, + "loss": 0.3274, + "step": 49154 + }, + { + "epoch": 3.410936090486434, + "grad_norm": 0.7241479158401489, + "learning_rate": 1.1160732830988152e-05, + "loss": 0.2029, + "step": 49155 + }, + { + "epoch": 3.4110054819235307, + "grad_norm": 0.8541005849838257, + "learning_rate": 1.1158153552661187e-05, + "loss": 0.3225, + "step": 49156 + }, + { + "epoch": 3.4110748733606275, + "grad_norm": 1.3639079332351685, + "learning_rate": 1.11555745547976e-05, + "loss": 0.3095, + "step": 49157 + }, + { + "epoch": 3.411144264797724, + "grad_norm": 1.6151463985443115, + "learning_rate": 1.1152995837405555e-05, + "loss": 0.4481, + "step": 49158 + }, + { + "epoch": 3.4112136562348208, + "grad_norm": 1.7652350664138794, + "learning_rate": 1.1150417400493196e-05, + "loss": 0.4407, + "step": 49159 + }, + { + "epoch": 3.411283047671917, + "grad_norm": 1.0770992040634155, + "learning_rate": 1.1147839244068658e-05, + "loss": 0.2391, + "step": 49160 + }, + { + "epoch": 3.411352439109014, + "grad_norm": 1.216219186782837, + "learning_rate": 1.1145261368140092e-05, + "loss": 0.234, + "step": 49161 + }, + { + "epoch": 3.411421830546111, + "grad_norm": 0.7666634321212769, + "learning_rate": 1.1142683772715623e-05, + "loss": 0.1894, + "step": 49162 + }, + { + "epoch": 3.411491221983207, + "grad_norm": 0.8032666444778442, + "learning_rate": 1.1140106457803402e-05, + "loss": 0.2791, + "step": 49163 + }, + { + "epoch": 3.411560613420304, + "grad_norm": 0.8514783382415771, + "learning_rate": 1.113752942341154e-05, + "loss": 0.2526, + "step": 49164 + }, + { + "epoch": 3.4116300048574004, + "grad_norm": 0.9449464678764343, + "learning_rate": 1.1134952669548192e-05, + "loss": 0.2768, + "step": 49165 + }, + { + "epoch": 3.4116993962944973, + "grad_norm": 1.091837763786316, + "learning_rate": 1.1132376196221484e-05, + "loss": 0.4233, + "step": 49166 + }, + { + "epoch": 3.411768787731594, + "grad_norm": 1.5355901718139648, + "learning_rate": 1.1129800003439562e-05, + "loss": 0.2175, + "step": 49167 + }, + { + "epoch": 3.4118381791686905, + "grad_norm": 1.0924067497253418, + "learning_rate": 1.1127224091210542e-05, + "loss": 0.2557, + "step": 49168 + }, + { + "epoch": 3.4119075706057873, + "grad_norm": 0.787295937538147, + "learning_rate": 1.1124648459542563e-05, + "loss": 0.3395, + "step": 49169 + }, + { + "epoch": 3.4119769620428837, + "grad_norm": 0.6726049184799194, + "learning_rate": 1.1122073108443786e-05, + "loss": 0.2095, + "step": 49170 + }, + { + "epoch": 3.4120463534799805, + "grad_norm": 0.918700635433197, + "learning_rate": 1.1119498037922282e-05, + "loss": 0.2025, + "step": 49171 + }, + { + "epoch": 3.4121157449170774, + "grad_norm": 1.3199747800827026, + "learning_rate": 1.1116923247986222e-05, + "loss": 0.1907, + "step": 49172 + }, + { + "epoch": 3.4121851363541738, + "grad_norm": 0.794860303401947, + "learning_rate": 1.111434873864371e-05, + "loss": 0.2442, + "step": 49173 + }, + { + "epoch": 3.4122545277912706, + "grad_norm": 0.867233395576477, + "learning_rate": 1.1111774509902895e-05, + "loss": 0.1792, + "step": 49174 + }, + { + "epoch": 3.412323919228367, + "grad_norm": 1.008722186088562, + "learning_rate": 1.1109200561771894e-05, + "loss": 0.3254, + "step": 49175 + }, + { + "epoch": 3.412393310665464, + "grad_norm": 0.7918968796730042, + "learning_rate": 1.1106626894258831e-05, + "loss": 0.3133, + "step": 49176 + }, + { + "epoch": 3.4124627021025606, + "grad_norm": 0.8229545950889587, + "learning_rate": 1.1104053507371859e-05, + "loss": 0.2494, + "step": 49177 + }, + { + "epoch": 3.412532093539657, + "grad_norm": 0.7981412410736084, + "learning_rate": 1.1101480401119058e-05, + "loss": 0.2101, + "step": 49178 + }, + { + "epoch": 3.412601484976754, + "grad_norm": 1.0238440036773682, + "learning_rate": 1.1098907575508554e-05, + "loss": 0.2996, + "step": 49179 + }, + { + "epoch": 3.4126708764138507, + "grad_norm": 0.9978506565093994, + "learning_rate": 1.1096335030548499e-05, + "loss": 0.2559, + "step": 49180 + }, + { + "epoch": 3.412740267850947, + "grad_norm": 0.8906485438346863, + "learning_rate": 1.1093762766246995e-05, + "loss": 0.2426, + "step": 49181 + }, + { + "epoch": 3.412809659288044, + "grad_norm": 0.5559861063957214, + "learning_rate": 1.109119078261217e-05, + "loss": 0.1678, + "step": 49182 + }, + { + "epoch": 3.4128790507251408, + "grad_norm": 1.0544557571411133, + "learning_rate": 1.1088619079652152e-05, + "loss": 0.2893, + "step": 49183 + }, + { + "epoch": 3.412948442162237, + "grad_norm": 0.805675208568573, + "learning_rate": 1.1086047657375032e-05, + "loss": 0.1902, + "step": 49184 + }, + { + "epoch": 3.413017833599334, + "grad_norm": 0.7798198461532593, + "learning_rate": 1.1083476515788949e-05, + "loss": 0.1724, + "step": 49185 + }, + { + "epoch": 3.4130872250364304, + "grad_norm": 0.8765312433242798, + "learning_rate": 1.1080905654902007e-05, + "loss": 0.2365, + "step": 49186 + }, + { + "epoch": 3.413156616473527, + "grad_norm": 1.425474762916565, + "learning_rate": 1.1078335074722358e-05, + "loss": 0.337, + "step": 49187 + }, + { + "epoch": 3.413226007910624, + "grad_norm": 1.0533456802368164, + "learning_rate": 1.1075764775258057e-05, + "loss": 0.3599, + "step": 49188 + }, + { + "epoch": 3.4132953993477204, + "grad_norm": 0.8233643770217896, + "learning_rate": 1.1073194756517258e-05, + "loss": 0.2398, + "step": 49189 + }, + { + "epoch": 3.4133647907848172, + "grad_norm": 0.9729403257369995, + "learning_rate": 1.1070625018508063e-05, + "loss": 0.3005, + "step": 49190 + }, + { + "epoch": 3.4134341822219136, + "grad_norm": 0.9027039408683777, + "learning_rate": 1.106805556123861e-05, + "loss": 0.2105, + "step": 49191 + }, + { + "epoch": 3.4135035736590105, + "grad_norm": 0.9578970074653625, + "learning_rate": 1.1065486384716962e-05, + "loss": 0.2514, + "step": 49192 + }, + { + "epoch": 3.4135729650961073, + "grad_norm": 0.9759321808815002, + "learning_rate": 1.1062917488951251e-05, + "loss": 0.2466, + "step": 49193 + }, + { + "epoch": 3.4136423565332037, + "grad_norm": 1.0799598693847656, + "learning_rate": 1.10603488739496e-05, + "loss": 0.3269, + "step": 49194 + }, + { + "epoch": 3.4137117479703005, + "grad_norm": 0.7606794834136963, + "learning_rate": 1.1057780539720109e-05, + "loss": 0.2264, + "step": 49195 + }, + { + "epoch": 3.413781139407397, + "grad_norm": 1.29877769947052, + "learning_rate": 1.1055212486270883e-05, + "loss": 0.2981, + "step": 49196 + }, + { + "epoch": 3.4138505308444937, + "grad_norm": 0.8147504329681396, + "learning_rate": 1.1052644713610049e-05, + "loss": 0.193, + "step": 49197 + }, + { + "epoch": 3.4139199222815906, + "grad_norm": 1.1847971677780151, + "learning_rate": 1.105007722174568e-05, + "loss": 0.247, + "step": 49198 + }, + { + "epoch": 3.413989313718687, + "grad_norm": 0.9963827133178711, + "learning_rate": 1.1047510010685891e-05, + "loss": 0.282, + "step": 49199 + }, + { + "epoch": 3.414058705155784, + "grad_norm": 1.0433279275894165, + "learning_rate": 1.1044943080438808e-05, + "loss": 0.3003, + "step": 49200 + }, + { + "epoch": 3.41412809659288, + "grad_norm": 0.8594927787780762, + "learning_rate": 1.1042376431012502e-05, + "loss": 0.2844, + "step": 49201 + }, + { + "epoch": 3.414197488029977, + "grad_norm": 1.1821236610412598, + "learning_rate": 1.1039810062415101e-05, + "loss": 0.3276, + "step": 49202 + }, + { + "epoch": 3.414266879467074, + "grad_norm": 0.7624760866165161, + "learning_rate": 1.10372439746547e-05, + "loss": 0.2518, + "step": 49203 + }, + { + "epoch": 3.4143362709041702, + "grad_norm": 0.7894708514213562, + "learning_rate": 1.103467816773941e-05, + "loss": 0.2067, + "step": 49204 + }, + { + "epoch": 3.414405662341267, + "grad_norm": 0.9988821744918823, + "learning_rate": 1.1032112641677306e-05, + "loss": 0.3024, + "step": 49205 + }, + { + "epoch": 3.414475053778364, + "grad_norm": 0.7917543053627014, + "learning_rate": 1.1029547396476492e-05, + "loss": 0.1699, + "step": 49206 + }, + { + "epoch": 3.4145444452154603, + "grad_norm": 0.8498562574386597, + "learning_rate": 1.1026982432145083e-05, + "loss": 0.2828, + "step": 49207 + }, + { + "epoch": 3.414613836652557, + "grad_norm": 0.8131399750709534, + "learning_rate": 1.1024417748691162e-05, + "loss": 0.2327, + "step": 49208 + }, + { + "epoch": 3.414683228089654, + "grad_norm": 1.0570836067199707, + "learning_rate": 1.1021853346122834e-05, + "loss": 0.2769, + "step": 49209 + }, + { + "epoch": 3.4147526195267504, + "grad_norm": 1.2207667827606201, + "learning_rate": 1.1019289224448193e-05, + "loss": 0.412, + "step": 49210 + }, + { + "epoch": 3.414822010963847, + "grad_norm": 0.9390620589256287, + "learning_rate": 1.1016725383675353e-05, + "loss": 0.2638, + "step": 49211 + }, + { + "epoch": 3.4148914024009436, + "grad_norm": 0.7030231952667236, + "learning_rate": 1.1014161823812363e-05, + "loss": 0.1748, + "step": 49212 + }, + { + "epoch": 3.4149607938380404, + "grad_norm": 1.0170773267745972, + "learning_rate": 1.1011598544867341e-05, + "loss": 0.306, + "step": 49213 + }, + { + "epoch": 3.4150301852751372, + "grad_norm": 1.5173113346099854, + "learning_rate": 1.1009035546848379e-05, + "loss": 0.4965, + "step": 49214 + }, + { + "epoch": 3.4150995767122336, + "grad_norm": 0.9636831283569336, + "learning_rate": 1.1006472829763559e-05, + "loss": 0.2988, + "step": 49215 + }, + { + "epoch": 3.4151689681493305, + "grad_norm": 1.0184494256973267, + "learning_rate": 1.1003910393620987e-05, + "loss": 0.2234, + "step": 49216 + }, + { + "epoch": 3.415238359586427, + "grad_norm": 0.9170677661895752, + "learning_rate": 1.1001348238428743e-05, + "loss": 0.263, + "step": 49217 + }, + { + "epoch": 3.4153077510235237, + "grad_norm": 1.0391217470169067, + "learning_rate": 1.0998786364194947e-05, + "loss": 0.2027, + "step": 49218 + }, + { + "epoch": 3.4153771424606205, + "grad_norm": 0.8259015083312988, + "learning_rate": 1.0996224770927622e-05, + "loss": 0.2113, + "step": 49219 + }, + { + "epoch": 3.415446533897717, + "grad_norm": 0.8868549466133118, + "learning_rate": 1.0993663458634895e-05, + "loss": 0.2444, + "step": 49220 + }, + { + "epoch": 3.4155159253348137, + "grad_norm": 1.14894700050354, + "learning_rate": 1.099110242732484e-05, + "loss": 0.3682, + "step": 49221 + }, + { + "epoch": 3.41558531677191, + "grad_norm": 1.1496633291244507, + "learning_rate": 1.0988541677005571e-05, + "loss": 0.3126, + "step": 49222 + }, + { + "epoch": 3.415654708209007, + "grad_norm": 1.1243987083435059, + "learning_rate": 1.0985981207685136e-05, + "loss": 0.239, + "step": 49223 + }, + { + "epoch": 3.415724099646104, + "grad_norm": 0.9627781510353088, + "learning_rate": 1.098342101937162e-05, + "loss": 0.3143, + "step": 49224 + }, + { + "epoch": 3.4157934910832, + "grad_norm": 1.2567297220230103, + "learning_rate": 1.0980861112073149e-05, + "loss": 0.2433, + "step": 49225 + }, + { + "epoch": 3.415862882520297, + "grad_norm": 1.2713432312011719, + "learning_rate": 1.0978301485797738e-05, + "loss": 0.3563, + "step": 49226 + }, + { + "epoch": 3.415932273957394, + "grad_norm": 1.0770819187164307, + "learning_rate": 1.0975742140553502e-05, + "loss": 0.2513, + "step": 49227 + }, + { + "epoch": 3.4160016653944902, + "grad_norm": 0.906382143497467, + "learning_rate": 1.0973183076348525e-05, + "loss": 0.2554, + "step": 49228 + }, + { + "epoch": 3.416071056831587, + "grad_norm": 0.9253160953521729, + "learning_rate": 1.0970624293190878e-05, + "loss": 0.1952, + "step": 49229 + }, + { + "epoch": 3.4161404482686835, + "grad_norm": 0.7552793622016907, + "learning_rate": 1.0968065791088644e-05, + "loss": 0.2532, + "step": 49230 + }, + { + "epoch": 3.4162098397057803, + "grad_norm": 0.807074785232544, + "learning_rate": 1.0965507570049893e-05, + "loss": 0.2597, + "step": 49231 + }, + { + "epoch": 3.416279231142877, + "grad_norm": 0.8620643019676208, + "learning_rate": 1.0962949630082731e-05, + "loss": 0.2614, + "step": 49232 + }, + { + "epoch": 3.4163486225799735, + "grad_norm": 0.9775822162628174, + "learning_rate": 1.0960391971195183e-05, + "loss": 0.3115, + "step": 49233 + }, + { + "epoch": 3.4164180140170703, + "grad_norm": 1.0068413019180298, + "learning_rate": 1.0957834593395345e-05, + "loss": 0.3946, + "step": 49234 + }, + { + "epoch": 3.416487405454167, + "grad_norm": 0.9983062744140625, + "learning_rate": 1.0955277496691284e-05, + "loss": 0.3916, + "step": 49235 + }, + { + "epoch": 3.4165567968912636, + "grad_norm": 0.8198236227035522, + "learning_rate": 1.095272068109109e-05, + "loss": 0.2122, + "step": 49236 + }, + { + "epoch": 3.4166261883283604, + "grad_norm": 0.921730101108551, + "learning_rate": 1.0950164146602827e-05, + "loss": 0.17, + "step": 49237 + }, + { + "epoch": 3.416695579765457, + "grad_norm": 0.7210169434547424, + "learning_rate": 1.0947607893234569e-05, + "loss": 0.2018, + "step": 49238 + }, + { + "epoch": 3.4167649712025536, + "grad_norm": 0.852935791015625, + "learning_rate": 1.0945051920994388e-05, + "loss": 0.241, + "step": 49239 + }, + { + "epoch": 3.4168343626396505, + "grad_norm": 0.9548144936561584, + "learning_rate": 1.094249622989033e-05, + "loss": 0.2024, + "step": 49240 + }, + { + "epoch": 3.416903754076747, + "grad_norm": 1.0095134973526, + "learning_rate": 1.0939940819930483e-05, + "loss": 0.2641, + "step": 49241 + }, + { + "epoch": 3.4169731455138437, + "grad_norm": 0.7383982539176941, + "learning_rate": 1.0937385691122903e-05, + "loss": 0.2365, + "step": 49242 + }, + { + "epoch": 3.41704253695094, + "grad_norm": 0.8469067811965942, + "learning_rate": 1.0934830843475674e-05, + "loss": 0.1853, + "step": 49243 + }, + { + "epoch": 3.417111928388037, + "grad_norm": 0.9159548878669739, + "learning_rate": 1.0932276276996845e-05, + "loss": 0.2575, + "step": 49244 + }, + { + "epoch": 3.4171813198251337, + "grad_norm": 0.6970676183700562, + "learning_rate": 1.0929721991694486e-05, + "loss": 0.1882, + "step": 49245 + }, + { + "epoch": 3.41725071126223, + "grad_norm": 0.946785569190979, + "learning_rate": 1.0927167987576681e-05, + "loss": 0.2524, + "step": 49246 + }, + { + "epoch": 3.417320102699327, + "grad_norm": 0.7660281658172607, + "learning_rate": 1.0924614264651445e-05, + "loss": 0.2691, + "step": 49247 + }, + { + "epoch": 3.4173894941364233, + "grad_norm": 1.1973376274108887, + "learning_rate": 1.0922060822926871e-05, + "loss": 0.4295, + "step": 49248 + }, + { + "epoch": 3.41745888557352, + "grad_norm": 1.1235651969909668, + "learning_rate": 1.0919507662411022e-05, + "loss": 0.2671, + "step": 49249 + }, + { + "epoch": 3.417528277010617, + "grad_norm": 0.8421255350112915, + "learning_rate": 1.0916954783111955e-05, + "loss": 0.2527, + "step": 49250 + }, + { + "epoch": 3.4175976684477134, + "grad_norm": 0.8300194144248962, + "learning_rate": 1.0914402185037719e-05, + "loss": 0.1658, + "step": 49251 + }, + { + "epoch": 3.4176670598848102, + "grad_norm": 0.9503161311149597, + "learning_rate": 1.0911849868196377e-05, + "loss": 0.2619, + "step": 49252 + }, + { + "epoch": 3.417736451321907, + "grad_norm": 0.7121657133102417, + "learning_rate": 1.090929783259601e-05, + "loss": 0.1823, + "step": 49253 + }, + { + "epoch": 3.4178058427590035, + "grad_norm": 0.7813912630081177, + "learning_rate": 1.0906746078244634e-05, + "loss": 0.1785, + "step": 49254 + }, + { + "epoch": 3.4178752341961003, + "grad_norm": 0.6116968393325806, + "learning_rate": 1.0904194605150331e-05, + "loss": 0.2304, + "step": 49255 + }, + { + "epoch": 3.4179446256331967, + "grad_norm": 0.8812462091445923, + "learning_rate": 1.0901643413321138e-05, + "loss": 0.2332, + "step": 49256 + }, + { + "epoch": 3.4180140170702935, + "grad_norm": 1.0262359380722046, + "learning_rate": 1.089909250276515e-05, + "loss": 0.2845, + "step": 49257 + }, + { + "epoch": 3.4180834085073903, + "grad_norm": 1.0455846786499023, + "learning_rate": 1.0896541873490362e-05, + "loss": 0.2499, + "step": 49258 + }, + { + "epoch": 3.4181527999444867, + "grad_norm": 1.1076314449310303, + "learning_rate": 1.0893991525504854e-05, + "loss": 0.3025, + "step": 49259 + }, + { + "epoch": 3.4182221913815836, + "grad_norm": 1.3787739276885986, + "learning_rate": 1.0891441458816698e-05, + "loss": 0.2337, + "step": 49260 + }, + { + "epoch": 3.4182915828186804, + "grad_norm": 0.6326662302017212, + "learning_rate": 1.08888916734339e-05, + "loss": 0.12, + "step": 49261 + }, + { + "epoch": 3.418360974255777, + "grad_norm": 0.7611204385757446, + "learning_rate": 1.0886342169364539e-05, + "loss": 0.2684, + "step": 49262 + }, + { + "epoch": 3.4184303656928736, + "grad_norm": 0.7882281541824341, + "learning_rate": 1.0883792946616645e-05, + "loss": 0.2219, + "step": 49263 + }, + { + "epoch": 3.41849975712997, + "grad_norm": 0.7854518890380859, + "learning_rate": 1.0881244005198287e-05, + "loss": 0.2576, + "step": 49264 + }, + { + "epoch": 3.418569148567067, + "grad_norm": 0.7668372988700867, + "learning_rate": 1.0878695345117495e-05, + "loss": 0.219, + "step": 49265 + }, + { + "epoch": 3.4186385400041637, + "grad_norm": 0.7063999176025391, + "learning_rate": 1.0876146966382351e-05, + "loss": 0.2031, + "step": 49266 + }, + { + "epoch": 3.41870793144126, + "grad_norm": 0.6562079787254333, + "learning_rate": 1.0873598869000845e-05, + "loss": 0.2574, + "step": 49267 + }, + { + "epoch": 3.418777322878357, + "grad_norm": 1.0179188251495361, + "learning_rate": 1.0871051052981051e-05, + "loss": 0.206, + "step": 49268 + }, + { + "epoch": 3.4188467143154533, + "grad_norm": 0.9675671458244324, + "learning_rate": 1.0868503518331008e-05, + "loss": 0.3821, + "step": 49269 + }, + { + "epoch": 3.41891610575255, + "grad_norm": 0.686539888381958, + "learning_rate": 1.0865956265058753e-05, + "loss": 0.3409, + "step": 49270 + }, + { + "epoch": 3.418985497189647, + "grad_norm": 0.8690307140350342, + "learning_rate": 1.0863409293172333e-05, + "loss": 0.2677, + "step": 49271 + }, + { + "epoch": 3.4190548886267433, + "grad_norm": 1.0607082843780518, + "learning_rate": 1.0860862602679789e-05, + "loss": 0.1986, + "step": 49272 + }, + { + "epoch": 3.41912428006384, + "grad_norm": 0.5438799858093262, + "learning_rate": 1.085831619358918e-05, + "loss": 0.1547, + "step": 49273 + }, + { + "epoch": 3.4191936715009366, + "grad_norm": 0.9592445492744446, + "learning_rate": 1.08557700659085e-05, + "loss": 0.1616, + "step": 49274 + }, + { + "epoch": 3.4192630629380334, + "grad_norm": 1.1140996217727661, + "learning_rate": 1.0853224219645818e-05, + "loss": 0.2597, + "step": 49275 + }, + { + "epoch": 3.4193324543751302, + "grad_norm": 0.8644548058509827, + "learning_rate": 1.0850678654809165e-05, + "loss": 0.2202, + "step": 49276 + }, + { + "epoch": 3.4194018458122266, + "grad_norm": 1.1124211549758911, + "learning_rate": 1.0848133371406577e-05, + "loss": 0.2638, + "step": 49277 + }, + { + "epoch": 3.4194712372493234, + "grad_norm": 0.629656970500946, + "learning_rate": 1.084558836944608e-05, + "loss": 0.2158, + "step": 49278 + }, + { + "epoch": 3.4195406286864203, + "grad_norm": 1.0280089378356934, + "learning_rate": 1.0843043648935724e-05, + "loss": 0.2571, + "step": 49279 + }, + { + "epoch": 3.4196100201235167, + "grad_norm": 1.079443097114563, + "learning_rate": 1.0840499209883547e-05, + "loss": 0.2285, + "step": 49280 + }, + { + "epoch": 3.4196794115606135, + "grad_norm": 1.0319857597351074, + "learning_rate": 1.0837955052297554e-05, + "loss": 0.3724, + "step": 49281 + }, + { + "epoch": 3.41974880299771, + "grad_norm": 0.8019259572029114, + "learning_rate": 1.0835411176185795e-05, + "loss": 0.229, + "step": 49282 + }, + { + "epoch": 3.4198181944348067, + "grad_norm": 0.7231982946395874, + "learning_rate": 1.0832867581556305e-05, + "loss": 0.1348, + "step": 49283 + }, + { + "epoch": 3.4198875858719036, + "grad_norm": 0.6910269856452942, + "learning_rate": 1.0830324268417092e-05, + "loss": 0.2658, + "step": 49284 + }, + { + "epoch": 3.419956977309, + "grad_norm": 0.8191846609115601, + "learning_rate": 1.0827781236776202e-05, + "loss": 0.1973, + "step": 49285 + }, + { + "epoch": 3.420026368746097, + "grad_norm": 0.8343964219093323, + "learning_rate": 1.0825238486641675e-05, + "loss": 0.3329, + "step": 49286 + }, + { + "epoch": 3.4200957601831936, + "grad_norm": 0.7076783776283264, + "learning_rate": 1.0822696018021527e-05, + "loss": 0.1378, + "step": 49287 + }, + { + "epoch": 3.42016515162029, + "grad_norm": 1.1849796772003174, + "learning_rate": 1.0820153830923773e-05, + "loss": 0.2988, + "step": 49288 + }, + { + "epoch": 3.420234543057387, + "grad_norm": 1.206888198852539, + "learning_rate": 1.0817611925356442e-05, + "loss": 0.2979, + "step": 49289 + }, + { + "epoch": 3.4203039344944832, + "grad_norm": 1.2039506435394287, + "learning_rate": 1.0815070301327557e-05, + "loss": 0.4436, + "step": 49290 + }, + { + "epoch": 3.42037332593158, + "grad_norm": 1.0436292886734009, + "learning_rate": 1.0812528958845159e-05, + "loss": 0.2029, + "step": 49291 + }, + { + "epoch": 3.420442717368677, + "grad_norm": 1.0438575744628906, + "learning_rate": 1.0809987897917273e-05, + "loss": 0.293, + "step": 49292 + }, + { + "epoch": 3.4205121088057733, + "grad_norm": 0.9949218034744263, + "learning_rate": 1.0807447118551883e-05, + "loss": 0.2498, + "step": 49293 + }, + { + "epoch": 3.42058150024287, + "grad_norm": 0.8226456046104431, + "learning_rate": 1.0804906620757039e-05, + "loss": 0.2214, + "step": 49294 + }, + { + "epoch": 3.4206508916799665, + "grad_norm": 0.8751475811004639, + "learning_rate": 1.0802366404540776e-05, + "loss": 0.2046, + "step": 49295 + }, + { + "epoch": 3.4207202831170633, + "grad_norm": 0.7730920314788818, + "learning_rate": 1.0799826469911078e-05, + "loss": 0.2125, + "step": 49296 + }, + { + "epoch": 3.42078967455416, + "grad_norm": 1.148252010345459, + "learning_rate": 1.0797286816875973e-05, + "loss": 0.3017, + "step": 49297 + }, + { + "epoch": 3.4208590659912566, + "grad_norm": 0.8818536996841431, + "learning_rate": 1.0794747445443487e-05, + "loss": 0.1717, + "step": 49298 + }, + { + "epoch": 3.4209284574283534, + "grad_norm": 0.8663843870162964, + "learning_rate": 1.0792208355621624e-05, + "loss": 0.198, + "step": 49299 + }, + { + "epoch": 3.4209978488654498, + "grad_norm": 0.7087990045547485, + "learning_rate": 1.0789669547418423e-05, + "loss": 0.1565, + "step": 49300 + }, + { + "epoch": 3.4210672403025466, + "grad_norm": 0.638721227645874, + "learning_rate": 1.0787131020841902e-05, + "loss": 0.1989, + "step": 49301 + }, + { + "epoch": 3.4211366317396434, + "grad_norm": 0.8906516432762146, + "learning_rate": 1.0784592775900026e-05, + "loss": 0.3097, + "step": 49302 + }, + { + "epoch": 3.42120602317674, + "grad_norm": 0.7550854086875916, + "learning_rate": 1.078205481260085e-05, + "loss": 0.2351, + "step": 49303 + }, + { + "epoch": 3.4212754146138367, + "grad_norm": 0.8393189311027527, + "learning_rate": 1.0779517130952377e-05, + "loss": 0.2918, + "step": 49304 + }, + { + "epoch": 3.4213448060509335, + "grad_norm": 0.7387322783470154, + "learning_rate": 1.0776979730962611e-05, + "loss": 0.213, + "step": 49305 + }, + { + "epoch": 3.42141419748803, + "grad_norm": 0.6708134412765503, + "learning_rate": 1.0774442612639569e-05, + "loss": 0.2108, + "step": 49306 + }, + { + "epoch": 3.4214835889251267, + "grad_norm": 0.9318408370018005, + "learning_rate": 1.0771905775991265e-05, + "loss": 0.2187, + "step": 49307 + }, + { + "epoch": 3.4215529803622236, + "grad_norm": 0.689655065536499, + "learning_rate": 1.0769369221025717e-05, + "loss": 0.1633, + "step": 49308 + }, + { + "epoch": 3.42162237179932, + "grad_norm": 0.7719144225120544, + "learning_rate": 1.0766832947750905e-05, + "loss": 0.2528, + "step": 49309 + }, + { + "epoch": 3.4216917632364168, + "grad_norm": 1.1194968223571777, + "learning_rate": 1.0764296956174847e-05, + "loss": 0.2349, + "step": 49310 + }, + { + "epoch": 3.421761154673513, + "grad_norm": 1.0387704372406006, + "learning_rate": 1.0761761246305545e-05, + "loss": 0.3042, + "step": 49311 + }, + { + "epoch": 3.42183054611061, + "grad_norm": 0.8793666362762451, + "learning_rate": 1.0759225818151008e-05, + "loss": 0.1724, + "step": 49312 + }, + { + "epoch": 3.421899937547707, + "grad_norm": 1.619120478630066, + "learning_rate": 1.0756690671719239e-05, + "loss": 0.4542, + "step": 49313 + }, + { + "epoch": 3.421969328984803, + "grad_norm": 1.1680659055709839, + "learning_rate": 1.0754155807018252e-05, + "loss": 0.3161, + "step": 49314 + }, + { + "epoch": 3.4220387204219, + "grad_norm": 0.9622633457183838, + "learning_rate": 1.0751621224056052e-05, + "loss": 0.2119, + "step": 49315 + }, + { + "epoch": 3.4221081118589964, + "grad_norm": 1.2839255332946777, + "learning_rate": 1.0749086922840623e-05, + "loss": 0.3415, + "step": 49316 + }, + { + "epoch": 3.4221775032960933, + "grad_norm": 1.0969806909561157, + "learning_rate": 1.074655290337996e-05, + "loss": 0.1695, + "step": 49317 + }, + { + "epoch": 3.42224689473319, + "grad_norm": 1.054107904434204, + "learning_rate": 1.0744019165682073e-05, + "loss": 0.3929, + "step": 49318 + }, + { + "epoch": 3.4223162861702865, + "grad_norm": 1.3077070713043213, + "learning_rate": 1.0741485709754972e-05, + "loss": 0.2261, + "step": 49319 + }, + { + "epoch": 3.4223856776073833, + "grad_norm": 0.8484697341918945, + "learning_rate": 1.073895253560664e-05, + "loss": 0.3093, + "step": 49320 + }, + { + "epoch": 3.4224550690444797, + "grad_norm": 0.8408461809158325, + "learning_rate": 1.0736419643245067e-05, + "loss": 0.2316, + "step": 49321 + }, + { + "epoch": 3.4225244604815765, + "grad_norm": 2.1025445461273193, + "learning_rate": 1.073388703267829e-05, + "loss": 0.2744, + "step": 49322 + }, + { + "epoch": 3.4225938519186734, + "grad_norm": 0.8488897085189819, + "learning_rate": 1.0731354703914265e-05, + "loss": 0.259, + "step": 49323 + }, + { + "epoch": 3.4226632433557698, + "grad_norm": 0.672394871711731, + "learning_rate": 1.072882265696098e-05, + "loss": 0.159, + "step": 49324 + }, + { + "epoch": 3.4227326347928666, + "grad_norm": 0.7657946348190308, + "learning_rate": 1.0726290891826452e-05, + "loss": 0.2589, + "step": 49325 + }, + { + "epoch": 3.422802026229963, + "grad_norm": 0.8340896964073181, + "learning_rate": 1.0723759408518685e-05, + "loss": 0.2608, + "step": 49326 + }, + { + "epoch": 3.42287141766706, + "grad_norm": 0.9098958969116211, + "learning_rate": 1.072122820704563e-05, + "loss": 0.2453, + "step": 49327 + }, + { + "epoch": 3.4229408091041567, + "grad_norm": 1.5532652139663696, + "learning_rate": 1.0718697287415302e-05, + "loss": 0.4374, + "step": 49328 + }, + { + "epoch": 3.423010200541253, + "grad_norm": 0.9153653383255005, + "learning_rate": 1.0716166649635694e-05, + "loss": 0.1983, + "step": 49329 + }, + { + "epoch": 3.42307959197835, + "grad_norm": 1.1636784076690674, + "learning_rate": 1.0713636293714802e-05, + "loss": 0.492, + "step": 49330 + }, + { + "epoch": 3.4231489834154467, + "grad_norm": 0.8990610241889954, + "learning_rate": 1.0711106219660571e-05, + "loss": 0.2905, + "step": 49331 + }, + { + "epoch": 3.423218374852543, + "grad_norm": 0.7507341504096985, + "learning_rate": 1.0708576427481032e-05, + "loss": 0.2126, + "step": 49332 + }, + { + "epoch": 3.42328776628964, + "grad_norm": 0.7454273104667664, + "learning_rate": 1.0706046917184153e-05, + "loss": 0.227, + "step": 49333 + }, + { + "epoch": 3.4233571577267368, + "grad_norm": 0.67568439245224, + "learning_rate": 1.0703517688777931e-05, + "loss": 0.142, + "step": 49334 + }, + { + "epoch": 3.423426549163833, + "grad_norm": 1.5569982528686523, + "learning_rate": 1.0700988742270357e-05, + "loss": 0.3253, + "step": 49335 + }, + { + "epoch": 3.42349594060093, + "grad_norm": 0.8506060838699341, + "learning_rate": 1.069846007766938e-05, + "loss": 0.1831, + "step": 49336 + }, + { + "epoch": 3.4235653320380264, + "grad_norm": 0.6605414152145386, + "learning_rate": 1.0695931694983008e-05, + "loss": 0.1681, + "step": 49337 + }, + { + "epoch": 3.423634723475123, + "grad_norm": 1.1067864894866943, + "learning_rate": 1.069340359421921e-05, + "loss": 0.2047, + "step": 49338 + }, + { + "epoch": 3.42370411491222, + "grad_norm": 0.8148035407066345, + "learning_rate": 1.0690875775385977e-05, + "loss": 0.1679, + "step": 49339 + }, + { + "epoch": 3.4237735063493164, + "grad_norm": 0.9775962829589844, + "learning_rate": 1.0688348238491297e-05, + "loss": 0.2388, + "step": 49340 + }, + { + "epoch": 3.4238428977864133, + "grad_norm": 1.0808899402618408, + "learning_rate": 1.0685820983543139e-05, + "loss": 0.2611, + "step": 49341 + }, + { + "epoch": 3.4239122892235097, + "grad_norm": 0.8139563798904419, + "learning_rate": 1.0683294010549494e-05, + "loss": 0.1799, + "step": 49342 + }, + { + "epoch": 3.4239816806606065, + "grad_norm": 0.9436152577400208, + "learning_rate": 1.0680767319518303e-05, + "loss": 0.2184, + "step": 49343 + }, + { + "epoch": 3.4240510720977033, + "grad_norm": 0.8458937406539917, + "learning_rate": 1.0678240910457571e-05, + "loss": 0.223, + "step": 49344 + }, + { + "epoch": 3.4241204635347997, + "grad_norm": 0.8365634083747864, + "learning_rate": 1.0675714783375279e-05, + "loss": 0.2507, + "step": 49345 + }, + { + "epoch": 3.4241898549718965, + "grad_norm": 1.158961296081543, + "learning_rate": 1.0673188938279377e-05, + "loss": 0.2536, + "step": 49346 + }, + { + "epoch": 3.424259246408993, + "grad_norm": 0.7838602662086487, + "learning_rate": 1.067066337517787e-05, + "loss": 0.2029, + "step": 49347 + }, + { + "epoch": 3.4243286378460898, + "grad_norm": 0.9223832488059998, + "learning_rate": 1.0668138094078705e-05, + "loss": 0.2623, + "step": 49348 + }, + { + "epoch": 3.4243980292831866, + "grad_norm": 1.1586464643478394, + "learning_rate": 1.0665613094989879e-05, + "loss": 0.3206, + "step": 49349 + }, + { + "epoch": 3.424467420720283, + "grad_norm": 0.7201059460639954, + "learning_rate": 1.0663088377919338e-05, + "loss": 0.2594, + "step": 49350 + }, + { + "epoch": 3.42453681215738, + "grad_norm": 1.1886169910430908, + "learning_rate": 1.0660563942875057e-05, + "loss": 0.3435, + "step": 49351 + }, + { + "epoch": 3.424606203594476, + "grad_norm": 0.6425696611404419, + "learning_rate": 1.0658039789865005e-05, + "loss": 0.247, + "step": 49352 + }, + { + "epoch": 3.424675595031573, + "grad_norm": 1.0548138618469238, + "learning_rate": 1.0655515918897163e-05, + "loss": 0.2442, + "step": 49353 + }, + { + "epoch": 3.42474498646867, + "grad_norm": 0.7607158422470093, + "learning_rate": 1.0652992329979483e-05, + "loss": 0.1969, + "step": 49354 + }, + { + "epoch": 3.4248143779057663, + "grad_norm": 1.0239177942276, + "learning_rate": 1.0650469023119947e-05, + "loss": 0.3042, + "step": 49355 + }, + { + "epoch": 3.424883769342863, + "grad_norm": 1.253272294998169, + "learning_rate": 1.0647945998326537e-05, + "loss": 0.2332, + "step": 49356 + }, + { + "epoch": 3.42495316077996, + "grad_norm": 0.8019903302192688, + "learning_rate": 1.0645423255607168e-05, + "loss": 0.174, + "step": 49357 + }, + { + "epoch": 3.4250225522170563, + "grad_norm": 0.9187259674072266, + "learning_rate": 1.0642900794969834e-05, + "loss": 0.2038, + "step": 49358 + }, + { + "epoch": 3.425091943654153, + "grad_norm": 0.6706386208534241, + "learning_rate": 1.0640378616422497e-05, + "loss": 0.2097, + "step": 49359 + }, + { + "epoch": 3.42516133509125, + "grad_norm": 0.7076700329780579, + "learning_rate": 1.0637856719973116e-05, + "loss": 0.2451, + "step": 49360 + }, + { + "epoch": 3.4252307265283464, + "grad_norm": 0.9128900170326233, + "learning_rate": 1.0635335105629673e-05, + "loss": 0.1823, + "step": 49361 + }, + { + "epoch": 3.425300117965443, + "grad_norm": 1.0081353187561035, + "learning_rate": 1.0632813773400085e-05, + "loss": 0.2056, + "step": 49362 + }, + { + "epoch": 3.4253695094025396, + "grad_norm": 0.9186200499534607, + "learning_rate": 1.0630292723292345e-05, + "loss": 0.237, + "step": 49363 + }, + { + "epoch": 3.4254389008396364, + "grad_norm": 0.7555276155471802, + "learning_rate": 1.0627771955314393e-05, + "loss": 0.1709, + "step": 49364 + }, + { + "epoch": 3.4255082922767333, + "grad_norm": 0.883600115776062, + "learning_rate": 1.0625251469474217e-05, + "loss": 0.3081, + "step": 49365 + }, + { + "epoch": 3.4255776837138296, + "grad_norm": 0.8265511393547058, + "learning_rate": 1.0622731265779729e-05, + "loss": 0.1783, + "step": 49366 + }, + { + "epoch": 3.4256470751509265, + "grad_norm": 0.9612892270088196, + "learning_rate": 1.0620211344238917e-05, + "loss": 0.1902, + "step": 49367 + }, + { + "epoch": 3.425716466588023, + "grad_norm": 1.2008932828903198, + "learning_rate": 1.0617691704859723e-05, + "loss": 0.153, + "step": 49368 + }, + { + "epoch": 3.4257858580251197, + "grad_norm": 1.844415307044983, + "learning_rate": 1.0615172347650116e-05, + "loss": 0.2421, + "step": 49369 + }, + { + "epoch": 3.4258552494622165, + "grad_norm": 0.772560715675354, + "learning_rate": 1.0612653272618044e-05, + "loss": 0.2382, + "step": 49370 + }, + { + "epoch": 3.425924640899313, + "grad_norm": 0.9147597551345825, + "learning_rate": 1.0610134479771438e-05, + "loss": 0.2187, + "step": 49371 + }, + { + "epoch": 3.4259940323364098, + "grad_norm": 0.6179968118667603, + "learning_rate": 1.0607615969118279e-05, + "loss": 0.1872, + "step": 49372 + }, + { + "epoch": 3.426063423773506, + "grad_norm": 0.7733766436576843, + "learning_rate": 1.0605097740666491e-05, + "loss": 0.2525, + "step": 49373 + }, + { + "epoch": 3.426132815210603, + "grad_norm": 0.7638909816741943, + "learning_rate": 1.060257979442404e-05, + "loss": 0.1759, + "step": 49374 + }, + { + "epoch": 3.4262022066477, + "grad_norm": 1.1071866750717163, + "learning_rate": 1.060006213039888e-05, + "loss": 0.2241, + "step": 49375 + }, + { + "epoch": 3.426271598084796, + "grad_norm": 0.651296079158783, + "learning_rate": 1.0597544748598942e-05, + "loss": 0.1252, + "step": 49376 + }, + { + "epoch": 3.426340989521893, + "grad_norm": 0.9294683933258057, + "learning_rate": 1.0595027649032208e-05, + "loss": 0.2225, + "step": 49377 + }, + { + "epoch": 3.4264103809589894, + "grad_norm": 0.9132443070411682, + "learning_rate": 1.0592510831706582e-05, + "loss": 0.2882, + "step": 49378 + }, + { + "epoch": 3.4264797723960863, + "grad_norm": 0.9547825455665588, + "learning_rate": 1.0589994296630024e-05, + "loss": 0.3196, + "step": 49379 + }, + { + "epoch": 3.426549163833183, + "grad_norm": 0.8280970454216003, + "learning_rate": 1.0587478043810473e-05, + "loss": 0.2665, + "step": 49380 + }, + { + "epoch": 3.4266185552702795, + "grad_norm": 0.7947278618812561, + "learning_rate": 1.0584962073255889e-05, + "loss": 0.2538, + "step": 49381 + }, + { + "epoch": 3.4266879467073763, + "grad_norm": 0.8594609498977661, + "learning_rate": 1.0582446384974199e-05, + "loss": 0.3126, + "step": 49382 + }, + { + "epoch": 3.426757338144473, + "grad_norm": 1.133049726486206, + "learning_rate": 1.0579930978973362e-05, + "loss": 0.2581, + "step": 49383 + }, + { + "epoch": 3.4268267295815695, + "grad_norm": 0.958644449710846, + "learning_rate": 1.057741585526132e-05, + "loss": 0.2274, + "step": 49384 + }, + { + "epoch": 3.4268961210186664, + "grad_norm": 0.8601563572883606, + "learning_rate": 1.0574901013845984e-05, + "loss": 0.1814, + "step": 49385 + }, + { + "epoch": 3.426965512455763, + "grad_norm": 0.6750144958496094, + "learning_rate": 1.0572386454735305e-05, + "loss": 0.255, + "step": 49386 + }, + { + "epoch": 3.4270349038928596, + "grad_norm": 1.0756241083145142, + "learning_rate": 1.0569872177937223e-05, + "loss": 0.2485, + "step": 49387 + }, + { + "epoch": 3.4271042953299564, + "grad_norm": 0.841076135635376, + "learning_rate": 1.0567358183459686e-05, + "loss": 0.2179, + "step": 49388 + }, + { + "epoch": 3.427173686767053, + "grad_norm": 1.0952306985855103, + "learning_rate": 1.056484447131063e-05, + "loss": 0.3454, + "step": 49389 + }, + { + "epoch": 3.4272430782041496, + "grad_norm": 0.6748052835464478, + "learning_rate": 1.0562331041497974e-05, + "loss": 0.2145, + "step": 49390 + }, + { + "epoch": 3.4273124696412465, + "grad_norm": 1.0209494829177856, + "learning_rate": 1.055981789402969e-05, + "loss": 0.3061, + "step": 49391 + }, + { + "epoch": 3.427381861078343, + "grad_norm": 1.117216944694519, + "learning_rate": 1.0557305028913656e-05, + "loss": 0.2907, + "step": 49392 + }, + { + "epoch": 3.4274512525154397, + "grad_norm": 0.7649822235107422, + "learning_rate": 1.0554792446157834e-05, + "loss": 0.2046, + "step": 49393 + }, + { + "epoch": 3.427520643952536, + "grad_norm": 1.4888291358947754, + "learning_rate": 1.0552280145770155e-05, + "loss": 0.2964, + "step": 49394 + }, + { + "epoch": 3.427590035389633, + "grad_norm": 1.0562870502471924, + "learning_rate": 1.0549768127758553e-05, + "loss": 0.4867, + "step": 49395 + }, + { + "epoch": 3.4276594268267297, + "grad_norm": 0.8162364959716797, + "learning_rate": 1.0547256392130967e-05, + "loss": 0.1223, + "step": 49396 + }, + { + "epoch": 3.427728818263826, + "grad_norm": 1.0796821117401123, + "learning_rate": 1.0544744938895301e-05, + "loss": 0.2763, + "step": 49397 + }, + { + "epoch": 3.427798209700923, + "grad_norm": 1.0301471948623657, + "learning_rate": 1.0542233768059496e-05, + "loss": 0.289, + "step": 49398 + }, + { + "epoch": 3.4278676011380194, + "grad_norm": 1.2545009851455688, + "learning_rate": 1.0539722879631475e-05, + "loss": 0.3095, + "step": 49399 + }, + { + "epoch": 3.427936992575116, + "grad_norm": 0.7118389010429382, + "learning_rate": 1.0537212273619202e-05, + "loss": 0.3024, + "step": 49400 + }, + { + "epoch": 3.428006384012213, + "grad_norm": 1.0282999277114868, + "learning_rate": 1.0534701950030534e-05, + "loss": 0.2981, + "step": 49401 + }, + { + "epoch": 3.4280757754493094, + "grad_norm": 1.223204255104065, + "learning_rate": 1.0532191908873445e-05, + "loss": 0.2782, + "step": 49402 + }, + { + "epoch": 3.4281451668864062, + "grad_norm": 0.6721897125244141, + "learning_rate": 1.0529682150155839e-05, + "loss": 0.2316, + "step": 49403 + }, + { + "epoch": 3.4282145583235026, + "grad_norm": 0.9255769848823547, + "learning_rate": 1.0527172673885655e-05, + "loss": 0.2989, + "step": 49404 + }, + { + "epoch": 3.4282839497605995, + "grad_norm": 0.7405681610107422, + "learning_rate": 1.0524663480070817e-05, + "loss": 0.2256, + "step": 49405 + }, + { + "epoch": 3.4283533411976963, + "grad_norm": 9.15955924987793, + "learning_rate": 1.0522154568719212e-05, + "loss": 0.2947, + "step": 49406 + }, + { + "epoch": 3.4284227326347927, + "grad_norm": 1.131941795349121, + "learning_rate": 1.0519645939838796e-05, + "loss": 0.2115, + "step": 49407 + }, + { + "epoch": 3.4284921240718895, + "grad_norm": 0.607718825340271, + "learning_rate": 1.0517137593437465e-05, + "loss": 0.2193, + "step": 49408 + }, + { + "epoch": 3.4285615155089864, + "grad_norm": 1.1292681694030762, + "learning_rate": 1.0514629529523157e-05, + "loss": 0.3364, + "step": 49409 + }, + { + "epoch": 3.4286309069460827, + "grad_norm": 0.8200529217720032, + "learning_rate": 1.0512121748103776e-05, + "loss": 0.2057, + "step": 49410 + }, + { + "epoch": 3.4287002983831796, + "grad_norm": 0.8279457092285156, + "learning_rate": 1.050961424918726e-05, + "loss": 0.2175, + "step": 49411 + }, + { + "epoch": 3.4287696898202764, + "grad_norm": 0.8576617240905762, + "learning_rate": 1.0507107032781494e-05, + "loss": 0.233, + "step": 49412 + }, + { + "epoch": 3.428839081257373, + "grad_norm": 0.7522841691970825, + "learning_rate": 1.0504600098894413e-05, + "loss": 0.3432, + "step": 49413 + }, + { + "epoch": 3.4289084726944696, + "grad_norm": 0.6790022850036621, + "learning_rate": 1.0502093447533923e-05, + "loss": 0.1821, + "step": 49414 + }, + { + "epoch": 3.428977864131566, + "grad_norm": 1.0425138473510742, + "learning_rate": 1.049958707870794e-05, + "loss": 0.3082, + "step": 49415 + }, + { + "epoch": 3.429047255568663, + "grad_norm": 0.8232097625732422, + "learning_rate": 1.0497080992424369e-05, + "loss": 0.1508, + "step": 49416 + }, + { + "epoch": 3.4291166470057597, + "grad_norm": 0.7265322804450989, + "learning_rate": 1.0494575188691135e-05, + "loss": 0.2123, + "step": 49417 + }, + { + "epoch": 3.429186038442856, + "grad_norm": 0.7190196514129639, + "learning_rate": 1.0492069667516168e-05, + "loss": 0.1772, + "step": 49418 + }, + { + "epoch": 3.429255429879953, + "grad_norm": 0.8861119151115417, + "learning_rate": 1.0489564428907329e-05, + "loss": 0.2001, + "step": 49419 + }, + { + "epoch": 3.4293248213170493, + "grad_norm": 2.1825454235076904, + "learning_rate": 1.048705947287254e-05, + "loss": 0.3283, + "step": 49420 + }, + { + "epoch": 3.429394212754146, + "grad_norm": 1.1454883813858032, + "learning_rate": 1.0484554799419733e-05, + "loss": 0.3443, + "step": 49421 + }, + { + "epoch": 3.429463604191243, + "grad_norm": 1.057496428489685, + "learning_rate": 1.0482050408556798e-05, + "loss": 0.2822, + "step": 49422 + }, + { + "epoch": 3.4295329956283394, + "grad_norm": 0.7533576488494873, + "learning_rate": 1.0479546300291631e-05, + "loss": 0.3095, + "step": 49423 + }, + { + "epoch": 3.429602387065436, + "grad_norm": 0.8005817532539368, + "learning_rate": 1.047704247463217e-05, + "loss": 0.291, + "step": 49424 + }, + { + "epoch": 3.4296717785025326, + "grad_norm": 1.2870429754257202, + "learning_rate": 1.0474538931586309e-05, + "loss": 0.191, + "step": 49425 + }, + { + "epoch": 3.4297411699396294, + "grad_norm": 0.7111440896987915, + "learning_rate": 1.0472035671161917e-05, + "loss": 0.1555, + "step": 49426 + }, + { + "epoch": 3.4298105613767262, + "grad_norm": 0.9074242115020752, + "learning_rate": 1.0469532693366923e-05, + "loss": 0.2279, + "step": 49427 + }, + { + "epoch": 3.4298799528138226, + "grad_norm": 0.8162553906440735, + "learning_rate": 1.0467029998209232e-05, + "loss": 0.2984, + "step": 49428 + }, + { + "epoch": 3.4299493442509195, + "grad_norm": 0.5603182911872864, + "learning_rate": 1.0464527585696737e-05, + "loss": 0.1671, + "step": 49429 + }, + { + "epoch": 3.430018735688016, + "grad_norm": 0.9015000462532043, + "learning_rate": 1.0462025455837343e-05, + "loss": 0.1951, + "step": 49430 + }, + { + "epoch": 3.4300881271251127, + "grad_norm": 1.099650502204895, + "learning_rate": 1.0459523608638967e-05, + "loss": 0.2226, + "step": 49431 + }, + { + "epoch": 3.4301575185622095, + "grad_norm": 0.8092737197875977, + "learning_rate": 1.0457022044109454e-05, + "loss": 0.1737, + "step": 49432 + }, + { + "epoch": 3.430226909999306, + "grad_norm": 0.9131715297698975, + "learning_rate": 1.0454520762256748e-05, + "loss": 0.2766, + "step": 49433 + }, + { + "epoch": 3.4302963014364027, + "grad_norm": 0.9621526598930359, + "learning_rate": 1.0452019763088727e-05, + "loss": 0.2179, + "step": 49434 + }, + { + "epoch": 3.4303656928734996, + "grad_norm": 0.9545508027076721, + "learning_rate": 1.044951904661331e-05, + "loss": 0.4151, + "step": 49435 + }, + { + "epoch": 3.430435084310596, + "grad_norm": 1.3857170343399048, + "learning_rate": 1.0447018612838344e-05, + "loss": 0.4322, + "step": 49436 + }, + { + "epoch": 3.430504475747693, + "grad_norm": 0.6743165254592896, + "learning_rate": 1.0444518461771757e-05, + "loss": 0.2722, + "step": 49437 + }, + { + "epoch": 3.4305738671847896, + "grad_norm": 0.6567783355712891, + "learning_rate": 1.0442018593421444e-05, + "loss": 0.2427, + "step": 49438 + }, + { + "epoch": 3.430643258621886, + "grad_norm": 0.8048530220985413, + "learning_rate": 1.0439519007795296e-05, + "loss": 0.2092, + "step": 49439 + }, + { + "epoch": 3.430712650058983, + "grad_norm": 1.1224433183670044, + "learning_rate": 1.0437019704901175e-05, + "loss": 0.2656, + "step": 49440 + }, + { + "epoch": 3.4307820414960792, + "grad_norm": 0.8256992101669312, + "learning_rate": 1.0434520684746996e-05, + "loss": 0.2737, + "step": 49441 + }, + { + "epoch": 3.430851432933176, + "grad_norm": 0.7262440323829651, + "learning_rate": 1.0432021947340642e-05, + "loss": 0.2222, + "step": 49442 + }, + { + "epoch": 3.430920824370273, + "grad_norm": 0.7360460758209229, + "learning_rate": 1.0429523492689997e-05, + "loss": 0.1482, + "step": 49443 + }, + { + "epoch": 3.4309902158073693, + "grad_norm": 0.9910228848457336, + "learning_rate": 1.0427025320802963e-05, + "loss": 0.2222, + "step": 49444 + }, + { + "epoch": 3.431059607244466, + "grad_norm": 0.9782713055610657, + "learning_rate": 1.0424527431687415e-05, + "loss": 0.2225, + "step": 49445 + }, + { + "epoch": 3.4311289986815625, + "grad_norm": 0.8049613237380981, + "learning_rate": 1.0422029825351264e-05, + "loss": 0.1681, + "step": 49446 + }, + { + "epoch": 3.4311983901186593, + "grad_norm": 1.216939091682434, + "learning_rate": 1.0419532501802343e-05, + "loss": 0.2641, + "step": 49447 + }, + { + "epoch": 3.431267781555756, + "grad_norm": 1.0406675338745117, + "learning_rate": 1.0417035461048574e-05, + "loss": 0.2796, + "step": 49448 + }, + { + "epoch": 3.4313371729928526, + "grad_norm": 0.7899605631828308, + "learning_rate": 1.0414538703097832e-05, + "loss": 0.2401, + "step": 49449 + }, + { + "epoch": 3.4314065644299494, + "grad_norm": 1.2520606517791748, + "learning_rate": 1.0412042227957985e-05, + "loss": 0.3569, + "step": 49450 + }, + { + "epoch": 3.431475955867046, + "grad_norm": 1.104774832725525, + "learning_rate": 1.0409546035636931e-05, + "loss": 0.2256, + "step": 49451 + }, + { + "epoch": 3.4315453473041426, + "grad_norm": 1.4287728071212769, + "learning_rate": 1.0407050126142548e-05, + "loss": 0.2776, + "step": 49452 + }, + { + "epoch": 3.4316147387412395, + "grad_norm": 0.8434779047966003, + "learning_rate": 1.0404554499482743e-05, + "loss": 0.2841, + "step": 49453 + }, + { + "epoch": 3.431684130178336, + "grad_norm": 1.571722388267517, + "learning_rate": 1.0402059155665334e-05, + "loss": 0.1541, + "step": 49454 + }, + { + "epoch": 3.4317535216154327, + "grad_norm": 0.8831202387809753, + "learning_rate": 1.0399564094698222e-05, + "loss": 0.3052, + "step": 49455 + }, + { + "epoch": 3.4318229130525295, + "grad_norm": 0.8598207831382751, + "learning_rate": 1.0397069316589304e-05, + "loss": 0.1984, + "step": 49456 + }, + { + "epoch": 3.431892304489626, + "grad_norm": 1.0036883354187012, + "learning_rate": 1.0394574821346437e-05, + "loss": 0.2869, + "step": 49457 + }, + { + "epoch": 3.4319616959267227, + "grad_norm": 1.331790804862976, + "learning_rate": 1.0392080608977506e-05, + "loss": 0.5063, + "step": 49458 + }, + { + "epoch": 3.432031087363819, + "grad_norm": 0.5284387469291687, + "learning_rate": 1.038958667949037e-05, + "loss": 0.1772, + "step": 49459 + }, + { + "epoch": 3.432100478800916, + "grad_norm": 1.149817705154419, + "learning_rate": 1.0387093032892936e-05, + "loss": 0.2872, + "step": 49460 + }, + { + "epoch": 3.432169870238013, + "grad_norm": 1.5392560958862305, + "learning_rate": 1.038459966919304e-05, + "loss": 0.4217, + "step": 49461 + }, + { + "epoch": 3.432239261675109, + "grad_norm": 0.7543219923973083, + "learning_rate": 1.0382106588398555e-05, + "loss": 0.1774, + "step": 49462 + }, + { + "epoch": 3.432308653112206, + "grad_norm": 0.5852826833724976, + "learning_rate": 1.0379613790517363e-05, + "loss": 0.1611, + "step": 49463 + }, + { + "epoch": 3.432378044549303, + "grad_norm": 1.3286449909210205, + "learning_rate": 1.0377121275557333e-05, + "loss": 0.2623, + "step": 49464 + }, + { + "epoch": 3.4324474359863992, + "grad_norm": 0.6875396966934204, + "learning_rate": 1.037462904352634e-05, + "loss": 0.1905, + "step": 49465 + }, + { + "epoch": 3.432516827423496, + "grad_norm": 0.6699708700180054, + "learning_rate": 1.0372137094432266e-05, + "loss": 0.1573, + "step": 49466 + }, + { + "epoch": 3.4325862188605925, + "grad_norm": 0.9884459376335144, + "learning_rate": 1.0369645428282926e-05, + "loss": 0.3113, + "step": 49467 + }, + { + "epoch": 3.4326556102976893, + "grad_norm": 0.8006325364112854, + "learning_rate": 1.0367154045086224e-05, + "loss": 0.2496, + "step": 49468 + }, + { + "epoch": 3.432725001734786, + "grad_norm": 1.2974934577941895, + "learning_rate": 1.0364662944850035e-05, + "loss": 0.2416, + "step": 49469 + }, + { + "epoch": 3.4327943931718825, + "grad_norm": 1.5559957027435303, + "learning_rate": 1.0362172127582182e-05, + "loss": 0.3343, + "step": 49470 + }, + { + "epoch": 3.4328637846089793, + "grad_norm": 1.0299780368804932, + "learning_rate": 1.0359681593290548e-05, + "loss": 0.2819, + "step": 49471 + }, + { + "epoch": 3.4329331760460757, + "grad_norm": 1.0953327417373657, + "learning_rate": 1.0357191341983008e-05, + "loss": 0.197, + "step": 49472 + }, + { + "epoch": 3.4330025674831726, + "grad_norm": 1.2464492321014404, + "learning_rate": 1.0354701373667408e-05, + "loss": 0.2544, + "step": 49473 + }, + { + "epoch": 3.4330719589202694, + "grad_norm": 0.7879282832145691, + "learning_rate": 1.0352211688351642e-05, + "loss": 0.26, + "step": 49474 + }, + { + "epoch": 3.433141350357366, + "grad_norm": 0.9885644316673279, + "learning_rate": 1.0349722286043518e-05, + "loss": 0.2791, + "step": 49475 + }, + { + "epoch": 3.4332107417944626, + "grad_norm": 0.7439317107200623, + "learning_rate": 1.0347233166750924e-05, + "loss": 0.2199, + "step": 49476 + }, + { + "epoch": 3.433280133231559, + "grad_norm": 0.6027300357818604, + "learning_rate": 1.0344744330481715e-05, + "loss": 0.1605, + "step": 49477 + }, + { + "epoch": 3.433349524668656, + "grad_norm": 1.6140559911727905, + "learning_rate": 1.0342255777243737e-05, + "loss": 0.2411, + "step": 49478 + }, + { + "epoch": 3.4334189161057527, + "grad_norm": 0.8933934569358826, + "learning_rate": 1.0339767507044861e-05, + "loss": 0.2799, + "step": 49479 + }, + { + "epoch": 3.433488307542849, + "grad_norm": 0.91643887758255, + "learning_rate": 1.0337279519892929e-05, + "loss": 0.2438, + "step": 49480 + }, + { + "epoch": 3.433557698979946, + "grad_norm": 0.8522992134094238, + "learning_rate": 1.0334791815795831e-05, + "loss": 0.2526, + "step": 49481 + }, + { + "epoch": 3.4336270904170427, + "grad_norm": 0.8397441506385803, + "learning_rate": 1.0332304394761372e-05, + "loss": 0.2881, + "step": 49482 + }, + { + "epoch": 3.433696481854139, + "grad_norm": 0.7692822813987732, + "learning_rate": 1.0329817256797425e-05, + "loss": 0.2571, + "step": 49483 + }, + { + "epoch": 3.433765873291236, + "grad_norm": 0.8609503507614136, + "learning_rate": 1.0327330401911838e-05, + "loss": 0.2512, + "step": 49484 + }, + { + "epoch": 3.4338352647283323, + "grad_norm": 0.8685590624809265, + "learning_rate": 1.0324843830112462e-05, + "loss": 0.2725, + "step": 49485 + }, + { + "epoch": 3.433904656165429, + "grad_norm": 1.116329312324524, + "learning_rate": 1.0322357541407157e-05, + "loss": 0.2671, + "step": 49486 + }, + { + "epoch": 3.433974047602526, + "grad_norm": 0.8527453541755676, + "learning_rate": 1.0319871535803782e-05, + "loss": 0.2158, + "step": 49487 + }, + { + "epoch": 3.4340434390396224, + "grad_norm": 1.0214353799819946, + "learning_rate": 1.0317385813310143e-05, + "loss": 0.2195, + "step": 49488 + }, + { + "epoch": 3.4341128304767192, + "grad_norm": 0.912874162197113, + "learning_rate": 1.0314900373934123e-05, + "loss": 0.2076, + "step": 49489 + }, + { + "epoch": 3.434182221913816, + "grad_norm": 1.1834970712661743, + "learning_rate": 1.0312415217683546e-05, + "loss": 0.3394, + "step": 49490 + }, + { + "epoch": 3.4342516133509124, + "grad_norm": 1.4434406757354736, + "learning_rate": 1.0309930344566265e-05, + "loss": 0.3614, + "step": 49491 + }, + { + "epoch": 3.4343210047880093, + "grad_norm": 1.2540749311447144, + "learning_rate": 1.030744575459014e-05, + "loss": 0.3014, + "step": 49492 + }, + { + "epoch": 3.4343903962251057, + "grad_norm": 0.7692523002624512, + "learning_rate": 1.0304961447762995e-05, + "loss": 0.2469, + "step": 49493 + }, + { + "epoch": 3.4344597876622025, + "grad_norm": 1.1835803985595703, + "learning_rate": 1.0302477424092704e-05, + "loss": 0.2278, + "step": 49494 + }, + { + "epoch": 3.4345291790992993, + "grad_norm": 0.8660014271736145, + "learning_rate": 1.0299993683587061e-05, + "loss": 0.2546, + "step": 49495 + }, + { + "epoch": 3.4345985705363957, + "grad_norm": 1.0993664264678955, + "learning_rate": 1.0297510226253937e-05, + "loss": 0.2565, + "step": 49496 + }, + { + "epoch": 3.4346679619734926, + "grad_norm": 1.0668461322784424, + "learning_rate": 1.0295027052101169e-05, + "loss": 0.2538, + "step": 49497 + }, + { + "epoch": 3.434737353410589, + "grad_norm": 1.1388435363769531, + "learning_rate": 1.0292544161136585e-05, + "loss": 0.2731, + "step": 49498 + }, + { + "epoch": 3.4348067448476858, + "grad_norm": 0.9548549652099609, + "learning_rate": 1.0290061553368024e-05, + "loss": 0.3386, + "step": 49499 + }, + { + "epoch": 3.4348761362847826, + "grad_norm": 1.0401206016540527, + "learning_rate": 1.0287579228803346e-05, + "loss": 0.275, + "step": 49500 + }, + { + "epoch": 3.434945527721879, + "grad_norm": 1.133705973625183, + "learning_rate": 1.0285097187450388e-05, + "loss": 0.191, + "step": 49501 + }, + { + "epoch": 3.435014919158976, + "grad_norm": 0.9019893407821655, + "learning_rate": 1.0282615429316945e-05, + "loss": 0.2262, + "step": 49502 + }, + { + "epoch": 3.435084310596072, + "grad_norm": 0.8560287356376648, + "learning_rate": 1.0280133954410876e-05, + "loss": 0.2739, + "step": 49503 + }, + { + "epoch": 3.435153702033169, + "grad_norm": 0.8679932355880737, + "learning_rate": 1.0277652762740032e-05, + "loss": 0.2559, + "step": 49504 + }, + { + "epoch": 3.435223093470266, + "grad_norm": 0.9605547785758972, + "learning_rate": 1.0275171854312215e-05, + "loss": 0.2091, + "step": 49505 + }, + { + "epoch": 3.4352924849073623, + "grad_norm": 1.3138309717178345, + "learning_rate": 1.0272691229135268e-05, + "loss": 0.2937, + "step": 49506 + }, + { + "epoch": 3.435361876344459, + "grad_norm": 1.01674222946167, + "learning_rate": 1.0270210887217025e-05, + "loss": 0.4036, + "step": 49507 + }, + { + "epoch": 3.435431267781556, + "grad_norm": 1.0657720565795898, + "learning_rate": 1.0267730828565336e-05, + "loss": 0.3651, + "step": 49508 + }, + { + "epoch": 3.4355006592186523, + "grad_norm": 1.008225917816162, + "learning_rate": 1.0265251053187985e-05, + "loss": 0.3419, + "step": 49509 + }, + { + "epoch": 3.435570050655749, + "grad_norm": 1.3015888929367065, + "learning_rate": 1.0262771561092832e-05, + "loss": 0.3095, + "step": 49510 + }, + { + "epoch": 3.4356394420928456, + "grad_norm": 1.2131054401397705, + "learning_rate": 1.0260292352287693e-05, + "loss": 0.2765, + "step": 49511 + }, + { + "epoch": 3.4357088335299424, + "grad_norm": 1.1525838375091553, + "learning_rate": 1.0257813426780404e-05, + "loss": 0.3634, + "step": 49512 + }, + { + "epoch": 3.435778224967039, + "grad_norm": 1.1714884042739868, + "learning_rate": 1.0255334784578773e-05, + "loss": 0.3839, + "step": 49513 + }, + { + "epoch": 3.4358476164041356, + "grad_norm": 0.8403964042663574, + "learning_rate": 1.0252856425690649e-05, + "loss": 0.1973, + "step": 49514 + }, + { + "epoch": 3.4359170078412324, + "grad_norm": 0.7720906734466553, + "learning_rate": 1.0250378350123868e-05, + "loss": 0.2298, + "step": 49515 + }, + { + "epoch": 3.4359863992783293, + "grad_norm": 0.7355112433433533, + "learning_rate": 1.0247900557886203e-05, + "loss": 0.2262, + "step": 49516 + }, + { + "epoch": 3.4360557907154257, + "grad_norm": 1.2157225608825684, + "learning_rate": 1.0245423048985502e-05, + "loss": 0.3282, + "step": 49517 + }, + { + "epoch": 3.4361251821525225, + "grad_norm": 0.8292527198791504, + "learning_rate": 1.0242945823429594e-05, + "loss": 0.3016, + "step": 49518 + }, + { + "epoch": 3.436194573589619, + "grad_norm": 0.7524645924568176, + "learning_rate": 1.0240468881226284e-05, + "loss": 0.2459, + "step": 49519 + }, + { + "epoch": 3.4362639650267157, + "grad_norm": 0.9231749773025513, + "learning_rate": 1.0237992222383409e-05, + "loss": 0.2473, + "step": 49520 + }, + { + "epoch": 3.4363333564638125, + "grad_norm": 1.11090886592865, + "learning_rate": 1.0235515846908772e-05, + "loss": 0.3356, + "step": 49521 + }, + { + "epoch": 3.436402747900909, + "grad_norm": 0.8136181235313416, + "learning_rate": 1.0233039754810214e-05, + "loss": 0.2408, + "step": 49522 + }, + { + "epoch": 3.4364721393380058, + "grad_norm": 1.780773401260376, + "learning_rate": 1.0230563946095528e-05, + "loss": 0.3315, + "step": 49523 + }, + { + "epoch": 3.436541530775102, + "grad_norm": 0.9785588383674622, + "learning_rate": 1.0228088420772519e-05, + "loss": 0.2184, + "step": 49524 + }, + { + "epoch": 3.436610922212199, + "grad_norm": 0.7864631414413452, + "learning_rate": 1.0225613178849036e-05, + "loss": 0.2281, + "step": 49525 + }, + { + "epoch": 3.436680313649296, + "grad_norm": 0.6888351440429688, + "learning_rate": 1.0223138220332863e-05, + "loss": 0.1887, + "step": 49526 + }, + { + "epoch": 3.436749705086392, + "grad_norm": 1.0389611721038818, + "learning_rate": 1.0220663545231834e-05, + "loss": 0.2032, + "step": 49527 + }, + { + "epoch": 3.436819096523489, + "grad_norm": 0.9498452544212341, + "learning_rate": 1.0218189153553748e-05, + "loss": 0.1627, + "step": 49528 + }, + { + "epoch": 3.4368884879605854, + "grad_norm": 1.181961178779602, + "learning_rate": 1.0215715045306451e-05, + "loss": 0.3415, + "step": 49529 + }, + { + "epoch": 3.4369578793976823, + "grad_norm": 0.8585386872291565, + "learning_rate": 1.0213241220497694e-05, + "loss": 0.2105, + "step": 49530 + }, + { + "epoch": 3.437027270834779, + "grad_norm": 0.8226932883262634, + "learning_rate": 1.0210767679135325e-05, + "loss": 0.1653, + "step": 49531 + }, + { + "epoch": 3.4370966622718755, + "grad_norm": 0.9311769008636475, + "learning_rate": 1.0208294421227139e-05, + "loss": 0.1973, + "step": 49532 + }, + { + "epoch": 3.4371660537089723, + "grad_norm": 1.0876095294952393, + "learning_rate": 1.020582144678095e-05, + "loss": 0.237, + "step": 49533 + }, + { + "epoch": 3.437235445146069, + "grad_norm": 0.8546052575111389, + "learning_rate": 1.0203348755804576e-05, + "loss": 0.1366, + "step": 49534 + }, + { + "epoch": 3.4373048365831655, + "grad_norm": 0.8748703002929688, + "learning_rate": 1.0200876348305821e-05, + "loss": 0.2843, + "step": 49535 + }, + { + "epoch": 3.4373742280202624, + "grad_norm": 0.9792153835296631, + "learning_rate": 1.0198404224292457e-05, + "loss": 0.1852, + "step": 49536 + }, + { + "epoch": 3.4374436194573588, + "grad_norm": 1.1949512958526611, + "learning_rate": 1.0195932383772322e-05, + "loss": 0.352, + "step": 49537 + }, + { + "epoch": 3.4375130108944556, + "grad_norm": 1.0565673112869263, + "learning_rate": 1.0193460826753209e-05, + "loss": 0.2828, + "step": 49538 + }, + { + "epoch": 3.4375824023315524, + "grad_norm": 0.8984784483909607, + "learning_rate": 1.0190989553242935e-05, + "loss": 0.2105, + "step": 49539 + }, + { + "epoch": 3.437651793768649, + "grad_norm": 0.8781810402870178, + "learning_rate": 1.0188518563249271e-05, + "loss": 0.27, + "step": 49540 + }, + { + "epoch": 3.4377211852057457, + "grad_norm": 1.174943208694458, + "learning_rate": 1.0186047856780034e-05, + "loss": 0.356, + "step": 49541 + }, + { + "epoch": 3.4377905766428425, + "grad_norm": 0.8912452459335327, + "learning_rate": 1.0183577433843017e-05, + "loss": 0.2586, + "step": 49542 + }, + { + "epoch": 3.437859968079939, + "grad_norm": 1.0710512399673462, + "learning_rate": 1.0181107294446046e-05, + "loss": 0.3609, + "step": 49543 + }, + { + "epoch": 3.4379293595170357, + "grad_norm": 0.6591023206710815, + "learning_rate": 1.0178637438596883e-05, + "loss": 0.223, + "step": 49544 + }, + { + "epoch": 3.437998750954132, + "grad_norm": 1.1183536052703857, + "learning_rate": 1.0176167866303344e-05, + "loss": 0.3463, + "step": 49545 + }, + { + "epoch": 3.438068142391229, + "grad_norm": 1.0585479736328125, + "learning_rate": 1.017369857757322e-05, + "loss": 0.2654, + "step": 49546 + }, + { + "epoch": 3.4381375338283258, + "grad_norm": 0.9446368217468262, + "learning_rate": 1.0171229572414298e-05, + "loss": 0.3435, + "step": 49547 + }, + { + "epoch": 3.438206925265422, + "grad_norm": 0.9076143503189087, + "learning_rate": 1.0168760850834391e-05, + "loss": 0.2773, + "step": 49548 + }, + { + "epoch": 3.438276316702519, + "grad_norm": 0.8519745469093323, + "learning_rate": 1.0166292412841283e-05, + "loss": 0.263, + "step": 49549 + }, + { + "epoch": 3.4383457081396154, + "grad_norm": 1.1523665189743042, + "learning_rate": 1.016382425844279e-05, + "loss": 0.2556, + "step": 49550 + }, + { + "epoch": 3.438415099576712, + "grad_norm": 1.0067338943481445, + "learning_rate": 1.0161356387646658e-05, + "loss": 0.1463, + "step": 49551 + }, + { + "epoch": 3.438484491013809, + "grad_norm": 0.7544053792953491, + "learning_rate": 1.0158888800460698e-05, + "loss": 0.1817, + "step": 49552 + }, + { + "epoch": 3.4385538824509054, + "grad_norm": 0.7704216837882996, + "learning_rate": 1.0156421496892699e-05, + "loss": 0.1904, + "step": 49553 + }, + { + "epoch": 3.4386232738880023, + "grad_norm": 0.7176565527915955, + "learning_rate": 1.0153954476950467e-05, + "loss": 0.2303, + "step": 49554 + }, + { + "epoch": 3.4386926653250987, + "grad_norm": 1.2897179126739502, + "learning_rate": 1.0151487740641762e-05, + "loss": 0.3033, + "step": 49555 + }, + { + "epoch": 3.4387620567621955, + "grad_norm": 0.9941942095756531, + "learning_rate": 1.0149021287974392e-05, + "loss": 0.3047, + "step": 49556 + }, + { + "epoch": 3.4388314481992923, + "grad_norm": 1.046241283416748, + "learning_rate": 1.0146555118956157e-05, + "loss": 0.2833, + "step": 49557 + }, + { + "epoch": 3.4389008396363887, + "grad_norm": 0.9745771884918213, + "learning_rate": 1.014408923359481e-05, + "loss": 0.2707, + "step": 49558 + }, + { + "epoch": 3.4389702310734855, + "grad_norm": 1.0636554956436157, + "learning_rate": 1.0141623631898135e-05, + "loss": 0.3415, + "step": 49559 + }, + { + "epoch": 3.4390396225105824, + "grad_norm": 1.210822343826294, + "learning_rate": 1.0139158313873942e-05, + "loss": 0.2823, + "step": 49560 + }, + { + "epoch": 3.4391090139476788, + "grad_norm": 0.842802107334137, + "learning_rate": 1.0136693279529985e-05, + "loss": 0.2231, + "step": 49561 + }, + { + "epoch": 3.4391784053847756, + "grad_norm": 0.9389598965644836, + "learning_rate": 1.013422852887408e-05, + "loss": 0.26, + "step": 49562 + }, + { + "epoch": 3.4392477968218724, + "grad_norm": 2.128779172897339, + "learning_rate": 1.0131764061913973e-05, + "loss": 0.2359, + "step": 49563 + }, + { + "epoch": 3.439317188258969, + "grad_norm": 0.7210208773612976, + "learning_rate": 1.0129299878657483e-05, + "loss": 0.2023, + "step": 49564 + }, + { + "epoch": 3.4393865796960656, + "grad_norm": 1.1590564250946045, + "learning_rate": 1.0126835979112348e-05, + "loss": 0.2413, + "step": 49565 + }, + { + "epoch": 3.439455971133162, + "grad_norm": 0.7810443043708801, + "learning_rate": 1.0124372363286372e-05, + "loss": 0.2545, + "step": 49566 + }, + { + "epoch": 3.439525362570259, + "grad_norm": 0.6094443202018738, + "learning_rate": 1.0121909031187316e-05, + "loss": 0.2984, + "step": 49567 + }, + { + "epoch": 3.4395947540073557, + "grad_norm": 0.7735592126846313, + "learning_rate": 1.0119445982822972e-05, + "loss": 0.2091, + "step": 49568 + }, + { + "epoch": 3.439664145444452, + "grad_norm": 1.1794953346252441, + "learning_rate": 1.0116983218201103e-05, + "loss": 0.2879, + "step": 49569 + }, + { + "epoch": 3.439733536881549, + "grad_norm": 0.800809919834137, + "learning_rate": 1.0114520737329513e-05, + "loss": 0.2273, + "step": 49570 + }, + { + "epoch": 3.4398029283186453, + "grad_norm": 1.0445332527160645, + "learning_rate": 1.0112058540215929e-05, + "loss": 0.24, + "step": 49571 + }, + { + "epoch": 3.439872319755742, + "grad_norm": 0.8500863313674927, + "learning_rate": 1.0109596626868145e-05, + "loss": 0.2728, + "step": 49572 + }, + { + "epoch": 3.439941711192839, + "grad_norm": 1.0308769941329956, + "learning_rate": 1.0107134997293944e-05, + "loss": 0.3662, + "step": 49573 + }, + { + "epoch": 3.4400111026299354, + "grad_norm": 1.1845605373382568, + "learning_rate": 1.0104673651501107e-05, + "loss": 0.2904, + "step": 49574 + }, + { + "epoch": 3.440080494067032, + "grad_norm": 1.0164549350738525, + "learning_rate": 1.0102212589497361e-05, + "loss": 0.2373, + "step": 49575 + }, + { + "epoch": 3.4401498855041286, + "grad_norm": 0.6826834678649902, + "learning_rate": 1.0099751811290504e-05, + "loss": 0.2833, + "step": 49576 + }, + { + "epoch": 3.4402192769412254, + "grad_norm": 0.8310059309005737, + "learning_rate": 1.0097291316888313e-05, + "loss": 0.1452, + "step": 49577 + }, + { + "epoch": 3.4402886683783223, + "grad_norm": 1.4083449840545654, + "learning_rate": 1.009483110629853e-05, + "loss": 0.3668, + "step": 49578 + }, + { + "epoch": 3.4403580598154186, + "grad_norm": 0.7903534770011902, + "learning_rate": 1.0092371179528925e-05, + "loss": 0.2252, + "step": 49579 + }, + { + "epoch": 3.4404274512525155, + "grad_norm": 0.7005519270896912, + "learning_rate": 1.0089911536587283e-05, + "loss": 0.2103, + "step": 49580 + }, + { + "epoch": 3.440496842689612, + "grad_norm": 1.0013071298599243, + "learning_rate": 1.008745217748136e-05, + "loss": 0.2903, + "step": 49581 + }, + { + "epoch": 3.4405662341267087, + "grad_norm": 0.8922540545463562, + "learning_rate": 1.008499310221891e-05, + "loss": 0.2628, + "step": 49582 + }, + { + "epoch": 3.4406356255638055, + "grad_norm": 0.7979122400283813, + "learning_rate": 1.0082534310807712e-05, + "loss": 0.1842, + "step": 49583 + }, + { + "epoch": 3.440705017000902, + "grad_norm": 1.0716191530227661, + "learning_rate": 1.0080075803255529e-05, + "loss": 0.2765, + "step": 49584 + }, + { + "epoch": 3.4407744084379988, + "grad_norm": 0.6887196898460388, + "learning_rate": 1.0077617579570108e-05, + "loss": 0.1837, + "step": 49585 + }, + { + "epoch": 3.4408437998750956, + "grad_norm": 1.3108488321304321, + "learning_rate": 1.0075159639759214e-05, + "loss": 0.2754, + "step": 49586 + }, + { + "epoch": 3.440913191312192, + "grad_norm": 0.6458554267883301, + "learning_rate": 1.0072701983830602e-05, + "loss": 0.1627, + "step": 49587 + }, + { + "epoch": 3.440982582749289, + "grad_norm": 1.1364599466323853, + "learning_rate": 1.0070244611792034e-05, + "loss": 0.3279, + "step": 49588 + }, + { + "epoch": 3.4410519741863856, + "grad_norm": 0.805115282535553, + "learning_rate": 1.0067787523651274e-05, + "loss": 0.2048, + "step": 49589 + }, + { + "epoch": 3.441121365623482, + "grad_norm": 0.9991111159324646, + "learning_rate": 1.0065330719416077e-05, + "loss": 0.1917, + "step": 49590 + }, + { + "epoch": 3.441190757060579, + "grad_norm": 0.7231420874595642, + "learning_rate": 1.0062874199094207e-05, + "loss": 0.2244, + "step": 49591 + }, + { + "epoch": 3.4412601484976753, + "grad_norm": 1.0099562406539917, + "learning_rate": 1.0060417962693403e-05, + "loss": 0.2626, + "step": 49592 + }, + { + "epoch": 3.441329539934772, + "grad_norm": 0.7939198613166809, + "learning_rate": 1.0057962010221411e-05, + "loss": 0.2086, + "step": 49593 + }, + { + "epoch": 3.441398931371869, + "grad_norm": 0.5993824601173401, + "learning_rate": 1.0055506341686005e-05, + "loss": 0.1374, + "step": 49594 + }, + { + "epoch": 3.4414683228089653, + "grad_norm": 0.7646799683570862, + "learning_rate": 1.0053050957094933e-05, + "loss": 0.2307, + "step": 49595 + }, + { + "epoch": 3.441537714246062, + "grad_norm": 0.8111246824264526, + "learning_rate": 1.0050595856455936e-05, + "loss": 0.2633, + "step": 49596 + }, + { + "epoch": 3.4416071056831585, + "grad_norm": 1.2888637781143188, + "learning_rate": 1.0048141039776782e-05, + "loss": 0.2027, + "step": 49597 + }, + { + "epoch": 3.4416764971202554, + "grad_norm": 0.6921356916427612, + "learning_rate": 1.0045686507065221e-05, + "loss": 0.1229, + "step": 49598 + }, + { + "epoch": 3.441745888557352, + "grad_norm": 0.950969934463501, + "learning_rate": 1.0043232258328972e-05, + "loss": 0.2274, + "step": 49599 + }, + { + "epoch": 3.4418152799944486, + "grad_norm": 0.8323538899421692, + "learning_rate": 1.0040778293575802e-05, + "loss": 0.1867, + "step": 49600 + }, + { + "epoch": 3.4418846714315454, + "grad_norm": 0.8168020844459534, + "learning_rate": 1.0038324612813466e-05, + "loss": 0.275, + "step": 49601 + }, + { + "epoch": 3.441954062868642, + "grad_norm": 1.0084000825881958, + "learning_rate": 1.0035871216049686e-05, + "loss": 0.2134, + "step": 49602 + }, + { + "epoch": 3.4420234543057386, + "grad_norm": 1.0791181325912476, + "learning_rate": 1.0033418103292236e-05, + "loss": 0.4913, + "step": 49603 + }, + { + "epoch": 3.4420928457428355, + "grad_norm": 0.7558085322380066, + "learning_rate": 1.0030965274548842e-05, + "loss": 0.2753, + "step": 49604 + }, + { + "epoch": 3.442162237179932, + "grad_norm": 1.0279793739318848, + "learning_rate": 1.0028512729827277e-05, + "loss": 0.3451, + "step": 49605 + }, + { + "epoch": 3.4422316286170287, + "grad_norm": 0.659984827041626, + "learning_rate": 1.0026060469135235e-05, + "loss": 0.2485, + "step": 49606 + }, + { + "epoch": 3.442301020054125, + "grad_norm": 1.2035901546478271, + "learning_rate": 1.0023608492480473e-05, + "loss": 0.5649, + "step": 49607 + }, + { + "epoch": 3.442370411491222, + "grad_norm": 0.9641889333724976, + "learning_rate": 1.0021156799870757e-05, + "loss": 0.1983, + "step": 49608 + }, + { + "epoch": 3.4424398029283187, + "grad_norm": 1.10298752784729, + "learning_rate": 1.0018705391313809e-05, + "loss": 0.2768, + "step": 49609 + }, + { + "epoch": 3.442509194365415, + "grad_norm": 0.9977391958236694, + "learning_rate": 1.001625426681736e-05, + "loss": 0.3288, + "step": 49610 + }, + { + "epoch": 3.442578585802512, + "grad_norm": 1.5840388536453247, + "learning_rate": 1.0013803426389158e-05, + "loss": 0.351, + "step": 49611 + }, + { + "epoch": 3.442647977239609, + "grad_norm": 0.666562020778656, + "learning_rate": 1.001135287003696e-05, + "loss": 0.2408, + "step": 49612 + }, + { + "epoch": 3.442717368676705, + "grad_norm": 0.7138808965682983, + "learning_rate": 1.0008902597768455e-05, + "loss": 0.2107, + "step": 49613 + }, + { + "epoch": 3.442786760113802, + "grad_norm": 1.0065211057662964, + "learning_rate": 1.00064526095914e-05, + "loss": 0.3608, + "step": 49614 + }, + { + "epoch": 3.442856151550899, + "grad_norm": 0.8402608036994934, + "learning_rate": 1.0004002905513543e-05, + "loss": 0.2648, + "step": 49615 + }, + { + "epoch": 3.4429255429879952, + "grad_norm": 1.128014087677002, + "learning_rate": 1.0001553485542603e-05, + "loss": 0.2349, + "step": 49616 + }, + { + "epoch": 3.442994934425092, + "grad_norm": 0.9747245907783508, + "learning_rate": 9.999104349686317e-06, + "loss": 0.2695, + "step": 49617 + }, + { + "epoch": 3.4430643258621885, + "grad_norm": 1.5896837711334229, + "learning_rate": 9.996655497952412e-06, + "loss": 0.3765, + "step": 49618 + }, + { + "epoch": 3.4431337172992853, + "grad_norm": 0.6696596145629883, + "learning_rate": 9.99420693034865e-06, + "loss": 0.1803, + "step": 49619 + }, + { + "epoch": 3.443203108736382, + "grad_norm": 0.8133335709571838, + "learning_rate": 9.991758646882709e-06, + "loss": 0.2847, + "step": 49620 + }, + { + "epoch": 3.4432725001734785, + "grad_norm": 0.7669590711593628, + "learning_rate": 9.989310647562345e-06, + "loss": 0.1502, + "step": 49621 + }, + { + "epoch": 3.4433418916105754, + "grad_norm": 0.6333447694778442, + "learning_rate": 9.98686293239528e-06, + "loss": 0.1827, + "step": 49622 + }, + { + "epoch": 3.4434112830476717, + "grad_norm": 0.7508252263069153, + "learning_rate": 9.984415501389256e-06, + "loss": 0.244, + "step": 49623 + }, + { + "epoch": 3.4434806744847686, + "grad_norm": 0.7910941243171692, + "learning_rate": 9.981968354551986e-06, + "loss": 0.1961, + "step": 49624 + }, + { + "epoch": 3.4435500659218654, + "grad_norm": 0.6339153051376343, + "learning_rate": 9.979521491891186e-06, + "loss": 0.2348, + "step": 49625 + }, + { + "epoch": 3.443619457358962, + "grad_norm": 0.8328506946563721, + "learning_rate": 9.97707491341462e-06, + "loss": 0.2492, + "step": 49626 + }, + { + "epoch": 3.4436888487960586, + "grad_norm": 0.8487417101860046, + "learning_rate": 9.974628619129967e-06, + "loss": 0.2896, + "step": 49627 + }, + { + "epoch": 3.443758240233155, + "grad_norm": 0.6106002926826477, + "learning_rate": 9.972182609044966e-06, + "loss": 0.1872, + "step": 49628 + }, + { + "epoch": 3.443827631670252, + "grad_norm": 0.8116193413734436, + "learning_rate": 9.969736883167336e-06, + "loss": 0.2568, + "step": 49629 + }, + { + "epoch": 3.4438970231073487, + "grad_norm": 0.8219232559204102, + "learning_rate": 9.9672914415048e-06, + "loss": 0.3338, + "step": 49630 + }, + { + "epoch": 3.443966414544445, + "grad_norm": 0.654873788356781, + "learning_rate": 9.964846284065076e-06, + "loss": 0.1417, + "step": 49631 + }, + { + "epoch": 3.444035805981542, + "grad_norm": 0.813115119934082, + "learning_rate": 9.962401410855893e-06, + "loss": 0.2182, + "step": 49632 + }, + { + "epoch": 3.4441051974186383, + "grad_norm": 1.8695504665374756, + "learning_rate": 9.959956821884985e-06, + "loss": 0.4297, + "step": 49633 + }, + { + "epoch": 3.444174588855735, + "grad_norm": 0.6756591796875, + "learning_rate": 9.957512517160029e-06, + "loss": 0.0993, + "step": 49634 + }, + { + "epoch": 3.444243980292832, + "grad_norm": 0.9917395710945129, + "learning_rate": 9.955068496688746e-06, + "loss": 0.2009, + "step": 49635 + }, + { + "epoch": 3.4443133717299284, + "grad_norm": 0.7708418369293213, + "learning_rate": 9.952624760478868e-06, + "loss": 0.2392, + "step": 49636 + }, + { + "epoch": 3.444382763167025, + "grad_norm": 0.8085978627204895, + "learning_rate": 9.95018130853812e-06, + "loss": 0.1885, + "step": 49637 + }, + { + "epoch": 3.444452154604122, + "grad_norm": 0.7495578527450562, + "learning_rate": 9.947738140874185e-06, + "loss": 0.2184, + "step": 49638 + }, + { + "epoch": 3.4445215460412184, + "grad_norm": 0.8202717900276184, + "learning_rate": 9.945295257494801e-06, + "loss": 0.2472, + "step": 49639 + }, + { + "epoch": 3.4445909374783152, + "grad_norm": 0.9021688103675842, + "learning_rate": 9.942852658407686e-06, + "loss": 0.2337, + "step": 49640 + }, + { + "epoch": 3.444660328915412, + "grad_norm": 1.123265027999878, + "learning_rate": 9.94041034362052e-06, + "loss": 0.4539, + "step": 49641 + }, + { + "epoch": 3.4447297203525085, + "grad_norm": 0.6468656659126282, + "learning_rate": 9.937968313141033e-06, + "loss": 0.2332, + "step": 49642 + }, + { + "epoch": 3.4447991117896053, + "grad_norm": 0.8673118352890015, + "learning_rate": 9.935526566976927e-06, + "loss": 0.2491, + "step": 49643 + }, + { + "epoch": 3.4448685032267017, + "grad_norm": 1.198550820350647, + "learning_rate": 9.933085105135942e-06, + "loss": 0.2298, + "step": 49644 + }, + { + "epoch": 3.4449378946637985, + "grad_norm": 0.8809673190116882, + "learning_rate": 9.930643927625727e-06, + "loss": 0.2675, + "step": 49645 + }, + { + "epoch": 3.4450072861008953, + "grad_norm": 0.9920812845230103, + "learning_rate": 9.92820303445402e-06, + "loss": 0.2439, + "step": 49646 + }, + { + "epoch": 3.4450766775379917, + "grad_norm": 1.00556218624115, + "learning_rate": 9.925762425628549e-06, + "loss": 0.1885, + "step": 49647 + }, + { + "epoch": 3.4451460689750886, + "grad_norm": 1.0166274309158325, + "learning_rate": 9.923322101156985e-06, + "loss": 0.216, + "step": 49648 + }, + { + "epoch": 3.445215460412185, + "grad_norm": 0.9394785165786743, + "learning_rate": 9.920882061047033e-06, + "loss": 0.2404, + "step": 49649 + }, + { + "epoch": 3.445284851849282, + "grad_norm": 1.3620282411575317, + "learning_rate": 9.918442305306397e-06, + "loss": 0.2241, + "step": 49650 + }, + { + "epoch": 3.4453542432863786, + "grad_norm": 0.8560095429420471, + "learning_rate": 9.916002833942795e-06, + "loss": 0.2617, + "step": 49651 + }, + { + "epoch": 3.445423634723475, + "grad_norm": 1.0772007703781128, + "learning_rate": 9.913563646963919e-06, + "loss": 0.2874, + "step": 49652 + }, + { + "epoch": 3.445493026160572, + "grad_norm": 1.033919334411621, + "learning_rate": 9.911124744377499e-06, + "loss": 0.4721, + "step": 49653 + }, + { + "epoch": 3.4455624175976682, + "grad_norm": 0.821022093296051, + "learning_rate": 9.90868612619118e-06, + "loss": 0.2603, + "step": 49654 + }, + { + "epoch": 3.445631809034765, + "grad_norm": 1.0183218717575073, + "learning_rate": 9.906247792412682e-06, + "loss": 0.2311, + "step": 49655 + }, + { + "epoch": 3.445701200471862, + "grad_norm": 1.0900083780288696, + "learning_rate": 9.903809743049708e-06, + "loss": 0.2887, + "step": 49656 + }, + { + "epoch": 3.4457705919089583, + "grad_norm": 0.9759231805801392, + "learning_rate": 9.901371978109953e-06, + "loss": 0.3246, + "step": 49657 + }, + { + "epoch": 3.445839983346055, + "grad_norm": 0.7041773796081543, + "learning_rate": 9.89893449760112e-06, + "loss": 0.1793, + "step": 49658 + }, + { + "epoch": 3.4459093747831515, + "grad_norm": 0.9665282964706421, + "learning_rate": 9.896497301530894e-06, + "loss": 0.3271, + "step": 49659 + }, + { + "epoch": 3.4459787662202483, + "grad_norm": 1.006278395652771, + "learning_rate": 9.89406038990699e-06, + "loss": 0.2391, + "step": 49660 + }, + { + "epoch": 3.446048157657345, + "grad_norm": 0.8753325939178467, + "learning_rate": 9.891623762737068e-06, + "loss": 0.3026, + "step": 49661 + }, + { + "epoch": 3.4461175490944416, + "grad_norm": 0.9071268439292908, + "learning_rate": 9.889187420028834e-06, + "loss": 0.226, + "step": 49662 + }, + { + "epoch": 3.4461869405315384, + "grad_norm": 0.883303165435791, + "learning_rate": 9.88675136178997e-06, + "loss": 0.2872, + "step": 49663 + }, + { + "epoch": 3.4462563319686352, + "grad_norm": 0.5567183494567871, + "learning_rate": 9.884315588028192e-06, + "loss": 0.1256, + "step": 49664 + }, + { + "epoch": 3.4463257234057316, + "grad_norm": 0.8410788178443909, + "learning_rate": 9.881880098751161e-06, + "loss": 0.2128, + "step": 49665 + }, + { + "epoch": 3.4463951148428285, + "grad_norm": 0.8703410029411316, + "learning_rate": 9.879444893966593e-06, + "loss": 0.286, + "step": 49666 + }, + { + "epoch": 3.4464645062799253, + "grad_norm": 0.9110333323478699, + "learning_rate": 9.877009973682173e-06, + "loss": 0.4079, + "step": 49667 + }, + { + "epoch": 3.4465338977170217, + "grad_norm": 1.26655113697052, + "learning_rate": 9.874575337905556e-06, + "loss": 0.255, + "step": 49668 + }, + { + "epoch": 3.4466032891541185, + "grad_norm": 1.2071332931518555, + "learning_rate": 9.87214098664445e-06, + "loss": 0.2804, + "step": 49669 + }, + { + "epoch": 3.446672680591215, + "grad_norm": 0.7544286251068115, + "learning_rate": 9.86970691990654e-06, + "loss": 0.1884, + "step": 49670 + }, + { + "epoch": 3.4467420720283117, + "grad_norm": 0.9173119068145752, + "learning_rate": 9.867273137699506e-06, + "loss": 0.2496, + "step": 49671 + }, + { + "epoch": 3.4468114634654086, + "grad_norm": 1.0404692888259888, + "learning_rate": 9.864839640031032e-06, + "loss": 0.2037, + "step": 49672 + }, + { + "epoch": 3.446880854902505, + "grad_norm": 0.882132351398468, + "learning_rate": 9.862406426908799e-06, + "loss": 0.1981, + "step": 49673 + }, + { + "epoch": 3.446950246339602, + "grad_norm": 0.8195147514343262, + "learning_rate": 9.859973498340513e-06, + "loss": 0.2234, + "step": 49674 + }, + { + "epoch": 3.447019637776698, + "grad_norm": 0.7102932929992676, + "learning_rate": 9.857540854333813e-06, + "loss": 0.2338, + "step": 49675 + }, + { + "epoch": 3.447089029213795, + "grad_norm": 1.3194098472595215, + "learning_rate": 9.855108494896403e-06, + "loss": 0.4027, + "step": 49676 + }, + { + "epoch": 3.447158420650892, + "grad_norm": 0.6874359250068665, + "learning_rate": 9.852676420035955e-06, + "loss": 0.1777, + "step": 49677 + }, + { + "epoch": 3.4472278120879882, + "grad_norm": 1.0336568355560303, + "learning_rate": 9.850244629760163e-06, + "loss": 0.2138, + "step": 49678 + }, + { + "epoch": 3.447297203525085, + "grad_norm": 1.3226364850997925, + "learning_rate": 9.847813124076677e-06, + "loss": 0.2974, + "step": 49679 + }, + { + "epoch": 3.4473665949621815, + "grad_norm": 0.9126654267311096, + "learning_rate": 9.845381902993179e-06, + "loss": 0.2106, + "step": 49680 + }, + { + "epoch": 3.4474359863992783, + "grad_norm": 0.9492445588111877, + "learning_rate": 9.84295096651735e-06, + "loss": 0.2719, + "step": 49681 + }, + { + "epoch": 3.447505377836375, + "grad_norm": 0.7241997718811035, + "learning_rate": 9.840520314656887e-06, + "loss": 0.1513, + "step": 49682 + }, + { + "epoch": 3.4475747692734715, + "grad_norm": 0.9203122854232788, + "learning_rate": 9.838089947419426e-06, + "loss": 0.2736, + "step": 49683 + }, + { + "epoch": 3.4476441607105683, + "grad_norm": 0.96748948097229, + "learning_rate": 9.835659864812653e-06, + "loss": 0.2015, + "step": 49684 + }, + { + "epoch": 3.4477135521476647, + "grad_norm": 0.8048897981643677, + "learning_rate": 9.833230066844235e-06, + "loss": 0.2523, + "step": 49685 + }, + { + "epoch": 3.4477829435847616, + "grad_norm": 0.7553611993789673, + "learning_rate": 9.830800553521846e-06, + "loss": 0.2436, + "step": 49686 + }, + { + "epoch": 3.4478523350218584, + "grad_norm": 1.0142772197723389, + "learning_rate": 9.82837132485317e-06, + "loss": 0.3313, + "step": 49687 + }, + { + "epoch": 3.447921726458955, + "grad_norm": 0.9381904602050781, + "learning_rate": 9.825942380845876e-06, + "loss": 0.2723, + "step": 49688 + }, + { + "epoch": 3.4479911178960516, + "grad_norm": 0.7249301671981812, + "learning_rate": 9.823513721507605e-06, + "loss": 0.2166, + "step": 49689 + }, + { + "epoch": 3.4480605093331484, + "grad_norm": 0.9837833046913147, + "learning_rate": 9.821085346846027e-06, + "loss": 0.3307, + "step": 49690 + }, + { + "epoch": 3.448129900770245, + "grad_norm": 1.1397887468338013, + "learning_rate": 9.818657256868835e-06, + "loss": 0.2946, + "step": 49691 + }, + { + "epoch": 3.4481992922073417, + "grad_norm": 0.8700296878814697, + "learning_rate": 9.81622945158367e-06, + "loss": 0.1797, + "step": 49692 + }, + { + "epoch": 3.4482686836444385, + "grad_norm": 0.8105472326278687, + "learning_rate": 9.8138019309982e-06, + "loss": 0.247, + "step": 49693 + }, + { + "epoch": 3.448338075081535, + "grad_norm": 1.0820612907409668, + "learning_rate": 9.811374695120102e-06, + "loss": 0.4314, + "step": 49694 + }, + { + "epoch": 3.4484074665186317, + "grad_norm": 0.9571723341941833, + "learning_rate": 9.808947743957053e-06, + "loss": 0.2956, + "step": 49695 + }, + { + "epoch": 3.448476857955728, + "grad_norm": 1.1885316371917725, + "learning_rate": 9.806521077516661e-06, + "loss": 0.3905, + "step": 49696 + }, + { + "epoch": 3.448546249392825, + "grad_norm": 0.9654516577720642, + "learning_rate": 9.804094695806631e-06, + "loss": 0.2576, + "step": 49697 + }, + { + "epoch": 3.448615640829922, + "grad_norm": 1.0279546976089478, + "learning_rate": 9.801668598834601e-06, + "loss": 0.3117, + "step": 49698 + }, + { + "epoch": 3.448685032267018, + "grad_norm": 1.3755260705947876, + "learning_rate": 9.799242786608242e-06, + "loss": 0.2617, + "step": 49699 + }, + { + "epoch": 3.448754423704115, + "grad_norm": 1.2266826629638672, + "learning_rate": 9.796817259135204e-06, + "loss": 0.1874, + "step": 49700 + }, + { + "epoch": 3.4488238151412114, + "grad_norm": 1.13010835647583, + "learning_rate": 9.794392016423148e-06, + "loss": 0.2819, + "step": 49701 + }, + { + "epoch": 3.4488932065783082, + "grad_norm": 1.048444390296936, + "learning_rate": 9.791967058479756e-06, + "loss": 0.3823, + "step": 49702 + }, + { + "epoch": 3.448962598015405, + "grad_norm": 0.7024252414703369, + "learning_rate": 9.789542385312633e-06, + "loss": 0.2785, + "step": 49703 + }, + { + "epoch": 3.4490319894525014, + "grad_norm": 1.256156086921692, + "learning_rate": 9.787117996929462e-06, + "loss": 0.2673, + "step": 49704 + }, + { + "epoch": 3.4491013808895983, + "grad_norm": 1.2523789405822754, + "learning_rate": 9.784693893337893e-06, + "loss": 0.2858, + "step": 49705 + }, + { + "epoch": 3.4491707723266947, + "grad_norm": 0.807529866695404, + "learning_rate": 9.782270074545575e-06, + "loss": 0.2505, + "step": 49706 + }, + { + "epoch": 3.4492401637637915, + "grad_norm": 0.8404669761657715, + "learning_rate": 9.779846540560167e-06, + "loss": 0.2651, + "step": 49707 + }, + { + "epoch": 3.4493095552008883, + "grad_norm": 1.0891107320785522, + "learning_rate": 9.777423291389309e-06, + "loss": 0.344, + "step": 49708 + }, + { + "epoch": 3.4493789466379847, + "grad_norm": 1.1544184684753418, + "learning_rate": 9.775000327040683e-06, + "loss": 0.3266, + "step": 49709 + }, + { + "epoch": 3.4494483380750816, + "grad_norm": 1.810049295425415, + "learning_rate": 9.772577647521896e-06, + "loss": 0.4193, + "step": 49710 + }, + { + "epoch": 3.4495177295121784, + "grad_norm": 1.1344661712646484, + "learning_rate": 9.770155252840607e-06, + "loss": 0.3401, + "step": 49711 + }, + { + "epoch": 3.4495871209492748, + "grad_norm": 0.8068015575408936, + "learning_rate": 9.767733143004476e-06, + "loss": 0.2377, + "step": 49712 + }, + { + "epoch": 3.4496565123863716, + "grad_norm": 0.8719074726104736, + "learning_rate": 9.765311318021164e-06, + "loss": 0.1522, + "step": 49713 + }, + { + "epoch": 3.449725903823468, + "grad_norm": 1.1793813705444336, + "learning_rate": 9.762889777898265e-06, + "loss": 0.2832, + "step": 49714 + }, + { + "epoch": 3.449795295260565, + "grad_norm": 1.2291477918624878, + "learning_rate": 9.760468522643462e-06, + "loss": 0.3061, + "step": 49715 + }, + { + "epoch": 3.4498646866976617, + "grad_norm": 0.6158174276351929, + "learning_rate": 9.758047552264383e-06, + "loss": 0.2256, + "step": 49716 + }, + { + "epoch": 3.449934078134758, + "grad_norm": 0.9465678930282593, + "learning_rate": 9.755626866768696e-06, + "loss": 0.2218, + "step": 49717 + }, + { + "epoch": 3.450003469571855, + "grad_norm": 1.1385085582733154, + "learning_rate": 9.753206466164011e-06, + "loss": 0.3223, + "step": 49718 + }, + { + "epoch": 3.4500728610089517, + "grad_norm": 0.9878481030464172, + "learning_rate": 9.750786350457986e-06, + "loss": 0.3176, + "step": 49719 + }, + { + "epoch": 3.450142252446048, + "grad_norm": 0.7278270125389099, + "learning_rate": 9.748366519658259e-06, + "loss": 0.2473, + "step": 49720 + }, + { + "epoch": 3.450211643883145, + "grad_norm": 1.084905982017517, + "learning_rate": 9.745946973772458e-06, + "loss": 0.3632, + "step": 49721 + }, + { + "epoch": 3.4502810353202413, + "grad_norm": 0.8939449191093445, + "learning_rate": 9.743527712808254e-06, + "loss": 0.2825, + "step": 49722 + }, + { + "epoch": 3.450350426757338, + "grad_norm": 0.9604856371879578, + "learning_rate": 9.741108736773241e-06, + "loss": 0.2479, + "step": 49723 + }, + { + "epoch": 3.450419818194435, + "grad_norm": 0.8273600935935974, + "learning_rate": 9.73869004567508e-06, + "loss": 0.2451, + "step": 49724 + }, + { + "epoch": 3.4504892096315314, + "grad_norm": 0.909137487411499, + "learning_rate": 9.7362716395214e-06, + "loss": 0.2169, + "step": 49725 + }, + { + "epoch": 3.450558601068628, + "grad_norm": 0.8189495801925659, + "learning_rate": 9.733853518319835e-06, + "loss": 0.1023, + "step": 49726 + }, + { + "epoch": 3.4506279925057246, + "grad_norm": 0.9668214321136475, + "learning_rate": 9.731435682078028e-06, + "loss": 0.2431, + "step": 49727 + }, + { + "epoch": 3.4506973839428214, + "grad_norm": 1.2362630367279053, + "learning_rate": 9.729018130803602e-06, + "loss": 0.3507, + "step": 49728 + }, + { + "epoch": 3.4507667753799183, + "grad_norm": 0.7456159591674805, + "learning_rate": 9.726600864504209e-06, + "loss": 0.2521, + "step": 49729 + }, + { + "epoch": 3.4508361668170147, + "grad_norm": 0.8820549249649048, + "learning_rate": 9.724183883187443e-06, + "loss": 0.2151, + "step": 49730 + }, + { + "epoch": 3.4509055582541115, + "grad_norm": 0.8994966149330139, + "learning_rate": 9.721767186860964e-06, + "loss": 0.2976, + "step": 49731 + }, + { + "epoch": 3.450974949691208, + "grad_norm": 1.871720314025879, + "learning_rate": 9.719350775532376e-06, + "loss": 0.3385, + "step": 49732 + }, + { + "epoch": 3.4510443411283047, + "grad_norm": 1.2457308769226074, + "learning_rate": 9.716934649209342e-06, + "loss": 0.2659, + "step": 49733 + }, + { + "epoch": 3.4511137325654015, + "grad_norm": 0.8158884048461914, + "learning_rate": 9.714518807899453e-06, + "loss": 0.3371, + "step": 49734 + }, + { + "epoch": 3.451183124002498, + "grad_norm": 0.6374475955963135, + "learning_rate": 9.712103251610371e-06, + "loss": 0.1664, + "step": 49735 + }, + { + "epoch": 3.4512525154395948, + "grad_norm": 1.0786960124969482, + "learning_rate": 9.709687980349713e-06, + "loss": 0.3107, + "step": 49736 + }, + { + "epoch": 3.4513219068766916, + "grad_norm": 0.8598957657814026, + "learning_rate": 9.707272994125071e-06, + "loss": 0.1839, + "step": 49737 + }, + { + "epoch": 3.451391298313788, + "grad_norm": 0.9819722175598145, + "learning_rate": 9.704858292944096e-06, + "loss": 0.217, + "step": 49738 + }, + { + "epoch": 3.451460689750885, + "grad_norm": 1.3485743999481201, + "learning_rate": 9.702443876814404e-06, + "loss": 0.322, + "step": 49739 + }, + { + "epoch": 3.451530081187981, + "grad_norm": 0.8643184900283813, + "learning_rate": 9.700029745743632e-06, + "loss": 0.2445, + "step": 49740 + }, + { + "epoch": 3.451599472625078, + "grad_norm": 0.8950276374816895, + "learning_rate": 9.697615899739377e-06, + "loss": 0.2463, + "step": 49741 + }, + { + "epoch": 3.451668864062175, + "grad_norm": 1.2672401666641235, + "learning_rate": 9.695202338809273e-06, + "loss": 0.2072, + "step": 49742 + }, + { + "epoch": 3.4517382554992713, + "grad_norm": 0.6712360978126526, + "learning_rate": 9.692789062960962e-06, + "loss": 0.189, + "step": 49743 + }, + { + "epoch": 3.451807646936368, + "grad_norm": 0.7059313058853149, + "learning_rate": 9.690376072202022e-06, + "loss": 0.1528, + "step": 49744 + }, + { + "epoch": 3.451877038373465, + "grad_norm": 0.8635037541389465, + "learning_rate": 9.687963366540076e-06, + "loss": 0.2956, + "step": 49745 + }, + { + "epoch": 3.4519464298105613, + "grad_norm": 0.7932635545730591, + "learning_rate": 9.685550945982758e-06, + "loss": 0.2565, + "step": 49746 + }, + { + "epoch": 3.452015821247658, + "grad_norm": 0.9210079908370972, + "learning_rate": 9.683138810537683e-06, + "loss": 0.2747, + "step": 49747 + }, + { + "epoch": 3.4520852126847545, + "grad_norm": 1.162268877029419, + "learning_rate": 9.680726960212471e-06, + "loss": 0.2895, + "step": 49748 + }, + { + "epoch": 3.4521546041218514, + "grad_norm": 1.2492884397506714, + "learning_rate": 9.678315395014714e-06, + "loss": 0.1563, + "step": 49749 + }, + { + "epoch": 3.452223995558948, + "grad_norm": 0.7467436194419861, + "learning_rate": 9.675904114952028e-06, + "loss": 0.2402, + "step": 49750 + }, + { + "epoch": 3.4522933869960446, + "grad_norm": 1.6283392906188965, + "learning_rate": 9.673493120032028e-06, + "loss": 0.396, + "step": 49751 + }, + { + "epoch": 3.4523627784331414, + "grad_norm": 0.7567373514175415, + "learning_rate": 9.671082410262367e-06, + "loss": 0.2067, + "step": 49752 + }, + { + "epoch": 3.452432169870238, + "grad_norm": 0.9827160239219666, + "learning_rate": 9.66867198565059e-06, + "loss": 0.2108, + "step": 49753 + }, + { + "epoch": 3.4525015613073347, + "grad_norm": 0.7456661462783813, + "learning_rate": 9.666261846204328e-06, + "loss": 0.3117, + "step": 49754 + }, + { + "epoch": 3.4525709527444315, + "grad_norm": 0.950980544090271, + "learning_rate": 9.663851991931206e-06, + "loss": 0.1976, + "step": 49755 + }, + { + "epoch": 3.452640344181528, + "grad_norm": 0.9177631735801697, + "learning_rate": 9.661442422838817e-06, + "loss": 0.3132, + "step": 49756 + }, + { + "epoch": 3.4527097356186247, + "grad_norm": 1.1151753664016724, + "learning_rate": 9.659033138934791e-06, + "loss": 0.4037, + "step": 49757 + }, + { + "epoch": 3.452779127055721, + "grad_norm": 1.028493881225586, + "learning_rate": 9.656624140226712e-06, + "loss": 0.3155, + "step": 49758 + }, + { + "epoch": 3.452848518492818, + "grad_norm": 0.5758605599403381, + "learning_rate": 9.654215426722169e-06, + "loss": 0.1631, + "step": 49759 + }, + { + "epoch": 3.4529179099299148, + "grad_norm": 0.9112266898155212, + "learning_rate": 9.651806998428803e-06, + "loss": 0.2652, + "step": 49760 + }, + { + "epoch": 3.452987301367011, + "grad_norm": 1.078818917274475, + "learning_rate": 9.649398855354186e-06, + "loss": 0.44, + "step": 49761 + }, + { + "epoch": 3.453056692804108, + "grad_norm": 0.7288292646408081, + "learning_rate": 9.646990997505945e-06, + "loss": 0.1749, + "step": 49762 + }, + { + "epoch": 3.453126084241205, + "grad_norm": 1.1432305574417114, + "learning_rate": 9.644583424891661e-06, + "loss": 0.2948, + "step": 49763 + }, + { + "epoch": 3.453195475678301, + "grad_norm": 1.0636018514633179, + "learning_rate": 9.642176137518965e-06, + "loss": 0.2695, + "step": 49764 + }, + { + "epoch": 3.453264867115398, + "grad_norm": 0.8749802708625793, + "learning_rate": 9.639769135395415e-06, + "loss": 0.2588, + "step": 49765 + }, + { + "epoch": 3.4533342585524944, + "grad_norm": 0.7229369282722473, + "learning_rate": 9.63736241852864e-06, + "loss": 0.25, + "step": 49766 + }, + { + "epoch": 3.4534036499895913, + "grad_norm": 1.276063323020935, + "learning_rate": 9.634955986926219e-06, + "loss": 0.303, + "step": 49767 + }, + { + "epoch": 3.453473041426688, + "grad_norm": 0.8043195605278015, + "learning_rate": 9.632549840595762e-06, + "loss": 0.2503, + "step": 49768 + }, + { + "epoch": 3.4535424328637845, + "grad_norm": 0.8648539185523987, + "learning_rate": 9.630143979544859e-06, + "loss": 0.2708, + "step": 49769 + }, + { + "epoch": 3.4536118243008813, + "grad_norm": 0.7675965428352356, + "learning_rate": 9.627738403781106e-06, + "loss": 0.205, + "step": 49770 + }, + { + "epoch": 3.453681215737978, + "grad_norm": 0.6966642737388611, + "learning_rate": 9.625333113312118e-06, + "loss": 0.1917, + "step": 49771 + }, + { + "epoch": 3.4537506071750745, + "grad_norm": 0.785161554813385, + "learning_rate": 9.622928108145445e-06, + "loss": 0.1598, + "step": 49772 + }, + { + "epoch": 3.4538199986121714, + "grad_norm": 0.9446335434913635, + "learning_rate": 9.620523388288705e-06, + "loss": 0.4249, + "step": 49773 + }, + { + "epoch": 3.4538893900492678, + "grad_norm": 0.730541467666626, + "learning_rate": 9.61811895374949e-06, + "loss": 0.1396, + "step": 49774 + }, + { + "epoch": 3.4539587814863646, + "grad_norm": 1.2417737245559692, + "learning_rate": 9.61571480453538e-06, + "loss": 0.3022, + "step": 49775 + }, + { + "epoch": 3.4540281729234614, + "grad_norm": 1.0379700660705566, + "learning_rate": 9.613310940653975e-06, + "loss": 0.324, + "step": 49776 + }, + { + "epoch": 3.454097564360558, + "grad_norm": 1.1347312927246094, + "learning_rate": 9.610907362112864e-06, + "loss": 0.3684, + "step": 49777 + }, + { + "epoch": 3.4541669557976546, + "grad_norm": 0.8327099084854126, + "learning_rate": 9.608504068919655e-06, + "loss": 0.2168, + "step": 49778 + }, + { + "epoch": 3.454236347234751, + "grad_norm": 0.9554723501205444, + "learning_rate": 9.606101061081884e-06, + "loss": 0.3921, + "step": 49779 + }, + { + "epoch": 3.454305738671848, + "grad_norm": 0.8915382623672485, + "learning_rate": 9.603698338607159e-06, + "loss": 0.2483, + "step": 49780 + }, + { + "epoch": 3.4543751301089447, + "grad_norm": 0.7217444777488708, + "learning_rate": 9.601295901503082e-06, + "loss": 0.2222, + "step": 49781 + }, + { + "epoch": 3.454444521546041, + "grad_norm": 1.0747301578521729, + "learning_rate": 9.598893749777215e-06, + "loss": 0.2345, + "step": 49782 + }, + { + "epoch": 3.454513912983138, + "grad_norm": 1.0576781034469604, + "learning_rate": 9.596491883437186e-06, + "loss": 0.1923, + "step": 49783 + }, + { + "epoch": 3.4545833044202343, + "grad_norm": 1.2576018571853638, + "learning_rate": 9.59409030249051e-06, + "loss": 0.4155, + "step": 49784 + }, + { + "epoch": 3.454652695857331, + "grad_norm": 1.4780157804489136, + "learning_rate": 9.591689006944804e-06, + "loss": 0.3054, + "step": 49785 + }, + { + "epoch": 3.454722087294428, + "grad_norm": 1.0796130895614624, + "learning_rate": 9.58928799680765e-06, + "loss": 0.2452, + "step": 49786 + }, + { + "epoch": 3.4547914787315244, + "grad_norm": 0.8341469168663025, + "learning_rate": 9.586887272086641e-06, + "loss": 0.2048, + "step": 49787 + }, + { + "epoch": 3.454860870168621, + "grad_norm": 1.0242360830307007, + "learning_rate": 9.58448683278933e-06, + "loss": 0.2733, + "step": 49788 + }, + { + "epoch": 3.454930261605718, + "grad_norm": 0.9722371101379395, + "learning_rate": 9.582086678923297e-06, + "loss": 0.2923, + "step": 49789 + }, + { + "epoch": 3.4549996530428144, + "grad_norm": 0.8318002820014954, + "learning_rate": 9.579686810496124e-06, + "loss": 0.2513, + "step": 49790 + }, + { + "epoch": 3.4550690444799113, + "grad_norm": 1.0502320528030396, + "learning_rate": 9.577287227515386e-06, + "loss": 0.4302, + "step": 49791 + }, + { + "epoch": 3.455138435917008, + "grad_norm": 0.8636131882667542, + "learning_rate": 9.574887929988686e-06, + "loss": 0.2844, + "step": 49792 + }, + { + "epoch": 3.4552078273541045, + "grad_norm": 0.9058598875999451, + "learning_rate": 9.57248891792355e-06, + "loss": 0.3888, + "step": 49793 + }, + { + "epoch": 3.4552772187912013, + "grad_norm": 0.8792263269424438, + "learning_rate": 9.570090191327574e-06, + "loss": 0.2382, + "step": 49794 + }, + { + "epoch": 3.4553466102282977, + "grad_norm": 0.9506641626358032, + "learning_rate": 9.56769175020834e-06, + "loss": 0.2031, + "step": 49795 + }, + { + "epoch": 3.4554160016653945, + "grad_norm": 1.0530078411102295, + "learning_rate": 9.565293594573399e-06, + "loss": 0.2394, + "step": 49796 + }, + { + "epoch": 3.4554853931024914, + "grad_norm": 1.0363248586654663, + "learning_rate": 9.562895724430343e-06, + "loss": 0.2554, + "step": 49797 + }, + { + "epoch": 3.4555547845395878, + "grad_norm": 0.9278787970542908, + "learning_rate": 9.560498139786756e-06, + "loss": 0.2475, + "step": 49798 + }, + { + "epoch": 3.4556241759766846, + "grad_norm": 0.9787781834602356, + "learning_rate": 9.558100840650153e-06, + "loss": 0.2451, + "step": 49799 + }, + { + "epoch": 3.455693567413781, + "grad_norm": 0.9843927621841431, + "learning_rate": 9.55570382702814e-06, + "loss": 0.2434, + "step": 49800 + }, + { + "epoch": 3.455762958850878, + "grad_norm": 1.0113584995269775, + "learning_rate": 9.553307098928266e-06, + "loss": 0.3105, + "step": 49801 + }, + { + "epoch": 3.4558323502879746, + "grad_norm": 0.9287357926368713, + "learning_rate": 9.550910656358126e-06, + "loss": 0.2357, + "step": 49802 + }, + { + "epoch": 3.455901741725071, + "grad_norm": 1.0175281763076782, + "learning_rate": 9.548514499325245e-06, + "loss": 0.2748, + "step": 49803 + }, + { + "epoch": 3.455971133162168, + "grad_norm": 0.8105982542037964, + "learning_rate": 9.54611862783723e-06, + "loss": 0.2225, + "step": 49804 + }, + { + "epoch": 3.4560405245992643, + "grad_norm": 0.9750003218650818, + "learning_rate": 9.543723041901632e-06, + "loss": 0.3217, + "step": 49805 + }, + { + "epoch": 3.456109916036361, + "grad_norm": 0.8577063083648682, + "learning_rate": 9.541327741525985e-06, + "loss": 0.2572, + "step": 49806 + }, + { + "epoch": 3.456179307473458, + "grad_norm": 0.5815478563308716, + "learning_rate": 9.538932726717875e-06, + "loss": 0.1626, + "step": 49807 + }, + { + "epoch": 3.4562486989105543, + "grad_norm": 0.9843627214431763, + "learning_rate": 9.536537997484863e-06, + "loss": 0.2313, + "step": 49808 + }, + { + "epoch": 3.456318090347651, + "grad_norm": 1.0629929304122925, + "learning_rate": 9.534143553834497e-06, + "loss": 0.2617, + "step": 49809 + }, + { + "epoch": 3.4563874817847475, + "grad_norm": 0.9748387932777405, + "learning_rate": 9.531749395774347e-06, + "loss": 0.235, + "step": 49810 + }, + { + "epoch": 3.4564568732218444, + "grad_norm": 1.2355669736862183, + "learning_rate": 9.529355523311967e-06, + "loss": 0.245, + "step": 49811 + }, + { + "epoch": 3.456526264658941, + "grad_norm": 0.8136806488037109, + "learning_rate": 9.526961936454936e-06, + "loss": 0.1756, + "step": 49812 + }, + { + "epoch": 3.4565956560960376, + "grad_norm": 0.9211755394935608, + "learning_rate": 9.524568635210774e-06, + "loss": 0.1985, + "step": 49813 + }, + { + "epoch": 3.4566650475331344, + "grad_norm": 0.6812688708305359, + "learning_rate": 9.522175619587037e-06, + "loss": 0.2666, + "step": 49814 + }, + { + "epoch": 3.4567344389702312, + "grad_norm": 0.9875174164772034, + "learning_rate": 9.51978288959131e-06, + "loss": 0.2065, + "step": 49815 + }, + { + "epoch": 3.4568038304073276, + "grad_norm": 1.441338062286377, + "learning_rate": 9.517390445231123e-06, + "loss": 0.419, + "step": 49816 + }, + { + "epoch": 3.4568732218444245, + "grad_norm": 0.9251071214675903, + "learning_rate": 9.514998286514043e-06, + "loss": 0.2543, + "step": 49817 + }, + { + "epoch": 3.4569426132815213, + "grad_norm": 1.1428724527359009, + "learning_rate": 9.512606413447623e-06, + "loss": 0.3736, + "step": 49818 + }, + { + "epoch": 3.4570120047186177, + "grad_norm": 0.8195733428001404, + "learning_rate": 9.510214826039399e-06, + "loss": 0.2615, + "step": 49819 + }, + { + "epoch": 3.4570813961557145, + "grad_norm": 0.9637439250946045, + "learning_rate": 9.507823524296922e-06, + "loss": 0.2726, + "step": 49820 + }, + { + "epoch": 3.457150787592811, + "grad_norm": 0.9514171481132507, + "learning_rate": 9.505432508227762e-06, + "loss": 0.2272, + "step": 49821 + }, + { + "epoch": 3.4572201790299077, + "grad_norm": 0.9982649683952332, + "learning_rate": 9.503041777839438e-06, + "loss": 0.2107, + "step": 49822 + }, + { + "epoch": 3.4572895704670046, + "grad_norm": 0.7417208552360535, + "learning_rate": 9.50065133313951e-06, + "loss": 0.1536, + "step": 49823 + }, + { + "epoch": 3.457358961904101, + "grad_norm": 0.8833249807357788, + "learning_rate": 9.498261174135537e-06, + "loss": 0.2289, + "step": 49824 + }, + { + "epoch": 3.457428353341198, + "grad_norm": 1.658791422843933, + "learning_rate": 9.495871300835036e-06, + "loss": 0.2471, + "step": 49825 + }, + { + "epoch": 3.457497744778294, + "grad_norm": 0.796461284160614, + "learning_rate": 9.4934817132456e-06, + "loss": 0.2298, + "step": 49826 + }, + { + "epoch": 3.457567136215391, + "grad_norm": 0.8546710014343262, + "learning_rate": 9.491092411374713e-06, + "loss": 0.1936, + "step": 49827 + }, + { + "epoch": 3.457636527652488, + "grad_norm": 0.8736621141433716, + "learning_rate": 9.488703395229958e-06, + "loss": 0.2429, + "step": 49828 + }, + { + "epoch": 3.4577059190895842, + "grad_norm": 1.1323950290679932, + "learning_rate": 9.486314664818851e-06, + "loss": 0.3377, + "step": 49829 + }, + { + "epoch": 3.457775310526681, + "grad_norm": 1.6098207235336304, + "learning_rate": 9.483926220148964e-06, + "loss": 0.2621, + "step": 49830 + }, + { + "epoch": 3.4578447019637775, + "grad_norm": 0.7776321768760681, + "learning_rate": 9.481538061227802e-06, + "loss": 0.2112, + "step": 49831 + }, + { + "epoch": 3.4579140934008743, + "grad_norm": 0.7369669079780579, + "learning_rate": 9.479150188062935e-06, + "loss": 0.2301, + "step": 49832 + }, + { + "epoch": 3.457983484837971, + "grad_norm": 1.117943525314331, + "learning_rate": 9.476762600661903e-06, + "loss": 0.2238, + "step": 49833 + }, + { + "epoch": 3.4580528762750675, + "grad_norm": 0.6957017779350281, + "learning_rate": 9.474375299032212e-06, + "loss": 0.3178, + "step": 49834 + }, + { + "epoch": 3.4581222677121644, + "grad_norm": 0.8096706867218018, + "learning_rate": 9.471988283181409e-06, + "loss": 0.2737, + "step": 49835 + }, + { + "epoch": 3.4581916591492607, + "grad_norm": 1.2303367853164673, + "learning_rate": 9.469601553117047e-06, + "loss": 0.3668, + "step": 49836 + }, + { + "epoch": 3.4582610505863576, + "grad_norm": 0.6628034710884094, + "learning_rate": 9.467215108846638e-06, + "loss": 0.1934, + "step": 49837 + }, + { + "epoch": 3.4583304420234544, + "grad_norm": 1.4976863861083984, + "learning_rate": 9.464828950377736e-06, + "loss": 0.2905, + "step": 49838 + }, + { + "epoch": 3.458399833460551, + "grad_norm": 0.91181480884552, + "learning_rate": 9.462443077717864e-06, + "loss": 0.2465, + "step": 49839 + }, + { + "epoch": 3.4584692248976476, + "grad_norm": 0.8895658254623413, + "learning_rate": 9.460057490874563e-06, + "loss": 0.2159, + "step": 49840 + }, + { + "epoch": 3.4585386163347445, + "grad_norm": 0.7219280004501343, + "learning_rate": 9.457672189855348e-06, + "loss": 0.1966, + "step": 49841 + }, + { + "epoch": 3.458608007771841, + "grad_norm": 1.0323649644851685, + "learning_rate": 9.45528717466776e-06, + "loss": 0.3628, + "step": 49842 + }, + { + "epoch": 3.4586773992089377, + "grad_norm": 0.9118618965148926, + "learning_rate": 9.452902445319311e-06, + "loss": 0.2615, + "step": 49843 + }, + { + "epoch": 3.4587467906460345, + "grad_norm": 0.8434988260269165, + "learning_rate": 9.450518001817554e-06, + "loss": 0.2844, + "step": 49844 + }, + { + "epoch": 3.458816182083131, + "grad_norm": 1.210363745689392, + "learning_rate": 9.448133844170003e-06, + "loss": 0.3036, + "step": 49845 + }, + { + "epoch": 3.4588855735202277, + "grad_norm": 0.8903262615203857, + "learning_rate": 9.445749972384188e-06, + "loss": 0.2406, + "step": 49846 + }, + { + "epoch": 3.458954964957324, + "grad_norm": 0.7875683307647705, + "learning_rate": 9.443366386467656e-06, + "loss": 0.232, + "step": 49847 + }, + { + "epoch": 3.459024356394421, + "grad_norm": 0.94517582654953, + "learning_rate": 9.440983086427891e-06, + "loss": 0.2681, + "step": 49848 + }, + { + "epoch": 3.459093747831518, + "grad_norm": 1.1156249046325684, + "learning_rate": 9.43860007227243e-06, + "loss": 0.2964, + "step": 49849 + }, + { + "epoch": 3.459163139268614, + "grad_norm": 1.2508327960968018, + "learning_rate": 9.436217344008802e-06, + "loss": 0.3861, + "step": 49850 + }, + { + "epoch": 3.459232530705711, + "grad_norm": 0.8405372500419617, + "learning_rate": 9.433834901644533e-06, + "loss": 0.1693, + "step": 49851 + }, + { + "epoch": 3.4593019221428074, + "grad_norm": 1.2802842855453491, + "learning_rate": 9.43145274518713e-06, + "loss": 0.3707, + "step": 49852 + }, + { + "epoch": 3.4593713135799042, + "grad_norm": 0.8846372961997986, + "learning_rate": 9.429070874644152e-06, + "loss": 0.2782, + "step": 49853 + }, + { + "epoch": 3.459440705017001, + "grad_norm": 1.229680061340332, + "learning_rate": 9.426689290023072e-06, + "loss": 0.3009, + "step": 49854 + }, + { + "epoch": 3.4595100964540975, + "grad_norm": 1.0036720037460327, + "learning_rate": 9.424307991331416e-06, + "loss": 0.3553, + "step": 49855 + }, + { + "epoch": 3.4595794878911943, + "grad_norm": 1.2271251678466797, + "learning_rate": 9.421926978576733e-06, + "loss": 0.386, + "step": 49856 + }, + { + "epoch": 3.4596488793282907, + "grad_norm": 1.111006259918213, + "learning_rate": 9.41954625176651e-06, + "loss": 0.3291, + "step": 49857 + }, + { + "epoch": 3.4597182707653875, + "grad_norm": 1.364454984664917, + "learning_rate": 9.417165810908257e-06, + "loss": 0.2767, + "step": 49858 + }, + { + "epoch": 3.4597876622024843, + "grad_norm": 4.938511848449707, + "learning_rate": 9.414785656009507e-06, + "loss": 0.2569, + "step": 49859 + }, + { + "epoch": 3.4598570536395807, + "grad_norm": 0.9582135677337646, + "learning_rate": 9.41240578707776e-06, + "loss": 0.2403, + "step": 49860 + }, + { + "epoch": 3.4599264450766776, + "grad_norm": 0.7202541828155518, + "learning_rate": 9.41002620412057e-06, + "loss": 0.1782, + "step": 49861 + }, + { + "epoch": 3.459995836513774, + "grad_norm": 1.1027907133102417, + "learning_rate": 9.407646907145385e-06, + "loss": 0.2933, + "step": 49862 + }, + { + "epoch": 3.460065227950871, + "grad_norm": 0.7840169072151184, + "learning_rate": 9.405267896159765e-06, + "loss": 0.2097, + "step": 49863 + }, + { + "epoch": 3.4601346193879676, + "grad_norm": 0.8743595480918884, + "learning_rate": 9.402889171171192e-06, + "loss": 0.1825, + "step": 49864 + }, + { + "epoch": 3.460204010825064, + "grad_norm": 0.9331600666046143, + "learning_rate": 9.400510732187184e-06, + "loss": 0.2899, + "step": 49865 + }, + { + "epoch": 3.460273402262161, + "grad_norm": 1.1113206148147583, + "learning_rate": 9.398132579215258e-06, + "loss": 0.1861, + "step": 49866 + }, + { + "epoch": 3.4603427936992577, + "grad_norm": 0.8091443181037903, + "learning_rate": 9.395754712262916e-06, + "loss": 0.2348, + "step": 49867 + }, + { + "epoch": 3.460412185136354, + "grad_norm": 0.6339735984802246, + "learning_rate": 9.393377131337677e-06, + "loss": 0.2296, + "step": 49868 + }, + { + "epoch": 3.460481576573451, + "grad_norm": 1.1307185888290405, + "learning_rate": 9.390999836447024e-06, + "loss": 0.3204, + "step": 49869 + }, + { + "epoch": 3.4605509680105477, + "grad_norm": 0.5863255858421326, + "learning_rate": 9.38862282759846e-06, + "loss": 0.1996, + "step": 49870 + }, + { + "epoch": 3.460620359447644, + "grad_norm": 1.0292401313781738, + "learning_rate": 9.386246104799511e-06, + "loss": 0.2281, + "step": 49871 + }, + { + "epoch": 3.460689750884741, + "grad_norm": 0.8395382761955261, + "learning_rate": 9.383869668057666e-06, + "loss": 0.2087, + "step": 49872 + }, + { + "epoch": 3.4607591423218373, + "grad_norm": 1.3166090250015259, + "learning_rate": 9.381493517380436e-06, + "loss": 0.3468, + "step": 49873 + }, + { + "epoch": 3.460828533758934, + "grad_norm": 1.097583532333374, + "learning_rate": 9.379117652775327e-06, + "loss": 0.4081, + "step": 49874 + }, + { + "epoch": 3.460897925196031, + "grad_norm": 1.170095443725586, + "learning_rate": 9.376742074249812e-06, + "loss": 0.2897, + "step": 49875 + }, + { + "epoch": 3.4609673166331274, + "grad_norm": 1.306835412979126, + "learning_rate": 9.374366781811417e-06, + "loss": 0.3168, + "step": 49876 + }, + { + "epoch": 3.4610367080702242, + "grad_norm": 1.0134085416793823, + "learning_rate": 9.371991775467626e-06, + "loss": 0.2154, + "step": 49877 + }, + { + "epoch": 3.4611060995073206, + "grad_norm": 1.2137213945388794, + "learning_rate": 9.369617055225942e-06, + "loss": 0.3723, + "step": 49878 + }, + { + "epoch": 3.4611754909444175, + "grad_norm": 0.9739444255828857, + "learning_rate": 9.367242621093863e-06, + "loss": 0.3114, + "step": 49879 + }, + { + "epoch": 3.4612448823815143, + "grad_norm": 1.096418023109436, + "learning_rate": 9.36486847307888e-06, + "loss": 0.2405, + "step": 49880 + }, + { + "epoch": 3.4613142738186107, + "grad_norm": 0.9056729674339294, + "learning_rate": 9.362494611188511e-06, + "loss": 0.2772, + "step": 49881 + }, + { + "epoch": 3.4613836652557075, + "grad_norm": 0.6169939041137695, + "learning_rate": 9.360121035430214e-06, + "loss": 0.2207, + "step": 49882 + }, + { + "epoch": 3.461453056692804, + "grad_norm": 1.0766336917877197, + "learning_rate": 9.357747745811508e-06, + "loss": 0.3347, + "step": 49883 + }, + { + "epoch": 3.4615224481299007, + "grad_norm": 1.5428062677383423, + "learning_rate": 9.355374742339862e-06, + "loss": 0.2952, + "step": 49884 + }, + { + "epoch": 3.4615918395669976, + "grad_norm": 1.071038842201233, + "learning_rate": 9.353002025022784e-06, + "loss": 0.3192, + "step": 49885 + }, + { + "epoch": 3.461661231004094, + "grad_norm": 1.4731769561767578, + "learning_rate": 9.350629593867765e-06, + "loss": 0.2892, + "step": 49886 + }, + { + "epoch": 3.461730622441191, + "grad_norm": 0.935177206993103, + "learning_rate": 9.348257448882314e-06, + "loss": 0.2642, + "step": 49887 + }, + { + "epoch": 3.461800013878287, + "grad_norm": 0.8168140649795532, + "learning_rate": 9.345885590073867e-06, + "loss": 0.2676, + "step": 49888 + }, + { + "epoch": 3.461869405315384, + "grad_norm": 1.0221996307373047, + "learning_rate": 9.34351401744995e-06, + "loss": 0.2822, + "step": 49889 + }, + { + "epoch": 3.461938796752481, + "grad_norm": 0.8174107670783997, + "learning_rate": 9.341142731018037e-06, + "loss": 0.3242, + "step": 49890 + }, + { + "epoch": 3.4620081881895772, + "grad_norm": 0.7795341610908508, + "learning_rate": 9.338771730785634e-06, + "loss": 0.2154, + "step": 49891 + }, + { + "epoch": 3.462077579626674, + "grad_norm": 1.000248670578003, + "learning_rate": 9.336401016760199e-06, + "loss": 0.2165, + "step": 49892 + }, + { + "epoch": 3.462146971063771, + "grad_norm": 0.902973473072052, + "learning_rate": 9.334030588949228e-06, + "loss": 0.3313, + "step": 49893 + }, + { + "epoch": 3.4622163625008673, + "grad_norm": 0.9709751009941101, + "learning_rate": 9.331660447360192e-06, + "loss": 0.2203, + "step": 49894 + }, + { + "epoch": 3.462285753937964, + "grad_norm": 0.5336166620254517, + "learning_rate": 9.329290592000617e-06, + "loss": 0.1444, + "step": 49895 + }, + { + "epoch": 3.462355145375061, + "grad_norm": 0.8317033648490906, + "learning_rate": 9.32692102287792e-06, + "loss": 0.2947, + "step": 49896 + }, + { + "epoch": 3.4624245368121573, + "grad_norm": 0.8953993916511536, + "learning_rate": 9.324551739999621e-06, + "loss": 0.332, + "step": 49897 + }, + { + "epoch": 3.462493928249254, + "grad_norm": 1.1729884147644043, + "learning_rate": 9.322182743373187e-06, + "loss": 0.3743, + "step": 49898 + }, + { + "epoch": 3.4625633196863506, + "grad_norm": 0.7406057715415955, + "learning_rate": 9.319814033006103e-06, + "loss": 0.182, + "step": 49899 + }, + { + "epoch": 3.4626327111234474, + "grad_norm": 1.1200779676437378, + "learning_rate": 9.317445608905851e-06, + "loss": 0.2462, + "step": 49900 + }, + { + "epoch": 3.4627021025605442, + "grad_norm": 0.9685900211334229, + "learning_rate": 9.315077471079891e-06, + "loss": 0.3373, + "step": 49901 + }, + { + "epoch": 3.4627714939976406, + "grad_norm": 0.8584092855453491, + "learning_rate": 9.312709619535742e-06, + "loss": 0.2106, + "step": 49902 + }, + { + "epoch": 3.4628408854347374, + "grad_norm": 0.9838305711746216, + "learning_rate": 9.310342054280818e-06, + "loss": 0.239, + "step": 49903 + }, + { + "epoch": 3.462910276871834, + "grad_norm": 1.6028944253921509, + "learning_rate": 9.307974775322625e-06, + "loss": 0.3851, + "step": 49904 + }, + { + "epoch": 3.4629796683089307, + "grad_norm": 1.4019701480865479, + "learning_rate": 9.305607782668635e-06, + "loss": 0.3921, + "step": 49905 + }, + { + "epoch": 3.4630490597460275, + "grad_norm": 0.909614086151123, + "learning_rate": 9.303241076326318e-06, + "loss": 0.1659, + "step": 49906 + }, + { + "epoch": 3.463118451183124, + "grad_norm": 0.9483920335769653, + "learning_rate": 9.300874656303149e-06, + "loss": 0.2809, + "step": 49907 + }, + { + "epoch": 3.4631878426202207, + "grad_norm": 0.9958083629608154, + "learning_rate": 9.298508522606597e-06, + "loss": 0.2117, + "step": 49908 + }, + { + "epoch": 3.463257234057317, + "grad_norm": 1.0206362009048462, + "learning_rate": 9.296142675244146e-06, + "loss": 0.235, + "step": 49909 + }, + { + "epoch": 3.463326625494414, + "grad_norm": 0.7404093146324158, + "learning_rate": 9.293777114223222e-06, + "loss": 0.2657, + "step": 49910 + }, + { + "epoch": 3.463396016931511, + "grad_norm": 1.0771030187606812, + "learning_rate": 9.291411839551335e-06, + "loss": 0.211, + "step": 49911 + }, + { + "epoch": 3.463465408368607, + "grad_norm": 0.6976713538169861, + "learning_rate": 9.289046851235928e-06, + "loss": 0.2051, + "step": 49912 + }, + { + "epoch": 3.463534799805704, + "grad_norm": 1.6310207843780518, + "learning_rate": 9.286682149284476e-06, + "loss": 0.2578, + "step": 49913 + }, + { + "epoch": 3.4636041912428004, + "grad_norm": 0.6958329677581787, + "learning_rate": 9.28431773370444e-06, + "loss": 0.2019, + "step": 49914 + }, + { + "epoch": 3.4636735826798972, + "grad_norm": 0.8671005964279175, + "learning_rate": 9.281953604503291e-06, + "loss": 0.2769, + "step": 49915 + }, + { + "epoch": 3.463742974116994, + "grad_norm": 1.1157177686691284, + "learning_rate": 9.279589761688511e-06, + "loss": 0.4536, + "step": 49916 + }, + { + "epoch": 3.4638123655540904, + "grad_norm": 0.9630977511405945, + "learning_rate": 9.27722620526752e-06, + "loss": 0.256, + "step": 49917 + }, + { + "epoch": 3.4638817569911873, + "grad_norm": 0.9535853266716003, + "learning_rate": 9.274862935247796e-06, + "loss": 0.263, + "step": 49918 + }, + { + "epoch": 3.463951148428284, + "grad_norm": 1.0214613676071167, + "learning_rate": 9.272499951636816e-06, + "loss": 0.2104, + "step": 49919 + }, + { + "epoch": 3.4640205398653805, + "grad_norm": 0.8328747749328613, + "learning_rate": 9.270137254442013e-06, + "loss": 0.2314, + "step": 49920 + }, + { + "epoch": 3.4640899313024773, + "grad_norm": 0.7972052693367004, + "learning_rate": 9.267774843670874e-06, + "loss": 0.2446, + "step": 49921 + }, + { + "epoch": 3.464159322739574, + "grad_norm": 1.0366809368133545, + "learning_rate": 9.265412719330845e-06, + "loss": 0.2176, + "step": 49922 + }, + { + "epoch": 3.4642287141766706, + "grad_norm": 0.6966308355331421, + "learning_rate": 9.263050881429369e-06, + "loss": 0.1882, + "step": 49923 + }, + { + "epoch": 3.4642981056137674, + "grad_norm": 1.1456844806671143, + "learning_rate": 9.260689329973927e-06, + "loss": 0.2565, + "step": 49924 + }, + { + "epoch": 3.4643674970508638, + "grad_norm": 0.8495386242866516, + "learning_rate": 9.258328064971944e-06, + "loss": 0.2775, + "step": 49925 + }, + { + "epoch": 3.4644368884879606, + "grad_norm": 1.040222406387329, + "learning_rate": 9.255967086430917e-06, + "loss": 0.3485, + "step": 49926 + }, + { + "epoch": 3.4645062799250574, + "grad_norm": 0.644110918045044, + "learning_rate": 9.253606394358249e-06, + "loss": 0.1634, + "step": 49927 + }, + { + "epoch": 3.464575671362154, + "grad_norm": 0.8534029126167297, + "learning_rate": 9.251245988761426e-06, + "loss": 0.3541, + "step": 49928 + }, + { + "epoch": 3.4646450627992507, + "grad_norm": 0.461210161447525, + "learning_rate": 9.248885869647883e-06, + "loss": 0.1029, + "step": 49929 + }, + { + "epoch": 3.464714454236347, + "grad_norm": 1.7716882228851318, + "learning_rate": 9.246526037025104e-06, + "loss": 0.4363, + "step": 49930 + }, + { + "epoch": 3.464783845673444, + "grad_norm": 0.9750798940658569, + "learning_rate": 9.244166490900485e-06, + "loss": 0.2718, + "step": 49931 + }, + { + "epoch": 3.4648532371105407, + "grad_norm": 1.1194430589675903, + "learning_rate": 9.241807231281519e-06, + "loss": 0.178, + "step": 49932 + }, + { + "epoch": 3.464922628547637, + "grad_norm": 0.7600377202033997, + "learning_rate": 9.23944825817562e-06, + "loss": 0.2518, + "step": 49933 + }, + { + "epoch": 3.464992019984734, + "grad_norm": 1.2151373624801636, + "learning_rate": 9.237089571590262e-06, + "loss": 0.2797, + "step": 49934 + }, + { + "epoch": 3.4650614114218303, + "grad_norm": 0.7677578926086426, + "learning_rate": 9.234731171532884e-06, + "loss": 0.2253, + "step": 49935 + }, + { + "epoch": 3.465130802858927, + "grad_norm": 0.7701170444488525, + "learning_rate": 9.232373058010924e-06, + "loss": 0.225, + "step": 49936 + }, + { + "epoch": 3.465200194296024, + "grad_norm": 1.0299838781356812, + "learning_rate": 9.230015231031852e-06, + "loss": 0.2838, + "step": 49937 + }, + { + "epoch": 3.4652695857331204, + "grad_norm": 0.9722386002540588, + "learning_rate": 9.227657690603064e-06, + "loss": 0.2975, + "step": 49938 + }, + { + "epoch": 3.465338977170217, + "grad_norm": 1.202641487121582, + "learning_rate": 9.225300436732042e-06, + "loss": 0.3112, + "step": 49939 + }, + { + "epoch": 3.465408368607314, + "grad_norm": 1.020621418952942, + "learning_rate": 9.222943469426204e-06, + "loss": 0.2752, + "step": 49940 + }, + { + "epoch": 3.4654777600444104, + "grad_norm": 1.9261828660964966, + "learning_rate": 9.220586788693009e-06, + "loss": 0.3732, + "step": 49941 + }, + { + "epoch": 3.4655471514815073, + "grad_norm": 1.2403950691223145, + "learning_rate": 9.218230394539884e-06, + "loss": 0.2814, + "step": 49942 + }, + { + "epoch": 3.4656165429186037, + "grad_norm": 0.9996337294578552, + "learning_rate": 9.21587428697427e-06, + "loss": 0.3672, + "step": 49943 + }, + { + "epoch": 3.4656859343557005, + "grad_norm": 0.8343231678009033, + "learning_rate": 9.213518466003634e-06, + "loss": 0.1827, + "step": 49944 + }, + { + "epoch": 3.4657553257927973, + "grad_norm": 1.4300673007965088, + "learning_rate": 9.211162931635364e-06, + "loss": 0.3153, + "step": 49945 + }, + { + "epoch": 3.4658247172298937, + "grad_norm": 0.9458383321762085, + "learning_rate": 9.208807683876919e-06, + "loss": 0.2001, + "step": 49946 + }, + { + "epoch": 3.4658941086669905, + "grad_norm": 1.0345032215118408, + "learning_rate": 9.206452722735737e-06, + "loss": 0.2871, + "step": 49947 + }, + { + "epoch": 3.4659635001040874, + "grad_norm": 0.8241439461708069, + "learning_rate": 9.204098048219256e-06, + "loss": 0.17, + "step": 49948 + }, + { + "epoch": 3.4660328915411838, + "grad_norm": 1.0173940658569336, + "learning_rate": 9.201743660334895e-06, + "loss": 0.1963, + "step": 49949 + }, + { + "epoch": 3.4661022829782806, + "grad_norm": 0.7858508825302124, + "learning_rate": 9.199389559090111e-06, + "loss": 0.2763, + "step": 49950 + }, + { + "epoch": 3.466171674415377, + "grad_norm": 0.8228570818901062, + "learning_rate": 9.19703574449231e-06, + "loss": 0.2584, + "step": 49951 + }, + { + "epoch": 3.466241065852474, + "grad_norm": 0.8831660747528076, + "learning_rate": 9.194682216548922e-06, + "loss": 0.2479, + "step": 49952 + }, + { + "epoch": 3.4663104572895707, + "grad_norm": 1.1466971635818481, + "learning_rate": 9.192328975267395e-06, + "loss": 0.2268, + "step": 49953 + }, + { + "epoch": 3.466379848726667, + "grad_norm": 0.9419638514518738, + "learning_rate": 9.189976020655145e-06, + "loss": 0.2493, + "step": 49954 + }, + { + "epoch": 3.466449240163764, + "grad_norm": 0.7032532095909119, + "learning_rate": 9.18762335271961e-06, + "loss": 0.1926, + "step": 49955 + }, + { + "epoch": 3.4665186316008603, + "grad_norm": 0.7873355746269226, + "learning_rate": 9.185270971468207e-06, + "loss": 0.1669, + "step": 49956 + }, + { + "epoch": 3.466588023037957, + "grad_norm": 0.7057201266288757, + "learning_rate": 9.182918876908386e-06, + "loss": 0.2029, + "step": 49957 + }, + { + "epoch": 3.466657414475054, + "grad_norm": 0.9790684580802917, + "learning_rate": 9.18056706904753e-06, + "loss": 0.3636, + "step": 49958 + }, + { + "epoch": 3.4667268059121503, + "grad_norm": 0.8349452018737793, + "learning_rate": 9.178215547893098e-06, + "loss": 0.2391, + "step": 49959 + }, + { + "epoch": 3.466796197349247, + "grad_norm": 0.9086911678314209, + "learning_rate": 9.175864313452499e-06, + "loss": 0.2096, + "step": 49960 + }, + { + "epoch": 3.4668655887863435, + "grad_norm": 0.9297946095466614, + "learning_rate": 9.173513365733167e-06, + "loss": 0.386, + "step": 49961 + }, + { + "epoch": 3.4669349802234404, + "grad_norm": 0.8872904181480408, + "learning_rate": 9.171162704742509e-06, + "loss": 0.183, + "step": 49962 + }, + { + "epoch": 3.467004371660537, + "grad_norm": 0.7596148252487183, + "learning_rate": 9.168812330487942e-06, + "loss": 0.2501, + "step": 49963 + }, + { + "epoch": 3.4670737630976336, + "grad_norm": 1.0527710914611816, + "learning_rate": 9.166462242976926e-06, + "loss": 0.2808, + "step": 49964 + }, + { + "epoch": 3.4671431545347304, + "grad_norm": 0.9330195784568787, + "learning_rate": 9.164112442216822e-06, + "loss": 0.3494, + "step": 49965 + }, + { + "epoch": 3.4672125459718273, + "grad_norm": 0.893310546875, + "learning_rate": 9.161762928215079e-06, + "loss": 0.2718, + "step": 49966 + }, + { + "epoch": 3.4672819374089237, + "grad_norm": 0.6942929625511169, + "learning_rate": 9.159413700979113e-06, + "loss": 0.2515, + "step": 49967 + }, + { + "epoch": 3.4673513288460205, + "grad_norm": 1.055983066558838, + "learning_rate": 9.157064760516344e-06, + "loss": 0.226, + "step": 49968 + }, + { + "epoch": 3.467420720283117, + "grad_norm": 1.081769347190857, + "learning_rate": 9.154716106834171e-06, + "loss": 0.1903, + "step": 49969 + }, + { + "epoch": 3.4674901117202137, + "grad_norm": 1.0886141061782837, + "learning_rate": 9.152367739940026e-06, + "loss": 0.3652, + "step": 49970 + }, + { + "epoch": 3.4675595031573105, + "grad_norm": 1.0443497896194458, + "learning_rate": 9.150019659841335e-06, + "loss": 0.2318, + "step": 49971 + }, + { + "epoch": 3.467628894594407, + "grad_norm": 0.8532385230064392, + "learning_rate": 9.14767186654547e-06, + "loss": 0.2359, + "step": 49972 + }, + { + "epoch": 3.4676982860315038, + "grad_norm": 0.795922577381134, + "learning_rate": 9.14532436005986e-06, + "loss": 0.2784, + "step": 49973 + }, + { + "epoch": 3.4677676774686006, + "grad_norm": 1.1023149490356445, + "learning_rate": 9.142977140391929e-06, + "loss": 0.306, + "step": 49974 + }, + { + "epoch": 3.467837068905697, + "grad_norm": 1.1992757320404053, + "learning_rate": 9.140630207549073e-06, + "loss": 0.3214, + "step": 49975 + }, + { + "epoch": 3.467906460342794, + "grad_norm": 0.8800418972969055, + "learning_rate": 9.138283561538697e-06, + "loss": 0.241, + "step": 49976 + }, + { + "epoch": 3.46797585177989, + "grad_norm": 0.9787912964820862, + "learning_rate": 9.135937202368227e-06, + "loss": 0.2708, + "step": 49977 + }, + { + "epoch": 3.468045243216987, + "grad_norm": 0.8443401455879211, + "learning_rate": 9.133591130045082e-06, + "loss": 0.2157, + "step": 49978 + }, + { + "epoch": 3.468114634654084, + "grad_norm": 0.7299520373344421, + "learning_rate": 9.131245344576623e-06, + "loss": 0.1975, + "step": 49979 + }, + { + "epoch": 3.4681840260911803, + "grad_norm": 0.8162726759910583, + "learning_rate": 9.128899845970284e-06, + "loss": 0.1895, + "step": 49980 + }, + { + "epoch": 3.468253417528277, + "grad_norm": 0.7852600812911987, + "learning_rate": 9.126554634233464e-06, + "loss": 0.2759, + "step": 49981 + }, + { + "epoch": 3.4683228089653735, + "grad_norm": 0.9030627608299255, + "learning_rate": 9.124209709373565e-06, + "loss": 0.2475, + "step": 49982 + }, + { + "epoch": 3.4683922004024703, + "grad_norm": 0.8817022442817688, + "learning_rate": 9.121865071398005e-06, + "loss": 0.2659, + "step": 49983 + }, + { + "epoch": 3.468461591839567, + "grad_norm": 0.7940007448196411, + "learning_rate": 9.119520720314156e-06, + "loss": 0.2302, + "step": 49984 + }, + { + "epoch": 3.4685309832766635, + "grad_norm": 0.8436826467514038, + "learning_rate": 9.117176656129467e-06, + "loss": 0.2257, + "step": 49985 + }, + { + "epoch": 3.4686003747137604, + "grad_norm": 0.9573782086372375, + "learning_rate": 9.114832878851276e-06, + "loss": 0.3562, + "step": 49986 + }, + { + "epoch": 3.4686697661508568, + "grad_norm": 0.8049353361129761, + "learning_rate": 9.112489388487022e-06, + "loss": 0.2422, + "step": 49987 + }, + { + "epoch": 3.4687391575879536, + "grad_norm": 1.1166880130767822, + "learning_rate": 9.110146185044099e-06, + "loss": 0.2857, + "step": 49988 + }, + { + "epoch": 3.4688085490250504, + "grad_norm": 1.6171561479568481, + "learning_rate": 9.107803268529891e-06, + "loss": 0.3967, + "step": 49989 + }, + { + "epoch": 3.468877940462147, + "grad_norm": 0.8978496193885803, + "learning_rate": 9.1054606389518e-06, + "loss": 0.1738, + "step": 49990 + }, + { + "epoch": 3.4689473318992436, + "grad_norm": 1.23128080368042, + "learning_rate": 9.103118296317226e-06, + "loss": 0.2851, + "step": 49991 + }, + { + "epoch": 3.4690167233363405, + "grad_norm": 0.855094850063324, + "learning_rate": 9.10077624063359e-06, + "loss": 0.3164, + "step": 49992 + }, + { + "epoch": 3.469086114773437, + "grad_norm": 0.8595947623252869, + "learning_rate": 9.098434471908224e-06, + "loss": 0.1611, + "step": 49993 + }, + { + "epoch": 3.4691555062105337, + "grad_norm": 0.6987476348876953, + "learning_rate": 9.096092990148563e-06, + "loss": 0.1572, + "step": 49994 + }, + { + "epoch": 3.46922489764763, + "grad_norm": 0.9155997633934021, + "learning_rate": 9.093751795361993e-06, + "loss": 0.2942, + "step": 49995 + }, + { + "epoch": 3.469294289084727, + "grad_norm": 0.8032954931259155, + "learning_rate": 9.091410887555906e-06, + "loss": 0.2081, + "step": 49996 + }, + { + "epoch": 3.4693636805218238, + "grad_norm": 0.8328197598457336, + "learning_rate": 9.089070266737687e-06, + "loss": 0.2671, + "step": 49997 + }, + { + "epoch": 3.46943307195892, + "grad_norm": 0.9195760488510132, + "learning_rate": 9.086729932914705e-06, + "loss": 0.1586, + "step": 49998 + }, + { + "epoch": 3.469502463396017, + "grad_norm": 1.2125418186187744, + "learning_rate": 9.084389886094402e-06, + "loss": 0.3488, + "step": 49999 + }, + { + "epoch": 3.469571854833114, + "grad_norm": 1.4593156576156616, + "learning_rate": 9.082050126284114e-06, + "loss": 0.3675, + "step": 50000 + }, + { + "epoch": 3.46964124627021, + "grad_norm": 0.9003570675849915, + "learning_rate": 9.079710653491236e-06, + "loss": 0.3055, + "step": 50001 + }, + { + "epoch": 3.469710637707307, + "grad_norm": 0.9849160313606262, + "learning_rate": 9.077371467723172e-06, + "loss": 0.2678, + "step": 50002 + }, + { + "epoch": 3.4697800291444034, + "grad_norm": 0.9435535073280334, + "learning_rate": 9.075032568987296e-06, + "loss": 0.3474, + "step": 50003 + }, + { + "epoch": 3.4698494205815003, + "grad_norm": 0.9759212732315063, + "learning_rate": 9.07269395729099e-06, + "loss": 0.3064, + "step": 50004 + }, + { + "epoch": 3.469918812018597, + "grad_norm": 0.84591144323349, + "learning_rate": 9.070355632641636e-06, + "loss": 0.2592, + "step": 50005 + }, + { + "epoch": 3.4699882034556935, + "grad_norm": 0.7720643281936646, + "learning_rate": 9.06801759504664e-06, + "loss": 0.1616, + "step": 50006 + }, + { + "epoch": 3.4700575948927903, + "grad_norm": 0.5526191592216492, + "learning_rate": 9.065679844513342e-06, + "loss": 0.1841, + "step": 50007 + }, + { + "epoch": 3.4701269863298867, + "grad_norm": 1.0547120571136475, + "learning_rate": 9.063342381049145e-06, + "loss": 0.3549, + "step": 50008 + }, + { + "epoch": 3.4701963777669835, + "grad_norm": 0.6525936126708984, + "learning_rate": 9.061005204661433e-06, + "loss": 0.1727, + "step": 50009 + }, + { + "epoch": 3.4702657692040804, + "grad_norm": 0.6194751262664795, + "learning_rate": 9.058668315357565e-06, + "loss": 0.1878, + "step": 50010 + }, + { + "epoch": 3.4703351606411768, + "grad_norm": 1.0443825721740723, + "learning_rate": 9.05633171314494e-06, + "loss": 0.366, + "step": 50011 + }, + { + "epoch": 3.4704045520782736, + "grad_norm": 0.999318540096283, + "learning_rate": 9.053995398030912e-06, + "loss": 0.1901, + "step": 50012 + }, + { + "epoch": 3.47047394351537, + "grad_norm": 0.7914519906044006, + "learning_rate": 9.051659370022902e-06, + "loss": 0.2397, + "step": 50013 + }, + { + "epoch": 3.470543334952467, + "grad_norm": 0.5605064630508423, + "learning_rate": 9.049323629128226e-06, + "loss": 0.22, + "step": 50014 + }, + { + "epoch": 3.4706127263895636, + "grad_norm": 0.8067977428436279, + "learning_rate": 9.046988175354276e-06, + "loss": 0.1386, + "step": 50015 + }, + { + "epoch": 3.47068211782666, + "grad_norm": 1.3037657737731934, + "learning_rate": 9.044653008708449e-06, + "loss": 0.3922, + "step": 50016 + }, + { + "epoch": 3.470751509263757, + "grad_norm": 0.8295517563819885, + "learning_rate": 9.04231812919808e-06, + "loss": 0.2787, + "step": 50017 + }, + { + "epoch": 3.4708209007008537, + "grad_norm": 1.0060425996780396, + "learning_rate": 9.039983536830576e-06, + "loss": 0.2545, + "step": 50018 + }, + { + "epoch": 3.47089029213795, + "grad_norm": 0.9296687245368958, + "learning_rate": 9.037649231613277e-06, + "loss": 0.2449, + "step": 50019 + }, + { + "epoch": 3.470959683575047, + "grad_norm": 1.0216326713562012, + "learning_rate": 9.035315213553597e-06, + "loss": 0.2629, + "step": 50020 + }, + { + "epoch": 3.4710290750121437, + "grad_norm": 1.241590142250061, + "learning_rate": 9.032981482658843e-06, + "loss": 0.2963, + "step": 50021 + }, + { + "epoch": 3.47109846644924, + "grad_norm": 1.5560230016708374, + "learning_rate": 9.030648038936418e-06, + "loss": 0.273, + "step": 50022 + }, + { + "epoch": 3.471167857886337, + "grad_norm": 0.7980877757072449, + "learning_rate": 9.028314882393684e-06, + "loss": 0.2934, + "step": 50023 + }, + { + "epoch": 3.4712372493234334, + "grad_norm": 1.2375500202178955, + "learning_rate": 9.025982013038004e-06, + "loss": 0.3624, + "step": 50024 + }, + { + "epoch": 3.47130664076053, + "grad_norm": 1.2450405359268188, + "learning_rate": 9.023649430876746e-06, + "loss": 0.2553, + "step": 50025 + }, + { + "epoch": 3.471376032197627, + "grad_norm": 0.6187612414360046, + "learning_rate": 9.021317135917285e-06, + "loss": 0.1567, + "step": 50026 + }, + { + "epoch": 3.4714454236347234, + "grad_norm": 1.0594642162322998, + "learning_rate": 9.018985128166958e-06, + "loss": 0.2693, + "step": 50027 + }, + { + "epoch": 3.4715148150718202, + "grad_norm": 1.2003097534179688, + "learning_rate": 9.016653407633135e-06, + "loss": 0.2183, + "step": 50028 + }, + { + "epoch": 3.4715842065089166, + "grad_norm": 0.7106826901435852, + "learning_rate": 9.014321974323182e-06, + "loss": 0.1289, + "step": 50029 + }, + { + "epoch": 3.4716535979460135, + "grad_norm": 0.6886901259422302, + "learning_rate": 9.011990828244476e-06, + "loss": 0.2707, + "step": 50030 + }, + { + "epoch": 3.4717229893831103, + "grad_norm": 1.1489816904067993, + "learning_rate": 9.009659969404338e-06, + "loss": 0.2859, + "step": 50031 + }, + { + "epoch": 3.4717923808202067, + "grad_norm": 1.068692684173584, + "learning_rate": 9.007329397810149e-06, + "loss": 0.2413, + "step": 50032 + }, + { + "epoch": 3.4718617722573035, + "grad_norm": 0.6428796052932739, + "learning_rate": 9.004999113469259e-06, + "loss": 0.1992, + "step": 50033 + }, + { + "epoch": 3.4719311636944, + "grad_norm": 1.009320616722107, + "learning_rate": 9.00266911638905e-06, + "loss": 0.1711, + "step": 50034 + }, + { + "epoch": 3.4720005551314967, + "grad_norm": 1.1344224214553833, + "learning_rate": 9.000339406576852e-06, + "loss": 0.2946, + "step": 50035 + }, + { + "epoch": 3.4720699465685936, + "grad_norm": 0.9346086978912354, + "learning_rate": 8.998009984040012e-06, + "loss": 0.2411, + "step": 50036 + }, + { + "epoch": 3.47213933800569, + "grad_norm": 0.749988853931427, + "learning_rate": 8.995680848785892e-06, + "loss": 0.2178, + "step": 50037 + }, + { + "epoch": 3.472208729442787, + "grad_norm": 1.1783056259155273, + "learning_rate": 8.993352000821864e-06, + "loss": 0.3903, + "step": 50038 + }, + { + "epoch": 3.472278120879883, + "grad_norm": 1.5863624811172485, + "learning_rate": 8.991023440155256e-06, + "loss": 0.2126, + "step": 50039 + }, + { + "epoch": 3.47234751231698, + "grad_norm": 1.0490833520889282, + "learning_rate": 8.988695166793437e-06, + "loss": 0.2427, + "step": 50040 + }, + { + "epoch": 3.472416903754077, + "grad_norm": 1.0287950038909912, + "learning_rate": 8.986367180743748e-06, + "loss": 0.2585, + "step": 50041 + }, + { + "epoch": 3.4724862951911732, + "grad_norm": 1.0922399759292603, + "learning_rate": 8.984039482013529e-06, + "loss": 0.2717, + "step": 50042 + }, + { + "epoch": 3.47255568662827, + "grad_norm": 0.8300145268440247, + "learning_rate": 8.981712070610137e-06, + "loss": 0.2511, + "step": 50043 + }, + { + "epoch": 3.472625078065367, + "grad_norm": 0.9857034683227539, + "learning_rate": 8.979384946540915e-06, + "loss": 0.3466, + "step": 50044 + }, + { + "epoch": 3.4726944695024633, + "grad_norm": 0.9001913070678711, + "learning_rate": 8.97705810981323e-06, + "loss": 0.2676, + "step": 50045 + }, + { + "epoch": 3.47276386093956, + "grad_norm": 0.6010140180587769, + "learning_rate": 8.974731560434402e-06, + "loss": 0.221, + "step": 50046 + }, + { + "epoch": 3.472833252376657, + "grad_norm": 2.003659248352051, + "learning_rate": 8.9724052984118e-06, + "loss": 0.2112, + "step": 50047 + }, + { + "epoch": 3.4729026438137534, + "grad_norm": 1.4814279079437256, + "learning_rate": 8.970079323752733e-06, + "loss": 0.2724, + "step": 50048 + }, + { + "epoch": 3.47297203525085, + "grad_norm": 0.9156795740127563, + "learning_rate": 8.967753636464582e-06, + "loss": 0.272, + "step": 50049 + }, + { + "epoch": 3.4730414266879466, + "grad_norm": 1.0338488817214966, + "learning_rate": 8.96542823655465e-06, + "loss": 0.2868, + "step": 50050 + }, + { + "epoch": 3.4731108181250434, + "grad_norm": 1.025503158569336, + "learning_rate": 8.963103124030315e-06, + "loss": 0.1933, + "step": 50051 + }, + { + "epoch": 3.4731802095621402, + "grad_norm": 1.2590945959091187, + "learning_rate": 8.960778298898898e-06, + "loss": 0.4118, + "step": 50052 + }, + { + "epoch": 3.4732496009992366, + "grad_norm": 1.2906557321548462, + "learning_rate": 8.958453761167729e-06, + "loss": 0.5607, + "step": 50053 + }, + { + "epoch": 3.4733189924363335, + "grad_norm": 0.677863597869873, + "learning_rate": 8.956129510844191e-06, + "loss": 0.2037, + "step": 50054 + }, + { + "epoch": 3.47338838387343, + "grad_norm": 0.8338835835456848, + "learning_rate": 8.953805547935567e-06, + "loss": 0.2502, + "step": 50055 + }, + { + "epoch": 3.4734577753105267, + "grad_norm": 0.8759499788284302, + "learning_rate": 8.951481872449218e-06, + "loss": 0.2608, + "step": 50056 + }, + { + "epoch": 3.4735271667476235, + "grad_norm": 0.5840532183647156, + "learning_rate": 8.949158484392473e-06, + "loss": 0.1889, + "step": 50057 + }, + { + "epoch": 3.47359655818472, + "grad_norm": 0.7092726826667786, + "learning_rate": 8.946835383772667e-06, + "loss": 0.1926, + "step": 50058 + }, + { + "epoch": 3.4736659496218167, + "grad_norm": 0.9313302636146545, + "learning_rate": 8.944512570597141e-06, + "loss": 0.2878, + "step": 50059 + }, + { + "epoch": 3.473735341058913, + "grad_norm": 0.9426036477088928, + "learning_rate": 8.942190044873222e-06, + "loss": 0.314, + "step": 50060 + }, + { + "epoch": 3.47380473249601, + "grad_norm": 0.7111430764198303, + "learning_rate": 8.939867806608271e-06, + "loss": 0.2267, + "step": 50061 + }, + { + "epoch": 3.473874123933107, + "grad_norm": 0.930547297000885, + "learning_rate": 8.93754585580957e-06, + "loss": 0.2459, + "step": 50062 + }, + { + "epoch": 3.473943515370203, + "grad_norm": 0.6832945942878723, + "learning_rate": 8.93522419248447e-06, + "loss": 0.2673, + "step": 50063 + }, + { + "epoch": 3.4740129068073, + "grad_norm": 0.709347128868103, + "learning_rate": 8.93290281664031e-06, + "loss": 0.2165, + "step": 50064 + }, + { + "epoch": 3.4740822982443964, + "grad_norm": 0.9156220555305481, + "learning_rate": 8.930581728284415e-06, + "loss": 0.2587, + "step": 50065 + }, + { + "epoch": 3.4741516896814932, + "grad_norm": 0.829328715801239, + "learning_rate": 8.928260927424093e-06, + "loss": 0.1979, + "step": 50066 + }, + { + "epoch": 3.47422108111859, + "grad_norm": 1.2737512588500977, + "learning_rate": 8.925940414066691e-06, + "loss": 0.4887, + "step": 50067 + }, + { + "epoch": 3.4742904725556865, + "grad_norm": 0.8965374827384949, + "learning_rate": 8.923620188219528e-06, + "loss": 0.2077, + "step": 50068 + }, + { + "epoch": 3.4743598639927833, + "grad_norm": 0.8185339570045471, + "learning_rate": 8.921300249889941e-06, + "loss": 0.187, + "step": 50069 + }, + { + "epoch": 3.47442925542988, + "grad_norm": 0.9923223853111267, + "learning_rate": 8.918980599085225e-06, + "loss": 0.1986, + "step": 50070 + }, + { + "epoch": 3.4744986468669765, + "grad_norm": 0.8261917233467102, + "learning_rate": 8.916661235812718e-06, + "loss": 0.2306, + "step": 50071 + }, + { + "epoch": 3.4745680383040733, + "grad_norm": 0.9507869482040405, + "learning_rate": 8.914342160079747e-06, + "loss": 0.2356, + "step": 50072 + }, + { + "epoch": 3.47463742974117, + "grad_norm": 1.1288312673568726, + "learning_rate": 8.91202337189364e-06, + "loss": 0.4877, + "step": 50073 + }, + { + "epoch": 3.4747068211782666, + "grad_norm": 0.7820106744766235, + "learning_rate": 8.909704871261693e-06, + "loss": 0.2324, + "step": 50074 + }, + { + "epoch": 3.4747762126153634, + "grad_norm": 0.5994094014167786, + "learning_rate": 8.907386658191264e-06, + "loss": 0.1811, + "step": 50075 + }, + { + "epoch": 3.47484560405246, + "grad_norm": 0.639511227607727, + "learning_rate": 8.905068732689625e-06, + "loss": 0.1272, + "step": 50076 + }, + { + "epoch": 3.4749149954895566, + "grad_norm": 0.85586017370224, + "learning_rate": 8.902751094764116e-06, + "loss": 0.3415, + "step": 50077 + }, + { + "epoch": 3.4749843869266535, + "grad_norm": 0.951140820980072, + "learning_rate": 8.900433744422054e-06, + "loss": 0.3471, + "step": 50078 + }, + { + "epoch": 3.47505377836375, + "grad_norm": 1.1220921277999878, + "learning_rate": 8.898116681670755e-06, + "loss": 0.2452, + "step": 50079 + }, + { + "epoch": 3.4751231698008467, + "grad_norm": 0.695396363735199, + "learning_rate": 8.895799906517533e-06, + "loss": 0.2342, + "step": 50080 + }, + { + "epoch": 3.475192561237943, + "grad_norm": 0.9233425855636597, + "learning_rate": 8.893483418969695e-06, + "loss": 0.2142, + "step": 50081 + }, + { + "epoch": 3.47526195267504, + "grad_norm": 0.7053220272064209, + "learning_rate": 8.89116721903458e-06, + "loss": 0.2453, + "step": 50082 + }, + { + "epoch": 3.4753313441121367, + "grad_norm": 1.0258922576904297, + "learning_rate": 8.88885130671946e-06, + "loss": 0.1419, + "step": 50083 + }, + { + "epoch": 3.475400735549233, + "grad_norm": 1.0434083938598633, + "learning_rate": 8.88653568203166e-06, + "loss": 0.2667, + "step": 50084 + }, + { + "epoch": 3.47547012698633, + "grad_norm": 0.829907238483429, + "learning_rate": 8.884220344978512e-06, + "loss": 0.1971, + "step": 50085 + }, + { + "epoch": 3.4755395184234263, + "grad_norm": 1.0055763721466064, + "learning_rate": 8.881905295567294e-06, + "loss": 0.3246, + "step": 50086 + }, + { + "epoch": 3.475608909860523, + "grad_norm": 1.0082213878631592, + "learning_rate": 8.879590533805337e-06, + "loss": 0.239, + "step": 50087 + }, + { + "epoch": 3.47567830129762, + "grad_norm": 0.8746071457862854, + "learning_rate": 8.877276059699936e-06, + "loss": 0.2549, + "step": 50088 + }, + { + "epoch": 3.4757476927347164, + "grad_norm": 0.954498291015625, + "learning_rate": 8.874961873258424e-06, + "loss": 0.2989, + "step": 50089 + }, + { + "epoch": 3.4758170841718132, + "grad_norm": 0.8996133804321289, + "learning_rate": 8.872647974488069e-06, + "loss": 0.2341, + "step": 50090 + }, + { + "epoch": 3.4758864756089096, + "grad_norm": 1.1252272129058838, + "learning_rate": 8.870334363396183e-06, + "loss": 0.243, + "step": 50091 + }, + { + "epoch": 3.4759558670460065, + "grad_norm": 1.0991744995117188, + "learning_rate": 8.868021039990082e-06, + "loss": 0.1989, + "step": 50092 + }, + { + "epoch": 3.4760252584831033, + "grad_norm": 0.9467124342918396, + "learning_rate": 8.865708004277073e-06, + "loss": 0.3341, + "step": 50093 + }, + { + "epoch": 3.4760946499201997, + "grad_norm": 0.662769615650177, + "learning_rate": 8.86339525626444e-06, + "loss": 0.2318, + "step": 50094 + }, + { + "epoch": 3.4761640413572965, + "grad_norm": 0.7110011577606201, + "learning_rate": 8.861082795959496e-06, + "loss": 0.2353, + "step": 50095 + }, + { + "epoch": 3.4762334327943933, + "grad_norm": 0.8389377593994141, + "learning_rate": 8.858770623369561e-06, + "loss": 0.1781, + "step": 50096 + }, + { + "epoch": 3.4763028242314897, + "grad_norm": 0.6589994430541992, + "learning_rate": 8.856458738501893e-06, + "loss": 0.2097, + "step": 50097 + }, + { + "epoch": 3.4763722156685866, + "grad_norm": 1.08086359500885, + "learning_rate": 8.854147141363811e-06, + "loss": 0.3608, + "step": 50098 + }, + { + "epoch": 3.4764416071056834, + "grad_norm": 0.8958352208137512, + "learning_rate": 8.851835831962618e-06, + "loss": 0.2053, + "step": 50099 + }, + { + "epoch": 3.47651099854278, + "grad_norm": 0.8744810819625854, + "learning_rate": 8.84952481030561e-06, + "loss": 0.2361, + "step": 50100 + }, + { + "epoch": 3.4765803899798766, + "grad_norm": 0.8016396164894104, + "learning_rate": 8.847214076400068e-06, + "loss": 0.1366, + "step": 50101 + }, + { + "epoch": 3.476649781416973, + "grad_norm": 0.6941695809364319, + "learning_rate": 8.844903630253288e-06, + "loss": 0.174, + "step": 50102 + }, + { + "epoch": 3.47671917285407, + "grad_norm": 0.8651891946792603, + "learning_rate": 8.842593471872573e-06, + "loss": 0.2873, + "step": 50103 + }, + { + "epoch": 3.4767885642911667, + "grad_norm": 1.57549250125885, + "learning_rate": 8.840283601265242e-06, + "loss": 0.301, + "step": 50104 + }, + { + "epoch": 3.476857955728263, + "grad_norm": 0.8703059554100037, + "learning_rate": 8.837974018438533e-06, + "loss": 0.3246, + "step": 50105 + }, + { + "epoch": 3.47692734716536, + "grad_norm": 0.970428466796875, + "learning_rate": 8.835664723399761e-06, + "loss": 0.3612, + "step": 50106 + }, + { + "epoch": 3.4769967386024563, + "grad_norm": 0.9711107015609741, + "learning_rate": 8.833355716156222e-06, + "loss": 0.222, + "step": 50107 + }, + { + "epoch": 3.477066130039553, + "grad_norm": 0.9141433835029602, + "learning_rate": 8.831046996715197e-06, + "loss": 0.2502, + "step": 50108 + }, + { + "epoch": 3.47713552147665, + "grad_norm": 0.6591401100158691, + "learning_rate": 8.828738565084005e-06, + "loss": 0.1874, + "step": 50109 + }, + { + "epoch": 3.4772049129137463, + "grad_norm": 1.0473346710205078, + "learning_rate": 8.826430421269893e-06, + "loss": 0.3131, + "step": 50110 + }, + { + "epoch": 3.477274304350843, + "grad_norm": 0.6469184756278992, + "learning_rate": 8.824122565280145e-06, + "loss": 0.1804, + "step": 50111 + }, + { + "epoch": 3.4773436957879396, + "grad_norm": 0.8103820085525513, + "learning_rate": 8.82181499712208e-06, + "loss": 0.3053, + "step": 50112 + }, + { + "epoch": 3.4774130872250364, + "grad_norm": 1.3614267110824585, + "learning_rate": 8.819507716802955e-06, + "loss": 0.387, + "step": 50113 + }, + { + "epoch": 3.4774824786621332, + "grad_norm": 0.9274186491966248, + "learning_rate": 8.817200724330077e-06, + "loss": 0.3571, + "step": 50114 + }, + { + "epoch": 3.4775518700992296, + "grad_norm": 1.1601144075393677, + "learning_rate": 8.814894019710707e-06, + "loss": 0.2954, + "step": 50115 + }, + { + "epoch": 3.4776212615363264, + "grad_norm": 0.8037429451942444, + "learning_rate": 8.81258760295216e-06, + "loss": 0.1991, + "step": 50116 + }, + { + "epoch": 3.477690652973423, + "grad_norm": 0.8730477690696716, + "learning_rate": 8.810281474061676e-06, + "loss": 0.2629, + "step": 50117 + }, + { + "epoch": 3.4777600444105197, + "grad_norm": 1.100984811782837, + "learning_rate": 8.80797563304655e-06, + "loss": 0.434, + "step": 50118 + }, + { + "epoch": 3.4778294358476165, + "grad_norm": 0.7815274596214294, + "learning_rate": 8.805670079914063e-06, + "loss": 0.1534, + "step": 50119 + }, + { + "epoch": 3.477898827284713, + "grad_norm": 0.7727642059326172, + "learning_rate": 8.8033648146715e-06, + "loss": 0.2388, + "step": 50120 + }, + { + "epoch": 3.4779682187218097, + "grad_norm": 0.9115703701972961, + "learning_rate": 8.801059837326132e-06, + "loss": 0.2811, + "step": 50121 + }, + { + "epoch": 3.4780376101589066, + "grad_norm": 0.8206939697265625, + "learning_rate": 8.79875514788524e-06, + "loss": 0.2341, + "step": 50122 + }, + { + "epoch": 3.478107001596003, + "grad_norm": 1.2856383323669434, + "learning_rate": 8.796450746356112e-06, + "loss": 0.2775, + "step": 50123 + }, + { + "epoch": 3.4781763930331, + "grad_norm": 0.7698783278465271, + "learning_rate": 8.794146632745992e-06, + "loss": 0.1777, + "step": 50124 + }, + { + "epoch": 3.4782457844701966, + "grad_norm": 1.3426891565322876, + "learning_rate": 8.791842807062167e-06, + "loss": 0.3964, + "step": 50125 + }, + { + "epoch": 3.478315175907293, + "grad_norm": 1.0421842336654663, + "learning_rate": 8.789539269311908e-06, + "loss": 0.2198, + "step": 50126 + }, + { + "epoch": 3.47838456734439, + "grad_norm": 0.999614417552948, + "learning_rate": 8.787236019502498e-06, + "loss": 0.2439, + "step": 50127 + }, + { + "epoch": 3.478453958781486, + "grad_norm": 0.7764596343040466, + "learning_rate": 8.784933057641198e-06, + "loss": 0.2307, + "step": 50128 + }, + { + "epoch": 3.478523350218583, + "grad_norm": 0.8518458008766174, + "learning_rate": 8.78263038373528e-06, + "loss": 0.2851, + "step": 50129 + }, + { + "epoch": 3.47859274165568, + "grad_norm": 0.8851873874664307, + "learning_rate": 8.780327997792037e-06, + "loss": 0.2268, + "step": 50130 + }, + { + "epoch": 3.4786621330927763, + "grad_norm": 0.7374415993690491, + "learning_rate": 8.778025899818699e-06, + "loss": 0.2195, + "step": 50131 + }, + { + "epoch": 3.478731524529873, + "grad_norm": 1.1894028186798096, + "learning_rate": 8.775724089822545e-06, + "loss": 0.1896, + "step": 50132 + }, + { + "epoch": 3.4788009159669695, + "grad_norm": 0.9294034838676453, + "learning_rate": 8.77342256781084e-06, + "loss": 0.2557, + "step": 50133 + }, + { + "epoch": 3.4788703074040663, + "grad_norm": 0.5777519345283508, + "learning_rate": 8.771121333790865e-06, + "loss": 0.1548, + "step": 50134 + }, + { + "epoch": 3.478939698841163, + "grad_norm": 1.6026538610458374, + "learning_rate": 8.768820387769894e-06, + "loss": 0.3549, + "step": 50135 + }, + { + "epoch": 3.4790090902782596, + "grad_norm": 1.0345783233642578, + "learning_rate": 8.76651972975514e-06, + "loss": 0.227, + "step": 50136 + }, + { + "epoch": 3.4790784817153564, + "grad_norm": 1.0287503004074097, + "learning_rate": 8.76421935975391e-06, + "loss": 0.3257, + "step": 50137 + }, + { + "epoch": 3.4791478731524528, + "grad_norm": 0.674139678478241, + "learning_rate": 8.761919277773456e-06, + "loss": 0.2049, + "step": 50138 + }, + { + "epoch": 3.4792172645895496, + "grad_norm": 0.7787673473358154, + "learning_rate": 8.759619483821057e-06, + "loss": 0.2661, + "step": 50139 + }, + { + "epoch": 3.4792866560266464, + "grad_norm": 0.806316614151001, + "learning_rate": 8.757319977903932e-06, + "loss": 0.2077, + "step": 50140 + }, + { + "epoch": 3.479356047463743, + "grad_norm": 0.7924919128417969, + "learning_rate": 8.755020760029353e-06, + "loss": 0.1755, + "step": 50141 + }, + { + "epoch": 3.4794254389008397, + "grad_norm": 0.7524957656860352, + "learning_rate": 8.752721830204602e-06, + "loss": 0.1747, + "step": 50142 + }, + { + "epoch": 3.479494830337936, + "grad_norm": 1.0706242322921753, + "learning_rate": 8.750423188436907e-06, + "loss": 0.2602, + "step": 50143 + }, + { + "epoch": 3.479564221775033, + "grad_norm": 0.7936206459999084, + "learning_rate": 8.748124834733562e-06, + "loss": 0.1666, + "step": 50144 + }, + { + "epoch": 3.4796336132121297, + "grad_norm": 1.192436695098877, + "learning_rate": 8.745826769101784e-06, + "loss": 0.4013, + "step": 50145 + }, + { + "epoch": 3.479703004649226, + "grad_norm": 0.793814480304718, + "learning_rate": 8.743528991548844e-06, + "loss": 0.2824, + "step": 50146 + }, + { + "epoch": 3.479772396086323, + "grad_norm": 0.9917550086975098, + "learning_rate": 8.741231502082003e-06, + "loss": 0.1844, + "step": 50147 + }, + { + "epoch": 3.4798417875234198, + "grad_norm": 0.8228799700737, + "learning_rate": 8.73893430070849e-06, + "loss": 0.2086, + "step": 50148 + }, + { + "epoch": 3.479911178960516, + "grad_norm": 1.0554072856903076, + "learning_rate": 8.736637387435587e-06, + "loss": 0.2664, + "step": 50149 + }, + { + "epoch": 3.479980570397613, + "grad_norm": 1.740294337272644, + "learning_rate": 8.734340762270521e-06, + "loss": 0.2376, + "step": 50150 + }, + { + "epoch": 3.48004996183471, + "grad_norm": 0.7415263056755066, + "learning_rate": 8.732044425220576e-06, + "loss": 0.1987, + "step": 50151 + }, + { + "epoch": 3.480119353271806, + "grad_norm": 0.8147706985473633, + "learning_rate": 8.729748376292957e-06, + "loss": 0.3385, + "step": 50152 + }, + { + "epoch": 3.480188744708903, + "grad_norm": 1.0463634729385376, + "learning_rate": 8.727452615494935e-06, + "loss": 0.2139, + "step": 50153 + }, + { + "epoch": 3.4802581361459994, + "grad_norm": 0.9271494150161743, + "learning_rate": 8.725157142833751e-06, + "loss": 0.177, + "step": 50154 + }, + { + "epoch": 3.4803275275830963, + "grad_norm": 0.750580906867981, + "learning_rate": 8.722861958316663e-06, + "loss": 0.2237, + "step": 50155 + }, + { + "epoch": 3.480396919020193, + "grad_norm": 1.1505813598632812, + "learning_rate": 8.720567061950913e-06, + "loss": 0.2408, + "step": 50156 + }, + { + "epoch": 3.4804663104572895, + "grad_norm": 1.025027871131897, + "learning_rate": 8.718272453743726e-06, + "loss": 0.2492, + "step": 50157 + }, + { + "epoch": 3.4805357018943863, + "grad_norm": 0.8243492245674133, + "learning_rate": 8.715978133702396e-06, + "loss": 0.2526, + "step": 50158 + }, + { + "epoch": 3.4806050933314827, + "grad_norm": 0.9699327945709229, + "learning_rate": 8.713684101834108e-06, + "loss": 0.2121, + "step": 50159 + }, + { + "epoch": 3.4806744847685795, + "grad_norm": 0.8326755166053772, + "learning_rate": 8.711390358146132e-06, + "loss": 0.2611, + "step": 50160 + }, + { + "epoch": 3.4807438762056764, + "grad_norm": 0.7344263792037964, + "learning_rate": 8.709096902645697e-06, + "loss": 0.1963, + "step": 50161 + }, + { + "epoch": 3.4808132676427728, + "grad_norm": 0.9135846495628357, + "learning_rate": 8.706803735340051e-06, + "loss": 0.2899, + "step": 50162 + }, + { + "epoch": 3.4808826590798696, + "grad_norm": 0.7647179365158081, + "learning_rate": 8.704510856236448e-06, + "loss": 0.2402, + "step": 50163 + }, + { + "epoch": 3.480952050516966, + "grad_norm": 1.1202362775802612, + "learning_rate": 8.702218265342099e-06, + "loss": 0.2629, + "step": 50164 + }, + { + "epoch": 3.481021441954063, + "grad_norm": 0.7144201993942261, + "learning_rate": 8.699925962664279e-06, + "loss": 0.1989, + "step": 50165 + }, + { + "epoch": 3.4810908333911597, + "grad_norm": 0.8398622274398804, + "learning_rate": 8.697633948210182e-06, + "loss": 0.3017, + "step": 50166 + }, + { + "epoch": 3.481160224828256, + "grad_norm": 0.760253369808197, + "learning_rate": 8.695342221987057e-06, + "loss": 0.226, + "step": 50167 + }, + { + "epoch": 3.481229616265353, + "grad_norm": 1.0405688285827637, + "learning_rate": 8.693050784002144e-06, + "loss": 0.2542, + "step": 50168 + }, + { + "epoch": 3.4812990077024497, + "grad_norm": 0.9904788732528687, + "learning_rate": 8.69075963426268e-06, + "loss": 0.3185, + "step": 50169 + }, + { + "epoch": 3.481368399139546, + "grad_norm": 1.0876306295394897, + "learning_rate": 8.688468772775904e-06, + "loss": 0.3027, + "step": 50170 + }, + { + "epoch": 3.481437790576643, + "grad_norm": 0.8336295485496521, + "learning_rate": 8.686178199549022e-06, + "loss": 0.2876, + "step": 50171 + }, + { + "epoch": 3.4815071820137393, + "grad_norm": 0.9937582015991211, + "learning_rate": 8.683887914589284e-06, + "loss": 0.1924, + "step": 50172 + }, + { + "epoch": 3.481576573450836, + "grad_norm": 0.922598659992218, + "learning_rate": 8.681597917903938e-06, + "loss": 0.2178, + "step": 50173 + }, + { + "epoch": 3.481645964887933, + "grad_norm": 1.155422329902649, + "learning_rate": 8.67930820950017e-06, + "loss": 0.3078, + "step": 50174 + }, + { + "epoch": 3.4817153563250294, + "grad_norm": 0.8571503758430481, + "learning_rate": 8.677018789385238e-06, + "loss": 0.2304, + "step": 50175 + }, + { + "epoch": 3.481784747762126, + "grad_norm": 0.8138239979743958, + "learning_rate": 8.674729657566349e-06, + "loss": 0.2186, + "step": 50176 + }, + { + "epoch": 3.481854139199223, + "grad_norm": 0.7238038182258606, + "learning_rate": 8.672440814050753e-06, + "loss": 0.1977, + "step": 50177 + }, + { + "epoch": 3.4819235306363194, + "grad_norm": 3.438920021057129, + "learning_rate": 8.670152258845687e-06, + "loss": 0.3632, + "step": 50178 + }, + { + "epoch": 3.4819929220734163, + "grad_norm": 0.6573824882507324, + "learning_rate": 8.667863991958325e-06, + "loss": 0.2598, + "step": 50179 + }, + { + "epoch": 3.4820623135105127, + "grad_norm": 0.7888036966323853, + "learning_rate": 8.665576013395926e-06, + "loss": 0.2607, + "step": 50180 + }, + { + "epoch": 3.4821317049476095, + "grad_norm": 0.615902304649353, + "learning_rate": 8.663288323165708e-06, + "loss": 0.1487, + "step": 50181 + }, + { + "epoch": 3.4822010963847063, + "grad_norm": 1.24152410030365, + "learning_rate": 8.661000921274886e-06, + "loss": 0.3897, + "step": 50182 + }, + { + "epoch": 3.4822704878218027, + "grad_norm": 0.9315531849861145, + "learning_rate": 8.65871380773069e-06, + "loss": 0.2403, + "step": 50183 + }, + { + "epoch": 3.4823398792588995, + "grad_norm": 1.3793784379959106, + "learning_rate": 8.656426982540345e-06, + "loss": 0.32, + "step": 50184 + }, + { + "epoch": 3.482409270695996, + "grad_norm": 0.8516339063644409, + "learning_rate": 8.654140445711068e-06, + "loss": 0.1472, + "step": 50185 + }, + { + "epoch": 3.4824786621330928, + "grad_norm": 1.0741878747940063, + "learning_rate": 8.651854197250053e-06, + "loss": 0.1945, + "step": 50186 + }, + { + "epoch": 3.4825480535701896, + "grad_norm": 0.6955037117004395, + "learning_rate": 8.649568237164551e-06, + "loss": 0.1748, + "step": 50187 + }, + { + "epoch": 3.482617445007286, + "grad_norm": 0.9472651481628418, + "learning_rate": 8.647282565461745e-06, + "loss": 0.3535, + "step": 50188 + }, + { + "epoch": 3.482686836444383, + "grad_norm": 1.3057639598846436, + "learning_rate": 8.644997182148883e-06, + "loss": 0.4264, + "step": 50189 + }, + { + "epoch": 3.482756227881479, + "grad_norm": 1.046463966369629, + "learning_rate": 8.642712087233163e-06, + "loss": 0.3327, + "step": 50190 + }, + { + "epoch": 3.482825619318576, + "grad_norm": 0.7704046368598938, + "learning_rate": 8.640427280721796e-06, + "loss": 0.1947, + "step": 50191 + }, + { + "epoch": 3.482895010755673, + "grad_norm": 0.9143346548080444, + "learning_rate": 8.638142762622014e-06, + "loss": 0.2341, + "step": 50192 + }, + { + "epoch": 3.4829644021927693, + "grad_norm": 0.9661046266555786, + "learning_rate": 8.635858532941011e-06, + "loss": 0.273, + "step": 50193 + }, + { + "epoch": 3.483033793629866, + "grad_norm": 0.7657408118247986, + "learning_rate": 8.633574591686e-06, + "loss": 0.2124, + "step": 50194 + }, + { + "epoch": 3.483103185066963, + "grad_norm": 1.2223467826843262, + "learning_rate": 8.63129093886419e-06, + "loss": 0.2875, + "step": 50195 + }, + { + "epoch": 3.4831725765040593, + "grad_norm": 1.021144986152649, + "learning_rate": 8.629007574482784e-06, + "loss": 0.2389, + "step": 50196 + }, + { + "epoch": 3.483241967941156, + "grad_norm": 0.812339186668396, + "learning_rate": 8.62672449854901e-06, + "loss": 0.2012, + "step": 50197 + }, + { + "epoch": 3.4833113593782525, + "grad_norm": 1.1292716264724731, + "learning_rate": 8.624441711070063e-06, + "loss": 0.4306, + "step": 50198 + }, + { + "epoch": 3.4833807508153494, + "grad_norm": 0.7917577028274536, + "learning_rate": 8.622159212053182e-06, + "loss": 0.2334, + "step": 50199 + }, + { + "epoch": 3.483450142252446, + "grad_norm": 0.6523983478546143, + "learning_rate": 8.619877001505505e-06, + "loss": 0.1587, + "step": 50200 + }, + { + "epoch": 3.4835195336895426, + "grad_norm": 1.0860378742218018, + "learning_rate": 8.617595079434292e-06, + "loss": 0.1937, + "step": 50201 + }, + { + "epoch": 3.4835889251266394, + "grad_norm": 0.71172034740448, + "learning_rate": 8.615313445846718e-06, + "loss": 0.2173, + "step": 50202 + }, + { + "epoch": 3.4836583165637363, + "grad_norm": 1.178661823272705, + "learning_rate": 8.613032100749996e-06, + "loss": 0.3157, + "step": 50203 + }, + { + "epoch": 3.4837277080008326, + "grad_norm": 0.7374966740608215, + "learning_rate": 8.610751044151333e-06, + "loss": 0.2437, + "step": 50204 + }, + { + "epoch": 3.4837970994379295, + "grad_norm": 0.9719739556312561, + "learning_rate": 8.608470276057945e-06, + "loss": 0.2501, + "step": 50205 + }, + { + "epoch": 3.483866490875026, + "grad_norm": 0.8199664354324341, + "learning_rate": 8.606189796476994e-06, + "loss": 0.2194, + "step": 50206 + }, + { + "epoch": 3.4839358823121227, + "grad_norm": 1.4548665285110474, + "learning_rate": 8.603909605415695e-06, + "loss": 0.3176, + "step": 50207 + }, + { + "epoch": 3.4840052737492195, + "grad_norm": 0.8675327301025391, + "learning_rate": 8.601629702881275e-06, + "loss": 0.2509, + "step": 50208 + }, + { + "epoch": 3.484074665186316, + "grad_norm": 0.7788988351821899, + "learning_rate": 8.599350088880876e-06, + "loss": 0.2497, + "step": 50209 + }, + { + "epoch": 3.4841440566234128, + "grad_norm": 0.9650663137435913, + "learning_rate": 8.597070763421733e-06, + "loss": 0.2595, + "step": 50210 + }, + { + "epoch": 3.484213448060509, + "grad_norm": 1.015386700630188, + "learning_rate": 8.594791726511031e-06, + "loss": 0.3087, + "step": 50211 + }, + { + "epoch": 3.484282839497606, + "grad_norm": 0.6756023168563843, + "learning_rate": 8.592512978155964e-06, + "loss": 0.1234, + "step": 50212 + }, + { + "epoch": 3.484352230934703, + "grad_norm": 1.0727334022521973, + "learning_rate": 8.59023451836375e-06, + "loss": 0.4295, + "step": 50213 + }, + { + "epoch": 3.484421622371799, + "grad_norm": 0.8061996102333069, + "learning_rate": 8.587956347141545e-06, + "loss": 0.2718, + "step": 50214 + }, + { + "epoch": 3.484491013808896, + "grad_norm": 0.7831860184669495, + "learning_rate": 8.585678464496549e-06, + "loss": 0.1371, + "step": 50215 + }, + { + "epoch": 3.4845604052459924, + "grad_norm": 1.0818806886672974, + "learning_rate": 8.583400870435964e-06, + "loss": 0.3139, + "step": 50216 + }, + { + "epoch": 3.4846297966830893, + "grad_norm": 1.350589394569397, + "learning_rate": 8.581123564966976e-06, + "loss": 0.3159, + "step": 50217 + }, + { + "epoch": 3.484699188120186, + "grad_norm": 0.9527158737182617, + "learning_rate": 8.578846548096776e-06, + "loss": 0.2154, + "step": 50218 + }, + { + "epoch": 3.4847685795572825, + "grad_norm": 1.0693899393081665, + "learning_rate": 8.576569819832548e-06, + "loss": 0.2436, + "step": 50219 + }, + { + "epoch": 3.4848379709943793, + "grad_norm": 0.7364044189453125, + "learning_rate": 8.574293380181497e-06, + "loss": 0.1722, + "step": 50220 + }, + { + "epoch": 3.484907362431476, + "grad_norm": 0.6967012882232666, + "learning_rate": 8.572017229150786e-06, + "loss": 0.1994, + "step": 50221 + }, + { + "epoch": 3.4849767538685725, + "grad_norm": 1.305169939994812, + "learning_rate": 8.569741366747596e-06, + "loss": 0.1984, + "step": 50222 + }, + { + "epoch": 3.4850461453056694, + "grad_norm": 0.7328160405158997, + "learning_rate": 8.567465792979135e-06, + "loss": 0.1987, + "step": 50223 + }, + { + "epoch": 3.4851155367427658, + "grad_norm": 1.0212187767028809, + "learning_rate": 8.565190507852571e-06, + "loss": 0.305, + "step": 50224 + }, + { + "epoch": 3.4851849281798626, + "grad_norm": 1.5312350988388062, + "learning_rate": 8.562915511375091e-06, + "loss": 0.1829, + "step": 50225 + }, + { + "epoch": 3.4852543196169594, + "grad_norm": 0.9865469932556152, + "learning_rate": 8.560640803553887e-06, + "loss": 0.2498, + "step": 50226 + }, + { + "epoch": 3.485323711054056, + "grad_norm": 0.7517971992492676, + "learning_rate": 8.558366384396143e-06, + "loss": 0.2402, + "step": 50227 + }, + { + "epoch": 3.4853931024911526, + "grad_norm": 1.105216145515442, + "learning_rate": 8.55609225390901e-06, + "loss": 0.1645, + "step": 50228 + }, + { + "epoch": 3.4854624939282495, + "grad_norm": 0.7844298481941223, + "learning_rate": 8.55381841209968e-06, + "loss": 0.2449, + "step": 50229 + }, + { + "epoch": 3.485531885365346, + "grad_norm": 0.8612788319587708, + "learning_rate": 8.551544858975336e-06, + "loss": 0.2114, + "step": 50230 + }, + { + "epoch": 3.4856012768024427, + "grad_norm": 0.6368989944458008, + "learning_rate": 8.549271594543163e-06, + "loss": 0.2336, + "step": 50231 + }, + { + "epoch": 3.485670668239539, + "grad_norm": 0.7998385429382324, + "learning_rate": 8.54699861881032e-06, + "loss": 0.2003, + "step": 50232 + }, + { + "epoch": 3.485740059676636, + "grad_norm": 1.1269451379776, + "learning_rate": 8.544725931784004e-06, + "loss": 0.3889, + "step": 50233 + }, + { + "epoch": 3.4858094511137327, + "grad_norm": 0.7582205533981323, + "learning_rate": 8.542453533471384e-06, + "loss": 0.1743, + "step": 50234 + }, + { + "epoch": 3.485878842550829, + "grad_norm": 0.9728862643241882, + "learning_rate": 8.540181423879612e-06, + "loss": 0.2425, + "step": 50235 + }, + { + "epoch": 3.485948233987926, + "grad_norm": 0.6742426753044128, + "learning_rate": 8.53790960301587e-06, + "loss": 0.2194, + "step": 50236 + }, + { + "epoch": 3.4860176254250224, + "grad_norm": 1.3608341217041016, + "learning_rate": 8.535638070887341e-06, + "loss": 0.1606, + "step": 50237 + }, + { + "epoch": 3.486087016862119, + "grad_norm": 0.8555762767791748, + "learning_rate": 8.5333668275012e-06, + "loss": 0.2477, + "step": 50238 + }, + { + "epoch": 3.486156408299216, + "grad_norm": 1.1150665283203125, + "learning_rate": 8.531095872864603e-06, + "loss": 0.1781, + "step": 50239 + }, + { + "epoch": 3.4862257997363124, + "grad_norm": 0.7714614868164062, + "learning_rate": 8.528825206984736e-06, + "loss": 0.2199, + "step": 50240 + }, + { + "epoch": 3.4862951911734092, + "grad_norm": 0.8170703053474426, + "learning_rate": 8.526554829868737e-06, + "loss": 0.2383, + "step": 50241 + }, + { + "epoch": 3.4863645826105056, + "grad_norm": 1.2012546062469482, + "learning_rate": 8.524284741523791e-06, + "loss": 0.2307, + "step": 50242 + }, + { + "epoch": 3.4864339740476025, + "grad_norm": 0.8756215572357178, + "learning_rate": 8.522014941957091e-06, + "loss": 0.1683, + "step": 50243 + }, + { + "epoch": 3.4865033654846993, + "grad_norm": 1.3386361598968506, + "learning_rate": 8.519745431175753e-06, + "loss": 0.3753, + "step": 50244 + }, + { + "epoch": 3.4865727569217957, + "grad_norm": 0.8803573250770569, + "learning_rate": 8.51747620918697e-06, + "loss": 0.1633, + "step": 50245 + }, + { + "epoch": 3.4866421483588925, + "grad_norm": 0.7576704621315002, + "learning_rate": 8.515207275997893e-06, + "loss": 0.1724, + "step": 50246 + }, + { + "epoch": 3.4867115397959894, + "grad_norm": 0.7697163224220276, + "learning_rate": 8.512938631615698e-06, + "loss": 0.2507, + "step": 50247 + }, + { + "epoch": 3.4867809312330857, + "grad_norm": 0.6912786364555359, + "learning_rate": 8.51067027604756e-06, + "loss": 0.2089, + "step": 50248 + }, + { + "epoch": 3.4868503226701826, + "grad_norm": 0.8399727940559387, + "learning_rate": 8.508402209300593e-06, + "loss": 0.1743, + "step": 50249 + }, + { + "epoch": 3.4869197141072794, + "grad_norm": 0.9181108474731445, + "learning_rate": 8.506134431381996e-06, + "loss": 0.2924, + "step": 50250 + }, + { + "epoch": 3.486989105544376, + "grad_norm": 0.6626285314559937, + "learning_rate": 8.503866942298911e-06, + "loss": 0.2429, + "step": 50251 + }, + { + "epoch": 3.4870584969814726, + "grad_norm": 0.6597544550895691, + "learning_rate": 8.501599742058508e-06, + "loss": 0.2001, + "step": 50252 + }, + { + "epoch": 3.487127888418569, + "grad_norm": 1.089357614517212, + "learning_rate": 8.499332830667938e-06, + "loss": 0.2015, + "step": 50253 + }, + { + "epoch": 3.487197279855666, + "grad_norm": 1.2103592157363892, + "learning_rate": 8.497066208134375e-06, + "loss": 0.3718, + "step": 50254 + }, + { + "epoch": 3.4872666712927627, + "grad_norm": 0.7058897018432617, + "learning_rate": 8.494799874464942e-06, + "loss": 0.217, + "step": 50255 + }, + { + "epoch": 3.487336062729859, + "grad_norm": 0.9625611305236816, + "learning_rate": 8.492533829666804e-06, + "loss": 0.2433, + "step": 50256 + }, + { + "epoch": 3.487405454166956, + "grad_norm": 0.8824744820594788, + "learning_rate": 8.490268073747122e-06, + "loss": 0.1993, + "step": 50257 + }, + { + "epoch": 3.4874748456040523, + "grad_norm": 0.9682384133338928, + "learning_rate": 8.488002606713052e-06, + "loss": 0.293, + "step": 50258 + }, + { + "epoch": 3.487544237041149, + "grad_norm": 0.8213931322097778, + "learning_rate": 8.485737428571727e-06, + "loss": 0.2258, + "step": 50259 + }, + { + "epoch": 3.487613628478246, + "grad_norm": 0.8194085359573364, + "learning_rate": 8.483472539330329e-06, + "loss": 0.2701, + "step": 50260 + }, + { + "epoch": 3.4876830199153424, + "grad_norm": 0.5882789492607117, + "learning_rate": 8.481207938995994e-06, + "loss": 0.1282, + "step": 50261 + }, + { + "epoch": 3.487752411352439, + "grad_norm": 0.6213819980621338, + "learning_rate": 8.478943627575852e-06, + "loss": 0.202, + "step": 50262 + }, + { + "epoch": 3.4878218027895356, + "grad_norm": 0.8146981596946716, + "learning_rate": 8.476679605077075e-06, + "loss": 0.2741, + "step": 50263 + }, + { + "epoch": 3.4878911942266324, + "grad_norm": 0.8558934926986694, + "learning_rate": 8.474415871506791e-06, + "loss": 0.2963, + "step": 50264 + }, + { + "epoch": 3.4879605856637292, + "grad_norm": 0.7513675689697266, + "learning_rate": 8.47215242687216e-06, + "loss": 0.2441, + "step": 50265 + }, + { + "epoch": 3.4880299771008256, + "grad_norm": 0.6282334327697754, + "learning_rate": 8.46988927118033e-06, + "loss": 0.2073, + "step": 50266 + }, + { + "epoch": 3.4880993685379225, + "grad_norm": 0.9327831268310547, + "learning_rate": 8.467626404438433e-06, + "loss": 0.4229, + "step": 50267 + }, + { + "epoch": 3.488168759975019, + "grad_norm": 0.8526215553283691, + "learning_rate": 8.465363826653649e-06, + "loss": 0.2443, + "step": 50268 + }, + { + "epoch": 3.4882381514121157, + "grad_norm": 1.3524774312973022, + "learning_rate": 8.463101537833063e-06, + "loss": 0.2893, + "step": 50269 + }, + { + "epoch": 3.4883075428492125, + "grad_norm": 1.0337040424346924, + "learning_rate": 8.460839537983844e-06, + "loss": 0.2515, + "step": 50270 + }, + { + "epoch": 3.488376934286309, + "grad_norm": 0.9994046688079834, + "learning_rate": 8.458577827113146e-06, + "loss": 0.2423, + "step": 50271 + }, + { + "epoch": 3.4884463257234057, + "grad_norm": 1.0342662334442139, + "learning_rate": 8.456316405228081e-06, + "loss": 0.2578, + "step": 50272 + }, + { + "epoch": 3.4885157171605026, + "grad_norm": 0.9081639647483826, + "learning_rate": 8.454055272335815e-06, + "loss": 0.2944, + "step": 50273 + }, + { + "epoch": 3.488585108597599, + "grad_norm": 0.9074745774269104, + "learning_rate": 8.451794428443493e-06, + "loss": 0.2425, + "step": 50274 + }, + { + "epoch": 3.488654500034696, + "grad_norm": 0.5622698068618774, + "learning_rate": 8.449533873558212e-06, + "loss": 0.1001, + "step": 50275 + }, + { + "epoch": 3.4887238914717926, + "grad_norm": 1.2158985137939453, + "learning_rate": 8.447273607687134e-06, + "loss": 0.2934, + "step": 50276 + }, + { + "epoch": 3.488793282908889, + "grad_norm": 0.9368364810943604, + "learning_rate": 8.445013630837395e-06, + "loss": 0.2907, + "step": 50277 + }, + { + "epoch": 3.488862674345986, + "grad_norm": 0.6490146517753601, + "learning_rate": 8.442753943016135e-06, + "loss": 0.157, + "step": 50278 + }, + { + "epoch": 3.4889320657830822, + "grad_norm": 1.0444018840789795, + "learning_rate": 8.44049454423046e-06, + "loss": 0.2427, + "step": 50279 + }, + { + "epoch": 3.489001457220179, + "grad_norm": 0.9116461873054504, + "learning_rate": 8.43823543448753e-06, + "loss": 0.2065, + "step": 50280 + }, + { + "epoch": 3.489070848657276, + "grad_norm": 1.432958960533142, + "learning_rate": 8.435976613794461e-06, + "loss": 0.1991, + "step": 50281 + }, + { + "epoch": 3.4891402400943723, + "grad_norm": 1.1171818971633911, + "learning_rate": 8.433718082158404e-06, + "loss": 0.3749, + "step": 50282 + }, + { + "epoch": 3.489209631531469, + "grad_norm": 0.850837230682373, + "learning_rate": 8.431459839586464e-06, + "loss": 0.2987, + "step": 50283 + }, + { + "epoch": 3.4892790229685655, + "grad_norm": 0.6658048033714294, + "learning_rate": 8.42920188608578e-06, + "loss": 0.1871, + "step": 50284 + }, + { + "epoch": 3.4893484144056623, + "grad_norm": 1.1901592016220093, + "learning_rate": 8.42694422166348e-06, + "loss": 0.2912, + "step": 50285 + }, + { + "epoch": 3.489417805842759, + "grad_norm": 0.8072784543037415, + "learning_rate": 8.424686846326702e-06, + "loss": 0.2666, + "step": 50286 + }, + { + "epoch": 3.4894871972798556, + "grad_norm": 1.0368658304214478, + "learning_rate": 8.42242976008255e-06, + "loss": 0.2738, + "step": 50287 + }, + { + "epoch": 3.4895565887169524, + "grad_norm": 0.9601749181747437, + "learning_rate": 8.420172962938167e-06, + "loss": 0.2316, + "step": 50288 + }, + { + "epoch": 3.489625980154049, + "grad_norm": 0.9169789552688599, + "learning_rate": 8.417916454900698e-06, + "loss": 0.25, + "step": 50289 + }, + { + "epoch": 3.4896953715911456, + "grad_norm": 0.7330768704414368, + "learning_rate": 8.415660235977218e-06, + "loss": 0.2043, + "step": 50290 + }, + { + "epoch": 3.4897647630282425, + "grad_norm": 0.9160228967666626, + "learning_rate": 8.413404306174866e-06, + "loss": 0.3328, + "step": 50291 + }, + { + "epoch": 3.489834154465339, + "grad_norm": 0.7584800720214844, + "learning_rate": 8.411148665500779e-06, + "loss": 0.2849, + "step": 50292 + }, + { + "epoch": 3.4899035459024357, + "grad_norm": 1.0815094709396362, + "learning_rate": 8.408893313962063e-06, + "loss": 0.3504, + "step": 50293 + }, + { + "epoch": 3.489972937339532, + "grad_norm": 0.9514367580413818, + "learning_rate": 8.406638251565845e-06, + "loss": 0.2001, + "step": 50294 + }, + { + "epoch": 3.490042328776629, + "grad_norm": 0.9102886319160461, + "learning_rate": 8.404383478319255e-06, + "loss": 0.1791, + "step": 50295 + }, + { + "epoch": 3.4901117202137257, + "grad_norm": 1.094558835029602, + "learning_rate": 8.402128994229408e-06, + "loss": 0.301, + "step": 50296 + }, + { + "epoch": 3.490181111650822, + "grad_norm": 0.9081610441207886, + "learning_rate": 8.399874799303387e-06, + "loss": 0.2108, + "step": 50297 + }, + { + "epoch": 3.490250503087919, + "grad_norm": 0.647994875907898, + "learning_rate": 8.397620893548341e-06, + "loss": 0.2373, + "step": 50298 + }, + { + "epoch": 3.490319894525016, + "grad_norm": 0.7502053380012512, + "learning_rate": 8.395367276971377e-06, + "loss": 0.1631, + "step": 50299 + }, + { + "epoch": 3.490389285962112, + "grad_norm": 0.7694737911224365, + "learning_rate": 8.393113949579613e-06, + "loss": 0.2095, + "step": 50300 + }, + { + "epoch": 3.490458677399209, + "grad_norm": 0.7744950652122498, + "learning_rate": 8.39086091138016e-06, + "loss": 0.3245, + "step": 50301 + }, + { + "epoch": 3.490528068836306, + "grad_norm": 0.7561649680137634, + "learning_rate": 8.388608162380129e-06, + "loss": 0.3152, + "step": 50302 + }, + { + "epoch": 3.4905974602734022, + "grad_norm": 1.7714329957962036, + "learning_rate": 8.386355702586657e-06, + "loss": 0.2389, + "step": 50303 + }, + { + "epoch": 3.490666851710499, + "grad_norm": 1.4533405303955078, + "learning_rate": 8.384103532006805e-06, + "loss": 0.1892, + "step": 50304 + }, + { + "epoch": 3.4907362431475955, + "grad_norm": 0.6963905692100525, + "learning_rate": 8.381851650647709e-06, + "loss": 0.224, + "step": 50305 + }, + { + "epoch": 3.4908056345846923, + "grad_norm": 0.8056900501251221, + "learning_rate": 8.379600058516479e-06, + "loss": 0.2702, + "step": 50306 + }, + { + "epoch": 3.490875026021789, + "grad_norm": 0.5765171051025391, + "learning_rate": 8.377348755620228e-06, + "loss": 0.1043, + "step": 50307 + }, + { + "epoch": 3.4909444174588855, + "grad_norm": 2.009396553039551, + "learning_rate": 8.375097741966043e-06, + "loss": 0.2501, + "step": 50308 + }, + { + "epoch": 3.4910138088959823, + "grad_norm": 0.7592931389808655, + "learning_rate": 8.372847017561069e-06, + "loss": 0.2004, + "step": 50309 + }, + { + "epoch": 3.4910832003330787, + "grad_norm": 1.1222379207611084, + "learning_rate": 8.370596582412372e-06, + "loss": 0.315, + "step": 50310 + }, + { + "epoch": 3.4911525917701756, + "grad_norm": 0.6451804637908936, + "learning_rate": 8.368346436527074e-06, + "loss": 0.177, + "step": 50311 + }, + { + "epoch": 3.4912219832072724, + "grad_norm": 1.0273380279541016, + "learning_rate": 8.366096579912264e-06, + "loss": 0.3304, + "step": 50312 + }, + { + "epoch": 3.491291374644369, + "grad_norm": 0.7277399301528931, + "learning_rate": 8.363847012575076e-06, + "loss": 0.2643, + "step": 50313 + }, + { + "epoch": 3.4913607660814656, + "grad_norm": 0.8089639544487, + "learning_rate": 8.361597734522575e-06, + "loss": 0.2072, + "step": 50314 + }, + { + "epoch": 3.491430157518562, + "grad_norm": 1.210545539855957, + "learning_rate": 8.359348745761886e-06, + "loss": 0.2761, + "step": 50315 + }, + { + "epoch": 3.491499548955659, + "grad_norm": 0.6297590136528015, + "learning_rate": 8.357100046300093e-06, + "loss": 0.2092, + "step": 50316 + }, + { + "epoch": 3.4915689403927557, + "grad_norm": 1.067047119140625, + "learning_rate": 8.354851636144323e-06, + "loss": 0.3637, + "step": 50317 + }, + { + "epoch": 3.491638331829852, + "grad_norm": 1.1636384725570679, + "learning_rate": 8.352603515301626e-06, + "loss": 0.3177, + "step": 50318 + }, + { + "epoch": 3.491707723266949, + "grad_norm": 1.1184922456741333, + "learning_rate": 8.350355683779143e-06, + "loss": 0.3166, + "step": 50319 + }, + { + "epoch": 3.4917771147040453, + "grad_norm": 1.234220027923584, + "learning_rate": 8.348108141583943e-06, + "loss": 0.2994, + "step": 50320 + }, + { + "epoch": 3.491846506141142, + "grad_norm": 0.8392522931098938, + "learning_rate": 8.345860888723134e-06, + "loss": 0.1949, + "step": 50321 + }, + { + "epoch": 3.491915897578239, + "grad_norm": 0.975615918636322, + "learning_rate": 8.343613925203808e-06, + "loss": 0.2942, + "step": 50322 + }, + { + "epoch": 3.4919852890153353, + "grad_norm": 1.0530695915222168, + "learning_rate": 8.341367251033062e-06, + "loss": 0.2167, + "step": 50323 + }, + { + "epoch": 3.492054680452432, + "grad_norm": 1.1633453369140625, + "learning_rate": 8.339120866217997e-06, + "loss": 0.3578, + "step": 50324 + }, + { + "epoch": 3.492124071889529, + "grad_norm": 1.1282917261123657, + "learning_rate": 8.33687477076568e-06, + "loss": 0.2672, + "step": 50325 + }, + { + "epoch": 3.4921934633266254, + "grad_norm": 1.1142239570617676, + "learning_rate": 8.334628964683211e-06, + "loss": 0.3646, + "step": 50326 + }, + { + "epoch": 3.4922628547637222, + "grad_norm": 1.0580394268035889, + "learning_rate": 8.332383447977688e-06, + "loss": 0.298, + "step": 50327 + }, + { + "epoch": 3.492332246200819, + "grad_norm": 0.8958662748336792, + "learning_rate": 8.330138220656192e-06, + "loss": 0.3174, + "step": 50328 + }, + { + "epoch": 3.4924016376379154, + "grad_norm": 1.116720199584961, + "learning_rate": 8.327893282725808e-06, + "loss": 0.286, + "step": 50329 + }, + { + "epoch": 3.4924710290750123, + "grad_norm": 0.6659228205680847, + "learning_rate": 8.32564863419365e-06, + "loss": 0.154, + "step": 50330 + }, + { + "epoch": 3.4925404205121087, + "grad_norm": 0.8414700627326965, + "learning_rate": 8.32340427506677e-06, + "loss": 0.2476, + "step": 50331 + }, + { + "epoch": 3.4926098119492055, + "grad_norm": 1.0676864385604858, + "learning_rate": 8.321160205352264e-06, + "loss": 0.2358, + "step": 50332 + }, + { + "epoch": 3.4926792033863023, + "grad_norm": 0.7556572556495667, + "learning_rate": 8.318916425057221e-06, + "loss": 0.2286, + "step": 50333 + }, + { + "epoch": 3.4927485948233987, + "grad_norm": 0.7107539772987366, + "learning_rate": 8.316672934188719e-06, + "loss": 0.1768, + "step": 50334 + }, + { + "epoch": 3.4928179862604956, + "grad_norm": 0.901868999004364, + "learning_rate": 8.314429732753848e-06, + "loss": 0.3156, + "step": 50335 + }, + { + "epoch": 3.492887377697592, + "grad_norm": 0.9075155258178711, + "learning_rate": 8.31218682075967e-06, + "loss": 0.2226, + "step": 50336 + }, + { + "epoch": 3.4929567691346888, + "grad_norm": 0.731522798538208, + "learning_rate": 8.309944198213315e-06, + "loss": 0.2915, + "step": 50337 + }, + { + "epoch": 3.4930261605717856, + "grad_norm": 1.0014108419418335, + "learning_rate": 8.307701865121808e-06, + "loss": 0.3307, + "step": 50338 + }, + { + "epoch": 3.493095552008882, + "grad_norm": 0.9768662452697754, + "learning_rate": 8.305459821492246e-06, + "loss": 0.3579, + "step": 50339 + }, + { + "epoch": 3.493164943445979, + "grad_norm": 0.9918225407600403, + "learning_rate": 8.303218067331697e-06, + "loss": 0.2034, + "step": 50340 + }, + { + "epoch": 3.493234334883075, + "grad_norm": 1.038987636566162, + "learning_rate": 8.300976602647259e-06, + "loss": 0.2099, + "step": 50341 + }, + { + "epoch": 3.493303726320172, + "grad_norm": 0.7251759171485901, + "learning_rate": 8.298735427446003e-06, + "loss": 0.2134, + "step": 50342 + }, + { + "epoch": 3.493373117757269, + "grad_norm": 1.0146942138671875, + "learning_rate": 8.296494541735e-06, + "loss": 0.2437, + "step": 50343 + }, + { + "epoch": 3.4934425091943653, + "grad_norm": 1.0019978284835815, + "learning_rate": 8.294253945521335e-06, + "loss": 0.2606, + "step": 50344 + }, + { + "epoch": 3.493511900631462, + "grad_norm": 0.791731059551239, + "learning_rate": 8.292013638812057e-06, + "loss": 0.2075, + "step": 50345 + }, + { + "epoch": 3.4935812920685585, + "grad_norm": 0.778976321220398, + "learning_rate": 8.28977362161425e-06, + "loss": 0.2677, + "step": 50346 + }, + { + "epoch": 3.4936506835056553, + "grad_norm": 0.9498056769371033, + "learning_rate": 8.287533893934996e-06, + "loss": 0.2336, + "step": 50347 + }, + { + "epoch": 3.493720074942752, + "grad_norm": 0.9221312403678894, + "learning_rate": 8.285294455781367e-06, + "loss": 0.1542, + "step": 50348 + }, + { + "epoch": 3.4937894663798486, + "grad_norm": 1.1655266284942627, + "learning_rate": 8.283055307160404e-06, + "loss": 0.2866, + "step": 50349 + }, + { + "epoch": 3.4938588578169454, + "grad_norm": 1.2365052700042725, + "learning_rate": 8.280816448079187e-06, + "loss": 0.2452, + "step": 50350 + }, + { + "epoch": 3.493928249254042, + "grad_norm": 0.9698172211647034, + "learning_rate": 8.278577878544824e-06, + "loss": 0.1924, + "step": 50351 + }, + { + "epoch": 3.4939976406911386, + "grad_norm": 1.0178210735321045, + "learning_rate": 8.27633959856432e-06, + "loss": 0.394, + "step": 50352 + }, + { + "epoch": 3.4940670321282354, + "grad_norm": 1.2457399368286133, + "learning_rate": 8.274101608144779e-06, + "loss": 0.4918, + "step": 50353 + }, + { + "epoch": 3.4941364235653323, + "grad_norm": 0.9576226472854614, + "learning_rate": 8.271863907293243e-06, + "loss": 0.3312, + "step": 50354 + }, + { + "epoch": 3.4942058150024287, + "grad_norm": 0.793731689453125, + "learning_rate": 8.269626496016803e-06, + "loss": 0.2038, + "step": 50355 + }, + { + "epoch": 3.4942752064395255, + "grad_norm": 0.6978375911712646, + "learning_rate": 8.267389374322509e-06, + "loss": 0.1945, + "step": 50356 + }, + { + "epoch": 3.494344597876622, + "grad_norm": 0.9280793070793152, + "learning_rate": 8.265152542217413e-06, + "loss": 0.2478, + "step": 50357 + }, + { + "epoch": 3.4944139893137187, + "grad_norm": 0.907071590423584, + "learning_rate": 8.26291599970862e-06, + "loss": 0.151, + "step": 50358 + }, + { + "epoch": 3.4944833807508155, + "grad_norm": 1.1708954572677612, + "learning_rate": 8.260679746803135e-06, + "loss": 0.3204, + "step": 50359 + }, + { + "epoch": 3.494552772187912, + "grad_norm": 0.9591355919837952, + "learning_rate": 8.25844378350803e-06, + "loss": 0.2238, + "step": 50360 + }, + { + "epoch": 3.4946221636250088, + "grad_norm": 0.8016872406005859, + "learning_rate": 8.256208109830387e-06, + "loss": 0.2656, + "step": 50361 + }, + { + "epoch": 3.494691555062105, + "grad_norm": 0.8287999033927917, + "learning_rate": 8.253972725777248e-06, + "loss": 0.1788, + "step": 50362 + }, + { + "epoch": 3.494760946499202, + "grad_norm": 0.8555790781974792, + "learning_rate": 8.25173763135566e-06, + "loss": 0.2336, + "step": 50363 + }, + { + "epoch": 3.494830337936299, + "grad_norm": 0.9715436100959778, + "learning_rate": 8.249502826572709e-06, + "loss": 0.2472, + "step": 50364 + }, + { + "epoch": 3.494899729373395, + "grad_norm": 1.269709587097168, + "learning_rate": 8.247268311435442e-06, + "loss": 0.2315, + "step": 50365 + }, + { + "epoch": 3.494969120810492, + "grad_norm": 0.7475132942199707, + "learning_rate": 8.245034085950887e-06, + "loss": 0.2061, + "step": 50366 + }, + { + "epoch": 3.4950385122475884, + "grad_norm": 0.7083020806312561, + "learning_rate": 8.242800150126117e-06, + "loss": 0.2091, + "step": 50367 + }, + { + "epoch": 3.4951079036846853, + "grad_norm": 0.7541873455047607, + "learning_rate": 8.240566503968172e-06, + "loss": 0.118, + "step": 50368 + }, + { + "epoch": 3.495177295121782, + "grad_norm": 0.8453771471977234, + "learning_rate": 8.238333147484124e-06, + "loss": 0.1815, + "step": 50369 + }, + { + "epoch": 3.4952466865588785, + "grad_norm": 0.9531381130218506, + "learning_rate": 8.236100080680997e-06, + "loss": 0.2676, + "step": 50370 + }, + { + "epoch": 3.4953160779959753, + "grad_norm": 1.1109918355941772, + "learning_rate": 8.233867303565867e-06, + "loss": 0.4309, + "step": 50371 + }, + { + "epoch": 3.4953854694330717, + "grad_norm": 0.8217290639877319, + "learning_rate": 8.231634816145795e-06, + "loss": 0.1845, + "step": 50372 + }, + { + "epoch": 3.4954548608701685, + "grad_norm": 0.716674268245697, + "learning_rate": 8.229402618427772e-06, + "loss": 0.1623, + "step": 50373 + }, + { + "epoch": 3.4955242523072654, + "grad_norm": 0.9123156070709229, + "learning_rate": 8.227170710418885e-06, + "loss": 0.2247, + "step": 50374 + }, + { + "epoch": 3.4955936437443618, + "grad_norm": 0.8569532632827759, + "learning_rate": 8.224939092126171e-06, + "loss": 0.2273, + "step": 50375 + }, + { + "epoch": 3.4956630351814586, + "grad_norm": 0.6539653539657593, + "learning_rate": 8.22270776355668e-06, + "loss": 0.113, + "step": 50376 + }, + { + "epoch": 3.4957324266185554, + "grad_norm": 1.1128242015838623, + "learning_rate": 8.220476724717452e-06, + "loss": 0.3073, + "step": 50377 + }, + { + "epoch": 3.495801818055652, + "grad_norm": 0.9985487461090088, + "learning_rate": 8.218245975615524e-06, + "loss": 0.173, + "step": 50378 + }, + { + "epoch": 3.4958712094927487, + "grad_norm": 0.7626644968986511, + "learning_rate": 8.216015516257969e-06, + "loss": 0.2465, + "step": 50379 + }, + { + "epoch": 3.4959406009298455, + "grad_norm": 0.7603464722633362, + "learning_rate": 8.213785346651781e-06, + "loss": 0.2337, + "step": 50380 + }, + { + "epoch": 3.496009992366942, + "grad_norm": 0.9732644557952881, + "learning_rate": 8.211555466804021e-06, + "loss": 0.219, + "step": 50381 + }, + { + "epoch": 3.4960793838040387, + "grad_norm": 0.9331871271133423, + "learning_rate": 8.209325876721741e-06, + "loss": 0.2942, + "step": 50382 + }, + { + "epoch": 3.496148775241135, + "grad_norm": 0.6296330690383911, + "learning_rate": 8.207096576411988e-06, + "loss": 0.1073, + "step": 50383 + }, + { + "epoch": 3.496218166678232, + "grad_norm": 0.9504103660583496, + "learning_rate": 8.204867565881758e-06, + "loss": 0.2138, + "step": 50384 + }, + { + "epoch": 3.4962875581153288, + "grad_norm": 1.2997426986694336, + "learning_rate": 8.202638845138111e-06, + "loss": 0.3892, + "step": 50385 + }, + { + "epoch": 3.496356949552425, + "grad_norm": 0.770826518535614, + "learning_rate": 8.200410414188097e-06, + "loss": 0.178, + "step": 50386 + }, + { + "epoch": 3.496426340989522, + "grad_norm": 1.0551782846450806, + "learning_rate": 8.198182273038724e-06, + "loss": 0.2035, + "step": 50387 + }, + { + "epoch": 3.4964957324266184, + "grad_norm": 0.6845373511314392, + "learning_rate": 8.195954421697038e-06, + "loss": 0.195, + "step": 50388 + }, + { + "epoch": 3.496565123863715, + "grad_norm": 0.8528683185577393, + "learning_rate": 8.19372686017007e-06, + "loss": 0.2765, + "step": 50389 + }, + { + "epoch": 3.496634515300812, + "grad_norm": 1.2387388944625854, + "learning_rate": 8.191499588464858e-06, + "loss": 0.274, + "step": 50390 + }, + { + "epoch": 3.4967039067379084, + "grad_norm": 0.7494475245475769, + "learning_rate": 8.189272606588427e-06, + "loss": 0.2333, + "step": 50391 + }, + { + "epoch": 3.4967732981750053, + "grad_norm": 0.788806676864624, + "learning_rate": 8.187045914547808e-06, + "loss": 0.1564, + "step": 50392 + }, + { + "epoch": 3.4968426896121017, + "grad_norm": 1.017333984375, + "learning_rate": 8.18481951235005e-06, + "loss": 0.2119, + "step": 50393 + }, + { + "epoch": 3.4969120810491985, + "grad_norm": 0.6864269375801086, + "learning_rate": 8.182593400002159e-06, + "loss": 0.198, + "step": 50394 + }, + { + "epoch": 3.4969814724862953, + "grad_norm": 0.9501200914382935, + "learning_rate": 8.180367577511162e-06, + "loss": 0.2724, + "step": 50395 + }, + { + "epoch": 3.4970508639233917, + "grad_norm": 1.1569229364395142, + "learning_rate": 8.178142044884085e-06, + "loss": 0.1917, + "step": 50396 + }, + { + "epoch": 3.4971202553604885, + "grad_norm": 0.7975890636444092, + "learning_rate": 8.17591680212796e-06, + "loss": 0.2216, + "step": 50397 + }, + { + "epoch": 3.4971896467975854, + "grad_norm": 1.1634924411773682, + "learning_rate": 8.17369184924982e-06, + "loss": 0.1518, + "step": 50398 + }, + { + "epoch": 3.4972590382346818, + "grad_norm": 0.9193847179412842, + "learning_rate": 8.171467186256676e-06, + "loss": 0.2642, + "step": 50399 + }, + { + "epoch": 3.4973284296717786, + "grad_norm": 1.2770960330963135, + "learning_rate": 8.169242813155576e-06, + "loss": 0.1402, + "step": 50400 + }, + { + "epoch": 3.497397821108875, + "grad_norm": 0.81501704454422, + "learning_rate": 8.167018729953501e-06, + "loss": 0.2095, + "step": 50401 + }, + { + "epoch": 3.497467212545972, + "grad_norm": 0.6151701807975769, + "learning_rate": 8.164794936657494e-06, + "loss": 0.2615, + "step": 50402 + }, + { + "epoch": 3.4975366039830686, + "grad_norm": 1.1297996044158936, + "learning_rate": 8.162571433274579e-06, + "loss": 0.244, + "step": 50403 + }, + { + "epoch": 3.497605995420165, + "grad_norm": 2.6627140045166016, + "learning_rate": 8.160348219811763e-06, + "loss": 0.1777, + "step": 50404 + }, + { + "epoch": 3.497675386857262, + "grad_norm": 1.0910978317260742, + "learning_rate": 8.158125296276086e-06, + "loss": 0.2302, + "step": 50405 + }, + { + "epoch": 3.4977447782943587, + "grad_norm": 0.7592564821243286, + "learning_rate": 8.155902662674564e-06, + "loss": 0.2613, + "step": 50406 + }, + { + "epoch": 3.497814169731455, + "grad_norm": 0.8674781322479248, + "learning_rate": 8.153680319014178e-06, + "loss": 0.2352, + "step": 50407 + }, + { + "epoch": 3.497883561168552, + "grad_norm": 0.8274816274642944, + "learning_rate": 8.15145826530197e-06, + "loss": 0.2334, + "step": 50408 + }, + { + "epoch": 3.4979529526056483, + "grad_norm": 1.022236943244934, + "learning_rate": 8.149236501544954e-06, + "loss": 0.2492, + "step": 50409 + }, + { + "epoch": 3.498022344042745, + "grad_norm": 0.7964975833892822, + "learning_rate": 8.147015027750137e-06, + "loss": 0.2184, + "step": 50410 + }, + { + "epoch": 3.498091735479842, + "grad_norm": 1.6839165687561035, + "learning_rate": 8.144793843924536e-06, + "loss": 0.4364, + "step": 50411 + }, + { + "epoch": 3.4981611269169384, + "grad_norm": 0.6435463428497314, + "learning_rate": 8.142572950075167e-06, + "loss": 0.2366, + "step": 50412 + }, + { + "epoch": 3.498230518354035, + "grad_norm": 0.7617978453636169, + "learning_rate": 8.140352346209057e-06, + "loss": 0.2469, + "step": 50413 + }, + { + "epoch": 3.4982999097911316, + "grad_norm": 1.0799777507781982, + "learning_rate": 8.138132032333179e-06, + "loss": 0.296, + "step": 50414 + }, + { + "epoch": 3.4983693012282284, + "grad_norm": 0.5600749850273132, + "learning_rate": 8.13591200845456e-06, + "loss": 0.1612, + "step": 50415 + }, + { + "epoch": 3.4984386926653253, + "grad_norm": 1.020338535308838, + "learning_rate": 8.133692274580207e-06, + "loss": 0.2494, + "step": 50416 + }, + { + "epoch": 3.4985080841024216, + "grad_norm": 0.8858692646026611, + "learning_rate": 8.131472830717146e-06, + "loss": 0.2683, + "step": 50417 + }, + { + "epoch": 3.4985774755395185, + "grad_norm": 0.9602231979370117, + "learning_rate": 8.129253676872351e-06, + "loss": 0.3853, + "step": 50418 + }, + { + "epoch": 3.498646866976615, + "grad_norm": 0.7880119681358337, + "learning_rate": 8.127034813052836e-06, + "loss": 0.2335, + "step": 50419 + }, + { + "epoch": 3.4987162584137117, + "grad_norm": 0.7149924635887146, + "learning_rate": 8.124816239265621e-06, + "loss": 0.1978, + "step": 50420 + }, + { + "epoch": 3.4987856498508085, + "grad_norm": 0.9833321571350098, + "learning_rate": 8.12259795551772e-06, + "loss": 0.2784, + "step": 50421 + }, + { + "epoch": 3.498855041287905, + "grad_norm": 1.0624083280563354, + "learning_rate": 8.120379961816093e-06, + "loss": 0.2636, + "step": 50422 + }, + { + "epoch": 3.4989244327250018, + "grad_norm": 0.90647953748703, + "learning_rate": 8.118162258167772e-06, + "loss": 0.3104, + "step": 50423 + }, + { + "epoch": 3.4989938241620986, + "grad_norm": 1.214655876159668, + "learning_rate": 8.115944844579748e-06, + "loss": 0.3072, + "step": 50424 + }, + { + "epoch": 3.499063215599195, + "grad_norm": 1.4350627660751343, + "learning_rate": 8.113727721059028e-06, + "loss": 0.4792, + "step": 50425 + }, + { + "epoch": 3.499132607036292, + "grad_norm": 0.7636995911598206, + "learning_rate": 8.111510887612616e-06, + "loss": 0.2038, + "step": 50426 + }, + { + "epoch": 3.499201998473388, + "grad_norm": 1.252998948097229, + "learning_rate": 8.10929434424751e-06, + "loss": 0.3667, + "step": 50427 + }, + { + "epoch": 3.499271389910485, + "grad_norm": 0.7643749713897705, + "learning_rate": 8.107078090970689e-06, + "loss": 0.1629, + "step": 50428 + }, + { + "epoch": 3.499340781347582, + "grad_norm": 0.6022384762763977, + "learning_rate": 8.104862127789159e-06, + "loss": 0.1647, + "step": 50429 + }, + { + "epoch": 3.4994101727846783, + "grad_norm": 0.949867308139801, + "learning_rate": 8.102646454709917e-06, + "loss": 0.2144, + "step": 50430 + }, + { + "epoch": 3.499479564221775, + "grad_norm": 0.9488383531570435, + "learning_rate": 8.100431071739956e-06, + "loss": 0.2674, + "step": 50431 + }, + { + "epoch": 3.499548955658872, + "grad_norm": 0.9466694593429565, + "learning_rate": 8.09821597888627e-06, + "loss": 0.2342, + "step": 50432 + }, + { + "epoch": 3.4996183470959683, + "grad_norm": 1.6029038429260254, + "learning_rate": 8.096001176155864e-06, + "loss": 0.5075, + "step": 50433 + }, + { + "epoch": 3.499687738533065, + "grad_norm": 1.511516809463501, + "learning_rate": 8.093786663555725e-06, + "loss": 0.2646, + "step": 50434 + }, + { + "epoch": 3.4997571299701615, + "grad_norm": 0.8808608055114746, + "learning_rate": 8.091572441092831e-06, + "loss": 0.2213, + "step": 50435 + }, + { + "epoch": 3.4998265214072584, + "grad_norm": 1.01521897315979, + "learning_rate": 8.08935850877417e-06, + "loss": 0.263, + "step": 50436 + }, + { + "epoch": 3.499895912844355, + "grad_norm": 0.7433643341064453, + "learning_rate": 8.087144866606743e-06, + "loss": 0.2025, + "step": 50437 + }, + { + "epoch": 3.4999653042814516, + "grad_norm": 0.8419930338859558, + "learning_rate": 8.08493151459755e-06, + "loss": 0.1828, + "step": 50438 + }, + { + "epoch": 3.5000346957185484, + "grad_norm": 1.092612385749817, + "learning_rate": 8.082718452753546e-06, + "loss": 0.2235, + "step": 50439 + }, + { + "epoch": 3.500104087155645, + "grad_norm": 0.8448650240898132, + "learning_rate": 8.080505681081741e-06, + "loss": 0.278, + "step": 50440 + }, + { + "epoch": 3.5001734785927416, + "grad_norm": 1.0095444917678833, + "learning_rate": 8.078293199589137e-06, + "loss": 0.2156, + "step": 50441 + }, + { + "epoch": 3.5002428700298385, + "grad_norm": 0.6778953075408936, + "learning_rate": 8.076081008282676e-06, + "loss": 0.2111, + "step": 50442 + }, + { + "epoch": 3.500312261466935, + "grad_norm": 1.0373188257217407, + "learning_rate": 8.073869107169362e-06, + "loss": 0.1814, + "step": 50443 + }, + { + "epoch": 3.5003816529040317, + "grad_norm": 1.0653659105300903, + "learning_rate": 8.071657496256168e-06, + "loss": 0.3085, + "step": 50444 + }, + { + "epoch": 3.500451044341128, + "grad_norm": 0.8052967190742493, + "learning_rate": 8.0694461755501e-06, + "loss": 0.266, + "step": 50445 + }, + { + "epoch": 3.500520435778225, + "grad_norm": 1.0275204181671143, + "learning_rate": 8.067235145058116e-06, + "loss": 0.2372, + "step": 50446 + }, + { + "epoch": 3.5005898272153217, + "grad_norm": 0.842606782913208, + "learning_rate": 8.065024404787214e-06, + "loss": 0.224, + "step": 50447 + }, + { + "epoch": 3.500659218652418, + "grad_norm": 1.2608165740966797, + "learning_rate": 8.062813954744375e-06, + "loss": 0.3179, + "step": 50448 + }, + { + "epoch": 3.500728610089515, + "grad_norm": 0.963641345500946, + "learning_rate": 8.06060379493655e-06, + "loss": 0.1913, + "step": 50449 + }, + { + "epoch": 3.5007980015266114, + "grad_norm": 1.0296547412872314, + "learning_rate": 8.058393925370721e-06, + "loss": 0.2026, + "step": 50450 + }, + { + "epoch": 3.500867392963708, + "grad_norm": 1.2252798080444336, + "learning_rate": 8.056184346053886e-06, + "loss": 0.2964, + "step": 50451 + }, + { + "epoch": 3.500936784400805, + "grad_norm": 0.998492419719696, + "learning_rate": 8.053975056993025e-06, + "loss": 0.2598, + "step": 50452 + }, + { + "epoch": 3.501006175837902, + "grad_norm": 0.5317165851593018, + "learning_rate": 8.051766058195087e-06, + "loss": 0.1524, + "step": 50453 + }, + { + "epoch": 3.5010755672749982, + "grad_norm": 1.2412346601486206, + "learning_rate": 8.049557349667048e-06, + "loss": 0.234, + "step": 50454 + }, + { + "epoch": 3.501144958712095, + "grad_norm": 1.3063359260559082, + "learning_rate": 8.047348931415889e-06, + "loss": 0.4298, + "step": 50455 + }, + { + "epoch": 3.5012143501491915, + "grad_norm": 0.809767484664917, + "learning_rate": 8.045140803448592e-06, + "loss": 0.2164, + "step": 50456 + }, + { + "epoch": 3.5012837415862883, + "grad_norm": 0.7948267459869385, + "learning_rate": 8.04293296577211e-06, + "loss": 0.2077, + "step": 50457 + }, + { + "epoch": 3.501353133023385, + "grad_norm": 0.76202791929245, + "learning_rate": 8.040725418393413e-06, + "loss": 0.1538, + "step": 50458 + }, + { + "epoch": 3.5014225244604815, + "grad_norm": 1.14947509765625, + "learning_rate": 8.038518161319475e-06, + "loss": 0.3636, + "step": 50459 + }, + { + "epoch": 3.5014919158975784, + "grad_norm": 1.4748897552490234, + "learning_rate": 8.036311194557267e-06, + "loss": 0.3268, + "step": 50460 + }, + { + "epoch": 3.5015613073346747, + "grad_norm": 0.9570795893669128, + "learning_rate": 8.034104518113761e-06, + "loss": 0.2706, + "step": 50461 + }, + { + "epoch": 3.5016306987717716, + "grad_norm": 0.9986246228218079, + "learning_rate": 8.031898131995929e-06, + "loss": 0.3463, + "step": 50462 + }, + { + "epoch": 3.5017000902088684, + "grad_norm": 0.9227957129478455, + "learning_rate": 8.029692036210711e-06, + "loss": 0.2222, + "step": 50463 + }, + { + "epoch": 3.501769481645965, + "grad_norm": 1.4031546115875244, + "learning_rate": 8.02748623076508e-06, + "loss": 0.3843, + "step": 50464 + }, + { + "epoch": 3.5018388730830616, + "grad_norm": 1.3601802587509155, + "learning_rate": 8.025280715666006e-06, + "loss": 0.4981, + "step": 50465 + }, + { + "epoch": 3.501908264520158, + "grad_norm": 1.0317918062210083, + "learning_rate": 8.023075490920451e-06, + "loss": 0.3458, + "step": 50466 + }, + { + "epoch": 3.501977655957255, + "grad_norm": 0.6892472505569458, + "learning_rate": 8.020870556535376e-06, + "loss": 0.1992, + "step": 50467 + }, + { + "epoch": 3.5020470473943517, + "grad_norm": 1.1878089904785156, + "learning_rate": 8.018665912517743e-06, + "loss": 0.2531, + "step": 50468 + }, + { + "epoch": 3.502116438831448, + "grad_norm": 0.6049138903617859, + "learning_rate": 8.016461558874522e-06, + "loss": 0.2145, + "step": 50469 + }, + { + "epoch": 3.502185830268545, + "grad_norm": 0.8742662072181702, + "learning_rate": 8.014257495612643e-06, + "loss": 0.2278, + "step": 50470 + }, + { + "epoch": 3.5022552217056413, + "grad_norm": 0.9746453762054443, + "learning_rate": 8.012053722739077e-06, + "loss": 0.236, + "step": 50471 + }, + { + "epoch": 3.502324613142738, + "grad_norm": 0.9365457892417908, + "learning_rate": 8.009850240260786e-06, + "loss": 0.3768, + "step": 50472 + }, + { + "epoch": 3.502394004579835, + "grad_norm": 0.6269471645355225, + "learning_rate": 8.00764704818473e-06, + "loss": 0.2158, + "step": 50473 + }, + { + "epoch": 3.5024633960169314, + "grad_norm": 0.6705036163330078, + "learning_rate": 8.00544414651786e-06, + "loss": 0.1742, + "step": 50474 + }, + { + "epoch": 3.502532787454028, + "grad_norm": 0.8814608454704285, + "learning_rate": 8.003241535267125e-06, + "loss": 0.2131, + "step": 50475 + }, + { + "epoch": 3.5026021788911246, + "grad_norm": 0.9355766177177429, + "learning_rate": 8.00103921443951e-06, + "loss": 0.2524, + "step": 50476 + }, + { + "epoch": 3.5026715703282214, + "grad_norm": 1.8181357383728027, + "learning_rate": 7.998837184041908e-06, + "loss": 0.2678, + "step": 50477 + }, + { + "epoch": 3.5027409617653182, + "grad_norm": 0.7559579610824585, + "learning_rate": 7.996635444081314e-06, + "loss": 0.1698, + "step": 50478 + }, + { + "epoch": 3.502810353202415, + "grad_norm": 0.9628387093544006, + "learning_rate": 7.994433994564665e-06, + "loss": 0.2295, + "step": 50479 + }, + { + "epoch": 3.5028797446395115, + "grad_norm": 0.9253610968589783, + "learning_rate": 7.992232835498914e-06, + "loss": 0.2208, + "step": 50480 + }, + { + "epoch": 3.5029491360766083, + "grad_norm": 1.2692828178405762, + "learning_rate": 7.990031966890999e-06, + "loss": 0.2781, + "step": 50481 + }, + { + "epoch": 3.5030185275137047, + "grad_norm": 0.977226972579956, + "learning_rate": 7.98783138874789e-06, + "loss": 0.2869, + "step": 50482 + }, + { + "epoch": 3.5030879189508015, + "grad_norm": 1.0806174278259277, + "learning_rate": 7.98563110107654e-06, + "loss": 0.1792, + "step": 50483 + }, + { + "epoch": 3.5031573103878983, + "grad_norm": 1.3839765787124634, + "learning_rate": 7.983431103883855e-06, + "loss": 0.2383, + "step": 50484 + }, + { + "epoch": 3.5032267018249947, + "grad_norm": 0.9408178925514221, + "learning_rate": 7.981231397176802e-06, + "loss": 0.2614, + "step": 50485 + }, + { + "epoch": 3.5032960932620916, + "grad_norm": 1.214047908782959, + "learning_rate": 7.979031980962327e-06, + "loss": 0.3648, + "step": 50486 + }, + { + "epoch": 3.503365484699188, + "grad_norm": 1.320966124534607, + "learning_rate": 7.976832855247385e-06, + "loss": 0.3289, + "step": 50487 + }, + { + "epoch": 3.503434876136285, + "grad_norm": 0.718731164932251, + "learning_rate": 7.974634020038885e-06, + "loss": 0.1415, + "step": 50488 + }, + { + "epoch": 3.5035042675733816, + "grad_norm": 1.1767470836639404, + "learning_rate": 7.972435475343798e-06, + "loss": 0.3821, + "step": 50489 + }, + { + "epoch": 3.503573659010478, + "grad_norm": 0.7991542220115662, + "learning_rate": 7.970237221169052e-06, + "loss": 0.2282, + "step": 50490 + }, + { + "epoch": 3.503643050447575, + "grad_norm": 0.9588238000869751, + "learning_rate": 7.968039257521598e-06, + "loss": 0.2165, + "step": 50491 + }, + { + "epoch": 3.5037124418846712, + "grad_norm": 0.7889772653579712, + "learning_rate": 7.965841584408362e-06, + "loss": 0.2949, + "step": 50492 + }, + { + "epoch": 3.503781833321768, + "grad_norm": 0.5148622393608093, + "learning_rate": 7.963644201836284e-06, + "loss": 0.244, + "step": 50493 + }, + { + "epoch": 3.503851224758865, + "grad_norm": 0.659593939781189, + "learning_rate": 7.961447109812292e-06, + "loss": 0.1285, + "step": 50494 + }, + { + "epoch": 3.5039206161959613, + "grad_norm": 0.9939000010490417, + "learning_rate": 7.959250308343347e-06, + "loss": 0.2053, + "step": 50495 + }, + { + "epoch": 3.503990007633058, + "grad_norm": 0.9192301630973816, + "learning_rate": 7.957053797436375e-06, + "loss": 0.2479, + "step": 50496 + }, + { + "epoch": 3.5040593990701545, + "grad_norm": 0.7963985204696655, + "learning_rate": 7.954857577098295e-06, + "loss": 0.211, + "step": 50497 + }, + { + "epoch": 3.5041287905072513, + "grad_norm": 1.3654391765594482, + "learning_rate": 7.952661647336057e-06, + "loss": 0.3586, + "step": 50498 + }, + { + "epoch": 3.504198181944348, + "grad_norm": 0.8281428217887878, + "learning_rate": 7.950466008156576e-06, + "loss": 0.1987, + "step": 50499 + }, + { + "epoch": 3.5042675733814446, + "grad_norm": 0.8161572813987732, + "learning_rate": 7.948270659566791e-06, + "loss": 0.2011, + "step": 50500 + }, + { + "epoch": 3.5043369648185414, + "grad_norm": 0.846771776676178, + "learning_rate": 7.946075601573645e-06, + "loss": 0.2348, + "step": 50501 + }, + { + "epoch": 3.504406356255638, + "grad_norm": 0.6911315321922302, + "learning_rate": 7.94388083418406e-06, + "loss": 0.2134, + "step": 50502 + }, + { + "epoch": 3.5044757476927346, + "grad_norm": 0.6877068281173706, + "learning_rate": 7.941686357404977e-06, + "loss": 0.1478, + "step": 50503 + }, + { + "epoch": 3.5045451391298315, + "grad_norm": 1.058098316192627, + "learning_rate": 7.93949217124329e-06, + "loss": 0.3443, + "step": 50504 + }, + { + "epoch": 3.5046145305669283, + "grad_norm": 0.8574216961860657, + "learning_rate": 7.93729827570594e-06, + "loss": 0.2125, + "step": 50505 + }, + { + "epoch": 3.5046839220040247, + "grad_norm": 0.6506261825561523, + "learning_rate": 7.935104670799864e-06, + "loss": 0.1493, + "step": 50506 + }, + { + "epoch": 3.5047533134411215, + "grad_norm": 0.8644168972969055, + "learning_rate": 7.93291135653198e-06, + "loss": 0.2472, + "step": 50507 + }, + { + "epoch": 3.504822704878218, + "grad_norm": 1.0055800676345825, + "learning_rate": 7.930718332909215e-06, + "loss": 0.2155, + "step": 50508 + }, + { + "epoch": 3.5048920963153147, + "grad_norm": 0.6372470855712891, + "learning_rate": 7.928525599938485e-06, + "loss": 0.2869, + "step": 50509 + }, + { + "epoch": 3.5049614877524116, + "grad_norm": 1.0986590385437012, + "learning_rate": 7.926333157626742e-06, + "loss": 0.4123, + "step": 50510 + }, + { + "epoch": 3.505030879189508, + "grad_norm": 0.7228705286979675, + "learning_rate": 7.924141005980867e-06, + "loss": 0.1865, + "step": 50511 + }, + { + "epoch": 3.505100270626605, + "grad_norm": 0.7214722037315369, + "learning_rate": 7.92194914500779e-06, + "loss": 0.2115, + "step": 50512 + }, + { + "epoch": 3.505169662063701, + "grad_norm": 0.8133465051651001, + "learning_rate": 7.919757574714437e-06, + "loss": 0.2469, + "step": 50513 + }, + { + "epoch": 3.505239053500798, + "grad_norm": 0.745223879814148, + "learning_rate": 7.917566295107726e-06, + "loss": 0.2385, + "step": 50514 + }, + { + "epoch": 3.505308444937895, + "grad_norm": 0.7661387324333191, + "learning_rate": 7.915375306194572e-06, + "loss": 0.2686, + "step": 50515 + }, + { + "epoch": 3.5053778363749912, + "grad_norm": 1.230610966682434, + "learning_rate": 7.913184607981905e-06, + "loss": 0.3672, + "step": 50516 + }, + { + "epoch": 3.505447227812088, + "grad_norm": 0.7460528612136841, + "learning_rate": 7.91099420047664e-06, + "loss": 0.2415, + "step": 50517 + }, + { + "epoch": 3.5055166192491845, + "grad_norm": 0.7047571539878845, + "learning_rate": 7.908804083685661e-06, + "loss": 0.1925, + "step": 50518 + }, + { + "epoch": 3.5055860106862813, + "grad_norm": 1.1379101276397705, + "learning_rate": 7.906614257615907e-06, + "loss": 0.215, + "step": 50519 + }, + { + "epoch": 3.505655402123378, + "grad_norm": 1.506395936012268, + "learning_rate": 7.904424722274283e-06, + "loss": 0.2453, + "step": 50520 + }, + { + "epoch": 3.5057247935604745, + "grad_norm": 0.7803564071655273, + "learning_rate": 7.902235477667708e-06, + "loss": 0.2087, + "step": 50521 + }, + { + "epoch": 3.5057941849975713, + "grad_norm": 1.0006494522094727, + "learning_rate": 7.900046523803106e-06, + "loss": 0.2817, + "step": 50522 + }, + { + "epoch": 3.5058635764346677, + "grad_norm": 0.8895450830459595, + "learning_rate": 7.897857860687362e-06, + "loss": 0.1581, + "step": 50523 + }, + { + "epoch": 3.5059329678717646, + "grad_norm": 0.4792340099811554, + "learning_rate": 7.895669488327396e-06, + "loss": 0.1511, + "step": 50524 + }, + { + "epoch": 3.5060023593088614, + "grad_norm": 0.8982037901878357, + "learning_rate": 7.893481406730118e-06, + "loss": 0.3023, + "step": 50525 + }, + { + "epoch": 3.506071750745958, + "grad_norm": 0.8198739886283875, + "learning_rate": 7.89129361590243e-06, + "loss": 0.1182, + "step": 50526 + }, + { + "epoch": 3.5061411421830546, + "grad_norm": 0.9976308345794678, + "learning_rate": 7.88910611585123e-06, + "loss": 0.2255, + "step": 50527 + }, + { + "epoch": 3.506210533620151, + "grad_norm": 0.804442286491394, + "learning_rate": 7.886918906583441e-06, + "loss": 0.1983, + "step": 50528 + }, + { + "epoch": 3.506279925057248, + "grad_norm": 0.6254669427871704, + "learning_rate": 7.884731988105965e-06, + "loss": 0.1558, + "step": 50529 + }, + { + "epoch": 3.5063493164943447, + "grad_norm": 0.9098355174064636, + "learning_rate": 7.882545360425696e-06, + "loss": 0.337, + "step": 50530 + }, + { + "epoch": 3.5064187079314415, + "grad_norm": 0.926915168762207, + "learning_rate": 7.880359023549567e-06, + "loss": 0.2913, + "step": 50531 + }, + { + "epoch": 3.506488099368538, + "grad_norm": 0.9190410375595093, + "learning_rate": 7.878172977484432e-06, + "loss": 0.2885, + "step": 50532 + }, + { + "epoch": 3.5065574908056347, + "grad_norm": 0.8908940553665161, + "learning_rate": 7.875987222237225e-06, + "loss": 0.2067, + "step": 50533 + }, + { + "epoch": 3.506626882242731, + "grad_norm": 1.7516300678253174, + "learning_rate": 7.873801757814837e-06, + "loss": 0.23, + "step": 50534 + }, + { + "epoch": 3.506696273679828, + "grad_norm": 0.8271651864051819, + "learning_rate": 7.871616584224162e-06, + "loss": 0.1957, + "step": 50535 + }, + { + "epoch": 3.506765665116925, + "grad_norm": 0.9064487814903259, + "learning_rate": 7.869431701472108e-06, + "loss": 0.315, + "step": 50536 + }, + { + "epoch": 3.506835056554021, + "grad_norm": 1.4791686534881592, + "learning_rate": 7.867247109565567e-06, + "loss": 0.2994, + "step": 50537 + }, + { + "epoch": 3.506904447991118, + "grad_norm": 0.8552004098892212, + "learning_rate": 7.865062808511459e-06, + "loss": 0.2595, + "step": 50538 + }, + { + "epoch": 3.5069738394282144, + "grad_norm": 1.1213631629943848, + "learning_rate": 7.862878798316642e-06, + "loss": 0.2076, + "step": 50539 + }, + { + "epoch": 3.5070432308653112, + "grad_norm": 0.7113051414489746, + "learning_rate": 7.860695078988023e-06, + "loss": 0.2059, + "step": 50540 + }, + { + "epoch": 3.507112622302408, + "grad_norm": 0.6709768772125244, + "learning_rate": 7.858511650532496e-06, + "loss": 0.2389, + "step": 50541 + }, + { + "epoch": 3.5071820137395044, + "grad_norm": 1.0574603080749512, + "learning_rate": 7.856328512956956e-06, + "loss": 0.3408, + "step": 50542 + }, + { + "epoch": 3.5072514051766013, + "grad_norm": 0.8192856907844543, + "learning_rate": 7.854145666268308e-06, + "loss": 0.2355, + "step": 50543 + }, + { + "epoch": 3.5073207966136977, + "grad_norm": 1.2172496318817139, + "learning_rate": 7.851963110473426e-06, + "loss": 0.2548, + "step": 50544 + }, + { + "epoch": 3.5073901880507945, + "grad_norm": 1.1680805683135986, + "learning_rate": 7.849780845579213e-06, + "loss": 0.3428, + "step": 50545 + }, + { + "epoch": 3.5074595794878913, + "grad_norm": 1.1664774417877197, + "learning_rate": 7.847598871592543e-06, + "loss": 0.2499, + "step": 50546 + }, + { + "epoch": 3.5075289709249877, + "grad_norm": 1.0231808423995972, + "learning_rate": 7.845417188520309e-06, + "loss": 0.2136, + "step": 50547 + }, + { + "epoch": 3.5075983623620846, + "grad_norm": 1.0921822786331177, + "learning_rate": 7.843235796369408e-06, + "loss": 0.3567, + "step": 50548 + }, + { + "epoch": 3.507667753799181, + "grad_norm": 0.9132093787193298, + "learning_rate": 7.84105469514671e-06, + "loss": 0.3063, + "step": 50549 + }, + { + "epoch": 3.5077371452362778, + "grad_norm": 0.7600334286689758, + "learning_rate": 7.838873884859121e-06, + "loss": 0.2243, + "step": 50550 + }, + { + "epoch": 3.5078065366733746, + "grad_norm": 0.7286460995674133, + "learning_rate": 7.836693365513504e-06, + "loss": 0.2541, + "step": 50551 + }, + { + "epoch": 3.5078759281104714, + "grad_norm": 1.076263427734375, + "learning_rate": 7.834513137116772e-06, + "loss": 0.3223, + "step": 50552 + }, + { + "epoch": 3.507945319547568, + "grad_norm": 0.9402633905410767, + "learning_rate": 7.83233319967578e-06, + "loss": 0.2827, + "step": 50553 + }, + { + "epoch": 3.508014710984664, + "grad_norm": 0.9485777020454407, + "learning_rate": 7.830153553197406e-06, + "loss": 0.1964, + "step": 50554 + }, + { + "epoch": 3.508084102421761, + "grad_norm": 0.7606766223907471, + "learning_rate": 7.827974197688548e-06, + "loss": 0.1774, + "step": 50555 + }, + { + "epoch": 3.508153493858858, + "grad_norm": 0.7713515758514404, + "learning_rate": 7.825795133156078e-06, + "loss": 0.1639, + "step": 50556 + }, + { + "epoch": 3.5082228852959547, + "grad_norm": 1.1890928745269775, + "learning_rate": 7.8236163596069e-06, + "loss": 0.2063, + "step": 50557 + }, + { + "epoch": 3.508292276733051, + "grad_norm": 0.8126600384712219, + "learning_rate": 7.821437877047855e-06, + "loss": 0.2531, + "step": 50558 + }, + { + "epoch": 3.508361668170148, + "grad_norm": 1.015611171722412, + "learning_rate": 7.819259685485835e-06, + "loss": 0.2967, + "step": 50559 + }, + { + "epoch": 3.5084310596072443, + "grad_norm": 0.9382142424583435, + "learning_rate": 7.817081784927737e-06, + "loss": 0.2606, + "step": 50560 + }, + { + "epoch": 3.508500451044341, + "grad_norm": 0.9422228932380676, + "learning_rate": 7.814904175380399e-06, + "loss": 0.1884, + "step": 50561 + }, + { + "epoch": 3.508569842481438, + "grad_norm": 0.8030144572257996, + "learning_rate": 7.812726856850705e-06, + "loss": 0.2437, + "step": 50562 + }, + { + "epoch": 3.5086392339185344, + "grad_norm": 0.6423381567001343, + "learning_rate": 7.810549829345547e-06, + "loss": 0.183, + "step": 50563 + }, + { + "epoch": 3.508708625355631, + "grad_norm": 1.022349238395691, + "learning_rate": 7.80837309287178e-06, + "loss": 0.2624, + "step": 50564 + }, + { + "epoch": 3.5087780167927276, + "grad_norm": 0.9483939409255981, + "learning_rate": 7.806196647436303e-06, + "loss": 0.3783, + "step": 50565 + }, + { + "epoch": 3.5088474082298244, + "grad_norm": 0.541459321975708, + "learning_rate": 7.804020493045949e-06, + "loss": 0.1811, + "step": 50566 + }, + { + "epoch": 3.5089167996669213, + "grad_norm": 1.1391812562942505, + "learning_rate": 7.801844629707611e-06, + "loss": 0.3446, + "step": 50567 + }, + { + "epoch": 3.5089861911040177, + "grad_norm": 0.8651574850082397, + "learning_rate": 7.799669057428149e-06, + "loss": 0.1752, + "step": 50568 + }, + { + "epoch": 3.5090555825411145, + "grad_norm": 1.020487904548645, + "learning_rate": 7.797493776214427e-06, + "loss": 0.2474, + "step": 50569 + }, + { + "epoch": 3.509124973978211, + "grad_norm": 1.0206732749938965, + "learning_rate": 7.795318786073325e-06, + "loss": 0.3378, + "step": 50570 + }, + { + "epoch": 3.5091943654153077, + "grad_norm": 0.9868839383125305, + "learning_rate": 7.793144087011705e-06, + "loss": 0.2792, + "step": 50571 + }, + { + "epoch": 3.5092637568524045, + "grad_norm": 1.3659073114395142, + "learning_rate": 7.790969679036453e-06, + "loss": 0.2352, + "step": 50572 + }, + { + "epoch": 3.509333148289501, + "grad_norm": 0.9002314805984497, + "learning_rate": 7.788795562154383e-06, + "loss": 0.2897, + "step": 50573 + }, + { + "epoch": 3.5094025397265978, + "grad_norm": 0.9629836678504944, + "learning_rate": 7.7866217363724e-06, + "loss": 0.2236, + "step": 50574 + }, + { + "epoch": 3.509471931163694, + "grad_norm": 0.5846979022026062, + "learning_rate": 7.784448201697348e-06, + "loss": 0.1185, + "step": 50575 + }, + { + "epoch": 3.509541322600791, + "grad_norm": 0.8219379782676697, + "learning_rate": 7.782274958136093e-06, + "loss": 0.2333, + "step": 50576 + }, + { + "epoch": 3.509610714037888, + "grad_norm": 0.8358895182609558, + "learning_rate": 7.78010200569549e-06, + "loss": 0.2707, + "step": 50577 + }, + { + "epoch": 3.5096801054749847, + "grad_norm": 0.7139101624488831, + "learning_rate": 7.777929344382417e-06, + "loss": 0.2177, + "step": 50578 + }, + { + "epoch": 3.509749496912081, + "grad_norm": 1.0042710304260254, + "learning_rate": 7.77575697420373e-06, + "loss": 0.3228, + "step": 50579 + }, + { + "epoch": 3.5098188883491774, + "grad_norm": 0.8366632461547852, + "learning_rate": 7.773584895166264e-06, + "loss": 0.2208, + "step": 50580 + }, + { + "epoch": 3.5098882797862743, + "grad_norm": 0.9762701392173767, + "learning_rate": 7.771413107276893e-06, + "loss": 0.2556, + "step": 50581 + }, + { + "epoch": 3.509957671223371, + "grad_norm": 0.8480355739593506, + "learning_rate": 7.769241610542465e-06, + "loss": 0.2611, + "step": 50582 + }, + { + "epoch": 3.510027062660468, + "grad_norm": 1.2370187044143677, + "learning_rate": 7.767070404969845e-06, + "loss": 0.2735, + "step": 50583 + }, + { + "epoch": 3.5100964540975643, + "grad_norm": 0.8214558362960815, + "learning_rate": 7.76489949056588e-06, + "loss": 0.2606, + "step": 50584 + }, + { + "epoch": 3.510165845534661, + "grad_norm": 1.4879063367843628, + "learning_rate": 7.762728867337433e-06, + "loss": 0.5461, + "step": 50585 + }, + { + "epoch": 3.5102352369717575, + "grad_norm": 0.7333166599273682, + "learning_rate": 7.760558535291363e-06, + "loss": 0.2406, + "step": 50586 + }, + { + "epoch": 3.5103046284088544, + "grad_norm": 0.7442348599433899, + "learning_rate": 7.758388494434487e-06, + "loss": 0.1889, + "step": 50587 + }, + { + "epoch": 3.510374019845951, + "grad_norm": 0.9352176785469055, + "learning_rate": 7.75621874477367e-06, + "loss": 0.2801, + "step": 50588 + }, + { + "epoch": 3.5104434112830476, + "grad_norm": 1.0649595260620117, + "learning_rate": 7.754049286315779e-06, + "loss": 0.1274, + "step": 50589 + }, + { + "epoch": 3.5105128027201444, + "grad_norm": 0.8669300079345703, + "learning_rate": 7.751880119067657e-06, + "loss": 0.294, + "step": 50590 + }, + { + "epoch": 3.510582194157241, + "grad_norm": 0.6500921845436096, + "learning_rate": 7.74971124303613e-06, + "loss": 0.2522, + "step": 50591 + }, + { + "epoch": 3.5106515855943377, + "grad_norm": 0.7132775783538818, + "learning_rate": 7.74754265822809e-06, + "loss": 0.1985, + "step": 50592 + }, + { + "epoch": 3.5107209770314345, + "grad_norm": 1.0240962505340576, + "learning_rate": 7.745374364650338e-06, + "loss": 0.2865, + "step": 50593 + }, + { + "epoch": 3.510790368468531, + "grad_norm": 0.825442910194397, + "learning_rate": 7.74320636230973e-06, + "loss": 0.186, + "step": 50594 + }, + { + "epoch": 3.5108597599056277, + "grad_norm": 0.9733646512031555, + "learning_rate": 7.74103865121314e-06, + "loss": 0.2261, + "step": 50595 + }, + { + "epoch": 3.510929151342724, + "grad_norm": 1.1617228984832764, + "learning_rate": 7.738871231367362e-06, + "loss": 0.3416, + "step": 50596 + }, + { + "epoch": 3.510998542779821, + "grad_norm": 1.1771430969238281, + "learning_rate": 7.736704102779268e-06, + "loss": 0.2549, + "step": 50597 + }, + { + "epoch": 3.5110679342169178, + "grad_norm": 0.6873403787612915, + "learning_rate": 7.734537265455689e-06, + "loss": 0.1875, + "step": 50598 + }, + { + "epoch": 3.511137325654014, + "grad_norm": 0.9362647533416748, + "learning_rate": 7.732370719403482e-06, + "loss": 0.2919, + "step": 50599 + }, + { + "epoch": 3.511206717091111, + "grad_norm": 1.315497875213623, + "learning_rate": 7.730204464629486e-06, + "loss": 0.3362, + "step": 50600 + }, + { + "epoch": 3.5112761085282074, + "grad_norm": 1.6392947435379028, + "learning_rate": 7.728038501140511e-06, + "loss": 0.352, + "step": 50601 + }, + { + "epoch": 3.511345499965304, + "grad_norm": 0.8623930811882019, + "learning_rate": 7.725872828943415e-06, + "loss": 0.2889, + "step": 50602 + }, + { + "epoch": 3.511414891402401, + "grad_norm": 0.7211084961891174, + "learning_rate": 7.723707448045026e-06, + "loss": 0.2213, + "step": 50603 + }, + { + "epoch": 3.511484282839498, + "grad_norm": 1.3013774156570435, + "learning_rate": 7.721542358452194e-06, + "loss": 0.1888, + "step": 50604 + }, + { + "epoch": 3.5115536742765943, + "grad_norm": 1.1625367403030396, + "learning_rate": 7.719377560171748e-06, + "loss": 0.3002, + "step": 50605 + }, + { + "epoch": 3.511623065713691, + "grad_norm": 0.773230254650116, + "learning_rate": 7.717213053210515e-06, + "loss": 0.2523, + "step": 50606 + }, + { + "epoch": 3.5116924571507875, + "grad_norm": 0.8683587908744812, + "learning_rate": 7.715048837575356e-06, + "loss": 0.1956, + "step": 50607 + }, + { + "epoch": 3.5117618485878843, + "grad_norm": 1.223005771636963, + "learning_rate": 7.712884913273055e-06, + "loss": 0.2685, + "step": 50608 + }, + { + "epoch": 3.511831240024981, + "grad_norm": 0.5511196851730347, + "learning_rate": 7.710721280310473e-06, + "loss": 0.1799, + "step": 50609 + }, + { + "epoch": 3.5119006314620775, + "grad_norm": 0.8657209873199463, + "learning_rate": 7.708557938694438e-06, + "loss": 0.3668, + "step": 50610 + }, + { + "epoch": 3.5119700228991744, + "grad_norm": 1.3017653226852417, + "learning_rate": 7.706394888431779e-06, + "loss": 0.3281, + "step": 50611 + }, + { + "epoch": 3.5120394143362708, + "grad_norm": 0.7808641791343689, + "learning_rate": 7.704232129529331e-06, + "loss": 0.2294, + "step": 50612 + }, + { + "epoch": 3.5121088057733676, + "grad_norm": 0.7002483606338501, + "learning_rate": 7.702069661993905e-06, + "loss": 0.1993, + "step": 50613 + }, + { + "epoch": 3.5121781972104644, + "grad_norm": 0.7983700037002563, + "learning_rate": 7.699907485832347e-06, + "loss": 0.1309, + "step": 50614 + }, + { + "epoch": 3.512247588647561, + "grad_norm": 0.7326998114585876, + "learning_rate": 7.697745601051465e-06, + "loss": 0.171, + "step": 50615 + }, + { + "epoch": 3.5123169800846576, + "grad_norm": 0.7545817494392395, + "learning_rate": 7.695584007658085e-06, + "loss": 0.2302, + "step": 50616 + }, + { + "epoch": 3.512386371521754, + "grad_norm": 0.9983240365982056, + "learning_rate": 7.693422705659049e-06, + "loss": 0.2275, + "step": 50617 + }, + { + "epoch": 3.512455762958851, + "grad_norm": 0.8014011383056641, + "learning_rate": 7.691261695061157e-06, + "loss": 0.2302, + "step": 50618 + }, + { + "epoch": 3.5125251543959477, + "grad_norm": 0.8050525784492493, + "learning_rate": 7.689100975871243e-06, + "loss": 0.2031, + "step": 50619 + }, + { + "epoch": 3.512594545833044, + "grad_norm": 1.1650047302246094, + "learning_rate": 7.686940548096121e-06, + "loss": 0.3342, + "step": 50620 + }, + { + "epoch": 3.512663937270141, + "grad_norm": 1.4535682201385498, + "learning_rate": 7.684780411742643e-06, + "loss": 0.3325, + "step": 50621 + }, + { + "epoch": 3.5127333287072373, + "grad_norm": 0.8895981311798096, + "learning_rate": 7.682620566817589e-06, + "loss": 0.2406, + "step": 50622 + }, + { + "epoch": 3.512802720144334, + "grad_norm": 0.6790342330932617, + "learning_rate": 7.68046101332779e-06, + "loss": 0.2564, + "step": 50623 + }, + { + "epoch": 3.512872111581431, + "grad_norm": 1.0169868469238281, + "learning_rate": 7.678301751280059e-06, + "loss": 0.2379, + "step": 50624 + }, + { + "epoch": 3.5129415030185274, + "grad_norm": 0.7542202472686768, + "learning_rate": 7.676142780681217e-06, + "loss": 0.1241, + "step": 50625 + }, + { + "epoch": 3.513010894455624, + "grad_norm": 0.8864225745201111, + "learning_rate": 7.673984101538101e-06, + "loss": 0.2931, + "step": 50626 + }, + { + "epoch": 3.5130802858927206, + "grad_norm": 0.8665932416915894, + "learning_rate": 7.671825713857484e-06, + "loss": 0.2322, + "step": 50627 + }, + { + "epoch": 3.5131496773298174, + "grad_norm": 0.8802230954170227, + "learning_rate": 7.669667617646203e-06, + "loss": 0.3444, + "step": 50628 + }, + { + "epoch": 3.5132190687669143, + "grad_norm": 0.8588577508926392, + "learning_rate": 7.667509812911077e-06, + "loss": 0.2573, + "step": 50629 + }, + { + "epoch": 3.513288460204011, + "grad_norm": 0.7488147020339966, + "learning_rate": 7.665352299658923e-06, + "loss": 0.2351, + "step": 50630 + }, + { + "epoch": 3.5133578516411075, + "grad_norm": 0.7736696004867554, + "learning_rate": 7.663195077896524e-06, + "loss": 0.2748, + "step": 50631 + }, + { + "epoch": 3.5134272430782043, + "grad_norm": 1.5231502056121826, + "learning_rate": 7.661038147630694e-06, + "loss": 0.2683, + "step": 50632 + }, + { + "epoch": 3.5134966345153007, + "grad_norm": 0.8999350070953369, + "learning_rate": 7.658881508868265e-06, + "loss": 0.1251, + "step": 50633 + }, + { + "epoch": 3.5135660259523975, + "grad_norm": 1.1669621467590332, + "learning_rate": 7.65672516161603e-06, + "loss": 0.3223, + "step": 50634 + }, + { + "epoch": 3.5136354173894944, + "grad_norm": 1.6526914834976196, + "learning_rate": 7.654569105880816e-06, + "loss": 0.2849, + "step": 50635 + }, + { + "epoch": 3.5137048088265908, + "grad_norm": 0.7294878363609314, + "learning_rate": 7.652413341669395e-06, + "loss": 0.2312, + "step": 50636 + }, + { + "epoch": 3.5137742002636876, + "grad_norm": 1.0727875232696533, + "learning_rate": 7.650257868988586e-06, + "loss": 0.2434, + "step": 50637 + }, + { + "epoch": 3.513843591700784, + "grad_norm": 1.0672591924667358, + "learning_rate": 7.648102687845205e-06, + "loss": 0.3247, + "step": 50638 + }, + { + "epoch": 3.513912983137881, + "grad_norm": 1.1986956596374512, + "learning_rate": 7.645947798246056e-06, + "loss": 0.3113, + "step": 50639 + }, + { + "epoch": 3.5139823745749776, + "grad_norm": 0.7936862111091614, + "learning_rate": 7.643793200197925e-06, + "loss": 0.1629, + "step": 50640 + }, + { + "epoch": 3.514051766012074, + "grad_norm": 0.7874159216880798, + "learning_rate": 7.641638893707636e-06, + "loss": 0.1945, + "step": 50641 + }, + { + "epoch": 3.514121157449171, + "grad_norm": 0.6047307848930359, + "learning_rate": 7.639484878781976e-06, + "loss": 0.1341, + "step": 50642 + }, + { + "epoch": 3.5141905488862673, + "grad_norm": 1.000362753868103, + "learning_rate": 7.637331155427729e-06, + "loss": 0.1996, + "step": 50643 + }, + { + "epoch": 3.514259940323364, + "grad_norm": 0.6538587212562561, + "learning_rate": 7.63517772365172e-06, + "loss": 0.2182, + "step": 50644 + }, + { + "epoch": 3.514329331760461, + "grad_norm": 0.8190261721611023, + "learning_rate": 7.633024583460746e-06, + "loss": 0.3748, + "step": 50645 + }, + { + "epoch": 3.5143987231975573, + "grad_norm": 0.9150679707527161, + "learning_rate": 7.63087173486159e-06, + "loss": 0.2697, + "step": 50646 + }, + { + "epoch": 3.514468114634654, + "grad_norm": 0.8979218006134033, + "learning_rate": 7.628719177861055e-06, + "loss": 0.1705, + "step": 50647 + }, + { + "epoch": 3.5145375060717505, + "grad_norm": 1.2579562664031982, + "learning_rate": 7.626566912465949e-06, + "loss": 0.2389, + "step": 50648 + }, + { + "epoch": 3.5146068975088474, + "grad_norm": 0.7096962332725525, + "learning_rate": 7.624414938683044e-06, + "loss": 0.2122, + "step": 50649 + }, + { + "epoch": 3.514676288945944, + "grad_norm": 0.9007768034934998, + "learning_rate": 7.622263256519135e-06, + "loss": 0.1943, + "step": 50650 + }, + { + "epoch": 3.5147456803830406, + "grad_norm": 0.778082013130188, + "learning_rate": 7.620111865981039e-06, + "loss": 0.2309, + "step": 50651 + }, + { + "epoch": 3.5148150718201374, + "grad_norm": 0.5862938165664673, + "learning_rate": 7.617960767075516e-06, + "loss": 0.2142, + "step": 50652 + }, + { + "epoch": 3.514884463257234, + "grad_norm": 0.821875274181366, + "learning_rate": 7.615809959809383e-06, + "loss": 0.2186, + "step": 50653 + }, + { + "epoch": 3.5149538546943306, + "grad_norm": 0.7771594524383545, + "learning_rate": 7.613659444189414e-06, + "loss": 0.2393, + "step": 50654 + }, + { + "epoch": 3.5150232461314275, + "grad_norm": 0.8589150905609131, + "learning_rate": 7.611509220222413e-06, + "loss": 0.2373, + "step": 50655 + }, + { + "epoch": 3.5150926375685243, + "grad_norm": 0.6673151850700378, + "learning_rate": 7.609359287915152e-06, + "loss": 0.2202, + "step": 50656 + }, + { + "epoch": 3.5151620290056207, + "grad_norm": 0.6081060171127319, + "learning_rate": 7.607209647274427e-06, + "loss": 0.1785, + "step": 50657 + }, + { + "epoch": 3.5152314204427175, + "grad_norm": 0.7350118160247803, + "learning_rate": 7.60506029830701e-06, + "loss": 0.2443, + "step": 50658 + }, + { + "epoch": 3.515300811879814, + "grad_norm": 0.9316354990005493, + "learning_rate": 7.6029112410197055e-06, + "loss": 0.3123, + "step": 50659 + }, + { + "epoch": 3.5153702033169107, + "grad_norm": 0.8519639372825623, + "learning_rate": 7.600762475419288e-06, + "loss": 0.3045, + "step": 50660 + }, + { + "epoch": 3.5154395947540076, + "grad_norm": 1.4585151672363281, + "learning_rate": 7.598614001512561e-06, + "loss": 0.3219, + "step": 50661 + }, + { + "epoch": 3.515508986191104, + "grad_norm": 0.7946017980575562, + "learning_rate": 7.596465819306275e-06, + "loss": 0.2427, + "step": 50662 + }, + { + "epoch": 3.515578377628201, + "grad_norm": 0.779751181602478, + "learning_rate": 7.594317928807215e-06, + "loss": 0.2159, + "step": 50663 + }, + { + "epoch": 3.515647769065297, + "grad_norm": 0.6813624501228333, + "learning_rate": 7.592170330022186e-06, + "loss": 0.1843, + "step": 50664 + }, + { + "epoch": 3.515717160502394, + "grad_norm": 0.7808755040168762, + "learning_rate": 7.59002302295796e-06, + "loss": 0.1762, + "step": 50665 + }, + { + "epoch": 3.515786551939491, + "grad_norm": 0.800557017326355, + "learning_rate": 7.5878760076212975e-06, + "loss": 0.2903, + "step": 50666 + }, + { + "epoch": 3.5158559433765872, + "grad_norm": 0.8826613426208496, + "learning_rate": 7.585729284018994e-06, + "loss": 0.3083, + "step": 50667 + }, + { + "epoch": 3.515925334813684, + "grad_norm": 0.8231254816055298, + "learning_rate": 7.583582852157811e-06, + "loss": 0.1393, + "step": 50668 + }, + { + "epoch": 3.5159947262507805, + "grad_norm": 0.6887618899345398, + "learning_rate": 7.581436712044565e-06, + "loss": 0.2171, + "step": 50669 + }, + { + "epoch": 3.5160641176878773, + "grad_norm": 0.8603971004486084, + "learning_rate": 7.579290863685973e-06, + "loss": 0.2073, + "step": 50670 + }, + { + "epoch": 3.516133509124974, + "grad_norm": 0.7761813402175903, + "learning_rate": 7.57714530708884e-06, + "loss": 0.2129, + "step": 50671 + }, + { + "epoch": 3.5162029005620705, + "grad_norm": 0.9308432340621948, + "learning_rate": 7.5750000422599276e-06, + "loss": 0.2919, + "step": 50672 + }, + { + "epoch": 3.5162722919991674, + "grad_norm": 0.7939059138298035, + "learning_rate": 7.572855069206031e-06, + "loss": 0.2853, + "step": 50673 + }, + { + "epoch": 3.5163416834362637, + "grad_norm": 1.303048849105835, + "learning_rate": 7.5707103879338994e-06, + "loss": 0.2543, + "step": 50674 + }, + { + "epoch": 3.5164110748733606, + "grad_norm": 0.7732577919960022, + "learning_rate": 7.568565998450317e-06, + "loss": 0.2247, + "step": 50675 + }, + { + "epoch": 3.5164804663104574, + "grad_norm": 0.9152975082397461, + "learning_rate": 7.5664219007620556e-06, + "loss": 0.367, + "step": 50676 + }, + { + "epoch": 3.516549857747554, + "grad_norm": 0.7408136129379272, + "learning_rate": 7.564278094875865e-06, + "loss": 0.1919, + "step": 50677 + }, + { + "epoch": 3.5166192491846506, + "grad_norm": 1.0288870334625244, + "learning_rate": 7.562134580798519e-06, + "loss": 0.2011, + "step": 50678 + }, + { + "epoch": 3.516688640621747, + "grad_norm": 1.245581030845642, + "learning_rate": 7.5599913585368e-06, + "loss": 0.3063, + "step": 50679 + }, + { + "epoch": 3.516758032058844, + "grad_norm": 1.1352463960647583, + "learning_rate": 7.557848428097447e-06, + "loss": 0.2488, + "step": 50680 + }, + { + "epoch": 3.5168274234959407, + "grad_norm": 1.1406584978103638, + "learning_rate": 7.555705789487255e-06, + "loss": 0.3095, + "step": 50681 + }, + { + "epoch": 3.5168968149330375, + "grad_norm": 1.0604724884033203, + "learning_rate": 7.553563442712974e-06, + "loss": 0.2271, + "step": 50682 + }, + { + "epoch": 3.516966206370134, + "grad_norm": 0.8845476508140564, + "learning_rate": 7.551421387781377e-06, + "loss": 0.2747, + "step": 50683 + }, + { + "epoch": 3.5170355978072307, + "grad_norm": 1.3188337087631226, + "learning_rate": 7.549279624699202e-06, + "loss": 0.3785, + "step": 50684 + }, + { + "epoch": 3.517104989244327, + "grad_norm": 0.9001861214637756, + "learning_rate": 7.547138153473232e-06, + "loss": 0.1994, + "step": 50685 + }, + { + "epoch": 3.517174380681424, + "grad_norm": 0.7761908769607544, + "learning_rate": 7.5449969741102075e-06, + "loss": 0.2573, + "step": 50686 + }, + { + "epoch": 3.517243772118521, + "grad_norm": 0.7393847703933716, + "learning_rate": 7.5428560866169114e-06, + "loss": 0.184, + "step": 50687 + }, + { + "epoch": 3.517313163555617, + "grad_norm": 0.8089772462844849, + "learning_rate": 7.540715491000083e-06, + "loss": 0.2112, + "step": 50688 + }, + { + "epoch": 3.517382554992714, + "grad_norm": 0.8416760563850403, + "learning_rate": 7.538575187266494e-06, + "loss": 0.2651, + "step": 50689 + }, + { + "epoch": 3.5174519464298104, + "grad_norm": 0.8548483848571777, + "learning_rate": 7.536435175422918e-06, + "loss": 0.2815, + "step": 50690 + }, + { + "epoch": 3.5175213378669072, + "grad_norm": 0.7551080584526062, + "learning_rate": 7.534295455476059e-06, + "loss": 0.1362, + "step": 50691 + }, + { + "epoch": 3.517590729304004, + "grad_norm": 0.8920493721961975, + "learning_rate": 7.532156027432713e-06, + "loss": 0.2096, + "step": 50692 + }, + { + "epoch": 3.5176601207411005, + "grad_norm": 1.1592220067977905, + "learning_rate": 7.530016891299618e-06, + "loss": 0.3307, + "step": 50693 + }, + { + "epoch": 3.5177295121781973, + "grad_norm": 0.7447257041931152, + "learning_rate": 7.527878047083525e-06, + "loss": 0.217, + "step": 50694 + }, + { + "epoch": 3.5177989036152937, + "grad_norm": 0.7008333206176758, + "learning_rate": 7.525739494791195e-06, + "loss": 0.265, + "step": 50695 + }, + { + "epoch": 3.5178682950523905, + "grad_norm": 0.7665037512779236, + "learning_rate": 7.5236012344294e-06, + "loss": 0.1189, + "step": 50696 + }, + { + "epoch": 3.5179376864894873, + "grad_norm": 0.6775622367858887, + "learning_rate": 7.521463266004847e-06, + "loss": 0.2218, + "step": 50697 + }, + { + "epoch": 3.5180070779265837, + "grad_norm": 1.1471635103225708, + "learning_rate": 7.519325589524306e-06, + "loss": 0.3514, + "step": 50698 + }, + { + "epoch": 3.5180764693636806, + "grad_norm": 0.8236123919487, + "learning_rate": 7.517188204994519e-06, + "loss": 0.2406, + "step": 50699 + }, + { + "epoch": 3.518145860800777, + "grad_norm": 0.8683608174324036, + "learning_rate": 7.515051112422255e-06, + "loss": 0.1615, + "step": 50700 + }, + { + "epoch": 3.518215252237874, + "grad_norm": 1.5358983278274536, + "learning_rate": 7.512914311814235e-06, + "loss": 0.3539, + "step": 50701 + }, + { + "epoch": 3.5182846436749706, + "grad_norm": 0.8733861446380615, + "learning_rate": 7.510777803177205e-06, + "loss": 0.2291, + "step": 50702 + }, + { + "epoch": 3.518354035112067, + "grad_norm": 0.6875607371330261, + "learning_rate": 7.508641586517928e-06, + "loss": 0.2515, + "step": 50703 + }, + { + "epoch": 3.518423426549164, + "grad_norm": 0.8003062009811401, + "learning_rate": 7.506505661843155e-06, + "loss": 0.25, + "step": 50704 + }, + { + "epoch": 3.5184928179862602, + "grad_norm": 0.6010258793830872, + "learning_rate": 7.504370029159591e-06, + "loss": 0.1142, + "step": 50705 + }, + { + "epoch": 3.518562209423357, + "grad_norm": 0.8524079918861389, + "learning_rate": 7.502234688473997e-06, + "loss": 0.1726, + "step": 50706 + }, + { + "epoch": 3.518631600860454, + "grad_norm": 0.8559367060661316, + "learning_rate": 7.500099639793123e-06, + "loss": 0.2797, + "step": 50707 + }, + { + "epoch": 3.5187009922975507, + "grad_norm": 0.8164495229721069, + "learning_rate": 7.497964883123698e-06, + "loss": 0.2107, + "step": 50708 + }, + { + "epoch": 3.518770383734647, + "grad_norm": 1.254226565361023, + "learning_rate": 7.49583041847246e-06, + "loss": 0.2012, + "step": 50709 + }, + { + "epoch": 3.518839775171744, + "grad_norm": 0.6802076697349548, + "learning_rate": 7.49369624584616e-06, + "loss": 0.1983, + "step": 50710 + }, + { + "epoch": 3.5189091666088403, + "grad_norm": 1.0396370887756348, + "learning_rate": 7.491562365251548e-06, + "loss": 0.2376, + "step": 50711 + }, + { + "epoch": 3.518978558045937, + "grad_norm": 0.8571039438247681, + "learning_rate": 7.4894287766953175e-06, + "loss": 0.2198, + "step": 50712 + }, + { + "epoch": 3.519047949483034, + "grad_norm": 0.8062736988067627, + "learning_rate": 7.487295480184231e-06, + "loss": 0.196, + "step": 50713 + }, + { + "epoch": 3.5191173409201304, + "grad_norm": 0.8051261901855469, + "learning_rate": 7.4851624757250164e-06, + "loss": 0.2742, + "step": 50714 + }, + { + "epoch": 3.5191867323572272, + "grad_norm": 0.9545313715934753, + "learning_rate": 7.483029763324412e-06, + "loss": 0.2566, + "step": 50715 + }, + { + "epoch": 3.5192561237943236, + "grad_norm": 1.2444647550582886, + "learning_rate": 7.480897342989146e-06, + "loss": 0.57, + "step": 50716 + }, + { + "epoch": 3.5193255152314205, + "grad_norm": 1.05133855342865, + "learning_rate": 7.47876521472598e-06, + "loss": 0.2688, + "step": 50717 + }, + { + "epoch": 3.5193949066685173, + "grad_norm": 0.826421320438385, + "learning_rate": 7.476633378541586e-06, + "loss": 0.2239, + "step": 50718 + }, + { + "epoch": 3.5194642981056137, + "grad_norm": 1.5281364917755127, + "learning_rate": 7.474501834442737e-06, + "loss": 0.3172, + "step": 50719 + }, + { + "epoch": 3.5195336895427105, + "grad_norm": 0.820713996887207, + "learning_rate": 7.47237058243615e-06, + "loss": 0.2139, + "step": 50720 + }, + { + "epoch": 3.519603080979807, + "grad_norm": 0.9403354525566101, + "learning_rate": 7.470239622528552e-06, + "loss": 0.2309, + "step": 50721 + }, + { + "epoch": 3.5196724724169037, + "grad_norm": 1.0431160926818848, + "learning_rate": 7.46810895472666e-06, + "loss": 0.238, + "step": 50722 + }, + { + "epoch": 3.5197418638540006, + "grad_norm": 0.9106937050819397, + "learning_rate": 7.465978579037225e-06, + "loss": 0.2468, + "step": 50723 + }, + { + "epoch": 3.519811255291097, + "grad_norm": 0.7263315916061401, + "learning_rate": 7.463848495466974e-06, + "loss": 0.2333, + "step": 50724 + }, + { + "epoch": 3.519880646728194, + "grad_norm": 0.8030470013618469, + "learning_rate": 7.461718704022602e-06, + "loss": 0.2259, + "step": 50725 + }, + { + "epoch": 3.51995003816529, + "grad_norm": 1.076192021369934, + "learning_rate": 7.459589204710849e-06, + "loss": 0.1784, + "step": 50726 + }, + { + "epoch": 3.520019429602387, + "grad_norm": 0.909333348274231, + "learning_rate": 7.45745999753843e-06, + "loss": 0.248, + "step": 50727 + }, + { + "epoch": 3.520088821039484, + "grad_norm": 0.8221942186355591, + "learning_rate": 7.455331082512085e-06, + "loss": 0.1795, + "step": 50728 + }, + { + "epoch": 3.5201582124765802, + "grad_norm": 1.2074306011199951, + "learning_rate": 7.4532024596385085e-06, + "loss": 0.3272, + "step": 50729 + }, + { + "epoch": 3.520227603913677, + "grad_norm": 1.0091079473495483, + "learning_rate": 7.45107412892444e-06, + "loss": 0.2678, + "step": 50730 + }, + { + "epoch": 3.5202969953507735, + "grad_norm": 1.69401216506958, + "learning_rate": 7.448946090376607e-06, + "loss": 0.3673, + "step": 50731 + }, + { + "epoch": 3.5203663867878703, + "grad_norm": 0.7660315632820129, + "learning_rate": 7.446818344001705e-06, + "loss": 0.2099, + "step": 50732 + }, + { + "epoch": 3.520435778224967, + "grad_norm": 0.65011066198349, + "learning_rate": 7.444690889806449e-06, + "loss": 0.2538, + "step": 50733 + }, + { + "epoch": 3.520505169662064, + "grad_norm": 0.7554769515991211, + "learning_rate": 7.442563727797569e-06, + "loss": 0.1967, + "step": 50734 + }, + { + "epoch": 3.5205745610991603, + "grad_norm": 1.4305754899978638, + "learning_rate": 7.440436857981803e-06, + "loss": 0.294, + "step": 50735 + }, + { + "epoch": 3.520643952536257, + "grad_norm": 1.1830686330795288, + "learning_rate": 7.438310280365812e-06, + "loss": 0.3095, + "step": 50736 + }, + { + "epoch": 3.5207133439733536, + "grad_norm": 0.7983309626579285, + "learning_rate": 7.436183994956336e-06, + "loss": 0.1005, + "step": 50737 + }, + { + "epoch": 3.5207827354104504, + "grad_norm": 0.7392286062240601, + "learning_rate": 7.434058001760102e-06, + "loss": 0.1874, + "step": 50738 + }, + { + "epoch": 3.5208521268475472, + "grad_norm": 0.8671544194221497, + "learning_rate": 7.431932300783795e-06, + "loss": 0.2786, + "step": 50739 + }, + { + "epoch": 3.5209215182846436, + "grad_norm": 0.8456865549087524, + "learning_rate": 7.429806892034141e-06, + "loss": 0.2322, + "step": 50740 + }, + { + "epoch": 3.5209909097217404, + "grad_norm": 0.8012625575065613, + "learning_rate": 7.427681775517848e-06, + "loss": 0.1919, + "step": 50741 + }, + { + "epoch": 3.521060301158837, + "grad_norm": 1.143373966217041, + "learning_rate": 7.425556951241608e-06, + "loss": 0.3707, + "step": 50742 + }, + { + "epoch": 3.5211296925959337, + "grad_norm": 0.5736784934997559, + "learning_rate": 7.423432419212162e-06, + "loss": 0.1285, + "step": 50743 + }, + { + "epoch": 3.5211990840330305, + "grad_norm": 0.7330282330513, + "learning_rate": 7.421308179436193e-06, + "loss": 0.2954, + "step": 50744 + }, + { + "epoch": 3.521268475470127, + "grad_norm": 0.858713686466217, + "learning_rate": 7.4191842319204285e-06, + "loss": 0.3162, + "step": 50745 + }, + { + "epoch": 3.5213378669072237, + "grad_norm": 0.8336383700370789, + "learning_rate": 7.417060576671531e-06, + "loss": 0.2567, + "step": 50746 + }, + { + "epoch": 3.52140725834432, + "grad_norm": 0.638919472694397, + "learning_rate": 7.414937213696249e-06, + "loss": 0.1761, + "step": 50747 + }, + { + "epoch": 3.521476649781417, + "grad_norm": 1.8993815183639526, + "learning_rate": 7.412814143001257e-06, + "loss": 0.2339, + "step": 50748 + }, + { + "epoch": 3.521546041218514, + "grad_norm": 1.0722315311431885, + "learning_rate": 7.41069136459327e-06, + "loss": 0.3303, + "step": 50749 + }, + { + "epoch": 3.52161543265561, + "grad_norm": 0.830101728439331, + "learning_rate": 7.408568878478994e-06, + "loss": 0.1698, + "step": 50750 + }, + { + "epoch": 3.521684824092707, + "grad_norm": 1.0750372409820557, + "learning_rate": 7.4064466846651244e-06, + "loss": 0.342, + "step": 50751 + }, + { + "epoch": 3.5217542155298034, + "grad_norm": 1.036428689956665, + "learning_rate": 7.404324783158367e-06, + "loss": 0.2846, + "step": 50752 + }, + { + "epoch": 3.5218236069669, + "grad_norm": 1.0945566892623901, + "learning_rate": 7.402203173965405e-06, + "loss": 0.3202, + "step": 50753 + }, + { + "epoch": 3.521892998403997, + "grad_norm": 1.019407868385315, + "learning_rate": 7.400081857092944e-06, + "loss": 0.2454, + "step": 50754 + }, + { + "epoch": 3.5219623898410934, + "grad_norm": 0.633504331111908, + "learning_rate": 7.397960832547668e-06, + "loss": 0.1706, + "step": 50755 + }, + { + "epoch": 3.5220317812781903, + "grad_norm": 1.351157784461975, + "learning_rate": 7.395840100336304e-06, + "loss": 0.2905, + "step": 50756 + }, + { + "epoch": 3.5221011727152867, + "grad_norm": 1.4123401641845703, + "learning_rate": 7.393719660465515e-06, + "loss": 0.3212, + "step": 50757 + }, + { + "epoch": 3.5221705641523835, + "grad_norm": 1.0449333190917969, + "learning_rate": 7.3915995129420154e-06, + "loss": 0.2589, + "step": 50758 + }, + { + "epoch": 3.5222399555894803, + "grad_norm": 0.9067304134368896, + "learning_rate": 7.389479657772513e-06, + "loss": 0.2227, + "step": 50759 + }, + { + "epoch": 3.522309347026577, + "grad_norm": 1.161402702331543, + "learning_rate": 7.387360094963647e-06, + "loss": 0.1801, + "step": 50760 + }, + { + "epoch": 3.5223787384636736, + "grad_norm": 1.248011827468872, + "learning_rate": 7.385240824522155e-06, + "loss": 0.4057, + "step": 50761 + }, + { + "epoch": 3.5224481299007704, + "grad_norm": 0.9757800698280334, + "learning_rate": 7.3831218464546995e-06, + "loss": 0.1665, + "step": 50762 + }, + { + "epoch": 3.5225175213378668, + "grad_norm": 0.8455264568328857, + "learning_rate": 7.381003160767996e-06, + "loss": 0.1973, + "step": 50763 + }, + { + "epoch": 3.5225869127749636, + "grad_norm": 0.6013631820678711, + "learning_rate": 7.3788847674687075e-06, + "loss": 0.1326, + "step": 50764 + }, + { + "epoch": 3.5226563042120604, + "grad_norm": 0.9939972758293152, + "learning_rate": 7.37676666656354e-06, + "loss": 0.2711, + "step": 50765 + }, + { + "epoch": 3.522725695649157, + "grad_norm": 0.8520157933235168, + "learning_rate": 7.374648858059196e-06, + "loss": 0.3386, + "step": 50766 + }, + { + "epoch": 3.5227950870862537, + "grad_norm": 0.8991496562957764, + "learning_rate": 7.372531341962319e-06, + "loss": 0.1953, + "step": 50767 + }, + { + "epoch": 3.52286447852335, + "grad_norm": 0.783169686794281, + "learning_rate": 7.370414118279612e-06, + "loss": 0.2125, + "step": 50768 + }, + { + "epoch": 3.522933869960447, + "grad_norm": 0.6978095173835754, + "learning_rate": 7.368297187017769e-06, + "loss": 0.2625, + "step": 50769 + }, + { + "epoch": 3.5230032613975437, + "grad_norm": 1.1190505027770996, + "learning_rate": 7.366180548183455e-06, + "loss": 0.2483, + "step": 50770 + }, + { + "epoch": 3.52307265283464, + "grad_norm": 1.0781441926956177, + "learning_rate": 7.36406420178336e-06, + "loss": 0.271, + "step": 50771 + }, + { + "epoch": 3.523142044271737, + "grad_norm": 0.6901012063026428, + "learning_rate": 7.3619481478241716e-06, + "loss": 0.2099, + "step": 50772 + }, + { + "epoch": 3.5232114357088333, + "grad_norm": 0.9677932262420654, + "learning_rate": 7.359832386312582e-06, + "loss": 0.2575, + "step": 50773 + }, + { + "epoch": 3.52328082714593, + "grad_norm": 0.8336142897605896, + "learning_rate": 7.357716917255231e-06, + "loss": 0.2737, + "step": 50774 + }, + { + "epoch": 3.523350218583027, + "grad_norm": 1.1326483488082886, + "learning_rate": 7.355601740658824e-06, + "loss": 0.2337, + "step": 50775 + }, + { + "epoch": 3.5234196100201234, + "grad_norm": 0.8895008563995361, + "learning_rate": 7.353486856530034e-06, + "loss": 0.2193, + "step": 50776 + }, + { + "epoch": 3.52348900145722, + "grad_norm": 0.974266529083252, + "learning_rate": 7.351372264875534e-06, + "loss": 0.2541, + "step": 50777 + }, + { + "epoch": 3.5235583928943166, + "grad_norm": 0.7724314332008362, + "learning_rate": 7.349257965701995e-06, + "loss": 0.2018, + "step": 50778 + }, + { + "epoch": 3.5236277843314134, + "grad_norm": 1.223752498626709, + "learning_rate": 7.347143959016112e-06, + "loss": 0.2571, + "step": 50779 + }, + { + "epoch": 3.5236971757685103, + "grad_norm": 0.8473122715950012, + "learning_rate": 7.345030244824558e-06, + "loss": 0.2209, + "step": 50780 + }, + { + "epoch": 3.523766567205607, + "grad_norm": 0.8375546932220459, + "learning_rate": 7.342916823133972e-06, + "loss": 0.3426, + "step": 50781 + }, + { + "epoch": 3.5238359586427035, + "grad_norm": 0.7749916911125183, + "learning_rate": 7.340803693951048e-06, + "loss": 0.1906, + "step": 50782 + }, + { + "epoch": 3.5239053500798, + "grad_norm": 1.0426310300827026, + "learning_rate": 7.33869085728246e-06, + "loss": 0.2538, + "step": 50783 + }, + { + "epoch": 3.5239747415168967, + "grad_norm": 0.8488715887069702, + "learning_rate": 7.336578313134879e-06, + "loss": 0.2876, + "step": 50784 + }, + { + "epoch": 3.5240441329539935, + "grad_norm": 0.7332643270492554, + "learning_rate": 7.334466061514967e-06, + "loss": 0.2212, + "step": 50785 + }, + { + "epoch": 3.5241135243910904, + "grad_norm": 0.794573962688446, + "learning_rate": 7.332354102429384e-06, + "loss": 0.2089, + "step": 50786 + }, + { + "epoch": 3.5241829158281868, + "grad_norm": 0.9851252436637878, + "learning_rate": 7.330242435884837e-06, + "loss": 0.2477, + "step": 50787 + }, + { + "epoch": 3.5242523072652836, + "grad_norm": 1.149886131286621, + "learning_rate": 7.328131061887933e-06, + "loss": 0.4188, + "step": 50788 + }, + { + "epoch": 3.52432169870238, + "grad_norm": 0.7182433605194092, + "learning_rate": 7.326019980445376e-06, + "loss": 0.1682, + "step": 50789 + }, + { + "epoch": 3.524391090139477, + "grad_norm": 0.6880061030387878, + "learning_rate": 7.323909191563827e-06, + "loss": 0.1983, + "step": 50790 + }, + { + "epoch": 3.5244604815765737, + "grad_norm": 0.6682506799697876, + "learning_rate": 7.321798695249937e-06, + "loss": 0.1632, + "step": 50791 + }, + { + "epoch": 3.52452987301367, + "grad_norm": 0.7059348821640015, + "learning_rate": 7.319688491510379e-06, + "loss": 0.1979, + "step": 50792 + }, + { + "epoch": 3.524599264450767, + "grad_norm": 0.839359700679779, + "learning_rate": 7.317578580351825e-06, + "loss": 0.3087, + "step": 50793 + }, + { + "epoch": 3.5246686558878633, + "grad_norm": 1.0496045351028442, + "learning_rate": 7.315468961780903e-06, + "loss": 0.3347, + "step": 50794 + }, + { + "epoch": 3.52473804732496, + "grad_norm": 1.125801682472229, + "learning_rate": 7.313359635804295e-06, + "loss": 0.2214, + "step": 50795 + }, + { + "epoch": 3.524807438762057, + "grad_norm": 0.7903860211372375, + "learning_rate": 7.3112506024286655e-06, + "loss": 0.229, + "step": 50796 + }, + { + "epoch": 3.5248768301991533, + "grad_norm": 0.9432170987129211, + "learning_rate": 7.309141861660651e-06, + "loss": 0.2786, + "step": 50797 + }, + { + "epoch": 3.52494622163625, + "grad_norm": 0.9007219076156616, + "learning_rate": 7.307033413506936e-06, + "loss": 0.2148, + "step": 50798 + }, + { + "epoch": 3.5250156130733465, + "grad_norm": 0.9287539124488831, + "learning_rate": 7.304925257974149e-06, + "loss": 0.2502, + "step": 50799 + }, + { + "epoch": 3.5250850045104434, + "grad_norm": 0.8935098648071289, + "learning_rate": 7.3028173950689835e-06, + "loss": 0.2337, + "step": 50800 + }, + { + "epoch": 3.52515439594754, + "grad_norm": 0.8142207264900208, + "learning_rate": 7.300709824798047e-06, + "loss": 0.1556, + "step": 50801 + }, + { + "epoch": 3.5252237873846366, + "grad_norm": 0.9828412532806396, + "learning_rate": 7.29860254716801e-06, + "loss": 0.415, + "step": 50802 + }, + { + "epoch": 3.5252931788217334, + "grad_norm": 1.7471818923950195, + "learning_rate": 7.296495562185534e-06, + "loss": 0.4529, + "step": 50803 + }, + { + "epoch": 3.52536257025883, + "grad_norm": 0.954423725605011, + "learning_rate": 7.294388869857283e-06, + "loss": 0.1623, + "step": 50804 + }, + { + "epoch": 3.5254319616959267, + "grad_norm": 1.406938910484314, + "learning_rate": 7.292282470189882e-06, + "loss": 0.281, + "step": 50805 + }, + { + "epoch": 3.5255013531330235, + "grad_norm": 1.1951916217803955, + "learning_rate": 7.290176363189971e-06, + "loss": 0.2994, + "step": 50806 + }, + { + "epoch": 3.5255707445701203, + "grad_norm": 1.2267998456954956, + "learning_rate": 7.2880705488642345e-06, + "loss": 0.2644, + "step": 50807 + }, + { + "epoch": 3.5256401360072167, + "grad_norm": 0.7204971313476562, + "learning_rate": 7.28596502721931e-06, + "loss": 0.2303, + "step": 50808 + }, + { + "epoch": 3.525709527444313, + "grad_norm": 0.8003344535827637, + "learning_rate": 7.283859798261827e-06, + "loss": 0.1797, + "step": 50809 + }, + { + "epoch": 3.52577891888141, + "grad_norm": 0.7519683837890625, + "learning_rate": 7.281754861998436e-06, + "loss": 0.2659, + "step": 50810 + }, + { + "epoch": 3.5258483103185068, + "grad_norm": 0.6216838955879211, + "learning_rate": 7.279650218435785e-06, + "loss": 0.2056, + "step": 50811 + }, + { + "epoch": 3.5259177017556036, + "grad_norm": 1.3635191917419434, + "learning_rate": 7.2775458675805266e-06, + "loss": 0.4384, + "step": 50812 + }, + { + "epoch": 3.5259870931927, + "grad_norm": 1.3733593225479126, + "learning_rate": 7.275441809439299e-06, + "loss": 0.3306, + "step": 50813 + }, + { + "epoch": 3.526056484629797, + "grad_norm": 0.7908677458763123, + "learning_rate": 7.2733380440187514e-06, + "loss": 0.2636, + "step": 50814 + }, + { + "epoch": 3.526125876066893, + "grad_norm": 0.8701903820037842, + "learning_rate": 7.271234571325502e-06, + "loss": 0.2532, + "step": 50815 + }, + { + "epoch": 3.52619526750399, + "grad_norm": 0.8321870565414429, + "learning_rate": 7.269131391366213e-06, + "loss": 0.203, + "step": 50816 + }, + { + "epoch": 3.526264658941087, + "grad_norm": 0.6997989416122437, + "learning_rate": 7.267028504147522e-06, + "loss": 0.1758, + "step": 50817 + }, + { + "epoch": 3.5263340503781833, + "grad_norm": 1.1006866693496704, + "learning_rate": 7.264925909676057e-06, + "loss": 0.3953, + "step": 50818 + }, + { + "epoch": 3.52640344181528, + "grad_norm": 0.7510331869125366, + "learning_rate": 7.2628236079584575e-06, + "loss": 0.197, + "step": 50819 + }, + { + "epoch": 3.5264728332523765, + "grad_norm": 1.1271194219589233, + "learning_rate": 7.260721599001364e-06, + "loss": 0.2475, + "step": 50820 + }, + { + "epoch": 3.5265422246894733, + "grad_norm": 1.4521005153656006, + "learning_rate": 7.258619882811435e-06, + "loss": 0.3025, + "step": 50821 + }, + { + "epoch": 3.52661161612657, + "grad_norm": 0.9401691555976868, + "learning_rate": 7.256518459395267e-06, + "loss": 0.1673, + "step": 50822 + }, + { + "epoch": 3.5266810075636665, + "grad_norm": 1.002315878868103, + "learning_rate": 7.254417328759499e-06, + "loss": 0.2448, + "step": 50823 + }, + { + "epoch": 3.5267503990007634, + "grad_norm": 0.9841986894607544, + "learning_rate": 7.252316490910793e-06, + "loss": 0.2439, + "step": 50824 + }, + { + "epoch": 3.5268197904378598, + "grad_norm": 0.5363581776618958, + "learning_rate": 7.250215945855754e-06, + "loss": 0.1327, + "step": 50825 + }, + { + "epoch": 3.5268891818749566, + "grad_norm": 1.725702166557312, + "learning_rate": 7.248115693601021e-06, + "loss": 0.2799, + "step": 50826 + }, + { + "epoch": 3.5269585733120534, + "grad_norm": 0.5429952144622803, + "learning_rate": 7.246015734153222e-06, + "loss": 0.213, + "step": 50827 + }, + { + "epoch": 3.52702796474915, + "grad_norm": 1.4286915063858032, + "learning_rate": 7.243916067519008e-06, + "loss": 0.3101, + "step": 50828 + }, + { + "epoch": 3.5270973561862466, + "grad_norm": 0.8937935829162598, + "learning_rate": 7.241816693704984e-06, + "loss": 0.1935, + "step": 50829 + }, + { + "epoch": 3.527166747623343, + "grad_norm": 0.3995019197463989, + "learning_rate": 7.239717612717767e-06, + "loss": 0.1197, + "step": 50830 + }, + { + "epoch": 3.52723613906044, + "grad_norm": 1.1369396448135376, + "learning_rate": 7.2376188245640075e-06, + "loss": 0.2103, + "step": 50831 + }, + { + "epoch": 3.5273055304975367, + "grad_norm": 0.6192894577980042, + "learning_rate": 7.235520329250323e-06, + "loss": 0.28, + "step": 50832 + }, + { + "epoch": 3.5273749219346335, + "grad_norm": 1.2410035133361816, + "learning_rate": 7.23342212678334e-06, + "loss": 0.249, + "step": 50833 + }, + { + "epoch": 3.52744431337173, + "grad_norm": 1.4645161628723145, + "learning_rate": 7.231324217169677e-06, + "loss": 0.5959, + "step": 50834 + }, + { + "epoch": 3.5275137048088268, + "grad_norm": 0.7270916700363159, + "learning_rate": 7.229226600415973e-06, + "loss": 0.151, + "step": 50835 + }, + { + "epoch": 3.527583096245923, + "grad_norm": 0.8436102271080017, + "learning_rate": 7.227129276528832e-06, + "loss": 0.2949, + "step": 50836 + }, + { + "epoch": 3.52765248768302, + "grad_norm": 1.0460915565490723, + "learning_rate": 7.2250322455148735e-06, + "loss": 0.4759, + "step": 50837 + }, + { + "epoch": 3.527721879120117, + "grad_norm": 1.2480602264404297, + "learning_rate": 7.222935507380724e-06, + "loss": 0.3377, + "step": 50838 + }, + { + "epoch": 3.527791270557213, + "grad_norm": 1.504714846611023, + "learning_rate": 7.220839062133011e-06, + "loss": 0.2456, + "step": 50839 + }, + { + "epoch": 3.52786066199431, + "grad_norm": 0.9557342529296875, + "learning_rate": 7.2187429097783425e-06, + "loss": 0.2951, + "step": 50840 + }, + { + "epoch": 3.5279300534314064, + "grad_norm": 0.9034571051597595, + "learning_rate": 7.216647050323333e-06, + "loss": 0.2457, + "step": 50841 + }, + { + "epoch": 3.5279994448685033, + "grad_norm": 1.0978467464447021, + "learning_rate": 7.214551483774601e-06, + "loss": 0.2389, + "step": 50842 + }, + { + "epoch": 3.5280688363056, + "grad_norm": 0.9197219014167786, + "learning_rate": 7.212456210138785e-06, + "loss": 0.2051, + "step": 50843 + }, + { + "epoch": 3.5281382277426965, + "grad_norm": 1.181834101676941, + "learning_rate": 7.210361229422458e-06, + "loss": 0.2088, + "step": 50844 + }, + { + "epoch": 3.5282076191797933, + "grad_norm": 0.8952240943908691, + "learning_rate": 7.208266541632258e-06, + "loss": 0.2722, + "step": 50845 + }, + { + "epoch": 3.5282770106168897, + "grad_norm": 1.0936686992645264, + "learning_rate": 7.206172146774803e-06, + "loss": 0.2088, + "step": 50846 + }, + { + "epoch": 3.5283464020539865, + "grad_norm": 1.203514575958252, + "learning_rate": 7.204078044856688e-06, + "loss": 0.3382, + "step": 50847 + }, + { + "epoch": 3.5284157934910834, + "grad_norm": 0.9052132964134216, + "learning_rate": 7.201984235884529e-06, + "loss": 0.2688, + "step": 50848 + }, + { + "epoch": 3.5284851849281798, + "grad_norm": 0.8540293574333191, + "learning_rate": 7.199890719864966e-06, + "loss": 0.1647, + "step": 50849 + }, + { + "epoch": 3.5285545763652766, + "grad_norm": 0.8612220287322998, + "learning_rate": 7.1977974968045595e-06, + "loss": 0.1576, + "step": 50850 + }, + { + "epoch": 3.528623967802373, + "grad_norm": 0.8779913783073425, + "learning_rate": 7.1957045667099375e-06, + "loss": 0.2447, + "step": 50851 + }, + { + "epoch": 3.52869335923947, + "grad_norm": 0.9658576250076294, + "learning_rate": 7.193611929587707e-06, + "loss": 0.2351, + "step": 50852 + }, + { + "epoch": 3.5287627506765666, + "grad_norm": 0.8140224814414978, + "learning_rate": 7.191519585444473e-06, + "loss": 0.2556, + "step": 50853 + }, + { + "epoch": 3.528832142113663, + "grad_norm": 0.6508554816246033, + "learning_rate": 7.189427534286852e-06, + "loss": 0.1974, + "step": 50854 + }, + { + "epoch": 3.52890153355076, + "grad_norm": 2.0038018226623535, + "learning_rate": 7.18733577612144e-06, + "loss": 0.1714, + "step": 50855 + }, + { + "epoch": 3.5289709249878563, + "grad_norm": 0.9005401730537415, + "learning_rate": 7.185244310954853e-06, + "loss": 0.2532, + "step": 50856 + }, + { + "epoch": 3.529040316424953, + "grad_norm": 0.7101171016693115, + "learning_rate": 7.183153138793663e-06, + "loss": 0.1814, + "step": 50857 + }, + { + "epoch": 3.52910970786205, + "grad_norm": 0.820889413356781, + "learning_rate": 7.1810622596445e-06, + "loss": 0.2963, + "step": 50858 + }, + { + "epoch": 3.5291790992991467, + "grad_norm": 0.7866686582565308, + "learning_rate": 7.178971673513946e-06, + "loss": 0.2093, + "step": 50859 + }, + { + "epoch": 3.529248490736243, + "grad_norm": 0.7832288146018982, + "learning_rate": 7.176881380408607e-06, + "loss": 0.1616, + "step": 50860 + }, + { + "epoch": 3.52931788217334, + "grad_norm": 1.309774398803711, + "learning_rate": 7.174791380335089e-06, + "loss": 0.2108, + "step": 50861 + }, + { + "epoch": 3.5293872736104364, + "grad_norm": 0.8511424660682678, + "learning_rate": 7.1727016732999864e-06, + "loss": 0.2089, + "step": 50862 + }, + { + "epoch": 3.529456665047533, + "grad_norm": 0.9624049067497253, + "learning_rate": 7.170612259309906e-06, + "loss": 0.21, + "step": 50863 + }, + { + "epoch": 3.52952605648463, + "grad_norm": 0.9044337868690491, + "learning_rate": 7.168523138371419e-06, + "loss": 0.247, + "step": 50864 + }, + { + "epoch": 3.5295954479217264, + "grad_norm": 1.0694687366485596, + "learning_rate": 7.166434310491132e-06, + "loss": 0.3719, + "step": 50865 + }, + { + "epoch": 3.5296648393588232, + "grad_norm": 0.8753448128700256, + "learning_rate": 7.164345775675652e-06, + "loss": 0.2252, + "step": 50866 + }, + { + "epoch": 3.5297342307959196, + "grad_norm": 0.9897847175598145, + "learning_rate": 7.162257533931549e-06, + "loss": 0.2952, + "step": 50867 + }, + { + "epoch": 3.5298036222330165, + "grad_norm": 0.7060150504112244, + "learning_rate": 7.16016958526543e-06, + "loss": 0.1778, + "step": 50868 + }, + { + "epoch": 3.5298730136701133, + "grad_norm": 0.8283201456069946, + "learning_rate": 7.158081929683902e-06, + "loss": 0.265, + "step": 50869 + }, + { + "epoch": 3.5299424051072097, + "grad_norm": 0.879734218120575, + "learning_rate": 7.155994567193525e-06, + "loss": 0.1876, + "step": 50870 + }, + { + "epoch": 3.5300117965443065, + "grad_norm": 0.7946597337722778, + "learning_rate": 7.153907497800893e-06, + "loss": 0.2805, + "step": 50871 + }, + { + "epoch": 3.530081187981403, + "grad_norm": 0.6088172197341919, + "learning_rate": 7.151820721512614e-06, + "loss": 0.1717, + "step": 50872 + }, + { + "epoch": 3.5301505794184997, + "grad_norm": 0.967464029788971, + "learning_rate": 7.149734238335248e-06, + "loss": 0.237, + "step": 50873 + }, + { + "epoch": 3.5302199708555966, + "grad_norm": 0.6823973059654236, + "learning_rate": 7.1476480482754346e-06, + "loss": 0.2211, + "step": 50874 + }, + { + "epoch": 3.530289362292693, + "grad_norm": 1.6259328126907349, + "learning_rate": 7.145562151339691e-06, + "loss": 0.248, + "step": 50875 + }, + { + "epoch": 3.53035875372979, + "grad_norm": 0.8411945700645447, + "learning_rate": 7.143476547534645e-06, + "loss": 0.201, + "step": 50876 + }, + { + "epoch": 3.530428145166886, + "grad_norm": 0.9203308820724487, + "learning_rate": 7.141391236866868e-06, + "loss": 0.2562, + "step": 50877 + }, + { + "epoch": 3.530497536603983, + "grad_norm": 0.5972353219985962, + "learning_rate": 7.139306219342956e-06, + "loss": 0.1674, + "step": 50878 + }, + { + "epoch": 3.53056692804108, + "grad_norm": 0.9333573579788208, + "learning_rate": 7.13722149496947e-06, + "loss": 0.2597, + "step": 50879 + }, + { + "epoch": 3.5306363194781762, + "grad_norm": 1.2583856582641602, + "learning_rate": 7.1351370637529945e-06, + "loss": 0.3742, + "step": 50880 + }, + { + "epoch": 3.530705710915273, + "grad_norm": 1.2188148498535156, + "learning_rate": 7.133052925700123e-06, + "loss": 0.238, + "step": 50881 + }, + { + "epoch": 3.5307751023523695, + "grad_norm": 0.9957390427589417, + "learning_rate": 7.130969080817429e-06, + "loss": 0.306, + "step": 50882 + }, + { + "epoch": 3.5308444937894663, + "grad_norm": 0.9214867949485779, + "learning_rate": 7.128885529111507e-06, + "loss": 0.294, + "step": 50883 + }, + { + "epoch": 3.530913885226563, + "grad_norm": 1.1064811944961548, + "learning_rate": 7.126802270588895e-06, + "loss": 0.365, + "step": 50884 + }, + { + "epoch": 3.53098327666366, + "grad_norm": 0.9313248991966248, + "learning_rate": 7.1247193052562e-06, + "loss": 0.2288, + "step": 50885 + }, + { + "epoch": 3.5310526681007564, + "grad_norm": 1.359588861465454, + "learning_rate": 7.1226366331199945e-06, + "loss": 0.2608, + "step": 50886 + }, + { + "epoch": 3.531122059537853, + "grad_norm": 0.9251299500465393, + "learning_rate": 7.12055425418684e-06, + "loss": 0.2133, + "step": 50887 + }, + { + "epoch": 3.5311914509749496, + "grad_norm": 0.997686505317688, + "learning_rate": 7.118472168463319e-06, + "loss": 0.2545, + "step": 50888 + }, + { + "epoch": 3.5312608424120464, + "grad_norm": 0.6966528296470642, + "learning_rate": 7.116390375956017e-06, + "loss": 0.2157, + "step": 50889 + }, + { + "epoch": 3.5313302338491432, + "grad_norm": 0.8314130306243896, + "learning_rate": 7.114308876671505e-06, + "loss": 0.2541, + "step": 50890 + }, + { + "epoch": 3.5313996252862396, + "grad_norm": 0.6750229597091675, + "learning_rate": 7.112227670616322e-06, + "loss": 0.1338, + "step": 50891 + }, + { + "epoch": 3.5314690167233365, + "grad_norm": 0.9653483033180237, + "learning_rate": 7.110146757797065e-06, + "loss": 0.3513, + "step": 50892 + }, + { + "epoch": 3.531538408160433, + "grad_norm": 0.6737700700759888, + "learning_rate": 7.1080661382202925e-06, + "loss": 0.1948, + "step": 50893 + }, + { + "epoch": 3.5316077995975297, + "grad_norm": 1.0524132251739502, + "learning_rate": 7.105985811892579e-06, + "loss": 0.3151, + "step": 50894 + }, + { + "epoch": 3.5316771910346265, + "grad_norm": 0.6920551061630249, + "learning_rate": 7.1039057788204965e-06, + "loss": 0.2149, + "step": 50895 + }, + { + "epoch": 3.531746582471723, + "grad_norm": 0.8131616711616516, + "learning_rate": 7.101826039010595e-06, + "loss": 0.2031, + "step": 50896 + }, + { + "epoch": 3.5318159739088197, + "grad_norm": 0.9221621155738831, + "learning_rate": 7.09974659246948e-06, + "loss": 0.2389, + "step": 50897 + }, + { + "epoch": 3.531885365345916, + "grad_norm": 1.029379963874817, + "learning_rate": 7.097667439203659e-06, + "loss": 0.2212, + "step": 50898 + }, + { + "epoch": 3.531954756783013, + "grad_norm": 1.0737199783325195, + "learning_rate": 7.095588579219725e-06, + "loss": 0.2355, + "step": 50899 + }, + { + "epoch": 3.53202414822011, + "grad_norm": 1.1455899477005005, + "learning_rate": 7.093510012524241e-06, + "loss": 0.1862, + "step": 50900 + }, + { + "epoch": 3.532093539657206, + "grad_norm": 0.8569173216819763, + "learning_rate": 7.091431739123766e-06, + "loss": 0.2463, + "step": 50901 + }, + { + "epoch": 3.532162931094303, + "grad_norm": 0.9940087795257568, + "learning_rate": 7.089353759024864e-06, + "loss": 0.2165, + "step": 50902 + }, + { + "epoch": 3.5322323225313994, + "grad_norm": 0.8237994909286499, + "learning_rate": 7.087276072234095e-06, + "loss": 0.2567, + "step": 50903 + }, + { + "epoch": 3.5323017139684962, + "grad_norm": 0.9543784856796265, + "learning_rate": 7.085198678758021e-06, + "loss": 0.2908, + "step": 50904 + }, + { + "epoch": 3.532371105405593, + "grad_norm": 0.8652216196060181, + "learning_rate": 7.083121578603191e-06, + "loss": 0.2087, + "step": 50905 + }, + { + "epoch": 3.5324404968426895, + "grad_norm": 1.2169163227081299, + "learning_rate": 7.081044771776158e-06, + "loss": 0.4057, + "step": 50906 + }, + { + "epoch": 3.5325098882797863, + "grad_norm": 0.6041097044944763, + "learning_rate": 7.078968258283492e-06, + "loss": 0.1484, + "step": 50907 + }, + { + "epoch": 3.5325792797168827, + "grad_norm": 0.7529955506324768, + "learning_rate": 7.076892038131732e-06, + "loss": 0.2346, + "step": 50908 + }, + { + "epoch": 3.5326486711539795, + "grad_norm": 1.2329192161560059, + "learning_rate": 7.074816111327465e-06, + "loss": 0.3082, + "step": 50909 + }, + { + "epoch": 3.5327180625910763, + "grad_norm": 0.9185081720352173, + "learning_rate": 7.0727404778772155e-06, + "loss": 0.2436, + "step": 50910 + }, + { + "epoch": 3.532787454028173, + "grad_norm": 0.9820639491081238, + "learning_rate": 7.070665137787525e-06, + "loss": 0.302, + "step": 50911 + }, + { + "epoch": 3.5328568454652696, + "grad_norm": 0.704603910446167, + "learning_rate": 7.068590091064997e-06, + "loss": 0.182, + "step": 50912 + }, + { + "epoch": 3.5329262369023664, + "grad_norm": 1.2598234415054321, + "learning_rate": 7.066515337716117e-06, + "loss": 0.2333, + "step": 50913 + }, + { + "epoch": 3.532995628339463, + "grad_norm": 0.8946248888969421, + "learning_rate": 7.064440877747469e-06, + "loss": 0.2148, + "step": 50914 + }, + { + "epoch": 3.5330650197765596, + "grad_norm": 3.0432093143463135, + "learning_rate": 7.062366711165602e-06, + "loss": 0.3025, + "step": 50915 + }, + { + "epoch": 3.5331344112136565, + "grad_norm": 1.2116974592208862, + "learning_rate": 7.060292837977067e-06, + "loss": 0.284, + "step": 50916 + }, + { + "epoch": 3.533203802650753, + "grad_norm": 1.1411207914352417, + "learning_rate": 7.058219258188392e-06, + "loss": 0.3075, + "step": 50917 + }, + { + "epoch": 3.5332731940878497, + "grad_norm": 0.7057231664657593, + "learning_rate": 7.05614597180616e-06, + "loss": 0.2115, + "step": 50918 + }, + { + "epoch": 3.533342585524946, + "grad_norm": 0.915667712688446, + "learning_rate": 7.054072978836868e-06, + "loss": 0.2542, + "step": 50919 + }, + { + "epoch": 3.533411976962043, + "grad_norm": 0.8585625290870667, + "learning_rate": 7.052000279287085e-06, + "loss": 0.3863, + "step": 50920 + }, + { + "epoch": 3.5334813683991397, + "grad_norm": 1.0204898118972778, + "learning_rate": 7.0499278731633645e-06, + "loss": 0.2324, + "step": 50921 + }, + { + "epoch": 3.533550759836236, + "grad_norm": 1.1623228788375854, + "learning_rate": 7.047855760472222e-06, + "loss": 0.2825, + "step": 50922 + }, + { + "epoch": 3.533620151273333, + "grad_norm": 0.9963346123695374, + "learning_rate": 7.04578394122023e-06, + "loss": 0.2972, + "step": 50923 + }, + { + "epoch": 3.5336895427104293, + "grad_norm": 0.8704785704612732, + "learning_rate": 7.043712415413906e-06, + "loss": 0.2706, + "step": 50924 + }, + { + "epoch": 3.533758934147526, + "grad_norm": 0.7330912947654724, + "learning_rate": 7.04164118305981e-06, + "loss": 0.1818, + "step": 50925 + }, + { + "epoch": 3.533828325584623, + "grad_norm": 1.1298168897628784, + "learning_rate": 7.039570244164451e-06, + "loss": 0.3213, + "step": 50926 + }, + { + "epoch": 3.5338977170217194, + "grad_norm": 0.8299727439880371, + "learning_rate": 7.037499598734387e-06, + "loss": 0.3051, + "step": 50927 + }, + { + "epoch": 3.5339671084588162, + "grad_norm": 0.7853378057479858, + "learning_rate": 7.035429246776148e-06, + "loss": 0.2571, + "step": 50928 + }, + { + "epoch": 3.5340364998959126, + "grad_norm": 1.2030243873596191, + "learning_rate": 7.0333591882962735e-06, + "loss": 0.293, + "step": 50929 + }, + { + "epoch": 3.5341058913330095, + "grad_norm": 0.989844799041748, + "learning_rate": 7.031289423301302e-06, + "loss": 0.2425, + "step": 50930 + }, + { + "epoch": 3.5341752827701063, + "grad_norm": 0.8204370737075806, + "learning_rate": 7.029219951797761e-06, + "loss": 0.2573, + "step": 50931 + }, + { + "epoch": 3.5342446742072027, + "grad_norm": 1.1462900638580322, + "learning_rate": 7.027150773792201e-06, + "loss": 0.248, + "step": 50932 + }, + { + "epoch": 3.5343140656442995, + "grad_norm": 0.896367073059082, + "learning_rate": 7.02508188929113e-06, + "loss": 0.2556, + "step": 50933 + }, + { + "epoch": 3.534383457081396, + "grad_norm": 0.8067660331726074, + "learning_rate": 7.023013298301085e-06, + "loss": 0.2039, + "step": 50934 + }, + { + "epoch": 3.5344528485184927, + "grad_norm": 0.9430893063545227, + "learning_rate": 7.020945000828594e-06, + "loss": 0.2042, + "step": 50935 + }, + { + "epoch": 3.5345222399555896, + "grad_norm": 0.6609677672386169, + "learning_rate": 7.018876996880197e-06, + "loss": 0.2331, + "step": 50936 + }, + { + "epoch": 3.5345916313926864, + "grad_norm": 0.7880761623382568, + "learning_rate": 7.016809286462412e-06, + "loss": 0.2651, + "step": 50937 + }, + { + "epoch": 3.534661022829783, + "grad_norm": 0.8358461260795593, + "learning_rate": 7.014741869581775e-06, + "loss": 0.2296, + "step": 50938 + }, + { + "epoch": 3.5347304142668796, + "grad_norm": 0.8402161002159119, + "learning_rate": 7.012674746244829e-06, + "loss": 0.2499, + "step": 50939 + }, + { + "epoch": 3.534799805703976, + "grad_norm": 0.9472066164016724, + "learning_rate": 7.0106079164580676e-06, + "loss": 0.2044, + "step": 50940 + }, + { + "epoch": 3.534869197141073, + "grad_norm": 0.7677598595619202, + "learning_rate": 7.0085413802280175e-06, + "loss": 0.2426, + "step": 50941 + }, + { + "epoch": 3.5349385885781697, + "grad_norm": 0.9469983577728271, + "learning_rate": 7.006475137561219e-06, + "loss": 0.2332, + "step": 50942 + }, + { + "epoch": 3.535007980015266, + "grad_norm": 0.9411315321922302, + "learning_rate": 7.004409188464189e-06, + "loss": 0.2606, + "step": 50943 + }, + { + "epoch": 3.535077371452363, + "grad_norm": 0.8582317233085632, + "learning_rate": 7.0023435329434674e-06, + "loss": 0.2374, + "step": 50944 + }, + { + "epoch": 3.5351467628894593, + "grad_norm": 0.992544949054718, + "learning_rate": 7.0002781710055365e-06, + "loss": 0.3479, + "step": 50945 + }, + { + "epoch": 3.535216154326556, + "grad_norm": 0.666736364364624, + "learning_rate": 6.998213102656936e-06, + "loss": 0.2775, + "step": 50946 + }, + { + "epoch": 3.535285545763653, + "grad_norm": 0.5731365084648132, + "learning_rate": 6.996148327904206e-06, + "loss": 0.1242, + "step": 50947 + }, + { + "epoch": 3.5353549372007493, + "grad_norm": 0.9272359609603882, + "learning_rate": 6.99408384675384e-06, + "loss": 0.2292, + "step": 50948 + }, + { + "epoch": 3.535424328637846, + "grad_norm": 0.8102331161499023, + "learning_rate": 6.992019659212346e-06, + "loss": 0.2486, + "step": 50949 + }, + { + "epoch": 3.5354937200749426, + "grad_norm": 1.0498380661010742, + "learning_rate": 6.98995576528626e-06, + "loss": 0.323, + "step": 50950 + }, + { + "epoch": 3.5355631115120394, + "grad_norm": 0.8196518421173096, + "learning_rate": 6.98789216498209e-06, + "loss": 0.216, + "step": 50951 + }, + { + "epoch": 3.5356325029491362, + "grad_norm": 0.7559342980384827, + "learning_rate": 6.985828858306376e-06, + "loss": 0.3017, + "step": 50952 + }, + { + "epoch": 3.5357018943862326, + "grad_norm": 1.3071545362472534, + "learning_rate": 6.9837658452655885e-06, + "loss": 0.3221, + "step": 50953 + }, + { + "epoch": 3.5357712858233294, + "grad_norm": 1.584025263786316, + "learning_rate": 6.981703125866257e-06, + "loss": 0.3741, + "step": 50954 + }, + { + "epoch": 3.535840677260426, + "grad_norm": 0.8353179693222046, + "learning_rate": 6.979640700114898e-06, + "loss": 0.1841, + "step": 50955 + }, + { + "epoch": 3.5359100686975227, + "grad_norm": 1.4200035333633423, + "learning_rate": 6.977578568018028e-06, + "loss": 0.2249, + "step": 50956 + }, + { + "epoch": 3.5359794601346195, + "grad_norm": 0.8566252589225769, + "learning_rate": 6.975516729582143e-06, + "loss": 0.1851, + "step": 50957 + }, + { + "epoch": 3.536048851571716, + "grad_norm": 1.2587122917175293, + "learning_rate": 6.97345518481376e-06, + "loss": 0.3333, + "step": 50958 + }, + { + "epoch": 3.5361182430088127, + "grad_norm": 0.8043529391288757, + "learning_rate": 6.971393933719394e-06, + "loss": 0.1729, + "step": 50959 + }, + { + "epoch": 3.536187634445909, + "grad_norm": 0.49997323751449585, + "learning_rate": 6.969332976305542e-06, + "loss": 0.1631, + "step": 50960 + }, + { + "epoch": 3.536257025883006, + "grad_norm": 0.6785615086555481, + "learning_rate": 6.9672723125787096e-06, + "loss": 0.2071, + "step": 50961 + }, + { + "epoch": 3.5363264173201028, + "grad_norm": 1.0447142124176025, + "learning_rate": 6.965211942545402e-06, + "loss": 0.2288, + "step": 50962 + }, + { + "epoch": 3.5363958087571996, + "grad_norm": 1.1387040615081787, + "learning_rate": 6.963151866212125e-06, + "loss": 0.3732, + "step": 50963 + }, + { + "epoch": 3.536465200194296, + "grad_norm": 0.6395941376686096, + "learning_rate": 6.961092083585386e-06, + "loss": 0.1733, + "step": 50964 + }, + { + "epoch": 3.536534591631393, + "grad_norm": 1.003395915031433, + "learning_rate": 6.959032594671677e-06, + "loss": 0.2664, + "step": 50965 + }, + { + "epoch": 3.536603983068489, + "grad_norm": 0.5799301862716675, + "learning_rate": 6.956973399477529e-06, + "loss": 0.1259, + "step": 50966 + }, + { + "epoch": 3.536673374505586, + "grad_norm": 0.7392657399177551, + "learning_rate": 6.954914498009402e-06, + "loss": 0.2594, + "step": 50967 + }, + { + "epoch": 3.536742765942683, + "grad_norm": 0.68059241771698, + "learning_rate": 6.9528558902738236e-06, + "loss": 0.2606, + "step": 50968 + }, + { + "epoch": 3.5368121573797793, + "grad_norm": 0.9851245284080505, + "learning_rate": 6.950797576277279e-06, + "loss": 0.2427, + "step": 50969 + }, + { + "epoch": 3.536881548816876, + "grad_norm": 1.2062591314315796, + "learning_rate": 6.948739556026262e-06, + "loss": 0.3239, + "step": 50970 + }, + { + "epoch": 3.5369509402539725, + "grad_norm": 0.4842660129070282, + "learning_rate": 6.946681829527291e-06, + "loss": 0.1131, + "step": 50971 + }, + { + "epoch": 3.5370203316910693, + "grad_norm": 1.2074732780456543, + "learning_rate": 6.944624396786847e-06, + "loss": 0.3103, + "step": 50972 + }, + { + "epoch": 3.537089723128166, + "grad_norm": 1.13558828830719, + "learning_rate": 6.9425672578114385e-06, + "loss": 0.2998, + "step": 50973 + }, + { + "epoch": 3.5371591145652626, + "grad_norm": 1.1131443977355957, + "learning_rate": 6.940510412607537e-06, + "loss": 0.2547, + "step": 50974 + }, + { + "epoch": 3.5372285060023594, + "grad_norm": 1.024876594543457, + "learning_rate": 6.938453861181638e-06, + "loss": 0.309, + "step": 50975 + }, + { + "epoch": 3.5372978974394558, + "grad_norm": 0.8472862839698792, + "learning_rate": 6.936397603540256e-06, + "loss": 0.2004, + "step": 50976 + }, + { + "epoch": 3.5373672888765526, + "grad_norm": 1.0686308145523071, + "learning_rate": 6.9343416396898565e-06, + "loss": 0.2874, + "step": 50977 + }, + { + "epoch": 3.5374366803136494, + "grad_norm": 1.143541693687439, + "learning_rate": 6.9322859696369645e-06, + "loss": 0.236, + "step": 50978 + }, + { + "epoch": 3.537506071750746, + "grad_norm": 0.9788593649864197, + "learning_rate": 6.930230593388043e-06, + "loss": 0.3269, + "step": 50979 + }, + { + "epoch": 3.5375754631878427, + "grad_norm": 2.0245628356933594, + "learning_rate": 6.928175510949575e-06, + "loss": 0.4069, + "step": 50980 + }, + { + "epoch": 3.537644854624939, + "grad_norm": 1.3075138330459595, + "learning_rate": 6.926120722328056e-06, + "loss": 0.2661, + "step": 50981 + }, + { + "epoch": 3.537714246062036, + "grad_norm": 1.109374761581421, + "learning_rate": 6.924066227529991e-06, + "loss": 0.2469, + "step": 50982 + }, + { + "epoch": 3.5377836374991327, + "grad_norm": 1.280511736869812, + "learning_rate": 6.922012026561841e-06, + "loss": 0.3332, + "step": 50983 + }, + { + "epoch": 3.537853028936229, + "grad_norm": 1.0790998935699463, + "learning_rate": 6.919958119430092e-06, + "loss": 0.4096, + "step": 50984 + }, + { + "epoch": 3.537922420373326, + "grad_norm": 0.8764034509658813, + "learning_rate": 6.917904506141237e-06, + "loss": 0.1738, + "step": 50985 + }, + { + "epoch": 3.5379918118104223, + "grad_norm": 1.0579586029052734, + "learning_rate": 6.915851186701761e-06, + "loss": 0.3671, + "step": 50986 + }, + { + "epoch": 3.538061203247519, + "grad_norm": 0.8269590139389038, + "learning_rate": 6.913798161118168e-06, + "loss": 0.2815, + "step": 50987 + }, + { + "epoch": 3.538130594684616, + "grad_norm": 0.8872512578964233, + "learning_rate": 6.911745429396887e-06, + "loss": 0.2164, + "step": 50988 + }, + { + "epoch": 3.538199986121713, + "grad_norm": 0.949144184589386, + "learning_rate": 6.909692991544425e-06, + "loss": 0.2817, + "step": 50989 + }, + { + "epoch": 3.538269377558809, + "grad_norm": 1.137818455696106, + "learning_rate": 6.907640847567265e-06, + "loss": 0.3393, + "step": 50990 + }, + { + "epoch": 3.538338768995906, + "grad_norm": 0.9046066403388977, + "learning_rate": 6.905588997471879e-06, + "loss": 0.3832, + "step": 50991 + }, + { + "epoch": 3.5384081604330024, + "grad_norm": 0.8092138767242432, + "learning_rate": 6.903537441264751e-06, + "loss": 0.2622, + "step": 50992 + }, + { + "epoch": 3.5384775518700993, + "grad_norm": 0.8197333216667175, + "learning_rate": 6.901486178952354e-06, + "loss": 0.2129, + "step": 50993 + }, + { + "epoch": 3.538546943307196, + "grad_norm": 0.6979339122772217, + "learning_rate": 6.899435210541183e-06, + "loss": 0.2117, + "step": 50994 + }, + { + "epoch": 3.5386163347442925, + "grad_norm": 1.0987094640731812, + "learning_rate": 6.897384536037677e-06, + "loss": 0.341, + "step": 50995 + }, + { + "epoch": 3.5386857261813893, + "grad_norm": 0.7102974057197571, + "learning_rate": 6.895334155448319e-06, + "loss": 0.1655, + "step": 50996 + }, + { + "epoch": 3.5387551176184857, + "grad_norm": 0.9154794812202454, + "learning_rate": 6.893284068779593e-06, + "loss": 0.2256, + "step": 50997 + }, + { + "epoch": 3.5388245090555825, + "grad_norm": 1.006314992904663, + "learning_rate": 6.891234276037961e-06, + "loss": 0.232, + "step": 50998 + }, + { + "epoch": 3.5388939004926794, + "grad_norm": 1.447400450706482, + "learning_rate": 6.889184777229907e-06, + "loss": 0.481, + "step": 50999 + }, + { + "epoch": 3.5389632919297758, + "grad_norm": 0.678876519203186, + "learning_rate": 6.887135572361891e-06, + "loss": 0.2631, + "step": 51000 + }, + { + "epoch": 3.5390326833668726, + "grad_norm": 0.8246655464172363, + "learning_rate": 6.88508666144041e-06, + "loss": 0.1703, + "step": 51001 + }, + { + "epoch": 3.539102074803969, + "grad_norm": 1.0467387437820435, + "learning_rate": 6.883038044471879e-06, + "loss": 0.2464, + "step": 51002 + }, + { + "epoch": 3.539171466241066, + "grad_norm": 0.8526728749275208, + "learning_rate": 6.8809897214627935e-06, + "loss": 0.2694, + "step": 51003 + }, + { + "epoch": 3.5392408576781627, + "grad_norm": 1.183964490890503, + "learning_rate": 6.878941692419627e-06, + "loss": 0.4365, + "step": 51004 + }, + { + "epoch": 3.539310249115259, + "grad_norm": 0.8470232486724854, + "learning_rate": 6.876893957348829e-06, + "loss": 0.2912, + "step": 51005 + }, + { + "epoch": 3.539379640552356, + "grad_norm": 1.1173416376113892, + "learning_rate": 6.8748465162568835e-06, + "loss": 0.2834, + "step": 51006 + }, + { + "epoch": 3.5394490319894523, + "grad_norm": 1.1100622415542603, + "learning_rate": 6.87279936915024e-06, + "loss": 0.2525, + "step": 51007 + }, + { + "epoch": 3.539518423426549, + "grad_norm": 0.6650618314743042, + "learning_rate": 6.8707525160353726e-06, + "loss": 0.1684, + "step": 51008 + }, + { + "epoch": 3.539587814863646, + "grad_norm": 0.8517285585403442, + "learning_rate": 6.86870595691872e-06, + "loss": 0.1716, + "step": 51009 + }, + { + "epoch": 3.5396572063007428, + "grad_norm": 1.1907176971435547, + "learning_rate": 6.866659691806765e-06, + "loss": 0.3054, + "step": 51010 + }, + { + "epoch": 3.539726597737839, + "grad_norm": 0.7605210542678833, + "learning_rate": 6.864613720705959e-06, + "loss": 0.2484, + "step": 51011 + }, + { + "epoch": 3.5397959891749355, + "grad_norm": 1.7417405843734741, + "learning_rate": 6.8625680436227526e-06, + "loss": 0.3373, + "step": 51012 + }, + { + "epoch": 3.5398653806120324, + "grad_norm": 0.7977391481399536, + "learning_rate": 6.86052266056364e-06, + "loss": 0.3026, + "step": 51013 + }, + { + "epoch": 3.539934772049129, + "grad_norm": 1.0013660192489624, + "learning_rate": 6.858477571535027e-06, + "loss": 0.2959, + "step": 51014 + }, + { + "epoch": 3.540004163486226, + "grad_norm": 0.6039825081825256, + "learning_rate": 6.8564327765433975e-06, + "loss": 0.1332, + "step": 51015 + }, + { + "epoch": 3.5400735549233224, + "grad_norm": 0.994111955165863, + "learning_rate": 6.854388275595202e-06, + "loss": 0.2151, + "step": 51016 + }, + { + "epoch": 3.5401429463604193, + "grad_norm": 0.8720662593841553, + "learning_rate": 6.852344068696914e-06, + "loss": 0.2172, + "step": 51017 + }, + { + "epoch": 3.5402123377975157, + "grad_norm": 0.8356975317001343, + "learning_rate": 6.8503001558549494e-06, + "loss": 0.2328, + "step": 51018 + }, + { + "epoch": 3.5402817292346125, + "grad_norm": 1.1828467845916748, + "learning_rate": 6.848256537075781e-06, + "loss": 0.3407, + "step": 51019 + }, + { + "epoch": 3.5403511206717093, + "grad_norm": 1.133840560913086, + "learning_rate": 6.84621321236586e-06, + "loss": 0.2549, + "step": 51020 + }, + { + "epoch": 3.5404205121088057, + "grad_norm": 0.8290736079216003, + "learning_rate": 6.8441701817316465e-06, + "loss": 0.1842, + "step": 51021 + }, + { + "epoch": 3.5404899035459025, + "grad_norm": 1.0319887399673462, + "learning_rate": 6.842127445179569e-06, + "loss": 0.306, + "step": 51022 + }, + { + "epoch": 3.540559294982999, + "grad_norm": 1.4697058200836182, + "learning_rate": 6.8400850027160905e-06, + "loss": 0.352, + "step": 51023 + }, + { + "epoch": 3.5406286864200958, + "grad_norm": 0.711254358291626, + "learning_rate": 6.838042854347648e-06, + "loss": 0.215, + "step": 51024 + }, + { + "epoch": 3.5406980778571926, + "grad_norm": 1.6522225141525269, + "learning_rate": 6.836001000080694e-06, + "loss": 0.3658, + "step": 51025 + }, + { + "epoch": 3.540767469294289, + "grad_norm": 1.3030458688735962, + "learning_rate": 6.833959439921678e-06, + "loss": 0.2278, + "step": 51026 + }, + { + "epoch": 3.540836860731386, + "grad_norm": 0.8043470978736877, + "learning_rate": 6.83191817387705e-06, + "loss": 0.2347, + "step": 51027 + }, + { + "epoch": 3.540906252168482, + "grad_norm": 0.9223414659500122, + "learning_rate": 6.829877201953261e-06, + "loss": 0.1917, + "step": 51028 + }, + { + "epoch": 3.540975643605579, + "grad_norm": 0.7725226879119873, + "learning_rate": 6.827836524156716e-06, + "loss": 0.2087, + "step": 51029 + }, + { + "epoch": 3.541045035042676, + "grad_norm": 0.7750669121742249, + "learning_rate": 6.8257961404938896e-06, + "loss": 0.2289, + "step": 51030 + }, + { + "epoch": 3.5411144264797723, + "grad_norm": 1.6952757835388184, + "learning_rate": 6.823756050971208e-06, + "loss": 0.1534, + "step": 51031 + }, + { + "epoch": 3.541183817916869, + "grad_norm": 0.7981616258621216, + "learning_rate": 6.821716255595123e-06, + "loss": 0.231, + "step": 51032 + }, + { + "epoch": 3.5412532093539655, + "grad_norm": 1.5927355289459229, + "learning_rate": 6.819676754372062e-06, + "loss": 0.2351, + "step": 51033 + }, + { + "epoch": 3.5413226007910623, + "grad_norm": 0.7108435034751892, + "learning_rate": 6.8176375473084755e-06, + "loss": 0.199, + "step": 51034 + }, + { + "epoch": 3.541391992228159, + "grad_norm": 0.8561211228370667, + "learning_rate": 6.815598634410814e-06, + "loss": 0.3192, + "step": 51035 + }, + { + "epoch": 3.541461383665256, + "grad_norm": 0.9321022629737854, + "learning_rate": 6.813560015685483e-06, + "loss": 0.2277, + "step": 51036 + }, + { + "epoch": 3.5415307751023524, + "grad_norm": 0.5780085325241089, + "learning_rate": 6.811521691138923e-06, + "loss": 0.1756, + "step": 51037 + }, + { + "epoch": 3.5416001665394488, + "grad_norm": 0.5926253795623779, + "learning_rate": 6.809483660777582e-06, + "loss": 0.1827, + "step": 51038 + }, + { + "epoch": 3.5416695579765456, + "grad_norm": 0.8138514161109924, + "learning_rate": 6.807445924607891e-06, + "loss": 0.2245, + "step": 51039 + }, + { + "epoch": 3.5417389494136424, + "grad_norm": 1.0807795524597168, + "learning_rate": 6.805408482636278e-06, + "loss": 0.2279, + "step": 51040 + }, + { + "epoch": 3.5418083408507393, + "grad_norm": 1.2032735347747803, + "learning_rate": 6.803371334869169e-06, + "loss": 0.2357, + "step": 51041 + }, + { + "epoch": 3.5418777322878356, + "grad_norm": 0.8805821537971497, + "learning_rate": 6.801334481313038e-06, + "loss": 0.2544, + "step": 51042 + }, + { + "epoch": 3.5419471237249325, + "grad_norm": 1.2345386743545532, + "learning_rate": 6.799297921974246e-06, + "loss": 0.3267, + "step": 51043 + }, + { + "epoch": 3.542016515162029, + "grad_norm": 0.8023751974105835, + "learning_rate": 6.797261656859267e-06, + "loss": 0.1944, + "step": 51044 + }, + { + "epoch": 3.5420859065991257, + "grad_norm": 1.0001652240753174, + "learning_rate": 6.795225685974526e-06, + "loss": 0.198, + "step": 51045 + }, + { + "epoch": 3.5421552980362225, + "grad_norm": 1.1876475811004639, + "learning_rate": 6.793190009326433e-06, + "loss": 0.3265, + "step": 51046 + }, + { + "epoch": 3.542224689473319, + "grad_norm": 0.78272944688797, + "learning_rate": 6.791154626921425e-06, + "loss": 0.2104, + "step": 51047 + }, + { + "epoch": 3.5422940809104158, + "grad_norm": 0.7791900038719177, + "learning_rate": 6.7891195387659536e-06, + "loss": 0.2217, + "step": 51048 + }, + { + "epoch": 3.542363472347512, + "grad_norm": 1.1568374633789062, + "learning_rate": 6.78708474486639e-06, + "loss": 0.3583, + "step": 51049 + }, + { + "epoch": 3.542432863784609, + "grad_norm": 0.9606889486312866, + "learning_rate": 6.785050245229196e-06, + "loss": 0.3449, + "step": 51050 + }, + { + "epoch": 3.542502255221706, + "grad_norm": 0.6932074427604675, + "learning_rate": 6.783016039860779e-06, + "loss": 0.1731, + "step": 51051 + }, + { + "epoch": 3.542571646658802, + "grad_norm": 0.882561206817627, + "learning_rate": 6.7809821287675765e-06, + "loss": 0.227, + "step": 51052 + }, + { + "epoch": 3.542641038095899, + "grad_norm": 1.0196245908737183, + "learning_rate": 6.778948511955985e-06, + "loss": 0.3101, + "step": 51053 + }, + { + "epoch": 3.5427104295329954, + "grad_norm": 1.218050241470337, + "learning_rate": 6.776915189432442e-06, + "loss": 0.1723, + "step": 51054 + }, + { + "epoch": 3.5427798209700923, + "grad_norm": 0.7315186262130737, + "learning_rate": 6.774882161203355e-06, + "loss": 0.1518, + "step": 51055 + }, + { + "epoch": 3.542849212407189, + "grad_norm": 0.7562776207923889, + "learning_rate": 6.772849427275163e-06, + "loss": 0.2501, + "step": 51056 + }, + { + "epoch": 3.5429186038442855, + "grad_norm": 1.0500123500823975, + "learning_rate": 6.7708169876542605e-06, + "loss": 0.24, + "step": 51057 + }, + { + "epoch": 3.5429879952813823, + "grad_norm": 0.6737921833992004, + "learning_rate": 6.768784842347075e-06, + "loss": 0.1905, + "step": 51058 + }, + { + "epoch": 3.5430573867184787, + "grad_norm": 0.7450748085975647, + "learning_rate": 6.766752991360015e-06, + "loss": 0.2349, + "step": 51059 + }, + { + "epoch": 3.5431267781555755, + "grad_norm": 0.7367643713951111, + "learning_rate": 6.7647214346994945e-06, + "loss": 0.2388, + "step": 51060 + }, + { + "epoch": 3.5431961695926724, + "grad_norm": 0.7644174098968506, + "learning_rate": 6.762690172371933e-06, + "loss": 0.1815, + "step": 51061 + }, + { + "epoch": 3.543265561029769, + "grad_norm": 0.889812171459198, + "learning_rate": 6.760659204383746e-06, + "loss": 0.2818, + "step": 51062 + }, + { + "epoch": 3.5433349524668656, + "grad_norm": 0.882264256477356, + "learning_rate": 6.758628530741351e-06, + "loss": 0.1972, + "step": 51063 + }, + { + "epoch": 3.5434043439039624, + "grad_norm": 0.6979143619537354, + "learning_rate": 6.756598151451133e-06, + "loss": 0.1968, + "step": 51064 + }, + { + "epoch": 3.543473735341059, + "grad_norm": 0.9444001913070679, + "learning_rate": 6.754568066519507e-06, + "loss": 0.2814, + "step": 51065 + }, + { + "epoch": 3.5435431267781556, + "grad_norm": 1.0736883878707886, + "learning_rate": 6.752538275952902e-06, + "loss": 0.3753, + "step": 51066 + }, + { + "epoch": 3.5436125182152525, + "grad_norm": 1.2394428253173828, + "learning_rate": 6.750508779757714e-06, + "loss": 0.2548, + "step": 51067 + }, + { + "epoch": 3.543681909652349, + "grad_norm": 1.2399524450302124, + "learning_rate": 6.748479577940348e-06, + "loss": 0.3396, + "step": 51068 + }, + { + "epoch": 3.5437513010894457, + "grad_norm": 1.0559957027435303, + "learning_rate": 6.746450670507221e-06, + "loss": 0.2411, + "step": 51069 + }, + { + "epoch": 3.543820692526542, + "grad_norm": 0.7820118069648743, + "learning_rate": 6.744422057464739e-06, + "loss": 0.2124, + "step": 51070 + }, + { + "epoch": 3.543890083963639, + "grad_norm": 0.6497828960418701, + "learning_rate": 6.742393738819286e-06, + "loss": 0.1729, + "step": 51071 + }, + { + "epoch": 3.5439594754007357, + "grad_norm": 0.9566515684127808, + "learning_rate": 6.740365714577268e-06, + "loss": 0.2626, + "step": 51072 + }, + { + "epoch": 3.544028866837832, + "grad_norm": 1.0916951894760132, + "learning_rate": 6.738337984745102e-06, + "loss": 0.2341, + "step": 51073 + }, + { + "epoch": 3.544098258274929, + "grad_norm": 0.9577705264091492, + "learning_rate": 6.7363105493291725e-06, + "loss": 0.2604, + "step": 51074 + }, + { + "epoch": 3.5441676497120254, + "grad_norm": 0.9423045516014099, + "learning_rate": 6.734283408335895e-06, + "loss": 0.2622, + "step": 51075 + }, + { + "epoch": 3.544237041149122, + "grad_norm": 0.9165242910385132, + "learning_rate": 6.732256561771666e-06, + "loss": 0.2271, + "step": 51076 + }, + { + "epoch": 3.544306432586219, + "grad_norm": 1.0050702095031738, + "learning_rate": 6.730230009642902e-06, + "loss": 0.3075, + "step": 51077 + }, + { + "epoch": 3.5443758240233154, + "grad_norm": 0.8920319080352783, + "learning_rate": 6.728203751955964e-06, + "loss": 0.2456, + "step": 51078 + }, + { + "epoch": 3.5444452154604122, + "grad_norm": 0.9936482906341553, + "learning_rate": 6.726177788717259e-06, + "loss": 0.2409, + "step": 51079 + }, + { + "epoch": 3.5445146068975086, + "grad_norm": 0.7271818518638611, + "learning_rate": 6.724152119933191e-06, + "loss": 0.1849, + "step": 51080 + }, + { + "epoch": 3.5445839983346055, + "grad_norm": 0.9646584987640381, + "learning_rate": 6.722126745610158e-06, + "loss": 0.352, + "step": 51081 + }, + { + "epoch": 3.5446533897717023, + "grad_norm": 1.0648314952850342, + "learning_rate": 6.720101665754541e-06, + "loss": 0.2103, + "step": 51082 + }, + { + "epoch": 3.5447227812087987, + "grad_norm": 0.8160833120346069, + "learning_rate": 6.718076880372759e-06, + "loss": 0.2269, + "step": 51083 + }, + { + "epoch": 3.5447921726458955, + "grad_norm": 0.8182950615882874, + "learning_rate": 6.716052389471172e-06, + "loss": 0.2774, + "step": 51084 + }, + { + "epoch": 3.544861564082992, + "grad_norm": 1.1871591806411743, + "learning_rate": 6.714028193056188e-06, + "loss": 0.4031, + "step": 51085 + }, + { + "epoch": 3.5449309555200887, + "grad_norm": 0.7847808599472046, + "learning_rate": 6.7120042911341885e-06, + "loss": 0.1881, + "step": 51086 + }, + { + "epoch": 3.5450003469571856, + "grad_norm": 0.7851496338844299, + "learning_rate": 6.709980683711581e-06, + "loss": 0.2793, + "step": 51087 + }, + { + "epoch": 3.5450697383942824, + "grad_norm": 0.645289957523346, + "learning_rate": 6.707957370794726e-06, + "loss": 0.1702, + "step": 51088 + }, + { + "epoch": 3.545139129831379, + "grad_norm": 0.7867923974990845, + "learning_rate": 6.705934352390031e-06, + "loss": 0.1844, + "step": 51089 + }, + { + "epoch": 3.5452085212684756, + "grad_norm": 0.7207759618759155, + "learning_rate": 6.7039116285038674e-06, + "loss": 0.2797, + "step": 51090 + }, + { + "epoch": 3.545277912705572, + "grad_norm": 1.3889409303665161, + "learning_rate": 6.701889199142653e-06, + "loss": 0.3345, + "step": 51091 + }, + { + "epoch": 3.545347304142669, + "grad_norm": 1.1660188436508179, + "learning_rate": 6.699867064312737e-06, + "loss": 0.326, + "step": 51092 + }, + { + "epoch": 3.5454166955797657, + "grad_norm": 1.495346188545227, + "learning_rate": 6.697845224020516e-06, + "loss": 0.2172, + "step": 51093 + }, + { + "epoch": 3.545486087016862, + "grad_norm": 1.0936973094940186, + "learning_rate": 6.695823678272373e-06, + "loss": 0.3284, + "step": 51094 + }, + { + "epoch": 3.545555478453959, + "grad_norm": 0.7738147974014282, + "learning_rate": 6.69380242707468e-06, + "loss": 0.2178, + "step": 51095 + }, + { + "epoch": 3.5456248698910553, + "grad_norm": 1.172074556350708, + "learning_rate": 6.691781470433833e-06, + "loss": 0.4106, + "step": 51096 + }, + { + "epoch": 3.545694261328152, + "grad_norm": 0.7322342991828918, + "learning_rate": 6.6897608083562265e-06, + "loss": 0.2365, + "step": 51097 + }, + { + "epoch": 3.545763652765249, + "grad_norm": 0.912869930267334, + "learning_rate": 6.687740440848189e-06, + "loss": 0.1809, + "step": 51098 + }, + { + "epoch": 3.5458330442023454, + "grad_norm": 0.8515197038650513, + "learning_rate": 6.685720367916148e-06, + "loss": 0.2246, + "step": 51099 + }, + { + "epoch": 3.545902435639442, + "grad_norm": 1.3086631298065186, + "learning_rate": 6.683700589566444e-06, + "loss": 0.2703, + "step": 51100 + }, + { + "epoch": 3.5459718270765386, + "grad_norm": 0.8720991015434265, + "learning_rate": 6.681681105805482e-06, + "loss": 0.1972, + "step": 51101 + }, + { + "epoch": 3.5460412185136354, + "grad_norm": 1.4054981470108032, + "learning_rate": 6.6796619166396235e-06, + "loss": 0.422, + "step": 51102 + }, + { + "epoch": 3.5461106099507322, + "grad_norm": 1.0082967281341553, + "learning_rate": 6.677643022075242e-06, + "loss": 0.3106, + "step": 51103 + }, + { + "epoch": 3.5461800013878286, + "grad_norm": 1.160941243171692, + "learning_rate": 6.675624422118731e-06, + "loss": 0.1907, + "step": 51104 + }, + { + "epoch": 3.5462493928249255, + "grad_norm": 0.8864461183547974, + "learning_rate": 6.673606116776432e-06, + "loss": 0.2438, + "step": 51105 + }, + { + "epoch": 3.546318784262022, + "grad_norm": 0.9067732095718384, + "learning_rate": 6.671588106054727e-06, + "loss": 0.204, + "step": 51106 + }, + { + "epoch": 3.5463881756991187, + "grad_norm": 0.9617617130279541, + "learning_rate": 6.6695703899600006e-06, + "loss": 0.2907, + "step": 51107 + }, + { + "epoch": 3.5464575671362155, + "grad_norm": 0.8057088255882263, + "learning_rate": 6.6675529684986025e-06, + "loss": 0.2574, + "step": 51108 + }, + { + "epoch": 3.546526958573312, + "grad_norm": 0.9746614694595337, + "learning_rate": 6.665535841676906e-06, + "loss": 0.2355, + "step": 51109 + }, + { + "epoch": 3.5465963500104087, + "grad_norm": 0.9986549615859985, + "learning_rate": 6.663519009501296e-06, + "loss": 0.2772, + "step": 51110 + }, + { + "epoch": 3.546665741447505, + "grad_norm": 0.8589446544647217, + "learning_rate": 6.6615024719781425e-06, + "loss": 0.1899, + "step": 51111 + }, + { + "epoch": 3.546735132884602, + "grad_norm": 0.9686750173568726, + "learning_rate": 6.659486229113776e-06, + "loss": 0.272, + "step": 51112 + }, + { + "epoch": 3.546804524321699, + "grad_norm": 1.0293341875076294, + "learning_rate": 6.6574702809145796e-06, + "loss": 0.3013, + "step": 51113 + }, + { + "epoch": 3.5468739157587956, + "grad_norm": 0.8786483407020569, + "learning_rate": 6.655454627386925e-06, + "loss": 0.2139, + "step": 51114 + }, + { + "epoch": 3.546943307195892, + "grad_norm": 0.6009624004364014, + "learning_rate": 6.653439268537165e-06, + "loss": 0.1361, + "step": 51115 + }, + { + "epoch": 3.547012698632989, + "grad_norm": 1.3556996583938599, + "learning_rate": 6.65142420437167e-06, + "loss": 0.3016, + "step": 51116 + }, + { + "epoch": 3.5470820900700852, + "grad_norm": 0.9814666509628296, + "learning_rate": 6.649409434896791e-06, + "loss": 0.3363, + "step": 51117 + }, + { + "epoch": 3.547151481507182, + "grad_norm": 0.6201812028884888, + "learning_rate": 6.647394960118925e-06, + "loss": 0.1522, + "step": 51118 + }, + { + "epoch": 3.547220872944279, + "grad_norm": 0.9584101438522339, + "learning_rate": 6.6453807800443745e-06, + "loss": 0.3403, + "step": 51119 + }, + { + "epoch": 3.5472902643813753, + "grad_norm": 0.8477770686149597, + "learning_rate": 6.643366894679537e-06, + "loss": 0.2215, + "step": 51120 + }, + { + "epoch": 3.547359655818472, + "grad_norm": 1.0640255212783813, + "learning_rate": 6.641353304030773e-06, + "loss": 0.3825, + "step": 51121 + }, + { + "epoch": 3.5474290472555685, + "grad_norm": 0.7294631600379944, + "learning_rate": 6.6393400081044e-06, + "loss": 0.2992, + "step": 51122 + }, + { + "epoch": 3.5474984386926653, + "grad_norm": 0.9055079221725464, + "learning_rate": 6.637327006906802e-06, + "loss": 0.366, + "step": 51123 + }, + { + "epoch": 3.547567830129762, + "grad_norm": 0.8408535718917847, + "learning_rate": 6.6353143004443285e-06, + "loss": 0.3449, + "step": 51124 + }, + { + "epoch": 3.5476372215668586, + "grad_norm": 0.9548054933547974, + "learning_rate": 6.633301888723364e-06, + "loss": 0.1634, + "step": 51125 + }, + { + "epoch": 3.5477066130039554, + "grad_norm": 0.8748893737792969, + "learning_rate": 6.631289771750204e-06, + "loss": 0.1822, + "step": 51126 + }, + { + "epoch": 3.547776004441052, + "grad_norm": 0.6464183330535889, + "learning_rate": 6.629277949531232e-06, + "loss": 0.203, + "step": 51127 + }, + { + "epoch": 3.5478453958781486, + "grad_norm": 0.7619118690490723, + "learning_rate": 6.627266422072798e-06, + "loss": 0.1336, + "step": 51128 + }, + { + "epoch": 3.5479147873152455, + "grad_norm": 0.709913432598114, + "learning_rate": 6.625255189381252e-06, + "loss": 0.2133, + "step": 51129 + }, + { + "epoch": 3.547984178752342, + "grad_norm": 1.1538339853286743, + "learning_rate": 6.623244251462946e-06, + "loss": 0.2501, + "step": 51130 + }, + { + "epoch": 3.5480535701894387, + "grad_norm": 1.192015290260315, + "learning_rate": 6.621233608324217e-06, + "loss": 0.3079, + "step": 51131 + }, + { + "epoch": 3.548122961626535, + "grad_norm": 0.9474421143531799, + "learning_rate": 6.619223259971441e-06, + "loss": 0.2602, + "step": 51132 + }, + { + "epoch": 3.548192353063632, + "grad_norm": 0.6564323902130127, + "learning_rate": 6.617213206410922e-06, + "loss": 0.1621, + "step": 51133 + }, + { + "epoch": 3.5482617445007287, + "grad_norm": 1.3229788541793823, + "learning_rate": 6.615203447649021e-06, + "loss": 0.4285, + "step": 51134 + }, + { + "epoch": 3.548331135937825, + "grad_norm": 0.8528203368186951, + "learning_rate": 6.613193983692101e-06, + "loss": 0.2385, + "step": 51135 + }, + { + "epoch": 3.548400527374922, + "grad_norm": 0.6554955244064331, + "learning_rate": 6.611184814546489e-06, + "loss": 0.2328, + "step": 51136 + }, + { + "epoch": 3.5484699188120183, + "grad_norm": 1.1340653896331787, + "learning_rate": 6.609175940218526e-06, + "loss": 0.3151, + "step": 51137 + }, + { + "epoch": 3.548539310249115, + "grad_norm": 0.932664155960083, + "learning_rate": 6.6071673607145614e-06, + "loss": 0.2775, + "step": 51138 + }, + { + "epoch": 3.548608701686212, + "grad_norm": 1.610221028327942, + "learning_rate": 6.605159076040946e-06, + "loss": 0.2426, + "step": 51139 + }, + { + "epoch": 3.548678093123309, + "grad_norm": 0.89173823595047, + "learning_rate": 6.6031510862039956e-06, + "loss": 0.2779, + "step": 51140 + }, + { + "epoch": 3.5487474845604052, + "grad_norm": 0.5788527131080627, + "learning_rate": 6.601143391210063e-06, + "loss": 0.1447, + "step": 51141 + }, + { + "epoch": 3.548816875997502, + "grad_norm": 0.7578595876693726, + "learning_rate": 6.599135991065486e-06, + "loss": 0.2726, + "step": 51142 + }, + { + "epoch": 3.5488862674345985, + "grad_norm": 1.0127512216567993, + "learning_rate": 6.597128885776604e-06, + "loss": 0.3572, + "step": 51143 + }, + { + "epoch": 3.5489556588716953, + "grad_norm": 0.9433327913284302, + "learning_rate": 6.595122075349736e-06, + "loss": 0.3123, + "step": 51144 + }, + { + "epoch": 3.549025050308792, + "grad_norm": 0.6667119264602661, + "learning_rate": 6.593115559791241e-06, + "loss": 0.184, + "step": 51145 + }, + { + "epoch": 3.5490944417458885, + "grad_norm": 0.8131645321846008, + "learning_rate": 6.591109339107459e-06, + "loss": 0.1873, + "step": 51146 + }, + { + "epoch": 3.5491638331829853, + "grad_norm": 1.0891048908233643, + "learning_rate": 6.589103413304687e-06, + "loss": 0.3905, + "step": 51147 + }, + { + "epoch": 3.5492332246200817, + "grad_norm": 1.0555775165557861, + "learning_rate": 6.587097782389273e-06, + "loss": 0.2156, + "step": 51148 + }, + { + "epoch": 3.5493026160571786, + "grad_norm": 1.0762666463851929, + "learning_rate": 6.585092446367558e-06, + "loss": 0.2539, + "step": 51149 + }, + { + "epoch": 3.5493720074942754, + "grad_norm": 1.1010180711746216, + "learning_rate": 6.583087405245869e-06, + "loss": 0.313, + "step": 51150 + }, + { + "epoch": 3.549441398931372, + "grad_norm": 1.0551905632019043, + "learning_rate": 6.581082659030535e-06, + "loss": 0.3242, + "step": 51151 + }, + { + "epoch": 3.5495107903684686, + "grad_norm": 0.8365760445594788, + "learning_rate": 6.579078207727885e-06, + "loss": 0.2499, + "step": 51152 + }, + { + "epoch": 3.549580181805565, + "grad_norm": 0.8261616826057434, + "learning_rate": 6.577074051344267e-06, + "loss": 0.2504, + "step": 51153 + }, + { + "epoch": 3.549649573242662, + "grad_norm": 0.857301652431488, + "learning_rate": 6.575070189885968e-06, + "loss": 0.2107, + "step": 51154 + }, + { + "epoch": 3.5497189646797587, + "grad_norm": 0.6375064253807068, + "learning_rate": 6.573066623359325e-06, + "loss": 0.1849, + "step": 51155 + }, + { + "epoch": 3.549788356116855, + "grad_norm": 1.3313437700271606, + "learning_rate": 6.571063351770701e-06, + "loss": 0.2958, + "step": 51156 + }, + { + "epoch": 3.549857747553952, + "grad_norm": 0.6553141474723816, + "learning_rate": 6.569060375126357e-06, + "loss": 0.2645, + "step": 51157 + }, + { + "epoch": 3.5499271389910483, + "grad_norm": 0.8941603302955627, + "learning_rate": 6.567057693432665e-06, + "loss": 0.3083, + "step": 51158 + }, + { + "epoch": 3.549996530428145, + "grad_norm": 0.8608345985412598, + "learning_rate": 6.565055306695922e-06, + "loss": 0.1797, + "step": 51159 + }, + { + "epoch": 3.550065921865242, + "grad_norm": 0.9772217273712158, + "learning_rate": 6.563053214922476e-06, + "loss": 0.1668, + "step": 51160 + }, + { + "epoch": 3.5501353133023383, + "grad_norm": 1.280641794204712, + "learning_rate": 6.561051418118625e-06, + "loss": 0.2891, + "step": 51161 + }, + { + "epoch": 3.550204704739435, + "grad_norm": 0.8837267160415649, + "learning_rate": 6.559049916290683e-06, + "loss": 0.2898, + "step": 51162 + }, + { + "epoch": 3.5502740961765316, + "grad_norm": 0.90700364112854, + "learning_rate": 6.55704870944498e-06, + "loss": 0.2137, + "step": 51163 + }, + { + "epoch": 3.5503434876136284, + "grad_norm": 0.9803888201713562, + "learning_rate": 6.555047797587832e-06, + "loss": 0.3135, + "step": 51164 + }, + { + "epoch": 3.5504128790507252, + "grad_norm": 0.9867779016494751, + "learning_rate": 6.553047180725558e-06, + "loss": 0.3419, + "step": 51165 + }, + { + "epoch": 3.550482270487822, + "grad_norm": 1.0261095762252808, + "learning_rate": 6.551046858864473e-06, + "loss": 0.2195, + "step": 51166 + }, + { + "epoch": 3.5505516619249184, + "grad_norm": 0.6490401029586792, + "learning_rate": 6.549046832010908e-06, + "loss": 0.1664, + "step": 51167 + }, + { + "epoch": 3.5506210533620153, + "grad_norm": 0.972137451171875, + "learning_rate": 6.547047100171144e-06, + "loss": 0.2505, + "step": 51168 + }, + { + "epoch": 3.5506904447991117, + "grad_norm": 1.4225813150405884, + "learning_rate": 6.545047663351511e-06, + "loss": 0.2812, + "step": 51169 + }, + { + "epoch": 3.5507598362362085, + "grad_norm": 1.0797520875930786, + "learning_rate": 6.543048521558315e-06, + "loss": 0.3227, + "step": 51170 + }, + { + "epoch": 3.5508292276733053, + "grad_norm": 0.7870006561279297, + "learning_rate": 6.541049674797872e-06, + "loss": 0.3128, + "step": 51171 + }, + { + "epoch": 3.5508986191104017, + "grad_norm": 0.9706848859786987, + "learning_rate": 6.5390511230765e-06, + "loss": 0.2992, + "step": 51172 + }, + { + "epoch": 3.5509680105474986, + "grad_norm": 0.7769346237182617, + "learning_rate": 6.537052866400506e-06, + "loss": 0.2395, + "step": 51173 + }, + { + "epoch": 3.551037401984595, + "grad_norm": 0.8008509874343872, + "learning_rate": 6.535054904776183e-06, + "loss": 0.2531, + "step": 51174 + }, + { + "epoch": 3.5511067934216918, + "grad_norm": 1.046849250793457, + "learning_rate": 6.53305723820985e-06, + "loss": 0.2969, + "step": 51175 + }, + { + "epoch": 3.5511761848587886, + "grad_norm": 1.0746339559555054, + "learning_rate": 6.5310598667078e-06, + "loss": 0.3344, + "step": 51176 + }, + { + "epoch": 3.551245576295885, + "grad_norm": 0.9563846588134766, + "learning_rate": 6.529062790276364e-06, + "loss": 0.2236, + "step": 51177 + }, + { + "epoch": 3.551314967732982, + "grad_norm": 0.8775118589401245, + "learning_rate": 6.527066008921823e-06, + "loss": 0.225, + "step": 51178 + }, + { + "epoch": 3.551384359170078, + "grad_norm": 0.8419276475906372, + "learning_rate": 6.525069522650485e-06, + "loss": 0.2629, + "step": 51179 + }, + { + "epoch": 3.551453750607175, + "grad_norm": 0.8925566077232361, + "learning_rate": 6.523073331468676e-06, + "loss": 0.2233, + "step": 51180 + }, + { + "epoch": 3.551523142044272, + "grad_norm": 0.6386193037033081, + "learning_rate": 6.521077435382672e-06, + "loss": 0.2773, + "step": 51181 + }, + { + "epoch": 3.5515925334813683, + "grad_norm": 0.9569957256317139, + "learning_rate": 6.519081834398777e-06, + "loss": 0.263, + "step": 51182 + }, + { + "epoch": 3.551661924918465, + "grad_norm": 0.8116995096206665, + "learning_rate": 6.517086528523286e-06, + "loss": 0.2516, + "step": 51183 + }, + { + "epoch": 3.5517313163555615, + "grad_norm": 0.8295696377754211, + "learning_rate": 6.515091517762517e-06, + "loss": 0.2171, + "step": 51184 + }, + { + "epoch": 3.5518007077926583, + "grad_norm": 1.386395812034607, + "learning_rate": 6.513096802122753e-06, + "loss": 0.3281, + "step": 51185 + }, + { + "epoch": 3.551870099229755, + "grad_norm": 0.8212913274765015, + "learning_rate": 6.511102381610301e-06, + "loss": 0.2936, + "step": 51186 + }, + { + "epoch": 3.5519394906668516, + "grad_norm": 1.0400480031967163, + "learning_rate": 6.5091082562314665e-06, + "loss": 0.3042, + "step": 51187 + }, + { + "epoch": 3.5520088821039484, + "grad_norm": 0.8645501136779785, + "learning_rate": 6.507114425992511e-06, + "loss": 0.2268, + "step": 51188 + }, + { + "epoch": 3.5520782735410448, + "grad_norm": 0.6314453482627869, + "learning_rate": 6.505120890899752e-06, + "loss": 0.169, + "step": 51189 + }, + { + "epoch": 3.5521476649781416, + "grad_norm": 0.7400316596031189, + "learning_rate": 6.503127650959473e-06, + "loss": 0.1952, + "step": 51190 + }, + { + "epoch": 3.5522170564152384, + "grad_norm": 1.1332203149795532, + "learning_rate": 6.501134706177991e-06, + "loss": 0.2783, + "step": 51191 + }, + { + "epoch": 3.5522864478523353, + "grad_norm": 1.1246020793914795, + "learning_rate": 6.499142056561569e-06, + "loss": 0.1493, + "step": 51192 + }, + { + "epoch": 3.5523558392894317, + "grad_norm": 1.1227083206176758, + "learning_rate": 6.497149702116501e-06, + "loss": 0.2549, + "step": 51193 + }, + { + "epoch": 3.5524252307265285, + "grad_norm": 0.99555504322052, + "learning_rate": 6.4951576428490815e-06, + "loss": 0.2466, + "step": 51194 + }, + { + "epoch": 3.552494622163625, + "grad_norm": 1.1400761604309082, + "learning_rate": 6.493165878765628e-06, + "loss": 0.2913, + "step": 51195 + }, + { + "epoch": 3.5525640136007217, + "grad_norm": 0.9300441741943359, + "learning_rate": 6.491174409872369e-06, + "loss": 0.2516, + "step": 51196 + }, + { + "epoch": 3.5526334050378185, + "grad_norm": 0.7088517546653748, + "learning_rate": 6.489183236175633e-06, + "loss": 0.181, + "step": 51197 + }, + { + "epoch": 3.552702796474915, + "grad_norm": 0.775537371635437, + "learning_rate": 6.4871923576816926e-06, + "loss": 0.2114, + "step": 51198 + }, + { + "epoch": 3.5527721879120118, + "grad_norm": 1.1521201133728027, + "learning_rate": 6.485201774396843e-06, + "loss": 0.2917, + "step": 51199 + }, + { + "epoch": 3.552841579349108, + "grad_norm": 2.1792285442352295, + "learning_rate": 6.483211486327357e-06, + "loss": 0.218, + "step": 51200 + }, + { + "epoch": 3.552910970786205, + "grad_norm": 0.8598959445953369, + "learning_rate": 6.481221493479539e-06, + "loss": 0.2691, + "step": 51201 + }, + { + "epoch": 3.552980362223302, + "grad_norm": 1.139319658279419, + "learning_rate": 6.479231795859642e-06, + "loss": 0.2814, + "step": 51202 + }, + { + "epoch": 3.553049753660398, + "grad_norm": 0.8415973782539368, + "learning_rate": 6.4772423934739475e-06, + "loss": 0.3003, + "step": 51203 + }, + { + "epoch": 3.553119145097495, + "grad_norm": 1.8258134126663208, + "learning_rate": 6.475253286328764e-06, + "loss": 0.3388, + "step": 51204 + }, + { + "epoch": 3.5531885365345914, + "grad_norm": 1.0029135942459106, + "learning_rate": 6.47326447443034e-06, + "loss": 0.3476, + "step": 51205 + }, + { + "epoch": 3.5532579279716883, + "grad_norm": 0.7835546731948853, + "learning_rate": 6.471275957784984e-06, + "loss": 0.2813, + "step": 51206 + }, + { + "epoch": 3.553327319408785, + "grad_norm": 1.4111729860305786, + "learning_rate": 6.469287736398944e-06, + "loss": 0.2932, + "step": 51207 + }, + { + "epoch": 3.5533967108458815, + "grad_norm": 0.8801189661026001, + "learning_rate": 6.467299810278537e-06, + "loss": 0.2517, + "step": 51208 + }, + { + "epoch": 3.5534661022829783, + "grad_norm": 1.0851620435714722, + "learning_rate": 6.4653121794299946e-06, + "loss": 0.3404, + "step": 51209 + }, + { + "epoch": 3.5535354937200747, + "grad_norm": 0.7679246068000793, + "learning_rate": 6.463324843859608e-06, + "loss": 0.2305, + "step": 51210 + }, + { + "epoch": 3.5536048851571715, + "grad_norm": 0.7776464223861694, + "learning_rate": 6.4613378035736525e-06, + "loss": 0.2745, + "step": 51211 + }, + { + "epoch": 3.5536742765942684, + "grad_norm": 0.6475105285644531, + "learning_rate": 6.4593510585784e-06, + "loss": 0.1615, + "step": 51212 + }, + { + "epoch": 3.5537436680313648, + "grad_norm": 0.931357741355896, + "learning_rate": 6.457364608880123e-06, + "loss": 0.1945, + "step": 51213 + }, + { + "epoch": 3.5538130594684616, + "grad_norm": 1.3096429109573364, + "learning_rate": 6.455378454485095e-06, + "loss": 0.3338, + "step": 51214 + }, + { + "epoch": 3.553882450905558, + "grad_norm": 0.9191434979438782, + "learning_rate": 6.453392595399599e-06, + "loss": 0.2206, + "step": 51215 + }, + { + "epoch": 3.553951842342655, + "grad_norm": 0.9246770143508911, + "learning_rate": 6.451407031629874e-06, + "loss": 0.2559, + "step": 51216 + }, + { + "epoch": 3.5540212337797517, + "grad_norm": 0.8071944117546082, + "learning_rate": 6.449421763182206e-06, + "loss": 0.3965, + "step": 51217 + }, + { + "epoch": 3.5540906252168485, + "grad_norm": 1.2658737897872925, + "learning_rate": 6.447436790062855e-06, + "loss": 0.2937, + "step": 51218 + }, + { + "epoch": 3.554160016653945, + "grad_norm": 0.9148926734924316, + "learning_rate": 6.445452112278094e-06, + "loss": 0.2684, + "step": 51219 + }, + { + "epoch": 3.5542294080910417, + "grad_norm": 1.132169246673584, + "learning_rate": 6.443467729834185e-06, + "loss": 0.465, + "step": 51220 + }, + { + "epoch": 3.554298799528138, + "grad_norm": 1.1324416399002075, + "learning_rate": 6.4414836427374004e-06, + "loss": 0.2607, + "step": 51221 + }, + { + "epoch": 3.554368190965235, + "grad_norm": 0.8039971590042114, + "learning_rate": 6.439499850994013e-06, + "loss": 0.3066, + "step": 51222 + }, + { + "epoch": 3.5544375824023318, + "grad_norm": 0.7413690686225891, + "learning_rate": 6.43751635461024e-06, + "loss": 0.2058, + "step": 51223 + }, + { + "epoch": 3.554506973839428, + "grad_norm": 0.9352670907974243, + "learning_rate": 6.435533153592388e-06, + "loss": 0.217, + "step": 51224 + }, + { + "epoch": 3.554576365276525, + "grad_norm": 0.7636247277259827, + "learning_rate": 6.433550247946696e-06, + "loss": 0.3272, + "step": 51225 + }, + { + "epoch": 3.5546457567136214, + "grad_norm": 1.2564451694488525, + "learning_rate": 6.431567637679436e-06, + "loss": 0.2477, + "step": 51226 + }, + { + "epoch": 3.554715148150718, + "grad_norm": 0.7744031548500061, + "learning_rate": 6.429585322796861e-06, + "loss": 0.1775, + "step": 51227 + }, + { + "epoch": 3.554784539587815, + "grad_norm": 1.9200745820999146, + "learning_rate": 6.427603303305219e-06, + "loss": 0.2527, + "step": 51228 + }, + { + "epoch": 3.5548539310249114, + "grad_norm": 1.4025565385818481, + "learning_rate": 6.425621579210783e-06, + "loss": 0.3856, + "step": 51229 + }, + { + "epoch": 3.5549233224620083, + "grad_norm": 0.8683961629867554, + "learning_rate": 6.4236401505198165e-06, + "loss": 0.2513, + "step": 51230 + }, + { + "epoch": 3.5549927138991047, + "grad_norm": 0.7463911175727844, + "learning_rate": 6.421659017238546e-06, + "loss": 0.1966, + "step": 51231 + }, + { + "epoch": 3.5550621053362015, + "grad_norm": 0.9993826746940613, + "learning_rate": 6.419678179373245e-06, + "loss": 0.2596, + "step": 51232 + }, + { + "epoch": 3.5551314967732983, + "grad_norm": 0.5709076523780823, + "learning_rate": 6.417697636930153e-06, + "loss": 0.171, + "step": 51233 + }, + { + "epoch": 3.5552008882103947, + "grad_norm": 0.7576900124549866, + "learning_rate": 6.415717389915543e-06, + "loss": 0.3441, + "step": 51234 + }, + { + "epoch": 3.5552702796474915, + "grad_norm": 0.8561438322067261, + "learning_rate": 6.413737438335643e-06, + "loss": 0.2671, + "step": 51235 + }, + { + "epoch": 3.555339671084588, + "grad_norm": 1.0695998668670654, + "learning_rate": 6.411757782196737e-06, + "loss": 0.314, + "step": 51236 + }, + { + "epoch": 3.5554090625216848, + "grad_norm": 1.032643437385559, + "learning_rate": 6.409778421505041e-06, + "loss": 0.2791, + "step": 51237 + }, + { + "epoch": 3.5554784539587816, + "grad_norm": 0.9165934324264526, + "learning_rate": 6.407799356266819e-06, + "loss": 0.228, + "step": 51238 + }, + { + "epoch": 3.5555478453958784, + "grad_norm": 1.8955016136169434, + "learning_rate": 6.405820586488309e-06, + "loss": 0.3063, + "step": 51239 + }, + { + "epoch": 3.555617236832975, + "grad_norm": 0.6670595407485962, + "learning_rate": 6.403842112175762e-06, + "loss": 0.2085, + "step": 51240 + }, + { + "epoch": 3.555686628270071, + "grad_norm": 1.2656118869781494, + "learning_rate": 6.401863933335428e-06, + "loss": 0.2003, + "step": 51241 + }, + { + "epoch": 3.555756019707168, + "grad_norm": 0.939402163028717, + "learning_rate": 6.399886049973558e-06, + "loss": 0.3354, + "step": 51242 + }, + { + "epoch": 3.555825411144265, + "grad_norm": 0.7378374338150024, + "learning_rate": 6.397908462096391e-06, + "loss": 0.157, + "step": 51243 + }, + { + "epoch": 3.5558948025813617, + "grad_norm": 0.9380426406860352, + "learning_rate": 6.395931169710168e-06, + "loss": 0.3361, + "step": 51244 + }, + { + "epoch": 3.555964194018458, + "grad_norm": 0.6546607613563538, + "learning_rate": 6.393954172821115e-06, + "loss": 0.2068, + "step": 51245 + }, + { + "epoch": 3.556033585455555, + "grad_norm": 2.0166678428649902, + "learning_rate": 6.391977471435495e-06, + "loss": 0.309, + "step": 51246 + }, + { + "epoch": 3.5561029768926513, + "grad_norm": 0.8998991250991821, + "learning_rate": 6.390001065559537e-06, + "loss": 0.2896, + "step": 51247 + }, + { + "epoch": 3.556172368329748, + "grad_norm": 0.717723548412323, + "learning_rate": 6.3880249551994895e-06, + "loss": 0.2145, + "step": 51248 + }, + { + "epoch": 3.556241759766845, + "grad_norm": 1.1490516662597656, + "learning_rate": 6.3860491403616054e-06, + "loss": 0.1772, + "step": 51249 + }, + { + "epoch": 3.5563111512039414, + "grad_norm": 1.2440227270126343, + "learning_rate": 6.384073621052078e-06, + "loss": 0.251, + "step": 51250 + }, + { + "epoch": 3.556380542641038, + "grad_norm": 0.8549318909645081, + "learning_rate": 6.382098397277181e-06, + "loss": 0.344, + "step": 51251 + }, + { + "epoch": 3.5564499340781346, + "grad_norm": 0.9629436135292053, + "learning_rate": 6.380123469043131e-06, + "loss": 0.1845, + "step": 51252 + }, + { + "epoch": 3.5565193255152314, + "grad_norm": 0.8187049627304077, + "learning_rate": 6.378148836356168e-06, + "loss": 0.1848, + "step": 51253 + }, + { + "epoch": 3.5565887169523283, + "grad_norm": 1.0623438358306885, + "learning_rate": 6.376174499222521e-06, + "loss": 0.3279, + "step": 51254 + }, + { + "epoch": 3.5566581083894246, + "grad_norm": 0.9279444813728333, + "learning_rate": 6.374200457648439e-06, + "loss": 0.3296, + "step": 51255 + }, + { + "epoch": 3.5567274998265215, + "grad_norm": 0.782364547252655, + "learning_rate": 6.37222671164015e-06, + "loss": 0.2185, + "step": 51256 + }, + { + "epoch": 3.556796891263618, + "grad_norm": 1.4526036977767944, + "learning_rate": 6.370253261203874e-06, + "loss": 0.3173, + "step": 51257 + }, + { + "epoch": 3.5568662827007147, + "grad_norm": 0.995008111000061, + "learning_rate": 6.368280106345836e-06, + "loss": 0.2929, + "step": 51258 + }, + { + "epoch": 3.5569356741378115, + "grad_norm": 0.8881266713142395, + "learning_rate": 6.366307247072278e-06, + "loss": 0.1604, + "step": 51259 + }, + { + "epoch": 3.557005065574908, + "grad_norm": 1.0470815896987915, + "learning_rate": 6.3643346833894275e-06, + "loss": 0.2481, + "step": 51260 + }, + { + "epoch": 3.5570744570120048, + "grad_norm": 0.9702434539794922, + "learning_rate": 6.362362415303524e-06, + "loss": 0.372, + "step": 51261 + }, + { + "epoch": 3.557143848449101, + "grad_norm": 2.833861827850342, + "learning_rate": 6.360390442820763e-06, + "loss": 0.3474, + "step": 51262 + }, + { + "epoch": 3.557213239886198, + "grad_norm": 1.0034531354904175, + "learning_rate": 6.358418765947383e-06, + "loss": 0.4128, + "step": 51263 + }, + { + "epoch": 3.557282631323295, + "grad_norm": 1.2898451089859009, + "learning_rate": 6.3564473846896235e-06, + "loss": 0.2451, + "step": 51264 + }, + { + "epoch": 3.5573520227603916, + "grad_norm": 1.3466472625732422, + "learning_rate": 6.35447629905368e-06, + "loss": 0.3636, + "step": 51265 + }, + { + "epoch": 3.557421414197488, + "grad_norm": 1.0001766681671143, + "learning_rate": 6.352505509045792e-06, + "loss": 0.4269, + "step": 51266 + }, + { + "epoch": 3.5574908056345844, + "grad_norm": 0.7672621011734009, + "learning_rate": 6.350535014672176e-06, + "loss": 0.2247, + "step": 51267 + }, + { + "epoch": 3.5575601970716813, + "grad_norm": 0.923778772354126, + "learning_rate": 6.348564815939062e-06, + "loss": 0.2287, + "step": 51268 + }, + { + "epoch": 3.557629588508778, + "grad_norm": 0.8306114077568054, + "learning_rate": 6.346594912852654e-06, + "loss": 0.3564, + "step": 51269 + }, + { + "epoch": 3.557698979945875, + "grad_norm": 1.0937379598617554, + "learning_rate": 6.344625305419205e-06, + "loss": 0.2882, + "step": 51270 + }, + { + "epoch": 3.5577683713829713, + "grad_norm": 0.7120669484138489, + "learning_rate": 6.342655993644886e-06, + "loss": 0.2768, + "step": 51271 + }, + { + "epoch": 3.557837762820068, + "grad_norm": 0.7338957786560059, + "learning_rate": 6.340686977535937e-06, + "loss": 0.2487, + "step": 51272 + }, + { + "epoch": 3.5579071542571645, + "grad_norm": 0.713874876499176, + "learning_rate": 6.338718257098575e-06, + "loss": 0.2395, + "step": 51273 + }, + { + "epoch": 3.5579765456942614, + "grad_norm": 0.7781686782836914, + "learning_rate": 6.3367498323390065e-06, + "loss": 0.1341, + "step": 51274 + }, + { + "epoch": 3.558045937131358, + "grad_norm": 1.8667553663253784, + "learning_rate": 6.3347817032634596e-06, + "loss": 0.4309, + "step": 51275 + }, + { + "epoch": 3.5581153285684546, + "grad_norm": 0.6072824597358704, + "learning_rate": 6.33281386987814e-06, + "loss": 0.2176, + "step": 51276 + }, + { + "epoch": 3.5581847200055514, + "grad_norm": 1.255595088005066, + "learning_rate": 6.3308463321892665e-06, + "loss": 0.3909, + "step": 51277 + }, + { + "epoch": 3.558254111442648, + "grad_norm": 0.8101757168769836, + "learning_rate": 6.328879090203033e-06, + "loss": 0.1981, + "step": 51278 + }, + { + "epoch": 3.5583235028797446, + "grad_norm": 1.3291881084442139, + "learning_rate": 6.326912143925667e-06, + "loss": 0.4474, + "step": 51279 + }, + { + "epoch": 3.5583928943168415, + "grad_norm": 1.0003957748413086, + "learning_rate": 6.324945493363365e-06, + "loss": 0.2956, + "step": 51280 + }, + { + "epoch": 3.558462285753938, + "grad_norm": 0.9163344502449036, + "learning_rate": 6.322979138522334e-06, + "loss": 0.2107, + "step": 51281 + }, + { + "epoch": 3.5585316771910347, + "grad_norm": 0.6794609427452087, + "learning_rate": 6.3210130794088e-06, + "loss": 0.2286, + "step": 51282 + }, + { + "epoch": 3.558601068628131, + "grad_norm": 0.7038641571998596, + "learning_rate": 6.3190473160289585e-06, + "loss": 0.1799, + "step": 51283 + }, + { + "epoch": 3.558670460065228, + "grad_norm": 0.7907413840293884, + "learning_rate": 6.317081848389028e-06, + "loss": 0.2141, + "step": 51284 + }, + { + "epoch": 3.5587398515023247, + "grad_norm": 0.7917993068695068, + "learning_rate": 6.315116676495181e-06, + "loss": 0.182, + "step": 51285 + }, + { + "epoch": 3.558809242939421, + "grad_norm": 1.0760494470596313, + "learning_rate": 6.313151800353656e-06, + "loss": 0.1998, + "step": 51286 + }, + { + "epoch": 3.558878634376518, + "grad_norm": 0.8332764506340027, + "learning_rate": 6.311187219970627e-06, + "loss": 0.283, + "step": 51287 + }, + { + "epoch": 3.5589480258136144, + "grad_norm": 0.8407773375511169, + "learning_rate": 6.30922293535231e-06, + "loss": 0.3658, + "step": 51288 + }, + { + "epoch": 3.559017417250711, + "grad_norm": 0.8399393558502197, + "learning_rate": 6.307258946504913e-06, + "loss": 0.1853, + "step": 51289 + }, + { + "epoch": 3.559086808687808, + "grad_norm": 0.8990064859390259, + "learning_rate": 6.305295253434629e-06, + "loss": 0.1701, + "step": 51290 + }, + { + "epoch": 3.559156200124905, + "grad_norm": 0.6627094149589539, + "learning_rate": 6.303331856147665e-06, + "loss": 0.1658, + "step": 51291 + }, + { + "epoch": 3.5592255915620012, + "grad_norm": 1.0563303232192993, + "learning_rate": 6.301368754650206e-06, + "loss": 0.1817, + "step": 51292 + }, + { + "epoch": 3.5592949829990976, + "grad_norm": 1.2160953283309937, + "learning_rate": 6.299405948948445e-06, + "loss": 0.361, + "step": 51293 + }, + { + "epoch": 3.5593643744361945, + "grad_norm": 0.7977386116981506, + "learning_rate": 6.2974434390486e-06, + "loss": 0.2114, + "step": 51294 + }, + { + "epoch": 3.5594337658732913, + "grad_norm": 0.5858573913574219, + "learning_rate": 6.2954812249568455e-06, + "loss": 0.1205, + "step": 51295 + }, + { + "epoch": 3.559503157310388, + "grad_norm": 1.105329155921936, + "learning_rate": 6.293519306679407e-06, + "loss": 0.3055, + "step": 51296 + }, + { + "epoch": 3.5595725487474845, + "grad_norm": 1.0177372694015503, + "learning_rate": 6.291557684222427e-06, + "loss": 0.319, + "step": 51297 + }, + { + "epoch": 3.5596419401845814, + "grad_norm": 0.6737303733825684, + "learning_rate": 6.289596357592142e-06, + "loss": 0.2615, + "step": 51298 + }, + { + "epoch": 3.5597113316216777, + "grad_norm": 0.7764929533004761, + "learning_rate": 6.287635326794738e-06, + "loss": 0.2055, + "step": 51299 + }, + { + "epoch": 3.5597807230587746, + "grad_norm": 0.9190666079521179, + "learning_rate": 6.285674591836388e-06, + "loss": 0.1549, + "step": 51300 + }, + { + "epoch": 3.5598501144958714, + "grad_norm": 0.9295719265937805, + "learning_rate": 6.283714152723275e-06, + "loss": 0.2468, + "step": 51301 + }, + { + "epoch": 3.559919505932968, + "grad_norm": 0.8650051951408386, + "learning_rate": 6.281754009461615e-06, + "loss": 0.3341, + "step": 51302 + }, + { + "epoch": 3.5599888973700646, + "grad_norm": 1.7939064502716064, + "learning_rate": 6.279794162057584e-06, + "loss": 0.3727, + "step": 51303 + }, + { + "epoch": 3.560058288807161, + "grad_norm": 0.9917320609092712, + "learning_rate": 6.277834610517364e-06, + "loss": 0.2677, + "step": 51304 + }, + { + "epoch": 3.560127680244258, + "grad_norm": 1.012137770652771, + "learning_rate": 6.2758753548471715e-06, + "loss": 0.2636, + "step": 51305 + }, + { + "epoch": 3.5601970716813547, + "grad_norm": 0.9967319369316101, + "learning_rate": 6.2739163950531476e-06, + "loss": 0.2223, + "step": 51306 + }, + { + "epoch": 3.560266463118451, + "grad_norm": 1.1236480474472046, + "learning_rate": 6.271957731141487e-06, + "loss": 0.2698, + "step": 51307 + }, + { + "epoch": 3.560335854555548, + "grad_norm": 1.1014033555984497, + "learning_rate": 6.269999363118384e-06, + "loss": 0.2974, + "step": 51308 + }, + { + "epoch": 3.5604052459926443, + "grad_norm": 0.6952957510948181, + "learning_rate": 6.2680412909900236e-06, + "loss": 0.1919, + "step": 51309 + }, + { + "epoch": 3.560474637429741, + "grad_norm": 0.7017853856086731, + "learning_rate": 6.266083514762577e-06, + "loss": 0.2389, + "step": 51310 + }, + { + "epoch": 3.560544028866838, + "grad_norm": 0.9875441193580627, + "learning_rate": 6.26412603444223e-06, + "loss": 0.1653, + "step": 51311 + }, + { + "epoch": 3.5606134203039344, + "grad_norm": 0.7462546825408936, + "learning_rate": 6.2621688500351885e-06, + "loss": 0.1515, + "step": 51312 + }, + { + "epoch": 3.560682811741031, + "grad_norm": 0.9506749510765076, + "learning_rate": 6.26021196154758e-06, + "loss": 0.1802, + "step": 51313 + }, + { + "epoch": 3.5607522031781276, + "grad_norm": 0.9459896683692932, + "learning_rate": 6.2582553689856e-06, + "loss": 0.21, + "step": 51314 + }, + { + "epoch": 3.5608215946152244, + "grad_norm": 1.2462910413742065, + "learning_rate": 6.256299072355443e-06, + "loss": 0.3461, + "step": 51315 + }, + { + "epoch": 3.5608909860523212, + "grad_norm": 0.8022552728652954, + "learning_rate": 6.254343071663271e-06, + "loss": 0.1375, + "step": 51316 + }, + { + "epoch": 3.560960377489418, + "grad_norm": 0.9297425746917725, + "learning_rate": 6.252387366915258e-06, + "loss": 0.2145, + "step": 51317 + }, + { + "epoch": 3.5610297689265145, + "grad_norm": 0.9091915488243103, + "learning_rate": 6.250431958117575e-06, + "loss": 0.2294, + "step": 51318 + }, + { + "epoch": 3.5610991603636113, + "grad_norm": 0.8923014402389526, + "learning_rate": 6.248476845276429e-06, + "loss": 0.2097, + "step": 51319 + }, + { + "epoch": 3.5611685518007077, + "grad_norm": 1.0254731178283691, + "learning_rate": 6.246522028397939e-06, + "loss": 0.2619, + "step": 51320 + }, + { + "epoch": 3.5612379432378045, + "grad_norm": 0.8310444355010986, + "learning_rate": 6.244567507488308e-06, + "loss": 0.2574, + "step": 51321 + }, + { + "epoch": 3.5613073346749013, + "grad_norm": 0.9093831181526184, + "learning_rate": 6.2426132825537e-06, + "loss": 0.2938, + "step": 51322 + }, + { + "epoch": 3.5613767261119977, + "grad_norm": 0.6374697089195251, + "learning_rate": 6.240659353600287e-06, + "loss": 0.1092, + "step": 51323 + }, + { + "epoch": 3.5614461175490946, + "grad_norm": 0.8671587109565735, + "learning_rate": 6.23870572063423e-06, + "loss": 0.2264, + "step": 51324 + }, + { + "epoch": 3.561515508986191, + "grad_norm": 0.855375349521637, + "learning_rate": 6.236752383661715e-06, + "loss": 0.3998, + "step": 51325 + }, + { + "epoch": 3.561584900423288, + "grad_norm": 0.5873433947563171, + "learning_rate": 6.2347993426888796e-06, + "loss": 0.2498, + "step": 51326 + }, + { + "epoch": 3.5616542918603846, + "grad_norm": 0.783440351486206, + "learning_rate": 6.232846597721898e-06, + "loss": 0.176, + "step": 51327 + }, + { + "epoch": 3.561723683297481, + "grad_norm": 1.3137935400009155, + "learning_rate": 6.230894148766952e-06, + "loss": 0.2587, + "step": 51328 + }, + { + "epoch": 3.561793074734578, + "grad_norm": 1.268290638923645, + "learning_rate": 6.228941995830184e-06, + "loss": 0.3107, + "step": 51329 + }, + { + "epoch": 3.5618624661716742, + "grad_norm": 0.9876279830932617, + "learning_rate": 6.2269901389177766e-06, + "loss": 0.241, + "step": 51330 + }, + { + "epoch": 3.561931857608771, + "grad_norm": 1.2393946647644043, + "learning_rate": 6.225038578035891e-06, + "loss": 0.2885, + "step": 51331 + }, + { + "epoch": 3.562001249045868, + "grad_norm": 1.021032452583313, + "learning_rate": 6.223087313190657e-06, + "loss": 0.3312, + "step": 51332 + }, + { + "epoch": 3.5620706404829643, + "grad_norm": 0.8049306273460388, + "learning_rate": 6.221136344388267e-06, + "loss": 0.2628, + "step": 51333 + }, + { + "epoch": 3.562140031920061, + "grad_norm": 0.878865659236908, + "learning_rate": 6.219185671634875e-06, + "loss": 0.2053, + "step": 51334 + }, + { + "epoch": 3.5622094233571575, + "grad_norm": 0.95111083984375, + "learning_rate": 6.217235294936629e-06, + "loss": 0.2566, + "step": 51335 + }, + { + "epoch": 3.5622788147942543, + "grad_norm": 1.2164453268051147, + "learning_rate": 6.215285214299682e-06, + "loss": 0.3781, + "step": 51336 + }, + { + "epoch": 3.562348206231351, + "grad_norm": 0.9486513733863831, + "learning_rate": 6.213335429730205e-06, + "loss": 0.2588, + "step": 51337 + }, + { + "epoch": 3.5624175976684476, + "grad_norm": 0.8111321330070496, + "learning_rate": 6.211385941234338e-06, + "loss": 0.1705, + "step": 51338 + }, + { + "epoch": 3.5624869891055444, + "grad_norm": 1.3581886291503906, + "learning_rate": 6.2094367488182644e-06, + "loss": 0.276, + "step": 51339 + }, + { + "epoch": 3.562556380542641, + "grad_norm": 0.646577000617981, + "learning_rate": 6.207487852488103e-06, + "loss": 0.2004, + "step": 51340 + }, + { + "epoch": 3.5626257719797376, + "grad_norm": 0.9833005666732788, + "learning_rate": 6.205539252250025e-06, + "loss": 0.2579, + "step": 51341 + }, + { + "epoch": 3.5626951634168345, + "grad_norm": 0.9861702919006348, + "learning_rate": 6.203590948110172e-06, + "loss": 0.221, + "step": 51342 + }, + { + "epoch": 3.5627645548539313, + "grad_norm": 0.643463671207428, + "learning_rate": 6.201642940074704e-06, + "loss": 0.3127, + "step": 51343 + }, + { + "epoch": 3.5628339462910277, + "grad_norm": 1.071976900100708, + "learning_rate": 6.19969522814976e-06, + "loss": 0.2594, + "step": 51344 + }, + { + "epoch": 3.5629033377281245, + "grad_norm": 1.15886390209198, + "learning_rate": 6.1977478123414925e-06, + "loss": 0.3593, + "step": 51345 + }, + { + "epoch": 3.562972729165221, + "grad_norm": 0.964090883731842, + "learning_rate": 6.195800692656073e-06, + "loss": 0.2238, + "step": 51346 + }, + { + "epoch": 3.5630421206023177, + "grad_norm": 0.9756112694740295, + "learning_rate": 6.19385386909962e-06, + "loss": 0.343, + "step": 51347 + }, + { + "epoch": 3.5631115120394146, + "grad_norm": 0.9299501180648804, + "learning_rate": 6.191907341678271e-06, + "loss": 0.2339, + "step": 51348 + }, + { + "epoch": 3.563180903476511, + "grad_norm": 0.8956248164176941, + "learning_rate": 6.189961110398201e-06, + "loss": 0.205, + "step": 51349 + }, + { + "epoch": 3.563250294913608, + "grad_norm": 0.90663743019104, + "learning_rate": 6.188015175265527e-06, + "loss": 0.2761, + "step": 51350 + }, + { + "epoch": 3.563319686350704, + "grad_norm": 0.7579056024551392, + "learning_rate": 6.186069536286409e-06, + "loss": 0.2077, + "step": 51351 + }, + { + "epoch": 3.563389077787801, + "grad_norm": 0.772178590297699, + "learning_rate": 6.184124193466989e-06, + "loss": 0.1606, + "step": 51352 + }, + { + "epoch": 3.563458469224898, + "grad_norm": 0.757156491279602, + "learning_rate": 6.182179146813416e-06, + "loss": 0.1266, + "step": 51353 + }, + { + "epoch": 3.5635278606619942, + "grad_norm": 1.325262188911438, + "learning_rate": 6.180234396331808e-06, + "loss": 0.3729, + "step": 51354 + }, + { + "epoch": 3.563597252099091, + "grad_norm": 0.9369901418685913, + "learning_rate": 6.1782899420283145e-06, + "loss": 0.2303, + "step": 51355 + }, + { + "epoch": 3.5636666435361875, + "grad_norm": 0.708217978477478, + "learning_rate": 6.176345783909077e-06, + "loss": 0.1688, + "step": 51356 + }, + { + "epoch": 3.5637360349732843, + "grad_norm": 0.9429049491882324, + "learning_rate": 6.1744019219802215e-06, + "loss": 0.2787, + "step": 51357 + }, + { + "epoch": 3.563805426410381, + "grad_norm": 1.0447897911071777, + "learning_rate": 6.17245835624789e-06, + "loss": 0.2155, + "step": 51358 + }, + { + "epoch": 3.5638748178474775, + "grad_norm": 0.7082905173301697, + "learning_rate": 6.170515086718232e-06, + "loss": 0.2344, + "step": 51359 + }, + { + "epoch": 3.5639442092845743, + "grad_norm": 1.7124494314193726, + "learning_rate": 6.168572113397386e-06, + "loss": 0.3404, + "step": 51360 + }, + { + "epoch": 3.5640136007216707, + "grad_norm": 1.2461044788360596, + "learning_rate": 6.166629436291449e-06, + "loss": 0.2562, + "step": 51361 + }, + { + "epoch": 3.5640829921587676, + "grad_norm": 0.7871841788291931, + "learning_rate": 6.1646870554065815e-06, + "loss": 0.3398, + "step": 51362 + }, + { + "epoch": 3.5641523835958644, + "grad_norm": 1.084972620010376, + "learning_rate": 6.162744970748912e-06, + "loss": 0.317, + "step": 51363 + }, + { + "epoch": 3.564221775032961, + "grad_norm": 0.7540543675422668, + "learning_rate": 6.160803182324571e-06, + "loss": 0.2926, + "step": 51364 + }, + { + "epoch": 3.5642911664700576, + "grad_norm": 0.5870683193206787, + "learning_rate": 6.1588616901396945e-06, + "loss": 0.1178, + "step": 51365 + }, + { + "epoch": 3.564360557907154, + "grad_norm": 0.9523805379867554, + "learning_rate": 6.156920494200391e-06, + "loss": 0.2763, + "step": 51366 + }, + { + "epoch": 3.564429949344251, + "grad_norm": 1.0849981307983398, + "learning_rate": 6.1549795945127995e-06, + "loss": 0.2642, + "step": 51367 + }, + { + "epoch": 3.5644993407813477, + "grad_norm": 1.0262401103973389, + "learning_rate": 6.15303899108306e-06, + "loss": 0.2771, + "step": 51368 + }, + { + "epoch": 3.5645687322184445, + "grad_norm": 0.7624838352203369, + "learning_rate": 6.151098683917289e-06, + "loss": 0.287, + "step": 51369 + }, + { + "epoch": 3.564638123655541, + "grad_norm": 0.9556900262832642, + "learning_rate": 6.149158673021604e-06, + "loss": 0.281, + "step": 51370 + }, + { + "epoch": 3.5647075150926377, + "grad_norm": 0.6837893128395081, + "learning_rate": 6.1472189584021345e-06, + "loss": 0.2398, + "step": 51371 + }, + { + "epoch": 3.564776906529734, + "grad_norm": 0.6825289726257324, + "learning_rate": 6.145279540065007e-06, + "loss": 0.1645, + "step": 51372 + }, + { + "epoch": 3.564846297966831, + "grad_norm": 0.7762882113456726, + "learning_rate": 6.143340418016341e-06, + "loss": 0.2762, + "step": 51373 + }, + { + "epoch": 3.564915689403928, + "grad_norm": 1.1282670497894287, + "learning_rate": 6.1414015922622855e-06, + "loss": 0.2912, + "step": 51374 + }, + { + "epoch": 3.564985080841024, + "grad_norm": 1.0780528783798218, + "learning_rate": 6.139463062808903e-06, + "loss": 0.3605, + "step": 51375 + }, + { + "epoch": 3.565054472278121, + "grad_norm": 1.0594549179077148, + "learning_rate": 6.1375248296623555e-06, + "loss": 0.2064, + "step": 51376 + }, + { + "epoch": 3.5651238637152174, + "grad_norm": 1.1362370252609253, + "learning_rate": 6.1355868928287485e-06, + "loss": 0.3314, + "step": 51377 + }, + { + "epoch": 3.5651932551523142, + "grad_norm": 0.9039732813835144, + "learning_rate": 6.133649252314211e-06, + "loss": 0.1389, + "step": 51378 + }, + { + "epoch": 3.565262646589411, + "grad_norm": 0.8571991920471191, + "learning_rate": 6.131711908124849e-06, + "loss": 0.3904, + "step": 51379 + }, + { + "epoch": 3.5653320380265074, + "grad_norm": 0.9249658584594727, + "learning_rate": 6.129774860266779e-06, + "loss": 0.3272, + "step": 51380 + }, + { + "epoch": 3.5654014294636043, + "grad_norm": 0.9941613674163818, + "learning_rate": 6.127838108746132e-06, + "loss": 0.2326, + "step": 51381 + }, + { + "epoch": 3.5654708209007007, + "grad_norm": 0.6391299962997437, + "learning_rate": 6.1259016535689995e-06, + "loss": 0.1341, + "step": 51382 + }, + { + "epoch": 3.5655402123377975, + "grad_norm": 0.6261831521987915, + "learning_rate": 6.123965494741502e-06, + "loss": 0.2093, + "step": 51383 + }, + { + "epoch": 3.5656096037748943, + "grad_norm": 1.0209144353866577, + "learning_rate": 6.1220296322697545e-06, + "loss": 0.3381, + "step": 51384 + }, + { + "epoch": 3.5656789952119907, + "grad_norm": 1.1931287050247192, + "learning_rate": 6.120094066159865e-06, + "loss": 0.361, + "step": 51385 + }, + { + "epoch": 3.5657483866490876, + "grad_norm": 0.7260229587554932, + "learning_rate": 6.118158796417939e-06, + "loss": 0.1971, + "step": 51386 + }, + { + "epoch": 3.565817778086184, + "grad_norm": 1.2687526941299438, + "learning_rate": 6.116223823050104e-06, + "loss": 0.3385, + "step": 51387 + }, + { + "epoch": 3.5658871695232808, + "grad_norm": 1.0095473527908325, + "learning_rate": 6.114289146062468e-06, + "loss": 0.3142, + "step": 51388 + }, + { + "epoch": 3.5659565609603776, + "grad_norm": 0.9858906269073486, + "learning_rate": 6.112354765461126e-06, + "loss": 0.2587, + "step": 51389 + }, + { + "epoch": 3.566025952397474, + "grad_norm": 0.8619805574417114, + "learning_rate": 6.110420681252171e-06, + "loss": 0.2388, + "step": 51390 + }, + { + "epoch": 3.566095343834571, + "grad_norm": 0.7534229159355164, + "learning_rate": 6.1084868934417334e-06, + "loss": 0.186, + "step": 51391 + }, + { + "epoch": 3.566164735271667, + "grad_norm": 0.6713239550590515, + "learning_rate": 6.1065534020359085e-06, + "loss": 0.2323, + "step": 51392 + }, + { + "epoch": 3.566234126708764, + "grad_norm": 0.9043782949447632, + "learning_rate": 6.104620207040801e-06, + "loss": 0.2693, + "step": 51393 + }, + { + "epoch": 3.566303518145861, + "grad_norm": 0.703119158744812, + "learning_rate": 6.102687308462507e-06, + "loss": 0.1982, + "step": 51394 + }, + { + "epoch": 3.5663729095829577, + "grad_norm": 1.1931401491165161, + "learning_rate": 6.100754706307154e-06, + "loss": 0.3078, + "step": 51395 + }, + { + "epoch": 3.566442301020054, + "grad_norm": 0.9964532256126404, + "learning_rate": 6.098822400580806e-06, + "loss": 0.3878, + "step": 51396 + }, + { + "epoch": 3.566511692457151, + "grad_norm": 0.6427058577537537, + "learning_rate": 6.096890391289589e-06, + "loss": 0.1916, + "step": 51397 + }, + { + "epoch": 3.5665810838942473, + "grad_norm": 1.124592661857605, + "learning_rate": 6.094958678439588e-06, + "loss": 0.3904, + "step": 51398 + }, + { + "epoch": 3.566650475331344, + "grad_norm": 0.9669375419616699, + "learning_rate": 6.09302726203691e-06, + "loss": 0.3219, + "step": 51399 + }, + { + "epoch": 3.566719866768441, + "grad_norm": 0.9762160181999207, + "learning_rate": 6.09109614208766e-06, + "loss": 0.3059, + "step": 51400 + }, + { + "epoch": 3.5667892582055374, + "grad_norm": 0.877863347530365, + "learning_rate": 6.089165318597922e-06, + "loss": 0.144, + "step": 51401 + }, + { + "epoch": 3.566858649642634, + "grad_norm": 1.0041948556900024, + "learning_rate": 6.087234791573781e-06, + "loss": 0.2071, + "step": 51402 + }, + { + "epoch": 3.5669280410797306, + "grad_norm": 0.7579694390296936, + "learning_rate": 6.0853045610213435e-06, + "loss": 0.1878, + "step": 51403 + }, + { + "epoch": 3.5669974325168274, + "grad_norm": 0.9624242782592773, + "learning_rate": 6.0833746269467255e-06, + "loss": 0.2232, + "step": 51404 + }, + { + "epoch": 3.5670668239539243, + "grad_norm": 0.7276006937026978, + "learning_rate": 6.081444989355978e-06, + "loss": 0.2358, + "step": 51405 + }, + { + "epoch": 3.5671362153910207, + "grad_norm": 1.0155348777770996, + "learning_rate": 6.079515648255219e-06, + "loss": 0.2023, + "step": 51406 + }, + { + "epoch": 3.5672056068281175, + "grad_norm": 0.7946224808692932, + "learning_rate": 6.07758660365052e-06, + "loss": 0.2177, + "step": 51407 + }, + { + "epoch": 3.567274998265214, + "grad_norm": 0.9879103899002075, + "learning_rate": 6.075657855548011e-06, + "loss": 0.2791, + "step": 51408 + }, + { + "epoch": 3.5673443897023107, + "grad_norm": 0.7630360722541809, + "learning_rate": 6.073729403953732e-06, + "loss": 0.3149, + "step": 51409 + }, + { + "epoch": 3.5674137811394075, + "grad_norm": 0.6365960836410522, + "learning_rate": 6.071801248873788e-06, + "loss": 0.1944, + "step": 51410 + }, + { + "epoch": 3.567483172576504, + "grad_norm": 0.967704713344574, + "learning_rate": 6.069873390314273e-06, + "loss": 0.3106, + "step": 51411 + }, + { + "epoch": 3.5675525640136008, + "grad_norm": 0.9844874739646912, + "learning_rate": 6.067945828281274e-06, + "loss": 0.3759, + "step": 51412 + }, + { + "epoch": 3.567621955450697, + "grad_norm": 0.6159204244613647, + "learning_rate": 6.066018562780862e-06, + "loss": 0.1788, + "step": 51413 + }, + { + "epoch": 3.567691346887794, + "grad_norm": 1.153333067893982, + "learning_rate": 6.064091593819143e-06, + "loss": 0.1951, + "step": 51414 + }, + { + "epoch": 3.567760738324891, + "grad_norm": 1.2720162868499756, + "learning_rate": 6.062164921402191e-06, + "loss": 0.2841, + "step": 51415 + }, + { + "epoch": 3.567830129761987, + "grad_norm": 1.0120017528533936, + "learning_rate": 6.060238545536079e-06, + "loss": 0.2358, + "step": 51416 + }, + { + "epoch": 3.567899521199084, + "grad_norm": 1.5596692562103271, + "learning_rate": 6.05831246622689e-06, + "loss": 0.1543, + "step": 51417 + }, + { + "epoch": 3.5679689126361804, + "grad_norm": 1.3965210914611816, + "learning_rate": 6.056386683480719e-06, + "loss": 0.2615, + "step": 51418 + }, + { + "epoch": 3.5680383040732773, + "grad_norm": 0.7756149172782898, + "learning_rate": 6.054461197303629e-06, + "loss": 0.2218, + "step": 51419 + }, + { + "epoch": 3.568107695510374, + "grad_norm": 0.7490004897117615, + "learning_rate": 6.052536007701703e-06, + "loss": 0.2187, + "step": 51420 + }, + { + "epoch": 3.568177086947471, + "grad_norm": 1.032894253730774, + "learning_rate": 6.050611114681026e-06, + "loss": 0.2537, + "step": 51421 + }, + { + "epoch": 3.5682464783845673, + "grad_norm": 0.9110926985740662, + "learning_rate": 6.0486865182476815e-06, + "loss": 0.2119, + "step": 51422 + }, + { + "epoch": 3.568315869821664, + "grad_norm": 0.9598128199577332, + "learning_rate": 6.0467622184077195e-06, + "loss": 0.292, + "step": 51423 + }, + { + "epoch": 3.5683852612587605, + "grad_norm": 1.3163257837295532, + "learning_rate": 6.044838215167226e-06, + "loss": 0.3991, + "step": 51424 + }, + { + "epoch": 3.5684546526958574, + "grad_norm": 0.7728427052497864, + "learning_rate": 6.042914508532271e-06, + "loss": 0.1902, + "step": 51425 + }, + { + "epoch": 3.568524044132954, + "grad_norm": 1.1770310401916504, + "learning_rate": 6.040991098508941e-06, + "loss": 0.1818, + "step": 51426 + }, + { + "epoch": 3.5685934355700506, + "grad_norm": 0.666244387626648, + "learning_rate": 6.039067985103308e-06, + "loss": 0.1579, + "step": 51427 + }, + { + "epoch": 3.5686628270071474, + "grad_norm": 0.7050934433937073, + "learning_rate": 6.037145168321423e-06, + "loss": 0.2087, + "step": 51428 + }, + { + "epoch": 3.568732218444244, + "grad_norm": 0.9156745672225952, + "learning_rate": 6.035222648169392e-06, + "loss": 0.2307, + "step": 51429 + }, + { + "epoch": 3.5688016098813407, + "grad_norm": 0.9766915440559387, + "learning_rate": 6.033300424653243e-06, + "loss": 0.3012, + "step": 51430 + }, + { + "epoch": 3.5688710013184375, + "grad_norm": 1.153479814529419, + "learning_rate": 6.031378497779072e-06, + "loss": 0.2588, + "step": 51431 + }, + { + "epoch": 3.568940392755534, + "grad_norm": 0.6727672815322876, + "learning_rate": 6.0294568675529295e-06, + "loss": 0.2253, + "step": 51432 + }, + { + "epoch": 3.5690097841926307, + "grad_norm": 1.0959110260009766, + "learning_rate": 6.027535533980888e-06, + "loss": 0.2518, + "step": 51433 + }, + { + "epoch": 3.569079175629727, + "grad_norm": 0.9942406415939331, + "learning_rate": 6.02561449706901e-06, + "loss": 0.2523, + "step": 51434 + }, + { + "epoch": 3.569148567066824, + "grad_norm": 1.15646493434906, + "learning_rate": 6.02369375682339e-06, + "loss": 0.315, + "step": 51435 + }, + { + "epoch": 3.5692179585039208, + "grad_norm": 1.3554373979568481, + "learning_rate": 6.0217733132500346e-06, + "loss": 0.2919, + "step": 51436 + }, + { + "epoch": 3.569287349941017, + "grad_norm": 1.0490657091140747, + "learning_rate": 6.019853166355049e-06, + "loss": 0.3068, + "step": 51437 + }, + { + "epoch": 3.569356741378114, + "grad_norm": 0.7098843455314636, + "learning_rate": 6.017933316144486e-06, + "loss": 0.1893, + "step": 51438 + }, + { + "epoch": 3.5694261328152104, + "grad_norm": 0.7448325157165527, + "learning_rate": 6.016013762624406e-06, + "loss": 0.207, + "step": 51439 + }, + { + "epoch": 3.569495524252307, + "grad_norm": 1.143978238105774, + "learning_rate": 6.01409450580086e-06, + "loss": 0.2408, + "step": 51440 + }, + { + "epoch": 3.569564915689404, + "grad_norm": 1.1475515365600586, + "learning_rate": 6.01217554567991e-06, + "loss": 0.3031, + "step": 51441 + }, + { + "epoch": 3.5696343071265004, + "grad_norm": 1.0165023803710938, + "learning_rate": 6.0102568822676285e-06, + "loss": 0.2475, + "step": 51442 + }, + { + "epoch": 3.5697036985635973, + "grad_norm": 1.0383846759796143, + "learning_rate": 6.0083385155700665e-06, + "loss": 0.2697, + "step": 51443 + }, + { + "epoch": 3.5697730900006936, + "grad_norm": 0.8751335740089417, + "learning_rate": 6.006420445593253e-06, + "loss": 0.2599, + "step": 51444 + }, + { + "epoch": 3.5698424814377905, + "grad_norm": 0.7158989310264587, + "learning_rate": 6.0045026723432705e-06, + "loss": 0.1489, + "step": 51445 + }, + { + "epoch": 3.5699118728748873, + "grad_norm": 0.8988571763038635, + "learning_rate": 6.002585195826171e-06, + "loss": 0.3534, + "step": 51446 + }, + { + "epoch": 3.569981264311984, + "grad_norm": 1.1837716102600098, + "learning_rate": 6.000668016048006e-06, + "loss": 0.3437, + "step": 51447 + }, + { + "epoch": 3.5700506557490805, + "grad_norm": 0.5789409279823303, + "learning_rate": 5.998751133014824e-06, + "loss": 0.2057, + "step": 51448 + }, + { + "epoch": 3.5701200471861774, + "grad_norm": 0.8459409475326538, + "learning_rate": 5.996834546732677e-06, + "loss": 0.2942, + "step": 51449 + }, + { + "epoch": 3.5701894386232738, + "grad_norm": 1.0083500146865845, + "learning_rate": 5.994918257207627e-06, + "loss": 0.3316, + "step": 51450 + }, + { + "epoch": 3.5702588300603706, + "grad_norm": 0.7552492618560791, + "learning_rate": 5.993002264445702e-06, + "loss": 0.2094, + "step": 51451 + }, + { + "epoch": 3.5703282214974674, + "grad_norm": 0.8546078205108643, + "learning_rate": 5.991086568452964e-06, + "loss": 0.3216, + "step": 51452 + }, + { + "epoch": 3.570397612934564, + "grad_norm": 0.7499123811721802, + "learning_rate": 5.989171169235463e-06, + "loss": 0.1448, + "step": 51453 + }, + { + "epoch": 3.5704670043716606, + "grad_norm": 1.168531060218811, + "learning_rate": 5.98725606679923e-06, + "loss": 0.3627, + "step": 51454 + }, + { + "epoch": 3.570536395808757, + "grad_norm": 0.7466926574707031, + "learning_rate": 5.9853412611503345e-06, + "loss": 0.1887, + "step": 51455 + }, + { + "epoch": 3.570605787245854, + "grad_norm": 0.8601219058036804, + "learning_rate": 5.983426752294807e-06, + "loss": 0.2153, + "step": 51456 + }, + { + "epoch": 3.5706751786829507, + "grad_norm": 1.126654028892517, + "learning_rate": 5.9815125402386985e-06, + "loss": 0.2638, + "step": 51457 + }, + { + "epoch": 3.570744570120047, + "grad_norm": 0.9812729358673096, + "learning_rate": 5.979598624988036e-06, + "loss": 0.2991, + "step": 51458 + }, + { + "epoch": 3.570813961557144, + "grad_norm": 0.9855859875679016, + "learning_rate": 5.977685006548872e-06, + "loss": 0.3716, + "step": 51459 + }, + { + "epoch": 3.5708833529942403, + "grad_norm": 0.7488444447517395, + "learning_rate": 5.975771684927256e-06, + "loss": 0.2766, + "step": 51460 + }, + { + "epoch": 3.570952744431337, + "grad_norm": 0.9161871075630188, + "learning_rate": 5.973858660129217e-06, + "loss": 0.2718, + "step": 51461 + }, + { + "epoch": 3.571022135868434, + "grad_norm": 0.8998867869377136, + "learning_rate": 5.971945932160783e-06, + "loss": 0.2871, + "step": 51462 + }, + { + "epoch": 3.5710915273055304, + "grad_norm": 1.1032798290252686, + "learning_rate": 5.970033501028027e-06, + "loss": 0.2984, + "step": 51463 + }, + { + "epoch": 3.571160918742627, + "grad_norm": 1.016086459159851, + "learning_rate": 5.968121366736967e-06, + "loss": 0.2107, + "step": 51464 + }, + { + "epoch": 3.5712303101797236, + "grad_norm": 0.7794323563575745, + "learning_rate": 5.9662095292936315e-06, + "loss": 0.2619, + "step": 51465 + }, + { + "epoch": 3.5712997016168204, + "grad_norm": 0.8277110457420349, + "learning_rate": 5.964297988704048e-06, + "loss": 0.2641, + "step": 51466 + }, + { + "epoch": 3.5713690930539173, + "grad_norm": 0.7967990636825562, + "learning_rate": 5.96238674497428e-06, + "loss": 0.2674, + "step": 51467 + }, + { + "epoch": 3.5714384844910136, + "grad_norm": 0.47853928804397583, + "learning_rate": 5.9604757981103434e-06, + "loss": 0.1491, + "step": 51468 + }, + { + "epoch": 3.5715078759281105, + "grad_norm": 0.8994473218917847, + "learning_rate": 5.958565148118278e-06, + "loss": 0.2441, + "step": 51469 + }, + { + "epoch": 3.571577267365207, + "grad_norm": 1.133501410484314, + "learning_rate": 5.956654795004113e-06, + "loss": 0.3744, + "step": 51470 + }, + { + "epoch": 3.5716466588023037, + "grad_norm": 0.7013941407203674, + "learning_rate": 5.954744738773877e-06, + "loss": 0.2443, + "step": 51471 + }, + { + "epoch": 3.5717160502394005, + "grad_norm": 0.8666144013404846, + "learning_rate": 5.952834979433597e-06, + "loss": 0.3266, + "step": 51472 + }, + { + "epoch": 3.5717854416764974, + "grad_norm": 0.9497785568237305, + "learning_rate": 5.950925516989314e-06, + "loss": 0.2684, + "step": 51473 + }, + { + "epoch": 3.5718548331135938, + "grad_norm": 1.5882049798965454, + "learning_rate": 5.949016351447034e-06, + "loss": 0.4131, + "step": 51474 + }, + { + "epoch": 3.5719242245506906, + "grad_norm": 0.5768918395042419, + "learning_rate": 5.947107482812808e-06, + "loss": 0.1231, + "step": 51475 + }, + { + "epoch": 3.571993615987787, + "grad_norm": 0.7510012984275818, + "learning_rate": 5.945198911092642e-06, + "loss": 0.235, + "step": 51476 + }, + { + "epoch": 3.572063007424884, + "grad_norm": 1.060967206954956, + "learning_rate": 5.943290636292575e-06, + "loss": 0.336, + "step": 51477 + }, + { + "epoch": 3.5721323988619806, + "grad_norm": 0.8894834518432617, + "learning_rate": 5.9413826584186375e-06, + "loss": 0.2521, + "step": 51478 + }, + { + "epoch": 3.572201790299077, + "grad_norm": 1.0408881902694702, + "learning_rate": 5.939474977476822e-06, + "loss": 0.3909, + "step": 51479 + }, + { + "epoch": 3.572271181736174, + "grad_norm": 1.3836334943771362, + "learning_rate": 5.937567593473181e-06, + "loss": 0.2824, + "step": 51480 + }, + { + "epoch": 3.5723405731732703, + "grad_norm": 0.6137698292732239, + "learning_rate": 5.935660506413721e-06, + "loss": 0.1911, + "step": 51481 + }, + { + "epoch": 3.572409964610367, + "grad_norm": 1.236634373664856, + "learning_rate": 5.933753716304469e-06, + "loss": 0.3583, + "step": 51482 + }, + { + "epoch": 3.572479356047464, + "grad_norm": 1.0340399742126465, + "learning_rate": 5.931847223151443e-06, + "loss": 0.2565, + "step": 51483 + }, + { + "epoch": 3.5725487474845603, + "grad_norm": 0.9770065546035767, + "learning_rate": 5.929941026960684e-06, + "loss": 0.3901, + "step": 51484 + }, + { + "epoch": 3.572618138921657, + "grad_norm": 1.0047448873519897, + "learning_rate": 5.928035127738163e-06, + "loss": 0.2911, + "step": 51485 + }, + { + "epoch": 3.5726875303587535, + "grad_norm": 0.789510190486908, + "learning_rate": 5.926129525489921e-06, + "loss": 0.2179, + "step": 51486 + }, + { + "epoch": 3.5727569217958504, + "grad_norm": 0.7742142081260681, + "learning_rate": 5.9242242202219745e-06, + "loss": 0.211, + "step": 51487 + }, + { + "epoch": 3.572826313232947, + "grad_norm": 0.6750882863998413, + "learning_rate": 5.922319211940342e-06, + "loss": 0.1632, + "step": 51488 + }, + { + "epoch": 3.5728957046700436, + "grad_norm": 0.9252640604972839, + "learning_rate": 5.9204145006510284e-06, + "loss": 0.2788, + "step": 51489 + }, + { + "epoch": 3.5729650961071404, + "grad_norm": 0.8491544723510742, + "learning_rate": 5.91851008636004e-06, + "loss": 0.2476, + "step": 51490 + }, + { + "epoch": 3.573034487544237, + "grad_norm": 1.0650995969772339, + "learning_rate": 5.91660596907343e-06, + "loss": 0.2392, + "step": 51491 + }, + { + "epoch": 3.5731038789813336, + "grad_norm": 1.2189674377441406, + "learning_rate": 5.914702148797158e-06, + "loss": 0.2406, + "step": 51492 + }, + { + "epoch": 3.5731732704184305, + "grad_norm": 1.8842642307281494, + "learning_rate": 5.912798625537264e-06, + "loss": 0.468, + "step": 51493 + }, + { + "epoch": 3.5732426618555273, + "grad_norm": 0.7036843299865723, + "learning_rate": 5.910895399299743e-06, + "loss": 0.1951, + "step": 51494 + }, + { + "epoch": 3.5733120532926237, + "grad_norm": 0.8526234030723572, + "learning_rate": 5.908992470090602e-06, + "loss": 0.2958, + "step": 51495 + }, + { + "epoch": 3.57338144472972, + "grad_norm": 1.0053362846374512, + "learning_rate": 5.907089837915858e-06, + "loss": 0.4449, + "step": 51496 + }, + { + "epoch": 3.573450836166817, + "grad_norm": 0.7576597929000854, + "learning_rate": 5.905187502781517e-06, + "loss": 0.2017, + "step": 51497 + }, + { + "epoch": 3.5735202276039137, + "grad_norm": 1.1625484228134155, + "learning_rate": 5.903285464693598e-06, + "loss": 0.2195, + "step": 51498 + }, + { + "epoch": 3.5735896190410106, + "grad_norm": 0.7757992744445801, + "learning_rate": 5.901383723658083e-06, + "loss": 0.2853, + "step": 51499 + }, + { + "epoch": 3.573659010478107, + "grad_norm": 1.0390537977218628, + "learning_rate": 5.899482279680968e-06, + "loss": 0.3019, + "step": 51500 + }, + { + "epoch": 3.573728401915204, + "grad_norm": 0.5255516767501831, + "learning_rate": 5.8975811327682706e-06, + "loss": 0.1786, + "step": 51501 + }, + { + "epoch": 3.5737977933523, + "grad_norm": 0.76971834897995, + "learning_rate": 5.895680282925997e-06, + "loss": 0.2337, + "step": 51502 + }, + { + "epoch": 3.573867184789397, + "grad_norm": 0.8558154106140137, + "learning_rate": 5.893779730160143e-06, + "loss": 0.3681, + "step": 51503 + }, + { + "epoch": 3.573936576226494, + "grad_norm": 0.979539155960083, + "learning_rate": 5.891879474476714e-06, + "loss": 0.16, + "step": 51504 + }, + { + "epoch": 3.5740059676635902, + "grad_norm": 1.1252070665359497, + "learning_rate": 5.889979515881705e-06, + "loss": 0.2609, + "step": 51505 + }, + { + "epoch": 3.574075359100687, + "grad_norm": 0.8592262864112854, + "learning_rate": 5.888079854381101e-06, + "loss": 0.2678, + "step": 51506 + }, + { + "epoch": 3.5741447505377835, + "grad_norm": 1.1162080764770508, + "learning_rate": 5.886180489980908e-06, + "loss": 0.1835, + "step": 51507 + }, + { + "epoch": 3.5742141419748803, + "grad_norm": 0.676594078540802, + "learning_rate": 5.884281422687143e-06, + "loss": 0.1921, + "step": 51508 + }, + { + "epoch": 3.574283533411977, + "grad_norm": 0.8341657519340515, + "learning_rate": 5.882382652505769e-06, + "loss": 0.173, + "step": 51509 + }, + { + "epoch": 3.5743529248490735, + "grad_norm": 0.8246192932128906, + "learning_rate": 5.880484179442791e-06, + "loss": 0.1535, + "step": 51510 + }, + { + "epoch": 3.5744223162861704, + "grad_norm": 0.9031786918640137, + "learning_rate": 5.878586003504205e-06, + "loss": 0.2552, + "step": 51511 + }, + { + "epoch": 3.5744917077232667, + "grad_norm": 0.8417367339134216, + "learning_rate": 5.876688124696017e-06, + "loss": 0.3064, + "step": 51512 + }, + { + "epoch": 3.5745610991603636, + "grad_norm": 0.9631388783454895, + "learning_rate": 5.874790543024189e-06, + "loss": 0.2255, + "step": 51513 + }, + { + "epoch": 3.5746304905974604, + "grad_norm": 0.6470048427581787, + "learning_rate": 5.872893258494727e-06, + "loss": 0.1371, + "step": 51514 + }, + { + "epoch": 3.574699882034557, + "grad_norm": 0.9665395021438599, + "learning_rate": 5.870996271113627e-06, + "loss": 0.1936, + "step": 51515 + }, + { + "epoch": 3.5747692734716536, + "grad_norm": 0.583802342414856, + "learning_rate": 5.869099580886861e-06, + "loss": 0.1358, + "step": 51516 + }, + { + "epoch": 3.57483866490875, + "grad_norm": 1.127446174621582, + "learning_rate": 5.867203187820424e-06, + "loss": 0.2153, + "step": 51517 + }, + { + "epoch": 3.574908056345847, + "grad_norm": 1.3719080686569214, + "learning_rate": 5.865307091920313e-06, + "loss": 0.2585, + "step": 51518 + }, + { + "epoch": 3.5749774477829437, + "grad_norm": 1.1459347009658813, + "learning_rate": 5.863411293192522e-06, + "loss": 0.2572, + "step": 51519 + }, + { + "epoch": 3.5750468392200405, + "grad_norm": 0.6552416682243347, + "learning_rate": 5.861515791643002e-06, + "loss": 0.2577, + "step": 51520 + }, + { + "epoch": 3.575116230657137, + "grad_norm": 1.3927475214004517, + "learning_rate": 5.859620587277748e-06, + "loss": 0.3988, + "step": 51521 + }, + { + "epoch": 3.5751856220942333, + "grad_norm": 1.1366207599639893, + "learning_rate": 5.857725680102755e-06, + "loss": 0.2936, + "step": 51522 + }, + { + "epoch": 3.57525501353133, + "grad_norm": 0.5962337255477905, + "learning_rate": 5.8558310701240075e-06, + "loss": 0.1479, + "step": 51523 + }, + { + "epoch": 3.575324404968427, + "grad_norm": 1.0801993608474731, + "learning_rate": 5.853936757347467e-06, + "loss": 0.3156, + "step": 51524 + }, + { + "epoch": 3.575393796405524, + "grad_norm": 1.1866673231124878, + "learning_rate": 5.852042741779129e-06, + "loss": 0.2988, + "step": 51525 + }, + { + "epoch": 3.57546318784262, + "grad_norm": 0.5522202849388123, + "learning_rate": 5.850149023424989e-06, + "loss": 0.1314, + "step": 51526 + }, + { + "epoch": 3.575532579279717, + "grad_norm": 0.9239606857299805, + "learning_rate": 5.848255602290975e-06, + "loss": 0.233, + "step": 51527 + }, + { + "epoch": 3.5756019707168134, + "grad_norm": 0.8739354014396667, + "learning_rate": 5.846362478383105e-06, + "loss": 0.268, + "step": 51528 + }, + { + "epoch": 3.5756713621539102, + "grad_norm": 2.6970887184143066, + "learning_rate": 5.844469651707352e-06, + "loss": 0.2503, + "step": 51529 + }, + { + "epoch": 3.575740753591007, + "grad_norm": 0.6866276264190674, + "learning_rate": 5.842577122269677e-06, + "loss": 0.2112, + "step": 51530 + }, + { + "epoch": 3.5758101450281035, + "grad_norm": 0.8509052395820618, + "learning_rate": 5.840684890076053e-06, + "loss": 0.1928, + "step": 51531 + }, + { + "epoch": 3.5758795364652003, + "grad_norm": 1.0894696712493896, + "learning_rate": 5.8387929551324775e-06, + "loss": 0.2676, + "step": 51532 + }, + { + "epoch": 3.5759489279022967, + "grad_norm": 0.9868451356887817, + "learning_rate": 5.8369013174449095e-06, + "loss": 0.2078, + "step": 51533 + }, + { + "epoch": 3.5760183193393935, + "grad_norm": 1.2839927673339844, + "learning_rate": 5.8350099770193125e-06, + "loss": 0.2131, + "step": 51534 + }, + { + "epoch": 3.5760877107764903, + "grad_norm": 0.9460684061050415, + "learning_rate": 5.833118933861659e-06, + "loss": 0.3456, + "step": 51535 + }, + { + "epoch": 3.5761571022135867, + "grad_norm": 1.1270328760147095, + "learning_rate": 5.831228187977933e-06, + "loss": 0.4003, + "step": 51536 + }, + { + "epoch": 3.5762264936506836, + "grad_norm": 1.1177270412445068, + "learning_rate": 5.829337739374085e-06, + "loss": 0.1963, + "step": 51537 + }, + { + "epoch": 3.57629588508778, + "grad_norm": 0.833315372467041, + "learning_rate": 5.8274475880561e-06, + "loss": 0.2549, + "step": 51538 + }, + { + "epoch": 3.576365276524877, + "grad_norm": 0.9749574661254883, + "learning_rate": 5.825557734029929e-06, + "loss": 0.3503, + "step": 51539 + }, + { + "epoch": 3.5764346679619736, + "grad_norm": 1.1671724319458008, + "learning_rate": 5.8236681773015665e-06, + "loss": 0.2795, + "step": 51540 + }, + { + "epoch": 3.57650405939907, + "grad_norm": 1.2844700813293457, + "learning_rate": 5.82177891787693e-06, + "loss": 0.2809, + "step": 51541 + }, + { + "epoch": 3.576573450836167, + "grad_norm": 1.2333710193634033, + "learning_rate": 5.819889955762026e-06, + "loss": 0.2294, + "step": 51542 + }, + { + "epoch": 3.5766428422732632, + "grad_norm": 0.8616039156913757, + "learning_rate": 5.818001290962805e-06, + "loss": 0.246, + "step": 51543 + }, + { + "epoch": 3.57671223371036, + "grad_norm": 1.0328782796859741, + "learning_rate": 5.816112923485228e-06, + "loss": 0.2863, + "step": 51544 + }, + { + "epoch": 3.576781625147457, + "grad_norm": 0.7831310033798218, + "learning_rate": 5.8142248533352374e-06, + "loss": 0.2047, + "step": 51545 + }, + { + "epoch": 3.5768510165845537, + "grad_norm": 1.1498225927352905, + "learning_rate": 5.812337080518826e-06, + "loss": 0.2243, + "step": 51546 + }, + { + "epoch": 3.57692040802165, + "grad_norm": 0.8842519521713257, + "learning_rate": 5.810449605041946e-06, + "loss": 0.364, + "step": 51547 + }, + { + "epoch": 3.576989799458747, + "grad_norm": 0.9716346263885498, + "learning_rate": 5.8085624269105355e-06, + "loss": 0.3162, + "step": 51548 + }, + { + "epoch": 3.5770591908958433, + "grad_norm": 0.9008537530899048, + "learning_rate": 5.806675546130558e-06, + "loss": 0.2654, + "step": 51549 + }, + { + "epoch": 3.57712858233294, + "grad_norm": 0.8937724828720093, + "learning_rate": 5.804788962707985e-06, + "loss": 0.2879, + "step": 51550 + }, + { + "epoch": 3.577197973770037, + "grad_norm": 0.9198893904685974, + "learning_rate": 5.8029026766487585e-06, + "loss": 0.1819, + "step": 51551 + }, + { + "epoch": 3.5772673652071334, + "grad_norm": 0.7832980751991272, + "learning_rate": 5.801016687958849e-06, + "loss": 0.1802, + "step": 51552 + }, + { + "epoch": 3.5773367566442302, + "grad_norm": 1.1204771995544434, + "learning_rate": 5.799130996644198e-06, + "loss": 0.1831, + "step": 51553 + }, + { + "epoch": 3.5774061480813266, + "grad_norm": 1.0083589553833008, + "learning_rate": 5.797245602710777e-06, + "loss": 0.2627, + "step": 51554 + }, + { + "epoch": 3.5774755395184235, + "grad_norm": 1.0587869882583618, + "learning_rate": 5.795360506164493e-06, + "loss": 0.3144, + "step": 51555 + }, + { + "epoch": 3.5775449309555203, + "grad_norm": 0.7338546514511108, + "learning_rate": 5.793475707011342e-06, + "loss": 0.2072, + "step": 51556 + }, + { + "epoch": 3.5776143223926167, + "grad_norm": 0.7184389233589172, + "learning_rate": 5.7915912052572514e-06, + "loss": 0.2557, + "step": 51557 + }, + { + "epoch": 3.5776837138297135, + "grad_norm": 0.7317013740539551, + "learning_rate": 5.789707000908184e-06, + "loss": 0.2086, + "step": 51558 + }, + { + "epoch": 3.57775310526681, + "grad_norm": 0.9261500239372253, + "learning_rate": 5.787823093970068e-06, + "loss": 0.1943, + "step": 51559 + }, + { + "epoch": 3.5778224967039067, + "grad_norm": 0.9323372840881348, + "learning_rate": 5.785939484448888e-06, + "loss": 0.2364, + "step": 51560 + }, + { + "epoch": 3.5778918881410036, + "grad_norm": 0.7417411804199219, + "learning_rate": 5.784056172350549e-06, + "loss": 0.1779, + "step": 51561 + }, + { + "epoch": 3.5779612795781, + "grad_norm": 0.8131251931190491, + "learning_rate": 5.782173157681026e-06, + "loss": 0.2173, + "step": 51562 + }, + { + "epoch": 3.578030671015197, + "grad_norm": 0.8133764266967773, + "learning_rate": 5.780290440446234e-06, + "loss": 0.18, + "step": 51563 + }, + { + "epoch": 3.578100062452293, + "grad_norm": 0.9246103167533875, + "learning_rate": 5.778408020652138e-06, + "loss": 0.257, + "step": 51564 + }, + { + "epoch": 3.57816945388939, + "grad_norm": 1.1910524368286133, + "learning_rate": 5.776525898304675e-06, + "loss": 0.1549, + "step": 51565 + }, + { + "epoch": 3.578238845326487, + "grad_norm": 0.5937407612800598, + "learning_rate": 5.774644073409796e-06, + "loss": 0.1836, + "step": 51566 + }, + { + "epoch": 3.5783082367635832, + "grad_norm": 0.9217613339424133, + "learning_rate": 5.7727625459734424e-06, + "loss": 0.2736, + "step": 51567 + }, + { + "epoch": 3.57837762820068, + "grad_norm": 1.004677176475525, + "learning_rate": 5.770881316001531e-06, + "loss": 0.3214, + "step": 51568 + }, + { + "epoch": 3.5784470196377764, + "grad_norm": 0.9110960960388184, + "learning_rate": 5.7690003835000116e-06, + "loss": 0.2413, + "step": 51569 + }, + { + "epoch": 3.5785164110748733, + "grad_norm": 0.9192642569541931, + "learning_rate": 5.767119748474825e-06, + "loss": 0.146, + "step": 51570 + }, + { + "epoch": 3.57858580251197, + "grad_norm": 0.7743070125579834, + "learning_rate": 5.765239410931911e-06, + "loss": 0.2439, + "step": 51571 + }, + { + "epoch": 3.578655193949067, + "grad_norm": 0.6962907314300537, + "learning_rate": 5.763359370877208e-06, + "loss": 0.2064, + "step": 51572 + }, + { + "epoch": 3.5787245853861633, + "grad_norm": 0.7723226547241211, + "learning_rate": 5.761479628316635e-06, + "loss": 0.3732, + "step": 51573 + }, + { + "epoch": 3.57879397682326, + "grad_norm": 0.8680897951126099, + "learning_rate": 5.759600183256164e-06, + "loss": 0.2421, + "step": 51574 + }, + { + "epoch": 3.5788633682603566, + "grad_norm": 0.9056901335716248, + "learning_rate": 5.75772103570168e-06, + "loss": 0.1967, + "step": 51575 + }, + { + "epoch": 3.5789327596974534, + "grad_norm": 0.7450051307678223, + "learning_rate": 5.755842185659132e-06, + "loss": 0.1174, + "step": 51576 + }, + { + "epoch": 3.5790021511345502, + "grad_norm": 0.7242051959037781, + "learning_rate": 5.753963633134462e-06, + "loss": 0.1923, + "step": 51577 + }, + { + "epoch": 3.5790715425716466, + "grad_norm": 1.8026326894760132, + "learning_rate": 5.752085378133609e-06, + "loss": 0.3282, + "step": 51578 + }, + { + "epoch": 3.5791409340087434, + "grad_norm": 0.8132685422897339, + "learning_rate": 5.750207420662479e-06, + "loss": 0.1598, + "step": 51579 + }, + { + "epoch": 3.57921032544584, + "grad_norm": 0.8841106295585632, + "learning_rate": 5.748329760727e-06, + "loss": 0.2079, + "step": 51580 + }, + { + "epoch": 3.5792797168829367, + "grad_norm": 0.8721357583999634, + "learning_rate": 5.746452398333113e-06, + "loss": 0.1772, + "step": 51581 + }, + { + "epoch": 3.5793491083200335, + "grad_norm": 1.0408868789672852, + "learning_rate": 5.744575333486746e-06, + "loss": 0.212, + "step": 51582 + }, + { + "epoch": 3.57941849975713, + "grad_norm": 0.7962291240692139, + "learning_rate": 5.742698566193816e-06, + "loss": 0.1767, + "step": 51583 + }, + { + "epoch": 3.5794878911942267, + "grad_norm": 1.1048365831375122, + "learning_rate": 5.740822096460241e-06, + "loss": 0.2876, + "step": 51584 + }, + { + "epoch": 3.579557282631323, + "grad_norm": 0.9462204575538635, + "learning_rate": 5.738945924291961e-06, + "loss": 0.264, + "step": 51585 + }, + { + "epoch": 3.57962667406842, + "grad_norm": 0.8428036570549011, + "learning_rate": 5.737070049694893e-06, + "loss": 0.2988, + "step": 51586 + }, + { + "epoch": 3.5796960655055168, + "grad_norm": 1.120790958404541, + "learning_rate": 5.735194472674954e-06, + "loss": 0.2, + "step": 51587 + }, + { + "epoch": 3.579765456942613, + "grad_norm": 0.8492388129234314, + "learning_rate": 5.733319193238085e-06, + "loss": 0.3427, + "step": 51588 + }, + { + "epoch": 3.57983484837971, + "grad_norm": 0.7103350758552551, + "learning_rate": 5.73144421139018e-06, + "loss": 0.2009, + "step": 51589 + }, + { + "epoch": 3.5799042398168064, + "grad_norm": 2.224201202392578, + "learning_rate": 5.729569527137169e-06, + "loss": 0.2121, + "step": 51590 + }, + { + "epoch": 3.579973631253903, + "grad_norm": 0.8165731430053711, + "learning_rate": 5.727695140484957e-06, + "loss": 0.2359, + "step": 51591 + }, + { + "epoch": 3.580043022691, + "grad_norm": 0.9368600845336914, + "learning_rate": 5.725821051439484e-06, + "loss": 0.2623, + "step": 51592 + }, + { + "epoch": 3.5801124141280964, + "grad_norm": 1.1225979328155518, + "learning_rate": 5.723947260006657e-06, + "loss": 0.2739, + "step": 51593 + }, + { + "epoch": 3.5801818055651933, + "grad_norm": 0.6696231961250305, + "learning_rate": 5.722073766192393e-06, + "loss": 0.1625, + "step": 51594 + }, + { + "epoch": 3.5802511970022897, + "grad_norm": 0.8207347989082336, + "learning_rate": 5.72020057000261e-06, + "loss": 0.1852, + "step": 51595 + }, + { + "epoch": 3.5803205884393865, + "grad_norm": 0.8609098196029663, + "learning_rate": 5.7183276714432024e-06, + "loss": 0.1767, + "step": 51596 + }, + { + "epoch": 3.5803899798764833, + "grad_norm": 0.8604956865310669, + "learning_rate": 5.7164550705200995e-06, + "loss": 0.1636, + "step": 51597 + }, + { + "epoch": 3.58045937131358, + "grad_norm": 1.2002464532852173, + "learning_rate": 5.714582767239207e-06, + "loss": 0.3627, + "step": 51598 + }, + { + "epoch": 3.5805287627506766, + "grad_norm": 0.7933488488197327, + "learning_rate": 5.712710761606432e-06, + "loss": 0.2078, + "step": 51599 + }, + { + "epoch": 3.5805981541877734, + "grad_norm": 1.6174265146255493, + "learning_rate": 5.710839053627704e-06, + "loss": 0.3737, + "step": 51600 + }, + { + "epoch": 3.5806675456248698, + "grad_norm": 0.7892276644706726, + "learning_rate": 5.708967643308905e-06, + "loss": 0.1991, + "step": 51601 + }, + { + "epoch": 3.5807369370619666, + "grad_norm": 0.727104902267456, + "learning_rate": 5.707096530655975e-06, + "loss": 0.1269, + "step": 51602 + }, + { + "epoch": 3.5808063284990634, + "grad_norm": 0.6453808546066284, + "learning_rate": 5.705225715674789e-06, + "loss": 0.188, + "step": 51603 + }, + { + "epoch": 3.58087571993616, + "grad_norm": 0.7524245381355286, + "learning_rate": 5.703355198371263e-06, + "loss": 0.2203, + "step": 51604 + }, + { + "epoch": 3.5809451113732567, + "grad_norm": 1.1559007167816162, + "learning_rate": 5.701484978751304e-06, + "loss": 0.2336, + "step": 51605 + }, + { + "epoch": 3.581014502810353, + "grad_norm": 1.6260449886322021, + "learning_rate": 5.699615056820806e-06, + "loss": 0.2953, + "step": 51606 + }, + { + "epoch": 3.58108389424745, + "grad_norm": 1.0286277532577515, + "learning_rate": 5.6977454325857e-06, + "loss": 0.2481, + "step": 51607 + }, + { + "epoch": 3.5811532856845467, + "grad_norm": 0.9846267700195312, + "learning_rate": 5.6958761060518565e-06, + "loss": 0.4091, + "step": 51608 + }, + { + "epoch": 3.581222677121643, + "grad_norm": 0.7747650742530823, + "learning_rate": 5.694007077225216e-06, + "loss": 0.2104, + "step": 51609 + }, + { + "epoch": 3.58129206855874, + "grad_norm": 1.450397253036499, + "learning_rate": 5.6921383461116305e-06, + "loss": 0.2831, + "step": 51610 + }, + { + "epoch": 3.5813614599958363, + "grad_norm": 0.8607847094535828, + "learning_rate": 5.6902699127170275e-06, + "loss": 0.2312, + "step": 51611 + }, + { + "epoch": 3.581430851432933, + "grad_norm": 1.4185079336166382, + "learning_rate": 5.688401777047303e-06, + "loss": 0.3581, + "step": 51612 + }, + { + "epoch": 3.58150024287003, + "grad_norm": 0.6362334489822388, + "learning_rate": 5.686533939108363e-06, + "loss": 0.1853, + "step": 51613 + }, + { + "epoch": 3.5815696343071264, + "grad_norm": 0.7125116586685181, + "learning_rate": 5.684666398906069e-06, + "loss": 0.1935, + "step": 51614 + }, + { + "epoch": 3.581639025744223, + "grad_norm": 0.7145470380783081, + "learning_rate": 5.68279915644635e-06, + "loss": 0.1683, + "step": 51615 + }, + { + "epoch": 3.5817084171813196, + "grad_norm": 1.050846815109253, + "learning_rate": 5.680932211735113e-06, + "loss": 0.1917, + "step": 51616 + }, + { + "epoch": 3.5817778086184164, + "grad_norm": 0.782599151134491, + "learning_rate": 5.679065564778208e-06, + "loss": 0.2639, + "step": 51617 + }, + { + "epoch": 3.5818472000555133, + "grad_norm": 0.8623230457305908, + "learning_rate": 5.677199215581552e-06, + "loss": 0.2541, + "step": 51618 + }, + { + "epoch": 3.5819165914926097, + "grad_norm": 0.9161871075630188, + "learning_rate": 5.675333164151031e-06, + "loss": 0.2775, + "step": 51619 + }, + { + "epoch": 3.5819859829297065, + "grad_norm": 1.326545000076294, + "learning_rate": 5.673467410492539e-06, + "loss": 0.2739, + "step": 51620 + }, + { + "epoch": 3.582055374366803, + "grad_norm": 0.7478650212287903, + "learning_rate": 5.671601954611971e-06, + "loss": 0.1634, + "step": 51621 + }, + { + "epoch": 3.5821247658038997, + "grad_norm": 0.8336227536201477, + "learning_rate": 5.669736796515201e-06, + "loss": 0.3345, + "step": 51622 + }, + { + "epoch": 3.5821941572409965, + "grad_norm": 0.8535602688789368, + "learning_rate": 5.667871936208158e-06, + "loss": 0.1363, + "step": 51623 + }, + { + "epoch": 3.5822635486780934, + "grad_norm": 1.2038654088974, + "learning_rate": 5.666007373696669e-06, + "loss": 0.278, + "step": 51624 + }, + { + "epoch": 3.5823329401151898, + "grad_norm": 1.0851800441741943, + "learning_rate": 5.6641431089866636e-06, + "loss": 0.2814, + "step": 51625 + }, + { + "epoch": 3.5824023315522866, + "grad_norm": 0.9451051950454712, + "learning_rate": 5.6622791420840036e-06, + "loss": 0.2083, + "step": 51626 + }, + { + "epoch": 3.582471722989383, + "grad_norm": 0.851095974445343, + "learning_rate": 5.660415472994585e-06, + "loss": 0.2559, + "step": 51627 + }, + { + "epoch": 3.58254111442648, + "grad_norm": 0.8984389305114746, + "learning_rate": 5.65855210172429e-06, + "loss": 0.2667, + "step": 51628 + }, + { + "epoch": 3.5826105058635767, + "grad_norm": 1.1062685251235962, + "learning_rate": 5.656689028279005e-06, + "loss": 0.3015, + "step": 51629 + }, + { + "epoch": 3.582679897300673, + "grad_norm": 0.6962354183197021, + "learning_rate": 5.654826252664614e-06, + "loss": 0.2575, + "step": 51630 + }, + { + "epoch": 3.58274928873777, + "grad_norm": 1.081221342086792, + "learning_rate": 5.6529637748869765e-06, + "loss": 0.2513, + "step": 51631 + }, + { + "epoch": 3.5828186801748663, + "grad_norm": 0.8204379677772522, + "learning_rate": 5.651101594951991e-06, + "loss": 0.2996, + "step": 51632 + }, + { + "epoch": 3.582888071611963, + "grad_norm": 0.9230974316596985, + "learning_rate": 5.649239712865528e-06, + "loss": 0.2533, + "step": 51633 + }, + { + "epoch": 3.58295746304906, + "grad_norm": 0.7033126354217529, + "learning_rate": 5.647378128633474e-06, + "loss": 0.2533, + "step": 51634 + }, + { + "epoch": 3.5830268544861563, + "grad_norm": 1.055955410003662, + "learning_rate": 5.6455168422617e-06, + "loss": 0.2123, + "step": 51635 + }, + { + "epoch": 3.583096245923253, + "grad_norm": 0.7935937643051147, + "learning_rate": 5.643655853756091e-06, + "loss": 0.2029, + "step": 51636 + }, + { + "epoch": 3.5831656373603495, + "grad_norm": 0.8147938251495361, + "learning_rate": 5.641795163122498e-06, + "loss": 0.1599, + "step": 51637 + }, + { + "epoch": 3.5832350287974464, + "grad_norm": 0.9659389853477478, + "learning_rate": 5.639934770366817e-06, + "loss": 0.2856, + "step": 51638 + }, + { + "epoch": 3.583304420234543, + "grad_norm": 0.7732385993003845, + "learning_rate": 5.638074675494909e-06, + "loss": 0.1265, + "step": 51639 + }, + { + "epoch": 3.5833738116716396, + "grad_norm": 0.8888916373252869, + "learning_rate": 5.636214878512647e-06, + "loss": 0.2088, + "step": 51640 + }, + { + "epoch": 3.5834432031087364, + "grad_norm": 0.8548680543899536, + "learning_rate": 5.634355379425915e-06, + "loss": 0.3614, + "step": 51641 + }, + { + "epoch": 3.583512594545833, + "grad_norm": 0.8853455185890198, + "learning_rate": 5.632496178240565e-06, + "loss": 0.2788, + "step": 51642 + }, + { + "epoch": 3.5835819859829297, + "grad_norm": 0.7634934186935425, + "learning_rate": 5.630637274962502e-06, + "loss": 0.2217, + "step": 51643 + }, + { + "epoch": 3.5836513774200265, + "grad_norm": 0.9285545945167542, + "learning_rate": 5.628778669597545e-06, + "loss": 0.288, + "step": 51644 + }, + { + "epoch": 3.583720768857123, + "grad_norm": 0.9781999588012695, + "learning_rate": 5.626920362151589e-06, + "loss": 0.2177, + "step": 51645 + }, + { + "epoch": 3.5837901602942197, + "grad_norm": 0.8591026663780212, + "learning_rate": 5.6250623526304944e-06, + "loss": 0.1952, + "step": 51646 + }, + { + "epoch": 3.583859551731316, + "grad_norm": 0.8384235501289368, + "learning_rate": 5.623204641040125e-06, + "loss": 0.2609, + "step": 51647 + }, + { + "epoch": 3.583928943168413, + "grad_norm": 0.8564715385437012, + "learning_rate": 5.621347227386375e-06, + "loss": 0.219, + "step": 51648 + }, + { + "epoch": 3.5839983346055098, + "grad_norm": 0.693752110004425, + "learning_rate": 5.619490111675052e-06, + "loss": 0.1527, + "step": 51649 + }, + { + "epoch": 3.5840677260426066, + "grad_norm": 0.9134982824325562, + "learning_rate": 5.617633293912061e-06, + "loss": 0.1902, + "step": 51650 + }, + { + "epoch": 3.584137117479703, + "grad_norm": 0.9734229445457458, + "learning_rate": 5.6157767741032655e-06, + "loss": 0.2945, + "step": 51651 + }, + { + "epoch": 3.5842065089168, + "grad_norm": 1.1744290590286255, + "learning_rate": 5.6139205522544925e-06, + "loss": 0.1777, + "step": 51652 + }, + { + "epoch": 3.584275900353896, + "grad_norm": 0.8527517318725586, + "learning_rate": 5.612064628371616e-06, + "loss": 0.1817, + "step": 51653 + }, + { + "epoch": 3.584345291790993, + "grad_norm": 0.996173620223999, + "learning_rate": 5.610209002460509e-06, + "loss": 0.2086, + "step": 51654 + }, + { + "epoch": 3.58441468322809, + "grad_norm": 1.0166897773742676, + "learning_rate": 5.608353674527023e-06, + "loss": 0.264, + "step": 51655 + }, + { + "epoch": 3.5844840746651863, + "grad_norm": 0.8631578087806702, + "learning_rate": 5.6064986445770075e-06, + "loss": 0.243, + "step": 51656 + }, + { + "epoch": 3.584553466102283, + "grad_norm": 0.8093588352203369, + "learning_rate": 5.6046439126163365e-06, + "loss": 0.23, + "step": 51657 + }, + { + "epoch": 3.5846228575393795, + "grad_norm": 0.8934723734855652, + "learning_rate": 5.602789478650838e-06, + "loss": 0.2992, + "step": 51658 + }, + { + "epoch": 3.5846922489764763, + "grad_norm": 0.9948782920837402, + "learning_rate": 5.600935342686386e-06, + "loss": 0.3379, + "step": 51659 + }, + { + "epoch": 3.584761640413573, + "grad_norm": 0.9724931716918945, + "learning_rate": 5.599081504728831e-06, + "loss": 0.3103, + "step": 51660 + }, + { + "epoch": 3.5848310318506695, + "grad_norm": 0.8705523014068604, + "learning_rate": 5.5972279647840135e-06, + "loss": 0.225, + "step": 51661 + }, + { + "epoch": 3.5849004232877664, + "grad_norm": 2.2811288833618164, + "learning_rate": 5.595374722857805e-06, + "loss": 0.3142, + "step": 51662 + }, + { + "epoch": 3.5849698147248628, + "grad_norm": 0.8776967525482178, + "learning_rate": 5.593521778956046e-06, + "loss": 0.3111, + "step": 51663 + }, + { + "epoch": 3.5850392061619596, + "grad_norm": 1.3634217977523804, + "learning_rate": 5.591669133084587e-06, + "loss": 0.3033, + "step": 51664 + }, + { + "epoch": 3.5851085975990564, + "grad_norm": 1.0165717601776123, + "learning_rate": 5.589816785249269e-06, + "loss": 0.2318, + "step": 51665 + }, + { + "epoch": 3.585177989036153, + "grad_norm": 0.840589165687561, + "learning_rate": 5.587964735455953e-06, + "loss": 0.286, + "step": 51666 + }, + { + "epoch": 3.5852473804732496, + "grad_norm": 0.9435800313949585, + "learning_rate": 5.5861129837104675e-06, + "loss": 0.1765, + "step": 51667 + }, + { + "epoch": 3.585316771910346, + "grad_norm": 1.176418423652649, + "learning_rate": 5.5842615300186754e-06, + "loss": 0.3383, + "step": 51668 + }, + { + "epoch": 3.585386163347443, + "grad_norm": 1.4512174129486084, + "learning_rate": 5.582410374386404e-06, + "loss": 0.3011, + "step": 51669 + }, + { + "epoch": 3.5854555547845397, + "grad_norm": 1.0410499572753906, + "learning_rate": 5.580559516819517e-06, + "loss": 0.1897, + "step": 51670 + }, + { + "epoch": 3.585524946221636, + "grad_norm": 0.8647339940071106, + "learning_rate": 5.578708957323864e-06, + "loss": 0.2278, + "step": 51671 + }, + { + "epoch": 3.585594337658733, + "grad_norm": 0.7533950805664062, + "learning_rate": 5.576858695905263e-06, + "loss": 0.2159, + "step": 51672 + }, + { + "epoch": 3.5856637290958293, + "grad_norm": 0.8655288815498352, + "learning_rate": 5.575008732569553e-06, + "loss": 0.1834, + "step": 51673 + }, + { + "epoch": 3.585733120532926, + "grad_norm": 0.8960856199264526, + "learning_rate": 5.573159067322598e-06, + "loss": 0.1963, + "step": 51674 + }, + { + "epoch": 3.585802511970023, + "grad_norm": 0.825541615486145, + "learning_rate": 5.571309700170213e-06, + "loss": 0.2433, + "step": 51675 + }, + { + "epoch": 3.58587190340712, + "grad_norm": 0.9533430337905884, + "learning_rate": 5.569460631118251e-06, + "loss": 0.1675, + "step": 51676 + }, + { + "epoch": 3.585941294844216, + "grad_norm": 0.8106095790863037, + "learning_rate": 5.56761186017255e-06, + "loss": 0.1982, + "step": 51677 + }, + { + "epoch": 3.586010686281313, + "grad_norm": 1.0785130262374878, + "learning_rate": 5.565763387338952e-06, + "loss": 0.1974, + "step": 51678 + }, + { + "epoch": 3.5860800777184094, + "grad_norm": 0.7570303082466125, + "learning_rate": 5.563915212623261e-06, + "loss": 0.2449, + "step": 51679 + }, + { + "epoch": 3.5861494691555063, + "grad_norm": 0.9575263857841492, + "learning_rate": 5.56206733603134e-06, + "loss": 0.2361, + "step": 51680 + }, + { + "epoch": 3.586218860592603, + "grad_norm": 0.8182405829429626, + "learning_rate": 5.560219757569018e-06, + "loss": 0.249, + "step": 51681 + }, + { + "epoch": 3.5862882520296995, + "grad_norm": 0.873882532119751, + "learning_rate": 5.558372477242124e-06, + "loss": 0.3391, + "step": 51682 + }, + { + "epoch": 3.5863576434667963, + "grad_norm": 0.9048212170600891, + "learning_rate": 5.556525495056508e-06, + "loss": 0.2629, + "step": 51683 + }, + { + "epoch": 3.5864270349038927, + "grad_norm": 0.9319739937782288, + "learning_rate": 5.554678811017966e-06, + "loss": 0.3002, + "step": 51684 + }, + { + "epoch": 3.5864964263409895, + "grad_norm": 1.012581467628479, + "learning_rate": 5.552832425132348e-06, + "loss": 0.2564, + "step": 51685 + }, + { + "epoch": 3.5865658177780864, + "grad_norm": 0.8335639834403992, + "learning_rate": 5.550986337405484e-06, + "loss": 0.2841, + "step": 51686 + }, + { + "epoch": 3.5866352092151828, + "grad_norm": 0.7737265825271606, + "learning_rate": 5.5491405478431906e-06, + "loss": 0.2476, + "step": 51687 + }, + { + "epoch": 3.5867046006522796, + "grad_norm": 1.6600613594055176, + "learning_rate": 5.547295056451307e-06, + "loss": 0.2748, + "step": 51688 + }, + { + "epoch": 3.586773992089376, + "grad_norm": 0.8625996708869934, + "learning_rate": 5.545449863235652e-06, + "loss": 0.4073, + "step": 51689 + }, + { + "epoch": 3.586843383526473, + "grad_norm": 1.3111931085586548, + "learning_rate": 5.543604968202054e-06, + "loss": 0.2143, + "step": 51690 + }, + { + "epoch": 3.5869127749635696, + "grad_norm": 1.0137346982955933, + "learning_rate": 5.54176037135633e-06, + "loss": 0.2957, + "step": 51691 + }, + { + "epoch": 3.586982166400666, + "grad_norm": 1.0927895307540894, + "learning_rate": 5.539916072704332e-06, + "loss": 0.3209, + "step": 51692 + }, + { + "epoch": 3.587051557837763, + "grad_norm": 0.7673420310020447, + "learning_rate": 5.538072072251832e-06, + "loss": 0.2445, + "step": 51693 + }, + { + "epoch": 3.5871209492748592, + "grad_norm": 0.9221831560134888, + "learning_rate": 5.536228370004692e-06, + "loss": 0.2273, + "step": 51694 + }, + { + "epoch": 3.587190340711956, + "grad_norm": 1.1150236129760742, + "learning_rate": 5.534384965968709e-06, + "loss": 0.2007, + "step": 51695 + }, + { + "epoch": 3.587259732149053, + "grad_norm": 0.7643255591392517, + "learning_rate": 5.5325418601497205e-06, + "loss": 0.2135, + "step": 51696 + }, + { + "epoch": 3.5873291235861493, + "grad_norm": 1.0722293853759766, + "learning_rate": 5.530699052553534e-06, + "loss": 0.2846, + "step": 51697 + }, + { + "epoch": 3.587398515023246, + "grad_norm": 0.8718449473381042, + "learning_rate": 5.528856543185967e-06, + "loss": 0.204, + "step": 51698 + }, + { + "epoch": 3.5874679064603425, + "grad_norm": 0.9700788259506226, + "learning_rate": 5.52701433205286e-06, + "loss": 0.2122, + "step": 51699 + }, + { + "epoch": 3.5875372978974394, + "grad_norm": 1.106027603149414, + "learning_rate": 5.525172419159996e-06, + "loss": 0.4292, + "step": 51700 + }, + { + "epoch": 3.587606689334536, + "grad_norm": 1.4925124645233154, + "learning_rate": 5.523330804513194e-06, + "loss": 0.3673, + "step": 51701 + }, + { + "epoch": 3.587676080771633, + "grad_norm": 1.137400507926941, + "learning_rate": 5.521489488118281e-06, + "loss": 0.2519, + "step": 51702 + }, + { + "epoch": 3.5877454722087294, + "grad_norm": 0.9848760366439819, + "learning_rate": 5.519648469981054e-06, + "loss": 0.2595, + "step": 51703 + }, + { + "epoch": 3.5878148636458262, + "grad_norm": 1.6973289251327515, + "learning_rate": 5.517807750107351e-06, + "loss": 0.2444, + "step": 51704 + }, + { + "epoch": 3.5878842550829226, + "grad_norm": 1.5259617567062378, + "learning_rate": 5.5159673285029574e-06, + "loss": 0.2841, + "step": 51705 + }, + { + "epoch": 3.5879536465200195, + "grad_norm": 2.644375801086426, + "learning_rate": 5.5141272051737025e-06, + "loss": 0.2712, + "step": 51706 + }, + { + "epoch": 3.5880230379571163, + "grad_norm": 0.7447394132614136, + "learning_rate": 5.5122873801253805e-06, + "loss": 0.2813, + "step": 51707 + }, + { + "epoch": 3.5880924293942127, + "grad_norm": 0.8738107085227966, + "learning_rate": 5.510447853363799e-06, + "loss": 0.2781, + "step": 51708 + }, + { + "epoch": 3.5881618208313095, + "grad_norm": 0.8434457778930664, + "learning_rate": 5.5086086248947734e-06, + "loss": 0.2422, + "step": 51709 + }, + { + "epoch": 3.588231212268406, + "grad_norm": 0.97432941198349, + "learning_rate": 5.506769694724101e-06, + "loss": 0.2421, + "step": 51710 + }, + { + "epoch": 3.5883006037055027, + "grad_norm": 1.1374355554580688, + "learning_rate": 5.504931062857599e-06, + "loss": 0.2339, + "step": 51711 + }, + { + "epoch": 3.5883699951425996, + "grad_norm": 0.665338933467865, + "learning_rate": 5.503092729301085e-06, + "loss": 0.2333, + "step": 51712 + }, + { + "epoch": 3.588439386579696, + "grad_norm": 1.3475419282913208, + "learning_rate": 5.50125469406032e-06, + "loss": 0.3859, + "step": 51713 + }, + { + "epoch": 3.588508778016793, + "grad_norm": 0.7982243895530701, + "learning_rate": 5.499416957141135e-06, + "loss": 0.1952, + "step": 51714 + }, + { + "epoch": 3.588578169453889, + "grad_norm": 1.0178016424179077, + "learning_rate": 5.497579518549323e-06, + "loss": 0.3315, + "step": 51715 + }, + { + "epoch": 3.588647560890986, + "grad_norm": 0.9964006543159485, + "learning_rate": 5.4957423782906915e-06, + "loss": 0.3334, + "step": 51716 + }, + { + "epoch": 3.588716952328083, + "grad_norm": 1.0220307111740112, + "learning_rate": 5.493905536371058e-06, + "loss": 0.2425, + "step": 51717 + }, + { + "epoch": 3.5887863437651792, + "grad_norm": 0.9106454849243164, + "learning_rate": 5.492068992796173e-06, + "loss": 0.3729, + "step": 51718 + }, + { + "epoch": 3.588855735202276, + "grad_norm": 1.1025254726409912, + "learning_rate": 5.4902327475718665e-06, + "loss": 0.3531, + "step": 51719 + }, + { + "epoch": 3.5889251266393725, + "grad_norm": 1.064992904663086, + "learning_rate": 5.488396800703932e-06, + "loss": 0.2582, + "step": 51720 + }, + { + "epoch": 3.5889945180764693, + "grad_norm": 1.4592944383621216, + "learning_rate": 5.4865611521981776e-06, + "loss": 0.295, + "step": 51721 + }, + { + "epoch": 3.589063909513566, + "grad_norm": 0.9274242520332336, + "learning_rate": 5.484725802060364e-06, + "loss": 0.2726, + "step": 51722 + }, + { + "epoch": 3.589133300950663, + "grad_norm": 1.0222949981689453, + "learning_rate": 5.48289075029631e-06, + "loss": 0.3052, + "step": 51723 + }, + { + "epoch": 3.5892026923877594, + "grad_norm": 0.7510730028152466, + "learning_rate": 5.481055996911811e-06, + "loss": 0.2416, + "step": 51724 + }, + { + "epoch": 3.5892720838248557, + "grad_norm": 1.6980849504470825, + "learning_rate": 5.479221541912638e-06, + "loss": 0.3622, + "step": 51725 + }, + { + "epoch": 3.5893414752619526, + "grad_norm": 0.8660241365432739, + "learning_rate": 5.477387385304622e-06, + "loss": 0.2012, + "step": 51726 + }, + { + "epoch": 3.5894108666990494, + "grad_norm": 1.170608639717102, + "learning_rate": 5.4755535270935025e-06, + "loss": 0.3141, + "step": 51727 + }, + { + "epoch": 3.5894802581361462, + "grad_norm": 0.6976094841957092, + "learning_rate": 5.473719967285096e-06, + "loss": 0.2014, + "step": 51728 + }, + { + "epoch": 3.5895496495732426, + "grad_norm": 0.7151244878768921, + "learning_rate": 5.471886705885199e-06, + "loss": 0.2594, + "step": 51729 + }, + { + "epoch": 3.5896190410103395, + "grad_norm": 0.7332584857940674, + "learning_rate": 5.470053742899572e-06, + "loss": 0.2061, + "step": 51730 + }, + { + "epoch": 3.589688432447436, + "grad_norm": 0.8972859978675842, + "learning_rate": 5.468221078334035e-06, + "loss": 0.2557, + "step": 51731 + }, + { + "epoch": 3.5897578238845327, + "grad_norm": 0.7086901068687439, + "learning_rate": 5.466388712194348e-06, + "loss": 0.226, + "step": 51732 + }, + { + "epoch": 3.5898272153216295, + "grad_norm": 0.8338697552680969, + "learning_rate": 5.464556644486318e-06, + "loss": 0.1581, + "step": 51733 + }, + { + "epoch": 3.589896606758726, + "grad_norm": 0.9787591695785522, + "learning_rate": 5.462724875215697e-06, + "loss": 0.1869, + "step": 51734 + }, + { + "epoch": 3.5899659981958227, + "grad_norm": 1.009650468826294, + "learning_rate": 5.46089340438829e-06, + "loss": 0.2469, + "step": 51735 + }, + { + "epoch": 3.590035389632919, + "grad_norm": 1.248623251914978, + "learning_rate": 5.459062232009871e-06, + "loss": 0.232, + "step": 51736 + }, + { + "epoch": 3.590104781070016, + "grad_norm": 0.9280574917793274, + "learning_rate": 5.457231358086223e-06, + "loss": 0.3031, + "step": 51737 + }, + { + "epoch": 3.590174172507113, + "grad_norm": 0.6815729737281799, + "learning_rate": 5.4554007826231325e-06, + "loss": 0.1557, + "step": 51738 + }, + { + "epoch": 3.590243563944209, + "grad_norm": 0.9362959861755371, + "learning_rate": 5.4535705056263705e-06, + "loss": 0.1599, + "step": 51739 + }, + { + "epoch": 3.590312955381306, + "grad_norm": 0.968592643737793, + "learning_rate": 5.451740527101723e-06, + "loss": 0.2599, + "step": 51740 + }, + { + "epoch": 3.5903823468184024, + "grad_norm": 0.8650992512702942, + "learning_rate": 5.449910847054962e-06, + "loss": 0.2281, + "step": 51741 + }, + { + "epoch": 3.5904517382554992, + "grad_norm": 0.6691127419471741, + "learning_rate": 5.4480814654918496e-06, + "loss": 0.1846, + "step": 51742 + }, + { + "epoch": 3.590521129692596, + "grad_norm": 0.8669589161872864, + "learning_rate": 5.446252382418182e-06, + "loss": 0.2163, + "step": 51743 + }, + { + "epoch": 3.5905905211296925, + "grad_norm": 1.0175501108169556, + "learning_rate": 5.44442359783972e-06, + "loss": 0.2533, + "step": 51744 + }, + { + "epoch": 3.5906599125667893, + "grad_norm": 1.0097333192825317, + "learning_rate": 5.442595111762239e-06, + "loss": 0.2686, + "step": 51745 + }, + { + "epoch": 3.5907293040038857, + "grad_norm": 0.8455225229263306, + "learning_rate": 5.4407669241915204e-06, + "loss": 0.271, + "step": 51746 + }, + { + "epoch": 3.5907986954409825, + "grad_norm": 0.7438333630561829, + "learning_rate": 5.43893903513335e-06, + "loss": 0.2168, + "step": 51747 + }, + { + "epoch": 3.5908680868780793, + "grad_norm": 1.123168706893921, + "learning_rate": 5.437111444593457e-06, + "loss": 0.2707, + "step": 51748 + }, + { + "epoch": 3.590937478315176, + "grad_norm": 0.8221660256385803, + "learning_rate": 5.435284152577624e-06, + "loss": 0.178, + "step": 51749 + }, + { + "epoch": 3.5910068697522726, + "grad_norm": 1.0464625358581543, + "learning_rate": 5.433457159091637e-06, + "loss": 0.2623, + "step": 51750 + }, + { + "epoch": 3.591076261189369, + "grad_norm": 1.0226209163665771, + "learning_rate": 5.431630464141257e-06, + "loss": 0.2107, + "step": 51751 + }, + { + "epoch": 3.591145652626466, + "grad_norm": 1.0347620248794556, + "learning_rate": 5.429804067732258e-06, + "loss": 0.2234, + "step": 51752 + }, + { + "epoch": 3.5912150440635626, + "grad_norm": 0.6956655979156494, + "learning_rate": 5.427977969870379e-06, + "loss": 0.2206, + "step": 51753 + }, + { + "epoch": 3.5912844355006595, + "grad_norm": 1.0997151136398315, + "learning_rate": 5.426152170561405e-06, + "loss": 0.3393, + "step": 51754 + }, + { + "epoch": 3.591353826937756, + "grad_norm": 0.8376298546791077, + "learning_rate": 5.4243266698110865e-06, + "loss": 0.2367, + "step": 51755 + }, + { + "epoch": 3.5914232183748527, + "grad_norm": 0.9924122095108032, + "learning_rate": 5.422501467625218e-06, + "loss": 0.2423, + "step": 51756 + }, + { + "epoch": 3.591492609811949, + "grad_norm": 1.6425111293792725, + "learning_rate": 5.4206765640095305e-06, + "loss": 0.4806, + "step": 51757 + }, + { + "epoch": 3.591562001249046, + "grad_norm": 1.2205853462219238, + "learning_rate": 5.418851958969784e-06, + "loss": 0.3, + "step": 51758 + }, + { + "epoch": 3.5916313926861427, + "grad_norm": 1.5184358358383179, + "learning_rate": 5.417027652511753e-06, + "loss": 0.3887, + "step": 51759 + }, + { + "epoch": 3.591700784123239, + "grad_norm": 0.9570134878158569, + "learning_rate": 5.4152036446411865e-06, + "loss": 0.284, + "step": 51760 + }, + { + "epoch": 3.591770175560336, + "grad_norm": 0.8097816109657288, + "learning_rate": 5.413379935363871e-06, + "loss": 0.1822, + "step": 51761 + }, + { + "epoch": 3.5918395669974323, + "grad_norm": 0.5825318694114685, + "learning_rate": 5.4115565246855235e-06, + "loss": 0.1023, + "step": 51762 + }, + { + "epoch": 3.591908958434529, + "grad_norm": 1.3164199590682983, + "learning_rate": 5.409733412611917e-06, + "loss": 0.3868, + "step": 51763 + }, + { + "epoch": 3.591978349871626, + "grad_norm": 0.8938092589378357, + "learning_rate": 5.407910599148802e-06, + "loss": 0.1916, + "step": 51764 + }, + { + "epoch": 3.5920477413087224, + "grad_norm": 0.7207573652267456, + "learning_rate": 5.406088084301941e-06, + "loss": 0.2443, + "step": 51765 + }, + { + "epoch": 3.5921171327458192, + "grad_norm": 0.6586291790008545, + "learning_rate": 5.4042658680770964e-06, + "loss": 0.1947, + "step": 51766 + }, + { + "epoch": 3.5921865241829156, + "grad_norm": 1.2106720209121704, + "learning_rate": 5.402443950479996e-06, + "loss": 0.3574, + "step": 51767 + }, + { + "epoch": 3.5922559156200125, + "grad_norm": 0.6621008515357971, + "learning_rate": 5.400622331516425e-06, + "loss": 0.1381, + "step": 51768 + }, + { + "epoch": 3.5923253070571093, + "grad_norm": 1.1657651662826538, + "learning_rate": 5.3988010111921006e-06, + "loss": 0.4016, + "step": 51769 + }, + { + "epoch": 3.5923946984942057, + "grad_norm": 0.7024118304252625, + "learning_rate": 5.396979989512785e-06, + "loss": 0.1382, + "step": 51770 + }, + { + "epoch": 3.5924640899313025, + "grad_norm": 0.8301867842674255, + "learning_rate": 5.395159266484217e-06, + "loss": 0.2044, + "step": 51771 + }, + { + "epoch": 3.592533481368399, + "grad_norm": 0.942220151424408, + "learning_rate": 5.393338842112172e-06, + "loss": 0.2501, + "step": 51772 + }, + { + "epoch": 3.5926028728054957, + "grad_norm": 1.4151246547698975, + "learning_rate": 5.391518716402367e-06, + "loss": 0.2978, + "step": 51773 + }, + { + "epoch": 3.5926722642425926, + "grad_norm": 0.8169059753417969, + "learning_rate": 5.3896988893605615e-06, + "loss": 0.1877, + "step": 51774 + }, + { + "epoch": 3.5927416556796894, + "grad_norm": 0.730584442615509, + "learning_rate": 5.38787936099252e-06, + "loss": 0.2461, + "step": 51775 + }, + { + "epoch": 3.592811047116786, + "grad_norm": 0.8374143838882446, + "learning_rate": 5.38606013130395e-06, + "loss": 0.3023, + "step": 51776 + }, + { + "epoch": 3.5928804385538826, + "grad_norm": 0.8811444640159607, + "learning_rate": 5.384241200300611e-06, + "loss": 0.2247, + "step": 51777 + }, + { + "epoch": 3.592949829990979, + "grad_norm": 0.9055702686309814, + "learning_rate": 5.382422567988254e-06, + "loss": 0.2265, + "step": 51778 + }, + { + "epoch": 3.593019221428076, + "grad_norm": 1.2025190591812134, + "learning_rate": 5.380604234372599e-06, + "loss": 0.2338, + "step": 51779 + }, + { + "epoch": 3.5930886128651727, + "grad_norm": 0.8350955843925476, + "learning_rate": 5.378786199459396e-06, + "loss": 0.2774, + "step": 51780 + }, + { + "epoch": 3.593158004302269, + "grad_norm": 1.3619575500488281, + "learning_rate": 5.3769684632543945e-06, + "loss": 0.3309, + "step": 51781 + }, + { + "epoch": 3.593227395739366, + "grad_norm": 0.9869834780693054, + "learning_rate": 5.375151025763336e-06, + "loss": 0.2447, + "step": 51782 + }, + { + "epoch": 3.5932967871764623, + "grad_norm": 0.8798393607139587, + "learning_rate": 5.3733338869919384e-06, + "loss": 0.2126, + "step": 51783 + }, + { + "epoch": 3.593366178613559, + "grad_norm": 1.0147095918655396, + "learning_rate": 5.371517046945951e-06, + "loss": 0.3287, + "step": 51784 + }, + { + "epoch": 3.593435570050656, + "grad_norm": 0.9090687036514282, + "learning_rate": 5.369700505631092e-06, + "loss": 0.3063, + "step": 51785 + }, + { + "epoch": 3.5935049614877523, + "grad_norm": 0.9673483967781067, + "learning_rate": 5.367884263053113e-06, + "loss": 0.1856, + "step": 51786 + }, + { + "epoch": 3.593574352924849, + "grad_norm": 0.898426353931427, + "learning_rate": 5.3660683192177655e-06, + "loss": 0.2563, + "step": 51787 + }, + { + "epoch": 3.5936437443619456, + "grad_norm": 0.7684614062309265, + "learning_rate": 5.3642526741307434e-06, + "loss": 0.1903, + "step": 51788 + }, + { + "epoch": 3.5937131357990424, + "grad_norm": 1.1532304286956787, + "learning_rate": 5.3624373277977865e-06, + "loss": 0.343, + "step": 51789 + }, + { + "epoch": 3.5937825272361392, + "grad_norm": 1.1742433309555054, + "learning_rate": 5.360622280224648e-06, + "loss": 0.2698, + "step": 51790 + }, + { + "epoch": 3.5938519186732356, + "grad_norm": 0.6882551908493042, + "learning_rate": 5.358807531417054e-06, + "loss": 0.1866, + "step": 51791 + }, + { + "epoch": 3.5939213101103324, + "grad_norm": 1.0858781337738037, + "learning_rate": 5.356993081380701e-06, + "loss": 0.3581, + "step": 51792 + }, + { + "epoch": 3.593990701547429, + "grad_norm": 1.3525362014770508, + "learning_rate": 5.3551789301213514e-06, + "loss": 0.2893, + "step": 51793 + }, + { + "epoch": 3.5940600929845257, + "grad_norm": 0.5846958160400391, + "learning_rate": 5.353365077644723e-06, + "loss": 0.1278, + "step": 51794 + }, + { + "epoch": 3.5941294844216225, + "grad_norm": 1.2929372787475586, + "learning_rate": 5.351551523956544e-06, + "loss": 0.3695, + "step": 51795 + }, + { + "epoch": 3.594198875858719, + "grad_norm": 0.8230983018875122, + "learning_rate": 5.349738269062532e-06, + "loss": 0.1929, + "step": 51796 + }, + { + "epoch": 3.5942682672958157, + "grad_norm": 0.7714146971702576, + "learning_rate": 5.347925312968405e-06, + "loss": 0.1832, + "step": 51797 + }, + { + "epoch": 3.594337658732912, + "grad_norm": 0.6716261506080627, + "learning_rate": 5.3461126556799025e-06, + "loss": 0.1243, + "step": 51798 + }, + { + "epoch": 3.594407050170009, + "grad_norm": 0.956480860710144, + "learning_rate": 5.344300297202742e-06, + "loss": 0.3237, + "step": 51799 + }, + { + "epoch": 3.5944764416071058, + "grad_norm": 0.8612903356552124, + "learning_rate": 5.342488237542642e-06, + "loss": 0.3057, + "step": 51800 + }, + { + "epoch": 3.5945458330442026, + "grad_norm": 1.0438196659088135, + "learning_rate": 5.3406764767053194e-06, + "loss": 0.3231, + "step": 51801 + }, + { + "epoch": 3.594615224481299, + "grad_norm": 1.2986825704574585, + "learning_rate": 5.3388650146965145e-06, + "loss": 0.251, + "step": 51802 + }, + { + "epoch": 3.594684615918396, + "grad_norm": 1.0755770206451416, + "learning_rate": 5.337053851521923e-06, + "loss": 0.2848, + "step": 51803 + }, + { + "epoch": 3.594754007355492, + "grad_norm": 0.8097466826438904, + "learning_rate": 5.335242987187261e-06, + "loss": 0.2023, + "step": 51804 + }, + { + "epoch": 3.594823398792589, + "grad_norm": 0.9136380553245544, + "learning_rate": 5.333432421698259e-06, + "loss": 0.2723, + "step": 51805 + }, + { + "epoch": 3.594892790229686, + "grad_norm": 0.7776740789413452, + "learning_rate": 5.331622155060623e-06, + "loss": 0.2486, + "step": 51806 + }, + { + "epoch": 3.5949621816667823, + "grad_norm": 1.0661932229995728, + "learning_rate": 5.32981218728007e-06, + "loss": 0.3563, + "step": 51807 + }, + { + "epoch": 3.595031573103879, + "grad_norm": 1.101384162902832, + "learning_rate": 5.328002518362318e-06, + "loss": 0.3157, + "step": 51808 + }, + { + "epoch": 3.5951009645409755, + "grad_norm": 1.4667913913726807, + "learning_rate": 5.3261931483130965e-06, + "loss": 0.3238, + "step": 51809 + }, + { + "epoch": 3.5951703559780723, + "grad_norm": 1.237635612487793, + "learning_rate": 5.324384077138089e-06, + "loss": 0.2045, + "step": 51810 + }, + { + "epoch": 3.595239747415169, + "grad_norm": 0.7618380188941956, + "learning_rate": 5.322575304843003e-06, + "loss": 0.1766, + "step": 51811 + }, + { + "epoch": 3.5953091388522656, + "grad_norm": 1.2455977201461792, + "learning_rate": 5.320766831433577e-06, + "loss": 0.3927, + "step": 51812 + }, + { + "epoch": 3.5953785302893624, + "grad_norm": 0.6821718215942383, + "learning_rate": 5.318958656915496e-06, + "loss": 0.1898, + "step": 51813 + }, + { + "epoch": 3.5954479217264588, + "grad_norm": 0.7164047956466675, + "learning_rate": 5.3171507812944775e-06, + "loss": 0.2276, + "step": 51814 + }, + { + "epoch": 3.5955173131635556, + "grad_norm": 1.0211288928985596, + "learning_rate": 5.315343204576228e-06, + "loss": 0.2973, + "step": 51815 + }, + { + "epoch": 3.5955867046006524, + "grad_norm": 1.0963855981826782, + "learning_rate": 5.313535926766478e-06, + "loss": 0.2972, + "step": 51816 + }, + { + "epoch": 3.595656096037749, + "grad_norm": 0.9881361722946167, + "learning_rate": 5.311728947870887e-06, + "loss": 0.2503, + "step": 51817 + }, + { + "epoch": 3.5957254874748457, + "grad_norm": 0.9311528205871582, + "learning_rate": 5.309922267895184e-06, + "loss": 0.3947, + "step": 51818 + }, + { + "epoch": 3.595794878911942, + "grad_norm": 0.69172203540802, + "learning_rate": 5.3081158868450664e-06, + "loss": 0.1758, + "step": 51819 + }, + { + "epoch": 3.595864270349039, + "grad_norm": 1.2307336330413818, + "learning_rate": 5.30630980472625e-06, + "loss": 0.2744, + "step": 51820 + }, + { + "epoch": 3.5959336617861357, + "grad_norm": 1.0631499290466309, + "learning_rate": 5.304504021544421e-06, + "loss": 0.3027, + "step": 51821 + }, + { + "epoch": 3.596003053223232, + "grad_norm": 0.9301023483276367, + "learning_rate": 5.302698537305295e-06, + "loss": 0.1976, + "step": 51822 + }, + { + "epoch": 3.596072444660329, + "grad_norm": 0.7603663802146912, + "learning_rate": 5.300893352014558e-06, + "loss": 0.2674, + "step": 51823 + }, + { + "epoch": 3.5961418360974253, + "grad_norm": 0.6976187825202942, + "learning_rate": 5.299088465677904e-06, + "loss": 0.1842, + "step": 51824 + }, + { + "epoch": 3.596211227534522, + "grad_norm": 0.9993417263031006, + "learning_rate": 5.297283878301051e-06, + "loss": 0.2319, + "step": 51825 + }, + { + "epoch": 3.596280618971619, + "grad_norm": 0.9617129564285278, + "learning_rate": 5.295479589889696e-06, + "loss": 0.2191, + "step": 51826 + }, + { + "epoch": 3.596350010408716, + "grad_norm": 1.4176650047302246, + "learning_rate": 5.2936756004494995e-06, + "loss": 0.4057, + "step": 51827 + }, + { + "epoch": 3.596419401845812, + "grad_norm": 0.7547321915626526, + "learning_rate": 5.29187190998619e-06, + "loss": 0.2449, + "step": 51828 + }, + { + "epoch": 3.596488793282909, + "grad_norm": 0.8685699105262756, + "learning_rate": 5.290068518505453e-06, + "loss": 0.2016, + "step": 51829 + }, + { + "epoch": 3.5965581847200054, + "grad_norm": 0.7954232096672058, + "learning_rate": 5.288265426012995e-06, + "loss": 0.2589, + "step": 51830 + }, + { + "epoch": 3.5966275761571023, + "grad_norm": 0.7147279381752014, + "learning_rate": 5.286462632514489e-06, + "loss": 0.1312, + "step": 51831 + }, + { + "epoch": 3.596696967594199, + "grad_norm": 1.101159930229187, + "learning_rate": 5.284660138015618e-06, + "loss": 0.2754, + "step": 51832 + }, + { + "epoch": 3.5967663590312955, + "grad_norm": 0.9604930877685547, + "learning_rate": 5.28285794252209e-06, + "loss": 0.2478, + "step": 51833 + }, + { + "epoch": 3.5968357504683923, + "grad_norm": 0.6869267821311951, + "learning_rate": 5.28105604603959e-06, + "loss": 0.1616, + "step": 51834 + }, + { + "epoch": 3.5969051419054887, + "grad_norm": 0.8754156231880188, + "learning_rate": 5.279254448573812e-06, + "loss": 0.2059, + "step": 51835 + }, + { + "epoch": 3.5969745333425855, + "grad_norm": 0.8754040598869324, + "learning_rate": 5.27745315013044e-06, + "loss": 0.1971, + "step": 51836 + }, + { + "epoch": 3.5970439247796824, + "grad_norm": 0.8407045602798462, + "learning_rate": 5.2756521507151605e-06, + "loss": 0.2469, + "step": 51837 + }, + { + "epoch": 3.5971133162167788, + "grad_norm": 0.6706312298774719, + "learning_rate": 5.273851450333656e-06, + "loss": 0.159, + "step": 51838 + }, + { + "epoch": 3.5971827076538756, + "grad_norm": 0.9047030210494995, + "learning_rate": 5.2720510489916e-06, + "loss": 0.2183, + "step": 51839 + }, + { + "epoch": 3.597252099090972, + "grad_norm": 1.1003726720809937, + "learning_rate": 5.270250946694699e-06, + "loss": 0.373, + "step": 51840 + }, + { + "epoch": 3.597321490528069, + "grad_norm": 0.9815126657485962, + "learning_rate": 5.268451143448616e-06, + "loss": 0.3633, + "step": 51841 + }, + { + "epoch": 3.5973908819651657, + "grad_norm": 1.1517407894134521, + "learning_rate": 5.266651639259046e-06, + "loss": 0.3537, + "step": 51842 + }, + { + "epoch": 3.597460273402262, + "grad_norm": 0.8999436497688293, + "learning_rate": 5.264852434131662e-06, + "loss": 0.3538, + "step": 51843 + }, + { + "epoch": 3.597529664839359, + "grad_norm": 1.0714102983474731, + "learning_rate": 5.26305352807217e-06, + "loss": 0.3443, + "step": 51844 + }, + { + "epoch": 3.5975990562764553, + "grad_norm": 0.97496098279953, + "learning_rate": 5.2612549210862e-06, + "loss": 0.3326, + "step": 51845 + }, + { + "epoch": 3.597668447713552, + "grad_norm": 1.2883434295654297, + "learning_rate": 5.25945661317947e-06, + "loss": 0.3235, + "step": 51846 + }, + { + "epoch": 3.597737839150649, + "grad_norm": 0.8837651014328003, + "learning_rate": 5.25765860435764e-06, + "loss": 0.1936, + "step": 51847 + }, + { + "epoch": 3.5978072305877453, + "grad_norm": 0.9210836291313171, + "learning_rate": 5.255860894626386e-06, + "loss": 0.2167, + "step": 51848 + }, + { + "epoch": 3.597876622024842, + "grad_norm": 0.5658008456230164, + "learning_rate": 5.254063483991389e-06, + "loss": 0.1784, + "step": 51849 + }, + { + "epoch": 3.5979460134619385, + "grad_norm": 1.2251650094985962, + "learning_rate": 5.252266372458314e-06, + "loss": 0.3352, + "step": 51850 + }, + { + "epoch": 3.5980154048990354, + "grad_norm": 1.3825291395187378, + "learning_rate": 5.250469560032867e-06, + "loss": 0.265, + "step": 51851 + }, + { + "epoch": 3.598084796336132, + "grad_norm": 0.933195948600769, + "learning_rate": 5.248673046720676e-06, + "loss": 0.2105, + "step": 51852 + }, + { + "epoch": 3.598154187773229, + "grad_norm": 2.1508283615112305, + "learning_rate": 5.246876832527426e-06, + "loss": 0.3232, + "step": 51853 + }, + { + "epoch": 3.5982235792103254, + "grad_norm": 1.1664663553237915, + "learning_rate": 5.2450809174588e-06, + "loss": 0.276, + "step": 51854 + }, + { + "epoch": 3.5982929706474223, + "grad_norm": 0.7260521650314331, + "learning_rate": 5.243285301520451e-06, + "loss": 0.2439, + "step": 51855 + }, + { + "epoch": 3.5983623620845187, + "grad_norm": 0.9110891819000244, + "learning_rate": 5.241489984718062e-06, + "loss": 0.2205, + "step": 51856 + }, + { + "epoch": 3.5984317535216155, + "grad_norm": 0.9538052678108215, + "learning_rate": 5.239694967057307e-06, + "loss": 0.1995, + "step": 51857 + }, + { + "epoch": 3.5985011449587123, + "grad_norm": 0.9037191271781921, + "learning_rate": 5.237900248543837e-06, + "loss": 0.23, + "step": 51858 + }, + { + "epoch": 3.5985705363958087, + "grad_norm": 0.762636125087738, + "learning_rate": 5.236105829183302e-06, + "loss": 0.1757, + "step": 51859 + }, + { + "epoch": 3.5986399278329055, + "grad_norm": 0.8314512372016907, + "learning_rate": 5.234311708981421e-06, + "loss": 0.2324, + "step": 51860 + }, + { + "epoch": 3.598709319270002, + "grad_norm": 1.0833834409713745, + "learning_rate": 5.232517887943799e-06, + "loss": 0.2624, + "step": 51861 + }, + { + "epoch": 3.5987787107070988, + "grad_norm": 0.6640296578407288, + "learning_rate": 5.230724366076123e-06, + "loss": 0.162, + "step": 51862 + }, + { + "epoch": 3.5988481021441956, + "grad_norm": 0.7194384336471558, + "learning_rate": 5.228931143384052e-06, + "loss": 0.2145, + "step": 51863 + }, + { + "epoch": 3.598917493581292, + "grad_norm": 0.7698809504508972, + "learning_rate": 5.227138219873262e-06, + "loss": 0.2143, + "step": 51864 + }, + { + "epoch": 3.598986885018389, + "grad_norm": 0.6230709552764893, + "learning_rate": 5.225345595549403e-06, + "loss": 0.2012, + "step": 51865 + }, + { + "epoch": 3.599056276455485, + "grad_norm": 1.256913185119629, + "learning_rate": 5.223553270418114e-06, + "loss": 0.2524, + "step": 51866 + }, + { + "epoch": 3.599125667892582, + "grad_norm": 1.0415598154067993, + "learning_rate": 5.22176124448508e-06, + "loss": 0.3792, + "step": 51867 + }, + { + "epoch": 3.599195059329679, + "grad_norm": 0.809829831123352, + "learning_rate": 5.219969517755952e-06, + "loss": 0.2706, + "step": 51868 + }, + { + "epoch": 3.5992644507667753, + "grad_norm": 1.1751891374588013, + "learning_rate": 5.218178090236381e-06, + "loss": 0.2837, + "step": 51869 + }, + { + "epoch": 3.599333842203872, + "grad_norm": 0.9700779914855957, + "learning_rate": 5.2163869619320185e-06, + "loss": 0.2751, + "step": 51870 + }, + { + "epoch": 3.5994032336409685, + "grad_norm": 0.8305085897445679, + "learning_rate": 5.214596132848548e-06, + "loss": 0.2706, + "step": 51871 + }, + { + "epoch": 3.5994726250780653, + "grad_norm": 1.126739740371704, + "learning_rate": 5.2128056029915775e-06, + "loss": 0.2468, + "step": 51872 + }, + { + "epoch": 3.599542016515162, + "grad_norm": 0.9157993793487549, + "learning_rate": 5.21101537236679e-06, + "loss": 0.1977, + "step": 51873 + }, + { + "epoch": 3.5996114079522585, + "grad_norm": 0.8829342126846313, + "learning_rate": 5.209225440979826e-06, + "loss": 0.2281, + "step": 51874 + }, + { + "epoch": 3.5996807993893554, + "grad_norm": 1.2609528303146362, + "learning_rate": 5.207435808836336e-06, + "loss": 0.2581, + "step": 51875 + }, + { + "epoch": 3.5997501908264518, + "grad_norm": 1.1875238418579102, + "learning_rate": 5.2056464759419835e-06, + "loss": 0.3252, + "step": 51876 + }, + { + "epoch": 3.5998195822635486, + "grad_norm": 0.8472784161567688, + "learning_rate": 5.203857442302396e-06, + "loss": 0.2604, + "step": 51877 + }, + { + "epoch": 3.5998889737006454, + "grad_norm": 1.1438322067260742, + "learning_rate": 5.202068707923258e-06, + "loss": 0.3763, + "step": 51878 + }, + { + "epoch": 3.5999583651377423, + "grad_norm": 0.7962341904640198, + "learning_rate": 5.2002802728101654e-06, + "loss": 0.2375, + "step": 51879 + }, + { + "epoch": 3.6000277565748386, + "grad_norm": 0.9085586667060852, + "learning_rate": 5.198492136968791e-06, + "loss": 0.2297, + "step": 51880 + }, + { + "epoch": 3.6000971480119355, + "grad_norm": 0.6422693133354187, + "learning_rate": 5.1967043004047865e-06, + "loss": 0.2426, + "step": 51881 + }, + { + "epoch": 3.600166539449032, + "grad_norm": 1.2682737112045288, + "learning_rate": 5.19491676312378e-06, + "loss": 0.3849, + "step": 51882 + }, + { + "epoch": 3.6002359308861287, + "grad_norm": 0.7615280151367188, + "learning_rate": 5.193129525131424e-06, + "loss": 0.2602, + "step": 51883 + }, + { + "epoch": 3.6003053223232255, + "grad_norm": 0.7528914213180542, + "learning_rate": 5.191342586433356e-06, + "loss": 0.1591, + "step": 51884 + }, + { + "epoch": 3.600374713760322, + "grad_norm": 0.904135525226593, + "learning_rate": 5.189555947035241e-06, + "loss": 0.2465, + "step": 51885 + }, + { + "epoch": 3.6004441051974188, + "grad_norm": 1.0229905843734741, + "learning_rate": 5.187769606942672e-06, + "loss": 0.292, + "step": 51886 + }, + { + "epoch": 3.600513496634515, + "grad_norm": 0.8640795946121216, + "learning_rate": 5.185983566161323e-06, + "loss": 0.3482, + "step": 51887 + }, + { + "epoch": 3.600582888071612, + "grad_norm": 0.8944249153137207, + "learning_rate": 5.1841978246968235e-06, + "loss": 0.2779, + "step": 51888 + }, + { + "epoch": 3.600652279508709, + "grad_norm": 0.9422365427017212, + "learning_rate": 5.182412382554802e-06, + "loss": 0.3587, + "step": 51889 + }, + { + "epoch": 3.600721670945805, + "grad_norm": 1.2487868070602417, + "learning_rate": 5.180627239740909e-06, + "loss": 0.2706, + "step": 51890 + }, + { + "epoch": 3.600791062382902, + "grad_norm": 1.2079806327819824, + "learning_rate": 5.178842396260775e-06, + "loss": 0.3078, + "step": 51891 + }, + { + "epoch": 3.6008604538199984, + "grad_norm": 0.768135130405426, + "learning_rate": 5.177057852120048e-06, + "loss": 0.2079, + "step": 51892 + }, + { + "epoch": 3.6009298452570953, + "grad_norm": 0.9414743185043335, + "learning_rate": 5.175273607324338e-06, + "loss": 0.2376, + "step": 51893 + }, + { + "epoch": 3.600999236694192, + "grad_norm": 1.3015936613082886, + "learning_rate": 5.173489661879283e-06, + "loss": 0.3435, + "step": 51894 + }, + { + "epoch": 3.6010686281312885, + "grad_norm": 0.7563140988349915, + "learning_rate": 5.171706015790534e-06, + "loss": 0.2057, + "step": 51895 + }, + { + "epoch": 3.6011380195683853, + "grad_norm": 0.8635666370391846, + "learning_rate": 5.169922669063698e-06, + "loss": 0.2877, + "step": 51896 + }, + { + "epoch": 3.6012074110054817, + "grad_norm": 0.5636315941810608, + "learning_rate": 5.168139621704405e-06, + "loss": 0.1758, + "step": 51897 + }, + { + "epoch": 3.6012768024425785, + "grad_norm": 0.6959212422370911, + "learning_rate": 5.166356873718303e-06, + "loss": 0.1955, + "step": 51898 + }, + { + "epoch": 3.6013461938796754, + "grad_norm": 0.9093437790870667, + "learning_rate": 5.1645744251110125e-06, + "loss": 0.2969, + "step": 51899 + }, + { + "epoch": 3.6014155853167718, + "grad_norm": 0.5443366169929504, + "learning_rate": 5.1627922758881504e-06, + "loss": 0.0909, + "step": 51900 + }, + { + "epoch": 3.6014849767538686, + "grad_norm": 0.9442828893661499, + "learning_rate": 5.161010426055346e-06, + "loss": 0.2032, + "step": 51901 + }, + { + "epoch": 3.601554368190965, + "grad_norm": 0.6686946153640747, + "learning_rate": 5.159228875618227e-06, + "loss": 0.1444, + "step": 51902 + }, + { + "epoch": 3.601623759628062, + "grad_norm": 1.1806325912475586, + "learning_rate": 5.157447624582423e-06, + "loss": 0.3233, + "step": 51903 + }, + { + "epoch": 3.6016931510651586, + "grad_norm": 0.7764715552330017, + "learning_rate": 5.155666672953552e-06, + "loss": 0.1923, + "step": 51904 + }, + { + "epoch": 3.6017625425022555, + "grad_norm": 1.4504212141036987, + "learning_rate": 5.153886020737242e-06, + "loss": 0.3852, + "step": 51905 + }, + { + "epoch": 3.601831933939352, + "grad_norm": 1.1574742794036865, + "learning_rate": 5.152105667939122e-06, + "loss": 0.2624, + "step": 51906 + }, + { + "epoch": 3.6019013253764487, + "grad_norm": 0.7912337183952332, + "learning_rate": 5.1503256145647774e-06, + "loss": 0.2389, + "step": 51907 + }, + { + "epoch": 3.601970716813545, + "grad_norm": 0.9482065439224243, + "learning_rate": 5.148545860619858e-06, + "loss": 0.2337, + "step": 51908 + }, + { + "epoch": 3.602040108250642, + "grad_norm": 0.848294734954834, + "learning_rate": 5.146766406109971e-06, + "loss": 0.2138, + "step": 51909 + }, + { + "epoch": 3.6021094996877387, + "grad_norm": 0.8275083303451538, + "learning_rate": 5.144987251040745e-06, + "loss": 0.2108, + "step": 51910 + }, + { + "epoch": 3.602178891124835, + "grad_norm": 1.1482902765274048, + "learning_rate": 5.1432083954177755e-06, + "loss": 0.2433, + "step": 51911 + }, + { + "epoch": 3.602248282561932, + "grad_norm": 0.9904579520225525, + "learning_rate": 5.141429839246703e-06, + "loss": 0.2782, + "step": 51912 + }, + { + "epoch": 3.6023176739990284, + "grad_norm": 0.9695112705230713, + "learning_rate": 5.1396515825331446e-06, + "loss": 0.233, + "step": 51913 + }, + { + "epoch": 3.602387065436125, + "grad_norm": 0.9151971340179443, + "learning_rate": 5.137873625282685e-06, + "loss": 0.2398, + "step": 51914 + }, + { + "epoch": 3.602456456873222, + "grad_norm": 0.9755166172981262, + "learning_rate": 5.136095967500942e-06, + "loss": 0.2311, + "step": 51915 + }, + { + "epoch": 3.6025258483103184, + "grad_norm": 0.8132280111312866, + "learning_rate": 5.134318609193556e-06, + "loss": 0.1565, + "step": 51916 + }, + { + "epoch": 3.6025952397474152, + "grad_norm": 1.301347017288208, + "learning_rate": 5.132541550366099e-06, + "loss": 0.2818, + "step": 51917 + }, + { + "epoch": 3.6026646311845116, + "grad_norm": 1.0041615962982178, + "learning_rate": 5.1307647910242225e-06, + "loss": 0.1708, + "step": 51918 + }, + { + "epoch": 3.6027340226216085, + "grad_norm": 0.6958204507827759, + "learning_rate": 5.128988331173501e-06, + "loss": 0.2081, + "step": 51919 + }, + { + "epoch": 3.6028034140587053, + "grad_norm": 0.9365288019180298, + "learning_rate": 5.127212170819573e-06, + "loss": 0.2776, + "step": 51920 + }, + { + "epoch": 3.6028728054958017, + "grad_norm": 0.9067739844322205, + "learning_rate": 5.1254363099680125e-06, + "loss": 0.2481, + "step": 51921 + }, + { + "epoch": 3.6029421969328985, + "grad_norm": 0.8187549710273743, + "learning_rate": 5.123660748624448e-06, + "loss": 0.2296, + "step": 51922 + }, + { + "epoch": 3.603011588369995, + "grad_norm": 1.2092105150222778, + "learning_rate": 5.1218854867944755e-06, + "loss": 0.2587, + "step": 51923 + }, + { + "epoch": 3.6030809798070917, + "grad_norm": 0.7484418153762817, + "learning_rate": 5.1201105244837015e-06, + "loss": 0.1872, + "step": 51924 + }, + { + "epoch": 3.6031503712441886, + "grad_norm": 0.8700698018074036, + "learning_rate": 5.118335861697732e-06, + "loss": 0.2226, + "step": 51925 + }, + { + "epoch": 3.603219762681285, + "grad_norm": 0.9519550800323486, + "learning_rate": 5.116561498442185e-06, + "loss": 0.2838, + "step": 51926 + }, + { + "epoch": 3.603289154118382, + "grad_norm": 0.5798589587211609, + "learning_rate": 5.114787434722623e-06, + "loss": 0.1091, + "step": 51927 + }, + { + "epoch": 3.603358545555478, + "grad_norm": 1.1662843227386475, + "learning_rate": 5.113013670544675e-06, + "loss": 0.251, + "step": 51928 + }, + { + "epoch": 3.603427936992575, + "grad_norm": 0.8722065091133118, + "learning_rate": 5.111240205913936e-06, + "loss": 0.2519, + "step": 51929 + }, + { + "epoch": 3.603497328429672, + "grad_norm": 0.7342366576194763, + "learning_rate": 5.109467040836014e-06, + "loss": 0.2688, + "step": 51930 + }, + { + "epoch": 3.6035667198667687, + "grad_norm": 0.7879807949066162, + "learning_rate": 5.1076941753164794e-06, + "loss": 0.2245, + "step": 51931 + }, + { + "epoch": 3.603636111303865, + "grad_norm": 0.9835085272789001, + "learning_rate": 5.1059216093609415e-06, + "loss": 0.3016, + "step": 51932 + }, + { + "epoch": 3.603705502740962, + "grad_norm": 0.9408406615257263, + "learning_rate": 5.1041493429750065e-06, + "loss": 0.2834, + "step": 51933 + }, + { + "epoch": 3.6037748941780583, + "grad_norm": 0.7642804384231567, + "learning_rate": 5.102377376164269e-06, + "loss": 0.201, + "step": 51934 + }, + { + "epoch": 3.603844285615155, + "grad_norm": 0.7312577962875366, + "learning_rate": 5.100605708934303e-06, + "loss": 0.2448, + "step": 51935 + }, + { + "epoch": 3.603913677052252, + "grad_norm": 0.7543826699256897, + "learning_rate": 5.098834341290726e-06, + "loss": 0.2696, + "step": 51936 + }, + { + "epoch": 3.6039830684893484, + "grad_norm": 1.0968670845031738, + "learning_rate": 5.0970632732390996e-06, + "loss": 0.2712, + "step": 51937 + }, + { + "epoch": 3.604052459926445, + "grad_norm": 1.0202828645706177, + "learning_rate": 5.095292504785054e-06, + "loss": 0.1915, + "step": 51938 + }, + { + "epoch": 3.6041218513635416, + "grad_norm": 0.730530858039856, + "learning_rate": 5.093522035934139e-06, + "loss": 0.2284, + "step": 51939 + }, + { + "epoch": 3.6041912428006384, + "grad_norm": 0.8072035312652588, + "learning_rate": 5.091751866691996e-06, + "loss": 0.2264, + "step": 51940 + }, + { + "epoch": 3.6042606342377352, + "grad_norm": 1.2223434448242188, + "learning_rate": 5.089981997064153e-06, + "loss": 0.2561, + "step": 51941 + }, + { + "epoch": 3.6043300256748316, + "grad_norm": 0.9566478729248047, + "learning_rate": 5.088212427056227e-06, + "loss": 0.3138, + "step": 51942 + }, + { + "epoch": 3.6043994171119285, + "grad_norm": 0.5976535677909851, + "learning_rate": 5.086443156673814e-06, + "loss": 0.2061, + "step": 51943 + }, + { + "epoch": 3.604468808549025, + "grad_norm": 0.9484288692474365, + "learning_rate": 5.084674185922478e-06, + "loss": 0.3114, + "step": 51944 + }, + { + "epoch": 3.6045381999861217, + "grad_norm": 0.6893547773361206, + "learning_rate": 5.082905514807812e-06, + "loss": 0.1769, + "step": 51945 + }, + { + "epoch": 3.6046075914232185, + "grad_norm": 0.8207785487174988, + "learning_rate": 5.0811371433354014e-06, + "loss": 0.1754, + "step": 51946 + }, + { + "epoch": 3.604676982860315, + "grad_norm": 0.8865926265716553, + "learning_rate": 5.079369071510853e-06, + "loss": 0.1966, + "step": 51947 + }, + { + "epoch": 3.6047463742974117, + "grad_norm": 1.1466437578201294, + "learning_rate": 5.077601299339696e-06, + "loss": 0.5381, + "step": 51948 + }, + { + "epoch": 3.604815765734508, + "grad_norm": 1.0289969444274902, + "learning_rate": 5.075833826827547e-06, + "loss": 0.328, + "step": 51949 + }, + { + "epoch": 3.604885157171605, + "grad_norm": 0.6434247493743896, + "learning_rate": 5.07406665397997e-06, + "loss": 0.1602, + "step": 51950 + }, + { + "epoch": 3.604954548608702, + "grad_norm": 1.1727639436721802, + "learning_rate": 5.072299780802559e-06, + "loss": 0.2261, + "step": 51951 + }, + { + "epoch": 3.6050239400457986, + "grad_norm": 0.7529560327529907, + "learning_rate": 5.070533207300887e-06, + "loss": 0.1864, + "step": 51952 + }, + { + "epoch": 3.605093331482895, + "grad_norm": 0.8786073327064514, + "learning_rate": 5.0687669334805174e-06, + "loss": 0.2061, + "step": 51953 + }, + { + "epoch": 3.6051627229199914, + "grad_norm": 1.0473430156707764, + "learning_rate": 5.067000959347057e-06, + "loss": 0.3019, + "step": 51954 + }, + { + "epoch": 3.6052321143570882, + "grad_norm": 0.6336240768432617, + "learning_rate": 5.065235284906045e-06, + "loss": 0.1588, + "step": 51955 + }, + { + "epoch": 3.605301505794185, + "grad_norm": 0.9647603034973145, + "learning_rate": 5.063469910163066e-06, + "loss": 0.1725, + "step": 51956 + }, + { + "epoch": 3.605370897231282, + "grad_norm": 0.7894446849822998, + "learning_rate": 5.061704835123704e-06, + "loss": 0.2744, + "step": 51957 + }, + { + "epoch": 3.6054402886683783, + "grad_norm": 0.8694725632667542, + "learning_rate": 5.059940059793522e-06, + "loss": 0.2822, + "step": 51958 + }, + { + "epoch": 3.605509680105475, + "grad_norm": 0.6912366151809692, + "learning_rate": 5.058175584178093e-06, + "loss": 0.1368, + "step": 51959 + }, + { + "epoch": 3.6055790715425715, + "grad_norm": 0.7551513314247131, + "learning_rate": 5.056411408282991e-06, + "loss": 0.2109, + "step": 51960 + }, + { + "epoch": 3.6056484629796683, + "grad_norm": 1.1697168350219727, + "learning_rate": 5.054647532113799e-06, + "loss": 0.2957, + "step": 51961 + }, + { + "epoch": 3.605717854416765, + "grad_norm": 0.8202170729637146, + "learning_rate": 5.0528839556760465e-06, + "loss": 0.1903, + "step": 51962 + }, + { + "epoch": 3.6057872458538616, + "grad_norm": 1.0804119110107422, + "learning_rate": 5.051120678975329e-06, + "loss": 0.2206, + "step": 51963 + }, + { + "epoch": 3.6058566372909584, + "grad_norm": 0.751274824142456, + "learning_rate": 5.04935770201721e-06, + "loss": 0.261, + "step": 51964 + }, + { + "epoch": 3.605926028728055, + "grad_norm": 1.706539273262024, + "learning_rate": 5.047595024807261e-06, + "loss": 0.4616, + "step": 51965 + }, + { + "epoch": 3.6059954201651516, + "grad_norm": 1.0237548351287842, + "learning_rate": 5.045832647351023e-06, + "loss": 0.2519, + "step": 51966 + }, + { + "epoch": 3.6060648116022485, + "grad_norm": 1.189072608947754, + "learning_rate": 5.0440705696540805e-06, + "loss": 0.4097, + "step": 51967 + }, + { + "epoch": 3.606134203039345, + "grad_norm": 0.8734267950057983, + "learning_rate": 5.042308791721983e-06, + "loss": 0.2935, + "step": 51968 + }, + { + "epoch": 3.6062035944764417, + "grad_norm": 0.9757077693939209, + "learning_rate": 5.040547313560318e-06, + "loss": 0.2236, + "step": 51969 + }, + { + "epoch": 3.606272985913538, + "grad_norm": 0.6234020590782166, + "learning_rate": 5.038786135174623e-06, + "loss": 0.2109, + "step": 51970 + }, + { + "epoch": 3.606342377350635, + "grad_norm": 0.7601417303085327, + "learning_rate": 5.037025256570449e-06, + "loss": 0.1952, + "step": 51971 + }, + { + "epoch": 3.6064117687877317, + "grad_norm": 0.9534751772880554, + "learning_rate": 5.035264677753371e-06, + "loss": 0.2293, + "step": 51972 + }, + { + "epoch": 3.606481160224828, + "grad_norm": 1.153911828994751, + "learning_rate": 5.03350439872895e-06, + "loss": 0.2728, + "step": 51973 + }, + { + "epoch": 3.606550551661925, + "grad_norm": 1.277953028678894, + "learning_rate": 5.031744419502738e-06, + "loss": 0.373, + "step": 51974 + }, + { + "epoch": 3.6066199430990213, + "grad_norm": 1.0570687055587769, + "learning_rate": 5.029984740080296e-06, + "loss": 0.2634, + "step": 51975 + }, + { + "epoch": 3.606689334536118, + "grad_norm": 0.9929234385490417, + "learning_rate": 5.028225360467165e-06, + "loss": 0.2598, + "step": 51976 + }, + { + "epoch": 3.606758725973215, + "grad_norm": 1.020709753036499, + "learning_rate": 5.026466280668907e-06, + "loss": 0.3241, + "step": 51977 + }, + { + "epoch": 3.606828117410312, + "grad_norm": 0.7863398194313049, + "learning_rate": 5.024707500691073e-06, + "loss": 0.2296, + "step": 51978 + }, + { + "epoch": 3.6068975088474082, + "grad_norm": 0.7806274890899658, + "learning_rate": 5.022949020539225e-06, + "loss": 0.1873, + "step": 51979 + }, + { + "epoch": 3.6069669002845046, + "grad_norm": 0.9869720935821533, + "learning_rate": 5.021190840218903e-06, + "loss": 0.167, + "step": 51980 + }, + { + "epoch": 3.6070362917216015, + "grad_norm": 0.9136248826980591, + "learning_rate": 5.0194329597356595e-06, + "loss": 0.2957, + "step": 51981 + }, + { + "epoch": 3.6071056831586983, + "grad_norm": 0.5824637413024902, + "learning_rate": 5.017675379095066e-06, + "loss": 0.1717, + "step": 51982 + }, + { + "epoch": 3.607175074595795, + "grad_norm": 1.261392593383789, + "learning_rate": 5.015918098302641e-06, + "loss": 0.2897, + "step": 51983 + }, + { + "epoch": 3.6072444660328915, + "grad_norm": 2.1382596492767334, + "learning_rate": 5.014161117363936e-06, + "loss": 0.4093, + "step": 51984 + }, + { + "epoch": 3.6073138574699883, + "grad_norm": 1.0973138809204102, + "learning_rate": 5.0124044362845015e-06, + "loss": 0.2831, + "step": 51985 + }, + { + "epoch": 3.6073832489070847, + "grad_norm": 0.8108941912651062, + "learning_rate": 5.010648055069889e-06, + "loss": 0.1823, + "step": 51986 + }, + { + "epoch": 3.6074526403441816, + "grad_norm": 0.7957366704940796, + "learning_rate": 5.008891973725638e-06, + "loss": 0.2451, + "step": 51987 + }, + { + "epoch": 3.6075220317812784, + "grad_norm": 0.7720680832862854, + "learning_rate": 5.0071361922573005e-06, + "loss": 0.1568, + "step": 51988 + }, + { + "epoch": 3.607591423218375, + "grad_norm": 1.0059818029403687, + "learning_rate": 5.005380710670416e-06, + "loss": 0.2776, + "step": 51989 + }, + { + "epoch": 3.6076608146554716, + "grad_norm": 0.4654051959514618, + "learning_rate": 5.003625528970513e-06, + "loss": 0.1439, + "step": 51990 + }, + { + "epoch": 3.607730206092568, + "grad_norm": 1.412567377090454, + "learning_rate": 5.001870647163143e-06, + "loss": 0.3191, + "step": 51991 + }, + { + "epoch": 3.607799597529665, + "grad_norm": 1.1393779516220093, + "learning_rate": 5.000116065253857e-06, + "loss": 0.3008, + "step": 51992 + }, + { + "epoch": 3.6078689889667617, + "grad_norm": 1.091982126235962, + "learning_rate": 4.998361783248162e-06, + "loss": 0.2884, + "step": 51993 + }, + { + "epoch": 3.607938380403858, + "grad_norm": 1.105528712272644, + "learning_rate": 4.996607801151631e-06, + "loss": 0.2221, + "step": 51994 + }, + { + "epoch": 3.608007771840955, + "grad_norm": 0.789711594581604, + "learning_rate": 4.9948541189697825e-06, + "loss": 0.2174, + "step": 51995 + }, + { + "epoch": 3.6080771632780513, + "grad_norm": 0.7822428941726685, + "learning_rate": 4.993100736708179e-06, + "loss": 0.2279, + "step": 51996 + }, + { + "epoch": 3.608146554715148, + "grad_norm": 1.16972017288208, + "learning_rate": 4.9913476543723026e-06, + "loss": 0.239, + "step": 51997 + }, + { + "epoch": 3.608215946152245, + "grad_norm": 0.7713211178779602, + "learning_rate": 4.9895948719677286e-06, + "loss": 0.2993, + "step": 51998 + }, + { + "epoch": 3.6082853375893413, + "grad_norm": 1.0984848737716675, + "learning_rate": 4.987842389499986e-06, + "loss": 0.2116, + "step": 51999 + }, + { + "epoch": 3.608354729026438, + "grad_norm": 1.3086282014846802, + "learning_rate": 4.986090206974603e-06, + "loss": 0.2731, + "step": 52000 + }, + { + "epoch": 3.608354729026438, + "eval_loss": 0.4687287211418152, + "eval_runtime": 3827.9228, + "eval_samples_per_second": 6.341, + "eval_steps_per_second": 0.793, + "step": 52000 + }, + { + "epoch": 3.6084241204635346, + "grad_norm": 0.8029067516326904, + "learning_rate": 4.984338324397109e-06, + "loss": 0.1867, + "step": 52001 + }, + { + "epoch": 3.6084935119006314, + "grad_norm": 1.1279412508010864, + "learning_rate": 4.982586741773021e-06, + "loss": 0.345, + "step": 52002 + }, + { + "epoch": 3.6085629033377282, + "grad_norm": 0.890004575252533, + "learning_rate": 4.9808354591079025e-06, + "loss": 0.2936, + "step": 52003 + }, + { + "epoch": 3.608632294774825, + "grad_norm": 1.545305848121643, + "learning_rate": 4.979084476407248e-06, + "loss": 0.2274, + "step": 52004 + }, + { + "epoch": 3.6087016862119214, + "grad_norm": 0.8350717425346375, + "learning_rate": 4.977333793676608e-06, + "loss": 0.2145, + "step": 52005 + }, + { + "epoch": 3.6087710776490183, + "grad_norm": 0.8323938250541687, + "learning_rate": 4.975583410921491e-06, + "loss": 0.2241, + "step": 52006 + }, + { + "epoch": 3.6088404690861147, + "grad_norm": 1.05226731300354, + "learning_rate": 4.973833328147437e-06, + "loss": 0.2347, + "step": 52007 + }, + { + "epoch": 3.6089098605232115, + "grad_norm": 0.7743309736251831, + "learning_rate": 4.972083545359962e-06, + "loss": 0.3505, + "step": 52008 + }, + { + "epoch": 3.6089792519603083, + "grad_norm": 0.8246489763259888, + "learning_rate": 4.970334062564596e-06, + "loss": 0.2176, + "step": 52009 + }, + { + "epoch": 3.6090486433974047, + "grad_norm": 1.0036951303482056, + "learning_rate": 4.968584879766869e-06, + "loss": 0.2989, + "step": 52010 + }, + { + "epoch": 3.6091180348345016, + "grad_norm": 0.9305649995803833, + "learning_rate": 4.9668359969722856e-06, + "loss": 0.2222, + "step": 52011 + }, + { + "epoch": 3.609187426271598, + "grad_norm": 1.3850749731063843, + "learning_rate": 4.965087414186376e-06, + "loss": 0.3814, + "step": 52012 + }, + { + "epoch": 3.6092568177086948, + "grad_norm": 1.2020469903945923, + "learning_rate": 4.9633391314146575e-06, + "loss": 0.3549, + "step": 52013 + }, + { + "epoch": 3.6093262091457916, + "grad_norm": 0.7255478501319885, + "learning_rate": 4.961591148662648e-06, + "loss": 0.182, + "step": 52014 + }, + { + "epoch": 3.609395600582888, + "grad_norm": 1.3415676355361938, + "learning_rate": 4.959843465935865e-06, + "loss": 0.2608, + "step": 52015 + }, + { + "epoch": 3.609464992019985, + "grad_norm": 0.8762152194976807, + "learning_rate": 4.9580960832398495e-06, + "loss": 0.1473, + "step": 52016 + }, + { + "epoch": 3.609534383457081, + "grad_norm": 0.8332724571228027, + "learning_rate": 4.956349000580085e-06, + "loss": 0.1906, + "step": 52017 + }, + { + "epoch": 3.609603774894178, + "grad_norm": 0.7063843011856079, + "learning_rate": 4.954602217962101e-06, + "loss": 0.1852, + "step": 52018 + }, + { + "epoch": 3.609673166331275, + "grad_norm": 0.8847043514251709, + "learning_rate": 4.952855735391415e-06, + "loss": 0.2865, + "step": 52019 + }, + { + "epoch": 3.6097425577683713, + "grad_norm": 0.6348775029182434, + "learning_rate": 4.951109552873523e-06, + "loss": 0.1515, + "step": 52020 + }, + { + "epoch": 3.609811949205468, + "grad_norm": 0.9409544467926025, + "learning_rate": 4.949363670413954e-06, + "loss": 0.3169, + "step": 52021 + }, + { + "epoch": 3.6098813406425645, + "grad_norm": 0.7660780549049377, + "learning_rate": 4.947618088018224e-06, + "loss": 0.212, + "step": 52022 + }, + { + "epoch": 3.6099507320796613, + "grad_norm": 0.9828561544418335, + "learning_rate": 4.945872805691843e-06, + "loss": 0.3062, + "step": 52023 + }, + { + "epoch": 3.610020123516758, + "grad_norm": 0.5180569887161255, + "learning_rate": 4.9441278234403034e-06, + "loss": 0.1609, + "step": 52024 + }, + { + "epoch": 3.6100895149538546, + "grad_norm": 0.8971554040908813, + "learning_rate": 4.942383141269113e-06, + "loss": 0.2048, + "step": 52025 + }, + { + "epoch": 3.6101589063909514, + "grad_norm": 1.1119141578674316, + "learning_rate": 4.940638759183802e-06, + "loss": 0.3032, + "step": 52026 + }, + { + "epoch": 3.6102282978280478, + "grad_norm": 1.238294005393982, + "learning_rate": 4.9388946771898645e-06, + "loss": 0.3796, + "step": 52027 + }, + { + "epoch": 3.6102976892651446, + "grad_norm": 0.5689239501953125, + "learning_rate": 4.937150895292797e-06, + "loss": 0.1961, + "step": 52028 + }, + { + "epoch": 3.6103670807022414, + "grad_norm": 1.0862475633621216, + "learning_rate": 4.935407413498127e-06, + "loss": 0.2519, + "step": 52029 + }, + { + "epoch": 3.6104364721393383, + "grad_norm": 0.8431236147880554, + "learning_rate": 4.933664231811353e-06, + "loss": 0.2777, + "step": 52030 + }, + { + "epoch": 3.6105058635764347, + "grad_norm": 0.623072624206543, + "learning_rate": 4.931921350237956e-06, + "loss": 0.1529, + "step": 52031 + }, + { + "epoch": 3.6105752550135315, + "grad_norm": 0.7911587953567505, + "learning_rate": 4.930178768783455e-06, + "loss": 0.151, + "step": 52032 + }, + { + "epoch": 3.610644646450628, + "grad_norm": 0.7402799725532532, + "learning_rate": 4.928436487453358e-06, + "loss": 0.2183, + "step": 52033 + }, + { + "epoch": 3.6107140378877247, + "grad_norm": 1.5916420221328735, + "learning_rate": 4.9266945062531485e-06, + "loss": 0.4039, + "step": 52034 + }, + { + "epoch": 3.6107834293248215, + "grad_norm": 0.6008829474449158, + "learning_rate": 4.924952825188345e-06, + "loss": 0.1575, + "step": 52035 + }, + { + "epoch": 3.610852820761918, + "grad_norm": 1.2749354839324951, + "learning_rate": 4.92321144426443e-06, + "loss": 0.2423, + "step": 52036 + }, + { + "epoch": 3.6109222121990148, + "grad_norm": 0.9342559576034546, + "learning_rate": 4.921470363486902e-06, + "loss": 0.2463, + "step": 52037 + }, + { + "epoch": 3.610991603636111, + "grad_norm": 1.1089743375778198, + "learning_rate": 4.919729582861276e-06, + "loss": 0.2628, + "step": 52038 + }, + { + "epoch": 3.611060995073208, + "grad_norm": 1.0520826578140259, + "learning_rate": 4.917989102393017e-06, + "loss": 0.3647, + "step": 52039 + }, + { + "epoch": 3.611130386510305, + "grad_norm": 1.0536736249923706, + "learning_rate": 4.91624892208763e-06, + "loss": 0.315, + "step": 52040 + }, + { + "epoch": 3.611199777947401, + "grad_norm": 0.8305380940437317, + "learning_rate": 4.914509041950621e-06, + "loss": 0.1756, + "step": 52041 + }, + { + "epoch": 3.611269169384498, + "grad_norm": 0.8791204690933228, + "learning_rate": 4.912769461987465e-06, + "loss": 0.2176, + "step": 52042 + }, + { + "epoch": 3.6113385608215944, + "grad_norm": 0.8543229699134827, + "learning_rate": 4.911030182203669e-06, + "loss": 0.2582, + "step": 52043 + }, + { + "epoch": 3.6114079522586913, + "grad_norm": 0.8270686268806458, + "learning_rate": 4.909291202604738e-06, + "loss": 0.2206, + "step": 52044 + }, + { + "epoch": 3.611477343695788, + "grad_norm": 0.8331142067909241, + "learning_rate": 4.907552523196124e-06, + "loss": 0.2141, + "step": 52045 + }, + { + "epoch": 3.6115467351328845, + "grad_norm": 1.1341875791549683, + "learning_rate": 4.905814143983334e-06, + "loss": 0.2708, + "step": 52046 + }, + { + "epoch": 3.6116161265699813, + "grad_norm": 0.9156262278556824, + "learning_rate": 4.904076064971852e-06, + "loss": 0.1884, + "step": 52047 + }, + { + "epoch": 3.6116855180070777, + "grad_norm": 0.8184095025062561, + "learning_rate": 4.9023382861671745e-06, + "loss": 0.2357, + "step": 52048 + }, + { + "epoch": 3.6117549094441745, + "grad_norm": 0.8458063006401062, + "learning_rate": 4.900600807574773e-06, + "loss": 0.1925, + "step": 52049 + }, + { + "epoch": 3.6118243008812714, + "grad_norm": 0.8891326189041138, + "learning_rate": 4.898863629200145e-06, + "loss": 0.2675, + "step": 52050 + }, + { + "epoch": 3.6118936923183678, + "grad_norm": 0.5857526063919067, + "learning_rate": 4.897126751048786e-06, + "loss": 0.1376, + "step": 52051 + }, + { + "epoch": 3.6119630837554646, + "grad_norm": 1.2939077615737915, + "learning_rate": 4.8953901731261575e-06, + "loss": 0.3987, + "step": 52052 + }, + { + "epoch": 3.612032475192561, + "grad_norm": 1.0004838705062866, + "learning_rate": 4.8936538954377445e-06, + "loss": 0.2033, + "step": 52053 + }, + { + "epoch": 3.612101866629658, + "grad_norm": 0.7342423796653748, + "learning_rate": 4.89191791798902e-06, + "loss": 0.1593, + "step": 52054 + }, + { + "epoch": 3.6121712580667547, + "grad_norm": 0.8892377614974976, + "learning_rate": 4.89018224078549e-06, + "loss": 0.3259, + "step": 52055 + }, + { + "epoch": 3.6122406495038515, + "grad_norm": 1.1775521039962769, + "learning_rate": 4.888446863832608e-06, + "loss": 0.2496, + "step": 52056 + }, + { + "epoch": 3.612310040940948, + "grad_norm": 0.5455511212348938, + "learning_rate": 4.886711787135878e-06, + "loss": 0.167, + "step": 52057 + }, + { + "epoch": 3.6123794323780447, + "grad_norm": 1.2547796964645386, + "learning_rate": 4.884977010700764e-06, + "loss": 0.1743, + "step": 52058 + }, + { + "epoch": 3.612448823815141, + "grad_norm": 1.2499728202819824, + "learning_rate": 4.883242534532739e-06, + "loss": 0.3015, + "step": 52059 + }, + { + "epoch": 3.612518215252238, + "grad_norm": 0.9969820976257324, + "learning_rate": 4.881508358637277e-06, + "loss": 0.2575, + "step": 52060 + }, + { + "epoch": 3.6125876066893348, + "grad_norm": 0.6252544522285461, + "learning_rate": 4.879774483019861e-06, + "loss": 0.1297, + "step": 52061 + }, + { + "epoch": 3.612656998126431, + "grad_norm": 0.7509046792984009, + "learning_rate": 4.878040907685954e-06, + "loss": 0.1436, + "step": 52062 + }, + { + "epoch": 3.612726389563528, + "grad_norm": 0.6669710278511047, + "learning_rate": 4.876307632641042e-06, + "loss": 0.2353, + "step": 52063 + }, + { + "epoch": 3.6127957810006244, + "grad_norm": 0.7830537557601929, + "learning_rate": 4.874574657890585e-06, + "loss": 0.1518, + "step": 52064 + }, + { + "epoch": 3.612865172437721, + "grad_norm": 0.7325148582458496, + "learning_rate": 4.87284198344008e-06, + "loss": 0.2308, + "step": 52065 + }, + { + "epoch": 3.612934563874818, + "grad_norm": 0.8979803919792175, + "learning_rate": 4.871109609294955e-06, + "loss": 0.1675, + "step": 52066 + }, + { + "epoch": 3.6130039553119144, + "grad_norm": 0.9229792952537537, + "learning_rate": 4.869377535460706e-06, + "loss": 0.1543, + "step": 52067 + }, + { + "epoch": 3.6130733467490113, + "grad_norm": 1.0861339569091797, + "learning_rate": 4.867645761942785e-06, + "loss": 0.4172, + "step": 52068 + }, + { + "epoch": 3.6131427381861076, + "grad_norm": 0.8197620511054993, + "learning_rate": 4.865914288746687e-06, + "loss": 0.2447, + "step": 52069 + }, + { + "epoch": 3.6132121296232045, + "grad_norm": 0.7787448167800903, + "learning_rate": 4.864183115877852e-06, + "loss": 0.3106, + "step": 52070 + }, + { + "epoch": 3.6132815210603013, + "grad_norm": 0.9797319769859314, + "learning_rate": 4.862452243341753e-06, + "loss": 0.2651, + "step": 52071 + }, + { + "epoch": 3.6133509124973977, + "grad_norm": 1.0120916366577148, + "learning_rate": 4.860721671143853e-06, + "loss": 0.2025, + "step": 52072 + }, + { + "epoch": 3.6134203039344945, + "grad_norm": 1.2704602479934692, + "learning_rate": 4.858991399289625e-06, + "loss": 0.2654, + "step": 52073 + }, + { + "epoch": 3.613489695371591, + "grad_norm": 1.887503981590271, + "learning_rate": 4.85726142778451e-06, + "loss": 0.3749, + "step": 52074 + }, + { + "epoch": 3.6135590868086878, + "grad_norm": 0.9121191501617432, + "learning_rate": 4.8555317566339905e-06, + "loss": 0.3195, + "step": 52075 + }, + { + "epoch": 3.6136284782457846, + "grad_norm": 0.8896672129631042, + "learning_rate": 4.853802385843509e-06, + "loss": 0.1693, + "step": 52076 + }, + { + "epoch": 3.613697869682881, + "grad_norm": 0.5661521553993225, + "learning_rate": 4.852073315418537e-06, + "loss": 0.1868, + "step": 52077 + }, + { + "epoch": 3.613767261119978, + "grad_norm": 0.7860503196716309, + "learning_rate": 4.850344545364538e-06, + "loss": 0.1837, + "step": 52078 + }, + { + "epoch": 3.613836652557074, + "grad_norm": 0.581171989440918, + "learning_rate": 4.848616075686963e-06, + "loss": 0.1674, + "step": 52079 + }, + { + "epoch": 3.613906043994171, + "grad_norm": 0.869451642036438, + "learning_rate": 4.846887906391262e-06, + "loss": 0.1582, + "step": 52080 + }, + { + "epoch": 3.613975435431268, + "grad_norm": 0.7536153793334961, + "learning_rate": 4.845160037482888e-06, + "loss": 0.2442, + "step": 52081 + }, + { + "epoch": 3.6140448268683647, + "grad_norm": 0.8561946749687195, + "learning_rate": 4.843432468967313e-06, + "loss": 0.239, + "step": 52082 + }, + { + "epoch": 3.614114218305461, + "grad_norm": 0.7404789328575134, + "learning_rate": 4.841705200849977e-06, + "loss": 0.2177, + "step": 52083 + }, + { + "epoch": 3.614183609742558, + "grad_norm": 1.213633418083191, + "learning_rate": 4.839978233136333e-06, + "loss": 0.2945, + "step": 52084 + }, + { + "epoch": 3.6142530011796543, + "grad_norm": 0.9095597863197327, + "learning_rate": 4.838251565831853e-06, + "loss": 0.2762, + "step": 52085 + }, + { + "epoch": 3.614322392616751, + "grad_norm": 1.1770330667495728, + "learning_rate": 4.836525198941965e-06, + "loss": 0.3051, + "step": 52086 + }, + { + "epoch": 3.614391784053848, + "grad_norm": 0.9944003820419312, + "learning_rate": 4.834799132472123e-06, + "loss": 0.1719, + "step": 52087 + }, + { + "epoch": 3.6144611754909444, + "grad_norm": 1.023672103881836, + "learning_rate": 4.833073366427787e-06, + "loss": 0.2948, + "step": 52088 + }, + { + "epoch": 3.614530566928041, + "grad_norm": 0.9322360754013062, + "learning_rate": 4.831347900814387e-06, + "loss": 0.249, + "step": 52089 + }, + { + "epoch": 3.6145999583651376, + "grad_norm": 0.8972591161727905, + "learning_rate": 4.829622735637374e-06, + "loss": 0.1871, + "step": 52090 + }, + { + "epoch": 3.6146693498022344, + "grad_norm": 1.2288755178451538, + "learning_rate": 4.82789787090221e-06, + "loss": 0.2349, + "step": 52091 + }, + { + "epoch": 3.6147387412393313, + "grad_norm": 1.1804594993591309, + "learning_rate": 4.826173306614345e-06, + "loss": 0.2704, + "step": 52092 + }, + { + "epoch": 3.6148081326764276, + "grad_norm": 0.8629146218299866, + "learning_rate": 4.824449042779189e-06, + "loss": 0.2665, + "step": 52093 + }, + { + "epoch": 3.6148775241135245, + "grad_norm": 0.9298919439315796, + "learning_rate": 4.822725079402201e-06, + "loss": 0.1936, + "step": 52094 + }, + { + "epoch": 3.614946915550621, + "grad_norm": 1.0946414470672607, + "learning_rate": 4.821001416488835e-06, + "loss": 0.1637, + "step": 52095 + }, + { + "epoch": 3.6150163069877177, + "grad_norm": 1.1197139024734497, + "learning_rate": 4.819278054044529e-06, + "loss": 0.2223, + "step": 52096 + }, + { + "epoch": 3.6150856984248145, + "grad_norm": 0.6589378714561462, + "learning_rate": 4.817554992074713e-06, + "loss": 0.1811, + "step": 52097 + }, + { + "epoch": 3.615155089861911, + "grad_norm": 0.9493675827980042, + "learning_rate": 4.815832230584838e-06, + "loss": 0.2687, + "step": 52098 + }, + { + "epoch": 3.6152244812990078, + "grad_norm": 0.7794312834739685, + "learning_rate": 4.814109769580344e-06, + "loss": 0.1774, + "step": 52099 + }, + { + "epoch": 3.615293872736104, + "grad_norm": 0.8245791792869568, + "learning_rate": 4.812387609066649e-06, + "loss": 0.3049, + "step": 52100 + }, + { + "epoch": 3.615363264173201, + "grad_norm": 0.5326295495033264, + "learning_rate": 4.810665749049204e-06, + "loss": 0.154, + "step": 52101 + }, + { + "epoch": 3.615432655610298, + "grad_norm": 1.8708103895187378, + "learning_rate": 4.8089441895334375e-06, + "loss": 0.362, + "step": 52102 + }, + { + "epoch": 3.615502047047394, + "grad_norm": 0.7575514912605286, + "learning_rate": 4.8072229305248016e-06, + "loss": 0.1799, + "step": 52103 + }, + { + "epoch": 3.615571438484491, + "grad_norm": 1.0312367677688599, + "learning_rate": 4.805501972028725e-06, + "loss": 0.2633, + "step": 52104 + }, + { + "epoch": 3.6156408299215874, + "grad_norm": 0.9033887386322021, + "learning_rate": 4.803781314050615e-06, + "loss": 0.2785, + "step": 52105 + }, + { + "epoch": 3.6157102213586843, + "grad_norm": 0.9025610685348511, + "learning_rate": 4.802060956595933e-06, + "loss": 0.2189, + "step": 52106 + }, + { + "epoch": 3.615779612795781, + "grad_norm": 0.9332075715065002, + "learning_rate": 4.800340899670097e-06, + "loss": 0.1793, + "step": 52107 + }, + { + "epoch": 3.615849004232878, + "grad_norm": 1.402907371520996, + "learning_rate": 4.798621143278548e-06, + "loss": 0.2189, + "step": 52108 + }, + { + "epoch": 3.6159183956699743, + "grad_norm": 1.3296504020690918, + "learning_rate": 4.796901687426703e-06, + "loss": 0.3718, + "step": 52109 + }, + { + "epoch": 3.615987787107071, + "grad_norm": 0.8706786036491394, + "learning_rate": 4.795182532119991e-06, + "loss": 0.2971, + "step": 52110 + }, + { + "epoch": 3.6160571785441675, + "grad_norm": 0.8350399732589722, + "learning_rate": 4.79346367736383e-06, + "loss": 0.3337, + "step": 52111 + }, + { + "epoch": 3.6161265699812644, + "grad_norm": 0.8216656446456909, + "learning_rate": 4.791745123163671e-06, + "loss": 0.2305, + "step": 52112 + }, + { + "epoch": 3.616195961418361, + "grad_norm": 1.4718087911605835, + "learning_rate": 4.790026869524933e-06, + "loss": 0.3974, + "step": 52113 + }, + { + "epoch": 3.6162653528554576, + "grad_norm": 1.198455810546875, + "learning_rate": 4.788308916453021e-06, + "loss": 0.2843, + "step": 52114 + }, + { + "epoch": 3.6163347442925544, + "grad_norm": 1.1496206521987915, + "learning_rate": 4.786591263953377e-06, + "loss": 0.2647, + "step": 52115 + }, + { + "epoch": 3.616404135729651, + "grad_norm": 0.9327420592308044, + "learning_rate": 4.784873912031418e-06, + "loss": 0.2113, + "step": 52116 + }, + { + "epoch": 3.6164735271667476, + "grad_norm": 0.7419897317886353, + "learning_rate": 4.783156860692561e-06, + "loss": 0.1889, + "step": 52117 + }, + { + "epoch": 3.6165429186038445, + "grad_norm": 1.0149259567260742, + "learning_rate": 4.781440109942226e-06, + "loss": 0.3879, + "step": 52118 + }, + { + "epoch": 3.616612310040941, + "grad_norm": 0.9485196471214294, + "learning_rate": 4.77972365978584e-06, + "loss": 0.4535, + "step": 52119 + }, + { + "epoch": 3.6166817014780377, + "grad_norm": 0.9668565988540649, + "learning_rate": 4.778007510228832e-06, + "loss": 0.2273, + "step": 52120 + }, + { + "epoch": 3.616751092915134, + "grad_norm": 0.8602322340011597, + "learning_rate": 4.7762916612766e-06, + "loss": 0.2593, + "step": 52121 + }, + { + "epoch": 3.616820484352231, + "grad_norm": 0.9494712352752686, + "learning_rate": 4.774576112934559e-06, + "loss": 0.1501, + "step": 52122 + }, + { + "epoch": 3.6168898757893277, + "grad_norm": 0.7704013586044312, + "learning_rate": 4.77286086520814e-06, + "loss": 0.2374, + "step": 52123 + }, + { + "epoch": 3.616959267226424, + "grad_norm": 0.8443625569343567, + "learning_rate": 4.771145918102748e-06, + "loss": 0.3547, + "step": 52124 + }, + { + "epoch": 3.617028658663521, + "grad_norm": 0.5953601598739624, + "learning_rate": 4.769431271623792e-06, + "loss": 0.1849, + "step": 52125 + }, + { + "epoch": 3.6170980501006174, + "grad_norm": 0.7255520224571228, + "learning_rate": 4.7677169257767e-06, + "loss": 0.1521, + "step": 52126 + }, + { + "epoch": 3.617167441537714, + "grad_norm": 1.0099189281463623, + "learning_rate": 4.766002880566889e-06, + "loss": 0.2815, + "step": 52127 + }, + { + "epoch": 3.617236832974811, + "grad_norm": 1.2685657739639282, + "learning_rate": 4.764289135999744e-06, + "loss": 0.2991, + "step": 52128 + }, + { + "epoch": 3.6173062244119074, + "grad_norm": 0.8103811144828796, + "learning_rate": 4.762575692080695e-06, + "loss": 0.2171, + "step": 52129 + }, + { + "epoch": 3.6173756158490042, + "grad_norm": 0.9858807325363159, + "learning_rate": 4.760862548815137e-06, + "loss": 0.3595, + "step": 52130 + }, + { + "epoch": 3.6174450072861006, + "grad_norm": 1.048393964767456, + "learning_rate": 4.759149706208487e-06, + "loss": 0.3052, + "step": 52131 + }, + { + "epoch": 3.6175143987231975, + "grad_norm": 0.7622799873352051, + "learning_rate": 4.757437164266154e-06, + "loss": 0.288, + "step": 52132 + }, + { + "epoch": 3.6175837901602943, + "grad_norm": 1.2412291765213013, + "learning_rate": 4.755724922993543e-06, + "loss": 0.2746, + "step": 52133 + }, + { + "epoch": 3.617653181597391, + "grad_norm": 0.840214192867279, + "learning_rate": 4.754012982396072e-06, + "loss": 0.2141, + "step": 52134 + }, + { + "epoch": 3.6177225730344875, + "grad_norm": 1.1942274570465088, + "learning_rate": 4.752301342479115e-06, + "loss": 0.2507, + "step": 52135 + }, + { + "epoch": 3.6177919644715844, + "grad_norm": 0.9008787274360657, + "learning_rate": 4.75059000324809e-06, + "loss": 0.256, + "step": 52136 + }, + { + "epoch": 3.6178613559086807, + "grad_norm": 1.2331202030181885, + "learning_rate": 4.748878964708403e-06, + "loss": 0.211, + "step": 52137 + }, + { + "epoch": 3.6179307473457776, + "grad_norm": 0.7556912302970886, + "learning_rate": 4.747168226865451e-06, + "loss": 0.287, + "step": 52138 + }, + { + "epoch": 3.6180001387828744, + "grad_norm": 0.8324945569038391, + "learning_rate": 4.745457789724661e-06, + "loss": 0.3579, + "step": 52139 + }, + { + "epoch": 3.618069530219971, + "grad_norm": 1.374661922454834, + "learning_rate": 4.743747653291386e-06, + "loss": 0.171, + "step": 52140 + }, + { + "epoch": 3.6181389216570676, + "grad_norm": 1.1350528001785278, + "learning_rate": 4.742037817571054e-06, + "loss": 0.3047, + "step": 52141 + }, + { + "epoch": 3.618208313094164, + "grad_norm": 1.256488561630249, + "learning_rate": 4.740328282569051e-06, + "loss": 0.3181, + "step": 52142 + }, + { + "epoch": 3.618277704531261, + "grad_norm": 0.7093169093132019, + "learning_rate": 4.7386190482907935e-06, + "loss": 0.1976, + "step": 52143 + }, + { + "epoch": 3.6183470959683577, + "grad_norm": 1.1871131658554077, + "learning_rate": 4.736910114741655e-06, + "loss": 0.3408, + "step": 52144 + }, + { + "epoch": 3.618416487405454, + "grad_norm": 0.679844856262207, + "learning_rate": 4.735201481927032e-06, + "loss": 0.1506, + "step": 52145 + }, + { + "epoch": 3.618485878842551, + "grad_norm": 0.9478061199188232, + "learning_rate": 4.73349314985233e-06, + "loss": 0.2364, + "step": 52146 + }, + { + "epoch": 3.6185552702796473, + "grad_norm": 1.0329707860946655, + "learning_rate": 4.731785118522936e-06, + "loss": 0.2585, + "step": 52147 + }, + { + "epoch": 3.618624661716744, + "grad_norm": 0.6713009476661682, + "learning_rate": 4.730077387944254e-06, + "loss": 0.1381, + "step": 52148 + }, + { + "epoch": 3.618694053153841, + "grad_norm": 0.7712783217430115, + "learning_rate": 4.728369958121659e-06, + "loss": 0.2199, + "step": 52149 + }, + { + "epoch": 3.6187634445909374, + "grad_norm": 0.9913530349731445, + "learning_rate": 4.7266628290605355e-06, + "loss": 0.2674, + "step": 52150 + }, + { + "epoch": 3.618832836028034, + "grad_norm": 1.2138394117355347, + "learning_rate": 4.72495600076629e-06, + "loss": 0.3057, + "step": 52151 + }, + { + "epoch": 3.6189022274651306, + "grad_norm": 0.8352497816085815, + "learning_rate": 4.7232494732443065e-06, + "loss": 0.2093, + "step": 52152 + }, + { + "epoch": 3.6189716189022274, + "grad_norm": 0.9471281170845032, + "learning_rate": 4.721543246499971e-06, + "loss": 0.2619, + "step": 52153 + }, + { + "epoch": 3.6190410103393242, + "grad_norm": 1.264260172843933, + "learning_rate": 4.719837320538667e-06, + "loss": 0.2539, + "step": 52154 + }, + { + "epoch": 3.6191104017764206, + "grad_norm": 0.5834839940071106, + "learning_rate": 4.71813169536579e-06, + "loss": 0.1267, + "step": 52155 + }, + { + "epoch": 3.6191797932135175, + "grad_norm": 0.7128828763961792, + "learning_rate": 4.7164263709867045e-06, + "loss": 0.2248, + "step": 52156 + }, + { + "epoch": 3.619249184650614, + "grad_norm": 0.9110264182090759, + "learning_rate": 4.714721347406814e-06, + "loss": 0.2606, + "step": 52157 + }, + { + "epoch": 3.6193185760877107, + "grad_norm": 0.7862228155136108, + "learning_rate": 4.713016624631483e-06, + "loss": 0.1736, + "step": 52158 + }, + { + "epoch": 3.6193879675248075, + "grad_norm": 1.2802734375, + "learning_rate": 4.711312202666107e-06, + "loss": 0.3694, + "step": 52159 + }, + { + "epoch": 3.6194573589619043, + "grad_norm": 1.0516026020050049, + "learning_rate": 4.709608081516071e-06, + "loss": 0.2998, + "step": 52160 + }, + { + "epoch": 3.6195267503990007, + "grad_norm": 0.7165473103523254, + "learning_rate": 4.7079042611867356e-06, + "loss": 0.1644, + "step": 52161 + }, + { + "epoch": 3.6195961418360976, + "grad_norm": 1.0712716579437256, + "learning_rate": 4.706200741683508e-06, + "loss": 0.2527, + "step": 52162 + }, + { + "epoch": 3.619665533273194, + "grad_norm": 0.920430600643158, + "learning_rate": 4.704497523011731e-06, + "loss": 0.1918, + "step": 52163 + }, + { + "epoch": 3.619734924710291, + "grad_norm": 0.7213340997695923, + "learning_rate": 4.702794605176808e-06, + "loss": 0.2016, + "step": 52164 + }, + { + "epoch": 3.6198043161473876, + "grad_norm": 1.1309888362884521, + "learning_rate": 4.701091988184103e-06, + "loss": 0.2402, + "step": 52165 + }, + { + "epoch": 3.619873707584484, + "grad_norm": 1.1069445610046387, + "learning_rate": 4.69938967203899e-06, + "loss": 0.3175, + "step": 52166 + }, + { + "epoch": 3.619943099021581, + "grad_norm": 0.843416154384613, + "learning_rate": 4.697687656746841e-06, + "loss": 0.131, + "step": 52167 + }, + { + "epoch": 3.6200124904586772, + "grad_norm": 0.6960630416870117, + "learning_rate": 4.695985942313064e-06, + "loss": 0.2202, + "step": 52168 + }, + { + "epoch": 3.620081881895774, + "grad_norm": 0.8868682384490967, + "learning_rate": 4.694284528742976e-06, + "loss": 0.1852, + "step": 52169 + }, + { + "epoch": 3.620151273332871, + "grad_norm": 1.2065720558166504, + "learning_rate": 4.692583416041985e-06, + "loss": 0.2765, + "step": 52170 + }, + { + "epoch": 3.6202206647699673, + "grad_norm": 0.933879017829895, + "learning_rate": 4.6908826042154406e-06, + "loss": 0.2576, + "step": 52171 + }, + { + "epoch": 3.620290056207064, + "grad_norm": 0.9621863961219788, + "learning_rate": 4.689182093268718e-06, + "loss": 0.2358, + "step": 52172 + }, + { + "epoch": 3.6203594476441605, + "grad_norm": 0.6331132054328918, + "learning_rate": 4.687481883207201e-06, + "loss": 0.1553, + "step": 52173 + }, + { + "epoch": 3.6204288390812573, + "grad_norm": 0.6336368322372437, + "learning_rate": 4.6857819740362515e-06, + "loss": 0.1158, + "step": 52174 + }, + { + "epoch": 3.620498230518354, + "grad_norm": 0.816886842250824, + "learning_rate": 4.684082365761222e-06, + "loss": 0.2155, + "step": 52175 + }, + { + "epoch": 3.6205676219554506, + "grad_norm": 1.426523208618164, + "learning_rate": 4.682383058387485e-06, + "loss": 0.4223, + "step": 52176 + }, + { + "epoch": 3.6206370133925474, + "grad_norm": 0.6299039125442505, + "learning_rate": 4.6806840519204036e-06, + "loss": 0.2692, + "step": 52177 + }, + { + "epoch": 3.620706404829644, + "grad_norm": 0.8006047010421753, + "learning_rate": 4.678985346365361e-06, + "loss": 0.2923, + "step": 52178 + }, + { + "epoch": 3.6207757962667406, + "grad_norm": 0.8870963454246521, + "learning_rate": 4.677286941727688e-06, + "loss": 0.2226, + "step": 52179 + }, + { + "epoch": 3.6208451877038375, + "grad_norm": 0.9735704064369202, + "learning_rate": 4.6755888380127564e-06, + "loss": 0.3825, + "step": 52180 + }, + { + "epoch": 3.6209145791409343, + "grad_norm": 0.7508691549301147, + "learning_rate": 4.67389103522593e-06, + "loss": 0.2325, + "step": 52181 + }, + { + "epoch": 3.6209839705780307, + "grad_norm": 0.715562641620636, + "learning_rate": 4.672193533372582e-06, + "loss": 0.2422, + "step": 52182 + }, + { + "epoch": 3.621053362015127, + "grad_norm": 0.86064612865448, + "learning_rate": 4.670496332458052e-06, + "loss": 0.1797, + "step": 52183 + }, + { + "epoch": 3.621122753452224, + "grad_norm": 1.0127463340759277, + "learning_rate": 4.668799432487703e-06, + "loss": 0.2732, + "step": 52184 + }, + { + "epoch": 3.6211921448893207, + "grad_norm": 1.0127110481262207, + "learning_rate": 4.667102833466897e-06, + "loss": 0.2642, + "step": 52185 + }, + { + "epoch": 3.6212615363264176, + "grad_norm": 0.8860352039337158, + "learning_rate": 4.6654065354009845e-06, + "loss": 0.1877, + "step": 52186 + }, + { + "epoch": 3.621330927763514, + "grad_norm": 0.7422643899917603, + "learning_rate": 4.663710538295319e-06, + "loss": 0.2183, + "step": 52187 + }, + { + "epoch": 3.621400319200611, + "grad_norm": 1.2780251502990723, + "learning_rate": 4.66201484215526e-06, + "loss": 0.3732, + "step": 52188 + }, + { + "epoch": 3.621469710637707, + "grad_norm": 0.9577349424362183, + "learning_rate": 4.660319446986172e-06, + "loss": 0.2639, + "step": 52189 + }, + { + "epoch": 3.621539102074804, + "grad_norm": 1.0480886697769165, + "learning_rate": 4.6586243527933836e-06, + "loss": 0.2543, + "step": 52190 + }, + { + "epoch": 3.621608493511901, + "grad_norm": 1.361889362335205, + "learning_rate": 4.656929559582246e-06, + "loss": 0.3108, + "step": 52191 + }, + { + "epoch": 3.6216778849489972, + "grad_norm": 0.8024996519088745, + "learning_rate": 4.6552350673581325e-06, + "loss": 0.2462, + "step": 52192 + }, + { + "epoch": 3.621747276386094, + "grad_norm": 0.886640191078186, + "learning_rate": 4.653540876126372e-06, + "loss": 0.4099, + "step": 52193 + }, + { + "epoch": 3.6218166678231904, + "grad_norm": 0.9638923406600952, + "learning_rate": 4.651846985892328e-06, + "loss": 0.2953, + "step": 52194 + }, + { + "epoch": 3.6218860592602873, + "grad_norm": 1.0154563188552856, + "learning_rate": 4.650153396661339e-06, + "loss": 0.2567, + "step": 52195 + }, + { + "epoch": 3.621955450697384, + "grad_norm": 0.6255817413330078, + "learning_rate": 4.648460108438757e-06, + "loss": 0.1875, + "step": 52196 + }, + { + "epoch": 3.6220248421344805, + "grad_norm": 0.7253646850585938, + "learning_rate": 4.646767121229923e-06, + "loss": 0.2382, + "step": 52197 + }, + { + "epoch": 3.6220942335715773, + "grad_norm": 0.7267038822174072, + "learning_rate": 4.645074435040175e-06, + "loss": 0.1631, + "step": 52198 + }, + { + "epoch": 3.6221636250086737, + "grad_norm": 0.8673741817474365, + "learning_rate": 4.643382049874867e-06, + "loss": 0.3033, + "step": 52199 + }, + { + "epoch": 3.6222330164457706, + "grad_norm": 0.9175138473510742, + "learning_rate": 4.641689965739338e-06, + "loss": 0.2391, + "step": 52200 + }, + { + "epoch": 3.6223024078828674, + "grad_norm": 0.7806283235549927, + "learning_rate": 4.639998182638927e-06, + "loss": 0.2551, + "step": 52201 + }, + { + "epoch": 3.622371799319964, + "grad_norm": 1.027359962463379, + "learning_rate": 4.638306700578976e-06, + "loss": 0.3472, + "step": 52202 + }, + { + "epoch": 3.6224411907570606, + "grad_norm": 1.1504502296447754, + "learning_rate": 4.636615519564846e-06, + "loss": 0.3804, + "step": 52203 + }, + { + "epoch": 3.622510582194157, + "grad_norm": 0.6386018395423889, + "learning_rate": 4.634924639601845e-06, + "loss": 0.2481, + "step": 52204 + }, + { + "epoch": 3.622579973631254, + "grad_norm": 1.1669185161590576, + "learning_rate": 4.633234060695313e-06, + "loss": 0.3529, + "step": 52205 + }, + { + "epoch": 3.6226493650683507, + "grad_norm": 1.0108108520507812, + "learning_rate": 4.6315437828506e-06, + "loss": 0.2204, + "step": 52206 + }, + { + "epoch": 3.6227187565054475, + "grad_norm": 0.681696891784668, + "learning_rate": 4.629853806073048e-06, + "loss": 0.1512, + "step": 52207 + }, + { + "epoch": 3.622788147942544, + "grad_norm": 1.0779383182525635, + "learning_rate": 4.628164130367973e-06, + "loss": 0.183, + "step": 52208 + }, + { + "epoch": 3.6228575393796403, + "grad_norm": 0.7829223275184631, + "learning_rate": 4.626474755740728e-06, + "loss": 0.2121, + "step": 52209 + }, + { + "epoch": 3.622926930816737, + "grad_norm": 0.9652488827705383, + "learning_rate": 4.624785682196631e-06, + "loss": 0.2817, + "step": 52210 + }, + { + "epoch": 3.622996322253834, + "grad_norm": 1.332741379737854, + "learning_rate": 4.62309690974102e-06, + "loss": 0.3879, + "step": 52211 + }, + { + "epoch": 3.623065713690931, + "grad_norm": 0.9380783438682556, + "learning_rate": 4.621408438379238e-06, + "loss": 0.2991, + "step": 52212 + }, + { + "epoch": 3.623135105128027, + "grad_norm": 0.896447479724884, + "learning_rate": 4.619720268116589e-06, + "loss": 0.1778, + "step": 52213 + }, + { + "epoch": 3.623204496565124, + "grad_norm": 1.2031573057174683, + "learning_rate": 4.618032398958428e-06, + "loss": 0.2607, + "step": 52214 + }, + { + "epoch": 3.6232738880022204, + "grad_norm": 0.7058300375938416, + "learning_rate": 4.616344830910058e-06, + "loss": 0.1885, + "step": 52215 + }, + { + "epoch": 3.623343279439317, + "grad_norm": 1.1890146732330322, + "learning_rate": 4.614657563976832e-06, + "loss": 0.4645, + "step": 52216 + }, + { + "epoch": 3.623412670876414, + "grad_norm": 0.6029397249221802, + "learning_rate": 4.612970598164068e-06, + "loss": 0.2186, + "step": 52217 + }, + { + "epoch": 3.6234820623135104, + "grad_norm": 0.751585066318512, + "learning_rate": 4.611283933477084e-06, + "loss": 0.1584, + "step": 52218 + }, + { + "epoch": 3.6235514537506073, + "grad_norm": 1.5378258228302002, + "learning_rate": 4.609597569921198e-06, + "loss": 0.1463, + "step": 52219 + }, + { + "epoch": 3.6236208451877037, + "grad_norm": 0.728348970413208, + "learning_rate": 4.607911507501761e-06, + "loss": 0.1519, + "step": 52220 + }, + { + "epoch": 3.6236902366248005, + "grad_norm": 0.869388222694397, + "learning_rate": 4.606225746224069e-06, + "loss": 0.2217, + "step": 52221 + }, + { + "epoch": 3.6237596280618973, + "grad_norm": 0.6152468323707581, + "learning_rate": 4.604540286093451e-06, + "loss": 0.2035, + "step": 52222 + }, + { + "epoch": 3.6238290194989937, + "grad_norm": 0.8100934028625488, + "learning_rate": 4.6028551271152464e-06, + "loss": 0.2475, + "step": 52223 + }, + { + "epoch": 3.6238984109360906, + "grad_norm": 1.295433759689331, + "learning_rate": 4.601170269294763e-06, + "loss": 0.2891, + "step": 52224 + }, + { + "epoch": 3.623967802373187, + "grad_norm": 1.1590137481689453, + "learning_rate": 4.599485712637297e-06, + "loss": 0.2908, + "step": 52225 + }, + { + "epoch": 3.6240371938102838, + "grad_norm": 1.0886930227279663, + "learning_rate": 4.5978014571481985e-06, + "loss": 0.2432, + "step": 52226 + }, + { + "epoch": 3.6241065852473806, + "grad_norm": 1.0322070121765137, + "learning_rate": 4.596117502832764e-06, + "loss": 0.2299, + "step": 52227 + }, + { + "epoch": 3.624175976684477, + "grad_norm": 0.914868175983429, + "learning_rate": 4.594433849696323e-06, + "loss": 0.1864, + "step": 52228 + }, + { + "epoch": 3.624245368121574, + "grad_norm": 1.2272039651870728, + "learning_rate": 4.59275049774418e-06, + "loss": 0.2111, + "step": 52229 + }, + { + "epoch": 3.62431475955867, + "grad_norm": 0.9486129283905029, + "learning_rate": 4.591067446981667e-06, + "loss": 0.2729, + "step": 52230 + }, + { + "epoch": 3.624384150995767, + "grad_norm": 0.8189377784729004, + "learning_rate": 4.589384697414089e-06, + "loss": 0.1547, + "step": 52231 + }, + { + "epoch": 3.624453542432864, + "grad_norm": 0.7581535577774048, + "learning_rate": 4.587702249046733e-06, + "loss": 0.1918, + "step": 52232 + }, + { + "epoch": 3.6245229338699607, + "grad_norm": 0.9892758131027222, + "learning_rate": 4.586020101884936e-06, + "loss": 0.1907, + "step": 52233 + }, + { + "epoch": 3.624592325307057, + "grad_norm": 0.7848137021064758, + "learning_rate": 4.584338255934006e-06, + "loss": 0.1698, + "step": 52234 + }, + { + "epoch": 3.6246617167441535, + "grad_norm": 1.0743827819824219, + "learning_rate": 4.582656711199251e-06, + "loss": 0.3705, + "step": 52235 + }, + { + "epoch": 3.6247311081812503, + "grad_norm": 1.027270793914795, + "learning_rate": 4.580975467685977e-06, + "loss": 0.1558, + "step": 52236 + }, + { + "epoch": 3.624800499618347, + "grad_norm": 0.7736921310424805, + "learning_rate": 4.579294525399491e-06, + "loss": 0.1682, + "step": 52237 + }, + { + "epoch": 3.624869891055444, + "grad_norm": 1.1381422281265259, + "learning_rate": 4.577613884345122e-06, + "loss": 0.2672, + "step": 52238 + }, + { + "epoch": 3.6249392824925404, + "grad_norm": 0.7827662229537964, + "learning_rate": 4.575933544528133e-06, + "loss": 0.1893, + "step": 52239 + }, + { + "epoch": 3.625008673929637, + "grad_norm": 0.755563497543335, + "learning_rate": 4.574253505953851e-06, + "loss": 0.3329, + "step": 52240 + }, + { + "epoch": 3.6250780653667336, + "grad_norm": 0.9425761699676514, + "learning_rate": 4.572573768627586e-06, + "loss": 0.3469, + "step": 52241 + }, + { + "epoch": 3.6251474568038304, + "grad_norm": 0.9860048294067383, + "learning_rate": 4.570894332554632e-06, + "loss": 0.322, + "step": 52242 + }, + { + "epoch": 3.6252168482409273, + "grad_norm": 0.93816077709198, + "learning_rate": 4.569215197740295e-06, + "loss": 0.3033, + "step": 52243 + }, + { + "epoch": 3.6252862396780237, + "grad_norm": 0.8195582032203674, + "learning_rate": 4.567536364189873e-06, + "loss": 0.1726, + "step": 52244 + }, + { + "epoch": 3.6253556311151205, + "grad_norm": 0.7224617004394531, + "learning_rate": 4.5658578319086705e-06, + "loss": 0.1684, + "step": 52245 + }, + { + "epoch": 3.625425022552217, + "grad_norm": 0.7880829572677612, + "learning_rate": 4.564179600901974e-06, + "loss": 0.2072, + "step": 52246 + }, + { + "epoch": 3.6254944139893137, + "grad_norm": 0.9408438205718994, + "learning_rate": 4.562501671175101e-06, + "loss": 0.3746, + "step": 52247 + }, + { + "epoch": 3.6255638054264105, + "grad_norm": 0.7550655603408813, + "learning_rate": 4.560824042733325e-06, + "loss": 0.2235, + "step": 52248 + }, + { + "epoch": 3.625633196863507, + "grad_norm": 0.8673001527786255, + "learning_rate": 4.5591467155819636e-06, + "loss": 0.2074, + "step": 52249 + }, + { + "epoch": 3.6257025883006038, + "grad_norm": 0.9048520922660828, + "learning_rate": 4.5574696897262906e-06, + "loss": 0.2865, + "step": 52250 + }, + { + "epoch": 3.6257719797377, + "grad_norm": 0.7758841514587402, + "learning_rate": 4.5557929651716346e-06, + "loss": 0.1554, + "step": 52251 + }, + { + "epoch": 3.625841371174797, + "grad_norm": 0.6988406777381897, + "learning_rate": 4.554116541923248e-06, + "loss": 0.1779, + "step": 52252 + }, + { + "epoch": 3.625910762611894, + "grad_norm": 1.1122616529464722, + "learning_rate": 4.552440419986437e-06, + "loss": 0.2497, + "step": 52253 + }, + { + "epoch": 3.62598015404899, + "grad_norm": 0.9181423783302307, + "learning_rate": 4.550764599366508e-06, + "loss": 0.3132, + "step": 52254 + }, + { + "epoch": 3.626049545486087, + "grad_norm": 0.8230186104774475, + "learning_rate": 4.5490890800687354e-06, + "loss": 0.3011, + "step": 52255 + }, + { + "epoch": 3.6261189369231834, + "grad_norm": 1.0555229187011719, + "learning_rate": 4.5474138620984154e-06, + "loss": 0.3318, + "step": 52256 + }, + { + "epoch": 3.6261883283602803, + "grad_norm": 0.9413148760795593, + "learning_rate": 4.545738945460842e-06, + "loss": 0.2144, + "step": 52257 + }, + { + "epoch": 3.626257719797377, + "grad_norm": 0.973350465297699, + "learning_rate": 4.5440643301613016e-06, + "loss": 0.2748, + "step": 52258 + }, + { + "epoch": 3.626327111234474, + "grad_norm": 0.7265670895576477, + "learning_rate": 4.542390016205056e-06, + "loss": 0.1607, + "step": 52259 + }, + { + "epoch": 3.6263965026715703, + "grad_norm": 0.8889951109886169, + "learning_rate": 4.540716003597423e-06, + "loss": 0.165, + "step": 52260 + }, + { + "epoch": 3.626465894108667, + "grad_norm": 0.7984238862991333, + "learning_rate": 4.539042292343665e-06, + "loss": 0.2392, + "step": 52261 + }, + { + "epoch": 3.6265352855457635, + "grad_norm": 1.7027297019958496, + "learning_rate": 4.537368882449078e-06, + "loss": 0.3104, + "step": 52262 + }, + { + "epoch": 3.6266046769828604, + "grad_norm": 1.1816788911819458, + "learning_rate": 4.535695773918946e-06, + "loss": 0.2558, + "step": 52263 + }, + { + "epoch": 3.626674068419957, + "grad_norm": 0.634111225605011, + "learning_rate": 4.534022966758544e-06, + "loss": 0.1286, + "step": 52264 + }, + { + "epoch": 3.6267434598570536, + "grad_norm": 1.3430390357971191, + "learning_rate": 4.5323504609731675e-06, + "loss": 0.3442, + "step": 52265 + }, + { + "epoch": 3.6268128512941504, + "grad_norm": 1.1703050136566162, + "learning_rate": 4.5306782565680664e-06, + "loss": 0.405, + "step": 52266 + }, + { + "epoch": 3.626882242731247, + "grad_norm": 1.0178349018096924, + "learning_rate": 4.5290063535485375e-06, + "loss": 0.334, + "step": 52267 + }, + { + "epoch": 3.6269516341683437, + "grad_norm": 0.8038607835769653, + "learning_rate": 4.527334751919865e-06, + "loss": 0.2029, + "step": 52268 + }, + { + "epoch": 3.6270210256054405, + "grad_norm": 0.5663026571273804, + "learning_rate": 4.525663451687323e-06, + "loss": 0.1272, + "step": 52269 + }, + { + "epoch": 3.627090417042537, + "grad_norm": 1.022838830947876, + "learning_rate": 4.523992452856174e-06, + "loss": 0.2285, + "step": 52270 + }, + { + "epoch": 3.6271598084796337, + "grad_norm": 0.8823330998420715, + "learning_rate": 4.522321755431713e-06, + "loss": 0.2796, + "step": 52271 + }, + { + "epoch": 3.62722919991673, + "grad_norm": 0.7773358225822449, + "learning_rate": 4.5206513594192144e-06, + "loss": 0.197, + "step": 52272 + }, + { + "epoch": 3.627298591353827, + "grad_norm": 0.7554326057434082, + "learning_rate": 4.518981264823929e-06, + "loss": 0.2519, + "step": 52273 + }, + { + "epoch": 3.6273679827909238, + "grad_norm": 0.7096425294876099, + "learning_rate": 4.517311471651142e-06, + "loss": 0.1603, + "step": 52274 + }, + { + "epoch": 3.62743737422802, + "grad_norm": 0.6852319240570068, + "learning_rate": 4.5156419799061266e-06, + "loss": 0.2959, + "step": 52275 + }, + { + "epoch": 3.627506765665117, + "grad_norm": 0.9094350337982178, + "learning_rate": 4.513972789594145e-06, + "loss": 0.2964, + "step": 52276 + }, + { + "epoch": 3.6275761571022134, + "grad_norm": 1.1900700330734253, + "learning_rate": 4.512303900720471e-06, + "loss": 0.2231, + "step": 52277 + }, + { + "epoch": 3.62764554853931, + "grad_norm": 0.6419341564178467, + "learning_rate": 4.510635313290379e-06, + "loss": 0.1761, + "step": 52278 + }, + { + "epoch": 3.627714939976407, + "grad_norm": 0.9827044010162354, + "learning_rate": 4.508967027309152e-06, + "loss": 0.2772, + "step": 52279 + }, + { + "epoch": 3.6277843314135034, + "grad_norm": 0.5594169497489929, + "learning_rate": 4.50729904278202e-06, + "loss": 0.1309, + "step": 52280 + }, + { + "epoch": 3.6278537228506003, + "grad_norm": 0.8143011927604675, + "learning_rate": 4.505631359714258e-06, + "loss": 0.1809, + "step": 52281 + }, + { + "epoch": 3.6279231142876966, + "grad_norm": 0.8444747924804688, + "learning_rate": 4.503963978111159e-06, + "loss": 0.1674, + "step": 52282 + }, + { + "epoch": 3.6279925057247935, + "grad_norm": 1.1045072078704834, + "learning_rate": 4.502296897977953e-06, + "loss": 0.3013, + "step": 52283 + }, + { + "epoch": 3.6280618971618903, + "grad_norm": 0.8379454612731934, + "learning_rate": 4.500630119319915e-06, + "loss": 0.2011, + "step": 52284 + }, + { + "epoch": 3.628131288598987, + "grad_norm": 0.7655956745147705, + "learning_rate": 4.498963642142306e-06, + "loss": 0.2734, + "step": 52285 + }, + { + "epoch": 3.6282006800360835, + "grad_norm": 0.7019896507263184, + "learning_rate": 4.497297466450412e-06, + "loss": 0.2237, + "step": 52286 + }, + { + "epoch": 3.6282700714731804, + "grad_norm": 0.9065969586372375, + "learning_rate": 4.4956315922494495e-06, + "loss": 0.2231, + "step": 52287 + }, + { + "epoch": 3.6283394629102768, + "grad_norm": 0.8966037034988403, + "learning_rate": 4.493966019544693e-06, + "loss": 0.2706, + "step": 52288 + }, + { + "epoch": 3.6284088543473736, + "grad_norm": 0.9299456477165222, + "learning_rate": 4.492300748341405e-06, + "loss": 0.3213, + "step": 52289 + }, + { + "epoch": 3.6284782457844704, + "grad_norm": 0.8140555024147034, + "learning_rate": 4.490635778644847e-06, + "loss": 0.2177, + "step": 52290 + }, + { + "epoch": 3.628547637221567, + "grad_norm": 0.9465065002441406, + "learning_rate": 4.488971110460272e-06, + "loss": 0.2253, + "step": 52291 + }, + { + "epoch": 3.6286170286586636, + "grad_norm": 1.178024411201477, + "learning_rate": 4.487306743792929e-06, + "loss": 0.2137, + "step": 52292 + }, + { + "epoch": 3.62868642009576, + "grad_norm": 0.92107093334198, + "learning_rate": 4.4856426786480945e-06, + "loss": 0.3214, + "step": 52293 + }, + { + "epoch": 3.628755811532857, + "grad_norm": 1.0317589044570923, + "learning_rate": 4.483978915030984e-06, + "loss": 0.3439, + "step": 52294 + }, + { + "epoch": 3.6288252029699537, + "grad_norm": 1.8576215505599976, + "learning_rate": 4.482315452946873e-06, + "loss": 0.342, + "step": 52295 + }, + { + "epoch": 3.62889459440705, + "grad_norm": 0.8857404589653015, + "learning_rate": 4.480652292401011e-06, + "loss": 0.2632, + "step": 52296 + }, + { + "epoch": 3.628963985844147, + "grad_norm": 1.3773835897445679, + "learning_rate": 4.47898943339865e-06, + "loss": 0.2723, + "step": 52297 + }, + { + "epoch": 3.6290333772812433, + "grad_norm": 1.333033800125122, + "learning_rate": 4.477326875945043e-06, + "loss": 0.336, + "step": 52298 + }, + { + "epoch": 3.62910276871834, + "grad_norm": 0.9299736022949219, + "learning_rate": 4.475664620045416e-06, + "loss": 0.2242, + "step": 52299 + }, + { + "epoch": 3.629172160155437, + "grad_norm": 0.653140127658844, + "learning_rate": 4.474002665705057e-06, + "loss": 0.1899, + "step": 52300 + }, + { + "epoch": 3.6292415515925334, + "grad_norm": 0.9141993522644043, + "learning_rate": 4.472341012929171e-06, + "loss": 0.3224, + "step": 52301 + }, + { + "epoch": 3.62931094302963, + "grad_norm": 1.0382450819015503, + "learning_rate": 4.470679661723032e-06, + "loss": 0.3187, + "step": 52302 + }, + { + "epoch": 3.6293803344667266, + "grad_norm": 1.064516544342041, + "learning_rate": 4.469018612091857e-06, + "loss": 0.2003, + "step": 52303 + }, + { + "epoch": 3.6294497259038234, + "grad_norm": 0.810023844242096, + "learning_rate": 4.467357864040922e-06, + "loss": 0.1358, + "step": 52304 + }, + { + "epoch": 3.6295191173409203, + "grad_norm": 0.8437938094139099, + "learning_rate": 4.4656974175754426e-06, + "loss": 0.1694, + "step": 52305 + }, + { + "epoch": 3.6295885087780166, + "grad_norm": 0.6728675365447998, + "learning_rate": 4.464037272700683e-06, + "loss": 0.1628, + "step": 52306 + }, + { + "epoch": 3.6296579002151135, + "grad_norm": 1.147371530532837, + "learning_rate": 4.4623774294218826e-06, + "loss": 0.2434, + "step": 52307 + }, + { + "epoch": 3.62972729165221, + "grad_norm": 0.8234036564826965, + "learning_rate": 4.460717887744248e-06, + "loss": 0.2853, + "step": 52308 + }, + { + "epoch": 3.6297966830893067, + "grad_norm": 1.4943830966949463, + "learning_rate": 4.459058647673053e-06, + "loss": 0.3537, + "step": 52309 + }, + { + "epoch": 3.6298660745264035, + "grad_norm": 0.7641923427581787, + "learning_rate": 4.457399709213528e-06, + "loss": 0.223, + "step": 52310 + }, + { + "epoch": 3.6299354659635004, + "grad_norm": 1.022472620010376, + "learning_rate": 4.4557410723709e-06, + "loss": 0.2487, + "step": 52311 + }, + { + "epoch": 3.6300048574005968, + "grad_norm": 1.132699966430664, + "learning_rate": 4.454082737150411e-06, + "loss": 0.4652, + "step": 52312 + }, + { + "epoch": 3.6300742488376936, + "grad_norm": 1.0466907024383545, + "learning_rate": 4.452424703557323e-06, + "loss": 0.2636, + "step": 52313 + }, + { + "epoch": 3.63014364027479, + "grad_norm": 0.7245398759841919, + "learning_rate": 4.450766971596821e-06, + "loss": 0.216, + "step": 52314 + }, + { + "epoch": 3.630213031711887, + "grad_norm": 0.7171481847763062, + "learning_rate": 4.449109541274166e-06, + "loss": 0.1627, + "step": 52315 + }, + { + "epoch": 3.6302824231489836, + "grad_norm": 0.9900006055831909, + "learning_rate": 4.447452412594577e-06, + "loss": 0.2214, + "step": 52316 + }, + { + "epoch": 3.63035181458608, + "grad_norm": 0.8730252385139465, + "learning_rate": 4.445795585563317e-06, + "loss": 0.3164, + "step": 52317 + }, + { + "epoch": 3.630421206023177, + "grad_norm": 0.8308306932449341, + "learning_rate": 4.444139060185582e-06, + "loss": 0.2299, + "step": 52318 + }, + { + "epoch": 3.6304905974602732, + "grad_norm": 1.0577242374420166, + "learning_rate": 4.44248283646661e-06, + "loss": 0.2825, + "step": 52319 + }, + { + "epoch": 3.63055998889737, + "grad_norm": 1.1641885042190552, + "learning_rate": 4.440826914411633e-06, + "loss": 0.3342, + "step": 52320 + }, + { + "epoch": 3.630629380334467, + "grad_norm": 0.6313067078590393, + "learning_rate": 4.4391712940258905e-06, + "loss": 0.2089, + "step": 52321 + }, + { + "epoch": 3.6306987717715633, + "grad_norm": 1.134373664855957, + "learning_rate": 4.4375159753145875e-06, + "loss": 0.4458, + "step": 52322 + }, + { + "epoch": 3.63076816320866, + "grad_norm": 1.126670002937317, + "learning_rate": 4.435860958282956e-06, + "loss": 0.3503, + "step": 52323 + }, + { + "epoch": 3.6308375546457565, + "grad_norm": 1.0739006996154785, + "learning_rate": 4.434206242936223e-06, + "loss": 0.2982, + "step": 52324 + }, + { + "epoch": 3.6309069460828534, + "grad_norm": 0.8795959949493408, + "learning_rate": 4.4325518292796185e-06, + "loss": 0.2793, + "step": 52325 + }, + { + "epoch": 3.63097633751995, + "grad_norm": 0.907866358757019, + "learning_rate": 4.430897717318361e-06, + "loss": 0.2482, + "step": 52326 + }, + { + "epoch": 3.6310457289570466, + "grad_norm": 1.4211281538009644, + "learning_rate": 4.429243907057679e-06, + "loss": 0.2441, + "step": 52327 + }, + { + "epoch": 3.6311151203941434, + "grad_norm": 0.8855316638946533, + "learning_rate": 4.427590398502768e-06, + "loss": 0.2257, + "step": 52328 + }, + { + "epoch": 3.63118451183124, + "grad_norm": 0.8084078431129456, + "learning_rate": 4.425937191658869e-06, + "loss": 0.2276, + "step": 52329 + }, + { + "epoch": 3.6312539032683366, + "grad_norm": 0.831112265586853, + "learning_rate": 4.4242842865312e-06, + "loss": 0.1954, + "step": 52330 + }, + { + "epoch": 3.6313232947054335, + "grad_norm": 1.4287467002868652, + "learning_rate": 4.4226316831249785e-06, + "loss": 0.3676, + "step": 52331 + }, + { + "epoch": 3.63139268614253, + "grad_norm": 0.9742304682731628, + "learning_rate": 4.420979381445423e-06, + "loss": 0.2822, + "step": 52332 + }, + { + "epoch": 3.6314620775796267, + "grad_norm": 0.9230012893676758, + "learning_rate": 4.41932738149774e-06, + "loss": 0.3512, + "step": 52333 + }, + { + "epoch": 3.631531469016723, + "grad_norm": 1.4514968395233154, + "learning_rate": 4.4176756832871705e-06, + "loss": 0.2712, + "step": 52334 + }, + { + "epoch": 3.63160086045382, + "grad_norm": 0.9738816618919373, + "learning_rate": 4.416024286818887e-06, + "loss": 0.2846, + "step": 52335 + }, + { + "epoch": 3.6316702518909167, + "grad_norm": 1.050268292427063, + "learning_rate": 4.414373192098131e-06, + "loss": 0.218, + "step": 52336 + }, + { + "epoch": 3.6317396433280136, + "grad_norm": 0.6958178877830505, + "learning_rate": 4.412722399130109e-06, + "loss": 0.2144, + "step": 52337 + }, + { + "epoch": 3.63180903476511, + "grad_norm": 0.8745349645614624, + "learning_rate": 4.411071907920028e-06, + "loss": 0.2082, + "step": 52338 + }, + { + "epoch": 3.631878426202207, + "grad_norm": 1.1356323957443237, + "learning_rate": 4.409421718473106e-06, + "loss": 0.2046, + "step": 52339 + }, + { + "epoch": 3.631947817639303, + "grad_norm": 0.9653481841087341, + "learning_rate": 4.407771830794549e-06, + "loss": 0.1654, + "step": 52340 + }, + { + "epoch": 3.6320172090764, + "grad_norm": 0.9466253519058228, + "learning_rate": 4.406122244889577e-06, + "loss": 0.2316, + "step": 52341 + }, + { + "epoch": 3.632086600513497, + "grad_norm": 1.1687562465667725, + "learning_rate": 4.4044729607633724e-06, + "loss": 0.4209, + "step": 52342 + }, + { + "epoch": 3.6321559919505932, + "grad_norm": 0.7093731164932251, + "learning_rate": 4.402823978421156e-06, + "loss": 0.1766, + "step": 52343 + }, + { + "epoch": 3.63222538338769, + "grad_norm": 1.1086251735687256, + "learning_rate": 4.401175297868132e-06, + "loss": 0.4087, + "step": 52344 + }, + { + "epoch": 3.6322947748247865, + "grad_norm": 0.9896039962768555, + "learning_rate": 4.399526919109509e-06, + "loss": 0.2793, + "step": 52345 + }, + { + "epoch": 3.6323641662618833, + "grad_norm": 1.1707707643508911, + "learning_rate": 4.397878842150482e-06, + "loss": 0.3315, + "step": 52346 + }, + { + "epoch": 3.63243355769898, + "grad_norm": 0.9646037817001343, + "learning_rate": 4.39623106699627e-06, + "loss": 0.2365, + "step": 52347 + }, + { + "epoch": 3.6325029491360765, + "grad_norm": 1.438986897468567, + "learning_rate": 4.394583593652068e-06, + "loss": 0.2561, + "step": 52348 + }, + { + "epoch": 3.6325723405731734, + "grad_norm": 1.2843009233474731, + "learning_rate": 4.392936422123073e-06, + "loss": 0.2265, + "step": 52349 + }, + { + "epoch": 3.6326417320102697, + "grad_norm": 0.6748862862586975, + "learning_rate": 4.3912895524144685e-06, + "loss": 0.1265, + "step": 52350 + }, + { + "epoch": 3.6327111234473666, + "grad_norm": 0.8807591199874878, + "learning_rate": 4.389642984531484e-06, + "loss": 0.2026, + "step": 52351 + }, + { + "epoch": 3.6327805148844634, + "grad_norm": 0.9469696283340454, + "learning_rate": 4.387996718479303e-06, + "loss": 0.337, + "step": 52352 + }, + { + "epoch": 3.63284990632156, + "grad_norm": 1.0774332284927368, + "learning_rate": 4.386350754263125e-06, + "loss": 0.299, + "step": 52353 + }, + { + "epoch": 3.6329192977586566, + "grad_norm": 0.8441876173019409, + "learning_rate": 4.384705091888131e-06, + "loss": 0.2431, + "step": 52354 + }, + { + "epoch": 3.632988689195753, + "grad_norm": 1.1944377422332764, + "learning_rate": 4.383059731359551e-06, + "loss": 0.3056, + "step": 52355 + }, + { + "epoch": 3.63305808063285, + "grad_norm": 0.9753178358078003, + "learning_rate": 4.381414672682549e-06, + "loss": 0.2465, + "step": 52356 + }, + { + "epoch": 3.6331274720699467, + "grad_norm": 1.1339387893676758, + "learning_rate": 4.3797699158623196e-06, + "loss": 0.2732, + "step": 52357 + }, + { + "epoch": 3.633196863507043, + "grad_norm": 0.6608996391296387, + "learning_rate": 4.378125460904059e-06, + "loss": 0.1741, + "step": 52358 + }, + { + "epoch": 3.63326625494414, + "grad_norm": 1.0478278398513794, + "learning_rate": 4.376481307812974e-06, + "loss": 0.2999, + "step": 52359 + }, + { + "epoch": 3.6333356463812363, + "grad_norm": 0.9558433294296265, + "learning_rate": 4.374837456594239e-06, + "loss": 0.2839, + "step": 52360 + }, + { + "epoch": 3.633405037818333, + "grad_norm": 1.0186442136764526, + "learning_rate": 4.373193907253037e-06, + "loss": 0.2841, + "step": 52361 + }, + { + "epoch": 3.63347442925543, + "grad_norm": 0.8455929160118103, + "learning_rate": 4.371550659794588e-06, + "loss": 0.2608, + "step": 52362 + }, + { + "epoch": 3.633543820692527, + "grad_norm": 1.3513073921203613, + "learning_rate": 4.369907714224053e-06, + "loss": 0.1842, + "step": 52363 + }, + { + "epoch": 3.633613212129623, + "grad_norm": 1.1165941953659058, + "learning_rate": 4.368265070546606e-06, + "loss": 0.2557, + "step": 52364 + }, + { + "epoch": 3.63368260356672, + "grad_norm": 0.8567587733268738, + "learning_rate": 4.366622728767466e-06, + "loss": 0.21, + "step": 52365 + }, + { + "epoch": 3.6337519950038164, + "grad_norm": 1.4013971090316772, + "learning_rate": 4.364980688891795e-06, + "loss": 0.4648, + "step": 52366 + }, + { + "epoch": 3.6338213864409132, + "grad_norm": 1.1490851640701294, + "learning_rate": 4.363338950924789e-06, + "loss": 0.2039, + "step": 52367 + }, + { + "epoch": 3.63389077787801, + "grad_norm": 1.3228968381881714, + "learning_rate": 4.361697514871632e-06, + "loss": 0.31, + "step": 52368 + }, + { + "epoch": 3.6339601693151065, + "grad_norm": 0.7291303277015686, + "learning_rate": 4.360056380737499e-06, + "loss": 0.1441, + "step": 52369 + }, + { + "epoch": 3.6340295607522033, + "grad_norm": 0.7811009883880615, + "learning_rate": 4.3584155485275615e-06, + "loss": 0.2692, + "step": 52370 + }, + { + "epoch": 3.6340989521892997, + "grad_norm": 0.805019199848175, + "learning_rate": 4.356775018247017e-06, + "loss": 0.2092, + "step": 52371 + }, + { + "epoch": 3.6341683436263965, + "grad_norm": 0.9653880596160889, + "learning_rate": 4.355134789901039e-06, + "loss": 0.1201, + "step": 52372 + }, + { + "epoch": 3.6342377350634933, + "grad_norm": 0.7437793612480164, + "learning_rate": 4.353494863494789e-06, + "loss": 0.2143, + "step": 52373 + }, + { + "epoch": 3.6343071265005897, + "grad_norm": 0.7946255207061768, + "learning_rate": 4.351855239033475e-06, + "loss": 0.1896, + "step": 52374 + }, + { + "epoch": 3.6343765179376866, + "grad_norm": 0.9970809817314148, + "learning_rate": 4.350215916522249e-06, + "loss": 0.2853, + "step": 52375 + }, + { + "epoch": 3.634445909374783, + "grad_norm": 1.5506770610809326, + "learning_rate": 4.348576895966305e-06, + "loss": 0.3474, + "step": 52376 + }, + { + "epoch": 3.63451530081188, + "grad_norm": 1.0831838846206665, + "learning_rate": 4.346938177370796e-06, + "loss": 0.2281, + "step": 52377 + }, + { + "epoch": 3.6345846922489766, + "grad_norm": 0.7539243102073669, + "learning_rate": 4.345299760740906e-06, + "loss": 0.2167, + "step": 52378 + }, + { + "epoch": 3.634654083686073, + "grad_norm": 0.6930408477783203, + "learning_rate": 4.343661646081798e-06, + "loss": 0.2183, + "step": 52379 + }, + { + "epoch": 3.63472347512317, + "grad_norm": 0.8650979995727539, + "learning_rate": 4.3420238333986675e-06, + "loss": 0.2258, + "step": 52380 + }, + { + "epoch": 3.6347928665602662, + "grad_norm": 1.105294942855835, + "learning_rate": 4.340386322696654e-06, + "loss": 0.236, + "step": 52381 + }, + { + "epoch": 3.634862257997363, + "grad_norm": 1.0906975269317627, + "learning_rate": 4.338749113980956e-06, + "loss": 0.2938, + "step": 52382 + }, + { + "epoch": 3.63493164943446, + "grad_norm": 1.1567295789718628, + "learning_rate": 4.337112207256733e-06, + "loss": 0.2821, + "step": 52383 + }, + { + "epoch": 3.6350010408715563, + "grad_norm": 1.0957908630371094, + "learning_rate": 4.335475602529138e-06, + "loss": 0.3148, + "step": 52384 + }, + { + "epoch": 3.635070432308653, + "grad_norm": 0.9215257167816162, + "learning_rate": 4.333839299803344e-06, + "loss": 0.2047, + "step": 52385 + }, + { + "epoch": 3.6351398237457495, + "grad_norm": 0.8190645575523376, + "learning_rate": 4.332203299084525e-06, + "loss": 0.3285, + "step": 52386 + }, + { + "epoch": 3.6352092151828463, + "grad_norm": 0.7398127913475037, + "learning_rate": 4.330567600377844e-06, + "loss": 0.2409, + "step": 52387 + }, + { + "epoch": 3.635278606619943, + "grad_norm": 1.0074865818023682, + "learning_rate": 4.328932203688463e-06, + "loss": 0.2382, + "step": 52388 + }, + { + "epoch": 3.63534799805704, + "grad_norm": 1.4400566816329956, + "learning_rate": 4.327297109021533e-06, + "loss": 0.2599, + "step": 52389 + }, + { + "epoch": 3.6354173894941364, + "grad_norm": 1.089272379875183, + "learning_rate": 4.32566231638224e-06, + "loss": 0.2196, + "step": 52390 + }, + { + "epoch": 3.6354867809312332, + "grad_norm": 1.1263757944107056, + "learning_rate": 4.324027825775723e-06, + "loss": 0.3607, + "step": 52391 + }, + { + "epoch": 3.6355561723683296, + "grad_norm": 1.2119436264038086, + "learning_rate": 4.322393637207145e-06, + "loss": 0.273, + "step": 52392 + }, + { + "epoch": 3.6356255638054265, + "grad_norm": 1.1118402481079102, + "learning_rate": 4.3207597506816685e-06, + "loss": 0.2596, + "step": 52393 + }, + { + "epoch": 3.6356949552425233, + "grad_norm": 0.6525857448577881, + "learning_rate": 4.319126166204446e-06, + "loss": 0.2754, + "step": 52394 + }, + { + "epoch": 3.6357643466796197, + "grad_norm": 0.9338521361351013, + "learning_rate": 4.3174928837806496e-06, + "loss": 0.3212, + "step": 52395 + }, + { + "epoch": 3.6358337381167165, + "grad_norm": 1.2208025455474854, + "learning_rate": 4.315859903415431e-06, + "loss": 0.3553, + "step": 52396 + }, + { + "epoch": 3.635903129553813, + "grad_norm": 0.9224134087562561, + "learning_rate": 4.314227225113954e-06, + "loss": 0.2573, + "step": 52397 + }, + { + "epoch": 3.6359725209909097, + "grad_norm": 0.9706383943557739, + "learning_rate": 4.312594848881335e-06, + "loss": 0.2754, + "step": 52398 + }, + { + "epoch": 3.6360419124280066, + "grad_norm": 0.9477553367614746, + "learning_rate": 4.310962774722771e-06, + "loss": 0.2371, + "step": 52399 + }, + { + "epoch": 3.636111303865103, + "grad_norm": 0.5877189040184021, + "learning_rate": 4.30933100264338e-06, + "loss": 0.1859, + "step": 52400 + }, + { + "epoch": 3.6361806953022, + "grad_norm": 0.8833422660827637, + "learning_rate": 4.307699532648335e-06, + "loss": 0.2496, + "step": 52401 + }, + { + "epoch": 3.636250086739296, + "grad_norm": 0.8650872111320496, + "learning_rate": 4.306068364742777e-06, + "loss": 0.2234, + "step": 52402 + }, + { + "epoch": 3.636319478176393, + "grad_norm": 0.8432390093803406, + "learning_rate": 4.3044374989318795e-06, + "loss": 0.2467, + "step": 52403 + }, + { + "epoch": 3.63638886961349, + "grad_norm": 0.6485939025878906, + "learning_rate": 4.302806935220749e-06, + "loss": 0.2065, + "step": 52404 + }, + { + "epoch": 3.6364582610505862, + "grad_norm": 1.1370189189910889, + "learning_rate": 4.301176673614549e-06, + "loss": 0.4972, + "step": 52405 + }, + { + "epoch": 3.636527652487683, + "grad_norm": 0.5732446312904358, + "learning_rate": 4.2995467141184405e-06, + "loss": 0.1828, + "step": 52406 + }, + { + "epoch": 3.6365970439247794, + "grad_norm": 0.8472148776054382, + "learning_rate": 4.297917056737555e-06, + "loss": 0.2389, + "step": 52407 + }, + { + "epoch": 3.6366664353618763, + "grad_norm": 1.1948997974395752, + "learning_rate": 4.296287701477053e-06, + "loss": 0.414, + "step": 52408 + }, + { + "epoch": 3.636735826798973, + "grad_norm": 1.1650097370147705, + "learning_rate": 4.2946586483420536e-06, + "loss": 0.3316, + "step": 52409 + }, + { + "epoch": 3.6368052182360695, + "grad_norm": 0.7007487416267395, + "learning_rate": 4.29302989733773e-06, + "loss": 0.2894, + "step": 52410 + }, + { + "epoch": 3.6368746096731663, + "grad_norm": 0.9079873561859131, + "learning_rate": 4.2914014484692015e-06, + "loss": 0.2365, + "step": 52411 + }, + { + "epoch": 3.6369440011102627, + "grad_norm": 1.0991379022598267, + "learning_rate": 4.289773301741607e-06, + "loss": 0.196, + "step": 52412 + }, + { + "epoch": 3.6370133925473596, + "grad_norm": 1.1891539096832275, + "learning_rate": 4.288145457160086e-06, + "loss": 0.2616, + "step": 52413 + }, + { + "epoch": 3.6370827839844564, + "grad_norm": 0.8817440271377563, + "learning_rate": 4.286517914729804e-06, + "loss": 0.1935, + "step": 52414 + }, + { + "epoch": 3.6371521754215532, + "grad_norm": 0.7709471583366394, + "learning_rate": 4.284890674455866e-06, + "loss": 0.2316, + "step": 52415 + }, + { + "epoch": 3.6372215668586496, + "grad_norm": 0.9476889371871948, + "learning_rate": 4.2832637363434235e-06, + "loss": 0.2945, + "step": 52416 + }, + { + "epoch": 3.6372909582957464, + "grad_norm": 1.6911296844482422, + "learning_rate": 4.281637100397629e-06, + "loss": 0.3261, + "step": 52417 + }, + { + "epoch": 3.637360349732843, + "grad_norm": 0.7360129356384277, + "learning_rate": 4.280010766623588e-06, + "loss": 0.3006, + "step": 52418 + }, + { + "epoch": 3.6374297411699397, + "grad_norm": 1.6646910905838013, + "learning_rate": 4.278384735026453e-06, + "loss": 0.3264, + "step": 52419 + }, + { + "epoch": 3.6374991326070365, + "grad_norm": 0.7221375703811646, + "learning_rate": 4.2767590056113415e-06, + "loss": 0.2208, + "step": 52420 + }, + { + "epoch": 3.637568524044133, + "grad_norm": 1.0290992259979248, + "learning_rate": 4.275133578383406e-06, + "loss": 0.1883, + "step": 52421 + }, + { + "epoch": 3.6376379154812297, + "grad_norm": 0.9339109063148499, + "learning_rate": 4.273508453347763e-06, + "loss": 0.264, + "step": 52422 + }, + { + "epoch": 3.637707306918326, + "grad_norm": 0.7521885633468628, + "learning_rate": 4.2718836305095545e-06, + "loss": 0.2523, + "step": 52423 + }, + { + "epoch": 3.637776698355423, + "grad_norm": 0.8446951508522034, + "learning_rate": 4.270259109873887e-06, + "loss": 0.2522, + "step": 52424 + }, + { + "epoch": 3.6378460897925198, + "grad_norm": 0.6572237014770508, + "learning_rate": 4.268634891445933e-06, + "loss": 0.1862, + "step": 52425 + }, + { + "epoch": 3.637915481229616, + "grad_norm": 0.8397490382194519, + "learning_rate": 4.267010975230767e-06, + "loss": 0.2497, + "step": 52426 + }, + { + "epoch": 3.637984872666713, + "grad_norm": 1.271338701248169, + "learning_rate": 4.2653873612335414e-06, + "loss": 0.3309, + "step": 52427 + }, + { + "epoch": 3.6380542641038094, + "grad_norm": 1.1279913187026978, + "learning_rate": 4.263764049459384e-06, + "loss": 0.3498, + "step": 52428 + }, + { + "epoch": 3.638123655540906, + "grad_norm": 1.1277835369110107, + "learning_rate": 4.262141039913415e-06, + "loss": 0.272, + "step": 52429 + }, + { + "epoch": 3.638193046978003, + "grad_norm": 1.465672492980957, + "learning_rate": 4.26051833260076e-06, + "loss": 0.169, + "step": 52430 + }, + { + "epoch": 3.6382624384150994, + "grad_norm": 0.7960318922996521, + "learning_rate": 4.258895927526551e-06, + "loss": 0.2123, + "step": 52431 + }, + { + "epoch": 3.6383318298521963, + "grad_norm": 0.6890260577201843, + "learning_rate": 4.257273824695896e-06, + "loss": 0.3306, + "step": 52432 + }, + { + "epoch": 3.6384012212892927, + "grad_norm": 0.6680052876472473, + "learning_rate": 4.25565202411391e-06, + "loss": 0.2145, + "step": 52433 + }, + { + "epoch": 3.6384706127263895, + "grad_norm": 1.8317581415176392, + "learning_rate": 4.254030525785724e-06, + "loss": 0.3371, + "step": 52434 + }, + { + "epoch": 3.6385400041634863, + "grad_norm": 0.813218891620636, + "learning_rate": 4.252409329716456e-06, + "loss": 0.2393, + "step": 52435 + }, + { + "epoch": 3.638609395600583, + "grad_norm": 0.8364167213439941, + "learning_rate": 4.250788435911224e-06, + "loss": 0.2636, + "step": 52436 + }, + { + "epoch": 3.6386787870376796, + "grad_norm": 0.7308928966522217, + "learning_rate": 4.2491678443751345e-06, + "loss": 0.2002, + "step": 52437 + }, + { + "epoch": 3.638748178474776, + "grad_norm": 0.8976613283157349, + "learning_rate": 4.24754755511334e-06, + "loss": 0.2983, + "step": 52438 + }, + { + "epoch": 3.6388175699118728, + "grad_norm": 0.8347076773643494, + "learning_rate": 4.245927568130903e-06, + "loss": 0.1598, + "step": 52439 + }, + { + "epoch": 3.6388869613489696, + "grad_norm": 1.1641160249710083, + "learning_rate": 4.2443078834329745e-06, + "loss": 0.2931, + "step": 52440 + }, + { + "epoch": 3.6389563527860664, + "grad_norm": 0.6302094459533691, + "learning_rate": 4.24268850102465e-06, + "loss": 0.2228, + "step": 52441 + }, + { + "epoch": 3.639025744223163, + "grad_norm": 0.7941071391105652, + "learning_rate": 4.241069420911048e-06, + "loss": 0.1649, + "step": 52442 + }, + { + "epoch": 3.6390951356602597, + "grad_norm": 0.7582663893699646, + "learning_rate": 4.239450643097275e-06, + "loss": 0.1769, + "step": 52443 + }, + { + "epoch": 3.639164527097356, + "grad_norm": 0.6503702998161316, + "learning_rate": 4.237832167588462e-06, + "loss": 0.2258, + "step": 52444 + }, + { + "epoch": 3.639233918534453, + "grad_norm": 0.8410419821739197, + "learning_rate": 4.236213994389704e-06, + "loss": 0.1983, + "step": 52445 + }, + { + "epoch": 3.6393033099715497, + "grad_norm": 1.0524342060089111, + "learning_rate": 4.234596123506096e-06, + "loss": 0.2374, + "step": 52446 + }, + { + "epoch": 3.639372701408646, + "grad_norm": 0.7730682492256165, + "learning_rate": 4.232978554942757e-06, + "loss": 0.2163, + "step": 52447 + }, + { + "epoch": 3.639442092845743, + "grad_norm": 0.7991638779640198, + "learning_rate": 4.231361288704794e-06, + "loss": 0.379, + "step": 52448 + }, + { + "epoch": 3.6395114842828393, + "grad_norm": 0.6419838070869446, + "learning_rate": 4.229744324797314e-06, + "loss": 0.2151, + "step": 52449 + }, + { + "epoch": 3.639580875719936, + "grad_norm": 1.0048530101776123, + "learning_rate": 4.228127663225412e-06, + "loss": 0.2657, + "step": 52450 + }, + { + "epoch": 3.639650267157033, + "grad_norm": 0.9744511246681213, + "learning_rate": 4.226511303994208e-06, + "loss": 0.225, + "step": 52451 + }, + { + "epoch": 3.6397196585941294, + "grad_norm": 0.9145474433898926, + "learning_rate": 4.224895247108807e-06, + "loss": 0.2107, + "step": 52452 + }, + { + "epoch": 3.639789050031226, + "grad_norm": 0.709636390209198, + "learning_rate": 4.223279492574284e-06, + "loss": 0.1891, + "step": 52453 + }, + { + "epoch": 3.6398584414683226, + "grad_norm": 0.6785569787025452, + "learning_rate": 4.221664040395756e-06, + "loss": 0.1191, + "step": 52454 + }, + { + "epoch": 3.6399278329054194, + "grad_norm": 0.8631740808486938, + "learning_rate": 4.2200488905783206e-06, + "loss": 0.1994, + "step": 52455 + }, + { + "epoch": 3.6399972243425163, + "grad_norm": 0.8629910349845886, + "learning_rate": 4.218434043127095e-06, + "loss": 0.1949, + "step": 52456 + }, + { + "epoch": 3.6400666157796127, + "grad_norm": 0.9291731715202332, + "learning_rate": 4.216819498047153e-06, + "loss": 0.2542, + "step": 52457 + }, + { + "epoch": 3.6401360072167095, + "grad_norm": 1.189994215965271, + "learning_rate": 4.2152052553435904e-06, + "loss": 0.2057, + "step": 52458 + }, + { + "epoch": 3.640205398653806, + "grad_norm": 1.3394224643707275, + "learning_rate": 4.213591315021514e-06, + "loss": 0.3043, + "step": 52459 + }, + { + "epoch": 3.6402747900909027, + "grad_norm": 0.6954841017723083, + "learning_rate": 4.211977677086032e-06, + "loss": 0.2832, + "step": 52460 + }, + { + "epoch": 3.6403441815279995, + "grad_norm": 0.7604264616966248, + "learning_rate": 4.210364341542206e-06, + "loss": 0.2106, + "step": 52461 + }, + { + "epoch": 3.6404135729650964, + "grad_norm": 1.128056526184082, + "learning_rate": 4.208751308395154e-06, + "loss": 0.2922, + "step": 52462 + }, + { + "epoch": 3.6404829644021928, + "grad_norm": 1.233444094657898, + "learning_rate": 4.207138577649949e-06, + "loss": 0.3515, + "step": 52463 + }, + { + "epoch": 3.640552355839289, + "grad_norm": 0.8787344098091125, + "learning_rate": 4.2055261493117e-06, + "loss": 0.22, + "step": 52464 + }, + { + "epoch": 3.640621747276386, + "grad_norm": 0.7083876729011536, + "learning_rate": 4.203914023385491e-06, + "loss": 0.253, + "step": 52465 + }, + { + "epoch": 3.640691138713483, + "grad_norm": 0.674351692199707, + "learning_rate": 4.202302199876418e-06, + "loss": 0.2241, + "step": 52466 + }, + { + "epoch": 3.6407605301505797, + "grad_norm": 0.6310250163078308, + "learning_rate": 4.200690678789554e-06, + "loss": 0.1829, + "step": 52467 + }, + { + "epoch": 3.640829921587676, + "grad_norm": 0.8792396783828735, + "learning_rate": 4.199079460129996e-06, + "loss": 0.2843, + "step": 52468 + }, + { + "epoch": 3.640899313024773, + "grad_norm": 0.8438823223114014, + "learning_rate": 4.197468543902827e-06, + "loss": 0.1583, + "step": 52469 + }, + { + "epoch": 3.6409687044618693, + "grad_norm": 0.7861741185188293, + "learning_rate": 4.195857930113134e-06, + "loss": 0.2026, + "step": 52470 + }, + { + "epoch": 3.641038095898966, + "grad_norm": 1.0024135112762451, + "learning_rate": 4.194247618766001e-06, + "loss": 0.2298, + "step": 52471 + }, + { + "epoch": 3.641107487336063, + "grad_norm": 1.057131290435791, + "learning_rate": 4.192637609866523e-06, + "loss": 0.3911, + "step": 52472 + }, + { + "epoch": 3.6411768787731593, + "grad_norm": 0.775271475315094, + "learning_rate": 4.1910279034197754e-06, + "loss": 0.1892, + "step": 52473 + }, + { + "epoch": 3.641246270210256, + "grad_norm": 0.7275716662406921, + "learning_rate": 4.189418499430819e-06, + "loss": 0.2244, + "step": 52474 + }, + { + "epoch": 3.6413156616473525, + "grad_norm": 0.8524147272109985, + "learning_rate": 4.1878093979047625e-06, + "loss": 0.3192, + "step": 52475 + }, + { + "epoch": 3.6413850530844494, + "grad_norm": 1.1745033264160156, + "learning_rate": 4.186200598846679e-06, + "loss": 0.3563, + "step": 52476 + }, + { + "epoch": 3.641454444521546, + "grad_norm": 0.753790020942688, + "learning_rate": 4.184592102261631e-06, + "loss": 0.2294, + "step": 52477 + }, + { + "epoch": 3.6415238359586426, + "grad_norm": 0.9149582386016846, + "learning_rate": 4.182983908154714e-06, + "loss": 0.2484, + "step": 52478 + }, + { + "epoch": 3.6415932273957394, + "grad_norm": 0.9101495146751404, + "learning_rate": 4.181376016531013e-06, + "loss": 0.2691, + "step": 52479 + }, + { + "epoch": 3.641662618832836, + "grad_norm": 1.0201491117477417, + "learning_rate": 4.179768427395581e-06, + "loss": 0.2365, + "step": 52480 + }, + { + "epoch": 3.6417320102699327, + "grad_norm": 0.8526895642280579, + "learning_rate": 4.1781611407535e-06, + "loss": 0.1787, + "step": 52481 + }, + { + "epoch": 3.6418014017070295, + "grad_norm": 1.2286587953567505, + "learning_rate": 4.176554156609847e-06, + "loss": 0.2394, + "step": 52482 + }, + { + "epoch": 3.641870793144126, + "grad_norm": 0.6906652450561523, + "learning_rate": 4.174947474969703e-06, + "loss": 0.1343, + "step": 52483 + }, + { + "epoch": 3.6419401845812227, + "grad_norm": 1.0088930130004883, + "learning_rate": 4.173341095838124e-06, + "loss": 0.3038, + "step": 52484 + }, + { + "epoch": 3.642009576018319, + "grad_norm": 1.1745655536651611, + "learning_rate": 4.171735019220191e-06, + "loss": 0.2611, + "step": 52485 + }, + { + "epoch": 3.642078967455416, + "grad_norm": 1.7688636779785156, + "learning_rate": 4.170129245120979e-06, + "loss": 0.2355, + "step": 52486 + }, + { + "epoch": 3.6421483588925128, + "grad_norm": 1.316643238067627, + "learning_rate": 4.168523773545541e-06, + "loss": 0.338, + "step": 52487 + }, + { + "epoch": 3.6422177503296096, + "grad_norm": 0.9058511257171631, + "learning_rate": 4.166918604498959e-06, + "loss": 0.2422, + "step": 52488 + }, + { + "epoch": 3.642287141766706, + "grad_norm": 0.7852573990821838, + "learning_rate": 4.165313737986299e-06, + "loss": 0.2067, + "step": 52489 + }, + { + "epoch": 3.642356533203803, + "grad_norm": 0.6199080348014832, + "learning_rate": 4.16370917401262e-06, + "loss": 0.1613, + "step": 52490 + }, + { + "epoch": 3.642425924640899, + "grad_norm": 0.9062686562538147, + "learning_rate": 4.162104912582998e-06, + "loss": 0.2654, + "step": 52491 + }, + { + "epoch": 3.642495316077996, + "grad_norm": 1.0203871726989746, + "learning_rate": 4.160500953702484e-06, + "loss": 0.2862, + "step": 52492 + }, + { + "epoch": 3.642564707515093, + "grad_norm": 0.9074485301971436, + "learning_rate": 4.158897297376152e-06, + "loss": 0.2467, + "step": 52493 + }, + { + "epoch": 3.6426340989521893, + "grad_norm": 0.7271792888641357, + "learning_rate": 4.157293943609064e-06, + "loss": 0.1637, + "step": 52494 + }, + { + "epoch": 3.642703490389286, + "grad_norm": 0.6246678233146667, + "learning_rate": 4.155690892406283e-06, + "loss": 0.2139, + "step": 52495 + }, + { + "epoch": 3.6427728818263825, + "grad_norm": 0.9170033931732178, + "learning_rate": 4.1540881437728604e-06, + "loss": 0.3244, + "step": 52496 + }, + { + "epoch": 3.6428422732634793, + "grad_norm": 0.6776948571205139, + "learning_rate": 4.15248569771386e-06, + "loss": 0.2499, + "step": 52497 + }, + { + "epoch": 3.642911664700576, + "grad_norm": 1.0156017541885376, + "learning_rate": 4.150883554234342e-06, + "loss": 0.1832, + "step": 52498 + }, + { + "epoch": 3.6429810561376725, + "grad_norm": 0.8274765610694885, + "learning_rate": 4.149281713339359e-06, + "loss": 0.2128, + "step": 52499 + }, + { + "epoch": 3.6430504475747694, + "grad_norm": 0.7911663055419922, + "learning_rate": 4.147680175033996e-06, + "loss": 0.1834, + "step": 52500 + }, + { + "epoch": 3.6431198390118658, + "grad_norm": 0.8715351223945618, + "learning_rate": 4.146078939323262e-06, + "loss": 0.2982, + "step": 52501 + }, + { + "epoch": 3.6431892304489626, + "grad_norm": 1.0350953340530396, + "learning_rate": 4.1444780062122494e-06, + "loss": 0.2519, + "step": 52502 + }, + { + "epoch": 3.6432586218860594, + "grad_norm": 0.871137261390686, + "learning_rate": 4.142877375705989e-06, + "loss": 0.253, + "step": 52503 + }, + { + "epoch": 3.643328013323156, + "grad_norm": 0.9455471038818359, + "learning_rate": 4.141277047809544e-06, + "loss": 0.2303, + "step": 52504 + }, + { + "epoch": 3.6433974047602526, + "grad_norm": 0.8095040321350098, + "learning_rate": 4.139677022527966e-06, + "loss": 0.2113, + "step": 52505 + }, + { + "epoch": 3.643466796197349, + "grad_norm": 1.1507304906845093, + "learning_rate": 4.138077299866306e-06, + "loss": 0.3485, + "step": 52506 + }, + { + "epoch": 3.643536187634446, + "grad_norm": 0.8429651260375977, + "learning_rate": 4.136477879829615e-06, + "loss": 0.1864, + "step": 52507 + }, + { + "epoch": 3.6436055790715427, + "grad_norm": 1.1227906942367554, + "learning_rate": 4.1348787624229446e-06, + "loss": 0.1771, + "step": 52508 + }, + { + "epoch": 3.643674970508639, + "grad_norm": 0.9887437224388123, + "learning_rate": 4.133279947651336e-06, + "loss": 0.259, + "step": 52509 + }, + { + "epoch": 3.643744361945736, + "grad_norm": 1.1400277614593506, + "learning_rate": 4.1316814355198405e-06, + "loss": 0.2013, + "step": 52510 + }, + { + "epoch": 3.6438137533828323, + "grad_norm": 0.7854551672935486, + "learning_rate": 4.130083226033499e-06, + "loss": 0.216, + "step": 52511 + }, + { + "epoch": 3.643883144819929, + "grad_norm": 0.755085289478302, + "learning_rate": 4.128485319197361e-06, + "loss": 0.2522, + "step": 52512 + }, + { + "epoch": 3.643952536257026, + "grad_norm": 0.9310125112533569, + "learning_rate": 4.1268877150164806e-06, + "loss": 0.2437, + "step": 52513 + }, + { + "epoch": 3.644021927694123, + "grad_norm": 1.0088772773742676, + "learning_rate": 4.125290413495897e-06, + "loss": 0.2286, + "step": 52514 + }, + { + "epoch": 3.644091319131219, + "grad_norm": 1.2126679420471191, + "learning_rate": 4.123693414640639e-06, + "loss": 0.2982, + "step": 52515 + }, + { + "epoch": 3.644160710568316, + "grad_norm": 0.7529183030128479, + "learning_rate": 4.122096718455748e-06, + "loss": 0.206, + "step": 52516 + }, + { + "epoch": 3.6442301020054124, + "grad_norm": 0.8695913553237915, + "learning_rate": 4.120500324946286e-06, + "loss": 0.2208, + "step": 52517 + }, + { + "epoch": 3.6442994934425093, + "grad_norm": 0.6055811643600464, + "learning_rate": 4.118904234117282e-06, + "loss": 0.2019, + "step": 52518 + }, + { + "epoch": 3.644368884879606, + "grad_norm": 0.8467153310775757, + "learning_rate": 4.1173084459737664e-06, + "loss": 0.2666, + "step": 52519 + }, + { + "epoch": 3.6444382763167025, + "grad_norm": 1.169858455657959, + "learning_rate": 4.1157129605207786e-06, + "loss": 0.3485, + "step": 52520 + }, + { + "epoch": 3.6445076677537993, + "grad_norm": 1.1280580759048462, + "learning_rate": 4.114117777763382e-06, + "loss": 0.2979, + "step": 52521 + }, + { + "epoch": 3.6445770591908957, + "grad_norm": 0.9656127691268921, + "learning_rate": 4.112522897706572e-06, + "loss": 0.2229, + "step": 52522 + }, + { + "epoch": 3.6446464506279925, + "grad_norm": 1.1725473403930664, + "learning_rate": 4.110928320355412e-06, + "loss": 0.2594, + "step": 52523 + }, + { + "epoch": 3.6447158420650894, + "grad_norm": 1.1113027334213257, + "learning_rate": 4.109334045714919e-06, + "loss": 0.324, + "step": 52524 + }, + { + "epoch": 3.6447852335021858, + "grad_norm": 0.9031866788864136, + "learning_rate": 4.107740073790134e-06, + "loss": 0.2864, + "step": 52525 + }, + { + "epoch": 3.6448546249392826, + "grad_norm": 0.7612910866737366, + "learning_rate": 4.106146404586097e-06, + "loss": 0.2727, + "step": 52526 + }, + { + "epoch": 3.644924016376379, + "grad_norm": 0.827181875705719, + "learning_rate": 4.104553038107817e-06, + "loss": 0.1711, + "step": 52527 + }, + { + "epoch": 3.644993407813476, + "grad_norm": 0.8759703040122986, + "learning_rate": 4.102959974360343e-06, + "loss": 0.1894, + "step": 52528 + }, + { + "epoch": 3.6450627992505726, + "grad_norm": 0.8325121998786926, + "learning_rate": 4.1013672133486944e-06, + "loss": 0.3531, + "step": 52529 + }, + { + "epoch": 3.645132190687669, + "grad_norm": 0.7958627343177795, + "learning_rate": 4.099774755077923e-06, + "loss": 0.2688, + "step": 52530 + }, + { + "epoch": 3.645201582124766, + "grad_norm": 1.075870394706726, + "learning_rate": 4.0981825995530135e-06, + "loss": 0.3487, + "step": 52531 + }, + { + "epoch": 3.6452709735618622, + "grad_norm": 0.9494314193725586, + "learning_rate": 4.096590746779016e-06, + "loss": 0.2781, + "step": 52532 + }, + { + "epoch": 3.645340364998959, + "grad_norm": 0.9918287992477417, + "learning_rate": 4.094999196760962e-06, + "loss": 0.2549, + "step": 52533 + }, + { + "epoch": 3.645409756436056, + "grad_norm": 0.6830200552940369, + "learning_rate": 4.093407949503858e-06, + "loss": 0.2603, + "step": 52534 + }, + { + "epoch": 3.6454791478731523, + "grad_norm": 0.7208458781242371, + "learning_rate": 4.091817005012766e-06, + "loss": 0.2093, + "step": 52535 + }, + { + "epoch": 3.645548539310249, + "grad_norm": 0.6143688559532166, + "learning_rate": 4.0902263632926485e-06, + "loss": 0.2122, + "step": 52536 + }, + { + "epoch": 3.6456179307473455, + "grad_norm": 1.236505150794983, + "learning_rate": 4.088636024348569e-06, + "loss": 0.3008, + "step": 52537 + }, + { + "epoch": 3.6456873221844424, + "grad_norm": 0.7275641560554504, + "learning_rate": 4.087045988185545e-06, + "loss": 0.1911, + "step": 52538 + }, + { + "epoch": 3.645756713621539, + "grad_norm": 0.7728639841079712, + "learning_rate": 4.085456254808584e-06, + "loss": 0.2588, + "step": 52539 + }, + { + "epoch": 3.645826105058636, + "grad_norm": 0.8163113594055176, + "learning_rate": 4.083866824222704e-06, + "loss": 0.2017, + "step": 52540 + }, + { + "epoch": 3.6458954964957324, + "grad_norm": 0.8081895709037781, + "learning_rate": 4.082277696432923e-06, + "loss": 0.2155, + "step": 52541 + }, + { + "epoch": 3.6459648879328292, + "grad_norm": 0.7549625635147095, + "learning_rate": 4.080688871444294e-06, + "loss": 0.2242, + "step": 52542 + }, + { + "epoch": 3.6460342793699256, + "grad_norm": 0.6661500930786133, + "learning_rate": 4.079100349261777e-06, + "loss": 0.2484, + "step": 52543 + }, + { + "epoch": 3.6461036708070225, + "grad_norm": 0.7815737724304199, + "learning_rate": 4.077512129890404e-06, + "loss": 0.2824, + "step": 52544 + }, + { + "epoch": 3.6461730622441193, + "grad_norm": 0.7753537893295288, + "learning_rate": 4.075924213335214e-06, + "loss": 0.2266, + "step": 52545 + }, + { + "epoch": 3.6462424536812157, + "grad_norm": 0.6838996410369873, + "learning_rate": 4.074336599601191e-06, + "loss": 0.2479, + "step": 52546 + }, + { + "epoch": 3.6463118451183125, + "grad_norm": 0.6542237401008606, + "learning_rate": 4.072749288693356e-06, + "loss": 0.2053, + "step": 52547 + }, + { + "epoch": 3.646381236555409, + "grad_norm": 0.9855672717094421, + "learning_rate": 4.071162280616725e-06, + "loss": 0.2189, + "step": 52548 + }, + { + "epoch": 3.6464506279925057, + "grad_norm": 1.0329219102859497, + "learning_rate": 4.069575575376316e-06, + "loss": 0.2766, + "step": 52549 + }, + { + "epoch": 3.6465200194296026, + "grad_norm": 0.7535037994384766, + "learning_rate": 4.067989172977116e-06, + "loss": 0.2239, + "step": 52550 + }, + { + "epoch": 3.646589410866699, + "grad_norm": 1.236804485321045, + "learning_rate": 4.066403073424152e-06, + "loss": 0.4043, + "step": 52551 + }, + { + "epoch": 3.646658802303796, + "grad_norm": 1.2905718088150024, + "learning_rate": 4.06481727672241e-06, + "loss": 0.214, + "step": 52552 + }, + { + "epoch": 3.646728193740892, + "grad_norm": 0.616075336933136, + "learning_rate": 4.0632317828769194e-06, + "loss": 0.1861, + "step": 52553 + }, + { + "epoch": 3.646797585177989, + "grad_norm": 1.0205124616622925, + "learning_rate": 4.061646591892665e-06, + "loss": 0.3309, + "step": 52554 + }, + { + "epoch": 3.646866976615086, + "grad_norm": 0.8152270913124084, + "learning_rate": 4.060061703774687e-06, + "loss": 0.2518, + "step": 52555 + }, + { + "epoch": 3.6469363680521822, + "grad_norm": 1.2560069561004639, + "learning_rate": 4.0584771185279366e-06, + "loss": 0.335, + "step": 52556 + }, + { + "epoch": 3.647005759489279, + "grad_norm": 1.170482873916626, + "learning_rate": 4.056892836157456e-06, + "loss": 0.2717, + "step": 52557 + }, + { + "epoch": 3.6470751509263755, + "grad_norm": 0.829165518283844, + "learning_rate": 4.0553088566682276e-06, + "loss": 0.2124, + "step": 52558 + }, + { + "epoch": 3.6471445423634723, + "grad_norm": 0.7582311034202576, + "learning_rate": 4.053725180065249e-06, + "loss": 0.3042, + "step": 52559 + }, + { + "epoch": 3.647213933800569, + "grad_norm": 0.9910750985145569, + "learning_rate": 4.052141806353538e-06, + "loss": 0.262, + "step": 52560 + }, + { + "epoch": 3.6472833252376655, + "grad_norm": 1.0738581418991089, + "learning_rate": 4.05055873553809e-06, + "loss": 0.3653, + "step": 52561 + }, + { + "epoch": 3.6473527166747624, + "grad_norm": 0.7442823648452759, + "learning_rate": 4.048975967623892e-06, + "loss": 0.1934, + "step": 52562 + }, + { + "epoch": 3.6474221081118587, + "grad_norm": 0.8207558393478394, + "learning_rate": 4.047393502615937e-06, + "loss": 0.185, + "step": 52563 + }, + { + "epoch": 3.6474914995489556, + "grad_norm": 0.6979028582572937, + "learning_rate": 4.045811340519245e-06, + "loss": 0.1803, + "step": 52564 + }, + { + "epoch": 3.6475608909860524, + "grad_norm": 0.8227285146713257, + "learning_rate": 4.044229481338779e-06, + "loss": 0.2454, + "step": 52565 + }, + { + "epoch": 3.6476302824231492, + "grad_norm": 0.6177524924278259, + "learning_rate": 4.042647925079546e-06, + "loss": 0.2048, + "step": 52566 + }, + { + "epoch": 3.6476996738602456, + "grad_norm": 0.7399325370788574, + "learning_rate": 4.041066671746541e-06, + "loss": 0.2682, + "step": 52567 + }, + { + "epoch": 3.6477690652973425, + "grad_norm": 0.7898706793785095, + "learning_rate": 4.03948572134476e-06, + "loss": 0.2213, + "step": 52568 + }, + { + "epoch": 3.647838456734439, + "grad_norm": 0.9270410537719727, + "learning_rate": 4.0379050738792e-06, + "loss": 0.3783, + "step": 52569 + }, + { + "epoch": 3.6479078481715357, + "grad_norm": 0.708538293838501, + "learning_rate": 4.036324729354824e-06, + "loss": 0.1771, + "step": 52570 + }, + { + "epoch": 3.6479772396086325, + "grad_norm": 0.6957582235336304, + "learning_rate": 4.034744687776637e-06, + "loss": 0.1401, + "step": 52571 + }, + { + "epoch": 3.648046631045729, + "grad_norm": 0.8682592511177063, + "learning_rate": 4.033164949149626e-06, + "loss": 0.2699, + "step": 52572 + }, + { + "epoch": 3.6481160224828257, + "grad_norm": 0.7908797264099121, + "learning_rate": 4.031585513478786e-06, + "loss": 0.2145, + "step": 52573 + }, + { + "epoch": 3.648185413919922, + "grad_norm": 0.982096254825592, + "learning_rate": 4.030006380769091e-06, + "loss": 0.158, + "step": 52574 + }, + { + "epoch": 3.648254805357019, + "grad_norm": 0.7277545928955078, + "learning_rate": 4.028427551025538e-06, + "loss": 0.2045, + "step": 52575 + }, + { + "epoch": 3.648324196794116, + "grad_norm": 0.6687835454940796, + "learning_rate": 4.02684902425311e-06, + "loss": 0.2202, + "step": 52576 + }, + { + "epoch": 3.648393588231212, + "grad_norm": 0.7036941647529602, + "learning_rate": 4.02527080045677e-06, + "loss": 0.1989, + "step": 52577 + }, + { + "epoch": 3.648462979668309, + "grad_norm": 0.7946994304656982, + "learning_rate": 4.023692879641517e-06, + "loss": 0.2398, + "step": 52578 + }, + { + "epoch": 3.6485323711054054, + "grad_norm": 0.787498950958252, + "learning_rate": 4.022115261812331e-06, + "loss": 0.1615, + "step": 52579 + }, + { + "epoch": 3.6486017625425022, + "grad_norm": 0.8422709107398987, + "learning_rate": 4.02053794697419e-06, + "loss": 0.2638, + "step": 52580 + }, + { + "epoch": 3.648671153979599, + "grad_norm": 0.7661809325218201, + "learning_rate": 4.018960935132076e-06, + "loss": 0.1731, + "step": 52581 + }, + { + "epoch": 3.6487405454166955, + "grad_norm": 0.9543895125389099, + "learning_rate": 4.017384226290977e-06, + "loss": 0.2529, + "step": 52582 + }, + { + "epoch": 3.6488099368537923, + "grad_norm": 0.6171339750289917, + "learning_rate": 4.0158078204558525e-06, + "loss": 0.1816, + "step": 52583 + }, + { + "epoch": 3.6488793282908887, + "grad_norm": 0.814397931098938, + "learning_rate": 4.0142317176316894e-06, + "loss": 0.2195, + "step": 52584 + }, + { + "epoch": 3.6489487197279855, + "grad_norm": 1.0327805280685425, + "learning_rate": 4.01265591782346e-06, + "loss": 0.2887, + "step": 52585 + }, + { + "epoch": 3.6490181111650823, + "grad_norm": 0.8630198836326599, + "learning_rate": 4.011080421036129e-06, + "loss": 0.2451, + "step": 52586 + }, + { + "epoch": 3.6490875026021787, + "grad_norm": 0.860249400138855, + "learning_rate": 4.00950522727469e-06, + "loss": 0.2509, + "step": 52587 + }, + { + "epoch": 3.6491568940392756, + "grad_norm": 0.9187492728233337, + "learning_rate": 4.007930336544097e-06, + "loss": 0.2613, + "step": 52588 + }, + { + "epoch": 3.649226285476372, + "grad_norm": 1.5226153135299683, + "learning_rate": 4.006355748849344e-06, + "loss": 0.4719, + "step": 52589 + }, + { + "epoch": 3.649295676913469, + "grad_norm": 1.0031710863113403, + "learning_rate": 4.004781464195384e-06, + "loss": 0.2423, + "step": 52590 + }, + { + "epoch": 3.6493650683505656, + "grad_norm": 0.6121841669082642, + "learning_rate": 4.00320748258719e-06, + "loss": 0.1618, + "step": 52591 + }, + { + "epoch": 3.6494344597876625, + "grad_norm": 0.6395224928855896, + "learning_rate": 4.001633804029736e-06, + "loss": 0.1593, + "step": 52592 + }, + { + "epoch": 3.649503851224759, + "grad_norm": 1.0417262315750122, + "learning_rate": 4.000060428527974e-06, + "loss": 0.2537, + "step": 52593 + }, + { + "epoch": 3.6495732426618557, + "grad_norm": 0.8274183869361877, + "learning_rate": 3.998487356086889e-06, + "loss": 0.246, + "step": 52594 + }, + { + "epoch": 3.649642634098952, + "grad_norm": 0.6725167036056519, + "learning_rate": 3.9969145867114424e-06, + "loss": 0.143, + "step": 52595 + }, + { + "epoch": 3.649712025536049, + "grad_norm": 1.1633943319320679, + "learning_rate": 3.995342120406609e-06, + "loss": 0.2878, + "step": 52596 + }, + { + "epoch": 3.6497814169731457, + "grad_norm": 1.4206368923187256, + "learning_rate": 3.993769957177329e-06, + "loss": 0.3015, + "step": 52597 + }, + { + "epoch": 3.649850808410242, + "grad_norm": 0.706488847732544, + "learning_rate": 3.992198097028588e-06, + "loss": 0.233, + "step": 52598 + }, + { + "epoch": 3.649920199847339, + "grad_norm": 0.9687461853027344, + "learning_rate": 3.9906265399653365e-06, + "loss": 0.3716, + "step": 52599 + }, + { + "epoch": 3.6499895912844353, + "grad_norm": 0.8727436065673828, + "learning_rate": 3.989055285992538e-06, + "loss": 0.1911, + "step": 52600 + }, + { + "epoch": 3.650058982721532, + "grad_norm": 0.8844355940818787, + "learning_rate": 3.987484335115144e-06, + "loss": 0.2507, + "step": 52601 + }, + { + "epoch": 3.650128374158629, + "grad_norm": 0.8614605069160461, + "learning_rate": 3.9859136873381275e-06, + "loss": 0.2827, + "step": 52602 + }, + { + "epoch": 3.6501977655957254, + "grad_norm": 0.928248405456543, + "learning_rate": 3.98434334266643e-06, + "loss": 0.2843, + "step": 52603 + }, + { + "epoch": 3.6502671570328222, + "grad_norm": 0.9715300798416138, + "learning_rate": 3.982773301105047e-06, + "loss": 0.191, + "step": 52604 + }, + { + "epoch": 3.6503365484699186, + "grad_norm": 1.6563992500305176, + "learning_rate": 3.981203562658897e-06, + "loss": 0.2154, + "step": 52605 + }, + { + "epoch": 3.6504059399070155, + "grad_norm": 1.1098893880844116, + "learning_rate": 3.979634127332932e-06, + "loss": 0.2672, + "step": 52606 + }, + { + "epoch": 3.6504753313441123, + "grad_norm": 0.8628815412521362, + "learning_rate": 3.978064995132136e-06, + "loss": 0.1899, + "step": 52607 + }, + { + "epoch": 3.6505447227812087, + "grad_norm": 1.1203254461288452, + "learning_rate": 3.9764961660614385e-06, + "loss": 0.2311, + "step": 52608 + }, + { + "epoch": 3.6506141142183055, + "grad_norm": 0.7917584180831909, + "learning_rate": 3.974927640125814e-06, + "loss": 0.1971, + "step": 52609 + }, + { + "epoch": 3.650683505655402, + "grad_norm": 0.6873453259468079, + "learning_rate": 3.973359417330192e-06, + "loss": 0.2767, + "step": 52610 + }, + { + "epoch": 3.6507528970924987, + "grad_norm": 1.2533938884735107, + "learning_rate": 3.971791497679545e-06, + "loss": 0.2639, + "step": 52611 + }, + { + "epoch": 3.6508222885295956, + "grad_norm": 1.0085850954055786, + "learning_rate": 3.970223881178803e-06, + "loss": 0.3339, + "step": 52612 + }, + { + "epoch": 3.650891679966692, + "grad_norm": 0.8450230956077576, + "learning_rate": 3.968656567832929e-06, + "loss": 0.2434, + "step": 52613 + }, + { + "epoch": 3.650961071403789, + "grad_norm": 1.1059644222259521, + "learning_rate": 3.967089557646853e-06, + "loss": 0.2711, + "step": 52614 + }, + { + "epoch": 3.651030462840885, + "grad_norm": 0.8992096781730652, + "learning_rate": 3.965522850625536e-06, + "loss": 0.2444, + "step": 52615 + }, + { + "epoch": 3.651099854277982, + "grad_norm": 1.152604579925537, + "learning_rate": 3.96395644677392e-06, + "loss": 0.2758, + "step": 52616 + }, + { + "epoch": 3.651169245715079, + "grad_norm": 0.6889544129371643, + "learning_rate": 3.9623903460969555e-06, + "loss": 0.1842, + "step": 52617 + }, + { + "epoch": 3.6512386371521757, + "grad_norm": 0.6382239460945129, + "learning_rate": 3.960824548599595e-06, + "loss": 0.1952, + "step": 52618 + }, + { + "epoch": 3.651308028589272, + "grad_norm": 0.9255589246749878, + "learning_rate": 3.9592590542867565e-06, + "loss": 0.1929, + "step": 52619 + }, + { + "epoch": 3.651377420026369, + "grad_norm": 0.7362818717956543, + "learning_rate": 3.957693863163392e-06, + "loss": 0.1964, + "step": 52620 + }, + { + "epoch": 3.6514468114634653, + "grad_norm": 0.8391072750091553, + "learning_rate": 3.956128975234441e-06, + "loss": 0.2136, + "step": 52621 + }, + { + "epoch": 3.651516202900562, + "grad_norm": 0.6805461049079895, + "learning_rate": 3.954564390504855e-06, + "loss": 0.1716, + "step": 52622 + }, + { + "epoch": 3.651585594337659, + "grad_norm": 0.8736404180526733, + "learning_rate": 3.953000108979566e-06, + "loss": 0.2194, + "step": 52623 + }, + { + "epoch": 3.6516549857747553, + "grad_norm": 0.9125484228134155, + "learning_rate": 3.951436130663511e-06, + "loss": 0.2024, + "step": 52624 + }, + { + "epoch": 3.651724377211852, + "grad_norm": 1.1934622526168823, + "learning_rate": 3.949872455561632e-06, + "loss": 0.2893, + "step": 52625 + }, + { + "epoch": 3.6517937686489486, + "grad_norm": 1.0321385860443115, + "learning_rate": 3.948309083678858e-06, + "loss": 0.3401, + "step": 52626 + }, + { + "epoch": 3.6518631600860454, + "grad_norm": 0.9430120587348938, + "learning_rate": 3.946746015020131e-06, + "loss": 0.2801, + "step": 52627 + }, + { + "epoch": 3.6519325515231422, + "grad_norm": 0.9782058596611023, + "learning_rate": 3.945183249590378e-06, + "loss": 0.2526, + "step": 52628 + }, + { + "epoch": 3.6520019429602386, + "grad_norm": 0.822281539440155, + "learning_rate": 3.943620787394531e-06, + "loss": 0.2593, + "step": 52629 + }, + { + "epoch": 3.6520713343973354, + "grad_norm": 0.8308938145637512, + "learning_rate": 3.942058628437539e-06, + "loss": 0.2453, + "step": 52630 + }, + { + "epoch": 3.652140725834432, + "grad_norm": 0.7942308783531189, + "learning_rate": 3.940496772724323e-06, + "loss": 0.2555, + "step": 52631 + }, + { + "epoch": 3.6522101172715287, + "grad_norm": 0.8300301432609558, + "learning_rate": 3.9389352202598095e-06, + "loss": 0.333, + "step": 52632 + }, + { + "epoch": 3.6522795087086255, + "grad_norm": 0.8790302276611328, + "learning_rate": 3.937373971048941e-06, + "loss": 0.2555, + "step": 52633 + }, + { + "epoch": 3.652348900145722, + "grad_norm": 0.936795175075531, + "learning_rate": 3.935813025096635e-06, + "loss": 0.2806, + "step": 52634 + }, + { + "epoch": 3.6524182915828187, + "grad_norm": 1.226846694946289, + "learning_rate": 3.934252382407822e-06, + "loss": 0.2399, + "step": 52635 + }, + { + "epoch": 3.652487683019915, + "grad_norm": 0.785735011100769, + "learning_rate": 3.932692042987418e-06, + "loss": 0.2432, + "step": 52636 + }, + { + "epoch": 3.652557074457012, + "grad_norm": 1.0602415800094604, + "learning_rate": 3.931132006840366e-06, + "loss": 0.2323, + "step": 52637 + }, + { + "epoch": 3.6526264658941088, + "grad_norm": 0.6603343486785889, + "learning_rate": 3.9295722739715935e-06, + "loss": 0.1621, + "step": 52638 + }, + { + "epoch": 3.652695857331205, + "grad_norm": 0.9817532896995544, + "learning_rate": 3.928012844385998e-06, + "loss": 0.2261, + "step": 52639 + }, + { + "epoch": 3.652765248768302, + "grad_norm": 0.6978317499160767, + "learning_rate": 3.92645371808853e-06, + "loss": 0.1986, + "step": 52640 + }, + { + "epoch": 3.6528346402053984, + "grad_norm": 1.3084534406661987, + "learning_rate": 3.924894895084097e-06, + "loss": 0.2657, + "step": 52641 + }, + { + "epoch": 3.652904031642495, + "grad_norm": 0.9324058294296265, + "learning_rate": 3.923336375377617e-06, + "loss": 0.3408, + "step": 52642 + }, + { + "epoch": 3.652973423079592, + "grad_norm": 1.2977538108825684, + "learning_rate": 3.921778158974032e-06, + "loss": 0.2989, + "step": 52643 + }, + { + "epoch": 3.653042814516689, + "grad_norm": 0.9441542625427246, + "learning_rate": 3.920220245878237e-06, + "loss": 0.3632, + "step": 52644 + }, + { + "epoch": 3.6531122059537853, + "grad_norm": 1.0113697052001953, + "learning_rate": 3.918662636095172e-06, + "loss": 0.3003, + "step": 52645 + }, + { + "epoch": 3.653181597390882, + "grad_norm": 1.165130376815796, + "learning_rate": 3.917105329629722e-06, + "loss": 0.294, + "step": 52646 + }, + { + "epoch": 3.6532509888279785, + "grad_norm": 0.9208811521530151, + "learning_rate": 3.915548326486829e-06, + "loss": 0.2975, + "step": 52647 + }, + { + "epoch": 3.6533203802650753, + "grad_norm": 1.5396554470062256, + "learning_rate": 3.913991626671409e-06, + "loss": 0.2214, + "step": 52648 + }, + { + "epoch": 3.653389771702172, + "grad_norm": 1.0781641006469727, + "learning_rate": 3.912435230188361e-06, + "loss": 0.2458, + "step": 52649 + }, + { + "epoch": 3.6534591631392686, + "grad_norm": 1.0575194358825684, + "learning_rate": 3.910879137042611e-06, + "loss": 0.3323, + "step": 52650 + }, + { + "epoch": 3.6535285545763654, + "grad_norm": 0.8318018913269043, + "learning_rate": 3.909323347239069e-06, + "loss": 0.3241, + "step": 52651 + }, + { + "epoch": 3.6535979460134618, + "grad_norm": 1.1964136362075806, + "learning_rate": 3.907767860782652e-06, + "loss": 0.2931, + "step": 52652 + }, + { + "epoch": 3.6536673374505586, + "grad_norm": 0.8021435141563416, + "learning_rate": 3.906212677678245e-06, + "loss": 0.2631, + "step": 52653 + }, + { + "epoch": 3.6537367288876554, + "grad_norm": 0.7243555784225464, + "learning_rate": 3.904657797930789e-06, + "loss": 0.2283, + "step": 52654 + }, + { + "epoch": 3.653806120324752, + "grad_norm": 0.8504951000213623, + "learning_rate": 3.903103221545168e-06, + "loss": 0.1532, + "step": 52655 + }, + { + "epoch": 3.6538755117618487, + "grad_norm": 0.8852255344390869, + "learning_rate": 3.901548948526312e-06, + "loss": 0.2503, + "step": 52656 + }, + { + "epoch": 3.653944903198945, + "grad_norm": 0.8699684143066406, + "learning_rate": 3.899994978879107e-06, + "loss": 0.2287, + "step": 52657 + }, + { + "epoch": 3.654014294636042, + "grad_norm": 0.6812434196472168, + "learning_rate": 3.898441312608469e-06, + "loss": 0.1618, + "step": 52658 + }, + { + "epoch": 3.6540836860731387, + "grad_norm": 0.6846200227737427, + "learning_rate": 3.896887949719319e-06, + "loss": 0.2151, + "step": 52659 + }, + { + "epoch": 3.654153077510235, + "grad_norm": 0.9215379953384399, + "learning_rate": 3.895334890216529e-06, + "loss": 0.2705, + "step": 52660 + }, + { + "epoch": 3.654222468947332, + "grad_norm": 0.9322764873504639, + "learning_rate": 3.893782134105017e-06, + "loss": 0.2508, + "step": 52661 + }, + { + "epoch": 3.6542918603844283, + "grad_norm": 0.6799325942993164, + "learning_rate": 3.892229681389681e-06, + "loss": 0.2053, + "step": 52662 + }, + { + "epoch": 3.654361251821525, + "grad_norm": 1.3703004121780396, + "learning_rate": 3.890677532075427e-06, + "loss": 0.2807, + "step": 52663 + }, + { + "epoch": 3.654430643258622, + "grad_norm": 0.7424569129943848, + "learning_rate": 3.889125686167161e-06, + "loss": 0.1482, + "step": 52664 + }, + { + "epoch": 3.654500034695719, + "grad_norm": 0.7153061032295227, + "learning_rate": 3.887574143669781e-06, + "loss": 0.2407, + "step": 52665 + }, + { + "epoch": 3.654569426132815, + "grad_norm": 0.7571722865104675, + "learning_rate": 3.88602290458816e-06, + "loss": 0.3027, + "step": 52666 + }, + { + "epoch": 3.6546388175699116, + "grad_norm": 0.7005161643028259, + "learning_rate": 3.884471968927217e-06, + "loss": 0.2398, + "step": 52667 + }, + { + "epoch": 3.6547082090070084, + "grad_norm": 0.9935612678527832, + "learning_rate": 3.882921336691847e-06, + "loss": 0.2264, + "step": 52668 + }, + { + "epoch": 3.6547776004441053, + "grad_norm": 1.3707878589630127, + "learning_rate": 3.881371007886958e-06, + "loss": 0.3632, + "step": 52669 + }, + { + "epoch": 3.654846991881202, + "grad_norm": 0.8076624870300293, + "learning_rate": 3.879820982517412e-06, + "loss": 0.263, + "step": 52670 + }, + { + "epoch": 3.6549163833182985, + "grad_norm": 0.9034048318862915, + "learning_rate": 3.878271260588118e-06, + "loss": 0.252, + "step": 52671 + }, + { + "epoch": 3.6549857747553953, + "grad_norm": 1.1637709140777588, + "learning_rate": 3.87672184210397e-06, + "loss": 0.3009, + "step": 52672 + }, + { + "epoch": 3.6550551661924917, + "grad_norm": 1.3908665180206299, + "learning_rate": 3.875172727069876e-06, + "loss": 0.2292, + "step": 52673 + }, + { + "epoch": 3.6551245576295885, + "grad_norm": 1.169650912284851, + "learning_rate": 3.873623915490698e-06, + "loss": 0.5008, + "step": 52674 + }, + { + "epoch": 3.6551939490666854, + "grad_norm": 0.6524322628974915, + "learning_rate": 3.872075407371323e-06, + "loss": 0.2721, + "step": 52675 + }, + { + "epoch": 3.6552633405037818, + "grad_norm": 1.200284719467163, + "learning_rate": 3.870527202716668e-06, + "loss": 0.2325, + "step": 52676 + }, + { + "epoch": 3.6553327319408786, + "grad_norm": 1.1160686016082764, + "learning_rate": 3.868979301531594e-06, + "loss": 0.2782, + "step": 52677 + }, + { + "epoch": 3.655402123377975, + "grad_norm": 1.0153861045837402, + "learning_rate": 3.867431703821012e-06, + "loss": 0.1568, + "step": 52678 + }, + { + "epoch": 3.655471514815072, + "grad_norm": 1.3176015615463257, + "learning_rate": 3.865884409589782e-06, + "loss": 0.4045, + "step": 52679 + }, + { + "epoch": 3.6555409062521687, + "grad_norm": 0.9521582126617432, + "learning_rate": 3.864337418842823e-06, + "loss": 0.2392, + "step": 52680 + }, + { + "epoch": 3.655610297689265, + "grad_norm": 0.6770145893096924, + "learning_rate": 3.862790731584975e-06, + "loss": 0.2159, + "step": 52681 + }, + { + "epoch": 3.655679689126362, + "grad_norm": 0.8120859265327454, + "learning_rate": 3.861244347821147e-06, + "loss": 0.3038, + "step": 52682 + }, + { + "epoch": 3.6557490805634583, + "grad_norm": 0.9124694466590881, + "learning_rate": 3.859698267556222e-06, + "loss": 0.2635, + "step": 52683 + }, + { + "epoch": 3.655818472000555, + "grad_norm": 0.8050738573074341, + "learning_rate": 3.858152490795075e-06, + "loss": 0.2327, + "step": 52684 + }, + { + "epoch": 3.655887863437652, + "grad_norm": 0.5454455614089966, + "learning_rate": 3.85660701754258e-06, + "loss": 0.163, + "step": 52685 + }, + { + "epoch": 3.6559572548747483, + "grad_norm": 0.70222008228302, + "learning_rate": 3.855061847803621e-06, + "loss": 0.1731, + "step": 52686 + }, + { + "epoch": 3.656026646311845, + "grad_norm": 0.6317734718322754, + "learning_rate": 3.853516981583094e-06, + "loss": 0.1998, + "step": 52687 + }, + { + "epoch": 3.6560960377489415, + "grad_norm": 0.8767004013061523, + "learning_rate": 3.851972418885841e-06, + "loss": 0.2629, + "step": 52688 + }, + { + "epoch": 3.6561654291860384, + "grad_norm": 0.8575160503387451, + "learning_rate": 3.850428159716768e-06, + "loss": 0.2836, + "step": 52689 + }, + { + "epoch": 3.656234820623135, + "grad_norm": 0.7404438853263855, + "learning_rate": 3.848884204080727e-06, + "loss": 0.261, + "step": 52690 + }, + { + "epoch": 3.656304212060232, + "grad_norm": 0.7841260433197021, + "learning_rate": 3.847340551982603e-06, + "loss": 0.1815, + "step": 52691 + }, + { + "epoch": 3.6563736034973284, + "grad_norm": 0.7411857843399048, + "learning_rate": 3.84579720342727e-06, + "loss": 0.1958, + "step": 52692 + }, + { + "epoch": 3.656442994934425, + "grad_norm": 0.980502188205719, + "learning_rate": 3.844254158419602e-06, + "loss": 0.3057, + "step": 52693 + }, + { + "epoch": 3.6565123863715217, + "grad_norm": 0.8588389158248901, + "learning_rate": 3.842711416964473e-06, + "loss": 0.2631, + "step": 52694 + }, + { + "epoch": 3.6565817778086185, + "grad_norm": 0.8545160889625549, + "learning_rate": 3.8411689790667445e-06, + "loss": 0.1786, + "step": 52695 + }, + { + "epoch": 3.6566511692457153, + "grad_norm": 0.9475885033607483, + "learning_rate": 3.839626844731292e-06, + "loss": 0.2932, + "step": 52696 + }, + { + "epoch": 3.6567205606828117, + "grad_norm": 0.8554646968841553, + "learning_rate": 3.838085013962967e-06, + "loss": 0.2499, + "step": 52697 + }, + { + "epoch": 3.6567899521199085, + "grad_norm": 1.3468037843704224, + "learning_rate": 3.836543486766664e-06, + "loss": 0.4353, + "step": 52698 + }, + { + "epoch": 3.656859343557005, + "grad_norm": 0.6641599535942078, + "learning_rate": 3.8350022631472245e-06, + "loss": 0.2242, + "step": 52699 + }, + { + "epoch": 3.6569287349941018, + "grad_norm": 0.9991735816001892, + "learning_rate": 3.833461343109546e-06, + "loss": 0.2674, + "step": 52700 + }, + { + "epoch": 3.6569981264311986, + "grad_norm": 0.94752436876297, + "learning_rate": 3.831920726658456e-06, + "loss": 0.1507, + "step": 52701 + }, + { + "epoch": 3.657067517868295, + "grad_norm": 0.8261562585830688, + "learning_rate": 3.83038041379884e-06, + "loss": 0.1624, + "step": 52702 + }, + { + "epoch": 3.657136909305392, + "grad_norm": 0.8198181986808777, + "learning_rate": 3.828840404535549e-06, + "loss": 0.1158, + "step": 52703 + }, + { + "epoch": 3.657206300742488, + "grad_norm": 0.8170338869094849, + "learning_rate": 3.827300698873471e-06, + "loss": 0.3153, + "step": 52704 + }, + { + "epoch": 3.657275692179585, + "grad_norm": 0.959362268447876, + "learning_rate": 3.825761296817432e-06, + "loss": 0.1832, + "step": 52705 + }, + { + "epoch": 3.657345083616682, + "grad_norm": 0.7728179693222046, + "learning_rate": 3.8242221983722956e-06, + "loss": 0.2904, + "step": 52706 + }, + { + "epoch": 3.6574144750537783, + "grad_norm": 1.2953604459762573, + "learning_rate": 3.822683403542959e-06, + "loss": 0.4169, + "step": 52707 + }, + { + "epoch": 3.657483866490875, + "grad_norm": 0.6324697732925415, + "learning_rate": 3.821144912334229e-06, + "loss": 0.2022, + "step": 52708 + }, + { + "epoch": 3.6575532579279715, + "grad_norm": 0.7509033679962158, + "learning_rate": 3.8196067247509905e-06, + "loss": 0.1392, + "step": 52709 + }, + { + "epoch": 3.6576226493650683, + "grad_norm": 1.3341318368911743, + "learning_rate": 3.818068840798094e-06, + "loss": 0.252, + "step": 52710 + }, + { + "epoch": 3.657692040802165, + "grad_norm": 0.9755740165710449, + "learning_rate": 3.816531260480394e-06, + "loss": 0.3349, + "step": 52711 + }, + { + "epoch": 3.6577614322392615, + "grad_norm": 1.2427504062652588, + "learning_rate": 3.814993983802751e-06, + "loss": 0.2257, + "step": 52712 + }, + { + "epoch": 3.6578308236763584, + "grad_norm": 0.9887919425964355, + "learning_rate": 3.813457010770005e-06, + "loss": 0.283, + "step": 52713 + }, + { + "epoch": 3.6579002151134548, + "grad_norm": 0.744551420211792, + "learning_rate": 3.811920341387032e-06, + "loss": 0.1899, + "step": 52714 + }, + { + "epoch": 3.6579696065505516, + "grad_norm": 0.5849030613899231, + "learning_rate": 3.810383975658649e-06, + "loss": 0.1231, + "step": 52715 + }, + { + "epoch": 3.6580389979876484, + "grad_norm": 1.2116870880126953, + "learning_rate": 3.8088479135897305e-06, + "loss": 0.3, + "step": 52716 + }, + { + "epoch": 3.6581083894247453, + "grad_norm": 1.1354525089263916, + "learning_rate": 3.8073121551851166e-06, + "loss": 0.372, + "step": 52717 + }, + { + "epoch": 3.6581777808618416, + "grad_norm": 1.1263906955718994, + "learning_rate": 3.805776700449648e-06, + "loss": 0.256, + "step": 52718 + }, + { + "epoch": 3.6582471722989385, + "grad_norm": 1.3722777366638184, + "learning_rate": 3.8042415493881988e-06, + "loss": 0.251, + "step": 52719 + }, + { + "epoch": 3.658316563736035, + "grad_norm": 0.691522479057312, + "learning_rate": 3.8027067020055874e-06, + "loss": 0.2506, + "step": 52720 + }, + { + "epoch": 3.6583859551731317, + "grad_norm": 0.8269989490509033, + "learning_rate": 3.8011721583066875e-06, + "loss": 0.2868, + "step": 52721 + }, + { + "epoch": 3.6584553466102285, + "grad_norm": 0.8082358837127686, + "learning_rate": 3.799637918296306e-06, + "loss": 0.2972, + "step": 52722 + }, + { + "epoch": 3.658524738047325, + "grad_norm": 0.7972755432128906, + "learning_rate": 3.7981039819793175e-06, + "loss": 0.2141, + "step": 52723 + }, + { + "epoch": 3.6585941294844218, + "grad_norm": 1.1272920370101929, + "learning_rate": 3.7965703493605397e-06, + "loss": 0.3381, + "step": 52724 + }, + { + "epoch": 3.658663520921518, + "grad_norm": 1.2631138563156128, + "learning_rate": 3.795037020444836e-06, + "loss": 0.3444, + "step": 52725 + }, + { + "epoch": 3.658732912358615, + "grad_norm": 5.442626953125, + "learning_rate": 3.793503995237047e-06, + "loss": 0.1924, + "step": 52726 + }, + { + "epoch": 3.658802303795712, + "grad_norm": 0.9774461388587952, + "learning_rate": 3.79197127374199e-06, + "loss": 0.2165, + "step": 52727 + }, + { + "epoch": 3.658871695232808, + "grad_norm": 1.081315040588379, + "learning_rate": 3.790438855964529e-06, + "loss": 0.2956, + "step": 52728 + }, + { + "epoch": 3.658941086669905, + "grad_norm": 1.2174967527389526, + "learning_rate": 3.7889067419094928e-06, + "loss": 0.2541, + "step": 52729 + }, + { + "epoch": 3.6590104781070014, + "grad_norm": 0.8734613060951233, + "learning_rate": 3.7873749315816996e-06, + "loss": 0.2027, + "step": 52730 + }, + { + "epoch": 3.6590798695440983, + "grad_norm": 1.3964290618896484, + "learning_rate": 3.785843424986013e-06, + "loss": 0.2908, + "step": 52731 + }, + { + "epoch": 3.659149260981195, + "grad_norm": 0.6318169236183167, + "learning_rate": 3.784312222127251e-06, + "loss": 0.1614, + "step": 52732 + }, + { + "epoch": 3.6592186524182915, + "grad_norm": 1.0833442211151123, + "learning_rate": 3.782781323010243e-06, + "loss": 0.2108, + "step": 52733 + }, + { + "epoch": 3.6592880438553883, + "grad_norm": 0.7539622187614441, + "learning_rate": 3.78125072763984e-06, + "loss": 0.2063, + "step": 52734 + }, + { + "epoch": 3.6593574352924847, + "grad_norm": 0.6773257851600647, + "learning_rate": 3.7797204360208836e-06, + "loss": 0.2015, + "step": 52735 + }, + { + "epoch": 3.6594268267295815, + "grad_norm": 0.7910909652709961, + "learning_rate": 3.7781904481581586e-06, + "loss": 0.2761, + "step": 52736 + }, + { + "epoch": 3.6594962181666784, + "grad_norm": 1.0136054754257202, + "learning_rate": 3.7766607640565388e-06, + "loss": 0.2702, + "step": 52737 + }, + { + "epoch": 3.6595656096037747, + "grad_norm": 0.8671572208404541, + "learning_rate": 3.7751313837208315e-06, + "loss": 0.2244, + "step": 52738 + }, + { + "epoch": 3.6596350010408716, + "grad_norm": 1.4746307134628296, + "learning_rate": 3.7736023071558767e-06, + "loss": 0.2608, + "step": 52739 + }, + { + "epoch": 3.659704392477968, + "grad_norm": 0.7300363779067993, + "learning_rate": 3.7720735343664826e-06, + "loss": 0.1849, + "step": 52740 + }, + { + "epoch": 3.659773783915065, + "grad_norm": 0.958463728427887, + "learning_rate": 3.770545065357489e-06, + "loss": 0.1959, + "step": 52741 + }, + { + "epoch": 3.6598431753521616, + "grad_norm": 0.7839449644088745, + "learning_rate": 3.7690169001337373e-06, + "loss": 0.2745, + "step": 52742 + }, + { + "epoch": 3.6599125667892585, + "grad_norm": 1.084510087966919, + "learning_rate": 3.767489038700023e-06, + "loss": 0.3084, + "step": 52743 + }, + { + "epoch": 3.659981958226355, + "grad_norm": 1.3546128273010254, + "learning_rate": 3.7659614810611752e-06, + "loss": 0.2657, + "step": 52744 + }, + { + "epoch": 3.6600513496634517, + "grad_norm": 0.9089767336845398, + "learning_rate": 3.7644342272220244e-06, + "loss": 0.1761, + "step": 52745 + }, + { + "epoch": 3.660120741100548, + "grad_norm": 0.9890543222427368, + "learning_rate": 3.7629072771873887e-06, + "loss": 0.2763, + "step": 52746 + }, + { + "epoch": 3.660190132537645, + "grad_norm": 1.0915641784667969, + "learning_rate": 3.761380630962097e-06, + "loss": 0.3574, + "step": 52747 + }, + { + "epoch": 3.6602595239747417, + "grad_norm": 0.6599135398864746, + "learning_rate": 3.7598542885509458e-06, + "loss": 0.1673, + "step": 52748 + }, + { + "epoch": 3.660328915411838, + "grad_norm": 1.1935904026031494, + "learning_rate": 3.758328249958787e-06, + "loss": 0.3267, + "step": 52749 + }, + { + "epoch": 3.660398306848935, + "grad_norm": 0.9411699175834656, + "learning_rate": 3.756802515190416e-06, + "loss": 0.2299, + "step": 52750 + }, + { + "epoch": 3.6604676982860314, + "grad_norm": 1.3615411520004272, + "learning_rate": 3.7552770842506414e-06, + "loss": 0.3425, + "step": 52751 + }, + { + "epoch": 3.660537089723128, + "grad_norm": 0.6227725744247437, + "learning_rate": 3.7537519571443023e-06, + "loss": 0.1456, + "step": 52752 + }, + { + "epoch": 3.660606481160225, + "grad_norm": 0.8161095380783081, + "learning_rate": 3.7522271338761848e-06, + "loss": 0.3454, + "step": 52753 + }, + { + "epoch": 3.6606758725973214, + "grad_norm": 0.8777706623077393, + "learning_rate": 3.7507026144511292e-06, + "loss": 0.2419, + "step": 52754 + }, + { + "epoch": 3.6607452640344182, + "grad_norm": 0.8335351347923279, + "learning_rate": 3.7491783988739426e-06, + "loss": 0.3156, + "step": 52755 + }, + { + "epoch": 3.6608146554715146, + "grad_norm": 0.7461448311805725, + "learning_rate": 3.7476544871494436e-06, + "loss": 0.2077, + "step": 52756 + }, + { + "epoch": 3.6608840469086115, + "grad_norm": 0.9339038133621216, + "learning_rate": 3.7461308792824168e-06, + "loss": 0.2456, + "step": 52757 + }, + { + "epoch": 3.6609534383457083, + "grad_norm": 0.7871794700622559, + "learning_rate": 3.7446075752776923e-06, + "loss": 0.1943, + "step": 52758 + }, + { + "epoch": 3.6610228297828047, + "grad_norm": 0.8335361480712891, + "learning_rate": 3.7430845751400656e-06, + "loss": 0.2902, + "step": 52759 + }, + { + "epoch": 3.6610922212199015, + "grad_norm": 1.455260992050171, + "learning_rate": 3.7415618788743556e-06, + "loss": 0.2709, + "step": 52760 + }, + { + "epoch": 3.661161612656998, + "grad_norm": 1.0004637241363525, + "learning_rate": 3.7400394864853694e-06, + "loss": 0.2908, + "step": 52761 + }, + { + "epoch": 3.6612310040940947, + "grad_norm": 0.5980512499809265, + "learning_rate": 3.738517397977903e-06, + "loss": 0.1213, + "step": 52762 + }, + { + "epoch": 3.6613003955311916, + "grad_norm": 0.8517219424247742, + "learning_rate": 3.7369956133567864e-06, + "loss": 0.3116, + "step": 52763 + }, + { + "epoch": 3.661369786968288, + "grad_norm": 1.053046703338623, + "learning_rate": 3.7354741326268038e-06, + "loss": 0.2088, + "step": 52764 + }, + { + "epoch": 3.661439178405385, + "grad_norm": 1.2560701370239258, + "learning_rate": 3.733952955792741e-06, + "loss": 0.3148, + "step": 52765 + }, + { + "epoch": 3.661508569842481, + "grad_norm": 0.6848070621490479, + "learning_rate": 3.7324320828594383e-06, + "loss": 0.1505, + "step": 52766 + }, + { + "epoch": 3.661577961279578, + "grad_norm": 0.9998874068260193, + "learning_rate": 3.7309115138316698e-06, + "loss": 0.2649, + "step": 52767 + }, + { + "epoch": 3.661647352716675, + "grad_norm": 0.872134268283844, + "learning_rate": 3.7293912487142426e-06, + "loss": 0.2254, + "step": 52768 + }, + { + "epoch": 3.6617167441537717, + "grad_norm": 0.7014914155006409, + "learning_rate": 3.7278712875119637e-06, + "loss": 0.2053, + "step": 52769 + }, + { + "epoch": 3.661786135590868, + "grad_norm": 0.8253743648529053, + "learning_rate": 3.72635163022963e-06, + "loss": 0.1798, + "step": 52770 + }, + { + "epoch": 3.661855527027965, + "grad_norm": 0.8485817909240723, + "learning_rate": 3.7248322768720256e-06, + "loss": 0.283, + "step": 52771 + }, + { + "epoch": 3.6619249184650613, + "grad_norm": 0.7362350225448608, + "learning_rate": 3.7233132274439586e-06, + "loss": 0.3, + "step": 52772 + }, + { + "epoch": 3.661994309902158, + "grad_norm": 0.9017701745033264, + "learning_rate": 3.7217944819502137e-06, + "loss": 0.2144, + "step": 52773 + }, + { + "epoch": 3.662063701339255, + "grad_norm": 0.9472851753234863, + "learning_rate": 3.7202760403956094e-06, + "loss": 0.2771, + "step": 52774 + }, + { + "epoch": 3.6621330927763514, + "grad_norm": 1.119800329208374, + "learning_rate": 3.7187579027849194e-06, + "loss": 0.218, + "step": 52775 + }, + { + "epoch": 3.662202484213448, + "grad_norm": 0.5417465567588806, + "learning_rate": 3.717240069122929e-06, + "loss": 0.1835, + "step": 52776 + }, + { + "epoch": 3.6622718756505446, + "grad_norm": 0.8826377987861633, + "learning_rate": 3.715722539414457e-06, + "loss": 0.2864, + "step": 52777 + }, + { + "epoch": 3.6623412670876414, + "grad_norm": 0.715610921382904, + "learning_rate": 3.7142053136642653e-06, + "loss": 0.2644, + "step": 52778 + }, + { + "epoch": 3.6624106585247382, + "grad_norm": 1.0298442840576172, + "learning_rate": 3.7126883918771615e-06, + "loss": 0.2232, + "step": 52779 + }, + { + "epoch": 3.6624800499618346, + "grad_norm": 0.9557578563690186, + "learning_rate": 3.71117177405792e-06, + "loss": 0.217, + "step": 52780 + }, + { + "epoch": 3.6625494413989315, + "grad_norm": 0.9020882248878479, + "learning_rate": 3.7096554602113364e-06, + "loss": 0.1948, + "step": 52781 + }, + { + "epoch": 3.662618832836028, + "grad_norm": 1.0987536907196045, + "learning_rate": 3.7081394503422075e-06, + "loss": 0.237, + "step": 52782 + }, + { + "epoch": 3.6626882242731247, + "grad_norm": 1.1245042085647583, + "learning_rate": 3.7066237444553174e-06, + "loss": 0.3922, + "step": 52783 + }, + { + "epoch": 3.6627576157102215, + "grad_norm": 0.7975564002990723, + "learning_rate": 3.7051083425554413e-06, + "loss": 0.1544, + "step": 52784 + }, + { + "epoch": 3.662827007147318, + "grad_norm": 0.8209479451179504, + "learning_rate": 3.703593244647352e-06, + "loss": 0.276, + "step": 52785 + }, + { + "epoch": 3.6628963985844147, + "grad_norm": 0.6106798648834229, + "learning_rate": 3.7020784507358574e-06, + "loss": 0.2437, + "step": 52786 + }, + { + "epoch": 3.662965790021511, + "grad_norm": 0.7851235866546631, + "learning_rate": 3.7005639608257202e-06, + "loss": 0.286, + "step": 52787 + }, + { + "epoch": 3.663035181458608, + "grad_norm": 1.0925766229629517, + "learning_rate": 3.6990497749217366e-06, + "loss": 0.3176, + "step": 52788 + }, + { + "epoch": 3.663104572895705, + "grad_norm": 0.9179455637931824, + "learning_rate": 3.6975358930286698e-06, + "loss": 0.2874, + "step": 52789 + }, + { + "epoch": 3.663173964332801, + "grad_norm": 1.0025218725204468, + "learning_rate": 3.6960223151513373e-06, + "loss": 0.2666, + "step": 52790 + }, + { + "epoch": 3.663243355769898, + "grad_norm": 0.8558415770530701, + "learning_rate": 3.6945090412944583e-06, + "loss": 0.1931, + "step": 52791 + }, + { + "epoch": 3.6633127472069944, + "grad_norm": 1.0300118923187256, + "learning_rate": 3.692996071462851e-06, + "loss": 0.2184, + "step": 52792 + }, + { + "epoch": 3.6633821386440912, + "grad_norm": 0.6613882184028625, + "learning_rate": 3.691483405661289e-06, + "loss": 0.2974, + "step": 52793 + }, + { + "epoch": 3.663451530081188, + "grad_norm": 0.8538778424263, + "learning_rate": 3.6899710438945247e-06, + "loss": 0.2159, + "step": 52794 + }, + { + "epoch": 3.663520921518285, + "grad_norm": 0.7582238912582397, + "learning_rate": 3.6884589861673534e-06, + "loss": 0.2025, + "step": 52795 + }, + { + "epoch": 3.6635903129553813, + "grad_norm": 0.7398366332054138, + "learning_rate": 3.68694723248455e-06, + "loss": 0.1988, + "step": 52796 + }, + { + "epoch": 3.663659704392478, + "grad_norm": 0.8430434465408325, + "learning_rate": 3.6854357828508877e-06, + "loss": 0.2246, + "step": 52797 + }, + { + "epoch": 3.6637290958295745, + "grad_norm": 0.7777190208435059, + "learning_rate": 3.6839246372711076e-06, + "loss": 0.2274, + "step": 52798 + }, + { + "epoch": 3.6637984872666713, + "grad_norm": 1.108371376991272, + "learning_rate": 3.682413795750017e-06, + "loss": 0.2361, + "step": 52799 + }, + { + "epoch": 3.663867878703768, + "grad_norm": 0.6828113198280334, + "learning_rate": 3.6809032582923673e-06, + "loss": 0.1627, + "step": 52800 + }, + { + "epoch": 3.6639372701408646, + "grad_norm": 1.2730222940444946, + "learning_rate": 3.679393024902922e-06, + "loss": 0.5062, + "step": 52801 + }, + { + "epoch": 3.6640066615779614, + "grad_norm": 0.9516423940658569, + "learning_rate": 3.677883095586465e-06, + "loss": 0.3145, + "step": 52802 + }, + { + "epoch": 3.664076053015058, + "grad_norm": 0.9863996505737305, + "learning_rate": 3.6763734703477606e-06, + "loss": 0.2127, + "step": 52803 + }, + { + "epoch": 3.6641454444521546, + "grad_norm": 1.2361347675323486, + "learning_rate": 3.67486414919157e-06, + "loss": 0.332, + "step": 52804 + }, + { + "epoch": 3.6642148358892515, + "grad_norm": 0.9991285800933838, + "learning_rate": 3.6733551321226577e-06, + "loss": 0.3377, + "step": 52805 + }, + { + "epoch": 3.664284227326348, + "grad_norm": 0.9163216948509216, + "learning_rate": 3.671846419145786e-06, + "loss": 0.2322, + "step": 52806 + }, + { + "epoch": 3.6643536187634447, + "grad_norm": 0.8034060001373291, + "learning_rate": 3.670338010265706e-06, + "loss": 0.1789, + "step": 52807 + }, + { + "epoch": 3.664423010200541, + "grad_norm": 0.9118898510932922, + "learning_rate": 3.6688299054872145e-06, + "loss": 0.3199, + "step": 52808 + }, + { + "epoch": 3.664492401637638, + "grad_norm": 1.0485773086547852, + "learning_rate": 3.6673221048150407e-06, + "loss": 0.2316, + "step": 52809 + }, + { + "epoch": 3.6645617930747347, + "grad_norm": 0.5661081075668335, + "learning_rate": 3.665814608253948e-06, + "loss": 0.2411, + "step": 52810 + }, + { + "epoch": 3.664631184511831, + "grad_norm": 0.7651076316833496, + "learning_rate": 3.66430741580871e-06, + "loss": 0.2058, + "step": 52811 + }, + { + "epoch": 3.664700575948928, + "grad_norm": 0.6741513609886169, + "learning_rate": 3.6628005274840783e-06, + "loss": 0.1994, + "step": 52812 + }, + { + "epoch": 3.6647699673860243, + "grad_norm": 0.6185786724090576, + "learning_rate": 3.661293943284805e-06, + "loss": 0.1986, + "step": 52813 + }, + { + "epoch": 3.664839358823121, + "grad_norm": 0.8053078055381775, + "learning_rate": 3.6597876632156414e-06, + "loss": 0.1797, + "step": 52814 + }, + { + "epoch": 3.664908750260218, + "grad_norm": 1.0105183124542236, + "learning_rate": 3.658281687281351e-06, + "loss": 0.2378, + "step": 52815 + }, + { + "epoch": 3.6649781416973144, + "grad_norm": 1.0512404441833496, + "learning_rate": 3.656776015486696e-06, + "loss": 0.2188, + "step": 52816 + }, + { + "epoch": 3.6650475331344112, + "grad_norm": 1.0362000465393066, + "learning_rate": 3.6552706478364174e-06, + "loss": 0.2473, + "step": 52817 + }, + { + "epoch": 3.6651169245715076, + "grad_norm": 1.0602971315383911, + "learning_rate": 3.6537655843352782e-06, + "loss": 0.3441, + "step": 52818 + }, + { + "epoch": 3.6651863160086045, + "grad_norm": 1.43651282787323, + "learning_rate": 3.6522608249880186e-06, + "loss": 0.2978, + "step": 52819 + }, + { + "epoch": 3.6652557074457013, + "grad_norm": 1.0958476066589355, + "learning_rate": 3.6507563697993906e-06, + "loss": 0.1992, + "step": 52820 + }, + { + "epoch": 3.665325098882798, + "grad_norm": 0.7819609045982361, + "learning_rate": 3.6492522187741352e-06, + "loss": 0.1916, + "step": 52821 + }, + { + "epoch": 3.6653944903198945, + "grad_norm": 0.7498773336410522, + "learning_rate": 3.647748371917026e-06, + "loss": 0.171, + "step": 52822 + }, + { + "epoch": 3.6654638817569913, + "grad_norm": 1.0513660907745361, + "learning_rate": 3.6462448292327924e-06, + "loss": 0.3538, + "step": 52823 + }, + { + "epoch": 3.6655332731940877, + "grad_norm": 0.7785619497299194, + "learning_rate": 3.6447415907261862e-06, + "loss": 0.1995, + "step": 52824 + }, + { + "epoch": 3.6656026646311846, + "grad_norm": 0.8644651174545288, + "learning_rate": 3.6432386564019592e-06, + "loss": 0.2339, + "step": 52825 + }, + { + "epoch": 3.6656720560682814, + "grad_norm": 0.6970366835594177, + "learning_rate": 3.6417360262648416e-06, + "loss": 0.1677, + "step": 52826 + }, + { + "epoch": 3.665741447505378, + "grad_norm": 0.7459127902984619, + "learning_rate": 3.640233700319573e-06, + "loss": 0.197, + "step": 52827 + }, + { + "epoch": 3.6658108389424746, + "grad_norm": 0.8802746534347534, + "learning_rate": 3.638731678570917e-06, + "loss": 0.3478, + "step": 52828 + }, + { + "epoch": 3.665880230379571, + "grad_norm": 0.9514691829681396, + "learning_rate": 3.637229961023603e-06, + "loss": 0.2273, + "step": 52829 + }, + { + "epoch": 3.665949621816668, + "grad_norm": 0.6574222445487976, + "learning_rate": 3.635728547682371e-06, + "loss": 0.1471, + "step": 52830 + }, + { + "epoch": 3.6660190132537647, + "grad_norm": 0.6288377046585083, + "learning_rate": 3.6342274385519738e-06, + "loss": 0.1309, + "step": 52831 + }, + { + "epoch": 3.666088404690861, + "grad_norm": 0.7203860282897949, + "learning_rate": 3.63272663363714e-06, + "loss": 0.1813, + "step": 52832 + }, + { + "epoch": 3.666157796127958, + "grad_norm": 1.0637108087539673, + "learning_rate": 3.6312261329426e-06, + "loss": 0.2325, + "step": 52833 + }, + { + "epoch": 3.6662271875650543, + "grad_norm": 0.6329671144485474, + "learning_rate": 3.629725936473094e-06, + "loss": 0.2017, + "step": 52834 + }, + { + "epoch": 3.666296579002151, + "grad_norm": 0.5443372130393982, + "learning_rate": 3.628226044233374e-06, + "loss": 0.2118, + "step": 52835 + }, + { + "epoch": 3.666365970439248, + "grad_norm": 1.1411933898925781, + "learning_rate": 3.626726456228158e-06, + "loss": 0.2823, + "step": 52836 + }, + { + "epoch": 3.6664353618763443, + "grad_norm": 1.1495394706726074, + "learning_rate": 3.6252271724621754e-06, + "loss": 0.2568, + "step": 52837 + }, + { + "epoch": 3.666504753313441, + "grad_norm": 1.3892102241516113, + "learning_rate": 3.623728192940179e-06, + "loss": 0.2282, + "step": 52838 + }, + { + "epoch": 3.6665741447505376, + "grad_norm": 0.8660702705383301, + "learning_rate": 3.6222295176668975e-06, + "loss": 0.2575, + "step": 52839 + }, + { + "epoch": 3.6666435361876344, + "grad_norm": 1.1516629457473755, + "learning_rate": 3.6207311466470493e-06, + "loss": 0.2603, + "step": 52840 + }, + { + "epoch": 3.666712927624731, + "grad_norm": 0.7940815091133118, + "learning_rate": 3.619233079885365e-06, + "loss": 0.1951, + "step": 52841 + }, + { + "epoch": 3.6667823190618276, + "grad_norm": 1.055883526802063, + "learning_rate": 3.6177353173865835e-06, + "loss": 0.2947, + "step": 52842 + }, + { + "epoch": 3.6668517104989244, + "grad_norm": 1.0975536108016968, + "learning_rate": 3.616237859155447e-06, + "loss": 0.2978, + "step": 52843 + }, + { + "epoch": 3.666921101936021, + "grad_norm": 0.7814816236495972, + "learning_rate": 3.6147407051966396e-06, + "loss": 0.2482, + "step": 52844 + }, + { + "epoch": 3.6669904933731177, + "grad_norm": 0.740532398223877, + "learning_rate": 3.6132438555149252e-06, + "loss": 0.1971, + "step": 52845 + }, + { + "epoch": 3.6670598848102145, + "grad_norm": 0.8329312205314636, + "learning_rate": 3.61174731011501e-06, + "loss": 0.2769, + "step": 52846 + }, + { + "epoch": 3.6671292762473113, + "grad_norm": 0.7974249720573425, + "learning_rate": 3.6102510690016354e-06, + "loss": 0.2236, + "step": 52847 + }, + { + "epoch": 3.6671986676844077, + "grad_norm": 1.1413917541503906, + "learning_rate": 3.6087551321795087e-06, + "loss": 0.2651, + "step": 52848 + }, + { + "epoch": 3.6672680591215046, + "grad_norm": 0.5307501554489136, + "learning_rate": 3.607259499653359e-06, + "loss": 0.126, + "step": 52849 + }, + { + "epoch": 3.667337450558601, + "grad_norm": 0.8924290537834167, + "learning_rate": 3.6057641714279054e-06, + "loss": 0.2446, + "step": 52850 + }, + { + "epoch": 3.6674068419956978, + "grad_norm": 1.1058319807052612, + "learning_rate": 3.604269147507866e-06, + "loss": 0.3003, + "step": 52851 + }, + { + "epoch": 3.6674762334327946, + "grad_norm": 0.9531375169754028, + "learning_rate": 3.6027744278979703e-06, + "loss": 0.1562, + "step": 52852 + }, + { + "epoch": 3.667545624869891, + "grad_norm": 0.9836112260818481, + "learning_rate": 3.6012800126029367e-06, + "loss": 0.3024, + "step": 52853 + }, + { + "epoch": 3.667615016306988, + "grad_norm": 1.2447441816329956, + "learning_rate": 3.5997859016274725e-06, + "loss": 0.4458, + "step": 52854 + }, + { + "epoch": 3.667684407744084, + "grad_norm": 0.8001025915145874, + "learning_rate": 3.5982920949762968e-06, + "loss": 0.2904, + "step": 52855 + }, + { + "epoch": 3.667753799181181, + "grad_norm": 0.5568177700042725, + "learning_rate": 3.596798592654138e-06, + "loss": 0.1817, + "step": 52856 + }, + { + "epoch": 3.667823190618278, + "grad_norm": 1.1110326051712036, + "learning_rate": 3.5953053946656934e-06, + "loss": 0.2512, + "step": 52857 + }, + { + "epoch": 3.6678925820553743, + "grad_norm": 0.7500618696212769, + "learning_rate": 3.593812501015681e-06, + "loss": 0.1642, + "step": 52858 + }, + { + "epoch": 3.667961973492471, + "grad_norm": 0.8113080859184265, + "learning_rate": 3.59231991170883e-06, + "loss": 0.2385, + "step": 52859 + }, + { + "epoch": 3.6680313649295675, + "grad_norm": 1.3406553268432617, + "learning_rate": 3.5908276267498262e-06, + "loss": 0.1414, + "step": 52860 + }, + { + "epoch": 3.6681007563666643, + "grad_norm": 0.9350559711456299, + "learning_rate": 3.5893356461433993e-06, + "loss": 0.2144, + "step": 52861 + }, + { + "epoch": 3.668170147803761, + "grad_norm": 0.5711258053779602, + "learning_rate": 3.587843969894256e-06, + "loss": 0.1998, + "step": 52862 + }, + { + "epoch": 3.6682395392408575, + "grad_norm": 1.0950944423675537, + "learning_rate": 3.5863525980071035e-06, + "loss": 0.3671, + "step": 52863 + }, + { + "epoch": 3.6683089306779544, + "grad_norm": 0.8268890380859375, + "learning_rate": 3.5848615304866494e-06, + "loss": 0.2304, + "step": 52864 + }, + { + "epoch": 3.6683783221150508, + "grad_norm": 0.8021008968353271, + "learning_rate": 3.5833707673375903e-06, + "loss": 0.1985, + "step": 52865 + }, + { + "epoch": 3.6684477135521476, + "grad_norm": 0.6858654618263245, + "learning_rate": 3.581880308564667e-06, + "loss": 0.2374, + "step": 52866 + }, + { + "epoch": 3.6685171049892444, + "grad_norm": 0.8606309294700623, + "learning_rate": 3.5803901541725416e-06, + "loss": 0.2323, + "step": 52867 + }, + { + "epoch": 3.668586496426341, + "grad_norm": 0.8299965858459473, + "learning_rate": 3.5789003041659443e-06, + "loss": 0.1998, + "step": 52868 + }, + { + "epoch": 3.6686558878634377, + "grad_norm": 0.7189550399780273, + "learning_rate": 3.5774107585495708e-06, + "loss": 0.1605, + "step": 52869 + }, + { + "epoch": 3.668725279300534, + "grad_norm": 0.9201632738113403, + "learning_rate": 3.575921517328118e-06, + "loss": 0.2203, + "step": 52870 + }, + { + "epoch": 3.668794670737631, + "grad_norm": 0.9051681160926819, + "learning_rate": 3.574432580506304e-06, + "loss": 0.2602, + "step": 52871 + }, + { + "epoch": 3.6688640621747277, + "grad_norm": 1.027970790863037, + "learning_rate": 3.572943948088814e-06, + "loss": 0.1901, + "step": 52872 + }, + { + "epoch": 3.6689334536118245, + "grad_norm": 0.6885640025138855, + "learning_rate": 3.571455620080366e-06, + "loss": 0.2091, + "step": 52873 + }, + { + "epoch": 3.669002845048921, + "grad_norm": 1.1076538562774658, + "learning_rate": 3.5699675964856238e-06, + "loss": 0.2361, + "step": 52874 + }, + { + "epoch": 3.6690722364860178, + "grad_norm": 1.274052619934082, + "learning_rate": 3.5684798773093165e-06, + "loss": 0.3133, + "step": 52875 + }, + { + "epoch": 3.669141627923114, + "grad_norm": 1.2637856006622314, + "learning_rate": 3.566992462556129e-06, + "loss": 0.3314, + "step": 52876 + }, + { + "epoch": 3.669211019360211, + "grad_norm": 0.8653051257133484, + "learning_rate": 3.565505352230747e-06, + "loss": 0.2414, + "step": 52877 + }, + { + "epoch": 3.669280410797308, + "grad_norm": 1.146427035331726, + "learning_rate": 3.5640185463378995e-06, + "loss": 0.2277, + "step": 52878 + }, + { + "epoch": 3.669349802234404, + "grad_norm": 0.821749210357666, + "learning_rate": 3.5625320448822386e-06, + "loss": 0.2505, + "step": 52879 + }, + { + "epoch": 3.669419193671501, + "grad_norm": 1.2128407955169678, + "learning_rate": 3.5610458478684826e-06, + "loss": 0.3057, + "step": 52880 + }, + { + "epoch": 3.6694885851085974, + "grad_norm": 0.9315517544746399, + "learning_rate": 3.559559955301306e-06, + "loss": 0.3274, + "step": 52881 + }, + { + "epoch": 3.6695579765456943, + "grad_norm": 1.0222057104110718, + "learning_rate": 3.5580743671854267e-06, + "loss": 0.2887, + "step": 52882 + }, + { + "epoch": 3.669627367982791, + "grad_norm": 0.9175231456756592, + "learning_rate": 3.556589083525508e-06, + "loss": 0.2519, + "step": 52883 + }, + { + "epoch": 3.6696967594198875, + "grad_norm": 0.8046415448188782, + "learning_rate": 3.5551041043262455e-06, + "loss": 0.242, + "step": 52884 + }, + { + "epoch": 3.6697661508569843, + "grad_norm": 0.7526302933692932, + "learning_rate": 3.5536194295923363e-06, + "loss": 0.3133, + "step": 52885 + }, + { + "epoch": 3.6698355422940807, + "grad_norm": 0.8020963072776794, + "learning_rate": 3.552135059328454e-06, + "loss": 0.23, + "step": 52886 + }, + { + "epoch": 3.6699049337311775, + "grad_norm": 1.3132686614990234, + "learning_rate": 3.550650993539306e-06, + "loss": 0.2521, + "step": 52887 + }, + { + "epoch": 3.6699743251682744, + "grad_norm": 1.177404761314392, + "learning_rate": 3.549167232229544e-06, + "loss": 0.2317, + "step": 52888 + }, + { + "epoch": 3.6700437166053708, + "grad_norm": 0.6774498224258423, + "learning_rate": 3.547683775403876e-06, + "loss": 0.1823, + "step": 52889 + }, + { + "epoch": 3.6701131080424676, + "grad_norm": 0.9578673243522644, + "learning_rate": 3.546200623066975e-06, + "loss": 0.3176, + "step": 52890 + }, + { + "epoch": 3.670182499479564, + "grad_norm": 0.7897270321846008, + "learning_rate": 3.5447177752235383e-06, + "loss": 0.2687, + "step": 52891 + }, + { + "epoch": 3.670251890916661, + "grad_norm": 0.7574769854545593, + "learning_rate": 3.543235231878228e-06, + "loss": 0.2063, + "step": 52892 + }, + { + "epoch": 3.6703212823537577, + "grad_norm": 0.920184314250946, + "learning_rate": 3.5417529930357297e-06, + "loss": 0.243, + "step": 52893 + }, + { + "epoch": 3.6703906737908545, + "grad_norm": 0.79635089635849, + "learning_rate": 3.5402710587007504e-06, + "loss": 0.1587, + "step": 52894 + }, + { + "epoch": 3.670460065227951, + "grad_norm": 0.8602080345153809, + "learning_rate": 3.53878942887792e-06, + "loss": 0.1817, + "step": 52895 + }, + { + "epoch": 3.6705294566650473, + "grad_norm": 0.8309805989265442, + "learning_rate": 3.5373081035719457e-06, + "loss": 0.2728, + "step": 52896 + }, + { + "epoch": 3.670598848102144, + "grad_norm": 0.8174852728843689, + "learning_rate": 3.5358270827874906e-06, + "loss": 0.2487, + "step": 52897 + }, + { + "epoch": 3.670668239539241, + "grad_norm": 0.8489221930503845, + "learning_rate": 3.534346366529251e-06, + "loss": 0.219, + "step": 52898 + }, + { + "epoch": 3.6707376309763378, + "grad_norm": 0.8498846292495728, + "learning_rate": 3.5328659548018785e-06, + "loss": 0.3127, + "step": 52899 + }, + { + "epoch": 3.670807022413434, + "grad_norm": 0.8649004101753235, + "learning_rate": 3.531385847610058e-06, + "loss": 0.2245, + "step": 52900 + }, + { + "epoch": 3.670876413850531, + "grad_norm": 0.8090595006942749, + "learning_rate": 3.529906044958475e-06, + "loss": 0.1235, + "step": 52901 + }, + { + "epoch": 3.6709458052876274, + "grad_norm": 1.034088373184204, + "learning_rate": 3.5284265468517707e-06, + "loss": 0.2811, + "step": 52902 + }, + { + "epoch": 3.671015196724724, + "grad_norm": 0.8440188765525818, + "learning_rate": 3.5269473532946296e-06, + "loss": 0.2781, + "step": 52903 + }, + { + "epoch": 3.671084588161821, + "grad_norm": 0.7787423729896545, + "learning_rate": 3.5254684642917257e-06, + "loss": 0.2515, + "step": 52904 + }, + { + "epoch": 3.6711539795989174, + "grad_norm": 0.7229148149490356, + "learning_rate": 3.5239898798477223e-06, + "loss": 0.1406, + "step": 52905 + }, + { + "epoch": 3.6712233710360143, + "grad_norm": 1.1923850774765015, + "learning_rate": 3.522511599967293e-06, + "loss": 0.1879, + "step": 52906 + }, + { + "epoch": 3.6712927624731106, + "grad_norm": 0.7115556597709656, + "learning_rate": 3.521033624655101e-06, + "loss": 0.1864, + "step": 52907 + }, + { + "epoch": 3.6713621539102075, + "grad_norm": 0.8031206727027893, + "learning_rate": 3.519555953915821e-06, + "loss": 0.1823, + "step": 52908 + }, + { + "epoch": 3.6714315453473043, + "grad_norm": 1.0298633575439453, + "learning_rate": 3.5180785877540924e-06, + "loss": 0.2903, + "step": 52909 + }, + { + "epoch": 3.6715009367844007, + "grad_norm": 0.7786320447921753, + "learning_rate": 3.516601526174601e-06, + "loss": 0.2311, + "step": 52910 + }, + { + "epoch": 3.6715703282214975, + "grad_norm": 0.8493711948394775, + "learning_rate": 3.51512476918201e-06, + "loss": 0.2593, + "step": 52911 + }, + { + "epoch": 3.671639719658594, + "grad_norm": 0.774604856967926, + "learning_rate": 3.5136483167809708e-06, + "loss": 0.244, + "step": 52912 + }, + { + "epoch": 3.6717091110956908, + "grad_norm": 1.0319122076034546, + "learning_rate": 3.512172168976158e-06, + "loss": 0.1663, + "step": 52913 + }, + { + "epoch": 3.6717785025327876, + "grad_norm": 0.7995699644088745, + "learning_rate": 3.5106963257722114e-06, + "loss": 0.2239, + "step": 52914 + }, + { + "epoch": 3.671847893969884, + "grad_norm": 0.6420539617538452, + "learning_rate": 3.509220787173806e-06, + "loss": 0.1838, + "step": 52915 + }, + { + "epoch": 3.671917285406981, + "grad_norm": 1.1153733730316162, + "learning_rate": 3.507745553185593e-06, + "loss": 0.2925, + "step": 52916 + }, + { + "epoch": 3.671986676844077, + "grad_norm": 0.6170710921287537, + "learning_rate": 3.5062706238122466e-06, + "loss": 0.1516, + "step": 52917 + }, + { + "epoch": 3.672056068281174, + "grad_norm": 1.257214903831482, + "learning_rate": 3.504795999058397e-06, + "loss": 0.4063, + "step": 52918 + }, + { + "epoch": 3.672125459718271, + "grad_norm": 0.7013484239578247, + "learning_rate": 3.503321678928706e-06, + "loss": 0.2463, + "step": 52919 + }, + { + "epoch": 3.6721948511553677, + "grad_norm": 1.0476346015930176, + "learning_rate": 3.5018476634278375e-06, + "loss": 0.3916, + "step": 52920 + }, + { + "epoch": 3.672264242592464, + "grad_norm": 0.8568995594978333, + "learning_rate": 3.5003739525604427e-06, + "loss": 0.206, + "step": 52921 + }, + { + "epoch": 3.6723336340295605, + "grad_norm": 1.1877797842025757, + "learning_rate": 3.498900546331174e-06, + "loss": 0.3796, + "step": 52922 + }, + { + "epoch": 3.6724030254666573, + "grad_norm": 1.2116031646728516, + "learning_rate": 3.497427444744672e-06, + "loss": 0.4253, + "step": 52923 + }, + { + "epoch": 3.672472416903754, + "grad_norm": 0.892752468585968, + "learning_rate": 3.4959546478055995e-06, + "loss": 0.3366, + "step": 52924 + }, + { + "epoch": 3.672541808340851, + "grad_norm": 0.7295418381690979, + "learning_rate": 3.494482155518597e-06, + "loss": 0.2072, + "step": 52925 + }, + { + "epoch": 3.6726111997779474, + "grad_norm": 0.7274441719055176, + "learning_rate": 3.4930099678883166e-06, + "loss": 0.1842, + "step": 52926 + }, + { + "epoch": 3.672680591215044, + "grad_norm": 0.7220281958580017, + "learning_rate": 3.4915380849194102e-06, + "loss": 0.1916, + "step": 52927 + }, + { + "epoch": 3.6727499826521406, + "grad_norm": 1.329685926437378, + "learning_rate": 3.4900665066165185e-06, + "loss": 0.1954, + "step": 52928 + }, + { + "epoch": 3.6728193740892374, + "grad_norm": 1.2052279710769653, + "learning_rate": 3.488595232984304e-06, + "loss": 0.2951, + "step": 52929 + }, + { + "epoch": 3.6728887655263343, + "grad_norm": 0.6552150249481201, + "learning_rate": 3.487124264027386e-06, + "loss": 0.1522, + "step": 52930 + }, + { + "epoch": 3.6729581569634306, + "grad_norm": 1.011803150177002, + "learning_rate": 3.485653599750416e-06, + "loss": 0.3356, + "step": 52931 + }, + { + "epoch": 3.6730275484005275, + "grad_norm": 0.9841328263282776, + "learning_rate": 3.4841832401580456e-06, + "loss": 0.2514, + "step": 52932 + }, + { + "epoch": 3.673096939837624, + "grad_norm": 0.9599246382713318, + "learning_rate": 3.4827131852549044e-06, + "loss": 0.2139, + "step": 52933 + }, + { + "epoch": 3.6731663312747207, + "grad_norm": 0.8763537406921387, + "learning_rate": 3.481243435045645e-06, + "loss": 0.2271, + "step": 52934 + }, + { + "epoch": 3.6732357227118175, + "grad_norm": 1.0999492406845093, + "learning_rate": 3.479773989534907e-06, + "loss": 0.2487, + "step": 52935 + }, + { + "epoch": 3.673305114148914, + "grad_norm": 0.9176873564720154, + "learning_rate": 3.4783048487273095e-06, + "loss": 0.2471, + "step": 52936 + }, + { + "epoch": 3.6733745055860108, + "grad_norm": 0.7847295999526978, + "learning_rate": 3.4768360126275156e-06, + "loss": 0.2354, + "step": 52937 + }, + { + "epoch": 3.673443897023107, + "grad_norm": 0.843348503112793, + "learning_rate": 3.475367481240144e-06, + "loss": 0.2128, + "step": 52938 + }, + { + "epoch": 3.673513288460204, + "grad_norm": 0.7523249387741089, + "learning_rate": 3.4738992545698456e-06, + "loss": 0.2521, + "step": 52939 + }, + { + "epoch": 3.673582679897301, + "grad_norm": 1.060494303703308, + "learning_rate": 3.47243133262124e-06, + "loss": 0.2285, + "step": 52940 + }, + { + "epoch": 3.673652071334397, + "grad_norm": 0.5878137946128845, + "learning_rate": 3.4709637153989784e-06, + "loss": 0.126, + "step": 52941 + }, + { + "epoch": 3.673721462771494, + "grad_norm": 1.1671658754348755, + "learning_rate": 3.4694964029076905e-06, + "loss": 0.3401, + "step": 52942 + }, + { + "epoch": 3.6737908542085904, + "grad_norm": 1.0914603471755981, + "learning_rate": 3.468029395151995e-06, + "loss": 0.476, + "step": 52943 + }, + { + "epoch": 3.6738602456456873, + "grad_norm": 0.9191989302635193, + "learning_rate": 3.466562692136521e-06, + "loss": 0.2602, + "step": 52944 + }, + { + "epoch": 3.673929637082784, + "grad_norm": 0.6279016137123108, + "learning_rate": 3.4650962938659214e-06, + "loss": 0.1263, + "step": 52945 + }, + { + "epoch": 3.673999028519881, + "grad_norm": 0.8715167045593262, + "learning_rate": 3.463630200344803e-06, + "loss": 0.196, + "step": 52946 + }, + { + "epoch": 3.6740684199569773, + "grad_norm": 0.7335593700408936, + "learning_rate": 3.4621644115778063e-06, + "loss": 0.2544, + "step": 52947 + }, + { + "epoch": 3.6741378113940737, + "grad_norm": 1.1148093938827515, + "learning_rate": 3.4606989275695723e-06, + "loss": 0.3206, + "step": 52948 + }, + { + "epoch": 3.6742072028311705, + "grad_norm": 0.9492875337600708, + "learning_rate": 3.4592337483246972e-06, + "loss": 0.3397, + "step": 52949 + }, + { + "epoch": 3.6742765942682674, + "grad_norm": 0.9348167777061462, + "learning_rate": 3.457768873847822e-06, + "loss": 0.2245, + "step": 52950 + }, + { + "epoch": 3.674345985705364, + "grad_norm": 0.80461585521698, + "learning_rate": 3.456304304143576e-06, + "loss": 0.2052, + "step": 52951 + }, + { + "epoch": 3.6744153771424606, + "grad_norm": 0.8027936220169067, + "learning_rate": 3.454840039216567e-06, + "loss": 0.2253, + "step": 52952 + }, + { + "epoch": 3.6744847685795574, + "grad_norm": 1.1896467208862305, + "learning_rate": 3.4533760790714243e-06, + "loss": 0.3813, + "step": 52953 + }, + { + "epoch": 3.674554160016654, + "grad_norm": 1.0051095485687256, + "learning_rate": 3.4519124237127776e-06, + "loss": 0.3565, + "step": 52954 + }, + { + "epoch": 3.6746235514537506, + "grad_norm": 1.076108455657959, + "learning_rate": 3.4504490731452344e-06, + "loss": 0.4213, + "step": 52955 + }, + { + "epoch": 3.6746929428908475, + "grad_norm": 0.6456583738327026, + "learning_rate": 3.4489860273734464e-06, + "loss": 0.1317, + "step": 52956 + }, + { + "epoch": 3.674762334327944, + "grad_norm": 0.8344406485557556, + "learning_rate": 3.447523286401988e-06, + "loss": 0.2081, + "step": 52957 + }, + { + "epoch": 3.6748317257650407, + "grad_norm": 0.5521381497383118, + "learning_rate": 3.4460608502354887e-06, + "loss": 0.1845, + "step": 52958 + }, + { + "epoch": 3.674901117202137, + "grad_norm": 0.6945587992668152, + "learning_rate": 3.444598718878578e-06, + "loss": 0.194, + "step": 52959 + }, + { + "epoch": 3.674970508639234, + "grad_norm": 0.6672000885009766, + "learning_rate": 3.443136892335874e-06, + "loss": 0.1201, + "step": 52960 + }, + { + "epoch": 3.6750399000763307, + "grad_norm": 0.9510357975959778, + "learning_rate": 3.4416753706119743e-06, + "loss": 0.3366, + "step": 52961 + }, + { + "epoch": 3.675109291513427, + "grad_norm": 0.7010124325752258, + "learning_rate": 3.4402141537115074e-06, + "loss": 0.2634, + "step": 52962 + }, + { + "epoch": 3.675178682950524, + "grad_norm": 1.2287957668304443, + "learning_rate": 3.438753241639092e-06, + "loss": 0.1803, + "step": 52963 + }, + { + "epoch": 3.6752480743876204, + "grad_norm": 0.7537556886672974, + "learning_rate": 3.437292634399325e-06, + "loss": 0.1739, + "step": 52964 + }, + { + "epoch": 3.675317465824717, + "grad_norm": 0.9274728894233704, + "learning_rate": 3.4358323319968132e-06, + "loss": 0.2372, + "step": 52965 + }, + { + "epoch": 3.675386857261814, + "grad_norm": 0.6901623010635376, + "learning_rate": 3.4343723344361757e-06, + "loss": 0.2512, + "step": 52966 + }, + { + "epoch": 3.6754562486989104, + "grad_norm": 0.9513075947761536, + "learning_rate": 3.432912641722019e-06, + "loss": 0.2982, + "step": 52967 + }, + { + "epoch": 3.6755256401360072, + "grad_norm": 0.7967249751091003, + "learning_rate": 3.431453253858963e-06, + "loss": 0.1693, + "step": 52968 + }, + { + "epoch": 3.6755950315731036, + "grad_norm": 0.8480421304702759, + "learning_rate": 3.4299941708515913e-06, + "loss": 0.2131, + "step": 52969 + }, + { + "epoch": 3.6756644230102005, + "grad_norm": 0.8240949511528015, + "learning_rate": 3.428535392704546e-06, + "loss": 0.1296, + "step": 52970 + }, + { + "epoch": 3.6757338144472973, + "grad_norm": 1.034045696258545, + "learning_rate": 3.4270769194223895e-06, + "loss": 0.2733, + "step": 52971 + }, + { + "epoch": 3.675803205884394, + "grad_norm": 2.745941638946533, + "learning_rate": 3.4256187510097627e-06, + "loss": 0.3681, + "step": 52972 + }, + { + "epoch": 3.6758725973214905, + "grad_norm": 0.8854989409446716, + "learning_rate": 3.4241608874712394e-06, + "loss": 0.1167, + "step": 52973 + }, + { + "epoch": 3.6759419887585874, + "grad_norm": 1.0732579231262207, + "learning_rate": 3.4227033288114386e-06, + "loss": 0.3371, + "step": 52974 + }, + { + "epoch": 3.6760113801956837, + "grad_norm": 0.8137110471725464, + "learning_rate": 3.421246075034956e-06, + "loss": 0.3045, + "step": 52975 + }, + { + "epoch": 3.6760807716327806, + "grad_norm": 0.8819783926010132, + "learning_rate": 3.4197891261464e-06, + "loss": 0.2754, + "step": 52976 + }, + { + "epoch": 3.6761501630698774, + "grad_norm": 0.8555946350097656, + "learning_rate": 3.418332482150366e-06, + "loss": 0.2037, + "step": 52977 + }, + { + "epoch": 3.676219554506974, + "grad_norm": 0.9190030097961426, + "learning_rate": 3.4168761430514507e-06, + "loss": 0.3191, + "step": 52978 + }, + { + "epoch": 3.6762889459440706, + "grad_norm": 0.9143438935279846, + "learning_rate": 3.4154201088542506e-06, + "loss": 0.2254, + "step": 52979 + }, + { + "epoch": 3.676358337381167, + "grad_norm": 1.523876428604126, + "learning_rate": 3.41396437956335e-06, + "loss": 0.555, + "step": 52980 + }, + { + "epoch": 3.676427728818264, + "grad_norm": 0.9033908247947693, + "learning_rate": 3.412508955183369e-06, + "loss": 0.3176, + "step": 52981 + }, + { + "epoch": 3.6764971202553607, + "grad_norm": 0.75307697057724, + "learning_rate": 3.4110538357188915e-06, + "loss": 0.1943, + "step": 52982 + }, + { + "epoch": 3.676566511692457, + "grad_norm": 1.336175560951233, + "learning_rate": 3.4095990211745143e-06, + "loss": 0.2593, + "step": 52983 + }, + { + "epoch": 3.676635903129554, + "grad_norm": 0.875470757484436, + "learning_rate": 3.408144511554823e-06, + "loss": 0.2151, + "step": 52984 + }, + { + "epoch": 3.6767052945666503, + "grad_norm": 0.6017130613327026, + "learning_rate": 3.406690306864413e-06, + "loss": 0.2599, + "step": 52985 + }, + { + "epoch": 3.676774686003747, + "grad_norm": 0.7595446705818176, + "learning_rate": 3.405236407107881e-06, + "loss": 0.2298, + "step": 52986 + }, + { + "epoch": 3.676844077440844, + "grad_norm": 0.9449142813682556, + "learning_rate": 3.403782812289802e-06, + "loss": 0.3852, + "step": 52987 + }, + { + "epoch": 3.6769134688779403, + "grad_norm": 1.468275547027588, + "learning_rate": 3.402329522414771e-06, + "loss": 0.2637, + "step": 52988 + }, + { + "epoch": 3.676982860315037, + "grad_norm": 0.8833169341087341, + "learning_rate": 3.400876537487385e-06, + "loss": 0.2446, + "step": 52989 + }, + { + "epoch": 3.6770522517521336, + "grad_norm": 0.7084528207778931, + "learning_rate": 3.3994238575122182e-06, + "loss": 0.211, + "step": 52990 + }, + { + "epoch": 3.6771216431892304, + "grad_norm": 3.334298610687256, + "learning_rate": 3.397971482493878e-06, + "loss": 0.4141, + "step": 52991 + }, + { + "epoch": 3.6771910346263272, + "grad_norm": 0.9316712617874146, + "learning_rate": 3.396519412436916e-06, + "loss": 0.3232, + "step": 52992 + }, + { + "epoch": 3.6772604260634236, + "grad_norm": 0.9062972068786621, + "learning_rate": 3.3950676473459397e-06, + "loss": 0.2826, + "step": 52993 + }, + { + "epoch": 3.6773298175005205, + "grad_norm": 0.7964937090873718, + "learning_rate": 3.3936161872255233e-06, + "loss": 0.2157, + "step": 52994 + }, + { + "epoch": 3.677399208937617, + "grad_norm": 0.7885552644729614, + "learning_rate": 3.3921650320802522e-06, + "loss": 0.1653, + "step": 52995 + }, + { + "epoch": 3.6774686003747137, + "grad_norm": 1.1003550291061401, + "learning_rate": 3.3907141819147114e-06, + "loss": 0.2763, + "step": 52996 + }, + { + "epoch": 3.6775379918118105, + "grad_norm": 0.7813948392868042, + "learning_rate": 3.389263636733475e-06, + "loss": 0.2462, + "step": 52997 + }, + { + "epoch": 3.6776073832489073, + "grad_norm": 1.1417930126190186, + "learning_rate": 3.3878133965411396e-06, + "loss": 0.2526, + "step": 52998 + }, + { + "epoch": 3.6776767746860037, + "grad_norm": 0.8541239500045776, + "learning_rate": 3.3863634613422456e-06, + "loss": 0.3329, + "step": 52999 + }, + { + "epoch": 3.6777461661231006, + "grad_norm": 1.0207961797714233, + "learning_rate": 3.384913831141401e-06, + "loss": 0.3319, + "step": 53000 + }, + { + "epoch": 3.677815557560197, + "grad_norm": 1.1750884056091309, + "learning_rate": 3.383464505943179e-06, + "loss": 0.2951, + "step": 53001 + }, + { + "epoch": 3.677884948997294, + "grad_norm": 0.9782153367996216, + "learning_rate": 3.3820154857521434e-06, + "loss": 0.3383, + "step": 53002 + }, + { + "epoch": 3.6779543404343906, + "grad_norm": 1.1001789569854736, + "learning_rate": 3.3805667705728795e-06, + "loss": 0.1735, + "step": 53003 + }, + { + "epoch": 3.678023731871487, + "grad_norm": 1.5704745054244995, + "learning_rate": 3.3791183604099496e-06, + "loss": 0.4274, + "step": 53004 + }, + { + "epoch": 3.678093123308584, + "grad_norm": 1.445233702659607, + "learning_rate": 3.3776702552679505e-06, + "loss": 0.2584, + "step": 53005 + }, + { + "epoch": 3.6781625147456802, + "grad_norm": 0.9490081071853638, + "learning_rate": 3.376222455151412e-06, + "loss": 0.2135, + "step": 53006 + }, + { + "epoch": 3.678231906182777, + "grad_norm": 0.838209331035614, + "learning_rate": 3.3747749600649413e-06, + "loss": 0.1994, + "step": 53007 + }, + { + "epoch": 3.678301297619874, + "grad_norm": 0.5744810700416565, + "learning_rate": 3.3733277700130906e-06, + "loss": 0.2072, + "step": 53008 + }, + { + "epoch": 3.6783706890569703, + "grad_norm": 1.3918976783752441, + "learning_rate": 3.3718808850004446e-06, + "loss": 0.451, + "step": 53009 + }, + { + "epoch": 3.678440080494067, + "grad_norm": 0.7477777600288391, + "learning_rate": 3.3704343050315446e-06, + "loss": 0.2143, + "step": 53010 + }, + { + "epoch": 3.6785094719311635, + "grad_norm": 0.9950288534164429, + "learning_rate": 3.3689880301109866e-06, + "loss": 0.3324, + "step": 53011 + }, + { + "epoch": 3.6785788633682603, + "grad_norm": 0.7586365342140198, + "learning_rate": 3.3675420602433225e-06, + "loss": 0.1677, + "step": 53012 + }, + { + "epoch": 3.678648254805357, + "grad_norm": 1.0219314098358154, + "learning_rate": 3.366096395433105e-06, + "loss": 0.2155, + "step": 53013 + }, + { + "epoch": 3.6787176462424536, + "grad_norm": 0.5803285837173462, + "learning_rate": 3.364651035684907e-06, + "loss": 0.1541, + "step": 53014 + }, + { + "epoch": 3.6787870376795504, + "grad_norm": 0.9605403542518616, + "learning_rate": 3.3632059810033033e-06, + "loss": 0.1977, + "step": 53015 + }, + { + "epoch": 3.678856429116647, + "grad_norm": 0.8435836434364319, + "learning_rate": 3.361761231392846e-06, + "loss": 0.1812, + "step": 53016 + }, + { + "epoch": 3.6789258205537436, + "grad_norm": 0.8200467824935913, + "learning_rate": 3.3603167868580977e-06, + "loss": 0.2081, + "step": 53017 + }, + { + "epoch": 3.6789952119908405, + "grad_norm": 0.8452894687652588, + "learning_rate": 3.3588726474036215e-06, + "loss": 0.2381, + "step": 53018 + }, + { + "epoch": 3.679064603427937, + "grad_norm": 0.9002548456192017, + "learning_rate": 3.357428813033958e-06, + "loss": 0.2636, + "step": 53019 + }, + { + "epoch": 3.6791339948650337, + "grad_norm": 1.2047784328460693, + "learning_rate": 3.355985283753682e-06, + "loss": 0.3749, + "step": 53020 + }, + { + "epoch": 3.67920338630213, + "grad_norm": 0.8235478401184082, + "learning_rate": 3.3545420595673673e-06, + "loss": 0.2197, + "step": 53021 + }, + { + "epoch": 3.679272777739227, + "grad_norm": 0.650657594203949, + "learning_rate": 3.353099140479532e-06, + "loss": 0.2329, + "step": 53022 + }, + { + "epoch": 3.6793421691763237, + "grad_norm": 2.2117505073547363, + "learning_rate": 3.3516565264947508e-06, + "loss": 0.2455, + "step": 53023 + }, + { + "epoch": 3.6794115606134206, + "grad_norm": 0.9236979484558105, + "learning_rate": 3.3502142176175755e-06, + "loss": 0.2552, + "step": 53024 + }, + { + "epoch": 3.679480952050517, + "grad_norm": 0.9373269081115723, + "learning_rate": 3.348772213852569e-06, + "loss": 0.2975, + "step": 53025 + }, + { + "epoch": 3.679550343487614, + "grad_norm": 1.0748085975646973, + "learning_rate": 3.347330515204261e-06, + "loss": 0.2924, + "step": 53026 + }, + { + "epoch": 3.67961973492471, + "grad_norm": 0.8891423344612122, + "learning_rate": 3.345889121677226e-06, + "loss": 0.1993, + "step": 53027 + }, + { + "epoch": 3.679689126361807, + "grad_norm": 0.8194762468338013, + "learning_rate": 3.344448033276004e-06, + "loss": 0.237, + "step": 53028 + }, + { + "epoch": 3.679758517798904, + "grad_norm": 1.3133599758148193, + "learning_rate": 3.343007250005148e-06, + "loss": 0.2114, + "step": 53029 + }, + { + "epoch": 3.6798279092360002, + "grad_norm": 0.8713664412498474, + "learning_rate": 3.341566771869198e-06, + "loss": 0.2393, + "step": 53030 + }, + { + "epoch": 3.679897300673097, + "grad_norm": 0.9174558520317078, + "learning_rate": 3.340126598872706e-06, + "loss": 0.2508, + "step": 53031 + }, + { + "epoch": 3.6799666921101934, + "grad_norm": 0.9388232231140137, + "learning_rate": 3.3386867310202353e-06, + "loss": 0.261, + "step": 53032 + }, + { + "epoch": 3.6800360835472903, + "grad_norm": 1.282896876335144, + "learning_rate": 3.337247168316304e-06, + "loss": 0.2365, + "step": 53033 + }, + { + "epoch": 3.680105474984387, + "grad_norm": 1.1165770292282104, + "learning_rate": 3.335807910765465e-06, + "loss": 0.2794, + "step": 53034 + }, + { + "epoch": 3.6801748664214835, + "grad_norm": 0.43421992659568787, + "learning_rate": 3.3343689583722693e-06, + "loss": 0.1092, + "step": 53035 + }, + { + "epoch": 3.6802442578585803, + "grad_norm": 1.0758110284805298, + "learning_rate": 3.332930311141247e-06, + "loss": 0.2458, + "step": 53036 + }, + { + "epoch": 3.6803136492956767, + "grad_norm": 1.2842028141021729, + "learning_rate": 3.3314919690769608e-06, + "loss": 0.2292, + "step": 53037 + }, + { + "epoch": 3.6803830407327736, + "grad_norm": 1.2029582262039185, + "learning_rate": 3.3300539321839297e-06, + "loss": 0.2463, + "step": 53038 + }, + { + "epoch": 3.6804524321698704, + "grad_norm": 0.8546351194381714, + "learning_rate": 3.3286162004667165e-06, + "loss": 0.1399, + "step": 53039 + }, + { + "epoch": 3.680521823606967, + "grad_norm": 2.5567288398742676, + "learning_rate": 3.32717877392984e-06, + "loss": 0.1839, + "step": 53040 + }, + { + "epoch": 3.6805912150440636, + "grad_norm": 1.0038492679595947, + "learning_rate": 3.3257416525778294e-06, + "loss": 0.3724, + "step": 53041 + }, + { + "epoch": 3.68066060648116, + "grad_norm": 0.6670615673065186, + "learning_rate": 3.3243048364152484e-06, + "loss": 0.22, + "step": 53042 + }, + { + "epoch": 3.680729997918257, + "grad_norm": 1.0345996618270874, + "learning_rate": 3.3228683254466152e-06, + "loss": 0.1364, + "step": 53043 + }, + { + "epoch": 3.6807993893553537, + "grad_norm": 1.347092866897583, + "learning_rate": 3.321432119676482e-06, + "loss": 0.3831, + "step": 53044 + }, + { + "epoch": 3.68086878079245, + "grad_norm": 0.666997492313385, + "learning_rate": 3.3199962191093557e-06, + "loss": 0.2095, + "step": 53045 + }, + { + "epoch": 3.680938172229547, + "grad_norm": 0.6339298486709595, + "learning_rate": 3.3185606237498e-06, + "loss": 0.1905, + "step": 53046 + }, + { + "epoch": 3.6810075636666433, + "grad_norm": 0.8258510231971741, + "learning_rate": 3.3171253336023224e-06, + "loss": 0.2692, + "step": 53047 + }, + { + "epoch": 3.68107695510374, + "grad_norm": 1.316977858543396, + "learning_rate": 3.315690348671463e-06, + "loss": 0.433, + "step": 53048 + }, + { + "epoch": 3.681146346540837, + "grad_norm": 0.7872257828712463, + "learning_rate": 3.314255668961752e-06, + "loss": 0.2373, + "step": 53049 + }, + { + "epoch": 3.6812157379779338, + "grad_norm": 0.7231152057647705, + "learning_rate": 3.312821294477719e-06, + "loss": 0.1514, + "step": 53050 + }, + { + "epoch": 3.68128512941503, + "grad_norm": 0.8491038680076599, + "learning_rate": 3.311387225223883e-06, + "loss": 0.236, + "step": 53051 + }, + { + "epoch": 3.681354520852127, + "grad_norm": 0.7074378728866577, + "learning_rate": 3.3099534612048068e-06, + "loss": 0.2243, + "step": 53052 + }, + { + "epoch": 3.6814239122892234, + "grad_norm": 0.8475281000137329, + "learning_rate": 3.308520002424964e-06, + "loss": 0.2627, + "step": 53053 + }, + { + "epoch": 3.68149330372632, + "grad_norm": 0.9538949728012085, + "learning_rate": 3.3070868488889075e-06, + "loss": 0.3207, + "step": 53054 + }, + { + "epoch": 3.681562695163417, + "grad_norm": 1.0910899639129639, + "learning_rate": 3.305654000601166e-06, + "loss": 0.241, + "step": 53055 + }, + { + "epoch": 3.6816320866005134, + "grad_norm": 0.8729639053344727, + "learning_rate": 3.3042214575662588e-06, + "loss": 0.2158, + "step": 53056 + }, + { + "epoch": 3.6817014780376103, + "grad_norm": 0.6677645444869995, + "learning_rate": 3.3027892197887045e-06, + "loss": 0.1707, + "step": 53057 + }, + { + "epoch": 3.6817708694747067, + "grad_norm": 1.0474580526351929, + "learning_rate": 3.3013572872730216e-06, + "loss": 0.2961, + "step": 53058 + }, + { + "epoch": 3.6818402609118035, + "grad_norm": 0.9008971452713013, + "learning_rate": 3.2999256600237282e-06, + "loss": 0.312, + "step": 53059 + }, + { + "epoch": 3.6819096523489003, + "grad_norm": 1.0384284257888794, + "learning_rate": 3.298494338045366e-06, + "loss": 0.3829, + "step": 53060 + }, + { + "epoch": 3.6819790437859967, + "grad_norm": 0.7265560030937195, + "learning_rate": 3.297063321342431e-06, + "loss": 0.1825, + "step": 53061 + }, + { + "epoch": 3.6820484352230936, + "grad_norm": 0.927967369556427, + "learning_rate": 3.2956326099194414e-06, + "loss": 0.19, + "step": 53062 + }, + { + "epoch": 3.68211782666019, + "grad_norm": 1.3833037614822388, + "learning_rate": 3.2942022037809163e-06, + "loss": 0.5122, + "step": 53063 + }, + { + "epoch": 3.6821872180972868, + "grad_norm": 1.1055907011032104, + "learning_rate": 3.292772102931385e-06, + "loss": 0.1873, + "step": 53064 + }, + { + "epoch": 3.6822566095343836, + "grad_norm": 1.263216495513916, + "learning_rate": 3.291342307375345e-06, + "loss": 0.2083, + "step": 53065 + }, + { + "epoch": 3.68232600097148, + "grad_norm": 1.3440686464309692, + "learning_rate": 3.2899128171173245e-06, + "loss": 0.2256, + "step": 53066 + }, + { + "epoch": 3.682395392408577, + "grad_norm": 1.029899001121521, + "learning_rate": 3.288483632161832e-06, + "loss": 0.2536, + "step": 53067 + }, + { + "epoch": 3.682464783845673, + "grad_norm": 1.3241157531738281, + "learning_rate": 3.2870547525133634e-06, + "loss": 0.276, + "step": 53068 + }, + { + "epoch": 3.68253417528277, + "grad_norm": 1.0453498363494873, + "learning_rate": 3.2856261781764486e-06, + "loss": 0.2338, + "step": 53069 + }, + { + "epoch": 3.682603566719867, + "grad_norm": 1.2837598323822021, + "learning_rate": 3.284197909155595e-06, + "loss": 0.3239, + "step": 53070 + }, + { + "epoch": 3.6826729581569633, + "grad_norm": 0.5556972026824951, + "learning_rate": 3.2827699454552996e-06, + "loss": 0.2682, + "step": 53071 + }, + { + "epoch": 3.68274234959406, + "grad_norm": 1.1179684400558472, + "learning_rate": 3.28134228708008e-06, + "loss": 0.1591, + "step": 53072 + }, + { + "epoch": 3.6828117410311565, + "grad_norm": 0.7433308362960815, + "learning_rate": 3.279914934034434e-06, + "loss": 0.2113, + "step": 53073 + }, + { + "epoch": 3.6828811324682533, + "grad_norm": 0.6965965628623962, + "learning_rate": 3.2784878863229007e-06, + "loss": 0.2229, + "step": 53074 + }, + { + "epoch": 3.68295052390535, + "grad_norm": 0.9469629526138306, + "learning_rate": 3.277061143949944e-06, + "loss": 0.2859, + "step": 53075 + }, + { + "epoch": 3.683019915342447, + "grad_norm": 0.6553966403007507, + "learning_rate": 3.275634706920072e-06, + "loss": 0.1906, + "step": 53076 + }, + { + "epoch": 3.6830893067795434, + "grad_norm": 0.49294766783714294, + "learning_rate": 3.2742085752378136e-06, + "loss": 0.1878, + "step": 53077 + }, + { + "epoch": 3.68315869821664, + "grad_norm": 1.1821638345718384, + "learning_rate": 3.2727827489076436e-06, + "loss": 0.2573, + "step": 53078 + }, + { + "epoch": 3.6832280896537366, + "grad_norm": 1.0262465476989746, + "learning_rate": 3.271357227934091e-06, + "loss": 0.2405, + "step": 53079 + }, + { + "epoch": 3.6832974810908334, + "grad_norm": 1.2265993356704712, + "learning_rate": 3.2699320123216306e-06, + "loss": 0.2866, + "step": 53080 + }, + { + "epoch": 3.6833668725279303, + "grad_norm": 0.8288832306861877, + "learning_rate": 3.2685071020747806e-06, + "loss": 0.2894, + "step": 53081 + }, + { + "epoch": 3.6834362639650267, + "grad_norm": 1.7340104579925537, + "learning_rate": 3.267082497198026e-06, + "loss": 0.3221, + "step": 53082 + }, + { + "epoch": 3.6835056554021235, + "grad_norm": 0.9040590524673462, + "learning_rate": 3.2656581976958645e-06, + "loss": 0.2691, + "step": 53083 + }, + { + "epoch": 3.68357504683922, + "grad_norm": 0.6544065475463867, + "learning_rate": 3.2642342035728023e-06, + "loss": 0.1557, + "step": 53084 + }, + { + "epoch": 3.6836444382763167, + "grad_norm": 0.9910534024238586, + "learning_rate": 3.2628105148333256e-06, + "loss": 0.2234, + "step": 53085 + }, + { + "epoch": 3.6837138297134135, + "grad_norm": 0.8477397561073303, + "learning_rate": 3.2613871314819412e-06, + "loss": 0.3053, + "step": 53086 + }, + { + "epoch": 3.68378322115051, + "grad_norm": 0.8829293251037598, + "learning_rate": 3.259964053523135e-06, + "loss": 0.1893, + "step": 53087 + }, + { + "epoch": 3.6838526125876068, + "grad_norm": 0.7194540500640869, + "learning_rate": 3.2585412809613913e-06, + "loss": 0.222, + "step": 53088 + }, + { + "epoch": 3.683922004024703, + "grad_norm": 0.7246642112731934, + "learning_rate": 3.2571188138012077e-06, + "loss": 0.202, + "step": 53089 + }, + { + "epoch": 3.6839913954618, + "grad_norm": 0.6853764653205872, + "learning_rate": 3.255696652047069e-06, + "loss": 0.1912, + "step": 53090 + }, + { + "epoch": 3.684060786898897, + "grad_norm": 0.9740429520606995, + "learning_rate": 3.2542747957034935e-06, + "loss": 0.2911, + "step": 53091 + }, + { + "epoch": 3.684130178335993, + "grad_norm": 0.8467182517051697, + "learning_rate": 3.2528532447749225e-06, + "loss": 0.2097, + "step": 53092 + }, + { + "epoch": 3.68419956977309, + "grad_norm": 1.2254588603973389, + "learning_rate": 3.251431999265886e-06, + "loss": 0.2087, + "step": 53093 + }, + { + "epoch": 3.6842689612101864, + "grad_norm": 0.5963166356086731, + "learning_rate": 3.2500110591808574e-06, + "loss": 0.2271, + "step": 53094 + }, + { + "epoch": 3.6843383526472833, + "grad_norm": 0.7637585997581482, + "learning_rate": 3.2485904245243004e-06, + "loss": 0.2296, + "step": 53095 + }, + { + "epoch": 3.68440774408438, + "grad_norm": 0.9670653939247131, + "learning_rate": 3.247170095300733e-06, + "loss": 0.2184, + "step": 53096 + }, + { + "epoch": 3.6844771355214765, + "grad_norm": 0.9317824244499207, + "learning_rate": 3.2457500715146193e-06, + "loss": 0.2553, + "step": 53097 + }, + { + "epoch": 3.6845465269585733, + "grad_norm": 0.7503345012664795, + "learning_rate": 3.2443303531704437e-06, + "loss": 0.1664, + "step": 53098 + }, + { + "epoch": 3.6846159183956697, + "grad_norm": 0.9783247709274292, + "learning_rate": 3.242910940272692e-06, + "loss": 0.3701, + "step": 53099 + }, + { + "epoch": 3.6846853098327665, + "grad_norm": 0.7322473526000977, + "learning_rate": 3.2414918328258383e-06, + "loss": 0.2472, + "step": 53100 + }, + { + "epoch": 3.6847547012698634, + "grad_norm": 1.4327553510665894, + "learning_rate": 3.2400730308343896e-06, + "loss": 0.2944, + "step": 53101 + }, + { + "epoch": 3.68482409270696, + "grad_norm": 1.0110132694244385, + "learning_rate": 3.2386545343027872e-06, + "loss": 0.2905, + "step": 53102 + }, + { + "epoch": 3.6848934841440566, + "grad_norm": 0.7260684370994568, + "learning_rate": 3.237236343235528e-06, + "loss": 0.154, + "step": 53103 + }, + { + "epoch": 3.6849628755811534, + "grad_norm": 0.841407835483551, + "learning_rate": 3.235818457637085e-06, + "loss": 0.1868, + "step": 53104 + }, + { + "epoch": 3.68503226701825, + "grad_norm": 1.0194618701934814, + "learning_rate": 3.234400877511945e-06, + "loss": 0.3064, + "step": 53105 + }, + { + "epoch": 3.6851016584553467, + "grad_norm": 1.4829903841018677, + "learning_rate": 3.2329836028645697e-06, + "loss": 0.3396, + "step": 53106 + }, + { + "epoch": 3.6851710498924435, + "grad_norm": 0.7730224132537842, + "learning_rate": 3.231566633699434e-06, + "loss": 0.2744, + "step": 53107 + }, + { + "epoch": 3.68524044132954, + "grad_norm": 1.858962059020996, + "learning_rate": 3.230149970021035e-06, + "loss": 0.4384, + "step": 53108 + }, + { + "epoch": 3.6853098327666367, + "grad_norm": 0.9625529050827026, + "learning_rate": 3.2287336118338008e-06, + "loss": 0.2591, + "step": 53109 + }, + { + "epoch": 3.685379224203733, + "grad_norm": 0.8134593963623047, + "learning_rate": 3.22731755914224e-06, + "loss": 0.275, + "step": 53110 + }, + { + "epoch": 3.68544861564083, + "grad_norm": 1.1238635778427124, + "learning_rate": 3.2259018119508044e-06, + "loss": 0.2663, + "step": 53111 + }, + { + "epoch": 3.6855180070779268, + "grad_norm": 1.050626516342163, + "learning_rate": 3.224486370263968e-06, + "loss": 0.3388, + "step": 53112 + }, + { + "epoch": 3.685587398515023, + "grad_norm": 1.064745306968689, + "learning_rate": 3.2230712340862057e-06, + "loss": 0.3159, + "step": 53113 + }, + { + "epoch": 3.68565678995212, + "grad_norm": 0.7135249972343445, + "learning_rate": 3.221656403421969e-06, + "loss": 0.2122, + "step": 53114 + }, + { + "epoch": 3.6857261813892164, + "grad_norm": 0.6907045245170593, + "learning_rate": 3.220241878275754e-06, + "loss": 0.181, + "step": 53115 + }, + { + "epoch": 3.685795572826313, + "grad_norm": 0.8561713695526123, + "learning_rate": 3.2188276586519906e-06, + "loss": 0.1814, + "step": 53116 + }, + { + "epoch": 3.68586496426341, + "grad_norm": 0.8497563600540161, + "learning_rate": 3.2174137445551642e-06, + "loss": 0.2682, + "step": 53117 + }, + { + "epoch": 3.6859343557005064, + "grad_norm": 1.4793734550476074, + "learning_rate": 3.2160001359897275e-06, + "loss": 0.3058, + "step": 53118 + }, + { + "epoch": 3.6860037471376033, + "grad_norm": 0.646090030670166, + "learning_rate": 3.2145868329601535e-06, + "loss": 0.1616, + "step": 53119 + }, + { + "epoch": 3.6860731385746996, + "grad_norm": 0.615540087223053, + "learning_rate": 3.2131738354709063e-06, + "loss": 0.1411, + "step": 53120 + }, + { + "epoch": 3.6861425300117965, + "grad_norm": 0.7067459225654602, + "learning_rate": 3.211761143526426e-06, + "loss": 0.2348, + "step": 53121 + }, + { + "epoch": 3.6862119214488933, + "grad_norm": 1.4252640008926392, + "learning_rate": 3.2103487571312098e-06, + "loss": 0.3212, + "step": 53122 + }, + { + "epoch": 3.6862813128859897, + "grad_norm": 0.9739653468132019, + "learning_rate": 3.2089366762896755e-06, + "loss": 0.2667, + "step": 53123 + }, + { + "epoch": 3.6863507043230865, + "grad_norm": 0.649370014667511, + "learning_rate": 3.207524901006298e-06, + "loss": 0.1706, + "step": 53124 + }, + { + "epoch": 3.686420095760183, + "grad_norm": 0.9913697242736816, + "learning_rate": 3.20611343128554e-06, + "loss": 0.3307, + "step": 53125 + }, + { + "epoch": 3.6864894871972798, + "grad_norm": 0.7340237498283386, + "learning_rate": 3.2047022671318537e-06, + "loss": 0.2433, + "step": 53126 + }, + { + "epoch": 3.6865588786343766, + "grad_norm": 1.0974040031433105, + "learning_rate": 3.2032914085496914e-06, + "loss": 0.3302, + "step": 53127 + }, + { + "epoch": 3.6866282700714734, + "grad_norm": 0.9431537389755249, + "learning_rate": 3.201880855543504e-06, + "loss": 0.2297, + "step": 53128 + }, + { + "epoch": 3.68669766150857, + "grad_norm": 0.798783004283905, + "learning_rate": 3.200470608117756e-06, + "loss": 0.2845, + "step": 53129 + }, + { + "epoch": 3.6867670529456666, + "grad_norm": 1.1247754096984863, + "learning_rate": 3.1990606662768763e-06, + "loss": 0.269, + "step": 53130 + }, + { + "epoch": 3.686836444382763, + "grad_norm": 1.1667025089263916, + "learning_rate": 3.1976510300253393e-06, + "loss": 0.285, + "step": 53131 + }, + { + "epoch": 3.68690583581986, + "grad_norm": 1.0819600820541382, + "learning_rate": 3.1962416993675858e-06, + "loss": 0.219, + "step": 53132 + }, + { + "epoch": 3.6869752272569567, + "grad_norm": 0.9738649129867554, + "learning_rate": 3.194832674308068e-06, + "loss": 0.2143, + "step": 53133 + }, + { + "epoch": 3.687044618694053, + "grad_norm": 1.1643320322036743, + "learning_rate": 3.1934239548512267e-06, + "loss": 0.3785, + "step": 53134 + }, + { + "epoch": 3.68711401013115, + "grad_norm": 0.6966156363487244, + "learning_rate": 3.192015541001514e-06, + "loss": 0.2021, + "step": 53135 + }, + { + "epoch": 3.6871834015682463, + "grad_norm": 0.8496273159980774, + "learning_rate": 3.190607432763393e-06, + "loss": 0.1728, + "step": 53136 + }, + { + "epoch": 3.687252793005343, + "grad_norm": 0.7575773596763611, + "learning_rate": 3.1891996301412707e-06, + "loss": 0.2105, + "step": 53137 + }, + { + "epoch": 3.68732218444244, + "grad_norm": 0.8518279194831848, + "learning_rate": 3.1877921331396223e-06, + "loss": 0.3946, + "step": 53138 + }, + { + "epoch": 3.6873915758795364, + "grad_norm": 0.8212687969207764, + "learning_rate": 3.186384941762888e-06, + "loss": 0.2365, + "step": 53139 + }, + { + "epoch": 3.687460967316633, + "grad_norm": 0.9805021286010742, + "learning_rate": 3.1849780560154863e-06, + "loss": 0.2802, + "step": 53140 + }, + { + "epoch": 3.6875303587537296, + "grad_norm": 0.6013533473014832, + "learning_rate": 3.183571475901892e-06, + "loss": 0.1827, + "step": 53141 + }, + { + "epoch": 3.6875997501908264, + "grad_norm": 0.7652095556259155, + "learning_rate": 3.1821652014265235e-06, + "loss": 0.2014, + "step": 53142 + }, + { + "epoch": 3.6876691416279233, + "grad_norm": 1.1694000959396362, + "learning_rate": 3.180759232593833e-06, + "loss": 0.3904, + "step": 53143 + }, + { + "epoch": 3.6877385330650196, + "grad_norm": 0.9899067282676697, + "learning_rate": 3.17935356940825e-06, + "loss": 0.3404, + "step": 53144 + }, + { + "epoch": 3.6878079245021165, + "grad_norm": 1.377396821975708, + "learning_rate": 3.1779482118742042e-06, + "loss": 0.1684, + "step": 53145 + }, + { + "epoch": 3.687877315939213, + "grad_norm": 0.69785076379776, + "learning_rate": 3.176543159996148e-06, + "loss": 0.1221, + "step": 53146 + }, + { + "epoch": 3.6879467073763097, + "grad_norm": 0.7980772256851196, + "learning_rate": 3.175138413778522e-06, + "loss": 0.1657, + "step": 53147 + }, + { + "epoch": 3.6880160988134065, + "grad_norm": 0.7661815881729126, + "learning_rate": 3.1737339732257344e-06, + "loss": 0.2067, + "step": 53148 + }, + { + "epoch": 3.6880854902505034, + "grad_norm": 0.5803782343864441, + "learning_rate": 3.172329838342236e-06, + "loss": 0.1783, + "step": 53149 + }, + { + "epoch": 3.6881548816875998, + "grad_norm": 2.006718873977661, + "learning_rate": 3.17092600913248e-06, + "loss": 0.2273, + "step": 53150 + }, + { + "epoch": 3.688224273124696, + "grad_norm": 0.7591059803962708, + "learning_rate": 3.169522485600862e-06, + "loss": 0.3463, + "step": 53151 + }, + { + "epoch": 3.688293664561793, + "grad_norm": 0.9552137851715088, + "learning_rate": 3.1681192677518234e-06, + "loss": 0.2579, + "step": 53152 + }, + { + "epoch": 3.68836305599889, + "grad_norm": 0.6923174858093262, + "learning_rate": 3.166716355589794e-06, + "loss": 0.1657, + "step": 53153 + }, + { + "epoch": 3.6884324474359866, + "grad_norm": 1.0173630714416504, + "learning_rate": 3.165313749119214e-06, + "loss": 0.2707, + "step": 53154 + }, + { + "epoch": 3.688501838873083, + "grad_norm": 0.7280455827713013, + "learning_rate": 3.163911448344503e-06, + "loss": 0.1932, + "step": 53155 + }, + { + "epoch": 3.68857123031018, + "grad_norm": 0.6330099701881409, + "learning_rate": 3.1625094532700795e-06, + "loss": 0.1688, + "step": 53156 + }, + { + "epoch": 3.6886406217472762, + "grad_norm": 0.8353087306022644, + "learning_rate": 3.161107763900395e-06, + "loss": 0.2377, + "step": 53157 + }, + { + "epoch": 3.688710013184373, + "grad_norm": 0.6926595568656921, + "learning_rate": 3.1597063802398576e-06, + "loss": 0.1811, + "step": 53158 + }, + { + "epoch": 3.68877940462147, + "grad_norm": 0.8460388779640198, + "learning_rate": 3.158305302292874e-06, + "loss": 0.2477, + "step": 53159 + }, + { + "epoch": 3.6888487960585663, + "grad_norm": 0.9872559905052185, + "learning_rate": 3.156904530063909e-06, + "loss": 0.2551, + "step": 53160 + }, + { + "epoch": 3.688918187495663, + "grad_norm": 1.110102891921997, + "learning_rate": 3.155504063557346e-06, + "loss": 0.31, + "step": 53161 + }, + { + "epoch": 3.6889875789327595, + "grad_norm": 0.9430922865867615, + "learning_rate": 3.1541039027776165e-06, + "loss": 0.3448, + "step": 53162 + }, + { + "epoch": 3.6890569703698564, + "grad_norm": 1.0975909233093262, + "learning_rate": 3.152704047729149e-06, + "loss": 0.3145, + "step": 53163 + }, + { + "epoch": 3.689126361806953, + "grad_norm": 0.7537703514099121, + "learning_rate": 3.151304498416363e-06, + "loss": 0.1969, + "step": 53164 + }, + { + "epoch": 3.6891957532440496, + "grad_norm": 0.8163285851478577, + "learning_rate": 3.149905254843666e-06, + "loss": 0.208, + "step": 53165 + }, + { + "epoch": 3.6892651446811464, + "grad_norm": 1.0905265808105469, + "learning_rate": 3.148506317015487e-06, + "loss": 0.203, + "step": 53166 + }, + { + "epoch": 3.689334536118243, + "grad_norm": 0.9308817386627197, + "learning_rate": 3.1471076849362236e-06, + "loss": 0.3188, + "step": 53167 + }, + { + "epoch": 3.6894039275553396, + "grad_norm": 1.055237889289856, + "learning_rate": 3.145709358610316e-06, + "loss": 0.2808, + "step": 53168 + }, + { + "epoch": 3.6894733189924365, + "grad_norm": 0.893912136554718, + "learning_rate": 3.1443113380421606e-06, + "loss": 0.2273, + "step": 53169 + }, + { + "epoch": 3.689542710429533, + "grad_norm": 1.0194416046142578, + "learning_rate": 3.142913623236188e-06, + "loss": 0.2731, + "step": 53170 + }, + { + "epoch": 3.6896121018666297, + "grad_norm": 0.7009082436561584, + "learning_rate": 3.1415162141967933e-06, + "loss": 0.2155, + "step": 53171 + }, + { + "epoch": 3.689681493303726, + "grad_norm": 2.058814287185669, + "learning_rate": 3.1401191109283857e-06, + "loss": 0.209, + "step": 53172 + }, + { + "epoch": 3.689750884740823, + "grad_norm": 0.7113879323005676, + "learning_rate": 3.1387223134353936e-06, + "loss": 0.132, + "step": 53173 + }, + { + "epoch": 3.6898202761779197, + "grad_norm": 1.034628987312317, + "learning_rate": 3.1373258217222146e-06, + "loss": 0.4488, + "step": 53174 + }, + { + "epoch": 3.6898896676150166, + "grad_norm": 1.5402847528457642, + "learning_rate": 3.1359296357932556e-06, + "loss": 0.4316, + "step": 53175 + }, + { + "epoch": 3.689959059052113, + "grad_norm": 1.3656134605407715, + "learning_rate": 3.1345337556529353e-06, + "loss": 0.4386, + "step": 53176 + }, + { + "epoch": 3.6900284504892094, + "grad_norm": 0.8375118374824524, + "learning_rate": 3.1331381813056616e-06, + "loss": 0.3279, + "step": 53177 + }, + { + "epoch": 3.690097841926306, + "grad_norm": 0.7648465037345886, + "learning_rate": 3.1317429127558196e-06, + "loss": 0.2394, + "step": 53178 + }, + { + "epoch": 3.690167233363403, + "grad_norm": 1.2133705615997314, + "learning_rate": 3.1303479500078283e-06, + "loss": 0.2741, + "step": 53179 + }, + { + "epoch": 3.6902366248005, + "grad_norm": 1.0260142087936401, + "learning_rate": 3.1289532930660837e-06, + "loss": 0.2606, + "step": 53180 + }, + { + "epoch": 3.6903060162375962, + "grad_norm": 0.9680162072181702, + "learning_rate": 3.127558941934994e-06, + "loss": 0.1847, + "step": 53181 + }, + { + "epoch": 3.690375407674693, + "grad_norm": 0.8592121601104736, + "learning_rate": 3.126164896618966e-06, + "loss": 0.2151, + "step": 53182 + }, + { + "epoch": 3.6904447991117895, + "grad_norm": 0.7755826115608215, + "learning_rate": 3.124771157122397e-06, + "loss": 0.1984, + "step": 53183 + }, + { + "epoch": 3.6905141905488863, + "grad_norm": 1.2661030292510986, + "learning_rate": 3.1233777234496942e-06, + "loss": 0.2212, + "step": 53184 + }, + { + "epoch": 3.690583581985983, + "grad_norm": 0.7708819508552551, + "learning_rate": 3.1219845956052428e-06, + "loss": 0.1589, + "step": 53185 + }, + { + "epoch": 3.6906529734230795, + "grad_norm": 0.6757275462150574, + "learning_rate": 3.1205917735934397e-06, + "loss": 0.1406, + "step": 53186 + }, + { + "epoch": 3.6907223648601764, + "grad_norm": 1.095456600189209, + "learning_rate": 3.1191992574186925e-06, + "loss": 0.212, + "step": 53187 + }, + { + "epoch": 3.6907917562972727, + "grad_norm": 0.7414220571517944, + "learning_rate": 3.117807047085397e-06, + "loss": 0.1284, + "step": 53188 + }, + { + "epoch": 3.6908611477343696, + "grad_norm": 1.0577294826507568, + "learning_rate": 3.116415142597939e-06, + "loss": 0.2228, + "step": 53189 + }, + { + "epoch": 3.6909305391714664, + "grad_norm": 0.8824016451835632, + "learning_rate": 3.1150235439607156e-06, + "loss": 0.22, + "step": 53190 + }, + { + "epoch": 3.690999930608563, + "grad_norm": 1.2542601823806763, + "learning_rate": 3.1136322511781335e-06, + "loss": 0.3491, + "step": 53191 + }, + { + "epoch": 3.6910693220456596, + "grad_norm": 0.8875364661216736, + "learning_rate": 3.1122412642545675e-06, + "loss": 0.3309, + "step": 53192 + }, + { + "epoch": 3.691138713482756, + "grad_norm": 1.51616632938385, + "learning_rate": 3.1108505831944023e-06, + "loss": 0.26, + "step": 53193 + }, + { + "epoch": 3.691208104919853, + "grad_norm": 0.6366727352142334, + "learning_rate": 3.1094602080020576e-06, + "loss": 0.2464, + "step": 53194 + }, + { + "epoch": 3.6912774963569497, + "grad_norm": 0.8808900713920593, + "learning_rate": 3.1080701386819067e-06, + "loss": 0.4349, + "step": 53195 + }, + { + "epoch": 3.691346887794046, + "grad_norm": 0.8517668843269348, + "learning_rate": 3.1066803752383245e-06, + "loss": 0.2639, + "step": 53196 + }, + { + "epoch": 3.691416279231143, + "grad_norm": 0.8960620760917664, + "learning_rate": 3.1052909176757185e-06, + "loss": 0.3678, + "step": 53197 + }, + { + "epoch": 3.6914856706682393, + "grad_norm": 0.8934839963912964, + "learning_rate": 3.1039017659984625e-06, + "loss": 0.2385, + "step": 53198 + }, + { + "epoch": 3.691555062105336, + "grad_norm": 0.9726189374923706, + "learning_rate": 3.1025129202109538e-06, + "loss": 0.3164, + "step": 53199 + }, + { + "epoch": 3.691624453542433, + "grad_norm": 1.2813199758529663, + "learning_rate": 3.101124380317566e-06, + "loss": 0.4048, + "step": 53200 + }, + { + "epoch": 3.69169384497953, + "grad_norm": 0.944027304649353, + "learning_rate": 3.099736146322685e-06, + "loss": 0.3585, + "step": 53201 + }, + { + "epoch": 3.691763236416626, + "grad_norm": 0.9717517495155334, + "learning_rate": 3.0983482182306846e-06, + "loss": 0.3617, + "step": 53202 + }, + { + "epoch": 3.691832627853723, + "grad_norm": 0.8927412629127502, + "learning_rate": 3.0969605960459612e-06, + "loss": 0.27, + "step": 53203 + }, + { + "epoch": 3.6919020192908194, + "grad_norm": 1.138867974281311, + "learning_rate": 3.095573279772901e-06, + "loss": 0.2288, + "step": 53204 + }, + { + "epoch": 3.6919714107279162, + "grad_norm": 1.1153371334075928, + "learning_rate": 3.0941862694158775e-06, + "loss": 0.2393, + "step": 53205 + }, + { + "epoch": 3.692040802165013, + "grad_norm": 0.7357097864151001, + "learning_rate": 3.092799564979243e-06, + "loss": 0.211, + "step": 53206 + }, + { + "epoch": 3.6921101936021095, + "grad_norm": 1.112199068069458, + "learning_rate": 3.091413166467405e-06, + "loss": 0.2991, + "step": 53207 + }, + { + "epoch": 3.6921795850392063, + "grad_norm": 0.8441143035888672, + "learning_rate": 3.090027073884738e-06, + "loss": 0.2428, + "step": 53208 + }, + { + "epoch": 3.6922489764763027, + "grad_norm": 0.9374582767486572, + "learning_rate": 3.088641287235605e-06, + "loss": 0.2259, + "step": 53209 + }, + { + "epoch": 3.6923183679133995, + "grad_norm": 1.3691152334213257, + "learning_rate": 3.087255806524392e-06, + "loss": 0.1898, + "step": 53210 + }, + { + "epoch": 3.6923877593504963, + "grad_norm": 0.6924369931221008, + "learning_rate": 3.085870631755472e-06, + "loss": 0.2144, + "step": 53211 + }, + { + "epoch": 3.6924571507875927, + "grad_norm": 0.5112026929855347, + "learning_rate": 3.0844857629332093e-06, + "loss": 0.142, + "step": 53212 + }, + { + "epoch": 3.6925265422246896, + "grad_norm": 0.7685482501983643, + "learning_rate": 3.0831012000619887e-06, + "loss": 0.2309, + "step": 53213 + }, + { + "epoch": 3.692595933661786, + "grad_norm": 1.8273502588272095, + "learning_rate": 3.081716943146162e-06, + "loss": 0.2196, + "step": 53214 + }, + { + "epoch": 3.692665325098883, + "grad_norm": 0.8023099303245544, + "learning_rate": 3.080332992190116e-06, + "loss": 0.2876, + "step": 53215 + }, + { + "epoch": 3.6927347165359796, + "grad_norm": 0.9941473603248596, + "learning_rate": 3.0789493471982013e-06, + "loss": 0.2008, + "step": 53216 + }, + { + "epoch": 3.692804107973076, + "grad_norm": 0.9117312431335449, + "learning_rate": 3.077566008174815e-06, + "loss": 0.2661, + "step": 53217 + }, + { + "epoch": 3.692873499410173, + "grad_norm": 0.751123309135437, + "learning_rate": 3.0761829751242975e-06, + "loss": 0.2405, + "step": 53218 + }, + { + "epoch": 3.6929428908472692, + "grad_norm": 0.9681656956672668, + "learning_rate": 3.0748002480510463e-06, + "loss": 0.2481, + "step": 53219 + }, + { + "epoch": 3.693012282284366, + "grad_norm": 0.78863126039505, + "learning_rate": 3.0734178269593793e-06, + "loss": 0.1858, + "step": 53220 + }, + { + "epoch": 3.693081673721463, + "grad_norm": 0.9379422068595886, + "learning_rate": 3.0720357118536936e-06, + "loss": 0.1913, + "step": 53221 + }, + { + "epoch": 3.6931510651585593, + "grad_norm": 0.8558135628700256, + "learning_rate": 3.070653902738352e-06, + "loss": 0.2828, + "step": 53222 + }, + { + "epoch": 3.693220456595656, + "grad_norm": 0.8354923129081726, + "learning_rate": 3.0692723996177063e-06, + "loss": 0.1873, + "step": 53223 + }, + { + "epoch": 3.6932898480327525, + "grad_norm": 0.676770031452179, + "learning_rate": 3.0678912024961203e-06, + "loss": 0.2409, + "step": 53224 + }, + { + "epoch": 3.6933592394698493, + "grad_norm": 0.7647985816001892, + "learning_rate": 3.066510311377957e-06, + "loss": 0.2283, + "step": 53225 + }, + { + "epoch": 3.693428630906946, + "grad_norm": 0.71435546875, + "learning_rate": 3.065129726267579e-06, + "loss": 0.2479, + "step": 53226 + }, + { + "epoch": 3.693498022344043, + "grad_norm": 0.8258420825004578, + "learning_rate": 3.0637494471693286e-06, + "loss": 0.2511, + "step": 53227 + }, + { + "epoch": 3.6935674137811394, + "grad_norm": 0.9397872090339661, + "learning_rate": 3.0623694740875787e-06, + "loss": 0.2332, + "step": 53228 + }, + { + "epoch": 3.6936368052182362, + "grad_norm": 0.9195442795753479, + "learning_rate": 3.060989807026682e-06, + "loss": 0.3365, + "step": 53229 + }, + { + "epoch": 3.6937061966553326, + "grad_norm": 1.0143415927886963, + "learning_rate": 3.0596104459910124e-06, + "loss": 0.2889, + "step": 53230 + }, + { + "epoch": 3.6937755880924295, + "grad_norm": 0.8503257036209106, + "learning_rate": 3.058231390984889e-06, + "loss": 0.2884, + "step": 53231 + }, + { + "epoch": 3.6938449795295263, + "grad_norm": 1.1379636526107788, + "learning_rate": 3.0568526420126753e-06, + "loss": 0.2392, + "step": 53232 + }, + { + "epoch": 3.6939143709666227, + "grad_norm": 1.2262142896652222, + "learning_rate": 3.0554741990787337e-06, + "loss": 0.3182, + "step": 53233 + }, + { + "epoch": 3.6939837624037195, + "grad_norm": 0.8988277316093445, + "learning_rate": 3.054096062187428e-06, + "loss": 0.1778, + "step": 53234 + }, + { + "epoch": 3.694053153840816, + "grad_norm": 0.9984171986579895, + "learning_rate": 3.052718231343077e-06, + "loss": 0.2648, + "step": 53235 + }, + { + "epoch": 3.6941225452779127, + "grad_norm": 1.1448087692260742, + "learning_rate": 3.0513407065500433e-06, + "loss": 0.3662, + "step": 53236 + }, + { + "epoch": 3.6941919367150096, + "grad_norm": 0.8865941166877747, + "learning_rate": 3.0499634878126902e-06, + "loss": 0.2619, + "step": 53237 + }, + { + "epoch": 3.694261328152106, + "grad_norm": 1.1753952503204346, + "learning_rate": 3.0485865751353483e-06, + "loss": 0.3706, + "step": 53238 + }, + { + "epoch": 3.694330719589203, + "grad_norm": 0.9928656816482544, + "learning_rate": 3.0472099685223686e-06, + "loss": 0.3368, + "step": 53239 + }, + { + "epoch": 3.694400111026299, + "grad_norm": 0.7863038778305054, + "learning_rate": 3.045833667978104e-06, + "loss": 0.3315, + "step": 53240 + }, + { + "epoch": 3.694469502463396, + "grad_norm": 0.9198148250579834, + "learning_rate": 3.0444576735068953e-06, + "loss": 0.2484, + "step": 53241 + }, + { + "epoch": 3.694538893900493, + "grad_norm": 0.818840503692627, + "learning_rate": 3.043081985113072e-06, + "loss": 0.2027, + "step": 53242 + }, + { + "epoch": 3.6946082853375892, + "grad_norm": 0.8228681683540344, + "learning_rate": 3.0417066028009976e-06, + "loss": 0.2409, + "step": 53243 + }, + { + "epoch": 3.694677676774686, + "grad_norm": 1.0744073390960693, + "learning_rate": 3.040331526575002e-06, + "loss": 0.259, + "step": 53244 + }, + { + "epoch": 3.6947470682117824, + "grad_norm": 0.9382361769676208, + "learning_rate": 3.038956756439437e-06, + "loss": 0.2629, + "step": 53245 + }, + { + "epoch": 3.6948164596488793, + "grad_norm": 0.9915516972541809, + "learning_rate": 3.0375822923986443e-06, + "loss": 0.2936, + "step": 53246 + }, + { + "epoch": 3.694885851085976, + "grad_norm": 0.9480922818183899, + "learning_rate": 3.0362081344569416e-06, + "loss": 0.2376, + "step": 53247 + }, + { + "epoch": 3.6949552425230725, + "grad_norm": 1.2135883569717407, + "learning_rate": 3.0348342826186815e-06, + "loss": 0.4287, + "step": 53248 + }, + { + "epoch": 3.6950246339601693, + "grad_norm": 0.9044777154922485, + "learning_rate": 3.033460736888194e-06, + "loss": 0.2735, + "step": 53249 + }, + { + "epoch": 3.6950940253972657, + "grad_norm": 0.7802573442459106, + "learning_rate": 3.032087497269831e-06, + "loss": 0.2664, + "step": 53250 + }, + { + "epoch": 3.6951634168343626, + "grad_norm": 1.1485637426376343, + "learning_rate": 3.0307145637679112e-06, + "loss": 0.2363, + "step": 53251 + }, + { + "epoch": 3.6952328082714594, + "grad_norm": 1.0217881202697754, + "learning_rate": 3.029341936386776e-06, + "loss": 0.3698, + "step": 53252 + }, + { + "epoch": 3.6953021997085562, + "grad_norm": 0.7866340279579163, + "learning_rate": 3.027969615130766e-06, + "loss": 0.2289, + "step": 53253 + }, + { + "epoch": 3.6953715911456526, + "grad_norm": 0.7239991426467896, + "learning_rate": 3.0265976000041997e-06, + "loss": 0.1976, + "step": 53254 + }, + { + "epoch": 3.6954409825827494, + "grad_norm": 1.0832324028015137, + "learning_rate": 3.0252258910114073e-06, + "loss": 0.2865, + "step": 53255 + }, + { + "epoch": 3.695510374019846, + "grad_norm": 1.1002819538116455, + "learning_rate": 3.02385448815673e-06, + "loss": 0.2547, + "step": 53256 + }, + { + "epoch": 3.6955797654569427, + "grad_norm": 1.0423941612243652, + "learning_rate": 3.022483391444486e-06, + "loss": 0.2424, + "step": 53257 + }, + { + "epoch": 3.6956491568940395, + "grad_norm": 0.8143038749694824, + "learning_rate": 3.021112600879017e-06, + "loss": 0.235, + "step": 53258 + }, + { + "epoch": 3.695718548331136, + "grad_norm": 1.1193466186523438, + "learning_rate": 3.019742116464652e-06, + "loss": 0.3822, + "step": 53259 + }, + { + "epoch": 3.6957879397682327, + "grad_norm": 0.5924639701843262, + "learning_rate": 3.0183719382057107e-06, + "loss": 0.1495, + "step": 53260 + }, + { + "epoch": 3.695857331205329, + "grad_norm": 0.8030403256416321, + "learning_rate": 3.0170020661065e-06, + "loss": 0.2565, + "step": 53261 + }, + { + "epoch": 3.695926722642426, + "grad_norm": 0.9155058860778809, + "learning_rate": 3.0156325001713727e-06, + "loss": 0.1601, + "step": 53262 + }, + { + "epoch": 3.6959961140795228, + "grad_norm": 0.6605204343795776, + "learning_rate": 3.0142632404046354e-06, + "loss": 0.1741, + "step": 53263 + }, + { + "epoch": 3.696065505516619, + "grad_norm": 0.9002630114555359, + "learning_rate": 3.0128942868106192e-06, + "loss": 0.2385, + "step": 53264 + }, + { + "epoch": 3.696134896953716, + "grad_norm": 0.941719114780426, + "learning_rate": 3.011525639393653e-06, + "loss": 0.2493, + "step": 53265 + }, + { + "epoch": 3.6962042883908124, + "grad_norm": 1.2698489427566528, + "learning_rate": 3.010157298158034e-06, + "loss": 0.4079, + "step": 53266 + }, + { + "epoch": 3.696273679827909, + "grad_norm": 0.9291955828666687, + "learning_rate": 3.0087892631081026e-06, + "loss": 0.2883, + "step": 53267 + }, + { + "epoch": 3.696343071265006, + "grad_norm": 0.6829808354377747, + "learning_rate": 3.0074215342481673e-06, + "loss": 0.2509, + "step": 53268 + }, + { + "epoch": 3.6964124627021024, + "grad_norm": 0.793329119682312, + "learning_rate": 3.006054111582557e-06, + "loss": 0.225, + "step": 53269 + }, + { + "epoch": 3.6964818541391993, + "grad_norm": 0.7394638657569885, + "learning_rate": 3.004686995115569e-06, + "loss": 0.304, + "step": 53270 + }, + { + "epoch": 3.6965512455762957, + "grad_norm": 0.9071756601333618, + "learning_rate": 3.0033201848515324e-06, + "loss": 0.3265, + "step": 53271 + }, + { + "epoch": 3.6966206370133925, + "grad_norm": 0.6788227558135986, + "learning_rate": 3.0019536807947667e-06, + "loss": 0.2003, + "step": 53272 + }, + { + "epoch": 3.6966900284504893, + "grad_norm": 1.3771260976791382, + "learning_rate": 3.0005874829495682e-06, + "loss": 0.3058, + "step": 53273 + }, + { + "epoch": 3.6967594198875857, + "grad_norm": 0.816203236579895, + "learning_rate": 2.999221591320278e-06, + "loss": 0.2396, + "step": 53274 + }, + { + "epoch": 3.6968288113246826, + "grad_norm": 1.0783934593200684, + "learning_rate": 2.9978560059111814e-06, + "loss": 0.3256, + "step": 53275 + }, + { + "epoch": 3.696898202761779, + "grad_norm": 0.749467134475708, + "learning_rate": 2.996490726726597e-06, + "loss": 0.2587, + "step": 53276 + }, + { + "epoch": 3.6969675941988758, + "grad_norm": 0.9475373029708862, + "learning_rate": 2.995125753770833e-06, + "loss": 0.2361, + "step": 53277 + }, + { + "epoch": 3.6970369856359726, + "grad_norm": 0.9246235489845276, + "learning_rate": 2.993761087048208e-06, + "loss": 0.3502, + "step": 53278 + }, + { + "epoch": 3.6971063770730694, + "grad_norm": 1.0169081687927246, + "learning_rate": 2.9923967265630293e-06, + "loss": 0.3446, + "step": 53279 + }, + { + "epoch": 3.697175768510166, + "grad_norm": 1.1081290245056152, + "learning_rate": 2.9910326723195938e-06, + "loss": 0.2244, + "step": 53280 + }, + { + "epoch": 3.6972451599472627, + "grad_norm": 0.8215798735618591, + "learning_rate": 2.9896689243222197e-06, + "loss": 0.238, + "step": 53281 + }, + { + "epoch": 3.697314551384359, + "grad_norm": 0.4448131322860718, + "learning_rate": 2.9883054825751934e-06, + "loss": 0.1513, + "step": 53282 + }, + { + "epoch": 3.697383942821456, + "grad_norm": 1.07033109664917, + "learning_rate": 2.9869423470828328e-06, + "loss": 0.3208, + "step": 53283 + }, + { + "epoch": 3.6974533342585527, + "grad_norm": 1.1104382276535034, + "learning_rate": 2.985579517849446e-06, + "loss": 0.1751, + "step": 53284 + }, + { + "epoch": 3.697522725695649, + "grad_norm": 0.8424106240272522, + "learning_rate": 2.9842169948793296e-06, + "loss": 0.2277, + "step": 53285 + }, + { + "epoch": 3.697592117132746, + "grad_norm": 0.8317028284072876, + "learning_rate": 2.9828547781767804e-06, + "loss": 0.2474, + "step": 53286 + }, + { + "epoch": 3.6976615085698423, + "grad_norm": 0.915660560131073, + "learning_rate": 2.981492867746094e-06, + "loss": 0.2757, + "step": 53287 + }, + { + "epoch": 3.697730900006939, + "grad_norm": 0.9326115846633911, + "learning_rate": 2.9801312635916013e-06, + "loss": 0.2769, + "step": 53288 + }, + { + "epoch": 3.697800291444036, + "grad_norm": 0.7926054000854492, + "learning_rate": 2.9787699657175648e-06, + "loss": 0.2143, + "step": 53289 + }, + { + "epoch": 3.6978696828811324, + "grad_norm": 1.0023900270462036, + "learning_rate": 2.9774089741282927e-06, + "loss": 0.2583, + "step": 53290 + }, + { + "epoch": 3.697939074318229, + "grad_norm": 1.0074094533920288, + "learning_rate": 2.9760482888280817e-06, + "loss": 0.2546, + "step": 53291 + }, + { + "epoch": 3.6980084657553256, + "grad_norm": 0.7023229002952576, + "learning_rate": 2.974687909821239e-06, + "loss": 0.2275, + "step": 53292 + }, + { + "epoch": 3.6980778571924224, + "grad_norm": 1.143316388130188, + "learning_rate": 2.973327837112039e-06, + "loss": 0.2987, + "step": 53293 + }, + { + "epoch": 3.6981472486295193, + "grad_norm": 1.6460387706756592, + "learning_rate": 2.971968070704789e-06, + "loss": 0.2696, + "step": 53294 + }, + { + "epoch": 3.6982166400666157, + "grad_norm": 0.8269684910774231, + "learning_rate": 2.9706086106037868e-06, + "loss": 0.2109, + "step": 53295 + }, + { + "epoch": 3.6982860315037125, + "grad_norm": 0.7768124341964722, + "learning_rate": 2.9692494568133055e-06, + "loss": 0.231, + "step": 53296 + }, + { + "epoch": 3.698355422940809, + "grad_norm": 0.6660850048065186, + "learning_rate": 2.9678906093376533e-06, + "loss": 0.139, + "step": 53297 + }, + { + "epoch": 3.6984248143779057, + "grad_norm": 0.9967797994613647, + "learning_rate": 2.9665320681811047e-06, + "loss": 0.2259, + "step": 53298 + }, + { + "epoch": 3.6984942058150025, + "grad_norm": 0.7635016441345215, + "learning_rate": 2.9651738333479563e-06, + "loss": 0.2278, + "step": 53299 + }, + { + "epoch": 3.698563597252099, + "grad_norm": 1.1421360969543457, + "learning_rate": 2.9638159048425153e-06, + "loss": 0.3281, + "step": 53300 + }, + { + "epoch": 3.6986329886891958, + "grad_norm": 1.527700662612915, + "learning_rate": 2.9624582826690337e-06, + "loss": 0.2979, + "step": 53301 + }, + { + "epoch": 3.698702380126292, + "grad_norm": 1.1935091018676758, + "learning_rate": 2.961100966831809e-06, + "loss": 0.2282, + "step": 53302 + }, + { + "epoch": 3.698771771563389, + "grad_norm": 0.6639683842658997, + "learning_rate": 2.9597439573351484e-06, + "loss": 0.1333, + "step": 53303 + }, + { + "epoch": 3.698841163000486, + "grad_norm": 0.7232223749160767, + "learning_rate": 2.9583872541833035e-06, + "loss": 0.1513, + "step": 53304 + }, + { + "epoch": 3.6989105544375827, + "grad_norm": 0.996777355670929, + "learning_rate": 2.9570308573805607e-06, + "loss": 0.2758, + "step": 53305 + }, + { + "epoch": 3.698979945874679, + "grad_norm": 0.925794780254364, + "learning_rate": 2.955674766931227e-06, + "loss": 0.2246, + "step": 53306 + }, + { + "epoch": 3.699049337311776, + "grad_norm": 0.8611172437667847, + "learning_rate": 2.9543189828395658e-06, + "loss": 0.2886, + "step": 53307 + }, + { + "epoch": 3.6991187287488723, + "grad_norm": 0.8884036540985107, + "learning_rate": 2.9529635051098513e-06, + "loss": 0.3368, + "step": 53308 + }, + { + "epoch": 3.699188120185969, + "grad_norm": 0.9401483535766602, + "learning_rate": 2.9516083337463918e-06, + "loss": 0.282, + "step": 53309 + }, + { + "epoch": 3.699257511623066, + "grad_norm": 0.40443292260169983, + "learning_rate": 2.950253468753428e-06, + "loss": 0.1042, + "step": 53310 + }, + { + "epoch": 3.6993269030601623, + "grad_norm": 1.0036606788635254, + "learning_rate": 2.948898910135256e-06, + "loss": 0.2733, + "step": 53311 + }, + { + "epoch": 3.699396294497259, + "grad_norm": 0.7387688159942627, + "learning_rate": 2.947544657896151e-06, + "loss": 0.0847, + "step": 53312 + }, + { + "epoch": 3.6994656859343555, + "grad_norm": 0.8388752937316895, + "learning_rate": 2.9461907120403866e-06, + "loss": 0.2134, + "step": 53313 + }, + { + "epoch": 3.6995350773714524, + "grad_norm": 1.433327078819275, + "learning_rate": 2.9448370725722486e-06, + "loss": 0.2338, + "step": 53314 + }, + { + "epoch": 3.699604468808549, + "grad_norm": 0.7422366142272949, + "learning_rate": 2.9434837394959893e-06, + "loss": 0.1669, + "step": 53315 + }, + { + "epoch": 3.6996738602456456, + "grad_norm": 0.818678617477417, + "learning_rate": 2.9421307128158937e-06, + "loss": 0.2001, + "step": 53316 + }, + { + "epoch": 3.6997432516827424, + "grad_norm": 0.7722411751747131, + "learning_rate": 2.9407779925362367e-06, + "loss": 0.2361, + "step": 53317 + }, + { + "epoch": 3.699812643119839, + "grad_norm": 0.9175211191177368, + "learning_rate": 2.93942557866127e-06, + "loss": 0.2994, + "step": 53318 + }, + { + "epoch": 3.6998820345569357, + "grad_norm": 1.0607720613479614, + "learning_rate": 2.9380734711952795e-06, + "loss": 0.374, + "step": 53319 + }, + { + "epoch": 3.6999514259940325, + "grad_norm": 0.870692253112793, + "learning_rate": 2.9367216701425284e-06, + "loss": 0.2579, + "step": 53320 + }, + { + "epoch": 3.700020817431129, + "grad_norm": 0.7413992881774902, + "learning_rate": 2.9353701755072904e-06, + "loss": 0.2226, + "step": 53321 + }, + { + "epoch": 3.7000902088682257, + "grad_norm": 0.7158269882202148, + "learning_rate": 2.9340189872938405e-06, + "loss": 0.2636, + "step": 53322 + }, + { + "epoch": 3.700159600305322, + "grad_norm": 1.130905032157898, + "learning_rate": 2.9326681055064085e-06, + "loss": 0.3973, + "step": 53323 + }, + { + "epoch": 3.700228991742419, + "grad_norm": 0.8774191737174988, + "learning_rate": 2.9313175301492913e-06, + "loss": 0.1795, + "step": 53324 + }, + { + "epoch": 3.7002983831795158, + "grad_norm": 1.0632359981536865, + "learning_rate": 2.92996726122674e-06, + "loss": 0.2494, + "step": 53325 + }, + { + "epoch": 3.700367774616612, + "grad_norm": 0.7118811011314392, + "learning_rate": 2.9286172987430195e-06, + "loss": 0.1255, + "step": 53326 + }, + { + "epoch": 3.700437166053709, + "grad_norm": 1.1617552042007446, + "learning_rate": 2.9272676427023915e-06, + "loss": 0.2589, + "step": 53327 + }, + { + "epoch": 3.7005065574908054, + "grad_norm": 0.8050180673599243, + "learning_rate": 2.925918293109109e-06, + "loss": 0.2369, + "step": 53328 + }, + { + "epoch": 3.700575948927902, + "grad_norm": 0.6494535207748413, + "learning_rate": 2.924569249967457e-06, + "loss": 0.1599, + "step": 53329 + }, + { + "epoch": 3.700645340364999, + "grad_norm": 1.2593207359313965, + "learning_rate": 2.923220513281677e-06, + "loss": 0.3244, + "step": 53330 + }, + { + "epoch": 3.700714731802096, + "grad_norm": 0.9898051619529724, + "learning_rate": 2.9218720830560096e-06, + "loss": 0.3158, + "step": 53331 + }, + { + "epoch": 3.7007841232391923, + "grad_norm": 0.5177956819534302, + "learning_rate": 2.920523959294741e-06, + "loss": 0.1004, + "step": 53332 + }, + { + "epoch": 3.700853514676289, + "grad_norm": 0.8171534538269043, + "learning_rate": 2.9191761420021112e-06, + "loss": 0.3599, + "step": 53333 + }, + { + "epoch": 3.7009229061133855, + "grad_norm": 0.9204766154289246, + "learning_rate": 2.9178286311823734e-06, + "loss": 0.3402, + "step": 53334 + }, + { + "epoch": 3.7009922975504823, + "grad_norm": 0.7910600304603577, + "learning_rate": 2.916481426839801e-06, + "loss": 0.2321, + "step": 53335 + }, + { + "epoch": 3.701061688987579, + "grad_norm": 1.1426150798797607, + "learning_rate": 2.9151345289786247e-06, + "loss": 0.1826, + "step": 53336 + }, + { + "epoch": 3.7011310804246755, + "grad_norm": 1.0360774993896484, + "learning_rate": 2.9137879376031075e-06, + "loss": 0.3739, + "step": 53337 + }, + { + "epoch": 3.7012004718617724, + "grad_norm": 0.9775790572166443, + "learning_rate": 2.9124416527175014e-06, + "loss": 0.3603, + "step": 53338 + }, + { + "epoch": 3.7012698632988688, + "grad_norm": 1.0093090534210205, + "learning_rate": 2.9110956743260477e-06, + "loss": 0.2248, + "step": 53339 + }, + { + "epoch": 3.7013392547359656, + "grad_norm": 1.1004213094711304, + "learning_rate": 2.9097500024329983e-06, + "loss": 0.3589, + "step": 53340 + }, + { + "epoch": 3.7014086461730624, + "grad_norm": 0.786092221736908, + "learning_rate": 2.9084046370426055e-06, + "loss": 0.2787, + "step": 53341 + }, + { + "epoch": 3.701478037610159, + "grad_norm": 1.206119418144226, + "learning_rate": 2.9070595781591103e-06, + "loss": 0.3266, + "step": 53342 + }, + { + "epoch": 3.7015474290472556, + "grad_norm": 0.8627561926841736, + "learning_rate": 2.905714825786776e-06, + "loss": 0.19, + "step": 53343 + }, + { + "epoch": 3.701616820484352, + "grad_norm": 0.8412311673164368, + "learning_rate": 2.904370379929822e-06, + "loss": 0.2602, + "step": 53344 + }, + { + "epoch": 3.701686211921449, + "grad_norm": 0.8069278597831726, + "learning_rate": 2.90302624059251e-06, + "loss": 0.1857, + "step": 53345 + }, + { + "epoch": 3.7017556033585457, + "grad_norm": 0.9673998355865479, + "learning_rate": 2.9016824077790826e-06, + "loss": 0.2442, + "step": 53346 + }, + { + "epoch": 3.701824994795642, + "grad_norm": 1.0338871479034424, + "learning_rate": 2.9003388814937693e-06, + "loss": 0.2482, + "step": 53347 + }, + { + "epoch": 3.701894386232739, + "grad_norm": 1.0119843482971191, + "learning_rate": 2.898995661740822e-06, + "loss": 0.263, + "step": 53348 + }, + { + "epoch": 3.7019637776698353, + "grad_norm": 1.3716204166412354, + "learning_rate": 2.897652748524482e-06, + "loss": 0.4093, + "step": 53349 + }, + { + "epoch": 3.702033169106932, + "grad_norm": 1.347731351852417, + "learning_rate": 2.8963101418490014e-06, + "loss": 0.3351, + "step": 53350 + }, + { + "epoch": 3.702102560544029, + "grad_norm": 0.7625125050544739, + "learning_rate": 2.8949678417185875e-06, + "loss": 0.1682, + "step": 53351 + }, + { + "epoch": 3.7021719519811254, + "grad_norm": 1.2886215448379517, + "learning_rate": 2.8936258481374932e-06, + "loss": 0.2178, + "step": 53352 + }, + { + "epoch": 3.702241343418222, + "grad_norm": 1.2940009832382202, + "learning_rate": 2.892284161109948e-06, + "loss": 0.335, + "step": 53353 + }, + { + "epoch": 3.7023107348553186, + "grad_norm": 1.0184146165847778, + "learning_rate": 2.8909427806402043e-06, + "loss": 0.2483, + "step": 53354 + }, + { + "epoch": 3.7023801262924154, + "grad_norm": 0.7301997542381287, + "learning_rate": 2.889601706732492e-06, + "loss": 0.2352, + "step": 53355 + }, + { + "epoch": 3.7024495177295123, + "grad_norm": 1.1161201000213623, + "learning_rate": 2.88826093939103e-06, + "loss": 0.2222, + "step": 53356 + }, + { + "epoch": 3.702518909166609, + "grad_norm": 1.0319509506225586, + "learning_rate": 2.886920478620081e-06, + "loss": 0.2414, + "step": 53357 + }, + { + "epoch": 3.7025883006037055, + "grad_norm": 1.0159026384353638, + "learning_rate": 2.8855803244238423e-06, + "loss": 0.2844, + "step": 53358 + }, + { + "epoch": 3.7026576920408023, + "grad_norm": 1.0448933839797974, + "learning_rate": 2.8842404768065546e-06, + "loss": 0.2913, + "step": 53359 + }, + { + "epoch": 3.7027270834778987, + "grad_norm": 0.7109065651893616, + "learning_rate": 2.882900935772459e-06, + "loss": 0.2368, + "step": 53360 + }, + { + "epoch": 3.7027964749149955, + "grad_norm": 1.2333251237869263, + "learning_rate": 2.8815617013257743e-06, + "loss": 0.3713, + "step": 53361 + }, + { + "epoch": 3.7028658663520924, + "grad_norm": 0.7821398973464966, + "learning_rate": 2.880222773470731e-06, + "loss": 0.1778, + "step": 53362 + }, + { + "epoch": 3.7029352577891887, + "grad_norm": 0.8718967437744141, + "learning_rate": 2.8788841522115585e-06, + "loss": 0.3225, + "step": 53363 + }, + { + "epoch": 3.7030046492262856, + "grad_norm": 0.9734160900115967, + "learning_rate": 2.8775458375524977e-06, + "loss": 0.2463, + "step": 53364 + }, + { + "epoch": 3.703074040663382, + "grad_norm": 0.8798762559890747, + "learning_rate": 2.8762078294977348e-06, + "loss": 0.2357, + "step": 53365 + }, + { + "epoch": 3.703143432100479, + "grad_norm": 0.8796873688697815, + "learning_rate": 2.874870128051521e-06, + "loss": 0.2213, + "step": 53366 + }, + { + "epoch": 3.7032128235375756, + "grad_norm": 0.7928626537322998, + "learning_rate": 2.873532733218076e-06, + "loss": 0.1992, + "step": 53367 + }, + { + "epoch": 3.703282214974672, + "grad_norm": 0.6457917094230652, + "learning_rate": 2.8721956450016072e-06, + "loss": 0.1918, + "step": 53368 + }, + { + "epoch": 3.703351606411769, + "grad_norm": 0.7667796611785889, + "learning_rate": 2.8708588634063673e-06, + "loss": 0.2885, + "step": 53369 + }, + { + "epoch": 3.7034209978488652, + "grad_norm": 0.6003396511077881, + "learning_rate": 2.869522388436552e-06, + "loss": 0.2074, + "step": 53370 + }, + { + "epoch": 3.703490389285962, + "grad_norm": 1.1422234773635864, + "learning_rate": 2.8681862200963694e-06, + "loss": 0.293, + "step": 53371 + }, + { + "epoch": 3.703559780723059, + "grad_norm": 0.5669041275978088, + "learning_rate": 2.866850358390072e-06, + "loss": 0.1228, + "step": 53372 + }, + { + "epoch": 3.7036291721601553, + "grad_norm": 0.8655934929847717, + "learning_rate": 2.865514803321856e-06, + "loss": 0.2686, + "step": 53373 + }, + { + "epoch": 3.703698563597252, + "grad_norm": 0.9018146395683289, + "learning_rate": 2.8641795548959294e-06, + "loss": 0.3125, + "step": 53374 + }, + { + "epoch": 3.7037679550343485, + "grad_norm": 0.7973605990409851, + "learning_rate": 2.862844613116522e-06, + "loss": 0.2018, + "step": 53375 + }, + { + "epoch": 3.7038373464714454, + "grad_norm": 1.182747483253479, + "learning_rate": 2.8615099779878417e-06, + "loss": 0.4246, + "step": 53376 + }, + { + "epoch": 3.703906737908542, + "grad_norm": 0.9533163905143738, + "learning_rate": 2.8601756495141073e-06, + "loss": 0.2872, + "step": 53377 + }, + { + "epoch": 3.703976129345639, + "grad_norm": 1.1679565906524658, + "learning_rate": 2.858841627699538e-06, + "loss": 0.252, + "step": 53378 + }, + { + "epoch": 3.7040455207827354, + "grad_norm": 0.5207454562187195, + "learning_rate": 2.8575079125483294e-06, + "loss": 0.2142, + "step": 53379 + }, + { + "epoch": 3.704114912219832, + "grad_norm": 1.1838009357452393, + "learning_rate": 2.85617450406469e-06, + "loss": 0.3302, + "step": 53380 + }, + { + "epoch": 3.7041843036569286, + "grad_norm": 0.7125097513198853, + "learning_rate": 2.854841402252839e-06, + "loss": 0.2525, + "step": 53381 + }, + { + "epoch": 3.7042536950940255, + "grad_norm": 1.104551911354065, + "learning_rate": 2.8535086071169836e-06, + "loss": 0.3147, + "step": 53382 + }, + { + "epoch": 3.7043230865311223, + "grad_norm": 0.7622400522232056, + "learning_rate": 2.8521761186613315e-06, + "loss": 0.221, + "step": 53383 + }, + { + "epoch": 3.7043924779682187, + "grad_norm": 0.8528696298599243, + "learning_rate": 2.8508439368900906e-06, + "loss": 0.2191, + "step": 53384 + }, + { + "epoch": 3.7044618694053155, + "grad_norm": 0.9845252633094788, + "learning_rate": 2.849512061807469e-06, + "loss": 0.1775, + "step": 53385 + }, + { + "epoch": 3.704531260842412, + "grad_norm": 0.9189397096633911, + "learning_rate": 2.8481804934176626e-06, + "loss": 0.1741, + "step": 53386 + }, + { + "epoch": 3.7046006522795087, + "grad_norm": 0.8869960904121399, + "learning_rate": 2.8468492317248683e-06, + "loss": 0.2221, + "step": 53387 + }, + { + "epoch": 3.7046700437166056, + "grad_norm": 1.086305856704712, + "learning_rate": 2.8455182767333056e-06, + "loss": 0.2268, + "step": 53388 + }, + { + "epoch": 3.704739435153702, + "grad_norm": 0.6674150228500366, + "learning_rate": 2.84418762844717e-06, + "loss": 0.2696, + "step": 53389 + }, + { + "epoch": 3.704808826590799, + "grad_norm": 0.9333735108375549, + "learning_rate": 2.8428572868706594e-06, + "loss": 0.198, + "step": 53390 + }, + { + "epoch": 3.704878218027895, + "grad_norm": 0.822521984577179, + "learning_rate": 2.8415272520079807e-06, + "loss": 0.2358, + "step": 53391 + }, + { + "epoch": 3.704947609464992, + "grad_norm": 1.1947405338287354, + "learning_rate": 2.8401975238633307e-06, + "loss": 0.1644, + "step": 53392 + }, + { + "epoch": 3.705017000902089, + "grad_norm": 0.8931753635406494, + "learning_rate": 2.838868102440906e-06, + "loss": 0.1703, + "step": 53393 + }, + { + "epoch": 3.7050863923391852, + "grad_norm": 0.8596600890159607, + "learning_rate": 2.8375389877448922e-06, + "loss": 0.2254, + "step": 53394 + }, + { + "epoch": 3.705155783776282, + "grad_norm": 0.8059784173965454, + "learning_rate": 2.836210179779497e-06, + "loss": 0.2725, + "step": 53395 + }, + { + "epoch": 3.7052251752133785, + "grad_norm": 0.9516646265983582, + "learning_rate": 2.8348816785489176e-06, + "loss": 0.2109, + "step": 53396 + }, + { + "epoch": 3.7052945666504753, + "grad_norm": 1.1484953165054321, + "learning_rate": 2.8335534840573387e-06, + "loss": 0.2912, + "step": 53397 + }, + { + "epoch": 3.705363958087572, + "grad_norm": 0.7717800736427307, + "learning_rate": 2.8322255963089682e-06, + "loss": 0.1924, + "step": 53398 + }, + { + "epoch": 3.7054333495246685, + "grad_norm": 0.8287308216094971, + "learning_rate": 2.830898015307981e-06, + "loss": 0.1923, + "step": 53399 + }, + { + "epoch": 3.7055027409617654, + "grad_norm": 0.8986636400222778, + "learning_rate": 2.8295707410585626e-06, + "loss": 0.2582, + "step": 53400 + }, + { + "epoch": 3.7055721323988617, + "grad_norm": 0.727361798286438, + "learning_rate": 2.8282437735649313e-06, + "loss": 0.1806, + "step": 53401 + }, + { + "epoch": 3.7056415238359586, + "grad_norm": 0.9546692371368408, + "learning_rate": 2.8269171128312513e-06, + "loss": 0.276, + "step": 53402 + }, + { + "epoch": 3.7057109152730554, + "grad_norm": 1.1780165433883667, + "learning_rate": 2.8255907588617292e-06, + "loss": 0.2908, + "step": 53403 + }, + { + "epoch": 3.7057803067101522, + "grad_norm": 0.8816895484924316, + "learning_rate": 2.8242647116605402e-06, + "loss": 0.2393, + "step": 53404 + }, + { + "epoch": 3.7058496981472486, + "grad_norm": 1.1640123128890991, + "learning_rate": 2.82293897123187e-06, + "loss": 0.3672, + "step": 53405 + }, + { + "epoch": 3.705919089584345, + "grad_norm": 0.6532887816429138, + "learning_rate": 2.8216135375799034e-06, + "loss": 0.2166, + "step": 53406 + }, + { + "epoch": 3.705988481021442, + "grad_norm": 0.6930813789367676, + "learning_rate": 2.820288410708838e-06, + "loss": 0.1885, + "step": 53407 + }, + { + "epoch": 3.7060578724585387, + "grad_norm": 0.6582324504852295, + "learning_rate": 2.8189635906228473e-06, + "loss": 0.2406, + "step": 53408 + }, + { + "epoch": 3.7061272638956355, + "grad_norm": 0.6765602827072144, + "learning_rate": 2.8176390773261065e-06, + "loss": 0.1943, + "step": 53409 + }, + { + "epoch": 3.706196655332732, + "grad_norm": 0.9148696064949036, + "learning_rate": 2.8163148708228006e-06, + "loss": 0.2072, + "step": 53410 + }, + { + "epoch": 3.7062660467698287, + "grad_norm": 0.9299629330635071, + "learning_rate": 2.814990971117115e-06, + "loss": 0.2417, + "step": 53411 + }, + { + "epoch": 3.706335438206925, + "grad_norm": 0.6425341367721558, + "learning_rate": 2.813667378213247e-06, + "loss": 0.2106, + "step": 53412 + }, + { + "epoch": 3.706404829644022, + "grad_norm": 1.0754306316375732, + "learning_rate": 2.8123440921153375e-06, + "loss": 0.2227, + "step": 53413 + }, + { + "epoch": 3.706474221081119, + "grad_norm": 1.0210272073745728, + "learning_rate": 2.811021112827572e-06, + "loss": 0.2738, + "step": 53414 + }, + { + "epoch": 3.706543612518215, + "grad_norm": 1.0384693145751953, + "learning_rate": 2.8096984403541473e-06, + "loss": 0.2529, + "step": 53415 + }, + { + "epoch": 3.706613003955312, + "grad_norm": 1.1552571058273315, + "learning_rate": 2.808376074699226e-06, + "loss": 0.2851, + "step": 53416 + }, + { + "epoch": 3.7066823953924084, + "grad_norm": 1.066648006439209, + "learning_rate": 2.8070540158669945e-06, + "loss": 0.3839, + "step": 53417 + }, + { + "epoch": 3.7067517868295052, + "grad_norm": 0.6990994811058044, + "learning_rate": 2.805732263861605e-06, + "loss": 0.2144, + "step": 53418 + }, + { + "epoch": 3.706821178266602, + "grad_norm": 0.8542277216911316, + "learning_rate": 2.804410818687253e-06, + "loss": 0.2179, + "step": 53419 + }, + { + "epoch": 3.7068905697036985, + "grad_norm": 1.0318747758865356, + "learning_rate": 2.803089680348092e-06, + "loss": 0.1347, + "step": 53420 + }, + { + "epoch": 3.7069599611407953, + "grad_norm": 0.6333833336830139, + "learning_rate": 2.8017688488483074e-06, + "loss": 0.207, + "step": 53421 + }, + { + "epoch": 3.7070293525778917, + "grad_norm": 0.7743300199508667, + "learning_rate": 2.800448324192051e-06, + "loss": 0.2471, + "step": 53422 + }, + { + "epoch": 3.7070987440149885, + "grad_norm": 0.8491235971450806, + "learning_rate": 2.7991281063835083e-06, + "loss": 0.1683, + "step": 53423 + }, + { + "epoch": 3.7071681354520853, + "grad_norm": 1.143385648727417, + "learning_rate": 2.797808195426832e-06, + "loss": 0.2557, + "step": 53424 + }, + { + "epoch": 3.7072375268891817, + "grad_norm": 1.0345609188079834, + "learning_rate": 2.7964885913262073e-06, + "loss": 0.3431, + "step": 53425 + }, + { + "epoch": 3.7073069183262786, + "grad_norm": 0.9215732216835022, + "learning_rate": 2.795169294085798e-06, + "loss": 0.2404, + "step": 53426 + }, + { + "epoch": 3.707376309763375, + "grad_norm": 1.1000306606292725, + "learning_rate": 2.793850303709744e-06, + "loss": 0.2489, + "step": 53427 + }, + { + "epoch": 3.707445701200472, + "grad_norm": 0.8433422446250916, + "learning_rate": 2.7925316202022434e-06, + "loss": 0.2665, + "step": 53428 + }, + { + "epoch": 3.7075150926375686, + "grad_norm": 1.3141841888427734, + "learning_rate": 2.7912132435674256e-06, + "loss": 0.2952, + "step": 53429 + }, + { + "epoch": 3.7075844840746655, + "grad_norm": 0.9353423118591309, + "learning_rate": 2.789895173809476e-06, + "loss": 0.203, + "step": 53430 + }, + { + "epoch": 3.707653875511762, + "grad_norm": 0.8049646019935608, + "learning_rate": 2.7885774109325467e-06, + "loss": 0.1989, + "step": 53431 + }, + { + "epoch": 3.7077232669488587, + "grad_norm": 0.8489254117012024, + "learning_rate": 2.787259954940802e-06, + "loss": 0.2654, + "step": 53432 + }, + { + "epoch": 3.707792658385955, + "grad_norm": 1.1527072191238403, + "learning_rate": 2.7859428058384043e-06, + "loss": 0.2312, + "step": 53433 + }, + { + "epoch": 3.707862049823052, + "grad_norm": 1.0908621549606323, + "learning_rate": 2.784625963629506e-06, + "loss": 0.2563, + "step": 53434 + }, + { + "epoch": 3.7079314412601487, + "grad_norm": 1.131691813468933, + "learning_rate": 2.783309428318259e-06, + "loss": 0.2522, + "step": 53435 + }, + { + "epoch": 3.708000832697245, + "grad_norm": 0.8344026803970337, + "learning_rate": 2.7819931999088276e-06, + "loss": 0.1845, + "step": 53436 + }, + { + "epoch": 3.708070224134342, + "grad_norm": 0.974381148815155, + "learning_rate": 2.7806772784053526e-06, + "loss": 0.271, + "step": 53437 + }, + { + "epoch": 3.7081396155714383, + "grad_norm": 0.9306786060333252, + "learning_rate": 2.779361663812008e-06, + "loss": 0.2899, + "step": 53438 + }, + { + "epoch": 3.708209007008535, + "grad_norm": 0.8939138054847717, + "learning_rate": 2.7780463561329574e-06, + "loss": 0.2646, + "step": 53439 + }, + { + "epoch": 3.708278398445632, + "grad_norm": 0.6772498488426208, + "learning_rate": 2.7767313553723086e-06, + "loss": 0.1798, + "step": 53440 + }, + { + "epoch": 3.7083477898827284, + "grad_norm": 0.6179294586181641, + "learning_rate": 2.775416661534247e-06, + "loss": 0.0828, + "step": 53441 + }, + { + "epoch": 3.7084171813198252, + "grad_norm": 1.252742052078247, + "learning_rate": 2.7741022746229252e-06, + "loss": 0.2896, + "step": 53442 + }, + { + "epoch": 3.7084865727569216, + "grad_norm": 0.9672442078590393, + "learning_rate": 2.772788194642473e-06, + "loss": 0.3629, + "step": 53443 + }, + { + "epoch": 3.7085559641940185, + "grad_norm": 0.6507073044776917, + "learning_rate": 2.7714744215970535e-06, + "loss": 0.1961, + "step": 53444 + }, + { + "epoch": 3.7086253556311153, + "grad_norm": 1.0337731838226318, + "learning_rate": 2.770160955490797e-06, + "loss": 0.3934, + "step": 53445 + }, + { + "epoch": 3.7086947470682117, + "grad_norm": 0.6816456317901611, + "learning_rate": 2.7688477963278668e-06, + "loss": 0.2427, + "step": 53446 + }, + { + "epoch": 3.7087641385053085, + "grad_norm": 0.9425927400588989, + "learning_rate": 2.7675349441124156e-06, + "loss": 0.2767, + "step": 53447 + }, + { + "epoch": 3.708833529942405, + "grad_norm": 0.9319326877593994, + "learning_rate": 2.7662223988485614e-06, + "loss": 0.2598, + "step": 53448 + }, + { + "epoch": 3.7089029213795017, + "grad_norm": 0.6959130764007568, + "learning_rate": 2.764910160540468e-06, + "loss": 0.1038, + "step": 53449 + }, + { + "epoch": 3.7089723128165986, + "grad_norm": 1.496842384338379, + "learning_rate": 2.763598229192266e-06, + "loss": 0.3313, + "step": 53450 + }, + { + "epoch": 3.709041704253695, + "grad_norm": 0.8951818346977234, + "learning_rate": 2.7622866048081066e-06, + "loss": 0.1926, + "step": 53451 + }, + { + "epoch": 3.709111095690792, + "grad_norm": 0.6732591390609741, + "learning_rate": 2.76097528739212e-06, + "loss": 0.2394, + "step": 53452 + }, + { + "epoch": 3.709180487127888, + "grad_norm": 1.0787614583969116, + "learning_rate": 2.7596642769484592e-06, + "loss": 0.3061, + "step": 53453 + }, + { + "epoch": 3.709249878564985, + "grad_norm": 1.0830328464508057, + "learning_rate": 2.758353573481265e-06, + "loss": 0.3263, + "step": 53454 + }, + { + "epoch": 3.709319270002082, + "grad_norm": 0.7662463784217834, + "learning_rate": 2.757043176994656e-06, + "loss": 0.2113, + "step": 53455 + }, + { + "epoch": 3.7093886614391787, + "grad_norm": 1.4766871929168701, + "learning_rate": 2.7557330874927733e-06, + "loss": 0.3014, + "step": 53456 + }, + { + "epoch": 3.709458052876275, + "grad_norm": 0.9075803160667419, + "learning_rate": 2.7544233049797695e-06, + "loss": 0.2496, + "step": 53457 + }, + { + "epoch": 3.709527444313372, + "grad_norm": 0.8047435283660889, + "learning_rate": 2.7531138294597636e-06, + "loss": 0.1246, + "step": 53458 + }, + { + "epoch": 3.7095968357504683, + "grad_norm": 0.5927491784095764, + "learning_rate": 2.7518046609368964e-06, + "loss": 0.1709, + "step": 53459 + }, + { + "epoch": 3.709666227187565, + "grad_norm": 0.7491317987442017, + "learning_rate": 2.750495799415298e-06, + "loss": 0.2345, + "step": 53460 + }, + { + "epoch": 3.709735618624662, + "grad_norm": 0.8778272271156311, + "learning_rate": 2.7491872448991096e-06, + "loss": 0.1929, + "step": 53461 + }, + { + "epoch": 3.7098050100617583, + "grad_norm": 0.7946631908416748, + "learning_rate": 2.747878997392439e-06, + "loss": 0.2757, + "step": 53462 + }, + { + "epoch": 3.709874401498855, + "grad_norm": 1.2398746013641357, + "learning_rate": 2.7465710568994386e-06, + "loss": 0.2308, + "step": 53463 + }, + { + "epoch": 3.7099437929359516, + "grad_norm": 0.7279679775238037, + "learning_rate": 2.745263423424227e-06, + "loss": 0.1937, + "step": 53464 + }, + { + "epoch": 3.7100131843730484, + "grad_norm": 0.747250497341156, + "learning_rate": 2.7439560969709343e-06, + "loss": 0.2543, + "step": 53465 + }, + { + "epoch": 3.7100825758101452, + "grad_norm": 0.9367614388465881, + "learning_rate": 2.742649077543691e-06, + "loss": 0.2524, + "step": 53466 + }, + { + "epoch": 3.7101519672472416, + "grad_norm": 0.7935405373573303, + "learning_rate": 2.7413423651466262e-06, + "loss": 0.2376, + "step": 53467 + }, + { + "epoch": 3.7102213586843384, + "grad_norm": 0.9997909665107727, + "learning_rate": 2.7400359597838597e-06, + "loss": 0.2002, + "step": 53468 + }, + { + "epoch": 3.710290750121435, + "grad_norm": 0.7666861414909363, + "learning_rate": 2.7387298614595102e-06, + "loss": 0.2061, + "step": 53469 + }, + { + "epoch": 3.7103601415585317, + "grad_norm": 1.1065040826797485, + "learning_rate": 2.7374240701777076e-06, + "loss": 0.2166, + "step": 53470 + }, + { + "epoch": 3.7104295329956285, + "grad_norm": 1.029942274093628, + "learning_rate": 2.736118585942571e-06, + "loss": 0.2466, + "step": 53471 + }, + { + "epoch": 3.710498924432725, + "grad_norm": 2.234466075897217, + "learning_rate": 2.734813408758219e-06, + "loss": 0.2305, + "step": 53472 + }, + { + "epoch": 3.7105683158698217, + "grad_norm": 0.8832914233207703, + "learning_rate": 2.733508538628782e-06, + "loss": 0.1579, + "step": 53473 + }, + { + "epoch": 3.710637707306918, + "grad_norm": 0.6858533620834351, + "learning_rate": 2.7322039755583785e-06, + "loss": 0.2514, + "step": 53474 + }, + { + "epoch": 3.710707098744015, + "grad_norm": 0.6930400133132935, + "learning_rate": 2.7308997195511164e-06, + "loss": 0.183, + "step": 53475 + }, + { + "epoch": 3.7107764901811118, + "grad_norm": 0.773495614528656, + "learning_rate": 2.7295957706111154e-06, + "loss": 0.2374, + "step": 53476 + }, + { + "epoch": 3.710845881618208, + "grad_norm": 0.9907159805297852, + "learning_rate": 2.7282921287424934e-06, + "loss": 0.1937, + "step": 53477 + }, + { + "epoch": 3.710915273055305, + "grad_norm": 0.6568994522094727, + "learning_rate": 2.726988793949381e-06, + "loss": 0.1886, + "step": 53478 + }, + { + "epoch": 3.7109846644924014, + "grad_norm": 0.8103972673416138, + "learning_rate": 2.725685766235875e-06, + "loss": 0.1599, + "step": 53479 + }, + { + "epoch": 3.711054055929498, + "grad_norm": 0.7353747487068176, + "learning_rate": 2.7243830456060825e-06, + "loss": 0.195, + "step": 53480 + }, + { + "epoch": 3.711123447366595, + "grad_norm": 1.0937325954437256, + "learning_rate": 2.7230806320641457e-06, + "loss": 0.4381, + "step": 53481 + }, + { + "epoch": 3.711192838803692, + "grad_norm": 0.8201819658279419, + "learning_rate": 2.721778525614138e-06, + "loss": 0.2387, + "step": 53482 + }, + { + "epoch": 3.7112622302407883, + "grad_norm": 1.096113681793213, + "learning_rate": 2.7204767262602014e-06, + "loss": 0.3665, + "step": 53483 + }, + { + "epoch": 3.711331621677885, + "grad_norm": 0.8021572828292847, + "learning_rate": 2.7191752340064212e-06, + "loss": 0.211, + "step": 53484 + }, + { + "epoch": 3.7114010131149815, + "grad_norm": 1.1115800142288208, + "learning_rate": 2.7178740488569277e-06, + "loss": 0.3137, + "step": 53485 + }, + { + "epoch": 3.7114704045520783, + "grad_norm": 0.7738127708435059, + "learning_rate": 2.716573170815806e-06, + "loss": 0.261, + "step": 53486 + }, + { + "epoch": 3.711539795989175, + "grad_norm": 2.4484291076660156, + "learning_rate": 2.7152725998871864e-06, + "loss": 0.2624, + "step": 53487 + }, + { + "epoch": 3.7116091874262715, + "grad_norm": 0.6297059655189514, + "learning_rate": 2.713972336075177e-06, + "loss": 0.1873, + "step": 53488 + }, + { + "epoch": 3.7116785788633684, + "grad_norm": 0.915729284286499, + "learning_rate": 2.7126723793838514e-06, + "loss": 0.3055, + "step": 53489 + }, + { + "epoch": 3.7117479703004648, + "grad_norm": 0.8160518407821655, + "learning_rate": 2.7113727298173297e-06, + "loss": 0.2036, + "step": 53490 + }, + { + "epoch": 3.7118173617375616, + "grad_norm": 0.682732880115509, + "learning_rate": 2.710073387379719e-06, + "loss": 0.208, + "step": 53491 + }, + { + "epoch": 3.7118867531746584, + "grad_norm": 1.2826625108718872, + "learning_rate": 2.7087743520751275e-06, + "loss": 0.3421, + "step": 53492 + }, + { + "epoch": 3.711956144611755, + "grad_norm": 0.9430918097496033, + "learning_rate": 2.7074756239076404e-06, + "loss": 0.2172, + "step": 53493 + }, + { + "epoch": 3.7120255360488517, + "grad_norm": 1.3131380081176758, + "learning_rate": 2.706177202881366e-06, + "loss": 0.2143, + "step": 53494 + }, + { + "epoch": 3.712094927485948, + "grad_norm": 0.6917241811752319, + "learning_rate": 2.7048790890004116e-06, + "loss": 0.1929, + "step": 53495 + }, + { + "epoch": 3.712164318923045, + "grad_norm": 0.9099398255348206, + "learning_rate": 2.7035812822688524e-06, + "loss": 0.178, + "step": 53496 + }, + { + "epoch": 3.7122337103601417, + "grad_norm": 0.8164353966712952, + "learning_rate": 2.7022837826908064e-06, + "loss": 0.2358, + "step": 53497 + }, + { + "epoch": 3.712303101797238, + "grad_norm": 0.7114750146865845, + "learning_rate": 2.700986590270349e-06, + "loss": 0.2384, + "step": 53498 + }, + { + "epoch": 3.712372493234335, + "grad_norm": 0.7654846906661987, + "learning_rate": 2.699689705011599e-06, + "loss": 0.1648, + "step": 53499 + }, + { + "epoch": 3.7124418846714313, + "grad_norm": 0.8984987735748291, + "learning_rate": 2.6983931269186412e-06, + "loss": 0.1551, + "step": 53500 + }, + { + "epoch": 3.712511276108528, + "grad_norm": 0.9311047792434692, + "learning_rate": 2.6970968559955624e-06, + "loss": 0.2195, + "step": 53501 + }, + { + "epoch": 3.712580667545625, + "grad_norm": 0.8747280240058899, + "learning_rate": 2.6958008922464696e-06, + "loss": 0.3224, + "step": 53502 + }, + { + "epoch": 3.7126500589827214, + "grad_norm": 0.7482489943504333, + "learning_rate": 2.6945052356754374e-06, + "loss": 0.246, + "step": 53503 + }, + { + "epoch": 3.712719450419818, + "grad_norm": 0.5572304725646973, + "learning_rate": 2.693209886286563e-06, + "loss": 0.1469, + "step": 53504 + }, + { + "epoch": 3.7127888418569146, + "grad_norm": 0.9073674082756042, + "learning_rate": 2.6919148440839313e-06, + "loss": 0.1742, + "step": 53505 + }, + { + "epoch": 3.7128582332940114, + "grad_norm": 0.654228150844574, + "learning_rate": 2.6906201090716286e-06, + "loss": 0.1667, + "step": 53506 + }, + { + "epoch": 3.7129276247311083, + "grad_norm": 1.1228435039520264, + "learning_rate": 2.6893256812537627e-06, + "loss": 0.2561, + "step": 53507 + }, + { + "epoch": 3.712997016168205, + "grad_norm": 0.7118674516677856, + "learning_rate": 2.6880315606343964e-06, + "loss": 0.164, + "step": 53508 + }, + { + "epoch": 3.7130664076053015, + "grad_norm": 0.925494372844696, + "learning_rate": 2.6867377472176268e-06, + "loss": 0.2079, + "step": 53509 + }, + { + "epoch": 3.7131357990423983, + "grad_norm": 0.6921838521957397, + "learning_rate": 2.68544424100754e-06, + "loss": 0.1817, + "step": 53510 + }, + { + "epoch": 3.7132051904794947, + "grad_norm": 0.7535250782966614, + "learning_rate": 2.6841510420082096e-06, + "loss": 0.1561, + "step": 53511 + }, + { + "epoch": 3.7132745819165915, + "grad_norm": 0.9613112807273865, + "learning_rate": 2.6828581502237325e-06, + "loss": 0.167, + "step": 53512 + }, + { + "epoch": 3.7133439733536884, + "grad_norm": 0.7689197063446045, + "learning_rate": 2.6815655656581728e-06, + "loss": 0.1905, + "step": 53513 + }, + { + "epoch": 3.7134133647907848, + "grad_norm": 0.8933525681495667, + "learning_rate": 2.6802732883156266e-06, + "loss": 0.2398, + "step": 53514 + }, + { + "epoch": 3.7134827562278816, + "grad_norm": 1.284014105796814, + "learning_rate": 2.6789813182001576e-06, + "loss": 0.4185, + "step": 53515 + }, + { + "epoch": 3.713552147664978, + "grad_norm": 0.7565107345581055, + "learning_rate": 2.6776896553158625e-06, + "loss": 0.209, + "step": 53516 + }, + { + "epoch": 3.713621539102075, + "grad_norm": 1.0940855741500854, + "learning_rate": 2.676398299666805e-06, + "loss": 0.2449, + "step": 53517 + }, + { + "epoch": 3.7136909305391717, + "grad_norm": 0.7278448939323425, + "learning_rate": 2.6751072512570586e-06, + "loss": 0.2607, + "step": 53518 + }, + { + "epoch": 3.713760321976268, + "grad_norm": 0.8401092886924744, + "learning_rate": 2.6738165100907097e-06, + "loss": 0.2128, + "step": 53519 + }, + { + "epoch": 3.713829713413365, + "grad_norm": 1.0269908905029297, + "learning_rate": 2.672526076171833e-06, + "loss": 0.3615, + "step": 53520 + }, + { + "epoch": 3.7138991048504613, + "grad_norm": 0.9644733667373657, + "learning_rate": 2.6712359495045024e-06, + "loss": 0.1911, + "step": 53521 + }, + { + "epoch": 3.713968496287558, + "grad_norm": 1.266374945640564, + "learning_rate": 2.6699461300927818e-06, + "loss": 0.426, + "step": 53522 + }, + { + "epoch": 3.714037887724655, + "grad_norm": 0.8473988771438599, + "learning_rate": 2.6686566179407678e-06, + "loss": 0.2738, + "step": 53523 + }, + { + "epoch": 3.7141072791617513, + "grad_norm": 0.6631972789764404, + "learning_rate": 2.6673674130524906e-06, + "loss": 0.2154, + "step": 53524 + }, + { + "epoch": 3.714176670598848, + "grad_norm": 0.9738932847976685, + "learning_rate": 2.666078515432047e-06, + "loss": 0.2278, + "step": 53525 + }, + { + "epoch": 3.7142460620359445, + "grad_norm": 0.9439594745635986, + "learning_rate": 2.6647899250834997e-06, + "loss": 0.2178, + "step": 53526 + }, + { + "epoch": 3.7143154534730414, + "grad_norm": 0.7208015322685242, + "learning_rate": 2.6635016420109236e-06, + "loss": 0.2813, + "step": 53527 + }, + { + "epoch": 3.714384844910138, + "grad_norm": 0.8167102932929993, + "learning_rate": 2.662213666218372e-06, + "loss": 0.1739, + "step": 53528 + }, + { + "epoch": 3.7144542363472346, + "grad_norm": 0.730078935623169, + "learning_rate": 2.6609259977099178e-06, + "loss": 0.2371, + "step": 53529 + }, + { + "epoch": 3.7145236277843314, + "grad_norm": 0.9390745162963867, + "learning_rate": 2.6596386364896366e-06, + "loss": 0.2679, + "step": 53530 + }, + { + "epoch": 3.714593019221428, + "grad_norm": 0.949694812297821, + "learning_rate": 2.65835158256158e-06, + "loss": 0.2367, + "step": 53531 + }, + { + "epoch": 3.7146624106585246, + "grad_norm": 1.0022028684616089, + "learning_rate": 2.657064835929812e-06, + "loss": 0.2238, + "step": 53532 + }, + { + "epoch": 3.7147318020956215, + "grad_norm": 1.031520128250122, + "learning_rate": 2.655778396598396e-06, + "loss": 0.2703, + "step": 53533 + }, + { + "epoch": 3.7148011935327183, + "grad_norm": 1.107925295829773, + "learning_rate": 2.6544922645713843e-06, + "loss": 0.289, + "step": 53534 + }, + { + "epoch": 3.7148705849698147, + "grad_norm": 0.7521624565124512, + "learning_rate": 2.6532064398528623e-06, + "loss": 0.1804, + "step": 53535 + }, + { + "epoch": 3.7149399764069115, + "grad_norm": 0.8379143476486206, + "learning_rate": 2.65192092244686e-06, + "loss": 0.244, + "step": 53536 + }, + { + "epoch": 3.715009367844008, + "grad_norm": 0.8082336783409119, + "learning_rate": 2.650635712357463e-06, + "loss": 0.2059, + "step": 53537 + }, + { + "epoch": 3.7150787592811048, + "grad_norm": 0.8029707670211792, + "learning_rate": 2.649350809588713e-06, + "loss": 0.204, + "step": 53538 + }, + { + "epoch": 3.7151481507182016, + "grad_norm": 0.8984634876251221, + "learning_rate": 2.648066214144651e-06, + "loss": 0.2976, + "step": 53539 + }, + { + "epoch": 3.715217542155298, + "grad_norm": 0.6711752414703369, + "learning_rate": 2.6467819260293625e-06, + "loss": 0.256, + "step": 53540 + }, + { + "epoch": 3.715286933592395, + "grad_norm": 1.3386255502700806, + "learning_rate": 2.6454979452468885e-06, + "loss": 0.238, + "step": 53541 + }, + { + "epoch": 3.715356325029491, + "grad_norm": 0.9244033694267273, + "learning_rate": 2.644214271801271e-06, + "loss": 0.2809, + "step": 53542 + }, + { + "epoch": 3.715425716466588, + "grad_norm": 0.7351270914077759, + "learning_rate": 2.6429309056965833e-06, + "loss": 0.1648, + "step": 53543 + }, + { + "epoch": 3.715495107903685, + "grad_norm": 0.6610926389694214, + "learning_rate": 2.641647846936879e-06, + "loss": 0.2249, + "step": 53544 + }, + { + "epoch": 3.7155644993407813, + "grad_norm": 0.7383273839950562, + "learning_rate": 2.640365095526176e-06, + "loss": 0.1877, + "step": 53545 + }, + { + "epoch": 3.715633890777878, + "grad_norm": 0.8510856628417969, + "learning_rate": 2.6390826514685495e-06, + "loss": 0.2426, + "step": 53546 + }, + { + "epoch": 3.7157032822149745, + "grad_norm": 0.8789490461349487, + "learning_rate": 2.6378005147680517e-06, + "loss": 0.2715, + "step": 53547 + }, + { + "epoch": 3.7157726736520713, + "grad_norm": 0.7012153267860413, + "learning_rate": 2.636518685428724e-06, + "loss": 0.1147, + "step": 53548 + }, + { + "epoch": 3.715842065089168, + "grad_norm": 0.5935274362564087, + "learning_rate": 2.635237163454596e-06, + "loss": 0.1997, + "step": 53549 + }, + { + "epoch": 3.7159114565262645, + "grad_norm": 0.672487735748291, + "learning_rate": 2.633955948849731e-06, + "loss": 0.1491, + "step": 53550 + }, + { + "epoch": 3.7159808479633614, + "grad_norm": 0.5976373553276062, + "learning_rate": 2.6326750416181822e-06, + "loss": 0.1419, + "step": 53551 + }, + { + "epoch": 3.7160502394004578, + "grad_norm": 1.0962400436401367, + "learning_rate": 2.6313944417639678e-06, + "loss": 0.238, + "step": 53552 + }, + { + "epoch": 3.7161196308375546, + "grad_norm": 0.901162326335907, + "learning_rate": 2.630114149291152e-06, + "loss": 0.2496, + "step": 53553 + }, + { + "epoch": 3.7161890222746514, + "grad_norm": 0.6890872716903687, + "learning_rate": 2.628834164203764e-06, + "loss": 0.2317, + "step": 53554 + }, + { + "epoch": 3.716258413711748, + "grad_norm": 0.9397916197776794, + "learning_rate": 2.6275544865058565e-06, + "loss": 0.212, + "step": 53555 + }, + { + "epoch": 3.7163278051488446, + "grad_norm": 0.8574752807617188, + "learning_rate": 2.6262751162014487e-06, + "loss": 0.221, + "step": 53556 + }, + { + "epoch": 3.716397196585941, + "grad_norm": 0.873950183391571, + "learning_rate": 2.624996053294615e-06, + "loss": 0.1982, + "step": 53557 + }, + { + "epoch": 3.716466588023038, + "grad_norm": 3.1158876419067383, + "learning_rate": 2.623717297789352e-06, + "loss": 0.3195, + "step": 53558 + }, + { + "epoch": 3.7165359794601347, + "grad_norm": 0.9165059328079224, + "learning_rate": 2.6224388496897236e-06, + "loss": 0.2542, + "step": 53559 + }, + { + "epoch": 3.7166053708972315, + "grad_norm": 0.8800368309020996, + "learning_rate": 2.621160708999748e-06, + "loss": 0.3178, + "step": 53560 + }, + { + "epoch": 3.716674762334328, + "grad_norm": 0.8365986943244934, + "learning_rate": 2.6198828757234783e-06, + "loss": 0.2157, + "step": 53561 + }, + { + "epoch": 3.7167441537714248, + "grad_norm": 0.8668221235275269, + "learning_rate": 2.6186053498649443e-06, + "loss": 0.2857, + "step": 53562 + }, + { + "epoch": 3.716813545208521, + "grad_norm": 1.075365424156189, + "learning_rate": 2.617328131428165e-06, + "loss": 0.3319, + "step": 53563 + }, + { + "epoch": 3.716882936645618, + "grad_norm": 0.8143429160118103, + "learning_rate": 2.616051220417193e-06, + "loss": 0.1687, + "step": 53564 + }, + { + "epoch": 3.716952328082715, + "grad_norm": 0.7302149534225464, + "learning_rate": 2.6147746168360467e-06, + "loss": 0.2062, + "step": 53565 + }, + { + "epoch": 3.717021719519811, + "grad_norm": 0.9590327739715576, + "learning_rate": 2.6134983206887563e-06, + "loss": 0.2815, + "step": 53566 + }, + { + "epoch": 3.717091110956908, + "grad_norm": 1.5016762018203735, + "learning_rate": 2.6122223319793416e-06, + "loss": 0.2427, + "step": 53567 + }, + { + "epoch": 3.7171605023940044, + "grad_norm": 1.051835536956787, + "learning_rate": 2.610946650711854e-06, + "loss": 0.268, + "step": 53568 + }, + { + "epoch": 3.7172298938311013, + "grad_norm": 1.155423641204834, + "learning_rate": 2.609671276890302e-06, + "loss": 0.293, + "step": 53569 + }, + { + "epoch": 3.717299285268198, + "grad_norm": 1.2359062433242798, + "learning_rate": 2.6083962105187267e-06, + "loss": 0.4045, + "step": 53570 + }, + { + "epoch": 3.7173686767052945, + "grad_norm": 0.7439399361610413, + "learning_rate": 2.6071214516011465e-06, + "loss": 0.2384, + "step": 53571 + }, + { + "epoch": 3.7174380681423913, + "grad_norm": 0.7663937211036682, + "learning_rate": 2.6058470001415815e-06, + "loss": 0.1609, + "step": 53572 + }, + { + "epoch": 3.7175074595794877, + "grad_norm": 0.8010006546974182, + "learning_rate": 2.6045728561440496e-06, + "loss": 0.1933, + "step": 53573 + }, + { + "epoch": 3.7175768510165845, + "grad_norm": 0.9747952222824097, + "learning_rate": 2.603299019612582e-06, + "loss": 0.2503, + "step": 53574 + }, + { + "epoch": 3.7176462424536814, + "grad_norm": 0.793226957321167, + "learning_rate": 2.6020254905512077e-06, + "loss": 0.2257, + "step": 53575 + }, + { + "epoch": 3.7177156338907777, + "grad_norm": 0.715286135673523, + "learning_rate": 2.600752268963924e-06, + "loss": 0.1748, + "step": 53576 + }, + { + "epoch": 3.7177850253278746, + "grad_norm": 0.5939645171165466, + "learning_rate": 2.5994793548547834e-06, + "loss": 0.1677, + "step": 53577 + }, + { + "epoch": 3.717854416764971, + "grad_norm": 1.1953766345977783, + "learning_rate": 2.5982067482277827e-06, + "loss": 0.3492, + "step": 53578 + }, + { + "epoch": 3.717923808202068, + "grad_norm": 1.0815637111663818, + "learning_rate": 2.5969344490869407e-06, + "loss": 0.2985, + "step": 53579 + }, + { + "epoch": 3.7179931996391646, + "grad_norm": 0.9293866753578186, + "learning_rate": 2.5956624574362652e-06, + "loss": 0.3031, + "step": 53580 + }, + { + "epoch": 3.718062591076261, + "grad_norm": 1.101306676864624, + "learning_rate": 2.594390773279787e-06, + "loss": 0.3234, + "step": 53581 + }, + { + "epoch": 3.718131982513358, + "grad_norm": 0.7358130216598511, + "learning_rate": 2.593119396621535e-06, + "loss": 0.2456, + "step": 53582 + }, + { + "epoch": 3.7182013739504542, + "grad_norm": 1.3965638875961304, + "learning_rate": 2.591848327465485e-06, + "loss": 0.3221, + "step": 53583 + }, + { + "epoch": 3.718270765387551, + "grad_norm": 1.0690423250198364, + "learning_rate": 2.590577565815666e-06, + "loss": 0.3291, + "step": 53584 + }, + { + "epoch": 3.718340156824648, + "grad_norm": 0.6548054814338684, + "learning_rate": 2.589307111676098e-06, + "loss": 0.1957, + "step": 53585 + }, + { + "epoch": 3.7184095482617447, + "grad_norm": 1.6486119031906128, + "learning_rate": 2.5880369650507997e-06, + "loss": 0.1924, + "step": 53586 + }, + { + "epoch": 3.718478939698841, + "grad_norm": 1.1150442361831665, + "learning_rate": 2.5867671259437455e-06, + "loss": 0.3049, + "step": 53587 + }, + { + "epoch": 3.718548331135938, + "grad_norm": 0.8723235130310059, + "learning_rate": 2.5854975943589764e-06, + "loss": 0.1458, + "step": 53588 + }, + { + "epoch": 3.7186177225730344, + "grad_norm": 1.0422981977462769, + "learning_rate": 2.584228370300479e-06, + "loss": 0.2475, + "step": 53589 + }, + { + "epoch": 3.718687114010131, + "grad_norm": 0.991155207157135, + "learning_rate": 2.582959453772271e-06, + "loss": 0.2072, + "step": 53590 + }, + { + "epoch": 3.718756505447228, + "grad_norm": 0.6940633654594421, + "learning_rate": 2.5816908447783615e-06, + "loss": 0.1895, + "step": 53591 + }, + { + "epoch": 3.7188258968843244, + "grad_norm": 0.9290223121643066, + "learning_rate": 2.580422543322769e-06, + "loss": 0.2357, + "step": 53592 + }, + { + "epoch": 3.7188952883214212, + "grad_norm": 0.8469317555427551, + "learning_rate": 2.579154549409457e-06, + "loss": 0.2378, + "step": 53593 + }, + { + "epoch": 3.7189646797585176, + "grad_norm": 0.5864863991737366, + "learning_rate": 2.5778868630424556e-06, + "loss": 0.1236, + "step": 53594 + }, + { + "epoch": 3.7190340711956145, + "grad_norm": 0.8777565956115723, + "learning_rate": 2.576619484225762e-06, + "loss": 0.2726, + "step": 53595 + }, + { + "epoch": 3.7191034626327113, + "grad_norm": 0.7381688356399536, + "learning_rate": 2.575352412963383e-06, + "loss": 0.2453, + "step": 53596 + }, + { + "epoch": 3.7191728540698077, + "grad_norm": 1.299285650253296, + "learning_rate": 2.5740856492593056e-06, + "loss": 0.4038, + "step": 53597 + }, + { + "epoch": 3.7192422455069045, + "grad_norm": 0.7522048354148865, + "learning_rate": 2.572819193117537e-06, + "loss": 0.1746, + "step": 53598 + }, + { + "epoch": 3.719311636944001, + "grad_norm": 0.8101289868354797, + "learning_rate": 2.5715530445420854e-06, + "loss": 0.2574, + "step": 53599 + }, + { + "epoch": 3.7193810283810977, + "grad_norm": 0.7629454731941223, + "learning_rate": 2.5702872035369363e-06, + "loss": 0.1903, + "step": 53600 + }, + { + "epoch": 3.7194504198181946, + "grad_norm": 0.6601265668869019, + "learning_rate": 2.5690216701060756e-06, + "loss": 0.0954, + "step": 53601 + }, + { + "epoch": 3.719519811255291, + "grad_norm": 0.6997973918914795, + "learning_rate": 2.567756444253511e-06, + "loss": 0.2008, + "step": 53602 + }, + { + "epoch": 3.719589202692388, + "grad_norm": 1.17392897605896, + "learning_rate": 2.5664915259832388e-06, + "loss": 0.2942, + "step": 53603 + }, + { + "epoch": 3.719658594129484, + "grad_norm": 1.6294580698013306, + "learning_rate": 2.5652269152992457e-06, + "loss": 0.2766, + "step": 53604 + }, + { + "epoch": 3.719727985566581, + "grad_norm": 1.0438859462738037, + "learning_rate": 2.563962612205528e-06, + "loss": 0.3376, + "step": 53605 + }, + { + "epoch": 3.719797377003678, + "grad_norm": 1.0197067260742188, + "learning_rate": 2.562698616706083e-06, + "loss": 0.3824, + "step": 53606 + }, + { + "epoch": 3.7198667684407747, + "grad_norm": 0.8019470572471619, + "learning_rate": 2.5614349288048844e-06, + "loss": 0.1898, + "step": 53607 + }, + { + "epoch": 3.719936159877871, + "grad_norm": 0.633819580078125, + "learning_rate": 2.5601715485059406e-06, + "loss": 0.2393, + "step": 53608 + }, + { + "epoch": 3.7200055513149675, + "grad_norm": 1.3092267513275146, + "learning_rate": 2.5589084758132153e-06, + "loss": 0.3337, + "step": 53609 + }, + { + "epoch": 3.7200749427520643, + "grad_norm": 0.7490355968475342, + "learning_rate": 2.557645710730716e-06, + "loss": 0.1847, + "step": 53610 + }, + { + "epoch": 3.720144334189161, + "grad_norm": 1.0427485704421997, + "learning_rate": 2.556383253262429e-06, + "loss": 0.3135, + "step": 53611 + }, + { + "epoch": 3.720213725626258, + "grad_norm": 1.0077768564224243, + "learning_rate": 2.555121103412328e-06, + "loss": 0.2804, + "step": 53612 + }, + { + "epoch": 3.7202831170633544, + "grad_norm": 0.838379979133606, + "learning_rate": 2.5538592611844212e-06, + "loss": 0.3159, + "step": 53613 + }, + { + "epoch": 3.720352508500451, + "grad_norm": 1.7921321392059326, + "learning_rate": 2.5525977265826506e-06, + "loss": 0.364, + "step": 53614 + }, + { + "epoch": 3.7204218999375476, + "grad_norm": 1.2592476606369019, + "learning_rate": 2.5513364996110343e-06, + "loss": 0.2744, + "step": 53615 + }, + { + "epoch": 3.7204912913746444, + "grad_norm": 0.9475111961364746, + "learning_rate": 2.5500755802735475e-06, + "loss": 0.3004, + "step": 53616 + }, + { + "epoch": 3.7205606828117412, + "grad_norm": 1.012007236480713, + "learning_rate": 2.548814968574165e-06, + "loss": 0.2856, + "step": 53617 + }, + { + "epoch": 3.7206300742488376, + "grad_norm": 0.7606760263442993, + "learning_rate": 2.54755466451686e-06, + "loss": 0.2832, + "step": 53618 + }, + { + "epoch": 3.7206994656859345, + "grad_norm": 0.8741509318351746, + "learning_rate": 2.5462946681056197e-06, + "loss": 0.2692, + "step": 53619 + }, + { + "epoch": 3.720768857123031, + "grad_norm": 1.041600227355957, + "learning_rate": 2.5450349793444186e-06, + "loss": 0.2259, + "step": 53620 + }, + { + "epoch": 3.7208382485601277, + "grad_norm": 0.8742949962615967, + "learning_rate": 2.5437755982372526e-06, + "loss": 0.2755, + "step": 53621 + }, + { + "epoch": 3.7209076399972245, + "grad_norm": 1.224568486213684, + "learning_rate": 2.5425165247880634e-06, + "loss": 0.3099, + "step": 53622 + }, + { + "epoch": 3.720977031434321, + "grad_norm": 0.6495367288589478, + "learning_rate": 2.5412577590008368e-06, + "loss": 0.1535, + "step": 53623 + }, + { + "epoch": 3.7210464228714177, + "grad_norm": 1.2276577949523926, + "learning_rate": 2.5399993008795584e-06, + "loss": 0.3355, + "step": 53624 + }, + { + "epoch": 3.721115814308514, + "grad_norm": 0.790425181388855, + "learning_rate": 2.538741150428192e-06, + "loss": 0.2851, + "step": 53625 + }, + { + "epoch": 3.721185205745611, + "grad_norm": 1.0305171012878418, + "learning_rate": 2.5374833076507343e-06, + "loss": 0.2015, + "step": 53626 + }, + { + "epoch": 3.721254597182708, + "grad_norm": 0.9808995723724365, + "learning_rate": 2.536225772551115e-06, + "loss": 0.2897, + "step": 53627 + }, + { + "epoch": 3.721323988619804, + "grad_norm": 0.9952573180198669, + "learning_rate": 2.5349685451333205e-06, + "loss": 0.2128, + "step": 53628 + }, + { + "epoch": 3.721393380056901, + "grad_norm": 0.7425248622894287, + "learning_rate": 2.5337116254013248e-06, + "loss": 0.2802, + "step": 53629 + }, + { + "epoch": 3.7214627714939974, + "grad_norm": 0.7369589805603027, + "learning_rate": 2.5324550133590916e-06, + "loss": 0.3101, + "step": 53630 + }, + { + "epoch": 3.7215321629310942, + "grad_norm": 0.9733442664146423, + "learning_rate": 2.531198709010585e-06, + "loss": 0.3076, + "step": 53631 + }, + { + "epoch": 3.721601554368191, + "grad_norm": 2.287482976913452, + "learning_rate": 2.5299427123597896e-06, + "loss": 0.2725, + "step": 53632 + }, + { + "epoch": 3.721670945805288, + "grad_norm": 0.9211487770080566, + "learning_rate": 2.5286870234106586e-06, + "loss": 0.2389, + "step": 53633 + }, + { + "epoch": 3.7217403372423843, + "grad_norm": 1.5817266702651978, + "learning_rate": 2.527431642167144e-06, + "loss": 0.3955, + "step": 53634 + }, + { + "epoch": 3.7218097286794807, + "grad_norm": 1.6395739316940308, + "learning_rate": 2.5261765686332094e-06, + "loss": 0.2614, + "step": 53635 + }, + { + "epoch": 3.7218791201165775, + "grad_norm": 0.4571129381656647, + "learning_rate": 2.5249218028128406e-06, + "loss": 0.0923, + "step": 53636 + }, + { + "epoch": 3.7219485115536743, + "grad_norm": 0.8847830891609192, + "learning_rate": 2.523667344709979e-06, + "loss": 0.2093, + "step": 53637 + }, + { + "epoch": 3.722017902990771, + "grad_norm": 0.7065118551254272, + "learning_rate": 2.5224131943285878e-06, + "loss": 0.1567, + "step": 53638 + }, + { + "epoch": 3.7220872944278676, + "grad_norm": 0.937961757183075, + "learning_rate": 2.5211593516726307e-06, + "loss": 0.1849, + "step": 53639 + }, + { + "epoch": 3.7221566858649644, + "grad_norm": 1.3868952989578247, + "learning_rate": 2.519905816746082e-06, + "loss": 0.3468, + "step": 53640 + }, + { + "epoch": 3.722226077302061, + "grad_norm": 0.904203474521637, + "learning_rate": 2.5186525895528614e-06, + "loss": 0.3029, + "step": 53641 + }, + { + "epoch": 3.7222954687391576, + "grad_norm": 1.0413578748703003, + "learning_rate": 2.5173996700969426e-06, + "loss": 0.2661, + "step": 53642 + }, + { + "epoch": 3.7223648601762545, + "grad_norm": 0.9971304535865784, + "learning_rate": 2.51614705838229e-06, + "loss": 0.2238, + "step": 53643 + }, + { + "epoch": 3.722434251613351, + "grad_norm": 1.2631564140319824, + "learning_rate": 2.5148947544128444e-06, + "loss": 0.2315, + "step": 53644 + }, + { + "epoch": 3.7225036430504477, + "grad_norm": 0.8376457691192627, + "learning_rate": 2.5136427581925694e-06, + "loss": 0.2081, + "step": 53645 + }, + { + "epoch": 3.722573034487544, + "grad_norm": 1.0326545238494873, + "learning_rate": 2.512391069725417e-06, + "loss": 0.3572, + "step": 53646 + }, + { + "epoch": 3.722642425924641, + "grad_norm": 0.5300071835517883, + "learning_rate": 2.5111396890153405e-06, + "loss": 0.2063, + "step": 53647 + }, + { + "epoch": 3.7227118173617377, + "grad_norm": 1.0941663980484009, + "learning_rate": 2.5098886160662804e-06, + "loss": 0.237, + "step": 53648 + }, + { + "epoch": 3.722781208798834, + "grad_norm": 1.0546809434890747, + "learning_rate": 2.5086378508821893e-06, + "loss": 0.2102, + "step": 53649 + }, + { + "epoch": 3.722850600235931, + "grad_norm": 0.6984122395515442, + "learning_rate": 2.507387393467009e-06, + "loss": 0.1791, + "step": 53650 + }, + { + "epoch": 3.7229199916730273, + "grad_norm": 0.9529337882995605, + "learning_rate": 2.5061372438247023e-06, + "loss": 0.1817, + "step": 53651 + }, + { + "epoch": 3.722989383110124, + "grad_norm": 0.6303490996360779, + "learning_rate": 2.5048874019592215e-06, + "loss": 0.2431, + "step": 53652 + }, + { + "epoch": 3.723058774547221, + "grad_norm": 0.8673107624053955, + "learning_rate": 2.5036378678744864e-06, + "loss": 0.372, + "step": 53653 + }, + { + "epoch": 3.7231281659843174, + "grad_norm": 0.6538466811180115, + "learning_rate": 2.502388641574449e-06, + "loss": 0.2042, + "step": 53654 + }, + { + "epoch": 3.7231975574214142, + "grad_norm": 0.8234646916389465, + "learning_rate": 2.501139723063084e-06, + "loss": 0.2552, + "step": 53655 + }, + { + "epoch": 3.7232669488585106, + "grad_norm": 0.819368302822113, + "learning_rate": 2.499891112344288e-06, + "loss": 0.2268, + "step": 53656 + }, + { + "epoch": 3.7233363402956074, + "grad_norm": 0.7976694703102112, + "learning_rate": 2.4986428094220247e-06, + "loss": 0.2301, + "step": 53657 + }, + { + "epoch": 3.7234057317327043, + "grad_norm": 0.6895728707313538, + "learning_rate": 2.4973948143002355e-06, + "loss": 0.2279, + "step": 53658 + }, + { + "epoch": 3.723475123169801, + "grad_norm": 1.0015621185302734, + "learning_rate": 2.4961471269828506e-06, + "loss": 0.2365, + "step": 53659 + }, + { + "epoch": 3.7235445146068975, + "grad_norm": 0.9846001863479614, + "learning_rate": 2.4948997474738223e-06, + "loss": 0.2331, + "step": 53660 + }, + { + "epoch": 3.7236139060439943, + "grad_norm": 0.9933878183364868, + "learning_rate": 2.4936526757771027e-06, + "loss": 0.3159, + "step": 53661 + }, + { + "epoch": 3.7236832974810907, + "grad_norm": 1.024671196937561, + "learning_rate": 2.492405911896589e-06, + "loss": 0.1927, + "step": 53662 + }, + { + "epoch": 3.7237526889181876, + "grad_norm": 0.9526066780090332, + "learning_rate": 2.491159455836234e-06, + "loss": 0.2904, + "step": 53663 + }, + { + "epoch": 3.7238220803552844, + "grad_norm": 0.6394450664520264, + "learning_rate": 2.489913307599978e-06, + "loss": 0.1774, + "step": 53664 + }, + { + "epoch": 3.723891471792381, + "grad_norm": 0.8719181418418884, + "learning_rate": 2.488667467191752e-06, + "loss": 0.3316, + "step": 53665 + }, + { + "epoch": 3.7239608632294776, + "grad_norm": 0.8275408148765564, + "learning_rate": 2.487421934615497e-06, + "loss": 0.2202, + "step": 53666 + }, + { + "epoch": 3.724030254666574, + "grad_norm": 0.6713863611221313, + "learning_rate": 2.486176709875132e-06, + "loss": 0.1875, + "step": 53667 + }, + { + "epoch": 3.724099646103671, + "grad_norm": 1.217596173286438, + "learning_rate": 2.4849317929745985e-06, + "loss": 0.4732, + "step": 53668 + }, + { + "epoch": 3.7241690375407677, + "grad_norm": 1.1061464548110962, + "learning_rate": 2.4836871839178155e-06, + "loss": 0.3696, + "step": 53669 + }, + { + "epoch": 3.724238428977864, + "grad_norm": 0.8637010455131531, + "learning_rate": 2.4824428827087244e-06, + "loss": 0.2342, + "step": 53670 + }, + { + "epoch": 3.724307820414961, + "grad_norm": 0.7077743411064148, + "learning_rate": 2.481198889351244e-06, + "loss": 0.2099, + "step": 53671 + }, + { + "epoch": 3.7243772118520573, + "grad_norm": 0.6106438636779785, + "learning_rate": 2.4799552038492935e-06, + "loss": 0.2412, + "step": 53672 + }, + { + "epoch": 3.724446603289154, + "grad_norm": 0.9378376603126526, + "learning_rate": 2.4787118262068255e-06, + "loss": 0.2845, + "step": 53673 + }, + { + "epoch": 3.724515994726251, + "grad_norm": 1.2641758918762207, + "learning_rate": 2.4774687564277366e-06, + "loss": 0.2244, + "step": 53674 + }, + { + "epoch": 3.7245853861633473, + "grad_norm": 0.953283429145813, + "learning_rate": 2.476225994515979e-06, + "loss": 0.2643, + "step": 53675 + }, + { + "epoch": 3.724654777600444, + "grad_norm": 0.7409673929214478, + "learning_rate": 2.4749835404754616e-06, + "loss": 0.17, + "step": 53676 + }, + { + "epoch": 3.7247241690375406, + "grad_norm": 0.8757656812667847, + "learning_rate": 2.4737413943100917e-06, + "loss": 0.216, + "step": 53677 + }, + { + "epoch": 3.7247935604746374, + "grad_norm": 0.794505774974823, + "learning_rate": 2.4724995560238106e-06, + "loss": 0.268, + "step": 53678 + }, + { + "epoch": 3.724862951911734, + "grad_norm": 0.867813229560852, + "learning_rate": 2.4712580256205377e-06, + "loss": 0.2086, + "step": 53679 + }, + { + "epoch": 3.7249323433488306, + "grad_norm": 1.7056106328964233, + "learning_rate": 2.4700168031041805e-06, + "loss": 0.3205, + "step": 53680 + }, + { + "epoch": 3.7250017347859274, + "grad_norm": 1.167962670326233, + "learning_rate": 2.468775888478658e-06, + "loss": 0.26, + "step": 53681 + }, + { + "epoch": 3.725071126223024, + "grad_norm": 0.5396034717559814, + "learning_rate": 2.4675352817479234e-06, + "loss": 0.1334, + "step": 53682 + }, + { + "epoch": 3.7251405176601207, + "grad_norm": 0.6656848788261414, + "learning_rate": 2.4662949829158398e-06, + "loss": 0.1795, + "step": 53683 + }, + { + "epoch": 3.7252099090972175, + "grad_norm": 1.0716241598129272, + "learning_rate": 2.4650549919863485e-06, + "loss": 0.3684, + "step": 53684 + }, + { + "epoch": 3.7252793005343143, + "grad_norm": 0.7707404494285583, + "learning_rate": 2.4638153089633688e-06, + "loss": 0.2115, + "step": 53685 + }, + { + "epoch": 3.7253486919714107, + "grad_norm": 1.1053097248077393, + "learning_rate": 2.462575933850797e-06, + "loss": 0.3113, + "step": 53686 + }, + { + "epoch": 3.7254180834085076, + "grad_norm": 1.3978898525238037, + "learning_rate": 2.461336866652575e-06, + "loss": 0.2693, + "step": 53687 + }, + { + "epoch": 3.725487474845604, + "grad_norm": 0.7886305451393127, + "learning_rate": 2.460098107372588e-06, + "loss": 0.2575, + "step": 53688 + }, + { + "epoch": 3.7255568662827008, + "grad_norm": 0.9399693012237549, + "learning_rate": 2.458859656014756e-06, + "loss": 0.3201, + "step": 53689 + }, + { + "epoch": 3.7256262577197976, + "grad_norm": 1.0878485441207886, + "learning_rate": 2.457621512582986e-06, + "loss": 0.2624, + "step": 53690 + }, + { + "epoch": 3.725695649156894, + "grad_norm": 1.1007035970687866, + "learning_rate": 2.4563836770811867e-06, + "loss": 0.2812, + "step": 53691 + }, + { + "epoch": 3.725765040593991, + "grad_norm": 0.857801079750061, + "learning_rate": 2.4551461495132656e-06, + "loss": 0.2341, + "step": 53692 + }, + { + "epoch": 3.725834432031087, + "grad_norm": 0.820700466632843, + "learning_rate": 2.453908929883131e-06, + "loss": 0.2368, + "step": 53693 + }, + { + "epoch": 3.725903823468184, + "grad_norm": 1.2755382061004639, + "learning_rate": 2.4526720181946793e-06, + "loss": 0.2359, + "step": 53694 + }, + { + "epoch": 3.725973214905281, + "grad_norm": 1.235496163368225, + "learning_rate": 2.4514354144518305e-06, + "loss": 0.2914, + "step": 53695 + }, + { + "epoch": 3.7260426063423773, + "grad_norm": 1.0564875602722168, + "learning_rate": 2.4501991186585027e-06, + "loss": 0.2679, + "step": 53696 + }, + { + "epoch": 3.726111997779474, + "grad_norm": 0.9527149796485901, + "learning_rate": 2.44896313081856e-06, + "loss": 0.3769, + "step": 53697 + }, + { + "epoch": 3.7261813892165705, + "grad_norm": 0.9473699927330017, + "learning_rate": 2.4477274509359216e-06, + "loss": 0.1791, + "step": 53698 + }, + { + "epoch": 3.7262507806536673, + "grad_norm": 0.9166672229766846, + "learning_rate": 2.4464920790144842e-06, + "loss": 0.1182, + "step": 53699 + }, + { + "epoch": 3.726320172090764, + "grad_norm": 0.7266594767570496, + "learning_rate": 2.4452570150581668e-06, + "loss": 0.1419, + "step": 53700 + }, + { + "epoch": 3.7263895635278605, + "grad_norm": 0.6665005087852478, + "learning_rate": 2.444022259070844e-06, + "loss": 0.1636, + "step": 53701 + }, + { + "epoch": 3.7264589549649574, + "grad_norm": 0.8450223803520203, + "learning_rate": 2.442787811056435e-06, + "loss": 0.2306, + "step": 53702 + }, + { + "epoch": 3.7265283464020538, + "grad_norm": 1.1366702318191528, + "learning_rate": 2.4415536710188145e-06, + "loss": 0.2444, + "step": 53703 + }, + { + "epoch": 3.7265977378391506, + "grad_norm": 0.6926429867744446, + "learning_rate": 2.440319838961902e-06, + "loss": 0.2975, + "step": 53704 + }, + { + "epoch": 3.7266671292762474, + "grad_norm": 1.0315515995025635, + "learning_rate": 2.4390863148895716e-06, + "loss": 0.322, + "step": 53705 + }, + { + "epoch": 3.726736520713344, + "grad_norm": 0.7044632434844971, + "learning_rate": 2.43785309880572e-06, + "loss": 0.182, + "step": 53706 + }, + { + "epoch": 3.7268059121504407, + "grad_norm": 0.6568784713745117, + "learning_rate": 2.436620190714256e-06, + "loss": 0.2513, + "step": 53707 + }, + { + "epoch": 3.726875303587537, + "grad_norm": 0.7985777854919434, + "learning_rate": 2.4353875906190647e-06, + "loss": 0.1235, + "step": 53708 + }, + { + "epoch": 3.726944695024634, + "grad_norm": 0.88540118932724, + "learning_rate": 2.4341552985240434e-06, + "loss": 0.2311, + "step": 53709 + }, + { + "epoch": 3.7270140864617307, + "grad_norm": 0.8476812243461609, + "learning_rate": 2.4329233144330554e-06, + "loss": 0.1719, + "step": 53710 + }, + { + "epoch": 3.7270834778988275, + "grad_norm": 1.2191118001937866, + "learning_rate": 2.4316916383500197e-06, + "loss": 0.2583, + "step": 53711 + }, + { + "epoch": 3.727152869335924, + "grad_norm": 0.8092365264892578, + "learning_rate": 2.4304602702788114e-06, + "loss": 0.2705, + "step": 53712 + }, + { + "epoch": 3.7272222607730208, + "grad_norm": 0.8279839158058167, + "learning_rate": 2.4292292102233163e-06, + "loss": 0.2927, + "step": 53713 + }, + { + "epoch": 3.727291652210117, + "grad_norm": 0.9251711964607239, + "learning_rate": 2.4279984581874306e-06, + "loss": 0.2915, + "step": 53714 + }, + { + "epoch": 3.727361043647214, + "grad_norm": 1.1471425294876099, + "learning_rate": 2.426768014175029e-06, + "loss": 0.2356, + "step": 53715 + }, + { + "epoch": 3.727430435084311, + "grad_norm": 1.1998014450073242, + "learning_rate": 2.4255378781900097e-06, + "loss": 0.3279, + "step": 53716 + }, + { + "epoch": 3.727499826521407, + "grad_norm": 0.7485150694847107, + "learning_rate": 2.424308050236235e-06, + "loss": 0.2962, + "step": 53717 + }, + { + "epoch": 3.727569217958504, + "grad_norm": 0.7716354727745056, + "learning_rate": 2.4230785303176018e-06, + "loss": 0.1572, + "step": 53718 + }, + { + "epoch": 3.7276386093956004, + "grad_norm": 0.8117040991783142, + "learning_rate": 2.4218493184379853e-06, + "loss": 0.2067, + "step": 53719 + }, + { + "epoch": 3.7277080008326973, + "grad_norm": 1.0024237632751465, + "learning_rate": 2.4206204146012713e-06, + "loss": 0.3574, + "step": 53720 + }, + { + "epoch": 3.727777392269794, + "grad_norm": 0.8261936902999878, + "learning_rate": 2.419391818811334e-06, + "loss": 0.2217, + "step": 53721 + }, + { + "epoch": 3.7278467837068905, + "grad_norm": 1.560671329498291, + "learning_rate": 2.4181635310720706e-06, + "loss": 0.2307, + "step": 53722 + }, + { + "epoch": 3.7279161751439873, + "grad_norm": 1.5849719047546387, + "learning_rate": 2.4169355513873338e-06, + "loss": 0.2926, + "step": 53723 + }, + { + "epoch": 3.7279855665810837, + "grad_norm": 0.7994822859764099, + "learning_rate": 2.415707879761009e-06, + "loss": 0.2057, + "step": 53724 + }, + { + "epoch": 3.7280549580181805, + "grad_norm": 1.009108066558838, + "learning_rate": 2.4144805161969817e-06, + "loss": 0.2287, + "step": 53725 + }, + { + "epoch": 3.7281243494552774, + "grad_norm": 0.9254602193832397, + "learning_rate": 2.4132534606991054e-06, + "loss": 0.1594, + "step": 53726 + }, + { + "epoch": 3.7281937408923738, + "grad_norm": 0.662730872631073, + "learning_rate": 2.412026713271265e-06, + "loss": 0.2198, + "step": 53727 + }, + { + "epoch": 3.7282631323294706, + "grad_norm": 1.1814321279525757, + "learning_rate": 2.4108002739173353e-06, + "loss": 0.2983, + "step": 53728 + }, + { + "epoch": 3.728332523766567, + "grad_norm": 0.670569121837616, + "learning_rate": 2.409574142641191e-06, + "loss": 0.2521, + "step": 53729 + }, + { + "epoch": 3.728401915203664, + "grad_norm": 0.6928852200508118, + "learning_rate": 2.408348319446696e-06, + "loss": 0.2127, + "step": 53730 + }, + { + "epoch": 3.7284713066407607, + "grad_norm": 0.9351518750190735, + "learning_rate": 2.4071228043377247e-06, + "loss": 0.1853, + "step": 53731 + }, + { + "epoch": 3.728540698077857, + "grad_norm": 0.8124568462371826, + "learning_rate": 2.4058975973181407e-06, + "loss": 0.1752, + "step": 53732 + }, + { + "epoch": 3.728610089514954, + "grad_norm": 0.8701272010803223, + "learning_rate": 2.4046726983918187e-06, + "loss": 0.3329, + "step": 53733 + }, + { + "epoch": 3.7286794809520503, + "grad_norm": 0.9254705905914307, + "learning_rate": 2.4034481075626114e-06, + "loss": 0.187, + "step": 53734 + }, + { + "epoch": 3.728748872389147, + "grad_norm": 0.9558954238891602, + "learning_rate": 2.402223824834404e-06, + "loss": 0.1829, + "step": 53735 + }, + { + "epoch": 3.728818263826244, + "grad_norm": 0.6938510537147522, + "learning_rate": 2.4009998502110498e-06, + "loss": 0.1769, + "step": 53736 + }, + { + "epoch": 3.7288876552633408, + "grad_norm": 0.5982946157455444, + "learning_rate": 2.399776183696434e-06, + "loss": 0.2225, + "step": 53737 + }, + { + "epoch": 3.728957046700437, + "grad_norm": 0.8558070063591003, + "learning_rate": 2.398552825294376e-06, + "loss": 0.2747, + "step": 53738 + }, + { + "epoch": 3.729026438137534, + "grad_norm": 0.9423138499259949, + "learning_rate": 2.397329775008772e-06, + "loss": 0.2914, + "step": 53739 + }, + { + "epoch": 3.7290958295746304, + "grad_norm": 0.7664811611175537, + "learning_rate": 2.3961070328434753e-06, + "loss": 0.1958, + "step": 53740 + }, + { + "epoch": 3.729165221011727, + "grad_norm": 0.9206449389457703, + "learning_rate": 2.3948845988023384e-06, + "loss": 0.2646, + "step": 53741 + }, + { + "epoch": 3.729234612448824, + "grad_norm": 0.6367307305335999, + "learning_rate": 2.3936624728892244e-06, + "loss": 0.1653, + "step": 53742 + }, + { + "epoch": 3.7293040038859204, + "grad_norm": 1.1276111602783203, + "learning_rate": 2.392440655108008e-06, + "loss": 0.2683, + "step": 53743 + }, + { + "epoch": 3.7293733953230173, + "grad_norm": 0.8219624161720276, + "learning_rate": 2.391219145462531e-06, + "loss": 0.2122, + "step": 53744 + }, + { + "epoch": 3.7294427867601136, + "grad_norm": 0.7939244508743286, + "learning_rate": 2.3899979439566346e-06, + "loss": 0.3129, + "step": 53745 + }, + { + "epoch": 3.7295121781972105, + "grad_norm": 1.221096158027649, + "learning_rate": 2.388777050594204e-06, + "loss": 0.2831, + "step": 53746 + }, + { + "epoch": 3.7295815696343073, + "grad_norm": 0.7519833445549011, + "learning_rate": 2.3875564653790707e-06, + "loss": 0.3987, + "step": 53747 + }, + { + "epoch": 3.7296509610714037, + "grad_norm": 0.84687739610672, + "learning_rate": 2.386336188315097e-06, + "loss": 0.3106, + "step": 53748 + }, + { + "epoch": 3.7297203525085005, + "grad_norm": 0.9555209875106812, + "learning_rate": 2.385116219406136e-06, + "loss": 0.2747, + "step": 53749 + }, + { + "epoch": 3.729789743945597, + "grad_norm": 1.0116318464279175, + "learning_rate": 2.3838965586560403e-06, + "loss": 0.2403, + "step": 53750 + }, + { + "epoch": 3.7298591353826938, + "grad_norm": 0.8619539737701416, + "learning_rate": 2.382677206068662e-06, + "loss": 0.2208, + "step": 53751 + }, + { + "epoch": 3.7299285268197906, + "grad_norm": 0.7924817800521851, + "learning_rate": 2.3814581616478427e-06, + "loss": 0.2393, + "step": 53752 + }, + { + "epoch": 3.729997918256887, + "grad_norm": 0.8157413005828857, + "learning_rate": 2.3802394253974235e-06, + "loss": 0.2585, + "step": 53753 + }, + { + "epoch": 3.730067309693984, + "grad_norm": 0.83403480052948, + "learning_rate": 2.3790209973212684e-06, + "loss": 0.222, + "step": 53754 + }, + { + "epoch": 3.73013670113108, + "grad_norm": 0.852594792842865, + "learning_rate": 2.3778028774232184e-06, + "loss": 0.2123, + "step": 53755 + }, + { + "epoch": 3.730206092568177, + "grad_norm": 0.6546127200126648, + "learning_rate": 2.3765850657071266e-06, + "loss": 0.1908, + "step": 53756 + }, + { + "epoch": 3.730275484005274, + "grad_norm": 0.9949448704719543, + "learning_rate": 2.3753675621768224e-06, + "loss": 0.1911, + "step": 53757 + }, + { + "epoch": 3.7303448754423703, + "grad_norm": 1.830954670906067, + "learning_rate": 2.3741503668361585e-06, + "loss": 0.1938, + "step": 53758 + }, + { + "epoch": 3.730414266879467, + "grad_norm": 1.0046197175979614, + "learning_rate": 2.3729334796889768e-06, + "loss": 0.3658, + "step": 53759 + }, + { + "epoch": 3.7304836583165635, + "grad_norm": 0.6363102197647095, + "learning_rate": 2.3717169007391184e-06, + "loss": 0.1782, + "step": 53760 + }, + { + "epoch": 3.7305530497536603, + "grad_norm": 0.7556812763214111, + "learning_rate": 2.3705006299904132e-06, + "loss": 0.4156, + "step": 53761 + }, + { + "epoch": 3.730622441190757, + "grad_norm": 1.3346670866012573, + "learning_rate": 2.369284667446714e-06, + "loss": 0.4181, + "step": 53762 + }, + { + "epoch": 3.730691832627854, + "grad_norm": 0.798835277557373, + "learning_rate": 2.368069013111862e-06, + "loss": 0.269, + "step": 53763 + }, + { + "epoch": 3.7307612240649504, + "grad_norm": 0.9420284628868103, + "learning_rate": 2.3668536669896767e-06, + "loss": 0.2614, + "step": 53764 + }, + { + "epoch": 3.730830615502047, + "grad_norm": 0.9495645761489868, + "learning_rate": 2.3656386290840215e-06, + "loss": 0.3223, + "step": 53765 + }, + { + "epoch": 3.7309000069391436, + "grad_norm": 0.7697654962539673, + "learning_rate": 2.3644238993987155e-06, + "loss": 0.3236, + "step": 53766 + }, + { + "epoch": 3.7309693983762404, + "grad_norm": 0.894429624080658, + "learning_rate": 2.363209477937589e-06, + "loss": 0.2872, + "step": 53767 + }, + { + "epoch": 3.7310387898133373, + "grad_norm": 0.7664215564727783, + "learning_rate": 2.3619953647044834e-06, + "loss": 0.172, + "step": 53768 + }, + { + "epoch": 3.7311081812504336, + "grad_norm": 1.0610830783843994, + "learning_rate": 2.360781559703229e-06, + "loss": 0.2364, + "step": 53769 + }, + { + "epoch": 3.7311775726875305, + "grad_norm": 0.7067539095878601, + "learning_rate": 2.359568062937656e-06, + "loss": 0.2298, + "step": 53770 + }, + { + "epoch": 3.731246964124627, + "grad_norm": 0.9076757431030273, + "learning_rate": 2.3583548744115946e-06, + "loss": 0.1901, + "step": 53771 + }, + { + "epoch": 3.7313163555617237, + "grad_norm": 0.7698354721069336, + "learning_rate": 2.3571419941288976e-06, + "loss": 0.1794, + "step": 53772 + }, + { + "epoch": 3.7313857469988205, + "grad_norm": 1.0085662603378296, + "learning_rate": 2.3559294220933616e-06, + "loss": 0.2495, + "step": 53773 + }, + { + "epoch": 3.731455138435917, + "grad_norm": 1.0483119487762451, + "learning_rate": 2.354717158308828e-06, + "loss": 0.2318, + "step": 53774 + }, + { + "epoch": 3.7315245298730138, + "grad_norm": 0.9763997793197632, + "learning_rate": 2.353505202779127e-06, + "loss": 0.2927, + "step": 53775 + }, + { + "epoch": 3.73159392131011, + "grad_norm": 0.6987702250480652, + "learning_rate": 2.3522935555080782e-06, + "loss": 0.1581, + "step": 53776 + }, + { + "epoch": 3.731663312747207, + "grad_norm": 0.6820730566978455, + "learning_rate": 2.3510822164995005e-06, + "loss": 0.2286, + "step": 53777 + }, + { + "epoch": 3.731732704184304, + "grad_norm": 1.462743878364563, + "learning_rate": 2.3498711857572463e-06, + "loss": 0.3832, + "step": 53778 + }, + { + "epoch": 3.7318020956214, + "grad_norm": 1.1131385564804077, + "learning_rate": 2.348660463285113e-06, + "loss": 0.1494, + "step": 53779 + }, + { + "epoch": 3.731871487058497, + "grad_norm": 1.093126893043518, + "learning_rate": 2.347450049086919e-06, + "loss": 0.2258, + "step": 53780 + }, + { + "epoch": 3.7319408784955934, + "grad_norm": 1.3216840028762817, + "learning_rate": 2.346239943166506e-06, + "loss": 0.2452, + "step": 53781 + }, + { + "epoch": 3.7320102699326902, + "grad_norm": 0.6151869893074036, + "learning_rate": 2.3450301455276714e-06, + "loss": 0.239, + "step": 53782 + }, + { + "epoch": 3.732079661369787, + "grad_norm": 0.9077761173248291, + "learning_rate": 2.343820656174256e-06, + "loss": 0.2966, + "step": 53783 + }, + { + "epoch": 3.7321490528068835, + "grad_norm": 0.628440797328949, + "learning_rate": 2.342611475110068e-06, + "loss": 0.2099, + "step": 53784 + }, + { + "epoch": 3.7322184442439803, + "grad_norm": 1.6524052619934082, + "learning_rate": 2.3414026023389267e-06, + "loss": 0.3266, + "step": 53785 + }, + { + "epoch": 3.7322878356810767, + "grad_norm": 1.4839781522750854, + "learning_rate": 2.34019403786464e-06, + "loss": 0.2979, + "step": 53786 + }, + { + "epoch": 3.7323572271181735, + "grad_norm": 0.6987505555152893, + "learning_rate": 2.3389857816910387e-06, + "loss": 0.1859, + "step": 53787 + }, + { + "epoch": 3.7324266185552704, + "grad_norm": 1.1723968982696533, + "learning_rate": 2.337777833821919e-06, + "loss": 0.2851, + "step": 53788 + }, + { + "epoch": 3.732496009992367, + "grad_norm": 0.8904389142990112, + "learning_rate": 2.3365701942611118e-06, + "loss": 0.1692, + "step": 53789 + }, + { + "epoch": 3.7325654014294636, + "grad_norm": 0.7282647490501404, + "learning_rate": 2.3353628630124135e-06, + "loss": 0.2002, + "step": 53790 + }, + { + "epoch": 3.7326347928665604, + "grad_norm": 0.9125795960426331, + "learning_rate": 2.334155840079655e-06, + "loss": 0.2743, + "step": 53791 + }, + { + "epoch": 3.732704184303657, + "grad_norm": 0.7938042283058167, + "learning_rate": 2.3329491254666325e-06, + "loss": 0.1933, + "step": 53792 + }, + { + "epoch": 3.7327735757407536, + "grad_norm": 0.7042882442474365, + "learning_rate": 2.3317427191771545e-06, + "loss": 0.1503, + "step": 53793 + }, + { + "epoch": 3.7328429671778505, + "grad_norm": 0.7966794967651367, + "learning_rate": 2.330536621215029e-06, + "loss": 0.1806, + "step": 53794 + }, + { + "epoch": 3.732912358614947, + "grad_norm": 0.8678264617919922, + "learning_rate": 2.329330831584076e-06, + "loss": 0.3021, + "step": 53795 + }, + { + "epoch": 3.7329817500520437, + "grad_norm": 0.9687070250511169, + "learning_rate": 2.3281253502880908e-06, + "loss": 0.2057, + "step": 53796 + }, + { + "epoch": 3.73305114148914, + "grad_norm": 0.9453646540641785, + "learning_rate": 2.3269201773308823e-06, + "loss": 0.2238, + "step": 53797 + }, + { + "epoch": 3.733120532926237, + "grad_norm": 0.9389950037002563, + "learning_rate": 2.3257153127162478e-06, + "loss": 0.2673, + "step": 53798 + }, + { + "epoch": 3.7331899243633337, + "grad_norm": 0.9985119104385376, + "learning_rate": 2.3245107564480063e-06, + "loss": 0.1973, + "step": 53799 + }, + { + "epoch": 3.73325931580043, + "grad_norm": 0.8892092704772949, + "learning_rate": 2.3233065085299544e-06, + "loss": 0.2952, + "step": 53800 + }, + { + "epoch": 3.733328707237527, + "grad_norm": 1.203202724456787, + "learning_rate": 2.322102568965878e-06, + "loss": 0.3911, + "step": 53801 + }, + { + "epoch": 3.7333980986746234, + "grad_norm": 0.7172217965126038, + "learning_rate": 2.320898937759597e-06, + "loss": 0.2086, + "step": 53802 + }, + { + "epoch": 3.73346749011172, + "grad_norm": 0.9544864892959595, + "learning_rate": 2.3196956149149074e-06, + "loss": 0.2849, + "step": 53803 + }, + { + "epoch": 3.733536881548817, + "grad_norm": 0.7725539803504944, + "learning_rate": 2.3184926004355958e-06, + "loss": 0.188, + "step": 53804 + }, + { + "epoch": 3.7336062729859134, + "grad_norm": 1.0912420749664307, + "learning_rate": 2.3172898943254807e-06, + "loss": 0.2877, + "step": 53805 + }, + { + "epoch": 3.7336756644230102, + "grad_norm": 0.7281365990638733, + "learning_rate": 2.3160874965883594e-06, + "loss": 0.2361, + "step": 53806 + }, + { + "epoch": 3.7337450558601066, + "grad_norm": 1.1216607093811035, + "learning_rate": 2.3148854072279956e-06, + "loss": 0.2709, + "step": 53807 + }, + { + "epoch": 3.7338144472972035, + "grad_norm": 0.6415630578994751, + "learning_rate": 2.3136836262482087e-06, + "loss": 0.1459, + "step": 53808 + }, + { + "epoch": 3.7338838387343003, + "grad_norm": 0.876387894153595, + "learning_rate": 2.312482153652795e-06, + "loss": 0.267, + "step": 53809 + }, + { + "epoch": 3.7339532301713967, + "grad_norm": 0.9308969974517822, + "learning_rate": 2.311280989445541e-06, + "loss": 0.1655, + "step": 53810 + }, + { + "epoch": 3.7340226216084935, + "grad_norm": 0.853239119052887, + "learning_rate": 2.3100801336302323e-06, + "loss": 0.2348, + "step": 53811 + }, + { + "epoch": 3.73409201304559, + "grad_norm": 1.1390541791915894, + "learning_rate": 2.308879586210666e-06, + "loss": 0.2997, + "step": 53812 + }, + { + "epoch": 3.7341614044826867, + "grad_norm": 1.1382033824920654, + "learning_rate": 2.30767934719065e-06, + "loss": 0.301, + "step": 53813 + }, + { + "epoch": 3.7342307959197836, + "grad_norm": 1.2000247240066528, + "learning_rate": 2.3064794165739365e-06, + "loss": 0.2996, + "step": 53814 + }, + { + "epoch": 3.7343001873568804, + "grad_norm": 0.9938194751739502, + "learning_rate": 2.3052797943643457e-06, + "loss": 0.248, + "step": 53815 + }, + { + "epoch": 3.734369578793977, + "grad_norm": 0.7656276822090149, + "learning_rate": 2.30408048056564e-06, + "loss": 0.2236, + "step": 53816 + }, + { + "epoch": 3.7344389702310736, + "grad_norm": 1.2417606115341187, + "learning_rate": 2.3028814751816287e-06, + "loss": 0.3924, + "step": 53817 + }, + { + "epoch": 3.73450836166817, + "grad_norm": 0.908271849155426, + "learning_rate": 2.301682778216074e-06, + "loss": 0.1861, + "step": 53818 + }, + { + "epoch": 3.734577753105267, + "grad_norm": 1.169358253479004, + "learning_rate": 2.3004843896727744e-06, + "loss": 0.331, + "step": 53819 + }, + { + "epoch": 3.7346471445423637, + "grad_norm": 0.9135819673538208, + "learning_rate": 2.299286309555526e-06, + "loss": 0.2619, + "step": 53820 + }, + { + "epoch": 3.73471653597946, + "grad_norm": 0.8897545337677002, + "learning_rate": 2.2980885378680816e-06, + "loss": 0.2385, + "step": 53821 + }, + { + "epoch": 3.734785927416557, + "grad_norm": 0.9966946840286255, + "learning_rate": 2.2968910746142382e-06, + "loss": 0.3594, + "step": 53822 + }, + { + "epoch": 3.7348553188536533, + "grad_norm": 0.965549111366272, + "learning_rate": 2.29569391979777e-06, + "loss": 0.2956, + "step": 53823 + }, + { + "epoch": 3.73492471029075, + "grad_norm": 0.8188734650611877, + "learning_rate": 2.2944970734224637e-06, + "loss": 0.2534, + "step": 53824 + }, + { + "epoch": 3.734994101727847, + "grad_norm": 0.6152007579803467, + "learning_rate": 2.293300535492093e-06, + "loss": 0.1333, + "step": 53825 + }, + { + "epoch": 3.7350634931649433, + "grad_norm": 0.6674963235855103, + "learning_rate": 2.2921043060104454e-06, + "loss": 0.2065, + "step": 53826 + }, + { + "epoch": 3.73513288460204, + "grad_norm": 0.7754181623458862, + "learning_rate": 2.290908384981283e-06, + "loss": 0.1777, + "step": 53827 + }, + { + "epoch": 3.7352022760391366, + "grad_norm": 0.6399956345558167, + "learning_rate": 2.2897127724083813e-06, + "loss": 0.1853, + "step": 53828 + }, + { + "epoch": 3.7352716674762334, + "grad_norm": 0.586581289768219, + "learning_rate": 2.2885174682955256e-06, + "loss": 0.1181, + "step": 53829 + }, + { + "epoch": 3.7353410589133302, + "grad_norm": 0.8648563623428345, + "learning_rate": 2.2873224726464804e-06, + "loss": 0.2629, + "step": 53830 + }, + { + "epoch": 3.7354104503504266, + "grad_norm": 0.8028373718261719, + "learning_rate": 2.2861277854650197e-06, + "loss": 0.206, + "step": 53831 + }, + { + "epoch": 3.7354798417875235, + "grad_norm": 1.0064659118652344, + "learning_rate": 2.2849334067549187e-06, + "loss": 0.299, + "step": 53832 + }, + { + "epoch": 3.73554923322462, + "grad_norm": 2.060638904571533, + "learning_rate": 2.2837393365199522e-06, + "loss": 0.3239, + "step": 53833 + }, + { + "epoch": 3.7356186246617167, + "grad_norm": 0.6743695139884949, + "learning_rate": 2.2825455747638836e-06, + "loss": 0.1841, + "step": 53834 + }, + { + "epoch": 3.7356880160988135, + "grad_norm": 1.0585838556289673, + "learning_rate": 2.2813521214904767e-06, + "loss": 0.2572, + "step": 53835 + }, + { + "epoch": 3.7357574075359103, + "grad_norm": 0.8830171823501587, + "learning_rate": 2.280158976703495e-06, + "loss": 0.2746, + "step": 53836 + }, + { + "epoch": 3.7358267989730067, + "grad_norm": 0.9352030158042908, + "learning_rate": 2.278966140406724e-06, + "loss": 0.1808, + "step": 53837 + }, + { + "epoch": 3.735896190410103, + "grad_norm": 0.8154942989349365, + "learning_rate": 2.277773612603906e-06, + "loss": 0.1449, + "step": 53838 + }, + { + "epoch": 3.7359655818472, + "grad_norm": 0.8647306561470032, + "learning_rate": 2.276581393298838e-06, + "loss": 0.1769, + "step": 53839 + }, + { + "epoch": 3.736034973284297, + "grad_norm": 0.6964065432548523, + "learning_rate": 2.275389482495249e-06, + "loss": 0.2244, + "step": 53840 + }, + { + "epoch": 3.7361043647213936, + "grad_norm": 0.7915312051773071, + "learning_rate": 2.2741978801969264e-06, + "loss": 0.3, + "step": 53841 + }, + { + "epoch": 3.73617375615849, + "grad_norm": 0.7864819765090942, + "learning_rate": 2.2730065864076224e-06, + "loss": 0.2036, + "step": 53842 + }, + { + "epoch": 3.736243147595587, + "grad_norm": 0.6939536929130554, + "learning_rate": 2.2718156011310997e-06, + "loss": 0.1444, + "step": 53843 + }, + { + "epoch": 3.7363125390326832, + "grad_norm": 1.1632304191589355, + "learning_rate": 2.270624924371112e-06, + "loss": 0.2415, + "step": 53844 + }, + { + "epoch": 3.73638193046978, + "grad_norm": 1.0238193273544312, + "learning_rate": 2.2694345561314225e-06, + "loss": 0.1781, + "step": 53845 + }, + { + "epoch": 3.736451321906877, + "grad_norm": 0.7401649355888367, + "learning_rate": 2.2682444964157833e-06, + "loss": 0.2186, + "step": 53846 + }, + { + "epoch": 3.7365207133439733, + "grad_norm": 2.197633981704712, + "learning_rate": 2.267054745227959e-06, + "loss": 0.2733, + "step": 53847 + }, + { + "epoch": 3.73659010478107, + "grad_norm": 0.6734001040458679, + "learning_rate": 2.2658653025717234e-06, + "loss": 0.2305, + "step": 53848 + }, + { + "epoch": 3.7366594962181665, + "grad_norm": 0.8671436309814453, + "learning_rate": 2.2646761684507965e-06, + "loss": 0.2926, + "step": 53849 + }, + { + "epoch": 3.7367288876552633, + "grad_norm": 1.0407103300094604, + "learning_rate": 2.263487342868942e-06, + "loss": 0.2877, + "step": 53850 + }, + { + "epoch": 3.73679827909236, + "grad_norm": 1.0516291856765747, + "learning_rate": 2.2622988258299227e-06, + "loss": 0.3127, + "step": 53851 + }, + { + "epoch": 3.7368676705294566, + "grad_norm": 1.2526228427886963, + "learning_rate": 2.261110617337492e-06, + "loss": 0.3153, + "step": 53852 + }, + { + "epoch": 3.7369370619665534, + "grad_norm": 0.7475927472114563, + "learning_rate": 2.2599227173953906e-06, + "loss": 0.2573, + "step": 53853 + }, + { + "epoch": 3.73700645340365, + "grad_norm": 0.9814368486404419, + "learning_rate": 2.258735126007383e-06, + "loss": 0.2858, + "step": 53854 + }, + { + "epoch": 3.7370758448407466, + "grad_norm": 0.8350462913513184, + "learning_rate": 2.257547843177199e-06, + "loss": 0.1805, + "step": 53855 + }, + { + "epoch": 3.7371452362778435, + "grad_norm": 0.6917315125465393, + "learning_rate": 2.2563608689085915e-06, + "loss": 0.2663, + "step": 53856 + }, + { + "epoch": 3.73721462771494, + "grad_norm": 0.6886811852455139, + "learning_rate": 2.2551742032053125e-06, + "loss": 0.2188, + "step": 53857 + }, + { + "epoch": 3.7372840191520367, + "grad_norm": 0.8697521090507507, + "learning_rate": 2.253987846071115e-06, + "loss": 0.188, + "step": 53858 + }, + { + "epoch": 3.737353410589133, + "grad_norm": 1.0623503923416138, + "learning_rate": 2.2528017975097405e-06, + "loss": 0.3129, + "step": 53859 + }, + { + "epoch": 3.73742280202623, + "grad_norm": 1.1254925727844238, + "learning_rate": 2.251616057524919e-06, + "loss": 0.2093, + "step": 53860 + }, + { + "epoch": 3.7374921934633267, + "grad_norm": 0.7861533761024475, + "learning_rate": 2.2504306261204146e-06, + "loss": 0.2845, + "step": 53861 + }, + { + "epoch": 3.7375615849004236, + "grad_norm": 0.8559955954551697, + "learning_rate": 2.2492455032999572e-06, + "loss": 0.2428, + "step": 53862 + }, + { + "epoch": 3.73763097633752, + "grad_norm": 1.386825680732727, + "learning_rate": 2.2480606890672884e-06, + "loss": 0.316, + "step": 53863 + }, + { + "epoch": 3.7377003677746163, + "grad_norm": 0.5738962292671204, + "learning_rate": 2.2468761834261496e-06, + "loss": 0.2415, + "step": 53864 + }, + { + "epoch": 3.737769759211713, + "grad_norm": 1.201409935951233, + "learning_rate": 2.245691986380294e-06, + "loss": 0.2339, + "step": 53865 + }, + { + "epoch": 3.73783915064881, + "grad_norm": 0.8183301687240601, + "learning_rate": 2.244508097933429e-06, + "loss": 0.1913, + "step": 53866 + }, + { + "epoch": 3.737908542085907, + "grad_norm": 0.7974933981895447, + "learning_rate": 2.243324518089318e-06, + "loss": 0.1706, + "step": 53867 + }, + { + "epoch": 3.7379779335230032, + "grad_norm": 0.8591254949569702, + "learning_rate": 2.2421412468517035e-06, + "loss": 0.2382, + "step": 53868 + }, + { + "epoch": 3.7380473249601, + "grad_norm": 0.8316085338592529, + "learning_rate": 2.2409582842242927e-06, + "loss": 0.2057, + "step": 53869 + }, + { + "epoch": 3.7381167163971964, + "grad_norm": 0.8933539986610413, + "learning_rate": 2.2397756302108274e-06, + "loss": 0.2112, + "step": 53870 + }, + { + "epoch": 3.7381861078342933, + "grad_norm": 1.0832369327545166, + "learning_rate": 2.2385932848150492e-06, + "loss": 0.28, + "step": 53871 + }, + { + "epoch": 3.73825549927139, + "grad_norm": 0.8840938210487366, + "learning_rate": 2.2374112480406996e-06, + "loss": 0.2832, + "step": 53872 + }, + { + "epoch": 3.7383248907084865, + "grad_norm": 1.1135642528533936, + "learning_rate": 2.2362295198914974e-06, + "loss": 0.2872, + "step": 53873 + }, + { + "epoch": 3.7383942821455833, + "grad_norm": 1.0275585651397705, + "learning_rate": 2.235048100371173e-06, + "loss": 0.232, + "step": 53874 + }, + { + "epoch": 3.7384636735826797, + "grad_norm": 1.024527907371521, + "learning_rate": 2.2338669894834685e-06, + "loss": 0.1829, + "step": 53875 + }, + { + "epoch": 3.7385330650197766, + "grad_norm": 1.0187363624572754, + "learning_rate": 2.2326861872320915e-06, + "loss": 0.1508, + "step": 53876 + }, + { + "epoch": 3.7386024564568734, + "grad_norm": 0.6356583833694458, + "learning_rate": 2.2315056936207836e-06, + "loss": 0.2561, + "step": 53877 + }, + { + "epoch": 3.73867184789397, + "grad_norm": 0.7707359194755554, + "learning_rate": 2.230325508653264e-06, + "loss": 0.212, + "step": 53878 + }, + { + "epoch": 3.7387412393310666, + "grad_norm": 0.6299737691879272, + "learning_rate": 2.229145632333274e-06, + "loss": 0.1728, + "step": 53879 + }, + { + "epoch": 3.738810630768163, + "grad_norm": 0.7310993671417236, + "learning_rate": 2.2279660646645217e-06, + "loss": 0.1938, + "step": 53880 + }, + { + "epoch": 3.73888002220526, + "grad_norm": 0.7678991556167603, + "learning_rate": 2.226786805650738e-06, + "loss": 0.1819, + "step": 53881 + }, + { + "epoch": 3.7389494136423567, + "grad_norm": 0.7431215643882751, + "learning_rate": 2.2256078552956528e-06, + "loss": 0.2137, + "step": 53882 + }, + { + "epoch": 3.739018805079453, + "grad_norm": 0.8660544157028198, + "learning_rate": 2.224429213602963e-06, + "loss": 0.2624, + "step": 53883 + }, + { + "epoch": 3.73908819651655, + "grad_norm": 0.6576476693153381, + "learning_rate": 2.2232508805764217e-06, + "loss": 0.2386, + "step": 53884 + }, + { + "epoch": 3.7391575879536463, + "grad_norm": 0.637941837310791, + "learning_rate": 2.2220728562197255e-06, + "loss": 0.1558, + "step": 53885 + }, + { + "epoch": 3.739226979390743, + "grad_norm": 1.0147719383239746, + "learning_rate": 2.220895140536594e-06, + "loss": 0.1967, + "step": 53886 + }, + { + "epoch": 3.73929637082784, + "grad_norm": 0.8821101784706116, + "learning_rate": 2.219717733530757e-06, + "loss": 0.22, + "step": 53887 + }, + { + "epoch": 3.7393657622649368, + "grad_norm": 1.166361689567566, + "learning_rate": 2.2185406352059346e-06, + "loss": 0.2668, + "step": 53888 + }, + { + "epoch": 3.739435153702033, + "grad_norm": 0.7358620762825012, + "learning_rate": 2.2173638455658343e-06, + "loss": 0.2078, + "step": 53889 + }, + { + "epoch": 3.7395045451391296, + "grad_norm": 1.0911740064620972, + "learning_rate": 2.216187364614164e-06, + "loss": 0.3165, + "step": 53890 + }, + { + "epoch": 3.7395739365762264, + "grad_norm": 1.5396164655685425, + "learning_rate": 2.2150111923546433e-06, + "loss": 0.2672, + "step": 53891 + }, + { + "epoch": 3.739643328013323, + "grad_norm": 0.7621417045593262, + "learning_rate": 2.2138353287909917e-06, + "loss": 0.1194, + "step": 53892 + }, + { + "epoch": 3.73971271945042, + "grad_norm": 0.9374271035194397, + "learning_rate": 2.212659773926917e-06, + "loss": 0.2053, + "step": 53893 + }, + { + "epoch": 3.7397821108875164, + "grad_norm": 1.102323055267334, + "learning_rate": 2.211484527766128e-06, + "loss": 0.3707, + "step": 53894 + }, + { + "epoch": 3.7398515023246133, + "grad_norm": 1.1857500076293945, + "learning_rate": 2.2103095903123427e-06, + "loss": 0.3277, + "step": 53895 + }, + { + "epoch": 3.7399208937617097, + "grad_norm": 0.9177632331848145, + "learning_rate": 2.2091349615692593e-06, + "loss": 0.2542, + "step": 53896 + }, + { + "epoch": 3.7399902851988065, + "grad_norm": 0.8363624811172485, + "learning_rate": 2.207960641540596e-06, + "loss": 0.2942, + "step": 53897 + }, + { + "epoch": 3.7400596766359033, + "grad_norm": 0.8332200646400452, + "learning_rate": 2.206786630230051e-06, + "loss": 0.3095, + "step": 53898 + }, + { + "epoch": 3.7401290680729997, + "grad_norm": 0.8693105578422546, + "learning_rate": 2.205612927641343e-06, + "loss": 0.2315, + "step": 53899 + }, + { + "epoch": 3.7401984595100966, + "grad_norm": 1.0552815198898315, + "learning_rate": 2.2044395337781575e-06, + "loss": 0.3048, + "step": 53900 + }, + { + "epoch": 3.740267850947193, + "grad_norm": 1.0060232877731323, + "learning_rate": 2.2032664486442033e-06, + "loss": 0.2663, + "step": 53901 + }, + { + "epoch": 3.7403372423842898, + "grad_norm": 0.9353331923484802, + "learning_rate": 2.2020936722431996e-06, + "loss": 0.2322, + "step": 53902 + }, + { + "epoch": 3.7404066338213866, + "grad_norm": 1.004417061805725, + "learning_rate": 2.200921204578854e-06, + "loss": 0.2936, + "step": 53903 + }, + { + "epoch": 3.740476025258483, + "grad_norm": 0.7803023457527161, + "learning_rate": 2.199749045654831e-06, + "loss": 0.1977, + "step": 53904 + }, + { + "epoch": 3.74054541669558, + "grad_norm": 1.3829925060272217, + "learning_rate": 2.19857719547486e-06, + "loss": 0.3989, + "step": 53905 + }, + { + "epoch": 3.740614808132676, + "grad_norm": 0.6335625052452087, + "learning_rate": 2.197405654042639e-06, + "loss": 0.1625, + "step": 53906 + }, + { + "epoch": 3.740684199569773, + "grad_norm": 1.2393178939819336, + "learning_rate": 2.1962344213618534e-06, + "loss": 0.2806, + "step": 53907 + }, + { + "epoch": 3.74075359100687, + "grad_norm": 0.8353100419044495, + "learning_rate": 2.195063497436212e-06, + "loss": 0.229, + "step": 53908 + }, + { + "epoch": 3.7408229824439663, + "grad_norm": 1.198698878288269, + "learning_rate": 2.1938928822693993e-06, + "loss": 0.3126, + "step": 53909 + }, + { + "epoch": 3.740892373881063, + "grad_norm": 0.63909912109375, + "learning_rate": 2.1927225758651358e-06, + "loss": 0.1593, + "step": 53910 + }, + { + "epoch": 3.7409617653181595, + "grad_norm": 0.6811312437057495, + "learning_rate": 2.1915525782270964e-06, + "loss": 0.1791, + "step": 53911 + }, + { + "epoch": 3.7410311567552563, + "grad_norm": 1.0215325355529785, + "learning_rate": 2.190382889358966e-06, + "loss": 0.2156, + "step": 53912 + }, + { + "epoch": 3.741100548192353, + "grad_norm": 0.6156725287437439, + "learning_rate": 2.1892135092644538e-06, + "loss": 0.1622, + "step": 53913 + }, + { + "epoch": 3.74116993962945, + "grad_norm": 0.6662071943283081, + "learning_rate": 2.188044437947245e-06, + "loss": 0.1995, + "step": 53914 + }, + { + "epoch": 3.7412393310665464, + "grad_norm": 1.1331552267074585, + "learning_rate": 2.1868756754110377e-06, + "loss": 0.3051, + "step": 53915 + }, + { + "epoch": 3.741308722503643, + "grad_norm": 0.9102522134780884, + "learning_rate": 2.1857072216595053e-06, + "loss": 0.2689, + "step": 53916 + }, + { + "epoch": 3.7413781139407396, + "grad_norm": 0.945619523525238, + "learning_rate": 2.184539076696357e-06, + "loss": 0.1993, + "step": 53917 + }, + { + "epoch": 3.7414475053778364, + "grad_norm": 0.8373525738716125, + "learning_rate": 2.183371240525267e-06, + "loss": 0.2342, + "step": 53918 + }, + { + "epoch": 3.7415168968149333, + "grad_norm": 0.9562064409255981, + "learning_rate": 2.1822037131499217e-06, + "loss": 0.3015, + "step": 53919 + }, + { + "epoch": 3.7415862882520297, + "grad_norm": 0.7226595282554626, + "learning_rate": 2.181036494574018e-06, + "loss": 0.222, + "step": 53920 + }, + { + "epoch": 3.7416556796891265, + "grad_norm": 1.0701556205749512, + "learning_rate": 2.1798695848012308e-06, + "loss": 0.2199, + "step": 53921 + }, + { + "epoch": 3.741725071126223, + "grad_norm": 0.8804560303688049, + "learning_rate": 2.1787029838352345e-06, + "loss": 0.2464, + "step": 53922 + }, + { + "epoch": 3.7417944625633197, + "grad_norm": 1.025713324546814, + "learning_rate": 2.177536691679738e-06, + "loss": 0.2536, + "step": 53923 + }, + { + "epoch": 3.7418638540004165, + "grad_norm": 0.9759231805801392, + "learning_rate": 2.1763707083384046e-06, + "loss": 0.3062, + "step": 53924 + }, + { + "epoch": 3.741933245437513, + "grad_norm": 1.3138892650604248, + "learning_rate": 2.17520503381492e-06, + "loss": 0.4841, + "step": 53925 + }, + { + "epoch": 3.7420026368746098, + "grad_norm": 1.119925618171692, + "learning_rate": 2.1740396681129592e-06, + "loss": 0.2896, + "step": 53926 + }, + { + "epoch": 3.742072028311706, + "grad_norm": 1.0598561763763428, + "learning_rate": 2.1728746112362086e-06, + "loss": 0.2847, + "step": 53927 + }, + { + "epoch": 3.742141419748803, + "grad_norm": 0.8531886339187622, + "learning_rate": 2.1717098631883316e-06, + "loss": 0.2662, + "step": 53928 + }, + { + "epoch": 3.7422108111859, + "grad_norm": 1.0658907890319824, + "learning_rate": 2.170545423973025e-06, + "loss": 0.3359, + "step": 53929 + }, + { + "epoch": 3.742280202622996, + "grad_norm": 0.8244253396987915, + "learning_rate": 2.169381293593975e-06, + "loss": 0.2381, + "step": 53930 + }, + { + "epoch": 3.742349594060093, + "grad_norm": 1.2256152629852295, + "learning_rate": 2.168217472054812e-06, + "loss": 0.3481, + "step": 53931 + }, + { + "epoch": 3.7424189854971894, + "grad_norm": 0.9609412550926208, + "learning_rate": 2.1670539593592444e-06, + "loss": 0.2563, + "step": 53932 + }, + { + "epoch": 3.7424883769342863, + "grad_norm": 0.7684838771820068, + "learning_rate": 2.1658907555109353e-06, + "loss": 0.2263, + "step": 53933 + }, + { + "epoch": 3.742557768371383, + "grad_norm": 0.8311857581138611, + "learning_rate": 2.1647278605135714e-06, + "loss": 0.1715, + "step": 53934 + }, + { + "epoch": 3.7426271598084795, + "grad_norm": 0.6558826565742493, + "learning_rate": 2.1635652743707933e-06, + "loss": 0.1894, + "step": 53935 + }, + { + "epoch": 3.7426965512455763, + "grad_norm": 0.897697925567627, + "learning_rate": 2.162402997086288e-06, + "loss": 0.3798, + "step": 53936 + }, + { + "epoch": 3.7427659426826727, + "grad_norm": 1.0362656116485596, + "learning_rate": 2.16124102866373e-06, + "loss": 0.2828, + "step": 53937 + }, + { + "epoch": 3.7428353341197695, + "grad_norm": 0.9193057417869568, + "learning_rate": 2.1600793691067934e-06, + "loss": 0.189, + "step": 53938 + }, + { + "epoch": 3.7429047255568664, + "grad_norm": 0.751466691493988, + "learning_rate": 2.1589180184191205e-06, + "loss": 0.1853, + "step": 53939 + }, + { + "epoch": 3.742974116993963, + "grad_norm": 0.765242874622345, + "learning_rate": 2.157756976604397e-06, + "loss": 0.239, + "step": 53940 + }, + { + "epoch": 3.7430435084310596, + "grad_norm": 1.114770531654358, + "learning_rate": 2.156596243666287e-06, + "loss": 0.1651, + "step": 53941 + }, + { + "epoch": 3.7431128998681564, + "grad_norm": 0.8143153786659241, + "learning_rate": 2.155435819608442e-06, + "loss": 0.2178, + "step": 53942 + }, + { + "epoch": 3.743182291305253, + "grad_norm": 1.1024316549301147, + "learning_rate": 2.1542757044345387e-06, + "loss": 0.2356, + "step": 53943 + }, + { + "epoch": 3.7432516827423497, + "grad_norm": 1.2022936344146729, + "learning_rate": 2.1531158981482393e-06, + "loss": 0.3908, + "step": 53944 + }, + { + "epoch": 3.7433210741794465, + "grad_norm": 1.1166179180145264, + "learning_rate": 2.1519564007531967e-06, + "loss": 0.232, + "step": 53945 + }, + { + "epoch": 3.743390465616543, + "grad_norm": 0.7069976925849915, + "learning_rate": 2.150797212253064e-06, + "loss": 0.1721, + "step": 53946 + }, + { + "epoch": 3.7434598570536397, + "grad_norm": 1.0499297380447388, + "learning_rate": 2.149638332651527e-06, + "loss": 0.3591, + "step": 53947 + }, + { + "epoch": 3.743529248490736, + "grad_norm": 0.906865656375885, + "learning_rate": 2.148479761952216e-06, + "loss": 0.2659, + "step": 53948 + }, + { + "epoch": 3.743598639927833, + "grad_norm": 1.592523455619812, + "learning_rate": 2.1473215001588164e-06, + "loss": 0.2141, + "step": 53949 + }, + { + "epoch": 3.7436680313649298, + "grad_norm": 0.7624203562736511, + "learning_rate": 2.14616354727496e-06, + "loss": 0.2364, + "step": 53950 + }, + { + "epoch": 3.743737422802026, + "grad_norm": 1.098612904548645, + "learning_rate": 2.14500590330432e-06, + "loss": 0.3273, + "step": 53951 + }, + { + "epoch": 3.743806814239123, + "grad_norm": 1.03670072555542, + "learning_rate": 2.1438485682505393e-06, + "loss": 0.2675, + "step": 53952 + }, + { + "epoch": 3.7438762056762194, + "grad_norm": 1.3025171756744385, + "learning_rate": 2.14269154211727e-06, + "loss": 0.3242, + "step": 53953 + }, + { + "epoch": 3.743945597113316, + "grad_norm": 1.2132056951522827, + "learning_rate": 2.141534824908176e-06, + "loss": 0.2289, + "step": 53954 + }, + { + "epoch": 3.744014988550413, + "grad_norm": 1.0199168920516968, + "learning_rate": 2.140378416626898e-06, + "loss": 0.3095, + "step": 53955 + }, + { + "epoch": 3.7440843799875094, + "grad_norm": 0.8134964108467102, + "learning_rate": 2.1392223172770896e-06, + "loss": 0.2025, + "step": 53956 + }, + { + "epoch": 3.7441537714246063, + "grad_norm": 0.9803938865661621, + "learning_rate": 2.1380665268624035e-06, + "loss": 0.2969, + "step": 53957 + }, + { + "epoch": 3.7442231628617026, + "grad_norm": 0.8845942616462708, + "learning_rate": 2.136911045386503e-06, + "loss": 0.2099, + "step": 53958 + }, + { + "epoch": 3.7442925542987995, + "grad_norm": 1.2832387685775757, + "learning_rate": 2.135755872853007e-06, + "loss": 0.3373, + "step": 53959 + }, + { + "epoch": 3.7443619457358963, + "grad_norm": 0.8800534605979919, + "learning_rate": 2.134601009265569e-06, + "loss": 0.176, + "step": 53960 + }, + { + "epoch": 3.7444313371729927, + "grad_norm": 0.9807775020599365, + "learning_rate": 2.133446454627852e-06, + "loss": 0.2601, + "step": 53961 + }, + { + "epoch": 3.7445007286100895, + "grad_norm": 0.9753558039665222, + "learning_rate": 2.132292208943487e-06, + "loss": 0.3551, + "step": 53962 + }, + { + "epoch": 3.744570120047186, + "grad_norm": 1.113999605178833, + "learning_rate": 2.1311382722161155e-06, + "loss": 0.2635, + "step": 53963 + }, + { + "epoch": 3.7446395114842828, + "grad_norm": 1.0613723993301392, + "learning_rate": 2.1299846444493897e-06, + "loss": 0.151, + "step": 53964 + }, + { + "epoch": 3.7447089029213796, + "grad_norm": 0.7770950198173523, + "learning_rate": 2.1288313256469518e-06, + "loss": 0.2207, + "step": 53965 + }, + { + "epoch": 3.7447782943584764, + "grad_norm": 0.9798986911773682, + "learning_rate": 2.1276783158124314e-06, + "loss": 0.1716, + "step": 53966 + }, + { + "epoch": 3.744847685795573, + "grad_norm": 1.0837780237197876, + "learning_rate": 2.126525614949482e-06, + "loss": 0.3349, + "step": 53967 + }, + { + "epoch": 3.7449170772326696, + "grad_norm": 0.7894635796546936, + "learning_rate": 2.125373223061733e-06, + "loss": 0.2287, + "step": 53968 + }, + { + "epoch": 3.744986468669766, + "grad_norm": 1.1939573287963867, + "learning_rate": 2.1242211401528267e-06, + "loss": 0.3523, + "step": 53969 + }, + { + "epoch": 3.745055860106863, + "grad_norm": 0.8595273494720459, + "learning_rate": 2.1230693662263934e-06, + "loss": 0.3636, + "step": 53970 + }, + { + "epoch": 3.7451252515439597, + "grad_norm": 0.872330904006958, + "learning_rate": 2.1219179012860748e-06, + "loss": 0.2924, + "step": 53971 + }, + { + "epoch": 3.745194642981056, + "grad_norm": 0.982765257358551, + "learning_rate": 2.120766745335512e-06, + "loss": 0.3188, + "step": 53972 + }, + { + "epoch": 3.745264034418153, + "grad_norm": 0.5690633654594421, + "learning_rate": 2.1196158983783356e-06, + "loss": 0.1573, + "step": 53973 + }, + { + "epoch": 3.7453334258552493, + "grad_norm": 0.8875654935836792, + "learning_rate": 2.1184653604181648e-06, + "loss": 0.2546, + "step": 53974 + }, + { + "epoch": 3.745402817292346, + "grad_norm": 1.253304123878479, + "learning_rate": 2.1173151314586416e-06, + "loss": 0.2192, + "step": 53975 + }, + { + "epoch": 3.745472208729443, + "grad_norm": 0.694534182548523, + "learning_rate": 2.116165211503407e-06, + "loss": 0.219, + "step": 53976 + }, + { + "epoch": 3.7455416001665394, + "grad_norm": 0.8537280559539795, + "learning_rate": 2.1150156005560694e-06, + "loss": 0.2639, + "step": 53977 + }, + { + "epoch": 3.745610991603636, + "grad_norm": 0.8888135552406311, + "learning_rate": 2.113866298620282e-06, + "loss": 0.2554, + "step": 53978 + }, + { + "epoch": 3.7456803830407326, + "grad_norm": 0.937140166759491, + "learning_rate": 2.112717305699663e-06, + "loss": 0.3187, + "step": 53979 + }, + { + "epoch": 3.7457497744778294, + "grad_norm": 1.2524722814559937, + "learning_rate": 2.1115686217978324e-06, + "loss": 0.2749, + "step": 53980 + }, + { + "epoch": 3.7458191659149263, + "grad_norm": 1.2180705070495605, + "learning_rate": 2.110420246918421e-06, + "loss": 0.1805, + "step": 53981 + }, + { + "epoch": 3.7458885573520226, + "grad_norm": 1.6521639823913574, + "learning_rate": 2.1092721810650585e-06, + "loss": 0.4013, + "step": 53982 + }, + { + "epoch": 3.7459579487891195, + "grad_norm": 0.7396224141120911, + "learning_rate": 2.1081244242413644e-06, + "loss": 0.211, + "step": 53983 + }, + { + "epoch": 3.746027340226216, + "grad_norm": 1.4160306453704834, + "learning_rate": 2.1069769764509695e-06, + "loss": 0.3246, + "step": 53984 + }, + { + "epoch": 3.7460967316633127, + "grad_norm": 0.673157811164856, + "learning_rate": 2.1058298376974816e-06, + "loss": 0.201, + "step": 53985 + }, + { + "epoch": 3.7461661231004095, + "grad_norm": 0.7670906186103821, + "learning_rate": 2.1046830079845426e-06, + "loss": 0.2542, + "step": 53986 + }, + { + "epoch": 3.746235514537506, + "grad_norm": 1.4020050764083862, + "learning_rate": 2.1035364873157605e-06, + "loss": 0.2663, + "step": 53987 + }, + { + "epoch": 3.7463049059746028, + "grad_norm": 0.9859849810600281, + "learning_rate": 2.102390275694743e-06, + "loss": 0.3831, + "step": 53988 + }, + { + "epoch": 3.746374297411699, + "grad_norm": 0.8672106266021729, + "learning_rate": 2.101244373125133e-06, + "loss": 0.2636, + "step": 53989 + }, + { + "epoch": 3.746443688848796, + "grad_norm": 0.5203759074211121, + "learning_rate": 2.1000987796105265e-06, + "loss": 0.1668, + "step": 53990 + }, + { + "epoch": 3.746513080285893, + "grad_norm": 0.9285572171211243, + "learning_rate": 2.098953495154554e-06, + "loss": 0.3565, + "step": 53991 + }, + { + "epoch": 3.7465824717229896, + "grad_norm": 1.0397251844406128, + "learning_rate": 2.097808519760824e-06, + "loss": 0.3189, + "step": 53992 + }, + { + "epoch": 3.746651863160086, + "grad_norm": 0.985035240650177, + "learning_rate": 2.0966638534329675e-06, + "loss": 0.2575, + "step": 53993 + }, + { + "epoch": 3.746721254597183, + "grad_norm": 1.0912145376205444, + "learning_rate": 2.0955194961745693e-06, + "loss": 0.2411, + "step": 53994 + }, + { + "epoch": 3.7467906460342792, + "grad_norm": 0.7598931789398193, + "learning_rate": 2.0943754479892606e-06, + "loss": 0.2103, + "step": 53995 + }, + { + "epoch": 3.746860037471376, + "grad_norm": 1.106379747390747, + "learning_rate": 2.0932317088806495e-06, + "loss": 0.3244, + "step": 53996 + }, + { + "epoch": 3.746929428908473, + "grad_norm": 0.9129778742790222, + "learning_rate": 2.0920882788523556e-06, + "loss": 0.2752, + "step": 53997 + }, + { + "epoch": 3.7469988203455693, + "grad_norm": 0.4884258508682251, + "learning_rate": 2.0909451579079642e-06, + "loss": 0.1653, + "step": 53998 + }, + { + "epoch": 3.747068211782666, + "grad_norm": 0.8957690000534058, + "learning_rate": 2.0898023460511063e-06, + "loss": 0.2548, + "step": 53999 + }, + { + "epoch": 3.7471376032197625, + "grad_norm": 1.050081491470337, + "learning_rate": 2.0886598432853897e-06, + "loss": 0.2966, + "step": 54000 + }, + { + "epoch": 3.7472069946568594, + "grad_norm": 1.1249362230300903, + "learning_rate": 2.087517649614412e-06, + "loss": 0.402, + "step": 54001 + }, + { + "epoch": 3.747276386093956, + "grad_norm": 1.3883285522460938, + "learning_rate": 2.0863757650417703e-06, + "loss": 0.255, + "step": 54002 + }, + { + "epoch": 3.7473457775310526, + "grad_norm": 1.366918921470642, + "learning_rate": 2.085234189571084e-06, + "loss": 0.3593, + "step": 54003 + }, + { + "epoch": 3.7474151689681494, + "grad_norm": 1.3452740907669067, + "learning_rate": 2.0840929232059712e-06, + "loss": 0.2389, + "step": 54004 + }, + { + "epoch": 3.747484560405246, + "grad_norm": 0.9109042882919312, + "learning_rate": 2.0829519659499974e-06, + "loss": 0.2568, + "step": 54005 + }, + { + "epoch": 3.7475539518423426, + "grad_norm": 0.9798374176025391, + "learning_rate": 2.081811317806792e-06, + "loss": 0.2335, + "step": 54006 + }, + { + "epoch": 3.7476233432794395, + "grad_norm": 0.8960714340209961, + "learning_rate": 2.080670978779942e-06, + "loss": 0.3074, + "step": 54007 + }, + { + "epoch": 3.747692734716536, + "grad_norm": 0.566628098487854, + "learning_rate": 2.079530948873054e-06, + "loss": 0.1204, + "step": 54008 + }, + { + "epoch": 3.7477621261536327, + "grad_norm": 0.9553033709526062, + "learning_rate": 2.0783912280897264e-06, + "loss": 0.2651, + "step": 54009 + }, + { + "epoch": 3.747831517590729, + "grad_norm": 1.0106170177459717, + "learning_rate": 2.0772518164335565e-06, + "loss": 0.4954, + "step": 54010 + }, + { + "epoch": 3.747900909027826, + "grad_norm": 0.748685359954834, + "learning_rate": 2.0761127139081406e-06, + "loss": 0.1611, + "step": 54011 + }, + { + "epoch": 3.7479703004649227, + "grad_norm": 0.8059960007667542, + "learning_rate": 2.074973920517076e-06, + "loss": 0.1805, + "step": 54012 + }, + { + "epoch": 3.748039691902019, + "grad_norm": 0.9168770909309387, + "learning_rate": 2.073835436263971e-06, + "loss": 0.2535, + "step": 54013 + }, + { + "epoch": 3.748109083339116, + "grad_norm": 0.8533908128738403, + "learning_rate": 2.07269726115239e-06, + "loss": 0.2672, + "step": 54014 + }, + { + "epoch": 3.7481784747762124, + "grad_norm": 0.8082066178321838, + "learning_rate": 2.0715593951859402e-06, + "loss": 0.2242, + "step": 54015 + }, + { + "epoch": 3.748247866213309, + "grad_norm": 0.6186346411705017, + "learning_rate": 2.0704218383682195e-06, + "loss": 0.1565, + "step": 54016 + }, + { + "epoch": 3.748317257650406, + "grad_norm": 0.6102051138877869, + "learning_rate": 2.0692845907028133e-06, + "loss": 0.1073, + "step": 54017 + }, + { + "epoch": 3.748386649087503, + "grad_norm": 0.9044213891029358, + "learning_rate": 2.06814765219332e-06, + "loss": 0.2905, + "step": 54018 + }, + { + "epoch": 3.7484560405245992, + "grad_norm": 0.994997501373291, + "learning_rate": 2.0670110228433126e-06, + "loss": 0.2045, + "step": 54019 + }, + { + "epoch": 3.748525431961696, + "grad_norm": 0.7499203681945801, + "learning_rate": 2.065874702656401e-06, + "loss": 0.1888, + "step": 54020 + }, + { + "epoch": 3.7485948233987925, + "grad_norm": 0.8646500110626221, + "learning_rate": 2.0647386916361588e-06, + "loss": 0.2422, + "step": 54021 + }, + { + "epoch": 3.7486642148358893, + "grad_norm": 0.8050246834754944, + "learning_rate": 2.063602989786162e-06, + "loss": 0.2308, + "step": 54022 + }, + { + "epoch": 3.748733606272986, + "grad_norm": 0.5520590543746948, + "learning_rate": 2.0624675971100184e-06, + "loss": 0.1621, + "step": 54023 + }, + { + "epoch": 3.7488029977100825, + "grad_norm": 1.3338205814361572, + "learning_rate": 2.0613325136113026e-06, + "loss": 0.2469, + "step": 54024 + }, + { + "epoch": 3.7488723891471794, + "grad_norm": 0.8882614970207214, + "learning_rate": 2.0601977392935902e-06, + "loss": 0.237, + "step": 54025 + }, + { + "epoch": 3.7489417805842757, + "grad_norm": 1.2427420616149902, + "learning_rate": 2.0590632741604665e-06, + "loss": 0.246, + "step": 54026 + }, + { + "epoch": 3.7490111720213726, + "grad_norm": 0.9590967893600464, + "learning_rate": 2.0579291182155404e-06, + "loss": 0.3192, + "step": 54027 + }, + { + "epoch": 3.7490805634584694, + "grad_norm": 1.5022904872894287, + "learning_rate": 2.0567952714623418e-06, + "loss": 0.3899, + "step": 54028 + }, + { + "epoch": 3.749149954895566, + "grad_norm": 0.6617381572723389, + "learning_rate": 2.0556617339044902e-06, + "loss": 0.2681, + "step": 54029 + }, + { + "epoch": 3.7492193463326626, + "grad_norm": 0.8823610544204712, + "learning_rate": 2.0545285055455497e-06, + "loss": 0.2057, + "step": 54030 + }, + { + "epoch": 3.749288737769759, + "grad_norm": 0.6993985176086426, + "learning_rate": 2.053395586389095e-06, + "loss": 0.1977, + "step": 54031 + }, + { + "epoch": 3.749358129206856, + "grad_norm": 1.2854664325714111, + "learning_rate": 2.0522629764387124e-06, + "loss": 0.2897, + "step": 54032 + }, + { + "epoch": 3.7494275206439527, + "grad_norm": 0.9765544533729553, + "learning_rate": 2.0511306756979655e-06, + "loss": 0.2893, + "step": 54033 + }, + { + "epoch": 3.749496912081049, + "grad_norm": 0.949601411819458, + "learning_rate": 2.04999868417044e-06, + "loss": 0.3011, + "step": 54034 + }, + { + "epoch": 3.749566303518146, + "grad_norm": 0.6628440022468567, + "learning_rate": 2.0488670018597e-06, + "loss": 0.1749, + "step": 54035 + }, + { + "epoch": 3.7496356949552423, + "grad_norm": 0.8023876547813416, + "learning_rate": 2.0477356287693206e-06, + "loss": 0.1749, + "step": 54036 + }, + { + "epoch": 3.749705086392339, + "grad_norm": 0.8158652782440186, + "learning_rate": 2.0466045649028766e-06, + "loss": 0.3134, + "step": 54037 + }, + { + "epoch": 3.749774477829436, + "grad_norm": 1.0560970306396484, + "learning_rate": 2.0454738102639425e-06, + "loss": 0.3261, + "step": 54038 + }, + { + "epoch": 3.7498438692665323, + "grad_norm": 0.6797865629196167, + "learning_rate": 2.044343364856094e-06, + "loss": 0.1645, + "step": 54039 + }, + { + "epoch": 3.749913260703629, + "grad_norm": 0.6961197257041931, + "learning_rate": 2.043213228682872e-06, + "loss": 0.2366, + "step": 54040 + }, + { + "epoch": 3.7499826521407256, + "grad_norm": 0.8085980415344238, + "learning_rate": 2.042083401747863e-06, + "loss": 0.232, + "step": 54041 + }, + { + "epoch": 3.7500520435778224, + "grad_norm": 1.099845290184021, + "learning_rate": 2.0409538840546307e-06, + "loss": 0.3555, + "step": 54042 + }, + { + "epoch": 3.7501214350149192, + "grad_norm": 0.6446261405944824, + "learning_rate": 2.03982467560675e-06, + "loss": 0.1819, + "step": 54043 + }, + { + "epoch": 3.750190826452016, + "grad_norm": 0.927412748336792, + "learning_rate": 2.038695776407762e-06, + "loss": 0.2856, + "step": 54044 + }, + { + "epoch": 3.7502602178891125, + "grad_norm": 0.9095216393470764, + "learning_rate": 2.0375671864612534e-06, + "loss": 0.2285, + "step": 54045 + }, + { + "epoch": 3.7503296093262093, + "grad_norm": 0.8420848846435547, + "learning_rate": 2.0364389057707768e-06, + "loss": 0.171, + "step": 54046 + }, + { + "epoch": 3.7503990007633057, + "grad_norm": 0.9948099255561829, + "learning_rate": 2.0353109343398956e-06, + "loss": 0.1985, + "step": 54047 + }, + { + "epoch": 3.7504683922004025, + "grad_norm": 1.0223650932312012, + "learning_rate": 2.0341832721721855e-06, + "loss": 0.3337, + "step": 54048 + }, + { + "epoch": 3.7505377836374993, + "grad_norm": 0.9332689642906189, + "learning_rate": 2.0330559192711873e-06, + "loss": 0.3054, + "step": 54049 + }, + { + "epoch": 3.7506071750745957, + "grad_norm": 1.097042202949524, + "learning_rate": 2.031928875640454e-06, + "loss": 0.2171, + "step": 54050 + }, + { + "epoch": 3.7506765665116926, + "grad_norm": 0.9751644134521484, + "learning_rate": 2.030802141283561e-06, + "loss": 0.2915, + "step": 54051 + }, + { + "epoch": 3.750745957948789, + "grad_norm": 1.144019603729248, + "learning_rate": 2.0296757162040602e-06, + "loss": 0.3185, + "step": 54052 + }, + { + "epoch": 3.750815349385886, + "grad_norm": 2.19543194770813, + "learning_rate": 2.028549600405505e-06, + "loss": 0.2395, + "step": 54053 + }, + { + "epoch": 3.7508847408229826, + "grad_norm": 0.6073634624481201, + "learning_rate": 2.0274237938914476e-06, + "loss": 0.1526, + "step": 54054 + }, + { + "epoch": 3.750954132260079, + "grad_norm": 0.8599225282669067, + "learning_rate": 2.026298296665463e-06, + "loss": 0.2099, + "step": 54055 + }, + { + "epoch": 3.751023523697176, + "grad_norm": 0.7884583473205566, + "learning_rate": 2.025173108731082e-06, + "loss": 0.1999, + "step": 54056 + }, + { + "epoch": 3.7510929151342722, + "grad_norm": 0.7115729451179504, + "learning_rate": 2.024048230091846e-06, + "loss": 0.1688, + "step": 54057 + }, + { + "epoch": 3.751162306571369, + "grad_norm": 0.9460783004760742, + "learning_rate": 2.0229236607513415e-06, + "loss": 0.3165, + "step": 54058 + }, + { + "epoch": 3.751231698008466, + "grad_norm": 0.9287911057472229, + "learning_rate": 2.0217994007130868e-06, + "loss": 0.2883, + "step": 54059 + }, + { + "epoch": 3.7513010894455623, + "grad_norm": 0.535328209400177, + "learning_rate": 2.0206754499806466e-06, + "loss": 0.139, + "step": 54060 + }, + { + "epoch": 3.751370480882659, + "grad_norm": 0.6047463417053223, + "learning_rate": 2.0195518085575626e-06, + "loss": 0.2166, + "step": 54061 + }, + { + "epoch": 3.7514398723197555, + "grad_norm": 0.7274322509765625, + "learning_rate": 2.018428476447398e-06, + "loss": 0.1986, + "step": 54062 + }, + { + "epoch": 3.7515092637568523, + "grad_norm": 0.715263843536377, + "learning_rate": 2.017305453653684e-06, + "loss": 0.1889, + "step": 54063 + }, + { + "epoch": 3.751578655193949, + "grad_norm": 0.9363061785697937, + "learning_rate": 2.0161827401799614e-06, + "loss": 0.3584, + "step": 54064 + }, + { + "epoch": 3.7516480466310456, + "grad_norm": 0.8577204942703247, + "learning_rate": 2.015060336029784e-06, + "loss": 0.203, + "step": 54065 + }, + { + "epoch": 3.7517174380681424, + "grad_norm": 1.0029971599578857, + "learning_rate": 2.013938241206692e-06, + "loss": 0.442, + "step": 54066 + }, + { + "epoch": 3.751786829505239, + "grad_norm": 0.721026599407196, + "learning_rate": 2.012816455714228e-06, + "loss": 0.2201, + "step": 54067 + }, + { + "epoch": 3.7518562209423356, + "grad_norm": 0.9248238205909729, + "learning_rate": 2.011694979555934e-06, + "loss": 0.2349, + "step": 54068 + }, + { + "epoch": 3.7519256123794325, + "grad_norm": 0.7379553914070129, + "learning_rate": 2.010573812735361e-06, + "loss": 0.2066, + "step": 54069 + }, + { + "epoch": 3.7519950038165293, + "grad_norm": 0.897034764289856, + "learning_rate": 2.0094529552560302e-06, + "loss": 0.3334, + "step": 54070 + }, + { + "epoch": 3.7520643952536257, + "grad_norm": 0.808591365814209, + "learning_rate": 2.0083324071214826e-06, + "loss": 0.2401, + "step": 54071 + }, + { + "epoch": 3.7521337866907225, + "grad_norm": 0.9034290909767151, + "learning_rate": 2.0072121683352596e-06, + "loss": 0.2262, + "step": 54072 + }, + { + "epoch": 3.752203178127819, + "grad_norm": 0.8885973691940308, + "learning_rate": 2.0060922389009027e-06, + "loss": 0.3, + "step": 54073 + }, + { + "epoch": 3.7522725695649157, + "grad_norm": 0.7662637233734131, + "learning_rate": 2.0049726188219543e-06, + "loss": 0.2725, + "step": 54074 + }, + { + "epoch": 3.7523419610020126, + "grad_norm": 1.1336809396743774, + "learning_rate": 2.003853308101922e-06, + "loss": 0.3847, + "step": 54075 + }, + { + "epoch": 3.752411352439109, + "grad_norm": 0.8291364908218384, + "learning_rate": 2.002734306744358e-06, + "loss": 0.2021, + "step": 54076 + }, + { + "epoch": 3.752480743876206, + "grad_norm": 1.126539707183838, + "learning_rate": 2.0016156147528053e-06, + "loss": 0.2761, + "step": 54077 + }, + { + "epoch": 3.752550135313302, + "grad_norm": 1.309891700744629, + "learning_rate": 2.000497232130771e-06, + "loss": 0.4452, + "step": 54078 + }, + { + "epoch": 3.752619526750399, + "grad_norm": 0.8887557983398438, + "learning_rate": 1.9993791588817866e-06, + "loss": 0.2603, + "step": 54079 + }, + { + "epoch": 3.752688918187496, + "grad_norm": 1.4231826066970825, + "learning_rate": 1.9982613950094044e-06, + "loss": 0.2784, + "step": 54080 + }, + { + "epoch": 3.7527583096245922, + "grad_norm": 0.9095187187194824, + "learning_rate": 1.9971439405171433e-06, + "loss": 0.3651, + "step": 54081 + }, + { + "epoch": 3.752827701061689, + "grad_norm": 0.9822902679443359, + "learning_rate": 1.996026795408512e-06, + "loss": 0.3143, + "step": 54082 + }, + { + "epoch": 3.7528970924987854, + "grad_norm": 1.0653278827667236, + "learning_rate": 1.9949099596870746e-06, + "loss": 0.3901, + "step": 54083 + }, + { + "epoch": 3.7529664839358823, + "grad_norm": 0.7259154915809631, + "learning_rate": 1.9937934333563278e-06, + "loss": 0.2443, + "step": 54084 + }, + { + "epoch": 3.753035875372979, + "grad_norm": 1.1180516481399536, + "learning_rate": 1.9926772164198026e-06, + "loss": 0.2647, + "step": 54085 + }, + { + "epoch": 3.7531052668100755, + "grad_norm": 0.8027560710906982, + "learning_rate": 1.991561308881018e-06, + "loss": 0.1656, + "step": 54086 + }, + { + "epoch": 3.7531746582471723, + "grad_norm": 1.0506179332733154, + "learning_rate": 1.990445710743516e-06, + "loss": 0.1803, + "step": 54087 + }, + { + "epoch": 3.7532440496842687, + "grad_norm": 1.2795408964157104, + "learning_rate": 1.9893304220107935e-06, + "loss": 0.2903, + "step": 54088 + }, + { + "epoch": 3.7533134411213656, + "grad_norm": 1.0477286577224731, + "learning_rate": 1.988215442686392e-06, + "loss": 0.2455, + "step": 54089 + }, + { + "epoch": 3.7533828325584624, + "grad_norm": 0.8461871147155762, + "learning_rate": 1.98710077277382e-06, + "loss": 0.289, + "step": 54090 + }, + { + "epoch": 3.7534522239955592, + "grad_norm": 1.2868247032165527, + "learning_rate": 1.985986412276597e-06, + "loss": 0.4798, + "step": 54091 + }, + { + "epoch": 3.7535216154326556, + "grad_norm": 1.3792833089828491, + "learning_rate": 1.9848723611982534e-06, + "loss": 0.2138, + "step": 54092 + }, + { + "epoch": 3.753591006869752, + "grad_norm": 0.9664590358734131, + "learning_rate": 1.9837586195422865e-06, + "loss": 0.342, + "step": 54093 + }, + { + "epoch": 3.753660398306849, + "grad_norm": 1.147667407989502, + "learning_rate": 1.9826451873122155e-06, + "loss": 0.3488, + "step": 54094 + }, + { + "epoch": 3.7537297897439457, + "grad_norm": 0.9474658370018005, + "learning_rate": 1.981532064511571e-06, + "loss": 0.2652, + "step": 54095 + }, + { + "epoch": 3.7537991811810425, + "grad_norm": 0.9106594920158386, + "learning_rate": 1.980419251143861e-06, + "loss": 0.1742, + "step": 54096 + }, + { + "epoch": 3.753868572618139, + "grad_norm": 0.8961667418479919, + "learning_rate": 1.979306747212595e-06, + "loss": 0.1764, + "step": 54097 + }, + { + "epoch": 3.7539379640552357, + "grad_norm": 0.9471021890640259, + "learning_rate": 1.9781945527212796e-06, + "loss": 0.3348, + "step": 54098 + }, + { + "epoch": 3.754007355492332, + "grad_norm": 0.7957243323326111, + "learning_rate": 1.9770826676734242e-06, + "loss": 0.2785, + "step": 54099 + }, + { + "epoch": 3.754076746929429, + "grad_norm": 1.219286322593689, + "learning_rate": 1.9759710920725594e-06, + "loss": 0.3219, + "step": 54100 + }, + { + "epoch": 3.7541461383665258, + "grad_norm": 1.0881032943725586, + "learning_rate": 1.9748598259221705e-06, + "loss": 0.3198, + "step": 54101 + }, + { + "epoch": 3.754215529803622, + "grad_norm": 0.945976734161377, + "learning_rate": 1.973748869225778e-06, + "loss": 0.3572, + "step": 54102 + }, + { + "epoch": 3.754284921240719, + "grad_norm": 0.895182192325592, + "learning_rate": 1.9726382219869e-06, + "loss": 0.247, + "step": 54103 + }, + { + "epoch": 3.7543543126778154, + "grad_norm": 0.8665211796760559, + "learning_rate": 1.971527884209012e-06, + "loss": 0.1925, + "step": 54104 + }, + { + "epoch": 3.754423704114912, + "grad_norm": 0.9299649596214294, + "learning_rate": 1.9704178558956453e-06, + "loss": 0.2315, + "step": 54105 + }, + { + "epoch": 3.754493095552009, + "grad_norm": 0.9942921996116638, + "learning_rate": 1.9693081370502965e-06, + "loss": 0.2785, + "step": 54106 + }, + { + "epoch": 3.7545624869891054, + "grad_norm": 0.8628849983215332, + "learning_rate": 1.9681987276764623e-06, + "loss": 0.2015, + "step": 54107 + }, + { + "epoch": 3.7546318784262023, + "grad_norm": 0.9308982491493225, + "learning_rate": 1.967089627777663e-06, + "loss": 0.3251, + "step": 54108 + }, + { + "epoch": 3.7547012698632987, + "grad_norm": 1.211602807044983, + "learning_rate": 1.965980837357384e-06, + "loss": 0.1295, + "step": 54109 + }, + { + "epoch": 3.7547706613003955, + "grad_norm": 1.098705768585205, + "learning_rate": 1.964872356419123e-06, + "loss": 0.282, + "step": 54110 + }, + { + "epoch": 3.7548400527374923, + "grad_norm": 0.907936155796051, + "learning_rate": 1.9637641849663878e-06, + "loss": 0.2527, + "step": 54111 + }, + { + "epoch": 3.7549094441745887, + "grad_norm": 0.7634331583976746, + "learning_rate": 1.962656323002687e-06, + "loss": 0.3235, + "step": 54112 + }, + { + "epoch": 3.7549788356116856, + "grad_norm": 0.9990032315254211, + "learning_rate": 1.961548770531485e-06, + "loss": 0.2349, + "step": 54113 + }, + { + "epoch": 3.755048227048782, + "grad_norm": 0.9371129870414734, + "learning_rate": 1.9604415275563116e-06, + "loss": 0.2121, + "step": 54114 + }, + { + "epoch": 3.7551176184858788, + "grad_norm": 0.7239062786102295, + "learning_rate": 1.9593345940806417e-06, + "loss": 0.2316, + "step": 54115 + }, + { + "epoch": 3.7551870099229756, + "grad_norm": 0.9564937949180603, + "learning_rate": 1.958227970107984e-06, + "loss": 0.2953, + "step": 54116 + }, + { + "epoch": 3.7552564013600724, + "grad_norm": 0.8457499742507935, + "learning_rate": 1.957121655641836e-06, + "loss": 0.267, + "step": 54117 + }, + { + "epoch": 3.755325792797169, + "grad_norm": 0.532036304473877, + "learning_rate": 1.9560156506856607e-06, + "loss": 0.1628, + "step": 54118 + }, + { + "epoch": 3.755395184234265, + "grad_norm": 0.8315739631652832, + "learning_rate": 1.9549099552429785e-06, + "loss": 0.2282, + "step": 54119 + }, + { + "epoch": 3.755464575671362, + "grad_norm": 1.4175068140029907, + "learning_rate": 1.9538045693172636e-06, + "loss": 0.2346, + "step": 54120 + }, + { + "epoch": 3.755533967108459, + "grad_norm": 0.6431819200515747, + "learning_rate": 1.9526994929120133e-06, + "loss": 0.1748, + "step": 54121 + }, + { + "epoch": 3.7556033585455557, + "grad_norm": 0.8926326036453247, + "learning_rate": 1.9515947260307255e-06, + "loss": 0.2461, + "step": 54122 + }, + { + "epoch": 3.755672749982652, + "grad_norm": 0.7724372148513794, + "learning_rate": 1.9504902686768632e-06, + "loss": 0.27, + "step": 54123 + }, + { + "epoch": 3.755742141419749, + "grad_norm": 0.7177446484565735, + "learning_rate": 1.9493861208539464e-06, + "loss": 0.2405, + "step": 54124 + }, + { + "epoch": 3.7558115328568453, + "grad_norm": 0.7788817286491394, + "learning_rate": 1.948282282565428e-06, + "loss": 0.2725, + "step": 54125 + }, + { + "epoch": 3.755880924293942, + "grad_norm": 0.9732003808021545, + "learning_rate": 1.9471787538148047e-06, + "loss": 0.2841, + "step": 54126 + }, + { + "epoch": 3.755950315731039, + "grad_norm": 0.8050129413604736, + "learning_rate": 1.9460755346055628e-06, + "loss": 0.2189, + "step": 54127 + }, + { + "epoch": 3.7560197071681354, + "grad_norm": 0.6574059724807739, + "learning_rate": 1.944972624941188e-06, + "loss": 0.2335, + "step": 54128 + }, + { + "epoch": 3.756089098605232, + "grad_norm": 0.6759178638458252, + "learning_rate": 1.943870024825156e-06, + "loss": 0.2083, + "step": 54129 + }, + { + "epoch": 3.7561584900423286, + "grad_norm": 0.7755627632141113, + "learning_rate": 1.942767734260942e-06, + "loss": 0.2439, + "step": 54130 + }, + { + "epoch": 3.7562278814794254, + "grad_norm": 1.0421335697174072, + "learning_rate": 1.9416657532520534e-06, + "loss": 0.3496, + "step": 54131 + }, + { + "epoch": 3.7562972729165223, + "grad_norm": 1.9999898672103882, + "learning_rate": 1.940564081801932e-06, + "loss": 0.3705, + "step": 54132 + }, + { + "epoch": 3.7563666643536187, + "grad_norm": 0.9855790138244629, + "learning_rate": 1.939462719914076e-06, + "loss": 0.2552, + "step": 54133 + }, + { + "epoch": 3.7564360557907155, + "grad_norm": 0.9734733700752258, + "learning_rate": 1.938361667591959e-06, + "loss": 0.1947, + "step": 54134 + }, + { + "epoch": 3.756505447227812, + "grad_norm": 0.8407430648803711, + "learning_rate": 1.9372609248390685e-06, + "loss": 0.1452, + "step": 54135 + }, + { + "epoch": 3.7565748386649087, + "grad_norm": 0.7137047648429871, + "learning_rate": 1.9361604916588562e-06, + "loss": 0.217, + "step": 54136 + }, + { + "epoch": 3.7566442301020055, + "grad_norm": 0.7741791009902954, + "learning_rate": 1.9350603680548087e-06, + "loss": 0.1743, + "step": 54137 + }, + { + "epoch": 3.756713621539102, + "grad_norm": 1.3062050342559814, + "learning_rate": 1.9339605540304007e-06, + "loss": 0.5523, + "step": 54138 + }, + { + "epoch": 3.7567830129761988, + "grad_norm": 0.8660224676132202, + "learning_rate": 1.9328610495890964e-06, + "loss": 0.2849, + "step": 54139 + }, + { + "epoch": 3.756852404413295, + "grad_norm": 1.3344480991363525, + "learning_rate": 1.931761854734382e-06, + "loss": 0.4445, + "step": 54140 + }, + { + "epoch": 3.756921795850392, + "grad_norm": 0.6331169009208679, + "learning_rate": 1.9306629694697097e-06, + "loss": 0.127, + "step": 54141 + }, + { + "epoch": 3.756991187287489, + "grad_norm": 0.9504539370536804, + "learning_rate": 1.929564393798555e-06, + "loss": 0.2136, + "step": 54142 + }, + { + "epoch": 3.7570605787245857, + "grad_norm": 0.6212565302848816, + "learning_rate": 1.928466127724393e-06, + "loss": 0.1534, + "step": 54143 + }, + { + "epoch": 3.757129970161682, + "grad_norm": 1.3153016567230225, + "learning_rate": 1.9273681712506877e-06, + "loss": 0.4917, + "step": 54144 + }, + { + "epoch": 3.757199361598779, + "grad_norm": 0.7286112904548645, + "learning_rate": 1.926270524380891e-06, + "loss": 0.2091, + "step": 54145 + }, + { + "epoch": 3.7572687530358753, + "grad_norm": 0.6970857977867126, + "learning_rate": 1.92517318711849e-06, + "loss": 0.1979, + "step": 54146 + }, + { + "epoch": 3.757338144472972, + "grad_norm": 1.2874119281768799, + "learning_rate": 1.9240761594669366e-06, + "loss": 0.1796, + "step": 54147 + }, + { + "epoch": 3.757407535910069, + "grad_norm": 1.0022625923156738, + "learning_rate": 1.9229794414296955e-06, + "loss": 0.2336, + "step": 54148 + }, + { + "epoch": 3.7574769273471653, + "grad_norm": 1.0261938571929932, + "learning_rate": 1.921883033010219e-06, + "loss": 0.2739, + "step": 54149 + }, + { + "epoch": 3.757546318784262, + "grad_norm": 0.9988771677017212, + "learning_rate": 1.920786934211982e-06, + "loss": 0.3738, + "step": 54150 + }, + { + "epoch": 3.7576157102213585, + "grad_norm": 0.8170203566551208, + "learning_rate": 1.9196911450384493e-06, + "loss": 0.2084, + "step": 54151 + }, + { + "epoch": 3.7576851016584554, + "grad_norm": 1.2259315252304077, + "learning_rate": 1.9185956654930727e-06, + "loss": 0.2647, + "step": 54152 + }, + { + "epoch": 3.757754493095552, + "grad_norm": 0.5867131948471069, + "learning_rate": 1.9175004955793054e-06, + "loss": 0.118, + "step": 54153 + }, + { + "epoch": 3.7578238845326486, + "grad_norm": 0.7399874329566956, + "learning_rate": 1.916405635300611e-06, + "loss": 0.2625, + "step": 54154 + }, + { + "epoch": 3.7578932759697454, + "grad_norm": 0.7185850739479065, + "learning_rate": 1.9153110846604427e-06, + "loss": 0.161, + "step": 54155 + }, + { + "epoch": 3.757962667406842, + "grad_norm": 0.7000589966773987, + "learning_rate": 1.914216843662253e-06, + "loss": 0.2344, + "step": 54156 + }, + { + "epoch": 3.7580320588439386, + "grad_norm": 0.9471980929374695, + "learning_rate": 1.9131229123095063e-06, + "loss": 0.223, + "step": 54157 + }, + { + "epoch": 3.7581014502810355, + "grad_norm": 1.2063279151916504, + "learning_rate": 1.9120292906056546e-06, + "loss": 0.2973, + "step": 54158 + }, + { + "epoch": 3.758170841718132, + "grad_norm": 1.0704395771026611, + "learning_rate": 1.9109359785541404e-06, + "loss": 0.2606, + "step": 54159 + }, + { + "epoch": 3.7582402331552287, + "grad_norm": 0.6525079607963562, + "learning_rate": 1.909842976158427e-06, + "loss": 0.1312, + "step": 54160 + }, + { + "epoch": 3.758309624592325, + "grad_norm": 0.8340815901756287, + "learning_rate": 1.9087502834219452e-06, + "loss": 0.2811, + "step": 54161 + }, + { + "epoch": 3.758379016029422, + "grad_norm": 1.1781491041183472, + "learning_rate": 1.9076579003481696e-06, + "loss": 0.4437, + "step": 54162 + }, + { + "epoch": 3.7584484074665188, + "grad_norm": 1.1335277557373047, + "learning_rate": 1.9065658269405318e-06, + "loss": 0.1972, + "step": 54163 + }, + { + "epoch": 3.758517798903615, + "grad_norm": 0.8551485538482666, + "learning_rate": 1.9054740632024837e-06, + "loss": 0.1839, + "step": 54164 + }, + { + "epoch": 3.758587190340712, + "grad_norm": 0.8324335217475891, + "learning_rate": 1.9043826091374894e-06, + "loss": 0.2442, + "step": 54165 + }, + { + "epoch": 3.7586565817778084, + "grad_norm": 1.061416506767273, + "learning_rate": 1.9032914647489576e-06, + "loss": 0.3252, + "step": 54166 + }, + { + "epoch": 3.758725973214905, + "grad_norm": 0.6473374962806702, + "learning_rate": 1.9022006300403627e-06, + "loss": 0.1454, + "step": 54167 + }, + { + "epoch": 3.758795364652002, + "grad_norm": 0.8586477637290955, + "learning_rate": 1.9011101050151359e-06, + "loss": 0.2395, + "step": 54168 + }, + { + "epoch": 3.758864756089099, + "grad_norm": 1.0414484739303589, + "learning_rate": 1.9000198896767186e-06, + "loss": 0.2924, + "step": 54169 + }, + { + "epoch": 3.7589341475261953, + "grad_norm": 0.8010509610176086, + "learning_rate": 1.8989299840285635e-06, + "loss": 0.279, + "step": 54170 + }, + { + "epoch": 3.759003538963292, + "grad_norm": 1.2351268529891968, + "learning_rate": 1.8978403880741014e-06, + "loss": 0.5563, + "step": 54171 + }, + { + "epoch": 3.7590729304003885, + "grad_norm": 0.9326993227005005, + "learning_rate": 1.896751101816785e-06, + "loss": 0.2159, + "step": 54172 + }, + { + "epoch": 3.7591423218374853, + "grad_norm": 1.0382626056671143, + "learning_rate": 1.8956621252600337e-06, + "loss": 0.2759, + "step": 54173 + }, + { + "epoch": 3.759211713274582, + "grad_norm": 0.6419837474822998, + "learning_rate": 1.8945734584072893e-06, + "loss": 0.2118, + "step": 54174 + }, + { + "epoch": 3.7592811047116785, + "grad_norm": 0.7910164594650269, + "learning_rate": 1.8934851012620047e-06, + "loss": 0.2223, + "step": 54175 + }, + { + "epoch": 3.7593504961487754, + "grad_norm": 1.0173141956329346, + "learning_rate": 1.8923970538275993e-06, + "loss": 0.5903, + "step": 54176 + }, + { + "epoch": 3.7594198875858718, + "grad_norm": 0.753501296043396, + "learning_rate": 1.8913093161075147e-06, + "loss": 0.2183, + "step": 54177 + }, + { + "epoch": 3.7594892790229686, + "grad_norm": 0.8098652362823486, + "learning_rate": 1.8902218881051925e-06, + "loss": 0.1767, + "step": 54178 + }, + { + "epoch": 3.7595586704600654, + "grad_norm": 1.0544148683547974, + "learning_rate": 1.8891347698240414e-06, + "loss": 0.2384, + "step": 54179 + }, + { + "epoch": 3.759628061897162, + "grad_norm": 1.2954078912734985, + "learning_rate": 1.888047961267525e-06, + "loss": 0.3003, + "step": 54180 + }, + { + "epoch": 3.7596974533342586, + "grad_norm": 0.9061128497123718, + "learning_rate": 1.8869614624390407e-06, + "loss": 0.2731, + "step": 54181 + }, + { + "epoch": 3.759766844771355, + "grad_norm": 0.848919689655304, + "learning_rate": 1.8858752733420527e-06, + "loss": 0.187, + "step": 54182 + }, + { + "epoch": 3.759836236208452, + "grad_norm": 1.129554033279419, + "learning_rate": 1.8847893939799687e-06, + "loss": 0.3219, + "step": 54183 + }, + { + "epoch": 3.7599056276455487, + "grad_norm": 1.1648157835006714, + "learning_rate": 1.8837038243562199e-06, + "loss": 0.4538, + "step": 54184 + }, + { + "epoch": 3.759975019082645, + "grad_norm": 0.8014290928840637, + "learning_rate": 1.8826185644742256e-06, + "loss": 0.1736, + "step": 54185 + }, + { + "epoch": 3.760044410519742, + "grad_norm": 0.7065460085868835, + "learning_rate": 1.8815336143374385e-06, + "loss": 0.1907, + "step": 54186 + }, + { + "epoch": 3.7601138019568383, + "grad_norm": 0.7763152122497559, + "learning_rate": 1.880448973949256e-06, + "loss": 0.2182, + "step": 54187 + }, + { + "epoch": 3.760183193393935, + "grad_norm": 1.076045036315918, + "learning_rate": 1.8793646433131084e-06, + "loss": 0.229, + "step": 54188 + }, + { + "epoch": 3.760252584831032, + "grad_norm": 0.7920147180557251, + "learning_rate": 1.8782806224324267e-06, + "loss": 0.2373, + "step": 54189 + }, + { + "epoch": 3.7603219762681284, + "grad_norm": 0.7196626663208008, + "learning_rate": 1.8771969113106192e-06, + "loss": 0.2227, + "step": 54190 + }, + { + "epoch": 3.760391367705225, + "grad_norm": 1.04707932472229, + "learning_rate": 1.8761135099511273e-06, + "loss": 0.2262, + "step": 54191 + }, + { + "epoch": 3.7604607591423216, + "grad_norm": 1.1003168821334839, + "learning_rate": 1.8750304183573597e-06, + "loss": 0.2923, + "step": 54192 + }, + { + "epoch": 3.7605301505794184, + "grad_norm": 0.9141578674316406, + "learning_rate": 1.8739476365327468e-06, + "loss": 0.2915, + "step": 54193 + }, + { + "epoch": 3.7605995420165153, + "grad_norm": 0.7811575531959534, + "learning_rate": 1.872865164480675e-06, + "loss": 0.1649, + "step": 54194 + }, + { + "epoch": 3.760668933453612, + "grad_norm": 0.7756574153900146, + "learning_rate": 1.8717830022045968e-06, + "loss": 0.1956, + "step": 54195 + }, + { + "epoch": 3.7607383248907085, + "grad_norm": 0.9915605783462524, + "learning_rate": 1.8707011497079097e-06, + "loss": 0.2481, + "step": 54196 + }, + { + "epoch": 3.7608077163278053, + "grad_norm": 0.7891096472740173, + "learning_rate": 1.869619606994033e-06, + "loss": 0.2016, + "step": 54197 + }, + { + "epoch": 3.7608771077649017, + "grad_norm": 1.2874561548233032, + "learning_rate": 1.8685383740663865e-06, + "loss": 0.2892, + "step": 54198 + }, + { + "epoch": 3.7609464992019985, + "grad_norm": 1.1685062646865845, + "learning_rate": 1.867457450928367e-06, + "loss": 0.4009, + "step": 54199 + }, + { + "epoch": 3.7610158906390954, + "grad_norm": 0.6505409479141235, + "learning_rate": 1.8663768375834168e-06, + "loss": 0.2864, + "step": 54200 + }, + { + "epoch": 3.7610852820761917, + "grad_norm": 0.6316568851470947, + "learning_rate": 1.8652965340349216e-06, + "loss": 0.2079, + "step": 54201 + }, + { + "epoch": 3.7611546735132886, + "grad_norm": 0.784103274345398, + "learning_rate": 1.8642165402863009e-06, + "loss": 0.2332, + "step": 54202 + }, + { + "epoch": 3.761224064950385, + "grad_norm": 1.0712451934814453, + "learning_rate": 1.8631368563409524e-06, + "loss": 0.3093, + "step": 54203 + }, + { + "epoch": 3.761293456387482, + "grad_norm": 0.7411912083625793, + "learning_rate": 1.8620574822023063e-06, + "loss": 0.1534, + "step": 54204 + }, + { + "epoch": 3.7613628478245786, + "grad_norm": 0.5651435852050781, + "learning_rate": 1.86097841787376e-06, + "loss": 0.199, + "step": 54205 + }, + { + "epoch": 3.761432239261675, + "grad_norm": 1.4942682981491089, + "learning_rate": 1.8598996633587107e-06, + "loss": 0.282, + "step": 54206 + }, + { + "epoch": 3.761501630698772, + "grad_norm": 0.932353138923645, + "learning_rate": 1.8588212186605781e-06, + "loss": 0.265, + "step": 54207 + }, + { + "epoch": 3.7615710221358682, + "grad_norm": 0.8838093876838684, + "learning_rate": 1.8577430837827592e-06, + "loss": 0.1886, + "step": 54208 + }, + { + "epoch": 3.761640413572965, + "grad_norm": 0.6960949897766113, + "learning_rate": 1.8566652587286515e-06, + "loss": 0.1822, + "step": 54209 + }, + { + "epoch": 3.761709805010062, + "grad_norm": 0.957548975944519, + "learning_rate": 1.8555877435016745e-06, + "loss": 0.2618, + "step": 54210 + }, + { + "epoch": 3.7617791964471583, + "grad_norm": 0.8312119245529175, + "learning_rate": 1.854510538105214e-06, + "loss": 0.2643, + "step": 54211 + }, + { + "epoch": 3.761848587884255, + "grad_norm": 0.5277271866798401, + "learning_rate": 1.853433642542679e-06, + "loss": 0.1678, + "step": 54212 + }, + { + "epoch": 3.7619179793213515, + "grad_norm": 0.6568945646286011, + "learning_rate": 1.8523570568174663e-06, + "loss": 0.2832, + "step": 54213 + }, + { + "epoch": 3.7619873707584484, + "grad_norm": 0.7747892141342163, + "learning_rate": 1.8512807809329847e-06, + "loss": 0.2509, + "step": 54214 + }, + { + "epoch": 3.762056762195545, + "grad_norm": 1.1316276788711548, + "learning_rate": 1.8502048148926088e-06, + "loss": 0.2575, + "step": 54215 + }, + { + "epoch": 3.7621261536326416, + "grad_norm": 0.7473551034927368, + "learning_rate": 1.8491291586997472e-06, + "loss": 0.1827, + "step": 54216 + }, + { + "epoch": 3.7621955450697384, + "grad_norm": 0.98028165102005, + "learning_rate": 1.8480538123578084e-06, + "loss": 0.3311, + "step": 54217 + }, + { + "epoch": 3.762264936506835, + "grad_norm": 0.74310702085495, + "learning_rate": 1.8469787758701784e-06, + "loss": 0.1926, + "step": 54218 + }, + { + "epoch": 3.7623343279439316, + "grad_norm": 0.7740806937217712, + "learning_rate": 1.8459040492402325e-06, + "loss": 0.2079, + "step": 54219 + }, + { + "epoch": 3.7624037193810285, + "grad_norm": 0.606454074382782, + "learning_rate": 1.8448296324713897e-06, + "loss": 0.1618, + "step": 54220 + }, + { + "epoch": 3.7624731108181253, + "grad_norm": 0.8650951385498047, + "learning_rate": 1.8437555255670368e-06, + "loss": 0.2028, + "step": 54221 + }, + { + "epoch": 3.7625425022552217, + "grad_norm": 0.6816703081130981, + "learning_rate": 1.8426817285305486e-06, + "loss": 0.2119, + "step": 54222 + }, + { + "epoch": 3.7626118936923185, + "grad_norm": 0.7758598923683167, + "learning_rate": 1.8416082413653336e-06, + "loss": 0.2416, + "step": 54223 + }, + { + "epoch": 3.762681285129415, + "grad_norm": 0.9725542664527893, + "learning_rate": 1.8405350640747665e-06, + "loss": 0.1938, + "step": 54224 + }, + { + "epoch": 3.7627506765665117, + "grad_norm": 0.7109574675559998, + "learning_rate": 1.8394621966622449e-06, + "loss": 0.1497, + "step": 54225 + }, + { + "epoch": 3.7628200680036086, + "grad_norm": 0.9425584077835083, + "learning_rate": 1.8383896391311438e-06, + "loss": 0.2209, + "step": 54226 + }, + { + "epoch": 3.762889459440705, + "grad_norm": 0.8954468369483948, + "learning_rate": 1.8373173914848718e-06, + "loss": 0.2803, + "step": 54227 + }, + { + "epoch": 3.762958850877802, + "grad_norm": 1.3211935758590698, + "learning_rate": 1.8362454537267925e-06, + "loss": 0.2387, + "step": 54228 + }, + { + "epoch": 3.763028242314898, + "grad_norm": 0.7294021844863892, + "learning_rate": 1.8351738258603035e-06, + "loss": 0.298, + "step": 54229 + }, + { + "epoch": 3.763097633751995, + "grad_norm": 1.1128227710723877, + "learning_rate": 1.8341025078887796e-06, + "loss": 0.2455, + "step": 54230 + }, + { + "epoch": 3.763167025189092, + "grad_norm": 0.5558878183364868, + "learning_rate": 1.8330314998155963e-06, + "loss": 0.1789, + "step": 54231 + }, + { + "epoch": 3.7632364166261882, + "grad_norm": 0.9802871942520142, + "learning_rate": 1.8319608016441503e-06, + "loss": 0.1623, + "step": 54232 + }, + { + "epoch": 3.763305808063285, + "grad_norm": 0.7472841739654541, + "learning_rate": 1.8308904133778059e-06, + "loss": 0.156, + "step": 54233 + }, + { + "epoch": 3.7633751995003815, + "grad_norm": 0.5728902220726013, + "learning_rate": 1.8298203350199606e-06, + "loss": 0.1548, + "step": 54234 + }, + { + "epoch": 3.7634445909374783, + "grad_norm": 1.7962802648544312, + "learning_rate": 1.828750566573978e-06, + "loss": 0.1638, + "step": 54235 + }, + { + "epoch": 3.763513982374575, + "grad_norm": 1.381272554397583, + "learning_rate": 1.8276811080432444e-06, + "loss": 0.3062, + "step": 54236 + }, + { + "epoch": 3.7635833738116715, + "grad_norm": 0.8653880953788757, + "learning_rate": 1.8266119594311238e-06, + "loss": 0.3532, + "step": 54237 + }, + { + "epoch": 3.7636527652487684, + "grad_norm": 0.7341206073760986, + "learning_rate": 1.8255431207410023e-06, + "loss": 0.2001, + "step": 54238 + }, + { + "epoch": 3.7637221566858647, + "grad_norm": 0.7083163261413574, + "learning_rate": 1.8244745919762551e-06, + "loss": 0.1975, + "step": 54239 + }, + { + "epoch": 3.7637915481229616, + "grad_norm": 0.7248185873031616, + "learning_rate": 1.8234063731402461e-06, + "loss": 0.1191, + "step": 54240 + }, + { + "epoch": 3.7638609395600584, + "grad_norm": 1.0510231256484985, + "learning_rate": 1.8223384642363617e-06, + "loss": 0.2819, + "step": 54241 + }, + { + "epoch": 3.763930330997155, + "grad_norm": 0.7413129210472107, + "learning_rate": 1.8212708652679543e-06, + "loss": 0.2319, + "step": 54242 + }, + { + "epoch": 3.7639997224342516, + "grad_norm": 1.0493943691253662, + "learning_rate": 1.8202035762384106e-06, + "loss": 0.2038, + "step": 54243 + }, + { + "epoch": 3.764069113871348, + "grad_norm": 0.9124100208282471, + "learning_rate": 1.8191365971510831e-06, + "loss": 0.3037, + "step": 54244 + }, + { + "epoch": 3.764138505308445, + "grad_norm": 0.8909865617752075, + "learning_rate": 1.8180699280093582e-06, + "loss": 0.162, + "step": 54245 + }, + { + "epoch": 3.7642078967455417, + "grad_norm": 1.1480293273925781, + "learning_rate": 1.8170035688165886e-06, + "loss": 0.3378, + "step": 54246 + }, + { + "epoch": 3.7642772881826385, + "grad_norm": 0.9640744924545288, + "learning_rate": 1.8159375195761496e-06, + "loss": 0.3144, + "step": 54247 + }, + { + "epoch": 3.764346679619735, + "grad_norm": 0.8048461079597473, + "learning_rate": 1.814871780291416e-06, + "loss": 0.2675, + "step": 54248 + }, + { + "epoch": 3.7644160710568317, + "grad_norm": 0.8837399482727051, + "learning_rate": 1.8138063509657299e-06, + "loss": 0.3105, + "step": 54249 + }, + { + "epoch": 3.764485462493928, + "grad_norm": 0.7567567229270935, + "learning_rate": 1.8127412316024661e-06, + "loss": 0.2081, + "step": 54250 + }, + { + "epoch": 3.764554853931025, + "grad_norm": 1.0557529926300049, + "learning_rate": 1.8116764222049997e-06, + "loss": 0.3094, + "step": 54251 + }, + { + "epoch": 3.764624245368122, + "grad_norm": 0.8193162679672241, + "learning_rate": 1.8106119227766616e-06, + "loss": 0.2275, + "step": 54252 + }, + { + "epoch": 3.764693636805218, + "grad_norm": 1.2081071138381958, + "learning_rate": 1.8095477333208377e-06, + "loss": 0.2221, + "step": 54253 + }, + { + "epoch": 3.764763028242315, + "grad_norm": 0.950695812702179, + "learning_rate": 1.8084838538408698e-06, + "loss": 0.2166, + "step": 54254 + }, + { + "epoch": 3.7648324196794114, + "grad_norm": 0.5770792961120605, + "learning_rate": 1.8074202843401444e-06, + "loss": 0.1265, + "step": 54255 + }, + { + "epoch": 3.7649018111165082, + "grad_norm": 1.4464796781539917, + "learning_rate": 1.8063570248219808e-06, + "loss": 0.4348, + "step": 54256 + }, + { + "epoch": 3.764971202553605, + "grad_norm": 0.7422195672988892, + "learning_rate": 1.8052940752897652e-06, + "loss": 0.2509, + "step": 54257 + }, + { + "epoch": 3.7650405939907015, + "grad_norm": 0.7812992334365845, + "learning_rate": 1.8042314357468393e-06, + "loss": 0.2896, + "step": 54258 + }, + { + "epoch": 3.7651099854277983, + "grad_norm": 0.8048163652420044, + "learning_rate": 1.8031691061965672e-06, + "loss": 0.2643, + "step": 54259 + }, + { + "epoch": 3.7651793768648947, + "grad_norm": 1.1307531595230103, + "learning_rate": 1.8021070866422906e-06, + "loss": 0.2752, + "step": 54260 + }, + { + "epoch": 3.7652487683019915, + "grad_norm": 0.8783122301101685, + "learning_rate": 1.8010453770873737e-06, + "loss": 0.2227, + "step": 54261 + }, + { + "epoch": 3.7653181597390883, + "grad_norm": 1.2211798429489136, + "learning_rate": 1.799983977535169e-06, + "loss": 0.3166, + "step": 54262 + }, + { + "epoch": 3.7653875511761847, + "grad_norm": 0.6747930645942688, + "learning_rate": 1.7989228879890185e-06, + "loss": 0.2539, + "step": 54263 + }, + { + "epoch": 3.7654569426132816, + "grad_norm": 1.07253897190094, + "learning_rate": 1.797862108452264e-06, + "loss": 0.3394, + "step": 54264 + }, + { + "epoch": 3.765526334050378, + "grad_norm": 0.8899404406547546, + "learning_rate": 1.7968016389282805e-06, + "loss": 0.2522, + "step": 54265 + }, + { + "epoch": 3.765595725487475, + "grad_norm": 0.742444634437561, + "learning_rate": 1.7957414794203876e-06, + "loss": 0.1869, + "step": 54266 + }, + { + "epoch": 3.7656651169245716, + "grad_norm": 0.842837393283844, + "learning_rate": 1.7946816299319492e-06, + "loss": 0.2903, + "step": 54267 + }, + { + "epoch": 3.765734508361668, + "grad_norm": 0.8152700066566467, + "learning_rate": 1.793622090466307e-06, + "loss": 0.1909, + "step": 54268 + }, + { + "epoch": 3.765803899798765, + "grad_norm": 0.8216423392295837, + "learning_rate": 1.7925628610268142e-06, + "loss": 0.172, + "step": 54269 + }, + { + "epoch": 3.7658732912358612, + "grad_norm": 0.8473383784294128, + "learning_rate": 1.791503941616801e-06, + "loss": 0.2681, + "step": 54270 + }, + { + "epoch": 3.765942682672958, + "grad_norm": 0.8589827418327332, + "learning_rate": 1.7904453322396097e-06, + "loss": 0.1374, + "step": 54271 + }, + { + "epoch": 3.766012074110055, + "grad_norm": 1.2174477577209473, + "learning_rate": 1.7893870328985928e-06, + "loss": 0.2083, + "step": 54272 + }, + { + "epoch": 3.7660814655471517, + "grad_norm": 2.4426021575927734, + "learning_rate": 1.7883290435970812e-06, + "loss": 0.229, + "step": 54273 + }, + { + "epoch": 3.766150856984248, + "grad_norm": 1.1888415813446045, + "learning_rate": 1.7872713643384275e-06, + "loss": 0.2025, + "step": 54274 + }, + { + "epoch": 3.766220248421345, + "grad_norm": 1.035040259361267, + "learning_rate": 1.7862139951259515e-06, + "loss": 0.2946, + "step": 54275 + }, + { + "epoch": 3.7662896398584413, + "grad_norm": 1.0286152362823486, + "learning_rate": 1.7851569359630282e-06, + "loss": 0.1884, + "step": 54276 + }, + { + "epoch": 3.766359031295538, + "grad_norm": 0.9162867069244385, + "learning_rate": 1.7841001868529438e-06, + "loss": 0.2304, + "step": 54277 + }, + { + "epoch": 3.766428422732635, + "grad_norm": 0.8297374844551086, + "learning_rate": 1.7830437477990736e-06, + "loss": 0.1825, + "step": 54278 + }, + { + "epoch": 3.7664978141697314, + "grad_norm": 1.143259882926941, + "learning_rate": 1.7819876188047368e-06, + "loss": 0.3144, + "step": 54279 + }, + { + "epoch": 3.7665672056068282, + "grad_norm": 1.3521348237991333, + "learning_rate": 1.7809317998732645e-06, + "loss": 0.3198, + "step": 54280 + }, + { + "epoch": 3.7666365970439246, + "grad_norm": 0.5707776546478271, + "learning_rate": 1.779876291007998e-06, + "loss": 0.1658, + "step": 54281 + }, + { + "epoch": 3.7667059884810214, + "grad_norm": 0.7008519172668457, + "learning_rate": 1.7788210922122572e-06, + "loss": 0.2028, + "step": 54282 + }, + { + "epoch": 3.7667753799181183, + "grad_norm": 0.9671236872673035, + "learning_rate": 1.777766203489406e-06, + "loss": 0.254, + "step": 54283 + }, + { + "epoch": 3.7668447713552147, + "grad_norm": 0.8954342007637024, + "learning_rate": 1.7767116248427308e-06, + "loss": 0.3022, + "step": 54284 + }, + { + "epoch": 3.7669141627923115, + "grad_norm": 0.814755916595459, + "learning_rate": 1.775657356275584e-06, + "loss": 0.2766, + "step": 54285 + }, + { + "epoch": 3.766983554229408, + "grad_norm": 0.9866303205490112, + "learning_rate": 1.774603397791308e-06, + "loss": 0.287, + "step": 54286 + }, + { + "epoch": 3.7670529456665047, + "grad_norm": 0.6737961173057556, + "learning_rate": 1.7735497493931997e-06, + "loss": 0.1482, + "step": 54287 + }, + { + "epoch": 3.7671223371036016, + "grad_norm": 0.7772544026374817, + "learning_rate": 1.7724964110845898e-06, + "loss": 0.3285, + "step": 54288 + }, + { + "epoch": 3.767191728540698, + "grad_norm": 1.3594845533370972, + "learning_rate": 1.7714433828688205e-06, + "loss": 0.2661, + "step": 54289 + }, + { + "epoch": 3.767261119977795, + "grad_norm": 0.7988885641098022, + "learning_rate": 1.7703906647492108e-06, + "loss": 0.1501, + "step": 54290 + }, + { + "epoch": 3.767330511414891, + "grad_norm": 1.1498373746871948, + "learning_rate": 1.7693382567290807e-06, + "loss": 0.3445, + "step": 54291 + }, + { + "epoch": 3.767399902851988, + "grad_norm": 0.8173119425773621, + "learning_rate": 1.7682861588117494e-06, + "loss": 0.1478, + "step": 54292 + }, + { + "epoch": 3.767469294289085, + "grad_norm": 0.903563916683197, + "learning_rate": 1.7672343710005369e-06, + "loss": 0.2319, + "step": 54293 + }, + { + "epoch": 3.7675386857261812, + "grad_norm": 0.894083559513092, + "learning_rate": 1.7661828932987733e-06, + "loss": 0.2405, + "step": 54294 + }, + { + "epoch": 3.767608077163278, + "grad_norm": 0.8335221409797668, + "learning_rate": 1.7651317257097677e-06, + "loss": 0.2831, + "step": 54295 + }, + { + "epoch": 3.7676774686003744, + "grad_norm": 0.9048369526863098, + "learning_rate": 1.7640808682368503e-06, + "loss": 0.2531, + "step": 54296 + }, + { + "epoch": 3.7677468600374713, + "grad_norm": 0.9154291749000549, + "learning_rate": 1.763030320883341e-06, + "loss": 0.1785, + "step": 54297 + }, + { + "epoch": 3.767816251474568, + "grad_norm": 0.9383236169815063, + "learning_rate": 1.761980083652537e-06, + "loss": 0.1811, + "step": 54298 + }, + { + "epoch": 3.767885642911665, + "grad_norm": 0.8688547611236572, + "learning_rate": 1.7609301565477576e-06, + "loss": 0.2197, + "step": 54299 + }, + { + "epoch": 3.7679550343487613, + "grad_norm": 0.9968734979629517, + "learning_rate": 1.759880539572334e-06, + "loss": 0.2827, + "step": 54300 + }, + { + "epoch": 3.768024425785858, + "grad_norm": 1.2818223237991333, + "learning_rate": 1.758831232729563e-06, + "loss": 0.3859, + "step": 54301 + }, + { + "epoch": 3.7680938172229546, + "grad_norm": 1.1306488513946533, + "learning_rate": 1.7577822360227647e-06, + "loss": 0.172, + "step": 54302 + }, + { + "epoch": 3.7681632086600514, + "grad_norm": 0.7598012089729309, + "learning_rate": 1.7567335494552583e-06, + "loss": 0.2398, + "step": 54303 + }, + { + "epoch": 3.768232600097148, + "grad_norm": 0.9574382305145264, + "learning_rate": 1.7556851730303414e-06, + "loss": 0.2836, + "step": 54304 + }, + { + "epoch": 3.7683019915342446, + "grad_norm": 1.1657159328460693, + "learning_rate": 1.7546371067513223e-06, + "loss": 0.2392, + "step": 54305 + }, + { + "epoch": 3.7683713829713414, + "grad_norm": 0.7389960289001465, + "learning_rate": 1.7535893506215205e-06, + "loss": 0.1627, + "step": 54306 + }, + { + "epoch": 3.768440774408438, + "grad_norm": 0.6409910321235657, + "learning_rate": 1.7525419046442337e-06, + "loss": 0.1583, + "step": 54307 + }, + { + "epoch": 3.7685101658455347, + "grad_norm": 1.4167978763580322, + "learning_rate": 1.7514947688227813e-06, + "loss": 0.3805, + "step": 54308 + }, + { + "epoch": 3.7685795572826315, + "grad_norm": 0.8949046730995178, + "learning_rate": 1.7504479431604492e-06, + "loss": 0.226, + "step": 54309 + }, + { + "epoch": 3.768648948719728, + "grad_norm": 0.9595112204551697, + "learning_rate": 1.7494014276605685e-06, + "loss": 0.3708, + "step": 54310 + }, + { + "epoch": 3.7687183401568247, + "grad_norm": 0.8401142358779907, + "learning_rate": 1.7483552223264254e-06, + "loss": 0.2516, + "step": 54311 + }, + { + "epoch": 3.768787731593921, + "grad_norm": 0.6257270574569702, + "learning_rate": 1.7473093271613173e-06, + "loss": 0.1362, + "step": 54312 + }, + { + "epoch": 3.768857123031018, + "grad_norm": 0.9482378959655762, + "learning_rate": 1.7462637421685525e-06, + "loss": 0.3381, + "step": 54313 + }, + { + "epoch": 3.7689265144681148, + "grad_norm": 1.189188003540039, + "learning_rate": 1.7452184673514393e-06, + "loss": 0.2995, + "step": 54314 + }, + { + "epoch": 3.768995905905211, + "grad_norm": 0.6186415553092957, + "learning_rate": 1.7441735027132754e-06, + "loss": 0.2344, + "step": 54315 + }, + { + "epoch": 3.769065297342308, + "grad_norm": 1.1560320854187012, + "learning_rate": 1.7431288482573472e-06, + "loss": 0.2429, + "step": 54316 + }, + { + "epoch": 3.7691346887794044, + "grad_norm": 1.048592448234558, + "learning_rate": 1.7420845039869738e-06, + "loss": 0.2589, + "step": 54317 + }, + { + "epoch": 3.769204080216501, + "grad_norm": 1.3788081407546997, + "learning_rate": 1.7410404699054307e-06, + "loss": 0.2209, + "step": 54318 + }, + { + "epoch": 3.769273471653598, + "grad_norm": 0.991989016532898, + "learning_rate": 1.739996746016026e-06, + "loss": 0.1942, + "step": 54319 + }, + { + "epoch": 3.769342863090695, + "grad_norm": 1.5303884744644165, + "learning_rate": 1.7389533323220463e-06, + "loss": 0.3091, + "step": 54320 + }, + { + "epoch": 3.7694122545277913, + "grad_norm": 1.3218555450439453, + "learning_rate": 1.7379102288268e-06, + "loss": 0.3768, + "step": 54321 + }, + { + "epoch": 3.7694816459648877, + "grad_norm": 1.057008147239685, + "learning_rate": 1.7368674355335623e-06, + "loss": 0.288, + "step": 54322 + }, + { + "epoch": 3.7695510374019845, + "grad_norm": 1.2374917268753052, + "learning_rate": 1.7358249524456304e-06, + "loss": 0.3178, + "step": 54323 + }, + { + "epoch": 3.7696204288390813, + "grad_norm": 0.8971812725067139, + "learning_rate": 1.7347827795663019e-06, + "loss": 0.1991, + "step": 54324 + }, + { + "epoch": 3.769689820276178, + "grad_norm": 0.7794851064682007, + "learning_rate": 1.733740916898874e-06, + "loss": 0.1931, + "step": 54325 + }, + { + "epoch": 3.7697592117132745, + "grad_norm": 1.0735937356948853, + "learning_rate": 1.7326993644466216e-06, + "loss": 0.249, + "step": 54326 + }, + { + "epoch": 3.7698286031503714, + "grad_norm": 0.7281861305236816, + "learning_rate": 1.7316581222128204e-06, + "loss": 0.1367, + "step": 54327 + }, + { + "epoch": 3.7698979945874678, + "grad_norm": 1.0631648302078247, + "learning_rate": 1.7306171902007895e-06, + "loss": 0.4073, + "step": 54328 + }, + { + "epoch": 3.7699673860245646, + "grad_norm": 0.7553849816322327, + "learning_rate": 1.7295765684137822e-06, + "loss": 0.2533, + "step": 54329 + }, + { + "epoch": 3.7700367774616614, + "grad_norm": 0.5388505458831787, + "learning_rate": 1.7285362568551177e-06, + "loss": 0.1859, + "step": 54330 + }, + { + "epoch": 3.770106168898758, + "grad_norm": 0.7734389901161194, + "learning_rate": 1.7274962555280605e-06, + "loss": 0.1932, + "step": 54331 + }, + { + "epoch": 3.7701755603358547, + "grad_norm": 1.4173822402954102, + "learning_rate": 1.7264565644358854e-06, + "loss": 0.3192, + "step": 54332 + }, + { + "epoch": 3.770244951772951, + "grad_norm": 0.8093607425689697, + "learning_rate": 1.72541718358189e-06, + "loss": 0.2745, + "step": 54333 + }, + { + "epoch": 3.770314343210048, + "grad_norm": 0.7942162752151489, + "learning_rate": 1.7243781129693493e-06, + "loss": 0.2057, + "step": 54334 + }, + { + "epoch": 3.7703837346471447, + "grad_norm": 1.0253307819366455, + "learning_rate": 1.7233393526015496e-06, + "loss": 0.2506, + "step": 54335 + }, + { + "epoch": 3.770453126084241, + "grad_norm": 0.7429133653640747, + "learning_rate": 1.722300902481755e-06, + "loss": 0.1079, + "step": 54336 + }, + { + "epoch": 3.770522517521338, + "grad_norm": 0.7411149740219116, + "learning_rate": 1.7212627626132626e-06, + "loss": 0.2095, + "step": 54337 + }, + { + "epoch": 3.7705919089584343, + "grad_norm": 0.9825676679611206, + "learning_rate": 1.720224932999348e-06, + "loss": 0.1953, + "step": 54338 + }, + { + "epoch": 3.770661300395531, + "grad_norm": 0.9268697500228882, + "learning_rate": 1.7191874136432751e-06, + "loss": 0.1745, + "step": 54339 + }, + { + "epoch": 3.770730691832628, + "grad_norm": 1.047695279121399, + "learning_rate": 1.7181502045483189e-06, + "loss": 0.2226, + "step": 54340 + }, + { + "epoch": 3.7708000832697244, + "grad_norm": 0.9113742113113403, + "learning_rate": 1.7171133057177657e-06, + "loss": 0.2786, + "step": 54341 + }, + { + "epoch": 3.770869474706821, + "grad_norm": 0.6381495594978333, + "learning_rate": 1.7160767171548797e-06, + "loss": 0.1943, + "step": 54342 + }, + { + "epoch": 3.7709388661439176, + "grad_norm": 0.8960424065589905, + "learning_rate": 1.715040438862936e-06, + "loss": 0.1715, + "step": 54343 + }, + { + "epoch": 3.7710082575810144, + "grad_norm": 0.8946181535720825, + "learning_rate": 1.7140044708451986e-06, + "loss": 0.3003, + "step": 54344 + }, + { + "epoch": 3.7710776490181113, + "grad_norm": 1.9535367488861084, + "learning_rate": 1.7129688131049647e-06, + "loss": 0.4735, + "step": 54345 + }, + { + "epoch": 3.771147040455208, + "grad_norm": 1.302553415298462, + "learning_rate": 1.7119334656454766e-06, + "loss": 0.3266, + "step": 54346 + }, + { + "epoch": 3.7712164318923045, + "grad_norm": 0.6889173984527588, + "learning_rate": 1.710898428470009e-06, + "loss": 0.2183, + "step": 54347 + }, + { + "epoch": 3.771285823329401, + "grad_norm": 0.9833865761756897, + "learning_rate": 1.7098637015818264e-06, + "loss": 0.2774, + "step": 54348 + }, + { + "epoch": 3.7713552147664977, + "grad_norm": 0.6287636756896973, + "learning_rate": 1.7088292849842036e-06, + "loss": 0.1783, + "step": 54349 + }, + { + "epoch": 3.7714246062035945, + "grad_norm": 0.9526016712188721, + "learning_rate": 1.7077951786804047e-06, + "loss": 0.2774, + "step": 54350 + }, + { + "epoch": 3.7714939976406914, + "grad_norm": 1.480248212814331, + "learning_rate": 1.7067613826736938e-06, + "loss": 0.31, + "step": 54351 + }, + { + "epoch": 3.7715633890777878, + "grad_norm": 0.8697469234466553, + "learning_rate": 1.7057278969673351e-06, + "loss": 0.2934, + "step": 54352 + }, + { + "epoch": 3.7716327805148846, + "grad_norm": 0.9734950661659241, + "learning_rate": 1.7046947215645925e-06, + "loss": 0.1795, + "step": 54353 + }, + { + "epoch": 3.771702171951981, + "grad_norm": 1.0525115728378296, + "learning_rate": 1.7036618564687078e-06, + "loss": 0.1981, + "step": 54354 + }, + { + "epoch": 3.771771563389078, + "grad_norm": 1.1123162508010864, + "learning_rate": 1.7026293016829674e-06, + "loss": 0.3199, + "step": 54355 + }, + { + "epoch": 3.7718409548261747, + "grad_norm": 0.9165191054344177, + "learning_rate": 1.7015970572106355e-06, + "loss": 0.33, + "step": 54356 + }, + { + "epoch": 3.771910346263271, + "grad_norm": 1.1394648551940918, + "learning_rate": 1.7005651230549425e-06, + "loss": 0.1777, + "step": 54357 + }, + { + "epoch": 3.771979737700368, + "grad_norm": 0.8090115189552307, + "learning_rate": 1.6995334992191637e-06, + "loss": 0.2481, + "step": 54358 + }, + { + "epoch": 3.7720491291374643, + "grad_norm": 1.0401594638824463, + "learning_rate": 1.6985021857065408e-06, + "loss": 0.3408, + "step": 54359 + }, + { + "epoch": 3.772118520574561, + "grad_norm": 0.8455870151519775, + "learning_rate": 1.6974711825203604e-06, + "loss": 0.186, + "step": 54360 + }, + { + "epoch": 3.772187912011658, + "grad_norm": 1.2274501323699951, + "learning_rate": 1.696440489663853e-06, + "loss": 0.2165, + "step": 54361 + }, + { + "epoch": 3.7722573034487543, + "grad_norm": 0.8222661018371582, + "learning_rate": 1.6954101071402717e-06, + "loss": 0.2425, + "step": 54362 + }, + { + "epoch": 3.772326694885851, + "grad_norm": 0.6785026788711548, + "learning_rate": 1.6943800349528805e-06, + "loss": 0.2956, + "step": 54363 + }, + { + "epoch": 3.7723960863229475, + "grad_norm": 1.2478781938552856, + "learning_rate": 1.6933502731049212e-06, + "loss": 0.2152, + "step": 54364 + }, + { + "epoch": 3.7724654777600444, + "grad_norm": 0.9799895882606506, + "learning_rate": 1.6923208215996467e-06, + "loss": 0.2492, + "step": 54365 + }, + { + "epoch": 3.772534869197141, + "grad_norm": 1.0317134857177734, + "learning_rate": 1.691291680440321e-06, + "loss": 0.2482, + "step": 54366 + }, + { + "epoch": 3.7726042606342376, + "grad_norm": 1.1848033666610718, + "learning_rate": 1.6902628496301753e-06, + "loss": 0.2708, + "step": 54367 + }, + { + "epoch": 3.7726736520713344, + "grad_norm": 0.541927695274353, + "learning_rate": 1.6892343291724622e-06, + "loss": 0.1664, + "step": 54368 + }, + { + "epoch": 3.772743043508431, + "grad_norm": 1.4489593505859375, + "learning_rate": 1.6882061190704234e-06, + "loss": 0.3888, + "step": 54369 + }, + { + "epoch": 3.7728124349455276, + "grad_norm": 0.8687088489532471, + "learning_rate": 1.6871782193273234e-06, + "loss": 0.2237, + "step": 54370 + }, + { + "epoch": 3.7728818263826245, + "grad_norm": 0.7889779806137085, + "learning_rate": 1.6861506299463925e-06, + "loss": 0.2873, + "step": 54371 + }, + { + "epoch": 3.7729512178197213, + "grad_norm": 0.8076956272125244, + "learning_rate": 1.6851233509308729e-06, + "loss": 0.2902, + "step": 54372 + }, + { + "epoch": 3.7730206092568177, + "grad_norm": 0.7962747812271118, + "learning_rate": 1.6840963822840172e-06, + "loss": 0.1777, + "step": 54373 + }, + { + "epoch": 3.7730900006939145, + "grad_norm": 0.814664363861084, + "learning_rate": 1.6830697240090676e-06, + "loss": 0.2003, + "step": 54374 + }, + { + "epoch": 3.773159392131011, + "grad_norm": 0.7107864618301392, + "learning_rate": 1.6820433761092435e-06, + "loss": 0.1617, + "step": 54375 + }, + { + "epoch": 3.7732287835681078, + "grad_norm": 0.6736572980880737, + "learning_rate": 1.6810173385878202e-06, + "loss": 0.1405, + "step": 54376 + }, + { + "epoch": 3.7732981750052046, + "grad_norm": 1.1667660474777222, + "learning_rate": 1.6799916114480063e-06, + "loss": 0.4099, + "step": 54377 + }, + { + "epoch": 3.773367566442301, + "grad_norm": 0.8101360201835632, + "learning_rate": 1.6789661946930545e-06, + "loss": 0.3284, + "step": 54378 + }, + { + "epoch": 3.773436957879398, + "grad_norm": 0.5928885340690613, + "learning_rate": 1.6779410883261958e-06, + "loss": 0.1636, + "step": 54379 + }, + { + "epoch": 3.773506349316494, + "grad_norm": 0.8402543663978577, + "learning_rate": 1.676916292350683e-06, + "loss": 0.2005, + "step": 54380 + }, + { + "epoch": 3.773575740753591, + "grad_norm": 1.005484938621521, + "learning_rate": 1.675891806769725e-06, + "loss": 0.2817, + "step": 54381 + }, + { + "epoch": 3.773645132190688, + "grad_norm": 0.6403902173042297, + "learning_rate": 1.674867631586563e-06, + "loss": 0.1483, + "step": 54382 + }, + { + "epoch": 3.7737145236277843, + "grad_norm": 1.1524392366409302, + "learning_rate": 1.6738437668044504e-06, + "loss": 0.2477, + "step": 54383 + }, + { + "epoch": 3.773783915064881, + "grad_norm": 0.7952528595924377, + "learning_rate": 1.6728202124265958e-06, + "loss": 0.202, + "step": 54384 + }, + { + "epoch": 3.7738533065019775, + "grad_norm": 0.9767169952392578, + "learning_rate": 1.6717969684562406e-06, + "loss": 0.349, + "step": 54385 + }, + { + "epoch": 3.7739226979390743, + "grad_norm": 0.6814472675323486, + "learning_rate": 1.6707740348966161e-06, + "loss": 0.2537, + "step": 54386 + }, + { + "epoch": 3.773992089376171, + "grad_norm": 0.7159781455993652, + "learning_rate": 1.6697514117509639e-06, + "loss": 0.253, + "step": 54387 + }, + { + "epoch": 3.7740614808132675, + "grad_norm": 1.0381529331207275, + "learning_rate": 1.6687290990224813e-06, + "loss": 0.181, + "step": 54388 + }, + { + "epoch": 3.7741308722503644, + "grad_norm": 0.8509081602096558, + "learning_rate": 1.6677070967144103e-06, + "loss": 0.1769, + "step": 54389 + }, + { + "epoch": 3.7742002636874608, + "grad_norm": 0.820517361164093, + "learning_rate": 1.6666854048299818e-06, + "loss": 0.2135, + "step": 54390 + }, + { + "epoch": 3.7742696551245576, + "grad_norm": 1.133272647857666, + "learning_rate": 1.6656640233724262e-06, + "loss": 0.366, + "step": 54391 + }, + { + "epoch": 3.7743390465616544, + "grad_norm": 0.8832557201385498, + "learning_rate": 1.6646429523449636e-06, + "loss": 0.2519, + "step": 54392 + }, + { + "epoch": 3.774408437998751, + "grad_norm": 0.8597891926765442, + "learning_rate": 1.6636221917508021e-06, + "loss": 0.4017, + "step": 54393 + }, + { + "epoch": 3.7744778294358476, + "grad_norm": 0.8562857508659363, + "learning_rate": 1.6626017415931838e-06, + "loss": 0.2126, + "step": 54394 + }, + { + "epoch": 3.774547220872944, + "grad_norm": 1.2746086120605469, + "learning_rate": 1.6615816018753172e-06, + "loss": 0.2164, + "step": 54395 + }, + { + "epoch": 3.774616612310041, + "grad_norm": 0.8190780282020569, + "learning_rate": 1.660561772600433e-06, + "loss": 0.2279, + "step": 54396 + }, + { + "epoch": 3.7746860037471377, + "grad_norm": 0.8167960047721863, + "learning_rate": 1.6595422537717398e-06, + "loss": 0.2515, + "step": 54397 + }, + { + "epoch": 3.7747553951842345, + "grad_norm": 0.991408109664917, + "learning_rate": 1.6585230453924682e-06, + "loss": 0.2363, + "step": 54398 + }, + { + "epoch": 3.774824786621331, + "grad_norm": 0.5706322193145752, + "learning_rate": 1.657504147465827e-06, + "loss": 0.146, + "step": 54399 + }, + { + "epoch": 3.7748941780584278, + "grad_norm": 0.7787472009658813, + "learning_rate": 1.6564855599950358e-06, + "loss": 0.2688, + "step": 54400 + }, + { + "epoch": 3.774963569495524, + "grad_norm": 0.9926820397377014, + "learning_rate": 1.655467282983314e-06, + "loss": 0.1663, + "step": 54401 + }, + { + "epoch": 3.775032960932621, + "grad_norm": 1.2360304594039917, + "learning_rate": 1.6544493164338593e-06, + "loss": 0.3371, + "step": 54402 + }, + { + "epoch": 3.775102352369718, + "grad_norm": 0.8187994956970215, + "learning_rate": 1.6534316603499133e-06, + "loss": 0.2858, + "step": 54403 + }, + { + "epoch": 3.775171743806814, + "grad_norm": 0.8538100719451904, + "learning_rate": 1.6524143147346627e-06, + "loss": 0.2695, + "step": 54404 + }, + { + "epoch": 3.775241135243911, + "grad_norm": 0.7111048102378845, + "learning_rate": 1.6513972795913268e-06, + "loss": 0.2124, + "step": 54405 + }, + { + "epoch": 3.7753105266810074, + "grad_norm": 0.809055507183075, + "learning_rate": 1.6503805549231254e-06, + "loss": 0.1349, + "step": 54406 + }, + { + "epoch": 3.7753799181181042, + "grad_norm": 0.9141103029251099, + "learning_rate": 1.649364140733267e-06, + "loss": 0.2889, + "step": 54407 + }, + { + "epoch": 3.775449309555201, + "grad_norm": 1.5403159856796265, + "learning_rate": 1.6483480370249494e-06, + "loss": 0.2039, + "step": 54408 + }, + { + "epoch": 3.7755187009922975, + "grad_norm": 0.8605997562408447, + "learning_rate": 1.6473322438013805e-06, + "loss": 0.2537, + "step": 54409 + }, + { + "epoch": 3.7755880924293943, + "grad_norm": 1.0417262315750122, + "learning_rate": 1.6463167610657803e-06, + "loss": 0.4333, + "step": 54410 + }, + { + "epoch": 3.7756574838664907, + "grad_norm": 0.7847189903259277, + "learning_rate": 1.6453015888213462e-06, + "loss": 0.1812, + "step": 54411 + }, + { + "epoch": 3.7757268753035875, + "grad_norm": 1.0691571235656738, + "learning_rate": 1.6442867270712758e-06, + "loss": 0.3078, + "step": 54412 + }, + { + "epoch": 3.7757962667406844, + "grad_norm": 1.2852284908294678, + "learning_rate": 1.6432721758187886e-06, + "loss": 0.2479, + "step": 54413 + }, + { + "epoch": 3.7758656581777807, + "grad_norm": 0.8195958137512207, + "learning_rate": 1.642257935067093e-06, + "loss": 0.2145, + "step": 54414 + }, + { + "epoch": 3.7759350496148776, + "grad_norm": 0.798585057258606, + "learning_rate": 1.6412440048193534e-06, + "loss": 0.231, + "step": 54415 + }, + { + "epoch": 3.776004441051974, + "grad_norm": 0.9430088400840759, + "learning_rate": 1.6402303850788115e-06, + "loss": 0.1891, + "step": 54416 + }, + { + "epoch": 3.776073832489071, + "grad_norm": 1.04603910446167, + "learning_rate": 1.6392170758486424e-06, + "loss": 0.3468, + "step": 54417 + }, + { + "epoch": 3.7761432239261676, + "grad_norm": 1.158394455909729, + "learning_rate": 1.638204077132055e-06, + "loss": 0.3004, + "step": 54418 + }, + { + "epoch": 3.776212615363264, + "grad_norm": 0.9445933103561401, + "learning_rate": 1.6371913889322354e-06, + "loss": 0.1583, + "step": 54419 + }, + { + "epoch": 3.776282006800361, + "grad_norm": 0.872571587562561, + "learning_rate": 1.6361790112524034e-06, + "loss": 0.2614, + "step": 54420 + }, + { + "epoch": 3.7763513982374572, + "grad_norm": 0.897761881351471, + "learning_rate": 1.6351669440957452e-06, + "loss": 0.2586, + "step": 54421 + }, + { + "epoch": 3.776420789674554, + "grad_norm": 0.5805349946022034, + "learning_rate": 1.6341551874654358e-06, + "loss": 0.2703, + "step": 54422 + }, + { + "epoch": 3.776490181111651, + "grad_norm": 0.8592770099639893, + "learning_rate": 1.6331437413646956e-06, + "loss": 0.2458, + "step": 54423 + }, + { + "epoch": 3.7765595725487477, + "grad_norm": 1.1676238775253296, + "learning_rate": 1.63213260579671e-06, + "loss": 0.2662, + "step": 54424 + }, + { + "epoch": 3.776628963985844, + "grad_norm": 0.73078852891922, + "learning_rate": 1.6311217807646661e-06, + "loss": 0.1477, + "step": 54425 + }, + { + "epoch": 3.776698355422941, + "grad_norm": 1.5360336303710938, + "learning_rate": 1.630111266271761e-06, + "loss": 0.3405, + "step": 54426 + }, + { + "epoch": 3.7767677468600374, + "grad_norm": 0.8752927780151367, + "learning_rate": 1.6291010623211812e-06, + "loss": 0.2484, + "step": 54427 + }, + { + "epoch": 3.776837138297134, + "grad_norm": 0.8273256421089172, + "learning_rate": 1.6280911689161128e-06, + "loss": 0.2471, + "step": 54428 + }, + { + "epoch": 3.776906529734231, + "grad_norm": 0.9376543760299683, + "learning_rate": 1.6270815860597533e-06, + "loss": 0.2633, + "step": 54429 + }, + { + "epoch": 3.7769759211713274, + "grad_norm": 0.979447066783905, + "learning_rate": 1.626072313755278e-06, + "loss": 0.2527, + "step": 54430 + }, + { + "epoch": 3.7770453126084242, + "grad_norm": 1.7900915145874023, + "learning_rate": 1.6250633520058845e-06, + "loss": 0.277, + "step": 54431 + }, + { + "epoch": 3.7771147040455206, + "grad_norm": 0.7114477157592773, + "learning_rate": 1.6240547008147477e-06, + "loss": 0.2107, + "step": 54432 + }, + { + "epoch": 3.7771840954826175, + "grad_norm": 0.8921208381652832, + "learning_rate": 1.6230463601850542e-06, + "loss": 0.2338, + "step": 54433 + }, + { + "epoch": 3.7772534869197143, + "grad_norm": 0.8285020589828491, + "learning_rate": 1.6220383301199905e-06, + "loss": 0.2262, + "step": 54434 + }, + { + "epoch": 3.7773228783568107, + "grad_norm": 0.7072665095329285, + "learning_rate": 1.6210306106227423e-06, + "loss": 0.2521, + "step": 54435 + }, + { + "epoch": 3.7773922697939075, + "grad_norm": 0.8134350776672363, + "learning_rate": 1.6200232016964855e-06, + "loss": 0.252, + "step": 54436 + }, + { + "epoch": 3.777461661231004, + "grad_norm": 0.8332840800285339, + "learning_rate": 1.619016103344395e-06, + "loss": 0.2738, + "step": 54437 + }, + { + "epoch": 3.7775310526681007, + "grad_norm": 0.6962894797325134, + "learning_rate": 1.6180093155696574e-06, + "loss": 0.1783, + "step": 54438 + }, + { + "epoch": 3.7776004441051976, + "grad_norm": 0.6976459622383118, + "learning_rate": 1.6170028383754586e-06, + "loss": 0.2454, + "step": 54439 + }, + { + "epoch": 3.777669835542294, + "grad_norm": 0.719194769859314, + "learning_rate": 1.6159966717649632e-06, + "loss": 0.2063, + "step": 54440 + }, + { + "epoch": 3.777739226979391, + "grad_norm": 1.0631892681121826, + "learning_rate": 1.6149908157413463e-06, + "loss": 0.2919, + "step": 54441 + }, + { + "epoch": 3.777808618416487, + "grad_norm": 0.6251022815704346, + "learning_rate": 1.613985270307794e-06, + "loss": 0.2083, + "step": 54442 + }, + { + "epoch": 3.777878009853584, + "grad_norm": 0.954646110534668, + "learning_rate": 1.6129800354674819e-06, + "loss": 0.483, + "step": 54443 + }, + { + "epoch": 3.777947401290681, + "grad_norm": 1.0609233379364014, + "learning_rate": 1.6119751112235625e-06, + "loss": 0.1475, + "step": 54444 + }, + { + "epoch": 3.7780167927277772, + "grad_norm": 0.58182692527771, + "learning_rate": 1.6109704975792229e-06, + "loss": 0.2466, + "step": 54445 + }, + { + "epoch": 3.778086184164874, + "grad_norm": 0.6978595852851868, + "learning_rate": 1.6099661945376487e-06, + "loss": 0.2442, + "step": 54446 + }, + { + "epoch": 3.7781555756019705, + "grad_norm": 1.326074242591858, + "learning_rate": 1.6089622021019824e-06, + "loss": 0.3351, + "step": 54447 + }, + { + "epoch": 3.7782249670390673, + "grad_norm": 0.8149693608283997, + "learning_rate": 1.60795852027541e-06, + "loss": 0.2775, + "step": 54448 + }, + { + "epoch": 3.778294358476164, + "grad_norm": 1.1714776754379272, + "learning_rate": 1.6069551490611067e-06, + "loss": 0.2257, + "step": 54449 + }, + { + "epoch": 3.778363749913261, + "grad_norm": 1.1689536571502686, + "learning_rate": 1.6059520884622259e-06, + "loss": 0.2419, + "step": 54450 + }, + { + "epoch": 3.7784331413503573, + "grad_norm": 0.9677242636680603, + "learning_rate": 1.6049493384819314e-06, + "loss": 0.3003, + "step": 54451 + }, + { + "epoch": 3.778502532787454, + "grad_norm": 0.8202659487724304, + "learning_rate": 1.6039468991233985e-06, + "loss": 0.2632, + "step": 54452 + }, + { + "epoch": 3.7785719242245506, + "grad_norm": 1.1333532333374023, + "learning_rate": 1.6029447703897916e-06, + "loss": 0.2421, + "step": 54453 + }, + { + "epoch": 3.7786413156616474, + "grad_norm": 1.4697128534317017, + "learning_rate": 1.6019429522842632e-06, + "loss": 0.2254, + "step": 54454 + }, + { + "epoch": 3.7787107070987442, + "grad_norm": 1.0786164999008179, + "learning_rate": 1.600941444809989e-06, + "loss": 0.3381, + "step": 54455 + }, + { + "epoch": 3.7787800985358406, + "grad_norm": 1.0505138635635376, + "learning_rate": 1.5999402479701442e-06, + "loss": 0.2838, + "step": 54456 + }, + { + "epoch": 3.7788494899729375, + "grad_norm": 1.0297858715057373, + "learning_rate": 1.5989393617678483e-06, + "loss": 0.2396, + "step": 54457 + }, + { + "epoch": 3.778918881410034, + "grad_norm": 1.1552073955535889, + "learning_rate": 1.597938786206299e-06, + "loss": 0.3255, + "step": 54458 + }, + { + "epoch": 3.7789882728471307, + "grad_norm": 1.0667695999145508, + "learning_rate": 1.5969385212886267e-06, + "loss": 0.288, + "step": 54459 + }, + { + "epoch": 3.7790576642842275, + "grad_norm": 1.0055288076400757, + "learning_rate": 1.5959385670180182e-06, + "loss": 0.2845, + "step": 54460 + }, + { + "epoch": 3.779127055721324, + "grad_norm": 1.102575659751892, + "learning_rate": 1.5949389233976043e-06, + "loss": 0.3971, + "step": 54461 + }, + { + "epoch": 3.7791964471584207, + "grad_norm": 1.0876200199127197, + "learning_rate": 1.5939395904305487e-06, + "loss": 0.3313, + "step": 54462 + }, + { + "epoch": 3.779265838595517, + "grad_norm": 1.2232835292816162, + "learning_rate": 1.5929405681200161e-06, + "loss": 0.1728, + "step": 54463 + }, + { + "epoch": 3.779335230032614, + "grad_norm": 0.8324691653251648, + "learning_rate": 1.591941856469159e-06, + "loss": 0.2368, + "step": 54464 + }, + { + "epoch": 3.779404621469711, + "grad_norm": 0.8591905832290649, + "learning_rate": 1.5909434554811086e-06, + "loss": 0.2252, + "step": 54465 + }, + { + "epoch": 3.779474012906807, + "grad_norm": 1.253714680671692, + "learning_rate": 1.5899453651590401e-06, + "loss": 0.1607, + "step": 54466 + }, + { + "epoch": 3.779543404343904, + "grad_norm": 0.8364745378494263, + "learning_rate": 1.5889475855060953e-06, + "loss": 0.2378, + "step": 54467 + }, + { + "epoch": 3.7796127957810004, + "grad_norm": 0.7763926982879639, + "learning_rate": 1.5879501165254162e-06, + "loss": 0.2494, + "step": 54468 + }, + { + "epoch": 3.7796821872180972, + "grad_norm": 0.7003236413002014, + "learning_rate": 1.586952958220178e-06, + "loss": 0.1657, + "step": 54469 + }, + { + "epoch": 3.779751578655194, + "grad_norm": 1.1553274393081665, + "learning_rate": 1.5859561105935005e-06, + "loss": 0.3169, + "step": 54470 + }, + { + "epoch": 3.7798209700922905, + "grad_norm": 0.9118553996086121, + "learning_rate": 1.5849595736485368e-06, + "loss": 0.3627, + "step": 54471 + }, + { + "epoch": 3.7798903615293873, + "grad_norm": 0.6615209579467773, + "learning_rate": 1.5839633473884507e-06, + "loss": 0.1885, + "step": 54472 + }, + { + "epoch": 3.7799597529664837, + "grad_norm": 0.6653012633323669, + "learning_rate": 1.5829674318163624e-06, + "loss": 0.118, + "step": 54473 + }, + { + "epoch": 3.7800291444035805, + "grad_norm": 1.3164856433868408, + "learning_rate": 1.5819718269354356e-06, + "loss": 0.3124, + "step": 54474 + }, + { + "epoch": 3.7800985358406773, + "grad_norm": 1.0539218187332153, + "learning_rate": 1.5809765327487902e-06, + "loss": 0.2733, + "step": 54475 + }, + { + "epoch": 3.780167927277774, + "grad_norm": 0.8221397995948792, + "learning_rate": 1.5799815492596014e-06, + "loss": 0.2527, + "step": 54476 + }, + { + "epoch": 3.7802373187148706, + "grad_norm": 0.709206759929657, + "learning_rate": 1.5789868764709892e-06, + "loss": 0.1396, + "step": 54477 + }, + { + "epoch": 3.7803067101519674, + "grad_norm": 0.7469072341918945, + "learning_rate": 1.5779925143860951e-06, + "loss": 0.1346, + "step": 54478 + }, + { + "epoch": 3.780376101589064, + "grad_norm": 0.850879430770874, + "learning_rate": 1.5769984630080503e-06, + "loss": 0.2513, + "step": 54479 + }, + { + "epoch": 3.7804454930261606, + "grad_norm": 0.9896536469459534, + "learning_rate": 1.5760047223400077e-06, + "loss": 0.309, + "step": 54480 + }, + { + "epoch": 3.7805148844632575, + "grad_norm": 0.8679407238960266, + "learning_rate": 1.5750112923850979e-06, + "loss": 0.252, + "step": 54481 + }, + { + "epoch": 3.780584275900354, + "grad_norm": 0.9856100678443909, + "learning_rate": 1.574018173146463e-06, + "loss": 0.3283, + "step": 54482 + }, + { + "epoch": 3.7806536673374507, + "grad_norm": 0.7161268591880798, + "learning_rate": 1.5730253646272342e-06, + "loss": 0.2016, + "step": 54483 + }, + { + "epoch": 3.780723058774547, + "grad_norm": 1.2692346572875977, + "learning_rate": 1.5720328668305418e-06, + "loss": 0.2825, + "step": 54484 + }, + { + "epoch": 3.780792450211644, + "grad_norm": 1.3839316368103027, + "learning_rate": 1.5710406797595168e-06, + "loss": 0.2499, + "step": 54485 + }, + { + "epoch": 3.7808618416487407, + "grad_norm": 0.7833033204078674, + "learning_rate": 1.570048803417301e-06, + "loss": 0.272, + "step": 54486 + }, + { + "epoch": 3.780931233085837, + "grad_norm": 0.8630238771438599, + "learning_rate": 1.5690572378070146e-06, + "loss": 0.2168, + "step": 54487 + }, + { + "epoch": 3.781000624522934, + "grad_norm": 1.0262587070465088, + "learning_rate": 1.568065982931799e-06, + "loss": 0.2106, + "step": 54488 + }, + { + "epoch": 3.7810700159600303, + "grad_norm": 0.8609241843223572, + "learning_rate": 1.5670750387947853e-06, + "loss": 0.1967, + "step": 54489 + }, + { + "epoch": 3.781139407397127, + "grad_norm": 1.4591253995895386, + "learning_rate": 1.5660844053990931e-06, + "loss": 0.3156, + "step": 54490 + }, + { + "epoch": 3.781208798834224, + "grad_norm": 0.8264771699905396, + "learning_rate": 1.5650940827478532e-06, + "loss": 0.2249, + "step": 54491 + }, + { + "epoch": 3.7812781902713204, + "grad_norm": 0.669137179851532, + "learning_rate": 1.5641040708441856e-06, + "loss": 0.1891, + "step": 54492 + }, + { + "epoch": 3.7813475817084172, + "grad_norm": 1.1275954246520996, + "learning_rate": 1.5631143696912209e-06, + "loss": 0.2795, + "step": 54493 + }, + { + "epoch": 3.7814169731455136, + "grad_norm": 0.9701135158538818, + "learning_rate": 1.562124979292079e-06, + "loss": 0.2214, + "step": 54494 + }, + { + "epoch": 3.7814863645826104, + "grad_norm": 0.8714630603790283, + "learning_rate": 1.5611358996499016e-06, + "loss": 0.3518, + "step": 54495 + }, + { + "epoch": 3.7815557560197073, + "grad_norm": 0.6497825384140015, + "learning_rate": 1.5601471307677862e-06, + "loss": 0.1977, + "step": 54496 + }, + { + "epoch": 3.7816251474568037, + "grad_norm": 0.8735904097557068, + "learning_rate": 1.559158672648864e-06, + "loss": 0.2214, + "step": 54497 + }, + { + "epoch": 3.7816945388939005, + "grad_norm": 1.3256700038909912, + "learning_rate": 1.558170525296254e-06, + "loss": 0.3056, + "step": 54498 + }, + { + "epoch": 3.781763930330997, + "grad_norm": 1.0177148580551147, + "learning_rate": 1.5571826887130992e-06, + "loss": 0.2438, + "step": 54499 + }, + { + "epoch": 3.7818333217680937, + "grad_norm": 0.6250877976417542, + "learning_rate": 1.556195162902474e-06, + "loss": 0.1519, + "step": 54500 + }, + { + "epoch": 3.7819027132051906, + "grad_norm": 0.5697463750839233, + "learning_rate": 1.5552079478675319e-06, + "loss": 0.2134, + "step": 54501 + }, + { + "epoch": 3.7819721046422874, + "grad_norm": 0.963771402835846, + "learning_rate": 1.5542210436113703e-06, + "loss": 0.377, + "step": 54502 + }, + { + "epoch": 3.782041496079384, + "grad_norm": 1.3753242492675781, + "learning_rate": 1.5532344501371088e-06, + "loss": 0.1883, + "step": 54503 + }, + { + "epoch": 3.7821108875164806, + "grad_norm": 0.8999343514442444, + "learning_rate": 1.5522481674478785e-06, + "loss": 0.1874, + "step": 54504 + }, + { + "epoch": 3.782180278953577, + "grad_norm": 1.1534361839294434, + "learning_rate": 1.5512621955467655e-06, + "loss": 0.3633, + "step": 54505 + }, + { + "epoch": 3.782249670390674, + "grad_norm": 0.6137667298316956, + "learning_rate": 1.5502765344369009e-06, + "loss": 0.2051, + "step": 54506 + }, + { + "epoch": 3.7823190618277707, + "grad_norm": 0.9770963788032532, + "learning_rate": 1.5492911841213819e-06, + "loss": 0.2208, + "step": 54507 + }, + { + "epoch": 3.782388453264867, + "grad_norm": 0.8643152117729187, + "learning_rate": 1.5483061446033398e-06, + "loss": 0.1772, + "step": 54508 + }, + { + "epoch": 3.782457844701964, + "grad_norm": 0.7366170287132263, + "learning_rate": 1.5473214158858606e-06, + "loss": 0.1478, + "step": 54509 + }, + { + "epoch": 3.7825272361390603, + "grad_norm": 1.0380985736846924, + "learning_rate": 1.5463369979720755e-06, + "loss": 0.2252, + "step": 54510 + }, + { + "epoch": 3.782596627576157, + "grad_norm": 0.7954990863800049, + "learning_rate": 1.5453528908650816e-06, + "loss": 0.2067, + "step": 54511 + }, + { + "epoch": 3.782666019013254, + "grad_norm": 1.2735230922698975, + "learning_rate": 1.5443690945679878e-06, + "loss": 0.3559, + "step": 54512 + }, + { + "epoch": 3.7827354104503503, + "grad_norm": 1.0690265893936157, + "learning_rate": 1.5433856090838915e-06, + "loss": 0.2599, + "step": 54513 + }, + { + "epoch": 3.782804801887447, + "grad_norm": 0.9859316945075989, + "learning_rate": 1.5424024344159016e-06, + "loss": 0.2779, + "step": 54514 + }, + { + "epoch": 3.7828741933245436, + "grad_norm": 0.799727201461792, + "learning_rate": 1.5414195705671265e-06, + "loss": 0.215, + "step": 54515 + }, + { + "epoch": 3.7829435847616404, + "grad_norm": 0.7599495649337769, + "learning_rate": 1.5404370175406635e-06, + "loss": 0.2445, + "step": 54516 + }, + { + "epoch": 3.783012976198737, + "grad_norm": 0.8650110960006714, + "learning_rate": 1.5394547753396217e-06, + "loss": 0.2335, + "step": 54517 + }, + { + "epoch": 3.7830823676358336, + "grad_norm": 0.7885535359382629, + "learning_rate": 1.5384728439670982e-06, + "loss": 0.1505, + "step": 54518 + }, + { + "epoch": 3.7831517590729304, + "grad_norm": 0.6132099628448486, + "learning_rate": 1.5374912234261907e-06, + "loss": 0.1804, + "step": 54519 + }, + { + "epoch": 3.783221150510027, + "grad_norm": 0.9264833927154541, + "learning_rate": 1.536509913720008e-06, + "loss": 0.2668, + "step": 54520 + }, + { + "epoch": 3.7832905419471237, + "grad_norm": 1.027267336845398, + "learning_rate": 1.535528914851625e-06, + "loss": 0.4246, + "step": 54521 + }, + { + "epoch": 3.7833599333842205, + "grad_norm": 1.1829643249511719, + "learning_rate": 1.5345482268241617e-06, + "loss": 0.3898, + "step": 54522 + }, + { + "epoch": 3.783429324821317, + "grad_norm": 0.8949524164199829, + "learning_rate": 1.5335678496407046e-06, + "loss": 0.2401, + "step": 54523 + }, + { + "epoch": 3.7834987162584137, + "grad_norm": 0.7237903475761414, + "learning_rate": 1.53258778330434e-06, + "loss": 0.1874, + "step": 54524 + }, + { + "epoch": 3.78356810769551, + "grad_norm": 0.9309185147285461, + "learning_rate": 1.5316080278181877e-06, + "loss": 0.2027, + "step": 54525 + }, + { + "epoch": 3.783637499132607, + "grad_norm": 0.8639493584632874, + "learning_rate": 1.5306285831853228e-06, + "loss": 0.1717, + "step": 54526 + }, + { + "epoch": 3.7837068905697038, + "grad_norm": 0.7017540335655212, + "learning_rate": 1.529649449408832e-06, + "loss": 0.1376, + "step": 54527 + }, + { + "epoch": 3.7837762820068006, + "grad_norm": 0.7057580947875977, + "learning_rate": 1.5286706264918128e-06, + "loss": 0.2267, + "step": 54528 + }, + { + "epoch": 3.783845673443897, + "grad_norm": 0.6444911360740662, + "learning_rate": 1.5276921144373513e-06, + "loss": 0.1854, + "step": 54529 + }, + { + "epoch": 3.783915064880994, + "grad_norm": 3.581634283065796, + "learning_rate": 1.5267139132485564e-06, + "loss": 0.4094, + "step": 54530 + }, + { + "epoch": 3.78398445631809, + "grad_norm": 1.3452457189559937, + "learning_rate": 1.5257360229284922e-06, + "loss": 0.3993, + "step": 54531 + }, + { + "epoch": 3.784053847755187, + "grad_norm": 1.008147120475769, + "learning_rate": 1.5247584434802454e-06, + "loss": 0.3137, + "step": 54532 + }, + { + "epoch": 3.784123239192284, + "grad_norm": 0.8796692490577698, + "learning_rate": 1.5237811749069241e-06, + "loss": 0.2536, + "step": 54533 + }, + { + "epoch": 3.7841926306293803, + "grad_norm": 0.7686660289764404, + "learning_rate": 1.522804217211593e-06, + "loss": 0.2966, + "step": 54534 + }, + { + "epoch": 3.784262022066477, + "grad_norm": 0.7842747569084167, + "learning_rate": 1.5218275703973496e-06, + "loss": 0.2902, + "step": 54535 + }, + { + "epoch": 3.7843314135035735, + "grad_norm": 0.7860211730003357, + "learning_rate": 1.5208512344672687e-06, + "loss": 0.2395, + "step": 54536 + }, + { + "epoch": 3.7844008049406703, + "grad_norm": 0.7194384336471558, + "learning_rate": 1.5198752094244262e-06, + "loss": 0.2095, + "step": 54537 + }, + { + "epoch": 3.784470196377767, + "grad_norm": 1.0002610683441162, + "learning_rate": 1.5188994952719193e-06, + "loss": 0.2674, + "step": 54538 + }, + { + "epoch": 3.7845395878148635, + "grad_norm": 0.6182748079299927, + "learning_rate": 1.5179240920128235e-06, + "loss": 0.2202, + "step": 54539 + }, + { + "epoch": 3.7846089792519604, + "grad_norm": 0.9258273243904114, + "learning_rate": 1.516948999650214e-06, + "loss": 0.2701, + "step": 54540 + }, + { + "epoch": 3.7846783706890568, + "grad_norm": 0.6939378976821899, + "learning_rate": 1.5159742181871662e-06, + "loss": 0.2129, + "step": 54541 + }, + { + "epoch": 3.7847477621261536, + "grad_norm": 0.8125394582748413, + "learning_rate": 1.5149997476267664e-06, + "loss": 0.1976, + "step": 54542 + }, + { + "epoch": 3.7848171535632504, + "grad_norm": 0.7935453653335571, + "learning_rate": 1.514025587972079e-06, + "loss": 0.2349, + "step": 54543 + }, + { + "epoch": 3.784886545000347, + "grad_norm": 0.8670608401298523, + "learning_rate": 1.5130517392261901e-06, + "loss": 0.2787, + "step": 54544 + }, + { + "epoch": 3.7849559364374437, + "grad_norm": 0.8339297771453857, + "learning_rate": 1.5120782013921753e-06, + "loss": 0.1818, + "step": 54545 + }, + { + "epoch": 3.78502532787454, + "grad_norm": 0.8761916160583496, + "learning_rate": 1.5111049744730987e-06, + "loss": 0.3175, + "step": 54546 + }, + { + "epoch": 3.785094719311637, + "grad_norm": 0.7869735956192017, + "learning_rate": 1.5101320584720358e-06, + "loss": 0.2225, + "step": 54547 + }, + { + "epoch": 3.7851641107487337, + "grad_norm": 0.6843913197517395, + "learning_rate": 1.5091594533920505e-06, + "loss": 0.2606, + "step": 54548 + }, + { + "epoch": 3.7852335021858305, + "grad_norm": 0.8216644525527954, + "learning_rate": 1.5081871592362296e-06, + "loss": 0.3076, + "step": 54549 + }, + { + "epoch": 3.785302893622927, + "grad_norm": 1.320137619972229, + "learning_rate": 1.507215176007637e-06, + "loss": 0.2509, + "step": 54550 + }, + { + "epoch": 3.7853722850600233, + "grad_norm": 0.8837334513664246, + "learning_rate": 1.506243503709337e-06, + "loss": 0.2171, + "step": 54551 + }, + { + "epoch": 3.78544167649712, + "grad_norm": 1.12666916847229, + "learning_rate": 1.505272142344405e-06, + "loss": 0.3047, + "step": 54552 + }, + { + "epoch": 3.785511067934217, + "grad_norm": 0.9909575581550598, + "learning_rate": 1.504301091915894e-06, + "loss": 0.3354, + "step": 54553 + }, + { + "epoch": 3.785580459371314, + "grad_norm": 0.9370793104171753, + "learning_rate": 1.5033303524268795e-06, + "loss": 0.1531, + "step": 54554 + }, + { + "epoch": 3.78564985080841, + "grad_norm": 1.11983060836792, + "learning_rate": 1.5023599238804253e-06, + "loss": 0.3785, + "step": 54555 + }, + { + "epoch": 3.785719242245507, + "grad_norm": 1.1811914443969727, + "learning_rate": 1.501389806279585e-06, + "loss": 0.3175, + "step": 54556 + }, + { + "epoch": 3.7857886336826034, + "grad_norm": 1.0066624879837036, + "learning_rate": 1.5004199996274338e-06, + "loss": 0.1824, + "step": 54557 + }, + { + "epoch": 3.7858580251197003, + "grad_norm": 1.0029282569885254, + "learning_rate": 1.4994505039270358e-06, + "loss": 0.1255, + "step": 54558 + }, + { + "epoch": 3.785927416556797, + "grad_norm": 1.1188900470733643, + "learning_rate": 1.498481319181444e-06, + "loss": 0.3922, + "step": 54559 + }, + { + "epoch": 3.7859968079938935, + "grad_norm": 0.7463017702102661, + "learning_rate": 1.4975124453937117e-06, + "loss": 0.2028, + "step": 54560 + }, + { + "epoch": 3.7860661994309903, + "grad_norm": 0.8181827664375305, + "learning_rate": 1.4965438825669032e-06, + "loss": 0.2394, + "step": 54561 + }, + { + "epoch": 3.7861355908680867, + "grad_norm": 0.7580156922340393, + "learning_rate": 1.4955756307040714e-06, + "loss": 0.1602, + "step": 54562 + }, + { + "epoch": 3.7862049823051835, + "grad_norm": 0.4992707669734955, + "learning_rate": 1.4946076898082807e-06, + "loss": 0.1139, + "step": 54563 + }, + { + "epoch": 3.7862743737422804, + "grad_norm": 0.7239187955856323, + "learning_rate": 1.4936400598825951e-06, + "loss": 0.1705, + "step": 54564 + }, + { + "epoch": 3.7863437651793768, + "grad_norm": 0.7138850688934326, + "learning_rate": 1.4926727409300567e-06, + "loss": 0.2034, + "step": 54565 + }, + { + "epoch": 3.7864131566164736, + "grad_norm": 0.6733489632606506, + "learning_rate": 1.4917057329537187e-06, + "loss": 0.1786, + "step": 54566 + }, + { + "epoch": 3.78648254805357, + "grad_norm": 0.6947928071022034, + "learning_rate": 1.4907390359566342e-06, + "loss": 0.1785, + "step": 54567 + }, + { + "epoch": 3.786551939490667, + "grad_norm": 0.984074056148529, + "learning_rate": 1.4897726499418563e-06, + "loss": 0.3319, + "step": 54568 + }, + { + "epoch": 3.7866213309277637, + "grad_norm": 2.888986587524414, + "learning_rate": 1.488806574912449e-06, + "loss": 0.3773, + "step": 54569 + }, + { + "epoch": 3.78669072236486, + "grad_norm": 0.8946305513381958, + "learning_rate": 1.4878408108714436e-06, + "loss": 0.1253, + "step": 54570 + }, + { + "epoch": 3.786760113801957, + "grad_norm": 1.715781807899475, + "learning_rate": 1.486875357821893e-06, + "loss": 0.3461, + "step": 54571 + }, + { + "epoch": 3.7868295052390533, + "grad_norm": 0.7319163680076599, + "learning_rate": 1.4859102157668392e-06, + "loss": 0.169, + "step": 54572 + }, + { + "epoch": 3.78689889667615, + "grad_norm": 0.7301703095436096, + "learning_rate": 1.4849453847093575e-06, + "loss": 0.2153, + "step": 54573 + }, + { + "epoch": 3.786968288113247, + "grad_norm": 1.031882405281067, + "learning_rate": 1.4839808646524567e-06, + "loss": 0.2445, + "step": 54574 + }, + { + "epoch": 3.7870376795503438, + "grad_norm": 0.8244395852088928, + "learning_rate": 1.4830166555992008e-06, + "loss": 0.2072, + "step": 54575 + }, + { + "epoch": 3.78710707098744, + "grad_norm": 0.9861648678779602, + "learning_rate": 1.4820527575526323e-06, + "loss": 0.2874, + "step": 54576 + }, + { + "epoch": 3.7871764624245365, + "grad_norm": 0.8696613311767578, + "learning_rate": 1.4810891705157924e-06, + "loss": 0.1802, + "step": 54577 + }, + { + "epoch": 3.7872458538616334, + "grad_norm": 0.7410342693328857, + "learning_rate": 1.480125894491724e-06, + "loss": 0.2553, + "step": 54578 + }, + { + "epoch": 3.78731524529873, + "grad_norm": 1.0550988912582397, + "learning_rate": 1.4791629294834687e-06, + "loss": 0.1904, + "step": 54579 + }, + { + "epoch": 3.787384636735827, + "grad_norm": 0.7377369999885559, + "learning_rate": 1.4782002754940683e-06, + "loss": 0.187, + "step": 54580 + }, + { + "epoch": 3.7874540281729234, + "grad_norm": 0.7320515513420105, + "learning_rate": 1.477237932526565e-06, + "loss": 0.1576, + "step": 54581 + }, + { + "epoch": 3.7875234196100203, + "grad_norm": 0.9214165210723877, + "learning_rate": 1.4762759005839788e-06, + "loss": 0.2494, + "step": 54582 + }, + { + "epoch": 3.7875928110471166, + "grad_norm": 0.7838494181632996, + "learning_rate": 1.4753141796693627e-06, + "loss": 0.2494, + "step": 54583 + }, + { + "epoch": 3.7876622024842135, + "grad_norm": 0.8029395937919617, + "learning_rate": 1.4743527697857473e-06, + "loss": 0.1999, + "step": 54584 + }, + { + "epoch": 3.7877315939213103, + "grad_norm": 0.9104120135307312, + "learning_rate": 1.473391670936164e-06, + "loss": 0.2523, + "step": 54585 + }, + { + "epoch": 3.7878009853584067, + "grad_norm": 0.9201350808143616, + "learning_rate": 1.4724308831236543e-06, + "loss": 0.2551, + "step": 54586 + }, + { + "epoch": 3.7878703767955035, + "grad_norm": 0.865222692489624, + "learning_rate": 1.4714704063512608e-06, + "loss": 0.2963, + "step": 54587 + }, + { + "epoch": 3.7879397682326, + "grad_norm": 0.92696613073349, + "learning_rate": 1.4705102406219918e-06, + "loss": 0.3073, + "step": 54588 + }, + { + "epoch": 3.7880091596696968, + "grad_norm": 0.8619269728660583, + "learning_rate": 1.4695503859389004e-06, + "loss": 0.295, + "step": 54589 + }, + { + "epoch": 3.7880785511067936, + "grad_norm": 0.6234501004219055, + "learning_rate": 1.4685908423049954e-06, + "loss": 0.1492, + "step": 54590 + }, + { + "epoch": 3.78814794254389, + "grad_norm": 0.7329809069633484, + "learning_rate": 1.4676316097233188e-06, + "loss": 0.2771, + "step": 54591 + }, + { + "epoch": 3.788217333980987, + "grad_norm": 0.8677687048912048, + "learning_rate": 1.4666726881969018e-06, + "loss": 0.2984, + "step": 54592 + }, + { + "epoch": 3.788286725418083, + "grad_norm": 1.0066094398498535, + "learning_rate": 1.4657140777287638e-06, + "loss": 0.258, + "step": 54593 + }, + { + "epoch": 3.78835611685518, + "grad_norm": 0.8090328574180603, + "learning_rate": 1.4647557783219467e-06, + "loss": 0.2633, + "step": 54594 + }, + { + "epoch": 3.788425508292277, + "grad_norm": 0.8973150849342346, + "learning_rate": 1.4637977899794485e-06, + "loss": 0.1833, + "step": 54595 + }, + { + "epoch": 3.7884948997293733, + "grad_norm": 0.9876359701156616, + "learning_rate": 1.4628401127043113e-06, + "loss": 0.1943, + "step": 54596 + }, + { + "epoch": 3.78856429116647, + "grad_norm": 0.8073984384536743, + "learning_rate": 1.4618827464995544e-06, + "loss": 0.2333, + "step": 54597 + }, + { + "epoch": 3.7886336826035665, + "grad_norm": 0.5467286109924316, + "learning_rate": 1.460925691368209e-06, + "loss": 0.1345, + "step": 54598 + }, + { + "epoch": 3.7887030740406633, + "grad_norm": 1.2368879318237305, + "learning_rate": 1.4599689473132837e-06, + "loss": 0.1912, + "step": 54599 + }, + { + "epoch": 3.78877246547776, + "grad_norm": 0.7107058763504028, + "learning_rate": 1.4590125143378097e-06, + "loss": 0.2036, + "step": 54600 + }, + { + "epoch": 3.788841856914857, + "grad_norm": 0.8725764751434326, + "learning_rate": 1.4580563924447955e-06, + "loss": 0.2748, + "step": 54601 + }, + { + "epoch": 3.7889112483519534, + "grad_norm": 0.8066111207008362, + "learning_rate": 1.4571005816372607e-06, + "loss": 0.2018, + "step": 54602 + }, + { + "epoch": 3.78898063978905, + "grad_norm": 1.0413340330123901, + "learning_rate": 1.4561450819182365e-06, + "loss": 0.1661, + "step": 54603 + }, + { + "epoch": 3.7890500312261466, + "grad_norm": 0.8565329909324646, + "learning_rate": 1.4551898932907205e-06, + "loss": 0.2168, + "step": 54604 + }, + { + "epoch": 3.7891194226632434, + "grad_norm": 0.937714695930481, + "learning_rate": 1.4542350157577322e-06, + "loss": 0.3839, + "step": 54605 + }, + { + "epoch": 3.7891888141003403, + "grad_norm": 1.0916993618011475, + "learning_rate": 1.4532804493222918e-06, + "loss": 0.3189, + "step": 54606 + }, + { + "epoch": 3.7892582055374366, + "grad_norm": 1.3377856016159058, + "learning_rate": 1.4523261939874188e-06, + "loss": 0.4, + "step": 54607 + }, + { + "epoch": 3.7893275969745335, + "grad_norm": 0.9432755708694458, + "learning_rate": 1.4513722497561222e-06, + "loss": 0.1982, + "step": 54608 + }, + { + "epoch": 3.78939698841163, + "grad_norm": 0.7271336317062378, + "learning_rate": 1.4504186166313994e-06, + "loss": 0.2454, + "step": 54609 + }, + { + "epoch": 3.7894663798487267, + "grad_norm": 1.122499942779541, + "learning_rate": 1.4494652946162702e-06, + "loss": 0.187, + "step": 54610 + }, + { + "epoch": 3.7895357712858235, + "grad_norm": 0.8372901678085327, + "learning_rate": 1.4485122837137544e-06, + "loss": 0.1835, + "step": 54611 + }, + { + "epoch": 3.78960516272292, + "grad_norm": 0.5412973165512085, + "learning_rate": 1.4475595839268386e-06, + "loss": 0.1166, + "step": 54612 + }, + { + "epoch": 3.7896745541600168, + "grad_norm": 0.6196601390838623, + "learning_rate": 1.4466071952585536e-06, + "loss": 0.1876, + "step": 54613 + }, + { + "epoch": 3.789743945597113, + "grad_norm": 0.8871336579322815, + "learning_rate": 1.445655117711886e-06, + "loss": 0.2143, + "step": 54614 + }, + { + "epoch": 3.78981333703421, + "grad_norm": 1.7317166328430176, + "learning_rate": 1.4447033512898666e-06, + "loss": 0.3227, + "step": 54615 + }, + { + "epoch": 3.789882728471307, + "grad_norm": 0.8623643517494202, + "learning_rate": 1.443751895995471e-06, + "loss": 0.1789, + "step": 54616 + }, + { + "epoch": 3.789952119908403, + "grad_norm": 0.7506604194641113, + "learning_rate": 1.4428007518317189e-06, + "loss": 0.2186, + "step": 54617 + }, + { + "epoch": 3.7900215113455, + "grad_norm": 0.7613530158996582, + "learning_rate": 1.4418499188016188e-06, + "loss": 0.2348, + "step": 54618 + }, + { + "epoch": 3.7900909027825964, + "grad_norm": 0.7016152739524841, + "learning_rate": 1.4408993969081463e-06, + "loss": 0.3029, + "step": 54619 + }, + { + "epoch": 3.7901602942196932, + "grad_norm": 1.2085338830947876, + "learning_rate": 1.4399491861543325e-06, + "loss": 0.2606, + "step": 54620 + }, + { + "epoch": 3.79022968565679, + "grad_norm": 1.3361144065856934, + "learning_rate": 1.4389992865431633e-06, + "loss": 0.2795, + "step": 54621 + }, + { + "epoch": 3.7902990770938865, + "grad_norm": 0.8680000305175781, + "learning_rate": 1.438049698077637e-06, + "loss": 0.2607, + "step": 54622 + }, + { + "epoch": 3.7903684685309833, + "grad_norm": 0.746712327003479, + "learning_rate": 1.4371004207607508e-06, + "loss": 0.2383, + "step": 54623 + }, + { + "epoch": 3.7904378599680797, + "grad_norm": 0.837694525718689, + "learning_rate": 1.4361514545955025e-06, + "loss": 0.1737, + "step": 54624 + }, + { + "epoch": 3.7905072514051765, + "grad_norm": 0.8452736139297485, + "learning_rate": 1.4352027995848893e-06, + "loss": 0.3114, + "step": 54625 + }, + { + "epoch": 3.7905766428422734, + "grad_norm": 1.3134292364120483, + "learning_rate": 1.4342544557318981e-06, + "loss": 0.2346, + "step": 54626 + }, + { + "epoch": 3.79064603427937, + "grad_norm": 0.7177923321723938, + "learning_rate": 1.4333064230395376e-06, + "loss": 0.201, + "step": 54627 + }, + { + "epoch": 3.7907154257164666, + "grad_norm": 0.9728248119354248, + "learning_rate": 1.4323587015107942e-06, + "loss": 0.2966, + "step": 54628 + }, + { + "epoch": 3.7907848171535634, + "grad_norm": 1.1319102048873901, + "learning_rate": 1.4314112911486544e-06, + "loss": 0.2126, + "step": 54629 + }, + { + "epoch": 3.79085420859066, + "grad_norm": 0.6974346041679382, + "learning_rate": 1.4304641919561157e-06, + "loss": 0.2211, + "step": 54630 + }, + { + "epoch": 3.7909236000277566, + "grad_norm": 0.8371039032936096, + "learning_rate": 1.4295174039361647e-06, + "loss": 0.287, + "step": 54631 + }, + { + "epoch": 3.7909929914648535, + "grad_norm": 0.8636883497238159, + "learning_rate": 1.428570927091788e-06, + "loss": 0.1882, + "step": 54632 + }, + { + "epoch": 3.79106238290195, + "grad_norm": 0.8010101914405823, + "learning_rate": 1.427624761425983e-06, + "loss": 0.2279, + "step": 54633 + }, + { + "epoch": 3.7911317743390467, + "grad_norm": 0.6911409497261047, + "learning_rate": 1.4266789069417253e-06, + "loss": 0.2341, + "step": 54634 + }, + { + "epoch": 3.791201165776143, + "grad_norm": 0.8444812893867493, + "learning_rate": 1.4257333636420122e-06, + "loss": 0.1836, + "step": 54635 + }, + { + "epoch": 3.79127055721324, + "grad_norm": 1.0269263982772827, + "learning_rate": 1.4247881315298195e-06, + "loss": 0.2348, + "step": 54636 + }, + { + "epoch": 3.7913399486503367, + "grad_norm": 0.8288685083389282, + "learning_rate": 1.4238432106081223e-06, + "loss": 0.186, + "step": 54637 + }, + { + "epoch": 3.791409340087433, + "grad_norm": 0.7976416349411011, + "learning_rate": 1.4228986008799293e-06, + "loss": 0.2537, + "step": 54638 + }, + { + "epoch": 3.79147873152453, + "grad_norm": 0.9450541734695435, + "learning_rate": 1.4219543023482052e-06, + "loss": 0.1909, + "step": 54639 + }, + { + "epoch": 3.7915481229616264, + "grad_norm": 0.8184773921966553, + "learning_rate": 1.421010315015936e-06, + "loss": 0.2135, + "step": 54640 + }, + { + "epoch": 3.791617514398723, + "grad_norm": 0.6307477355003357, + "learning_rate": 1.4200666388860861e-06, + "loss": 0.1808, + "step": 54641 + }, + { + "epoch": 3.79168690583582, + "grad_norm": 1.062937617301941, + "learning_rate": 1.4191232739616644e-06, + "loss": 0.2143, + "step": 54642 + }, + { + "epoch": 3.7917562972729164, + "grad_norm": 1.1273051500320435, + "learning_rate": 1.4181802202456352e-06, + "loss": 0.2172, + "step": 54643 + }, + { + "epoch": 3.7918256887100132, + "grad_norm": 0.9102226495742798, + "learning_rate": 1.4172374777409625e-06, + "loss": 0.2345, + "step": 54644 + }, + { + "epoch": 3.7918950801471096, + "grad_norm": 0.9571439623832703, + "learning_rate": 1.416295046450633e-06, + "loss": 0.1798, + "step": 54645 + }, + { + "epoch": 3.7919644715842065, + "grad_norm": 0.8870162963867188, + "learning_rate": 1.4153529263776223e-06, + "loss": 0.2084, + "step": 54646 + }, + { + "epoch": 3.7920338630213033, + "grad_norm": 0.6562129855155945, + "learning_rate": 1.4144111175249052e-06, + "loss": 0.1856, + "step": 54647 + }, + { + "epoch": 3.7921032544583997, + "grad_norm": 1.6297831535339355, + "learning_rate": 1.4134696198954577e-06, + "loss": 0.3143, + "step": 54648 + }, + { + "epoch": 3.7921726458954965, + "grad_norm": 0.7059202194213867, + "learning_rate": 1.412528433492255e-06, + "loss": 0.2404, + "step": 54649 + }, + { + "epoch": 3.792242037332593, + "grad_norm": 0.8994964361190796, + "learning_rate": 1.4115875583182503e-06, + "loss": 0.225, + "step": 54650 + }, + { + "epoch": 3.7923114287696897, + "grad_norm": 0.8879457116127014, + "learning_rate": 1.41064699437643e-06, + "loss": 0.2592, + "step": 54651 + }, + { + "epoch": 3.7923808202067866, + "grad_norm": 0.9094051122665405, + "learning_rate": 1.4097067416697585e-06, + "loss": 0.2026, + "step": 54652 + }, + { + "epoch": 3.7924502116438834, + "grad_norm": 0.8285399675369263, + "learning_rate": 1.4087668002012e-06, + "loss": 0.2146, + "step": 54653 + }, + { + "epoch": 3.79251960308098, + "grad_norm": 0.9315792918205261, + "learning_rate": 1.4078271699737299e-06, + "loss": 0.2361, + "step": 54654 + }, + { + "epoch": 3.7925889945180766, + "grad_norm": 0.9917213916778564, + "learning_rate": 1.4068878509903016e-06, + "loss": 0.4118, + "step": 54655 + }, + { + "epoch": 3.792658385955173, + "grad_norm": 0.7141126394271851, + "learning_rate": 1.4059488432539125e-06, + "loss": 0.1884, + "step": 54656 + }, + { + "epoch": 3.79272777739227, + "grad_norm": 1.0766465663909912, + "learning_rate": 1.4050101467674826e-06, + "loss": 0.2075, + "step": 54657 + }, + { + "epoch": 3.7927971688293667, + "grad_norm": 0.6016058921813965, + "learning_rate": 1.4040717615339981e-06, + "loss": 0.1789, + "step": 54658 + }, + { + "epoch": 3.792866560266463, + "grad_norm": 0.9401628971099854, + "learning_rate": 1.4031336875564239e-06, + "loss": 0.1374, + "step": 54659 + }, + { + "epoch": 3.79293595170356, + "grad_norm": 0.8907730579376221, + "learning_rate": 1.4021959248377236e-06, + "loss": 0.2257, + "step": 54660 + }, + { + "epoch": 3.7930053431406563, + "grad_norm": 0.6980615258216858, + "learning_rate": 1.4012584733808398e-06, + "loss": 0.2029, + "step": 54661 + }, + { + "epoch": 3.793074734577753, + "grad_norm": 0.7578168511390686, + "learning_rate": 1.4003213331887477e-06, + "loss": 0.2618, + "step": 54662 + }, + { + "epoch": 3.79314412601485, + "grad_norm": 1.761014461517334, + "learning_rate": 1.3993845042644115e-06, + "loss": 0.4078, + "step": 54663 + }, + { + "epoch": 3.7932135174519463, + "grad_norm": 0.8799289464950562, + "learning_rate": 1.3984479866107737e-06, + "loss": 0.1836, + "step": 54664 + }, + { + "epoch": 3.793282908889043, + "grad_norm": 0.9664332866668701, + "learning_rate": 1.397511780230787e-06, + "loss": 0.3033, + "step": 54665 + }, + { + "epoch": 3.7933523003261396, + "grad_norm": 1.1206203699111938, + "learning_rate": 1.396575885127427e-06, + "loss": 0.2561, + "step": 54666 + }, + { + "epoch": 3.7934216917632364, + "grad_norm": 0.9434934854507446, + "learning_rate": 1.3956403013036245e-06, + "loss": 0.2556, + "step": 54667 + }, + { + "epoch": 3.7934910832003332, + "grad_norm": 1.0859402418136597, + "learning_rate": 1.3947050287623553e-06, + "loss": 0.2938, + "step": 54668 + }, + { + "epoch": 3.7935604746374296, + "grad_norm": 0.6412935853004456, + "learning_rate": 1.39377006750655e-06, + "loss": 0.2045, + "step": 54669 + }, + { + "epoch": 3.7936298660745265, + "grad_norm": 0.8729853630065918, + "learning_rate": 1.3928354175391844e-06, + "loss": 0.3206, + "step": 54670 + }, + { + "epoch": 3.793699257511623, + "grad_norm": 0.8855711221694946, + "learning_rate": 1.3919010788632003e-06, + "loss": 0.2793, + "step": 54671 + }, + { + "epoch": 3.7937686489487197, + "grad_norm": 0.8800867199897766, + "learning_rate": 1.3909670514815288e-06, + "loss": 0.1684, + "step": 54672 + }, + { + "epoch": 3.7938380403858165, + "grad_norm": 1.0424964427947998, + "learning_rate": 1.3900333353971561e-06, + "loss": 0.2409, + "step": 54673 + }, + { + "epoch": 3.793907431822913, + "grad_norm": 0.8098788261413574, + "learning_rate": 1.3890999306129915e-06, + "loss": 0.2698, + "step": 54674 + }, + { + "epoch": 3.7939768232600097, + "grad_norm": 0.9802905321121216, + "learning_rate": 1.3881668371319878e-06, + "loss": 0.2953, + "step": 54675 + }, + { + "epoch": 3.794046214697106, + "grad_norm": 1.088773250579834, + "learning_rate": 1.3872340549571094e-06, + "loss": 0.3419, + "step": 54676 + }, + { + "epoch": 3.794115606134203, + "grad_norm": 0.954037070274353, + "learning_rate": 1.3863015840913095e-06, + "loss": 0.2571, + "step": 54677 + }, + { + "epoch": 3.7941849975713, + "grad_norm": 0.6192890405654907, + "learning_rate": 1.385369424537497e-06, + "loss": 0.1568, + "step": 54678 + }, + { + "epoch": 3.7942543890083966, + "grad_norm": 1.4371029138565063, + "learning_rate": 1.384437576298625e-06, + "loss": 0.4363, + "step": 54679 + }, + { + "epoch": 3.794323780445493, + "grad_norm": 0.7011480331420898, + "learning_rate": 1.3835060393776466e-06, + "loss": 0.1652, + "step": 54680 + }, + { + "epoch": 3.79439317188259, + "grad_norm": 0.8208434581756592, + "learning_rate": 1.3825748137775042e-06, + "loss": 0.2306, + "step": 54681 + }, + { + "epoch": 3.7944625633196862, + "grad_norm": 1.9302563667297363, + "learning_rate": 1.3816438995011283e-06, + "loss": 0.2915, + "step": 54682 + }, + { + "epoch": 3.794531954756783, + "grad_norm": 0.8258903622627258, + "learning_rate": 1.3807132965514614e-06, + "loss": 0.281, + "step": 54683 + }, + { + "epoch": 3.79460134619388, + "grad_norm": 0.7006315588951111, + "learning_rate": 1.3797830049314454e-06, + "loss": 0.2089, + "step": 54684 + }, + { + "epoch": 3.7946707376309763, + "grad_norm": 0.8755828142166138, + "learning_rate": 1.378853024644e-06, + "loss": 0.2567, + "step": 54685 + }, + { + "epoch": 3.794740129068073, + "grad_norm": 1.3791369199752808, + "learning_rate": 1.3779233556920902e-06, + "loss": 0.2378, + "step": 54686 + }, + { + "epoch": 3.7948095205051695, + "grad_norm": 0.8372135758399963, + "learning_rate": 1.3769939980786129e-06, + "loss": 0.2605, + "step": 54687 + }, + { + "epoch": 3.7948789119422663, + "grad_norm": 0.8137003779411316, + "learning_rate": 1.3760649518065438e-06, + "loss": 0.272, + "step": 54688 + }, + { + "epoch": 3.794948303379363, + "grad_norm": 0.9219874739646912, + "learning_rate": 1.3751362168787806e-06, + "loss": 0.2166, + "step": 54689 + }, + { + "epoch": 3.7950176948164596, + "grad_norm": 0.6423596739768982, + "learning_rate": 1.3742077932982766e-06, + "loss": 0.1423, + "step": 54690 + }, + { + "epoch": 3.7950870862535564, + "grad_norm": 0.6958073377609253, + "learning_rate": 1.3732796810679627e-06, + "loss": 0.2446, + "step": 54691 + }, + { + "epoch": 3.795156477690653, + "grad_norm": 0.7756326198577881, + "learning_rate": 1.3723518801907475e-06, + "loss": 0.1893, + "step": 54692 + }, + { + "epoch": 3.7952258691277496, + "grad_norm": 0.9013763070106506, + "learning_rate": 1.3714243906695845e-06, + "loss": 0.2569, + "step": 54693 + }, + { + "epoch": 3.7952952605648465, + "grad_norm": 1.009289026260376, + "learning_rate": 1.3704972125073824e-06, + "loss": 0.2501, + "step": 54694 + }, + { + "epoch": 3.795364652001943, + "grad_norm": 0.9878154993057251, + "learning_rate": 1.369570345707083e-06, + "loss": 0.2442, + "step": 54695 + }, + { + "epoch": 3.7954340434390397, + "grad_norm": 0.90193110704422, + "learning_rate": 1.3686437902716066e-06, + "loss": 0.2956, + "step": 54696 + }, + { + "epoch": 3.795503434876136, + "grad_norm": 0.8747655153274536, + "learning_rate": 1.3677175462038838e-06, + "loss": 0.2012, + "step": 54697 + }, + { + "epoch": 3.795572826313233, + "grad_norm": 0.9450642466545105, + "learning_rate": 1.3667916135068238e-06, + "loss": 0.2519, + "step": 54698 + }, + { + "epoch": 3.7956422177503297, + "grad_norm": 0.9478567838668823, + "learning_rate": 1.3658659921833682e-06, + "loss": 0.2933, + "step": 54699 + }, + { + "epoch": 3.795711609187426, + "grad_norm": 0.7341459393501282, + "learning_rate": 1.3649406822364264e-06, + "loss": 0.2161, + "step": 54700 + }, + { + "epoch": 3.795781000624523, + "grad_norm": 0.7433846592903137, + "learning_rate": 1.3640156836689178e-06, + "loss": 0.1788, + "step": 54701 + }, + { + "epoch": 3.7958503920616193, + "grad_norm": 0.9705455899238586, + "learning_rate": 1.3630909964837734e-06, + "loss": 0.3787, + "step": 54702 + }, + { + "epoch": 3.795919783498716, + "grad_norm": 0.8344094753265381, + "learning_rate": 1.3621666206839023e-06, + "loss": 0.2398, + "step": 54703 + }, + { + "epoch": 3.795989174935813, + "grad_norm": 1.8649579286575317, + "learning_rate": 1.3612425562722353e-06, + "loss": 0.3238, + "step": 54704 + }, + { + "epoch": 3.79605856637291, + "grad_norm": 1.1978681087493896, + "learning_rate": 1.36031880325167e-06, + "loss": 0.346, + "step": 54705 + }, + { + "epoch": 3.7961279578100062, + "grad_norm": 0.9481082558631897, + "learning_rate": 1.3593953616251487e-06, + "loss": 0.2758, + "step": 54706 + }, + { + "epoch": 3.796197349247103, + "grad_norm": 0.8667999505996704, + "learning_rate": 1.3584722313955578e-06, + "loss": 0.23, + "step": 54707 + }, + { + "epoch": 3.7962667406841994, + "grad_norm": 0.669838011264801, + "learning_rate": 1.3575494125658394e-06, + "loss": 0.1972, + "step": 54708 + }, + { + "epoch": 3.7963361321212963, + "grad_norm": 1.826604962348938, + "learning_rate": 1.3566269051388803e-06, + "loss": 0.3048, + "step": 54709 + }, + { + "epoch": 3.796405523558393, + "grad_norm": 0.9424715042114258, + "learning_rate": 1.3557047091176112e-06, + "loss": 0.3137, + "step": 54710 + }, + { + "epoch": 3.7964749149954895, + "grad_norm": 1.063335657119751, + "learning_rate": 1.35478282450493e-06, + "loss": 0.3239, + "step": 54711 + }, + { + "epoch": 3.7965443064325863, + "grad_norm": 0.6708208918571472, + "learning_rate": 1.3538612513037674e-06, + "loss": 0.2258, + "step": 54712 + }, + { + "epoch": 3.7966136978696827, + "grad_norm": 1.086449146270752, + "learning_rate": 1.3529399895170104e-06, + "loss": 0.2677, + "step": 54713 + }, + { + "epoch": 3.7966830893067796, + "grad_norm": 0.8892894983291626, + "learning_rate": 1.3520190391475784e-06, + "loss": 0.3008, + "step": 54714 + }, + { + "epoch": 3.7967524807438764, + "grad_norm": 1.018349528312683, + "learning_rate": 1.3510984001983806e-06, + "loss": 0.1907, + "step": 54715 + }, + { + "epoch": 3.796821872180973, + "grad_norm": 2.2365193367004395, + "learning_rate": 1.3501780726723145e-06, + "loss": 0.3288, + "step": 54716 + }, + { + "epoch": 3.7968912636180696, + "grad_norm": 0.7330357432365417, + "learning_rate": 1.349258056572289e-06, + "loss": 0.2141, + "step": 54717 + }, + { + "epoch": 3.796960655055166, + "grad_norm": 1.0758893489837646, + "learning_rate": 1.3483383519012237e-06, + "loss": 0.3556, + "step": 54718 + }, + { + "epoch": 3.797030046492263, + "grad_norm": 0.6115373373031616, + "learning_rate": 1.3474189586619945e-06, + "loss": 0.1548, + "step": 54719 + }, + { + "epoch": 3.7970994379293597, + "grad_norm": 0.9220356345176697, + "learning_rate": 1.346499876857521e-06, + "loss": 0.2116, + "step": 54720 + }, + { + "epoch": 3.797168829366456, + "grad_norm": 0.7524557113647461, + "learning_rate": 1.345581106490701e-06, + "loss": 0.2017, + "step": 54721 + }, + { + "epoch": 3.797238220803553, + "grad_norm": 0.7786087393760681, + "learning_rate": 1.344662647564432e-06, + "loss": 0.1616, + "step": 54722 + }, + { + "epoch": 3.7973076122406493, + "grad_norm": 0.9532673358917236, + "learning_rate": 1.3437445000816229e-06, + "loss": 0.2895, + "step": 54723 + }, + { + "epoch": 3.797377003677746, + "grad_norm": 1.1912139654159546, + "learning_rate": 1.3428266640451602e-06, + "loss": 0.2031, + "step": 54724 + }, + { + "epoch": 3.797446395114843, + "grad_norm": 1.0570257902145386, + "learning_rate": 1.341909139457964e-06, + "loss": 0.2468, + "step": 54725 + }, + { + "epoch": 3.7975157865519393, + "grad_norm": 0.7971048355102539, + "learning_rate": 1.3409919263228988e-06, + "loss": 0.3131, + "step": 54726 + }, + { + "epoch": 3.797585177989036, + "grad_norm": 0.7140514850616455, + "learning_rate": 1.340075024642884e-06, + "loss": 0.1353, + "step": 54727 + }, + { + "epoch": 3.7976545694261326, + "grad_norm": 1.176493763923645, + "learning_rate": 1.3391584344207952e-06, + "loss": 0.2575, + "step": 54728 + }, + { + "epoch": 3.7977239608632294, + "grad_norm": 1.1447629928588867, + "learning_rate": 1.3382421556595414e-06, + "loss": 0.2896, + "step": 54729 + }, + { + "epoch": 3.797793352300326, + "grad_norm": 0.9486809372901917, + "learning_rate": 1.3373261883620092e-06, + "loss": 0.3107, + "step": 54730 + }, + { + "epoch": 3.797862743737423, + "grad_norm": 0.8596304059028625, + "learning_rate": 1.3364105325310961e-06, + "loss": 0.1753, + "step": 54731 + }, + { + "epoch": 3.7979321351745194, + "grad_norm": 0.9178811311721802, + "learning_rate": 1.3354951881696887e-06, + "loss": 0.289, + "step": 54732 + }, + { + "epoch": 3.7980015266116163, + "grad_norm": 0.7519652247428894, + "learning_rate": 1.3345801552806737e-06, + "loss": 0.2118, + "step": 54733 + }, + { + "epoch": 3.7980709180487127, + "grad_norm": 0.8978502154350281, + "learning_rate": 1.3336654338669375e-06, + "loss": 0.1688, + "step": 54734 + }, + { + "epoch": 3.7981403094858095, + "grad_norm": 0.9946563839912415, + "learning_rate": 1.3327510239313668e-06, + "loss": 0.2916, + "step": 54735 + }, + { + "epoch": 3.7982097009229063, + "grad_norm": 0.9058042168617249, + "learning_rate": 1.3318369254768593e-06, + "loss": 0.2035, + "step": 54736 + }, + { + "epoch": 3.7982790923600027, + "grad_norm": 0.9031720757484436, + "learning_rate": 1.3309231385063014e-06, + "loss": 0.2685, + "step": 54737 + }, + { + "epoch": 3.7983484837970996, + "grad_norm": 0.9833064079284668, + "learning_rate": 1.330009663022569e-06, + "loss": 0.2538, + "step": 54738 + }, + { + "epoch": 3.798417875234196, + "grad_norm": 0.8468460440635681, + "learning_rate": 1.3290964990285482e-06, + "loss": 0.2486, + "step": 54739 + }, + { + "epoch": 3.7984872666712928, + "grad_norm": 0.9262343049049377, + "learning_rate": 1.3281836465271148e-06, + "loss": 0.2254, + "step": 54740 + }, + { + "epoch": 3.7985566581083896, + "grad_norm": 0.5704818964004517, + "learning_rate": 1.3272711055211661e-06, + "loss": 0.2488, + "step": 54741 + }, + { + "epoch": 3.798626049545486, + "grad_norm": 0.8412162661552429, + "learning_rate": 1.326358876013567e-06, + "loss": 0.2393, + "step": 54742 + }, + { + "epoch": 3.798695440982583, + "grad_norm": 0.9851971864700317, + "learning_rate": 1.325446958007215e-06, + "loss": 0.2677, + "step": 54743 + }, + { + "epoch": 3.798764832419679, + "grad_norm": 0.9287739396095276, + "learning_rate": 1.3245353515049741e-06, + "loss": 0.216, + "step": 54744 + }, + { + "epoch": 3.798834223856776, + "grad_norm": 0.7626466751098633, + "learning_rate": 1.3236240565097202e-06, + "loss": 0.1428, + "step": 54745 + }, + { + "epoch": 3.798903615293873, + "grad_norm": 0.8687757849693298, + "learning_rate": 1.3227130730243509e-06, + "loss": 0.2754, + "step": 54746 + }, + { + "epoch": 3.7989730067309693, + "grad_norm": 1.1737217903137207, + "learning_rate": 1.3218024010517193e-06, + "loss": 0.2069, + "step": 54747 + }, + { + "epoch": 3.799042398168066, + "grad_norm": 1.0394508838653564, + "learning_rate": 1.3208920405947013e-06, + "loss": 0.3396, + "step": 54748 + }, + { + "epoch": 3.7991117896051625, + "grad_norm": 0.7304611802101135, + "learning_rate": 1.319981991656183e-06, + "loss": 0.1626, + "step": 54749 + }, + { + "epoch": 3.7991811810422593, + "grad_norm": 0.9888449311256409, + "learning_rate": 1.3190722542390398e-06, + "loss": 0.3662, + "step": 54750 + }, + { + "epoch": 3.799250572479356, + "grad_norm": 0.8723782300949097, + "learning_rate": 1.3181628283461256e-06, + "loss": 0.2093, + "step": 54751 + }, + { + "epoch": 3.7993199639164525, + "grad_norm": 0.6979687213897705, + "learning_rate": 1.3172537139803265e-06, + "loss": 0.2032, + "step": 54752 + }, + { + "epoch": 3.7993893553535494, + "grad_norm": 0.6930502653121948, + "learning_rate": 1.316344911144518e-06, + "loss": 0.2297, + "step": 54753 + }, + { + "epoch": 3.7994587467906458, + "grad_norm": 1.0633212327957153, + "learning_rate": 1.3154364198415425e-06, + "loss": 0.2681, + "step": 54754 + }, + { + "epoch": 3.7995281382277426, + "grad_norm": 0.9564958810806274, + "learning_rate": 1.3145282400742975e-06, + "loss": 0.3402, + "step": 54755 + }, + { + "epoch": 3.7995975296648394, + "grad_norm": 0.8341977596282959, + "learning_rate": 1.3136203718456253e-06, + "loss": 0.2348, + "step": 54756 + }, + { + "epoch": 3.7996669211019363, + "grad_norm": 1.770782232284546, + "learning_rate": 1.3127128151584123e-06, + "loss": 0.3345, + "step": 54757 + }, + { + "epoch": 3.7997363125390327, + "grad_norm": 0.949748694896698, + "learning_rate": 1.3118055700155119e-06, + "loss": 0.2541, + "step": 54758 + }, + { + "epoch": 3.7998057039761295, + "grad_norm": 0.9440030455589294, + "learning_rate": 1.3108986364197884e-06, + "loss": 0.2867, + "step": 54759 + }, + { + "epoch": 3.799875095413226, + "grad_norm": 0.9144826531410217, + "learning_rate": 1.3099920143741173e-06, + "loss": 0.2292, + "step": 54760 + }, + { + "epoch": 3.7999444868503227, + "grad_norm": 1.0434229373931885, + "learning_rate": 1.309085703881352e-06, + "loss": 0.2375, + "step": 54761 + }, + { + "epoch": 3.8000138782874195, + "grad_norm": 0.8366252779960632, + "learning_rate": 1.3081797049443457e-06, + "loss": 0.2797, + "step": 54762 + }, + { + "epoch": 3.800083269724516, + "grad_norm": 1.8462051153182983, + "learning_rate": 1.3072740175659625e-06, + "loss": 0.2472, + "step": 54763 + }, + { + "epoch": 3.8001526611616128, + "grad_norm": 0.8329187035560608, + "learning_rate": 1.3063686417490672e-06, + "loss": 0.1821, + "step": 54764 + }, + { + "epoch": 3.800222052598709, + "grad_norm": 1.0522031784057617, + "learning_rate": 1.3054635774965129e-06, + "loss": 0.2366, + "step": 54765 + }, + { + "epoch": 3.800291444035806, + "grad_norm": 0.9036983251571655, + "learning_rate": 1.304558824811164e-06, + "loss": 0.2588, + "step": 54766 + }, + { + "epoch": 3.800360835472903, + "grad_norm": 1.0206286907196045, + "learning_rate": 1.3036543836958847e-06, + "loss": 0.3077, + "step": 54767 + }, + { + "epoch": 3.800430226909999, + "grad_norm": 1.392970323562622, + "learning_rate": 1.3027502541534953e-06, + "loss": 0.2873, + "step": 54768 + }, + { + "epoch": 3.800499618347096, + "grad_norm": 1.1826475858688354, + "learning_rate": 1.301846436186882e-06, + "loss": 0.3209, + "step": 54769 + }, + { + "epoch": 3.8005690097841924, + "grad_norm": 1.037584900856018, + "learning_rate": 1.3009429297988873e-06, + "loss": 0.3478, + "step": 54770 + }, + { + "epoch": 3.8006384012212893, + "grad_norm": 1.3730522394180298, + "learning_rate": 1.3000397349923643e-06, + "loss": 0.4883, + "step": 54771 + }, + { + "epoch": 3.800707792658386, + "grad_norm": 1.0854671001434326, + "learning_rate": 1.2991368517701663e-06, + "loss": 0.2482, + "step": 54772 + }, + { + "epoch": 3.8007771840954825, + "grad_norm": 0.9055025577545166, + "learning_rate": 1.2982342801351466e-06, + "loss": 0.1834, + "step": 54773 + }, + { + "epoch": 3.8008465755325793, + "grad_norm": 1.4191795587539673, + "learning_rate": 1.2973320200901473e-06, + "loss": 0.2536, + "step": 54774 + }, + { + "epoch": 3.8009159669696757, + "grad_norm": 0.833045482635498, + "learning_rate": 1.2964300716380107e-06, + "loss": 0.168, + "step": 54775 + }, + { + "epoch": 3.8009853584067725, + "grad_norm": 0.6953366994857788, + "learning_rate": 1.295528434781601e-06, + "loss": 0.2305, + "step": 54776 + }, + { + "epoch": 3.8010547498438694, + "grad_norm": 2.0674755573272705, + "learning_rate": 1.2946271095237495e-06, + "loss": 0.2311, + "step": 54777 + }, + { + "epoch": 3.8011241412809658, + "grad_norm": 1.005906581878662, + "learning_rate": 1.2937260958673203e-06, + "loss": 0.1651, + "step": 54778 + }, + { + "epoch": 3.8011935327180626, + "grad_norm": 0.7902872562408447, + "learning_rate": 1.2928253938151335e-06, + "loss": 0.2244, + "step": 54779 + }, + { + "epoch": 3.801262924155159, + "grad_norm": 0.8355684280395508, + "learning_rate": 1.2919250033700537e-06, + "loss": 0.2992, + "step": 54780 + }, + { + "epoch": 3.801332315592256, + "grad_norm": 1.1148663759231567, + "learning_rate": 1.2910249245349115e-06, + "loss": 0.3167, + "step": 54781 + }, + { + "epoch": 3.8014017070293527, + "grad_norm": 0.8950576782226562, + "learning_rate": 1.2901251573125495e-06, + "loss": 0.2504, + "step": 54782 + }, + { + "epoch": 3.8014710984664495, + "grad_norm": 0.7494627833366394, + "learning_rate": 1.2892257017058095e-06, + "loss": 0.16, + "step": 54783 + }, + { + "epoch": 3.801540489903546, + "grad_norm": 0.8765825033187866, + "learning_rate": 1.288326557717534e-06, + "loss": 0.2708, + "step": 54784 + }, + { + "epoch": 3.8016098813406427, + "grad_norm": 0.9535136222839355, + "learning_rate": 1.287427725350554e-06, + "loss": 0.2098, + "step": 54785 + }, + { + "epoch": 3.801679272777739, + "grad_norm": 0.8609575033187866, + "learning_rate": 1.2865292046077116e-06, + "loss": 0.3475, + "step": 54786 + }, + { + "epoch": 3.801748664214836, + "grad_norm": 0.8368092775344849, + "learning_rate": 1.2856309954918489e-06, + "loss": 0.2286, + "step": 54787 + }, + { + "epoch": 3.8018180556519328, + "grad_norm": 1.1481550931930542, + "learning_rate": 1.284733098005797e-06, + "loss": 0.3059, + "step": 54788 + }, + { + "epoch": 3.801887447089029, + "grad_norm": 0.7882388234138489, + "learning_rate": 1.2838355121523871e-06, + "loss": 0.3618, + "step": 54789 + }, + { + "epoch": 3.801956838526126, + "grad_norm": 0.7461569309234619, + "learning_rate": 1.28293823793445e-06, + "loss": 0.2222, + "step": 54790 + }, + { + "epoch": 3.8020262299632224, + "grad_norm": 0.819765567779541, + "learning_rate": 1.2820412753548283e-06, + "loss": 0.1984, + "step": 54791 + }, + { + "epoch": 3.802095621400319, + "grad_norm": 0.6783134937286377, + "learning_rate": 1.2811446244163527e-06, + "loss": 0.1511, + "step": 54792 + }, + { + "epoch": 3.802165012837416, + "grad_norm": 2.213191270828247, + "learning_rate": 1.2802482851218433e-06, + "loss": 0.2743, + "step": 54793 + }, + { + "epoch": 3.8022344042745124, + "grad_norm": 0.7118245959281921, + "learning_rate": 1.2793522574741423e-06, + "loss": 0.1899, + "step": 54794 + }, + { + "epoch": 3.8023037957116093, + "grad_norm": 1.25244140625, + "learning_rate": 1.2784565414760697e-06, + "loss": 0.2618, + "step": 54795 + }, + { + "epoch": 3.8023731871487056, + "grad_norm": 0.8597304821014404, + "learning_rate": 1.2775611371304564e-06, + "loss": 0.1783, + "step": 54796 + }, + { + "epoch": 3.8024425785858025, + "grad_norm": 0.9473494291305542, + "learning_rate": 1.2766660444401224e-06, + "loss": 0.2011, + "step": 54797 + }, + { + "epoch": 3.8025119700228993, + "grad_norm": 0.7692173719406128, + "learning_rate": 1.27577126340791e-06, + "loss": 0.2012, + "step": 54798 + }, + { + "epoch": 3.8025813614599957, + "grad_norm": 0.8330919742584229, + "learning_rate": 1.274876794036628e-06, + "loss": 0.2862, + "step": 54799 + }, + { + "epoch": 3.8026507528970925, + "grad_norm": 1.0079320669174194, + "learning_rate": 1.2739826363291074e-06, + "loss": 0.2246, + "step": 54800 + }, + { + "epoch": 3.802720144334189, + "grad_norm": 0.8275830149650574, + "learning_rate": 1.2730887902881684e-06, + "loss": 0.1758, + "step": 54801 + }, + { + "epoch": 3.8027895357712858, + "grad_norm": 0.9020031690597534, + "learning_rate": 1.2721952559166306e-06, + "loss": 0.2019, + "step": 54802 + }, + { + "epoch": 3.8028589272083826, + "grad_norm": 0.747164785861969, + "learning_rate": 1.2713020332173252e-06, + "loss": 0.2405, + "step": 54803 + }, + { + "epoch": 3.8029283186454794, + "grad_norm": 1.2369487285614014, + "learning_rate": 1.2704091221930615e-06, + "loss": 0.2093, + "step": 54804 + }, + { + "epoch": 3.802997710082576, + "grad_norm": 0.8129143118858337, + "learning_rate": 1.2695165228466589e-06, + "loss": 0.2578, + "step": 54805 + }, + { + "epoch": 3.803067101519672, + "grad_norm": 0.7183769941329956, + "learning_rate": 1.2686242351809374e-06, + "loss": 0.2578, + "step": 54806 + }, + { + "epoch": 3.803136492956769, + "grad_norm": 0.8435985445976257, + "learning_rate": 1.2677322591987173e-06, + "loss": 0.1941, + "step": 54807 + }, + { + "epoch": 3.803205884393866, + "grad_norm": 1.0566213130950928, + "learning_rate": 1.2668405949028183e-06, + "loss": 0.4507, + "step": 54808 + }, + { + "epoch": 3.8032752758309627, + "grad_norm": 0.8696014285087585, + "learning_rate": 1.2659492422960384e-06, + "loss": 0.3184, + "step": 54809 + }, + { + "epoch": 3.803344667268059, + "grad_norm": 0.9644085168838501, + "learning_rate": 1.2650582013811973e-06, + "loss": 0.2321, + "step": 54810 + }, + { + "epoch": 3.803414058705156, + "grad_norm": 1.0693867206573486, + "learning_rate": 1.2641674721611153e-06, + "loss": 0.3254, + "step": 54811 + }, + { + "epoch": 3.8034834501422523, + "grad_norm": 0.7766101360321045, + "learning_rate": 1.2632770546386008e-06, + "loss": 0.1389, + "step": 54812 + }, + { + "epoch": 3.803552841579349, + "grad_norm": 0.8627116084098816, + "learning_rate": 1.2623869488164742e-06, + "loss": 0.2748, + "step": 54813 + }, + { + "epoch": 3.803622233016446, + "grad_norm": 0.6626262068748474, + "learning_rate": 1.2614971546975329e-06, + "loss": 0.1676, + "step": 54814 + }, + { + "epoch": 3.8036916244535424, + "grad_norm": 0.7084934711456299, + "learning_rate": 1.260607672284586e-06, + "loss": 0.2084, + "step": 54815 + }, + { + "epoch": 3.803761015890639, + "grad_norm": 1.1666066646575928, + "learning_rate": 1.2597185015804535e-06, + "loss": 0.2879, + "step": 54816 + }, + { + "epoch": 3.8038304073277356, + "grad_norm": 0.9152224659919739, + "learning_rate": 1.258829642587922e-06, + "loss": 0.2139, + "step": 54817 + }, + { + "epoch": 3.8038997987648324, + "grad_norm": 0.9028743505477905, + "learning_rate": 1.2579410953098114e-06, + "loss": 0.2595, + "step": 54818 + }, + { + "epoch": 3.8039691902019293, + "grad_norm": 0.8553205132484436, + "learning_rate": 1.2570528597489306e-06, + "loss": 0.2062, + "step": 54819 + }, + { + "epoch": 3.8040385816390256, + "grad_norm": 1.0336583852767944, + "learning_rate": 1.2561649359080662e-06, + "loss": 0.3122, + "step": 54820 + }, + { + "epoch": 3.8041079730761225, + "grad_norm": 0.6199270486831665, + "learning_rate": 1.2552773237900383e-06, + "loss": 0.2486, + "step": 54821 + }, + { + "epoch": 3.804177364513219, + "grad_norm": 0.78577721118927, + "learning_rate": 1.2543900233976558e-06, + "loss": 0.2024, + "step": 54822 + }, + { + "epoch": 3.8042467559503157, + "grad_norm": 1.0672614574432373, + "learning_rate": 1.253503034733694e-06, + "loss": 0.2793, + "step": 54823 + }, + { + "epoch": 3.8043161473874125, + "grad_norm": 0.8405542373657227, + "learning_rate": 1.252616357800962e-06, + "loss": 0.3443, + "step": 54824 + }, + { + "epoch": 3.804385538824509, + "grad_norm": 1.0949324369430542, + "learning_rate": 1.2517299926022686e-06, + "loss": 0.3269, + "step": 54825 + }, + { + "epoch": 3.8044549302616057, + "grad_norm": 1.3232290744781494, + "learning_rate": 1.2508439391404114e-06, + "loss": 0.4088, + "step": 54826 + }, + { + "epoch": 3.804524321698702, + "grad_norm": 1.02474844455719, + "learning_rate": 1.2499581974181774e-06, + "loss": 0.2994, + "step": 54827 + }, + { + "epoch": 3.804593713135799, + "grad_norm": 1.298152208328247, + "learning_rate": 1.2490727674383752e-06, + "loss": 0.5128, + "step": 54828 + }, + { + "epoch": 3.804663104572896, + "grad_norm": 2.058673858642578, + "learning_rate": 1.2481876492037914e-06, + "loss": 0.3338, + "step": 54829 + }, + { + "epoch": 3.8047324960099926, + "grad_norm": 0.9948810935020447, + "learning_rate": 1.247302842717213e-06, + "loss": 0.2384, + "step": 54830 + }, + { + "epoch": 3.804801887447089, + "grad_norm": 0.8200640082359314, + "learning_rate": 1.2464183479814484e-06, + "loss": 0.3096, + "step": 54831 + }, + { + "epoch": 3.8048712788841854, + "grad_norm": 0.6365424990653992, + "learning_rate": 1.2455341649992846e-06, + "loss": 0.2197, + "step": 54832 + }, + { + "epoch": 3.8049406703212822, + "grad_norm": 0.9416064023971558, + "learning_rate": 1.2446502937735083e-06, + "loss": 0.2273, + "step": 54833 + }, + { + "epoch": 3.805010061758379, + "grad_norm": 0.7091141939163208, + "learning_rate": 1.243766734306917e-06, + "loss": 0.1994, + "step": 54834 + }, + { + "epoch": 3.805079453195476, + "grad_norm": 0.8173285722732544, + "learning_rate": 1.2428834866022864e-06, + "loss": 0.2904, + "step": 54835 + }, + { + "epoch": 3.8051488446325723, + "grad_norm": 1.0760724544525146, + "learning_rate": 1.2420005506624365e-06, + "loss": 0.3019, + "step": 54836 + }, + { + "epoch": 3.805218236069669, + "grad_norm": 0.6498292088508606, + "learning_rate": 1.2411179264901097e-06, + "loss": 0.2341, + "step": 54837 + }, + { + "epoch": 3.8052876275067655, + "grad_norm": 0.6352695226669312, + "learning_rate": 1.2402356140881255e-06, + "loss": 0.1983, + "step": 54838 + }, + { + "epoch": 3.8053570189438624, + "grad_norm": 0.9267755150794983, + "learning_rate": 1.23935361345926e-06, + "loss": 0.2585, + "step": 54839 + }, + { + "epoch": 3.805426410380959, + "grad_norm": 0.9572348594665527, + "learning_rate": 1.2384719246062993e-06, + "loss": 0.297, + "step": 54840 + }, + { + "epoch": 3.8054958018180556, + "grad_norm": 1.0411710739135742, + "learning_rate": 1.2375905475320192e-06, + "loss": 0.3294, + "step": 54841 + }, + { + "epoch": 3.8055651932551524, + "grad_norm": 1.0078644752502441, + "learning_rate": 1.2367094822392066e-06, + "loss": 0.2719, + "step": 54842 + }, + { + "epoch": 3.805634584692249, + "grad_norm": 0.9738138914108276, + "learning_rate": 1.235828728730648e-06, + "loss": 0.3089, + "step": 54843 + }, + { + "epoch": 3.8057039761293456, + "grad_norm": 0.835785448551178, + "learning_rate": 1.234948287009119e-06, + "loss": 0.2081, + "step": 54844 + }, + { + "epoch": 3.8057733675664425, + "grad_norm": 0.8203831315040588, + "learning_rate": 1.2340681570773949e-06, + "loss": 0.19, + "step": 54845 + }, + { + "epoch": 3.805842759003539, + "grad_norm": 0.6501784920692444, + "learning_rate": 1.2331883389382516e-06, + "loss": 0.13, + "step": 54846 + }, + { + "epoch": 3.8059121504406357, + "grad_norm": 0.710655927658081, + "learning_rate": 1.2323088325944976e-06, + "loss": 0.236, + "step": 54847 + }, + { + "epoch": 3.805981541877732, + "grad_norm": 0.8592118620872498, + "learning_rate": 1.2314296380488644e-06, + "loss": 0.4346, + "step": 54848 + }, + { + "epoch": 3.806050933314829, + "grad_norm": 0.831579327583313, + "learning_rate": 1.2305507553041495e-06, + "loss": 0.2482, + "step": 54849 + }, + { + "epoch": 3.8061203247519257, + "grad_norm": 0.7754706144332886, + "learning_rate": 1.2296721843631397e-06, + "loss": 0.2288, + "step": 54850 + }, + { + "epoch": 3.806189716189022, + "grad_norm": 0.9597982168197632, + "learning_rate": 1.2287939252285884e-06, + "loss": 0.2698, + "step": 54851 + }, + { + "epoch": 3.806259107626119, + "grad_norm": 0.7843778729438782, + "learning_rate": 1.2279159779032712e-06, + "loss": 0.2312, + "step": 54852 + }, + { + "epoch": 3.8063284990632154, + "grad_norm": 0.9796881675720215, + "learning_rate": 1.2270383423899634e-06, + "loss": 0.2125, + "step": 54853 + }, + { + "epoch": 3.806397890500312, + "grad_norm": 0.8645321726799011, + "learning_rate": 1.2261610186914407e-06, + "loss": 0.2468, + "step": 54854 + }, + { + "epoch": 3.806467281937409, + "grad_norm": 0.933324933052063, + "learning_rate": 1.2252840068104675e-06, + "loss": 0.3427, + "step": 54855 + }, + { + "epoch": 3.806536673374506, + "grad_norm": 0.9136168360710144, + "learning_rate": 1.2244073067498197e-06, + "loss": 0.1917, + "step": 54856 + }, + { + "epoch": 3.8066060648116022, + "grad_norm": 0.6796775460243225, + "learning_rate": 1.22353091851225e-06, + "loss": 0.1521, + "step": 54857 + }, + { + "epoch": 3.806675456248699, + "grad_norm": 1.7840768098831177, + "learning_rate": 1.2226548421005236e-06, + "loss": 0.3182, + "step": 54858 + }, + { + "epoch": 3.8067448476857955, + "grad_norm": 1.2466533184051514, + "learning_rate": 1.2217790775174265e-06, + "loss": 0.2393, + "step": 54859 + }, + { + "epoch": 3.8068142391228923, + "grad_norm": 0.9975258111953735, + "learning_rate": 1.2209036247657123e-06, + "loss": 0.2738, + "step": 54860 + }, + { + "epoch": 3.806883630559989, + "grad_norm": 0.735541582107544, + "learning_rate": 1.2200284838481346e-06, + "loss": 0.2181, + "step": 54861 + }, + { + "epoch": 3.8069530219970855, + "grad_norm": 0.8509876728057861, + "learning_rate": 1.2191536547674797e-06, + "loss": 0.1854, + "step": 54862 + }, + { + "epoch": 3.8070224134341824, + "grad_norm": 0.9082764983177185, + "learning_rate": 1.2182791375264901e-06, + "loss": 0.2062, + "step": 54863 + }, + { + "epoch": 3.8070918048712787, + "grad_norm": 0.6015109419822693, + "learning_rate": 1.2174049321279412e-06, + "loss": 0.1724, + "step": 54864 + }, + { + "epoch": 3.8071611963083756, + "grad_norm": 0.6408658027648926, + "learning_rate": 1.2165310385745753e-06, + "loss": 0.19, + "step": 54865 + }, + { + "epoch": 3.8072305877454724, + "grad_norm": 0.8082396388053894, + "learning_rate": 1.215657456869157e-06, + "loss": 0.1543, + "step": 54866 + }, + { + "epoch": 3.807299979182569, + "grad_norm": 2.022921323776245, + "learning_rate": 1.2147841870144505e-06, + "loss": 0.2406, + "step": 54867 + }, + { + "epoch": 3.8073693706196656, + "grad_norm": 0.8409140706062317, + "learning_rate": 1.2139112290132094e-06, + "loss": 0.233, + "step": 54868 + }, + { + "epoch": 3.807438762056762, + "grad_norm": 0.8642407655715942, + "learning_rate": 1.2130385828681868e-06, + "loss": 0.1906, + "step": 54869 + }, + { + "epoch": 3.807508153493859, + "grad_norm": 0.980782687664032, + "learning_rate": 1.2121662485821473e-06, + "loss": 0.25, + "step": 54870 + }, + { + "epoch": 3.8075775449309557, + "grad_norm": 0.8932204842567444, + "learning_rate": 1.2112942261578331e-06, + "loss": 0.2586, + "step": 54871 + }, + { + "epoch": 3.807646936368052, + "grad_norm": 1.0409901142120361, + "learning_rate": 1.2104225155979976e-06, + "loss": 0.2728, + "step": 54872 + }, + { + "epoch": 3.807716327805149, + "grad_norm": 0.7484800815582275, + "learning_rate": 1.209551116905394e-06, + "loss": 0.2096, + "step": 54873 + }, + { + "epoch": 3.8077857192422453, + "grad_norm": 1.0615959167480469, + "learning_rate": 1.2086800300827873e-06, + "loss": 0.2898, + "step": 54874 + }, + { + "epoch": 3.807855110679342, + "grad_norm": 0.7153548002243042, + "learning_rate": 1.2078092551329078e-06, + "loss": 0.1996, + "step": 54875 + }, + { + "epoch": 3.807924502116439, + "grad_norm": 0.5628870725631714, + "learning_rate": 1.2069387920585096e-06, + "loss": 0.1553, + "step": 54876 + }, + { + "epoch": 3.8079938935535353, + "grad_norm": 1.2596224546432495, + "learning_rate": 1.2060686408623568e-06, + "loss": 0.2055, + "step": 54877 + }, + { + "epoch": 3.808063284990632, + "grad_norm": 1.4129654169082642, + "learning_rate": 1.2051988015471694e-06, + "loss": 0.2719, + "step": 54878 + }, + { + "epoch": 3.8081326764277286, + "grad_norm": 0.8729109168052673, + "learning_rate": 1.2043292741157118e-06, + "loss": 0.2237, + "step": 54879 + }, + { + "epoch": 3.8082020678648254, + "grad_norm": 1.239699125289917, + "learning_rate": 1.2034600585707267e-06, + "loss": 0.3467, + "step": 54880 + }, + { + "epoch": 3.8082714593019222, + "grad_norm": 1.0976353883743286, + "learning_rate": 1.202591154914956e-06, + "loss": 0.1831, + "step": 54881 + }, + { + "epoch": 3.808340850739019, + "grad_norm": 0.6953972578048706, + "learning_rate": 1.2017225631511419e-06, + "loss": 0.1716, + "step": 54882 + }, + { + "epoch": 3.8084102421761155, + "grad_norm": 0.820401668548584, + "learning_rate": 1.200854283282027e-06, + "loss": 0.2386, + "step": 54883 + }, + { + "epoch": 3.8084796336132123, + "grad_norm": 1.0217068195343018, + "learning_rate": 1.199986315310353e-06, + "loss": 0.3443, + "step": 54884 + }, + { + "epoch": 3.8085490250503087, + "grad_norm": 0.8812357187271118, + "learning_rate": 1.1991186592388516e-06, + "loss": 0.219, + "step": 54885 + }, + { + "epoch": 3.8086184164874055, + "grad_norm": 0.8699870705604553, + "learning_rate": 1.1982513150702757e-06, + "loss": 0.1998, + "step": 54886 + }, + { + "epoch": 3.8086878079245023, + "grad_norm": 0.8356900215148926, + "learning_rate": 1.1973842828073568e-06, + "loss": 0.2785, + "step": 54887 + }, + { + "epoch": 3.8087571993615987, + "grad_norm": 0.7005043625831604, + "learning_rate": 1.1965175624528368e-06, + "loss": 0.2194, + "step": 54888 + }, + { + "epoch": 3.8088265907986956, + "grad_norm": 0.6635180711746216, + "learning_rate": 1.1956511540094363e-06, + "loss": 0.2081, + "step": 54889 + }, + { + "epoch": 3.808895982235792, + "grad_norm": 1.044981598854065, + "learning_rate": 1.194785057479908e-06, + "loss": 0.3105, + "step": 54890 + }, + { + "epoch": 3.808965373672889, + "grad_norm": 1.359689712524414, + "learning_rate": 1.1939192728669835e-06, + "loss": 0.383, + "step": 54891 + }, + { + "epoch": 3.8090347651099856, + "grad_norm": 1.1524914503097534, + "learning_rate": 1.1930538001733938e-06, + "loss": 0.2377, + "step": 54892 + }, + { + "epoch": 3.809104156547082, + "grad_norm": 0.5702936053276062, + "learning_rate": 1.1921886394018589e-06, + "loss": 0.2145, + "step": 54893 + }, + { + "epoch": 3.809173547984179, + "grad_norm": 0.7936190962791443, + "learning_rate": 1.1913237905551323e-06, + "loss": 0.257, + "step": 54894 + }, + { + "epoch": 3.8092429394212752, + "grad_norm": 0.7509891390800476, + "learning_rate": 1.1904592536359227e-06, + "loss": 0.1466, + "step": 54895 + }, + { + "epoch": 3.809312330858372, + "grad_norm": 3.298004388809204, + "learning_rate": 1.1895950286469726e-06, + "loss": 0.3172, + "step": 54896 + }, + { + "epoch": 3.809381722295469, + "grad_norm": 0.7998113036155701, + "learning_rate": 1.1887311155910018e-06, + "loss": 0.2324, + "step": 54897 + }, + { + "epoch": 3.8094511137325653, + "grad_norm": 0.894985020160675, + "learning_rate": 1.187867514470753e-06, + "loss": 0.1839, + "step": 54898 + }, + { + "epoch": 3.809520505169662, + "grad_norm": 1.0338120460510254, + "learning_rate": 1.1870042252889346e-06, + "loss": 0.4067, + "step": 54899 + }, + { + "epoch": 3.8095898966067585, + "grad_norm": 0.7354503870010376, + "learning_rate": 1.1861412480482892e-06, + "loss": 0.3305, + "step": 54900 + }, + { + "epoch": 3.8096592880438553, + "grad_norm": 1.0606745481491089, + "learning_rate": 1.1852785827515144e-06, + "loss": 0.2778, + "step": 54901 + }, + { + "epoch": 3.809728679480952, + "grad_norm": 1.0289286375045776, + "learning_rate": 1.184416229401364e-06, + "loss": 0.3172, + "step": 54902 + }, + { + "epoch": 3.8097980709180486, + "grad_norm": 0.8842047452926636, + "learning_rate": 1.1835541880005464e-06, + "loss": 0.2718, + "step": 54903 + }, + { + "epoch": 3.8098674623551454, + "grad_norm": 0.565683126449585, + "learning_rate": 1.1826924585517707e-06, + "loss": 0.1836, + "step": 54904 + }, + { + "epoch": 3.809936853792242, + "grad_norm": 1.0574575662612915, + "learning_rate": 1.1818310410577904e-06, + "loss": 0.3282, + "step": 54905 + }, + { + "epoch": 3.8100062452293386, + "grad_norm": 0.6923700571060181, + "learning_rate": 1.1809699355212921e-06, + "loss": 0.1957, + "step": 54906 + }, + { + "epoch": 3.8100756366664355, + "grad_norm": 0.6388658285140991, + "learning_rate": 1.1801091419450073e-06, + "loss": 0.1912, + "step": 54907 + }, + { + "epoch": 3.8101450281035323, + "grad_norm": 1.0283790826797485, + "learning_rate": 1.1792486603316555e-06, + "loss": 0.2421, + "step": 54908 + }, + { + "epoch": 3.8102144195406287, + "grad_norm": 0.8785754442214966, + "learning_rate": 1.178388490683946e-06, + "loss": 0.236, + "step": 54909 + }, + { + "epoch": 3.8102838109777255, + "grad_norm": 0.8738521337509155, + "learning_rate": 1.1775286330045987e-06, + "loss": 0.24, + "step": 54910 + }, + { + "epoch": 3.810353202414822, + "grad_norm": 0.8741692900657654, + "learning_rate": 1.1766690872963227e-06, + "loss": 0.287, + "step": 54911 + }, + { + "epoch": 3.8104225938519187, + "grad_norm": 0.7548696398735046, + "learning_rate": 1.17580985356186e-06, + "loss": 0.1885, + "step": 54912 + }, + { + "epoch": 3.8104919852890156, + "grad_norm": 0.7814151644706726, + "learning_rate": 1.1749509318038753e-06, + "loss": 0.1539, + "step": 54913 + }, + { + "epoch": 3.810561376726112, + "grad_norm": 0.9416444301605225, + "learning_rate": 1.174092322025111e-06, + "loss": 0.2631, + "step": 54914 + }, + { + "epoch": 3.810630768163209, + "grad_norm": 0.9394712448120117, + "learning_rate": 1.1732340242282758e-06, + "loss": 0.3088, + "step": 54915 + }, + { + "epoch": 3.810700159600305, + "grad_norm": 0.8650161027908325, + "learning_rate": 1.1723760384160676e-06, + "loss": 0.3887, + "step": 54916 + }, + { + "epoch": 3.810769551037402, + "grad_norm": 0.9425433874130249, + "learning_rate": 1.1715183645912176e-06, + "loss": 0.3706, + "step": 54917 + }, + { + "epoch": 3.810838942474499, + "grad_norm": 0.920409619808197, + "learning_rate": 1.1706610027564013e-06, + "loss": 0.3455, + "step": 54918 + }, + { + "epoch": 3.8109083339115952, + "grad_norm": 1.0708537101745605, + "learning_rate": 1.16980395291435e-06, + "loss": 0.1895, + "step": 54919 + }, + { + "epoch": 3.810977725348692, + "grad_norm": 0.8205761313438416, + "learning_rate": 1.1689472150677616e-06, + "loss": 0.2453, + "step": 54920 + }, + { + "epoch": 3.8110471167857884, + "grad_norm": 0.7704923152923584, + "learning_rate": 1.168090789219345e-06, + "loss": 0.1792, + "step": 54921 + }, + { + "epoch": 3.8111165082228853, + "grad_norm": 0.7438368201255798, + "learning_rate": 1.1672346753717866e-06, + "loss": 0.2305, + "step": 54922 + }, + { + "epoch": 3.811185899659982, + "grad_norm": 0.8418996334075928, + "learning_rate": 1.1663788735278069e-06, + "loss": 0.2024, + "step": 54923 + }, + { + "epoch": 3.8112552910970785, + "grad_norm": 0.6454266905784607, + "learning_rate": 1.1655233836901037e-06, + "loss": 0.2176, + "step": 54924 + }, + { + "epoch": 3.8113246825341753, + "grad_norm": 0.7525631189346313, + "learning_rate": 1.1646682058613855e-06, + "loss": 0.2696, + "step": 54925 + }, + { + "epoch": 3.8113940739712717, + "grad_norm": 0.853488028049469, + "learning_rate": 1.1638133400443285e-06, + "loss": 0.1983, + "step": 54926 + }, + { + "epoch": 3.8114634654083686, + "grad_norm": 0.7876091003417969, + "learning_rate": 1.1629587862416413e-06, + "loss": 0.2159, + "step": 54927 + }, + { + "epoch": 3.8115328568454654, + "grad_norm": 1.07789146900177, + "learning_rate": 1.162104544456033e-06, + "loss": 0.3151, + "step": 54928 + }, + { + "epoch": 3.8116022482825618, + "grad_norm": 0.9637660980224609, + "learning_rate": 1.1612506146902014e-06, + "loss": 0.2034, + "step": 54929 + }, + { + "epoch": 3.8116716397196586, + "grad_norm": 0.8669813871383667, + "learning_rate": 1.160396996946822e-06, + "loss": 0.1581, + "step": 54930 + }, + { + "epoch": 3.811741031156755, + "grad_norm": 0.8705934286117554, + "learning_rate": 1.1595436912286039e-06, + "loss": 0.2418, + "step": 54931 + }, + { + "epoch": 3.811810422593852, + "grad_norm": 1.0207648277282715, + "learning_rate": 1.158690697538256e-06, + "loss": 0.2843, + "step": 54932 + }, + { + "epoch": 3.8118798140309487, + "grad_norm": 0.6726106405258179, + "learning_rate": 1.157838015878432e-06, + "loss": 0.2288, + "step": 54933 + }, + { + "epoch": 3.8119492054680455, + "grad_norm": 0.8209892511367798, + "learning_rate": 1.1569856462518514e-06, + "loss": 0.2352, + "step": 54934 + }, + { + "epoch": 3.812018596905142, + "grad_norm": 1.0952855348587036, + "learning_rate": 1.1561335886612012e-06, + "loss": 0.4233, + "step": 54935 + }, + { + "epoch": 3.8120879883422387, + "grad_norm": 0.8796757459640503, + "learning_rate": 1.1552818431091682e-06, + "loss": 0.2248, + "step": 54936 + }, + { + "epoch": 3.812157379779335, + "grad_norm": 1.7079179286956787, + "learning_rate": 1.1544304095984394e-06, + "loss": 0.2374, + "step": 54937 + }, + { + "epoch": 3.812226771216432, + "grad_norm": 0.9824504256248474, + "learning_rate": 1.153579288131712e-06, + "loss": 0.3008, + "step": 54938 + }, + { + "epoch": 3.8122961626535288, + "grad_norm": 0.6749600768089294, + "learning_rate": 1.1527284787116732e-06, + "loss": 0.179, + "step": 54939 + }, + { + "epoch": 3.812365554090625, + "grad_norm": 0.5393670201301575, + "learning_rate": 1.1518779813409875e-06, + "loss": 0.1651, + "step": 54940 + }, + { + "epoch": 3.812434945527722, + "grad_norm": 0.9814626574516296, + "learning_rate": 1.1510277960223637e-06, + "loss": 0.301, + "step": 54941 + }, + { + "epoch": 3.8125043369648184, + "grad_norm": 0.8655776977539062, + "learning_rate": 1.1501779227584663e-06, + "loss": 0.2724, + "step": 54942 + }, + { + "epoch": 3.812573728401915, + "grad_norm": 1.3849762678146362, + "learning_rate": 1.1493283615519934e-06, + "loss": 0.2293, + "step": 54943 + }, + { + "epoch": 3.812643119839012, + "grad_norm": 0.8023849129676819, + "learning_rate": 1.1484791124056315e-06, + "loss": 0.2274, + "step": 54944 + }, + { + "epoch": 3.8127125112761084, + "grad_norm": 0.8971165418624878, + "learning_rate": 1.1476301753220342e-06, + "loss": 0.1771, + "step": 54945 + }, + { + "epoch": 3.8127819027132053, + "grad_norm": 0.527270495891571, + "learning_rate": 1.1467815503039215e-06, + "loss": 0.1554, + "step": 54946 + }, + { + "epoch": 3.8128512941503017, + "grad_norm": 1.0287621021270752, + "learning_rate": 1.1459332373539356e-06, + "loss": 0.2107, + "step": 54947 + }, + { + "epoch": 3.8129206855873985, + "grad_norm": 1.0768064260482788, + "learning_rate": 1.1450852364747743e-06, + "loss": 0.3037, + "step": 54948 + }, + { + "epoch": 3.8129900770244953, + "grad_norm": 0.6176892518997192, + "learning_rate": 1.1442375476691026e-06, + "loss": 0.1326, + "step": 54949 + }, + { + "epoch": 3.8130594684615917, + "grad_norm": 0.8817666172981262, + "learning_rate": 1.1433901709396067e-06, + "loss": 0.1716, + "step": 54950 + }, + { + "epoch": 3.8131288598986885, + "grad_norm": 1.1077711582183838, + "learning_rate": 1.1425431062889625e-06, + "loss": 0.222, + "step": 54951 + }, + { + "epoch": 3.813198251335785, + "grad_norm": 0.7581732273101807, + "learning_rate": 1.1416963537198455e-06, + "loss": 0.1244, + "step": 54952 + }, + { + "epoch": 3.8132676427728818, + "grad_norm": 0.5523532032966614, + "learning_rate": 1.1408499132349094e-06, + "loss": 0.2411, + "step": 54953 + }, + { + "epoch": 3.8133370342099786, + "grad_norm": 0.9598371386528015, + "learning_rate": 1.1400037848368517e-06, + "loss": 0.2552, + "step": 54954 + }, + { + "epoch": 3.813406425647075, + "grad_norm": 0.8434988260269165, + "learning_rate": 1.139157968528326e-06, + "loss": 0.2867, + "step": 54955 + }, + { + "epoch": 3.813475817084172, + "grad_norm": 0.9781166315078735, + "learning_rate": 1.138312464312008e-06, + "loss": 0.2266, + "step": 54956 + }, + { + "epoch": 3.813545208521268, + "grad_norm": 1.2214155197143555, + "learning_rate": 1.1374672721905733e-06, + "loss": 0.2744, + "step": 54957 + }, + { + "epoch": 3.813614599958365, + "grad_norm": 0.8490088582038879, + "learning_rate": 1.1366223921666753e-06, + "loss": 0.1916, + "step": 54958 + }, + { + "epoch": 3.813683991395462, + "grad_norm": 0.6850140690803528, + "learning_rate": 1.1357778242430006e-06, + "loss": 0.252, + "step": 54959 + }, + { + "epoch": 3.8137533828325587, + "grad_norm": 0.8666943907737732, + "learning_rate": 1.1349335684222029e-06, + "loss": 0.2809, + "step": 54960 + }, + { + "epoch": 3.813822774269655, + "grad_norm": 0.9838248491287231, + "learning_rate": 1.1340896247069465e-06, + "loss": 0.2863, + "step": 54961 + }, + { + "epoch": 3.813892165706752, + "grad_norm": 0.6943295001983643, + "learning_rate": 1.1332459930998962e-06, + "loss": 0.2361, + "step": 54962 + }, + { + "epoch": 3.8139615571438483, + "grad_norm": 0.7674679160118103, + "learning_rate": 1.1324026736037164e-06, + "loss": 0.3314, + "step": 54963 + }, + { + "epoch": 3.814030948580945, + "grad_norm": 1.4977869987487793, + "learning_rate": 1.1315596662210715e-06, + "loss": 0.3182, + "step": 54964 + }, + { + "epoch": 3.814100340018042, + "grad_norm": 0.8820518851280212, + "learning_rate": 1.1307169709546262e-06, + "loss": 0.2681, + "step": 54965 + }, + { + "epoch": 3.8141697314551384, + "grad_norm": 1.069006323814392, + "learning_rate": 1.1298745878070337e-06, + "loss": 0.2614, + "step": 54966 + }, + { + "epoch": 3.814239122892235, + "grad_norm": 0.8073067665100098, + "learning_rate": 1.129032516780959e-06, + "loss": 0.1588, + "step": 54967 + }, + { + "epoch": 3.8143085143293316, + "grad_norm": 0.911872923374176, + "learning_rate": 1.1281907578790551e-06, + "loss": 0.3878, + "step": 54968 + }, + { + "epoch": 3.8143779057664284, + "grad_norm": 0.9598478674888611, + "learning_rate": 1.1273493111039757e-06, + "loss": 0.2503, + "step": 54969 + }, + { + "epoch": 3.8144472972035253, + "grad_norm": 1.0923843383789062, + "learning_rate": 1.1265081764583962e-06, + "loss": 0.2835, + "step": 54970 + }, + { + "epoch": 3.8145166886406217, + "grad_norm": 0.938495397567749, + "learning_rate": 1.125667353944948e-06, + "loss": 0.3298, + "step": 54971 + }, + { + "epoch": 3.8145860800777185, + "grad_norm": 0.7459365725517273, + "learning_rate": 1.1248268435662956e-06, + "loss": 0.2613, + "step": 54972 + }, + { + "epoch": 3.814655471514815, + "grad_norm": 0.9394201040267944, + "learning_rate": 1.1239866453251036e-06, + "loss": 0.1337, + "step": 54973 + }, + { + "epoch": 3.8147248629519117, + "grad_norm": 1.0491505861282349, + "learning_rate": 1.1231467592240142e-06, + "loss": 0.4481, + "step": 54974 + }, + { + "epoch": 3.8147942543890085, + "grad_norm": 0.8078936338424683, + "learning_rate": 1.1223071852656697e-06, + "loss": 0.2108, + "step": 54975 + }, + { + "epoch": 3.814863645826105, + "grad_norm": 0.948390543460846, + "learning_rate": 1.1214679234527349e-06, + "loss": 0.2077, + "step": 54976 + }, + { + "epoch": 3.8149330372632018, + "grad_norm": 0.6695556640625, + "learning_rate": 1.1206289737878517e-06, + "loss": 0.1512, + "step": 54977 + }, + { + "epoch": 3.815002428700298, + "grad_norm": 0.9661774635314941, + "learning_rate": 1.1197903362736627e-06, + "loss": 0.2206, + "step": 54978 + }, + { + "epoch": 3.815071820137395, + "grad_norm": 0.9059375524520874, + "learning_rate": 1.1189520109128326e-06, + "loss": 0.1824, + "step": 54979 + }, + { + "epoch": 3.815141211574492, + "grad_norm": 1.1185450553894043, + "learning_rate": 1.1181139977079925e-06, + "loss": 0.3021, + "step": 54980 + }, + { + "epoch": 3.815210603011588, + "grad_norm": 0.7971089482307434, + "learning_rate": 1.1172762966618066e-06, + "loss": 0.2321, + "step": 54981 + }, + { + "epoch": 3.815279994448685, + "grad_norm": 0.9120199680328369, + "learning_rate": 1.1164389077768955e-06, + "loss": 0.1726, + "step": 54982 + }, + { + "epoch": 3.8153493858857814, + "grad_norm": 0.8690612316131592, + "learning_rate": 1.1156018310559236e-06, + "loss": 0.2512, + "step": 54983 + }, + { + "epoch": 3.8154187773228783, + "grad_norm": 0.44788485765457153, + "learning_rate": 1.1147650665015109e-06, + "loss": 0.1319, + "step": 54984 + }, + { + "epoch": 3.815488168759975, + "grad_norm": 0.7634284496307373, + "learning_rate": 1.1139286141163218e-06, + "loss": 0.1851, + "step": 54985 + }, + { + "epoch": 3.815557560197072, + "grad_norm": 0.7998407483100891, + "learning_rate": 1.113092473902988e-06, + "loss": 0.1964, + "step": 54986 + }, + { + "epoch": 3.8156269516341683, + "grad_norm": 0.7522568702697754, + "learning_rate": 1.1122566458641514e-06, + "loss": 0.1682, + "step": 54987 + }, + { + "epoch": 3.815696343071265, + "grad_norm": 0.9859260320663452, + "learning_rate": 1.1114211300024325e-06, + "loss": 0.2845, + "step": 54988 + }, + { + "epoch": 3.8157657345083615, + "grad_norm": 0.509558379650116, + "learning_rate": 1.1105859263204954e-06, + "loss": 0.0878, + "step": 54989 + }, + { + "epoch": 3.8158351259454584, + "grad_norm": 0.8491736650466919, + "learning_rate": 1.1097510348209716e-06, + "loss": 0.2115, + "step": 54990 + }, + { + "epoch": 3.815904517382555, + "grad_norm": 0.9232484102249146, + "learning_rate": 1.1089164555064812e-06, + "loss": 0.2157, + "step": 54991 + }, + { + "epoch": 3.8159739088196516, + "grad_norm": 0.6303938031196594, + "learning_rate": 1.1080821883796666e-06, + "loss": 0.2386, + "step": 54992 + }, + { + "epoch": 3.8160433002567484, + "grad_norm": 0.8583338260650635, + "learning_rate": 1.1072482334431589e-06, + "loss": 0.3364, + "step": 54993 + }, + { + "epoch": 3.816112691693845, + "grad_norm": 0.6923225522041321, + "learning_rate": 1.1064145906996004e-06, + "loss": 0.1818, + "step": 54994 + }, + { + "epoch": 3.8161820831309416, + "grad_norm": 1.0284154415130615, + "learning_rate": 1.1055812601516225e-06, + "loss": 0.2638, + "step": 54995 + }, + { + "epoch": 3.8162514745680385, + "grad_norm": 0.9242666363716125, + "learning_rate": 1.1047482418018452e-06, + "loss": 0.2527, + "step": 54996 + }, + { + "epoch": 3.816320866005135, + "grad_norm": 0.9581491351127625, + "learning_rate": 1.1039155356528996e-06, + "loss": 0.1833, + "step": 54997 + }, + { + "epoch": 3.8163902574422317, + "grad_norm": 0.9255644083023071, + "learning_rate": 1.1030831417074284e-06, + "loss": 0.223, + "step": 54998 + }, + { + "epoch": 3.816459648879328, + "grad_norm": 0.908723771572113, + "learning_rate": 1.1022510599680403e-06, + "loss": 0.3237, + "step": 54999 + }, + { + "epoch": 3.816529040316425, + "grad_norm": 0.6753847599029541, + "learning_rate": 1.1014192904373666e-06, + "loss": 0.2064, + "step": 55000 + }, + { + "epoch": 3.8165984317535218, + "grad_norm": 0.7168920040130615, + "learning_rate": 1.1005878331180496e-06, + "loss": 0.2694, + "step": 55001 + }, + { + "epoch": 3.816667823190618, + "grad_norm": 1.2259507179260254, + "learning_rate": 1.0997566880127098e-06, + "loss": 0.3084, + "step": 55002 + }, + { + "epoch": 3.816737214627715, + "grad_norm": 0.8392354249954224, + "learning_rate": 1.0989258551239445e-06, + "loss": 0.2307, + "step": 55003 + }, + { + "epoch": 3.8168066060648114, + "grad_norm": 0.9546611309051514, + "learning_rate": 1.0980953344544077e-06, + "loss": 0.3772, + "step": 55004 + }, + { + "epoch": 3.816875997501908, + "grad_norm": 0.6613355875015259, + "learning_rate": 1.0972651260067078e-06, + "loss": 0.1773, + "step": 55005 + }, + { + "epoch": 3.816945388939005, + "grad_norm": 0.9561030864715576, + "learning_rate": 1.0964352297834546e-06, + "loss": 0.2159, + "step": 55006 + }, + { + "epoch": 3.8170147803761014, + "grad_norm": 1.1946866512298584, + "learning_rate": 1.0956056457872898e-06, + "loss": 0.3841, + "step": 55007 + }, + { + "epoch": 3.8170841718131983, + "grad_norm": 1.0582400560379028, + "learning_rate": 1.094776374020834e-06, + "loss": 0.1834, + "step": 55008 + }, + { + "epoch": 3.8171535632502946, + "grad_norm": 1.00166654586792, + "learning_rate": 1.0939474144866847e-06, + "loss": 0.247, + "step": 55009 + }, + { + "epoch": 3.8172229546873915, + "grad_norm": 2.554657459259033, + "learning_rate": 1.0931187671874621e-06, + "loss": 0.1995, + "step": 55010 + }, + { + "epoch": 3.8172923461244883, + "grad_norm": 1.1514958143234253, + "learning_rate": 1.0922904321257977e-06, + "loss": 0.2599, + "step": 55011 + }, + { + "epoch": 3.817361737561585, + "grad_norm": 0.8574109673500061, + "learning_rate": 1.0914624093042891e-06, + "loss": 0.3301, + "step": 55012 + }, + { + "epoch": 3.8174311289986815, + "grad_norm": 0.8081821799278259, + "learning_rate": 1.0906346987255566e-06, + "loss": 0.3125, + "step": 55013 + }, + { + "epoch": 3.8175005204357784, + "grad_norm": 0.9026225209236145, + "learning_rate": 1.0898073003922204e-06, + "loss": 0.2098, + "step": 55014 + }, + { + "epoch": 3.8175699118728748, + "grad_norm": 0.9559450149536133, + "learning_rate": 1.0889802143068894e-06, + "loss": 0.2501, + "step": 55015 + }, + { + "epoch": 3.8176393033099716, + "grad_norm": 0.7707690596580505, + "learning_rate": 1.0881534404721727e-06, + "loss": 0.27, + "step": 55016 + }, + { + "epoch": 3.8177086947470684, + "grad_norm": 0.8998335003852844, + "learning_rate": 1.0873269788906793e-06, + "loss": 0.2047, + "step": 55017 + }, + { + "epoch": 3.817778086184165, + "grad_norm": 1.2570327520370483, + "learning_rate": 1.086500829565007e-06, + "loss": 0.4462, + "step": 55018 + }, + { + "epoch": 3.8178474776212616, + "grad_norm": 0.697902262210846, + "learning_rate": 1.0856749924977872e-06, + "loss": 0.228, + "step": 55019 + }, + { + "epoch": 3.817916869058358, + "grad_norm": 2.113715648651123, + "learning_rate": 1.0848494676916066e-06, + "loss": 0.2887, + "step": 55020 + }, + { + "epoch": 3.817986260495455, + "grad_norm": 1.146703839302063, + "learning_rate": 1.0840242551490853e-06, + "loss": 0.1604, + "step": 55021 + }, + { + "epoch": 3.8180556519325517, + "grad_norm": 1.3477776050567627, + "learning_rate": 1.0831993548728326e-06, + "loss": 0.3899, + "step": 55022 + }, + { + "epoch": 3.818125043369648, + "grad_norm": 1.0818302631378174, + "learning_rate": 1.082374766865435e-06, + "loss": 0.2936, + "step": 55023 + }, + { + "epoch": 3.818194434806745, + "grad_norm": 1.05577552318573, + "learning_rate": 1.0815504911295016e-06, + "loss": 0.2933, + "step": 55024 + }, + { + "epoch": 3.8182638262438413, + "grad_norm": 0.8662980198860168, + "learning_rate": 1.0807265276676414e-06, + "loss": 0.316, + "step": 55025 + }, + { + "epoch": 3.818333217680938, + "grad_norm": 0.9848813414573669, + "learning_rate": 1.0799028764824415e-06, + "loss": 0.2706, + "step": 55026 + }, + { + "epoch": 3.818402609118035, + "grad_norm": 0.6806034445762634, + "learning_rate": 1.0790795375765105e-06, + "loss": 0.2219, + "step": 55027 + }, + { + "epoch": 3.8184720005551314, + "grad_norm": 0.8335594534873962, + "learning_rate": 1.0782565109524578e-06, + "loss": 0.1976, + "step": 55028 + }, + { + "epoch": 3.818541391992228, + "grad_norm": 1.228415846824646, + "learning_rate": 1.07743379661287e-06, + "loss": 0.3491, + "step": 55029 + }, + { + "epoch": 3.8186107834293246, + "grad_norm": 0.7983219027519226, + "learning_rate": 1.0766113945603451e-06, + "loss": 0.216, + "step": 55030 + }, + { + "epoch": 3.8186801748664214, + "grad_norm": 0.8378092050552368, + "learning_rate": 1.075789304797481e-06, + "loss": 0.2237, + "step": 55031 + }, + { + "epoch": 3.8187495663035183, + "grad_norm": 0.8207027912139893, + "learning_rate": 1.0749675273268757e-06, + "loss": 0.2041, + "step": 55032 + }, + { + "epoch": 3.818818957740615, + "grad_norm": 0.59365314245224, + "learning_rate": 1.0741460621511158e-06, + "loss": 0.1902, + "step": 55033 + }, + { + "epoch": 3.8188883491777115, + "grad_norm": 1.1133688688278198, + "learning_rate": 1.0733249092727992e-06, + "loss": 0.2793, + "step": 55034 + }, + { + "epoch": 3.818957740614808, + "grad_norm": 0.9461538791656494, + "learning_rate": 1.0725040686945243e-06, + "loss": 0.1528, + "step": 55035 + }, + { + "epoch": 3.8190271320519047, + "grad_norm": 0.7854909896850586, + "learning_rate": 1.0716835404188773e-06, + "loss": 0.2425, + "step": 55036 + }, + { + "epoch": 3.8190965234890015, + "grad_norm": 0.808457612991333, + "learning_rate": 1.0708633244484344e-06, + "loss": 0.2385, + "step": 55037 + }, + { + "epoch": 3.8191659149260984, + "grad_norm": 0.9563588500022888, + "learning_rate": 1.0700434207858046e-06, + "loss": 0.3621, + "step": 55038 + }, + { + "epoch": 3.8192353063631947, + "grad_norm": 1.111936330795288, + "learning_rate": 1.0692238294335744e-06, + "loss": 0.3841, + "step": 55039 + }, + { + "epoch": 3.8193046978002916, + "grad_norm": 1.65395987033844, + "learning_rate": 1.0684045503943197e-06, + "loss": 0.386, + "step": 55040 + }, + { + "epoch": 3.819374089237388, + "grad_norm": 0.8374661803245544, + "learning_rate": 1.0675855836706384e-06, + "loss": 0.1725, + "step": 55041 + }, + { + "epoch": 3.819443480674485, + "grad_norm": 0.9878285527229309, + "learning_rate": 1.0667669292651062e-06, + "loss": 0.2157, + "step": 55042 + }, + { + "epoch": 3.8195128721115816, + "grad_norm": 0.9242804050445557, + "learning_rate": 1.065948587180321e-06, + "loss": 0.1446, + "step": 55043 + }, + { + "epoch": 3.819582263548678, + "grad_norm": 0.9964492917060852, + "learning_rate": 1.0651305574188586e-06, + "loss": 0.2408, + "step": 55044 + }, + { + "epoch": 3.819651654985775, + "grad_norm": 0.8561482429504395, + "learning_rate": 1.0643128399832947e-06, + "loss": 0.2395, + "step": 55045 + }, + { + "epoch": 3.8197210464228712, + "grad_norm": 0.7229714393615723, + "learning_rate": 1.063495434876216e-06, + "loss": 0.2213, + "step": 55046 + }, + { + "epoch": 3.819790437859968, + "grad_norm": 0.7618684768676758, + "learning_rate": 1.0626783421001985e-06, + "loss": 0.2241, + "step": 55047 + }, + { + "epoch": 3.819859829297065, + "grad_norm": 0.9979142546653748, + "learning_rate": 1.0618615616578286e-06, + "loss": 0.1808, + "step": 55048 + }, + { + "epoch": 3.8199292207341613, + "grad_norm": 0.7545778155326843, + "learning_rate": 1.0610450935516824e-06, + "loss": 0.2534, + "step": 55049 + }, + { + "epoch": 3.819998612171258, + "grad_norm": 0.7690707445144653, + "learning_rate": 1.0602289377843577e-06, + "loss": 0.2577, + "step": 55050 + }, + { + "epoch": 3.8200680036083545, + "grad_norm": 1.0230151414871216, + "learning_rate": 1.0594130943583858e-06, + "loss": 0.3717, + "step": 55051 + }, + { + "epoch": 3.8201373950454514, + "grad_norm": 0.9228535294532776, + "learning_rate": 1.0585975632763757e-06, + "loss": 0.2794, + "step": 55052 + }, + { + "epoch": 3.820206786482548, + "grad_norm": 0.8099803924560547, + "learning_rate": 1.057782344540892e-06, + "loss": 0.2504, + "step": 55053 + }, + { + "epoch": 3.8202761779196446, + "grad_norm": 0.9379760026931763, + "learning_rate": 1.0569674381545213e-06, + "loss": 0.3506, + "step": 55054 + }, + { + "epoch": 3.8203455693567414, + "grad_norm": 0.7522284984588623, + "learning_rate": 1.0561528441198064e-06, + "loss": 0.2529, + "step": 55055 + }, + { + "epoch": 3.820414960793838, + "grad_norm": 0.8303512930870056, + "learning_rate": 1.0553385624393564e-06, + "loss": 0.2801, + "step": 55056 + }, + { + "epoch": 3.8204843522309346, + "grad_norm": 1.3995057344436646, + "learning_rate": 1.0545245931157134e-06, + "loss": 0.2564, + "step": 55057 + }, + { + "epoch": 3.8205537436680315, + "grad_norm": 1.2063703536987305, + "learning_rate": 1.053710936151453e-06, + "loss": 0.2891, + "step": 55058 + }, + { + "epoch": 3.8206231351051283, + "grad_norm": 0.9735943078994751, + "learning_rate": 1.0528975915491513e-06, + "loss": 0.2482, + "step": 55059 + }, + { + "epoch": 3.8206925265422247, + "grad_norm": 1.186025857925415, + "learning_rate": 1.0520845593113838e-06, + "loss": 0.2474, + "step": 55060 + }, + { + "epoch": 3.820761917979321, + "grad_norm": 0.880962610244751, + "learning_rate": 1.0512718394406928e-06, + "loss": 0.2242, + "step": 55061 + }, + { + "epoch": 3.820831309416418, + "grad_norm": 0.9694333076477051, + "learning_rate": 1.0504594319396544e-06, + "loss": 0.2024, + "step": 55062 + }, + { + "epoch": 3.8209007008535147, + "grad_norm": 1.026057481765747, + "learning_rate": 1.0496473368108327e-06, + "loss": 0.2881, + "step": 55063 + }, + { + "epoch": 3.8209700922906116, + "grad_norm": 0.8208848237991333, + "learning_rate": 1.048835554056804e-06, + "loss": 0.1756, + "step": 55064 + }, + { + "epoch": 3.821039483727708, + "grad_norm": 0.7706705331802368, + "learning_rate": 1.0480240836801215e-06, + "loss": 0.276, + "step": 55065 + }, + { + "epoch": 3.821108875164805, + "grad_norm": 0.938513457775116, + "learning_rate": 1.0472129256833384e-06, + "loss": 0.2559, + "step": 55066 + }, + { + "epoch": 3.821178266601901, + "grad_norm": 0.8348444700241089, + "learning_rate": 1.04640208006902e-06, + "loss": 0.1795, + "step": 55067 + }, + { + "epoch": 3.821247658038998, + "grad_norm": 0.8546695709228516, + "learning_rate": 1.0455915468397415e-06, + "loss": 0.239, + "step": 55068 + }, + { + "epoch": 3.821317049476095, + "grad_norm": 0.8843579888343811, + "learning_rate": 1.0447813259980344e-06, + "loss": 0.2455, + "step": 55069 + }, + { + "epoch": 3.8213864409131912, + "grad_norm": 0.8986918330192566, + "learning_rate": 1.0439714175464854e-06, + "loss": 0.3438, + "step": 55070 + }, + { + "epoch": 3.821455832350288, + "grad_norm": 0.8603711724281311, + "learning_rate": 1.043161821487637e-06, + "loss": 0.2101, + "step": 55071 + }, + { + "epoch": 3.8215252237873845, + "grad_norm": 0.7388714551925659, + "learning_rate": 1.0423525378240428e-06, + "loss": 0.1933, + "step": 55072 + }, + { + "epoch": 3.8215946152244813, + "grad_norm": 1.106198787689209, + "learning_rate": 1.0415435665582563e-06, + "loss": 0.2571, + "step": 55073 + }, + { + "epoch": 3.821664006661578, + "grad_norm": 0.8827593922615051, + "learning_rate": 1.040734907692842e-06, + "loss": 0.2241, + "step": 55074 + }, + { + "epoch": 3.8217333980986745, + "grad_norm": 0.8846341967582703, + "learning_rate": 1.0399265612303422e-06, + "loss": 0.2358, + "step": 55075 + }, + { + "epoch": 3.8218027895357713, + "grad_norm": 0.6652249693870544, + "learning_rate": 1.0391185271733107e-06, + "loss": 0.2295, + "step": 55076 + }, + { + "epoch": 3.8218721809728677, + "grad_norm": 1.0937398672103882, + "learning_rate": 1.0383108055243007e-06, + "loss": 0.2829, + "step": 55077 + }, + { + "epoch": 3.8219415724099646, + "grad_norm": 1.0172274112701416, + "learning_rate": 1.037503396285866e-06, + "loss": 0.2564, + "step": 55078 + }, + { + "epoch": 3.8220109638470614, + "grad_norm": 0.920428991317749, + "learning_rate": 1.036696299460549e-06, + "loss": 0.3178, + "step": 55079 + }, + { + "epoch": 3.822080355284158, + "grad_norm": 0.68839430809021, + "learning_rate": 1.035889515050903e-06, + "loss": 0.1631, + "step": 55080 + }, + { + "epoch": 3.8221497467212546, + "grad_norm": 0.6780847311019897, + "learning_rate": 1.0350830430594705e-06, + "loss": 0.2329, + "step": 55081 + }, + { + "epoch": 3.822219138158351, + "grad_norm": 0.8302220702171326, + "learning_rate": 1.034276883488794e-06, + "loss": 0.3075, + "step": 55082 + }, + { + "epoch": 3.822288529595448, + "grad_norm": 1.198939323425293, + "learning_rate": 1.0334710363414268e-06, + "loss": 0.2278, + "step": 55083 + }, + { + "epoch": 3.8223579210325447, + "grad_norm": 0.8295087814331055, + "learning_rate": 1.0326655016199116e-06, + "loss": 0.2591, + "step": 55084 + }, + { + "epoch": 3.8224273124696415, + "grad_norm": 0.7204179763793945, + "learning_rate": 1.0318602793267906e-06, + "loss": 0.1477, + "step": 55085 + }, + { + "epoch": 3.822496703906738, + "grad_norm": 1.490774393081665, + "learning_rate": 1.0310553694645952e-06, + "loss": 0.321, + "step": 55086 + }, + { + "epoch": 3.8225660953438347, + "grad_norm": 1.0751739740371704, + "learning_rate": 1.0302507720358789e-06, + "loss": 0.2146, + "step": 55087 + }, + { + "epoch": 3.822635486780931, + "grad_norm": 0.8033207654953003, + "learning_rate": 1.029446487043184e-06, + "loss": 0.2325, + "step": 55088 + }, + { + "epoch": 3.822704878218028, + "grad_norm": 0.6881071329116821, + "learning_rate": 1.0286425144890311e-06, + "loss": 0.1825, + "step": 55089 + }, + { + "epoch": 3.822774269655125, + "grad_norm": 0.8223838806152344, + "learning_rate": 1.0278388543759842e-06, + "loss": 0.1034, + "step": 55090 + }, + { + "epoch": 3.822843661092221, + "grad_norm": 0.7738524079322815, + "learning_rate": 1.027035506706564e-06, + "loss": 0.2752, + "step": 55091 + }, + { + "epoch": 3.822913052529318, + "grad_norm": 0.6514158844947815, + "learning_rate": 1.0262324714833127e-06, + "loss": 0.1609, + "step": 55092 + }, + { + "epoch": 3.8229824439664144, + "grad_norm": 1.1296594142913818, + "learning_rate": 1.0254297487087505e-06, + "loss": 0.2834, + "step": 55093 + }, + { + "epoch": 3.8230518354035112, + "grad_norm": 1.3467799425125122, + "learning_rate": 1.024627338385431e-06, + "loss": 0.3162, + "step": 55094 + }, + { + "epoch": 3.823121226840608, + "grad_norm": 1.138554573059082, + "learning_rate": 1.0238252405158855e-06, + "loss": 0.3099, + "step": 55095 + }, + { + "epoch": 3.8231906182777045, + "grad_norm": 0.7407802939414978, + "learning_rate": 1.0230234551026342e-06, + "loss": 0.231, + "step": 55096 + }, + { + "epoch": 3.8232600097148013, + "grad_norm": 1.1522274017333984, + "learning_rate": 1.0222219821482194e-06, + "loss": 0.2476, + "step": 55097 + }, + { + "epoch": 3.8233294011518977, + "grad_norm": 1.1999050378799438, + "learning_rate": 1.0214208216551613e-06, + "loss": 0.3726, + "step": 55098 + }, + { + "epoch": 3.8233987925889945, + "grad_norm": 0.6807160377502441, + "learning_rate": 1.0206199736259914e-06, + "loss": 0.1433, + "step": 55099 + }, + { + "epoch": 3.8234681840260913, + "grad_norm": 0.8034153580665588, + "learning_rate": 1.019819438063252e-06, + "loss": 0.1527, + "step": 55100 + }, + { + "epoch": 3.8235375754631877, + "grad_norm": 0.607103168964386, + "learning_rate": 1.019019214969441e-06, + "loss": 0.1697, + "step": 55101 + }, + { + "epoch": 3.8236069669002846, + "grad_norm": 0.6485708951950073, + "learning_rate": 1.0182193043471122e-06, + "loss": 0.1263, + "step": 55102 + }, + { + "epoch": 3.823676358337381, + "grad_norm": 1.0108379125595093, + "learning_rate": 1.0174197061987856e-06, + "loss": 0.3308, + "step": 55103 + }, + { + "epoch": 3.823745749774478, + "grad_norm": 1.0342456102371216, + "learning_rate": 1.0166204205269702e-06, + "loss": 0.2816, + "step": 55104 + }, + { + "epoch": 3.8238151412115746, + "grad_norm": 0.684905469417572, + "learning_rate": 1.0158214473342088e-06, + "loss": 0.2254, + "step": 55105 + }, + { + "epoch": 3.823884532648671, + "grad_norm": 0.6688528060913086, + "learning_rate": 1.01502278662301e-06, + "loss": 0.1329, + "step": 55106 + }, + { + "epoch": 3.823953924085768, + "grad_norm": 0.8483918905258179, + "learning_rate": 1.0142244383959054e-06, + "loss": 0.3047, + "step": 55107 + }, + { + "epoch": 3.8240233155228642, + "grad_norm": 0.7718603610992432, + "learning_rate": 1.0134264026554042e-06, + "loss": 0.255, + "step": 55108 + }, + { + "epoch": 3.824092706959961, + "grad_norm": 0.927733838558197, + "learning_rate": 1.0126286794040263e-06, + "loss": 0.3633, + "step": 55109 + }, + { + "epoch": 3.824162098397058, + "grad_norm": 0.7390912771224976, + "learning_rate": 1.0118312686443032e-06, + "loss": 0.1842, + "step": 55110 + }, + { + "epoch": 3.8242314898341547, + "grad_norm": 0.573838472366333, + "learning_rate": 1.011034170378744e-06, + "loss": 0.1692, + "step": 55111 + }, + { + "epoch": 3.824300881271251, + "grad_norm": 1.4193077087402344, + "learning_rate": 1.010237384609869e-06, + "loss": 0.4282, + "step": 55112 + }, + { + "epoch": 3.824370272708348, + "grad_norm": 1.1823017597198486, + "learning_rate": 1.009440911340187e-06, + "loss": 0.4198, + "step": 55113 + }, + { + "epoch": 3.8244396641454443, + "grad_norm": 0.7817990183830261, + "learning_rate": 1.0086447505722075e-06, + "loss": 0.2028, + "step": 55114 + }, + { + "epoch": 3.824509055582541, + "grad_norm": 1.0615689754486084, + "learning_rate": 1.0078489023084503e-06, + "loss": 0.3751, + "step": 55115 + }, + { + "epoch": 3.824578447019638, + "grad_norm": 0.8229068517684937, + "learning_rate": 1.007053366551436e-06, + "loss": 0.2862, + "step": 55116 + }, + { + "epoch": 3.8246478384567344, + "grad_norm": 0.9146926999092102, + "learning_rate": 1.0062581433036733e-06, + "loss": 0.2494, + "step": 55117 + }, + { + "epoch": 3.8247172298938312, + "grad_norm": 0.9276338219642639, + "learning_rate": 1.0054632325676605e-06, + "loss": 0.2755, + "step": 55118 + }, + { + "epoch": 3.8247866213309276, + "grad_norm": 0.8273259401321411, + "learning_rate": 1.0046686343459177e-06, + "loss": 0.2672, + "step": 55119 + }, + { + "epoch": 3.8248560127680244, + "grad_norm": 1.3407920598983765, + "learning_rate": 1.003874348640954e-06, + "loss": 0.2043, + "step": 55120 + }, + { + "epoch": 3.8249254042051213, + "grad_norm": 1.2338600158691406, + "learning_rate": 1.0030803754552675e-06, + "loss": 0.2634, + "step": 55121 + }, + { + "epoch": 3.8249947956422177, + "grad_norm": 0.729752779006958, + "learning_rate": 1.0022867147913672e-06, + "loss": 0.2027, + "step": 55122 + }, + { + "epoch": 3.8250641870793145, + "grad_norm": 0.8468868732452393, + "learning_rate": 1.0014933666517623e-06, + "loss": 0.3215, + "step": 55123 + }, + { + "epoch": 3.825133578516411, + "grad_norm": 0.8752301931381226, + "learning_rate": 1.0007003310389619e-06, + "loss": 0.2869, + "step": 55124 + }, + { + "epoch": 3.8252029699535077, + "grad_norm": 0.8685005307197571, + "learning_rate": 9.999076079554637e-07, + "loss": 0.3171, + "step": 55125 + }, + { + "epoch": 3.8252723613906046, + "grad_norm": 0.7783370614051819, + "learning_rate": 9.99115197403777e-07, + "loss": 0.2587, + "step": 55126 + }, + { + "epoch": 3.825341752827701, + "grad_norm": 1.0820719003677368, + "learning_rate": 9.98323099386389e-07, + "loss": 0.3491, + "step": 55127 + }, + { + "epoch": 3.825411144264798, + "grad_norm": 0.894416332244873, + "learning_rate": 9.975313139058085e-07, + "loss": 0.235, + "step": 55128 + }, + { + "epoch": 3.825480535701894, + "grad_norm": 0.819290041923523, + "learning_rate": 9.967398409645446e-07, + "loss": 0.2771, + "step": 55129 + }, + { + "epoch": 3.825549927138991, + "grad_norm": 1.8392513990402222, + "learning_rate": 9.95948680565084e-07, + "loss": 0.2058, + "step": 55130 + }, + { + "epoch": 3.825619318576088, + "grad_norm": 1.0123194456100464, + "learning_rate": 9.951578327099143e-07, + "loss": 0.2695, + "step": 55131 + }, + { + "epoch": 3.8256887100131842, + "grad_norm": 0.7973282933235168, + "learning_rate": 9.94367297401555e-07, + "loss": 0.1907, + "step": 55132 + }, + { + "epoch": 3.825758101450281, + "grad_norm": 0.8516255617141724, + "learning_rate": 9.935770746424933e-07, + "loss": 0.3567, + "step": 55133 + }, + { + "epoch": 3.8258274928873774, + "grad_norm": 0.7115416526794434, + "learning_rate": 9.92787164435216e-07, + "loss": 0.2089, + "step": 55134 + }, + { + "epoch": 3.8258968843244743, + "grad_norm": 1.4161432981491089, + "learning_rate": 9.919975667822322e-07, + "loss": 0.2633, + "step": 55135 + }, + { + "epoch": 3.825966275761571, + "grad_norm": 0.8984365463256836, + "learning_rate": 9.91208281686007e-07, + "loss": 0.2827, + "step": 55136 + }, + { + "epoch": 3.826035667198668, + "grad_norm": 1.0114686489105225, + "learning_rate": 9.9041930914906e-07, + "loss": 0.3013, + "step": 55137 + }, + { + "epoch": 3.8261050586357643, + "grad_norm": 0.7362402081489563, + "learning_rate": 9.896306491738782e-07, + "loss": 0.1626, + "step": 55138 + }, + { + "epoch": 3.826174450072861, + "grad_norm": 0.5486963987350464, + "learning_rate": 9.888423017629378e-07, + "loss": 0.1574, + "step": 55139 + }, + { + "epoch": 3.8262438415099576, + "grad_norm": 0.8535784482955933, + "learning_rate": 9.880542669187364e-07, + "loss": 0.3749, + "step": 55140 + }, + { + "epoch": 3.8263132329470544, + "grad_norm": 1.9539700746536255, + "learning_rate": 9.872665446437612e-07, + "loss": 0.2061, + "step": 55141 + }, + { + "epoch": 3.826382624384151, + "grad_norm": 0.8284995555877686, + "learning_rate": 9.86479134940499e-07, + "loss": 0.2498, + "step": 55142 + }, + { + "epoch": 3.8264520158212476, + "grad_norm": 1.0742894411087036, + "learning_rate": 9.856920378114253e-07, + "loss": 0.3232, + "step": 55143 + }, + { + "epoch": 3.8265214072583444, + "grad_norm": 1.0458145141601562, + "learning_rate": 9.849052532590498e-07, + "loss": 0.2884, + "step": 55144 + }, + { + "epoch": 3.826590798695441, + "grad_norm": 0.931516706943512, + "learning_rate": 9.841187812858254e-07, + "loss": 0.1684, + "step": 55145 + }, + { + "epoch": 3.8266601901325377, + "grad_norm": 0.9046058654785156, + "learning_rate": 9.833326218942618e-07, + "loss": 0.2009, + "step": 55146 + }, + { + "epoch": 3.8267295815696345, + "grad_norm": 0.9425764679908752, + "learning_rate": 9.825467750868233e-07, + "loss": 0.1936, + "step": 55147 + }, + { + "epoch": 3.826798973006731, + "grad_norm": 0.768010675907135, + "learning_rate": 9.81761240865997e-07, + "loss": 0.2295, + "step": 55148 + }, + { + "epoch": 3.8268683644438277, + "grad_norm": 1.0727332830429077, + "learning_rate": 9.809760192342588e-07, + "loss": 0.2417, + "step": 55149 + }, + { + "epoch": 3.826937755880924, + "grad_norm": 0.7274088859558105, + "learning_rate": 9.801911101940952e-07, + "loss": 0.2893, + "step": 55150 + }, + { + "epoch": 3.827007147318021, + "grad_norm": 0.775861382484436, + "learning_rate": 9.794065137479825e-07, + "loss": 0.1856, + "step": 55151 + }, + { + "epoch": 3.8270765387551178, + "grad_norm": 0.7181543707847595, + "learning_rate": 9.786222298983848e-07, + "loss": 0.1875, + "step": 55152 + }, + { + "epoch": 3.827145930192214, + "grad_norm": 0.8010556101799011, + "learning_rate": 9.778382586478007e-07, + "loss": 0.2875, + "step": 55153 + }, + { + "epoch": 3.827215321629311, + "grad_norm": 0.5947840213775635, + "learning_rate": 9.770545999986835e-07, + "loss": 0.1132, + "step": 55154 + }, + { + "epoch": 3.8272847130664074, + "grad_norm": 0.6779213547706604, + "learning_rate": 9.762712539535201e-07, + "loss": 0.1745, + "step": 55155 + }, + { + "epoch": 3.827354104503504, + "grad_norm": 0.9090043306350708, + "learning_rate": 9.754882205147752e-07, + "loss": 0.3457, + "step": 55156 + }, + { + "epoch": 3.827423495940601, + "grad_norm": 0.982402503490448, + "learning_rate": 9.74705499684936e-07, + "loss": 0.2204, + "step": 55157 + }, + { + "epoch": 3.8274928873776974, + "grad_norm": 0.9152733087539673, + "learning_rate": 9.739230914664444e-07, + "loss": 0.1862, + "step": 55158 + }, + { + "epoch": 3.8275622788147943, + "grad_norm": 0.6208578944206238, + "learning_rate": 9.731409958617986e-07, + "loss": 0.1195, + "step": 55159 + }, + { + "epoch": 3.8276316702518907, + "grad_norm": 0.7319139242172241, + "learning_rate": 9.723592128734638e-07, + "loss": 0.1947, + "step": 55160 + }, + { + "epoch": 3.8277010616889875, + "grad_norm": 0.7381781339645386, + "learning_rate": 9.71577742503893e-07, + "loss": 0.1956, + "step": 55161 + }, + { + "epoch": 3.8277704531260843, + "grad_norm": 0.9311976432800293, + "learning_rate": 9.707965847555733e-07, + "loss": 0.2234, + "step": 55162 + }, + { + "epoch": 3.827839844563181, + "grad_norm": 0.930316150188446, + "learning_rate": 9.700157396309473e-07, + "loss": 0.3208, + "step": 55163 + }, + { + "epoch": 3.8279092360002775, + "grad_norm": 1.0417704582214355, + "learning_rate": 9.692352071324907e-07, + "loss": 0.2997, + "step": 55164 + }, + { + "epoch": 3.8279786274373744, + "grad_norm": 0.8641461730003357, + "learning_rate": 9.684549872626791e-07, + "loss": 0.2987, + "step": 55165 + }, + { + "epoch": 3.8280480188744708, + "grad_norm": 0.7624304294586182, + "learning_rate": 9.676750800239555e-07, + "loss": 0.1987, + "step": 55166 + }, + { + "epoch": 3.8281174103115676, + "grad_norm": 1.1376622915267944, + "learning_rate": 9.66895485418795e-07, + "loss": 0.2435, + "step": 55167 + }, + { + "epoch": 3.8281868017486644, + "grad_norm": 0.6841593980789185, + "learning_rate": 9.661162034496629e-07, + "loss": 0.2002, + "step": 55168 + }, + { + "epoch": 3.828256193185761, + "grad_norm": 1.0661531686782837, + "learning_rate": 9.653372341190014e-07, + "loss": 0.1875, + "step": 55169 + }, + { + "epoch": 3.8283255846228577, + "grad_norm": 0.6302365064620972, + "learning_rate": 9.645585774292865e-07, + "loss": 0.1879, + "step": 55170 + }, + { + "epoch": 3.828394976059954, + "grad_norm": 0.9169801473617554, + "learning_rate": 9.637802333829715e-07, + "loss": 0.2517, + "step": 55171 + }, + { + "epoch": 3.828464367497051, + "grad_norm": 2.0707244873046875, + "learning_rate": 9.630022019824991e-07, + "loss": 0.2804, + "step": 55172 + }, + { + "epoch": 3.8285337589341477, + "grad_norm": 0.99476557970047, + "learning_rate": 9.622244832303452e-07, + "loss": 0.3524, + "step": 55173 + }, + { + "epoch": 3.828603150371244, + "grad_norm": 0.8683261871337891, + "learning_rate": 9.614470771289741e-07, + "loss": 0.1925, + "step": 55174 + }, + { + "epoch": 3.828672541808341, + "grad_norm": 0.7972335815429688, + "learning_rate": 9.606699836808064e-07, + "loss": 0.2099, + "step": 55175 + }, + { + "epoch": 3.8287419332454373, + "grad_norm": 0.9956696033477783, + "learning_rate": 9.598932028883178e-07, + "loss": 0.1522, + "step": 55176 + }, + { + "epoch": 3.828811324682534, + "grad_norm": 0.832685112953186, + "learning_rate": 9.59116734753951e-07, + "loss": 0.215, + "step": 55177 + }, + { + "epoch": 3.828880716119631, + "grad_norm": 0.9406111240386963, + "learning_rate": 9.583405792801592e-07, + "loss": 0.3567, + "step": 55178 + }, + { + "epoch": 3.8289501075567274, + "grad_norm": 0.7198193669319153, + "learning_rate": 9.575647364694074e-07, + "loss": 0.1741, + "step": 55179 + }, + { + "epoch": 3.829019498993824, + "grad_norm": 0.8006766438484192, + "learning_rate": 9.567892063241158e-07, + "loss": 0.2847, + "step": 55180 + }, + { + "epoch": 3.8290888904309206, + "grad_norm": 0.6852611899375916, + "learning_rate": 9.560139888467712e-07, + "loss": 0.2192, + "step": 55181 + }, + { + "epoch": 3.8291582818680174, + "grad_norm": 1.5206063985824585, + "learning_rate": 9.55239084039783e-07, + "loss": 0.3996, + "step": 55182 + }, + { + "epoch": 3.8292276733051143, + "grad_norm": 0.6824431419372559, + "learning_rate": 9.544644919056266e-07, + "loss": 0.1974, + "step": 55183 + }, + { + "epoch": 3.8292970647422107, + "grad_norm": 1.237334966659546, + "learning_rate": 9.536902124467228e-07, + "loss": 0.3343, + "step": 55184 + }, + { + "epoch": 3.8293664561793075, + "grad_norm": 1.1316927671432495, + "learning_rate": 9.52916245665525e-07, + "loss": 0.3079, + "step": 55185 + }, + { + "epoch": 3.829435847616404, + "grad_norm": 0.8420085310935974, + "learning_rate": 9.521425915644866e-07, + "loss": 0.2462, + "step": 55186 + }, + { + "epoch": 3.8295052390535007, + "grad_norm": 0.5543677806854248, + "learning_rate": 9.513692501460391e-07, + "loss": 0.0913, + "step": 55187 + }, + { + "epoch": 3.8295746304905975, + "grad_norm": 0.7962295413017273, + "learning_rate": 9.505962214126362e-07, + "loss": 0.2292, + "step": 55188 + }, + { + "epoch": 3.8296440219276944, + "grad_norm": 0.8313288688659668, + "learning_rate": 9.498235053666982e-07, + "loss": 0.2069, + "step": 55189 + }, + { + "epoch": 3.8297134133647908, + "grad_norm": 1.1484143733978271, + "learning_rate": 9.490511020106785e-07, + "loss": 0.2865, + "step": 55190 + }, + { + "epoch": 3.8297828048018876, + "grad_norm": 0.643709659576416, + "learning_rate": 9.482790113470086e-07, + "loss": 0.183, + "step": 55191 + }, + { + "epoch": 3.829852196238984, + "grad_norm": 0.9425880312919617, + "learning_rate": 9.475072333781421e-07, + "loss": 0.295, + "step": 55192 + }, + { + "epoch": 3.829921587676081, + "grad_norm": 0.8096069693565369, + "learning_rate": 9.467357681064882e-07, + "loss": 0.2004, + "step": 55193 + }, + { + "epoch": 3.8299909791131777, + "grad_norm": 1.2495760917663574, + "learning_rate": 9.459646155345003e-07, + "loss": 0.3318, + "step": 55194 + }, + { + "epoch": 3.830060370550274, + "grad_norm": 0.8359748721122742, + "learning_rate": 9.451937756646212e-07, + "loss": 0.1915, + "step": 55195 + }, + { + "epoch": 3.830129761987371, + "grad_norm": 0.8477252125740051, + "learning_rate": 9.44423248499271e-07, + "loss": 0.1998, + "step": 55196 + }, + { + "epoch": 3.8301991534244673, + "grad_norm": 0.6348325610160828, + "learning_rate": 9.4365303404087e-07, + "loss": 0.1911, + "step": 55197 + }, + { + "epoch": 3.830268544861564, + "grad_norm": 0.7102113962173462, + "learning_rate": 9.428831322918719e-07, + "loss": 0.1975, + "step": 55198 + }, + { + "epoch": 3.830337936298661, + "grad_norm": 0.9117962121963501, + "learning_rate": 9.421135432547079e-07, + "loss": 0.3467, + "step": 55199 + }, + { + "epoch": 3.8304073277357573, + "grad_norm": 1.335379958152771, + "learning_rate": 9.413442669317873e-07, + "loss": 0.3313, + "step": 55200 + }, + { + "epoch": 3.830476719172854, + "grad_norm": 1.1669096946716309, + "learning_rate": 9.405753033255527e-07, + "loss": 0.2588, + "step": 55201 + }, + { + "epoch": 3.8305461106099505, + "grad_norm": 0.6771953701972961, + "learning_rate": 9.398066524384241e-07, + "loss": 0.1956, + "step": 55202 + }, + { + "epoch": 3.8306155020470474, + "grad_norm": 1.0048694610595703, + "learning_rate": 9.390383142728442e-07, + "loss": 0.333, + "step": 55203 + }, + { + "epoch": 3.830684893484144, + "grad_norm": 0.9661850929260254, + "learning_rate": 9.382702888312222e-07, + "loss": 0.2552, + "step": 55204 + }, + { + "epoch": 3.8307542849212406, + "grad_norm": 1.099138855934143, + "learning_rate": 9.375025761159783e-07, + "loss": 0.2702, + "step": 55205 + }, + { + "epoch": 3.8308236763583374, + "grad_norm": 0.8017958402633667, + "learning_rate": 9.36735176129555e-07, + "loss": 0.1676, + "step": 55206 + }, + { + "epoch": 3.830893067795434, + "grad_norm": 0.9463434219360352, + "learning_rate": 9.359680888743616e-07, + "loss": 0.2075, + "step": 55207 + }, + { + "epoch": 3.8309624592325306, + "grad_norm": 0.9687593579292297, + "learning_rate": 9.352013143528293e-07, + "loss": 0.1111, + "step": 55208 + }, + { + "epoch": 3.8310318506696275, + "grad_norm": 0.8155796527862549, + "learning_rate": 9.344348525673675e-07, + "loss": 0.2432, + "step": 55209 + }, + { + "epoch": 3.831101242106724, + "grad_norm": 0.9663738012313843, + "learning_rate": 9.336687035204072e-07, + "loss": 0.2075, + "step": 55210 + }, + { + "epoch": 3.8311706335438207, + "grad_norm": 0.6372959017753601, + "learning_rate": 9.329028672143469e-07, + "loss": 0.1715, + "step": 55211 + }, + { + "epoch": 3.831240024980917, + "grad_norm": 1.109689712524414, + "learning_rate": 9.32137343651629e-07, + "loss": 0.3001, + "step": 55212 + }, + { + "epoch": 3.831309416418014, + "grad_norm": 0.9763675928115845, + "learning_rate": 9.313721328346625e-07, + "loss": 0.1999, + "step": 55213 + }, + { + "epoch": 3.8313788078551108, + "grad_norm": 0.9930465817451477, + "learning_rate": 9.306072347658457e-07, + "loss": 0.3335, + "step": 55214 + }, + { + "epoch": 3.8314481992922076, + "grad_norm": 0.7484434843063354, + "learning_rate": 9.298426494476209e-07, + "loss": 0.1876, + "step": 55215 + }, + { + "epoch": 3.831517590729304, + "grad_norm": 0.6847000122070312, + "learning_rate": 9.290783768823863e-07, + "loss": 0.1495, + "step": 55216 + }, + { + "epoch": 3.831586982166401, + "grad_norm": 1.126311182975769, + "learning_rate": 9.283144170725622e-07, + "loss": 0.3924, + "step": 55217 + }, + { + "epoch": 3.831656373603497, + "grad_norm": 1.047763705253601, + "learning_rate": 9.275507700205465e-07, + "loss": 0.3614, + "step": 55218 + }, + { + "epoch": 3.831725765040594, + "grad_norm": 1.1718571186065674, + "learning_rate": 9.267874357287598e-07, + "loss": 0.3528, + "step": 55219 + }, + { + "epoch": 3.831795156477691, + "grad_norm": 0.9831006526947021, + "learning_rate": 9.260244141996111e-07, + "loss": 0.329, + "step": 55220 + }, + { + "epoch": 3.8318645479147873, + "grad_norm": 0.9098299145698547, + "learning_rate": 9.252617054354983e-07, + "loss": 0.3285, + "step": 55221 + }, + { + "epoch": 3.831933939351884, + "grad_norm": 1.250685453414917, + "learning_rate": 9.244993094388532e-07, + "loss": 0.33, + "step": 55222 + }, + { + "epoch": 3.8320033307889805, + "grad_norm": 0.5828015804290771, + "learning_rate": 9.237372262120736e-07, + "loss": 0.1954, + "step": 55223 + }, + { + "epoch": 3.8320727222260773, + "grad_norm": 0.8120279312133789, + "learning_rate": 9.229754557575466e-07, + "loss": 0.1822, + "step": 55224 + }, + { + "epoch": 3.832142113663174, + "grad_norm": 1.0534268617630005, + "learning_rate": 9.222139980777034e-07, + "loss": 0.3305, + "step": 55225 + }, + { + "epoch": 3.8322115051002705, + "grad_norm": 1.005446434020996, + "learning_rate": 9.2145285317492e-07, + "loss": 0.2609, + "step": 55226 + }, + { + "epoch": 3.8322808965373674, + "grad_norm": 0.7080990076065063, + "learning_rate": 9.206920210516279e-07, + "loss": 0.1337, + "step": 55227 + }, + { + "epoch": 3.8323502879744638, + "grad_norm": 0.7020970582962036, + "learning_rate": 9.199315017102139e-07, + "loss": 0.1724, + "step": 55228 + }, + { + "epoch": 3.8324196794115606, + "grad_norm": 0.7664270401000977, + "learning_rate": 9.191712951530762e-07, + "loss": 0.2979, + "step": 55229 + }, + { + "epoch": 3.8324890708486574, + "grad_norm": 0.9819355010986328, + "learning_rate": 9.18411401382624e-07, + "loss": 0.1994, + "step": 55230 + }, + { + "epoch": 3.832558462285754, + "grad_norm": 1.3425874710083008, + "learning_rate": 9.176518204012441e-07, + "loss": 0.2598, + "step": 55231 + }, + { + "epoch": 3.8326278537228506, + "grad_norm": 0.9268794059753418, + "learning_rate": 9.168925522113459e-07, + "loss": 0.205, + "step": 55232 + }, + { + "epoch": 3.832697245159947, + "grad_norm": 0.8400740623474121, + "learning_rate": 9.161335968153272e-07, + "loss": 0.3427, + "step": 55233 + }, + { + "epoch": 3.832766636597044, + "grad_norm": 0.7053459286689758, + "learning_rate": 9.153749542155754e-07, + "loss": 0.2731, + "step": 55234 + }, + { + "epoch": 3.8328360280341407, + "grad_norm": 0.6474102139472961, + "learning_rate": 9.146166244144882e-07, + "loss": 0.2736, + "step": 55235 + }, + { + "epoch": 3.832905419471237, + "grad_norm": 0.8733741044998169, + "learning_rate": 9.138586074144528e-07, + "loss": 0.2462, + "step": 55236 + }, + { + "epoch": 3.832974810908334, + "grad_norm": 0.6539693474769592, + "learning_rate": 9.131009032178784e-07, + "loss": 0.1892, + "step": 55237 + }, + { + "epoch": 3.8330442023454303, + "grad_norm": 0.6072558164596558, + "learning_rate": 9.123435118271517e-07, + "loss": 0.1408, + "step": 55238 + }, + { + "epoch": 3.833113593782527, + "grad_norm": 0.8537411689758301, + "learning_rate": 9.1158643324466e-07, + "loss": 0.2182, + "step": 55239 + }, + { + "epoch": 3.833182985219624, + "grad_norm": 0.7827514410018921, + "learning_rate": 9.1082966747279e-07, + "loss": 0.286, + "step": 55240 + }, + { + "epoch": 3.833252376656721, + "grad_norm": 1.1282477378845215, + "learning_rate": 9.10073214513929e-07, + "loss": 0.249, + "step": 55241 + }, + { + "epoch": 3.833321768093817, + "grad_norm": 0.813630998134613, + "learning_rate": 9.093170743704749e-07, + "loss": 0.194, + "step": 55242 + }, + { + "epoch": 3.833391159530914, + "grad_norm": 0.7696774005889893, + "learning_rate": 9.085612470448146e-07, + "loss": 0.2493, + "step": 55243 + }, + { + "epoch": 3.8334605509680104, + "grad_norm": 1.5041996240615845, + "learning_rate": 9.078057325393241e-07, + "loss": 0.2483, + "step": 55244 + }, + { + "epoch": 3.8335299424051072, + "grad_norm": 1.0211889743804932, + "learning_rate": 9.070505308563903e-07, + "loss": 0.2337, + "step": 55245 + }, + { + "epoch": 3.833599333842204, + "grad_norm": 1.157923698425293, + "learning_rate": 9.062956419984004e-07, + "loss": 0.3134, + "step": 55246 + }, + { + "epoch": 3.8336687252793005, + "grad_norm": 0.8601852655410767, + "learning_rate": 9.05541065967741e-07, + "loss": 0.2881, + "step": 55247 + }, + { + "epoch": 3.8337381167163973, + "grad_norm": 0.9478349089622498, + "learning_rate": 9.047868027667994e-07, + "loss": 0.262, + "step": 55248 + }, + { + "epoch": 3.8338075081534937, + "grad_norm": 0.7235567569732666, + "learning_rate": 9.040328523979402e-07, + "loss": 0.1813, + "step": 55249 + }, + { + "epoch": 3.8338768995905905, + "grad_norm": 0.8576454520225525, + "learning_rate": 9.032792148635616e-07, + "loss": 0.2627, + "step": 55250 + }, + { + "epoch": 3.8339462910276874, + "grad_norm": 0.8088516592979431, + "learning_rate": 9.025258901660282e-07, + "loss": 0.2356, + "step": 55251 + }, + { + "epoch": 3.8340156824647837, + "grad_norm": 0.9903583526611328, + "learning_rate": 9.017728783077161e-07, + "loss": 0.2721, + "step": 55252 + }, + { + "epoch": 3.8340850739018806, + "grad_norm": 0.855217456817627, + "learning_rate": 9.010201792910122e-07, + "loss": 0.2747, + "step": 55253 + }, + { + "epoch": 3.834154465338977, + "grad_norm": 1.3312695026397705, + "learning_rate": 9.002677931182923e-07, + "loss": 0.3157, + "step": 55254 + }, + { + "epoch": 3.834223856776074, + "grad_norm": 0.742875874042511, + "learning_rate": 8.995157197919324e-07, + "loss": 0.2008, + "step": 55255 + }, + { + "epoch": 3.8342932482131706, + "grad_norm": 0.8490808606147766, + "learning_rate": 8.987639593142972e-07, + "loss": 0.2388, + "step": 55256 + }, + { + "epoch": 3.834362639650267, + "grad_norm": 0.9417130351066589, + "learning_rate": 8.980125116877736e-07, + "loss": 0.2549, + "step": 55257 + }, + { + "epoch": 3.834432031087364, + "grad_norm": 0.7331492304801941, + "learning_rate": 8.972613769147265e-07, + "loss": 0.1745, + "step": 55258 + }, + { + "epoch": 3.8345014225244602, + "grad_norm": 0.7443121075630188, + "learning_rate": 8.965105549975206e-07, + "loss": 0.2343, + "step": 55259 + }, + { + "epoch": 3.834570813961557, + "grad_norm": 0.9502336382865906, + "learning_rate": 8.957600459385317e-07, + "loss": 0.1797, + "step": 55260 + }, + { + "epoch": 3.834640205398654, + "grad_norm": 0.8734113574028015, + "learning_rate": 8.950098497401361e-07, + "loss": 0.2313, + "step": 55261 + }, + { + "epoch": 3.8347095968357507, + "grad_norm": 1.0041260719299316, + "learning_rate": 8.942599664046869e-07, + "loss": 0.2889, + "step": 55262 + }, + { + "epoch": 3.834778988272847, + "grad_norm": 0.8468008637428284, + "learning_rate": 8.935103959345714e-07, + "loss": 0.2567, + "step": 55263 + }, + { + "epoch": 3.8348483797099435, + "grad_norm": 0.8634976148605347, + "learning_rate": 8.927611383321432e-07, + "loss": 0.2701, + "step": 55264 + }, + { + "epoch": 3.8349177711470404, + "grad_norm": 0.8002203702926636, + "learning_rate": 8.920121935997672e-07, + "loss": 0.2613, + "step": 55265 + }, + { + "epoch": 3.834987162584137, + "grad_norm": 1.091971516609192, + "learning_rate": 8.91263561739808e-07, + "loss": 0.349, + "step": 55266 + }, + { + "epoch": 3.835056554021234, + "grad_norm": 0.8233826160430908, + "learning_rate": 8.905152427546303e-07, + "loss": 0.1941, + "step": 55267 + }, + { + "epoch": 3.8351259454583304, + "grad_norm": 0.8862621188163757, + "learning_rate": 8.897672366465992e-07, + "loss": 0.2623, + "step": 55268 + }, + { + "epoch": 3.8351953368954272, + "grad_norm": 1.413594365119934, + "learning_rate": 8.890195434180792e-07, + "loss": 0.2535, + "step": 55269 + }, + { + "epoch": 3.8352647283325236, + "grad_norm": 0.8041036128997803, + "learning_rate": 8.882721630714241e-07, + "loss": 0.2469, + "step": 55270 + }, + { + "epoch": 3.8353341197696205, + "grad_norm": 1.2195998430252075, + "learning_rate": 8.875250956089875e-07, + "loss": 0.2505, + "step": 55271 + }, + { + "epoch": 3.8354035112067173, + "grad_norm": 0.6500198245048523, + "learning_rate": 8.867783410331454e-07, + "loss": 0.2328, + "step": 55272 + }, + { + "epoch": 3.8354729026438137, + "grad_norm": 1.2391045093536377, + "learning_rate": 8.860318993462402e-07, + "loss": 0.2567, + "step": 55273 + }, + { + "epoch": 3.8355422940809105, + "grad_norm": 1.3377420902252197, + "learning_rate": 8.852857705506368e-07, + "loss": 0.3232, + "step": 55274 + }, + { + "epoch": 3.835611685518007, + "grad_norm": 0.6710251569747925, + "learning_rate": 8.845399546486777e-07, + "loss": 0.1721, + "step": 55275 + }, + { + "epoch": 3.8356810769551037, + "grad_norm": 0.8157769441604614, + "learning_rate": 8.837944516427276e-07, + "loss": 0.3243, + "step": 55276 + }, + { + "epoch": 3.8357504683922006, + "grad_norm": 0.9581127166748047, + "learning_rate": 8.830492615351404e-07, + "loss": 0.3052, + "step": 55277 + }, + { + "epoch": 3.835819859829297, + "grad_norm": 1.50115966796875, + "learning_rate": 8.823043843282697e-07, + "loss": 0.2853, + "step": 55278 + }, + { + "epoch": 3.835889251266394, + "grad_norm": 1.007311463356018, + "learning_rate": 8.81559820024469e-07, + "loss": 0.1759, + "step": 55279 + }, + { + "epoch": 3.83595864270349, + "grad_norm": 1.1599225997924805, + "learning_rate": 8.808155686260699e-07, + "loss": 0.2828, + "step": 55280 + }, + { + "epoch": 3.836028034140587, + "grad_norm": 0.6175311803817749, + "learning_rate": 8.800716301354484e-07, + "loss": 0.1951, + "step": 55281 + }, + { + "epoch": 3.836097425577684, + "grad_norm": 0.843923807144165, + "learning_rate": 8.793280045549357e-07, + "loss": 0.2071, + "step": 55282 + }, + { + "epoch": 3.8361668170147802, + "grad_norm": 1.2889182567596436, + "learning_rate": 8.785846918868856e-07, + "loss": 0.1854, + "step": 55283 + }, + { + "epoch": 3.836236208451877, + "grad_norm": 1.0990583896636963, + "learning_rate": 8.778416921336408e-07, + "loss": 0.2485, + "step": 55284 + }, + { + "epoch": 3.8363055998889735, + "grad_norm": 1.0273877382278442, + "learning_rate": 8.770990052975548e-07, + "loss": 0.1811, + "step": 55285 + }, + { + "epoch": 3.8363749913260703, + "grad_norm": 0.9209500551223755, + "learning_rate": 8.76356631380959e-07, + "loss": 0.368, + "step": 55286 + }, + { + "epoch": 3.836444382763167, + "grad_norm": 0.9021344780921936, + "learning_rate": 8.756145703862073e-07, + "loss": 0.3242, + "step": 55287 + }, + { + "epoch": 3.836513774200264, + "grad_norm": 1.3455930948257446, + "learning_rate": 8.748728223156421e-07, + "loss": 0.3142, + "step": 55288 + }, + { + "epoch": 3.8365831656373603, + "grad_norm": 0.9122578501701355, + "learning_rate": 8.741313871716062e-07, + "loss": 0.2513, + "step": 55289 + }, + { + "epoch": 3.8366525570744567, + "grad_norm": 0.8501376509666443, + "learning_rate": 8.733902649564308e-07, + "loss": 0.2289, + "step": 55290 + }, + { + "epoch": 3.8367219485115536, + "grad_norm": 0.8115879893302917, + "learning_rate": 8.726494556724585e-07, + "loss": 0.217, + "step": 55291 + }, + { + "epoch": 3.8367913399486504, + "grad_norm": 1.2352532148361206, + "learning_rate": 8.71908959322043e-07, + "loss": 0.3578, + "step": 55292 + }, + { + "epoch": 3.8368607313857472, + "grad_norm": 0.8930625915527344, + "learning_rate": 8.711687759075049e-07, + "loss": 0.2503, + "step": 55293 + }, + { + "epoch": 3.8369301228228436, + "grad_norm": 0.7922177314758301, + "learning_rate": 8.704289054311865e-07, + "loss": 0.2394, + "step": 55294 + }, + { + "epoch": 3.8369995142599405, + "grad_norm": 1.277306079864502, + "learning_rate": 8.696893478954193e-07, + "loss": 0.247, + "step": 55295 + }, + { + "epoch": 3.837068905697037, + "grad_norm": 1.123133659362793, + "learning_rate": 8.689501033025349e-07, + "loss": 0.2046, + "step": 55296 + }, + { + "epoch": 3.8371382971341337, + "grad_norm": 1.0988988876342773, + "learning_rate": 8.682111716548869e-07, + "loss": 0.2707, + "step": 55297 + }, + { + "epoch": 3.8372076885712305, + "grad_norm": 0.6204537749290466, + "learning_rate": 8.674725529547844e-07, + "loss": 0.1835, + "step": 55298 + }, + { + "epoch": 3.837277080008327, + "grad_norm": 1.5221645832061768, + "learning_rate": 8.667342472045814e-07, + "loss": 0.1713, + "step": 55299 + }, + { + "epoch": 3.8373464714454237, + "grad_norm": 1.075589656829834, + "learning_rate": 8.65996254406587e-07, + "loss": 0.3104, + "step": 55300 + }, + { + "epoch": 3.83741586288252, + "grad_norm": 1.0690016746520996, + "learning_rate": 8.652585745631437e-07, + "loss": 0.2222, + "step": 55301 + }, + { + "epoch": 3.837485254319617, + "grad_norm": 1.020287036895752, + "learning_rate": 8.645212076765829e-07, + "loss": 0.3256, + "step": 55302 + }, + { + "epoch": 3.837554645756714, + "grad_norm": 0.8358242511749268, + "learning_rate": 8.637841537492142e-07, + "loss": 0.2482, + "step": 55303 + }, + { + "epoch": 3.83762403719381, + "grad_norm": 0.8751879334449768, + "learning_rate": 8.63047412783391e-07, + "loss": 0.2486, + "step": 55304 + }, + { + "epoch": 3.837693428630907, + "grad_norm": 0.911849319934845, + "learning_rate": 8.623109847814225e-07, + "loss": 0.1985, + "step": 55305 + }, + { + "epoch": 3.8377628200680034, + "grad_norm": 0.7921527028083801, + "learning_rate": 8.615748697456294e-07, + "loss": 0.193, + "step": 55306 + }, + { + "epoch": 3.8378322115051002, + "grad_norm": 0.7550333142280579, + "learning_rate": 8.608390676783429e-07, + "loss": 0.255, + "step": 55307 + }, + { + "epoch": 3.837901602942197, + "grad_norm": 1.0518879890441895, + "learning_rate": 8.601035785818945e-07, + "loss": 0.1547, + "step": 55308 + }, + { + "epoch": 3.8379709943792935, + "grad_norm": 0.6343330144882202, + "learning_rate": 8.593684024585824e-07, + "loss": 0.2592, + "step": 55309 + }, + { + "epoch": 3.8380403858163903, + "grad_norm": 0.9901443123817444, + "learning_rate": 8.586335393107492e-07, + "loss": 0.2174, + "step": 55310 + }, + { + "epoch": 3.8381097772534867, + "grad_norm": 0.9801491498947144, + "learning_rate": 8.578989891407041e-07, + "loss": 0.2511, + "step": 55311 + }, + { + "epoch": 3.8381791686905835, + "grad_norm": 0.9427891373634338, + "learning_rate": 8.571647519507786e-07, + "loss": 0.269, + "step": 55312 + }, + { + "epoch": 3.8382485601276803, + "grad_norm": 0.8441314101219177, + "learning_rate": 8.564308277432709e-07, + "loss": 0.4172, + "step": 55313 + }, + { + "epoch": 3.838317951564777, + "grad_norm": 0.8977153897285461, + "learning_rate": 8.556972165205124e-07, + "loss": 0.2462, + "step": 55314 + }, + { + "epoch": 3.8383873430018736, + "grad_norm": 0.6440163254737854, + "learning_rate": 8.549639182848123e-07, + "loss": 0.2339, + "step": 55315 + }, + { + "epoch": 3.8384567344389704, + "grad_norm": 0.7588403820991516, + "learning_rate": 8.542309330384912e-07, + "loss": 0.2677, + "step": 55316 + }, + { + "epoch": 3.838526125876067, + "grad_norm": 0.8761698603630066, + "learning_rate": 8.534982607838582e-07, + "loss": 0.1893, + "step": 55317 + }, + { + "epoch": 3.8385955173131636, + "grad_norm": 0.7163114547729492, + "learning_rate": 8.527659015232336e-07, + "loss": 0.2276, + "step": 55318 + }, + { + "epoch": 3.8386649087502605, + "grad_norm": 1.2814451456069946, + "learning_rate": 8.520338552589269e-07, + "loss": 0.4715, + "step": 55319 + }, + { + "epoch": 3.838734300187357, + "grad_norm": 1.0292037725448608, + "learning_rate": 8.51302121993236e-07, + "loss": 0.1991, + "step": 55320 + }, + { + "epoch": 3.8388036916244537, + "grad_norm": 1.2573249340057373, + "learning_rate": 8.505707017284814e-07, + "loss": 0.3073, + "step": 55321 + }, + { + "epoch": 3.83887308306155, + "grad_norm": 0.8952530026435852, + "learning_rate": 8.498395944669724e-07, + "loss": 0.2162, + "step": 55322 + }, + { + "epoch": 3.838942474498647, + "grad_norm": 0.9735583662986755, + "learning_rate": 8.491088002110181e-07, + "loss": 0.2754, + "step": 55323 + }, + { + "epoch": 3.8390118659357437, + "grad_norm": 1.13045334815979, + "learning_rate": 8.483783189629168e-07, + "loss": 0.3098, + "step": 55324 + }, + { + "epoch": 3.83908125737284, + "grad_norm": 0.708659827709198, + "learning_rate": 8.47648150724989e-07, + "loss": 0.1178, + "step": 55325 + }, + { + "epoch": 3.839150648809937, + "grad_norm": 0.7972570061683655, + "learning_rate": 8.469182954995325e-07, + "loss": 0.21, + "step": 55326 + }, + { + "epoch": 3.8392200402470333, + "grad_norm": 0.9250450134277344, + "learning_rate": 8.461887532888568e-07, + "loss": 0.2733, + "step": 55327 + }, + { + "epoch": 3.83928943168413, + "grad_norm": 0.7303457856178284, + "learning_rate": 8.45459524095249e-07, + "loss": 0.2016, + "step": 55328 + }, + { + "epoch": 3.839358823121227, + "grad_norm": 0.7100400924682617, + "learning_rate": 8.447306079210182e-07, + "loss": 0.1353, + "step": 55329 + }, + { + "epoch": 3.8394282145583234, + "grad_norm": 1.0505925416946411, + "learning_rate": 8.440020047684738e-07, + "loss": 0.2374, + "step": 55330 + }, + { + "epoch": 3.8394976059954202, + "grad_norm": 0.7567861080169678, + "learning_rate": 8.432737146399139e-07, + "loss": 0.2075, + "step": 55331 + }, + { + "epoch": 3.8395669974325166, + "grad_norm": 0.5567798614501953, + "learning_rate": 8.425457375376367e-07, + "loss": 0.1975, + "step": 55332 + }, + { + "epoch": 3.8396363888696134, + "grad_norm": 1.0080854892730713, + "learning_rate": 8.418180734639402e-07, + "loss": 0.2604, + "step": 55333 + }, + { + "epoch": 3.8397057803067103, + "grad_norm": 2.8159983158111572, + "learning_rate": 8.410907224211117e-07, + "loss": 0.4062, + "step": 55334 + }, + { + "epoch": 3.8397751717438067, + "grad_norm": 1.0136499404907227, + "learning_rate": 8.403636844114493e-07, + "loss": 0.3098, + "step": 55335 + }, + { + "epoch": 3.8398445631809035, + "grad_norm": 0.9680507779121399, + "learning_rate": 8.396369594372621e-07, + "loss": 0.1726, + "step": 55336 + }, + { + "epoch": 3.839913954618, + "grad_norm": 0.6727930307388306, + "learning_rate": 8.389105475008374e-07, + "loss": 0.1823, + "step": 55337 + }, + { + "epoch": 3.8399833460550967, + "grad_norm": 1.1373353004455566, + "learning_rate": 8.381844486044621e-07, + "loss": 0.2632, + "step": 55338 + }, + { + "epoch": 3.8400527374921936, + "grad_norm": 0.6987746357917786, + "learning_rate": 8.374586627504455e-07, + "loss": 0.1322, + "step": 55339 + }, + { + "epoch": 3.8401221289292904, + "grad_norm": 1.0029891729354858, + "learning_rate": 8.367331899410524e-07, + "loss": 0.3143, + "step": 55340 + }, + { + "epoch": 3.840191520366387, + "grad_norm": 0.9709065556526184, + "learning_rate": 8.360080301786033e-07, + "loss": 0.3405, + "step": 55341 + }, + { + "epoch": 3.8402609118034836, + "grad_norm": 0.5980968475341797, + "learning_rate": 8.352831834653629e-07, + "loss": 0.1451, + "step": 55342 + }, + { + "epoch": 3.84033030324058, + "grad_norm": 1.4643656015396118, + "learning_rate": 8.345586498036295e-07, + "loss": 0.3523, + "step": 55343 + }, + { + "epoch": 3.840399694677677, + "grad_norm": 1.0062743425369263, + "learning_rate": 8.33834429195679e-07, + "loss": 0.3391, + "step": 55344 + }, + { + "epoch": 3.8404690861147737, + "grad_norm": 0.7131629586219788, + "learning_rate": 8.331105216438206e-07, + "loss": 0.2144, + "step": 55345 + }, + { + "epoch": 3.84053847755187, + "grad_norm": 0.7626742124557495, + "learning_rate": 8.323869271503193e-07, + "loss": 0.2148, + "step": 55346 + }, + { + "epoch": 3.840607868988967, + "grad_norm": 0.7701038718223572, + "learning_rate": 8.316636457174731e-07, + "loss": 0.2516, + "step": 55347 + }, + { + "epoch": 3.8406772604260633, + "grad_norm": 0.7395179271697998, + "learning_rate": 8.309406773475581e-07, + "loss": 0.2646, + "step": 55348 + }, + { + "epoch": 3.84074665186316, + "grad_norm": 0.9410377144813538, + "learning_rate": 8.302180220428502e-07, + "loss": 0.2029, + "step": 55349 + }, + { + "epoch": 3.840816043300257, + "grad_norm": 0.6946863532066345, + "learning_rate": 8.294956798056475e-07, + "loss": 0.2258, + "step": 55350 + }, + { + "epoch": 3.8408854347373533, + "grad_norm": 0.9037210941314697, + "learning_rate": 8.287736506382149e-07, + "loss": 0.2538, + "step": 55351 + }, + { + "epoch": 3.84095482617445, + "grad_norm": 0.7257502675056458, + "learning_rate": 8.280519345428395e-07, + "loss": 0.2818, + "step": 55352 + }, + { + "epoch": 3.8410242176115466, + "grad_norm": 0.9226384162902832, + "learning_rate": 8.273305315217972e-07, + "loss": 0.2976, + "step": 55353 + }, + { + "epoch": 3.8410936090486434, + "grad_norm": 0.93133944272995, + "learning_rate": 8.26609441577375e-07, + "loss": 0.2147, + "step": 55354 + }, + { + "epoch": 3.84116300048574, + "grad_norm": 1.1361653804779053, + "learning_rate": 8.258886647118269e-07, + "loss": 0.3195, + "step": 55355 + }, + { + "epoch": 3.8412323919228366, + "grad_norm": 1.1966774463653564, + "learning_rate": 8.251682009274508e-07, + "loss": 0.2554, + "step": 55356 + }, + { + "epoch": 3.8413017833599334, + "grad_norm": 0.8276739120483398, + "learning_rate": 8.244480502265118e-07, + "loss": 0.2596, + "step": 55357 + }, + { + "epoch": 3.84137117479703, + "grad_norm": 0.9900087714195251, + "learning_rate": 8.237282126112745e-07, + "loss": 0.2358, + "step": 55358 + }, + { + "epoch": 3.8414405662341267, + "grad_norm": 0.7158469557762146, + "learning_rate": 8.230086880840371e-07, + "loss": 0.1842, + "step": 55359 + }, + { + "epoch": 3.8415099576712235, + "grad_norm": 1.0801151990890503, + "learning_rate": 8.222894766470423e-07, + "loss": 0.2762, + "step": 55360 + }, + { + "epoch": 3.84157934910832, + "grad_norm": 0.7810486555099487, + "learning_rate": 8.215705783025773e-07, + "loss": 0.2034, + "step": 55361 + }, + { + "epoch": 3.8416487405454167, + "grad_norm": 0.9915131330490112, + "learning_rate": 8.208519930529068e-07, + "loss": 0.3094, + "step": 55362 + }, + { + "epoch": 3.841718131982513, + "grad_norm": 0.8569521903991699, + "learning_rate": 8.201337209002957e-07, + "loss": 0.2461, + "step": 55363 + }, + { + "epoch": 3.84178752341961, + "grad_norm": 0.6191455125808716, + "learning_rate": 8.1941576184702e-07, + "loss": 0.1317, + "step": 55364 + }, + { + "epoch": 3.8418569148567068, + "grad_norm": 0.5820751190185547, + "learning_rate": 8.186981158953333e-07, + "loss": 0.1691, + "step": 55365 + }, + { + "epoch": 3.8419263062938036, + "grad_norm": 0.6669967174530029, + "learning_rate": 8.179807830475228e-07, + "loss": 0.1421, + "step": 55366 + }, + { + "epoch": 3.8419956977309, + "grad_norm": 0.7753250598907471, + "learning_rate": 8.172637633058311e-07, + "loss": 0.2523, + "step": 55367 + }, + { + "epoch": 3.842065089167997, + "grad_norm": 1.01949942111969, + "learning_rate": 8.165470566725342e-07, + "loss": 0.3962, + "step": 55368 + }, + { + "epoch": 3.842134480605093, + "grad_norm": 1.5932459831237793, + "learning_rate": 8.158306631498858e-07, + "loss": 0.3705, + "step": 55369 + }, + { + "epoch": 3.84220387204219, + "grad_norm": 1.0968408584594727, + "learning_rate": 8.151145827401618e-07, + "loss": 0.3391, + "step": 55370 + }, + { + "epoch": 3.842273263479287, + "grad_norm": 0.7376092076301575, + "learning_rate": 8.14398815445605e-07, + "loss": 0.1711, + "step": 55371 + }, + { + "epoch": 3.8423426549163833, + "grad_norm": 0.9078981280326843, + "learning_rate": 8.136833612684803e-07, + "loss": 0.251, + "step": 55372 + }, + { + "epoch": 3.84241204635348, + "grad_norm": 0.9547996520996094, + "learning_rate": 8.129682202110634e-07, + "loss": 0.2746, + "step": 55373 + }, + { + "epoch": 3.8424814377905765, + "grad_norm": 1.0195852518081665, + "learning_rate": 8.122533922755859e-07, + "loss": 0.2575, + "step": 55374 + }, + { + "epoch": 3.8425508292276733, + "grad_norm": 1.7838685512542725, + "learning_rate": 8.11538877464324e-07, + "loss": 0.3914, + "step": 55375 + }, + { + "epoch": 3.84262022066477, + "grad_norm": 1.310317039489746, + "learning_rate": 8.108246757795202e-07, + "loss": 0.3271, + "step": 55376 + }, + { + "epoch": 3.8426896121018665, + "grad_norm": 1.137161374092102, + "learning_rate": 8.101107872234393e-07, + "loss": 0.328, + "step": 55377 + }, + { + "epoch": 3.8427590035389634, + "grad_norm": 0.874898374080658, + "learning_rate": 8.093972117983351e-07, + "loss": 0.3171, + "step": 55378 + }, + { + "epoch": 3.8428283949760598, + "grad_norm": 0.8884127140045166, + "learning_rate": 8.086839495064391e-07, + "loss": 0.2369, + "step": 55379 + }, + { + "epoch": 3.8428977864131566, + "grad_norm": 1.1472927331924438, + "learning_rate": 8.079710003500273e-07, + "loss": 0.2848, + "step": 55380 + }, + { + "epoch": 3.8429671778502534, + "grad_norm": 0.7149012684822083, + "learning_rate": 8.072583643313536e-07, + "loss": 0.1966, + "step": 55381 + }, + { + "epoch": 3.84303656928735, + "grad_norm": 0.9364978075027466, + "learning_rate": 8.065460414526493e-07, + "loss": 0.2224, + "step": 55382 + }, + { + "epoch": 3.8431059607244467, + "grad_norm": 0.9337639808654785, + "learning_rate": 8.058340317161683e-07, + "loss": 0.2693, + "step": 55383 + }, + { + "epoch": 3.843175352161543, + "grad_norm": 0.7689082622528076, + "learning_rate": 8.051223351241533e-07, + "loss": 0.2272, + "step": 55384 + }, + { + "epoch": 3.84324474359864, + "grad_norm": 0.8849804997444153, + "learning_rate": 8.04410951678869e-07, + "loss": 0.2165, + "step": 55385 + }, + { + "epoch": 3.8433141350357367, + "grad_norm": 0.6676241159439087, + "learning_rate": 8.036998813825469e-07, + "loss": 0.2341, + "step": 55386 + }, + { + "epoch": 3.843383526472833, + "grad_norm": 1.2068901062011719, + "learning_rate": 8.02989124237441e-07, + "loss": 0.326, + "step": 55387 + }, + { + "epoch": 3.84345291790993, + "grad_norm": 1.263961911201477, + "learning_rate": 8.022786802457827e-07, + "loss": 0.3086, + "step": 55388 + }, + { + "epoch": 3.8435223093470263, + "grad_norm": 0.9370603561401367, + "learning_rate": 8.015685494098257e-07, + "loss": 0.4187, + "step": 55389 + }, + { + "epoch": 3.843591700784123, + "grad_norm": 0.8202654719352722, + "learning_rate": 8.008587317318017e-07, + "loss": 0.1881, + "step": 55390 + }, + { + "epoch": 3.84366109222122, + "grad_norm": 1.0098013877868652, + "learning_rate": 8.001492272139644e-07, + "loss": 0.2868, + "step": 55391 + }, + { + "epoch": 3.843730483658317, + "grad_norm": 0.5387315154075623, + "learning_rate": 7.994400358585452e-07, + "loss": 0.1187, + "step": 55392 + }, + { + "epoch": 3.843799875095413, + "grad_norm": 0.8778043389320374, + "learning_rate": 7.987311576677758e-07, + "loss": 0.3139, + "step": 55393 + }, + { + "epoch": 3.84386926653251, + "grad_norm": 0.8152715563774109, + "learning_rate": 7.9802259264391e-07, + "loss": 0.2897, + "step": 55394 + }, + { + "epoch": 3.8439386579696064, + "grad_norm": 1.3941256999969482, + "learning_rate": 7.973143407891792e-07, + "loss": 0.5035, + "step": 55395 + }, + { + "epoch": 3.8440080494067033, + "grad_norm": 1.000472903251648, + "learning_rate": 7.96606402105815e-07, + "loss": 0.2742, + "step": 55396 + }, + { + "epoch": 3.8440774408438, + "grad_norm": 0.8237993717193604, + "learning_rate": 7.958987765960601e-07, + "loss": 0.2164, + "step": 55397 + }, + { + "epoch": 3.8441468322808965, + "grad_norm": 1.2688202857971191, + "learning_rate": 7.95191464262135e-07, + "loss": 0.1949, + "step": 55398 + }, + { + "epoch": 3.8442162237179933, + "grad_norm": 0.7149372696876526, + "learning_rate": 7.944844651062821e-07, + "loss": 0.1997, + "step": 55399 + }, + { + "epoch": 3.8442856151550897, + "grad_norm": 1.0865707397460938, + "learning_rate": 7.937777791307333e-07, + "loss": 0.2795, + "step": 55400 + }, + { + "epoch": 3.8443550065921865, + "grad_norm": 0.5551755428314209, + "learning_rate": 7.930714063377087e-07, + "loss": 0.115, + "step": 55401 + }, + { + "epoch": 3.8444243980292834, + "grad_norm": 1.0964922904968262, + "learning_rate": 7.923653467294622e-07, + "loss": 0.255, + "step": 55402 + }, + { + "epoch": 3.8444937894663798, + "grad_norm": 1.0648155212402344, + "learning_rate": 7.916596003082033e-07, + "loss": 0.3231, + "step": 55403 + }, + { + "epoch": 3.8445631809034766, + "grad_norm": 0.8135519623756409, + "learning_rate": 7.909541670761633e-07, + "loss": 0.2187, + "step": 55404 + }, + { + "epoch": 3.844632572340573, + "grad_norm": 0.8160704374313354, + "learning_rate": 7.902490470355739e-07, + "loss": 0.2297, + "step": 55405 + }, + { + "epoch": 3.84470196377767, + "grad_norm": 1.0113400220870972, + "learning_rate": 7.895442401886555e-07, + "loss": 0.3322, + "step": 55406 + }, + { + "epoch": 3.8447713552147667, + "grad_norm": 0.9192177057266235, + "learning_rate": 7.888397465376285e-07, + "loss": 0.3201, + "step": 55407 + }, + { + "epoch": 3.844840746651863, + "grad_norm": 0.725298285484314, + "learning_rate": 7.881355660847356e-07, + "loss": 0.1247, + "step": 55408 + }, + { + "epoch": 3.84491013808896, + "grad_norm": 1.1055697202682495, + "learning_rate": 7.874316988321861e-07, + "loss": 0.2145, + "step": 55409 + }, + { + "epoch": 3.8449795295260563, + "grad_norm": 1.3121230602264404, + "learning_rate": 7.867281447822006e-07, + "loss": 0.2798, + "step": 55410 + }, + { + "epoch": 3.845048920963153, + "grad_norm": 0.6741521954536438, + "learning_rate": 7.860249039369994e-07, + "loss": 0.1734, + "step": 55411 + }, + { + "epoch": 3.84511831240025, + "grad_norm": 0.9769169092178345, + "learning_rate": 7.85321976298814e-07, + "loss": 0.2184, + "step": 55412 + }, + { + "epoch": 3.8451877038373463, + "grad_norm": 0.9650087356567383, + "learning_rate": 7.846193618698538e-07, + "loss": 0.3084, + "step": 55413 + }, + { + "epoch": 3.845257095274443, + "grad_norm": 0.6253988742828369, + "learning_rate": 7.839170606523394e-07, + "loss": 0.12, + "step": 55414 + }, + { + "epoch": 3.8453264867115395, + "grad_norm": 0.8375356793403625, + "learning_rate": 7.8321507264848e-07, + "loss": 0.2281, + "step": 55415 + }, + { + "epoch": 3.8453958781486364, + "grad_norm": 0.8594423532485962, + "learning_rate": 7.82513397860507e-07, + "loss": 0.2041, + "step": 55416 + }, + { + "epoch": 3.845465269585733, + "grad_norm": 1.2066172361373901, + "learning_rate": 7.8181203629063e-07, + "loss": 0.2467, + "step": 55417 + }, + { + "epoch": 3.84553466102283, + "grad_norm": 0.6332009434700012, + "learning_rate": 7.811109879410583e-07, + "loss": 0.1984, + "step": 55418 + }, + { + "epoch": 3.8456040524599264, + "grad_norm": 0.6409661173820496, + "learning_rate": 7.804102528140012e-07, + "loss": 0.1943, + "step": 55419 + }, + { + "epoch": 3.8456734438970233, + "grad_norm": 0.7221017479896545, + "learning_rate": 7.79709830911679e-07, + "loss": 0.2075, + "step": 55420 + }, + { + "epoch": 3.8457428353341196, + "grad_norm": 1.193861961364746, + "learning_rate": 7.790097222363124e-07, + "loss": 0.3053, + "step": 55421 + }, + { + "epoch": 3.8458122267712165, + "grad_norm": 1.0904324054718018, + "learning_rate": 7.783099267900884e-07, + "loss": 0.4212, + "step": 55422 + }, + { + "epoch": 3.8458816182083133, + "grad_norm": 1.0267034769058228, + "learning_rate": 7.776104445752386e-07, + "loss": 0.291, + "step": 55423 + }, + { + "epoch": 3.8459510096454097, + "grad_norm": 0.9737129807472229, + "learning_rate": 7.7691127559395e-07, + "loss": 0.3078, + "step": 55424 + }, + { + "epoch": 3.8460204010825065, + "grad_norm": 1.4500569105148315, + "learning_rate": 7.762124198484433e-07, + "loss": 0.4251, + "step": 55425 + }, + { + "epoch": 3.846089792519603, + "grad_norm": 1.3425662517547607, + "learning_rate": 7.755138773409277e-07, + "loss": 0.2372, + "step": 55426 + }, + { + "epoch": 3.8461591839566998, + "grad_norm": 0.7468223571777344, + "learning_rate": 7.748156480736013e-07, + "loss": 0.1409, + "step": 55427 + }, + { + "epoch": 3.8462285753937966, + "grad_norm": 0.75553959608078, + "learning_rate": 7.741177320486626e-07, + "loss": 0.1801, + "step": 55428 + }, + { + "epoch": 3.846297966830893, + "grad_norm": 0.9327181577682495, + "learning_rate": 7.734201292683207e-07, + "loss": 0.3086, + "step": 55429 + }, + { + "epoch": 3.84636735826799, + "grad_norm": 1.5541951656341553, + "learning_rate": 7.727228397347963e-07, + "loss": 0.3059, + "step": 55430 + }, + { + "epoch": 3.846436749705086, + "grad_norm": 0.8271318078041077, + "learning_rate": 7.720258634502542e-07, + "loss": 0.2077, + "step": 55431 + }, + { + "epoch": 3.846506141142183, + "grad_norm": 0.835269570350647, + "learning_rate": 7.713292004169259e-07, + "loss": 0.2278, + "step": 55432 + }, + { + "epoch": 3.84657553257928, + "grad_norm": 0.8298839330673218, + "learning_rate": 7.706328506369875e-07, + "loss": 0.2203, + "step": 55433 + }, + { + "epoch": 3.8466449240163763, + "grad_norm": 0.7732005715370178, + "learning_rate": 7.699368141126595e-07, + "loss": 0.1679, + "step": 55434 + }, + { + "epoch": 3.846714315453473, + "grad_norm": 0.780523419380188, + "learning_rate": 7.692410908461178e-07, + "loss": 0.1841, + "step": 55435 + }, + { + "epoch": 3.8467837068905695, + "grad_norm": 1.0097318887710571, + "learning_rate": 7.68545680839583e-07, + "loss": 0.3533, + "step": 55436 + }, + { + "epoch": 3.8468530983276663, + "grad_norm": 0.7263970971107483, + "learning_rate": 7.67850584095231e-07, + "loss": 0.2214, + "step": 55437 + }, + { + "epoch": 3.846922489764763, + "grad_norm": 0.7756307125091553, + "learning_rate": 7.671558006152602e-07, + "loss": 0.3282, + "step": 55438 + }, + { + "epoch": 3.8469918812018595, + "grad_norm": 0.9680033922195435, + "learning_rate": 7.664613304018575e-07, + "loss": 0.3466, + "step": 55439 + }, + { + "epoch": 3.8470612726389564, + "grad_norm": 0.8047435283660889, + "learning_rate": 7.657671734572325e-07, + "loss": 0.2431, + "step": 55440 + }, + { + "epoch": 3.8471306640760528, + "grad_norm": 1.3153553009033203, + "learning_rate": 7.650733297835611e-07, + "loss": 0.2298, + "step": 55441 + }, + { + "epoch": 3.8472000555131496, + "grad_norm": 1.0595122575759888, + "learning_rate": 7.643797993830415e-07, + "loss": 0.2669, + "step": 55442 + }, + { + "epoch": 3.8472694469502464, + "grad_norm": 0.8981009125709534, + "learning_rate": 7.63686582257872e-07, + "loss": 0.2593, + "step": 55443 + }, + { + "epoch": 3.8473388383873433, + "grad_norm": 0.9216489791870117, + "learning_rate": 7.629936784102176e-07, + "loss": 0.3205, + "step": 55444 + }, + { + "epoch": 3.8474082298244396, + "grad_norm": 0.7481803894042969, + "learning_rate": 7.623010878422876e-07, + "loss": 0.1301, + "step": 55445 + }, + { + "epoch": 3.8474776212615365, + "grad_norm": 0.807076632976532, + "learning_rate": 7.61608810556258e-07, + "loss": 0.3006, + "step": 55446 + }, + { + "epoch": 3.847547012698633, + "grad_norm": 0.9766877889633179, + "learning_rate": 7.609168465543159e-07, + "loss": 0.2992, + "step": 55447 + }, + { + "epoch": 3.8476164041357297, + "grad_norm": 0.9232653379440308, + "learning_rate": 7.602251958386375e-07, + "loss": 0.2841, + "step": 55448 + }, + { + "epoch": 3.8476857955728265, + "grad_norm": 0.9931731820106506, + "learning_rate": 7.595338584114208e-07, + "loss": 0.2835, + "step": 55449 + }, + { + "epoch": 3.847755187009923, + "grad_norm": 0.9039966464042664, + "learning_rate": 7.588428342748311e-07, + "loss": 0.2097, + "step": 55450 + }, + { + "epoch": 3.8478245784470197, + "grad_norm": 1.1078695058822632, + "learning_rate": 7.581521234310774e-07, + "loss": 0.1926, + "step": 55451 + }, + { + "epoch": 3.847893969884116, + "grad_norm": 1.1708219051361084, + "learning_rate": 7.574617258823136e-07, + "loss": 0.2854, + "step": 55452 + }, + { + "epoch": 3.847963361321213, + "grad_norm": 1.0723886489868164, + "learning_rate": 7.567716416307269e-07, + "loss": 0.3849, + "step": 55453 + }, + { + "epoch": 3.84803275275831, + "grad_norm": 0.8272262215614319, + "learning_rate": 7.560818706785045e-07, + "loss": 0.1969, + "step": 55454 + }, + { + "epoch": 3.848102144195406, + "grad_norm": 1.0298882722854614, + "learning_rate": 7.553924130278112e-07, + "loss": 0.2778, + "step": 55455 + }, + { + "epoch": 3.848171535632503, + "grad_norm": 1.005057692527771, + "learning_rate": 7.547032686808342e-07, + "loss": 0.2782, + "step": 55456 + }, + { + "epoch": 3.8482409270695994, + "grad_norm": 0.7915965914726257, + "learning_rate": 7.540144376397384e-07, + "loss": 0.1613, + "step": 55457 + }, + { + "epoch": 3.8483103185066962, + "grad_norm": 3.3203797340393066, + "learning_rate": 7.533259199067111e-07, + "loss": 0.2548, + "step": 55458 + }, + { + "epoch": 3.848379709943793, + "grad_norm": 0.8009393215179443, + "learning_rate": 7.52637715483917e-07, + "loss": 0.1888, + "step": 55459 + }, + { + "epoch": 3.8484491013808895, + "grad_norm": 1.2860240936279297, + "learning_rate": 7.519498243735324e-07, + "loss": 0.2431, + "step": 55460 + }, + { + "epoch": 3.8485184928179863, + "grad_norm": 0.7790263891220093, + "learning_rate": 7.51262246577722e-07, + "loss": 0.2207, + "step": 55461 + }, + { + "epoch": 3.8485878842550827, + "grad_norm": 0.6109247803688049, + "learning_rate": 7.505749820986618e-07, + "loss": 0.221, + "step": 55462 + }, + { + "epoch": 3.8486572756921795, + "grad_norm": 1.1639201641082764, + "learning_rate": 7.49888030938517e-07, + "loss": 0.4305, + "step": 55463 + }, + { + "epoch": 3.8487266671292764, + "grad_norm": 1.2463939189910889, + "learning_rate": 7.492013930994745e-07, + "loss": 0.232, + "step": 55464 + }, + { + "epoch": 3.8487960585663727, + "grad_norm": 0.8143099546432495, + "learning_rate": 7.485150685836772e-07, + "loss": 0.1725, + "step": 55465 + }, + { + "epoch": 3.8488654500034696, + "grad_norm": 1.1168551445007324, + "learning_rate": 7.478290573933122e-07, + "loss": 0.3216, + "step": 55466 + }, + { + "epoch": 3.848934841440566, + "grad_norm": 0.7941579818725586, + "learning_rate": 7.471433595305222e-07, + "loss": 0.2073, + "step": 55467 + }, + { + "epoch": 3.849004232877663, + "grad_norm": 0.7758721113204956, + "learning_rate": 7.464579749974943e-07, + "loss": 0.2129, + "step": 55468 + }, + { + "epoch": 3.8490736243147596, + "grad_norm": 0.6967142224311829, + "learning_rate": 7.457729037963823e-07, + "loss": 0.2195, + "step": 55469 + }, + { + "epoch": 3.8491430157518565, + "grad_norm": 0.7464134693145752, + "learning_rate": 7.450881459293513e-07, + "loss": 0.2135, + "step": 55470 + }, + { + "epoch": 3.849212407188953, + "grad_norm": 1.3202067613601685, + "learning_rate": 7.444037013985549e-07, + "loss": 0.298, + "step": 55471 + }, + { + "epoch": 3.8492817986260497, + "grad_norm": 1.4153648614883423, + "learning_rate": 7.437195702061694e-07, + "loss": 0.3112, + "step": 55472 + }, + { + "epoch": 3.849351190063146, + "grad_norm": 0.7854339480400085, + "learning_rate": 7.430357523543374e-07, + "loss": 0.2121, + "step": 55473 + }, + { + "epoch": 3.849420581500243, + "grad_norm": 0.9005748629570007, + "learning_rate": 7.423522478452239e-07, + "loss": 0.2322, + "step": 55474 + }, + { + "epoch": 3.8494899729373397, + "grad_norm": 0.7872240543365479, + "learning_rate": 7.416690566809936e-07, + "loss": 0.2063, + "step": 55475 + }, + { + "epoch": 3.849559364374436, + "grad_norm": 0.7383013367652893, + "learning_rate": 7.409861788638006e-07, + "loss": 0.2679, + "step": 55476 + }, + { + "epoch": 3.849628755811533, + "grad_norm": 0.656500518321991, + "learning_rate": 7.403036143957876e-07, + "loss": 0.1535, + "step": 55477 + }, + { + "epoch": 3.8496981472486294, + "grad_norm": 1.0299006700515747, + "learning_rate": 7.396213632791305e-07, + "loss": 0.2659, + "step": 55478 + }, + { + "epoch": 3.849767538685726, + "grad_norm": 0.952555239200592, + "learning_rate": 7.38939425515972e-07, + "loss": 0.2538, + "step": 55479 + }, + { + "epoch": 3.849836930122823, + "grad_norm": 0.6210874915122986, + "learning_rate": 7.382578011084551e-07, + "loss": 0.1692, + "step": 55480 + }, + { + "epoch": 3.8499063215599194, + "grad_norm": 0.7697571516036987, + "learning_rate": 7.375764900587556e-07, + "loss": 0.2026, + "step": 55481 + }, + { + "epoch": 3.8499757129970162, + "grad_norm": 1.4456732273101807, + "learning_rate": 7.368954923690052e-07, + "loss": 0.2807, + "step": 55482 + }, + { + "epoch": 3.8500451044341126, + "grad_norm": 0.49573543667793274, + "learning_rate": 7.362148080413578e-07, + "loss": 0.1174, + "step": 55483 + }, + { + "epoch": 3.8501144958712095, + "grad_norm": 0.7608249187469482, + "learning_rate": 7.35534437077956e-07, + "loss": 0.2022, + "step": 55484 + }, + { + "epoch": 3.8501838873083063, + "grad_norm": 0.9056046009063721, + "learning_rate": 7.34854379480976e-07, + "loss": 0.2698, + "step": 55485 + }, + { + "epoch": 3.8502532787454027, + "grad_norm": 1.2021410465240479, + "learning_rate": 7.34174635252527e-07, + "loss": 0.3128, + "step": 55486 + }, + { + "epoch": 3.8503226701824995, + "grad_norm": 0.6823760867118835, + "learning_rate": 7.33495204394774e-07, + "loss": 0.2182, + "step": 55487 + }, + { + "epoch": 3.850392061619596, + "grad_norm": 1.2315858602523804, + "learning_rate": 7.328160869098599e-07, + "loss": 0.18, + "step": 55488 + }, + { + "epoch": 3.8504614530566927, + "grad_norm": 0.7266910076141357, + "learning_rate": 7.321372827999273e-07, + "loss": 0.2163, + "step": 55489 + }, + { + "epoch": 3.8505308444937896, + "grad_norm": 0.7320114970207214, + "learning_rate": 7.31458792067119e-07, + "loss": 0.3036, + "step": 55490 + }, + { + "epoch": 3.8506002359308864, + "grad_norm": 1.0531260967254639, + "learning_rate": 7.307806147135777e-07, + "loss": 0.2956, + "step": 55491 + }, + { + "epoch": 3.850669627367983, + "grad_norm": 0.8301653265953064, + "learning_rate": 7.301027507414571e-07, + "loss": 0.1853, + "step": 55492 + }, + { + "epoch": 3.850739018805079, + "grad_norm": 1.1775484085083008, + "learning_rate": 7.294252001528668e-07, + "loss": 0.2219, + "step": 55493 + }, + { + "epoch": 3.850808410242176, + "grad_norm": 0.7068482041358948, + "learning_rate": 7.287479629499716e-07, + "loss": 0.2522, + "step": 55494 + }, + { + "epoch": 3.850877801679273, + "grad_norm": 0.9393096566200256, + "learning_rate": 7.280710391349033e-07, + "loss": 0.1982, + "step": 55495 + }, + { + "epoch": 3.8509471931163697, + "grad_norm": 0.6364030838012695, + "learning_rate": 7.273944287097933e-07, + "loss": 0.1262, + "step": 55496 + }, + { + "epoch": 3.851016584553466, + "grad_norm": 1.0680550336837769, + "learning_rate": 7.267181316767846e-07, + "loss": 0.1724, + "step": 55497 + }, + { + "epoch": 3.851085975990563, + "grad_norm": 0.912752628326416, + "learning_rate": 7.260421480380086e-07, + "loss": 0.2898, + "step": 55498 + }, + { + "epoch": 3.8511553674276593, + "grad_norm": 0.8182405233383179, + "learning_rate": 7.253664777956082e-07, + "loss": 0.2414, + "step": 55499 + }, + { + "epoch": 3.851224758864756, + "grad_norm": 0.8418715596199036, + "learning_rate": 7.246911209516927e-07, + "loss": 0.2533, + "step": 55500 + }, + { + "epoch": 3.851294150301853, + "grad_norm": 0.7658770084381104, + "learning_rate": 7.240160775084159e-07, + "loss": 0.269, + "step": 55501 + }, + { + "epoch": 3.8513635417389493, + "grad_norm": 1.0935481786727905, + "learning_rate": 7.233413474678985e-07, + "loss": 0.3465, + "step": 55502 + }, + { + "epoch": 3.851432933176046, + "grad_norm": 1.0312516689300537, + "learning_rate": 7.22666930832272e-07, + "loss": 0.303, + "step": 55503 + }, + { + "epoch": 3.8515023246131426, + "grad_norm": 0.9974145889282227, + "learning_rate": 7.219928276036792e-07, + "loss": 0.2243, + "step": 55504 + }, + { + "epoch": 3.8515717160502394, + "grad_norm": 0.8756589293479919, + "learning_rate": 7.213190377842294e-07, + "loss": 0.2788, + "step": 55505 + }, + { + "epoch": 3.8516411074873362, + "grad_norm": 0.7740480899810791, + "learning_rate": 7.206455613760654e-07, + "loss": 0.2137, + "step": 55506 + }, + { + "epoch": 3.8517104989244326, + "grad_norm": 1.050571084022522, + "learning_rate": 7.199723983812967e-07, + "loss": 0.2928, + "step": 55507 + }, + { + "epoch": 3.8517798903615295, + "grad_norm": 1.4715070724487305, + "learning_rate": 7.192995488020549e-07, + "loss": 0.3228, + "step": 55508 + }, + { + "epoch": 3.851849281798626, + "grad_norm": 0.8983787894248962, + "learning_rate": 7.186270126404604e-07, + "loss": 0.2533, + "step": 55509 + }, + { + "epoch": 3.8519186732357227, + "grad_norm": 0.7119376063346863, + "learning_rate": 7.179547898986561e-07, + "loss": 0.2048, + "step": 55510 + }, + { + "epoch": 3.8519880646728195, + "grad_norm": 1.1953010559082031, + "learning_rate": 7.172828805787402e-07, + "loss": 0.3667, + "step": 55511 + }, + { + "epoch": 3.852057456109916, + "grad_norm": 0.999017596244812, + "learning_rate": 7.166112846828443e-07, + "loss": 0.2814, + "step": 55512 + }, + { + "epoch": 3.8521268475470127, + "grad_norm": 0.7134909629821777, + "learning_rate": 7.159400022130891e-07, + "loss": 0.1468, + "step": 55513 + }, + { + "epoch": 3.852196238984109, + "grad_norm": 0.8587083220481873, + "learning_rate": 7.15269033171595e-07, + "loss": 0.2701, + "step": 55514 + }, + { + "epoch": 3.852265630421206, + "grad_norm": 0.8960656523704529, + "learning_rate": 7.145983775604715e-07, + "loss": 0.2772, + "step": 55515 + }, + { + "epoch": 3.852335021858303, + "grad_norm": 0.7757728695869446, + "learning_rate": 7.139280353818501e-07, + "loss": 0.1954, + "step": 55516 + }, + { + "epoch": 3.8524044132953996, + "grad_norm": 0.7016050815582275, + "learning_rate": 7.132580066378292e-07, + "loss": 0.2461, + "step": 55517 + }, + { + "epoch": 3.852473804732496, + "grad_norm": 1.0388067960739136, + "learning_rate": 7.125882913305404e-07, + "loss": 0.2284, + "step": 55518 + }, + { + "epoch": 3.8525431961695924, + "grad_norm": 0.9356892108917236, + "learning_rate": 7.119188894620931e-07, + "loss": 0.2509, + "step": 55519 + }, + { + "epoch": 3.8526125876066892, + "grad_norm": 1.5934041738510132, + "learning_rate": 7.112498010345969e-07, + "loss": 0.2631, + "step": 55520 + }, + { + "epoch": 3.852681979043786, + "grad_norm": 1.0883073806762695, + "learning_rate": 7.105810260501611e-07, + "loss": 0.2828, + "step": 55521 + }, + { + "epoch": 3.852751370480883, + "grad_norm": 0.8250778317451477, + "learning_rate": 7.09912564510895e-07, + "loss": 0.1718, + "step": 55522 + }, + { + "epoch": 3.8528207619179793, + "grad_norm": 0.8999878764152527, + "learning_rate": 7.092444164189194e-07, + "loss": 0.2897, + "step": 55523 + }, + { + "epoch": 3.852890153355076, + "grad_norm": 0.9069812893867493, + "learning_rate": 7.085765817763434e-07, + "loss": 0.2627, + "step": 55524 + }, + { + "epoch": 3.8529595447921725, + "grad_norm": 1.224478006362915, + "learning_rate": 7.079090605852768e-07, + "loss": 0.3557, + "step": 55525 + }, + { + "epoch": 3.8530289362292693, + "grad_norm": 0.7759560346603394, + "learning_rate": 7.072418528478065e-07, + "loss": 0.3104, + "step": 55526 + }, + { + "epoch": 3.853098327666366, + "grad_norm": 1.1239418983459473, + "learning_rate": 7.065749585660641e-07, + "loss": 0.2123, + "step": 55527 + }, + { + "epoch": 3.8531677191034626, + "grad_norm": 1.336663007736206, + "learning_rate": 7.059083777421482e-07, + "loss": 0.2105, + "step": 55528 + }, + { + "epoch": 3.8532371105405594, + "grad_norm": 0.9120129942893982, + "learning_rate": 7.052421103781459e-07, + "loss": 0.2916, + "step": 55529 + }, + { + "epoch": 3.853306501977656, + "grad_norm": 1.252282738685608, + "learning_rate": 7.045761564761777e-07, + "loss": 0.3476, + "step": 55530 + }, + { + "epoch": 3.8533758934147526, + "grad_norm": 0.7077292799949646, + "learning_rate": 7.03910516038353e-07, + "loss": 0.2162, + "step": 55531 + }, + { + "epoch": 3.8534452848518495, + "grad_norm": 0.6769384145736694, + "learning_rate": 7.03245189066748e-07, + "loss": 0.1795, + "step": 55532 + }, + { + "epoch": 3.853514676288946, + "grad_norm": 1.1014797687530518, + "learning_rate": 7.025801755634831e-07, + "loss": 0.3389, + "step": 55533 + }, + { + "epoch": 3.8535840677260427, + "grad_norm": 0.9278388023376465, + "learning_rate": 7.019154755306567e-07, + "loss": 0.2556, + "step": 55534 + }, + { + "epoch": 3.853653459163139, + "grad_norm": 0.9836105704307556, + "learning_rate": 7.012510889703561e-07, + "loss": 0.3476, + "step": 55535 + }, + { + "epoch": 3.853722850600236, + "grad_norm": 0.8638614416122437, + "learning_rate": 7.005870158846906e-07, + "loss": 0.1971, + "step": 55536 + }, + { + "epoch": 3.8537922420373327, + "grad_norm": 0.9179856777191162, + "learning_rate": 6.999232562757474e-07, + "loss": 0.2922, + "step": 55537 + }, + { + "epoch": 3.853861633474429, + "grad_norm": 1.0112855434417725, + "learning_rate": 6.992598101456249e-07, + "loss": 0.1778, + "step": 55538 + }, + { + "epoch": 3.853931024911526, + "grad_norm": 0.7519378066062927, + "learning_rate": 6.985966774964214e-07, + "loss": 0.2047, + "step": 55539 + }, + { + "epoch": 3.8540004163486223, + "grad_norm": 1.8704274892807007, + "learning_rate": 6.979338583302242e-07, + "loss": 0.235, + "step": 55540 + }, + { + "epoch": 3.854069807785719, + "grad_norm": 0.9672514200210571, + "learning_rate": 6.972713526491315e-07, + "loss": 0.1798, + "step": 55541 + }, + { + "epoch": 3.854139199222816, + "grad_norm": 1.1187056303024292, + "learning_rate": 6.966091604552305e-07, + "loss": 0.1967, + "step": 55542 + }, + { + "epoch": 3.854208590659913, + "grad_norm": 0.5588282346725464, + "learning_rate": 6.959472817506085e-07, + "loss": 0.1503, + "step": 55543 + }, + { + "epoch": 3.8542779820970092, + "grad_norm": 0.9696874618530273, + "learning_rate": 6.952857165373638e-07, + "loss": 0.2846, + "step": 55544 + }, + { + "epoch": 3.8543473735341056, + "grad_norm": 0.9997681975364685, + "learning_rate": 6.946244648175837e-07, + "loss": 0.2016, + "step": 55545 + }, + { + "epoch": 3.8544167649712024, + "grad_norm": 0.6911051273345947, + "learning_rate": 6.939635265933442e-07, + "loss": 0.2129, + "step": 55546 + }, + { + "epoch": 3.8544861564082993, + "grad_norm": 1.2217888832092285, + "learning_rate": 6.933029018667548e-07, + "loss": 0.2846, + "step": 55547 + }, + { + "epoch": 3.854555547845396, + "grad_norm": 0.6957465410232544, + "learning_rate": 6.926425906398693e-07, + "loss": 0.2389, + "step": 55548 + }, + { + "epoch": 3.8546249392824925, + "grad_norm": 0.8908408880233765, + "learning_rate": 6.919825929147971e-07, + "loss": 0.2854, + "step": 55549 + }, + { + "epoch": 3.8546943307195893, + "grad_norm": 0.8076246976852417, + "learning_rate": 6.913229086936146e-07, + "loss": 0.1678, + "step": 55550 + }, + { + "epoch": 3.8547637221566857, + "grad_norm": 0.8491769433021545, + "learning_rate": 6.906635379784088e-07, + "loss": 0.2289, + "step": 55551 + }, + { + "epoch": 3.8548331135937826, + "grad_norm": 0.6700794696807861, + "learning_rate": 6.900044807712447e-07, + "loss": 0.1963, + "step": 55552 + }, + { + "epoch": 3.8549025050308794, + "grad_norm": 0.9364868998527527, + "learning_rate": 6.893457370742206e-07, + "loss": 0.2204, + "step": 55553 + }, + { + "epoch": 3.854971896467976, + "grad_norm": 0.606724739074707, + "learning_rate": 6.886873068894128e-07, + "loss": 0.1899, + "step": 55554 + }, + { + "epoch": 3.8550412879050726, + "grad_norm": 1.618974208831787, + "learning_rate": 6.880291902188974e-07, + "loss": 0.4346, + "step": 55555 + }, + { + "epoch": 3.855110679342169, + "grad_norm": 0.6907756924629211, + "learning_rate": 6.873713870647502e-07, + "loss": 0.3016, + "step": 55556 + }, + { + "epoch": 3.855180070779266, + "grad_norm": 0.9164389967918396, + "learning_rate": 6.867138974290366e-07, + "loss": 0.2465, + "step": 55557 + }, + { + "epoch": 3.8552494622163627, + "grad_norm": 0.8177456855773926, + "learning_rate": 6.860567213138547e-07, + "loss": 0.2381, + "step": 55558 + }, + { + "epoch": 3.855318853653459, + "grad_norm": 1.0026730298995972, + "learning_rate": 6.853998587212696e-07, + "loss": 0.2785, + "step": 55559 + }, + { + "epoch": 3.855388245090556, + "grad_norm": 0.8599623441696167, + "learning_rate": 6.847433096533573e-07, + "loss": 0.2092, + "step": 55560 + }, + { + "epoch": 3.8554576365276523, + "grad_norm": 1.0194153785705566, + "learning_rate": 6.84087074112183e-07, + "loss": 0.3203, + "step": 55561 + }, + { + "epoch": 3.855527027964749, + "grad_norm": 0.7338367700576782, + "learning_rate": 6.834311520998226e-07, + "loss": 0.2355, + "step": 55562 + }, + { + "epoch": 3.855596419401846, + "grad_norm": 0.8693176507949829, + "learning_rate": 6.827755436183525e-07, + "loss": 0.1821, + "step": 55563 + }, + { + "epoch": 3.8556658108389423, + "grad_norm": 0.7628862857818604, + "learning_rate": 6.821202486698374e-07, + "loss": 0.2693, + "step": 55564 + }, + { + "epoch": 3.855735202276039, + "grad_norm": 0.8836972117424011, + "learning_rate": 6.814652672563315e-07, + "loss": 0.1902, + "step": 55565 + }, + { + "epoch": 3.8558045937131356, + "grad_norm": 1.055315375328064, + "learning_rate": 6.808105993799329e-07, + "loss": 0.3308, + "step": 55566 + }, + { + "epoch": 3.8558739851502324, + "grad_norm": 0.8661463856697083, + "learning_rate": 6.801562450426846e-07, + "loss": 0.2307, + "step": 55567 + }, + { + "epoch": 3.855943376587329, + "grad_norm": 0.8273311853408813, + "learning_rate": 6.795022042466625e-07, + "loss": 0.2231, + "step": 55568 + }, + { + "epoch": 3.856012768024426, + "grad_norm": 1.0926052331924438, + "learning_rate": 6.788484769939318e-07, + "loss": 0.3298, + "step": 55569 + }, + { + "epoch": 3.8560821594615224, + "grad_norm": 0.7453609108924866, + "learning_rate": 6.781950632865464e-07, + "loss": 0.2333, + "step": 55570 + }, + { + "epoch": 3.8561515508986193, + "grad_norm": 1.055100679397583, + "learning_rate": 6.775419631265711e-07, + "loss": 0.2124, + "step": 55571 + }, + { + "epoch": 3.8562209423357157, + "grad_norm": 1.1394078731536865, + "learning_rate": 6.768891765160823e-07, + "loss": 0.1663, + "step": 55572 + }, + { + "epoch": 3.8562903337728125, + "grad_norm": 1.0042234659194946, + "learning_rate": 6.762367034571338e-07, + "loss": 0.2414, + "step": 55573 + }, + { + "epoch": 3.8563597252099093, + "grad_norm": 0.6694386601448059, + "learning_rate": 6.755845439517683e-07, + "loss": 0.1703, + "step": 55574 + }, + { + "epoch": 3.8564291166470057, + "grad_norm": 0.8601635098457336, + "learning_rate": 6.749326980020731e-07, + "loss": 0.2665, + "step": 55575 + }, + { + "epoch": 3.8564985080841025, + "grad_norm": 0.797493577003479, + "learning_rate": 6.7428116561008e-07, + "loss": 0.2827, + "step": 55576 + }, + { + "epoch": 3.856567899521199, + "grad_norm": 0.6609406471252441, + "learning_rate": 6.736299467778651e-07, + "loss": 0.1612, + "step": 55577 + }, + { + "epoch": 3.8566372909582958, + "grad_norm": 1.3603147268295288, + "learning_rate": 6.72979041507471e-07, + "loss": 0.2024, + "step": 55578 + }, + { + "epoch": 3.8567066823953926, + "grad_norm": 0.867501974105835, + "learning_rate": 6.72328449800963e-07, + "loss": 0.2505, + "step": 55579 + }, + { + "epoch": 3.856776073832489, + "grad_norm": 1.66901433467865, + "learning_rate": 6.716781716603837e-07, + "loss": 0.4452, + "step": 55580 + }, + { + "epoch": 3.856845465269586, + "grad_norm": 0.8851764798164368, + "learning_rate": 6.710282070877982e-07, + "loss": 0.2736, + "step": 55581 + }, + { + "epoch": 3.856914856706682, + "grad_norm": 1.2741953134536743, + "learning_rate": 6.703785560852493e-07, + "loss": 0.2829, + "step": 55582 + }, + { + "epoch": 3.856984248143779, + "grad_norm": 0.974835991859436, + "learning_rate": 6.69729218654791e-07, + "loss": 0.29, + "step": 55583 + }, + { + "epoch": 3.857053639580876, + "grad_norm": 1.107635498046875, + "learning_rate": 6.69080194798477e-07, + "loss": 0.199, + "step": 55584 + }, + { + "epoch": 3.8571230310179723, + "grad_norm": 1.3302239179611206, + "learning_rate": 6.684314845183504e-07, + "loss": 0.3102, + "step": 55585 + }, + { + "epoch": 3.857192422455069, + "grad_norm": 0.9543046355247498, + "learning_rate": 6.677830878164648e-07, + "loss": 0.3633, + "step": 55586 + }, + { + "epoch": 3.8572618138921655, + "grad_norm": 0.8094434142112732, + "learning_rate": 6.671350046948521e-07, + "loss": 0.1927, + "step": 55587 + }, + { + "epoch": 3.8573312053292623, + "grad_norm": 0.7955819964408875, + "learning_rate": 6.664872351555774e-07, + "loss": 0.1571, + "step": 55588 + }, + { + "epoch": 3.857400596766359, + "grad_norm": 0.5890509486198425, + "learning_rate": 6.658397792006832e-07, + "loss": 0.1535, + "step": 55589 + }, + { + "epoch": 3.8574699882034555, + "grad_norm": 1.0431569814682007, + "learning_rate": 6.651926368322015e-07, + "loss": 0.2334, + "step": 55590 + }, + { + "epoch": 3.8575393796405524, + "grad_norm": 0.9384815096855164, + "learning_rate": 6.645458080521749e-07, + "loss": 0.2703, + "step": 55591 + }, + { + "epoch": 3.8576087710776488, + "grad_norm": 0.8842535018920898, + "learning_rate": 6.638992928626575e-07, + "loss": 0.266, + "step": 55592 + }, + { + "epoch": 3.8576781625147456, + "grad_norm": 0.7905877828598022, + "learning_rate": 6.632530912656921e-07, + "loss": 0.2263, + "step": 55593 + }, + { + "epoch": 3.8577475539518424, + "grad_norm": 1.2766605615615845, + "learning_rate": 6.62607203263299e-07, + "loss": 0.2413, + "step": 55594 + }, + { + "epoch": 3.8578169453889393, + "grad_norm": 0.677349865436554, + "learning_rate": 6.619616288575326e-07, + "loss": 0.1773, + "step": 55595 + }, + { + "epoch": 3.8578863368260357, + "grad_norm": 0.709591805934906, + "learning_rate": 6.613163680504353e-07, + "loss": 0.2431, + "step": 55596 + }, + { + "epoch": 3.8579557282631325, + "grad_norm": 0.8186882734298706, + "learning_rate": 6.60671420844039e-07, + "loss": 0.2614, + "step": 55597 + }, + { + "epoch": 3.858025119700229, + "grad_norm": 0.7033360600471497, + "learning_rate": 6.600267872403642e-07, + "loss": 0.2243, + "step": 55598 + }, + { + "epoch": 3.8580945111373257, + "grad_norm": 0.9051797986030579, + "learning_rate": 6.59382467241465e-07, + "loss": 0.2369, + "step": 55599 + }, + { + "epoch": 3.8581639025744225, + "grad_norm": 1.0033419132232666, + "learning_rate": 6.58738460849373e-07, + "loss": 0.2133, + "step": 55600 + }, + { + "epoch": 3.858233294011519, + "grad_norm": 1.2725483179092407, + "learning_rate": 6.580947680661198e-07, + "loss": 0.4476, + "step": 55601 + }, + { + "epoch": 3.8583026854486158, + "grad_norm": 1.0431499481201172, + "learning_rate": 6.574513888937262e-07, + "loss": 0.3111, + "step": 55602 + }, + { + "epoch": 3.858372076885712, + "grad_norm": 0.9775819778442383, + "learning_rate": 6.568083233342459e-07, + "loss": 0.1924, + "step": 55603 + }, + { + "epoch": 3.858441468322809, + "grad_norm": 1.1492632627487183, + "learning_rate": 6.561655713896886e-07, + "loss": 0.3091, + "step": 55604 + }, + { + "epoch": 3.858510859759906, + "grad_norm": 0.6353903412818909, + "learning_rate": 6.555231330620859e-07, + "loss": 0.1751, + "step": 55605 + }, + { + "epoch": 3.858580251197002, + "grad_norm": 0.8035536408424377, + "learning_rate": 6.548810083534695e-07, + "loss": 0.3078, + "step": 55606 + }, + { + "epoch": 3.858649642634099, + "grad_norm": 0.9137665033340454, + "learning_rate": 6.542391972658824e-07, + "loss": 0.3439, + "step": 55607 + }, + { + "epoch": 3.8587190340711954, + "grad_norm": 0.6560341715812683, + "learning_rate": 6.535976998013227e-07, + "loss": 0.195, + "step": 55608 + }, + { + "epoch": 3.8587884255082923, + "grad_norm": 0.9715626835823059, + "learning_rate": 6.529565159618334e-07, + "loss": 0.2747, + "step": 55609 + }, + { + "epoch": 3.858857816945389, + "grad_norm": 0.769008457660675, + "learning_rate": 6.52315645749435e-07, + "loss": 0.1991, + "step": 55610 + }, + { + "epoch": 3.8589272083824855, + "grad_norm": 1.4270737171173096, + "learning_rate": 6.516750891661483e-07, + "loss": 0.2717, + "step": 55611 + }, + { + "epoch": 3.8589965998195823, + "grad_norm": 0.6479899883270264, + "learning_rate": 6.510348462139936e-07, + "loss": 0.2087, + "step": 55612 + }, + { + "epoch": 3.8590659912566787, + "grad_norm": 0.9090011715888977, + "learning_rate": 6.503949168949919e-07, + "loss": 0.346, + "step": 55613 + }, + { + "epoch": 3.8591353826937755, + "grad_norm": 0.7686859965324402, + "learning_rate": 6.497553012111746e-07, + "loss": 0.2146, + "step": 55614 + }, + { + "epoch": 3.8592047741308724, + "grad_norm": 1.0228592157363892, + "learning_rate": 6.491159991645513e-07, + "loss": 0.3149, + "step": 55615 + }, + { + "epoch": 3.8592741655679688, + "grad_norm": 1.0027949810028076, + "learning_rate": 6.484770107571425e-07, + "loss": 0.1729, + "step": 55616 + }, + { + "epoch": 3.8593435570050656, + "grad_norm": 0.9029875993728638, + "learning_rate": 6.478383359909689e-07, + "loss": 0.3557, + "step": 55617 + }, + { + "epoch": 3.859412948442162, + "grad_norm": 1.1043026447296143, + "learning_rate": 6.471999748680401e-07, + "loss": 0.4498, + "step": 55618 + }, + { + "epoch": 3.859482339879259, + "grad_norm": 0.9959505796432495, + "learning_rate": 6.465619273903656e-07, + "loss": 0.3235, + "step": 55619 + }, + { + "epoch": 3.8595517313163556, + "grad_norm": 0.9297699928283691, + "learning_rate": 6.459241935599769e-07, + "loss": 0.273, + "step": 55620 + }, + { + "epoch": 3.8596211227534525, + "grad_norm": 0.8135722875595093, + "learning_rate": 6.452867733788725e-07, + "loss": 0.17, + "step": 55621 + }, + { + "epoch": 3.859690514190549, + "grad_norm": 0.8787634372711182, + "learning_rate": 6.446496668490731e-07, + "loss": 0.2374, + "step": 55622 + }, + { + "epoch": 3.8597599056276457, + "grad_norm": 0.6960019469261169, + "learning_rate": 6.440128739725881e-07, + "loss": 0.175, + "step": 55623 + }, + { + "epoch": 3.859829297064742, + "grad_norm": 0.8099287748336792, + "learning_rate": 6.43376394751427e-07, + "loss": 0.195, + "step": 55624 + }, + { + "epoch": 3.859898688501839, + "grad_norm": 0.7739891409873962, + "learning_rate": 6.427402291876105e-07, + "loss": 0.2663, + "step": 55625 + }, + { + "epoch": 3.8599680799389358, + "grad_norm": 0.7213617563247681, + "learning_rate": 6.421043772831148e-07, + "loss": 0.1819, + "step": 55626 + }, + { + "epoch": 3.860037471376032, + "grad_norm": 1.0909944772720337, + "learning_rate": 6.414688390399826e-07, + "loss": 0.2588, + "step": 55627 + }, + { + "epoch": 3.860106862813129, + "grad_norm": 0.8519433736801147, + "learning_rate": 6.408336144602012e-07, + "loss": 0.2212, + "step": 55628 + }, + { + "epoch": 3.8601762542502254, + "grad_norm": 0.9305636882781982, + "learning_rate": 6.40198703545769e-07, + "loss": 0.1708, + "step": 55629 + }, + { + "epoch": 3.860245645687322, + "grad_norm": 0.6527231335639954, + "learning_rate": 6.395641062987179e-07, + "loss": 0.1659, + "step": 55630 + }, + { + "epoch": 3.860315037124419, + "grad_norm": 1.1547291278839111, + "learning_rate": 6.38929822721035e-07, + "loss": 0.3789, + "step": 55631 + }, + { + "epoch": 3.8603844285615154, + "grad_norm": 0.7880418300628662, + "learning_rate": 6.382958528147076e-07, + "loss": 0.1637, + "step": 55632 + }, + { + "epoch": 3.8604538199986123, + "grad_norm": 0.8427911996841431, + "learning_rate": 6.376621965817676e-07, + "loss": 0.2927, + "step": 55633 + }, + { + "epoch": 3.8605232114357086, + "grad_norm": 0.6492425203323364, + "learning_rate": 6.370288540241909e-07, + "loss": 0.269, + "step": 55634 + }, + { + "epoch": 3.8605926028728055, + "grad_norm": 1.0607163906097412, + "learning_rate": 6.363958251439872e-07, + "loss": 0.252, + "step": 55635 + }, + { + "epoch": 3.8606619943099023, + "grad_norm": 0.9539787173271179, + "learning_rate": 6.357631099431549e-07, + "loss": 0.3327, + "step": 55636 + }, + { + "epoch": 3.8607313857469987, + "grad_norm": 0.6944581270217896, + "learning_rate": 6.351307084236924e-07, + "loss": 0.2639, + "step": 55637 + }, + { + "epoch": 3.8608007771840955, + "grad_norm": 0.8874754905700684, + "learning_rate": 6.34498620587587e-07, + "loss": 0.2324, + "step": 55638 + }, + { + "epoch": 3.860870168621192, + "grad_norm": 0.8263513445854187, + "learning_rate": 6.33866846436848e-07, + "loss": 0.1985, + "step": 55639 + }, + { + "epoch": 3.8609395600582888, + "grad_norm": 0.7213523387908936, + "learning_rate": 6.332353859734519e-07, + "loss": 0.1787, + "step": 55640 + }, + { + "epoch": 3.8610089514953856, + "grad_norm": 1.1796252727508545, + "learning_rate": 6.326042391994191e-07, + "loss": 0.3645, + "step": 55641 + }, + { + "epoch": 3.861078342932482, + "grad_norm": 1.065689206123352, + "learning_rate": 6.319734061167148e-07, + "loss": 0.4715, + "step": 55642 + }, + { + "epoch": 3.861147734369579, + "grad_norm": 0.9559558629989624, + "learning_rate": 6.313428867273375e-07, + "loss": 0.3199, + "step": 55643 + }, + { + "epoch": 3.861217125806675, + "grad_norm": 0.6239063739776611, + "learning_rate": 6.307126810332964e-07, + "loss": 0.1802, + "step": 55644 + }, + { + "epoch": 3.861286517243772, + "grad_norm": 1.0010243654251099, + "learning_rate": 6.300827890365568e-07, + "loss": 0.3131, + "step": 55645 + }, + { + "epoch": 3.861355908680869, + "grad_norm": 1.0953608751296997, + "learning_rate": 6.294532107391171e-07, + "loss": 0.4109, + "step": 55646 + }, + { + "epoch": 3.8614253001179657, + "grad_norm": 0.615346372127533, + "learning_rate": 6.288239461429646e-07, + "loss": 0.2369, + "step": 55647 + }, + { + "epoch": 3.861494691555062, + "grad_norm": 0.8897688388824463, + "learning_rate": 6.281949952500865e-07, + "loss": 0.2457, + "step": 55648 + }, + { + "epoch": 3.861564082992159, + "grad_norm": 0.7701563239097595, + "learning_rate": 6.275663580624702e-07, + "loss": 0.1995, + "step": 55649 + }, + { + "epoch": 3.8616334744292553, + "grad_norm": 1.0386755466461182, + "learning_rate": 6.269380345820918e-07, + "loss": 0.2636, + "step": 55650 + }, + { + "epoch": 3.861702865866352, + "grad_norm": 1.3309673070907593, + "learning_rate": 6.2631002481095e-07, + "loss": 0.2919, + "step": 55651 + }, + { + "epoch": 3.861772257303449, + "grad_norm": 1.0233681201934814, + "learning_rate": 6.256823287510094e-07, + "loss": 0.3515, + "step": 55652 + }, + { + "epoch": 3.8618416487405454, + "grad_norm": 0.8926147222518921, + "learning_rate": 6.250549464042688e-07, + "loss": 0.3345, + "step": 55653 + }, + { + "epoch": 3.861911040177642, + "grad_norm": 0.8635886311531067, + "learning_rate": 6.244278777726931e-07, + "loss": 0.2186, + "step": 55654 + }, + { + "epoch": 3.8619804316147386, + "grad_norm": 0.7393778562545776, + "learning_rate": 6.238011228582696e-07, + "loss": 0.1729, + "step": 55655 + }, + { + "epoch": 3.8620498230518354, + "grad_norm": 1.2137629985809326, + "learning_rate": 6.231746816629747e-07, + "loss": 0.3385, + "step": 55656 + }, + { + "epoch": 3.8621192144889323, + "grad_norm": 0.9358405470848083, + "learning_rate": 6.225485541887954e-07, + "loss": 0.2917, + "step": 55657 + }, + { + "epoch": 3.8621886059260286, + "grad_norm": 0.7977643609046936, + "learning_rate": 6.21922740437697e-07, + "loss": 0.2625, + "step": 55658 + }, + { + "epoch": 3.8622579973631255, + "grad_norm": 0.8431075215339661, + "learning_rate": 6.212972404116557e-07, + "loss": 0.2543, + "step": 55659 + }, + { + "epoch": 3.862327388800222, + "grad_norm": 0.9040797352790833, + "learning_rate": 6.206720541126477e-07, + "loss": 0.1625, + "step": 55660 + }, + { + "epoch": 3.8623967802373187, + "grad_norm": 1.0953717231750488, + "learning_rate": 6.200471815426489e-07, + "loss": 0.3311, + "step": 55661 + }, + { + "epoch": 3.8624661716744155, + "grad_norm": 0.7794107794761658, + "learning_rate": 6.194226227036359e-07, + "loss": 0.2029, + "step": 55662 + }, + { + "epoch": 3.862535563111512, + "grad_norm": 1.0050930976867676, + "learning_rate": 6.187983775975625e-07, + "loss": 0.3045, + "step": 55663 + }, + { + "epoch": 3.8626049545486087, + "grad_norm": 0.8743815422058105, + "learning_rate": 6.181744462264272e-07, + "loss": 0.2582, + "step": 55664 + }, + { + "epoch": 3.862674345985705, + "grad_norm": 1.2242257595062256, + "learning_rate": 6.175508285921839e-07, + "loss": 0.2391, + "step": 55665 + }, + { + "epoch": 3.862743737422802, + "grad_norm": 0.9818958044052124, + "learning_rate": 6.169275246967976e-07, + "loss": 0.1792, + "step": 55666 + }, + { + "epoch": 3.862813128859899, + "grad_norm": 0.8825499415397644, + "learning_rate": 6.163045345422336e-07, + "loss": 0.212, + "step": 55667 + }, + { + "epoch": 3.862882520296995, + "grad_norm": 0.7441277503967285, + "learning_rate": 6.15681858130468e-07, + "loss": 0.2304, + "step": 55668 + }, + { + "epoch": 3.862951911734092, + "grad_norm": 0.7614662647247314, + "learning_rate": 6.150594954634659e-07, + "loss": 0.2201, + "step": 55669 + }, + { + "epoch": 3.8630213031711884, + "grad_norm": 1.0222578048706055, + "learning_rate": 6.144374465431924e-07, + "loss": 0.2465, + "step": 55670 + }, + { + "epoch": 3.8630906946082852, + "grad_norm": 1.0470606088638306, + "learning_rate": 6.138157113716125e-07, + "loss": 0.2683, + "step": 55671 + }, + { + "epoch": 3.863160086045382, + "grad_norm": 0.9029218554496765, + "learning_rate": 6.131942899506804e-07, + "loss": 0.2424, + "step": 55672 + }, + { + "epoch": 3.863229477482479, + "grad_norm": 2.013427734375, + "learning_rate": 6.125731822823722e-07, + "loss": 0.2618, + "step": 55673 + }, + { + "epoch": 3.8632988689195753, + "grad_norm": 1.3389732837677002, + "learning_rate": 6.119523883686306e-07, + "loss": 0.3641, + "step": 55674 + }, + { + "epoch": 3.863368260356672, + "grad_norm": 0.77860027551651, + "learning_rate": 6.11331908211421e-07, + "loss": 0.198, + "step": 55675 + }, + { + "epoch": 3.8634376517937685, + "grad_norm": 0.8509315252304077, + "learning_rate": 6.107117418127084e-07, + "loss": 0.1868, + "step": 55676 + }, + { + "epoch": 3.8635070432308654, + "grad_norm": 0.893963634967804, + "learning_rate": 6.100918891744468e-07, + "loss": 0.2315, + "step": 55677 + }, + { + "epoch": 3.863576434667962, + "grad_norm": 0.6647753715515137, + "learning_rate": 6.094723502985899e-07, + "loss": 0.1771, + "step": 55678 + }, + { + "epoch": 3.8636458261050586, + "grad_norm": 1.1480411291122437, + "learning_rate": 6.088531251871144e-07, + "loss": 0.2495, + "step": 55679 + }, + { + "epoch": 3.8637152175421554, + "grad_norm": 0.6489607095718384, + "learning_rate": 6.082342138419406e-07, + "loss": 0.1532, + "step": 55680 + }, + { + "epoch": 3.863784608979252, + "grad_norm": 0.7557215690612793, + "learning_rate": 6.076156162650448e-07, + "loss": 0.3057, + "step": 55681 + }, + { + "epoch": 3.8638540004163486, + "grad_norm": 0.6316784024238586, + "learning_rate": 6.069973324583811e-07, + "loss": 0.1686, + "step": 55682 + }, + { + "epoch": 3.8639233918534455, + "grad_norm": 0.8144202828407288, + "learning_rate": 6.063793624238811e-07, + "loss": 0.201, + "step": 55683 + }, + { + "epoch": 3.863992783290542, + "grad_norm": 0.6451818943023682, + "learning_rate": 6.057617061635213e-07, + "loss": 0.1938, + "step": 55684 + }, + { + "epoch": 3.8640621747276387, + "grad_norm": 0.9892510175704956, + "learning_rate": 6.051443636792331e-07, + "loss": 0.2755, + "step": 55685 + }, + { + "epoch": 3.864131566164735, + "grad_norm": 0.8070095777511597, + "learning_rate": 6.045273349729818e-07, + "loss": 0.2324, + "step": 55686 + }, + { + "epoch": 3.864200957601832, + "grad_norm": 0.9630341529846191, + "learning_rate": 6.039106200466993e-07, + "loss": 0.3675, + "step": 55687 + }, + { + "epoch": 3.8642703490389287, + "grad_norm": 1.057112693786621, + "learning_rate": 6.032942189023394e-07, + "loss": 0.2922, + "step": 55688 + }, + { + "epoch": 3.864339740476025, + "grad_norm": 1.1182132959365845, + "learning_rate": 6.02678131541845e-07, + "loss": 0.3191, + "step": 55689 + }, + { + "epoch": 3.864409131913122, + "grad_norm": 0.8828107714653015, + "learning_rate": 6.020623579671702e-07, + "loss": 0.2392, + "step": 55690 + }, + { + "epoch": 3.8644785233502184, + "grad_norm": 1.3152920007705688, + "learning_rate": 6.014468981802468e-07, + "loss": 0.2674, + "step": 55691 + }, + { + "epoch": 3.864547914787315, + "grad_norm": 0.6941110491752625, + "learning_rate": 6.008317521830287e-07, + "loss": 0.1725, + "step": 55692 + }, + { + "epoch": 3.864617306224412, + "grad_norm": 1.7319397926330566, + "learning_rate": 6.002169199774477e-07, + "loss": 0.4087, + "step": 55693 + }, + { + "epoch": 3.8646866976615084, + "grad_norm": 0.6301384568214417, + "learning_rate": 5.996024015654467e-07, + "loss": 0.2235, + "step": 55694 + }, + { + "epoch": 3.8647560890986052, + "grad_norm": 0.710506021976471, + "learning_rate": 5.989881969489797e-07, + "loss": 0.1568, + "step": 55695 + }, + { + "epoch": 3.8648254805357016, + "grad_norm": 0.8659495115280151, + "learning_rate": 5.983743061299563e-07, + "loss": 0.2756, + "step": 55696 + }, + { + "epoch": 3.8648948719727985, + "grad_norm": 0.8784129619598389, + "learning_rate": 5.977607291103416e-07, + "loss": 0.3053, + "step": 55697 + }, + { + "epoch": 3.8649642634098953, + "grad_norm": 0.9638804197311401, + "learning_rate": 5.971474658920672e-07, + "loss": 0.2324, + "step": 55698 + }, + { + "epoch": 3.865033654846992, + "grad_norm": 0.7847364544868469, + "learning_rate": 5.965345164770652e-07, + "loss": 0.2277, + "step": 55699 + }, + { + "epoch": 3.8651030462840885, + "grad_norm": 0.8936609625816345, + "learning_rate": 5.959218808672673e-07, + "loss": 0.3066, + "step": 55700 + }, + { + "epoch": 3.8651724377211854, + "grad_norm": 1.4891867637634277, + "learning_rate": 5.953095590646163e-07, + "loss": 0.2353, + "step": 55701 + }, + { + "epoch": 3.8652418291582817, + "grad_norm": 1.44503653049469, + "learning_rate": 5.946975510710328e-07, + "loss": 0.2867, + "step": 55702 + }, + { + "epoch": 3.8653112205953786, + "grad_norm": 1.1838513612747192, + "learning_rate": 5.940858568884488e-07, + "loss": 0.2927, + "step": 55703 + }, + { + "epoch": 3.8653806120324754, + "grad_norm": 0.7753657698631287, + "learning_rate": 5.934744765188182e-07, + "loss": 0.2303, + "step": 55704 + }, + { + "epoch": 3.865450003469572, + "grad_norm": 0.8463159203529358, + "learning_rate": 5.928634099640507e-07, + "loss": 0.2168, + "step": 55705 + }, + { + "epoch": 3.8655193949066686, + "grad_norm": 0.8382138609886169, + "learning_rate": 5.922526572260778e-07, + "loss": 0.2344, + "step": 55706 + }, + { + "epoch": 3.865588786343765, + "grad_norm": 1.270204782485962, + "learning_rate": 5.916422183068316e-07, + "loss": 0.2694, + "step": 55707 + }, + { + "epoch": 3.865658177780862, + "grad_norm": 0.7540831565856934, + "learning_rate": 5.910320932082325e-07, + "loss": 0.1994, + "step": 55708 + }, + { + "epoch": 3.8657275692179587, + "grad_norm": 0.9615955352783203, + "learning_rate": 5.904222819322236e-07, + "loss": 0.2184, + "step": 55709 + }, + { + "epoch": 3.865796960655055, + "grad_norm": 0.8638392686843872, + "learning_rate": 5.898127844807033e-07, + "loss": 0.3451, + "step": 55710 + }, + { + "epoch": 3.865866352092152, + "grad_norm": 0.9419261813163757, + "learning_rate": 5.892036008556145e-07, + "loss": 0.2248, + "step": 55711 + }, + { + "epoch": 3.8659357435292483, + "grad_norm": 0.9664697051048279, + "learning_rate": 5.88594731058878e-07, + "loss": 0.2549, + "step": 55712 + }, + { + "epoch": 3.866005134966345, + "grad_norm": 0.8792849183082581, + "learning_rate": 5.879861750924253e-07, + "loss": 0.3515, + "step": 55713 + }, + { + "epoch": 3.866074526403442, + "grad_norm": 0.9061967134475708, + "learning_rate": 5.873779329581553e-07, + "loss": 0.1898, + "step": 55714 + }, + { + "epoch": 3.8661439178405383, + "grad_norm": 0.8640848994255066, + "learning_rate": 5.867700046579883e-07, + "loss": 0.2351, + "step": 55715 + }, + { + "epoch": 3.866213309277635, + "grad_norm": 1.3862509727478027, + "learning_rate": 5.861623901938674e-07, + "loss": 0.3226, + "step": 55716 + }, + { + "epoch": 3.8662827007147316, + "grad_norm": 0.7211065292358398, + "learning_rate": 5.85555089567702e-07, + "loss": 0.1721, + "step": 55717 + }, + { + "epoch": 3.8663520921518284, + "grad_norm": 0.8294837474822998, + "learning_rate": 5.849481027813908e-07, + "loss": 0.2154, + "step": 55718 + }, + { + "epoch": 3.8664214835889252, + "grad_norm": 0.8040881752967834, + "learning_rate": 5.843414298368766e-07, + "loss": 0.258, + "step": 55719 + }, + { + "epoch": 3.8664908750260216, + "grad_norm": 0.8693048357963562, + "learning_rate": 5.837350707360578e-07, + "loss": 0.2267, + "step": 55720 + }, + { + "epoch": 3.8665602664631185, + "grad_norm": 0.9798302054405212, + "learning_rate": 5.831290254808552e-07, + "loss": 0.3293, + "step": 55721 + }, + { + "epoch": 3.866629657900215, + "grad_norm": 1.8273688554763794, + "learning_rate": 5.825232940731784e-07, + "loss": 0.467, + "step": 55722 + }, + { + "epoch": 3.8666990493373117, + "grad_norm": 1.0868443250656128, + "learning_rate": 5.819178765149369e-07, + "loss": 0.2606, + "step": 55723 + }, + { + "epoch": 3.8667684407744085, + "grad_norm": 0.678135335445404, + "learning_rate": 5.813127728080514e-07, + "loss": 0.2383, + "step": 55724 + }, + { + "epoch": 3.8668378322115053, + "grad_norm": 0.9472784996032715, + "learning_rate": 5.807079829544204e-07, + "loss": 0.2689, + "step": 55725 + }, + { + "epoch": 3.8669072236486017, + "grad_norm": 0.7233505845069885, + "learning_rate": 5.801035069559646e-07, + "loss": 0.2059, + "step": 55726 + }, + { + "epoch": 3.8669766150856986, + "grad_norm": 0.9440191984176636, + "learning_rate": 5.794993448145825e-07, + "loss": 0.2465, + "step": 55727 + }, + { + "epoch": 3.867046006522795, + "grad_norm": 0.9330465197563171, + "learning_rate": 5.788954965321947e-07, + "loss": 0.2869, + "step": 55728 + }, + { + "epoch": 3.867115397959892, + "grad_norm": 0.7848411202430725, + "learning_rate": 5.782919621106997e-07, + "loss": 0.3105, + "step": 55729 + }, + { + "epoch": 3.8671847893969886, + "grad_norm": 0.7842294573783875, + "learning_rate": 5.77688741551996e-07, + "loss": 0.2085, + "step": 55730 + }, + { + "epoch": 3.867254180834085, + "grad_norm": 0.9425410628318787, + "learning_rate": 5.770858348579933e-07, + "loss": 0.2438, + "step": 55731 + }, + { + "epoch": 3.867323572271182, + "grad_norm": 0.8986322283744812, + "learning_rate": 5.764832420305898e-07, + "loss": 0.3092, + "step": 55732 + }, + { + "epoch": 3.8673929637082782, + "grad_norm": 0.7915800213813782, + "learning_rate": 5.758809630717066e-07, + "loss": 0.1951, + "step": 55733 + }, + { + "epoch": 3.867462355145375, + "grad_norm": 0.7360067963600159, + "learning_rate": 5.752789979832196e-07, + "loss": 0.2058, + "step": 55734 + }, + { + "epoch": 3.867531746582472, + "grad_norm": 1.059323787689209, + "learning_rate": 5.746773467670496e-07, + "loss": 0.2761, + "step": 55735 + }, + { + "epoch": 3.8676011380195683, + "grad_norm": 0.7779903411865234, + "learning_rate": 5.740760094250841e-07, + "loss": 0.1571, + "step": 55736 + }, + { + "epoch": 3.867670529456665, + "grad_norm": 0.7572636604309082, + "learning_rate": 5.734749859592214e-07, + "loss": 0.1624, + "step": 55737 + }, + { + "epoch": 3.8677399208937615, + "grad_norm": 0.6734554767608643, + "learning_rate": 5.728742763713712e-07, + "loss": 0.2029, + "step": 55738 + }, + { + "epoch": 3.8678093123308583, + "grad_norm": 1.0210040807724, + "learning_rate": 5.722738806634209e-07, + "loss": 0.3187, + "step": 55739 + }, + { + "epoch": 3.867878703767955, + "grad_norm": 0.6205387115478516, + "learning_rate": 5.716737988372578e-07, + "loss": 0.1571, + "step": 55740 + }, + { + "epoch": 3.8679480952050516, + "grad_norm": 0.951809287071228, + "learning_rate": 5.710740308947915e-07, + "loss": 0.283, + "step": 55741 + }, + { + "epoch": 3.8680174866421484, + "grad_norm": 0.8112169504165649, + "learning_rate": 5.704745768379094e-07, + "loss": 0.2152, + "step": 55742 + }, + { + "epoch": 3.868086878079245, + "grad_norm": 1.0346559286117554, + "learning_rate": 5.698754366684988e-07, + "loss": 0.251, + "step": 55743 + }, + { + "epoch": 3.8681562695163416, + "grad_norm": 1.071765661239624, + "learning_rate": 5.692766103884584e-07, + "loss": 0.2643, + "step": 55744 + }, + { + "epoch": 3.8682256609534384, + "grad_norm": 0.5954322814941406, + "learning_rate": 5.686780979996642e-07, + "loss": 0.1515, + "step": 55745 + }, + { + "epoch": 3.8682950523905353, + "grad_norm": 0.8430383801460266, + "learning_rate": 5.680798995040259e-07, + "loss": 0.182, + "step": 55746 + }, + { + "epoch": 3.8683644438276317, + "grad_norm": 0.8297972679138184, + "learning_rate": 5.67482014903431e-07, + "loss": 0.2623, + "step": 55747 + }, + { + "epoch": 3.868433835264728, + "grad_norm": 0.6320863962173462, + "learning_rate": 5.668844441997556e-07, + "loss": 0.177, + "step": 55748 + }, + { + "epoch": 3.868503226701825, + "grad_norm": 0.8448682427406311, + "learning_rate": 5.662871873948872e-07, + "loss": 0.1914, + "step": 55749 + }, + { + "epoch": 3.8685726181389217, + "grad_norm": 0.6025377511978149, + "learning_rate": 5.656902444907131e-07, + "loss": 0.139, + "step": 55750 + }, + { + "epoch": 3.8686420095760186, + "grad_norm": 0.8056721091270447, + "learning_rate": 5.650936154891096e-07, + "loss": 0.2996, + "step": 55751 + }, + { + "epoch": 3.868711401013115, + "grad_norm": 0.6591221690177917, + "learning_rate": 5.644973003919751e-07, + "loss": 0.1613, + "step": 55752 + }, + { + "epoch": 3.868780792450212, + "grad_norm": 1.294439435005188, + "learning_rate": 5.639012992011972e-07, + "loss": 0.3266, + "step": 55753 + }, + { + "epoch": 3.868850183887308, + "grad_norm": 0.9146231412887573, + "learning_rate": 5.633056119186298e-07, + "loss": 0.2554, + "step": 55754 + }, + { + "epoch": 3.868919575324405, + "grad_norm": 0.7127030491828918, + "learning_rate": 5.627102385461824e-07, + "loss": 0.2476, + "step": 55755 + }, + { + "epoch": 3.868988966761502, + "grad_norm": 1.2032557725906372, + "learning_rate": 5.621151790857205e-07, + "loss": 0.3112, + "step": 55756 + }, + { + "epoch": 3.8690583581985982, + "grad_norm": 0.996741771697998, + "learning_rate": 5.6152043353912e-07, + "loss": 0.2511, + "step": 55757 + }, + { + "epoch": 3.869127749635695, + "grad_norm": 0.6738576889038086, + "learning_rate": 5.609260019082574e-07, + "loss": 0.3234, + "step": 55758 + }, + { + "epoch": 3.8691971410727914, + "grad_norm": 1.1335372924804688, + "learning_rate": 5.6033188419502e-07, + "loss": 0.3275, + "step": 55759 + }, + { + "epoch": 3.8692665325098883, + "grad_norm": 0.8095943927764893, + "learning_rate": 5.59738080401273e-07, + "loss": 0.1943, + "step": 55760 + }, + { + "epoch": 3.869335923946985, + "grad_norm": 0.9111264944076538, + "learning_rate": 5.591445905289039e-07, + "loss": 0.2382, + "step": 55761 + }, + { + "epoch": 3.8694053153840815, + "grad_norm": 0.9174496531486511, + "learning_rate": 5.585514145797776e-07, + "loss": 0.2751, + "step": 55762 + }, + { + "epoch": 3.8694747068211783, + "grad_norm": 0.8837311863899231, + "learning_rate": 5.579585525557596e-07, + "loss": 0.2428, + "step": 55763 + }, + { + "epoch": 3.8695440982582747, + "grad_norm": 0.9181462526321411, + "learning_rate": 5.57366004458737e-07, + "loss": 0.4095, + "step": 55764 + }, + { + "epoch": 3.8696134896953716, + "grad_norm": 0.999370276927948, + "learning_rate": 5.567737702905751e-07, + "loss": 0.2218, + "step": 55765 + }, + { + "epoch": 3.8696828811324684, + "grad_norm": 1.3154746294021606, + "learning_rate": 5.561818500531279e-07, + "loss": 0.2503, + "step": 55766 + }, + { + "epoch": 3.8697522725695648, + "grad_norm": 0.837144672870636, + "learning_rate": 5.55590243748283e-07, + "loss": 0.3794, + "step": 55767 + }, + { + "epoch": 3.8698216640066616, + "grad_norm": 0.8013516664505005, + "learning_rate": 5.549989513779052e-07, + "loss": 0.228, + "step": 55768 + }, + { + "epoch": 3.869891055443758, + "grad_norm": 1.6183522939682007, + "learning_rate": 5.544079729438601e-07, + "loss": 0.3935, + "step": 55769 + }, + { + "epoch": 3.869960446880855, + "grad_norm": 0.8422895073890686, + "learning_rate": 5.538173084480014e-07, + "loss": 0.1294, + "step": 55770 + }, + { + "epoch": 3.8700298383179517, + "grad_norm": 0.7943530678749084, + "learning_rate": 5.532269578922055e-07, + "loss": 0.2157, + "step": 55771 + }, + { + "epoch": 3.8700992297550485, + "grad_norm": 1.1949735879898071, + "learning_rate": 5.526369212783378e-07, + "loss": 0.2067, + "step": 55772 + }, + { + "epoch": 3.870168621192145, + "grad_norm": 0.9913954734802246, + "learning_rate": 5.52047198608252e-07, + "loss": 0.262, + "step": 55773 + }, + { + "epoch": 3.8702380126292413, + "grad_norm": 0.7748383283615112, + "learning_rate": 5.514577898838247e-07, + "loss": 0.1883, + "step": 55774 + }, + { + "epoch": 3.870307404066338, + "grad_norm": 1.0360356569290161, + "learning_rate": 5.508686951069098e-07, + "loss": 0.2809, + "step": 55775 + }, + { + "epoch": 3.870376795503435, + "grad_norm": 0.6634125709533691, + "learning_rate": 5.502799142793502e-07, + "loss": 0.1689, + "step": 55776 + }, + { + "epoch": 3.8704461869405318, + "grad_norm": 0.8818268775939941, + "learning_rate": 5.496914474030224e-07, + "loss": 0.2719, + "step": 55777 + }, + { + "epoch": 3.870515578377628, + "grad_norm": 1.0879236459732056, + "learning_rate": 5.491032944797803e-07, + "loss": 0.3898, + "step": 55778 + }, + { + "epoch": 3.870584969814725, + "grad_norm": 0.6548583507537842, + "learning_rate": 5.485154555114891e-07, + "loss": 0.1483, + "step": 55779 + }, + { + "epoch": 3.8706543612518214, + "grad_norm": 0.9437021613121033, + "learning_rate": 5.479279304999918e-07, + "loss": 0.2228, + "step": 55780 + }, + { + "epoch": 3.870723752688918, + "grad_norm": 0.9369896650314331, + "learning_rate": 5.473407194471425e-07, + "loss": 0.1708, + "step": 55781 + }, + { + "epoch": 3.870793144126015, + "grad_norm": 0.6340360045433044, + "learning_rate": 5.467538223548063e-07, + "loss": 0.201, + "step": 55782 + }, + { + "epoch": 3.8708625355631114, + "grad_norm": 0.7849039435386658, + "learning_rate": 5.461672392248263e-07, + "loss": 0.1954, + "step": 55783 + }, + { + "epoch": 3.8709319270002083, + "grad_norm": 1.1113057136535645, + "learning_rate": 5.455809700590564e-07, + "loss": 0.2616, + "step": 55784 + }, + { + "epoch": 3.8710013184373047, + "grad_norm": 0.6470054388046265, + "learning_rate": 5.449950148593508e-07, + "loss": 0.1995, + "step": 55785 + }, + { + "epoch": 3.8710707098744015, + "grad_norm": 0.95774245262146, + "learning_rate": 5.444093736275524e-07, + "loss": 0.2528, + "step": 55786 + }, + { + "epoch": 3.8711401013114983, + "grad_norm": 0.6811963319778442, + "learning_rate": 5.438240463655153e-07, + "loss": 0.1348, + "step": 55787 + }, + { + "epoch": 3.8712094927485947, + "grad_norm": 0.9171746969223022, + "learning_rate": 5.432390330750936e-07, + "loss": 0.2821, + "step": 55788 + }, + { + "epoch": 3.8712788841856915, + "grad_norm": 0.8214308023452759, + "learning_rate": 5.426543337581302e-07, + "loss": 0.3092, + "step": 55789 + }, + { + "epoch": 3.871348275622788, + "grad_norm": 0.9094278812408447, + "learning_rate": 5.420699484164571e-07, + "loss": 0.2438, + "step": 55790 + }, + { + "epoch": 3.8714176670598848, + "grad_norm": 1.0011107921600342, + "learning_rate": 5.414858770519282e-07, + "loss": 0.2183, + "step": 55791 + }, + { + "epoch": 3.8714870584969816, + "grad_norm": 1.1747585535049438, + "learning_rate": 5.409021196663977e-07, + "loss": 0.4077, + "step": 55792 + }, + { + "epoch": 3.871556449934078, + "grad_norm": 0.8582077622413635, + "learning_rate": 5.403186762616974e-07, + "loss": 0.2281, + "step": 55793 + }, + { + "epoch": 3.871625841371175, + "grad_norm": 1.000250220298767, + "learning_rate": 5.397355468396703e-07, + "loss": 0.2503, + "step": 55794 + }, + { + "epoch": 3.871695232808271, + "grad_norm": 0.9377021193504333, + "learning_rate": 5.391527314021594e-07, + "loss": 0.2973, + "step": 55795 + }, + { + "epoch": 3.871764624245368, + "grad_norm": 1.2374463081359863, + "learning_rate": 5.385702299510076e-07, + "loss": 0.2935, + "step": 55796 + }, + { + "epoch": 3.871834015682465, + "grad_norm": 1.364754557609558, + "learning_rate": 5.379880424880357e-07, + "loss": 0.2374, + "step": 55797 + }, + { + "epoch": 3.8719034071195617, + "grad_norm": 0.9533836245536804, + "learning_rate": 5.374061690151089e-07, + "loss": 0.3387, + "step": 55798 + }, + { + "epoch": 3.871972798556658, + "grad_norm": 0.706741452217102, + "learning_rate": 5.368246095340479e-07, + "loss": 0.1182, + "step": 55799 + }, + { + "epoch": 3.872042189993755, + "grad_norm": 0.9312388896942139, + "learning_rate": 5.362433640466958e-07, + "loss": 0.207, + "step": 55800 + }, + { + "epoch": 3.8721115814308513, + "grad_norm": 1.1026487350463867, + "learning_rate": 5.356624325548842e-07, + "loss": 0.5337, + "step": 55801 + }, + { + "epoch": 3.872180972867948, + "grad_norm": 0.7096052765846252, + "learning_rate": 5.350818150604342e-07, + "loss": 0.1953, + "step": 55802 + }, + { + "epoch": 3.872250364305045, + "grad_norm": 0.9983834624290466, + "learning_rate": 5.345015115651997e-07, + "loss": 0.2672, + "step": 55803 + }, + { + "epoch": 3.8723197557421414, + "grad_norm": 1.3702523708343506, + "learning_rate": 5.339215220710014e-07, + "loss": 0.2581, + "step": 55804 + }, + { + "epoch": 3.872389147179238, + "grad_norm": 1.0239219665527344, + "learning_rate": 5.333418465796825e-07, + "loss": 0.2259, + "step": 55805 + }, + { + "epoch": 3.8724585386163346, + "grad_norm": 1.069131851196289, + "learning_rate": 5.327624850930524e-07, + "loss": 0.5426, + "step": 55806 + }, + { + "epoch": 3.8725279300534314, + "grad_norm": 0.7727813720703125, + "learning_rate": 5.321834376129431e-07, + "loss": 0.3186, + "step": 55807 + }, + { + "epoch": 3.8725973214905283, + "grad_norm": 0.9246786236763, + "learning_rate": 5.316047041412087e-07, + "loss": 0.31, + "step": 55808 + }, + { + "epoch": 3.8726667129276247, + "grad_norm": 0.868133544921875, + "learning_rate": 5.310262846796476e-07, + "loss": 0.2661, + "step": 55809 + }, + { + "epoch": 3.8727361043647215, + "grad_norm": 0.9406101107597351, + "learning_rate": 5.30448179230092e-07, + "loss": 0.3257, + "step": 55810 + }, + { + "epoch": 3.872805495801818, + "grad_norm": 0.9705154895782471, + "learning_rate": 5.298703877943845e-07, + "loss": 0.2516, + "step": 55811 + }, + { + "epoch": 3.8728748872389147, + "grad_norm": 1.173585295677185, + "learning_rate": 5.292929103743238e-07, + "loss": 0.3113, + "step": 55812 + }, + { + "epoch": 3.8729442786760115, + "grad_norm": 0.7350037097930908, + "learning_rate": 5.287157469717419e-07, + "loss": 0.2347, + "step": 55813 + }, + { + "epoch": 3.873013670113108, + "grad_norm": 0.8220683336257935, + "learning_rate": 5.281388975884705e-07, + "loss": 0.1921, + "step": 55814 + }, + { + "epoch": 3.8730830615502048, + "grad_norm": 0.692931056022644, + "learning_rate": 5.275623622263192e-07, + "loss": 0.2445, + "step": 55815 + }, + { + "epoch": 3.873152452987301, + "grad_norm": 1.4439525604248047, + "learning_rate": 5.269861408871202e-07, + "loss": 0.1739, + "step": 55816 + }, + { + "epoch": 3.873221844424398, + "grad_norm": 0.9599428176879883, + "learning_rate": 5.264102335726828e-07, + "loss": 0.2654, + "step": 55817 + }, + { + "epoch": 3.873291235861495, + "grad_norm": 1.3500078916549683, + "learning_rate": 5.258346402848279e-07, + "loss": 0.3491, + "step": 55818 + }, + { + "epoch": 3.873360627298591, + "grad_norm": 1.0148078203201294, + "learning_rate": 5.252593610253653e-07, + "loss": 0.3058, + "step": 55819 + }, + { + "epoch": 3.873430018735688, + "grad_norm": 1.1309051513671875, + "learning_rate": 5.246843957961268e-07, + "loss": 0.2175, + "step": 55820 + }, + { + "epoch": 3.8734994101727844, + "grad_norm": 0.9504892230033875, + "learning_rate": 5.241097445989107e-07, + "loss": 0.2896, + "step": 55821 + }, + { + "epoch": 3.8735688016098813, + "grad_norm": 0.7243127226829529, + "learning_rate": 5.23535407435538e-07, + "loss": 0.1831, + "step": 55822 + }, + { + "epoch": 3.873638193046978, + "grad_norm": 0.8907527327537537, + "learning_rate": 5.229613843078296e-07, + "loss": 0.2023, + "step": 55823 + }, + { + "epoch": 3.873707584484075, + "grad_norm": 0.836132824420929, + "learning_rate": 5.223876752175839e-07, + "loss": 0.2686, + "step": 55824 + }, + { + "epoch": 3.8737769759211713, + "grad_norm": 1.0745437145233154, + "learning_rate": 5.218142801666215e-07, + "loss": 0.2489, + "step": 55825 + }, + { + "epoch": 3.873846367358268, + "grad_norm": 0.9158615469932556, + "learning_rate": 5.212411991567523e-07, + "loss": 0.2518, + "step": 55826 + }, + { + "epoch": 3.8739157587953645, + "grad_norm": 0.9764126539230347, + "learning_rate": 5.206684321897859e-07, + "loss": 0.2103, + "step": 55827 + }, + { + "epoch": 3.8739851502324614, + "grad_norm": 0.958160936832428, + "learning_rate": 5.200959792675209e-07, + "loss": 0.2422, + "step": 55828 + }, + { + "epoch": 3.874054541669558, + "grad_norm": 0.9280698895454407, + "learning_rate": 5.195238403917669e-07, + "loss": 0.1845, + "step": 55829 + }, + { + "epoch": 3.8741239331066546, + "grad_norm": 1.039078712463379, + "learning_rate": 5.189520155643446e-07, + "loss": 0.2882, + "step": 55830 + }, + { + "epoch": 3.8741933245437514, + "grad_norm": 0.6331185698509216, + "learning_rate": 5.183805047870416e-07, + "loss": 0.1996, + "step": 55831 + }, + { + "epoch": 3.874262715980848, + "grad_norm": 0.9354984760284424, + "learning_rate": 5.178093080616786e-07, + "loss": 0.2526, + "step": 55832 + }, + { + "epoch": 3.8743321074179446, + "grad_norm": 0.8879692554473877, + "learning_rate": 5.17238425390043e-07, + "loss": 0.1707, + "step": 55833 + }, + { + "epoch": 3.8744014988550415, + "grad_norm": 1.0048460960388184, + "learning_rate": 5.166678567739447e-07, + "loss": 0.2909, + "step": 55834 + }, + { + "epoch": 3.874470890292138, + "grad_norm": 0.7322627305984497, + "learning_rate": 5.16097602215182e-07, + "loss": 0.2163, + "step": 55835 + }, + { + "epoch": 3.8745402817292347, + "grad_norm": 1.1325737237930298, + "learning_rate": 5.155276617155536e-07, + "loss": 0.2599, + "step": 55836 + }, + { + "epoch": 3.874609673166331, + "grad_norm": 0.887016236782074, + "learning_rate": 5.14958035276869e-07, + "loss": 0.2901, + "step": 55837 + }, + { + "epoch": 3.874679064603428, + "grad_norm": 0.5636023283004761, + "learning_rate": 5.143887229009158e-07, + "loss": 0.1786, + "step": 55838 + }, + { + "epoch": 3.8747484560405248, + "grad_norm": 0.8911367654800415, + "learning_rate": 5.138197245895037e-07, + "loss": 0.2881, + "step": 55839 + }, + { + "epoch": 3.874817847477621, + "grad_norm": 0.8665220141410828, + "learning_rate": 5.132510403444091e-07, + "loss": 0.1536, + "step": 55840 + }, + { + "epoch": 3.874887238914718, + "grad_norm": 0.6300826072692871, + "learning_rate": 5.126826701674414e-07, + "loss": 0.178, + "step": 55841 + }, + { + "epoch": 3.8749566303518144, + "grad_norm": 1.0464537143707275, + "learning_rate": 5.121146140603883e-07, + "loss": 0.2866, + "step": 55842 + }, + { + "epoch": 3.875026021788911, + "grad_norm": 1.1648775339126587, + "learning_rate": 5.115468720250482e-07, + "loss": 0.2746, + "step": 55843 + }, + { + "epoch": 3.875095413226008, + "grad_norm": 0.8525948524475098, + "learning_rate": 5.109794440632198e-07, + "loss": 0.2095, + "step": 55844 + }, + { + "epoch": 3.8751648046631044, + "grad_norm": 1.063071846961975, + "learning_rate": 5.104123301766795e-07, + "loss": 0.4879, + "step": 55845 + }, + { + "epoch": 3.8752341961002013, + "grad_norm": 0.8934267163276672, + "learning_rate": 5.098455303672256e-07, + "loss": 0.1497, + "step": 55846 + }, + { + "epoch": 3.8753035875372976, + "grad_norm": 1.6281728744506836, + "learning_rate": 5.092790446366458e-07, + "loss": 0.1742, + "step": 55847 + }, + { + "epoch": 3.8753729789743945, + "grad_norm": 1.2172417640686035, + "learning_rate": 5.087128729867274e-07, + "loss": 0.2429, + "step": 55848 + }, + { + "epoch": 3.8754423704114913, + "grad_norm": 0.7666298151016235, + "learning_rate": 5.08147015419258e-07, + "loss": 0.1763, + "step": 55849 + }, + { + "epoch": 3.875511761848588, + "grad_norm": 1.616335153579712, + "learning_rate": 5.075814719360361e-07, + "loss": 0.2252, + "step": 55850 + }, + { + "epoch": 3.8755811532856845, + "grad_norm": 0.980634331703186, + "learning_rate": 5.07016242538827e-07, + "loss": 0.2988, + "step": 55851 + }, + { + "epoch": 3.8756505447227814, + "grad_norm": 0.6987365484237671, + "learning_rate": 5.064513272294291e-07, + "loss": 0.2248, + "step": 55852 + }, + { + "epoch": 3.8757199361598778, + "grad_norm": 1.2979868650436401, + "learning_rate": 5.058867260096189e-07, + "loss": 0.366, + "step": 55853 + }, + { + "epoch": 3.8757893275969746, + "grad_norm": 0.8785693049430847, + "learning_rate": 5.053224388811729e-07, + "loss": 0.2528, + "step": 55854 + }, + { + "epoch": 3.8758587190340714, + "grad_norm": 0.7163004875183105, + "learning_rate": 5.047584658458893e-07, + "loss": 0.2437, + "step": 55855 + }, + { + "epoch": 3.875928110471168, + "grad_norm": 1.0631102323532104, + "learning_rate": 5.041948069055336e-07, + "loss": 0.3787, + "step": 55856 + }, + { + "epoch": 3.8759975019082646, + "grad_norm": 0.8612929582595825, + "learning_rate": 5.036314620618931e-07, + "loss": 0.2673, + "step": 55857 + }, + { + "epoch": 3.876066893345361, + "grad_norm": 1.2795851230621338, + "learning_rate": 5.030684313167555e-07, + "loss": 0.2685, + "step": 55858 + }, + { + "epoch": 3.876136284782458, + "grad_norm": 1.0662435293197632, + "learning_rate": 5.025057146718748e-07, + "loss": 0.2996, + "step": 55859 + }, + { + "epoch": 3.8762056762195547, + "grad_norm": 0.5670832395553589, + "learning_rate": 5.019433121290495e-07, + "loss": 0.1513, + "step": 55860 + }, + { + "epoch": 3.876275067656651, + "grad_norm": 1.0968408584594727, + "learning_rate": 5.01381223690034e-07, + "loss": 0.3132, + "step": 55861 + }, + { + "epoch": 3.876344459093748, + "grad_norm": 0.7946087121963501, + "learning_rate": 5.008194493566154e-07, + "loss": 0.1829, + "step": 55862 + }, + { + "epoch": 3.8764138505308443, + "grad_norm": 1.392264485359192, + "learning_rate": 5.002579891305703e-07, + "loss": 0.4506, + "step": 55863 + }, + { + "epoch": 3.876483241967941, + "grad_norm": 0.8809707164764404, + "learning_rate": 4.99696843013664e-07, + "loss": 0.2332, + "step": 55864 + }, + { + "epoch": 3.876552633405038, + "grad_norm": 1.5567874908447266, + "learning_rate": 4.991360110076837e-07, + "loss": 0.2251, + "step": 55865 + }, + { + "epoch": 3.8766220248421344, + "grad_norm": 0.938831090927124, + "learning_rate": 4.985754931143727e-07, + "loss": 0.3049, + "step": 55866 + }, + { + "epoch": 3.876691416279231, + "grad_norm": 0.8603353500366211, + "learning_rate": 4.980152893355183e-07, + "loss": 0.2575, + "step": 55867 + }, + { + "epoch": 3.8767608077163276, + "grad_norm": 0.8880322575569153, + "learning_rate": 4.974553996728748e-07, + "loss": 0.3156, + "step": 55868 + }, + { + "epoch": 3.8768301991534244, + "grad_norm": 0.8782685995101929, + "learning_rate": 4.968958241282406e-07, + "loss": 0.1495, + "step": 55869 + }, + { + "epoch": 3.8768995905905212, + "grad_norm": 0.8684247136116028, + "learning_rate": 4.963365627033479e-07, + "loss": 0.2564, + "step": 55870 + }, + { + "epoch": 3.8769689820276176, + "grad_norm": 0.9065418243408203, + "learning_rate": 4.957776153999727e-07, + "loss": 0.2587, + "step": 55871 + }, + { + "epoch": 3.8770383734647145, + "grad_norm": 0.8887814283370972, + "learning_rate": 4.952189822199027e-07, + "loss": 0.2164, + "step": 55872 + }, + { + "epoch": 3.877107764901811, + "grad_norm": 1.1238901615142822, + "learning_rate": 4.946606631648587e-07, + "loss": 0.2731, + "step": 55873 + }, + { + "epoch": 3.8771771563389077, + "grad_norm": 0.9164473414421082, + "learning_rate": 4.941026582366393e-07, + "loss": 0.2834, + "step": 55874 + }, + { + "epoch": 3.8772465477760045, + "grad_norm": 0.6961625218391418, + "learning_rate": 4.935449674369875e-07, + "loss": 0.1267, + "step": 55875 + }, + { + "epoch": 3.8773159392131014, + "grad_norm": 0.8253269791603088, + "learning_rate": 4.929875907676685e-07, + "loss": 0.1657, + "step": 55876 + }, + { + "epoch": 3.8773853306501977, + "grad_norm": 1.1099393367767334, + "learning_rate": 4.924305282304475e-07, + "loss": 0.3389, + "step": 55877 + }, + { + "epoch": 3.8774547220872946, + "grad_norm": 1.0265840291976929, + "learning_rate": 4.918737798270789e-07, + "loss": 0.3972, + "step": 55878 + }, + { + "epoch": 3.877524113524391, + "grad_norm": 0.893835723400116, + "learning_rate": 4.913173455593167e-07, + "loss": 0.2138, + "step": 55879 + }, + { + "epoch": 3.877593504961488, + "grad_norm": 0.8278747200965881, + "learning_rate": 4.907612254289262e-07, + "loss": 0.2459, + "step": 55880 + }, + { + "epoch": 3.8776628963985846, + "grad_norm": 0.7194278836250305, + "learning_rate": 4.902054194376504e-07, + "loss": 0.2033, + "step": 55881 + }, + { + "epoch": 3.877732287835681, + "grad_norm": 1.092578411102295, + "learning_rate": 4.896499275872435e-07, + "loss": 0.2509, + "step": 55882 + }, + { + "epoch": 3.877801679272778, + "grad_norm": 0.7709886431694031, + "learning_rate": 4.890947498794817e-07, + "loss": 0.2372, + "step": 55883 + }, + { + "epoch": 3.8778710707098742, + "grad_norm": 1.0135247707366943, + "learning_rate": 4.885398863160973e-07, + "loss": 0.2242, + "step": 55884 + }, + { + "epoch": 3.877940462146971, + "grad_norm": 0.7328283190727234, + "learning_rate": 4.87985336898844e-07, + "loss": 0.158, + "step": 55885 + }, + { + "epoch": 3.878009853584068, + "grad_norm": 0.8875659108161926, + "learning_rate": 4.874311016294764e-07, + "loss": 0.273, + "step": 55886 + }, + { + "epoch": 3.8780792450211643, + "grad_norm": 1.0200964212417603, + "learning_rate": 4.868771805097483e-07, + "loss": 0.2436, + "step": 55887 + }, + { + "epoch": 3.878148636458261, + "grad_norm": 1.2600243091583252, + "learning_rate": 4.863235735413918e-07, + "loss": 0.3576, + "step": 55888 + }, + { + "epoch": 3.8782180278953575, + "grad_norm": 0.8289868235588074, + "learning_rate": 4.857702807261722e-07, + "loss": 0.2023, + "step": 55889 + }, + { + "epoch": 3.8782874193324544, + "grad_norm": 0.8526975512504578, + "learning_rate": 4.852173020658324e-07, + "loss": 0.2268, + "step": 55890 + }, + { + "epoch": 3.878356810769551, + "grad_norm": 0.9506742358207703, + "learning_rate": 4.846646375621156e-07, + "loss": 0.3209, + "step": 55891 + }, + { + "epoch": 3.8784262022066476, + "grad_norm": 0.8926920294761658, + "learning_rate": 4.841122872167647e-07, + "loss": 0.2393, + "step": 55892 + }, + { + "epoch": 3.8784955936437444, + "grad_norm": 0.8949624300003052, + "learning_rate": 4.83560251031534e-07, + "loss": 0.2995, + "step": 55893 + }, + { + "epoch": 3.878564985080841, + "grad_norm": 1.4545267820358276, + "learning_rate": 4.830085290081443e-07, + "loss": 0.1878, + "step": 55894 + }, + { + "epoch": 3.8786343765179376, + "grad_norm": 0.9893654584884644, + "learning_rate": 4.824571211483608e-07, + "loss": 0.1692, + "step": 55895 + }, + { + "epoch": 3.8787037679550345, + "grad_norm": 0.7947813868522644, + "learning_rate": 4.819060274539045e-07, + "loss": 0.2416, + "step": 55896 + }, + { + "epoch": 3.878773159392131, + "grad_norm": 0.9514596462249756, + "learning_rate": 4.813552479265294e-07, + "loss": 0.2875, + "step": 55897 + }, + { + "epoch": 3.8788425508292277, + "grad_norm": 1.1384860277175903, + "learning_rate": 4.808047825679674e-07, + "loss": 0.2836, + "step": 55898 + }, + { + "epoch": 3.878911942266324, + "grad_norm": 1.1703442335128784, + "learning_rate": 4.802546313799617e-07, + "loss": 0.2212, + "step": 55899 + }, + { + "epoch": 3.878981333703421, + "grad_norm": 0.5952149033546448, + "learning_rate": 4.797047943642441e-07, + "loss": 0.1381, + "step": 55900 + }, + { + "epoch": 3.8790507251405177, + "grad_norm": 0.6190743446350098, + "learning_rate": 4.791552715225467e-07, + "loss": 0.1611, + "step": 55901 + }, + { + "epoch": 3.8791201165776146, + "grad_norm": 1.0098072290420532, + "learning_rate": 4.786060628566236e-07, + "loss": 0.268, + "step": 55902 + }, + { + "epoch": 3.879189508014711, + "grad_norm": 1.7064292430877686, + "learning_rate": 4.780571683681844e-07, + "loss": 0.2647, + "step": 55903 + }, + { + "epoch": 3.879258899451808, + "grad_norm": 0.6620919704437256, + "learning_rate": 4.775085880589725e-07, + "loss": 0.1787, + "step": 55904 + }, + { + "epoch": 3.879328290888904, + "grad_norm": 0.8537572026252747, + "learning_rate": 4.769603219307195e-07, + "loss": 0.3227, + "step": 55905 + }, + { + "epoch": 3.879397682326001, + "grad_norm": 1.1781132221221924, + "learning_rate": 4.764123699851575e-07, + "loss": 0.2829, + "step": 55906 + }, + { + "epoch": 3.879467073763098, + "grad_norm": 0.6958587169647217, + "learning_rate": 4.758647322240184e-07, + "loss": 0.1843, + "step": 55907 + }, + { + "epoch": 3.8795364652001942, + "grad_norm": 1.19441556930542, + "learning_rate": 4.753174086490231e-07, + "loss": 0.2217, + "step": 55908 + }, + { + "epoch": 3.879605856637291, + "grad_norm": 0.695641815662384, + "learning_rate": 4.7477039926190346e-07, + "loss": 0.1562, + "step": 55909 + }, + { + "epoch": 3.8796752480743875, + "grad_norm": 0.8176726698875427, + "learning_rate": 4.7422370406438044e-07, + "loss": 0.1749, + "step": 55910 + }, + { + "epoch": 3.8797446395114843, + "grad_norm": 1.2308944463729858, + "learning_rate": 4.73677323058197e-07, + "loss": 0.2422, + "step": 55911 + }, + { + "epoch": 3.879814030948581, + "grad_norm": 0.6338667869567871, + "learning_rate": 4.7313125624506294e-07, + "loss": 0.2181, + "step": 55912 + }, + { + "epoch": 3.8798834223856775, + "grad_norm": 1.011927604675293, + "learning_rate": 4.7258550362669905e-07, + "loss": 0.2369, + "step": 55913 + }, + { + "epoch": 3.8799528138227743, + "grad_norm": 0.872565746307373, + "learning_rate": 4.7204006520484847e-07, + "loss": 0.2391, + "step": 55914 + }, + { + "epoch": 3.8800222052598707, + "grad_norm": 0.6798509359359741, + "learning_rate": 4.714949409812097e-07, + "loss": 0.2404, + "step": 55915 + }, + { + "epoch": 3.8800915966969676, + "grad_norm": 0.9125922918319702, + "learning_rate": 4.709501309575259e-07, + "loss": 0.1451, + "step": 55916 + }, + { + "epoch": 3.8801609881340644, + "grad_norm": 0.6790828704833984, + "learning_rate": 4.704056351354957e-07, + "loss": 0.1832, + "step": 55917 + }, + { + "epoch": 3.880230379571161, + "grad_norm": 0.9396604299545288, + "learning_rate": 4.6986145351683995e-07, + "loss": 0.1662, + "step": 55918 + }, + { + "epoch": 3.8802997710082576, + "grad_norm": 0.9393832087516785, + "learning_rate": 4.6931758610330165e-07, + "loss": 0.1637, + "step": 55919 + }, + { + "epoch": 3.880369162445354, + "grad_norm": 1.4357233047485352, + "learning_rate": 4.687740328965684e-07, + "loss": 0.3411, + "step": 55920 + }, + { + "epoch": 3.880438553882451, + "grad_norm": 0.9562806487083435, + "learning_rate": 4.682307938983721e-07, + "loss": 0.2636, + "step": 55921 + }, + { + "epoch": 3.8805079453195477, + "grad_norm": 0.8313521146774292, + "learning_rate": 4.676878691104225e-07, + "loss": 0.2233, + "step": 55922 + }, + { + "epoch": 3.880577336756644, + "grad_norm": 0.7142647504806519, + "learning_rate": 4.6714525853442935e-07, + "loss": 0.215, + "step": 55923 + }, + { + "epoch": 3.880646728193741, + "grad_norm": 0.9242008328437805, + "learning_rate": 4.666029621721135e-07, + "loss": 0.2116, + "step": 55924 + }, + { + "epoch": 3.8807161196308373, + "grad_norm": 1.2720023393630981, + "learning_rate": 4.660609800251847e-07, + "loss": 0.3642, + "step": 55925 + }, + { + "epoch": 3.880785511067934, + "grad_norm": 0.8305222988128662, + "learning_rate": 4.655193120953527e-07, + "loss": 0.2543, + "step": 55926 + }, + { + "epoch": 3.880854902505031, + "grad_norm": 0.972586989402771, + "learning_rate": 4.649779583843383e-07, + "loss": 0.1933, + "step": 55927 + }, + { + "epoch": 3.880924293942128, + "grad_norm": 0.5440709590911865, + "learning_rate": 4.644369188938291e-07, + "loss": 0.106, + "step": 55928 + }, + { + "epoch": 3.880993685379224, + "grad_norm": 0.8819732666015625, + "learning_rate": 4.6389619362554593e-07, + "loss": 0.2142, + "step": 55929 + }, + { + "epoch": 3.881063076816321, + "grad_norm": 0.7235203385353088, + "learning_rate": 4.6335578258119847e-07, + "loss": 0.2107, + "step": 55930 + }, + { + "epoch": 3.8811324682534174, + "grad_norm": 0.7609231472015381, + "learning_rate": 4.628156857624855e-07, + "loss": 0.1867, + "step": 55931 + }, + { + "epoch": 3.8812018596905142, + "grad_norm": 0.7543022632598877, + "learning_rate": 4.6227590317110547e-07, + "loss": 0.1847, + "step": 55932 + }, + { + "epoch": 3.881271251127611, + "grad_norm": 0.9539359211921692, + "learning_rate": 4.617364348087794e-07, + "loss": 0.2167, + "step": 55933 + }, + { + "epoch": 3.8813406425647075, + "grad_norm": 0.8277298808097839, + "learning_rate": 4.611972806772169e-07, + "loss": 0.2029, + "step": 55934 + }, + { + "epoch": 3.8814100340018043, + "grad_norm": 0.8318000435829163, + "learning_rate": 4.606584407780834e-07, + "loss": 0.2327, + "step": 55935 + }, + { + "epoch": 3.8814794254389007, + "grad_norm": 0.7213408350944519, + "learning_rate": 4.601199151131219e-07, + "loss": 0.2446, + "step": 55936 + }, + { + "epoch": 3.8815488168759975, + "grad_norm": 0.8307501077651978, + "learning_rate": 4.5958170368399777e-07, + "loss": 0.2217, + "step": 55937 + }, + { + "epoch": 3.8816182083130943, + "grad_norm": 1.2786792516708374, + "learning_rate": 4.590438064924318e-07, + "loss": 0.2955, + "step": 55938 + }, + { + "epoch": 3.8816875997501907, + "grad_norm": 1.0289803743362427, + "learning_rate": 4.585062235401116e-07, + "loss": 0.2648, + "step": 55939 + }, + { + "epoch": 3.8817569911872876, + "grad_norm": 0.8002169728279114, + "learning_rate": 4.579689548287469e-07, + "loss": 0.2042, + "step": 55940 + }, + { + "epoch": 3.881826382624384, + "grad_norm": 0.8805561661720276, + "learning_rate": 4.574320003600141e-07, + "loss": 0.2468, + "step": 55941 + }, + { + "epoch": 3.881895774061481, + "grad_norm": 0.6384586691856384, + "learning_rate": 4.5689536013563403e-07, + "loss": 0.2203, + "step": 55942 + }, + { + "epoch": 3.8819651654985776, + "grad_norm": 0.6836208701133728, + "learning_rate": 4.563590341572721e-07, + "loss": 0.1475, + "step": 55943 + }, + { + "epoch": 3.882034556935674, + "grad_norm": 0.6992855668067932, + "learning_rate": 4.558230224266491e-07, + "loss": 0.237, + "step": 55944 + }, + { + "epoch": 3.882103948372771, + "grad_norm": 0.9310255646705627, + "learning_rate": 4.5528732494543036e-07, + "loss": 0.1989, + "step": 55945 + }, + { + "epoch": 3.8821733398098672, + "grad_norm": 0.9243097305297852, + "learning_rate": 4.5475194171532564e-07, + "loss": 0.2275, + "step": 55946 + }, + { + "epoch": 3.882242731246964, + "grad_norm": 0.8896243572235107, + "learning_rate": 4.542168727380225e-07, + "loss": 0.1824, + "step": 55947 + }, + { + "epoch": 3.882312122684061, + "grad_norm": 0.7557735443115234, + "learning_rate": 4.5368211801519734e-07, + "loss": 0.1716, + "step": 55948 + }, + { + "epoch": 3.8823815141211573, + "grad_norm": 0.865259051322937, + "learning_rate": 4.531476775485599e-07, + "loss": 0.2255, + "step": 55949 + }, + { + "epoch": 3.882450905558254, + "grad_norm": 0.7710756659507751, + "learning_rate": 4.5261355133977556e-07, + "loss": 0.1943, + "step": 55950 + }, + { + "epoch": 3.8825202969953505, + "grad_norm": 0.8371801972389221, + "learning_rate": 4.520797393905429e-07, + "loss": 0.2292, + "step": 55951 + }, + { + "epoch": 3.8825896884324473, + "grad_norm": 0.6765573024749756, + "learning_rate": 4.515462417025496e-07, + "loss": 0.282, + "step": 55952 + }, + { + "epoch": 3.882659079869544, + "grad_norm": 0.9286421537399292, + "learning_rate": 4.510130582774719e-07, + "loss": 0.2106, + "step": 55953 + }, + { + "epoch": 3.882728471306641, + "grad_norm": 0.779001772403717, + "learning_rate": 4.504801891169974e-07, + "loss": 0.1932, + "step": 55954 + }, + { + "epoch": 3.8827978627437374, + "grad_norm": 1.0708273649215698, + "learning_rate": 4.4994763422281373e-07, + "loss": 0.2086, + "step": 55955 + }, + { + "epoch": 3.8828672541808342, + "grad_norm": 0.806293785572052, + "learning_rate": 4.4941539359658614e-07, + "loss": 0.2053, + "step": 55956 + }, + { + "epoch": 3.8829366456179306, + "grad_norm": 0.9149811267852783, + "learning_rate": 4.488834672400022e-07, + "loss": 0.2041, + "step": 55957 + }, + { + "epoch": 3.8830060370550274, + "grad_norm": 2.148555040359497, + "learning_rate": 4.4835185515474945e-07, + "loss": 0.2223, + "step": 55958 + }, + { + "epoch": 3.8830754284921243, + "grad_norm": 0.7164919376373291, + "learning_rate": 4.478205573425043e-07, + "loss": 0.2788, + "step": 55959 + }, + { + "epoch": 3.8831448199292207, + "grad_norm": 0.6545617580413818, + "learning_rate": 4.472895738049321e-07, + "loss": 0.1977, + "step": 55960 + }, + { + "epoch": 3.8832142113663175, + "grad_norm": 0.7183432579040527, + "learning_rate": 4.467589045437093e-07, + "loss": 0.2024, + "step": 55961 + }, + { + "epoch": 3.883283602803414, + "grad_norm": 0.8131530284881592, + "learning_rate": 4.462285495605345e-07, + "loss": 0.2, + "step": 55962 + }, + { + "epoch": 3.8833529942405107, + "grad_norm": 0.7147852778434753, + "learning_rate": 4.456985088570509e-07, + "loss": 0.2081, + "step": 55963 + }, + { + "epoch": 3.8834223856776076, + "grad_norm": 0.755236804485321, + "learning_rate": 4.4516878243495707e-07, + "loss": 0.1831, + "step": 55964 + }, + { + "epoch": 3.883491777114704, + "grad_norm": 1.1778639554977417, + "learning_rate": 4.446393702959073e-07, + "loss": 0.256, + "step": 55965 + }, + { + "epoch": 3.883561168551801, + "grad_norm": 1.2244806289672852, + "learning_rate": 4.441102724415891e-07, + "loss": 0.2329, + "step": 55966 + }, + { + "epoch": 3.883630559988897, + "grad_norm": 0.8832805752754211, + "learning_rate": 4.435814888736567e-07, + "loss": 0.2455, + "step": 55967 + }, + { + "epoch": 3.883699951425994, + "grad_norm": 1.2439640760421753, + "learning_rate": 4.4305301959378654e-07, + "loss": 0.2432, + "step": 55968 + }, + { + "epoch": 3.883769342863091, + "grad_norm": 0.7058325409889221, + "learning_rate": 4.4252486460365506e-07, + "loss": 0.1518, + "step": 55969 + }, + { + "epoch": 3.8838387343001872, + "grad_norm": 0.8266645669937134, + "learning_rate": 4.419970239049165e-07, + "loss": 0.1831, + "step": 55970 + }, + { + "epoch": 3.883908125737284, + "grad_norm": 0.7371298670768738, + "learning_rate": 4.414694974992473e-07, + "loss": 0.1935, + "step": 55971 + }, + { + "epoch": 3.8839775171743804, + "grad_norm": 0.7779441475868225, + "learning_rate": 4.409422853883016e-07, + "loss": 0.2204, + "step": 55972 + }, + { + "epoch": 3.8840469086114773, + "grad_norm": 1.0166324377059937, + "learning_rate": 4.40415387573756e-07, + "loss": 0.4268, + "step": 55973 + }, + { + "epoch": 3.884116300048574, + "grad_norm": 0.6745567917823792, + "learning_rate": 4.398888040572646e-07, + "loss": 0.1505, + "step": 55974 + }, + { + "epoch": 3.884185691485671, + "grad_norm": 0.9387059211730957, + "learning_rate": 4.393625348404928e-07, + "loss": 0.3392, + "step": 55975 + }, + { + "epoch": 3.8842550829227673, + "grad_norm": 0.9218236207962036, + "learning_rate": 4.388365799250949e-07, + "loss": 0.1778, + "step": 55976 + }, + { + "epoch": 3.8843244743598637, + "grad_norm": 0.6184720993041992, + "learning_rate": 4.3831093931274715e-07, + "loss": 0.1337, + "step": 55977 + }, + { + "epoch": 3.8843938657969606, + "grad_norm": 1.0416308641433716, + "learning_rate": 4.377856130051039e-07, + "loss": 0.3135, + "step": 55978 + }, + { + "epoch": 3.8844632572340574, + "grad_norm": 1.181915044784546, + "learning_rate": 4.3726060100381937e-07, + "loss": 0.2246, + "step": 55979 + }, + { + "epoch": 3.884532648671154, + "grad_norm": 0.8332251310348511, + "learning_rate": 4.367359033105478e-07, + "loss": 0.2918, + "step": 55980 + }, + { + "epoch": 3.8846020401082506, + "grad_norm": 0.8969279527664185, + "learning_rate": 4.3621151992694343e-07, + "loss": 0.2587, + "step": 55981 + }, + { + "epoch": 3.8846714315453474, + "grad_norm": 0.8639728426933289, + "learning_rate": 4.3568745085467156e-07, + "loss": 0.2739, + "step": 55982 + }, + { + "epoch": 3.884740822982444, + "grad_norm": 0.9613266587257385, + "learning_rate": 4.3516369609538645e-07, + "loss": 0.2887, + "step": 55983 + }, + { + "epoch": 3.8848102144195407, + "grad_norm": 1.1665823459625244, + "learning_rate": 4.346402556507423e-07, + "loss": 0.4714, + "step": 55984 + }, + { + "epoch": 3.8848796058566375, + "grad_norm": 0.8681078553199768, + "learning_rate": 4.3411712952238225e-07, + "loss": 0.3164, + "step": 55985 + }, + { + "epoch": 3.884948997293734, + "grad_norm": 0.7940236330032349, + "learning_rate": 4.3359431771196057e-07, + "loss": 0.2255, + "step": 55986 + }, + { + "epoch": 3.8850183887308307, + "grad_norm": 0.7265913486480713, + "learning_rate": 4.330718202211315e-07, + "loss": 0.3522, + "step": 55987 + }, + { + "epoch": 3.885087780167927, + "grad_norm": 0.9334302544593811, + "learning_rate": 4.3254963705153804e-07, + "loss": 0.3033, + "step": 55988 + }, + { + "epoch": 3.885157171605024, + "grad_norm": 0.6779069304466248, + "learning_rate": 4.320277682048346e-07, + "loss": 0.2716, + "step": 55989 + }, + { + "epoch": 3.8852265630421208, + "grad_norm": 1.0480554103851318, + "learning_rate": 4.315062136826753e-07, + "loss": 0.2352, + "step": 55990 + }, + { + "epoch": 3.885295954479217, + "grad_norm": 0.8738195896148682, + "learning_rate": 4.309849734866922e-07, + "loss": 0.2939, + "step": 55991 + }, + { + "epoch": 3.885365345916314, + "grad_norm": 1.224826455116272, + "learning_rate": 4.3046404761853953e-07, + "loss": 0.4338, + "step": 55992 + }, + { + "epoch": 3.8854347373534104, + "grad_norm": 0.7678879499435425, + "learning_rate": 4.299434360798604e-07, + "loss": 0.2239, + "step": 55993 + }, + { + "epoch": 3.885504128790507, + "grad_norm": 0.6822006702423096, + "learning_rate": 4.29423138872298e-07, + "loss": 0.1379, + "step": 55994 + }, + { + "epoch": 3.885573520227604, + "grad_norm": 0.8975771069526672, + "learning_rate": 4.289031559974954e-07, + "loss": 0.2501, + "step": 55995 + }, + { + "epoch": 3.8856429116647004, + "grad_norm": 1.0884060859680176, + "learning_rate": 4.283834874570847e-07, + "loss": 0.2462, + "step": 55996 + }, + { + "epoch": 3.8857123031017973, + "grad_norm": 1.1713964939117432, + "learning_rate": 4.2786413325272e-07, + "loss": 0.314, + "step": 55997 + }, + { + "epoch": 3.8857816945388937, + "grad_norm": 0.574475109577179, + "learning_rate": 4.2734509338604454e-07, + "loss": 0.1084, + "step": 55998 + }, + { + "epoch": 3.8858510859759905, + "grad_norm": 1.1308361291885376, + "learning_rate": 4.2682636785867925e-07, + "loss": 0.3202, + "step": 55999 + }, + { + "epoch": 3.8859204774130873, + "grad_norm": 0.6099843382835388, + "learning_rate": 4.263079566722672e-07, + "loss": 0.1563, + "step": 56000 + }, + { + "epoch": 3.8859204774130873, + "eval_loss": 0.46615734696388245, + "eval_runtime": 3859.4135, + "eval_samples_per_second": 6.289, + "eval_steps_per_second": 0.786, + "step": 56000 + }, + { + "epoch": 3.885989868850184, + "grad_norm": 0.7821841835975647, + "learning_rate": 4.2578985982845155e-07, + "loss": 0.2123, + "step": 56001 + }, + { + "epoch": 3.8860592602872805, + "grad_norm": 0.8068131804466248, + "learning_rate": 4.252720773288643e-07, + "loss": 0.1886, + "step": 56002 + }, + { + "epoch": 3.886128651724377, + "grad_norm": 1.23777174949646, + "learning_rate": 4.2475460917514866e-07, + "loss": 0.3228, + "step": 56003 + }, + { + "epoch": 3.8861980431614738, + "grad_norm": 0.7412698268890381, + "learning_rate": 4.2423745536891437e-07, + "loss": 0.2555, + "step": 56004 + }, + { + "epoch": 3.8862674345985706, + "grad_norm": 0.763307511806488, + "learning_rate": 4.237206159118268e-07, + "loss": 0.146, + "step": 56005 + }, + { + "epoch": 3.8863368260356674, + "grad_norm": 0.970374584197998, + "learning_rate": 4.232040908054846e-07, + "loss": 0.2414, + "step": 56006 + }, + { + "epoch": 3.886406217472764, + "grad_norm": 0.6377127170562744, + "learning_rate": 4.2268788005154213e-07, + "loss": 0.2136, + "step": 56007 + }, + { + "epoch": 3.8864756089098607, + "grad_norm": 0.641086995601654, + "learning_rate": 4.2217198365162024e-07, + "loss": 0.2479, + "step": 56008 + }, + { + "epoch": 3.886545000346957, + "grad_norm": 0.6739086508750916, + "learning_rate": 4.21656401607351e-07, + "loss": 0.1831, + "step": 56009 + }, + { + "epoch": 3.886614391784054, + "grad_norm": 1.223793625831604, + "learning_rate": 4.2114113392035525e-07, + "loss": 0.2326, + "step": 56010 + }, + { + "epoch": 3.8866837832211507, + "grad_norm": 0.992163896560669, + "learning_rate": 4.206261805922651e-07, + "loss": 0.3132, + "step": 56011 + }, + { + "epoch": 3.886753174658247, + "grad_norm": 0.9724528789520264, + "learning_rate": 4.2011154162470146e-07, + "loss": 0.2386, + "step": 56012 + }, + { + "epoch": 3.886822566095344, + "grad_norm": 0.7895357012748718, + "learning_rate": 4.1959721701929635e-07, + "loss": 0.183, + "step": 56013 + }, + { + "epoch": 3.8868919575324403, + "grad_norm": 1.0066715478897095, + "learning_rate": 4.1908320677767064e-07, + "loss": 0.3449, + "step": 56014 + }, + { + "epoch": 3.886961348969537, + "grad_norm": 0.9615362286567688, + "learning_rate": 4.1856951090143426e-07, + "loss": 0.1636, + "step": 56015 + }, + { + "epoch": 3.887030740406634, + "grad_norm": 1.964808702468872, + "learning_rate": 4.1805612939223027e-07, + "loss": 0.3313, + "step": 56016 + }, + { + "epoch": 3.8871001318437304, + "grad_norm": 1.0744662284851074, + "learning_rate": 4.175430622516685e-07, + "loss": 0.2525, + "step": 56017 + }, + { + "epoch": 3.887169523280827, + "grad_norm": 1.4110201597213745, + "learning_rate": 4.170303094813699e-07, + "loss": 0.3008, + "step": 56018 + }, + { + "epoch": 3.8872389147179236, + "grad_norm": 1.3846235275268555, + "learning_rate": 4.1651787108295537e-07, + "loss": 0.2073, + "step": 56019 + }, + { + "epoch": 3.8873083061550204, + "grad_norm": 0.7931117415428162, + "learning_rate": 4.1600574705803474e-07, + "loss": 0.2145, + "step": 56020 + }, + { + "epoch": 3.8873776975921173, + "grad_norm": 1.192976474761963, + "learning_rate": 4.1549393740822897e-07, + "loss": 0.2938, + "step": 56021 + }, + { + "epoch": 3.8874470890292137, + "grad_norm": 0.9579289555549622, + "learning_rate": 4.149824421351589e-07, + "loss": 0.347, + "step": 56022 + }, + { + "epoch": 3.8875164804663105, + "grad_norm": 0.8322994709014893, + "learning_rate": 4.1447126124043443e-07, + "loss": 0.184, + "step": 56023 + }, + { + "epoch": 3.887585871903407, + "grad_norm": 0.8470122814178467, + "learning_rate": 4.139603947256765e-07, + "loss": 0.1994, + "step": 56024 + }, + { + "epoch": 3.8876552633405037, + "grad_norm": 1.1500566005706787, + "learning_rate": 4.1344984259248377e-07, + "loss": 0.2369, + "step": 56025 + }, + { + "epoch": 3.8877246547776005, + "grad_norm": 0.8578557372093201, + "learning_rate": 4.1293960484247716e-07, + "loss": 0.2486, + "step": 56026 + }, + { + "epoch": 3.8877940462146974, + "grad_norm": 0.8683314919471741, + "learning_rate": 4.124296814772666e-07, + "loss": 0.1454, + "step": 56027 + }, + { + "epoch": 3.8878634376517938, + "grad_norm": 0.8521549105644226, + "learning_rate": 4.119200724984618e-07, + "loss": 0.1647, + "step": 56028 + }, + { + "epoch": 3.8879328290888906, + "grad_norm": 1.4508514404296875, + "learning_rate": 4.1141077790766146e-07, + "loss": 0.2791, + "step": 56029 + }, + { + "epoch": 3.888002220525987, + "grad_norm": 1.0154446363449097, + "learning_rate": 4.109017977064977e-07, + "loss": 0.1751, + "step": 56030 + }, + { + "epoch": 3.888071611963084, + "grad_norm": 0.7633887529373169, + "learning_rate": 4.1039313189655816e-07, + "loss": 0.2374, + "step": 56031 + }, + { + "epoch": 3.8881410034001807, + "grad_norm": 0.6503026485443115, + "learning_rate": 4.0988478047944144e-07, + "loss": 0.1767, + "step": 56032 + }, + { + "epoch": 3.888210394837277, + "grad_norm": 0.8114141225814819, + "learning_rate": 4.093767434567797e-07, + "loss": 0.253, + "step": 56033 + }, + { + "epoch": 3.888279786274374, + "grad_norm": 0.4732615053653717, + "learning_rate": 4.0886902083014933e-07, + "loss": 0.0938, + "step": 56034 + }, + { + "epoch": 3.8883491777114703, + "grad_norm": 0.9346365928649902, + "learning_rate": 4.083616126011713e-07, + "loss": 0.1904, + "step": 56035 + }, + { + "epoch": 3.888418569148567, + "grad_norm": 0.6858022212982178, + "learning_rate": 4.078545187714333e-07, + "loss": 0.1062, + "step": 56036 + }, + { + "epoch": 3.888487960585664, + "grad_norm": 0.6587725281715393, + "learning_rate": 4.073477393425562e-07, + "loss": 0.1755, + "step": 56037 + }, + { + "epoch": 3.8885573520227603, + "grad_norm": 0.9585922360420227, + "learning_rate": 4.068412743161276e-07, + "loss": 0.2614, + "step": 56038 + }, + { + "epoch": 3.888626743459857, + "grad_norm": 0.9697242379188538, + "learning_rate": 4.0633512369373514e-07, + "loss": 0.3367, + "step": 56039 + }, + { + "epoch": 3.8886961348969535, + "grad_norm": 0.8809716105461121, + "learning_rate": 4.058292874769887e-07, + "loss": 0.296, + "step": 56040 + }, + { + "epoch": 3.8887655263340504, + "grad_norm": 0.8060727119445801, + "learning_rate": 4.053237656674869e-07, + "loss": 0.299, + "step": 56041 + }, + { + "epoch": 3.888834917771147, + "grad_norm": 0.997525155544281, + "learning_rate": 4.0481855826682846e-07, + "loss": 0.2357, + "step": 56042 + }, + { + "epoch": 3.8889043092082436, + "grad_norm": 0.8284834623336792, + "learning_rate": 4.043136652766011e-07, + "loss": 0.1783, + "step": 56043 + }, + { + "epoch": 3.8889737006453404, + "grad_norm": 0.7928547263145447, + "learning_rate": 4.0380908669839236e-07, + "loss": 0.2475, + "step": 56044 + }, + { + "epoch": 3.889043092082437, + "grad_norm": 0.8341358304023743, + "learning_rate": 4.0330482253381206e-07, + "loss": 0.2207, + "step": 56045 + }, + { + "epoch": 3.8891124835195336, + "grad_norm": 0.5440618395805359, + "learning_rate": 4.028008727844368e-07, + "loss": 0.1321, + "step": 56046 + }, + { + "epoch": 3.8891818749566305, + "grad_norm": 0.9997164607048035, + "learning_rate": 4.022972374518652e-07, + "loss": 0.4001, + "step": 56047 + }, + { + "epoch": 3.889251266393727, + "grad_norm": 1.1065373420715332, + "learning_rate": 4.0179391653768494e-07, + "loss": 0.2207, + "step": 56048 + }, + { + "epoch": 3.8893206578308237, + "grad_norm": 1.2940572500228882, + "learning_rate": 4.0129091004348364e-07, + "loss": 0.2443, + "step": 56049 + }, + { + "epoch": 3.88939004926792, + "grad_norm": 1.000086784362793, + "learning_rate": 4.0078821797086e-07, + "loss": 0.2516, + "step": 56050 + }, + { + "epoch": 3.889459440705017, + "grad_norm": 0.898690938949585, + "learning_rate": 4.002858403213794e-07, + "loss": 0.3511, + "step": 56051 + }, + { + "epoch": 3.8895288321421138, + "grad_norm": 0.8164345622062683, + "learning_rate": 3.997837770966517e-07, + "loss": 0.2606, + "step": 56052 + }, + { + "epoch": 3.8895982235792106, + "grad_norm": 1.4656739234924316, + "learning_rate": 3.9928202829825344e-07, + "loss": 0.4001, + "step": 56053 + }, + { + "epoch": 3.889667615016307, + "grad_norm": 0.9761154055595398, + "learning_rate": 3.987805939277611e-07, + "loss": 0.2275, + "step": 56054 + }, + { + "epoch": 3.889737006453404, + "grad_norm": 0.9724193811416626, + "learning_rate": 3.9827947398675123e-07, + "loss": 0.2036, + "step": 56055 + }, + { + "epoch": 3.8898063978905, + "grad_norm": 0.8236521482467651, + "learning_rate": 3.9777866847683364e-07, + "loss": 0.2751, + "step": 56056 + }, + { + "epoch": 3.889875789327597, + "grad_norm": 0.9587699174880981, + "learning_rate": 3.972781773995626e-07, + "loss": 0.2982, + "step": 56057 + }, + { + "epoch": 3.889945180764694, + "grad_norm": 1.241531491279602, + "learning_rate": 3.9677800075653693e-07, + "loss": 0.2532, + "step": 56058 + }, + { + "epoch": 3.8900145722017903, + "grad_norm": 0.6868055462837219, + "learning_rate": 3.9627813854932196e-07, + "loss": 0.2323, + "step": 56059 + }, + { + "epoch": 3.890083963638887, + "grad_norm": 0.644579291343689, + "learning_rate": 3.957785907794942e-07, + "loss": 0.2628, + "step": 56060 + }, + { + "epoch": 3.8901533550759835, + "grad_norm": 1.0928754806518555, + "learning_rate": 3.952793574486524e-07, + "loss": 0.2001, + "step": 56061 + }, + { + "epoch": 3.8902227465130803, + "grad_norm": 1.0419672727584839, + "learning_rate": 3.947804385583398e-07, + "loss": 0.2292, + "step": 56062 + }, + { + "epoch": 3.890292137950177, + "grad_norm": 0.8567889332771301, + "learning_rate": 3.9428183411016615e-07, + "loss": 0.2237, + "step": 56063 + }, + { + "epoch": 3.8903615293872735, + "grad_norm": 1.374735951423645, + "learning_rate": 3.9378354410567473e-07, + "loss": 0.3404, + "step": 56064 + }, + { + "epoch": 3.8904309208243704, + "grad_norm": 0.8653237223625183, + "learning_rate": 3.9328556854645315e-07, + "loss": 0.2987, + "step": 56065 + }, + { + "epoch": 3.8905003122614668, + "grad_norm": 0.8202363848686218, + "learning_rate": 3.92787907434089e-07, + "loss": 0.2315, + "step": 56066 + }, + { + "epoch": 3.8905697036985636, + "grad_norm": 1.2377476692199707, + "learning_rate": 3.9229056077012547e-07, + "loss": 0.3784, + "step": 56067 + }, + { + "epoch": 3.8906390951356604, + "grad_norm": 0.9025752544403076, + "learning_rate": 3.9179352855613915e-07, + "loss": 0.4265, + "step": 56068 + }, + { + "epoch": 3.890708486572757, + "grad_norm": 0.7680599093437195, + "learning_rate": 3.912968107937065e-07, + "loss": 0.2088, + "step": 56069 + }, + { + "epoch": 3.8907778780098536, + "grad_norm": 0.9828600883483887, + "learning_rate": 3.9080040748439297e-07, + "loss": 0.2067, + "step": 56070 + }, + { + "epoch": 3.89084726944695, + "grad_norm": 1.0396276712417603, + "learning_rate": 3.903043186297639e-07, + "loss": 0.3947, + "step": 56071 + }, + { + "epoch": 3.890916660884047, + "grad_norm": 0.7083011269569397, + "learning_rate": 3.898085442313959e-07, + "loss": 0.1648, + "step": 56072 + }, + { + "epoch": 3.8909860523211437, + "grad_norm": 1.0163003206253052, + "learning_rate": 3.893130842908432e-07, + "loss": 0.284, + "step": 56073 + }, + { + "epoch": 3.89105544375824, + "grad_norm": 0.9000205397605896, + "learning_rate": 3.888179388096713e-07, + "loss": 0.3481, + "step": 56074 + }, + { + "epoch": 3.891124835195337, + "grad_norm": 0.695815920829773, + "learning_rate": 3.8832310778943446e-07, + "loss": 0.1571, + "step": 56075 + }, + { + "epoch": 3.8911942266324333, + "grad_norm": 1.037290096282959, + "learning_rate": 3.8782859123172034e-07, + "loss": 0.2068, + "step": 56076 + }, + { + "epoch": 3.89126361806953, + "grad_norm": 1.5509042739868164, + "learning_rate": 3.8733438913806097e-07, + "loss": 0.2692, + "step": 56077 + }, + { + "epoch": 3.891333009506627, + "grad_norm": 0.6836903095245361, + "learning_rate": 3.86840501510044e-07, + "loss": 0.2109, + "step": 56078 + }, + { + "epoch": 3.891402400943724, + "grad_norm": 1.0802562236785889, + "learning_rate": 3.863469283492016e-07, + "loss": 0.2918, + "step": 56079 + }, + { + "epoch": 3.89147179238082, + "grad_norm": 0.7932372689247131, + "learning_rate": 3.858536696571102e-07, + "loss": 0.2583, + "step": 56080 + }, + { + "epoch": 3.891541183817917, + "grad_norm": 0.9981545805931091, + "learning_rate": 3.853607254353131e-07, + "loss": 0.2555, + "step": 56081 + }, + { + "epoch": 3.8916105752550134, + "grad_norm": 0.983151376247406, + "learning_rate": 3.848680956853867e-07, + "loss": 0.1995, + "step": 56082 + }, + { + "epoch": 3.8916799666921102, + "grad_norm": 0.595227837562561, + "learning_rate": 3.843757804088632e-07, + "loss": 0.1618, + "step": 56083 + }, + { + "epoch": 3.891749358129207, + "grad_norm": 1.0153226852416992, + "learning_rate": 3.8388377960731915e-07, + "loss": 0.2788, + "step": 56084 + }, + { + "epoch": 3.8918187495663035, + "grad_norm": 0.7489423155784607, + "learning_rate": 3.8339209328228656e-07, + "loss": 0.2115, + "step": 56085 + }, + { + "epoch": 3.8918881410034003, + "grad_norm": 0.9846563339233398, + "learning_rate": 3.829007214353308e-07, + "loss": 0.2747, + "step": 56086 + }, + { + "epoch": 3.8919575324404967, + "grad_norm": 0.685814619064331, + "learning_rate": 3.8240966406800636e-07, + "loss": 0.1934, + "step": 56087 + }, + { + "epoch": 3.8920269238775935, + "grad_norm": 0.5797142386436462, + "learning_rate": 3.8191892118184524e-07, + "loss": 0.2215, + "step": 56088 + }, + { + "epoch": 3.8920963153146904, + "grad_norm": 0.6895716190338135, + "learning_rate": 3.814284927784129e-07, + "loss": 0.1694, + "step": 56089 + }, + { + "epoch": 3.8921657067517867, + "grad_norm": 1.3502277135849, + "learning_rate": 3.809383788592524e-07, + "loss": 0.2961, + "step": 56090 + }, + { + "epoch": 3.8922350981888836, + "grad_norm": 1.089820384979248, + "learning_rate": 3.8044857942591826e-07, + "loss": 0.2946, + "step": 56091 + }, + { + "epoch": 3.89230448962598, + "grad_norm": 0.8076882362365723, + "learning_rate": 3.799590944799425e-07, + "loss": 0.2763, + "step": 56092 + }, + { + "epoch": 3.892373881063077, + "grad_norm": 1.017497181892395, + "learning_rate": 3.794699240228794e-07, + "loss": 0.2304, + "step": 56093 + }, + { + "epoch": 3.8924432725001736, + "grad_norm": 0.7160163521766663, + "learning_rate": 3.789810680562722e-07, + "loss": 0.2746, + "step": 56094 + }, + { + "epoch": 3.89251266393727, + "grad_norm": 1.133230447769165, + "learning_rate": 3.7849252658166414e-07, + "loss": 0.2187, + "step": 56095 + }, + { + "epoch": 3.892582055374367, + "grad_norm": 1.1371220350265503, + "learning_rate": 3.7800429960059835e-07, + "loss": 0.3627, + "step": 56096 + }, + { + "epoch": 3.8926514468114632, + "grad_norm": 1.00388503074646, + "learning_rate": 3.7751638711460703e-07, + "loss": 0.2584, + "step": 56097 + }, + { + "epoch": 3.89272083824856, + "grad_norm": 0.9295024871826172, + "learning_rate": 3.7702878912524445e-07, + "loss": 0.2046, + "step": 56098 + }, + { + "epoch": 3.892790229685657, + "grad_norm": 0.6046764850616455, + "learning_rate": 3.7654150563404266e-07, + "loss": 0.1795, + "step": 56099 + }, + { + "epoch": 3.8928596211227533, + "grad_norm": 1.1099528074264526, + "learning_rate": 3.76054536642545e-07, + "loss": 0.3425, + "step": 56100 + }, + { + "epoch": 3.89292901255985, + "grad_norm": 0.7724878191947937, + "learning_rate": 3.7556788215228346e-07, + "loss": 0.2172, + "step": 56101 + }, + { + "epoch": 3.8929984039969465, + "grad_norm": 0.6433420777320862, + "learning_rate": 3.750815421647902e-07, + "loss": 0.1763, + "step": 56102 + }, + { + "epoch": 3.8930677954340434, + "grad_norm": 0.969740629196167, + "learning_rate": 3.7459551668160844e-07, + "loss": 0.2766, + "step": 56103 + }, + { + "epoch": 3.89313718687114, + "grad_norm": 1.041438341140747, + "learning_rate": 3.7410980570427024e-07, + "loss": 0.1733, + "step": 56104 + }, + { + "epoch": 3.893206578308237, + "grad_norm": 0.8227667212486267, + "learning_rate": 3.736244092343077e-07, + "loss": 0.3158, + "step": 56105 + }, + { + "epoch": 3.8932759697453334, + "grad_norm": 0.9350919127464294, + "learning_rate": 3.7313932727325305e-07, + "loss": 0.2374, + "step": 56106 + }, + { + "epoch": 3.8933453611824302, + "grad_norm": 1.0749804973602295, + "learning_rate": 3.726545598226494e-07, + "loss": 0.2467, + "step": 56107 + }, + { + "epoch": 3.8934147526195266, + "grad_norm": 0.7843742966651917, + "learning_rate": 3.7217010688400667e-07, + "loss": 0.0944, + "step": 56108 + }, + { + "epoch": 3.8934841440566235, + "grad_norm": 0.9951914548873901, + "learning_rate": 3.7168596845885696e-07, + "loss": 0.2378, + "step": 56109 + }, + { + "epoch": 3.8935535354937203, + "grad_norm": 0.912575900554657, + "learning_rate": 3.7120214454874345e-07, + "loss": 0.228, + "step": 56110 + }, + { + "epoch": 3.8936229269308167, + "grad_norm": 0.9768416285514832, + "learning_rate": 3.7071863515518723e-07, + "loss": 0.2812, + "step": 56111 + }, + { + "epoch": 3.8936923183679135, + "grad_norm": 1.1130459308624268, + "learning_rate": 3.702354402797092e-07, + "loss": 0.3694, + "step": 56112 + }, + { + "epoch": 3.89376170980501, + "grad_norm": 1.074495553970337, + "learning_rate": 3.6975255992384163e-07, + "loss": 0.2139, + "step": 56113 + }, + { + "epoch": 3.8938311012421067, + "grad_norm": 0.7333975434303284, + "learning_rate": 3.6926999408910533e-07, + "loss": 0.165, + "step": 56114 + }, + { + "epoch": 3.8939004926792036, + "grad_norm": 0.9318671226501465, + "learning_rate": 3.6878774277703253e-07, + "loss": 0.2394, + "step": 56115 + }, + { + "epoch": 3.8939698841163, + "grad_norm": 0.8001603484153748, + "learning_rate": 3.68305805989122e-07, + "loss": 0.2647, + "step": 56116 + }, + { + "epoch": 3.894039275553397, + "grad_norm": 0.8471692204475403, + "learning_rate": 3.6782418372692807e-07, + "loss": 0.1779, + "step": 56117 + }, + { + "epoch": 3.894108666990493, + "grad_norm": 0.9818302989006042, + "learning_rate": 3.6734287599193837e-07, + "loss": 0.3569, + "step": 56118 + }, + { + "epoch": 3.89417805842759, + "grad_norm": 1.280613660812378, + "learning_rate": 3.668618827856962e-07, + "loss": 0.2865, + "step": 56119 + }, + { + "epoch": 3.894247449864687, + "grad_norm": 1.0252314805984497, + "learning_rate": 3.6638120410970035e-07, + "loss": 0.3252, + "step": 56120 + }, + { + "epoch": 3.8943168413017832, + "grad_norm": 1.0698550939559937, + "learning_rate": 3.6590083996549393e-07, + "loss": 0.3208, + "step": 56121 + }, + { + "epoch": 3.89438623273888, + "grad_norm": 0.6543436646461487, + "learning_rate": 3.654207903545759e-07, + "loss": 0.1394, + "step": 56122 + }, + { + "epoch": 3.8944556241759765, + "grad_norm": 0.9936522841453552, + "learning_rate": 3.64941055278456e-07, + "loss": 0.2125, + "step": 56123 + }, + { + "epoch": 3.8945250156130733, + "grad_norm": 1.7522553205490112, + "learning_rate": 3.6446163473866643e-07, + "loss": 0.3834, + "step": 56124 + }, + { + "epoch": 3.89459440705017, + "grad_norm": 0.8233502507209778, + "learning_rate": 3.63982528736706e-07, + "loss": 0.2404, + "step": 56125 + }, + { + "epoch": 3.8946637984872665, + "grad_norm": 0.8184723258018494, + "learning_rate": 3.635037372740846e-07, + "loss": 0.2737, + "step": 56126 + }, + { + "epoch": 3.8947331899243633, + "grad_norm": 0.9552616477012634, + "learning_rate": 3.630252603523343e-07, + "loss": 0.1962, + "step": 56127 + }, + { + "epoch": 3.8948025813614597, + "grad_norm": 1.4089863300323486, + "learning_rate": 3.6254709797295396e-07, + "loss": 0.3883, + "step": 56128 + }, + { + "epoch": 3.8948719727985566, + "grad_norm": 1.3595274686813354, + "learning_rate": 3.6206925013745343e-07, + "loss": 0.2805, + "step": 56129 + }, + { + "epoch": 3.8949413642356534, + "grad_norm": 1.6327217817306519, + "learning_rate": 3.615917168473315e-07, + "loss": 0.3531, + "step": 56130 + }, + { + "epoch": 3.8950107556727502, + "grad_norm": 1.242441177368164, + "learning_rate": 3.6111449810410924e-07, + "loss": 0.2834, + "step": 56131 + }, + { + "epoch": 3.8950801471098466, + "grad_norm": 0.7874962687492371, + "learning_rate": 3.606375939092854e-07, + "loss": 0.177, + "step": 56132 + }, + { + "epoch": 3.8951495385469435, + "grad_norm": 0.9987038969993591, + "learning_rate": 3.6016100426436995e-07, + "loss": 0.235, + "step": 56133 + }, + { + "epoch": 3.89521892998404, + "grad_norm": 0.7793640494346619, + "learning_rate": 3.596847291708616e-07, + "loss": 0.2005, + "step": 56134 + }, + { + "epoch": 3.8952883214211367, + "grad_norm": 0.9444321393966675, + "learning_rate": 3.5920876863028143e-07, + "loss": 0.2068, + "step": 56135 + }, + { + "epoch": 3.8953577128582335, + "grad_norm": 1.1537890434265137, + "learning_rate": 3.5873312264410596e-07, + "loss": 0.3348, + "step": 56136 + }, + { + "epoch": 3.89542710429533, + "grad_norm": 0.7552225589752197, + "learning_rate": 3.582577912138563e-07, + "loss": 0.2619, + "step": 56137 + }, + { + "epoch": 3.8954964957324267, + "grad_norm": 1.3856407403945923, + "learning_rate": 3.5778277434102004e-07, + "loss": 0.4287, + "step": 56138 + }, + { + "epoch": 3.895565887169523, + "grad_norm": 1.3656049966812134, + "learning_rate": 3.5730807202710713e-07, + "loss": 0.2992, + "step": 56139 + }, + { + "epoch": 3.89563527860662, + "grad_norm": 0.7409583330154419, + "learning_rate": 3.568336842736164e-07, + "loss": 0.2473, + "step": 56140 + }, + { + "epoch": 3.895704670043717, + "grad_norm": 0.929979681968689, + "learning_rate": 3.563596110820355e-07, + "loss": 0.2511, + "step": 56141 + }, + { + "epoch": 3.895774061480813, + "grad_norm": 1.0759670734405518, + "learning_rate": 3.558858524538633e-07, + "loss": 0.3829, + "step": 56142 + }, + { + "epoch": 3.89584345291791, + "grad_norm": 0.7267866134643555, + "learning_rate": 3.5541240839060964e-07, + "loss": 0.2432, + "step": 56143 + }, + { + "epoch": 3.8959128443550064, + "grad_norm": 0.8551252484321594, + "learning_rate": 3.549392788937511e-07, + "loss": 0.2162, + "step": 56144 + }, + { + "epoch": 3.8959822357921032, + "grad_norm": 1.1539952754974365, + "learning_rate": 3.5446646396478657e-07, + "loss": 0.2942, + "step": 56145 + }, + { + "epoch": 3.8960516272292, + "grad_norm": 0.7617080211639404, + "learning_rate": 3.5399396360520365e-07, + "loss": 0.3051, + "step": 56146 + }, + { + "epoch": 3.8961210186662965, + "grad_norm": 0.8572301268577576, + "learning_rate": 3.535217778165123e-07, + "loss": 0.358, + "step": 56147 + }, + { + "epoch": 3.8961904101033933, + "grad_norm": 0.6590167284011841, + "learning_rate": 3.530499066001891e-07, + "loss": 0.2433, + "step": 56148 + }, + { + "epoch": 3.8962598015404897, + "grad_norm": 0.6878646016120911, + "learning_rate": 3.525783499577218e-07, + "loss": 0.1617, + "step": 56149 + }, + { + "epoch": 3.8963291929775865, + "grad_norm": 0.8405160307884216, + "learning_rate": 3.5210710789060907e-07, + "loss": 0.1811, + "step": 56150 + }, + { + "epoch": 3.8963985844146833, + "grad_norm": 1.0611190795898438, + "learning_rate": 3.5163618040033874e-07, + "loss": 0.203, + "step": 56151 + }, + { + "epoch": 3.8964679758517797, + "grad_norm": 0.7331975102424622, + "learning_rate": 3.511655674883763e-07, + "loss": 0.1528, + "step": 56152 + }, + { + "epoch": 3.8965373672888766, + "grad_norm": 0.9614686965942383, + "learning_rate": 3.506952691562315e-07, + "loss": 0.2287, + "step": 56153 + }, + { + "epoch": 3.896606758725973, + "grad_norm": 0.9553678035736084, + "learning_rate": 3.5022528540538116e-07, + "loss": 0.3261, + "step": 56154 + }, + { + "epoch": 3.89667615016307, + "grad_norm": 0.9940676093101501, + "learning_rate": 3.497556162373128e-07, + "loss": 0.36, + "step": 56155 + }, + { + "epoch": 3.8967455416001666, + "grad_norm": 1.0661296844482422, + "learning_rate": 3.4928626165349197e-07, + "loss": 0.2891, + "step": 56156 + }, + { + "epoch": 3.8968149330372635, + "grad_norm": 0.9471421837806702, + "learning_rate": 3.488172216554175e-07, + "loss": 0.2028, + "step": 56157 + }, + { + "epoch": 3.89688432447436, + "grad_norm": 0.8616228103637695, + "learning_rate": 3.48348496244566e-07, + "loss": 0.2271, + "step": 56158 + }, + { + "epoch": 3.8969537159114567, + "grad_norm": 0.6527584791183472, + "learning_rate": 3.47880085422414e-07, + "loss": 0.1916, + "step": 56159 + }, + { + "epoch": 3.897023107348553, + "grad_norm": 1.2608762979507446, + "learning_rate": 3.4741198919044924e-07, + "loss": 0.2185, + "step": 56160 + }, + { + "epoch": 3.89709249878565, + "grad_norm": 0.9806725382804871, + "learning_rate": 3.4694420755013723e-07, + "loss": 0.2576, + "step": 56161 + }, + { + "epoch": 3.8971618902227467, + "grad_norm": 0.894902765750885, + "learning_rate": 3.464767405029656e-07, + "loss": 0.3783, + "step": 56162 + }, + { + "epoch": 3.897231281659843, + "grad_norm": 0.9523651003837585, + "learning_rate": 3.4600958805039994e-07, + "loss": 0.1875, + "step": 56163 + }, + { + "epoch": 3.89730067309694, + "grad_norm": 1.0844554901123047, + "learning_rate": 3.455427501939168e-07, + "loss": 0.3123, + "step": 56164 + }, + { + "epoch": 3.8973700645340363, + "grad_norm": 0.661984384059906, + "learning_rate": 3.4507622693500387e-07, + "loss": 0.1639, + "step": 56165 + }, + { + "epoch": 3.897439455971133, + "grad_norm": 0.961739718914032, + "learning_rate": 3.4461001827511553e-07, + "loss": 0.2752, + "step": 56166 + }, + { + "epoch": 3.89750884740823, + "grad_norm": 0.6079230308532715, + "learning_rate": 3.441441242157284e-07, + "loss": 0.1223, + "step": 56167 + }, + { + "epoch": 3.8975782388453264, + "grad_norm": 0.86934494972229, + "learning_rate": 3.436785447583191e-07, + "loss": 0.2107, + "step": 56168 + }, + { + "epoch": 3.8976476302824232, + "grad_norm": 0.6109306812286377, + "learning_rate": 3.4321327990436416e-07, + "loss": 0.2334, + "step": 56169 + }, + { + "epoch": 3.8977170217195196, + "grad_norm": 0.9547168612480164, + "learning_rate": 3.427483296553069e-07, + "loss": 0.2882, + "step": 56170 + }, + { + "epoch": 3.8977864131566164, + "grad_norm": 0.8401981592178345, + "learning_rate": 3.4228369401264617e-07, + "loss": 0.2286, + "step": 56171 + }, + { + "epoch": 3.8978558045937133, + "grad_norm": 0.9983282685279846, + "learning_rate": 3.418193729778252e-07, + "loss": 0.3448, + "step": 56172 + }, + { + "epoch": 3.8979251960308097, + "grad_norm": 0.8145874738693237, + "learning_rate": 3.413553665523095e-07, + "loss": 0.1873, + "step": 56173 + }, + { + "epoch": 3.8979945874679065, + "grad_norm": 0.8969906568527222, + "learning_rate": 3.408916747375868e-07, + "loss": 0.2872, + "step": 56174 + }, + { + "epoch": 3.898063978905003, + "grad_norm": 0.9045701622962952, + "learning_rate": 3.404282975351003e-07, + "loss": 0.2031, + "step": 56175 + }, + { + "epoch": 3.8981333703420997, + "grad_norm": 0.9762121438980103, + "learning_rate": 3.399652349463156e-07, + "loss": 0.1462, + "step": 56176 + }, + { + "epoch": 3.8982027617791966, + "grad_norm": 0.6349328756332397, + "learning_rate": 3.3950248697270924e-07, + "loss": 0.176, + "step": 56177 + }, + { + "epoch": 3.898272153216293, + "grad_norm": 1.1386327743530273, + "learning_rate": 3.3904005361572453e-07, + "loss": 0.4796, + "step": 56178 + }, + { + "epoch": 3.89834154465339, + "grad_norm": 1.2944504022598267, + "learning_rate": 3.3857793487681587e-07, + "loss": 0.2982, + "step": 56179 + }, + { + "epoch": 3.898410936090486, + "grad_norm": 1.6491016149520874, + "learning_rate": 3.3811613075747097e-07, + "loss": 0.346, + "step": 56180 + }, + { + "epoch": 3.898480327527583, + "grad_norm": 0.8702815175056458, + "learning_rate": 3.3765464125912195e-07, + "loss": 0.2794, + "step": 56181 + }, + { + "epoch": 3.89854971896468, + "grad_norm": 1.1147109270095825, + "learning_rate": 3.3719346638323433e-07, + "loss": 0.3023, + "step": 56182 + }, + { + "epoch": 3.8986191104017767, + "grad_norm": 0.9464730620384216, + "learning_rate": 3.3673260613127364e-07, + "loss": 0.2504, + "step": 56183 + }, + { + "epoch": 3.898688501838873, + "grad_norm": 1.373763084411621, + "learning_rate": 3.3627206050467206e-07, + "loss": 0.3711, + "step": 56184 + }, + { + "epoch": 3.89875789327597, + "grad_norm": 0.6153469681739807, + "learning_rate": 3.358118295049062e-07, + "loss": 0.1486, + "step": 56185 + }, + { + "epoch": 3.8988272847130663, + "grad_norm": 0.7343494892120361, + "learning_rate": 3.3535191313340817e-07, + "loss": 0.2019, + "step": 56186 + }, + { + "epoch": 3.898896676150163, + "grad_norm": 1.0443923473358154, + "learning_rate": 3.3489231139165466e-07, + "loss": 0.1421, + "step": 56187 + }, + { + "epoch": 3.89896606758726, + "grad_norm": 1.0866214036941528, + "learning_rate": 3.3443302428106673e-07, + "loss": 0.2525, + "step": 56188 + }, + { + "epoch": 3.8990354590243563, + "grad_norm": 0.8540158271789551, + "learning_rate": 3.339740518031209e-07, + "loss": 0.2811, + "step": 56189 + }, + { + "epoch": 3.899104850461453, + "grad_norm": 0.7706876993179321, + "learning_rate": 3.335153939592606e-07, + "loss": 0.1684, + "step": 56190 + }, + { + "epoch": 3.8991742418985496, + "grad_norm": 1.054661750793457, + "learning_rate": 3.330570507509179e-07, + "loss": 0.3226, + "step": 56191 + }, + { + "epoch": 3.8992436333356464, + "grad_norm": 0.9636918902397156, + "learning_rate": 3.325990221795583e-07, + "loss": 0.1798, + "step": 56192 + }, + { + "epoch": 3.899313024772743, + "grad_norm": 1.0167869329452515, + "learning_rate": 3.3214130824661405e-07, + "loss": 0.2212, + "step": 56193 + }, + { + "epoch": 3.8993824162098396, + "grad_norm": 1.1692949533462524, + "learning_rate": 3.316839089535395e-07, + "loss": 0.2969, + "step": 56194 + }, + { + "epoch": 3.8994518076469364, + "grad_norm": 1.215030312538147, + "learning_rate": 3.3122682430176687e-07, + "loss": 0.2552, + "step": 56195 + }, + { + "epoch": 3.899521199084033, + "grad_norm": 0.9254948496818542, + "learning_rate": 3.3077005429276167e-07, + "loss": 0.2124, + "step": 56196 + }, + { + "epoch": 3.8995905905211297, + "grad_norm": 0.9453226923942566, + "learning_rate": 3.303135989279449e-07, + "loss": 0.238, + "step": 56197 + }, + { + "epoch": 3.8996599819582265, + "grad_norm": 0.9901356101036072, + "learning_rate": 3.298574582087599e-07, + "loss": 0.2349, + "step": 56198 + }, + { + "epoch": 3.899729373395323, + "grad_norm": 1.0683953762054443, + "learning_rate": 3.294016321366611e-07, + "loss": 0.2998, + "step": 56199 + }, + { + "epoch": 3.8997987648324197, + "grad_norm": 0.5917509198188782, + "learning_rate": 3.289461207130695e-07, + "loss": 0.1631, + "step": 56200 + }, + { + "epoch": 3.899868156269516, + "grad_norm": 1.8100550174713135, + "learning_rate": 3.284909239394285e-07, + "loss": 0.1506, + "step": 56201 + }, + { + "epoch": 3.899937547706613, + "grad_norm": 1.2345824241638184, + "learning_rate": 3.280360418171813e-07, + "loss": 0.3455, + "step": 56202 + }, + { + "epoch": 3.9000069391437098, + "grad_norm": 1.2033764123916626, + "learning_rate": 3.275814743477601e-07, + "loss": 0.3811, + "step": 56203 + }, + { + "epoch": 3.9000763305808066, + "grad_norm": 0.9081977009773254, + "learning_rate": 3.271272215326082e-07, + "loss": 0.2233, + "step": 56204 + }, + { + "epoch": 3.900145722017903, + "grad_norm": 0.9701923131942749, + "learning_rate": 3.2667328337314676e-07, + "loss": 0.2622, + "step": 56205 + }, + { + "epoch": 3.9002151134549994, + "grad_norm": 0.9113882184028625, + "learning_rate": 3.2621965987080784e-07, + "loss": 0.2635, + "step": 56206 + }, + { + "epoch": 3.900284504892096, + "grad_norm": 0.9180646538734436, + "learning_rate": 3.257663510270348e-07, + "loss": 0.277, + "step": 56207 + }, + { + "epoch": 3.900353896329193, + "grad_norm": 0.5670560598373413, + "learning_rate": 3.253133568432487e-07, + "loss": 0.1364, + "step": 56208 + }, + { + "epoch": 3.90042328776629, + "grad_norm": 1.1279380321502686, + "learning_rate": 3.248606773208818e-07, + "loss": 0.2508, + "step": 56209 + }, + { + "epoch": 3.9004926792033863, + "grad_norm": 0.9934288263320923, + "learning_rate": 3.2440831246136614e-07, + "loss": 0.2496, + "step": 56210 + }, + { + "epoch": 3.900562070640483, + "grad_norm": 0.9589166045188904, + "learning_rate": 3.239562622661341e-07, + "loss": 0.297, + "step": 56211 + }, + { + "epoch": 3.9006314620775795, + "grad_norm": 0.9292318224906921, + "learning_rate": 3.235045267366066e-07, + "loss": 0.2698, + "step": 56212 + }, + { + "epoch": 3.9007008535146763, + "grad_norm": 0.9664949774742126, + "learning_rate": 3.230531058742159e-07, + "loss": 0.2625, + "step": 56213 + }, + { + "epoch": 3.900770244951773, + "grad_norm": 0.826894998550415, + "learning_rate": 3.2260199968037195e-07, + "loss": 0.1697, + "step": 56214 + }, + { + "epoch": 3.9008396363888695, + "grad_norm": 1.3579660654067993, + "learning_rate": 3.22151208156507e-07, + "loss": 0.2587, + "step": 56215 + }, + { + "epoch": 3.9009090278259664, + "grad_norm": 0.8381237983703613, + "learning_rate": 3.217007313040532e-07, + "loss": 0.233, + "step": 56216 + }, + { + "epoch": 3.9009784192630628, + "grad_norm": 0.6835240721702576, + "learning_rate": 3.212505691244316e-07, + "loss": 0.2046, + "step": 56217 + }, + { + "epoch": 3.9010478107001596, + "grad_norm": 0.8470348119735718, + "learning_rate": 3.208007216190412e-07, + "loss": 0.2315, + "step": 56218 + }, + { + "epoch": 3.9011172021372564, + "grad_norm": 0.6657252907752991, + "learning_rate": 3.203511887893251e-07, + "loss": 0.2467, + "step": 56219 + }, + { + "epoch": 3.901186593574353, + "grad_norm": 1.1154696941375732, + "learning_rate": 3.199019706366935e-07, + "loss": 0.3092, + "step": 56220 + }, + { + "epoch": 3.9012559850114497, + "grad_norm": 0.9338440299034119, + "learning_rate": 3.1945306716257837e-07, + "loss": 0.2363, + "step": 56221 + }, + { + "epoch": 3.901325376448546, + "grad_norm": 1.042283296585083, + "learning_rate": 3.190044783683788e-07, + "loss": 0.2832, + "step": 56222 + }, + { + "epoch": 3.901394767885643, + "grad_norm": 0.6334726810455322, + "learning_rate": 3.185562042555046e-07, + "loss": 0.1521, + "step": 56223 + }, + { + "epoch": 3.9014641593227397, + "grad_norm": 1.0100644826889038, + "learning_rate": 3.1810824482539916e-07, + "loss": 0.2882, + "step": 56224 + }, + { + "epoch": 3.901533550759836, + "grad_norm": 1.0161892175674438, + "learning_rate": 3.176606000794502e-07, + "loss": 0.2334, + "step": 56225 + }, + { + "epoch": 3.901602942196933, + "grad_norm": 0.9782673120498657, + "learning_rate": 3.1721327001909e-07, + "loss": 0.2021, + "step": 56226 + }, + { + "epoch": 3.9016723336340293, + "grad_norm": 0.9924870729446411, + "learning_rate": 3.167662546457173e-07, + "loss": 0.3048, + "step": 56227 + }, + { + "epoch": 3.901741725071126, + "grad_norm": 1.002687931060791, + "learning_rate": 3.163195539607422e-07, + "loss": 0.254, + "step": 56228 + }, + { + "epoch": 3.901811116508223, + "grad_norm": 1.0122120380401611, + "learning_rate": 3.158731679655857e-07, + "loss": 0.2579, + "step": 56229 + }, + { + "epoch": 3.90188050794532, + "grad_norm": 0.9120526313781738, + "learning_rate": 3.1542709666165795e-07, + "loss": 0.297, + "step": 56230 + }, + { + "epoch": 3.901949899382416, + "grad_norm": 0.9272521734237671, + "learning_rate": 3.149813400503576e-07, + "loss": 0.2457, + "step": 56231 + }, + { + "epoch": 3.9020192908195126, + "grad_norm": 1.2056536674499512, + "learning_rate": 3.145358981330948e-07, + "loss": 0.3748, + "step": 56232 + }, + { + "epoch": 3.9020886822566094, + "grad_norm": 0.5936416983604431, + "learning_rate": 3.140907709112684e-07, + "loss": 0.1871, + "step": 56233 + }, + { + "epoch": 3.9021580736937063, + "grad_norm": 0.9351323246955872, + "learning_rate": 3.136459583862994e-07, + "loss": 0.2121, + "step": 56234 + }, + { + "epoch": 3.902227465130803, + "grad_norm": 0.908204197883606, + "learning_rate": 3.132014605595868e-07, + "loss": 0.2784, + "step": 56235 + }, + { + "epoch": 3.9022968565678995, + "grad_norm": 0.4869561791419983, + "learning_rate": 3.1275727743252935e-07, + "loss": 0.091, + "step": 56236 + }, + { + "epoch": 3.9023662480049963, + "grad_norm": 0.7781142592430115, + "learning_rate": 3.1231340900652607e-07, + "loss": 0.2462, + "step": 56237 + }, + { + "epoch": 3.9024356394420927, + "grad_norm": 0.9133692383766174, + "learning_rate": 3.1186985528299795e-07, + "loss": 0.1875, + "step": 56238 + }, + { + "epoch": 3.9025050308791895, + "grad_norm": 1.2059767246246338, + "learning_rate": 3.114266162633217e-07, + "loss": 0.3875, + "step": 56239 + }, + { + "epoch": 3.9025744223162864, + "grad_norm": 1.2601100206375122, + "learning_rate": 3.109836919488962e-07, + "loss": 0.2334, + "step": 56240 + }, + { + "epoch": 3.9026438137533828, + "grad_norm": 1.1649956703186035, + "learning_rate": 3.1054108234114255e-07, + "loss": 0.2053, + "step": 56241 + }, + { + "epoch": 3.9027132051904796, + "grad_norm": 1.0811570882797241, + "learning_rate": 3.100987874414374e-07, + "loss": 0.2164, + "step": 56242 + }, + { + "epoch": 3.902782596627576, + "grad_norm": 1.4000102281570435, + "learning_rate": 3.0965680725119074e-07, + "loss": 0.2721, + "step": 56243 + }, + { + "epoch": 3.902851988064673, + "grad_norm": 1.63154935836792, + "learning_rate": 3.092151417717792e-07, + "loss": 0.2884, + "step": 56244 + }, + { + "epoch": 3.9029213795017696, + "grad_norm": 0.8136187195777893, + "learning_rate": 3.08773791004624e-07, + "loss": 0.1798, + "step": 56245 + }, + { + "epoch": 3.902990770938866, + "grad_norm": 0.9690682888031006, + "learning_rate": 3.083327549510906e-07, + "loss": 0.2642, + "step": 56246 + }, + { + "epoch": 3.903060162375963, + "grad_norm": 0.791845440864563, + "learning_rate": 3.078920336126001e-07, + "loss": 0.2273, + "step": 56247 + }, + { + "epoch": 3.9031295538130593, + "grad_norm": 0.694419801235199, + "learning_rate": 3.074516269905181e-07, + "loss": 0.2486, + "step": 56248 + }, + { + "epoch": 3.903198945250156, + "grad_norm": 0.6974227428436279, + "learning_rate": 3.0701153508624347e-07, + "loss": 0.2274, + "step": 56249 + }, + { + "epoch": 3.903268336687253, + "grad_norm": 0.9741861820220947, + "learning_rate": 3.06571757901164e-07, + "loss": 0.2311, + "step": 56250 + }, + { + "epoch": 3.9033377281243493, + "grad_norm": 0.7751935720443726, + "learning_rate": 3.0613229543667853e-07, + "loss": 0.2136, + "step": 56251 + }, + { + "epoch": 3.903407119561446, + "grad_norm": 0.9294031858444214, + "learning_rate": 3.056931476941638e-07, + "loss": 0.2856, + "step": 56252 + }, + { + "epoch": 3.9034765109985425, + "grad_norm": 0.9222835898399353, + "learning_rate": 3.0525431467500753e-07, + "loss": 0.235, + "step": 56253 + }, + { + "epoch": 3.9035459024356394, + "grad_norm": 0.8165687918663025, + "learning_rate": 3.048157963805975e-07, + "loss": 0.2113, + "step": 56254 + }, + { + "epoch": 3.903615293872736, + "grad_norm": 0.7142754197120667, + "learning_rate": 3.0437759281232157e-07, + "loss": 0.1844, + "step": 56255 + }, + { + "epoch": 3.903684685309833, + "grad_norm": 0.7443076372146606, + "learning_rate": 3.0393970397155636e-07, + "loss": 0.2066, + "step": 56256 + }, + { + "epoch": 3.9037540767469294, + "grad_norm": 0.982338547706604, + "learning_rate": 3.0350212985967855e-07, + "loss": 0.2751, + "step": 56257 + }, + { + "epoch": 3.9038234681840263, + "grad_norm": 0.7874717116355896, + "learning_rate": 3.0306487047808694e-07, + "loss": 0.2675, + "step": 56258 + }, + { + "epoch": 3.9038928596211226, + "grad_norm": 0.6876092553138733, + "learning_rate": 3.0262792582815835e-07, + "loss": 0.196, + "step": 56259 + }, + { + "epoch": 3.9039622510582195, + "grad_norm": 0.6529186964035034, + "learning_rate": 3.021912959112583e-07, + "loss": 0.1504, + "step": 56260 + }, + { + "epoch": 3.9040316424953163, + "grad_norm": 0.958529531955719, + "learning_rate": 3.017549807287634e-07, + "loss": 0.3505, + "step": 56261 + }, + { + "epoch": 3.9041010339324127, + "grad_norm": 1.0072013139724731, + "learning_rate": 3.013189802820726e-07, + "loss": 0.3622, + "step": 56262 + }, + { + "epoch": 3.9041704253695095, + "grad_norm": 0.7532488107681274, + "learning_rate": 3.008832945725515e-07, + "loss": 0.1774, + "step": 56263 + }, + { + "epoch": 3.904239816806606, + "grad_norm": 1.0081958770751953, + "learning_rate": 3.004479236015656e-07, + "loss": 0.2813, + "step": 56264 + }, + { + "epoch": 3.9043092082437028, + "grad_norm": 0.7831618189811707, + "learning_rate": 3.000128673705027e-07, + "loss": 0.2046, + "step": 56265 + }, + { + "epoch": 3.9043785996807996, + "grad_norm": 0.723903238773346, + "learning_rate": 2.995781258807395e-07, + "loss": 0.2742, + "step": 56266 + }, + { + "epoch": 3.904447991117896, + "grad_norm": 1.4589935541152954, + "learning_rate": 2.991436991336416e-07, + "loss": 0.2868, + "step": 56267 + }, + { + "epoch": 3.904517382554993, + "grad_norm": 0.9904220104217529, + "learning_rate": 2.987095871305634e-07, + "loss": 0.2263, + "step": 56268 + }, + { + "epoch": 3.904586773992089, + "grad_norm": 0.9957239627838135, + "learning_rate": 2.9827578987290384e-07, + "loss": 0.2968, + "step": 56269 + }, + { + "epoch": 3.904656165429186, + "grad_norm": 1.140378475189209, + "learning_rate": 2.9784230736201736e-07, + "loss": 0.3064, + "step": 56270 + }, + { + "epoch": 3.904725556866283, + "grad_norm": 1.3146497011184692, + "learning_rate": 2.974091395992695e-07, + "loss": 0.3064, + "step": 56271 + }, + { + "epoch": 3.9047949483033793, + "grad_norm": 1.0530719757080078, + "learning_rate": 2.9697628658603703e-07, + "loss": 0.2995, + "step": 56272 + }, + { + "epoch": 3.904864339740476, + "grad_norm": 0.8297377228736877, + "learning_rate": 2.965437483236966e-07, + "loss": 0.2605, + "step": 56273 + }, + { + "epoch": 3.9049337311775725, + "grad_norm": 0.709181010723114, + "learning_rate": 2.961115248135804e-07, + "loss": 0.2329, + "step": 56274 + }, + { + "epoch": 3.9050031226146693, + "grad_norm": 0.936617910861969, + "learning_rate": 2.956796160570763e-07, + "loss": 0.3178, + "step": 56275 + }, + { + "epoch": 3.905072514051766, + "grad_norm": 0.735500156879425, + "learning_rate": 2.952480220555498e-07, + "loss": 0.2194, + "step": 56276 + }, + { + "epoch": 3.9051419054888625, + "grad_norm": 0.9910640120506287, + "learning_rate": 2.9481674281035544e-07, + "loss": 0.2344, + "step": 56277 + }, + { + "epoch": 3.9052112969259594, + "grad_norm": 1.6337534189224243, + "learning_rate": 2.9438577832285875e-07, + "loss": 0.3515, + "step": 56278 + }, + { + "epoch": 3.9052806883630558, + "grad_norm": 0.9197758436203003, + "learning_rate": 2.939551285944142e-07, + "loss": 0.31, + "step": 56279 + }, + { + "epoch": 3.9053500798001526, + "grad_norm": 0.8489149808883667, + "learning_rate": 2.935247936263874e-07, + "loss": 0.2215, + "step": 56280 + }, + { + "epoch": 3.9054194712372494, + "grad_norm": 1.055421233177185, + "learning_rate": 2.9309477342013283e-07, + "loss": 0.4387, + "step": 56281 + }, + { + "epoch": 3.9054888626743463, + "grad_norm": 1.1621675491333008, + "learning_rate": 2.9266506797701596e-07, + "loss": 0.3148, + "step": 56282 + }, + { + "epoch": 3.9055582541114426, + "grad_norm": 0.6568042635917664, + "learning_rate": 2.922356772983803e-07, + "loss": 0.2146, + "step": 56283 + }, + { + "epoch": 3.9056276455485395, + "grad_norm": 0.8516273498535156, + "learning_rate": 2.9180660138558027e-07, + "loss": 0.1808, + "step": 56284 + }, + { + "epoch": 3.905697036985636, + "grad_norm": 1.3549739122390747, + "learning_rate": 2.913778402399925e-07, + "loss": 0.1875, + "step": 56285 + }, + { + "epoch": 3.9057664284227327, + "grad_norm": 0.8692783713340759, + "learning_rate": 2.909493938629493e-07, + "loss": 0.3332, + "step": 56286 + }, + { + "epoch": 3.9058358198598295, + "grad_norm": 0.8168304562568665, + "learning_rate": 2.9052126225581624e-07, + "loss": 0.2606, + "step": 56287 + }, + { + "epoch": 3.905905211296926, + "grad_norm": 0.6488193869590759, + "learning_rate": 2.900934454199367e-07, + "loss": 0.1525, + "step": 56288 + }, + { + "epoch": 3.9059746027340227, + "grad_norm": 0.9413829445838928, + "learning_rate": 2.8966594335666506e-07, + "loss": 0.1854, + "step": 56289 + }, + { + "epoch": 3.906043994171119, + "grad_norm": 1.288785457611084, + "learning_rate": 2.8923875606735596e-07, + "loss": 0.3963, + "step": 56290 + }, + { + "epoch": 3.906113385608216, + "grad_norm": 1.1550767421722412, + "learning_rate": 2.8881188355334154e-07, + "loss": 0.2029, + "step": 56291 + }, + { + "epoch": 3.906182777045313, + "grad_norm": 0.8843160271644592, + "learning_rate": 2.883853258159763e-07, + "loss": 0.2339, + "step": 56292 + }, + { + "epoch": 3.906252168482409, + "grad_norm": 0.9924190044403076, + "learning_rate": 2.879590828566148e-07, + "loss": 0.2589, + "step": 56293 + }, + { + "epoch": 3.906321559919506, + "grad_norm": 0.6132408976554871, + "learning_rate": 2.8753315467660027e-07, + "loss": 0.1787, + "step": 56294 + }, + { + "epoch": 3.9063909513566024, + "grad_norm": 1.5104731321334839, + "learning_rate": 2.8710754127727614e-07, + "loss": 0.2143, + "step": 56295 + }, + { + "epoch": 3.9064603427936992, + "grad_norm": 0.9150873422622681, + "learning_rate": 2.8668224265997467e-07, + "loss": 0.2654, + "step": 56296 + }, + { + "epoch": 3.906529734230796, + "grad_norm": 1.0247433185577393, + "learning_rate": 2.8625725882606145e-07, + "loss": 0.1723, + "step": 56297 + }, + { + "epoch": 3.9065991256678925, + "grad_norm": 0.6362959146499634, + "learning_rate": 2.858325897768577e-07, + "loss": 0.2454, + "step": 56298 + }, + { + "epoch": 3.9066685171049893, + "grad_norm": 0.7118480801582336, + "learning_rate": 2.8540823551370665e-07, + "loss": 0.2101, + "step": 56299 + }, + { + "epoch": 3.9067379085420857, + "grad_norm": 0.8140109181404114, + "learning_rate": 2.849841960379629e-07, + "loss": 0.1788, + "step": 56300 + }, + { + "epoch": 3.9068072999791825, + "grad_norm": 0.7976267337799072, + "learning_rate": 2.845604713509475e-07, + "loss": 0.2365, + "step": 56301 + }, + { + "epoch": 3.9068766914162794, + "grad_norm": 0.9099182486534119, + "learning_rate": 2.84137061454004e-07, + "loss": 0.2624, + "step": 56302 + }, + { + "epoch": 3.9069460828533757, + "grad_norm": 0.9655677080154419, + "learning_rate": 2.837139663484756e-07, + "loss": 0.2484, + "step": 56303 + }, + { + "epoch": 3.9070154742904726, + "grad_norm": 1.121433138847351, + "learning_rate": 2.832911860356946e-07, + "loss": 0.3863, + "step": 56304 + }, + { + "epoch": 3.907084865727569, + "grad_norm": 0.7191218733787537, + "learning_rate": 2.828687205169933e-07, + "loss": 0.2234, + "step": 56305 + }, + { + "epoch": 3.907154257164666, + "grad_norm": 1.103912353515625, + "learning_rate": 2.82446569793704e-07, + "loss": 0.4257, + "step": 56306 + }, + { + "epoch": 3.9072236486017626, + "grad_norm": 1.408435344696045, + "learning_rate": 2.8202473386717e-07, + "loss": 0.4853, + "step": 56307 + }, + { + "epoch": 3.9072930400388595, + "grad_norm": 0.7155053019523621, + "learning_rate": 2.816032127387014e-07, + "loss": 0.1516, + "step": 56308 + }, + { + "epoch": 3.907362431475956, + "grad_norm": 1.230041742324829, + "learning_rate": 2.8118200640964155e-07, + "loss": 0.2306, + "step": 56309 + }, + { + "epoch": 3.9074318229130527, + "grad_norm": 0.5984331965446472, + "learning_rate": 2.807611148813338e-07, + "loss": 0.1944, + "step": 56310 + }, + { + "epoch": 3.907501214350149, + "grad_norm": 0.8474584221839905, + "learning_rate": 2.803405381550772e-07, + "loss": 0.2971, + "step": 56311 + }, + { + "epoch": 3.907570605787246, + "grad_norm": 0.8063851594924927, + "learning_rate": 2.799202762322262e-07, + "loss": 0.2369, + "step": 56312 + }, + { + "epoch": 3.9076399972243427, + "grad_norm": 0.8567660450935364, + "learning_rate": 2.7950032911410186e-07, + "loss": 0.1784, + "step": 56313 + }, + { + "epoch": 3.907709388661439, + "grad_norm": 0.9360913634300232, + "learning_rate": 2.7908069680201433e-07, + "loss": 0.1496, + "step": 56314 + }, + { + "epoch": 3.907778780098536, + "grad_norm": 0.9169263243675232, + "learning_rate": 2.7866137929730693e-07, + "loss": 0.2858, + "step": 56315 + }, + { + "epoch": 3.9078481715356324, + "grad_norm": 1.1539263725280762, + "learning_rate": 2.782423766012898e-07, + "loss": 0.3146, + "step": 56316 + }, + { + "epoch": 3.907917562972729, + "grad_norm": 0.9821920394897461, + "learning_rate": 2.778236887152952e-07, + "loss": 0.2426, + "step": 56317 + }, + { + "epoch": 3.907986954409826, + "grad_norm": 0.6536295413970947, + "learning_rate": 2.774053156406442e-07, + "loss": 0.2598, + "step": 56318 + }, + { + "epoch": 3.9080563458469224, + "grad_norm": 0.7400389909744263, + "learning_rate": 2.7698725737864696e-07, + "loss": 0.2354, + "step": 56319 + }, + { + "epoch": 3.9081257372840192, + "grad_norm": 0.8902008533477783, + "learning_rate": 2.7656951393063566e-07, + "loss": 0.2577, + "step": 56320 + }, + { + "epoch": 3.9081951287211156, + "grad_norm": 0.960178792476654, + "learning_rate": 2.761520852979316e-07, + "loss": 0.334, + "step": 56321 + }, + { + "epoch": 3.9082645201582125, + "grad_norm": 0.9119406342506409, + "learning_rate": 2.757349714818336e-07, + "loss": 0.2767, + "step": 56322 + }, + { + "epoch": 3.9083339115953093, + "grad_norm": 0.6200910806655884, + "learning_rate": 2.7531817248367396e-07, + "loss": 0.1187, + "step": 56323 + }, + { + "epoch": 3.9084033030324057, + "grad_norm": 0.7955770492553711, + "learning_rate": 2.749016883047628e-07, + "loss": 0.2708, + "step": 56324 + }, + { + "epoch": 3.9084726944695025, + "grad_norm": 0.9623458981513977, + "learning_rate": 2.7448551894642126e-07, + "loss": 0.2058, + "step": 56325 + }, + { + "epoch": 3.908542085906599, + "grad_norm": 1.0835994482040405, + "learning_rate": 2.740696644099594e-07, + "loss": 0.2836, + "step": 56326 + }, + { + "epoch": 3.9086114773436957, + "grad_norm": 0.6547645330429077, + "learning_rate": 2.736541246966873e-07, + "loss": 0.1573, + "step": 56327 + }, + { + "epoch": 3.9086808687807926, + "grad_norm": 0.7577893137931824, + "learning_rate": 2.7323889980791496e-07, + "loss": 0.2162, + "step": 56328 + }, + { + "epoch": 3.908750260217889, + "grad_norm": 0.7565094232559204, + "learning_rate": 2.728239897449636e-07, + "loss": 0.1894, + "step": 56329 + }, + { + "epoch": 3.908819651654986, + "grad_norm": 1.0848424434661865, + "learning_rate": 2.7240939450913214e-07, + "loss": 0.2455, + "step": 56330 + }, + { + "epoch": 3.908889043092082, + "grad_norm": 0.8312956094741821, + "learning_rate": 2.7199511410173073e-07, + "loss": 0.202, + "step": 56331 + }, + { + "epoch": 3.908958434529179, + "grad_norm": 0.9732490181922913, + "learning_rate": 2.715811485240693e-07, + "loss": 0.2199, + "step": 56332 + }, + { + "epoch": 3.909027825966276, + "grad_norm": 0.894615113735199, + "learning_rate": 2.7116749777745807e-07, + "loss": 0.2118, + "step": 56333 + }, + { + "epoch": 3.9090972174033727, + "grad_norm": 0.6543933749198914, + "learning_rate": 2.70754161863207e-07, + "loss": 0.1658, + "step": 56334 + }, + { + "epoch": 3.909166608840469, + "grad_norm": 0.6852511763572693, + "learning_rate": 2.7034114078261506e-07, + "loss": 0.1892, + "step": 56335 + }, + { + "epoch": 3.909236000277566, + "grad_norm": 0.7365608811378479, + "learning_rate": 2.6992843453697015e-07, + "loss": 0.2011, + "step": 56336 + }, + { + "epoch": 3.9093053917146623, + "grad_norm": 1.0955721139907837, + "learning_rate": 2.695160431276045e-07, + "loss": 0.2978, + "step": 56337 + }, + { + "epoch": 3.909374783151759, + "grad_norm": 0.9855559468269348, + "learning_rate": 2.6910396655579484e-07, + "loss": 0.3027, + "step": 56338 + }, + { + "epoch": 3.909444174588856, + "grad_norm": 1.8178876638412476, + "learning_rate": 2.686922048228624e-07, + "loss": 0.3816, + "step": 56339 + }, + { + "epoch": 3.9095135660259523, + "grad_norm": 0.7310680150985718, + "learning_rate": 2.68280757930095e-07, + "loss": 0.2004, + "step": 56340 + }, + { + "epoch": 3.909582957463049, + "grad_norm": 0.6416459083557129, + "learning_rate": 2.678696258787916e-07, + "loss": 0.1852, + "step": 56341 + }, + { + "epoch": 3.9096523489001456, + "grad_norm": 0.7737231850624084, + "learning_rate": 2.674588086702623e-07, + "loss": 0.2403, + "step": 56342 + }, + { + "epoch": 3.9097217403372424, + "grad_norm": 1.1106395721435547, + "learning_rate": 2.670483063057838e-07, + "loss": 0.2809, + "step": 56343 + }, + { + "epoch": 3.9097911317743392, + "grad_norm": 1.0730223655700684, + "learning_rate": 2.6663811878666625e-07, + "loss": 0.2712, + "step": 56344 + }, + { + "epoch": 3.9098605232114356, + "grad_norm": 0.7997673153877258, + "learning_rate": 2.6622824611419736e-07, + "loss": 0.2316, + "step": 56345 + }, + { + "epoch": 3.9099299146485325, + "grad_norm": 0.9981173276901245, + "learning_rate": 2.6581868828967625e-07, + "loss": 0.181, + "step": 56346 + }, + { + "epoch": 3.909999306085629, + "grad_norm": 0.687584638595581, + "learning_rate": 2.6540944531440184e-07, + "loss": 0.1738, + "step": 56347 + }, + { + "epoch": 3.9100686975227257, + "grad_norm": 0.967576265335083, + "learning_rate": 2.650005171896508e-07, + "loss": 0.2882, + "step": 56348 + }, + { + "epoch": 3.9101380889598225, + "grad_norm": 1.1302757263183594, + "learning_rate": 2.645919039167222e-07, + "loss": 0.1965, + "step": 56349 + }, + { + "epoch": 3.910207480396919, + "grad_norm": 1.1160959005355835, + "learning_rate": 2.641836054969038e-07, + "loss": 0.3711, + "step": 56350 + }, + { + "epoch": 3.9102768718340157, + "grad_norm": 0.8388429284095764, + "learning_rate": 2.637756219314946e-07, + "loss": 0.2002, + "step": 56351 + }, + { + "epoch": 3.910346263271112, + "grad_norm": 1.0595039129257202, + "learning_rate": 2.633679532217603e-07, + "loss": 0.2366, + "step": 56352 + }, + { + "epoch": 3.910415654708209, + "grad_norm": 0.8427596092224121, + "learning_rate": 2.62960599369011e-07, + "loss": 0.2396, + "step": 56353 + }, + { + "epoch": 3.910485046145306, + "grad_norm": 0.914185643196106, + "learning_rate": 2.6255356037452327e-07, + "loss": 0.1916, + "step": 56354 + }, + { + "epoch": 3.910554437582402, + "grad_norm": 0.9740790128707886, + "learning_rate": 2.6214683623958513e-07, + "loss": 0.2818, + "step": 56355 + }, + { + "epoch": 3.910623829019499, + "grad_norm": 0.7317238450050354, + "learning_rate": 2.6174042696547327e-07, + "loss": 0.2034, + "step": 56356 + }, + { + "epoch": 3.9106932204565954, + "grad_norm": 0.895991325378418, + "learning_rate": 2.613343325534756e-07, + "loss": 0.2623, + "step": 56357 + }, + { + "epoch": 3.9107626118936922, + "grad_norm": 0.7397224307060242, + "learning_rate": 2.6092855300487996e-07, + "loss": 0.1816, + "step": 56358 + }, + { + "epoch": 3.910832003330789, + "grad_norm": 0.9286419749259949, + "learning_rate": 2.6052308832095194e-07, + "loss": 0.2147, + "step": 56359 + }, + { + "epoch": 3.910901394767886, + "grad_norm": 0.7292764782905579, + "learning_rate": 2.601179385029795e-07, + "loss": 0.1869, + "step": 56360 + }, + { + "epoch": 3.9109707862049823, + "grad_norm": 0.7665177583694458, + "learning_rate": 2.597131035522504e-07, + "loss": 0.3045, + "step": 56361 + }, + { + "epoch": 3.911040177642079, + "grad_norm": 1.0224852561950684, + "learning_rate": 2.593085834700304e-07, + "loss": 0.354, + "step": 56362 + }, + { + "epoch": 3.9111095690791755, + "grad_norm": 0.7958087921142578, + "learning_rate": 2.5890437825760725e-07, + "loss": 0.1842, + "step": 56363 + }, + { + "epoch": 3.9111789605162723, + "grad_norm": 1.0983586311340332, + "learning_rate": 2.585004879162467e-07, + "loss": 0.2986, + "step": 56364 + }, + { + "epoch": 3.911248351953369, + "grad_norm": 0.8563225269317627, + "learning_rate": 2.5809691244722544e-07, + "loss": 0.1845, + "step": 56365 + }, + { + "epoch": 3.9113177433904656, + "grad_norm": 1.1595584154129028, + "learning_rate": 2.5769365185182025e-07, + "loss": 0.2834, + "step": 56366 + }, + { + "epoch": 3.9113871348275624, + "grad_norm": 0.8824114203453064, + "learning_rate": 2.572907061313079e-07, + "loss": 0.2652, + "step": 56367 + }, + { + "epoch": 3.911456526264659, + "grad_norm": 0.6132080554962158, + "learning_rate": 2.5688807528695403e-07, + "loss": 0.1676, + "step": 56368 + }, + { + "epoch": 3.9115259177017556, + "grad_norm": 0.8387066125869751, + "learning_rate": 2.5648575932002425e-07, + "loss": 0.2582, + "step": 56369 + }, + { + "epoch": 3.9115953091388524, + "grad_norm": 0.7921238541603088, + "learning_rate": 2.5608375823180653e-07, + "loss": 0.1062, + "step": 56370 + }, + { + "epoch": 3.911664700575949, + "grad_norm": 0.9617961645126343, + "learning_rate": 2.5568207202356644e-07, + "loss": 0.3965, + "step": 56371 + }, + { + "epoch": 3.9117340920130457, + "grad_norm": 1.0428928136825562, + "learning_rate": 2.552807006965474e-07, + "loss": 0.2172, + "step": 56372 + }, + { + "epoch": 3.911803483450142, + "grad_norm": 0.7234799265861511, + "learning_rate": 2.5487964425204847e-07, + "loss": 0.307, + "step": 56373 + }, + { + "epoch": 3.911872874887239, + "grad_norm": 0.8669964671134949, + "learning_rate": 2.54478902691313e-07, + "loss": 0.1684, + "step": 56374 + }, + { + "epoch": 3.9119422663243357, + "grad_norm": 0.8247870206832886, + "learning_rate": 2.5407847601562896e-07, + "loss": 0.2828, + "step": 56375 + }, + { + "epoch": 3.912011657761432, + "grad_norm": 0.8076925873756409, + "learning_rate": 2.536783642262286e-07, + "loss": 0.1786, + "step": 56376 + }, + { + "epoch": 3.912081049198529, + "grad_norm": 0.6334965229034424, + "learning_rate": 2.532785673244109e-07, + "loss": 0.1741, + "step": 56377 + }, + { + "epoch": 3.9121504406356253, + "grad_norm": 1.110129475593567, + "learning_rate": 2.5287908531140826e-07, + "loss": 0.2981, + "step": 56378 + }, + { + "epoch": 3.912219832072722, + "grad_norm": 0.9954756498336792, + "learning_rate": 2.5247991818849735e-07, + "loss": 0.2901, + "step": 56379 + }, + { + "epoch": 3.912289223509819, + "grad_norm": 1.0955593585968018, + "learning_rate": 2.5208106595693284e-07, + "loss": 0.2557, + "step": 56380 + }, + { + "epoch": 3.9123586149469154, + "grad_norm": 1.6239824295043945, + "learning_rate": 2.5168252861796914e-07, + "loss": 0.2751, + "step": 56381 + }, + { + "epoch": 3.9124280063840122, + "grad_norm": 1.1693264245986938, + "learning_rate": 2.5128430617288314e-07, + "loss": 0.3303, + "step": 56382 + }, + { + "epoch": 3.9124973978211086, + "grad_norm": 0.9410558938980103, + "learning_rate": 2.508863986229182e-07, + "loss": 0.3957, + "step": 56383 + }, + { + "epoch": 3.9125667892582054, + "grad_norm": 0.7405950427055359, + "learning_rate": 2.504888059693178e-07, + "loss": 0.2203, + "step": 56384 + }, + { + "epoch": 3.9126361806953023, + "grad_norm": 1.2328464984893799, + "learning_rate": 2.500915282133698e-07, + "loss": 0.3091, + "step": 56385 + }, + { + "epoch": 3.912705572132399, + "grad_norm": 0.808037519454956, + "learning_rate": 2.496945653562954e-07, + "loss": 0.198, + "step": 56386 + }, + { + "epoch": 3.9127749635694955, + "grad_norm": 0.9139724969863892, + "learning_rate": 2.492979173993715e-07, + "loss": 0.2945, + "step": 56387 + }, + { + "epoch": 3.9128443550065923, + "grad_norm": 1.1566715240478516, + "learning_rate": 2.4890158434383025e-07, + "loss": 0.3327, + "step": 56388 + }, + { + "epoch": 3.9129137464436887, + "grad_norm": 1.6776819229125977, + "learning_rate": 2.485055661909375e-07, + "loss": 0.467, + "step": 56389 + }, + { + "epoch": 3.9129831378807856, + "grad_norm": 1.4953724145889282, + "learning_rate": 2.481098629419476e-07, + "loss": 0.3058, + "step": 56390 + }, + { + "epoch": 3.9130525293178824, + "grad_norm": 1.0967695713043213, + "learning_rate": 2.4771447459809305e-07, + "loss": 0.3354, + "step": 56391 + }, + { + "epoch": 3.9131219207549788, + "grad_norm": 0.9131262302398682, + "learning_rate": 2.4731940116063945e-07, + "loss": 0.2283, + "step": 56392 + }, + { + "epoch": 3.9131913121920756, + "grad_norm": 0.5892130732536316, + "learning_rate": 2.4692464263081915e-07, + "loss": 0.1938, + "step": 56393 + }, + { + "epoch": 3.913260703629172, + "grad_norm": 0.8469977974891663, + "learning_rate": 2.465301990098756e-07, + "loss": 0.2978, + "step": 56394 + }, + { + "epoch": 3.913330095066269, + "grad_norm": 0.8946477770805359, + "learning_rate": 2.4613607029907446e-07, + "loss": 0.211, + "step": 56395 + }, + { + "epoch": 3.9133994865033657, + "grad_norm": 0.7067086100578308, + "learning_rate": 2.457422564996481e-07, + "loss": 0.2175, + "step": 56396 + }, + { + "epoch": 3.913468877940462, + "grad_norm": 0.8866575956344604, + "learning_rate": 2.4534875761283993e-07, + "loss": 0.2937, + "step": 56397 + }, + { + "epoch": 3.913538269377559, + "grad_norm": 0.7290319204330444, + "learning_rate": 2.4495557363988233e-07, + "loss": 0.2794, + "step": 56398 + }, + { + "epoch": 3.9136076608146553, + "grad_norm": 1.605936050415039, + "learning_rate": 2.445627045820409e-07, + "loss": 0.5354, + "step": 56399 + }, + { + "epoch": 3.913677052251752, + "grad_norm": 1.2654832601547241, + "learning_rate": 2.441701504405258e-07, + "loss": 0.2168, + "step": 56400 + }, + { + "epoch": 3.913746443688849, + "grad_norm": 0.9962652921676636, + "learning_rate": 2.4377791121660276e-07, + "loss": 0.3376, + "step": 56401 + }, + { + "epoch": 3.9138158351259453, + "grad_norm": 1.0908304452896118, + "learning_rate": 2.4338598691149296e-07, + "loss": 0.1422, + "step": 56402 + }, + { + "epoch": 3.913885226563042, + "grad_norm": 0.8431016206741333, + "learning_rate": 2.4299437752643985e-07, + "loss": 0.3036, + "step": 56403 + }, + { + "epoch": 3.9139546180001386, + "grad_norm": 1.2726815938949585, + "learning_rate": 2.4260308306268687e-07, + "loss": 0.197, + "step": 56404 + }, + { + "epoch": 3.9140240094372354, + "grad_norm": 1.2750556468963623, + "learning_rate": 2.422121035214553e-07, + "loss": 0.6079, + "step": 56405 + }, + { + "epoch": 3.914093400874332, + "grad_norm": 1.2703275680541992, + "learning_rate": 2.4182143890398856e-07, + "loss": 0.3599, + "step": 56406 + }, + { + "epoch": 3.9141627923114286, + "grad_norm": 1.2637677192687988, + "learning_rate": 2.4143108921150793e-07, + "loss": 0.3928, + "step": 56407 + }, + { + "epoch": 3.9142321837485254, + "grad_norm": 1.0325616598129272, + "learning_rate": 2.410410544452679e-07, + "loss": 0.2839, + "step": 56408 + }, + { + "epoch": 3.914301575185622, + "grad_norm": 0.9002541899681091, + "learning_rate": 2.406513346064787e-07, + "loss": 0.2654, + "step": 56409 + }, + { + "epoch": 3.9143709666227187, + "grad_norm": 0.7973004579544067, + "learning_rate": 2.402619296963837e-07, + "loss": 0.1888, + "step": 56410 + }, + { + "epoch": 3.9144403580598155, + "grad_norm": 0.8906163573265076, + "learning_rate": 2.3987283971621534e-07, + "loss": 0.2055, + "step": 56411 + }, + { + "epoch": 3.9145097494969123, + "grad_norm": 0.937517523765564, + "learning_rate": 2.3948406466718366e-07, + "loss": 0.3268, + "step": 56412 + }, + { + "epoch": 3.9145791409340087, + "grad_norm": 0.8699719905853271, + "learning_rate": 2.39095604550521e-07, + "loss": 0.1915, + "step": 56413 + }, + { + "epoch": 3.9146485323711055, + "grad_norm": 0.8718901872634888, + "learning_rate": 2.3870745936747094e-07, + "loss": 0.2822, + "step": 56414 + }, + { + "epoch": 3.914717923808202, + "grad_norm": 1.181151270866394, + "learning_rate": 2.3831962911924356e-07, + "loss": 0.3417, + "step": 56415 + }, + { + "epoch": 3.9147873152452988, + "grad_norm": 0.668610692024231, + "learning_rate": 2.3793211380706004e-07, + "loss": 0.1702, + "step": 56416 + }, + { + "epoch": 3.9148567066823956, + "grad_norm": 1.073417067527771, + "learning_rate": 2.375449134321528e-07, + "loss": 0.3804, + "step": 56417 + }, + { + "epoch": 3.914926098119492, + "grad_norm": 0.6770346760749817, + "learning_rate": 2.3715802799575415e-07, + "loss": 0.1732, + "step": 56418 + }, + { + "epoch": 3.914995489556589, + "grad_norm": 1.1491914987564087, + "learning_rate": 2.3677145749906314e-07, + "loss": 0.6052, + "step": 56419 + }, + { + "epoch": 3.915064880993685, + "grad_norm": 0.7955648303031921, + "learning_rate": 2.3638520194330104e-07, + "loss": 0.2539, + "step": 56420 + }, + { + "epoch": 3.915134272430782, + "grad_norm": 1.0360745191574097, + "learning_rate": 2.359992613297113e-07, + "loss": 0.2296, + "step": 56421 + }, + { + "epoch": 3.915203663867879, + "grad_norm": 0.9348540902137756, + "learning_rate": 2.356136356594929e-07, + "loss": 0.3465, + "step": 56422 + }, + { + "epoch": 3.9152730553049753, + "grad_norm": 0.7397494912147522, + "learning_rate": 2.3522832493386714e-07, + "loss": 0.2174, + "step": 56423 + }, + { + "epoch": 3.915342446742072, + "grad_norm": 0.8426457643508911, + "learning_rate": 2.3484332915404417e-07, + "loss": 0.2385, + "step": 56424 + }, + { + "epoch": 3.9154118381791685, + "grad_norm": 0.9087085723876953, + "learning_rate": 2.344586483212563e-07, + "loss": 0.2202, + "step": 56425 + }, + { + "epoch": 3.9154812296162653, + "grad_norm": 0.8015590906143188, + "learning_rate": 2.340742824367026e-07, + "loss": 0.1795, + "step": 56426 + }, + { + "epoch": 3.915550621053362, + "grad_norm": 0.8966382741928101, + "learning_rate": 2.336902315015932e-07, + "loss": 0.2295, + "step": 56427 + }, + { + "epoch": 3.9156200124904585, + "grad_norm": 0.8628377318382263, + "learning_rate": 2.3330649551716045e-07, + "loss": 0.2323, + "step": 56428 + }, + { + "epoch": 3.9156894039275554, + "grad_norm": 0.7383934855461121, + "learning_rate": 2.3292307448460338e-07, + "loss": 0.1377, + "step": 56429 + }, + { + "epoch": 3.9157587953646518, + "grad_norm": 0.7774084210395813, + "learning_rate": 2.3253996840513216e-07, + "loss": 0.1518, + "step": 56430 + }, + { + "epoch": 3.9158281868017486, + "grad_norm": 0.7459072470664978, + "learning_rate": 2.321571772799458e-07, + "loss": 0.1208, + "step": 56431 + }, + { + "epoch": 3.9158975782388454, + "grad_norm": 1.2672885656356812, + "learning_rate": 2.3177470111027667e-07, + "loss": 0.2696, + "step": 56432 + }, + { + "epoch": 3.915966969675942, + "grad_norm": 0.700706422328949, + "learning_rate": 2.313925398973127e-07, + "loss": 0.212, + "step": 56433 + }, + { + "epoch": 3.9160363611130387, + "grad_norm": 0.6069801449775696, + "learning_rate": 2.3101069364227512e-07, + "loss": 0.188, + "step": 56434 + }, + { + "epoch": 3.916105752550135, + "grad_norm": 0.7615882158279419, + "learning_rate": 2.306291623463519e-07, + "loss": 0.1639, + "step": 56435 + }, + { + "epoch": 3.916175143987232, + "grad_norm": 0.640145480632782, + "learning_rate": 2.3024794601076428e-07, + "loss": 0.1954, + "step": 56436 + }, + { + "epoch": 3.9162445354243287, + "grad_norm": 1.060491681098938, + "learning_rate": 2.2986704463671128e-07, + "loss": 0.2141, + "step": 56437 + }, + { + "epoch": 3.9163139268614255, + "grad_norm": 0.9247714281082153, + "learning_rate": 2.2948645822538084e-07, + "loss": 0.2727, + "step": 56438 + }, + { + "epoch": 3.916383318298522, + "grad_norm": 0.8012987971305847, + "learning_rate": 2.2910618677799423e-07, + "loss": 0.1849, + "step": 56439 + }, + { + "epoch": 3.9164527097356188, + "grad_norm": 1.0336002111434937, + "learning_rate": 2.2872623029573938e-07, + "loss": 0.2729, + "step": 56440 + }, + { + "epoch": 3.916522101172715, + "grad_norm": 0.636468768119812, + "learning_rate": 2.283465887798264e-07, + "loss": 0.1322, + "step": 56441 + }, + { + "epoch": 3.916591492609812, + "grad_norm": 0.9388958215713501, + "learning_rate": 2.2796726223144327e-07, + "loss": 0.2694, + "step": 56442 + }, + { + "epoch": 3.916660884046909, + "grad_norm": 1.0836974382400513, + "learning_rate": 2.2758825065178901e-07, + "loss": 0.191, + "step": 56443 + }, + { + "epoch": 3.916730275484005, + "grad_norm": 1.3240331411361694, + "learning_rate": 2.2720955404207377e-07, + "loss": 0.3908, + "step": 56444 + }, + { + "epoch": 3.916799666921102, + "grad_norm": 0.9915441870689392, + "learning_rate": 2.268311724034744e-07, + "loss": 0.3118, + "step": 56445 + }, + { + "epoch": 3.9168690583581984, + "grad_norm": 0.7479676008224487, + "learning_rate": 2.264531057371899e-07, + "loss": 0.1676, + "step": 56446 + }, + { + "epoch": 3.9169384497952953, + "grad_norm": 0.99481201171875, + "learning_rate": 2.2607535404441937e-07, + "loss": 0.2721, + "step": 56447 + }, + { + "epoch": 3.917007841232392, + "grad_norm": 1.2826545238494873, + "learning_rate": 2.256979173263618e-07, + "loss": 0.3945, + "step": 56448 + }, + { + "epoch": 3.9170772326694885, + "grad_norm": 1.1314759254455566, + "learning_rate": 2.2532079558418296e-07, + "loss": 0.3288, + "step": 56449 + }, + { + "epoch": 3.9171466241065853, + "grad_norm": 0.9875580668449402, + "learning_rate": 2.2494398881910406e-07, + "loss": 0.2661, + "step": 56450 + }, + { + "epoch": 3.9172160155436817, + "grad_norm": 0.857837438583374, + "learning_rate": 2.2456749703229086e-07, + "loss": 0.3211, + "step": 56451 + }, + { + "epoch": 3.9172854069807785, + "grad_norm": 0.5551295876502991, + "learning_rate": 2.2419132022495347e-07, + "loss": 0.1326, + "step": 56452 + }, + { + "epoch": 3.9173547984178754, + "grad_norm": 0.8243195414543152, + "learning_rate": 2.238154583982688e-07, + "loss": 0.2728, + "step": 56453 + }, + { + "epoch": 3.9174241898549718, + "grad_norm": 1.0439249277114868, + "learning_rate": 2.234399115534136e-07, + "loss": 0.2364, + "step": 56454 + }, + { + "epoch": 3.9174935812920686, + "grad_norm": 0.8357101678848267, + "learning_rate": 2.2306467969158696e-07, + "loss": 0.2021, + "step": 56455 + }, + { + "epoch": 3.917562972729165, + "grad_norm": 0.9256139397621155, + "learning_rate": 2.226897628139657e-07, + "loss": 0.2147, + "step": 56456 + }, + { + "epoch": 3.917632364166262, + "grad_norm": 0.8031517863273621, + "learning_rate": 2.2231516092173777e-07, + "loss": 0.2311, + "step": 56457 + }, + { + "epoch": 3.9177017556033586, + "grad_norm": 0.8644092082977295, + "learning_rate": 2.2194087401608e-07, + "loss": 0.2988, + "step": 56458 + }, + { + "epoch": 3.9177711470404555, + "grad_norm": 1.2627569437026978, + "learning_rate": 2.215669020981803e-07, + "loss": 0.3347, + "step": 56459 + }, + { + "epoch": 3.917840538477552, + "grad_norm": 0.6488372087478638, + "learning_rate": 2.2119324516922668e-07, + "loss": 0.2633, + "step": 56460 + }, + { + "epoch": 3.9179099299146483, + "grad_norm": 1.0060452222824097, + "learning_rate": 2.2081990323038482e-07, + "loss": 0.2844, + "step": 56461 + }, + { + "epoch": 3.917979321351745, + "grad_norm": 0.7632331848144531, + "learning_rate": 2.2044687628283157e-07, + "loss": 0.1732, + "step": 56462 + }, + { + "epoch": 3.918048712788842, + "grad_norm": 1.028523564338684, + "learning_rate": 2.2007416432775485e-07, + "loss": 0.225, + "step": 56463 + }, + { + "epoch": 3.9181181042259388, + "grad_norm": 0.8799155354499817, + "learning_rate": 2.1970176736633153e-07, + "loss": 0.2398, + "step": 56464 + }, + { + "epoch": 3.918187495663035, + "grad_norm": 0.7417190074920654, + "learning_rate": 2.193296853997273e-07, + "loss": 0.1704, + "step": 56465 + }, + { + "epoch": 3.918256887100132, + "grad_norm": 0.7434720396995544, + "learning_rate": 2.1895791842913017e-07, + "loss": 0.171, + "step": 56466 + }, + { + "epoch": 3.9183262785372284, + "grad_norm": 0.9208416938781738, + "learning_rate": 2.185864664556947e-07, + "loss": 0.2192, + "step": 56467 + }, + { + "epoch": 3.918395669974325, + "grad_norm": 1.1077172756195068, + "learning_rate": 2.1821532948061997e-07, + "loss": 0.2848, + "step": 56468 + }, + { + "epoch": 3.918465061411422, + "grad_norm": 0.9322451949119568, + "learning_rate": 2.1784450750504947e-07, + "loss": 0.3015, + "step": 56469 + }, + { + "epoch": 3.9185344528485184, + "grad_norm": 0.8285787105560303, + "learning_rate": 2.174740005301712e-07, + "loss": 0.1979, + "step": 56470 + }, + { + "epoch": 3.9186038442856153, + "grad_norm": 1.9204890727996826, + "learning_rate": 2.1710380855715085e-07, + "loss": 0.3168, + "step": 56471 + }, + { + "epoch": 3.9186732357227116, + "grad_norm": 1.0785582065582275, + "learning_rate": 2.1673393158715416e-07, + "loss": 0.2584, + "step": 56472 + }, + { + "epoch": 3.9187426271598085, + "grad_norm": 1.1562952995300293, + "learning_rate": 2.1636436962135797e-07, + "loss": 0.3848, + "step": 56473 + }, + { + "epoch": 3.9188120185969053, + "grad_norm": 1.1271531581878662, + "learning_rate": 2.1599512266091693e-07, + "loss": 0.2771, + "step": 56474 + }, + { + "epoch": 3.9188814100340017, + "grad_norm": 0.699138343334198, + "learning_rate": 2.1562619070700784e-07, + "loss": 0.2238, + "step": 56475 + }, + { + "epoch": 3.9189508014710985, + "grad_norm": 0.8211471438407898, + "learning_rate": 2.1525757376077426e-07, + "loss": 0.1792, + "step": 56476 + }, + { + "epoch": 3.919020192908195, + "grad_norm": 1.208993911743164, + "learning_rate": 2.1488927182341522e-07, + "loss": 0.3149, + "step": 56477 + }, + { + "epoch": 3.9190895843452918, + "grad_norm": 1.113314151763916, + "learning_rate": 2.1452128489606315e-07, + "loss": 0.2427, + "step": 56478 + }, + { + "epoch": 3.9191589757823886, + "grad_norm": 0.6763826012611389, + "learning_rate": 2.1415361297989488e-07, + "loss": 0.1361, + "step": 56479 + }, + { + "epoch": 3.919228367219485, + "grad_norm": 1.1944348812103271, + "learning_rate": 2.1378625607606505e-07, + "loss": 0.3302, + "step": 56480 + }, + { + "epoch": 3.919297758656582, + "grad_norm": 0.8991948962211609, + "learning_rate": 2.134192141857283e-07, + "loss": 0.1755, + "step": 56481 + }, + { + "epoch": 3.919367150093678, + "grad_norm": 0.7217617630958557, + "learning_rate": 2.1305248731005034e-07, + "loss": 0.2634, + "step": 56482 + }, + { + "epoch": 3.919436541530775, + "grad_norm": 1.3961135149002075, + "learning_rate": 2.126860754501969e-07, + "loss": 0.378, + "step": 56483 + }, + { + "epoch": 3.919505932967872, + "grad_norm": 0.8436673879623413, + "learning_rate": 2.1231997860731157e-07, + "loss": 0.2781, + "step": 56484 + }, + { + "epoch": 3.9195753244049687, + "grad_norm": 1.0653736591339111, + "learning_rate": 2.119541967825489e-07, + "loss": 0.2417, + "step": 56485 + }, + { + "epoch": 3.919644715842065, + "grad_norm": 0.7910282015800476, + "learning_rate": 2.1158872997707469e-07, + "loss": 0.2291, + "step": 56486 + }, + { + "epoch": 3.9197141072791615, + "grad_norm": 0.7541069984436035, + "learning_rate": 2.1122357819204354e-07, + "loss": 0.2692, + "step": 56487 + }, + { + "epoch": 3.9197834987162583, + "grad_norm": 0.8097547292709351, + "learning_rate": 2.1085874142859897e-07, + "loss": 0.2784, + "step": 56488 + }, + { + "epoch": 3.919852890153355, + "grad_norm": 0.8745982646942139, + "learning_rate": 2.1049421968788451e-07, + "loss": 0.2228, + "step": 56489 + }, + { + "epoch": 3.919922281590452, + "grad_norm": 0.8004868030548096, + "learning_rate": 2.1013001297107704e-07, + "loss": 0.2439, + "step": 56490 + }, + { + "epoch": 3.9199916730275484, + "grad_norm": 0.9504187703132629, + "learning_rate": 2.0976612127929785e-07, + "loss": 0.2479, + "step": 56491 + }, + { + "epoch": 3.920061064464645, + "grad_norm": 0.6932913661003113, + "learning_rate": 2.0940254461371267e-07, + "loss": 0.1753, + "step": 56492 + }, + { + "epoch": 3.9201304559017416, + "grad_norm": 0.9011749029159546, + "learning_rate": 2.0903928297547614e-07, + "loss": 0.327, + "step": 56493 + }, + { + "epoch": 3.9201998473388384, + "grad_norm": 1.07989501953125, + "learning_rate": 2.086763363657207e-07, + "loss": 0.2785, + "step": 56494 + }, + { + "epoch": 3.9202692387759352, + "grad_norm": 0.9488984942436218, + "learning_rate": 2.0831370478558987e-07, + "loss": 0.2787, + "step": 56495 + }, + { + "epoch": 3.9203386302130316, + "grad_norm": 0.7210641503334045, + "learning_rate": 2.0795138823623827e-07, + "loss": 0.2566, + "step": 56496 + }, + { + "epoch": 3.9204080216501285, + "grad_norm": 0.9828294515609741, + "learning_rate": 2.0758938671879834e-07, + "loss": 0.2834, + "step": 56497 + }, + { + "epoch": 3.920477413087225, + "grad_norm": 0.9655436873435974, + "learning_rate": 2.0722770023443583e-07, + "loss": 0.2898, + "step": 56498 + }, + { + "epoch": 3.9205468045243217, + "grad_norm": 0.6074063181877136, + "learning_rate": 2.0686632878427203e-07, + "loss": 0.17, + "step": 56499 + }, + { + "epoch": 3.9206161959614185, + "grad_norm": 0.8801295161247253, + "learning_rate": 2.065052723694505e-07, + "loss": 0.1894, + "step": 56500 + }, + { + "epoch": 3.920685587398515, + "grad_norm": 1.0611236095428467, + "learning_rate": 2.0614453099110366e-07, + "loss": 0.1827, + "step": 56501 + }, + { + "epoch": 3.9207549788356117, + "grad_norm": 1.3409627676010132, + "learning_rate": 2.057841046503972e-07, + "loss": 0.4031, + "step": 56502 + }, + { + "epoch": 3.920824370272708, + "grad_norm": 1.0361148118972778, + "learning_rate": 2.0542399334844143e-07, + "loss": 0.2517, + "step": 56503 + }, + { + "epoch": 3.920893761709805, + "grad_norm": 0.696245014667511, + "learning_rate": 2.0506419708639092e-07, + "loss": 0.3549, + "step": 56504 + }, + { + "epoch": 3.920963153146902, + "grad_norm": 0.8746050596237183, + "learning_rate": 2.0470471586536698e-07, + "loss": 0.2376, + "step": 56505 + }, + { + "epoch": 3.921032544583998, + "grad_norm": 1.0813268423080444, + "learning_rate": 2.0434554968651322e-07, + "loss": 0.2697, + "step": 56506 + }, + { + "epoch": 3.921101936021095, + "grad_norm": 0.4849586486816406, + "learning_rate": 2.0398669855097307e-07, + "loss": 0.1121, + "step": 56507 + }, + { + "epoch": 3.9211713274581914, + "grad_norm": 0.7734493017196655, + "learning_rate": 2.0362816245985684e-07, + "loss": 0.2037, + "step": 56508 + }, + { + "epoch": 3.9212407188952882, + "grad_norm": 1.3780579566955566, + "learning_rate": 2.03269941414308e-07, + "loss": 0.3633, + "step": 56509 + }, + { + "epoch": 3.921310110332385, + "grad_norm": 0.7463282942771912, + "learning_rate": 2.02912035415459e-07, + "loss": 0.1507, + "step": 56510 + }, + { + "epoch": 3.921379501769482, + "grad_norm": 0.7371051907539368, + "learning_rate": 2.0255444446444227e-07, + "loss": 0.255, + "step": 56511 + }, + { + "epoch": 3.9214488932065783, + "grad_norm": 0.9791058897972107, + "learning_rate": 2.0219716856237913e-07, + "loss": 0.3864, + "step": 56512 + }, + { + "epoch": 3.921518284643675, + "grad_norm": 1.0276070833206177, + "learning_rate": 2.0184020771040203e-07, + "loss": 0.3425, + "step": 56513 + }, + { + "epoch": 3.9215876760807715, + "grad_norm": 0.8860967755317688, + "learning_rate": 2.0148356190963226e-07, + "loss": 0.2243, + "step": 56514 + }, + { + "epoch": 3.9216570675178684, + "grad_norm": 0.7892023324966431, + "learning_rate": 2.0112723116121333e-07, + "loss": 0.2384, + "step": 56515 + }, + { + "epoch": 3.921726458954965, + "grad_norm": 1.1325926780700684, + "learning_rate": 2.0077121546624445e-07, + "loss": 0.2635, + "step": 56516 + }, + { + "epoch": 3.9217958503920616, + "grad_norm": 1.1600587368011475, + "learning_rate": 2.0041551482586906e-07, + "loss": 0.3255, + "step": 56517 + }, + { + "epoch": 3.9218652418291584, + "grad_norm": 1.0791270732879639, + "learning_rate": 2.0006012924119745e-07, + "loss": 0.2436, + "step": 56518 + }, + { + "epoch": 3.921934633266255, + "grad_norm": 1.317922592163086, + "learning_rate": 1.99705058713362e-07, + "loss": 0.2387, + "step": 56519 + }, + { + "epoch": 3.9220040247033516, + "grad_norm": 0.7464017868041992, + "learning_rate": 1.9935030324347293e-07, + "loss": 0.1352, + "step": 56520 + }, + { + "epoch": 3.9220734161404485, + "grad_norm": 0.6288790702819824, + "learning_rate": 1.989958628326516e-07, + "loss": 0.2167, + "step": 56521 + }, + { + "epoch": 3.922142807577545, + "grad_norm": 0.9263238310813904, + "learning_rate": 1.9864173748203042e-07, + "loss": 0.3788, + "step": 56522 + }, + { + "epoch": 3.9222121990146417, + "grad_norm": 0.7573798894882202, + "learning_rate": 1.982879271927196e-07, + "loss": 0.1867, + "step": 56523 + }, + { + "epoch": 3.922281590451738, + "grad_norm": 0.9100185036659241, + "learning_rate": 1.979344319658294e-07, + "loss": 0.2368, + "step": 56524 + }, + { + "epoch": 3.922350981888835, + "grad_norm": 0.755361795425415, + "learning_rate": 1.9758125180248112e-07, + "loss": 0.2243, + "step": 56525 + }, + { + "epoch": 3.9224203733259317, + "grad_norm": 1.4827162027359009, + "learning_rate": 1.9722838670379607e-07, + "loss": 0.3003, + "step": 56526 + }, + { + "epoch": 3.922489764763028, + "grad_norm": 0.755016028881073, + "learning_rate": 1.968758366708734e-07, + "loss": 0.2464, + "step": 56527 + }, + { + "epoch": 3.922559156200125, + "grad_norm": 0.73623126745224, + "learning_rate": 1.965236017048344e-07, + "loss": 0.2127, + "step": 56528 + }, + { + "epoch": 3.9226285476372214, + "grad_norm": 0.9146315455436707, + "learning_rate": 1.9617168180680044e-07, + "loss": 0.3247, + "step": 56529 + }, + { + "epoch": 3.922697939074318, + "grad_norm": 0.7464296221733093, + "learning_rate": 1.9582007697785954e-07, + "loss": 0.1434, + "step": 56530 + }, + { + "epoch": 3.922767330511415, + "grad_norm": 0.7754212617874146, + "learning_rate": 1.9546878721914408e-07, + "loss": 0.2358, + "step": 56531 + }, + { + "epoch": 3.9228367219485114, + "grad_norm": 0.8281263709068298, + "learning_rate": 1.951178125317532e-07, + "loss": 0.2472, + "step": 56532 + }, + { + "epoch": 3.9229061133856082, + "grad_norm": 0.8161576986312866, + "learning_rate": 1.9476715291679716e-07, + "loss": 0.1573, + "step": 56533 + }, + { + "epoch": 3.9229755048227046, + "grad_norm": 1.3152731657028198, + "learning_rate": 1.944168083753861e-07, + "loss": 0.1454, + "step": 56534 + }, + { + "epoch": 3.9230448962598015, + "grad_norm": 0.9658110737800598, + "learning_rate": 1.9406677890860814e-07, + "loss": 0.2724, + "step": 56535 + }, + { + "epoch": 3.9231142876968983, + "grad_norm": 1.1598474979400635, + "learning_rate": 1.9371706451759564e-07, + "loss": 0.2625, + "step": 56536 + }, + { + "epoch": 3.923183679133995, + "grad_norm": 1.0099502801895142, + "learning_rate": 1.9336766520342554e-07, + "loss": 0.3, + "step": 56537 + }, + { + "epoch": 3.9232530705710915, + "grad_norm": 0.9162040948867798, + "learning_rate": 1.9301858096723023e-07, + "loss": 0.2762, + "step": 56538 + }, + { + "epoch": 3.9233224620081883, + "grad_norm": 1.1423176527023315, + "learning_rate": 1.926698118100867e-07, + "loss": 0.2338, + "step": 56539 + }, + { + "epoch": 3.9233918534452847, + "grad_norm": 0.835501492023468, + "learning_rate": 1.92321357733094e-07, + "loss": 0.1769, + "step": 56540 + }, + { + "epoch": 3.9234612448823816, + "grad_norm": 0.839754581451416, + "learning_rate": 1.919732187373735e-07, + "loss": 0.22, + "step": 56541 + }, + { + "epoch": 3.9235306363194784, + "grad_norm": 0.947597086429596, + "learning_rate": 1.916253948240132e-07, + "loss": 0.2513, + "step": 56542 + }, + { + "epoch": 3.923600027756575, + "grad_norm": 0.8985292315483093, + "learning_rate": 1.9127788599411223e-07, + "loss": 0.3181, + "step": 56543 + }, + { + "epoch": 3.9236694191936716, + "grad_norm": 1.1786975860595703, + "learning_rate": 1.9093069224876968e-07, + "loss": 0.3156, + "step": 56544 + }, + { + "epoch": 3.923738810630768, + "grad_norm": 0.9133560061454773, + "learning_rate": 1.9058381358907363e-07, + "loss": 0.1273, + "step": 56545 + }, + { + "epoch": 3.923808202067865, + "grad_norm": 0.9337758421897888, + "learning_rate": 1.9023725001612313e-07, + "loss": 0.256, + "step": 56546 + }, + { + "epoch": 3.9238775935049617, + "grad_norm": 0.8200876712799072, + "learning_rate": 1.8989100153101736e-07, + "loss": 0.2008, + "step": 56547 + }, + { + "epoch": 3.923946984942058, + "grad_norm": 0.9259414076805115, + "learning_rate": 1.8954506813484429e-07, + "loss": 0.412, + "step": 56548 + }, + { + "epoch": 3.924016376379155, + "grad_norm": 0.938166081905365, + "learning_rate": 1.891994498287031e-07, + "loss": 0.2705, + "step": 56549 + }, + { + "epoch": 3.9240857678162513, + "grad_norm": 1.0950586795806885, + "learning_rate": 1.888541466136706e-07, + "loss": 0.2138, + "step": 56550 + }, + { + "epoch": 3.924155159253348, + "grad_norm": 0.8746383190155029, + "learning_rate": 1.8850915849085716e-07, + "loss": 0.2667, + "step": 56551 + }, + { + "epoch": 3.924224550690445, + "grad_norm": 1.1155296564102173, + "learning_rate": 1.8816448546132848e-07, + "loss": 0.2645, + "step": 56552 + }, + { + "epoch": 3.9242939421275413, + "grad_norm": 0.591846227645874, + "learning_rate": 1.8782012752618373e-07, + "loss": 0.1203, + "step": 56553 + }, + { + "epoch": 3.924363333564638, + "grad_norm": 1.3582277297973633, + "learning_rate": 1.874760846865109e-07, + "loss": 0.2825, + "step": 56554 + }, + { + "epoch": 3.9244327250017346, + "grad_norm": 0.7359658479690552, + "learning_rate": 1.8713235694339804e-07, + "loss": 0.1743, + "step": 56555 + }, + { + "epoch": 3.9245021164388314, + "grad_norm": 0.7297472357749939, + "learning_rate": 1.8678894429793315e-07, + "loss": 0.1763, + "step": 56556 + }, + { + "epoch": 3.9245715078759282, + "grad_norm": 0.7798914313316345, + "learning_rate": 1.8644584675119315e-07, + "loss": 0.2235, + "step": 56557 + }, + { + "epoch": 3.9246408993130246, + "grad_norm": 1.0292950868606567, + "learning_rate": 1.8610306430425495e-07, + "loss": 0.2433, + "step": 56558 + }, + { + "epoch": 3.9247102907501215, + "grad_norm": 1.3586583137512207, + "learning_rate": 1.8576059695821768e-07, + "loss": 0.4551, + "step": 56559 + }, + { + "epoch": 3.924779682187218, + "grad_norm": 1.0064826011657715, + "learning_rate": 1.8541844471414716e-07, + "loss": 0.3861, + "step": 56560 + }, + { + "epoch": 3.9248490736243147, + "grad_norm": 0.7622839212417603, + "learning_rate": 1.8507660757313138e-07, + "loss": 0.1956, + "step": 56561 + }, + { + "epoch": 3.9249184650614115, + "grad_norm": 0.8887988924980164, + "learning_rate": 1.847350855362584e-07, + "loss": 0.1294, + "step": 56562 + }, + { + "epoch": 3.9249878564985083, + "grad_norm": 0.8598105907440186, + "learning_rate": 1.843938786045829e-07, + "loss": 0.2308, + "step": 56563 + }, + { + "epoch": 3.9250572479356047, + "grad_norm": 0.9559141397476196, + "learning_rate": 1.8405298677920402e-07, + "loss": 0.2798, + "step": 56564 + }, + { + "epoch": 3.9251266393727016, + "grad_norm": 0.7352197170257568, + "learning_rate": 1.8371241006118755e-07, + "loss": 0.2308, + "step": 56565 + }, + { + "epoch": 3.925196030809798, + "grad_norm": 1.5256731510162354, + "learning_rate": 1.8337214845159934e-07, + "loss": 0.3094, + "step": 56566 + }, + { + "epoch": 3.925265422246895, + "grad_norm": 0.9605564475059509, + "learning_rate": 1.8303220195153848e-07, + "loss": 0.2808, + "step": 56567 + }, + { + "epoch": 3.9253348136839916, + "grad_norm": 0.7566048502922058, + "learning_rate": 1.826925705620486e-07, + "loss": 0.1535, + "step": 56568 + }, + { + "epoch": 3.925404205121088, + "grad_norm": 0.871092677116394, + "learning_rate": 1.823532542842399e-07, + "loss": 0.2744, + "step": 56569 + }, + { + "epoch": 3.925473596558185, + "grad_norm": 0.8183643221855164, + "learning_rate": 1.820142531191449e-07, + "loss": 0.2598, + "step": 56570 + }, + { + "epoch": 3.9255429879952812, + "grad_norm": 0.7998250126838684, + "learning_rate": 1.816755670678516e-07, + "loss": 0.2237, + "step": 56571 + }, + { + "epoch": 3.925612379432378, + "grad_norm": 0.594132661819458, + "learning_rate": 1.8133719613142584e-07, + "loss": 0.0868, + "step": 56572 + }, + { + "epoch": 3.925681770869475, + "grad_norm": 0.8920541405677795, + "learning_rate": 1.8099914031094455e-07, + "loss": 0.2524, + "step": 56573 + }, + { + "epoch": 3.9257511623065713, + "grad_norm": 0.9686788320541382, + "learning_rate": 1.8066139960746243e-07, + "loss": 0.1713, + "step": 56574 + }, + { + "epoch": 3.925820553743668, + "grad_norm": 0.9126613736152649, + "learning_rate": 1.8032397402205637e-07, + "loss": 0.1408, + "step": 56575 + }, + { + "epoch": 3.9258899451807645, + "grad_norm": 0.6777435541152954, + "learning_rate": 1.7998686355578108e-07, + "loss": 0.1393, + "step": 56576 + }, + { + "epoch": 3.9259593366178613, + "grad_norm": 1.2121102809906006, + "learning_rate": 1.7965006820970241e-07, + "loss": 0.249, + "step": 56577 + }, + { + "epoch": 3.926028728054958, + "grad_norm": 1.3980705738067627, + "learning_rate": 1.7931358798489727e-07, + "loss": 0.2695, + "step": 56578 + }, + { + "epoch": 3.9260981194920546, + "grad_norm": 0.8359333276748657, + "learning_rate": 1.7897742288240928e-07, + "loss": 0.1934, + "step": 56579 + }, + { + "epoch": 3.9261675109291514, + "grad_norm": 0.9002202749252319, + "learning_rate": 1.786415729033153e-07, + "loss": 0.227, + "step": 56580 + }, + { + "epoch": 3.926236902366248, + "grad_norm": 0.7377066612243652, + "learning_rate": 1.7830603804865899e-07, + "loss": 0.2033, + "step": 56581 + }, + { + "epoch": 3.9263062938033446, + "grad_norm": 0.9161905646324158, + "learning_rate": 1.7797081831951724e-07, + "loss": 0.2605, + "step": 56582 + }, + { + "epoch": 3.9263756852404414, + "grad_norm": 0.823612630367279, + "learning_rate": 1.7763591371693365e-07, + "loss": 0.1985, + "step": 56583 + }, + { + "epoch": 3.926445076677538, + "grad_norm": 0.9882715940475464, + "learning_rate": 1.773013242419741e-07, + "loss": 0.2787, + "step": 56584 + }, + { + "epoch": 3.9265144681146347, + "grad_norm": 0.8683743476867676, + "learning_rate": 1.7696704989569323e-07, + "loss": 0.2791, + "step": 56585 + }, + { + "epoch": 3.926583859551731, + "grad_norm": 0.693531334400177, + "learning_rate": 1.766330906791458e-07, + "loss": 0.2279, + "step": 56586 + }, + { + "epoch": 3.926653250988828, + "grad_norm": 0.847604513168335, + "learning_rate": 1.7629944659338648e-07, + "loss": 0.1566, + "step": 56587 + }, + { + "epoch": 3.9267226424259247, + "grad_norm": 0.825639545917511, + "learning_rate": 1.7596611763947001e-07, + "loss": 0.2303, + "step": 56588 + }, + { + "epoch": 3.9267920338630216, + "grad_norm": 1.0039833784103394, + "learning_rate": 1.7563310381845112e-07, + "loss": 0.2484, + "step": 56589 + }, + { + "epoch": 3.926861425300118, + "grad_norm": 0.6462528705596924, + "learning_rate": 1.7530040513137336e-07, + "loss": 0.1683, + "step": 56590 + }, + { + "epoch": 3.926930816737215, + "grad_norm": 0.904718816280365, + "learning_rate": 1.7496802157928039e-07, + "loss": 0.3752, + "step": 56591 + }, + { + "epoch": 3.927000208174311, + "grad_norm": 1.0039687156677246, + "learning_rate": 1.74635953163238e-07, + "loss": 0.2554, + "step": 56592 + }, + { + "epoch": 3.927069599611408, + "grad_norm": 0.9952673316001892, + "learning_rate": 1.743041998842898e-07, + "loss": 0.2566, + "step": 56593 + }, + { + "epoch": 3.927138991048505, + "grad_norm": 0.939738392829895, + "learning_rate": 1.7397276174347944e-07, + "loss": 0.1808, + "step": 56594 + }, + { + "epoch": 3.9272083824856012, + "grad_norm": 1.1060048341751099, + "learning_rate": 1.7364163874186158e-07, + "loss": 0.3579, + "step": 56595 + }, + { + "epoch": 3.927277773922698, + "grad_norm": 0.8333747386932373, + "learning_rate": 1.7331083088046873e-07, + "loss": 0.3063, + "step": 56596 + }, + { + "epoch": 3.9273471653597944, + "grad_norm": 0.9539262652397156, + "learning_rate": 1.7298033816035563e-07, + "loss": 0.2445, + "step": 56597 + }, + { + "epoch": 3.9274165567968913, + "grad_norm": 2.2634246349334717, + "learning_rate": 1.726501605825548e-07, + "loss": 0.2188, + "step": 56598 + }, + { + "epoch": 3.927485948233988, + "grad_norm": 0.8039681315422058, + "learning_rate": 1.7232029814812088e-07, + "loss": 0.1989, + "step": 56599 + }, + { + "epoch": 3.9275553396710845, + "grad_norm": 1.0682066679000854, + "learning_rate": 1.7199075085808648e-07, + "loss": 0.2351, + "step": 56600 + }, + { + "epoch": 3.9276247311081813, + "grad_norm": 0.7784832715988159, + "learning_rate": 1.7166151871349512e-07, + "loss": 0.2339, + "step": 56601 + }, + { + "epoch": 3.9276941225452777, + "grad_norm": 0.8883951306343079, + "learning_rate": 1.7133260171539045e-07, + "loss": 0.2418, + "step": 56602 + }, + { + "epoch": 3.9277635139823746, + "grad_norm": 0.7746461033821106, + "learning_rate": 1.7100399986480498e-07, + "loss": 0.2504, + "step": 56603 + }, + { + "epoch": 3.9278329054194714, + "grad_norm": 0.9817802906036377, + "learning_rate": 1.706757131627712e-07, + "loss": 0.2837, + "step": 56604 + }, + { + "epoch": 3.9279022968565678, + "grad_norm": 0.6898108124732971, + "learning_rate": 1.7034774161034383e-07, + "loss": 0.2371, + "step": 56605 + }, + { + "epoch": 3.9279716882936646, + "grad_norm": 1.1736795902252197, + "learning_rate": 1.7002008520853319e-07, + "loss": 0.337, + "step": 56606 + }, + { + "epoch": 3.928041079730761, + "grad_norm": 0.7160977721214294, + "learning_rate": 1.6969274395839396e-07, + "loss": 0.1346, + "step": 56607 + }, + { + "epoch": 3.928110471167858, + "grad_norm": 0.8069671392440796, + "learning_rate": 1.6936571786095866e-07, + "loss": 0.2397, + "step": 56608 + }, + { + "epoch": 3.9281798626049547, + "grad_norm": 1.232840895652771, + "learning_rate": 1.690390069172376e-07, + "loss": 0.4233, + "step": 56609 + }, + { + "epoch": 3.928249254042051, + "grad_norm": 0.967275083065033, + "learning_rate": 1.6871261112828552e-07, + "loss": 0.405, + "step": 56610 + }, + { + "epoch": 3.928318645479148, + "grad_norm": 0.9239306449890137, + "learning_rate": 1.6838653049513486e-07, + "loss": 0.2625, + "step": 56611 + }, + { + "epoch": 3.9283880369162443, + "grad_norm": 0.9396577477455139, + "learning_rate": 1.6806076501879599e-07, + "loss": 0.2468, + "step": 56612 + }, + { + "epoch": 3.928457428353341, + "grad_norm": 2.1159873008728027, + "learning_rate": 1.6773531470030134e-07, + "loss": 0.3016, + "step": 56613 + }, + { + "epoch": 3.928526819790438, + "grad_norm": 0.7496570348739624, + "learning_rate": 1.674101795406946e-07, + "loss": 0.2906, + "step": 56614 + }, + { + "epoch": 3.9285962112275348, + "grad_norm": 0.9328353404998779, + "learning_rate": 1.6708535954098604e-07, + "loss": 0.2726, + "step": 56615 + }, + { + "epoch": 3.928665602664631, + "grad_norm": 0.9420620799064636, + "learning_rate": 1.6676085470219706e-07, + "loss": 0.2636, + "step": 56616 + }, + { + "epoch": 3.928734994101728, + "grad_norm": 1.3692600727081299, + "learning_rate": 1.6643666502537126e-07, + "loss": 0.2557, + "step": 56617 + }, + { + "epoch": 3.9288043855388244, + "grad_norm": 0.7346453666687012, + "learning_rate": 1.6611279051153006e-07, + "loss": 0.1413, + "step": 56618 + }, + { + "epoch": 3.928873776975921, + "grad_norm": 0.6627330183982849, + "learning_rate": 1.6578923116167267e-07, + "loss": 0.2658, + "step": 56619 + }, + { + "epoch": 3.928943168413018, + "grad_norm": 1.3036302328109741, + "learning_rate": 1.654659869768427e-07, + "loss": 0.3716, + "step": 56620 + }, + { + "epoch": 3.9290125598501144, + "grad_norm": 0.7398916482925415, + "learning_rate": 1.651430579580504e-07, + "loss": 0.1617, + "step": 56621 + }, + { + "epoch": 3.9290819512872113, + "grad_norm": 1.0217657089233398, + "learning_rate": 1.6482044410631726e-07, + "loss": 0.2396, + "step": 56622 + }, + { + "epoch": 3.9291513427243077, + "grad_norm": 0.7952768802642822, + "learning_rate": 1.6449814542266461e-07, + "loss": 0.3192, + "step": 56623 + }, + { + "epoch": 3.9292207341614045, + "grad_norm": 1.6926424503326416, + "learning_rate": 1.6417616190811392e-07, + "loss": 0.2846, + "step": 56624 + }, + { + "epoch": 3.9292901255985013, + "grad_norm": 1.1820101737976074, + "learning_rate": 1.6385449356367543e-07, + "loss": 0.2578, + "step": 56625 + }, + { + "epoch": 3.9293595170355977, + "grad_norm": 0.7402614951133728, + "learning_rate": 1.635331403903595e-07, + "loss": 0.2133, + "step": 56626 + }, + { + "epoch": 3.9294289084726945, + "grad_norm": 0.8131775856018066, + "learning_rate": 1.632121023891875e-07, + "loss": 0.1592, + "step": 56627 + }, + { + "epoch": 3.929498299909791, + "grad_norm": 0.7633105516433716, + "learning_rate": 1.6289137956116973e-07, + "loss": 0.1808, + "step": 56628 + }, + { + "epoch": 3.9295676913468878, + "grad_norm": 0.9456487894058228, + "learning_rate": 1.6257097190732762e-07, + "loss": 0.2904, + "step": 56629 + }, + { + "epoch": 3.9296370827839846, + "grad_norm": 0.6919397115707397, + "learning_rate": 1.6225087942866035e-07, + "loss": 0.2021, + "step": 56630 + }, + { + "epoch": 3.929706474221081, + "grad_norm": 0.9362261891365051, + "learning_rate": 1.6193110212618935e-07, + "loss": 0.321, + "step": 56631 + }, + { + "epoch": 3.929775865658178, + "grad_norm": 0.7799896597862244, + "learning_rate": 1.6161164000091379e-07, + "loss": 0.1081, + "step": 56632 + }, + { + "epoch": 3.929845257095274, + "grad_norm": 1.314056158065796, + "learning_rate": 1.6129249305384398e-07, + "loss": 0.253, + "step": 56633 + }, + { + "epoch": 3.929914648532371, + "grad_norm": 0.9670397043228149, + "learning_rate": 1.6097366128600132e-07, + "loss": 0.2757, + "step": 56634 + }, + { + "epoch": 3.929984039969468, + "grad_norm": 0.8507041335105896, + "learning_rate": 1.6065514469837395e-07, + "loss": 0.2444, + "step": 56635 + }, + { + "epoch": 3.9300534314065643, + "grad_norm": 0.8464433550834656, + "learning_rate": 1.6033694329197212e-07, + "loss": 0.228, + "step": 56636 + }, + { + "epoch": 3.930122822843661, + "grad_norm": 0.859531044960022, + "learning_rate": 1.6001905706780617e-07, + "loss": 0.188, + "step": 56637 + }, + { + "epoch": 3.9301922142807575, + "grad_norm": 0.7106176018714905, + "learning_rate": 1.597014860268753e-07, + "loss": 0.1624, + "step": 56638 + }, + { + "epoch": 3.9302616057178543, + "grad_norm": 0.8449882864952087, + "learning_rate": 1.5938423017017867e-07, + "loss": 0.2317, + "step": 56639 + }, + { + "epoch": 3.930330997154951, + "grad_norm": 1.7599438428878784, + "learning_rate": 1.5906728949872661e-07, + "loss": 0.2336, + "step": 56640 + }, + { + "epoch": 3.930400388592048, + "grad_norm": 0.7711350917816162, + "learning_rate": 1.5875066401351834e-07, + "loss": 0.3189, + "step": 56641 + }, + { + "epoch": 3.9304697800291444, + "grad_norm": 0.7848764061927795, + "learning_rate": 1.5843435371555304e-07, + "loss": 0.2067, + "step": 56642 + }, + { + "epoch": 3.930539171466241, + "grad_norm": 0.9270566701889038, + "learning_rate": 1.581183586058188e-07, + "loss": 0.2404, + "step": 56643 + }, + { + "epoch": 3.9306085629033376, + "grad_norm": 0.9741631150245667, + "learning_rate": 1.5780267868532594e-07, + "loss": 0.3391, + "step": 56644 + }, + { + "epoch": 3.9306779543404344, + "grad_norm": 0.7363245487213135, + "learning_rate": 1.5748731395506256e-07, + "loss": 0.2667, + "step": 56645 + }, + { + "epoch": 3.9307473457775313, + "grad_norm": 0.7398361563682556, + "learning_rate": 1.5717226441603895e-07, + "loss": 0.2228, + "step": 56646 + }, + { + "epoch": 3.9308167372146277, + "grad_norm": 0.7107536792755127, + "learning_rate": 1.56857530069221e-07, + "loss": 0.1456, + "step": 56647 + }, + { + "epoch": 3.9308861286517245, + "grad_norm": 0.8881219029426575, + "learning_rate": 1.5654311091563013e-07, + "loss": 0.2809, + "step": 56648 + }, + { + "epoch": 3.930955520088821, + "grad_norm": 0.9066975712776184, + "learning_rate": 1.5622900695624332e-07, + "loss": 0.2478, + "step": 56649 + }, + { + "epoch": 3.9310249115259177, + "grad_norm": 0.789383053779602, + "learning_rate": 1.559152181920598e-07, + "loss": 0.2343, + "step": 56650 + }, + { + "epoch": 3.9310943029630145, + "grad_norm": 0.8485754132270813, + "learning_rate": 1.5560174462407873e-07, + "loss": 0.2151, + "step": 56651 + }, + { + "epoch": 3.931163694400111, + "grad_norm": 1.0340619087219238, + "learning_rate": 1.5528858625326604e-07, + "loss": 0.2633, + "step": 56652 + }, + { + "epoch": 3.9312330858372078, + "grad_norm": 1.145980715751648, + "learning_rate": 1.54975743080632e-07, + "loss": 0.3193, + "step": 56653 + }, + { + "epoch": 3.931302477274304, + "grad_norm": 0.7952861189842224, + "learning_rate": 1.5466321510714255e-07, + "loss": 0.18, + "step": 56654 + }, + { + "epoch": 3.931371868711401, + "grad_norm": 0.8400622010231018, + "learning_rate": 1.5435100233380794e-07, + "loss": 0.188, + "step": 56655 + }, + { + "epoch": 3.931441260148498, + "grad_norm": 1.1737165451049805, + "learning_rate": 1.540391047615941e-07, + "loss": 0.3428, + "step": 56656 + }, + { + "epoch": 3.931510651585594, + "grad_norm": 1.5736182928085327, + "learning_rate": 1.5372752239150024e-07, + "loss": 0.2182, + "step": 56657 + }, + { + "epoch": 3.931580043022691, + "grad_norm": 0.7889024019241333, + "learning_rate": 1.534162552245033e-07, + "loss": 0.241, + "step": 56658 + }, + { + "epoch": 3.9316494344597874, + "grad_norm": 0.8371791839599609, + "learning_rate": 1.5310530326159144e-07, + "loss": 0.3297, + "step": 56659 + }, + { + "epoch": 3.9317188258968843, + "grad_norm": 1.1654936075210571, + "learning_rate": 1.5279466650374163e-07, + "loss": 0.2479, + "step": 56660 + }, + { + "epoch": 3.931788217333981, + "grad_norm": 0.6912656426429749, + "learning_rate": 1.5248434495193087e-07, + "loss": 0.1674, + "step": 56661 + }, + { + "epoch": 3.9318576087710775, + "grad_norm": 1.3230770826339722, + "learning_rate": 1.5217433860713615e-07, + "loss": 0.2148, + "step": 56662 + }, + { + "epoch": 3.9319270002081743, + "grad_norm": 0.8896827101707458, + "learning_rate": 1.5186464747034557e-07, + "loss": 0.239, + "step": 56663 + }, + { + "epoch": 3.9319963916452707, + "grad_norm": 0.6545361876487732, + "learning_rate": 1.5155527154254722e-07, + "loss": 0.154, + "step": 56664 + }, + { + "epoch": 3.9320657830823675, + "grad_norm": 1.0339946746826172, + "learning_rate": 1.5124621082468483e-07, + "loss": 0.3072, + "step": 56665 + }, + { + "epoch": 3.9321351745194644, + "grad_norm": 0.8430866599082947, + "learning_rate": 1.5093746531776863e-07, + "loss": 0.3138, + "step": 56666 + }, + { + "epoch": 3.932204565956561, + "grad_norm": 1.0958706140518188, + "learning_rate": 1.506290350227535e-07, + "loss": 0.2308, + "step": 56667 + }, + { + "epoch": 3.9322739573936576, + "grad_norm": 0.8566051721572876, + "learning_rate": 1.5032091994061636e-07, + "loss": 0.2841, + "step": 56668 + }, + { + "epoch": 3.9323433488307544, + "grad_norm": 1.0069752931594849, + "learning_rate": 1.5001312007233425e-07, + "loss": 0.2378, + "step": 56669 + }, + { + "epoch": 3.932412740267851, + "grad_norm": 0.6096144318580627, + "learning_rate": 1.4970563541887307e-07, + "loss": 0.1352, + "step": 56670 + }, + { + "epoch": 3.9324821317049476, + "grad_norm": 0.9150409698486328, + "learning_rate": 1.4939846598119867e-07, + "loss": 0.2988, + "step": 56671 + }, + { + "epoch": 3.9325515231420445, + "grad_norm": 0.9064040184020996, + "learning_rate": 1.4909161176029917e-07, + "loss": 0.2592, + "step": 56672 + }, + { + "epoch": 3.932620914579141, + "grad_norm": 0.7264096736907959, + "learning_rate": 1.487850727571294e-07, + "loss": 0.1917, + "step": 56673 + }, + { + "epoch": 3.9326903060162377, + "grad_norm": 0.8300895690917969, + "learning_rate": 1.484788489726663e-07, + "loss": 0.2453, + "step": 56674 + }, + { + "epoch": 3.932759697453334, + "grad_norm": 0.9929112195968628, + "learning_rate": 1.4817294040786466e-07, + "loss": 0.2456, + "step": 56675 + }, + { + "epoch": 3.932829088890431, + "grad_norm": 1.0568523406982422, + "learning_rate": 1.4786734706370153e-07, + "loss": 0.2314, + "step": 56676 + }, + { + "epoch": 3.9328984803275278, + "grad_norm": 0.7226382493972778, + "learning_rate": 1.4756206894113168e-07, + "loss": 0.1847, + "step": 56677 + }, + { + "epoch": 3.932967871764624, + "grad_norm": 1.0681493282318115, + "learning_rate": 1.4725710604112098e-07, + "loss": 0.2917, + "step": 56678 + }, + { + "epoch": 3.933037263201721, + "grad_norm": 0.7520702481269836, + "learning_rate": 1.4695245836463535e-07, + "loss": 0.2509, + "step": 56679 + }, + { + "epoch": 3.9331066546388174, + "grad_norm": 0.9957179427146912, + "learning_rate": 1.4664812591264066e-07, + "loss": 0.3081, + "step": 56680 + }, + { + "epoch": 3.933176046075914, + "grad_norm": 1.181640625, + "learning_rate": 1.463441086860917e-07, + "loss": 0.3273, + "step": 56681 + }, + { + "epoch": 3.933245437513011, + "grad_norm": 0.7171018719673157, + "learning_rate": 1.4604040668594331e-07, + "loss": 0.1531, + "step": 56682 + }, + { + "epoch": 3.9333148289501074, + "grad_norm": 0.7272637486457825, + "learning_rate": 1.4573701991316136e-07, + "loss": 0.1816, + "step": 56683 + }, + { + "epoch": 3.9333842203872043, + "grad_norm": 0.9237533211708069, + "learning_rate": 1.454339483687006e-07, + "loss": 0.2023, + "step": 56684 + }, + { + "epoch": 3.9334536118243006, + "grad_norm": 0.8569719791412354, + "learning_rate": 1.45131192053527e-07, + "loss": 0.127, + "step": 56685 + }, + { + "epoch": 3.9335230032613975, + "grad_norm": 1.3145688772201538, + "learning_rate": 1.4482875096858416e-07, + "loss": 0.4146, + "step": 56686 + }, + { + "epoch": 3.9335923946984943, + "grad_norm": 0.5604365468025208, + "learning_rate": 1.4452662511483805e-07, + "loss": 0.2365, + "step": 56687 + }, + { + "epoch": 3.933661786135591, + "grad_norm": 0.9323527216911316, + "learning_rate": 1.4422481449323234e-07, + "loss": 0.1856, + "step": 56688 + }, + { + "epoch": 3.9337311775726875, + "grad_norm": 0.5965030193328857, + "learning_rate": 1.4392331910472178e-07, + "loss": 0.1435, + "step": 56689 + }, + { + "epoch": 3.933800569009784, + "grad_norm": 1.4668123722076416, + "learning_rate": 1.4362213895025012e-07, + "loss": 0.3654, + "step": 56690 + }, + { + "epoch": 3.9338699604468808, + "grad_norm": 0.79596346616745, + "learning_rate": 1.433212740307943e-07, + "loss": 0.1471, + "step": 56691 + }, + { + "epoch": 3.9339393518839776, + "grad_norm": 0.9192934632301331, + "learning_rate": 1.4302072434727586e-07, + "loss": 0.2493, + "step": 56692 + }, + { + "epoch": 3.9340087433210744, + "grad_norm": 0.9199220538139343, + "learning_rate": 1.4272048990066067e-07, + "loss": 0.1918, + "step": 56693 + }, + { + "epoch": 3.934078134758171, + "grad_norm": 0.9117647409439087, + "learning_rate": 1.424205706918924e-07, + "loss": 0.1842, + "step": 56694 + }, + { + "epoch": 3.9341475261952676, + "grad_norm": 0.8424246907234192, + "learning_rate": 1.4212096672191476e-07, + "loss": 0.2382, + "step": 56695 + }, + { + "epoch": 3.934216917632364, + "grad_norm": 0.8480830788612366, + "learning_rate": 1.4182167799168257e-07, + "loss": 0.1956, + "step": 56696 + }, + { + "epoch": 3.934286309069461, + "grad_norm": 0.7456002831459045, + "learning_rate": 1.4152270450212833e-07, + "loss": 0.17, + "step": 56697 + }, + { + "epoch": 3.9343557005065577, + "grad_norm": 0.919315755367279, + "learning_rate": 1.4122404625419583e-07, + "loss": 0.3399, + "step": 56698 + }, + { + "epoch": 3.934425091943654, + "grad_norm": 0.8932241797447205, + "learning_rate": 1.409257032488398e-07, + "loss": 0.209, + "step": 56699 + }, + { + "epoch": 3.934494483380751, + "grad_norm": 0.6194241046905518, + "learning_rate": 1.4062767548699285e-07, + "loss": 0.1844, + "step": 56700 + }, + { + "epoch": 3.9345638748178473, + "grad_norm": 0.8130941987037659, + "learning_rate": 1.4032996296959865e-07, + "loss": 0.2966, + "step": 56701 + }, + { + "epoch": 3.934633266254944, + "grad_norm": 0.7863736748695374, + "learning_rate": 1.4003256569760093e-07, + "loss": 0.295, + "step": 56702 + }, + { + "epoch": 3.934702657692041, + "grad_norm": 0.9565919637680054, + "learning_rate": 1.3973548367193224e-07, + "loss": 0.3003, + "step": 56703 + }, + { + "epoch": 3.9347720491291374, + "grad_norm": 1.3202592134475708, + "learning_rate": 1.3943871689352517e-07, + "loss": 0.2559, + "step": 56704 + }, + { + "epoch": 3.934841440566234, + "grad_norm": 1.3049321174621582, + "learning_rate": 1.3914226536333452e-07, + "loss": 0.3835, + "step": 56705 + }, + { + "epoch": 3.9349108320033306, + "grad_norm": 0.6731938123703003, + "learning_rate": 1.3884612908228178e-07, + "loss": 0.2383, + "step": 56706 + }, + { + "epoch": 3.9349802234404274, + "grad_norm": 1.8350471258163452, + "learning_rate": 1.3855030805129953e-07, + "loss": 0.3537, + "step": 56707 + }, + { + "epoch": 3.9350496148775242, + "grad_norm": 0.9654465317726135, + "learning_rate": 1.3825480227134257e-07, + "loss": 0.2646, + "step": 56708 + }, + { + "epoch": 3.9351190063146206, + "grad_norm": 0.8284081816673279, + "learning_rate": 1.3795961174332127e-07, + "loss": 0.3039, + "step": 56709 + }, + { + "epoch": 3.9351883977517175, + "grad_norm": 1.0306202173233032, + "learning_rate": 1.3766473646816824e-07, + "loss": 0.2211, + "step": 56710 + }, + { + "epoch": 3.935257789188814, + "grad_norm": 1.1406627893447876, + "learning_rate": 1.3737017644682716e-07, + "loss": 0.1968, + "step": 56711 + }, + { + "epoch": 3.9353271806259107, + "grad_norm": 2.2684366703033447, + "learning_rate": 1.3707593168023058e-07, + "loss": 0.3775, + "step": 56712 + }, + { + "epoch": 3.9353965720630075, + "grad_norm": 1.2778115272521973, + "learning_rate": 1.3678200216928893e-07, + "loss": 0.2729, + "step": 56713 + }, + { + "epoch": 3.9354659635001044, + "grad_norm": 0.6600664854049683, + "learning_rate": 1.364883879149459e-07, + "loss": 0.1159, + "step": 56714 + }, + { + "epoch": 3.9355353549372007, + "grad_norm": 0.9090803265571594, + "learning_rate": 1.3619508891811183e-07, + "loss": 0.2107, + "step": 56715 + }, + { + "epoch": 3.935604746374297, + "grad_norm": 1.1375136375427246, + "learning_rate": 1.3590210517974156e-07, + "loss": 0.4016, + "step": 56716 + }, + { + "epoch": 3.935674137811394, + "grad_norm": 0.7847921848297119, + "learning_rate": 1.3560943670072323e-07, + "loss": 0.2129, + "step": 56717 + }, + { + "epoch": 3.935743529248491, + "grad_norm": 0.74251788854599, + "learning_rate": 1.3531708348201167e-07, + "loss": 0.2522, + "step": 56718 + }, + { + "epoch": 3.9358129206855876, + "grad_norm": 0.6864027380943298, + "learning_rate": 1.3502504552451722e-07, + "loss": 0.1754, + "step": 56719 + }, + { + "epoch": 3.935882312122684, + "grad_norm": 1.0357246398925781, + "learning_rate": 1.347333228291503e-07, + "loss": 0.2358, + "step": 56720 + }, + { + "epoch": 3.935951703559781, + "grad_norm": 0.7900442481040955, + "learning_rate": 1.3444191539685457e-07, + "loss": 0.1693, + "step": 56721 + }, + { + "epoch": 3.9360210949968772, + "grad_norm": 0.8229185342788696, + "learning_rate": 1.3415082322854044e-07, + "loss": 0.2389, + "step": 56722 + }, + { + "epoch": 3.936090486433974, + "grad_norm": 1.8058910369873047, + "learning_rate": 1.338600463251183e-07, + "loss": 0.3095, + "step": 56723 + }, + { + "epoch": 3.936159877871071, + "grad_norm": 0.6378788948059082, + "learning_rate": 1.3356958468752068e-07, + "loss": 0.1457, + "step": 56724 + }, + { + "epoch": 3.9362292693081673, + "grad_norm": 1.0386810302734375, + "learning_rate": 1.3327943831664692e-07, + "loss": 0.2082, + "step": 56725 + }, + { + "epoch": 3.936298660745264, + "grad_norm": 1.0820144414901733, + "learning_rate": 1.3298960721342957e-07, + "loss": 0.3505, + "step": 56726 + }, + { + "epoch": 3.9363680521823605, + "grad_norm": 0.7114025354385376, + "learning_rate": 1.3270009137877903e-07, + "loss": 0.2066, + "step": 56727 + }, + { + "epoch": 3.9364374436194574, + "grad_norm": 1.0204554796218872, + "learning_rate": 1.324108908136057e-07, + "loss": 0.2271, + "step": 56728 + }, + { + "epoch": 3.936506835056554, + "grad_norm": 0.9449129700660706, + "learning_rate": 1.3212200551883102e-07, + "loss": 0.2043, + "step": 56729 + }, + { + "epoch": 3.9365762264936506, + "grad_norm": 0.836639940738678, + "learning_rate": 1.3183343549535432e-07, + "loss": 0.2529, + "step": 56730 + }, + { + "epoch": 3.9366456179307474, + "grad_norm": 1.1933587789535522, + "learning_rate": 1.3154518074409705e-07, + "loss": 0.3326, + "step": 56731 + }, + { + "epoch": 3.936715009367844, + "grad_norm": 0.8376119136810303, + "learning_rate": 1.3125724126595852e-07, + "loss": 0.2511, + "step": 56732 + }, + { + "epoch": 3.9367844008049406, + "grad_norm": 1.1458890438079834, + "learning_rate": 1.3096961706184908e-07, + "loss": 0.2992, + "step": 56733 + }, + { + "epoch": 3.9368537922420375, + "grad_norm": 0.6564017534255981, + "learning_rate": 1.306823081326902e-07, + "loss": 0.1956, + "step": 56734 + }, + { + "epoch": 3.936923183679134, + "grad_norm": 0.9753631949424744, + "learning_rate": 1.3039531447938126e-07, + "loss": 0.283, + "step": 56735 + }, + { + "epoch": 3.9369925751162307, + "grad_norm": 0.4928426444530487, + "learning_rate": 1.3010863610282143e-07, + "loss": 0.1612, + "step": 56736 + }, + { + "epoch": 3.937061966553327, + "grad_norm": 0.8006313443183899, + "learning_rate": 1.2982227300392115e-07, + "loss": 0.1865, + "step": 56737 + }, + { + "epoch": 3.937131357990424, + "grad_norm": 3.0958337783813477, + "learning_rate": 1.2953622518359076e-07, + "loss": 0.3185, + "step": 56738 + }, + { + "epoch": 3.9372007494275207, + "grad_norm": 1.0175583362579346, + "learning_rate": 1.292504926427296e-07, + "loss": 0.3459, + "step": 56739 + }, + { + "epoch": 3.9372701408646176, + "grad_norm": 0.828605055809021, + "learning_rate": 1.2896507538222579e-07, + "loss": 0.1763, + "step": 56740 + }, + { + "epoch": 3.937339532301714, + "grad_norm": 0.8866250514984131, + "learning_rate": 1.2867997340300085e-07, + "loss": 0.3152, + "step": 56741 + }, + { + "epoch": 3.937408923738811, + "grad_norm": 0.9140639901161194, + "learning_rate": 1.2839518670594298e-07, + "loss": 0.2038, + "step": 56742 + }, + { + "epoch": 3.937478315175907, + "grad_norm": 0.8951842188835144, + "learning_rate": 1.281107152919625e-07, + "loss": 0.3593, + "step": 56743 + }, + { + "epoch": 3.937547706613004, + "grad_norm": 0.6922780871391296, + "learning_rate": 1.2782655916193654e-07, + "loss": 0.2616, + "step": 56744 + }, + { + "epoch": 3.937617098050101, + "grad_norm": 0.772813618183136, + "learning_rate": 1.2754271831678654e-07, + "loss": 0.2329, + "step": 56745 + }, + { + "epoch": 3.9376864894871972, + "grad_norm": 0.8747655749320984, + "learning_rate": 1.2725919275738962e-07, + "loss": 0.2273, + "step": 56746 + }, + { + "epoch": 3.937755880924294, + "grad_norm": 1.034787893295288, + "learning_rate": 1.2697598248464504e-07, + "loss": 0.3226, + "step": 56747 + }, + { + "epoch": 3.9378252723613905, + "grad_norm": 0.791892409324646, + "learning_rate": 1.2669308749946317e-07, + "loss": 0.2637, + "step": 56748 + }, + { + "epoch": 3.9378946637984873, + "grad_norm": 0.7298795580863953, + "learning_rate": 1.2641050780271003e-07, + "loss": 0.1748, + "step": 56749 + }, + { + "epoch": 3.937964055235584, + "grad_norm": 0.8557602167129517, + "learning_rate": 1.2612824339529594e-07, + "loss": 0.2294, + "step": 56750 + }, + { + "epoch": 3.9380334466726805, + "grad_norm": 0.7151828408241272, + "learning_rate": 1.2584629427810912e-07, + "loss": 0.1944, + "step": 56751 + }, + { + "epoch": 3.9381028381097773, + "grad_norm": 0.8793318271636963, + "learning_rate": 1.2556466045203775e-07, + "loss": 0.2884, + "step": 56752 + }, + { + "epoch": 3.9381722295468737, + "grad_norm": 1.2471439838409424, + "learning_rate": 1.2528334191797e-07, + "loss": 0.2631, + "step": 56753 + }, + { + "epoch": 3.9382416209839706, + "grad_norm": 0.68524569272995, + "learning_rate": 1.25002338676794e-07, + "loss": 0.2116, + "step": 56754 + }, + { + "epoch": 3.9383110124210674, + "grad_norm": 1.1892009973526, + "learning_rate": 1.2472165072939802e-07, + "loss": 0.2927, + "step": 56755 + }, + { + "epoch": 3.938380403858164, + "grad_norm": 0.9670370817184448, + "learning_rate": 1.2444127807667016e-07, + "loss": 0.2844, + "step": 56756 + }, + { + "epoch": 3.9384497952952606, + "grad_norm": 1.4399949312210083, + "learning_rate": 1.2416122071948755e-07, + "loss": 0.2931, + "step": 56757 + }, + { + "epoch": 3.938519186732357, + "grad_norm": 0.9044219851493835, + "learning_rate": 1.2388147865874944e-07, + "loss": 0.2028, + "step": 56758 + }, + { + "epoch": 3.938588578169454, + "grad_norm": 0.9828928709030151, + "learning_rate": 1.2360205189532182e-07, + "loss": 0.3157, + "step": 56759 + }, + { + "epoch": 3.9386579696065507, + "grad_norm": 1.1867611408233643, + "learning_rate": 1.2332294043010395e-07, + "loss": 0.3569, + "step": 56760 + }, + { + "epoch": 3.938727361043647, + "grad_norm": 0.8957512974739075, + "learning_rate": 1.2304414426396182e-07, + "loss": 0.2041, + "step": 56761 + }, + { + "epoch": 3.938796752480744, + "grad_norm": 1.321192741394043, + "learning_rate": 1.2276566339777252e-07, + "loss": 0.3031, + "step": 56762 + }, + { + "epoch": 3.9388661439178403, + "grad_norm": 0.6390706896781921, + "learning_rate": 1.224874978324353e-07, + "loss": 0.2211, + "step": 56763 + }, + { + "epoch": 3.938935535354937, + "grad_norm": 1.182521939277649, + "learning_rate": 1.2220964756881613e-07, + "loss": 0.4026, + "step": 56764 + }, + { + "epoch": 3.939004926792034, + "grad_norm": 0.7458346486091614, + "learning_rate": 1.2193211260779214e-07, + "loss": 0.2163, + "step": 56765 + }, + { + "epoch": 3.939074318229131, + "grad_norm": 3.094621419906616, + "learning_rate": 1.2165489295024036e-07, + "loss": 0.3204, + "step": 56766 + }, + { + "epoch": 3.939143709666227, + "grad_norm": 1.5001637935638428, + "learning_rate": 1.2137798859703787e-07, + "loss": 0.2113, + "step": 56767 + }, + { + "epoch": 3.939213101103324, + "grad_norm": 0.8757429122924805, + "learning_rate": 1.2110139954905063e-07, + "loss": 0.1795, + "step": 56768 + }, + { + "epoch": 3.9392824925404204, + "grad_norm": 0.6125153303146362, + "learning_rate": 1.2082512580716687e-07, + "loss": 0.1595, + "step": 56769 + }, + { + "epoch": 3.9393518839775172, + "grad_norm": 0.9088496565818787, + "learning_rate": 1.2054916737225253e-07, + "loss": 0.2276, + "step": 56770 + }, + { + "epoch": 3.939421275414614, + "grad_norm": 1.0159716606140137, + "learning_rate": 1.2027352424516247e-07, + "loss": 0.3157, + "step": 56771 + }, + { + "epoch": 3.9394906668517105, + "grad_norm": 0.9638099670410156, + "learning_rate": 1.199981964267849e-07, + "loss": 0.2613, + "step": 56772 + }, + { + "epoch": 3.9395600582888073, + "grad_norm": 0.7645508050918579, + "learning_rate": 1.1972318391799686e-07, + "loss": 0.2131, + "step": 56773 + }, + { + "epoch": 3.9396294497259037, + "grad_norm": 0.9040126204490662, + "learning_rate": 1.1944848671964214e-07, + "loss": 0.2945, + "step": 56774 + }, + { + "epoch": 3.9396988411630005, + "grad_norm": 0.6607257127761841, + "learning_rate": 1.1917410483259784e-07, + "loss": 0.1838, + "step": 56775 + }, + { + "epoch": 3.9397682326000973, + "grad_norm": 0.7872775197029114, + "learning_rate": 1.1890003825774099e-07, + "loss": 0.2314, + "step": 56776 + }, + { + "epoch": 3.9398376240371937, + "grad_norm": 0.47304216027259827, + "learning_rate": 1.1862628699592649e-07, + "loss": 0.1164, + "step": 56777 + }, + { + "epoch": 3.9399070154742906, + "grad_norm": 1.2142049074172974, + "learning_rate": 1.183528510480203e-07, + "loss": 0.3586, + "step": 56778 + }, + { + "epoch": 3.939976406911387, + "grad_norm": 0.9118441939353943, + "learning_rate": 1.180797304148884e-07, + "loss": 0.2705, + "step": 56779 + }, + { + "epoch": 3.940045798348484, + "grad_norm": 0.90663743019104, + "learning_rate": 1.1780692509738566e-07, + "loss": 0.2479, + "step": 56780 + }, + { + "epoch": 3.9401151897855806, + "grad_norm": 2.0240657329559326, + "learning_rate": 1.1753443509637807e-07, + "loss": 0.2986, + "step": 56781 + }, + { + "epoch": 3.940184581222677, + "grad_norm": 0.8067831993103027, + "learning_rate": 1.1726226041273159e-07, + "loss": 0.1686, + "step": 56782 + }, + { + "epoch": 3.940253972659774, + "grad_norm": 0.935839056968689, + "learning_rate": 1.1699040104728997e-07, + "loss": 0.2404, + "step": 56783 + }, + { + "epoch": 3.9403233640968702, + "grad_norm": 0.7871552109718323, + "learning_rate": 1.1671885700093033e-07, + "loss": 0.1797, + "step": 56784 + }, + { + "epoch": 3.940392755533967, + "grad_norm": 1.195631742477417, + "learning_rate": 1.164476282744964e-07, + "loss": 0.3124, + "step": 56785 + }, + { + "epoch": 3.940462146971064, + "grad_norm": 0.6741659045219421, + "learning_rate": 1.1617671486884307e-07, + "loss": 0.1481, + "step": 56786 + }, + { + "epoch": 3.9405315384081603, + "grad_norm": 1.0889836549758911, + "learning_rate": 1.1590611678483631e-07, + "loss": 0.247, + "step": 56787 + }, + { + "epoch": 3.940600929845257, + "grad_norm": 0.8499217629432678, + "learning_rate": 1.15635834023331e-07, + "loss": 0.3613, + "step": 56788 + }, + { + "epoch": 3.9406703212823535, + "grad_norm": 1.1326168775558472, + "learning_rate": 1.153658665851709e-07, + "loss": 0.2781, + "step": 56789 + }, + { + "epoch": 3.9407397127194503, + "grad_norm": 1.0609956979751587, + "learning_rate": 1.1509621447121088e-07, + "loss": 0.2963, + "step": 56790 + }, + { + "epoch": 3.940809104156547, + "grad_norm": 0.7251200675964355, + "learning_rate": 1.1482687768230582e-07, + "loss": 0.23, + "step": 56791 + }, + { + "epoch": 3.940878495593644, + "grad_norm": 1.5367236137390137, + "learning_rate": 1.1455785621929948e-07, + "loss": 0.2314, + "step": 56792 + }, + { + "epoch": 3.9409478870307404, + "grad_norm": 1.147077202796936, + "learning_rate": 1.1428915008304675e-07, + "loss": 0.2307, + "step": 56793 + }, + { + "epoch": 3.9410172784678372, + "grad_norm": 0.8562558889389038, + "learning_rate": 1.1402075927439137e-07, + "loss": 0.3091, + "step": 56794 + }, + { + "epoch": 3.9410866699049336, + "grad_norm": 0.8585337400436401, + "learning_rate": 1.1375268379418824e-07, + "loss": 0.2444, + "step": 56795 + }, + { + "epoch": 3.9411560613420304, + "grad_norm": 0.9907830357551575, + "learning_rate": 1.1348492364327001e-07, + "loss": 0.2575, + "step": 56796 + }, + { + "epoch": 3.9412254527791273, + "grad_norm": 0.7431693077087402, + "learning_rate": 1.1321747882250266e-07, + "loss": 0.1529, + "step": 56797 + }, + { + "epoch": 3.9412948442162237, + "grad_norm": 1.0731714963912964, + "learning_rate": 1.1295034933270776e-07, + "loss": 0.2562, + "step": 56798 + }, + { + "epoch": 3.9413642356533205, + "grad_norm": 0.9131544828414917, + "learning_rate": 1.1268353517475127e-07, + "loss": 0.2905, + "step": 56799 + }, + { + "epoch": 3.941433627090417, + "grad_norm": 0.8668481111526489, + "learning_rate": 1.1241703634945478e-07, + "loss": 0.3236, + "step": 56800 + }, + { + "epoch": 3.9415030185275137, + "grad_norm": 0.978009819984436, + "learning_rate": 1.1215085285767313e-07, + "loss": 0.2469, + "step": 56801 + }, + { + "epoch": 3.9415724099646106, + "grad_norm": 0.8520021438598633, + "learning_rate": 1.1188498470023901e-07, + "loss": 0.2419, + "step": 56802 + }, + { + "epoch": 3.941641801401707, + "grad_norm": 1.0656489133834839, + "learning_rate": 1.1161943187799617e-07, + "loss": 0.3253, + "step": 56803 + }, + { + "epoch": 3.941711192838804, + "grad_norm": 1.087517499923706, + "learning_rate": 1.113541943917773e-07, + "loss": 0.4832, + "step": 56804 + }, + { + "epoch": 3.9417805842759, + "grad_norm": 0.8189701437950134, + "learning_rate": 1.1108927224243726e-07, + "loss": 0.2034, + "step": 56805 + }, + { + "epoch": 3.941849975712997, + "grad_norm": 0.7941866517066956, + "learning_rate": 1.1082466543078651e-07, + "loss": 0.2003, + "step": 56806 + }, + { + "epoch": 3.941919367150094, + "grad_norm": 0.8295549154281616, + "learning_rate": 1.1056037395767993e-07, + "loss": 0.2911, + "step": 56807 + }, + { + "epoch": 3.9419887585871902, + "grad_norm": 0.6713841557502747, + "learning_rate": 1.1029639782393908e-07, + "loss": 0.3376, + "step": 56808 + }, + { + "epoch": 3.942058150024287, + "grad_norm": 0.8677299618721008, + "learning_rate": 1.1003273703040773e-07, + "loss": 0.3684, + "step": 56809 + }, + { + "epoch": 3.9421275414613834, + "grad_norm": 0.635482907295227, + "learning_rate": 1.0976939157790744e-07, + "loss": 0.3067, + "step": 56810 + }, + { + "epoch": 3.9421969328984803, + "grad_norm": 1.2246285676956177, + "learning_rate": 1.0950636146728199e-07, + "loss": 0.2434, + "step": 56811 + }, + { + "epoch": 3.942266324335577, + "grad_norm": 0.5731025338172913, + "learning_rate": 1.0924364669935295e-07, + "loss": 0.1755, + "step": 56812 + }, + { + "epoch": 3.9423357157726735, + "grad_norm": 1.5091525316238403, + "learning_rate": 1.0898124727495295e-07, + "loss": 0.3315, + "step": 56813 + }, + { + "epoch": 3.9424051072097703, + "grad_norm": 0.9863656163215637, + "learning_rate": 1.0871916319491471e-07, + "loss": 0.2805, + "step": 56814 + }, + { + "epoch": 3.9424744986468667, + "grad_norm": 1.012914776802063, + "learning_rate": 1.0845739446004865e-07, + "loss": 0.2173, + "step": 56815 + }, + { + "epoch": 3.9425438900839636, + "grad_norm": 0.9471990466117859, + "learning_rate": 1.0819594107120967e-07, + "loss": 0.1544, + "step": 56816 + }, + { + "epoch": 3.9426132815210604, + "grad_norm": 0.7433015704154968, + "learning_rate": 1.0793480302919711e-07, + "loss": 0.1516, + "step": 56817 + }, + { + "epoch": 3.942682672958157, + "grad_norm": 0.7419657111167908, + "learning_rate": 1.0767398033484366e-07, + "loss": 0.1786, + "step": 56818 + }, + { + "epoch": 3.9427520643952536, + "grad_norm": 1.00011146068573, + "learning_rate": 1.0741347298897087e-07, + "loss": 0.2693, + "step": 56819 + }, + { + "epoch": 3.9428214558323504, + "grad_norm": 0.9299345016479492, + "learning_rate": 1.071532809924114e-07, + "loss": 0.3473, + "step": 56820 + }, + { + "epoch": 3.942890847269447, + "grad_norm": 0.9614426493644714, + "learning_rate": 1.0689340434597573e-07, + "loss": 0.3147, + "step": 56821 + }, + { + "epoch": 3.9429602387065437, + "grad_norm": 0.8726586103439331, + "learning_rate": 1.066338430504854e-07, + "loss": 0.1856, + "step": 56822 + }, + { + "epoch": 3.9430296301436405, + "grad_norm": 1.1171120405197144, + "learning_rate": 1.0637459710676201e-07, + "loss": 0.3584, + "step": 56823 + }, + { + "epoch": 3.943099021580737, + "grad_norm": 0.8770831227302551, + "learning_rate": 1.061156665156271e-07, + "loss": 0.3333, + "step": 56824 + }, + { + "epoch": 3.9431684130178337, + "grad_norm": 0.8511716723442078, + "learning_rate": 1.0585705127790224e-07, + "loss": 0.2564, + "step": 56825 + }, + { + "epoch": 3.94323780445493, + "grad_norm": 0.7655284404754639, + "learning_rate": 1.055987513943868e-07, + "loss": 0.2809, + "step": 56826 + }, + { + "epoch": 3.943307195892027, + "grad_norm": 1.9057366847991943, + "learning_rate": 1.0534076686591343e-07, + "loss": 0.2594, + "step": 56827 + }, + { + "epoch": 3.9433765873291238, + "grad_norm": 0.881499171257019, + "learning_rate": 1.0508309769328151e-07, + "loss": 0.2457, + "step": 56828 + }, + { + "epoch": 3.94344597876622, + "grad_norm": 0.8251357674598694, + "learning_rate": 1.0482574387731258e-07, + "loss": 0.2109, + "step": 56829 + }, + { + "epoch": 3.943515370203317, + "grad_norm": 0.7137382626533508, + "learning_rate": 1.0456870541882824e-07, + "loss": 0.1906, + "step": 56830 + }, + { + "epoch": 3.9435847616404134, + "grad_norm": 0.8963717818260193, + "learning_rate": 1.0431198231861671e-07, + "loss": 0.1704, + "step": 56831 + }, + { + "epoch": 3.94365415307751, + "grad_norm": 0.5927493572235107, + "learning_rate": 1.040555745775107e-07, + "loss": 0.1795, + "step": 56832 + }, + { + "epoch": 3.943723544514607, + "grad_norm": 0.6521167755126953, + "learning_rate": 1.0379948219630952e-07, + "loss": 0.1404, + "step": 56833 + }, + { + "epoch": 3.9437929359517034, + "grad_norm": 0.8115959167480469, + "learning_rate": 1.0354370517582368e-07, + "loss": 0.1822, + "step": 56834 + }, + { + "epoch": 3.9438623273888003, + "grad_norm": 0.5404078960418701, + "learning_rate": 1.0328824351686362e-07, + "loss": 0.178, + "step": 56835 + }, + { + "epoch": 3.9439317188258967, + "grad_norm": 0.6872166395187378, + "learning_rate": 1.0303309722022869e-07, + "loss": 0.197, + "step": 56836 + }, + { + "epoch": 3.9440011102629935, + "grad_norm": 0.6277918219566345, + "learning_rate": 1.0277826628672937e-07, + "loss": 0.1743, + "step": 56837 + }, + { + "epoch": 3.9440705017000903, + "grad_norm": 0.7624670267105103, + "learning_rate": 1.0252375071716503e-07, + "loss": 0.1985, + "step": 56838 + }, + { + "epoch": 3.9441398931371867, + "grad_norm": 0.8216585516929626, + "learning_rate": 1.022695505123461e-07, + "loss": 0.2704, + "step": 56839 + }, + { + "epoch": 3.9442092845742835, + "grad_norm": 1.763545274734497, + "learning_rate": 1.0201566567307197e-07, + "loss": 0.3992, + "step": 56840 + }, + { + "epoch": 3.94427867601138, + "grad_norm": 1.0088040828704834, + "learning_rate": 1.0176209620014199e-07, + "loss": 0.2028, + "step": 56841 + }, + { + "epoch": 3.9443480674484768, + "grad_norm": 0.8921161890029907, + "learning_rate": 1.0150884209436661e-07, + "loss": 0.2276, + "step": 56842 + }, + { + "epoch": 3.9444174588855736, + "grad_norm": 0.6792568564414978, + "learning_rate": 1.0125590335653412e-07, + "loss": 0.1628, + "step": 56843 + }, + { + "epoch": 3.9444868503226704, + "grad_norm": 0.6467457413673401, + "learning_rate": 1.0100327998744385e-07, + "loss": 0.1986, + "step": 56844 + }, + { + "epoch": 3.944556241759767, + "grad_norm": 1.6158583164215088, + "learning_rate": 1.0075097198789518e-07, + "loss": 0.3783, + "step": 56845 + }, + { + "epoch": 3.9446256331968637, + "grad_norm": 1.2298200130462646, + "learning_rate": 1.0049897935868747e-07, + "loss": 0.2405, + "step": 56846 + }, + { + "epoch": 3.94469502463396, + "grad_norm": 1.390190839767456, + "learning_rate": 1.0024730210060895e-07, + "loss": 0.2991, + "step": 56847 + }, + { + "epoch": 3.944764416071057, + "grad_norm": 1.7380541563034058, + "learning_rate": 9.999594021445901e-08, + "loss": 0.1989, + "step": 56848 + }, + { + "epoch": 3.9448338075081537, + "grad_norm": 0.6941360235214233, + "learning_rate": 9.974489370103702e-08, + "loss": 0.2236, + "step": 56849 + }, + { + "epoch": 3.94490319894525, + "grad_norm": 0.7933136820793152, + "learning_rate": 9.94941625611312e-08, + "loss": 0.2487, + "step": 56850 + }, + { + "epoch": 3.944972590382347, + "grad_norm": 0.6062384843826294, + "learning_rate": 9.924374679552984e-08, + "loss": 0.1909, + "step": 56851 + }, + { + "epoch": 3.9450419818194433, + "grad_norm": 0.8353455662727356, + "learning_rate": 9.899364640502118e-08, + "loss": 0.1881, + "step": 56852 + }, + { + "epoch": 3.94511137325654, + "grad_norm": 0.9535917639732361, + "learning_rate": 9.874386139039349e-08, + "loss": 0.2673, + "step": 56853 + }, + { + "epoch": 3.945180764693637, + "grad_norm": 0.777595043182373, + "learning_rate": 9.849439175245723e-08, + "loss": 0.228, + "step": 56854 + }, + { + "epoch": 3.9452501561307334, + "grad_norm": 0.7666022777557373, + "learning_rate": 9.824523749196734e-08, + "loss": 0.2167, + "step": 56855 + }, + { + "epoch": 3.94531954756783, + "grad_norm": 0.8741218447685242, + "learning_rate": 9.799639860973431e-08, + "loss": 0.3252, + "step": 56856 + }, + { + "epoch": 3.9453889390049266, + "grad_norm": 1.5515539646148682, + "learning_rate": 9.774787510653527e-08, + "loss": 0.3303, + "step": 56857 + }, + { + "epoch": 3.9454583304420234, + "grad_norm": 1.1822887659072876, + "learning_rate": 9.749966698315849e-08, + "loss": 0.4864, + "step": 56858 + }, + { + "epoch": 3.9455277218791203, + "grad_norm": 1.017049789428711, + "learning_rate": 9.725177424038112e-08, + "loss": 0.2001, + "step": 56859 + }, + { + "epoch": 3.9455971133162167, + "grad_norm": 1.3141052722930908, + "learning_rate": 9.700419687899143e-08, + "loss": 0.2975, + "step": 56860 + }, + { + "epoch": 3.9456665047533135, + "grad_norm": 0.6265122294425964, + "learning_rate": 9.675693489976655e-08, + "loss": 0.1645, + "step": 56861 + }, + { + "epoch": 3.94573589619041, + "grad_norm": 0.8818703889846802, + "learning_rate": 9.650998830349478e-08, + "loss": 0.2796, + "step": 56862 + }, + { + "epoch": 3.9458052876275067, + "grad_norm": 0.6868970394134521, + "learning_rate": 9.626335709094215e-08, + "loss": 0.1742, + "step": 56863 + }, + { + "epoch": 3.9458746790646035, + "grad_norm": 0.8683327436447144, + "learning_rate": 9.601704126289691e-08, + "loss": 0.209, + "step": 56864 + }, + { + "epoch": 3.9459440705017, + "grad_norm": 0.8546562194824219, + "learning_rate": 9.577104082013621e-08, + "loss": 0.1902, + "step": 56865 + }, + { + "epoch": 3.9460134619387968, + "grad_norm": 1.1646397113800049, + "learning_rate": 9.552535576342614e-08, + "loss": 0.3767, + "step": 56866 + }, + { + "epoch": 3.946082853375893, + "grad_norm": 0.8369969725608826, + "learning_rate": 9.527998609356603e-08, + "loss": 0.2637, + "step": 56867 + }, + { + "epoch": 3.94615224481299, + "grad_norm": 0.7244618535041809, + "learning_rate": 9.503493181131085e-08, + "loss": 0.1892, + "step": 56868 + }, + { + "epoch": 3.946221636250087, + "grad_norm": 0.8996216058731079, + "learning_rate": 9.479019291743774e-08, + "loss": 0.2257, + "step": 56869 + }, + { + "epoch": 3.9462910276871837, + "grad_norm": 1.0305759906768799, + "learning_rate": 9.454576941272386e-08, + "loss": 0.4339, + "step": 56870 + }, + { + "epoch": 3.94636041912428, + "grad_norm": 0.9751927852630615, + "learning_rate": 9.430166129793527e-08, + "loss": 0.2223, + "step": 56871 + }, + { + "epoch": 3.946429810561377, + "grad_norm": 0.7936865091323853, + "learning_rate": 9.405786857383803e-08, + "loss": 0.3095, + "step": 56872 + }, + { + "epoch": 3.9464992019984733, + "grad_norm": 0.5771082043647766, + "learning_rate": 9.381439124122038e-08, + "loss": 0.1386, + "step": 56873 + }, + { + "epoch": 3.94656859343557, + "grad_norm": 1.0346616506576538, + "learning_rate": 9.357122930083729e-08, + "loss": 0.2103, + "step": 56874 + }, + { + "epoch": 3.946637984872667, + "grad_norm": 0.9696524739265442, + "learning_rate": 9.33283827534548e-08, + "loss": 0.3557, + "step": 56875 + }, + { + "epoch": 3.9467073763097633, + "grad_norm": 1.1188898086547852, + "learning_rate": 9.308585159983896e-08, + "loss": 0.2261, + "step": 56876 + }, + { + "epoch": 3.94677676774686, + "grad_norm": 0.8654164671897888, + "learning_rate": 9.284363584076694e-08, + "loss": 0.2581, + "step": 56877 + }, + { + "epoch": 3.9468461591839565, + "grad_norm": 0.8678085803985596, + "learning_rate": 9.260173547699369e-08, + "loss": 0.2095, + "step": 56878 + }, + { + "epoch": 3.9469155506210534, + "grad_norm": 0.7844933271408081, + "learning_rate": 9.236015050928526e-08, + "loss": 0.2276, + "step": 56879 + }, + { + "epoch": 3.94698494205815, + "grad_norm": 1.603201150894165, + "learning_rate": 9.21188809384077e-08, + "loss": 0.3047, + "step": 56880 + }, + { + "epoch": 3.9470543334952466, + "grad_norm": 0.7024959325790405, + "learning_rate": 9.187792676511597e-08, + "loss": 0.1954, + "step": 56881 + }, + { + "epoch": 3.9471237249323434, + "grad_norm": 0.7177896499633789, + "learning_rate": 9.163728799017613e-08, + "loss": 0.2076, + "step": 56882 + }, + { + "epoch": 3.94719311636944, + "grad_norm": 0.8592459559440613, + "learning_rate": 9.13969646143431e-08, + "loss": 0.3037, + "step": 56883 + }, + { + "epoch": 3.9472625078065366, + "grad_norm": 0.7000651359558105, + "learning_rate": 9.115695663838298e-08, + "loss": 0.2461, + "step": 56884 + }, + { + "epoch": 3.9473318992436335, + "grad_norm": 1.1817437410354614, + "learning_rate": 9.091726406305068e-08, + "loss": 0.2512, + "step": 56885 + }, + { + "epoch": 3.94740129068073, + "grad_norm": 0.8980718851089478, + "learning_rate": 9.067788688909008e-08, + "loss": 0.1633, + "step": 56886 + }, + { + "epoch": 3.9474706821178267, + "grad_norm": 0.7800050377845764, + "learning_rate": 9.04388251172783e-08, + "loss": 0.2385, + "step": 56887 + }, + { + "epoch": 3.947540073554923, + "grad_norm": 0.9199396371841431, + "learning_rate": 9.020007874834812e-08, + "loss": 0.1998, + "step": 56888 + }, + { + "epoch": 3.94760946499202, + "grad_norm": 0.9621707201004028, + "learning_rate": 8.996164778307669e-08, + "loss": 0.3246, + "step": 56889 + }, + { + "epoch": 3.9476788564291168, + "grad_norm": 0.9755388498306274, + "learning_rate": 8.972353222219677e-08, + "loss": 0.2085, + "step": 56890 + }, + { + "epoch": 3.947748247866213, + "grad_norm": 0.6444093585014343, + "learning_rate": 8.948573206647437e-08, + "loss": 0.1841, + "step": 56891 + }, + { + "epoch": 3.94781763930331, + "grad_norm": 0.903191864490509, + "learning_rate": 8.924824731665338e-08, + "loss": 0.2731, + "step": 56892 + }, + { + "epoch": 3.9478870307404064, + "grad_norm": 0.7818691730499268, + "learning_rate": 8.901107797347763e-08, + "loss": 0.2322, + "step": 56893 + }, + { + "epoch": 3.947956422177503, + "grad_norm": 0.7905694842338562, + "learning_rate": 8.877422403770208e-08, + "loss": 0.2833, + "step": 56894 + }, + { + "epoch": 3.9480258136146, + "grad_norm": 0.9951932430267334, + "learning_rate": 8.853768551008168e-08, + "loss": 0.3241, + "step": 56895 + }, + { + "epoch": 3.948095205051697, + "grad_norm": 0.8706287741661072, + "learning_rate": 8.830146239134918e-08, + "loss": 0.2398, + "step": 56896 + }, + { + "epoch": 3.9481645964887933, + "grad_norm": 1.1489177942276, + "learning_rate": 8.806555468225952e-08, + "loss": 0.3173, + "step": 56897 + }, + { + "epoch": 3.94823398792589, + "grad_norm": 1.152843713760376, + "learning_rate": 8.782996238355656e-08, + "loss": 0.2637, + "step": 56898 + }, + { + "epoch": 3.9483033793629865, + "grad_norm": 0.7439231872558594, + "learning_rate": 8.759468549598415e-08, + "loss": 0.1874, + "step": 56899 + }, + { + "epoch": 3.9483727708000833, + "grad_norm": 1.2203878164291382, + "learning_rate": 8.735972402027503e-08, + "loss": 0.3158, + "step": 56900 + }, + { + "epoch": 3.94844216223718, + "grad_norm": 0.893233060836792, + "learning_rate": 8.712507795718416e-08, + "loss": 0.2154, + "step": 56901 + }, + { + "epoch": 3.9485115536742765, + "grad_norm": 1.345268726348877, + "learning_rate": 8.689074730745539e-08, + "loss": 0.2392, + "step": 56902 + }, + { + "epoch": 3.9485809451113734, + "grad_norm": 0.6467905640602112, + "learning_rate": 8.665673207181036e-08, + "loss": 0.181, + "step": 56903 + }, + { + "epoch": 3.9486503365484698, + "grad_norm": 0.735929548740387, + "learning_rate": 8.642303225099291e-08, + "loss": 0.1684, + "step": 56904 + }, + { + "epoch": 3.9487197279855666, + "grad_norm": 1.0454131364822388, + "learning_rate": 8.6189647845758e-08, + "loss": 0.2144, + "step": 56905 + }, + { + "epoch": 3.9487891194226634, + "grad_norm": 1.379931092262268, + "learning_rate": 8.595657885682728e-08, + "loss": 0.2927, + "step": 56906 + }, + { + "epoch": 3.94885851085976, + "grad_norm": 1.005521535873413, + "learning_rate": 8.57238252849335e-08, + "loss": 0.1978, + "step": 56907 + }, + { + "epoch": 3.9489279022968566, + "grad_norm": 0.9182401299476624, + "learning_rate": 8.54913871308205e-08, + "loss": 0.1707, + "step": 56908 + }, + { + "epoch": 3.948997293733953, + "grad_norm": 0.6612358689308167, + "learning_rate": 8.525926439522103e-08, + "loss": 0.1975, + "step": 56909 + }, + { + "epoch": 3.94906668517105, + "grad_norm": 2.2309188842773438, + "learning_rate": 8.502745707886783e-08, + "loss": 0.3919, + "step": 56910 + }, + { + "epoch": 3.9491360766081467, + "grad_norm": 0.7957470417022705, + "learning_rate": 8.479596518248256e-08, + "loss": 0.1925, + "step": 56911 + }, + { + "epoch": 3.949205468045243, + "grad_norm": 0.8907338976860046, + "learning_rate": 8.456478870680907e-08, + "loss": 0.2283, + "step": 56912 + }, + { + "epoch": 3.94927485948234, + "grad_norm": 0.7573046088218689, + "learning_rate": 8.433392765258008e-08, + "loss": 0.1683, + "step": 56913 + }, + { + "epoch": 3.9493442509194363, + "grad_norm": 1.0381308794021606, + "learning_rate": 8.410338202050616e-08, + "loss": 0.2119, + "step": 56914 + }, + { + "epoch": 3.949413642356533, + "grad_norm": 0.7523839473724365, + "learning_rate": 8.387315181132004e-08, + "loss": 0.1934, + "step": 56915 + }, + { + "epoch": 3.94948303379363, + "grad_norm": 0.9550448656082153, + "learning_rate": 8.36432370257656e-08, + "loss": 0.233, + "step": 56916 + }, + { + "epoch": 3.949552425230727, + "grad_norm": 1.4972076416015625, + "learning_rate": 8.341363766455335e-08, + "loss": 0.3016, + "step": 56917 + }, + { + "epoch": 3.949621816667823, + "grad_norm": 0.8838626742362976, + "learning_rate": 8.318435372841604e-08, + "loss": 0.2186, + "step": 56918 + }, + { + "epoch": 3.9496912081049196, + "grad_norm": 1.0229703187942505, + "learning_rate": 8.295538521806424e-08, + "loss": 0.2877, + "step": 56919 + }, + { + "epoch": 3.9497605995420164, + "grad_norm": 1.0181591510772705, + "learning_rate": 8.272673213424176e-08, + "loss": 0.2735, + "step": 56920 + }, + { + "epoch": 3.9498299909791132, + "grad_norm": 1.2647440433502197, + "learning_rate": 8.249839447764806e-08, + "loss": 0.3441, + "step": 56921 + }, + { + "epoch": 3.94989938241621, + "grad_norm": 0.9738195538520813, + "learning_rate": 8.22703722490159e-08, + "loss": 0.2583, + "step": 56922 + }, + { + "epoch": 3.9499687738533065, + "grad_norm": 0.6741347908973694, + "learning_rate": 8.204266544906691e-08, + "loss": 0.1396, + "step": 56923 + }, + { + "epoch": 3.9500381652904033, + "grad_norm": 0.9317405223846436, + "learning_rate": 8.181527407851164e-08, + "loss": 0.2648, + "step": 56924 + }, + { + "epoch": 3.9501075567274997, + "grad_norm": 0.884994387626648, + "learning_rate": 8.158819813808283e-08, + "loss": 0.3149, + "step": 56925 + }, + { + "epoch": 3.9501769481645965, + "grad_norm": 1.9370307922363281, + "learning_rate": 8.136143762847992e-08, + "loss": 0.3577, + "step": 56926 + }, + { + "epoch": 3.9502463396016934, + "grad_norm": 0.8704413175582886, + "learning_rate": 8.113499255042456e-08, + "loss": 0.2815, + "step": 56927 + }, + { + "epoch": 3.9503157310387897, + "grad_norm": 0.7795339226722717, + "learning_rate": 8.090886290463839e-08, + "loss": 0.1935, + "step": 56928 + }, + { + "epoch": 3.9503851224758866, + "grad_norm": 0.6946286559104919, + "learning_rate": 8.068304869182087e-08, + "loss": 0.2743, + "step": 56929 + }, + { + "epoch": 3.950454513912983, + "grad_norm": 0.8963982462882996, + "learning_rate": 8.04575499127047e-08, + "loss": 0.2373, + "step": 56930 + }, + { + "epoch": 3.95052390535008, + "grad_norm": 0.8390260934829712, + "learning_rate": 8.023236656798938e-08, + "loss": 0.2366, + "step": 56931 + }, + { + "epoch": 3.9505932967871766, + "grad_norm": 0.7995516061782837, + "learning_rate": 8.000749865838541e-08, + "loss": 0.2437, + "step": 56932 + }, + { + "epoch": 3.950662688224273, + "grad_norm": 0.8373805284500122, + "learning_rate": 7.978294618460336e-08, + "loss": 0.2621, + "step": 56933 + }, + { + "epoch": 3.95073207966137, + "grad_norm": 0.7538414597511292, + "learning_rate": 7.955870914735375e-08, + "loss": 0.2395, + "step": 56934 + }, + { + "epoch": 3.9508014710984662, + "grad_norm": 0.682843804359436, + "learning_rate": 7.933478754734713e-08, + "loss": 0.2102, + "step": 56935 + }, + { + "epoch": 3.950870862535563, + "grad_norm": 1.1275629997253418, + "learning_rate": 7.911118138528295e-08, + "loss": 0.2702, + "step": 56936 + }, + { + "epoch": 3.95094025397266, + "grad_norm": 1.0895947217941284, + "learning_rate": 7.888789066188284e-08, + "loss": 0.2857, + "step": 56937 + }, + { + "epoch": 3.9510096454097563, + "grad_norm": 0.6599977612495422, + "learning_rate": 7.866491537783516e-08, + "loss": 0.1659, + "step": 56938 + }, + { + "epoch": 3.951079036846853, + "grad_norm": 1.162617802619934, + "learning_rate": 7.844225553383933e-08, + "loss": 0.2202, + "step": 56939 + }, + { + "epoch": 3.9511484282839495, + "grad_norm": 1.6532012224197388, + "learning_rate": 7.8219911130617e-08, + "loss": 0.3857, + "step": 56940 + }, + { + "epoch": 3.9512178197210464, + "grad_norm": 0.9073059558868408, + "learning_rate": 7.799788216886762e-08, + "loss": 0.3151, + "step": 56941 + }, + { + "epoch": 3.951287211158143, + "grad_norm": 1.018707275390625, + "learning_rate": 7.77761686492684e-08, + "loss": 0.2363, + "step": 56942 + }, + { + "epoch": 3.95135660259524, + "grad_norm": 1.0885642766952515, + "learning_rate": 7.755477057255211e-08, + "loss": 0.2694, + "step": 56943 + }, + { + "epoch": 3.9514259940323364, + "grad_norm": 1.4816871881484985, + "learning_rate": 7.7333687939396e-08, + "loss": 0.413, + "step": 56944 + }, + { + "epoch": 3.951495385469433, + "grad_norm": 0.7819545865058899, + "learning_rate": 7.711292075051058e-08, + "loss": 0.2091, + "step": 56945 + }, + { + "epoch": 3.9515647769065296, + "grad_norm": 0.9376164674758911, + "learning_rate": 7.68924690065731e-08, + "loss": 0.2392, + "step": 56946 + }, + { + "epoch": 3.9516341683436265, + "grad_norm": 0.6834603548049927, + "learning_rate": 7.66723327083052e-08, + "loss": 0.2243, + "step": 56947 + }, + { + "epoch": 3.9517035597807233, + "grad_norm": 0.9152267575263977, + "learning_rate": 7.645251185638414e-08, + "loss": 0.2645, + "step": 56948 + }, + { + "epoch": 3.9517729512178197, + "grad_norm": 0.8854783177375793, + "learning_rate": 7.623300645150932e-08, + "loss": 0.177, + "step": 56949 + }, + { + "epoch": 3.9518423426549165, + "grad_norm": 1.0070828199386597, + "learning_rate": 7.601381649438022e-08, + "loss": 0.2631, + "step": 56950 + }, + { + "epoch": 3.951911734092013, + "grad_norm": 0.8031067252159119, + "learning_rate": 7.579494198567405e-08, + "loss": 0.1576, + "step": 56951 + }, + { + "epoch": 3.9519811255291097, + "grad_norm": 0.9905197620391846, + "learning_rate": 7.557638292609026e-08, + "loss": 0.3193, + "step": 56952 + }, + { + "epoch": 3.9520505169662066, + "grad_norm": 1.1760132312774658, + "learning_rate": 7.535813931632829e-08, + "loss": 0.3164, + "step": 56953 + }, + { + "epoch": 3.952119908403303, + "grad_norm": 0.8692094683647156, + "learning_rate": 7.514021115705428e-08, + "loss": 0.2574, + "step": 56954 + }, + { + "epoch": 3.9521892998404, + "grad_norm": 0.8465292453765869, + "learning_rate": 7.492259844897876e-08, + "loss": 0.2537, + "step": 56955 + }, + { + "epoch": 3.952258691277496, + "grad_norm": 1.368439793586731, + "learning_rate": 7.470530119277897e-08, + "loss": 0.2965, + "step": 56956 + }, + { + "epoch": 3.952328082714593, + "grad_norm": 0.8691703081130981, + "learning_rate": 7.448831938913214e-08, + "loss": 0.2229, + "step": 56957 + }, + { + "epoch": 3.95239747415169, + "grad_norm": 0.7801973223686218, + "learning_rate": 7.427165303873773e-08, + "loss": 0.2725, + "step": 56958 + }, + { + "epoch": 3.9524668655887862, + "grad_norm": 0.803302526473999, + "learning_rate": 7.405530214227297e-08, + "loss": 0.3096, + "step": 56959 + }, + { + "epoch": 3.952536257025883, + "grad_norm": 0.9706740379333496, + "learning_rate": 7.383926670042618e-08, + "loss": 0.2879, + "step": 56960 + }, + { + "epoch": 3.9526056484629795, + "grad_norm": 0.9449148178100586, + "learning_rate": 7.362354671386352e-08, + "loss": 0.1475, + "step": 56961 + }, + { + "epoch": 3.9526750399000763, + "grad_norm": 1.3003535270690918, + "learning_rate": 7.34081421832955e-08, + "loss": 0.2889, + "step": 56962 + }, + { + "epoch": 3.952744431337173, + "grad_norm": 0.6730390191078186, + "learning_rate": 7.319305310936608e-08, + "loss": 0.1822, + "step": 56963 + }, + { + "epoch": 3.9528138227742695, + "grad_norm": 0.8863288164138794, + "learning_rate": 7.29782794927858e-08, + "loss": 0.3147, + "step": 56964 + }, + { + "epoch": 3.9528832142113663, + "grad_norm": 1.4158495664596558, + "learning_rate": 7.276382133422077e-08, + "loss": 0.3437, + "step": 56965 + }, + { + "epoch": 3.9529526056484627, + "grad_norm": 0.8114379644393921, + "learning_rate": 7.254967863433715e-08, + "loss": 0.166, + "step": 56966 + }, + { + "epoch": 3.9530219970855596, + "grad_norm": 0.7861487865447998, + "learning_rate": 7.233585139382327e-08, + "loss": 0.1553, + "step": 56967 + }, + { + "epoch": 3.9530913885226564, + "grad_norm": 0.7396108508110046, + "learning_rate": 7.212233961335634e-08, + "loss": 0.2066, + "step": 56968 + }, + { + "epoch": 3.9531607799597532, + "grad_norm": 0.832703173160553, + "learning_rate": 7.190914329359144e-08, + "loss": 0.1982, + "step": 56969 + }, + { + "epoch": 3.9532301713968496, + "grad_norm": 0.9115283489227295, + "learning_rate": 7.169626243522798e-08, + "loss": 0.1518, + "step": 56970 + }, + { + "epoch": 3.9532995628339465, + "grad_norm": 0.8632514476776123, + "learning_rate": 7.1483697038921e-08, + "loss": 0.2665, + "step": 56971 + }, + { + "epoch": 3.953368954271043, + "grad_norm": 0.8964032530784607, + "learning_rate": 7.127144710534772e-08, + "loss": 0.1906, + "step": 56972 + }, + { + "epoch": 3.9534383457081397, + "grad_norm": 1.2465460300445557, + "learning_rate": 7.10595126351854e-08, + "loss": 0.3187, + "step": 56973 + }, + { + "epoch": 3.9535077371452365, + "grad_norm": 0.8861004114151001, + "learning_rate": 7.084789362907796e-08, + "loss": 0.1972, + "step": 56974 + }, + { + "epoch": 3.953577128582333, + "grad_norm": 0.773708701133728, + "learning_rate": 7.063659008772483e-08, + "loss": 0.2216, + "step": 56975 + }, + { + "epoch": 3.9536465200194297, + "grad_norm": 0.9585771560668945, + "learning_rate": 7.042560201176996e-08, + "loss": 0.2855, + "step": 56976 + }, + { + "epoch": 3.953715911456526, + "grad_norm": 0.7927048206329346, + "learning_rate": 7.021492940189056e-08, + "loss": 0.2683, + "step": 56977 + }, + { + "epoch": 3.953785302893623, + "grad_norm": 0.9988738298416138, + "learning_rate": 7.000457225874169e-08, + "loss": 0.2393, + "step": 56978 + }, + { + "epoch": 3.95385469433072, + "grad_norm": 0.7691332697868347, + "learning_rate": 6.979453058300056e-08, + "loss": 0.2194, + "step": 56979 + }, + { + "epoch": 3.953924085767816, + "grad_norm": 0.7593153715133667, + "learning_rate": 6.958480437532222e-08, + "loss": 0.2369, + "step": 56980 + }, + { + "epoch": 3.953993477204913, + "grad_norm": 0.8663933873176575, + "learning_rate": 6.93753936363728e-08, + "loss": 0.1657, + "step": 56981 + }, + { + "epoch": 3.9540628686420094, + "grad_norm": 0.9249807596206665, + "learning_rate": 6.916629836680732e-08, + "loss": 0.222, + "step": 56982 + }, + { + "epoch": 3.9541322600791062, + "grad_norm": 0.9224390983581543, + "learning_rate": 6.895751856729194e-08, + "loss": 0.2362, + "step": 56983 + }, + { + "epoch": 3.954201651516203, + "grad_norm": 1.1624374389648438, + "learning_rate": 6.874905423848166e-08, + "loss": 0.185, + "step": 56984 + }, + { + "epoch": 3.9542710429532995, + "grad_norm": 0.7772195935249329, + "learning_rate": 6.854090538104263e-08, + "loss": 0.2525, + "step": 56985 + }, + { + "epoch": 3.9543404343903963, + "grad_norm": 0.8056564927101135, + "learning_rate": 6.833307199561878e-08, + "loss": 0.2095, + "step": 56986 + }, + { + "epoch": 3.9544098258274927, + "grad_norm": 1.017774224281311, + "learning_rate": 6.812555408287624e-08, + "loss": 0.3213, + "step": 56987 + }, + { + "epoch": 3.9544792172645895, + "grad_norm": 0.6703678965568542, + "learning_rate": 6.791835164347005e-08, + "loss": 0.1352, + "step": 56988 + }, + { + "epoch": 3.9545486087016863, + "grad_norm": 1.0080589056015015, + "learning_rate": 6.771146467804412e-08, + "loss": 0.2452, + "step": 56989 + }, + { + "epoch": 3.9546180001387827, + "grad_norm": 1.2099381685256958, + "learning_rate": 6.75048931872646e-08, + "loss": 0.2714, + "step": 56990 + }, + { + "epoch": 3.9546873915758796, + "grad_norm": 0.9672263264656067, + "learning_rate": 6.729863717177543e-08, + "loss": 0.2201, + "step": 56991 + }, + { + "epoch": 3.954756783012976, + "grad_norm": 1.0276471376419067, + "learning_rate": 6.709269663223161e-08, + "loss": 0.2554, + "step": 56992 + }, + { + "epoch": 3.954826174450073, + "grad_norm": 0.8948322534561157, + "learning_rate": 6.68870715692771e-08, + "loss": 0.1769, + "step": 56993 + }, + { + "epoch": 3.9548955658871696, + "grad_norm": 1.0483283996582031, + "learning_rate": 6.6681761983578e-08, + "loss": 0.2678, + "step": 56994 + }, + { + "epoch": 3.9549649573242665, + "grad_norm": 1.0981048345565796, + "learning_rate": 6.647676787575607e-08, + "loss": 0.2833, + "step": 56995 + }, + { + "epoch": 3.955034348761363, + "grad_norm": 0.7665389776229858, + "learning_rate": 6.627208924647743e-08, + "loss": 0.1647, + "step": 56996 + }, + { + "epoch": 3.9551037401984597, + "grad_norm": 1.0368661880493164, + "learning_rate": 6.6067726096386e-08, + "loss": 0.3022, + "step": 56997 + }, + { + "epoch": 3.955173131635556, + "grad_norm": 0.8976597189903259, + "learning_rate": 6.586367842612573e-08, + "loss": 0.4546, + "step": 56998 + }, + { + "epoch": 3.955242523072653, + "grad_norm": 0.8247143030166626, + "learning_rate": 6.565994623632943e-08, + "loss": 0.2532, + "step": 56999 + }, + { + "epoch": 3.9553119145097497, + "grad_norm": 1.0342680215835571, + "learning_rate": 6.545652952766324e-08, + "loss": 0.3126, + "step": 57000 + }, + { + "epoch": 3.955381305946846, + "grad_norm": 1.7884758710861206, + "learning_rate": 6.525342830074888e-08, + "loss": 0.3869, + "step": 57001 + }, + { + "epoch": 3.955450697383943, + "grad_norm": 3.291559934616089, + "learning_rate": 6.505064255623027e-08, + "loss": 0.2626, + "step": 57002 + }, + { + "epoch": 3.9555200888210393, + "grad_norm": 1.3208966255187988, + "learning_rate": 6.484817229476248e-08, + "loss": 0.2829, + "step": 57003 + }, + { + "epoch": 3.955589480258136, + "grad_norm": 0.943096935749054, + "learning_rate": 6.46460175169672e-08, + "loss": 0.2274, + "step": 57004 + }, + { + "epoch": 3.955658871695233, + "grad_norm": 0.8501906394958496, + "learning_rate": 6.444417822348836e-08, + "loss": 0.2513, + "step": 57005 + }, + { + "epoch": 3.9557282631323294, + "grad_norm": 0.6544477343559265, + "learning_rate": 6.42426544149699e-08, + "loss": 0.2596, + "step": 57006 + }, + { + "epoch": 3.9557976545694262, + "grad_norm": 0.7151722311973572, + "learning_rate": 6.404144609204465e-08, + "loss": 0.1461, + "step": 57007 + }, + { + "epoch": 3.9558670460065226, + "grad_norm": 1.6260892152786255, + "learning_rate": 6.384055325533434e-08, + "loss": 0.3532, + "step": 57008 + }, + { + "epoch": 3.9559364374436194, + "grad_norm": 0.906325101852417, + "learning_rate": 6.363997590549398e-08, + "loss": 0.3283, + "step": 57009 + }, + { + "epoch": 3.9560058288807163, + "grad_norm": 1.096393346786499, + "learning_rate": 6.34397140431453e-08, + "loss": 0.1795, + "step": 57010 + }, + { + "epoch": 3.9560752203178127, + "grad_norm": 0.7893424034118652, + "learning_rate": 6.323976766892114e-08, + "loss": 0.2104, + "step": 57011 + }, + { + "epoch": 3.9561446117549095, + "grad_norm": 0.868158221244812, + "learning_rate": 6.304013678346543e-08, + "loss": 0.2159, + "step": 57012 + }, + { + "epoch": 3.956214003192006, + "grad_norm": 1.1590684652328491, + "learning_rate": 6.284082138738878e-08, + "loss": 0.3095, + "step": 57013 + }, + { + "epoch": 3.9562833946291027, + "grad_norm": 0.9563534259796143, + "learning_rate": 6.264182148132402e-08, + "loss": 0.1777, + "step": 57014 + }, + { + "epoch": 3.9563527860661996, + "grad_norm": 1.5995512008666992, + "learning_rate": 6.244313706590399e-08, + "loss": 0.3864, + "step": 57015 + }, + { + "epoch": 3.956422177503296, + "grad_norm": 0.7493828535079956, + "learning_rate": 6.22447681417615e-08, + "loss": 0.2479, + "step": 57016 + }, + { + "epoch": 3.9564915689403928, + "grad_norm": 0.6999644637107849, + "learning_rate": 6.204671470951828e-08, + "loss": 0.1792, + "step": 57017 + }, + { + "epoch": 3.956560960377489, + "grad_norm": 1.1258692741394043, + "learning_rate": 6.184897676979607e-08, + "loss": 0.3149, + "step": 57018 + }, + { + "epoch": 3.956630351814586, + "grad_norm": 1.4856085777282715, + "learning_rate": 6.165155432322767e-08, + "loss": 0.2053, + "step": 57019 + }, + { + "epoch": 3.956699743251683, + "grad_norm": 0.9063770174980164, + "learning_rate": 6.145444737042372e-08, + "loss": 0.2588, + "step": 57020 + }, + { + "epoch": 3.9567691346887797, + "grad_norm": 0.9234777092933655, + "learning_rate": 6.125765591200594e-08, + "loss": 0.3043, + "step": 57021 + }, + { + "epoch": 3.956838526125876, + "grad_norm": 0.6952356100082397, + "learning_rate": 6.106117994861826e-08, + "loss": 0.1948, + "step": 57022 + }, + { + "epoch": 3.956907917562973, + "grad_norm": 0.8552746772766113, + "learning_rate": 6.08650194808491e-08, + "loss": 0.26, + "step": 57023 + }, + { + "epoch": 3.9569773090000693, + "grad_norm": 1.2503650188446045, + "learning_rate": 6.06691745093424e-08, + "loss": 0.2175, + "step": 57024 + }, + { + "epoch": 3.957046700437166, + "grad_norm": 0.8312410712242126, + "learning_rate": 6.047364503470876e-08, + "loss": 0.2242, + "step": 57025 + }, + { + "epoch": 3.957116091874263, + "grad_norm": 0.8669193387031555, + "learning_rate": 6.027843105756991e-08, + "loss": 0.2346, + "step": 57026 + }, + { + "epoch": 3.9571854833113593, + "grad_norm": 1.4373087882995605, + "learning_rate": 6.008353257853649e-08, + "loss": 0.3238, + "step": 57027 + }, + { + "epoch": 3.957254874748456, + "grad_norm": 1.2080155611038208, + "learning_rate": 5.988894959821911e-08, + "loss": 0.3348, + "step": 57028 + }, + { + "epoch": 3.9573242661855526, + "grad_norm": 0.7432239055633545, + "learning_rate": 5.96946821172284e-08, + "loss": 0.2734, + "step": 57029 + }, + { + "epoch": 3.9573936576226494, + "grad_norm": 1.1177771091461182, + "learning_rate": 5.950073013619717e-08, + "loss": 0.3891, + "step": 57030 + }, + { + "epoch": 3.957463049059746, + "grad_norm": 0.9683877229690552, + "learning_rate": 5.9307093655724954e-08, + "loss": 0.3062, + "step": 57031 + }, + { + "epoch": 3.9575324404968426, + "grad_norm": 1.1732656955718994, + "learning_rate": 5.9113772676422376e-08, + "loss": 0.282, + "step": 57032 + }, + { + "epoch": 3.9576018319339394, + "grad_norm": 0.9694169163703918, + "learning_rate": 5.892076719891115e-08, + "loss": 0.2825, + "step": 57033 + }, + { + "epoch": 3.957671223371036, + "grad_norm": 0.7597578763961792, + "learning_rate": 5.872807722377971e-08, + "loss": 0.2253, + "step": 57034 + }, + { + "epoch": 3.9577406148081327, + "grad_norm": 0.7268590331077576, + "learning_rate": 5.853570275166087e-08, + "loss": 0.182, + "step": 57035 + }, + { + "epoch": 3.9578100062452295, + "grad_norm": 0.5999310612678528, + "learning_rate": 5.834364378314306e-08, + "loss": 0.1747, + "step": 57036 + }, + { + "epoch": 3.957879397682326, + "grad_norm": 0.8554619550704956, + "learning_rate": 5.8151900318836884e-08, + "loss": 0.2023, + "step": 57037 + }, + { + "epoch": 3.9579487891194227, + "grad_norm": 0.623141884803772, + "learning_rate": 5.796047235935298e-08, + "loss": 0.1979, + "step": 57038 + }, + { + "epoch": 3.958018180556519, + "grad_norm": 0.698543906211853, + "learning_rate": 5.7769359905290865e-08, + "loss": 0.2515, + "step": 57039 + }, + { + "epoch": 3.958087571993616, + "grad_norm": 1.2756075859069824, + "learning_rate": 5.757856295725006e-08, + "loss": 0.3523, + "step": 57040 + }, + { + "epoch": 3.9581569634307128, + "grad_norm": 0.6408989429473877, + "learning_rate": 5.7388081515841186e-08, + "loss": 0.1907, + "step": 57041 + }, + { + "epoch": 3.958226354867809, + "grad_norm": 0.6612829566001892, + "learning_rate": 5.7197915581674866e-08, + "loss": 0.1226, + "step": 57042 + }, + { + "epoch": 3.958295746304906, + "grad_norm": 1.0114635229110718, + "learning_rate": 5.700806515532841e-08, + "loss": 0.1817, + "step": 57043 + }, + { + "epoch": 3.9583651377420024, + "grad_norm": 1.9487956762313843, + "learning_rate": 5.681853023741246e-08, + "loss": 0.2681, + "step": 57044 + }, + { + "epoch": 3.958434529179099, + "grad_norm": 0.7854605913162231, + "learning_rate": 5.662931082852652e-08, + "loss": 0.2391, + "step": 57045 + }, + { + "epoch": 3.958503920616196, + "grad_norm": 0.7671117186546326, + "learning_rate": 5.6440406929259006e-08, + "loss": 0.2412, + "step": 57046 + }, + { + "epoch": 3.958573312053293, + "grad_norm": 1.0569151639938354, + "learning_rate": 5.625181854022055e-08, + "loss": 0.1987, + "step": 57047 + }, + { + "epoch": 3.9586427034903893, + "grad_norm": 0.9001344442367554, + "learning_rate": 5.606354566199956e-08, + "loss": 0.3301, + "step": 57048 + }, + { + "epoch": 3.958712094927486, + "grad_norm": 0.8892960548400879, + "learning_rate": 5.587558829518447e-08, + "loss": 0.2467, + "step": 57049 + }, + { + "epoch": 3.9587814863645825, + "grad_norm": 0.7884272336959839, + "learning_rate": 5.568794644038589e-08, + "loss": 0.2504, + "step": 57050 + }, + { + "epoch": 3.9588508778016793, + "grad_norm": 1.1730715036392212, + "learning_rate": 5.550062009818113e-08, + "loss": 0.4812, + "step": 57051 + }, + { + "epoch": 3.958920269238776, + "grad_norm": 1.1489965915679932, + "learning_rate": 5.531360926915863e-08, + "loss": 0.3548, + "step": 57052 + }, + { + "epoch": 3.9589896606758725, + "grad_norm": 0.9321498870849609, + "learning_rate": 5.5126913953917894e-08, + "loss": 0.2829, + "step": 57053 + }, + { + "epoch": 3.9590590521129694, + "grad_norm": 0.9936897158622742, + "learning_rate": 5.494053415304734e-08, + "loss": 0.174, + "step": 57054 + }, + { + "epoch": 3.9591284435500658, + "grad_norm": 1.603071689605713, + "learning_rate": 5.475446986713539e-08, + "loss": 0.348, + "step": 57055 + }, + { + "epoch": 3.9591978349871626, + "grad_norm": 1.117608666419983, + "learning_rate": 5.456872109675937e-08, + "loss": 0.1985, + "step": 57056 + }, + { + "epoch": 3.9592672264242594, + "grad_norm": 0.8916561007499695, + "learning_rate": 5.438328784252988e-08, + "loss": 0.2864, + "step": 57057 + }, + { + "epoch": 3.959336617861356, + "grad_norm": 2.281212329864502, + "learning_rate": 5.419817010500206e-08, + "loss": 0.2599, + "step": 57058 + }, + { + "epoch": 3.9594060092984527, + "grad_norm": 1.02318274974823, + "learning_rate": 5.40133678847754e-08, + "loss": 0.282, + "step": 57059 + }, + { + "epoch": 3.959475400735549, + "grad_norm": 0.7532090544700623, + "learning_rate": 5.382888118243834e-08, + "loss": 0.2942, + "step": 57060 + }, + { + "epoch": 3.959544792172646, + "grad_norm": 0.980690062046051, + "learning_rate": 5.3644709998568185e-08, + "loss": 0.2409, + "step": 57061 + }, + { + "epoch": 3.9596141836097427, + "grad_norm": 0.6799367070198059, + "learning_rate": 5.346085433374226e-08, + "loss": 0.1718, + "step": 57062 + }, + { + "epoch": 3.959683575046839, + "grad_norm": 0.9706527590751648, + "learning_rate": 5.3277314188548974e-08, + "loss": 0.2926, + "step": 57063 + }, + { + "epoch": 3.959752966483936, + "grad_norm": 0.8519181609153748, + "learning_rate": 5.309408956356565e-08, + "loss": 0.1987, + "step": 57064 + }, + { + "epoch": 3.9598223579210323, + "grad_norm": 1.1590181589126587, + "learning_rate": 5.29111804593585e-08, + "loss": 0.1692, + "step": 57065 + }, + { + "epoch": 3.959891749358129, + "grad_norm": 0.9833095073699951, + "learning_rate": 5.272858687651594e-08, + "loss": 0.3277, + "step": 57066 + }, + { + "epoch": 3.959961140795226, + "grad_norm": 0.9837543368339539, + "learning_rate": 5.254630881561529e-08, + "loss": 0.2406, + "step": 57067 + }, + { + "epoch": 3.9600305322323224, + "grad_norm": 0.6606717705726624, + "learning_rate": 5.236434627723385e-08, + "loss": 0.255, + "step": 57068 + }, + { + "epoch": 3.960099923669419, + "grad_norm": 1.126556396484375, + "learning_rate": 5.218269926193786e-08, + "loss": 0.3173, + "step": 57069 + }, + { + "epoch": 3.9601693151065156, + "grad_norm": 0.8913215398788452, + "learning_rate": 5.200136777030462e-08, + "loss": 0.2544, + "step": 57070 + }, + { + "epoch": 3.9602387065436124, + "grad_norm": 1.0234203338623047, + "learning_rate": 5.182035180291145e-08, + "loss": 0.2639, + "step": 57071 + }, + { + "epoch": 3.9603080979807093, + "grad_norm": 0.7886465191841125, + "learning_rate": 5.1639651360324556e-08, + "loss": 0.1749, + "step": 57072 + }, + { + "epoch": 3.960377489417806, + "grad_norm": 0.8355259299278259, + "learning_rate": 5.145926644311017e-08, + "loss": 0.2343, + "step": 57073 + }, + { + "epoch": 3.9604468808549025, + "grad_norm": 0.8635177612304688, + "learning_rate": 5.127919705184558e-08, + "loss": 0.2333, + "step": 57074 + }, + { + "epoch": 3.9605162722919993, + "grad_norm": 0.8088424801826477, + "learning_rate": 5.109944318709703e-08, + "loss": 0.2508, + "step": 57075 + }, + { + "epoch": 3.9605856637290957, + "grad_norm": 0.7939692735671997, + "learning_rate": 5.0920004849430716e-08, + "loss": 0.1982, + "step": 57076 + }, + { + "epoch": 3.9606550551661925, + "grad_norm": 0.9637049436569214, + "learning_rate": 5.074088203941285e-08, + "loss": 0.3196, + "step": 57077 + }, + { + "epoch": 3.9607244466032894, + "grad_norm": 0.8025941848754883, + "learning_rate": 5.056207475760966e-08, + "loss": 0.2711, + "step": 57078 + }, + { + "epoch": 3.9607938380403858, + "grad_norm": 1.4199867248535156, + "learning_rate": 5.0383583004587344e-08, + "loss": 0.2545, + "step": 57079 + }, + { + "epoch": 3.9608632294774826, + "grad_norm": 1.021238088607788, + "learning_rate": 5.020540678091212e-08, + "loss": 0.2522, + "step": 57080 + }, + { + "epoch": 3.960932620914579, + "grad_norm": 0.8687349557876587, + "learning_rate": 5.002754608713911e-08, + "loss": 0.2574, + "step": 57081 + }, + { + "epoch": 3.961002012351676, + "grad_norm": 0.9124884009361267, + "learning_rate": 4.985000092382341e-08, + "loss": 0.249, + "step": 57082 + }, + { + "epoch": 3.9610714037887726, + "grad_norm": 1.2390568256378174, + "learning_rate": 4.9672771291542355e-08, + "loss": 0.3458, + "step": 57083 + }, + { + "epoch": 3.961140795225869, + "grad_norm": 0.9051392674446106, + "learning_rate": 4.9495857190851034e-08, + "loss": 0.3049, + "step": 57084 + }, + { + "epoch": 3.961210186662966, + "grad_norm": 1.282503604888916, + "learning_rate": 4.9319258622304574e-08, + "loss": 0.3991, + "step": 57085 + }, + { + "epoch": 3.9612795781000623, + "grad_norm": 0.5600297451019287, + "learning_rate": 4.9142975586458085e-08, + "loss": 0.1529, + "step": 57086 + }, + { + "epoch": 3.961348969537159, + "grad_norm": 1.0180054903030396, + "learning_rate": 4.896700808386667e-08, + "loss": 0.3316, + "step": 57087 + }, + { + "epoch": 3.961418360974256, + "grad_norm": 1.0280232429504395, + "learning_rate": 4.8791356115096555e-08, + "loss": 0.2964, + "step": 57088 + }, + { + "epoch": 3.9614877524113523, + "grad_norm": 0.8422904014587402, + "learning_rate": 4.8616019680691736e-08, + "loss": 0.1828, + "step": 57089 + }, + { + "epoch": 3.961557143848449, + "grad_norm": 0.852855384349823, + "learning_rate": 4.844099878121844e-08, + "loss": 0.2142, + "step": 57090 + }, + { + "epoch": 3.9616265352855455, + "grad_norm": 0.8849018812179565, + "learning_rate": 4.826629341720956e-08, + "loss": 0.3156, + "step": 57091 + }, + { + "epoch": 3.9616959267226424, + "grad_norm": 0.8238602876663208, + "learning_rate": 4.8091903589231327e-08, + "loss": 0.2274, + "step": 57092 + }, + { + "epoch": 3.961765318159739, + "grad_norm": 1.0579304695129395, + "learning_rate": 4.791782929782773e-08, + "loss": 0.3235, + "step": 57093 + }, + { + "epoch": 3.9618347095968356, + "grad_norm": 0.917389452457428, + "learning_rate": 4.7744070543553896e-08, + "loss": 0.3357, + "step": 57094 + }, + { + "epoch": 3.9619041010339324, + "grad_norm": 0.9513282179832458, + "learning_rate": 4.757062732695383e-08, + "loss": 0.2503, + "step": 57095 + }, + { + "epoch": 3.961973492471029, + "grad_norm": 0.9028638005256653, + "learning_rate": 4.739749964857154e-08, + "loss": 0.1744, + "step": 57096 + }, + { + "epoch": 3.9620428839081256, + "grad_norm": 0.8618525862693787, + "learning_rate": 4.722468750897324e-08, + "loss": 0.2546, + "step": 57097 + }, + { + "epoch": 3.9621122753452225, + "grad_norm": 1.5760893821716309, + "learning_rate": 4.705219090868074e-08, + "loss": 0.3968, + "step": 57098 + }, + { + "epoch": 3.9621816667823193, + "grad_norm": 0.7150571346282959, + "learning_rate": 4.688000984824914e-08, + "loss": 0.2076, + "step": 57099 + }, + { + "epoch": 3.9622510582194157, + "grad_norm": 0.8181416988372803, + "learning_rate": 4.670814432822246e-08, + "loss": 0.2972, + "step": 57100 + }, + { + "epoch": 3.9623204496565125, + "grad_norm": 1.0365841388702393, + "learning_rate": 4.65365943491336e-08, + "loss": 0.2822, + "step": 57101 + }, + { + "epoch": 3.962389841093609, + "grad_norm": 0.9630046486854553, + "learning_rate": 4.636535991153767e-08, + "loss": 0.2583, + "step": 57102 + }, + { + "epoch": 3.9624592325307058, + "grad_norm": 0.8143907189369202, + "learning_rate": 4.619444101596759e-08, + "loss": 0.2499, + "step": 57103 + }, + { + "epoch": 3.9625286239678026, + "grad_norm": 0.8891222476959229, + "learning_rate": 4.602383766296736e-08, + "loss": 0.1986, + "step": 57104 + }, + { + "epoch": 3.962598015404899, + "grad_norm": 1.0792524814605713, + "learning_rate": 4.585354985306989e-08, + "loss": 0.2753, + "step": 57105 + }, + { + "epoch": 3.962667406841996, + "grad_norm": 1.127954125404358, + "learning_rate": 4.568357758681918e-08, + "loss": 0.2565, + "step": 57106 + }, + { + "epoch": 3.962736798279092, + "grad_norm": 0.758467435836792, + "learning_rate": 4.551392086474815e-08, + "loss": 0.1471, + "step": 57107 + }, + { + "epoch": 3.962806189716189, + "grad_norm": 0.7753663659095764, + "learning_rate": 4.53445796873897e-08, + "loss": 0.2541, + "step": 57108 + }, + { + "epoch": 3.962875581153286, + "grad_norm": 0.6257915496826172, + "learning_rate": 4.5175554055287836e-08, + "loss": 0.2638, + "step": 57109 + }, + { + "epoch": 3.9629449725903823, + "grad_norm": 1.2775899171829224, + "learning_rate": 4.500684396896437e-08, + "loss": 0.332, + "step": 57110 + }, + { + "epoch": 3.963014364027479, + "grad_norm": 0.7969328165054321, + "learning_rate": 4.4838449428963314e-08, + "loss": 0.2957, + "step": 57111 + }, + { + "epoch": 3.9630837554645755, + "grad_norm": 0.8827259540557861, + "learning_rate": 4.4670370435806466e-08, + "loss": 0.1644, + "step": 57112 + }, + { + "epoch": 3.9631531469016723, + "grad_norm": 0.8664671778678894, + "learning_rate": 4.450260699002673e-08, + "loss": 0.2125, + "step": 57113 + }, + { + "epoch": 3.963222538338769, + "grad_norm": 0.9247782826423645, + "learning_rate": 4.433515909215702e-08, + "loss": 0.1473, + "step": 57114 + }, + { + "epoch": 3.9632919297758655, + "grad_norm": 1.105466365814209, + "learning_rate": 4.4168026742719136e-08, + "loss": 0.2999, + "step": 57115 + }, + { + "epoch": 3.9633613212129624, + "grad_norm": 0.894433856010437, + "learning_rate": 4.400120994224599e-08, + "loss": 0.2344, + "step": 57116 + }, + { + "epoch": 3.9634307126500588, + "grad_norm": 0.8001218438148499, + "learning_rate": 4.3834708691270485e-08, + "loss": 0.2564, + "step": 57117 + }, + { + "epoch": 3.9635001040871556, + "grad_norm": 1.5542104244232178, + "learning_rate": 4.366852299031443e-08, + "loss": 0.3279, + "step": 57118 + }, + { + "epoch": 3.9635694955242524, + "grad_norm": 1.0034849643707275, + "learning_rate": 4.3502652839888526e-08, + "loss": 0.2875, + "step": 57119 + }, + { + "epoch": 3.963638886961349, + "grad_norm": 0.9107617735862732, + "learning_rate": 4.3337098240536774e-08, + "loss": 0.215, + "step": 57120 + }, + { + "epoch": 3.9637082783984456, + "grad_norm": 0.6924505829811096, + "learning_rate": 4.317185919276989e-08, + "loss": 0.2081, + "step": 57121 + }, + { + "epoch": 3.963777669835542, + "grad_norm": 1.0501848459243774, + "learning_rate": 4.3006935697109676e-08, + "loss": 0.2179, + "step": 57122 + }, + { + "epoch": 3.963847061272639, + "grad_norm": 1.2562308311462402, + "learning_rate": 4.284232775408903e-08, + "loss": 0.3049, + "step": 57123 + }, + { + "epoch": 3.9639164527097357, + "grad_norm": 1.0647615194320679, + "learning_rate": 4.267803536420756e-08, + "loss": 0.3625, + "step": 57124 + }, + { + "epoch": 3.9639858441468325, + "grad_norm": 0.9500256776809692, + "learning_rate": 4.251405852799817e-08, + "loss": 0.2782, + "step": 57125 + }, + { + "epoch": 3.964055235583929, + "grad_norm": 0.8907211422920227, + "learning_rate": 4.235039724597156e-08, + "loss": 0.2234, + "step": 57126 + }, + { + "epoch": 3.9641246270210257, + "grad_norm": 1.7518031597137451, + "learning_rate": 4.218705151864954e-08, + "loss": 0.5572, + "step": 57127 + }, + { + "epoch": 3.964194018458122, + "grad_norm": 0.7365005612373352, + "learning_rate": 4.202402134654282e-08, + "loss": 0.2691, + "step": 57128 + }, + { + "epoch": 3.964263409895219, + "grad_norm": 1.325566291809082, + "learning_rate": 4.186130673017319e-08, + "loss": 0.3144, + "step": 57129 + }, + { + "epoch": 3.964332801332316, + "grad_norm": 1.2483936548233032, + "learning_rate": 4.169890767005136e-08, + "loss": 0.2572, + "step": 57130 + }, + { + "epoch": 3.964402192769412, + "grad_norm": 0.7056350708007812, + "learning_rate": 4.1536824166688024e-08, + "loss": 0.1886, + "step": 57131 + }, + { + "epoch": 3.964471584206509, + "grad_norm": 0.8723031282424927, + "learning_rate": 4.13750562205939e-08, + "loss": 0.2633, + "step": 57132 + }, + { + "epoch": 3.9645409756436054, + "grad_norm": 1.3519368171691895, + "learning_rate": 4.121360383227968e-08, + "loss": 0.2235, + "step": 57133 + }, + { + "epoch": 3.9646103670807022, + "grad_norm": 1.3861312866210938, + "learning_rate": 4.105246700226717e-08, + "loss": 0.3829, + "step": 57134 + }, + { + "epoch": 3.964679758517799, + "grad_norm": 1.6461468935012817, + "learning_rate": 4.089164573104487e-08, + "loss": 0.2822, + "step": 57135 + }, + { + "epoch": 3.9647491499548955, + "grad_norm": 1.0508280992507935, + "learning_rate": 4.073114001913458e-08, + "loss": 0.2847, + "step": 57136 + }, + { + "epoch": 3.9648185413919923, + "grad_norm": 0.8144168853759766, + "learning_rate": 4.0570949867035915e-08, + "loss": 0.26, + "step": 57137 + }, + { + "epoch": 3.9648879328290887, + "grad_norm": 0.8566292524337769, + "learning_rate": 4.0411075275248456e-08, + "loss": 0.2652, + "step": 57138 + }, + { + "epoch": 3.9649573242661855, + "grad_norm": 0.8875882625579834, + "learning_rate": 4.025151624430512e-08, + "loss": 0.4206, + "step": 57139 + }, + { + "epoch": 3.9650267157032824, + "grad_norm": 0.9101676344871521, + "learning_rate": 4.00922727746722e-08, + "loss": 0.246, + "step": 57140 + }, + { + "epoch": 3.9650961071403787, + "grad_norm": 0.8270441293716431, + "learning_rate": 3.9933344866882603e-08, + "loss": 0.2237, + "step": 57141 + }, + { + "epoch": 3.9651654985774756, + "grad_norm": 0.7329089641571045, + "learning_rate": 3.977473252141373e-08, + "loss": 0.233, + "step": 57142 + }, + { + "epoch": 3.965234890014572, + "grad_norm": 0.8897101283073425, + "learning_rate": 3.961643573878737e-08, + "loss": 0.2487, + "step": 57143 + }, + { + "epoch": 3.965304281451669, + "grad_norm": 0.7953745126724243, + "learning_rate": 3.945845451949204e-08, + "loss": 0.2404, + "step": 57144 + }, + { + "epoch": 3.9653736728887656, + "grad_norm": 1.4105724096298218, + "learning_rate": 3.930078886403843e-08, + "loss": 0.3694, + "step": 57145 + }, + { + "epoch": 3.9654430643258625, + "grad_norm": 1.3908188343048096, + "learning_rate": 3.914343877290394e-08, + "loss": 0.2908, + "step": 57146 + }, + { + "epoch": 3.965512455762959, + "grad_norm": 0.9317730069160461, + "learning_rate": 3.8986404246599275e-08, + "loss": 0.2951, + "step": 57147 + }, + { + "epoch": 3.9655818472000552, + "grad_norm": 0.8701613545417786, + "learning_rate": 3.8829685285612925e-08, + "loss": 0.3706, + "step": 57148 + }, + { + "epoch": 3.965651238637152, + "grad_norm": 0.5700460076332092, + "learning_rate": 3.86732818904445e-08, + "loss": 0.1658, + "step": 57149 + }, + { + "epoch": 3.965720630074249, + "grad_norm": 0.8304933309555054, + "learning_rate": 3.85171940615936e-08, + "loss": 0.3052, + "step": 57150 + }, + { + "epoch": 3.9657900215113457, + "grad_norm": 1.1350808143615723, + "learning_rate": 3.836142179953761e-08, + "loss": 0.2273, + "step": 57151 + }, + { + "epoch": 3.965859412948442, + "grad_norm": 0.8379684686660767, + "learning_rate": 3.8205965104787244e-08, + "loss": 0.189, + "step": 57152 + }, + { + "epoch": 3.965928804385539, + "grad_norm": 0.642093300819397, + "learning_rate": 3.8050823977808793e-08, + "loss": 0.1801, + "step": 57153 + }, + { + "epoch": 3.9659981958226354, + "grad_norm": 0.9685921669006348, + "learning_rate": 3.789599841911296e-08, + "loss": 0.2194, + "step": 57154 + }, + { + "epoch": 3.966067587259732, + "grad_norm": 1.1030460596084595, + "learning_rate": 3.774148842918823e-08, + "loss": 0.2688, + "step": 57155 + }, + { + "epoch": 3.966136978696829, + "grad_norm": 0.7917542457580566, + "learning_rate": 3.758729400851202e-08, + "loss": 0.2588, + "step": 57156 + }, + { + "epoch": 3.9662063701339254, + "grad_norm": 4.847786903381348, + "learning_rate": 3.7433415157572816e-08, + "loss": 0.3106, + "step": 57157 + }, + { + "epoch": 3.9662757615710222, + "grad_norm": 0.9595717191696167, + "learning_rate": 3.727985187685912e-08, + "loss": 0.2152, + "step": 57158 + }, + { + "epoch": 3.9663451530081186, + "grad_norm": 1.052506685256958, + "learning_rate": 3.7126604166848323e-08, + "loss": 0.2012, + "step": 57159 + }, + { + "epoch": 3.9664145444452155, + "grad_norm": 1.03705632686615, + "learning_rate": 3.697367202804003e-08, + "loss": 0.3118, + "step": 57160 + }, + { + "epoch": 3.9664839358823123, + "grad_norm": 0.9636155366897583, + "learning_rate": 3.6821055460900534e-08, + "loss": 0.2368, + "step": 57161 + }, + { + "epoch": 3.9665533273194087, + "grad_norm": 0.9909364581108093, + "learning_rate": 3.6668754465929435e-08, + "loss": 0.1976, + "step": 57162 + }, + { + "epoch": 3.9666227187565055, + "grad_norm": 1.0461478233337402, + "learning_rate": 3.6516769043593025e-08, + "loss": 0.2451, + "step": 57163 + }, + { + "epoch": 3.966692110193602, + "grad_norm": 0.6748255491256714, + "learning_rate": 3.63650991943687e-08, + "loss": 0.168, + "step": 57164 + }, + { + "epoch": 3.9667615016306987, + "grad_norm": 0.877075731754303, + "learning_rate": 3.6213744918756065e-08, + "loss": 0.2146, + "step": 57165 + }, + { + "epoch": 3.9668308930677956, + "grad_norm": 1.4401602745056152, + "learning_rate": 3.6062706217210306e-08, + "loss": 0.4027, + "step": 57166 + }, + { + "epoch": 3.966900284504892, + "grad_norm": 0.7313410639762878, + "learning_rate": 3.5911983090208826e-08, + "loss": 0.3135, + "step": 57167 + }, + { + "epoch": 3.966969675941989, + "grad_norm": 0.9103754162788391, + "learning_rate": 3.576157553824011e-08, + "loss": 0.3007, + "step": 57168 + }, + { + "epoch": 3.967039067379085, + "grad_norm": 0.8424109816551208, + "learning_rate": 3.5611483561781565e-08, + "loss": 0.2023, + "step": 57169 + }, + { + "epoch": 3.967108458816182, + "grad_norm": 0.8999086022377014, + "learning_rate": 3.5461707161299484e-08, + "loss": 0.2141, + "step": 57170 + }, + { + "epoch": 3.967177850253279, + "grad_norm": 1.0361778736114502, + "learning_rate": 3.5312246337260155e-08, + "loss": 0.3049, + "step": 57171 + }, + { + "epoch": 3.9672472416903757, + "grad_norm": 0.8581795692443848, + "learning_rate": 3.516310109014098e-08, + "loss": 0.1758, + "step": 57172 + }, + { + "epoch": 3.967316633127472, + "grad_norm": 0.7426983118057251, + "learning_rate": 3.501427142041935e-08, + "loss": 0.2511, + "step": 57173 + }, + { + "epoch": 3.9673860245645685, + "grad_norm": 0.6454123854637146, + "learning_rate": 3.486575732855046e-08, + "loss": 0.1987, + "step": 57174 + }, + { + "epoch": 3.9674554160016653, + "grad_norm": 0.747606635093689, + "learning_rate": 3.471755881502281e-08, + "loss": 0.2246, + "step": 57175 + }, + { + "epoch": 3.967524807438762, + "grad_norm": 0.9136533737182617, + "learning_rate": 3.456967588029159e-08, + "loss": 0.2798, + "step": 57176 + }, + { + "epoch": 3.967594198875859, + "grad_norm": 1.0212287902832031, + "learning_rate": 3.442210852482308e-08, + "loss": 0.2362, + "step": 57177 + }, + { + "epoch": 3.9676635903129553, + "grad_norm": 0.6911206841468811, + "learning_rate": 3.4274856749083594e-08, + "loss": 0.1378, + "step": 57178 + }, + { + "epoch": 3.967732981750052, + "grad_norm": 0.9492454528808594, + "learning_rate": 3.412792055353941e-08, + "loss": 0.323, + "step": 57179 + }, + { + "epoch": 3.9678023731871486, + "grad_norm": 1.3806253671646118, + "learning_rate": 3.398129993865684e-08, + "loss": 0.2377, + "step": 57180 + }, + { + "epoch": 3.9678717646242454, + "grad_norm": 0.953353226184845, + "learning_rate": 3.383499490490216e-08, + "loss": 0.192, + "step": 57181 + }, + { + "epoch": 3.9679411560613422, + "grad_norm": 0.704200804233551, + "learning_rate": 3.368900545273057e-08, + "loss": 0.1431, + "step": 57182 + }, + { + "epoch": 3.9680105474984386, + "grad_norm": 1.1008763313293457, + "learning_rate": 3.354333158260836e-08, + "loss": 0.238, + "step": 57183 + }, + { + "epoch": 3.9680799389355355, + "grad_norm": 0.9423325657844543, + "learning_rate": 3.339797329497962e-08, + "loss": 0.3343, + "step": 57184 + }, + { + "epoch": 3.968149330372632, + "grad_norm": 0.5794802904129028, + "learning_rate": 3.325293059032175e-08, + "loss": 0.168, + "step": 57185 + }, + { + "epoch": 3.9682187218097287, + "grad_norm": 0.7836495041847229, + "learning_rate": 3.310820346908994e-08, + "loss": 0.1776, + "step": 57186 + }, + { + "epoch": 3.9682881132468255, + "grad_norm": 0.7997156381607056, + "learning_rate": 3.2963791931728274e-08, + "loss": 0.3029, + "step": 57187 + }, + { + "epoch": 3.968357504683922, + "grad_norm": 1.330369472503662, + "learning_rate": 3.281969597871415e-08, + "loss": 0.2972, + "step": 57188 + }, + { + "epoch": 3.9684268961210187, + "grad_norm": 0.9674791693687439, + "learning_rate": 3.267591561048056e-08, + "loss": 0.167, + "step": 57189 + }, + { + "epoch": 3.968496287558115, + "grad_norm": 0.7450001239776611, + "learning_rate": 3.253245082749379e-08, + "loss": 0.1701, + "step": 57190 + }, + { + "epoch": 3.968565678995212, + "grad_norm": 0.8903398513793945, + "learning_rate": 3.238930163019793e-08, + "loss": 0.2199, + "step": 57191 + }, + { + "epoch": 3.968635070432309, + "grad_norm": 1.2366873025894165, + "learning_rate": 3.224646801905929e-08, + "loss": 0.3557, + "step": 57192 + }, + { + "epoch": 3.968704461869405, + "grad_norm": 1.0478379726409912, + "learning_rate": 3.210394999451083e-08, + "loss": 0.2462, + "step": 57193 + }, + { + "epoch": 3.968773853306502, + "grad_norm": 0.78907710313797, + "learning_rate": 3.196174755701886e-08, + "loss": 0.2425, + "step": 57194 + }, + { + "epoch": 3.9688432447435984, + "grad_norm": 0.8843491077423096, + "learning_rate": 3.181986070702747e-08, + "loss": 0.1893, + "step": 57195 + }, + { + "epoch": 3.9689126361806952, + "grad_norm": 0.8384473919868469, + "learning_rate": 3.1678289444969645e-08, + "loss": 0.1872, + "step": 57196 + }, + { + "epoch": 3.968982027617792, + "grad_norm": 0.9367199540138245, + "learning_rate": 3.1537033771322774e-08, + "loss": 0.2343, + "step": 57197 + }, + { + "epoch": 3.969051419054889, + "grad_norm": 0.7875101566314697, + "learning_rate": 3.139609368649765e-08, + "loss": 0.2793, + "step": 57198 + }, + { + "epoch": 3.9691208104919853, + "grad_norm": 0.8084964156150818, + "learning_rate": 3.125546919097166e-08, + "loss": 0.1754, + "step": 57199 + }, + { + "epoch": 3.9691902019290817, + "grad_norm": 2.932049512863159, + "learning_rate": 3.11151602851667e-08, + "loss": 0.3162, + "step": 57200 + }, + { + "epoch": 3.9692595933661785, + "grad_norm": 0.995131254196167, + "learning_rate": 3.0975166969526846e-08, + "loss": 0.2467, + "step": 57201 + }, + { + "epoch": 3.9693289848032753, + "grad_norm": 0.7426882386207581, + "learning_rate": 3.08354892445073e-08, + "loss": 0.2861, + "step": 57202 + }, + { + "epoch": 3.969398376240372, + "grad_norm": 1.2499583959579468, + "learning_rate": 3.069612711054104e-08, + "loss": 0.2745, + "step": 57203 + }, + { + "epoch": 3.9694677676774686, + "grad_norm": 0.8414176106452942, + "learning_rate": 3.055708056807216e-08, + "loss": 0.2788, + "step": 57204 + }, + { + "epoch": 3.9695371591145654, + "grad_norm": 0.8263970613479614, + "learning_rate": 3.0418349617522545e-08, + "loss": 0.2916, + "step": 57205 + }, + { + "epoch": 3.969606550551662, + "grad_norm": 0.8438743352890015, + "learning_rate": 3.027993425935849e-08, + "loss": 0.2671, + "step": 57206 + }, + { + "epoch": 3.9696759419887586, + "grad_norm": 0.687579333782196, + "learning_rate": 3.014183449399077e-08, + "loss": 0.2036, + "step": 57207 + }, + { + "epoch": 3.9697453334258554, + "grad_norm": 0.9698508977890015, + "learning_rate": 3.000405032187459e-08, + "loss": 0.2582, + "step": 57208 + }, + { + "epoch": 3.969814724862952, + "grad_norm": 0.7848283052444458, + "learning_rate": 2.986658174343182e-08, + "loss": 0.1915, + "step": 57209 + }, + { + "epoch": 3.9698841163000487, + "grad_norm": 1.8347915410995483, + "learning_rate": 2.9729428759106558e-08, + "loss": 0.2415, + "step": 57210 + }, + { + "epoch": 3.969953507737145, + "grad_norm": 0.8545162081718445, + "learning_rate": 2.959259136932069e-08, + "loss": 0.3172, + "step": 57211 + }, + { + "epoch": 3.970022899174242, + "grad_norm": 0.9162880778312683, + "learning_rate": 2.9456069574518296e-08, + "loss": 0.2769, + "step": 57212 + }, + { + "epoch": 3.9700922906113387, + "grad_norm": 0.9946667551994324, + "learning_rate": 2.931986337512127e-08, + "loss": 0.1526, + "step": 57213 + }, + { + "epoch": 3.970161682048435, + "grad_norm": 1.147202491760254, + "learning_rate": 2.9183972771562596e-08, + "loss": 0.261, + "step": 57214 + }, + { + "epoch": 3.970231073485532, + "grad_norm": 0.8698785305023193, + "learning_rate": 2.9048397764275258e-08, + "loss": 0.3209, + "step": 57215 + }, + { + "epoch": 3.9703004649226283, + "grad_norm": 0.8553287386894226, + "learning_rate": 2.8913138353681145e-08, + "loss": 0.1948, + "step": 57216 + }, + { + "epoch": 3.970369856359725, + "grad_norm": 0.8770666718482971, + "learning_rate": 2.877819454020214e-08, + "loss": 0.2524, + "step": 57217 + }, + { + "epoch": 3.970439247796822, + "grad_norm": 1.1379314661026, + "learning_rate": 2.8643566324282334e-08, + "loss": 0.2005, + "step": 57218 + }, + { + "epoch": 3.9705086392339184, + "grad_norm": 0.8192735314369202, + "learning_rate": 2.8509253706332505e-08, + "loss": 0.2629, + "step": 57219 + }, + { + "epoch": 3.9705780306710152, + "grad_norm": 0.8523711562156677, + "learning_rate": 2.8375256686774542e-08, + "loss": 0.2782, + "step": 57220 + }, + { + "epoch": 3.9706474221081116, + "grad_norm": 0.8764416575431824, + "learning_rate": 2.824157526604143e-08, + "loss": 0.2103, + "step": 57221 + }, + { + "epoch": 3.9707168135452084, + "grad_norm": 0.6124439835548401, + "learning_rate": 2.8108209444555055e-08, + "loss": 0.1928, + "step": 57222 + }, + { + "epoch": 3.9707862049823053, + "grad_norm": 0.8710631132125854, + "learning_rate": 2.79751592227262e-08, + "loss": 0.1811, + "step": 57223 + }, + { + "epoch": 3.970855596419402, + "grad_norm": 0.7997090220451355, + "learning_rate": 2.7842424600987847e-08, + "loss": 0.26, + "step": 57224 + }, + { + "epoch": 3.9709249878564985, + "grad_norm": 1.3086423873901367, + "learning_rate": 2.7710005579739683e-08, + "loss": 0.3989, + "step": 57225 + }, + { + "epoch": 3.9709943792935953, + "grad_norm": 0.9428240060806274, + "learning_rate": 2.7577902159425794e-08, + "loss": 0.3475, + "step": 57226 + }, + { + "epoch": 3.9710637707306917, + "grad_norm": 1.4015905857086182, + "learning_rate": 2.744611434043476e-08, + "loss": 0.3144, + "step": 57227 + }, + { + "epoch": 3.9711331621677886, + "grad_norm": 0.816274106502533, + "learning_rate": 2.7314642123210665e-08, + "loss": 0.3325, + "step": 57228 + }, + { + "epoch": 3.9712025536048854, + "grad_norm": 1.1813260316848755, + "learning_rate": 2.7183485508153195e-08, + "loss": 0.39, + "step": 57229 + }, + { + "epoch": 3.9712719450419818, + "grad_norm": 1.0630292892456055, + "learning_rate": 2.705264449567313e-08, + "loss": 0.2571, + "step": 57230 + }, + { + "epoch": 3.9713413364790786, + "grad_norm": 0.8155795335769653, + "learning_rate": 2.6922119086192355e-08, + "loss": 0.1621, + "step": 57231 + }, + { + "epoch": 3.971410727916175, + "grad_norm": 1.30192232131958, + "learning_rate": 2.679190928012165e-08, + "loss": 0.2578, + "step": 57232 + }, + { + "epoch": 3.971480119353272, + "grad_norm": 0.9049877524375916, + "learning_rate": 2.6662015077860703e-08, + "loss": 0.2429, + "step": 57233 + }, + { + "epoch": 3.9715495107903687, + "grad_norm": 1.0380829572677612, + "learning_rate": 2.6532436479831392e-08, + "loss": 0.3812, + "step": 57234 + }, + { + "epoch": 3.971618902227465, + "grad_norm": 0.7644748091697693, + "learning_rate": 2.64031734864445e-08, + "loss": 0.1216, + "step": 57235 + }, + { + "epoch": 3.971688293664562, + "grad_norm": 0.9089861512184143, + "learning_rate": 2.6274226098099707e-08, + "loss": 0.1881, + "step": 57236 + }, + { + "epoch": 3.9717576851016583, + "grad_norm": 0.9027430415153503, + "learning_rate": 2.6145594315218902e-08, + "loss": 0.2075, + "step": 57237 + }, + { + "epoch": 3.971827076538755, + "grad_norm": 0.9932993054389954, + "learning_rate": 2.601727813819066e-08, + "loss": 0.2235, + "step": 57238 + }, + { + "epoch": 3.971896467975852, + "grad_norm": 0.8697934150695801, + "learning_rate": 2.5889277567425763e-08, + "loss": 0.2796, + "step": 57239 + }, + { + "epoch": 3.9719658594129483, + "grad_norm": 0.7655712366104126, + "learning_rate": 2.576159260332389e-08, + "loss": 0.1911, + "step": 57240 + }, + { + "epoch": 3.972035250850045, + "grad_norm": 1.583865761756897, + "learning_rate": 2.563422324629583e-08, + "loss": 0.4614, + "step": 57241 + }, + { + "epoch": 3.9721046422871416, + "grad_norm": 0.49719953536987305, + "learning_rate": 2.5507169496752357e-08, + "loss": 0.1286, + "step": 57242 + }, + { + "epoch": 3.9721740337242384, + "grad_norm": 1.0084596872329712, + "learning_rate": 2.5380431355070955e-08, + "loss": 0.2833, + "step": 57243 + }, + { + "epoch": 3.972243425161335, + "grad_norm": 0.9194489121437073, + "learning_rate": 2.5254008821673504e-08, + "loss": 0.3332, + "step": 57244 + }, + { + "epoch": 3.9723128165984316, + "grad_norm": 1.152511477470398, + "learning_rate": 2.5127901896937478e-08, + "loss": 0.3666, + "step": 57245 + }, + { + "epoch": 3.9723822080355284, + "grad_norm": 0.7835365533828735, + "learning_rate": 2.5002110581284764e-08, + "loss": 0.1815, + "step": 57246 + }, + { + "epoch": 3.972451599472625, + "grad_norm": 1.1589552164077759, + "learning_rate": 2.487663487509284e-08, + "loss": 0.2902, + "step": 57247 + }, + { + "epoch": 3.9725209909097217, + "grad_norm": 0.6754534840583801, + "learning_rate": 2.4751474778772488e-08, + "loss": 0.2139, + "step": 57248 + }, + { + "epoch": 3.9725903823468185, + "grad_norm": 1.2363777160644531, + "learning_rate": 2.4626630292712283e-08, + "loss": 0.2164, + "step": 57249 + }, + { + "epoch": 3.9726597737839153, + "grad_norm": 0.8803032636642456, + "learning_rate": 2.4502101417300806e-08, + "loss": 0.3113, + "step": 57250 + }, + { + "epoch": 3.9727291652210117, + "grad_norm": 0.9673585295677185, + "learning_rate": 2.4377888152948834e-08, + "loss": 0.253, + "step": 57251 + }, + { + "epoch": 3.9727985566581085, + "grad_norm": 1.011950969696045, + "learning_rate": 2.425399050002275e-08, + "loss": 0.2357, + "step": 57252 + }, + { + "epoch": 3.972867948095205, + "grad_norm": 1.0126889944076538, + "learning_rate": 2.413040845893333e-08, + "loss": 0.2908, + "step": 57253 + }, + { + "epoch": 3.9729373395323018, + "grad_norm": 0.7788828611373901, + "learning_rate": 2.400714203006915e-08, + "loss": 0.2229, + "step": 57254 + }, + { + "epoch": 3.9730067309693986, + "grad_norm": 1.1234625577926636, + "learning_rate": 2.3884191213807695e-08, + "loss": 0.236, + "step": 57255 + }, + { + "epoch": 3.973076122406495, + "grad_norm": 0.790025532245636, + "learning_rate": 2.3761556010548636e-08, + "loss": 0.1969, + "step": 57256 + }, + { + "epoch": 3.973145513843592, + "grad_norm": 0.9593062400817871, + "learning_rate": 2.3639236420669454e-08, + "loss": 0.3672, + "step": 57257 + }, + { + "epoch": 3.973214905280688, + "grad_norm": 1.108511209487915, + "learning_rate": 2.3517232444569824e-08, + "loss": 0.3155, + "step": 57258 + }, + { + "epoch": 3.973284296717785, + "grad_norm": 1.2909435033798218, + "learning_rate": 2.339554408262723e-08, + "loss": 0.3755, + "step": 57259 + }, + { + "epoch": 3.973353688154882, + "grad_norm": 1.1132056713104248, + "learning_rate": 2.327417133521914e-08, + "loss": 0.2673, + "step": 57260 + }, + { + "epoch": 3.9734230795919783, + "grad_norm": 1.0144819021224976, + "learning_rate": 2.315311420273414e-08, + "loss": 0.3765, + "step": 57261 + }, + { + "epoch": 3.973492471029075, + "grad_norm": 0.6941230297088623, + "learning_rate": 2.30323726855608e-08, + "loss": 0.1955, + "step": 57262 + }, + { + "epoch": 3.9735618624661715, + "grad_norm": 0.8030107021331787, + "learning_rate": 2.2911946784076598e-08, + "loss": 0.1575, + "step": 57263 + }, + { + "epoch": 3.9736312539032683, + "grad_norm": 0.8103191256523132, + "learning_rate": 2.2791836498647913e-08, + "loss": 0.1771, + "step": 57264 + }, + { + "epoch": 3.973700645340365, + "grad_norm": 1.0090429782867432, + "learning_rate": 2.267204182967442e-08, + "loss": 0.2121, + "step": 57265 + }, + { + "epoch": 3.9737700367774615, + "grad_norm": 0.8350775837898254, + "learning_rate": 2.255256277752249e-08, + "loss": 0.1991, + "step": 57266 + }, + { + "epoch": 3.9738394282145584, + "grad_norm": 0.9537824392318726, + "learning_rate": 2.2433399342580707e-08, + "loss": 0.2411, + "step": 57267 + }, + { + "epoch": 3.9739088196516548, + "grad_norm": 0.6237894296646118, + "learning_rate": 2.2314551525215442e-08, + "loss": 0.1431, + "step": 57268 + }, + { + "epoch": 3.9739782110887516, + "grad_norm": 0.5317535996437073, + "learning_rate": 2.219601932579307e-08, + "loss": 0.1535, + "step": 57269 + }, + { + "epoch": 3.9740476025258484, + "grad_norm": 0.974487841129303, + "learning_rate": 2.207780274470217e-08, + "loss": 0.2848, + "step": 57270 + }, + { + "epoch": 3.974116993962945, + "grad_norm": 0.9322883486747742, + "learning_rate": 2.1959901782320213e-08, + "loss": 0.3464, + "step": 57271 + }, + { + "epoch": 3.9741863854000417, + "grad_norm": 1.7569419145584106, + "learning_rate": 2.1842316439002475e-08, + "loss": 0.2472, + "step": 57272 + }, + { + "epoch": 3.974255776837138, + "grad_norm": 0.7823396325111389, + "learning_rate": 2.1725046715137532e-08, + "loss": 0.1756, + "step": 57273 + }, + { + "epoch": 3.974325168274235, + "grad_norm": 0.8733628392219543, + "learning_rate": 2.1608092611080654e-08, + "loss": 0.3145, + "step": 57274 + }, + { + "epoch": 3.9743945597113317, + "grad_norm": 0.6588965058326721, + "learning_rate": 2.149145412720932e-08, + "loss": 0.1384, + "step": 57275 + }, + { + "epoch": 3.9744639511484285, + "grad_norm": 1.17023766040802, + "learning_rate": 2.1375131263889902e-08, + "loss": 0.2594, + "step": 57276 + }, + { + "epoch": 3.974533342585525, + "grad_norm": 0.7772045135498047, + "learning_rate": 2.1259124021488774e-08, + "loss": 0.2051, + "step": 57277 + }, + { + "epoch": 3.9746027340226218, + "grad_norm": 0.8838437795639038, + "learning_rate": 2.114343240037231e-08, + "loss": 0.3484, + "step": 57278 + }, + { + "epoch": 3.974672125459718, + "grad_norm": 1.0207487344741821, + "learning_rate": 2.102805640090688e-08, + "loss": 0.2597, + "step": 57279 + }, + { + "epoch": 3.974741516896815, + "grad_norm": 0.9046889543533325, + "learning_rate": 2.0912996023458864e-08, + "loss": 0.1886, + "step": 57280 + }, + { + "epoch": 3.974810908333912, + "grad_norm": 0.8551197648048401, + "learning_rate": 2.0798251268383527e-08, + "loss": 0.1762, + "step": 57281 + }, + { + "epoch": 3.974880299771008, + "grad_norm": 0.9729191660881042, + "learning_rate": 2.0683822136047247e-08, + "loss": 0.2275, + "step": 57282 + }, + { + "epoch": 3.974949691208105, + "grad_norm": 0.5909157991409302, + "learning_rate": 2.0569708626816396e-08, + "loss": 0.1397, + "step": 57283 + }, + { + "epoch": 3.9750190826452014, + "grad_norm": 0.7906000018119812, + "learning_rate": 2.0455910741046248e-08, + "loss": 0.2382, + "step": 57284 + }, + { + "epoch": 3.9750884740822983, + "grad_norm": 0.7568568587303162, + "learning_rate": 2.034242847910317e-08, + "loss": 0.2018, + "step": 57285 + }, + { + "epoch": 3.975157865519395, + "grad_norm": 0.8814495205879211, + "learning_rate": 2.0229261841331336e-08, + "loss": 0.1773, + "step": 57286 + }, + { + "epoch": 3.9752272569564915, + "grad_norm": 1.1358222961425781, + "learning_rate": 2.011641082809712e-08, + "loss": 0.2457, + "step": 57287 + }, + { + "epoch": 3.9752966483935883, + "grad_norm": 0.8163488507270813, + "learning_rate": 2.0003875439755794e-08, + "loss": 0.1996, + "step": 57288 + }, + { + "epoch": 3.9753660398306847, + "grad_norm": 1.207836627960205, + "learning_rate": 1.9891655676662622e-08, + "loss": 0.1953, + "step": 57289 + }, + { + "epoch": 3.9754354312677815, + "grad_norm": 0.9341457486152649, + "learning_rate": 1.9779751539172886e-08, + "loss": 0.3067, + "step": 57290 + }, + { + "epoch": 3.9755048227048784, + "grad_norm": 1.0474742650985718, + "learning_rate": 1.9668163027641852e-08, + "loss": 0.3488, + "step": 57291 + }, + { + "epoch": 3.9755742141419748, + "grad_norm": 0.6708047389984131, + "learning_rate": 1.9556890142413687e-08, + "loss": 0.1729, + "step": 57292 + }, + { + "epoch": 3.9756436055790716, + "grad_norm": 1.2895194292068481, + "learning_rate": 1.9445932883843665e-08, + "loss": 0.2697, + "step": 57293 + }, + { + "epoch": 3.975712997016168, + "grad_norm": 0.6393311619758606, + "learning_rate": 1.9335291252287057e-08, + "loss": 0.1397, + "step": 57294 + }, + { + "epoch": 3.975782388453265, + "grad_norm": 1.3945989608764648, + "learning_rate": 1.9224965248088033e-08, + "loss": 0.1787, + "step": 57295 + }, + { + "epoch": 3.9758517798903616, + "grad_norm": 1.2519322633743286, + "learning_rate": 1.9114954871590763e-08, + "loss": 0.3658, + "step": 57296 + }, + { + "epoch": 3.975921171327458, + "grad_norm": 0.8499801158905029, + "learning_rate": 1.9005260123161615e-08, + "loss": 0.2586, + "step": 57297 + }, + { + "epoch": 3.975990562764555, + "grad_norm": 1.0293197631835938, + "learning_rate": 1.889588100312256e-08, + "loss": 0.1626, + "step": 57298 + }, + { + "epoch": 3.9760599542016513, + "grad_norm": 1.1160553693771362, + "learning_rate": 1.878681751182887e-08, + "loss": 0.2015, + "step": 57299 + }, + { + "epoch": 3.976129345638748, + "grad_norm": 0.9537289142608643, + "learning_rate": 1.8678069649624708e-08, + "loss": 0.1891, + "step": 57300 + }, + { + "epoch": 3.976198737075845, + "grad_norm": 0.4401837885379791, + "learning_rate": 1.8569637416854245e-08, + "loss": 0.145, + "step": 57301 + }, + { + "epoch": 3.9762681285129418, + "grad_norm": 0.9073216915130615, + "learning_rate": 1.8461520813861654e-08, + "loss": 0.2985, + "step": 57302 + }, + { + "epoch": 3.976337519950038, + "grad_norm": 0.8332874178886414, + "learning_rate": 1.835371984098e-08, + "loss": 0.2671, + "step": 57303 + }, + { + "epoch": 3.976406911387135, + "grad_norm": 0.7945040464401245, + "learning_rate": 1.824623449856455e-08, + "loss": 0.2789, + "step": 57304 + }, + { + "epoch": 3.9764763028242314, + "grad_norm": 0.8825367093086243, + "learning_rate": 1.8139064786937275e-08, + "loss": 0.301, + "step": 57305 + }, + { + "epoch": 3.976545694261328, + "grad_norm": 0.6587364077568054, + "learning_rate": 1.8032210706453446e-08, + "loss": 0.2243, + "step": 57306 + }, + { + "epoch": 3.976615085698425, + "grad_norm": 0.7102433443069458, + "learning_rate": 1.7925672257435024e-08, + "loss": 0.2231, + "step": 57307 + }, + { + "epoch": 3.9766844771355214, + "grad_norm": 1.1060200929641724, + "learning_rate": 1.7819449440226177e-08, + "loss": 0.2443, + "step": 57308 + }, + { + "epoch": 3.9767538685726183, + "grad_norm": 0.7005428075790405, + "learning_rate": 1.771354225517108e-08, + "loss": 0.2119, + "step": 57309 + }, + { + "epoch": 3.9768232600097146, + "grad_norm": 0.6715770363807678, + "learning_rate": 1.760795070258059e-08, + "loss": 0.1748, + "step": 57310 + }, + { + "epoch": 3.9768926514468115, + "grad_norm": 3.475968360900879, + "learning_rate": 1.750267478280998e-08, + "loss": 0.3477, + "step": 57311 + }, + { + "epoch": 3.9769620428839083, + "grad_norm": 1.0930116176605225, + "learning_rate": 1.7397714496181216e-08, + "loss": 0.2112, + "step": 57312 + }, + { + "epoch": 3.9770314343210047, + "grad_norm": 1.1621880531311035, + "learning_rate": 1.7293069843027365e-08, + "loss": 0.2783, + "step": 57313 + }, + { + "epoch": 3.9771008257581015, + "grad_norm": 0.7774105668067932, + "learning_rate": 1.7188740823681492e-08, + "loss": 0.2443, + "step": 57314 + }, + { + "epoch": 3.977170217195198, + "grad_norm": 1.1707552671432495, + "learning_rate": 1.7084727438465566e-08, + "loss": 0.3018, + "step": 57315 + }, + { + "epoch": 3.9772396086322948, + "grad_norm": 1.0531367063522339, + "learning_rate": 1.6981029687712645e-08, + "loss": 0.2475, + "step": 57316 + }, + { + "epoch": 3.9773090000693916, + "grad_norm": 0.8386579751968384, + "learning_rate": 1.6877647571755806e-08, + "loss": 0.2477, + "step": 57317 + }, + { + "epoch": 3.977378391506488, + "grad_norm": 1.0137784481048584, + "learning_rate": 1.6774581090917008e-08, + "loss": 0.2615, + "step": 57318 + }, + { + "epoch": 3.977447782943585, + "grad_norm": 0.9976581335067749, + "learning_rate": 1.6671830245507114e-08, + "loss": 0.2909, + "step": 57319 + }, + { + "epoch": 3.977517174380681, + "grad_norm": 0.8566461801528931, + "learning_rate": 1.6569395035881395e-08, + "loss": 0.2275, + "step": 57320 + }, + { + "epoch": 3.977586565817778, + "grad_norm": 0.7448234558105469, + "learning_rate": 1.6467275462339616e-08, + "loss": 0.2259, + "step": 57321 + }, + { + "epoch": 3.977655957254875, + "grad_norm": 1.0112264156341553, + "learning_rate": 1.6365471525203734e-08, + "loss": 0.3044, + "step": 57322 + }, + { + "epoch": 3.9777253486919713, + "grad_norm": 1.1835538148880005, + "learning_rate": 1.6263983224806823e-08, + "loss": 0.3341, + "step": 57323 + }, + { + "epoch": 3.977794740129068, + "grad_norm": 0.9031230807304382, + "learning_rate": 1.6162810561470843e-08, + "loss": 0.3015, + "step": 57324 + }, + { + "epoch": 3.9778641315661645, + "grad_norm": 0.7984960675239563, + "learning_rate": 1.6061953535495556e-08, + "loss": 0.15, + "step": 57325 + }, + { + "epoch": 3.9779335230032613, + "grad_norm": 0.5796375274658203, + "learning_rate": 1.596141214722513e-08, + "loss": 0.1926, + "step": 57326 + }, + { + "epoch": 3.978002914440358, + "grad_norm": 0.8794702291488647, + "learning_rate": 1.5861186396959326e-08, + "loss": 0.3143, + "step": 57327 + }, + { + "epoch": 3.978072305877455, + "grad_norm": 0.6632204055786133, + "learning_rate": 1.576127628502011e-08, + "loss": 0.1134, + "step": 57328 + }, + { + "epoch": 3.9781416973145514, + "grad_norm": 0.8724068999290466, + "learning_rate": 1.5661681811718343e-08, + "loss": 0.2635, + "step": 57329 + }, + { + "epoch": 3.978211088751648, + "grad_norm": 0.6270472407341003, + "learning_rate": 1.5562402977375988e-08, + "loss": 0.2552, + "step": 57330 + }, + { + "epoch": 3.9782804801887446, + "grad_norm": 0.7530008554458618, + "learning_rate": 1.5463439782303912e-08, + "loss": 0.2542, + "step": 57331 + }, + { + "epoch": 3.9783498716258414, + "grad_norm": 0.7896750569343567, + "learning_rate": 1.536479222680187e-08, + "loss": 0.2006, + "step": 57332 + }, + { + "epoch": 3.9784192630629382, + "grad_norm": 1.1202746629714966, + "learning_rate": 1.5266460311202936e-08, + "loss": 0.2908, + "step": 57333 + }, + { + "epoch": 3.9784886545000346, + "grad_norm": 1.0242294073104858, + "learning_rate": 1.5168444035806862e-08, + "loss": 0.1405, + "step": 57334 + }, + { + "epoch": 3.9785580459371315, + "grad_norm": 0.9067761301994324, + "learning_rate": 1.5070743400924514e-08, + "loss": 0.2219, + "step": 57335 + }, + { + "epoch": 3.978627437374228, + "grad_norm": 0.6807822585105896, + "learning_rate": 1.4973358406866757e-08, + "loss": 0.1907, + "step": 57336 + }, + { + "epoch": 3.9786968288113247, + "grad_norm": 1.2708675861358643, + "learning_rate": 1.4876289053933345e-08, + "loss": 0.2992, + "step": 57337 + }, + { + "epoch": 3.9787662202484215, + "grad_norm": 1.0121031999588013, + "learning_rate": 1.4779535342435146e-08, + "loss": 0.1487, + "step": 57338 + }, + { + "epoch": 3.978835611685518, + "grad_norm": 1.064138412475586, + "learning_rate": 1.4683097272671919e-08, + "loss": 0.2693, + "step": 57339 + }, + { + "epoch": 3.9789050031226147, + "grad_norm": 0.9981639981269836, + "learning_rate": 1.4586974844965628e-08, + "loss": 0.3031, + "step": 57340 + }, + { + "epoch": 3.978974394559711, + "grad_norm": 1.101993203163147, + "learning_rate": 1.449116805959383e-08, + "loss": 0.3012, + "step": 57341 + }, + { + "epoch": 3.979043785996808, + "grad_norm": 1.0070624351501465, + "learning_rate": 1.4395676916878486e-08, + "loss": 0.2089, + "step": 57342 + }, + { + "epoch": 3.979113177433905, + "grad_norm": 0.7343869209289551, + "learning_rate": 1.4300501417119361e-08, + "loss": 0.1563, + "step": 57343 + }, + { + "epoch": 3.979182568871001, + "grad_norm": 0.71697998046875, + "learning_rate": 1.4205641560616212e-08, + "loss": 0.304, + "step": 57344 + }, + { + "epoch": 3.979251960308098, + "grad_norm": 0.6535866856575012, + "learning_rate": 1.4111097347657698e-08, + "loss": 0.2454, + "step": 57345 + }, + { + "epoch": 3.9793213517451944, + "grad_norm": 0.740083634853363, + "learning_rate": 1.4016868778554681e-08, + "loss": 0.1594, + "step": 57346 + }, + { + "epoch": 3.9793907431822912, + "grad_norm": 0.8025801181793213, + "learning_rate": 1.392295585359582e-08, + "loss": 0.2404, + "step": 57347 + }, + { + "epoch": 3.979460134619388, + "grad_norm": 0.8506478071212769, + "learning_rate": 1.3829358573080875e-08, + "loss": 0.207, + "step": 57348 + }, + { + "epoch": 3.9795295260564845, + "grad_norm": 0.8296893835067749, + "learning_rate": 1.3736076937309605e-08, + "loss": 0.272, + "step": 57349 + }, + { + "epoch": 3.9795989174935813, + "grad_norm": 2.880312204360962, + "learning_rate": 1.364311094657067e-08, + "loss": 0.3533, + "step": 57350 + }, + { + "epoch": 3.9796683089306777, + "grad_norm": 0.7950751781463623, + "learning_rate": 1.3550460601163828e-08, + "loss": 0.2619, + "step": 57351 + }, + { + "epoch": 3.9797377003677745, + "grad_norm": 1.0890440940856934, + "learning_rate": 1.3458125901377738e-08, + "loss": 0.2725, + "step": 57352 + }, + { + "epoch": 3.9798070918048714, + "grad_norm": 0.8790885806083679, + "learning_rate": 1.3366106847501058e-08, + "loss": 0.2405, + "step": 57353 + }, + { + "epoch": 3.979876483241968, + "grad_norm": 0.9144247770309448, + "learning_rate": 1.3274403439822447e-08, + "loss": 0.3257, + "step": 57354 + }, + { + "epoch": 3.9799458746790646, + "grad_norm": 1.037001371383667, + "learning_rate": 1.3183015678641664e-08, + "loss": 0.279, + "step": 57355 + }, + { + "epoch": 3.9800152661161614, + "grad_norm": 1.1190552711486816, + "learning_rate": 1.3091943564236265e-08, + "loss": 0.2961, + "step": 57356 + }, + { + "epoch": 3.980084657553258, + "grad_norm": 0.5713832974433899, + "learning_rate": 1.3001187096906009e-08, + "loss": 0.1271, + "step": 57357 + }, + { + "epoch": 3.9801540489903546, + "grad_norm": 2.3055176734924316, + "learning_rate": 1.2910746276928455e-08, + "loss": 0.3182, + "step": 57358 + }, + { + "epoch": 3.9802234404274515, + "grad_norm": 0.739206075668335, + "learning_rate": 1.2820621104592256e-08, + "loss": 0.148, + "step": 57359 + }, + { + "epoch": 3.980292831864548, + "grad_norm": 1.4319795370101929, + "learning_rate": 1.2730811580174973e-08, + "loss": 0.3258, + "step": 57360 + }, + { + "epoch": 3.9803622233016447, + "grad_norm": 0.7989477515220642, + "learning_rate": 1.2641317703965262e-08, + "loss": 0.2079, + "step": 57361 + }, + { + "epoch": 3.980431614738741, + "grad_norm": 0.9474452137947083, + "learning_rate": 1.2552139476251779e-08, + "loss": 0.2118, + "step": 57362 + }, + { + "epoch": 3.980501006175838, + "grad_norm": 1.1612745523452759, + "learning_rate": 1.2463276897312081e-08, + "loss": 0.255, + "step": 57363 + }, + { + "epoch": 3.9805703976129347, + "grad_norm": 0.9387468099594116, + "learning_rate": 1.2374729967423727e-08, + "loss": 0.328, + "step": 57364 + }, + { + "epoch": 3.980639789050031, + "grad_norm": 0.8219150304794312, + "learning_rate": 1.2286498686864268e-08, + "loss": 0.2077, + "step": 57365 + }, + { + "epoch": 3.980709180487128, + "grad_norm": 0.8166916966438293, + "learning_rate": 1.2198583055922364e-08, + "loss": 0.242, + "step": 57366 + }, + { + "epoch": 3.9807785719242244, + "grad_norm": 1.171695351600647, + "learning_rate": 1.2110983074853367e-08, + "loss": 0.2711, + "step": 57367 + }, + { + "epoch": 3.980847963361321, + "grad_norm": 0.9613250494003296, + "learning_rate": 1.2023698743968138e-08, + "loss": 0.3084, + "step": 57368 + }, + { + "epoch": 3.980917354798418, + "grad_norm": 1.0804345607757568, + "learning_rate": 1.1936730063510925e-08, + "loss": 0.2435, + "step": 57369 + }, + { + "epoch": 3.9809867462355144, + "grad_norm": 0.8980635404586792, + "learning_rate": 1.185007703377039e-08, + "loss": 0.3333, + "step": 57370 + }, + { + "epoch": 3.9810561376726112, + "grad_norm": 0.7449690699577332, + "learning_rate": 1.1763739655024086e-08, + "loss": 0.1533, + "step": 57371 + }, + { + "epoch": 3.9811255291097076, + "grad_norm": 1.359796166419983, + "learning_rate": 1.1677717927538467e-08, + "loss": 0.2744, + "step": 57372 + }, + { + "epoch": 3.9811949205468045, + "grad_norm": 0.5213130116462708, + "learning_rate": 1.1592011851579986e-08, + "loss": 0.09, + "step": 57373 + }, + { + "epoch": 3.9812643119839013, + "grad_norm": 0.9835290312767029, + "learning_rate": 1.15066214274262e-08, + "loss": 0.2582, + "step": 57374 + }, + { + "epoch": 3.9813337034209977, + "grad_norm": 0.966793417930603, + "learning_rate": 1.1421546655343562e-08, + "loss": 0.2213, + "step": 57375 + }, + { + "epoch": 3.9814030948580945, + "grad_norm": 1.0516616106033325, + "learning_rate": 1.1336787535609628e-08, + "loss": 0.2373, + "step": 57376 + }, + { + "epoch": 3.981472486295191, + "grad_norm": 0.6579204201698303, + "learning_rate": 1.1252344068479748e-08, + "loss": 0.1671, + "step": 57377 + }, + { + "epoch": 3.9815418777322877, + "grad_norm": 0.8880387544631958, + "learning_rate": 1.1168216254220376e-08, + "loss": 0.2434, + "step": 57378 + }, + { + "epoch": 3.9816112691693846, + "grad_norm": 0.8673749566078186, + "learning_rate": 1.108440409310907e-08, + "loss": 0.2795, + "step": 57379 + }, + { + "epoch": 3.9816806606064814, + "grad_norm": 0.6612104177474976, + "learning_rate": 1.1000907585401176e-08, + "loss": 0.1918, + "step": 57380 + }, + { + "epoch": 3.981750052043578, + "grad_norm": 1.0045381784439087, + "learning_rate": 1.0917726731363153e-08, + "loss": 0.2134, + "step": 57381 + }, + { + "epoch": 3.9818194434806746, + "grad_norm": 0.8490645885467529, + "learning_rate": 1.0834861531261453e-08, + "loss": 0.181, + "step": 57382 + }, + { + "epoch": 3.981888834917771, + "grad_norm": 1.0046169757843018, + "learning_rate": 1.0752311985340324e-08, + "loss": 0.1785, + "step": 57383 + }, + { + "epoch": 3.981958226354868, + "grad_norm": 0.8642206192016602, + "learning_rate": 1.0670078093888425e-08, + "loss": 0.2826, + "step": 57384 + }, + { + "epoch": 3.9820276177919647, + "grad_norm": 0.6266523599624634, + "learning_rate": 1.0588159857138902e-08, + "loss": 0.2153, + "step": 57385 + }, + { + "epoch": 3.982097009229061, + "grad_norm": 0.8091100454330444, + "learning_rate": 1.0506557275369311e-08, + "loss": 0.2708, + "step": 57386 + }, + { + "epoch": 3.982166400666158, + "grad_norm": 0.9138015508651733, + "learning_rate": 1.0425270348823901e-08, + "loss": 0.2126, + "step": 57387 + }, + { + "epoch": 3.9822357921032543, + "grad_norm": 0.9082412123680115, + "learning_rate": 1.0344299077769127e-08, + "loss": 0.3109, + "step": 57388 + }, + { + "epoch": 3.982305183540351, + "grad_norm": 0.726660966873169, + "learning_rate": 1.0263643462449235e-08, + "loss": 0.1877, + "step": 57389 + }, + { + "epoch": 3.982374574977448, + "grad_norm": 0.8411453366279602, + "learning_rate": 1.0183303503130682e-08, + "loss": 0.2058, + "step": 57390 + }, + { + "epoch": 3.9824439664145443, + "grad_norm": 0.6407175660133362, + "learning_rate": 1.0103279200068816e-08, + "loss": 0.1563, + "step": 57391 + }, + { + "epoch": 3.982513357851641, + "grad_norm": 0.7902677655220032, + "learning_rate": 1.0023570553496786e-08, + "loss": 0.1871, + "step": 57392 + }, + { + "epoch": 3.9825827492887376, + "grad_norm": 0.761950671672821, + "learning_rate": 9.944177563692148e-09, + "loss": 0.245, + "step": 57393 + }, + { + "epoch": 3.9826521407258344, + "grad_norm": 1.0763720273971558, + "learning_rate": 9.865100230888047e-09, + "loss": 0.2688, + "step": 57394 + }, + { + "epoch": 3.9827215321629312, + "grad_norm": 1.1421177387237549, + "learning_rate": 9.786338555339835e-09, + "loss": 0.2095, + "step": 57395 + }, + { + "epoch": 3.9827909236000276, + "grad_norm": 1.5137064456939697, + "learning_rate": 9.707892537291762e-09, + "loss": 0.3501, + "step": 57396 + }, + { + "epoch": 3.9828603150371245, + "grad_norm": 1.4026784896850586, + "learning_rate": 9.62976217699918e-09, + "loss": 0.3373, + "step": 57397 + }, + { + "epoch": 3.982929706474221, + "grad_norm": 1.0409905910491943, + "learning_rate": 9.551947474695233e-09, + "loss": 0.2296, + "step": 57398 + }, + { + "epoch": 3.9829990979113177, + "grad_norm": 0.9176102876663208, + "learning_rate": 9.474448430646377e-09, + "loss": 0.2778, + "step": 57399 + }, + { + "epoch": 3.9830684893484145, + "grad_norm": 0.8490959405899048, + "learning_rate": 9.397265045074655e-09, + "loss": 0.2725, + "step": 57400 + }, + { + "epoch": 3.9831378807855113, + "grad_norm": 0.8054481744766235, + "learning_rate": 9.320397318246522e-09, + "loss": 0.3063, + "step": 57401 + }, + { + "epoch": 3.9832072722226077, + "grad_norm": 0.5517603754997253, + "learning_rate": 9.24384525038402e-09, + "loss": 0.1433, + "step": 57402 + }, + { + "epoch": 3.983276663659704, + "grad_norm": 0.7858282923698425, + "learning_rate": 9.167608841742502e-09, + "loss": 0.2509, + "step": 57403 + }, + { + "epoch": 3.983346055096801, + "grad_norm": 1.4637534618377686, + "learning_rate": 9.091688092555118e-09, + "loss": 0.293, + "step": 57404 + }, + { + "epoch": 3.983415446533898, + "grad_norm": 1.0322705507278442, + "learning_rate": 9.016083003066111e-09, + "loss": 0.2107, + "step": 57405 + }, + { + "epoch": 3.9834848379709946, + "grad_norm": 0.8083885312080383, + "learning_rate": 8.940793573508632e-09, + "loss": 0.2169, + "step": 57406 + }, + { + "epoch": 3.983554229408091, + "grad_norm": 1.0587785243988037, + "learning_rate": 8.865819804126928e-09, + "loss": 0.2449, + "step": 57407 + }, + { + "epoch": 3.983623620845188, + "grad_norm": 1.0601937770843506, + "learning_rate": 8.791161695154149e-09, + "loss": 0.25, + "step": 57408 + }, + { + "epoch": 3.9836930122822842, + "grad_norm": 1.5895074605941772, + "learning_rate": 8.716819246823439e-09, + "loss": 0.2286, + "step": 57409 + }, + { + "epoch": 3.983762403719381, + "grad_norm": 0.9278198480606079, + "learning_rate": 8.642792459379046e-09, + "loss": 0.2417, + "step": 57410 + }, + { + "epoch": 3.983831795156478, + "grad_norm": 2.235224962234497, + "learning_rate": 8.569081333043017e-09, + "loss": 0.358, + "step": 57411 + }, + { + "epoch": 3.9839011865935743, + "grad_norm": 1.56239914894104, + "learning_rate": 8.4956858680596e-09, + "loss": 0.3398, + "step": 57412 + }, + { + "epoch": 3.983970578030671, + "grad_norm": 1.282165288925171, + "learning_rate": 8.422606064650839e-09, + "loss": 0.4317, + "step": 57413 + }, + { + "epoch": 3.9840399694677675, + "grad_norm": 0.8234702944755554, + "learning_rate": 8.349841923049884e-09, + "loss": 0.2781, + "step": 57414 + }, + { + "epoch": 3.9841093609048643, + "grad_norm": 1.0313911437988281, + "learning_rate": 8.277393443478775e-09, + "loss": 0.1437, + "step": 57415 + }, + { + "epoch": 3.984178752341961, + "grad_norm": 0.9656465649604797, + "learning_rate": 8.205260626192868e-09, + "loss": 0.2987, + "step": 57416 + }, + { + "epoch": 3.9842481437790576, + "grad_norm": 0.7343744039535522, + "learning_rate": 8.133443471380897e-09, + "loss": 0.191, + "step": 57417 + }, + { + "epoch": 3.9843175352161544, + "grad_norm": 0.8660436868667603, + "learning_rate": 8.061941979309317e-09, + "loss": 0.2998, + "step": 57418 + }, + { + "epoch": 3.984386926653251, + "grad_norm": 1.274627685546875, + "learning_rate": 7.990756150177969e-09, + "loss": 0.4391, + "step": 57419 + }, + { + "epoch": 3.9844563180903476, + "grad_norm": 1.1752562522888184, + "learning_rate": 7.919885984208896e-09, + "loss": 0.4154, + "step": 57420 + }, + { + "epoch": 3.9845257095274444, + "grad_norm": 1.1057348251342773, + "learning_rate": 7.849331481646349e-09, + "loss": 0.2226, + "step": 57421 + }, + { + "epoch": 3.984595100964541, + "grad_norm": 0.7371277213096619, + "learning_rate": 7.77909264270127e-09, + "loss": 0.2189, + "step": 57422 + }, + { + "epoch": 3.9846644924016377, + "grad_norm": 0.7274978160858154, + "learning_rate": 7.7091694675957e-09, + "loss": 0.1654, + "step": 57423 + }, + { + "epoch": 3.984733883838734, + "grad_norm": 0.738031268119812, + "learning_rate": 7.639561956551688e-09, + "loss": 0.2601, + "step": 57424 + }, + { + "epoch": 3.984803275275831, + "grad_norm": 2.067382574081421, + "learning_rate": 7.570270109791278e-09, + "loss": 0.2596, + "step": 57425 + }, + { + "epoch": 3.9848726667129277, + "grad_norm": 1.0702649354934692, + "learning_rate": 7.50129392752541e-09, + "loss": 0.2246, + "step": 57426 + }, + { + "epoch": 3.9849420581500246, + "grad_norm": 0.5403940081596375, + "learning_rate": 7.432633409976131e-09, + "loss": 0.1664, + "step": 57427 + }, + { + "epoch": 3.985011449587121, + "grad_norm": 0.8266937136650085, + "learning_rate": 7.364288557365484e-09, + "loss": 0.34, + "step": 57428 + }, + { + "epoch": 3.9850808410242173, + "grad_norm": 0.8834382891654968, + "learning_rate": 7.29625936989331e-09, + "loss": 0.3037, + "step": 57429 + }, + { + "epoch": 3.985150232461314, + "grad_norm": 1.6184229850769043, + "learning_rate": 7.228545847792756e-09, + "loss": 0.3593, + "step": 57430 + }, + { + "epoch": 3.985219623898411, + "grad_norm": 0.7651874423027039, + "learning_rate": 7.161147991274764e-09, + "loss": 0.2711, + "step": 57431 + }, + { + "epoch": 3.985289015335508, + "grad_norm": 0.5046980381011963, + "learning_rate": 7.094065800539174e-09, + "loss": 0.1194, + "step": 57432 + }, + { + "epoch": 3.9853584067726042, + "grad_norm": 0.6823732256889343, + "learning_rate": 7.027299275819133e-09, + "loss": 0.2175, + "step": 57433 + }, + { + "epoch": 3.985427798209701, + "grad_norm": 0.8116682171821594, + "learning_rate": 6.960848417303378e-09, + "loss": 0.2205, + "step": 57434 + }, + { + "epoch": 3.9854971896467974, + "grad_norm": 1.2290658950805664, + "learning_rate": 6.8947132252028535e-09, + "loss": 0.3326, + "step": 57435 + }, + { + "epoch": 3.9855665810838943, + "grad_norm": 0.807571530342102, + "learning_rate": 6.828893699750705e-09, + "loss": 0.2655, + "step": 57436 + }, + { + "epoch": 3.985635972520991, + "grad_norm": 0.6584019660949707, + "learning_rate": 6.763389841124568e-09, + "loss": 0.1115, + "step": 57437 + }, + { + "epoch": 3.9857053639580875, + "grad_norm": 1.3404436111450195, + "learning_rate": 6.6982016495575895e-09, + "loss": 0.3766, + "step": 57438 + }, + { + "epoch": 3.9857747553951843, + "grad_norm": 0.8292844891548157, + "learning_rate": 6.633329125238508e-09, + "loss": 0.281, + "step": 57439 + }, + { + "epoch": 3.9858441468322807, + "grad_norm": 0.8044368028640747, + "learning_rate": 6.568772268367163e-09, + "loss": 0.2959, + "step": 57440 + }, + { + "epoch": 3.9859135382693776, + "grad_norm": 0.7008633613586426, + "learning_rate": 6.504531079165599e-09, + "loss": 0.1001, + "step": 57441 + }, + { + "epoch": 3.9859829297064744, + "grad_norm": 1.3457492589950562, + "learning_rate": 6.440605557833656e-09, + "loss": 0.3406, + "step": 57442 + }, + { + "epoch": 3.9860523211435708, + "grad_norm": 1.1976571083068848, + "learning_rate": 6.376995704560074e-09, + "loss": 0.1779, + "step": 57443 + }, + { + "epoch": 3.9861217125806676, + "grad_norm": 1.4044007062911987, + "learning_rate": 6.313701519555792e-09, + "loss": 0.3545, + "step": 57444 + }, + { + "epoch": 3.986191104017764, + "grad_norm": 0.7132455110549927, + "learning_rate": 6.250723003009551e-09, + "loss": 0.2284, + "step": 57445 + }, + { + "epoch": 3.986260495454861, + "grad_norm": 0.9270786046981812, + "learning_rate": 6.188060155132292e-09, + "loss": 0.2759, + "step": 57446 + }, + { + "epoch": 3.9863298868919577, + "grad_norm": 0.9889554381370544, + "learning_rate": 6.125712976123854e-09, + "loss": 0.3019, + "step": 57447 + }, + { + "epoch": 3.986399278329054, + "grad_norm": 0.9581155776977539, + "learning_rate": 6.0636814661729765e-09, + "loss": 0.1806, + "step": 57448 + }, + { + "epoch": 3.986468669766151, + "grad_norm": 0.6717770099639893, + "learning_rate": 6.001965625468398e-09, + "loss": 0.1698, + "step": 57449 + }, + { + "epoch": 3.9865380612032473, + "grad_norm": 0.9440622329711914, + "learning_rate": 5.9405654542210585e-09, + "loss": 0.2106, + "step": 57450 + }, + { + "epoch": 3.986607452640344, + "grad_norm": 1.0775337219238281, + "learning_rate": 5.879480952619698e-09, + "loss": 0.2117, + "step": 57451 + }, + { + "epoch": 3.986676844077441, + "grad_norm": 0.7278926372528076, + "learning_rate": 5.8187121208530536e-09, + "loss": 0.1779, + "step": 57452 + }, + { + "epoch": 3.9867462355145378, + "grad_norm": 0.8020764589309692, + "learning_rate": 5.758258959120966e-09, + "loss": 0.2137, + "step": 57453 + }, + { + "epoch": 3.986815626951634, + "grad_norm": 1.089568853378296, + "learning_rate": 5.69812146760107e-09, + "loss": 0.2856, + "step": 57454 + }, + { + "epoch": 3.986885018388731, + "grad_norm": 0.610450267791748, + "learning_rate": 5.638299646493206e-09, + "loss": 0.1562, + "step": 57455 + }, + { + "epoch": 3.9869544098258274, + "grad_norm": 0.6763036847114563, + "learning_rate": 5.578793495986112e-09, + "loss": 0.134, + "step": 57456 + }, + { + "epoch": 3.987023801262924, + "grad_norm": 1.030753254890442, + "learning_rate": 5.519603016268527e-09, + "loss": 0.241, + "step": 57457 + }, + { + "epoch": 3.987093192700021, + "grad_norm": 1.2002394199371338, + "learning_rate": 5.460728207518084e-09, + "loss": 0.2931, + "step": 57458 + }, + { + "epoch": 3.9871625841371174, + "grad_norm": 0.914884626865387, + "learning_rate": 5.402169069923524e-09, + "loss": 0.2414, + "step": 57459 + }, + { + "epoch": 3.9872319755742143, + "grad_norm": 1.3981376886367798, + "learning_rate": 5.343925603673583e-09, + "loss": 0.3113, + "step": 57460 + }, + { + "epoch": 3.9873013670113107, + "grad_norm": 0.7083063125610352, + "learning_rate": 5.285997808957e-09, + "loss": 0.1616, + "step": 57461 + }, + { + "epoch": 3.9873707584484075, + "grad_norm": 0.8324449062347412, + "learning_rate": 5.228385685951409e-09, + "loss": 0.2467, + "step": 57462 + }, + { + "epoch": 3.9874401498855043, + "grad_norm": 0.9249290227890015, + "learning_rate": 5.1710892348344474e-09, + "loss": 0.2215, + "step": 57463 + }, + { + "epoch": 3.9875095413226007, + "grad_norm": 1.715044617652893, + "learning_rate": 5.11410845578375e-09, + "loss": 0.2463, + "step": 57464 + }, + { + "epoch": 3.9875789327596975, + "grad_norm": 0.8976508975028992, + "learning_rate": 5.057443348999158e-09, + "loss": 0.1967, + "step": 57465 + }, + { + "epoch": 3.987648324196794, + "grad_norm": 1.092340111732483, + "learning_rate": 5.001093914636101e-09, + "loss": 0.2252, + "step": 57466 + }, + { + "epoch": 3.9877177156338908, + "grad_norm": 0.8169909119606018, + "learning_rate": 4.945060152894421e-09, + "loss": 0.1798, + "step": 57467 + }, + { + "epoch": 3.9877871070709876, + "grad_norm": 0.7991346716880798, + "learning_rate": 4.889342063929548e-09, + "loss": 0.2064, + "step": 57468 + }, + { + "epoch": 3.987856498508084, + "grad_norm": 0.8661382794380188, + "learning_rate": 4.833939647930219e-09, + "loss": 0.235, + "step": 57469 + }, + { + "epoch": 3.987925889945181, + "grad_norm": 0.7459543943405151, + "learning_rate": 4.778852905074072e-09, + "loss": 0.2365, + "step": 57470 + }, + { + "epoch": 3.987995281382277, + "grad_norm": 0.6792100071907043, + "learning_rate": 4.724081835527638e-09, + "loss": 0.191, + "step": 57471 + }, + { + "epoch": 3.988064672819374, + "grad_norm": 1.1332155466079712, + "learning_rate": 4.669626439457453e-09, + "loss": 0.335, + "step": 57472 + }, + { + "epoch": 3.988134064256471, + "grad_norm": 1.0270192623138428, + "learning_rate": 4.615486717052253e-09, + "loss": 0.2477, + "step": 57473 + }, + { + "epoch": 3.9882034556935673, + "grad_norm": 0.8280321359634399, + "learning_rate": 4.56166266846747e-09, + "loss": 0.2624, + "step": 57474 + }, + { + "epoch": 3.988272847130664, + "grad_norm": 0.9067953824996948, + "learning_rate": 4.508154293891842e-09, + "loss": 0.2979, + "step": 57475 + }, + { + "epoch": 3.9883422385677605, + "grad_norm": 1.8846118450164795, + "learning_rate": 4.454961593469698e-09, + "loss": 0.3141, + "step": 57476 + }, + { + "epoch": 3.9884116300048573, + "grad_norm": 1.0219855308532715, + "learning_rate": 4.402084567389775e-09, + "loss": 0.263, + "step": 57477 + }, + { + "epoch": 3.988481021441954, + "grad_norm": 0.8531995415687561, + "learning_rate": 4.349523215807505e-09, + "loss": 0.2646, + "step": 57478 + }, + { + "epoch": 3.988550412879051, + "grad_norm": 0.7600283026695251, + "learning_rate": 4.297277538889421e-09, + "loss": 0.1764, + "step": 57479 + }, + { + "epoch": 3.9886198043161474, + "grad_norm": 1.158611536026001, + "learning_rate": 4.245347536802058e-09, + "loss": 0.2974, + "step": 57480 + }, + { + "epoch": 3.988689195753244, + "grad_norm": 0.893911600112915, + "learning_rate": 4.193733209723049e-09, + "loss": 0.2595, + "step": 57481 + }, + { + "epoch": 3.9887585871903406, + "grad_norm": 1.5853908061981201, + "learning_rate": 4.142434557785624e-09, + "loss": 0.2523, + "step": 57482 + }, + { + "epoch": 3.9888279786274374, + "grad_norm": 0.9672744870185852, + "learning_rate": 4.0914515811785185e-09, + "loss": 0.2286, + "step": 57483 + }, + { + "epoch": 3.9888973700645343, + "grad_norm": 1.084376573562622, + "learning_rate": 4.040784280057164e-09, + "loss": 0.3138, + "step": 57484 + }, + { + "epoch": 3.9889667615016307, + "grad_norm": 1.350608229637146, + "learning_rate": 3.99043265456589e-09, + "loss": 0.2612, + "step": 57485 + }, + { + "epoch": 3.9890361529387275, + "grad_norm": 0.7187651991844177, + "learning_rate": 3.940396704882332e-09, + "loss": 0.1886, + "step": 57486 + }, + { + "epoch": 3.989105544375824, + "grad_norm": 0.8289912343025208, + "learning_rate": 3.890676431161921e-09, + "loss": 0.2308, + "step": 57487 + }, + { + "epoch": 3.9891749358129207, + "grad_norm": 0.6612673401832581, + "learning_rate": 3.841271833548987e-09, + "loss": 0.1663, + "step": 57488 + }, + { + "epoch": 3.9892443272500175, + "grad_norm": 1.9014811515808105, + "learning_rate": 3.792182912210062e-09, + "loss": 0.2735, + "step": 57489 + }, + { + "epoch": 3.989313718687114, + "grad_norm": 0.9045340418815613, + "learning_rate": 3.743409667300579e-09, + "loss": 0.1797, + "step": 57490 + }, + { + "epoch": 3.9893831101242108, + "grad_norm": 0.8886857628822327, + "learning_rate": 3.6949520989648656e-09, + "loss": 0.3307, + "step": 57491 + }, + { + "epoch": 3.989452501561307, + "grad_norm": 1.4331045150756836, + "learning_rate": 3.6468102073583533e-09, + "loss": 0.5008, + "step": 57492 + }, + { + "epoch": 3.989521892998404, + "grad_norm": 1.1469630002975464, + "learning_rate": 3.5989839926475753e-09, + "loss": 0.2903, + "step": 57493 + }, + { + "epoch": 3.989591284435501, + "grad_norm": 1.0934369564056396, + "learning_rate": 3.5514734549768614e-09, + "loss": 0.1863, + "step": 57494 + }, + { + "epoch": 3.989660675872597, + "grad_norm": 0.7105444669723511, + "learning_rate": 3.5042785944794377e-09, + "loss": 0.1793, + "step": 57495 + }, + { + "epoch": 3.989730067309694, + "grad_norm": 0.9797165393829346, + "learning_rate": 3.4573994113218376e-09, + "loss": 0.2622, + "step": 57496 + }, + { + "epoch": 3.9897994587467904, + "grad_norm": 0.8610832691192627, + "learning_rate": 3.4108359056483905e-09, + "loss": 0.3025, + "step": 57497 + }, + { + "epoch": 3.9898688501838873, + "grad_norm": 1.0825440883636475, + "learning_rate": 3.3645880776034257e-09, + "loss": 0.3157, + "step": 57498 + }, + { + "epoch": 3.989938241620984, + "grad_norm": 0.8375012874603271, + "learning_rate": 3.3186559273423734e-09, + "loss": 0.2772, + "step": 57499 + }, + { + "epoch": 3.9900076330580805, + "grad_norm": 0.6839343905448914, + "learning_rate": 3.2730394549984613e-09, + "loss": 0.2062, + "step": 57500 + }, + { + "epoch": 3.9900770244951773, + "grad_norm": 1.0667167901992798, + "learning_rate": 3.2277386607160176e-09, + "loss": 0.3626, + "step": 57501 + }, + { + "epoch": 3.9901464159322737, + "grad_norm": 0.8982194066047668, + "learning_rate": 3.182753544639372e-09, + "loss": 0.2489, + "step": 57502 + }, + { + "epoch": 3.9902158073693705, + "grad_norm": 0.7471563816070557, + "learning_rate": 3.138084106923955e-09, + "loss": 0.2069, + "step": 57503 + }, + { + "epoch": 3.9902851988064674, + "grad_norm": 0.8293487429618835, + "learning_rate": 3.0937303476918922e-09, + "loss": 0.2123, + "step": 57504 + }, + { + "epoch": 3.990354590243564, + "grad_norm": 1.204964280128479, + "learning_rate": 3.049692267098614e-09, + "loss": 0.2006, + "step": 57505 + }, + { + "epoch": 3.9904239816806606, + "grad_norm": 1.395445704460144, + "learning_rate": 3.005969865266245e-09, + "loss": 0.2514, + "step": 57506 + }, + { + "epoch": 3.9904933731177574, + "grad_norm": 0.8411269187927246, + "learning_rate": 2.9625631423502167e-09, + "loss": 0.251, + "step": 57507 + }, + { + "epoch": 3.990562764554854, + "grad_norm": 0.7388738393783569, + "learning_rate": 2.919472098472653e-09, + "loss": 0.2671, + "step": 57508 + }, + { + "epoch": 3.9906321559919506, + "grad_norm": 1.0807863473892212, + "learning_rate": 2.876696733777884e-09, + "loss": 0.2638, + "step": 57509 + }, + { + "epoch": 3.9907015474290475, + "grad_norm": 1.0259106159210205, + "learning_rate": 2.8342370483991353e-09, + "loss": 0.2344, + "step": 57510 + }, + { + "epoch": 3.990770938866144, + "grad_norm": 1.0469094514846802, + "learning_rate": 2.7920930424696345e-09, + "loss": 0.2663, + "step": 57511 + }, + { + "epoch": 3.9908403303032407, + "grad_norm": 1.0493240356445312, + "learning_rate": 2.75026471613371e-09, + "loss": 0.2018, + "step": 57512 + }, + { + "epoch": 3.990909721740337, + "grad_norm": 0.7044068574905396, + "learning_rate": 2.7087520695023848e-09, + "loss": 0.1877, + "step": 57513 + }, + { + "epoch": 3.990979113177434, + "grad_norm": 0.9794332981109619, + "learning_rate": 2.6675551027199873e-09, + "loss": 0.3212, + "step": 57514 + }, + { + "epoch": 3.9910485046145308, + "grad_norm": 0.7807164192199707, + "learning_rate": 2.6266738159197447e-09, + "loss": 0.2365, + "step": 57515 + }, + { + "epoch": 3.991117896051627, + "grad_norm": 0.7700945734977722, + "learning_rate": 2.586108209223781e-09, + "loss": 0.1823, + "step": 57516 + }, + { + "epoch": 3.991187287488724, + "grad_norm": 0.8688541054725647, + "learning_rate": 2.545858282754221e-09, + "loss": 0.2084, + "step": 57517 + }, + { + "epoch": 3.9912566789258204, + "grad_norm": 1.6427885293960571, + "learning_rate": 2.505924036655394e-09, + "loss": 0.2338, + "step": 57518 + }, + { + "epoch": 3.991326070362917, + "grad_norm": 1.1354219913482666, + "learning_rate": 2.4663054710383216e-09, + "loss": 0.3123, + "step": 57519 + }, + { + "epoch": 3.991395461800014, + "grad_norm": 1.064096450805664, + "learning_rate": 2.427002586025129e-09, + "loss": 0.319, + "step": 57520 + }, + { + "epoch": 3.9914648532371104, + "grad_norm": 0.9227451086044312, + "learning_rate": 2.3880153817601447e-09, + "loss": 0.1928, + "step": 57521 + }, + { + "epoch": 3.9915342446742073, + "grad_norm": 1.1093488931655884, + "learning_rate": 2.3493438583543915e-09, + "loss": 0.2768, + "step": 57522 + }, + { + "epoch": 3.9916036361113036, + "grad_norm": 0.7522839903831482, + "learning_rate": 2.3109880159188914e-09, + "loss": 0.1487, + "step": 57523 + }, + { + "epoch": 3.9916730275484005, + "grad_norm": 1.2039070129394531, + "learning_rate": 2.272947854597973e-09, + "loss": 0.3332, + "step": 57524 + }, + { + "epoch": 3.9917424189854973, + "grad_norm": 1.0296556949615479, + "learning_rate": 2.235223374491557e-09, + "loss": 0.2288, + "step": 57525 + }, + { + "epoch": 3.9918118104225937, + "grad_norm": 0.7693877220153809, + "learning_rate": 2.1978145757328705e-09, + "loss": 0.1385, + "step": 57526 + }, + { + "epoch": 3.9918812018596905, + "grad_norm": 0.9305456280708313, + "learning_rate": 2.160721458432935e-09, + "loss": 0.1964, + "step": 57527 + }, + { + "epoch": 3.991950593296787, + "grad_norm": 1.2390327453613281, + "learning_rate": 2.123944022702773e-09, + "loss": 0.2482, + "step": 57528 + }, + { + "epoch": 3.9920199847338838, + "grad_norm": 1.0606000423431396, + "learning_rate": 2.0874822686756113e-09, + "loss": 0.2835, + "step": 57529 + }, + { + "epoch": 3.9920893761709806, + "grad_norm": 0.6982511281967163, + "learning_rate": 2.05133619645137e-09, + "loss": 0.1806, + "step": 57530 + }, + { + "epoch": 3.9921587676080774, + "grad_norm": 0.9248186349868774, + "learning_rate": 2.0155058061410713e-09, + "loss": 0.2223, + "step": 57531 + }, + { + "epoch": 3.992228159045174, + "grad_norm": 0.9685547351837158, + "learning_rate": 1.9799910978779424e-09, + "loss": 0.3214, + "step": 57532 + }, + { + "epoch": 3.9922975504822706, + "grad_norm": 0.8831853270530701, + "learning_rate": 1.9447920717619027e-09, + "loss": 0.1933, + "step": 57533 + }, + { + "epoch": 3.992366941919367, + "grad_norm": 0.8110149502754211, + "learning_rate": 1.9099087278928728e-09, + "loss": 0.1995, + "step": 57534 + }, + { + "epoch": 3.992436333356464, + "grad_norm": 0.5977107286453247, + "learning_rate": 1.8753410664040795e-09, + "loss": 0.1165, + "step": 57535 + }, + { + "epoch": 3.9925057247935607, + "grad_norm": 1.5297825336456299, + "learning_rate": 1.8410890873843401e-09, + "loss": 0.326, + "step": 57536 + }, + { + "epoch": 3.992575116230657, + "grad_norm": 1.480267882347107, + "learning_rate": 1.8071527909557794e-09, + "loss": 0.2006, + "step": 57537 + }, + { + "epoch": 3.992644507667754, + "grad_norm": 0.4488738477230072, + "learning_rate": 1.7735321772183177e-09, + "loss": 0.1024, + "step": 57538 + }, + { + "epoch": 3.9927138991048503, + "grad_norm": 0.8972042798995972, + "learning_rate": 1.7402272462829771e-09, + "loss": 0.2301, + "step": 57539 + }, + { + "epoch": 3.992783290541947, + "grad_norm": 1.0169850587844849, + "learning_rate": 1.7072379982496777e-09, + "loss": 0.2843, + "step": 57540 + }, + { + "epoch": 3.992852681979044, + "grad_norm": 0.7762569189071655, + "learning_rate": 1.6745644332183397e-09, + "loss": 0.2036, + "step": 57541 + }, + { + "epoch": 3.9929220734161404, + "grad_norm": 1.5094497203826904, + "learning_rate": 1.642206551299985e-09, + "loss": 0.2725, + "step": 57542 + }, + { + "epoch": 3.992991464853237, + "grad_norm": 1.0788551568984985, + "learning_rate": 1.6101643526056364e-09, + "loss": 0.2795, + "step": 57543 + }, + { + "epoch": 3.9930608562903336, + "grad_norm": 1.1001108884811401, + "learning_rate": 1.5784378372130094e-09, + "loss": 0.2129, + "step": 57544 + }, + { + "epoch": 3.9931302477274304, + "grad_norm": 0.7408100962638855, + "learning_rate": 1.5470270052442282e-09, + "loss": 0.1784, + "step": 57545 + }, + { + "epoch": 3.9931996391645272, + "grad_norm": 1.0119576454162598, + "learning_rate": 1.5159318567881108e-09, + "loss": 0.3219, + "step": 57546 + }, + { + "epoch": 3.9932690306016236, + "grad_norm": 0.7648897171020508, + "learning_rate": 1.4851523919445776e-09, + "loss": 0.1881, + "step": 57547 + }, + { + "epoch": 3.9933384220387205, + "grad_norm": 0.6651782989501953, + "learning_rate": 1.4546886108135481e-09, + "loss": 0.1927, + "step": 57548 + }, + { + "epoch": 3.993407813475817, + "grad_norm": 1.1674400568008423, + "learning_rate": 1.4245405134838408e-09, + "loss": 0.2718, + "step": 57549 + }, + { + "epoch": 3.9934772049129137, + "grad_norm": 1.4856292009353638, + "learning_rate": 1.3947081000553752e-09, + "loss": 0.2967, + "step": 57550 + }, + { + "epoch": 3.9935465963500105, + "grad_norm": 1.166468858718872, + "learning_rate": 1.3651913706169694e-09, + "loss": 0.2924, + "step": 57551 + }, + { + "epoch": 3.993615987787107, + "grad_norm": 0.6616620421409607, + "learning_rate": 1.3359903252685435e-09, + "loss": 0.207, + "step": 57552 + }, + { + "epoch": 3.9936853792242037, + "grad_norm": 1.301758885383606, + "learning_rate": 1.3071049640989152e-09, + "loss": 0.3016, + "step": 57553 + }, + { + "epoch": 3.9937547706613, + "grad_norm": 0.8060429692268372, + "learning_rate": 1.2785352872080048e-09, + "loss": 0.1674, + "step": 57554 + }, + { + "epoch": 3.993824162098397, + "grad_norm": 0.828428328037262, + "learning_rate": 1.2502812946735276e-09, + "loss": 0.1808, + "step": 57555 + }, + { + "epoch": 3.993893553535494, + "grad_norm": 0.9730542302131653, + "learning_rate": 1.2223429865954038e-09, + "loss": 0.2528, + "step": 57556 + }, + { + "epoch": 3.9939629449725906, + "grad_norm": 1.2394144535064697, + "learning_rate": 1.1947203630513493e-09, + "loss": 0.3688, + "step": 57557 + }, + { + "epoch": 3.994032336409687, + "grad_norm": 0.7360147833824158, + "learning_rate": 1.1674134241412838e-09, + "loss": 0.2006, + "step": 57558 + }, + { + "epoch": 3.994101727846784, + "grad_norm": 1.008041501045227, + "learning_rate": 1.1404221699318207e-09, + "loss": 0.2206, + "step": 57559 + }, + { + "epoch": 3.9941711192838802, + "grad_norm": 0.9445322751998901, + "learning_rate": 1.1137466005228803e-09, + "loss": 0.2738, + "step": 57560 + }, + { + "epoch": 3.994240510720977, + "grad_norm": 0.9449391961097717, + "learning_rate": 1.0873867160032802e-09, + "loss": 0.2411, + "step": 57561 + }, + { + "epoch": 3.994309902158074, + "grad_norm": 0.6362403035163879, + "learning_rate": 1.061342516439634e-09, + "loss": 0.1976, + "step": 57562 + }, + { + "epoch": 3.9943792935951703, + "grad_norm": 0.9706304669380188, + "learning_rate": 1.0356140019318616e-09, + "loss": 0.2184, + "step": 57563 + }, + { + "epoch": 3.994448685032267, + "grad_norm": 0.8386002779006958, + "learning_rate": 1.0102011725465766e-09, + "loss": 0.1672, + "step": 57564 + }, + { + "epoch": 3.9945180764693635, + "grad_norm": 0.999629020690918, + "learning_rate": 9.851040283725966e-10, + "loss": 0.2298, + "step": 57565 + }, + { + "epoch": 3.9945874679064604, + "grad_norm": 0.6996283531188965, + "learning_rate": 9.603225694876372e-10, + "loss": 0.1973, + "step": 57566 + }, + { + "epoch": 3.994656859343557, + "grad_norm": 1.1619056463241577, + "learning_rate": 9.35856795958312e-10, + "loss": 0.447, + "step": 57567 + }, + { + "epoch": 3.9947262507806536, + "grad_norm": 0.8552354574203491, + "learning_rate": 9.11706707884541e-10, + "loss": 0.2029, + "step": 57568 + }, + { + "epoch": 3.9947956422177504, + "grad_norm": 0.9669915437698364, + "learning_rate": 8.878723053329374e-10, + "loss": 0.2666, + "step": 57569 + }, + { + "epoch": 3.994865033654847, + "grad_norm": 0.9520514607429504, + "learning_rate": 8.643535883701148e-10, + "loss": 0.2383, + "step": 57570 + }, + { + "epoch": 3.9949344250919436, + "grad_norm": 0.8810482025146484, + "learning_rate": 8.411505570737888e-10, + "loss": 0.2398, + "step": 57571 + }, + { + "epoch": 3.9950038165290405, + "grad_norm": 0.8010657429695129, + "learning_rate": 8.182632115216749e-10, + "loss": 0.2279, + "step": 57572 + }, + { + "epoch": 3.995073207966137, + "grad_norm": 1.447357177734375, + "learning_rate": 7.956915517914887e-10, + "loss": 0.2302, + "step": 57573 + }, + { + "epoch": 3.9951425994032337, + "grad_norm": 0.9941118359565735, + "learning_rate": 7.734355779387414e-10, + "loss": 0.3261, + "step": 57574 + }, + { + "epoch": 3.99521199084033, + "grad_norm": 0.7378653287887573, + "learning_rate": 7.514952900411487e-10, + "loss": 0.2278, + "step": 57575 + }, + { + "epoch": 3.995281382277427, + "grad_norm": 0.9032348394393921, + "learning_rate": 7.298706881764261e-10, + "loss": 0.3192, + "step": 57576 + }, + { + "epoch": 3.9953507737145237, + "grad_norm": 0.8589177131652832, + "learning_rate": 7.085617724000848e-10, + "loss": 0.237, + "step": 57577 + }, + { + "epoch": 3.99542016515162, + "grad_norm": 1.0595698356628418, + "learning_rate": 6.875685427898404e-10, + "loss": 0.302, + "step": 57578 + }, + { + "epoch": 3.995489556588717, + "grad_norm": 0.9998068809509277, + "learning_rate": 6.66890999401204e-10, + "loss": 0.2736, + "step": 57579 + }, + { + "epoch": 3.9955589480258134, + "grad_norm": 0.8228049278259277, + "learning_rate": 6.465291423007891e-10, + "loss": 0.2125, + "step": 57580 + }, + { + "epoch": 3.99562833946291, + "grad_norm": 1.1395145654678345, + "learning_rate": 6.264829715663111e-10, + "loss": 0.2557, + "step": 57581 + }, + { + "epoch": 3.995697730900007, + "grad_norm": 1.0216277837753296, + "learning_rate": 6.067524872532815e-10, + "loss": 0.3218, + "step": 57582 + }, + { + "epoch": 3.995767122337104, + "grad_norm": 0.6658577919006348, + "learning_rate": 5.873376894172111e-10, + "loss": 0.2177, + "step": 57583 + }, + { + "epoch": 3.9958365137742002, + "grad_norm": 1.516359806060791, + "learning_rate": 5.682385781247135e-10, + "loss": 0.3616, + "step": 57584 + }, + { + "epoch": 3.995905905211297, + "grad_norm": 0.9351863265037537, + "learning_rate": 5.49455153442402e-10, + "loss": 0.2327, + "step": 57585 + }, + { + "epoch": 3.9959752966483935, + "grad_norm": 0.8148224949836731, + "learning_rate": 5.309874154146854e-10, + "loss": 0.1678, + "step": 57586 + }, + { + "epoch": 3.9960446880854903, + "grad_norm": 0.993964672088623, + "learning_rate": 5.128353641081774e-10, + "loss": 0.2554, + "step": 57587 + }, + { + "epoch": 3.996114079522587, + "grad_norm": 0.9939743876457214, + "learning_rate": 4.94998999589491e-10, + "loss": 0.2319, + "step": 57588 + }, + { + "epoch": 3.9961834709596835, + "grad_norm": 0.6436840891838074, + "learning_rate": 4.774783219030354e-10, + "loss": 0.1711, + "step": 57589 + }, + { + "epoch": 3.9962528623967803, + "grad_norm": 0.7326321601867676, + "learning_rate": 4.602733311043217e-10, + "loss": 0.1876, + "step": 57590 + }, + { + "epoch": 3.9963222538338767, + "grad_norm": 0.9428609013557434, + "learning_rate": 4.4338402724886096e-10, + "loss": 0.3027, + "step": 57591 + }, + { + "epoch": 3.9963916452709736, + "grad_norm": 0.7450880408287048, + "learning_rate": 4.2681041039216443e-10, + "loss": 0.1465, + "step": 57592 + }, + { + "epoch": 3.9964610367080704, + "grad_norm": 0.7523030042648315, + "learning_rate": 4.1055248058974315e-10, + "loss": 0.2453, + "step": 57593 + }, + { + "epoch": 3.996530428145167, + "grad_norm": 0.8382644057273865, + "learning_rate": 3.9461023788600613e-10, + "loss": 0.1495, + "step": 57594 + }, + { + "epoch": 3.9965998195822636, + "grad_norm": 1.1009384393692017, + "learning_rate": 3.789836823364645e-10, + "loss": 0.3531, + "step": 57595 + }, + { + "epoch": 3.99666921101936, + "grad_norm": 1.3378829956054688, + "learning_rate": 3.636728139966294e-10, + "loss": 0.3025, + "step": 57596 + }, + { + "epoch": 3.996738602456457, + "grad_norm": 1.29506516456604, + "learning_rate": 3.486776328887054e-10, + "loss": 0.31, + "step": 57597 + }, + { + "epoch": 3.9968079938935537, + "grad_norm": 0.9868634939193726, + "learning_rate": 3.3399813909040787e-10, + "loss": 0.2926, + "step": 57598 + }, + { + "epoch": 3.99687738533065, + "grad_norm": 0.6946884393692017, + "learning_rate": 3.196343326350437e-10, + "loss": 0.154, + "step": 57599 + }, + { + "epoch": 3.996946776767747, + "grad_norm": 0.815168559551239, + "learning_rate": 3.055862135670218e-10, + "loss": 0.2054, + "step": 57600 + }, + { + "epoch": 3.9970161682048433, + "grad_norm": 0.9359409213066101, + "learning_rate": 2.91853781930751e-10, + "loss": 0.1572, + "step": 57601 + }, + { + "epoch": 3.99708555964194, + "grad_norm": 0.8252009153366089, + "learning_rate": 2.784370377706402e-10, + "loss": 0.2646, + "step": 57602 + }, + { + "epoch": 3.997154951079037, + "grad_norm": 0.7764968276023865, + "learning_rate": 2.6533598113109847e-10, + "loss": 0.2636, + "step": 57603 + }, + { + "epoch": 3.9972243425161333, + "grad_norm": 1.0384795665740967, + "learning_rate": 2.525506120454324e-10, + "loss": 0.1726, + "step": 57604 + }, + { + "epoch": 3.99729373395323, + "grad_norm": 0.9828444123268127, + "learning_rate": 2.4008093056915317e-10, + "loss": 0.2142, + "step": 57605 + }, + { + "epoch": 3.9973631253903266, + "grad_norm": 0.727164089679718, + "learning_rate": 2.279269367244652e-10, + "loss": 0.1882, + "step": 57606 + }, + { + "epoch": 3.9974325168274234, + "grad_norm": 0.815008819103241, + "learning_rate": 2.1608863056687967e-10, + "loss": 0.2366, + "step": 57607 + }, + { + "epoch": 3.9975019082645202, + "grad_norm": 1.1990150213241577, + "learning_rate": 2.0456601211860105e-10, + "loss": 0.2444, + "step": 57608 + }, + { + "epoch": 3.997571299701617, + "grad_norm": 0.9643526673316956, + "learning_rate": 1.9335908142403824e-10, + "loss": 0.1726, + "step": 57609 + }, + { + "epoch": 3.9976406911387135, + "grad_norm": 0.7271166443824768, + "learning_rate": 1.8246783851649796e-10, + "loss": 0.2116, + "step": 57610 + }, + { + "epoch": 3.9977100825758103, + "grad_norm": 0.6954725384712219, + "learning_rate": 1.7189228342928688e-10, + "loss": 0.168, + "step": 57611 + }, + { + "epoch": 3.9977794740129067, + "grad_norm": 0.8391544222831726, + "learning_rate": 1.616324162068139e-10, + "loss": 0.2279, + "step": 57612 + }, + { + "epoch": 3.9978488654500035, + "grad_norm": 1.214706301689148, + "learning_rate": 1.516882368601813e-10, + "loss": 0.3254, + "step": 57613 + }, + { + "epoch": 3.9979182568871003, + "grad_norm": 0.9733776450157166, + "learning_rate": 1.4205974543379797e-10, + "loss": 0.2788, + "step": 57614 + }, + { + "epoch": 3.9979876483241967, + "grad_norm": 1.033565878868103, + "learning_rate": 1.327469419609706e-10, + "loss": 0.2565, + "step": 57615 + }, + { + "epoch": 3.9980570397612936, + "grad_norm": 0.8680600523948669, + "learning_rate": 1.2374982646390365e-10, + "loss": 0.2527, + "step": 57616 + }, + { + "epoch": 3.99812643119839, + "grad_norm": 0.8874149918556213, + "learning_rate": 1.1506839897590382e-10, + "loss": 0.3248, + "step": 57617 + }, + { + "epoch": 3.998195822635487, + "grad_norm": 0.9586488008499146, + "learning_rate": 1.067026595302778e-10, + "loss": 0.1991, + "step": 57618 + }, + { + "epoch": 3.9982652140725836, + "grad_norm": 0.7002377510070801, + "learning_rate": 9.865260813812782e-11, + "loss": 0.1531, + "step": 57619 + }, + { + "epoch": 3.99833460550968, + "grad_norm": 0.9489929676055908, + "learning_rate": 9.091824483276057e-11, + "loss": 0.2284, + "step": 57620 + }, + { + "epoch": 3.998403996946777, + "grad_norm": 0.817343533039093, + "learning_rate": 8.349956963638051e-11, + "loss": 0.2536, + "step": 57621 + }, + { + "epoch": 3.9984733883838732, + "grad_norm": 2.083327293395996, + "learning_rate": 7.639658257119209e-11, + "loss": 0.2551, + "step": 57622 + }, + { + "epoch": 3.99854277982097, + "grad_norm": 1.0419162511825562, + "learning_rate": 6.960928367050202e-11, + "loss": 0.2524, + "step": 57623 + }, + { + "epoch": 3.998612171258067, + "grad_norm": 1.0582138299942017, + "learning_rate": 6.313767294541251e-11, + "loss": 0.3308, + "step": 57624 + }, + { + "epoch": 3.9986815626951633, + "grad_norm": 0.589280903339386, + "learning_rate": 5.6981750418128034e-11, + "loss": 0.1577, + "step": 57625 + }, + { + "epoch": 3.99875095413226, + "grad_norm": 0.7170749306678772, + "learning_rate": 5.114151611085305e-11, + "loss": 0.222, + "step": 57626 + }, + { + "epoch": 3.9988203455693565, + "grad_norm": 1.046629786491394, + "learning_rate": 4.5616970034689786e-11, + "loss": 0.2956, + "step": 57627 + }, + { + "epoch": 3.9988897370064533, + "grad_norm": 2.3132529258728027, + "learning_rate": 4.040811221184271e-11, + "loss": 0.2904, + "step": 57628 + }, + { + "epoch": 3.99895912844355, + "grad_norm": 1.143376111984253, + "learning_rate": 3.5514942664516274e-11, + "loss": 0.3016, + "step": 57629 + }, + { + "epoch": 3.999028519880647, + "grad_norm": 1.0581592321395874, + "learning_rate": 3.093746140381271e-11, + "loss": 0.285, + "step": 57630 + }, + { + "epoch": 3.9990979113177434, + "grad_norm": 0.9380515217781067, + "learning_rate": 2.6675668440834245e-11, + "loss": 0.1868, + "step": 57631 + }, + { + "epoch": 3.99916730275484, + "grad_norm": 0.9669721722602844, + "learning_rate": 2.2729563786683118e-11, + "loss": 0.2651, + "step": 57632 + }, + { + "epoch": 3.9992366941919366, + "grad_norm": 1.7612582445144653, + "learning_rate": 1.9099147463563783e-11, + "loss": 0.3919, + "step": 57633 + }, + { + "epoch": 3.9993060856290334, + "grad_norm": 0.8698136210441589, + "learning_rate": 1.5784419482578472e-11, + "loss": 0.2335, + "step": 57634 + }, + { + "epoch": 3.9993754770661303, + "grad_norm": 1.1492148637771606, + "learning_rate": 1.2785379843727186e-11, + "loss": 0.4027, + "step": 57635 + }, + { + "epoch": 3.9994448685032267, + "grad_norm": 1.024625301361084, + "learning_rate": 1.0102028569214384e-11, + "loss": 0.2667, + "step": 57636 + }, + { + "epoch": 3.9995142599403235, + "grad_norm": 1.1529743671417236, + "learning_rate": 7.734365647937836e-12, + "loss": 0.2153, + "step": 57637 + }, + { + "epoch": 3.99958365137742, + "grad_norm": 0.865763247013092, + "learning_rate": 5.682391113204233e-12, + "loss": 0.2596, + "step": 57638 + }, + { + "epoch": 3.9996530428145167, + "grad_norm": 0.7477474808692932, + "learning_rate": 3.946104942809115e-12, + "loss": 0.176, + "step": 57639 + }, + { + "epoch": 3.9997224342516136, + "grad_norm": 1.3747143745422363, + "learning_rate": 2.5255071700591716e-12, + "loss": 0.2084, + "step": 57640 + }, + { + "epoch": 3.99979182568871, + "grad_norm": 1.0712988376617432, + "learning_rate": 1.4205977838521734e-12, + "loss": 0.2504, + "step": 57641 + }, + { + "epoch": 3.999861217125807, + "grad_norm": 0.5856016874313354, + "learning_rate": 6.313767952903504e-13, + "loss": 0.1616, + "step": 57642 + }, + { + "epoch": 3.999930608562903, + "grad_norm": 0.6224274635314941, + "learning_rate": 1.5784420437370273e-13, + "loss": 0.1642, + "step": 57643 + }, + { + "epoch": 4.0, + "grad_norm": 0.7567241191864014, + "learning_rate": 0.0, + "loss": 0.1734, + "step": 57644 + }, + { + "epoch": 4.0, + "step": 57644, + "total_flos": 1.963722719928464e+19, + "train_loss": 0.5722225170174615, + "train_runtime": 438130.943, + "train_samples_per_second": 4.21, + "train_steps_per_second": 0.132 + } + ], + "logging_steps": 1.0, + "max_steps": 57644, + "num_input_tokens_seen": 0, + "num_train_epochs": 4, + "save_steps": 10000, + "stateful_callbacks": { + "TrainerControl": { + "args": { + "should_epoch_stop": false, + "should_evaluate": false, + "should_log": false, + "should_save": true, + "should_training_stop": true + }, + "attributes": {} + } + }, + "total_flos": 1.963722719928464e+19, + "train_batch_size": 1, + "trial_name": null, + "trial_params": null +}